{ "best_metric": null, "best_model_checkpoint": null, "epoch": 3.0, "eval_steps": 500, "global_step": 27672, "is_hyper_param_search": false, "is_local_process_zero": true, "is_world_process_zero": true, "log_history": [ { "epoch": 0.00010841283607979184, "grad_norm": 5.2149457931518555, "learning_rate": 1.0000000000000002e-06, "loss": 3.1196, "step": 1 }, { "epoch": 0.0002168256721595837, "grad_norm": 5.145304203033447, "learning_rate": 2.0000000000000003e-06, "loss": 3.0598, "step": 2 }, { "epoch": 0.00032523850823937554, "grad_norm": 4.993410110473633, "learning_rate": 3e-06, "loss": 3.1034, "step": 3 }, { "epoch": 0.0004336513443191674, "grad_norm": 5.154311656951904, "learning_rate": 4.000000000000001e-06, "loss": 3.0236, "step": 4 }, { "epoch": 0.0005420641803989592, "grad_norm": 4.919717311859131, "learning_rate": 5e-06, "loss": 2.8792, "step": 5 }, { "epoch": 0.0006504770164787511, "grad_norm": 4.886139392852783, "learning_rate": 6e-06, "loss": 2.9656, "step": 6 }, { "epoch": 0.000758889852558543, "grad_norm": 4.83964204788208, "learning_rate": 7.000000000000001e-06, "loss": 2.9581, "step": 7 }, { "epoch": 0.0008673026886383347, "grad_norm": 5.021399021148682, "learning_rate": 8.000000000000001e-06, "loss": 2.7555, "step": 8 }, { "epoch": 0.0009757155247181266, "grad_norm": 5.11181116104126, "learning_rate": 9e-06, "loss": 2.8945, "step": 9 }, { "epoch": 0.0010841283607979184, "grad_norm": 4.489777565002441, "learning_rate": 1e-05, "loss": 2.6895, "step": 10 }, { "epoch": 0.0011925411968777104, "grad_norm": 4.633260726928711, "learning_rate": 1.1000000000000001e-05, "loss": 2.5648, "step": 11 }, { "epoch": 0.0013009540329575022, "grad_norm": 5.830043315887451, "learning_rate": 1.2e-05, "loss": 2.538, "step": 12 }, { "epoch": 0.001409366869037294, "grad_norm": 5.258893013000488, "learning_rate": 1.3000000000000001e-05, "loss": 2.4865, "step": 13 }, { "epoch": 0.001517779705117086, "grad_norm": 4.351803779602051, "learning_rate": 1.4000000000000001e-05, "loss": 2.4987, "step": 14 }, { "epoch": 0.0016261925411968777, "grad_norm": 4.373904228210449, "learning_rate": 1.5e-05, "loss": 2.2326, "step": 15 }, { "epoch": 0.0017346053772766695, "grad_norm": 3.848686933517456, "learning_rate": 1.6000000000000003e-05, "loss": 2.1614, "step": 16 }, { "epoch": 0.0018430182133564615, "grad_norm": 3.730391263961792, "learning_rate": 1.7000000000000003e-05, "loss": 2.0308, "step": 17 }, { "epoch": 0.0019514310494362533, "grad_norm": 4.63187313079834, "learning_rate": 1.8e-05, "loss": 1.9656, "step": 18 }, { "epoch": 0.002059843885516045, "grad_norm": 3.5389740467071533, "learning_rate": 1.9e-05, "loss": 1.7762, "step": 19 }, { "epoch": 0.002168256721595837, "grad_norm": 3.3306868076324463, "learning_rate": 2e-05, "loss": 1.7181, "step": 20 }, { "epoch": 0.0022766695576756286, "grad_norm": 3.258305788040161, "learning_rate": 2.1e-05, "loss": 1.6326, "step": 21 }, { "epoch": 0.002385082393755421, "grad_norm": 3.082594633102417, "learning_rate": 2.2000000000000003e-05, "loss": 1.4544, "step": 22 }, { "epoch": 0.0024934952298352126, "grad_norm": 3.4173619747161865, "learning_rate": 2.3000000000000003e-05, "loss": 1.4869, "step": 23 }, { "epoch": 0.0026019080659150044, "grad_norm": 2.7845571041107178, "learning_rate": 2.4e-05, "loss": 1.2558, "step": 24 }, { "epoch": 0.002710320901994796, "grad_norm": 3.3685572147369385, "learning_rate": 2.5e-05, "loss": 1.5883, "step": 25 }, { "epoch": 0.002818733738074588, "grad_norm": 6.6605987548828125, "learning_rate": 2.6000000000000002e-05, "loss": 1.0951, "step": 26 }, { "epoch": 0.0029271465741543797, "grad_norm": 2.962818145751953, "learning_rate": 2.7000000000000002e-05, "loss": 1.0049, "step": 27 }, { "epoch": 0.003035559410234172, "grad_norm": 2.4423131942749023, "learning_rate": 2.8000000000000003e-05, "loss": 0.8532, "step": 28 }, { "epoch": 0.0031439722463139637, "grad_norm": 2.50779128074646, "learning_rate": 2.9e-05, "loss": 0.8296, "step": 29 }, { "epoch": 0.0032523850823937554, "grad_norm": 3.367466449737549, "learning_rate": 3e-05, "loss": 0.7453, "step": 30 }, { "epoch": 0.0033607979184735472, "grad_norm": 2.1686596870422363, "learning_rate": 3.1e-05, "loss": 0.6795, "step": 31 }, { "epoch": 0.003469210754553339, "grad_norm": 1.8389191627502441, "learning_rate": 3.2000000000000005e-05, "loss": 0.5234, "step": 32 }, { "epoch": 0.0035776235906331308, "grad_norm": 1.7388216257095337, "learning_rate": 3.3e-05, "loss": 0.4599, "step": 33 }, { "epoch": 0.003686036426712923, "grad_norm": 2.851001262664795, "learning_rate": 3.4000000000000007e-05, "loss": 0.5464, "step": 34 }, { "epoch": 0.0037944492627927148, "grad_norm": 1.556646466255188, "learning_rate": 3.5e-05, "loss": 0.3974, "step": 35 }, { "epoch": 0.0039028620988725065, "grad_norm": 1.4102658033370972, "learning_rate": 3.6e-05, "loss": 0.3245, "step": 36 }, { "epoch": 0.004011274934952298, "grad_norm": 1.2429163455963135, "learning_rate": 3.7e-05, "loss": 0.2853, "step": 37 }, { "epoch": 0.00411968777103209, "grad_norm": 1.0415282249450684, "learning_rate": 3.8e-05, "loss": 0.2274, "step": 38 }, { "epoch": 0.004228100607111882, "grad_norm": 1.3674875497817993, "learning_rate": 3.9000000000000006e-05, "loss": 0.248, "step": 39 }, { "epoch": 0.004336513443191674, "grad_norm": 1.2795597314834595, "learning_rate": 4e-05, "loss": 0.1984, "step": 40 }, { "epoch": 0.004444926279271465, "grad_norm": 1.1633061170578003, "learning_rate": 4.1e-05, "loss": 0.1958, "step": 41 }, { "epoch": 0.004553339115351257, "grad_norm": 0.6090769171714783, "learning_rate": 4.2e-05, "loss": 0.155, "step": 42 }, { "epoch": 0.00466175195143105, "grad_norm": 0.8438916206359863, "learning_rate": 4.3e-05, "loss": 0.1872, "step": 43 }, { "epoch": 0.004770164787510842, "grad_norm": 1.139715313911438, "learning_rate": 4.4000000000000006e-05, "loss": 0.1684, "step": 44 }, { "epoch": 0.004878577623590633, "grad_norm": 0.8619150519371033, "learning_rate": 4.5e-05, "loss": 0.165, "step": 45 }, { "epoch": 0.004986990459670425, "grad_norm": 0.8248367309570312, "learning_rate": 4.600000000000001e-05, "loss": 0.1924, "step": 46 }, { "epoch": 0.005095403295750217, "grad_norm": 0.5267935991287231, "learning_rate": 4.7e-05, "loss": 0.1393, "step": 47 }, { "epoch": 0.005203816131830009, "grad_norm": 0.3814792335033417, "learning_rate": 4.8e-05, "loss": 0.1195, "step": 48 }, { "epoch": 0.0053122289679098005, "grad_norm": 0.5827663540840149, "learning_rate": 4.9e-05, "loss": 0.1406, "step": 49 }, { "epoch": 0.005420641803989592, "grad_norm": 0.539864182472229, "learning_rate": 5e-05, "loss": 0.1461, "step": 50 }, { "epoch": 0.005529054640069384, "grad_norm": 0.3016558587551117, "learning_rate": 5.1000000000000006e-05, "loss": 0.1306, "step": 51 }, { "epoch": 0.005637467476149176, "grad_norm": 0.4037388265132904, "learning_rate": 5.2000000000000004e-05, "loss": 0.1222, "step": 52 }, { "epoch": 0.005745880312228968, "grad_norm": 0.3812527656555176, "learning_rate": 5.300000000000001e-05, "loss": 0.156, "step": 53 }, { "epoch": 0.005854293148308759, "grad_norm": 0.297986775636673, "learning_rate": 5.4000000000000005e-05, "loss": 0.1472, "step": 54 }, { "epoch": 0.005962705984388552, "grad_norm": 2.845261573791504, "learning_rate": 5.500000000000001e-05, "loss": 0.5234, "step": 55 }, { "epoch": 0.006071118820468344, "grad_norm": 1.1167420148849487, "learning_rate": 5.6000000000000006e-05, "loss": 0.138, "step": 56 }, { "epoch": 0.0061795316565481356, "grad_norm": 0.31902506947517395, "learning_rate": 5.6999999999999996e-05, "loss": 0.1428, "step": 57 }, { "epoch": 0.006287944492627927, "grad_norm": 0.22591254115104675, "learning_rate": 5.8e-05, "loss": 0.1234, "step": 58 }, { "epoch": 0.006396357328707719, "grad_norm": 0.15402370691299438, "learning_rate": 5.9e-05, "loss": 0.116, "step": 59 }, { "epoch": 0.006504770164787511, "grad_norm": 2.796264886856079, "learning_rate": 6e-05, "loss": 0.3957, "step": 60 }, { "epoch": 0.006613183000867303, "grad_norm": 0.3829329311847687, "learning_rate": 6.1e-05, "loss": 0.1321, "step": 61 }, { "epoch": 0.0067215958369470944, "grad_norm": 0.16280348598957062, "learning_rate": 6.2e-05, "loss": 0.1219, "step": 62 }, { "epoch": 0.006830008673026886, "grad_norm": 0.3204944133758545, "learning_rate": 6.3e-05, "loss": 0.1402, "step": 63 }, { "epoch": 0.006938421509106678, "grad_norm": 0.24803131818771362, "learning_rate": 6.400000000000001e-05, "loss": 0.1319, "step": 64 }, { "epoch": 0.00704683434518647, "grad_norm": 0.16616782546043396, "learning_rate": 6.500000000000001e-05, "loss": 0.1399, "step": 65 }, { "epoch": 0.0071552471812662615, "grad_norm": 0.16086168587207794, "learning_rate": 6.6e-05, "loss": 0.1292, "step": 66 }, { "epoch": 0.007263660017346054, "grad_norm": 0.24584974348545074, "learning_rate": 6.7e-05, "loss": 0.1378, "step": 67 }, { "epoch": 0.007372072853425846, "grad_norm": 0.1571667492389679, "learning_rate": 6.800000000000001e-05, "loss": 0.1208, "step": 68 }, { "epoch": 0.007480485689505638, "grad_norm": 0.15028952062129974, "learning_rate": 6.9e-05, "loss": 0.1102, "step": 69 }, { "epoch": 0.0075888985255854295, "grad_norm": 0.40434566140174866, "learning_rate": 7e-05, "loss": 0.1356, "step": 70 }, { "epoch": 0.007697311361665221, "grad_norm": 0.13718268275260925, "learning_rate": 7.1e-05, "loss": 0.1067, "step": 71 }, { "epoch": 0.007805724197745013, "grad_norm": 0.15419070422649384, "learning_rate": 7.2e-05, "loss": 0.1496, "step": 72 }, { "epoch": 0.007914137033824806, "grad_norm": 0.21107347309589386, "learning_rate": 7.3e-05, "loss": 0.1296, "step": 73 }, { "epoch": 0.008022549869904597, "grad_norm": 0.1417638510465622, "learning_rate": 7.4e-05, "loss": 0.1277, "step": 74 }, { "epoch": 0.00813096270598439, "grad_norm": 0.2572252154350281, "learning_rate": 7.500000000000001e-05, "loss": 0.131, "step": 75 }, { "epoch": 0.00823937554206418, "grad_norm": 0.25398099422454834, "learning_rate": 7.6e-05, "loss": 0.1481, "step": 76 }, { "epoch": 0.008347788378143973, "grad_norm": 0.4150233864784241, "learning_rate": 7.7e-05, "loss": 0.1221, "step": 77 }, { "epoch": 0.008456201214223764, "grad_norm": 0.135957732796669, "learning_rate": 7.800000000000001e-05, "loss": 0.1215, "step": 78 }, { "epoch": 0.008564614050303556, "grad_norm": 0.17039060592651367, "learning_rate": 7.900000000000001e-05, "loss": 0.0949, "step": 79 }, { "epoch": 0.008673026886383347, "grad_norm": 0.11901222169399261, "learning_rate": 8e-05, "loss": 0.1236, "step": 80 }, { "epoch": 0.00878143972246314, "grad_norm": 0.4118669629096985, "learning_rate": 8.1e-05, "loss": 0.1278, "step": 81 }, { "epoch": 0.00888985255854293, "grad_norm": 0.6734925508499146, "learning_rate": 8.2e-05, "loss": 0.1469, "step": 82 }, { "epoch": 0.008998265394622723, "grad_norm": 2.939640522003174, "learning_rate": 8.3e-05, "loss": 0.1998, "step": 83 }, { "epoch": 0.009106678230702514, "grad_norm": 4.037971496582031, "learning_rate": 8.4e-05, "loss": 0.2291, "step": 84 }, { "epoch": 0.009215091066782307, "grad_norm": 0.2463056892156601, "learning_rate": 8.5e-05, "loss": 0.1188, "step": 85 }, { "epoch": 0.0093235039028621, "grad_norm": 0.139509379863739, "learning_rate": 8.6e-05, "loss": 0.1181, "step": 86 }, { "epoch": 0.00943191673894189, "grad_norm": 0.41287559270858765, "learning_rate": 8.7e-05, "loss": 0.1365, "step": 87 }, { "epoch": 0.009540329575021683, "grad_norm": 0.20919878780841827, "learning_rate": 8.800000000000001e-05, "loss": 0.1294, "step": 88 }, { "epoch": 0.009648742411101474, "grad_norm": 0.16807663440704346, "learning_rate": 8.900000000000001e-05, "loss": 0.1007, "step": 89 }, { "epoch": 0.009757155247181267, "grad_norm": 0.14114335179328918, "learning_rate": 9e-05, "loss": 0.1225, "step": 90 }, { "epoch": 0.009865568083261058, "grad_norm": 0.2461656928062439, "learning_rate": 9.1e-05, "loss": 0.146, "step": 91 }, { "epoch": 0.00997398091934085, "grad_norm": 0.13559341430664062, "learning_rate": 9.200000000000001e-05, "loss": 0.1219, "step": 92 }, { "epoch": 0.010082393755420641, "grad_norm": 0.12956395745277405, "learning_rate": 9.300000000000001e-05, "loss": 0.1221, "step": 93 }, { "epoch": 0.010190806591500434, "grad_norm": 0.14083503186702728, "learning_rate": 9.4e-05, "loss": 0.1394, "step": 94 }, { "epoch": 0.010299219427580225, "grad_norm": 0.1253279596567154, "learning_rate": 9.5e-05, "loss": 0.1228, "step": 95 }, { "epoch": 0.010407632263660017, "grad_norm": 0.17557112872600555, "learning_rate": 9.6e-05, "loss": 0.1346, "step": 96 }, { "epoch": 0.010516045099739808, "grad_norm": 0.1422986537218094, "learning_rate": 9.7e-05, "loss": 0.1365, "step": 97 }, { "epoch": 0.010624457935819601, "grad_norm": 0.18533742427825928, "learning_rate": 9.8e-05, "loss": 0.1508, "step": 98 }, { "epoch": 0.010732870771899394, "grad_norm": 0.22691737115383148, "learning_rate": 9.900000000000001e-05, "loss": 0.129, "step": 99 }, { "epoch": 0.010841283607979185, "grad_norm": 0.17348608374595642, "learning_rate": 0.0001, "loss": 0.1347, "step": 100 }, { "epoch": 0.010949696444058977, "grad_norm": 0.12170619517564774, "learning_rate": 9.999637313216306e-05, "loss": 0.1325, "step": 101 }, { "epoch": 0.011058109280138768, "grad_norm": 0.14500796794891357, "learning_rate": 9.999274626432613e-05, "loss": 0.1323, "step": 102 }, { "epoch": 0.01116652211621856, "grad_norm": 0.14567314088344574, "learning_rate": 9.998911939648919e-05, "loss": 0.1459, "step": 103 }, { "epoch": 0.011274934952298352, "grad_norm": 0.2874372601509094, "learning_rate": 9.998549252865226e-05, "loss": 0.1298, "step": 104 }, { "epoch": 0.011383347788378144, "grad_norm": 0.2724505662918091, "learning_rate": 9.998186566081532e-05, "loss": 0.1238, "step": 105 }, { "epoch": 0.011491760624457935, "grad_norm": 0.17930521070957184, "learning_rate": 9.997823879297839e-05, "loss": 0.1393, "step": 106 }, { "epoch": 0.011600173460537728, "grad_norm": 0.1797245293855667, "learning_rate": 9.997461192514145e-05, "loss": 0.1415, "step": 107 }, { "epoch": 0.011708586296617519, "grad_norm": 0.1458282619714737, "learning_rate": 9.997098505730452e-05, "loss": 0.1238, "step": 108 }, { "epoch": 0.011816999132697311, "grad_norm": 0.1360117793083191, "learning_rate": 9.996735818946759e-05, "loss": 0.1415, "step": 109 }, { "epoch": 0.011925411968777104, "grad_norm": 0.15984034538269043, "learning_rate": 9.996373132163065e-05, "loss": 0.1352, "step": 110 }, { "epoch": 0.012033824804856895, "grad_norm": 0.16881564259529114, "learning_rate": 9.996010445379372e-05, "loss": 0.1362, "step": 111 }, { "epoch": 0.012142237640936688, "grad_norm": 0.11185746639966965, "learning_rate": 9.995647758595677e-05, "loss": 0.1248, "step": 112 }, { "epoch": 0.012250650477016478, "grad_norm": 0.12183712422847748, "learning_rate": 9.995285071811984e-05, "loss": 0.137, "step": 113 }, { "epoch": 0.012359063313096271, "grad_norm": 0.13424690067768097, "learning_rate": 9.99492238502829e-05, "loss": 0.1361, "step": 114 }, { "epoch": 0.012467476149176062, "grad_norm": 0.3294023275375366, "learning_rate": 9.994559698244596e-05, "loss": 0.1365, "step": 115 }, { "epoch": 0.012575888985255855, "grad_norm": 2.6743650436401367, "learning_rate": 9.994197011460903e-05, "loss": 0.299, "step": 116 }, { "epoch": 0.012684301821335646, "grad_norm": 0.16263447701931, "learning_rate": 9.993834324677209e-05, "loss": 0.1297, "step": 117 }, { "epoch": 0.012792714657415438, "grad_norm": 0.18966206908226013, "learning_rate": 9.993471637893516e-05, "loss": 0.1536, "step": 118 }, { "epoch": 0.01290112749349523, "grad_norm": 0.3197293281555176, "learning_rate": 9.993108951109822e-05, "loss": 0.1541, "step": 119 }, { "epoch": 0.013009540329575022, "grad_norm": 0.16033799946308136, "learning_rate": 9.992746264326129e-05, "loss": 0.1196, "step": 120 }, { "epoch": 0.013117953165654813, "grad_norm": 0.1324940174818039, "learning_rate": 9.992383577542434e-05, "loss": 0.1415, "step": 121 }, { "epoch": 0.013226366001734605, "grad_norm": 0.26303631067276, "learning_rate": 9.992020890758742e-05, "loss": 0.1313, "step": 122 }, { "epoch": 0.013334778837814398, "grad_norm": 0.1303134560585022, "learning_rate": 9.991658203975047e-05, "loss": 0.1159, "step": 123 }, { "epoch": 0.013443191673894189, "grad_norm": 0.13192713260650635, "learning_rate": 9.991295517191353e-05, "loss": 0.1214, "step": 124 }, { "epoch": 0.013551604509973982, "grad_norm": 0.10305740684270859, "learning_rate": 9.99093283040766e-05, "loss": 0.1343, "step": 125 }, { "epoch": 0.013660017346053772, "grad_norm": 0.16451941430568695, "learning_rate": 9.990570143623967e-05, "loss": 0.1384, "step": 126 }, { "epoch": 0.013768430182133565, "grad_norm": 0.21125568449497223, "learning_rate": 9.990207456840274e-05, "loss": 0.127, "step": 127 }, { "epoch": 0.013876843018213356, "grad_norm": 0.17478352785110474, "learning_rate": 9.98984477005658e-05, "loss": 0.1131, "step": 128 }, { "epoch": 0.013985255854293149, "grad_norm": 0.15261723101139069, "learning_rate": 9.989482083272886e-05, "loss": 0.1276, "step": 129 }, { "epoch": 0.01409366869037294, "grad_norm": 0.18892721831798553, "learning_rate": 9.989119396489193e-05, "loss": 0.135, "step": 130 }, { "epoch": 0.014202081526452732, "grad_norm": 0.14765827357769012, "learning_rate": 9.988756709705499e-05, "loss": 0.1181, "step": 131 }, { "epoch": 0.014310494362532523, "grad_norm": 0.14829866588115692, "learning_rate": 9.988394022921806e-05, "loss": 0.1278, "step": 132 }, { "epoch": 0.014418907198612316, "grad_norm": 0.14578723907470703, "learning_rate": 9.988031336138111e-05, "loss": 0.1282, "step": 133 }, { "epoch": 0.014527320034692108, "grad_norm": 0.2550632357597351, "learning_rate": 9.987668649354418e-05, "loss": 0.1653, "step": 134 }, { "epoch": 0.0146357328707719, "grad_norm": 0.15899217128753662, "learning_rate": 9.987305962570724e-05, "loss": 0.1392, "step": 135 }, { "epoch": 0.014744145706851692, "grad_norm": 0.1267155110836029, "learning_rate": 9.986943275787031e-05, "loss": 0.1366, "step": 136 }, { "epoch": 0.014852558542931483, "grad_norm": 2.0227127075195312, "learning_rate": 9.986580589003337e-05, "loss": 0.1538, "step": 137 }, { "epoch": 0.014960971379011275, "grad_norm": 2.0106492042541504, "learning_rate": 9.986217902219643e-05, "loss": 0.2843, "step": 138 }, { "epoch": 0.015069384215091066, "grad_norm": 0.12389489263296127, "learning_rate": 9.98585521543595e-05, "loss": 0.1325, "step": 139 }, { "epoch": 0.015177797051170859, "grad_norm": 0.11544619500637054, "learning_rate": 9.985492528652256e-05, "loss": 0.1196, "step": 140 }, { "epoch": 0.01528620988725065, "grad_norm": 0.13035991787910461, "learning_rate": 9.985129841868563e-05, "loss": 0.1286, "step": 141 }, { "epoch": 0.015394622723330443, "grad_norm": 0.09631640464067459, "learning_rate": 9.984767155084868e-05, "loss": 0.1321, "step": 142 }, { "epoch": 0.015503035559410233, "grad_norm": 0.09737260639667511, "learning_rate": 9.984404468301175e-05, "loss": 0.126, "step": 143 }, { "epoch": 0.015611448395490026, "grad_norm": 0.11527895927429199, "learning_rate": 9.984041781517483e-05, "loss": 0.1233, "step": 144 }, { "epoch": 0.015719861231569817, "grad_norm": 0.08487590402364731, "learning_rate": 9.983679094733788e-05, "loss": 0.1306, "step": 145 }, { "epoch": 0.01582827406764961, "grad_norm": 0.10155870020389557, "learning_rate": 9.983316407950095e-05, "loss": 0.1358, "step": 146 }, { "epoch": 0.015936686903729402, "grad_norm": 0.15308935940265656, "learning_rate": 9.982953721166401e-05, "loss": 0.1347, "step": 147 }, { "epoch": 0.016045099739809193, "grad_norm": 0.11481768637895584, "learning_rate": 9.982591034382708e-05, "loss": 0.1376, "step": 148 }, { "epoch": 0.016153512575888984, "grad_norm": 0.13576391339302063, "learning_rate": 9.982228347599014e-05, "loss": 0.1333, "step": 149 }, { "epoch": 0.01626192541196878, "grad_norm": 0.094816654920578, "learning_rate": 9.981865660815321e-05, "loss": 0.1196, "step": 150 }, { "epoch": 0.01637033824804857, "grad_norm": 0.20657342672348022, "learning_rate": 9.981502974031627e-05, "loss": 0.1382, "step": 151 }, { "epoch": 0.01647875108412836, "grad_norm": 0.11815548688173294, "learning_rate": 9.981140287247934e-05, "loss": 0.1171, "step": 152 }, { "epoch": 0.01658716392020815, "grad_norm": 0.09440053254365921, "learning_rate": 9.98077760046424e-05, "loss": 0.1192, "step": 153 }, { "epoch": 0.016695576756287946, "grad_norm": 0.12479250878095627, "learning_rate": 9.980414913680545e-05, "loss": 0.1274, "step": 154 }, { "epoch": 0.016803989592367737, "grad_norm": 0.11923300474882126, "learning_rate": 9.980052226896852e-05, "loss": 0.1236, "step": 155 }, { "epoch": 0.016912402428447527, "grad_norm": 0.13512563705444336, "learning_rate": 9.979689540113158e-05, "loss": 0.1139, "step": 156 }, { "epoch": 0.01702081526452732, "grad_norm": 0.09181004017591476, "learning_rate": 9.979326853329465e-05, "loss": 0.126, "step": 157 }, { "epoch": 0.017129228100607113, "grad_norm": 0.10445132851600647, "learning_rate": 9.978964166545771e-05, "loss": 0.1247, "step": 158 }, { "epoch": 0.017237640936686904, "grad_norm": 0.1570311337709427, "learning_rate": 9.978601479762078e-05, "loss": 0.1346, "step": 159 }, { "epoch": 0.017346053772766695, "grad_norm": 0.09328942745923996, "learning_rate": 9.978238792978384e-05, "loss": 0.1376, "step": 160 }, { "epoch": 0.01745446660884649, "grad_norm": 0.12497913092374802, "learning_rate": 9.977876106194691e-05, "loss": 0.1278, "step": 161 }, { "epoch": 0.01756287944492628, "grad_norm": 0.11073092371225357, "learning_rate": 9.977513419410998e-05, "loss": 0.1407, "step": 162 }, { "epoch": 0.01767129228100607, "grad_norm": 0.14218127727508545, "learning_rate": 9.977150732627304e-05, "loss": 0.1351, "step": 163 }, { "epoch": 0.01777970511708586, "grad_norm": 2.7658016681671143, "learning_rate": 9.976788045843611e-05, "loss": 0.1483, "step": 164 }, { "epoch": 0.017888117953165656, "grad_norm": 0.1440364122390747, "learning_rate": 9.976425359059916e-05, "loss": 0.1237, "step": 165 }, { "epoch": 0.017996530789245447, "grad_norm": 0.13687509298324585, "learning_rate": 9.976062672276224e-05, "loss": 0.1305, "step": 166 }, { "epoch": 0.018104943625325238, "grad_norm": 0.1770717054605484, "learning_rate": 9.975699985492529e-05, "loss": 0.138, "step": 167 }, { "epoch": 0.01821335646140503, "grad_norm": 0.13232694566249847, "learning_rate": 9.975337298708835e-05, "loss": 0.1214, "step": 168 }, { "epoch": 0.018321769297484823, "grad_norm": 0.11508698016405106, "learning_rate": 9.974974611925142e-05, "loss": 0.1223, "step": 169 }, { "epoch": 0.018430182133564614, "grad_norm": 0.1289430558681488, "learning_rate": 9.974611925141448e-05, "loss": 0.1239, "step": 170 }, { "epoch": 0.018538594969644405, "grad_norm": 0.10615533590316772, "learning_rate": 9.974249238357755e-05, "loss": 0.1268, "step": 171 }, { "epoch": 0.0186470078057242, "grad_norm": 0.1109352856874466, "learning_rate": 9.973886551574061e-05, "loss": 0.1519, "step": 172 }, { "epoch": 0.01875542064180399, "grad_norm": 0.13500618934631348, "learning_rate": 9.973523864790368e-05, "loss": 0.1201, "step": 173 }, { "epoch": 0.01886383347788378, "grad_norm": 0.10082557052373886, "learning_rate": 9.973161178006674e-05, "loss": 0.1171, "step": 174 }, { "epoch": 0.018972246313963572, "grad_norm": 0.10272687673568726, "learning_rate": 9.97279849122298e-05, "loss": 0.1324, "step": 175 }, { "epoch": 0.019080659150043366, "grad_norm": 0.11438053101301193, "learning_rate": 9.972435804439286e-05, "loss": 0.1267, "step": 176 }, { "epoch": 0.019189071986123157, "grad_norm": 0.13000522553920746, "learning_rate": 9.972073117655592e-05, "loss": 0.1514, "step": 177 }, { "epoch": 0.019297484822202948, "grad_norm": 0.12262362241744995, "learning_rate": 9.9717104308719e-05, "loss": 0.1122, "step": 178 }, { "epoch": 0.01940589765828274, "grad_norm": 0.1190502941608429, "learning_rate": 9.971347744088206e-05, "loss": 0.1308, "step": 179 }, { "epoch": 0.019514310494362534, "grad_norm": 0.10970817506313324, "learning_rate": 9.970985057304513e-05, "loss": 0.1281, "step": 180 }, { "epoch": 0.019622723330442324, "grad_norm": 0.09598585963249207, "learning_rate": 9.970622370520819e-05, "loss": 0.1079, "step": 181 }, { "epoch": 0.019731136166522115, "grad_norm": 0.15805421769618988, "learning_rate": 9.970259683737125e-05, "loss": 0.1476, "step": 182 }, { "epoch": 0.01983954900260191, "grad_norm": 0.324884831905365, "learning_rate": 9.969896996953432e-05, "loss": 0.159, "step": 183 }, { "epoch": 0.0199479618386817, "grad_norm": 0.38055792450904846, "learning_rate": 9.969534310169738e-05, "loss": 0.1273, "step": 184 }, { "epoch": 0.02005637467476149, "grad_norm": 0.11114319413900375, "learning_rate": 9.969171623386045e-05, "loss": 0.1118, "step": 185 }, { "epoch": 0.020164787510841282, "grad_norm": 0.6280705332756042, "learning_rate": 9.96880893660235e-05, "loss": 0.1303, "step": 186 }, { "epoch": 0.020273200346921077, "grad_norm": 0.15535187721252441, "learning_rate": 9.968446249818658e-05, "loss": 0.1196, "step": 187 }, { "epoch": 0.020381613183000868, "grad_norm": 0.1150515228509903, "learning_rate": 9.968083563034963e-05, "loss": 0.1342, "step": 188 }, { "epoch": 0.02049002601908066, "grad_norm": 0.11491328477859497, "learning_rate": 9.96772087625127e-05, "loss": 0.1237, "step": 189 }, { "epoch": 0.02059843885516045, "grad_norm": 0.17929574847221375, "learning_rate": 9.967358189467576e-05, "loss": 0.1312, "step": 190 }, { "epoch": 0.020706851691240244, "grad_norm": 0.13127996027469635, "learning_rate": 9.966995502683882e-05, "loss": 0.1489, "step": 191 }, { "epoch": 0.020815264527320035, "grad_norm": 0.10639460384845734, "learning_rate": 9.966632815900189e-05, "loss": 0.1323, "step": 192 }, { "epoch": 0.020923677363399826, "grad_norm": 0.12502805888652802, "learning_rate": 9.966270129116495e-05, "loss": 0.118, "step": 193 }, { "epoch": 0.021032090199479617, "grad_norm": 0.1380777359008789, "learning_rate": 9.965907442332802e-05, "loss": 0.1225, "step": 194 }, { "epoch": 0.02114050303555941, "grad_norm": 0.13549555838108063, "learning_rate": 9.965544755549107e-05, "loss": 0.1109, "step": 195 }, { "epoch": 0.021248915871639202, "grad_norm": 0.109560027718544, "learning_rate": 9.965182068765415e-05, "loss": 0.1252, "step": 196 }, { "epoch": 0.021357328707718993, "grad_norm": 0.1355946660041809, "learning_rate": 9.964819381981722e-05, "loss": 0.1156, "step": 197 }, { "epoch": 0.021465741543798787, "grad_norm": 0.09955989569425583, "learning_rate": 9.964456695198027e-05, "loss": 0.135, "step": 198 }, { "epoch": 0.021574154379878578, "grad_norm": 0.11638107895851135, "learning_rate": 9.964094008414334e-05, "loss": 0.1311, "step": 199 }, { "epoch": 0.02168256721595837, "grad_norm": 0.1709141731262207, "learning_rate": 9.96373132163064e-05, "loss": 0.1413, "step": 200 }, { "epoch": 0.02179098005203816, "grad_norm": 0.12056425213813782, "learning_rate": 9.963368634846947e-05, "loss": 0.124, "step": 201 }, { "epoch": 0.021899392888117954, "grad_norm": 0.14599131047725677, "learning_rate": 9.963005948063253e-05, "loss": 0.1193, "step": 202 }, { "epoch": 0.022007805724197745, "grad_norm": 0.13976173102855682, "learning_rate": 9.96264326127956e-05, "loss": 0.1126, "step": 203 }, { "epoch": 0.022116218560277536, "grad_norm": 0.1221957877278328, "learning_rate": 9.962280574495866e-05, "loss": 0.1229, "step": 204 }, { "epoch": 0.022224631396357327, "grad_norm": 0.1677732765674591, "learning_rate": 9.961917887712172e-05, "loss": 0.1288, "step": 205 }, { "epoch": 0.02233304423243712, "grad_norm": 0.16093505918979645, "learning_rate": 9.961555200928479e-05, "loss": 0.1096, "step": 206 }, { "epoch": 0.022441457068516912, "grad_norm": 0.15370431542396545, "learning_rate": 9.961192514144784e-05, "loss": 0.1336, "step": 207 }, { "epoch": 0.022549869904596703, "grad_norm": 0.20025312900543213, "learning_rate": 9.960829827361091e-05, "loss": 0.1197, "step": 208 }, { "epoch": 0.022658282740676498, "grad_norm": 0.30901384353637695, "learning_rate": 9.960467140577397e-05, "loss": 0.1379, "step": 209 }, { "epoch": 0.02276669557675629, "grad_norm": 0.1538674235343933, "learning_rate": 9.960104453793704e-05, "loss": 0.1329, "step": 210 }, { "epoch": 0.02287510841283608, "grad_norm": 0.1657291054725647, "learning_rate": 9.95974176701001e-05, "loss": 0.1294, "step": 211 }, { "epoch": 0.02298352124891587, "grad_norm": 0.20337271690368652, "learning_rate": 9.959379080226317e-05, "loss": 0.1351, "step": 212 }, { "epoch": 0.023091934084995665, "grad_norm": 0.2516423165798187, "learning_rate": 9.959016393442624e-05, "loss": 0.1299, "step": 213 }, { "epoch": 0.023200346921075456, "grad_norm": 0.24394896626472473, "learning_rate": 9.95865370665893e-05, "loss": 0.116, "step": 214 }, { "epoch": 0.023308759757155247, "grad_norm": 0.27288946509361267, "learning_rate": 9.958291019875237e-05, "loss": 0.1341, "step": 215 }, { "epoch": 0.023417172593235037, "grad_norm": 0.2672813832759857, "learning_rate": 9.957928333091543e-05, "loss": 0.1217, "step": 216 }, { "epoch": 0.023525585429314832, "grad_norm": 0.2543863356113434, "learning_rate": 9.95756564630785e-05, "loss": 0.109, "step": 217 }, { "epoch": 0.023633998265394623, "grad_norm": 0.25781533122062683, "learning_rate": 9.957202959524156e-05, "loss": 0.0908, "step": 218 }, { "epoch": 0.023742411101474414, "grad_norm": 0.3181903064250946, "learning_rate": 9.956840272740461e-05, "loss": 0.1084, "step": 219 }, { "epoch": 0.023850823937554208, "grad_norm": 0.3477591872215271, "learning_rate": 9.956477585956768e-05, "loss": 0.1162, "step": 220 }, { "epoch": 0.023959236773634, "grad_norm": 0.33487313985824585, "learning_rate": 9.956114899173074e-05, "loss": 0.1155, "step": 221 }, { "epoch": 0.02406764960971379, "grad_norm": 0.593386709690094, "learning_rate": 9.955752212389381e-05, "loss": 0.1271, "step": 222 }, { "epoch": 0.02417606244579358, "grad_norm": 0.6811215281486511, "learning_rate": 9.955389525605687e-05, "loss": 0.1323, "step": 223 }, { "epoch": 0.024284475281873375, "grad_norm": 0.8933523893356323, "learning_rate": 9.955026838821994e-05, "loss": 0.1702, "step": 224 }, { "epoch": 0.024392888117953166, "grad_norm": 0.322263240814209, "learning_rate": 9.9546641520383e-05, "loss": 0.1087, "step": 225 }, { "epoch": 0.024501300954032957, "grad_norm": 0.1742347925901413, "learning_rate": 9.954301465254607e-05, "loss": 0.1065, "step": 226 }, { "epoch": 0.024609713790112748, "grad_norm": 0.34003663063049316, "learning_rate": 9.953938778470913e-05, "loss": 0.1387, "step": 227 }, { "epoch": 0.024718126626192542, "grad_norm": 0.14683124423027039, "learning_rate": 9.953576091687218e-05, "loss": 0.1122, "step": 228 }, { "epoch": 0.024826539462272333, "grad_norm": 0.21677912771701813, "learning_rate": 9.953213404903525e-05, "loss": 0.122, "step": 229 }, { "epoch": 0.024934952298352124, "grad_norm": 0.1505880057811737, "learning_rate": 9.952850718119832e-05, "loss": 0.124, "step": 230 }, { "epoch": 0.02504336513443192, "grad_norm": 0.16343818604946136, "learning_rate": 9.95248803133614e-05, "loss": 0.1269, "step": 231 }, { "epoch": 0.02515177797051171, "grad_norm": 0.1292470097541809, "learning_rate": 9.952125344552445e-05, "loss": 0.1203, "step": 232 }, { "epoch": 0.0252601908065915, "grad_norm": 0.13108700513839722, "learning_rate": 9.951762657768751e-05, "loss": 0.114, "step": 233 }, { "epoch": 0.02536860364267129, "grad_norm": 0.16893452405929565, "learning_rate": 9.951399970985058e-05, "loss": 0.1337, "step": 234 }, { "epoch": 0.025477016478751086, "grad_norm": 0.12745606899261475, "learning_rate": 9.951037284201364e-05, "loss": 0.1153, "step": 235 }, { "epoch": 0.025585429314830876, "grad_norm": 0.17183400690555573, "learning_rate": 9.950674597417671e-05, "loss": 0.1201, "step": 236 }, { "epoch": 0.025693842150910667, "grad_norm": 0.13243946433067322, "learning_rate": 9.950311910633977e-05, "loss": 0.1258, "step": 237 }, { "epoch": 0.02580225498699046, "grad_norm": 0.12578503787517548, "learning_rate": 9.949949223850284e-05, "loss": 0.1278, "step": 238 }, { "epoch": 0.025910667823070253, "grad_norm": 0.17366163432598114, "learning_rate": 9.94958653706659e-05, "loss": 0.1483, "step": 239 }, { "epoch": 0.026019080659150044, "grad_norm": 0.19561327993869781, "learning_rate": 9.949223850282897e-05, "loss": 0.119, "step": 240 }, { "epoch": 0.026127493495229834, "grad_norm": 0.19692310690879822, "learning_rate": 9.948861163499202e-05, "loss": 0.1207, "step": 241 }, { "epoch": 0.026235906331309625, "grad_norm": 0.12393520027399063, "learning_rate": 9.948498476715508e-05, "loss": 0.1148, "step": 242 }, { "epoch": 0.02634431916738942, "grad_norm": 0.22765368223190308, "learning_rate": 9.948135789931815e-05, "loss": 0.1437, "step": 243 }, { "epoch": 0.02645273200346921, "grad_norm": 0.1626598984003067, "learning_rate": 9.947773103148121e-05, "loss": 0.1208, "step": 244 }, { "epoch": 0.026561144839549, "grad_norm": 0.13885174691677094, "learning_rate": 9.947410416364428e-05, "loss": 0.1117, "step": 245 }, { "epoch": 0.026669557675628796, "grad_norm": 0.1614433228969574, "learning_rate": 9.947047729580734e-05, "loss": 0.1188, "step": 246 }, { "epoch": 0.026777970511708587, "grad_norm": 0.1626748889684677, "learning_rate": 9.946685042797042e-05, "loss": 0.1406, "step": 247 }, { "epoch": 0.026886383347788378, "grad_norm": 0.1448822021484375, "learning_rate": 9.946322356013348e-05, "loss": 0.1098, "step": 248 }, { "epoch": 0.02699479618386817, "grad_norm": 0.2266145497560501, "learning_rate": 9.945959669229654e-05, "loss": 0.1275, "step": 249 }, { "epoch": 0.027103209019947963, "grad_norm": 0.1260368824005127, "learning_rate": 9.945596982445961e-05, "loss": 0.1197, "step": 250 }, { "epoch": 0.027211621856027754, "grad_norm": 0.15092907845973969, "learning_rate": 9.945234295662266e-05, "loss": 0.1092, "step": 251 }, { "epoch": 0.027320034692107545, "grad_norm": 0.19695094227790833, "learning_rate": 9.944871608878574e-05, "loss": 0.1052, "step": 252 }, { "epoch": 0.027428447528187336, "grad_norm": 0.2634161710739136, "learning_rate": 9.944508922094879e-05, "loss": 0.1417, "step": 253 }, { "epoch": 0.02753686036426713, "grad_norm": 0.16970281302928925, "learning_rate": 9.944146235311186e-05, "loss": 0.1049, "step": 254 }, { "epoch": 0.02764527320034692, "grad_norm": 0.3959355354309082, "learning_rate": 9.943783548527492e-05, "loss": 0.1356, "step": 255 }, { "epoch": 0.027753686036426712, "grad_norm": 0.3768247365951538, "learning_rate": 9.943420861743799e-05, "loss": 0.1579, "step": 256 }, { "epoch": 0.027862098872506506, "grad_norm": 0.3114122152328491, "learning_rate": 9.943058174960105e-05, "loss": 0.1403, "step": 257 }, { "epoch": 0.027970511708586297, "grad_norm": 0.4493565559387207, "learning_rate": 9.94269548817641e-05, "loss": 0.131, "step": 258 }, { "epoch": 0.028078924544666088, "grad_norm": 0.19529375433921814, "learning_rate": 9.942332801392718e-05, "loss": 0.1131, "step": 259 }, { "epoch": 0.02818733738074588, "grad_norm": 0.3056773245334625, "learning_rate": 9.941970114609023e-05, "loss": 0.0923, "step": 260 }, { "epoch": 0.028295750216825673, "grad_norm": 1.9880496263504028, "learning_rate": 9.94160742782533e-05, "loss": 0.482, "step": 261 }, { "epoch": 0.028404163052905464, "grad_norm": 0.348686158657074, "learning_rate": 9.941244741041636e-05, "loss": 0.1402, "step": 262 }, { "epoch": 0.028512575888985255, "grad_norm": 0.4005283713340759, "learning_rate": 9.940882054257943e-05, "loss": 0.1029, "step": 263 }, { "epoch": 0.028620988725065046, "grad_norm": 0.32057470083236694, "learning_rate": 9.940519367474249e-05, "loss": 0.1151, "step": 264 }, { "epoch": 0.02872940156114484, "grad_norm": 0.7335758209228516, "learning_rate": 9.940156680690556e-05, "loss": 0.1861, "step": 265 }, { "epoch": 0.02883781439722463, "grad_norm": 0.6374050974845886, "learning_rate": 9.939793993906863e-05, "loss": 0.1254, "step": 266 }, { "epoch": 0.028946227233304422, "grad_norm": 0.6786290407180786, "learning_rate": 9.939431307123169e-05, "loss": 0.1433, "step": 267 }, { "epoch": 0.029054640069384217, "grad_norm": 0.4154859185218811, "learning_rate": 9.939068620339476e-05, "loss": 0.154, "step": 268 }, { "epoch": 0.029163052905464008, "grad_norm": 0.2123202085494995, "learning_rate": 9.938705933555782e-05, "loss": 0.1128, "step": 269 }, { "epoch": 0.0292714657415438, "grad_norm": 0.2835502028465271, "learning_rate": 9.938343246772089e-05, "loss": 0.1268, "step": 270 }, { "epoch": 0.02937987857762359, "grad_norm": 0.20654474198818207, "learning_rate": 9.937980559988395e-05, "loss": 0.1125, "step": 271 }, { "epoch": 0.029488291413703384, "grad_norm": 0.23350019752979279, "learning_rate": 9.9376178732047e-05, "loss": 0.1146, "step": 272 }, { "epoch": 0.029596704249783175, "grad_norm": 0.38479769229888916, "learning_rate": 9.937255186421007e-05, "loss": 0.1306, "step": 273 }, { "epoch": 0.029705117085862966, "grad_norm": 0.19484680891036987, "learning_rate": 9.936892499637313e-05, "loss": 0.0888, "step": 274 }, { "epoch": 0.029813529921942757, "grad_norm": 0.21060413122177124, "learning_rate": 9.93652981285362e-05, "loss": 0.1178, "step": 275 }, { "epoch": 0.02992194275802255, "grad_norm": 0.289014607667923, "learning_rate": 9.936167126069926e-05, "loss": 0.1047, "step": 276 }, { "epoch": 0.030030355594102342, "grad_norm": 0.30091777443885803, "learning_rate": 9.935804439286233e-05, "loss": 0.1264, "step": 277 }, { "epoch": 0.030138768430182133, "grad_norm": 0.3205192983150482, "learning_rate": 9.935441752502539e-05, "loss": 0.1284, "step": 278 }, { "epoch": 0.030247181266261924, "grad_norm": 0.38515546917915344, "learning_rate": 9.935079065718846e-05, "loss": 0.1289, "step": 279 }, { "epoch": 0.030355594102341718, "grad_norm": 0.1982385665178299, "learning_rate": 9.934716378935152e-05, "loss": 0.105, "step": 280 }, { "epoch": 0.03046400693842151, "grad_norm": 0.45646801590919495, "learning_rate": 9.934353692151457e-05, "loss": 0.1044, "step": 281 }, { "epoch": 0.0305724197745013, "grad_norm": 0.2734658718109131, "learning_rate": 9.933991005367766e-05, "loss": 0.1149, "step": 282 }, { "epoch": 0.030680832610581094, "grad_norm": 0.5013412237167358, "learning_rate": 9.933628318584072e-05, "loss": 0.1296, "step": 283 }, { "epoch": 0.030789245446660885, "grad_norm": 0.26244890689849854, "learning_rate": 9.933265631800379e-05, "loss": 0.1053, "step": 284 }, { "epoch": 0.030897658282740676, "grad_norm": 0.36835768818855286, "learning_rate": 9.932902945016684e-05, "loss": 0.1198, "step": 285 }, { "epoch": 0.031006071118820467, "grad_norm": 0.47366178035736084, "learning_rate": 9.93254025823299e-05, "loss": 0.1465, "step": 286 }, { "epoch": 0.03111448395490026, "grad_norm": 0.28092706203460693, "learning_rate": 9.932177571449297e-05, "loss": 0.0801, "step": 287 }, { "epoch": 0.031222896790980052, "grad_norm": 0.4462004005908966, "learning_rate": 9.931814884665603e-05, "loss": 0.1102, "step": 288 }, { "epoch": 0.03133130962705984, "grad_norm": 0.33671125769615173, "learning_rate": 9.93145219788191e-05, "loss": 0.0959, "step": 289 }, { "epoch": 0.031439722463139634, "grad_norm": 0.4191037118434906, "learning_rate": 9.931089511098216e-05, "loss": 0.1125, "step": 290 }, { "epoch": 0.031548135299219425, "grad_norm": 0.5278077125549316, "learning_rate": 9.930726824314523e-05, "loss": 0.1316, "step": 291 }, { "epoch": 0.03165654813529922, "grad_norm": 0.4065866470336914, "learning_rate": 9.930364137530829e-05, "loss": 0.1204, "step": 292 }, { "epoch": 0.031764960971379014, "grad_norm": 0.30495551228523254, "learning_rate": 9.930001450747136e-05, "loss": 0.0854, "step": 293 }, { "epoch": 0.031873373807458805, "grad_norm": 0.4601772427558899, "learning_rate": 9.929638763963441e-05, "loss": 0.1129, "step": 294 }, { "epoch": 0.031981786643538596, "grad_norm": 0.5568308234214783, "learning_rate": 9.929276077179747e-05, "loss": 0.1256, "step": 295 }, { "epoch": 0.032090199479618386, "grad_norm": 0.5129754543304443, "learning_rate": 9.928913390396054e-05, "loss": 0.1115, "step": 296 }, { "epoch": 0.03219861231569818, "grad_norm": 0.4528515934944153, "learning_rate": 9.92855070361236e-05, "loss": 0.1192, "step": 297 }, { "epoch": 0.03230702515177797, "grad_norm": 0.29402443766593933, "learning_rate": 9.928188016828667e-05, "loss": 0.0873, "step": 298 }, { "epoch": 0.03241543798785776, "grad_norm": 0.4222450256347656, "learning_rate": 9.927825330044974e-05, "loss": 0.0952, "step": 299 }, { "epoch": 0.03252385082393756, "grad_norm": 0.5187278389930725, "learning_rate": 9.92746264326128e-05, "loss": 0.1351, "step": 300 }, { "epoch": 0.03263226366001735, "grad_norm": 0.3391411006450653, "learning_rate": 9.927099956477587e-05, "loss": 0.1053, "step": 301 }, { "epoch": 0.03274067649609714, "grad_norm": 0.4005550444126129, "learning_rate": 9.926737269693893e-05, "loss": 0.131, "step": 302 }, { "epoch": 0.03284908933217693, "grad_norm": 0.3632897436618805, "learning_rate": 9.9263745829102e-05, "loss": 0.1021, "step": 303 }, { "epoch": 0.03295750216825672, "grad_norm": 0.5410080552101135, "learning_rate": 9.926011896126506e-05, "loss": 0.1291, "step": 304 }, { "epoch": 0.03306591500433651, "grad_norm": 0.552513062953949, "learning_rate": 9.925649209342813e-05, "loss": 0.1358, "step": 305 }, { "epoch": 0.0331743278404163, "grad_norm": 0.4205426573753357, "learning_rate": 9.925286522559118e-05, "loss": 0.1162, "step": 306 }, { "epoch": 0.0332827406764961, "grad_norm": 0.5091808438301086, "learning_rate": 9.924923835775425e-05, "loss": 0.1267, "step": 307 }, { "epoch": 0.03339115351257589, "grad_norm": 0.4898209273815155, "learning_rate": 9.924561148991731e-05, "loss": 0.1235, "step": 308 }, { "epoch": 0.03349956634865568, "grad_norm": 0.3675508201122284, "learning_rate": 9.924198462208037e-05, "loss": 0.1246, "step": 309 }, { "epoch": 0.03360797918473547, "grad_norm": 0.40086376667022705, "learning_rate": 9.923835775424344e-05, "loss": 0.0914, "step": 310 }, { "epoch": 0.033716392020815264, "grad_norm": 0.3581847846508026, "learning_rate": 9.92347308864065e-05, "loss": 0.1163, "step": 311 }, { "epoch": 0.033824804856895055, "grad_norm": 0.30943164229393005, "learning_rate": 9.923110401856957e-05, "loss": 0.1208, "step": 312 }, { "epoch": 0.033933217692974846, "grad_norm": 0.29413482546806335, "learning_rate": 9.922747715073263e-05, "loss": 0.1193, "step": 313 }, { "epoch": 0.03404163052905464, "grad_norm": 0.3196796476840973, "learning_rate": 9.92238502828957e-05, "loss": 0.1169, "step": 314 }, { "epoch": 0.034150043365134435, "grad_norm": 0.4101099967956543, "learning_rate": 9.922022341505875e-05, "loss": 0.1097, "step": 315 }, { "epoch": 0.034258456201214225, "grad_norm": 0.3664577007293701, "learning_rate": 9.921659654722182e-05, "loss": 0.1191, "step": 316 }, { "epoch": 0.034366869037294016, "grad_norm": 0.2721303105354309, "learning_rate": 9.92129696793849e-05, "loss": 0.1184, "step": 317 }, { "epoch": 0.03447528187337381, "grad_norm": 0.4380200505256653, "learning_rate": 9.920934281154795e-05, "loss": 0.1121, "step": 318 }, { "epoch": 0.0345836947094536, "grad_norm": 0.28323331475257874, "learning_rate": 9.920571594371102e-05, "loss": 0.1035, "step": 319 }, { "epoch": 0.03469210754553339, "grad_norm": 0.3159482777118683, "learning_rate": 9.920208907587408e-05, "loss": 0.0765, "step": 320 }, { "epoch": 0.03480052038161318, "grad_norm": 7.653610706329346, "learning_rate": 9.919846220803715e-05, "loss": 0.1942, "step": 321 }, { "epoch": 0.03490893321769298, "grad_norm": 0.3445223569869995, "learning_rate": 9.919483534020021e-05, "loss": 0.0997, "step": 322 }, { "epoch": 0.03501734605377277, "grad_norm": 0.42163312435150146, "learning_rate": 9.919120847236327e-05, "loss": 0.1103, "step": 323 }, { "epoch": 0.03512575888985256, "grad_norm": 0.5415775775909424, "learning_rate": 9.918758160452634e-05, "loss": 0.0956, "step": 324 }, { "epoch": 0.03523417172593235, "grad_norm": 0.4903617799282074, "learning_rate": 9.91839547366894e-05, "loss": 0.1017, "step": 325 }, { "epoch": 0.03534258456201214, "grad_norm": 0.4802268147468567, "learning_rate": 9.918032786885247e-05, "loss": 0.0761, "step": 326 }, { "epoch": 0.03545099739809193, "grad_norm": 0.44067537784576416, "learning_rate": 9.917670100101552e-05, "loss": 0.1025, "step": 327 }, { "epoch": 0.03555941023417172, "grad_norm": 0.549714207649231, "learning_rate": 9.91730741331786e-05, "loss": 0.0896, "step": 328 }, { "epoch": 0.03566782307025152, "grad_norm": 0.7183760404586792, "learning_rate": 9.916944726534165e-05, "loss": 0.1347, "step": 329 }, { "epoch": 0.03577623590633131, "grad_norm": 0.4663676619529724, "learning_rate": 9.916582039750472e-05, "loss": 0.1178, "step": 330 }, { "epoch": 0.0358846487424111, "grad_norm": 0.7179635763168335, "learning_rate": 9.916219352966778e-05, "loss": 0.1448, "step": 331 }, { "epoch": 0.035993061578490894, "grad_norm": 1.1327431201934814, "learning_rate": 9.915856666183084e-05, "loss": 0.1243, "step": 332 }, { "epoch": 0.036101474414570685, "grad_norm": 0.6648373007774353, "learning_rate": 9.915493979399391e-05, "loss": 0.1396, "step": 333 }, { "epoch": 0.036209887250650476, "grad_norm": 0.6019517183303833, "learning_rate": 9.915131292615698e-05, "loss": 0.1079, "step": 334 }, { "epoch": 0.03631830008673027, "grad_norm": 0.5720371007919312, "learning_rate": 9.914768605832005e-05, "loss": 0.1167, "step": 335 }, { "epoch": 0.03642671292281006, "grad_norm": 0.39956673979759216, "learning_rate": 9.91440591904831e-05, "loss": 0.0824, "step": 336 }, { "epoch": 0.036535125758889855, "grad_norm": 0.688218355178833, "learning_rate": 9.914043232264618e-05, "loss": 0.144, "step": 337 }, { "epoch": 0.036643538594969646, "grad_norm": 0.6920549273490906, "learning_rate": 9.913680545480923e-05, "loss": 0.1175, "step": 338 }, { "epoch": 0.03675195143104944, "grad_norm": 0.468303918838501, "learning_rate": 9.913317858697229e-05, "loss": 0.1262, "step": 339 }, { "epoch": 0.03686036426712923, "grad_norm": 0.36143118143081665, "learning_rate": 9.912955171913536e-05, "loss": 0.0949, "step": 340 }, { "epoch": 0.03696877710320902, "grad_norm": 0.27765920758247375, "learning_rate": 9.912592485129842e-05, "loss": 0.092, "step": 341 }, { "epoch": 0.03707718993928881, "grad_norm": 0.3729793429374695, "learning_rate": 9.912229798346149e-05, "loss": 0.0977, "step": 342 }, { "epoch": 0.0371856027753686, "grad_norm": 0.3784249722957611, "learning_rate": 9.911867111562455e-05, "loss": 0.0966, "step": 343 }, { "epoch": 0.0372940156114484, "grad_norm": 0.35683998465538025, "learning_rate": 9.911504424778762e-05, "loss": 0.0962, "step": 344 }, { "epoch": 0.03740242844752819, "grad_norm": 0.34519296884536743, "learning_rate": 9.911141737995068e-05, "loss": 0.0892, "step": 345 }, { "epoch": 0.03751084128360798, "grad_norm": 0.42365941405296326, "learning_rate": 9.910779051211373e-05, "loss": 0.1128, "step": 346 }, { "epoch": 0.03761925411968777, "grad_norm": 0.4882600009441376, "learning_rate": 9.91041636442768e-05, "loss": 0.1019, "step": 347 }, { "epoch": 0.03772766695576756, "grad_norm": 0.5526384711265564, "learning_rate": 9.910053677643986e-05, "loss": 0.0902, "step": 348 }, { "epoch": 0.03783607979184735, "grad_norm": 0.7365081310272217, "learning_rate": 9.909690990860293e-05, "loss": 0.113, "step": 349 }, { "epoch": 0.037944492627927144, "grad_norm": 1.2827476263046265, "learning_rate": 9.909328304076599e-05, "loss": 0.1137, "step": 350 }, { "epoch": 0.038052905464006935, "grad_norm": 0.6406203508377075, "learning_rate": 9.908965617292907e-05, "loss": 0.1272, "step": 351 }, { "epoch": 0.03816131830008673, "grad_norm": 0.6827720999717712, "learning_rate": 9.908602930509213e-05, "loss": 0.0836, "step": 352 }, { "epoch": 0.038269731136166524, "grad_norm": 0.3137438893318176, "learning_rate": 9.908240243725519e-05, "loss": 0.0615, "step": 353 }, { "epoch": 0.038378143972246315, "grad_norm": 0.4964936375617981, "learning_rate": 9.907877556941826e-05, "loss": 0.1075, "step": 354 }, { "epoch": 0.038486556808326106, "grad_norm": 0.8072117567062378, "learning_rate": 9.907514870158132e-05, "loss": 0.12, "step": 355 }, { "epoch": 0.038594969644405897, "grad_norm": 0.45715269446372986, "learning_rate": 9.907152183374439e-05, "loss": 0.1139, "step": 356 }, { "epoch": 0.03870338248048569, "grad_norm": 0.6667345762252808, "learning_rate": 9.906789496590745e-05, "loss": 0.0911, "step": 357 }, { "epoch": 0.03881179531656548, "grad_norm": 0.40916937589645386, "learning_rate": 9.906426809807052e-05, "loss": 0.0783, "step": 358 }, { "epoch": 0.038920208152645276, "grad_norm": 0.6383957266807556, "learning_rate": 9.906064123023357e-05, "loss": 0.1206, "step": 359 }, { "epoch": 0.03902862098872507, "grad_norm": 1.4137427806854248, "learning_rate": 9.905701436239665e-05, "loss": 0.0945, "step": 360 }, { "epoch": 0.03913703382480486, "grad_norm": 0.5083688497543335, "learning_rate": 9.90533874945597e-05, "loss": 0.0727, "step": 361 }, { "epoch": 0.03924544666088465, "grad_norm": 0.47528916597366333, "learning_rate": 9.904976062672276e-05, "loss": 0.0823, "step": 362 }, { "epoch": 0.03935385949696444, "grad_norm": 0.41194164752960205, "learning_rate": 9.904613375888583e-05, "loss": 0.0818, "step": 363 }, { "epoch": 0.03946227233304423, "grad_norm": 0.9217107892036438, "learning_rate": 9.904250689104889e-05, "loss": 0.0703, "step": 364 }, { "epoch": 0.03957068516912402, "grad_norm": 0.5941664576530457, "learning_rate": 9.903888002321196e-05, "loss": 0.0735, "step": 365 }, { "epoch": 0.03967909800520382, "grad_norm": 0.5896227359771729, "learning_rate": 9.903525315537502e-05, "loss": 0.1272, "step": 366 }, { "epoch": 0.03978751084128361, "grad_norm": 5.148974418640137, "learning_rate": 9.903162628753809e-05, "loss": 0.3588, "step": 367 }, { "epoch": 0.0398959236773634, "grad_norm": 0.8603709936141968, "learning_rate": 9.902799941970116e-05, "loss": 0.1278, "step": 368 }, { "epoch": 0.04000433651344319, "grad_norm": 0.9749218821525574, "learning_rate": 9.902437255186422e-05, "loss": 0.1439, "step": 369 }, { "epoch": 0.04011274934952298, "grad_norm": 0.8671954274177551, "learning_rate": 9.902074568402729e-05, "loss": 0.0788, "step": 370 }, { "epoch": 0.040221162185602774, "grad_norm": 0.43109431862831116, "learning_rate": 9.901711881619034e-05, "loss": 0.0794, "step": 371 }, { "epoch": 0.040329575021682565, "grad_norm": 0.7063488960266113, "learning_rate": 9.901349194835341e-05, "loss": 0.1204, "step": 372 }, { "epoch": 0.040437987857762356, "grad_norm": 0.2846534848213196, "learning_rate": 9.900986508051647e-05, "loss": 0.0638, "step": 373 }, { "epoch": 0.040546400693842154, "grad_norm": 0.608674943447113, "learning_rate": 9.900623821267954e-05, "loss": 0.1279, "step": 374 }, { "epoch": 0.040654813529921945, "grad_norm": 0.5713672637939453, "learning_rate": 9.90026113448426e-05, "loss": 0.1212, "step": 375 }, { "epoch": 0.040763226366001735, "grad_norm": 0.6724212169647217, "learning_rate": 9.899898447700566e-05, "loss": 0.0797, "step": 376 }, { "epoch": 0.040871639202081526, "grad_norm": 1.4908140897750854, "learning_rate": 9.899535760916873e-05, "loss": 0.1207, "step": 377 }, { "epoch": 0.04098005203816132, "grad_norm": 0.5947487354278564, "learning_rate": 9.899173074133179e-05, "loss": 0.1046, "step": 378 }, { "epoch": 0.04108846487424111, "grad_norm": 0.3444848358631134, "learning_rate": 9.898810387349486e-05, "loss": 0.1106, "step": 379 }, { "epoch": 0.0411968777103209, "grad_norm": 0.7131757736206055, "learning_rate": 9.898447700565791e-05, "loss": 0.129, "step": 380 }, { "epoch": 0.0413052905464007, "grad_norm": 0.761031985282898, "learning_rate": 9.898085013782098e-05, "loss": 0.098, "step": 381 }, { "epoch": 0.04141370338248049, "grad_norm": 0.27112630009651184, "learning_rate": 9.897722326998404e-05, "loss": 0.0833, "step": 382 }, { "epoch": 0.04152211621856028, "grad_norm": 0.6176376342773438, "learning_rate": 9.897359640214711e-05, "loss": 0.1342, "step": 383 }, { "epoch": 0.04163052905464007, "grad_norm": 0.2600693702697754, "learning_rate": 9.896996953431017e-05, "loss": 0.0827, "step": 384 }, { "epoch": 0.04173894189071986, "grad_norm": 0.3771299719810486, "learning_rate": 9.896634266647323e-05, "loss": 0.1059, "step": 385 }, { "epoch": 0.04184735472679965, "grad_norm": 0.5059844255447388, "learning_rate": 9.896271579863631e-05, "loss": 0.0977, "step": 386 }, { "epoch": 0.04195576756287944, "grad_norm": 0.37870562076568604, "learning_rate": 9.895908893079937e-05, "loss": 0.1062, "step": 387 }, { "epoch": 0.04206418039895923, "grad_norm": 0.7182376384735107, "learning_rate": 9.895546206296244e-05, "loss": 0.1223, "step": 388 }, { "epoch": 0.04217259323503903, "grad_norm": 0.4013010859489441, "learning_rate": 9.89518351951255e-05, "loss": 0.1061, "step": 389 }, { "epoch": 0.04228100607111882, "grad_norm": 0.3709091246128082, "learning_rate": 9.894820832728855e-05, "loss": 0.0945, "step": 390 }, { "epoch": 0.04238941890719861, "grad_norm": 0.5855059027671814, "learning_rate": 9.894458145945163e-05, "loss": 0.1041, "step": 391 }, { "epoch": 0.042497831743278404, "grad_norm": 0.5612392425537109, "learning_rate": 9.894095459161468e-05, "loss": 0.0862, "step": 392 }, { "epoch": 0.042606244579358195, "grad_norm": 0.41481488943099976, "learning_rate": 9.893732772377775e-05, "loss": 0.095, "step": 393 }, { "epoch": 0.042714657415437986, "grad_norm": 0.6477811932563782, "learning_rate": 9.893370085594081e-05, "loss": 0.0834, "step": 394 }, { "epoch": 0.04282307025151778, "grad_norm": 0.48517316579818726, "learning_rate": 9.893007398810388e-05, "loss": 0.0671, "step": 395 }, { "epoch": 0.042931483087597574, "grad_norm": 0.6010326743125916, "learning_rate": 9.892644712026694e-05, "loss": 0.1015, "step": 396 }, { "epoch": 0.043039895923677365, "grad_norm": 0.44709518551826477, "learning_rate": 9.892282025243001e-05, "loss": 0.0723, "step": 397 }, { "epoch": 0.043148308759757156, "grad_norm": 0.7340505123138428, "learning_rate": 9.891919338459307e-05, "loss": 0.0926, "step": 398 }, { "epoch": 0.04325672159583695, "grad_norm": 0.8364338278770447, "learning_rate": 9.891556651675612e-05, "loss": 0.0833, "step": 399 }, { "epoch": 0.04336513443191674, "grad_norm": 0.5436657667160034, "learning_rate": 9.89119396489192e-05, "loss": 0.1428, "step": 400 }, { "epoch": 0.04347354726799653, "grad_norm": 0.33888542652130127, "learning_rate": 9.890831278108225e-05, "loss": 0.0705, "step": 401 }, { "epoch": 0.04358196010407632, "grad_norm": 0.3756312131881714, "learning_rate": 9.890468591324532e-05, "loss": 0.0673, "step": 402 }, { "epoch": 0.04369037294015612, "grad_norm": 0.8623625636100769, "learning_rate": 9.89010590454084e-05, "loss": 0.1066, "step": 403 }, { "epoch": 0.04379878577623591, "grad_norm": 11.913687705993652, "learning_rate": 9.889743217757145e-05, "loss": 0.1577, "step": 404 }, { "epoch": 0.0439071986123157, "grad_norm": 1.269169569015503, "learning_rate": 9.889380530973452e-05, "loss": 0.1281, "step": 405 }, { "epoch": 0.04401561144839549, "grad_norm": 0.45170003175735474, "learning_rate": 9.889017844189758e-05, "loss": 0.1096, "step": 406 }, { "epoch": 0.04412402428447528, "grad_norm": 0.5787107348442078, "learning_rate": 9.888655157406065e-05, "loss": 0.0953, "step": 407 }, { "epoch": 0.04423243712055507, "grad_norm": 0.6357136964797974, "learning_rate": 9.888292470622371e-05, "loss": 0.0721, "step": 408 }, { "epoch": 0.04434084995663486, "grad_norm": 0.5221465826034546, "learning_rate": 9.887929783838678e-05, "loss": 0.1308, "step": 409 }, { "epoch": 0.044449262792714654, "grad_norm": 0.4664268493652344, "learning_rate": 9.887567097054984e-05, "loss": 0.0928, "step": 410 }, { "epoch": 0.04455767562879445, "grad_norm": 0.4967299699783325, "learning_rate": 9.887204410271291e-05, "loss": 0.102, "step": 411 }, { "epoch": 0.04466608846487424, "grad_norm": 0.4468687176704407, "learning_rate": 9.886841723487596e-05, "loss": 0.0925, "step": 412 }, { "epoch": 0.044774501300954034, "grad_norm": 0.4192066788673401, "learning_rate": 9.886479036703902e-05, "loss": 0.0772, "step": 413 }, { "epoch": 0.044882914137033825, "grad_norm": 0.38501182198524475, "learning_rate": 9.886116349920209e-05, "loss": 0.0908, "step": 414 }, { "epoch": 0.044991326973113616, "grad_norm": 0.49038606882095337, "learning_rate": 9.885753663136515e-05, "loss": 0.095, "step": 415 }, { "epoch": 0.045099739809193407, "grad_norm": 3.823589563369751, "learning_rate": 9.885390976352822e-05, "loss": 0.1025, "step": 416 }, { "epoch": 0.0452081526452732, "grad_norm": 0.3266782760620117, "learning_rate": 9.885028289569128e-05, "loss": 0.0826, "step": 417 }, { "epoch": 0.045316565481352995, "grad_norm": 0.6704066395759583, "learning_rate": 9.884665602785435e-05, "loss": 0.112, "step": 418 }, { "epoch": 0.045424978317432786, "grad_norm": 0.6655594706535339, "learning_rate": 9.884302916001741e-05, "loss": 0.0843, "step": 419 }, { "epoch": 0.04553339115351258, "grad_norm": 0.2981758415699005, "learning_rate": 9.883940229218048e-05, "loss": 0.1045, "step": 420 }, { "epoch": 0.04564180398959237, "grad_norm": 0.4019629657268524, "learning_rate": 9.883577542434355e-05, "loss": 0.0783, "step": 421 }, { "epoch": 0.04575021682567216, "grad_norm": 0.5087241530418396, "learning_rate": 9.88321485565066e-05, "loss": 0.0832, "step": 422 }, { "epoch": 0.04585862966175195, "grad_norm": 0.3728336989879608, "learning_rate": 9.882852168866968e-05, "loss": 0.0962, "step": 423 }, { "epoch": 0.04596704249783174, "grad_norm": 0.5468832850456238, "learning_rate": 9.882489482083273e-05, "loss": 0.0897, "step": 424 }, { "epoch": 0.04607545533391154, "grad_norm": 0.604433000087738, "learning_rate": 9.88212679529958e-05, "loss": 0.098, "step": 425 }, { "epoch": 0.04618386816999133, "grad_norm": 0.5347226858139038, "learning_rate": 9.881764108515886e-05, "loss": 0.0705, "step": 426 }, { "epoch": 0.04629228100607112, "grad_norm": 0.34253522753715515, "learning_rate": 9.881401421732192e-05, "loss": 0.0834, "step": 427 }, { "epoch": 0.04640069384215091, "grad_norm": 0.5860463976860046, "learning_rate": 9.881038734948499e-05, "loss": 0.066, "step": 428 }, { "epoch": 0.0465091066782307, "grad_norm": 0.3532988131046295, "learning_rate": 9.880676048164805e-05, "loss": 0.0692, "step": 429 }, { "epoch": 0.04661751951431049, "grad_norm": 0.6648280024528503, "learning_rate": 9.880313361381112e-05, "loss": 0.1061, "step": 430 }, { "epoch": 0.046725932350390284, "grad_norm": 0.6974600553512573, "learning_rate": 9.879950674597418e-05, "loss": 0.0877, "step": 431 }, { "epoch": 0.046834345186470075, "grad_norm": 0.6680644750595093, "learning_rate": 9.879587987813725e-05, "loss": 0.0724, "step": 432 }, { "epoch": 0.04694275802254987, "grad_norm": 0.44691362977027893, "learning_rate": 9.87922530103003e-05, "loss": 0.0859, "step": 433 }, { "epoch": 0.047051170858629664, "grad_norm": 0.6795225739479065, "learning_rate": 9.878862614246338e-05, "loss": 0.0741, "step": 434 }, { "epoch": 0.047159583694709455, "grad_norm": 0.3884204924106598, "learning_rate": 9.878499927462643e-05, "loss": 0.0726, "step": 435 }, { "epoch": 0.047267996530789246, "grad_norm": 0.950244128704071, "learning_rate": 9.878137240678949e-05, "loss": 0.1633, "step": 436 }, { "epoch": 0.047376409366869036, "grad_norm": 0.6694636344909668, "learning_rate": 9.877774553895257e-05, "loss": 0.1065, "step": 437 }, { "epoch": 0.04748482220294883, "grad_norm": 1.0101580619812012, "learning_rate": 9.877411867111563e-05, "loss": 0.0935, "step": 438 }, { "epoch": 0.04759323503902862, "grad_norm": 0.6740319132804871, "learning_rate": 9.87704918032787e-05, "loss": 0.095, "step": 439 }, { "epoch": 0.047701647875108416, "grad_norm": 0.5629370212554932, "learning_rate": 9.876686493544176e-05, "loss": 0.0822, "step": 440 }, { "epoch": 0.04781006071118821, "grad_norm": 0.578711211681366, "learning_rate": 9.876323806760483e-05, "loss": 0.0687, "step": 441 }, { "epoch": 0.047918473547268, "grad_norm": 0.4391231834888458, "learning_rate": 9.875961119976789e-05, "loss": 0.0741, "step": 442 }, { "epoch": 0.04802688638334779, "grad_norm": 0.45237937569618225, "learning_rate": 9.875598433193095e-05, "loss": 0.0852, "step": 443 }, { "epoch": 0.04813529921942758, "grad_norm": 0.2997264862060547, "learning_rate": 9.875235746409402e-05, "loss": 0.0797, "step": 444 }, { "epoch": 0.04824371205550737, "grad_norm": 0.5066566467285156, "learning_rate": 9.874873059625707e-05, "loss": 0.0828, "step": 445 }, { "epoch": 0.04835212489158716, "grad_norm": 0.5992841720581055, "learning_rate": 9.874510372842014e-05, "loss": 0.1034, "step": 446 }, { "epoch": 0.04846053772766695, "grad_norm": 0.3061962127685547, "learning_rate": 9.87414768605832e-05, "loss": 0.062, "step": 447 }, { "epoch": 0.04856895056374675, "grad_norm": 0.49855178594589233, "learning_rate": 9.873784999274627e-05, "loss": 0.0858, "step": 448 }, { "epoch": 0.04867736339982654, "grad_norm": 0.644451916217804, "learning_rate": 9.873422312490933e-05, "loss": 0.1142, "step": 449 }, { "epoch": 0.04878577623590633, "grad_norm": 1.8879961967468262, "learning_rate": 9.873059625707239e-05, "loss": 0.0488, "step": 450 }, { "epoch": 0.04889418907198612, "grad_norm": 0.4717314839363098, "learning_rate": 9.872696938923546e-05, "loss": 0.0618, "step": 451 }, { "epoch": 0.049002601908065914, "grad_norm": 0.7027629613876343, "learning_rate": 9.872334252139852e-05, "loss": 0.106, "step": 452 }, { "epoch": 0.049111014744145705, "grad_norm": 1.407902479171753, "learning_rate": 9.871971565356159e-05, "loss": 0.1219, "step": 453 }, { "epoch": 0.049219427580225496, "grad_norm": 0.9413109421730042, "learning_rate": 9.871608878572464e-05, "loss": 0.087, "step": 454 }, { "epoch": 0.049327840416305294, "grad_norm": 0.8595862984657288, "learning_rate": 9.871246191788773e-05, "loss": 0.0882, "step": 455 }, { "epoch": 0.049436253252385085, "grad_norm": 0.7531812787055969, "learning_rate": 9.870883505005079e-05, "loss": 0.0627, "step": 456 }, { "epoch": 0.049544666088464875, "grad_norm": 1.0773489475250244, "learning_rate": 9.870520818221384e-05, "loss": 0.1009, "step": 457 }, { "epoch": 0.049653078924544666, "grad_norm": 0.9800229072570801, "learning_rate": 9.870158131437691e-05, "loss": 0.0995, "step": 458 }, { "epoch": 0.04976149176062446, "grad_norm": 0.5466852188110352, "learning_rate": 9.869795444653997e-05, "loss": 0.0917, "step": 459 }, { "epoch": 0.04986990459670425, "grad_norm": 0.7765229940414429, "learning_rate": 9.869432757870304e-05, "loss": 0.1152, "step": 460 }, { "epoch": 0.04997831743278404, "grad_norm": 0.6875452399253845, "learning_rate": 9.86907007108661e-05, "loss": 0.1128, "step": 461 }, { "epoch": 0.05008673026886384, "grad_norm": 0.6983433365821838, "learning_rate": 9.868707384302917e-05, "loss": 0.064, "step": 462 }, { "epoch": 0.05019514310494363, "grad_norm": 0.6512218713760376, "learning_rate": 9.868344697519223e-05, "loss": 0.0583, "step": 463 }, { "epoch": 0.05030355594102342, "grad_norm": 0.901423990726471, "learning_rate": 9.86798201073553e-05, "loss": 0.0921, "step": 464 }, { "epoch": 0.05041196877710321, "grad_norm": 0.9559782147407532, "learning_rate": 9.867619323951836e-05, "loss": 0.1313, "step": 465 }, { "epoch": 0.050520381613183, "grad_norm": 1.508342981338501, "learning_rate": 9.867256637168141e-05, "loss": 0.0996, "step": 466 }, { "epoch": 0.05062879444926279, "grad_norm": 0.2931099534034729, "learning_rate": 9.866893950384448e-05, "loss": 0.0568, "step": 467 }, { "epoch": 0.05073720728534258, "grad_norm": 0.5752550363540649, "learning_rate": 9.866531263600754e-05, "loss": 0.0834, "step": 468 }, { "epoch": 0.05084562012142237, "grad_norm": 0.4710371494293213, "learning_rate": 9.866168576817061e-05, "loss": 0.0819, "step": 469 }, { "epoch": 0.05095403295750217, "grad_norm": 0.5449542999267578, "learning_rate": 9.865805890033367e-05, "loss": 0.1233, "step": 470 }, { "epoch": 0.05106244579358196, "grad_norm": 0.4189887046813965, "learning_rate": 9.865443203249674e-05, "loss": 0.0652, "step": 471 }, { "epoch": 0.05117085862966175, "grad_norm": 0.738837480545044, "learning_rate": 9.865080516465981e-05, "loss": 0.0819, "step": 472 }, { "epoch": 0.051279271465741544, "grad_norm": 0.4283820390701294, "learning_rate": 9.864717829682287e-05, "loss": 0.0911, "step": 473 }, { "epoch": 0.051387684301821335, "grad_norm": 0.5088993310928345, "learning_rate": 9.864355142898594e-05, "loss": 0.0891, "step": 474 }, { "epoch": 0.051496097137901126, "grad_norm": 0.5188904404640198, "learning_rate": 9.8639924561149e-05, "loss": 0.1123, "step": 475 }, { "epoch": 0.05160450997398092, "grad_norm": 0.8196998238563538, "learning_rate": 9.863629769331207e-05, "loss": 0.1376, "step": 476 }, { "epoch": 0.051712922810060714, "grad_norm": 0.5134260058403015, "learning_rate": 9.863267082547512e-05, "loss": 0.1136, "step": 477 }, { "epoch": 0.051821335646140505, "grad_norm": 0.29203909635543823, "learning_rate": 9.86290439576382e-05, "loss": 0.0709, "step": 478 }, { "epoch": 0.051929748482220296, "grad_norm": 0.35212627053260803, "learning_rate": 9.862541708980125e-05, "loss": 0.0556, "step": 479 }, { "epoch": 0.05203816131830009, "grad_norm": 2.8972604274749756, "learning_rate": 9.862179022196431e-05, "loss": 0.1139, "step": 480 }, { "epoch": 0.05214657415437988, "grad_norm": 0.44803881645202637, "learning_rate": 9.861816335412738e-05, "loss": 0.0914, "step": 481 }, { "epoch": 0.05225498699045967, "grad_norm": 0.4926321506500244, "learning_rate": 9.861453648629044e-05, "loss": 0.0864, "step": 482 }, { "epoch": 0.05236339982653946, "grad_norm": 0.2670271396636963, "learning_rate": 9.861090961845351e-05, "loss": 0.093, "step": 483 }, { "epoch": 0.05247181266261925, "grad_norm": 0.5771321058273315, "learning_rate": 9.860728275061657e-05, "loss": 0.1179, "step": 484 }, { "epoch": 0.05258022549869905, "grad_norm": 0.382963627576828, "learning_rate": 9.860365588277964e-05, "loss": 0.0995, "step": 485 }, { "epoch": 0.05268863833477884, "grad_norm": 0.2559579014778137, "learning_rate": 9.86000290149427e-05, "loss": 0.0789, "step": 486 }, { "epoch": 0.05279705117085863, "grad_norm": 0.3600384593009949, "learning_rate": 9.859640214710577e-05, "loss": 0.0888, "step": 487 }, { "epoch": 0.05290546400693842, "grad_norm": 0.341256707906723, "learning_rate": 9.859277527926882e-05, "loss": 0.0818, "step": 488 }, { "epoch": 0.05301387684301821, "grad_norm": 0.3630164563655853, "learning_rate": 9.85891484114319e-05, "loss": 0.0703, "step": 489 }, { "epoch": 0.053122289679098, "grad_norm": 0.613624095916748, "learning_rate": 9.858552154359497e-05, "loss": 0.087, "step": 490 }, { "epoch": 0.053230702515177794, "grad_norm": 0.47455093264579773, "learning_rate": 9.858189467575802e-05, "loss": 0.0467, "step": 491 }, { "epoch": 0.05333911535125759, "grad_norm": 0.5173048973083496, "learning_rate": 9.85782678079211e-05, "loss": 0.0683, "step": 492 }, { "epoch": 0.05344752818733738, "grad_norm": 0.9313958883285522, "learning_rate": 9.857464094008415e-05, "loss": 0.1153, "step": 493 }, { "epoch": 0.053555941023417174, "grad_norm": 0.5522825717926025, "learning_rate": 9.857101407224721e-05, "loss": 0.0937, "step": 494 }, { "epoch": 0.053664353859496965, "grad_norm": 0.6051600575447083, "learning_rate": 9.856738720441028e-05, "loss": 0.0949, "step": 495 }, { "epoch": 0.053772766695576756, "grad_norm": 0.7443764209747314, "learning_rate": 9.856376033657334e-05, "loss": 0.0976, "step": 496 }, { "epoch": 0.053881179531656546, "grad_norm": 0.6347344517707825, "learning_rate": 9.856013346873641e-05, "loss": 0.093, "step": 497 }, { "epoch": 0.05398959236773634, "grad_norm": 0.2586113512516022, "learning_rate": 9.855650660089946e-05, "loss": 0.0655, "step": 498 }, { "epoch": 0.054098005203816135, "grad_norm": 0.3769162595272064, "learning_rate": 9.855287973306254e-05, "loss": 0.0603, "step": 499 }, { "epoch": 0.054206418039895926, "grad_norm": 0.7792385220527649, "learning_rate": 9.854925286522559e-05, "loss": 0.1087, "step": 500 }, { "epoch": 0.05431483087597572, "grad_norm": 0.5074377655982971, "learning_rate": 9.854562599738866e-05, "loss": 0.0732, "step": 501 }, { "epoch": 0.05442324371205551, "grad_norm": 0.34421876072883606, "learning_rate": 9.854199912955172e-05, "loss": 0.0594, "step": 502 }, { "epoch": 0.0545316565481353, "grad_norm": 0.856195330619812, "learning_rate": 9.853837226171478e-05, "loss": 0.0935, "step": 503 }, { "epoch": 0.05464006938421509, "grad_norm": 0.8137935400009155, "learning_rate": 9.853474539387785e-05, "loss": 0.1045, "step": 504 }, { "epoch": 0.05474848222029488, "grad_norm": 0.9750550985336304, "learning_rate": 9.85311185260409e-05, "loss": 0.1014, "step": 505 }, { "epoch": 0.05485689505637467, "grad_norm": 0.529441237449646, "learning_rate": 9.852749165820399e-05, "loss": 0.1089, "step": 506 }, { "epoch": 0.05496530789245447, "grad_norm": 0.6982451677322388, "learning_rate": 9.852386479036705e-05, "loss": 0.0991, "step": 507 }, { "epoch": 0.05507372072853426, "grad_norm": 0.7897246479988098, "learning_rate": 9.85202379225301e-05, "loss": 0.0504, "step": 508 }, { "epoch": 0.05518213356461405, "grad_norm": 0.5432852506637573, "learning_rate": 9.851661105469318e-05, "loss": 0.0954, "step": 509 }, { "epoch": 0.05529054640069384, "grad_norm": 0.6229777932167053, "learning_rate": 9.851298418685623e-05, "loss": 0.1081, "step": 510 }, { "epoch": 0.05539895923677363, "grad_norm": 0.31619054079055786, "learning_rate": 9.85093573190193e-05, "loss": 0.0827, "step": 511 }, { "epoch": 0.055507372072853424, "grad_norm": 0.3880375325679779, "learning_rate": 9.850573045118236e-05, "loss": 0.0495, "step": 512 }, { "epoch": 0.055615784908933215, "grad_norm": 0.4559656083583832, "learning_rate": 9.850210358334543e-05, "loss": 0.0767, "step": 513 }, { "epoch": 0.05572419774501301, "grad_norm": 0.6036137938499451, "learning_rate": 9.849847671550849e-05, "loss": 0.0923, "step": 514 }, { "epoch": 0.055832610581092804, "grad_norm": 0.4437588155269623, "learning_rate": 9.849484984767156e-05, "loss": 0.0658, "step": 515 }, { "epoch": 0.055941023417172595, "grad_norm": 0.4130304455757141, "learning_rate": 9.849122297983462e-05, "loss": 0.0616, "step": 516 }, { "epoch": 0.056049436253252385, "grad_norm": 0.7252734303474426, "learning_rate": 9.848759611199768e-05, "loss": 0.0809, "step": 517 }, { "epoch": 0.056157849089332176, "grad_norm": 0.5020862221717834, "learning_rate": 9.848396924416075e-05, "loss": 0.0712, "step": 518 }, { "epoch": 0.05626626192541197, "grad_norm": 0.6643249988555908, "learning_rate": 9.84803423763238e-05, "loss": 0.0844, "step": 519 }, { "epoch": 0.05637467476149176, "grad_norm": 1.3853152990341187, "learning_rate": 9.847671550848687e-05, "loss": 0.0809, "step": 520 }, { "epoch": 0.05648308759757155, "grad_norm": 0.7065362930297852, "learning_rate": 9.847308864064993e-05, "loss": 0.0826, "step": 521 }, { "epoch": 0.05659150043365135, "grad_norm": 0.4625021815299988, "learning_rate": 9.8469461772813e-05, "loss": 0.0721, "step": 522 }, { "epoch": 0.05669991326973114, "grad_norm": 0.8727527856826782, "learning_rate": 9.846583490497606e-05, "loss": 0.0549, "step": 523 }, { "epoch": 0.05680832610581093, "grad_norm": 0.7443035244941711, "learning_rate": 9.846220803713913e-05, "loss": 0.0873, "step": 524 }, { "epoch": 0.05691673894189072, "grad_norm": 0.6058226823806763, "learning_rate": 9.84585811693022e-05, "loss": 0.0901, "step": 525 }, { "epoch": 0.05702515177797051, "grad_norm": 0.47636687755584717, "learning_rate": 9.845495430146526e-05, "loss": 0.074, "step": 526 }, { "epoch": 0.0571335646140503, "grad_norm": 1.9641203880310059, "learning_rate": 9.845132743362833e-05, "loss": 0.0616, "step": 527 }, { "epoch": 0.05724197745013009, "grad_norm": 0.40451177954673767, "learning_rate": 9.844770056579139e-05, "loss": 0.0389, "step": 528 }, { "epoch": 0.05735039028620989, "grad_norm": 0.30972975492477417, "learning_rate": 9.844407369795446e-05, "loss": 0.0464, "step": 529 }, { "epoch": 0.05745880312228968, "grad_norm": 0.9994449019432068, "learning_rate": 9.844044683011752e-05, "loss": 0.065, "step": 530 }, { "epoch": 0.05756721595836947, "grad_norm": 1.3773900270462036, "learning_rate": 9.843681996228057e-05, "loss": 0.0932, "step": 531 }, { "epoch": 0.05767562879444926, "grad_norm": 1.254870057106018, "learning_rate": 9.843319309444364e-05, "loss": 0.0742, "step": 532 }, { "epoch": 0.057784041630529054, "grad_norm": 0.5909129977226257, "learning_rate": 9.84295662266067e-05, "loss": 0.0472, "step": 533 }, { "epoch": 0.057892454466608845, "grad_norm": 1.4020891189575195, "learning_rate": 9.842593935876977e-05, "loss": 0.1184, "step": 534 }, { "epoch": 0.058000867302688636, "grad_norm": 0.8392055034637451, "learning_rate": 9.842231249093283e-05, "loss": 0.1038, "step": 535 }, { "epoch": 0.058109280138768434, "grad_norm": 1.461511492729187, "learning_rate": 9.84186856230959e-05, "loss": 0.1511, "step": 536 }, { "epoch": 0.058217692974848224, "grad_norm": 0.506489634513855, "learning_rate": 9.841505875525896e-05, "loss": 0.0737, "step": 537 }, { "epoch": 0.058326105810928015, "grad_norm": 0.7033230662345886, "learning_rate": 9.841143188742203e-05, "loss": 0.1179, "step": 538 }, { "epoch": 0.058434518647007806, "grad_norm": 2.3751018047332764, "learning_rate": 9.840780501958509e-05, "loss": 0.1147, "step": 539 }, { "epoch": 0.0585429314830876, "grad_norm": 0.5978088974952698, "learning_rate": 9.840417815174814e-05, "loss": 0.0895, "step": 540 }, { "epoch": 0.05865134431916739, "grad_norm": 0.5447893738746643, "learning_rate": 9.840055128391123e-05, "loss": 0.0787, "step": 541 }, { "epoch": 0.05875975715524718, "grad_norm": 0.6191588640213013, "learning_rate": 9.839692441607429e-05, "loss": 0.0899, "step": 542 }, { "epoch": 0.05886816999132697, "grad_norm": 0.36324232816696167, "learning_rate": 9.839329754823736e-05, "loss": 0.0646, "step": 543 }, { "epoch": 0.05897658282740677, "grad_norm": 0.37839475274086, "learning_rate": 9.838967068040041e-05, "loss": 0.0655, "step": 544 }, { "epoch": 0.05908499566348656, "grad_norm": 0.5093445777893066, "learning_rate": 9.838604381256348e-05, "loss": 0.109, "step": 545 }, { "epoch": 0.05919340849956635, "grad_norm": 0.4167039692401886, "learning_rate": 9.838241694472654e-05, "loss": 0.09, "step": 546 }, { "epoch": 0.05930182133564614, "grad_norm": 0.5804490447044373, "learning_rate": 9.83787900768896e-05, "loss": 0.0845, "step": 547 }, { "epoch": 0.05941023417172593, "grad_norm": 0.27093636989593506, "learning_rate": 9.837516320905267e-05, "loss": 0.0863, "step": 548 }, { "epoch": 0.05951864700780572, "grad_norm": 0.46436411142349243, "learning_rate": 9.837153634121573e-05, "loss": 0.0703, "step": 549 }, { "epoch": 0.05962705984388551, "grad_norm": 0.4904249310493469, "learning_rate": 9.83679094733788e-05, "loss": 0.0472, "step": 550 }, { "epoch": 0.05973547267996531, "grad_norm": 0.3777318298816681, "learning_rate": 9.836428260554186e-05, "loss": 0.0933, "step": 551 }, { "epoch": 0.0598438855160451, "grad_norm": 0.4155241847038269, "learning_rate": 9.836065573770493e-05, "loss": 0.0655, "step": 552 }, { "epoch": 0.05995229835212489, "grad_norm": 0.29939472675323486, "learning_rate": 9.835702886986798e-05, "loss": 0.0677, "step": 553 }, { "epoch": 0.060060711188204684, "grad_norm": 0.47613203525543213, "learning_rate": 9.835340200203105e-05, "loss": 0.0936, "step": 554 }, { "epoch": 0.060169124024284475, "grad_norm": 0.5967857241630554, "learning_rate": 9.834977513419411e-05, "loss": 0.0618, "step": 555 }, { "epoch": 0.060277536860364266, "grad_norm": 0.36915236711502075, "learning_rate": 9.834614826635717e-05, "loss": 0.0357, "step": 556 }, { "epoch": 0.060385949696444056, "grad_norm": 0.6596335172653198, "learning_rate": 9.834252139852024e-05, "loss": 0.1122, "step": 557 }, { "epoch": 0.06049436253252385, "grad_norm": 0.444679856300354, "learning_rate": 9.833889453068331e-05, "loss": 0.0821, "step": 558 }, { "epoch": 0.060602775368603645, "grad_norm": 0.6635786890983582, "learning_rate": 9.833526766284638e-05, "loss": 0.0713, "step": 559 }, { "epoch": 0.060711188204683436, "grad_norm": 0.4808369278907776, "learning_rate": 9.833164079500944e-05, "loss": 0.059, "step": 560 }, { "epoch": 0.06081960104076323, "grad_norm": 0.39627790451049805, "learning_rate": 9.83280139271725e-05, "loss": 0.0528, "step": 561 }, { "epoch": 0.06092801387684302, "grad_norm": 0.9373733997344971, "learning_rate": 9.832438705933557e-05, "loss": 0.0706, "step": 562 }, { "epoch": 0.06103642671292281, "grad_norm": 0.5507509708404541, "learning_rate": 9.832076019149862e-05, "loss": 0.084, "step": 563 }, { "epoch": 0.0611448395490026, "grad_norm": 0.5078014731407166, "learning_rate": 9.83171333236617e-05, "loss": 0.0929, "step": 564 }, { "epoch": 0.06125325238508239, "grad_norm": 0.9355130195617676, "learning_rate": 9.831350645582475e-05, "loss": 0.0606, "step": 565 }, { "epoch": 0.06136166522116219, "grad_norm": 0.4562567174434662, "learning_rate": 9.830987958798782e-05, "loss": 0.0449, "step": 566 }, { "epoch": 0.06147007805724198, "grad_norm": 0.511061429977417, "learning_rate": 9.830625272015088e-05, "loss": 0.063, "step": 567 }, { "epoch": 0.06157849089332177, "grad_norm": 0.74825119972229, "learning_rate": 9.830262585231395e-05, "loss": 0.0512, "step": 568 }, { "epoch": 0.06168690372940156, "grad_norm": 0.6711365580558777, "learning_rate": 9.829899898447701e-05, "loss": 0.0868, "step": 569 }, { "epoch": 0.06179531656548135, "grad_norm": 1.0652861595153809, "learning_rate": 9.829537211664007e-05, "loss": 0.078, "step": 570 }, { "epoch": 0.06190372940156114, "grad_norm": 0.315992534160614, "learning_rate": 9.829174524880314e-05, "loss": 0.05, "step": 571 }, { "epoch": 0.062012142237640934, "grad_norm": 0.6739448308944702, "learning_rate": 9.82881183809662e-05, "loss": 0.0763, "step": 572 }, { "epoch": 0.06212055507372073, "grad_norm": 0.8512603640556335, "learning_rate": 9.828449151312927e-05, "loss": 0.0756, "step": 573 }, { "epoch": 0.06222896790980052, "grad_norm": 0.4056899845600128, "learning_rate": 9.828086464529232e-05, "loss": 0.0577, "step": 574 }, { "epoch": 0.062337380745880314, "grad_norm": 0.7571991682052612, "learning_rate": 9.82772377774554e-05, "loss": 0.0656, "step": 575 }, { "epoch": 0.062445793581960105, "grad_norm": 0.649224042892456, "learning_rate": 9.827361090961846e-05, "loss": 0.064, "step": 576 }, { "epoch": 0.0625542064180399, "grad_norm": 0.6711054444313049, "learning_rate": 9.826998404178152e-05, "loss": 0.107, "step": 577 }, { "epoch": 0.06266261925411969, "grad_norm": 1.0159608125686646, "learning_rate": 9.826635717394459e-05, "loss": 0.1495, "step": 578 }, { "epoch": 0.06277103209019948, "grad_norm": 0.39365890622138977, "learning_rate": 9.826273030610765e-05, "loss": 0.0382, "step": 579 }, { "epoch": 0.06287944492627927, "grad_norm": 0.6661434769630432, "learning_rate": 9.825910343827072e-05, "loss": 0.0833, "step": 580 }, { "epoch": 0.06298785776235906, "grad_norm": 0.2923911511898041, "learning_rate": 9.825547657043378e-05, "loss": 0.0507, "step": 581 }, { "epoch": 0.06309627059843885, "grad_norm": 0.33691978454589844, "learning_rate": 9.825184970259685e-05, "loss": 0.0431, "step": 582 }, { "epoch": 0.06320468343451864, "grad_norm": 0.4326940178871155, "learning_rate": 9.82482228347599e-05, "loss": 0.0847, "step": 583 }, { "epoch": 0.06331309627059845, "grad_norm": 0.550110399723053, "learning_rate": 9.824459596692296e-05, "loss": 0.0646, "step": 584 }, { "epoch": 0.06342150910667824, "grad_norm": 0.7317106127738953, "learning_rate": 9.824096909908603e-05, "loss": 0.115, "step": 585 }, { "epoch": 0.06352992194275803, "grad_norm": 0.5634461045265198, "learning_rate": 9.823734223124909e-05, "loss": 0.0605, "step": 586 }, { "epoch": 0.06363833477883782, "grad_norm": 0.4618016481399536, "learning_rate": 9.823371536341216e-05, "loss": 0.0832, "step": 587 }, { "epoch": 0.06374674761491761, "grad_norm": 0.38754329085350037, "learning_rate": 9.823008849557522e-05, "loss": 0.0375, "step": 588 }, { "epoch": 0.0638551604509974, "grad_norm": 0.5410308837890625, "learning_rate": 9.822646162773829e-05, "loss": 0.0667, "step": 589 }, { "epoch": 0.06396357328707719, "grad_norm": 0.33950918912887573, "learning_rate": 9.822283475990135e-05, "loss": 0.0785, "step": 590 }, { "epoch": 0.06407198612315698, "grad_norm": 0.8712959289550781, "learning_rate": 9.821920789206442e-05, "loss": 0.1581, "step": 591 }, { "epoch": 0.06418039895923677, "grad_norm": 0.5965842604637146, "learning_rate": 9.821558102422748e-05, "loss": 0.0668, "step": 592 }, { "epoch": 0.06428881179531656, "grad_norm": 0.8079363703727722, "learning_rate": 9.821195415639055e-05, "loss": 0.0666, "step": 593 }, { "epoch": 0.06439722463139635, "grad_norm": 0.4757165014743805, "learning_rate": 9.820832728855362e-05, "loss": 0.0928, "step": 594 }, { "epoch": 0.06450563746747615, "grad_norm": 0.9121995568275452, "learning_rate": 9.820470042071668e-05, "loss": 0.1125, "step": 595 }, { "epoch": 0.06461405030355594, "grad_norm": 0.6448156833648682, "learning_rate": 9.820107355287975e-05, "loss": 0.0904, "step": 596 }, { "epoch": 0.06472246313963573, "grad_norm": 0.36286067962646484, "learning_rate": 9.81974466850428e-05, "loss": 0.0836, "step": 597 }, { "epoch": 0.06483087597571552, "grad_norm": 0.6097002625465393, "learning_rate": 9.819381981720586e-05, "loss": 0.0893, "step": 598 }, { "epoch": 0.06493928881179532, "grad_norm": 0.7048788666725159, "learning_rate": 9.819019294936893e-05, "loss": 0.088, "step": 599 }, { "epoch": 0.06504770164787511, "grad_norm": 0.5463560223579407, "learning_rate": 9.818656608153199e-05, "loss": 0.0783, "step": 600 }, { "epoch": 0.0651561144839549, "grad_norm": 0.791601300239563, "learning_rate": 9.818293921369506e-05, "loss": 0.105, "step": 601 }, { "epoch": 0.0652645273200347, "grad_norm": 0.4508003294467926, "learning_rate": 9.817931234585812e-05, "loss": 0.0907, "step": 602 }, { "epoch": 0.06537294015611449, "grad_norm": 0.474373459815979, "learning_rate": 9.817568547802119e-05, "loss": 0.1042, "step": 603 }, { "epoch": 0.06548135299219428, "grad_norm": 0.5300912857055664, "learning_rate": 9.817205861018425e-05, "loss": 0.0787, "step": 604 }, { "epoch": 0.06558976582827407, "grad_norm": 0.4836699962615967, "learning_rate": 9.816843174234732e-05, "loss": 0.0748, "step": 605 }, { "epoch": 0.06569817866435386, "grad_norm": 0.2948440909385681, "learning_rate": 9.816480487451037e-05, "loss": 0.0357, "step": 606 }, { "epoch": 0.06580659150043365, "grad_norm": 0.5303706526756287, "learning_rate": 9.816117800667343e-05, "loss": 0.0589, "step": 607 }, { "epoch": 0.06591500433651344, "grad_norm": 0.3880583643913269, "learning_rate": 9.81575511388365e-05, "loss": 0.103, "step": 608 }, { "epoch": 0.06602341717259323, "grad_norm": 0.5458276271820068, "learning_rate": 9.815392427099956e-05, "loss": 0.0959, "step": 609 }, { "epoch": 0.06613183000867302, "grad_norm": 0.5764269828796387, "learning_rate": 9.815029740316264e-05, "loss": 0.0846, "step": 610 }, { "epoch": 0.06624024284475281, "grad_norm": 0.5074121356010437, "learning_rate": 9.81466705353257e-05, "loss": 0.0904, "step": 611 }, { "epoch": 0.0663486556808326, "grad_norm": 0.7112545371055603, "learning_rate": 9.814304366748876e-05, "loss": 0.0938, "step": 612 }, { "epoch": 0.0664570685169124, "grad_norm": 0.23493683338165283, "learning_rate": 9.813941679965183e-05, "loss": 0.0294, "step": 613 }, { "epoch": 0.0665654813529922, "grad_norm": 0.3845883011817932, "learning_rate": 9.813578993181489e-05, "loss": 0.0664, "step": 614 }, { "epoch": 0.06667389418907199, "grad_norm": 0.767769992351532, "learning_rate": 9.813216306397796e-05, "loss": 0.0985, "step": 615 }, { "epoch": 0.06678230702515178, "grad_norm": 0.3877485394477844, "learning_rate": 9.812853619614102e-05, "loss": 0.0709, "step": 616 }, { "epoch": 0.06689071986123157, "grad_norm": 0.35518679022789, "learning_rate": 9.812490932830409e-05, "loss": 0.0602, "step": 617 }, { "epoch": 0.06699913269731136, "grad_norm": 0.5975604057312012, "learning_rate": 9.812128246046714e-05, "loss": 0.083, "step": 618 }, { "epoch": 0.06710754553339116, "grad_norm": 0.4724676311016083, "learning_rate": 9.811765559263021e-05, "loss": 0.0337, "step": 619 }, { "epoch": 0.06721595836947095, "grad_norm": 0.8134536743164062, "learning_rate": 9.811402872479327e-05, "loss": 0.0831, "step": 620 }, { "epoch": 0.06732437120555074, "grad_norm": 0.42264291644096375, "learning_rate": 9.811040185695633e-05, "loss": 0.0365, "step": 621 }, { "epoch": 0.06743278404163053, "grad_norm": 0.6570221781730652, "learning_rate": 9.81067749891194e-05, "loss": 0.099, "step": 622 }, { "epoch": 0.06754119687771032, "grad_norm": 1.6128069162368774, "learning_rate": 9.810314812128246e-05, "loss": 0.0958, "step": 623 }, { "epoch": 0.06764960971379011, "grad_norm": 0.49546995759010315, "learning_rate": 9.809952125344553e-05, "loss": 0.0611, "step": 624 }, { "epoch": 0.0677580225498699, "grad_norm": 0.3756049871444702, "learning_rate": 9.809589438560859e-05, "loss": 0.0475, "step": 625 }, { "epoch": 0.06786643538594969, "grad_norm": 0.6570595502853394, "learning_rate": 9.809226751777166e-05, "loss": 0.0519, "step": 626 }, { "epoch": 0.06797484822202948, "grad_norm": 0.40993157029151917, "learning_rate": 9.808864064993473e-05, "loss": 0.0746, "step": 627 }, { "epoch": 0.06808326105810927, "grad_norm": 0.7181292772293091, "learning_rate": 9.808501378209778e-05, "loss": 0.1222, "step": 628 }, { "epoch": 0.06819167389418908, "grad_norm": 0.514165997505188, "learning_rate": 9.808138691426086e-05, "loss": 0.0637, "step": 629 }, { "epoch": 0.06830008673026887, "grad_norm": 1.0207935571670532, "learning_rate": 9.807776004642391e-05, "loss": 0.0767, "step": 630 }, { "epoch": 0.06840849956634866, "grad_norm": 0.7278145551681519, "learning_rate": 9.807413317858698e-05, "loss": 0.0833, "step": 631 }, { "epoch": 0.06851691240242845, "grad_norm": 0.8430814743041992, "learning_rate": 9.807050631075004e-05, "loss": 0.101, "step": 632 }, { "epoch": 0.06862532523850824, "grad_norm": 0.560275673866272, "learning_rate": 9.806687944291311e-05, "loss": 0.068, "step": 633 }, { "epoch": 0.06873373807458803, "grad_norm": 0.4498143196105957, "learning_rate": 9.806325257507617e-05, "loss": 0.0621, "step": 634 }, { "epoch": 0.06884215091066782, "grad_norm": 0.8359224796295166, "learning_rate": 9.805962570723924e-05, "loss": 0.0944, "step": 635 }, { "epoch": 0.06895056374674761, "grad_norm": 0.38813966512680054, "learning_rate": 9.80559988394023e-05, "loss": 0.0641, "step": 636 }, { "epoch": 0.0690589765828274, "grad_norm": 0.23207935690879822, "learning_rate": 9.805237197156535e-05, "loss": 0.0419, "step": 637 }, { "epoch": 0.0691673894189072, "grad_norm": 0.46650636196136475, "learning_rate": 9.804874510372843e-05, "loss": 0.0864, "step": 638 }, { "epoch": 0.06927580225498699, "grad_norm": 1.4075285196304321, "learning_rate": 9.804511823589148e-05, "loss": 0.0863, "step": 639 }, { "epoch": 0.06938421509106678, "grad_norm": 0.8136749863624573, "learning_rate": 9.804149136805455e-05, "loss": 0.0643, "step": 640 }, { "epoch": 0.06949262792714657, "grad_norm": 0.5936011075973511, "learning_rate": 9.803786450021761e-05, "loss": 0.0641, "step": 641 }, { "epoch": 0.06960104076322636, "grad_norm": 0.44310900568962097, "learning_rate": 9.803423763238068e-05, "loss": 0.0734, "step": 642 }, { "epoch": 0.06970945359930615, "grad_norm": 0.4348272383213043, "learning_rate": 9.803061076454374e-05, "loss": 0.0509, "step": 643 }, { "epoch": 0.06981786643538596, "grad_norm": 0.33767902851104736, "learning_rate": 9.80269838967068e-05, "loss": 0.0442, "step": 644 }, { "epoch": 0.06992627927146575, "grad_norm": 0.46737632155418396, "learning_rate": 9.802335702886988e-05, "loss": 0.0985, "step": 645 }, { "epoch": 0.07003469210754554, "grad_norm": 0.8100427985191345, "learning_rate": 9.801973016103294e-05, "loss": 0.0825, "step": 646 }, { "epoch": 0.07014310494362533, "grad_norm": 1.120666742324829, "learning_rate": 9.801610329319601e-05, "loss": 0.1222, "step": 647 }, { "epoch": 0.07025151777970512, "grad_norm": 0.556582510471344, "learning_rate": 9.801247642535907e-05, "loss": 0.0818, "step": 648 }, { "epoch": 0.07035993061578491, "grad_norm": 0.7664306163787842, "learning_rate": 9.800884955752214e-05, "loss": 0.0605, "step": 649 }, { "epoch": 0.0704683434518647, "grad_norm": 0.6349785327911377, "learning_rate": 9.80052226896852e-05, "loss": 0.0613, "step": 650 }, { "epoch": 0.07057675628794449, "grad_norm": 0.7575204372406006, "learning_rate": 9.800159582184825e-05, "loss": 0.0713, "step": 651 }, { "epoch": 0.07068516912402428, "grad_norm": 0.48733457922935486, "learning_rate": 9.799796895401132e-05, "loss": 0.0883, "step": 652 }, { "epoch": 0.07079358196010407, "grad_norm": 0.3797195255756378, "learning_rate": 9.799434208617438e-05, "loss": 0.0582, "step": 653 }, { "epoch": 0.07090199479618386, "grad_norm": 0.5248768329620361, "learning_rate": 9.799071521833745e-05, "loss": 0.0899, "step": 654 }, { "epoch": 0.07101040763226366, "grad_norm": 0.4632895588874817, "learning_rate": 9.798708835050051e-05, "loss": 0.0786, "step": 655 }, { "epoch": 0.07111882046834345, "grad_norm": 0.5752514004707336, "learning_rate": 9.798346148266358e-05, "loss": 0.0668, "step": 656 }, { "epoch": 0.07122723330442324, "grad_norm": 0.3810531198978424, "learning_rate": 9.797983461482664e-05, "loss": 0.06, "step": 657 }, { "epoch": 0.07133564614050304, "grad_norm": 0.848351240158081, "learning_rate": 9.797620774698971e-05, "loss": 0.1116, "step": 658 }, { "epoch": 0.07144405897658283, "grad_norm": 0.43770384788513184, "learning_rate": 9.797258087915276e-05, "loss": 0.0662, "step": 659 }, { "epoch": 0.07155247181266262, "grad_norm": 0.3448054790496826, "learning_rate": 9.796895401131582e-05, "loss": 0.0595, "step": 660 }, { "epoch": 0.07166088464874242, "grad_norm": 0.5571114420890808, "learning_rate": 9.796532714347889e-05, "loss": 0.0557, "step": 661 }, { "epoch": 0.0717692974848222, "grad_norm": 0.8095369338989258, "learning_rate": 9.796170027564196e-05, "loss": 0.0809, "step": 662 }, { "epoch": 0.071877710320902, "grad_norm": 1.0758482217788696, "learning_rate": 9.795807340780503e-05, "loss": 0.0997, "step": 663 }, { "epoch": 0.07198612315698179, "grad_norm": 0.855961799621582, "learning_rate": 9.795444653996809e-05, "loss": 0.0995, "step": 664 }, { "epoch": 0.07209453599306158, "grad_norm": 0.7401790022850037, "learning_rate": 9.795081967213115e-05, "loss": 0.0689, "step": 665 }, { "epoch": 0.07220294882914137, "grad_norm": 1.0213289260864258, "learning_rate": 9.794719280429422e-05, "loss": 0.0889, "step": 666 }, { "epoch": 0.07231136166522116, "grad_norm": 0.3955745995044708, "learning_rate": 9.794356593645728e-05, "loss": 0.0349, "step": 667 }, { "epoch": 0.07241977450130095, "grad_norm": 0.6165972948074341, "learning_rate": 9.793993906862035e-05, "loss": 0.0714, "step": 668 }, { "epoch": 0.07252818733738074, "grad_norm": 0.4667428731918335, "learning_rate": 9.79363122007834e-05, "loss": 0.0527, "step": 669 }, { "epoch": 0.07263660017346053, "grad_norm": 0.5425849556922913, "learning_rate": 9.793268533294648e-05, "loss": 0.0804, "step": 670 }, { "epoch": 0.07274501300954032, "grad_norm": 0.7196183204650879, "learning_rate": 9.792905846510953e-05, "loss": 0.0827, "step": 671 }, { "epoch": 0.07285342584562012, "grad_norm": 0.7280611395835876, "learning_rate": 9.79254315972726e-05, "loss": 0.0699, "step": 672 }, { "epoch": 0.07296183868169992, "grad_norm": 0.5346488952636719, "learning_rate": 9.792180472943566e-05, "loss": 0.044, "step": 673 }, { "epoch": 0.07307025151777971, "grad_norm": 0.41916605830192566, "learning_rate": 9.791817786159872e-05, "loss": 0.0397, "step": 674 }, { "epoch": 0.0731786643538595, "grad_norm": 0.7157227396965027, "learning_rate": 9.791455099376179e-05, "loss": 0.0743, "step": 675 }, { "epoch": 0.07328707718993929, "grad_norm": 0.7560863494873047, "learning_rate": 9.791092412592485e-05, "loss": 0.0975, "step": 676 }, { "epoch": 0.07339549002601908, "grad_norm": 0.6880209445953369, "learning_rate": 9.790729725808792e-05, "loss": 0.0659, "step": 677 }, { "epoch": 0.07350390286209887, "grad_norm": 0.6112470030784607, "learning_rate": 9.790367039025098e-05, "loss": 0.1041, "step": 678 }, { "epoch": 0.07361231569817867, "grad_norm": 0.7357256412506104, "learning_rate": 9.790004352241405e-05, "loss": 0.1092, "step": 679 }, { "epoch": 0.07372072853425846, "grad_norm": 0.6230786442756653, "learning_rate": 9.789641665457712e-05, "loss": 0.0775, "step": 680 }, { "epoch": 0.07382914137033825, "grad_norm": 0.4532262086868286, "learning_rate": 9.789278978674018e-05, "loss": 0.047, "step": 681 }, { "epoch": 0.07393755420641804, "grad_norm": 0.6085010766983032, "learning_rate": 9.788916291890325e-05, "loss": 0.0524, "step": 682 }, { "epoch": 0.07404596704249783, "grad_norm": 0.7158128619194031, "learning_rate": 9.78855360510663e-05, "loss": 0.0675, "step": 683 }, { "epoch": 0.07415437987857762, "grad_norm": 0.44022417068481445, "learning_rate": 9.788190918322937e-05, "loss": 0.0377, "step": 684 }, { "epoch": 0.07426279271465741, "grad_norm": 0.6066200137138367, "learning_rate": 9.787828231539243e-05, "loss": 0.0803, "step": 685 }, { "epoch": 0.0743712055507372, "grad_norm": 0.42551538348197937, "learning_rate": 9.78746554475555e-05, "loss": 0.0494, "step": 686 }, { "epoch": 0.07447961838681699, "grad_norm": 0.3404230773448944, "learning_rate": 9.787102857971856e-05, "loss": 0.0488, "step": 687 }, { "epoch": 0.0745880312228968, "grad_norm": 0.5352039337158203, "learning_rate": 9.786740171188162e-05, "loss": 0.0962, "step": 688 }, { "epoch": 0.07469644405897659, "grad_norm": 0.9265494346618652, "learning_rate": 9.786377484404469e-05, "loss": 0.0915, "step": 689 }, { "epoch": 0.07480485689505638, "grad_norm": 1.0869154930114746, "learning_rate": 9.786014797620775e-05, "loss": 0.0944, "step": 690 }, { "epoch": 0.07491326973113617, "grad_norm": 0.8128892779350281, "learning_rate": 9.785652110837082e-05, "loss": 0.1038, "step": 691 }, { "epoch": 0.07502168256721596, "grad_norm": 0.7345775365829468, "learning_rate": 9.785289424053387e-05, "loss": 0.1233, "step": 692 }, { "epoch": 0.07513009540329575, "grad_norm": 0.739715576171875, "learning_rate": 9.784926737269694e-05, "loss": 0.1369, "step": 693 }, { "epoch": 0.07523850823937554, "grad_norm": 0.2000858634710312, "learning_rate": 9.784564050486e-05, "loss": 0.0207, "step": 694 }, { "epoch": 0.07534692107545533, "grad_norm": 0.4452505111694336, "learning_rate": 9.784201363702307e-05, "loss": 0.0558, "step": 695 }, { "epoch": 0.07545533391153512, "grad_norm": 0.5214964151382446, "learning_rate": 9.783838676918613e-05, "loss": 0.0728, "step": 696 }, { "epoch": 0.07556374674761492, "grad_norm": 0.4578377604484558, "learning_rate": 9.78347599013492e-05, "loss": 0.0612, "step": 697 }, { "epoch": 0.0756721595836947, "grad_norm": 0.34610775113105774, "learning_rate": 9.783113303351227e-05, "loss": 0.0865, "step": 698 }, { "epoch": 0.0757805724197745, "grad_norm": 0.35405513644218445, "learning_rate": 9.782750616567533e-05, "loss": 0.0418, "step": 699 }, { "epoch": 0.07588898525585429, "grad_norm": 0.4706442952156067, "learning_rate": 9.78238792978384e-05, "loss": 0.0858, "step": 700 }, { "epoch": 0.07599739809193408, "grad_norm": 0.8495997786521912, "learning_rate": 9.782025243000146e-05, "loss": 0.0951, "step": 701 }, { "epoch": 0.07610581092801387, "grad_norm": 0.562516987323761, "learning_rate": 9.781662556216451e-05, "loss": 0.1305, "step": 702 }, { "epoch": 0.07621422376409367, "grad_norm": 0.31649264693260193, "learning_rate": 9.781299869432759e-05, "loss": 0.0805, "step": 703 }, { "epoch": 0.07632263660017347, "grad_norm": 0.4714401364326477, "learning_rate": 9.780937182649064e-05, "loss": 0.0961, "step": 704 }, { "epoch": 0.07643104943625326, "grad_norm": 0.7049551010131836, "learning_rate": 9.780574495865371e-05, "loss": 0.1153, "step": 705 }, { "epoch": 0.07653946227233305, "grad_norm": 0.5214707851409912, "learning_rate": 9.780211809081677e-05, "loss": 0.0789, "step": 706 }, { "epoch": 0.07664787510841284, "grad_norm": 0.40077534317970276, "learning_rate": 9.779849122297984e-05, "loss": 0.0754, "step": 707 }, { "epoch": 0.07675628794449263, "grad_norm": 0.4077340066432953, "learning_rate": 9.77948643551429e-05, "loss": 0.0624, "step": 708 }, { "epoch": 0.07686470078057242, "grad_norm": 0.5211284160614014, "learning_rate": 9.779123748730597e-05, "loss": 0.1239, "step": 709 }, { "epoch": 0.07697311361665221, "grad_norm": 0.33230477571487427, "learning_rate": 9.778761061946903e-05, "loss": 0.0557, "step": 710 }, { "epoch": 0.077081526452732, "grad_norm": 0.46741440892219543, "learning_rate": 9.778398375163208e-05, "loss": 0.0674, "step": 711 }, { "epoch": 0.07718993928881179, "grad_norm": 0.33459538221359253, "learning_rate": 9.778035688379516e-05, "loss": 0.0534, "step": 712 }, { "epoch": 0.07729835212489158, "grad_norm": 0.4303707182407379, "learning_rate": 9.777673001595821e-05, "loss": 0.049, "step": 713 }, { "epoch": 0.07740676496097137, "grad_norm": 0.5940755605697632, "learning_rate": 9.77731031481213e-05, "loss": 0.0949, "step": 714 }, { "epoch": 0.07751517779705117, "grad_norm": 0.2606859803199768, "learning_rate": 9.776947628028435e-05, "loss": 0.0581, "step": 715 }, { "epoch": 0.07762359063313096, "grad_norm": 0.2850607633590698, "learning_rate": 9.776584941244741e-05, "loss": 0.0661, "step": 716 }, { "epoch": 0.07773200346921076, "grad_norm": 0.3100925385951996, "learning_rate": 9.776222254461048e-05, "loss": 0.0643, "step": 717 }, { "epoch": 0.07784041630529055, "grad_norm": 0.3656713664531708, "learning_rate": 9.775859567677354e-05, "loss": 0.057, "step": 718 }, { "epoch": 0.07794882914137034, "grad_norm": 0.39168959856033325, "learning_rate": 9.775496880893661e-05, "loss": 0.0415, "step": 719 }, { "epoch": 0.07805724197745013, "grad_norm": 0.7509340047836304, "learning_rate": 9.775134194109967e-05, "loss": 0.0965, "step": 720 }, { "epoch": 0.07816565481352993, "grad_norm": 0.4651702642440796, "learning_rate": 9.774771507326274e-05, "loss": 0.0656, "step": 721 }, { "epoch": 0.07827406764960972, "grad_norm": 0.6109954118728638, "learning_rate": 9.77440882054258e-05, "loss": 0.0651, "step": 722 }, { "epoch": 0.0783824804856895, "grad_norm": 0.4757864475250244, "learning_rate": 9.774046133758887e-05, "loss": 0.0476, "step": 723 }, { "epoch": 0.0784908933217693, "grad_norm": 0.907374382019043, "learning_rate": 9.773683446975193e-05, "loss": 0.0951, "step": 724 }, { "epoch": 0.07859930615784909, "grad_norm": 1.2199500799179077, "learning_rate": 9.773320760191498e-05, "loss": 0.1367, "step": 725 }, { "epoch": 0.07870771899392888, "grad_norm": 0.4963178038597107, "learning_rate": 9.772958073407805e-05, "loss": 0.0594, "step": 726 }, { "epoch": 0.07881613183000867, "grad_norm": 0.7674083709716797, "learning_rate": 9.772595386624111e-05, "loss": 0.0831, "step": 727 }, { "epoch": 0.07892454466608846, "grad_norm": 1.5545756816864014, "learning_rate": 9.772232699840418e-05, "loss": 0.4178, "step": 728 }, { "epoch": 0.07903295750216825, "grad_norm": 0.2678217887878418, "learning_rate": 9.771870013056724e-05, "loss": 0.0357, "step": 729 }, { "epoch": 0.07914137033824804, "grad_norm": 1.0386303663253784, "learning_rate": 9.771507326273031e-05, "loss": 0.0455, "step": 730 }, { "epoch": 0.07924978317432783, "grad_norm": 0.8436711430549622, "learning_rate": 9.771144639489338e-05, "loss": 0.0735, "step": 731 }, { "epoch": 0.07935819601040764, "grad_norm": 0.49298909306526184, "learning_rate": 9.770781952705644e-05, "loss": 0.058, "step": 732 }, { "epoch": 0.07946660884648743, "grad_norm": 0.30943506956100464, "learning_rate": 9.770419265921951e-05, "loss": 0.0298, "step": 733 }, { "epoch": 0.07957502168256722, "grad_norm": 0.48199930787086487, "learning_rate": 9.770056579138257e-05, "loss": 0.063, "step": 734 }, { "epoch": 0.07968343451864701, "grad_norm": 0.7698824405670166, "learning_rate": 9.769693892354564e-05, "loss": 0.1091, "step": 735 }, { "epoch": 0.0797918473547268, "grad_norm": 1.0148462057113647, "learning_rate": 9.76933120557087e-05, "loss": 0.0882, "step": 736 }, { "epoch": 0.0799002601908066, "grad_norm": 0.4573986232280731, "learning_rate": 9.768968518787177e-05, "loss": 0.0498, "step": 737 }, { "epoch": 0.08000867302688638, "grad_norm": 0.5934492945671082, "learning_rate": 9.768605832003482e-05, "loss": 0.0417, "step": 738 }, { "epoch": 0.08011708586296618, "grad_norm": 0.32592394948005676, "learning_rate": 9.76824314521979e-05, "loss": 0.0343, "step": 739 }, { "epoch": 0.08022549869904597, "grad_norm": 0.8502817153930664, "learning_rate": 9.767880458436095e-05, "loss": 0.0631, "step": 740 }, { "epoch": 0.08033391153512576, "grad_norm": 0.8614488840103149, "learning_rate": 9.767517771652401e-05, "loss": 0.0721, "step": 741 }, { "epoch": 0.08044232437120555, "grad_norm": 0.42982733249664307, "learning_rate": 9.767155084868708e-05, "loss": 0.0591, "step": 742 }, { "epoch": 0.08055073720728534, "grad_norm": 0.49320876598358154, "learning_rate": 9.766792398085014e-05, "loss": 0.0641, "step": 743 }, { "epoch": 0.08065915004336513, "grad_norm": 0.39727783203125, "learning_rate": 9.766429711301321e-05, "loss": 0.0576, "step": 744 }, { "epoch": 0.08076756287944492, "grad_norm": 0.7650752663612366, "learning_rate": 9.766067024517626e-05, "loss": 0.0683, "step": 745 }, { "epoch": 0.08087597571552471, "grad_norm": 0.4262329041957855, "learning_rate": 9.765704337733934e-05, "loss": 0.04, "step": 746 }, { "epoch": 0.08098438855160452, "grad_norm": 0.49667853116989136, "learning_rate": 9.765341650950239e-05, "loss": 0.0453, "step": 747 }, { "epoch": 0.08109280138768431, "grad_norm": 0.5553999543190002, "learning_rate": 9.764978964166546e-05, "loss": 0.0753, "step": 748 }, { "epoch": 0.0812012142237641, "grad_norm": 0.47543224692344666, "learning_rate": 9.764616277382853e-05, "loss": 0.0876, "step": 749 }, { "epoch": 0.08130962705984389, "grad_norm": 0.7772607207298279, "learning_rate": 9.764253590599159e-05, "loss": 0.0418, "step": 750 }, { "epoch": 0.08141803989592368, "grad_norm": 0.718052089214325, "learning_rate": 9.763890903815466e-05, "loss": 0.09, "step": 751 }, { "epoch": 0.08152645273200347, "grad_norm": 0.8017333745956421, "learning_rate": 9.763528217031772e-05, "loss": 0.0934, "step": 752 }, { "epoch": 0.08163486556808326, "grad_norm": 1.052696704864502, "learning_rate": 9.763165530248079e-05, "loss": 0.0646, "step": 753 }, { "epoch": 0.08174327840416305, "grad_norm": 0.3310949206352234, "learning_rate": 9.762802843464385e-05, "loss": 0.051, "step": 754 }, { "epoch": 0.08185169124024284, "grad_norm": 0.6501230001449585, "learning_rate": 9.76244015668069e-05, "loss": 0.0925, "step": 755 }, { "epoch": 0.08196010407632263, "grad_norm": 0.37495529651641846, "learning_rate": 9.762077469896998e-05, "loss": 0.0684, "step": 756 }, { "epoch": 0.08206851691240243, "grad_norm": 0.49605247378349304, "learning_rate": 9.761714783113303e-05, "loss": 0.0676, "step": 757 }, { "epoch": 0.08217692974848222, "grad_norm": 0.30230438709259033, "learning_rate": 9.76135209632961e-05, "loss": 0.0427, "step": 758 }, { "epoch": 0.08228534258456201, "grad_norm": 0.5996018648147583, "learning_rate": 9.760989409545916e-05, "loss": 0.0734, "step": 759 }, { "epoch": 0.0823937554206418, "grad_norm": 0.6038388013839722, "learning_rate": 9.760626722762223e-05, "loss": 0.1342, "step": 760 }, { "epoch": 0.08250216825672159, "grad_norm": 0.8810690641403198, "learning_rate": 9.760264035978529e-05, "loss": 0.0917, "step": 761 }, { "epoch": 0.0826105810928014, "grad_norm": 0.30079731345176697, "learning_rate": 9.759901349194836e-05, "loss": 0.0529, "step": 762 }, { "epoch": 0.08271899392888118, "grad_norm": 0.5820724964141846, "learning_rate": 9.759538662411142e-05, "loss": 0.1028, "step": 763 }, { "epoch": 0.08282740676496098, "grad_norm": 0.734008252620697, "learning_rate": 9.759175975627448e-05, "loss": 0.0791, "step": 764 }, { "epoch": 0.08293581960104077, "grad_norm": 0.48477989435195923, "learning_rate": 9.758813288843755e-05, "loss": 0.1128, "step": 765 }, { "epoch": 0.08304423243712056, "grad_norm": 0.37036988139152527, "learning_rate": 9.758450602060062e-05, "loss": 0.071, "step": 766 }, { "epoch": 0.08315264527320035, "grad_norm": 0.34174132347106934, "learning_rate": 9.758087915276369e-05, "loss": 0.0755, "step": 767 }, { "epoch": 0.08326105810928014, "grad_norm": 0.4900585412979126, "learning_rate": 9.757725228492675e-05, "loss": 0.0919, "step": 768 }, { "epoch": 0.08336947094535993, "grad_norm": 0.2560671865940094, "learning_rate": 9.75736254170898e-05, "loss": 0.0473, "step": 769 }, { "epoch": 0.08347788378143972, "grad_norm": 0.4855523407459259, "learning_rate": 9.756999854925287e-05, "loss": 0.0954, "step": 770 }, { "epoch": 0.08358629661751951, "grad_norm": 0.4060906767845154, "learning_rate": 9.756637168141593e-05, "loss": 0.0626, "step": 771 }, { "epoch": 0.0836947094535993, "grad_norm": 0.22328414022922516, "learning_rate": 9.7562744813579e-05, "loss": 0.0409, "step": 772 }, { "epoch": 0.0838031222896791, "grad_norm": 0.49351561069488525, "learning_rate": 9.755911794574206e-05, "loss": 0.0837, "step": 773 }, { "epoch": 0.08391153512575888, "grad_norm": 0.3897554278373718, "learning_rate": 9.755549107790513e-05, "loss": 0.0611, "step": 774 }, { "epoch": 0.08401994796183868, "grad_norm": 0.5700775980949402, "learning_rate": 9.755186421006819e-05, "loss": 0.1286, "step": 775 }, { "epoch": 0.08412836079791847, "grad_norm": 0.45939067006111145, "learning_rate": 9.754823734223126e-05, "loss": 0.1014, "step": 776 }, { "epoch": 0.08423677363399827, "grad_norm": 0.4438456594944, "learning_rate": 9.754461047439432e-05, "loss": 0.1079, "step": 777 }, { "epoch": 0.08434518647007806, "grad_norm": 0.40705356001853943, "learning_rate": 9.754098360655737e-05, "loss": 0.0377, "step": 778 }, { "epoch": 0.08445359930615785, "grad_norm": 0.33517545461654663, "learning_rate": 9.753735673872044e-05, "loss": 0.0507, "step": 779 }, { "epoch": 0.08456201214223764, "grad_norm": 1.0031466484069824, "learning_rate": 9.75337298708835e-05, "loss": 0.0667, "step": 780 }, { "epoch": 0.08467042497831744, "grad_norm": 0.4472539722919464, "learning_rate": 9.753010300304657e-05, "loss": 0.0806, "step": 781 }, { "epoch": 0.08477883781439723, "grad_norm": 0.4558166265487671, "learning_rate": 9.752647613520963e-05, "loss": 0.084, "step": 782 }, { "epoch": 0.08488725065047702, "grad_norm": 0.3395930528640747, "learning_rate": 9.75228492673727e-05, "loss": 0.0564, "step": 783 }, { "epoch": 0.08499566348655681, "grad_norm": 0.3377668857574463, "learning_rate": 9.751922239953577e-05, "loss": 0.0499, "step": 784 }, { "epoch": 0.0851040763226366, "grad_norm": 1.1375479698181152, "learning_rate": 9.751559553169883e-05, "loss": 0.0515, "step": 785 }, { "epoch": 0.08521248915871639, "grad_norm": 0.5352253317832947, "learning_rate": 9.75119686638619e-05, "loss": 0.094, "step": 786 }, { "epoch": 0.08532090199479618, "grad_norm": 0.4564967453479767, "learning_rate": 9.750834179602496e-05, "loss": 0.0842, "step": 787 }, { "epoch": 0.08542931483087597, "grad_norm": 0.4958781898021698, "learning_rate": 9.750471492818803e-05, "loss": 0.0688, "step": 788 }, { "epoch": 0.08553772766695576, "grad_norm": 0.3173529803752899, "learning_rate": 9.750108806035109e-05, "loss": 0.0485, "step": 789 }, { "epoch": 0.08564614050303555, "grad_norm": 0.5061728954315186, "learning_rate": 9.749746119251416e-05, "loss": 0.0681, "step": 790 }, { "epoch": 0.08575455333911536, "grad_norm": 0.3447498679161072, "learning_rate": 9.749383432467721e-05, "loss": 0.0436, "step": 791 }, { "epoch": 0.08586296617519515, "grad_norm": 1.8145049810409546, "learning_rate": 9.749020745684027e-05, "loss": 0.0518, "step": 792 }, { "epoch": 0.08597137901127494, "grad_norm": 0.8952356576919556, "learning_rate": 9.748658058900334e-05, "loss": 0.0477, "step": 793 }, { "epoch": 0.08607979184735473, "grad_norm": 0.3479878306388855, "learning_rate": 9.74829537211664e-05, "loss": 0.0372, "step": 794 }, { "epoch": 0.08618820468343452, "grad_norm": 0.4077695608139038, "learning_rate": 9.747932685332947e-05, "loss": 0.0192, "step": 795 }, { "epoch": 0.08629661751951431, "grad_norm": 0.5917627215385437, "learning_rate": 9.747569998549253e-05, "loss": 0.0736, "step": 796 }, { "epoch": 0.0864050303555941, "grad_norm": 0.7282764911651611, "learning_rate": 9.74720731176556e-05, "loss": 0.0637, "step": 797 }, { "epoch": 0.0865134431916739, "grad_norm": 0.8097841739654541, "learning_rate": 9.746844624981866e-05, "loss": 0.097, "step": 798 }, { "epoch": 0.08662185602775369, "grad_norm": 0.5764357447624207, "learning_rate": 9.746481938198173e-05, "loss": 0.0848, "step": 799 }, { "epoch": 0.08673026886383348, "grad_norm": 0.30036312341690063, "learning_rate": 9.74611925141448e-05, "loss": 0.0392, "step": 800 }, { "epoch": 0.08683868169991327, "grad_norm": 1.0626250505447388, "learning_rate": 9.745756564630785e-05, "loss": 0.0325, "step": 801 }, { "epoch": 0.08694709453599306, "grad_norm": 0.662631094455719, "learning_rate": 9.745393877847093e-05, "loss": 0.054, "step": 802 }, { "epoch": 0.08705550737207285, "grad_norm": 0.6435604095458984, "learning_rate": 9.745031191063398e-05, "loss": 0.0967, "step": 803 }, { "epoch": 0.08716392020815264, "grad_norm": 0.613777220249176, "learning_rate": 9.744668504279705e-05, "loss": 0.0794, "step": 804 }, { "epoch": 0.08727233304423243, "grad_norm": 0.5607749223709106, "learning_rate": 9.744305817496011e-05, "loss": 0.0531, "step": 805 }, { "epoch": 0.08738074588031224, "grad_norm": 0.7848803400993347, "learning_rate": 9.743943130712317e-05, "loss": 0.0812, "step": 806 }, { "epoch": 0.08748915871639203, "grad_norm": 0.4831368923187256, "learning_rate": 9.743580443928624e-05, "loss": 0.0596, "step": 807 }, { "epoch": 0.08759757155247182, "grad_norm": 0.4351140856742859, "learning_rate": 9.74321775714493e-05, "loss": 0.0761, "step": 808 }, { "epoch": 0.08770598438855161, "grad_norm": 0.844457745552063, "learning_rate": 9.742855070361237e-05, "loss": 0.0406, "step": 809 }, { "epoch": 0.0878143972246314, "grad_norm": 1.1111137866973877, "learning_rate": 9.742492383577542e-05, "loss": 0.0659, "step": 810 }, { "epoch": 0.08792281006071119, "grad_norm": 0.7056833505630493, "learning_rate": 9.74212969679385e-05, "loss": 0.0914, "step": 811 }, { "epoch": 0.08803122289679098, "grad_norm": 0.480078786611557, "learning_rate": 9.741767010010155e-05, "loss": 0.0853, "step": 812 }, { "epoch": 0.08813963573287077, "grad_norm": 0.6462282538414001, "learning_rate": 9.741404323226462e-05, "loss": 0.0614, "step": 813 }, { "epoch": 0.08824804856895056, "grad_norm": 0.40191173553466797, "learning_rate": 9.741041636442768e-05, "loss": 0.0742, "step": 814 }, { "epoch": 0.08835646140503035, "grad_norm": 0.502816379070282, "learning_rate": 9.740678949659074e-05, "loss": 0.0839, "step": 815 }, { "epoch": 0.08846487424111014, "grad_norm": 0.326243132352829, "learning_rate": 9.740316262875381e-05, "loss": 0.0743, "step": 816 }, { "epoch": 0.08857328707718994, "grad_norm": 0.534164309501648, "learning_rate": 9.739953576091688e-05, "loss": 0.1028, "step": 817 }, { "epoch": 0.08868169991326973, "grad_norm": 0.3953757584095001, "learning_rate": 9.739590889307995e-05, "loss": 0.0616, "step": 818 }, { "epoch": 0.08879011274934952, "grad_norm": 0.33386358618736267, "learning_rate": 9.739228202524301e-05, "loss": 0.0632, "step": 819 }, { "epoch": 0.08889852558542931, "grad_norm": 0.735598087310791, "learning_rate": 9.738865515740608e-05, "loss": 0.0903, "step": 820 }, { "epoch": 0.08900693842150911, "grad_norm": 0.3369658589363098, "learning_rate": 9.738502828956914e-05, "loss": 0.0434, "step": 821 }, { "epoch": 0.0891153512575889, "grad_norm": 0.6425350904464722, "learning_rate": 9.73814014217322e-05, "loss": 0.0924, "step": 822 }, { "epoch": 0.0892237640936687, "grad_norm": 0.39782389998435974, "learning_rate": 9.737777455389526e-05, "loss": 0.0644, "step": 823 }, { "epoch": 0.08933217692974849, "grad_norm": 0.6085790991783142, "learning_rate": 9.737414768605832e-05, "loss": 0.0552, "step": 824 }, { "epoch": 0.08944058976582828, "grad_norm": 0.34661659598350525, "learning_rate": 9.737052081822139e-05, "loss": 0.066, "step": 825 }, { "epoch": 0.08954900260190807, "grad_norm": 0.5092345476150513, "learning_rate": 9.736689395038445e-05, "loss": 0.0586, "step": 826 }, { "epoch": 0.08965741543798786, "grad_norm": 0.4188010096549988, "learning_rate": 9.736326708254752e-05, "loss": 0.0549, "step": 827 }, { "epoch": 0.08976582827406765, "grad_norm": 0.38009151816368103, "learning_rate": 9.735964021471058e-05, "loss": 0.0686, "step": 828 }, { "epoch": 0.08987424111014744, "grad_norm": 0.2897365689277649, "learning_rate": 9.735601334687364e-05, "loss": 0.0561, "step": 829 }, { "epoch": 0.08998265394622723, "grad_norm": 0.24453873932361603, "learning_rate": 9.73523864790367e-05, "loss": 0.0359, "step": 830 }, { "epoch": 0.09009106678230702, "grad_norm": 0.38017839193344116, "learning_rate": 9.734875961119976e-05, "loss": 0.067, "step": 831 }, { "epoch": 0.09019947961838681, "grad_norm": 0.5346794128417969, "learning_rate": 9.734513274336283e-05, "loss": 0.0352, "step": 832 }, { "epoch": 0.0903078924544666, "grad_norm": 0.42592763900756836, "learning_rate": 9.734150587552589e-05, "loss": 0.0296, "step": 833 }, { "epoch": 0.0904163052905464, "grad_norm": 0.9858507513999939, "learning_rate": 9.733787900768896e-05, "loss": 0.0883, "step": 834 }, { "epoch": 0.09052471812662619, "grad_norm": 0.6708551645278931, "learning_rate": 9.733425213985203e-05, "loss": 0.0577, "step": 835 }, { "epoch": 0.09063313096270599, "grad_norm": 0.319521963596344, "learning_rate": 9.733062527201509e-05, "loss": 0.0481, "step": 836 }, { "epoch": 0.09074154379878578, "grad_norm": 0.5769931674003601, "learning_rate": 9.732699840417816e-05, "loss": 0.068, "step": 837 }, { "epoch": 0.09084995663486557, "grad_norm": 0.9463897943496704, "learning_rate": 9.732337153634122e-05, "loss": 0.0668, "step": 838 }, { "epoch": 0.09095836947094536, "grad_norm": 0.6062118411064148, "learning_rate": 9.731974466850429e-05, "loss": 0.052, "step": 839 }, { "epoch": 0.09106678230702515, "grad_norm": 0.376173198223114, "learning_rate": 9.731611780066735e-05, "loss": 0.0433, "step": 840 }, { "epoch": 0.09117519514310495, "grad_norm": 1.1057008504867554, "learning_rate": 9.731249093283042e-05, "loss": 0.0597, "step": 841 }, { "epoch": 0.09128360797918474, "grad_norm": 1.006273627281189, "learning_rate": 9.730886406499348e-05, "loss": 0.0683, "step": 842 }, { "epoch": 0.09139202081526453, "grad_norm": 0.4022150933742523, "learning_rate": 9.730523719715655e-05, "loss": 0.0205, "step": 843 }, { "epoch": 0.09150043365134432, "grad_norm": 1.3433021306991577, "learning_rate": 9.73016103293196e-05, "loss": 0.062, "step": 844 }, { "epoch": 0.09160884648742411, "grad_norm": 1.1420273780822754, "learning_rate": 9.729798346148266e-05, "loss": 0.0774, "step": 845 }, { "epoch": 0.0917172593235039, "grad_norm": 0.8533089756965637, "learning_rate": 9.729435659364573e-05, "loss": 0.1202, "step": 846 }, { "epoch": 0.09182567215958369, "grad_norm": 0.47172269225120544, "learning_rate": 9.729072972580879e-05, "loss": 0.0595, "step": 847 }, { "epoch": 0.09193408499566348, "grad_norm": 0.43485456705093384, "learning_rate": 9.728710285797186e-05, "loss": 0.0819, "step": 848 }, { "epoch": 0.09204249783174327, "grad_norm": 1.1267482042312622, "learning_rate": 9.728347599013492e-05, "loss": 0.0761, "step": 849 }, { "epoch": 0.09215091066782308, "grad_norm": 0.6753718256950378, "learning_rate": 9.727984912229799e-05, "loss": 0.0566, "step": 850 }, { "epoch": 0.09225932350390287, "grad_norm": 0.27950170636177063, "learning_rate": 9.727622225446105e-05, "loss": 0.0353, "step": 851 }, { "epoch": 0.09236773633998266, "grad_norm": 0.6082629561424255, "learning_rate": 9.727259538662412e-05, "loss": 0.0764, "step": 852 }, { "epoch": 0.09247614917606245, "grad_norm": 0.19624917209148407, "learning_rate": 9.726896851878719e-05, "loss": 0.0294, "step": 853 }, { "epoch": 0.09258456201214224, "grad_norm": 0.40439170598983765, "learning_rate": 9.726534165095025e-05, "loss": 0.0432, "step": 854 }, { "epoch": 0.09269297484822203, "grad_norm": 0.6954824328422546, "learning_rate": 9.726171478311332e-05, "loss": 0.0615, "step": 855 }, { "epoch": 0.09280138768430182, "grad_norm": 1.2729642391204834, "learning_rate": 9.725808791527637e-05, "loss": 0.099, "step": 856 }, { "epoch": 0.09290980052038161, "grad_norm": 0.45091718435287476, "learning_rate": 9.725446104743944e-05, "loss": 0.0588, "step": 857 }, { "epoch": 0.0930182133564614, "grad_norm": 0.39465346932411194, "learning_rate": 9.72508341796025e-05, "loss": 0.0445, "step": 858 }, { "epoch": 0.0931266261925412, "grad_norm": 0.38908788561820984, "learning_rate": 9.724720731176556e-05, "loss": 0.0604, "step": 859 }, { "epoch": 0.09323503902862099, "grad_norm": 2.2615561485290527, "learning_rate": 9.724358044392863e-05, "loss": 0.1029, "step": 860 }, { "epoch": 0.09334345186470078, "grad_norm": 0.631105899810791, "learning_rate": 9.723995357609169e-05, "loss": 0.082, "step": 861 }, { "epoch": 0.09345186470078057, "grad_norm": 0.3978806436061859, "learning_rate": 9.723632670825476e-05, "loss": 0.0549, "step": 862 }, { "epoch": 0.09356027753686036, "grad_norm": 0.23657569289207458, "learning_rate": 9.723269984041782e-05, "loss": 0.0399, "step": 863 }, { "epoch": 0.09366869037294015, "grad_norm": 0.9511329531669617, "learning_rate": 9.722907297258089e-05, "loss": 0.0723, "step": 864 }, { "epoch": 0.09377710320901995, "grad_norm": 0.8120002746582031, "learning_rate": 9.722544610474394e-05, "loss": 0.0952, "step": 865 }, { "epoch": 0.09388551604509975, "grad_norm": 0.7257368564605713, "learning_rate": 9.722181923690701e-05, "loss": 0.0731, "step": 866 }, { "epoch": 0.09399392888117954, "grad_norm": 0.5041123032569885, "learning_rate": 9.721819236907007e-05, "loss": 0.0627, "step": 867 }, { "epoch": 0.09410234171725933, "grad_norm": 0.5649366974830627, "learning_rate": 9.721456550123313e-05, "loss": 0.0814, "step": 868 }, { "epoch": 0.09421075455333912, "grad_norm": 0.9321578145027161, "learning_rate": 9.721093863339621e-05, "loss": 0.0818, "step": 869 }, { "epoch": 0.09431916738941891, "grad_norm": 0.3531586229801178, "learning_rate": 9.720731176555927e-05, "loss": 0.0637, "step": 870 }, { "epoch": 0.0944275802254987, "grad_norm": 0.45406007766723633, "learning_rate": 9.720368489772234e-05, "loss": 0.0603, "step": 871 }, { "epoch": 0.09453599306157849, "grad_norm": 0.43389394879341125, "learning_rate": 9.72000580298854e-05, "loss": 0.0483, "step": 872 }, { "epoch": 0.09464440589765828, "grad_norm": 0.32966193556785583, "learning_rate": 9.719643116204846e-05, "loss": 0.0397, "step": 873 }, { "epoch": 0.09475281873373807, "grad_norm": 0.7599174380302429, "learning_rate": 9.719280429421153e-05, "loss": 0.0681, "step": 874 }, { "epoch": 0.09486123156981786, "grad_norm": 0.32885488867759705, "learning_rate": 9.718917742637458e-05, "loss": 0.0488, "step": 875 }, { "epoch": 0.09496964440589765, "grad_norm": 0.2927687168121338, "learning_rate": 9.718555055853766e-05, "loss": 0.0431, "step": 876 }, { "epoch": 0.09507805724197745, "grad_norm": 0.38131454586982727, "learning_rate": 9.718192369070071e-05, "loss": 0.0383, "step": 877 }, { "epoch": 0.09518647007805724, "grad_norm": 0.7769386172294617, "learning_rate": 9.717829682286378e-05, "loss": 0.0574, "step": 878 }, { "epoch": 0.09529488291413703, "grad_norm": 0.6583757400512695, "learning_rate": 9.717466995502684e-05, "loss": 0.041, "step": 879 }, { "epoch": 0.09540329575021683, "grad_norm": 1.05270254611969, "learning_rate": 9.717104308718991e-05, "loss": 0.0627, "step": 880 }, { "epoch": 0.09551170858629662, "grad_norm": 0.32431045174598694, "learning_rate": 9.716741621935297e-05, "loss": 0.0487, "step": 881 }, { "epoch": 0.09562012142237641, "grad_norm": 0.7411895394325256, "learning_rate": 9.716378935151603e-05, "loss": 0.063, "step": 882 }, { "epoch": 0.0957285342584562, "grad_norm": 0.6136487722396851, "learning_rate": 9.71601624836791e-05, "loss": 0.0439, "step": 883 }, { "epoch": 0.095836947094536, "grad_norm": 0.41457998752593994, "learning_rate": 9.715653561584215e-05, "loss": 0.0536, "step": 884 }, { "epoch": 0.09594535993061579, "grad_norm": 1.583255648612976, "learning_rate": 9.715290874800523e-05, "loss": 0.1277, "step": 885 }, { "epoch": 0.09605377276669558, "grad_norm": 1.1374704837799072, "learning_rate": 9.714928188016828e-05, "loss": 0.0895, "step": 886 }, { "epoch": 0.09616218560277537, "grad_norm": 0.7820709943771362, "learning_rate": 9.714565501233135e-05, "loss": 0.1178, "step": 887 }, { "epoch": 0.09627059843885516, "grad_norm": 0.48252615332603455, "learning_rate": 9.714202814449442e-05, "loss": 0.0588, "step": 888 }, { "epoch": 0.09637901127493495, "grad_norm": 0.2123064547777176, "learning_rate": 9.713840127665748e-05, "loss": 0.0211, "step": 889 }, { "epoch": 0.09648742411101474, "grad_norm": 0.7018849849700928, "learning_rate": 9.713477440882055e-05, "loss": 0.0719, "step": 890 }, { "epoch": 0.09659583694709453, "grad_norm": 0.24949340522289276, "learning_rate": 9.713114754098361e-05, "loss": 0.0222, "step": 891 }, { "epoch": 0.09670424978317432, "grad_norm": 0.6435171961784363, "learning_rate": 9.712752067314668e-05, "loss": 0.035, "step": 892 }, { "epoch": 0.09681266261925411, "grad_norm": 0.47615543007850647, "learning_rate": 9.712389380530974e-05, "loss": 0.0649, "step": 893 }, { "epoch": 0.0969210754553339, "grad_norm": 0.4020293951034546, "learning_rate": 9.712026693747281e-05, "loss": 0.075, "step": 894 }, { "epoch": 0.09702948829141371, "grad_norm": 0.3078588843345642, "learning_rate": 9.711664006963587e-05, "loss": 0.0501, "step": 895 }, { "epoch": 0.0971379011274935, "grad_norm": 0.8299762010574341, "learning_rate": 9.711301320179892e-05, "loss": 0.0672, "step": 896 }, { "epoch": 0.09724631396357329, "grad_norm": 0.6053393483161926, "learning_rate": 9.7109386333962e-05, "loss": 0.0349, "step": 897 }, { "epoch": 0.09735472679965308, "grad_norm": 0.26172468066215515, "learning_rate": 9.710575946612505e-05, "loss": 0.0496, "step": 898 }, { "epoch": 0.09746313963573287, "grad_norm": 0.43970707058906555, "learning_rate": 9.710213259828812e-05, "loss": 0.0567, "step": 899 }, { "epoch": 0.09757155247181266, "grad_norm": 0.45930415391921997, "learning_rate": 9.709850573045118e-05, "loss": 0.0572, "step": 900 }, { "epoch": 0.09767996530789246, "grad_norm": 0.7568166255950928, "learning_rate": 9.709487886261425e-05, "loss": 0.0612, "step": 901 }, { "epoch": 0.09778837814397225, "grad_norm": 0.3107117712497711, "learning_rate": 9.709125199477731e-05, "loss": 0.0246, "step": 902 }, { "epoch": 0.09789679098005204, "grad_norm": 0.461914598941803, "learning_rate": 9.708762512694038e-05, "loss": 0.0651, "step": 903 }, { "epoch": 0.09800520381613183, "grad_norm": 0.4231007397174835, "learning_rate": 9.708399825910345e-05, "loss": 0.0553, "step": 904 }, { "epoch": 0.09811361665221162, "grad_norm": 0.40642669796943665, "learning_rate": 9.708037139126651e-05, "loss": 0.0551, "step": 905 }, { "epoch": 0.09822202948829141, "grad_norm": 0.5104663968086243, "learning_rate": 9.707674452342958e-05, "loss": 0.0784, "step": 906 }, { "epoch": 0.0983304423243712, "grad_norm": 0.6891834139823914, "learning_rate": 9.707311765559264e-05, "loss": 0.0634, "step": 907 }, { "epoch": 0.09843885516045099, "grad_norm": 0.2433965504169464, "learning_rate": 9.706949078775571e-05, "loss": 0.0214, "step": 908 }, { "epoch": 0.09854726799653078, "grad_norm": 0.22759389877319336, "learning_rate": 9.706586391991876e-05, "loss": 0.0194, "step": 909 }, { "epoch": 0.09865568083261059, "grad_norm": 0.5285367369651794, "learning_rate": 9.706223705208182e-05, "loss": 0.0427, "step": 910 }, { "epoch": 0.09876409366869038, "grad_norm": 0.6779706478118896, "learning_rate": 9.705861018424489e-05, "loss": 0.0802, "step": 911 }, { "epoch": 0.09887250650477017, "grad_norm": 0.3551822006702423, "learning_rate": 9.705498331640795e-05, "loss": 0.0531, "step": 912 }, { "epoch": 0.09898091934084996, "grad_norm": 0.41520559787750244, "learning_rate": 9.705135644857102e-05, "loss": 0.0369, "step": 913 }, { "epoch": 0.09908933217692975, "grad_norm": 0.7583938837051392, "learning_rate": 9.704772958073408e-05, "loss": 0.0595, "step": 914 }, { "epoch": 0.09919774501300954, "grad_norm": 0.5787798762321472, "learning_rate": 9.704410271289715e-05, "loss": 0.0624, "step": 915 }, { "epoch": 0.09930615784908933, "grad_norm": 0.5549026727676392, "learning_rate": 9.70404758450602e-05, "loss": 0.0514, "step": 916 }, { "epoch": 0.09941457068516912, "grad_norm": 0.7451505661010742, "learning_rate": 9.703684897722328e-05, "loss": 0.0702, "step": 917 }, { "epoch": 0.09952298352124891, "grad_norm": 0.5709109902381897, "learning_rate": 9.703322210938633e-05, "loss": 0.0502, "step": 918 }, { "epoch": 0.0996313963573287, "grad_norm": 0.4457632601261139, "learning_rate": 9.702959524154939e-05, "loss": 0.0223, "step": 919 }, { "epoch": 0.0997398091934085, "grad_norm": 0.6904404163360596, "learning_rate": 9.702596837371246e-05, "loss": 0.0341, "step": 920 }, { "epoch": 0.09984822202948829, "grad_norm": 0.313882440328598, "learning_rate": 9.702234150587553e-05, "loss": 0.0293, "step": 921 }, { "epoch": 0.09995663486556808, "grad_norm": 0.38196036219596863, "learning_rate": 9.70187146380386e-05, "loss": 0.0223, "step": 922 }, { "epoch": 0.10006504770164787, "grad_norm": 0.8842722773551941, "learning_rate": 9.701508777020166e-05, "loss": 0.0662, "step": 923 }, { "epoch": 0.10017346053772767, "grad_norm": 0.616826057434082, "learning_rate": 9.701146090236473e-05, "loss": 0.0507, "step": 924 }, { "epoch": 0.10028187337380746, "grad_norm": 0.4196354150772095, "learning_rate": 9.700783403452779e-05, "loss": 0.034, "step": 925 }, { "epoch": 0.10039028620988726, "grad_norm": 0.6138347387313843, "learning_rate": 9.700420716669085e-05, "loss": 0.0799, "step": 926 }, { "epoch": 0.10049869904596705, "grad_norm": 0.9117764830589294, "learning_rate": 9.700058029885392e-05, "loss": 0.1189, "step": 927 }, { "epoch": 0.10060711188204684, "grad_norm": 1.1315587759017944, "learning_rate": 9.699695343101698e-05, "loss": 0.1218, "step": 928 }, { "epoch": 0.10071552471812663, "grad_norm": 0.44362807273864746, "learning_rate": 9.699332656318005e-05, "loss": 0.0936, "step": 929 }, { "epoch": 0.10082393755420642, "grad_norm": 0.7190647125244141, "learning_rate": 9.69896996953431e-05, "loss": 0.1026, "step": 930 }, { "epoch": 0.10093235039028621, "grad_norm": 0.472698837518692, "learning_rate": 9.698607282750617e-05, "loss": 0.0665, "step": 931 }, { "epoch": 0.101040763226366, "grad_norm": 0.3025296926498413, "learning_rate": 9.698244595966923e-05, "loss": 0.0429, "step": 932 }, { "epoch": 0.10114917606244579, "grad_norm": 0.20290270447731018, "learning_rate": 9.697881909183229e-05, "loss": 0.0337, "step": 933 }, { "epoch": 0.10125758889852558, "grad_norm": 0.4849521815776825, "learning_rate": 9.697519222399536e-05, "loss": 0.0873, "step": 934 }, { "epoch": 0.10136600173460537, "grad_norm": 0.524376630783081, "learning_rate": 9.697156535615842e-05, "loss": 0.0575, "step": 935 }, { "epoch": 0.10147441457068516, "grad_norm": 0.5084156394004822, "learning_rate": 9.696793848832149e-05, "loss": 0.0942, "step": 936 }, { "epoch": 0.10158282740676496, "grad_norm": 0.4380047917366028, "learning_rate": 9.696431162048455e-05, "loss": 0.0772, "step": 937 }, { "epoch": 0.10169124024284475, "grad_norm": 0.5654634833335876, "learning_rate": 9.696068475264763e-05, "loss": 0.0735, "step": 938 }, { "epoch": 0.10179965307892455, "grad_norm": 0.5022498369216919, "learning_rate": 9.695705788481069e-05, "loss": 0.0656, "step": 939 }, { "epoch": 0.10190806591500434, "grad_norm": 0.3460265100002289, "learning_rate": 9.695343101697374e-05, "loss": 0.0311, "step": 940 }, { "epoch": 0.10201647875108413, "grad_norm": 0.43353426456451416, "learning_rate": 9.694980414913682e-05, "loss": 0.0373, "step": 941 }, { "epoch": 0.10212489158716392, "grad_norm": 0.48588454723358154, "learning_rate": 9.694617728129987e-05, "loss": 0.0645, "step": 942 }, { "epoch": 0.10223330442324371, "grad_norm": 0.4235095977783203, "learning_rate": 9.694255041346294e-05, "loss": 0.0875, "step": 943 }, { "epoch": 0.1023417172593235, "grad_norm": 0.4058656096458435, "learning_rate": 9.6938923545626e-05, "loss": 0.0825, "step": 944 }, { "epoch": 0.1024501300954033, "grad_norm": 0.39487597346305847, "learning_rate": 9.693529667778907e-05, "loss": 0.0543, "step": 945 }, { "epoch": 0.10255854293148309, "grad_norm": 0.45582959055900574, "learning_rate": 9.693166980995213e-05, "loss": 0.0642, "step": 946 }, { "epoch": 0.10266695576756288, "grad_norm": 0.6082140207290649, "learning_rate": 9.69280429421152e-05, "loss": 0.0551, "step": 947 }, { "epoch": 0.10277536860364267, "grad_norm": 0.30615606904029846, "learning_rate": 9.692441607427826e-05, "loss": 0.0663, "step": 948 }, { "epoch": 0.10288378143972246, "grad_norm": 0.3150961399078369, "learning_rate": 9.692078920644131e-05, "loss": 0.0565, "step": 949 }, { "epoch": 0.10299219427580225, "grad_norm": 0.42448800802230835, "learning_rate": 9.691716233860439e-05, "loss": 0.0407, "step": 950 }, { "epoch": 0.10310060711188204, "grad_norm": 0.43817025423049927, "learning_rate": 9.691353547076744e-05, "loss": 0.0513, "step": 951 }, { "epoch": 0.10320901994796183, "grad_norm": 0.6212766170501709, "learning_rate": 9.690990860293051e-05, "loss": 0.0552, "step": 952 }, { "epoch": 0.10331743278404162, "grad_norm": 0.39654386043548584, "learning_rate": 9.690628173509357e-05, "loss": 0.0446, "step": 953 }, { "epoch": 0.10342584562012143, "grad_norm": 0.37387293577194214, "learning_rate": 9.690265486725664e-05, "loss": 0.043, "step": 954 }, { "epoch": 0.10353425845620122, "grad_norm": 0.31893083453178406, "learning_rate": 9.68990279994197e-05, "loss": 0.0372, "step": 955 }, { "epoch": 0.10364267129228101, "grad_norm": 0.1671503484249115, "learning_rate": 9.689540113158277e-05, "loss": 0.0259, "step": 956 }, { "epoch": 0.1037510841283608, "grad_norm": 0.5798497796058655, "learning_rate": 9.689177426374584e-05, "loss": 0.0592, "step": 957 }, { "epoch": 0.10385949696444059, "grad_norm": 0.596616804599762, "learning_rate": 9.68881473959089e-05, "loss": 0.0776, "step": 958 }, { "epoch": 0.10396790980052038, "grad_norm": 0.6153174638748169, "learning_rate": 9.688452052807197e-05, "loss": 0.0329, "step": 959 }, { "epoch": 0.10407632263660017, "grad_norm": 0.4208217263221741, "learning_rate": 9.688089366023503e-05, "loss": 0.068, "step": 960 }, { "epoch": 0.10418473547267997, "grad_norm": 0.707732617855072, "learning_rate": 9.68772667923981e-05, "loss": 0.0751, "step": 961 }, { "epoch": 0.10429314830875976, "grad_norm": 0.5705692768096924, "learning_rate": 9.687363992456115e-05, "loss": 0.0371, "step": 962 }, { "epoch": 0.10440156114483955, "grad_norm": 0.8788425922393799, "learning_rate": 9.687001305672421e-05, "loss": 0.0865, "step": 963 }, { "epoch": 0.10450997398091934, "grad_norm": 0.2891666889190674, "learning_rate": 9.686638618888728e-05, "loss": 0.0394, "step": 964 }, { "epoch": 0.10461838681699913, "grad_norm": 0.9409440159797668, "learning_rate": 9.686275932105034e-05, "loss": 0.0905, "step": 965 }, { "epoch": 0.10472679965307892, "grad_norm": 0.6850751042366028, "learning_rate": 9.685913245321341e-05, "loss": 0.0729, "step": 966 }, { "epoch": 0.10483521248915871, "grad_norm": 0.749049723148346, "learning_rate": 9.685550558537647e-05, "loss": 0.0768, "step": 967 }, { "epoch": 0.1049436253252385, "grad_norm": 0.48270711302757263, "learning_rate": 9.685187871753954e-05, "loss": 0.0717, "step": 968 }, { "epoch": 0.1050520381613183, "grad_norm": 0.458347886800766, "learning_rate": 9.68482518497026e-05, "loss": 0.0535, "step": 969 }, { "epoch": 0.1051604509973981, "grad_norm": 0.44173625111579895, "learning_rate": 9.684462498186567e-05, "loss": 0.1134, "step": 970 }, { "epoch": 0.10526886383347789, "grad_norm": 0.6913609504699707, "learning_rate": 9.684099811402873e-05, "loss": 0.0412, "step": 971 }, { "epoch": 0.10537727666955768, "grad_norm": 0.20152771472930908, "learning_rate": 9.683737124619178e-05, "loss": 0.0294, "step": 972 }, { "epoch": 0.10548568950563747, "grad_norm": 0.5906622409820557, "learning_rate": 9.683374437835487e-05, "loss": 0.0499, "step": 973 }, { "epoch": 0.10559410234171726, "grad_norm": 0.32683467864990234, "learning_rate": 9.683011751051792e-05, "loss": 0.0341, "step": 974 }, { "epoch": 0.10570251517779705, "grad_norm": 0.5487158298492432, "learning_rate": 9.6826490642681e-05, "loss": 0.0823, "step": 975 }, { "epoch": 0.10581092801387684, "grad_norm": 0.7268614768981934, "learning_rate": 9.682286377484405e-05, "loss": 0.0982, "step": 976 }, { "epoch": 0.10591934084995663, "grad_norm": 0.5675257444381714, "learning_rate": 9.681923690700711e-05, "loss": 0.0911, "step": 977 }, { "epoch": 0.10602775368603642, "grad_norm": 0.5872218608856201, "learning_rate": 9.681561003917018e-05, "loss": 0.0772, "step": 978 }, { "epoch": 0.10613616652211622, "grad_norm": 0.33188268542289734, "learning_rate": 9.681198317133324e-05, "loss": 0.0348, "step": 979 }, { "epoch": 0.106244579358196, "grad_norm": 0.418849915266037, "learning_rate": 9.680835630349631e-05, "loss": 0.0495, "step": 980 }, { "epoch": 0.1063529921942758, "grad_norm": 0.35065004229545593, "learning_rate": 9.680472943565937e-05, "loss": 0.0598, "step": 981 }, { "epoch": 0.10646140503035559, "grad_norm": 0.26746946573257446, "learning_rate": 9.680110256782244e-05, "loss": 0.0394, "step": 982 }, { "epoch": 0.10656981786643538, "grad_norm": 0.5505760312080383, "learning_rate": 9.67974756999855e-05, "loss": 0.0336, "step": 983 }, { "epoch": 0.10667823070251518, "grad_norm": 0.41639742255210876, "learning_rate": 9.679384883214857e-05, "loss": 0.0811, "step": 984 }, { "epoch": 0.10678664353859497, "grad_norm": 0.691643476486206, "learning_rate": 9.679022196431162e-05, "loss": 0.0475, "step": 985 }, { "epoch": 0.10689505637467477, "grad_norm": 0.407133013010025, "learning_rate": 9.678659509647468e-05, "loss": 0.0666, "step": 986 }, { "epoch": 0.10700346921075456, "grad_norm": 0.3540577292442322, "learning_rate": 9.678296822863775e-05, "loss": 0.0394, "step": 987 }, { "epoch": 0.10711188204683435, "grad_norm": 0.56771320104599, "learning_rate": 9.677934136080081e-05, "loss": 0.0853, "step": 988 }, { "epoch": 0.10722029488291414, "grad_norm": 0.8499017953872681, "learning_rate": 9.677571449296388e-05, "loss": 0.0878, "step": 989 }, { "epoch": 0.10732870771899393, "grad_norm": 0.35677117109298706, "learning_rate": 9.677208762512695e-05, "loss": 0.0232, "step": 990 }, { "epoch": 0.10743712055507372, "grad_norm": 0.46847012639045715, "learning_rate": 9.676846075729001e-05, "loss": 0.0662, "step": 991 }, { "epoch": 0.10754553339115351, "grad_norm": 0.7161316871643066, "learning_rate": 9.676483388945308e-05, "loss": 0.1027, "step": 992 }, { "epoch": 0.1076539462272333, "grad_norm": 0.6095179915428162, "learning_rate": 9.676120702161614e-05, "loss": 0.0644, "step": 993 }, { "epoch": 0.10776235906331309, "grad_norm": 0.5793284177780151, "learning_rate": 9.67575801537792e-05, "loss": 0.0993, "step": 994 }, { "epoch": 0.10787077189939288, "grad_norm": 0.30059322714805603, "learning_rate": 9.675395328594226e-05, "loss": 0.0386, "step": 995 }, { "epoch": 0.10797918473547267, "grad_norm": 0.41009074449539185, "learning_rate": 9.675032641810533e-05, "loss": 0.0623, "step": 996 }, { "epoch": 0.10808759757155247, "grad_norm": 0.322340190410614, "learning_rate": 9.674669955026839e-05, "loss": 0.051, "step": 997 }, { "epoch": 0.10819601040763227, "grad_norm": 0.30558645725250244, "learning_rate": 9.674307268243146e-05, "loss": 0.0377, "step": 998 }, { "epoch": 0.10830442324371206, "grad_norm": 0.39534226059913635, "learning_rate": 9.673944581459452e-05, "loss": 0.0504, "step": 999 }, { "epoch": 0.10841283607979185, "grad_norm": 0.29280000925064087, "learning_rate": 9.673581894675758e-05, "loss": 0.0439, "step": 1000 }, { "epoch": 0.10852124891587164, "grad_norm": 0.3014381527900696, "learning_rate": 9.673219207892065e-05, "loss": 0.0276, "step": 1001 }, { "epoch": 0.10862966175195143, "grad_norm": 0.19772356748580933, "learning_rate": 9.67285652110837e-05, "loss": 0.0186, "step": 1002 }, { "epoch": 0.10873807458803122, "grad_norm": 0.46341603994369507, "learning_rate": 9.672493834324678e-05, "loss": 0.0535, "step": 1003 }, { "epoch": 0.10884648742411102, "grad_norm": 0.33851319551467896, "learning_rate": 9.672131147540983e-05, "loss": 0.0465, "step": 1004 }, { "epoch": 0.1089549002601908, "grad_norm": 0.939196765422821, "learning_rate": 9.67176846075729e-05, "loss": 0.0915, "step": 1005 }, { "epoch": 0.1090633130962706, "grad_norm": 0.39463311433792114, "learning_rate": 9.671405773973596e-05, "loss": 0.0348, "step": 1006 }, { "epoch": 0.10917172593235039, "grad_norm": 0.4738016426563263, "learning_rate": 9.671043087189903e-05, "loss": 0.0793, "step": 1007 }, { "epoch": 0.10928013876843018, "grad_norm": 0.442687451839447, "learning_rate": 9.67068040040621e-05, "loss": 0.0539, "step": 1008 }, { "epoch": 0.10938855160450997, "grad_norm": 0.40828937292099, "learning_rate": 9.670317713622516e-05, "loss": 0.0569, "step": 1009 }, { "epoch": 0.10949696444058976, "grad_norm": 0.44929635524749756, "learning_rate": 9.669955026838823e-05, "loss": 0.0819, "step": 1010 }, { "epoch": 0.10960537727666955, "grad_norm": 0.7317262887954712, "learning_rate": 9.669592340055129e-05, "loss": 0.0772, "step": 1011 }, { "epoch": 0.10971379011274934, "grad_norm": 0.4149087369441986, "learning_rate": 9.669229653271436e-05, "loss": 0.0616, "step": 1012 }, { "epoch": 0.10982220294882915, "grad_norm": 0.864345133304596, "learning_rate": 9.668866966487742e-05, "loss": 0.0664, "step": 1013 }, { "epoch": 0.10993061578490894, "grad_norm": 0.2655927836894989, "learning_rate": 9.668504279704047e-05, "loss": 0.0436, "step": 1014 }, { "epoch": 0.11003902862098873, "grad_norm": 0.2810893654823303, "learning_rate": 9.668141592920355e-05, "loss": 0.0637, "step": 1015 }, { "epoch": 0.11014744145706852, "grad_norm": 0.7106830477714539, "learning_rate": 9.66777890613666e-05, "loss": 0.0596, "step": 1016 }, { "epoch": 0.11025585429314831, "grad_norm": 0.16435138881206512, "learning_rate": 9.667416219352967e-05, "loss": 0.0207, "step": 1017 }, { "epoch": 0.1103642671292281, "grad_norm": 0.6856909990310669, "learning_rate": 9.667053532569273e-05, "loss": 0.0554, "step": 1018 }, { "epoch": 0.1104726799653079, "grad_norm": 0.4375952184200287, "learning_rate": 9.66669084578558e-05, "loss": 0.0233, "step": 1019 }, { "epoch": 0.11058109280138768, "grad_norm": 0.5433666110038757, "learning_rate": 9.666328159001886e-05, "loss": 0.0734, "step": 1020 }, { "epoch": 0.11068950563746748, "grad_norm": 0.26406222581863403, "learning_rate": 9.665965472218193e-05, "loss": 0.041, "step": 1021 }, { "epoch": 0.11079791847354727, "grad_norm": 0.6800373792648315, "learning_rate": 9.665602785434499e-05, "loss": 0.095, "step": 1022 }, { "epoch": 0.11090633130962706, "grad_norm": 0.43484964966773987, "learning_rate": 9.665240098650805e-05, "loss": 0.0622, "step": 1023 }, { "epoch": 0.11101474414570685, "grad_norm": 0.2864947021007538, "learning_rate": 9.664877411867112e-05, "loss": 0.019, "step": 1024 }, { "epoch": 0.11112315698178664, "grad_norm": 0.4075694680213928, "learning_rate": 9.664514725083419e-05, "loss": 0.0483, "step": 1025 }, { "epoch": 0.11123156981786643, "grad_norm": 0.8093576431274414, "learning_rate": 9.664152038299726e-05, "loss": 0.0778, "step": 1026 }, { "epoch": 0.11133998265394622, "grad_norm": 0.34847694635391235, "learning_rate": 9.663789351516032e-05, "loss": 0.0502, "step": 1027 }, { "epoch": 0.11144839549002603, "grad_norm": 0.6136441826820374, "learning_rate": 9.663426664732339e-05, "loss": 0.0718, "step": 1028 }, { "epoch": 0.11155680832610582, "grad_norm": 0.43427225947380066, "learning_rate": 9.663063977948644e-05, "loss": 0.0539, "step": 1029 }, { "epoch": 0.11166522116218561, "grad_norm": 0.6257562041282654, "learning_rate": 9.66270129116495e-05, "loss": 0.0733, "step": 1030 }, { "epoch": 0.1117736339982654, "grad_norm": 0.22273865342140198, "learning_rate": 9.662338604381257e-05, "loss": 0.0347, "step": 1031 }, { "epoch": 0.11188204683434519, "grad_norm": 0.5254814028739929, "learning_rate": 9.661975917597563e-05, "loss": 0.0523, "step": 1032 }, { "epoch": 0.11199045967042498, "grad_norm": 0.386201947927475, "learning_rate": 9.66161323081387e-05, "loss": 0.0252, "step": 1033 }, { "epoch": 0.11209887250650477, "grad_norm": 0.5358994603157043, "learning_rate": 9.661250544030176e-05, "loss": 0.0893, "step": 1034 }, { "epoch": 0.11220728534258456, "grad_norm": 1.1614570617675781, "learning_rate": 9.660887857246483e-05, "loss": 0.0472, "step": 1035 }, { "epoch": 0.11231569817866435, "grad_norm": 0.6813406944274902, "learning_rate": 9.660525170462789e-05, "loss": 0.0518, "step": 1036 }, { "epoch": 0.11242411101474414, "grad_norm": 0.6659958362579346, "learning_rate": 9.660162483679096e-05, "loss": 0.0546, "step": 1037 }, { "epoch": 0.11253252385082393, "grad_norm": 0.8444514274597168, "learning_rate": 9.659799796895401e-05, "loss": 0.1285, "step": 1038 }, { "epoch": 0.11264093668690373, "grad_norm": 0.34226834774017334, "learning_rate": 9.659437110111707e-05, "loss": 0.0471, "step": 1039 }, { "epoch": 0.11274934952298352, "grad_norm": 0.5110864639282227, "learning_rate": 9.659074423328014e-05, "loss": 0.0544, "step": 1040 }, { "epoch": 0.11285776235906331, "grad_norm": 0.26054611802101135, "learning_rate": 9.65871173654432e-05, "loss": 0.0351, "step": 1041 }, { "epoch": 0.1129661751951431, "grad_norm": 0.5149602890014648, "learning_rate": 9.658349049760628e-05, "loss": 0.0245, "step": 1042 }, { "epoch": 0.1130745880312229, "grad_norm": 0.2102397084236145, "learning_rate": 9.657986362976934e-05, "loss": 0.0279, "step": 1043 }, { "epoch": 0.1131830008673027, "grad_norm": 0.42370739579200745, "learning_rate": 9.65762367619324e-05, "loss": 0.0405, "step": 1044 }, { "epoch": 0.11329141370338248, "grad_norm": 0.6443818211555481, "learning_rate": 9.657260989409547e-05, "loss": 0.0264, "step": 1045 }, { "epoch": 0.11339982653946228, "grad_norm": 0.7578242421150208, "learning_rate": 9.656898302625853e-05, "loss": 0.0412, "step": 1046 }, { "epoch": 0.11350823937554207, "grad_norm": 0.3388437032699585, "learning_rate": 9.65653561584216e-05, "loss": 0.031, "step": 1047 }, { "epoch": 0.11361665221162186, "grad_norm": 0.5147340893745422, "learning_rate": 9.656172929058465e-05, "loss": 0.0901, "step": 1048 }, { "epoch": 0.11372506504770165, "grad_norm": 0.7063462138175964, "learning_rate": 9.655810242274773e-05, "loss": 0.0504, "step": 1049 }, { "epoch": 0.11383347788378144, "grad_norm": 0.8521038889884949, "learning_rate": 9.655447555491078e-05, "loss": 0.064, "step": 1050 }, { "epoch": 0.11394189071986123, "grad_norm": 0.7780292630195618, "learning_rate": 9.655084868707385e-05, "loss": 0.1055, "step": 1051 }, { "epoch": 0.11405030355594102, "grad_norm": 0.701786458492279, "learning_rate": 9.654722181923691e-05, "loss": 0.0893, "step": 1052 }, { "epoch": 0.11415871639202081, "grad_norm": 0.4510385990142822, "learning_rate": 9.654359495139997e-05, "loss": 0.0688, "step": 1053 }, { "epoch": 0.1142671292281006, "grad_norm": 0.9456707835197449, "learning_rate": 9.653996808356304e-05, "loss": 0.0765, "step": 1054 }, { "epoch": 0.1143755420641804, "grad_norm": 0.2837821841239929, "learning_rate": 9.65363412157261e-05, "loss": 0.041, "step": 1055 }, { "epoch": 0.11448395490026018, "grad_norm": 0.3080383539199829, "learning_rate": 9.653271434788917e-05, "loss": 0.036, "step": 1056 }, { "epoch": 0.11459236773633999, "grad_norm": 0.6853926777839661, "learning_rate": 9.652908748005222e-05, "loss": 0.134, "step": 1057 }, { "epoch": 0.11470078057241978, "grad_norm": 0.5523648858070374, "learning_rate": 9.65254606122153e-05, "loss": 0.0644, "step": 1058 }, { "epoch": 0.11480919340849957, "grad_norm": 1.0360603332519531, "learning_rate": 9.652183374437837e-05, "loss": 0.1219, "step": 1059 }, { "epoch": 0.11491760624457936, "grad_norm": 0.3034614026546478, "learning_rate": 9.651820687654142e-05, "loss": 0.0369, "step": 1060 }, { "epoch": 0.11502601908065915, "grad_norm": 0.24202661216259003, "learning_rate": 9.65145800087045e-05, "loss": 0.0485, "step": 1061 }, { "epoch": 0.11513443191673894, "grad_norm": 0.35690969228744507, "learning_rate": 9.651095314086755e-05, "loss": 0.0805, "step": 1062 }, { "epoch": 0.11524284475281873, "grad_norm": 0.3328808844089508, "learning_rate": 9.650732627303062e-05, "loss": 0.0683, "step": 1063 }, { "epoch": 0.11535125758889853, "grad_norm": 0.5619276762008667, "learning_rate": 9.650369940519368e-05, "loss": 0.0561, "step": 1064 }, { "epoch": 0.11545967042497832, "grad_norm": 0.3078155517578125, "learning_rate": 9.650007253735675e-05, "loss": 0.038, "step": 1065 }, { "epoch": 0.11556808326105811, "grad_norm": 0.24095912277698517, "learning_rate": 9.649644566951981e-05, "loss": 0.0537, "step": 1066 }, { "epoch": 0.1156764960971379, "grad_norm": 0.5911895632743835, "learning_rate": 9.649281880168287e-05, "loss": 0.0868, "step": 1067 }, { "epoch": 0.11578490893321769, "grad_norm": 0.28748205304145813, "learning_rate": 9.648919193384594e-05, "loss": 0.0609, "step": 1068 }, { "epoch": 0.11589332176929748, "grad_norm": 0.3909411132335663, "learning_rate": 9.6485565066009e-05, "loss": 0.082, "step": 1069 }, { "epoch": 0.11600173460537727, "grad_norm": 0.2703455686569214, "learning_rate": 9.648193819817206e-05, "loss": 0.0459, "step": 1070 }, { "epoch": 0.11611014744145706, "grad_norm": 0.5466304421424866, "learning_rate": 9.647831133033512e-05, "loss": 0.0921, "step": 1071 }, { "epoch": 0.11621856027753687, "grad_norm": 0.3206017315387726, "learning_rate": 9.647468446249819e-05, "loss": 0.09, "step": 1072 }, { "epoch": 0.11632697311361666, "grad_norm": 0.48652854561805725, "learning_rate": 9.647105759466125e-05, "loss": 0.0883, "step": 1073 }, { "epoch": 0.11643538594969645, "grad_norm": 0.34245389699935913, "learning_rate": 9.646743072682432e-05, "loss": 0.0519, "step": 1074 }, { "epoch": 0.11654379878577624, "grad_norm": 0.43592971563339233, "learning_rate": 9.646380385898738e-05, "loss": 0.0492, "step": 1075 }, { "epoch": 0.11665221162185603, "grad_norm": 0.3599115014076233, "learning_rate": 9.646017699115044e-05, "loss": 0.0658, "step": 1076 }, { "epoch": 0.11676062445793582, "grad_norm": 0.2649509012699127, "learning_rate": 9.645655012331352e-05, "loss": 0.0395, "step": 1077 }, { "epoch": 0.11686903729401561, "grad_norm": 0.5017350912094116, "learning_rate": 9.645292325547658e-05, "loss": 0.0808, "step": 1078 }, { "epoch": 0.1169774501300954, "grad_norm": 0.3130677342414856, "learning_rate": 9.644929638763965e-05, "loss": 0.0432, "step": 1079 }, { "epoch": 0.1170858629661752, "grad_norm": 0.8074643015861511, "learning_rate": 9.64456695198027e-05, "loss": 0.0615, "step": 1080 }, { "epoch": 0.11719427580225499, "grad_norm": 0.41579389572143555, "learning_rate": 9.644204265196576e-05, "loss": 0.0576, "step": 1081 }, { "epoch": 0.11730268863833478, "grad_norm": 0.6546041369438171, "learning_rate": 9.643841578412883e-05, "loss": 0.0892, "step": 1082 }, { "epoch": 0.11741110147441457, "grad_norm": 0.4711214303970337, "learning_rate": 9.643478891629189e-05, "loss": 0.0794, "step": 1083 }, { "epoch": 0.11751951431049436, "grad_norm": 0.2520419955253601, "learning_rate": 9.643116204845496e-05, "loss": 0.0448, "step": 1084 }, { "epoch": 0.11762792714657415, "grad_norm": 0.33411410450935364, "learning_rate": 9.642753518061802e-05, "loss": 0.0353, "step": 1085 }, { "epoch": 0.11773633998265394, "grad_norm": 0.5164233446121216, "learning_rate": 9.642390831278109e-05, "loss": 0.0349, "step": 1086 }, { "epoch": 0.11784475281873374, "grad_norm": 0.4973407983779907, "learning_rate": 9.642028144494415e-05, "loss": 0.0487, "step": 1087 }, { "epoch": 0.11795316565481354, "grad_norm": 0.5558212399482727, "learning_rate": 9.641665457710722e-05, "loss": 0.0708, "step": 1088 }, { "epoch": 0.11806157849089333, "grad_norm": 0.3955412805080414, "learning_rate": 9.641302770927028e-05, "loss": 0.0624, "step": 1089 }, { "epoch": 0.11816999132697312, "grad_norm": 0.5058233141899109, "learning_rate": 9.640940084143333e-05, "loss": 0.0368, "step": 1090 }, { "epoch": 0.11827840416305291, "grad_norm": 0.2904975712299347, "learning_rate": 9.64057739735964e-05, "loss": 0.0345, "step": 1091 }, { "epoch": 0.1183868169991327, "grad_norm": 0.41082143783569336, "learning_rate": 9.640214710575946e-05, "loss": 0.0352, "step": 1092 }, { "epoch": 0.11849522983521249, "grad_norm": 0.7524197101593018, "learning_rate": 9.639852023792253e-05, "loss": 0.1402, "step": 1093 }, { "epoch": 0.11860364267129228, "grad_norm": 0.3644576668739319, "learning_rate": 9.63948933700856e-05, "loss": 0.0597, "step": 1094 }, { "epoch": 0.11871205550737207, "grad_norm": 0.3779633045196533, "learning_rate": 9.639126650224866e-05, "loss": 0.08, "step": 1095 }, { "epoch": 0.11882046834345186, "grad_norm": 0.7365357875823975, "learning_rate": 9.638763963441173e-05, "loss": 0.0828, "step": 1096 }, { "epoch": 0.11892888117953165, "grad_norm": 0.4770847260951996, "learning_rate": 9.638401276657479e-05, "loss": 0.0713, "step": 1097 }, { "epoch": 0.11903729401561144, "grad_norm": 0.8440389037132263, "learning_rate": 9.638038589873786e-05, "loss": 0.1357, "step": 1098 }, { "epoch": 0.11914570685169124, "grad_norm": 0.20506581664085388, "learning_rate": 9.637675903090092e-05, "loss": 0.0316, "step": 1099 }, { "epoch": 0.11925411968777103, "grad_norm": 0.5697807669639587, "learning_rate": 9.637313216306399e-05, "loss": 0.0726, "step": 1100 }, { "epoch": 0.11936253252385082, "grad_norm": 0.49167659878730774, "learning_rate": 9.636950529522705e-05, "loss": 0.0542, "step": 1101 }, { "epoch": 0.11947094535993062, "grad_norm": 0.44727978110313416, "learning_rate": 9.636587842739012e-05, "loss": 0.0439, "step": 1102 }, { "epoch": 0.11957935819601041, "grad_norm": 0.22475802898406982, "learning_rate": 9.636225155955317e-05, "loss": 0.0317, "step": 1103 }, { "epoch": 0.1196877710320902, "grad_norm": 0.36919456720352173, "learning_rate": 9.635862469171623e-05, "loss": 0.0594, "step": 1104 }, { "epoch": 0.11979618386817, "grad_norm": 0.44343405961990356, "learning_rate": 9.63549978238793e-05, "loss": 0.0513, "step": 1105 }, { "epoch": 0.11990459670424979, "grad_norm": 0.33381888270378113, "learning_rate": 9.635137095604236e-05, "loss": 0.0378, "step": 1106 }, { "epoch": 0.12001300954032958, "grad_norm": 0.3900810182094574, "learning_rate": 9.634774408820543e-05, "loss": 0.0681, "step": 1107 }, { "epoch": 0.12012142237640937, "grad_norm": 0.14854711294174194, "learning_rate": 9.634411722036849e-05, "loss": 0.0233, "step": 1108 }, { "epoch": 0.12022983521248916, "grad_norm": 0.42253413796424866, "learning_rate": 9.634049035253156e-05, "loss": 0.0583, "step": 1109 }, { "epoch": 0.12033824804856895, "grad_norm": 0.31772515177726746, "learning_rate": 9.633686348469462e-05, "loss": 0.0448, "step": 1110 }, { "epoch": 0.12044666088464874, "grad_norm": 0.4631556570529938, "learning_rate": 9.633323661685769e-05, "loss": 0.0374, "step": 1111 }, { "epoch": 0.12055507372072853, "grad_norm": 0.8563886284828186, "learning_rate": 9.632960974902076e-05, "loss": 0.104, "step": 1112 }, { "epoch": 0.12066348655680832, "grad_norm": 0.2868839204311371, "learning_rate": 9.632598288118381e-05, "loss": 0.0314, "step": 1113 }, { "epoch": 0.12077189939288811, "grad_norm": 0.5871885418891907, "learning_rate": 9.632235601334689e-05, "loss": 0.0688, "step": 1114 }, { "epoch": 0.1208803122289679, "grad_norm": 0.22705204784870148, "learning_rate": 9.631872914550994e-05, "loss": 0.0214, "step": 1115 }, { "epoch": 0.1209887250650477, "grad_norm": 1.576602578163147, "learning_rate": 9.631510227767301e-05, "loss": 0.052, "step": 1116 }, { "epoch": 0.1210971379011275, "grad_norm": 0.4248339831829071, "learning_rate": 9.631147540983607e-05, "loss": 0.0442, "step": 1117 }, { "epoch": 0.12120555073720729, "grad_norm": 0.4488738775253296, "learning_rate": 9.630784854199913e-05, "loss": 0.0497, "step": 1118 }, { "epoch": 0.12131396357328708, "grad_norm": 0.4074333608150482, "learning_rate": 9.63042216741622e-05, "loss": 0.0311, "step": 1119 }, { "epoch": 0.12142237640936687, "grad_norm": 0.6941240429878235, "learning_rate": 9.630059480632526e-05, "loss": 0.053, "step": 1120 }, { "epoch": 0.12153078924544666, "grad_norm": 0.4593442976474762, "learning_rate": 9.629696793848833e-05, "loss": 0.0762, "step": 1121 }, { "epoch": 0.12163920208152645, "grad_norm": 0.4900933504104614, "learning_rate": 9.629334107065138e-05, "loss": 0.0502, "step": 1122 }, { "epoch": 0.12174761491760624, "grad_norm": 0.6677107810974121, "learning_rate": 9.628971420281446e-05, "loss": 0.0655, "step": 1123 }, { "epoch": 0.12185602775368604, "grad_norm": 0.44758182764053345, "learning_rate": 9.628608733497751e-05, "loss": 0.0549, "step": 1124 }, { "epoch": 0.12196444058976583, "grad_norm": 0.46068090200424194, "learning_rate": 9.628246046714058e-05, "loss": 0.0592, "step": 1125 }, { "epoch": 0.12207285342584562, "grad_norm": 0.8612934350967407, "learning_rate": 9.627883359930364e-05, "loss": 0.0683, "step": 1126 }, { "epoch": 0.12218126626192541, "grad_norm": 0.6874939203262329, "learning_rate": 9.62752067314667e-05, "loss": 0.067, "step": 1127 }, { "epoch": 0.1222896790980052, "grad_norm": 0.8695457577705383, "learning_rate": 9.627157986362978e-05, "loss": 0.1046, "step": 1128 }, { "epoch": 0.12239809193408499, "grad_norm": 0.5003030896186829, "learning_rate": 9.626795299579284e-05, "loss": 0.0639, "step": 1129 }, { "epoch": 0.12250650477016478, "grad_norm": 0.19334985315799713, "learning_rate": 9.626432612795591e-05, "loss": 0.0236, "step": 1130 }, { "epoch": 0.12261491760624459, "grad_norm": 0.3804379403591156, "learning_rate": 9.626069926011897e-05, "loss": 0.0551, "step": 1131 }, { "epoch": 0.12272333044232438, "grad_norm": 0.510175347328186, "learning_rate": 9.625707239228204e-05, "loss": 0.0694, "step": 1132 }, { "epoch": 0.12283174327840417, "grad_norm": 0.30530816316604614, "learning_rate": 9.62534455244451e-05, "loss": 0.0221, "step": 1133 }, { "epoch": 0.12294015611448396, "grad_norm": 0.33630865812301636, "learning_rate": 9.624981865660815e-05, "loss": 0.0412, "step": 1134 }, { "epoch": 0.12304856895056375, "grad_norm": 0.2986772358417511, "learning_rate": 9.624619178877122e-05, "loss": 0.045, "step": 1135 }, { "epoch": 0.12315698178664354, "grad_norm": 0.3045162856578827, "learning_rate": 9.624256492093428e-05, "loss": 0.0615, "step": 1136 }, { "epoch": 0.12326539462272333, "grad_norm": 0.5213388204574585, "learning_rate": 9.623893805309735e-05, "loss": 0.0808, "step": 1137 }, { "epoch": 0.12337380745880312, "grad_norm": 0.9870851039886475, "learning_rate": 9.623531118526041e-05, "loss": 0.0762, "step": 1138 }, { "epoch": 0.12348222029488291, "grad_norm": 0.27978208661079407, "learning_rate": 9.623168431742348e-05, "loss": 0.0387, "step": 1139 }, { "epoch": 0.1235906331309627, "grad_norm": 0.5431694388389587, "learning_rate": 9.622805744958654e-05, "loss": 0.0498, "step": 1140 }, { "epoch": 0.1236990459670425, "grad_norm": 0.42921552062034607, "learning_rate": 9.622443058174961e-05, "loss": 0.0919, "step": 1141 }, { "epoch": 0.12380745880312229, "grad_norm": 0.5430335998535156, "learning_rate": 9.622080371391267e-05, "loss": 0.0504, "step": 1142 }, { "epoch": 0.12391587163920208, "grad_norm": 0.31388768553733826, "learning_rate": 9.621717684607572e-05, "loss": 0.0286, "step": 1143 }, { "epoch": 0.12402428447528187, "grad_norm": 0.41246864199638367, "learning_rate": 9.62135499782388e-05, "loss": 0.0384, "step": 1144 }, { "epoch": 0.12413269731136166, "grad_norm": 0.36065179109573364, "learning_rate": 9.620992311040185e-05, "loss": 0.0701, "step": 1145 }, { "epoch": 0.12424111014744146, "grad_norm": 0.7904974222183228, "learning_rate": 9.620629624256494e-05, "loss": 0.1013, "step": 1146 }, { "epoch": 0.12434952298352125, "grad_norm": 0.4501514434814453, "learning_rate": 9.6202669374728e-05, "loss": 0.0346, "step": 1147 }, { "epoch": 0.12445793581960105, "grad_norm": 0.3448121249675751, "learning_rate": 9.619904250689105e-05, "loss": 0.0508, "step": 1148 }, { "epoch": 0.12456634865568084, "grad_norm": 0.6327135562896729, "learning_rate": 9.619541563905412e-05, "loss": 0.064, "step": 1149 }, { "epoch": 0.12467476149176063, "grad_norm": 0.5862367749214172, "learning_rate": 9.619178877121718e-05, "loss": 0.0555, "step": 1150 }, { "epoch": 0.12478317432784042, "grad_norm": 0.4727838635444641, "learning_rate": 9.618816190338025e-05, "loss": 0.0715, "step": 1151 }, { "epoch": 0.12489158716392021, "grad_norm": 0.3256281018257141, "learning_rate": 9.618453503554331e-05, "loss": 0.0579, "step": 1152 }, { "epoch": 0.125, "grad_norm": 0.5066164135932922, "learning_rate": 9.618090816770638e-05, "loss": 0.0315, "step": 1153 }, { "epoch": 0.1251084128360798, "grad_norm": 0.37879952788352966, "learning_rate": 9.617728129986944e-05, "loss": 0.0755, "step": 1154 }, { "epoch": 0.12521682567215958, "grad_norm": 0.2536046802997589, "learning_rate": 9.617365443203251e-05, "loss": 0.0323, "step": 1155 }, { "epoch": 0.12532523850823937, "grad_norm": 0.19555824995040894, "learning_rate": 9.617002756419556e-05, "loss": 0.0174, "step": 1156 }, { "epoch": 0.12543365134431916, "grad_norm": 0.4799111485481262, "learning_rate": 9.616640069635862e-05, "loss": 0.0549, "step": 1157 }, { "epoch": 0.12554206418039895, "grad_norm": 0.2737760543823242, "learning_rate": 9.616277382852169e-05, "loss": 0.0417, "step": 1158 }, { "epoch": 0.12565047701647875, "grad_norm": 0.6211292743682861, "learning_rate": 9.615914696068475e-05, "loss": 0.0845, "step": 1159 }, { "epoch": 0.12575888985255854, "grad_norm": 0.4503001868724823, "learning_rate": 9.615552009284782e-05, "loss": 0.0671, "step": 1160 }, { "epoch": 0.12586730268863833, "grad_norm": 0.6739040613174438, "learning_rate": 9.615189322501088e-05, "loss": 0.0742, "step": 1161 }, { "epoch": 0.12597571552471812, "grad_norm": 0.6330134272575378, "learning_rate": 9.614826635717395e-05, "loss": 0.0833, "step": 1162 }, { "epoch": 0.1260841283607979, "grad_norm": 0.7808976173400879, "learning_rate": 9.614463948933702e-05, "loss": 0.0501, "step": 1163 }, { "epoch": 0.1261925411968777, "grad_norm": 0.40998101234436035, "learning_rate": 9.614101262150008e-05, "loss": 0.087, "step": 1164 }, { "epoch": 0.1263009540329575, "grad_norm": 0.49272751808166504, "learning_rate": 9.613738575366315e-05, "loss": 0.0603, "step": 1165 }, { "epoch": 0.12640936686903728, "grad_norm": 0.26840364933013916, "learning_rate": 9.61337588858262e-05, "loss": 0.0454, "step": 1166 }, { "epoch": 0.12651777970511707, "grad_norm": 0.7110151052474976, "learning_rate": 9.613013201798928e-05, "loss": 0.0514, "step": 1167 }, { "epoch": 0.1266261925411969, "grad_norm": 0.35713645815849304, "learning_rate": 9.612650515015233e-05, "loss": 0.0707, "step": 1168 }, { "epoch": 0.12673460537727668, "grad_norm": 0.4958434998989105, "learning_rate": 9.61228782823154e-05, "loss": 0.0694, "step": 1169 }, { "epoch": 0.12684301821335647, "grad_norm": 0.3090367913246155, "learning_rate": 9.611925141447846e-05, "loss": 0.0438, "step": 1170 }, { "epoch": 0.12695143104943626, "grad_norm": 0.39219459891319275, "learning_rate": 9.611562454664152e-05, "loss": 0.0409, "step": 1171 }, { "epoch": 0.12705984388551606, "grad_norm": 0.5226238369941711, "learning_rate": 9.611199767880459e-05, "loss": 0.0718, "step": 1172 }, { "epoch": 0.12716825672159585, "grad_norm": 0.3973537087440491, "learning_rate": 9.610837081096765e-05, "loss": 0.0659, "step": 1173 }, { "epoch": 0.12727666955767564, "grad_norm": 0.3895716071128845, "learning_rate": 9.610474394313072e-05, "loss": 0.0466, "step": 1174 }, { "epoch": 0.12738508239375543, "grad_norm": 0.293111115694046, "learning_rate": 9.610111707529378e-05, "loss": 0.0474, "step": 1175 }, { "epoch": 0.12749349522983522, "grad_norm": 0.446269690990448, "learning_rate": 9.609749020745685e-05, "loss": 0.0712, "step": 1176 }, { "epoch": 0.127601908065915, "grad_norm": 0.3032476305961609, "learning_rate": 9.60938633396199e-05, "loss": 0.043, "step": 1177 }, { "epoch": 0.1277103209019948, "grad_norm": 0.8579301834106445, "learning_rate": 9.609023647178297e-05, "loss": 0.0767, "step": 1178 }, { "epoch": 0.1278187337380746, "grad_norm": 0.7708007097244263, "learning_rate": 9.608660960394603e-05, "loss": 0.0658, "step": 1179 }, { "epoch": 0.12792714657415438, "grad_norm": 0.5880667567253113, "learning_rate": 9.60829827361091e-05, "loss": 0.1029, "step": 1180 }, { "epoch": 0.12803555941023417, "grad_norm": 0.3740796148777008, "learning_rate": 9.607935586827217e-05, "loss": 0.0559, "step": 1181 }, { "epoch": 0.12814397224631396, "grad_norm": 0.2880842089653015, "learning_rate": 9.607572900043523e-05, "loss": 0.0361, "step": 1182 }, { "epoch": 0.12825238508239376, "grad_norm": 0.2802221477031708, "learning_rate": 9.60721021325983e-05, "loss": 0.045, "step": 1183 }, { "epoch": 0.12836079791847355, "grad_norm": 0.21193091571331024, "learning_rate": 9.606847526476136e-05, "loss": 0.0255, "step": 1184 }, { "epoch": 0.12846921075455334, "grad_norm": 0.5297073125839233, "learning_rate": 9.606484839692442e-05, "loss": 0.0607, "step": 1185 }, { "epoch": 0.12857762359063313, "grad_norm": 0.3950366675853729, "learning_rate": 9.606122152908749e-05, "loss": 0.0381, "step": 1186 }, { "epoch": 0.12868603642671292, "grad_norm": 0.8230805397033691, "learning_rate": 9.605759466125054e-05, "loss": 0.0668, "step": 1187 }, { "epoch": 0.1287944492627927, "grad_norm": 0.5135222673416138, "learning_rate": 9.605396779341362e-05, "loss": 0.0611, "step": 1188 }, { "epoch": 0.1289028620988725, "grad_norm": 0.6739736795425415, "learning_rate": 9.605034092557667e-05, "loss": 0.1162, "step": 1189 }, { "epoch": 0.1290112749349523, "grad_norm": 0.3432231843471527, "learning_rate": 9.604671405773974e-05, "loss": 0.0443, "step": 1190 }, { "epoch": 0.12911968777103208, "grad_norm": 1.2788175344467163, "learning_rate": 9.60430871899028e-05, "loss": 0.0255, "step": 1191 }, { "epoch": 0.12922810060711187, "grad_norm": 0.3601071834564209, "learning_rate": 9.603946032206587e-05, "loss": 0.0408, "step": 1192 }, { "epoch": 0.12933651344319166, "grad_norm": 0.17279046773910522, "learning_rate": 9.603583345422893e-05, "loss": 0.0177, "step": 1193 }, { "epoch": 0.12944492627927146, "grad_norm": 0.3093559145927429, "learning_rate": 9.603220658639199e-05, "loss": 0.0366, "step": 1194 }, { "epoch": 0.12955333911535125, "grad_norm": 0.5528412461280823, "learning_rate": 9.602857971855506e-05, "loss": 0.0363, "step": 1195 }, { "epoch": 0.12966175195143104, "grad_norm": 1.0343393087387085, "learning_rate": 9.602495285071811e-05, "loss": 0.0457, "step": 1196 }, { "epoch": 0.12977016478751083, "grad_norm": 0.8255594372749329, "learning_rate": 9.602132598288119e-05, "loss": 0.0798, "step": 1197 }, { "epoch": 0.12987857762359065, "grad_norm": 0.3795504570007324, "learning_rate": 9.601769911504426e-05, "loss": 0.032, "step": 1198 }, { "epoch": 0.12998699045967044, "grad_norm": 0.34373024106025696, "learning_rate": 9.601407224720731e-05, "loss": 0.034, "step": 1199 }, { "epoch": 0.13009540329575023, "grad_norm": 0.5035457611083984, "learning_rate": 9.601044537937038e-05, "loss": 0.054, "step": 1200 }, { "epoch": 0.13020381613183002, "grad_norm": 0.4402303993701935, "learning_rate": 9.600681851153344e-05, "loss": 0.0596, "step": 1201 }, { "epoch": 0.1303122289679098, "grad_norm": 0.349561870098114, "learning_rate": 9.600319164369651e-05, "loss": 0.0614, "step": 1202 }, { "epoch": 0.1304206418039896, "grad_norm": 0.5552339553833008, "learning_rate": 9.599956477585957e-05, "loss": 0.0648, "step": 1203 }, { "epoch": 0.1305290546400694, "grad_norm": 0.363908976316452, "learning_rate": 9.599593790802264e-05, "loss": 0.0608, "step": 1204 }, { "epoch": 0.13063746747614918, "grad_norm": 0.8982036113739014, "learning_rate": 9.59923110401857e-05, "loss": 0.0361, "step": 1205 }, { "epoch": 0.13074588031222897, "grad_norm": 0.527721107006073, "learning_rate": 9.598868417234877e-05, "loss": 0.0467, "step": 1206 }, { "epoch": 0.13085429314830876, "grad_norm": 0.5363091230392456, "learning_rate": 9.598505730451183e-05, "loss": 0.1053, "step": 1207 }, { "epoch": 0.13096270598438856, "grad_norm": 0.498750239610672, "learning_rate": 9.598143043667488e-05, "loss": 0.0804, "step": 1208 }, { "epoch": 0.13107111882046835, "grad_norm": 0.7476757764816284, "learning_rate": 9.597780356883796e-05, "loss": 0.0417, "step": 1209 }, { "epoch": 0.13117953165654814, "grad_norm": 0.4144784212112427, "learning_rate": 9.597417670100101e-05, "loss": 0.0384, "step": 1210 }, { "epoch": 0.13128794449262793, "grad_norm": 0.4602672755718231, "learning_rate": 9.597054983316408e-05, "loss": 0.0379, "step": 1211 }, { "epoch": 0.13139635732870772, "grad_norm": 0.35204827785491943, "learning_rate": 9.596692296532714e-05, "loss": 0.0417, "step": 1212 }, { "epoch": 0.1315047701647875, "grad_norm": 0.42149457335472107, "learning_rate": 9.596329609749021e-05, "loss": 0.0738, "step": 1213 }, { "epoch": 0.1316131830008673, "grad_norm": 0.675961971282959, "learning_rate": 9.595966922965327e-05, "loss": 0.0792, "step": 1214 }, { "epoch": 0.1317215958369471, "grad_norm": 0.3462047576904297, "learning_rate": 9.595604236181634e-05, "loss": 0.0668, "step": 1215 }, { "epoch": 0.13183000867302688, "grad_norm": 0.47030436992645264, "learning_rate": 9.595241549397941e-05, "loss": 0.0415, "step": 1216 }, { "epoch": 0.13193842150910667, "grad_norm": 0.5969009399414062, "learning_rate": 9.594878862614247e-05, "loss": 0.0834, "step": 1217 }, { "epoch": 0.13204683434518646, "grad_norm": 0.6370078325271606, "learning_rate": 9.594516175830554e-05, "loss": 0.0598, "step": 1218 }, { "epoch": 0.13215524718126626, "grad_norm": 0.6134319305419922, "learning_rate": 9.59415348904686e-05, "loss": 0.07, "step": 1219 }, { "epoch": 0.13226366001734605, "grad_norm": 0.2707040309906006, "learning_rate": 9.593790802263167e-05, "loss": 0.0364, "step": 1220 }, { "epoch": 0.13237207285342584, "grad_norm": 0.18400220572948456, "learning_rate": 9.593428115479472e-05, "loss": 0.0374, "step": 1221 }, { "epoch": 0.13248048568950563, "grad_norm": 0.5761657953262329, "learning_rate": 9.59306542869578e-05, "loss": 0.0583, "step": 1222 }, { "epoch": 0.13258889852558542, "grad_norm": 0.7344638705253601, "learning_rate": 9.592702741912085e-05, "loss": 0.0682, "step": 1223 }, { "epoch": 0.1326973113616652, "grad_norm": 0.6177888512611389, "learning_rate": 9.592340055128391e-05, "loss": 0.035, "step": 1224 }, { "epoch": 0.132805724197745, "grad_norm": 0.33119484782218933, "learning_rate": 9.591977368344698e-05, "loss": 0.0447, "step": 1225 }, { "epoch": 0.1329141370338248, "grad_norm": 0.6532973647117615, "learning_rate": 9.591614681561004e-05, "loss": 0.0779, "step": 1226 }, { "epoch": 0.1330225498699046, "grad_norm": 0.34535303711891174, "learning_rate": 9.591251994777311e-05, "loss": 0.0297, "step": 1227 }, { "epoch": 0.1331309627059844, "grad_norm": 0.32436808943748474, "learning_rate": 9.590889307993617e-05, "loss": 0.0579, "step": 1228 }, { "epoch": 0.1332393755420642, "grad_norm": 0.11228697746992111, "learning_rate": 9.590526621209924e-05, "loss": 0.0152, "step": 1229 }, { "epoch": 0.13334778837814398, "grad_norm": 0.870183527469635, "learning_rate": 9.59016393442623e-05, "loss": 0.0648, "step": 1230 }, { "epoch": 0.13345620121422377, "grad_norm": 0.46509164571762085, "learning_rate": 9.589801247642535e-05, "loss": 0.0752, "step": 1231 }, { "epoch": 0.13356461405030357, "grad_norm": 0.5498208403587341, "learning_rate": 9.589438560858844e-05, "loss": 0.0669, "step": 1232 }, { "epoch": 0.13367302688638336, "grad_norm": 0.7491757869720459, "learning_rate": 9.58907587407515e-05, "loss": 0.0687, "step": 1233 }, { "epoch": 0.13378143972246315, "grad_norm": 0.5180199146270752, "learning_rate": 9.588713187291456e-05, "loss": 0.077, "step": 1234 }, { "epoch": 0.13388985255854294, "grad_norm": 0.4257039725780487, "learning_rate": 9.588350500507762e-05, "loss": 0.0675, "step": 1235 }, { "epoch": 0.13399826539462273, "grad_norm": 0.21727582812309265, "learning_rate": 9.587987813724069e-05, "loss": 0.0307, "step": 1236 }, { "epoch": 0.13410667823070252, "grad_norm": 0.5025773644447327, "learning_rate": 9.587625126940375e-05, "loss": 0.065, "step": 1237 }, { "epoch": 0.1342150910667823, "grad_norm": 0.7487730383872986, "learning_rate": 9.587262440156681e-05, "loss": 0.0858, "step": 1238 }, { "epoch": 0.1343235039028621, "grad_norm": 0.3483811020851135, "learning_rate": 9.586899753372988e-05, "loss": 0.0667, "step": 1239 }, { "epoch": 0.1344319167389419, "grad_norm": 0.4406965672969818, "learning_rate": 9.586537066589294e-05, "loss": 0.0641, "step": 1240 }, { "epoch": 0.13454032957502168, "grad_norm": 0.4858853816986084, "learning_rate": 9.5861743798056e-05, "loss": 0.0759, "step": 1241 }, { "epoch": 0.13464874241110147, "grad_norm": 0.24395979940891266, "learning_rate": 9.585811693021906e-05, "loss": 0.0236, "step": 1242 }, { "epoch": 0.13475715524718127, "grad_norm": 0.4374445080757141, "learning_rate": 9.585449006238213e-05, "loss": 0.0355, "step": 1243 }, { "epoch": 0.13486556808326106, "grad_norm": 0.5723076462745667, "learning_rate": 9.585086319454519e-05, "loss": 0.0745, "step": 1244 }, { "epoch": 0.13497398091934085, "grad_norm": 0.31408777832984924, "learning_rate": 9.584723632670826e-05, "loss": 0.072, "step": 1245 }, { "epoch": 0.13508239375542064, "grad_norm": 0.3335530757904053, "learning_rate": 9.584360945887132e-05, "loss": 0.059, "step": 1246 }, { "epoch": 0.13519080659150043, "grad_norm": 0.46966230869293213, "learning_rate": 9.583998259103438e-05, "loss": 0.0392, "step": 1247 }, { "epoch": 0.13529921942758022, "grad_norm": 0.6355646252632141, "learning_rate": 9.583635572319745e-05, "loss": 0.0615, "step": 1248 }, { "epoch": 0.13540763226366, "grad_norm": 0.44642746448516846, "learning_rate": 9.583272885536052e-05, "loss": 0.0436, "step": 1249 }, { "epoch": 0.1355160450997398, "grad_norm": 0.5792426466941833, "learning_rate": 9.582910198752359e-05, "loss": 0.1113, "step": 1250 }, { "epoch": 0.1356244579358196, "grad_norm": 0.5328561067581177, "learning_rate": 9.582547511968665e-05, "loss": 0.0527, "step": 1251 }, { "epoch": 0.13573287077189938, "grad_norm": 0.6370123028755188, "learning_rate": 9.58218482518497e-05, "loss": 0.083, "step": 1252 }, { "epoch": 0.13584128360797917, "grad_norm": 0.38879576325416565, "learning_rate": 9.581822138401278e-05, "loss": 0.0722, "step": 1253 }, { "epoch": 0.13594969644405897, "grad_norm": 0.41254448890686035, "learning_rate": 9.581459451617583e-05, "loss": 0.0615, "step": 1254 }, { "epoch": 0.13605810928013876, "grad_norm": 0.36497166752815247, "learning_rate": 9.58109676483389e-05, "loss": 0.0621, "step": 1255 }, { "epoch": 0.13616652211621855, "grad_norm": 0.43743112683296204, "learning_rate": 9.580734078050196e-05, "loss": 0.0669, "step": 1256 }, { "epoch": 0.13627493495229837, "grad_norm": 0.45158714056015015, "learning_rate": 9.580371391266503e-05, "loss": 0.0695, "step": 1257 }, { "epoch": 0.13638334778837816, "grad_norm": 0.2229820191860199, "learning_rate": 9.580008704482809e-05, "loss": 0.024, "step": 1258 }, { "epoch": 0.13649176062445795, "grad_norm": 0.32688108086586, "learning_rate": 9.579646017699116e-05, "loss": 0.0476, "step": 1259 }, { "epoch": 0.13660017346053774, "grad_norm": 0.21876952052116394, "learning_rate": 9.579283330915422e-05, "loss": 0.0343, "step": 1260 }, { "epoch": 0.13670858629661753, "grad_norm": 0.31202512979507446, "learning_rate": 9.578920644131727e-05, "loss": 0.0532, "step": 1261 }, { "epoch": 0.13681699913269732, "grad_norm": 0.8256633877754211, "learning_rate": 9.578557957348035e-05, "loss": 0.1094, "step": 1262 }, { "epoch": 0.1369254119687771, "grad_norm": 0.4062594473361969, "learning_rate": 9.57819527056434e-05, "loss": 0.0478, "step": 1263 }, { "epoch": 0.1370338248048569, "grad_norm": 0.24568818509578705, "learning_rate": 9.577832583780647e-05, "loss": 0.0348, "step": 1264 }, { "epoch": 0.1371422376409367, "grad_norm": 0.4583929181098938, "learning_rate": 9.577469896996953e-05, "loss": 0.0829, "step": 1265 }, { "epoch": 0.13725065047701648, "grad_norm": 0.4696405827999115, "learning_rate": 9.57710721021326e-05, "loss": 0.064, "step": 1266 }, { "epoch": 0.13735906331309627, "grad_norm": 0.7320578098297119, "learning_rate": 9.576744523429567e-05, "loss": 0.0715, "step": 1267 }, { "epoch": 0.13746747614917607, "grad_norm": 0.387717068195343, "learning_rate": 9.576381836645873e-05, "loss": 0.0519, "step": 1268 }, { "epoch": 0.13757588898525586, "grad_norm": 0.19843441247940063, "learning_rate": 9.57601914986218e-05, "loss": 0.0283, "step": 1269 }, { "epoch": 0.13768430182133565, "grad_norm": 0.4005407691001892, "learning_rate": 9.575656463078486e-05, "loss": 0.0375, "step": 1270 }, { "epoch": 0.13779271465741544, "grad_norm": 0.5356683731079102, "learning_rate": 9.575293776294793e-05, "loss": 0.0716, "step": 1271 }, { "epoch": 0.13790112749349523, "grad_norm": 0.5216037034988403, "learning_rate": 9.574931089511099e-05, "loss": 0.0949, "step": 1272 }, { "epoch": 0.13800954032957502, "grad_norm": 0.5377545356750488, "learning_rate": 9.574568402727406e-05, "loss": 0.0834, "step": 1273 }, { "epoch": 0.1381179531656548, "grad_norm": 0.5542801022529602, "learning_rate": 9.574205715943712e-05, "loss": 0.0673, "step": 1274 }, { "epoch": 0.1382263660017346, "grad_norm": 0.3287280797958374, "learning_rate": 9.573843029160017e-05, "loss": 0.0357, "step": 1275 }, { "epoch": 0.1383347788378144, "grad_norm": 0.2556573450565338, "learning_rate": 9.573480342376324e-05, "loss": 0.0283, "step": 1276 }, { "epoch": 0.13844319167389418, "grad_norm": 0.33271223306655884, "learning_rate": 9.57311765559263e-05, "loss": 0.0378, "step": 1277 }, { "epoch": 0.13855160450997397, "grad_norm": 0.458473801612854, "learning_rate": 9.572754968808937e-05, "loss": 0.0619, "step": 1278 }, { "epoch": 0.13866001734605377, "grad_norm": 0.824306309223175, "learning_rate": 9.572392282025243e-05, "loss": 0.1113, "step": 1279 }, { "epoch": 0.13876843018213356, "grad_norm": 0.4162353575229645, "learning_rate": 9.57202959524155e-05, "loss": 0.0582, "step": 1280 }, { "epoch": 0.13887684301821335, "grad_norm": 0.34343624114990234, "learning_rate": 9.571666908457856e-05, "loss": 0.0376, "step": 1281 }, { "epoch": 0.13898525585429314, "grad_norm": 0.37254148721694946, "learning_rate": 9.571304221674163e-05, "loss": 0.0574, "step": 1282 }, { "epoch": 0.13909366869037293, "grad_norm": 0.2847936451435089, "learning_rate": 9.570941534890469e-05, "loss": 0.0383, "step": 1283 }, { "epoch": 0.13920208152645272, "grad_norm": 0.4064705967903137, "learning_rate": 9.570578848106776e-05, "loss": 0.0267, "step": 1284 }, { "epoch": 0.1393104943625325, "grad_norm": 0.2821447253227234, "learning_rate": 9.570216161323083e-05, "loss": 0.0492, "step": 1285 }, { "epoch": 0.1394189071986123, "grad_norm": 0.5394465923309326, "learning_rate": 9.569853474539388e-05, "loss": 0.0558, "step": 1286 }, { "epoch": 0.13952732003469212, "grad_norm": 0.5257329940795898, "learning_rate": 9.569490787755696e-05, "loss": 0.0494, "step": 1287 }, { "epoch": 0.1396357328707719, "grad_norm": 0.34042149782180786, "learning_rate": 9.569128100972001e-05, "loss": 0.0451, "step": 1288 }, { "epoch": 0.1397441457068517, "grad_norm": 0.5714598298072815, "learning_rate": 9.568765414188307e-05, "loss": 0.0393, "step": 1289 }, { "epoch": 0.1398525585429315, "grad_norm": 0.5229496955871582, "learning_rate": 9.568402727404614e-05, "loss": 0.0546, "step": 1290 }, { "epoch": 0.13996097137901128, "grad_norm": 0.48946869373321533, "learning_rate": 9.56804004062092e-05, "loss": 0.0286, "step": 1291 }, { "epoch": 0.14006938421509108, "grad_norm": 0.3668719232082367, "learning_rate": 9.567677353837227e-05, "loss": 0.0708, "step": 1292 }, { "epoch": 0.14017779705117087, "grad_norm": 0.659062922000885, "learning_rate": 9.567314667053533e-05, "loss": 0.0945, "step": 1293 }, { "epoch": 0.14028620988725066, "grad_norm": 0.8438855409622192, "learning_rate": 9.56695198026984e-05, "loss": 0.0728, "step": 1294 }, { "epoch": 0.14039462272333045, "grad_norm": 0.5448653697967529, "learning_rate": 9.566589293486145e-05, "loss": 0.0772, "step": 1295 }, { "epoch": 0.14050303555941024, "grad_norm": 0.781132698059082, "learning_rate": 9.566226606702453e-05, "loss": 0.0523, "step": 1296 }, { "epoch": 0.14061144839549003, "grad_norm": 0.5782954096794128, "learning_rate": 9.565863919918758e-05, "loss": 0.1191, "step": 1297 }, { "epoch": 0.14071986123156982, "grad_norm": 0.659436821937561, "learning_rate": 9.565501233135064e-05, "loss": 0.072, "step": 1298 }, { "epoch": 0.1408282740676496, "grad_norm": 0.6154519319534302, "learning_rate": 9.565138546351371e-05, "loss": 0.0665, "step": 1299 }, { "epoch": 0.1409366869037294, "grad_norm": 0.4577277898788452, "learning_rate": 9.564775859567677e-05, "loss": 0.0402, "step": 1300 }, { "epoch": 0.1410450997398092, "grad_norm": 0.7377919554710388, "learning_rate": 9.564413172783985e-05, "loss": 0.0503, "step": 1301 }, { "epoch": 0.14115351257588898, "grad_norm": 0.5277714729309082, "learning_rate": 9.564050486000291e-05, "loss": 0.0506, "step": 1302 }, { "epoch": 0.14126192541196878, "grad_norm": 0.8885176181793213, "learning_rate": 9.563687799216597e-05, "loss": 0.0573, "step": 1303 }, { "epoch": 0.14137033824804857, "grad_norm": 0.3927386999130249, "learning_rate": 9.563325112432904e-05, "loss": 0.0423, "step": 1304 }, { "epoch": 0.14147875108412836, "grad_norm": 0.3000117540359497, "learning_rate": 9.56296242564921e-05, "loss": 0.0481, "step": 1305 }, { "epoch": 0.14158716392020815, "grad_norm": 0.47545531392097473, "learning_rate": 9.562599738865517e-05, "loss": 0.047, "step": 1306 }, { "epoch": 0.14169557675628794, "grad_norm": 0.6635836362838745, "learning_rate": 9.562237052081822e-05, "loss": 0.0823, "step": 1307 }, { "epoch": 0.14180398959236773, "grad_norm": 0.36036038398742676, "learning_rate": 9.56187436529813e-05, "loss": 0.0261, "step": 1308 }, { "epoch": 0.14191240242844752, "grad_norm": 0.30975475907325745, "learning_rate": 9.561511678514435e-05, "loss": 0.0358, "step": 1309 }, { "epoch": 0.1420208152645273, "grad_norm": 0.2665340006351471, "learning_rate": 9.561148991730742e-05, "loss": 0.0412, "step": 1310 }, { "epoch": 0.1421292281006071, "grad_norm": 0.4613702893257141, "learning_rate": 9.560786304947048e-05, "loss": 0.0543, "step": 1311 }, { "epoch": 0.1422376409366869, "grad_norm": 0.40219053626060486, "learning_rate": 9.560423618163354e-05, "loss": 0.0363, "step": 1312 }, { "epoch": 0.14234605377276668, "grad_norm": 0.31459948420524597, "learning_rate": 9.560060931379661e-05, "loss": 0.0319, "step": 1313 }, { "epoch": 0.14245446660884648, "grad_norm": 0.663811206817627, "learning_rate": 9.559698244595967e-05, "loss": 0.0962, "step": 1314 }, { "epoch": 0.14256287944492627, "grad_norm": 0.33005788922309875, "learning_rate": 9.559335557812274e-05, "loss": 0.0443, "step": 1315 }, { "epoch": 0.14267129228100608, "grad_norm": 0.9027496576309204, "learning_rate": 9.55897287102858e-05, "loss": 0.1154, "step": 1316 }, { "epoch": 0.14277970511708588, "grad_norm": 0.298437237739563, "learning_rate": 9.558610184244886e-05, "loss": 0.0307, "step": 1317 }, { "epoch": 0.14288811795316567, "grad_norm": 0.23619206249713898, "learning_rate": 9.558247497461194e-05, "loss": 0.0345, "step": 1318 }, { "epoch": 0.14299653078924546, "grad_norm": 0.5800207853317261, "learning_rate": 9.557884810677499e-05, "loss": 0.0531, "step": 1319 }, { "epoch": 0.14310494362532525, "grad_norm": 0.7626429796218872, "learning_rate": 9.557522123893806e-05, "loss": 0.0614, "step": 1320 }, { "epoch": 0.14321335646140504, "grad_norm": 0.2945078909397125, "learning_rate": 9.557159437110112e-05, "loss": 0.0631, "step": 1321 }, { "epoch": 0.14332176929748483, "grad_norm": 0.4350839853286743, "learning_rate": 9.556796750326419e-05, "loss": 0.0568, "step": 1322 }, { "epoch": 0.14343018213356462, "grad_norm": 0.7993150353431702, "learning_rate": 9.556434063542725e-05, "loss": 0.0802, "step": 1323 }, { "epoch": 0.1435385949696444, "grad_norm": 0.4413595199584961, "learning_rate": 9.556071376759032e-05, "loss": 0.0443, "step": 1324 }, { "epoch": 0.1436470078057242, "grad_norm": 0.7501711845397949, "learning_rate": 9.555708689975338e-05, "loss": 0.0576, "step": 1325 }, { "epoch": 0.143755420641804, "grad_norm": 0.2671109437942505, "learning_rate": 9.555346003191645e-05, "loss": 0.0455, "step": 1326 }, { "epoch": 0.14386383347788378, "grad_norm": 0.3708999454975128, "learning_rate": 9.55498331640795e-05, "loss": 0.0414, "step": 1327 }, { "epoch": 0.14397224631396358, "grad_norm": 0.46768835186958313, "learning_rate": 9.554620629624256e-05, "loss": 0.0716, "step": 1328 }, { "epoch": 0.14408065915004337, "grad_norm": 0.26567694544792175, "learning_rate": 9.554257942840563e-05, "loss": 0.0392, "step": 1329 }, { "epoch": 0.14418907198612316, "grad_norm": 0.48286449909210205, "learning_rate": 9.553895256056869e-05, "loss": 0.0562, "step": 1330 }, { "epoch": 0.14429748482220295, "grad_norm": 0.5072507858276367, "learning_rate": 9.553532569273176e-05, "loss": 0.1087, "step": 1331 }, { "epoch": 0.14440589765828274, "grad_norm": 0.3791542947292328, "learning_rate": 9.553169882489482e-05, "loss": 0.0529, "step": 1332 }, { "epoch": 0.14451431049436253, "grad_norm": 0.3028676509857178, "learning_rate": 9.552807195705789e-05, "loss": 0.0324, "step": 1333 }, { "epoch": 0.14462272333044232, "grad_norm": 0.3846468925476074, "learning_rate": 9.552444508922095e-05, "loss": 0.0646, "step": 1334 }, { "epoch": 0.1447311361665221, "grad_norm": 0.293069988489151, "learning_rate": 9.5520818221384e-05, "loss": 0.0344, "step": 1335 }, { "epoch": 0.1448395490026019, "grad_norm": 0.20472249388694763, "learning_rate": 9.551719135354709e-05, "loss": 0.0348, "step": 1336 }, { "epoch": 0.1449479618386817, "grad_norm": 0.6099933981895447, "learning_rate": 9.551356448571015e-05, "loss": 0.0961, "step": 1337 }, { "epoch": 0.14505637467476148, "grad_norm": 0.4512979984283447, "learning_rate": 9.550993761787322e-05, "loss": 0.041, "step": 1338 }, { "epoch": 0.14516478751084128, "grad_norm": 0.4677542746067047, "learning_rate": 9.550631075003628e-05, "loss": 0.0685, "step": 1339 }, { "epoch": 0.14527320034692107, "grad_norm": 0.45087194442749023, "learning_rate": 9.550268388219935e-05, "loss": 0.0528, "step": 1340 }, { "epoch": 0.14538161318300086, "grad_norm": 0.3354012966156006, "learning_rate": 9.54990570143624e-05, "loss": 0.0497, "step": 1341 }, { "epoch": 0.14549002601908065, "grad_norm": 0.33477815985679626, "learning_rate": 9.549543014652546e-05, "loss": 0.053, "step": 1342 }, { "epoch": 0.14559843885516044, "grad_norm": 0.4956231415271759, "learning_rate": 9.549180327868853e-05, "loss": 0.0728, "step": 1343 }, { "epoch": 0.14570685169124023, "grad_norm": 0.4538578391075134, "learning_rate": 9.548817641085159e-05, "loss": 0.0552, "step": 1344 }, { "epoch": 0.14581526452732002, "grad_norm": 0.26172780990600586, "learning_rate": 9.548454954301466e-05, "loss": 0.0549, "step": 1345 }, { "epoch": 0.14592367736339984, "grad_norm": 0.8753830790519714, "learning_rate": 9.548092267517772e-05, "loss": 0.09, "step": 1346 }, { "epoch": 0.14603209019947963, "grad_norm": 0.6490553021430969, "learning_rate": 9.547729580734079e-05, "loss": 0.0244, "step": 1347 }, { "epoch": 0.14614050303555942, "grad_norm": 0.41384607553482056, "learning_rate": 9.547366893950385e-05, "loss": 0.031, "step": 1348 }, { "epoch": 0.1462489158716392, "grad_norm": 0.23539981245994568, "learning_rate": 9.547004207166692e-05, "loss": 0.0237, "step": 1349 }, { "epoch": 0.146357328707719, "grad_norm": 0.2573741376399994, "learning_rate": 9.546641520382997e-05, "loss": 0.0271, "step": 1350 }, { "epoch": 0.1464657415437988, "grad_norm": 0.5606499314308167, "learning_rate": 9.546278833599303e-05, "loss": 0.0702, "step": 1351 }, { "epoch": 0.14657415437987859, "grad_norm": 0.2380709946155548, "learning_rate": 9.54591614681561e-05, "loss": 0.0549, "step": 1352 }, { "epoch": 0.14668256721595838, "grad_norm": 0.1850307583808899, "learning_rate": 9.545553460031917e-05, "loss": 0.0162, "step": 1353 }, { "epoch": 0.14679098005203817, "grad_norm": 0.33358240127563477, "learning_rate": 9.545190773248224e-05, "loss": 0.0511, "step": 1354 }, { "epoch": 0.14689939288811796, "grad_norm": 0.475469708442688, "learning_rate": 9.54482808646453e-05, "loss": 0.079, "step": 1355 }, { "epoch": 0.14700780572419775, "grad_norm": 0.6480074524879456, "learning_rate": 9.544465399680836e-05, "loss": 0.0725, "step": 1356 }, { "epoch": 0.14711621856027754, "grad_norm": 0.3656986653804779, "learning_rate": 9.544102712897143e-05, "loss": 0.0212, "step": 1357 }, { "epoch": 0.14722463139635733, "grad_norm": 0.40748587250709534, "learning_rate": 9.543740026113449e-05, "loss": 0.0482, "step": 1358 }, { "epoch": 0.14733304423243712, "grad_norm": 0.609171986579895, "learning_rate": 9.543377339329756e-05, "loss": 0.0634, "step": 1359 }, { "epoch": 0.1474414570685169, "grad_norm": 0.7126306295394897, "learning_rate": 9.543014652546061e-05, "loss": 0.0818, "step": 1360 }, { "epoch": 0.1475498699045967, "grad_norm": 0.21463842689990997, "learning_rate": 9.542651965762369e-05, "loss": 0.0341, "step": 1361 }, { "epoch": 0.1476582827406765, "grad_norm": 0.5000882148742676, "learning_rate": 9.542289278978674e-05, "loss": 0.0539, "step": 1362 }, { "epoch": 0.14776669557675629, "grad_norm": 1.1101163625717163, "learning_rate": 9.541926592194981e-05, "loss": 0.0469, "step": 1363 }, { "epoch": 0.14787510841283608, "grad_norm": 0.8564966917037964, "learning_rate": 9.541563905411287e-05, "loss": 0.0962, "step": 1364 }, { "epoch": 0.14798352124891587, "grad_norm": 0.37970802187919617, "learning_rate": 9.541201218627593e-05, "loss": 0.0277, "step": 1365 }, { "epoch": 0.14809193408499566, "grad_norm": 0.3622520864009857, "learning_rate": 9.5408385318439e-05, "loss": 0.0534, "step": 1366 }, { "epoch": 0.14820034692107545, "grad_norm": 0.5071144104003906, "learning_rate": 9.540475845060206e-05, "loss": 0.0415, "step": 1367 }, { "epoch": 0.14830875975715524, "grad_norm": 0.3760180175304413, "learning_rate": 9.540113158276513e-05, "loss": 0.0491, "step": 1368 }, { "epoch": 0.14841717259323503, "grad_norm": 0.7546484470367432, "learning_rate": 9.539750471492818e-05, "loss": 0.0734, "step": 1369 }, { "epoch": 0.14852558542931482, "grad_norm": 0.6264544129371643, "learning_rate": 9.539387784709126e-05, "loss": 0.085, "step": 1370 }, { "epoch": 0.1486339982653946, "grad_norm": 0.4021918475627899, "learning_rate": 9.539025097925433e-05, "loss": 0.0387, "step": 1371 }, { "epoch": 0.1487424111014744, "grad_norm": 0.45272091031074524, "learning_rate": 9.538662411141738e-05, "loss": 0.0581, "step": 1372 }, { "epoch": 0.1488508239375542, "grad_norm": 0.3352506756782532, "learning_rate": 9.538299724358045e-05, "loss": 0.0272, "step": 1373 }, { "epoch": 0.14895923677363399, "grad_norm": 0.5296693444252014, "learning_rate": 9.537937037574351e-05, "loss": 0.0452, "step": 1374 }, { "epoch": 0.1490676496097138, "grad_norm": 0.16752566397190094, "learning_rate": 9.537574350790658e-05, "loss": 0.0233, "step": 1375 }, { "epoch": 0.1491760624457936, "grad_norm": 0.3537997305393219, "learning_rate": 9.537211664006964e-05, "loss": 0.0494, "step": 1376 }, { "epoch": 0.14928447528187339, "grad_norm": 1.2500019073486328, "learning_rate": 9.536848977223271e-05, "loss": 0.069, "step": 1377 }, { "epoch": 0.14939288811795318, "grad_norm": 0.19117988646030426, "learning_rate": 9.536486290439577e-05, "loss": 0.0225, "step": 1378 }, { "epoch": 0.14950130095403297, "grad_norm": 0.301390677690506, "learning_rate": 9.536123603655883e-05, "loss": 0.0347, "step": 1379 }, { "epoch": 0.14960971379011276, "grad_norm": 0.46149447560310364, "learning_rate": 9.53576091687219e-05, "loss": 0.066, "step": 1380 }, { "epoch": 0.14971812662619255, "grad_norm": 0.560149610042572, "learning_rate": 9.535398230088495e-05, "loss": 0.0725, "step": 1381 }, { "epoch": 0.14982653946227234, "grad_norm": 0.1496129035949707, "learning_rate": 9.535035543304802e-05, "loss": 0.0163, "step": 1382 }, { "epoch": 0.14993495229835213, "grad_norm": 0.4102928340435028, "learning_rate": 9.534672856521108e-05, "loss": 0.054, "step": 1383 }, { "epoch": 0.15004336513443192, "grad_norm": 0.4762820899486542, "learning_rate": 9.534310169737415e-05, "loss": 0.0591, "step": 1384 }, { "epoch": 0.1501517779705117, "grad_norm": 0.44790491461753845, "learning_rate": 9.533947482953721e-05, "loss": 0.0349, "step": 1385 }, { "epoch": 0.1502601908065915, "grad_norm": 0.43596768379211426, "learning_rate": 9.533584796170028e-05, "loss": 0.0397, "step": 1386 }, { "epoch": 0.1503686036426713, "grad_norm": 0.35297468304634094, "learning_rate": 9.533222109386334e-05, "loss": 0.055, "step": 1387 }, { "epoch": 0.15047701647875109, "grad_norm": 0.8118528127670288, "learning_rate": 9.532859422602641e-05, "loss": 0.0936, "step": 1388 }, { "epoch": 0.15058542931483088, "grad_norm": 0.07744184136390686, "learning_rate": 9.532496735818948e-05, "loss": 0.0063, "step": 1389 }, { "epoch": 0.15069384215091067, "grad_norm": 0.2861754298210144, "learning_rate": 9.532134049035254e-05, "loss": 0.0234, "step": 1390 }, { "epoch": 0.15080225498699046, "grad_norm": 0.4514746069908142, "learning_rate": 9.531771362251561e-05, "loss": 0.0333, "step": 1391 }, { "epoch": 0.15091066782307025, "grad_norm": 0.5310697555541992, "learning_rate": 9.531408675467867e-05, "loss": 0.0779, "step": 1392 }, { "epoch": 0.15101908065915004, "grad_norm": 0.3118395209312439, "learning_rate": 9.531045988684172e-05, "loss": 0.0531, "step": 1393 }, { "epoch": 0.15112749349522983, "grad_norm": 0.5869510173797607, "learning_rate": 9.53068330190048e-05, "loss": 0.0572, "step": 1394 }, { "epoch": 0.15123590633130962, "grad_norm": 0.37360361218452454, "learning_rate": 9.530320615116785e-05, "loss": 0.0488, "step": 1395 }, { "epoch": 0.1513443191673894, "grad_norm": 0.24995121359825134, "learning_rate": 9.529957928333092e-05, "loss": 0.0618, "step": 1396 }, { "epoch": 0.1514527320034692, "grad_norm": 0.36885714530944824, "learning_rate": 9.529595241549398e-05, "loss": 0.0257, "step": 1397 }, { "epoch": 0.151561144839549, "grad_norm": 0.48295697569847107, "learning_rate": 9.529232554765705e-05, "loss": 0.0923, "step": 1398 }, { "epoch": 0.15166955767562879, "grad_norm": 0.33724483847618103, "learning_rate": 9.528869867982011e-05, "loss": 0.0425, "step": 1399 }, { "epoch": 0.15177797051170858, "grad_norm": 0.5600839853286743, "learning_rate": 9.528507181198318e-05, "loss": 0.0591, "step": 1400 }, { "epoch": 0.15188638334778837, "grad_norm": 0.5269072651863098, "learning_rate": 9.528144494414624e-05, "loss": 0.0879, "step": 1401 }, { "epoch": 0.15199479618386816, "grad_norm": 0.46257102489471436, "learning_rate": 9.52778180763093e-05, "loss": 0.0394, "step": 1402 }, { "epoch": 0.15210320901994795, "grad_norm": 0.32932010293006897, "learning_rate": 9.527419120847236e-05, "loss": 0.0375, "step": 1403 }, { "epoch": 0.15221162185602774, "grad_norm": 0.4264088273048401, "learning_rate": 9.527056434063542e-05, "loss": 0.0712, "step": 1404 }, { "epoch": 0.15232003469210756, "grad_norm": 0.40528959035873413, "learning_rate": 9.52669374727985e-05, "loss": 0.0336, "step": 1405 }, { "epoch": 0.15242844752818735, "grad_norm": 0.6963746547698975, "learning_rate": 9.526331060496156e-05, "loss": 0.0531, "step": 1406 }, { "epoch": 0.15253686036426714, "grad_norm": 0.3229958713054657, "learning_rate": 9.525968373712463e-05, "loss": 0.0256, "step": 1407 }, { "epoch": 0.15264527320034693, "grad_norm": 0.4879133999347687, "learning_rate": 9.525605686928769e-05, "loss": 0.0478, "step": 1408 }, { "epoch": 0.15275368603642672, "grad_norm": 0.4019128382205963, "learning_rate": 9.525243000145075e-05, "loss": 0.0701, "step": 1409 }, { "epoch": 0.1528620988725065, "grad_norm": 0.6161134243011475, "learning_rate": 9.524880313361382e-05, "loss": 0.0801, "step": 1410 }, { "epoch": 0.1529705117085863, "grad_norm": 0.5334380865097046, "learning_rate": 9.524517626577688e-05, "loss": 0.0472, "step": 1411 }, { "epoch": 0.1530789245446661, "grad_norm": 0.2528248429298401, "learning_rate": 9.524154939793995e-05, "loss": 0.041, "step": 1412 }, { "epoch": 0.15318733738074589, "grad_norm": 0.27591273188591003, "learning_rate": 9.5237922530103e-05, "loss": 0.0522, "step": 1413 }, { "epoch": 0.15329575021682568, "grad_norm": 0.2630448639392853, "learning_rate": 9.523429566226608e-05, "loss": 0.0396, "step": 1414 }, { "epoch": 0.15340416305290547, "grad_norm": 0.48943522572517395, "learning_rate": 9.523066879442913e-05, "loss": 0.0435, "step": 1415 }, { "epoch": 0.15351257588898526, "grad_norm": 0.2718656659126282, "learning_rate": 9.522704192659219e-05, "loss": 0.0264, "step": 1416 }, { "epoch": 0.15362098872506505, "grad_norm": 0.8637377023696899, "learning_rate": 9.522341505875526e-05, "loss": 0.0608, "step": 1417 }, { "epoch": 0.15372940156114484, "grad_norm": 0.31759771704673767, "learning_rate": 9.521978819091832e-05, "loss": 0.0783, "step": 1418 }, { "epoch": 0.15383781439722463, "grad_norm": 0.42066383361816406, "learning_rate": 9.521616132308139e-05, "loss": 0.0293, "step": 1419 }, { "epoch": 0.15394622723330442, "grad_norm": 0.47564762830734253, "learning_rate": 9.521253445524445e-05, "loss": 0.0549, "step": 1420 }, { "epoch": 0.1540546400693842, "grad_norm": 0.2927757203578949, "learning_rate": 9.520890758740752e-05, "loss": 0.0479, "step": 1421 }, { "epoch": 0.154163052905464, "grad_norm": 1.0324068069458008, "learning_rate": 9.520528071957059e-05, "loss": 0.0367, "step": 1422 }, { "epoch": 0.1542714657415438, "grad_norm": 0.41889047622680664, "learning_rate": 9.520165385173365e-05, "loss": 0.0629, "step": 1423 }, { "epoch": 0.15437987857762359, "grad_norm": 0.6649196147918701, "learning_rate": 9.519802698389672e-05, "loss": 0.0595, "step": 1424 }, { "epoch": 0.15448829141370338, "grad_norm": 0.5573312640190125, "learning_rate": 9.519440011605977e-05, "loss": 0.0525, "step": 1425 }, { "epoch": 0.15459670424978317, "grad_norm": 0.40047138929367065, "learning_rate": 9.519077324822285e-05, "loss": 0.0589, "step": 1426 }, { "epoch": 0.15470511708586296, "grad_norm": 0.4289607107639313, "learning_rate": 9.51871463803859e-05, "loss": 0.0514, "step": 1427 }, { "epoch": 0.15481352992194275, "grad_norm": 0.38458022475242615, "learning_rate": 9.518351951254897e-05, "loss": 0.0361, "step": 1428 }, { "epoch": 0.15492194275802254, "grad_norm": 0.3483280539512634, "learning_rate": 9.517989264471203e-05, "loss": 0.0371, "step": 1429 }, { "epoch": 0.15503035559410233, "grad_norm": 0.5003224015235901, "learning_rate": 9.51762657768751e-05, "loss": 0.0575, "step": 1430 }, { "epoch": 0.15513876843018212, "grad_norm": 1.7155276536941528, "learning_rate": 9.517263890903816e-05, "loss": 0.1033, "step": 1431 }, { "epoch": 0.1552471812662619, "grad_norm": 0.2909815311431885, "learning_rate": 9.516901204120122e-05, "loss": 0.0578, "step": 1432 }, { "epoch": 0.1553555941023417, "grad_norm": 0.40195217728614807, "learning_rate": 9.516538517336429e-05, "loss": 0.05, "step": 1433 }, { "epoch": 0.15546400693842152, "grad_norm": 0.46327805519104004, "learning_rate": 9.516175830552734e-05, "loss": 0.0306, "step": 1434 }, { "epoch": 0.1555724197745013, "grad_norm": 0.1320207417011261, "learning_rate": 9.515813143769042e-05, "loss": 0.0218, "step": 1435 }, { "epoch": 0.1556808326105811, "grad_norm": 0.6198722720146179, "learning_rate": 9.515450456985347e-05, "loss": 0.0355, "step": 1436 }, { "epoch": 0.1557892454466609, "grad_norm": 0.33038198947906494, "learning_rate": 9.515087770201654e-05, "loss": 0.0299, "step": 1437 }, { "epoch": 0.1558976582827407, "grad_norm": 0.37934666872024536, "learning_rate": 9.51472508341796e-05, "loss": 0.0408, "step": 1438 }, { "epoch": 0.15600607111882048, "grad_norm": 0.48958009481430054, "learning_rate": 9.514362396634267e-05, "loss": 0.037, "step": 1439 }, { "epoch": 0.15611448395490027, "grad_norm": 0.365776389837265, "learning_rate": 9.513999709850574e-05, "loss": 0.0445, "step": 1440 }, { "epoch": 0.15622289679098006, "grad_norm": 0.7802051901817322, "learning_rate": 9.51363702306688e-05, "loss": 0.0884, "step": 1441 }, { "epoch": 0.15633130962705985, "grad_norm": 0.24273881316184998, "learning_rate": 9.513274336283187e-05, "loss": 0.0386, "step": 1442 }, { "epoch": 0.15643972246313964, "grad_norm": 0.19777856767177582, "learning_rate": 9.512911649499493e-05, "loss": 0.0187, "step": 1443 }, { "epoch": 0.15654813529921943, "grad_norm": 0.6213618516921997, "learning_rate": 9.5125489627158e-05, "loss": 0.0564, "step": 1444 }, { "epoch": 0.15665654813529922, "grad_norm": 0.476057767868042, "learning_rate": 9.512186275932106e-05, "loss": 0.0478, "step": 1445 }, { "epoch": 0.156764960971379, "grad_norm": 0.5006645917892456, "learning_rate": 9.511823589148411e-05, "loss": 0.0643, "step": 1446 }, { "epoch": 0.1568733738074588, "grad_norm": 0.2916336953639984, "learning_rate": 9.511460902364718e-05, "loss": 0.0436, "step": 1447 }, { "epoch": 0.1569817866435386, "grad_norm": 0.3582410216331482, "learning_rate": 9.511098215581024e-05, "loss": 0.0378, "step": 1448 }, { "epoch": 0.1570901994796184, "grad_norm": 0.23052923381328583, "learning_rate": 9.510735528797331e-05, "loss": 0.0255, "step": 1449 }, { "epoch": 0.15719861231569818, "grad_norm": 0.6609007120132446, "learning_rate": 9.510372842013637e-05, "loss": 0.122, "step": 1450 }, { "epoch": 0.15730702515177797, "grad_norm": 0.2877033054828644, "learning_rate": 9.510010155229944e-05, "loss": 0.0306, "step": 1451 }, { "epoch": 0.15741543798785776, "grad_norm": 0.5633376836776733, "learning_rate": 9.50964746844625e-05, "loss": 0.0369, "step": 1452 }, { "epoch": 0.15752385082393755, "grad_norm": 0.42575234174728394, "learning_rate": 9.509284781662557e-05, "loss": 0.045, "step": 1453 }, { "epoch": 0.15763226366001734, "grad_norm": 0.9778749346733093, "learning_rate": 9.508922094878863e-05, "loss": 0.1132, "step": 1454 }, { "epoch": 0.15774067649609713, "grad_norm": 0.3596076965332031, "learning_rate": 9.508559408095168e-05, "loss": 0.0283, "step": 1455 }, { "epoch": 0.15784908933217692, "grad_norm": 0.3099343776702881, "learning_rate": 9.508196721311476e-05, "loss": 0.0301, "step": 1456 }, { "epoch": 0.1579575021682567, "grad_norm": 0.5093673467636108, "learning_rate": 9.507834034527783e-05, "loss": 0.0638, "step": 1457 }, { "epoch": 0.1580659150043365, "grad_norm": 1.108784794807434, "learning_rate": 9.50747134774409e-05, "loss": 0.0288, "step": 1458 }, { "epoch": 0.1581743278404163, "grad_norm": 0.2153153270483017, "learning_rate": 9.507108660960395e-05, "loss": 0.0243, "step": 1459 }, { "epoch": 0.1582827406764961, "grad_norm": 0.3715519607067108, "learning_rate": 9.506745974176701e-05, "loss": 0.0427, "step": 1460 }, { "epoch": 0.15839115351257588, "grad_norm": 0.825432538986206, "learning_rate": 9.506383287393008e-05, "loss": 0.0353, "step": 1461 }, { "epoch": 0.15849956634865567, "grad_norm": 1.5420198440551758, "learning_rate": 9.506020600609314e-05, "loss": 0.0934, "step": 1462 }, { "epoch": 0.15860797918473546, "grad_norm": 0.7082664966583252, "learning_rate": 9.505657913825621e-05, "loss": 0.0964, "step": 1463 }, { "epoch": 0.15871639202081528, "grad_norm": 0.4342944920063019, "learning_rate": 9.505295227041927e-05, "loss": 0.0953, "step": 1464 }, { "epoch": 0.15882480485689507, "grad_norm": 0.5342485904693604, "learning_rate": 9.504932540258234e-05, "loss": 0.0943, "step": 1465 }, { "epoch": 0.15893321769297486, "grad_norm": 0.17211145162582397, "learning_rate": 9.50456985347454e-05, "loss": 0.0216, "step": 1466 }, { "epoch": 0.15904163052905465, "grad_norm": 0.3388943672180176, "learning_rate": 9.504207166690847e-05, "loss": 0.0558, "step": 1467 }, { "epoch": 0.15915004336513444, "grad_norm": 0.5470564961433411, "learning_rate": 9.503844479907152e-05, "loss": 0.08, "step": 1468 }, { "epoch": 0.15925845620121423, "grad_norm": 0.2828742265701294, "learning_rate": 9.503481793123458e-05, "loss": 0.0351, "step": 1469 }, { "epoch": 0.15936686903729402, "grad_norm": 0.3321884572505951, "learning_rate": 9.503119106339765e-05, "loss": 0.0446, "step": 1470 }, { "epoch": 0.15947528187337381, "grad_norm": 0.40556788444519043, "learning_rate": 9.502756419556071e-05, "loss": 0.0637, "step": 1471 }, { "epoch": 0.1595836947094536, "grad_norm": 0.3963378965854645, "learning_rate": 9.502393732772378e-05, "loss": 0.0746, "step": 1472 }, { "epoch": 0.1596921075455334, "grad_norm": 0.181504487991333, "learning_rate": 9.502031045988684e-05, "loss": 0.0186, "step": 1473 }, { "epoch": 0.1598005203816132, "grad_norm": 0.950205385684967, "learning_rate": 9.501668359204991e-05, "loss": 0.0581, "step": 1474 }, { "epoch": 0.15990893321769298, "grad_norm": 0.5250534415245056, "learning_rate": 9.501305672421298e-05, "loss": 0.049, "step": 1475 }, { "epoch": 0.16001734605377277, "grad_norm": 0.2294691652059555, "learning_rate": 9.500942985637604e-05, "loss": 0.0296, "step": 1476 }, { "epoch": 0.16012575888985256, "grad_norm": 0.23640356957912445, "learning_rate": 9.500580298853911e-05, "loss": 0.0429, "step": 1477 }, { "epoch": 0.16023417172593235, "grad_norm": 1.2072938680648804, "learning_rate": 9.500217612070217e-05, "loss": 0.0432, "step": 1478 }, { "epoch": 0.16034258456201214, "grad_norm": 0.2170291692018509, "learning_rate": 9.499854925286524e-05, "loss": 0.0339, "step": 1479 }, { "epoch": 0.16045099739809193, "grad_norm": 0.41862499713897705, "learning_rate": 9.49949223850283e-05, "loss": 0.0412, "step": 1480 }, { "epoch": 0.16055941023417172, "grad_norm": 0.2473115622997284, "learning_rate": 9.499129551719136e-05, "loss": 0.0376, "step": 1481 }, { "epoch": 0.16066782307025151, "grad_norm": 0.30975329875946045, "learning_rate": 9.498766864935442e-05, "loss": 0.0437, "step": 1482 }, { "epoch": 0.1607762359063313, "grad_norm": 0.3620662987232208, "learning_rate": 9.498404178151748e-05, "loss": 0.0744, "step": 1483 }, { "epoch": 0.1608846487424111, "grad_norm": 0.596795916557312, "learning_rate": 9.498041491368055e-05, "loss": 0.0274, "step": 1484 }, { "epoch": 0.1609930615784909, "grad_norm": 0.4898509681224823, "learning_rate": 9.497678804584361e-05, "loss": 0.0594, "step": 1485 }, { "epoch": 0.16110147441457068, "grad_norm": 0.3298223316669464, "learning_rate": 9.497316117800668e-05, "loss": 0.0421, "step": 1486 }, { "epoch": 0.16120988725065047, "grad_norm": 0.2819550633430481, "learning_rate": 9.496953431016974e-05, "loss": 0.0341, "step": 1487 }, { "epoch": 0.16131830008673026, "grad_norm": 0.334166944026947, "learning_rate": 9.49659074423328e-05, "loss": 0.0363, "step": 1488 }, { "epoch": 0.16142671292281005, "grad_norm": 0.7973310351371765, "learning_rate": 9.496228057449586e-05, "loss": 0.0579, "step": 1489 }, { "epoch": 0.16153512575888984, "grad_norm": 0.44213151931762695, "learning_rate": 9.495865370665893e-05, "loss": 0.0568, "step": 1490 }, { "epoch": 0.16164353859496963, "grad_norm": 0.4109446108341217, "learning_rate": 9.4955026838822e-05, "loss": 0.0435, "step": 1491 }, { "epoch": 0.16175195143104942, "grad_norm": 0.7189295291900635, "learning_rate": 9.495139997098506e-05, "loss": 0.0836, "step": 1492 }, { "epoch": 0.16186036426712924, "grad_norm": 0.258068323135376, "learning_rate": 9.494777310314813e-05, "loss": 0.0142, "step": 1493 }, { "epoch": 0.16196877710320903, "grad_norm": 0.5511718988418579, "learning_rate": 9.494414623531119e-05, "loss": 0.1007, "step": 1494 }, { "epoch": 0.16207718993928882, "grad_norm": 1.3523662090301514, "learning_rate": 9.494051936747426e-05, "loss": 0.0382, "step": 1495 }, { "epoch": 0.16218560277536861, "grad_norm": 0.13311147689819336, "learning_rate": 9.493689249963732e-05, "loss": 0.0125, "step": 1496 }, { "epoch": 0.1622940156114484, "grad_norm": 0.6536365151405334, "learning_rate": 9.493326563180038e-05, "loss": 0.0717, "step": 1497 }, { "epoch": 0.1624024284475282, "grad_norm": 0.7201508283615112, "learning_rate": 9.492963876396345e-05, "loss": 0.0481, "step": 1498 }, { "epoch": 0.162510841283608, "grad_norm": 0.20771054923534393, "learning_rate": 9.49260118961265e-05, "loss": 0.0162, "step": 1499 }, { "epoch": 0.16261925411968778, "grad_norm": 0.6322606205940247, "learning_rate": 9.492238502828958e-05, "loss": 0.0807, "step": 1500 }, { "epoch": 0.16272766695576757, "grad_norm": 0.7414185404777527, "learning_rate": 9.491875816045263e-05, "loss": 0.0953, "step": 1501 }, { "epoch": 0.16283607979184736, "grad_norm": 1.6962560415267944, "learning_rate": 9.49151312926157e-05, "loss": 0.0745, "step": 1502 }, { "epoch": 0.16294449262792715, "grad_norm": 0.1941341906785965, "learning_rate": 9.491150442477876e-05, "loss": 0.0139, "step": 1503 }, { "epoch": 0.16305290546400694, "grad_norm": 0.23785236477851868, "learning_rate": 9.490787755694183e-05, "loss": 0.0158, "step": 1504 }, { "epoch": 0.16316131830008673, "grad_norm": 0.863698422908783, "learning_rate": 9.490425068910489e-05, "loss": 0.038, "step": 1505 }, { "epoch": 0.16326973113616652, "grad_norm": 1.0171359777450562, "learning_rate": 9.490062382126795e-05, "loss": 0.0653, "step": 1506 }, { "epoch": 0.16337814397224631, "grad_norm": 0.2217540293931961, "learning_rate": 9.489699695343102e-05, "loss": 0.0166, "step": 1507 }, { "epoch": 0.1634865568083261, "grad_norm": 0.43776875734329224, "learning_rate": 9.489337008559409e-05, "loss": 0.0406, "step": 1508 }, { "epoch": 0.1635949696444059, "grad_norm": 0.11917433887720108, "learning_rate": 9.488974321775716e-05, "loss": 0.0213, "step": 1509 }, { "epoch": 0.1637033824804857, "grad_norm": 0.6341823935508728, "learning_rate": 9.488611634992022e-05, "loss": 0.0365, "step": 1510 }, { "epoch": 0.16381179531656548, "grad_norm": 0.8378426432609558, "learning_rate": 9.488248948208329e-05, "loss": 0.0528, "step": 1511 }, { "epoch": 0.16392020815264527, "grad_norm": 0.583910346031189, "learning_rate": 9.487886261424634e-05, "loss": 0.0407, "step": 1512 }, { "epoch": 0.16402862098872506, "grad_norm": 0.6240293383598328, "learning_rate": 9.48752357464094e-05, "loss": 0.0543, "step": 1513 }, { "epoch": 0.16413703382480485, "grad_norm": 0.31282973289489746, "learning_rate": 9.487160887857247e-05, "loss": 0.0229, "step": 1514 }, { "epoch": 0.16424544666088464, "grad_norm": 0.8052083253860474, "learning_rate": 9.486798201073553e-05, "loss": 0.086, "step": 1515 }, { "epoch": 0.16435385949696443, "grad_norm": 0.6075556874275208, "learning_rate": 9.48643551428986e-05, "loss": 0.0476, "step": 1516 }, { "epoch": 0.16446227233304422, "grad_norm": 0.7858022451400757, "learning_rate": 9.486072827506166e-05, "loss": 0.0868, "step": 1517 }, { "epoch": 0.16457068516912401, "grad_norm": 0.3549613058567047, "learning_rate": 9.485710140722473e-05, "loss": 0.0314, "step": 1518 }, { "epoch": 0.1646790980052038, "grad_norm": 0.4351697266101837, "learning_rate": 9.485347453938779e-05, "loss": 0.0426, "step": 1519 }, { "epoch": 0.1647875108412836, "grad_norm": 0.4363429546356201, "learning_rate": 9.484984767155086e-05, "loss": 0.0424, "step": 1520 }, { "epoch": 0.1648959236773634, "grad_norm": 0.6406906843185425, "learning_rate": 9.484622080371392e-05, "loss": 0.0505, "step": 1521 }, { "epoch": 0.16500433651344318, "grad_norm": 0.6622992753982544, "learning_rate": 9.484259393587697e-05, "loss": 0.0551, "step": 1522 }, { "epoch": 0.165112749349523, "grad_norm": 0.7622795701026917, "learning_rate": 9.483896706804004e-05, "loss": 0.065, "step": 1523 }, { "epoch": 0.1652211621856028, "grad_norm": 0.5220192670822144, "learning_rate": 9.48353402002031e-05, "loss": 0.0762, "step": 1524 }, { "epoch": 0.16532957502168258, "grad_norm": 0.2795522212982178, "learning_rate": 9.483171333236617e-05, "loss": 0.0302, "step": 1525 }, { "epoch": 0.16543798785776237, "grad_norm": 0.2795417010784149, "learning_rate": 9.482808646452924e-05, "loss": 0.0497, "step": 1526 }, { "epoch": 0.16554640069384216, "grad_norm": 0.5665898323059082, "learning_rate": 9.48244595966923e-05, "loss": 0.0395, "step": 1527 }, { "epoch": 0.16565481352992195, "grad_norm": 0.28009942173957825, "learning_rate": 9.482083272885537e-05, "loss": 0.0227, "step": 1528 }, { "epoch": 0.16576322636600174, "grad_norm": 0.7075133919715881, "learning_rate": 9.481720586101843e-05, "loss": 0.0498, "step": 1529 }, { "epoch": 0.16587163920208153, "grad_norm": 0.43796512484550476, "learning_rate": 9.48135789931815e-05, "loss": 0.1021, "step": 1530 }, { "epoch": 0.16598005203816132, "grad_norm": 0.3274950087070465, "learning_rate": 9.480995212534456e-05, "loss": 0.0438, "step": 1531 }, { "epoch": 0.16608846487424112, "grad_norm": 0.6830554604530334, "learning_rate": 9.480632525750763e-05, "loss": 0.0524, "step": 1532 }, { "epoch": 0.1661968777103209, "grad_norm": 0.2747766077518463, "learning_rate": 9.480269838967068e-05, "loss": 0.0272, "step": 1533 }, { "epoch": 0.1663052905464007, "grad_norm": 0.3891919255256653, "learning_rate": 9.479907152183376e-05, "loss": 0.0756, "step": 1534 }, { "epoch": 0.1664137033824805, "grad_norm": 0.5927941799163818, "learning_rate": 9.479544465399681e-05, "loss": 0.0653, "step": 1535 }, { "epoch": 0.16652211621856028, "grad_norm": 1.1291686296463013, "learning_rate": 9.479181778615987e-05, "loss": 0.0269, "step": 1536 }, { "epoch": 0.16663052905464007, "grad_norm": 0.5013661980628967, "learning_rate": 9.478819091832294e-05, "loss": 0.0965, "step": 1537 }, { "epoch": 0.16673894189071986, "grad_norm": 0.5230159759521484, "learning_rate": 9.4784564050486e-05, "loss": 0.0787, "step": 1538 }, { "epoch": 0.16684735472679965, "grad_norm": 0.501700758934021, "learning_rate": 9.478093718264907e-05, "loss": 0.06, "step": 1539 }, { "epoch": 0.16695576756287944, "grad_norm": 0.28285107016563416, "learning_rate": 9.477731031481213e-05, "loss": 0.0457, "step": 1540 }, { "epoch": 0.16706418039895923, "grad_norm": 0.5409829020500183, "learning_rate": 9.47736834469752e-05, "loss": 0.0834, "step": 1541 }, { "epoch": 0.16717259323503902, "grad_norm": 0.2523553669452667, "learning_rate": 9.477005657913825e-05, "loss": 0.0441, "step": 1542 }, { "epoch": 0.16728100607111882, "grad_norm": 0.46573299169540405, "learning_rate": 9.476642971130133e-05, "loss": 0.0596, "step": 1543 }, { "epoch": 0.1673894189071986, "grad_norm": 0.43319621682167053, "learning_rate": 9.47628028434644e-05, "loss": 0.0485, "step": 1544 }, { "epoch": 0.1674978317432784, "grad_norm": 0.2801234722137451, "learning_rate": 9.475917597562745e-05, "loss": 0.0314, "step": 1545 }, { "epoch": 0.1676062445793582, "grad_norm": 0.22000491619110107, "learning_rate": 9.475554910779052e-05, "loss": 0.036, "step": 1546 }, { "epoch": 0.16771465741543798, "grad_norm": 0.187096506357193, "learning_rate": 9.475192223995358e-05, "loss": 0.0225, "step": 1547 }, { "epoch": 0.16782307025151777, "grad_norm": 0.32470521330833435, "learning_rate": 9.474829537211665e-05, "loss": 0.0301, "step": 1548 }, { "epoch": 0.16793148308759756, "grad_norm": 0.4263068735599518, "learning_rate": 9.474466850427971e-05, "loss": 0.0536, "step": 1549 }, { "epoch": 0.16803989592367735, "grad_norm": 1.2310733795166016, "learning_rate": 9.474104163644277e-05, "loss": 0.039, "step": 1550 }, { "epoch": 0.16814830875975714, "grad_norm": 0.5292829275131226, "learning_rate": 9.473741476860584e-05, "loss": 0.0867, "step": 1551 }, { "epoch": 0.16825672159583693, "grad_norm": 0.2715952396392822, "learning_rate": 9.47337879007689e-05, "loss": 0.0345, "step": 1552 }, { "epoch": 0.16836513443191675, "grad_norm": 0.39036309719085693, "learning_rate": 9.473016103293197e-05, "loss": 0.0308, "step": 1553 }, { "epoch": 0.16847354726799654, "grad_norm": 0.6295433640480042, "learning_rate": 9.472653416509502e-05, "loss": 0.0518, "step": 1554 }, { "epoch": 0.16858196010407633, "grad_norm": 0.5718704462051392, "learning_rate": 9.47229072972581e-05, "loss": 0.0534, "step": 1555 }, { "epoch": 0.16869037294015612, "grad_norm": 0.3228178918361664, "learning_rate": 9.471928042942115e-05, "loss": 0.0531, "step": 1556 }, { "epoch": 0.16879878577623592, "grad_norm": 0.15516869723796844, "learning_rate": 9.471565356158422e-05, "loss": 0.0223, "step": 1557 }, { "epoch": 0.1689071986123157, "grad_norm": 0.6124467849731445, "learning_rate": 9.471202669374728e-05, "loss": 0.0624, "step": 1558 }, { "epoch": 0.1690156114483955, "grad_norm": 0.8622007369995117, "learning_rate": 9.470839982591034e-05, "loss": 0.0951, "step": 1559 }, { "epoch": 0.1691240242844753, "grad_norm": 0.2796047627925873, "learning_rate": 9.470477295807342e-05, "loss": 0.0301, "step": 1560 }, { "epoch": 0.16923243712055508, "grad_norm": 0.46068474650382996, "learning_rate": 9.470114609023648e-05, "loss": 0.0762, "step": 1561 }, { "epoch": 0.16934084995663487, "grad_norm": 0.7719026803970337, "learning_rate": 9.469751922239955e-05, "loss": 0.0825, "step": 1562 }, { "epoch": 0.16944926279271466, "grad_norm": 0.4014297425746918, "learning_rate": 9.469389235456261e-05, "loss": 0.0638, "step": 1563 }, { "epoch": 0.16955767562879445, "grad_norm": 0.38381022214889526, "learning_rate": 9.469026548672566e-05, "loss": 0.0294, "step": 1564 }, { "epoch": 0.16966608846487424, "grad_norm": 0.4201887845993042, "learning_rate": 9.468663861888874e-05, "loss": 0.0443, "step": 1565 }, { "epoch": 0.16977450130095403, "grad_norm": 0.47182953357696533, "learning_rate": 9.468301175105179e-05, "loss": 0.0329, "step": 1566 }, { "epoch": 0.16988291413703382, "grad_norm": 0.3314196467399597, "learning_rate": 9.467938488321486e-05, "loss": 0.0363, "step": 1567 }, { "epoch": 0.16999132697311362, "grad_norm": 0.34189194440841675, "learning_rate": 9.467575801537792e-05, "loss": 0.0516, "step": 1568 }, { "epoch": 0.1700997398091934, "grad_norm": 0.959563136100769, "learning_rate": 9.467213114754099e-05, "loss": 0.1067, "step": 1569 }, { "epoch": 0.1702081526452732, "grad_norm": 0.9041455388069153, "learning_rate": 9.466850427970405e-05, "loss": 0.0629, "step": 1570 }, { "epoch": 0.170316565481353, "grad_norm": 0.5494480133056641, "learning_rate": 9.466487741186712e-05, "loss": 0.0544, "step": 1571 }, { "epoch": 0.17042497831743278, "grad_norm": 0.5482295751571655, "learning_rate": 9.466125054403018e-05, "loss": 0.0691, "step": 1572 }, { "epoch": 0.17053339115351257, "grad_norm": 0.530862033367157, "learning_rate": 9.465762367619324e-05, "loss": 0.074, "step": 1573 }, { "epoch": 0.17064180398959236, "grad_norm": 0.09958586096763611, "learning_rate": 9.46539968083563e-05, "loss": 0.0091, "step": 1574 }, { "epoch": 0.17075021682567215, "grad_norm": 0.2524740993976593, "learning_rate": 9.465036994051936e-05, "loss": 0.0338, "step": 1575 }, { "epoch": 0.17085862966175194, "grad_norm": 0.5087155699729919, "learning_rate": 9.464674307268243e-05, "loss": 0.059, "step": 1576 }, { "epoch": 0.17096704249783173, "grad_norm": 0.23385575413703918, "learning_rate": 9.464311620484549e-05, "loss": 0.0476, "step": 1577 }, { "epoch": 0.17107545533391152, "grad_norm": 0.18326866626739502, "learning_rate": 9.463948933700856e-05, "loss": 0.0257, "step": 1578 }, { "epoch": 0.17118386816999132, "grad_norm": 0.3821328580379486, "learning_rate": 9.463586246917163e-05, "loss": 0.0598, "step": 1579 }, { "epoch": 0.1712922810060711, "grad_norm": 0.45496320724487305, "learning_rate": 9.463223560133469e-05, "loss": 0.0578, "step": 1580 }, { "epoch": 0.1714006938421509, "grad_norm": 0.5930209159851074, "learning_rate": 9.462860873349776e-05, "loss": 0.0685, "step": 1581 }, { "epoch": 0.17150910667823072, "grad_norm": 0.5400509238243103, "learning_rate": 9.462498186566082e-05, "loss": 0.0718, "step": 1582 }, { "epoch": 0.1716175195143105, "grad_norm": 0.27089348435401917, "learning_rate": 9.462135499782389e-05, "loss": 0.0276, "step": 1583 }, { "epoch": 0.1717259323503903, "grad_norm": 0.319387286901474, "learning_rate": 9.461772812998695e-05, "loss": 0.0501, "step": 1584 }, { "epoch": 0.1718343451864701, "grad_norm": 0.6763713955879211, "learning_rate": 9.461410126215002e-05, "loss": 0.0721, "step": 1585 }, { "epoch": 0.17194275802254988, "grad_norm": 0.27863195538520813, "learning_rate": 9.461047439431308e-05, "loss": 0.0251, "step": 1586 }, { "epoch": 0.17205117085862967, "grad_norm": 0.5114193558692932, "learning_rate": 9.460684752647613e-05, "loss": 0.0555, "step": 1587 }, { "epoch": 0.17215958369470946, "grad_norm": 0.6153342723846436, "learning_rate": 9.46032206586392e-05, "loss": 0.0598, "step": 1588 }, { "epoch": 0.17226799653078925, "grad_norm": 0.6137434840202332, "learning_rate": 9.459959379080226e-05, "loss": 0.0354, "step": 1589 }, { "epoch": 0.17237640936686904, "grad_norm": 0.2963103950023651, "learning_rate": 9.459596692296533e-05, "loss": 0.0512, "step": 1590 }, { "epoch": 0.17248482220294883, "grad_norm": 0.45662468671798706, "learning_rate": 9.459234005512839e-05, "loss": 0.0714, "step": 1591 }, { "epoch": 0.17259323503902863, "grad_norm": 0.1940794587135315, "learning_rate": 9.458871318729146e-05, "loss": 0.0337, "step": 1592 }, { "epoch": 0.17270164787510842, "grad_norm": 0.31634122133255005, "learning_rate": 9.458508631945452e-05, "loss": 0.0357, "step": 1593 }, { "epoch": 0.1728100607111882, "grad_norm": 0.4973314702510834, "learning_rate": 9.458145945161759e-05, "loss": 0.0866, "step": 1594 }, { "epoch": 0.172918473547268, "grad_norm": 0.3998546004295349, "learning_rate": 9.457783258378066e-05, "loss": 0.0561, "step": 1595 }, { "epoch": 0.1730268863833478, "grad_norm": 0.36891692876815796, "learning_rate": 9.457420571594372e-05, "loss": 0.0379, "step": 1596 }, { "epoch": 0.17313529921942758, "grad_norm": 0.39716723561286926, "learning_rate": 9.457057884810679e-05, "loss": 0.0317, "step": 1597 }, { "epoch": 0.17324371205550737, "grad_norm": 0.4747673273086548, "learning_rate": 9.456695198026984e-05, "loss": 0.1099, "step": 1598 }, { "epoch": 0.17335212489158716, "grad_norm": 0.5950831174850464, "learning_rate": 9.456332511243292e-05, "loss": 0.0638, "step": 1599 }, { "epoch": 0.17346053772766695, "grad_norm": 0.5598370432853699, "learning_rate": 9.455969824459597e-05, "loss": 0.0677, "step": 1600 }, { "epoch": 0.17356895056374674, "grad_norm": 0.39566120505332947, "learning_rate": 9.455607137675903e-05, "loss": 0.0478, "step": 1601 }, { "epoch": 0.17367736339982653, "grad_norm": 0.3367222845554352, "learning_rate": 9.45524445089221e-05, "loss": 0.0483, "step": 1602 }, { "epoch": 0.17378577623590633, "grad_norm": 0.10986971110105515, "learning_rate": 9.454881764108516e-05, "loss": 0.0238, "step": 1603 }, { "epoch": 0.17389418907198612, "grad_norm": 1.2054309844970703, "learning_rate": 9.454519077324823e-05, "loss": 0.0877, "step": 1604 }, { "epoch": 0.1740026019080659, "grad_norm": 0.41892871260643005, "learning_rate": 9.454156390541129e-05, "loss": 0.0448, "step": 1605 }, { "epoch": 0.1741110147441457, "grad_norm": 0.6014031767845154, "learning_rate": 9.453793703757436e-05, "loss": 0.0883, "step": 1606 }, { "epoch": 0.1742194275802255, "grad_norm": 0.5079967379570007, "learning_rate": 9.453431016973741e-05, "loss": 0.037, "step": 1607 }, { "epoch": 0.17432784041630528, "grad_norm": 0.27857574820518494, "learning_rate": 9.453068330190049e-05, "loss": 0.0323, "step": 1608 }, { "epoch": 0.17443625325238507, "grad_norm": 0.35405534505844116, "learning_rate": 9.452705643406354e-05, "loss": 0.0418, "step": 1609 }, { "epoch": 0.17454466608846486, "grad_norm": 0.28971371054649353, "learning_rate": 9.45234295662266e-05, "loss": 0.0495, "step": 1610 }, { "epoch": 0.17465307892454465, "grad_norm": 0.6155475974082947, "learning_rate": 9.451980269838967e-05, "loss": 0.0872, "step": 1611 }, { "epoch": 0.17476149176062447, "grad_norm": 0.17738784849643707, "learning_rate": 9.451617583055274e-05, "loss": 0.0271, "step": 1612 }, { "epoch": 0.17486990459670426, "grad_norm": 0.8395654559135437, "learning_rate": 9.451254896271581e-05, "loss": 0.0609, "step": 1613 }, { "epoch": 0.17497831743278405, "grad_norm": 0.6144455671310425, "learning_rate": 9.450892209487887e-05, "loss": 0.0742, "step": 1614 }, { "epoch": 0.17508673026886384, "grad_norm": 0.3138125538825989, "learning_rate": 9.450529522704194e-05, "loss": 0.0516, "step": 1615 }, { "epoch": 0.17519514310494363, "grad_norm": 0.36960533261299133, "learning_rate": 9.4501668359205e-05, "loss": 0.0866, "step": 1616 }, { "epoch": 0.17530355594102343, "grad_norm": 0.41426315903663635, "learning_rate": 9.449804149136806e-05, "loss": 0.0859, "step": 1617 }, { "epoch": 0.17541196877710322, "grad_norm": 0.5607994794845581, "learning_rate": 9.449441462353113e-05, "loss": 0.0789, "step": 1618 }, { "epoch": 0.175520381613183, "grad_norm": 0.31626030802726746, "learning_rate": 9.449078775569418e-05, "loss": 0.0432, "step": 1619 }, { "epoch": 0.1756287944492628, "grad_norm": 0.4389789402484894, "learning_rate": 9.448716088785725e-05, "loss": 0.058, "step": 1620 }, { "epoch": 0.1757372072853426, "grad_norm": 0.4649680554866791, "learning_rate": 9.448353402002031e-05, "loss": 0.0301, "step": 1621 }, { "epoch": 0.17584562012142238, "grad_norm": 0.2269514501094818, "learning_rate": 9.447990715218338e-05, "loss": 0.0332, "step": 1622 }, { "epoch": 0.17595403295750217, "grad_norm": 0.39677169919013977, "learning_rate": 9.447628028434644e-05, "loss": 0.0486, "step": 1623 }, { "epoch": 0.17606244579358196, "grad_norm": 0.44188040494918823, "learning_rate": 9.447265341650951e-05, "loss": 0.0563, "step": 1624 }, { "epoch": 0.17617085862966175, "grad_norm": 0.5511540174484253, "learning_rate": 9.446902654867257e-05, "loss": 0.08, "step": 1625 }, { "epoch": 0.17627927146574154, "grad_norm": 0.32744160294532776, "learning_rate": 9.446539968083563e-05, "loss": 0.0354, "step": 1626 }, { "epoch": 0.17638768430182133, "grad_norm": 0.40258997678756714, "learning_rate": 9.44617728129987e-05, "loss": 0.0643, "step": 1627 }, { "epoch": 0.17649609713790113, "grad_norm": 0.3681642711162567, "learning_rate": 9.445814594516175e-05, "loss": 0.0608, "step": 1628 }, { "epoch": 0.17660450997398092, "grad_norm": 0.36383455991744995, "learning_rate": 9.445451907732484e-05, "loss": 0.026, "step": 1629 }, { "epoch": 0.1767129228100607, "grad_norm": 0.2944760322570801, "learning_rate": 9.44508922094879e-05, "loss": 0.0422, "step": 1630 }, { "epoch": 0.1768213356461405, "grad_norm": 0.23198865354061127, "learning_rate": 9.444726534165095e-05, "loss": 0.0317, "step": 1631 }, { "epoch": 0.1769297484822203, "grad_norm": 0.23432078957557678, "learning_rate": 9.444363847381402e-05, "loss": 0.0249, "step": 1632 }, { "epoch": 0.17703816131830008, "grad_norm": 0.17181357741355896, "learning_rate": 9.444001160597708e-05, "loss": 0.0257, "step": 1633 }, { "epoch": 0.17714657415437987, "grad_norm": 0.38664427399635315, "learning_rate": 9.443638473814015e-05, "loss": 0.0373, "step": 1634 }, { "epoch": 0.17725498699045966, "grad_norm": 0.6739829778671265, "learning_rate": 9.443275787030321e-05, "loss": 0.0881, "step": 1635 }, { "epoch": 0.17736339982653945, "grad_norm": 0.25608569383621216, "learning_rate": 9.442913100246628e-05, "loss": 0.0354, "step": 1636 }, { "epoch": 0.17747181266261924, "grad_norm": 0.5413661003112793, "learning_rate": 9.442550413462934e-05, "loss": 0.0788, "step": 1637 }, { "epoch": 0.17758022549869903, "grad_norm": 0.5011325478553772, "learning_rate": 9.442187726679241e-05, "loss": 0.0294, "step": 1638 }, { "epoch": 0.17768863833477883, "grad_norm": 0.378164678812027, "learning_rate": 9.441825039895547e-05, "loss": 0.0872, "step": 1639 }, { "epoch": 0.17779705117085862, "grad_norm": 0.49268031120300293, "learning_rate": 9.441462353111852e-05, "loss": 0.0515, "step": 1640 }, { "epoch": 0.17790546400693844, "grad_norm": 0.2956886291503906, "learning_rate": 9.44109966632816e-05, "loss": 0.0379, "step": 1641 }, { "epoch": 0.17801387684301823, "grad_norm": 0.8353269100189209, "learning_rate": 9.440736979544465e-05, "loss": 0.0428, "step": 1642 }, { "epoch": 0.17812228967909802, "grad_norm": 0.29232293367385864, "learning_rate": 9.440374292760772e-05, "loss": 0.0543, "step": 1643 }, { "epoch": 0.1782307025151778, "grad_norm": 0.45864325761795044, "learning_rate": 9.440011605977078e-05, "loss": 0.0535, "step": 1644 }, { "epoch": 0.1783391153512576, "grad_norm": 0.16938918828964233, "learning_rate": 9.439648919193385e-05, "loss": 0.0269, "step": 1645 }, { "epoch": 0.1784475281873374, "grad_norm": 0.5248890519142151, "learning_rate": 9.439286232409691e-05, "loss": 0.0754, "step": 1646 }, { "epoch": 0.17855594102341718, "grad_norm": 0.38232266902923584, "learning_rate": 9.438923545625998e-05, "loss": 0.0471, "step": 1647 }, { "epoch": 0.17866435385949697, "grad_norm": 0.7047879695892334, "learning_rate": 9.438560858842305e-05, "loss": 0.0412, "step": 1648 }, { "epoch": 0.17877276669557676, "grad_norm": 0.45175641775131226, "learning_rate": 9.438198172058611e-05, "loss": 0.0831, "step": 1649 }, { "epoch": 0.17888117953165655, "grad_norm": 0.2651156485080719, "learning_rate": 9.437835485274918e-05, "loss": 0.049, "step": 1650 }, { "epoch": 0.17898959236773634, "grad_norm": 0.4831756055355072, "learning_rate": 9.437472798491224e-05, "loss": 0.0432, "step": 1651 }, { "epoch": 0.17909800520381614, "grad_norm": 0.1762629598379135, "learning_rate": 9.43711011170753e-05, "loss": 0.0181, "step": 1652 }, { "epoch": 0.17920641803989593, "grad_norm": 0.5326229333877563, "learning_rate": 9.436747424923836e-05, "loss": 0.0535, "step": 1653 }, { "epoch": 0.17931483087597572, "grad_norm": 0.40076273679733276, "learning_rate": 9.436384738140142e-05, "loss": 0.0166, "step": 1654 }, { "epoch": 0.1794232437120555, "grad_norm": 0.49699166417121887, "learning_rate": 9.436022051356449e-05, "loss": 0.0626, "step": 1655 }, { "epoch": 0.1795316565481353, "grad_norm": 0.32893210649490356, "learning_rate": 9.435659364572755e-05, "loss": 0.0343, "step": 1656 }, { "epoch": 0.1796400693842151, "grad_norm": 0.3006039559841156, "learning_rate": 9.435296677789062e-05, "loss": 0.0498, "step": 1657 }, { "epoch": 0.17974848222029488, "grad_norm": 0.7991642951965332, "learning_rate": 9.434933991005368e-05, "loss": 0.1011, "step": 1658 }, { "epoch": 0.17985689505637467, "grad_norm": 0.17739906907081604, "learning_rate": 9.434571304221675e-05, "loss": 0.0164, "step": 1659 }, { "epoch": 0.17996530789245446, "grad_norm": 0.6155792474746704, "learning_rate": 9.43420861743798e-05, "loss": 0.0616, "step": 1660 }, { "epoch": 0.18007372072853425, "grad_norm": 0.7620417475700378, "learning_rate": 9.433845930654288e-05, "loss": 0.1098, "step": 1661 }, { "epoch": 0.18018213356461404, "grad_norm": 0.24539415538311005, "learning_rate": 9.433483243870593e-05, "loss": 0.024, "step": 1662 }, { "epoch": 0.18029054640069384, "grad_norm": 0.689421534538269, "learning_rate": 9.433120557086899e-05, "loss": 0.1151, "step": 1663 }, { "epoch": 0.18039895923677363, "grad_norm": 0.3999794125556946, "learning_rate": 9.432757870303208e-05, "loss": 0.0503, "step": 1664 }, { "epoch": 0.18050737207285342, "grad_norm": 0.4964592456817627, "learning_rate": 9.432395183519513e-05, "loss": 0.0568, "step": 1665 }, { "epoch": 0.1806157849089332, "grad_norm": 0.4744388163089752, "learning_rate": 9.43203249673582e-05, "loss": 0.0398, "step": 1666 }, { "epoch": 0.180724197745013, "grad_norm": 0.4164460301399231, "learning_rate": 9.431669809952126e-05, "loss": 0.0927, "step": 1667 }, { "epoch": 0.1808326105810928, "grad_norm": 0.21132755279541016, "learning_rate": 9.431307123168432e-05, "loss": 0.0454, "step": 1668 }, { "epoch": 0.18094102341717258, "grad_norm": 0.6797118186950684, "learning_rate": 9.430944436384739e-05, "loss": 0.074, "step": 1669 }, { "epoch": 0.18104943625325237, "grad_norm": 0.2315083146095276, "learning_rate": 9.430581749601045e-05, "loss": 0.0308, "step": 1670 }, { "epoch": 0.1811578490893322, "grad_norm": 0.11606279015541077, "learning_rate": 9.430219062817352e-05, "loss": 0.0158, "step": 1671 }, { "epoch": 0.18126626192541198, "grad_norm": 0.35420283675193787, "learning_rate": 9.429856376033657e-05, "loss": 0.036, "step": 1672 }, { "epoch": 0.18137467476149177, "grad_norm": 1.3004660606384277, "learning_rate": 9.429493689249965e-05, "loss": 0.0527, "step": 1673 }, { "epoch": 0.18148308759757156, "grad_norm": 0.1873827576637268, "learning_rate": 9.42913100246627e-05, "loss": 0.0383, "step": 1674 }, { "epoch": 0.18159150043365135, "grad_norm": 0.3884720206260681, "learning_rate": 9.428768315682577e-05, "loss": 0.0557, "step": 1675 }, { "epoch": 0.18169991326973114, "grad_norm": 0.30040982365608215, "learning_rate": 9.428405628898883e-05, "loss": 0.0337, "step": 1676 }, { "epoch": 0.18180832610581094, "grad_norm": 0.4800068438053131, "learning_rate": 9.428042942115189e-05, "loss": 0.0342, "step": 1677 }, { "epoch": 0.18191673894189073, "grad_norm": 0.2462698519229889, "learning_rate": 9.427680255331496e-05, "loss": 0.0393, "step": 1678 }, { "epoch": 0.18202515177797052, "grad_norm": 0.23939475417137146, "learning_rate": 9.427317568547802e-05, "loss": 0.0285, "step": 1679 }, { "epoch": 0.1821335646140503, "grad_norm": 0.6387245059013367, "learning_rate": 9.426954881764109e-05, "loss": 0.0842, "step": 1680 }, { "epoch": 0.1822419774501301, "grad_norm": 0.4819486141204834, "learning_rate": 9.426592194980416e-05, "loss": 0.0552, "step": 1681 }, { "epoch": 0.1823503902862099, "grad_norm": 0.6936518549919128, "learning_rate": 9.426229508196722e-05, "loss": 0.0576, "step": 1682 }, { "epoch": 0.18245880312228968, "grad_norm": 0.2696215510368347, "learning_rate": 9.425866821413029e-05, "loss": 0.0354, "step": 1683 }, { "epoch": 0.18256721595836947, "grad_norm": 0.1676282435655594, "learning_rate": 9.425504134629334e-05, "loss": 0.0256, "step": 1684 }, { "epoch": 0.18267562879444926, "grad_norm": 0.42029738426208496, "learning_rate": 9.425141447845641e-05, "loss": 0.0387, "step": 1685 }, { "epoch": 0.18278404163052905, "grad_norm": 0.40893903374671936, "learning_rate": 9.424778761061947e-05, "loss": 0.052, "step": 1686 }, { "epoch": 0.18289245446660884, "grad_norm": 0.34906089305877686, "learning_rate": 9.424416074278254e-05, "loss": 0.0407, "step": 1687 }, { "epoch": 0.18300086730268864, "grad_norm": 0.48614054918289185, "learning_rate": 9.42405338749456e-05, "loss": 0.0165, "step": 1688 }, { "epoch": 0.18310928013876843, "grad_norm": 0.3107841908931732, "learning_rate": 9.423690700710867e-05, "loss": 0.0321, "step": 1689 }, { "epoch": 0.18321769297484822, "grad_norm": 0.4405224621295929, "learning_rate": 9.423328013927173e-05, "loss": 0.0804, "step": 1690 }, { "epoch": 0.183326105810928, "grad_norm": 0.5180060863494873, "learning_rate": 9.422965327143479e-05, "loss": 0.0268, "step": 1691 }, { "epoch": 0.1834345186470078, "grad_norm": 0.5635162591934204, "learning_rate": 9.422602640359786e-05, "loss": 0.0519, "step": 1692 }, { "epoch": 0.1835429314830876, "grad_norm": 0.40152379870414734, "learning_rate": 9.422239953576091e-05, "loss": 0.0253, "step": 1693 }, { "epoch": 0.18365134431916738, "grad_norm": 0.5406116843223572, "learning_rate": 9.421877266792399e-05, "loss": 0.057, "step": 1694 }, { "epoch": 0.18375975715524717, "grad_norm": 0.42775416374206543, "learning_rate": 9.421514580008704e-05, "loss": 0.0472, "step": 1695 }, { "epoch": 0.18386816999132696, "grad_norm": 0.9126606583595276, "learning_rate": 9.421151893225011e-05, "loss": 0.098, "step": 1696 }, { "epoch": 0.18397658282740675, "grad_norm": 0.685379147529602, "learning_rate": 9.420789206441317e-05, "loss": 0.0672, "step": 1697 }, { "epoch": 0.18408499566348654, "grad_norm": 0.40178313851356506, "learning_rate": 9.420426519657624e-05, "loss": 0.0676, "step": 1698 }, { "epoch": 0.18419340849956634, "grad_norm": 0.5056653618812561, "learning_rate": 9.420063832873931e-05, "loss": 0.0582, "step": 1699 }, { "epoch": 0.18430182133564615, "grad_norm": 0.8724767565727234, "learning_rate": 9.419701146090237e-05, "loss": 0.056, "step": 1700 }, { "epoch": 0.18441023417172595, "grad_norm": 0.6742528080940247, "learning_rate": 9.419338459306544e-05, "loss": 0.0958, "step": 1701 }, { "epoch": 0.18451864700780574, "grad_norm": 0.35299915075302124, "learning_rate": 9.41897577252285e-05, "loss": 0.0348, "step": 1702 }, { "epoch": 0.18462705984388553, "grad_norm": 0.1752498596906662, "learning_rate": 9.418613085739157e-05, "loss": 0.0202, "step": 1703 }, { "epoch": 0.18473547267996532, "grad_norm": 0.46785008907318115, "learning_rate": 9.418250398955463e-05, "loss": 0.0708, "step": 1704 }, { "epoch": 0.1848438855160451, "grad_norm": 0.43793871998786926, "learning_rate": 9.41788771217177e-05, "loss": 0.0477, "step": 1705 }, { "epoch": 0.1849522983521249, "grad_norm": 0.44780710339546204, "learning_rate": 9.417525025388075e-05, "loss": 0.0582, "step": 1706 }, { "epoch": 0.1850607111882047, "grad_norm": 0.23035211861133575, "learning_rate": 9.417162338604381e-05, "loss": 0.0526, "step": 1707 }, { "epoch": 0.18516912402428448, "grad_norm": 0.31387895345687866, "learning_rate": 9.416799651820688e-05, "loss": 0.0457, "step": 1708 }, { "epoch": 0.18527753686036427, "grad_norm": 0.47530514001846313, "learning_rate": 9.416436965036994e-05, "loss": 0.0382, "step": 1709 }, { "epoch": 0.18538594969644406, "grad_norm": 0.2898920774459839, "learning_rate": 9.416074278253301e-05, "loss": 0.0508, "step": 1710 }, { "epoch": 0.18549436253252385, "grad_norm": 0.17148622870445251, "learning_rate": 9.415711591469607e-05, "loss": 0.035, "step": 1711 }, { "epoch": 0.18560277536860365, "grad_norm": 0.42143723368644714, "learning_rate": 9.415348904685914e-05, "loss": 0.032, "step": 1712 }, { "epoch": 0.18571118820468344, "grad_norm": 0.272123247385025, "learning_rate": 9.41498621790222e-05, "loss": 0.0559, "step": 1713 }, { "epoch": 0.18581960104076323, "grad_norm": 0.4592066705226898, "learning_rate": 9.414623531118525e-05, "loss": 0.0916, "step": 1714 }, { "epoch": 0.18592801387684302, "grad_norm": 0.628169596195221, "learning_rate": 9.414260844334832e-05, "loss": 0.0936, "step": 1715 }, { "epoch": 0.1860364267129228, "grad_norm": 0.3510935604572296, "learning_rate": 9.41389815755114e-05, "loss": 0.044, "step": 1716 }, { "epoch": 0.1861448395490026, "grad_norm": 0.4033823013305664, "learning_rate": 9.413535470767447e-05, "loss": 0.0383, "step": 1717 }, { "epoch": 0.1862532523850824, "grad_norm": 0.348004549741745, "learning_rate": 9.413172783983752e-05, "loss": 0.0397, "step": 1718 }, { "epoch": 0.18636166522116218, "grad_norm": 0.5487136840820312, "learning_rate": 9.41281009720006e-05, "loss": 0.0508, "step": 1719 }, { "epoch": 0.18647007805724197, "grad_norm": 0.22317267954349518, "learning_rate": 9.412447410416365e-05, "loss": 0.0363, "step": 1720 }, { "epoch": 0.18657849089332176, "grad_norm": 0.299382746219635, "learning_rate": 9.412084723632671e-05, "loss": 0.0407, "step": 1721 }, { "epoch": 0.18668690372940155, "grad_norm": 0.5961348414421082, "learning_rate": 9.411722036848978e-05, "loss": 0.0538, "step": 1722 }, { "epoch": 0.18679531656548135, "grad_norm": 0.35178130865097046, "learning_rate": 9.411359350065284e-05, "loss": 0.046, "step": 1723 }, { "epoch": 0.18690372940156114, "grad_norm": 0.4621918499469757, "learning_rate": 9.410996663281591e-05, "loss": 0.0527, "step": 1724 }, { "epoch": 0.18701214223764093, "grad_norm": 0.5317115783691406, "learning_rate": 9.410633976497897e-05, "loss": 0.0528, "step": 1725 }, { "epoch": 0.18712055507372072, "grad_norm": 0.3011169135570526, "learning_rate": 9.410271289714204e-05, "loss": 0.028, "step": 1726 }, { "epoch": 0.1872289679098005, "grad_norm": 0.13742555677890778, "learning_rate": 9.40990860293051e-05, "loss": 0.0089, "step": 1727 }, { "epoch": 0.1873373807458803, "grad_norm": 0.2893613874912262, "learning_rate": 9.409545916146816e-05, "loss": 0.0505, "step": 1728 }, { "epoch": 0.1874457935819601, "grad_norm": 0.5887779593467712, "learning_rate": 9.409183229363122e-05, "loss": 0.0468, "step": 1729 }, { "epoch": 0.1875542064180399, "grad_norm": 0.6292153596878052, "learning_rate": 9.408820542579428e-05, "loss": 0.0827, "step": 1730 }, { "epoch": 0.1876626192541197, "grad_norm": 0.2926650047302246, "learning_rate": 9.408457855795735e-05, "loss": 0.0402, "step": 1731 }, { "epoch": 0.1877710320901995, "grad_norm": 0.3386939465999603, "learning_rate": 9.408095169012041e-05, "loss": 0.0438, "step": 1732 }, { "epoch": 0.18787944492627928, "grad_norm": 0.19810537993907928, "learning_rate": 9.407732482228349e-05, "loss": 0.0355, "step": 1733 }, { "epoch": 0.18798785776235907, "grad_norm": 0.318817138671875, "learning_rate": 9.407369795444655e-05, "loss": 0.047, "step": 1734 }, { "epoch": 0.18809627059843886, "grad_norm": 0.4530656635761261, "learning_rate": 9.40700710866096e-05, "loss": 0.0663, "step": 1735 }, { "epoch": 0.18820468343451865, "grad_norm": 0.634161114692688, "learning_rate": 9.406644421877268e-05, "loss": 0.1108, "step": 1736 }, { "epoch": 0.18831309627059845, "grad_norm": 0.30598556995391846, "learning_rate": 9.406281735093573e-05, "loss": 0.0424, "step": 1737 }, { "epoch": 0.18842150910667824, "grad_norm": 0.5976157188415527, "learning_rate": 9.40591904830988e-05, "loss": 0.0871, "step": 1738 }, { "epoch": 0.18852992194275803, "grad_norm": 0.4925602674484253, "learning_rate": 9.405556361526186e-05, "loss": 0.0592, "step": 1739 }, { "epoch": 0.18863833477883782, "grad_norm": 0.3338387906551361, "learning_rate": 9.405193674742493e-05, "loss": 0.0553, "step": 1740 }, { "epoch": 0.1887467476149176, "grad_norm": 0.4219960570335388, "learning_rate": 9.404830987958799e-05, "loss": 0.045, "step": 1741 }, { "epoch": 0.1888551604509974, "grad_norm": 0.32500895857810974, "learning_rate": 9.404468301175106e-05, "loss": 0.0589, "step": 1742 }, { "epoch": 0.1889635732870772, "grad_norm": 0.5895605683326721, "learning_rate": 9.404105614391412e-05, "loss": 0.0568, "step": 1743 }, { "epoch": 0.18907198612315698, "grad_norm": 0.4430801272392273, "learning_rate": 9.403742927607718e-05, "loss": 0.0336, "step": 1744 }, { "epoch": 0.18918039895923677, "grad_norm": 0.6304055452346802, "learning_rate": 9.403380240824025e-05, "loss": 0.0613, "step": 1745 }, { "epoch": 0.18928881179531656, "grad_norm": 0.3509046137332916, "learning_rate": 9.40301755404033e-05, "loss": 0.0551, "step": 1746 }, { "epoch": 0.18939722463139635, "grad_norm": 0.3137136697769165, "learning_rate": 9.402654867256638e-05, "loss": 0.0458, "step": 1747 }, { "epoch": 0.18950563746747615, "grad_norm": 0.368389368057251, "learning_rate": 9.402292180472943e-05, "loss": 0.0309, "step": 1748 }, { "epoch": 0.18961405030355594, "grad_norm": 0.5438107252120972, "learning_rate": 9.40192949368925e-05, "loss": 0.0393, "step": 1749 }, { "epoch": 0.18972246313963573, "grad_norm": 0.2634875476360321, "learning_rate": 9.401566806905557e-05, "loss": 0.0188, "step": 1750 }, { "epoch": 0.18983087597571552, "grad_norm": 0.4801265597343445, "learning_rate": 9.401204120121863e-05, "loss": 0.0434, "step": 1751 }, { "epoch": 0.1899392888117953, "grad_norm": 0.6381727457046509, "learning_rate": 9.40084143333817e-05, "loss": 0.0573, "step": 1752 }, { "epoch": 0.1900477016478751, "grad_norm": 0.5331783294677734, "learning_rate": 9.400478746554476e-05, "loss": 0.0427, "step": 1753 }, { "epoch": 0.1901561144839549, "grad_norm": 0.5207101106643677, "learning_rate": 9.400116059770783e-05, "loss": 0.044, "step": 1754 }, { "epoch": 0.19026452732003468, "grad_norm": 0.3818296790122986, "learning_rate": 9.399753372987089e-05, "loss": 0.0333, "step": 1755 }, { "epoch": 0.19037294015611447, "grad_norm": 0.6458210349082947, "learning_rate": 9.399390686203396e-05, "loss": 0.069, "step": 1756 }, { "epoch": 0.19048135299219426, "grad_norm": 0.2808574438095093, "learning_rate": 9.399027999419702e-05, "loss": 0.0365, "step": 1757 }, { "epoch": 0.19058976582827405, "grad_norm": 0.2502758502960205, "learning_rate": 9.398665312636007e-05, "loss": 0.0265, "step": 1758 }, { "epoch": 0.19069817866435385, "grad_norm": 0.3387150764465332, "learning_rate": 9.398302625852315e-05, "loss": 0.0487, "step": 1759 }, { "epoch": 0.19080659150043366, "grad_norm": 0.4685699939727783, "learning_rate": 9.39793993906862e-05, "loss": 0.0581, "step": 1760 }, { "epoch": 0.19091500433651346, "grad_norm": 0.3766342103481293, "learning_rate": 9.397577252284927e-05, "loss": 0.0419, "step": 1761 }, { "epoch": 0.19102341717259325, "grad_norm": 0.40269753336906433, "learning_rate": 9.397214565501233e-05, "loss": 0.0415, "step": 1762 }, { "epoch": 0.19113183000867304, "grad_norm": 0.7218931317329407, "learning_rate": 9.39685187871754e-05, "loss": 0.0512, "step": 1763 }, { "epoch": 0.19124024284475283, "grad_norm": 0.32761675119400024, "learning_rate": 9.396489191933846e-05, "loss": 0.0299, "step": 1764 }, { "epoch": 0.19134865568083262, "grad_norm": 0.4015944004058838, "learning_rate": 9.396126505150153e-05, "loss": 0.0373, "step": 1765 }, { "epoch": 0.1914570685169124, "grad_norm": 0.5936662554740906, "learning_rate": 9.395763818366459e-05, "loss": 0.028, "step": 1766 }, { "epoch": 0.1915654813529922, "grad_norm": 0.8945732116699219, "learning_rate": 9.395401131582764e-05, "loss": 0.0359, "step": 1767 }, { "epoch": 0.191673894189072, "grad_norm": 0.3143032491207123, "learning_rate": 9.395038444799073e-05, "loss": 0.0502, "step": 1768 }, { "epoch": 0.19178230702515178, "grad_norm": 0.372952938079834, "learning_rate": 9.394675758015379e-05, "loss": 0.0547, "step": 1769 }, { "epoch": 0.19189071986123157, "grad_norm": 0.8493732810020447, "learning_rate": 9.394313071231686e-05, "loss": 0.0819, "step": 1770 }, { "epoch": 0.19199913269731136, "grad_norm": 0.47509750723838806, "learning_rate": 9.393950384447991e-05, "loss": 0.0482, "step": 1771 }, { "epoch": 0.19210754553339116, "grad_norm": 0.7896063327789307, "learning_rate": 9.393587697664297e-05, "loss": 0.0793, "step": 1772 }, { "epoch": 0.19221595836947095, "grad_norm": 0.5003307461738586, "learning_rate": 9.393225010880604e-05, "loss": 0.0466, "step": 1773 }, { "epoch": 0.19232437120555074, "grad_norm": 0.25171908736228943, "learning_rate": 9.39286232409691e-05, "loss": 0.0306, "step": 1774 }, { "epoch": 0.19243278404163053, "grad_norm": 0.5143831372261047, "learning_rate": 9.392499637313217e-05, "loss": 0.0256, "step": 1775 }, { "epoch": 0.19254119687771032, "grad_norm": 0.42183831334114075, "learning_rate": 9.392136950529523e-05, "loss": 0.0627, "step": 1776 }, { "epoch": 0.1926496097137901, "grad_norm": 1.0465004444122314, "learning_rate": 9.39177426374583e-05, "loss": 0.0632, "step": 1777 }, { "epoch": 0.1927580225498699, "grad_norm": 0.76795893907547, "learning_rate": 9.391411576962136e-05, "loss": 0.0468, "step": 1778 }, { "epoch": 0.1928664353859497, "grad_norm": 0.672337532043457, "learning_rate": 9.391048890178443e-05, "loss": 0.0464, "step": 1779 }, { "epoch": 0.19297484822202948, "grad_norm": 0.2614632546901703, "learning_rate": 9.390686203394748e-05, "loss": 0.0322, "step": 1780 }, { "epoch": 0.19308326105810927, "grad_norm": 0.46075984835624695, "learning_rate": 9.390323516611054e-05, "loss": 0.0453, "step": 1781 }, { "epoch": 0.19319167389418906, "grad_norm": 0.48972222208976746, "learning_rate": 9.389960829827361e-05, "loss": 0.0476, "step": 1782 }, { "epoch": 0.19330008673026886, "grad_norm": 0.4070713222026825, "learning_rate": 9.389598143043667e-05, "loss": 0.045, "step": 1783 }, { "epoch": 0.19340849956634865, "grad_norm": 0.43030551075935364, "learning_rate": 9.389235456259974e-05, "loss": 0.0524, "step": 1784 }, { "epoch": 0.19351691240242844, "grad_norm": 0.3491775095462799, "learning_rate": 9.388872769476281e-05, "loss": 0.0367, "step": 1785 }, { "epoch": 0.19362532523850823, "grad_norm": 0.17380057275295258, "learning_rate": 9.388510082692587e-05, "loss": 0.025, "step": 1786 }, { "epoch": 0.19373373807458802, "grad_norm": 0.5190564393997192, "learning_rate": 9.388147395908894e-05, "loss": 0.0647, "step": 1787 }, { "epoch": 0.1938421509106678, "grad_norm": 0.48446062207221985, "learning_rate": 9.3877847091252e-05, "loss": 0.0925, "step": 1788 }, { "epoch": 0.19395056374674763, "grad_norm": 0.19932372868061066, "learning_rate": 9.387422022341507e-05, "loss": 0.0244, "step": 1789 }, { "epoch": 0.19405897658282742, "grad_norm": 0.5299738049507141, "learning_rate": 9.387059335557813e-05, "loss": 0.0825, "step": 1790 }, { "epoch": 0.1941673894189072, "grad_norm": 0.5692341923713684, "learning_rate": 9.38669664877412e-05, "loss": 0.0772, "step": 1791 }, { "epoch": 0.194275802254987, "grad_norm": 0.30201300978660583, "learning_rate": 9.386333961990425e-05, "loss": 0.0267, "step": 1792 }, { "epoch": 0.1943842150910668, "grad_norm": 0.5255452990531921, "learning_rate": 9.385971275206732e-05, "loss": 0.0798, "step": 1793 }, { "epoch": 0.19449262792714658, "grad_norm": 0.20280250906944275, "learning_rate": 9.385608588423038e-05, "loss": 0.0348, "step": 1794 }, { "epoch": 0.19460104076322637, "grad_norm": 0.14222648739814758, "learning_rate": 9.385245901639344e-05, "loss": 0.0171, "step": 1795 }, { "epoch": 0.19470945359930616, "grad_norm": 0.6473680138587952, "learning_rate": 9.384883214855651e-05, "loss": 0.0677, "step": 1796 }, { "epoch": 0.19481786643538596, "grad_norm": 0.5106446743011475, "learning_rate": 9.384520528071957e-05, "loss": 0.0736, "step": 1797 }, { "epoch": 0.19492627927146575, "grad_norm": 0.4638446867465973, "learning_rate": 9.384157841288264e-05, "loss": 0.0772, "step": 1798 }, { "epoch": 0.19503469210754554, "grad_norm": 0.39626818895339966, "learning_rate": 9.38379515450457e-05, "loss": 0.0631, "step": 1799 }, { "epoch": 0.19514310494362533, "grad_norm": 0.3832210302352905, "learning_rate": 9.383432467720877e-05, "loss": 0.0601, "step": 1800 }, { "epoch": 0.19525151777970512, "grad_norm": 0.5199958682060242, "learning_rate": 9.383069780937182e-05, "loss": 0.065, "step": 1801 }, { "epoch": 0.1953599306157849, "grad_norm": 0.4475412368774414, "learning_rate": 9.38270709415349e-05, "loss": 0.0587, "step": 1802 }, { "epoch": 0.1954683434518647, "grad_norm": 0.5371076464653015, "learning_rate": 9.382344407369797e-05, "loss": 0.0879, "step": 1803 }, { "epoch": 0.1955767562879445, "grad_norm": 1.9953722953796387, "learning_rate": 9.381981720586102e-05, "loss": 0.1395, "step": 1804 }, { "epoch": 0.19568516912402428, "grad_norm": 0.24921630322933197, "learning_rate": 9.38161903380241e-05, "loss": 0.0482, "step": 1805 }, { "epoch": 0.19579358196010407, "grad_norm": 0.31781813502311707, "learning_rate": 9.381256347018715e-05, "loss": 0.0335, "step": 1806 }, { "epoch": 0.19590199479618386, "grad_norm": 0.24781683087348938, "learning_rate": 9.380893660235022e-05, "loss": 0.0243, "step": 1807 }, { "epoch": 0.19601040763226366, "grad_norm": 0.44554710388183594, "learning_rate": 9.380530973451328e-05, "loss": 0.0302, "step": 1808 }, { "epoch": 0.19611882046834345, "grad_norm": 0.27471086382865906, "learning_rate": 9.380168286667635e-05, "loss": 0.033, "step": 1809 }, { "epoch": 0.19622723330442324, "grad_norm": 0.13258038461208344, "learning_rate": 9.379805599883941e-05, "loss": 0.0099, "step": 1810 }, { "epoch": 0.19633564614050303, "grad_norm": 0.22391842305660248, "learning_rate": 9.379442913100246e-05, "loss": 0.0266, "step": 1811 }, { "epoch": 0.19644405897658282, "grad_norm": 0.6068840622901917, "learning_rate": 9.379080226316554e-05, "loss": 0.0572, "step": 1812 }, { "epoch": 0.1965524718126626, "grad_norm": 0.31140512228012085, "learning_rate": 9.378717539532859e-05, "loss": 0.0718, "step": 1813 }, { "epoch": 0.1966608846487424, "grad_norm": 0.332276850938797, "learning_rate": 9.378354852749166e-05, "loss": 0.0378, "step": 1814 }, { "epoch": 0.1967692974848222, "grad_norm": 0.40307891368865967, "learning_rate": 9.377992165965472e-05, "loss": 0.0291, "step": 1815 }, { "epoch": 0.19687771032090198, "grad_norm": 0.40074869990348816, "learning_rate": 9.377629479181779e-05, "loss": 0.0498, "step": 1816 }, { "epoch": 0.19698612315698177, "grad_norm": 0.858356237411499, "learning_rate": 9.377266792398085e-05, "loss": 0.0546, "step": 1817 }, { "epoch": 0.19709453599306156, "grad_norm": 0.48382213711738586, "learning_rate": 9.376904105614391e-05, "loss": 0.0625, "step": 1818 }, { "epoch": 0.19720294882914138, "grad_norm": 0.32564395666122437, "learning_rate": 9.376541418830699e-05, "loss": 0.0538, "step": 1819 }, { "epoch": 0.19731136166522117, "grad_norm": 0.17918644845485687, "learning_rate": 9.376178732047005e-05, "loss": 0.0201, "step": 1820 }, { "epoch": 0.19741977450130097, "grad_norm": 0.48444122076034546, "learning_rate": 9.375816045263312e-05, "loss": 0.0491, "step": 1821 }, { "epoch": 0.19752818733738076, "grad_norm": 0.400017648935318, "learning_rate": 9.375453358479618e-05, "loss": 0.0505, "step": 1822 }, { "epoch": 0.19763660017346055, "grad_norm": 0.45567697286605835, "learning_rate": 9.375090671695925e-05, "loss": 0.0939, "step": 1823 }, { "epoch": 0.19774501300954034, "grad_norm": 0.4346919059753418, "learning_rate": 9.37472798491223e-05, "loss": 0.0384, "step": 1824 }, { "epoch": 0.19785342584562013, "grad_norm": 0.6133635640144348, "learning_rate": 9.374365298128536e-05, "loss": 0.0167, "step": 1825 }, { "epoch": 0.19796183868169992, "grad_norm": 0.292467325925827, "learning_rate": 9.374002611344843e-05, "loss": 0.033, "step": 1826 }, { "epoch": 0.1980702515177797, "grad_norm": 0.42414119839668274, "learning_rate": 9.373639924561149e-05, "loss": 0.0446, "step": 1827 }, { "epoch": 0.1981786643538595, "grad_norm": 0.4876108467578888, "learning_rate": 9.373277237777456e-05, "loss": 0.0995, "step": 1828 }, { "epoch": 0.1982870771899393, "grad_norm": 0.24194657802581787, "learning_rate": 9.372914550993762e-05, "loss": 0.048, "step": 1829 }, { "epoch": 0.19839549002601908, "grad_norm": 0.22618354856967926, "learning_rate": 9.372551864210069e-05, "loss": 0.0318, "step": 1830 }, { "epoch": 0.19850390286209887, "grad_norm": 1.9662270545959473, "learning_rate": 9.372189177426375e-05, "loss": 0.0371, "step": 1831 }, { "epoch": 0.19861231569817867, "grad_norm": 0.1263163983821869, "learning_rate": 9.371826490642682e-05, "loss": 0.0131, "step": 1832 }, { "epoch": 0.19872072853425846, "grad_norm": 0.572857677936554, "learning_rate": 9.371463803858988e-05, "loss": 0.0657, "step": 1833 }, { "epoch": 0.19882914137033825, "grad_norm": 0.3986412584781647, "learning_rate": 9.371101117075293e-05, "loss": 0.0594, "step": 1834 }, { "epoch": 0.19893755420641804, "grad_norm": 0.7520245909690857, "learning_rate": 9.3707384302916e-05, "loss": 0.1042, "step": 1835 }, { "epoch": 0.19904596704249783, "grad_norm": 0.5121050477027893, "learning_rate": 9.370375743507906e-05, "loss": 0.0313, "step": 1836 }, { "epoch": 0.19915437987857762, "grad_norm": 0.48687031865119934, "learning_rate": 9.370013056724215e-05, "loss": 0.0347, "step": 1837 }, { "epoch": 0.1992627927146574, "grad_norm": 1.0892257690429688, "learning_rate": 9.36965036994052e-05, "loss": 0.0835, "step": 1838 }, { "epoch": 0.1993712055507372, "grad_norm": 0.4647352993488312, "learning_rate": 9.369287683156826e-05, "loss": 0.0943, "step": 1839 }, { "epoch": 0.199479618386817, "grad_norm": 0.18648584187030792, "learning_rate": 9.368924996373133e-05, "loss": 0.0276, "step": 1840 }, { "epoch": 0.19958803122289678, "grad_norm": 0.48943811655044556, "learning_rate": 9.368562309589439e-05, "loss": 0.0733, "step": 1841 }, { "epoch": 0.19969644405897657, "grad_norm": 0.2518700361251831, "learning_rate": 9.368199622805746e-05, "loss": 0.022, "step": 1842 }, { "epoch": 0.19980485689505637, "grad_norm": 0.22989657521247864, "learning_rate": 9.367836936022052e-05, "loss": 0.0257, "step": 1843 }, { "epoch": 0.19991326973113616, "grad_norm": 0.2759210169315338, "learning_rate": 9.367474249238359e-05, "loss": 0.0536, "step": 1844 }, { "epoch": 0.20002168256721595, "grad_norm": 0.45636287331581116, "learning_rate": 9.367111562454664e-05, "loss": 0.0768, "step": 1845 }, { "epoch": 0.20013009540329574, "grad_norm": 0.45659497380256653, "learning_rate": 9.366748875670972e-05, "loss": 0.0613, "step": 1846 }, { "epoch": 0.20023850823937553, "grad_norm": 0.49154365062713623, "learning_rate": 9.366386188887277e-05, "loss": 0.0648, "step": 1847 }, { "epoch": 0.20034692107545535, "grad_norm": 0.4541441798210144, "learning_rate": 9.366023502103583e-05, "loss": 0.0347, "step": 1848 }, { "epoch": 0.20045533391153514, "grad_norm": 0.4657798707485199, "learning_rate": 9.36566081531989e-05, "loss": 0.0492, "step": 1849 }, { "epoch": 0.20056374674761493, "grad_norm": 0.2851807177066803, "learning_rate": 9.365298128536196e-05, "loss": 0.0354, "step": 1850 }, { "epoch": 0.20067215958369472, "grad_norm": 0.35157591104507446, "learning_rate": 9.364935441752503e-05, "loss": 0.0637, "step": 1851 }, { "epoch": 0.2007805724197745, "grad_norm": 0.9336418509483337, "learning_rate": 9.364572754968809e-05, "loss": 0.0705, "step": 1852 }, { "epoch": 0.2008889852558543, "grad_norm": 0.26079463958740234, "learning_rate": 9.364210068185116e-05, "loss": 0.0274, "step": 1853 }, { "epoch": 0.2009973980919341, "grad_norm": 0.39248549938201904, "learning_rate": 9.363847381401423e-05, "loss": 0.074, "step": 1854 }, { "epoch": 0.20110581092801388, "grad_norm": 0.32014042139053345, "learning_rate": 9.363484694617729e-05, "loss": 0.0659, "step": 1855 }, { "epoch": 0.20121422376409367, "grad_norm": 0.14842045307159424, "learning_rate": 9.363122007834036e-05, "loss": 0.0274, "step": 1856 }, { "epoch": 0.20132263660017347, "grad_norm": 0.7381591200828552, "learning_rate": 9.362759321050341e-05, "loss": 0.0664, "step": 1857 }, { "epoch": 0.20143104943625326, "grad_norm": 0.3155174255371094, "learning_rate": 9.362396634266648e-05, "loss": 0.0439, "step": 1858 }, { "epoch": 0.20153946227233305, "grad_norm": 0.38256970047950745, "learning_rate": 9.362033947482954e-05, "loss": 0.0315, "step": 1859 }, { "epoch": 0.20164787510841284, "grad_norm": 0.40885215997695923, "learning_rate": 9.361671260699261e-05, "loss": 0.0442, "step": 1860 }, { "epoch": 0.20175628794449263, "grad_norm": 0.3554856479167938, "learning_rate": 9.361308573915567e-05, "loss": 0.0746, "step": 1861 }, { "epoch": 0.20186470078057242, "grad_norm": 0.3837777376174927, "learning_rate": 9.360945887131873e-05, "loss": 0.0386, "step": 1862 }, { "epoch": 0.2019731136166522, "grad_norm": 0.6471855044364929, "learning_rate": 9.36058320034818e-05, "loss": 0.0766, "step": 1863 }, { "epoch": 0.202081526452732, "grad_norm": 0.3381631076335907, "learning_rate": 9.360220513564486e-05, "loss": 0.0607, "step": 1864 }, { "epoch": 0.2021899392888118, "grad_norm": 0.3351169526576996, "learning_rate": 9.359857826780793e-05, "loss": 0.0566, "step": 1865 }, { "epoch": 0.20229835212489158, "grad_norm": 0.7021968364715576, "learning_rate": 9.359495139997098e-05, "loss": 0.0669, "step": 1866 }, { "epoch": 0.20240676496097137, "grad_norm": 0.38997822999954224, "learning_rate": 9.359132453213405e-05, "loss": 0.0432, "step": 1867 }, { "epoch": 0.20251517779705117, "grad_norm": 0.48798519372940063, "learning_rate": 9.358769766429711e-05, "loss": 0.0619, "step": 1868 }, { "epoch": 0.20262359063313096, "grad_norm": 0.15924888849258423, "learning_rate": 9.358407079646018e-05, "loss": 0.0166, "step": 1869 }, { "epoch": 0.20273200346921075, "grad_norm": 0.3038497567176819, "learning_rate": 9.358044392862324e-05, "loss": 0.044, "step": 1870 }, { "epoch": 0.20284041630529054, "grad_norm": 0.8013809323310852, "learning_rate": 9.357681706078631e-05, "loss": 0.0955, "step": 1871 }, { "epoch": 0.20294882914137033, "grad_norm": 0.5087195038795471, "learning_rate": 9.357319019294938e-05, "loss": 0.0642, "step": 1872 }, { "epoch": 0.20305724197745012, "grad_norm": 0.203656405210495, "learning_rate": 9.356956332511244e-05, "loss": 0.0315, "step": 1873 }, { "epoch": 0.2031656548135299, "grad_norm": 0.2593593895435333, "learning_rate": 9.356593645727551e-05, "loss": 0.0289, "step": 1874 }, { "epoch": 0.2032740676496097, "grad_norm": 0.426727831363678, "learning_rate": 9.356230958943857e-05, "loss": 0.0465, "step": 1875 }, { "epoch": 0.2033824804856895, "grad_norm": 0.3820761442184448, "learning_rate": 9.355868272160163e-05, "loss": 0.0558, "step": 1876 }, { "epoch": 0.20349089332176928, "grad_norm": 0.28153982758522034, "learning_rate": 9.35550558537647e-05, "loss": 0.0609, "step": 1877 }, { "epoch": 0.2035993061578491, "grad_norm": 0.3275964856147766, "learning_rate": 9.355142898592775e-05, "loss": 0.0346, "step": 1878 }, { "epoch": 0.2037077189939289, "grad_norm": 0.4195462167263031, "learning_rate": 9.354780211809082e-05, "loss": 0.0495, "step": 1879 }, { "epoch": 0.20381613183000868, "grad_norm": 0.5680849552154541, "learning_rate": 9.354417525025388e-05, "loss": 0.0996, "step": 1880 }, { "epoch": 0.20392454466608848, "grad_norm": 0.23448312282562256, "learning_rate": 9.354054838241695e-05, "loss": 0.0263, "step": 1881 }, { "epoch": 0.20403295750216827, "grad_norm": 0.4722982347011566, "learning_rate": 9.353692151458001e-05, "loss": 0.0858, "step": 1882 }, { "epoch": 0.20414137033824806, "grad_norm": 0.6240682601928711, "learning_rate": 9.353329464674308e-05, "loss": 0.0324, "step": 1883 }, { "epoch": 0.20424978317432785, "grad_norm": 0.19755949079990387, "learning_rate": 9.352966777890614e-05, "loss": 0.0225, "step": 1884 }, { "epoch": 0.20435819601040764, "grad_norm": 0.33061882853507996, "learning_rate": 9.35260409110692e-05, "loss": 0.0522, "step": 1885 }, { "epoch": 0.20446660884648743, "grad_norm": 0.45137664675712585, "learning_rate": 9.352241404323227e-05, "loss": 0.0413, "step": 1886 }, { "epoch": 0.20457502168256722, "grad_norm": 0.4625245928764343, "learning_rate": 9.351878717539532e-05, "loss": 0.0398, "step": 1887 }, { "epoch": 0.204683434518647, "grad_norm": 0.3881548345088959, "learning_rate": 9.35151603075584e-05, "loss": 0.0379, "step": 1888 }, { "epoch": 0.2047918473547268, "grad_norm": 0.7585275173187256, "learning_rate": 9.351153343972147e-05, "loss": 0.0723, "step": 1889 }, { "epoch": 0.2049002601908066, "grad_norm": 0.4297526180744171, "learning_rate": 9.350790657188454e-05, "loss": 0.0409, "step": 1890 }, { "epoch": 0.20500867302688638, "grad_norm": 0.4526948928833008, "learning_rate": 9.35042797040476e-05, "loss": 0.0756, "step": 1891 }, { "epoch": 0.20511708586296618, "grad_norm": 0.262084037065506, "learning_rate": 9.350065283621065e-05, "loss": 0.0193, "step": 1892 }, { "epoch": 0.20522549869904597, "grad_norm": 0.3909560441970825, "learning_rate": 9.349702596837372e-05, "loss": 0.0456, "step": 1893 }, { "epoch": 0.20533391153512576, "grad_norm": 0.4774170219898224, "learning_rate": 9.349339910053678e-05, "loss": 0.0346, "step": 1894 }, { "epoch": 0.20544232437120555, "grad_norm": 0.43926987051963806, "learning_rate": 9.348977223269985e-05, "loss": 0.0621, "step": 1895 }, { "epoch": 0.20555073720728534, "grad_norm": 0.5073149800300598, "learning_rate": 9.348614536486291e-05, "loss": 0.1097, "step": 1896 }, { "epoch": 0.20565915004336513, "grad_norm": 0.5036872029304504, "learning_rate": 9.348251849702598e-05, "loss": 0.0777, "step": 1897 }, { "epoch": 0.20576756287944492, "grad_norm": 0.29218193888664246, "learning_rate": 9.347889162918904e-05, "loss": 0.0282, "step": 1898 }, { "epoch": 0.2058759757155247, "grad_norm": 0.28458407521247864, "learning_rate": 9.347526476135209e-05, "loss": 0.0275, "step": 1899 }, { "epoch": 0.2059843885516045, "grad_norm": 0.46748659014701843, "learning_rate": 9.347163789351516e-05, "loss": 0.0248, "step": 1900 }, { "epoch": 0.2060928013876843, "grad_norm": 0.3344772160053253, "learning_rate": 9.346801102567822e-05, "loss": 0.0361, "step": 1901 }, { "epoch": 0.20620121422376408, "grad_norm": 0.2716803252696991, "learning_rate": 9.346438415784129e-05, "loss": 0.027, "step": 1902 }, { "epoch": 0.20630962705984388, "grad_norm": 0.5873724222183228, "learning_rate": 9.346075729000435e-05, "loss": 0.0452, "step": 1903 }, { "epoch": 0.20641803989592367, "grad_norm": 0.41145795583724976, "learning_rate": 9.345713042216742e-05, "loss": 0.0295, "step": 1904 }, { "epoch": 0.20652645273200346, "grad_norm": 0.42718350887298584, "learning_rate": 9.345350355433048e-05, "loss": 0.0813, "step": 1905 }, { "epoch": 0.20663486556808325, "grad_norm": 0.17665347456932068, "learning_rate": 9.344987668649355e-05, "loss": 0.0189, "step": 1906 }, { "epoch": 0.20674327840416307, "grad_norm": 0.26521551609039307, "learning_rate": 9.344624981865662e-05, "loss": 0.0455, "step": 1907 }, { "epoch": 0.20685169124024286, "grad_norm": 0.22423161566257477, "learning_rate": 9.344262295081968e-05, "loss": 0.0201, "step": 1908 }, { "epoch": 0.20696010407632265, "grad_norm": 0.5513858199119568, "learning_rate": 9.343899608298275e-05, "loss": 0.0929, "step": 1909 }, { "epoch": 0.20706851691240244, "grad_norm": 0.48707398772239685, "learning_rate": 9.34353692151458e-05, "loss": 0.0424, "step": 1910 }, { "epoch": 0.20717692974848223, "grad_norm": 0.5838791728019714, "learning_rate": 9.343174234730888e-05, "loss": 0.0497, "step": 1911 }, { "epoch": 0.20728534258456202, "grad_norm": 0.14672298729419708, "learning_rate": 9.342811547947193e-05, "loss": 0.0188, "step": 1912 }, { "epoch": 0.2073937554206418, "grad_norm": 0.8320350646972656, "learning_rate": 9.3424488611635e-05, "loss": 0.0834, "step": 1913 }, { "epoch": 0.2075021682567216, "grad_norm": 0.5759221315383911, "learning_rate": 9.342086174379806e-05, "loss": 0.0618, "step": 1914 }, { "epoch": 0.2076105810928014, "grad_norm": 0.35868197679519653, "learning_rate": 9.341723487596112e-05, "loss": 0.0322, "step": 1915 }, { "epoch": 0.20771899392888118, "grad_norm": 0.26712003350257874, "learning_rate": 9.341360800812419e-05, "loss": 0.0294, "step": 1916 }, { "epoch": 0.20782740676496098, "grad_norm": 0.3016356825828552, "learning_rate": 9.340998114028725e-05, "loss": 0.0331, "step": 1917 }, { "epoch": 0.20793581960104077, "grad_norm": 0.38017281889915466, "learning_rate": 9.340635427245032e-05, "loss": 0.0423, "step": 1918 }, { "epoch": 0.20804423243712056, "grad_norm": 0.43762651085853577, "learning_rate": 9.340272740461337e-05, "loss": 0.0707, "step": 1919 }, { "epoch": 0.20815264527320035, "grad_norm": 0.3226134181022644, "learning_rate": 9.339910053677645e-05, "loss": 0.0352, "step": 1920 }, { "epoch": 0.20826105810928014, "grad_norm": 0.5491843223571777, "learning_rate": 9.33954736689395e-05, "loss": 0.0904, "step": 1921 }, { "epoch": 0.20836947094535993, "grad_norm": 0.4931027293205261, "learning_rate": 9.339184680110257e-05, "loss": 0.0594, "step": 1922 }, { "epoch": 0.20847788378143972, "grad_norm": 0.37040796875953674, "learning_rate": 9.338821993326564e-05, "loss": 0.0613, "step": 1923 }, { "epoch": 0.2085862966175195, "grad_norm": 0.43482205271720886, "learning_rate": 9.33845930654287e-05, "loss": 0.0486, "step": 1924 }, { "epoch": 0.2086947094535993, "grad_norm": 0.5036557912826538, "learning_rate": 9.338096619759177e-05, "loss": 0.0583, "step": 1925 }, { "epoch": 0.2088031222896791, "grad_norm": 0.31751376390457153, "learning_rate": 9.337733932975483e-05, "loss": 0.024, "step": 1926 }, { "epoch": 0.20891153512575888, "grad_norm": 0.7284859418869019, "learning_rate": 9.33737124619179e-05, "loss": 0.0407, "step": 1927 }, { "epoch": 0.20901994796183868, "grad_norm": 0.3653315603733063, "learning_rate": 9.337008559408096e-05, "loss": 0.0328, "step": 1928 }, { "epoch": 0.20912836079791847, "grad_norm": 0.6908511519432068, "learning_rate": 9.336645872624402e-05, "loss": 0.03, "step": 1929 }, { "epoch": 0.20923677363399826, "grad_norm": 0.21029339730739594, "learning_rate": 9.336283185840709e-05, "loss": 0.0195, "step": 1930 }, { "epoch": 0.20934518647007805, "grad_norm": 0.2991075813770294, "learning_rate": 9.335920499057014e-05, "loss": 0.0231, "step": 1931 }, { "epoch": 0.20945359930615784, "grad_norm": 0.1648888885974884, "learning_rate": 9.335557812273321e-05, "loss": 0.0475, "step": 1932 }, { "epoch": 0.20956201214223763, "grad_norm": 0.5876095294952393, "learning_rate": 9.335195125489627e-05, "loss": 0.0549, "step": 1933 }, { "epoch": 0.20967042497831742, "grad_norm": 0.4994608759880066, "learning_rate": 9.334832438705934e-05, "loss": 0.0625, "step": 1934 }, { "epoch": 0.2097788378143972, "grad_norm": 0.36402660608291626, "learning_rate": 9.33446975192224e-05, "loss": 0.0219, "step": 1935 }, { "epoch": 0.209887250650477, "grad_norm": 1.262856125831604, "learning_rate": 9.334107065138547e-05, "loss": 0.063, "step": 1936 }, { "epoch": 0.20999566348655682, "grad_norm": 0.6673805713653564, "learning_rate": 9.333744378354853e-05, "loss": 0.0992, "step": 1937 }, { "epoch": 0.2101040763226366, "grad_norm": 0.7277297973632812, "learning_rate": 9.333381691571159e-05, "loss": 0.0536, "step": 1938 }, { "epoch": 0.2102124891587164, "grad_norm": 1.5479930639266968, "learning_rate": 9.333019004787466e-05, "loss": 0.0966, "step": 1939 }, { "epoch": 0.2103209019947962, "grad_norm": 0.5650029182434082, "learning_rate": 9.332656318003773e-05, "loss": 0.0333, "step": 1940 }, { "epoch": 0.21042931483087599, "grad_norm": 0.2785952091217041, "learning_rate": 9.33229363122008e-05, "loss": 0.0262, "step": 1941 }, { "epoch": 0.21053772766695578, "grad_norm": 0.37343353033065796, "learning_rate": 9.331930944436386e-05, "loss": 0.0279, "step": 1942 }, { "epoch": 0.21064614050303557, "grad_norm": 0.3643409311771393, "learning_rate": 9.331568257652691e-05, "loss": 0.0373, "step": 1943 }, { "epoch": 0.21075455333911536, "grad_norm": 0.6808802485466003, "learning_rate": 9.331205570868998e-05, "loss": 0.0393, "step": 1944 }, { "epoch": 0.21086296617519515, "grad_norm": 0.2568645775318146, "learning_rate": 9.330842884085304e-05, "loss": 0.0179, "step": 1945 }, { "epoch": 0.21097137901127494, "grad_norm": 0.4336816370487213, "learning_rate": 9.330480197301611e-05, "loss": 0.0696, "step": 1946 }, { "epoch": 0.21107979184735473, "grad_norm": 0.1279820203781128, "learning_rate": 9.330117510517917e-05, "loss": 0.0137, "step": 1947 }, { "epoch": 0.21118820468343452, "grad_norm": 0.2864289879798889, "learning_rate": 9.329754823734224e-05, "loss": 0.0436, "step": 1948 }, { "epoch": 0.2112966175195143, "grad_norm": 0.29142990708351135, "learning_rate": 9.32939213695053e-05, "loss": 0.0483, "step": 1949 }, { "epoch": 0.2114050303555941, "grad_norm": 0.3214363157749176, "learning_rate": 9.329029450166837e-05, "loss": 0.0631, "step": 1950 }, { "epoch": 0.2115134431916739, "grad_norm": 0.6036995649337769, "learning_rate": 9.328666763383143e-05, "loss": 0.0721, "step": 1951 }, { "epoch": 0.21162185602775369, "grad_norm": 1.1853080987930298, "learning_rate": 9.328304076599448e-05, "loss": 0.0758, "step": 1952 }, { "epoch": 0.21173026886383348, "grad_norm": 0.3483993113040924, "learning_rate": 9.327941389815755e-05, "loss": 0.0376, "step": 1953 }, { "epoch": 0.21183868169991327, "grad_norm": 0.3704655170440674, "learning_rate": 9.327578703032061e-05, "loss": 0.0595, "step": 1954 }, { "epoch": 0.21194709453599306, "grad_norm": 0.40023234486579895, "learning_rate": 9.327216016248368e-05, "loss": 0.0337, "step": 1955 }, { "epoch": 0.21205550737207285, "grad_norm": 0.62827467918396, "learning_rate": 9.326853329464674e-05, "loss": 0.0718, "step": 1956 }, { "epoch": 0.21216392020815264, "grad_norm": 0.13466645777225494, "learning_rate": 9.326490642680981e-05, "loss": 0.0153, "step": 1957 }, { "epoch": 0.21227233304423243, "grad_norm": 0.2497139424085617, "learning_rate": 9.326127955897288e-05, "loss": 0.0414, "step": 1958 }, { "epoch": 0.21238074588031222, "grad_norm": 0.19813598692417145, "learning_rate": 9.325765269113594e-05, "loss": 0.0366, "step": 1959 }, { "epoch": 0.212489158716392, "grad_norm": 1.7429890632629395, "learning_rate": 9.325402582329901e-05, "loss": 0.0391, "step": 1960 }, { "epoch": 0.2125975715524718, "grad_norm": 0.3129916489124298, "learning_rate": 9.325039895546207e-05, "loss": 0.0547, "step": 1961 }, { "epoch": 0.2127059843885516, "grad_norm": 0.38655999302864075, "learning_rate": 9.324677208762514e-05, "loss": 0.0508, "step": 1962 }, { "epoch": 0.21281439722463139, "grad_norm": 0.5178228616714478, "learning_rate": 9.32431452197882e-05, "loss": 0.0942, "step": 1963 }, { "epoch": 0.21292281006071118, "grad_norm": 0.23647397756576538, "learning_rate": 9.323951835195127e-05, "loss": 0.0297, "step": 1964 }, { "epoch": 0.21303122289679097, "grad_norm": 0.18773286044597626, "learning_rate": 9.323589148411432e-05, "loss": 0.0244, "step": 1965 }, { "epoch": 0.21313963573287076, "grad_norm": 0.32494184374809265, "learning_rate": 9.323226461627738e-05, "loss": 0.0454, "step": 1966 }, { "epoch": 0.21324804856895058, "grad_norm": 0.3791286051273346, "learning_rate": 9.322863774844045e-05, "loss": 0.0499, "step": 1967 }, { "epoch": 0.21335646140503037, "grad_norm": 0.31107109785079956, "learning_rate": 9.322501088060351e-05, "loss": 0.0436, "step": 1968 }, { "epoch": 0.21346487424111016, "grad_norm": 0.4234563708305359, "learning_rate": 9.322138401276658e-05, "loss": 0.0491, "step": 1969 }, { "epoch": 0.21357328707718995, "grad_norm": 0.4621122181415558, "learning_rate": 9.321775714492964e-05, "loss": 0.0512, "step": 1970 }, { "epoch": 0.21368169991326974, "grad_norm": 0.5846109986305237, "learning_rate": 9.321413027709271e-05, "loss": 0.0658, "step": 1971 }, { "epoch": 0.21379011274934953, "grad_norm": 0.4290162920951843, "learning_rate": 9.321050340925577e-05, "loss": 0.05, "step": 1972 }, { "epoch": 0.21389852558542932, "grad_norm": 0.3445437252521515, "learning_rate": 9.320687654141884e-05, "loss": 0.0247, "step": 1973 }, { "epoch": 0.2140069384215091, "grad_norm": 0.5294432640075684, "learning_rate": 9.32032496735819e-05, "loss": 0.0401, "step": 1974 }, { "epoch": 0.2141153512575889, "grad_norm": 0.23980940878391266, "learning_rate": 9.319962280574496e-05, "loss": 0.0195, "step": 1975 }, { "epoch": 0.2142237640936687, "grad_norm": 0.5338749289512634, "learning_rate": 9.319599593790804e-05, "loss": 0.0299, "step": 1976 }, { "epoch": 0.21433217692974849, "grad_norm": 0.35832518339157104, "learning_rate": 9.319236907007109e-05, "loss": 0.1091, "step": 1977 }, { "epoch": 0.21444058976582828, "grad_norm": 0.9058534502983093, "learning_rate": 9.318874220223416e-05, "loss": 0.1324, "step": 1978 }, { "epoch": 0.21454900260190807, "grad_norm": 1.0465757846832275, "learning_rate": 9.318511533439722e-05, "loss": 0.0665, "step": 1979 }, { "epoch": 0.21465741543798786, "grad_norm": 0.355084091424942, "learning_rate": 9.318148846656028e-05, "loss": 0.0422, "step": 1980 }, { "epoch": 0.21476582827406765, "grad_norm": 0.39710819721221924, "learning_rate": 9.317786159872335e-05, "loss": 0.0705, "step": 1981 }, { "epoch": 0.21487424111014744, "grad_norm": 0.42569661140441895, "learning_rate": 9.31742347308864e-05, "loss": 0.0574, "step": 1982 }, { "epoch": 0.21498265394622723, "grad_norm": 0.46068868041038513, "learning_rate": 9.317060786304948e-05, "loss": 0.0228, "step": 1983 }, { "epoch": 0.21509106678230702, "grad_norm": 0.5827615857124329, "learning_rate": 9.316698099521253e-05, "loss": 0.0619, "step": 1984 }, { "epoch": 0.2151994796183868, "grad_norm": 0.30977073311805725, "learning_rate": 9.31633541273756e-05, "loss": 0.0305, "step": 1985 }, { "epoch": 0.2153078924544666, "grad_norm": 0.3299778401851654, "learning_rate": 9.315972725953866e-05, "loss": 0.0392, "step": 1986 }, { "epoch": 0.2154163052905464, "grad_norm": 0.26937228441238403, "learning_rate": 9.315610039170173e-05, "loss": 0.0329, "step": 1987 }, { "epoch": 0.21552471812662619, "grad_norm": 0.4172801077365875, "learning_rate": 9.315247352386479e-05, "loss": 0.1212, "step": 1988 }, { "epoch": 0.21563313096270598, "grad_norm": 0.3894600570201874, "learning_rate": 9.314884665602785e-05, "loss": 0.0583, "step": 1989 }, { "epoch": 0.21574154379878577, "grad_norm": 0.526125967502594, "learning_rate": 9.314521978819092e-05, "loss": 0.1014, "step": 1990 }, { "epoch": 0.21584995663486556, "grad_norm": 0.3266701400279999, "learning_rate": 9.314159292035398e-05, "loss": 0.0368, "step": 1991 }, { "epoch": 0.21595836947094535, "grad_norm": 0.5061041116714478, "learning_rate": 9.313796605251706e-05, "loss": 0.1097, "step": 1992 }, { "epoch": 0.21606678230702514, "grad_norm": 0.25241971015930176, "learning_rate": 9.313433918468012e-05, "loss": 0.0486, "step": 1993 }, { "epoch": 0.21617519514310493, "grad_norm": 0.5856354832649231, "learning_rate": 9.313071231684319e-05, "loss": 0.069, "step": 1994 }, { "epoch": 0.21628360797918472, "grad_norm": 0.49345654249191284, "learning_rate": 9.312708544900625e-05, "loss": 0.0326, "step": 1995 }, { "epoch": 0.21639202081526454, "grad_norm": 0.5274739265441895, "learning_rate": 9.31234585811693e-05, "loss": 0.0522, "step": 1996 }, { "epoch": 0.21650043365134433, "grad_norm": 0.4152091145515442, "learning_rate": 9.311983171333237e-05, "loss": 0.0443, "step": 1997 }, { "epoch": 0.21660884648742412, "grad_norm": 0.1927947849035263, "learning_rate": 9.311620484549543e-05, "loss": 0.0325, "step": 1998 }, { "epoch": 0.2167172593235039, "grad_norm": 0.470066100358963, "learning_rate": 9.31125779776585e-05, "loss": 0.0678, "step": 1999 }, { "epoch": 0.2168256721595837, "grad_norm": 0.3526821732521057, "learning_rate": 9.310895110982156e-05, "loss": 0.0417, "step": 2000 }, { "epoch": 0.2169340849956635, "grad_norm": 0.26506105065345764, "learning_rate": 9.310532424198463e-05, "loss": 0.0605, "step": 2001 }, { "epoch": 0.2170424978317433, "grad_norm": 0.318325936794281, "learning_rate": 9.310169737414769e-05, "loss": 0.0443, "step": 2002 }, { "epoch": 0.21715091066782308, "grad_norm": 0.5461663603782654, "learning_rate": 9.309807050631075e-05, "loss": 0.0458, "step": 2003 }, { "epoch": 0.21725932350390287, "grad_norm": 0.24779468774795532, "learning_rate": 9.309444363847382e-05, "loss": 0.0367, "step": 2004 }, { "epoch": 0.21736773633998266, "grad_norm": 0.2493133395910263, "learning_rate": 9.309081677063687e-05, "loss": 0.0236, "step": 2005 }, { "epoch": 0.21747614917606245, "grad_norm": 0.16816268861293793, "learning_rate": 9.308718990279995e-05, "loss": 0.0327, "step": 2006 }, { "epoch": 0.21758456201214224, "grad_norm": 0.2724100351333618, "learning_rate": 9.3083563034963e-05, "loss": 0.0428, "step": 2007 }, { "epoch": 0.21769297484822203, "grad_norm": 0.3877571225166321, "learning_rate": 9.307993616712607e-05, "loss": 0.0855, "step": 2008 }, { "epoch": 0.21780138768430182, "grad_norm": 0.23539209365844727, "learning_rate": 9.307630929928914e-05, "loss": 0.0396, "step": 2009 }, { "epoch": 0.2179098005203816, "grad_norm": 0.4527754783630371, "learning_rate": 9.30726824314522e-05, "loss": 0.1014, "step": 2010 }, { "epoch": 0.2180182133564614, "grad_norm": 0.5078362226486206, "learning_rate": 9.306905556361527e-05, "loss": 0.0683, "step": 2011 }, { "epoch": 0.2181266261925412, "grad_norm": 0.5069011449813843, "learning_rate": 9.306542869577833e-05, "loss": 0.0674, "step": 2012 }, { "epoch": 0.218235039028621, "grad_norm": 0.261705607175827, "learning_rate": 9.30618018279414e-05, "loss": 0.0364, "step": 2013 }, { "epoch": 0.21834345186470078, "grad_norm": 0.41356968879699707, "learning_rate": 9.305817496010446e-05, "loss": 0.0156, "step": 2014 }, { "epoch": 0.21845186470078057, "grad_norm": 0.4267340898513794, "learning_rate": 9.305454809226753e-05, "loss": 0.071, "step": 2015 }, { "epoch": 0.21856027753686036, "grad_norm": 0.39078420400619507, "learning_rate": 9.305092122443059e-05, "loss": 0.0194, "step": 2016 }, { "epoch": 0.21866869037294015, "grad_norm": 0.565072238445282, "learning_rate": 9.304729435659366e-05, "loss": 0.058, "step": 2017 }, { "epoch": 0.21877710320901994, "grad_norm": 0.681118369102478, "learning_rate": 9.304366748875671e-05, "loss": 0.0883, "step": 2018 }, { "epoch": 0.21888551604509973, "grad_norm": 0.44532233476638794, "learning_rate": 9.304004062091977e-05, "loss": 0.0492, "step": 2019 }, { "epoch": 0.21899392888117952, "grad_norm": 0.680164635181427, "learning_rate": 9.303641375308284e-05, "loss": 0.0445, "step": 2020 }, { "epoch": 0.2191023417172593, "grad_norm": 0.4873110055923462, "learning_rate": 9.30327868852459e-05, "loss": 0.0342, "step": 2021 }, { "epoch": 0.2192107545533391, "grad_norm": 0.5417220592498779, "learning_rate": 9.302916001740897e-05, "loss": 0.0573, "step": 2022 }, { "epoch": 0.2193191673894189, "grad_norm": 0.08102676272392273, "learning_rate": 9.302553314957203e-05, "loss": 0.0078, "step": 2023 }, { "epoch": 0.2194275802254987, "grad_norm": 0.2713383138179779, "learning_rate": 9.30219062817351e-05, "loss": 0.0363, "step": 2024 }, { "epoch": 0.21953599306157848, "grad_norm": 0.23534895479679108, "learning_rate": 9.301827941389816e-05, "loss": 0.021, "step": 2025 }, { "epoch": 0.2196444058976583, "grad_norm": 0.17799195647239685, "learning_rate": 9.301465254606123e-05, "loss": 0.0284, "step": 2026 }, { "epoch": 0.2197528187337381, "grad_norm": 0.23366951942443848, "learning_rate": 9.30110256782243e-05, "loss": 0.0218, "step": 2027 }, { "epoch": 0.21986123156981788, "grad_norm": 0.3334996700286865, "learning_rate": 9.300739881038736e-05, "loss": 0.0345, "step": 2028 }, { "epoch": 0.21996964440589767, "grad_norm": 0.8152299523353577, "learning_rate": 9.300377194255043e-05, "loss": 0.0667, "step": 2029 }, { "epoch": 0.22007805724197746, "grad_norm": 0.16408337652683258, "learning_rate": 9.300014507471348e-05, "loss": 0.0167, "step": 2030 }, { "epoch": 0.22018647007805725, "grad_norm": 0.47593995928764343, "learning_rate": 9.299651820687655e-05, "loss": 0.0754, "step": 2031 }, { "epoch": 0.22029488291413704, "grad_norm": 0.4983815550804138, "learning_rate": 9.299289133903961e-05, "loss": 0.0452, "step": 2032 }, { "epoch": 0.22040329575021683, "grad_norm": 0.5426188111305237, "learning_rate": 9.298926447120267e-05, "loss": 0.0341, "step": 2033 }, { "epoch": 0.22051170858629662, "grad_norm": 0.3079219162464142, "learning_rate": 9.298563760336574e-05, "loss": 0.0497, "step": 2034 }, { "epoch": 0.22062012142237641, "grad_norm": 0.38061290979385376, "learning_rate": 9.29820107355288e-05, "loss": 0.0696, "step": 2035 }, { "epoch": 0.2207285342584562, "grad_norm": 0.3049779236316681, "learning_rate": 9.297838386769187e-05, "loss": 0.0258, "step": 2036 }, { "epoch": 0.220836947094536, "grad_norm": 0.47467365860939026, "learning_rate": 9.297475699985493e-05, "loss": 0.055, "step": 2037 }, { "epoch": 0.2209453599306158, "grad_norm": 0.6399380564689636, "learning_rate": 9.2971130132018e-05, "loss": 0.0568, "step": 2038 }, { "epoch": 0.22105377276669558, "grad_norm": 0.517789900302887, "learning_rate": 9.296750326418105e-05, "loss": 0.0235, "step": 2039 }, { "epoch": 0.22116218560277537, "grad_norm": 0.22483183443546295, "learning_rate": 9.296387639634412e-05, "loss": 0.0228, "step": 2040 }, { "epoch": 0.22127059843885516, "grad_norm": 0.37257707118988037, "learning_rate": 9.296024952850718e-05, "loss": 0.0508, "step": 2041 }, { "epoch": 0.22137901127493495, "grad_norm": 0.45642444491386414, "learning_rate": 9.295662266067024e-05, "loss": 0.0369, "step": 2042 }, { "epoch": 0.22148742411101474, "grad_norm": 0.4069623649120331, "learning_rate": 9.295299579283331e-05, "loss": 0.0641, "step": 2043 }, { "epoch": 0.22159583694709453, "grad_norm": 0.371986448764801, "learning_rate": 9.294936892499638e-05, "loss": 0.0443, "step": 2044 }, { "epoch": 0.22170424978317432, "grad_norm": 1.1435928344726562, "learning_rate": 9.294574205715945e-05, "loss": 0.0868, "step": 2045 }, { "epoch": 0.22181266261925411, "grad_norm": 0.5018096566200256, "learning_rate": 9.294211518932251e-05, "loss": 0.0512, "step": 2046 }, { "epoch": 0.2219210754553339, "grad_norm": 0.4253140985965729, "learning_rate": 9.293848832148557e-05, "loss": 0.0518, "step": 2047 }, { "epoch": 0.2220294882914137, "grad_norm": 0.12727078795433044, "learning_rate": 9.293486145364864e-05, "loss": 0.0177, "step": 2048 }, { "epoch": 0.2221379011274935, "grad_norm": 0.6369535326957703, "learning_rate": 9.29312345858117e-05, "loss": 0.052, "step": 2049 }, { "epoch": 0.22224631396357328, "grad_norm": 1.071216106414795, "learning_rate": 9.292760771797477e-05, "loss": 0.0573, "step": 2050 }, { "epoch": 0.22235472679965307, "grad_norm": 0.33717718720436096, "learning_rate": 9.292398085013782e-05, "loss": 0.0447, "step": 2051 }, { "epoch": 0.22246313963573286, "grad_norm": 0.4153399169445038, "learning_rate": 9.29203539823009e-05, "loss": 0.0572, "step": 2052 }, { "epoch": 0.22257155247181265, "grad_norm": 0.3084106147289276, "learning_rate": 9.291672711446395e-05, "loss": 0.0244, "step": 2053 }, { "epoch": 0.22267996530789244, "grad_norm": 0.462554395198822, "learning_rate": 9.291310024662702e-05, "loss": 0.0528, "step": 2054 }, { "epoch": 0.22278837814397226, "grad_norm": 0.5281354784965515, "learning_rate": 9.290947337879008e-05, "loss": 0.0721, "step": 2055 }, { "epoch": 0.22289679098005205, "grad_norm": 0.45039641857147217, "learning_rate": 9.290584651095314e-05, "loss": 0.0547, "step": 2056 }, { "epoch": 0.22300520381613184, "grad_norm": 0.3299214243888855, "learning_rate": 9.290221964311621e-05, "loss": 0.0312, "step": 2057 }, { "epoch": 0.22311361665221163, "grad_norm": 0.6597036719322205, "learning_rate": 9.289859277527927e-05, "loss": 0.0431, "step": 2058 }, { "epoch": 0.22322202948829142, "grad_norm": 0.6994830369949341, "learning_rate": 9.289496590744234e-05, "loss": 0.0577, "step": 2059 }, { "epoch": 0.22333044232437121, "grad_norm": 0.8101363778114319, "learning_rate": 9.28913390396054e-05, "loss": 0.1212, "step": 2060 }, { "epoch": 0.223438855160451, "grad_norm": 0.13681025803089142, "learning_rate": 9.288771217176846e-05, "loss": 0.0251, "step": 2061 }, { "epoch": 0.2235472679965308, "grad_norm": 0.6985074877738953, "learning_rate": 9.288408530393154e-05, "loss": 0.0811, "step": 2062 }, { "epoch": 0.2236556808326106, "grad_norm": 0.3495241105556488, "learning_rate": 9.288045843609459e-05, "loss": 0.0349, "step": 2063 }, { "epoch": 0.22376409366869038, "grad_norm": 0.26422178745269775, "learning_rate": 9.287683156825766e-05, "loss": 0.0332, "step": 2064 }, { "epoch": 0.22387250650477017, "grad_norm": 0.29859238862991333, "learning_rate": 9.287320470042072e-05, "loss": 0.0383, "step": 2065 }, { "epoch": 0.22398091934084996, "grad_norm": 0.577338457107544, "learning_rate": 9.286957783258379e-05, "loss": 0.0595, "step": 2066 }, { "epoch": 0.22408933217692975, "grad_norm": 0.211703822016716, "learning_rate": 9.286595096474685e-05, "loss": 0.0286, "step": 2067 }, { "epoch": 0.22419774501300954, "grad_norm": 0.2809743881225586, "learning_rate": 9.286232409690992e-05, "loss": 0.0266, "step": 2068 }, { "epoch": 0.22430615784908933, "grad_norm": 0.5065445899963379, "learning_rate": 9.285869722907298e-05, "loss": 0.0375, "step": 2069 }, { "epoch": 0.22441457068516912, "grad_norm": 0.44904834032058716, "learning_rate": 9.285507036123603e-05, "loss": 0.0415, "step": 2070 }, { "epoch": 0.22452298352124891, "grad_norm": 0.5633551478385925, "learning_rate": 9.28514434933991e-05, "loss": 0.0926, "step": 2071 }, { "epoch": 0.2246313963573287, "grad_norm": 0.4515957534313202, "learning_rate": 9.284781662556216e-05, "loss": 0.0496, "step": 2072 }, { "epoch": 0.2247398091934085, "grad_norm": 0.8696802258491516, "learning_rate": 9.284418975772523e-05, "loss": 0.0831, "step": 2073 }, { "epoch": 0.2248482220294883, "grad_norm": 0.23669485747814178, "learning_rate": 9.284056288988829e-05, "loss": 0.0182, "step": 2074 }, { "epoch": 0.22495663486556808, "grad_norm": 0.5366795063018799, "learning_rate": 9.283693602205136e-05, "loss": 0.0723, "step": 2075 }, { "epoch": 0.22506504770164787, "grad_norm": 0.21057887375354767, "learning_rate": 9.283330915421442e-05, "loss": 0.0437, "step": 2076 }, { "epoch": 0.22517346053772766, "grad_norm": 0.46507585048675537, "learning_rate": 9.282968228637749e-05, "loss": 0.0766, "step": 2077 }, { "epoch": 0.22528187337380745, "grad_norm": 0.4542667269706726, "learning_rate": 9.282605541854055e-05, "loss": 0.0533, "step": 2078 }, { "epoch": 0.22539028620988724, "grad_norm": 0.2374955713748932, "learning_rate": 9.282242855070362e-05, "loss": 0.0378, "step": 2079 }, { "epoch": 0.22549869904596703, "grad_norm": 0.3306155502796173, "learning_rate": 9.281880168286669e-05, "loss": 0.0542, "step": 2080 }, { "epoch": 0.22560711188204682, "grad_norm": 0.3315508961677551, "learning_rate": 9.281517481502975e-05, "loss": 0.0343, "step": 2081 }, { "epoch": 0.22571552471812661, "grad_norm": 0.3537636697292328, "learning_rate": 9.281154794719282e-05, "loss": 0.0445, "step": 2082 }, { "epoch": 0.2258239375542064, "grad_norm": 0.4250458776950836, "learning_rate": 9.280792107935587e-05, "loss": 0.0816, "step": 2083 }, { "epoch": 0.2259323503902862, "grad_norm": 0.20099525153636932, "learning_rate": 9.280429421151893e-05, "loss": 0.0431, "step": 2084 }, { "epoch": 0.22604076322636601, "grad_norm": 0.24469682574272156, "learning_rate": 9.2800667343682e-05, "loss": 0.0474, "step": 2085 }, { "epoch": 0.2261491760624458, "grad_norm": 0.7020736932754517, "learning_rate": 9.279704047584506e-05, "loss": 0.055, "step": 2086 }, { "epoch": 0.2262575888985256, "grad_norm": 0.23550735414028168, "learning_rate": 9.279341360800813e-05, "loss": 0.027, "step": 2087 }, { "epoch": 0.2263660017346054, "grad_norm": 0.3687695562839508, "learning_rate": 9.278978674017119e-05, "loss": 0.0513, "step": 2088 }, { "epoch": 0.22647441457068518, "grad_norm": 0.39353346824645996, "learning_rate": 9.278615987233426e-05, "loss": 0.0391, "step": 2089 }, { "epoch": 0.22658282740676497, "grad_norm": 0.3673529624938965, "learning_rate": 9.278253300449732e-05, "loss": 0.0499, "step": 2090 }, { "epoch": 0.22669124024284476, "grad_norm": 0.38415461778640747, "learning_rate": 9.277890613666039e-05, "loss": 0.0653, "step": 2091 }, { "epoch": 0.22679965307892455, "grad_norm": 0.28803813457489014, "learning_rate": 9.277527926882344e-05, "loss": 0.0324, "step": 2092 }, { "epoch": 0.22690806591500434, "grad_norm": 0.3575214445590973, "learning_rate": 9.27716524009865e-05, "loss": 0.0596, "step": 2093 }, { "epoch": 0.22701647875108413, "grad_norm": 0.3560801148414612, "learning_rate": 9.276802553314957e-05, "loss": 0.046, "step": 2094 }, { "epoch": 0.22712489158716392, "grad_norm": 0.3937680423259735, "learning_rate": 9.276439866531263e-05, "loss": 0.0703, "step": 2095 }, { "epoch": 0.22723330442324371, "grad_norm": 0.21523115038871765, "learning_rate": 9.276077179747571e-05, "loss": 0.024, "step": 2096 }, { "epoch": 0.2273417172593235, "grad_norm": 0.3401196300983429, "learning_rate": 9.275714492963877e-05, "loss": 0.0309, "step": 2097 }, { "epoch": 0.2274501300954033, "grad_norm": 0.4123780429363251, "learning_rate": 9.275351806180184e-05, "loss": 0.0599, "step": 2098 }, { "epoch": 0.2275585429314831, "grad_norm": 0.49257493019104004, "learning_rate": 9.27498911939649e-05, "loss": 0.0798, "step": 2099 }, { "epoch": 0.22766695576756288, "grad_norm": 0.3743464946746826, "learning_rate": 9.274626432612796e-05, "loss": 0.075, "step": 2100 }, { "epoch": 0.22777536860364267, "grad_norm": 0.641219437122345, "learning_rate": 9.274263745829103e-05, "loss": 0.0468, "step": 2101 }, { "epoch": 0.22788378143972246, "grad_norm": 0.2668560743331909, "learning_rate": 9.273901059045409e-05, "loss": 0.0386, "step": 2102 }, { "epoch": 0.22799219427580225, "grad_norm": 0.2569079101085663, "learning_rate": 9.273538372261716e-05, "loss": 0.0505, "step": 2103 }, { "epoch": 0.22810060711188204, "grad_norm": 0.31482207775115967, "learning_rate": 9.273175685478021e-05, "loss": 0.0572, "step": 2104 }, { "epoch": 0.22820901994796183, "grad_norm": 0.26452139019966125, "learning_rate": 9.272812998694328e-05, "loss": 0.0349, "step": 2105 }, { "epoch": 0.22831743278404162, "grad_norm": 0.46460169553756714, "learning_rate": 9.272450311910634e-05, "loss": 0.0417, "step": 2106 }, { "epoch": 0.22842584562012141, "grad_norm": 0.29989174008369446, "learning_rate": 9.272087625126941e-05, "loss": 0.0433, "step": 2107 }, { "epoch": 0.2285342584562012, "grad_norm": 0.6182345151901245, "learning_rate": 9.271724938343247e-05, "loss": 0.0442, "step": 2108 }, { "epoch": 0.228642671292281, "grad_norm": 0.445405513048172, "learning_rate": 9.271362251559553e-05, "loss": 0.0624, "step": 2109 }, { "epoch": 0.2287510841283608, "grad_norm": 0.7221338748931885, "learning_rate": 9.27099956477586e-05, "loss": 0.1164, "step": 2110 }, { "epoch": 0.22885949696444058, "grad_norm": 0.3793623745441437, "learning_rate": 9.270636877992166e-05, "loss": 0.0438, "step": 2111 }, { "epoch": 0.22896790980052037, "grad_norm": 0.4866873025894165, "learning_rate": 9.270274191208473e-05, "loss": 0.0571, "step": 2112 }, { "epoch": 0.22907632263660016, "grad_norm": 0.17491857707500458, "learning_rate": 9.26991150442478e-05, "loss": 0.0307, "step": 2113 }, { "epoch": 0.22918473547267998, "grad_norm": 1.0368331670761108, "learning_rate": 9.269548817641085e-05, "loss": 0.0554, "step": 2114 }, { "epoch": 0.22929314830875977, "grad_norm": 0.3780848979949951, "learning_rate": 9.269186130857393e-05, "loss": 0.0448, "step": 2115 }, { "epoch": 0.22940156114483956, "grad_norm": 0.35883793234825134, "learning_rate": 9.268823444073698e-05, "loss": 0.0509, "step": 2116 }, { "epoch": 0.22950997398091935, "grad_norm": 0.5177151560783386, "learning_rate": 9.268460757290005e-05, "loss": 0.0562, "step": 2117 }, { "epoch": 0.22961838681699914, "grad_norm": 0.3208698332309723, "learning_rate": 9.268098070506311e-05, "loss": 0.0261, "step": 2118 }, { "epoch": 0.22972679965307893, "grad_norm": 0.2623363435268402, "learning_rate": 9.267735383722618e-05, "loss": 0.0439, "step": 2119 }, { "epoch": 0.22983521248915872, "grad_norm": 0.42818769812583923, "learning_rate": 9.267372696938924e-05, "loss": 0.0472, "step": 2120 }, { "epoch": 0.22994362532523852, "grad_norm": 0.39621734619140625, "learning_rate": 9.267010010155231e-05, "loss": 0.0468, "step": 2121 }, { "epoch": 0.2300520381613183, "grad_norm": 0.43232518434524536, "learning_rate": 9.266647323371537e-05, "loss": 0.0645, "step": 2122 }, { "epoch": 0.2301604509973981, "grad_norm": 1.7440800666809082, "learning_rate": 9.266284636587843e-05, "loss": 0.0969, "step": 2123 }, { "epoch": 0.2302688638334779, "grad_norm": 0.2093127816915512, "learning_rate": 9.26592194980415e-05, "loss": 0.0297, "step": 2124 }, { "epoch": 0.23037727666955768, "grad_norm": 0.21554014086723328, "learning_rate": 9.265559263020455e-05, "loss": 0.0229, "step": 2125 }, { "epoch": 0.23048568950563747, "grad_norm": 0.40915051102638245, "learning_rate": 9.265196576236762e-05, "loss": 0.1051, "step": 2126 }, { "epoch": 0.23059410234171726, "grad_norm": 0.5166277289390564, "learning_rate": 9.264833889453068e-05, "loss": 0.0563, "step": 2127 }, { "epoch": 0.23070251517779705, "grad_norm": 0.48162880539894104, "learning_rate": 9.264471202669375e-05, "loss": 0.0761, "step": 2128 }, { "epoch": 0.23081092801387684, "grad_norm": 0.3449465334415436, "learning_rate": 9.264108515885681e-05, "loss": 0.0846, "step": 2129 }, { "epoch": 0.23091934084995663, "grad_norm": 0.14263641834259033, "learning_rate": 9.263745829101988e-05, "loss": 0.0126, "step": 2130 }, { "epoch": 0.23102775368603642, "grad_norm": 0.2890443503856659, "learning_rate": 9.263383142318295e-05, "loss": 0.0231, "step": 2131 }, { "epoch": 0.23113616652211622, "grad_norm": 0.17674456536769867, "learning_rate": 9.263020455534601e-05, "loss": 0.0248, "step": 2132 }, { "epoch": 0.231244579358196, "grad_norm": 0.5336780548095703, "learning_rate": 9.262657768750908e-05, "loss": 0.0633, "step": 2133 }, { "epoch": 0.2313529921942758, "grad_norm": 0.42820197343826294, "learning_rate": 9.262295081967214e-05, "loss": 0.0447, "step": 2134 }, { "epoch": 0.2314614050303556, "grad_norm": 0.8316314220428467, "learning_rate": 9.261932395183521e-05, "loss": 0.0593, "step": 2135 }, { "epoch": 0.23156981786643538, "grad_norm": 0.6067508459091187, "learning_rate": 9.261569708399827e-05, "loss": 0.022, "step": 2136 }, { "epoch": 0.23167823070251517, "grad_norm": 0.9593392014503479, "learning_rate": 9.261207021616132e-05, "loss": 0.0431, "step": 2137 }, { "epoch": 0.23178664353859496, "grad_norm": 0.3051692545413971, "learning_rate": 9.26084433483244e-05, "loss": 0.0369, "step": 2138 }, { "epoch": 0.23189505637467475, "grad_norm": 0.1833340972661972, "learning_rate": 9.260481648048745e-05, "loss": 0.0264, "step": 2139 }, { "epoch": 0.23200346921075454, "grad_norm": 0.35129955410957336, "learning_rate": 9.260118961265052e-05, "loss": 0.0295, "step": 2140 }, { "epoch": 0.23211188204683433, "grad_norm": 0.4565686285495758, "learning_rate": 9.259756274481358e-05, "loss": 0.0202, "step": 2141 }, { "epoch": 0.23222029488291412, "grad_norm": 0.7469648122787476, "learning_rate": 9.259393587697665e-05, "loss": 0.0409, "step": 2142 }, { "epoch": 0.23232870771899392, "grad_norm": 0.4394645690917969, "learning_rate": 9.259030900913971e-05, "loss": 0.0681, "step": 2143 }, { "epoch": 0.23243712055507373, "grad_norm": 0.7802729606628418, "learning_rate": 9.258668214130278e-05, "loss": 0.0998, "step": 2144 }, { "epoch": 0.23254553339115352, "grad_norm": 0.8011483550071716, "learning_rate": 9.258305527346584e-05, "loss": 0.065, "step": 2145 }, { "epoch": 0.23265394622723332, "grad_norm": 0.598229706287384, "learning_rate": 9.257942840562889e-05, "loss": 0.0369, "step": 2146 }, { "epoch": 0.2327623590633131, "grad_norm": 0.25019145011901855, "learning_rate": 9.257580153779196e-05, "loss": 0.0345, "step": 2147 }, { "epoch": 0.2328707718993929, "grad_norm": 0.2192545384168625, "learning_rate": 9.257217466995503e-05, "loss": 0.0161, "step": 2148 }, { "epoch": 0.2329791847354727, "grad_norm": 0.19189830124378204, "learning_rate": 9.25685478021181e-05, "loss": 0.0321, "step": 2149 }, { "epoch": 0.23308759757155248, "grad_norm": 1.0183638334274292, "learning_rate": 9.256492093428116e-05, "loss": 0.0652, "step": 2150 }, { "epoch": 0.23319601040763227, "grad_norm": 0.39530789852142334, "learning_rate": 9.256129406644422e-05, "loss": 0.0238, "step": 2151 }, { "epoch": 0.23330442324371206, "grad_norm": 0.8855409622192383, "learning_rate": 9.255766719860729e-05, "loss": 0.0671, "step": 2152 }, { "epoch": 0.23341283607979185, "grad_norm": 0.5177617073059082, "learning_rate": 9.255404033077035e-05, "loss": 0.0412, "step": 2153 }, { "epoch": 0.23352124891587164, "grad_norm": 0.694980263710022, "learning_rate": 9.255041346293342e-05, "loss": 0.0733, "step": 2154 }, { "epoch": 0.23362966175195143, "grad_norm": 0.6779690384864807, "learning_rate": 9.254678659509648e-05, "loss": 0.0494, "step": 2155 }, { "epoch": 0.23373807458803122, "grad_norm": 0.21759222447872162, "learning_rate": 9.254315972725955e-05, "loss": 0.0353, "step": 2156 }, { "epoch": 0.23384648742411102, "grad_norm": 0.16968651115894318, "learning_rate": 9.25395328594226e-05, "loss": 0.0155, "step": 2157 }, { "epoch": 0.2339549002601908, "grad_norm": 0.3169737160205841, "learning_rate": 9.253590599158568e-05, "loss": 0.0347, "step": 2158 }, { "epoch": 0.2340633130962706, "grad_norm": 0.7531782984733582, "learning_rate": 9.253227912374873e-05, "loss": 0.121, "step": 2159 }, { "epoch": 0.2341717259323504, "grad_norm": 0.33500802516937256, "learning_rate": 9.252865225591179e-05, "loss": 0.0499, "step": 2160 }, { "epoch": 0.23428013876843018, "grad_norm": 0.20476169884204865, "learning_rate": 9.252502538807486e-05, "loss": 0.0367, "step": 2161 }, { "epoch": 0.23438855160450997, "grad_norm": 0.5140121579170227, "learning_rate": 9.252139852023792e-05, "loss": 0.0381, "step": 2162 }, { "epoch": 0.23449696444058976, "grad_norm": 0.4234408736228943, "learning_rate": 9.251777165240099e-05, "loss": 0.0873, "step": 2163 }, { "epoch": 0.23460537727666955, "grad_norm": 0.29693377017974854, "learning_rate": 9.251414478456405e-05, "loss": 0.0385, "step": 2164 }, { "epoch": 0.23471379011274934, "grad_norm": 0.3998181223869324, "learning_rate": 9.251051791672712e-05, "loss": 0.0464, "step": 2165 }, { "epoch": 0.23482220294882913, "grad_norm": 0.6737385988235474, "learning_rate": 9.250689104889019e-05, "loss": 0.0416, "step": 2166 }, { "epoch": 0.23493061578490892, "grad_norm": 0.40438327193260193, "learning_rate": 9.250326418105325e-05, "loss": 0.0522, "step": 2167 }, { "epoch": 0.23503902862098872, "grad_norm": 0.6396470665931702, "learning_rate": 9.249963731321632e-05, "loss": 0.0441, "step": 2168 }, { "epoch": 0.2351474414570685, "grad_norm": 0.8747183680534363, "learning_rate": 9.249601044537937e-05, "loss": 0.1038, "step": 2169 }, { "epoch": 0.2352558542931483, "grad_norm": 0.2276766002178192, "learning_rate": 9.249238357754244e-05, "loss": 0.0462, "step": 2170 }, { "epoch": 0.2353642671292281, "grad_norm": 0.47201621532440186, "learning_rate": 9.24887567097055e-05, "loss": 0.0408, "step": 2171 }, { "epoch": 0.23547267996530788, "grad_norm": 0.5147926211357117, "learning_rate": 9.248512984186857e-05, "loss": 0.0618, "step": 2172 }, { "epoch": 0.2355810928013877, "grad_norm": 0.13043370842933655, "learning_rate": 9.248150297403163e-05, "loss": 0.0137, "step": 2173 }, { "epoch": 0.2356895056374675, "grad_norm": 0.9553658366203308, "learning_rate": 9.247787610619469e-05, "loss": 0.0741, "step": 2174 }, { "epoch": 0.23579791847354728, "grad_norm": 0.22011467814445496, "learning_rate": 9.247424923835776e-05, "loss": 0.0274, "step": 2175 }, { "epoch": 0.23590633130962707, "grad_norm": 0.3260127902030945, "learning_rate": 9.247062237052082e-05, "loss": 0.0416, "step": 2176 }, { "epoch": 0.23601474414570686, "grad_norm": 0.3988880217075348, "learning_rate": 9.246699550268389e-05, "loss": 0.0378, "step": 2177 }, { "epoch": 0.23612315698178665, "grad_norm": 0.6174565553665161, "learning_rate": 9.246336863484694e-05, "loss": 0.0372, "step": 2178 }, { "epoch": 0.23623156981786644, "grad_norm": 0.2149813175201416, "learning_rate": 9.245974176701001e-05, "loss": 0.028, "step": 2179 }, { "epoch": 0.23633998265394623, "grad_norm": 0.4709712862968445, "learning_rate": 9.245611489917307e-05, "loss": 0.0832, "step": 2180 }, { "epoch": 0.23644839549002603, "grad_norm": 0.3107605576515198, "learning_rate": 9.245248803133614e-05, "loss": 0.0344, "step": 2181 }, { "epoch": 0.23655680832610582, "grad_norm": 0.3252129852771759, "learning_rate": 9.244886116349921e-05, "loss": 0.0181, "step": 2182 }, { "epoch": 0.2366652211621856, "grad_norm": 0.3653825521469116, "learning_rate": 9.244523429566227e-05, "loss": 0.0386, "step": 2183 }, { "epoch": 0.2367736339982654, "grad_norm": 0.2986644208431244, "learning_rate": 9.244160742782534e-05, "loss": 0.0422, "step": 2184 }, { "epoch": 0.2368820468343452, "grad_norm": 0.40322959423065186, "learning_rate": 9.24379805599884e-05, "loss": 0.049, "step": 2185 }, { "epoch": 0.23699045967042498, "grad_norm": 0.4374748766422272, "learning_rate": 9.243435369215147e-05, "loss": 0.0638, "step": 2186 }, { "epoch": 0.23709887250650477, "grad_norm": 0.9200934171676636, "learning_rate": 9.243072682431453e-05, "loss": 0.0748, "step": 2187 }, { "epoch": 0.23720728534258456, "grad_norm": 0.42615842819213867, "learning_rate": 9.242709995647759e-05, "loss": 0.0307, "step": 2188 }, { "epoch": 0.23731569817866435, "grad_norm": 0.4539109170436859, "learning_rate": 9.242347308864066e-05, "loss": 0.0572, "step": 2189 }, { "epoch": 0.23742411101474414, "grad_norm": 0.26159122586250305, "learning_rate": 9.241984622080371e-05, "loss": 0.0224, "step": 2190 }, { "epoch": 0.23753252385082393, "grad_norm": 0.18158991634845734, "learning_rate": 9.241621935296678e-05, "loss": 0.0237, "step": 2191 }, { "epoch": 0.23764093668690373, "grad_norm": 0.3788294494152069, "learning_rate": 9.241259248512984e-05, "loss": 0.0406, "step": 2192 }, { "epoch": 0.23774934952298352, "grad_norm": 0.3223222494125366, "learning_rate": 9.240896561729291e-05, "loss": 0.0273, "step": 2193 }, { "epoch": 0.2378577623590633, "grad_norm": 0.532886266708374, "learning_rate": 9.240533874945597e-05, "loss": 0.0958, "step": 2194 }, { "epoch": 0.2379661751951431, "grad_norm": 0.39134737849235535, "learning_rate": 9.240171188161904e-05, "loss": 0.0303, "step": 2195 }, { "epoch": 0.2380745880312229, "grad_norm": 0.4840128719806671, "learning_rate": 9.23980850137821e-05, "loss": 0.0617, "step": 2196 }, { "epoch": 0.23818300086730268, "grad_norm": 0.3928483724594116, "learning_rate": 9.239445814594516e-05, "loss": 0.0441, "step": 2197 }, { "epoch": 0.23829141370338247, "grad_norm": 0.6941282749176025, "learning_rate": 9.239083127810823e-05, "loss": 0.0529, "step": 2198 }, { "epoch": 0.23839982653946226, "grad_norm": 0.3803955316543579, "learning_rate": 9.23872044102713e-05, "loss": 0.0352, "step": 2199 }, { "epoch": 0.23850823937554205, "grad_norm": 0.38564741611480713, "learning_rate": 9.238357754243437e-05, "loss": 0.0553, "step": 2200 }, { "epoch": 0.23861665221162184, "grad_norm": 0.2975044250488281, "learning_rate": 9.237995067459743e-05, "loss": 0.0346, "step": 2201 }, { "epoch": 0.23872506504770163, "grad_norm": 0.6887633204460144, "learning_rate": 9.23763238067605e-05, "loss": 0.1568, "step": 2202 }, { "epoch": 0.23883347788378145, "grad_norm": 0.20004458725452423, "learning_rate": 9.237269693892355e-05, "loss": 0.0211, "step": 2203 }, { "epoch": 0.23894189071986124, "grad_norm": 0.3219969570636749, "learning_rate": 9.236907007108661e-05, "loss": 0.0417, "step": 2204 }, { "epoch": 0.23905030355594103, "grad_norm": 0.3583080470561981, "learning_rate": 9.236544320324968e-05, "loss": 0.0285, "step": 2205 }, { "epoch": 0.23915871639202083, "grad_norm": 0.6279677748680115, "learning_rate": 9.236181633541274e-05, "loss": 0.0448, "step": 2206 }, { "epoch": 0.23926712922810062, "grad_norm": 0.42097607254981995, "learning_rate": 9.235818946757581e-05, "loss": 0.0405, "step": 2207 }, { "epoch": 0.2393755420641804, "grad_norm": 0.18063652515411377, "learning_rate": 9.235456259973887e-05, "loss": 0.0104, "step": 2208 }, { "epoch": 0.2394839549002602, "grad_norm": 0.33439821004867554, "learning_rate": 9.235093573190194e-05, "loss": 0.0325, "step": 2209 }, { "epoch": 0.23959236773634, "grad_norm": 0.35783782601356506, "learning_rate": 9.2347308864065e-05, "loss": 0.0516, "step": 2210 }, { "epoch": 0.23970078057241978, "grad_norm": 0.5379393696784973, "learning_rate": 9.234368199622807e-05, "loss": 0.0407, "step": 2211 }, { "epoch": 0.23980919340849957, "grad_norm": 0.3883744180202484, "learning_rate": 9.234005512839112e-05, "loss": 0.0608, "step": 2212 }, { "epoch": 0.23991760624457936, "grad_norm": 0.3775283992290497, "learning_rate": 9.233642826055418e-05, "loss": 0.0326, "step": 2213 }, { "epoch": 0.24002601908065915, "grad_norm": 0.5462762713432312, "learning_rate": 9.233280139271725e-05, "loss": 0.0848, "step": 2214 }, { "epoch": 0.24013443191673894, "grad_norm": 0.4027344286441803, "learning_rate": 9.232917452488031e-05, "loss": 0.029, "step": 2215 }, { "epoch": 0.24024284475281873, "grad_norm": 0.26893067359924316, "learning_rate": 9.232554765704338e-05, "loss": 0.0313, "step": 2216 }, { "epoch": 0.24035125758889853, "grad_norm": 0.8573342561721802, "learning_rate": 9.232192078920645e-05, "loss": 0.0573, "step": 2217 }, { "epoch": 0.24045967042497832, "grad_norm": 0.6084429621696472, "learning_rate": 9.231829392136951e-05, "loss": 0.0686, "step": 2218 }, { "epoch": 0.2405680832610581, "grad_norm": 0.38235539197921753, "learning_rate": 9.231466705353258e-05, "loss": 0.0415, "step": 2219 }, { "epoch": 0.2406764960971379, "grad_norm": 0.3278276026248932, "learning_rate": 9.231104018569564e-05, "loss": 0.0313, "step": 2220 }, { "epoch": 0.2407849089332177, "grad_norm": 0.16433972120285034, "learning_rate": 9.230741331785871e-05, "loss": 0.0048, "step": 2221 }, { "epoch": 0.24089332176929748, "grad_norm": 0.4144666790962219, "learning_rate": 9.230378645002176e-05, "loss": 0.0537, "step": 2222 }, { "epoch": 0.24100173460537727, "grad_norm": 0.4651981592178345, "learning_rate": 9.230015958218484e-05, "loss": 0.0448, "step": 2223 }, { "epoch": 0.24111014744145706, "grad_norm": 0.7850832343101501, "learning_rate": 9.229653271434789e-05, "loss": 0.0167, "step": 2224 }, { "epoch": 0.24121856027753685, "grad_norm": 0.9463955760002136, "learning_rate": 9.229290584651096e-05, "loss": 0.1196, "step": 2225 }, { "epoch": 0.24132697311361664, "grad_norm": 0.3008718192577362, "learning_rate": 9.228927897867402e-05, "loss": 0.0568, "step": 2226 }, { "epoch": 0.24143538594969643, "grad_norm": 0.38360437750816345, "learning_rate": 9.228565211083708e-05, "loss": 0.0537, "step": 2227 }, { "epoch": 0.24154379878577623, "grad_norm": 0.18962392210960388, "learning_rate": 9.228202524300015e-05, "loss": 0.0252, "step": 2228 }, { "epoch": 0.24165221162185602, "grad_norm": 0.7646730542182922, "learning_rate": 9.22783983751632e-05, "loss": 0.0521, "step": 2229 }, { "epoch": 0.2417606244579358, "grad_norm": 0.19419440627098083, "learning_rate": 9.227477150732628e-05, "loss": 0.0302, "step": 2230 }, { "epoch": 0.2418690372940156, "grad_norm": 0.20669853687286377, "learning_rate": 9.227114463948933e-05, "loss": 0.0314, "step": 2231 }, { "epoch": 0.2419774501300954, "grad_norm": 0.24497845768928528, "learning_rate": 9.22675177716524e-05, "loss": 0.0229, "step": 2232 }, { "epoch": 0.2420858629661752, "grad_norm": 0.3122895061969757, "learning_rate": 9.226389090381546e-05, "loss": 0.0428, "step": 2233 }, { "epoch": 0.242194275802255, "grad_norm": 0.5348412394523621, "learning_rate": 9.226026403597853e-05, "loss": 0.0549, "step": 2234 }, { "epoch": 0.2423026886383348, "grad_norm": 0.7543100118637085, "learning_rate": 9.22566371681416e-05, "loss": 0.031, "step": 2235 }, { "epoch": 0.24241110147441458, "grad_norm": 0.3667091727256775, "learning_rate": 9.225301030030466e-05, "loss": 0.0708, "step": 2236 }, { "epoch": 0.24251951431049437, "grad_norm": 0.2774553894996643, "learning_rate": 9.224938343246773e-05, "loss": 0.0277, "step": 2237 }, { "epoch": 0.24262792714657416, "grad_norm": 0.5296193957328796, "learning_rate": 9.224575656463079e-05, "loss": 0.0565, "step": 2238 }, { "epoch": 0.24273633998265395, "grad_norm": 0.3948482573032379, "learning_rate": 9.224212969679386e-05, "loss": 0.0549, "step": 2239 }, { "epoch": 0.24284475281873374, "grad_norm": 0.5694299340248108, "learning_rate": 9.223850282895692e-05, "loss": 0.0737, "step": 2240 }, { "epoch": 0.24295316565481354, "grad_norm": 0.35481706261634827, "learning_rate": 9.223487596111998e-05, "loss": 0.0384, "step": 2241 }, { "epoch": 0.24306157849089333, "grad_norm": 0.17302171885967255, "learning_rate": 9.223124909328305e-05, "loss": 0.0177, "step": 2242 }, { "epoch": 0.24316999132697312, "grad_norm": 0.25252434611320496, "learning_rate": 9.22276222254461e-05, "loss": 0.0148, "step": 2243 }, { "epoch": 0.2432784041630529, "grad_norm": 0.30216553807258606, "learning_rate": 9.222399535760918e-05, "loss": 0.0505, "step": 2244 }, { "epoch": 0.2433868169991327, "grad_norm": 0.47149983048439026, "learning_rate": 9.222036848977223e-05, "loss": 0.07, "step": 2245 }, { "epoch": 0.2434952298352125, "grad_norm": 0.374504417181015, "learning_rate": 9.22167416219353e-05, "loss": 0.0364, "step": 2246 }, { "epoch": 0.24360364267129228, "grad_norm": 0.6487348675727844, "learning_rate": 9.221311475409836e-05, "loss": 0.0749, "step": 2247 }, { "epoch": 0.24371205550737207, "grad_norm": 0.37291207909584045, "learning_rate": 9.220948788626143e-05, "loss": 0.0536, "step": 2248 }, { "epoch": 0.24382046834345186, "grad_norm": 0.3189755380153656, "learning_rate": 9.220586101842449e-05, "loss": 0.0413, "step": 2249 }, { "epoch": 0.24392888117953165, "grad_norm": 0.7017539739608765, "learning_rate": 9.220223415058755e-05, "loss": 0.0342, "step": 2250 }, { "epoch": 0.24403729401561144, "grad_norm": 0.44124045968055725, "learning_rate": 9.219860728275063e-05, "loss": 0.0513, "step": 2251 }, { "epoch": 0.24414570685169124, "grad_norm": 0.33239325881004333, "learning_rate": 9.219498041491369e-05, "loss": 0.0416, "step": 2252 }, { "epoch": 0.24425411968777103, "grad_norm": 0.3043954372406006, "learning_rate": 9.219135354707676e-05, "loss": 0.0353, "step": 2253 }, { "epoch": 0.24436253252385082, "grad_norm": 0.5296938419342041, "learning_rate": 9.218772667923982e-05, "loss": 0.0642, "step": 2254 }, { "epoch": 0.2444709453599306, "grad_norm": 0.46936655044555664, "learning_rate": 9.218409981140287e-05, "loss": 0.0567, "step": 2255 }, { "epoch": 0.2445793581960104, "grad_norm": 0.34640753269195557, "learning_rate": 9.218047294356594e-05, "loss": 0.0906, "step": 2256 }, { "epoch": 0.2446877710320902, "grad_norm": 0.4440969228744507, "learning_rate": 9.2176846075729e-05, "loss": 0.0314, "step": 2257 }, { "epoch": 0.24479618386816998, "grad_norm": 0.4970456659793854, "learning_rate": 9.217321920789207e-05, "loss": 0.0941, "step": 2258 }, { "epoch": 0.24490459670424977, "grad_norm": 0.36565613746643066, "learning_rate": 9.216959234005513e-05, "loss": 0.0231, "step": 2259 }, { "epoch": 0.24501300954032956, "grad_norm": 0.35963955521583557, "learning_rate": 9.21659654722182e-05, "loss": 0.0608, "step": 2260 }, { "epoch": 0.24512142237640935, "grad_norm": 0.339214950799942, "learning_rate": 9.216233860438126e-05, "loss": 0.0433, "step": 2261 }, { "epoch": 0.24522983521248917, "grad_norm": 0.19883672893047333, "learning_rate": 9.215871173654433e-05, "loss": 0.0211, "step": 2262 }, { "epoch": 0.24533824804856896, "grad_norm": 0.5441051125526428, "learning_rate": 9.215508486870739e-05, "loss": 0.0269, "step": 2263 }, { "epoch": 0.24544666088464875, "grad_norm": 0.2706221342086792, "learning_rate": 9.215145800087044e-05, "loss": 0.0415, "step": 2264 }, { "epoch": 0.24555507372072854, "grad_norm": 0.3725849688053131, "learning_rate": 9.214783113303351e-05, "loss": 0.0434, "step": 2265 }, { "epoch": 0.24566348655680834, "grad_norm": 0.31270620226860046, "learning_rate": 9.214420426519657e-05, "loss": 0.0272, "step": 2266 }, { "epoch": 0.24577189939288813, "grad_norm": 0.29971006512641907, "learning_rate": 9.214057739735964e-05, "loss": 0.0304, "step": 2267 }, { "epoch": 0.24588031222896792, "grad_norm": 0.42278191447257996, "learning_rate": 9.21369505295227e-05, "loss": 0.0587, "step": 2268 }, { "epoch": 0.2459887250650477, "grad_norm": 1.1053974628448486, "learning_rate": 9.213332366168577e-05, "loss": 0.0927, "step": 2269 }, { "epoch": 0.2460971379011275, "grad_norm": 0.561114490032196, "learning_rate": 9.212969679384884e-05, "loss": 0.0326, "step": 2270 }, { "epoch": 0.2462055507372073, "grad_norm": 0.3055378198623657, "learning_rate": 9.21260699260119e-05, "loss": 0.0386, "step": 2271 }, { "epoch": 0.24631396357328708, "grad_norm": 1.0297971963882446, "learning_rate": 9.212244305817497e-05, "loss": 0.0617, "step": 2272 }, { "epoch": 0.24642237640936687, "grad_norm": 0.3521058261394501, "learning_rate": 9.211881619033803e-05, "loss": 0.0488, "step": 2273 }, { "epoch": 0.24653078924544666, "grad_norm": 0.3926519453525543, "learning_rate": 9.21151893225011e-05, "loss": 0.0458, "step": 2274 }, { "epoch": 0.24663920208152645, "grad_norm": 0.4825918674468994, "learning_rate": 9.211156245466416e-05, "loss": 0.0483, "step": 2275 }, { "epoch": 0.24674761491760624, "grad_norm": 0.5310332775115967, "learning_rate": 9.210793558682723e-05, "loss": 0.077, "step": 2276 }, { "epoch": 0.24685602775368604, "grad_norm": 0.9879180192947388, "learning_rate": 9.210430871899028e-05, "loss": 0.0658, "step": 2277 }, { "epoch": 0.24696444058976583, "grad_norm": 0.6258630156517029, "learning_rate": 9.210068185115334e-05, "loss": 0.1137, "step": 2278 }, { "epoch": 0.24707285342584562, "grad_norm": 0.5338617563247681, "learning_rate": 9.209705498331641e-05, "loss": 0.0695, "step": 2279 }, { "epoch": 0.2471812662619254, "grad_norm": 0.14581042528152466, "learning_rate": 9.209342811547947e-05, "loss": 0.0192, "step": 2280 }, { "epoch": 0.2472896790980052, "grad_norm": 0.48002874851226807, "learning_rate": 9.208980124764254e-05, "loss": 0.086, "step": 2281 }, { "epoch": 0.247398091934085, "grad_norm": 0.6991795301437378, "learning_rate": 9.20861743798056e-05, "loss": 0.0892, "step": 2282 }, { "epoch": 0.24750650477016478, "grad_norm": 0.3790808916091919, "learning_rate": 9.208254751196867e-05, "loss": 0.0449, "step": 2283 }, { "epoch": 0.24761491760624457, "grad_norm": 0.47851407527923584, "learning_rate": 9.207892064413173e-05, "loss": 0.0666, "step": 2284 }, { "epoch": 0.24772333044232436, "grad_norm": 0.37605369091033936, "learning_rate": 9.20752937762948e-05, "loss": 0.0548, "step": 2285 }, { "epoch": 0.24783174327840415, "grad_norm": 0.12000230699777603, "learning_rate": 9.207166690845787e-05, "loss": 0.0175, "step": 2286 }, { "epoch": 0.24794015611448394, "grad_norm": 0.6438184380531311, "learning_rate": 9.206804004062092e-05, "loss": 0.0811, "step": 2287 }, { "epoch": 0.24804856895056374, "grad_norm": 0.2643124461174011, "learning_rate": 9.2064413172784e-05, "loss": 0.0531, "step": 2288 }, { "epoch": 0.24815698178664353, "grad_norm": 0.24015061557292938, "learning_rate": 9.206078630494705e-05, "loss": 0.0555, "step": 2289 }, { "epoch": 0.24826539462272332, "grad_norm": 0.36641937494277954, "learning_rate": 9.205715943711012e-05, "loss": 0.069, "step": 2290 }, { "epoch": 0.2483738074588031, "grad_norm": 0.4545617699623108, "learning_rate": 9.205353256927318e-05, "loss": 0.0499, "step": 2291 }, { "epoch": 0.24848222029488293, "grad_norm": 0.3525034785270691, "learning_rate": 9.204990570143625e-05, "loss": 0.05, "step": 2292 }, { "epoch": 0.24859063313096272, "grad_norm": 0.2836877107620239, "learning_rate": 9.204627883359931e-05, "loss": 0.0732, "step": 2293 }, { "epoch": 0.2486990459670425, "grad_norm": 0.15913976728916168, "learning_rate": 9.204265196576237e-05, "loss": 0.0343, "step": 2294 }, { "epoch": 0.2488074588031223, "grad_norm": 0.29264119267463684, "learning_rate": 9.203902509792544e-05, "loss": 0.0238, "step": 2295 }, { "epoch": 0.2489158716392021, "grad_norm": 0.3332155644893646, "learning_rate": 9.20353982300885e-05, "loss": 0.0442, "step": 2296 }, { "epoch": 0.24902428447528188, "grad_norm": 0.27117717266082764, "learning_rate": 9.203177136225157e-05, "loss": 0.0261, "step": 2297 }, { "epoch": 0.24913269731136167, "grad_norm": 0.22853617370128632, "learning_rate": 9.202814449441462e-05, "loss": 0.0426, "step": 2298 }, { "epoch": 0.24924111014744146, "grad_norm": 0.3978085219860077, "learning_rate": 9.20245176265777e-05, "loss": 0.0568, "step": 2299 }, { "epoch": 0.24934952298352125, "grad_norm": 0.3845672309398651, "learning_rate": 9.202089075874075e-05, "loss": 0.0785, "step": 2300 }, { "epoch": 0.24945793581960105, "grad_norm": 0.2648000419139862, "learning_rate": 9.201726389090381e-05, "loss": 0.0338, "step": 2301 }, { "epoch": 0.24956634865568084, "grad_norm": 0.21758300065994263, "learning_rate": 9.201363702306688e-05, "loss": 0.0232, "step": 2302 }, { "epoch": 0.24967476149176063, "grad_norm": 0.31997424364089966, "learning_rate": 9.201001015522995e-05, "loss": 0.0725, "step": 2303 }, { "epoch": 0.24978317432784042, "grad_norm": 6.2783331871032715, "learning_rate": 9.200638328739302e-05, "loss": 0.2405, "step": 2304 }, { "epoch": 0.2498915871639202, "grad_norm": 0.23867784440517426, "learning_rate": 9.200275641955608e-05, "loss": 0.0458, "step": 2305 }, { "epoch": 0.25, "grad_norm": 0.2902816832065582, "learning_rate": 9.199912955171915e-05, "loss": 0.0636, "step": 2306 }, { "epoch": 0.2501084128360798, "grad_norm": 0.4652940332889557, "learning_rate": 9.199550268388221e-05, "loss": 0.0468, "step": 2307 }, { "epoch": 0.2502168256721596, "grad_norm": 0.1330506056547165, "learning_rate": 9.199187581604526e-05, "loss": 0.0275, "step": 2308 }, { "epoch": 0.2503252385082394, "grad_norm": 0.4126858711242676, "learning_rate": 9.198824894820834e-05, "loss": 0.0442, "step": 2309 }, { "epoch": 0.25043365134431916, "grad_norm": 0.717414379119873, "learning_rate": 9.198462208037139e-05, "loss": 0.0778, "step": 2310 }, { "epoch": 0.250542064180399, "grad_norm": 0.1233363077044487, "learning_rate": 9.198099521253446e-05, "loss": 0.0312, "step": 2311 }, { "epoch": 0.25065047701647875, "grad_norm": 0.5405213236808777, "learning_rate": 9.197736834469752e-05, "loss": 0.074, "step": 2312 }, { "epoch": 0.25075888985255856, "grad_norm": 0.15010975301265717, "learning_rate": 9.197374147686059e-05, "loss": 0.0153, "step": 2313 }, { "epoch": 0.2508673026886383, "grad_norm": 0.09147949516773224, "learning_rate": 9.197011460902365e-05, "loss": 0.0147, "step": 2314 }, { "epoch": 0.25097571552471815, "grad_norm": 0.3372172713279724, "learning_rate": 9.196648774118672e-05, "loss": 0.0336, "step": 2315 }, { "epoch": 0.2510841283607979, "grad_norm": 1.3974580764770508, "learning_rate": 9.196286087334978e-05, "loss": 0.0662, "step": 2316 }, { "epoch": 0.2511925411968777, "grad_norm": 0.26718607544898987, "learning_rate": 9.195923400551283e-05, "loss": 0.0516, "step": 2317 }, { "epoch": 0.2513009540329575, "grad_norm": 0.452704519033432, "learning_rate": 9.19556071376759e-05, "loss": 0.0403, "step": 2318 }, { "epoch": 0.2514093668690373, "grad_norm": 0.220447838306427, "learning_rate": 9.195198026983896e-05, "loss": 0.0207, "step": 2319 }, { "epoch": 0.2515177797051171, "grad_norm": 0.5965515375137329, "learning_rate": 9.194835340200205e-05, "loss": 0.0656, "step": 2320 }, { "epoch": 0.2516261925411969, "grad_norm": 0.5257496237754822, "learning_rate": 9.19447265341651e-05, "loss": 0.0559, "step": 2321 }, { "epoch": 0.25173460537727665, "grad_norm": 0.2586042582988739, "learning_rate": 9.194109966632816e-05, "loss": 0.035, "step": 2322 }, { "epoch": 0.2518430182133565, "grad_norm": 0.5939616560935974, "learning_rate": 9.193747279849123e-05, "loss": 0.0508, "step": 2323 }, { "epoch": 0.25195143104943624, "grad_norm": 1.063650369644165, "learning_rate": 9.193384593065429e-05, "loss": 0.0706, "step": 2324 }, { "epoch": 0.25205984388551606, "grad_norm": 0.4501287341117859, "learning_rate": 9.193021906281736e-05, "loss": 0.0718, "step": 2325 }, { "epoch": 0.2521682567215958, "grad_norm": 0.1899581104516983, "learning_rate": 9.192659219498042e-05, "loss": 0.0202, "step": 2326 }, { "epoch": 0.25227666955767564, "grad_norm": 0.4009004533290863, "learning_rate": 9.192296532714349e-05, "loss": 0.0229, "step": 2327 }, { "epoch": 0.2523850823937554, "grad_norm": 0.6034108996391296, "learning_rate": 9.191933845930655e-05, "loss": 0.0293, "step": 2328 }, { "epoch": 0.2524934952298352, "grad_norm": 0.27473142743110657, "learning_rate": 9.191571159146962e-05, "loss": 0.0382, "step": 2329 }, { "epoch": 0.252601908065915, "grad_norm": 0.5316789746284485, "learning_rate": 9.191208472363267e-05, "loss": 0.061, "step": 2330 }, { "epoch": 0.2527103209019948, "grad_norm": 1.0575159788131714, "learning_rate": 9.190845785579573e-05, "loss": 0.0769, "step": 2331 }, { "epoch": 0.25281873373807456, "grad_norm": 0.10944218188524246, "learning_rate": 9.19048309879588e-05, "loss": 0.0112, "step": 2332 }, { "epoch": 0.2529271465741544, "grad_norm": 0.8666626811027527, "learning_rate": 9.190120412012186e-05, "loss": 0.0483, "step": 2333 }, { "epoch": 0.25303555941023415, "grad_norm": 0.5942503809928894, "learning_rate": 9.189757725228493e-05, "loss": 0.0312, "step": 2334 }, { "epoch": 0.25314397224631396, "grad_norm": 0.5293735265731812, "learning_rate": 9.189395038444799e-05, "loss": 0.0712, "step": 2335 }, { "epoch": 0.2532523850823938, "grad_norm": 0.3060941994190216, "learning_rate": 9.189032351661106e-05, "loss": 0.026, "step": 2336 }, { "epoch": 0.25336079791847355, "grad_norm": 0.6548306345939636, "learning_rate": 9.188669664877412e-05, "loss": 0.069, "step": 2337 }, { "epoch": 0.25346921075455336, "grad_norm": 0.21706396341323853, "learning_rate": 9.188306978093719e-05, "loss": 0.0155, "step": 2338 }, { "epoch": 0.2535776235906331, "grad_norm": 0.6150100231170654, "learning_rate": 9.187944291310026e-05, "loss": 0.0357, "step": 2339 }, { "epoch": 0.25368603642671295, "grad_norm": 0.8121621012687683, "learning_rate": 9.187581604526332e-05, "loss": 0.0871, "step": 2340 }, { "epoch": 0.2537944492627927, "grad_norm": 0.3199698030948639, "learning_rate": 9.187218917742639e-05, "loss": 0.0594, "step": 2341 }, { "epoch": 0.25390286209887253, "grad_norm": 0.43372291326522827, "learning_rate": 9.186856230958944e-05, "loss": 0.0424, "step": 2342 }, { "epoch": 0.2540112749349523, "grad_norm": 0.5337413549423218, "learning_rate": 9.186493544175251e-05, "loss": 0.0577, "step": 2343 }, { "epoch": 0.2541196877710321, "grad_norm": 0.3392901122570038, "learning_rate": 9.186130857391557e-05, "loss": 0.0316, "step": 2344 }, { "epoch": 0.2542281006071119, "grad_norm": 0.4535839259624481, "learning_rate": 9.185768170607863e-05, "loss": 0.0317, "step": 2345 }, { "epoch": 0.2543365134431917, "grad_norm": 0.2393874228000641, "learning_rate": 9.18540548382417e-05, "loss": 0.0334, "step": 2346 }, { "epoch": 0.25444492627927146, "grad_norm": 0.26099833846092224, "learning_rate": 9.185042797040476e-05, "loss": 0.0285, "step": 2347 }, { "epoch": 0.2545533391153513, "grad_norm": 0.40844249725341797, "learning_rate": 9.184680110256783e-05, "loss": 0.0505, "step": 2348 }, { "epoch": 0.25466175195143104, "grad_norm": 0.23149175941944122, "learning_rate": 9.184317423473089e-05, "loss": 0.0307, "step": 2349 }, { "epoch": 0.25477016478751086, "grad_norm": 0.34552738070487976, "learning_rate": 9.183954736689396e-05, "loss": 0.0513, "step": 2350 }, { "epoch": 0.2548785776235906, "grad_norm": 0.48869943618774414, "learning_rate": 9.183592049905701e-05, "loss": 0.0596, "step": 2351 }, { "epoch": 0.25498699045967044, "grad_norm": 0.4906224310398102, "learning_rate": 9.183229363122008e-05, "loss": 0.0361, "step": 2352 }, { "epoch": 0.2550954032957502, "grad_norm": 0.24860073626041412, "learning_rate": 9.182866676338314e-05, "loss": 0.0495, "step": 2353 }, { "epoch": 0.25520381613183, "grad_norm": 0.18434005975723267, "learning_rate": 9.18250398955462e-05, "loss": 0.0169, "step": 2354 }, { "epoch": 0.2553122289679098, "grad_norm": 0.4744463562965393, "learning_rate": 9.182141302770928e-05, "loss": 0.0567, "step": 2355 }, { "epoch": 0.2554206418039896, "grad_norm": 0.4323098659515381, "learning_rate": 9.181778615987234e-05, "loss": 0.0283, "step": 2356 }, { "epoch": 0.25552905464006936, "grad_norm": 0.24856692552566528, "learning_rate": 9.181415929203541e-05, "loss": 0.0299, "step": 2357 }, { "epoch": 0.2556374674761492, "grad_norm": 0.3753272593021393, "learning_rate": 9.181053242419847e-05, "loss": 0.0453, "step": 2358 }, { "epoch": 0.25574588031222895, "grad_norm": 0.9261971116065979, "learning_rate": 9.180690555636153e-05, "loss": 0.1204, "step": 2359 }, { "epoch": 0.25585429314830876, "grad_norm": 0.7622228264808655, "learning_rate": 9.18032786885246e-05, "loss": 0.1187, "step": 2360 }, { "epoch": 0.2559627059843885, "grad_norm": 0.4398307502269745, "learning_rate": 9.179965182068766e-05, "loss": 0.034, "step": 2361 }, { "epoch": 0.25607111882046835, "grad_norm": 0.3651435375213623, "learning_rate": 9.179602495285073e-05, "loss": 0.0275, "step": 2362 }, { "epoch": 0.2561795316565481, "grad_norm": 0.24356675148010254, "learning_rate": 9.179239808501378e-05, "loss": 0.0408, "step": 2363 }, { "epoch": 0.25628794449262793, "grad_norm": 0.28073224425315857, "learning_rate": 9.178877121717685e-05, "loss": 0.0352, "step": 2364 }, { "epoch": 0.2563963573287077, "grad_norm": 0.2065797746181488, "learning_rate": 9.178514434933991e-05, "loss": 0.0355, "step": 2365 }, { "epoch": 0.2565047701647875, "grad_norm": 1.3733820915222168, "learning_rate": 9.178151748150298e-05, "loss": 0.0636, "step": 2366 }, { "epoch": 0.25661318300086733, "grad_norm": 0.622096061706543, "learning_rate": 9.177789061366604e-05, "loss": 0.0629, "step": 2367 }, { "epoch": 0.2567215958369471, "grad_norm": 0.3289145529270172, "learning_rate": 9.17742637458291e-05, "loss": 0.0254, "step": 2368 }, { "epoch": 0.2568300086730269, "grad_norm": 0.45163217186927795, "learning_rate": 9.177063687799217e-05, "loss": 0.0696, "step": 2369 }, { "epoch": 0.2569384215091067, "grad_norm": 0.4371854364871979, "learning_rate": 9.176701001015523e-05, "loss": 0.0504, "step": 2370 }, { "epoch": 0.2570468343451865, "grad_norm": 0.502467930316925, "learning_rate": 9.17633831423183e-05, "loss": 0.0688, "step": 2371 }, { "epoch": 0.25715524718126626, "grad_norm": 0.1945924460887909, "learning_rate": 9.175975627448137e-05, "loss": 0.0221, "step": 2372 }, { "epoch": 0.2572636600173461, "grad_norm": 0.5437453985214233, "learning_rate": 9.175612940664442e-05, "loss": 0.0468, "step": 2373 }, { "epoch": 0.25737207285342584, "grad_norm": 0.5185596346855164, "learning_rate": 9.17525025388075e-05, "loss": 0.0811, "step": 2374 }, { "epoch": 0.25748048568950566, "grad_norm": 0.1999468058347702, "learning_rate": 9.174887567097055e-05, "loss": 0.0286, "step": 2375 }, { "epoch": 0.2575888985255854, "grad_norm": 0.4777458608150482, "learning_rate": 9.174524880313362e-05, "loss": 0.0506, "step": 2376 }, { "epoch": 0.25769731136166524, "grad_norm": 0.4733850955963135, "learning_rate": 9.174162193529668e-05, "loss": 0.0377, "step": 2377 }, { "epoch": 0.257805724197745, "grad_norm": 0.7610703110694885, "learning_rate": 9.173799506745975e-05, "loss": 0.0573, "step": 2378 }, { "epoch": 0.2579141370338248, "grad_norm": 0.4296698272228241, "learning_rate": 9.173436819962281e-05, "loss": 0.0496, "step": 2379 }, { "epoch": 0.2580225498699046, "grad_norm": 0.3298565149307251, "learning_rate": 9.173074133178588e-05, "loss": 0.0487, "step": 2380 }, { "epoch": 0.2581309627059844, "grad_norm": 0.3538247346878052, "learning_rate": 9.172711446394894e-05, "loss": 0.0282, "step": 2381 }, { "epoch": 0.25823937554206416, "grad_norm": 0.22327737510204315, "learning_rate": 9.1723487596112e-05, "loss": 0.0276, "step": 2382 }, { "epoch": 0.258347788378144, "grad_norm": 0.21268628537654877, "learning_rate": 9.171986072827507e-05, "loss": 0.0231, "step": 2383 }, { "epoch": 0.25845620121422375, "grad_norm": 0.5954232811927795, "learning_rate": 9.171623386043812e-05, "loss": 0.0743, "step": 2384 }, { "epoch": 0.25856461405030357, "grad_norm": 0.29073670506477356, "learning_rate": 9.17126069926012e-05, "loss": 0.0299, "step": 2385 }, { "epoch": 0.25867302688638333, "grad_norm": 0.4999508857727051, "learning_rate": 9.170898012476425e-05, "loss": 0.0783, "step": 2386 }, { "epoch": 0.25878143972246315, "grad_norm": 0.2955767512321472, "learning_rate": 9.170535325692732e-05, "loss": 0.0208, "step": 2387 }, { "epoch": 0.2588898525585429, "grad_norm": 0.552577793598175, "learning_rate": 9.170172638909038e-05, "loss": 0.042, "step": 2388 }, { "epoch": 0.25899826539462273, "grad_norm": 0.2858025133609772, "learning_rate": 9.169809952125345e-05, "loss": 0.0384, "step": 2389 }, { "epoch": 0.2591066782307025, "grad_norm": 0.3311827778816223, "learning_rate": 9.169447265341652e-05, "loss": 0.0313, "step": 2390 }, { "epoch": 0.2592150910667823, "grad_norm": 0.6231037974357605, "learning_rate": 9.169084578557958e-05, "loss": 0.0515, "step": 2391 }, { "epoch": 0.2593235039028621, "grad_norm": 0.6280006766319275, "learning_rate": 9.168721891774265e-05, "loss": 0.0301, "step": 2392 }, { "epoch": 0.2594319167389419, "grad_norm": 0.32513418793678284, "learning_rate": 9.16835920499057e-05, "loss": 0.0318, "step": 2393 }, { "epoch": 0.25954032957502166, "grad_norm": 0.4505733251571655, "learning_rate": 9.167996518206878e-05, "loss": 0.0592, "step": 2394 }, { "epoch": 0.2596487424111015, "grad_norm": 0.17731308937072754, "learning_rate": 9.167633831423183e-05, "loss": 0.0235, "step": 2395 }, { "epoch": 0.2597571552471813, "grad_norm": 0.7983034253120422, "learning_rate": 9.16727114463949e-05, "loss": 0.0468, "step": 2396 }, { "epoch": 0.25986556808326106, "grad_norm": 0.5094042420387268, "learning_rate": 9.166908457855796e-05, "loss": 0.0301, "step": 2397 }, { "epoch": 0.2599739809193409, "grad_norm": 0.406196653842926, "learning_rate": 9.166545771072102e-05, "loss": 0.0335, "step": 2398 }, { "epoch": 0.26008239375542064, "grad_norm": 0.6367236375808716, "learning_rate": 9.166183084288409e-05, "loss": 0.0227, "step": 2399 }, { "epoch": 0.26019080659150046, "grad_norm": 0.6711035966873169, "learning_rate": 9.165820397504715e-05, "loss": 0.0832, "step": 2400 }, { "epoch": 0.2602992194275802, "grad_norm": 0.7680823802947998, "learning_rate": 9.165457710721022e-05, "loss": 0.0859, "step": 2401 }, { "epoch": 0.26040763226366004, "grad_norm": 0.7667432427406311, "learning_rate": 9.165095023937328e-05, "loss": 0.041, "step": 2402 }, { "epoch": 0.2605160450997398, "grad_norm": 0.41212716698646545, "learning_rate": 9.164732337153635e-05, "loss": 0.024, "step": 2403 }, { "epoch": 0.2606244579358196, "grad_norm": 0.3275209665298462, "learning_rate": 9.16436965036994e-05, "loss": 0.0195, "step": 2404 }, { "epoch": 0.2607328707718994, "grad_norm": 0.12238983809947968, "learning_rate": 9.164006963586248e-05, "loss": 0.0095, "step": 2405 }, { "epoch": 0.2608412836079792, "grad_norm": 0.5685873031616211, "learning_rate": 9.163644276802553e-05, "loss": 0.044, "step": 2406 }, { "epoch": 0.26094969644405897, "grad_norm": 0.35234636068344116, "learning_rate": 9.16328159001886e-05, "loss": 0.031, "step": 2407 }, { "epoch": 0.2610581092801388, "grad_norm": 0.2635282874107361, "learning_rate": 9.162918903235167e-05, "loss": 0.0357, "step": 2408 }, { "epoch": 0.26116652211621855, "grad_norm": 0.3835148811340332, "learning_rate": 9.162556216451473e-05, "loss": 0.0251, "step": 2409 }, { "epoch": 0.26127493495229837, "grad_norm": 0.5095087289810181, "learning_rate": 9.16219352966778e-05, "loss": 0.0712, "step": 2410 }, { "epoch": 0.26138334778837813, "grad_norm": 2.5144505500793457, "learning_rate": 9.161830842884086e-05, "loss": 0.0571, "step": 2411 }, { "epoch": 0.26149176062445795, "grad_norm": 0.350697785615921, "learning_rate": 9.161468156100392e-05, "loss": 0.0402, "step": 2412 }, { "epoch": 0.2616001734605377, "grad_norm": 0.5164029002189636, "learning_rate": 9.161105469316699e-05, "loss": 0.0555, "step": 2413 }, { "epoch": 0.26170858629661753, "grad_norm": 0.8786524534225464, "learning_rate": 9.160742782533005e-05, "loss": 0.0897, "step": 2414 }, { "epoch": 0.2618169991326973, "grad_norm": 0.24521200358867645, "learning_rate": 9.160380095749312e-05, "loss": 0.0184, "step": 2415 }, { "epoch": 0.2619254119687771, "grad_norm": 0.4134199619293213, "learning_rate": 9.160017408965617e-05, "loss": 0.0508, "step": 2416 }, { "epoch": 0.2620338248048569, "grad_norm": 0.48895159363746643, "learning_rate": 9.159654722181924e-05, "loss": 0.0672, "step": 2417 }, { "epoch": 0.2621422376409367, "grad_norm": 0.3456800878047943, "learning_rate": 9.15929203539823e-05, "loss": 0.0398, "step": 2418 }, { "epoch": 0.26225065047701646, "grad_norm": 0.29590654373168945, "learning_rate": 9.158929348614537e-05, "loss": 0.0265, "step": 2419 }, { "epoch": 0.2623590633130963, "grad_norm": 33.67499923706055, "learning_rate": 9.158566661830843e-05, "loss": 0.1635, "step": 2420 }, { "epoch": 0.26246747614917604, "grad_norm": 0.5805784463882446, "learning_rate": 9.158203975047149e-05, "loss": 0.0354, "step": 2421 }, { "epoch": 0.26257588898525586, "grad_norm": 0.46936193108558655, "learning_rate": 9.157841288263456e-05, "loss": 0.0683, "step": 2422 }, { "epoch": 0.2626843018213356, "grad_norm": 0.4793723523616791, "learning_rate": 9.157478601479762e-05, "loss": 0.0724, "step": 2423 }, { "epoch": 0.26279271465741544, "grad_norm": 0.4899129569530487, "learning_rate": 9.15711591469607e-05, "loss": 0.0678, "step": 2424 }, { "epoch": 0.26290112749349526, "grad_norm": 1.1686248779296875, "learning_rate": 9.156753227912376e-05, "loss": 0.1302, "step": 2425 }, { "epoch": 0.263009540329575, "grad_norm": 0.2786605656147003, "learning_rate": 9.156390541128682e-05, "loss": 0.0263, "step": 2426 }, { "epoch": 0.26311795316565484, "grad_norm": 0.7944961786270142, "learning_rate": 9.156027854344989e-05, "loss": 0.0746, "step": 2427 }, { "epoch": 0.2632263660017346, "grad_norm": 0.4984382390975952, "learning_rate": 9.155665167561294e-05, "loss": 0.0475, "step": 2428 }, { "epoch": 0.2633347788378144, "grad_norm": 0.3490331172943115, "learning_rate": 9.155302480777601e-05, "loss": 0.0428, "step": 2429 }, { "epoch": 0.2634431916738942, "grad_norm": 0.2467910200357437, "learning_rate": 9.154939793993907e-05, "loss": 0.0432, "step": 2430 }, { "epoch": 0.263551604509974, "grad_norm": 0.3327254354953766, "learning_rate": 9.154577107210214e-05, "loss": 0.0473, "step": 2431 }, { "epoch": 0.26366001734605377, "grad_norm": 0.11903300881385803, "learning_rate": 9.15421442042652e-05, "loss": 0.0227, "step": 2432 }, { "epoch": 0.2637684301821336, "grad_norm": 0.38054773211479187, "learning_rate": 9.153851733642827e-05, "loss": 0.0547, "step": 2433 }, { "epoch": 0.26387684301821335, "grad_norm": 0.4839518070220947, "learning_rate": 9.153489046859133e-05, "loss": 0.0578, "step": 2434 }, { "epoch": 0.26398525585429317, "grad_norm": 0.4414035379886627, "learning_rate": 9.153126360075439e-05, "loss": 0.0533, "step": 2435 }, { "epoch": 0.26409366869037293, "grad_norm": 0.407220721244812, "learning_rate": 9.152763673291746e-05, "loss": 0.0743, "step": 2436 }, { "epoch": 0.26420208152645275, "grad_norm": 1.0041085481643677, "learning_rate": 9.152400986508051e-05, "loss": 0.0638, "step": 2437 }, { "epoch": 0.2643104943625325, "grad_norm": 0.3689354360103607, "learning_rate": 9.152038299724358e-05, "loss": 0.0494, "step": 2438 }, { "epoch": 0.26441890719861233, "grad_norm": 0.23533906042575836, "learning_rate": 9.151675612940664e-05, "loss": 0.0184, "step": 2439 }, { "epoch": 0.2645273200346921, "grad_norm": 0.7655043005943298, "learning_rate": 9.151312926156971e-05, "loss": 0.0381, "step": 2440 }, { "epoch": 0.2646357328707719, "grad_norm": 0.20135702192783356, "learning_rate": 9.150950239373278e-05, "loss": 0.0421, "step": 2441 }, { "epoch": 0.2647441457068517, "grad_norm": 0.39959901571273804, "learning_rate": 9.150587552589584e-05, "loss": 0.0474, "step": 2442 }, { "epoch": 0.2648525585429315, "grad_norm": 0.35269618034362793, "learning_rate": 9.150224865805891e-05, "loss": 0.0538, "step": 2443 }, { "epoch": 0.26496097137901126, "grad_norm": 0.29498758912086487, "learning_rate": 9.149862179022197e-05, "loss": 0.0463, "step": 2444 }, { "epoch": 0.2650693842150911, "grad_norm": 0.35475122928619385, "learning_rate": 9.149499492238504e-05, "loss": 0.0497, "step": 2445 }, { "epoch": 0.26517779705117084, "grad_norm": 0.467098593711853, "learning_rate": 9.14913680545481e-05, "loss": 0.0468, "step": 2446 }, { "epoch": 0.26528620988725066, "grad_norm": 0.18039041757583618, "learning_rate": 9.148774118671117e-05, "loss": 0.0256, "step": 2447 }, { "epoch": 0.2653946227233304, "grad_norm": 0.4538613557815552, "learning_rate": 9.148411431887423e-05, "loss": 0.059, "step": 2448 }, { "epoch": 0.26550303555941024, "grad_norm": 0.4096740186214447, "learning_rate": 9.148048745103728e-05, "loss": 0.0738, "step": 2449 }, { "epoch": 0.26561144839549, "grad_norm": 0.2274821251630783, "learning_rate": 9.147686058320035e-05, "loss": 0.0351, "step": 2450 }, { "epoch": 0.2657198612315698, "grad_norm": 0.33684873580932617, "learning_rate": 9.147323371536341e-05, "loss": 0.0378, "step": 2451 }, { "epoch": 0.2658282740676496, "grad_norm": 0.5945419073104858, "learning_rate": 9.146960684752648e-05, "loss": 0.0548, "step": 2452 }, { "epoch": 0.2659366869037294, "grad_norm": 2.023660182952881, "learning_rate": 9.146597997968954e-05, "loss": 0.0442, "step": 2453 }, { "epoch": 0.2660450997398092, "grad_norm": 0.37552410364151, "learning_rate": 9.146235311185261e-05, "loss": 0.0746, "step": 2454 }, { "epoch": 0.266153512575889, "grad_norm": 0.5309873819351196, "learning_rate": 9.145872624401567e-05, "loss": 0.0506, "step": 2455 }, { "epoch": 0.2662619254119688, "grad_norm": 0.5665913224220276, "learning_rate": 9.145509937617874e-05, "loss": 0.0787, "step": 2456 }, { "epoch": 0.26637033824804857, "grad_norm": 0.5467194318771362, "learning_rate": 9.14514725083418e-05, "loss": 0.0599, "step": 2457 }, { "epoch": 0.2664787510841284, "grad_norm": 0.18255701661109924, "learning_rate": 9.144784564050485e-05, "loss": 0.0244, "step": 2458 }, { "epoch": 0.26658716392020815, "grad_norm": 0.4780253767967224, "learning_rate": 9.144421877266794e-05, "loss": 0.0456, "step": 2459 }, { "epoch": 0.26669557675628797, "grad_norm": 0.1667231023311615, "learning_rate": 9.1440591904831e-05, "loss": 0.0159, "step": 2460 }, { "epoch": 0.26680398959236773, "grad_norm": 0.8383904695510864, "learning_rate": 9.143696503699407e-05, "loss": 0.1112, "step": 2461 }, { "epoch": 0.26691240242844755, "grad_norm": 0.27962788939476013, "learning_rate": 9.143333816915712e-05, "loss": 0.0515, "step": 2462 }, { "epoch": 0.2670208152645273, "grad_norm": 0.39601844549179077, "learning_rate": 9.142971130132018e-05, "loss": 0.0595, "step": 2463 }, { "epoch": 0.26712922810060713, "grad_norm": 0.3448551893234253, "learning_rate": 9.142608443348325e-05, "loss": 0.0343, "step": 2464 }, { "epoch": 0.2672376409366869, "grad_norm": 0.2555754482746124, "learning_rate": 9.142245756564631e-05, "loss": 0.0282, "step": 2465 }, { "epoch": 0.2673460537727667, "grad_norm": 0.6948689222335815, "learning_rate": 9.141883069780938e-05, "loss": 0.0432, "step": 2466 }, { "epoch": 0.2674544666088465, "grad_norm": 0.4467616081237793, "learning_rate": 9.141520382997244e-05, "loss": 0.0965, "step": 2467 }, { "epoch": 0.2675628794449263, "grad_norm": 0.6602675914764404, "learning_rate": 9.141157696213551e-05, "loss": 0.0815, "step": 2468 }, { "epoch": 0.26767129228100606, "grad_norm": 0.3196755647659302, "learning_rate": 9.140795009429856e-05, "loss": 0.043, "step": 2469 }, { "epoch": 0.2677797051170859, "grad_norm": 0.17496903240680695, "learning_rate": 9.140432322646164e-05, "loss": 0.0224, "step": 2470 }, { "epoch": 0.26788811795316564, "grad_norm": 0.4211765229701996, "learning_rate": 9.140069635862469e-05, "loss": 0.0582, "step": 2471 }, { "epoch": 0.26799653078924546, "grad_norm": 0.474197119474411, "learning_rate": 9.139706949078775e-05, "loss": 0.0481, "step": 2472 }, { "epoch": 0.2681049436253252, "grad_norm": 0.3673250675201416, "learning_rate": 9.139344262295082e-05, "loss": 0.0462, "step": 2473 }, { "epoch": 0.26821335646140504, "grad_norm": 0.6218931674957275, "learning_rate": 9.138981575511388e-05, "loss": 0.0673, "step": 2474 }, { "epoch": 0.2683217692974848, "grad_norm": 0.40302377939224243, "learning_rate": 9.138618888727695e-05, "loss": 0.0686, "step": 2475 }, { "epoch": 0.2684301821335646, "grad_norm": 0.5213481783866882, "learning_rate": 9.138256201944002e-05, "loss": 0.0662, "step": 2476 }, { "epoch": 0.2685385949696444, "grad_norm": 0.5094539523124695, "learning_rate": 9.137893515160309e-05, "loss": 0.0793, "step": 2477 }, { "epoch": 0.2686470078057242, "grad_norm": 0.4936218559741974, "learning_rate": 9.137530828376615e-05, "loss": 0.0696, "step": 2478 }, { "epoch": 0.26875542064180397, "grad_norm": 0.664956271648407, "learning_rate": 9.13716814159292e-05, "loss": 0.0888, "step": 2479 }, { "epoch": 0.2688638334778838, "grad_norm": 0.3536316454410553, "learning_rate": 9.136805454809228e-05, "loss": 0.0559, "step": 2480 }, { "epoch": 0.26897224631396355, "grad_norm": 0.42799508571624756, "learning_rate": 9.136442768025533e-05, "loss": 0.0591, "step": 2481 }, { "epoch": 0.26908065915004337, "grad_norm": 0.24504846334457397, "learning_rate": 9.13608008124184e-05, "loss": 0.0348, "step": 2482 }, { "epoch": 0.26918907198612313, "grad_norm": 0.1999277025461197, "learning_rate": 9.135717394458146e-05, "loss": 0.0316, "step": 2483 }, { "epoch": 0.26929748482220295, "grad_norm": 0.30711865425109863, "learning_rate": 9.135354707674453e-05, "loss": 0.0449, "step": 2484 }, { "epoch": 0.26940589765828277, "grad_norm": 0.5376157164573669, "learning_rate": 9.134992020890759e-05, "loss": 0.0444, "step": 2485 }, { "epoch": 0.26951431049436253, "grad_norm": 0.43979987502098083, "learning_rate": 9.134629334107065e-05, "loss": 0.0509, "step": 2486 }, { "epoch": 0.26962272333044235, "grad_norm": 0.32541173696517944, "learning_rate": 9.134266647323372e-05, "loss": 0.0624, "step": 2487 }, { "epoch": 0.2697311361665221, "grad_norm": 0.43163394927978516, "learning_rate": 9.133903960539678e-05, "loss": 0.0734, "step": 2488 }, { "epoch": 0.26983954900260193, "grad_norm": 0.22378727793693542, "learning_rate": 9.133541273755985e-05, "loss": 0.033, "step": 2489 }, { "epoch": 0.2699479618386817, "grad_norm": 0.3974356949329376, "learning_rate": 9.13317858697229e-05, "loss": 0.0309, "step": 2490 }, { "epoch": 0.2700563746747615, "grad_norm": 0.23825868964195251, "learning_rate": 9.132815900188598e-05, "loss": 0.0355, "step": 2491 }, { "epoch": 0.2701647875108413, "grad_norm": 0.2521940767765045, "learning_rate": 9.132453213404903e-05, "loss": 0.0525, "step": 2492 }, { "epoch": 0.2702732003469211, "grad_norm": 0.26804620027542114, "learning_rate": 9.13209052662121e-05, "loss": 0.0409, "step": 2493 }, { "epoch": 0.27038161318300086, "grad_norm": 0.3457334339618683, "learning_rate": 9.131727839837517e-05, "loss": 0.0564, "step": 2494 }, { "epoch": 0.2704900260190807, "grad_norm": 0.17983323335647583, "learning_rate": 9.131365153053823e-05, "loss": 0.0153, "step": 2495 }, { "epoch": 0.27059843885516044, "grad_norm": 0.49407097697257996, "learning_rate": 9.13100246627013e-05, "loss": 0.063, "step": 2496 }, { "epoch": 0.27070685169124026, "grad_norm": 0.3142303228378296, "learning_rate": 9.130639779486436e-05, "loss": 0.0508, "step": 2497 }, { "epoch": 0.27081526452732, "grad_norm": 0.4539966881275177, "learning_rate": 9.130277092702743e-05, "loss": 0.0433, "step": 2498 }, { "epoch": 0.27092367736339984, "grad_norm": 0.2775405943393707, "learning_rate": 9.129914405919049e-05, "loss": 0.0642, "step": 2499 }, { "epoch": 0.2710320901994796, "grad_norm": 0.27799299359321594, "learning_rate": 9.129551719135356e-05, "loss": 0.0539, "step": 2500 }, { "epoch": 0.2711405030355594, "grad_norm": 0.4122450351715088, "learning_rate": 9.129189032351662e-05, "loss": 0.0425, "step": 2501 }, { "epoch": 0.2712489158716392, "grad_norm": 0.39920416474342346, "learning_rate": 9.128826345567967e-05, "loss": 0.0673, "step": 2502 }, { "epoch": 0.271357328707719, "grad_norm": 0.37151122093200684, "learning_rate": 9.128463658784274e-05, "loss": 0.0526, "step": 2503 }, { "epoch": 0.27146574154379877, "grad_norm": 0.16547620296478271, "learning_rate": 9.12810097200058e-05, "loss": 0.0103, "step": 2504 }, { "epoch": 0.2715741543798786, "grad_norm": 0.3999404013156891, "learning_rate": 9.127738285216887e-05, "loss": 0.0447, "step": 2505 }, { "epoch": 0.27168256721595835, "grad_norm": 0.26081109046936035, "learning_rate": 9.127375598433193e-05, "loss": 0.0267, "step": 2506 }, { "epoch": 0.27179098005203817, "grad_norm": 0.3903443515300751, "learning_rate": 9.1270129116495e-05, "loss": 0.041, "step": 2507 }, { "epoch": 0.27189939288811793, "grad_norm": 0.5522079467773438, "learning_rate": 9.126650224865806e-05, "loss": 0.0532, "step": 2508 }, { "epoch": 0.27200780572419775, "grad_norm": 0.23253120481967926, "learning_rate": 9.126287538082113e-05, "loss": 0.0174, "step": 2509 }, { "epoch": 0.2721162185602775, "grad_norm": 0.49002882838249207, "learning_rate": 9.12592485129842e-05, "loss": 0.0283, "step": 2510 }, { "epoch": 0.27222463139635733, "grad_norm": 0.483772873878479, "learning_rate": 9.125562164514726e-05, "loss": 0.0526, "step": 2511 }, { "epoch": 0.2723330442324371, "grad_norm": 0.5514647364616394, "learning_rate": 9.125199477731033e-05, "loss": 0.0322, "step": 2512 }, { "epoch": 0.2724414570685169, "grad_norm": 0.4694764018058777, "learning_rate": 9.124836790947339e-05, "loss": 0.0527, "step": 2513 }, { "epoch": 0.27254986990459673, "grad_norm": 0.8708914518356323, "learning_rate": 9.124474104163646e-05, "loss": 0.1144, "step": 2514 }, { "epoch": 0.2726582827406765, "grad_norm": 0.38440558314323425, "learning_rate": 9.124111417379951e-05, "loss": 0.0654, "step": 2515 }, { "epoch": 0.2727666955767563, "grad_norm": 0.4454812705516815, "learning_rate": 9.123748730596257e-05, "loss": 0.0682, "step": 2516 }, { "epoch": 0.2728751084128361, "grad_norm": 0.5098534822463989, "learning_rate": 9.123386043812564e-05, "loss": 0.0245, "step": 2517 }, { "epoch": 0.2729835212489159, "grad_norm": 0.300409197807312, "learning_rate": 9.12302335702887e-05, "loss": 0.0385, "step": 2518 }, { "epoch": 0.27309193408499566, "grad_norm": 0.27132052183151245, "learning_rate": 9.122660670245177e-05, "loss": 0.0369, "step": 2519 }, { "epoch": 0.2732003469210755, "grad_norm": 0.28915441036224365, "learning_rate": 9.122297983461483e-05, "loss": 0.0248, "step": 2520 }, { "epoch": 0.27330875975715524, "grad_norm": 0.3323060870170593, "learning_rate": 9.12193529667779e-05, "loss": 0.0158, "step": 2521 }, { "epoch": 0.27341717259323506, "grad_norm": 0.49861496686935425, "learning_rate": 9.121572609894096e-05, "loss": 0.0851, "step": 2522 }, { "epoch": 0.2735255854293148, "grad_norm": 1.5338042974472046, "learning_rate": 9.121209923110403e-05, "loss": 0.0898, "step": 2523 }, { "epoch": 0.27363399826539464, "grad_norm": 0.14727887511253357, "learning_rate": 9.120847236326708e-05, "loss": 0.0115, "step": 2524 }, { "epoch": 0.2737424111014744, "grad_norm": 0.2936844527721405, "learning_rate": 9.120484549543014e-05, "loss": 0.0367, "step": 2525 }, { "epoch": 0.2738508239375542, "grad_norm": 0.28747642040252686, "learning_rate": 9.120121862759321e-05, "loss": 0.0479, "step": 2526 }, { "epoch": 0.273959236773634, "grad_norm": 0.270010769367218, "learning_rate": 9.119759175975627e-05, "loss": 0.0213, "step": 2527 }, { "epoch": 0.2740676496097138, "grad_norm": 0.32395845651626587, "learning_rate": 9.119396489191935e-05, "loss": 0.0546, "step": 2528 }, { "epoch": 0.27417606244579357, "grad_norm": 0.22359097003936768, "learning_rate": 9.119033802408241e-05, "loss": 0.0275, "step": 2529 }, { "epoch": 0.2742844752818734, "grad_norm": 0.5379827618598938, "learning_rate": 9.118671115624547e-05, "loss": 0.0501, "step": 2530 }, { "epoch": 0.27439288811795315, "grad_norm": 0.35231006145477295, "learning_rate": 9.118308428840854e-05, "loss": 0.041, "step": 2531 }, { "epoch": 0.27450130095403297, "grad_norm": 0.26983100175857544, "learning_rate": 9.11794574205716e-05, "loss": 0.0404, "step": 2532 }, { "epoch": 0.27460971379011273, "grad_norm": 0.45474210381507874, "learning_rate": 9.117583055273467e-05, "loss": 0.0645, "step": 2533 }, { "epoch": 0.27471812662619255, "grad_norm": 0.29373183846473694, "learning_rate": 9.117220368489772e-05, "loss": 0.0508, "step": 2534 }, { "epoch": 0.2748265394622723, "grad_norm": 0.31166180968284607, "learning_rate": 9.11685768170608e-05, "loss": 0.0602, "step": 2535 }, { "epoch": 0.27493495229835213, "grad_norm": 0.4650743901729584, "learning_rate": 9.116494994922385e-05, "loss": 0.0651, "step": 2536 }, { "epoch": 0.2750433651344319, "grad_norm": 0.3276517391204834, "learning_rate": 9.116132308138692e-05, "loss": 0.0391, "step": 2537 }, { "epoch": 0.2751517779705117, "grad_norm": 0.4233381450176239, "learning_rate": 9.115769621354998e-05, "loss": 0.063, "step": 2538 }, { "epoch": 0.2752601908065915, "grad_norm": 0.32253196835517883, "learning_rate": 9.115406934571304e-05, "loss": 0.0425, "step": 2539 }, { "epoch": 0.2753686036426713, "grad_norm": 0.6184103488922119, "learning_rate": 9.115044247787611e-05, "loss": 0.079, "step": 2540 }, { "epoch": 0.27547701647875106, "grad_norm": 0.2897958755493164, "learning_rate": 9.114681561003917e-05, "loss": 0.0424, "step": 2541 }, { "epoch": 0.2755854293148309, "grad_norm": 0.2662547826766968, "learning_rate": 9.114318874220224e-05, "loss": 0.0417, "step": 2542 }, { "epoch": 0.2756938421509107, "grad_norm": 0.3470768332481384, "learning_rate": 9.11395618743653e-05, "loss": 0.041, "step": 2543 }, { "epoch": 0.27580225498699046, "grad_norm": 0.4999983608722687, "learning_rate": 9.113593500652837e-05, "loss": 0.0932, "step": 2544 }, { "epoch": 0.2759106678230703, "grad_norm": 0.18268954753875732, "learning_rate": 9.113230813869144e-05, "loss": 0.0243, "step": 2545 }, { "epoch": 0.27601908065915004, "grad_norm": 0.25107112526893616, "learning_rate": 9.11286812708545e-05, "loss": 0.0343, "step": 2546 }, { "epoch": 0.27612749349522986, "grad_norm": 0.5664924383163452, "learning_rate": 9.112505440301756e-05, "loss": 0.0777, "step": 2547 }, { "epoch": 0.2762359063313096, "grad_norm": 0.3048820495605469, "learning_rate": 9.112142753518062e-05, "loss": 0.0435, "step": 2548 }, { "epoch": 0.27634431916738944, "grad_norm": 0.7642228007316589, "learning_rate": 9.111780066734369e-05, "loss": 0.0648, "step": 2549 }, { "epoch": 0.2764527320034692, "grad_norm": 0.2598980963230133, "learning_rate": 9.111417379950675e-05, "loss": 0.0234, "step": 2550 }, { "epoch": 0.276561144839549, "grad_norm": 0.3569332957267761, "learning_rate": 9.111054693166982e-05, "loss": 0.0408, "step": 2551 }, { "epoch": 0.2766695576756288, "grad_norm": 0.3753562867641449, "learning_rate": 9.110692006383288e-05, "loss": 0.0491, "step": 2552 }, { "epoch": 0.2767779705117086, "grad_norm": 0.17584702372550964, "learning_rate": 9.110329319599594e-05, "loss": 0.0373, "step": 2553 }, { "epoch": 0.27688638334778837, "grad_norm": 0.27503594756126404, "learning_rate": 9.109966632815901e-05, "loss": 0.0459, "step": 2554 }, { "epoch": 0.2769947961838682, "grad_norm": 0.28216737508773804, "learning_rate": 9.109603946032206e-05, "loss": 0.0315, "step": 2555 }, { "epoch": 0.27710320901994795, "grad_norm": 0.3350420594215393, "learning_rate": 9.109241259248514e-05, "loss": 0.0373, "step": 2556 }, { "epoch": 0.27721162185602777, "grad_norm": 0.36262497305870056, "learning_rate": 9.108878572464819e-05, "loss": 0.0379, "step": 2557 }, { "epoch": 0.27732003469210753, "grad_norm": 0.3573240041732788, "learning_rate": 9.108515885681126e-05, "loss": 0.0502, "step": 2558 }, { "epoch": 0.27742844752818735, "grad_norm": 0.5606093406677246, "learning_rate": 9.108153198897432e-05, "loss": 0.066, "step": 2559 }, { "epoch": 0.2775368603642671, "grad_norm": 0.2292628288269043, "learning_rate": 9.107790512113739e-05, "loss": 0.0245, "step": 2560 }, { "epoch": 0.27764527320034693, "grad_norm": 0.3990287482738495, "learning_rate": 9.107427825330045e-05, "loss": 0.0745, "step": 2561 }, { "epoch": 0.2777536860364267, "grad_norm": 0.35827407240867615, "learning_rate": 9.107065138546352e-05, "loss": 0.0285, "step": 2562 }, { "epoch": 0.2778620988725065, "grad_norm": 0.4562380909919739, "learning_rate": 9.106702451762659e-05, "loss": 0.0649, "step": 2563 }, { "epoch": 0.2779705117085863, "grad_norm": 0.43071335554122925, "learning_rate": 9.106339764978965e-05, "loss": 0.0448, "step": 2564 }, { "epoch": 0.2780789245446661, "grad_norm": 0.30776387453079224, "learning_rate": 9.105977078195272e-05, "loss": 0.0632, "step": 2565 }, { "epoch": 0.27818733738074586, "grad_norm": 0.4538935720920563, "learning_rate": 9.105614391411578e-05, "loss": 0.1131, "step": 2566 }, { "epoch": 0.2782957502168257, "grad_norm": 0.2092437595129013, "learning_rate": 9.105251704627883e-05, "loss": 0.0189, "step": 2567 }, { "epoch": 0.27840416305290544, "grad_norm": 0.3874198794364929, "learning_rate": 9.10488901784419e-05, "loss": 0.0306, "step": 2568 }, { "epoch": 0.27851257588898526, "grad_norm": 0.5830916166305542, "learning_rate": 9.104526331060496e-05, "loss": 0.0703, "step": 2569 }, { "epoch": 0.278620988725065, "grad_norm": 0.30284202098846436, "learning_rate": 9.104163644276803e-05, "loss": 0.0226, "step": 2570 }, { "epoch": 0.27872940156114484, "grad_norm": 0.5511496663093567, "learning_rate": 9.103800957493109e-05, "loss": 0.0698, "step": 2571 }, { "epoch": 0.2788378143972246, "grad_norm": 0.37731048464775085, "learning_rate": 9.103438270709416e-05, "loss": 0.0685, "step": 2572 }, { "epoch": 0.2789462272333044, "grad_norm": 0.654410719871521, "learning_rate": 9.103075583925722e-05, "loss": 0.0481, "step": 2573 }, { "epoch": 0.27905464006938424, "grad_norm": 0.36542561650276184, "learning_rate": 9.102712897142029e-05, "loss": 0.0475, "step": 2574 }, { "epoch": 0.279163052905464, "grad_norm": 0.5375925302505493, "learning_rate": 9.102350210358335e-05, "loss": 0.0634, "step": 2575 }, { "epoch": 0.2792714657415438, "grad_norm": 0.58146071434021, "learning_rate": 9.10198752357464e-05, "loss": 0.0629, "step": 2576 }, { "epoch": 0.2793798785776236, "grad_norm": 0.25490260124206543, "learning_rate": 9.101624836790947e-05, "loss": 0.0358, "step": 2577 }, { "epoch": 0.2794882914137034, "grad_norm": 0.286503404378891, "learning_rate": 9.101262150007253e-05, "loss": 0.0334, "step": 2578 }, { "epoch": 0.27959670424978317, "grad_norm": 0.2560926079750061, "learning_rate": 9.10089946322356e-05, "loss": 0.042, "step": 2579 }, { "epoch": 0.279705117085863, "grad_norm": 0.17831185460090637, "learning_rate": 9.100536776439867e-05, "loss": 0.0175, "step": 2580 }, { "epoch": 0.27981352992194275, "grad_norm": 0.28371837735176086, "learning_rate": 9.100174089656174e-05, "loss": 0.0439, "step": 2581 }, { "epoch": 0.27992194275802257, "grad_norm": 0.24835288524627686, "learning_rate": 9.09981140287248e-05, "loss": 0.0329, "step": 2582 }, { "epoch": 0.28003035559410233, "grad_norm": 0.37276995182037354, "learning_rate": 9.099448716088786e-05, "loss": 0.0477, "step": 2583 }, { "epoch": 0.28013876843018215, "grad_norm": 1.0158649682998657, "learning_rate": 9.099086029305093e-05, "loss": 0.0681, "step": 2584 }, { "epoch": 0.2802471812662619, "grad_norm": 0.6631553173065186, "learning_rate": 9.098723342521399e-05, "loss": 0.0503, "step": 2585 }, { "epoch": 0.28035559410234173, "grad_norm": 0.224182590842247, "learning_rate": 9.098360655737706e-05, "loss": 0.052, "step": 2586 }, { "epoch": 0.2804640069384215, "grad_norm": 0.26090964674949646, "learning_rate": 9.097997968954012e-05, "loss": 0.0268, "step": 2587 }, { "epoch": 0.2805724197745013, "grad_norm": 0.5998227596282959, "learning_rate": 9.097635282170319e-05, "loss": 0.0732, "step": 2588 }, { "epoch": 0.2806808326105811, "grad_norm": 0.5243447422981262, "learning_rate": 9.097272595386624e-05, "loss": 0.0485, "step": 2589 }, { "epoch": 0.2807892454466609, "grad_norm": 0.27681925892829895, "learning_rate": 9.096909908602931e-05, "loss": 0.0427, "step": 2590 }, { "epoch": 0.28089765828274066, "grad_norm": 0.366955429315567, "learning_rate": 9.096547221819237e-05, "loss": 0.0433, "step": 2591 }, { "epoch": 0.2810060711188205, "grad_norm": 0.4731999933719635, "learning_rate": 9.096184535035543e-05, "loss": 0.0648, "step": 2592 }, { "epoch": 0.28111448395490024, "grad_norm": 0.5858505964279175, "learning_rate": 9.09582184825185e-05, "loss": 0.0577, "step": 2593 }, { "epoch": 0.28122289679098006, "grad_norm": 0.6475706100463867, "learning_rate": 9.095459161468156e-05, "loss": 0.0753, "step": 2594 }, { "epoch": 0.2813313096270598, "grad_norm": 0.20170000195503235, "learning_rate": 9.095096474684463e-05, "loss": 0.0295, "step": 2595 }, { "epoch": 0.28143972246313964, "grad_norm": 0.3460483253002167, "learning_rate": 9.094733787900769e-05, "loss": 0.0504, "step": 2596 }, { "epoch": 0.2815481352992194, "grad_norm": 0.3697015345096588, "learning_rate": 9.094371101117076e-05, "loss": 0.0785, "step": 2597 }, { "epoch": 0.2816565481352992, "grad_norm": 0.6653920412063599, "learning_rate": 9.094008414333383e-05, "loss": 0.1024, "step": 2598 }, { "epoch": 0.281764960971379, "grad_norm": 0.5679699778556824, "learning_rate": 9.093645727549688e-05, "loss": 0.0474, "step": 2599 }, { "epoch": 0.2818733738074588, "grad_norm": 0.812318742275238, "learning_rate": 9.093283040765996e-05, "loss": 0.0579, "step": 2600 }, { "epoch": 0.28198178664353857, "grad_norm": 0.34248653054237366, "learning_rate": 9.092920353982301e-05, "loss": 0.0233, "step": 2601 }, { "epoch": 0.2820901994796184, "grad_norm": 0.5061943531036377, "learning_rate": 9.092557667198608e-05, "loss": 0.0237, "step": 2602 }, { "epoch": 0.2821986123156982, "grad_norm": 0.43182823061943054, "learning_rate": 9.092194980414914e-05, "loss": 0.0358, "step": 2603 }, { "epoch": 0.28230702515177797, "grad_norm": 0.4423953890800476, "learning_rate": 9.091832293631221e-05, "loss": 0.0376, "step": 2604 }, { "epoch": 0.2824154379878578, "grad_norm": 0.40004482865333557, "learning_rate": 9.091469606847527e-05, "loss": 0.0436, "step": 2605 }, { "epoch": 0.28252385082393755, "grad_norm": 0.43494853377342224, "learning_rate": 9.091106920063833e-05, "loss": 0.0568, "step": 2606 }, { "epoch": 0.28263226366001737, "grad_norm": 0.15698452293872833, "learning_rate": 9.09074423328014e-05, "loss": 0.0089, "step": 2607 }, { "epoch": 0.28274067649609713, "grad_norm": 0.2654987573623657, "learning_rate": 9.090381546496446e-05, "loss": 0.0277, "step": 2608 }, { "epoch": 0.28284908933217695, "grad_norm": 0.43953365087509155, "learning_rate": 9.090018859712753e-05, "loss": 0.049, "step": 2609 }, { "epoch": 0.2829575021682567, "grad_norm": 0.49121206998825073, "learning_rate": 9.089656172929058e-05, "loss": 0.0516, "step": 2610 }, { "epoch": 0.28306591500433653, "grad_norm": 0.38805803656578064, "learning_rate": 9.089293486145365e-05, "loss": 0.0384, "step": 2611 }, { "epoch": 0.2831743278404163, "grad_norm": 0.37442493438720703, "learning_rate": 9.088930799361671e-05, "loss": 0.0465, "step": 2612 }, { "epoch": 0.2832827406764961, "grad_norm": 0.476605623960495, "learning_rate": 9.088568112577978e-05, "loss": 0.0602, "step": 2613 }, { "epoch": 0.2833911535125759, "grad_norm": 0.48426705598831177, "learning_rate": 9.088205425794285e-05, "loss": 0.095, "step": 2614 }, { "epoch": 0.2834995663486557, "grad_norm": 0.19239012897014618, "learning_rate": 9.087842739010591e-05, "loss": 0.0109, "step": 2615 }, { "epoch": 0.28360797918473546, "grad_norm": 0.513535737991333, "learning_rate": 9.087480052226898e-05, "loss": 0.0437, "step": 2616 }, { "epoch": 0.2837163920208153, "grad_norm": 0.7335177063941956, "learning_rate": 9.087117365443204e-05, "loss": 0.0927, "step": 2617 }, { "epoch": 0.28382480485689504, "grad_norm": 0.7545831799507141, "learning_rate": 9.086754678659511e-05, "loss": 0.0617, "step": 2618 }, { "epoch": 0.28393321769297486, "grad_norm": 0.4059801995754242, "learning_rate": 9.086391991875817e-05, "loss": 0.0623, "step": 2619 }, { "epoch": 0.2840416305290546, "grad_norm": 0.5716904997825623, "learning_rate": 9.086029305092122e-05, "loss": 0.0536, "step": 2620 }, { "epoch": 0.28415004336513444, "grad_norm": 0.22646625339984894, "learning_rate": 9.08566661830843e-05, "loss": 0.0286, "step": 2621 }, { "epoch": 0.2842584562012142, "grad_norm": 0.9189560413360596, "learning_rate": 9.085303931524735e-05, "loss": 0.0375, "step": 2622 }, { "epoch": 0.284366869037294, "grad_norm": 0.3118526041507721, "learning_rate": 9.084941244741042e-05, "loss": 0.0283, "step": 2623 }, { "epoch": 0.2844752818733738, "grad_norm": 0.435563325881958, "learning_rate": 9.084578557957348e-05, "loss": 0.0359, "step": 2624 }, { "epoch": 0.2845836947094536, "grad_norm": 0.5185089111328125, "learning_rate": 9.084215871173655e-05, "loss": 0.0557, "step": 2625 }, { "epoch": 0.28469210754553337, "grad_norm": 0.47090378403663635, "learning_rate": 9.083853184389961e-05, "loss": 0.0414, "step": 2626 }, { "epoch": 0.2848005203816132, "grad_norm": 0.6218593716621399, "learning_rate": 9.083490497606268e-05, "loss": 0.0675, "step": 2627 }, { "epoch": 0.28490893321769295, "grad_norm": 0.9215378165245056, "learning_rate": 9.083127810822574e-05, "loss": 0.1032, "step": 2628 }, { "epoch": 0.28501734605377277, "grad_norm": 0.4273145794868469, "learning_rate": 9.08276512403888e-05, "loss": 0.0343, "step": 2629 }, { "epoch": 0.28512575888985253, "grad_norm": 0.10449683666229248, "learning_rate": 9.082402437255187e-05, "loss": 0.0276, "step": 2630 }, { "epoch": 0.28523417172593235, "grad_norm": 0.25756406784057617, "learning_rate": 9.082039750471494e-05, "loss": 0.058, "step": 2631 }, { "epoch": 0.28534258456201217, "grad_norm": 0.31825342774391174, "learning_rate": 9.081677063687801e-05, "loss": 0.0472, "step": 2632 }, { "epoch": 0.28545099739809193, "grad_norm": 0.42875316739082336, "learning_rate": 9.081314376904106e-05, "loss": 0.0308, "step": 2633 }, { "epoch": 0.28555941023417175, "grad_norm": 0.17754681408405304, "learning_rate": 9.080951690120412e-05, "loss": 0.0274, "step": 2634 }, { "epoch": 0.2856678230702515, "grad_norm": 0.9500636458396912, "learning_rate": 9.080589003336719e-05, "loss": 0.0419, "step": 2635 }, { "epoch": 0.28577623590633133, "grad_norm": 0.5935355424880981, "learning_rate": 9.080226316553025e-05, "loss": 0.0939, "step": 2636 }, { "epoch": 0.2858846487424111, "grad_norm": 0.5679813623428345, "learning_rate": 9.079863629769332e-05, "loss": 0.0513, "step": 2637 }, { "epoch": 0.2859930615784909, "grad_norm": 0.36366504430770874, "learning_rate": 9.079500942985638e-05, "loss": 0.0472, "step": 2638 }, { "epoch": 0.2861014744145707, "grad_norm": 0.2937197983264923, "learning_rate": 9.079138256201945e-05, "loss": 0.04, "step": 2639 }, { "epoch": 0.2862098872506505, "grad_norm": 0.2258549928665161, "learning_rate": 9.07877556941825e-05, "loss": 0.0483, "step": 2640 }, { "epoch": 0.28631830008673026, "grad_norm": 0.2574346363544464, "learning_rate": 9.078412882634558e-05, "loss": 0.0228, "step": 2641 }, { "epoch": 0.2864267129228101, "grad_norm": 0.2834804952144623, "learning_rate": 9.078050195850863e-05, "loss": 0.0759, "step": 2642 }, { "epoch": 0.28653512575888984, "grad_norm": 0.17979364097118378, "learning_rate": 9.077687509067169e-05, "loss": 0.0387, "step": 2643 }, { "epoch": 0.28664353859496966, "grad_norm": 0.27781710028648376, "learning_rate": 9.077324822283476e-05, "loss": 0.0375, "step": 2644 }, { "epoch": 0.2867519514310494, "grad_norm": 0.22579699754714966, "learning_rate": 9.076962135499782e-05, "loss": 0.0302, "step": 2645 }, { "epoch": 0.28686036426712924, "grad_norm": 0.2610301673412323, "learning_rate": 9.076599448716089e-05, "loss": 0.0325, "step": 2646 }, { "epoch": 0.286968777103209, "grad_norm": 0.252752423286438, "learning_rate": 9.076236761932395e-05, "loss": 0.0434, "step": 2647 }, { "epoch": 0.2870771899392888, "grad_norm": 0.5211769342422485, "learning_rate": 9.075874075148702e-05, "loss": 0.068, "step": 2648 }, { "epoch": 0.2871856027753686, "grad_norm": 0.39235764741897583, "learning_rate": 9.075511388365009e-05, "loss": 0.0192, "step": 2649 }, { "epoch": 0.2872940156114484, "grad_norm": 0.8088257908821106, "learning_rate": 9.075148701581315e-05, "loss": 0.0376, "step": 2650 }, { "epoch": 0.28740242844752817, "grad_norm": 0.43846091628074646, "learning_rate": 9.074786014797622e-05, "loss": 0.0879, "step": 2651 }, { "epoch": 0.287510841283608, "grad_norm": 0.2679232358932495, "learning_rate": 9.074423328013928e-05, "loss": 0.0391, "step": 2652 }, { "epoch": 0.28761925411968775, "grad_norm": 0.2751295566558838, "learning_rate": 9.074060641230235e-05, "loss": 0.0536, "step": 2653 }, { "epoch": 0.28772766695576757, "grad_norm": 0.5629369020462036, "learning_rate": 9.07369795444654e-05, "loss": 0.081, "step": 2654 }, { "epoch": 0.28783607979184733, "grad_norm": 0.5619133710861206, "learning_rate": 9.073335267662847e-05, "loss": 0.0981, "step": 2655 }, { "epoch": 0.28794449262792715, "grad_norm": 0.5324842929840088, "learning_rate": 9.072972580879153e-05, "loss": 0.0591, "step": 2656 }, { "epoch": 0.2880529054640069, "grad_norm": 0.1769053190946579, "learning_rate": 9.072609894095459e-05, "loss": 0.0327, "step": 2657 }, { "epoch": 0.28816131830008673, "grad_norm": 0.49760255217552185, "learning_rate": 9.072247207311766e-05, "loss": 0.0874, "step": 2658 }, { "epoch": 0.2882697311361665, "grad_norm": 0.3879537880420685, "learning_rate": 9.071884520528072e-05, "loss": 0.0417, "step": 2659 }, { "epoch": 0.2883781439722463, "grad_norm": 0.2792111039161682, "learning_rate": 9.071521833744379e-05, "loss": 0.0646, "step": 2660 }, { "epoch": 0.28848655680832613, "grad_norm": 0.12290972471237183, "learning_rate": 9.071159146960685e-05, "loss": 0.0315, "step": 2661 }, { "epoch": 0.2885949696444059, "grad_norm": 0.2958608567714691, "learning_rate": 9.070796460176992e-05, "loss": 0.0424, "step": 2662 }, { "epoch": 0.2887033824804857, "grad_norm": 0.5165923833847046, "learning_rate": 9.070433773393297e-05, "loss": 0.0842, "step": 2663 }, { "epoch": 0.2888117953165655, "grad_norm": 0.4894074499607086, "learning_rate": 9.070071086609604e-05, "loss": 0.0518, "step": 2664 }, { "epoch": 0.2889202081526453, "grad_norm": 0.3249136209487915, "learning_rate": 9.06970839982591e-05, "loss": 0.0376, "step": 2665 }, { "epoch": 0.28902862098872506, "grad_norm": 0.3854162395000458, "learning_rate": 9.069345713042217e-05, "loss": 0.0485, "step": 2666 }, { "epoch": 0.2891370338248049, "grad_norm": 0.22027896344661713, "learning_rate": 9.068983026258524e-05, "loss": 0.0209, "step": 2667 }, { "epoch": 0.28924544666088464, "grad_norm": 0.863584578037262, "learning_rate": 9.06862033947483e-05, "loss": 0.0881, "step": 2668 }, { "epoch": 0.28935385949696446, "grad_norm": 0.18116019666194916, "learning_rate": 9.068257652691137e-05, "loss": 0.0311, "step": 2669 }, { "epoch": 0.2894622723330442, "grad_norm": 0.35126781463623047, "learning_rate": 9.067894965907443e-05, "loss": 0.0626, "step": 2670 }, { "epoch": 0.28957068516912404, "grad_norm": 0.22768129408359528, "learning_rate": 9.067532279123749e-05, "loss": 0.0245, "step": 2671 }, { "epoch": 0.2896790980052038, "grad_norm": 0.25280770659446716, "learning_rate": 9.067169592340056e-05, "loss": 0.0292, "step": 2672 }, { "epoch": 0.2897875108412836, "grad_norm": 0.11219705641269684, "learning_rate": 9.066806905556362e-05, "loss": 0.0174, "step": 2673 }, { "epoch": 0.2898959236773634, "grad_norm": 0.3458138108253479, "learning_rate": 9.066444218772669e-05, "loss": 0.0829, "step": 2674 }, { "epoch": 0.2900043365134432, "grad_norm": 0.29190632700920105, "learning_rate": 9.066081531988974e-05, "loss": 0.0353, "step": 2675 }, { "epoch": 0.29011274934952297, "grad_norm": 0.33231106400489807, "learning_rate": 9.065718845205281e-05, "loss": 0.0523, "step": 2676 }, { "epoch": 0.2902211621856028, "grad_norm": 0.35112884640693665, "learning_rate": 9.065356158421587e-05, "loss": 0.0483, "step": 2677 }, { "epoch": 0.29032957502168255, "grad_norm": 0.3490806519985199, "learning_rate": 9.064993471637894e-05, "loss": 0.0593, "step": 2678 }, { "epoch": 0.29043798785776237, "grad_norm": 0.217361181974411, "learning_rate": 9.0646307848542e-05, "loss": 0.0327, "step": 2679 }, { "epoch": 0.29054640069384213, "grad_norm": 0.32634320855140686, "learning_rate": 9.064268098070506e-05, "loss": 0.0446, "step": 2680 }, { "epoch": 0.29065481352992195, "grad_norm": 0.41600751876831055, "learning_rate": 9.063905411286813e-05, "loss": 0.0294, "step": 2681 }, { "epoch": 0.2907632263660017, "grad_norm": 0.3839131295681, "learning_rate": 9.063542724503119e-05, "loss": 0.0325, "step": 2682 }, { "epoch": 0.29087163920208153, "grad_norm": 0.678829550743103, "learning_rate": 9.063180037719427e-05, "loss": 0.0746, "step": 2683 }, { "epoch": 0.2909800520381613, "grad_norm": 0.8251864910125732, "learning_rate": 9.062817350935733e-05, "loss": 0.0185, "step": 2684 }, { "epoch": 0.2910884648742411, "grad_norm": 0.1621316522359848, "learning_rate": 9.06245466415204e-05, "loss": 0.0178, "step": 2685 }, { "epoch": 0.2911968777103209, "grad_norm": 0.439689040184021, "learning_rate": 9.062091977368346e-05, "loss": 0.062, "step": 2686 }, { "epoch": 0.2913052905464007, "grad_norm": 0.1456812471151352, "learning_rate": 9.061729290584651e-05, "loss": 0.0232, "step": 2687 }, { "epoch": 0.29141370338248046, "grad_norm": 0.39846569299697876, "learning_rate": 9.061366603800958e-05, "loss": 0.0518, "step": 2688 }, { "epoch": 0.2915221162185603, "grad_norm": 0.5080788731575012, "learning_rate": 9.061003917017264e-05, "loss": 0.0787, "step": 2689 }, { "epoch": 0.29163052905464004, "grad_norm": 0.5120000243186951, "learning_rate": 9.060641230233571e-05, "loss": 0.0801, "step": 2690 }, { "epoch": 0.29173894189071986, "grad_norm": 0.30716922879219055, "learning_rate": 9.060278543449877e-05, "loss": 0.0294, "step": 2691 }, { "epoch": 0.2918473547267997, "grad_norm": 0.24011218547821045, "learning_rate": 9.059915856666184e-05, "loss": 0.0235, "step": 2692 }, { "epoch": 0.29195576756287944, "grad_norm": 0.7903215289115906, "learning_rate": 9.05955316988249e-05, "loss": 0.1047, "step": 2693 }, { "epoch": 0.29206418039895926, "grad_norm": 0.3887096643447876, "learning_rate": 9.059190483098797e-05, "loss": 0.0401, "step": 2694 }, { "epoch": 0.292172593235039, "grad_norm": 0.5612781047821045, "learning_rate": 9.058827796315103e-05, "loss": 0.0465, "step": 2695 }, { "epoch": 0.29228100607111884, "grad_norm": 0.20363616943359375, "learning_rate": 9.058465109531408e-05, "loss": 0.0332, "step": 2696 }, { "epoch": 0.2923894189071986, "grad_norm": 0.351898729801178, "learning_rate": 9.058102422747715e-05, "loss": 0.0339, "step": 2697 }, { "epoch": 0.2924978317432784, "grad_norm": 0.35007214546203613, "learning_rate": 9.057739735964021e-05, "loss": 0.0263, "step": 2698 }, { "epoch": 0.2926062445793582, "grad_norm": 0.3256678581237793, "learning_rate": 9.057377049180328e-05, "loss": 0.0434, "step": 2699 }, { "epoch": 0.292714657415438, "grad_norm": 0.510250985622406, "learning_rate": 9.057014362396635e-05, "loss": 0.0771, "step": 2700 }, { "epoch": 0.29282307025151777, "grad_norm": 0.6042733788490295, "learning_rate": 9.056651675612941e-05, "loss": 0.0559, "step": 2701 }, { "epoch": 0.2929314830875976, "grad_norm": 0.3256903886795044, "learning_rate": 9.056288988829248e-05, "loss": 0.0536, "step": 2702 }, { "epoch": 0.29303989592367735, "grad_norm": 0.2320714145898819, "learning_rate": 9.055926302045554e-05, "loss": 0.0302, "step": 2703 }, { "epoch": 0.29314830875975717, "grad_norm": 0.6266908049583435, "learning_rate": 9.055563615261861e-05, "loss": 0.0607, "step": 2704 }, { "epoch": 0.29325672159583693, "grad_norm": 0.46387144923210144, "learning_rate": 9.055200928478167e-05, "loss": 0.0473, "step": 2705 }, { "epoch": 0.29336513443191675, "grad_norm": 1.33230721950531, "learning_rate": 9.054838241694474e-05, "loss": 0.1159, "step": 2706 }, { "epoch": 0.2934735472679965, "grad_norm": 0.16393840312957764, "learning_rate": 9.05447555491078e-05, "loss": 0.0084, "step": 2707 }, { "epoch": 0.29358196010407633, "grad_norm": 0.5338529944419861, "learning_rate": 9.054112868127087e-05, "loss": 0.057, "step": 2708 }, { "epoch": 0.2936903729401561, "grad_norm": 0.5742285251617432, "learning_rate": 9.053750181343392e-05, "loss": 0.0934, "step": 2709 }, { "epoch": 0.2937987857762359, "grad_norm": 0.46576082706451416, "learning_rate": 9.053387494559698e-05, "loss": 0.072, "step": 2710 }, { "epoch": 0.2939071986123157, "grad_norm": 0.21570786833763123, "learning_rate": 9.053024807776005e-05, "loss": 0.0243, "step": 2711 }, { "epoch": 0.2940156114483955, "grad_norm": 0.18737202882766724, "learning_rate": 9.052662120992311e-05, "loss": 0.0261, "step": 2712 }, { "epoch": 0.29412402428447526, "grad_norm": 0.39504000544548035, "learning_rate": 9.052299434208618e-05, "loss": 0.0631, "step": 2713 }, { "epoch": 0.2942324371205551, "grad_norm": 0.25251251459121704, "learning_rate": 9.051936747424924e-05, "loss": 0.0302, "step": 2714 }, { "epoch": 0.29434084995663484, "grad_norm": 0.3620133697986603, "learning_rate": 9.051574060641231e-05, "loss": 0.06, "step": 2715 }, { "epoch": 0.29444926279271466, "grad_norm": 0.3094841241836548, "learning_rate": 9.051211373857536e-05, "loss": 0.0334, "step": 2716 }, { "epoch": 0.2945576756287944, "grad_norm": 0.29562702775001526, "learning_rate": 9.050848687073844e-05, "loss": 0.0458, "step": 2717 }, { "epoch": 0.29466608846487424, "grad_norm": 0.2524893581867218, "learning_rate": 9.05048600029015e-05, "loss": 0.0317, "step": 2718 }, { "epoch": 0.294774501300954, "grad_norm": 0.9074125289916992, "learning_rate": 9.050123313506456e-05, "loss": 0.0421, "step": 2719 }, { "epoch": 0.2948829141370338, "grad_norm": 0.325194388628006, "learning_rate": 9.049760626722763e-05, "loss": 0.0277, "step": 2720 }, { "epoch": 0.29499132697311364, "grad_norm": 0.45832279324531555, "learning_rate": 9.049397939939069e-05, "loss": 0.0481, "step": 2721 }, { "epoch": 0.2950997398091934, "grad_norm": 0.2501811981201172, "learning_rate": 9.049035253155376e-05, "loss": 0.054, "step": 2722 }, { "epoch": 0.2952081526452732, "grad_norm": 0.5063721537590027, "learning_rate": 9.048672566371682e-05, "loss": 0.0737, "step": 2723 }, { "epoch": 0.295316565481353, "grad_norm": 0.32085081934928894, "learning_rate": 9.048309879587988e-05, "loss": 0.0412, "step": 2724 }, { "epoch": 0.2954249783174328, "grad_norm": 0.2535497546195984, "learning_rate": 9.047947192804295e-05, "loss": 0.0302, "step": 2725 }, { "epoch": 0.29553339115351257, "grad_norm": 0.5275242328643799, "learning_rate": 9.0475845060206e-05, "loss": 0.0778, "step": 2726 }, { "epoch": 0.2956418039895924, "grad_norm": 0.16449186205863953, "learning_rate": 9.047221819236908e-05, "loss": 0.0235, "step": 2727 }, { "epoch": 0.29575021682567215, "grad_norm": 0.38211390376091003, "learning_rate": 9.046859132453213e-05, "loss": 0.0484, "step": 2728 }, { "epoch": 0.29585862966175197, "grad_norm": 0.15476198494434357, "learning_rate": 9.04649644566952e-05, "loss": 0.0195, "step": 2729 }, { "epoch": 0.29596704249783173, "grad_norm": 0.23658162355422974, "learning_rate": 9.046133758885826e-05, "loss": 0.0277, "step": 2730 }, { "epoch": 0.29607545533391155, "grad_norm": 0.43645355105400085, "learning_rate": 9.045771072102133e-05, "loss": 0.0426, "step": 2731 }, { "epoch": 0.2961838681699913, "grad_norm": 0.6751161813735962, "learning_rate": 9.045408385318439e-05, "loss": 0.069, "step": 2732 }, { "epoch": 0.29629228100607113, "grad_norm": 0.42222949862480164, "learning_rate": 9.045045698534745e-05, "loss": 0.0476, "step": 2733 }, { "epoch": 0.2964006938421509, "grad_norm": 0.5573660135269165, "learning_rate": 9.044683011751052e-05, "loss": 0.0544, "step": 2734 }, { "epoch": 0.2965091066782307, "grad_norm": 0.905345618724823, "learning_rate": 9.044320324967359e-05, "loss": 0.0982, "step": 2735 }, { "epoch": 0.2966175195143105, "grad_norm": 0.6030809879302979, "learning_rate": 9.043957638183666e-05, "loss": 0.0449, "step": 2736 }, { "epoch": 0.2967259323503903, "grad_norm": 0.9629055261611938, "learning_rate": 9.043594951399972e-05, "loss": 0.0497, "step": 2737 }, { "epoch": 0.29683434518647006, "grad_norm": 0.3934451639652252, "learning_rate": 9.043232264616278e-05, "loss": 0.0293, "step": 2738 }, { "epoch": 0.2969427580225499, "grad_norm": 0.9713699817657471, "learning_rate": 9.042869577832585e-05, "loss": 0.0657, "step": 2739 }, { "epoch": 0.29705117085862964, "grad_norm": 0.32237720489501953, "learning_rate": 9.04250689104889e-05, "loss": 0.0303, "step": 2740 }, { "epoch": 0.29715958369470946, "grad_norm": 0.5318268537521362, "learning_rate": 9.042144204265197e-05, "loss": 0.0889, "step": 2741 }, { "epoch": 0.2972679965307892, "grad_norm": 0.47936034202575684, "learning_rate": 9.041781517481503e-05, "loss": 0.0501, "step": 2742 }, { "epoch": 0.29737640936686904, "grad_norm": 0.25466153025627136, "learning_rate": 9.04141883069781e-05, "loss": 0.0333, "step": 2743 }, { "epoch": 0.2974848222029488, "grad_norm": 0.5633022785186768, "learning_rate": 9.041056143914116e-05, "loss": 0.0455, "step": 2744 }, { "epoch": 0.2975932350390286, "grad_norm": 0.49170732498168945, "learning_rate": 9.040693457130423e-05, "loss": 0.0849, "step": 2745 }, { "epoch": 0.2977016478751084, "grad_norm": 0.27422580122947693, "learning_rate": 9.040330770346729e-05, "loss": 0.0261, "step": 2746 }, { "epoch": 0.2978100607111882, "grad_norm": 0.6266526579856873, "learning_rate": 9.039968083563035e-05, "loss": 0.0369, "step": 2747 }, { "epoch": 0.29791847354726797, "grad_norm": 0.35758256912231445, "learning_rate": 9.039605396779342e-05, "loss": 0.0483, "step": 2748 }, { "epoch": 0.2980268863833478, "grad_norm": 0.5963480472564697, "learning_rate": 9.039242709995647e-05, "loss": 0.0542, "step": 2749 }, { "epoch": 0.2981352992194276, "grad_norm": 0.7166321873664856, "learning_rate": 9.038880023211954e-05, "loss": 0.0864, "step": 2750 }, { "epoch": 0.29824371205550737, "grad_norm": 0.1062794178724289, "learning_rate": 9.03851733642826e-05, "loss": 0.0178, "step": 2751 }, { "epoch": 0.2983521248915872, "grad_norm": 0.5207999348640442, "learning_rate": 9.038154649644567e-05, "loss": 0.0437, "step": 2752 }, { "epoch": 0.29846053772766695, "grad_norm": 0.17745015025138855, "learning_rate": 9.037791962860874e-05, "loss": 0.0379, "step": 2753 }, { "epoch": 0.29856895056374677, "grad_norm": 0.24936099350452423, "learning_rate": 9.03742927607718e-05, "loss": 0.0224, "step": 2754 }, { "epoch": 0.29867736339982653, "grad_norm": 0.43794333934783936, "learning_rate": 9.037066589293487e-05, "loss": 0.0773, "step": 2755 }, { "epoch": 0.29878577623590635, "grad_norm": 0.3571361303329468, "learning_rate": 9.036703902509793e-05, "loss": 0.0487, "step": 2756 }, { "epoch": 0.2988941890719861, "grad_norm": 0.5160982608795166, "learning_rate": 9.0363412157261e-05, "loss": 0.0957, "step": 2757 }, { "epoch": 0.29900260190806593, "grad_norm": 0.3628360331058502, "learning_rate": 9.035978528942406e-05, "loss": 0.0492, "step": 2758 }, { "epoch": 0.2991110147441457, "grad_norm": 0.6162056922912598, "learning_rate": 9.035615842158713e-05, "loss": 0.0634, "step": 2759 }, { "epoch": 0.2992194275802255, "grad_norm": 0.4134586453437805, "learning_rate": 9.035253155375019e-05, "loss": 0.0242, "step": 2760 }, { "epoch": 0.2993278404163053, "grad_norm": 0.20616382360458374, "learning_rate": 9.034890468591324e-05, "loss": 0.0369, "step": 2761 }, { "epoch": 0.2994362532523851, "grad_norm": 0.6618534922599792, "learning_rate": 9.034527781807631e-05, "loss": 0.0633, "step": 2762 }, { "epoch": 0.29954466608846486, "grad_norm": 0.23795263469219208, "learning_rate": 9.034165095023937e-05, "loss": 0.0455, "step": 2763 }, { "epoch": 0.2996530789245447, "grad_norm": 0.28956034779548645, "learning_rate": 9.033802408240244e-05, "loss": 0.0466, "step": 2764 }, { "epoch": 0.29976149176062444, "grad_norm": 0.40473616123199463, "learning_rate": 9.03343972145655e-05, "loss": 0.0677, "step": 2765 }, { "epoch": 0.29986990459670426, "grad_norm": 0.694398045539856, "learning_rate": 9.033077034672857e-05, "loss": 0.0808, "step": 2766 }, { "epoch": 0.299978317432784, "grad_norm": 0.2709556818008423, "learning_rate": 9.032714347889163e-05, "loss": 0.0286, "step": 2767 }, { "epoch": 0.30008673026886384, "grad_norm": 0.3309970796108246, "learning_rate": 9.03235166110547e-05, "loss": 0.0439, "step": 2768 }, { "epoch": 0.3001951431049436, "grad_norm": 0.3006606698036194, "learning_rate": 9.031988974321776e-05, "loss": 0.0396, "step": 2769 }, { "epoch": 0.3003035559410234, "grad_norm": 0.4811013340950012, "learning_rate": 9.031626287538083e-05, "loss": 0.0933, "step": 2770 }, { "epoch": 0.3004119687771032, "grad_norm": 0.4545406401157379, "learning_rate": 9.03126360075439e-05, "loss": 0.0401, "step": 2771 }, { "epoch": 0.300520381613183, "grad_norm": 0.40353789925575256, "learning_rate": 9.030900913970695e-05, "loss": 0.0696, "step": 2772 }, { "epoch": 0.30062879444926277, "grad_norm": 0.3430008888244629, "learning_rate": 9.030538227187003e-05, "loss": 0.0418, "step": 2773 }, { "epoch": 0.3007372072853426, "grad_norm": 0.4153067469596863, "learning_rate": 9.030175540403308e-05, "loss": 0.0646, "step": 2774 }, { "epoch": 0.30084562012142235, "grad_norm": 0.6443102955818176, "learning_rate": 9.029812853619615e-05, "loss": 0.0577, "step": 2775 }, { "epoch": 0.30095403295750217, "grad_norm": 0.6522417664527893, "learning_rate": 9.029450166835921e-05, "loss": 0.0368, "step": 2776 }, { "epoch": 0.30106244579358193, "grad_norm": 0.27165690064430237, "learning_rate": 9.029087480052227e-05, "loss": 0.0352, "step": 2777 }, { "epoch": 0.30117085862966175, "grad_norm": 0.3312934935092926, "learning_rate": 9.028724793268534e-05, "loss": 0.0443, "step": 2778 }, { "epoch": 0.3012792714657415, "grad_norm": 0.13680920004844666, "learning_rate": 9.02836210648484e-05, "loss": 0.0229, "step": 2779 }, { "epoch": 0.30138768430182133, "grad_norm": 0.3748157024383545, "learning_rate": 9.027999419701147e-05, "loss": 0.041, "step": 2780 }, { "epoch": 0.30149609713790115, "grad_norm": 0.3062927722930908, "learning_rate": 9.027636732917452e-05, "loss": 0.052, "step": 2781 }, { "epoch": 0.3016045099739809, "grad_norm": 0.3411240577697754, "learning_rate": 9.02727404613376e-05, "loss": 0.0327, "step": 2782 }, { "epoch": 0.30171292281006074, "grad_norm": 0.5990200042724609, "learning_rate": 9.026911359350065e-05, "loss": 0.0264, "step": 2783 }, { "epoch": 0.3018213356461405, "grad_norm": 0.4749259948730469, "learning_rate": 9.026548672566371e-05, "loss": 0.052, "step": 2784 }, { "epoch": 0.3019297484822203, "grad_norm": 0.2911134958267212, "learning_rate": 9.026185985782678e-05, "loss": 0.0439, "step": 2785 }, { "epoch": 0.3020381613183001, "grad_norm": 0.3850453197956085, "learning_rate": 9.025823298998984e-05, "loss": 0.0302, "step": 2786 }, { "epoch": 0.3021465741543799, "grad_norm": 0.454965740442276, "learning_rate": 9.025460612215292e-05, "loss": 0.0697, "step": 2787 }, { "epoch": 0.30225498699045966, "grad_norm": 0.5188993215560913, "learning_rate": 9.025097925431598e-05, "loss": 0.0543, "step": 2788 }, { "epoch": 0.3023633998265395, "grad_norm": 0.4772016406059265, "learning_rate": 9.024735238647905e-05, "loss": 0.0354, "step": 2789 }, { "epoch": 0.30247181266261924, "grad_norm": 0.1969473958015442, "learning_rate": 9.024372551864211e-05, "loss": 0.016, "step": 2790 }, { "epoch": 0.30258022549869906, "grad_norm": 0.2587556540966034, "learning_rate": 9.024009865080517e-05, "loss": 0.0337, "step": 2791 }, { "epoch": 0.3026886383347788, "grad_norm": 0.5463857650756836, "learning_rate": 9.023647178296824e-05, "loss": 0.0454, "step": 2792 }, { "epoch": 0.30279705117085864, "grad_norm": 0.5919232964515686, "learning_rate": 9.02328449151313e-05, "loss": 0.0673, "step": 2793 }, { "epoch": 0.3029054640069384, "grad_norm": 0.30096709728240967, "learning_rate": 9.022921804729437e-05, "loss": 0.023, "step": 2794 }, { "epoch": 0.3030138768430182, "grad_norm": 0.18787269294261932, "learning_rate": 9.022559117945742e-05, "loss": 0.0182, "step": 2795 }, { "epoch": 0.303122289679098, "grad_norm": 0.5893129706382751, "learning_rate": 9.02219643116205e-05, "loss": 0.0438, "step": 2796 }, { "epoch": 0.3032307025151778, "grad_norm": 0.5382670760154724, "learning_rate": 9.021833744378355e-05, "loss": 0.0438, "step": 2797 }, { "epoch": 0.30333911535125757, "grad_norm": 0.5657950639724731, "learning_rate": 9.021471057594662e-05, "loss": 0.0835, "step": 2798 }, { "epoch": 0.3034475281873374, "grad_norm": 0.2267863154411316, "learning_rate": 9.021108370810968e-05, "loss": 0.019, "step": 2799 }, { "epoch": 0.30355594102341715, "grad_norm": 0.41244107484817505, "learning_rate": 9.020745684027274e-05, "loss": 0.0551, "step": 2800 }, { "epoch": 0.30366435385949697, "grad_norm": 0.36998340487480164, "learning_rate": 9.020382997243581e-05, "loss": 0.032, "step": 2801 }, { "epoch": 0.30377276669557673, "grad_norm": 0.25497502088546753, "learning_rate": 9.020020310459886e-05, "loss": 0.0315, "step": 2802 }, { "epoch": 0.30388117953165655, "grad_norm": 0.46964341402053833, "learning_rate": 9.019657623676194e-05, "loss": 0.0443, "step": 2803 }, { "epoch": 0.3039895923677363, "grad_norm": 0.565481960773468, "learning_rate": 9.0192949368925e-05, "loss": 0.07, "step": 2804 }, { "epoch": 0.30409800520381614, "grad_norm": 0.4417359232902527, "learning_rate": 9.018932250108806e-05, "loss": 0.0349, "step": 2805 }, { "epoch": 0.3042064180398959, "grad_norm": 0.3168242573738098, "learning_rate": 9.018569563325113e-05, "loss": 0.0268, "step": 2806 }, { "epoch": 0.3043148308759757, "grad_norm": 0.6460436582565308, "learning_rate": 9.018206876541419e-05, "loss": 0.0691, "step": 2807 }, { "epoch": 0.3044232437120555, "grad_norm": 0.4253802001476288, "learning_rate": 9.017844189757726e-05, "loss": 0.0382, "step": 2808 }, { "epoch": 0.3045316565481353, "grad_norm": 0.6662473678588867, "learning_rate": 9.017481502974032e-05, "loss": 0.089, "step": 2809 }, { "epoch": 0.3046400693842151, "grad_norm": 0.19534733891487122, "learning_rate": 9.017118816190339e-05, "loss": 0.0184, "step": 2810 }, { "epoch": 0.3047484822202949, "grad_norm": 0.511174201965332, "learning_rate": 9.016756129406645e-05, "loss": 0.0509, "step": 2811 }, { "epoch": 0.3048568950563747, "grad_norm": 0.41722622513771057, "learning_rate": 9.016393442622952e-05, "loss": 0.0332, "step": 2812 }, { "epoch": 0.30496530789245446, "grad_norm": 0.5063440203666687, "learning_rate": 9.016030755839258e-05, "loss": 0.0442, "step": 2813 }, { "epoch": 0.3050737207285343, "grad_norm": 0.28340256214141846, "learning_rate": 9.015668069055563e-05, "loss": 0.0355, "step": 2814 }, { "epoch": 0.30518213356461404, "grad_norm": 0.2457297295331955, "learning_rate": 9.01530538227187e-05, "loss": 0.0212, "step": 2815 }, { "epoch": 0.30529054640069386, "grad_norm": 0.39005047082901, "learning_rate": 9.014942695488176e-05, "loss": 0.0433, "step": 2816 }, { "epoch": 0.3053989592367736, "grad_norm": 0.21222268044948578, "learning_rate": 9.014580008704483e-05, "loss": 0.0195, "step": 2817 }, { "epoch": 0.30550737207285344, "grad_norm": 0.2380683720111847, "learning_rate": 9.014217321920789e-05, "loss": 0.0308, "step": 2818 }, { "epoch": 0.3056157849089332, "grad_norm": 0.17062413692474365, "learning_rate": 9.013854635137096e-05, "loss": 0.0235, "step": 2819 }, { "epoch": 0.305724197745013, "grad_norm": 0.8285359740257263, "learning_rate": 9.013491948353402e-05, "loss": 0.0541, "step": 2820 }, { "epoch": 0.3058326105810928, "grad_norm": 0.26719534397125244, "learning_rate": 9.013129261569709e-05, "loss": 0.0175, "step": 2821 }, { "epoch": 0.3059410234171726, "grad_norm": 0.7203819751739502, "learning_rate": 9.012766574786016e-05, "loss": 0.0462, "step": 2822 }, { "epoch": 0.30604943625325237, "grad_norm": 0.556626558303833, "learning_rate": 9.012403888002322e-05, "loss": 0.0262, "step": 2823 }, { "epoch": 0.3061578490893322, "grad_norm": 0.8178575038909912, "learning_rate": 9.012041201218629e-05, "loss": 0.0411, "step": 2824 }, { "epoch": 0.30626626192541195, "grad_norm": 0.6142445206642151, "learning_rate": 9.011678514434935e-05, "loss": 0.0304, "step": 2825 }, { "epoch": 0.30637467476149177, "grad_norm": 0.3128412663936615, "learning_rate": 9.011315827651242e-05, "loss": 0.0187, "step": 2826 }, { "epoch": 0.30648308759757154, "grad_norm": 0.4747702181339264, "learning_rate": 9.010953140867547e-05, "loss": 0.0582, "step": 2827 }, { "epoch": 0.30659150043365135, "grad_norm": 0.39536046981811523, "learning_rate": 9.010590454083853e-05, "loss": 0.0382, "step": 2828 }, { "epoch": 0.3066999132697311, "grad_norm": 0.5521162152290344, "learning_rate": 9.01022776730016e-05, "loss": 0.0454, "step": 2829 }, { "epoch": 0.30680832610581094, "grad_norm": 0.21721479296684265, "learning_rate": 9.009865080516466e-05, "loss": 0.0219, "step": 2830 }, { "epoch": 0.3069167389418907, "grad_norm": 0.23925508558750153, "learning_rate": 9.009502393732773e-05, "loss": 0.0458, "step": 2831 }, { "epoch": 0.3070251517779705, "grad_norm": 0.40708696842193604, "learning_rate": 9.009139706949079e-05, "loss": 0.0585, "step": 2832 }, { "epoch": 0.3071335646140503, "grad_norm": 0.5553081035614014, "learning_rate": 9.008777020165386e-05, "loss": 0.0653, "step": 2833 }, { "epoch": 0.3072419774501301, "grad_norm": 0.3054531514644623, "learning_rate": 9.008414333381692e-05, "loss": 0.0435, "step": 2834 }, { "epoch": 0.30735039028620986, "grad_norm": 0.7275221943855286, "learning_rate": 9.008051646597999e-05, "loss": 0.052, "step": 2835 }, { "epoch": 0.3074588031222897, "grad_norm": 0.6649864912033081, "learning_rate": 9.007688959814304e-05, "loss": 0.0672, "step": 2836 }, { "epoch": 0.30756721595836944, "grad_norm": 0.18997447192668915, "learning_rate": 9.00732627303061e-05, "loss": 0.0096, "step": 2837 }, { "epoch": 0.30767562879444926, "grad_norm": 0.6225766539573669, "learning_rate": 9.006963586246917e-05, "loss": 0.0294, "step": 2838 }, { "epoch": 0.3077840416305291, "grad_norm": 0.48132002353668213, "learning_rate": 9.006600899463224e-05, "loss": 0.0535, "step": 2839 }, { "epoch": 0.30789245446660884, "grad_norm": 0.7115712761878967, "learning_rate": 9.006238212679531e-05, "loss": 0.0341, "step": 2840 }, { "epoch": 0.30800086730268866, "grad_norm": 0.3770350515842438, "learning_rate": 9.005875525895837e-05, "loss": 0.0377, "step": 2841 }, { "epoch": 0.3081092801387684, "grad_norm": 0.33109498023986816, "learning_rate": 9.005512839112143e-05, "loss": 0.0276, "step": 2842 }, { "epoch": 0.30821769297484825, "grad_norm": 0.6879532337188721, "learning_rate": 9.00515015232845e-05, "loss": 0.0498, "step": 2843 }, { "epoch": 0.308326105810928, "grad_norm": 0.679794430732727, "learning_rate": 9.004787465544756e-05, "loss": 0.0505, "step": 2844 }, { "epoch": 0.3084345186470078, "grad_norm": 0.9750697612762451, "learning_rate": 9.004424778761063e-05, "loss": 0.0815, "step": 2845 }, { "epoch": 0.3085429314830876, "grad_norm": 0.4320603609085083, "learning_rate": 9.004062091977368e-05, "loss": 0.0359, "step": 2846 }, { "epoch": 0.3086513443191674, "grad_norm": 0.5016499161720276, "learning_rate": 9.003699405193676e-05, "loss": 0.0347, "step": 2847 }, { "epoch": 0.30875975715524717, "grad_norm": 0.3151336908340454, "learning_rate": 9.003336718409981e-05, "loss": 0.0344, "step": 2848 }, { "epoch": 0.308868169991327, "grad_norm": 0.5558921694755554, "learning_rate": 9.002974031626288e-05, "loss": 0.0479, "step": 2849 }, { "epoch": 0.30897658282740675, "grad_norm": 0.3683154881000519, "learning_rate": 9.002611344842594e-05, "loss": 0.0278, "step": 2850 }, { "epoch": 0.3090849956634866, "grad_norm": 0.4904957413673401, "learning_rate": 9.0022486580589e-05, "loss": 0.042, "step": 2851 }, { "epoch": 0.30919340849956634, "grad_norm": 0.5734422206878662, "learning_rate": 9.001885971275207e-05, "loss": 0.0419, "step": 2852 }, { "epoch": 0.30930182133564615, "grad_norm": 0.3556376099586487, "learning_rate": 9.001523284491513e-05, "loss": 0.026, "step": 2853 }, { "epoch": 0.3094102341717259, "grad_norm": 0.2624008059501648, "learning_rate": 9.00116059770782e-05, "loss": 0.03, "step": 2854 }, { "epoch": 0.30951864700780574, "grad_norm": 0.5283095836639404, "learning_rate": 9.000797910924126e-05, "loss": 0.0979, "step": 2855 }, { "epoch": 0.3096270598438855, "grad_norm": 0.5141448378562927, "learning_rate": 9.000435224140433e-05, "loss": 0.0321, "step": 2856 }, { "epoch": 0.3097354726799653, "grad_norm": 0.8051297664642334, "learning_rate": 9.00007253735674e-05, "loss": 0.0553, "step": 2857 }, { "epoch": 0.3098438855160451, "grad_norm": 0.5311105847358704, "learning_rate": 8.999709850573045e-05, "loss": 0.0734, "step": 2858 }, { "epoch": 0.3099522983521249, "grad_norm": 0.583462655544281, "learning_rate": 8.999347163789353e-05, "loss": 0.0482, "step": 2859 }, { "epoch": 0.31006071118820466, "grad_norm": 0.12402807176113129, "learning_rate": 8.998984477005658e-05, "loss": 0.0166, "step": 2860 }, { "epoch": 0.3101691240242845, "grad_norm": 0.5662097334861755, "learning_rate": 8.998621790221965e-05, "loss": 0.0243, "step": 2861 }, { "epoch": 0.31027753686036424, "grad_norm": 0.21853652596473694, "learning_rate": 8.998259103438271e-05, "loss": 0.0318, "step": 2862 }, { "epoch": 0.31038594969644406, "grad_norm": 0.38048067688941956, "learning_rate": 8.997896416654578e-05, "loss": 0.0305, "step": 2863 }, { "epoch": 0.3104943625325238, "grad_norm": 0.7682182192802429, "learning_rate": 8.997533729870884e-05, "loss": 0.081, "step": 2864 }, { "epoch": 0.31060277536860365, "grad_norm": 0.4293581545352936, "learning_rate": 8.99717104308719e-05, "loss": 0.1062, "step": 2865 }, { "epoch": 0.3107111882046834, "grad_norm": 0.41559404134750366, "learning_rate": 8.996808356303497e-05, "loss": 0.0343, "step": 2866 }, { "epoch": 0.3108196010407632, "grad_norm": 0.11909820884466171, "learning_rate": 8.996445669519802e-05, "loss": 0.0091, "step": 2867 }, { "epoch": 0.31092801387684305, "grad_norm": 0.42183926701545715, "learning_rate": 8.99608298273611e-05, "loss": 0.0549, "step": 2868 }, { "epoch": 0.3110364267129228, "grad_norm": 0.1488541215658188, "learning_rate": 8.995720295952415e-05, "loss": 0.0149, "step": 2869 }, { "epoch": 0.3111448395490026, "grad_norm": 0.5102177262306213, "learning_rate": 8.995357609168722e-05, "loss": 0.0433, "step": 2870 }, { "epoch": 0.3112532523850824, "grad_norm": 0.23871159553527832, "learning_rate": 8.994994922385028e-05, "loss": 0.0358, "step": 2871 }, { "epoch": 0.3113616652211622, "grad_norm": 0.5875980854034424, "learning_rate": 8.994632235601335e-05, "loss": 0.0827, "step": 2872 }, { "epoch": 0.311470078057242, "grad_norm": 0.38368484377861023, "learning_rate": 8.994269548817642e-05, "loss": 0.0609, "step": 2873 }, { "epoch": 0.3115784908933218, "grad_norm": 0.5245661735534668, "learning_rate": 8.993906862033948e-05, "loss": 0.0705, "step": 2874 }, { "epoch": 0.31168690372940155, "grad_norm": 0.35067838430404663, "learning_rate": 8.993544175250255e-05, "loss": 0.0221, "step": 2875 }, { "epoch": 0.3117953165654814, "grad_norm": 0.2684086859226227, "learning_rate": 8.993181488466561e-05, "loss": 0.0382, "step": 2876 }, { "epoch": 0.31190372940156114, "grad_norm": 0.7030792832374573, "learning_rate": 8.992818801682868e-05, "loss": 0.0554, "step": 2877 }, { "epoch": 0.31201214223764095, "grad_norm": 0.6084462404251099, "learning_rate": 8.992456114899174e-05, "loss": 0.0322, "step": 2878 }, { "epoch": 0.3121205550737207, "grad_norm": 0.3318326473236084, "learning_rate": 8.992093428115481e-05, "loss": 0.0674, "step": 2879 }, { "epoch": 0.31222896790980054, "grad_norm": 0.49857181310653687, "learning_rate": 8.991730741331786e-05, "loss": 0.0545, "step": 2880 }, { "epoch": 0.3123373807458803, "grad_norm": 0.4010125398635864, "learning_rate": 8.991368054548092e-05, "loss": 0.0485, "step": 2881 }, { "epoch": 0.3124457935819601, "grad_norm": 0.254434198141098, "learning_rate": 8.991005367764399e-05, "loss": 0.0469, "step": 2882 }, { "epoch": 0.3125542064180399, "grad_norm": 0.30691054463386536, "learning_rate": 8.990642680980705e-05, "loss": 0.0389, "step": 2883 }, { "epoch": 0.3126626192541197, "grad_norm": 0.2501903176307678, "learning_rate": 8.990279994197012e-05, "loss": 0.04, "step": 2884 }, { "epoch": 0.31277103209019946, "grad_norm": 0.3916719853878021, "learning_rate": 8.989917307413318e-05, "loss": 0.0527, "step": 2885 }, { "epoch": 0.3128794449262793, "grad_norm": 0.2833918631076813, "learning_rate": 8.989554620629625e-05, "loss": 0.0469, "step": 2886 }, { "epoch": 0.31298785776235905, "grad_norm": 0.3070192337036133, "learning_rate": 8.98919193384593e-05, "loss": 0.047, "step": 2887 }, { "epoch": 0.31309627059843886, "grad_norm": 0.47077375650405884, "learning_rate": 8.988829247062236e-05, "loss": 0.0651, "step": 2888 }, { "epoch": 0.3132046834345186, "grad_norm": 0.3071298897266388, "learning_rate": 8.988466560278543e-05, "loss": 0.0402, "step": 2889 }, { "epoch": 0.31331309627059845, "grad_norm": 0.28754645586013794, "learning_rate": 8.988103873494849e-05, "loss": 0.0536, "step": 2890 }, { "epoch": 0.3134215091066782, "grad_norm": 0.775582492351532, "learning_rate": 8.987741186711158e-05, "loss": 0.0636, "step": 2891 }, { "epoch": 0.313529921942758, "grad_norm": 0.3034876883029938, "learning_rate": 8.987378499927463e-05, "loss": 0.0458, "step": 2892 }, { "epoch": 0.3136383347788378, "grad_norm": 0.3657785952091217, "learning_rate": 8.98701581314377e-05, "loss": 0.0263, "step": 2893 }, { "epoch": 0.3137467476149176, "grad_norm": 0.5100468993186951, "learning_rate": 8.986653126360076e-05, "loss": 0.0844, "step": 2894 }, { "epoch": 0.3138551604509974, "grad_norm": 0.3983466625213623, "learning_rate": 8.986290439576382e-05, "loss": 0.0256, "step": 2895 }, { "epoch": 0.3139635732870772, "grad_norm": 0.35169243812561035, "learning_rate": 8.985927752792689e-05, "loss": 0.0464, "step": 2896 }, { "epoch": 0.31407198612315695, "grad_norm": 0.4490346610546112, "learning_rate": 8.985565066008995e-05, "loss": 0.0561, "step": 2897 }, { "epoch": 0.3141803989592368, "grad_norm": 0.3176024258136749, "learning_rate": 8.985202379225302e-05, "loss": 0.034, "step": 2898 }, { "epoch": 0.3142888117953166, "grad_norm": 0.9590942859649658, "learning_rate": 8.984839692441608e-05, "loss": 0.0574, "step": 2899 }, { "epoch": 0.31439722463139635, "grad_norm": 0.3606038987636566, "learning_rate": 8.984477005657915e-05, "loss": 0.0742, "step": 2900 }, { "epoch": 0.3145056374674762, "grad_norm": 0.15189406275749207, "learning_rate": 8.98411431887422e-05, "loss": 0.0079, "step": 2901 }, { "epoch": 0.31461405030355594, "grad_norm": 0.32830506563186646, "learning_rate": 8.983751632090527e-05, "loss": 0.0164, "step": 2902 }, { "epoch": 0.31472246313963576, "grad_norm": 0.3095628023147583, "learning_rate": 8.983388945306833e-05, "loss": 0.0698, "step": 2903 }, { "epoch": 0.3148308759757155, "grad_norm": 0.4004982113838196, "learning_rate": 8.983026258523139e-05, "loss": 0.0642, "step": 2904 }, { "epoch": 0.31493928881179534, "grad_norm": 0.48695799708366394, "learning_rate": 8.982663571739446e-05, "loss": 0.1067, "step": 2905 }, { "epoch": 0.3150477016478751, "grad_norm": 0.14284774661064148, "learning_rate": 8.982300884955752e-05, "loss": 0.0212, "step": 2906 }, { "epoch": 0.3151561144839549, "grad_norm": 0.5934236645698547, "learning_rate": 8.981938198172059e-05, "loss": 0.0364, "step": 2907 }, { "epoch": 0.3152645273200347, "grad_norm": 0.3991644084453583, "learning_rate": 8.981575511388366e-05, "loss": 0.0778, "step": 2908 }, { "epoch": 0.3153729401561145, "grad_norm": 0.27238184213638306, "learning_rate": 8.981212824604672e-05, "loss": 0.0389, "step": 2909 }, { "epoch": 0.31548135299219426, "grad_norm": 0.30091431736946106, "learning_rate": 8.980850137820979e-05, "loss": 0.0219, "step": 2910 }, { "epoch": 0.3155897658282741, "grad_norm": 0.5259302854537964, "learning_rate": 8.980487451037285e-05, "loss": 0.0334, "step": 2911 }, { "epoch": 0.31569817866435385, "grad_norm": 0.3181382715702057, "learning_rate": 8.980124764253592e-05, "loss": 0.0418, "step": 2912 }, { "epoch": 0.31580659150043366, "grad_norm": 0.4088122248649597, "learning_rate": 8.979762077469897e-05, "loss": 0.0556, "step": 2913 }, { "epoch": 0.3159150043365134, "grad_norm": 0.13894310593605042, "learning_rate": 8.979399390686204e-05, "loss": 0.0119, "step": 2914 }, { "epoch": 0.31602341717259325, "grad_norm": 0.3658078908920288, "learning_rate": 8.97903670390251e-05, "loss": 0.061, "step": 2915 }, { "epoch": 0.316131830008673, "grad_norm": 0.4917811155319214, "learning_rate": 8.978674017118817e-05, "loss": 0.0596, "step": 2916 }, { "epoch": 0.31624024284475283, "grad_norm": 0.28571000695228577, "learning_rate": 8.978311330335123e-05, "loss": 0.0344, "step": 2917 }, { "epoch": 0.3163486556808326, "grad_norm": 0.278075635433197, "learning_rate": 8.977948643551429e-05, "loss": 0.0347, "step": 2918 }, { "epoch": 0.3164570685169124, "grad_norm": 0.2553616464138031, "learning_rate": 8.977585956767736e-05, "loss": 0.0349, "step": 2919 }, { "epoch": 0.3165654813529922, "grad_norm": 0.41422003507614136, "learning_rate": 8.977223269984042e-05, "loss": 0.0554, "step": 2920 }, { "epoch": 0.316673894189072, "grad_norm": 0.2623620629310608, "learning_rate": 8.976860583200349e-05, "loss": 0.0461, "step": 2921 }, { "epoch": 0.31678230702515175, "grad_norm": 0.4296587109565735, "learning_rate": 8.976497896416654e-05, "loss": 0.041, "step": 2922 }, { "epoch": 0.3168907198612316, "grad_norm": 0.41091835498809814, "learning_rate": 8.976135209632961e-05, "loss": 0.0463, "step": 2923 }, { "epoch": 0.31699913269731134, "grad_norm": 0.23736774921417236, "learning_rate": 8.975772522849267e-05, "loss": 0.0325, "step": 2924 }, { "epoch": 0.31710754553339116, "grad_norm": 0.5405881404876709, "learning_rate": 8.975409836065574e-05, "loss": 0.0217, "step": 2925 }, { "epoch": 0.3172159583694709, "grad_norm": 0.3299824297428131, "learning_rate": 8.975047149281881e-05, "loss": 0.0327, "step": 2926 }, { "epoch": 0.31732437120555074, "grad_norm": 0.23568841814994812, "learning_rate": 8.974684462498187e-05, "loss": 0.0168, "step": 2927 }, { "epoch": 0.31743278404163056, "grad_norm": 0.5387671589851379, "learning_rate": 8.974321775714494e-05, "loss": 0.0473, "step": 2928 }, { "epoch": 0.3175411968777103, "grad_norm": 0.6442615389823914, "learning_rate": 8.9739590889308e-05, "loss": 0.0538, "step": 2929 }, { "epoch": 0.31764960971379014, "grad_norm": 0.308485209941864, "learning_rate": 8.973596402147107e-05, "loss": 0.0532, "step": 2930 }, { "epoch": 0.3177580225498699, "grad_norm": 0.2068413943052292, "learning_rate": 8.973233715363413e-05, "loss": 0.0314, "step": 2931 }, { "epoch": 0.3178664353859497, "grad_norm": 0.2759053707122803, "learning_rate": 8.972871028579718e-05, "loss": 0.0277, "step": 2932 }, { "epoch": 0.3179748482220295, "grad_norm": 0.16160446405410767, "learning_rate": 8.972508341796026e-05, "loss": 0.0165, "step": 2933 }, { "epoch": 0.3180832610581093, "grad_norm": 0.7891277074813843, "learning_rate": 8.972145655012331e-05, "loss": 0.073, "step": 2934 }, { "epoch": 0.31819167389418906, "grad_norm": 0.3176671862602234, "learning_rate": 8.971782968228638e-05, "loss": 0.0208, "step": 2935 }, { "epoch": 0.3183000867302689, "grad_norm": 0.44093289971351624, "learning_rate": 8.971420281444944e-05, "loss": 0.0396, "step": 2936 }, { "epoch": 0.31840849956634865, "grad_norm": 0.5483407378196716, "learning_rate": 8.971057594661251e-05, "loss": 0.039, "step": 2937 }, { "epoch": 0.31851691240242846, "grad_norm": 0.5356181859970093, "learning_rate": 8.970694907877557e-05, "loss": 0.0505, "step": 2938 }, { "epoch": 0.31862532523850823, "grad_norm": 0.38092848658561707, "learning_rate": 8.970332221093864e-05, "loss": 0.0308, "step": 2939 }, { "epoch": 0.31873373807458805, "grad_norm": 0.36340856552124023, "learning_rate": 8.96996953431017e-05, "loss": 0.0511, "step": 2940 }, { "epoch": 0.3188421509106678, "grad_norm": 0.3364153206348419, "learning_rate": 8.969606847526475e-05, "loss": 0.0151, "step": 2941 }, { "epoch": 0.31895056374674763, "grad_norm": 0.7485520243644714, "learning_rate": 8.969244160742784e-05, "loss": 0.0391, "step": 2942 }, { "epoch": 0.3190589765828274, "grad_norm": 0.6064440608024597, "learning_rate": 8.96888147395909e-05, "loss": 0.0889, "step": 2943 }, { "epoch": 0.3191673894189072, "grad_norm": 0.33907291293144226, "learning_rate": 8.968518787175397e-05, "loss": 0.0525, "step": 2944 }, { "epoch": 0.319275802254987, "grad_norm": 0.6138657331466675, "learning_rate": 8.968156100391702e-05, "loss": 0.057, "step": 2945 }, { "epoch": 0.3193842150910668, "grad_norm": 0.4299224615097046, "learning_rate": 8.967793413608008e-05, "loss": 0.0587, "step": 2946 }, { "epoch": 0.31949262792714656, "grad_norm": 0.37792646884918213, "learning_rate": 8.967430726824315e-05, "loss": 0.0439, "step": 2947 }, { "epoch": 0.3196010407632264, "grad_norm": 0.3486732840538025, "learning_rate": 8.967068040040621e-05, "loss": 0.0433, "step": 2948 }, { "epoch": 0.31970945359930614, "grad_norm": 0.41059157252311707, "learning_rate": 8.966705353256928e-05, "loss": 0.0433, "step": 2949 }, { "epoch": 0.31981786643538596, "grad_norm": 0.347067266702652, "learning_rate": 8.966342666473234e-05, "loss": 0.0287, "step": 2950 }, { "epoch": 0.3199262792714657, "grad_norm": 0.4453713595867157, "learning_rate": 8.965979979689541e-05, "loss": 0.0568, "step": 2951 }, { "epoch": 0.32003469210754554, "grad_norm": 0.46576985716819763, "learning_rate": 8.965617292905847e-05, "loss": 0.0505, "step": 2952 }, { "epoch": 0.3201431049436253, "grad_norm": 0.6303920149803162, "learning_rate": 8.965254606122154e-05, "loss": 0.0391, "step": 2953 }, { "epoch": 0.3202515177797051, "grad_norm": 0.4383980631828308, "learning_rate": 8.96489191933846e-05, "loss": 0.0782, "step": 2954 }, { "epoch": 0.3203599306157849, "grad_norm": 0.14089111983776093, "learning_rate": 8.964529232554765e-05, "loss": 0.0137, "step": 2955 }, { "epoch": 0.3204683434518647, "grad_norm": 0.3333061635494232, "learning_rate": 8.964166545771072e-05, "loss": 0.035, "step": 2956 }, { "epoch": 0.3205767562879445, "grad_norm": 0.5541122555732727, "learning_rate": 8.963803858987378e-05, "loss": 0.0428, "step": 2957 }, { "epoch": 0.3206851691240243, "grad_norm": 0.4018808603286743, "learning_rate": 8.963441172203685e-05, "loss": 0.0379, "step": 2958 }, { "epoch": 0.3207935819601041, "grad_norm": 0.1807861477136612, "learning_rate": 8.963078485419991e-05, "loss": 0.0187, "step": 2959 }, { "epoch": 0.32090199479618386, "grad_norm": 0.3829610049724579, "learning_rate": 8.962715798636299e-05, "loss": 0.0476, "step": 2960 }, { "epoch": 0.3210104076322637, "grad_norm": 0.17488910257816315, "learning_rate": 8.962353111852605e-05, "loss": 0.0122, "step": 2961 }, { "epoch": 0.32111882046834345, "grad_norm": 0.19718392193317413, "learning_rate": 8.961990425068911e-05, "loss": 0.0233, "step": 2962 }, { "epoch": 0.32122723330442327, "grad_norm": 0.6571956872940063, "learning_rate": 8.961627738285218e-05, "loss": 0.0686, "step": 2963 }, { "epoch": 0.32133564614050303, "grad_norm": 0.38455191254615784, "learning_rate": 8.961265051501524e-05, "loss": 0.06, "step": 2964 }, { "epoch": 0.32144405897658285, "grad_norm": 0.45833539962768555, "learning_rate": 8.96090236471783e-05, "loss": 0.0415, "step": 2965 }, { "epoch": 0.3215524718126626, "grad_norm": 0.4982450306415558, "learning_rate": 8.960539677934136e-05, "loss": 0.0769, "step": 2966 }, { "epoch": 0.32166088464874243, "grad_norm": 0.7750599980354309, "learning_rate": 8.960176991150443e-05, "loss": 0.0844, "step": 2967 }, { "epoch": 0.3217692974848222, "grad_norm": 1.4284579753875732, "learning_rate": 8.959814304366749e-05, "loss": 0.0478, "step": 2968 }, { "epoch": 0.321877710320902, "grad_norm": 0.2254296988248825, "learning_rate": 8.959451617583055e-05, "loss": 0.0289, "step": 2969 }, { "epoch": 0.3219861231569818, "grad_norm": 0.19533836841583252, "learning_rate": 8.959088930799362e-05, "loss": 0.0204, "step": 2970 }, { "epoch": 0.3220945359930616, "grad_norm": 0.26827117800712585, "learning_rate": 8.958726244015668e-05, "loss": 0.0261, "step": 2971 }, { "epoch": 0.32220294882914136, "grad_norm": 0.686718225479126, "learning_rate": 8.958363557231975e-05, "loss": 0.0875, "step": 2972 }, { "epoch": 0.3223113616652212, "grad_norm": 0.3618023097515106, "learning_rate": 8.95800087044828e-05, "loss": 0.038, "step": 2973 }, { "epoch": 0.32241977450130094, "grad_norm": 0.3115321099758148, "learning_rate": 8.957638183664588e-05, "loss": 0.031, "step": 2974 }, { "epoch": 0.32252818733738076, "grad_norm": 0.33198821544647217, "learning_rate": 8.957275496880893e-05, "loss": 0.0481, "step": 2975 }, { "epoch": 0.3226366001734605, "grad_norm": 0.31797873973846436, "learning_rate": 8.9569128100972e-05, "loss": 0.026, "step": 2976 }, { "epoch": 0.32274501300954034, "grad_norm": 0.4703308641910553, "learning_rate": 8.956550123313508e-05, "loss": 0.0683, "step": 2977 }, { "epoch": 0.3228534258456201, "grad_norm": 0.1361035704612732, "learning_rate": 8.956187436529813e-05, "loss": 0.0326, "step": 2978 }, { "epoch": 0.3229618386816999, "grad_norm": 0.35255518555641174, "learning_rate": 8.95582474974612e-05, "loss": 0.041, "step": 2979 }, { "epoch": 0.3230702515177797, "grad_norm": 0.43825772404670715, "learning_rate": 8.955462062962426e-05, "loss": 0.0769, "step": 2980 }, { "epoch": 0.3231786643538595, "grad_norm": 0.5823899507522583, "learning_rate": 8.955099376178733e-05, "loss": 0.0244, "step": 2981 }, { "epoch": 0.32328707718993926, "grad_norm": 0.1754390448331833, "learning_rate": 8.954736689395039e-05, "loss": 0.0323, "step": 2982 }, { "epoch": 0.3233954900260191, "grad_norm": 0.4930426776409149, "learning_rate": 8.954374002611346e-05, "loss": 0.0816, "step": 2983 }, { "epoch": 0.32350390286209885, "grad_norm": 0.264548122882843, "learning_rate": 8.954011315827652e-05, "loss": 0.0187, "step": 2984 }, { "epoch": 0.32361231569817867, "grad_norm": 0.33921054005622864, "learning_rate": 8.953648629043958e-05, "loss": 0.0535, "step": 2985 }, { "epoch": 0.3237207285342585, "grad_norm": 0.4212264120578766, "learning_rate": 8.953285942260265e-05, "loss": 0.0878, "step": 2986 }, { "epoch": 0.32382914137033825, "grad_norm": 0.3361729383468628, "learning_rate": 8.95292325547657e-05, "loss": 0.0543, "step": 2987 }, { "epoch": 0.32393755420641807, "grad_norm": 0.33295056223869324, "learning_rate": 8.952560568692877e-05, "loss": 0.0361, "step": 2988 }, { "epoch": 0.32404596704249783, "grad_norm": 0.6167330145835876, "learning_rate": 8.952197881909183e-05, "loss": 0.1005, "step": 2989 }, { "epoch": 0.32415437987857765, "grad_norm": 0.26015275716781616, "learning_rate": 8.95183519512549e-05, "loss": 0.0261, "step": 2990 }, { "epoch": 0.3242627927146574, "grad_norm": 0.31390219926834106, "learning_rate": 8.951472508341796e-05, "loss": 0.0454, "step": 2991 }, { "epoch": 0.32437120555073723, "grad_norm": 0.45536643266677856, "learning_rate": 8.951109821558103e-05, "loss": 0.0473, "step": 2992 }, { "epoch": 0.324479618386817, "grad_norm": 0.3472692668437958, "learning_rate": 8.950747134774409e-05, "loss": 0.0662, "step": 2993 }, { "epoch": 0.3245880312228968, "grad_norm": 0.22382815182209015, "learning_rate": 8.950384447990716e-05, "loss": 0.0291, "step": 2994 }, { "epoch": 0.3246964440589766, "grad_norm": 0.1470162719488144, "learning_rate": 8.950021761207023e-05, "loss": 0.0232, "step": 2995 }, { "epoch": 0.3248048568950564, "grad_norm": 0.3024536073207855, "learning_rate": 8.949659074423329e-05, "loss": 0.0396, "step": 2996 }, { "epoch": 0.32491326973113616, "grad_norm": 0.5783097743988037, "learning_rate": 8.949296387639636e-05, "loss": 0.0427, "step": 2997 }, { "epoch": 0.325021682567216, "grad_norm": 0.3855043053627014, "learning_rate": 8.948933700855942e-05, "loss": 0.0704, "step": 2998 }, { "epoch": 0.32513009540329574, "grad_norm": 0.2625725567340851, "learning_rate": 8.948571014072247e-05, "loss": 0.0488, "step": 2999 }, { "epoch": 0.32523850823937556, "grad_norm": 0.19306419789791107, "learning_rate": 8.948208327288554e-05, "loss": 0.0336, "step": 3000 }, { "epoch": 0.3253469210754553, "grad_norm": 0.3231046795845032, "learning_rate": 8.94784564050486e-05, "loss": 0.0399, "step": 3001 }, { "epoch": 0.32545533391153514, "grad_norm": 0.7976841330528259, "learning_rate": 8.947482953721167e-05, "loss": 0.1038, "step": 3002 }, { "epoch": 0.3255637467476149, "grad_norm": 0.19214920699596405, "learning_rate": 8.947120266937473e-05, "loss": 0.015, "step": 3003 }, { "epoch": 0.3256721595836947, "grad_norm": 0.2684077322483063, "learning_rate": 8.94675758015378e-05, "loss": 0.0592, "step": 3004 }, { "epoch": 0.3257805724197745, "grad_norm": 0.17129698395729065, "learning_rate": 8.946394893370086e-05, "loss": 0.0278, "step": 3005 }, { "epoch": 0.3258889852558543, "grad_norm": 0.3745400011539459, "learning_rate": 8.946032206586393e-05, "loss": 0.0464, "step": 3006 }, { "epoch": 0.32599739809193407, "grad_norm": 0.2933962345123291, "learning_rate": 8.945669519802699e-05, "loss": 0.0467, "step": 3007 }, { "epoch": 0.3261058109280139, "grad_norm": 0.2852981388568878, "learning_rate": 8.945306833019004e-05, "loss": 0.019, "step": 3008 }, { "epoch": 0.32621422376409365, "grad_norm": 0.21002131700515747, "learning_rate": 8.944944146235311e-05, "loss": 0.0248, "step": 3009 }, { "epoch": 0.32632263660017347, "grad_norm": 0.37378862500190735, "learning_rate": 8.944581459451617e-05, "loss": 0.0926, "step": 3010 }, { "epoch": 0.32643104943625323, "grad_norm": 0.45568913221359253, "learning_rate": 8.944218772667926e-05, "loss": 0.0701, "step": 3011 }, { "epoch": 0.32653946227233305, "grad_norm": 0.5428444147109985, "learning_rate": 8.943856085884231e-05, "loss": 0.0579, "step": 3012 }, { "epoch": 0.3266478751084128, "grad_norm": 0.067023366689682, "learning_rate": 8.943493399100537e-05, "loss": 0.0092, "step": 3013 }, { "epoch": 0.32675628794449263, "grad_norm": 0.18736176192760468, "learning_rate": 8.943130712316844e-05, "loss": 0.0234, "step": 3014 }, { "epoch": 0.3268647007805724, "grad_norm": 2.1574888229370117, "learning_rate": 8.94276802553315e-05, "loss": 0.0542, "step": 3015 }, { "epoch": 0.3269731136166522, "grad_norm": 0.1498039811849594, "learning_rate": 8.942405338749457e-05, "loss": 0.0218, "step": 3016 }, { "epoch": 0.32708152645273203, "grad_norm": 0.172713965177536, "learning_rate": 8.942042651965763e-05, "loss": 0.0269, "step": 3017 }, { "epoch": 0.3271899392888118, "grad_norm": 0.4575663208961487, "learning_rate": 8.94167996518207e-05, "loss": 0.046, "step": 3018 }, { "epoch": 0.3272983521248916, "grad_norm": 0.4584386646747589, "learning_rate": 8.941317278398375e-05, "loss": 0.0381, "step": 3019 }, { "epoch": 0.3274067649609714, "grad_norm": 0.29930391907691956, "learning_rate": 8.940954591614683e-05, "loss": 0.0267, "step": 3020 }, { "epoch": 0.3275151777970512, "grad_norm": 0.24248500168323517, "learning_rate": 8.940591904830988e-05, "loss": 0.0357, "step": 3021 }, { "epoch": 0.32762359063313096, "grad_norm": 0.25806570053100586, "learning_rate": 8.940229218047294e-05, "loss": 0.0275, "step": 3022 }, { "epoch": 0.3277320034692108, "grad_norm": 0.6689634323120117, "learning_rate": 8.939866531263601e-05, "loss": 0.0602, "step": 3023 }, { "epoch": 0.32784041630529054, "grad_norm": 0.13040953874588013, "learning_rate": 8.939503844479907e-05, "loss": 0.0158, "step": 3024 }, { "epoch": 0.32794882914137036, "grad_norm": 0.2736049294471741, "learning_rate": 8.939141157696214e-05, "loss": 0.0261, "step": 3025 }, { "epoch": 0.3280572419774501, "grad_norm": 0.5157879590988159, "learning_rate": 8.93877847091252e-05, "loss": 0.0719, "step": 3026 }, { "epoch": 0.32816565481352994, "grad_norm": 0.253706157207489, "learning_rate": 8.938415784128827e-05, "loss": 0.0222, "step": 3027 }, { "epoch": 0.3282740676496097, "grad_norm": 0.5665046572685242, "learning_rate": 8.938053097345133e-05, "loss": 0.0442, "step": 3028 }, { "epoch": 0.3283824804856895, "grad_norm": 0.3605266809463501, "learning_rate": 8.93769041056144e-05, "loss": 0.0305, "step": 3029 }, { "epoch": 0.3284908933217693, "grad_norm": 0.16777680814266205, "learning_rate": 8.937327723777747e-05, "loss": 0.0238, "step": 3030 }, { "epoch": 0.3285993061578491, "grad_norm": 0.1794588416814804, "learning_rate": 8.936965036994052e-05, "loss": 0.0243, "step": 3031 }, { "epoch": 0.32870771899392887, "grad_norm": 0.6578787565231323, "learning_rate": 8.93660235021036e-05, "loss": 0.0659, "step": 3032 }, { "epoch": 0.3288161318300087, "grad_norm": 0.22610022127628326, "learning_rate": 8.936239663426665e-05, "loss": 0.0271, "step": 3033 }, { "epoch": 0.32892454466608845, "grad_norm": 0.23447608947753906, "learning_rate": 8.935876976642972e-05, "loss": 0.0345, "step": 3034 }, { "epoch": 0.32903295750216827, "grad_norm": 0.2737383544445038, "learning_rate": 8.935514289859278e-05, "loss": 0.0405, "step": 3035 }, { "epoch": 0.32914137033824803, "grad_norm": 0.643963634967804, "learning_rate": 8.935151603075584e-05, "loss": 0.0366, "step": 3036 }, { "epoch": 0.32924978317432785, "grad_norm": 0.21731607615947723, "learning_rate": 8.934788916291891e-05, "loss": 0.0238, "step": 3037 }, { "epoch": 0.3293581960104076, "grad_norm": 0.5377642512321472, "learning_rate": 8.934426229508197e-05, "loss": 0.0949, "step": 3038 }, { "epoch": 0.32946660884648743, "grad_norm": 0.44425228238105774, "learning_rate": 8.934063542724504e-05, "loss": 0.0658, "step": 3039 }, { "epoch": 0.3295750216825672, "grad_norm": 0.4875446557998657, "learning_rate": 8.93370085594081e-05, "loss": 0.0469, "step": 3040 }, { "epoch": 0.329683434518647, "grad_norm": 0.08218108117580414, "learning_rate": 8.933338169157117e-05, "loss": 0.0115, "step": 3041 }, { "epoch": 0.3297918473547268, "grad_norm": 0.3087231516838074, "learning_rate": 8.932975482373422e-05, "loss": 0.0693, "step": 3042 }, { "epoch": 0.3299002601908066, "grad_norm": 0.3570306897163391, "learning_rate": 8.93261279558973e-05, "loss": 0.0323, "step": 3043 }, { "epoch": 0.33000867302688636, "grad_norm": 0.16743890941143036, "learning_rate": 8.932250108806035e-05, "loss": 0.016, "step": 3044 }, { "epoch": 0.3301170858629662, "grad_norm": 0.3280717730522156, "learning_rate": 8.931887422022341e-05, "loss": 0.061, "step": 3045 }, { "epoch": 0.330225498699046, "grad_norm": 0.5361069440841675, "learning_rate": 8.931524735238649e-05, "loss": 0.0665, "step": 3046 }, { "epoch": 0.33033391153512576, "grad_norm": 0.4630265235900879, "learning_rate": 8.931162048454955e-05, "loss": 0.0285, "step": 3047 }, { "epoch": 0.3304423243712056, "grad_norm": 0.2649438977241516, "learning_rate": 8.930799361671262e-05, "loss": 0.0451, "step": 3048 }, { "epoch": 0.33055073720728534, "grad_norm": 0.34400197863578796, "learning_rate": 8.930436674887568e-05, "loss": 0.0464, "step": 3049 }, { "epoch": 0.33065915004336516, "grad_norm": 0.19419370591640472, "learning_rate": 8.930073988103874e-05, "loss": 0.0255, "step": 3050 }, { "epoch": 0.3307675628794449, "grad_norm": 0.5228205919265747, "learning_rate": 8.92971130132018e-05, "loss": 0.0977, "step": 3051 }, { "epoch": 0.33087597571552474, "grad_norm": 0.5105196833610535, "learning_rate": 8.929348614536486e-05, "loss": 0.0489, "step": 3052 }, { "epoch": 0.3309843885516045, "grad_norm": 0.6130709648132324, "learning_rate": 8.928985927752793e-05, "loss": 0.0507, "step": 3053 }, { "epoch": 0.3310928013876843, "grad_norm": 0.7311919927597046, "learning_rate": 8.928623240969099e-05, "loss": 0.0579, "step": 3054 }, { "epoch": 0.3312012142237641, "grad_norm": 0.38462674617767334, "learning_rate": 8.928260554185406e-05, "loss": 0.0375, "step": 3055 }, { "epoch": 0.3313096270598439, "grad_norm": 0.5019206404685974, "learning_rate": 8.927897867401712e-05, "loss": 0.1449, "step": 3056 }, { "epoch": 0.33141803989592367, "grad_norm": 0.38225895166397095, "learning_rate": 8.927535180618019e-05, "loss": 0.0378, "step": 3057 }, { "epoch": 0.3315264527320035, "grad_norm": 0.22786681354045868, "learning_rate": 8.927172493834325e-05, "loss": 0.0249, "step": 3058 }, { "epoch": 0.33163486556808325, "grad_norm": 0.488812118768692, "learning_rate": 8.92680980705063e-05, "loss": 0.0767, "step": 3059 }, { "epoch": 0.33174327840416307, "grad_norm": 0.5094360113143921, "learning_rate": 8.926447120266938e-05, "loss": 0.034, "step": 3060 }, { "epoch": 0.33185169124024283, "grad_norm": 0.6326332092285156, "learning_rate": 8.926084433483243e-05, "loss": 0.0538, "step": 3061 }, { "epoch": 0.33196010407632265, "grad_norm": 0.3917277455329895, "learning_rate": 8.92572174669955e-05, "loss": 0.0507, "step": 3062 }, { "epoch": 0.3320685169124024, "grad_norm": 0.08012113720178604, "learning_rate": 8.925359059915858e-05, "loss": 0.0086, "step": 3063 }, { "epoch": 0.33217692974848223, "grad_norm": 0.32387131452560425, "learning_rate": 8.924996373132165e-05, "loss": 0.0405, "step": 3064 }, { "epoch": 0.332285342584562, "grad_norm": 0.28315454721450806, "learning_rate": 8.92463368634847e-05, "loss": 0.0405, "step": 3065 }, { "epoch": 0.3323937554206418, "grad_norm": 0.31737956404685974, "learning_rate": 8.924270999564776e-05, "loss": 0.0633, "step": 3066 }, { "epoch": 0.3325021682567216, "grad_norm": 0.2442808598279953, "learning_rate": 8.923908312781083e-05, "loss": 0.0375, "step": 3067 }, { "epoch": 0.3326105810928014, "grad_norm": 0.30080968141555786, "learning_rate": 8.923545625997389e-05, "loss": 0.063, "step": 3068 }, { "epoch": 0.33271899392888116, "grad_norm": 0.6021973490715027, "learning_rate": 8.923182939213696e-05, "loss": 0.0777, "step": 3069 }, { "epoch": 0.332827406764961, "grad_norm": 0.5037634968757629, "learning_rate": 8.922820252430002e-05, "loss": 0.0653, "step": 3070 }, { "epoch": 0.33293581960104074, "grad_norm": 0.3896307647228241, "learning_rate": 8.922457565646309e-05, "loss": 0.0351, "step": 3071 }, { "epoch": 0.33304423243712056, "grad_norm": 0.3416818380355835, "learning_rate": 8.922094878862615e-05, "loss": 0.0436, "step": 3072 }, { "epoch": 0.3331526452732003, "grad_norm": 0.2977372407913208, "learning_rate": 8.92173219207892e-05, "loss": 0.0294, "step": 3073 }, { "epoch": 0.33326105810928014, "grad_norm": 0.40777313709259033, "learning_rate": 8.921369505295227e-05, "loss": 0.0394, "step": 3074 }, { "epoch": 0.33336947094535996, "grad_norm": 0.5249201655387878, "learning_rate": 8.921006818511533e-05, "loss": 0.0622, "step": 3075 }, { "epoch": 0.3334778837814397, "grad_norm": 0.3024139106273651, "learning_rate": 8.92064413172784e-05, "loss": 0.0436, "step": 3076 }, { "epoch": 0.33358629661751954, "grad_norm": 0.3911679983139038, "learning_rate": 8.920281444944146e-05, "loss": 0.0517, "step": 3077 }, { "epoch": 0.3336947094535993, "grad_norm": 0.5344608426094055, "learning_rate": 8.919918758160453e-05, "loss": 0.0543, "step": 3078 }, { "epoch": 0.3338031222896791, "grad_norm": 0.6152700781822205, "learning_rate": 8.919556071376759e-05, "loss": 0.0783, "step": 3079 }, { "epoch": 0.3339115351257589, "grad_norm": 0.6389613747596741, "learning_rate": 8.919193384593066e-05, "loss": 0.0669, "step": 3080 }, { "epoch": 0.3340199479618387, "grad_norm": 0.5607441663742065, "learning_rate": 8.918830697809373e-05, "loss": 0.0677, "step": 3081 }, { "epoch": 0.33412836079791847, "grad_norm": 0.29375529289245605, "learning_rate": 8.918468011025679e-05, "loss": 0.0269, "step": 3082 }, { "epoch": 0.3342367736339983, "grad_norm": 0.36664018034935, "learning_rate": 8.918105324241986e-05, "loss": 0.0587, "step": 3083 }, { "epoch": 0.33434518647007805, "grad_norm": 0.44292768836021423, "learning_rate": 8.917742637458291e-05, "loss": 0.0947, "step": 3084 }, { "epoch": 0.33445359930615787, "grad_norm": 0.2370891273021698, "learning_rate": 8.917379950674599e-05, "loss": 0.0221, "step": 3085 }, { "epoch": 0.33456201214223763, "grad_norm": 0.245186910033226, "learning_rate": 8.917017263890904e-05, "loss": 0.0338, "step": 3086 }, { "epoch": 0.33467042497831745, "grad_norm": 0.6006591320037842, "learning_rate": 8.916654577107211e-05, "loss": 0.0472, "step": 3087 }, { "epoch": 0.3347788378143972, "grad_norm": 0.4159989356994629, "learning_rate": 8.916291890323517e-05, "loss": 0.0808, "step": 3088 }, { "epoch": 0.33488725065047703, "grad_norm": 0.24360746145248413, "learning_rate": 8.915929203539823e-05, "loss": 0.0445, "step": 3089 }, { "epoch": 0.3349956634865568, "grad_norm": 0.570694625377655, "learning_rate": 8.91556651675613e-05, "loss": 0.0437, "step": 3090 }, { "epoch": 0.3351040763226366, "grad_norm": 0.5754511952400208, "learning_rate": 8.915203829972436e-05, "loss": 0.0461, "step": 3091 }, { "epoch": 0.3352124891587164, "grad_norm": 0.7809154391288757, "learning_rate": 8.914841143188743e-05, "loss": 0.1454, "step": 3092 }, { "epoch": 0.3353209019947962, "grad_norm": 0.41862261295318604, "learning_rate": 8.914478456405049e-05, "loss": 0.0776, "step": 3093 }, { "epoch": 0.33542931483087596, "grad_norm": 0.35804715752601624, "learning_rate": 8.914115769621356e-05, "loss": 0.0447, "step": 3094 }, { "epoch": 0.3355377276669558, "grad_norm": 0.17507107555866241, "learning_rate": 8.913753082837661e-05, "loss": 0.0084, "step": 3095 }, { "epoch": 0.33564614050303554, "grad_norm": 0.44478511810302734, "learning_rate": 8.913390396053968e-05, "loss": 0.0647, "step": 3096 }, { "epoch": 0.33575455333911536, "grad_norm": 0.19956493377685547, "learning_rate": 8.913027709270274e-05, "loss": 0.0369, "step": 3097 }, { "epoch": 0.3358629661751951, "grad_norm": 0.4330681562423706, "learning_rate": 8.912665022486581e-05, "loss": 0.0479, "step": 3098 }, { "epoch": 0.33597137901127494, "grad_norm": 0.4435941278934479, "learning_rate": 8.912302335702888e-05, "loss": 0.0605, "step": 3099 }, { "epoch": 0.3360797918473547, "grad_norm": 0.37504199147224426, "learning_rate": 8.911939648919194e-05, "loss": 0.0466, "step": 3100 }, { "epoch": 0.3361882046834345, "grad_norm": 0.1892721801996231, "learning_rate": 8.911576962135501e-05, "loss": 0.0252, "step": 3101 }, { "epoch": 0.3362966175195143, "grad_norm": 0.09227219969034195, "learning_rate": 8.911214275351807e-05, "loss": 0.012, "step": 3102 }, { "epoch": 0.3364050303555941, "grad_norm": 0.1881321668624878, "learning_rate": 8.910851588568113e-05, "loss": 0.0203, "step": 3103 }, { "epoch": 0.33651344319167387, "grad_norm": 0.30297380685806274, "learning_rate": 8.91048890178442e-05, "loss": 0.0314, "step": 3104 }, { "epoch": 0.3366218560277537, "grad_norm": 0.27398037910461426, "learning_rate": 8.910126215000725e-05, "loss": 0.0354, "step": 3105 }, { "epoch": 0.3367302688638335, "grad_norm": 0.3151945173740387, "learning_rate": 8.909763528217033e-05, "loss": 0.0347, "step": 3106 }, { "epoch": 0.33683868169991327, "grad_norm": 0.3107323944568634, "learning_rate": 8.909400841433338e-05, "loss": 0.0315, "step": 3107 }, { "epoch": 0.3369470945359931, "grad_norm": 0.32798072695732117, "learning_rate": 8.909038154649645e-05, "loss": 0.0495, "step": 3108 }, { "epoch": 0.33705550737207285, "grad_norm": 0.3779139816761017, "learning_rate": 8.908675467865951e-05, "loss": 0.0316, "step": 3109 }, { "epoch": 0.33716392020815267, "grad_norm": 0.3062788248062134, "learning_rate": 8.908312781082258e-05, "loss": 0.0182, "step": 3110 }, { "epoch": 0.33727233304423243, "grad_norm": 0.36887526512145996, "learning_rate": 8.907950094298564e-05, "loss": 0.0381, "step": 3111 }, { "epoch": 0.33738074588031225, "grad_norm": 0.39989668130874634, "learning_rate": 8.90758740751487e-05, "loss": 0.0606, "step": 3112 }, { "epoch": 0.337489158716392, "grad_norm": 0.7532022595405579, "learning_rate": 8.907224720731177e-05, "loss": 0.0695, "step": 3113 }, { "epoch": 0.33759757155247183, "grad_norm": 0.6355072855949402, "learning_rate": 8.906862033947482e-05, "loss": 0.0861, "step": 3114 }, { "epoch": 0.3377059843885516, "grad_norm": 0.10996527224779129, "learning_rate": 8.906499347163791e-05, "loss": 0.008, "step": 3115 }, { "epoch": 0.3378143972246314, "grad_norm": 0.5146927833557129, "learning_rate": 8.906136660380097e-05, "loss": 0.0693, "step": 3116 }, { "epoch": 0.3379228100607112, "grad_norm": 0.3871355354785919, "learning_rate": 8.905773973596402e-05, "loss": 0.0332, "step": 3117 }, { "epoch": 0.338031222896791, "grad_norm": 0.42834800481796265, "learning_rate": 8.90541128681271e-05, "loss": 0.0622, "step": 3118 }, { "epoch": 0.33813963573287076, "grad_norm": 0.37789830565452576, "learning_rate": 8.905048600029015e-05, "loss": 0.0354, "step": 3119 }, { "epoch": 0.3382480485689506, "grad_norm": 1.629905343055725, "learning_rate": 8.904685913245322e-05, "loss": 0.0684, "step": 3120 }, { "epoch": 0.33835646140503034, "grad_norm": 0.2601810097694397, "learning_rate": 8.904323226461628e-05, "loss": 0.035, "step": 3121 }, { "epoch": 0.33846487424111016, "grad_norm": 0.8270122408866882, "learning_rate": 8.903960539677935e-05, "loss": 0.0621, "step": 3122 }, { "epoch": 0.3385732870771899, "grad_norm": 0.288762629032135, "learning_rate": 8.903597852894241e-05, "loss": 0.0426, "step": 3123 }, { "epoch": 0.33868169991326974, "grad_norm": 0.4134001135826111, "learning_rate": 8.903235166110548e-05, "loss": 0.0682, "step": 3124 }, { "epoch": 0.3387901127493495, "grad_norm": 0.39978107810020447, "learning_rate": 8.902872479326854e-05, "loss": 0.0279, "step": 3125 }, { "epoch": 0.3388985255854293, "grad_norm": 0.2099224030971527, "learning_rate": 8.90250979254316e-05, "loss": 0.0252, "step": 3126 }, { "epoch": 0.3390069384215091, "grad_norm": 0.2907271981239319, "learning_rate": 8.902147105759466e-05, "loss": 0.0365, "step": 3127 }, { "epoch": 0.3391153512575889, "grad_norm": 0.3481152355670929, "learning_rate": 8.901784418975772e-05, "loss": 0.0792, "step": 3128 }, { "epoch": 0.33922376409366867, "grad_norm": 0.36313188076019287, "learning_rate": 8.901421732192079e-05, "loss": 0.0409, "step": 3129 }, { "epoch": 0.3393321769297485, "grad_norm": 0.4453877806663513, "learning_rate": 8.901059045408385e-05, "loss": 0.0287, "step": 3130 }, { "epoch": 0.33944058976582825, "grad_norm": 0.09736254066228867, "learning_rate": 8.900696358624692e-05, "loss": 0.0118, "step": 3131 }, { "epoch": 0.33954900260190807, "grad_norm": 3.429888963699341, "learning_rate": 8.900333671840999e-05, "loss": 0.2119, "step": 3132 }, { "epoch": 0.33965741543798783, "grad_norm": 0.3259426951408386, "learning_rate": 8.899970985057305e-05, "loss": 0.0519, "step": 3133 }, { "epoch": 0.33976582827406765, "grad_norm": 0.201762855052948, "learning_rate": 8.899608298273612e-05, "loss": 0.0356, "step": 3134 }, { "epoch": 0.33987424111014747, "grad_norm": 0.31089290976524353, "learning_rate": 8.899245611489918e-05, "loss": 0.0557, "step": 3135 }, { "epoch": 0.33998265394622723, "grad_norm": 0.3269970417022705, "learning_rate": 8.898882924706225e-05, "loss": 0.0249, "step": 3136 }, { "epoch": 0.34009106678230705, "grad_norm": 0.4653986096382141, "learning_rate": 8.89852023792253e-05, "loss": 0.076, "step": 3137 }, { "epoch": 0.3401994796183868, "grad_norm": 0.455291211605072, "learning_rate": 8.898157551138838e-05, "loss": 0.0857, "step": 3138 }, { "epoch": 0.34030789245446663, "grad_norm": 0.43603426218032837, "learning_rate": 8.897794864355143e-05, "loss": 0.0401, "step": 3139 }, { "epoch": 0.3404163052905464, "grad_norm": 0.23808273673057556, "learning_rate": 8.897432177571449e-05, "loss": 0.0266, "step": 3140 }, { "epoch": 0.3405247181266262, "grad_norm": 0.4085344076156616, "learning_rate": 8.897069490787756e-05, "loss": 0.0682, "step": 3141 }, { "epoch": 0.340633130962706, "grad_norm": 0.5709272027015686, "learning_rate": 8.896706804004062e-05, "loss": 0.0479, "step": 3142 }, { "epoch": 0.3407415437987858, "grad_norm": 0.11720207333564758, "learning_rate": 8.896344117220369e-05, "loss": 0.0394, "step": 3143 }, { "epoch": 0.34084995663486556, "grad_norm": 0.3006024658679962, "learning_rate": 8.895981430436675e-05, "loss": 0.0588, "step": 3144 }, { "epoch": 0.3409583694709454, "grad_norm": 0.21336078643798828, "learning_rate": 8.895618743652982e-05, "loss": 0.0347, "step": 3145 }, { "epoch": 0.34106678230702514, "grad_norm": 0.349517285823822, "learning_rate": 8.895256056869288e-05, "loss": 0.0279, "step": 3146 }, { "epoch": 0.34117519514310496, "grad_norm": 0.46404629945755005, "learning_rate": 8.894893370085595e-05, "loss": 0.0498, "step": 3147 }, { "epoch": 0.3412836079791847, "grad_norm": 0.8497287631034851, "learning_rate": 8.8945306833019e-05, "loss": 0.0932, "step": 3148 }, { "epoch": 0.34139202081526454, "grad_norm": 1.050001859664917, "learning_rate": 8.894167996518206e-05, "loss": 0.0692, "step": 3149 }, { "epoch": 0.3415004336513443, "grad_norm": 0.4923873543739319, "learning_rate": 8.893805309734515e-05, "loss": 0.0237, "step": 3150 }, { "epoch": 0.3416088464874241, "grad_norm": 0.1847352534532547, "learning_rate": 8.89344262295082e-05, "loss": 0.0233, "step": 3151 }, { "epoch": 0.3417172593235039, "grad_norm": 0.7028707265853882, "learning_rate": 8.893079936167127e-05, "loss": 0.0429, "step": 3152 }, { "epoch": 0.3418256721595837, "grad_norm": 0.1796208620071411, "learning_rate": 8.892717249383433e-05, "loss": 0.0295, "step": 3153 }, { "epoch": 0.34193408499566347, "grad_norm": 0.22598041594028473, "learning_rate": 8.892354562599739e-05, "loss": 0.0342, "step": 3154 }, { "epoch": 0.3420424978317433, "grad_norm": 0.15058930218219757, "learning_rate": 8.891991875816046e-05, "loss": 0.0215, "step": 3155 }, { "epoch": 0.34215091066782305, "grad_norm": 0.3342167139053345, "learning_rate": 8.891629189032352e-05, "loss": 0.0429, "step": 3156 }, { "epoch": 0.34225932350390287, "grad_norm": 0.1863754689693451, "learning_rate": 8.891266502248659e-05, "loss": 0.0312, "step": 3157 }, { "epoch": 0.34236773633998263, "grad_norm": 0.5014959573745728, "learning_rate": 8.890903815464965e-05, "loss": 0.0494, "step": 3158 }, { "epoch": 0.34247614917606245, "grad_norm": 0.3094920217990875, "learning_rate": 8.890541128681272e-05, "loss": 0.0521, "step": 3159 }, { "epoch": 0.3425845620121422, "grad_norm": 0.3013620972633362, "learning_rate": 8.890178441897577e-05, "loss": 0.0382, "step": 3160 }, { "epoch": 0.34269297484822203, "grad_norm": 0.33720260858535767, "learning_rate": 8.889815755113884e-05, "loss": 0.0371, "step": 3161 }, { "epoch": 0.3428013876843018, "grad_norm": 0.3551321029663086, "learning_rate": 8.88945306833019e-05, "loss": 0.0731, "step": 3162 }, { "epoch": 0.3429098005203816, "grad_norm": 0.669546902179718, "learning_rate": 8.889090381546496e-05, "loss": 0.063, "step": 3163 }, { "epoch": 0.34301821335646143, "grad_norm": 0.4320206642150879, "learning_rate": 8.888727694762803e-05, "loss": 0.0589, "step": 3164 }, { "epoch": 0.3431266261925412, "grad_norm": 0.4712567925453186, "learning_rate": 8.888365007979109e-05, "loss": 0.059, "step": 3165 }, { "epoch": 0.343235039028621, "grad_norm": 0.3496328592300415, "learning_rate": 8.888002321195416e-05, "loss": 0.0319, "step": 3166 }, { "epoch": 0.3433434518647008, "grad_norm": 0.37088730931282043, "learning_rate": 8.887639634411723e-05, "loss": 0.0417, "step": 3167 }, { "epoch": 0.3434518647007806, "grad_norm": 0.9036403298377991, "learning_rate": 8.88727694762803e-05, "loss": 0.0425, "step": 3168 }, { "epoch": 0.34356027753686036, "grad_norm": 0.24411870539188385, "learning_rate": 8.886914260844336e-05, "loss": 0.0242, "step": 3169 }, { "epoch": 0.3436686903729402, "grad_norm": 0.4587036967277527, "learning_rate": 8.886551574060641e-05, "loss": 0.0301, "step": 3170 }, { "epoch": 0.34377710320901994, "grad_norm": 0.17304272949695587, "learning_rate": 8.886188887276949e-05, "loss": 0.0104, "step": 3171 }, { "epoch": 0.34388551604509976, "grad_norm": 0.26573216915130615, "learning_rate": 8.885826200493254e-05, "loss": 0.0409, "step": 3172 }, { "epoch": 0.3439939288811795, "grad_norm": 0.15727949142456055, "learning_rate": 8.885463513709561e-05, "loss": 0.029, "step": 3173 }, { "epoch": 0.34410234171725934, "grad_norm": 0.463543176651001, "learning_rate": 8.885100826925867e-05, "loss": 0.0219, "step": 3174 }, { "epoch": 0.3442107545533391, "grad_norm": 0.37684476375579834, "learning_rate": 8.884738140142174e-05, "loss": 0.0573, "step": 3175 }, { "epoch": 0.3443191673894189, "grad_norm": 0.748266339302063, "learning_rate": 8.88437545335848e-05, "loss": 0.0944, "step": 3176 }, { "epoch": 0.3444275802254987, "grad_norm": 0.3708208501338959, "learning_rate": 8.884012766574787e-05, "loss": 0.0799, "step": 3177 }, { "epoch": 0.3445359930615785, "grad_norm": 0.39948996901512146, "learning_rate": 8.883650079791093e-05, "loss": 0.045, "step": 3178 }, { "epoch": 0.34464440589765827, "grad_norm": 0.3192654550075531, "learning_rate": 8.883287393007398e-05, "loss": 0.0519, "step": 3179 }, { "epoch": 0.3447528187337381, "grad_norm": 0.3375509977340698, "learning_rate": 8.882924706223706e-05, "loss": 0.0628, "step": 3180 }, { "epoch": 0.34486123156981785, "grad_norm": 0.7052115797996521, "learning_rate": 8.882562019440011e-05, "loss": 0.0165, "step": 3181 }, { "epoch": 0.34496964440589767, "grad_norm": 0.5987801551818848, "learning_rate": 8.882199332656318e-05, "loss": 0.0578, "step": 3182 }, { "epoch": 0.34507805724197743, "grad_norm": 0.3803057372570038, "learning_rate": 8.881836645872624e-05, "loss": 0.0713, "step": 3183 }, { "epoch": 0.34518647007805725, "grad_norm": 0.33783426880836487, "learning_rate": 8.881473959088931e-05, "loss": 0.0397, "step": 3184 }, { "epoch": 0.345294882914137, "grad_norm": 0.5873101353645325, "learning_rate": 8.881111272305238e-05, "loss": 0.036, "step": 3185 }, { "epoch": 0.34540329575021683, "grad_norm": 0.5583386421203613, "learning_rate": 8.880748585521544e-05, "loss": 0.0815, "step": 3186 }, { "epoch": 0.3455117085862966, "grad_norm": 0.6404353380203247, "learning_rate": 8.880385898737851e-05, "loss": 0.0658, "step": 3187 }, { "epoch": 0.3456201214223764, "grad_norm": 0.3173501789569855, "learning_rate": 8.880023211954157e-05, "loss": 0.0311, "step": 3188 }, { "epoch": 0.3457285342584562, "grad_norm": 0.221134215593338, "learning_rate": 8.879660525170464e-05, "loss": 0.0403, "step": 3189 }, { "epoch": 0.345836947094536, "grad_norm": 0.9836069941520691, "learning_rate": 8.87929783838677e-05, "loss": 0.0947, "step": 3190 }, { "epoch": 0.34594535993061576, "grad_norm": 0.4939429759979248, "learning_rate": 8.878935151603077e-05, "loss": 0.079, "step": 3191 }, { "epoch": 0.3460537727666956, "grad_norm": 0.4039759635925293, "learning_rate": 8.878572464819382e-05, "loss": 0.0476, "step": 3192 }, { "epoch": 0.3461621856027754, "grad_norm": 0.12960229814052582, "learning_rate": 8.878209778035688e-05, "loss": 0.018, "step": 3193 }, { "epoch": 0.34627059843885516, "grad_norm": 0.9260284900665283, "learning_rate": 8.877847091251995e-05, "loss": 0.0534, "step": 3194 }, { "epoch": 0.346379011274935, "grad_norm": 0.2225998044013977, "learning_rate": 8.877484404468301e-05, "loss": 0.0283, "step": 3195 }, { "epoch": 0.34648742411101474, "grad_norm": 0.46011853218078613, "learning_rate": 8.877121717684608e-05, "loss": 0.0367, "step": 3196 }, { "epoch": 0.34659583694709456, "grad_norm": 0.5061470866203308, "learning_rate": 8.876759030900914e-05, "loss": 0.0668, "step": 3197 }, { "epoch": 0.3467042497831743, "grad_norm": 0.218892440199852, "learning_rate": 8.876396344117221e-05, "loss": 0.0268, "step": 3198 }, { "epoch": 0.34681266261925414, "grad_norm": 0.7441394925117493, "learning_rate": 8.876033657333527e-05, "loss": 0.0676, "step": 3199 }, { "epoch": 0.3469210754553339, "grad_norm": 0.30446532368659973, "learning_rate": 8.875670970549834e-05, "loss": 0.042, "step": 3200 }, { "epoch": 0.3470294882914137, "grad_norm": 0.1745060533285141, "learning_rate": 8.875308283766141e-05, "loss": 0.0255, "step": 3201 }, { "epoch": 0.3471379011274935, "grad_norm": 0.5498632192611694, "learning_rate": 8.874945596982447e-05, "loss": 0.0614, "step": 3202 }, { "epoch": 0.3472463139635733, "grad_norm": 0.17397287487983704, "learning_rate": 8.874582910198754e-05, "loss": 0.0391, "step": 3203 }, { "epoch": 0.34735472679965307, "grad_norm": 0.19756270945072174, "learning_rate": 8.87422022341506e-05, "loss": 0.0434, "step": 3204 }, { "epoch": 0.3474631396357329, "grad_norm": 0.31471821665763855, "learning_rate": 8.873857536631366e-05, "loss": 0.0431, "step": 3205 }, { "epoch": 0.34757155247181265, "grad_norm": 0.7597004175186157, "learning_rate": 8.873494849847672e-05, "loss": 0.0895, "step": 3206 }, { "epoch": 0.34767996530789247, "grad_norm": 0.33207449316978455, "learning_rate": 8.873132163063978e-05, "loss": 0.0281, "step": 3207 }, { "epoch": 0.34778837814397223, "grad_norm": 0.4147017300128937, "learning_rate": 8.872769476280285e-05, "loss": 0.0434, "step": 3208 }, { "epoch": 0.34789679098005205, "grad_norm": 0.26630645990371704, "learning_rate": 8.872406789496591e-05, "loss": 0.0332, "step": 3209 }, { "epoch": 0.3480052038161318, "grad_norm": 0.24349050223827362, "learning_rate": 8.872044102712898e-05, "loss": 0.0749, "step": 3210 }, { "epoch": 0.34811361665221163, "grad_norm": 0.31825536489486694, "learning_rate": 8.871681415929204e-05, "loss": 0.022, "step": 3211 }, { "epoch": 0.3482220294882914, "grad_norm": 0.2022983431816101, "learning_rate": 8.871318729145511e-05, "loss": 0.0282, "step": 3212 }, { "epoch": 0.3483304423243712, "grad_norm": 0.22833961248397827, "learning_rate": 8.870956042361816e-05, "loss": 0.0308, "step": 3213 }, { "epoch": 0.348438855160451, "grad_norm": 0.25958457589149475, "learning_rate": 8.870593355578123e-05, "loss": 0.0256, "step": 3214 }, { "epoch": 0.3485472679965308, "grad_norm": 0.47968271374702454, "learning_rate": 8.870230668794429e-05, "loss": 0.083, "step": 3215 }, { "epoch": 0.34865568083261056, "grad_norm": 0.4256824851036072, "learning_rate": 8.869867982010735e-05, "loss": 0.0528, "step": 3216 }, { "epoch": 0.3487640936686904, "grad_norm": 0.6075900197029114, "learning_rate": 8.869505295227042e-05, "loss": 0.0763, "step": 3217 }, { "epoch": 0.34887250650477014, "grad_norm": 0.23024316132068634, "learning_rate": 8.869142608443348e-05, "loss": 0.0374, "step": 3218 }, { "epoch": 0.34898091934084996, "grad_norm": 0.2533360719680786, "learning_rate": 8.868779921659656e-05, "loss": 0.0344, "step": 3219 }, { "epoch": 0.3490893321769297, "grad_norm": 0.5123780369758606, "learning_rate": 8.868417234875962e-05, "loss": 0.036, "step": 3220 }, { "epoch": 0.34919774501300954, "grad_norm": 0.7971088886260986, "learning_rate": 8.868054548092268e-05, "loss": 0.039, "step": 3221 }, { "epoch": 0.3493061578490893, "grad_norm": 0.4235641062259674, "learning_rate": 8.867691861308575e-05, "loss": 0.0625, "step": 3222 }, { "epoch": 0.3494145706851691, "grad_norm": 0.43448084592819214, "learning_rate": 8.86732917452488e-05, "loss": 0.0547, "step": 3223 }, { "epoch": 0.34952298352124894, "grad_norm": 0.27193278074264526, "learning_rate": 8.866966487741188e-05, "loss": 0.0461, "step": 3224 }, { "epoch": 0.3496313963573287, "grad_norm": 0.6484523415565491, "learning_rate": 8.866603800957493e-05, "loss": 0.1035, "step": 3225 }, { "epoch": 0.3497398091934085, "grad_norm": 0.339155375957489, "learning_rate": 8.8662411141738e-05, "loss": 0.0318, "step": 3226 }, { "epoch": 0.3498482220294883, "grad_norm": 0.6604996919631958, "learning_rate": 8.865878427390106e-05, "loss": 0.0989, "step": 3227 }, { "epoch": 0.3499566348655681, "grad_norm": 0.25196677446365356, "learning_rate": 8.865515740606413e-05, "loss": 0.0224, "step": 3228 }, { "epoch": 0.35006504770164787, "grad_norm": 0.2748572528362274, "learning_rate": 8.865153053822719e-05, "loss": 0.0416, "step": 3229 }, { "epoch": 0.3501734605377277, "grad_norm": 2.4470462799072266, "learning_rate": 8.864790367039025e-05, "loss": 0.0582, "step": 3230 }, { "epoch": 0.35028187337380745, "grad_norm": 0.41232141852378845, "learning_rate": 8.864427680255332e-05, "loss": 0.0434, "step": 3231 }, { "epoch": 0.35039028620988727, "grad_norm": 0.5015106201171875, "learning_rate": 8.864064993471638e-05, "loss": 0.0853, "step": 3232 }, { "epoch": 0.35049869904596703, "grad_norm": 0.46533843874931335, "learning_rate": 8.863702306687945e-05, "loss": 0.0434, "step": 3233 }, { "epoch": 0.35060711188204685, "grad_norm": 0.2646383047103882, "learning_rate": 8.86333961990425e-05, "loss": 0.042, "step": 3234 }, { "epoch": 0.3507155247181266, "grad_norm": 0.31092140078544617, "learning_rate": 8.862976933120557e-05, "loss": 0.0468, "step": 3235 }, { "epoch": 0.35082393755420643, "grad_norm": 0.3265778124332428, "learning_rate": 8.862614246336865e-05, "loss": 0.0439, "step": 3236 }, { "epoch": 0.3509323503902862, "grad_norm": 0.259143590927124, "learning_rate": 8.86225155955317e-05, "loss": 0.0368, "step": 3237 }, { "epoch": 0.351040763226366, "grad_norm": 0.2545570731163025, "learning_rate": 8.861888872769477e-05, "loss": 0.0328, "step": 3238 }, { "epoch": 0.3511491760624458, "grad_norm": 0.25819262862205505, "learning_rate": 8.861526185985783e-05, "loss": 0.0412, "step": 3239 }, { "epoch": 0.3512575888985256, "grad_norm": 0.5508299469947815, "learning_rate": 8.86116349920209e-05, "loss": 0.0831, "step": 3240 }, { "epoch": 0.35136600173460536, "grad_norm": 0.35843169689178467, "learning_rate": 8.860800812418396e-05, "loss": 0.0258, "step": 3241 }, { "epoch": 0.3514744145706852, "grad_norm": 0.2585332691669464, "learning_rate": 8.860438125634703e-05, "loss": 0.0346, "step": 3242 }, { "epoch": 0.35158282740676494, "grad_norm": 0.4620535671710968, "learning_rate": 8.860075438851009e-05, "loss": 0.077, "step": 3243 }, { "epoch": 0.35169124024284476, "grad_norm": 0.25244107842445374, "learning_rate": 8.859712752067314e-05, "loss": 0.0407, "step": 3244 }, { "epoch": 0.3517996530789245, "grad_norm": 0.5529150366783142, "learning_rate": 8.859350065283622e-05, "loss": 0.095, "step": 3245 }, { "epoch": 0.35190806591500434, "grad_norm": 0.1336471438407898, "learning_rate": 8.858987378499927e-05, "loss": 0.0211, "step": 3246 }, { "epoch": 0.3520164787510841, "grad_norm": 0.4691210389137268, "learning_rate": 8.858624691716234e-05, "loss": 0.0752, "step": 3247 }, { "epoch": 0.3521248915871639, "grad_norm": 0.2747996747493744, "learning_rate": 8.85826200493254e-05, "loss": 0.0351, "step": 3248 }, { "epoch": 0.3522333044232437, "grad_norm": 0.31927111744880676, "learning_rate": 8.857899318148847e-05, "loss": 0.0556, "step": 3249 }, { "epoch": 0.3523417172593235, "grad_norm": 0.24899472296237946, "learning_rate": 8.857536631365153e-05, "loss": 0.0281, "step": 3250 }, { "epoch": 0.35245013009540327, "grad_norm": 0.4565246105194092, "learning_rate": 8.85717394458146e-05, "loss": 0.0812, "step": 3251 }, { "epoch": 0.3525585429314831, "grad_norm": 0.4324013292789459, "learning_rate": 8.856811257797766e-05, "loss": 0.077, "step": 3252 }, { "epoch": 0.3526669557675629, "grad_norm": 0.2611502408981323, "learning_rate": 8.856448571014073e-05, "loss": 0.0482, "step": 3253 }, { "epoch": 0.35277536860364267, "grad_norm": 0.43479523062705994, "learning_rate": 8.85608588423038e-05, "loss": 0.0501, "step": 3254 }, { "epoch": 0.3528837814397225, "grad_norm": 0.3390105366706848, "learning_rate": 8.855723197446686e-05, "loss": 0.045, "step": 3255 }, { "epoch": 0.35299219427580225, "grad_norm": 0.4794495105743408, "learning_rate": 8.855360510662993e-05, "loss": 0.0666, "step": 3256 }, { "epoch": 0.35310060711188207, "grad_norm": 0.1833738386631012, "learning_rate": 8.854997823879298e-05, "loss": 0.0296, "step": 3257 }, { "epoch": 0.35320901994796183, "grad_norm": 0.23682145774364471, "learning_rate": 8.854635137095604e-05, "loss": 0.0393, "step": 3258 }, { "epoch": 0.35331743278404165, "grad_norm": 0.33307287096977234, "learning_rate": 8.854272450311911e-05, "loss": 0.0395, "step": 3259 }, { "epoch": 0.3534258456201214, "grad_norm": 0.2854507863521576, "learning_rate": 8.853909763528217e-05, "loss": 0.0353, "step": 3260 }, { "epoch": 0.35353425845620123, "grad_norm": 0.19079960882663727, "learning_rate": 8.853547076744524e-05, "loss": 0.0195, "step": 3261 }, { "epoch": 0.353642671292281, "grad_norm": 0.39523038268089294, "learning_rate": 8.85318438996083e-05, "loss": 0.035, "step": 3262 }, { "epoch": 0.3537510841283608, "grad_norm": 0.3448050916194916, "learning_rate": 8.852821703177137e-05, "loss": 0.089, "step": 3263 }, { "epoch": 0.3538594969644406, "grad_norm": 0.3874160945415497, "learning_rate": 8.852459016393443e-05, "loss": 0.0651, "step": 3264 }, { "epoch": 0.3539679098005204, "grad_norm": 0.2765496075153351, "learning_rate": 8.85209632960975e-05, "loss": 0.0426, "step": 3265 }, { "epoch": 0.35407632263660016, "grad_norm": 0.25446128845214844, "learning_rate": 8.851733642826055e-05, "loss": 0.0346, "step": 3266 }, { "epoch": 0.35418473547268, "grad_norm": 0.6268018484115601, "learning_rate": 8.851370956042361e-05, "loss": 0.0829, "step": 3267 }, { "epoch": 0.35429314830875974, "grad_norm": 0.4529252052307129, "learning_rate": 8.851008269258668e-05, "loss": 0.0269, "step": 3268 }, { "epoch": 0.35440156114483956, "grad_norm": 0.30954527854919434, "learning_rate": 8.850645582474974e-05, "loss": 0.0371, "step": 3269 }, { "epoch": 0.3545099739809193, "grad_norm": 0.23762689530849457, "learning_rate": 8.850282895691281e-05, "loss": 0.024, "step": 3270 }, { "epoch": 0.35461838681699914, "grad_norm": 0.5489417910575867, "learning_rate": 8.849920208907588e-05, "loss": 0.0904, "step": 3271 }, { "epoch": 0.3547267996530789, "grad_norm": 0.43616461753845215, "learning_rate": 8.849557522123895e-05, "loss": 0.0455, "step": 3272 }, { "epoch": 0.3548352124891587, "grad_norm": 0.42503735423088074, "learning_rate": 8.849194835340201e-05, "loss": 0.0548, "step": 3273 }, { "epoch": 0.3549436253252385, "grad_norm": 0.4197847843170166, "learning_rate": 8.848832148556507e-05, "loss": 0.041, "step": 3274 }, { "epoch": 0.3550520381613183, "grad_norm": 0.32528162002563477, "learning_rate": 8.848469461772814e-05, "loss": 0.0673, "step": 3275 }, { "epoch": 0.35516045099739807, "grad_norm": 0.4280926585197449, "learning_rate": 8.84810677498912e-05, "loss": 0.0446, "step": 3276 }, { "epoch": 0.3552688638334779, "grad_norm": 0.2581334412097931, "learning_rate": 8.847744088205427e-05, "loss": 0.0732, "step": 3277 }, { "epoch": 0.35537727666955765, "grad_norm": 0.40504997968673706, "learning_rate": 8.847381401421732e-05, "loss": 0.0565, "step": 3278 }, { "epoch": 0.35548568950563747, "grad_norm": 0.3144279718399048, "learning_rate": 8.84701871463804e-05, "loss": 0.0529, "step": 3279 }, { "epoch": 0.35559410234171723, "grad_norm": 0.3662194609642029, "learning_rate": 8.846656027854345e-05, "loss": 0.0298, "step": 3280 }, { "epoch": 0.35570251517779705, "grad_norm": 0.2989228069782257, "learning_rate": 8.846293341070652e-05, "loss": 0.0342, "step": 3281 }, { "epoch": 0.35581092801387687, "grad_norm": 0.5452847480773926, "learning_rate": 8.845930654286958e-05, "loss": 0.0644, "step": 3282 }, { "epoch": 0.35591934084995663, "grad_norm": 0.4145817756652832, "learning_rate": 8.845567967503264e-05, "loss": 0.0676, "step": 3283 }, { "epoch": 0.35602775368603645, "grad_norm": 0.30728814005851746, "learning_rate": 8.845205280719571e-05, "loss": 0.0279, "step": 3284 }, { "epoch": 0.3561361665221162, "grad_norm": 0.27683544158935547, "learning_rate": 8.844842593935877e-05, "loss": 0.0478, "step": 3285 }, { "epoch": 0.35624457935819603, "grad_norm": 0.4037250876426697, "learning_rate": 8.844479907152184e-05, "loss": 0.0354, "step": 3286 }, { "epoch": 0.3563529921942758, "grad_norm": 0.5673474669456482, "learning_rate": 8.84411722036849e-05, "loss": 0.0461, "step": 3287 }, { "epoch": 0.3564614050303556, "grad_norm": 0.20740197598934174, "learning_rate": 8.843754533584797e-05, "loss": 0.0336, "step": 3288 }, { "epoch": 0.3565698178664354, "grad_norm": 0.27322161197662354, "learning_rate": 8.843391846801104e-05, "loss": 0.0499, "step": 3289 }, { "epoch": 0.3566782307025152, "grad_norm": 0.5050411820411682, "learning_rate": 8.84302916001741e-05, "loss": 0.0652, "step": 3290 }, { "epoch": 0.35678664353859496, "grad_norm": 0.8957710266113281, "learning_rate": 8.842666473233716e-05, "loss": 0.0697, "step": 3291 }, { "epoch": 0.3568950563746748, "grad_norm": 0.23455868661403656, "learning_rate": 8.842303786450022e-05, "loss": 0.0479, "step": 3292 }, { "epoch": 0.35700346921075454, "grad_norm": 0.4115082621574402, "learning_rate": 8.841941099666329e-05, "loss": 0.0705, "step": 3293 }, { "epoch": 0.35711188204683436, "grad_norm": 0.4914371967315674, "learning_rate": 8.841578412882635e-05, "loss": 0.0349, "step": 3294 }, { "epoch": 0.3572202948829141, "grad_norm": 0.36337536573410034, "learning_rate": 8.841215726098942e-05, "loss": 0.0509, "step": 3295 }, { "epoch": 0.35732870771899394, "grad_norm": 0.26286622881889343, "learning_rate": 8.840853039315248e-05, "loss": 0.0413, "step": 3296 }, { "epoch": 0.3574371205550737, "grad_norm": 0.39633870124816895, "learning_rate": 8.840490352531554e-05, "loss": 0.0462, "step": 3297 }, { "epoch": 0.3575455333911535, "grad_norm": 0.5710629820823669, "learning_rate": 8.84012766574786e-05, "loss": 0.0548, "step": 3298 }, { "epoch": 0.3576539462272333, "grad_norm": 0.24062563478946686, "learning_rate": 8.839764978964166e-05, "loss": 0.0263, "step": 3299 }, { "epoch": 0.3577623590633131, "grad_norm": 0.4812389016151428, "learning_rate": 8.839402292180473e-05, "loss": 0.042, "step": 3300 }, { "epoch": 0.35787077189939287, "grad_norm": 0.25963738560676575, "learning_rate": 8.839039605396779e-05, "loss": 0.037, "step": 3301 }, { "epoch": 0.3579791847354727, "grad_norm": 0.30986595153808594, "learning_rate": 8.838676918613086e-05, "loss": 0.0474, "step": 3302 }, { "epoch": 0.35808759757155245, "grad_norm": 0.6311991810798645, "learning_rate": 8.838314231829392e-05, "loss": 0.1036, "step": 3303 }, { "epoch": 0.35819601040763227, "grad_norm": 0.42560696601867676, "learning_rate": 8.837951545045699e-05, "loss": 0.0579, "step": 3304 }, { "epoch": 0.35830442324371203, "grad_norm": 0.5203664898872375, "learning_rate": 8.837588858262006e-05, "loss": 0.0344, "step": 3305 }, { "epoch": 0.35841283607979185, "grad_norm": 0.2990446984767914, "learning_rate": 8.837226171478312e-05, "loss": 0.0269, "step": 3306 }, { "epoch": 0.3585212489158716, "grad_norm": 0.5474587678909302, "learning_rate": 8.836863484694619e-05, "loss": 0.0578, "step": 3307 }, { "epoch": 0.35862966175195143, "grad_norm": 0.6988529562950134, "learning_rate": 8.836500797910925e-05, "loss": 0.0997, "step": 3308 }, { "epoch": 0.3587380745880312, "grad_norm": 0.12736253440380096, "learning_rate": 8.836138111127232e-05, "loss": 0.0136, "step": 3309 }, { "epoch": 0.358846487424111, "grad_norm": 0.1827748417854309, "learning_rate": 8.835775424343538e-05, "loss": 0.0223, "step": 3310 }, { "epoch": 0.3589549002601908, "grad_norm": 0.3114027678966522, "learning_rate": 8.835412737559843e-05, "loss": 0.0426, "step": 3311 }, { "epoch": 0.3590633130962706, "grad_norm": 0.20708151161670685, "learning_rate": 8.83505005077615e-05, "loss": 0.0243, "step": 3312 }, { "epoch": 0.3591717259323504, "grad_norm": 0.3721102476119995, "learning_rate": 8.834687363992456e-05, "loss": 0.0504, "step": 3313 }, { "epoch": 0.3592801387684302, "grad_norm": 0.20683972537517548, "learning_rate": 8.834324677208763e-05, "loss": 0.0257, "step": 3314 }, { "epoch": 0.35938855160451, "grad_norm": 0.4602315425872803, "learning_rate": 8.833961990425069e-05, "loss": 0.047, "step": 3315 }, { "epoch": 0.35949696444058976, "grad_norm": 0.22286449372768402, "learning_rate": 8.833599303641376e-05, "loss": 0.0296, "step": 3316 }, { "epoch": 0.3596053772766696, "grad_norm": 0.37060123682022095, "learning_rate": 8.833236616857682e-05, "loss": 0.0391, "step": 3317 }, { "epoch": 0.35971379011274934, "grad_norm": 0.43402570486068726, "learning_rate": 8.832873930073989e-05, "loss": 0.0436, "step": 3318 }, { "epoch": 0.35982220294882916, "grad_norm": 0.15368635952472687, "learning_rate": 8.832511243290295e-05, "loss": 0.0129, "step": 3319 }, { "epoch": 0.3599306157849089, "grad_norm": 0.5330312848091125, "learning_rate": 8.8321485565066e-05, "loss": 0.0487, "step": 3320 }, { "epoch": 0.36003902862098874, "grad_norm": 0.21739305555820465, "learning_rate": 8.831785869722907e-05, "loss": 0.0205, "step": 3321 }, { "epoch": 0.3601474414570685, "grad_norm": 0.18026858568191528, "learning_rate": 8.831423182939214e-05, "loss": 0.0226, "step": 3322 }, { "epoch": 0.3602558542931483, "grad_norm": 0.26258933544158936, "learning_rate": 8.831060496155522e-05, "loss": 0.0227, "step": 3323 }, { "epoch": 0.3603642671292281, "grad_norm": 0.5437384843826294, "learning_rate": 8.830697809371827e-05, "loss": 0.0562, "step": 3324 }, { "epoch": 0.3604726799653079, "grad_norm": 0.6714861989021301, "learning_rate": 8.830335122588133e-05, "loss": 0.0231, "step": 3325 }, { "epoch": 0.36058109280138767, "grad_norm": 0.4722808599472046, "learning_rate": 8.82997243580444e-05, "loss": 0.0306, "step": 3326 }, { "epoch": 0.3606895056374675, "grad_norm": 1.7132784128189087, "learning_rate": 8.829609749020746e-05, "loss": 0.0327, "step": 3327 }, { "epoch": 0.36079791847354725, "grad_norm": 0.5771552324295044, "learning_rate": 8.829247062237053e-05, "loss": 0.0496, "step": 3328 }, { "epoch": 0.36090633130962707, "grad_norm": 0.9006063342094421, "learning_rate": 8.828884375453359e-05, "loss": 0.0546, "step": 3329 }, { "epoch": 0.36101474414570683, "grad_norm": 0.47488364577293396, "learning_rate": 8.828521688669666e-05, "loss": 0.0429, "step": 3330 }, { "epoch": 0.36112315698178665, "grad_norm": 0.34938934445381165, "learning_rate": 8.828159001885971e-05, "loss": 0.0266, "step": 3331 }, { "epoch": 0.3612315698178664, "grad_norm": 0.23201701045036316, "learning_rate": 8.827796315102279e-05, "loss": 0.0128, "step": 3332 }, { "epoch": 0.36133998265394623, "grad_norm": 0.44798770546913147, "learning_rate": 8.827433628318584e-05, "loss": 0.0542, "step": 3333 }, { "epoch": 0.361448395490026, "grad_norm": 0.33516475558280945, "learning_rate": 8.82707094153489e-05, "loss": 0.0426, "step": 3334 }, { "epoch": 0.3615568083261058, "grad_norm": 0.07530798017978668, "learning_rate": 8.826708254751197e-05, "loss": 0.0083, "step": 3335 }, { "epoch": 0.3616652211621856, "grad_norm": 0.29797661304473877, "learning_rate": 8.826345567967503e-05, "loss": 0.0296, "step": 3336 }, { "epoch": 0.3617736339982654, "grad_norm": 0.3615247905254364, "learning_rate": 8.82598288118381e-05, "loss": 0.037, "step": 3337 }, { "epoch": 0.36188204683434516, "grad_norm": 0.8609235286712646, "learning_rate": 8.825620194400116e-05, "loss": 0.1017, "step": 3338 }, { "epoch": 0.361990459670425, "grad_norm": 0.599170446395874, "learning_rate": 8.825257507616423e-05, "loss": 0.0436, "step": 3339 }, { "epoch": 0.36209887250650474, "grad_norm": 0.8974870443344116, "learning_rate": 8.82489482083273e-05, "loss": 0.0781, "step": 3340 }, { "epoch": 0.36220728534258456, "grad_norm": 0.20424959063529968, "learning_rate": 8.824532134049036e-05, "loss": 0.0266, "step": 3341 }, { "epoch": 0.3623156981786644, "grad_norm": 0.8320082426071167, "learning_rate": 8.824169447265343e-05, "loss": 0.0362, "step": 3342 }, { "epoch": 0.36242411101474414, "grad_norm": 0.20451846718788147, "learning_rate": 8.823806760481648e-05, "loss": 0.0319, "step": 3343 }, { "epoch": 0.36253252385082396, "grad_norm": 0.4854263663291931, "learning_rate": 8.823444073697956e-05, "loss": 0.0156, "step": 3344 }, { "epoch": 0.3626409366869037, "grad_norm": 0.4580068290233612, "learning_rate": 8.823081386914261e-05, "loss": 0.0679, "step": 3345 }, { "epoch": 0.36274934952298354, "grad_norm": 0.14572244882583618, "learning_rate": 8.822718700130568e-05, "loss": 0.0178, "step": 3346 }, { "epoch": 0.3628577623590633, "grad_norm": 0.8788337707519531, "learning_rate": 8.822356013346874e-05, "loss": 0.0909, "step": 3347 }, { "epoch": 0.3629661751951431, "grad_norm": 1.077822208404541, "learning_rate": 8.82199332656318e-05, "loss": 0.0872, "step": 3348 }, { "epoch": 0.3630745880312229, "grad_norm": 0.3796424865722656, "learning_rate": 8.821630639779487e-05, "loss": 0.0531, "step": 3349 }, { "epoch": 0.3631830008673027, "grad_norm": 0.23098202049732208, "learning_rate": 8.821267952995793e-05, "loss": 0.0133, "step": 3350 }, { "epoch": 0.36329141370338247, "grad_norm": 1.0445702075958252, "learning_rate": 8.8209052662121e-05, "loss": 0.0876, "step": 3351 }, { "epoch": 0.3633998265394623, "grad_norm": 0.3995949327945709, "learning_rate": 8.820542579428405e-05, "loss": 0.0676, "step": 3352 }, { "epoch": 0.36350823937554205, "grad_norm": 0.39190319180488586, "learning_rate": 8.820179892644713e-05, "loss": 0.0454, "step": 3353 }, { "epoch": 0.36361665221162187, "grad_norm": 0.48245736956596375, "learning_rate": 8.819817205861018e-05, "loss": 0.1145, "step": 3354 }, { "epoch": 0.36372506504770163, "grad_norm": 0.5376883149147034, "learning_rate": 8.819454519077325e-05, "loss": 0.0777, "step": 3355 }, { "epoch": 0.36383347788378145, "grad_norm": 0.14333830773830414, "learning_rate": 8.819091832293631e-05, "loss": 0.0226, "step": 3356 }, { "epoch": 0.3639418907198612, "grad_norm": 0.21711499989032745, "learning_rate": 8.818729145509938e-05, "loss": 0.0205, "step": 3357 }, { "epoch": 0.36405030355594103, "grad_norm": 0.3325548768043518, "learning_rate": 8.818366458726245e-05, "loss": 0.0398, "step": 3358 }, { "epoch": 0.3641587163920208, "grad_norm": 0.2972761392593384, "learning_rate": 8.818003771942551e-05, "loss": 0.0346, "step": 3359 }, { "epoch": 0.3642671292281006, "grad_norm": 0.40032750368118286, "learning_rate": 8.817641085158858e-05, "loss": 0.0711, "step": 3360 }, { "epoch": 0.3643755420641804, "grad_norm": 0.15512771904468536, "learning_rate": 8.817278398375164e-05, "loss": 0.0265, "step": 3361 }, { "epoch": 0.3644839549002602, "grad_norm": 0.43025141954421997, "learning_rate": 8.816915711591471e-05, "loss": 0.056, "step": 3362 }, { "epoch": 0.36459236773633996, "grad_norm": 0.595583975315094, "learning_rate": 8.816553024807777e-05, "loss": 0.0455, "step": 3363 }, { "epoch": 0.3647007805724198, "grad_norm": 0.21277719736099243, "learning_rate": 8.816190338024082e-05, "loss": 0.0413, "step": 3364 }, { "epoch": 0.36480919340849954, "grad_norm": 0.3675801157951355, "learning_rate": 8.81582765124039e-05, "loss": 0.0482, "step": 3365 }, { "epoch": 0.36491760624457936, "grad_norm": 0.26170364022254944, "learning_rate": 8.815464964456695e-05, "loss": 0.0461, "step": 3366 }, { "epoch": 0.3650260190806591, "grad_norm": 0.22119587659835815, "learning_rate": 8.815102277673002e-05, "loss": 0.0471, "step": 3367 }, { "epoch": 0.36513443191673894, "grad_norm": 0.34265774488449097, "learning_rate": 8.814739590889308e-05, "loss": 0.0649, "step": 3368 }, { "epoch": 0.3652428447528187, "grad_norm": 0.22298917174339294, "learning_rate": 8.814376904105615e-05, "loss": 0.0342, "step": 3369 }, { "epoch": 0.3653512575888985, "grad_norm": 0.26338279247283936, "learning_rate": 8.814014217321921e-05, "loss": 0.0434, "step": 3370 }, { "epoch": 0.36545967042497834, "grad_norm": 0.19698601961135864, "learning_rate": 8.813651530538227e-05, "loss": 0.0364, "step": 3371 }, { "epoch": 0.3655680832610581, "grad_norm": 0.11825134605169296, "learning_rate": 8.813288843754534e-05, "loss": 0.0249, "step": 3372 }, { "epoch": 0.3656764960971379, "grad_norm": 0.19831906259059906, "learning_rate": 8.81292615697084e-05, "loss": 0.0356, "step": 3373 }, { "epoch": 0.3657849089332177, "grad_norm": 0.5261361598968506, "learning_rate": 8.812563470187148e-05, "loss": 0.044, "step": 3374 }, { "epoch": 0.3658933217692975, "grad_norm": 0.21148163080215454, "learning_rate": 8.812200783403454e-05, "loss": 0.0331, "step": 3375 }, { "epoch": 0.36600173460537727, "grad_norm": 0.2951394319534302, "learning_rate": 8.81183809661976e-05, "loss": 0.0521, "step": 3376 }, { "epoch": 0.3661101474414571, "grad_norm": 0.44791191816329956, "learning_rate": 8.811475409836066e-05, "loss": 0.0551, "step": 3377 }, { "epoch": 0.36621856027753685, "grad_norm": 0.40043145418167114, "learning_rate": 8.811112723052372e-05, "loss": 0.0813, "step": 3378 }, { "epoch": 0.36632697311361667, "grad_norm": 0.5132423043251038, "learning_rate": 8.810750036268679e-05, "loss": 0.0529, "step": 3379 }, { "epoch": 0.36643538594969643, "grad_norm": 0.4561932682991028, "learning_rate": 8.810387349484985e-05, "loss": 0.0547, "step": 3380 }, { "epoch": 0.36654379878577625, "grad_norm": 0.41193515062332153, "learning_rate": 8.810024662701292e-05, "loss": 0.0756, "step": 3381 }, { "epoch": 0.366652211621856, "grad_norm": 0.2905307114124298, "learning_rate": 8.809661975917598e-05, "loss": 0.031, "step": 3382 }, { "epoch": 0.36676062445793584, "grad_norm": 0.3709419071674347, "learning_rate": 8.809299289133905e-05, "loss": 0.0369, "step": 3383 }, { "epoch": 0.3668690372940156, "grad_norm": 0.4421576261520386, "learning_rate": 8.80893660235021e-05, "loss": 0.0515, "step": 3384 }, { "epoch": 0.3669774501300954, "grad_norm": 0.3017193078994751, "learning_rate": 8.808573915566518e-05, "loss": 0.0498, "step": 3385 }, { "epoch": 0.3670858629661752, "grad_norm": 0.3796077370643616, "learning_rate": 8.808211228782823e-05, "loss": 0.0584, "step": 3386 }, { "epoch": 0.367194275802255, "grad_norm": 0.270330548286438, "learning_rate": 8.807848541999129e-05, "loss": 0.0221, "step": 3387 }, { "epoch": 0.36730268863833476, "grad_norm": 0.12585873901844025, "learning_rate": 8.807485855215436e-05, "loss": 0.0228, "step": 3388 }, { "epoch": 0.3674111014744146, "grad_norm": 0.19073902070522308, "learning_rate": 8.807123168431742e-05, "loss": 0.0229, "step": 3389 }, { "epoch": 0.36751951431049434, "grad_norm": 0.48390331864356995, "learning_rate": 8.806760481648049e-05, "loss": 0.037, "step": 3390 }, { "epoch": 0.36762792714657416, "grad_norm": 0.26024505496025085, "learning_rate": 8.806397794864355e-05, "loss": 0.0456, "step": 3391 }, { "epoch": 0.3677363399826539, "grad_norm": 0.35752013325691223, "learning_rate": 8.806035108080662e-05, "loss": 0.0291, "step": 3392 }, { "epoch": 0.36784475281873374, "grad_norm": 0.17915669083595276, "learning_rate": 8.805672421296969e-05, "loss": 0.0202, "step": 3393 }, { "epoch": 0.3679531656548135, "grad_norm": 0.13280248641967773, "learning_rate": 8.805309734513275e-05, "loss": 0.0186, "step": 3394 }, { "epoch": 0.3680615784908933, "grad_norm": 0.2078247368335724, "learning_rate": 8.804947047729582e-05, "loss": 0.0368, "step": 3395 }, { "epoch": 0.3681699913269731, "grad_norm": 0.5527489185333252, "learning_rate": 8.804584360945888e-05, "loss": 0.0453, "step": 3396 }, { "epoch": 0.3682784041630529, "grad_norm": 0.14568690955638885, "learning_rate": 8.804221674162195e-05, "loss": 0.0161, "step": 3397 }, { "epoch": 0.36838681699913267, "grad_norm": 0.16123338043689728, "learning_rate": 8.8038589873785e-05, "loss": 0.0154, "step": 3398 }, { "epoch": 0.3684952298352125, "grad_norm": 0.19130787253379822, "learning_rate": 8.803496300594807e-05, "loss": 0.0217, "step": 3399 }, { "epoch": 0.3686036426712923, "grad_norm": 0.4682934284210205, "learning_rate": 8.803133613811113e-05, "loss": 0.0562, "step": 3400 }, { "epoch": 0.36871205550737207, "grad_norm": 0.2542538046836853, "learning_rate": 8.802770927027419e-05, "loss": 0.0341, "step": 3401 }, { "epoch": 0.3688204683434519, "grad_norm": 0.39456290006637573, "learning_rate": 8.802408240243726e-05, "loss": 0.0441, "step": 3402 }, { "epoch": 0.36892888117953165, "grad_norm": 0.33112356066703796, "learning_rate": 8.802045553460032e-05, "loss": 0.0409, "step": 3403 }, { "epoch": 0.3690372940156115, "grad_norm": 0.40854743123054504, "learning_rate": 8.801682866676339e-05, "loss": 0.0325, "step": 3404 }, { "epoch": 0.36914570685169124, "grad_norm": 0.6484625935554504, "learning_rate": 8.801320179892645e-05, "loss": 0.1072, "step": 3405 }, { "epoch": 0.36925411968777105, "grad_norm": 0.25428763031959534, "learning_rate": 8.800957493108952e-05, "loss": 0.026, "step": 3406 }, { "epoch": 0.3693625325238508, "grad_norm": 0.44736573100090027, "learning_rate": 8.800594806325257e-05, "loss": 0.0419, "step": 3407 }, { "epoch": 0.36947094535993064, "grad_norm": 0.4981745779514313, "learning_rate": 8.800232119541564e-05, "loss": 0.0527, "step": 3408 }, { "epoch": 0.3695793581960104, "grad_norm": 1.1187751293182373, "learning_rate": 8.799869432757872e-05, "loss": 0.0367, "step": 3409 }, { "epoch": 0.3696877710320902, "grad_norm": 0.5628572702407837, "learning_rate": 8.799506745974177e-05, "loss": 0.0463, "step": 3410 }, { "epoch": 0.36979618386817, "grad_norm": 0.5329844355583191, "learning_rate": 8.799144059190484e-05, "loss": 0.0401, "step": 3411 }, { "epoch": 0.3699045967042498, "grad_norm": 0.1575557440519333, "learning_rate": 8.79878137240679e-05, "loss": 0.0155, "step": 3412 }, { "epoch": 0.37001300954032956, "grad_norm": 0.8477303385734558, "learning_rate": 8.798418685623097e-05, "loss": 0.0568, "step": 3413 }, { "epoch": 0.3701214223764094, "grad_norm": 0.25547441840171814, "learning_rate": 8.798055998839403e-05, "loss": 0.0341, "step": 3414 }, { "epoch": 0.37022983521248914, "grad_norm": 0.44939935207366943, "learning_rate": 8.797693312055709e-05, "loss": 0.0438, "step": 3415 }, { "epoch": 0.37033824804856896, "grad_norm": 0.7346107959747314, "learning_rate": 8.797330625272016e-05, "loss": 0.0593, "step": 3416 }, { "epoch": 0.3704466608846487, "grad_norm": 0.4949691891670227, "learning_rate": 8.796967938488321e-05, "loss": 0.0363, "step": 3417 }, { "epoch": 0.37055507372072854, "grad_norm": 0.6340675354003906, "learning_rate": 8.796605251704629e-05, "loss": 0.0253, "step": 3418 }, { "epoch": 0.3706634865568083, "grad_norm": 0.13010984659194946, "learning_rate": 8.796242564920934e-05, "loss": 0.0206, "step": 3419 }, { "epoch": 0.3707718993928881, "grad_norm": 1.051421880722046, "learning_rate": 8.795879878137241e-05, "loss": 0.1097, "step": 3420 }, { "epoch": 0.3708803122289679, "grad_norm": 0.39127856492996216, "learning_rate": 8.795517191353547e-05, "loss": 0.0338, "step": 3421 }, { "epoch": 0.3709887250650477, "grad_norm": 0.21042388677597046, "learning_rate": 8.795154504569854e-05, "loss": 0.0147, "step": 3422 }, { "epoch": 0.37109713790112747, "grad_norm": 0.40156784653663635, "learning_rate": 8.79479181778616e-05, "loss": 0.0381, "step": 3423 }, { "epoch": 0.3712055507372073, "grad_norm": 0.4369764029979706, "learning_rate": 8.794429131002466e-05, "loss": 0.0573, "step": 3424 }, { "epoch": 0.37131396357328705, "grad_norm": 0.10323350876569748, "learning_rate": 8.794066444218773e-05, "loss": 0.0139, "step": 3425 }, { "epoch": 0.3714223764093669, "grad_norm": 0.3671139180660248, "learning_rate": 8.79370375743508e-05, "loss": 0.044, "step": 3426 }, { "epoch": 0.37153078924544664, "grad_norm": 0.5226618647575378, "learning_rate": 8.793341070651387e-05, "loss": 0.0407, "step": 3427 }, { "epoch": 0.37163920208152645, "grad_norm": 0.26241880655288696, "learning_rate": 8.792978383867693e-05, "loss": 0.0306, "step": 3428 }, { "epoch": 0.3717476149176062, "grad_norm": 0.45865434408187866, "learning_rate": 8.792615697083998e-05, "loss": 0.0397, "step": 3429 }, { "epoch": 0.37185602775368604, "grad_norm": 0.5726093053817749, "learning_rate": 8.792253010300305e-05, "loss": 0.0311, "step": 3430 }, { "epoch": 0.37196444058976585, "grad_norm": 0.41816529631614685, "learning_rate": 8.791890323516611e-05, "loss": 0.0471, "step": 3431 }, { "epoch": 0.3720728534258456, "grad_norm": 0.4028029441833496, "learning_rate": 8.791527636732918e-05, "loss": 0.0406, "step": 3432 }, { "epoch": 0.37218126626192544, "grad_norm": 0.27043911814689636, "learning_rate": 8.791164949949224e-05, "loss": 0.0278, "step": 3433 }, { "epoch": 0.3722896790980052, "grad_norm": 0.44074776768684387, "learning_rate": 8.790802263165531e-05, "loss": 0.0552, "step": 3434 }, { "epoch": 0.372398091934085, "grad_norm": 0.2892032265663147, "learning_rate": 8.790439576381837e-05, "loss": 0.0277, "step": 3435 }, { "epoch": 0.3725065047701648, "grad_norm": 0.23375362157821655, "learning_rate": 8.790076889598144e-05, "loss": 0.0271, "step": 3436 }, { "epoch": 0.3726149176062446, "grad_norm": 0.3486381769180298, "learning_rate": 8.78971420281445e-05, "loss": 0.0339, "step": 3437 }, { "epoch": 0.37272333044232436, "grad_norm": 0.30557531118392944, "learning_rate": 8.789351516030755e-05, "loss": 0.0387, "step": 3438 }, { "epoch": 0.3728317432784042, "grad_norm": 0.3690205514431, "learning_rate": 8.788988829247062e-05, "loss": 0.0264, "step": 3439 }, { "epoch": 0.37294015611448394, "grad_norm": 0.12638019025325775, "learning_rate": 8.788626142463368e-05, "loss": 0.0129, "step": 3440 }, { "epoch": 0.37304856895056376, "grad_norm": 0.7192649841308594, "learning_rate": 8.788263455679675e-05, "loss": 0.0603, "step": 3441 }, { "epoch": 0.3731569817866435, "grad_norm": 0.4093472957611084, "learning_rate": 8.787900768895981e-05, "loss": 0.018, "step": 3442 }, { "epoch": 0.37326539462272335, "grad_norm": 0.3177787959575653, "learning_rate": 8.787538082112288e-05, "loss": 0.0311, "step": 3443 }, { "epoch": 0.3733738074588031, "grad_norm": 0.5309745073318481, "learning_rate": 8.787175395328595e-05, "loss": 0.047, "step": 3444 }, { "epoch": 0.3734822202948829, "grad_norm": 0.5009461045265198, "learning_rate": 8.786812708544901e-05, "loss": 0.0461, "step": 3445 }, { "epoch": 0.3735906331309627, "grad_norm": 0.8946084380149841, "learning_rate": 8.786450021761208e-05, "loss": 0.039, "step": 3446 }, { "epoch": 0.3736990459670425, "grad_norm": 0.5871247053146362, "learning_rate": 8.786087334977514e-05, "loss": 0.0553, "step": 3447 }, { "epoch": 0.3738074588031223, "grad_norm": 0.2566714584827423, "learning_rate": 8.785724648193821e-05, "loss": 0.0223, "step": 3448 }, { "epoch": 0.3739158716392021, "grad_norm": 0.5166948437690735, "learning_rate": 8.785361961410127e-05, "loss": 0.0262, "step": 3449 }, { "epoch": 0.37402428447528185, "grad_norm": 0.7509914636611938, "learning_rate": 8.784999274626434e-05, "loss": 0.0885, "step": 3450 }, { "epoch": 0.3741326973113617, "grad_norm": 0.14440687000751495, "learning_rate": 8.78463658784274e-05, "loss": 0.0123, "step": 3451 }, { "epoch": 0.37424111014744144, "grad_norm": 0.6604138612747192, "learning_rate": 8.784273901059045e-05, "loss": 0.0313, "step": 3452 }, { "epoch": 0.37434952298352125, "grad_norm": 0.4168044924736023, "learning_rate": 8.783911214275352e-05, "loss": 0.0674, "step": 3453 }, { "epoch": 0.374457935819601, "grad_norm": 0.38709819316864014, "learning_rate": 8.783548527491658e-05, "loss": 0.0187, "step": 3454 }, { "epoch": 0.37456634865568084, "grad_norm": 0.3842430114746094, "learning_rate": 8.783185840707965e-05, "loss": 0.0236, "step": 3455 }, { "epoch": 0.3746747614917606, "grad_norm": 0.33206766843795776, "learning_rate": 8.782823153924271e-05, "loss": 0.0369, "step": 3456 }, { "epoch": 0.3747831743278404, "grad_norm": 0.29247838258743286, "learning_rate": 8.782460467140578e-05, "loss": 0.0122, "step": 3457 }, { "epoch": 0.3748915871639202, "grad_norm": 0.7154924273490906, "learning_rate": 8.782097780356884e-05, "loss": 0.0567, "step": 3458 }, { "epoch": 0.375, "grad_norm": 0.5755689144134521, "learning_rate": 8.781735093573191e-05, "loss": 0.0558, "step": 3459 }, { "epoch": 0.3751084128360798, "grad_norm": 0.8629852533340454, "learning_rate": 8.781372406789496e-05, "loss": 0.0684, "step": 3460 }, { "epoch": 0.3752168256721596, "grad_norm": 2.165003538131714, "learning_rate": 8.781009720005804e-05, "loss": 0.0303, "step": 3461 }, { "epoch": 0.3753252385082394, "grad_norm": 0.35877883434295654, "learning_rate": 8.78064703322211e-05, "loss": 0.0405, "step": 3462 }, { "epoch": 0.37543365134431916, "grad_norm": 0.577888548374176, "learning_rate": 8.780284346438416e-05, "loss": 0.0542, "step": 3463 }, { "epoch": 0.375542064180399, "grad_norm": 0.5693604350090027, "learning_rate": 8.779921659654723e-05, "loss": 0.0393, "step": 3464 }, { "epoch": 0.37565047701647875, "grad_norm": 0.46590614318847656, "learning_rate": 8.779558972871029e-05, "loss": 0.0323, "step": 3465 }, { "epoch": 0.37575888985255856, "grad_norm": 0.5969956517219543, "learning_rate": 8.779196286087336e-05, "loss": 0.049, "step": 3466 }, { "epoch": 0.3758673026886383, "grad_norm": 0.5812209248542786, "learning_rate": 8.778833599303642e-05, "loss": 0.079, "step": 3467 }, { "epoch": 0.37597571552471815, "grad_norm": 0.3189094662666321, "learning_rate": 8.778470912519948e-05, "loss": 0.0429, "step": 3468 }, { "epoch": 0.3760841283607979, "grad_norm": 0.3856557309627533, "learning_rate": 8.778108225736255e-05, "loss": 0.0424, "step": 3469 }, { "epoch": 0.3761925411968777, "grad_norm": 0.16388145089149475, "learning_rate": 8.77774553895256e-05, "loss": 0.0132, "step": 3470 }, { "epoch": 0.3763009540329575, "grad_norm": 0.16210122406482697, "learning_rate": 8.777382852168868e-05, "loss": 0.0098, "step": 3471 }, { "epoch": 0.3764093668690373, "grad_norm": 0.2907916307449341, "learning_rate": 8.777020165385173e-05, "loss": 0.0542, "step": 3472 }, { "epoch": 0.3765177797051171, "grad_norm": 0.49634483456611633, "learning_rate": 8.77665747860148e-05, "loss": 0.0233, "step": 3473 }, { "epoch": 0.3766261925411969, "grad_norm": 0.4400888979434967, "learning_rate": 8.776294791817786e-05, "loss": 0.0757, "step": 3474 }, { "epoch": 0.37673460537727665, "grad_norm": 0.3104017376899719, "learning_rate": 8.775932105034093e-05, "loss": 0.0377, "step": 3475 }, { "epoch": 0.3768430182133565, "grad_norm": 0.16561245918273926, "learning_rate": 8.775569418250399e-05, "loss": 0.0227, "step": 3476 }, { "epoch": 0.37695143104943624, "grad_norm": 0.5870659351348877, "learning_rate": 8.775206731466705e-05, "loss": 0.1016, "step": 3477 }, { "epoch": 0.37705984388551606, "grad_norm": 0.2538094222545624, "learning_rate": 8.774844044683013e-05, "loss": 0.0261, "step": 3478 }, { "epoch": 0.3771682567215958, "grad_norm": 0.23335614800453186, "learning_rate": 8.774481357899319e-05, "loss": 0.0314, "step": 3479 }, { "epoch": 0.37727666955767564, "grad_norm": 0.08263330161571503, "learning_rate": 8.774118671115626e-05, "loss": 0.0083, "step": 3480 }, { "epoch": 0.3773850823937554, "grad_norm": 0.31430724263191223, "learning_rate": 8.773755984331932e-05, "loss": 0.0506, "step": 3481 }, { "epoch": 0.3774934952298352, "grad_norm": 0.3211683928966522, "learning_rate": 8.773393297548237e-05, "loss": 0.0463, "step": 3482 }, { "epoch": 0.377601908065915, "grad_norm": 0.5720111727714539, "learning_rate": 8.773030610764545e-05, "loss": 0.0435, "step": 3483 }, { "epoch": 0.3777103209019948, "grad_norm": 0.5106040835380554, "learning_rate": 8.77266792398085e-05, "loss": 0.0702, "step": 3484 }, { "epoch": 0.37781873373807456, "grad_norm": 0.15713734924793243, "learning_rate": 8.772305237197157e-05, "loss": 0.0146, "step": 3485 }, { "epoch": 0.3779271465741544, "grad_norm": 0.5167803168296814, "learning_rate": 8.771942550413463e-05, "loss": 0.0578, "step": 3486 }, { "epoch": 0.37803555941023415, "grad_norm": 0.30776166915893555, "learning_rate": 8.77157986362977e-05, "loss": 0.047, "step": 3487 }, { "epoch": 0.37814397224631396, "grad_norm": 0.30737730860710144, "learning_rate": 8.771217176846076e-05, "loss": 0.0357, "step": 3488 }, { "epoch": 0.3782523850823938, "grad_norm": 0.1980017125606537, "learning_rate": 8.770854490062383e-05, "loss": 0.0209, "step": 3489 }, { "epoch": 0.37836079791847355, "grad_norm": 0.7038468718528748, "learning_rate": 8.770491803278689e-05, "loss": 0.0597, "step": 3490 }, { "epoch": 0.37846921075455336, "grad_norm": 0.27761396765708923, "learning_rate": 8.770129116494994e-05, "loss": 0.022, "step": 3491 }, { "epoch": 0.3785776235906331, "grad_norm": 0.1483113169670105, "learning_rate": 8.769766429711302e-05, "loss": 0.0247, "step": 3492 }, { "epoch": 0.37868603642671295, "grad_norm": 0.1969946175813675, "learning_rate": 8.769403742927607e-05, "loss": 0.0143, "step": 3493 }, { "epoch": 0.3787944492627927, "grad_norm": 0.5126808881759644, "learning_rate": 8.769041056143914e-05, "loss": 0.0972, "step": 3494 }, { "epoch": 0.37890286209887253, "grad_norm": 0.3716103732585907, "learning_rate": 8.768678369360221e-05, "loss": 0.051, "step": 3495 }, { "epoch": 0.3790112749349523, "grad_norm": 0.5909647345542908, "learning_rate": 8.768315682576527e-05, "loss": 0.0815, "step": 3496 }, { "epoch": 0.3791196877710321, "grad_norm": 0.45164087414741516, "learning_rate": 8.767952995792834e-05, "loss": 0.0653, "step": 3497 }, { "epoch": 0.3792281006071119, "grad_norm": 0.21037331223487854, "learning_rate": 8.76759030900914e-05, "loss": 0.0181, "step": 3498 }, { "epoch": 0.3793365134431917, "grad_norm": 0.37315165996551514, "learning_rate": 8.767227622225447e-05, "loss": 0.0263, "step": 3499 }, { "epoch": 0.37944492627927146, "grad_norm": 0.32697954773902893, "learning_rate": 8.766864935441753e-05, "loss": 0.0428, "step": 3500 }, { "epoch": 0.3795533391153513, "grad_norm": 0.1481177806854248, "learning_rate": 8.76650224865806e-05, "loss": 0.0149, "step": 3501 }, { "epoch": 0.37966175195143104, "grad_norm": 0.3555806577205658, "learning_rate": 8.766139561874366e-05, "loss": 0.0353, "step": 3502 }, { "epoch": 0.37977016478751086, "grad_norm": 0.5114687085151672, "learning_rate": 8.765776875090673e-05, "loss": 0.0413, "step": 3503 }, { "epoch": 0.3798785776235906, "grad_norm": 0.3546871840953827, "learning_rate": 8.765414188306978e-05, "loss": 0.0514, "step": 3504 }, { "epoch": 0.37998699045967044, "grad_norm": 0.34473976492881775, "learning_rate": 8.765051501523284e-05, "loss": 0.0465, "step": 3505 }, { "epoch": 0.3800954032957502, "grad_norm": 0.6813743710517883, "learning_rate": 8.764688814739591e-05, "loss": 0.035, "step": 3506 }, { "epoch": 0.38020381613183, "grad_norm": 0.3356536328792572, "learning_rate": 8.764326127955897e-05, "loss": 0.033, "step": 3507 }, { "epoch": 0.3803122289679098, "grad_norm": 0.3235568106174469, "learning_rate": 8.763963441172204e-05, "loss": 0.0475, "step": 3508 }, { "epoch": 0.3804206418039896, "grad_norm": 0.2823447287082672, "learning_rate": 8.76360075438851e-05, "loss": 0.0387, "step": 3509 }, { "epoch": 0.38052905464006936, "grad_norm": 0.2426774650812149, "learning_rate": 8.763238067604817e-05, "loss": 0.0286, "step": 3510 }, { "epoch": 0.3806374674761492, "grad_norm": 0.3617803752422333, "learning_rate": 8.762875380821123e-05, "loss": 0.0651, "step": 3511 }, { "epoch": 0.38074588031222895, "grad_norm": 0.09061134606599808, "learning_rate": 8.76251269403743e-05, "loss": 0.0185, "step": 3512 }, { "epoch": 0.38085429314830876, "grad_norm": 0.43153026700019836, "learning_rate": 8.762150007253737e-05, "loss": 0.0476, "step": 3513 }, { "epoch": 0.3809627059843885, "grad_norm": 0.7176603674888611, "learning_rate": 8.761787320470043e-05, "loss": 0.064, "step": 3514 }, { "epoch": 0.38107111882046835, "grad_norm": 0.42959073185920715, "learning_rate": 8.76142463368635e-05, "loss": 0.03, "step": 3515 }, { "epoch": 0.3811795316565481, "grad_norm": 0.33884942531585693, "learning_rate": 8.761061946902655e-05, "loss": 0.0299, "step": 3516 }, { "epoch": 0.38128794449262793, "grad_norm": 0.48492854833602905, "learning_rate": 8.760699260118962e-05, "loss": 0.0454, "step": 3517 }, { "epoch": 0.3813963573287077, "grad_norm": 0.5373601317405701, "learning_rate": 8.760336573335268e-05, "loss": 0.063, "step": 3518 }, { "epoch": 0.3815047701647875, "grad_norm": 0.33924561738967896, "learning_rate": 8.759973886551574e-05, "loss": 0.0528, "step": 3519 }, { "epoch": 0.38161318300086733, "grad_norm": 0.19377940893173218, "learning_rate": 8.759611199767881e-05, "loss": 0.0126, "step": 3520 }, { "epoch": 0.3817215958369471, "grad_norm": 0.23808996379375458, "learning_rate": 8.759248512984187e-05, "loss": 0.0127, "step": 3521 }, { "epoch": 0.3818300086730269, "grad_norm": 0.762806236743927, "learning_rate": 8.758885826200494e-05, "loss": 0.1037, "step": 3522 }, { "epoch": 0.3819384215091067, "grad_norm": 0.4584197998046875, "learning_rate": 8.7585231394168e-05, "loss": 0.0503, "step": 3523 }, { "epoch": 0.3820468343451865, "grad_norm": 0.2209583967924118, "learning_rate": 8.758160452633107e-05, "loss": 0.0224, "step": 3524 }, { "epoch": 0.38215524718126626, "grad_norm": 0.364596962928772, "learning_rate": 8.757797765849412e-05, "loss": 0.0628, "step": 3525 }, { "epoch": 0.3822636600173461, "grad_norm": 0.5840753316879272, "learning_rate": 8.75743507906572e-05, "loss": 0.0465, "step": 3526 }, { "epoch": 0.38237207285342584, "grad_norm": 0.9553930759429932, "learning_rate": 8.757072392282025e-05, "loss": 0.0866, "step": 3527 }, { "epoch": 0.38248048568950566, "grad_norm": 0.5427677035331726, "learning_rate": 8.756709705498331e-05, "loss": 0.0468, "step": 3528 }, { "epoch": 0.3825888985255854, "grad_norm": 0.2513088285923004, "learning_rate": 8.756347018714638e-05, "loss": 0.0234, "step": 3529 }, { "epoch": 0.38269731136166524, "grad_norm": 0.3569019138813019, "learning_rate": 8.755984331930945e-05, "loss": 0.037, "step": 3530 }, { "epoch": 0.382805724197745, "grad_norm": 0.32542192935943604, "learning_rate": 8.755621645147252e-05, "loss": 0.058, "step": 3531 }, { "epoch": 0.3829141370338248, "grad_norm": 0.6078004837036133, "learning_rate": 8.755258958363558e-05, "loss": 0.0514, "step": 3532 }, { "epoch": 0.3830225498699046, "grad_norm": 0.3301517367362976, "learning_rate": 8.754896271579864e-05, "loss": 0.0525, "step": 3533 }, { "epoch": 0.3831309627059844, "grad_norm": 0.3413067162036896, "learning_rate": 8.754533584796171e-05, "loss": 0.0574, "step": 3534 }, { "epoch": 0.38323937554206416, "grad_norm": 0.3778626024723053, "learning_rate": 8.754170898012477e-05, "loss": 0.0363, "step": 3535 }, { "epoch": 0.383347788378144, "grad_norm": 0.7210346460342407, "learning_rate": 8.753808211228784e-05, "loss": 0.1236, "step": 3536 }, { "epoch": 0.38345620121422375, "grad_norm": 0.33413729071617126, "learning_rate": 8.75344552444509e-05, "loss": 0.0403, "step": 3537 }, { "epoch": 0.38356461405030357, "grad_norm": 0.4541076123714447, "learning_rate": 8.753082837661396e-05, "loss": 0.0707, "step": 3538 }, { "epoch": 0.38367302688638333, "grad_norm": 0.4105997085571289, "learning_rate": 8.752720150877702e-05, "loss": 0.0663, "step": 3539 }, { "epoch": 0.38378143972246315, "grad_norm": 0.42348241806030273, "learning_rate": 8.752357464094009e-05, "loss": 0.0334, "step": 3540 }, { "epoch": 0.3838898525585429, "grad_norm": 0.3075544238090515, "learning_rate": 8.751994777310315e-05, "loss": 0.0427, "step": 3541 }, { "epoch": 0.38399826539462273, "grad_norm": 0.7029467225074768, "learning_rate": 8.751632090526621e-05, "loss": 0.0626, "step": 3542 }, { "epoch": 0.3841066782307025, "grad_norm": 0.4646613895893097, "learning_rate": 8.751269403742928e-05, "loss": 0.0541, "step": 3543 }, { "epoch": 0.3842150910667823, "grad_norm": 0.19607238471508026, "learning_rate": 8.750906716959234e-05, "loss": 0.0259, "step": 3544 }, { "epoch": 0.3843235039028621, "grad_norm": 0.2566176950931549, "learning_rate": 8.75054403017554e-05, "loss": 0.0325, "step": 3545 }, { "epoch": 0.3844319167389419, "grad_norm": 0.37522590160369873, "learning_rate": 8.750181343391846e-05, "loss": 0.035, "step": 3546 }, { "epoch": 0.38454032957502166, "grad_norm": 0.4406731128692627, "learning_rate": 8.749818656608155e-05, "loss": 0.0646, "step": 3547 }, { "epoch": 0.3846487424111015, "grad_norm": 0.2510683536529541, "learning_rate": 8.74945596982446e-05, "loss": 0.0345, "step": 3548 }, { "epoch": 0.3847571552471813, "grad_norm": 0.3627602458000183, "learning_rate": 8.749093283040766e-05, "loss": 0.0291, "step": 3549 }, { "epoch": 0.38486556808326106, "grad_norm": 0.22443994879722595, "learning_rate": 8.748730596257073e-05, "loss": 0.0397, "step": 3550 }, { "epoch": 0.3849739809193409, "grad_norm": 0.3784758746623993, "learning_rate": 8.748367909473379e-05, "loss": 0.0405, "step": 3551 }, { "epoch": 0.38508239375542064, "grad_norm": 0.6271148324012756, "learning_rate": 8.748005222689686e-05, "loss": 0.0648, "step": 3552 }, { "epoch": 0.38519080659150046, "grad_norm": 0.4583018124103546, "learning_rate": 8.747642535905992e-05, "loss": 0.0438, "step": 3553 }, { "epoch": 0.3852992194275802, "grad_norm": 0.972672700881958, "learning_rate": 8.747279849122299e-05, "loss": 0.0776, "step": 3554 }, { "epoch": 0.38540763226366004, "grad_norm": 1.1157408952713013, "learning_rate": 8.746917162338605e-05, "loss": 0.0567, "step": 3555 }, { "epoch": 0.3855160450997398, "grad_norm": 0.27885425090789795, "learning_rate": 8.74655447555491e-05, "loss": 0.0264, "step": 3556 }, { "epoch": 0.3856244579358196, "grad_norm": 0.2493070811033249, "learning_rate": 8.746191788771218e-05, "loss": 0.0431, "step": 3557 }, { "epoch": 0.3857328707718994, "grad_norm": 0.37009263038635254, "learning_rate": 8.745829101987523e-05, "loss": 0.0615, "step": 3558 }, { "epoch": 0.3858412836079792, "grad_norm": 0.5878933072090149, "learning_rate": 8.74546641520383e-05, "loss": 0.0576, "step": 3559 }, { "epoch": 0.38594969644405897, "grad_norm": 0.33282336592674255, "learning_rate": 8.745103728420136e-05, "loss": 0.0326, "step": 3560 }, { "epoch": 0.3860581092801388, "grad_norm": 0.6967782378196716, "learning_rate": 8.744741041636443e-05, "loss": 0.0468, "step": 3561 }, { "epoch": 0.38616652211621855, "grad_norm": 0.1759491264820099, "learning_rate": 8.744378354852749e-05, "loss": 0.0194, "step": 3562 }, { "epoch": 0.38627493495229837, "grad_norm": 0.30071547627449036, "learning_rate": 8.744015668069056e-05, "loss": 0.0661, "step": 3563 }, { "epoch": 0.38638334778837813, "grad_norm": 0.1915329247713089, "learning_rate": 8.743652981285363e-05, "loss": 0.0164, "step": 3564 }, { "epoch": 0.38649176062445795, "grad_norm": 0.23062454164028168, "learning_rate": 8.743290294501669e-05, "loss": 0.0209, "step": 3565 }, { "epoch": 0.3866001734605377, "grad_norm": 0.4795418083667755, "learning_rate": 8.742927607717976e-05, "loss": 0.0689, "step": 3566 }, { "epoch": 0.38670858629661753, "grad_norm": 0.23584285378456116, "learning_rate": 8.742564920934282e-05, "loss": 0.0274, "step": 3567 }, { "epoch": 0.3868169991326973, "grad_norm": 0.2678203582763672, "learning_rate": 8.742202234150589e-05, "loss": 0.0395, "step": 3568 }, { "epoch": 0.3869254119687771, "grad_norm": 1.1460691690444946, "learning_rate": 8.741839547366894e-05, "loss": 0.0383, "step": 3569 }, { "epoch": 0.3870338248048569, "grad_norm": 0.36496493220329285, "learning_rate": 8.741476860583202e-05, "loss": 0.0473, "step": 3570 }, { "epoch": 0.3871422376409367, "grad_norm": 0.5111027359962463, "learning_rate": 8.741114173799507e-05, "loss": 0.0577, "step": 3571 }, { "epoch": 0.38725065047701646, "grad_norm": 0.21474628150463104, "learning_rate": 8.740751487015813e-05, "loss": 0.0295, "step": 3572 }, { "epoch": 0.3873590633130963, "grad_norm": 0.13293255865573883, "learning_rate": 8.74038880023212e-05, "loss": 0.0087, "step": 3573 }, { "epoch": 0.38746747614917604, "grad_norm": 0.36900824308395386, "learning_rate": 8.740026113448426e-05, "loss": 0.0441, "step": 3574 }, { "epoch": 0.38757588898525586, "grad_norm": 0.4699130058288574, "learning_rate": 8.739663426664733e-05, "loss": 0.0403, "step": 3575 }, { "epoch": 0.3876843018213356, "grad_norm": 0.6507343053817749, "learning_rate": 8.739300739881039e-05, "loss": 0.0506, "step": 3576 }, { "epoch": 0.38779271465741544, "grad_norm": 0.7966361045837402, "learning_rate": 8.738938053097346e-05, "loss": 0.041, "step": 3577 }, { "epoch": 0.38790112749349526, "grad_norm": 0.6325270533561707, "learning_rate": 8.738575366313652e-05, "loss": 0.0654, "step": 3578 }, { "epoch": 0.388009540329575, "grad_norm": 0.2828538715839386, "learning_rate": 8.738212679529959e-05, "loss": 0.0131, "step": 3579 }, { "epoch": 0.38811795316565484, "grad_norm": 0.6680465340614319, "learning_rate": 8.737849992746264e-05, "loss": 0.0393, "step": 3580 }, { "epoch": 0.3882263660017346, "grad_norm": 0.6446187496185303, "learning_rate": 8.73748730596257e-05, "loss": 0.0838, "step": 3581 }, { "epoch": 0.3883347788378144, "grad_norm": 0.4499285817146301, "learning_rate": 8.737124619178879e-05, "loss": 0.0425, "step": 3582 }, { "epoch": 0.3884431916738942, "grad_norm": 0.2319527417421341, "learning_rate": 8.736761932395184e-05, "loss": 0.0278, "step": 3583 }, { "epoch": 0.388551604509974, "grad_norm": 0.40698885917663574, "learning_rate": 8.736399245611491e-05, "loss": 0.0241, "step": 3584 }, { "epoch": 0.38866001734605377, "grad_norm": 0.5144635438919067, "learning_rate": 8.736036558827797e-05, "loss": 0.0495, "step": 3585 }, { "epoch": 0.3887684301821336, "grad_norm": 0.49149322509765625, "learning_rate": 8.735673872044103e-05, "loss": 0.0834, "step": 3586 }, { "epoch": 0.38887684301821335, "grad_norm": 0.7088814377784729, "learning_rate": 8.73531118526041e-05, "loss": 0.0594, "step": 3587 }, { "epoch": 0.38898525585429317, "grad_norm": 0.4275140166282654, "learning_rate": 8.734948498476716e-05, "loss": 0.0448, "step": 3588 }, { "epoch": 0.38909366869037293, "grad_norm": 0.23952078819274902, "learning_rate": 8.734585811693023e-05, "loss": 0.0225, "step": 3589 }, { "epoch": 0.38920208152645275, "grad_norm": 0.47343388199806213, "learning_rate": 8.734223124909328e-05, "loss": 0.0216, "step": 3590 }, { "epoch": 0.3893104943625325, "grad_norm": 0.484605073928833, "learning_rate": 8.733860438125636e-05, "loss": 0.062, "step": 3591 }, { "epoch": 0.38941890719861233, "grad_norm": 0.24208348989486694, "learning_rate": 8.733497751341941e-05, "loss": 0.0228, "step": 3592 }, { "epoch": 0.3895273200346921, "grad_norm": 0.2801293134689331, "learning_rate": 8.733135064558248e-05, "loss": 0.0386, "step": 3593 }, { "epoch": 0.3896357328707719, "grad_norm": 0.7729371190071106, "learning_rate": 8.732772377774554e-05, "loss": 0.0981, "step": 3594 }, { "epoch": 0.3897441457068517, "grad_norm": 0.19481094181537628, "learning_rate": 8.73240969099086e-05, "loss": 0.0263, "step": 3595 }, { "epoch": 0.3898525585429315, "grad_norm": 0.3545750379562378, "learning_rate": 8.732047004207167e-05, "loss": 0.0497, "step": 3596 }, { "epoch": 0.38996097137901126, "grad_norm": 0.15033890306949615, "learning_rate": 8.731684317423473e-05, "loss": 0.0107, "step": 3597 }, { "epoch": 0.3900693842150911, "grad_norm": 0.18989987671375275, "learning_rate": 8.73132163063978e-05, "loss": 0.0319, "step": 3598 }, { "epoch": 0.39017779705117084, "grad_norm": 0.37215813994407654, "learning_rate": 8.730958943856087e-05, "loss": 0.0333, "step": 3599 }, { "epoch": 0.39028620988725066, "grad_norm": 0.576500415802002, "learning_rate": 8.730596257072393e-05, "loss": 0.0557, "step": 3600 }, { "epoch": 0.3903946227233304, "grad_norm": 0.445517897605896, "learning_rate": 8.7302335702887e-05, "loss": 0.0528, "step": 3601 }, { "epoch": 0.39050303555941024, "grad_norm": 0.2881935238838196, "learning_rate": 8.729870883505005e-05, "loss": 0.0361, "step": 3602 }, { "epoch": 0.39061144839549, "grad_norm": 0.32044196128845215, "learning_rate": 8.729508196721312e-05, "loss": 0.0236, "step": 3603 }, { "epoch": 0.3907198612315698, "grad_norm": 0.24216987192630768, "learning_rate": 8.729145509937618e-05, "loss": 0.0269, "step": 3604 }, { "epoch": 0.3908282740676496, "grad_norm": 0.32045862078666687, "learning_rate": 8.728782823153925e-05, "loss": 0.0322, "step": 3605 }, { "epoch": 0.3909366869037294, "grad_norm": 0.26174870133399963, "learning_rate": 8.728420136370231e-05, "loss": 0.062, "step": 3606 }, { "epoch": 0.3910450997398092, "grad_norm": 0.2569129467010498, "learning_rate": 8.728057449586538e-05, "loss": 0.0309, "step": 3607 }, { "epoch": 0.391153512575889, "grad_norm": 0.12344015389680862, "learning_rate": 8.727694762802844e-05, "loss": 0.0132, "step": 3608 }, { "epoch": 0.3912619254119688, "grad_norm": 0.39873284101486206, "learning_rate": 8.72733207601915e-05, "loss": 0.0335, "step": 3609 }, { "epoch": 0.39137033824804857, "grad_norm": 0.38546913862228394, "learning_rate": 8.726969389235457e-05, "loss": 0.0506, "step": 3610 }, { "epoch": 0.3914787510841284, "grad_norm": 0.6577634811401367, "learning_rate": 8.726606702451762e-05, "loss": 0.0883, "step": 3611 }, { "epoch": 0.39158716392020815, "grad_norm": 0.8751890659332275, "learning_rate": 8.72624401566807e-05, "loss": 0.0751, "step": 3612 }, { "epoch": 0.39169557675628797, "grad_norm": 0.42214083671569824, "learning_rate": 8.725881328884375e-05, "loss": 0.0473, "step": 3613 }, { "epoch": 0.39180398959236773, "grad_norm": 0.8242529630661011, "learning_rate": 8.725518642100682e-05, "loss": 0.0908, "step": 3614 }, { "epoch": 0.39191240242844755, "grad_norm": 8.580159187316895, "learning_rate": 8.725155955316988e-05, "loss": 0.0938, "step": 3615 }, { "epoch": 0.3920208152645273, "grad_norm": 0.2897946238517761, "learning_rate": 8.724793268533295e-05, "loss": 0.0293, "step": 3616 }, { "epoch": 0.39212922810060713, "grad_norm": 0.5291116833686829, "learning_rate": 8.724430581749602e-05, "loss": 0.0535, "step": 3617 }, { "epoch": 0.3922376409366869, "grad_norm": 0.3349181115627289, "learning_rate": 8.724067894965908e-05, "loss": 0.0484, "step": 3618 }, { "epoch": 0.3923460537727667, "grad_norm": 0.3551866412162781, "learning_rate": 8.723705208182215e-05, "loss": 0.0361, "step": 3619 }, { "epoch": 0.3924544666088465, "grad_norm": 0.4476538896560669, "learning_rate": 8.723342521398521e-05, "loss": 0.0546, "step": 3620 }, { "epoch": 0.3925628794449263, "grad_norm": 0.2796900272369385, "learning_rate": 8.722979834614828e-05, "loss": 0.0494, "step": 3621 }, { "epoch": 0.39267129228100606, "grad_norm": 0.2002355456352234, "learning_rate": 8.722617147831134e-05, "loss": 0.0304, "step": 3622 }, { "epoch": 0.3927797051170859, "grad_norm": 0.5282237529754639, "learning_rate": 8.722254461047439e-05, "loss": 0.0868, "step": 3623 }, { "epoch": 0.39288811795316564, "grad_norm": 0.330902636051178, "learning_rate": 8.721891774263746e-05, "loss": 0.0379, "step": 3624 }, { "epoch": 0.39299653078924546, "grad_norm": 0.41601863503456116, "learning_rate": 8.721529087480052e-05, "loss": 0.0906, "step": 3625 }, { "epoch": 0.3931049436253252, "grad_norm": 0.34091541171073914, "learning_rate": 8.721166400696359e-05, "loss": 0.0295, "step": 3626 }, { "epoch": 0.39321335646140504, "grad_norm": 0.13197150826454163, "learning_rate": 8.720803713912665e-05, "loss": 0.0105, "step": 3627 }, { "epoch": 0.3933217692974848, "grad_norm": 0.36439529061317444, "learning_rate": 8.720441027128972e-05, "loss": 0.0461, "step": 3628 }, { "epoch": 0.3934301821335646, "grad_norm": 0.1875290870666504, "learning_rate": 8.720078340345278e-05, "loss": 0.0356, "step": 3629 }, { "epoch": 0.3935385949696444, "grad_norm": 0.22673803567886353, "learning_rate": 8.719715653561585e-05, "loss": 0.0224, "step": 3630 }, { "epoch": 0.3936470078057242, "grad_norm": 0.43989697098731995, "learning_rate": 8.71935296677789e-05, "loss": 0.0329, "step": 3631 }, { "epoch": 0.39375542064180397, "grad_norm": 0.1507033407688141, "learning_rate": 8.718990279994196e-05, "loss": 0.0228, "step": 3632 }, { "epoch": 0.3938638334778838, "grad_norm": 0.2997307777404785, "learning_rate": 8.718627593210505e-05, "loss": 0.0492, "step": 3633 }, { "epoch": 0.39397224631396355, "grad_norm": 0.23072050511837006, "learning_rate": 8.71826490642681e-05, "loss": 0.0193, "step": 3634 }, { "epoch": 0.39408065915004337, "grad_norm": 0.23258596658706665, "learning_rate": 8.717902219643118e-05, "loss": 0.0482, "step": 3635 }, { "epoch": 0.39418907198612313, "grad_norm": 0.6331446766853333, "learning_rate": 8.717539532859423e-05, "loss": 0.095, "step": 3636 }, { "epoch": 0.39429748482220295, "grad_norm": 0.17575062811374664, "learning_rate": 8.717176846075729e-05, "loss": 0.0208, "step": 3637 }, { "epoch": 0.39440589765828277, "grad_norm": 0.16258183121681213, "learning_rate": 8.716814159292036e-05, "loss": 0.0212, "step": 3638 }, { "epoch": 0.39451431049436253, "grad_norm": 0.32066014409065247, "learning_rate": 8.716451472508342e-05, "loss": 0.0323, "step": 3639 }, { "epoch": 0.39462272333044235, "grad_norm": 0.4137991964817047, "learning_rate": 8.716088785724649e-05, "loss": 0.0332, "step": 3640 }, { "epoch": 0.3947311361665221, "grad_norm": 0.19799292087554932, "learning_rate": 8.715726098940955e-05, "loss": 0.0257, "step": 3641 }, { "epoch": 0.39483954900260193, "grad_norm": 0.1978387087583542, "learning_rate": 8.715363412157262e-05, "loss": 0.0245, "step": 3642 }, { "epoch": 0.3949479618386817, "grad_norm": 0.2959921658039093, "learning_rate": 8.715000725373568e-05, "loss": 0.072, "step": 3643 }, { "epoch": 0.3950563746747615, "grad_norm": 0.26639777421951294, "learning_rate": 8.714638038589875e-05, "loss": 0.0499, "step": 3644 }, { "epoch": 0.3951647875108413, "grad_norm": 0.38436034321784973, "learning_rate": 8.71427535180618e-05, "loss": 0.0485, "step": 3645 }, { "epoch": 0.3952732003469211, "grad_norm": 0.18723319470882416, "learning_rate": 8.713912665022486e-05, "loss": 0.0234, "step": 3646 }, { "epoch": 0.39538161318300086, "grad_norm": 0.46289458870887756, "learning_rate": 8.713549978238793e-05, "loss": 0.0786, "step": 3647 }, { "epoch": 0.3954900260190807, "grad_norm": 0.21517318487167358, "learning_rate": 8.713187291455099e-05, "loss": 0.034, "step": 3648 }, { "epoch": 0.39559843885516044, "grad_norm": 0.4629596173763275, "learning_rate": 8.712824604671406e-05, "loss": 0.041, "step": 3649 }, { "epoch": 0.39570685169124026, "grad_norm": 0.36091962456703186, "learning_rate": 8.712461917887712e-05, "loss": 0.0521, "step": 3650 }, { "epoch": 0.39581526452732, "grad_norm": 0.18899871408939362, "learning_rate": 8.71209923110402e-05, "loss": 0.0152, "step": 3651 }, { "epoch": 0.39592367736339984, "grad_norm": 0.36274340748786926, "learning_rate": 8.711736544320326e-05, "loss": 0.0346, "step": 3652 }, { "epoch": 0.3960320901994796, "grad_norm": 0.28173425793647766, "learning_rate": 8.711373857536632e-05, "loss": 0.051, "step": 3653 }, { "epoch": 0.3961405030355594, "grad_norm": 0.6390611529350281, "learning_rate": 8.711011170752939e-05, "loss": 0.0531, "step": 3654 }, { "epoch": 0.3962489158716392, "grad_norm": 0.3251732289791107, "learning_rate": 8.710648483969244e-05, "loss": 0.0207, "step": 3655 }, { "epoch": 0.396357328707719, "grad_norm": 0.808269202709198, "learning_rate": 8.710285797185552e-05, "loss": 0.0574, "step": 3656 }, { "epoch": 0.39646574154379877, "grad_norm": 0.3257475793361664, "learning_rate": 8.709923110401857e-05, "loss": 0.0275, "step": 3657 }, { "epoch": 0.3965741543798786, "grad_norm": 0.16766071319580078, "learning_rate": 8.709560423618164e-05, "loss": 0.0185, "step": 3658 }, { "epoch": 0.39668256721595835, "grad_norm": 0.3272079527378082, "learning_rate": 8.70919773683447e-05, "loss": 0.042, "step": 3659 }, { "epoch": 0.39679098005203817, "grad_norm": 0.2460934817790985, "learning_rate": 8.708835050050777e-05, "loss": 0.0238, "step": 3660 }, { "epoch": 0.39689939288811793, "grad_norm": 0.23901332914829254, "learning_rate": 8.708472363267083e-05, "loss": 0.0248, "step": 3661 }, { "epoch": 0.39700780572419775, "grad_norm": 0.39674854278564453, "learning_rate": 8.708109676483389e-05, "loss": 0.0513, "step": 3662 }, { "epoch": 0.3971162185602775, "grad_norm": 0.22472266852855682, "learning_rate": 8.707746989699696e-05, "loss": 0.022, "step": 3663 }, { "epoch": 0.39722463139635733, "grad_norm": 0.7075787782669067, "learning_rate": 8.707384302916001e-05, "loss": 0.0341, "step": 3664 }, { "epoch": 0.3973330442324371, "grad_norm": 0.40485256910324097, "learning_rate": 8.707021616132309e-05, "loss": 0.0691, "step": 3665 }, { "epoch": 0.3974414570685169, "grad_norm": 0.2368883490562439, "learning_rate": 8.706658929348614e-05, "loss": 0.0308, "step": 3666 }, { "epoch": 0.39754986990459673, "grad_norm": 0.6957392692565918, "learning_rate": 8.706296242564921e-05, "loss": 0.0265, "step": 3667 }, { "epoch": 0.3976582827406765, "grad_norm": 0.36986109614372253, "learning_rate": 8.705933555781228e-05, "loss": 0.0456, "step": 3668 }, { "epoch": 0.3977666955767563, "grad_norm": 0.33531996607780457, "learning_rate": 8.705570868997534e-05, "loss": 0.0288, "step": 3669 }, { "epoch": 0.3978751084128361, "grad_norm": 0.5410205721855164, "learning_rate": 8.705208182213841e-05, "loss": 0.0469, "step": 3670 }, { "epoch": 0.3979835212489159, "grad_norm": 0.27919381856918335, "learning_rate": 8.704845495430147e-05, "loss": 0.0359, "step": 3671 }, { "epoch": 0.39809193408499566, "grad_norm": 0.3713505268096924, "learning_rate": 8.704482808646454e-05, "loss": 0.0492, "step": 3672 }, { "epoch": 0.3982003469210755, "grad_norm": 0.18590471148490906, "learning_rate": 8.70412012186276e-05, "loss": 0.015, "step": 3673 }, { "epoch": 0.39830875975715524, "grad_norm": 0.5548122525215149, "learning_rate": 8.703757435079067e-05, "loss": 0.0737, "step": 3674 }, { "epoch": 0.39841717259323506, "grad_norm": 0.3243492841720581, "learning_rate": 8.703394748295373e-05, "loss": 0.0322, "step": 3675 }, { "epoch": 0.3985255854293148, "grad_norm": 0.3887299597263336, "learning_rate": 8.703032061511678e-05, "loss": 0.0497, "step": 3676 }, { "epoch": 0.39863399826539464, "grad_norm": 0.6522678732872009, "learning_rate": 8.702669374727985e-05, "loss": 0.0723, "step": 3677 }, { "epoch": 0.3987424111014744, "grad_norm": 0.6833291053771973, "learning_rate": 8.702306687944291e-05, "loss": 0.0657, "step": 3678 }, { "epoch": 0.3988508239375542, "grad_norm": 0.32393336296081543, "learning_rate": 8.701944001160598e-05, "loss": 0.0195, "step": 3679 }, { "epoch": 0.398959236773634, "grad_norm": 0.5402754545211792, "learning_rate": 8.701581314376904e-05, "loss": 0.0413, "step": 3680 }, { "epoch": 0.3990676496097138, "grad_norm": 0.20354601740837097, "learning_rate": 8.701218627593211e-05, "loss": 0.0213, "step": 3681 }, { "epoch": 0.39917606244579357, "grad_norm": 0.5664681196212769, "learning_rate": 8.700855940809517e-05, "loss": 0.0475, "step": 3682 }, { "epoch": 0.3992844752818734, "grad_norm": 0.3253028690814972, "learning_rate": 8.700493254025824e-05, "loss": 0.0626, "step": 3683 }, { "epoch": 0.39939288811795315, "grad_norm": 0.3694376051425934, "learning_rate": 8.70013056724213e-05, "loss": 0.0465, "step": 3684 }, { "epoch": 0.39950130095403297, "grad_norm": 0.10530354827642441, "learning_rate": 8.699767880458437e-05, "loss": 0.0147, "step": 3685 }, { "epoch": 0.39960971379011273, "grad_norm": 0.5960802435874939, "learning_rate": 8.699405193674744e-05, "loss": 0.0941, "step": 3686 }, { "epoch": 0.39971812662619255, "grad_norm": 0.18582040071487427, "learning_rate": 8.69904250689105e-05, "loss": 0.0269, "step": 3687 }, { "epoch": 0.3998265394622723, "grad_norm": 0.3351210057735443, "learning_rate": 8.698679820107357e-05, "loss": 0.0434, "step": 3688 }, { "epoch": 0.39993495229835213, "grad_norm": 0.6877956390380859, "learning_rate": 8.698317133323662e-05, "loss": 0.0668, "step": 3689 }, { "epoch": 0.4000433651344319, "grad_norm": 0.13703501224517822, "learning_rate": 8.697954446539968e-05, "loss": 0.0103, "step": 3690 }, { "epoch": 0.4001517779705117, "grad_norm": 0.2222721129655838, "learning_rate": 8.697591759756275e-05, "loss": 0.0369, "step": 3691 }, { "epoch": 0.4002601908065915, "grad_norm": 0.2619682848453522, "learning_rate": 8.697229072972581e-05, "loss": 0.0256, "step": 3692 }, { "epoch": 0.4003686036426713, "grad_norm": 0.387808620929718, "learning_rate": 8.696866386188888e-05, "loss": 0.0371, "step": 3693 }, { "epoch": 0.40047701647875106, "grad_norm": 0.3736640214920044, "learning_rate": 8.696503699405194e-05, "loss": 0.0363, "step": 3694 }, { "epoch": 0.4005854293148309, "grad_norm": 0.5897755026817322, "learning_rate": 8.696141012621501e-05, "loss": 0.0522, "step": 3695 }, { "epoch": 0.4006938421509107, "grad_norm": 0.2512015998363495, "learning_rate": 8.695778325837807e-05, "loss": 0.0217, "step": 3696 }, { "epoch": 0.40080225498699046, "grad_norm": 0.7843658924102783, "learning_rate": 8.695415639054114e-05, "loss": 0.0527, "step": 3697 }, { "epoch": 0.4009106678230703, "grad_norm": 0.23410217463970184, "learning_rate": 8.69505295227042e-05, "loss": 0.0253, "step": 3698 }, { "epoch": 0.40101908065915004, "grad_norm": 0.23923729360103607, "learning_rate": 8.694690265486725e-05, "loss": 0.0314, "step": 3699 }, { "epoch": 0.40112749349522986, "grad_norm": 0.39384499192237854, "learning_rate": 8.694327578703032e-05, "loss": 0.0258, "step": 3700 }, { "epoch": 0.4012359063313096, "grad_norm": 0.5868948698043823, "learning_rate": 8.693964891919338e-05, "loss": 0.0796, "step": 3701 }, { "epoch": 0.40134431916738944, "grad_norm": 0.9526852965354919, "learning_rate": 8.693602205135646e-05, "loss": 0.0633, "step": 3702 }, { "epoch": 0.4014527320034692, "grad_norm": 0.3963961601257324, "learning_rate": 8.693239518351952e-05, "loss": 0.0845, "step": 3703 }, { "epoch": 0.401561144839549, "grad_norm": 0.33938050270080566, "learning_rate": 8.692876831568258e-05, "loss": 0.0536, "step": 3704 }, { "epoch": 0.4016695576756288, "grad_norm": 0.36639782786369324, "learning_rate": 8.692514144784565e-05, "loss": 0.047, "step": 3705 }, { "epoch": 0.4017779705117086, "grad_norm": 0.6143946051597595, "learning_rate": 8.692151458000871e-05, "loss": 0.0538, "step": 3706 }, { "epoch": 0.40188638334778837, "grad_norm": 0.4430137574672699, "learning_rate": 8.691788771217178e-05, "loss": 0.0474, "step": 3707 }, { "epoch": 0.4019947961838682, "grad_norm": 0.29173511266708374, "learning_rate": 8.691426084433484e-05, "loss": 0.0516, "step": 3708 }, { "epoch": 0.40210320901994795, "grad_norm": 0.1832292079925537, "learning_rate": 8.69106339764979e-05, "loss": 0.0194, "step": 3709 }, { "epoch": 0.40221162185602777, "grad_norm": 0.2901257872581482, "learning_rate": 8.690700710866096e-05, "loss": 0.0427, "step": 3710 }, { "epoch": 0.40232003469210753, "grad_norm": 0.2658754885196686, "learning_rate": 8.690338024082403e-05, "loss": 0.0436, "step": 3711 }, { "epoch": 0.40242844752818735, "grad_norm": 0.51604163646698, "learning_rate": 8.689975337298709e-05, "loss": 0.0822, "step": 3712 }, { "epoch": 0.4025368603642671, "grad_norm": 0.4718897044658661, "learning_rate": 8.689612650515015e-05, "loss": 0.0303, "step": 3713 }, { "epoch": 0.40264527320034693, "grad_norm": 0.3450038731098175, "learning_rate": 8.689249963731322e-05, "loss": 0.0219, "step": 3714 }, { "epoch": 0.4027536860364267, "grad_norm": 0.23364995419979095, "learning_rate": 8.688887276947628e-05, "loss": 0.0288, "step": 3715 }, { "epoch": 0.4028620988725065, "grad_norm": 0.33137261867523193, "learning_rate": 8.688524590163935e-05, "loss": 0.045, "step": 3716 }, { "epoch": 0.4029705117085863, "grad_norm": 0.41660892963409424, "learning_rate": 8.68816190338024e-05, "loss": 0.0653, "step": 3717 }, { "epoch": 0.4030789245446661, "grad_norm": 0.4895144999027252, "learning_rate": 8.687799216596548e-05, "loss": 0.0956, "step": 3718 }, { "epoch": 0.40318733738074586, "grad_norm": 0.2744457721710205, "learning_rate": 8.687436529812853e-05, "loss": 0.035, "step": 3719 }, { "epoch": 0.4032957502168257, "grad_norm": 0.26676562428474426, "learning_rate": 8.68707384302916e-05, "loss": 0.0511, "step": 3720 }, { "epoch": 0.40340416305290544, "grad_norm": 0.22778938710689545, "learning_rate": 8.686711156245468e-05, "loss": 0.0467, "step": 3721 }, { "epoch": 0.40351257588898526, "grad_norm": 0.277509868144989, "learning_rate": 8.686348469461773e-05, "loss": 0.0485, "step": 3722 }, { "epoch": 0.403620988725065, "grad_norm": 0.3981328308582306, "learning_rate": 8.68598578267808e-05, "loss": 0.0705, "step": 3723 }, { "epoch": 0.40372940156114484, "grad_norm": 0.24591480195522308, "learning_rate": 8.685623095894386e-05, "loss": 0.0353, "step": 3724 }, { "epoch": 0.4038378143972246, "grad_norm": 0.19768665730953217, "learning_rate": 8.685260409110693e-05, "loss": 0.0405, "step": 3725 }, { "epoch": 0.4039462272333044, "grad_norm": 0.4276306629180908, "learning_rate": 8.684897722326999e-05, "loss": 0.0547, "step": 3726 }, { "epoch": 0.40405464006938424, "grad_norm": 0.15652169287204742, "learning_rate": 8.684535035543305e-05, "loss": 0.0294, "step": 3727 }, { "epoch": 0.404163052905464, "grad_norm": 0.26272132992744446, "learning_rate": 8.684172348759612e-05, "loss": 0.0409, "step": 3728 }, { "epoch": 0.4042714657415438, "grad_norm": 0.5597295165061951, "learning_rate": 8.683809661975917e-05, "loss": 0.0388, "step": 3729 }, { "epoch": 0.4043798785776236, "grad_norm": 0.31076687574386597, "learning_rate": 8.683446975192225e-05, "loss": 0.0405, "step": 3730 }, { "epoch": 0.4044882914137034, "grad_norm": 0.2457931488752365, "learning_rate": 8.68308428840853e-05, "loss": 0.0373, "step": 3731 }, { "epoch": 0.40459670424978317, "grad_norm": 0.23922351002693176, "learning_rate": 8.682721601624837e-05, "loss": 0.0692, "step": 3732 }, { "epoch": 0.404705117085863, "grad_norm": 0.2645537555217743, "learning_rate": 8.682358914841143e-05, "loss": 0.034, "step": 3733 }, { "epoch": 0.40481352992194275, "grad_norm": 0.30472567677497864, "learning_rate": 8.68199622805745e-05, "loss": 0.0362, "step": 3734 }, { "epoch": 0.40492194275802257, "grad_norm": 0.4718228578567505, "learning_rate": 8.681633541273756e-05, "loss": 0.0619, "step": 3735 }, { "epoch": 0.40503035559410233, "grad_norm": 0.5337140560150146, "learning_rate": 8.681270854490062e-05, "loss": 0.0524, "step": 3736 }, { "epoch": 0.40513876843018215, "grad_norm": 0.3262106478214264, "learning_rate": 8.68090816770637e-05, "loss": 0.0476, "step": 3737 }, { "epoch": 0.4052471812662619, "grad_norm": 0.3628188669681549, "learning_rate": 8.680545480922676e-05, "loss": 0.0407, "step": 3738 }, { "epoch": 0.40535559410234173, "grad_norm": 0.5774675607681274, "learning_rate": 8.680182794138983e-05, "loss": 0.0468, "step": 3739 }, { "epoch": 0.4054640069384215, "grad_norm": 0.19364377856254578, "learning_rate": 8.679820107355289e-05, "loss": 0.0308, "step": 3740 }, { "epoch": 0.4055724197745013, "grad_norm": 0.3714228868484497, "learning_rate": 8.679457420571594e-05, "loss": 0.0722, "step": 3741 }, { "epoch": 0.4056808326105811, "grad_norm": 0.13205134868621826, "learning_rate": 8.679094733787901e-05, "loss": 0.0246, "step": 3742 }, { "epoch": 0.4057892454466609, "grad_norm": 0.7120660543441772, "learning_rate": 8.678732047004207e-05, "loss": 0.0403, "step": 3743 }, { "epoch": 0.40589765828274066, "grad_norm": 0.7034572958946228, "learning_rate": 8.678369360220514e-05, "loss": 0.032, "step": 3744 }, { "epoch": 0.4060060711188205, "grad_norm": 0.3708285689353943, "learning_rate": 8.67800667343682e-05, "loss": 0.0322, "step": 3745 }, { "epoch": 0.40611448395490024, "grad_norm": 0.5885559916496277, "learning_rate": 8.677643986653127e-05, "loss": 0.0825, "step": 3746 }, { "epoch": 0.40622289679098006, "grad_norm": 0.17448489367961884, "learning_rate": 8.677281299869433e-05, "loss": 0.0152, "step": 3747 }, { "epoch": 0.4063313096270598, "grad_norm": 0.3470627963542938, "learning_rate": 8.67691861308574e-05, "loss": 0.0567, "step": 3748 }, { "epoch": 0.40643972246313964, "grad_norm": 0.3422091007232666, "learning_rate": 8.676555926302046e-05, "loss": 0.0426, "step": 3749 }, { "epoch": 0.4065481352992194, "grad_norm": 0.12008178234100342, "learning_rate": 8.676193239518351e-05, "loss": 0.0187, "step": 3750 }, { "epoch": 0.4066565481352992, "grad_norm": 0.14025025069713593, "learning_rate": 8.675830552734658e-05, "loss": 0.0209, "step": 3751 }, { "epoch": 0.406764960971379, "grad_norm": 0.7314813137054443, "learning_rate": 8.675467865950964e-05, "loss": 0.068, "step": 3752 }, { "epoch": 0.4068733738074588, "grad_norm": 0.11799655854701996, "learning_rate": 8.675105179167271e-05, "loss": 0.0252, "step": 3753 }, { "epoch": 0.40698178664353857, "grad_norm": 0.21357987821102142, "learning_rate": 8.674742492383578e-05, "loss": 0.0228, "step": 3754 }, { "epoch": 0.4070901994796184, "grad_norm": 0.2698986530303955, "learning_rate": 8.674379805599885e-05, "loss": 0.0255, "step": 3755 }, { "epoch": 0.4071986123156982, "grad_norm": 0.19331224262714386, "learning_rate": 8.674017118816191e-05, "loss": 0.0205, "step": 3756 }, { "epoch": 0.40730702515177797, "grad_norm": 0.2744101583957672, "learning_rate": 8.673654432032497e-05, "loss": 0.0336, "step": 3757 }, { "epoch": 0.4074154379878578, "grad_norm": 0.35996511578559875, "learning_rate": 8.673291745248804e-05, "loss": 0.0367, "step": 3758 }, { "epoch": 0.40752385082393755, "grad_norm": 0.36625179648399353, "learning_rate": 8.67292905846511e-05, "loss": 0.0575, "step": 3759 }, { "epoch": 0.40763226366001737, "grad_norm": 0.36417803168296814, "learning_rate": 8.672566371681417e-05, "loss": 0.053, "step": 3760 }, { "epoch": 0.40774067649609713, "grad_norm": 0.49608826637268066, "learning_rate": 8.672203684897723e-05, "loss": 0.0682, "step": 3761 }, { "epoch": 0.40784908933217695, "grad_norm": 0.8222116231918335, "learning_rate": 8.67184099811403e-05, "loss": 0.0281, "step": 3762 }, { "epoch": 0.4079575021682567, "grad_norm": 0.47834402322769165, "learning_rate": 8.671478311330335e-05, "loss": 0.0619, "step": 3763 }, { "epoch": 0.40806591500433653, "grad_norm": 0.21814332902431488, "learning_rate": 8.671115624546643e-05, "loss": 0.0258, "step": 3764 }, { "epoch": 0.4081743278404163, "grad_norm": 0.3798868656158447, "learning_rate": 8.670752937762948e-05, "loss": 0.0374, "step": 3765 }, { "epoch": 0.4082827406764961, "grad_norm": 0.40169623494148254, "learning_rate": 8.670390250979254e-05, "loss": 0.0297, "step": 3766 }, { "epoch": 0.4083911535125759, "grad_norm": 0.3965549170970917, "learning_rate": 8.670027564195561e-05, "loss": 0.0691, "step": 3767 }, { "epoch": 0.4084995663486557, "grad_norm": 0.40388330817222595, "learning_rate": 8.669664877411867e-05, "loss": 0.0402, "step": 3768 }, { "epoch": 0.40860797918473546, "grad_norm": 0.23254172503948212, "learning_rate": 8.669302190628174e-05, "loss": 0.019, "step": 3769 }, { "epoch": 0.4087163920208153, "grad_norm": 0.3011637330055237, "learning_rate": 8.66893950384448e-05, "loss": 0.052, "step": 3770 }, { "epoch": 0.40882480485689504, "grad_norm": 0.6700248718261719, "learning_rate": 8.668576817060787e-05, "loss": 0.0688, "step": 3771 }, { "epoch": 0.40893321769297486, "grad_norm": 0.5259939432144165, "learning_rate": 8.668214130277094e-05, "loss": 0.0834, "step": 3772 }, { "epoch": 0.4090416305290546, "grad_norm": 0.3116445541381836, "learning_rate": 8.6678514434934e-05, "loss": 0.0404, "step": 3773 }, { "epoch": 0.40915004336513444, "grad_norm": 0.3435659408569336, "learning_rate": 8.667488756709707e-05, "loss": 0.0338, "step": 3774 }, { "epoch": 0.4092584562012142, "grad_norm": 0.33792075514793396, "learning_rate": 8.667126069926012e-05, "loss": 0.0521, "step": 3775 }, { "epoch": 0.409366869037294, "grad_norm": 0.3175831735134125, "learning_rate": 8.66676338314232e-05, "loss": 0.0426, "step": 3776 }, { "epoch": 0.4094752818733738, "grad_norm": 0.5396484732627869, "learning_rate": 8.666400696358625e-05, "loss": 0.0434, "step": 3777 }, { "epoch": 0.4095836947094536, "grad_norm": 0.6143812537193298, "learning_rate": 8.666038009574932e-05, "loss": 0.0456, "step": 3778 }, { "epoch": 0.40969210754553337, "grad_norm": 0.2792835533618927, "learning_rate": 8.665675322791238e-05, "loss": 0.0454, "step": 3779 }, { "epoch": 0.4098005203816132, "grad_norm": 0.32423731684684753, "learning_rate": 8.665312636007544e-05, "loss": 0.032, "step": 3780 }, { "epoch": 0.40990893321769295, "grad_norm": 0.2995797097682953, "learning_rate": 8.664949949223851e-05, "loss": 0.0301, "step": 3781 }, { "epoch": 0.41001734605377277, "grad_norm": 0.32200634479522705, "learning_rate": 8.664587262440157e-05, "loss": 0.0262, "step": 3782 }, { "epoch": 0.41012575888985253, "grad_norm": 0.9001383781433105, "learning_rate": 8.664224575656464e-05, "loss": 0.0629, "step": 3783 }, { "epoch": 0.41023417172593235, "grad_norm": 0.8048892617225647, "learning_rate": 8.66386188887277e-05, "loss": 0.058, "step": 3784 }, { "epoch": 0.41034258456201217, "grad_norm": 0.34143802523612976, "learning_rate": 8.663499202089076e-05, "loss": 0.0329, "step": 3785 }, { "epoch": 0.41045099739809193, "grad_norm": 0.3984759449958801, "learning_rate": 8.663136515305382e-05, "loss": 0.038, "step": 3786 }, { "epoch": 0.41055941023417175, "grad_norm": 0.26960697770118713, "learning_rate": 8.662773828521689e-05, "loss": 0.0364, "step": 3787 }, { "epoch": 0.4106678230702515, "grad_norm": 0.16194656491279602, "learning_rate": 8.662411141737995e-05, "loss": 0.0189, "step": 3788 }, { "epoch": 0.41077623590633133, "grad_norm": 0.5384594798088074, "learning_rate": 8.662048454954302e-05, "loss": 0.0251, "step": 3789 }, { "epoch": 0.4108846487424111, "grad_norm": 0.18432502448558807, "learning_rate": 8.661685768170609e-05, "loss": 0.0108, "step": 3790 }, { "epoch": 0.4109930615784909, "grad_norm": 0.363352507352829, "learning_rate": 8.661323081386915e-05, "loss": 0.0291, "step": 3791 }, { "epoch": 0.4111014744145707, "grad_norm": 0.3676832914352417, "learning_rate": 8.660960394603222e-05, "loss": 0.0276, "step": 3792 }, { "epoch": 0.4112098872506505, "grad_norm": 0.6058133244514465, "learning_rate": 8.660597707819528e-05, "loss": 0.0468, "step": 3793 }, { "epoch": 0.41131830008673026, "grad_norm": 2.309950590133667, "learning_rate": 8.660235021035833e-05, "loss": 0.0979, "step": 3794 }, { "epoch": 0.4114267129228101, "grad_norm": 0.3051944375038147, "learning_rate": 8.65987233425214e-05, "loss": 0.0237, "step": 3795 }, { "epoch": 0.41153512575888984, "grad_norm": 1.0590970516204834, "learning_rate": 8.659509647468446e-05, "loss": 0.0953, "step": 3796 }, { "epoch": 0.41164353859496966, "grad_norm": 0.7834188938140869, "learning_rate": 8.659146960684753e-05, "loss": 0.1174, "step": 3797 }, { "epoch": 0.4117519514310494, "grad_norm": 0.18625541031360626, "learning_rate": 8.658784273901059e-05, "loss": 0.0405, "step": 3798 }, { "epoch": 0.41186036426712924, "grad_norm": 0.5851626992225647, "learning_rate": 8.658421587117366e-05, "loss": 0.0613, "step": 3799 }, { "epoch": 0.411968777103209, "grad_norm": 0.25545623898506165, "learning_rate": 8.658058900333672e-05, "loss": 0.0386, "step": 3800 }, { "epoch": 0.4120771899392888, "grad_norm": 0.30548885464668274, "learning_rate": 8.657696213549979e-05, "loss": 0.0508, "step": 3801 }, { "epoch": 0.4121856027753686, "grad_norm": 0.6790766716003418, "learning_rate": 8.657333526766285e-05, "loss": 0.0687, "step": 3802 }, { "epoch": 0.4122940156114484, "grad_norm": 0.14168208837509155, "learning_rate": 8.65697083998259e-05, "loss": 0.0176, "step": 3803 }, { "epoch": 0.41240242844752817, "grad_norm": 0.33947616815567017, "learning_rate": 8.656608153198898e-05, "loss": 0.0351, "step": 3804 }, { "epoch": 0.412510841283608, "grad_norm": 0.6388203501701355, "learning_rate": 8.656245466415203e-05, "loss": 0.0424, "step": 3805 }, { "epoch": 0.41261925411968775, "grad_norm": 0.572372555732727, "learning_rate": 8.655882779631512e-05, "loss": 0.0556, "step": 3806 }, { "epoch": 0.41272766695576757, "grad_norm": 0.21085935831069946, "learning_rate": 8.655520092847817e-05, "loss": 0.0221, "step": 3807 }, { "epoch": 0.41283607979184733, "grad_norm": 0.3664703965187073, "learning_rate": 8.655157406064123e-05, "loss": 0.0344, "step": 3808 }, { "epoch": 0.41294449262792715, "grad_norm": 0.2006453424692154, "learning_rate": 8.65479471928043e-05, "loss": 0.0209, "step": 3809 }, { "epoch": 0.4130529054640069, "grad_norm": 0.41787946224212646, "learning_rate": 8.654432032496736e-05, "loss": 0.0441, "step": 3810 }, { "epoch": 0.41316131830008673, "grad_norm": 0.8218799233436584, "learning_rate": 8.654069345713043e-05, "loss": 0.0422, "step": 3811 }, { "epoch": 0.4132697311361665, "grad_norm": 0.3966529667377472, "learning_rate": 8.653706658929349e-05, "loss": 0.0372, "step": 3812 }, { "epoch": 0.4133781439722463, "grad_norm": 0.45322567224502563, "learning_rate": 8.653343972145656e-05, "loss": 0.0376, "step": 3813 }, { "epoch": 0.41348655680832613, "grad_norm": 0.41667547821998596, "learning_rate": 8.652981285361962e-05, "loss": 0.0949, "step": 3814 }, { "epoch": 0.4135949696444059, "grad_norm": 0.5848134756088257, "learning_rate": 8.652618598578269e-05, "loss": 0.0408, "step": 3815 }, { "epoch": 0.4137033824804857, "grad_norm": 0.45893195271492004, "learning_rate": 8.652255911794574e-05, "loss": 0.0425, "step": 3816 }, { "epoch": 0.4138117953165655, "grad_norm": 0.38668516278266907, "learning_rate": 8.65189322501088e-05, "loss": 0.0573, "step": 3817 }, { "epoch": 0.4139202081526453, "grad_norm": 0.5062288641929626, "learning_rate": 8.651530538227187e-05, "loss": 0.1178, "step": 3818 }, { "epoch": 0.41402862098872506, "grad_norm": 0.47074246406555176, "learning_rate": 8.651167851443493e-05, "loss": 0.0524, "step": 3819 }, { "epoch": 0.4141370338248049, "grad_norm": 0.3670222759246826, "learning_rate": 8.6508051646598e-05, "loss": 0.0465, "step": 3820 }, { "epoch": 0.41424544666088464, "grad_norm": 0.2995717525482178, "learning_rate": 8.650442477876106e-05, "loss": 0.0353, "step": 3821 }, { "epoch": 0.41435385949696446, "grad_norm": 0.27599695324897766, "learning_rate": 8.650079791092413e-05, "loss": 0.0453, "step": 3822 }, { "epoch": 0.4144622723330442, "grad_norm": 0.4706856310367584, "learning_rate": 8.64971710430872e-05, "loss": 0.0487, "step": 3823 }, { "epoch": 0.41457068516912404, "grad_norm": 0.25003868341445923, "learning_rate": 8.649354417525026e-05, "loss": 0.0237, "step": 3824 }, { "epoch": 0.4146790980052038, "grad_norm": 0.23929893970489502, "learning_rate": 8.648991730741333e-05, "loss": 0.0445, "step": 3825 }, { "epoch": 0.4147875108412836, "grad_norm": 0.5516413450241089, "learning_rate": 8.648629043957639e-05, "loss": 0.0583, "step": 3826 }, { "epoch": 0.4148959236773634, "grad_norm": 0.30380916595458984, "learning_rate": 8.648266357173946e-05, "loss": 0.0216, "step": 3827 }, { "epoch": 0.4150043365134432, "grad_norm": 0.3120384216308594, "learning_rate": 8.647903670390251e-05, "loss": 0.0466, "step": 3828 }, { "epoch": 0.41511274934952297, "grad_norm": 0.1953946202993393, "learning_rate": 8.647540983606559e-05, "loss": 0.0384, "step": 3829 }, { "epoch": 0.4152211621856028, "grad_norm": 0.21715083718299866, "learning_rate": 8.647178296822864e-05, "loss": 0.026, "step": 3830 }, { "epoch": 0.41532957502168255, "grad_norm": 0.3320150375366211, "learning_rate": 8.64681561003917e-05, "loss": 0.0644, "step": 3831 }, { "epoch": 0.41543798785776237, "grad_norm": 0.5812113285064697, "learning_rate": 8.646452923255477e-05, "loss": 0.062, "step": 3832 }, { "epoch": 0.41554640069384213, "grad_norm": 0.5616656541824341, "learning_rate": 8.646090236471783e-05, "loss": 0.0306, "step": 3833 }, { "epoch": 0.41565481352992195, "grad_norm": 0.2690008580684662, "learning_rate": 8.64572754968809e-05, "loss": 0.0321, "step": 3834 }, { "epoch": 0.4157632263660017, "grad_norm": 0.776238739490509, "learning_rate": 8.645364862904396e-05, "loss": 0.0452, "step": 3835 }, { "epoch": 0.41587163920208153, "grad_norm": 0.30830395221710205, "learning_rate": 8.645002176120703e-05, "loss": 0.0393, "step": 3836 }, { "epoch": 0.4159800520381613, "grad_norm": 0.19327399134635925, "learning_rate": 8.644639489337008e-05, "loss": 0.0278, "step": 3837 }, { "epoch": 0.4160884648742411, "grad_norm": 0.4568084180355072, "learning_rate": 8.644276802553316e-05, "loss": 0.0646, "step": 3838 }, { "epoch": 0.4161968777103209, "grad_norm": 0.06871873140335083, "learning_rate": 8.643914115769621e-05, "loss": 0.0071, "step": 3839 }, { "epoch": 0.4163052905464007, "grad_norm": 0.41979801654815674, "learning_rate": 8.643551428985927e-05, "loss": 0.0264, "step": 3840 }, { "epoch": 0.41641370338248046, "grad_norm": 0.22650311887264252, "learning_rate": 8.643188742202235e-05, "loss": 0.0115, "step": 3841 }, { "epoch": 0.4165221162185603, "grad_norm": 0.4692320227622986, "learning_rate": 8.642826055418541e-05, "loss": 0.0475, "step": 3842 }, { "epoch": 0.41663052905464004, "grad_norm": 0.18074931204319, "learning_rate": 8.642463368634848e-05, "loss": 0.0146, "step": 3843 }, { "epoch": 0.41673894189071986, "grad_norm": 0.6234591007232666, "learning_rate": 8.642100681851154e-05, "loss": 0.0698, "step": 3844 }, { "epoch": 0.4168473547267997, "grad_norm": 0.5388893485069275, "learning_rate": 8.641737995067461e-05, "loss": 0.0578, "step": 3845 }, { "epoch": 0.41695576756287944, "grad_norm": 0.49771156907081604, "learning_rate": 8.641375308283767e-05, "loss": 0.0677, "step": 3846 }, { "epoch": 0.41706418039895926, "grad_norm": 0.3871364891529083, "learning_rate": 8.641012621500073e-05, "loss": 0.0638, "step": 3847 }, { "epoch": 0.417172593235039, "grad_norm": 0.3121882379055023, "learning_rate": 8.64064993471638e-05, "loss": 0.0369, "step": 3848 }, { "epoch": 0.41728100607111884, "grad_norm": 0.214106485247612, "learning_rate": 8.640287247932685e-05, "loss": 0.0244, "step": 3849 }, { "epoch": 0.4173894189071986, "grad_norm": 0.5322237610816956, "learning_rate": 8.639924561148992e-05, "loss": 0.0484, "step": 3850 }, { "epoch": 0.4174978317432784, "grad_norm": 0.30726906657218933, "learning_rate": 8.639561874365298e-05, "loss": 0.0657, "step": 3851 }, { "epoch": 0.4176062445793582, "grad_norm": 0.18480806052684784, "learning_rate": 8.639199187581605e-05, "loss": 0.027, "step": 3852 }, { "epoch": 0.417714657415438, "grad_norm": 0.4375605285167694, "learning_rate": 8.638836500797911e-05, "loss": 0.0448, "step": 3853 }, { "epoch": 0.41782307025151777, "grad_norm": 0.2671351730823517, "learning_rate": 8.638473814014217e-05, "loss": 0.061, "step": 3854 }, { "epoch": 0.4179314830875976, "grad_norm": 0.5890136957168579, "learning_rate": 8.638111127230524e-05, "loss": 0.0274, "step": 3855 }, { "epoch": 0.41803989592367735, "grad_norm": 1.0910711288452148, "learning_rate": 8.63774844044683e-05, "loss": 0.0595, "step": 3856 }, { "epoch": 0.41814830875975717, "grad_norm": 0.34106317162513733, "learning_rate": 8.637385753663137e-05, "loss": 0.0557, "step": 3857 }, { "epoch": 0.41825672159583693, "grad_norm": 0.2231365293264389, "learning_rate": 8.637023066879444e-05, "loss": 0.024, "step": 3858 }, { "epoch": 0.41836513443191675, "grad_norm": 0.2770988941192627, "learning_rate": 8.636660380095751e-05, "loss": 0.0228, "step": 3859 }, { "epoch": 0.4184735472679965, "grad_norm": 0.32597586512565613, "learning_rate": 8.636297693312057e-05, "loss": 0.0179, "step": 3860 }, { "epoch": 0.41858196010407633, "grad_norm": 0.42250925302505493, "learning_rate": 8.635935006528362e-05, "loss": 0.0228, "step": 3861 }, { "epoch": 0.4186903729401561, "grad_norm": 0.29897549748420715, "learning_rate": 8.63557231974467e-05, "loss": 0.0517, "step": 3862 }, { "epoch": 0.4187987857762359, "grad_norm": 0.3424343466758728, "learning_rate": 8.635209632960975e-05, "loss": 0.0346, "step": 3863 }, { "epoch": 0.4189071986123157, "grad_norm": 0.5911908745765686, "learning_rate": 8.634846946177282e-05, "loss": 0.0631, "step": 3864 }, { "epoch": 0.4190156114483955, "grad_norm": 0.4164368510246277, "learning_rate": 8.634484259393588e-05, "loss": 0.0596, "step": 3865 }, { "epoch": 0.41912402428447526, "grad_norm": 0.3519470989704132, "learning_rate": 8.634121572609895e-05, "loss": 0.0334, "step": 3866 }, { "epoch": 0.4192324371205551, "grad_norm": 0.4179629981517792, "learning_rate": 8.633758885826201e-05, "loss": 0.0363, "step": 3867 }, { "epoch": 0.41934084995663484, "grad_norm": 0.37374019622802734, "learning_rate": 8.633396199042508e-05, "loss": 0.0348, "step": 3868 }, { "epoch": 0.41944926279271466, "grad_norm": 0.22379685938358307, "learning_rate": 8.633033512258814e-05, "loss": 0.0287, "step": 3869 }, { "epoch": 0.4195576756287944, "grad_norm": 0.6418288946151733, "learning_rate": 8.632670825475119e-05, "loss": 0.0276, "step": 3870 }, { "epoch": 0.41966608846487424, "grad_norm": 0.49924322962760925, "learning_rate": 8.632308138691426e-05, "loss": 0.0336, "step": 3871 }, { "epoch": 0.419774501300954, "grad_norm": 0.2019163817167282, "learning_rate": 8.631945451907732e-05, "loss": 0.0283, "step": 3872 }, { "epoch": 0.4198829141370338, "grad_norm": 0.5731408596038818, "learning_rate": 8.631582765124039e-05, "loss": 0.0692, "step": 3873 }, { "epoch": 0.41999132697311364, "grad_norm": 0.6999542713165283, "learning_rate": 8.631220078340345e-05, "loss": 0.0487, "step": 3874 }, { "epoch": 0.4200997398091934, "grad_norm": 0.19134999811649323, "learning_rate": 8.630857391556652e-05, "loss": 0.0143, "step": 3875 }, { "epoch": 0.4202081526452732, "grad_norm": 0.2682434916496277, "learning_rate": 8.630494704772959e-05, "loss": 0.0171, "step": 3876 }, { "epoch": 0.420316565481353, "grad_norm": 0.15881194174289703, "learning_rate": 8.630132017989265e-05, "loss": 0.0185, "step": 3877 }, { "epoch": 0.4204249783174328, "grad_norm": 0.32311001420021057, "learning_rate": 8.629769331205572e-05, "loss": 0.042, "step": 3878 }, { "epoch": 0.42053339115351257, "grad_norm": 0.4031526446342468, "learning_rate": 8.629406644421878e-05, "loss": 0.0358, "step": 3879 }, { "epoch": 0.4206418039895924, "grad_norm": 0.2651035487651825, "learning_rate": 8.629043957638185e-05, "loss": 0.0242, "step": 3880 }, { "epoch": 0.42075021682567215, "grad_norm": 0.5062795877456665, "learning_rate": 8.62868127085449e-05, "loss": 0.0458, "step": 3881 }, { "epoch": 0.42085862966175197, "grad_norm": 0.5505315065383911, "learning_rate": 8.628318584070798e-05, "loss": 0.035, "step": 3882 }, { "epoch": 0.42096704249783173, "grad_norm": 0.1578635722398758, "learning_rate": 8.627955897287103e-05, "loss": 0.0307, "step": 3883 }, { "epoch": 0.42107545533391155, "grad_norm": 0.45241665840148926, "learning_rate": 8.627593210503409e-05, "loss": 0.0592, "step": 3884 }, { "epoch": 0.4211838681699913, "grad_norm": 0.3032105565071106, "learning_rate": 8.627230523719716e-05, "loss": 0.0231, "step": 3885 }, { "epoch": 0.42129228100607113, "grad_norm": 0.4474741518497467, "learning_rate": 8.626867836936022e-05, "loss": 0.0252, "step": 3886 }, { "epoch": 0.4214006938421509, "grad_norm": 0.861291229724884, "learning_rate": 8.626505150152329e-05, "loss": 0.0585, "step": 3887 }, { "epoch": 0.4215091066782307, "grad_norm": 0.7258090972900391, "learning_rate": 8.626142463368635e-05, "loss": 0.0445, "step": 3888 }, { "epoch": 0.4216175195143105, "grad_norm": 0.6524324417114258, "learning_rate": 8.625779776584942e-05, "loss": 0.1014, "step": 3889 }, { "epoch": 0.4217259323503903, "grad_norm": 0.9690038561820984, "learning_rate": 8.625417089801248e-05, "loss": 0.0403, "step": 3890 }, { "epoch": 0.42183434518647006, "grad_norm": 0.4249267876148224, "learning_rate": 8.625054403017555e-05, "loss": 0.0232, "step": 3891 }, { "epoch": 0.4219427580225499, "grad_norm": 0.6024538278579712, "learning_rate": 8.62469171623386e-05, "loss": 0.0694, "step": 3892 }, { "epoch": 0.42205117085862964, "grad_norm": 0.6166866421699524, "learning_rate": 8.624329029450167e-05, "loss": 0.0446, "step": 3893 }, { "epoch": 0.42215958369470946, "grad_norm": 0.24114710092544556, "learning_rate": 8.623966342666475e-05, "loss": 0.0305, "step": 3894 }, { "epoch": 0.4222679965307892, "grad_norm": 0.8443973660469055, "learning_rate": 8.62360365588278e-05, "loss": 0.1117, "step": 3895 }, { "epoch": 0.42237640936686904, "grad_norm": 0.19752319157123566, "learning_rate": 8.623240969099087e-05, "loss": 0.0197, "step": 3896 }, { "epoch": 0.4224848222029488, "grad_norm": 0.16830849647521973, "learning_rate": 8.622878282315393e-05, "loss": 0.0156, "step": 3897 }, { "epoch": 0.4225932350390286, "grad_norm": 0.3872740864753723, "learning_rate": 8.622515595531699e-05, "loss": 0.0618, "step": 3898 }, { "epoch": 0.4227016478751084, "grad_norm": 0.2828616797924042, "learning_rate": 8.622152908748006e-05, "loss": 0.024, "step": 3899 }, { "epoch": 0.4228100607111882, "grad_norm": 0.2577996253967285, "learning_rate": 8.621790221964312e-05, "loss": 0.0268, "step": 3900 }, { "epoch": 0.42291847354726797, "grad_norm": 0.589950680732727, "learning_rate": 8.621427535180619e-05, "loss": 0.087, "step": 3901 }, { "epoch": 0.4230268863833478, "grad_norm": 0.4421626627445221, "learning_rate": 8.621064848396924e-05, "loss": 0.0566, "step": 3902 }, { "epoch": 0.4231352992194276, "grad_norm": 0.4958118200302124, "learning_rate": 8.620702161613232e-05, "loss": 0.0295, "step": 3903 }, { "epoch": 0.42324371205550737, "grad_norm": 0.745366096496582, "learning_rate": 8.620339474829537e-05, "loss": 0.0786, "step": 3904 }, { "epoch": 0.4233521248915872, "grad_norm": 0.3047623634338379, "learning_rate": 8.619976788045844e-05, "loss": 0.0301, "step": 3905 }, { "epoch": 0.42346053772766695, "grad_norm": 0.1233011931180954, "learning_rate": 8.61961410126215e-05, "loss": 0.0094, "step": 3906 }, { "epoch": 0.42356895056374677, "grad_norm": 0.12627847492694855, "learning_rate": 8.619251414478456e-05, "loss": 0.0371, "step": 3907 }, { "epoch": 0.42367736339982653, "grad_norm": 0.28066587448120117, "learning_rate": 8.618888727694763e-05, "loss": 0.0289, "step": 3908 }, { "epoch": 0.42378577623590635, "grad_norm": 0.6019560694694519, "learning_rate": 8.618526040911069e-05, "loss": 0.0477, "step": 3909 }, { "epoch": 0.4238941890719861, "grad_norm": 0.7600803375244141, "learning_rate": 8.618163354127377e-05, "loss": 0.1088, "step": 3910 }, { "epoch": 0.42400260190806593, "grad_norm": 0.2648879587650299, "learning_rate": 8.617800667343683e-05, "loss": 0.0284, "step": 3911 }, { "epoch": 0.4241110147441457, "grad_norm": 0.2891012132167816, "learning_rate": 8.617437980559989e-05, "loss": 0.0286, "step": 3912 }, { "epoch": 0.4242194275802255, "grad_norm": 0.47422799468040466, "learning_rate": 8.617075293776296e-05, "loss": 0.0573, "step": 3913 }, { "epoch": 0.4243278404163053, "grad_norm": 0.3896055221557617, "learning_rate": 8.616712606992601e-05, "loss": 0.0347, "step": 3914 }, { "epoch": 0.4244362532523851, "grad_norm": 0.217759370803833, "learning_rate": 8.616349920208908e-05, "loss": 0.0293, "step": 3915 }, { "epoch": 0.42454466608846486, "grad_norm": 0.45633572340011597, "learning_rate": 8.615987233425214e-05, "loss": 0.0598, "step": 3916 }, { "epoch": 0.4246530789245447, "grad_norm": 0.21901750564575195, "learning_rate": 8.615624546641521e-05, "loss": 0.0321, "step": 3917 }, { "epoch": 0.42476149176062444, "grad_norm": 1.1457041501998901, "learning_rate": 8.615261859857827e-05, "loss": 0.0466, "step": 3918 }, { "epoch": 0.42486990459670426, "grad_norm": 0.2238645702600479, "learning_rate": 8.614899173074134e-05, "loss": 0.0426, "step": 3919 }, { "epoch": 0.424978317432784, "grad_norm": 0.625197172164917, "learning_rate": 8.61453648629044e-05, "loss": 0.0788, "step": 3920 }, { "epoch": 0.42508673026886384, "grad_norm": 0.9675303101539612, "learning_rate": 8.614173799506746e-05, "loss": 0.0697, "step": 3921 }, { "epoch": 0.4251951431049436, "grad_norm": 0.51080721616745, "learning_rate": 8.613811112723053e-05, "loss": 0.0707, "step": 3922 }, { "epoch": 0.4253035559410234, "grad_norm": 0.47163647413253784, "learning_rate": 8.613448425939358e-05, "loss": 0.0412, "step": 3923 }, { "epoch": 0.4254119687771032, "grad_norm": 0.27543166279792786, "learning_rate": 8.613085739155665e-05, "loss": 0.0574, "step": 3924 }, { "epoch": 0.425520381613183, "grad_norm": 0.4961376190185547, "learning_rate": 8.612723052371971e-05, "loss": 0.0506, "step": 3925 }, { "epoch": 0.42562879444926277, "grad_norm": 0.2064587026834488, "learning_rate": 8.612360365588278e-05, "loss": 0.0271, "step": 3926 }, { "epoch": 0.4257372072853426, "grad_norm": 0.24965456128120422, "learning_rate": 8.611997678804585e-05, "loss": 0.0282, "step": 3927 }, { "epoch": 0.42584562012142235, "grad_norm": 0.2629576027393341, "learning_rate": 8.611634992020891e-05, "loss": 0.0362, "step": 3928 }, { "epoch": 0.42595403295750217, "grad_norm": 0.36218178272247314, "learning_rate": 8.611272305237198e-05, "loss": 0.073, "step": 3929 }, { "epoch": 0.42606244579358193, "grad_norm": 0.240371435880661, "learning_rate": 8.610909618453504e-05, "loss": 0.0376, "step": 3930 }, { "epoch": 0.42617085862966175, "grad_norm": 0.2962785065174103, "learning_rate": 8.610546931669811e-05, "loss": 0.0472, "step": 3931 }, { "epoch": 0.4262792714657415, "grad_norm": 0.27039340138435364, "learning_rate": 8.610184244886117e-05, "loss": 0.0421, "step": 3932 }, { "epoch": 0.42638768430182133, "grad_norm": 0.6240186095237732, "learning_rate": 8.609821558102424e-05, "loss": 0.0641, "step": 3933 }, { "epoch": 0.42649609713790115, "grad_norm": 0.5247158408164978, "learning_rate": 8.60945887131873e-05, "loss": 0.0617, "step": 3934 }, { "epoch": 0.4266045099739809, "grad_norm": 0.26588231325149536, "learning_rate": 8.609096184535035e-05, "loss": 0.0354, "step": 3935 }, { "epoch": 0.42671292281006074, "grad_norm": 0.41521304845809937, "learning_rate": 8.608733497751342e-05, "loss": 0.043, "step": 3936 }, { "epoch": 0.4268213356461405, "grad_norm": 0.49133187532424927, "learning_rate": 8.608370810967648e-05, "loss": 0.0524, "step": 3937 }, { "epoch": 0.4269297484822203, "grad_norm": 0.39960262179374695, "learning_rate": 8.608008124183955e-05, "loss": 0.02, "step": 3938 }, { "epoch": 0.4270381613183001, "grad_norm": 0.46116769313812256, "learning_rate": 8.607645437400261e-05, "loss": 0.0931, "step": 3939 }, { "epoch": 0.4271465741543799, "grad_norm": 0.5043841600418091, "learning_rate": 8.607282750616568e-05, "loss": 0.0505, "step": 3940 }, { "epoch": 0.42725498699045966, "grad_norm": 0.28199103474617004, "learning_rate": 8.606920063832874e-05, "loss": 0.0602, "step": 3941 }, { "epoch": 0.4273633998265395, "grad_norm": 0.2337713986635208, "learning_rate": 8.606557377049181e-05, "loss": 0.0283, "step": 3942 }, { "epoch": 0.42747181266261924, "grad_norm": 0.20973841845989227, "learning_rate": 8.606194690265487e-05, "loss": 0.0191, "step": 3943 }, { "epoch": 0.42758022549869906, "grad_norm": 0.5474860668182373, "learning_rate": 8.605832003481794e-05, "loss": 0.0453, "step": 3944 }, { "epoch": 0.4276886383347788, "grad_norm": 0.2909141480922699, "learning_rate": 8.605469316698101e-05, "loss": 0.0458, "step": 3945 }, { "epoch": 0.42779705117085864, "grad_norm": 0.41368916630744934, "learning_rate": 8.605106629914407e-05, "loss": 0.0901, "step": 3946 }, { "epoch": 0.4279054640069384, "grad_norm": 0.3733879625797272, "learning_rate": 8.604743943130714e-05, "loss": 0.0456, "step": 3947 }, { "epoch": 0.4280138768430182, "grad_norm": 0.28545060753822327, "learning_rate": 8.60438125634702e-05, "loss": 0.0515, "step": 3948 }, { "epoch": 0.428122289679098, "grad_norm": 0.450692355632782, "learning_rate": 8.604018569563326e-05, "loss": 0.0388, "step": 3949 }, { "epoch": 0.4282307025151778, "grad_norm": 0.2458346039056778, "learning_rate": 8.603655882779632e-05, "loss": 0.0216, "step": 3950 }, { "epoch": 0.42833911535125757, "grad_norm": 0.4923168420791626, "learning_rate": 8.603293195995938e-05, "loss": 0.037, "step": 3951 }, { "epoch": 0.4284475281873374, "grad_norm": 0.22442446649074554, "learning_rate": 8.602930509212245e-05, "loss": 0.0388, "step": 3952 }, { "epoch": 0.42855594102341715, "grad_norm": 0.32294654846191406, "learning_rate": 8.602567822428551e-05, "loss": 0.0609, "step": 3953 }, { "epoch": 0.42866435385949697, "grad_norm": 0.1113891452550888, "learning_rate": 8.602205135644858e-05, "loss": 0.0138, "step": 3954 }, { "epoch": 0.42877276669557673, "grad_norm": 0.5607165694236755, "learning_rate": 8.601842448861164e-05, "loss": 0.079, "step": 3955 }, { "epoch": 0.42888117953165655, "grad_norm": 0.34831225872039795, "learning_rate": 8.60147976207747e-05, "loss": 0.0259, "step": 3956 }, { "epoch": 0.4289895923677363, "grad_norm": 0.4030137062072754, "learning_rate": 8.601117075293776e-05, "loss": 0.04, "step": 3957 }, { "epoch": 0.42909800520381614, "grad_norm": 0.3089020550251007, "learning_rate": 8.600754388510082e-05, "loss": 0.0266, "step": 3958 }, { "epoch": 0.4292064180398959, "grad_norm": 0.4050256609916687, "learning_rate": 8.600391701726389e-05, "loss": 0.0641, "step": 3959 }, { "epoch": 0.4293148308759757, "grad_norm": 0.3419824540615082, "learning_rate": 8.600029014942695e-05, "loss": 0.0535, "step": 3960 }, { "epoch": 0.4294232437120555, "grad_norm": 0.5922210216522217, "learning_rate": 8.599666328159002e-05, "loss": 0.0894, "step": 3961 }, { "epoch": 0.4295316565481353, "grad_norm": 0.2947966456413269, "learning_rate": 8.599303641375309e-05, "loss": 0.0503, "step": 3962 }, { "epoch": 0.4296400693842151, "grad_norm": 0.3074018359184265, "learning_rate": 8.598940954591616e-05, "loss": 0.0454, "step": 3963 }, { "epoch": 0.4297484822202949, "grad_norm": 0.2835844159126282, "learning_rate": 8.598578267807922e-05, "loss": 0.0399, "step": 3964 }, { "epoch": 0.4298568950563747, "grad_norm": 0.2516247034072876, "learning_rate": 8.598215581024228e-05, "loss": 0.0284, "step": 3965 }, { "epoch": 0.42996530789245446, "grad_norm": 0.7389783263206482, "learning_rate": 8.597852894240535e-05, "loss": 0.0664, "step": 3966 }, { "epoch": 0.4300737207285343, "grad_norm": 0.686247706413269, "learning_rate": 8.59749020745684e-05, "loss": 0.0429, "step": 3967 }, { "epoch": 0.43018213356461404, "grad_norm": 0.1967936009168625, "learning_rate": 8.597127520673148e-05, "loss": 0.0262, "step": 3968 }, { "epoch": 0.43029054640069386, "grad_norm": 0.36221843957901, "learning_rate": 8.596764833889453e-05, "loss": 0.032, "step": 3969 }, { "epoch": 0.4303989592367736, "grad_norm": 0.9260530471801758, "learning_rate": 8.59640214710576e-05, "loss": 0.0608, "step": 3970 }, { "epoch": 0.43050737207285344, "grad_norm": 0.6406823396682739, "learning_rate": 8.596039460322066e-05, "loss": 0.0494, "step": 3971 }, { "epoch": 0.4306157849089332, "grad_norm": 0.263818621635437, "learning_rate": 8.595676773538373e-05, "loss": 0.0285, "step": 3972 }, { "epoch": 0.430724197745013, "grad_norm": 0.33398234844207764, "learning_rate": 8.595314086754679e-05, "loss": 0.0343, "step": 3973 }, { "epoch": 0.4308326105810928, "grad_norm": 0.3922455310821533, "learning_rate": 8.594951399970985e-05, "loss": 0.0991, "step": 3974 }, { "epoch": 0.4309410234171726, "grad_norm": 0.2949748933315277, "learning_rate": 8.594588713187292e-05, "loss": 0.0279, "step": 3975 }, { "epoch": 0.43104943625325237, "grad_norm": 0.6000425219535828, "learning_rate": 8.594226026403597e-05, "loss": 0.0761, "step": 3976 }, { "epoch": 0.4311578490893322, "grad_norm": 0.25784438848495483, "learning_rate": 8.593863339619905e-05, "loss": 0.0299, "step": 3977 }, { "epoch": 0.43126626192541195, "grad_norm": 0.11562847346067429, "learning_rate": 8.59350065283621e-05, "loss": 0.0144, "step": 3978 }, { "epoch": 0.43137467476149177, "grad_norm": 0.49190932512283325, "learning_rate": 8.593137966052517e-05, "loss": 0.0233, "step": 3979 }, { "epoch": 0.43148308759757154, "grad_norm": 0.35163745284080505, "learning_rate": 8.592775279268824e-05, "loss": 0.0312, "step": 3980 }, { "epoch": 0.43159150043365135, "grad_norm": 0.6859341263771057, "learning_rate": 8.59241259248513e-05, "loss": 0.0531, "step": 3981 }, { "epoch": 0.4316999132697311, "grad_norm": 0.6092796921730042, "learning_rate": 8.592049905701437e-05, "loss": 0.043, "step": 3982 }, { "epoch": 0.43180832610581094, "grad_norm": 0.3680393099784851, "learning_rate": 8.591687218917743e-05, "loss": 0.0578, "step": 3983 }, { "epoch": 0.4319167389418907, "grad_norm": 0.41508251428604126, "learning_rate": 8.59132453213405e-05, "loss": 0.0379, "step": 3984 }, { "epoch": 0.4320251517779705, "grad_norm": 0.43647831678390503, "learning_rate": 8.590961845350356e-05, "loss": 0.0479, "step": 3985 }, { "epoch": 0.4321335646140503, "grad_norm": 0.4434424638748169, "learning_rate": 8.590599158566663e-05, "loss": 0.0578, "step": 3986 }, { "epoch": 0.4322419774501301, "grad_norm": 0.2062099128961563, "learning_rate": 8.590236471782969e-05, "loss": 0.0436, "step": 3987 }, { "epoch": 0.43235039028620986, "grad_norm": 0.504239022731781, "learning_rate": 8.589873784999274e-05, "loss": 0.032, "step": 3988 }, { "epoch": 0.4324588031222897, "grad_norm": 0.17579150199890137, "learning_rate": 8.589511098215581e-05, "loss": 0.0286, "step": 3989 }, { "epoch": 0.43256721595836944, "grad_norm": 0.2315654754638672, "learning_rate": 8.589148411431887e-05, "loss": 0.035, "step": 3990 }, { "epoch": 0.43267562879444926, "grad_norm": 0.2628055810928345, "learning_rate": 8.588785724648194e-05, "loss": 0.0321, "step": 3991 }, { "epoch": 0.4327840416305291, "grad_norm": 0.20080876350402832, "learning_rate": 8.5884230378645e-05, "loss": 0.0242, "step": 3992 }, { "epoch": 0.43289245446660884, "grad_norm": 0.1374453455209732, "learning_rate": 8.588060351080807e-05, "loss": 0.0086, "step": 3993 }, { "epoch": 0.43300086730268866, "grad_norm": 0.2749598026275635, "learning_rate": 8.587697664297113e-05, "loss": 0.0422, "step": 3994 }, { "epoch": 0.4331092801387684, "grad_norm": 0.7531274557113647, "learning_rate": 8.58733497751342e-05, "loss": 0.0657, "step": 3995 }, { "epoch": 0.43321769297484825, "grad_norm": 0.5064114332199097, "learning_rate": 8.586972290729727e-05, "loss": 0.0507, "step": 3996 }, { "epoch": 0.433326105810928, "grad_norm": 0.09696634858846664, "learning_rate": 8.586609603946033e-05, "loss": 0.0119, "step": 3997 }, { "epoch": 0.4334345186470078, "grad_norm": 0.2387237250804901, "learning_rate": 8.58624691716234e-05, "loss": 0.0272, "step": 3998 }, { "epoch": 0.4335429314830876, "grad_norm": 0.49527162313461304, "learning_rate": 8.585884230378646e-05, "loss": 0.0405, "step": 3999 }, { "epoch": 0.4336513443191674, "grad_norm": 0.2745403051376343, "learning_rate": 8.585521543594953e-05, "loss": 0.0143, "step": 4000 }, { "epoch": 0.43375975715524717, "grad_norm": 0.3295825123786926, "learning_rate": 8.585158856811258e-05, "loss": 0.0208, "step": 4001 }, { "epoch": 0.433868169991327, "grad_norm": 0.40524011850357056, "learning_rate": 8.584796170027564e-05, "loss": 0.0561, "step": 4002 }, { "epoch": 0.43397658282740675, "grad_norm": 0.3925645351409912, "learning_rate": 8.584433483243871e-05, "loss": 0.0409, "step": 4003 }, { "epoch": 0.4340849956634866, "grad_norm": 0.31001394987106323, "learning_rate": 8.584070796460177e-05, "loss": 0.0338, "step": 4004 }, { "epoch": 0.43419340849956634, "grad_norm": 0.10233473777770996, "learning_rate": 8.583708109676484e-05, "loss": 0.018, "step": 4005 }, { "epoch": 0.43430182133564615, "grad_norm": 0.3113214373588562, "learning_rate": 8.58334542289279e-05, "loss": 0.0452, "step": 4006 }, { "epoch": 0.4344102341717259, "grad_norm": 0.36922118067741394, "learning_rate": 8.582982736109097e-05, "loss": 0.0335, "step": 4007 }, { "epoch": 0.43451864700780574, "grad_norm": 0.32362332940101624, "learning_rate": 8.582620049325403e-05, "loss": 0.0242, "step": 4008 }, { "epoch": 0.4346270598438855, "grad_norm": 0.2126288264989853, "learning_rate": 8.58225736254171e-05, "loss": 0.0115, "step": 4009 }, { "epoch": 0.4347354726799653, "grad_norm": 0.5044100284576416, "learning_rate": 8.581894675758015e-05, "loss": 0.0586, "step": 4010 }, { "epoch": 0.4348438855160451, "grad_norm": 0.2910005450248718, "learning_rate": 8.581531988974321e-05, "loss": 0.0443, "step": 4011 }, { "epoch": 0.4349522983521249, "grad_norm": 0.1963512897491455, "learning_rate": 8.581169302190628e-05, "loss": 0.0251, "step": 4012 }, { "epoch": 0.43506071118820466, "grad_norm": 0.14275866746902466, "learning_rate": 8.580806615406935e-05, "loss": 0.018, "step": 4013 }, { "epoch": 0.4351691240242845, "grad_norm": 0.8892315626144409, "learning_rate": 8.580443928623242e-05, "loss": 0.0332, "step": 4014 }, { "epoch": 0.43527753686036424, "grad_norm": 0.3243987560272217, "learning_rate": 8.580081241839548e-05, "loss": 0.0285, "step": 4015 }, { "epoch": 0.43538594969644406, "grad_norm": 0.4683276414871216, "learning_rate": 8.579718555055854e-05, "loss": 0.0479, "step": 4016 }, { "epoch": 0.4354943625325238, "grad_norm": 0.5480728149414062, "learning_rate": 8.579355868272161e-05, "loss": 0.0517, "step": 4017 }, { "epoch": 0.43560277536860365, "grad_norm": 0.20097719132900238, "learning_rate": 8.578993181488467e-05, "loss": 0.031, "step": 4018 }, { "epoch": 0.4357111882046834, "grad_norm": 0.6252961158752441, "learning_rate": 8.578630494704774e-05, "loss": 0.0524, "step": 4019 }, { "epoch": 0.4358196010407632, "grad_norm": 0.3826223611831665, "learning_rate": 8.57826780792108e-05, "loss": 0.0284, "step": 4020 }, { "epoch": 0.43592801387684305, "grad_norm": 0.7023736238479614, "learning_rate": 8.577905121137387e-05, "loss": 0.0636, "step": 4021 }, { "epoch": 0.4360364267129228, "grad_norm": 0.28642842173576355, "learning_rate": 8.577542434353692e-05, "loss": 0.0267, "step": 4022 }, { "epoch": 0.4361448395490026, "grad_norm": 0.6841307282447815, "learning_rate": 8.57717974757e-05, "loss": 0.0449, "step": 4023 }, { "epoch": 0.4362532523850824, "grad_norm": 0.5502661466598511, "learning_rate": 8.576817060786305e-05, "loss": 0.0784, "step": 4024 }, { "epoch": 0.4363616652211622, "grad_norm": 0.34782880544662476, "learning_rate": 8.576454374002611e-05, "loss": 0.037, "step": 4025 }, { "epoch": 0.436470078057242, "grad_norm": 0.35271555185317993, "learning_rate": 8.576091687218918e-05, "loss": 0.0601, "step": 4026 }, { "epoch": 0.4365784908933218, "grad_norm": 0.22502829134464264, "learning_rate": 8.575729000435224e-05, "loss": 0.0214, "step": 4027 }, { "epoch": 0.43668690372940155, "grad_norm": 0.5550301671028137, "learning_rate": 8.575366313651531e-05, "loss": 0.052, "step": 4028 }, { "epoch": 0.4367953165654814, "grad_norm": 0.5299351811408997, "learning_rate": 8.575003626867837e-05, "loss": 0.0328, "step": 4029 }, { "epoch": 0.43690372940156114, "grad_norm": 0.09084856510162354, "learning_rate": 8.574640940084144e-05, "loss": 0.0067, "step": 4030 }, { "epoch": 0.43701214223764095, "grad_norm": 0.4052635729312897, "learning_rate": 8.574278253300451e-05, "loss": 0.0767, "step": 4031 }, { "epoch": 0.4371205550737207, "grad_norm": 0.18493413925170898, "learning_rate": 8.573915566516756e-05, "loss": 0.0172, "step": 4032 }, { "epoch": 0.43722896790980054, "grad_norm": 0.57831209897995, "learning_rate": 8.573552879733064e-05, "loss": 0.0368, "step": 4033 }, { "epoch": 0.4373373807458803, "grad_norm": 0.8223013877868652, "learning_rate": 8.573190192949369e-05, "loss": 0.0501, "step": 4034 }, { "epoch": 0.4374457935819601, "grad_norm": 0.4602910280227661, "learning_rate": 8.572827506165676e-05, "loss": 0.0246, "step": 4035 }, { "epoch": 0.4375542064180399, "grad_norm": 0.8403361439704895, "learning_rate": 8.572464819381982e-05, "loss": 0.0807, "step": 4036 }, { "epoch": 0.4376626192541197, "grad_norm": 0.35645362734794617, "learning_rate": 8.572102132598289e-05, "loss": 0.0351, "step": 4037 }, { "epoch": 0.43777103209019946, "grad_norm": 0.1486375331878662, "learning_rate": 8.571739445814595e-05, "loss": 0.0272, "step": 4038 }, { "epoch": 0.4378794449262793, "grad_norm": 0.4107159674167633, "learning_rate": 8.5713767590309e-05, "loss": 0.0257, "step": 4039 }, { "epoch": 0.43798785776235905, "grad_norm": 0.6124634146690369, "learning_rate": 8.571014072247208e-05, "loss": 0.0509, "step": 4040 }, { "epoch": 0.43809627059843886, "grad_norm": 0.40085142850875854, "learning_rate": 8.570651385463513e-05, "loss": 0.0434, "step": 4041 }, { "epoch": 0.4382046834345186, "grad_norm": 0.921410858631134, "learning_rate": 8.57028869867982e-05, "loss": 0.0513, "step": 4042 }, { "epoch": 0.43831309627059845, "grad_norm": 0.3117434084415436, "learning_rate": 8.569926011896126e-05, "loss": 0.0403, "step": 4043 }, { "epoch": 0.4384215091066782, "grad_norm": 0.5624102354049683, "learning_rate": 8.569563325112433e-05, "loss": 0.0556, "step": 4044 }, { "epoch": 0.438529921942758, "grad_norm": 0.48748132586479187, "learning_rate": 8.569200638328739e-05, "loss": 0.0228, "step": 4045 }, { "epoch": 0.4386383347788378, "grad_norm": 0.7095162272453308, "learning_rate": 8.568837951545046e-05, "loss": 0.077, "step": 4046 }, { "epoch": 0.4387467476149176, "grad_norm": 0.1740833967924118, "learning_rate": 8.568475264761352e-05, "loss": 0.0171, "step": 4047 }, { "epoch": 0.4388551604509974, "grad_norm": 0.4818199872970581, "learning_rate": 8.568112577977659e-05, "loss": 0.0452, "step": 4048 }, { "epoch": 0.4389635732870772, "grad_norm": 0.4771155118942261, "learning_rate": 8.567749891193966e-05, "loss": 0.0618, "step": 4049 }, { "epoch": 0.43907198612315695, "grad_norm": 0.2630375027656555, "learning_rate": 8.567387204410272e-05, "loss": 0.0183, "step": 4050 }, { "epoch": 0.4391803989592368, "grad_norm": 0.20045258104801178, "learning_rate": 8.567024517626579e-05, "loss": 0.0117, "step": 4051 }, { "epoch": 0.4392888117953166, "grad_norm": 0.6015576720237732, "learning_rate": 8.566661830842885e-05, "loss": 0.1026, "step": 4052 }, { "epoch": 0.43939722463139635, "grad_norm": 0.08151904493570328, "learning_rate": 8.566299144059192e-05, "loss": 0.0059, "step": 4053 }, { "epoch": 0.4395056374674762, "grad_norm": 0.7158002853393555, "learning_rate": 8.565936457275497e-05, "loss": 0.0384, "step": 4054 }, { "epoch": 0.43961405030355594, "grad_norm": 0.2951280176639557, "learning_rate": 8.565573770491803e-05, "loss": 0.029, "step": 4055 }, { "epoch": 0.43972246313963576, "grad_norm": 0.4414564371109009, "learning_rate": 8.56521108370811e-05, "loss": 0.0496, "step": 4056 }, { "epoch": 0.4398308759757155, "grad_norm": 0.377844899892807, "learning_rate": 8.564848396924416e-05, "loss": 0.0457, "step": 4057 }, { "epoch": 0.43993928881179534, "grad_norm": 0.4750308394432068, "learning_rate": 8.564485710140723e-05, "loss": 0.0582, "step": 4058 }, { "epoch": 0.4400477016478751, "grad_norm": 0.33055078983306885, "learning_rate": 8.564123023357029e-05, "loss": 0.0508, "step": 4059 }, { "epoch": 0.4401561144839549, "grad_norm": 0.2558012306690216, "learning_rate": 8.563760336573336e-05, "loss": 0.0419, "step": 4060 }, { "epoch": 0.4402645273200347, "grad_norm": 0.2845182418823242, "learning_rate": 8.563397649789642e-05, "loss": 0.0386, "step": 4061 }, { "epoch": 0.4403729401561145, "grad_norm": 0.4096927344799042, "learning_rate": 8.563034963005949e-05, "loss": 0.0609, "step": 4062 }, { "epoch": 0.44048135299219426, "grad_norm": 0.7176379561424255, "learning_rate": 8.562672276222255e-05, "loss": 0.0512, "step": 4063 }, { "epoch": 0.4405897658282741, "grad_norm": 0.44709670543670654, "learning_rate": 8.56230958943856e-05, "loss": 0.0473, "step": 4064 }, { "epoch": 0.44069817866435385, "grad_norm": 0.12844718992710114, "learning_rate": 8.561946902654869e-05, "loss": 0.0387, "step": 4065 }, { "epoch": 0.44080659150043366, "grad_norm": 0.44857513904571533, "learning_rate": 8.561584215871174e-05, "loss": 0.0449, "step": 4066 }, { "epoch": 0.4409150043365134, "grad_norm": 0.3449469208717346, "learning_rate": 8.561221529087481e-05, "loss": 0.0614, "step": 4067 }, { "epoch": 0.44102341717259325, "grad_norm": 0.43002545833587646, "learning_rate": 8.560858842303787e-05, "loss": 0.0567, "step": 4068 }, { "epoch": 0.441131830008673, "grad_norm": 0.3846532106399536, "learning_rate": 8.560496155520093e-05, "loss": 0.0764, "step": 4069 }, { "epoch": 0.44124024284475283, "grad_norm": 0.4150553047657013, "learning_rate": 8.5601334687364e-05, "loss": 0.032, "step": 4070 }, { "epoch": 0.4413486556808326, "grad_norm": 0.26388776302337646, "learning_rate": 8.559770781952706e-05, "loss": 0.0227, "step": 4071 }, { "epoch": 0.4414570685169124, "grad_norm": 0.21301984786987305, "learning_rate": 8.559408095169013e-05, "loss": 0.0286, "step": 4072 }, { "epoch": 0.4415654813529922, "grad_norm": 0.5360244512557983, "learning_rate": 8.559045408385319e-05, "loss": 0.0429, "step": 4073 }, { "epoch": 0.441673894189072, "grad_norm": 0.20291855931282043, "learning_rate": 8.558682721601626e-05, "loss": 0.042, "step": 4074 }, { "epoch": 0.44178230702515175, "grad_norm": 0.3679855465888977, "learning_rate": 8.558320034817931e-05, "loss": 0.046, "step": 4075 }, { "epoch": 0.4418907198612316, "grad_norm": 0.3664781451225281, "learning_rate": 8.557957348034239e-05, "loss": 0.018, "step": 4076 }, { "epoch": 0.44199913269731134, "grad_norm": 0.14520691335201263, "learning_rate": 8.557594661250544e-05, "loss": 0.0166, "step": 4077 }, { "epoch": 0.44210754553339116, "grad_norm": 0.3389027714729309, "learning_rate": 8.55723197446685e-05, "loss": 0.0491, "step": 4078 }, { "epoch": 0.4422159583694709, "grad_norm": 0.3223925828933716, "learning_rate": 8.556869287683157e-05, "loss": 0.047, "step": 4079 }, { "epoch": 0.44232437120555074, "grad_norm": 0.1808055341243744, "learning_rate": 8.556506600899463e-05, "loss": 0.0297, "step": 4080 }, { "epoch": 0.44243278404163056, "grad_norm": 0.4718661308288574, "learning_rate": 8.55614391411577e-05, "loss": 0.0451, "step": 4081 }, { "epoch": 0.4425411968777103, "grad_norm": 0.29737886786460876, "learning_rate": 8.555781227332076e-05, "loss": 0.018, "step": 4082 }, { "epoch": 0.44264960971379014, "grad_norm": 0.19872340559959412, "learning_rate": 8.555418540548383e-05, "loss": 0.0411, "step": 4083 }, { "epoch": 0.4427580225498699, "grad_norm": 0.49013853073120117, "learning_rate": 8.55505585376469e-05, "loss": 0.0418, "step": 4084 }, { "epoch": 0.4428664353859497, "grad_norm": 0.241975799202919, "learning_rate": 8.554693166980996e-05, "loss": 0.0202, "step": 4085 }, { "epoch": 0.4429748482220295, "grad_norm": 0.467851847410202, "learning_rate": 8.554330480197303e-05, "loss": 0.0558, "step": 4086 }, { "epoch": 0.4430832610581093, "grad_norm": 0.31674402952194214, "learning_rate": 8.553967793413608e-05, "loss": 0.0379, "step": 4087 }, { "epoch": 0.44319167389418906, "grad_norm": 0.31354624032974243, "learning_rate": 8.553605106629915e-05, "loss": 0.0468, "step": 4088 }, { "epoch": 0.4433000867302689, "grad_norm": 0.5462393164634705, "learning_rate": 8.553242419846221e-05, "loss": 0.0637, "step": 4089 }, { "epoch": 0.44340849956634865, "grad_norm": 0.3194388151168823, "learning_rate": 8.552879733062528e-05, "loss": 0.0385, "step": 4090 }, { "epoch": 0.44351691240242846, "grad_norm": 0.7715393304824829, "learning_rate": 8.552517046278834e-05, "loss": 0.0749, "step": 4091 }, { "epoch": 0.44362532523850823, "grad_norm": 0.3496835231781006, "learning_rate": 8.55215435949514e-05, "loss": 0.0336, "step": 4092 }, { "epoch": 0.44373373807458805, "grad_norm": 0.19617697596549988, "learning_rate": 8.551791672711447e-05, "loss": 0.0139, "step": 4093 }, { "epoch": 0.4438421509106678, "grad_norm": 0.36613887548446655, "learning_rate": 8.551428985927753e-05, "loss": 0.069, "step": 4094 }, { "epoch": 0.44395056374674763, "grad_norm": 0.3255968987941742, "learning_rate": 8.55106629914406e-05, "loss": 0.0504, "step": 4095 }, { "epoch": 0.4440589765828274, "grad_norm": 0.556951642036438, "learning_rate": 8.550703612360365e-05, "loss": 0.0218, "step": 4096 }, { "epoch": 0.4441673894189072, "grad_norm": 0.44371113181114197, "learning_rate": 8.550340925576672e-05, "loss": 0.0495, "step": 4097 }, { "epoch": 0.444275802254987, "grad_norm": 0.17161457240581512, "learning_rate": 8.549978238792978e-05, "loss": 0.0178, "step": 4098 }, { "epoch": 0.4443842150910668, "grad_norm": 0.1760113388299942, "learning_rate": 8.549615552009285e-05, "loss": 0.03, "step": 4099 }, { "epoch": 0.44449262792714656, "grad_norm": 0.42181599140167236, "learning_rate": 8.549252865225592e-05, "loss": 0.0378, "step": 4100 }, { "epoch": 0.4446010407632264, "grad_norm": 0.09217146039009094, "learning_rate": 8.548890178441898e-05, "loss": 0.0168, "step": 4101 }, { "epoch": 0.44470945359930614, "grad_norm": 0.33289870619773865, "learning_rate": 8.548527491658205e-05, "loss": 0.0434, "step": 4102 }, { "epoch": 0.44481786643538596, "grad_norm": 0.23678547143936157, "learning_rate": 8.548164804874511e-05, "loss": 0.041, "step": 4103 }, { "epoch": 0.4449262792714657, "grad_norm": 0.3215348422527313, "learning_rate": 8.547802118090818e-05, "loss": 0.0388, "step": 4104 }, { "epoch": 0.44503469210754554, "grad_norm": 0.22434912621974945, "learning_rate": 8.547439431307124e-05, "loss": 0.0244, "step": 4105 }, { "epoch": 0.4451431049436253, "grad_norm": 0.3200471103191376, "learning_rate": 8.54707674452343e-05, "loss": 0.0624, "step": 4106 }, { "epoch": 0.4452515177797051, "grad_norm": 0.07995064556598663, "learning_rate": 8.546714057739737e-05, "loss": 0.0186, "step": 4107 }, { "epoch": 0.4453599306157849, "grad_norm": 0.5927772521972656, "learning_rate": 8.546351370956042e-05, "loss": 0.0417, "step": 4108 }, { "epoch": 0.4454683434518647, "grad_norm": 0.619921863079071, "learning_rate": 8.54598868417235e-05, "loss": 0.0879, "step": 4109 }, { "epoch": 0.4455767562879445, "grad_norm": 0.5512534976005554, "learning_rate": 8.545625997388655e-05, "loss": 0.0496, "step": 4110 }, { "epoch": 0.4456851691240243, "grad_norm": 0.3345797061920166, "learning_rate": 8.545263310604962e-05, "loss": 0.0313, "step": 4111 }, { "epoch": 0.4457935819601041, "grad_norm": 0.3311845064163208, "learning_rate": 8.544900623821268e-05, "loss": 0.0342, "step": 4112 }, { "epoch": 0.44590199479618386, "grad_norm": 0.26701655983924866, "learning_rate": 8.544537937037575e-05, "loss": 0.0286, "step": 4113 }, { "epoch": 0.4460104076322637, "grad_norm": 0.6665899753570557, "learning_rate": 8.544175250253881e-05, "loss": 0.0324, "step": 4114 }, { "epoch": 0.44611882046834345, "grad_norm": 0.24588441848754883, "learning_rate": 8.543812563470186e-05, "loss": 0.0128, "step": 4115 }, { "epoch": 0.44622723330442327, "grad_norm": 0.35829418897628784, "learning_rate": 8.543449876686494e-05, "loss": 0.0309, "step": 4116 }, { "epoch": 0.44633564614050303, "grad_norm": 0.17166242003440857, "learning_rate": 8.5430871899028e-05, "loss": 0.0173, "step": 4117 }, { "epoch": 0.44644405897658285, "grad_norm": 0.7055026888847351, "learning_rate": 8.542724503119108e-05, "loss": 0.1108, "step": 4118 }, { "epoch": 0.4465524718126626, "grad_norm": 1.2069916725158691, "learning_rate": 8.542361816335413e-05, "loss": 0.1191, "step": 4119 }, { "epoch": 0.44666088464874243, "grad_norm": 0.2627407908439636, "learning_rate": 8.541999129551719e-05, "loss": 0.034, "step": 4120 }, { "epoch": 0.4467692974848222, "grad_norm": 0.448148638010025, "learning_rate": 8.541636442768026e-05, "loss": 0.0494, "step": 4121 }, { "epoch": 0.446877710320902, "grad_norm": 0.32246726751327515, "learning_rate": 8.541273755984332e-05, "loss": 0.0476, "step": 4122 }, { "epoch": 0.4469861231569818, "grad_norm": 0.1906174272298813, "learning_rate": 8.540911069200639e-05, "loss": 0.0244, "step": 4123 }, { "epoch": 0.4470945359930616, "grad_norm": 0.1635853350162506, "learning_rate": 8.540548382416945e-05, "loss": 0.0208, "step": 4124 }, { "epoch": 0.44720294882914136, "grad_norm": 0.19441524147987366, "learning_rate": 8.540185695633252e-05, "loss": 0.0374, "step": 4125 }, { "epoch": 0.4473113616652212, "grad_norm": 0.691192090511322, "learning_rate": 8.539823008849558e-05, "loss": 0.0662, "step": 4126 }, { "epoch": 0.44741977450130094, "grad_norm": 0.31999537348747253, "learning_rate": 8.539460322065865e-05, "loss": 0.0453, "step": 4127 }, { "epoch": 0.44752818733738076, "grad_norm": 0.3887704312801361, "learning_rate": 8.53909763528217e-05, "loss": 0.0777, "step": 4128 }, { "epoch": 0.4476366001734605, "grad_norm": 0.3949145972728729, "learning_rate": 8.538734948498476e-05, "loss": 0.048, "step": 4129 }, { "epoch": 0.44774501300954034, "grad_norm": 0.4715886116027832, "learning_rate": 8.538372261714783e-05, "loss": 0.052, "step": 4130 }, { "epoch": 0.4478534258456201, "grad_norm": 0.27944856882095337, "learning_rate": 8.538009574931089e-05, "loss": 0.0315, "step": 4131 }, { "epoch": 0.4479618386816999, "grad_norm": 0.1906326413154602, "learning_rate": 8.537646888147396e-05, "loss": 0.0179, "step": 4132 }, { "epoch": 0.4480702515177797, "grad_norm": 0.1360456496477127, "learning_rate": 8.537284201363702e-05, "loss": 0.0296, "step": 4133 }, { "epoch": 0.4481786643538595, "grad_norm": 0.5021013617515564, "learning_rate": 8.53692151458001e-05, "loss": 0.0253, "step": 4134 }, { "epoch": 0.44828707718993926, "grad_norm": 0.39821603894233704, "learning_rate": 8.536558827796316e-05, "loss": 0.0773, "step": 4135 }, { "epoch": 0.4483954900260191, "grad_norm": 0.35156017541885376, "learning_rate": 8.536196141012622e-05, "loss": 0.066, "step": 4136 }, { "epoch": 0.44850390286209885, "grad_norm": 0.3358915150165558, "learning_rate": 8.535833454228929e-05, "loss": 0.0336, "step": 4137 }, { "epoch": 0.44861231569817867, "grad_norm": 0.23437629640102386, "learning_rate": 8.535470767445235e-05, "loss": 0.0495, "step": 4138 }, { "epoch": 0.4487207285342585, "grad_norm": 0.20334693789482117, "learning_rate": 8.535108080661542e-05, "loss": 0.0473, "step": 4139 }, { "epoch": 0.44882914137033825, "grad_norm": 0.18530936539173126, "learning_rate": 8.534745393877847e-05, "loss": 0.0216, "step": 4140 }, { "epoch": 0.44893755420641807, "grad_norm": 0.29081130027770996, "learning_rate": 8.534382707094155e-05, "loss": 0.0636, "step": 4141 }, { "epoch": 0.44904596704249783, "grad_norm": 0.2646196484565735, "learning_rate": 8.53402002031046e-05, "loss": 0.0195, "step": 4142 }, { "epoch": 0.44915437987857765, "grad_norm": 0.29454556107521057, "learning_rate": 8.533657333526766e-05, "loss": 0.0179, "step": 4143 }, { "epoch": 0.4492627927146574, "grad_norm": 0.3935224115848541, "learning_rate": 8.533294646743073e-05, "loss": 0.0733, "step": 4144 }, { "epoch": 0.44937120555073723, "grad_norm": 0.5276819467544556, "learning_rate": 8.532931959959379e-05, "loss": 0.055, "step": 4145 }, { "epoch": 0.449479618386817, "grad_norm": 0.5620033740997314, "learning_rate": 8.532569273175686e-05, "loss": 0.0307, "step": 4146 }, { "epoch": 0.4495880312228968, "grad_norm": 0.28972166776657104, "learning_rate": 8.532206586391992e-05, "loss": 0.0609, "step": 4147 }, { "epoch": 0.4496964440589766, "grad_norm": 0.17146454751491547, "learning_rate": 8.531843899608299e-05, "loss": 0.0327, "step": 4148 }, { "epoch": 0.4498048568950564, "grad_norm": 0.38047724962234497, "learning_rate": 8.531481212824604e-05, "loss": 0.0765, "step": 4149 }, { "epoch": 0.44991326973113616, "grad_norm": 0.32074207067489624, "learning_rate": 8.531118526040912e-05, "loss": 0.0418, "step": 4150 }, { "epoch": 0.450021682567216, "grad_norm": 0.17276644706726074, "learning_rate": 8.530755839257217e-05, "loss": 0.0275, "step": 4151 }, { "epoch": 0.45013009540329574, "grad_norm": 0.9924317598342896, "learning_rate": 8.530393152473524e-05, "loss": 0.0464, "step": 4152 }, { "epoch": 0.45023850823937556, "grad_norm": 0.28578895330429077, "learning_rate": 8.530030465689831e-05, "loss": 0.0285, "step": 4153 }, { "epoch": 0.4503469210754553, "grad_norm": 0.20772330462932587, "learning_rate": 8.529667778906137e-05, "loss": 0.0429, "step": 4154 }, { "epoch": 0.45045533391153514, "grad_norm": 0.6418024301528931, "learning_rate": 8.529305092122444e-05, "loss": 0.1341, "step": 4155 }, { "epoch": 0.4505637467476149, "grad_norm": 0.850922703742981, "learning_rate": 8.52894240533875e-05, "loss": 0.1101, "step": 4156 }, { "epoch": 0.4506721595836947, "grad_norm": 0.2605677545070648, "learning_rate": 8.528579718555057e-05, "loss": 0.0261, "step": 4157 }, { "epoch": 0.4507805724197745, "grad_norm": 0.4215969443321228, "learning_rate": 8.528217031771363e-05, "loss": 0.0318, "step": 4158 }, { "epoch": 0.4508889852558543, "grad_norm": 0.12731365859508514, "learning_rate": 8.527854344987669e-05, "loss": 0.0223, "step": 4159 }, { "epoch": 0.45099739809193407, "grad_norm": 0.2672247588634491, "learning_rate": 8.527491658203976e-05, "loss": 0.0506, "step": 4160 }, { "epoch": 0.4511058109280139, "grad_norm": 0.40311798453330994, "learning_rate": 8.527128971420281e-05, "loss": 0.054, "step": 4161 }, { "epoch": 0.45121422376409365, "grad_norm": 0.2823902666568756, "learning_rate": 8.526766284636588e-05, "loss": 0.0353, "step": 4162 }, { "epoch": 0.45132263660017347, "grad_norm": 0.2190762758255005, "learning_rate": 8.526403597852894e-05, "loss": 0.0342, "step": 4163 }, { "epoch": 0.45143104943625323, "grad_norm": 0.32096341252326965, "learning_rate": 8.526040911069201e-05, "loss": 0.052, "step": 4164 }, { "epoch": 0.45153946227233305, "grad_norm": 0.3007563650608063, "learning_rate": 8.525678224285507e-05, "loss": 0.0661, "step": 4165 }, { "epoch": 0.4516478751084128, "grad_norm": 0.47781282663345337, "learning_rate": 8.525315537501814e-05, "loss": 0.0416, "step": 4166 }, { "epoch": 0.45175628794449263, "grad_norm": 0.3069694936275482, "learning_rate": 8.52495285071812e-05, "loss": 0.0386, "step": 4167 }, { "epoch": 0.4518647007805724, "grad_norm": 0.32571297883987427, "learning_rate": 8.524590163934426e-05, "loss": 0.0425, "step": 4168 }, { "epoch": 0.4519731136166522, "grad_norm": 0.11836902052164078, "learning_rate": 8.524227477150734e-05, "loss": 0.0123, "step": 4169 }, { "epoch": 0.45208152645273203, "grad_norm": 0.5023770928382874, "learning_rate": 8.52386479036704e-05, "loss": 0.1005, "step": 4170 }, { "epoch": 0.4521899392888118, "grad_norm": 1.2911903858184814, "learning_rate": 8.523502103583347e-05, "loss": 0.0665, "step": 4171 }, { "epoch": 0.4522983521248916, "grad_norm": 0.9342339038848877, "learning_rate": 8.523139416799653e-05, "loss": 0.0602, "step": 4172 }, { "epoch": 0.4524067649609714, "grad_norm": 0.17149953544139862, "learning_rate": 8.522776730015958e-05, "loss": 0.0272, "step": 4173 }, { "epoch": 0.4525151777970512, "grad_norm": 0.24223366379737854, "learning_rate": 8.522414043232265e-05, "loss": 0.0677, "step": 4174 }, { "epoch": 0.45262359063313096, "grad_norm": 0.18933968245983124, "learning_rate": 8.522051356448571e-05, "loss": 0.023, "step": 4175 }, { "epoch": 0.4527320034692108, "grad_norm": 0.43215829133987427, "learning_rate": 8.521688669664878e-05, "loss": 0.0696, "step": 4176 }, { "epoch": 0.45284041630529054, "grad_norm": 0.6493622064590454, "learning_rate": 8.521325982881184e-05, "loss": 0.0275, "step": 4177 }, { "epoch": 0.45294882914137036, "grad_norm": 0.24988721311092377, "learning_rate": 8.520963296097491e-05, "loss": 0.0298, "step": 4178 }, { "epoch": 0.4530572419774501, "grad_norm": 0.43301424384117126, "learning_rate": 8.520600609313797e-05, "loss": 0.0602, "step": 4179 }, { "epoch": 0.45316565481352994, "grad_norm": 0.29744312167167664, "learning_rate": 8.520237922530104e-05, "loss": 0.0352, "step": 4180 }, { "epoch": 0.4532740676496097, "grad_norm": 0.7441638112068176, "learning_rate": 8.51987523574641e-05, "loss": 0.0554, "step": 4181 }, { "epoch": 0.4533824804856895, "grad_norm": 0.17045654356479645, "learning_rate": 8.519512548962715e-05, "loss": 0.026, "step": 4182 }, { "epoch": 0.4534908933217693, "grad_norm": 0.4349575340747833, "learning_rate": 8.519149862179022e-05, "loss": 0.0883, "step": 4183 }, { "epoch": 0.4535993061578491, "grad_norm": 0.43738818168640137, "learning_rate": 8.518787175395328e-05, "loss": 0.0675, "step": 4184 }, { "epoch": 0.45370771899392887, "grad_norm": 0.28453177213668823, "learning_rate": 8.518424488611635e-05, "loss": 0.0398, "step": 4185 }, { "epoch": 0.4538161318300087, "grad_norm": 0.4588811695575714, "learning_rate": 8.518061801827942e-05, "loss": 0.0512, "step": 4186 }, { "epoch": 0.45392454466608845, "grad_norm": 0.29494160413742065, "learning_rate": 8.517699115044248e-05, "loss": 0.0554, "step": 4187 }, { "epoch": 0.45403295750216827, "grad_norm": 0.5588915348052979, "learning_rate": 8.517336428260555e-05, "loss": 0.0443, "step": 4188 }, { "epoch": 0.45414137033824803, "grad_norm": 0.24394603073596954, "learning_rate": 8.516973741476861e-05, "loss": 0.0279, "step": 4189 }, { "epoch": 0.45424978317432785, "grad_norm": 0.40250322222709656, "learning_rate": 8.516611054693168e-05, "loss": 0.0476, "step": 4190 }, { "epoch": 0.4543581960104076, "grad_norm": 0.45210087299346924, "learning_rate": 8.516248367909474e-05, "loss": 0.0867, "step": 4191 }, { "epoch": 0.45446660884648743, "grad_norm": 0.6930227875709534, "learning_rate": 8.515885681125781e-05, "loss": 0.0846, "step": 4192 }, { "epoch": 0.4545750216825672, "grad_norm": 0.28490278124809265, "learning_rate": 8.515522994342087e-05, "loss": 0.0521, "step": 4193 }, { "epoch": 0.454683434518647, "grad_norm": 0.6748343110084534, "learning_rate": 8.515160307558394e-05, "loss": 0.102, "step": 4194 }, { "epoch": 0.4547918473547268, "grad_norm": 0.2547931373119354, "learning_rate": 8.5147976207747e-05, "loss": 0.048, "step": 4195 }, { "epoch": 0.4549002601908066, "grad_norm": 0.2654092311859131, "learning_rate": 8.514434933991005e-05, "loss": 0.0192, "step": 4196 }, { "epoch": 0.45500867302688636, "grad_norm": 0.33237797021865845, "learning_rate": 8.514072247207312e-05, "loss": 0.0364, "step": 4197 }, { "epoch": 0.4551170858629662, "grad_norm": 0.22838440537452698, "learning_rate": 8.513709560423618e-05, "loss": 0.046, "step": 4198 }, { "epoch": 0.455225498699046, "grad_norm": 0.25945016741752625, "learning_rate": 8.513346873639925e-05, "loss": 0.0636, "step": 4199 }, { "epoch": 0.45533391153512576, "grad_norm": 0.4556787610054016, "learning_rate": 8.512984186856231e-05, "loss": 0.0726, "step": 4200 }, { "epoch": 0.4554423243712056, "grad_norm": 0.49491122364997864, "learning_rate": 8.512621500072538e-05, "loss": 0.0468, "step": 4201 }, { "epoch": 0.45555073720728534, "grad_norm": 0.37502020597457886, "learning_rate": 8.512258813288844e-05, "loss": 0.0257, "step": 4202 }, { "epoch": 0.45565915004336516, "grad_norm": 0.4411454200744629, "learning_rate": 8.51189612650515e-05, "loss": 0.0656, "step": 4203 }, { "epoch": 0.4557675628794449, "grad_norm": 0.2953217327594757, "learning_rate": 8.511533439721458e-05, "loss": 0.0401, "step": 4204 }, { "epoch": 0.45587597571552474, "grad_norm": 0.6753972172737122, "learning_rate": 8.511170752937763e-05, "loss": 0.0788, "step": 4205 }, { "epoch": 0.4559843885516045, "grad_norm": 0.2770872712135315, "learning_rate": 8.51080806615407e-05, "loss": 0.0447, "step": 4206 }, { "epoch": 0.4560928013876843, "grad_norm": 0.2141072154045105, "learning_rate": 8.510445379370376e-05, "loss": 0.0463, "step": 4207 }, { "epoch": 0.4562012142237641, "grad_norm": 0.1438177227973938, "learning_rate": 8.510082692586683e-05, "loss": 0.023, "step": 4208 }, { "epoch": 0.4563096270598439, "grad_norm": 0.2251461148262024, "learning_rate": 8.509720005802989e-05, "loss": 0.0414, "step": 4209 }, { "epoch": 0.45641803989592367, "grad_norm": 0.49140748381614685, "learning_rate": 8.509357319019295e-05, "loss": 0.0304, "step": 4210 }, { "epoch": 0.4565264527320035, "grad_norm": 0.23351497948169708, "learning_rate": 8.508994632235602e-05, "loss": 0.0391, "step": 4211 }, { "epoch": 0.45663486556808325, "grad_norm": 0.16035115718841553, "learning_rate": 8.508631945451908e-05, "loss": 0.0179, "step": 4212 }, { "epoch": 0.45674327840416307, "grad_norm": 0.33445999026298523, "learning_rate": 8.508269258668215e-05, "loss": 0.0468, "step": 4213 }, { "epoch": 0.45685169124024283, "grad_norm": 0.3540276885032654, "learning_rate": 8.50790657188452e-05, "loss": 0.0447, "step": 4214 }, { "epoch": 0.45696010407632265, "grad_norm": 0.29095038771629333, "learning_rate": 8.507543885100828e-05, "loss": 0.0478, "step": 4215 }, { "epoch": 0.4570685169124024, "grad_norm": 0.1808701604604721, "learning_rate": 8.507181198317133e-05, "loss": 0.0251, "step": 4216 }, { "epoch": 0.45717692974848223, "grad_norm": 0.18322475254535675, "learning_rate": 8.50681851153344e-05, "loss": 0.0266, "step": 4217 }, { "epoch": 0.457285342584562, "grad_norm": 0.31523340940475464, "learning_rate": 8.506455824749746e-05, "loss": 0.0399, "step": 4218 }, { "epoch": 0.4573937554206418, "grad_norm": 0.8370488286018372, "learning_rate": 8.506093137966052e-05, "loss": 0.0766, "step": 4219 }, { "epoch": 0.4575021682567216, "grad_norm": 0.5207453370094299, "learning_rate": 8.505730451182359e-05, "loss": 0.0426, "step": 4220 }, { "epoch": 0.4576105810928014, "grad_norm": 0.39313870668411255, "learning_rate": 8.505367764398666e-05, "loss": 0.0257, "step": 4221 }, { "epoch": 0.45771899392888116, "grad_norm": 0.21505704522132874, "learning_rate": 8.505005077614973e-05, "loss": 0.0416, "step": 4222 }, { "epoch": 0.457827406764961, "grad_norm": 0.35986506938934326, "learning_rate": 8.504642390831279e-05, "loss": 0.0386, "step": 4223 }, { "epoch": 0.45793581960104074, "grad_norm": 0.6519648432731628, "learning_rate": 8.504279704047585e-05, "loss": 0.0618, "step": 4224 }, { "epoch": 0.45804423243712056, "grad_norm": 0.28125274181365967, "learning_rate": 8.503917017263892e-05, "loss": 0.0513, "step": 4225 }, { "epoch": 0.4581526452732003, "grad_norm": 0.3347686529159546, "learning_rate": 8.503554330480197e-05, "loss": 0.0528, "step": 4226 }, { "epoch": 0.45826105810928014, "grad_norm": 0.3577844202518463, "learning_rate": 8.503191643696504e-05, "loss": 0.0273, "step": 4227 }, { "epoch": 0.45836947094535996, "grad_norm": 0.3211114704608917, "learning_rate": 8.50282895691281e-05, "loss": 0.0356, "step": 4228 }, { "epoch": 0.4584778837814397, "grad_norm": 0.3217271566390991, "learning_rate": 8.502466270129117e-05, "loss": 0.0584, "step": 4229 }, { "epoch": 0.45858629661751954, "grad_norm": 0.25921630859375, "learning_rate": 8.502103583345423e-05, "loss": 0.0499, "step": 4230 }, { "epoch": 0.4586947094535993, "grad_norm": 0.3445391058921814, "learning_rate": 8.50174089656173e-05, "loss": 0.0595, "step": 4231 }, { "epoch": 0.4588031222896791, "grad_norm": 0.3110892176628113, "learning_rate": 8.501378209778036e-05, "loss": 0.0507, "step": 4232 }, { "epoch": 0.4589115351257589, "grad_norm": 0.40083229541778564, "learning_rate": 8.501015522994342e-05, "loss": 0.0238, "step": 4233 }, { "epoch": 0.4590199479618387, "grad_norm": 0.3148486316204071, "learning_rate": 8.500652836210649e-05, "loss": 0.0724, "step": 4234 }, { "epoch": 0.45912836079791847, "grad_norm": 0.25253698229789734, "learning_rate": 8.500290149426954e-05, "loss": 0.0473, "step": 4235 }, { "epoch": 0.4592367736339983, "grad_norm": 0.24257194995880127, "learning_rate": 8.499927462643261e-05, "loss": 0.0126, "step": 4236 }, { "epoch": 0.45934518647007805, "grad_norm": 0.16573676466941833, "learning_rate": 8.499564775859567e-05, "loss": 0.0191, "step": 4237 }, { "epoch": 0.45945359930615787, "grad_norm": 0.11148770153522491, "learning_rate": 8.499202089075876e-05, "loss": 0.0217, "step": 4238 }, { "epoch": 0.45956201214223763, "grad_norm": 0.43114015460014343, "learning_rate": 8.498839402292181e-05, "loss": 0.0571, "step": 4239 }, { "epoch": 0.45967042497831745, "grad_norm": 0.3731444478034973, "learning_rate": 8.498476715508487e-05, "loss": 0.0418, "step": 4240 }, { "epoch": 0.4597788378143972, "grad_norm": 0.5988461971282959, "learning_rate": 8.498114028724794e-05, "loss": 0.1443, "step": 4241 }, { "epoch": 0.45988725065047703, "grad_norm": 0.18580108880996704, "learning_rate": 8.4977513419411e-05, "loss": 0.0245, "step": 4242 }, { "epoch": 0.4599956634865568, "grad_norm": 0.38073068857192993, "learning_rate": 8.497388655157407e-05, "loss": 0.0652, "step": 4243 }, { "epoch": 0.4601040763226366, "grad_norm": 0.1518864929676056, "learning_rate": 8.497025968373713e-05, "loss": 0.0189, "step": 4244 }, { "epoch": 0.4602124891587164, "grad_norm": 0.2959492802619934, "learning_rate": 8.49666328159002e-05, "loss": 0.0279, "step": 4245 }, { "epoch": 0.4603209019947962, "grad_norm": 0.26197558641433716, "learning_rate": 8.496300594806326e-05, "loss": 0.029, "step": 4246 }, { "epoch": 0.46042931483087596, "grad_norm": 0.26167431473731995, "learning_rate": 8.495937908022633e-05, "loss": 0.0299, "step": 4247 }, { "epoch": 0.4605377276669558, "grad_norm": 0.21900154650211334, "learning_rate": 8.495575221238938e-05, "loss": 0.0171, "step": 4248 }, { "epoch": 0.46064614050303554, "grad_norm": 0.6004929542541504, "learning_rate": 8.495212534455244e-05, "loss": 0.057, "step": 4249 }, { "epoch": 0.46075455333911536, "grad_norm": 0.33672967553138733, "learning_rate": 8.494849847671551e-05, "loss": 0.053, "step": 4250 }, { "epoch": 0.4608629661751951, "grad_norm": 0.27045920491218567, "learning_rate": 8.494487160887857e-05, "loss": 0.0291, "step": 4251 }, { "epoch": 0.46097137901127494, "grad_norm": 0.6619173288345337, "learning_rate": 8.494124474104164e-05, "loss": 0.0799, "step": 4252 }, { "epoch": 0.4610797918473547, "grad_norm": 0.43620237708091736, "learning_rate": 8.49376178732047e-05, "loss": 0.0635, "step": 4253 }, { "epoch": 0.4611882046834345, "grad_norm": 0.2683315873146057, "learning_rate": 8.493399100536777e-05, "loss": 0.0302, "step": 4254 }, { "epoch": 0.4612966175195143, "grad_norm": 0.30876290798187256, "learning_rate": 8.493036413753084e-05, "loss": 0.0473, "step": 4255 }, { "epoch": 0.4614050303555941, "grad_norm": 0.33546602725982666, "learning_rate": 8.49267372696939e-05, "loss": 0.0318, "step": 4256 }, { "epoch": 0.46151344319167387, "grad_norm": 0.25929418206214905, "learning_rate": 8.492311040185697e-05, "loss": 0.0387, "step": 4257 }, { "epoch": 0.4616218560277537, "grad_norm": 0.3892175853252411, "learning_rate": 8.491948353402003e-05, "loss": 0.0408, "step": 4258 }, { "epoch": 0.4617302688638335, "grad_norm": 0.9202831387519836, "learning_rate": 8.49158566661831e-05, "loss": 0.0791, "step": 4259 }, { "epoch": 0.46183868169991327, "grad_norm": 0.45920851826667786, "learning_rate": 8.491222979834615e-05, "loss": 0.0702, "step": 4260 }, { "epoch": 0.4619470945359931, "grad_norm": 0.5657903552055359, "learning_rate": 8.490860293050922e-05, "loss": 0.0646, "step": 4261 }, { "epoch": 0.46205550737207285, "grad_norm": 0.30390673875808716, "learning_rate": 8.490497606267228e-05, "loss": 0.0285, "step": 4262 }, { "epoch": 0.46216392020815267, "grad_norm": 0.6168684363365173, "learning_rate": 8.490134919483534e-05, "loss": 0.1021, "step": 4263 }, { "epoch": 0.46227233304423243, "grad_norm": 0.06338609755039215, "learning_rate": 8.489772232699841e-05, "loss": 0.0181, "step": 4264 }, { "epoch": 0.46238074588031225, "grad_norm": 0.1893874555826187, "learning_rate": 8.489409545916147e-05, "loss": 0.0248, "step": 4265 }, { "epoch": 0.462489158716392, "grad_norm": 0.3071732521057129, "learning_rate": 8.489046859132454e-05, "loss": 0.054, "step": 4266 }, { "epoch": 0.46259757155247183, "grad_norm": 0.4621378481388092, "learning_rate": 8.48868417234876e-05, "loss": 0.0582, "step": 4267 }, { "epoch": 0.4627059843885516, "grad_norm": 0.3876493573188782, "learning_rate": 8.488321485565067e-05, "loss": 0.0461, "step": 4268 }, { "epoch": 0.4628143972246314, "grad_norm": 0.4370889961719513, "learning_rate": 8.487958798781372e-05, "loss": 0.0314, "step": 4269 }, { "epoch": 0.4629228100607112, "grad_norm": 0.16808731853961945, "learning_rate": 8.48759611199768e-05, "loss": 0.0188, "step": 4270 }, { "epoch": 0.463031222896791, "grad_norm": 0.17476095259189606, "learning_rate": 8.487233425213985e-05, "loss": 0.0146, "step": 4271 }, { "epoch": 0.46313963573287076, "grad_norm": 0.1867215782403946, "learning_rate": 8.486870738430291e-05, "loss": 0.024, "step": 4272 }, { "epoch": 0.4632480485689506, "grad_norm": 0.554481029510498, "learning_rate": 8.4865080516466e-05, "loss": 0.0698, "step": 4273 }, { "epoch": 0.46335646140503034, "grad_norm": 0.25558435916900635, "learning_rate": 8.486145364862905e-05, "loss": 0.0375, "step": 4274 }, { "epoch": 0.46346487424111016, "grad_norm": 0.5796300172805786, "learning_rate": 8.485782678079212e-05, "loss": 0.0507, "step": 4275 }, { "epoch": 0.4635732870771899, "grad_norm": 0.585500180721283, "learning_rate": 8.485419991295518e-05, "loss": 0.0597, "step": 4276 }, { "epoch": 0.46368169991326974, "grad_norm": 0.10123475641012192, "learning_rate": 8.485057304511824e-05, "loss": 0.0187, "step": 4277 }, { "epoch": 0.4637901127493495, "grad_norm": 0.38703304529190063, "learning_rate": 8.484694617728131e-05, "loss": 0.0277, "step": 4278 }, { "epoch": 0.4638985255854293, "grad_norm": 0.6044986248016357, "learning_rate": 8.484331930944436e-05, "loss": 0.0973, "step": 4279 }, { "epoch": 0.4640069384215091, "grad_norm": 0.6716130971908569, "learning_rate": 8.483969244160744e-05, "loss": 0.0386, "step": 4280 }, { "epoch": 0.4641153512575889, "grad_norm": 0.512839138507843, "learning_rate": 8.483606557377049e-05, "loss": 0.0607, "step": 4281 }, { "epoch": 0.46422376409366867, "grad_norm": 0.5913105010986328, "learning_rate": 8.483243870593356e-05, "loss": 0.0522, "step": 4282 }, { "epoch": 0.4643321769297485, "grad_norm": 0.12485203146934509, "learning_rate": 8.482881183809662e-05, "loss": 0.0097, "step": 4283 }, { "epoch": 0.46444058976582825, "grad_norm": 0.41174575686454773, "learning_rate": 8.482518497025969e-05, "loss": 0.0586, "step": 4284 }, { "epoch": 0.46454900260190807, "grad_norm": 0.34753766655921936, "learning_rate": 8.482155810242275e-05, "loss": 0.0431, "step": 4285 }, { "epoch": 0.46465741543798783, "grad_norm": 0.2719433605670929, "learning_rate": 8.48179312345858e-05, "loss": 0.0609, "step": 4286 }, { "epoch": 0.46476582827406765, "grad_norm": 0.3366871476173401, "learning_rate": 8.481430436674888e-05, "loss": 0.0483, "step": 4287 }, { "epoch": 0.46487424111014747, "grad_norm": 0.3036992847919464, "learning_rate": 8.481067749891193e-05, "loss": 0.0351, "step": 4288 }, { "epoch": 0.46498265394622723, "grad_norm": 0.20483940839767456, "learning_rate": 8.4807050631075e-05, "loss": 0.0272, "step": 4289 }, { "epoch": 0.46509106678230705, "grad_norm": 0.6178062558174133, "learning_rate": 8.480342376323808e-05, "loss": 0.0762, "step": 4290 }, { "epoch": 0.4651994796183868, "grad_norm": 0.21222670376300812, "learning_rate": 8.479979689540113e-05, "loss": 0.0222, "step": 4291 }, { "epoch": 0.46530789245446663, "grad_norm": 0.2914944887161255, "learning_rate": 8.47961700275642e-05, "loss": 0.0227, "step": 4292 }, { "epoch": 0.4654163052905464, "grad_norm": 0.43247807025909424, "learning_rate": 8.479254315972726e-05, "loss": 0.0724, "step": 4293 }, { "epoch": 0.4655247181266262, "grad_norm": 0.2764831781387329, "learning_rate": 8.478891629189033e-05, "loss": 0.0306, "step": 4294 }, { "epoch": 0.465633130962706, "grad_norm": 0.1935051828622818, "learning_rate": 8.478528942405339e-05, "loss": 0.0186, "step": 4295 }, { "epoch": 0.4657415437987858, "grad_norm": 0.4326610267162323, "learning_rate": 8.478166255621646e-05, "loss": 0.0694, "step": 4296 }, { "epoch": 0.46584995663486556, "grad_norm": 0.2577207088470459, "learning_rate": 8.477803568837952e-05, "loss": 0.0236, "step": 4297 }, { "epoch": 0.4659583694709454, "grad_norm": 0.30958127975463867, "learning_rate": 8.477440882054259e-05, "loss": 0.0214, "step": 4298 }, { "epoch": 0.46606678230702514, "grad_norm": 0.3103061616420746, "learning_rate": 8.477078195270565e-05, "loss": 0.0464, "step": 4299 }, { "epoch": 0.46617519514310496, "grad_norm": 0.2743836045265198, "learning_rate": 8.47671550848687e-05, "loss": 0.0283, "step": 4300 }, { "epoch": 0.4662836079791847, "grad_norm": 0.46300750970840454, "learning_rate": 8.476352821703177e-05, "loss": 0.0434, "step": 4301 }, { "epoch": 0.46639202081526454, "grad_norm": 0.37066754698753357, "learning_rate": 8.475990134919483e-05, "loss": 0.0429, "step": 4302 }, { "epoch": 0.4665004336513443, "grad_norm": 0.40238329768180847, "learning_rate": 8.47562744813579e-05, "loss": 0.0431, "step": 4303 }, { "epoch": 0.4666088464874241, "grad_norm": 0.3258703947067261, "learning_rate": 8.475264761352096e-05, "loss": 0.033, "step": 4304 }, { "epoch": 0.4667172593235039, "grad_norm": 0.3735628128051758, "learning_rate": 8.474902074568403e-05, "loss": 0.0507, "step": 4305 }, { "epoch": 0.4668256721595837, "grad_norm": 0.7658980488777161, "learning_rate": 8.474539387784709e-05, "loss": 0.1086, "step": 4306 }, { "epoch": 0.46693408499566347, "grad_norm": 0.6550306677818298, "learning_rate": 8.474176701001016e-05, "loss": 0.0567, "step": 4307 }, { "epoch": 0.4670424978317433, "grad_norm": 0.23258039355278015, "learning_rate": 8.473814014217323e-05, "loss": 0.018, "step": 4308 }, { "epoch": 0.46715091066782305, "grad_norm": 0.3586238920688629, "learning_rate": 8.473451327433629e-05, "loss": 0.0605, "step": 4309 }, { "epoch": 0.46725932350390287, "grad_norm": 0.39529919624328613, "learning_rate": 8.473088640649936e-05, "loss": 0.0529, "step": 4310 }, { "epoch": 0.46736773633998263, "grad_norm": 0.500811755657196, "learning_rate": 8.472725953866242e-05, "loss": 0.046, "step": 4311 }, { "epoch": 0.46747614917606245, "grad_norm": 0.07806835323572159, "learning_rate": 8.472363267082549e-05, "loss": 0.0107, "step": 4312 }, { "epoch": 0.4675845620121422, "grad_norm": 0.21902801096439362, "learning_rate": 8.472000580298854e-05, "loss": 0.0146, "step": 4313 }, { "epoch": 0.46769297484822203, "grad_norm": 0.23929902911186218, "learning_rate": 8.47163789351516e-05, "loss": 0.031, "step": 4314 }, { "epoch": 0.4678013876843018, "grad_norm": 0.48215770721435547, "learning_rate": 8.471275206731467e-05, "loss": 0.0586, "step": 4315 }, { "epoch": 0.4679098005203816, "grad_norm": 0.20344799757003784, "learning_rate": 8.470912519947773e-05, "loss": 0.0196, "step": 4316 }, { "epoch": 0.46801821335646143, "grad_norm": 0.280662477016449, "learning_rate": 8.47054983316408e-05, "loss": 0.025, "step": 4317 }, { "epoch": 0.4681266261925412, "grad_norm": 0.5214265584945679, "learning_rate": 8.470187146380386e-05, "loss": 0.0351, "step": 4318 }, { "epoch": 0.468235039028621, "grad_norm": 0.3120885193347931, "learning_rate": 8.469824459596693e-05, "loss": 0.0265, "step": 4319 }, { "epoch": 0.4683434518647008, "grad_norm": 0.26570752263069153, "learning_rate": 8.469461772812999e-05, "loss": 0.0388, "step": 4320 }, { "epoch": 0.4684518647007806, "grad_norm": 0.3874374330043793, "learning_rate": 8.469099086029306e-05, "loss": 0.0375, "step": 4321 }, { "epoch": 0.46856027753686036, "grad_norm": 0.8956236243247986, "learning_rate": 8.468736399245611e-05, "loss": 0.0497, "step": 4322 }, { "epoch": 0.4686686903729402, "grad_norm": 0.1780468374490738, "learning_rate": 8.468373712461917e-05, "loss": 0.0209, "step": 4323 }, { "epoch": 0.46877710320901994, "grad_norm": 1.0340558290481567, "learning_rate": 8.468011025678226e-05, "loss": 0.0362, "step": 4324 }, { "epoch": 0.46888551604509976, "grad_norm": 0.36380329728126526, "learning_rate": 8.467648338894531e-05, "loss": 0.0357, "step": 4325 }, { "epoch": 0.4689939288811795, "grad_norm": 0.13956864178180695, "learning_rate": 8.467285652110838e-05, "loss": 0.0136, "step": 4326 }, { "epoch": 0.46910234171725934, "grad_norm": 0.6578590273857117, "learning_rate": 8.466922965327144e-05, "loss": 0.0867, "step": 4327 }, { "epoch": 0.4692107545533391, "grad_norm": 0.5097184181213379, "learning_rate": 8.46656027854345e-05, "loss": 0.0279, "step": 4328 }, { "epoch": 0.4693191673894189, "grad_norm": 0.7363327741622925, "learning_rate": 8.466197591759757e-05, "loss": 0.0764, "step": 4329 }, { "epoch": 0.4694275802254987, "grad_norm": 0.3353250026702881, "learning_rate": 8.465834904976063e-05, "loss": 0.0305, "step": 4330 }, { "epoch": 0.4695359930615785, "grad_norm": 0.2695058584213257, "learning_rate": 8.46547221819237e-05, "loss": 0.0276, "step": 4331 }, { "epoch": 0.46964440589765827, "grad_norm": 0.4134637713432312, "learning_rate": 8.465109531408676e-05, "loss": 0.0485, "step": 4332 }, { "epoch": 0.4697528187337381, "grad_norm": 0.323874831199646, "learning_rate": 8.464746844624983e-05, "loss": 0.0869, "step": 4333 }, { "epoch": 0.46986123156981785, "grad_norm": 0.18844564259052277, "learning_rate": 8.464384157841288e-05, "loss": 0.0162, "step": 4334 }, { "epoch": 0.46996964440589767, "grad_norm": 0.19500553607940674, "learning_rate": 8.464021471057595e-05, "loss": 0.0463, "step": 4335 }, { "epoch": 0.47007805724197743, "grad_norm": 0.5119425654411316, "learning_rate": 8.463658784273901e-05, "loss": 0.0639, "step": 4336 }, { "epoch": 0.47018647007805725, "grad_norm": 0.7334327697753906, "learning_rate": 8.463296097490207e-05, "loss": 0.0421, "step": 4337 }, { "epoch": 0.470294882914137, "grad_norm": 0.314724326133728, "learning_rate": 8.462933410706514e-05, "loss": 0.0502, "step": 4338 }, { "epoch": 0.47040329575021683, "grad_norm": 0.5908742547035217, "learning_rate": 8.46257072392282e-05, "loss": 0.0813, "step": 4339 }, { "epoch": 0.4705117085862966, "grad_norm": 0.2958660125732422, "learning_rate": 8.462208037139127e-05, "loss": 0.0361, "step": 4340 }, { "epoch": 0.4706201214223764, "grad_norm": 0.44438955187797546, "learning_rate": 8.461845350355433e-05, "loss": 0.0495, "step": 4341 }, { "epoch": 0.4707285342584562, "grad_norm": 0.2228957712650299, "learning_rate": 8.461482663571741e-05, "loss": 0.0418, "step": 4342 }, { "epoch": 0.470836947094536, "grad_norm": 0.34958651661872864, "learning_rate": 8.461119976788047e-05, "loss": 0.037, "step": 4343 }, { "epoch": 0.47094535993061576, "grad_norm": 0.6288406848907471, "learning_rate": 8.460757290004352e-05, "loss": 0.0758, "step": 4344 }, { "epoch": 0.4710537727666956, "grad_norm": 0.26983755826950073, "learning_rate": 8.46039460322066e-05, "loss": 0.0266, "step": 4345 }, { "epoch": 0.4711621856027754, "grad_norm": 0.3331318497657776, "learning_rate": 8.460031916436965e-05, "loss": 0.0358, "step": 4346 }, { "epoch": 0.47127059843885516, "grad_norm": 0.5864435434341431, "learning_rate": 8.459669229653272e-05, "loss": 0.0738, "step": 4347 }, { "epoch": 0.471379011274935, "grad_norm": 0.6070578098297119, "learning_rate": 8.459306542869578e-05, "loss": 0.0775, "step": 4348 }, { "epoch": 0.47148742411101474, "grad_norm": 0.5104352831840515, "learning_rate": 8.458943856085885e-05, "loss": 0.0376, "step": 4349 }, { "epoch": 0.47159583694709456, "grad_norm": 0.46931013464927673, "learning_rate": 8.458581169302191e-05, "loss": 0.0457, "step": 4350 }, { "epoch": 0.4717042497831743, "grad_norm": 0.4397044777870178, "learning_rate": 8.458218482518498e-05, "loss": 0.0546, "step": 4351 }, { "epoch": 0.47181266261925414, "grad_norm": 0.23205505311489105, "learning_rate": 8.457855795734804e-05, "loss": 0.0302, "step": 4352 }, { "epoch": 0.4719210754553339, "grad_norm": 0.5014005899429321, "learning_rate": 8.45749310895111e-05, "loss": 0.05, "step": 4353 }, { "epoch": 0.4720294882914137, "grad_norm": 0.3790452182292938, "learning_rate": 8.457130422167417e-05, "loss": 0.0443, "step": 4354 }, { "epoch": 0.4721379011274935, "grad_norm": 0.10306449979543686, "learning_rate": 8.456767735383722e-05, "loss": 0.02, "step": 4355 }, { "epoch": 0.4722463139635733, "grad_norm": 0.20902252197265625, "learning_rate": 8.45640504860003e-05, "loss": 0.0313, "step": 4356 }, { "epoch": 0.47235472679965307, "grad_norm": 0.05650416761636734, "learning_rate": 8.456042361816335e-05, "loss": 0.01, "step": 4357 }, { "epoch": 0.4724631396357329, "grad_norm": 0.17014741897583008, "learning_rate": 8.455679675032642e-05, "loss": 0.0256, "step": 4358 }, { "epoch": 0.47257155247181265, "grad_norm": 0.5353246331214905, "learning_rate": 8.455316988248949e-05, "loss": 0.0625, "step": 4359 }, { "epoch": 0.47267996530789247, "grad_norm": 0.3695594370365143, "learning_rate": 8.454954301465255e-05, "loss": 0.0595, "step": 4360 }, { "epoch": 0.47278837814397223, "grad_norm": 0.8014892339706421, "learning_rate": 8.454591614681562e-05, "loss": 0.1271, "step": 4361 }, { "epoch": 0.47289679098005205, "grad_norm": 0.1435050219297409, "learning_rate": 8.454228927897868e-05, "loss": 0.0209, "step": 4362 }, { "epoch": 0.4730052038161318, "grad_norm": 0.29985129833221436, "learning_rate": 8.453866241114175e-05, "loss": 0.0279, "step": 4363 }, { "epoch": 0.47311361665221163, "grad_norm": 0.6461267471313477, "learning_rate": 8.45350355433048e-05, "loss": 0.0344, "step": 4364 }, { "epoch": 0.4732220294882914, "grad_norm": 0.19410742819309235, "learning_rate": 8.453140867546788e-05, "loss": 0.0362, "step": 4365 }, { "epoch": 0.4733304423243712, "grad_norm": 0.8052530884742737, "learning_rate": 8.452778180763093e-05, "loss": 0.0647, "step": 4366 }, { "epoch": 0.473438855160451, "grad_norm": 0.3049134910106659, "learning_rate": 8.452415493979399e-05, "loss": 0.049, "step": 4367 }, { "epoch": 0.4735472679965308, "grad_norm": 0.4596753716468811, "learning_rate": 8.452052807195706e-05, "loss": 0.0198, "step": 4368 }, { "epoch": 0.47365568083261056, "grad_norm": 0.20375871658325195, "learning_rate": 8.451690120412012e-05, "loss": 0.0288, "step": 4369 }, { "epoch": 0.4737640936686904, "grad_norm": 0.4259602129459381, "learning_rate": 8.451327433628319e-05, "loss": 0.0674, "step": 4370 }, { "epoch": 0.47387250650477014, "grad_norm": 0.2883487045764923, "learning_rate": 8.450964746844625e-05, "loss": 0.0338, "step": 4371 }, { "epoch": 0.47398091934084996, "grad_norm": 0.780616819858551, "learning_rate": 8.450602060060932e-05, "loss": 0.0833, "step": 4372 }, { "epoch": 0.4740893321769297, "grad_norm": 0.5408708453178406, "learning_rate": 8.450239373277238e-05, "loss": 0.0631, "step": 4373 }, { "epoch": 0.47419774501300954, "grad_norm": 0.10766709595918655, "learning_rate": 8.449876686493545e-05, "loss": 0.0155, "step": 4374 }, { "epoch": 0.4743061578490893, "grad_norm": 0.26535114645957947, "learning_rate": 8.44951399970985e-05, "loss": 0.0306, "step": 4375 }, { "epoch": 0.4744145706851691, "grad_norm": 0.2823638319969177, "learning_rate": 8.449151312926158e-05, "loss": 0.0315, "step": 4376 }, { "epoch": 0.47452298352124894, "grad_norm": 0.42632031440734863, "learning_rate": 8.448788626142465e-05, "loss": 0.0386, "step": 4377 }, { "epoch": 0.4746313963573287, "grad_norm": 0.27583494782447815, "learning_rate": 8.44842593935877e-05, "loss": 0.027, "step": 4378 }, { "epoch": 0.4747398091934085, "grad_norm": 0.29971545934677124, "learning_rate": 8.448063252575078e-05, "loss": 0.0223, "step": 4379 }, { "epoch": 0.4748482220294883, "grad_norm": 0.4833117723464966, "learning_rate": 8.447700565791383e-05, "loss": 0.0371, "step": 4380 }, { "epoch": 0.4749566348655681, "grad_norm": 0.42832228541374207, "learning_rate": 8.447337879007689e-05, "loss": 0.0587, "step": 4381 }, { "epoch": 0.47506504770164787, "grad_norm": 0.26843512058258057, "learning_rate": 8.446975192223996e-05, "loss": 0.0292, "step": 4382 }, { "epoch": 0.4751734605377277, "grad_norm": 0.48775115609169006, "learning_rate": 8.446612505440302e-05, "loss": 0.0516, "step": 4383 }, { "epoch": 0.47528187337380745, "grad_norm": 0.08802639693021774, "learning_rate": 8.446249818656609e-05, "loss": 0.0087, "step": 4384 }, { "epoch": 0.47539028620988727, "grad_norm": 0.15991878509521484, "learning_rate": 8.445887131872915e-05, "loss": 0.0251, "step": 4385 }, { "epoch": 0.47549869904596703, "grad_norm": 0.551357090473175, "learning_rate": 8.445524445089222e-05, "loss": 0.0545, "step": 4386 }, { "epoch": 0.47560711188204685, "grad_norm": 0.44348639249801636, "learning_rate": 8.445161758305527e-05, "loss": 0.0481, "step": 4387 }, { "epoch": 0.4757155247181266, "grad_norm": 0.3240358829498291, "learning_rate": 8.444799071521835e-05, "loss": 0.0516, "step": 4388 }, { "epoch": 0.47582393755420643, "grad_norm": 0.9196059107780457, "learning_rate": 8.44443638473814e-05, "loss": 0.0625, "step": 4389 }, { "epoch": 0.4759323503902862, "grad_norm": 0.32382431626319885, "learning_rate": 8.444073697954446e-05, "loss": 0.0312, "step": 4390 }, { "epoch": 0.476040763226366, "grad_norm": 0.6086357831954956, "learning_rate": 8.443711011170753e-05, "loss": 0.0639, "step": 4391 }, { "epoch": 0.4761491760624458, "grad_norm": 0.2799977660179138, "learning_rate": 8.443348324387059e-05, "loss": 0.0219, "step": 4392 }, { "epoch": 0.4762575888985256, "grad_norm": 0.5530656576156616, "learning_rate": 8.442985637603366e-05, "loss": 0.1123, "step": 4393 }, { "epoch": 0.47636600173460536, "grad_norm": 0.44763797521591187, "learning_rate": 8.442622950819673e-05, "loss": 0.0491, "step": 4394 }, { "epoch": 0.4764744145706852, "grad_norm": 0.42627742886543274, "learning_rate": 8.442260264035979e-05, "loss": 0.023, "step": 4395 }, { "epoch": 0.47658282740676494, "grad_norm": 0.5666998028755188, "learning_rate": 8.441897577252286e-05, "loss": 0.0791, "step": 4396 }, { "epoch": 0.47669124024284476, "grad_norm": 0.22934490442276, "learning_rate": 8.441534890468592e-05, "loss": 0.0198, "step": 4397 }, { "epoch": 0.4767996530789245, "grad_norm": 0.5205461978912354, "learning_rate": 8.441172203684899e-05, "loss": 0.0314, "step": 4398 }, { "epoch": 0.47690806591500434, "grad_norm": 0.1519308239221573, "learning_rate": 8.440809516901204e-05, "loss": 0.0184, "step": 4399 }, { "epoch": 0.4770164787510841, "grad_norm": 0.1433185636997223, "learning_rate": 8.440446830117511e-05, "loss": 0.0085, "step": 4400 }, { "epoch": 0.4771248915871639, "grad_norm": 0.19299694895744324, "learning_rate": 8.440084143333817e-05, "loss": 0.0146, "step": 4401 }, { "epoch": 0.4772333044232437, "grad_norm": 0.3994062542915344, "learning_rate": 8.439721456550124e-05, "loss": 0.0312, "step": 4402 }, { "epoch": 0.4773417172593235, "grad_norm": 0.3933161795139313, "learning_rate": 8.43935876976643e-05, "loss": 0.0475, "step": 4403 }, { "epoch": 0.47745013009540327, "grad_norm": 0.22242915630340576, "learning_rate": 8.438996082982736e-05, "loss": 0.0446, "step": 4404 }, { "epoch": 0.4775585429314831, "grad_norm": 0.19023513793945312, "learning_rate": 8.438633396199043e-05, "loss": 0.0208, "step": 4405 }, { "epoch": 0.4776669557675629, "grad_norm": 0.12060050666332245, "learning_rate": 8.438270709415349e-05, "loss": 0.0208, "step": 4406 }, { "epoch": 0.47777536860364267, "grad_norm": 0.7759896516799927, "learning_rate": 8.437908022631656e-05, "loss": 0.0341, "step": 4407 }, { "epoch": 0.4778837814397225, "grad_norm": 0.5019484162330627, "learning_rate": 8.437545335847961e-05, "loss": 0.0438, "step": 4408 }, { "epoch": 0.47799219427580225, "grad_norm": 0.42369577288627625, "learning_rate": 8.437182649064268e-05, "loss": 0.0471, "step": 4409 }, { "epoch": 0.47810060711188207, "grad_norm": 0.5282171964645386, "learning_rate": 8.436819962280574e-05, "loss": 0.0581, "step": 4410 }, { "epoch": 0.47820901994796183, "grad_norm": 0.5248499512672424, "learning_rate": 8.436457275496881e-05, "loss": 0.0258, "step": 4411 }, { "epoch": 0.47831743278404165, "grad_norm": 0.38142314553260803, "learning_rate": 8.436094588713188e-05, "loss": 0.0493, "step": 4412 }, { "epoch": 0.4784258456201214, "grad_norm": 0.6359621286392212, "learning_rate": 8.435731901929494e-05, "loss": 0.0464, "step": 4413 }, { "epoch": 0.47853425845620123, "grad_norm": 0.2555594742298126, "learning_rate": 8.435369215145801e-05, "loss": 0.0493, "step": 4414 }, { "epoch": 0.478642671292281, "grad_norm": 0.34273001551628113, "learning_rate": 8.435006528362107e-05, "loss": 0.0446, "step": 4415 }, { "epoch": 0.4787510841283608, "grad_norm": 0.23897427320480347, "learning_rate": 8.434643841578414e-05, "loss": 0.0389, "step": 4416 }, { "epoch": 0.4788594969644406, "grad_norm": 0.18062560260295868, "learning_rate": 8.43428115479472e-05, "loss": 0.0311, "step": 4417 }, { "epoch": 0.4789679098005204, "grad_norm": 0.2601913511753082, "learning_rate": 8.433918468011025e-05, "loss": 0.0408, "step": 4418 }, { "epoch": 0.47907632263660016, "grad_norm": 0.318699449300766, "learning_rate": 8.433555781227333e-05, "loss": 0.0235, "step": 4419 }, { "epoch": 0.47918473547268, "grad_norm": 0.3557179868221283, "learning_rate": 8.433193094443638e-05, "loss": 0.0575, "step": 4420 }, { "epoch": 0.47929314830875974, "grad_norm": 0.5224906206130981, "learning_rate": 8.432830407659945e-05, "loss": 0.0623, "step": 4421 }, { "epoch": 0.47940156114483956, "grad_norm": 0.7573286294937134, "learning_rate": 8.432467720876251e-05, "loss": 0.055, "step": 4422 }, { "epoch": 0.4795099739809193, "grad_norm": 0.4405449628829956, "learning_rate": 8.432105034092558e-05, "loss": 0.0823, "step": 4423 }, { "epoch": 0.47961838681699914, "grad_norm": 0.2025834619998932, "learning_rate": 8.431742347308864e-05, "loss": 0.026, "step": 4424 }, { "epoch": 0.4797267996530789, "grad_norm": 0.2964954376220703, "learning_rate": 8.431379660525171e-05, "loss": 0.0198, "step": 4425 }, { "epoch": 0.4798352124891587, "grad_norm": 0.358963280916214, "learning_rate": 8.431016973741477e-05, "loss": 0.0277, "step": 4426 }, { "epoch": 0.4799436253252385, "grad_norm": 0.21915490925312042, "learning_rate": 8.430654286957783e-05, "loss": 0.0175, "step": 4427 }, { "epoch": 0.4800520381613183, "grad_norm": 0.4174489676952362, "learning_rate": 8.430291600174091e-05, "loss": 0.0223, "step": 4428 }, { "epoch": 0.48016045099739807, "grad_norm": 0.5356245040893555, "learning_rate": 8.429928913390397e-05, "loss": 0.0644, "step": 4429 }, { "epoch": 0.4802688638334779, "grad_norm": 0.2063075304031372, "learning_rate": 8.429566226606704e-05, "loss": 0.0207, "step": 4430 }, { "epoch": 0.48037727666955765, "grad_norm": 0.6597917675971985, "learning_rate": 8.42920353982301e-05, "loss": 0.0793, "step": 4431 }, { "epoch": 0.48048568950563747, "grad_norm": 0.5449238419532776, "learning_rate": 8.428840853039317e-05, "loss": 0.0664, "step": 4432 }, { "epoch": 0.48059410234171723, "grad_norm": 0.16118434071540833, "learning_rate": 8.428478166255622e-05, "loss": 0.0242, "step": 4433 }, { "epoch": 0.48070251517779705, "grad_norm": 0.20208214223384857, "learning_rate": 8.428115479471928e-05, "loss": 0.0336, "step": 4434 }, { "epoch": 0.48081092801387687, "grad_norm": 0.453362375497818, "learning_rate": 8.427752792688235e-05, "loss": 0.0503, "step": 4435 }, { "epoch": 0.48091934084995663, "grad_norm": 0.6244055032730103, "learning_rate": 8.427390105904541e-05, "loss": 0.122, "step": 4436 }, { "epoch": 0.48102775368603645, "grad_norm": 0.20928636193275452, "learning_rate": 8.427027419120848e-05, "loss": 0.0261, "step": 4437 }, { "epoch": 0.4811361665221162, "grad_norm": 0.35385048389434814, "learning_rate": 8.426664732337154e-05, "loss": 0.0261, "step": 4438 }, { "epoch": 0.48124457935819603, "grad_norm": 0.3240225315093994, "learning_rate": 8.426302045553461e-05, "loss": 0.0527, "step": 4439 }, { "epoch": 0.4813529921942758, "grad_norm": 0.666013240814209, "learning_rate": 8.425939358769767e-05, "loss": 0.0568, "step": 4440 }, { "epoch": 0.4814614050303556, "grad_norm": 0.20589207112789154, "learning_rate": 8.425576671986072e-05, "loss": 0.0255, "step": 4441 }, { "epoch": 0.4815698178664354, "grad_norm": 0.2381790578365326, "learning_rate": 8.42521398520238e-05, "loss": 0.0219, "step": 4442 }, { "epoch": 0.4816782307025152, "grad_norm": 0.2766909599304199, "learning_rate": 8.424851298418685e-05, "loss": 0.0226, "step": 4443 }, { "epoch": 0.48178664353859496, "grad_norm": 0.5800973176956177, "learning_rate": 8.424488611634992e-05, "loss": 0.036, "step": 4444 }, { "epoch": 0.4818950563746748, "grad_norm": 0.2576400637626648, "learning_rate": 8.424125924851299e-05, "loss": 0.0254, "step": 4445 }, { "epoch": 0.48200346921075454, "grad_norm": 0.16252917051315308, "learning_rate": 8.423763238067606e-05, "loss": 0.0127, "step": 4446 }, { "epoch": 0.48211188204683436, "grad_norm": 0.4080617427825928, "learning_rate": 8.423400551283912e-05, "loss": 0.0458, "step": 4447 }, { "epoch": 0.4822202948829141, "grad_norm": 0.5192016363143921, "learning_rate": 8.423037864500218e-05, "loss": 0.0408, "step": 4448 }, { "epoch": 0.48232870771899394, "grad_norm": 0.7250443696975708, "learning_rate": 8.422675177716525e-05, "loss": 0.0388, "step": 4449 }, { "epoch": 0.4824371205550737, "grad_norm": 0.261896550655365, "learning_rate": 8.42231249093283e-05, "loss": 0.0245, "step": 4450 }, { "epoch": 0.4825455333911535, "grad_norm": 1.1875059604644775, "learning_rate": 8.421949804149138e-05, "loss": 0.0731, "step": 4451 }, { "epoch": 0.4826539462272333, "grad_norm": 0.3171429932117462, "learning_rate": 8.421587117365443e-05, "loss": 0.0305, "step": 4452 }, { "epoch": 0.4827623590633131, "grad_norm": 0.2887498140335083, "learning_rate": 8.42122443058175e-05, "loss": 0.0191, "step": 4453 }, { "epoch": 0.48287077189939287, "grad_norm": 0.3107839822769165, "learning_rate": 8.420861743798056e-05, "loss": 0.0446, "step": 4454 }, { "epoch": 0.4829791847354727, "grad_norm": 0.45936471223831177, "learning_rate": 8.420499057014363e-05, "loss": 0.035, "step": 4455 }, { "epoch": 0.48308759757155245, "grad_norm": 0.4142511487007141, "learning_rate": 8.420136370230669e-05, "loss": 0.0346, "step": 4456 }, { "epoch": 0.48319601040763227, "grad_norm": 0.24797117710113525, "learning_rate": 8.419773683446975e-05, "loss": 0.0696, "step": 4457 }, { "epoch": 0.48330442324371203, "grad_norm": 0.2221348136663437, "learning_rate": 8.419410996663282e-05, "loss": 0.0402, "step": 4458 }, { "epoch": 0.48341283607979185, "grad_norm": 0.2581706941127777, "learning_rate": 8.419048309879588e-05, "loss": 0.0324, "step": 4459 }, { "epoch": 0.4835212489158716, "grad_norm": 0.5258331894874573, "learning_rate": 8.418685623095895e-05, "loss": 0.0751, "step": 4460 }, { "epoch": 0.48362966175195143, "grad_norm": 0.45409080386161804, "learning_rate": 8.4183229363122e-05, "loss": 0.1053, "step": 4461 }, { "epoch": 0.4837380745880312, "grad_norm": 0.450484037399292, "learning_rate": 8.417960249528508e-05, "loss": 0.0659, "step": 4462 }, { "epoch": 0.483846487424111, "grad_norm": 0.22675347328186035, "learning_rate": 8.417597562744815e-05, "loss": 0.0448, "step": 4463 }, { "epoch": 0.4839549002601908, "grad_norm": 0.4532434940338135, "learning_rate": 8.41723487596112e-05, "loss": 0.0854, "step": 4464 }, { "epoch": 0.4840633130962706, "grad_norm": 0.09071587771177292, "learning_rate": 8.416872189177427e-05, "loss": 0.0144, "step": 4465 }, { "epoch": 0.4841717259323504, "grad_norm": 0.5345737338066101, "learning_rate": 8.416509502393733e-05, "loss": 0.0399, "step": 4466 }, { "epoch": 0.4842801387684302, "grad_norm": 0.2155836522579193, "learning_rate": 8.41614681561004e-05, "loss": 0.0304, "step": 4467 }, { "epoch": 0.48438855160451, "grad_norm": 0.3148610591888428, "learning_rate": 8.415784128826346e-05, "loss": 0.0426, "step": 4468 }, { "epoch": 0.48449696444058976, "grad_norm": 0.31245556473731995, "learning_rate": 8.415421442042653e-05, "loss": 0.0445, "step": 4469 }, { "epoch": 0.4846053772766696, "grad_norm": 0.21645955741405487, "learning_rate": 8.415058755258959e-05, "loss": 0.0247, "step": 4470 }, { "epoch": 0.48471379011274934, "grad_norm": 0.27201011776924133, "learning_rate": 8.414696068475265e-05, "loss": 0.0371, "step": 4471 }, { "epoch": 0.48482220294882916, "grad_norm": 0.5530888438224792, "learning_rate": 8.414333381691572e-05, "loss": 0.0263, "step": 4472 }, { "epoch": 0.4849306157849089, "grad_norm": 0.41856691241264343, "learning_rate": 8.413970694907877e-05, "loss": 0.0775, "step": 4473 }, { "epoch": 0.48503902862098874, "grad_norm": 0.49442023038864136, "learning_rate": 8.413608008124184e-05, "loss": 0.0687, "step": 4474 }, { "epoch": 0.4851474414570685, "grad_norm": 0.31678101420402527, "learning_rate": 8.41324532134049e-05, "loss": 0.0319, "step": 4475 }, { "epoch": 0.4852558542931483, "grad_norm": 0.4011489450931549, "learning_rate": 8.412882634556797e-05, "loss": 0.0801, "step": 4476 }, { "epoch": 0.4853642671292281, "grad_norm": 0.3331425189971924, "learning_rate": 8.412519947773103e-05, "loss": 0.0379, "step": 4477 }, { "epoch": 0.4854726799653079, "grad_norm": 0.5088639855384827, "learning_rate": 8.41215726098941e-05, "loss": 0.0463, "step": 4478 }, { "epoch": 0.48558109280138767, "grad_norm": 0.25245553255081177, "learning_rate": 8.411794574205716e-05, "loss": 0.0408, "step": 4479 }, { "epoch": 0.4856895056374675, "grad_norm": 0.4811916649341583, "learning_rate": 8.411431887422023e-05, "loss": 0.0376, "step": 4480 }, { "epoch": 0.48579791847354725, "grad_norm": 0.1166800782084465, "learning_rate": 8.41106920063833e-05, "loss": 0.0105, "step": 4481 }, { "epoch": 0.48590633130962707, "grad_norm": 0.38613051176071167, "learning_rate": 8.410706513854636e-05, "loss": 0.0409, "step": 4482 }, { "epoch": 0.48601474414570683, "grad_norm": 1.1999120712280273, "learning_rate": 8.410343827070943e-05, "loss": 0.2679, "step": 4483 }, { "epoch": 0.48612315698178665, "grad_norm": 0.39704248309135437, "learning_rate": 8.409981140287249e-05, "loss": 0.0641, "step": 4484 }, { "epoch": 0.4862315698178664, "grad_norm": 0.2159249186515808, "learning_rate": 8.409618453503554e-05, "loss": 0.0269, "step": 4485 }, { "epoch": 0.48633998265394623, "grad_norm": 0.7466654777526855, "learning_rate": 8.409255766719861e-05, "loss": 0.1039, "step": 4486 }, { "epoch": 0.486448395490026, "grad_norm": 0.18132497370243073, "learning_rate": 8.408893079936167e-05, "loss": 0.0292, "step": 4487 }, { "epoch": 0.4865568083261058, "grad_norm": 0.30094149708747864, "learning_rate": 8.408530393152474e-05, "loss": 0.0602, "step": 4488 }, { "epoch": 0.4866652211621856, "grad_norm": 0.29565343260765076, "learning_rate": 8.40816770636878e-05, "loss": 0.0557, "step": 4489 }, { "epoch": 0.4867736339982654, "grad_norm": 0.22262319922447205, "learning_rate": 8.407805019585087e-05, "loss": 0.0216, "step": 4490 }, { "epoch": 0.48688204683434516, "grad_norm": 0.3137255907058716, "learning_rate": 8.407442332801393e-05, "loss": 0.0396, "step": 4491 }, { "epoch": 0.486990459670425, "grad_norm": 0.8918955326080322, "learning_rate": 8.4070796460177e-05, "loss": 0.0287, "step": 4492 }, { "epoch": 0.48709887250650474, "grad_norm": 0.6359989643096924, "learning_rate": 8.406716959234006e-05, "loss": 0.0254, "step": 4493 }, { "epoch": 0.48720728534258456, "grad_norm": 0.2777406871318817, "learning_rate": 8.406354272450311e-05, "loss": 0.0548, "step": 4494 }, { "epoch": 0.4873156981786644, "grad_norm": 0.2517516016960144, "learning_rate": 8.405991585666618e-05, "loss": 0.0157, "step": 4495 }, { "epoch": 0.48742411101474414, "grad_norm": 0.33770865201950073, "learning_rate": 8.405628898882924e-05, "loss": 0.0599, "step": 4496 }, { "epoch": 0.48753252385082396, "grad_norm": 0.4067402482032776, "learning_rate": 8.405266212099233e-05, "loss": 0.044, "step": 4497 }, { "epoch": 0.4876409366869037, "grad_norm": 0.38145875930786133, "learning_rate": 8.404903525315538e-05, "loss": 0.0354, "step": 4498 }, { "epoch": 0.48774934952298354, "grad_norm": 0.4342574179172516, "learning_rate": 8.404540838531844e-05, "loss": 0.0458, "step": 4499 }, { "epoch": 0.4878577623590633, "grad_norm": 0.2599869668483734, "learning_rate": 8.404178151748151e-05, "loss": 0.0458, "step": 4500 }, { "epoch": 0.4879661751951431, "grad_norm": 0.4605432152748108, "learning_rate": 8.403815464964457e-05, "loss": 0.0445, "step": 4501 }, { "epoch": 0.4880745880312229, "grad_norm": 0.4077160954475403, "learning_rate": 8.403452778180764e-05, "loss": 0.037, "step": 4502 }, { "epoch": 0.4881830008673027, "grad_norm": 0.24990440905094147, "learning_rate": 8.40309009139707e-05, "loss": 0.0227, "step": 4503 }, { "epoch": 0.48829141370338247, "grad_norm": 0.3229352831840515, "learning_rate": 8.402727404613377e-05, "loss": 0.0232, "step": 4504 }, { "epoch": 0.4883998265394623, "grad_norm": 0.49858322739601135, "learning_rate": 8.402364717829683e-05, "loss": 0.0304, "step": 4505 }, { "epoch": 0.48850823937554205, "grad_norm": 0.20180727541446686, "learning_rate": 8.40200203104599e-05, "loss": 0.0206, "step": 4506 }, { "epoch": 0.48861665221162187, "grad_norm": 0.356315016746521, "learning_rate": 8.401639344262295e-05, "loss": 0.0204, "step": 4507 }, { "epoch": 0.48872506504770163, "grad_norm": 0.8644310832023621, "learning_rate": 8.401276657478601e-05, "loss": 0.0732, "step": 4508 }, { "epoch": 0.48883347788378145, "grad_norm": 0.20588789880275726, "learning_rate": 8.400913970694908e-05, "loss": 0.0139, "step": 4509 }, { "epoch": 0.4889418907198612, "grad_norm": 0.826055645942688, "learning_rate": 8.400551283911214e-05, "loss": 0.0616, "step": 4510 }, { "epoch": 0.48905030355594103, "grad_norm": 0.6343889832496643, "learning_rate": 8.400188597127521e-05, "loss": 0.0449, "step": 4511 }, { "epoch": 0.4891587163920208, "grad_norm": 0.6751546263694763, "learning_rate": 8.399825910343827e-05, "loss": 0.1154, "step": 4512 }, { "epoch": 0.4892671292281006, "grad_norm": 0.7817211151123047, "learning_rate": 8.399463223560134e-05, "loss": 0.063, "step": 4513 }, { "epoch": 0.4893755420641804, "grad_norm": 0.2859460115432739, "learning_rate": 8.399100536776441e-05, "loss": 0.0379, "step": 4514 }, { "epoch": 0.4894839549002602, "grad_norm": 0.47875916957855225, "learning_rate": 8.398737849992747e-05, "loss": 0.0304, "step": 4515 }, { "epoch": 0.48959236773633996, "grad_norm": 0.3372167944908142, "learning_rate": 8.398375163209054e-05, "loss": 0.0455, "step": 4516 }, { "epoch": 0.4897007805724198, "grad_norm": 0.43185824155807495, "learning_rate": 8.39801247642536e-05, "loss": 0.0462, "step": 4517 }, { "epoch": 0.48980919340849954, "grad_norm": 0.618973433971405, "learning_rate": 8.397649789641667e-05, "loss": 0.0412, "step": 4518 }, { "epoch": 0.48991760624457936, "grad_norm": 0.32470637559890747, "learning_rate": 8.397287102857972e-05, "loss": 0.0358, "step": 4519 }, { "epoch": 0.4900260190806591, "grad_norm": 0.5689901113510132, "learning_rate": 8.39692441607428e-05, "loss": 0.0887, "step": 4520 }, { "epoch": 0.49013443191673894, "grad_norm": 0.8916002511978149, "learning_rate": 8.396561729290585e-05, "loss": 0.0387, "step": 4521 }, { "epoch": 0.4902428447528187, "grad_norm": 0.30844923853874207, "learning_rate": 8.396199042506891e-05, "loss": 0.0475, "step": 4522 }, { "epoch": 0.4903512575888985, "grad_norm": 0.4596771001815796, "learning_rate": 8.395836355723198e-05, "loss": 0.0386, "step": 4523 }, { "epoch": 0.49045967042497834, "grad_norm": 0.25807780027389526, "learning_rate": 8.395473668939504e-05, "loss": 0.0367, "step": 4524 }, { "epoch": 0.4905680832610581, "grad_norm": 0.3795608878135681, "learning_rate": 8.395110982155811e-05, "loss": 0.0501, "step": 4525 }, { "epoch": 0.4906764960971379, "grad_norm": 0.487235426902771, "learning_rate": 8.394748295372116e-05, "loss": 0.0459, "step": 4526 }, { "epoch": 0.4907849089332177, "grad_norm": 0.24642503261566162, "learning_rate": 8.394385608588424e-05, "loss": 0.0276, "step": 4527 }, { "epoch": 0.4908933217692975, "grad_norm": 0.2500433623790741, "learning_rate": 8.394022921804729e-05, "loss": 0.0415, "step": 4528 }, { "epoch": 0.49100173460537727, "grad_norm": 0.3252948820590973, "learning_rate": 8.393660235021036e-05, "loss": 0.0266, "step": 4529 }, { "epoch": 0.4911101474414571, "grad_norm": 0.3250909447669983, "learning_rate": 8.393297548237342e-05, "loss": 0.0498, "step": 4530 }, { "epoch": 0.49121856027753685, "grad_norm": 0.35975348949432373, "learning_rate": 8.392934861453648e-05, "loss": 0.0542, "step": 4531 }, { "epoch": 0.49132697311361667, "grad_norm": 0.35760748386383057, "learning_rate": 8.392572174669956e-05, "loss": 0.0378, "step": 4532 }, { "epoch": 0.49143538594969643, "grad_norm": 0.5663256645202637, "learning_rate": 8.392209487886262e-05, "loss": 0.0865, "step": 4533 }, { "epoch": 0.49154379878577625, "grad_norm": 0.21619907021522522, "learning_rate": 8.391846801102569e-05, "loss": 0.0215, "step": 4534 }, { "epoch": 0.491652211621856, "grad_norm": 0.790104866027832, "learning_rate": 8.391484114318875e-05, "loss": 0.0622, "step": 4535 }, { "epoch": 0.49176062445793584, "grad_norm": 0.4991038143634796, "learning_rate": 8.391121427535182e-05, "loss": 0.1151, "step": 4536 }, { "epoch": 0.4918690372940156, "grad_norm": 0.2793213725090027, "learning_rate": 8.390758740751488e-05, "loss": 0.0446, "step": 4537 }, { "epoch": 0.4919774501300954, "grad_norm": 0.5473126173019409, "learning_rate": 8.390396053967793e-05, "loss": 0.1007, "step": 4538 }, { "epoch": 0.4920858629661752, "grad_norm": 0.399648517370224, "learning_rate": 8.3900333671841e-05, "loss": 0.0521, "step": 4539 }, { "epoch": 0.492194275802255, "grad_norm": 0.13777226209640503, "learning_rate": 8.389670680400406e-05, "loss": 0.0277, "step": 4540 }, { "epoch": 0.49230268863833476, "grad_norm": 0.38762474060058594, "learning_rate": 8.389307993616713e-05, "loss": 0.0191, "step": 4541 }, { "epoch": 0.4924111014744146, "grad_norm": 0.13147719204425812, "learning_rate": 8.388945306833019e-05, "loss": 0.012, "step": 4542 }, { "epoch": 0.49251951431049434, "grad_norm": 0.43708300590515137, "learning_rate": 8.388582620049326e-05, "loss": 0.0575, "step": 4543 }, { "epoch": 0.49262792714657416, "grad_norm": 0.48623499274253845, "learning_rate": 8.388219933265632e-05, "loss": 0.0745, "step": 4544 }, { "epoch": 0.4927363399826539, "grad_norm": 0.5016024708747864, "learning_rate": 8.387857246481939e-05, "loss": 0.0478, "step": 4545 }, { "epoch": 0.49284475281873374, "grad_norm": 0.16956232488155365, "learning_rate": 8.387494559698245e-05, "loss": 0.0232, "step": 4546 }, { "epoch": 0.4929531656548135, "grad_norm": 0.1981762945652008, "learning_rate": 8.38713187291455e-05, "loss": 0.0325, "step": 4547 }, { "epoch": 0.4930615784908933, "grad_norm": 0.6961861848831177, "learning_rate": 8.386769186130857e-05, "loss": 0.0563, "step": 4548 }, { "epoch": 0.4931699913269731, "grad_norm": 0.24230700731277466, "learning_rate": 8.386406499347165e-05, "loss": 0.0311, "step": 4549 }, { "epoch": 0.4932784041630529, "grad_norm": 0.2750345468521118, "learning_rate": 8.386043812563472e-05, "loss": 0.0405, "step": 4550 }, { "epoch": 0.49338681699913267, "grad_norm": 0.16127239167690277, "learning_rate": 8.385681125779777e-05, "loss": 0.0154, "step": 4551 }, { "epoch": 0.4934952298352125, "grad_norm": 0.37781715393066406, "learning_rate": 8.385318438996083e-05, "loss": 0.0437, "step": 4552 }, { "epoch": 0.4936036426712923, "grad_norm": 0.18962880969047546, "learning_rate": 8.38495575221239e-05, "loss": 0.0253, "step": 4553 }, { "epoch": 0.49371205550737207, "grad_norm": 0.459671288728714, "learning_rate": 8.384593065428696e-05, "loss": 0.0715, "step": 4554 }, { "epoch": 0.4938204683434519, "grad_norm": 0.8755285739898682, "learning_rate": 8.384230378645003e-05, "loss": 0.0642, "step": 4555 }, { "epoch": 0.49392888117953165, "grad_norm": 0.23607848584651947, "learning_rate": 8.383867691861309e-05, "loss": 0.0258, "step": 4556 }, { "epoch": 0.4940372940156115, "grad_norm": 0.2925175726413727, "learning_rate": 8.383505005077616e-05, "loss": 0.0691, "step": 4557 }, { "epoch": 0.49414570685169124, "grad_norm": 0.3643830120563507, "learning_rate": 8.383142318293922e-05, "loss": 0.0331, "step": 4558 }, { "epoch": 0.49425411968777105, "grad_norm": 0.32695651054382324, "learning_rate": 8.382779631510229e-05, "loss": 0.0577, "step": 4559 }, { "epoch": 0.4943625325238508, "grad_norm": 0.2585712671279907, "learning_rate": 8.382416944726534e-05, "loss": 0.0224, "step": 4560 }, { "epoch": 0.49447094535993064, "grad_norm": 0.2906908094882965, "learning_rate": 8.38205425794284e-05, "loss": 0.0294, "step": 4561 }, { "epoch": 0.4945793581960104, "grad_norm": 0.35076725482940674, "learning_rate": 8.381691571159147e-05, "loss": 0.0551, "step": 4562 }, { "epoch": 0.4946877710320902, "grad_norm": 0.1533113270998001, "learning_rate": 8.381328884375453e-05, "loss": 0.0107, "step": 4563 }, { "epoch": 0.49479618386817, "grad_norm": 0.3445722162723541, "learning_rate": 8.38096619759176e-05, "loss": 0.0336, "step": 4564 }, { "epoch": 0.4949045967042498, "grad_norm": 0.4136016070842743, "learning_rate": 8.380603510808066e-05, "loss": 0.0487, "step": 4565 }, { "epoch": 0.49501300954032956, "grad_norm": 0.2903064787387848, "learning_rate": 8.380240824024373e-05, "loss": 0.0225, "step": 4566 }, { "epoch": 0.4951214223764094, "grad_norm": 0.3150586783885956, "learning_rate": 8.37987813724068e-05, "loss": 0.0182, "step": 4567 }, { "epoch": 0.49522983521248914, "grad_norm": 0.20230142772197723, "learning_rate": 8.379515450456986e-05, "loss": 0.0384, "step": 4568 }, { "epoch": 0.49533824804856896, "grad_norm": 0.5039125084877014, "learning_rate": 8.379152763673293e-05, "loss": 0.0206, "step": 4569 }, { "epoch": 0.4954466608846487, "grad_norm": 0.4606322646141052, "learning_rate": 8.378790076889599e-05, "loss": 0.0452, "step": 4570 }, { "epoch": 0.49555507372072854, "grad_norm": 0.19559143483638763, "learning_rate": 8.378427390105906e-05, "loss": 0.0188, "step": 4571 }, { "epoch": 0.4956634865568083, "grad_norm": 0.4437609016895294, "learning_rate": 8.378064703322211e-05, "loss": 0.0258, "step": 4572 }, { "epoch": 0.4957718993928881, "grad_norm": 0.5596209168434143, "learning_rate": 8.377702016538518e-05, "loss": 0.0868, "step": 4573 }, { "epoch": 0.4958803122289679, "grad_norm": 0.41351351141929626, "learning_rate": 8.377339329754824e-05, "loss": 0.0358, "step": 4574 }, { "epoch": 0.4959887250650477, "grad_norm": 0.20505209267139435, "learning_rate": 8.37697664297113e-05, "loss": 0.0135, "step": 4575 }, { "epoch": 0.49609713790112747, "grad_norm": 0.7804675698280334, "learning_rate": 8.376613956187437e-05, "loss": 0.0656, "step": 4576 }, { "epoch": 0.4962055507372073, "grad_norm": 0.42823952436447144, "learning_rate": 8.376251269403743e-05, "loss": 0.05, "step": 4577 }, { "epoch": 0.49631396357328705, "grad_norm": 1.3488539457321167, "learning_rate": 8.37588858262005e-05, "loss": 0.0617, "step": 4578 }, { "epoch": 0.4964223764093669, "grad_norm": 0.5747660994529724, "learning_rate": 8.375525895836356e-05, "loss": 0.0377, "step": 4579 }, { "epoch": 0.49653078924544664, "grad_norm": 0.3226012885570526, "learning_rate": 8.375163209052663e-05, "loss": 0.0302, "step": 4580 }, { "epoch": 0.49663920208152645, "grad_norm": 0.6969184875488281, "learning_rate": 8.374800522268968e-05, "loss": 0.0743, "step": 4581 }, { "epoch": 0.4967476149176062, "grad_norm": 0.3227577209472656, "learning_rate": 8.374437835485275e-05, "loss": 0.0212, "step": 4582 }, { "epoch": 0.49685602775368604, "grad_norm": 0.36040785908699036, "learning_rate": 8.374075148701581e-05, "loss": 0.0215, "step": 4583 }, { "epoch": 0.49696444058976585, "grad_norm": 0.6443099975585938, "learning_rate": 8.373712461917888e-05, "loss": 0.0542, "step": 4584 }, { "epoch": 0.4970728534258456, "grad_norm": 0.34403035044670105, "learning_rate": 8.373349775134195e-05, "loss": 0.061, "step": 4585 }, { "epoch": 0.49718126626192544, "grad_norm": 0.8497428894042969, "learning_rate": 8.372987088350501e-05, "loss": 0.1133, "step": 4586 }, { "epoch": 0.4972896790980052, "grad_norm": 0.4493556022644043, "learning_rate": 8.372624401566808e-05, "loss": 0.0142, "step": 4587 }, { "epoch": 0.497398091934085, "grad_norm": 0.6896657347679138, "learning_rate": 8.372261714783114e-05, "loss": 0.0821, "step": 4588 }, { "epoch": 0.4975065047701648, "grad_norm": 0.36809736490249634, "learning_rate": 8.37189902799942e-05, "loss": 0.0664, "step": 4589 }, { "epoch": 0.4976149176062446, "grad_norm": 0.31370294094085693, "learning_rate": 8.371536341215727e-05, "loss": 0.0312, "step": 4590 }, { "epoch": 0.49772333044232436, "grad_norm": 0.14182770252227783, "learning_rate": 8.371173654432032e-05, "loss": 0.0123, "step": 4591 }, { "epoch": 0.4978317432784042, "grad_norm": 0.3087765574455261, "learning_rate": 8.37081096764834e-05, "loss": 0.0309, "step": 4592 }, { "epoch": 0.49794015611448394, "grad_norm": 0.2770099341869354, "learning_rate": 8.370448280864645e-05, "loss": 0.0252, "step": 4593 }, { "epoch": 0.49804856895056376, "grad_norm": 0.470563143491745, "learning_rate": 8.370085594080952e-05, "loss": 0.0455, "step": 4594 }, { "epoch": 0.4981569817866435, "grad_norm": 0.546416699886322, "learning_rate": 8.369722907297258e-05, "loss": 0.04, "step": 4595 }, { "epoch": 0.49826539462272335, "grad_norm": 0.24755647778511047, "learning_rate": 8.369360220513565e-05, "loss": 0.0353, "step": 4596 }, { "epoch": 0.4983738074588031, "grad_norm": 0.8709942102432251, "learning_rate": 8.368997533729871e-05, "loss": 0.0687, "step": 4597 }, { "epoch": 0.4984822202948829, "grad_norm": 0.7066741585731506, "learning_rate": 8.368634846946177e-05, "loss": 0.0823, "step": 4598 }, { "epoch": 0.4985906331309627, "grad_norm": 0.4566640555858612, "learning_rate": 8.368272160162484e-05, "loss": 0.0374, "step": 4599 }, { "epoch": 0.4986990459670425, "grad_norm": 0.668103039264679, "learning_rate": 8.36790947337879e-05, "loss": 0.0548, "step": 4600 }, { "epoch": 0.4988074588031223, "grad_norm": 0.42343077063560486, "learning_rate": 8.367546786595098e-05, "loss": 0.0521, "step": 4601 }, { "epoch": 0.4989158716392021, "grad_norm": 0.1880522519350052, "learning_rate": 8.367184099811404e-05, "loss": 0.0167, "step": 4602 }, { "epoch": 0.49902428447528185, "grad_norm": 0.2322542667388916, "learning_rate": 8.36682141302771e-05, "loss": 0.0345, "step": 4603 }, { "epoch": 0.4991326973113617, "grad_norm": 0.4352782964706421, "learning_rate": 8.366458726244016e-05, "loss": 0.0395, "step": 4604 }, { "epoch": 0.49924111014744144, "grad_norm": 0.30548742413520813, "learning_rate": 8.366096039460322e-05, "loss": 0.0631, "step": 4605 }, { "epoch": 0.49934952298352125, "grad_norm": 0.1973070502281189, "learning_rate": 8.365733352676629e-05, "loss": 0.0312, "step": 4606 }, { "epoch": 0.499457935819601, "grad_norm": 0.37579524517059326, "learning_rate": 8.365370665892935e-05, "loss": 0.0566, "step": 4607 }, { "epoch": 0.49956634865568084, "grad_norm": 0.365862637758255, "learning_rate": 8.365007979109242e-05, "loss": 0.0582, "step": 4608 }, { "epoch": 0.4996747614917606, "grad_norm": 0.3678530156612396, "learning_rate": 8.364645292325548e-05, "loss": 0.0324, "step": 4609 }, { "epoch": 0.4997831743278404, "grad_norm": 0.22838696837425232, "learning_rate": 8.364282605541855e-05, "loss": 0.0318, "step": 4610 }, { "epoch": 0.4998915871639202, "grad_norm": 0.20888274908065796, "learning_rate": 8.363919918758161e-05, "loss": 0.047, "step": 4611 }, { "epoch": 0.5, "grad_norm": 0.3950856328010559, "learning_rate": 8.363557231974466e-05, "loss": 0.0515, "step": 4612 }, { "epoch": 0.5001084128360798, "grad_norm": 0.2753843665122986, "learning_rate": 8.363194545190774e-05, "loss": 0.0302, "step": 4613 }, { "epoch": 0.5002168256721596, "grad_norm": 0.33619505167007446, "learning_rate": 8.362831858407079e-05, "loss": 0.0616, "step": 4614 }, { "epoch": 0.5003252385082394, "grad_norm": 0.28620225191116333, "learning_rate": 8.362469171623386e-05, "loss": 0.0246, "step": 4615 }, { "epoch": 0.5004336513443192, "grad_norm": 0.5791827440261841, "learning_rate": 8.362106484839692e-05, "loss": 0.0458, "step": 4616 }, { "epoch": 0.5005420641803989, "grad_norm": 0.38432812690734863, "learning_rate": 8.361743798055999e-05, "loss": 0.0632, "step": 4617 }, { "epoch": 0.5006504770164788, "grad_norm": 0.206788569688797, "learning_rate": 8.361381111272306e-05, "loss": 0.0253, "step": 4618 }, { "epoch": 0.5007588898525586, "grad_norm": 0.2648043632507324, "learning_rate": 8.361018424488612e-05, "loss": 0.0514, "step": 4619 }, { "epoch": 0.5008673026886383, "grad_norm": 0.21958664059638977, "learning_rate": 8.360655737704919e-05, "loss": 0.0301, "step": 4620 }, { "epoch": 0.5009757155247181, "grad_norm": 0.3349491059780121, "learning_rate": 8.360293050921225e-05, "loss": 0.0447, "step": 4621 }, { "epoch": 0.501084128360798, "grad_norm": 0.31896457076072693, "learning_rate": 8.359930364137532e-05, "loss": 0.0414, "step": 4622 }, { "epoch": 0.5011925411968777, "grad_norm": 0.2460983544588089, "learning_rate": 8.359567677353838e-05, "loss": 0.0436, "step": 4623 }, { "epoch": 0.5013009540329575, "grad_norm": 0.5886080861091614, "learning_rate": 8.359204990570145e-05, "loss": 0.086, "step": 4624 }, { "epoch": 0.5014093668690373, "grad_norm": 0.30405759811401367, "learning_rate": 8.35884230378645e-05, "loss": 0.0403, "step": 4625 }, { "epoch": 0.5015177797051171, "grad_norm": 0.1910000890493393, "learning_rate": 8.358479617002756e-05, "loss": 0.0214, "step": 4626 }, { "epoch": 0.5016261925411969, "grad_norm": 0.30294638872146606, "learning_rate": 8.358116930219063e-05, "loss": 0.0359, "step": 4627 }, { "epoch": 0.5017346053772767, "grad_norm": 0.2239377796649933, "learning_rate": 8.357754243435369e-05, "loss": 0.0343, "step": 4628 }, { "epoch": 0.5018430182133564, "grad_norm": 0.3927932381629944, "learning_rate": 8.357391556651676e-05, "loss": 0.052, "step": 4629 }, { "epoch": 0.5019514310494363, "grad_norm": 0.4546128511428833, "learning_rate": 8.357028869867982e-05, "loss": 0.0645, "step": 4630 }, { "epoch": 0.502059843885516, "grad_norm": 0.2650679051876068, "learning_rate": 8.356666183084289e-05, "loss": 0.0381, "step": 4631 }, { "epoch": 0.5021682567215958, "grad_norm": 0.4032273292541504, "learning_rate": 8.356303496300595e-05, "loss": 0.0354, "step": 4632 }, { "epoch": 0.5022766695576756, "grad_norm": 0.5046197772026062, "learning_rate": 8.355940809516902e-05, "loss": 0.0633, "step": 4633 }, { "epoch": 0.5023850823937555, "grad_norm": 0.24953420460224152, "learning_rate": 8.355578122733207e-05, "loss": 0.0506, "step": 4634 }, { "epoch": 0.5024934952298352, "grad_norm": 0.4805174767971039, "learning_rate": 8.355215435949515e-05, "loss": 0.0482, "step": 4635 }, { "epoch": 0.502601908065915, "grad_norm": 0.4675922691822052, "learning_rate": 8.354852749165822e-05, "loss": 0.0526, "step": 4636 }, { "epoch": 0.5027103209019947, "grad_norm": 0.38110464811325073, "learning_rate": 8.354490062382127e-05, "loss": 0.0461, "step": 4637 }, { "epoch": 0.5028187337380746, "grad_norm": 0.29790621995925903, "learning_rate": 8.354127375598434e-05, "loss": 0.0413, "step": 4638 }, { "epoch": 0.5029271465741544, "grad_norm": 0.2460242211818695, "learning_rate": 8.35376468881474e-05, "loss": 0.0236, "step": 4639 }, { "epoch": 0.5030355594102341, "grad_norm": 0.35240378975868225, "learning_rate": 8.353402002031047e-05, "loss": 0.0232, "step": 4640 }, { "epoch": 0.5031439722463139, "grad_norm": 0.3339425325393677, "learning_rate": 8.353039315247353e-05, "loss": 0.0453, "step": 4641 }, { "epoch": 0.5032523850823938, "grad_norm": 0.291932076215744, "learning_rate": 8.352676628463659e-05, "loss": 0.0404, "step": 4642 }, { "epoch": 0.5033607979184735, "grad_norm": 0.14158406853675842, "learning_rate": 8.352313941679966e-05, "loss": 0.0155, "step": 4643 }, { "epoch": 0.5034692107545533, "grad_norm": 0.5230783820152283, "learning_rate": 8.351951254896272e-05, "loss": 0.0279, "step": 4644 }, { "epoch": 0.5035776235906332, "grad_norm": 0.42375850677490234, "learning_rate": 8.351588568112579e-05, "loss": 0.0727, "step": 4645 }, { "epoch": 0.503686036426713, "grad_norm": 0.3260565996170044, "learning_rate": 8.351225881328884e-05, "loss": 0.0435, "step": 4646 }, { "epoch": 0.5037944492627927, "grad_norm": 0.7143384218215942, "learning_rate": 8.350863194545191e-05, "loss": 0.0785, "step": 4647 }, { "epoch": 0.5039028620988725, "grad_norm": 0.42601314187049866, "learning_rate": 8.350500507761497e-05, "loss": 0.0402, "step": 4648 }, { "epoch": 0.5040112749349523, "grad_norm": 0.2174597829580307, "learning_rate": 8.350137820977804e-05, "loss": 0.0307, "step": 4649 }, { "epoch": 0.5041196877710321, "grad_norm": 0.32330387830734253, "learning_rate": 8.34977513419411e-05, "loss": 0.0492, "step": 4650 }, { "epoch": 0.5042281006071119, "grad_norm": 0.1530415266752243, "learning_rate": 8.349412447410416e-05, "loss": 0.0135, "step": 4651 }, { "epoch": 0.5043365134431916, "grad_norm": 0.31729191541671753, "learning_rate": 8.349049760626723e-05, "loss": 0.0514, "step": 4652 }, { "epoch": 0.5044449262792715, "grad_norm": 0.11472392827272415, "learning_rate": 8.34868707384303e-05, "loss": 0.0081, "step": 4653 }, { "epoch": 0.5045533391153513, "grad_norm": 0.164133220911026, "learning_rate": 8.348324387059337e-05, "loss": 0.0159, "step": 4654 }, { "epoch": 0.504661751951431, "grad_norm": 0.33855828642845154, "learning_rate": 8.347961700275643e-05, "loss": 0.0551, "step": 4655 }, { "epoch": 0.5047701647875108, "grad_norm": 0.21586258709430695, "learning_rate": 8.347599013491948e-05, "loss": 0.023, "step": 4656 }, { "epoch": 0.5048785776235907, "grad_norm": 0.7195724844932556, "learning_rate": 8.347236326708256e-05, "loss": 0.0576, "step": 4657 }, { "epoch": 0.5049869904596704, "grad_norm": 0.24896638095378876, "learning_rate": 8.346873639924561e-05, "loss": 0.0195, "step": 4658 }, { "epoch": 0.5050954032957502, "grad_norm": 0.7197834253311157, "learning_rate": 8.346510953140868e-05, "loss": 0.0714, "step": 4659 }, { "epoch": 0.50520381613183, "grad_norm": 0.36959928274154663, "learning_rate": 8.346148266357174e-05, "loss": 0.0496, "step": 4660 }, { "epoch": 0.5053122289679098, "grad_norm": 0.5365130305290222, "learning_rate": 8.345785579573481e-05, "loss": 0.0727, "step": 4661 }, { "epoch": 0.5054206418039896, "grad_norm": 0.2290496528148651, "learning_rate": 8.345422892789787e-05, "loss": 0.043, "step": 4662 }, { "epoch": 0.5055290546400694, "grad_norm": 0.20998530089855194, "learning_rate": 8.345060206006094e-05, "loss": 0.0353, "step": 4663 }, { "epoch": 0.5056374674761491, "grad_norm": 0.7879076600074768, "learning_rate": 8.3446975192224e-05, "loss": 0.0446, "step": 4664 }, { "epoch": 0.505745880312229, "grad_norm": 0.710987389087677, "learning_rate": 8.344334832438705e-05, "loss": 0.0431, "step": 4665 }, { "epoch": 0.5058542931483088, "grad_norm": 0.3076340854167938, "learning_rate": 8.343972145655013e-05, "loss": 0.065, "step": 4666 }, { "epoch": 0.5059627059843885, "grad_norm": 0.13970902562141418, "learning_rate": 8.343609458871318e-05, "loss": 0.0185, "step": 4667 }, { "epoch": 0.5060711188204683, "grad_norm": 0.16478921473026276, "learning_rate": 8.343246772087625e-05, "loss": 0.0207, "step": 4668 }, { "epoch": 0.5061795316565482, "grad_norm": 0.31920719146728516, "learning_rate": 8.342884085303931e-05, "loss": 0.0655, "step": 4669 }, { "epoch": 0.5062879444926279, "grad_norm": 0.5130345225334167, "learning_rate": 8.342521398520238e-05, "loss": 0.0187, "step": 4670 }, { "epoch": 0.5063963573287077, "grad_norm": 0.5106267333030701, "learning_rate": 8.342158711736545e-05, "loss": 0.0382, "step": 4671 }, { "epoch": 0.5065047701647876, "grad_norm": 0.20256179571151733, "learning_rate": 8.341796024952851e-05, "loss": 0.0257, "step": 4672 }, { "epoch": 0.5066131830008673, "grad_norm": 0.4249451458454132, "learning_rate": 8.341433338169158e-05, "loss": 0.0451, "step": 4673 }, { "epoch": 0.5067215958369471, "grad_norm": 0.4543616473674774, "learning_rate": 8.341070651385464e-05, "loss": 0.0782, "step": 4674 }, { "epoch": 0.5068300086730269, "grad_norm": 0.21028655767440796, "learning_rate": 8.340707964601771e-05, "loss": 0.028, "step": 4675 }, { "epoch": 0.5069384215091067, "grad_norm": 0.6850601434707642, "learning_rate": 8.340345277818077e-05, "loss": 0.0712, "step": 4676 }, { "epoch": 0.5070468343451865, "grad_norm": 0.5167320966720581, "learning_rate": 8.339982591034384e-05, "loss": 0.0213, "step": 4677 }, { "epoch": 0.5071552471812663, "grad_norm": 0.1316082775592804, "learning_rate": 8.33961990425069e-05, "loss": 0.0278, "step": 4678 }, { "epoch": 0.507263660017346, "grad_norm": 0.3721650242805481, "learning_rate": 8.339257217466995e-05, "loss": 0.0446, "step": 4679 }, { "epoch": 0.5073720728534259, "grad_norm": 0.2300952821969986, "learning_rate": 8.338894530683302e-05, "loss": 0.0174, "step": 4680 }, { "epoch": 0.5074804856895057, "grad_norm": 0.3330649435520172, "learning_rate": 8.338531843899608e-05, "loss": 0.0383, "step": 4681 }, { "epoch": 0.5075888985255854, "grad_norm": 0.38103345036506653, "learning_rate": 8.338169157115915e-05, "loss": 0.0267, "step": 4682 }, { "epoch": 0.5076973113616652, "grad_norm": 0.4055204391479492, "learning_rate": 8.337806470332221e-05, "loss": 0.0462, "step": 4683 }, { "epoch": 0.5078057241977451, "grad_norm": 0.6082165241241455, "learning_rate": 8.337443783548528e-05, "loss": 0.0694, "step": 4684 }, { "epoch": 0.5079141370338248, "grad_norm": 0.20729652047157288, "learning_rate": 8.337081096764834e-05, "loss": 0.021, "step": 4685 }, { "epoch": 0.5080225498699046, "grad_norm": 0.22685597836971283, "learning_rate": 8.336718409981141e-05, "loss": 0.0281, "step": 4686 }, { "epoch": 0.5081309627059843, "grad_norm": 0.2518862187862396, "learning_rate": 8.336355723197448e-05, "loss": 0.0429, "step": 4687 }, { "epoch": 0.5082393755420642, "grad_norm": 0.255595862865448, "learning_rate": 8.335993036413754e-05, "loss": 0.0294, "step": 4688 }, { "epoch": 0.508347788378144, "grad_norm": 0.46087756752967834, "learning_rate": 8.335630349630061e-05, "loss": 0.056, "step": 4689 }, { "epoch": 0.5084562012142237, "grad_norm": 0.5453871488571167, "learning_rate": 8.335267662846366e-05, "loss": 0.0407, "step": 4690 }, { "epoch": 0.5085646140503035, "grad_norm": 0.17082218825817108, "learning_rate": 8.334904976062674e-05, "loss": 0.02, "step": 4691 }, { "epoch": 0.5086730268863834, "grad_norm": 0.2133890837430954, "learning_rate": 8.334542289278979e-05, "loss": 0.027, "step": 4692 }, { "epoch": 0.5087814397224631, "grad_norm": 0.22760139405727386, "learning_rate": 8.334179602495285e-05, "loss": 0.0342, "step": 4693 }, { "epoch": 0.5088898525585429, "grad_norm": 1.0419347286224365, "learning_rate": 8.333816915711592e-05, "loss": 0.0771, "step": 4694 }, { "epoch": 0.5089982653946227, "grad_norm": 0.25439393520355225, "learning_rate": 8.333454228927898e-05, "loss": 0.0162, "step": 4695 }, { "epoch": 0.5091066782307025, "grad_norm": 0.36996138095855713, "learning_rate": 8.333091542144205e-05, "loss": 0.0528, "step": 4696 }, { "epoch": 0.5092150910667823, "grad_norm": 0.6388220191001892, "learning_rate": 8.33272885536051e-05, "loss": 0.0496, "step": 4697 }, { "epoch": 0.5093235039028621, "grad_norm": 0.21601435542106628, "learning_rate": 8.332366168576818e-05, "loss": 0.0245, "step": 4698 }, { "epoch": 0.5094319167389418, "grad_norm": 0.1154913604259491, "learning_rate": 8.332003481793123e-05, "loss": 0.0164, "step": 4699 }, { "epoch": 0.5095403295750217, "grad_norm": 0.48008203506469727, "learning_rate": 8.33164079500943e-05, "loss": 0.07, "step": 4700 }, { "epoch": 0.5096487424111015, "grad_norm": 0.2652197778224945, "learning_rate": 8.331278108225736e-05, "loss": 0.0542, "step": 4701 }, { "epoch": 0.5097571552471812, "grad_norm": 0.6139099597930908, "learning_rate": 8.330915421442042e-05, "loss": 0.0694, "step": 4702 }, { "epoch": 0.5098655680832611, "grad_norm": 0.5761639475822449, "learning_rate": 8.330552734658349e-05, "loss": 0.049, "step": 4703 }, { "epoch": 0.5099739809193409, "grad_norm": 0.42220935225486755, "learning_rate": 8.330190047874656e-05, "loss": 0.0453, "step": 4704 }, { "epoch": 0.5100823937554206, "grad_norm": 0.4917561411857605, "learning_rate": 8.329827361090963e-05, "loss": 0.034, "step": 4705 }, { "epoch": 0.5101908065915004, "grad_norm": 0.3134770095348358, "learning_rate": 8.329464674307269e-05, "loss": 0.0204, "step": 4706 }, { "epoch": 0.5102992194275803, "grad_norm": 0.35171085596084595, "learning_rate": 8.329101987523575e-05, "loss": 0.0513, "step": 4707 }, { "epoch": 0.51040763226366, "grad_norm": 0.30703461170196533, "learning_rate": 8.328739300739882e-05, "loss": 0.034, "step": 4708 }, { "epoch": 0.5105160450997398, "grad_norm": 0.25281307101249695, "learning_rate": 8.328376613956188e-05, "loss": 0.0141, "step": 4709 }, { "epoch": 0.5106244579358196, "grad_norm": 0.2853928208351135, "learning_rate": 8.328013927172495e-05, "loss": 0.0382, "step": 4710 }, { "epoch": 0.5107328707718994, "grad_norm": 0.2531490921974182, "learning_rate": 8.3276512403888e-05, "loss": 0.0257, "step": 4711 }, { "epoch": 0.5108412836079792, "grad_norm": 0.4599021375179291, "learning_rate": 8.327288553605107e-05, "loss": 0.0735, "step": 4712 }, { "epoch": 0.510949696444059, "grad_norm": 0.056335948407649994, "learning_rate": 8.326925866821413e-05, "loss": 0.003, "step": 4713 }, { "epoch": 0.5110581092801387, "grad_norm": 0.45105206966400146, "learning_rate": 8.32656318003772e-05, "loss": 0.0627, "step": 4714 }, { "epoch": 0.5111665221162186, "grad_norm": 0.25840747356414795, "learning_rate": 8.326200493254026e-05, "loss": 0.018, "step": 4715 }, { "epoch": 0.5112749349522984, "grad_norm": 0.28004905581474304, "learning_rate": 8.325837806470332e-05, "loss": 0.0173, "step": 4716 }, { "epoch": 0.5113833477883781, "grad_norm": 0.4711858928203583, "learning_rate": 8.325475119686639e-05, "loss": 0.0545, "step": 4717 }, { "epoch": 0.5114917606244579, "grad_norm": 0.1364394873380661, "learning_rate": 8.325112432902945e-05, "loss": 0.0174, "step": 4718 }, { "epoch": 0.5116001734605378, "grad_norm": 0.3097732961177826, "learning_rate": 8.324749746119252e-05, "loss": 0.0532, "step": 4719 }, { "epoch": 0.5117085862966175, "grad_norm": 0.3262961506843567, "learning_rate": 8.324387059335557e-05, "loss": 0.0394, "step": 4720 }, { "epoch": 0.5118169991326973, "grad_norm": 0.5161823034286499, "learning_rate": 8.324024372551864e-05, "loss": 0.0921, "step": 4721 }, { "epoch": 0.511925411968777, "grad_norm": 0.35506099462509155, "learning_rate": 8.323661685768172e-05, "loss": 0.0374, "step": 4722 }, { "epoch": 0.5120338248048569, "grad_norm": 0.2487083077430725, "learning_rate": 8.323298998984477e-05, "loss": 0.021, "step": 4723 }, { "epoch": 0.5121422376409367, "grad_norm": 0.6425115466117859, "learning_rate": 8.322936312200784e-05, "loss": 0.0313, "step": 4724 }, { "epoch": 0.5122506504770165, "grad_norm": 0.3499578535556793, "learning_rate": 8.32257362541709e-05, "loss": 0.0297, "step": 4725 }, { "epoch": 0.5123590633130962, "grad_norm": 0.2820475697517395, "learning_rate": 8.322210938633397e-05, "loss": 0.041, "step": 4726 }, { "epoch": 0.5124674761491761, "grad_norm": 0.6913605332374573, "learning_rate": 8.321848251849703e-05, "loss": 0.0384, "step": 4727 }, { "epoch": 0.5125758889852559, "grad_norm": 0.18281389772891998, "learning_rate": 8.32148556506601e-05, "loss": 0.0166, "step": 4728 }, { "epoch": 0.5126843018213356, "grad_norm": 0.4028238356113434, "learning_rate": 8.321122878282316e-05, "loss": 0.0215, "step": 4729 }, { "epoch": 0.5127927146574154, "grad_norm": 0.2833217680454254, "learning_rate": 8.320760191498623e-05, "loss": 0.0189, "step": 4730 }, { "epoch": 0.5129011274934953, "grad_norm": 0.36790671944618225, "learning_rate": 8.320397504714929e-05, "loss": 0.0426, "step": 4731 }, { "epoch": 0.513009540329575, "grad_norm": 0.40574967861175537, "learning_rate": 8.320034817931234e-05, "loss": 0.0422, "step": 4732 }, { "epoch": 0.5131179531656548, "grad_norm": 0.5091977715492249, "learning_rate": 8.319672131147541e-05, "loss": 0.0857, "step": 4733 }, { "epoch": 0.5132263660017347, "grad_norm": 0.6343412399291992, "learning_rate": 8.319309444363847e-05, "loss": 0.0967, "step": 4734 }, { "epoch": 0.5133347788378144, "grad_norm": 0.4707697629928589, "learning_rate": 8.318946757580154e-05, "loss": 0.0876, "step": 4735 }, { "epoch": 0.5134431916738942, "grad_norm": 0.5045757293701172, "learning_rate": 8.31858407079646e-05, "loss": 0.078, "step": 4736 }, { "epoch": 0.513551604509974, "grad_norm": 0.210001602768898, "learning_rate": 8.318221384012767e-05, "loss": 0.0185, "step": 4737 }, { "epoch": 0.5136600173460538, "grad_norm": 0.4370919167995453, "learning_rate": 8.317858697229073e-05, "loss": 0.0792, "step": 4738 }, { "epoch": 0.5137684301821336, "grad_norm": 0.2801370918750763, "learning_rate": 8.31749601044538e-05, "loss": 0.0267, "step": 4739 }, { "epoch": 0.5138768430182133, "grad_norm": 0.0978972315788269, "learning_rate": 8.317133323661687e-05, "loss": 0.0103, "step": 4740 }, { "epoch": 0.5139852558542931, "grad_norm": 0.9121981263160706, "learning_rate": 8.316770636877993e-05, "loss": 0.0589, "step": 4741 }, { "epoch": 0.514093668690373, "grad_norm": 0.5264039039611816, "learning_rate": 8.3164079500943e-05, "loss": 0.0368, "step": 4742 }, { "epoch": 0.5142020815264527, "grad_norm": 0.43542224168777466, "learning_rate": 8.316045263310606e-05, "loss": 0.0318, "step": 4743 }, { "epoch": 0.5143104943625325, "grad_norm": 0.4511164724826813, "learning_rate": 8.315682576526913e-05, "loss": 0.0777, "step": 4744 }, { "epoch": 0.5144189071986123, "grad_norm": 0.2934363782405853, "learning_rate": 8.315319889743218e-05, "loss": 0.0293, "step": 4745 }, { "epoch": 0.5145273200346921, "grad_norm": 0.17556264996528625, "learning_rate": 8.314957202959524e-05, "loss": 0.0286, "step": 4746 }, { "epoch": 0.5146357328707719, "grad_norm": 0.676131010055542, "learning_rate": 8.314594516175831e-05, "loss": 0.0427, "step": 4747 }, { "epoch": 0.5147441457068517, "grad_norm": 0.29925209283828735, "learning_rate": 8.314231829392137e-05, "loss": 0.026, "step": 4748 }, { "epoch": 0.5148525585429314, "grad_norm": 0.4053623080253601, "learning_rate": 8.313869142608444e-05, "loss": 0.0513, "step": 4749 }, { "epoch": 0.5149609713790113, "grad_norm": 0.22976791858673096, "learning_rate": 8.31350645582475e-05, "loss": 0.0279, "step": 4750 }, { "epoch": 0.5150693842150911, "grad_norm": 0.3965551257133484, "learning_rate": 8.313143769041057e-05, "loss": 0.0663, "step": 4751 }, { "epoch": 0.5151777970511708, "grad_norm": 0.09847661852836609, "learning_rate": 8.312781082257363e-05, "loss": 0.0084, "step": 4752 }, { "epoch": 0.5152862098872506, "grad_norm": 0.3110676407814026, "learning_rate": 8.31241839547367e-05, "loss": 0.0111, "step": 4753 }, { "epoch": 0.5153946227233305, "grad_norm": 0.22249336540699005, "learning_rate": 8.312055708689975e-05, "loss": 0.0351, "step": 4754 }, { "epoch": 0.5155030355594102, "grad_norm": 0.22117459774017334, "learning_rate": 8.311693021906281e-05, "loss": 0.0212, "step": 4755 }, { "epoch": 0.51561144839549, "grad_norm": 0.6851757168769836, "learning_rate": 8.31133033512259e-05, "loss": 0.0617, "step": 4756 }, { "epoch": 0.5157198612315698, "grad_norm": 0.25280630588531494, "learning_rate": 8.310967648338895e-05, "loss": 0.0247, "step": 4757 }, { "epoch": 0.5158282740676496, "grad_norm": 0.4473797380924225, "learning_rate": 8.310604961555202e-05, "loss": 0.0455, "step": 4758 }, { "epoch": 0.5159366869037294, "grad_norm": 0.26926010847091675, "learning_rate": 8.310242274771508e-05, "loss": 0.0275, "step": 4759 }, { "epoch": 0.5160450997398092, "grad_norm": 0.8233350515365601, "learning_rate": 8.309879587987814e-05, "loss": 0.042, "step": 4760 }, { "epoch": 0.516153512575889, "grad_norm": 0.6938257217407227, "learning_rate": 8.309516901204121e-05, "loss": 0.0961, "step": 4761 }, { "epoch": 0.5162619254119688, "grad_norm": 0.264456182718277, "learning_rate": 8.309154214420427e-05, "loss": 0.0255, "step": 4762 }, { "epoch": 0.5163703382480486, "grad_norm": 0.537473738193512, "learning_rate": 8.308791527636734e-05, "loss": 0.0201, "step": 4763 }, { "epoch": 0.5164787510841283, "grad_norm": 0.3568666875362396, "learning_rate": 8.30842884085304e-05, "loss": 0.0296, "step": 4764 }, { "epoch": 0.5165871639202082, "grad_norm": 0.3434988260269165, "learning_rate": 8.308066154069347e-05, "loss": 0.036, "step": 4765 }, { "epoch": 0.516695576756288, "grad_norm": 0.459424763917923, "learning_rate": 8.307703467285652e-05, "loss": 0.0403, "step": 4766 }, { "epoch": 0.5168039895923677, "grad_norm": 0.27612075209617615, "learning_rate": 8.30734078050196e-05, "loss": 0.0147, "step": 4767 }, { "epoch": 0.5169124024284475, "grad_norm": 0.6224313974380493, "learning_rate": 8.306978093718265e-05, "loss": 0.1044, "step": 4768 }, { "epoch": 0.5170208152645274, "grad_norm": 0.4348694384098053, "learning_rate": 8.306615406934571e-05, "loss": 0.0358, "step": 4769 }, { "epoch": 0.5171292281006071, "grad_norm": 0.3383919298648834, "learning_rate": 8.306252720150878e-05, "loss": 0.0766, "step": 4770 }, { "epoch": 0.5172376409366869, "grad_norm": 0.16074484586715698, "learning_rate": 8.305890033367184e-05, "loss": 0.0182, "step": 4771 }, { "epoch": 0.5173460537727667, "grad_norm": 0.15891927480697632, "learning_rate": 8.305527346583491e-05, "loss": 0.0254, "step": 4772 }, { "epoch": 0.5174544666088465, "grad_norm": 0.45294973254203796, "learning_rate": 8.305164659799796e-05, "loss": 0.0544, "step": 4773 }, { "epoch": 0.5175628794449263, "grad_norm": 0.4252510368824005, "learning_rate": 8.304801973016104e-05, "loss": 0.049, "step": 4774 }, { "epoch": 0.5176712922810061, "grad_norm": 0.509326159954071, "learning_rate": 8.30443928623241e-05, "loss": 0.0422, "step": 4775 }, { "epoch": 0.5177797051170858, "grad_norm": 0.2997463643550873, "learning_rate": 8.304076599448716e-05, "loss": 0.0212, "step": 4776 }, { "epoch": 0.5178881179531657, "grad_norm": 0.4994339644908905, "learning_rate": 8.303713912665023e-05, "loss": 0.0287, "step": 4777 }, { "epoch": 0.5179965307892455, "grad_norm": 0.3325309753417969, "learning_rate": 8.303351225881329e-05, "loss": 0.0553, "step": 4778 }, { "epoch": 0.5181049436253252, "grad_norm": 0.16542385518550873, "learning_rate": 8.302988539097636e-05, "loss": 0.0176, "step": 4779 }, { "epoch": 0.518213356461405, "grad_norm": 0.4998578429222107, "learning_rate": 8.302625852313942e-05, "loss": 0.0433, "step": 4780 }, { "epoch": 0.5183217692974849, "grad_norm": 0.28348660469055176, "learning_rate": 8.302263165530249e-05, "loss": 0.0338, "step": 4781 }, { "epoch": 0.5184301821335646, "grad_norm": 0.1720132827758789, "learning_rate": 8.301900478746555e-05, "loss": 0.0161, "step": 4782 }, { "epoch": 0.5185385949696444, "grad_norm": 0.45919111371040344, "learning_rate": 8.30153779196286e-05, "loss": 0.0384, "step": 4783 }, { "epoch": 0.5186470078057241, "grad_norm": 1.098098874092102, "learning_rate": 8.301175105179168e-05, "loss": 0.0645, "step": 4784 }, { "epoch": 0.518755420641804, "grad_norm": 0.28036370873451233, "learning_rate": 8.300812418395473e-05, "loss": 0.0464, "step": 4785 }, { "epoch": 0.5188638334778838, "grad_norm": 0.3674863874912262, "learning_rate": 8.30044973161178e-05, "loss": 0.0484, "step": 4786 }, { "epoch": 0.5189722463139635, "grad_norm": 0.3729671835899353, "learning_rate": 8.300087044828086e-05, "loss": 0.0444, "step": 4787 }, { "epoch": 0.5190806591500433, "grad_norm": 0.43676432967185974, "learning_rate": 8.299724358044393e-05, "loss": 0.0228, "step": 4788 }, { "epoch": 0.5191890719861232, "grad_norm": 0.6417896151542664, "learning_rate": 8.299361671260699e-05, "loss": 0.0506, "step": 4789 }, { "epoch": 0.519297484822203, "grad_norm": 0.5609805583953857, "learning_rate": 8.298998984477006e-05, "loss": 0.0568, "step": 4790 }, { "epoch": 0.5194058976582827, "grad_norm": 0.30478549003601074, "learning_rate": 8.298636297693313e-05, "loss": 0.0316, "step": 4791 }, { "epoch": 0.5195143104943626, "grad_norm": 0.19457821547985077, "learning_rate": 8.298273610909619e-05, "loss": 0.0294, "step": 4792 }, { "epoch": 0.5196227233304423, "grad_norm": 0.33825549483299255, "learning_rate": 8.297910924125926e-05, "loss": 0.0544, "step": 4793 }, { "epoch": 0.5197311361665221, "grad_norm": 0.46715623140335083, "learning_rate": 8.297548237342232e-05, "loss": 0.0647, "step": 4794 }, { "epoch": 0.5198395490026019, "grad_norm": 0.24847020208835602, "learning_rate": 8.297185550558539e-05, "loss": 0.0255, "step": 4795 }, { "epoch": 0.5199479618386817, "grad_norm": 0.3206416070461273, "learning_rate": 8.296822863774845e-05, "loss": 0.0426, "step": 4796 }, { "epoch": 0.5200563746747615, "grad_norm": 0.357977956533432, "learning_rate": 8.29646017699115e-05, "loss": 0.0277, "step": 4797 }, { "epoch": 0.5201647875108413, "grad_norm": 0.19530755281448364, "learning_rate": 8.296097490207457e-05, "loss": 0.0322, "step": 4798 }, { "epoch": 0.520273200346921, "grad_norm": 0.26246583461761475, "learning_rate": 8.295734803423763e-05, "loss": 0.0179, "step": 4799 }, { "epoch": 0.5203816131830009, "grad_norm": 0.12112228572368622, "learning_rate": 8.29537211664007e-05, "loss": 0.0106, "step": 4800 }, { "epoch": 0.5204900260190807, "grad_norm": 0.47113487124443054, "learning_rate": 8.295009429856376e-05, "loss": 0.0485, "step": 4801 }, { "epoch": 0.5205984388551604, "grad_norm": 0.39104416966438293, "learning_rate": 8.294646743072683e-05, "loss": 0.0499, "step": 4802 }, { "epoch": 0.5207068516912402, "grad_norm": 0.4354326128959656, "learning_rate": 8.294284056288989e-05, "loss": 0.0648, "step": 4803 }, { "epoch": 0.5208152645273201, "grad_norm": 0.1732151061296463, "learning_rate": 8.293921369505296e-05, "loss": 0.0133, "step": 4804 }, { "epoch": 0.5209236773633998, "grad_norm": 0.8345760107040405, "learning_rate": 8.293558682721602e-05, "loss": 0.0734, "step": 4805 }, { "epoch": 0.5210320901994796, "grad_norm": 0.17215842008590698, "learning_rate": 8.293195995937907e-05, "loss": 0.0183, "step": 4806 }, { "epoch": 0.5211405030355594, "grad_norm": 0.24584075808525085, "learning_rate": 8.292833309154214e-05, "loss": 0.0369, "step": 4807 }, { "epoch": 0.5212489158716392, "grad_norm": 0.5368770360946655, "learning_rate": 8.292470622370522e-05, "loss": 0.0392, "step": 4808 }, { "epoch": 0.521357328707719, "grad_norm": 0.2695907652378082, "learning_rate": 8.292107935586829e-05, "loss": 0.0285, "step": 4809 }, { "epoch": 0.5214657415437988, "grad_norm": 0.3416653275489807, "learning_rate": 8.291745248803134e-05, "loss": 0.0354, "step": 4810 }, { "epoch": 0.5215741543798785, "grad_norm": 0.39636293053627014, "learning_rate": 8.29138256201944e-05, "loss": 0.0381, "step": 4811 }, { "epoch": 0.5216825672159584, "grad_norm": 0.6305208802223206, "learning_rate": 8.291019875235747e-05, "loss": 0.0262, "step": 4812 }, { "epoch": 0.5217909800520382, "grad_norm": 0.28473249077796936, "learning_rate": 8.290657188452053e-05, "loss": 0.0413, "step": 4813 }, { "epoch": 0.5218993928881179, "grad_norm": 0.16478198766708374, "learning_rate": 8.29029450166836e-05, "loss": 0.0221, "step": 4814 }, { "epoch": 0.5220078057241977, "grad_norm": 0.4044928252696991, "learning_rate": 8.289931814884666e-05, "loss": 0.0631, "step": 4815 }, { "epoch": 0.5221162185602776, "grad_norm": 0.38007935881614685, "learning_rate": 8.289569128100973e-05, "loss": 0.0399, "step": 4816 }, { "epoch": 0.5222246313963573, "grad_norm": 0.14707714319229126, "learning_rate": 8.289206441317279e-05, "loss": 0.0236, "step": 4817 }, { "epoch": 0.5223330442324371, "grad_norm": 0.2246919721364975, "learning_rate": 8.288843754533586e-05, "loss": 0.0268, "step": 4818 }, { "epoch": 0.522441457068517, "grad_norm": 0.8004911541938782, "learning_rate": 8.288481067749891e-05, "loss": 0.0631, "step": 4819 }, { "epoch": 0.5225498699045967, "grad_norm": 0.3257129192352295, "learning_rate": 8.288118380966197e-05, "loss": 0.0492, "step": 4820 }, { "epoch": 0.5226582827406765, "grad_norm": 0.4000456929206848, "learning_rate": 8.287755694182504e-05, "loss": 0.0217, "step": 4821 }, { "epoch": 0.5227666955767563, "grad_norm": 0.47054412961006165, "learning_rate": 8.28739300739881e-05, "loss": 0.0709, "step": 4822 }, { "epoch": 0.5228751084128361, "grad_norm": 0.31467750668525696, "learning_rate": 8.287030320615117e-05, "loss": 0.0225, "step": 4823 }, { "epoch": 0.5229835212489159, "grad_norm": 0.30253005027770996, "learning_rate": 8.286667633831423e-05, "loss": 0.0344, "step": 4824 }, { "epoch": 0.5230919340849957, "grad_norm": 0.46626001596450806, "learning_rate": 8.286304947047731e-05, "loss": 0.0911, "step": 4825 }, { "epoch": 0.5232003469210754, "grad_norm": 0.35860612988471985, "learning_rate": 8.285942260264037e-05, "loss": 0.0281, "step": 4826 }, { "epoch": 0.5233087597571553, "grad_norm": 0.2629535496234894, "learning_rate": 8.285579573480343e-05, "loss": 0.023, "step": 4827 }, { "epoch": 0.5234171725932351, "grad_norm": 0.31479692459106445, "learning_rate": 8.28521688669665e-05, "loss": 0.057, "step": 4828 }, { "epoch": 0.5235255854293148, "grad_norm": 0.5704834461212158, "learning_rate": 8.284854199912955e-05, "loss": 0.0375, "step": 4829 }, { "epoch": 0.5236339982653946, "grad_norm": 0.18030719459056854, "learning_rate": 8.284491513129263e-05, "loss": 0.0324, "step": 4830 }, { "epoch": 0.5237424111014745, "grad_norm": 0.47308966517448425, "learning_rate": 8.284128826345568e-05, "loss": 0.0837, "step": 4831 }, { "epoch": 0.5238508239375542, "grad_norm": 0.41683265566825867, "learning_rate": 8.283766139561875e-05, "loss": 0.0511, "step": 4832 }, { "epoch": 0.523959236773634, "grad_norm": 0.5573986768722534, "learning_rate": 8.283403452778181e-05, "loss": 0.0708, "step": 4833 }, { "epoch": 0.5240676496097137, "grad_norm": 0.2586382031440735, "learning_rate": 8.283040765994488e-05, "loss": 0.0409, "step": 4834 }, { "epoch": 0.5241760624457936, "grad_norm": 0.5282539129257202, "learning_rate": 8.282678079210794e-05, "loss": 0.044, "step": 4835 }, { "epoch": 0.5242844752818734, "grad_norm": 0.363654762506485, "learning_rate": 8.2823153924271e-05, "loss": 0.0424, "step": 4836 }, { "epoch": 0.5243928881179531, "grad_norm": 0.13112889230251312, "learning_rate": 8.281952705643407e-05, "loss": 0.0195, "step": 4837 }, { "epoch": 0.5245013009540329, "grad_norm": 0.34668079018592834, "learning_rate": 8.281590018859712e-05, "loss": 0.026, "step": 4838 }, { "epoch": 0.5246097137901128, "grad_norm": 0.1735173463821411, "learning_rate": 8.28122733207602e-05, "loss": 0.0209, "step": 4839 }, { "epoch": 0.5247181266261925, "grad_norm": 0.1876860409975052, "learning_rate": 8.280864645292325e-05, "loss": 0.019, "step": 4840 }, { "epoch": 0.5248265394622723, "grad_norm": 0.17564265429973602, "learning_rate": 8.280501958508632e-05, "loss": 0.0361, "step": 4841 }, { "epoch": 0.5249349522983521, "grad_norm": 0.17607952654361725, "learning_rate": 8.280139271724938e-05, "loss": 0.0216, "step": 4842 }, { "epoch": 0.525043365134432, "grad_norm": 0.357992947101593, "learning_rate": 8.279776584941245e-05, "loss": 0.0281, "step": 4843 }, { "epoch": 0.5251517779705117, "grad_norm": 0.246158167719841, "learning_rate": 8.279413898157552e-05, "loss": 0.04, "step": 4844 }, { "epoch": 0.5252601908065915, "grad_norm": 0.17470575869083405, "learning_rate": 8.279051211373858e-05, "loss": 0.0173, "step": 4845 }, { "epoch": 0.5253686036426712, "grad_norm": 0.5254480242729187, "learning_rate": 8.278688524590165e-05, "loss": 0.0618, "step": 4846 }, { "epoch": 0.5254770164787511, "grad_norm": 0.3868779242038727, "learning_rate": 8.278325837806471e-05, "loss": 0.0509, "step": 4847 }, { "epoch": 0.5255854293148309, "grad_norm": 0.14849400520324707, "learning_rate": 8.277963151022778e-05, "loss": 0.0127, "step": 4848 }, { "epoch": 0.5256938421509106, "grad_norm": 0.39868608117103577, "learning_rate": 8.277600464239084e-05, "loss": 0.0386, "step": 4849 }, { "epoch": 0.5258022549869905, "grad_norm": 0.38340282440185547, "learning_rate": 8.27723777745539e-05, "loss": 0.0357, "step": 4850 }, { "epoch": 0.5259106678230703, "grad_norm": 0.3936082422733307, "learning_rate": 8.276875090671696e-05, "loss": 0.0342, "step": 4851 }, { "epoch": 0.52601908065915, "grad_norm": 0.28719887137413025, "learning_rate": 8.276512403888002e-05, "loss": 0.0344, "step": 4852 }, { "epoch": 0.5261274934952298, "grad_norm": 0.26961541175842285, "learning_rate": 8.276149717104309e-05, "loss": 0.0231, "step": 4853 }, { "epoch": 0.5262359063313097, "grad_norm": 0.22261925041675568, "learning_rate": 8.275787030320615e-05, "loss": 0.0224, "step": 4854 }, { "epoch": 0.5263443191673894, "grad_norm": 0.26170113682746887, "learning_rate": 8.275424343536922e-05, "loss": 0.0365, "step": 4855 }, { "epoch": 0.5264527320034692, "grad_norm": 0.5588855147361755, "learning_rate": 8.275061656753228e-05, "loss": 0.0721, "step": 4856 }, { "epoch": 0.526561144839549, "grad_norm": 0.41076046228408813, "learning_rate": 8.274698969969535e-05, "loss": 0.0458, "step": 4857 }, { "epoch": 0.5266695576756288, "grad_norm": 0.15856431424617767, "learning_rate": 8.274336283185841e-05, "loss": 0.016, "step": 4858 }, { "epoch": 0.5267779705117086, "grad_norm": 0.5364786386489868, "learning_rate": 8.273973596402146e-05, "loss": 0.0543, "step": 4859 }, { "epoch": 0.5268863833477884, "grad_norm": 0.3690131604671478, "learning_rate": 8.273610909618455e-05, "loss": 0.0893, "step": 4860 }, { "epoch": 0.5269947961838681, "grad_norm": 0.14992253482341766, "learning_rate": 8.27324822283476e-05, "loss": 0.0121, "step": 4861 }, { "epoch": 0.527103209019948, "grad_norm": 0.6667361855506897, "learning_rate": 8.272885536051068e-05, "loss": 0.0458, "step": 4862 }, { "epoch": 0.5272116218560278, "grad_norm": 0.7766482830047607, "learning_rate": 8.272522849267373e-05, "loss": 0.0728, "step": 4863 }, { "epoch": 0.5273200346921075, "grad_norm": 0.19396089017391205, "learning_rate": 8.272160162483679e-05, "loss": 0.0132, "step": 4864 }, { "epoch": 0.5274284475281873, "grad_norm": 0.5093610882759094, "learning_rate": 8.271797475699986e-05, "loss": 0.0458, "step": 4865 }, { "epoch": 0.5275368603642672, "grad_norm": 0.24303647875785828, "learning_rate": 8.271434788916292e-05, "loss": 0.0245, "step": 4866 }, { "epoch": 0.5276452732003469, "grad_norm": 0.313609778881073, "learning_rate": 8.271072102132599e-05, "loss": 0.0374, "step": 4867 }, { "epoch": 0.5277536860364267, "grad_norm": 0.6024829745292664, "learning_rate": 8.270709415348905e-05, "loss": 0.0404, "step": 4868 }, { "epoch": 0.5278620988725065, "grad_norm": 0.21719473600387573, "learning_rate": 8.270346728565212e-05, "loss": 0.0541, "step": 4869 }, { "epoch": 0.5279705117085863, "grad_norm": 0.2921391725540161, "learning_rate": 8.269984041781518e-05, "loss": 0.0216, "step": 4870 }, { "epoch": 0.5280789245446661, "grad_norm": 0.47894179821014404, "learning_rate": 8.269621354997825e-05, "loss": 0.0731, "step": 4871 }, { "epoch": 0.5281873373807459, "grad_norm": 0.33939215540885925, "learning_rate": 8.26925866821413e-05, "loss": 0.0336, "step": 4872 }, { "epoch": 0.5282957502168256, "grad_norm": 0.30717742443084717, "learning_rate": 8.268895981430436e-05, "loss": 0.0405, "step": 4873 }, { "epoch": 0.5284041630529055, "grad_norm": 0.2897239327430725, "learning_rate": 8.268533294646743e-05, "loss": 0.0394, "step": 4874 }, { "epoch": 0.5285125758889853, "grad_norm": 0.36648625135421753, "learning_rate": 8.268170607863049e-05, "loss": 0.0485, "step": 4875 }, { "epoch": 0.528620988725065, "grad_norm": 0.4383121430873871, "learning_rate": 8.267807921079356e-05, "loss": 0.033, "step": 4876 }, { "epoch": 0.5287294015611448, "grad_norm": 0.270753413438797, "learning_rate": 8.267445234295663e-05, "loss": 0.0411, "step": 4877 }, { "epoch": 0.5288378143972247, "grad_norm": 0.40241843461990356, "learning_rate": 8.267082547511969e-05, "loss": 0.0496, "step": 4878 }, { "epoch": 0.5289462272333044, "grad_norm": 0.2789740562438965, "learning_rate": 8.266719860728276e-05, "loss": 0.0168, "step": 4879 }, { "epoch": 0.5290546400693842, "grad_norm": 0.11547742784023285, "learning_rate": 8.266357173944582e-05, "loss": 0.0154, "step": 4880 }, { "epoch": 0.5291630529054641, "grad_norm": 0.18173764646053314, "learning_rate": 8.265994487160889e-05, "loss": 0.035, "step": 4881 }, { "epoch": 0.5292714657415438, "grad_norm": 0.2214156836271286, "learning_rate": 8.265631800377195e-05, "loss": 0.0184, "step": 4882 }, { "epoch": 0.5293798785776236, "grad_norm": 0.7365564703941345, "learning_rate": 8.265269113593502e-05, "loss": 0.0872, "step": 4883 }, { "epoch": 0.5294882914137033, "grad_norm": 0.5063217282295227, "learning_rate": 8.264906426809807e-05, "loss": 0.0219, "step": 4884 }, { "epoch": 0.5295967042497832, "grad_norm": 0.16922323405742645, "learning_rate": 8.264543740026114e-05, "loss": 0.022, "step": 4885 }, { "epoch": 0.529705117085863, "grad_norm": 0.17863908410072327, "learning_rate": 8.26418105324242e-05, "loss": 0.0203, "step": 4886 }, { "epoch": 0.5298135299219427, "grad_norm": 0.20058712363243103, "learning_rate": 8.263818366458726e-05, "loss": 0.0215, "step": 4887 }, { "epoch": 0.5299219427580225, "grad_norm": 0.4262186884880066, "learning_rate": 8.263455679675033e-05, "loss": 0.0518, "step": 4888 }, { "epoch": 0.5300303555941024, "grad_norm": 0.3082541823387146, "learning_rate": 8.263092992891339e-05, "loss": 0.0416, "step": 4889 }, { "epoch": 0.5301387684301822, "grad_norm": 0.4065204858779907, "learning_rate": 8.262730306107646e-05, "loss": 0.0734, "step": 4890 }, { "epoch": 0.5302471812662619, "grad_norm": 0.16251572966575623, "learning_rate": 8.262367619323952e-05, "loss": 0.0179, "step": 4891 }, { "epoch": 0.5303555941023417, "grad_norm": 0.28996363282203674, "learning_rate": 8.262004932540259e-05, "loss": 0.0586, "step": 4892 }, { "epoch": 0.5304640069384216, "grad_norm": 0.2094043344259262, "learning_rate": 8.261642245756564e-05, "loss": 0.0117, "step": 4893 }, { "epoch": 0.5305724197745013, "grad_norm": 0.413565069437027, "learning_rate": 8.261279558972871e-05, "loss": 0.0676, "step": 4894 }, { "epoch": 0.5306808326105811, "grad_norm": 0.5395053029060364, "learning_rate": 8.260916872189179e-05, "loss": 0.0404, "step": 4895 }, { "epoch": 0.5307892454466608, "grad_norm": 0.37867313623428345, "learning_rate": 8.260554185405484e-05, "loss": 0.0405, "step": 4896 }, { "epoch": 0.5308976582827407, "grad_norm": 0.5447547435760498, "learning_rate": 8.260191498621791e-05, "loss": 0.0769, "step": 4897 }, { "epoch": 0.5310060711188205, "grad_norm": 0.253915935754776, "learning_rate": 8.259828811838097e-05, "loss": 0.051, "step": 4898 }, { "epoch": 0.5311144839549002, "grad_norm": 0.261601984500885, "learning_rate": 8.259466125054404e-05, "loss": 0.0308, "step": 4899 }, { "epoch": 0.53122289679098, "grad_norm": 0.3823794722557068, "learning_rate": 8.25910343827071e-05, "loss": 0.0616, "step": 4900 }, { "epoch": 0.5313313096270599, "grad_norm": 0.42706358432769775, "learning_rate": 8.258740751487016e-05, "loss": 0.0505, "step": 4901 }, { "epoch": 0.5314397224631396, "grad_norm": 0.435102254152298, "learning_rate": 8.258378064703323e-05, "loss": 0.0613, "step": 4902 }, { "epoch": 0.5315481352992194, "grad_norm": 0.22084330022335052, "learning_rate": 8.258015377919628e-05, "loss": 0.0186, "step": 4903 }, { "epoch": 0.5316565481352992, "grad_norm": 0.44338470697402954, "learning_rate": 8.257652691135936e-05, "loss": 0.0629, "step": 4904 }, { "epoch": 0.531764960971379, "grad_norm": 0.15779009461402893, "learning_rate": 8.257290004352241e-05, "loss": 0.024, "step": 4905 }, { "epoch": 0.5318733738074588, "grad_norm": 0.49120861291885376, "learning_rate": 8.256927317568548e-05, "loss": 0.0372, "step": 4906 }, { "epoch": 0.5319817866435386, "grad_norm": 0.6910156607627869, "learning_rate": 8.256564630784854e-05, "loss": 0.0734, "step": 4907 }, { "epoch": 0.5320901994796184, "grad_norm": 0.22672595083713531, "learning_rate": 8.256201944001161e-05, "loss": 0.0242, "step": 4908 }, { "epoch": 0.5321986123156982, "grad_norm": 0.3912805914878845, "learning_rate": 8.255839257217467e-05, "loss": 0.0548, "step": 4909 }, { "epoch": 0.532307025151778, "grad_norm": 0.30250653624534607, "learning_rate": 8.255476570433773e-05, "loss": 0.0324, "step": 4910 }, { "epoch": 0.5324154379878577, "grad_norm": 0.2757471799850464, "learning_rate": 8.25511388365008e-05, "loss": 0.0384, "step": 4911 }, { "epoch": 0.5325238508239376, "grad_norm": 0.4727843105792999, "learning_rate": 8.254751196866387e-05, "loss": 0.0416, "step": 4912 }, { "epoch": 0.5326322636600174, "grad_norm": 0.3889321982860565, "learning_rate": 8.254388510082694e-05, "loss": 0.0358, "step": 4913 }, { "epoch": 0.5327406764960971, "grad_norm": 0.5336938500404358, "learning_rate": 8.254025823299e-05, "loss": 0.068, "step": 4914 }, { "epoch": 0.5328490893321769, "grad_norm": 0.20398303866386414, "learning_rate": 8.253663136515307e-05, "loss": 0.0241, "step": 4915 }, { "epoch": 0.5329575021682568, "grad_norm": 0.28098785877227783, "learning_rate": 8.253300449731613e-05, "loss": 0.0457, "step": 4916 }, { "epoch": 0.5330659150043365, "grad_norm": 0.37075838446617126, "learning_rate": 8.252937762947918e-05, "loss": 0.0677, "step": 4917 }, { "epoch": 0.5331743278404163, "grad_norm": 0.43254655599594116, "learning_rate": 8.252575076164225e-05, "loss": 0.0482, "step": 4918 }, { "epoch": 0.5332827406764961, "grad_norm": 0.5158696174621582, "learning_rate": 8.252212389380531e-05, "loss": 0.1103, "step": 4919 }, { "epoch": 0.5333911535125759, "grad_norm": 0.25578808784484863, "learning_rate": 8.251849702596838e-05, "loss": 0.0261, "step": 4920 }, { "epoch": 0.5334995663486557, "grad_norm": 0.2519551217556, "learning_rate": 8.251487015813144e-05, "loss": 0.0396, "step": 4921 }, { "epoch": 0.5336079791847355, "grad_norm": 0.1940820813179016, "learning_rate": 8.251124329029451e-05, "loss": 0.0216, "step": 4922 }, { "epoch": 0.5337163920208152, "grad_norm": 0.46884259581565857, "learning_rate": 8.250761642245757e-05, "loss": 0.0592, "step": 4923 }, { "epoch": 0.5338248048568951, "grad_norm": 0.2078745812177658, "learning_rate": 8.250398955462062e-05, "loss": 0.025, "step": 4924 }, { "epoch": 0.5339332176929749, "grad_norm": 0.7430997490882874, "learning_rate": 8.25003626867837e-05, "loss": 0.0874, "step": 4925 }, { "epoch": 0.5340416305290546, "grad_norm": 0.4778417944908142, "learning_rate": 8.249673581894675e-05, "loss": 0.0639, "step": 4926 }, { "epoch": 0.5341500433651344, "grad_norm": 0.48730412125587463, "learning_rate": 8.249310895110982e-05, "loss": 0.0746, "step": 4927 }, { "epoch": 0.5342584562012143, "grad_norm": 0.22737590968608856, "learning_rate": 8.248948208327288e-05, "loss": 0.064, "step": 4928 }, { "epoch": 0.534366869037294, "grad_norm": 0.4976080358028412, "learning_rate": 8.248585521543597e-05, "loss": 0.0897, "step": 4929 }, { "epoch": 0.5344752818733738, "grad_norm": 0.24014943838119507, "learning_rate": 8.248222834759902e-05, "loss": 0.0438, "step": 4930 }, { "epoch": 0.5345836947094535, "grad_norm": 0.11461769044399261, "learning_rate": 8.247860147976208e-05, "loss": 0.0219, "step": 4931 }, { "epoch": 0.5346921075455334, "grad_norm": 0.33684685826301575, "learning_rate": 8.247497461192515e-05, "loss": 0.0653, "step": 4932 }, { "epoch": 0.5348005203816132, "grad_norm": 0.1399817317724228, "learning_rate": 8.247134774408821e-05, "loss": 0.0241, "step": 4933 }, { "epoch": 0.534908933217693, "grad_norm": 0.10693899542093277, "learning_rate": 8.246772087625128e-05, "loss": 0.0094, "step": 4934 }, { "epoch": 0.5350173460537727, "grad_norm": 0.3880739212036133, "learning_rate": 8.246409400841434e-05, "loss": 0.0471, "step": 4935 }, { "epoch": 0.5351257588898526, "grad_norm": 0.29808565974235535, "learning_rate": 8.246046714057741e-05, "loss": 0.0503, "step": 4936 }, { "epoch": 0.5352341717259324, "grad_norm": 0.615374743938446, "learning_rate": 8.245684027274046e-05, "loss": 0.0727, "step": 4937 }, { "epoch": 0.5353425845620121, "grad_norm": 0.15696373581886292, "learning_rate": 8.245321340490354e-05, "loss": 0.0241, "step": 4938 }, { "epoch": 0.535450997398092, "grad_norm": 0.2775857746601105, "learning_rate": 8.244958653706659e-05, "loss": 0.035, "step": 4939 }, { "epoch": 0.5355594102341718, "grad_norm": 0.19249369204044342, "learning_rate": 8.244595966922965e-05, "loss": 0.0261, "step": 4940 }, { "epoch": 0.5356678230702515, "grad_norm": 0.4052659273147583, "learning_rate": 8.244233280139272e-05, "loss": 0.0266, "step": 4941 }, { "epoch": 0.5357762359063313, "grad_norm": 1.041006326675415, "learning_rate": 8.243870593355578e-05, "loss": 0.0451, "step": 4942 }, { "epoch": 0.5358846487424112, "grad_norm": 0.21035541594028473, "learning_rate": 8.243507906571885e-05, "loss": 0.0374, "step": 4943 }, { "epoch": 0.5359930615784909, "grad_norm": 0.2831461727619171, "learning_rate": 8.24314521978819e-05, "loss": 0.0478, "step": 4944 }, { "epoch": 0.5361014744145707, "grad_norm": 0.44008779525756836, "learning_rate": 8.242782533004498e-05, "loss": 0.0591, "step": 4945 }, { "epoch": 0.5362098872506504, "grad_norm": 0.33847013115882874, "learning_rate": 8.242419846220805e-05, "loss": 0.0376, "step": 4946 }, { "epoch": 0.5363183000867303, "grad_norm": 0.26841461658477783, "learning_rate": 8.24205715943711e-05, "loss": 0.0257, "step": 4947 }, { "epoch": 0.5364267129228101, "grad_norm": 0.31552061438560486, "learning_rate": 8.241694472653418e-05, "loss": 0.0323, "step": 4948 }, { "epoch": 0.5365351257588898, "grad_norm": 0.23732315003871918, "learning_rate": 8.241331785869723e-05, "loss": 0.02, "step": 4949 }, { "epoch": 0.5366435385949696, "grad_norm": 0.4361787438392639, "learning_rate": 8.24096909908603e-05, "loss": 0.0584, "step": 4950 }, { "epoch": 0.5367519514310495, "grad_norm": 0.3173162639141083, "learning_rate": 8.240606412302336e-05, "loss": 0.0449, "step": 4951 }, { "epoch": 0.5368603642671292, "grad_norm": 0.19194839894771576, "learning_rate": 8.240243725518643e-05, "loss": 0.0195, "step": 4952 }, { "epoch": 0.536968777103209, "grad_norm": 0.41419604420661926, "learning_rate": 8.239881038734949e-05, "loss": 0.0333, "step": 4953 }, { "epoch": 0.5370771899392888, "grad_norm": 0.4378613233566284, "learning_rate": 8.239518351951255e-05, "loss": 0.0472, "step": 4954 }, { "epoch": 0.5371856027753686, "grad_norm": 0.379518061876297, "learning_rate": 8.239155665167562e-05, "loss": 0.0399, "step": 4955 }, { "epoch": 0.5372940156114484, "grad_norm": 0.40027138590812683, "learning_rate": 8.238792978383868e-05, "loss": 0.0442, "step": 4956 }, { "epoch": 0.5374024284475282, "grad_norm": 0.18122133612632751, "learning_rate": 8.238430291600175e-05, "loss": 0.0234, "step": 4957 }, { "epoch": 0.5375108412836079, "grad_norm": 0.42233964800834656, "learning_rate": 8.23806760481648e-05, "loss": 0.089, "step": 4958 }, { "epoch": 0.5376192541196878, "grad_norm": 0.38276898860931396, "learning_rate": 8.237704918032787e-05, "loss": 0.0421, "step": 4959 }, { "epoch": 0.5377276669557676, "grad_norm": 0.1596709042787552, "learning_rate": 8.237342231249093e-05, "loss": 0.0111, "step": 4960 }, { "epoch": 0.5378360797918473, "grad_norm": 0.2636162042617798, "learning_rate": 8.2369795444654e-05, "loss": 0.0303, "step": 4961 }, { "epoch": 0.5379444926279271, "grad_norm": 0.593039333820343, "learning_rate": 8.236616857681706e-05, "loss": 0.1033, "step": 4962 }, { "epoch": 0.538052905464007, "grad_norm": 0.373393177986145, "learning_rate": 8.236254170898012e-05, "loss": 0.0315, "step": 4963 }, { "epoch": 0.5381613183000867, "grad_norm": 0.4823669493198395, "learning_rate": 8.23589148411432e-05, "loss": 0.0424, "step": 4964 }, { "epoch": 0.5382697311361665, "grad_norm": 0.29050761461257935, "learning_rate": 8.235528797330626e-05, "loss": 0.0308, "step": 4965 }, { "epoch": 0.5383781439722463, "grad_norm": 0.16097140312194824, "learning_rate": 8.235166110546933e-05, "loss": 0.0195, "step": 4966 }, { "epoch": 0.5384865568083261, "grad_norm": 0.4251975417137146, "learning_rate": 8.234803423763239e-05, "loss": 0.0722, "step": 4967 }, { "epoch": 0.5385949696444059, "grad_norm": 0.3136286437511444, "learning_rate": 8.234440736979544e-05, "loss": 0.0315, "step": 4968 }, { "epoch": 0.5387033824804857, "grad_norm": 0.3004434108734131, "learning_rate": 8.234078050195852e-05, "loss": 0.0223, "step": 4969 }, { "epoch": 0.5388117953165655, "grad_norm": 1.241472840309143, "learning_rate": 8.233715363412157e-05, "loss": 0.064, "step": 4970 }, { "epoch": 0.5389202081526453, "grad_norm": 0.4126659035682678, "learning_rate": 8.233352676628464e-05, "loss": 0.0201, "step": 4971 }, { "epoch": 0.5390286209887251, "grad_norm": 1.346846342086792, "learning_rate": 8.23298998984477e-05, "loss": 0.0629, "step": 4972 }, { "epoch": 0.5391370338248048, "grad_norm": 0.2903381586074829, "learning_rate": 8.232627303061077e-05, "loss": 0.036, "step": 4973 }, { "epoch": 0.5392454466608847, "grad_norm": 0.4029359817504883, "learning_rate": 8.232264616277383e-05, "loss": 0.0445, "step": 4974 }, { "epoch": 0.5393538594969645, "grad_norm": 0.2981298267841339, "learning_rate": 8.23190192949369e-05, "loss": 0.047, "step": 4975 }, { "epoch": 0.5394622723330442, "grad_norm": 0.4745813012123108, "learning_rate": 8.231539242709996e-05, "loss": 0.0778, "step": 4976 }, { "epoch": 0.539570685169124, "grad_norm": 0.27538925409317017, "learning_rate": 8.231176555926302e-05, "loss": 0.0558, "step": 4977 }, { "epoch": 0.5396790980052039, "grad_norm": 0.5440868139266968, "learning_rate": 8.230813869142609e-05, "loss": 0.038, "step": 4978 }, { "epoch": 0.5397875108412836, "grad_norm": 0.2947275638580322, "learning_rate": 8.230451182358914e-05, "loss": 0.0447, "step": 4979 }, { "epoch": 0.5398959236773634, "grad_norm": 0.31677037477493286, "learning_rate": 8.230088495575221e-05, "loss": 0.0703, "step": 4980 }, { "epoch": 0.5400043365134432, "grad_norm": 0.11648275703191757, "learning_rate": 8.229725808791529e-05, "loss": 0.0087, "step": 4981 }, { "epoch": 0.540112749349523, "grad_norm": 0.29393526911735535, "learning_rate": 8.229363122007834e-05, "loss": 0.061, "step": 4982 }, { "epoch": 0.5402211621856028, "grad_norm": 0.18457134068012238, "learning_rate": 8.229000435224141e-05, "loss": 0.0214, "step": 4983 }, { "epoch": 0.5403295750216826, "grad_norm": 0.13070450723171234, "learning_rate": 8.228637748440447e-05, "loss": 0.0162, "step": 4984 }, { "epoch": 0.5404379878577623, "grad_norm": 0.17517946660518646, "learning_rate": 8.228275061656754e-05, "loss": 0.0172, "step": 4985 }, { "epoch": 0.5405464006938422, "grad_norm": 0.21157993376255035, "learning_rate": 8.22791237487306e-05, "loss": 0.0338, "step": 4986 }, { "epoch": 0.540654813529922, "grad_norm": 0.38553375005722046, "learning_rate": 8.227549688089367e-05, "loss": 0.0568, "step": 4987 }, { "epoch": 0.5407632263660017, "grad_norm": 0.2817651331424713, "learning_rate": 8.227187001305673e-05, "loss": 0.0572, "step": 4988 }, { "epoch": 0.5408716392020815, "grad_norm": 0.08102432638406754, "learning_rate": 8.22682431452198e-05, "loss": 0.0079, "step": 4989 }, { "epoch": 0.5409800520381614, "grad_norm": 0.11019136756658554, "learning_rate": 8.226461627738286e-05, "loss": 0.0183, "step": 4990 }, { "epoch": 0.5410884648742411, "grad_norm": 0.20897488296031952, "learning_rate": 8.226098940954591e-05, "loss": 0.0298, "step": 4991 }, { "epoch": 0.5411968777103209, "grad_norm": 0.24629797041416168, "learning_rate": 8.225736254170898e-05, "loss": 0.0277, "step": 4992 }, { "epoch": 0.5413052905464006, "grad_norm": 0.6979343891143799, "learning_rate": 8.225373567387204e-05, "loss": 0.0693, "step": 4993 }, { "epoch": 0.5414137033824805, "grad_norm": 0.7898422479629517, "learning_rate": 8.225010880603511e-05, "loss": 0.1098, "step": 4994 }, { "epoch": 0.5415221162185603, "grad_norm": 0.07762346416711807, "learning_rate": 8.224648193819817e-05, "loss": 0.0145, "step": 4995 }, { "epoch": 0.54163052905464, "grad_norm": 0.444163054227829, "learning_rate": 8.224285507036124e-05, "loss": 0.0832, "step": 4996 }, { "epoch": 0.5417389418907199, "grad_norm": 0.6052323579788208, "learning_rate": 8.22392282025243e-05, "loss": 0.0664, "step": 4997 }, { "epoch": 0.5418473547267997, "grad_norm": 0.5439999103546143, "learning_rate": 8.223560133468737e-05, "loss": 0.0188, "step": 4998 }, { "epoch": 0.5419557675628794, "grad_norm": 0.27097436785697937, "learning_rate": 8.223197446685044e-05, "loss": 0.0308, "step": 4999 }, { "epoch": 0.5420641803989592, "grad_norm": 0.867646336555481, "learning_rate": 8.22283475990135e-05, "loss": 0.086, "step": 5000 }, { "epoch": 0.5421725932350391, "grad_norm": 0.32789406180381775, "learning_rate": 8.222472073117657e-05, "loss": 0.0427, "step": 5001 }, { "epoch": 0.5422810060711188, "grad_norm": 0.34510838985443115, "learning_rate": 8.222109386333962e-05, "loss": 0.0458, "step": 5002 }, { "epoch": 0.5423894189071986, "grad_norm": 0.08183885365724564, "learning_rate": 8.22174669955027e-05, "loss": 0.0058, "step": 5003 }, { "epoch": 0.5424978317432784, "grad_norm": 0.1734461486339569, "learning_rate": 8.221384012766575e-05, "loss": 0.0199, "step": 5004 }, { "epoch": 0.5426062445793582, "grad_norm": 0.16748104989528656, "learning_rate": 8.221021325982881e-05, "loss": 0.0198, "step": 5005 }, { "epoch": 0.542714657415438, "grad_norm": 0.42103105783462524, "learning_rate": 8.220658639199188e-05, "loss": 0.0544, "step": 5006 }, { "epoch": 0.5428230702515178, "grad_norm": 0.17801780998706818, "learning_rate": 8.220295952415494e-05, "loss": 0.0361, "step": 5007 }, { "epoch": 0.5429314830875975, "grad_norm": 0.17073124647140503, "learning_rate": 8.219933265631801e-05, "loss": 0.0176, "step": 5008 }, { "epoch": 0.5430398959236774, "grad_norm": 0.344300240278244, "learning_rate": 8.219570578848107e-05, "loss": 0.0381, "step": 5009 }, { "epoch": 0.5431483087597572, "grad_norm": 0.39851993322372437, "learning_rate": 8.219207892064414e-05, "loss": 0.039, "step": 5010 }, { "epoch": 0.5432567215958369, "grad_norm": 0.44091057777404785, "learning_rate": 8.21884520528072e-05, "loss": 0.0554, "step": 5011 }, { "epoch": 0.5433651344319167, "grad_norm": 0.3830838203430176, "learning_rate": 8.218482518497027e-05, "loss": 0.0534, "step": 5012 }, { "epoch": 0.5434735472679966, "grad_norm": 0.6394455432891846, "learning_rate": 8.218119831713332e-05, "loss": 0.0558, "step": 5013 }, { "epoch": 0.5435819601040763, "grad_norm": 0.49459153413772583, "learning_rate": 8.217757144929638e-05, "loss": 0.0573, "step": 5014 }, { "epoch": 0.5436903729401561, "grad_norm": 0.4699114263057709, "learning_rate": 8.217394458145946e-05, "loss": 0.0633, "step": 5015 }, { "epoch": 0.5437987857762359, "grad_norm": 0.3870169222354889, "learning_rate": 8.217031771362252e-05, "loss": 0.0334, "step": 5016 }, { "epoch": 0.5439071986123157, "grad_norm": 0.4779084622859955, "learning_rate": 8.216669084578559e-05, "loss": 0.0369, "step": 5017 }, { "epoch": 0.5440156114483955, "grad_norm": 0.38136231899261475, "learning_rate": 8.216306397794865e-05, "loss": 0.04, "step": 5018 }, { "epoch": 0.5441240242844753, "grad_norm": 0.3032535910606384, "learning_rate": 8.215943711011172e-05, "loss": 0.0537, "step": 5019 }, { "epoch": 0.544232437120555, "grad_norm": 0.349844753742218, "learning_rate": 8.215581024227478e-05, "loss": 0.0422, "step": 5020 }, { "epoch": 0.5443408499566349, "grad_norm": 0.3320123255252838, "learning_rate": 8.215218337443784e-05, "loss": 0.0314, "step": 5021 }, { "epoch": 0.5444492627927147, "grad_norm": 0.15971319377422333, "learning_rate": 8.21485565066009e-05, "loss": 0.0146, "step": 5022 }, { "epoch": 0.5445576756287944, "grad_norm": 0.41879406571388245, "learning_rate": 8.214492963876396e-05, "loss": 0.0552, "step": 5023 }, { "epoch": 0.5446660884648742, "grad_norm": 0.1754482537508011, "learning_rate": 8.214130277092703e-05, "loss": 0.0171, "step": 5024 }, { "epoch": 0.5447745013009541, "grad_norm": 0.3337056636810303, "learning_rate": 8.213767590309009e-05, "loss": 0.0354, "step": 5025 }, { "epoch": 0.5448829141370338, "grad_norm": 0.3314250707626343, "learning_rate": 8.213404903525316e-05, "loss": 0.0415, "step": 5026 }, { "epoch": 0.5449913269731136, "grad_norm": 0.34103870391845703, "learning_rate": 8.213042216741622e-05, "loss": 0.0244, "step": 5027 }, { "epoch": 0.5450997398091935, "grad_norm": 0.4408910572528839, "learning_rate": 8.212679529957928e-05, "loss": 0.0483, "step": 5028 }, { "epoch": 0.5452081526452732, "grad_norm": 0.307861328125, "learning_rate": 8.212316843174235e-05, "loss": 0.0493, "step": 5029 }, { "epoch": 0.545316565481353, "grad_norm": 0.5053229928016663, "learning_rate": 8.21195415639054e-05, "loss": 0.0633, "step": 5030 }, { "epoch": 0.5454249783174328, "grad_norm": 0.31059280037879944, "learning_rate": 8.211591469606848e-05, "loss": 0.0436, "step": 5031 }, { "epoch": 0.5455333911535126, "grad_norm": 0.18052968382835388, "learning_rate": 8.211228782823153e-05, "loss": 0.0351, "step": 5032 }, { "epoch": 0.5456418039895924, "grad_norm": 0.1350274235010147, "learning_rate": 8.210866096039462e-05, "loss": 0.0128, "step": 5033 }, { "epoch": 0.5457502168256722, "grad_norm": 0.6477147340774536, "learning_rate": 8.210503409255768e-05, "loss": 0.0429, "step": 5034 }, { "epoch": 0.5458586296617519, "grad_norm": 0.39807695150375366, "learning_rate": 8.210140722472073e-05, "loss": 0.0554, "step": 5035 }, { "epoch": 0.5459670424978318, "grad_norm": 0.5343183279037476, "learning_rate": 8.20977803568838e-05, "loss": 0.0448, "step": 5036 }, { "epoch": 0.5460754553339116, "grad_norm": 0.6684103608131409, "learning_rate": 8.209415348904686e-05, "loss": 0.0237, "step": 5037 }, { "epoch": 0.5461838681699913, "grad_norm": 0.5436792373657227, "learning_rate": 8.209052662120993e-05, "loss": 0.0659, "step": 5038 }, { "epoch": 0.5462922810060711, "grad_norm": 0.19376814365386963, "learning_rate": 8.208689975337299e-05, "loss": 0.0208, "step": 5039 }, { "epoch": 0.546400693842151, "grad_norm": 0.6329449415206909, "learning_rate": 8.208327288553606e-05, "loss": 0.0422, "step": 5040 }, { "epoch": 0.5465091066782307, "grad_norm": 0.4908706545829773, "learning_rate": 8.207964601769912e-05, "loss": 0.0474, "step": 5041 }, { "epoch": 0.5466175195143105, "grad_norm": 0.4015680253505707, "learning_rate": 8.207601914986219e-05, "loss": 0.0338, "step": 5042 }, { "epoch": 0.5467259323503902, "grad_norm": 0.4434688985347748, "learning_rate": 8.207239228202525e-05, "loss": 0.0438, "step": 5043 }, { "epoch": 0.5468343451864701, "grad_norm": 0.5352384448051453, "learning_rate": 8.20687654141883e-05, "loss": 0.056, "step": 5044 }, { "epoch": 0.5469427580225499, "grad_norm": 0.298116534948349, "learning_rate": 8.206513854635137e-05, "loss": 0.0277, "step": 5045 }, { "epoch": 0.5470511708586296, "grad_norm": 0.2869872450828552, "learning_rate": 8.206151167851443e-05, "loss": 0.0367, "step": 5046 }, { "epoch": 0.5471595836947094, "grad_norm": 0.20966880023479462, "learning_rate": 8.20578848106775e-05, "loss": 0.0238, "step": 5047 }, { "epoch": 0.5472679965307893, "grad_norm": 0.5370872616767883, "learning_rate": 8.205425794284056e-05, "loss": 0.0756, "step": 5048 }, { "epoch": 0.547376409366869, "grad_norm": 0.16967104375362396, "learning_rate": 8.205063107500363e-05, "loss": 0.0107, "step": 5049 }, { "epoch": 0.5474848222029488, "grad_norm": 0.2740628123283386, "learning_rate": 8.20470042071667e-05, "loss": 0.0425, "step": 5050 }, { "epoch": 0.5475932350390286, "grad_norm": 0.5039361119270325, "learning_rate": 8.204337733932976e-05, "loss": 0.0662, "step": 5051 }, { "epoch": 0.5477016478751084, "grad_norm": 0.5258177518844604, "learning_rate": 8.203975047149283e-05, "loss": 0.0353, "step": 5052 }, { "epoch": 0.5478100607111882, "grad_norm": 0.18076163530349731, "learning_rate": 8.203612360365589e-05, "loss": 0.0176, "step": 5053 }, { "epoch": 0.547918473547268, "grad_norm": 0.5156627297401428, "learning_rate": 8.203249673581896e-05, "loss": 0.0763, "step": 5054 }, { "epoch": 0.5480268863833477, "grad_norm": 0.2761991620063782, "learning_rate": 8.202886986798202e-05, "loss": 0.0382, "step": 5055 }, { "epoch": 0.5481352992194276, "grad_norm": 0.21451446413993835, "learning_rate": 8.202524300014509e-05, "loss": 0.031, "step": 5056 }, { "epoch": 0.5482437120555074, "grad_norm": 0.37310656905174255, "learning_rate": 8.202161613230814e-05, "loss": 0.0233, "step": 5057 }, { "epoch": 0.5483521248915871, "grad_norm": 0.29201215505599976, "learning_rate": 8.20179892644712e-05, "loss": 0.0433, "step": 5058 }, { "epoch": 0.548460537727667, "grad_norm": 0.22238881886005402, "learning_rate": 8.201436239663427e-05, "loss": 0.0158, "step": 5059 }, { "epoch": 0.5485689505637468, "grad_norm": 0.4832070767879486, "learning_rate": 8.201073552879733e-05, "loss": 0.0426, "step": 5060 }, { "epoch": 0.5486773633998265, "grad_norm": 0.5519117712974548, "learning_rate": 8.20071086609604e-05, "loss": 0.0171, "step": 5061 }, { "epoch": 0.5487857762359063, "grad_norm": 0.38845083117485046, "learning_rate": 8.200348179312346e-05, "loss": 0.0346, "step": 5062 }, { "epoch": 0.5488941890719862, "grad_norm": 0.41535425186157227, "learning_rate": 8.199985492528653e-05, "loss": 0.032, "step": 5063 }, { "epoch": 0.5490026019080659, "grad_norm": 0.14953716099262238, "learning_rate": 8.199622805744959e-05, "loss": 0.0109, "step": 5064 }, { "epoch": 0.5491110147441457, "grad_norm": 0.5400201082229614, "learning_rate": 8.199260118961266e-05, "loss": 0.0395, "step": 5065 }, { "epoch": 0.5492194275802255, "grad_norm": 0.16935484111309052, "learning_rate": 8.198897432177571e-05, "loss": 0.0275, "step": 5066 }, { "epoch": 0.5493278404163053, "grad_norm": 0.3899924159049988, "learning_rate": 8.198534745393878e-05, "loss": 0.0633, "step": 5067 }, { "epoch": 0.5494362532523851, "grad_norm": 0.4441688358783722, "learning_rate": 8.198172058610186e-05, "loss": 0.064, "step": 5068 }, { "epoch": 0.5495446660884649, "grad_norm": 0.4040725827217102, "learning_rate": 8.197809371826491e-05, "loss": 0.0582, "step": 5069 }, { "epoch": 0.5496530789245446, "grad_norm": 0.2668318450450897, "learning_rate": 8.197446685042798e-05, "loss": 0.0165, "step": 5070 }, { "epoch": 0.5497614917606245, "grad_norm": 0.8016176819801331, "learning_rate": 8.197083998259104e-05, "loss": 0.0465, "step": 5071 }, { "epoch": 0.5498699045967043, "grad_norm": 0.33423662185668945, "learning_rate": 8.19672131147541e-05, "loss": 0.0367, "step": 5072 }, { "epoch": 0.549978317432784, "grad_norm": 0.16830101609230042, "learning_rate": 8.196358624691717e-05, "loss": 0.0154, "step": 5073 }, { "epoch": 0.5500867302688638, "grad_norm": 0.33327892422676086, "learning_rate": 8.195995937908023e-05, "loss": 0.0131, "step": 5074 }, { "epoch": 0.5501951431049437, "grad_norm": 0.34975844621658325, "learning_rate": 8.19563325112433e-05, "loss": 0.0371, "step": 5075 }, { "epoch": 0.5503035559410234, "grad_norm": 0.8620101809501648, "learning_rate": 8.195270564340635e-05, "loss": 0.1127, "step": 5076 }, { "epoch": 0.5504119687771032, "grad_norm": 0.22272294759750366, "learning_rate": 8.194907877556943e-05, "loss": 0.0294, "step": 5077 }, { "epoch": 0.550520381613183, "grad_norm": 0.3920259475708008, "learning_rate": 8.194545190773248e-05, "loss": 0.0425, "step": 5078 }, { "epoch": 0.5506287944492628, "grad_norm": 0.4344950020313263, "learning_rate": 8.194182503989555e-05, "loss": 0.0426, "step": 5079 }, { "epoch": 0.5507372072853426, "grad_norm": 0.913858950138092, "learning_rate": 8.193819817205861e-05, "loss": 0.0242, "step": 5080 }, { "epoch": 0.5508456201214224, "grad_norm": 0.24013380706310272, "learning_rate": 8.193457130422167e-05, "loss": 0.0431, "step": 5081 }, { "epoch": 0.5509540329575021, "grad_norm": 0.32359328866004944, "learning_rate": 8.193094443638474e-05, "loss": 0.0673, "step": 5082 }, { "epoch": 0.551062445793582, "grad_norm": 0.18561100959777832, "learning_rate": 8.19273175685478e-05, "loss": 0.0209, "step": 5083 }, { "epoch": 0.5511708586296618, "grad_norm": 0.3106892704963684, "learning_rate": 8.192369070071087e-05, "loss": 0.031, "step": 5084 }, { "epoch": 0.5512792714657415, "grad_norm": 0.28742191195487976, "learning_rate": 8.192006383287394e-05, "loss": 0.0295, "step": 5085 }, { "epoch": 0.5513876843018214, "grad_norm": 0.5545744299888611, "learning_rate": 8.1916436965037e-05, "loss": 0.0557, "step": 5086 }, { "epoch": 0.5514960971379012, "grad_norm": 0.34328827261924744, "learning_rate": 8.191281009720007e-05, "loss": 0.0373, "step": 5087 }, { "epoch": 0.5516045099739809, "grad_norm": 0.15153783559799194, "learning_rate": 8.190918322936312e-05, "loss": 0.016, "step": 5088 }, { "epoch": 0.5517129228100607, "grad_norm": 0.3392362892627716, "learning_rate": 8.19055563615262e-05, "loss": 0.0338, "step": 5089 }, { "epoch": 0.5518213356461406, "grad_norm": 0.15043283998966217, "learning_rate": 8.190192949368925e-05, "loss": 0.0298, "step": 5090 }, { "epoch": 0.5519297484822203, "grad_norm": 0.5373132228851318, "learning_rate": 8.189830262585232e-05, "loss": 0.0655, "step": 5091 }, { "epoch": 0.5520381613183001, "grad_norm": 0.5646593570709229, "learning_rate": 8.189467575801538e-05, "loss": 0.0609, "step": 5092 }, { "epoch": 0.5521465741543798, "grad_norm": 0.22249601781368256, "learning_rate": 8.189104889017845e-05, "loss": 0.0302, "step": 5093 }, { "epoch": 0.5522549869904597, "grad_norm": 0.2993588149547577, "learning_rate": 8.188742202234151e-05, "loss": 0.027, "step": 5094 }, { "epoch": 0.5523633998265395, "grad_norm": 0.6744606494903564, "learning_rate": 8.188379515450457e-05, "loss": 0.0716, "step": 5095 }, { "epoch": 0.5524718126626192, "grad_norm": 0.35377344489097595, "learning_rate": 8.188016828666764e-05, "loss": 0.0419, "step": 5096 }, { "epoch": 0.552580225498699, "grad_norm": 0.17962487041950226, "learning_rate": 8.18765414188307e-05, "loss": 0.0075, "step": 5097 }, { "epoch": 0.5526886383347789, "grad_norm": 0.23123113811016083, "learning_rate": 8.187291455099377e-05, "loss": 0.0406, "step": 5098 }, { "epoch": 0.5527970511708586, "grad_norm": 0.4744049608707428, "learning_rate": 8.186928768315682e-05, "loss": 0.0425, "step": 5099 }, { "epoch": 0.5529054640069384, "grad_norm": 0.521194577217102, "learning_rate": 8.18656608153199e-05, "loss": 0.0335, "step": 5100 }, { "epoch": 0.5530138768430182, "grad_norm": 0.4267083406448364, "learning_rate": 8.186203394748295e-05, "loss": 0.0327, "step": 5101 }, { "epoch": 0.553122289679098, "grad_norm": 0.8724076747894287, "learning_rate": 8.185840707964602e-05, "loss": 0.0611, "step": 5102 }, { "epoch": 0.5532307025151778, "grad_norm": 0.1923256367444992, "learning_rate": 8.185478021180909e-05, "loss": 0.0243, "step": 5103 }, { "epoch": 0.5533391153512576, "grad_norm": 0.4630774259567261, "learning_rate": 8.185115334397215e-05, "loss": 0.0351, "step": 5104 }, { "epoch": 0.5534475281873373, "grad_norm": 0.2893059551715851, "learning_rate": 8.184752647613522e-05, "loss": 0.028, "step": 5105 }, { "epoch": 0.5535559410234172, "grad_norm": 0.2442486435174942, "learning_rate": 8.184389960829828e-05, "loss": 0.0459, "step": 5106 }, { "epoch": 0.553664353859497, "grad_norm": 0.6842983961105347, "learning_rate": 8.184027274046135e-05, "loss": 0.0743, "step": 5107 }, { "epoch": 0.5537727666955767, "grad_norm": 0.3280923366546631, "learning_rate": 8.18366458726244e-05, "loss": 0.0427, "step": 5108 }, { "epoch": 0.5538811795316565, "grad_norm": 0.5934504866600037, "learning_rate": 8.183301900478746e-05, "loss": 0.0516, "step": 5109 }, { "epoch": 0.5539895923677364, "grad_norm": 0.4056662321090698, "learning_rate": 8.182939213695053e-05, "loss": 0.0369, "step": 5110 }, { "epoch": 0.5540980052038161, "grad_norm": 0.3523842394351959, "learning_rate": 8.182576526911359e-05, "loss": 0.025, "step": 5111 }, { "epoch": 0.5542064180398959, "grad_norm": 0.6038060188293457, "learning_rate": 8.182213840127666e-05, "loss": 0.0692, "step": 5112 }, { "epoch": 0.5543148308759757, "grad_norm": 1.4925142526626587, "learning_rate": 8.181851153343972e-05, "loss": 0.0566, "step": 5113 }, { "epoch": 0.5544232437120555, "grad_norm": 0.13379134237766266, "learning_rate": 8.181488466560279e-05, "loss": 0.0161, "step": 5114 }, { "epoch": 0.5545316565481353, "grad_norm": 0.34219005703926086, "learning_rate": 8.181125779776585e-05, "loss": 0.0501, "step": 5115 }, { "epoch": 0.5546400693842151, "grad_norm": 0.3732353150844574, "learning_rate": 8.180763092992892e-05, "loss": 0.0597, "step": 5116 }, { "epoch": 0.5547484822202949, "grad_norm": 0.7330708503723145, "learning_rate": 8.180400406209198e-05, "loss": 0.071, "step": 5117 }, { "epoch": 0.5548568950563747, "grad_norm": 0.2612982988357544, "learning_rate": 8.180037719425503e-05, "loss": 0.0485, "step": 5118 }, { "epoch": 0.5549653078924545, "grad_norm": 0.3185209035873413, "learning_rate": 8.179675032641812e-05, "loss": 0.0359, "step": 5119 }, { "epoch": 0.5550737207285342, "grad_norm": 0.22991159558296204, "learning_rate": 8.179312345858118e-05, "loss": 0.0225, "step": 5120 }, { "epoch": 0.5551821335646141, "grad_norm": 0.3042062222957611, "learning_rate": 8.178949659074425e-05, "loss": 0.0259, "step": 5121 }, { "epoch": 0.5552905464006939, "grad_norm": 0.19615057110786438, "learning_rate": 8.17858697229073e-05, "loss": 0.0287, "step": 5122 }, { "epoch": 0.5553989592367736, "grad_norm": 0.7046289443969727, "learning_rate": 8.178224285507037e-05, "loss": 0.046, "step": 5123 }, { "epoch": 0.5555073720728534, "grad_norm": 0.43483954668045044, "learning_rate": 8.177861598723343e-05, "loss": 0.052, "step": 5124 }, { "epoch": 0.5556157849089333, "grad_norm": 0.4245229959487915, "learning_rate": 8.177498911939649e-05, "loss": 0.0731, "step": 5125 }, { "epoch": 0.555724197745013, "grad_norm": 0.801103949546814, "learning_rate": 8.177136225155956e-05, "loss": 0.0448, "step": 5126 }, { "epoch": 0.5558326105810928, "grad_norm": 0.6884977221488953, "learning_rate": 8.176773538372262e-05, "loss": 0.063, "step": 5127 }, { "epoch": 0.5559410234171726, "grad_norm": 0.4101681709289551, "learning_rate": 8.176410851588569e-05, "loss": 0.0362, "step": 5128 }, { "epoch": 0.5560494362532524, "grad_norm": 0.41564157605171204, "learning_rate": 8.176048164804875e-05, "loss": 0.0441, "step": 5129 }, { "epoch": 0.5561578490893322, "grad_norm": 0.34099337458610535, "learning_rate": 8.175685478021182e-05, "loss": 0.062, "step": 5130 }, { "epoch": 0.556266261925412, "grad_norm": 0.3829379081726074, "learning_rate": 8.175322791237487e-05, "loss": 0.0967, "step": 5131 }, { "epoch": 0.5563746747614917, "grad_norm": 0.14962241053581238, "learning_rate": 8.174960104453794e-05, "loss": 0.0176, "step": 5132 }, { "epoch": 0.5564830875975716, "grad_norm": 0.26255378127098083, "learning_rate": 8.1745974176701e-05, "loss": 0.0472, "step": 5133 }, { "epoch": 0.5565915004336514, "grad_norm": 0.34437501430511475, "learning_rate": 8.174234730886406e-05, "loss": 0.0235, "step": 5134 }, { "epoch": 0.5566999132697311, "grad_norm": 0.47438737750053406, "learning_rate": 8.173872044102713e-05, "loss": 0.0551, "step": 5135 }, { "epoch": 0.5568083261058109, "grad_norm": 0.2691137492656708, "learning_rate": 8.17350935731902e-05, "loss": 0.0451, "step": 5136 }, { "epoch": 0.5569167389418908, "grad_norm": 0.41192907094955444, "learning_rate": 8.173146670535327e-05, "loss": 0.0762, "step": 5137 }, { "epoch": 0.5570251517779705, "grad_norm": 0.1995760202407837, "learning_rate": 8.172783983751633e-05, "loss": 0.0293, "step": 5138 }, { "epoch": 0.5571335646140503, "grad_norm": 0.34721410274505615, "learning_rate": 8.172421296967939e-05, "loss": 0.047, "step": 5139 }, { "epoch": 0.55724197745013, "grad_norm": 0.33145564794540405, "learning_rate": 8.172058610184246e-05, "loss": 0.0364, "step": 5140 }, { "epoch": 0.5573503902862099, "grad_norm": 0.3314403295516968, "learning_rate": 8.171695923400551e-05, "loss": 0.0262, "step": 5141 }, { "epoch": 0.5574588031222897, "grad_norm": 0.3114314675331116, "learning_rate": 8.171333236616859e-05, "loss": 0.0438, "step": 5142 }, { "epoch": 0.5575672159583694, "grad_norm": 0.2669835388660431, "learning_rate": 8.170970549833164e-05, "loss": 0.0289, "step": 5143 }, { "epoch": 0.5576756287944492, "grad_norm": 0.3364713490009308, "learning_rate": 8.170607863049471e-05, "loss": 0.0435, "step": 5144 }, { "epoch": 0.5577840416305291, "grad_norm": 0.4769963026046753, "learning_rate": 8.170245176265777e-05, "loss": 0.0534, "step": 5145 }, { "epoch": 0.5578924544666088, "grad_norm": 0.36168330907821655, "learning_rate": 8.169882489482084e-05, "loss": 0.0511, "step": 5146 }, { "epoch": 0.5580008673026886, "grad_norm": 0.28954023122787476, "learning_rate": 8.16951980269839e-05, "loss": 0.0388, "step": 5147 }, { "epoch": 0.5581092801387685, "grad_norm": 0.4121232032775879, "learning_rate": 8.169157115914696e-05, "loss": 0.0485, "step": 5148 }, { "epoch": 0.5582176929748482, "grad_norm": 0.3439463675022125, "learning_rate": 8.168794429131003e-05, "loss": 0.0621, "step": 5149 }, { "epoch": 0.558326105810928, "grad_norm": 0.155788853764534, "learning_rate": 8.168431742347308e-05, "loss": 0.0146, "step": 5150 }, { "epoch": 0.5584345186470078, "grad_norm": 0.35717859864234924, "learning_rate": 8.168069055563616e-05, "loss": 0.0273, "step": 5151 }, { "epoch": 0.5585429314830876, "grad_norm": 0.2648393213748932, "learning_rate": 8.167706368779921e-05, "loss": 0.0279, "step": 5152 }, { "epoch": 0.5586513443191674, "grad_norm": 0.34052687883377075, "learning_rate": 8.167343681996228e-05, "loss": 0.0604, "step": 5153 }, { "epoch": 0.5587597571552472, "grad_norm": 0.3412224352359772, "learning_rate": 8.166980995212535e-05, "loss": 0.0925, "step": 5154 }, { "epoch": 0.5588681699913269, "grad_norm": 0.18122850358486176, "learning_rate": 8.166618308428841e-05, "loss": 0.0219, "step": 5155 }, { "epoch": 0.5589765828274068, "grad_norm": 0.36370864510536194, "learning_rate": 8.166255621645148e-05, "loss": 0.0708, "step": 5156 }, { "epoch": 0.5590849956634866, "grad_norm": 0.35321053862571716, "learning_rate": 8.165892934861454e-05, "loss": 0.0244, "step": 5157 }, { "epoch": 0.5591934084995663, "grad_norm": 0.26643553376197815, "learning_rate": 8.165530248077761e-05, "loss": 0.0458, "step": 5158 }, { "epoch": 0.5593018213356461, "grad_norm": 0.3367779850959778, "learning_rate": 8.165167561294067e-05, "loss": 0.0555, "step": 5159 }, { "epoch": 0.559410234171726, "grad_norm": 0.2842009961605072, "learning_rate": 8.164804874510374e-05, "loss": 0.0273, "step": 5160 }, { "epoch": 0.5595186470078057, "grad_norm": 0.5032281875610352, "learning_rate": 8.16444218772668e-05, "loss": 0.05, "step": 5161 }, { "epoch": 0.5596270598438855, "grad_norm": 0.27411243319511414, "learning_rate": 8.164079500942985e-05, "loss": 0.0255, "step": 5162 }, { "epoch": 0.5597354726799653, "grad_norm": 0.3939918279647827, "learning_rate": 8.163716814159293e-05, "loss": 0.0388, "step": 5163 }, { "epoch": 0.5598438855160451, "grad_norm": 0.633403480052948, "learning_rate": 8.163354127375598e-05, "loss": 0.0881, "step": 5164 }, { "epoch": 0.5599522983521249, "grad_norm": 0.35579702258110046, "learning_rate": 8.162991440591905e-05, "loss": 0.0442, "step": 5165 }, { "epoch": 0.5600607111882047, "grad_norm": 0.5493823885917664, "learning_rate": 8.162628753808211e-05, "loss": 0.0521, "step": 5166 }, { "epoch": 0.5601691240242844, "grad_norm": 1.2548588514328003, "learning_rate": 8.162266067024518e-05, "loss": 0.0493, "step": 5167 }, { "epoch": 0.5602775368603643, "grad_norm": 0.22046242654323578, "learning_rate": 8.161903380240824e-05, "loss": 0.0213, "step": 5168 }, { "epoch": 0.5603859496964441, "grad_norm": 0.2888741195201874, "learning_rate": 8.161540693457131e-05, "loss": 0.0264, "step": 5169 }, { "epoch": 0.5604943625325238, "grad_norm": 0.7293456196784973, "learning_rate": 8.161178006673437e-05, "loss": 0.0341, "step": 5170 }, { "epoch": 0.5606027753686036, "grad_norm": 0.5671135783195496, "learning_rate": 8.160815319889744e-05, "loss": 0.061, "step": 5171 }, { "epoch": 0.5607111882046835, "grad_norm": 0.3907470405101776, "learning_rate": 8.160452633106051e-05, "loss": 0.0456, "step": 5172 }, { "epoch": 0.5608196010407632, "grad_norm": 0.4122657775878906, "learning_rate": 8.160089946322357e-05, "loss": 0.0617, "step": 5173 }, { "epoch": 0.560928013876843, "grad_norm": 0.1681167632341385, "learning_rate": 8.159727259538664e-05, "loss": 0.014, "step": 5174 }, { "epoch": 0.5610364267129229, "grad_norm": 0.5000792741775513, "learning_rate": 8.15936457275497e-05, "loss": 0.0573, "step": 5175 }, { "epoch": 0.5611448395490026, "grad_norm": 0.2515805959701538, "learning_rate": 8.159001885971275e-05, "loss": 0.0414, "step": 5176 }, { "epoch": 0.5612532523850824, "grad_norm": 0.24355009198188782, "learning_rate": 8.158639199187582e-05, "loss": 0.0222, "step": 5177 }, { "epoch": 0.5613616652211622, "grad_norm": 0.2624334692955017, "learning_rate": 8.158276512403888e-05, "loss": 0.0199, "step": 5178 }, { "epoch": 0.561470078057242, "grad_norm": 0.6400004029273987, "learning_rate": 8.157913825620195e-05, "loss": 0.0529, "step": 5179 }, { "epoch": 0.5615784908933218, "grad_norm": 0.362374484539032, "learning_rate": 8.157551138836501e-05, "loss": 0.0338, "step": 5180 }, { "epoch": 0.5616869037294016, "grad_norm": 0.47874099016189575, "learning_rate": 8.157188452052808e-05, "loss": 0.0305, "step": 5181 }, { "epoch": 0.5617953165654813, "grad_norm": 0.4526420831680298, "learning_rate": 8.156825765269114e-05, "loss": 0.0231, "step": 5182 }, { "epoch": 0.5619037294015612, "grad_norm": 0.2225600630044937, "learning_rate": 8.156463078485421e-05, "loss": 0.0457, "step": 5183 }, { "epoch": 0.562012142237641, "grad_norm": 0.43415895104408264, "learning_rate": 8.156100391701726e-05, "loss": 0.0466, "step": 5184 }, { "epoch": 0.5621205550737207, "grad_norm": 0.43482738733291626, "learning_rate": 8.155737704918032e-05, "loss": 0.0166, "step": 5185 }, { "epoch": 0.5622289679098005, "grad_norm": 0.18619127571582794, "learning_rate": 8.155375018134339e-05, "loss": 0.0298, "step": 5186 }, { "epoch": 0.5623373807458804, "grad_norm": 0.6006852984428406, "learning_rate": 8.155012331350645e-05, "loss": 0.0426, "step": 5187 }, { "epoch": 0.5624457935819601, "grad_norm": 0.4210429787635803, "learning_rate": 8.154649644566953e-05, "loss": 0.0579, "step": 5188 }, { "epoch": 0.5625542064180399, "grad_norm": 0.2056678831577301, "learning_rate": 8.154286957783259e-05, "loss": 0.0199, "step": 5189 }, { "epoch": 0.5626626192541196, "grad_norm": 0.27605706453323364, "learning_rate": 8.153924270999565e-05, "loss": 0.0389, "step": 5190 }, { "epoch": 0.5627710320901995, "grad_norm": 0.3131779432296753, "learning_rate": 8.153561584215872e-05, "loss": 0.0418, "step": 5191 }, { "epoch": 0.5628794449262793, "grad_norm": 0.28942811489105225, "learning_rate": 8.153198897432178e-05, "loss": 0.037, "step": 5192 }, { "epoch": 0.562987857762359, "grad_norm": 0.43721655011177063, "learning_rate": 8.152836210648485e-05, "loss": 0.0672, "step": 5193 }, { "epoch": 0.5630962705984388, "grad_norm": 0.2739056646823883, "learning_rate": 8.15247352386479e-05, "loss": 0.0261, "step": 5194 }, { "epoch": 0.5632046834345187, "grad_norm": 0.2604607939720154, "learning_rate": 8.152110837081098e-05, "loss": 0.0241, "step": 5195 }, { "epoch": 0.5633130962705984, "grad_norm": 0.4086896479129791, "learning_rate": 8.151748150297403e-05, "loss": 0.0501, "step": 5196 }, { "epoch": 0.5634215091066782, "grad_norm": 0.4575456976890564, "learning_rate": 8.15138546351371e-05, "loss": 0.0749, "step": 5197 }, { "epoch": 0.563529921942758, "grad_norm": 0.8145149350166321, "learning_rate": 8.151022776730016e-05, "loss": 0.1079, "step": 5198 }, { "epoch": 0.5636383347788378, "grad_norm": 0.14889393746852875, "learning_rate": 8.150660089946322e-05, "loss": 0.0178, "step": 5199 }, { "epoch": 0.5637467476149176, "grad_norm": 0.4623258709907532, "learning_rate": 8.150297403162629e-05, "loss": 0.0621, "step": 5200 }, { "epoch": 0.5638551604509974, "grad_norm": 0.8794292211532593, "learning_rate": 8.149934716378935e-05, "loss": 0.0533, "step": 5201 }, { "epoch": 0.5639635732870771, "grad_norm": 0.6369359493255615, "learning_rate": 8.149572029595242e-05, "loss": 0.0786, "step": 5202 }, { "epoch": 0.564071986123157, "grad_norm": 0.43529900908470154, "learning_rate": 8.149209342811548e-05, "loss": 0.0367, "step": 5203 }, { "epoch": 0.5641803989592368, "grad_norm": 0.3689357340335846, "learning_rate": 8.148846656027855e-05, "loss": 0.0323, "step": 5204 }, { "epoch": 0.5642888117953165, "grad_norm": 0.608870267868042, "learning_rate": 8.148483969244162e-05, "loss": 0.1123, "step": 5205 }, { "epoch": 0.5643972246313964, "grad_norm": 0.10428805649280548, "learning_rate": 8.148121282460467e-05, "loss": 0.012, "step": 5206 }, { "epoch": 0.5645056374674762, "grad_norm": 0.5452612042427063, "learning_rate": 8.147758595676775e-05, "loss": 0.0817, "step": 5207 }, { "epoch": 0.5646140503035559, "grad_norm": 0.42221423983573914, "learning_rate": 8.14739590889308e-05, "loss": 0.0672, "step": 5208 }, { "epoch": 0.5647224631396357, "grad_norm": 0.7578380107879639, "learning_rate": 8.147033222109387e-05, "loss": 0.0697, "step": 5209 }, { "epoch": 0.5648308759757156, "grad_norm": 0.22025060653686523, "learning_rate": 8.146670535325693e-05, "loss": 0.0364, "step": 5210 }, { "epoch": 0.5649392888117953, "grad_norm": 0.2132735699415207, "learning_rate": 8.146307848542e-05, "loss": 0.027, "step": 5211 }, { "epoch": 0.5650477016478751, "grad_norm": 0.3899618685245514, "learning_rate": 8.145945161758306e-05, "loss": 0.0616, "step": 5212 }, { "epoch": 0.5651561144839549, "grad_norm": 0.8507844805717468, "learning_rate": 8.145582474974612e-05, "loss": 0.1238, "step": 5213 }, { "epoch": 0.5652645273200347, "grad_norm": 0.2415657788515091, "learning_rate": 8.145219788190919e-05, "loss": 0.0439, "step": 5214 }, { "epoch": 0.5653729401561145, "grad_norm": 0.364347368478775, "learning_rate": 8.144857101407224e-05, "loss": 0.04, "step": 5215 }, { "epoch": 0.5654813529921943, "grad_norm": 0.6466849446296692, "learning_rate": 8.144494414623532e-05, "loss": 0.0649, "step": 5216 }, { "epoch": 0.565589765828274, "grad_norm": 0.38078776001930237, "learning_rate": 8.144131727839837e-05, "loss": 0.0511, "step": 5217 }, { "epoch": 0.5656981786643539, "grad_norm": 0.1553592085838318, "learning_rate": 8.143769041056144e-05, "loss": 0.0218, "step": 5218 }, { "epoch": 0.5658065915004337, "grad_norm": 0.28276801109313965, "learning_rate": 8.14340635427245e-05, "loss": 0.0235, "step": 5219 }, { "epoch": 0.5659150043365134, "grad_norm": 0.5572587251663208, "learning_rate": 8.143043667488757e-05, "loss": 0.0725, "step": 5220 }, { "epoch": 0.5660234171725932, "grad_norm": 0.29369843006134033, "learning_rate": 8.142680980705063e-05, "loss": 0.0388, "step": 5221 }, { "epoch": 0.5661318300086731, "grad_norm": 0.2680255174636841, "learning_rate": 8.142318293921369e-05, "loss": 0.0363, "step": 5222 }, { "epoch": 0.5662402428447528, "grad_norm": 0.30848950147628784, "learning_rate": 8.141955607137677e-05, "loss": 0.0356, "step": 5223 }, { "epoch": 0.5663486556808326, "grad_norm": 0.26619306206703186, "learning_rate": 8.141592920353983e-05, "loss": 0.0239, "step": 5224 }, { "epoch": 0.5664570685169124, "grad_norm": 0.216560959815979, "learning_rate": 8.14123023357029e-05, "loss": 0.0287, "step": 5225 }, { "epoch": 0.5665654813529922, "grad_norm": 0.2761397063732147, "learning_rate": 8.140867546786596e-05, "loss": 0.0379, "step": 5226 }, { "epoch": 0.566673894189072, "grad_norm": 0.19510270655155182, "learning_rate": 8.140504860002903e-05, "loss": 0.0289, "step": 5227 }, { "epoch": 0.5667823070251518, "grad_norm": 0.24641691148281097, "learning_rate": 8.140142173219209e-05, "loss": 0.0474, "step": 5228 }, { "epoch": 0.5668907198612315, "grad_norm": 0.35403454303741455, "learning_rate": 8.139779486435514e-05, "loss": 0.0331, "step": 5229 }, { "epoch": 0.5669991326973114, "grad_norm": 0.20535406470298767, "learning_rate": 8.139416799651821e-05, "loss": 0.0264, "step": 5230 }, { "epoch": 0.5671075455333912, "grad_norm": 0.3518349528312683, "learning_rate": 8.139054112868127e-05, "loss": 0.0641, "step": 5231 }, { "epoch": 0.5672159583694709, "grad_norm": 0.20470575988292694, "learning_rate": 8.138691426084434e-05, "loss": 0.0238, "step": 5232 }, { "epoch": 0.5673243712055508, "grad_norm": 0.3551022708415985, "learning_rate": 8.13832873930074e-05, "loss": 0.0515, "step": 5233 }, { "epoch": 0.5674327840416306, "grad_norm": 0.3214525282382965, "learning_rate": 8.137966052517047e-05, "loss": 0.0485, "step": 5234 }, { "epoch": 0.5675411968777103, "grad_norm": 0.21679313480854034, "learning_rate": 8.137603365733353e-05, "loss": 0.0199, "step": 5235 }, { "epoch": 0.5676496097137901, "grad_norm": 0.2807379961013794, "learning_rate": 8.13724067894966e-05, "loss": 0.0391, "step": 5236 }, { "epoch": 0.56775802254987, "grad_norm": 0.30150648951530457, "learning_rate": 8.136877992165966e-05, "loss": 0.04, "step": 5237 }, { "epoch": 0.5678664353859497, "grad_norm": 0.25270456075668335, "learning_rate": 8.136515305382271e-05, "loss": 0.0639, "step": 5238 }, { "epoch": 0.5679748482220295, "grad_norm": 0.13822445273399353, "learning_rate": 8.136152618598578e-05, "loss": 0.0157, "step": 5239 }, { "epoch": 0.5680832610581092, "grad_norm": 0.42414504289627075, "learning_rate": 8.135789931814885e-05, "loss": 0.0532, "step": 5240 }, { "epoch": 0.5681916738941891, "grad_norm": 0.251887708902359, "learning_rate": 8.135427245031193e-05, "loss": 0.0366, "step": 5241 }, { "epoch": 0.5683000867302689, "grad_norm": 0.1552140861749649, "learning_rate": 8.135064558247498e-05, "loss": 0.0166, "step": 5242 }, { "epoch": 0.5684084995663486, "grad_norm": 0.10394520312547684, "learning_rate": 8.134701871463804e-05, "loss": 0.0078, "step": 5243 }, { "epoch": 0.5685169124024284, "grad_norm": 0.15910464525222778, "learning_rate": 8.134339184680111e-05, "loss": 0.0154, "step": 5244 }, { "epoch": 0.5686253252385083, "grad_norm": 0.09337130934000015, "learning_rate": 8.133976497896417e-05, "loss": 0.0139, "step": 5245 }, { "epoch": 0.568733738074588, "grad_norm": 0.2226593941450119, "learning_rate": 8.133613811112724e-05, "loss": 0.0351, "step": 5246 }, { "epoch": 0.5688421509106678, "grad_norm": 0.3156743347644806, "learning_rate": 8.13325112432903e-05, "loss": 0.0374, "step": 5247 }, { "epoch": 0.5689505637467476, "grad_norm": 0.2789621651172638, "learning_rate": 8.132888437545337e-05, "loss": 0.0309, "step": 5248 }, { "epoch": 0.5690589765828274, "grad_norm": 0.3831719160079956, "learning_rate": 8.132525750761642e-05, "loss": 0.0498, "step": 5249 }, { "epoch": 0.5691673894189072, "grad_norm": 0.549778938293457, "learning_rate": 8.13216306397795e-05, "loss": 0.027, "step": 5250 }, { "epoch": 0.569275802254987, "grad_norm": 0.3584209382534027, "learning_rate": 8.131800377194255e-05, "loss": 0.0299, "step": 5251 }, { "epoch": 0.5693842150910667, "grad_norm": 0.40739575028419495, "learning_rate": 8.131437690410561e-05, "loss": 0.0338, "step": 5252 }, { "epoch": 0.5694926279271466, "grad_norm": 0.780498206615448, "learning_rate": 8.131075003626868e-05, "loss": 0.0735, "step": 5253 }, { "epoch": 0.5696010407632264, "grad_norm": 3.34157395362854, "learning_rate": 8.130712316843174e-05, "loss": 0.6556, "step": 5254 }, { "epoch": 0.5697094535993061, "grad_norm": 0.7005985379219055, "learning_rate": 8.130349630059481e-05, "loss": 0.0451, "step": 5255 }, { "epoch": 0.5698178664353859, "grad_norm": 0.22002968192100525, "learning_rate": 8.129986943275787e-05, "loss": 0.0201, "step": 5256 }, { "epoch": 0.5699262792714658, "grad_norm": 0.33384785056114197, "learning_rate": 8.129624256492094e-05, "loss": 0.0309, "step": 5257 }, { "epoch": 0.5700346921075455, "grad_norm": 0.42263638973236084, "learning_rate": 8.129261569708401e-05, "loss": 0.0323, "step": 5258 }, { "epoch": 0.5701431049436253, "grad_norm": 0.16745230555534363, "learning_rate": 8.128898882924707e-05, "loss": 0.0147, "step": 5259 }, { "epoch": 0.5702515177797051, "grad_norm": 0.4315076768398285, "learning_rate": 8.128536196141014e-05, "loss": 0.0781, "step": 5260 }, { "epoch": 0.5703599306157849, "grad_norm": 0.45592349767684937, "learning_rate": 8.12817350935732e-05, "loss": 0.034, "step": 5261 }, { "epoch": 0.5704683434518647, "grad_norm": 0.5336349606513977, "learning_rate": 8.127810822573626e-05, "loss": 0.0484, "step": 5262 }, { "epoch": 0.5705767562879445, "grad_norm": 0.43486177921295166, "learning_rate": 8.127448135789932e-05, "loss": 0.0282, "step": 5263 }, { "epoch": 0.5706851691240243, "grad_norm": 0.5471165776252747, "learning_rate": 8.127085449006239e-05, "loss": 0.0538, "step": 5264 }, { "epoch": 0.5707935819601041, "grad_norm": 0.2901839315891266, "learning_rate": 8.126722762222545e-05, "loss": 0.0288, "step": 5265 }, { "epoch": 0.5709019947961839, "grad_norm": 0.9281606674194336, "learning_rate": 8.126360075438851e-05, "loss": 0.0761, "step": 5266 }, { "epoch": 0.5710104076322636, "grad_norm": 0.23855696618556976, "learning_rate": 8.125997388655158e-05, "loss": 0.0186, "step": 5267 }, { "epoch": 0.5711188204683435, "grad_norm": 0.5163277387619019, "learning_rate": 8.125634701871464e-05, "loss": 0.0489, "step": 5268 }, { "epoch": 0.5712272333044233, "grad_norm": 0.3724072575569153, "learning_rate": 8.12527201508777e-05, "loss": 0.0301, "step": 5269 }, { "epoch": 0.571335646140503, "grad_norm": 0.8863410949707031, "learning_rate": 8.124909328304076e-05, "loss": 0.0507, "step": 5270 }, { "epoch": 0.5714440589765828, "grad_norm": 0.45652031898498535, "learning_rate": 8.124546641520383e-05, "loss": 0.0578, "step": 5271 }, { "epoch": 0.5715524718126627, "grad_norm": 0.7618046402931213, "learning_rate": 8.124183954736689e-05, "loss": 0.0766, "step": 5272 }, { "epoch": 0.5716608846487424, "grad_norm": 0.493345707654953, "learning_rate": 8.123821267952996e-05, "loss": 0.0416, "step": 5273 }, { "epoch": 0.5717692974848222, "grad_norm": 0.31541022658348083, "learning_rate": 8.123458581169302e-05, "loss": 0.0267, "step": 5274 }, { "epoch": 0.571877710320902, "grad_norm": 0.8504419326782227, "learning_rate": 8.123095894385609e-05, "loss": 0.081, "step": 5275 }, { "epoch": 0.5719861231569818, "grad_norm": 0.5397558212280273, "learning_rate": 8.122733207601916e-05, "loss": 0.0704, "step": 5276 }, { "epoch": 0.5720945359930616, "grad_norm": 0.46310290694236755, "learning_rate": 8.122370520818222e-05, "loss": 0.0428, "step": 5277 }, { "epoch": 0.5722029488291414, "grad_norm": 0.3081398904323578, "learning_rate": 8.122007834034529e-05, "loss": 0.0454, "step": 5278 }, { "epoch": 0.5723113616652211, "grad_norm": 0.19788631796836853, "learning_rate": 8.121645147250835e-05, "loss": 0.0476, "step": 5279 }, { "epoch": 0.572419774501301, "grad_norm": 0.3389436900615692, "learning_rate": 8.12128246046714e-05, "loss": 0.0517, "step": 5280 }, { "epoch": 0.5725281873373808, "grad_norm": 0.30260321497917175, "learning_rate": 8.120919773683448e-05, "loss": 0.0376, "step": 5281 }, { "epoch": 0.5726366001734605, "grad_norm": 0.4231724143028259, "learning_rate": 8.120557086899753e-05, "loss": 0.061, "step": 5282 }, { "epoch": 0.5727450130095403, "grad_norm": 0.18182598054409027, "learning_rate": 8.12019440011606e-05, "loss": 0.03, "step": 5283 }, { "epoch": 0.5728534258456202, "grad_norm": 0.20291687548160553, "learning_rate": 8.119831713332366e-05, "loss": 0.0472, "step": 5284 }, { "epoch": 0.5729618386816999, "grad_norm": 0.30724114179611206, "learning_rate": 8.119469026548673e-05, "loss": 0.0331, "step": 5285 }, { "epoch": 0.5730702515177797, "grad_norm": 0.21707382798194885, "learning_rate": 8.119106339764979e-05, "loss": 0.03, "step": 5286 }, { "epoch": 0.5731786643538594, "grad_norm": 0.24200914800167084, "learning_rate": 8.118743652981286e-05, "loss": 0.0322, "step": 5287 }, { "epoch": 0.5732870771899393, "grad_norm": 0.17946119606494904, "learning_rate": 8.118380966197592e-05, "loss": 0.053, "step": 5288 }, { "epoch": 0.5733954900260191, "grad_norm": 0.4800375998020172, "learning_rate": 8.118018279413898e-05, "loss": 0.0618, "step": 5289 }, { "epoch": 0.5735039028620988, "grad_norm": 0.4436498284339905, "learning_rate": 8.117655592630205e-05, "loss": 0.0717, "step": 5290 }, { "epoch": 0.5736123156981786, "grad_norm": 0.22119303047657013, "learning_rate": 8.11729290584651e-05, "loss": 0.0399, "step": 5291 }, { "epoch": 0.5737207285342585, "grad_norm": 0.3156571090221405, "learning_rate": 8.116930219062819e-05, "loss": 0.0423, "step": 5292 }, { "epoch": 0.5738291413703382, "grad_norm": 0.14509502053260803, "learning_rate": 8.116567532279125e-05, "loss": 0.0234, "step": 5293 }, { "epoch": 0.573937554206418, "grad_norm": 0.42925578355789185, "learning_rate": 8.11620484549543e-05, "loss": 0.0797, "step": 5294 }, { "epoch": 0.5740459670424979, "grad_norm": 0.5813557505607605, "learning_rate": 8.115842158711737e-05, "loss": 0.0876, "step": 5295 }, { "epoch": 0.5741543798785776, "grad_norm": 0.19237706065177917, "learning_rate": 8.115479471928043e-05, "loss": 0.0307, "step": 5296 }, { "epoch": 0.5742627927146574, "grad_norm": 0.2941599488258362, "learning_rate": 8.11511678514435e-05, "loss": 0.0544, "step": 5297 }, { "epoch": 0.5743712055507372, "grad_norm": 0.5153059363365173, "learning_rate": 8.114754098360656e-05, "loss": 0.0353, "step": 5298 }, { "epoch": 0.574479618386817, "grad_norm": 0.5087245106697083, "learning_rate": 8.114391411576963e-05, "loss": 0.0527, "step": 5299 }, { "epoch": 0.5745880312228968, "grad_norm": 0.3234323561191559, "learning_rate": 8.114028724793269e-05, "loss": 0.0742, "step": 5300 }, { "epoch": 0.5746964440589766, "grad_norm": 0.5051529407501221, "learning_rate": 8.113666038009576e-05, "loss": 0.0806, "step": 5301 }, { "epoch": 0.5748048568950563, "grad_norm": 0.4303998351097107, "learning_rate": 8.113303351225882e-05, "loss": 0.0411, "step": 5302 }, { "epoch": 0.5749132697311362, "grad_norm": 0.27349910140037537, "learning_rate": 8.112940664442187e-05, "loss": 0.0355, "step": 5303 }, { "epoch": 0.575021682567216, "grad_norm": 0.5378819704055786, "learning_rate": 8.112577977658494e-05, "loss": 0.0386, "step": 5304 }, { "epoch": 0.5751300954032957, "grad_norm": 0.6328377723693848, "learning_rate": 8.1122152908748e-05, "loss": 0.0634, "step": 5305 }, { "epoch": 0.5752385082393755, "grad_norm": 0.4864225387573242, "learning_rate": 8.111852604091107e-05, "loss": 0.0658, "step": 5306 }, { "epoch": 0.5753469210754554, "grad_norm": 0.5014964938163757, "learning_rate": 8.111489917307413e-05, "loss": 0.0538, "step": 5307 }, { "epoch": 0.5754553339115351, "grad_norm": 0.40009549260139465, "learning_rate": 8.11112723052372e-05, "loss": 0.0446, "step": 5308 }, { "epoch": 0.5755637467476149, "grad_norm": 0.25853192806243896, "learning_rate": 8.110764543740027e-05, "loss": 0.0276, "step": 5309 }, { "epoch": 0.5756721595836947, "grad_norm": 0.2321384996175766, "learning_rate": 8.110401856956333e-05, "loss": 0.0585, "step": 5310 }, { "epoch": 0.5757805724197745, "grad_norm": 0.23395463824272156, "learning_rate": 8.11003917017264e-05, "loss": 0.0395, "step": 5311 }, { "epoch": 0.5758889852558543, "grad_norm": 0.5169625282287598, "learning_rate": 8.109676483388946e-05, "loss": 0.0728, "step": 5312 }, { "epoch": 0.5759973980919341, "grad_norm": 0.18028078973293304, "learning_rate": 8.109313796605253e-05, "loss": 0.0187, "step": 5313 }, { "epoch": 0.5761058109280138, "grad_norm": 0.43448013067245483, "learning_rate": 8.108951109821558e-05, "loss": 0.0427, "step": 5314 }, { "epoch": 0.5762142237640937, "grad_norm": 0.177777960896492, "learning_rate": 8.108588423037866e-05, "loss": 0.0323, "step": 5315 }, { "epoch": 0.5763226366001735, "grad_norm": 0.13561056554317474, "learning_rate": 8.108225736254171e-05, "loss": 0.0235, "step": 5316 }, { "epoch": 0.5764310494362532, "grad_norm": 0.11400356143712997, "learning_rate": 8.107863049470478e-05, "loss": 0.0213, "step": 5317 }, { "epoch": 0.576539462272333, "grad_norm": 0.19443513453006744, "learning_rate": 8.107500362686784e-05, "loss": 0.0237, "step": 5318 }, { "epoch": 0.5766478751084129, "grad_norm": 0.19250287115573883, "learning_rate": 8.10713767590309e-05, "loss": 0.0155, "step": 5319 }, { "epoch": 0.5767562879444926, "grad_norm": 0.4039694368839264, "learning_rate": 8.106774989119397e-05, "loss": 0.0458, "step": 5320 }, { "epoch": 0.5768647007805724, "grad_norm": 0.11983607709407806, "learning_rate": 8.106412302335703e-05, "loss": 0.0554, "step": 5321 }, { "epoch": 0.5769731136166523, "grad_norm": 0.3930826187133789, "learning_rate": 8.10604961555201e-05, "loss": 0.0393, "step": 5322 }, { "epoch": 0.577081526452732, "grad_norm": 0.522904098033905, "learning_rate": 8.105686928768315e-05, "loss": 0.0485, "step": 5323 }, { "epoch": 0.5771899392888118, "grad_norm": 0.20495429635047913, "learning_rate": 8.105324241984623e-05, "loss": 0.0207, "step": 5324 }, { "epoch": 0.5772983521248916, "grad_norm": 0.9416099190711975, "learning_rate": 8.104961555200928e-05, "loss": 0.0255, "step": 5325 }, { "epoch": 0.5774067649609714, "grad_norm": 0.4257793128490448, "learning_rate": 8.104598868417235e-05, "loss": 0.0575, "step": 5326 }, { "epoch": 0.5775151777970512, "grad_norm": 0.44332629442214966, "learning_rate": 8.104236181633542e-05, "loss": 0.0645, "step": 5327 }, { "epoch": 0.577623590633131, "grad_norm": 0.25791415572166443, "learning_rate": 8.103873494849848e-05, "loss": 0.0389, "step": 5328 }, { "epoch": 0.5777320034692107, "grad_norm": 0.5791416168212891, "learning_rate": 8.103510808066155e-05, "loss": 0.0993, "step": 5329 }, { "epoch": 0.5778404163052906, "grad_norm": 0.3116673231124878, "learning_rate": 8.103148121282461e-05, "loss": 0.0723, "step": 5330 }, { "epoch": 0.5779488291413704, "grad_norm": 0.346330463886261, "learning_rate": 8.102785434498768e-05, "loss": 0.0443, "step": 5331 }, { "epoch": 0.5780572419774501, "grad_norm": 0.3556581437587738, "learning_rate": 8.102422747715074e-05, "loss": 0.0727, "step": 5332 }, { "epoch": 0.5781656548135299, "grad_norm": 0.39226651191711426, "learning_rate": 8.10206006093138e-05, "loss": 0.036, "step": 5333 }, { "epoch": 0.5782740676496098, "grad_norm": 0.1491529494524002, "learning_rate": 8.101697374147687e-05, "loss": 0.0172, "step": 5334 }, { "epoch": 0.5783824804856895, "grad_norm": 0.2909109592437744, "learning_rate": 8.101334687363992e-05, "loss": 0.0351, "step": 5335 }, { "epoch": 0.5784908933217693, "grad_norm": 0.2945035994052887, "learning_rate": 8.1009720005803e-05, "loss": 0.0251, "step": 5336 }, { "epoch": 0.578599306157849, "grad_norm": 0.3391992747783661, "learning_rate": 8.100609313796605e-05, "loss": 0.0413, "step": 5337 }, { "epoch": 0.5787077189939289, "grad_norm": 0.23857983946800232, "learning_rate": 8.100246627012912e-05, "loss": 0.021, "step": 5338 }, { "epoch": 0.5788161318300087, "grad_norm": 0.4010336697101593, "learning_rate": 8.099883940229218e-05, "loss": 0.0427, "step": 5339 }, { "epoch": 0.5789245446660884, "grad_norm": 0.3265545964241028, "learning_rate": 8.099521253445525e-05, "loss": 0.059, "step": 5340 }, { "epoch": 0.5790329575021682, "grad_norm": 0.30485033988952637, "learning_rate": 8.099158566661831e-05, "loss": 0.0562, "step": 5341 }, { "epoch": 0.5791413703382481, "grad_norm": 0.4287644624710083, "learning_rate": 8.098795879878137e-05, "loss": 0.0416, "step": 5342 }, { "epoch": 0.5792497831743278, "grad_norm": 0.22322233021259308, "learning_rate": 8.098433193094444e-05, "loss": 0.0185, "step": 5343 }, { "epoch": 0.5793581960104076, "grad_norm": 0.3357980251312256, "learning_rate": 8.098070506310751e-05, "loss": 0.0514, "step": 5344 }, { "epoch": 0.5794666088464874, "grad_norm": 0.40520426630973816, "learning_rate": 8.097707819527058e-05, "loss": 0.0503, "step": 5345 }, { "epoch": 0.5795750216825672, "grad_norm": 0.20353379845619202, "learning_rate": 8.097345132743364e-05, "loss": 0.0353, "step": 5346 }, { "epoch": 0.579683434518647, "grad_norm": 0.4249284565448761, "learning_rate": 8.09698244595967e-05, "loss": 0.0637, "step": 5347 }, { "epoch": 0.5797918473547268, "grad_norm": 0.13695277273654938, "learning_rate": 8.096619759175976e-05, "loss": 0.006, "step": 5348 }, { "epoch": 0.5799002601908065, "grad_norm": 0.3612128496170044, "learning_rate": 8.096257072392282e-05, "loss": 0.0407, "step": 5349 }, { "epoch": 0.5800086730268864, "grad_norm": 0.23895145952701569, "learning_rate": 8.095894385608589e-05, "loss": 0.0146, "step": 5350 }, { "epoch": 0.5801170858629662, "grad_norm": 0.5493640303611755, "learning_rate": 8.095531698824895e-05, "loss": 0.1098, "step": 5351 }, { "epoch": 0.5802254986990459, "grad_norm": 0.3653828203678131, "learning_rate": 8.095169012041202e-05, "loss": 0.038, "step": 5352 }, { "epoch": 0.5803339115351258, "grad_norm": 0.4212530851364136, "learning_rate": 8.094806325257508e-05, "loss": 0.037, "step": 5353 }, { "epoch": 0.5804423243712056, "grad_norm": 0.1839897632598877, "learning_rate": 8.094443638473815e-05, "loss": 0.0168, "step": 5354 }, { "epoch": 0.5805507372072853, "grad_norm": 0.5679924488067627, "learning_rate": 8.09408095169012e-05, "loss": 0.0466, "step": 5355 }, { "epoch": 0.5806591500433651, "grad_norm": 0.19640709459781647, "learning_rate": 8.093718264906426e-05, "loss": 0.0384, "step": 5356 }, { "epoch": 0.580767562879445, "grad_norm": 0.3021010160446167, "learning_rate": 8.093355578122733e-05, "loss": 0.0534, "step": 5357 }, { "epoch": 0.5808759757155247, "grad_norm": 0.4642834961414337, "learning_rate": 8.092992891339039e-05, "loss": 0.1004, "step": 5358 }, { "epoch": 0.5809843885516045, "grad_norm": 0.2849769592285156, "learning_rate": 8.092630204555346e-05, "loss": 0.024, "step": 5359 }, { "epoch": 0.5810928013876843, "grad_norm": 0.11053113639354706, "learning_rate": 8.092267517771652e-05, "loss": 0.0229, "step": 5360 }, { "epoch": 0.5812012142237641, "grad_norm": 0.4688055217266083, "learning_rate": 8.091904830987959e-05, "loss": 0.0568, "step": 5361 }, { "epoch": 0.5813096270598439, "grad_norm": 0.34312310814857483, "learning_rate": 8.091542144204266e-05, "loss": 0.0683, "step": 5362 }, { "epoch": 0.5814180398959237, "grad_norm": 0.4092356264591217, "learning_rate": 8.091179457420572e-05, "loss": 0.0491, "step": 5363 }, { "epoch": 0.5815264527320034, "grad_norm": 0.5507818460464478, "learning_rate": 8.090816770636879e-05, "loss": 0.0535, "step": 5364 }, { "epoch": 0.5816348655680833, "grad_norm": 0.2281830608844757, "learning_rate": 8.090454083853185e-05, "loss": 0.0346, "step": 5365 }, { "epoch": 0.5817432784041631, "grad_norm": 0.37251943349838257, "learning_rate": 8.090091397069492e-05, "loss": 0.0221, "step": 5366 }, { "epoch": 0.5818516912402428, "grad_norm": 0.30811217427253723, "learning_rate": 8.089728710285798e-05, "loss": 0.0432, "step": 5367 }, { "epoch": 0.5819601040763226, "grad_norm": 0.1083659678697586, "learning_rate": 8.089366023502105e-05, "loss": 0.0212, "step": 5368 }, { "epoch": 0.5820685169124025, "grad_norm": 0.34632962942123413, "learning_rate": 8.08900333671841e-05, "loss": 0.0246, "step": 5369 }, { "epoch": 0.5821769297484822, "grad_norm": 0.29091736674308777, "learning_rate": 8.088640649934716e-05, "loss": 0.0345, "step": 5370 }, { "epoch": 0.582285342584562, "grad_norm": 0.2166033536195755, "learning_rate": 8.088277963151023e-05, "loss": 0.0265, "step": 5371 }, { "epoch": 0.5823937554206418, "grad_norm": 0.48474472761154175, "learning_rate": 8.087915276367329e-05, "loss": 0.069, "step": 5372 }, { "epoch": 0.5825021682567216, "grad_norm": 0.42109522223472595, "learning_rate": 8.087552589583636e-05, "loss": 0.0575, "step": 5373 }, { "epoch": 0.5826105810928014, "grad_norm": 0.25659918785095215, "learning_rate": 8.087189902799942e-05, "loss": 0.0241, "step": 5374 }, { "epoch": 0.5827189939288812, "grad_norm": 0.38867107033729553, "learning_rate": 8.086827216016249e-05, "loss": 0.029, "step": 5375 }, { "epoch": 0.5828274067649609, "grad_norm": 0.1746811717748642, "learning_rate": 8.086464529232555e-05, "loss": 0.019, "step": 5376 }, { "epoch": 0.5829358196010408, "grad_norm": 0.3539564609527588, "learning_rate": 8.086101842448862e-05, "loss": 0.0244, "step": 5377 }, { "epoch": 0.5830442324371206, "grad_norm": 0.3055027723312378, "learning_rate": 8.085739155665169e-05, "loss": 0.0569, "step": 5378 }, { "epoch": 0.5831526452732003, "grad_norm": 0.4492037296295166, "learning_rate": 8.085376468881474e-05, "loss": 0.0299, "step": 5379 }, { "epoch": 0.5832610581092801, "grad_norm": 0.3970131278038025, "learning_rate": 8.085013782097782e-05, "loss": 0.0674, "step": 5380 }, { "epoch": 0.58336947094536, "grad_norm": 0.4287155866622925, "learning_rate": 8.084651095314087e-05, "loss": 0.0372, "step": 5381 }, { "epoch": 0.5834778837814397, "grad_norm": 0.33700811862945557, "learning_rate": 8.084288408530394e-05, "loss": 0.017, "step": 5382 }, { "epoch": 0.5835862966175195, "grad_norm": 0.24678440392017365, "learning_rate": 8.0839257217467e-05, "loss": 0.0308, "step": 5383 }, { "epoch": 0.5836947094535994, "grad_norm": 0.15530313551425934, "learning_rate": 8.083563034963006e-05, "loss": 0.0187, "step": 5384 }, { "epoch": 0.5838031222896791, "grad_norm": 0.5577998161315918, "learning_rate": 8.083200348179313e-05, "loss": 0.0678, "step": 5385 }, { "epoch": 0.5839115351257589, "grad_norm": 0.3036440908908844, "learning_rate": 8.082837661395619e-05, "loss": 0.0621, "step": 5386 }, { "epoch": 0.5840199479618386, "grad_norm": 0.29690560698509216, "learning_rate": 8.082474974611926e-05, "loss": 0.05, "step": 5387 }, { "epoch": 0.5841283607979185, "grad_norm": 0.3753214180469513, "learning_rate": 8.082112287828231e-05, "loss": 0.0461, "step": 5388 }, { "epoch": 0.5842367736339983, "grad_norm": 0.4061228036880493, "learning_rate": 8.081749601044539e-05, "loss": 0.0237, "step": 5389 }, { "epoch": 0.584345186470078, "grad_norm": 0.15541614592075348, "learning_rate": 8.081386914260844e-05, "loss": 0.0236, "step": 5390 }, { "epoch": 0.5844535993061578, "grad_norm": 0.28563493490219116, "learning_rate": 8.081024227477151e-05, "loss": 0.0244, "step": 5391 }, { "epoch": 0.5845620121422377, "grad_norm": 0.28185704350471497, "learning_rate": 8.080661540693457e-05, "loss": 0.0274, "step": 5392 }, { "epoch": 0.5846704249783174, "grad_norm": 0.4550420045852661, "learning_rate": 8.080298853909763e-05, "loss": 0.0345, "step": 5393 }, { "epoch": 0.5847788378143972, "grad_norm": 0.28745579719543457, "learning_rate": 8.07993616712607e-05, "loss": 0.0342, "step": 5394 }, { "epoch": 0.584887250650477, "grad_norm": 0.2650185823440552, "learning_rate": 8.079573480342377e-05, "loss": 0.0552, "step": 5395 }, { "epoch": 0.5849956634865568, "grad_norm": 0.16801109910011292, "learning_rate": 8.079210793558684e-05, "loss": 0.0548, "step": 5396 }, { "epoch": 0.5851040763226366, "grad_norm": 1.738788366317749, "learning_rate": 8.07884810677499e-05, "loss": 0.0401, "step": 5397 }, { "epoch": 0.5852124891587164, "grad_norm": 0.5258959531784058, "learning_rate": 8.078485419991297e-05, "loss": 0.0508, "step": 5398 }, { "epoch": 0.5853209019947961, "grad_norm": 0.4017125070095062, "learning_rate": 8.078122733207603e-05, "loss": 0.0384, "step": 5399 }, { "epoch": 0.585429314830876, "grad_norm": 0.12017637491226196, "learning_rate": 8.077760046423908e-05, "loss": 0.0177, "step": 5400 }, { "epoch": 0.5855377276669558, "grad_norm": 0.40449392795562744, "learning_rate": 8.077397359640215e-05, "loss": 0.0367, "step": 5401 }, { "epoch": 0.5856461405030355, "grad_norm": 0.49046117067337036, "learning_rate": 8.077034672856521e-05, "loss": 0.0312, "step": 5402 }, { "epoch": 0.5857545533391153, "grad_norm": 0.12135002762079239, "learning_rate": 8.076671986072828e-05, "loss": 0.0107, "step": 5403 }, { "epoch": 0.5858629661751952, "grad_norm": 0.1473178267478943, "learning_rate": 8.076309299289134e-05, "loss": 0.0147, "step": 5404 }, { "epoch": 0.5859713790112749, "grad_norm": 0.21510154008865356, "learning_rate": 8.075946612505441e-05, "loss": 0.022, "step": 5405 }, { "epoch": 0.5860797918473547, "grad_norm": 0.736412763595581, "learning_rate": 8.075583925721747e-05, "loss": 0.0593, "step": 5406 }, { "epoch": 0.5861882046834345, "grad_norm": 0.8456293344497681, "learning_rate": 8.075221238938053e-05, "loss": 0.0747, "step": 5407 }, { "epoch": 0.5862966175195143, "grad_norm": 0.33750560879707336, "learning_rate": 8.07485855215436e-05, "loss": 0.0414, "step": 5408 }, { "epoch": 0.5864050303555941, "grad_norm": 0.30394843220710754, "learning_rate": 8.074495865370665e-05, "loss": 0.0291, "step": 5409 }, { "epoch": 0.5865134431916739, "grad_norm": 0.4295656085014343, "learning_rate": 8.074133178586973e-05, "loss": 0.0534, "step": 5410 }, { "epoch": 0.5866218560277537, "grad_norm": 0.4306900203227997, "learning_rate": 8.073770491803278e-05, "loss": 0.0328, "step": 5411 }, { "epoch": 0.5867302688638335, "grad_norm": 0.2340344935655594, "learning_rate": 8.073407805019585e-05, "loss": 0.0163, "step": 5412 }, { "epoch": 0.5868386816999133, "grad_norm": 0.25365543365478516, "learning_rate": 8.073045118235892e-05, "loss": 0.0348, "step": 5413 }, { "epoch": 0.586947094535993, "grad_norm": 0.33350327610969543, "learning_rate": 8.072682431452198e-05, "loss": 0.0478, "step": 5414 }, { "epoch": 0.5870555073720729, "grad_norm": 0.5571169257164001, "learning_rate": 8.072319744668505e-05, "loss": 0.0873, "step": 5415 }, { "epoch": 0.5871639202081527, "grad_norm": 0.2797391414642334, "learning_rate": 8.071957057884811e-05, "loss": 0.0246, "step": 5416 }, { "epoch": 0.5872723330442324, "grad_norm": 0.37990859150886536, "learning_rate": 8.071594371101118e-05, "loss": 0.0137, "step": 5417 }, { "epoch": 0.5873807458803122, "grad_norm": 0.3401181101799011, "learning_rate": 8.071231684317424e-05, "loss": 0.0429, "step": 5418 }, { "epoch": 0.5874891587163921, "grad_norm": 0.9259282350540161, "learning_rate": 8.070868997533731e-05, "loss": 0.0409, "step": 5419 }, { "epoch": 0.5875975715524718, "grad_norm": 0.6660481095314026, "learning_rate": 8.070506310750037e-05, "loss": 0.0554, "step": 5420 }, { "epoch": 0.5877059843885516, "grad_norm": 1.0822018384933472, "learning_rate": 8.070143623966344e-05, "loss": 0.0539, "step": 5421 }, { "epoch": 0.5878143972246314, "grad_norm": 0.3563940227031708, "learning_rate": 8.06978093718265e-05, "loss": 0.0226, "step": 5422 }, { "epoch": 0.5879228100607112, "grad_norm": 0.34738126397132874, "learning_rate": 8.069418250398955e-05, "loss": 0.0487, "step": 5423 }, { "epoch": 0.588031222896791, "grad_norm": 0.7508165240287781, "learning_rate": 8.069055563615262e-05, "loss": 0.0495, "step": 5424 }, { "epoch": 0.5881396357328708, "grad_norm": 0.3341361880302429, "learning_rate": 8.068692876831568e-05, "loss": 0.03, "step": 5425 }, { "epoch": 0.5882480485689505, "grad_norm": 0.4141741096973419, "learning_rate": 8.068330190047875e-05, "loss": 0.0456, "step": 5426 }, { "epoch": 0.5883564614050304, "grad_norm": 0.2716313898563385, "learning_rate": 8.067967503264181e-05, "loss": 0.0489, "step": 5427 }, { "epoch": 0.5884648742411102, "grad_norm": 0.37294018268585205, "learning_rate": 8.067604816480488e-05, "loss": 0.0612, "step": 5428 }, { "epoch": 0.5885732870771899, "grad_norm": 0.5383845567703247, "learning_rate": 8.067242129696794e-05, "loss": 0.0755, "step": 5429 }, { "epoch": 0.5886816999132697, "grad_norm": 0.6966280341148376, "learning_rate": 8.066879442913101e-05, "loss": 0.0354, "step": 5430 }, { "epoch": 0.5887901127493496, "grad_norm": 0.6592369079589844, "learning_rate": 8.066516756129408e-05, "loss": 0.0566, "step": 5431 }, { "epoch": 0.5888985255854293, "grad_norm": 0.3790893256664276, "learning_rate": 8.066154069345714e-05, "loss": 0.0608, "step": 5432 }, { "epoch": 0.5890069384215091, "grad_norm": 0.18163107335567474, "learning_rate": 8.06579138256202e-05, "loss": 0.0148, "step": 5433 }, { "epoch": 0.5891153512575888, "grad_norm": 0.334189772605896, "learning_rate": 8.065428695778326e-05, "loss": 0.0569, "step": 5434 }, { "epoch": 0.5892237640936687, "grad_norm": 0.5445743799209595, "learning_rate": 8.065066008994633e-05, "loss": 0.0206, "step": 5435 }, { "epoch": 0.5893321769297485, "grad_norm": 0.3999744653701782, "learning_rate": 8.064703322210939e-05, "loss": 0.0581, "step": 5436 }, { "epoch": 0.5894405897658282, "grad_norm": 0.1884453445672989, "learning_rate": 8.064340635427245e-05, "loss": 0.0264, "step": 5437 }, { "epoch": 0.589549002601908, "grad_norm": 0.532170832157135, "learning_rate": 8.063977948643552e-05, "loss": 0.0484, "step": 5438 }, { "epoch": 0.5896574154379879, "grad_norm": 0.6482970118522644, "learning_rate": 8.063615261859858e-05, "loss": 0.065, "step": 5439 }, { "epoch": 0.5897658282740676, "grad_norm": 0.22962751984596252, "learning_rate": 8.063252575076165e-05, "loss": 0.0223, "step": 5440 }, { "epoch": 0.5898742411101474, "grad_norm": 0.5648045539855957, "learning_rate": 8.06288988829247e-05, "loss": 0.0779, "step": 5441 }, { "epoch": 0.5899826539462273, "grad_norm": 0.12052998691797256, "learning_rate": 8.062527201508778e-05, "loss": 0.0207, "step": 5442 }, { "epoch": 0.590091066782307, "grad_norm": 0.27020063996315, "learning_rate": 8.062164514725083e-05, "loss": 0.0401, "step": 5443 }, { "epoch": 0.5901994796183868, "grad_norm": 0.1244555115699768, "learning_rate": 8.06180182794139e-05, "loss": 0.0153, "step": 5444 }, { "epoch": 0.5903078924544666, "grad_norm": 0.3031920790672302, "learning_rate": 8.061439141157696e-05, "loss": 0.0621, "step": 5445 }, { "epoch": 0.5904163052905465, "grad_norm": 0.4526123106479645, "learning_rate": 8.061076454374002e-05, "loss": 0.0515, "step": 5446 }, { "epoch": 0.5905247181266262, "grad_norm": 0.2875136435031891, "learning_rate": 8.06071376759031e-05, "loss": 0.0273, "step": 5447 }, { "epoch": 0.590633130962706, "grad_norm": 0.24240757524967194, "learning_rate": 8.060351080806616e-05, "loss": 0.0472, "step": 5448 }, { "epoch": 0.5907415437987857, "grad_norm": 0.33138707280158997, "learning_rate": 8.059988394022923e-05, "loss": 0.0311, "step": 5449 }, { "epoch": 0.5908499566348656, "grad_norm": 0.1566266566514969, "learning_rate": 8.059625707239229e-05, "loss": 0.0375, "step": 5450 }, { "epoch": 0.5909583694709454, "grad_norm": 0.16549378633499146, "learning_rate": 8.059263020455535e-05, "loss": 0.0318, "step": 5451 }, { "epoch": 0.5910667823070251, "grad_norm": 0.3711439371109009, "learning_rate": 8.058900333671842e-05, "loss": 0.0272, "step": 5452 }, { "epoch": 0.5911751951431049, "grad_norm": 0.4463425576686859, "learning_rate": 8.058537646888147e-05, "loss": 0.0704, "step": 5453 }, { "epoch": 0.5912836079791848, "grad_norm": 0.34794801473617554, "learning_rate": 8.058174960104455e-05, "loss": 0.0281, "step": 5454 }, { "epoch": 0.5913920208152645, "grad_norm": 0.2933025062084198, "learning_rate": 8.05781227332076e-05, "loss": 0.0511, "step": 5455 }, { "epoch": 0.5915004336513443, "grad_norm": 0.35325372219085693, "learning_rate": 8.057449586537067e-05, "loss": 0.0374, "step": 5456 }, { "epoch": 0.5916088464874241, "grad_norm": 0.1800861805677414, "learning_rate": 8.057086899753373e-05, "loss": 0.0246, "step": 5457 }, { "epoch": 0.5917172593235039, "grad_norm": 0.35123980045318604, "learning_rate": 8.05672421296968e-05, "loss": 0.0773, "step": 5458 }, { "epoch": 0.5918256721595837, "grad_norm": 0.19688966870307922, "learning_rate": 8.056361526185986e-05, "loss": 0.0217, "step": 5459 }, { "epoch": 0.5919340849956635, "grad_norm": 0.2608196437358856, "learning_rate": 8.055998839402292e-05, "loss": 0.0421, "step": 5460 }, { "epoch": 0.5920424978317432, "grad_norm": 0.13883034884929657, "learning_rate": 8.055636152618599e-05, "loss": 0.0268, "step": 5461 }, { "epoch": 0.5921509106678231, "grad_norm": 0.38480979204177856, "learning_rate": 8.055273465834905e-05, "loss": 0.0339, "step": 5462 }, { "epoch": 0.5922593235039029, "grad_norm": 0.2763393223285675, "learning_rate": 8.054910779051212e-05, "loss": 0.0614, "step": 5463 }, { "epoch": 0.5923677363399826, "grad_norm": 0.059098899364471436, "learning_rate": 8.054548092267517e-05, "loss": 0.0053, "step": 5464 }, { "epoch": 0.5924761491760624, "grad_norm": 0.26780009269714355, "learning_rate": 8.054185405483824e-05, "loss": 0.038, "step": 5465 }, { "epoch": 0.5925845620121423, "grad_norm": 0.12704195082187653, "learning_rate": 8.053822718700132e-05, "loss": 0.0097, "step": 5466 }, { "epoch": 0.592692974848222, "grad_norm": 0.590972900390625, "learning_rate": 8.053460031916437e-05, "loss": 0.065, "step": 5467 }, { "epoch": 0.5928013876843018, "grad_norm": 0.3259018659591675, "learning_rate": 8.053097345132744e-05, "loss": 0.0258, "step": 5468 }, { "epoch": 0.5929098005203816, "grad_norm": 0.22149577736854553, "learning_rate": 8.05273465834905e-05, "loss": 0.0199, "step": 5469 }, { "epoch": 0.5930182133564614, "grad_norm": 0.33539924025535583, "learning_rate": 8.052371971565357e-05, "loss": 0.0619, "step": 5470 }, { "epoch": 0.5931266261925412, "grad_norm": 0.6693722605705261, "learning_rate": 8.052009284781663e-05, "loss": 0.0775, "step": 5471 }, { "epoch": 0.593235039028621, "grad_norm": 0.1237543597817421, "learning_rate": 8.05164659799797e-05, "loss": 0.0084, "step": 5472 }, { "epoch": 0.5933434518647008, "grad_norm": 0.608147144317627, "learning_rate": 8.051283911214276e-05, "loss": 0.037, "step": 5473 }, { "epoch": 0.5934518647007806, "grad_norm": 0.39184993505477905, "learning_rate": 8.050921224430581e-05, "loss": 0.0863, "step": 5474 }, { "epoch": 0.5935602775368604, "grad_norm": 0.855853259563446, "learning_rate": 8.050558537646889e-05, "loss": 0.0505, "step": 5475 }, { "epoch": 0.5936686903729401, "grad_norm": 0.2314431369304657, "learning_rate": 8.050195850863194e-05, "loss": 0.0422, "step": 5476 }, { "epoch": 0.59377710320902, "grad_norm": 0.5951195955276489, "learning_rate": 8.049833164079501e-05, "loss": 0.0931, "step": 5477 }, { "epoch": 0.5938855160450998, "grad_norm": 0.3428257405757904, "learning_rate": 8.049470477295807e-05, "loss": 0.0271, "step": 5478 }, { "epoch": 0.5939939288811795, "grad_norm": 0.5598216652870178, "learning_rate": 8.049107790512114e-05, "loss": 0.092, "step": 5479 }, { "epoch": 0.5941023417172593, "grad_norm": 0.34074288606643677, "learning_rate": 8.04874510372842e-05, "loss": 0.0498, "step": 5480 }, { "epoch": 0.5942107545533392, "grad_norm": 0.3670804798603058, "learning_rate": 8.048382416944727e-05, "loss": 0.0586, "step": 5481 }, { "epoch": 0.5943191673894189, "grad_norm": 1.0873374938964844, "learning_rate": 8.048019730161034e-05, "loss": 0.0667, "step": 5482 }, { "epoch": 0.5944275802254987, "grad_norm": 0.14510478079319, "learning_rate": 8.04765704337734e-05, "loss": 0.0314, "step": 5483 }, { "epoch": 0.5945359930615784, "grad_norm": 0.537341296672821, "learning_rate": 8.047294356593647e-05, "loss": 0.0641, "step": 5484 }, { "epoch": 0.5946444058976583, "grad_norm": 0.3705018162727356, "learning_rate": 8.046931669809953e-05, "loss": 0.0388, "step": 5485 }, { "epoch": 0.5947528187337381, "grad_norm": 0.22431443631649017, "learning_rate": 8.04656898302626e-05, "loss": 0.0253, "step": 5486 }, { "epoch": 0.5948612315698178, "grad_norm": 0.24845148622989655, "learning_rate": 8.046206296242565e-05, "loss": 0.0293, "step": 5487 }, { "epoch": 0.5949696444058976, "grad_norm": 0.2427939474582672, "learning_rate": 8.045843609458871e-05, "loss": 0.0316, "step": 5488 }, { "epoch": 0.5950780572419775, "grad_norm": 0.13105812668800354, "learning_rate": 8.045480922675178e-05, "loss": 0.029, "step": 5489 }, { "epoch": 0.5951864700780573, "grad_norm": 0.19407470524311066, "learning_rate": 8.045118235891484e-05, "loss": 0.039, "step": 5490 }, { "epoch": 0.595294882914137, "grad_norm": 0.4579993784427643, "learning_rate": 8.044755549107791e-05, "loss": 0.0581, "step": 5491 }, { "epoch": 0.5954032957502168, "grad_norm": 0.24515563249588013, "learning_rate": 8.044392862324097e-05, "loss": 0.0259, "step": 5492 }, { "epoch": 0.5955117085862967, "grad_norm": 0.5379590392112732, "learning_rate": 8.044030175540404e-05, "loss": 0.0764, "step": 5493 }, { "epoch": 0.5956201214223764, "grad_norm": 0.2327466607093811, "learning_rate": 8.04366748875671e-05, "loss": 0.0338, "step": 5494 }, { "epoch": 0.5957285342584562, "grad_norm": 0.19519998133182526, "learning_rate": 8.043304801973017e-05, "loss": 0.0271, "step": 5495 }, { "epoch": 0.5958369470945359, "grad_norm": 0.21691422164440155, "learning_rate": 8.042942115189322e-05, "loss": 0.04, "step": 5496 }, { "epoch": 0.5959453599306158, "grad_norm": 0.5225193500518799, "learning_rate": 8.042579428405628e-05, "loss": 0.061, "step": 5497 }, { "epoch": 0.5960537727666956, "grad_norm": 0.21984095871448517, "learning_rate": 8.042216741621935e-05, "loss": 0.0319, "step": 5498 }, { "epoch": 0.5961621856027753, "grad_norm": 0.2997795343399048, "learning_rate": 8.041854054838242e-05, "loss": 0.0316, "step": 5499 }, { "epoch": 0.5962705984388552, "grad_norm": 0.25245222449302673, "learning_rate": 8.04149136805455e-05, "loss": 0.0396, "step": 5500 }, { "epoch": 0.596379011274935, "grad_norm": 0.29226386547088623, "learning_rate": 8.041128681270855e-05, "loss": 0.0245, "step": 5501 }, { "epoch": 0.5964874241110147, "grad_norm": 0.18324226140975952, "learning_rate": 8.040765994487162e-05, "loss": 0.0209, "step": 5502 }, { "epoch": 0.5965958369470945, "grad_norm": 0.1751331388950348, "learning_rate": 8.040403307703468e-05, "loss": 0.0216, "step": 5503 }, { "epoch": 0.5967042497831744, "grad_norm": 0.23235853016376495, "learning_rate": 8.040040620919774e-05, "loss": 0.0373, "step": 5504 }, { "epoch": 0.5968126626192541, "grad_norm": 0.1950826793909073, "learning_rate": 8.039677934136081e-05, "loss": 0.0139, "step": 5505 }, { "epoch": 0.5969210754553339, "grad_norm": 0.1482396125793457, "learning_rate": 8.039315247352387e-05, "loss": 0.037, "step": 5506 }, { "epoch": 0.5970294882914137, "grad_norm": 0.34403523802757263, "learning_rate": 8.038952560568694e-05, "loss": 0.0331, "step": 5507 }, { "epoch": 0.5971379011274935, "grad_norm": 0.45395782589912415, "learning_rate": 8.038589873785e-05, "loss": 0.0295, "step": 5508 }, { "epoch": 0.5972463139635733, "grad_norm": 0.31562599539756775, "learning_rate": 8.038227187001306e-05, "loss": 0.0182, "step": 5509 }, { "epoch": 0.5973547267996531, "grad_norm": 0.5634253025054932, "learning_rate": 8.037864500217612e-05, "loss": 0.0329, "step": 5510 }, { "epoch": 0.5974631396357328, "grad_norm": 0.20960114896297455, "learning_rate": 8.037501813433918e-05, "loss": 0.0337, "step": 5511 }, { "epoch": 0.5975715524718127, "grad_norm": 0.39058467745780945, "learning_rate": 8.037139126650225e-05, "loss": 0.0126, "step": 5512 }, { "epoch": 0.5976799653078925, "grad_norm": 0.09233947843313217, "learning_rate": 8.036776439866531e-05, "loss": 0.009, "step": 5513 }, { "epoch": 0.5977883781439722, "grad_norm": 0.3321674168109894, "learning_rate": 8.036413753082838e-05, "loss": 0.0191, "step": 5514 }, { "epoch": 0.597896790980052, "grad_norm": 0.4167500138282776, "learning_rate": 8.036051066299144e-05, "loss": 0.0573, "step": 5515 }, { "epoch": 0.5980052038161319, "grad_norm": 0.46626296639442444, "learning_rate": 8.035688379515452e-05, "loss": 0.0301, "step": 5516 }, { "epoch": 0.5981136166522116, "grad_norm": 0.25175872445106506, "learning_rate": 8.035325692731758e-05, "loss": 0.0399, "step": 5517 }, { "epoch": 0.5982220294882914, "grad_norm": 0.42160966992378235, "learning_rate": 8.034963005948063e-05, "loss": 0.0166, "step": 5518 }, { "epoch": 0.5983304423243712, "grad_norm": 0.5066404938697815, "learning_rate": 8.03460031916437e-05, "loss": 0.0697, "step": 5519 }, { "epoch": 0.598438855160451, "grad_norm": 0.22408755123615265, "learning_rate": 8.034237632380676e-05, "loss": 0.0131, "step": 5520 }, { "epoch": 0.5985472679965308, "grad_norm": 0.3542536497116089, "learning_rate": 8.033874945596983e-05, "loss": 0.0218, "step": 5521 }, { "epoch": 0.5986556808326106, "grad_norm": 0.10697367787361145, "learning_rate": 8.033512258813289e-05, "loss": 0.0041, "step": 5522 }, { "epoch": 0.5987640936686903, "grad_norm": 0.39721694588661194, "learning_rate": 8.033149572029596e-05, "loss": 0.0414, "step": 5523 }, { "epoch": 0.5988725065047702, "grad_norm": 0.3784637749195099, "learning_rate": 8.032786885245902e-05, "loss": 0.0247, "step": 5524 }, { "epoch": 0.59898091934085, "grad_norm": 0.2488410621881485, "learning_rate": 8.032424198462209e-05, "loss": 0.0173, "step": 5525 }, { "epoch": 0.5990893321769297, "grad_norm": 0.3414071500301361, "learning_rate": 8.032061511678515e-05, "loss": 0.0452, "step": 5526 }, { "epoch": 0.5991977450130095, "grad_norm": 0.348662793636322, "learning_rate": 8.03169882489482e-05, "loss": 0.0297, "step": 5527 }, { "epoch": 0.5993061578490894, "grad_norm": 0.14780667424201965, "learning_rate": 8.031336138111128e-05, "loss": 0.0139, "step": 5528 }, { "epoch": 0.5994145706851691, "grad_norm": 0.7619975209236145, "learning_rate": 8.030973451327433e-05, "loss": 0.0683, "step": 5529 }, { "epoch": 0.5995229835212489, "grad_norm": 0.2946886718273163, "learning_rate": 8.03061076454374e-05, "loss": 0.0129, "step": 5530 }, { "epoch": 0.5996313963573288, "grad_norm": 0.2122677117586136, "learning_rate": 8.030248077760046e-05, "loss": 0.0296, "step": 5531 }, { "epoch": 0.5997398091934085, "grad_norm": 0.3992992043495178, "learning_rate": 8.029885390976353e-05, "loss": 0.0461, "step": 5532 }, { "epoch": 0.5998482220294883, "grad_norm": 0.527847945690155, "learning_rate": 8.029522704192659e-05, "loss": 0.0865, "step": 5533 }, { "epoch": 0.599956634865568, "grad_norm": 0.07941339164972305, "learning_rate": 8.029160017408966e-05, "loss": 0.0102, "step": 5534 }, { "epoch": 0.6000650477016479, "grad_norm": 0.36722058057785034, "learning_rate": 8.028797330625273e-05, "loss": 0.0266, "step": 5535 }, { "epoch": 0.6001734605377277, "grad_norm": 0.37414073944091797, "learning_rate": 8.028434643841579e-05, "loss": 0.0681, "step": 5536 }, { "epoch": 0.6002818733738075, "grad_norm": 0.6249538064002991, "learning_rate": 8.028071957057886e-05, "loss": 0.0655, "step": 5537 }, { "epoch": 0.6003902862098872, "grad_norm": 0.5675567984580994, "learning_rate": 8.027709270274192e-05, "loss": 0.0589, "step": 5538 }, { "epoch": 0.6004986990459671, "grad_norm": 0.7605546116828918, "learning_rate": 8.027346583490499e-05, "loss": 0.0678, "step": 5539 }, { "epoch": 0.6006071118820469, "grad_norm": 0.3957746624946594, "learning_rate": 8.026983896706805e-05, "loss": 0.0076, "step": 5540 }, { "epoch": 0.6007155247181266, "grad_norm": 0.7911270260810852, "learning_rate": 8.02662120992311e-05, "loss": 0.0833, "step": 5541 }, { "epoch": 0.6008239375542064, "grad_norm": 0.43787914514541626, "learning_rate": 8.026258523139417e-05, "loss": 0.0345, "step": 5542 }, { "epoch": 0.6009323503902863, "grad_norm": 0.6957953572273254, "learning_rate": 8.025895836355723e-05, "loss": 0.0497, "step": 5543 }, { "epoch": 0.601040763226366, "grad_norm": 2.193366765975952, "learning_rate": 8.02553314957203e-05, "loss": 0.0846, "step": 5544 }, { "epoch": 0.6011491760624458, "grad_norm": 0.3160270154476166, "learning_rate": 8.025170462788336e-05, "loss": 0.0436, "step": 5545 }, { "epoch": 0.6012575888985255, "grad_norm": 0.35135316848754883, "learning_rate": 8.024807776004643e-05, "loss": 0.058, "step": 5546 }, { "epoch": 0.6013660017346054, "grad_norm": 0.21168570220470428, "learning_rate": 8.024445089220949e-05, "loss": 0.0274, "step": 5547 }, { "epoch": 0.6014744145706852, "grad_norm": 0.27138832211494446, "learning_rate": 8.024082402437256e-05, "loss": 0.0495, "step": 5548 }, { "epoch": 0.6015828274067649, "grad_norm": 0.10061583667993546, "learning_rate": 8.023719715653562e-05, "loss": 0.0127, "step": 5549 }, { "epoch": 0.6016912402428447, "grad_norm": 0.22372561693191528, "learning_rate": 8.023357028869867e-05, "loss": 0.0281, "step": 5550 }, { "epoch": 0.6017996530789246, "grad_norm": 0.47049716114997864, "learning_rate": 8.022994342086176e-05, "loss": 0.0756, "step": 5551 }, { "epoch": 0.6019080659150043, "grad_norm": 0.2572442591190338, "learning_rate": 8.022631655302481e-05, "loss": 0.0241, "step": 5552 }, { "epoch": 0.6020164787510841, "grad_norm": 0.458345890045166, "learning_rate": 8.022268968518789e-05, "loss": 0.0615, "step": 5553 }, { "epoch": 0.6021248915871639, "grad_norm": 0.8846042156219482, "learning_rate": 8.021906281735094e-05, "loss": 0.0566, "step": 5554 }, { "epoch": 0.6022333044232437, "grad_norm": 0.23061342537403107, "learning_rate": 8.0215435949514e-05, "loss": 0.0272, "step": 5555 }, { "epoch": 0.6023417172593235, "grad_norm": 0.6351096630096436, "learning_rate": 8.021180908167707e-05, "loss": 0.0471, "step": 5556 }, { "epoch": 0.6024501300954033, "grad_norm": 0.1510704606771469, "learning_rate": 8.020818221384013e-05, "loss": 0.0172, "step": 5557 }, { "epoch": 0.602558542931483, "grad_norm": 0.3766271770000458, "learning_rate": 8.02045553460032e-05, "loss": 0.0562, "step": 5558 }, { "epoch": 0.6026669557675629, "grad_norm": 0.25754237174987793, "learning_rate": 8.020092847816626e-05, "loss": 0.0531, "step": 5559 }, { "epoch": 0.6027753686036427, "grad_norm": 0.5642340183258057, "learning_rate": 8.019730161032933e-05, "loss": 0.0387, "step": 5560 }, { "epoch": 0.6028837814397224, "grad_norm": 0.3706069886684418, "learning_rate": 8.019367474249238e-05, "loss": 0.0483, "step": 5561 }, { "epoch": 0.6029921942758023, "grad_norm": 0.1556578278541565, "learning_rate": 8.019004787465546e-05, "loss": 0.0238, "step": 5562 }, { "epoch": 0.6031006071118821, "grad_norm": 0.4659157395362854, "learning_rate": 8.018642100681851e-05, "loss": 0.0491, "step": 5563 }, { "epoch": 0.6032090199479618, "grad_norm": 0.3537481725215912, "learning_rate": 8.018279413898157e-05, "loss": 0.0549, "step": 5564 }, { "epoch": 0.6033174327840416, "grad_norm": 0.31078892946243286, "learning_rate": 8.017916727114464e-05, "loss": 0.0322, "step": 5565 }, { "epoch": 0.6034258456201215, "grad_norm": 0.17806358635425568, "learning_rate": 8.01755404033077e-05, "loss": 0.0169, "step": 5566 }, { "epoch": 0.6035342584562012, "grad_norm": 0.32577574253082275, "learning_rate": 8.017191353547077e-05, "loss": 0.039, "step": 5567 }, { "epoch": 0.603642671292281, "grad_norm": 0.5753150582313538, "learning_rate": 8.016828666763384e-05, "loss": 0.0586, "step": 5568 }, { "epoch": 0.6037510841283608, "grad_norm": 0.2656347453594208, "learning_rate": 8.01646597997969e-05, "loss": 0.0386, "step": 5569 }, { "epoch": 0.6038594969644406, "grad_norm": 0.3230753540992737, "learning_rate": 8.016103293195997e-05, "loss": 0.0694, "step": 5570 }, { "epoch": 0.6039679098005204, "grad_norm": 0.288646936416626, "learning_rate": 8.015740606412303e-05, "loss": 0.0207, "step": 5571 }, { "epoch": 0.6040763226366002, "grad_norm": 0.2650168538093567, "learning_rate": 8.01537791962861e-05, "loss": 0.038, "step": 5572 }, { "epoch": 0.6041847354726799, "grad_norm": 0.25968384742736816, "learning_rate": 8.015015232844915e-05, "loss": 0.0433, "step": 5573 }, { "epoch": 0.6042931483087598, "grad_norm": 0.11839822679758072, "learning_rate": 8.014652546061222e-05, "loss": 0.0287, "step": 5574 }, { "epoch": 0.6044015611448396, "grad_norm": 0.2150588482618332, "learning_rate": 8.014289859277528e-05, "loss": 0.0423, "step": 5575 }, { "epoch": 0.6045099739809193, "grad_norm": 0.09946437180042267, "learning_rate": 8.013927172493835e-05, "loss": 0.0161, "step": 5576 }, { "epoch": 0.6046183868169991, "grad_norm": 0.07721502333879471, "learning_rate": 8.013564485710141e-05, "loss": 0.0161, "step": 5577 }, { "epoch": 0.604726799653079, "grad_norm": 0.4601961672306061, "learning_rate": 8.013201798926447e-05, "loss": 0.0499, "step": 5578 }, { "epoch": 0.6048352124891587, "grad_norm": 0.288381963968277, "learning_rate": 8.012839112142754e-05, "loss": 0.0284, "step": 5579 }, { "epoch": 0.6049436253252385, "grad_norm": 0.39888155460357666, "learning_rate": 8.01247642535906e-05, "loss": 0.0627, "step": 5580 }, { "epoch": 0.6050520381613183, "grad_norm": 0.08550253510475159, "learning_rate": 8.012113738575367e-05, "loss": 0.0131, "step": 5581 }, { "epoch": 0.6051604509973981, "grad_norm": 0.3315284550189972, "learning_rate": 8.011751051791672e-05, "loss": 0.0394, "step": 5582 }, { "epoch": 0.6052688638334779, "grad_norm": 0.3059811294078827, "learning_rate": 8.01138836500798e-05, "loss": 0.0429, "step": 5583 }, { "epoch": 0.6053772766695577, "grad_norm": 0.22930622100830078, "learning_rate": 8.011025678224285e-05, "loss": 0.0224, "step": 5584 }, { "epoch": 0.6054856895056374, "grad_norm": 0.17824716866016388, "learning_rate": 8.010662991440592e-05, "loss": 0.0118, "step": 5585 }, { "epoch": 0.6055941023417173, "grad_norm": 0.360912948846817, "learning_rate": 8.0103003046569e-05, "loss": 0.0266, "step": 5586 }, { "epoch": 0.605702515177797, "grad_norm": 0.33051979541778564, "learning_rate": 8.009937617873205e-05, "loss": 0.0511, "step": 5587 }, { "epoch": 0.6058109280138768, "grad_norm": 0.36076316237449646, "learning_rate": 8.009574931089512e-05, "loss": 0.0521, "step": 5588 }, { "epoch": 0.6059193408499567, "grad_norm": 0.16211234033107758, "learning_rate": 8.009212244305818e-05, "loss": 0.0217, "step": 5589 }, { "epoch": 0.6060277536860365, "grad_norm": 0.325829416513443, "learning_rate": 8.008849557522125e-05, "loss": 0.0365, "step": 5590 }, { "epoch": 0.6061361665221162, "grad_norm": 0.3724709451198578, "learning_rate": 8.008486870738431e-05, "loss": 0.0571, "step": 5591 }, { "epoch": 0.606244579358196, "grad_norm": 0.37215664982795715, "learning_rate": 8.008124183954737e-05, "loss": 0.0122, "step": 5592 }, { "epoch": 0.6063529921942759, "grad_norm": 0.4976608455181122, "learning_rate": 8.007761497171044e-05, "loss": 0.0395, "step": 5593 }, { "epoch": 0.6064614050303556, "grad_norm": 0.7193784117698669, "learning_rate": 8.00739881038735e-05, "loss": 0.088, "step": 5594 }, { "epoch": 0.6065698178664354, "grad_norm": 0.23745091259479523, "learning_rate": 8.007036123603656e-05, "loss": 0.0276, "step": 5595 }, { "epoch": 0.6066782307025151, "grad_norm": 0.3586480915546417, "learning_rate": 8.006673436819962e-05, "loss": 0.0417, "step": 5596 }, { "epoch": 0.606786643538595, "grad_norm": 0.6037939190864563, "learning_rate": 8.006310750036269e-05, "loss": 0.1178, "step": 5597 }, { "epoch": 0.6068950563746748, "grad_norm": 0.25686395168304443, "learning_rate": 8.005948063252575e-05, "loss": 0.0303, "step": 5598 }, { "epoch": 0.6070034692107545, "grad_norm": 0.3329368233680725, "learning_rate": 8.005585376468882e-05, "loss": 0.0554, "step": 5599 }, { "epoch": 0.6071118820468343, "grad_norm": 0.45794928073883057, "learning_rate": 8.005222689685188e-05, "loss": 0.0561, "step": 5600 }, { "epoch": 0.6072202948829142, "grad_norm": 0.47218841314315796, "learning_rate": 8.004860002901494e-05, "loss": 0.0274, "step": 5601 }, { "epoch": 0.6073287077189939, "grad_norm": 0.3343459367752075, "learning_rate": 8.0044973161178e-05, "loss": 0.0313, "step": 5602 }, { "epoch": 0.6074371205550737, "grad_norm": 0.23943813145160675, "learning_rate": 8.004134629334108e-05, "loss": 0.0427, "step": 5603 }, { "epoch": 0.6075455333911535, "grad_norm": 0.14728672802448273, "learning_rate": 8.003771942550415e-05, "loss": 0.0234, "step": 5604 }, { "epoch": 0.6076539462272333, "grad_norm": 0.15657038986682892, "learning_rate": 8.00340925576672e-05, "loss": 0.0329, "step": 5605 }, { "epoch": 0.6077623590633131, "grad_norm": 0.392792284488678, "learning_rate": 8.003046568983028e-05, "loss": 0.0835, "step": 5606 }, { "epoch": 0.6078707718993929, "grad_norm": 0.18775181472301483, "learning_rate": 8.002683882199333e-05, "loss": 0.0288, "step": 5607 }, { "epoch": 0.6079791847354726, "grad_norm": 0.18473775684833527, "learning_rate": 8.002321195415639e-05, "loss": 0.0257, "step": 5608 }, { "epoch": 0.6080875975715525, "grad_norm": 0.48266616463661194, "learning_rate": 8.001958508631946e-05, "loss": 0.0482, "step": 5609 }, { "epoch": 0.6081960104076323, "grad_norm": 0.1115310788154602, "learning_rate": 8.001595821848252e-05, "loss": 0.0188, "step": 5610 }, { "epoch": 0.608304423243712, "grad_norm": 0.5310218334197998, "learning_rate": 8.001233135064559e-05, "loss": 0.0814, "step": 5611 }, { "epoch": 0.6084128360797918, "grad_norm": 0.36669737100601196, "learning_rate": 8.000870448280865e-05, "loss": 0.0548, "step": 5612 }, { "epoch": 0.6085212489158717, "grad_norm": 0.18913988769054413, "learning_rate": 8.000507761497172e-05, "loss": 0.023, "step": 5613 }, { "epoch": 0.6086296617519514, "grad_norm": 0.2939317524433136, "learning_rate": 8.000145074713478e-05, "loss": 0.0478, "step": 5614 }, { "epoch": 0.6087380745880312, "grad_norm": 0.4832552671432495, "learning_rate": 7.999782387929785e-05, "loss": 0.0798, "step": 5615 }, { "epoch": 0.608846487424111, "grad_norm": 0.6519548296928406, "learning_rate": 7.99941970114609e-05, "loss": 0.0632, "step": 5616 }, { "epoch": 0.6089549002601908, "grad_norm": 0.1679883450269699, "learning_rate": 7.999057014362396e-05, "loss": 0.0351, "step": 5617 }, { "epoch": 0.6090633130962706, "grad_norm": 0.3872116208076477, "learning_rate": 7.998694327578703e-05, "loss": 0.0389, "step": 5618 }, { "epoch": 0.6091717259323504, "grad_norm": 0.38884061574935913, "learning_rate": 7.998331640795009e-05, "loss": 0.0424, "step": 5619 }, { "epoch": 0.6092801387684302, "grad_norm": 0.31710752844810486, "learning_rate": 7.997968954011317e-05, "loss": 0.0268, "step": 5620 }, { "epoch": 0.60938855160451, "grad_norm": 0.18718530237674713, "learning_rate": 7.997606267227623e-05, "loss": 0.0179, "step": 5621 }, { "epoch": 0.6094969644405898, "grad_norm": 0.2005053162574768, "learning_rate": 7.997243580443929e-05, "loss": 0.0304, "step": 5622 }, { "epoch": 0.6096053772766695, "grad_norm": 0.5765646696090698, "learning_rate": 7.996880893660236e-05, "loss": 0.0437, "step": 5623 }, { "epoch": 0.6097137901127494, "grad_norm": 0.22768078744411469, "learning_rate": 7.996518206876542e-05, "loss": 0.0526, "step": 5624 }, { "epoch": 0.6098222029488292, "grad_norm": 0.592021107673645, "learning_rate": 7.996155520092849e-05, "loss": 0.0568, "step": 5625 }, { "epoch": 0.6099306157849089, "grad_norm": 0.14369507133960724, "learning_rate": 7.995792833309154e-05, "loss": 0.0211, "step": 5626 }, { "epoch": 0.6100390286209887, "grad_norm": 0.15325725078582764, "learning_rate": 7.995430146525462e-05, "loss": 0.02, "step": 5627 }, { "epoch": 0.6101474414570686, "grad_norm": 0.2187616527080536, "learning_rate": 7.995067459741767e-05, "loss": 0.0327, "step": 5628 }, { "epoch": 0.6102558542931483, "grad_norm": 0.456645667552948, "learning_rate": 7.994704772958074e-05, "loss": 0.0484, "step": 5629 }, { "epoch": 0.6103642671292281, "grad_norm": 0.4862610399723053, "learning_rate": 7.99434208617438e-05, "loss": 0.0416, "step": 5630 }, { "epoch": 0.6104726799653079, "grad_norm": 0.1823609322309494, "learning_rate": 7.993979399390686e-05, "loss": 0.0151, "step": 5631 }, { "epoch": 0.6105810928013877, "grad_norm": 0.6405994296073914, "learning_rate": 7.993616712606993e-05, "loss": 0.0989, "step": 5632 }, { "epoch": 0.6106895056374675, "grad_norm": 0.38529011607170105, "learning_rate": 7.993254025823299e-05, "loss": 0.0571, "step": 5633 }, { "epoch": 0.6107979184735473, "grad_norm": 0.3367559015750885, "learning_rate": 7.992891339039606e-05, "loss": 0.0803, "step": 5634 }, { "epoch": 0.610906331309627, "grad_norm": 0.1490473598241806, "learning_rate": 7.992528652255911e-05, "loss": 0.0194, "step": 5635 }, { "epoch": 0.6110147441457069, "grad_norm": 0.21099136769771576, "learning_rate": 7.992165965472219e-05, "loss": 0.0437, "step": 5636 }, { "epoch": 0.6111231569817867, "grad_norm": 0.4973996579647064, "learning_rate": 7.991803278688526e-05, "loss": 0.0655, "step": 5637 }, { "epoch": 0.6112315698178664, "grad_norm": 0.15539968013763428, "learning_rate": 7.991440591904831e-05, "loss": 0.009, "step": 5638 }, { "epoch": 0.6113399826539462, "grad_norm": 0.35635682940483093, "learning_rate": 7.991077905121138e-05, "loss": 0.0308, "step": 5639 }, { "epoch": 0.611448395490026, "grad_norm": 0.462783545255661, "learning_rate": 7.990715218337444e-05, "loss": 0.0479, "step": 5640 }, { "epoch": 0.6115568083261058, "grad_norm": 0.2223564237356186, "learning_rate": 7.990352531553751e-05, "loss": 0.0339, "step": 5641 }, { "epoch": 0.6116652211621856, "grad_norm": 0.44467321038246155, "learning_rate": 7.989989844770057e-05, "loss": 0.0406, "step": 5642 }, { "epoch": 0.6117736339982653, "grad_norm": 0.44599542021751404, "learning_rate": 7.989627157986364e-05, "loss": 0.0684, "step": 5643 }, { "epoch": 0.6118820468343452, "grad_norm": 0.4954480826854706, "learning_rate": 7.98926447120267e-05, "loss": 0.0841, "step": 5644 }, { "epoch": 0.611990459670425, "grad_norm": 0.4059624969959259, "learning_rate": 7.988901784418976e-05, "loss": 0.0469, "step": 5645 }, { "epoch": 0.6120988725065047, "grad_norm": 0.2431773692369461, "learning_rate": 7.988539097635283e-05, "loss": 0.0426, "step": 5646 }, { "epoch": 0.6122072853425846, "grad_norm": 0.24652566015720367, "learning_rate": 7.988176410851588e-05, "loss": 0.034, "step": 5647 }, { "epoch": 0.6123156981786644, "grad_norm": 0.3212832808494568, "learning_rate": 7.987813724067896e-05, "loss": 0.0403, "step": 5648 }, { "epoch": 0.6124241110147441, "grad_norm": 0.09767455607652664, "learning_rate": 7.987451037284201e-05, "loss": 0.0139, "step": 5649 }, { "epoch": 0.6125325238508239, "grad_norm": 0.06664802134037018, "learning_rate": 7.987088350500508e-05, "loss": 0.0092, "step": 5650 }, { "epoch": 0.6126409366869038, "grad_norm": 0.49602335691452026, "learning_rate": 7.986725663716814e-05, "loss": 0.0377, "step": 5651 }, { "epoch": 0.6127493495229835, "grad_norm": 0.08853624761104584, "learning_rate": 7.986362976933121e-05, "loss": 0.0092, "step": 5652 }, { "epoch": 0.6128577623590633, "grad_norm": 0.2268165498971939, "learning_rate": 7.986000290149427e-05, "loss": 0.0255, "step": 5653 }, { "epoch": 0.6129661751951431, "grad_norm": 0.16262255609035492, "learning_rate": 7.985637603365733e-05, "loss": 0.0117, "step": 5654 }, { "epoch": 0.613074588031223, "grad_norm": 0.5244727730751038, "learning_rate": 7.985274916582041e-05, "loss": 0.0469, "step": 5655 }, { "epoch": 0.6131830008673027, "grad_norm": 0.354560911655426, "learning_rate": 7.984912229798347e-05, "loss": 0.05, "step": 5656 }, { "epoch": 0.6132914137033825, "grad_norm": 0.4928256571292877, "learning_rate": 7.984549543014654e-05, "loss": 0.1134, "step": 5657 }, { "epoch": 0.6133998265394622, "grad_norm": 0.533865213394165, "learning_rate": 7.98418685623096e-05, "loss": 0.0487, "step": 5658 }, { "epoch": 0.6135082393755421, "grad_norm": 0.8470390439033508, "learning_rate": 7.983824169447265e-05, "loss": 0.0578, "step": 5659 }, { "epoch": 0.6136166522116219, "grad_norm": 0.26093754172325134, "learning_rate": 7.983461482663572e-05, "loss": 0.0357, "step": 5660 }, { "epoch": 0.6137250650477016, "grad_norm": 0.17217926681041718, "learning_rate": 7.983098795879878e-05, "loss": 0.0244, "step": 5661 }, { "epoch": 0.6138334778837814, "grad_norm": 0.3634282052516937, "learning_rate": 7.982736109096185e-05, "loss": 0.0513, "step": 5662 }, { "epoch": 0.6139418907198613, "grad_norm": 0.24115142226219177, "learning_rate": 7.982373422312491e-05, "loss": 0.0265, "step": 5663 }, { "epoch": 0.614050303555941, "grad_norm": 0.49835994839668274, "learning_rate": 7.982010735528798e-05, "loss": 0.0712, "step": 5664 }, { "epoch": 0.6141587163920208, "grad_norm": 0.29479730129241943, "learning_rate": 7.981648048745104e-05, "loss": 0.0585, "step": 5665 }, { "epoch": 0.6142671292281006, "grad_norm": 0.2723807096481323, "learning_rate": 7.981285361961411e-05, "loss": 0.0371, "step": 5666 }, { "epoch": 0.6143755420641804, "grad_norm": 0.31256601214408875, "learning_rate": 7.980922675177717e-05, "loss": 0.0415, "step": 5667 }, { "epoch": 0.6144839549002602, "grad_norm": 0.29656898975372314, "learning_rate": 7.980559988394022e-05, "loss": 0.0316, "step": 5668 }, { "epoch": 0.61459236773634, "grad_norm": 0.40542951226234436, "learning_rate": 7.98019730161033e-05, "loss": 0.0374, "step": 5669 }, { "epoch": 0.6147007805724197, "grad_norm": 0.18264742195606232, "learning_rate": 7.979834614826635e-05, "loss": 0.0311, "step": 5670 }, { "epoch": 0.6148091934084996, "grad_norm": 0.35728469491004944, "learning_rate": 7.979471928042942e-05, "loss": 0.0326, "step": 5671 }, { "epoch": 0.6149176062445794, "grad_norm": 0.15457496047019958, "learning_rate": 7.97910924125925e-05, "loss": 0.0185, "step": 5672 }, { "epoch": 0.6150260190806591, "grad_norm": 0.043598197400569916, "learning_rate": 7.978746554475555e-05, "loss": 0.009, "step": 5673 }, { "epoch": 0.6151344319167389, "grad_norm": 0.3526691794395447, "learning_rate": 7.978383867691862e-05, "loss": 0.0494, "step": 5674 }, { "epoch": 0.6152428447528188, "grad_norm": 0.2717716693878174, "learning_rate": 7.978021180908168e-05, "loss": 0.0143, "step": 5675 }, { "epoch": 0.6153512575888985, "grad_norm": 0.162648007273674, "learning_rate": 7.977658494124475e-05, "loss": 0.0153, "step": 5676 }, { "epoch": 0.6154596704249783, "grad_norm": 0.4553583562374115, "learning_rate": 7.977295807340781e-05, "loss": 0.0426, "step": 5677 }, { "epoch": 0.6155680832610582, "grad_norm": 0.15353688597679138, "learning_rate": 7.976933120557088e-05, "loss": 0.0197, "step": 5678 }, { "epoch": 0.6156764960971379, "grad_norm": 0.35128480195999146, "learning_rate": 7.976570433773394e-05, "loss": 0.0212, "step": 5679 }, { "epoch": 0.6157849089332177, "grad_norm": 0.4453389048576355, "learning_rate": 7.9762077469897e-05, "loss": 0.0282, "step": 5680 }, { "epoch": 0.6158933217692975, "grad_norm": 0.14268295466899872, "learning_rate": 7.975845060206006e-05, "loss": 0.0154, "step": 5681 }, { "epoch": 0.6160017346053773, "grad_norm": 0.17865711450576782, "learning_rate": 7.975482373422312e-05, "loss": 0.0288, "step": 5682 }, { "epoch": 0.6161101474414571, "grad_norm": 0.25441282987594604, "learning_rate": 7.975119686638619e-05, "loss": 0.0278, "step": 5683 }, { "epoch": 0.6162185602775369, "grad_norm": 0.8841186761856079, "learning_rate": 7.974756999854925e-05, "loss": 0.0534, "step": 5684 }, { "epoch": 0.6163269731136166, "grad_norm": 0.40230247378349304, "learning_rate": 7.974394313071232e-05, "loss": 0.0213, "step": 5685 }, { "epoch": 0.6164353859496965, "grad_norm": 0.20328451693058014, "learning_rate": 7.974031626287538e-05, "loss": 0.0245, "step": 5686 }, { "epoch": 0.6165437987857763, "grad_norm": 0.41259700059890747, "learning_rate": 7.973668939503845e-05, "loss": 0.0283, "step": 5687 }, { "epoch": 0.616652211621856, "grad_norm": 0.1565302610397339, "learning_rate": 7.97330625272015e-05, "loss": 0.0119, "step": 5688 }, { "epoch": 0.6167606244579358, "grad_norm": 0.23647712171077728, "learning_rate": 7.972943565936458e-05, "loss": 0.0105, "step": 5689 }, { "epoch": 0.6168690372940157, "grad_norm": 0.5492196083068848, "learning_rate": 7.972580879152765e-05, "loss": 0.0333, "step": 5690 }, { "epoch": 0.6169774501300954, "grad_norm": 0.2925703525543213, "learning_rate": 7.97221819236907e-05, "loss": 0.0156, "step": 5691 }, { "epoch": 0.6170858629661752, "grad_norm": 1.2704683542251587, "learning_rate": 7.971855505585378e-05, "loss": 0.0587, "step": 5692 }, { "epoch": 0.6171942758022549, "grad_norm": 0.27352476119995117, "learning_rate": 7.971492818801683e-05, "loss": 0.0223, "step": 5693 }, { "epoch": 0.6173026886383348, "grad_norm": 0.10750465840101242, "learning_rate": 7.97113013201799e-05, "loss": 0.0122, "step": 5694 }, { "epoch": 0.6174111014744146, "grad_norm": 0.7349252700805664, "learning_rate": 7.970767445234296e-05, "loss": 0.0798, "step": 5695 }, { "epoch": 0.6175195143104943, "grad_norm": 0.3405705392360687, "learning_rate": 7.970404758450602e-05, "loss": 0.0195, "step": 5696 }, { "epoch": 0.6176279271465741, "grad_norm": 0.2368067353963852, "learning_rate": 7.970042071666909e-05, "loss": 0.0215, "step": 5697 }, { "epoch": 0.617736339982654, "grad_norm": 0.7993823885917664, "learning_rate": 7.969679384883215e-05, "loss": 0.0523, "step": 5698 }, { "epoch": 0.6178447528187337, "grad_norm": 0.3586909770965576, "learning_rate": 7.969316698099522e-05, "loss": 0.0218, "step": 5699 }, { "epoch": 0.6179531656548135, "grad_norm": 0.8940408825874329, "learning_rate": 7.968954011315827e-05, "loss": 0.0387, "step": 5700 }, { "epoch": 0.6180615784908933, "grad_norm": 0.41377922892570496, "learning_rate": 7.968591324532135e-05, "loss": 0.0493, "step": 5701 }, { "epoch": 0.6181699913269731, "grad_norm": 0.08318749815225601, "learning_rate": 7.96822863774844e-05, "loss": 0.0212, "step": 5702 }, { "epoch": 0.6182784041630529, "grad_norm": 0.5275754332542419, "learning_rate": 7.967865950964747e-05, "loss": 0.04, "step": 5703 }, { "epoch": 0.6183868169991327, "grad_norm": 0.20853054523468018, "learning_rate": 7.967503264181053e-05, "loss": 0.0144, "step": 5704 }, { "epoch": 0.6184952298352124, "grad_norm": 0.7871747612953186, "learning_rate": 7.967140577397359e-05, "loss": 0.0643, "step": 5705 }, { "epoch": 0.6186036426712923, "grad_norm": 0.7832171320915222, "learning_rate": 7.966777890613667e-05, "loss": 0.0451, "step": 5706 }, { "epoch": 0.6187120555073721, "grad_norm": 0.41886353492736816, "learning_rate": 7.966415203829973e-05, "loss": 0.0244, "step": 5707 }, { "epoch": 0.6188204683434518, "grad_norm": 0.4420672059059143, "learning_rate": 7.96605251704628e-05, "loss": 0.064, "step": 5708 }, { "epoch": 0.6189288811795317, "grad_norm": 0.6201158165931702, "learning_rate": 7.965689830262586e-05, "loss": 0.0552, "step": 5709 }, { "epoch": 0.6190372940156115, "grad_norm": 0.07447747886180878, "learning_rate": 7.965327143478893e-05, "loss": 0.0063, "step": 5710 }, { "epoch": 0.6191457068516912, "grad_norm": 0.43550702929496765, "learning_rate": 7.964964456695199e-05, "loss": 0.0441, "step": 5711 }, { "epoch": 0.619254119687771, "grad_norm": 0.36934953927993774, "learning_rate": 7.964601769911504e-05, "loss": 0.0567, "step": 5712 }, { "epoch": 0.6193625325238509, "grad_norm": 0.15589745342731476, "learning_rate": 7.964239083127812e-05, "loss": 0.0131, "step": 5713 }, { "epoch": 0.6194709453599306, "grad_norm": 0.28812074661254883, "learning_rate": 7.963876396344117e-05, "loss": 0.0294, "step": 5714 }, { "epoch": 0.6195793581960104, "grad_norm": 0.3873727023601532, "learning_rate": 7.963513709560424e-05, "loss": 0.0405, "step": 5715 }, { "epoch": 0.6196877710320902, "grad_norm": 0.2304670214653015, "learning_rate": 7.96315102277673e-05, "loss": 0.0193, "step": 5716 }, { "epoch": 0.61979618386817, "grad_norm": 0.6382226347923279, "learning_rate": 7.962788335993037e-05, "loss": 0.0839, "step": 5717 }, { "epoch": 0.6199045967042498, "grad_norm": 0.23639070987701416, "learning_rate": 7.962425649209343e-05, "loss": 0.0131, "step": 5718 }, { "epoch": 0.6200130095403296, "grad_norm": 0.3424123525619507, "learning_rate": 7.96206296242565e-05, "loss": 0.043, "step": 5719 }, { "epoch": 0.6201214223764093, "grad_norm": 0.2760935425758362, "learning_rate": 7.961700275641956e-05, "loss": 0.047, "step": 5720 }, { "epoch": 0.6202298352124892, "grad_norm": 0.8590841889381409, "learning_rate": 7.961337588858261e-05, "loss": 0.0534, "step": 5721 }, { "epoch": 0.620338248048569, "grad_norm": 0.15668384730815887, "learning_rate": 7.960974902074569e-05, "loss": 0.017, "step": 5722 }, { "epoch": 0.6204466608846487, "grad_norm": 0.48465561866760254, "learning_rate": 7.960612215290874e-05, "loss": 0.0638, "step": 5723 }, { "epoch": 0.6205550737207285, "grad_norm": 0.4256996512413025, "learning_rate": 7.960249528507183e-05, "loss": 0.0318, "step": 5724 }, { "epoch": 0.6206634865568084, "grad_norm": 0.1978650689125061, "learning_rate": 7.959886841723488e-05, "loss": 0.026, "step": 5725 }, { "epoch": 0.6207718993928881, "grad_norm": 0.24552172422409058, "learning_rate": 7.959524154939794e-05, "loss": 0.0228, "step": 5726 }, { "epoch": 0.6208803122289679, "grad_norm": 0.24283210933208466, "learning_rate": 7.959161468156101e-05, "loss": 0.0408, "step": 5727 }, { "epoch": 0.6209887250650477, "grad_norm": 0.3273749053478241, "learning_rate": 7.958798781372407e-05, "loss": 0.0319, "step": 5728 }, { "epoch": 0.6210971379011275, "grad_norm": 0.7094749808311462, "learning_rate": 7.958436094588714e-05, "loss": 0.1011, "step": 5729 }, { "epoch": 0.6212055507372073, "grad_norm": 0.20641937851905823, "learning_rate": 7.95807340780502e-05, "loss": 0.0193, "step": 5730 }, { "epoch": 0.621313963573287, "grad_norm": 0.0931582972407341, "learning_rate": 7.957710721021327e-05, "loss": 0.0092, "step": 5731 }, { "epoch": 0.6214223764093668, "grad_norm": 0.2039448469877243, "learning_rate": 7.957348034237633e-05, "loss": 0.0179, "step": 5732 }, { "epoch": 0.6215307892454467, "grad_norm": 0.4457314908504486, "learning_rate": 7.95698534745394e-05, "loss": 0.0474, "step": 5733 }, { "epoch": 0.6216392020815265, "grad_norm": 0.307184636592865, "learning_rate": 7.956622660670245e-05, "loss": 0.0374, "step": 5734 }, { "epoch": 0.6217476149176062, "grad_norm": 0.2799522578716278, "learning_rate": 7.956259973886551e-05, "loss": 0.0545, "step": 5735 }, { "epoch": 0.6218560277536861, "grad_norm": 0.41137784719467163, "learning_rate": 7.955897287102858e-05, "loss": 0.0333, "step": 5736 }, { "epoch": 0.6219644405897659, "grad_norm": 0.3512304425239563, "learning_rate": 7.955534600319164e-05, "loss": 0.0257, "step": 5737 }, { "epoch": 0.6220728534258456, "grad_norm": 0.31875964999198914, "learning_rate": 7.955171913535471e-05, "loss": 0.0475, "step": 5738 }, { "epoch": 0.6221812662619254, "grad_norm": 0.36578166484832764, "learning_rate": 7.954809226751777e-05, "loss": 0.0652, "step": 5739 }, { "epoch": 0.6222896790980053, "grad_norm": 0.7049809694290161, "learning_rate": 7.954446539968084e-05, "loss": 0.0513, "step": 5740 }, { "epoch": 0.622398091934085, "grad_norm": 0.5668942928314209, "learning_rate": 7.954083853184391e-05, "loss": 0.0383, "step": 5741 }, { "epoch": 0.6225065047701648, "grad_norm": 0.2207527756690979, "learning_rate": 7.953721166400697e-05, "loss": 0.0395, "step": 5742 }, { "epoch": 0.6226149176062445, "grad_norm": 0.3016826808452606, "learning_rate": 7.953358479617004e-05, "loss": 0.069, "step": 5743 }, { "epoch": 0.6227233304423244, "grad_norm": 0.09036839753389359, "learning_rate": 7.95299579283331e-05, "loss": 0.0086, "step": 5744 }, { "epoch": 0.6228317432784042, "grad_norm": 0.42142167687416077, "learning_rate": 7.952633106049617e-05, "loss": 0.0691, "step": 5745 }, { "epoch": 0.622940156114484, "grad_norm": 0.22428561747074127, "learning_rate": 7.952270419265922e-05, "loss": 0.0398, "step": 5746 }, { "epoch": 0.6230485689505637, "grad_norm": 0.45624133944511414, "learning_rate": 7.95190773248223e-05, "loss": 0.0453, "step": 5747 }, { "epoch": 0.6231569817866436, "grad_norm": 0.4804762601852417, "learning_rate": 7.951545045698535e-05, "loss": 0.0769, "step": 5748 }, { "epoch": 0.6232653946227233, "grad_norm": 0.41979289054870605, "learning_rate": 7.951182358914841e-05, "loss": 0.0425, "step": 5749 }, { "epoch": 0.6233738074588031, "grad_norm": 0.32180067896842957, "learning_rate": 7.950819672131148e-05, "loss": 0.0486, "step": 5750 }, { "epoch": 0.6234822202948829, "grad_norm": 0.6533520817756653, "learning_rate": 7.950456985347454e-05, "loss": 0.054, "step": 5751 }, { "epoch": 0.6235906331309627, "grad_norm": 0.3518964350223541, "learning_rate": 7.950094298563761e-05, "loss": 0.041, "step": 5752 }, { "epoch": 0.6236990459670425, "grad_norm": 0.1988135278224945, "learning_rate": 7.949731611780067e-05, "loss": 0.0426, "step": 5753 }, { "epoch": 0.6238074588031223, "grad_norm": 0.3116030693054199, "learning_rate": 7.949368924996374e-05, "loss": 0.019, "step": 5754 }, { "epoch": 0.623915871639202, "grad_norm": 0.43115800619125366, "learning_rate": 7.94900623821268e-05, "loss": 0.0447, "step": 5755 }, { "epoch": 0.6240242844752819, "grad_norm": 0.26110121607780457, "learning_rate": 7.948643551428986e-05, "loss": 0.0335, "step": 5756 }, { "epoch": 0.6241326973113617, "grad_norm": 0.4349040985107422, "learning_rate": 7.948280864645292e-05, "loss": 0.0703, "step": 5757 }, { "epoch": 0.6242411101474414, "grad_norm": 0.3218139111995697, "learning_rate": 7.947918177861599e-05, "loss": 0.0261, "step": 5758 }, { "epoch": 0.6243495229835212, "grad_norm": 0.5013877749443054, "learning_rate": 7.947555491077906e-05, "loss": 0.082, "step": 5759 }, { "epoch": 0.6244579358196011, "grad_norm": 0.3904710114002228, "learning_rate": 7.947192804294212e-05, "loss": 0.0717, "step": 5760 }, { "epoch": 0.6245663486556808, "grad_norm": 0.31565961241722107, "learning_rate": 7.946830117510519e-05, "loss": 0.0328, "step": 5761 }, { "epoch": 0.6246747614917606, "grad_norm": 0.18921011686325073, "learning_rate": 7.946467430726825e-05, "loss": 0.0231, "step": 5762 }, { "epoch": 0.6247831743278404, "grad_norm": 0.5018848776817322, "learning_rate": 7.946104743943131e-05, "loss": 0.0551, "step": 5763 }, { "epoch": 0.6248915871639202, "grad_norm": 0.17821139097213745, "learning_rate": 7.945742057159438e-05, "loss": 0.0266, "step": 5764 }, { "epoch": 0.625, "grad_norm": 0.8511120676994324, "learning_rate": 7.945379370375744e-05, "loss": 0.0503, "step": 5765 }, { "epoch": 0.6251084128360798, "grad_norm": 0.2347528338432312, "learning_rate": 7.94501668359205e-05, "loss": 0.025, "step": 5766 }, { "epoch": 0.6252168256721596, "grad_norm": 0.396995484828949, "learning_rate": 7.944653996808356e-05, "loss": 0.0291, "step": 5767 }, { "epoch": 0.6253252385082394, "grad_norm": 0.3299659788608551, "learning_rate": 7.944291310024663e-05, "loss": 0.0338, "step": 5768 }, { "epoch": 0.6254336513443192, "grad_norm": 0.3016301989555359, "learning_rate": 7.943928623240969e-05, "loss": 0.0574, "step": 5769 }, { "epoch": 0.6255420641803989, "grad_norm": 0.23031209409236908, "learning_rate": 7.943565936457276e-05, "loss": 0.0304, "step": 5770 }, { "epoch": 0.6256504770164788, "grad_norm": 0.19978760182857513, "learning_rate": 7.943203249673582e-05, "loss": 0.0331, "step": 5771 }, { "epoch": 0.6257588898525586, "grad_norm": 0.1890263557434082, "learning_rate": 7.942840562889888e-05, "loss": 0.0218, "step": 5772 }, { "epoch": 0.6258673026886383, "grad_norm": 0.14566686749458313, "learning_rate": 7.942477876106195e-05, "loss": 0.0191, "step": 5773 }, { "epoch": 0.6259757155247181, "grad_norm": 0.32012808322906494, "learning_rate": 7.9421151893225e-05, "loss": 0.022, "step": 5774 }, { "epoch": 0.626084128360798, "grad_norm": 0.13398221135139465, "learning_rate": 7.941752502538808e-05, "loss": 0.0073, "step": 5775 }, { "epoch": 0.6261925411968777, "grad_norm": 0.2866800129413605, "learning_rate": 7.941389815755115e-05, "loss": 0.0513, "step": 5776 }, { "epoch": 0.6263009540329575, "grad_norm": 0.2604081332683563, "learning_rate": 7.94102712897142e-05, "loss": 0.0206, "step": 5777 }, { "epoch": 0.6264093668690373, "grad_norm": 0.25405746698379517, "learning_rate": 7.940664442187728e-05, "loss": 0.0327, "step": 5778 }, { "epoch": 0.6265177797051171, "grad_norm": 0.4635196030139923, "learning_rate": 7.940301755404033e-05, "loss": 0.0742, "step": 5779 }, { "epoch": 0.6266261925411969, "grad_norm": 0.24513870477676392, "learning_rate": 7.93993906862034e-05, "loss": 0.0371, "step": 5780 }, { "epoch": 0.6267346053772767, "grad_norm": 0.8537316918373108, "learning_rate": 7.939576381836646e-05, "loss": 0.1135, "step": 5781 }, { "epoch": 0.6268430182133564, "grad_norm": 0.4744454324245453, "learning_rate": 7.939213695052953e-05, "loss": 0.0479, "step": 5782 }, { "epoch": 0.6269514310494363, "grad_norm": 0.4086226224899292, "learning_rate": 7.938851008269259e-05, "loss": 0.0756, "step": 5783 }, { "epoch": 0.627059843885516, "grad_norm": 0.2655890882015228, "learning_rate": 7.938488321485566e-05, "loss": 0.0815, "step": 5784 }, { "epoch": 0.6271682567215958, "grad_norm": 0.345862478017807, "learning_rate": 7.938125634701872e-05, "loss": 0.0412, "step": 5785 }, { "epoch": 0.6272766695576756, "grad_norm": 0.15326163172721863, "learning_rate": 7.937762947918177e-05, "loss": 0.0209, "step": 5786 }, { "epoch": 0.6273850823937555, "grad_norm": 0.11877606809139252, "learning_rate": 7.937400261134485e-05, "loss": 0.0571, "step": 5787 }, { "epoch": 0.6274934952298352, "grad_norm": 0.5667543411254883, "learning_rate": 7.93703757435079e-05, "loss": 0.0633, "step": 5788 }, { "epoch": 0.627601908065915, "grad_norm": 0.3726731240749359, "learning_rate": 7.936674887567097e-05, "loss": 0.0381, "step": 5789 }, { "epoch": 0.6277103209019947, "grad_norm": 0.23580363392829895, "learning_rate": 7.936312200783403e-05, "loss": 0.0328, "step": 5790 }, { "epoch": 0.6278187337380746, "grad_norm": 0.19443736970424652, "learning_rate": 7.93594951399971e-05, "loss": 0.0181, "step": 5791 }, { "epoch": 0.6279271465741544, "grad_norm": 0.43068599700927734, "learning_rate": 7.935586827216016e-05, "loss": 0.0606, "step": 5792 }, { "epoch": 0.6280355594102341, "grad_norm": 0.19740842282772064, "learning_rate": 7.935224140432323e-05, "loss": 0.0251, "step": 5793 }, { "epoch": 0.6281439722463139, "grad_norm": 0.18644392490386963, "learning_rate": 7.93486145364863e-05, "loss": 0.04, "step": 5794 }, { "epoch": 0.6282523850823938, "grad_norm": 0.43153083324432373, "learning_rate": 7.934498766864936e-05, "loss": 0.0905, "step": 5795 }, { "epoch": 0.6283607979184735, "grad_norm": 0.19783560931682587, "learning_rate": 7.934136080081243e-05, "loss": 0.029, "step": 5796 }, { "epoch": 0.6284692107545533, "grad_norm": 0.4007187485694885, "learning_rate": 7.933773393297549e-05, "loss": 0.0454, "step": 5797 }, { "epoch": 0.6285776235906332, "grad_norm": 0.4281143546104431, "learning_rate": 7.933410706513856e-05, "loss": 0.0366, "step": 5798 }, { "epoch": 0.628686036426713, "grad_norm": 0.17721964418888092, "learning_rate": 7.933048019730161e-05, "loss": 0.0235, "step": 5799 }, { "epoch": 0.6287944492627927, "grad_norm": 0.33259618282318115, "learning_rate": 7.932685332946469e-05, "loss": 0.0367, "step": 5800 }, { "epoch": 0.6289028620988725, "grad_norm": 0.18992803990840912, "learning_rate": 7.932322646162774e-05, "loss": 0.032, "step": 5801 }, { "epoch": 0.6290112749349523, "grad_norm": 0.08873609453439713, "learning_rate": 7.93195995937908e-05, "loss": 0.0083, "step": 5802 }, { "epoch": 0.6291196877710321, "grad_norm": 0.2903420627117157, "learning_rate": 7.931597272595387e-05, "loss": 0.0447, "step": 5803 }, { "epoch": 0.6292281006071119, "grad_norm": 0.28518912196159363, "learning_rate": 7.931234585811693e-05, "loss": 0.0216, "step": 5804 }, { "epoch": 0.6293365134431916, "grad_norm": 0.22417376935482025, "learning_rate": 7.930871899028e-05, "loss": 0.0176, "step": 5805 }, { "epoch": 0.6294449262792715, "grad_norm": 0.24550212919712067, "learning_rate": 7.930509212244306e-05, "loss": 0.0298, "step": 5806 }, { "epoch": 0.6295533391153513, "grad_norm": 0.5173618793487549, "learning_rate": 7.930146525460613e-05, "loss": 0.0677, "step": 5807 }, { "epoch": 0.629661751951431, "grad_norm": 0.26624464988708496, "learning_rate": 7.929783838676918e-05, "loss": 0.0263, "step": 5808 }, { "epoch": 0.6297701647875108, "grad_norm": 0.3858301043510437, "learning_rate": 7.929421151893224e-05, "loss": 0.0335, "step": 5809 }, { "epoch": 0.6298785776235907, "grad_norm": 0.27225518226623535, "learning_rate": 7.929058465109533e-05, "loss": 0.0354, "step": 5810 }, { "epoch": 0.6299869904596704, "grad_norm": 0.18839122354984283, "learning_rate": 7.928695778325838e-05, "loss": 0.022, "step": 5811 }, { "epoch": 0.6300954032957502, "grad_norm": 0.2010764330625534, "learning_rate": 7.928333091542145e-05, "loss": 0.0258, "step": 5812 }, { "epoch": 0.63020381613183, "grad_norm": 0.36893993616104126, "learning_rate": 7.927970404758451e-05, "loss": 0.0343, "step": 5813 }, { "epoch": 0.6303122289679098, "grad_norm": 0.2986082136631012, "learning_rate": 7.927607717974758e-05, "loss": 0.0373, "step": 5814 }, { "epoch": 0.6304206418039896, "grad_norm": 0.49625536799430847, "learning_rate": 7.927245031191064e-05, "loss": 0.0578, "step": 5815 }, { "epoch": 0.6305290546400694, "grad_norm": 0.3770374655723572, "learning_rate": 7.92688234440737e-05, "loss": 0.0531, "step": 5816 }, { "epoch": 0.6306374674761491, "grad_norm": 0.10833537578582764, "learning_rate": 7.926519657623677e-05, "loss": 0.0157, "step": 5817 }, { "epoch": 0.630745880312229, "grad_norm": 0.40515273809432983, "learning_rate": 7.926156970839983e-05, "loss": 0.0583, "step": 5818 }, { "epoch": 0.6308542931483088, "grad_norm": 0.25682416558265686, "learning_rate": 7.92579428405629e-05, "loss": 0.0299, "step": 5819 }, { "epoch": 0.6309627059843885, "grad_norm": 1.1128534078598022, "learning_rate": 7.925431597272595e-05, "loss": 0.0769, "step": 5820 }, { "epoch": 0.6310711188204683, "grad_norm": 0.07501790672540665, "learning_rate": 7.925068910488902e-05, "loss": 0.0035, "step": 5821 }, { "epoch": 0.6311795316565482, "grad_norm": 0.4941345751285553, "learning_rate": 7.924706223705208e-05, "loss": 0.0717, "step": 5822 }, { "epoch": 0.6312879444926279, "grad_norm": 0.5868847966194153, "learning_rate": 7.924343536921515e-05, "loss": 0.0369, "step": 5823 }, { "epoch": 0.6313963573287077, "grad_norm": 0.1901610940694809, "learning_rate": 7.923980850137821e-05, "loss": 0.0161, "step": 5824 }, { "epoch": 0.6315047701647876, "grad_norm": 0.3123680353164673, "learning_rate": 7.923618163354127e-05, "loss": 0.0467, "step": 5825 }, { "epoch": 0.6316131830008673, "grad_norm": 0.3037317991256714, "learning_rate": 7.923255476570434e-05, "loss": 0.03, "step": 5826 }, { "epoch": 0.6317215958369471, "grad_norm": 0.36754310131073, "learning_rate": 7.922892789786741e-05, "loss": 0.0811, "step": 5827 }, { "epoch": 0.6318300086730269, "grad_norm": 0.682695746421814, "learning_rate": 7.922530103003048e-05, "loss": 0.0965, "step": 5828 }, { "epoch": 0.6319384215091067, "grad_norm": 0.24574922025203705, "learning_rate": 7.922167416219354e-05, "loss": 0.0186, "step": 5829 }, { "epoch": 0.6320468343451865, "grad_norm": 0.3680063784122467, "learning_rate": 7.92180472943566e-05, "loss": 0.0455, "step": 5830 }, { "epoch": 0.6321552471812663, "grad_norm": 0.34775108098983765, "learning_rate": 7.921442042651967e-05, "loss": 0.0458, "step": 5831 }, { "epoch": 0.632263660017346, "grad_norm": 0.18104463815689087, "learning_rate": 7.921079355868272e-05, "loss": 0.0228, "step": 5832 }, { "epoch": 0.6323720728534259, "grad_norm": 0.27677232027053833, "learning_rate": 7.92071666908458e-05, "loss": 0.0261, "step": 5833 }, { "epoch": 0.6324804856895057, "grad_norm": 0.605229914188385, "learning_rate": 7.920353982300885e-05, "loss": 0.0842, "step": 5834 }, { "epoch": 0.6325888985255854, "grad_norm": 0.47977203130722046, "learning_rate": 7.919991295517192e-05, "loss": 0.0439, "step": 5835 }, { "epoch": 0.6326973113616652, "grad_norm": 0.3273431360721588, "learning_rate": 7.919628608733498e-05, "loss": 0.0373, "step": 5836 }, { "epoch": 0.6328057241977451, "grad_norm": 0.5382529497146606, "learning_rate": 7.919265921949805e-05, "loss": 0.0761, "step": 5837 }, { "epoch": 0.6329141370338248, "grad_norm": 0.4046858549118042, "learning_rate": 7.918903235166111e-05, "loss": 0.05, "step": 5838 }, { "epoch": 0.6330225498699046, "grad_norm": 0.4131948947906494, "learning_rate": 7.918540548382417e-05, "loss": 0.0408, "step": 5839 }, { "epoch": 0.6331309627059843, "grad_norm": 0.23321054875850677, "learning_rate": 7.918177861598724e-05, "loss": 0.027, "step": 5840 }, { "epoch": 0.6332393755420642, "grad_norm": 0.46484023332595825, "learning_rate": 7.91781517481503e-05, "loss": 0.026, "step": 5841 }, { "epoch": 0.633347788378144, "grad_norm": 0.2189226597547531, "learning_rate": 7.917452488031336e-05, "loss": 0.0166, "step": 5842 }, { "epoch": 0.6334562012142237, "grad_norm": 0.3048436641693115, "learning_rate": 7.917089801247642e-05, "loss": 0.0607, "step": 5843 }, { "epoch": 0.6335646140503035, "grad_norm": 0.333441823720932, "learning_rate": 7.916727114463949e-05, "loss": 0.0374, "step": 5844 }, { "epoch": 0.6336730268863834, "grad_norm": 0.37797704339027405, "learning_rate": 7.916364427680256e-05, "loss": 0.0836, "step": 5845 }, { "epoch": 0.6337814397224631, "grad_norm": 0.3997412919998169, "learning_rate": 7.916001740896562e-05, "loss": 0.0518, "step": 5846 }, { "epoch": 0.6338898525585429, "grad_norm": 0.14932160079479218, "learning_rate": 7.915639054112869e-05, "loss": 0.0165, "step": 5847 }, { "epoch": 0.6339982653946227, "grad_norm": 0.2556487023830414, "learning_rate": 7.915276367329175e-05, "loss": 0.0249, "step": 5848 }, { "epoch": 0.6341066782307025, "grad_norm": 0.45716768503189087, "learning_rate": 7.914913680545482e-05, "loss": 0.0627, "step": 5849 }, { "epoch": 0.6342150910667823, "grad_norm": 0.8863860368728638, "learning_rate": 7.914550993761788e-05, "loss": 0.0694, "step": 5850 }, { "epoch": 0.6343235039028621, "grad_norm": 0.17468293011188507, "learning_rate": 7.914188306978095e-05, "loss": 0.0192, "step": 5851 }, { "epoch": 0.6344319167389418, "grad_norm": 0.23774133622646332, "learning_rate": 7.9138256201944e-05, "loss": 0.0235, "step": 5852 }, { "epoch": 0.6345403295750217, "grad_norm": 0.27176719903945923, "learning_rate": 7.913462933410706e-05, "loss": 0.0293, "step": 5853 }, { "epoch": 0.6346487424111015, "grad_norm": 0.2345561534166336, "learning_rate": 7.913100246627013e-05, "loss": 0.0192, "step": 5854 }, { "epoch": 0.6347571552471812, "grad_norm": 0.6305859088897705, "learning_rate": 7.912737559843319e-05, "loss": 0.0312, "step": 5855 }, { "epoch": 0.6348655680832611, "grad_norm": 0.2491929680109024, "learning_rate": 7.912374873059626e-05, "loss": 0.0229, "step": 5856 }, { "epoch": 0.6349739809193409, "grad_norm": 0.29431456327438354, "learning_rate": 7.912012186275932e-05, "loss": 0.0277, "step": 5857 }, { "epoch": 0.6350823937554206, "grad_norm": 0.22130721807479858, "learning_rate": 7.911649499492239e-05, "loss": 0.0318, "step": 5858 }, { "epoch": 0.6351908065915004, "grad_norm": 0.30577993392944336, "learning_rate": 7.911286812708545e-05, "loss": 0.022, "step": 5859 }, { "epoch": 0.6352992194275803, "grad_norm": 0.9113962054252625, "learning_rate": 7.910924125924852e-05, "loss": 0.0734, "step": 5860 }, { "epoch": 0.63540763226366, "grad_norm": 0.781959593296051, "learning_rate": 7.910561439141158e-05, "loss": 0.0761, "step": 5861 }, { "epoch": 0.6355160450997398, "grad_norm": 0.29970839619636536, "learning_rate": 7.910198752357465e-05, "loss": 0.0552, "step": 5862 }, { "epoch": 0.6356244579358196, "grad_norm": 0.25770804286003113, "learning_rate": 7.909836065573772e-05, "loss": 0.0248, "step": 5863 }, { "epoch": 0.6357328707718994, "grad_norm": 0.24820777773857117, "learning_rate": 7.909473378790077e-05, "loss": 0.0184, "step": 5864 }, { "epoch": 0.6358412836079792, "grad_norm": 0.22403275966644287, "learning_rate": 7.909110692006385e-05, "loss": 0.0296, "step": 5865 }, { "epoch": 0.635949696444059, "grad_norm": 0.37573984265327454, "learning_rate": 7.90874800522269e-05, "loss": 0.0476, "step": 5866 }, { "epoch": 0.6360581092801387, "grad_norm": 0.25583022832870483, "learning_rate": 7.908385318438996e-05, "loss": 0.037, "step": 5867 }, { "epoch": 0.6361665221162186, "grad_norm": 0.3387969136238098, "learning_rate": 7.908022631655303e-05, "loss": 0.0516, "step": 5868 }, { "epoch": 0.6362749349522984, "grad_norm": 0.2666662931442261, "learning_rate": 7.907659944871609e-05, "loss": 0.0243, "step": 5869 }, { "epoch": 0.6363833477883781, "grad_norm": 0.11309655010700226, "learning_rate": 7.907297258087916e-05, "loss": 0.0096, "step": 5870 }, { "epoch": 0.6364917606244579, "grad_norm": 0.2070738673210144, "learning_rate": 7.906934571304222e-05, "loss": 0.0144, "step": 5871 }, { "epoch": 0.6366001734605378, "grad_norm": 0.5321459770202637, "learning_rate": 7.906571884520529e-05, "loss": 0.075, "step": 5872 }, { "epoch": 0.6367085862966175, "grad_norm": 0.2515612244606018, "learning_rate": 7.906209197736834e-05, "loss": 0.0403, "step": 5873 }, { "epoch": 0.6368169991326973, "grad_norm": 0.31177011132240295, "learning_rate": 7.905846510953142e-05, "loss": 0.0328, "step": 5874 }, { "epoch": 0.636925411968777, "grad_norm": 0.7153438925743103, "learning_rate": 7.905483824169447e-05, "loss": 0.0359, "step": 5875 }, { "epoch": 0.6370338248048569, "grad_norm": 0.5262812972068787, "learning_rate": 7.905121137385753e-05, "loss": 0.0602, "step": 5876 }, { "epoch": 0.6371422376409367, "grad_norm": 0.3847036063671112, "learning_rate": 7.90475845060206e-05, "loss": 0.0148, "step": 5877 }, { "epoch": 0.6372506504770165, "grad_norm": 0.7051178216934204, "learning_rate": 7.904395763818366e-05, "loss": 0.063, "step": 5878 }, { "epoch": 0.6373590633130962, "grad_norm": 0.390926331281662, "learning_rate": 7.904033077034674e-05, "loss": 0.0419, "step": 5879 }, { "epoch": 0.6374674761491761, "grad_norm": 0.7987107634544373, "learning_rate": 7.90367039025098e-05, "loss": 0.0939, "step": 5880 }, { "epoch": 0.6375758889852559, "grad_norm": 0.4061731994152069, "learning_rate": 7.903307703467286e-05, "loss": 0.0432, "step": 5881 }, { "epoch": 0.6376843018213356, "grad_norm": 0.16392560303211212, "learning_rate": 7.902945016683593e-05, "loss": 0.0052, "step": 5882 }, { "epoch": 0.6377927146574154, "grad_norm": 0.4797956049442291, "learning_rate": 7.902582329899899e-05, "loss": 0.0973, "step": 5883 }, { "epoch": 0.6379011274934953, "grad_norm": 0.42696526646614075, "learning_rate": 7.902219643116206e-05, "loss": 0.0539, "step": 5884 }, { "epoch": 0.638009540329575, "grad_norm": 0.1764466017484665, "learning_rate": 7.901856956332511e-05, "loss": 0.0228, "step": 5885 }, { "epoch": 0.6381179531656548, "grad_norm": 0.5622677803039551, "learning_rate": 7.901494269548818e-05, "loss": 0.0632, "step": 5886 }, { "epoch": 0.6382263660017347, "grad_norm": 0.45983806252479553, "learning_rate": 7.901131582765124e-05, "loss": 0.0489, "step": 5887 }, { "epoch": 0.6383347788378144, "grad_norm": 0.3956362307071686, "learning_rate": 7.900768895981431e-05, "loss": 0.0312, "step": 5888 }, { "epoch": 0.6384431916738942, "grad_norm": 0.45733627676963806, "learning_rate": 7.900406209197737e-05, "loss": 0.033, "step": 5889 }, { "epoch": 0.638551604509974, "grad_norm": 0.2740009129047394, "learning_rate": 7.900043522414043e-05, "loss": 0.0181, "step": 5890 }, { "epoch": 0.6386600173460538, "grad_norm": 0.794257640838623, "learning_rate": 7.89968083563035e-05, "loss": 0.0221, "step": 5891 }, { "epoch": 0.6387684301821336, "grad_norm": 0.4822302758693695, "learning_rate": 7.899318148846656e-05, "loss": 0.0891, "step": 5892 }, { "epoch": 0.6388768430182133, "grad_norm": 0.31448718905448914, "learning_rate": 7.898955462062963e-05, "loss": 0.0193, "step": 5893 }, { "epoch": 0.6389852558542931, "grad_norm": 0.44804754853248596, "learning_rate": 7.898592775279268e-05, "loss": 0.0864, "step": 5894 }, { "epoch": 0.639093668690373, "grad_norm": 0.23089450597763062, "learning_rate": 7.898230088495576e-05, "loss": 0.0263, "step": 5895 }, { "epoch": 0.6392020815264527, "grad_norm": 0.7050084471702576, "learning_rate": 7.897867401711883e-05, "loss": 0.0388, "step": 5896 }, { "epoch": 0.6393104943625325, "grad_norm": 0.19082897901535034, "learning_rate": 7.897504714928188e-05, "loss": 0.0188, "step": 5897 }, { "epoch": 0.6394189071986123, "grad_norm": 0.5291271805763245, "learning_rate": 7.897142028144495e-05, "loss": 0.0719, "step": 5898 }, { "epoch": 0.6395273200346921, "grad_norm": 0.558183491230011, "learning_rate": 7.896779341360801e-05, "loss": 0.0202, "step": 5899 }, { "epoch": 0.6396357328707719, "grad_norm": 0.06354497373104095, "learning_rate": 7.896416654577108e-05, "loss": 0.0102, "step": 5900 }, { "epoch": 0.6397441457068517, "grad_norm": 0.34011462330818176, "learning_rate": 7.896053967793414e-05, "loss": 0.0131, "step": 5901 }, { "epoch": 0.6398525585429314, "grad_norm": 0.20496800541877747, "learning_rate": 7.895691281009721e-05, "loss": 0.0266, "step": 5902 }, { "epoch": 0.6399609713790113, "grad_norm": 0.6510936617851257, "learning_rate": 7.895328594226027e-05, "loss": 0.0377, "step": 5903 }, { "epoch": 0.6400693842150911, "grad_norm": 0.21334129571914673, "learning_rate": 7.894965907442334e-05, "loss": 0.0244, "step": 5904 }, { "epoch": 0.6401777970511708, "grad_norm": 0.27141958475112915, "learning_rate": 7.89460322065864e-05, "loss": 0.0322, "step": 5905 }, { "epoch": 0.6402862098872506, "grad_norm": 0.16479924321174622, "learning_rate": 7.894240533874945e-05, "loss": 0.0257, "step": 5906 }, { "epoch": 0.6403946227233305, "grad_norm": 0.35204848647117615, "learning_rate": 7.893877847091252e-05, "loss": 0.0669, "step": 5907 }, { "epoch": 0.6405030355594102, "grad_norm": 0.3277839422225952, "learning_rate": 7.893515160307558e-05, "loss": 0.0308, "step": 5908 }, { "epoch": 0.64061144839549, "grad_norm": 0.3322208821773529, "learning_rate": 7.893152473523865e-05, "loss": 0.0758, "step": 5909 }, { "epoch": 0.6407198612315698, "grad_norm": 0.25932344794273376, "learning_rate": 7.892789786740171e-05, "loss": 0.0414, "step": 5910 }, { "epoch": 0.6408282740676496, "grad_norm": 0.3535904884338379, "learning_rate": 7.892427099956478e-05, "loss": 0.0731, "step": 5911 }, { "epoch": 0.6409366869037294, "grad_norm": 0.29897764325141907, "learning_rate": 7.892064413172784e-05, "loss": 0.0228, "step": 5912 }, { "epoch": 0.6410450997398092, "grad_norm": 0.410905122756958, "learning_rate": 7.89170172638909e-05, "loss": 0.0827, "step": 5913 }, { "epoch": 0.641153512575889, "grad_norm": 0.400680273771286, "learning_rate": 7.891339039605398e-05, "loss": 0.0576, "step": 5914 }, { "epoch": 0.6412619254119688, "grad_norm": 0.2300329953432083, "learning_rate": 7.890976352821704e-05, "loss": 0.0239, "step": 5915 }, { "epoch": 0.6413703382480486, "grad_norm": 0.2114139050245285, "learning_rate": 7.890613666038011e-05, "loss": 0.0278, "step": 5916 }, { "epoch": 0.6414787510841283, "grad_norm": 0.6936896443367004, "learning_rate": 7.890250979254317e-05, "loss": 0.0957, "step": 5917 }, { "epoch": 0.6415871639202082, "grad_norm": 0.11970752477645874, "learning_rate": 7.889888292470624e-05, "loss": 0.0117, "step": 5918 }, { "epoch": 0.641695576756288, "grad_norm": 0.2686890959739685, "learning_rate": 7.88952560568693e-05, "loss": 0.0469, "step": 5919 }, { "epoch": 0.6418039895923677, "grad_norm": 0.49923595786094666, "learning_rate": 7.889162918903235e-05, "loss": 0.0563, "step": 5920 }, { "epoch": 0.6419124024284475, "grad_norm": 0.355011522769928, "learning_rate": 7.888800232119542e-05, "loss": 0.0557, "step": 5921 }, { "epoch": 0.6420208152645274, "grad_norm": 0.45762184262275696, "learning_rate": 7.888437545335848e-05, "loss": 0.0457, "step": 5922 }, { "epoch": 0.6421292281006071, "grad_norm": 0.23181547224521637, "learning_rate": 7.888074858552155e-05, "loss": 0.0207, "step": 5923 }, { "epoch": 0.6422376409366869, "grad_norm": 0.1466079205274582, "learning_rate": 7.887712171768461e-05, "loss": 0.0161, "step": 5924 }, { "epoch": 0.6423460537727667, "grad_norm": 0.22411879897117615, "learning_rate": 7.887349484984768e-05, "loss": 0.0109, "step": 5925 }, { "epoch": 0.6424544666088465, "grad_norm": 0.26286303997039795, "learning_rate": 7.886986798201074e-05, "loss": 0.0253, "step": 5926 }, { "epoch": 0.6425628794449263, "grad_norm": 0.3745981752872467, "learning_rate": 7.88662411141738e-05, "loss": 0.074, "step": 5927 }, { "epoch": 0.6426712922810061, "grad_norm": 0.24029721319675446, "learning_rate": 7.886261424633686e-05, "loss": 0.0205, "step": 5928 }, { "epoch": 0.6427797051170858, "grad_norm": 0.40538981556892395, "learning_rate": 7.885898737849992e-05, "loss": 0.0554, "step": 5929 }, { "epoch": 0.6428881179531657, "grad_norm": 0.6848741173744202, "learning_rate": 7.885536051066299e-05, "loss": 0.0733, "step": 5930 }, { "epoch": 0.6429965307892455, "grad_norm": 0.17297770082950592, "learning_rate": 7.885173364282606e-05, "loss": 0.0135, "step": 5931 }, { "epoch": 0.6431049436253252, "grad_norm": 0.7999649047851562, "learning_rate": 7.884810677498913e-05, "loss": 0.0743, "step": 5932 }, { "epoch": 0.643213356461405, "grad_norm": 0.2190714329481125, "learning_rate": 7.884447990715219e-05, "loss": 0.0256, "step": 5933 }, { "epoch": 0.6433217692974849, "grad_norm": 0.4197913110256195, "learning_rate": 7.884085303931525e-05, "loss": 0.0472, "step": 5934 }, { "epoch": 0.6434301821335646, "grad_norm": 0.11376408487558365, "learning_rate": 7.883722617147832e-05, "loss": 0.0127, "step": 5935 }, { "epoch": 0.6435385949696444, "grad_norm": 0.246821328997612, "learning_rate": 7.883359930364138e-05, "loss": 0.0425, "step": 5936 }, { "epoch": 0.6436470078057241, "grad_norm": 0.27545052766799927, "learning_rate": 7.882997243580445e-05, "loss": 0.0304, "step": 5937 }, { "epoch": 0.643755420641804, "grad_norm": 0.22885733842849731, "learning_rate": 7.88263455679675e-05, "loss": 0.0359, "step": 5938 }, { "epoch": 0.6438638334778838, "grad_norm": 0.19828814268112183, "learning_rate": 7.882271870013058e-05, "loss": 0.046, "step": 5939 }, { "epoch": 0.6439722463139635, "grad_norm": 0.3718596398830414, "learning_rate": 7.881909183229363e-05, "loss": 0.027, "step": 5940 }, { "epoch": 0.6440806591500433, "grad_norm": 0.44352367520332336, "learning_rate": 7.88154649644567e-05, "loss": 0.0354, "step": 5941 }, { "epoch": 0.6441890719861232, "grad_norm": 0.1688929945230484, "learning_rate": 7.881183809661976e-05, "loss": 0.0225, "step": 5942 }, { "epoch": 0.644297484822203, "grad_norm": 0.2793000638484955, "learning_rate": 7.880821122878282e-05, "loss": 0.033, "step": 5943 }, { "epoch": 0.6444058976582827, "grad_norm": 0.22006137669086456, "learning_rate": 7.880458436094589e-05, "loss": 0.0289, "step": 5944 }, { "epoch": 0.6445143104943626, "grad_norm": 0.22157765924930573, "learning_rate": 7.880095749310895e-05, "loss": 0.0346, "step": 5945 }, { "epoch": 0.6446227233304423, "grad_norm": 0.3987046182155609, "learning_rate": 7.879733062527202e-05, "loss": 0.037, "step": 5946 }, { "epoch": 0.6447311361665221, "grad_norm": 0.8819200396537781, "learning_rate": 7.879370375743508e-05, "loss": 0.0309, "step": 5947 }, { "epoch": 0.6448395490026019, "grad_norm": 0.40419888496398926, "learning_rate": 7.879007688959815e-05, "loss": 0.0786, "step": 5948 }, { "epoch": 0.6449479618386817, "grad_norm": 0.28862982988357544, "learning_rate": 7.878645002176122e-05, "loss": 0.0367, "step": 5949 }, { "epoch": 0.6450563746747615, "grad_norm": 0.39923757314682007, "learning_rate": 7.878282315392427e-05, "loss": 0.0311, "step": 5950 }, { "epoch": 0.6451647875108413, "grad_norm": 0.4619520604610443, "learning_rate": 7.877919628608735e-05, "loss": 0.0738, "step": 5951 }, { "epoch": 0.645273200346921, "grad_norm": 0.25383907556533813, "learning_rate": 7.87755694182504e-05, "loss": 0.0406, "step": 5952 }, { "epoch": 0.6453816131830009, "grad_norm": 0.3832542300224304, "learning_rate": 7.877194255041347e-05, "loss": 0.0304, "step": 5953 }, { "epoch": 0.6454900260190807, "grad_norm": 0.40041929483413696, "learning_rate": 7.876831568257653e-05, "loss": 0.0465, "step": 5954 }, { "epoch": 0.6455984388551604, "grad_norm": 0.34086719155311584, "learning_rate": 7.87646888147396e-05, "loss": 0.0421, "step": 5955 }, { "epoch": 0.6457068516912402, "grad_norm": 0.32468438148498535, "learning_rate": 7.876106194690266e-05, "loss": 0.0314, "step": 5956 }, { "epoch": 0.6458152645273201, "grad_norm": 0.2398802936077118, "learning_rate": 7.875743507906572e-05, "loss": 0.0322, "step": 5957 }, { "epoch": 0.6459236773633998, "grad_norm": 0.4921908676624298, "learning_rate": 7.875380821122879e-05, "loss": 0.0242, "step": 5958 }, { "epoch": 0.6460320901994796, "grad_norm": 0.14696481823921204, "learning_rate": 7.875018134339184e-05, "loss": 0.0143, "step": 5959 }, { "epoch": 0.6461405030355594, "grad_norm": 0.378601610660553, "learning_rate": 7.874655447555492e-05, "loss": 0.0411, "step": 5960 }, { "epoch": 0.6462489158716392, "grad_norm": 0.40276584029197693, "learning_rate": 7.874292760771797e-05, "loss": 0.0394, "step": 5961 }, { "epoch": 0.646357328707719, "grad_norm": 0.39004412293434143, "learning_rate": 7.873930073988104e-05, "loss": 0.074, "step": 5962 }, { "epoch": 0.6464657415437988, "grad_norm": 0.5418204069137573, "learning_rate": 7.87356738720441e-05, "loss": 0.0414, "step": 5963 }, { "epoch": 0.6465741543798785, "grad_norm": 0.5613149404525757, "learning_rate": 7.873204700420717e-05, "loss": 0.0746, "step": 5964 }, { "epoch": 0.6466825672159584, "grad_norm": 0.6938505172729492, "learning_rate": 7.872842013637023e-05, "loss": 0.1141, "step": 5965 }, { "epoch": 0.6467909800520382, "grad_norm": 0.42647382616996765, "learning_rate": 7.87247932685333e-05, "loss": 0.053, "step": 5966 }, { "epoch": 0.6468993928881179, "grad_norm": 0.35625308752059937, "learning_rate": 7.872116640069637e-05, "loss": 0.0539, "step": 5967 }, { "epoch": 0.6470078057241977, "grad_norm": 0.13415920734405518, "learning_rate": 7.871753953285943e-05, "loss": 0.0095, "step": 5968 }, { "epoch": 0.6471162185602776, "grad_norm": 0.3885526955127716, "learning_rate": 7.87139126650225e-05, "loss": 0.0315, "step": 5969 }, { "epoch": 0.6472246313963573, "grad_norm": 0.37387776374816895, "learning_rate": 7.871028579718556e-05, "loss": 0.0538, "step": 5970 }, { "epoch": 0.6473330442324371, "grad_norm": 0.2090344876050949, "learning_rate": 7.870665892934861e-05, "loss": 0.0479, "step": 5971 }, { "epoch": 0.647441457068517, "grad_norm": 0.5804448127746582, "learning_rate": 7.870303206151168e-05, "loss": 0.0734, "step": 5972 }, { "epoch": 0.6475498699045967, "grad_norm": 0.4095907211303711, "learning_rate": 7.869940519367474e-05, "loss": 0.0511, "step": 5973 }, { "epoch": 0.6476582827406765, "grad_norm": 0.322884738445282, "learning_rate": 7.869577832583781e-05, "loss": 0.0344, "step": 5974 }, { "epoch": 0.6477666955767563, "grad_norm": 0.2617190182209015, "learning_rate": 7.869215145800087e-05, "loss": 0.0169, "step": 5975 }, { "epoch": 0.6478751084128361, "grad_norm": 0.20387569069862366, "learning_rate": 7.868852459016394e-05, "loss": 0.0191, "step": 5976 }, { "epoch": 0.6479835212489159, "grad_norm": 0.21636426448822021, "learning_rate": 7.8684897722327e-05, "loss": 0.0198, "step": 5977 }, { "epoch": 0.6480919340849957, "grad_norm": 0.4185662567615509, "learning_rate": 7.868127085449007e-05, "loss": 0.0311, "step": 5978 }, { "epoch": 0.6482003469210754, "grad_norm": 0.44216710329055786, "learning_rate": 7.867764398665313e-05, "loss": 0.0626, "step": 5979 }, { "epoch": 0.6483087597571553, "grad_norm": 0.5273969769477844, "learning_rate": 7.867401711881618e-05, "loss": 0.0873, "step": 5980 }, { "epoch": 0.6484171725932351, "grad_norm": 0.4169980585575104, "learning_rate": 7.867039025097925e-05, "loss": 0.039, "step": 5981 }, { "epoch": 0.6485255854293148, "grad_norm": 0.17719560861587524, "learning_rate": 7.866676338314231e-05, "loss": 0.0151, "step": 5982 }, { "epoch": 0.6486339982653946, "grad_norm": 0.25253456830978394, "learning_rate": 7.86631365153054e-05, "loss": 0.0347, "step": 5983 }, { "epoch": 0.6487424111014745, "grad_norm": 0.26026201248168945, "learning_rate": 7.865950964746845e-05, "loss": 0.0447, "step": 5984 }, { "epoch": 0.6488508239375542, "grad_norm": 0.5313456654548645, "learning_rate": 7.865588277963152e-05, "loss": 0.0589, "step": 5985 }, { "epoch": 0.648959236773634, "grad_norm": 0.30225470662117004, "learning_rate": 7.865225591179458e-05, "loss": 0.0303, "step": 5986 }, { "epoch": 0.6490676496097137, "grad_norm": 0.40378111600875854, "learning_rate": 7.864862904395764e-05, "loss": 0.05, "step": 5987 }, { "epoch": 0.6491760624457936, "grad_norm": 0.3630000650882721, "learning_rate": 7.864500217612071e-05, "loss": 0.0997, "step": 5988 }, { "epoch": 0.6492844752818734, "grad_norm": 0.4045170843601227, "learning_rate": 7.864137530828377e-05, "loss": 0.04, "step": 5989 }, { "epoch": 0.6493928881179531, "grad_norm": 0.3392059803009033, "learning_rate": 7.863774844044684e-05, "loss": 0.056, "step": 5990 }, { "epoch": 0.6495013009540329, "grad_norm": 0.291637659072876, "learning_rate": 7.86341215726099e-05, "loss": 0.0419, "step": 5991 }, { "epoch": 0.6496097137901128, "grad_norm": 0.4862002432346344, "learning_rate": 7.863049470477297e-05, "loss": 0.0805, "step": 5992 }, { "epoch": 0.6497181266261925, "grad_norm": 0.4240191578865051, "learning_rate": 7.862686783693602e-05, "loss": 0.057, "step": 5993 }, { "epoch": 0.6498265394622723, "grad_norm": 0.45031464099884033, "learning_rate": 7.862324096909908e-05, "loss": 0.0445, "step": 5994 }, { "epoch": 0.6499349522983521, "grad_norm": 0.315368115901947, "learning_rate": 7.861961410126215e-05, "loss": 0.0563, "step": 5995 }, { "epoch": 0.650043365134432, "grad_norm": 0.43244633078575134, "learning_rate": 7.861598723342521e-05, "loss": 0.0917, "step": 5996 }, { "epoch": 0.6501517779705117, "grad_norm": 0.2620636224746704, "learning_rate": 7.861236036558828e-05, "loss": 0.0298, "step": 5997 }, { "epoch": 0.6502601908065915, "grad_norm": 0.4814766049385071, "learning_rate": 7.860873349775134e-05, "loss": 0.0323, "step": 5998 }, { "epoch": 0.6503686036426712, "grad_norm": 0.12179948389530182, "learning_rate": 7.860510662991441e-05, "loss": 0.016, "step": 5999 }, { "epoch": 0.6504770164787511, "grad_norm": 0.28700965642929077, "learning_rate": 7.860147976207748e-05, "loss": 0.0362, "step": 6000 }, { "epoch": 0.6505854293148309, "grad_norm": 0.4291921854019165, "learning_rate": 7.859785289424054e-05, "loss": 0.0691, "step": 6001 }, { "epoch": 0.6506938421509106, "grad_norm": 0.15507782995700836, "learning_rate": 7.859422602640361e-05, "loss": 0.0231, "step": 6002 }, { "epoch": 0.6508022549869905, "grad_norm": 0.24212712049484253, "learning_rate": 7.859059915856666e-05, "loss": 0.0433, "step": 6003 }, { "epoch": 0.6509106678230703, "grad_norm": 0.14203958213329315, "learning_rate": 7.858697229072974e-05, "loss": 0.0119, "step": 6004 }, { "epoch": 0.65101908065915, "grad_norm": 0.19575192034244537, "learning_rate": 7.858334542289279e-05, "loss": 0.0452, "step": 6005 }, { "epoch": 0.6511274934952298, "grad_norm": 0.2702579200267792, "learning_rate": 7.857971855505586e-05, "loss": 0.0292, "step": 6006 }, { "epoch": 0.6512359063313097, "grad_norm": 0.2779538929462433, "learning_rate": 7.857609168721892e-05, "loss": 0.031, "step": 6007 }, { "epoch": 0.6513443191673894, "grad_norm": 0.378816694021225, "learning_rate": 7.857246481938199e-05, "loss": 0.0714, "step": 6008 }, { "epoch": 0.6514527320034692, "grad_norm": 0.5213127732276917, "learning_rate": 7.856883795154505e-05, "loss": 0.0698, "step": 6009 }, { "epoch": 0.651561144839549, "grad_norm": 0.2496594935655594, "learning_rate": 7.856521108370811e-05, "loss": 0.0271, "step": 6010 }, { "epoch": 0.6516695576756288, "grad_norm": 0.2376132309436798, "learning_rate": 7.856158421587118e-05, "loss": 0.038, "step": 6011 }, { "epoch": 0.6517779705117086, "grad_norm": 0.10898800939321518, "learning_rate": 7.855795734803424e-05, "loss": 0.0172, "step": 6012 }, { "epoch": 0.6518863833477884, "grad_norm": 0.3791745603084564, "learning_rate": 7.85543304801973e-05, "loss": 0.0379, "step": 6013 }, { "epoch": 0.6519947961838681, "grad_norm": 0.14682166278362274, "learning_rate": 7.855070361236036e-05, "loss": 0.0206, "step": 6014 }, { "epoch": 0.652103209019948, "grad_norm": 0.39771193265914917, "learning_rate": 7.854707674452343e-05, "loss": 0.0465, "step": 6015 }, { "epoch": 0.6522116218560278, "grad_norm": 0.15442878007888794, "learning_rate": 7.854344987668649e-05, "loss": 0.0198, "step": 6016 }, { "epoch": 0.6523200346921075, "grad_norm": 0.21283964812755585, "learning_rate": 7.853982300884956e-05, "loss": 0.0124, "step": 6017 }, { "epoch": 0.6524284475281873, "grad_norm": 0.27190035581588745, "learning_rate": 7.853619614101263e-05, "loss": 0.044, "step": 6018 }, { "epoch": 0.6525368603642672, "grad_norm": 0.41058769822120667, "learning_rate": 7.853256927317569e-05, "loss": 0.0552, "step": 6019 }, { "epoch": 0.6526452732003469, "grad_norm": 0.2390124797821045, "learning_rate": 7.852894240533876e-05, "loss": 0.0473, "step": 6020 }, { "epoch": 0.6527536860364267, "grad_norm": 0.22369813919067383, "learning_rate": 7.852531553750182e-05, "loss": 0.0202, "step": 6021 }, { "epoch": 0.6528620988725065, "grad_norm": 0.1629997193813324, "learning_rate": 7.852168866966489e-05, "loss": 0.028, "step": 6022 }, { "epoch": 0.6529705117085863, "grad_norm": 0.17625825107097626, "learning_rate": 7.851806180182795e-05, "loss": 0.0416, "step": 6023 }, { "epoch": 0.6530789245446661, "grad_norm": 0.2600859999656677, "learning_rate": 7.8514434933991e-05, "loss": 0.0376, "step": 6024 }, { "epoch": 0.6531873373807459, "grad_norm": 0.1944682002067566, "learning_rate": 7.851080806615408e-05, "loss": 0.0141, "step": 6025 }, { "epoch": 0.6532957502168256, "grad_norm": 0.1909976452589035, "learning_rate": 7.850718119831713e-05, "loss": 0.0327, "step": 6026 }, { "epoch": 0.6534041630529055, "grad_norm": 0.3078341782093048, "learning_rate": 7.85035543304802e-05, "loss": 0.0404, "step": 6027 }, { "epoch": 0.6535125758889853, "grad_norm": 0.31287774443626404, "learning_rate": 7.849992746264326e-05, "loss": 0.0286, "step": 6028 }, { "epoch": 0.653620988725065, "grad_norm": 0.3844403624534607, "learning_rate": 7.849630059480633e-05, "loss": 0.0599, "step": 6029 }, { "epoch": 0.6537294015611448, "grad_norm": 0.33116328716278076, "learning_rate": 7.849267372696939e-05, "loss": 0.0102, "step": 6030 }, { "epoch": 0.6538378143972247, "grad_norm": 0.27703970670700073, "learning_rate": 7.848904685913246e-05, "loss": 0.0252, "step": 6031 }, { "epoch": 0.6539462272333044, "grad_norm": 0.34848088026046753, "learning_rate": 7.848541999129552e-05, "loss": 0.0189, "step": 6032 }, { "epoch": 0.6540546400693842, "grad_norm": 0.2620788812637329, "learning_rate": 7.848179312345857e-05, "loss": 0.0111, "step": 6033 }, { "epoch": 0.6541630529054641, "grad_norm": 0.3615117073059082, "learning_rate": 7.847816625562165e-05, "loss": 0.0429, "step": 6034 }, { "epoch": 0.6542714657415438, "grad_norm": 0.2684210538864136, "learning_rate": 7.847453938778472e-05, "loss": 0.0246, "step": 6035 }, { "epoch": 0.6543798785776236, "grad_norm": 0.6461626887321472, "learning_rate": 7.847091251994779e-05, "loss": 0.0385, "step": 6036 }, { "epoch": 0.6544882914137033, "grad_norm": 0.36942678689956665, "learning_rate": 7.846728565211084e-05, "loss": 0.0364, "step": 6037 }, { "epoch": 0.6545967042497832, "grad_norm": 0.2570076584815979, "learning_rate": 7.84636587842739e-05, "loss": 0.0256, "step": 6038 }, { "epoch": 0.654705117085863, "grad_norm": 0.3703230619430542, "learning_rate": 7.846003191643697e-05, "loss": 0.0839, "step": 6039 }, { "epoch": 0.6548135299219427, "grad_norm": 0.34263989329338074, "learning_rate": 7.845640504860003e-05, "loss": 0.0299, "step": 6040 }, { "epoch": 0.6549219427580225, "grad_norm": 0.16394035518169403, "learning_rate": 7.84527781807631e-05, "loss": 0.0082, "step": 6041 }, { "epoch": 0.6550303555941024, "grad_norm": 0.3130708932876587, "learning_rate": 7.844915131292616e-05, "loss": 0.0347, "step": 6042 }, { "epoch": 0.6551387684301822, "grad_norm": 0.3154146373271942, "learning_rate": 7.844552444508923e-05, "loss": 0.0299, "step": 6043 }, { "epoch": 0.6552471812662619, "grad_norm": 0.2726418972015381, "learning_rate": 7.844189757725229e-05, "loss": 0.0666, "step": 6044 }, { "epoch": 0.6553555941023417, "grad_norm": 0.16779083013534546, "learning_rate": 7.843827070941536e-05, "loss": 0.0159, "step": 6045 }, { "epoch": 0.6554640069384216, "grad_norm": 0.25846555829048157, "learning_rate": 7.843464384157841e-05, "loss": 0.0603, "step": 6046 }, { "epoch": 0.6555724197745013, "grad_norm": 0.2890675365924835, "learning_rate": 7.843101697374147e-05, "loss": 0.0329, "step": 6047 }, { "epoch": 0.6556808326105811, "grad_norm": 0.3069167733192444, "learning_rate": 7.842739010590454e-05, "loss": 0.0293, "step": 6048 }, { "epoch": 0.6557892454466608, "grad_norm": 0.9417468905448914, "learning_rate": 7.84237632380676e-05, "loss": 0.0299, "step": 6049 }, { "epoch": 0.6558976582827407, "grad_norm": 0.28245243430137634, "learning_rate": 7.842013637023067e-05, "loss": 0.0192, "step": 6050 }, { "epoch": 0.6560060711188205, "grad_norm": 0.2665773034095764, "learning_rate": 7.841650950239373e-05, "loss": 0.0349, "step": 6051 }, { "epoch": 0.6561144839549002, "grad_norm": 0.27985501289367676, "learning_rate": 7.84128826345568e-05, "loss": 0.0369, "step": 6052 }, { "epoch": 0.65622289679098, "grad_norm": 0.7347788214683533, "learning_rate": 7.840925576671987e-05, "loss": 0.0391, "step": 6053 }, { "epoch": 0.6563313096270599, "grad_norm": 0.5877689719200134, "learning_rate": 7.840562889888293e-05, "loss": 0.0897, "step": 6054 }, { "epoch": 0.6564397224631396, "grad_norm": 0.3765967786312103, "learning_rate": 7.8402002031046e-05, "loss": 0.0289, "step": 6055 }, { "epoch": 0.6565481352992194, "grad_norm": 0.1725970357656479, "learning_rate": 7.839837516320906e-05, "loss": 0.0195, "step": 6056 }, { "epoch": 0.6566565481352992, "grad_norm": 0.3243428170681, "learning_rate": 7.839474829537213e-05, "loss": 0.0402, "step": 6057 }, { "epoch": 0.656764960971379, "grad_norm": 0.3880769908428192, "learning_rate": 7.839112142753518e-05, "loss": 0.0729, "step": 6058 }, { "epoch": 0.6568733738074588, "grad_norm": 0.4562867283821106, "learning_rate": 7.838749455969825e-05, "loss": 0.0626, "step": 6059 }, { "epoch": 0.6569817866435386, "grad_norm": 0.38841187953948975, "learning_rate": 7.838386769186131e-05, "loss": 0.0354, "step": 6060 }, { "epoch": 0.6570901994796184, "grad_norm": 0.2559155523777008, "learning_rate": 7.838024082402437e-05, "loss": 0.0264, "step": 6061 }, { "epoch": 0.6571986123156982, "grad_norm": 0.3192727267742157, "learning_rate": 7.837661395618744e-05, "loss": 0.0347, "step": 6062 }, { "epoch": 0.657307025151778, "grad_norm": 0.26862242817878723, "learning_rate": 7.83729870883505e-05, "loss": 0.0493, "step": 6063 }, { "epoch": 0.6574154379878577, "grad_norm": 0.4516458511352539, "learning_rate": 7.836936022051357e-05, "loss": 0.0261, "step": 6064 }, { "epoch": 0.6575238508239376, "grad_norm": 0.33939963579177856, "learning_rate": 7.836573335267663e-05, "loss": 0.0199, "step": 6065 }, { "epoch": 0.6576322636600174, "grad_norm": 0.5145288705825806, "learning_rate": 7.83621064848397e-05, "loss": 0.0546, "step": 6066 }, { "epoch": 0.6577406764960971, "grad_norm": 0.2713654339313507, "learning_rate": 7.835847961700275e-05, "loss": 0.0641, "step": 6067 }, { "epoch": 0.6578490893321769, "grad_norm": 0.34072065353393555, "learning_rate": 7.835485274916582e-05, "loss": 0.0406, "step": 6068 }, { "epoch": 0.6579575021682568, "grad_norm": 0.6824490427970886, "learning_rate": 7.83512258813289e-05, "loss": 0.0413, "step": 6069 }, { "epoch": 0.6580659150043365, "grad_norm": 0.4464612603187561, "learning_rate": 7.834759901349195e-05, "loss": 0.0451, "step": 6070 }, { "epoch": 0.6581743278404163, "grad_norm": 0.23380059003829956, "learning_rate": 7.834397214565502e-05, "loss": 0.0187, "step": 6071 }, { "epoch": 0.6582827406764961, "grad_norm": 0.3894174098968506, "learning_rate": 7.834034527781808e-05, "loss": 0.0134, "step": 6072 }, { "epoch": 0.6583911535125759, "grad_norm": 0.5963961482048035, "learning_rate": 7.833671840998115e-05, "loss": 0.0555, "step": 6073 }, { "epoch": 0.6584995663486557, "grad_norm": 0.2982780933380127, "learning_rate": 7.833309154214421e-05, "loss": 0.0193, "step": 6074 }, { "epoch": 0.6586079791847355, "grad_norm": 0.303038090467453, "learning_rate": 7.832946467430727e-05, "loss": 0.0418, "step": 6075 }, { "epoch": 0.6587163920208152, "grad_norm": 0.2338791787624359, "learning_rate": 7.832583780647034e-05, "loss": 0.0281, "step": 6076 }, { "epoch": 0.6588248048568951, "grad_norm": 0.2629540264606476, "learning_rate": 7.83222109386334e-05, "loss": 0.0423, "step": 6077 }, { "epoch": 0.6589332176929749, "grad_norm": 0.06111014634370804, "learning_rate": 7.831858407079647e-05, "loss": 0.0076, "step": 6078 }, { "epoch": 0.6590416305290546, "grad_norm": 0.43498867750167847, "learning_rate": 7.831495720295952e-05, "loss": 0.0846, "step": 6079 }, { "epoch": 0.6591500433651344, "grad_norm": 0.671898365020752, "learning_rate": 7.83113303351226e-05, "loss": 0.0854, "step": 6080 }, { "epoch": 0.6592584562012143, "grad_norm": 0.7318047285079956, "learning_rate": 7.830770346728565e-05, "loss": 0.043, "step": 6081 }, { "epoch": 0.659366869037294, "grad_norm": 0.4430082142353058, "learning_rate": 7.830407659944872e-05, "loss": 0.0584, "step": 6082 }, { "epoch": 0.6594752818733738, "grad_norm": 0.5903115272521973, "learning_rate": 7.830044973161178e-05, "loss": 0.0521, "step": 6083 }, { "epoch": 0.6595836947094535, "grad_norm": 0.24795959889888763, "learning_rate": 7.829682286377484e-05, "loss": 0.0303, "step": 6084 }, { "epoch": 0.6596921075455334, "grad_norm": 0.5429630279541016, "learning_rate": 7.829319599593791e-05, "loss": 0.029, "step": 6085 }, { "epoch": 0.6598005203816132, "grad_norm": 0.2803455889225006, "learning_rate": 7.828956912810097e-05, "loss": 0.0258, "step": 6086 }, { "epoch": 0.659908933217693, "grad_norm": 0.3422605097293854, "learning_rate": 7.828594226026405e-05, "loss": 0.042, "step": 6087 }, { "epoch": 0.6600173460537727, "grad_norm": 0.5430307984352112, "learning_rate": 7.828231539242711e-05, "loss": 0.0593, "step": 6088 }, { "epoch": 0.6601257588898526, "grad_norm": 0.5864975452423096, "learning_rate": 7.827868852459018e-05, "loss": 0.0962, "step": 6089 }, { "epoch": 0.6602341717259324, "grad_norm": 0.3254891037940979, "learning_rate": 7.827506165675324e-05, "loss": 0.0481, "step": 6090 }, { "epoch": 0.6603425845620121, "grad_norm": 0.3263438940048218, "learning_rate": 7.827143478891629e-05, "loss": 0.0267, "step": 6091 }, { "epoch": 0.660450997398092, "grad_norm": 0.2219131737947464, "learning_rate": 7.826780792107936e-05, "loss": 0.024, "step": 6092 }, { "epoch": 0.6605594102341718, "grad_norm": 0.4291214346885681, "learning_rate": 7.826418105324242e-05, "loss": 0.0847, "step": 6093 }, { "epoch": 0.6606678230702515, "grad_norm": 0.5333839654922485, "learning_rate": 7.826055418540549e-05, "loss": 0.0396, "step": 6094 }, { "epoch": 0.6607762359063313, "grad_norm": 0.30980995297431946, "learning_rate": 7.825692731756855e-05, "loss": 0.0593, "step": 6095 }, { "epoch": 0.6608846487424112, "grad_norm": 0.25216689705848694, "learning_rate": 7.825330044973162e-05, "loss": 0.0337, "step": 6096 }, { "epoch": 0.6609930615784909, "grad_norm": 0.33086860179901123, "learning_rate": 7.824967358189468e-05, "loss": 0.0355, "step": 6097 }, { "epoch": 0.6611014744145707, "grad_norm": 0.42359840869903564, "learning_rate": 7.824604671405773e-05, "loss": 0.0551, "step": 6098 }, { "epoch": 0.6612098872506504, "grad_norm": 0.30019327998161316, "learning_rate": 7.82424198462208e-05, "loss": 0.0342, "step": 6099 }, { "epoch": 0.6613183000867303, "grad_norm": 0.27991512417793274, "learning_rate": 7.823879297838386e-05, "loss": 0.0466, "step": 6100 }, { "epoch": 0.6614267129228101, "grad_norm": 0.5843364596366882, "learning_rate": 7.823516611054693e-05, "loss": 0.0355, "step": 6101 }, { "epoch": 0.6615351257588898, "grad_norm": 0.38298356533050537, "learning_rate": 7.823153924270999e-05, "loss": 0.0274, "step": 6102 }, { "epoch": 0.6616435385949696, "grad_norm": 0.09567996859550476, "learning_rate": 7.822791237487306e-05, "loss": 0.0121, "step": 6103 }, { "epoch": 0.6617519514310495, "grad_norm": 0.2893731892108917, "learning_rate": 7.822428550703613e-05, "loss": 0.0493, "step": 6104 }, { "epoch": 0.6618603642671292, "grad_norm": 0.26169687509536743, "learning_rate": 7.822065863919919e-05, "loss": 0.0336, "step": 6105 }, { "epoch": 0.661968777103209, "grad_norm": 0.15553133189678192, "learning_rate": 7.821703177136226e-05, "loss": 0.0242, "step": 6106 }, { "epoch": 0.6620771899392888, "grad_norm": 0.27795156836509705, "learning_rate": 7.821340490352532e-05, "loss": 0.028, "step": 6107 }, { "epoch": 0.6621856027753686, "grad_norm": 0.10216499119997025, "learning_rate": 7.820977803568839e-05, "loss": 0.0259, "step": 6108 }, { "epoch": 0.6622940156114484, "grad_norm": 0.5245131850242615, "learning_rate": 7.820615116785145e-05, "loss": 0.037, "step": 6109 }, { "epoch": 0.6624024284475282, "grad_norm": 0.3006308376789093, "learning_rate": 7.820252430001452e-05, "loss": 0.0345, "step": 6110 }, { "epoch": 0.6625108412836079, "grad_norm": 0.4545457065105438, "learning_rate": 7.819889743217757e-05, "loss": 0.0779, "step": 6111 }, { "epoch": 0.6626192541196878, "grad_norm": 0.2522529363632202, "learning_rate": 7.819527056434065e-05, "loss": 0.0498, "step": 6112 }, { "epoch": 0.6627276669557676, "grad_norm": 0.2036169320344925, "learning_rate": 7.81916436965037e-05, "loss": 0.05, "step": 6113 }, { "epoch": 0.6628360797918473, "grad_norm": 0.4350326359272003, "learning_rate": 7.818801682866676e-05, "loss": 0.0403, "step": 6114 }, { "epoch": 0.6629444926279271, "grad_norm": 0.2200058102607727, "learning_rate": 7.818438996082983e-05, "loss": 0.0307, "step": 6115 }, { "epoch": 0.663052905464007, "grad_norm": 0.15622949600219727, "learning_rate": 7.818076309299289e-05, "loss": 0.0219, "step": 6116 }, { "epoch": 0.6631613183000867, "grad_norm": 0.18201355636119843, "learning_rate": 7.817713622515596e-05, "loss": 0.0305, "step": 6117 }, { "epoch": 0.6632697311361665, "grad_norm": 0.5846558213233948, "learning_rate": 7.817350935731902e-05, "loss": 0.1082, "step": 6118 }, { "epoch": 0.6633781439722463, "grad_norm": 0.3698929250240326, "learning_rate": 7.816988248948209e-05, "loss": 0.051, "step": 6119 }, { "epoch": 0.6634865568083261, "grad_norm": 0.17924416065216064, "learning_rate": 7.816625562164514e-05, "loss": 0.0375, "step": 6120 }, { "epoch": 0.6635949696444059, "grad_norm": 0.2509281039237976, "learning_rate": 7.816262875380822e-05, "loss": 0.0207, "step": 6121 }, { "epoch": 0.6637033824804857, "grad_norm": 0.3123581111431122, "learning_rate": 7.815900188597129e-05, "loss": 0.0432, "step": 6122 }, { "epoch": 0.6638117953165655, "grad_norm": 0.33874499797821045, "learning_rate": 7.815537501813434e-05, "loss": 0.0466, "step": 6123 }, { "epoch": 0.6639202081526453, "grad_norm": 0.2824540138244629, "learning_rate": 7.815174815029741e-05, "loss": 0.0229, "step": 6124 }, { "epoch": 0.6640286209887251, "grad_norm": 0.29163607954978943, "learning_rate": 7.814812128246047e-05, "loss": 0.0246, "step": 6125 }, { "epoch": 0.6641370338248048, "grad_norm": 0.28546273708343506, "learning_rate": 7.814449441462354e-05, "loss": 0.0261, "step": 6126 }, { "epoch": 0.6642454466608847, "grad_norm": 0.08191238343715668, "learning_rate": 7.81408675467866e-05, "loss": 0.0101, "step": 6127 }, { "epoch": 0.6643538594969645, "grad_norm": 0.22249732911586761, "learning_rate": 7.813724067894966e-05, "loss": 0.029, "step": 6128 }, { "epoch": 0.6644622723330442, "grad_norm": 0.3194333016872406, "learning_rate": 7.813361381111273e-05, "loss": 0.0374, "step": 6129 }, { "epoch": 0.664570685169124, "grad_norm": 0.49428728222846985, "learning_rate": 7.812998694327579e-05, "loss": 0.0335, "step": 6130 }, { "epoch": 0.6646790980052039, "grad_norm": 0.5719051361083984, "learning_rate": 7.812636007543886e-05, "loss": 0.0807, "step": 6131 }, { "epoch": 0.6647875108412836, "grad_norm": 0.2193222939968109, "learning_rate": 7.812273320760191e-05, "loss": 0.0098, "step": 6132 }, { "epoch": 0.6648959236773634, "grad_norm": 0.35028350353240967, "learning_rate": 7.811910633976499e-05, "loss": 0.0538, "step": 6133 }, { "epoch": 0.6650043365134432, "grad_norm": 0.2869371473789215, "learning_rate": 7.811547947192804e-05, "loss": 0.0614, "step": 6134 }, { "epoch": 0.665112749349523, "grad_norm": 0.3022940158843994, "learning_rate": 7.811185260409111e-05, "loss": 0.0254, "step": 6135 }, { "epoch": 0.6652211621856028, "grad_norm": 0.2654815912246704, "learning_rate": 7.810822573625417e-05, "loss": 0.0446, "step": 6136 }, { "epoch": 0.6653295750216826, "grad_norm": 0.1975591778755188, "learning_rate": 7.810459886841723e-05, "loss": 0.0316, "step": 6137 }, { "epoch": 0.6654379878577623, "grad_norm": 0.2115669548511505, "learning_rate": 7.810097200058031e-05, "loss": 0.0209, "step": 6138 }, { "epoch": 0.6655464006938422, "grad_norm": 0.3184279501438141, "learning_rate": 7.809734513274337e-05, "loss": 0.0512, "step": 6139 }, { "epoch": 0.665654813529922, "grad_norm": 0.20173047482967377, "learning_rate": 7.809371826490644e-05, "loss": 0.0277, "step": 6140 }, { "epoch": 0.6657632263660017, "grad_norm": 0.1286093294620514, "learning_rate": 7.80900913970695e-05, "loss": 0.0134, "step": 6141 }, { "epoch": 0.6658716392020815, "grad_norm": 0.5085433721542358, "learning_rate": 7.808646452923256e-05, "loss": 0.0305, "step": 6142 }, { "epoch": 0.6659800520381614, "grad_norm": 0.2558194100856781, "learning_rate": 7.808283766139563e-05, "loss": 0.0178, "step": 6143 }, { "epoch": 0.6660884648742411, "grad_norm": 0.3172883093357086, "learning_rate": 7.807921079355868e-05, "loss": 0.0368, "step": 6144 }, { "epoch": 0.6661968777103209, "grad_norm": 0.2540688216686249, "learning_rate": 7.807558392572175e-05, "loss": 0.0171, "step": 6145 }, { "epoch": 0.6663052905464006, "grad_norm": 0.3776855766773224, "learning_rate": 7.807195705788481e-05, "loss": 0.0521, "step": 6146 }, { "epoch": 0.6664137033824805, "grad_norm": 0.6932459473609924, "learning_rate": 7.806833019004788e-05, "loss": 0.0669, "step": 6147 }, { "epoch": 0.6665221162185603, "grad_norm": 0.18753483891487122, "learning_rate": 7.806470332221094e-05, "loss": 0.0264, "step": 6148 }, { "epoch": 0.66663052905464, "grad_norm": 0.34163713455200195, "learning_rate": 7.806107645437401e-05, "loss": 0.0152, "step": 6149 }, { "epoch": 0.6667389418907199, "grad_norm": 0.34954139590263367, "learning_rate": 7.805744958653707e-05, "loss": 0.0395, "step": 6150 }, { "epoch": 0.6668473547267997, "grad_norm": 0.266071081161499, "learning_rate": 7.805382271870013e-05, "loss": 0.0358, "step": 6151 }, { "epoch": 0.6669557675628794, "grad_norm": 0.4093349874019623, "learning_rate": 7.80501958508632e-05, "loss": 0.0288, "step": 6152 }, { "epoch": 0.6670641803989592, "grad_norm": 0.5843667984008789, "learning_rate": 7.804656898302625e-05, "loss": 0.0613, "step": 6153 }, { "epoch": 0.6671725932350391, "grad_norm": 0.21425437927246094, "learning_rate": 7.804294211518932e-05, "loss": 0.031, "step": 6154 }, { "epoch": 0.6672810060711188, "grad_norm": 0.3517363965511322, "learning_rate": 7.803931524735238e-05, "loss": 0.0476, "step": 6155 }, { "epoch": 0.6673894189071986, "grad_norm": 0.35086384415626526, "learning_rate": 7.803568837951545e-05, "loss": 0.0265, "step": 6156 }, { "epoch": 0.6674978317432784, "grad_norm": 0.3405017852783203, "learning_rate": 7.803206151167852e-05, "loss": 0.0401, "step": 6157 }, { "epoch": 0.6676062445793582, "grad_norm": 0.12895074486732483, "learning_rate": 7.802843464384158e-05, "loss": 0.0219, "step": 6158 }, { "epoch": 0.667714657415438, "grad_norm": 0.3573389947414398, "learning_rate": 7.802480777600465e-05, "loss": 0.0356, "step": 6159 }, { "epoch": 0.6678230702515178, "grad_norm": 0.27341148257255554, "learning_rate": 7.802118090816771e-05, "loss": 0.0574, "step": 6160 }, { "epoch": 0.6679314830875975, "grad_norm": 0.13515913486480713, "learning_rate": 7.801755404033078e-05, "loss": 0.0123, "step": 6161 }, { "epoch": 0.6680398959236774, "grad_norm": 0.20027869939804077, "learning_rate": 7.801392717249384e-05, "loss": 0.0142, "step": 6162 }, { "epoch": 0.6681483087597572, "grad_norm": 0.7941882014274597, "learning_rate": 7.801030030465691e-05, "loss": 0.0279, "step": 6163 }, { "epoch": 0.6682567215958369, "grad_norm": 0.11616843193769455, "learning_rate": 7.800667343681997e-05, "loss": 0.0175, "step": 6164 }, { "epoch": 0.6683651344319167, "grad_norm": 0.1412217915058136, "learning_rate": 7.800304656898302e-05, "loss": 0.0136, "step": 6165 }, { "epoch": 0.6684735472679966, "grad_norm": 0.5003151297569275, "learning_rate": 7.79994197011461e-05, "loss": 0.0496, "step": 6166 }, { "epoch": 0.6685819601040763, "grad_norm": 0.7259065508842468, "learning_rate": 7.799579283330915e-05, "loss": 0.0342, "step": 6167 }, { "epoch": 0.6686903729401561, "grad_norm": 0.4229713976383209, "learning_rate": 7.799216596547222e-05, "loss": 0.0296, "step": 6168 }, { "epoch": 0.6687987857762359, "grad_norm": 0.45128875970840454, "learning_rate": 7.798853909763528e-05, "loss": 0.0784, "step": 6169 }, { "epoch": 0.6689071986123157, "grad_norm": 0.12641455233097076, "learning_rate": 7.798491222979835e-05, "loss": 0.0061, "step": 6170 }, { "epoch": 0.6690156114483955, "grad_norm": 0.46075019240379333, "learning_rate": 7.798128536196141e-05, "loss": 0.0332, "step": 6171 }, { "epoch": 0.6691240242844753, "grad_norm": 0.23157277703285217, "learning_rate": 7.797765849412448e-05, "loss": 0.0285, "step": 6172 }, { "epoch": 0.669232437120555, "grad_norm": 0.4204108417034149, "learning_rate": 7.797403162628755e-05, "loss": 0.04, "step": 6173 }, { "epoch": 0.6693408499566349, "grad_norm": 0.660685122013092, "learning_rate": 7.79704047584506e-05, "loss": 0.0519, "step": 6174 }, { "epoch": 0.6694492627927147, "grad_norm": 0.12633123993873596, "learning_rate": 7.796677789061368e-05, "loss": 0.0186, "step": 6175 }, { "epoch": 0.6695576756287944, "grad_norm": 0.09365138411521912, "learning_rate": 7.796315102277673e-05, "loss": 0.0132, "step": 6176 }, { "epoch": 0.6696660884648742, "grad_norm": 0.2327316552400589, "learning_rate": 7.79595241549398e-05, "loss": 0.0368, "step": 6177 }, { "epoch": 0.6697745013009541, "grad_norm": 0.335706889629364, "learning_rate": 7.795589728710286e-05, "loss": 0.0332, "step": 6178 }, { "epoch": 0.6698829141370338, "grad_norm": 0.4126955270767212, "learning_rate": 7.795227041926592e-05, "loss": 0.0269, "step": 6179 }, { "epoch": 0.6699913269731136, "grad_norm": 0.6819767355918884, "learning_rate": 7.794864355142899e-05, "loss": 0.0644, "step": 6180 }, { "epoch": 0.6700997398091935, "grad_norm": 0.386339008808136, "learning_rate": 7.794501668359205e-05, "loss": 0.0175, "step": 6181 }, { "epoch": 0.6702081526452732, "grad_norm": 0.4016406834125519, "learning_rate": 7.794138981575512e-05, "loss": 0.049, "step": 6182 }, { "epoch": 0.670316565481353, "grad_norm": 0.6704956293106079, "learning_rate": 7.793776294791818e-05, "loss": 0.0457, "step": 6183 }, { "epoch": 0.6704249783174328, "grad_norm": 0.7812233567237854, "learning_rate": 7.793413608008125e-05, "loss": 0.0525, "step": 6184 }, { "epoch": 0.6705333911535126, "grad_norm": 0.5132724046707153, "learning_rate": 7.79305092122443e-05, "loss": 0.0256, "step": 6185 }, { "epoch": 0.6706418039895924, "grad_norm": 0.21677614748477936, "learning_rate": 7.792688234440738e-05, "loss": 0.0275, "step": 6186 }, { "epoch": 0.6707502168256722, "grad_norm": 0.25431978702545166, "learning_rate": 7.792325547657043e-05, "loss": 0.0132, "step": 6187 }, { "epoch": 0.6708586296617519, "grad_norm": 0.28988194465637207, "learning_rate": 7.791962860873349e-05, "loss": 0.0275, "step": 6188 }, { "epoch": 0.6709670424978318, "grad_norm": 0.5572949051856995, "learning_rate": 7.791600174089656e-05, "loss": 0.0877, "step": 6189 }, { "epoch": 0.6710754553339116, "grad_norm": 0.38412752747535706, "learning_rate": 7.791237487305963e-05, "loss": 0.0331, "step": 6190 }, { "epoch": 0.6711838681699913, "grad_norm": 0.43238115310668945, "learning_rate": 7.79087480052227e-05, "loss": 0.0493, "step": 6191 }, { "epoch": 0.6712922810060711, "grad_norm": 0.35726436972618103, "learning_rate": 7.790512113738576e-05, "loss": 0.0406, "step": 6192 }, { "epoch": 0.671400693842151, "grad_norm": 0.4419625699520111, "learning_rate": 7.790149426954883e-05, "loss": 0.0376, "step": 6193 }, { "epoch": 0.6715091066782307, "grad_norm": 0.17137911915779114, "learning_rate": 7.789786740171189e-05, "loss": 0.0158, "step": 6194 }, { "epoch": 0.6716175195143105, "grad_norm": 0.43367689847946167, "learning_rate": 7.789424053387495e-05, "loss": 0.0732, "step": 6195 }, { "epoch": 0.6717259323503902, "grad_norm": 0.34548741579055786, "learning_rate": 7.789061366603802e-05, "loss": 0.0219, "step": 6196 }, { "epoch": 0.6718343451864701, "grad_norm": 0.4423328936100006, "learning_rate": 7.788698679820107e-05, "loss": 0.076, "step": 6197 }, { "epoch": 0.6719427580225499, "grad_norm": 0.7520508170127869, "learning_rate": 7.788335993036415e-05, "loss": 0.0538, "step": 6198 }, { "epoch": 0.6720511708586296, "grad_norm": 0.3724397122859955, "learning_rate": 7.78797330625272e-05, "loss": 0.0291, "step": 6199 }, { "epoch": 0.6721595836947094, "grad_norm": 0.4806876480579376, "learning_rate": 7.787610619469027e-05, "loss": 0.0532, "step": 6200 }, { "epoch": 0.6722679965307893, "grad_norm": 0.29527705907821655, "learning_rate": 7.787247932685333e-05, "loss": 0.029, "step": 6201 }, { "epoch": 0.672376409366869, "grad_norm": 0.24513499438762665, "learning_rate": 7.78688524590164e-05, "loss": 0.0562, "step": 6202 }, { "epoch": 0.6724848222029488, "grad_norm": 0.2463676631450653, "learning_rate": 7.786522559117946e-05, "loss": 0.0443, "step": 6203 }, { "epoch": 0.6725932350390286, "grad_norm": 0.24143241345882416, "learning_rate": 7.786159872334252e-05, "loss": 0.0303, "step": 6204 }, { "epoch": 0.6727016478751084, "grad_norm": 0.3680666983127594, "learning_rate": 7.785797185550559e-05, "loss": 0.0259, "step": 6205 }, { "epoch": 0.6728100607111882, "grad_norm": 0.21220877766609192, "learning_rate": 7.785434498766864e-05, "loss": 0.0388, "step": 6206 }, { "epoch": 0.672918473547268, "grad_norm": 0.2353401780128479, "learning_rate": 7.785071811983173e-05, "loss": 0.0188, "step": 6207 }, { "epoch": 0.6730268863833477, "grad_norm": 0.27736037969589233, "learning_rate": 7.784709125199479e-05, "loss": 0.0267, "step": 6208 }, { "epoch": 0.6731352992194276, "grad_norm": 0.26409873366355896, "learning_rate": 7.784346438415784e-05, "loss": 0.0239, "step": 6209 }, { "epoch": 0.6732437120555074, "grad_norm": 0.5223110914230347, "learning_rate": 7.783983751632091e-05, "loss": 0.0806, "step": 6210 }, { "epoch": 0.6733521248915871, "grad_norm": 0.25404641032218933, "learning_rate": 7.783621064848397e-05, "loss": 0.0188, "step": 6211 }, { "epoch": 0.673460537727667, "grad_norm": 0.28350770473480225, "learning_rate": 7.783258378064704e-05, "loss": 0.0147, "step": 6212 }, { "epoch": 0.6735689505637468, "grad_norm": 0.32935914397239685, "learning_rate": 7.78289569128101e-05, "loss": 0.0281, "step": 6213 }, { "epoch": 0.6736773633998265, "grad_norm": 0.4317731559276581, "learning_rate": 7.782533004497317e-05, "loss": 0.0361, "step": 6214 }, { "epoch": 0.6737857762359063, "grad_norm": 0.5353265404701233, "learning_rate": 7.782170317713623e-05, "loss": 0.0672, "step": 6215 }, { "epoch": 0.6738941890719862, "grad_norm": 0.32264697551727295, "learning_rate": 7.78180763092993e-05, "loss": 0.0603, "step": 6216 }, { "epoch": 0.6740026019080659, "grad_norm": 0.2426634430885315, "learning_rate": 7.781444944146236e-05, "loss": 0.042, "step": 6217 }, { "epoch": 0.6741110147441457, "grad_norm": 0.7086910009384155, "learning_rate": 7.781082257362541e-05, "loss": 0.0858, "step": 6218 }, { "epoch": 0.6742194275802255, "grad_norm": 0.2514409124851227, "learning_rate": 7.780719570578848e-05, "loss": 0.0515, "step": 6219 }, { "epoch": 0.6743278404163053, "grad_norm": 0.6619516611099243, "learning_rate": 7.780356883795154e-05, "loss": 0.1032, "step": 6220 }, { "epoch": 0.6744362532523851, "grad_norm": 0.22127309441566467, "learning_rate": 7.779994197011461e-05, "loss": 0.0127, "step": 6221 }, { "epoch": 0.6745446660884649, "grad_norm": 0.21141335368156433, "learning_rate": 7.779631510227767e-05, "loss": 0.0386, "step": 6222 }, { "epoch": 0.6746530789245446, "grad_norm": 0.1289149671792984, "learning_rate": 7.779268823444074e-05, "loss": 0.0103, "step": 6223 }, { "epoch": 0.6747614917606245, "grad_norm": 0.21544092893600464, "learning_rate": 7.77890613666038e-05, "loss": 0.0255, "step": 6224 }, { "epoch": 0.6748699045967043, "grad_norm": 0.4348982870578766, "learning_rate": 7.778543449876687e-05, "loss": 0.0961, "step": 6225 }, { "epoch": 0.674978317432784, "grad_norm": 0.7346396446228027, "learning_rate": 7.778180763092994e-05, "loss": 0.1205, "step": 6226 }, { "epoch": 0.6750867302688638, "grad_norm": 0.5709318518638611, "learning_rate": 7.7778180763093e-05, "loss": 0.091, "step": 6227 }, { "epoch": 0.6751951431049437, "grad_norm": 0.6939243078231812, "learning_rate": 7.777455389525607e-05, "loss": 0.0887, "step": 6228 }, { "epoch": 0.6753035559410234, "grad_norm": 0.3160358667373657, "learning_rate": 7.777092702741913e-05, "loss": 0.0629, "step": 6229 }, { "epoch": 0.6754119687771032, "grad_norm": 0.22207708656787872, "learning_rate": 7.77673001595822e-05, "loss": 0.0175, "step": 6230 }, { "epoch": 0.675520381613183, "grad_norm": 0.33793920278549194, "learning_rate": 7.776367329174525e-05, "loss": 0.0372, "step": 6231 }, { "epoch": 0.6756287944492628, "grad_norm": 0.16142511367797852, "learning_rate": 7.776004642390831e-05, "loss": 0.0275, "step": 6232 }, { "epoch": 0.6757372072853426, "grad_norm": 0.17406155169010162, "learning_rate": 7.775641955607138e-05, "loss": 0.0293, "step": 6233 }, { "epoch": 0.6758456201214224, "grad_norm": 0.16254234313964844, "learning_rate": 7.775279268823444e-05, "loss": 0.0207, "step": 6234 }, { "epoch": 0.6759540329575021, "grad_norm": 0.28303176164627075, "learning_rate": 7.774916582039751e-05, "loss": 0.039, "step": 6235 }, { "epoch": 0.676062445793582, "grad_norm": 0.341673344373703, "learning_rate": 7.774553895256057e-05, "loss": 0.0461, "step": 6236 }, { "epoch": 0.6761708586296618, "grad_norm": 0.4049296975135803, "learning_rate": 7.774191208472364e-05, "loss": 0.0489, "step": 6237 }, { "epoch": 0.6762792714657415, "grad_norm": 0.26519775390625, "learning_rate": 7.77382852168867e-05, "loss": 0.0304, "step": 6238 }, { "epoch": 0.6763876843018214, "grad_norm": 0.4477740526199341, "learning_rate": 7.773465834904977e-05, "loss": 0.0772, "step": 6239 }, { "epoch": 0.6764960971379012, "grad_norm": 0.07485321164131165, "learning_rate": 7.773103148121282e-05, "loss": 0.0092, "step": 6240 }, { "epoch": 0.6766045099739809, "grad_norm": 0.3142589032649994, "learning_rate": 7.772740461337588e-05, "loss": 0.042, "step": 6241 }, { "epoch": 0.6767129228100607, "grad_norm": 0.13977865874767303, "learning_rate": 7.772377774553897e-05, "loss": 0.0227, "step": 6242 }, { "epoch": 0.6768213356461406, "grad_norm": 0.43547117710113525, "learning_rate": 7.772015087770202e-05, "loss": 0.0642, "step": 6243 }, { "epoch": 0.6769297484822203, "grad_norm": 0.08828207850456238, "learning_rate": 7.77165240098651e-05, "loss": 0.0157, "step": 6244 }, { "epoch": 0.6770381613183001, "grad_norm": 0.1696087270975113, "learning_rate": 7.771289714202815e-05, "loss": 0.0274, "step": 6245 }, { "epoch": 0.6771465741543798, "grad_norm": 0.4479921758174896, "learning_rate": 7.770927027419121e-05, "loss": 0.0307, "step": 6246 }, { "epoch": 0.6772549869904597, "grad_norm": 0.24440860748291016, "learning_rate": 7.770564340635428e-05, "loss": 0.0308, "step": 6247 }, { "epoch": 0.6773633998265395, "grad_norm": 0.27144598960876465, "learning_rate": 7.770201653851734e-05, "loss": 0.0542, "step": 6248 }, { "epoch": 0.6774718126626192, "grad_norm": 0.26966071128845215, "learning_rate": 7.769838967068041e-05, "loss": 0.0455, "step": 6249 }, { "epoch": 0.677580225498699, "grad_norm": 0.2923489212989807, "learning_rate": 7.769476280284346e-05, "loss": 0.0488, "step": 6250 }, { "epoch": 0.6776886383347789, "grad_norm": 0.5426903367042542, "learning_rate": 7.769113593500654e-05, "loss": 0.0574, "step": 6251 }, { "epoch": 0.6777970511708586, "grad_norm": 0.29149821400642395, "learning_rate": 7.768750906716959e-05, "loss": 0.0448, "step": 6252 }, { "epoch": 0.6779054640069384, "grad_norm": 0.14547160267829895, "learning_rate": 7.768388219933266e-05, "loss": 0.0165, "step": 6253 }, { "epoch": 0.6780138768430182, "grad_norm": 0.42409253120422363, "learning_rate": 7.768025533149572e-05, "loss": 0.0509, "step": 6254 }, { "epoch": 0.678122289679098, "grad_norm": 0.308027982711792, "learning_rate": 7.767662846365878e-05, "loss": 0.0574, "step": 6255 }, { "epoch": 0.6782307025151778, "grad_norm": 0.18201695382595062, "learning_rate": 7.767300159582185e-05, "loss": 0.0246, "step": 6256 }, { "epoch": 0.6783391153512576, "grad_norm": 0.6744424104690552, "learning_rate": 7.766937472798491e-05, "loss": 0.0728, "step": 6257 }, { "epoch": 0.6784475281873373, "grad_norm": 0.21873055398464203, "learning_rate": 7.766574786014798e-05, "loss": 0.0249, "step": 6258 }, { "epoch": 0.6785559410234172, "grad_norm": 0.35899025201797485, "learning_rate": 7.766212099231105e-05, "loss": 0.0592, "step": 6259 }, { "epoch": 0.678664353859497, "grad_norm": 0.39648979902267456, "learning_rate": 7.76584941244741e-05, "loss": 0.0691, "step": 6260 }, { "epoch": 0.6787727666955767, "grad_norm": 0.2788846492767334, "learning_rate": 7.765486725663718e-05, "loss": 0.0425, "step": 6261 }, { "epoch": 0.6788811795316565, "grad_norm": 0.298380047082901, "learning_rate": 7.765124038880023e-05, "loss": 0.054, "step": 6262 }, { "epoch": 0.6789895923677364, "grad_norm": 0.21738936007022858, "learning_rate": 7.76476135209633e-05, "loss": 0.0219, "step": 6263 }, { "epoch": 0.6790980052038161, "grad_norm": 1.2542351484298706, "learning_rate": 7.764398665312636e-05, "loss": 0.2722, "step": 6264 }, { "epoch": 0.6792064180398959, "grad_norm": 0.09186681360006332, "learning_rate": 7.764035978528943e-05, "loss": 0.0129, "step": 6265 }, { "epoch": 0.6793148308759757, "grad_norm": 0.5464978814125061, "learning_rate": 7.763673291745249e-05, "loss": 0.0297, "step": 6266 }, { "epoch": 0.6794232437120555, "grad_norm": 0.38001585006713867, "learning_rate": 7.763310604961556e-05, "loss": 0.0514, "step": 6267 }, { "epoch": 0.6795316565481353, "grad_norm": 0.33226120471954346, "learning_rate": 7.762947918177862e-05, "loss": 0.0593, "step": 6268 }, { "epoch": 0.6796400693842151, "grad_norm": 0.13418756425380707, "learning_rate": 7.762585231394168e-05, "loss": 0.0126, "step": 6269 }, { "epoch": 0.6797484822202949, "grad_norm": 0.42256301641464233, "learning_rate": 7.762222544610475e-05, "loss": 0.0419, "step": 6270 }, { "epoch": 0.6798568950563747, "grad_norm": 0.2898159623146057, "learning_rate": 7.76185985782678e-05, "loss": 0.0273, "step": 6271 }, { "epoch": 0.6799653078924545, "grad_norm": 0.4171654284000397, "learning_rate": 7.761497171043088e-05, "loss": 0.0483, "step": 6272 }, { "epoch": 0.6800737207285342, "grad_norm": 0.5938629508018494, "learning_rate": 7.761134484259393e-05, "loss": 0.0716, "step": 6273 }, { "epoch": 0.6801821335646141, "grad_norm": 0.4289131760597229, "learning_rate": 7.7607717974757e-05, "loss": 0.031, "step": 6274 }, { "epoch": 0.6802905464006939, "grad_norm": 0.32348597049713135, "learning_rate": 7.760409110692006e-05, "loss": 0.0324, "step": 6275 }, { "epoch": 0.6803989592367736, "grad_norm": 0.20771361887454987, "learning_rate": 7.760046423908313e-05, "loss": 0.04, "step": 6276 }, { "epoch": 0.6805073720728534, "grad_norm": 0.24042439460754395, "learning_rate": 7.75968373712462e-05, "loss": 0.0228, "step": 6277 }, { "epoch": 0.6806157849089333, "grad_norm": 0.20259582996368408, "learning_rate": 7.759321050340926e-05, "loss": 0.013, "step": 6278 }, { "epoch": 0.680724197745013, "grad_norm": 0.3560478091239929, "learning_rate": 7.758958363557233e-05, "loss": 0.0463, "step": 6279 }, { "epoch": 0.6808326105810928, "grad_norm": 0.16042661666870117, "learning_rate": 7.758595676773539e-05, "loss": 0.0125, "step": 6280 }, { "epoch": 0.6809410234171726, "grad_norm": 0.39316505193710327, "learning_rate": 7.758232989989846e-05, "loss": 0.0219, "step": 6281 }, { "epoch": 0.6810494362532524, "grad_norm": 0.17342425882816315, "learning_rate": 7.757870303206152e-05, "loss": 0.0163, "step": 6282 }, { "epoch": 0.6811578490893322, "grad_norm": 0.43500715494155884, "learning_rate": 7.757507616422457e-05, "loss": 0.056, "step": 6283 }, { "epoch": 0.681266261925412, "grad_norm": 0.47184351086616516, "learning_rate": 7.757144929638764e-05, "loss": 0.048, "step": 6284 }, { "epoch": 0.6813746747614917, "grad_norm": 0.2391439974308014, "learning_rate": 7.75678224285507e-05, "loss": 0.0341, "step": 6285 }, { "epoch": 0.6814830875975716, "grad_norm": 0.33334845304489136, "learning_rate": 7.756419556071377e-05, "loss": 0.0366, "step": 6286 }, { "epoch": 0.6815915004336514, "grad_norm": 0.4853874146938324, "learning_rate": 7.756056869287683e-05, "loss": 0.048, "step": 6287 }, { "epoch": 0.6816999132697311, "grad_norm": 0.24616527557373047, "learning_rate": 7.75569418250399e-05, "loss": 0.0212, "step": 6288 }, { "epoch": 0.6818083261058109, "grad_norm": 0.31870779395103455, "learning_rate": 7.755331495720296e-05, "loss": 0.0296, "step": 6289 }, { "epoch": 0.6819167389418908, "grad_norm": 0.22297009825706482, "learning_rate": 7.754968808936603e-05, "loss": 0.0239, "step": 6290 }, { "epoch": 0.6820251517779705, "grad_norm": 0.3935336470603943, "learning_rate": 7.754606122152909e-05, "loss": 0.0395, "step": 6291 }, { "epoch": 0.6821335646140503, "grad_norm": 0.5757524967193604, "learning_rate": 7.754243435369214e-05, "loss": 0.0396, "step": 6292 }, { "epoch": 0.68224197745013, "grad_norm": 0.3158644139766693, "learning_rate": 7.753880748585521e-05, "loss": 0.0251, "step": 6293 }, { "epoch": 0.6823503902862099, "grad_norm": 0.2991614043712616, "learning_rate": 7.753518061801829e-05, "loss": 0.037, "step": 6294 }, { "epoch": 0.6824588031222897, "grad_norm": 0.23750634491443634, "learning_rate": 7.753155375018136e-05, "loss": 0.0294, "step": 6295 }, { "epoch": 0.6825672159583694, "grad_norm": 0.3161260187625885, "learning_rate": 7.752792688234441e-05, "loss": 0.0308, "step": 6296 }, { "epoch": 0.6826756287944492, "grad_norm": 0.22472533583641052, "learning_rate": 7.752430001450748e-05, "loss": 0.0177, "step": 6297 }, { "epoch": 0.6827840416305291, "grad_norm": 0.3281901776790619, "learning_rate": 7.752067314667054e-05, "loss": 0.0494, "step": 6298 }, { "epoch": 0.6828924544666088, "grad_norm": 0.4139805734157562, "learning_rate": 7.75170462788336e-05, "loss": 0.0702, "step": 6299 }, { "epoch": 0.6830008673026886, "grad_norm": 0.04714877903461456, "learning_rate": 7.751341941099667e-05, "loss": 0.0034, "step": 6300 }, { "epoch": 0.6831092801387685, "grad_norm": 0.2979462146759033, "learning_rate": 7.750979254315973e-05, "loss": 0.0419, "step": 6301 }, { "epoch": 0.6832176929748482, "grad_norm": 0.46067628264427185, "learning_rate": 7.75061656753228e-05, "loss": 0.039, "step": 6302 }, { "epoch": 0.683326105810928, "grad_norm": 0.1970362514257431, "learning_rate": 7.750253880748586e-05, "loss": 0.0176, "step": 6303 }, { "epoch": 0.6834345186470078, "grad_norm": 0.3651900589466095, "learning_rate": 7.749891193964893e-05, "loss": 0.0244, "step": 6304 }, { "epoch": 0.6835429314830876, "grad_norm": 0.21645912528038025, "learning_rate": 7.749528507181198e-05, "loss": 0.0415, "step": 6305 }, { "epoch": 0.6836513443191674, "grad_norm": 0.6755011081695557, "learning_rate": 7.749165820397505e-05, "loss": 0.098, "step": 6306 }, { "epoch": 0.6837597571552472, "grad_norm": 0.32557785511016846, "learning_rate": 7.748803133613811e-05, "loss": 0.0388, "step": 6307 }, { "epoch": 0.6838681699913269, "grad_norm": 0.6174945831298828, "learning_rate": 7.748440446830117e-05, "loss": 0.0841, "step": 6308 }, { "epoch": 0.6839765828274068, "grad_norm": 0.461149126291275, "learning_rate": 7.748077760046424e-05, "loss": 0.0316, "step": 6309 }, { "epoch": 0.6840849956634866, "grad_norm": 0.18584410846233368, "learning_rate": 7.74771507326273e-05, "loss": 0.0175, "step": 6310 }, { "epoch": 0.6841934084995663, "grad_norm": 0.36316394805908203, "learning_rate": 7.747352386479038e-05, "loss": 0.0418, "step": 6311 }, { "epoch": 0.6843018213356461, "grad_norm": 0.533566415309906, "learning_rate": 7.746989699695344e-05, "loss": 0.0408, "step": 6312 }, { "epoch": 0.684410234171726, "grad_norm": 0.2191110998392105, "learning_rate": 7.74662701291165e-05, "loss": 0.024, "step": 6313 }, { "epoch": 0.6845186470078057, "grad_norm": 0.2527755796909332, "learning_rate": 7.746264326127957e-05, "loss": 0.0278, "step": 6314 }, { "epoch": 0.6846270598438855, "grad_norm": 0.3116665780544281, "learning_rate": 7.745901639344263e-05, "loss": 0.0149, "step": 6315 }, { "epoch": 0.6847354726799653, "grad_norm": 1.0508906841278076, "learning_rate": 7.74553895256057e-05, "loss": 0.1042, "step": 6316 }, { "epoch": 0.6848438855160451, "grad_norm": 0.1978946179151535, "learning_rate": 7.745176265776875e-05, "loss": 0.0415, "step": 6317 }, { "epoch": 0.6849522983521249, "grad_norm": 0.12287262082099915, "learning_rate": 7.744813578993182e-05, "loss": 0.0094, "step": 6318 }, { "epoch": 0.6850607111882047, "grad_norm": 0.27735891938209534, "learning_rate": 7.744450892209488e-05, "loss": 0.0208, "step": 6319 }, { "epoch": 0.6851691240242844, "grad_norm": 0.15112990140914917, "learning_rate": 7.744088205425795e-05, "loss": 0.0305, "step": 6320 }, { "epoch": 0.6852775368603643, "grad_norm": 0.4617041349411011, "learning_rate": 7.743725518642101e-05, "loss": 0.0408, "step": 6321 }, { "epoch": 0.6853859496964441, "grad_norm": 0.2554444372653961, "learning_rate": 7.743362831858407e-05, "loss": 0.0565, "step": 6322 }, { "epoch": 0.6854943625325238, "grad_norm": 1.0163846015930176, "learning_rate": 7.743000145074714e-05, "loss": 0.1251, "step": 6323 }, { "epoch": 0.6856027753686036, "grad_norm": 0.38970649242401123, "learning_rate": 7.74263745829102e-05, "loss": 0.0385, "step": 6324 }, { "epoch": 0.6857111882046835, "grad_norm": 0.27875205874443054, "learning_rate": 7.742274771507327e-05, "loss": 0.0381, "step": 6325 }, { "epoch": 0.6858196010407632, "grad_norm": 0.1722932606935501, "learning_rate": 7.741912084723632e-05, "loss": 0.0171, "step": 6326 }, { "epoch": 0.685928013876843, "grad_norm": 0.4445430338382721, "learning_rate": 7.74154939793994e-05, "loss": 0.0534, "step": 6327 }, { "epoch": 0.6860364267129229, "grad_norm": 0.49877139925956726, "learning_rate": 7.741186711156247e-05, "loss": 0.0529, "step": 6328 }, { "epoch": 0.6861448395490026, "grad_norm": 0.12738743424415588, "learning_rate": 7.740824024372552e-05, "loss": 0.0143, "step": 6329 }, { "epoch": 0.6862532523850824, "grad_norm": 0.2450479418039322, "learning_rate": 7.74046133758886e-05, "loss": 0.0279, "step": 6330 }, { "epoch": 0.6863616652211622, "grad_norm": 0.30027836561203003, "learning_rate": 7.740098650805165e-05, "loss": 0.0175, "step": 6331 }, { "epoch": 0.686470078057242, "grad_norm": 0.34149566292762756, "learning_rate": 7.739735964021472e-05, "loss": 0.0518, "step": 6332 }, { "epoch": 0.6865784908933218, "grad_norm": 0.3168095648288727, "learning_rate": 7.739373277237778e-05, "loss": 0.0209, "step": 6333 }, { "epoch": 0.6866869037294016, "grad_norm": 0.35418465733528137, "learning_rate": 7.739010590454085e-05, "loss": 0.0555, "step": 6334 }, { "epoch": 0.6867953165654813, "grad_norm": 0.5367152094841003, "learning_rate": 7.738647903670391e-05, "loss": 0.0652, "step": 6335 }, { "epoch": 0.6869037294015612, "grad_norm": 0.30538803339004517, "learning_rate": 7.738285216886696e-05, "loss": 0.0458, "step": 6336 }, { "epoch": 0.687012142237641, "grad_norm": 0.33971312642097473, "learning_rate": 7.737922530103004e-05, "loss": 0.0421, "step": 6337 }, { "epoch": 0.6871205550737207, "grad_norm": 0.46832531690597534, "learning_rate": 7.737559843319309e-05, "loss": 0.0479, "step": 6338 }, { "epoch": 0.6872289679098005, "grad_norm": 0.3547801673412323, "learning_rate": 7.737197156535616e-05, "loss": 0.0276, "step": 6339 }, { "epoch": 0.6873373807458804, "grad_norm": 0.09114561975002289, "learning_rate": 7.736834469751922e-05, "loss": 0.0157, "step": 6340 }, { "epoch": 0.6874457935819601, "grad_norm": 0.22733157873153687, "learning_rate": 7.736471782968229e-05, "loss": 0.0625, "step": 6341 }, { "epoch": 0.6875542064180399, "grad_norm": 0.5667079091072083, "learning_rate": 7.736109096184535e-05, "loss": 0.0497, "step": 6342 }, { "epoch": 0.6876626192541196, "grad_norm": 0.3651130199432373, "learning_rate": 7.735746409400842e-05, "loss": 0.0447, "step": 6343 }, { "epoch": 0.6877710320901995, "grad_norm": 0.2917006313800812, "learning_rate": 7.735383722617148e-05, "loss": 0.0573, "step": 6344 }, { "epoch": 0.6878794449262793, "grad_norm": 0.3638843297958374, "learning_rate": 7.735021035833453e-05, "loss": 0.0593, "step": 6345 }, { "epoch": 0.687987857762359, "grad_norm": 0.1760440170764923, "learning_rate": 7.734658349049762e-05, "loss": 0.0314, "step": 6346 }, { "epoch": 0.6880962705984388, "grad_norm": 0.26311543583869934, "learning_rate": 7.734295662266068e-05, "loss": 0.0286, "step": 6347 }, { "epoch": 0.6882046834345187, "grad_norm": 0.23614634573459625, "learning_rate": 7.733932975482375e-05, "loss": 0.0339, "step": 6348 }, { "epoch": 0.6883130962705984, "grad_norm": 0.1855171173810959, "learning_rate": 7.73357028869868e-05, "loss": 0.0468, "step": 6349 }, { "epoch": 0.6884215091066782, "grad_norm": 0.12213863432407379, "learning_rate": 7.733207601914986e-05, "loss": 0.0133, "step": 6350 }, { "epoch": 0.688529921942758, "grad_norm": 0.46837374567985535, "learning_rate": 7.732844915131293e-05, "loss": 0.0427, "step": 6351 }, { "epoch": 0.6886383347788378, "grad_norm": 0.3603399097919464, "learning_rate": 7.732482228347599e-05, "loss": 0.0305, "step": 6352 }, { "epoch": 0.6887467476149176, "grad_norm": 0.3925051689147949, "learning_rate": 7.732119541563906e-05, "loss": 0.0543, "step": 6353 }, { "epoch": 0.6888551604509974, "grad_norm": 0.13661763072013855, "learning_rate": 7.731756854780212e-05, "loss": 0.0111, "step": 6354 }, { "epoch": 0.6889635732870771, "grad_norm": 0.2400522530078888, "learning_rate": 7.731394167996519e-05, "loss": 0.0397, "step": 6355 }, { "epoch": 0.689071986123157, "grad_norm": 0.10185416787862778, "learning_rate": 7.731031481212825e-05, "loss": 0.0164, "step": 6356 }, { "epoch": 0.6891803989592368, "grad_norm": 0.20003938674926758, "learning_rate": 7.730668794429132e-05, "loss": 0.036, "step": 6357 }, { "epoch": 0.6892888117953165, "grad_norm": 0.21694894134998322, "learning_rate": 7.730306107645437e-05, "loss": 0.0277, "step": 6358 }, { "epoch": 0.6893972246313964, "grad_norm": 0.1896558701992035, "learning_rate": 7.729943420861743e-05, "loss": 0.0177, "step": 6359 }, { "epoch": 0.6895056374674762, "grad_norm": 0.38637033104896545, "learning_rate": 7.72958073407805e-05, "loss": 0.0571, "step": 6360 }, { "epoch": 0.6896140503035559, "grad_norm": 0.19790613651275635, "learning_rate": 7.729218047294356e-05, "loss": 0.0422, "step": 6361 }, { "epoch": 0.6897224631396357, "grad_norm": 0.36668238043785095, "learning_rate": 7.728855360510663e-05, "loss": 0.0391, "step": 6362 }, { "epoch": 0.6898308759757156, "grad_norm": 0.13148915767669678, "learning_rate": 7.72849267372697e-05, "loss": 0.0127, "step": 6363 }, { "epoch": 0.6899392888117953, "grad_norm": 0.21324869990348816, "learning_rate": 7.728129986943276e-05, "loss": 0.0305, "step": 6364 }, { "epoch": 0.6900477016478751, "grad_norm": 0.28756213188171387, "learning_rate": 7.727767300159583e-05, "loss": 0.0258, "step": 6365 }, { "epoch": 0.6901561144839549, "grad_norm": 0.3251717984676361, "learning_rate": 7.727404613375889e-05, "loss": 0.0387, "step": 6366 }, { "epoch": 0.6902645273200347, "grad_norm": 0.3410429060459137, "learning_rate": 7.727041926592196e-05, "loss": 0.0722, "step": 6367 }, { "epoch": 0.6903729401561145, "grad_norm": 0.23998068273067474, "learning_rate": 7.726679239808502e-05, "loss": 0.0566, "step": 6368 }, { "epoch": 0.6904813529921943, "grad_norm": 0.2629043459892273, "learning_rate": 7.726316553024809e-05, "loss": 0.0301, "step": 6369 }, { "epoch": 0.690589765828274, "grad_norm": 0.3213164806365967, "learning_rate": 7.725953866241114e-05, "loss": 0.0485, "step": 6370 }, { "epoch": 0.6906981786643539, "grad_norm": 0.3718082308769226, "learning_rate": 7.725591179457421e-05, "loss": 0.0502, "step": 6371 }, { "epoch": 0.6908065915004337, "grad_norm": 0.33025461435317993, "learning_rate": 7.725228492673727e-05, "loss": 0.0414, "step": 6372 }, { "epoch": 0.6909150043365134, "grad_norm": 0.22583192586898804, "learning_rate": 7.724865805890033e-05, "loss": 0.0346, "step": 6373 }, { "epoch": 0.6910234171725932, "grad_norm": 0.19314555823802948, "learning_rate": 7.72450311910634e-05, "loss": 0.0662, "step": 6374 }, { "epoch": 0.6911318300086731, "grad_norm": 0.18399037420749664, "learning_rate": 7.724140432322646e-05, "loss": 0.0397, "step": 6375 }, { "epoch": 0.6912402428447528, "grad_norm": 0.6306026577949524, "learning_rate": 7.723777745538953e-05, "loss": 0.0452, "step": 6376 }, { "epoch": 0.6913486556808326, "grad_norm": 0.2805262804031372, "learning_rate": 7.723415058755259e-05, "loss": 0.0559, "step": 6377 }, { "epoch": 0.6914570685169124, "grad_norm": 0.5596450567245483, "learning_rate": 7.723052371971566e-05, "loss": 0.0459, "step": 6378 }, { "epoch": 0.6915654813529922, "grad_norm": 0.3369060158729553, "learning_rate": 7.722689685187871e-05, "loss": 0.0443, "step": 6379 }, { "epoch": 0.691673894189072, "grad_norm": 0.4420894682407379, "learning_rate": 7.722326998404179e-05, "loss": 0.0333, "step": 6380 }, { "epoch": 0.6917823070251518, "grad_norm": 0.2617650330066681, "learning_rate": 7.721964311620486e-05, "loss": 0.0202, "step": 6381 }, { "epoch": 0.6918907198612315, "grad_norm": 0.3561922311782837, "learning_rate": 7.721601624836791e-05, "loss": 0.0525, "step": 6382 }, { "epoch": 0.6919991326973114, "grad_norm": 0.18186001479625702, "learning_rate": 7.721238938053098e-05, "loss": 0.0116, "step": 6383 }, { "epoch": 0.6921075455333912, "grad_norm": 0.5198861360549927, "learning_rate": 7.720876251269404e-05, "loss": 0.1197, "step": 6384 }, { "epoch": 0.6922159583694709, "grad_norm": 0.05749013274908066, "learning_rate": 7.720513564485711e-05, "loss": 0.009, "step": 6385 }, { "epoch": 0.6923243712055508, "grad_norm": 0.32585379481315613, "learning_rate": 7.720150877702017e-05, "loss": 0.0604, "step": 6386 }, { "epoch": 0.6924327840416306, "grad_norm": 0.30175864696502686, "learning_rate": 7.719788190918324e-05, "loss": 0.0248, "step": 6387 }, { "epoch": 0.6925411968777103, "grad_norm": 0.30488264560699463, "learning_rate": 7.71942550413463e-05, "loss": 0.0441, "step": 6388 }, { "epoch": 0.6926496097137901, "grad_norm": 0.5642341375350952, "learning_rate": 7.719062817350936e-05, "loss": 0.0406, "step": 6389 }, { "epoch": 0.69275802254987, "grad_norm": 0.2683134973049164, "learning_rate": 7.718700130567243e-05, "loss": 0.0398, "step": 6390 }, { "epoch": 0.6928664353859497, "grad_norm": 0.5511724948883057, "learning_rate": 7.718337443783548e-05, "loss": 0.0594, "step": 6391 }, { "epoch": 0.6929748482220295, "grad_norm": 0.1453186720609665, "learning_rate": 7.717974756999855e-05, "loss": 0.0195, "step": 6392 }, { "epoch": 0.6930832610581092, "grad_norm": 0.31547272205352783, "learning_rate": 7.717612070216161e-05, "loss": 0.032, "step": 6393 }, { "epoch": 0.6931916738941891, "grad_norm": 0.4990149140357971, "learning_rate": 7.717249383432468e-05, "loss": 0.0362, "step": 6394 }, { "epoch": 0.6933000867302689, "grad_norm": 0.21579930186271667, "learning_rate": 7.716886696648774e-05, "loss": 0.0329, "step": 6395 }, { "epoch": 0.6934084995663486, "grad_norm": 0.08792085945606232, "learning_rate": 7.71652400986508e-05, "loss": 0.0068, "step": 6396 }, { "epoch": 0.6935169124024284, "grad_norm": 0.2454913705587387, "learning_rate": 7.716161323081388e-05, "loss": 0.0485, "step": 6397 }, { "epoch": 0.6936253252385083, "grad_norm": 0.3633298873901367, "learning_rate": 7.715798636297694e-05, "loss": 0.0272, "step": 6398 }, { "epoch": 0.693733738074588, "grad_norm": 0.345995157957077, "learning_rate": 7.715435949514001e-05, "loss": 0.027, "step": 6399 }, { "epoch": 0.6938421509106678, "grad_norm": 0.16426798701286316, "learning_rate": 7.715073262730307e-05, "loss": 0.0123, "step": 6400 }, { "epoch": 0.6939505637467476, "grad_norm": 0.24183864891529083, "learning_rate": 7.714710575946614e-05, "loss": 0.0532, "step": 6401 }, { "epoch": 0.6940589765828274, "grad_norm": 0.8331490755081177, "learning_rate": 7.71434788916292e-05, "loss": 0.0447, "step": 6402 }, { "epoch": 0.6941673894189072, "grad_norm": 0.2146994173526764, "learning_rate": 7.713985202379225e-05, "loss": 0.0103, "step": 6403 }, { "epoch": 0.694275802254987, "grad_norm": 0.6383541822433472, "learning_rate": 7.713622515595532e-05, "loss": 0.062, "step": 6404 }, { "epoch": 0.6943842150910667, "grad_norm": 0.33578988909721375, "learning_rate": 7.713259828811838e-05, "loss": 0.0392, "step": 6405 }, { "epoch": 0.6944926279271466, "grad_norm": 0.2052464783191681, "learning_rate": 7.712897142028145e-05, "loss": 0.031, "step": 6406 }, { "epoch": 0.6946010407632264, "grad_norm": 0.12096310406923294, "learning_rate": 7.712534455244451e-05, "loss": 0.0071, "step": 6407 }, { "epoch": 0.6947094535993061, "grad_norm": 0.31153708696365356, "learning_rate": 7.712171768460758e-05, "loss": 0.0239, "step": 6408 }, { "epoch": 0.6948178664353859, "grad_norm": 0.2211175262928009, "learning_rate": 7.711809081677064e-05, "loss": 0.0215, "step": 6409 }, { "epoch": 0.6949262792714658, "grad_norm": 0.874873697757721, "learning_rate": 7.711446394893371e-05, "loss": 0.0503, "step": 6410 }, { "epoch": 0.6950346921075455, "grad_norm": 0.4813063144683838, "learning_rate": 7.711083708109677e-05, "loss": 0.0444, "step": 6411 }, { "epoch": 0.6951431049436253, "grad_norm": 0.6703543663024902, "learning_rate": 7.710721021325982e-05, "loss": 0.0479, "step": 6412 }, { "epoch": 0.6952515177797051, "grad_norm": 0.4951585829257965, "learning_rate": 7.71035833454229e-05, "loss": 0.0615, "step": 6413 }, { "epoch": 0.6953599306157849, "grad_norm": 0.29901087284088135, "learning_rate": 7.709995647758595e-05, "loss": 0.0282, "step": 6414 }, { "epoch": 0.6954683434518647, "grad_norm": 0.4015164077281952, "learning_rate": 7.709632960974904e-05, "loss": 0.0429, "step": 6415 }, { "epoch": 0.6955767562879445, "grad_norm": 0.5331568121910095, "learning_rate": 7.709270274191209e-05, "loss": 0.0478, "step": 6416 }, { "epoch": 0.6956851691240243, "grad_norm": 0.5329536199569702, "learning_rate": 7.708907587407515e-05, "loss": 0.0789, "step": 6417 }, { "epoch": 0.6957935819601041, "grad_norm": 0.2870362401008606, "learning_rate": 7.708544900623822e-05, "loss": 0.0593, "step": 6418 }, { "epoch": 0.6959019947961839, "grad_norm": 0.19392268359661102, "learning_rate": 7.708182213840128e-05, "loss": 0.0188, "step": 6419 }, { "epoch": 0.6960104076322636, "grad_norm": 0.8279494047164917, "learning_rate": 7.707819527056435e-05, "loss": 0.0621, "step": 6420 }, { "epoch": 0.6961188204683435, "grad_norm": 0.33795639872550964, "learning_rate": 7.70745684027274e-05, "loss": 0.023, "step": 6421 }, { "epoch": 0.6962272333044233, "grad_norm": 0.1938067525625229, "learning_rate": 7.707094153489048e-05, "loss": 0.0142, "step": 6422 }, { "epoch": 0.696335646140503, "grad_norm": 0.3391783535480499, "learning_rate": 7.706731466705353e-05, "loss": 0.0296, "step": 6423 }, { "epoch": 0.6964440589765828, "grad_norm": 0.42494305968284607, "learning_rate": 7.70636877992166e-05, "loss": 0.0516, "step": 6424 }, { "epoch": 0.6965524718126627, "grad_norm": 0.3655935823917389, "learning_rate": 7.706006093137966e-05, "loss": 0.0375, "step": 6425 }, { "epoch": 0.6966608846487424, "grad_norm": 0.41145431995391846, "learning_rate": 7.705643406354272e-05, "loss": 0.067, "step": 6426 }, { "epoch": 0.6967692974848222, "grad_norm": 0.4282393455505371, "learning_rate": 7.705280719570579e-05, "loss": 0.0245, "step": 6427 }, { "epoch": 0.696877710320902, "grad_norm": 0.3898884654045105, "learning_rate": 7.704918032786885e-05, "loss": 0.0414, "step": 6428 }, { "epoch": 0.6969861231569818, "grad_norm": 0.24328960478305817, "learning_rate": 7.704555346003192e-05, "loss": 0.0196, "step": 6429 }, { "epoch": 0.6970945359930616, "grad_norm": 0.35939499735832214, "learning_rate": 7.704192659219498e-05, "loss": 0.0396, "step": 6430 }, { "epoch": 0.6972029488291414, "grad_norm": 0.293184757232666, "learning_rate": 7.703829972435805e-05, "loss": 0.0281, "step": 6431 }, { "epoch": 0.6973113616652211, "grad_norm": 0.1173609048128128, "learning_rate": 7.703467285652112e-05, "loss": 0.0105, "step": 6432 }, { "epoch": 0.697419774501301, "grad_norm": 0.5023528337478638, "learning_rate": 7.703104598868418e-05, "loss": 0.0406, "step": 6433 }, { "epoch": 0.6975281873373808, "grad_norm": 0.19314217567443848, "learning_rate": 7.702741912084725e-05, "loss": 0.0295, "step": 6434 }, { "epoch": 0.6976366001734605, "grad_norm": 0.4779038429260254, "learning_rate": 7.70237922530103e-05, "loss": 0.0558, "step": 6435 }, { "epoch": 0.6977450130095403, "grad_norm": 0.29308924078941345, "learning_rate": 7.702016538517337e-05, "loss": 0.0392, "step": 6436 }, { "epoch": 0.6978534258456202, "grad_norm": 0.43514931201934814, "learning_rate": 7.701653851733643e-05, "loss": 0.0603, "step": 6437 }, { "epoch": 0.6979618386816999, "grad_norm": 0.12270782887935638, "learning_rate": 7.70129116494995e-05, "loss": 0.0175, "step": 6438 }, { "epoch": 0.6980702515177797, "grad_norm": 0.5131700038909912, "learning_rate": 7.700928478166256e-05, "loss": 0.0404, "step": 6439 }, { "epoch": 0.6981786643538594, "grad_norm": 0.6177954077720642, "learning_rate": 7.700565791382562e-05, "loss": 0.0241, "step": 6440 }, { "epoch": 0.6982870771899393, "grad_norm": 0.27098625898361206, "learning_rate": 7.700203104598869e-05, "loss": 0.0221, "step": 6441 }, { "epoch": 0.6983954900260191, "grad_norm": 0.4351787269115448, "learning_rate": 7.699840417815175e-05, "loss": 0.0694, "step": 6442 }, { "epoch": 0.6985039028620988, "grad_norm": 0.538632869720459, "learning_rate": 7.699477731031482e-05, "loss": 0.0361, "step": 6443 }, { "epoch": 0.6986123156981786, "grad_norm": 0.1963910013437271, "learning_rate": 7.699115044247787e-05, "loss": 0.0245, "step": 6444 }, { "epoch": 0.6987207285342585, "grad_norm": 0.25140970945358276, "learning_rate": 7.698752357464095e-05, "loss": 0.0237, "step": 6445 }, { "epoch": 0.6988291413703382, "grad_norm": 0.13851910829544067, "learning_rate": 7.6983896706804e-05, "loss": 0.0272, "step": 6446 }, { "epoch": 0.698937554206418, "grad_norm": 0.16751106083393097, "learning_rate": 7.698026983896707e-05, "loss": 0.0181, "step": 6447 }, { "epoch": 0.6990459670424979, "grad_norm": 0.3078785836696625, "learning_rate": 7.697664297113013e-05, "loss": 0.0661, "step": 6448 }, { "epoch": 0.6991543798785776, "grad_norm": 0.3833017647266388, "learning_rate": 7.69730161032932e-05, "loss": 0.0587, "step": 6449 }, { "epoch": 0.6992627927146574, "grad_norm": 0.675800085067749, "learning_rate": 7.696938923545627e-05, "loss": 0.0276, "step": 6450 }, { "epoch": 0.6993712055507372, "grad_norm": 0.7095272541046143, "learning_rate": 7.696576236761933e-05, "loss": 0.0381, "step": 6451 }, { "epoch": 0.699479618386817, "grad_norm": 0.3280159533023834, "learning_rate": 7.69621354997824e-05, "loss": 0.0244, "step": 6452 }, { "epoch": 0.6995880312228968, "grad_norm": 0.45477932691574097, "learning_rate": 7.695850863194546e-05, "loss": 0.0431, "step": 6453 }, { "epoch": 0.6996964440589766, "grad_norm": 0.262395977973938, "learning_rate": 7.695488176410852e-05, "loss": 0.0188, "step": 6454 }, { "epoch": 0.6998048568950563, "grad_norm": 0.3662095069885254, "learning_rate": 7.695125489627159e-05, "loss": 0.0273, "step": 6455 }, { "epoch": 0.6999132697311362, "grad_norm": 0.8513947129249573, "learning_rate": 7.694762802843464e-05, "loss": 0.0367, "step": 6456 }, { "epoch": 0.700021682567216, "grad_norm": 0.4119862914085388, "learning_rate": 7.694400116059771e-05, "loss": 0.0453, "step": 6457 }, { "epoch": 0.7001300954032957, "grad_norm": 0.1282110959291458, "learning_rate": 7.694037429276077e-05, "loss": 0.0162, "step": 6458 }, { "epoch": 0.7002385082393755, "grad_norm": 0.5679630041122437, "learning_rate": 7.693674742492384e-05, "loss": 0.0687, "step": 6459 }, { "epoch": 0.7003469210754554, "grad_norm": 0.07421519607305527, "learning_rate": 7.69331205570869e-05, "loss": 0.01, "step": 6460 }, { "epoch": 0.7004553339115351, "grad_norm": 0.6338358521461487, "learning_rate": 7.692949368924997e-05, "loss": 0.0561, "step": 6461 }, { "epoch": 0.7005637467476149, "grad_norm": 0.27178633213043213, "learning_rate": 7.692586682141303e-05, "loss": 0.026, "step": 6462 }, { "epoch": 0.7006721595836947, "grad_norm": 0.2934666574001312, "learning_rate": 7.692223995357609e-05, "loss": 0.0413, "step": 6463 }, { "epoch": 0.7007805724197745, "grad_norm": 0.42692282795906067, "learning_rate": 7.691861308573916e-05, "loss": 0.0301, "step": 6464 }, { "epoch": 0.7008889852558543, "grad_norm": 0.5575594902038574, "learning_rate": 7.691498621790221e-05, "loss": 0.0643, "step": 6465 }, { "epoch": 0.7009973980919341, "grad_norm": 0.2986004054546356, "learning_rate": 7.691135935006528e-05, "loss": 0.0212, "step": 6466 }, { "epoch": 0.7011058109280138, "grad_norm": 0.8071054816246033, "learning_rate": 7.690773248222836e-05, "loss": 0.0351, "step": 6467 }, { "epoch": 0.7012142237640937, "grad_norm": 0.3651007413864136, "learning_rate": 7.690410561439143e-05, "loss": 0.0782, "step": 6468 }, { "epoch": 0.7013226366001735, "grad_norm": 0.11410496383905411, "learning_rate": 7.690047874655448e-05, "loss": 0.0108, "step": 6469 }, { "epoch": 0.7014310494362532, "grad_norm": 0.8038579821586609, "learning_rate": 7.689685187871754e-05, "loss": 0.0635, "step": 6470 }, { "epoch": 0.701539462272333, "grad_norm": 0.47272858023643494, "learning_rate": 7.689322501088061e-05, "loss": 0.0482, "step": 6471 }, { "epoch": 0.7016478751084129, "grad_norm": 0.28994694352149963, "learning_rate": 7.688959814304367e-05, "loss": 0.033, "step": 6472 }, { "epoch": 0.7017562879444926, "grad_norm": 0.6581401824951172, "learning_rate": 7.688597127520674e-05, "loss": 0.0509, "step": 6473 }, { "epoch": 0.7018647007805724, "grad_norm": 0.36661389470100403, "learning_rate": 7.68823444073698e-05, "loss": 0.0312, "step": 6474 }, { "epoch": 0.7019731136166523, "grad_norm": 0.3217940628528595, "learning_rate": 7.687871753953287e-05, "loss": 0.0448, "step": 6475 }, { "epoch": 0.702081526452732, "grad_norm": 0.2153676599264145, "learning_rate": 7.687509067169593e-05, "loss": 0.028, "step": 6476 }, { "epoch": 0.7021899392888118, "grad_norm": 0.17151719331741333, "learning_rate": 7.687146380385898e-05, "loss": 0.0398, "step": 6477 }, { "epoch": 0.7022983521248916, "grad_norm": 0.5361326336860657, "learning_rate": 7.686783693602205e-05, "loss": 0.0398, "step": 6478 }, { "epoch": 0.7024067649609714, "grad_norm": 0.43952929973602295, "learning_rate": 7.686421006818511e-05, "loss": 0.047, "step": 6479 }, { "epoch": 0.7025151777970512, "grad_norm": 0.20167991518974304, "learning_rate": 7.686058320034818e-05, "loss": 0.0224, "step": 6480 }, { "epoch": 0.702623590633131, "grad_norm": 0.29683470726013184, "learning_rate": 7.685695633251124e-05, "loss": 0.0495, "step": 6481 }, { "epoch": 0.7027320034692107, "grad_norm": 0.5101122260093689, "learning_rate": 7.685332946467431e-05, "loss": 0.027, "step": 6482 }, { "epoch": 0.7028404163052906, "grad_norm": 0.2950226962566376, "learning_rate": 7.684970259683737e-05, "loss": 0.0269, "step": 6483 }, { "epoch": 0.7029488291413704, "grad_norm": 0.3157431483268738, "learning_rate": 7.684607572900044e-05, "loss": 0.0334, "step": 6484 }, { "epoch": 0.7030572419774501, "grad_norm": 0.5886742472648621, "learning_rate": 7.684244886116351e-05, "loss": 0.0777, "step": 6485 }, { "epoch": 0.7031656548135299, "grad_norm": 0.2879157066345215, "learning_rate": 7.683882199332657e-05, "loss": 0.0319, "step": 6486 }, { "epoch": 0.7032740676496098, "grad_norm": 0.18294622004032135, "learning_rate": 7.683519512548964e-05, "loss": 0.0244, "step": 6487 }, { "epoch": 0.7033824804856895, "grad_norm": 0.43138179183006287, "learning_rate": 7.68315682576527e-05, "loss": 0.0295, "step": 6488 }, { "epoch": 0.7034908933217693, "grad_norm": 0.41934719681739807, "learning_rate": 7.682794138981577e-05, "loss": 0.0463, "step": 6489 }, { "epoch": 0.703599306157849, "grad_norm": 0.3656066358089447, "learning_rate": 7.682431452197882e-05, "loss": 0.0543, "step": 6490 }, { "epoch": 0.7037077189939289, "grad_norm": 0.24471358954906464, "learning_rate": 7.68206876541419e-05, "loss": 0.0555, "step": 6491 }, { "epoch": 0.7038161318300087, "grad_norm": 0.170718714594841, "learning_rate": 7.681706078630495e-05, "loss": 0.0181, "step": 6492 }, { "epoch": 0.7039245446660884, "grad_norm": 0.21038441359996796, "learning_rate": 7.681343391846801e-05, "loss": 0.0243, "step": 6493 }, { "epoch": 0.7040329575021682, "grad_norm": 0.1536237895488739, "learning_rate": 7.680980705063108e-05, "loss": 0.0225, "step": 6494 }, { "epoch": 0.7041413703382481, "grad_norm": 0.21454928815364838, "learning_rate": 7.680618018279414e-05, "loss": 0.0385, "step": 6495 }, { "epoch": 0.7042497831743278, "grad_norm": 0.4409516155719757, "learning_rate": 7.680255331495721e-05, "loss": 0.0352, "step": 6496 }, { "epoch": 0.7043581960104076, "grad_norm": 0.578396201133728, "learning_rate": 7.679892644712027e-05, "loss": 0.0947, "step": 6497 }, { "epoch": 0.7044666088464874, "grad_norm": 0.3528849482536316, "learning_rate": 7.679529957928334e-05, "loss": 0.0363, "step": 6498 }, { "epoch": 0.7045750216825672, "grad_norm": 0.4632219076156616, "learning_rate": 7.67916727114464e-05, "loss": 0.056, "step": 6499 }, { "epoch": 0.704683434518647, "grad_norm": 0.33464673161506653, "learning_rate": 7.678804584360946e-05, "loss": 0.0439, "step": 6500 }, { "epoch": 0.7047918473547268, "grad_norm": 0.15876615047454834, "learning_rate": 7.678441897577254e-05, "loss": 0.0228, "step": 6501 }, { "epoch": 0.7049002601908065, "grad_norm": 0.16979925334453583, "learning_rate": 7.678079210793559e-05, "loss": 0.0211, "step": 6502 }, { "epoch": 0.7050086730268864, "grad_norm": 0.16155719757080078, "learning_rate": 7.677716524009866e-05, "loss": 0.0266, "step": 6503 }, { "epoch": 0.7051170858629662, "grad_norm": 0.34327831864356995, "learning_rate": 7.677353837226172e-05, "loss": 0.0206, "step": 6504 }, { "epoch": 0.7052254986990459, "grad_norm": 0.359182208776474, "learning_rate": 7.676991150442479e-05, "loss": 0.0422, "step": 6505 }, { "epoch": 0.7053339115351258, "grad_norm": 0.35819146037101746, "learning_rate": 7.676628463658785e-05, "loss": 0.0346, "step": 6506 }, { "epoch": 0.7054423243712056, "grad_norm": 0.33883556723594666, "learning_rate": 7.67626577687509e-05, "loss": 0.0498, "step": 6507 }, { "epoch": 0.7055507372072853, "grad_norm": 0.25691521167755127, "learning_rate": 7.675903090091398e-05, "loss": 0.0316, "step": 6508 }, { "epoch": 0.7056591500433651, "grad_norm": 0.3748784363269806, "learning_rate": 7.675540403307703e-05, "loss": 0.0579, "step": 6509 }, { "epoch": 0.705767562879445, "grad_norm": 0.3584999442100525, "learning_rate": 7.67517771652401e-05, "loss": 0.091, "step": 6510 }, { "epoch": 0.7058759757155247, "grad_norm": 0.16959814727306366, "learning_rate": 7.674815029740316e-05, "loss": 0.0358, "step": 6511 }, { "epoch": 0.7059843885516045, "grad_norm": 0.4688699543476105, "learning_rate": 7.674452342956623e-05, "loss": 0.0312, "step": 6512 }, { "epoch": 0.7060928013876843, "grad_norm": 0.27517950534820557, "learning_rate": 7.674089656172929e-05, "loss": 0.0302, "step": 6513 }, { "epoch": 0.7062012142237641, "grad_norm": 0.33603882789611816, "learning_rate": 7.673726969389236e-05, "loss": 0.067, "step": 6514 }, { "epoch": 0.7063096270598439, "grad_norm": 0.32770460844039917, "learning_rate": 7.673364282605542e-05, "loss": 0.0472, "step": 6515 }, { "epoch": 0.7064180398959237, "grad_norm": 0.5572274327278137, "learning_rate": 7.673001595821848e-05, "loss": 0.0531, "step": 6516 }, { "epoch": 0.7065264527320034, "grad_norm": 0.2696411907672882, "learning_rate": 7.672638909038155e-05, "loss": 0.0411, "step": 6517 }, { "epoch": 0.7066348655680833, "grad_norm": 0.19971761107444763, "learning_rate": 7.672276222254462e-05, "loss": 0.0153, "step": 6518 }, { "epoch": 0.7067432784041631, "grad_norm": 0.3231155574321747, "learning_rate": 7.671913535470769e-05, "loss": 0.0351, "step": 6519 }, { "epoch": 0.7068516912402428, "grad_norm": 0.25341328978538513, "learning_rate": 7.671550848687075e-05, "loss": 0.0438, "step": 6520 }, { "epoch": 0.7069601040763226, "grad_norm": 0.3250117003917694, "learning_rate": 7.67118816190338e-05, "loss": 0.0336, "step": 6521 }, { "epoch": 0.7070685169124025, "grad_norm": 0.171620711684227, "learning_rate": 7.670825475119687e-05, "loss": 0.0257, "step": 6522 }, { "epoch": 0.7071769297484822, "grad_norm": 0.5329856872558594, "learning_rate": 7.670462788335993e-05, "loss": 0.0502, "step": 6523 }, { "epoch": 0.707285342584562, "grad_norm": 0.249166801571846, "learning_rate": 7.6701001015523e-05, "loss": 0.0516, "step": 6524 }, { "epoch": 0.7073937554206418, "grad_norm": 0.42480868101119995, "learning_rate": 7.669737414768606e-05, "loss": 0.0671, "step": 6525 }, { "epoch": 0.7075021682567216, "grad_norm": 0.18707023561000824, "learning_rate": 7.669374727984913e-05, "loss": 0.0212, "step": 6526 }, { "epoch": 0.7076105810928014, "grad_norm": 0.43873655796051025, "learning_rate": 7.669012041201219e-05, "loss": 0.0645, "step": 6527 }, { "epoch": 0.7077189939288812, "grad_norm": 0.3310990035533905, "learning_rate": 7.668649354417526e-05, "loss": 0.0325, "step": 6528 }, { "epoch": 0.7078274067649609, "grad_norm": 0.16654396057128906, "learning_rate": 7.668286667633832e-05, "loss": 0.0171, "step": 6529 }, { "epoch": 0.7079358196010408, "grad_norm": 0.10485703498125076, "learning_rate": 7.667923980850137e-05, "loss": 0.0155, "step": 6530 }, { "epoch": 0.7080442324371206, "grad_norm": 0.37123945355415344, "learning_rate": 7.667561294066444e-05, "loss": 0.0418, "step": 6531 }, { "epoch": 0.7081526452732003, "grad_norm": 0.3514217734336853, "learning_rate": 7.66719860728275e-05, "loss": 0.0419, "step": 6532 }, { "epoch": 0.7082610581092801, "grad_norm": 0.36130523681640625, "learning_rate": 7.666835920499057e-05, "loss": 0.0407, "step": 6533 }, { "epoch": 0.70836947094536, "grad_norm": 0.17812862992286682, "learning_rate": 7.666473233715363e-05, "loss": 0.0236, "step": 6534 }, { "epoch": 0.7084778837814397, "grad_norm": 0.31752780079841614, "learning_rate": 7.66611054693167e-05, "loss": 0.0466, "step": 6535 }, { "epoch": 0.7085862966175195, "grad_norm": 0.3363150954246521, "learning_rate": 7.665747860147977e-05, "loss": 0.0326, "step": 6536 }, { "epoch": 0.7086947094535994, "grad_norm": 0.40006938576698303, "learning_rate": 7.665385173364283e-05, "loss": 0.0181, "step": 6537 }, { "epoch": 0.7088031222896791, "grad_norm": 0.38127970695495605, "learning_rate": 7.66502248658059e-05, "loss": 0.0369, "step": 6538 }, { "epoch": 0.7089115351257589, "grad_norm": 0.21125145256519318, "learning_rate": 7.664659799796896e-05, "loss": 0.0238, "step": 6539 }, { "epoch": 0.7090199479618386, "grad_norm": 0.4114402234554291, "learning_rate": 7.664297113013203e-05, "loss": 0.0455, "step": 6540 }, { "epoch": 0.7091283607979185, "grad_norm": 0.22913460433483124, "learning_rate": 7.663934426229509e-05, "loss": 0.0146, "step": 6541 }, { "epoch": 0.7092367736339983, "grad_norm": 0.4918469488620758, "learning_rate": 7.663571739445816e-05, "loss": 0.1003, "step": 6542 }, { "epoch": 0.709345186470078, "grad_norm": 0.323658287525177, "learning_rate": 7.663209052662121e-05, "loss": 0.0374, "step": 6543 }, { "epoch": 0.7094535993061578, "grad_norm": 0.21377594769001007, "learning_rate": 7.662846365878427e-05, "loss": 0.0319, "step": 6544 }, { "epoch": 0.7095620121422377, "grad_norm": 0.3625768721103668, "learning_rate": 7.662483679094734e-05, "loss": 0.0849, "step": 6545 }, { "epoch": 0.7096704249783174, "grad_norm": 0.10902529209852219, "learning_rate": 7.66212099231104e-05, "loss": 0.0091, "step": 6546 }, { "epoch": 0.7097788378143972, "grad_norm": 0.05345830321311951, "learning_rate": 7.661758305527347e-05, "loss": 0.0037, "step": 6547 }, { "epoch": 0.709887250650477, "grad_norm": 0.19306178390979767, "learning_rate": 7.661395618743653e-05, "loss": 0.0263, "step": 6548 }, { "epoch": 0.7099956634865568, "grad_norm": 0.279623419046402, "learning_rate": 7.66103293195996e-05, "loss": 0.0244, "step": 6549 }, { "epoch": 0.7101040763226366, "grad_norm": 0.32590949535369873, "learning_rate": 7.660670245176266e-05, "loss": 0.054, "step": 6550 }, { "epoch": 0.7102124891587164, "grad_norm": 0.40914997458457947, "learning_rate": 7.660307558392573e-05, "loss": 0.0725, "step": 6551 }, { "epoch": 0.7103209019947961, "grad_norm": 0.2333599179983139, "learning_rate": 7.659944871608878e-05, "loss": 0.0307, "step": 6552 }, { "epoch": 0.710429314830876, "grad_norm": 0.22107473015785217, "learning_rate": 7.659582184825185e-05, "loss": 0.0201, "step": 6553 }, { "epoch": 0.7105377276669558, "grad_norm": 0.3446052074432373, "learning_rate": 7.659219498041493e-05, "loss": 0.0427, "step": 6554 }, { "epoch": 0.7106461405030355, "grad_norm": 0.1606953889131546, "learning_rate": 7.658856811257798e-05, "loss": 0.017, "step": 6555 }, { "epoch": 0.7107545533391153, "grad_norm": 0.36957409977912903, "learning_rate": 7.658494124474105e-05, "loss": 0.0182, "step": 6556 }, { "epoch": 0.7108629661751952, "grad_norm": 0.5438377857208252, "learning_rate": 7.658131437690411e-05, "loss": 0.0977, "step": 6557 }, { "epoch": 0.7109713790112749, "grad_norm": 0.40677309036254883, "learning_rate": 7.657768750906717e-05, "loss": 0.0336, "step": 6558 }, { "epoch": 0.7110797918473547, "grad_norm": 0.23769541084766388, "learning_rate": 7.657406064123024e-05, "loss": 0.0273, "step": 6559 }, { "epoch": 0.7111882046834345, "grad_norm": 0.2462046891450882, "learning_rate": 7.65704337733933e-05, "loss": 0.0235, "step": 6560 }, { "epoch": 0.7112966175195143, "grad_norm": 0.3425028622150421, "learning_rate": 7.656680690555637e-05, "loss": 0.0337, "step": 6561 }, { "epoch": 0.7114050303555941, "grad_norm": 0.18169376254081726, "learning_rate": 7.656318003771943e-05, "loss": 0.0223, "step": 6562 }, { "epoch": 0.7115134431916739, "grad_norm": 0.7630958557128906, "learning_rate": 7.65595531698825e-05, "loss": 0.0587, "step": 6563 }, { "epoch": 0.7116218560277537, "grad_norm": 0.9903728365898132, "learning_rate": 7.655592630204555e-05, "loss": 0.0264, "step": 6564 }, { "epoch": 0.7117302688638335, "grad_norm": 0.5681807398796082, "learning_rate": 7.655229943420862e-05, "loss": 0.0464, "step": 6565 }, { "epoch": 0.7118386816999133, "grad_norm": 0.492201566696167, "learning_rate": 7.654867256637168e-05, "loss": 0.0453, "step": 6566 }, { "epoch": 0.711947094535993, "grad_norm": 0.24984656274318695, "learning_rate": 7.654504569853474e-05, "loss": 0.0248, "step": 6567 }, { "epoch": 0.7120555073720729, "grad_norm": 0.2549762725830078, "learning_rate": 7.654141883069781e-05, "loss": 0.0353, "step": 6568 }, { "epoch": 0.7121639202081527, "grad_norm": 0.39272668957710266, "learning_rate": 7.653779196286087e-05, "loss": 0.0281, "step": 6569 }, { "epoch": 0.7122723330442324, "grad_norm": 0.41876715421676636, "learning_rate": 7.653416509502395e-05, "loss": 0.0236, "step": 6570 }, { "epoch": 0.7123807458803122, "grad_norm": 0.20516309142112732, "learning_rate": 7.653053822718701e-05, "loss": 0.0342, "step": 6571 }, { "epoch": 0.7124891587163921, "grad_norm": 0.21069307625293732, "learning_rate": 7.652691135935008e-05, "loss": 0.0229, "step": 6572 }, { "epoch": 0.7125975715524718, "grad_norm": 0.3640235662460327, "learning_rate": 7.652328449151314e-05, "loss": 0.0575, "step": 6573 }, { "epoch": 0.7127059843885516, "grad_norm": 0.5615851283073425, "learning_rate": 7.65196576236762e-05, "loss": 0.0663, "step": 6574 }, { "epoch": 0.7128143972246314, "grad_norm": 0.96229088306427, "learning_rate": 7.651603075583927e-05, "loss": 0.0477, "step": 6575 }, { "epoch": 0.7129228100607112, "grad_norm": 0.5237632989883423, "learning_rate": 7.651240388800232e-05, "loss": 0.0507, "step": 6576 }, { "epoch": 0.713031222896791, "grad_norm": 0.3959328234195709, "learning_rate": 7.65087770201654e-05, "loss": 0.0508, "step": 6577 }, { "epoch": 0.7131396357328708, "grad_norm": 0.4096010625362396, "learning_rate": 7.650515015232845e-05, "loss": 0.0709, "step": 6578 }, { "epoch": 0.7132480485689505, "grad_norm": 0.2254646271467209, "learning_rate": 7.650152328449152e-05, "loss": 0.0219, "step": 6579 }, { "epoch": 0.7133564614050304, "grad_norm": 0.5574548244476318, "learning_rate": 7.649789641665458e-05, "loss": 0.039, "step": 6580 }, { "epoch": 0.7134648742411102, "grad_norm": 0.30713626742362976, "learning_rate": 7.649426954881764e-05, "loss": 0.0385, "step": 6581 }, { "epoch": 0.7135732870771899, "grad_norm": 0.5891329646110535, "learning_rate": 7.649064268098071e-05, "loss": 0.1002, "step": 6582 }, { "epoch": 0.7136816999132697, "grad_norm": 0.11189734190702438, "learning_rate": 7.648701581314376e-05, "loss": 0.0193, "step": 6583 }, { "epoch": 0.7137901127493496, "grad_norm": 0.26728129386901855, "learning_rate": 7.648338894530684e-05, "loss": 0.0222, "step": 6584 }, { "epoch": 0.7138985255854293, "grad_norm": 0.4014642536640167, "learning_rate": 7.647976207746989e-05, "loss": 0.0315, "step": 6585 }, { "epoch": 0.7140069384215091, "grad_norm": 0.4041079878807068, "learning_rate": 7.647613520963296e-05, "loss": 0.0341, "step": 6586 }, { "epoch": 0.7141153512575888, "grad_norm": 0.26103150844573975, "learning_rate": 7.647250834179602e-05, "loss": 0.0452, "step": 6587 }, { "epoch": 0.7142237640936687, "grad_norm": 0.29903677105903625, "learning_rate": 7.646888147395909e-05, "loss": 0.0443, "step": 6588 }, { "epoch": 0.7143321769297485, "grad_norm": 0.3990892469882965, "learning_rate": 7.646525460612216e-05, "loss": 0.0562, "step": 6589 }, { "epoch": 0.7144405897658282, "grad_norm": 0.10987496376037598, "learning_rate": 7.646162773828522e-05, "loss": 0.0101, "step": 6590 }, { "epoch": 0.714549002601908, "grad_norm": 0.4233351945877075, "learning_rate": 7.645800087044829e-05, "loss": 0.0592, "step": 6591 }, { "epoch": 0.7146574154379879, "grad_norm": 0.2437610626220703, "learning_rate": 7.645437400261135e-05, "loss": 0.0201, "step": 6592 }, { "epoch": 0.7147658282740676, "grad_norm": 0.46349862217903137, "learning_rate": 7.645074713477442e-05, "loss": 0.0534, "step": 6593 }, { "epoch": 0.7148742411101474, "grad_norm": 0.38115400075912476, "learning_rate": 7.644712026693748e-05, "loss": 0.061, "step": 6594 }, { "epoch": 0.7149826539462273, "grad_norm": 0.1494133472442627, "learning_rate": 7.644349339910055e-05, "loss": 0.023, "step": 6595 }, { "epoch": 0.715091066782307, "grad_norm": 0.24822893738746643, "learning_rate": 7.64398665312636e-05, "loss": 0.0134, "step": 6596 }, { "epoch": 0.7151994796183868, "grad_norm": 0.4975694417953491, "learning_rate": 7.643623966342666e-05, "loss": 0.0233, "step": 6597 }, { "epoch": 0.7153078924544666, "grad_norm": 0.305327445268631, "learning_rate": 7.643261279558973e-05, "loss": 0.0277, "step": 6598 }, { "epoch": 0.7154163052905465, "grad_norm": 0.2809881567955017, "learning_rate": 7.642898592775279e-05, "loss": 0.0245, "step": 6599 }, { "epoch": 0.7155247181266262, "grad_norm": 0.18719929456710815, "learning_rate": 7.642535905991586e-05, "loss": 0.0135, "step": 6600 }, { "epoch": 0.715633130962706, "grad_norm": 0.19855234026908875, "learning_rate": 7.642173219207892e-05, "loss": 0.0191, "step": 6601 }, { "epoch": 0.7157415437987857, "grad_norm": 0.401533305644989, "learning_rate": 7.641810532424199e-05, "loss": 0.0775, "step": 6602 }, { "epoch": 0.7158499566348656, "grad_norm": 0.313150018453598, "learning_rate": 7.641447845640505e-05, "loss": 0.0476, "step": 6603 }, { "epoch": 0.7159583694709454, "grad_norm": 0.16926392912864685, "learning_rate": 7.641085158856812e-05, "loss": 0.0181, "step": 6604 }, { "epoch": 0.7160667823070251, "grad_norm": 0.3151359260082245, "learning_rate": 7.640722472073119e-05, "loss": 0.0342, "step": 6605 }, { "epoch": 0.7161751951431049, "grad_norm": 0.5396778583526611, "learning_rate": 7.640359785289425e-05, "loss": 0.0409, "step": 6606 }, { "epoch": 0.7162836079791848, "grad_norm": 0.41482940316200256, "learning_rate": 7.639997098505732e-05, "loss": 0.0745, "step": 6607 }, { "epoch": 0.7163920208152645, "grad_norm": 0.3121229112148285, "learning_rate": 7.639634411722037e-05, "loss": 0.0362, "step": 6608 }, { "epoch": 0.7165004336513443, "grad_norm": 0.3144085109233856, "learning_rate": 7.639271724938344e-05, "loss": 0.0386, "step": 6609 }, { "epoch": 0.7166088464874241, "grad_norm": 0.15438462793827057, "learning_rate": 7.63890903815465e-05, "loss": 0.0154, "step": 6610 }, { "epoch": 0.7167172593235039, "grad_norm": 0.22071407735347748, "learning_rate": 7.638546351370956e-05, "loss": 0.0402, "step": 6611 }, { "epoch": 0.7168256721595837, "grad_norm": 0.3446645140647888, "learning_rate": 7.638183664587263e-05, "loss": 0.0541, "step": 6612 }, { "epoch": 0.7169340849956635, "grad_norm": 0.15128451585769653, "learning_rate": 7.637820977803569e-05, "loss": 0.0186, "step": 6613 }, { "epoch": 0.7170424978317432, "grad_norm": 0.2803395390510559, "learning_rate": 7.637458291019876e-05, "loss": 0.0307, "step": 6614 }, { "epoch": 0.7171509106678231, "grad_norm": 0.4713580906391144, "learning_rate": 7.637095604236182e-05, "loss": 0.0745, "step": 6615 }, { "epoch": 0.7172593235039029, "grad_norm": 0.13489024341106415, "learning_rate": 7.636732917452489e-05, "loss": 0.0094, "step": 6616 }, { "epoch": 0.7173677363399826, "grad_norm": 0.6237086057662964, "learning_rate": 7.636370230668794e-05, "loss": 0.0407, "step": 6617 }, { "epoch": 0.7174761491760624, "grad_norm": 0.2631889581680298, "learning_rate": 7.636007543885102e-05, "loss": 0.0455, "step": 6618 }, { "epoch": 0.7175845620121423, "grad_norm": 0.4926675856113434, "learning_rate": 7.635644857101407e-05, "loss": 0.0342, "step": 6619 }, { "epoch": 0.717692974848222, "grad_norm": 0.7803299427032471, "learning_rate": 7.635282170317713e-05, "loss": 0.0666, "step": 6620 }, { "epoch": 0.7178013876843018, "grad_norm": 0.6711936593055725, "learning_rate": 7.63491948353402e-05, "loss": 0.0978, "step": 6621 }, { "epoch": 0.7179098005203816, "grad_norm": 0.5514792203903198, "learning_rate": 7.634556796750327e-05, "loss": 0.0652, "step": 6622 }, { "epoch": 0.7180182133564614, "grad_norm": 0.1891336739063263, "learning_rate": 7.634194109966634e-05, "loss": 0.0186, "step": 6623 }, { "epoch": 0.7181266261925412, "grad_norm": 0.12088364362716675, "learning_rate": 7.63383142318294e-05, "loss": 0.0203, "step": 6624 }, { "epoch": 0.718235039028621, "grad_norm": 0.33896130323410034, "learning_rate": 7.633468736399246e-05, "loss": 0.0285, "step": 6625 }, { "epoch": 0.7183434518647008, "grad_norm": 0.34318462014198303, "learning_rate": 7.633106049615553e-05, "loss": 0.0265, "step": 6626 }, { "epoch": 0.7184518647007806, "grad_norm": 0.1646905541419983, "learning_rate": 7.632743362831859e-05, "loss": 0.0169, "step": 6627 }, { "epoch": 0.7185602775368604, "grad_norm": 0.08928373456001282, "learning_rate": 7.632380676048166e-05, "loss": 0.0061, "step": 6628 }, { "epoch": 0.7186686903729401, "grad_norm": 0.1955932378768921, "learning_rate": 7.632017989264471e-05, "loss": 0.0223, "step": 6629 }, { "epoch": 0.71877710320902, "grad_norm": 0.17553389072418213, "learning_rate": 7.631655302480778e-05, "loss": 0.0357, "step": 6630 }, { "epoch": 0.7188855160450998, "grad_norm": 0.19777889549732208, "learning_rate": 7.631292615697084e-05, "loss": 0.0352, "step": 6631 }, { "epoch": 0.7189939288811795, "grad_norm": 0.15024779736995697, "learning_rate": 7.630929928913391e-05, "loss": 0.0118, "step": 6632 }, { "epoch": 0.7191023417172593, "grad_norm": 0.1715162992477417, "learning_rate": 7.630567242129697e-05, "loss": 0.0275, "step": 6633 }, { "epoch": 0.7192107545533392, "grad_norm": 0.39075058698654175, "learning_rate": 7.630204555346003e-05, "loss": 0.0395, "step": 6634 }, { "epoch": 0.7193191673894189, "grad_norm": 0.3719955384731293, "learning_rate": 7.62984186856231e-05, "loss": 0.0322, "step": 6635 }, { "epoch": 0.7194275802254987, "grad_norm": 0.6569964289665222, "learning_rate": 7.629479181778616e-05, "loss": 0.043, "step": 6636 }, { "epoch": 0.7195359930615784, "grad_norm": 0.5511305928230286, "learning_rate": 7.629116494994923e-05, "loss": 0.046, "step": 6637 }, { "epoch": 0.7196444058976583, "grad_norm": 0.28370434045791626, "learning_rate": 7.628753808211228e-05, "loss": 0.0168, "step": 6638 }, { "epoch": 0.7197528187337381, "grad_norm": 0.4850026071071625, "learning_rate": 7.628391121427535e-05, "loss": 0.0702, "step": 6639 }, { "epoch": 0.7198612315698178, "grad_norm": 0.39085397124290466, "learning_rate": 7.628028434643843e-05, "loss": 0.0276, "step": 6640 }, { "epoch": 0.7199696444058976, "grad_norm": 0.37655767798423767, "learning_rate": 7.627665747860148e-05, "loss": 0.0541, "step": 6641 }, { "epoch": 0.7200780572419775, "grad_norm": 0.12209933996200562, "learning_rate": 7.627303061076455e-05, "loss": 0.0134, "step": 6642 }, { "epoch": 0.7201864700780573, "grad_norm": 0.044633250683546066, "learning_rate": 7.626940374292761e-05, "loss": 0.0029, "step": 6643 }, { "epoch": 0.720294882914137, "grad_norm": 0.3475787937641144, "learning_rate": 7.626577687509068e-05, "loss": 0.0674, "step": 6644 }, { "epoch": 0.7204032957502168, "grad_norm": 0.29618239402770996, "learning_rate": 7.626215000725374e-05, "loss": 0.0424, "step": 6645 }, { "epoch": 0.7205117085862967, "grad_norm": 0.48840805888175964, "learning_rate": 7.625852313941681e-05, "loss": 0.0571, "step": 6646 }, { "epoch": 0.7206201214223764, "grad_norm": 0.2733021080493927, "learning_rate": 7.625489627157987e-05, "loss": 0.0121, "step": 6647 }, { "epoch": 0.7207285342584562, "grad_norm": 0.0857432633638382, "learning_rate": 7.625126940374292e-05, "loss": 0.0131, "step": 6648 }, { "epoch": 0.7208369470945359, "grad_norm": 0.3768181800842285, "learning_rate": 7.6247642535906e-05, "loss": 0.0571, "step": 6649 }, { "epoch": 0.7209453599306158, "grad_norm": 0.659064769744873, "learning_rate": 7.624401566806905e-05, "loss": 0.019, "step": 6650 }, { "epoch": 0.7210537727666956, "grad_norm": 0.5207784175872803, "learning_rate": 7.624038880023212e-05, "loss": 0.0851, "step": 6651 }, { "epoch": 0.7211621856027753, "grad_norm": 1.0390299558639526, "learning_rate": 7.623676193239518e-05, "loss": 0.0909, "step": 6652 }, { "epoch": 0.7212705984388552, "grad_norm": 0.26262491941452026, "learning_rate": 7.623313506455825e-05, "loss": 0.0204, "step": 6653 }, { "epoch": 0.721379011274935, "grad_norm": 0.26593470573425293, "learning_rate": 7.622950819672131e-05, "loss": 0.0522, "step": 6654 }, { "epoch": 0.7214874241110147, "grad_norm": 0.19384361803531647, "learning_rate": 7.622588132888438e-05, "loss": 0.0383, "step": 6655 }, { "epoch": 0.7215958369470945, "grad_norm": 0.33666732907295227, "learning_rate": 7.622225446104744e-05, "loss": 0.0503, "step": 6656 }, { "epoch": 0.7217042497831744, "grad_norm": 0.2988271415233612, "learning_rate": 7.621862759321051e-05, "loss": 0.0124, "step": 6657 }, { "epoch": 0.7218126626192541, "grad_norm": 0.25943195819854736, "learning_rate": 7.621500072537358e-05, "loss": 0.0492, "step": 6658 }, { "epoch": 0.7219210754553339, "grad_norm": 0.3535284996032715, "learning_rate": 7.621137385753664e-05, "loss": 0.0525, "step": 6659 }, { "epoch": 0.7220294882914137, "grad_norm": 0.4104549288749695, "learning_rate": 7.620774698969971e-05, "loss": 0.0334, "step": 6660 }, { "epoch": 0.7221379011274935, "grad_norm": 0.6727384924888611, "learning_rate": 7.620412012186276e-05, "loss": 0.0817, "step": 6661 }, { "epoch": 0.7222463139635733, "grad_norm": 0.9717214107513428, "learning_rate": 7.620049325402582e-05, "loss": 0.0591, "step": 6662 }, { "epoch": 0.7223547267996531, "grad_norm": 0.8350337743759155, "learning_rate": 7.619686638618889e-05, "loss": 0.0506, "step": 6663 }, { "epoch": 0.7224631396357328, "grad_norm": 0.22709625959396362, "learning_rate": 7.619323951835195e-05, "loss": 0.0309, "step": 6664 }, { "epoch": 0.7225715524718127, "grad_norm": 0.11457160860300064, "learning_rate": 7.618961265051502e-05, "loss": 0.0225, "step": 6665 }, { "epoch": 0.7226799653078925, "grad_norm": 0.4358973503112793, "learning_rate": 7.618598578267808e-05, "loss": 0.0291, "step": 6666 }, { "epoch": 0.7227883781439722, "grad_norm": 0.19702604413032532, "learning_rate": 7.618235891484115e-05, "loss": 0.0368, "step": 6667 }, { "epoch": 0.722896790980052, "grad_norm": 0.20147818326950073, "learning_rate": 7.61787320470042e-05, "loss": 0.0357, "step": 6668 }, { "epoch": 0.7230052038161319, "grad_norm": 0.2526392638683319, "learning_rate": 7.617510517916728e-05, "loss": 0.0282, "step": 6669 }, { "epoch": 0.7231136166522116, "grad_norm": 0.3066328167915344, "learning_rate": 7.617147831133033e-05, "loss": 0.0391, "step": 6670 }, { "epoch": 0.7232220294882914, "grad_norm": 0.5125746130943298, "learning_rate": 7.616785144349339e-05, "loss": 0.0428, "step": 6671 }, { "epoch": 0.7233304423243712, "grad_norm": 0.31382471323013306, "learning_rate": 7.616422457565646e-05, "loss": 0.0161, "step": 6672 }, { "epoch": 0.723438855160451, "grad_norm": 0.42437979578971863, "learning_rate": 7.616059770781952e-05, "loss": 0.0213, "step": 6673 }, { "epoch": 0.7235472679965308, "grad_norm": 0.6914111375808716, "learning_rate": 7.61569708399826e-05, "loss": 0.0518, "step": 6674 }, { "epoch": 0.7236556808326106, "grad_norm": 0.49367791414260864, "learning_rate": 7.615334397214566e-05, "loss": 0.0463, "step": 6675 }, { "epoch": 0.7237640936686903, "grad_norm": 0.19504311680793762, "learning_rate": 7.614971710430873e-05, "loss": 0.0351, "step": 6676 }, { "epoch": 0.7238725065047702, "grad_norm": 0.42585110664367676, "learning_rate": 7.614609023647179e-05, "loss": 0.0562, "step": 6677 }, { "epoch": 0.72398091934085, "grad_norm": 0.45825526118278503, "learning_rate": 7.614246336863485e-05, "loss": 0.0664, "step": 6678 }, { "epoch": 0.7240893321769297, "grad_norm": 0.6229981780052185, "learning_rate": 7.613883650079792e-05, "loss": 0.0843, "step": 6679 }, { "epoch": 0.7241977450130095, "grad_norm": 0.3772874176502228, "learning_rate": 7.613520963296098e-05, "loss": 0.0569, "step": 6680 }, { "epoch": 0.7243061578490894, "grad_norm": 0.20698687434196472, "learning_rate": 7.613158276512405e-05, "loss": 0.0248, "step": 6681 }, { "epoch": 0.7244145706851691, "grad_norm": 0.42828604578971863, "learning_rate": 7.61279558972871e-05, "loss": 0.0504, "step": 6682 }, { "epoch": 0.7245229835212489, "grad_norm": 0.3936188519001007, "learning_rate": 7.612432902945018e-05, "loss": 0.0686, "step": 6683 }, { "epoch": 0.7246313963573288, "grad_norm": 0.2421291470527649, "learning_rate": 7.612070216161323e-05, "loss": 0.0306, "step": 6684 }, { "epoch": 0.7247398091934085, "grad_norm": 0.11310853809118271, "learning_rate": 7.61170752937763e-05, "loss": 0.0178, "step": 6685 }, { "epoch": 0.7248482220294883, "grad_norm": 0.20698517560958862, "learning_rate": 7.611344842593936e-05, "loss": 0.0194, "step": 6686 }, { "epoch": 0.724956634865568, "grad_norm": 0.43418920040130615, "learning_rate": 7.610982155810242e-05, "loss": 0.063, "step": 6687 }, { "epoch": 0.7250650477016479, "grad_norm": 0.4397362172603607, "learning_rate": 7.610619469026549e-05, "loss": 0.0561, "step": 6688 }, { "epoch": 0.7251734605377277, "grad_norm": 0.4090607166290283, "learning_rate": 7.610256782242855e-05, "loss": 0.0624, "step": 6689 }, { "epoch": 0.7252818733738075, "grad_norm": 0.233744278550148, "learning_rate": 7.609894095459162e-05, "loss": 0.0352, "step": 6690 }, { "epoch": 0.7253902862098872, "grad_norm": 0.373325377702713, "learning_rate": 7.609531408675469e-05, "loss": 0.0641, "step": 6691 }, { "epoch": 0.7254986990459671, "grad_norm": 0.28858399391174316, "learning_rate": 7.609168721891775e-05, "loss": 0.0391, "step": 6692 }, { "epoch": 0.7256071118820469, "grad_norm": 0.5650936365127563, "learning_rate": 7.608806035108082e-05, "loss": 0.0511, "step": 6693 }, { "epoch": 0.7257155247181266, "grad_norm": 0.5546041131019592, "learning_rate": 7.608443348324387e-05, "loss": 0.0314, "step": 6694 }, { "epoch": 0.7258239375542064, "grad_norm": 0.5035632848739624, "learning_rate": 7.608080661540694e-05, "loss": 0.0454, "step": 6695 }, { "epoch": 0.7259323503902863, "grad_norm": 0.26342877745628357, "learning_rate": 7.607717974757e-05, "loss": 0.0428, "step": 6696 }, { "epoch": 0.726040763226366, "grad_norm": 0.2852594256401062, "learning_rate": 7.607355287973307e-05, "loss": 0.0407, "step": 6697 }, { "epoch": 0.7261491760624458, "grad_norm": 0.25562411546707153, "learning_rate": 7.606992601189613e-05, "loss": 0.0272, "step": 6698 }, { "epoch": 0.7262575888985255, "grad_norm": 0.09434258192777634, "learning_rate": 7.60662991440592e-05, "loss": 0.0072, "step": 6699 }, { "epoch": 0.7263660017346054, "grad_norm": 0.3142453134059906, "learning_rate": 7.606267227622226e-05, "loss": 0.037, "step": 6700 }, { "epoch": 0.7264744145706852, "grad_norm": 0.38207024335861206, "learning_rate": 7.605904540838532e-05, "loss": 0.05, "step": 6701 }, { "epoch": 0.7265828274067649, "grad_norm": 0.2638099491596222, "learning_rate": 7.605541854054839e-05, "loss": 0.0418, "step": 6702 }, { "epoch": 0.7266912402428447, "grad_norm": 0.2776723802089691, "learning_rate": 7.605179167271144e-05, "loss": 0.0394, "step": 6703 }, { "epoch": 0.7267996530789246, "grad_norm": 0.46638283133506775, "learning_rate": 7.604816480487451e-05, "loss": 0.0366, "step": 6704 }, { "epoch": 0.7269080659150043, "grad_norm": 0.31154027581214905, "learning_rate": 7.604453793703757e-05, "loss": 0.0304, "step": 6705 }, { "epoch": 0.7270164787510841, "grad_norm": 0.4406312108039856, "learning_rate": 7.604091106920064e-05, "loss": 0.0409, "step": 6706 }, { "epoch": 0.7271248915871639, "grad_norm": 0.17517603933811188, "learning_rate": 7.60372842013637e-05, "loss": 0.0254, "step": 6707 }, { "epoch": 0.7272333044232437, "grad_norm": 0.4353695809841156, "learning_rate": 7.603365733352677e-05, "loss": 0.0355, "step": 6708 }, { "epoch": 0.7273417172593235, "grad_norm": 0.34840017557144165, "learning_rate": 7.603003046568984e-05, "loss": 0.0369, "step": 6709 }, { "epoch": 0.7274501300954033, "grad_norm": 0.2573569118976593, "learning_rate": 7.60264035978529e-05, "loss": 0.0391, "step": 6710 }, { "epoch": 0.727558542931483, "grad_norm": 0.4577818214893341, "learning_rate": 7.602277673001597e-05, "loss": 0.0334, "step": 6711 }, { "epoch": 0.7276669557675629, "grad_norm": 0.30672404170036316, "learning_rate": 7.601914986217903e-05, "loss": 0.0316, "step": 6712 }, { "epoch": 0.7277753686036427, "grad_norm": 0.3721393644809723, "learning_rate": 7.60155229943421e-05, "loss": 0.0509, "step": 6713 }, { "epoch": 0.7278837814397224, "grad_norm": 0.2928743362426758, "learning_rate": 7.601189612650516e-05, "loss": 0.0313, "step": 6714 }, { "epoch": 0.7279921942758023, "grad_norm": 0.4699672758579254, "learning_rate": 7.600826925866821e-05, "loss": 0.0443, "step": 6715 }, { "epoch": 0.7281006071118821, "grad_norm": 0.5367786288261414, "learning_rate": 7.600464239083128e-05, "loss": 0.0663, "step": 6716 }, { "epoch": 0.7282090199479618, "grad_norm": 0.24552308022975922, "learning_rate": 7.600101552299434e-05, "loss": 0.024, "step": 6717 }, { "epoch": 0.7283174327840416, "grad_norm": 0.29080888628959656, "learning_rate": 7.599738865515741e-05, "loss": 0.0215, "step": 6718 }, { "epoch": 0.7284258456201215, "grad_norm": 0.3415171205997467, "learning_rate": 7.599376178732047e-05, "loss": 0.0503, "step": 6719 }, { "epoch": 0.7285342584562012, "grad_norm": 0.27457088232040405, "learning_rate": 7.599013491948354e-05, "loss": 0.036, "step": 6720 }, { "epoch": 0.728642671292281, "grad_norm": 0.42835915088653564, "learning_rate": 7.59865080516466e-05, "loss": 0.0337, "step": 6721 }, { "epoch": 0.7287510841283608, "grad_norm": 0.16969461739063263, "learning_rate": 7.598288118380967e-05, "loss": 0.0196, "step": 6722 }, { "epoch": 0.7288594969644406, "grad_norm": 0.15182077884674072, "learning_rate": 7.597925431597273e-05, "loss": 0.0267, "step": 6723 }, { "epoch": 0.7289679098005204, "grad_norm": 0.3005235195159912, "learning_rate": 7.597562744813578e-05, "loss": 0.0296, "step": 6724 }, { "epoch": 0.7290763226366002, "grad_norm": 0.4543246328830719, "learning_rate": 7.597200058029885e-05, "loss": 0.0426, "step": 6725 }, { "epoch": 0.7291847354726799, "grad_norm": 0.31730416417121887, "learning_rate": 7.596837371246192e-05, "loss": 0.0306, "step": 6726 }, { "epoch": 0.7292931483087598, "grad_norm": 0.23418226838111877, "learning_rate": 7.5964746844625e-05, "loss": 0.0316, "step": 6727 }, { "epoch": 0.7294015611448396, "grad_norm": 0.31363606452941895, "learning_rate": 7.596111997678805e-05, "loss": 0.037, "step": 6728 }, { "epoch": 0.7295099739809193, "grad_norm": 0.3246731460094452, "learning_rate": 7.595749310895111e-05, "loss": 0.0419, "step": 6729 }, { "epoch": 0.7296183868169991, "grad_norm": 0.21025310456752777, "learning_rate": 7.595386624111418e-05, "loss": 0.0224, "step": 6730 }, { "epoch": 0.729726799653079, "grad_norm": 0.4611712396144867, "learning_rate": 7.595023937327724e-05, "loss": 0.0268, "step": 6731 }, { "epoch": 0.7298352124891587, "grad_norm": 0.5714676976203918, "learning_rate": 7.594661250544031e-05, "loss": 0.0621, "step": 6732 }, { "epoch": 0.7299436253252385, "grad_norm": 0.8876472115516663, "learning_rate": 7.594298563760337e-05, "loss": 0.0667, "step": 6733 }, { "epoch": 0.7300520381613183, "grad_norm": 0.19091054797172546, "learning_rate": 7.593935876976644e-05, "loss": 0.0187, "step": 6734 }, { "epoch": 0.7301604509973981, "grad_norm": 0.3612716495990753, "learning_rate": 7.59357319019295e-05, "loss": 0.0275, "step": 6735 }, { "epoch": 0.7302688638334779, "grad_norm": 0.4717462956905365, "learning_rate": 7.593210503409257e-05, "loss": 0.0531, "step": 6736 }, { "epoch": 0.7303772766695577, "grad_norm": 0.2754324972629547, "learning_rate": 7.592847816625562e-05, "loss": 0.0472, "step": 6737 }, { "epoch": 0.7304856895056374, "grad_norm": 0.2159164547920227, "learning_rate": 7.592485129841868e-05, "loss": 0.0203, "step": 6738 }, { "epoch": 0.7305941023417173, "grad_norm": 0.33350998163223267, "learning_rate": 7.592122443058175e-05, "loss": 0.0286, "step": 6739 }, { "epoch": 0.730702515177797, "grad_norm": 1.1599451303482056, "learning_rate": 7.591759756274481e-05, "loss": 0.0881, "step": 6740 }, { "epoch": 0.7308109280138768, "grad_norm": 0.17153728008270264, "learning_rate": 7.591397069490788e-05, "loss": 0.0108, "step": 6741 }, { "epoch": 0.7309193408499567, "grad_norm": 0.17362840473651886, "learning_rate": 7.591034382707094e-05, "loss": 0.0077, "step": 6742 }, { "epoch": 0.7310277536860365, "grad_norm": 0.3639959394931793, "learning_rate": 7.590671695923401e-05, "loss": 0.0404, "step": 6743 }, { "epoch": 0.7311361665221162, "grad_norm": 0.4213394224643707, "learning_rate": 7.590309009139708e-05, "loss": 0.0561, "step": 6744 }, { "epoch": 0.731244579358196, "grad_norm": 0.41483065485954285, "learning_rate": 7.589946322356014e-05, "loss": 0.0539, "step": 6745 }, { "epoch": 0.7313529921942759, "grad_norm": 0.25218018889427185, "learning_rate": 7.589583635572321e-05, "loss": 0.033, "step": 6746 }, { "epoch": 0.7314614050303556, "grad_norm": 0.5462328791618347, "learning_rate": 7.589220948788626e-05, "loss": 0.0551, "step": 6747 }, { "epoch": 0.7315698178664354, "grad_norm": 0.2965540587902069, "learning_rate": 7.588858262004934e-05, "loss": 0.0423, "step": 6748 }, { "epoch": 0.7316782307025151, "grad_norm": 0.9481726288795471, "learning_rate": 7.588495575221239e-05, "loss": 0.0478, "step": 6749 }, { "epoch": 0.731786643538595, "grad_norm": 0.22053876519203186, "learning_rate": 7.588132888437546e-05, "loss": 0.0366, "step": 6750 }, { "epoch": 0.7318950563746748, "grad_norm": 0.6178433299064636, "learning_rate": 7.587770201653852e-05, "loss": 0.0324, "step": 6751 }, { "epoch": 0.7320034692107545, "grad_norm": 0.287946492433548, "learning_rate": 7.587407514870158e-05, "loss": 0.0517, "step": 6752 }, { "epoch": 0.7321118820468343, "grad_norm": 0.35102319717407227, "learning_rate": 7.587044828086465e-05, "loss": 0.0347, "step": 6753 }, { "epoch": 0.7322202948829142, "grad_norm": 0.2579370439052582, "learning_rate": 7.58668214130277e-05, "loss": 0.0406, "step": 6754 }, { "epoch": 0.7323287077189939, "grad_norm": 0.2874129116535187, "learning_rate": 7.586319454519078e-05, "loss": 0.0328, "step": 6755 }, { "epoch": 0.7324371205550737, "grad_norm": 0.6334273815155029, "learning_rate": 7.585956767735383e-05, "loss": 0.0173, "step": 6756 }, { "epoch": 0.7325455333911535, "grad_norm": 0.16688145697116852, "learning_rate": 7.58559408095169e-05, "loss": 0.0173, "step": 6757 }, { "epoch": 0.7326539462272333, "grad_norm": 0.3291058838367462, "learning_rate": 7.585231394167996e-05, "loss": 0.0227, "step": 6758 }, { "epoch": 0.7327623590633131, "grad_norm": 0.41927289962768555, "learning_rate": 7.584868707384303e-05, "loss": 0.0289, "step": 6759 }, { "epoch": 0.7328707718993929, "grad_norm": 0.23320958018302917, "learning_rate": 7.58450602060061e-05, "loss": 0.0243, "step": 6760 }, { "epoch": 0.7329791847354726, "grad_norm": 0.35490065813064575, "learning_rate": 7.584143333816916e-05, "loss": 0.0251, "step": 6761 }, { "epoch": 0.7330875975715525, "grad_norm": 0.2802858352661133, "learning_rate": 7.583780647033223e-05, "loss": 0.0238, "step": 6762 }, { "epoch": 0.7331960104076323, "grad_norm": 0.6079926490783691, "learning_rate": 7.583417960249529e-05, "loss": 0.0569, "step": 6763 }, { "epoch": 0.733304423243712, "grad_norm": 0.3504573106765747, "learning_rate": 7.583055273465836e-05, "loss": 0.0377, "step": 6764 }, { "epoch": 0.7334128360797918, "grad_norm": 0.17238324880599976, "learning_rate": 7.582692586682142e-05, "loss": 0.0239, "step": 6765 }, { "epoch": 0.7335212489158717, "grad_norm": 0.35654690861701965, "learning_rate": 7.582329899898448e-05, "loss": 0.0502, "step": 6766 }, { "epoch": 0.7336296617519514, "grad_norm": 0.4563639163970947, "learning_rate": 7.581967213114755e-05, "loss": 0.0585, "step": 6767 }, { "epoch": 0.7337380745880312, "grad_norm": 0.13119597733020782, "learning_rate": 7.58160452633106e-05, "loss": 0.0113, "step": 6768 }, { "epoch": 0.733846487424111, "grad_norm": 0.9264033436775208, "learning_rate": 7.581241839547367e-05, "loss": 0.0718, "step": 6769 }, { "epoch": 0.7339549002601908, "grad_norm": 0.19148872792720795, "learning_rate": 7.580879152763673e-05, "loss": 0.0109, "step": 6770 }, { "epoch": 0.7340633130962706, "grad_norm": 0.29367220401763916, "learning_rate": 7.58051646597998e-05, "loss": 0.0369, "step": 6771 }, { "epoch": 0.7341717259323504, "grad_norm": 0.2540549039840698, "learning_rate": 7.580153779196286e-05, "loss": 0.0268, "step": 6772 }, { "epoch": 0.7342801387684302, "grad_norm": 0.21317967772483826, "learning_rate": 7.579791092412593e-05, "loss": 0.0441, "step": 6773 }, { "epoch": 0.73438855160451, "grad_norm": 0.16909560561180115, "learning_rate": 7.579428405628899e-05, "loss": 0.0248, "step": 6774 }, { "epoch": 0.7344969644405898, "grad_norm": 0.5147163271903992, "learning_rate": 7.579065718845205e-05, "loss": 0.0729, "step": 6775 }, { "epoch": 0.7346053772766695, "grad_norm": 0.19537287950515747, "learning_rate": 7.578703032061512e-05, "loss": 0.0277, "step": 6776 }, { "epoch": 0.7347137901127494, "grad_norm": 0.3561232089996338, "learning_rate": 7.578340345277817e-05, "loss": 0.0336, "step": 6777 }, { "epoch": 0.7348222029488292, "grad_norm": 0.29498526453971863, "learning_rate": 7.577977658494126e-05, "loss": 0.0404, "step": 6778 }, { "epoch": 0.7349306157849089, "grad_norm": 0.2562641203403473, "learning_rate": 7.577614971710432e-05, "loss": 0.0337, "step": 6779 }, { "epoch": 0.7350390286209887, "grad_norm": 0.3229483664035797, "learning_rate": 7.577252284926739e-05, "loss": 0.0479, "step": 6780 }, { "epoch": 0.7351474414570686, "grad_norm": 0.29755327105522156, "learning_rate": 7.576889598143044e-05, "loss": 0.0173, "step": 6781 }, { "epoch": 0.7352558542931483, "grad_norm": 0.6765231490135193, "learning_rate": 7.57652691135935e-05, "loss": 0.1036, "step": 6782 }, { "epoch": 0.7353642671292281, "grad_norm": 0.19306680560112, "learning_rate": 7.576164224575657e-05, "loss": 0.0172, "step": 6783 }, { "epoch": 0.7354726799653079, "grad_norm": 0.3356556296348572, "learning_rate": 7.575801537791963e-05, "loss": 0.0524, "step": 6784 }, { "epoch": 0.7355810928013877, "grad_norm": 0.3074091374874115, "learning_rate": 7.57543885100827e-05, "loss": 0.0308, "step": 6785 }, { "epoch": 0.7356895056374675, "grad_norm": 0.14806139469146729, "learning_rate": 7.575076164224576e-05, "loss": 0.0275, "step": 6786 }, { "epoch": 0.7357979184735473, "grad_norm": 0.5816655158996582, "learning_rate": 7.574713477440883e-05, "loss": 0.0325, "step": 6787 }, { "epoch": 0.735906331309627, "grad_norm": 0.4536491632461548, "learning_rate": 7.574350790657189e-05, "loss": 0.0447, "step": 6788 }, { "epoch": 0.7360147441457069, "grad_norm": 0.3274644613265991, "learning_rate": 7.573988103873496e-05, "loss": 0.0347, "step": 6789 }, { "epoch": 0.7361231569817867, "grad_norm": 0.3325318396091461, "learning_rate": 7.573625417089801e-05, "loss": 0.0418, "step": 6790 }, { "epoch": 0.7362315698178664, "grad_norm": 0.1855386644601822, "learning_rate": 7.573262730306107e-05, "loss": 0.0182, "step": 6791 }, { "epoch": 0.7363399826539462, "grad_norm": 0.2671017050743103, "learning_rate": 7.572900043522414e-05, "loss": 0.03, "step": 6792 }, { "epoch": 0.736448395490026, "grad_norm": 0.29269686341285706, "learning_rate": 7.57253735673872e-05, "loss": 0.0309, "step": 6793 }, { "epoch": 0.7365568083261058, "grad_norm": 0.08589296787977219, "learning_rate": 7.572174669955027e-05, "loss": 0.0112, "step": 6794 }, { "epoch": 0.7366652211621856, "grad_norm": 0.35824471712112427, "learning_rate": 7.571811983171334e-05, "loss": 0.0202, "step": 6795 }, { "epoch": 0.7367736339982653, "grad_norm": 0.1423661857843399, "learning_rate": 7.57144929638764e-05, "loss": 0.0115, "step": 6796 }, { "epoch": 0.7368820468343452, "grad_norm": 0.23617400228977203, "learning_rate": 7.571086609603947e-05, "loss": 0.023, "step": 6797 }, { "epoch": 0.736990459670425, "grad_norm": 0.2636447250843048, "learning_rate": 7.570723922820253e-05, "loss": 0.0292, "step": 6798 }, { "epoch": 0.7370988725065047, "grad_norm": 0.5084050297737122, "learning_rate": 7.57036123603656e-05, "loss": 0.0559, "step": 6799 }, { "epoch": 0.7372072853425846, "grad_norm": 0.2451961636543274, "learning_rate": 7.569998549252866e-05, "loss": 0.0402, "step": 6800 }, { "epoch": 0.7373156981786644, "grad_norm": 0.29377439618110657, "learning_rate": 7.569635862469173e-05, "loss": 0.0428, "step": 6801 }, { "epoch": 0.7374241110147441, "grad_norm": 0.259431391954422, "learning_rate": 7.569273175685478e-05, "loss": 0.0204, "step": 6802 }, { "epoch": 0.7375325238508239, "grad_norm": 0.31444910168647766, "learning_rate": 7.568910488901785e-05, "loss": 0.0592, "step": 6803 }, { "epoch": 0.7376409366869038, "grad_norm": 0.3649120330810547, "learning_rate": 7.568547802118091e-05, "loss": 0.0525, "step": 6804 }, { "epoch": 0.7377493495229835, "grad_norm": 0.41496017575263977, "learning_rate": 7.568185115334397e-05, "loss": 0.0354, "step": 6805 }, { "epoch": 0.7378577623590633, "grad_norm": 0.3545296788215637, "learning_rate": 7.567822428550704e-05, "loss": 0.0542, "step": 6806 }, { "epoch": 0.7379661751951431, "grad_norm": 0.14454196393489838, "learning_rate": 7.56745974176701e-05, "loss": 0.0233, "step": 6807 }, { "epoch": 0.738074588031223, "grad_norm": 0.3672540485858917, "learning_rate": 7.567097054983317e-05, "loss": 0.0349, "step": 6808 }, { "epoch": 0.7381830008673027, "grad_norm": 0.268638014793396, "learning_rate": 7.566734368199623e-05, "loss": 0.0338, "step": 6809 }, { "epoch": 0.7382914137033825, "grad_norm": 0.17065554857254028, "learning_rate": 7.56637168141593e-05, "loss": 0.0237, "step": 6810 }, { "epoch": 0.7383998265394622, "grad_norm": 0.7507112622261047, "learning_rate": 7.566008994632235e-05, "loss": 0.0743, "step": 6811 }, { "epoch": 0.7385082393755421, "grad_norm": 0.23759245872497559, "learning_rate": 7.565646307848542e-05, "loss": 0.0354, "step": 6812 }, { "epoch": 0.7386166522116219, "grad_norm": 0.3925892412662506, "learning_rate": 7.56528362106485e-05, "loss": 0.0196, "step": 6813 }, { "epoch": 0.7387250650477016, "grad_norm": 0.5848203897476196, "learning_rate": 7.564920934281155e-05, "loss": 0.0445, "step": 6814 }, { "epoch": 0.7388334778837814, "grad_norm": 0.4412848949432373, "learning_rate": 7.564558247497462e-05, "loss": 0.038, "step": 6815 }, { "epoch": 0.7389418907198613, "grad_norm": 0.293239951133728, "learning_rate": 7.564195560713768e-05, "loss": 0.0348, "step": 6816 }, { "epoch": 0.739050303555941, "grad_norm": 0.11764238029718399, "learning_rate": 7.563832873930075e-05, "loss": 0.0114, "step": 6817 }, { "epoch": 0.7391587163920208, "grad_norm": 0.2195008546113968, "learning_rate": 7.563470187146381e-05, "loss": 0.0199, "step": 6818 }, { "epoch": 0.7392671292281006, "grad_norm": 0.6024237275123596, "learning_rate": 7.563107500362687e-05, "loss": 0.0616, "step": 6819 }, { "epoch": 0.7393755420641804, "grad_norm": 0.5640032887458801, "learning_rate": 7.562744813578994e-05, "loss": 0.0448, "step": 6820 }, { "epoch": 0.7394839549002602, "grad_norm": 0.24269403517246246, "learning_rate": 7.5623821267953e-05, "loss": 0.0364, "step": 6821 }, { "epoch": 0.73959236773634, "grad_norm": 0.3367210030555725, "learning_rate": 7.562019440011607e-05, "loss": 0.0417, "step": 6822 }, { "epoch": 0.7397007805724197, "grad_norm": 0.39930930733680725, "learning_rate": 7.561656753227912e-05, "loss": 0.0255, "step": 6823 }, { "epoch": 0.7398091934084996, "grad_norm": 0.2817835211753845, "learning_rate": 7.56129406644422e-05, "loss": 0.0237, "step": 6824 }, { "epoch": 0.7399176062445794, "grad_norm": 0.6797050833702087, "learning_rate": 7.560931379660525e-05, "loss": 0.0513, "step": 6825 }, { "epoch": 0.7400260190806591, "grad_norm": 0.3465253710746765, "learning_rate": 7.560568692876832e-05, "loss": 0.0369, "step": 6826 }, { "epoch": 0.7401344319167389, "grad_norm": 0.19734396040439606, "learning_rate": 7.560206006093138e-05, "loss": 0.02, "step": 6827 }, { "epoch": 0.7402428447528188, "grad_norm": 0.5001591444015503, "learning_rate": 7.559843319309444e-05, "loss": 0.0461, "step": 6828 }, { "epoch": 0.7403512575888985, "grad_norm": 0.5058146119117737, "learning_rate": 7.559480632525752e-05, "loss": 0.0123, "step": 6829 }, { "epoch": 0.7404596704249783, "grad_norm": 0.1959228515625, "learning_rate": 7.559117945742058e-05, "loss": 0.0332, "step": 6830 }, { "epoch": 0.7405680832610582, "grad_norm": 0.3007933497428894, "learning_rate": 7.558755258958365e-05, "loss": 0.0702, "step": 6831 }, { "epoch": 0.7406764960971379, "grad_norm": 0.2690722644329071, "learning_rate": 7.55839257217467e-05, "loss": 0.0416, "step": 6832 }, { "epoch": 0.7407849089332177, "grad_norm": 0.2758348882198334, "learning_rate": 7.558029885390976e-05, "loss": 0.0328, "step": 6833 }, { "epoch": 0.7408933217692975, "grad_norm": 0.25519391894340515, "learning_rate": 7.557667198607283e-05, "loss": 0.0374, "step": 6834 }, { "epoch": 0.7410017346053773, "grad_norm": 0.5329705476760864, "learning_rate": 7.557304511823589e-05, "loss": 0.0683, "step": 6835 }, { "epoch": 0.7411101474414571, "grad_norm": 0.21760767698287964, "learning_rate": 7.556941825039896e-05, "loss": 0.0234, "step": 6836 }, { "epoch": 0.7412185602775369, "grad_norm": 0.29842671751976013, "learning_rate": 7.556579138256202e-05, "loss": 0.0342, "step": 6837 }, { "epoch": 0.7413269731136166, "grad_norm": 0.25416433811187744, "learning_rate": 7.556216451472509e-05, "loss": 0.0389, "step": 6838 }, { "epoch": 0.7414353859496965, "grad_norm": 0.29963377118110657, "learning_rate": 7.555853764688815e-05, "loss": 0.0239, "step": 6839 }, { "epoch": 0.7415437987857763, "grad_norm": 0.36354559659957886, "learning_rate": 7.555491077905122e-05, "loss": 0.0514, "step": 6840 }, { "epoch": 0.741652211621856, "grad_norm": 0.4629923403263092, "learning_rate": 7.555128391121428e-05, "loss": 0.0554, "step": 6841 }, { "epoch": 0.7417606244579358, "grad_norm": 0.3779583275318146, "learning_rate": 7.554765704337733e-05, "loss": 0.0481, "step": 6842 }, { "epoch": 0.7418690372940157, "grad_norm": 0.4052160978317261, "learning_rate": 7.55440301755404e-05, "loss": 0.0404, "step": 6843 }, { "epoch": 0.7419774501300954, "grad_norm": 0.2512308657169342, "learning_rate": 7.554040330770346e-05, "loss": 0.03, "step": 6844 }, { "epoch": 0.7420858629661752, "grad_norm": 0.4530699849128723, "learning_rate": 7.553677643986653e-05, "loss": 0.0426, "step": 6845 }, { "epoch": 0.7421942758022549, "grad_norm": 0.39679154753685, "learning_rate": 7.553314957202959e-05, "loss": 0.0583, "step": 6846 }, { "epoch": 0.7423026886383348, "grad_norm": 0.3727567195892334, "learning_rate": 7.552952270419266e-05, "loss": 0.0343, "step": 6847 }, { "epoch": 0.7424111014744146, "grad_norm": 1.5090997219085693, "learning_rate": 7.552589583635573e-05, "loss": 0.0455, "step": 6848 }, { "epoch": 0.7425195143104943, "grad_norm": 0.3232608437538147, "learning_rate": 7.552226896851879e-05, "loss": 0.0169, "step": 6849 }, { "epoch": 0.7426279271465741, "grad_norm": 0.3155367076396942, "learning_rate": 7.551864210068186e-05, "loss": 0.0384, "step": 6850 }, { "epoch": 0.742736339982654, "grad_norm": 0.7787783145904541, "learning_rate": 7.551501523284492e-05, "loss": 0.098, "step": 6851 }, { "epoch": 0.7428447528187337, "grad_norm": 0.2983822524547577, "learning_rate": 7.551138836500799e-05, "loss": 0.0306, "step": 6852 }, { "epoch": 0.7429531656548135, "grad_norm": 0.4626123905181885, "learning_rate": 7.550776149717105e-05, "loss": 0.0743, "step": 6853 }, { "epoch": 0.7430615784908933, "grad_norm": 0.31785500049591064, "learning_rate": 7.550413462933412e-05, "loss": 0.0376, "step": 6854 }, { "epoch": 0.7431699913269731, "grad_norm": 0.44415318965911865, "learning_rate": 7.550050776149717e-05, "loss": 0.0367, "step": 6855 }, { "epoch": 0.7432784041630529, "grad_norm": 0.26916953921318054, "learning_rate": 7.549688089366023e-05, "loss": 0.0301, "step": 6856 }, { "epoch": 0.7433868169991327, "grad_norm": 0.5676373243331909, "learning_rate": 7.54932540258233e-05, "loss": 0.0401, "step": 6857 }, { "epoch": 0.7434952298352124, "grad_norm": 0.2854067087173462, "learning_rate": 7.548962715798636e-05, "loss": 0.0333, "step": 6858 }, { "epoch": 0.7436036426712923, "grad_norm": 0.9510315656661987, "learning_rate": 7.548600029014943e-05, "loss": 0.0645, "step": 6859 }, { "epoch": 0.7437120555073721, "grad_norm": 0.42200425267219543, "learning_rate": 7.548237342231249e-05, "loss": 0.0339, "step": 6860 }, { "epoch": 0.7438204683434518, "grad_norm": 0.2832563817501068, "learning_rate": 7.547874655447556e-05, "loss": 0.0459, "step": 6861 }, { "epoch": 0.7439288811795317, "grad_norm": 0.17489060759544373, "learning_rate": 7.547511968663862e-05, "loss": 0.0248, "step": 6862 }, { "epoch": 0.7440372940156115, "grad_norm": 0.5399864315986633, "learning_rate": 7.547149281880169e-05, "loss": 0.057, "step": 6863 }, { "epoch": 0.7441457068516912, "grad_norm": 0.0777546688914299, "learning_rate": 7.546786595096476e-05, "loss": 0.0063, "step": 6864 }, { "epoch": 0.744254119687771, "grad_norm": 0.29007643461227417, "learning_rate": 7.546423908312782e-05, "loss": 0.0432, "step": 6865 }, { "epoch": 0.7443625325238509, "grad_norm": 0.3099289834499359, "learning_rate": 7.546061221529089e-05, "loss": 0.0623, "step": 6866 }, { "epoch": 0.7444709453599306, "grad_norm": 0.2682158052921295, "learning_rate": 7.545698534745394e-05, "loss": 0.0419, "step": 6867 }, { "epoch": 0.7445793581960104, "grad_norm": 0.27263492345809937, "learning_rate": 7.545335847961701e-05, "loss": 0.0294, "step": 6868 }, { "epoch": 0.7446877710320902, "grad_norm": 0.3203475773334503, "learning_rate": 7.544973161178007e-05, "loss": 0.0576, "step": 6869 }, { "epoch": 0.74479618386817, "grad_norm": 0.25522148609161377, "learning_rate": 7.544610474394314e-05, "loss": 0.0149, "step": 6870 }, { "epoch": 0.7449045967042498, "grad_norm": 0.3737538158893585, "learning_rate": 7.54424778761062e-05, "loss": 0.0285, "step": 6871 }, { "epoch": 0.7450130095403296, "grad_norm": 0.45020589232444763, "learning_rate": 7.543885100826926e-05, "loss": 0.0895, "step": 6872 }, { "epoch": 0.7451214223764093, "grad_norm": 0.39400967955589294, "learning_rate": 7.543522414043233e-05, "loss": 0.0691, "step": 6873 }, { "epoch": 0.7452298352124892, "grad_norm": 0.3579266667366028, "learning_rate": 7.543159727259539e-05, "loss": 0.0558, "step": 6874 }, { "epoch": 0.745338248048569, "grad_norm": 0.15896904468536377, "learning_rate": 7.542797040475846e-05, "loss": 0.0266, "step": 6875 }, { "epoch": 0.7454466608846487, "grad_norm": 0.3348942697048187, "learning_rate": 7.542434353692151e-05, "loss": 0.0459, "step": 6876 }, { "epoch": 0.7455550737207285, "grad_norm": 0.16176757216453552, "learning_rate": 7.542071666908458e-05, "loss": 0.0219, "step": 6877 }, { "epoch": 0.7456634865568084, "grad_norm": 0.3223296105861664, "learning_rate": 7.541708980124764e-05, "loss": 0.0389, "step": 6878 }, { "epoch": 0.7457718993928881, "grad_norm": 0.3414480686187744, "learning_rate": 7.54134629334107e-05, "loss": 0.0572, "step": 6879 }, { "epoch": 0.7458803122289679, "grad_norm": 0.3060081899166107, "learning_rate": 7.540983606557377e-05, "loss": 0.0377, "step": 6880 }, { "epoch": 0.7459887250650477, "grad_norm": 0.30486464500427246, "learning_rate": 7.540620919773684e-05, "loss": 0.035, "step": 6881 }, { "epoch": 0.7460971379011275, "grad_norm": 0.2146032452583313, "learning_rate": 7.540258232989991e-05, "loss": 0.0364, "step": 6882 }, { "epoch": 0.7462055507372073, "grad_norm": 0.17710906267166138, "learning_rate": 7.539895546206297e-05, "loss": 0.0286, "step": 6883 }, { "epoch": 0.746313963573287, "grad_norm": 0.2384698987007141, "learning_rate": 7.539532859422604e-05, "loss": 0.0392, "step": 6884 }, { "epoch": 0.7464223764093668, "grad_norm": 0.6257478594779968, "learning_rate": 7.53917017263891e-05, "loss": 0.0567, "step": 6885 }, { "epoch": 0.7465307892454467, "grad_norm": 0.3177439272403717, "learning_rate": 7.538807485855215e-05, "loss": 0.0323, "step": 6886 }, { "epoch": 0.7466392020815265, "grad_norm": 0.24290020763874054, "learning_rate": 7.538444799071523e-05, "loss": 0.0605, "step": 6887 }, { "epoch": 0.7467476149176062, "grad_norm": 0.3426104784011841, "learning_rate": 7.538082112287828e-05, "loss": 0.0413, "step": 6888 }, { "epoch": 0.7468560277536861, "grad_norm": 0.3761512339115143, "learning_rate": 7.537719425504135e-05, "loss": 0.0641, "step": 6889 }, { "epoch": 0.7469644405897659, "grad_norm": 0.33565354347229004, "learning_rate": 7.537356738720441e-05, "loss": 0.0362, "step": 6890 }, { "epoch": 0.7470728534258456, "grad_norm": 0.11156967282295227, "learning_rate": 7.536994051936748e-05, "loss": 0.0115, "step": 6891 }, { "epoch": 0.7471812662619254, "grad_norm": 0.30946677923202515, "learning_rate": 7.536631365153054e-05, "loss": 0.0728, "step": 6892 }, { "epoch": 0.7472896790980053, "grad_norm": 0.39155539870262146, "learning_rate": 7.536268678369361e-05, "loss": 0.0534, "step": 6893 }, { "epoch": 0.747398091934085, "grad_norm": 0.6528749465942383, "learning_rate": 7.535905991585667e-05, "loss": 0.0567, "step": 6894 }, { "epoch": 0.7475065047701648, "grad_norm": 0.5183363556861877, "learning_rate": 7.535543304801972e-05, "loss": 0.0699, "step": 6895 }, { "epoch": 0.7476149176062445, "grad_norm": 0.29948243498802185, "learning_rate": 7.53518061801828e-05, "loss": 0.03, "step": 6896 }, { "epoch": 0.7477233304423244, "grad_norm": 0.13712772727012634, "learning_rate": 7.534817931234585e-05, "loss": 0.018, "step": 6897 }, { "epoch": 0.7478317432784042, "grad_norm": 0.18177589774131775, "learning_rate": 7.534455244450894e-05, "loss": 0.0148, "step": 6898 }, { "epoch": 0.747940156114484, "grad_norm": 0.38010865449905396, "learning_rate": 7.5340925576672e-05, "loss": 0.0506, "step": 6899 }, { "epoch": 0.7480485689505637, "grad_norm": 0.22138532996177673, "learning_rate": 7.533729870883505e-05, "loss": 0.0191, "step": 6900 }, { "epoch": 0.7481569817866436, "grad_norm": 0.2447098046541214, "learning_rate": 7.533367184099812e-05, "loss": 0.0523, "step": 6901 }, { "epoch": 0.7482653946227233, "grad_norm": 0.41160422563552856, "learning_rate": 7.533004497316118e-05, "loss": 0.0256, "step": 6902 }, { "epoch": 0.7483738074588031, "grad_norm": 0.35822466015815735, "learning_rate": 7.532641810532425e-05, "loss": 0.0616, "step": 6903 }, { "epoch": 0.7484822202948829, "grad_norm": 0.3869342803955078, "learning_rate": 7.532279123748731e-05, "loss": 0.0256, "step": 6904 }, { "epoch": 0.7485906331309627, "grad_norm": 0.2931506931781769, "learning_rate": 7.531916436965038e-05, "loss": 0.0417, "step": 6905 }, { "epoch": 0.7486990459670425, "grad_norm": 0.2343239188194275, "learning_rate": 7.531553750181344e-05, "loss": 0.0571, "step": 6906 }, { "epoch": 0.7488074588031223, "grad_norm": 0.2668052613735199, "learning_rate": 7.531191063397651e-05, "loss": 0.0509, "step": 6907 }, { "epoch": 0.748915871639202, "grad_norm": 0.15432021021842957, "learning_rate": 7.530828376613956e-05, "loss": 0.0157, "step": 6908 }, { "epoch": 0.7490242844752819, "grad_norm": 0.464392751455307, "learning_rate": 7.530465689830262e-05, "loss": 0.0671, "step": 6909 }, { "epoch": 0.7491326973113617, "grad_norm": 0.5186446905136108, "learning_rate": 7.530103003046569e-05, "loss": 0.0665, "step": 6910 }, { "epoch": 0.7492411101474414, "grad_norm": 0.11396578699350357, "learning_rate": 7.529740316262875e-05, "loss": 0.0234, "step": 6911 }, { "epoch": 0.7493495229835212, "grad_norm": 0.1232098788022995, "learning_rate": 7.529377629479182e-05, "loss": 0.0121, "step": 6912 }, { "epoch": 0.7494579358196011, "grad_norm": 0.3626996874809265, "learning_rate": 7.529014942695488e-05, "loss": 0.0394, "step": 6913 }, { "epoch": 0.7495663486556808, "grad_norm": 0.36309632658958435, "learning_rate": 7.528652255911795e-05, "loss": 0.0504, "step": 6914 }, { "epoch": 0.7496747614917606, "grad_norm": 0.7168014645576477, "learning_rate": 7.528289569128101e-05, "loss": 0.0775, "step": 6915 }, { "epoch": 0.7497831743278404, "grad_norm": 0.35282421112060547, "learning_rate": 7.527926882344408e-05, "loss": 0.0311, "step": 6916 }, { "epoch": 0.7498915871639202, "grad_norm": 0.649431586265564, "learning_rate": 7.527564195560715e-05, "loss": 0.0516, "step": 6917 }, { "epoch": 0.75, "grad_norm": 0.1792842596769333, "learning_rate": 7.52720150877702e-05, "loss": 0.0338, "step": 6918 }, { "epoch": 0.7501084128360798, "grad_norm": 0.29423028230667114, "learning_rate": 7.526838821993328e-05, "loss": 0.0403, "step": 6919 }, { "epoch": 0.7502168256721596, "grad_norm": 0.3246663212776184, "learning_rate": 7.526476135209633e-05, "loss": 0.0436, "step": 6920 }, { "epoch": 0.7503252385082394, "grad_norm": 0.472804456949234, "learning_rate": 7.52611344842594e-05, "loss": 0.0669, "step": 6921 }, { "epoch": 0.7504336513443192, "grad_norm": 0.1671767383813858, "learning_rate": 7.525750761642246e-05, "loss": 0.0201, "step": 6922 }, { "epoch": 0.7505420641803989, "grad_norm": 0.15504758059978485, "learning_rate": 7.525388074858552e-05, "loss": 0.0262, "step": 6923 }, { "epoch": 0.7506504770164788, "grad_norm": 0.32507601380348206, "learning_rate": 7.525025388074859e-05, "loss": 0.0297, "step": 6924 }, { "epoch": 0.7507588898525586, "grad_norm": 0.3317504823207855, "learning_rate": 7.524662701291165e-05, "loss": 0.0789, "step": 6925 }, { "epoch": 0.7508673026886383, "grad_norm": 0.34460073709487915, "learning_rate": 7.524300014507472e-05, "loss": 0.0625, "step": 6926 }, { "epoch": 0.7509757155247181, "grad_norm": 0.14853045344352722, "learning_rate": 7.523937327723778e-05, "loss": 0.018, "step": 6927 }, { "epoch": 0.751084128360798, "grad_norm": 0.581023097038269, "learning_rate": 7.523574640940085e-05, "loss": 0.0532, "step": 6928 }, { "epoch": 0.7511925411968777, "grad_norm": 0.4348589777946472, "learning_rate": 7.52321195415639e-05, "loss": 0.0664, "step": 6929 }, { "epoch": 0.7513009540329575, "grad_norm": 0.36096224188804626, "learning_rate": 7.522849267372698e-05, "loss": 0.042, "step": 6930 }, { "epoch": 0.7514093668690373, "grad_norm": 0.25156334042549133, "learning_rate": 7.522486580589003e-05, "loss": 0.0183, "step": 6931 }, { "epoch": 0.7515177797051171, "grad_norm": 0.40340912342071533, "learning_rate": 7.522123893805309e-05, "loss": 0.0459, "step": 6932 }, { "epoch": 0.7516261925411969, "grad_norm": 0.17541664838790894, "learning_rate": 7.521761207021617e-05, "loss": 0.0119, "step": 6933 }, { "epoch": 0.7517346053772767, "grad_norm": 0.09627119451761246, "learning_rate": 7.521398520237923e-05, "loss": 0.0125, "step": 6934 }, { "epoch": 0.7518430182133564, "grad_norm": 0.35040849447250366, "learning_rate": 7.52103583345423e-05, "loss": 0.0201, "step": 6935 }, { "epoch": 0.7519514310494363, "grad_norm": 0.26734688878059387, "learning_rate": 7.520673146670536e-05, "loss": 0.0361, "step": 6936 }, { "epoch": 0.752059843885516, "grad_norm": 0.14955735206604004, "learning_rate": 7.520310459886842e-05, "loss": 0.0236, "step": 6937 }, { "epoch": 0.7521682567215958, "grad_norm": 0.09404092282056808, "learning_rate": 7.519947773103149e-05, "loss": 0.0154, "step": 6938 }, { "epoch": 0.7522766695576756, "grad_norm": 0.3457745611667633, "learning_rate": 7.519585086319455e-05, "loss": 0.0409, "step": 6939 }, { "epoch": 0.7523850823937555, "grad_norm": 0.3304799199104309, "learning_rate": 7.519222399535762e-05, "loss": 0.0567, "step": 6940 }, { "epoch": 0.7524934952298352, "grad_norm": 0.16361938416957855, "learning_rate": 7.518859712752067e-05, "loss": 0.0308, "step": 6941 }, { "epoch": 0.752601908065915, "grad_norm": 0.34580540657043457, "learning_rate": 7.518497025968374e-05, "loss": 0.0442, "step": 6942 }, { "epoch": 0.7527103209019947, "grad_norm": 0.24449703097343445, "learning_rate": 7.51813433918468e-05, "loss": 0.0257, "step": 6943 }, { "epoch": 0.7528187337380746, "grad_norm": 0.31084227561950684, "learning_rate": 7.517771652400987e-05, "loss": 0.0496, "step": 6944 }, { "epoch": 0.7529271465741544, "grad_norm": 0.17256268858909607, "learning_rate": 7.517408965617293e-05, "loss": 0.0305, "step": 6945 }, { "epoch": 0.7530355594102341, "grad_norm": 0.2527485489845276, "learning_rate": 7.517046278833599e-05, "loss": 0.0339, "step": 6946 }, { "epoch": 0.7531439722463139, "grad_norm": 0.323839396238327, "learning_rate": 7.516683592049906e-05, "loss": 0.0442, "step": 6947 }, { "epoch": 0.7532523850823938, "grad_norm": 0.2694299817085266, "learning_rate": 7.516320905266212e-05, "loss": 0.0366, "step": 6948 }, { "epoch": 0.7533607979184735, "grad_norm": 0.2664513885974884, "learning_rate": 7.515958218482519e-05, "loss": 0.0577, "step": 6949 }, { "epoch": 0.7534692107545533, "grad_norm": 0.2585594952106476, "learning_rate": 7.515595531698826e-05, "loss": 0.0596, "step": 6950 }, { "epoch": 0.7535776235906332, "grad_norm": 0.3732403516769409, "learning_rate": 7.515232844915131e-05, "loss": 0.0509, "step": 6951 }, { "epoch": 0.753686036426713, "grad_norm": 0.45308810472488403, "learning_rate": 7.514870158131439e-05, "loss": 0.0549, "step": 6952 }, { "epoch": 0.7537944492627927, "grad_norm": 0.3853611350059509, "learning_rate": 7.514507471347744e-05, "loss": 0.0673, "step": 6953 }, { "epoch": 0.7539028620988725, "grad_norm": 0.16155485808849335, "learning_rate": 7.514144784564051e-05, "loss": 0.0243, "step": 6954 }, { "epoch": 0.7540112749349523, "grad_norm": 0.24926955997943878, "learning_rate": 7.513782097780357e-05, "loss": 0.0301, "step": 6955 }, { "epoch": 0.7541196877710321, "grad_norm": 0.35412395000457764, "learning_rate": 7.513419410996664e-05, "loss": 0.042, "step": 6956 }, { "epoch": 0.7542281006071119, "grad_norm": 0.3848608434200287, "learning_rate": 7.51305672421297e-05, "loss": 0.0506, "step": 6957 }, { "epoch": 0.7543365134431916, "grad_norm": 0.32462242245674133, "learning_rate": 7.512694037429277e-05, "loss": 0.0518, "step": 6958 }, { "epoch": 0.7544449262792715, "grad_norm": 0.1580457091331482, "learning_rate": 7.512331350645583e-05, "loss": 0.019, "step": 6959 }, { "epoch": 0.7545533391153513, "grad_norm": 0.4705762565135956, "learning_rate": 7.511968663861888e-05, "loss": 0.0712, "step": 6960 }, { "epoch": 0.754661751951431, "grad_norm": 0.3322925865650177, "learning_rate": 7.511605977078196e-05, "loss": 0.0421, "step": 6961 }, { "epoch": 0.7547701647875108, "grad_norm": 0.39699938893318176, "learning_rate": 7.511243290294501e-05, "loss": 0.0596, "step": 6962 }, { "epoch": 0.7548785776235907, "grad_norm": 0.24584458768367767, "learning_rate": 7.510880603510808e-05, "loss": 0.0514, "step": 6963 }, { "epoch": 0.7549869904596704, "grad_norm": 0.23059093952178955, "learning_rate": 7.510517916727114e-05, "loss": 0.0187, "step": 6964 }, { "epoch": 0.7550954032957502, "grad_norm": 0.18367137014865875, "learning_rate": 7.510155229943421e-05, "loss": 0.041, "step": 6965 }, { "epoch": 0.75520381613183, "grad_norm": 0.2696975767612457, "learning_rate": 7.509792543159727e-05, "loss": 0.0442, "step": 6966 }, { "epoch": 0.7553122289679098, "grad_norm": 0.4307419955730438, "learning_rate": 7.509429856376034e-05, "loss": 0.0404, "step": 6967 }, { "epoch": 0.7554206418039896, "grad_norm": 0.49499422311782837, "learning_rate": 7.509067169592341e-05, "loss": 0.0374, "step": 6968 }, { "epoch": 0.7555290546400694, "grad_norm": 0.17840838432312012, "learning_rate": 7.508704482808647e-05, "loss": 0.0135, "step": 6969 }, { "epoch": 0.7556374674761491, "grad_norm": 0.6672370433807373, "learning_rate": 7.508341796024954e-05, "loss": 0.0771, "step": 6970 }, { "epoch": 0.755745880312229, "grad_norm": 0.2318275272846222, "learning_rate": 7.50797910924126e-05, "loss": 0.0302, "step": 6971 }, { "epoch": 0.7558542931483088, "grad_norm": 0.11035683751106262, "learning_rate": 7.507616422457567e-05, "loss": 0.0186, "step": 6972 }, { "epoch": 0.7559627059843885, "grad_norm": 0.535320520401001, "learning_rate": 7.507253735673872e-05, "loss": 0.0179, "step": 6973 }, { "epoch": 0.7560711188204683, "grad_norm": 0.2947462797164917, "learning_rate": 7.50689104889018e-05, "loss": 0.054, "step": 6974 }, { "epoch": 0.7561795316565482, "grad_norm": 0.616535484790802, "learning_rate": 7.506528362106485e-05, "loss": 0.0368, "step": 6975 }, { "epoch": 0.7562879444926279, "grad_norm": 0.23880122601985931, "learning_rate": 7.506165675322791e-05, "loss": 0.0241, "step": 6976 }, { "epoch": 0.7563963573287077, "grad_norm": 0.2418154925107956, "learning_rate": 7.505802988539098e-05, "loss": 0.0474, "step": 6977 }, { "epoch": 0.7565047701647876, "grad_norm": 0.1887793242931366, "learning_rate": 7.505440301755404e-05, "loss": 0.0215, "step": 6978 }, { "epoch": 0.7566131830008673, "grad_norm": 1.1654118299484253, "learning_rate": 7.505077614971711e-05, "loss": 0.0942, "step": 6979 }, { "epoch": 0.7567215958369471, "grad_norm": 0.26994991302490234, "learning_rate": 7.504714928188017e-05, "loss": 0.0418, "step": 6980 }, { "epoch": 0.7568300086730269, "grad_norm": 0.3849394619464874, "learning_rate": 7.504352241404324e-05, "loss": 0.0462, "step": 6981 }, { "epoch": 0.7569384215091067, "grad_norm": 0.39727625250816345, "learning_rate": 7.50398955462063e-05, "loss": 0.0641, "step": 6982 }, { "epoch": 0.7570468343451865, "grad_norm": 0.12011098861694336, "learning_rate": 7.503626867836935e-05, "loss": 0.0098, "step": 6983 }, { "epoch": 0.7571552471812663, "grad_norm": 0.3418634831905365, "learning_rate": 7.503264181053242e-05, "loss": 0.0271, "step": 6984 }, { "epoch": 0.757263660017346, "grad_norm": 0.37122610211372375, "learning_rate": 7.50290149426955e-05, "loss": 0.0738, "step": 6985 }, { "epoch": 0.7573720728534259, "grad_norm": 0.3176431357860565, "learning_rate": 7.502538807485857e-05, "loss": 0.0355, "step": 6986 }, { "epoch": 0.7574804856895057, "grad_norm": 0.6418431401252747, "learning_rate": 7.502176120702162e-05, "loss": 0.0561, "step": 6987 }, { "epoch": 0.7575888985255854, "grad_norm": 0.681372344493866, "learning_rate": 7.50181343391847e-05, "loss": 0.0377, "step": 6988 }, { "epoch": 0.7576973113616652, "grad_norm": 0.32026875019073486, "learning_rate": 7.501450747134775e-05, "loss": 0.0384, "step": 6989 }, { "epoch": 0.7578057241977451, "grad_norm": 0.6220223903656006, "learning_rate": 7.501088060351081e-05, "loss": 0.0362, "step": 6990 }, { "epoch": 0.7579141370338248, "grad_norm": 0.12896619737148285, "learning_rate": 7.500725373567388e-05, "loss": 0.0114, "step": 6991 }, { "epoch": 0.7580225498699046, "grad_norm": 0.4303809106349945, "learning_rate": 7.500362686783694e-05, "loss": 0.0529, "step": 6992 }, { "epoch": 0.7581309627059843, "grad_norm": 0.29605042934417725, "learning_rate": 7.500000000000001e-05, "loss": 0.0625, "step": 6993 }, { "epoch": 0.7582393755420642, "grad_norm": 0.251390278339386, "learning_rate": 7.499637313216306e-05, "loss": 0.0236, "step": 6994 }, { "epoch": 0.758347788378144, "grad_norm": 0.24345411360263824, "learning_rate": 7.499274626432614e-05, "loss": 0.0326, "step": 6995 }, { "epoch": 0.7584562012142237, "grad_norm": 0.1303316056728363, "learning_rate": 7.498911939648919e-05, "loss": 0.0246, "step": 6996 }, { "epoch": 0.7585646140503035, "grad_norm": 0.3430146276950836, "learning_rate": 7.498549252865226e-05, "loss": 0.0294, "step": 6997 }, { "epoch": 0.7586730268863834, "grad_norm": 0.43264639377593994, "learning_rate": 7.498186566081532e-05, "loss": 0.0271, "step": 6998 }, { "epoch": 0.7587814397224631, "grad_norm": 0.22441859543323517, "learning_rate": 7.497823879297838e-05, "loss": 0.0343, "step": 6999 }, { "epoch": 0.7588898525585429, "grad_norm": 0.21478979289531708, "learning_rate": 7.497461192514145e-05, "loss": 0.0455, "step": 7000 }, { "epoch": 0.7589982653946227, "grad_norm": 0.6600359082221985, "learning_rate": 7.49709850573045e-05, "loss": 0.0715, "step": 7001 }, { "epoch": 0.7591066782307025, "grad_norm": 0.5258790254592896, "learning_rate": 7.496735818946759e-05, "loss": 0.0332, "step": 7002 }, { "epoch": 0.7592150910667823, "grad_norm": 0.30445805191993713, "learning_rate": 7.496373132163065e-05, "loss": 0.0347, "step": 7003 }, { "epoch": 0.7593235039028621, "grad_norm": 0.2799590229988098, "learning_rate": 7.49601044537937e-05, "loss": 0.0233, "step": 7004 }, { "epoch": 0.7594319167389418, "grad_norm": 0.22084598243236542, "learning_rate": 7.495647758595678e-05, "loss": 0.0255, "step": 7005 }, { "epoch": 0.7595403295750217, "grad_norm": 0.17118337750434875, "learning_rate": 7.495285071811983e-05, "loss": 0.0156, "step": 7006 }, { "epoch": 0.7596487424111015, "grad_norm": 0.19967439770698547, "learning_rate": 7.49492238502829e-05, "loss": 0.0137, "step": 7007 }, { "epoch": 0.7597571552471812, "grad_norm": 0.2750154137611389, "learning_rate": 7.494559698244596e-05, "loss": 0.0219, "step": 7008 }, { "epoch": 0.7598655680832611, "grad_norm": 0.4530794620513916, "learning_rate": 7.494197011460903e-05, "loss": 0.0794, "step": 7009 }, { "epoch": 0.7599739809193409, "grad_norm": 0.21658389270305634, "learning_rate": 7.493834324677209e-05, "loss": 0.016, "step": 7010 }, { "epoch": 0.7600823937554206, "grad_norm": 0.22015658020973206, "learning_rate": 7.493471637893516e-05, "loss": 0.0287, "step": 7011 }, { "epoch": 0.7601908065915004, "grad_norm": 0.3816288411617279, "learning_rate": 7.493108951109822e-05, "loss": 0.0153, "step": 7012 }, { "epoch": 0.7602992194275803, "grad_norm": 0.580112874507904, "learning_rate": 7.492746264326128e-05, "loss": 0.0728, "step": 7013 }, { "epoch": 0.76040763226366, "grad_norm": 0.4616205394268036, "learning_rate": 7.492383577542435e-05, "loss": 0.0317, "step": 7014 }, { "epoch": 0.7605160450997398, "grad_norm": 0.23420614004135132, "learning_rate": 7.49202089075874e-05, "loss": 0.0393, "step": 7015 }, { "epoch": 0.7606244579358196, "grad_norm": 0.3063719570636749, "learning_rate": 7.491658203975047e-05, "loss": 0.0476, "step": 7016 }, { "epoch": 0.7607328707718994, "grad_norm": 0.5364724397659302, "learning_rate": 7.491295517191353e-05, "loss": 0.0368, "step": 7017 }, { "epoch": 0.7608412836079792, "grad_norm": 0.2085946947336197, "learning_rate": 7.49093283040766e-05, "loss": 0.0096, "step": 7018 }, { "epoch": 0.760949696444059, "grad_norm": 0.36124274134635925, "learning_rate": 7.490570143623967e-05, "loss": 0.0277, "step": 7019 }, { "epoch": 0.7610581092801387, "grad_norm": 0.9059196710586548, "learning_rate": 7.490207456840273e-05, "loss": 0.0394, "step": 7020 }, { "epoch": 0.7611665221162186, "grad_norm": 0.4666779637336731, "learning_rate": 7.48984477005658e-05, "loss": 0.0275, "step": 7021 }, { "epoch": 0.7612749349522984, "grad_norm": 0.08168932795524597, "learning_rate": 7.489482083272886e-05, "loss": 0.014, "step": 7022 }, { "epoch": 0.7613833477883781, "grad_norm": 0.40417715907096863, "learning_rate": 7.489119396489193e-05, "loss": 0.0489, "step": 7023 }, { "epoch": 0.7614917606244579, "grad_norm": 0.24894165992736816, "learning_rate": 7.488756709705499e-05, "loss": 0.0165, "step": 7024 }, { "epoch": 0.7616001734605378, "grad_norm": 0.6091732978820801, "learning_rate": 7.488394022921806e-05, "loss": 0.0596, "step": 7025 }, { "epoch": 0.7617085862966175, "grad_norm": 0.6355217099189758, "learning_rate": 7.488031336138112e-05, "loss": 0.0688, "step": 7026 }, { "epoch": 0.7618169991326973, "grad_norm": 0.6702032089233398, "learning_rate": 7.487668649354417e-05, "loss": 0.0445, "step": 7027 }, { "epoch": 0.761925411968777, "grad_norm": 0.12571392953395844, "learning_rate": 7.487305962570724e-05, "loss": 0.0064, "step": 7028 }, { "epoch": 0.7620338248048569, "grad_norm": 0.6886851787567139, "learning_rate": 7.48694327578703e-05, "loss": 0.0933, "step": 7029 }, { "epoch": 0.7621422376409367, "grad_norm": 0.0675763264298439, "learning_rate": 7.486580589003337e-05, "loss": 0.0114, "step": 7030 }, { "epoch": 0.7622506504770165, "grad_norm": 0.43632394075393677, "learning_rate": 7.486217902219643e-05, "loss": 0.064, "step": 7031 }, { "epoch": 0.7623590633130962, "grad_norm": 1.3165295124053955, "learning_rate": 7.48585521543595e-05, "loss": 0.0448, "step": 7032 }, { "epoch": 0.7624674761491761, "grad_norm": 0.30719536542892456, "learning_rate": 7.485492528652256e-05, "loss": 0.0512, "step": 7033 }, { "epoch": 0.7625758889852559, "grad_norm": 0.3263193368911743, "learning_rate": 7.485129841868563e-05, "loss": 0.0649, "step": 7034 }, { "epoch": 0.7626843018213356, "grad_norm": 0.08481313288211823, "learning_rate": 7.484767155084869e-05, "loss": 0.0082, "step": 7035 }, { "epoch": 0.7627927146574154, "grad_norm": 0.3953388035297394, "learning_rate": 7.484404468301174e-05, "loss": 0.0519, "step": 7036 }, { "epoch": 0.7629011274934953, "grad_norm": 0.20591691136360168, "learning_rate": 7.484041781517483e-05, "loss": 0.0165, "step": 7037 }, { "epoch": 0.763009540329575, "grad_norm": 0.3103460371494293, "learning_rate": 7.483679094733788e-05, "loss": 0.0264, "step": 7038 }, { "epoch": 0.7631179531656548, "grad_norm": 0.30872562527656555, "learning_rate": 7.483316407950096e-05, "loss": 0.0269, "step": 7039 }, { "epoch": 0.7632263660017347, "grad_norm": 0.7539210319519043, "learning_rate": 7.482953721166401e-05, "loss": 0.0731, "step": 7040 }, { "epoch": 0.7633347788378144, "grad_norm": 0.6251130700111389, "learning_rate": 7.482591034382707e-05, "loss": 0.0573, "step": 7041 }, { "epoch": 0.7634431916738942, "grad_norm": 0.2589747905731201, "learning_rate": 7.482228347599014e-05, "loss": 0.0418, "step": 7042 }, { "epoch": 0.763551604509974, "grad_norm": 0.7574493288993835, "learning_rate": 7.48186566081532e-05, "loss": 0.0727, "step": 7043 }, { "epoch": 0.7636600173460538, "grad_norm": 0.1985381692647934, "learning_rate": 7.481502974031627e-05, "loss": 0.0217, "step": 7044 }, { "epoch": 0.7637684301821336, "grad_norm": 0.5321775078773499, "learning_rate": 7.481140287247933e-05, "loss": 0.1012, "step": 7045 }, { "epoch": 0.7638768430182133, "grad_norm": 0.3519481420516968, "learning_rate": 7.48077760046424e-05, "loss": 0.0509, "step": 7046 }, { "epoch": 0.7639852558542931, "grad_norm": 0.35434895753860474, "learning_rate": 7.480414913680546e-05, "loss": 0.0571, "step": 7047 }, { "epoch": 0.764093668690373, "grad_norm": 0.24519497156143188, "learning_rate": 7.480052226896853e-05, "loss": 0.0195, "step": 7048 }, { "epoch": 0.7642020815264527, "grad_norm": 0.15679879486560822, "learning_rate": 7.479689540113158e-05, "loss": 0.0346, "step": 7049 }, { "epoch": 0.7643104943625325, "grad_norm": 0.41777467727661133, "learning_rate": 7.479326853329464e-05, "loss": 0.0737, "step": 7050 }, { "epoch": 0.7644189071986123, "grad_norm": 0.43943560123443604, "learning_rate": 7.478964166545771e-05, "loss": 0.0663, "step": 7051 }, { "epoch": 0.7645273200346921, "grad_norm": 0.11695735901594162, "learning_rate": 7.478601479762077e-05, "loss": 0.0258, "step": 7052 }, { "epoch": 0.7646357328707719, "grad_norm": 0.2556397318840027, "learning_rate": 7.478238792978384e-05, "loss": 0.0241, "step": 7053 }, { "epoch": 0.7647441457068517, "grad_norm": 0.5872783660888672, "learning_rate": 7.477876106194691e-05, "loss": 0.0529, "step": 7054 }, { "epoch": 0.7648525585429314, "grad_norm": 0.2548779547214508, "learning_rate": 7.477513419410998e-05, "loss": 0.0408, "step": 7055 }, { "epoch": 0.7649609713790113, "grad_norm": 0.5855301022529602, "learning_rate": 7.477150732627304e-05, "loss": 0.0546, "step": 7056 }, { "epoch": 0.7650693842150911, "grad_norm": 0.14908398687839508, "learning_rate": 7.47678804584361e-05, "loss": 0.0257, "step": 7057 }, { "epoch": 0.7651777970511708, "grad_norm": 0.388949990272522, "learning_rate": 7.476425359059917e-05, "loss": 0.03, "step": 7058 }, { "epoch": 0.7652862098872506, "grad_norm": 0.24535733461380005, "learning_rate": 7.476062672276222e-05, "loss": 0.0174, "step": 7059 }, { "epoch": 0.7653946227233305, "grad_norm": 0.35340967774391174, "learning_rate": 7.47569998549253e-05, "loss": 0.0596, "step": 7060 }, { "epoch": 0.7655030355594102, "grad_norm": 0.22440582513809204, "learning_rate": 7.475337298708835e-05, "loss": 0.026, "step": 7061 }, { "epoch": 0.76561144839549, "grad_norm": 0.4181041717529297, "learning_rate": 7.474974611925142e-05, "loss": 0.0371, "step": 7062 }, { "epoch": 0.7657198612315698, "grad_norm": 0.4574912488460541, "learning_rate": 7.474611925141448e-05, "loss": 0.0876, "step": 7063 }, { "epoch": 0.7658282740676496, "grad_norm": 0.6636444926261902, "learning_rate": 7.474249238357754e-05, "loss": 0.0532, "step": 7064 }, { "epoch": 0.7659366869037294, "grad_norm": 0.11407391726970673, "learning_rate": 7.473886551574061e-05, "loss": 0.0146, "step": 7065 }, { "epoch": 0.7660450997398092, "grad_norm": 0.3031880855560303, "learning_rate": 7.473523864790367e-05, "loss": 0.0719, "step": 7066 }, { "epoch": 0.766153512575889, "grad_norm": 0.29055115580558777, "learning_rate": 7.473161178006674e-05, "loss": 0.0312, "step": 7067 }, { "epoch": 0.7662619254119688, "grad_norm": 0.5154882669448853, "learning_rate": 7.47279849122298e-05, "loss": 0.0448, "step": 7068 }, { "epoch": 0.7663703382480486, "grad_norm": 0.7223077416419983, "learning_rate": 7.472435804439287e-05, "loss": 0.0643, "step": 7069 }, { "epoch": 0.7664787510841283, "grad_norm": 0.24488858878612518, "learning_rate": 7.472073117655592e-05, "loss": 0.0285, "step": 7070 }, { "epoch": 0.7665871639202082, "grad_norm": 0.18223007023334503, "learning_rate": 7.4717104308719e-05, "loss": 0.0336, "step": 7071 }, { "epoch": 0.766695576756288, "grad_norm": 0.34622862935066223, "learning_rate": 7.471347744088206e-05, "loss": 0.0591, "step": 7072 }, { "epoch": 0.7668039895923677, "grad_norm": 0.2619072496891022, "learning_rate": 7.470985057304512e-05, "loss": 0.0432, "step": 7073 }, { "epoch": 0.7669124024284475, "grad_norm": 0.22132650017738342, "learning_rate": 7.470622370520819e-05, "loss": 0.0287, "step": 7074 }, { "epoch": 0.7670208152645274, "grad_norm": 0.4293718636035919, "learning_rate": 7.470259683737125e-05, "loss": 0.0487, "step": 7075 }, { "epoch": 0.7671292281006071, "grad_norm": 0.21703578531742096, "learning_rate": 7.469896996953432e-05, "loss": 0.0431, "step": 7076 }, { "epoch": 0.7672376409366869, "grad_norm": 0.45120012760162354, "learning_rate": 7.469534310169738e-05, "loss": 0.0464, "step": 7077 }, { "epoch": 0.7673460537727667, "grad_norm": 0.245408296585083, "learning_rate": 7.469171623386045e-05, "loss": 0.0359, "step": 7078 }, { "epoch": 0.7674544666088465, "grad_norm": 0.14564433693885803, "learning_rate": 7.46880893660235e-05, "loss": 0.0204, "step": 7079 }, { "epoch": 0.7675628794449263, "grad_norm": 0.06906106323003769, "learning_rate": 7.468446249818656e-05, "loss": 0.0051, "step": 7080 }, { "epoch": 0.7676712922810061, "grad_norm": 0.14917348325252533, "learning_rate": 7.468083563034963e-05, "loss": 0.0156, "step": 7081 }, { "epoch": 0.7677797051170858, "grad_norm": 0.22916537523269653, "learning_rate": 7.467720876251269e-05, "loss": 0.0165, "step": 7082 }, { "epoch": 0.7678881179531657, "grad_norm": 0.20941480994224548, "learning_rate": 7.467358189467576e-05, "loss": 0.028, "step": 7083 }, { "epoch": 0.7679965307892455, "grad_norm": 0.32024240493774414, "learning_rate": 7.466995502683882e-05, "loss": 0.0374, "step": 7084 }, { "epoch": 0.7681049436253252, "grad_norm": 0.8610240817070007, "learning_rate": 7.466632815900189e-05, "loss": 0.0378, "step": 7085 }, { "epoch": 0.768213356461405, "grad_norm": 0.3746137320995331, "learning_rate": 7.466270129116495e-05, "loss": 0.0682, "step": 7086 }, { "epoch": 0.7683217692974849, "grad_norm": 0.2574552595615387, "learning_rate": 7.465907442332802e-05, "loss": 0.0408, "step": 7087 }, { "epoch": 0.7684301821335646, "grad_norm": 0.21814389526844025, "learning_rate": 7.465544755549108e-05, "loss": 0.0136, "step": 7088 }, { "epoch": 0.7685385949696444, "grad_norm": 0.33759021759033203, "learning_rate": 7.465182068765415e-05, "loss": 0.0438, "step": 7089 }, { "epoch": 0.7686470078057241, "grad_norm": 0.20183609426021576, "learning_rate": 7.464819381981722e-05, "loss": 0.0175, "step": 7090 }, { "epoch": 0.768755420641804, "grad_norm": 0.35024526715278625, "learning_rate": 7.464456695198028e-05, "loss": 0.0314, "step": 7091 }, { "epoch": 0.7688638334778838, "grad_norm": 0.24890194833278656, "learning_rate": 7.464094008414335e-05, "loss": 0.0451, "step": 7092 }, { "epoch": 0.7689722463139635, "grad_norm": 0.2468639761209488, "learning_rate": 7.46373132163064e-05, "loss": 0.0228, "step": 7093 }, { "epoch": 0.7690806591500433, "grad_norm": 0.6086708307266235, "learning_rate": 7.463368634846946e-05, "loss": 0.041, "step": 7094 }, { "epoch": 0.7691890719861232, "grad_norm": 0.2792736887931824, "learning_rate": 7.463005948063253e-05, "loss": 0.0316, "step": 7095 }, { "epoch": 0.769297484822203, "grad_norm": 0.4731099605560303, "learning_rate": 7.462643261279559e-05, "loss": 0.0462, "step": 7096 }, { "epoch": 0.7694058976582827, "grad_norm": 0.2844598591327667, "learning_rate": 7.462280574495866e-05, "loss": 0.0236, "step": 7097 }, { "epoch": 0.7695143104943626, "grad_norm": 1.0460420846939087, "learning_rate": 7.461917887712172e-05, "loss": 0.034, "step": 7098 }, { "epoch": 0.7696227233304423, "grad_norm": 0.49694108963012695, "learning_rate": 7.461555200928479e-05, "loss": 0.0842, "step": 7099 }, { "epoch": 0.7697311361665221, "grad_norm": 0.7384533286094666, "learning_rate": 7.461192514144785e-05, "loss": 0.0436, "step": 7100 }, { "epoch": 0.7698395490026019, "grad_norm": 0.30250462889671326, "learning_rate": 7.460829827361092e-05, "loss": 0.0212, "step": 7101 }, { "epoch": 0.7699479618386817, "grad_norm": 0.942398726940155, "learning_rate": 7.460467140577397e-05, "loss": 0.0691, "step": 7102 }, { "epoch": 0.7700563746747615, "grad_norm": 0.4984097182750702, "learning_rate": 7.460104453793703e-05, "loss": 0.0651, "step": 7103 }, { "epoch": 0.7701647875108413, "grad_norm": 0.08207714557647705, "learning_rate": 7.45974176701001e-05, "loss": 0.0079, "step": 7104 }, { "epoch": 0.770273200346921, "grad_norm": 0.5222489833831787, "learning_rate": 7.459379080226316e-05, "loss": 0.043, "step": 7105 }, { "epoch": 0.7703816131830009, "grad_norm": 0.5929635763168335, "learning_rate": 7.459016393442624e-05, "loss": 0.0273, "step": 7106 }, { "epoch": 0.7704900260190807, "grad_norm": 0.1264401376247406, "learning_rate": 7.45865370665893e-05, "loss": 0.0108, "step": 7107 }, { "epoch": 0.7705984388551604, "grad_norm": 0.25873976945877075, "learning_rate": 7.458291019875236e-05, "loss": 0.0355, "step": 7108 }, { "epoch": 0.7707068516912402, "grad_norm": 0.4586288034915924, "learning_rate": 7.457928333091543e-05, "loss": 0.0297, "step": 7109 }, { "epoch": 0.7708152645273201, "grad_norm": 0.80437171459198, "learning_rate": 7.457565646307849e-05, "loss": 0.1145, "step": 7110 }, { "epoch": 0.7709236773633998, "grad_norm": 0.4535684883594513, "learning_rate": 7.457202959524156e-05, "loss": 0.0477, "step": 7111 }, { "epoch": 0.7710320901994796, "grad_norm": 0.19769185781478882, "learning_rate": 7.456840272740462e-05, "loss": 0.0257, "step": 7112 }, { "epoch": 0.7711405030355594, "grad_norm": 0.5574365258216858, "learning_rate": 7.456477585956769e-05, "loss": 0.0571, "step": 7113 }, { "epoch": 0.7712489158716392, "grad_norm": 0.500827968120575, "learning_rate": 7.456114899173074e-05, "loss": 0.0429, "step": 7114 }, { "epoch": 0.771357328707719, "grad_norm": 0.3971251845359802, "learning_rate": 7.455752212389381e-05, "loss": 0.0535, "step": 7115 }, { "epoch": 0.7714657415437988, "grad_norm": 0.2722477912902832, "learning_rate": 7.455389525605687e-05, "loss": 0.0352, "step": 7116 }, { "epoch": 0.7715741543798785, "grad_norm": 0.8734215497970581, "learning_rate": 7.455026838821993e-05, "loss": 0.0463, "step": 7117 }, { "epoch": 0.7716825672159584, "grad_norm": 0.21128477156162262, "learning_rate": 7.4546641520383e-05, "loss": 0.0284, "step": 7118 }, { "epoch": 0.7717909800520382, "grad_norm": 0.38558706641197205, "learning_rate": 7.454301465254606e-05, "loss": 0.052, "step": 7119 }, { "epoch": 0.7718993928881179, "grad_norm": 0.3689645230770111, "learning_rate": 7.453938778470913e-05, "loss": 0.0364, "step": 7120 }, { "epoch": 0.7720078057241977, "grad_norm": 0.609524130821228, "learning_rate": 7.453576091687219e-05, "loss": 0.0769, "step": 7121 }, { "epoch": 0.7721162185602776, "grad_norm": 0.1704222559928894, "learning_rate": 7.453213404903526e-05, "loss": 0.0268, "step": 7122 }, { "epoch": 0.7722246313963573, "grad_norm": 0.0930120125412941, "learning_rate": 7.452850718119833e-05, "loss": 0.0154, "step": 7123 }, { "epoch": 0.7723330442324371, "grad_norm": 0.21522697806358337, "learning_rate": 7.452488031336138e-05, "loss": 0.021, "step": 7124 }, { "epoch": 0.772441457068517, "grad_norm": 0.45823779702186584, "learning_rate": 7.452125344552446e-05, "loss": 0.0318, "step": 7125 }, { "epoch": 0.7725498699045967, "grad_norm": 0.2459833025932312, "learning_rate": 7.451762657768751e-05, "loss": 0.0183, "step": 7126 }, { "epoch": 0.7726582827406765, "grad_norm": 0.2281336486339569, "learning_rate": 7.451399970985058e-05, "loss": 0.0336, "step": 7127 }, { "epoch": 0.7727666955767563, "grad_norm": 0.23493137955665588, "learning_rate": 7.451037284201364e-05, "loss": 0.0382, "step": 7128 }, { "epoch": 0.7728751084128361, "grad_norm": 0.13222041726112366, "learning_rate": 7.450674597417671e-05, "loss": 0.0122, "step": 7129 }, { "epoch": 0.7729835212489159, "grad_norm": 0.23927484452724457, "learning_rate": 7.450311910633977e-05, "loss": 0.0238, "step": 7130 }, { "epoch": 0.7730919340849957, "grad_norm": 0.23027396202087402, "learning_rate": 7.449949223850283e-05, "loss": 0.0343, "step": 7131 }, { "epoch": 0.7732003469210754, "grad_norm": 0.36334237456321716, "learning_rate": 7.44958653706659e-05, "loss": 0.0453, "step": 7132 }, { "epoch": 0.7733087597571553, "grad_norm": 0.2412133663892746, "learning_rate": 7.449223850282895e-05, "loss": 0.0191, "step": 7133 }, { "epoch": 0.7734171725932351, "grad_norm": 0.3175528645515442, "learning_rate": 7.448861163499203e-05, "loss": 0.0141, "step": 7134 }, { "epoch": 0.7735255854293148, "grad_norm": 0.5424768328666687, "learning_rate": 7.448498476715508e-05, "loss": 0.0474, "step": 7135 }, { "epoch": 0.7736339982653946, "grad_norm": 0.3340384364128113, "learning_rate": 7.448135789931815e-05, "loss": 0.0302, "step": 7136 }, { "epoch": 0.7737424111014745, "grad_norm": 0.3679755926132202, "learning_rate": 7.447773103148121e-05, "loss": 0.0542, "step": 7137 }, { "epoch": 0.7738508239375542, "grad_norm": 0.28397414088249207, "learning_rate": 7.447410416364428e-05, "loss": 0.0408, "step": 7138 }, { "epoch": 0.773959236773634, "grad_norm": 0.2691001296043396, "learning_rate": 7.447047729580734e-05, "loss": 0.0382, "step": 7139 }, { "epoch": 0.7740676496097137, "grad_norm": 0.44023168087005615, "learning_rate": 7.446685042797041e-05, "loss": 0.0472, "step": 7140 }, { "epoch": 0.7741760624457936, "grad_norm": 0.1794734001159668, "learning_rate": 7.446322356013348e-05, "loss": 0.0205, "step": 7141 }, { "epoch": 0.7742844752818734, "grad_norm": 0.6303317546844482, "learning_rate": 7.445959669229654e-05, "loss": 0.0388, "step": 7142 }, { "epoch": 0.7743928881179531, "grad_norm": 0.18269819021224976, "learning_rate": 7.445596982445961e-05, "loss": 0.0184, "step": 7143 }, { "epoch": 0.7745013009540329, "grad_norm": 0.2719869911670685, "learning_rate": 7.445234295662267e-05, "loss": 0.0188, "step": 7144 }, { "epoch": 0.7746097137901128, "grad_norm": 0.3202539384365082, "learning_rate": 7.444871608878572e-05, "loss": 0.0417, "step": 7145 }, { "epoch": 0.7747181266261925, "grad_norm": 0.23747390508651733, "learning_rate": 7.44450892209488e-05, "loss": 0.0286, "step": 7146 }, { "epoch": 0.7748265394622723, "grad_norm": 0.3530946969985962, "learning_rate": 7.444146235311185e-05, "loss": 0.0417, "step": 7147 }, { "epoch": 0.7749349522983521, "grad_norm": 0.4968096613883972, "learning_rate": 7.443783548527492e-05, "loss": 0.0613, "step": 7148 }, { "epoch": 0.775043365134432, "grad_norm": 0.2886918783187866, "learning_rate": 7.443420861743798e-05, "loss": 0.0388, "step": 7149 }, { "epoch": 0.7751517779705117, "grad_norm": 0.3434288501739502, "learning_rate": 7.443058174960105e-05, "loss": 0.0209, "step": 7150 }, { "epoch": 0.7752601908065915, "grad_norm": 0.32596147060394287, "learning_rate": 7.442695488176411e-05, "loss": 0.03, "step": 7151 }, { "epoch": 0.7753686036426712, "grad_norm": 0.24297794699668884, "learning_rate": 7.442332801392718e-05, "loss": 0.0167, "step": 7152 }, { "epoch": 0.7754770164787511, "grad_norm": 0.6473504304885864, "learning_rate": 7.441970114609024e-05, "loss": 0.0719, "step": 7153 }, { "epoch": 0.7755854293148309, "grad_norm": 0.5481699109077454, "learning_rate": 7.44160742782533e-05, "loss": 0.071, "step": 7154 }, { "epoch": 0.7756938421509106, "grad_norm": 0.18406489491462708, "learning_rate": 7.441244741041636e-05, "loss": 0.0117, "step": 7155 }, { "epoch": 0.7758022549869905, "grad_norm": 0.2500079870223999, "learning_rate": 7.440882054257942e-05, "loss": 0.0241, "step": 7156 }, { "epoch": 0.7759106678230703, "grad_norm": 0.10877682268619537, "learning_rate": 7.440519367474249e-05, "loss": 0.0169, "step": 7157 }, { "epoch": 0.77601908065915, "grad_norm": 0.22869496047496796, "learning_rate": 7.440156680690556e-05, "loss": 0.0219, "step": 7158 }, { "epoch": 0.7761274934952298, "grad_norm": 0.2800387144088745, "learning_rate": 7.439793993906863e-05, "loss": 0.0303, "step": 7159 }, { "epoch": 0.7762359063313097, "grad_norm": 0.7687073945999146, "learning_rate": 7.439431307123169e-05, "loss": 0.0499, "step": 7160 }, { "epoch": 0.7763443191673894, "grad_norm": 0.4226357042789459, "learning_rate": 7.439068620339475e-05, "loss": 0.0683, "step": 7161 }, { "epoch": 0.7764527320034692, "grad_norm": 0.3095313608646393, "learning_rate": 7.438705933555782e-05, "loss": 0.0245, "step": 7162 }, { "epoch": 0.776561144839549, "grad_norm": 0.24501889944076538, "learning_rate": 7.438343246772088e-05, "loss": 0.024, "step": 7163 }, { "epoch": 0.7766695576756288, "grad_norm": 0.1266660839319229, "learning_rate": 7.437980559988395e-05, "loss": 0.0145, "step": 7164 }, { "epoch": 0.7767779705117086, "grad_norm": 0.43543073534965515, "learning_rate": 7.4376178732047e-05, "loss": 0.0605, "step": 7165 }, { "epoch": 0.7768863833477884, "grad_norm": 0.8212564587593079, "learning_rate": 7.437255186421008e-05, "loss": 0.027, "step": 7166 }, { "epoch": 0.7769947961838681, "grad_norm": 0.43951427936553955, "learning_rate": 7.436892499637313e-05, "loss": 0.0238, "step": 7167 }, { "epoch": 0.777103209019948, "grad_norm": 0.3394080698490143, "learning_rate": 7.436529812853619e-05, "loss": 0.0383, "step": 7168 }, { "epoch": 0.7772116218560278, "grad_norm": 0.3484438359737396, "learning_rate": 7.436167126069926e-05, "loss": 0.041, "step": 7169 }, { "epoch": 0.7773200346921075, "grad_norm": 0.39329463243484497, "learning_rate": 7.435804439286232e-05, "loss": 0.0616, "step": 7170 }, { "epoch": 0.7774284475281873, "grad_norm": 0.5697999000549316, "learning_rate": 7.435441752502539e-05, "loss": 0.0743, "step": 7171 }, { "epoch": 0.7775368603642672, "grad_norm": 0.29390662908554077, "learning_rate": 7.435079065718845e-05, "loss": 0.0306, "step": 7172 }, { "epoch": 0.7776452732003469, "grad_norm": 0.21425026655197144, "learning_rate": 7.434716378935152e-05, "loss": 0.0244, "step": 7173 }, { "epoch": 0.7777536860364267, "grad_norm": 0.1959531605243683, "learning_rate": 7.434353692151458e-05, "loss": 0.0157, "step": 7174 }, { "epoch": 0.7778620988725065, "grad_norm": 0.14934784173965454, "learning_rate": 7.433991005367765e-05, "loss": 0.0236, "step": 7175 }, { "epoch": 0.7779705117085863, "grad_norm": 0.6052179336547852, "learning_rate": 7.433628318584072e-05, "loss": 0.0674, "step": 7176 }, { "epoch": 0.7780789245446661, "grad_norm": 0.17120645940303802, "learning_rate": 7.433265631800378e-05, "loss": 0.0207, "step": 7177 }, { "epoch": 0.7781873373807459, "grad_norm": 0.39745432138442993, "learning_rate": 7.432902945016685e-05, "loss": 0.0205, "step": 7178 }, { "epoch": 0.7782957502168256, "grad_norm": 0.4664956331253052, "learning_rate": 7.43254025823299e-05, "loss": 0.0446, "step": 7179 }, { "epoch": 0.7784041630529055, "grad_norm": 0.3560604453086853, "learning_rate": 7.432177571449297e-05, "loss": 0.071, "step": 7180 }, { "epoch": 0.7785125758889853, "grad_norm": 0.19513922929763794, "learning_rate": 7.431814884665603e-05, "loss": 0.033, "step": 7181 }, { "epoch": 0.778620988725065, "grad_norm": 0.2709694802761078, "learning_rate": 7.43145219788191e-05, "loss": 0.0246, "step": 7182 }, { "epoch": 0.7787294015611448, "grad_norm": 0.3502906858921051, "learning_rate": 7.431089511098216e-05, "loss": 0.0421, "step": 7183 }, { "epoch": 0.7788378143972247, "grad_norm": 0.1140102744102478, "learning_rate": 7.430726824314522e-05, "loss": 0.0126, "step": 7184 }, { "epoch": 0.7789462272333044, "grad_norm": 0.22621113061904907, "learning_rate": 7.430364137530829e-05, "loss": 0.0286, "step": 7185 }, { "epoch": 0.7790546400693842, "grad_norm": 0.40052497386932373, "learning_rate": 7.430001450747135e-05, "loss": 0.0683, "step": 7186 }, { "epoch": 0.7791630529054641, "grad_norm": 0.15371030569076538, "learning_rate": 7.429638763963442e-05, "loss": 0.0162, "step": 7187 }, { "epoch": 0.7792714657415438, "grad_norm": 0.30838271975517273, "learning_rate": 7.429276077179747e-05, "loss": 0.0289, "step": 7188 }, { "epoch": 0.7793798785776236, "grad_norm": 0.8547186255455017, "learning_rate": 7.428913390396054e-05, "loss": 0.0519, "step": 7189 }, { "epoch": 0.7794882914137033, "grad_norm": 0.299966424703598, "learning_rate": 7.42855070361236e-05, "loss": 0.0289, "step": 7190 }, { "epoch": 0.7795967042497832, "grad_norm": 0.17029233276844025, "learning_rate": 7.428188016828667e-05, "loss": 0.018, "step": 7191 }, { "epoch": 0.779705117085863, "grad_norm": 0.729476273059845, "learning_rate": 7.427825330044974e-05, "loss": 0.0306, "step": 7192 }, { "epoch": 0.7798135299219427, "grad_norm": 0.47699543833732605, "learning_rate": 7.42746264326128e-05, "loss": 0.0761, "step": 7193 }, { "epoch": 0.7799219427580225, "grad_norm": 0.6549237370491028, "learning_rate": 7.427099956477587e-05, "loss": 0.0241, "step": 7194 }, { "epoch": 0.7800303555941024, "grad_norm": 0.2768939137458801, "learning_rate": 7.426737269693893e-05, "loss": 0.034, "step": 7195 }, { "epoch": 0.7801387684301822, "grad_norm": 0.2881155014038086, "learning_rate": 7.4263745829102e-05, "loss": 0.0429, "step": 7196 }, { "epoch": 0.7802471812662619, "grad_norm": 0.31461820006370544, "learning_rate": 7.426011896126506e-05, "loss": 0.0227, "step": 7197 }, { "epoch": 0.7803555941023417, "grad_norm": 0.2944423258304596, "learning_rate": 7.425649209342811e-05, "loss": 0.0322, "step": 7198 }, { "epoch": 0.7804640069384216, "grad_norm": 0.5422961711883545, "learning_rate": 7.425286522559119e-05, "loss": 0.0333, "step": 7199 }, { "epoch": 0.7805724197745013, "grad_norm": 0.603768527507782, "learning_rate": 7.424923835775424e-05, "loss": 0.0568, "step": 7200 }, { "epoch": 0.7806808326105811, "grad_norm": 0.22290576994419098, "learning_rate": 7.424561148991731e-05, "loss": 0.0334, "step": 7201 }, { "epoch": 0.7807892454466608, "grad_norm": 0.2255975753068924, "learning_rate": 7.424198462208037e-05, "loss": 0.0159, "step": 7202 }, { "epoch": 0.7808976582827407, "grad_norm": 0.5073437690734863, "learning_rate": 7.423835775424344e-05, "loss": 0.0437, "step": 7203 }, { "epoch": 0.7810060711188205, "grad_norm": 0.45269694924354553, "learning_rate": 7.42347308864065e-05, "loss": 0.0351, "step": 7204 }, { "epoch": 0.7811144839549002, "grad_norm": 0.13559989631175995, "learning_rate": 7.423110401856957e-05, "loss": 0.0157, "step": 7205 }, { "epoch": 0.78122289679098, "grad_norm": 0.2809396982192993, "learning_rate": 7.422747715073263e-05, "loss": 0.0144, "step": 7206 }, { "epoch": 0.7813313096270599, "grad_norm": 0.41694262623786926, "learning_rate": 7.422385028289568e-05, "loss": 0.0775, "step": 7207 }, { "epoch": 0.7814397224631396, "grad_norm": 1.0596554279327393, "learning_rate": 7.422022341505876e-05, "loss": 0.0778, "step": 7208 }, { "epoch": 0.7815481352992194, "grad_norm": 0.5451157689094543, "learning_rate": 7.421659654722183e-05, "loss": 0.0685, "step": 7209 }, { "epoch": 0.7816565481352992, "grad_norm": 0.45737722516059875, "learning_rate": 7.42129696793849e-05, "loss": 0.052, "step": 7210 }, { "epoch": 0.781764960971379, "grad_norm": 0.20584215223789215, "learning_rate": 7.420934281154795e-05, "loss": 0.0128, "step": 7211 }, { "epoch": 0.7818733738074588, "grad_norm": 0.19273388385772705, "learning_rate": 7.420571594371101e-05, "loss": 0.0245, "step": 7212 }, { "epoch": 0.7819817866435386, "grad_norm": 0.38246384263038635, "learning_rate": 7.420208907587408e-05, "loss": 0.0531, "step": 7213 }, { "epoch": 0.7820901994796184, "grad_norm": 0.17585623264312744, "learning_rate": 7.419846220803714e-05, "loss": 0.0248, "step": 7214 }, { "epoch": 0.7821986123156982, "grad_norm": 0.13533660769462585, "learning_rate": 7.419483534020021e-05, "loss": 0.017, "step": 7215 }, { "epoch": 0.782307025151778, "grad_norm": 0.17110575735569, "learning_rate": 7.419120847236327e-05, "loss": 0.0199, "step": 7216 }, { "epoch": 0.7824154379878577, "grad_norm": 0.44384825229644775, "learning_rate": 7.418758160452634e-05, "loss": 0.0288, "step": 7217 }, { "epoch": 0.7825238508239376, "grad_norm": 0.2930488586425781, "learning_rate": 7.41839547366894e-05, "loss": 0.0316, "step": 7218 }, { "epoch": 0.7826322636600174, "grad_norm": 0.6124654412269592, "learning_rate": 7.418032786885247e-05, "loss": 0.0431, "step": 7219 }, { "epoch": 0.7827406764960971, "grad_norm": 0.11760212481021881, "learning_rate": 7.417670100101552e-05, "loss": 0.0099, "step": 7220 }, { "epoch": 0.7828490893321769, "grad_norm": 0.2760685086250305, "learning_rate": 7.417307413317858e-05, "loss": 0.0447, "step": 7221 }, { "epoch": 0.7829575021682568, "grad_norm": 0.13917142152786255, "learning_rate": 7.416944726534165e-05, "loss": 0.0217, "step": 7222 }, { "epoch": 0.7830659150043365, "grad_norm": 0.3444758355617523, "learning_rate": 7.416582039750471e-05, "loss": 0.0559, "step": 7223 }, { "epoch": 0.7831743278404163, "grad_norm": 0.3594372570514679, "learning_rate": 7.416219352966778e-05, "loss": 0.0431, "step": 7224 }, { "epoch": 0.7832827406764961, "grad_norm": 0.2872641682624817, "learning_rate": 7.415856666183084e-05, "loss": 0.0435, "step": 7225 }, { "epoch": 0.7833911535125759, "grad_norm": 0.26364120841026306, "learning_rate": 7.415493979399391e-05, "loss": 0.0206, "step": 7226 }, { "epoch": 0.7834995663486557, "grad_norm": 0.3484843671321869, "learning_rate": 7.415131292615698e-05, "loss": 0.0477, "step": 7227 }, { "epoch": 0.7836079791847355, "grad_norm": 0.5383736491203308, "learning_rate": 7.414768605832004e-05, "loss": 0.0519, "step": 7228 }, { "epoch": 0.7837163920208152, "grad_norm": 0.29278305172920227, "learning_rate": 7.414405919048311e-05, "loss": 0.0381, "step": 7229 }, { "epoch": 0.7838248048568951, "grad_norm": 0.2837611734867096, "learning_rate": 7.414043232264617e-05, "loss": 0.0354, "step": 7230 }, { "epoch": 0.7839332176929749, "grad_norm": 0.42589420080184937, "learning_rate": 7.413680545480924e-05, "loss": 0.0204, "step": 7231 }, { "epoch": 0.7840416305290546, "grad_norm": 0.2606181800365448, "learning_rate": 7.41331785869723e-05, "loss": 0.0223, "step": 7232 }, { "epoch": 0.7841500433651344, "grad_norm": 0.15869085490703583, "learning_rate": 7.412955171913537e-05, "loss": 0.0172, "step": 7233 }, { "epoch": 0.7842584562012143, "grad_norm": 0.08185327798128128, "learning_rate": 7.412592485129842e-05, "loss": 0.0117, "step": 7234 }, { "epoch": 0.784366869037294, "grad_norm": 0.45643821358680725, "learning_rate": 7.412229798346148e-05, "loss": 0.0529, "step": 7235 }, { "epoch": 0.7844752818733738, "grad_norm": 0.17542262375354767, "learning_rate": 7.411867111562455e-05, "loss": 0.014, "step": 7236 }, { "epoch": 0.7845836947094535, "grad_norm": 0.1350775808095932, "learning_rate": 7.411504424778761e-05, "loss": 0.014, "step": 7237 }, { "epoch": 0.7846921075455334, "grad_norm": 0.3278946876525879, "learning_rate": 7.411141737995068e-05, "loss": 0.0204, "step": 7238 }, { "epoch": 0.7848005203816132, "grad_norm": 0.4759754538536072, "learning_rate": 7.410779051211374e-05, "loss": 0.0612, "step": 7239 }, { "epoch": 0.784908933217693, "grad_norm": 0.4207131862640381, "learning_rate": 7.410416364427681e-05, "loss": 0.0503, "step": 7240 }, { "epoch": 0.7850173460537727, "grad_norm": 0.598156750202179, "learning_rate": 7.410053677643986e-05, "loss": 0.0449, "step": 7241 }, { "epoch": 0.7851257588898526, "grad_norm": 0.12226894497871399, "learning_rate": 7.409690990860294e-05, "loss": 0.0091, "step": 7242 }, { "epoch": 0.7852341717259324, "grad_norm": 0.6434886455535889, "learning_rate": 7.409328304076599e-05, "loss": 0.0816, "step": 7243 }, { "epoch": 0.7853425845620121, "grad_norm": 0.18139223754405975, "learning_rate": 7.408965617292906e-05, "loss": 0.0224, "step": 7244 }, { "epoch": 0.785450997398092, "grad_norm": 0.26628509163856506, "learning_rate": 7.408602930509213e-05, "loss": 0.0155, "step": 7245 }, { "epoch": 0.7855594102341718, "grad_norm": 0.24155397713184357, "learning_rate": 7.408240243725519e-05, "loss": 0.0353, "step": 7246 }, { "epoch": 0.7856678230702515, "grad_norm": 0.2782159745693207, "learning_rate": 7.407877556941826e-05, "loss": 0.0287, "step": 7247 }, { "epoch": 0.7857762359063313, "grad_norm": 0.37496861815452576, "learning_rate": 7.407514870158132e-05, "loss": 0.0454, "step": 7248 }, { "epoch": 0.7858846487424112, "grad_norm": 0.2918858230113983, "learning_rate": 7.407152183374438e-05, "loss": 0.0501, "step": 7249 }, { "epoch": 0.7859930615784909, "grad_norm": 0.4293577969074249, "learning_rate": 7.406789496590745e-05, "loss": 0.0417, "step": 7250 }, { "epoch": 0.7861014744145707, "grad_norm": 0.33301085233688354, "learning_rate": 7.40642680980705e-05, "loss": 0.0364, "step": 7251 }, { "epoch": 0.7862098872506504, "grad_norm": 0.521061360836029, "learning_rate": 7.406064123023358e-05, "loss": 0.0534, "step": 7252 }, { "epoch": 0.7863183000867303, "grad_norm": 0.6612648367881775, "learning_rate": 7.405701436239663e-05, "loss": 0.0945, "step": 7253 }, { "epoch": 0.7864267129228101, "grad_norm": 0.038478344678878784, "learning_rate": 7.40533874945597e-05, "loss": 0.0035, "step": 7254 }, { "epoch": 0.7865351257588898, "grad_norm": 0.3602342903614044, "learning_rate": 7.404976062672276e-05, "loss": 0.0531, "step": 7255 }, { "epoch": 0.7866435385949696, "grad_norm": 0.640667736530304, "learning_rate": 7.404613375888583e-05, "loss": 0.0659, "step": 7256 }, { "epoch": 0.7867519514310495, "grad_norm": 0.2513673007488251, "learning_rate": 7.404250689104889e-05, "loss": 0.0369, "step": 7257 }, { "epoch": 0.7868603642671292, "grad_norm": 1.1986446380615234, "learning_rate": 7.403888002321195e-05, "loss": 0.0593, "step": 7258 }, { "epoch": 0.786968777103209, "grad_norm": 0.11764393001794815, "learning_rate": 7.403525315537502e-05, "loss": 0.0176, "step": 7259 }, { "epoch": 0.7870771899392888, "grad_norm": 0.3881261944770813, "learning_rate": 7.403162628753808e-05, "loss": 0.0918, "step": 7260 }, { "epoch": 0.7871856027753686, "grad_norm": 0.12952472269535065, "learning_rate": 7.402799941970116e-05, "loss": 0.0144, "step": 7261 }, { "epoch": 0.7872940156114484, "grad_norm": 0.6436315774917603, "learning_rate": 7.402437255186422e-05, "loss": 0.0651, "step": 7262 }, { "epoch": 0.7874024284475282, "grad_norm": 0.3623707890510559, "learning_rate": 7.402074568402729e-05, "loss": 0.0414, "step": 7263 }, { "epoch": 0.7875108412836079, "grad_norm": 0.5259140133857727, "learning_rate": 7.401711881619035e-05, "loss": 0.0665, "step": 7264 }, { "epoch": 0.7876192541196878, "grad_norm": 0.3023720681667328, "learning_rate": 7.40134919483534e-05, "loss": 0.0386, "step": 7265 }, { "epoch": 0.7877276669557676, "grad_norm": 0.5458096861839294, "learning_rate": 7.400986508051647e-05, "loss": 0.0177, "step": 7266 }, { "epoch": 0.7878360797918473, "grad_norm": 0.38717594742774963, "learning_rate": 7.400623821267953e-05, "loss": 0.044, "step": 7267 }, { "epoch": 0.7879444926279271, "grad_norm": 0.4177587926387787, "learning_rate": 7.40026113448426e-05, "loss": 0.0787, "step": 7268 }, { "epoch": 0.788052905464007, "grad_norm": 0.406723290681839, "learning_rate": 7.399898447700566e-05, "loss": 0.0269, "step": 7269 }, { "epoch": 0.7881613183000867, "grad_norm": 0.1359550654888153, "learning_rate": 7.399535760916873e-05, "loss": 0.0105, "step": 7270 }, { "epoch": 0.7882697311361665, "grad_norm": 0.18809296190738678, "learning_rate": 7.399173074133179e-05, "loss": 0.0269, "step": 7271 }, { "epoch": 0.7883781439722463, "grad_norm": 0.20454269647598267, "learning_rate": 7.398810387349486e-05, "loss": 0.0373, "step": 7272 }, { "epoch": 0.7884865568083261, "grad_norm": 0.1561601608991623, "learning_rate": 7.398447700565792e-05, "loss": 0.0248, "step": 7273 }, { "epoch": 0.7885949696444059, "grad_norm": 0.46235495805740356, "learning_rate": 7.398085013782097e-05, "loss": 0.0852, "step": 7274 }, { "epoch": 0.7887033824804857, "grad_norm": 0.31504306197166443, "learning_rate": 7.397722326998404e-05, "loss": 0.0287, "step": 7275 }, { "epoch": 0.7888117953165655, "grad_norm": 0.17338570952415466, "learning_rate": 7.39735964021471e-05, "loss": 0.0281, "step": 7276 }, { "epoch": 0.7889202081526453, "grad_norm": 0.23081882297992706, "learning_rate": 7.396996953431017e-05, "loss": 0.037, "step": 7277 }, { "epoch": 0.7890286209887251, "grad_norm": 0.23037871718406677, "learning_rate": 7.396634266647323e-05, "loss": 0.0223, "step": 7278 }, { "epoch": 0.7891370338248048, "grad_norm": 0.2216407209634781, "learning_rate": 7.39627157986363e-05, "loss": 0.022, "step": 7279 }, { "epoch": 0.7892454466608847, "grad_norm": 0.25289374589920044, "learning_rate": 7.395908893079937e-05, "loss": 0.0556, "step": 7280 }, { "epoch": 0.7893538594969645, "grad_norm": 0.15181827545166016, "learning_rate": 7.395546206296243e-05, "loss": 0.0167, "step": 7281 }, { "epoch": 0.7894622723330442, "grad_norm": 0.20733609795570374, "learning_rate": 7.39518351951255e-05, "loss": 0.0155, "step": 7282 }, { "epoch": 0.789570685169124, "grad_norm": 0.26906126737594604, "learning_rate": 7.394820832728856e-05, "loss": 0.0362, "step": 7283 }, { "epoch": 0.7896790980052039, "grad_norm": 0.20786511898040771, "learning_rate": 7.394458145945163e-05, "loss": 0.0145, "step": 7284 }, { "epoch": 0.7897875108412836, "grad_norm": 0.5623252391815186, "learning_rate": 7.394095459161469e-05, "loss": 0.051, "step": 7285 }, { "epoch": 0.7898959236773634, "grad_norm": 0.26715511083602905, "learning_rate": 7.393732772377776e-05, "loss": 0.0349, "step": 7286 }, { "epoch": 0.7900043365134432, "grad_norm": 0.21009916067123413, "learning_rate": 7.393370085594081e-05, "loss": 0.0381, "step": 7287 }, { "epoch": 0.790112749349523, "grad_norm": 0.15761931240558624, "learning_rate": 7.393007398810387e-05, "loss": 0.0256, "step": 7288 }, { "epoch": 0.7902211621856028, "grad_norm": 0.3317476212978363, "learning_rate": 7.392644712026694e-05, "loss": 0.0156, "step": 7289 }, { "epoch": 0.7903295750216826, "grad_norm": 0.3093043863773346, "learning_rate": 7.392282025243e-05, "loss": 0.0186, "step": 7290 }, { "epoch": 0.7904379878577623, "grad_norm": 0.3263915479183197, "learning_rate": 7.391919338459307e-05, "loss": 0.0341, "step": 7291 }, { "epoch": 0.7905464006938422, "grad_norm": 0.2564474642276764, "learning_rate": 7.391556651675613e-05, "loss": 0.0379, "step": 7292 }, { "epoch": 0.790654813529922, "grad_norm": 0.5123199820518494, "learning_rate": 7.39119396489192e-05, "loss": 0.0728, "step": 7293 }, { "epoch": 0.7907632263660017, "grad_norm": 0.3252577781677246, "learning_rate": 7.390831278108226e-05, "loss": 0.032, "step": 7294 }, { "epoch": 0.7908716392020815, "grad_norm": 0.30115190148353577, "learning_rate": 7.390468591324533e-05, "loss": 0.0316, "step": 7295 }, { "epoch": 0.7909800520381614, "grad_norm": 0.06517176330089569, "learning_rate": 7.39010590454084e-05, "loss": 0.0127, "step": 7296 }, { "epoch": 0.7910884648742411, "grad_norm": 0.26437196135520935, "learning_rate": 7.389743217757145e-05, "loss": 0.0235, "step": 7297 }, { "epoch": 0.7911968777103209, "grad_norm": 0.24685029685497284, "learning_rate": 7.389380530973453e-05, "loss": 0.0152, "step": 7298 }, { "epoch": 0.7913052905464006, "grad_norm": 0.08803793787956238, "learning_rate": 7.389017844189758e-05, "loss": 0.0052, "step": 7299 }, { "epoch": 0.7914137033824805, "grad_norm": 0.3724015951156616, "learning_rate": 7.388655157406065e-05, "loss": 0.0256, "step": 7300 }, { "epoch": 0.7915221162185603, "grad_norm": 0.37258291244506836, "learning_rate": 7.388292470622371e-05, "loss": 0.0307, "step": 7301 }, { "epoch": 0.79163052905464, "grad_norm": 0.4483797252178192, "learning_rate": 7.387929783838677e-05, "loss": 0.0309, "step": 7302 }, { "epoch": 0.7917389418907199, "grad_norm": 0.17920054495334625, "learning_rate": 7.387567097054984e-05, "loss": 0.0333, "step": 7303 }, { "epoch": 0.7918473547267997, "grad_norm": 0.4702807366847992, "learning_rate": 7.38720441027129e-05, "loss": 0.0166, "step": 7304 }, { "epoch": 0.7919557675628794, "grad_norm": 0.3617919087409973, "learning_rate": 7.386841723487597e-05, "loss": 0.043, "step": 7305 }, { "epoch": 0.7920641803989592, "grad_norm": 0.16755442321300507, "learning_rate": 7.386479036703902e-05, "loss": 0.0205, "step": 7306 }, { "epoch": 0.7921725932350391, "grad_norm": 0.5228714942932129, "learning_rate": 7.38611634992021e-05, "loss": 0.0455, "step": 7307 }, { "epoch": 0.7922810060711188, "grad_norm": 0.21019133925437927, "learning_rate": 7.385753663136515e-05, "loss": 0.0184, "step": 7308 }, { "epoch": 0.7923894189071986, "grad_norm": 0.461731493473053, "learning_rate": 7.385390976352822e-05, "loss": 0.0625, "step": 7309 }, { "epoch": 0.7924978317432784, "grad_norm": 0.2581698000431061, "learning_rate": 7.385028289569128e-05, "loss": 0.0284, "step": 7310 }, { "epoch": 0.7926062445793582, "grad_norm": 0.24974307417869568, "learning_rate": 7.384665602785434e-05, "loss": 0.0204, "step": 7311 }, { "epoch": 0.792714657415438, "grad_norm": 0.5050244331359863, "learning_rate": 7.384302916001741e-05, "loss": 0.0419, "step": 7312 }, { "epoch": 0.7928230702515178, "grad_norm": 0.42540866136550903, "learning_rate": 7.383940229218048e-05, "loss": 0.0218, "step": 7313 }, { "epoch": 0.7929314830875975, "grad_norm": 0.5899295806884766, "learning_rate": 7.383577542434355e-05, "loss": 0.1809, "step": 7314 }, { "epoch": 0.7930398959236774, "grad_norm": 0.11445443332195282, "learning_rate": 7.383214855650661e-05, "loss": 0.0145, "step": 7315 }, { "epoch": 0.7931483087597572, "grad_norm": 0.6362661123275757, "learning_rate": 7.382852168866967e-05, "loss": 0.0399, "step": 7316 }, { "epoch": 0.7932567215958369, "grad_norm": 0.2526831030845642, "learning_rate": 7.382489482083274e-05, "loss": 0.0185, "step": 7317 }, { "epoch": 0.7933651344319167, "grad_norm": 0.18559704720973969, "learning_rate": 7.38212679529958e-05, "loss": 0.016, "step": 7318 }, { "epoch": 0.7934735472679966, "grad_norm": 0.38809409737586975, "learning_rate": 7.381764108515886e-05, "loss": 0.067, "step": 7319 }, { "epoch": 0.7935819601040763, "grad_norm": 0.274531751871109, "learning_rate": 7.381401421732192e-05, "loss": 0.0256, "step": 7320 }, { "epoch": 0.7936903729401561, "grad_norm": 0.403812974691391, "learning_rate": 7.381038734948499e-05, "loss": 0.0757, "step": 7321 }, { "epoch": 0.7937987857762359, "grad_norm": 0.23820453882217407, "learning_rate": 7.380676048164805e-05, "loss": 0.0318, "step": 7322 }, { "epoch": 0.7939071986123157, "grad_norm": 0.34456026554107666, "learning_rate": 7.380313361381112e-05, "loss": 0.0195, "step": 7323 }, { "epoch": 0.7940156114483955, "grad_norm": 0.22996236383914948, "learning_rate": 7.379950674597418e-05, "loss": 0.0334, "step": 7324 }, { "epoch": 0.7941240242844753, "grad_norm": 0.3582976758480072, "learning_rate": 7.379587987813724e-05, "loss": 0.0697, "step": 7325 }, { "epoch": 0.794232437120555, "grad_norm": 0.35190698504447937, "learning_rate": 7.37922530103003e-05, "loss": 0.0781, "step": 7326 }, { "epoch": 0.7943408499566349, "grad_norm": 0.20953620970249176, "learning_rate": 7.378862614246336e-05, "loss": 0.0275, "step": 7327 }, { "epoch": 0.7944492627927147, "grad_norm": 0.5517176389694214, "learning_rate": 7.378499927462643e-05, "loss": 0.0349, "step": 7328 }, { "epoch": 0.7945576756287944, "grad_norm": 0.3880617320537567, "learning_rate": 7.378137240678949e-05, "loss": 0.052, "step": 7329 }, { "epoch": 0.7946660884648742, "grad_norm": 0.2776011526584625, "learning_rate": 7.377774553895256e-05, "loss": 0.0337, "step": 7330 }, { "epoch": 0.7947745013009541, "grad_norm": 0.2006671130657196, "learning_rate": 7.377411867111563e-05, "loss": 0.0387, "step": 7331 }, { "epoch": 0.7948829141370338, "grad_norm": 0.5251744985580444, "learning_rate": 7.377049180327869e-05, "loss": 0.0567, "step": 7332 }, { "epoch": 0.7949913269731136, "grad_norm": 0.30463817715644836, "learning_rate": 7.376686493544176e-05, "loss": 0.0288, "step": 7333 }, { "epoch": 0.7950997398091935, "grad_norm": 0.3730728328227997, "learning_rate": 7.376323806760482e-05, "loss": 0.058, "step": 7334 }, { "epoch": 0.7952081526452732, "grad_norm": 0.4557010531425476, "learning_rate": 7.375961119976789e-05, "loss": 0.0397, "step": 7335 }, { "epoch": 0.795316565481353, "grad_norm": 0.2618752419948578, "learning_rate": 7.375598433193095e-05, "loss": 0.0387, "step": 7336 }, { "epoch": 0.7954249783174328, "grad_norm": 0.47922834753990173, "learning_rate": 7.375235746409402e-05, "loss": 0.0254, "step": 7337 }, { "epoch": 0.7955333911535126, "grad_norm": 0.25336647033691406, "learning_rate": 7.374873059625708e-05, "loss": 0.0251, "step": 7338 }, { "epoch": 0.7956418039895924, "grad_norm": 0.6934643983840942, "learning_rate": 7.374510372842013e-05, "loss": 0.0222, "step": 7339 }, { "epoch": 0.7957502168256722, "grad_norm": 0.23765172064304352, "learning_rate": 7.37414768605832e-05, "loss": 0.0326, "step": 7340 }, { "epoch": 0.7958586296617519, "grad_norm": 0.5740375518798828, "learning_rate": 7.373784999274626e-05, "loss": 0.053, "step": 7341 }, { "epoch": 0.7959670424978318, "grad_norm": 0.18660202622413635, "learning_rate": 7.373422312490933e-05, "loss": 0.0355, "step": 7342 }, { "epoch": 0.7960754553339116, "grad_norm": 0.3179035782814026, "learning_rate": 7.373059625707239e-05, "loss": 0.054, "step": 7343 }, { "epoch": 0.7961838681699913, "grad_norm": 0.32056328654289246, "learning_rate": 7.372696938923546e-05, "loss": 0.0402, "step": 7344 }, { "epoch": 0.7962922810060711, "grad_norm": 0.11017689108848572, "learning_rate": 7.372334252139852e-05, "loss": 0.0105, "step": 7345 }, { "epoch": 0.796400693842151, "grad_norm": 0.3961530923843384, "learning_rate": 7.371971565356159e-05, "loss": 0.024, "step": 7346 }, { "epoch": 0.7965091066782307, "grad_norm": 0.6010503172874451, "learning_rate": 7.371608878572465e-05, "loss": 0.0387, "step": 7347 }, { "epoch": 0.7966175195143105, "grad_norm": 0.17030160129070282, "learning_rate": 7.371246191788772e-05, "loss": 0.0218, "step": 7348 }, { "epoch": 0.7967259323503902, "grad_norm": 0.47954675555229187, "learning_rate": 7.370883505005079e-05, "loss": 0.0434, "step": 7349 }, { "epoch": 0.7968343451864701, "grad_norm": 0.26571011543273926, "learning_rate": 7.370520818221385e-05, "loss": 0.0366, "step": 7350 }, { "epoch": 0.7969427580225499, "grad_norm": 0.28557443618774414, "learning_rate": 7.370158131437692e-05, "loss": 0.0527, "step": 7351 }, { "epoch": 0.7970511708586296, "grad_norm": 0.1511840969324112, "learning_rate": 7.369795444653997e-05, "loss": 0.0161, "step": 7352 }, { "epoch": 0.7971595836947094, "grad_norm": 0.3451473116874695, "learning_rate": 7.369432757870304e-05, "loss": 0.0614, "step": 7353 }, { "epoch": 0.7972679965307893, "grad_norm": 0.3516523241996765, "learning_rate": 7.36907007108661e-05, "loss": 0.028, "step": 7354 }, { "epoch": 0.797376409366869, "grad_norm": 0.3144444525241852, "learning_rate": 7.368707384302916e-05, "loss": 0.0526, "step": 7355 }, { "epoch": 0.7974848222029488, "grad_norm": 0.1362583339214325, "learning_rate": 7.368344697519223e-05, "loss": 0.027, "step": 7356 }, { "epoch": 0.7975932350390286, "grad_norm": 0.4373476803302765, "learning_rate": 7.367982010735529e-05, "loss": 0.0383, "step": 7357 }, { "epoch": 0.7977016478751084, "grad_norm": 0.573111891746521, "learning_rate": 7.367619323951836e-05, "loss": 0.0622, "step": 7358 }, { "epoch": 0.7978100607111882, "grad_norm": 0.14814786612987518, "learning_rate": 7.367256637168142e-05, "loss": 0.0148, "step": 7359 }, { "epoch": 0.797918473547268, "grad_norm": 0.2692908048629761, "learning_rate": 7.366893950384449e-05, "loss": 0.029, "step": 7360 }, { "epoch": 0.7980268863833477, "grad_norm": 0.6036495566368103, "learning_rate": 7.366531263600754e-05, "loss": 0.0525, "step": 7361 }, { "epoch": 0.7981352992194276, "grad_norm": 0.5803971886634827, "learning_rate": 7.36616857681706e-05, "loss": 0.1003, "step": 7362 }, { "epoch": 0.7982437120555074, "grad_norm": 0.22878298163414001, "learning_rate": 7.365805890033367e-05, "loss": 0.0292, "step": 7363 }, { "epoch": 0.7983521248915871, "grad_norm": 0.3773726522922516, "learning_rate": 7.365443203249673e-05, "loss": 0.0467, "step": 7364 }, { "epoch": 0.798460537727667, "grad_norm": 0.18094205856323242, "learning_rate": 7.365080516465981e-05, "loss": 0.0252, "step": 7365 }, { "epoch": 0.7985689505637468, "grad_norm": 0.38334888219833374, "learning_rate": 7.364717829682287e-05, "loss": 0.0335, "step": 7366 }, { "epoch": 0.7986773633998265, "grad_norm": 0.3454173803329468, "learning_rate": 7.364355142898594e-05, "loss": 0.0688, "step": 7367 }, { "epoch": 0.7987857762359063, "grad_norm": 0.291267067193985, "learning_rate": 7.3639924561149e-05, "loss": 0.046, "step": 7368 }, { "epoch": 0.7988941890719862, "grad_norm": 0.40837404131889343, "learning_rate": 7.363629769331206e-05, "loss": 0.0719, "step": 7369 }, { "epoch": 0.7990026019080659, "grad_norm": 0.4010698199272156, "learning_rate": 7.363267082547513e-05, "loss": 0.0621, "step": 7370 }, { "epoch": 0.7991110147441457, "grad_norm": 0.1405622959136963, "learning_rate": 7.362904395763818e-05, "loss": 0.014, "step": 7371 }, { "epoch": 0.7992194275802255, "grad_norm": 0.41009244322776794, "learning_rate": 7.362541708980126e-05, "loss": 0.0256, "step": 7372 }, { "epoch": 0.7993278404163053, "grad_norm": 0.18028105795383453, "learning_rate": 7.362179022196431e-05, "loss": 0.0298, "step": 7373 }, { "epoch": 0.7994362532523851, "grad_norm": 0.2506447434425354, "learning_rate": 7.361816335412738e-05, "loss": 0.0289, "step": 7374 }, { "epoch": 0.7995446660884649, "grad_norm": 0.46601295471191406, "learning_rate": 7.361453648629044e-05, "loss": 0.0492, "step": 7375 }, { "epoch": 0.7996530789245446, "grad_norm": 0.3238711655139923, "learning_rate": 7.361090961845351e-05, "loss": 0.0306, "step": 7376 }, { "epoch": 0.7997614917606245, "grad_norm": 0.36876246333122253, "learning_rate": 7.360728275061657e-05, "loss": 0.0504, "step": 7377 }, { "epoch": 0.7998699045967043, "grad_norm": 0.1120968833565712, "learning_rate": 7.360365588277963e-05, "loss": 0.0211, "step": 7378 }, { "epoch": 0.799978317432784, "grad_norm": 1.026045560836792, "learning_rate": 7.36000290149427e-05, "loss": 0.0392, "step": 7379 }, { "epoch": 0.8000867302688638, "grad_norm": 0.24012494087219238, "learning_rate": 7.359640214710575e-05, "loss": 0.0477, "step": 7380 }, { "epoch": 0.8001951431049437, "grad_norm": 0.145063579082489, "learning_rate": 7.359277527926883e-05, "loss": 0.0142, "step": 7381 }, { "epoch": 0.8003035559410234, "grad_norm": 0.367700457572937, "learning_rate": 7.35891484114319e-05, "loss": 0.0933, "step": 7382 }, { "epoch": 0.8004119687771032, "grad_norm": 0.6037709712982178, "learning_rate": 7.358552154359495e-05, "loss": 0.0476, "step": 7383 }, { "epoch": 0.800520381613183, "grad_norm": 0.610733687877655, "learning_rate": 7.358189467575802e-05, "loss": 0.0309, "step": 7384 }, { "epoch": 0.8006287944492628, "grad_norm": 0.19836963713169098, "learning_rate": 7.357826780792108e-05, "loss": 0.0218, "step": 7385 }, { "epoch": 0.8007372072853426, "grad_norm": 0.21520386636257172, "learning_rate": 7.357464094008415e-05, "loss": 0.0366, "step": 7386 }, { "epoch": 0.8008456201214224, "grad_norm": 0.349060982465744, "learning_rate": 7.357101407224721e-05, "loss": 0.027, "step": 7387 }, { "epoch": 0.8009540329575021, "grad_norm": 0.28542274236679077, "learning_rate": 7.356738720441028e-05, "loss": 0.0461, "step": 7388 }, { "epoch": 0.801062445793582, "grad_norm": 0.3311833143234253, "learning_rate": 7.356376033657334e-05, "loss": 0.0394, "step": 7389 }, { "epoch": 0.8011708586296618, "grad_norm": 0.27841612696647644, "learning_rate": 7.356013346873641e-05, "loss": 0.0355, "step": 7390 }, { "epoch": 0.8012792714657415, "grad_norm": 0.5297048091888428, "learning_rate": 7.355650660089947e-05, "loss": 0.0723, "step": 7391 }, { "epoch": 0.8013876843018214, "grad_norm": 0.2374127060174942, "learning_rate": 7.355287973306252e-05, "loss": 0.0325, "step": 7392 }, { "epoch": 0.8014960971379012, "grad_norm": 0.1990448385477066, "learning_rate": 7.35492528652256e-05, "loss": 0.0206, "step": 7393 }, { "epoch": 0.8016045099739809, "grad_norm": 0.388911634683609, "learning_rate": 7.354562599738865e-05, "loss": 0.022, "step": 7394 }, { "epoch": 0.8017129228100607, "grad_norm": 0.1914398968219757, "learning_rate": 7.354199912955172e-05, "loss": 0.0245, "step": 7395 }, { "epoch": 0.8018213356461406, "grad_norm": 0.16154681146144867, "learning_rate": 7.353837226171478e-05, "loss": 0.0208, "step": 7396 }, { "epoch": 0.8019297484822203, "grad_norm": 0.2342427521944046, "learning_rate": 7.353474539387785e-05, "loss": 0.0236, "step": 7397 }, { "epoch": 0.8020381613183001, "grad_norm": 0.26272520422935486, "learning_rate": 7.353111852604091e-05, "loss": 0.0327, "step": 7398 }, { "epoch": 0.8021465741543798, "grad_norm": 0.22531571984291077, "learning_rate": 7.352749165820398e-05, "loss": 0.0177, "step": 7399 }, { "epoch": 0.8022549869904597, "grad_norm": 0.1784120500087738, "learning_rate": 7.352386479036705e-05, "loss": 0.0283, "step": 7400 }, { "epoch": 0.8023633998265395, "grad_norm": 0.2747727036476135, "learning_rate": 7.352023792253011e-05, "loss": 0.0317, "step": 7401 }, { "epoch": 0.8024718126626192, "grad_norm": 0.9887645840644836, "learning_rate": 7.351661105469318e-05, "loss": 0.0325, "step": 7402 }, { "epoch": 0.802580225498699, "grad_norm": 0.11831552535295486, "learning_rate": 7.351298418685624e-05, "loss": 0.0211, "step": 7403 }, { "epoch": 0.8026886383347789, "grad_norm": 0.34447038173675537, "learning_rate": 7.35093573190193e-05, "loss": 0.0276, "step": 7404 }, { "epoch": 0.8027970511708586, "grad_norm": 0.2410094439983368, "learning_rate": 7.350573045118236e-05, "loss": 0.0522, "step": 7405 }, { "epoch": 0.8029054640069384, "grad_norm": 0.10847686231136322, "learning_rate": 7.350210358334542e-05, "loss": 0.0153, "step": 7406 }, { "epoch": 0.8030138768430182, "grad_norm": 0.3302958607673645, "learning_rate": 7.349847671550849e-05, "loss": 0.0612, "step": 7407 }, { "epoch": 0.803122289679098, "grad_norm": 0.30398616194725037, "learning_rate": 7.349484984767155e-05, "loss": 0.0218, "step": 7408 }, { "epoch": 0.8032307025151778, "grad_norm": 0.1475468873977661, "learning_rate": 7.349122297983462e-05, "loss": 0.0071, "step": 7409 }, { "epoch": 0.8033391153512576, "grad_norm": 0.6034858822822571, "learning_rate": 7.348759611199768e-05, "loss": 0.0757, "step": 7410 }, { "epoch": 0.8034475281873373, "grad_norm": 0.15717114508152008, "learning_rate": 7.348396924416075e-05, "loss": 0.0171, "step": 7411 }, { "epoch": 0.8035559410234172, "grad_norm": 0.3111347556114197, "learning_rate": 7.34803423763238e-05, "loss": 0.0239, "step": 7412 }, { "epoch": 0.803664353859497, "grad_norm": 0.08686503022909164, "learning_rate": 7.347671550848688e-05, "loss": 0.0086, "step": 7413 }, { "epoch": 0.8037727666955767, "grad_norm": 0.29656335711479187, "learning_rate": 7.347308864064993e-05, "loss": 0.0425, "step": 7414 }, { "epoch": 0.8038811795316565, "grad_norm": 0.32892340421676636, "learning_rate": 7.346946177281299e-05, "loss": 0.0332, "step": 7415 }, { "epoch": 0.8039895923677364, "grad_norm": 0.3386991322040558, "learning_rate": 7.346583490497606e-05, "loss": 0.0411, "step": 7416 }, { "epoch": 0.8040980052038161, "grad_norm": 0.10994729399681091, "learning_rate": 7.346220803713913e-05, "loss": 0.012, "step": 7417 }, { "epoch": 0.8042064180398959, "grad_norm": 0.40908312797546387, "learning_rate": 7.34585811693022e-05, "loss": 0.0287, "step": 7418 }, { "epoch": 0.8043148308759757, "grad_norm": 0.26964762806892395, "learning_rate": 7.345495430146526e-05, "loss": 0.0258, "step": 7419 }, { "epoch": 0.8044232437120555, "grad_norm": 0.4358203411102295, "learning_rate": 7.345132743362832e-05, "loss": 0.0224, "step": 7420 }, { "epoch": 0.8045316565481353, "grad_norm": 0.3367292284965515, "learning_rate": 7.344770056579139e-05, "loss": 0.0378, "step": 7421 }, { "epoch": 0.8046400693842151, "grad_norm": 0.12532487511634827, "learning_rate": 7.344407369795445e-05, "loss": 0.0066, "step": 7422 }, { "epoch": 0.8047484822202949, "grad_norm": 0.3253529369831085, "learning_rate": 7.344044683011752e-05, "loss": 0.0217, "step": 7423 }, { "epoch": 0.8048568950563747, "grad_norm": 0.3324473202228546, "learning_rate": 7.343681996228058e-05, "loss": 0.049, "step": 7424 }, { "epoch": 0.8049653078924545, "grad_norm": 0.19046126306056976, "learning_rate": 7.343319309444365e-05, "loss": 0.0273, "step": 7425 }, { "epoch": 0.8050737207285342, "grad_norm": 0.1264723390340805, "learning_rate": 7.34295662266067e-05, "loss": 0.0071, "step": 7426 }, { "epoch": 0.8051821335646141, "grad_norm": 0.2450721114873886, "learning_rate": 7.342593935876977e-05, "loss": 0.0064, "step": 7427 }, { "epoch": 0.8052905464006939, "grad_norm": 0.410247266292572, "learning_rate": 7.342231249093283e-05, "loss": 0.0543, "step": 7428 }, { "epoch": 0.8053989592367736, "grad_norm": 0.36160582304000854, "learning_rate": 7.341868562309589e-05, "loss": 0.0478, "step": 7429 }, { "epoch": 0.8055073720728534, "grad_norm": 0.39514583349227905, "learning_rate": 7.341505875525896e-05, "loss": 0.0532, "step": 7430 }, { "epoch": 0.8056157849089333, "grad_norm": 0.2651047706604004, "learning_rate": 7.341143188742202e-05, "loss": 0.0137, "step": 7431 }, { "epoch": 0.805724197745013, "grad_norm": 0.1494198739528656, "learning_rate": 7.340780501958509e-05, "loss": 0.0153, "step": 7432 }, { "epoch": 0.8058326105810928, "grad_norm": 0.4499221444129944, "learning_rate": 7.340417815174815e-05, "loss": 0.0528, "step": 7433 }, { "epoch": 0.8059410234171726, "grad_norm": 0.3145540654659271, "learning_rate": 7.340055128391122e-05, "loss": 0.0324, "step": 7434 }, { "epoch": 0.8060494362532524, "grad_norm": 0.756389856338501, "learning_rate": 7.339692441607429e-05, "loss": 0.0744, "step": 7435 }, { "epoch": 0.8061578490893322, "grad_norm": 0.585207998752594, "learning_rate": 7.339329754823734e-05, "loss": 0.0513, "step": 7436 }, { "epoch": 0.806266261925412, "grad_norm": 0.6186605095863342, "learning_rate": 7.338967068040042e-05, "loss": 0.0869, "step": 7437 }, { "epoch": 0.8063746747614917, "grad_norm": 0.5441632270812988, "learning_rate": 7.338604381256347e-05, "loss": 0.03, "step": 7438 }, { "epoch": 0.8064830875975716, "grad_norm": 0.7702534198760986, "learning_rate": 7.338241694472654e-05, "loss": 0.08, "step": 7439 }, { "epoch": 0.8065915004336514, "grad_norm": 0.5139950513839722, "learning_rate": 7.33787900768896e-05, "loss": 0.0579, "step": 7440 }, { "epoch": 0.8066999132697311, "grad_norm": 0.3862360417842865, "learning_rate": 7.337516320905267e-05, "loss": 0.0288, "step": 7441 }, { "epoch": 0.8068083261058109, "grad_norm": 0.16976982355117798, "learning_rate": 7.337153634121573e-05, "loss": 0.0225, "step": 7442 }, { "epoch": 0.8069167389418908, "grad_norm": 0.2179652899503708, "learning_rate": 7.336790947337879e-05, "loss": 0.0234, "step": 7443 }, { "epoch": 0.8070251517779705, "grad_norm": 0.1803155541419983, "learning_rate": 7.336428260554186e-05, "loss": 0.0265, "step": 7444 }, { "epoch": 0.8071335646140503, "grad_norm": 0.49314722418785095, "learning_rate": 7.336065573770491e-05, "loss": 0.1132, "step": 7445 }, { "epoch": 0.80724197745013, "grad_norm": 0.33251795172691345, "learning_rate": 7.335702886986799e-05, "loss": 0.0561, "step": 7446 }, { "epoch": 0.8073503902862099, "grad_norm": 0.43793433904647827, "learning_rate": 7.335340200203104e-05, "loss": 0.0731, "step": 7447 }, { "epoch": 0.8074588031222897, "grad_norm": 0.4824555814266205, "learning_rate": 7.334977513419411e-05, "loss": 0.053, "step": 7448 }, { "epoch": 0.8075672159583694, "grad_norm": 0.18207506835460663, "learning_rate": 7.334614826635717e-05, "loss": 0.0182, "step": 7449 }, { "epoch": 0.8076756287944492, "grad_norm": 0.7706271409988403, "learning_rate": 7.334252139852024e-05, "loss": 0.0951, "step": 7450 }, { "epoch": 0.8077840416305291, "grad_norm": 0.08495458960533142, "learning_rate": 7.333889453068331e-05, "loss": 0.0204, "step": 7451 }, { "epoch": 0.8078924544666088, "grad_norm": 0.846527636051178, "learning_rate": 7.333526766284637e-05, "loss": 0.0658, "step": 7452 }, { "epoch": 0.8080008673026886, "grad_norm": 0.23640795052051544, "learning_rate": 7.333164079500944e-05, "loss": 0.0285, "step": 7453 }, { "epoch": 0.8081092801387685, "grad_norm": 0.18344508111476898, "learning_rate": 7.33280139271725e-05, "loss": 0.0224, "step": 7454 }, { "epoch": 0.8082176929748482, "grad_norm": 0.12282070517539978, "learning_rate": 7.332438705933557e-05, "loss": 0.0237, "step": 7455 }, { "epoch": 0.808326105810928, "grad_norm": 0.3799947500228882, "learning_rate": 7.332076019149863e-05, "loss": 0.0464, "step": 7456 }, { "epoch": 0.8084345186470078, "grad_norm": 0.3850559592247009, "learning_rate": 7.33171333236617e-05, "loss": 0.0385, "step": 7457 }, { "epoch": 0.8085429314830876, "grad_norm": 0.08572619408369064, "learning_rate": 7.331350645582475e-05, "loss": 0.0113, "step": 7458 }, { "epoch": 0.8086513443191674, "grad_norm": 0.2228928506374359, "learning_rate": 7.330987958798781e-05, "loss": 0.0577, "step": 7459 }, { "epoch": 0.8087597571552472, "grad_norm": 0.7734867930412292, "learning_rate": 7.330625272015088e-05, "loss": 0.0914, "step": 7460 }, { "epoch": 0.8088681699913269, "grad_norm": 0.13318650424480438, "learning_rate": 7.330262585231394e-05, "loss": 0.0158, "step": 7461 }, { "epoch": 0.8089765828274068, "grad_norm": 0.2223590612411499, "learning_rate": 7.329899898447701e-05, "loss": 0.029, "step": 7462 }, { "epoch": 0.8090849956634866, "grad_norm": 0.26714375615119934, "learning_rate": 7.329537211664007e-05, "loss": 0.0331, "step": 7463 }, { "epoch": 0.8091934084995663, "grad_norm": 0.30206504464149475, "learning_rate": 7.329174524880314e-05, "loss": 0.0608, "step": 7464 }, { "epoch": 0.8093018213356461, "grad_norm": 0.1160939559340477, "learning_rate": 7.32881183809662e-05, "loss": 0.013, "step": 7465 }, { "epoch": 0.809410234171726, "grad_norm": 0.37328994274139404, "learning_rate": 7.328449151312925e-05, "loss": 0.0461, "step": 7466 }, { "epoch": 0.8095186470078057, "grad_norm": 0.2675226032733917, "learning_rate": 7.328086464529233e-05, "loss": 0.0351, "step": 7467 }, { "epoch": 0.8096270598438855, "grad_norm": 0.1425662636756897, "learning_rate": 7.327723777745538e-05, "loss": 0.017, "step": 7468 }, { "epoch": 0.8097354726799653, "grad_norm": 0.5372952222824097, "learning_rate": 7.327361090961847e-05, "loss": 0.0632, "step": 7469 }, { "epoch": 0.8098438855160451, "grad_norm": 0.23158186674118042, "learning_rate": 7.326998404178152e-05, "loss": 0.0659, "step": 7470 }, { "epoch": 0.8099522983521249, "grad_norm": 0.2349306344985962, "learning_rate": 7.32663571739446e-05, "loss": 0.0215, "step": 7471 }, { "epoch": 0.8100607111882047, "grad_norm": 0.36173638701438904, "learning_rate": 7.326273030610765e-05, "loss": 0.034, "step": 7472 }, { "epoch": 0.8101691240242844, "grad_norm": 0.2295965999364853, "learning_rate": 7.325910343827071e-05, "loss": 0.0173, "step": 7473 }, { "epoch": 0.8102775368603643, "grad_norm": 0.2653436064720154, "learning_rate": 7.325547657043378e-05, "loss": 0.0358, "step": 7474 }, { "epoch": 0.8103859496964441, "grad_norm": 0.43921685218811035, "learning_rate": 7.325184970259684e-05, "loss": 0.0838, "step": 7475 }, { "epoch": 0.8104943625325238, "grad_norm": 0.17340701818466187, "learning_rate": 7.324822283475991e-05, "loss": 0.0289, "step": 7476 }, { "epoch": 0.8106027753686036, "grad_norm": 0.19200940430164337, "learning_rate": 7.324459596692297e-05, "loss": 0.0341, "step": 7477 }, { "epoch": 0.8107111882046835, "grad_norm": 0.3301803469657898, "learning_rate": 7.324096909908604e-05, "loss": 0.043, "step": 7478 }, { "epoch": 0.8108196010407632, "grad_norm": 0.28235742449760437, "learning_rate": 7.32373422312491e-05, "loss": 0.0431, "step": 7479 }, { "epoch": 0.810928013876843, "grad_norm": 0.20300889015197754, "learning_rate": 7.323371536341217e-05, "loss": 0.0309, "step": 7480 }, { "epoch": 0.8110364267129229, "grad_norm": 0.2415986955165863, "learning_rate": 7.323008849557522e-05, "loss": 0.0329, "step": 7481 }, { "epoch": 0.8111448395490026, "grad_norm": 0.3968254327774048, "learning_rate": 7.322646162773828e-05, "loss": 0.0651, "step": 7482 }, { "epoch": 0.8112532523850824, "grad_norm": 0.06951986253261566, "learning_rate": 7.322283475990135e-05, "loss": 0.0067, "step": 7483 }, { "epoch": 0.8113616652211622, "grad_norm": 0.1465279459953308, "learning_rate": 7.321920789206441e-05, "loss": 0.0149, "step": 7484 }, { "epoch": 0.811470078057242, "grad_norm": 0.27606117725372314, "learning_rate": 7.321558102422748e-05, "loss": 0.0294, "step": 7485 }, { "epoch": 0.8115784908933218, "grad_norm": 0.1626424640417099, "learning_rate": 7.321195415639055e-05, "loss": 0.0247, "step": 7486 }, { "epoch": 0.8116869037294016, "grad_norm": 0.35490652918815613, "learning_rate": 7.320832728855361e-05, "loss": 0.0515, "step": 7487 }, { "epoch": 0.8117953165654813, "grad_norm": 0.14671355485916138, "learning_rate": 7.320470042071668e-05, "loss": 0.0204, "step": 7488 }, { "epoch": 0.8119037294015612, "grad_norm": 0.28021833300590515, "learning_rate": 7.320107355287974e-05, "loss": 0.0419, "step": 7489 }, { "epoch": 0.812012142237641, "grad_norm": 0.22648625075817108, "learning_rate": 7.31974466850428e-05, "loss": 0.0487, "step": 7490 }, { "epoch": 0.8121205550737207, "grad_norm": 0.38869839906692505, "learning_rate": 7.319381981720586e-05, "loss": 0.046, "step": 7491 }, { "epoch": 0.8122289679098005, "grad_norm": 0.20212163031101227, "learning_rate": 7.319019294936893e-05, "loss": 0.0252, "step": 7492 }, { "epoch": 0.8123373807458804, "grad_norm": 0.14972364902496338, "learning_rate": 7.318656608153199e-05, "loss": 0.0189, "step": 7493 }, { "epoch": 0.8124457935819601, "grad_norm": 0.12343054264783859, "learning_rate": 7.318293921369506e-05, "loss": 0.0138, "step": 7494 }, { "epoch": 0.8125542064180399, "grad_norm": 0.3541710674762726, "learning_rate": 7.317931234585812e-05, "loss": 0.0554, "step": 7495 }, { "epoch": 0.8126626192541196, "grad_norm": 0.4122332036495209, "learning_rate": 7.317568547802118e-05, "loss": 0.0665, "step": 7496 }, { "epoch": 0.8127710320901995, "grad_norm": 0.4868146777153015, "learning_rate": 7.317205861018425e-05, "loss": 0.0672, "step": 7497 }, { "epoch": 0.8128794449262793, "grad_norm": 0.2859930396080017, "learning_rate": 7.31684317423473e-05, "loss": 0.0438, "step": 7498 }, { "epoch": 0.812987857762359, "grad_norm": 0.3365119695663452, "learning_rate": 7.316480487451038e-05, "loss": 0.0565, "step": 7499 }, { "epoch": 0.8130962705984388, "grad_norm": 0.27608412504196167, "learning_rate": 7.316117800667343e-05, "loss": 0.0356, "step": 7500 }, { "epoch": 0.8132046834345187, "grad_norm": 0.18123388290405273, "learning_rate": 7.31575511388365e-05, "loss": 0.0261, "step": 7501 }, { "epoch": 0.8133130962705984, "grad_norm": 0.5361722111701965, "learning_rate": 7.315392427099956e-05, "loss": 0.1223, "step": 7502 }, { "epoch": 0.8134215091066782, "grad_norm": 0.17667199671268463, "learning_rate": 7.315029740316263e-05, "loss": 0.0176, "step": 7503 }, { "epoch": 0.813529921942758, "grad_norm": 0.41686323285102844, "learning_rate": 7.31466705353257e-05, "loss": 0.0313, "step": 7504 }, { "epoch": 0.8136383347788378, "grad_norm": 0.30848607420921326, "learning_rate": 7.314304366748876e-05, "loss": 0.0543, "step": 7505 }, { "epoch": 0.8137467476149176, "grad_norm": 0.3665293753147125, "learning_rate": 7.313941679965183e-05, "loss": 0.0865, "step": 7506 }, { "epoch": 0.8138551604509974, "grad_norm": 0.1244748905301094, "learning_rate": 7.313578993181489e-05, "loss": 0.0187, "step": 7507 }, { "epoch": 0.8139635732870771, "grad_norm": 0.1530434787273407, "learning_rate": 7.313216306397796e-05, "loss": 0.0336, "step": 7508 }, { "epoch": 0.814071986123157, "grad_norm": 0.13941481709480286, "learning_rate": 7.312853619614102e-05, "loss": 0.0266, "step": 7509 }, { "epoch": 0.8141803989592368, "grad_norm": 0.20575487613677979, "learning_rate": 7.312490932830407e-05, "loss": 0.0266, "step": 7510 }, { "epoch": 0.8142888117953165, "grad_norm": 0.48829010128974915, "learning_rate": 7.312128246046715e-05, "loss": 0.0731, "step": 7511 }, { "epoch": 0.8143972246313964, "grad_norm": 0.12699557840824127, "learning_rate": 7.31176555926302e-05, "loss": 0.0142, "step": 7512 }, { "epoch": 0.8145056374674762, "grad_norm": 0.40117162466049194, "learning_rate": 7.311402872479327e-05, "loss": 0.0577, "step": 7513 }, { "epoch": 0.8146140503035559, "grad_norm": 0.1836705356836319, "learning_rate": 7.311040185695633e-05, "loss": 0.0151, "step": 7514 }, { "epoch": 0.8147224631396357, "grad_norm": 0.241327702999115, "learning_rate": 7.31067749891194e-05, "loss": 0.0655, "step": 7515 }, { "epoch": 0.8148308759757156, "grad_norm": 0.1668068766593933, "learning_rate": 7.310314812128246e-05, "loss": 0.0194, "step": 7516 }, { "epoch": 0.8149392888117953, "grad_norm": 0.07138732075691223, "learning_rate": 7.309952125344553e-05, "loss": 0.0072, "step": 7517 }, { "epoch": 0.8150477016478751, "grad_norm": 0.5495093464851379, "learning_rate": 7.309589438560859e-05, "loss": 0.0372, "step": 7518 }, { "epoch": 0.8151561144839549, "grad_norm": 0.49308082461357117, "learning_rate": 7.309226751777164e-05, "loss": 0.0347, "step": 7519 }, { "epoch": 0.8152645273200347, "grad_norm": 0.4924364984035492, "learning_rate": 7.308864064993473e-05, "loss": 0.0469, "step": 7520 }, { "epoch": 0.8153729401561145, "grad_norm": 0.1352006196975708, "learning_rate": 7.308501378209779e-05, "loss": 0.0129, "step": 7521 }, { "epoch": 0.8154813529921943, "grad_norm": 0.31298214197158813, "learning_rate": 7.308138691426086e-05, "loss": 0.0221, "step": 7522 }, { "epoch": 0.815589765828274, "grad_norm": 0.18838410079479218, "learning_rate": 7.307776004642391e-05, "loss": 0.0287, "step": 7523 }, { "epoch": 0.8156981786643539, "grad_norm": 0.221892312169075, "learning_rate": 7.307413317858697e-05, "loss": 0.0236, "step": 7524 }, { "epoch": 0.8158065915004337, "grad_norm": 0.15865539014339447, "learning_rate": 7.307050631075004e-05, "loss": 0.0284, "step": 7525 }, { "epoch": 0.8159150043365134, "grad_norm": 0.6627461314201355, "learning_rate": 7.30668794429131e-05, "loss": 0.05, "step": 7526 }, { "epoch": 0.8160234171725932, "grad_norm": 0.2558284401893616, "learning_rate": 7.306325257507617e-05, "loss": 0.016, "step": 7527 }, { "epoch": 0.8161318300086731, "grad_norm": 0.7478334307670593, "learning_rate": 7.305962570723923e-05, "loss": 0.0223, "step": 7528 }, { "epoch": 0.8162402428447528, "grad_norm": 0.2999652922153473, "learning_rate": 7.30559988394023e-05, "loss": 0.0476, "step": 7529 }, { "epoch": 0.8163486556808326, "grad_norm": 0.32166245579719543, "learning_rate": 7.305237197156536e-05, "loss": 0.0301, "step": 7530 }, { "epoch": 0.8164570685169124, "grad_norm": 0.502267599105835, "learning_rate": 7.304874510372843e-05, "loss": 0.0395, "step": 7531 }, { "epoch": 0.8165654813529922, "grad_norm": 0.6629350781440735, "learning_rate": 7.304511823589149e-05, "loss": 0.073, "step": 7532 }, { "epoch": 0.816673894189072, "grad_norm": 0.2729581892490387, "learning_rate": 7.304149136805454e-05, "loss": 0.04, "step": 7533 }, { "epoch": 0.8167823070251518, "grad_norm": 0.22208033502101898, "learning_rate": 7.303786450021761e-05, "loss": 0.0181, "step": 7534 }, { "epoch": 0.8168907198612315, "grad_norm": 0.02967766486108303, "learning_rate": 7.303423763238067e-05, "loss": 0.0036, "step": 7535 }, { "epoch": 0.8169991326973114, "grad_norm": 1.4930270910263062, "learning_rate": 7.303061076454374e-05, "loss": 0.1146, "step": 7536 }, { "epoch": 0.8171075455333912, "grad_norm": 0.35057327151298523, "learning_rate": 7.30269838967068e-05, "loss": 0.0224, "step": 7537 }, { "epoch": 0.8172159583694709, "grad_norm": 0.6567506790161133, "learning_rate": 7.302335702886988e-05, "loss": 0.07, "step": 7538 }, { "epoch": 0.8173243712055508, "grad_norm": 0.4857943654060364, "learning_rate": 7.301973016103294e-05, "loss": 0.0564, "step": 7539 }, { "epoch": 0.8174327840416306, "grad_norm": 0.2761875092983246, "learning_rate": 7.3016103293196e-05, "loss": 0.0265, "step": 7540 }, { "epoch": 0.8175411968777103, "grad_norm": 0.22307029366493225, "learning_rate": 7.301247642535907e-05, "loss": 0.0327, "step": 7541 }, { "epoch": 0.8176496097137901, "grad_norm": 0.4052538275718689, "learning_rate": 7.300884955752213e-05, "loss": 0.0445, "step": 7542 }, { "epoch": 0.81775802254987, "grad_norm": 0.24129953980445862, "learning_rate": 7.30052226896852e-05, "loss": 0.0468, "step": 7543 }, { "epoch": 0.8178664353859497, "grad_norm": 0.05768120661377907, "learning_rate": 7.300159582184825e-05, "loss": 0.0056, "step": 7544 }, { "epoch": 0.8179748482220295, "grad_norm": 0.13526926934719086, "learning_rate": 7.299796895401133e-05, "loss": 0.0177, "step": 7545 }, { "epoch": 0.8180832610581092, "grad_norm": 0.2940336763858795, "learning_rate": 7.299434208617438e-05, "loss": 0.0475, "step": 7546 }, { "epoch": 0.8181916738941891, "grad_norm": 0.369683176279068, "learning_rate": 7.299071521833744e-05, "loss": 0.0276, "step": 7547 }, { "epoch": 0.8183000867302689, "grad_norm": 0.22955799102783203, "learning_rate": 7.298708835050051e-05, "loss": 0.0219, "step": 7548 }, { "epoch": 0.8184084995663486, "grad_norm": 0.4954235553741455, "learning_rate": 7.298346148266357e-05, "loss": 0.067, "step": 7549 }, { "epoch": 0.8185169124024284, "grad_norm": 0.2068738490343094, "learning_rate": 7.297983461482664e-05, "loss": 0.0202, "step": 7550 }, { "epoch": 0.8186253252385083, "grad_norm": 0.24453239142894745, "learning_rate": 7.29762077469897e-05, "loss": 0.0401, "step": 7551 }, { "epoch": 0.818733738074588, "grad_norm": 0.39147377014160156, "learning_rate": 7.297258087915277e-05, "loss": 0.0886, "step": 7552 }, { "epoch": 0.8188421509106678, "grad_norm": 0.18975681066513062, "learning_rate": 7.296895401131582e-05, "loss": 0.0119, "step": 7553 }, { "epoch": 0.8189505637467476, "grad_norm": 0.40577155351638794, "learning_rate": 7.29653271434789e-05, "loss": 0.0146, "step": 7554 }, { "epoch": 0.8190589765828274, "grad_norm": 0.20290020108222961, "learning_rate": 7.296170027564197e-05, "loss": 0.0317, "step": 7555 }, { "epoch": 0.8191673894189072, "grad_norm": 0.29417884349823, "learning_rate": 7.295807340780502e-05, "loss": 0.0326, "step": 7556 }, { "epoch": 0.819275802254987, "grad_norm": 0.37416544556617737, "learning_rate": 7.29544465399681e-05, "loss": 0.0589, "step": 7557 }, { "epoch": 0.8193842150910667, "grad_norm": 0.22675634920597076, "learning_rate": 7.295081967213115e-05, "loss": 0.0593, "step": 7558 }, { "epoch": 0.8194926279271466, "grad_norm": 0.38825008273124695, "learning_rate": 7.294719280429422e-05, "loss": 0.0108, "step": 7559 }, { "epoch": 0.8196010407632264, "grad_norm": 0.43887579441070557, "learning_rate": 7.294356593645728e-05, "loss": 0.0752, "step": 7560 }, { "epoch": 0.8197094535993061, "grad_norm": 0.5146497488021851, "learning_rate": 7.293993906862035e-05, "loss": 0.06, "step": 7561 }, { "epoch": 0.8198178664353859, "grad_norm": 0.36140844225883484, "learning_rate": 7.293631220078341e-05, "loss": 0.0281, "step": 7562 }, { "epoch": 0.8199262792714658, "grad_norm": 0.3911173641681671, "learning_rate": 7.293268533294647e-05, "loss": 0.0297, "step": 7563 }, { "epoch": 0.8200346921075455, "grad_norm": 0.09600687772035599, "learning_rate": 7.292905846510954e-05, "loss": 0.0107, "step": 7564 }, { "epoch": 0.8201431049436253, "grad_norm": 0.1983727514743805, "learning_rate": 7.29254315972726e-05, "loss": 0.0336, "step": 7565 }, { "epoch": 0.8202515177797051, "grad_norm": 0.18320055305957794, "learning_rate": 7.292180472943566e-05, "loss": 0.0207, "step": 7566 }, { "epoch": 0.8203599306157849, "grad_norm": 0.43059396743774414, "learning_rate": 7.291817786159872e-05, "loss": 0.0415, "step": 7567 }, { "epoch": 0.8204683434518647, "grad_norm": 0.2806225121021271, "learning_rate": 7.291455099376179e-05, "loss": 0.0399, "step": 7568 }, { "epoch": 0.8205767562879445, "grad_norm": 0.25700852274894714, "learning_rate": 7.291092412592485e-05, "loss": 0.0288, "step": 7569 }, { "epoch": 0.8206851691240243, "grad_norm": 0.1931845247745514, "learning_rate": 7.290729725808792e-05, "loss": 0.0294, "step": 7570 }, { "epoch": 0.8207935819601041, "grad_norm": 0.5342658162117004, "learning_rate": 7.290367039025098e-05, "loss": 0.0178, "step": 7571 }, { "epoch": 0.8209019947961839, "grad_norm": 0.33275729417800903, "learning_rate": 7.290004352241405e-05, "loss": 0.0353, "step": 7572 }, { "epoch": 0.8210104076322636, "grad_norm": 0.5622420310974121, "learning_rate": 7.289641665457712e-05, "loss": 0.0329, "step": 7573 }, { "epoch": 0.8211188204683435, "grad_norm": 0.4476664364337921, "learning_rate": 7.289278978674018e-05, "loss": 0.0543, "step": 7574 }, { "epoch": 0.8212272333044233, "grad_norm": 0.24192388355731964, "learning_rate": 7.288916291890325e-05, "loss": 0.0208, "step": 7575 }, { "epoch": 0.821335646140503, "grad_norm": 0.46264195442199707, "learning_rate": 7.28855360510663e-05, "loss": 0.0837, "step": 7576 }, { "epoch": 0.8214440589765828, "grad_norm": 0.26082339882850647, "learning_rate": 7.288190918322936e-05, "loss": 0.0328, "step": 7577 }, { "epoch": 0.8215524718126627, "grad_norm": 0.28885146975517273, "learning_rate": 7.287828231539243e-05, "loss": 0.0522, "step": 7578 }, { "epoch": 0.8216608846487424, "grad_norm": 0.14764192700386047, "learning_rate": 7.287465544755549e-05, "loss": 0.0089, "step": 7579 }, { "epoch": 0.8217692974848222, "grad_norm": 0.3042779862880707, "learning_rate": 7.287102857971856e-05, "loss": 0.0256, "step": 7580 }, { "epoch": 0.821877710320902, "grad_norm": 0.6460333466529846, "learning_rate": 7.286740171188162e-05, "loss": 0.046, "step": 7581 }, { "epoch": 0.8219861231569818, "grad_norm": 0.8627961874008179, "learning_rate": 7.286377484404469e-05, "loss": 0.063, "step": 7582 }, { "epoch": 0.8220945359930616, "grad_norm": 0.25967493653297424, "learning_rate": 7.286014797620775e-05, "loss": 0.0158, "step": 7583 }, { "epoch": 0.8222029488291414, "grad_norm": 0.5806087255477905, "learning_rate": 7.285652110837082e-05, "loss": 0.0571, "step": 7584 }, { "epoch": 0.8223113616652211, "grad_norm": 0.344874769449234, "learning_rate": 7.285289424053388e-05, "loss": 0.0458, "step": 7585 }, { "epoch": 0.822419774501301, "grad_norm": 0.8255389928817749, "learning_rate": 7.284926737269693e-05, "loss": 0.0598, "step": 7586 }, { "epoch": 0.8225281873373808, "grad_norm": 0.383847177028656, "learning_rate": 7.284564050486e-05, "loss": 0.0426, "step": 7587 }, { "epoch": 0.8226366001734605, "grad_norm": 0.3510738015174866, "learning_rate": 7.284201363702306e-05, "loss": 0.0328, "step": 7588 }, { "epoch": 0.8227450130095403, "grad_norm": 0.6992799639701843, "learning_rate": 7.283838676918613e-05, "loss": 0.0266, "step": 7589 }, { "epoch": 0.8228534258456202, "grad_norm": 0.029728449881076813, "learning_rate": 7.28347599013492e-05, "loss": 0.0028, "step": 7590 }, { "epoch": 0.8229618386816999, "grad_norm": 0.22403743863105774, "learning_rate": 7.283113303351226e-05, "loss": 0.0238, "step": 7591 }, { "epoch": 0.8230702515177797, "grad_norm": 0.27827173471450806, "learning_rate": 7.282750616567533e-05, "loss": 0.0463, "step": 7592 }, { "epoch": 0.8231786643538594, "grad_norm": 0.28219151496887207, "learning_rate": 7.282387929783839e-05, "loss": 0.0317, "step": 7593 }, { "epoch": 0.8232870771899393, "grad_norm": 0.3097004294395447, "learning_rate": 7.282025243000146e-05, "loss": 0.0458, "step": 7594 }, { "epoch": 0.8233954900260191, "grad_norm": 0.2005893439054489, "learning_rate": 7.281662556216452e-05, "loss": 0.0347, "step": 7595 }, { "epoch": 0.8235039028620988, "grad_norm": 0.33365198969841003, "learning_rate": 7.281299869432759e-05, "loss": 0.0252, "step": 7596 }, { "epoch": 0.8236123156981786, "grad_norm": 0.31605520844459534, "learning_rate": 7.280937182649065e-05, "loss": 0.0074, "step": 7597 }, { "epoch": 0.8237207285342585, "grad_norm": 0.3899793028831482, "learning_rate": 7.280574495865372e-05, "loss": 0.0431, "step": 7598 }, { "epoch": 0.8238291413703382, "grad_norm": 0.32459020614624023, "learning_rate": 7.280211809081677e-05, "loss": 0.0449, "step": 7599 }, { "epoch": 0.823937554206418, "grad_norm": 0.27271878719329834, "learning_rate": 7.279849122297983e-05, "loss": 0.0511, "step": 7600 }, { "epoch": 0.8240459670424979, "grad_norm": 0.43557628989219666, "learning_rate": 7.27948643551429e-05, "loss": 0.044, "step": 7601 }, { "epoch": 0.8241543798785776, "grad_norm": 0.0978422686457634, "learning_rate": 7.279123748730596e-05, "loss": 0.0069, "step": 7602 }, { "epoch": 0.8242627927146574, "grad_norm": 0.4341309666633606, "learning_rate": 7.278761061946903e-05, "loss": 0.0678, "step": 7603 }, { "epoch": 0.8243712055507372, "grad_norm": 0.2713147699832916, "learning_rate": 7.278398375163209e-05, "loss": 0.027, "step": 7604 }, { "epoch": 0.824479618386817, "grad_norm": 0.4759891629219055, "learning_rate": 7.278035688379516e-05, "loss": 0.0186, "step": 7605 }, { "epoch": 0.8245880312228968, "grad_norm": 0.20944248139858246, "learning_rate": 7.277673001595822e-05, "loss": 0.0135, "step": 7606 }, { "epoch": 0.8246964440589766, "grad_norm": 0.2637064456939697, "learning_rate": 7.277310314812129e-05, "loss": 0.0585, "step": 7607 }, { "epoch": 0.8248048568950563, "grad_norm": 0.9172323346138, "learning_rate": 7.276947628028436e-05, "loss": 0.0716, "step": 7608 }, { "epoch": 0.8249132697311362, "grad_norm": 0.5526087284088135, "learning_rate": 7.276584941244741e-05, "loss": 0.0592, "step": 7609 }, { "epoch": 0.825021682567216, "grad_norm": 0.37965914607048035, "learning_rate": 7.276222254461049e-05, "loss": 0.0323, "step": 7610 }, { "epoch": 0.8251300954032957, "grad_norm": 0.6999977231025696, "learning_rate": 7.275859567677354e-05, "loss": 0.0459, "step": 7611 }, { "epoch": 0.8252385082393755, "grad_norm": 0.18867115676403046, "learning_rate": 7.275496880893661e-05, "loss": 0.0398, "step": 7612 }, { "epoch": 0.8253469210754554, "grad_norm": 0.3508147597312927, "learning_rate": 7.275134194109967e-05, "loss": 0.0275, "step": 7613 }, { "epoch": 0.8254553339115351, "grad_norm": 0.32329943776130676, "learning_rate": 7.274771507326273e-05, "loss": 0.0241, "step": 7614 }, { "epoch": 0.8255637467476149, "grad_norm": 0.4392671287059784, "learning_rate": 7.27440882054258e-05, "loss": 0.0702, "step": 7615 }, { "epoch": 0.8256721595836947, "grad_norm": 0.1577150672674179, "learning_rate": 7.274046133758886e-05, "loss": 0.0291, "step": 7616 }, { "epoch": 0.8257805724197745, "grad_norm": 0.2995431423187256, "learning_rate": 7.273683446975193e-05, "loss": 0.0159, "step": 7617 }, { "epoch": 0.8258889852558543, "grad_norm": 0.45336413383483887, "learning_rate": 7.273320760191498e-05, "loss": 0.0641, "step": 7618 }, { "epoch": 0.8259973980919341, "grad_norm": 0.1640806496143341, "learning_rate": 7.272958073407806e-05, "loss": 0.0321, "step": 7619 }, { "epoch": 0.8261058109280138, "grad_norm": 0.5224006772041321, "learning_rate": 7.272595386624111e-05, "loss": 0.0786, "step": 7620 }, { "epoch": 0.8262142237640937, "grad_norm": 0.5423130989074707, "learning_rate": 7.272232699840418e-05, "loss": 0.0281, "step": 7621 }, { "epoch": 0.8263226366001735, "grad_norm": 0.5078732967376709, "learning_rate": 7.271870013056724e-05, "loss": 0.0429, "step": 7622 }, { "epoch": 0.8264310494362532, "grad_norm": 0.35491442680358887, "learning_rate": 7.27150732627303e-05, "loss": 0.0498, "step": 7623 }, { "epoch": 0.826539462272333, "grad_norm": 0.3926980793476105, "learning_rate": 7.271144639489338e-05, "loss": 0.0412, "step": 7624 }, { "epoch": 0.8266478751084129, "grad_norm": 0.20186413824558258, "learning_rate": 7.270781952705644e-05, "loss": 0.0247, "step": 7625 }, { "epoch": 0.8267562879444926, "grad_norm": 0.5276736617088318, "learning_rate": 7.270419265921951e-05, "loss": 0.088, "step": 7626 }, { "epoch": 0.8268647007805724, "grad_norm": 0.24710361659526825, "learning_rate": 7.270056579138257e-05, "loss": 0.0371, "step": 7627 }, { "epoch": 0.8269731136166523, "grad_norm": 0.45507845282554626, "learning_rate": 7.269693892354563e-05, "loss": 0.0608, "step": 7628 }, { "epoch": 0.827081526452732, "grad_norm": 0.13562992215156555, "learning_rate": 7.26933120557087e-05, "loss": 0.014, "step": 7629 }, { "epoch": 0.8271899392888118, "grad_norm": 0.282036155462265, "learning_rate": 7.268968518787175e-05, "loss": 0.0388, "step": 7630 }, { "epoch": 0.8272983521248916, "grad_norm": 0.15357856452465057, "learning_rate": 7.268605832003482e-05, "loss": 0.0308, "step": 7631 }, { "epoch": 0.8274067649609714, "grad_norm": 0.23564361035823822, "learning_rate": 7.268243145219788e-05, "loss": 0.0383, "step": 7632 }, { "epoch": 0.8275151777970512, "grad_norm": 0.3485215902328491, "learning_rate": 7.267880458436095e-05, "loss": 0.0695, "step": 7633 }, { "epoch": 0.827623590633131, "grad_norm": 0.15714992582798004, "learning_rate": 7.267517771652401e-05, "loss": 0.0207, "step": 7634 }, { "epoch": 0.8277320034692107, "grad_norm": 0.5115841627120972, "learning_rate": 7.267155084868708e-05, "loss": 0.0237, "step": 7635 }, { "epoch": 0.8278404163052906, "grad_norm": 0.4651440680027008, "learning_rate": 7.266792398085014e-05, "loss": 0.014, "step": 7636 }, { "epoch": 0.8279488291413704, "grad_norm": 0.22979803383350372, "learning_rate": 7.26642971130132e-05, "loss": 0.037, "step": 7637 }, { "epoch": 0.8280572419774501, "grad_norm": 0.5683718323707581, "learning_rate": 7.266067024517627e-05, "loss": 0.0626, "step": 7638 }, { "epoch": 0.8281656548135299, "grad_norm": 0.2929024398326874, "learning_rate": 7.265704337733932e-05, "loss": 0.026, "step": 7639 }, { "epoch": 0.8282740676496098, "grad_norm": 0.10958729684352875, "learning_rate": 7.26534165095024e-05, "loss": 0.0072, "step": 7640 }, { "epoch": 0.8283824804856895, "grad_norm": 0.38596463203430176, "learning_rate": 7.264978964166547e-05, "loss": 0.0291, "step": 7641 }, { "epoch": 0.8284908933217693, "grad_norm": 0.43195030093193054, "learning_rate": 7.264616277382854e-05, "loss": 0.0623, "step": 7642 }, { "epoch": 0.828599306157849, "grad_norm": 0.15091648697853088, "learning_rate": 7.26425359059916e-05, "loss": 0.0189, "step": 7643 }, { "epoch": 0.8287077189939289, "grad_norm": 0.11090678721666336, "learning_rate": 7.263890903815465e-05, "loss": 0.0175, "step": 7644 }, { "epoch": 0.8288161318300087, "grad_norm": 0.41583380103111267, "learning_rate": 7.263528217031772e-05, "loss": 0.0399, "step": 7645 }, { "epoch": 0.8289245446660884, "grad_norm": 0.22199712693691254, "learning_rate": 7.263165530248078e-05, "loss": 0.0373, "step": 7646 }, { "epoch": 0.8290329575021682, "grad_norm": 0.33269622921943665, "learning_rate": 7.262802843464385e-05, "loss": 0.0657, "step": 7647 }, { "epoch": 0.8291413703382481, "grad_norm": 0.3304653465747833, "learning_rate": 7.262440156680691e-05, "loss": 0.0224, "step": 7648 }, { "epoch": 0.8292497831743278, "grad_norm": 0.18700991570949554, "learning_rate": 7.262077469896998e-05, "loss": 0.0202, "step": 7649 }, { "epoch": 0.8293581960104076, "grad_norm": 0.9915993809700012, "learning_rate": 7.261714783113304e-05, "loss": 0.0534, "step": 7650 }, { "epoch": 0.8294666088464874, "grad_norm": 0.5934516787528992, "learning_rate": 7.26135209632961e-05, "loss": 0.0532, "step": 7651 }, { "epoch": 0.8295750216825672, "grad_norm": 0.30092141032218933, "learning_rate": 7.260989409545916e-05, "loss": 0.0297, "step": 7652 }, { "epoch": 0.829683434518647, "grad_norm": 0.2518352270126343, "learning_rate": 7.260626722762222e-05, "loss": 0.0314, "step": 7653 }, { "epoch": 0.8297918473547268, "grad_norm": 0.23362517356872559, "learning_rate": 7.260264035978529e-05, "loss": 0.0173, "step": 7654 }, { "epoch": 0.8299002601908065, "grad_norm": 0.24185234308242798, "learning_rate": 7.259901349194835e-05, "loss": 0.0329, "step": 7655 }, { "epoch": 0.8300086730268864, "grad_norm": 0.6023398637771606, "learning_rate": 7.259538662411142e-05, "loss": 0.0586, "step": 7656 }, { "epoch": 0.8301170858629662, "grad_norm": 0.23470637202262878, "learning_rate": 7.259175975627448e-05, "loss": 0.0221, "step": 7657 }, { "epoch": 0.8302254986990459, "grad_norm": 0.08354002982378006, "learning_rate": 7.258813288843755e-05, "loss": 0.0092, "step": 7658 }, { "epoch": 0.8303339115351258, "grad_norm": 0.5720191597938538, "learning_rate": 7.258450602060062e-05, "loss": 0.0909, "step": 7659 }, { "epoch": 0.8304423243712056, "grad_norm": 0.5047959685325623, "learning_rate": 7.258087915276368e-05, "loss": 0.0301, "step": 7660 }, { "epoch": 0.8305507372072853, "grad_norm": 0.256146103143692, "learning_rate": 7.257725228492675e-05, "loss": 0.0604, "step": 7661 }, { "epoch": 0.8306591500433651, "grad_norm": 0.3177536427974701, "learning_rate": 7.25736254170898e-05, "loss": 0.0478, "step": 7662 }, { "epoch": 0.830767562879445, "grad_norm": 0.2063198983669281, "learning_rate": 7.256999854925288e-05, "loss": 0.0319, "step": 7663 }, { "epoch": 0.8308759757155247, "grad_norm": 0.17931108176708221, "learning_rate": 7.256637168141593e-05, "loss": 0.0188, "step": 7664 }, { "epoch": 0.8309843885516045, "grad_norm": 0.5709831118583679, "learning_rate": 7.2562744813579e-05, "loss": 0.0244, "step": 7665 }, { "epoch": 0.8310928013876843, "grad_norm": 0.16320253908634186, "learning_rate": 7.255911794574206e-05, "loss": 0.0237, "step": 7666 }, { "epoch": 0.8312012142237641, "grad_norm": 0.48042920231819153, "learning_rate": 7.255549107790512e-05, "loss": 0.0628, "step": 7667 }, { "epoch": 0.8313096270598439, "grad_norm": 0.8737472295761108, "learning_rate": 7.255186421006819e-05, "loss": 0.0693, "step": 7668 }, { "epoch": 0.8314180398959237, "grad_norm": 0.6023978590965271, "learning_rate": 7.254823734223125e-05, "loss": 0.0786, "step": 7669 }, { "epoch": 0.8315264527320034, "grad_norm": 0.5379429459571838, "learning_rate": 7.254461047439432e-05, "loss": 0.0751, "step": 7670 }, { "epoch": 0.8316348655680833, "grad_norm": 0.0335531085729599, "learning_rate": 7.254098360655738e-05, "loss": 0.0031, "step": 7671 }, { "epoch": 0.8317432784041631, "grad_norm": 0.23005817830562592, "learning_rate": 7.253735673872045e-05, "loss": 0.0274, "step": 7672 }, { "epoch": 0.8318516912402428, "grad_norm": 0.7351809740066528, "learning_rate": 7.25337298708835e-05, "loss": 0.0405, "step": 7673 }, { "epoch": 0.8319601040763226, "grad_norm": 0.3613972067832947, "learning_rate": 7.253010300304657e-05, "loss": 0.0812, "step": 7674 }, { "epoch": 0.8320685169124025, "grad_norm": 0.06247837468981743, "learning_rate": 7.252647613520963e-05, "loss": 0.0045, "step": 7675 }, { "epoch": 0.8321769297484822, "grad_norm": 0.1859143227338791, "learning_rate": 7.25228492673727e-05, "loss": 0.0211, "step": 7676 }, { "epoch": 0.832285342584562, "grad_norm": 0.25938671827316284, "learning_rate": 7.251922239953577e-05, "loss": 0.0281, "step": 7677 }, { "epoch": 0.8323937554206418, "grad_norm": 0.18843218684196472, "learning_rate": 7.251559553169883e-05, "loss": 0.0264, "step": 7678 }, { "epoch": 0.8325021682567216, "grad_norm": 0.5077552199363708, "learning_rate": 7.25119686638619e-05, "loss": 0.0406, "step": 7679 }, { "epoch": 0.8326105810928014, "grad_norm": 0.1842397302389145, "learning_rate": 7.250834179602496e-05, "loss": 0.0274, "step": 7680 }, { "epoch": 0.8327189939288812, "grad_norm": 0.2672308385372162, "learning_rate": 7.250471492818802e-05, "loss": 0.0342, "step": 7681 }, { "epoch": 0.8328274067649609, "grad_norm": 0.12814436852931976, "learning_rate": 7.250108806035109e-05, "loss": 0.0138, "step": 7682 }, { "epoch": 0.8329358196010408, "grad_norm": 0.4410320222377777, "learning_rate": 7.249746119251414e-05, "loss": 0.0461, "step": 7683 }, { "epoch": 0.8330442324371206, "grad_norm": 0.2491450160741806, "learning_rate": 7.249383432467722e-05, "loss": 0.03, "step": 7684 }, { "epoch": 0.8331526452732003, "grad_norm": 0.05518636852502823, "learning_rate": 7.249020745684027e-05, "loss": 0.0054, "step": 7685 }, { "epoch": 0.8332610581092801, "grad_norm": 0.49745216965675354, "learning_rate": 7.248658058900334e-05, "loss": 0.058, "step": 7686 }, { "epoch": 0.83336947094536, "grad_norm": 0.1600218117237091, "learning_rate": 7.24829537211664e-05, "loss": 0.014, "step": 7687 }, { "epoch": 0.8334778837814397, "grad_norm": 0.19789700210094452, "learning_rate": 7.247932685332947e-05, "loss": 0.0112, "step": 7688 }, { "epoch": 0.8335862966175195, "grad_norm": 0.4788045287132263, "learning_rate": 7.247569998549253e-05, "loss": 0.0393, "step": 7689 }, { "epoch": 0.8336947094535994, "grad_norm": 0.39857038855552673, "learning_rate": 7.247207311765559e-05, "loss": 0.0348, "step": 7690 }, { "epoch": 0.8338031222896791, "grad_norm": 0.2583707571029663, "learning_rate": 7.246844624981866e-05, "loss": 0.0209, "step": 7691 }, { "epoch": 0.8339115351257589, "grad_norm": 1.2615491151809692, "learning_rate": 7.246481938198171e-05, "loss": 0.0464, "step": 7692 }, { "epoch": 0.8340199479618386, "grad_norm": 0.30406486988067627, "learning_rate": 7.24611925141448e-05, "loss": 0.0244, "step": 7693 }, { "epoch": 0.8341283607979185, "grad_norm": 0.14707812666893005, "learning_rate": 7.245756564630786e-05, "loss": 0.0216, "step": 7694 }, { "epoch": 0.8342367736339983, "grad_norm": 0.2743329703807831, "learning_rate": 7.245393877847091e-05, "loss": 0.0269, "step": 7695 }, { "epoch": 0.834345186470078, "grad_norm": 0.23212863504886627, "learning_rate": 7.245031191063398e-05, "loss": 0.0217, "step": 7696 }, { "epoch": 0.8344535993061578, "grad_norm": 0.2404371052980423, "learning_rate": 7.244668504279704e-05, "loss": 0.0368, "step": 7697 }, { "epoch": 0.8345620121422377, "grad_norm": 0.3286944031715393, "learning_rate": 7.244305817496011e-05, "loss": 0.0194, "step": 7698 }, { "epoch": 0.8346704249783174, "grad_norm": 0.36598360538482666, "learning_rate": 7.243943130712317e-05, "loss": 0.0431, "step": 7699 }, { "epoch": 0.8347788378143972, "grad_norm": 0.521105170249939, "learning_rate": 7.243580443928624e-05, "loss": 0.0849, "step": 7700 }, { "epoch": 0.834887250650477, "grad_norm": 0.4103184938430786, "learning_rate": 7.24321775714493e-05, "loss": 0.0237, "step": 7701 }, { "epoch": 0.8349956634865568, "grad_norm": 0.600766658782959, "learning_rate": 7.242855070361237e-05, "loss": 0.0475, "step": 7702 }, { "epoch": 0.8351040763226366, "grad_norm": 0.761058509349823, "learning_rate": 7.242492383577543e-05, "loss": 0.1009, "step": 7703 }, { "epoch": 0.8352124891587164, "grad_norm": 0.4342188835144043, "learning_rate": 7.242129696793848e-05, "loss": 0.0619, "step": 7704 }, { "epoch": 0.8353209019947961, "grad_norm": 0.5069863796234131, "learning_rate": 7.241767010010155e-05, "loss": 0.0159, "step": 7705 }, { "epoch": 0.835429314830876, "grad_norm": 0.34311509132385254, "learning_rate": 7.241404323226461e-05, "loss": 0.0311, "step": 7706 }, { "epoch": 0.8355377276669558, "grad_norm": 0.5759938359260559, "learning_rate": 7.241041636442768e-05, "loss": 0.0338, "step": 7707 }, { "epoch": 0.8356461405030355, "grad_norm": 0.2958800196647644, "learning_rate": 7.240678949659074e-05, "loss": 0.0448, "step": 7708 }, { "epoch": 0.8357545533391153, "grad_norm": 0.5025086998939514, "learning_rate": 7.240316262875381e-05, "loss": 0.049, "step": 7709 }, { "epoch": 0.8358629661751952, "grad_norm": 0.2396065592765808, "learning_rate": 7.239953576091688e-05, "loss": 0.0422, "step": 7710 }, { "epoch": 0.8359713790112749, "grad_norm": 0.3240523338317871, "learning_rate": 7.239590889307994e-05, "loss": 0.0556, "step": 7711 }, { "epoch": 0.8360797918473547, "grad_norm": 0.47776755690574646, "learning_rate": 7.239228202524301e-05, "loss": 0.0326, "step": 7712 }, { "epoch": 0.8361882046834345, "grad_norm": 0.3423064947128296, "learning_rate": 7.238865515740607e-05, "loss": 0.0215, "step": 7713 }, { "epoch": 0.8362966175195143, "grad_norm": 0.1231142207980156, "learning_rate": 7.238502828956914e-05, "loss": 0.0119, "step": 7714 }, { "epoch": 0.8364050303555941, "grad_norm": 0.21301862597465515, "learning_rate": 7.23814014217322e-05, "loss": 0.0309, "step": 7715 }, { "epoch": 0.8365134431916739, "grad_norm": 0.3820579946041107, "learning_rate": 7.237777455389527e-05, "loss": 0.0327, "step": 7716 }, { "epoch": 0.8366218560277537, "grad_norm": 0.20902273058891296, "learning_rate": 7.237414768605832e-05, "loss": 0.0176, "step": 7717 }, { "epoch": 0.8367302688638335, "grad_norm": 0.3812560737133026, "learning_rate": 7.237052081822138e-05, "loss": 0.063, "step": 7718 }, { "epoch": 0.8368386816999133, "grad_norm": 0.14436480402946472, "learning_rate": 7.236689395038445e-05, "loss": 0.0165, "step": 7719 }, { "epoch": 0.836947094535993, "grad_norm": 0.3278065323829651, "learning_rate": 7.236326708254751e-05, "loss": 0.0291, "step": 7720 }, { "epoch": 0.8370555073720729, "grad_norm": 0.6152099370956421, "learning_rate": 7.235964021471058e-05, "loss": 0.0802, "step": 7721 }, { "epoch": 0.8371639202081527, "grad_norm": 0.63901287317276, "learning_rate": 7.235601334687364e-05, "loss": 0.0744, "step": 7722 }, { "epoch": 0.8372723330442324, "grad_norm": 0.2949674427509308, "learning_rate": 7.235238647903671e-05, "loss": 0.0486, "step": 7723 }, { "epoch": 0.8373807458803122, "grad_norm": 0.26356151700019836, "learning_rate": 7.234875961119977e-05, "loss": 0.0275, "step": 7724 }, { "epoch": 0.8374891587163921, "grad_norm": 0.2811974287033081, "learning_rate": 7.234513274336284e-05, "loss": 0.0368, "step": 7725 }, { "epoch": 0.8375975715524718, "grad_norm": 0.17514389753341675, "learning_rate": 7.23415058755259e-05, "loss": 0.0167, "step": 7726 }, { "epoch": 0.8377059843885516, "grad_norm": 0.22184449434280396, "learning_rate": 7.233787900768895e-05, "loss": 0.0295, "step": 7727 }, { "epoch": 0.8378143972246314, "grad_norm": 0.6344422698020935, "learning_rate": 7.233425213985204e-05, "loss": 0.0958, "step": 7728 }, { "epoch": 0.8379228100607112, "grad_norm": 0.7018696665763855, "learning_rate": 7.23306252720151e-05, "loss": 0.032, "step": 7729 }, { "epoch": 0.838031222896791, "grad_norm": 0.37632593512535095, "learning_rate": 7.232699840417816e-05, "loss": 0.0687, "step": 7730 }, { "epoch": 0.8381396357328708, "grad_norm": 0.2560047209262848, "learning_rate": 7.232337153634122e-05, "loss": 0.023, "step": 7731 }, { "epoch": 0.8382480485689505, "grad_norm": 0.2046143263578415, "learning_rate": 7.231974466850428e-05, "loss": 0.0181, "step": 7732 }, { "epoch": 0.8383564614050304, "grad_norm": 0.13976946473121643, "learning_rate": 7.231611780066735e-05, "loss": 0.0203, "step": 7733 }, { "epoch": 0.8384648742411102, "grad_norm": 0.19605188071727753, "learning_rate": 7.231249093283041e-05, "loss": 0.0121, "step": 7734 }, { "epoch": 0.8385732870771899, "grad_norm": 0.7222315073013306, "learning_rate": 7.230886406499348e-05, "loss": 0.0621, "step": 7735 }, { "epoch": 0.8386816999132697, "grad_norm": 0.21643507480621338, "learning_rate": 7.230523719715654e-05, "loss": 0.0207, "step": 7736 }, { "epoch": 0.8387901127493496, "grad_norm": 0.11851871013641357, "learning_rate": 7.23016103293196e-05, "loss": 0.0081, "step": 7737 }, { "epoch": 0.8388985255854293, "grad_norm": 0.1467389315366745, "learning_rate": 7.229798346148266e-05, "loss": 0.0085, "step": 7738 }, { "epoch": 0.8390069384215091, "grad_norm": 0.24452264606952667, "learning_rate": 7.229435659364573e-05, "loss": 0.0506, "step": 7739 }, { "epoch": 0.8391153512575888, "grad_norm": 0.3951411247253418, "learning_rate": 7.229072972580879e-05, "loss": 0.0354, "step": 7740 }, { "epoch": 0.8392237640936687, "grad_norm": 0.31376510858535767, "learning_rate": 7.228710285797185e-05, "loss": 0.0523, "step": 7741 }, { "epoch": 0.8393321769297485, "grad_norm": 0.49256592988967896, "learning_rate": 7.228347599013492e-05, "loss": 0.0148, "step": 7742 }, { "epoch": 0.8394405897658282, "grad_norm": 0.17799997329711914, "learning_rate": 7.227984912229798e-05, "loss": 0.0118, "step": 7743 }, { "epoch": 0.839549002601908, "grad_norm": 0.5062814354896545, "learning_rate": 7.227622225446105e-05, "loss": 0.0745, "step": 7744 }, { "epoch": 0.8396574154379879, "grad_norm": 0.4152430593967438, "learning_rate": 7.227259538662412e-05, "loss": 0.0469, "step": 7745 }, { "epoch": 0.8397658282740676, "grad_norm": 0.4962094724178314, "learning_rate": 7.226896851878719e-05, "loss": 0.079, "step": 7746 }, { "epoch": 0.8398742411101474, "grad_norm": 0.5558749437332153, "learning_rate": 7.226534165095025e-05, "loss": 0.0549, "step": 7747 }, { "epoch": 0.8399826539462273, "grad_norm": 0.3227306306362152, "learning_rate": 7.22617147831133e-05, "loss": 0.0355, "step": 7748 }, { "epoch": 0.840091066782307, "grad_norm": 0.4521106481552124, "learning_rate": 7.225808791527638e-05, "loss": 0.0252, "step": 7749 }, { "epoch": 0.8401994796183868, "grad_norm": 0.3407823145389557, "learning_rate": 7.225446104743943e-05, "loss": 0.0506, "step": 7750 }, { "epoch": 0.8403078924544666, "grad_norm": 0.3061803877353668, "learning_rate": 7.22508341796025e-05, "loss": 0.0297, "step": 7751 }, { "epoch": 0.8404163052905465, "grad_norm": 0.6261972188949585, "learning_rate": 7.224720731176556e-05, "loss": 0.0641, "step": 7752 }, { "epoch": 0.8405247181266262, "grad_norm": 0.6112002730369568, "learning_rate": 7.224358044392863e-05, "loss": 0.0833, "step": 7753 }, { "epoch": 0.840633130962706, "grad_norm": 0.4673513174057007, "learning_rate": 7.223995357609169e-05, "loss": 0.0709, "step": 7754 }, { "epoch": 0.8407415437987857, "grad_norm": 0.3733965754508972, "learning_rate": 7.223632670825476e-05, "loss": 0.0618, "step": 7755 }, { "epoch": 0.8408499566348656, "grad_norm": 0.22400198876857758, "learning_rate": 7.223269984041782e-05, "loss": 0.0257, "step": 7756 }, { "epoch": 0.8409583694709454, "grad_norm": 0.3376397490501404, "learning_rate": 7.222907297258087e-05, "loss": 0.0599, "step": 7757 }, { "epoch": 0.8410667823070251, "grad_norm": 0.4923191964626312, "learning_rate": 7.222544610474395e-05, "loss": 0.0424, "step": 7758 }, { "epoch": 0.8411751951431049, "grad_norm": 0.48475348949432373, "learning_rate": 7.2221819236907e-05, "loss": 0.0531, "step": 7759 }, { "epoch": 0.8412836079791848, "grad_norm": 0.18162964284420013, "learning_rate": 7.221819236907007e-05, "loss": 0.0062, "step": 7760 }, { "epoch": 0.8413920208152645, "grad_norm": 0.3633188307285309, "learning_rate": 7.221456550123313e-05, "loss": 0.0356, "step": 7761 }, { "epoch": 0.8415004336513443, "grad_norm": 0.24030764400959015, "learning_rate": 7.22109386333962e-05, "loss": 0.047, "step": 7762 }, { "epoch": 0.8416088464874241, "grad_norm": 1.0637694597244263, "learning_rate": 7.220731176555927e-05, "loss": 0.0508, "step": 7763 }, { "epoch": 0.8417172593235039, "grad_norm": 0.1371147483587265, "learning_rate": 7.220368489772233e-05, "loss": 0.0249, "step": 7764 }, { "epoch": 0.8418256721595837, "grad_norm": 0.5060508847236633, "learning_rate": 7.22000580298854e-05, "loss": 0.0445, "step": 7765 }, { "epoch": 0.8419340849956635, "grad_norm": 0.4770796298980713, "learning_rate": 7.219643116204846e-05, "loss": 0.0327, "step": 7766 }, { "epoch": 0.8420424978317432, "grad_norm": 0.7028550505638123, "learning_rate": 7.219280429421153e-05, "loss": 0.0698, "step": 7767 }, { "epoch": 0.8421509106678231, "grad_norm": 0.20855948328971863, "learning_rate": 7.218917742637459e-05, "loss": 0.0298, "step": 7768 }, { "epoch": 0.8422593235039029, "grad_norm": 0.2575656771659851, "learning_rate": 7.218555055853766e-05, "loss": 0.0176, "step": 7769 }, { "epoch": 0.8423677363399826, "grad_norm": 0.2098340541124344, "learning_rate": 7.218192369070071e-05, "loss": 0.0163, "step": 7770 }, { "epoch": 0.8424761491760624, "grad_norm": 0.717974066734314, "learning_rate": 7.217829682286377e-05, "loss": 0.0826, "step": 7771 }, { "epoch": 0.8425845620121423, "grad_norm": 0.15614643692970276, "learning_rate": 7.217466995502684e-05, "loss": 0.017, "step": 7772 }, { "epoch": 0.842692974848222, "grad_norm": 0.20037631690502167, "learning_rate": 7.21710430871899e-05, "loss": 0.0359, "step": 7773 }, { "epoch": 0.8428013876843018, "grad_norm": 0.15765686333179474, "learning_rate": 7.216741621935297e-05, "loss": 0.0217, "step": 7774 }, { "epoch": 0.8429098005203816, "grad_norm": 0.47634947299957275, "learning_rate": 7.216378935151603e-05, "loss": 0.0347, "step": 7775 }, { "epoch": 0.8430182133564614, "grad_norm": 0.29555219411849976, "learning_rate": 7.21601624836791e-05, "loss": 0.0359, "step": 7776 }, { "epoch": 0.8431266261925412, "grad_norm": 0.21320247650146484, "learning_rate": 7.215653561584216e-05, "loss": 0.0282, "step": 7777 }, { "epoch": 0.843235039028621, "grad_norm": 0.2910270094871521, "learning_rate": 7.215290874800523e-05, "loss": 0.0454, "step": 7778 }, { "epoch": 0.8433434518647008, "grad_norm": 0.5004189014434814, "learning_rate": 7.214928188016829e-05, "loss": 0.0556, "step": 7779 }, { "epoch": 0.8434518647007806, "grad_norm": 0.25753822922706604, "learning_rate": 7.214565501233136e-05, "loss": 0.0525, "step": 7780 }, { "epoch": 0.8435602775368604, "grad_norm": 0.5094240307807922, "learning_rate": 7.214202814449443e-05, "loss": 0.0291, "step": 7781 }, { "epoch": 0.8436686903729401, "grad_norm": 0.5212437510490417, "learning_rate": 7.213840127665748e-05, "loss": 0.0357, "step": 7782 }, { "epoch": 0.84377710320902, "grad_norm": 0.2266434282064438, "learning_rate": 7.213477440882056e-05, "loss": 0.0302, "step": 7783 }, { "epoch": 0.8438855160450998, "grad_norm": 0.6374713182449341, "learning_rate": 7.213114754098361e-05, "loss": 0.0639, "step": 7784 }, { "epoch": 0.8439939288811795, "grad_norm": 0.26500818133354187, "learning_rate": 7.212752067314667e-05, "loss": 0.0468, "step": 7785 }, { "epoch": 0.8441023417172593, "grad_norm": 0.1986023336648941, "learning_rate": 7.212389380530974e-05, "loss": 0.0235, "step": 7786 }, { "epoch": 0.8442107545533392, "grad_norm": 0.36100465059280396, "learning_rate": 7.21202669374728e-05, "loss": 0.0585, "step": 7787 }, { "epoch": 0.8443191673894189, "grad_norm": 0.4836834967136383, "learning_rate": 7.211664006963587e-05, "loss": 0.0351, "step": 7788 }, { "epoch": 0.8444275802254987, "grad_norm": 0.21312978863716125, "learning_rate": 7.211301320179893e-05, "loss": 0.0258, "step": 7789 }, { "epoch": 0.8445359930615784, "grad_norm": 0.46299445629119873, "learning_rate": 7.2109386333962e-05, "loss": 0.0943, "step": 7790 }, { "epoch": 0.8446444058976583, "grad_norm": 0.43727418780326843, "learning_rate": 7.210575946612505e-05, "loss": 0.0401, "step": 7791 }, { "epoch": 0.8447528187337381, "grad_norm": 0.5648466348648071, "learning_rate": 7.210213259828813e-05, "loss": 0.0516, "step": 7792 }, { "epoch": 0.8448612315698178, "grad_norm": 0.45540204644203186, "learning_rate": 7.209850573045118e-05, "loss": 0.0274, "step": 7793 }, { "epoch": 0.8449696444058976, "grad_norm": 0.29791465401649475, "learning_rate": 7.209487886261424e-05, "loss": 0.0318, "step": 7794 }, { "epoch": 0.8450780572419775, "grad_norm": 0.09287901222705841, "learning_rate": 7.209125199477731e-05, "loss": 0.0148, "step": 7795 }, { "epoch": 0.8451864700780573, "grad_norm": 0.16957996785640717, "learning_rate": 7.208762512694037e-05, "loss": 0.0285, "step": 7796 }, { "epoch": 0.845294882914137, "grad_norm": 0.21988122165203094, "learning_rate": 7.208399825910345e-05, "loss": 0.0342, "step": 7797 }, { "epoch": 0.8454032957502168, "grad_norm": 0.34720146656036377, "learning_rate": 7.208037139126651e-05, "loss": 0.0298, "step": 7798 }, { "epoch": 0.8455117085862967, "grad_norm": 0.5192841291427612, "learning_rate": 7.207674452342957e-05, "loss": 0.0276, "step": 7799 }, { "epoch": 0.8456201214223764, "grad_norm": 0.21936936676502228, "learning_rate": 7.207311765559264e-05, "loss": 0.0184, "step": 7800 }, { "epoch": 0.8457285342584562, "grad_norm": 1.158479928970337, "learning_rate": 7.20694907877557e-05, "loss": 0.0955, "step": 7801 }, { "epoch": 0.8458369470945359, "grad_norm": 1.0409214496612549, "learning_rate": 7.206586391991877e-05, "loss": 0.0901, "step": 7802 }, { "epoch": 0.8459453599306158, "grad_norm": 0.24370260536670685, "learning_rate": 7.206223705208182e-05, "loss": 0.0356, "step": 7803 }, { "epoch": 0.8460537727666956, "grad_norm": 0.38403967022895813, "learning_rate": 7.20586101842449e-05, "loss": 0.0529, "step": 7804 }, { "epoch": 0.8461621856027753, "grad_norm": 0.17293792963027954, "learning_rate": 7.205498331640795e-05, "loss": 0.0326, "step": 7805 }, { "epoch": 0.8462705984388552, "grad_norm": 0.34503886103630066, "learning_rate": 7.205135644857102e-05, "loss": 0.0593, "step": 7806 }, { "epoch": 0.846379011274935, "grad_norm": 0.18210016191005707, "learning_rate": 7.204772958073408e-05, "loss": 0.0198, "step": 7807 }, { "epoch": 0.8464874241110147, "grad_norm": 0.14144490659236908, "learning_rate": 7.204410271289714e-05, "loss": 0.02, "step": 7808 }, { "epoch": 0.8465958369470945, "grad_norm": 0.4657178819179535, "learning_rate": 7.204047584506021e-05, "loss": 0.0754, "step": 7809 }, { "epoch": 0.8467042497831744, "grad_norm": 0.24873416125774384, "learning_rate": 7.203684897722327e-05, "loss": 0.0247, "step": 7810 }, { "epoch": 0.8468126626192541, "grad_norm": 0.2953799068927765, "learning_rate": 7.203322210938634e-05, "loss": 0.0359, "step": 7811 }, { "epoch": 0.8469210754553339, "grad_norm": 0.2727367579936981, "learning_rate": 7.20295952415494e-05, "loss": 0.052, "step": 7812 }, { "epoch": 0.8470294882914137, "grad_norm": 0.2444024533033371, "learning_rate": 7.202596837371246e-05, "loss": 0.0232, "step": 7813 }, { "epoch": 0.8471379011274935, "grad_norm": 0.36054080724716187, "learning_rate": 7.202234150587554e-05, "loss": 0.0748, "step": 7814 }, { "epoch": 0.8472463139635733, "grad_norm": 0.6222077012062073, "learning_rate": 7.201871463803859e-05, "loss": 0.0354, "step": 7815 }, { "epoch": 0.8473547267996531, "grad_norm": 0.26554590463638306, "learning_rate": 7.201508777020166e-05, "loss": 0.0354, "step": 7816 }, { "epoch": 0.8474631396357328, "grad_norm": 0.4039090871810913, "learning_rate": 7.201146090236472e-05, "loss": 0.0573, "step": 7817 }, { "epoch": 0.8475715524718127, "grad_norm": 0.1434599757194519, "learning_rate": 7.200783403452779e-05, "loss": 0.019, "step": 7818 }, { "epoch": 0.8476799653078925, "grad_norm": 0.18429294228553772, "learning_rate": 7.200420716669085e-05, "loss": 0.0184, "step": 7819 }, { "epoch": 0.8477883781439722, "grad_norm": 0.3511143624782562, "learning_rate": 7.200058029885392e-05, "loss": 0.0377, "step": 7820 }, { "epoch": 0.847896790980052, "grad_norm": 0.17374925315380096, "learning_rate": 7.199695343101698e-05, "loss": 0.0181, "step": 7821 }, { "epoch": 0.8480052038161319, "grad_norm": 0.29442068934440613, "learning_rate": 7.199332656318003e-05, "loss": 0.0264, "step": 7822 }, { "epoch": 0.8481136166522116, "grad_norm": 0.14637963473796844, "learning_rate": 7.19896996953431e-05, "loss": 0.0114, "step": 7823 }, { "epoch": 0.8482220294882914, "grad_norm": 0.5796815156936646, "learning_rate": 7.198607282750616e-05, "loss": 0.0295, "step": 7824 }, { "epoch": 0.8483304423243712, "grad_norm": 0.20966923236846924, "learning_rate": 7.198244595966923e-05, "loss": 0.031, "step": 7825 }, { "epoch": 0.848438855160451, "grad_norm": 0.20607620477676392, "learning_rate": 7.197881909183229e-05, "loss": 0.0176, "step": 7826 }, { "epoch": 0.8485472679965308, "grad_norm": 0.9750752449035645, "learning_rate": 7.197519222399536e-05, "loss": 0.0817, "step": 7827 }, { "epoch": 0.8486556808326106, "grad_norm": 0.20916321873664856, "learning_rate": 7.197156535615842e-05, "loss": 0.0317, "step": 7828 }, { "epoch": 0.8487640936686903, "grad_norm": 0.12911328673362732, "learning_rate": 7.196793848832149e-05, "loss": 0.0151, "step": 7829 }, { "epoch": 0.8488725065047702, "grad_norm": 0.3486631512641907, "learning_rate": 7.196431162048455e-05, "loss": 0.0328, "step": 7830 }, { "epoch": 0.84898091934085, "grad_norm": 0.11257288604974747, "learning_rate": 7.196068475264762e-05, "loss": 0.014, "step": 7831 }, { "epoch": 0.8490893321769297, "grad_norm": 0.29631367325782776, "learning_rate": 7.195705788481069e-05, "loss": 0.0326, "step": 7832 }, { "epoch": 0.8491977450130095, "grad_norm": 0.35641077160835266, "learning_rate": 7.195343101697375e-05, "loss": 0.0288, "step": 7833 }, { "epoch": 0.8493061578490894, "grad_norm": 0.3582817614078522, "learning_rate": 7.194980414913682e-05, "loss": 0.0418, "step": 7834 }, { "epoch": 0.8494145706851691, "grad_norm": 0.2956281304359436, "learning_rate": 7.194617728129988e-05, "loss": 0.0462, "step": 7835 }, { "epoch": 0.8495229835212489, "grad_norm": 0.3901764154434204, "learning_rate": 7.194255041346293e-05, "loss": 0.0376, "step": 7836 }, { "epoch": 0.8496313963573288, "grad_norm": 0.2098008692264557, "learning_rate": 7.1938923545626e-05, "loss": 0.022, "step": 7837 }, { "epoch": 0.8497398091934085, "grad_norm": 0.3821038007736206, "learning_rate": 7.193529667778906e-05, "loss": 0.0496, "step": 7838 }, { "epoch": 0.8498482220294883, "grad_norm": 0.12589268386363983, "learning_rate": 7.193166980995213e-05, "loss": 0.0153, "step": 7839 }, { "epoch": 0.849956634865568, "grad_norm": 0.14736531674861908, "learning_rate": 7.192804294211519e-05, "loss": 0.0131, "step": 7840 }, { "epoch": 0.8500650477016479, "grad_norm": 0.11629168689250946, "learning_rate": 7.192441607427826e-05, "loss": 0.0102, "step": 7841 }, { "epoch": 0.8501734605377277, "grad_norm": 0.34912553429603577, "learning_rate": 7.192078920644132e-05, "loss": 0.0342, "step": 7842 }, { "epoch": 0.8502818733738075, "grad_norm": 0.3228073716163635, "learning_rate": 7.191716233860439e-05, "loss": 0.0306, "step": 7843 }, { "epoch": 0.8503902862098872, "grad_norm": 0.42812439799308777, "learning_rate": 7.191353547076745e-05, "loss": 0.0623, "step": 7844 }, { "epoch": 0.8504986990459671, "grad_norm": 0.3359079658985138, "learning_rate": 7.19099086029305e-05, "loss": 0.0644, "step": 7845 }, { "epoch": 0.8506071118820469, "grad_norm": 0.22616176307201385, "learning_rate": 7.190628173509357e-05, "loss": 0.0546, "step": 7846 }, { "epoch": 0.8507155247181266, "grad_norm": 0.36027413606643677, "learning_rate": 7.190265486725663e-05, "loss": 0.0219, "step": 7847 }, { "epoch": 0.8508239375542064, "grad_norm": 0.3196447789669037, "learning_rate": 7.18990279994197e-05, "loss": 0.0392, "step": 7848 }, { "epoch": 0.8509323503902863, "grad_norm": 0.1384066492319107, "learning_rate": 7.189540113158277e-05, "loss": 0.0251, "step": 7849 }, { "epoch": 0.851040763226366, "grad_norm": 0.44724032282829285, "learning_rate": 7.189177426374584e-05, "loss": 0.0304, "step": 7850 }, { "epoch": 0.8511491760624458, "grad_norm": 0.515261709690094, "learning_rate": 7.18881473959089e-05, "loss": 0.0339, "step": 7851 }, { "epoch": 0.8512575888985255, "grad_norm": 0.3015488386154175, "learning_rate": 7.188452052807196e-05, "loss": 0.0275, "step": 7852 }, { "epoch": 0.8513660017346054, "grad_norm": 0.2611617147922516, "learning_rate": 7.188089366023503e-05, "loss": 0.0338, "step": 7853 }, { "epoch": 0.8514744145706852, "grad_norm": 0.4457266628742218, "learning_rate": 7.187726679239809e-05, "loss": 0.0438, "step": 7854 }, { "epoch": 0.8515828274067649, "grad_norm": 0.3273684084415436, "learning_rate": 7.187363992456116e-05, "loss": 0.0356, "step": 7855 }, { "epoch": 0.8516912402428447, "grad_norm": 0.4651291072368622, "learning_rate": 7.187001305672421e-05, "loss": 0.0509, "step": 7856 }, { "epoch": 0.8517996530789246, "grad_norm": 0.37336039543151855, "learning_rate": 7.186638618888729e-05, "loss": 0.0526, "step": 7857 }, { "epoch": 0.8519080659150043, "grad_norm": 0.4704641103744507, "learning_rate": 7.186275932105034e-05, "loss": 0.0359, "step": 7858 }, { "epoch": 0.8520164787510841, "grad_norm": 0.0480831079185009, "learning_rate": 7.185913245321341e-05, "loss": 0.0048, "step": 7859 }, { "epoch": 0.8521248915871639, "grad_norm": 0.4984639286994934, "learning_rate": 7.185550558537647e-05, "loss": 0.0519, "step": 7860 }, { "epoch": 0.8522333044232437, "grad_norm": 0.30409327149391174, "learning_rate": 7.185187871753953e-05, "loss": 0.0432, "step": 7861 }, { "epoch": 0.8523417172593235, "grad_norm": 0.49260279536247253, "learning_rate": 7.18482518497026e-05, "loss": 0.0479, "step": 7862 }, { "epoch": 0.8524501300954033, "grad_norm": 0.1185573935508728, "learning_rate": 7.184462498186566e-05, "loss": 0.0224, "step": 7863 }, { "epoch": 0.852558542931483, "grad_norm": 0.06405515968799591, "learning_rate": 7.184099811402873e-05, "loss": 0.013, "step": 7864 }, { "epoch": 0.8526669557675629, "grad_norm": 0.33205801248550415, "learning_rate": 7.183737124619178e-05, "loss": 0.0362, "step": 7865 }, { "epoch": 0.8527753686036427, "grad_norm": 0.1805790811777115, "learning_rate": 7.183374437835486e-05, "loss": 0.0103, "step": 7866 }, { "epoch": 0.8528837814397224, "grad_norm": 0.2606457471847534, "learning_rate": 7.183011751051793e-05, "loss": 0.0478, "step": 7867 }, { "epoch": 0.8529921942758023, "grad_norm": 0.13567788898944855, "learning_rate": 7.182649064268098e-05, "loss": 0.0217, "step": 7868 }, { "epoch": 0.8531006071118821, "grad_norm": 0.3856970965862274, "learning_rate": 7.182286377484405e-05, "loss": 0.0596, "step": 7869 }, { "epoch": 0.8532090199479618, "grad_norm": 0.2670956254005432, "learning_rate": 7.181923690700711e-05, "loss": 0.0432, "step": 7870 }, { "epoch": 0.8533174327840416, "grad_norm": 0.4268929958343506, "learning_rate": 7.181561003917018e-05, "loss": 0.0706, "step": 7871 }, { "epoch": 0.8534258456201215, "grad_norm": 0.34113067388534546, "learning_rate": 7.181198317133324e-05, "loss": 0.0173, "step": 7872 }, { "epoch": 0.8535342584562012, "grad_norm": 0.2809220552444458, "learning_rate": 7.180835630349631e-05, "loss": 0.0275, "step": 7873 }, { "epoch": 0.853642671292281, "grad_norm": 0.30818650126457214, "learning_rate": 7.180472943565937e-05, "loss": 0.0555, "step": 7874 }, { "epoch": 0.8537510841283608, "grad_norm": 0.4292754828929901, "learning_rate": 7.180110256782243e-05, "loss": 0.048, "step": 7875 }, { "epoch": 0.8538594969644406, "grad_norm": 0.5024129152297974, "learning_rate": 7.17974756999855e-05, "loss": 0.0725, "step": 7876 }, { "epoch": 0.8539679098005204, "grad_norm": 0.36688971519470215, "learning_rate": 7.179384883214855e-05, "loss": 0.0667, "step": 7877 }, { "epoch": 0.8540763226366002, "grad_norm": 0.2160617709159851, "learning_rate": 7.179022196431162e-05, "loss": 0.0288, "step": 7878 }, { "epoch": 0.8541847354726799, "grad_norm": 0.27176812291145325, "learning_rate": 7.178659509647468e-05, "loss": 0.0519, "step": 7879 }, { "epoch": 0.8542931483087598, "grad_norm": 0.20744864642620087, "learning_rate": 7.178296822863775e-05, "loss": 0.0479, "step": 7880 }, { "epoch": 0.8544015611448396, "grad_norm": 0.21954505145549774, "learning_rate": 7.177934136080081e-05, "loss": 0.0173, "step": 7881 }, { "epoch": 0.8545099739809193, "grad_norm": 0.1610851287841797, "learning_rate": 7.177571449296388e-05, "loss": 0.0201, "step": 7882 }, { "epoch": 0.8546183868169991, "grad_norm": 0.3231917917728424, "learning_rate": 7.177208762512695e-05, "loss": 0.0374, "step": 7883 }, { "epoch": 0.854726799653079, "grad_norm": 0.3376995623111725, "learning_rate": 7.176846075729001e-05, "loss": 0.0352, "step": 7884 }, { "epoch": 0.8548352124891587, "grad_norm": 0.293513685464859, "learning_rate": 7.176483388945308e-05, "loss": 0.0573, "step": 7885 }, { "epoch": 0.8549436253252385, "grad_norm": 0.17527370154857635, "learning_rate": 7.176120702161614e-05, "loss": 0.0231, "step": 7886 }, { "epoch": 0.8550520381613183, "grad_norm": 0.4152143895626068, "learning_rate": 7.175758015377921e-05, "loss": 0.0576, "step": 7887 }, { "epoch": 0.8551604509973981, "grad_norm": 0.2326209843158722, "learning_rate": 7.175395328594227e-05, "loss": 0.0366, "step": 7888 }, { "epoch": 0.8552688638334779, "grad_norm": 0.3531210422515869, "learning_rate": 7.175032641810532e-05, "loss": 0.055, "step": 7889 }, { "epoch": 0.8553772766695577, "grad_norm": 0.25465163588523865, "learning_rate": 7.17466995502684e-05, "loss": 0.0234, "step": 7890 }, { "epoch": 0.8554856895056374, "grad_norm": 0.525799036026001, "learning_rate": 7.174307268243145e-05, "loss": 0.0474, "step": 7891 }, { "epoch": 0.8555941023417173, "grad_norm": 0.45567506551742554, "learning_rate": 7.173944581459452e-05, "loss": 0.0484, "step": 7892 }, { "epoch": 0.855702515177797, "grad_norm": 0.8283542394638062, "learning_rate": 7.173581894675758e-05, "loss": 0.0417, "step": 7893 }, { "epoch": 0.8558109280138768, "grad_norm": 0.1507374346256256, "learning_rate": 7.173219207892065e-05, "loss": 0.0164, "step": 7894 }, { "epoch": 0.8559193408499567, "grad_norm": 0.5585081577301025, "learning_rate": 7.172856521108371e-05, "loss": 0.0594, "step": 7895 }, { "epoch": 0.8560277536860365, "grad_norm": 0.20491500198841095, "learning_rate": 7.172493834324678e-05, "loss": 0.0226, "step": 7896 }, { "epoch": 0.8561361665221162, "grad_norm": 0.31404662132263184, "learning_rate": 7.172131147540984e-05, "loss": 0.0602, "step": 7897 }, { "epoch": 0.856244579358196, "grad_norm": 0.30783724784851074, "learning_rate": 7.17176846075729e-05, "loss": 0.0402, "step": 7898 }, { "epoch": 0.8563529921942759, "grad_norm": 0.3256929814815521, "learning_rate": 7.171405773973596e-05, "loss": 0.0219, "step": 7899 }, { "epoch": 0.8564614050303556, "grad_norm": 0.8732761740684509, "learning_rate": 7.171043087189904e-05, "loss": 0.0751, "step": 7900 }, { "epoch": 0.8565698178664354, "grad_norm": 0.31488245725631714, "learning_rate": 7.17068040040621e-05, "loss": 0.0361, "step": 7901 }, { "epoch": 0.8566782307025151, "grad_norm": 0.2891712486743927, "learning_rate": 7.170317713622516e-05, "loss": 0.0347, "step": 7902 }, { "epoch": 0.856786643538595, "grad_norm": 0.29300957918167114, "learning_rate": 7.169955026838822e-05, "loss": 0.0574, "step": 7903 }, { "epoch": 0.8568950563746748, "grad_norm": 0.21209442615509033, "learning_rate": 7.169592340055129e-05, "loss": 0.0177, "step": 7904 }, { "epoch": 0.8570034692107545, "grad_norm": 0.5623254179954529, "learning_rate": 7.169229653271435e-05, "loss": 0.0509, "step": 7905 }, { "epoch": 0.8571118820468343, "grad_norm": 0.39123907685279846, "learning_rate": 7.168866966487742e-05, "loss": 0.0852, "step": 7906 }, { "epoch": 0.8572202948829142, "grad_norm": 0.7867835164070129, "learning_rate": 7.168504279704048e-05, "loss": 0.0721, "step": 7907 }, { "epoch": 0.8573287077189939, "grad_norm": 0.24907873570919037, "learning_rate": 7.168141592920355e-05, "loss": 0.0366, "step": 7908 }, { "epoch": 0.8574371205550737, "grad_norm": 0.1005852073431015, "learning_rate": 7.16777890613666e-05, "loss": 0.0128, "step": 7909 }, { "epoch": 0.8575455333911535, "grad_norm": 0.3272286057472229, "learning_rate": 7.167416219352968e-05, "loss": 0.033, "step": 7910 }, { "epoch": 0.8576539462272333, "grad_norm": 0.11837544292211533, "learning_rate": 7.167053532569273e-05, "loss": 0.0273, "step": 7911 }, { "epoch": 0.8577623590633131, "grad_norm": 0.3854758143424988, "learning_rate": 7.166690845785579e-05, "loss": 0.0708, "step": 7912 }, { "epoch": 0.8578707718993929, "grad_norm": 0.2086605578660965, "learning_rate": 7.166328159001886e-05, "loss": 0.0288, "step": 7913 }, { "epoch": 0.8579791847354726, "grad_norm": 0.34908896684646606, "learning_rate": 7.165965472218192e-05, "loss": 0.0516, "step": 7914 }, { "epoch": 0.8580875975715525, "grad_norm": 0.7499718070030212, "learning_rate": 7.165602785434499e-05, "loss": 0.0655, "step": 7915 }, { "epoch": 0.8581960104076323, "grad_norm": 0.15448668599128723, "learning_rate": 7.165240098650805e-05, "loss": 0.0179, "step": 7916 }, { "epoch": 0.858304423243712, "grad_norm": 0.34385886788368225, "learning_rate": 7.164877411867112e-05, "loss": 0.0658, "step": 7917 }, { "epoch": 0.8584128360797918, "grad_norm": 0.22843697667121887, "learning_rate": 7.164514725083419e-05, "loss": 0.0334, "step": 7918 }, { "epoch": 0.8585212489158717, "grad_norm": 0.20750950276851654, "learning_rate": 7.164152038299725e-05, "loss": 0.0377, "step": 7919 }, { "epoch": 0.8586296617519514, "grad_norm": 0.15393780171871185, "learning_rate": 7.163789351516032e-05, "loss": 0.0241, "step": 7920 }, { "epoch": 0.8587380745880312, "grad_norm": 0.27815577387809753, "learning_rate": 7.163426664732337e-05, "loss": 0.0252, "step": 7921 }, { "epoch": 0.858846487424111, "grad_norm": 0.12580548226833344, "learning_rate": 7.163063977948645e-05, "loss": 0.0168, "step": 7922 }, { "epoch": 0.8589549002601908, "grad_norm": 0.3659466803073883, "learning_rate": 7.16270129116495e-05, "loss": 0.0353, "step": 7923 }, { "epoch": 0.8590633130962706, "grad_norm": 0.3410642445087433, "learning_rate": 7.162338604381257e-05, "loss": 0.0331, "step": 7924 }, { "epoch": 0.8591717259323504, "grad_norm": 0.8570126891136169, "learning_rate": 7.161975917597563e-05, "loss": 0.0941, "step": 7925 }, { "epoch": 0.8592801387684302, "grad_norm": 0.3893715441226959, "learning_rate": 7.161613230813869e-05, "loss": 0.0579, "step": 7926 }, { "epoch": 0.85938855160451, "grad_norm": 0.39035898447036743, "learning_rate": 7.161250544030176e-05, "loss": 0.0499, "step": 7927 }, { "epoch": 0.8594969644405898, "grad_norm": 0.5164640545845032, "learning_rate": 7.160887857246482e-05, "loss": 0.0569, "step": 7928 }, { "epoch": 0.8596053772766695, "grad_norm": 0.1568494737148285, "learning_rate": 7.160525170462789e-05, "loss": 0.0273, "step": 7929 }, { "epoch": 0.8597137901127494, "grad_norm": 0.3551728129386902, "learning_rate": 7.160162483679094e-05, "loss": 0.0336, "step": 7930 }, { "epoch": 0.8598222029488292, "grad_norm": 0.4359820783138275, "learning_rate": 7.159799796895402e-05, "loss": 0.0556, "step": 7931 }, { "epoch": 0.8599306157849089, "grad_norm": 0.46160370111465454, "learning_rate": 7.159437110111707e-05, "loss": 0.0329, "step": 7932 }, { "epoch": 0.8600390286209887, "grad_norm": 0.5090187191963196, "learning_rate": 7.159074423328014e-05, "loss": 0.0452, "step": 7933 }, { "epoch": 0.8601474414570686, "grad_norm": 0.1514768749475479, "learning_rate": 7.15871173654432e-05, "loss": 0.0144, "step": 7934 }, { "epoch": 0.8602558542931483, "grad_norm": 0.2565727233886719, "learning_rate": 7.158349049760627e-05, "loss": 0.033, "step": 7935 }, { "epoch": 0.8603642671292281, "grad_norm": 0.3199026882648468, "learning_rate": 7.157986362976934e-05, "loss": 0.0274, "step": 7936 }, { "epoch": 0.8604726799653079, "grad_norm": 0.2423182725906372, "learning_rate": 7.15762367619324e-05, "loss": 0.0263, "step": 7937 }, { "epoch": 0.8605810928013877, "grad_norm": 0.14941741526126862, "learning_rate": 7.157260989409547e-05, "loss": 0.0163, "step": 7938 }, { "epoch": 0.8606895056374675, "grad_norm": 0.3979519307613373, "learning_rate": 7.156898302625853e-05, "loss": 0.0301, "step": 7939 }, { "epoch": 0.8607979184735473, "grad_norm": 0.2838309705257416, "learning_rate": 7.15653561584216e-05, "loss": 0.0363, "step": 7940 }, { "epoch": 0.860906331309627, "grad_norm": 0.4543308913707733, "learning_rate": 7.156172929058466e-05, "loss": 0.0567, "step": 7941 }, { "epoch": 0.8610147441457069, "grad_norm": 0.19811449944972992, "learning_rate": 7.155810242274771e-05, "loss": 0.0191, "step": 7942 }, { "epoch": 0.8611231569817867, "grad_norm": 0.43062201142311096, "learning_rate": 7.155447555491078e-05, "loss": 0.0468, "step": 7943 }, { "epoch": 0.8612315698178664, "grad_norm": 0.2551100552082062, "learning_rate": 7.155084868707384e-05, "loss": 0.0236, "step": 7944 }, { "epoch": 0.8613399826539462, "grad_norm": 0.47873517870903015, "learning_rate": 7.154722181923691e-05, "loss": 0.0524, "step": 7945 }, { "epoch": 0.861448395490026, "grad_norm": 0.43938761949539185, "learning_rate": 7.154359495139997e-05, "loss": 0.0531, "step": 7946 }, { "epoch": 0.8615568083261058, "grad_norm": 0.2492682784795761, "learning_rate": 7.153996808356304e-05, "loss": 0.0312, "step": 7947 }, { "epoch": 0.8616652211621856, "grad_norm": 0.08735242486000061, "learning_rate": 7.15363412157261e-05, "loss": 0.0182, "step": 7948 }, { "epoch": 0.8617736339982653, "grad_norm": 0.13736025989055634, "learning_rate": 7.153271434788916e-05, "loss": 0.0086, "step": 7949 }, { "epoch": 0.8618820468343452, "grad_norm": 0.7588695883750916, "learning_rate": 7.152908748005223e-05, "loss": 0.0298, "step": 7950 }, { "epoch": 0.861990459670425, "grad_norm": 0.21487945318222046, "learning_rate": 7.152546061221528e-05, "loss": 0.027, "step": 7951 }, { "epoch": 0.8620988725065047, "grad_norm": 0.30324938893318176, "learning_rate": 7.152183374437837e-05, "loss": 0.0484, "step": 7952 }, { "epoch": 0.8622072853425846, "grad_norm": 1.0095329284667969, "learning_rate": 7.151820687654143e-05, "loss": 0.0727, "step": 7953 }, { "epoch": 0.8623156981786644, "grad_norm": 0.294229656457901, "learning_rate": 7.15145800087045e-05, "loss": 0.0262, "step": 7954 }, { "epoch": 0.8624241110147441, "grad_norm": 0.3563244044780731, "learning_rate": 7.151095314086755e-05, "loss": 0.0346, "step": 7955 }, { "epoch": 0.8625325238508239, "grad_norm": 0.28398019075393677, "learning_rate": 7.150732627303061e-05, "loss": 0.0192, "step": 7956 }, { "epoch": 0.8626409366869038, "grad_norm": 0.09819366037845612, "learning_rate": 7.150369940519368e-05, "loss": 0.0114, "step": 7957 }, { "epoch": 0.8627493495229835, "grad_norm": 0.4365837574005127, "learning_rate": 7.150007253735674e-05, "loss": 0.0433, "step": 7958 }, { "epoch": 0.8628577623590633, "grad_norm": 0.4212956428527832, "learning_rate": 7.149644566951981e-05, "loss": 0.0333, "step": 7959 }, { "epoch": 0.8629661751951431, "grad_norm": 1.0507704019546509, "learning_rate": 7.149281880168287e-05, "loss": 0.0906, "step": 7960 }, { "epoch": 0.863074588031223, "grad_norm": 0.6130009889602661, "learning_rate": 7.148919193384594e-05, "loss": 0.0746, "step": 7961 }, { "epoch": 0.8631830008673027, "grad_norm": 0.15337888896465302, "learning_rate": 7.1485565066009e-05, "loss": 0.0236, "step": 7962 }, { "epoch": 0.8632914137033825, "grad_norm": 0.5049077868461609, "learning_rate": 7.148193819817207e-05, "loss": 0.0537, "step": 7963 }, { "epoch": 0.8633998265394622, "grad_norm": 0.48606643080711365, "learning_rate": 7.147831133033512e-05, "loss": 0.1082, "step": 7964 }, { "epoch": 0.8635082393755421, "grad_norm": 0.3304225206375122, "learning_rate": 7.147468446249818e-05, "loss": 0.0484, "step": 7965 }, { "epoch": 0.8636166522116219, "grad_norm": 0.16096839308738708, "learning_rate": 7.147105759466125e-05, "loss": 0.0275, "step": 7966 }, { "epoch": 0.8637250650477016, "grad_norm": 0.36994609236717224, "learning_rate": 7.146743072682431e-05, "loss": 0.0612, "step": 7967 }, { "epoch": 0.8638334778837814, "grad_norm": 0.10857696086168289, "learning_rate": 7.146380385898738e-05, "loss": 0.0158, "step": 7968 }, { "epoch": 0.8639418907198613, "grad_norm": 0.9116649031639099, "learning_rate": 7.146017699115044e-05, "loss": 0.0581, "step": 7969 }, { "epoch": 0.864050303555941, "grad_norm": 0.2500959634780884, "learning_rate": 7.145655012331351e-05, "loss": 0.025, "step": 7970 }, { "epoch": 0.8641587163920208, "grad_norm": 0.32265493273735046, "learning_rate": 7.145292325547658e-05, "loss": 0.0231, "step": 7971 }, { "epoch": 0.8642671292281006, "grad_norm": 0.2419857680797577, "learning_rate": 7.144929638763964e-05, "loss": 0.0346, "step": 7972 }, { "epoch": 0.8643755420641804, "grad_norm": 0.20072594285011292, "learning_rate": 7.144566951980271e-05, "loss": 0.0191, "step": 7973 }, { "epoch": 0.8644839549002602, "grad_norm": 0.035242341458797455, "learning_rate": 7.144204265196577e-05, "loss": 0.0028, "step": 7974 }, { "epoch": 0.86459236773634, "grad_norm": 0.24315012991428375, "learning_rate": 7.143841578412884e-05, "loss": 0.0465, "step": 7975 }, { "epoch": 0.8647007805724197, "grad_norm": 0.4030124843120575, "learning_rate": 7.14347889162919e-05, "loss": 0.0598, "step": 7976 }, { "epoch": 0.8648091934084996, "grad_norm": 0.3648602366447449, "learning_rate": 7.143116204845496e-05, "loss": 0.0298, "step": 7977 }, { "epoch": 0.8649176062445794, "grad_norm": 0.2855554223060608, "learning_rate": 7.142753518061802e-05, "loss": 0.0341, "step": 7978 }, { "epoch": 0.8650260190806591, "grad_norm": 0.337104469537735, "learning_rate": 7.142390831278108e-05, "loss": 0.0522, "step": 7979 }, { "epoch": 0.8651344319167389, "grad_norm": 0.18393033742904663, "learning_rate": 7.142028144494415e-05, "loss": 0.0234, "step": 7980 }, { "epoch": 0.8652428447528188, "grad_norm": 0.37916940450668335, "learning_rate": 7.141665457710721e-05, "loss": 0.0495, "step": 7981 }, { "epoch": 0.8653512575888985, "grad_norm": 0.2899214029312134, "learning_rate": 7.141302770927028e-05, "loss": 0.0231, "step": 7982 }, { "epoch": 0.8654596704249783, "grad_norm": 0.18525448441505432, "learning_rate": 7.140940084143334e-05, "loss": 0.0144, "step": 7983 }, { "epoch": 0.8655680832610582, "grad_norm": 0.2000824362039566, "learning_rate": 7.14057739735964e-05, "loss": 0.0322, "step": 7984 }, { "epoch": 0.8656764960971379, "grad_norm": 0.45712074637413025, "learning_rate": 7.140214710575946e-05, "loss": 0.0326, "step": 7985 }, { "epoch": 0.8657849089332177, "grad_norm": 0.22400972247123718, "learning_rate": 7.139852023792253e-05, "loss": 0.018, "step": 7986 }, { "epoch": 0.8658933217692975, "grad_norm": 0.39924025535583496, "learning_rate": 7.13948933700856e-05, "loss": 0.0586, "step": 7987 }, { "epoch": 0.8660017346053773, "grad_norm": 0.6624288558959961, "learning_rate": 7.139126650224866e-05, "loss": 0.0627, "step": 7988 }, { "epoch": 0.8661101474414571, "grad_norm": 0.3063311278820038, "learning_rate": 7.138763963441173e-05, "loss": 0.0264, "step": 7989 }, { "epoch": 0.8662185602775369, "grad_norm": 0.4561119079589844, "learning_rate": 7.138401276657479e-05, "loss": 0.0366, "step": 7990 }, { "epoch": 0.8663269731136166, "grad_norm": 0.10764472931623459, "learning_rate": 7.138038589873786e-05, "loss": 0.0128, "step": 7991 }, { "epoch": 0.8664353859496965, "grad_norm": 0.19515341520309448, "learning_rate": 7.137675903090092e-05, "loss": 0.0269, "step": 7992 }, { "epoch": 0.8665437987857763, "grad_norm": 0.11173844337463379, "learning_rate": 7.137313216306398e-05, "loss": 0.0081, "step": 7993 }, { "epoch": 0.866652211621856, "grad_norm": 0.4282270073890686, "learning_rate": 7.136950529522705e-05, "loss": 0.0353, "step": 7994 }, { "epoch": 0.8667606244579358, "grad_norm": 0.5149756669998169, "learning_rate": 7.13658784273901e-05, "loss": 0.0427, "step": 7995 }, { "epoch": 0.8668690372940157, "grad_norm": 0.29993996024131775, "learning_rate": 7.136225155955318e-05, "loss": 0.0565, "step": 7996 }, { "epoch": 0.8669774501300954, "grad_norm": 0.4831428527832031, "learning_rate": 7.135862469171623e-05, "loss": 0.0766, "step": 7997 }, { "epoch": 0.8670858629661752, "grad_norm": 0.22675131261348724, "learning_rate": 7.13549978238793e-05, "loss": 0.0274, "step": 7998 }, { "epoch": 0.8671942758022549, "grad_norm": 0.1505156010389328, "learning_rate": 7.135137095604236e-05, "loss": 0.0175, "step": 7999 }, { "epoch": 0.8673026886383348, "grad_norm": 0.33219173550605774, "learning_rate": 7.134774408820543e-05, "loss": 0.0378, "step": 8000 }, { "epoch": 0.8674111014744146, "grad_norm": 0.23868001997470856, "learning_rate": 7.134411722036849e-05, "loss": 0.0429, "step": 8001 }, { "epoch": 0.8675195143104943, "grad_norm": 0.4607146680355072, "learning_rate": 7.134049035253155e-05, "loss": 0.0298, "step": 8002 }, { "epoch": 0.8676279271465741, "grad_norm": 0.1467309147119522, "learning_rate": 7.133686348469462e-05, "loss": 0.0201, "step": 8003 }, { "epoch": 0.867736339982654, "grad_norm": 0.1720624417066574, "learning_rate": 7.133323661685769e-05, "loss": 0.0134, "step": 8004 }, { "epoch": 0.8678447528187337, "grad_norm": 0.19556832313537598, "learning_rate": 7.132960974902076e-05, "loss": 0.0302, "step": 8005 }, { "epoch": 0.8679531656548135, "grad_norm": 0.5503491163253784, "learning_rate": 7.132598288118382e-05, "loss": 0.0791, "step": 8006 }, { "epoch": 0.8680615784908933, "grad_norm": 0.3329686224460602, "learning_rate": 7.132235601334687e-05, "loss": 0.0348, "step": 8007 }, { "epoch": 0.8681699913269731, "grad_norm": 0.3066084682941437, "learning_rate": 7.131872914550994e-05, "loss": 0.0382, "step": 8008 }, { "epoch": 0.8682784041630529, "grad_norm": 0.1030585914850235, "learning_rate": 7.1315102277673e-05, "loss": 0.0114, "step": 8009 }, { "epoch": 0.8683868169991327, "grad_norm": 0.4329555928707123, "learning_rate": 7.131147540983607e-05, "loss": 0.0221, "step": 8010 }, { "epoch": 0.8684952298352124, "grad_norm": 0.2830892503261566, "learning_rate": 7.130784854199913e-05, "loss": 0.0253, "step": 8011 }, { "epoch": 0.8686036426712923, "grad_norm": 1.7539242506027222, "learning_rate": 7.13042216741622e-05, "loss": 0.0626, "step": 8012 }, { "epoch": 0.8687120555073721, "grad_norm": 0.2755059003829956, "learning_rate": 7.130059480632526e-05, "loss": 0.0204, "step": 8013 }, { "epoch": 0.8688204683434518, "grad_norm": 0.2902379333972931, "learning_rate": 7.129696793848833e-05, "loss": 0.0279, "step": 8014 }, { "epoch": 0.8689288811795317, "grad_norm": 0.8118016123771667, "learning_rate": 7.129334107065139e-05, "loss": 0.0384, "step": 8015 }, { "epoch": 0.8690372940156115, "grad_norm": 0.16833408176898956, "learning_rate": 7.128971420281444e-05, "loss": 0.0205, "step": 8016 }, { "epoch": 0.8691457068516912, "grad_norm": 0.36893659830093384, "learning_rate": 7.128608733497752e-05, "loss": 0.0698, "step": 8017 }, { "epoch": 0.869254119687771, "grad_norm": 0.2841933071613312, "learning_rate": 7.128246046714057e-05, "loss": 0.0284, "step": 8018 }, { "epoch": 0.8693625325238509, "grad_norm": 0.3607076406478882, "learning_rate": 7.127883359930364e-05, "loss": 0.0504, "step": 8019 }, { "epoch": 0.8694709453599306, "grad_norm": 0.4139125645160675, "learning_rate": 7.12752067314667e-05, "loss": 0.0157, "step": 8020 }, { "epoch": 0.8695793581960104, "grad_norm": 0.46597301959991455, "learning_rate": 7.127157986362977e-05, "loss": 0.0569, "step": 8021 }, { "epoch": 0.8696877710320902, "grad_norm": 0.1584467887878418, "learning_rate": 7.126795299579284e-05, "loss": 0.0116, "step": 8022 }, { "epoch": 0.86979618386817, "grad_norm": 0.4013975262641907, "learning_rate": 7.12643261279559e-05, "loss": 0.0779, "step": 8023 }, { "epoch": 0.8699045967042498, "grad_norm": 0.3885432481765747, "learning_rate": 7.126069926011897e-05, "loss": 0.0489, "step": 8024 }, { "epoch": 0.8700130095403296, "grad_norm": 0.3450610637664795, "learning_rate": 7.125707239228203e-05, "loss": 0.0196, "step": 8025 }, { "epoch": 0.8701214223764093, "grad_norm": 0.34087374806404114, "learning_rate": 7.12534455244451e-05, "loss": 0.0254, "step": 8026 }, { "epoch": 0.8702298352124892, "grad_norm": 0.3008173704147339, "learning_rate": 7.124981865660816e-05, "loss": 0.0315, "step": 8027 }, { "epoch": 0.870338248048569, "grad_norm": 0.2982686758041382, "learning_rate": 7.124619178877123e-05, "loss": 0.0323, "step": 8028 }, { "epoch": 0.8704466608846487, "grad_norm": 0.4210357367992401, "learning_rate": 7.124256492093428e-05, "loss": 0.0539, "step": 8029 }, { "epoch": 0.8705550737207285, "grad_norm": 0.1908760368824005, "learning_rate": 7.123893805309734e-05, "loss": 0.0354, "step": 8030 }, { "epoch": 0.8706634865568084, "grad_norm": 0.5217524170875549, "learning_rate": 7.123531118526041e-05, "loss": 0.0481, "step": 8031 }, { "epoch": 0.8707718993928881, "grad_norm": 0.3728947043418884, "learning_rate": 7.123168431742347e-05, "loss": 0.0422, "step": 8032 }, { "epoch": 0.8708803122289679, "grad_norm": 0.7814446687698364, "learning_rate": 7.122805744958654e-05, "loss": 0.0707, "step": 8033 }, { "epoch": 0.8709887250650477, "grad_norm": 0.24010112881660461, "learning_rate": 7.12244305817496e-05, "loss": 0.0291, "step": 8034 }, { "epoch": 0.8710971379011275, "grad_norm": 0.48629793524742126, "learning_rate": 7.122080371391267e-05, "loss": 0.0516, "step": 8035 }, { "epoch": 0.8712055507372073, "grad_norm": 0.08524948358535767, "learning_rate": 7.121717684607573e-05, "loss": 0.0059, "step": 8036 }, { "epoch": 0.871313963573287, "grad_norm": 0.30745479464530945, "learning_rate": 7.12135499782388e-05, "loss": 0.0233, "step": 8037 }, { "epoch": 0.8714223764093668, "grad_norm": 0.23810596764087677, "learning_rate": 7.120992311040185e-05, "loss": 0.0369, "step": 8038 }, { "epoch": 0.8715307892454467, "grad_norm": 0.48098477721214294, "learning_rate": 7.120629624256493e-05, "loss": 0.0562, "step": 8039 }, { "epoch": 0.8716392020815265, "grad_norm": 0.26720893383026123, "learning_rate": 7.1202669374728e-05, "loss": 0.0186, "step": 8040 }, { "epoch": 0.8717476149176062, "grad_norm": 0.49519845843315125, "learning_rate": 7.119904250689105e-05, "loss": 0.0703, "step": 8041 }, { "epoch": 0.8718560277536861, "grad_norm": 0.22815313935279846, "learning_rate": 7.119541563905412e-05, "loss": 0.0226, "step": 8042 }, { "epoch": 0.8719644405897659, "grad_norm": 0.18026262521743774, "learning_rate": 7.119178877121718e-05, "loss": 0.0187, "step": 8043 }, { "epoch": 0.8720728534258456, "grad_norm": 0.40698370337486267, "learning_rate": 7.118816190338025e-05, "loss": 0.0598, "step": 8044 }, { "epoch": 0.8721812662619254, "grad_norm": 0.17379072308540344, "learning_rate": 7.118453503554331e-05, "loss": 0.0194, "step": 8045 }, { "epoch": 0.8722896790980053, "grad_norm": 0.6058213710784912, "learning_rate": 7.118090816770637e-05, "loss": 0.0809, "step": 8046 }, { "epoch": 0.872398091934085, "grad_norm": 0.2547031044960022, "learning_rate": 7.117728129986944e-05, "loss": 0.0365, "step": 8047 }, { "epoch": 0.8725065047701648, "grad_norm": 0.2572499215602875, "learning_rate": 7.11736544320325e-05, "loss": 0.0376, "step": 8048 }, { "epoch": 0.8726149176062445, "grad_norm": 0.2003406435251236, "learning_rate": 7.117002756419557e-05, "loss": 0.0236, "step": 8049 }, { "epoch": 0.8727233304423244, "grad_norm": 0.26622846722602844, "learning_rate": 7.116640069635862e-05, "loss": 0.0497, "step": 8050 }, { "epoch": 0.8728317432784042, "grad_norm": 0.2606870234012604, "learning_rate": 7.11627738285217e-05, "loss": 0.0214, "step": 8051 }, { "epoch": 0.872940156114484, "grad_norm": 0.2851855456829071, "learning_rate": 7.115914696068475e-05, "loss": 0.0449, "step": 8052 }, { "epoch": 0.8730485689505637, "grad_norm": 0.38806647062301636, "learning_rate": 7.115552009284781e-05, "loss": 0.0366, "step": 8053 }, { "epoch": 0.8731569817866436, "grad_norm": 0.5030326247215271, "learning_rate": 7.115189322501088e-05, "loss": 0.0465, "step": 8054 }, { "epoch": 0.8732653946227233, "grad_norm": 0.23129448294639587, "learning_rate": 7.114826635717394e-05, "loss": 0.0357, "step": 8055 }, { "epoch": 0.8733738074588031, "grad_norm": 0.1925153285264969, "learning_rate": 7.114463948933702e-05, "loss": 0.0173, "step": 8056 }, { "epoch": 0.8734822202948829, "grad_norm": 0.6803992390632629, "learning_rate": 7.114101262150008e-05, "loss": 0.0435, "step": 8057 }, { "epoch": 0.8735906331309627, "grad_norm": 0.31794092059135437, "learning_rate": 7.113738575366315e-05, "loss": 0.0434, "step": 8058 }, { "epoch": 0.8736990459670425, "grad_norm": 0.6845824718475342, "learning_rate": 7.113375888582621e-05, "loss": 0.0869, "step": 8059 }, { "epoch": 0.8738074588031223, "grad_norm": 0.30070173740386963, "learning_rate": 7.113013201798926e-05, "loss": 0.0457, "step": 8060 }, { "epoch": 0.873915871639202, "grad_norm": 0.3198436200618744, "learning_rate": 7.112650515015234e-05, "loss": 0.042, "step": 8061 }, { "epoch": 0.8740242844752819, "grad_norm": 0.14461685717105865, "learning_rate": 7.112287828231539e-05, "loss": 0.0174, "step": 8062 }, { "epoch": 0.8741326973113617, "grad_norm": 0.6115966439247131, "learning_rate": 7.111925141447846e-05, "loss": 0.0597, "step": 8063 }, { "epoch": 0.8742411101474414, "grad_norm": 0.24214549362659454, "learning_rate": 7.111562454664152e-05, "loss": 0.0378, "step": 8064 }, { "epoch": 0.8743495229835212, "grad_norm": 0.21487218141555786, "learning_rate": 7.111199767880459e-05, "loss": 0.0487, "step": 8065 }, { "epoch": 0.8744579358196011, "grad_norm": 0.3486259877681732, "learning_rate": 7.110837081096765e-05, "loss": 0.0298, "step": 8066 }, { "epoch": 0.8745663486556808, "grad_norm": 0.35556137561798096, "learning_rate": 7.110474394313072e-05, "loss": 0.0384, "step": 8067 }, { "epoch": 0.8746747614917606, "grad_norm": 0.1762847900390625, "learning_rate": 7.110111707529378e-05, "loss": 0.0228, "step": 8068 }, { "epoch": 0.8747831743278404, "grad_norm": 0.4515279233455658, "learning_rate": 7.109749020745683e-05, "loss": 0.0523, "step": 8069 }, { "epoch": 0.8748915871639202, "grad_norm": 0.3801545202732086, "learning_rate": 7.10938633396199e-05, "loss": 0.0321, "step": 8070 }, { "epoch": 0.875, "grad_norm": 0.29871073365211487, "learning_rate": 7.109023647178296e-05, "loss": 0.0301, "step": 8071 }, { "epoch": 0.8751084128360798, "grad_norm": 0.2470184862613678, "learning_rate": 7.108660960394603e-05, "loss": 0.0164, "step": 8072 }, { "epoch": 0.8752168256721596, "grad_norm": 0.37081509828567505, "learning_rate": 7.10829827361091e-05, "loss": 0.0361, "step": 8073 }, { "epoch": 0.8753252385082394, "grad_norm": 0.4294189214706421, "learning_rate": 7.107935586827216e-05, "loss": 0.0773, "step": 8074 }, { "epoch": 0.8754336513443192, "grad_norm": 0.5692263245582581, "learning_rate": 7.107572900043523e-05, "loss": 0.043, "step": 8075 }, { "epoch": 0.8755420641803989, "grad_norm": 0.4069371819496155, "learning_rate": 7.107210213259829e-05, "loss": 0.0264, "step": 8076 }, { "epoch": 0.8756504770164788, "grad_norm": 0.30675914883613586, "learning_rate": 7.106847526476136e-05, "loss": 0.0366, "step": 8077 }, { "epoch": 0.8757588898525586, "grad_norm": 0.26685988903045654, "learning_rate": 7.106484839692442e-05, "loss": 0.0262, "step": 8078 }, { "epoch": 0.8758673026886383, "grad_norm": 0.6143907904624939, "learning_rate": 7.106122152908749e-05, "loss": 0.0861, "step": 8079 }, { "epoch": 0.8759757155247181, "grad_norm": 0.4012819230556488, "learning_rate": 7.105759466125055e-05, "loss": 0.0639, "step": 8080 }, { "epoch": 0.876084128360798, "grad_norm": 0.40267112851142883, "learning_rate": 7.105396779341362e-05, "loss": 0.0587, "step": 8081 }, { "epoch": 0.8761925411968777, "grad_norm": 0.18441525101661682, "learning_rate": 7.105034092557668e-05, "loss": 0.0183, "step": 8082 }, { "epoch": 0.8763009540329575, "grad_norm": 0.20805537700653076, "learning_rate": 7.104671405773973e-05, "loss": 0.0385, "step": 8083 }, { "epoch": 0.8764093668690373, "grad_norm": 0.4743345081806183, "learning_rate": 7.10430871899028e-05, "loss": 0.0395, "step": 8084 }, { "epoch": 0.8765177797051171, "grad_norm": 0.14474739134311676, "learning_rate": 7.103946032206586e-05, "loss": 0.023, "step": 8085 }, { "epoch": 0.8766261925411969, "grad_norm": 0.632892370223999, "learning_rate": 7.103583345422893e-05, "loss": 0.1033, "step": 8086 }, { "epoch": 0.8767346053772767, "grad_norm": 0.46803101897239685, "learning_rate": 7.103220658639199e-05, "loss": 0.0251, "step": 8087 }, { "epoch": 0.8768430182133564, "grad_norm": 0.2163287252187729, "learning_rate": 7.102857971855506e-05, "loss": 0.0235, "step": 8088 }, { "epoch": 0.8769514310494363, "grad_norm": 0.2953411340713501, "learning_rate": 7.102495285071812e-05, "loss": 0.0178, "step": 8089 }, { "epoch": 0.877059843885516, "grad_norm": 0.1749672293663025, "learning_rate": 7.102132598288119e-05, "loss": 0.0331, "step": 8090 }, { "epoch": 0.8771682567215958, "grad_norm": 0.3376023769378662, "learning_rate": 7.101769911504426e-05, "loss": 0.0257, "step": 8091 }, { "epoch": 0.8772766695576756, "grad_norm": 1.136953592300415, "learning_rate": 7.101407224720732e-05, "loss": 0.0386, "step": 8092 }, { "epoch": 0.8773850823937555, "grad_norm": 0.4961899518966675, "learning_rate": 7.101044537937039e-05, "loss": 0.0669, "step": 8093 }, { "epoch": 0.8774934952298352, "grad_norm": 0.33562859892845154, "learning_rate": 7.100681851153344e-05, "loss": 0.0418, "step": 8094 }, { "epoch": 0.877601908065915, "grad_norm": 0.28775548934936523, "learning_rate": 7.100319164369652e-05, "loss": 0.0401, "step": 8095 }, { "epoch": 0.8777103209019947, "grad_norm": 0.17261210083961487, "learning_rate": 7.099956477585957e-05, "loss": 0.0157, "step": 8096 }, { "epoch": 0.8778187337380746, "grad_norm": 0.43436869978904724, "learning_rate": 7.099593790802263e-05, "loss": 0.0646, "step": 8097 }, { "epoch": 0.8779271465741544, "grad_norm": 0.6392492055892944, "learning_rate": 7.09923110401857e-05, "loss": 0.0359, "step": 8098 }, { "epoch": 0.8780355594102341, "grad_norm": 0.5735266208648682, "learning_rate": 7.098868417234876e-05, "loss": 0.0549, "step": 8099 }, { "epoch": 0.8781439722463139, "grad_norm": 0.31968551874160767, "learning_rate": 7.098505730451183e-05, "loss": 0.0213, "step": 8100 }, { "epoch": 0.8782523850823938, "grad_norm": 0.2694088816642761, "learning_rate": 7.098143043667489e-05, "loss": 0.0303, "step": 8101 }, { "epoch": 0.8783607979184735, "grad_norm": 0.17949923872947693, "learning_rate": 7.097780356883796e-05, "loss": 0.0223, "step": 8102 }, { "epoch": 0.8784692107545533, "grad_norm": 0.2600996792316437, "learning_rate": 7.097417670100101e-05, "loss": 0.0283, "step": 8103 }, { "epoch": 0.8785776235906332, "grad_norm": 0.13693217933177948, "learning_rate": 7.097054983316409e-05, "loss": 0.0106, "step": 8104 }, { "epoch": 0.878686036426713, "grad_norm": 0.36504092812538147, "learning_rate": 7.096692296532714e-05, "loss": 0.0382, "step": 8105 }, { "epoch": 0.8787944492627927, "grad_norm": 0.2563374936580658, "learning_rate": 7.09632960974902e-05, "loss": 0.0531, "step": 8106 }, { "epoch": 0.8789028620988725, "grad_norm": 0.18524356186389923, "learning_rate": 7.095966922965327e-05, "loss": 0.0174, "step": 8107 }, { "epoch": 0.8790112749349523, "grad_norm": 0.32782894372940063, "learning_rate": 7.095604236181634e-05, "loss": 0.0487, "step": 8108 }, { "epoch": 0.8791196877710321, "grad_norm": 0.6587104797363281, "learning_rate": 7.095241549397941e-05, "loss": 0.0715, "step": 8109 }, { "epoch": 0.8792281006071119, "grad_norm": 0.27466318011283875, "learning_rate": 7.094878862614247e-05, "loss": 0.0476, "step": 8110 }, { "epoch": 0.8793365134431916, "grad_norm": 0.14494261145591736, "learning_rate": 7.094516175830553e-05, "loss": 0.0116, "step": 8111 }, { "epoch": 0.8794449262792715, "grad_norm": 0.27122390270233154, "learning_rate": 7.09415348904686e-05, "loss": 0.0517, "step": 8112 }, { "epoch": 0.8795533391153513, "grad_norm": 0.4347308576107025, "learning_rate": 7.093790802263166e-05, "loss": 0.0426, "step": 8113 }, { "epoch": 0.879661751951431, "grad_norm": 0.27542540431022644, "learning_rate": 7.093428115479473e-05, "loss": 0.0424, "step": 8114 }, { "epoch": 0.8797701647875108, "grad_norm": 0.46401751041412354, "learning_rate": 7.093065428695778e-05, "loss": 0.0733, "step": 8115 }, { "epoch": 0.8798785776235907, "grad_norm": 0.36383217573165894, "learning_rate": 7.092702741912085e-05, "loss": 0.0521, "step": 8116 }, { "epoch": 0.8799869904596704, "grad_norm": 0.1844838410615921, "learning_rate": 7.092340055128391e-05, "loss": 0.0297, "step": 8117 }, { "epoch": 0.8800954032957502, "grad_norm": 0.4264051020145416, "learning_rate": 7.091977368344698e-05, "loss": 0.047, "step": 8118 }, { "epoch": 0.88020381613183, "grad_norm": 0.678458034992218, "learning_rate": 7.091614681561004e-05, "loss": 0.1065, "step": 8119 }, { "epoch": 0.8803122289679098, "grad_norm": 0.31528475880622864, "learning_rate": 7.09125199477731e-05, "loss": 0.0455, "step": 8120 }, { "epoch": 0.8804206418039896, "grad_norm": 0.509768009185791, "learning_rate": 7.090889307993617e-05, "loss": 0.0299, "step": 8121 }, { "epoch": 0.8805290546400694, "grad_norm": 0.260131299495697, "learning_rate": 7.090526621209923e-05, "loss": 0.0207, "step": 8122 }, { "epoch": 0.8806374674761491, "grad_norm": 0.2998104989528656, "learning_rate": 7.09016393442623e-05, "loss": 0.0358, "step": 8123 }, { "epoch": 0.880745880312229, "grad_norm": 0.19059684872627258, "learning_rate": 7.089801247642535e-05, "loss": 0.0178, "step": 8124 }, { "epoch": 0.8808542931483088, "grad_norm": 0.5916192531585693, "learning_rate": 7.089438560858844e-05, "loss": 0.0829, "step": 8125 }, { "epoch": 0.8809627059843885, "grad_norm": 0.094899982213974, "learning_rate": 7.08907587407515e-05, "loss": 0.0125, "step": 8126 }, { "epoch": 0.8810711188204683, "grad_norm": 0.21044453978538513, "learning_rate": 7.088713187291455e-05, "loss": 0.0223, "step": 8127 }, { "epoch": 0.8811795316565482, "grad_norm": 0.1824406534433365, "learning_rate": 7.088350500507762e-05, "loss": 0.0345, "step": 8128 }, { "epoch": 0.8812879444926279, "grad_norm": 0.14196398854255676, "learning_rate": 7.087987813724068e-05, "loss": 0.0232, "step": 8129 }, { "epoch": 0.8813963573287077, "grad_norm": 0.17930682003498077, "learning_rate": 7.087625126940375e-05, "loss": 0.0328, "step": 8130 }, { "epoch": 0.8815047701647876, "grad_norm": 0.3228236138820648, "learning_rate": 7.087262440156681e-05, "loss": 0.0522, "step": 8131 }, { "epoch": 0.8816131830008673, "grad_norm": 0.24235489964485168, "learning_rate": 7.086899753372988e-05, "loss": 0.0413, "step": 8132 }, { "epoch": 0.8817215958369471, "grad_norm": 0.5313293933868408, "learning_rate": 7.086537066589294e-05, "loss": 0.038, "step": 8133 }, { "epoch": 0.8818300086730269, "grad_norm": 0.40506690740585327, "learning_rate": 7.0861743798056e-05, "loss": 0.0417, "step": 8134 }, { "epoch": 0.8819384215091067, "grad_norm": 0.20812781155109406, "learning_rate": 7.085811693021907e-05, "loss": 0.0309, "step": 8135 }, { "epoch": 0.8820468343451865, "grad_norm": 0.2213255763053894, "learning_rate": 7.085449006238212e-05, "loss": 0.0213, "step": 8136 }, { "epoch": 0.8821552471812663, "grad_norm": 0.31179308891296387, "learning_rate": 7.08508631945452e-05, "loss": 0.0432, "step": 8137 }, { "epoch": 0.882263660017346, "grad_norm": 0.17096178233623505, "learning_rate": 7.084723632670825e-05, "loss": 0.035, "step": 8138 }, { "epoch": 0.8823720728534259, "grad_norm": 0.18027029931545258, "learning_rate": 7.084360945887132e-05, "loss": 0.0086, "step": 8139 }, { "epoch": 0.8824804856895057, "grad_norm": 0.44602054357528687, "learning_rate": 7.083998259103438e-05, "loss": 0.0692, "step": 8140 }, { "epoch": 0.8825888985255854, "grad_norm": 0.3842267096042633, "learning_rate": 7.083635572319745e-05, "loss": 0.0338, "step": 8141 }, { "epoch": 0.8826973113616652, "grad_norm": 0.5500832796096802, "learning_rate": 7.083272885536052e-05, "loss": 0.0233, "step": 8142 }, { "epoch": 0.8828057241977451, "grad_norm": 0.3664986789226532, "learning_rate": 7.082910198752358e-05, "loss": 0.0403, "step": 8143 }, { "epoch": 0.8829141370338248, "grad_norm": 0.5106587409973145, "learning_rate": 7.082547511968665e-05, "loss": 0.0537, "step": 8144 }, { "epoch": 0.8830225498699046, "grad_norm": 0.4126334488391876, "learning_rate": 7.082184825184971e-05, "loss": 0.0385, "step": 8145 }, { "epoch": 0.8831309627059843, "grad_norm": 0.07790036499500275, "learning_rate": 7.081822138401278e-05, "loss": 0.008, "step": 8146 }, { "epoch": 0.8832393755420642, "grad_norm": 0.3500314950942993, "learning_rate": 7.081459451617584e-05, "loss": 0.0392, "step": 8147 }, { "epoch": 0.883347788378144, "grad_norm": 0.20336821675300598, "learning_rate": 7.08109676483389e-05, "loss": 0.0163, "step": 8148 }, { "epoch": 0.8834562012142237, "grad_norm": 0.3055873215198517, "learning_rate": 7.080734078050196e-05, "loss": 0.0285, "step": 8149 }, { "epoch": 0.8835646140503035, "grad_norm": 0.33167147636413574, "learning_rate": 7.080371391266502e-05, "loss": 0.0412, "step": 8150 }, { "epoch": 0.8836730268863834, "grad_norm": 0.2451801747083664, "learning_rate": 7.080008704482809e-05, "loss": 0.0315, "step": 8151 }, { "epoch": 0.8837814397224631, "grad_norm": 0.6072148084640503, "learning_rate": 7.079646017699115e-05, "loss": 0.047, "step": 8152 }, { "epoch": 0.8838898525585429, "grad_norm": 0.17054085433483124, "learning_rate": 7.079283330915422e-05, "loss": 0.0263, "step": 8153 }, { "epoch": 0.8839982653946227, "grad_norm": 0.11126386374235153, "learning_rate": 7.078920644131728e-05, "loss": 0.016, "step": 8154 }, { "epoch": 0.8841066782307025, "grad_norm": 0.18241485953330994, "learning_rate": 7.078557957348035e-05, "loss": 0.0188, "step": 8155 }, { "epoch": 0.8842150910667823, "grad_norm": 0.14572837948799133, "learning_rate": 7.07819527056434e-05, "loss": 0.0134, "step": 8156 }, { "epoch": 0.8843235039028621, "grad_norm": 0.6791537404060364, "learning_rate": 7.077832583780648e-05, "loss": 0.037, "step": 8157 }, { "epoch": 0.8844319167389418, "grad_norm": 0.154613196849823, "learning_rate": 7.077469896996953e-05, "loss": 0.0121, "step": 8158 }, { "epoch": 0.8845403295750217, "grad_norm": 0.16043546795845032, "learning_rate": 7.077107210213259e-05, "loss": 0.0128, "step": 8159 }, { "epoch": 0.8846487424111015, "grad_norm": 0.4167194068431854, "learning_rate": 7.076744523429568e-05, "loss": 0.0081, "step": 8160 }, { "epoch": 0.8847571552471812, "grad_norm": 0.3382686972618103, "learning_rate": 7.076381836645873e-05, "loss": 0.0653, "step": 8161 }, { "epoch": 0.8848655680832611, "grad_norm": 0.31522759795188904, "learning_rate": 7.07601914986218e-05, "loss": 0.02, "step": 8162 }, { "epoch": 0.8849739809193409, "grad_norm": 0.4218278229236603, "learning_rate": 7.075656463078486e-05, "loss": 0.0393, "step": 8163 }, { "epoch": 0.8850823937554206, "grad_norm": 0.29412466287612915, "learning_rate": 7.075293776294792e-05, "loss": 0.0205, "step": 8164 }, { "epoch": 0.8851908065915004, "grad_norm": 0.6271398663520813, "learning_rate": 7.074931089511099e-05, "loss": 0.0994, "step": 8165 }, { "epoch": 0.8852992194275803, "grad_norm": 0.46848002076148987, "learning_rate": 7.074568402727405e-05, "loss": 0.0383, "step": 8166 }, { "epoch": 0.88540763226366, "grad_norm": 0.24570734798908234, "learning_rate": 7.074205715943712e-05, "loss": 0.0089, "step": 8167 }, { "epoch": 0.8855160450997398, "grad_norm": 0.24102471768856049, "learning_rate": 7.073843029160017e-05, "loss": 0.027, "step": 8168 }, { "epoch": 0.8856244579358196, "grad_norm": 0.3993368148803711, "learning_rate": 7.073480342376325e-05, "loss": 0.0533, "step": 8169 }, { "epoch": 0.8857328707718994, "grad_norm": 0.2520166337490082, "learning_rate": 7.07311765559263e-05, "loss": 0.0304, "step": 8170 }, { "epoch": 0.8858412836079792, "grad_norm": 0.684141218662262, "learning_rate": 7.072754968808937e-05, "loss": 0.034, "step": 8171 }, { "epoch": 0.885949696444059, "grad_norm": 0.44469234347343445, "learning_rate": 7.072392282025243e-05, "loss": 0.0428, "step": 8172 }, { "epoch": 0.8860581092801387, "grad_norm": 0.19767187535762787, "learning_rate": 7.072029595241549e-05, "loss": 0.0328, "step": 8173 }, { "epoch": 0.8861665221162186, "grad_norm": 0.5684441328048706, "learning_rate": 7.071666908457856e-05, "loss": 0.0682, "step": 8174 }, { "epoch": 0.8862749349522984, "grad_norm": 0.3249090313911438, "learning_rate": 7.071304221674162e-05, "loss": 0.0434, "step": 8175 }, { "epoch": 0.8863833477883781, "grad_norm": 0.345826119184494, "learning_rate": 7.070941534890469e-05, "loss": 0.0108, "step": 8176 }, { "epoch": 0.8864917606244579, "grad_norm": 0.6738152503967285, "learning_rate": 7.070578848106776e-05, "loss": 0.0486, "step": 8177 }, { "epoch": 0.8866001734605378, "grad_norm": 0.3554690182209015, "learning_rate": 7.070216161323082e-05, "loss": 0.0435, "step": 8178 }, { "epoch": 0.8867085862966175, "grad_norm": 0.43763354420661926, "learning_rate": 7.069853474539389e-05, "loss": 0.0407, "step": 8179 }, { "epoch": 0.8868169991326973, "grad_norm": 0.4322910010814667, "learning_rate": 7.069490787755694e-05, "loss": 0.0282, "step": 8180 }, { "epoch": 0.886925411968777, "grad_norm": 0.2719101309776306, "learning_rate": 7.069128100972001e-05, "loss": 0.0279, "step": 8181 }, { "epoch": 0.8870338248048569, "grad_norm": 0.3026770353317261, "learning_rate": 7.068765414188307e-05, "loss": 0.0348, "step": 8182 }, { "epoch": 0.8871422376409367, "grad_norm": 0.5666779279708862, "learning_rate": 7.068402727404614e-05, "loss": 0.0551, "step": 8183 }, { "epoch": 0.8872506504770165, "grad_norm": 0.2301020324230194, "learning_rate": 7.06804004062092e-05, "loss": 0.0234, "step": 8184 }, { "epoch": 0.8873590633130962, "grad_norm": 0.5429239869117737, "learning_rate": 7.067677353837227e-05, "loss": 0.0469, "step": 8185 }, { "epoch": 0.8874674761491761, "grad_norm": 0.38419508934020996, "learning_rate": 7.067314667053533e-05, "loss": 0.0476, "step": 8186 }, { "epoch": 0.8875758889852559, "grad_norm": 0.3333520293235779, "learning_rate": 7.066951980269839e-05, "loss": 0.0252, "step": 8187 }, { "epoch": 0.8876843018213356, "grad_norm": 0.3884081542491913, "learning_rate": 7.066589293486146e-05, "loss": 0.046, "step": 8188 }, { "epoch": 0.8877927146574154, "grad_norm": 0.3107840120792389, "learning_rate": 7.066226606702451e-05, "loss": 0.0311, "step": 8189 }, { "epoch": 0.8879011274934953, "grad_norm": 0.3402217626571655, "learning_rate": 7.065863919918758e-05, "loss": 0.0457, "step": 8190 }, { "epoch": 0.888009540329575, "grad_norm": 0.23107226192951202, "learning_rate": 7.065501233135064e-05, "loss": 0.0296, "step": 8191 }, { "epoch": 0.8881179531656548, "grad_norm": 0.46799129247665405, "learning_rate": 7.065138546351371e-05, "loss": 0.0725, "step": 8192 }, { "epoch": 0.8882263660017347, "grad_norm": 0.06924562901258469, "learning_rate": 7.064775859567677e-05, "loss": 0.0074, "step": 8193 }, { "epoch": 0.8883347788378144, "grad_norm": 0.17283648252487183, "learning_rate": 7.064413172783984e-05, "loss": 0.0395, "step": 8194 }, { "epoch": 0.8884431916738942, "grad_norm": 0.44156086444854736, "learning_rate": 7.064050486000291e-05, "loss": 0.0551, "step": 8195 }, { "epoch": 0.888551604509974, "grad_norm": 0.15336260199546814, "learning_rate": 7.063687799216597e-05, "loss": 0.0152, "step": 8196 }, { "epoch": 0.8886600173460538, "grad_norm": 0.14703109860420227, "learning_rate": 7.063325112432904e-05, "loss": 0.0153, "step": 8197 }, { "epoch": 0.8887684301821336, "grad_norm": 0.4564245343208313, "learning_rate": 7.06296242564921e-05, "loss": 0.0473, "step": 8198 }, { "epoch": 0.8888768430182133, "grad_norm": 0.5835501551628113, "learning_rate": 7.062599738865517e-05, "loss": 0.0483, "step": 8199 }, { "epoch": 0.8889852558542931, "grad_norm": 0.5945451855659485, "learning_rate": 7.062237052081823e-05, "loss": 0.0899, "step": 8200 }, { "epoch": 0.889093668690373, "grad_norm": 0.2820247709751129, "learning_rate": 7.061874365298128e-05, "loss": 0.0309, "step": 8201 }, { "epoch": 0.8892020815264527, "grad_norm": 0.4314570426940918, "learning_rate": 7.061511678514435e-05, "loss": 0.074, "step": 8202 }, { "epoch": 0.8893104943625325, "grad_norm": 0.28447309136390686, "learning_rate": 7.061148991730741e-05, "loss": 0.0568, "step": 8203 }, { "epoch": 0.8894189071986123, "grad_norm": 0.16771911084651947, "learning_rate": 7.060786304947048e-05, "loss": 0.0196, "step": 8204 }, { "epoch": 0.8895273200346921, "grad_norm": 0.20545056462287903, "learning_rate": 7.060423618163354e-05, "loss": 0.0236, "step": 8205 }, { "epoch": 0.8896357328707719, "grad_norm": 0.3236832022666931, "learning_rate": 7.060060931379661e-05, "loss": 0.0346, "step": 8206 }, { "epoch": 0.8897441457068517, "grad_norm": 0.45304492115974426, "learning_rate": 7.059698244595967e-05, "loss": 0.0386, "step": 8207 }, { "epoch": 0.8898525585429314, "grad_norm": 0.29694893956184387, "learning_rate": 7.059335557812274e-05, "loss": 0.0361, "step": 8208 }, { "epoch": 0.8899609713790113, "grad_norm": 0.20633108913898468, "learning_rate": 7.05897287102858e-05, "loss": 0.0165, "step": 8209 }, { "epoch": 0.8900693842150911, "grad_norm": 0.22285667061805725, "learning_rate": 7.058610184244885e-05, "loss": 0.0219, "step": 8210 }, { "epoch": 0.8901777970511708, "grad_norm": 0.19446222484111786, "learning_rate": 7.058247497461194e-05, "loss": 0.0172, "step": 8211 }, { "epoch": 0.8902862098872506, "grad_norm": 0.6214487552642822, "learning_rate": 7.0578848106775e-05, "loss": 0.0797, "step": 8212 }, { "epoch": 0.8903946227233305, "grad_norm": 0.12392042577266693, "learning_rate": 7.057522123893807e-05, "loss": 0.0119, "step": 8213 }, { "epoch": 0.8905030355594102, "grad_norm": 0.27664420008659363, "learning_rate": 7.057159437110112e-05, "loss": 0.0469, "step": 8214 }, { "epoch": 0.89061144839549, "grad_norm": 0.2350059300661087, "learning_rate": 7.056796750326418e-05, "loss": 0.0319, "step": 8215 }, { "epoch": 0.8907198612315698, "grad_norm": 0.38998663425445557, "learning_rate": 7.056434063542725e-05, "loss": 0.0564, "step": 8216 }, { "epoch": 0.8908282740676496, "grad_norm": 0.66097092628479, "learning_rate": 7.056071376759031e-05, "loss": 0.0581, "step": 8217 }, { "epoch": 0.8909366869037294, "grad_norm": 0.38644662499427795, "learning_rate": 7.055708689975338e-05, "loss": 0.0556, "step": 8218 }, { "epoch": 0.8910450997398092, "grad_norm": 0.23037053644657135, "learning_rate": 7.055346003191644e-05, "loss": 0.0192, "step": 8219 }, { "epoch": 0.891153512575889, "grad_norm": 0.27667802572250366, "learning_rate": 7.054983316407951e-05, "loss": 0.0202, "step": 8220 }, { "epoch": 0.8912619254119688, "grad_norm": 0.5131875872612, "learning_rate": 7.054620629624257e-05, "loss": 0.0872, "step": 8221 }, { "epoch": 0.8913703382480486, "grad_norm": 0.49040675163269043, "learning_rate": 7.054257942840564e-05, "loss": 0.1233, "step": 8222 }, { "epoch": 0.8914787510841283, "grad_norm": 0.4718749523162842, "learning_rate": 7.05389525605687e-05, "loss": 0.0925, "step": 8223 }, { "epoch": 0.8915871639202082, "grad_norm": 0.4951261579990387, "learning_rate": 7.053532569273175e-05, "loss": 0.0466, "step": 8224 }, { "epoch": 0.891695576756288, "grad_norm": 0.23740556836128235, "learning_rate": 7.053169882489482e-05, "loss": 0.0306, "step": 8225 }, { "epoch": 0.8918039895923677, "grad_norm": 0.22809632122516632, "learning_rate": 7.052807195705788e-05, "loss": 0.0396, "step": 8226 }, { "epoch": 0.8919124024284475, "grad_norm": 0.2963050901889801, "learning_rate": 7.052444508922095e-05, "loss": 0.0266, "step": 8227 }, { "epoch": 0.8920208152645274, "grad_norm": 0.6207825541496277, "learning_rate": 7.052081822138401e-05, "loss": 0.0485, "step": 8228 }, { "epoch": 0.8921292281006071, "grad_norm": 0.26379457116127014, "learning_rate": 7.051719135354709e-05, "loss": 0.0554, "step": 8229 }, { "epoch": 0.8922376409366869, "grad_norm": 0.15220670402050018, "learning_rate": 7.051356448571015e-05, "loss": 0.0132, "step": 8230 }, { "epoch": 0.8923460537727667, "grad_norm": 0.33661630749702454, "learning_rate": 7.05099376178732e-05, "loss": 0.0284, "step": 8231 }, { "epoch": 0.8924544666088465, "grad_norm": 0.3452916145324707, "learning_rate": 7.050631075003628e-05, "loss": 0.0658, "step": 8232 }, { "epoch": 0.8925628794449263, "grad_norm": 0.444314569234848, "learning_rate": 7.050268388219933e-05, "loss": 0.0722, "step": 8233 }, { "epoch": 0.8926712922810061, "grad_norm": 0.2705520987510681, "learning_rate": 7.04990570143624e-05, "loss": 0.0333, "step": 8234 }, { "epoch": 0.8927797051170858, "grad_norm": 0.2028864175081253, "learning_rate": 7.049543014652546e-05, "loss": 0.0219, "step": 8235 }, { "epoch": 0.8928881179531657, "grad_norm": 0.3169020712375641, "learning_rate": 7.049180327868853e-05, "loss": 0.0284, "step": 8236 }, { "epoch": 0.8929965307892455, "grad_norm": 0.30865025520324707, "learning_rate": 7.048817641085159e-05, "loss": 0.0703, "step": 8237 }, { "epoch": 0.8931049436253252, "grad_norm": 0.5179864168167114, "learning_rate": 7.048454954301466e-05, "loss": 0.0635, "step": 8238 }, { "epoch": 0.893213356461405, "grad_norm": 0.5606207251548767, "learning_rate": 7.048092267517772e-05, "loss": 0.0622, "step": 8239 }, { "epoch": 0.8933217692974849, "grad_norm": 0.22008846700191498, "learning_rate": 7.047729580734078e-05, "loss": 0.0493, "step": 8240 }, { "epoch": 0.8934301821335646, "grad_norm": 0.3121587336063385, "learning_rate": 7.047366893950385e-05, "loss": 0.0499, "step": 8241 }, { "epoch": 0.8935385949696444, "grad_norm": 0.33632734417915344, "learning_rate": 7.04700420716669e-05, "loss": 0.0552, "step": 8242 }, { "epoch": 0.8936470078057241, "grad_norm": 0.29376330971717834, "learning_rate": 7.046641520382998e-05, "loss": 0.0275, "step": 8243 }, { "epoch": 0.893755420641804, "grad_norm": 0.33343246579170227, "learning_rate": 7.046278833599303e-05, "loss": 0.0859, "step": 8244 }, { "epoch": 0.8938638334778838, "grad_norm": 0.3419765532016754, "learning_rate": 7.04591614681561e-05, "loss": 0.043, "step": 8245 }, { "epoch": 0.8939722463139635, "grad_norm": 0.20148174464702606, "learning_rate": 7.045553460031917e-05, "loss": 0.0255, "step": 8246 }, { "epoch": 0.8940806591500433, "grad_norm": 0.17394332587718964, "learning_rate": 7.045190773248223e-05, "loss": 0.0451, "step": 8247 }, { "epoch": 0.8941890719861232, "grad_norm": 0.2394695281982422, "learning_rate": 7.04482808646453e-05, "loss": 0.0329, "step": 8248 }, { "epoch": 0.894297484822203, "grad_norm": 0.21213379502296448, "learning_rate": 7.044465399680836e-05, "loss": 0.0562, "step": 8249 }, { "epoch": 0.8944058976582827, "grad_norm": 0.19689150154590607, "learning_rate": 7.044102712897143e-05, "loss": 0.0326, "step": 8250 }, { "epoch": 0.8945143104943626, "grad_norm": 0.28431668877601624, "learning_rate": 7.043740026113449e-05, "loss": 0.0373, "step": 8251 }, { "epoch": 0.8946227233304423, "grad_norm": 0.3011299669742584, "learning_rate": 7.043377339329756e-05, "loss": 0.0455, "step": 8252 }, { "epoch": 0.8947311361665221, "grad_norm": 0.24834956228733063, "learning_rate": 7.043014652546062e-05, "loss": 0.0444, "step": 8253 }, { "epoch": 0.8948395490026019, "grad_norm": 0.21278467774391174, "learning_rate": 7.042651965762367e-05, "loss": 0.0244, "step": 8254 }, { "epoch": 0.8949479618386817, "grad_norm": 0.20883694291114807, "learning_rate": 7.042289278978674e-05, "loss": 0.0268, "step": 8255 }, { "epoch": 0.8950563746747615, "grad_norm": 0.21748237311840057, "learning_rate": 7.04192659219498e-05, "loss": 0.0283, "step": 8256 }, { "epoch": 0.8951647875108413, "grad_norm": 0.21991440653800964, "learning_rate": 7.041563905411287e-05, "loss": 0.0279, "step": 8257 }, { "epoch": 0.895273200346921, "grad_norm": 0.23750345408916473, "learning_rate": 7.041201218627593e-05, "loss": 0.0252, "step": 8258 }, { "epoch": 0.8953816131830009, "grad_norm": 0.29836800694465637, "learning_rate": 7.0408385318439e-05, "loss": 0.0348, "step": 8259 }, { "epoch": 0.8954900260190807, "grad_norm": 0.37698161602020264, "learning_rate": 7.040475845060206e-05, "loss": 0.0341, "step": 8260 }, { "epoch": 0.8955984388551604, "grad_norm": 0.6376975774765015, "learning_rate": 7.040113158276513e-05, "loss": 0.0769, "step": 8261 }, { "epoch": 0.8957068516912402, "grad_norm": 0.5348132848739624, "learning_rate": 7.039750471492819e-05, "loss": 0.0338, "step": 8262 }, { "epoch": 0.8958152645273201, "grad_norm": 0.45501139760017395, "learning_rate": 7.039387784709126e-05, "loss": 0.0472, "step": 8263 }, { "epoch": 0.8959236773633998, "grad_norm": 0.30665749311447144, "learning_rate": 7.039025097925433e-05, "loss": 0.0198, "step": 8264 }, { "epoch": 0.8960320901994796, "grad_norm": 0.16774147748947144, "learning_rate": 7.038662411141739e-05, "loss": 0.0156, "step": 8265 }, { "epoch": 0.8961405030355594, "grad_norm": 0.7512730956077576, "learning_rate": 7.038299724358046e-05, "loss": 0.0995, "step": 8266 }, { "epoch": 0.8962489158716392, "grad_norm": 0.4812290370464325, "learning_rate": 7.037937037574351e-05, "loss": 0.0807, "step": 8267 }, { "epoch": 0.896357328707719, "grad_norm": 0.4502752423286438, "learning_rate": 7.037574350790657e-05, "loss": 0.0642, "step": 8268 }, { "epoch": 0.8964657415437988, "grad_norm": 0.3846445679664612, "learning_rate": 7.037211664006964e-05, "loss": 0.0318, "step": 8269 }, { "epoch": 0.8965741543798785, "grad_norm": 0.8883446455001831, "learning_rate": 7.03684897722327e-05, "loss": 0.05, "step": 8270 }, { "epoch": 0.8966825672159584, "grad_norm": 0.1151951476931572, "learning_rate": 7.036486290439577e-05, "loss": 0.021, "step": 8271 }, { "epoch": 0.8967909800520382, "grad_norm": 0.20419900119304657, "learning_rate": 7.036123603655883e-05, "loss": 0.0265, "step": 8272 }, { "epoch": 0.8968993928881179, "grad_norm": 0.20882746577262878, "learning_rate": 7.03576091687219e-05, "loss": 0.0285, "step": 8273 }, { "epoch": 0.8970078057241977, "grad_norm": 0.470882385969162, "learning_rate": 7.035398230088496e-05, "loss": 0.054, "step": 8274 }, { "epoch": 0.8971162185602776, "grad_norm": 0.40031659603118896, "learning_rate": 7.035035543304803e-05, "loss": 0.0589, "step": 8275 }, { "epoch": 0.8972246313963573, "grad_norm": 0.6865810751914978, "learning_rate": 7.034672856521108e-05, "loss": 0.0812, "step": 8276 }, { "epoch": 0.8973330442324371, "grad_norm": 0.5806080102920532, "learning_rate": 7.034310169737414e-05, "loss": 0.043, "step": 8277 }, { "epoch": 0.897441457068517, "grad_norm": 0.3190026879310608, "learning_rate": 7.033947482953721e-05, "loss": 0.0585, "step": 8278 }, { "epoch": 0.8975498699045967, "grad_norm": 0.43860211968421936, "learning_rate": 7.033584796170027e-05, "loss": 0.0366, "step": 8279 }, { "epoch": 0.8976582827406765, "grad_norm": 0.2243112027645111, "learning_rate": 7.033222109386334e-05, "loss": 0.0111, "step": 8280 }, { "epoch": 0.8977666955767563, "grad_norm": 0.5386127829551697, "learning_rate": 7.032859422602641e-05, "loss": 0.1244, "step": 8281 }, { "epoch": 0.8978751084128361, "grad_norm": 0.08026695996522903, "learning_rate": 7.032496735818947e-05, "loss": 0.0055, "step": 8282 }, { "epoch": 0.8979835212489159, "grad_norm": 0.3256092667579651, "learning_rate": 7.032134049035254e-05, "loss": 0.0375, "step": 8283 }, { "epoch": 0.8980919340849957, "grad_norm": 0.4564916789531708, "learning_rate": 7.03177136225156e-05, "loss": 0.0374, "step": 8284 }, { "epoch": 0.8982003469210754, "grad_norm": 0.11612968891859055, "learning_rate": 7.031408675467867e-05, "loss": 0.0181, "step": 8285 }, { "epoch": 0.8983087597571553, "grad_norm": 0.13947726786136627, "learning_rate": 7.031045988684173e-05, "loss": 0.0173, "step": 8286 }, { "epoch": 0.8984171725932351, "grad_norm": 0.20994748175144196, "learning_rate": 7.03068330190048e-05, "loss": 0.024, "step": 8287 }, { "epoch": 0.8985255854293148, "grad_norm": 0.39349961280822754, "learning_rate": 7.030320615116785e-05, "loss": 0.0507, "step": 8288 }, { "epoch": 0.8986339982653946, "grad_norm": 0.3624572157859802, "learning_rate": 7.029957928333092e-05, "loss": 0.0417, "step": 8289 }, { "epoch": 0.8987424111014745, "grad_norm": 0.21162912249565125, "learning_rate": 7.029595241549398e-05, "loss": 0.0241, "step": 8290 }, { "epoch": 0.8988508239375542, "grad_norm": 0.30048054456710815, "learning_rate": 7.029232554765704e-05, "loss": 0.0258, "step": 8291 }, { "epoch": 0.898959236773634, "grad_norm": 0.40245017409324646, "learning_rate": 7.028869867982011e-05, "loss": 0.0686, "step": 8292 }, { "epoch": 0.8990676496097137, "grad_norm": 0.43842440843582153, "learning_rate": 7.028507181198317e-05, "loss": 0.0407, "step": 8293 }, { "epoch": 0.8991760624457936, "grad_norm": 0.30505844950675964, "learning_rate": 7.028144494414624e-05, "loss": 0.0368, "step": 8294 }, { "epoch": 0.8992844752818734, "grad_norm": 0.40943944454193115, "learning_rate": 7.02778180763093e-05, "loss": 0.0291, "step": 8295 }, { "epoch": 0.8993928881179531, "grad_norm": 0.41811662912368774, "learning_rate": 7.027419120847237e-05, "loss": 0.0406, "step": 8296 }, { "epoch": 0.8995013009540329, "grad_norm": 0.13854622840881348, "learning_rate": 7.027056434063542e-05, "loss": 0.0134, "step": 8297 }, { "epoch": 0.8996097137901128, "grad_norm": 0.23002628982067108, "learning_rate": 7.02669374727985e-05, "loss": 0.0387, "step": 8298 }, { "epoch": 0.8997181266261925, "grad_norm": 0.3356549143791199, "learning_rate": 7.026331060496157e-05, "loss": 0.052, "step": 8299 }, { "epoch": 0.8998265394622723, "grad_norm": 0.1895870417356491, "learning_rate": 7.025968373712462e-05, "loss": 0.0281, "step": 8300 }, { "epoch": 0.8999349522983521, "grad_norm": 0.17207922041416168, "learning_rate": 7.02560568692877e-05, "loss": 0.0237, "step": 8301 }, { "epoch": 0.900043365134432, "grad_norm": 0.17603568732738495, "learning_rate": 7.025243000145075e-05, "loss": 0.0248, "step": 8302 }, { "epoch": 0.9001517779705117, "grad_norm": 0.4848693907260895, "learning_rate": 7.024880313361382e-05, "loss": 0.0573, "step": 8303 }, { "epoch": 0.9002601908065915, "grad_norm": 0.2713419795036316, "learning_rate": 7.024517626577688e-05, "loss": 0.0208, "step": 8304 }, { "epoch": 0.9003686036426712, "grad_norm": 0.12172838300466537, "learning_rate": 7.024154939793994e-05, "loss": 0.0125, "step": 8305 }, { "epoch": 0.9004770164787511, "grad_norm": 0.14699982106685638, "learning_rate": 7.023792253010301e-05, "loss": 0.0241, "step": 8306 }, { "epoch": 0.9005854293148309, "grad_norm": 0.21671274304389954, "learning_rate": 7.023429566226606e-05, "loss": 0.0202, "step": 8307 }, { "epoch": 0.9006938421509106, "grad_norm": 0.5174499154090881, "learning_rate": 7.023066879442914e-05, "loss": 0.0558, "step": 8308 }, { "epoch": 0.9008022549869905, "grad_norm": 0.21171972155570984, "learning_rate": 7.022704192659219e-05, "loss": 0.0222, "step": 8309 }, { "epoch": 0.9009106678230703, "grad_norm": 0.3459501266479492, "learning_rate": 7.022341505875526e-05, "loss": 0.0517, "step": 8310 }, { "epoch": 0.90101908065915, "grad_norm": 0.2349177896976471, "learning_rate": 7.021978819091832e-05, "loss": 0.0255, "step": 8311 }, { "epoch": 0.9011274934952298, "grad_norm": 0.5124483108520508, "learning_rate": 7.021616132308139e-05, "loss": 0.0728, "step": 8312 }, { "epoch": 0.9012359063313097, "grad_norm": 0.42482447624206543, "learning_rate": 7.021253445524445e-05, "loss": 0.0349, "step": 8313 }, { "epoch": 0.9013443191673894, "grad_norm": 0.1775461882352829, "learning_rate": 7.020890758740751e-05, "loss": 0.0152, "step": 8314 }, { "epoch": 0.9014527320034692, "grad_norm": 0.23121285438537598, "learning_rate": 7.020528071957059e-05, "loss": 0.0133, "step": 8315 }, { "epoch": 0.901561144839549, "grad_norm": 0.24476328492164612, "learning_rate": 7.020165385173365e-05, "loss": 0.0395, "step": 8316 }, { "epoch": 0.9016695576756288, "grad_norm": 0.5687626004219055, "learning_rate": 7.019802698389672e-05, "loss": 0.0536, "step": 8317 }, { "epoch": 0.9017779705117086, "grad_norm": 0.355913907289505, "learning_rate": 7.019440011605978e-05, "loss": 0.0302, "step": 8318 }, { "epoch": 0.9018863833477884, "grad_norm": 0.24056828022003174, "learning_rate": 7.019077324822283e-05, "loss": 0.0275, "step": 8319 }, { "epoch": 0.9019947961838681, "grad_norm": 0.2729325592517853, "learning_rate": 7.01871463803859e-05, "loss": 0.0363, "step": 8320 }, { "epoch": 0.902103209019948, "grad_norm": 0.3642962872982025, "learning_rate": 7.018351951254896e-05, "loss": 0.0401, "step": 8321 }, { "epoch": 0.9022116218560278, "grad_norm": 0.6286858916282654, "learning_rate": 7.017989264471203e-05, "loss": 0.057, "step": 8322 }, { "epoch": 0.9023200346921075, "grad_norm": 0.2268257439136505, "learning_rate": 7.017626577687509e-05, "loss": 0.035, "step": 8323 }, { "epoch": 0.9024284475281873, "grad_norm": 0.15897871553897858, "learning_rate": 7.017263890903816e-05, "loss": 0.0184, "step": 8324 }, { "epoch": 0.9025368603642672, "grad_norm": 0.32843121886253357, "learning_rate": 7.016901204120122e-05, "loss": 0.0207, "step": 8325 }, { "epoch": 0.9026452732003469, "grad_norm": 0.41384363174438477, "learning_rate": 7.016538517336429e-05, "loss": 0.0302, "step": 8326 }, { "epoch": 0.9027536860364267, "grad_norm": 0.17230278253555298, "learning_rate": 7.016175830552735e-05, "loss": 0.0143, "step": 8327 }, { "epoch": 0.9028620988725065, "grad_norm": 0.45949581265449524, "learning_rate": 7.01581314376904e-05, "loss": 0.0213, "step": 8328 }, { "epoch": 0.9029705117085863, "grad_norm": 0.6311580538749695, "learning_rate": 7.015450456985348e-05, "loss": 0.1101, "step": 8329 }, { "epoch": 0.9030789245446661, "grad_norm": 0.16794809699058533, "learning_rate": 7.015087770201653e-05, "loss": 0.0249, "step": 8330 }, { "epoch": 0.9031873373807459, "grad_norm": 0.4560316205024719, "learning_rate": 7.01472508341796e-05, "loss": 0.0381, "step": 8331 }, { "epoch": 0.9032957502168256, "grad_norm": 0.28976646065711975, "learning_rate": 7.014362396634267e-05, "loss": 0.0518, "step": 8332 }, { "epoch": 0.9034041630529055, "grad_norm": 0.2874715030193329, "learning_rate": 7.013999709850575e-05, "loss": 0.0219, "step": 8333 }, { "epoch": 0.9035125758889853, "grad_norm": 0.8673073053359985, "learning_rate": 7.01363702306688e-05, "loss": 0.0574, "step": 8334 }, { "epoch": 0.903620988725065, "grad_norm": 0.21267716586589813, "learning_rate": 7.013274336283186e-05, "loss": 0.0269, "step": 8335 }, { "epoch": 0.9037294015611448, "grad_norm": 0.2617461085319519, "learning_rate": 7.012911649499493e-05, "loss": 0.0361, "step": 8336 }, { "epoch": 0.9038378143972247, "grad_norm": 0.39897575974464417, "learning_rate": 7.012548962715799e-05, "loss": 0.054, "step": 8337 }, { "epoch": 0.9039462272333044, "grad_norm": 0.5538889765739441, "learning_rate": 7.012186275932106e-05, "loss": 0.0495, "step": 8338 }, { "epoch": 0.9040546400693842, "grad_norm": 0.1778656244277954, "learning_rate": 7.011823589148412e-05, "loss": 0.0127, "step": 8339 }, { "epoch": 0.9041630529054641, "grad_norm": 0.5260095000267029, "learning_rate": 7.011460902364719e-05, "loss": 0.03, "step": 8340 }, { "epoch": 0.9042714657415438, "grad_norm": 0.1711447685956955, "learning_rate": 7.011098215581024e-05, "loss": 0.0173, "step": 8341 }, { "epoch": 0.9043798785776236, "grad_norm": 0.26656636595726013, "learning_rate": 7.010735528797332e-05, "loss": 0.0275, "step": 8342 }, { "epoch": 0.9044882914137033, "grad_norm": 0.955034613609314, "learning_rate": 7.010372842013637e-05, "loss": 0.0524, "step": 8343 }, { "epoch": 0.9045967042497832, "grad_norm": 0.2098240852355957, "learning_rate": 7.010010155229943e-05, "loss": 0.0227, "step": 8344 }, { "epoch": 0.904705117085863, "grad_norm": 0.9466928839683533, "learning_rate": 7.00964746844625e-05, "loss": 0.057, "step": 8345 }, { "epoch": 0.9048135299219427, "grad_norm": 0.10983464121818542, "learning_rate": 7.009284781662556e-05, "loss": 0.0106, "step": 8346 }, { "epoch": 0.9049219427580225, "grad_norm": 0.18485009670257568, "learning_rate": 7.008922094878863e-05, "loss": 0.0362, "step": 8347 }, { "epoch": 0.9050303555941024, "grad_norm": 0.28843578696250916, "learning_rate": 7.008559408095169e-05, "loss": 0.0508, "step": 8348 }, { "epoch": 0.9051387684301822, "grad_norm": 0.20833079516887665, "learning_rate": 7.008196721311476e-05, "loss": 0.0231, "step": 8349 }, { "epoch": 0.9052471812662619, "grad_norm": 0.25585493445396423, "learning_rate": 7.007834034527783e-05, "loss": 0.0299, "step": 8350 }, { "epoch": 0.9053555941023417, "grad_norm": 0.2623235583305359, "learning_rate": 7.007471347744089e-05, "loss": 0.0203, "step": 8351 }, { "epoch": 0.9054640069384216, "grad_norm": 0.2980567216873169, "learning_rate": 7.007108660960396e-05, "loss": 0.0285, "step": 8352 }, { "epoch": 0.9055724197745013, "grad_norm": 0.2792320251464844, "learning_rate": 7.006745974176701e-05, "loss": 0.0333, "step": 8353 }, { "epoch": 0.9056808326105811, "grad_norm": 0.054595693945884705, "learning_rate": 7.006383287393008e-05, "loss": 0.0049, "step": 8354 }, { "epoch": 0.9057892454466608, "grad_norm": 0.5972162485122681, "learning_rate": 7.006020600609314e-05, "loss": 0.0431, "step": 8355 }, { "epoch": 0.9058976582827407, "grad_norm": 0.7157241106033325, "learning_rate": 7.005657913825621e-05, "loss": 0.0761, "step": 8356 }, { "epoch": 0.9060060711188205, "grad_norm": 0.21860191226005554, "learning_rate": 7.005295227041927e-05, "loss": 0.0434, "step": 8357 }, { "epoch": 0.9061144839549002, "grad_norm": 0.4220033288002014, "learning_rate": 7.004932540258233e-05, "loss": 0.0425, "step": 8358 }, { "epoch": 0.90622289679098, "grad_norm": 0.29751014709472656, "learning_rate": 7.00456985347454e-05, "loss": 0.0177, "step": 8359 }, { "epoch": 0.9063313096270599, "grad_norm": 0.25495725870132446, "learning_rate": 7.004207166690846e-05, "loss": 0.0321, "step": 8360 }, { "epoch": 0.9064397224631396, "grad_norm": 0.46279531717300415, "learning_rate": 7.003844479907153e-05, "loss": 0.0313, "step": 8361 }, { "epoch": 0.9065481352992194, "grad_norm": 0.19487012922763824, "learning_rate": 7.003481793123458e-05, "loss": 0.0402, "step": 8362 }, { "epoch": 0.9066565481352992, "grad_norm": 0.25641772150993347, "learning_rate": 7.003119106339765e-05, "loss": 0.0282, "step": 8363 }, { "epoch": 0.906764960971379, "grad_norm": 0.21128974854946136, "learning_rate": 7.002756419556071e-05, "loss": 0.0172, "step": 8364 }, { "epoch": 0.9068733738074588, "grad_norm": 0.7915648221969604, "learning_rate": 7.002393732772378e-05, "loss": 0.0473, "step": 8365 }, { "epoch": 0.9069817866435386, "grad_norm": 0.48594430088996887, "learning_rate": 7.002031045988684e-05, "loss": 0.0498, "step": 8366 }, { "epoch": 0.9070901994796184, "grad_norm": 0.5653549432754517, "learning_rate": 7.001668359204991e-05, "loss": 0.0454, "step": 8367 }, { "epoch": 0.9071986123156982, "grad_norm": 0.4983610510826111, "learning_rate": 7.001305672421298e-05, "loss": 0.0562, "step": 8368 }, { "epoch": 0.907307025151778, "grad_norm": 0.4260022044181824, "learning_rate": 7.000942985637604e-05, "loss": 0.0493, "step": 8369 }, { "epoch": 0.9074154379878577, "grad_norm": 0.22600464522838593, "learning_rate": 7.000580298853911e-05, "loss": 0.0465, "step": 8370 }, { "epoch": 0.9075238508239376, "grad_norm": 0.24359185993671417, "learning_rate": 7.000217612070217e-05, "loss": 0.0374, "step": 8371 }, { "epoch": 0.9076322636600174, "grad_norm": 0.18999354541301727, "learning_rate": 6.999854925286522e-05, "loss": 0.0284, "step": 8372 }, { "epoch": 0.9077406764960971, "grad_norm": 0.376785010099411, "learning_rate": 6.99949223850283e-05, "loss": 0.0774, "step": 8373 }, { "epoch": 0.9078490893321769, "grad_norm": 0.24036946892738342, "learning_rate": 6.999129551719135e-05, "loss": 0.0164, "step": 8374 }, { "epoch": 0.9079575021682568, "grad_norm": 0.26629215478897095, "learning_rate": 6.998766864935442e-05, "loss": 0.0362, "step": 8375 }, { "epoch": 0.9080659150043365, "grad_norm": 0.45905718207359314, "learning_rate": 6.998404178151748e-05, "loss": 0.034, "step": 8376 }, { "epoch": 0.9081743278404163, "grad_norm": 0.3081991672515869, "learning_rate": 6.998041491368055e-05, "loss": 0.027, "step": 8377 }, { "epoch": 0.9082827406764961, "grad_norm": 0.21411588788032532, "learning_rate": 6.997678804584361e-05, "loss": 0.0259, "step": 8378 }, { "epoch": 0.9083911535125759, "grad_norm": 0.11020330339670181, "learning_rate": 6.997316117800668e-05, "loss": 0.0105, "step": 8379 }, { "epoch": 0.9084995663486557, "grad_norm": 0.44677841663360596, "learning_rate": 6.996953431016974e-05, "loss": 0.0691, "step": 8380 }, { "epoch": 0.9086079791847355, "grad_norm": 0.14835160970687866, "learning_rate": 6.99659074423328e-05, "loss": 0.0108, "step": 8381 }, { "epoch": 0.9087163920208152, "grad_norm": 0.5187487006187439, "learning_rate": 6.996228057449587e-05, "loss": 0.0761, "step": 8382 }, { "epoch": 0.9088248048568951, "grad_norm": 0.5202969908714294, "learning_rate": 6.995865370665892e-05, "loss": 0.0351, "step": 8383 }, { "epoch": 0.9089332176929749, "grad_norm": 0.3116361200809479, "learning_rate": 6.995502683882201e-05, "loss": 0.0153, "step": 8384 }, { "epoch": 0.9090416305290546, "grad_norm": 0.3722483217716217, "learning_rate": 6.995139997098507e-05, "loss": 0.0495, "step": 8385 }, { "epoch": 0.9091500433651344, "grad_norm": 0.163038969039917, "learning_rate": 6.994777310314812e-05, "loss": 0.0316, "step": 8386 }, { "epoch": 0.9092584562012143, "grad_norm": 0.18731962144374847, "learning_rate": 6.99441462353112e-05, "loss": 0.02, "step": 8387 }, { "epoch": 0.909366869037294, "grad_norm": 0.14739178121089935, "learning_rate": 6.994051936747425e-05, "loss": 0.0141, "step": 8388 }, { "epoch": 0.9094752818733738, "grad_norm": 0.3234034478664398, "learning_rate": 6.993689249963732e-05, "loss": 0.0267, "step": 8389 }, { "epoch": 0.9095836947094535, "grad_norm": 0.3826833963394165, "learning_rate": 6.993326563180038e-05, "loss": 0.0699, "step": 8390 }, { "epoch": 0.9096921075455334, "grad_norm": 0.31679344177246094, "learning_rate": 6.992963876396345e-05, "loss": 0.0224, "step": 8391 }, { "epoch": 0.9098005203816132, "grad_norm": 0.20082014799118042, "learning_rate": 6.992601189612651e-05, "loss": 0.0151, "step": 8392 }, { "epoch": 0.909908933217693, "grad_norm": 0.19795069098472595, "learning_rate": 6.992238502828958e-05, "loss": 0.0164, "step": 8393 }, { "epoch": 0.9100173460537727, "grad_norm": 0.22152388095855713, "learning_rate": 6.991875816045264e-05, "loss": 0.0286, "step": 8394 }, { "epoch": 0.9101257588898526, "grad_norm": 0.41577088832855225, "learning_rate": 6.991513129261569e-05, "loss": 0.0581, "step": 8395 }, { "epoch": 0.9102341717259324, "grad_norm": 0.35967618227005005, "learning_rate": 6.991150442477876e-05, "loss": 0.0516, "step": 8396 }, { "epoch": 0.9103425845620121, "grad_norm": 0.3649493455886841, "learning_rate": 6.990787755694182e-05, "loss": 0.0391, "step": 8397 }, { "epoch": 0.910450997398092, "grad_norm": 0.5219171643257141, "learning_rate": 6.990425068910489e-05, "loss": 0.0265, "step": 8398 }, { "epoch": 0.9105594102341718, "grad_norm": 0.2798958718776703, "learning_rate": 6.990062382126795e-05, "loss": 0.0382, "step": 8399 }, { "epoch": 0.9106678230702515, "grad_norm": 0.29953640699386597, "learning_rate": 6.989699695343102e-05, "loss": 0.0471, "step": 8400 }, { "epoch": 0.9107762359063313, "grad_norm": 0.5325639843940735, "learning_rate": 6.989337008559409e-05, "loss": 0.0366, "step": 8401 }, { "epoch": 0.9108846487424112, "grad_norm": 0.24843037128448486, "learning_rate": 6.988974321775715e-05, "loss": 0.0502, "step": 8402 }, { "epoch": 0.9109930615784909, "grad_norm": 0.4846254885196686, "learning_rate": 6.988611634992022e-05, "loss": 0.0748, "step": 8403 }, { "epoch": 0.9111014744145707, "grad_norm": 0.13851739466190338, "learning_rate": 6.988248948208328e-05, "loss": 0.0195, "step": 8404 }, { "epoch": 0.9112098872506504, "grad_norm": 0.4050498902797699, "learning_rate": 6.987886261424635e-05, "loss": 0.0334, "step": 8405 }, { "epoch": 0.9113183000867303, "grad_norm": 0.419210284948349, "learning_rate": 6.98752357464094e-05, "loss": 0.0467, "step": 8406 }, { "epoch": 0.9114267129228101, "grad_norm": 0.5411938428878784, "learning_rate": 6.987160887857248e-05, "loss": 0.0485, "step": 8407 }, { "epoch": 0.9115351257588898, "grad_norm": 0.32366612553596497, "learning_rate": 6.986798201073553e-05, "loss": 0.024, "step": 8408 }, { "epoch": 0.9116435385949696, "grad_norm": 0.18268011510372162, "learning_rate": 6.986435514289859e-05, "loss": 0.0195, "step": 8409 }, { "epoch": 0.9117519514310495, "grad_norm": 0.17186616361141205, "learning_rate": 6.986072827506166e-05, "loss": 0.0211, "step": 8410 }, { "epoch": 0.9118603642671292, "grad_norm": 0.375074565410614, "learning_rate": 6.985710140722472e-05, "loss": 0.0226, "step": 8411 }, { "epoch": 0.911968777103209, "grad_norm": 0.5121022462844849, "learning_rate": 6.985347453938779e-05, "loss": 0.0647, "step": 8412 }, { "epoch": 0.9120771899392888, "grad_norm": 0.4083606004714966, "learning_rate": 6.984984767155085e-05, "loss": 0.0591, "step": 8413 }, { "epoch": 0.9121856027753686, "grad_norm": 0.3021796941757202, "learning_rate": 6.984622080371392e-05, "loss": 0.0269, "step": 8414 }, { "epoch": 0.9122940156114484, "grad_norm": 0.35687926411628723, "learning_rate": 6.984259393587697e-05, "loss": 0.0471, "step": 8415 }, { "epoch": 0.9124024284475282, "grad_norm": 0.26462289690971375, "learning_rate": 6.983896706804005e-05, "loss": 0.0266, "step": 8416 }, { "epoch": 0.9125108412836079, "grad_norm": 0.18035486340522766, "learning_rate": 6.98353402002031e-05, "loss": 0.0209, "step": 8417 }, { "epoch": 0.9126192541196878, "grad_norm": 0.2668099105358124, "learning_rate": 6.983171333236616e-05, "loss": 0.0279, "step": 8418 }, { "epoch": 0.9127276669557676, "grad_norm": 0.206289604306221, "learning_rate": 6.982808646452924e-05, "loss": 0.0252, "step": 8419 }, { "epoch": 0.9128360797918473, "grad_norm": 0.24594545364379883, "learning_rate": 6.98244595966923e-05, "loss": 0.0296, "step": 8420 }, { "epoch": 0.9129444926279271, "grad_norm": 0.49371856451034546, "learning_rate": 6.982083272885537e-05, "loss": 0.0504, "step": 8421 }, { "epoch": 0.913052905464007, "grad_norm": 0.352736234664917, "learning_rate": 6.981720586101843e-05, "loss": 0.0561, "step": 8422 }, { "epoch": 0.9131613183000867, "grad_norm": 0.6873900294303894, "learning_rate": 6.98135789931815e-05, "loss": 0.0481, "step": 8423 }, { "epoch": 0.9132697311361665, "grad_norm": 0.6008899807929993, "learning_rate": 6.980995212534456e-05, "loss": 0.0628, "step": 8424 }, { "epoch": 0.9133781439722463, "grad_norm": 0.2897401452064514, "learning_rate": 6.980632525750762e-05, "loss": 0.0291, "step": 8425 }, { "epoch": 0.9134865568083261, "grad_norm": 0.17099598050117493, "learning_rate": 6.980269838967069e-05, "loss": 0.0351, "step": 8426 }, { "epoch": 0.9135949696444059, "grad_norm": 0.28777992725372314, "learning_rate": 6.979907152183374e-05, "loss": 0.0392, "step": 8427 }, { "epoch": 0.9137033824804857, "grad_norm": 0.13983656466007233, "learning_rate": 6.979544465399681e-05, "loss": 0.0319, "step": 8428 }, { "epoch": 0.9138117953165655, "grad_norm": 0.4353347718715668, "learning_rate": 6.979181778615987e-05, "loss": 0.0537, "step": 8429 }, { "epoch": 0.9139202081526453, "grad_norm": 0.19493913650512695, "learning_rate": 6.978819091832294e-05, "loss": 0.034, "step": 8430 }, { "epoch": 0.9140286209887251, "grad_norm": 0.3839786648750305, "learning_rate": 6.9784564050486e-05, "loss": 0.0234, "step": 8431 }, { "epoch": 0.9141370338248048, "grad_norm": 0.3136414587497711, "learning_rate": 6.978093718264906e-05, "loss": 0.039, "step": 8432 }, { "epoch": 0.9142454466608847, "grad_norm": 0.4469268321990967, "learning_rate": 6.977731031481213e-05, "loss": 0.0942, "step": 8433 }, { "epoch": 0.9143538594969645, "grad_norm": 0.31927260756492615, "learning_rate": 6.977368344697519e-05, "loss": 0.0506, "step": 8434 }, { "epoch": 0.9144622723330442, "grad_norm": 0.40770983695983887, "learning_rate": 6.977005657913826e-05, "loss": 0.0241, "step": 8435 }, { "epoch": 0.914570685169124, "grad_norm": 0.3489364683628082, "learning_rate": 6.976642971130133e-05, "loss": 0.0383, "step": 8436 }, { "epoch": 0.9146790980052039, "grad_norm": 0.21939237415790558, "learning_rate": 6.97628028434644e-05, "loss": 0.0218, "step": 8437 }, { "epoch": 0.9147875108412836, "grad_norm": 0.18937437236309052, "learning_rate": 6.975917597562746e-05, "loss": 0.0235, "step": 8438 }, { "epoch": 0.9148959236773634, "grad_norm": 0.4963741898536682, "learning_rate": 6.975554910779051e-05, "loss": 0.0343, "step": 8439 }, { "epoch": 0.9150043365134432, "grad_norm": 0.17277920246124268, "learning_rate": 6.975192223995358e-05, "loss": 0.0235, "step": 8440 }, { "epoch": 0.915112749349523, "grad_norm": 0.25626373291015625, "learning_rate": 6.974829537211664e-05, "loss": 0.0384, "step": 8441 }, { "epoch": 0.9152211621856028, "grad_norm": 0.37944722175598145, "learning_rate": 6.974466850427971e-05, "loss": 0.0286, "step": 8442 }, { "epoch": 0.9153295750216826, "grad_norm": 0.28265026211738586, "learning_rate": 6.974104163644277e-05, "loss": 0.0272, "step": 8443 }, { "epoch": 0.9154379878577623, "grad_norm": 0.4081977903842926, "learning_rate": 6.973741476860584e-05, "loss": 0.0522, "step": 8444 }, { "epoch": 0.9155464006938422, "grad_norm": 0.315731942653656, "learning_rate": 6.97337879007689e-05, "loss": 0.0438, "step": 8445 }, { "epoch": 0.915654813529922, "grad_norm": 0.4648219048976898, "learning_rate": 6.973016103293197e-05, "loss": 0.0496, "step": 8446 }, { "epoch": 0.9157632263660017, "grad_norm": 0.4858810007572174, "learning_rate": 6.972653416509503e-05, "loss": 0.0395, "step": 8447 }, { "epoch": 0.9158716392020815, "grad_norm": 0.22417782247066498, "learning_rate": 6.972290729725808e-05, "loss": 0.0201, "step": 8448 }, { "epoch": 0.9159800520381614, "grad_norm": 0.3627330958843231, "learning_rate": 6.971928042942115e-05, "loss": 0.0534, "step": 8449 }, { "epoch": 0.9160884648742411, "grad_norm": 0.2672104835510254, "learning_rate": 6.971565356158421e-05, "loss": 0.0407, "step": 8450 }, { "epoch": 0.9161968777103209, "grad_norm": 0.2385796755552292, "learning_rate": 6.971202669374728e-05, "loss": 0.0253, "step": 8451 }, { "epoch": 0.9163052905464006, "grad_norm": 0.3830993175506592, "learning_rate": 6.970839982591034e-05, "loss": 0.0494, "step": 8452 }, { "epoch": 0.9164137033824805, "grad_norm": 0.7835061550140381, "learning_rate": 6.970477295807341e-05, "loss": 0.0629, "step": 8453 }, { "epoch": 0.9165221162185603, "grad_norm": 0.1637381613254547, "learning_rate": 6.970114609023648e-05, "loss": 0.0237, "step": 8454 }, { "epoch": 0.91663052905464, "grad_norm": 0.11415746808052063, "learning_rate": 6.969751922239954e-05, "loss": 0.0106, "step": 8455 }, { "epoch": 0.9167389418907199, "grad_norm": 0.3122047483921051, "learning_rate": 6.969389235456261e-05, "loss": 0.0316, "step": 8456 }, { "epoch": 0.9168473547267997, "grad_norm": 0.19929414987564087, "learning_rate": 6.969026548672567e-05, "loss": 0.0189, "step": 8457 }, { "epoch": 0.9169557675628794, "grad_norm": 0.5917117595672607, "learning_rate": 6.968663861888874e-05, "loss": 0.0652, "step": 8458 }, { "epoch": 0.9170641803989592, "grad_norm": 0.2602751553058624, "learning_rate": 6.96830117510518e-05, "loss": 0.0277, "step": 8459 }, { "epoch": 0.9171725932350391, "grad_norm": 0.30673179030418396, "learning_rate": 6.967938488321487e-05, "loss": 0.0832, "step": 8460 }, { "epoch": 0.9172810060711188, "grad_norm": 0.24395522475242615, "learning_rate": 6.967575801537792e-05, "loss": 0.018, "step": 8461 }, { "epoch": 0.9173894189071986, "grad_norm": 0.19447898864746094, "learning_rate": 6.967213114754098e-05, "loss": 0.0254, "step": 8462 }, { "epoch": 0.9174978317432784, "grad_norm": 0.2134280949831009, "learning_rate": 6.966850427970405e-05, "loss": 0.0241, "step": 8463 }, { "epoch": 0.9176062445793582, "grad_norm": 0.25996047258377075, "learning_rate": 6.966487741186711e-05, "loss": 0.057, "step": 8464 }, { "epoch": 0.917714657415438, "grad_norm": 0.17591436207294464, "learning_rate": 6.966125054403018e-05, "loss": 0.0152, "step": 8465 }, { "epoch": 0.9178230702515178, "grad_norm": 0.2613908350467682, "learning_rate": 6.965762367619324e-05, "loss": 0.0296, "step": 8466 }, { "epoch": 0.9179314830875975, "grad_norm": 0.6729989051818848, "learning_rate": 6.965399680835631e-05, "loss": 0.0825, "step": 8467 }, { "epoch": 0.9180398959236774, "grad_norm": 0.37464025616645813, "learning_rate": 6.965036994051937e-05, "loss": 0.052, "step": 8468 }, { "epoch": 0.9181483087597572, "grad_norm": 0.3121888041496277, "learning_rate": 6.964674307268244e-05, "loss": 0.021, "step": 8469 }, { "epoch": 0.9182567215958369, "grad_norm": 0.36193394660949707, "learning_rate": 6.96431162048455e-05, "loss": 0.0351, "step": 8470 }, { "epoch": 0.9183651344319167, "grad_norm": 0.5042871832847595, "learning_rate": 6.963948933700856e-05, "loss": 0.06, "step": 8471 }, { "epoch": 0.9184735472679966, "grad_norm": 0.4628365933895111, "learning_rate": 6.963586246917164e-05, "loss": 0.0582, "step": 8472 }, { "epoch": 0.9185819601040763, "grad_norm": 0.3300859332084656, "learning_rate": 6.963223560133469e-05, "loss": 0.0234, "step": 8473 }, { "epoch": 0.9186903729401561, "grad_norm": 0.10445749759674072, "learning_rate": 6.962860873349776e-05, "loss": 0.0181, "step": 8474 }, { "epoch": 0.9187987857762359, "grad_norm": 0.5164387226104736, "learning_rate": 6.962498186566082e-05, "loss": 0.0519, "step": 8475 }, { "epoch": 0.9189071986123157, "grad_norm": 0.3478322923183441, "learning_rate": 6.962135499782388e-05, "loss": 0.0393, "step": 8476 }, { "epoch": 0.9190156114483955, "grad_norm": 0.33510732650756836, "learning_rate": 6.961772812998695e-05, "loss": 0.0643, "step": 8477 }, { "epoch": 0.9191240242844753, "grad_norm": 0.3184563219547272, "learning_rate": 6.961410126215e-05, "loss": 0.0565, "step": 8478 }, { "epoch": 0.919232437120555, "grad_norm": 0.23267696797847748, "learning_rate": 6.961047439431308e-05, "loss": 0.0488, "step": 8479 }, { "epoch": 0.9193408499566349, "grad_norm": 0.25596117973327637, "learning_rate": 6.960684752647613e-05, "loss": 0.0365, "step": 8480 }, { "epoch": 0.9194492627927147, "grad_norm": 0.33697229623794556, "learning_rate": 6.96032206586392e-05, "loss": 0.0481, "step": 8481 }, { "epoch": 0.9195576756287944, "grad_norm": 0.06115918606519699, "learning_rate": 6.959959379080226e-05, "loss": 0.0057, "step": 8482 }, { "epoch": 0.9196660884648742, "grad_norm": 0.299465149641037, "learning_rate": 6.959596692296533e-05, "loss": 0.054, "step": 8483 }, { "epoch": 0.9197745013009541, "grad_norm": 0.5075221657752991, "learning_rate": 6.959234005512839e-05, "loss": 0.0434, "step": 8484 }, { "epoch": 0.9198829141370338, "grad_norm": 0.2697845995426178, "learning_rate": 6.958871318729145e-05, "loss": 0.0338, "step": 8485 }, { "epoch": 0.9199913269731136, "grad_norm": 0.4064680337905884, "learning_rate": 6.958508631945452e-05, "loss": 0.043, "step": 8486 }, { "epoch": 0.9200997398091935, "grad_norm": 0.31664884090423584, "learning_rate": 6.958145945161758e-05, "loss": 0.0385, "step": 8487 }, { "epoch": 0.9202081526452732, "grad_norm": 0.15404443442821503, "learning_rate": 6.957783258378066e-05, "loss": 0.0245, "step": 8488 }, { "epoch": 0.920316565481353, "grad_norm": 0.24515676498413086, "learning_rate": 6.957420571594372e-05, "loss": 0.0424, "step": 8489 }, { "epoch": 0.9204249783174328, "grad_norm": 0.3887629806995392, "learning_rate": 6.957057884810678e-05, "loss": 0.0382, "step": 8490 }, { "epoch": 0.9205333911535126, "grad_norm": 0.43261000514030457, "learning_rate": 6.956695198026985e-05, "loss": 0.0354, "step": 8491 }, { "epoch": 0.9206418039895924, "grad_norm": 0.16100217401981354, "learning_rate": 6.95633251124329e-05, "loss": 0.0103, "step": 8492 }, { "epoch": 0.9207502168256722, "grad_norm": 0.4808197319507599, "learning_rate": 6.955969824459597e-05, "loss": 0.0588, "step": 8493 }, { "epoch": 0.9208586296617519, "grad_norm": 0.16812992095947266, "learning_rate": 6.955607137675903e-05, "loss": 0.0276, "step": 8494 }, { "epoch": 0.9209670424978318, "grad_norm": 0.4839973747730255, "learning_rate": 6.95524445089221e-05, "loss": 0.0583, "step": 8495 }, { "epoch": 0.9210754553339116, "grad_norm": 0.172659233212471, "learning_rate": 6.954881764108516e-05, "loss": 0.0222, "step": 8496 }, { "epoch": 0.9211838681699913, "grad_norm": 0.4159451723098755, "learning_rate": 6.954519077324823e-05, "loss": 0.0472, "step": 8497 }, { "epoch": 0.9212922810060711, "grad_norm": 0.39063388109207153, "learning_rate": 6.954156390541129e-05, "loss": 0.037, "step": 8498 }, { "epoch": 0.921400693842151, "grad_norm": 0.31533780694007874, "learning_rate": 6.953793703757435e-05, "loss": 0.0426, "step": 8499 }, { "epoch": 0.9215091066782307, "grad_norm": 0.2688896954059601, "learning_rate": 6.953431016973742e-05, "loss": 0.0328, "step": 8500 }, { "epoch": 0.9216175195143105, "grad_norm": 0.2261599898338318, "learning_rate": 6.953068330190047e-05, "loss": 0.0213, "step": 8501 }, { "epoch": 0.9217259323503902, "grad_norm": 0.2629387080669403, "learning_rate": 6.952705643406355e-05, "loss": 0.0159, "step": 8502 }, { "epoch": 0.9218343451864701, "grad_norm": 0.22552786767482758, "learning_rate": 6.95234295662266e-05, "loss": 0.0391, "step": 8503 }, { "epoch": 0.9219427580225499, "grad_norm": 0.40110158920288086, "learning_rate": 6.951980269838967e-05, "loss": 0.0434, "step": 8504 }, { "epoch": 0.9220511708586296, "grad_norm": 0.3268282115459442, "learning_rate": 6.951617583055274e-05, "loss": 0.0396, "step": 8505 }, { "epoch": 0.9221595836947094, "grad_norm": 0.3836536407470703, "learning_rate": 6.95125489627158e-05, "loss": 0.0471, "step": 8506 }, { "epoch": 0.9222679965307893, "grad_norm": 0.08251550793647766, "learning_rate": 6.950892209487887e-05, "loss": 0.0099, "step": 8507 }, { "epoch": 0.922376409366869, "grad_norm": 0.1719929426908493, "learning_rate": 6.950529522704193e-05, "loss": 0.016, "step": 8508 }, { "epoch": 0.9224848222029488, "grad_norm": 0.2300253063440323, "learning_rate": 6.9501668359205e-05, "loss": 0.0347, "step": 8509 }, { "epoch": 0.9225932350390286, "grad_norm": 0.31317293643951416, "learning_rate": 6.949804149136806e-05, "loss": 0.0733, "step": 8510 }, { "epoch": 0.9227016478751084, "grad_norm": 0.7592554092407227, "learning_rate": 6.949441462353113e-05, "loss": 0.0853, "step": 8511 }, { "epoch": 0.9228100607111882, "grad_norm": 0.22935816645622253, "learning_rate": 6.949078775569419e-05, "loss": 0.0195, "step": 8512 }, { "epoch": 0.922918473547268, "grad_norm": 0.23264430463314056, "learning_rate": 6.948716088785724e-05, "loss": 0.0203, "step": 8513 }, { "epoch": 0.9230268863833477, "grad_norm": 0.4174952507019043, "learning_rate": 6.948353402002031e-05, "loss": 0.0347, "step": 8514 }, { "epoch": 0.9231352992194276, "grad_norm": 0.2248845398426056, "learning_rate": 6.947990715218337e-05, "loss": 0.0253, "step": 8515 }, { "epoch": 0.9232437120555074, "grad_norm": 0.15609273314476013, "learning_rate": 6.947628028434644e-05, "loss": 0.02, "step": 8516 }, { "epoch": 0.9233521248915871, "grad_norm": 0.19325833022594452, "learning_rate": 6.94726534165095e-05, "loss": 0.0202, "step": 8517 }, { "epoch": 0.923460537727667, "grad_norm": 0.33237630128860474, "learning_rate": 6.946902654867257e-05, "loss": 0.0368, "step": 8518 }, { "epoch": 0.9235689505637468, "grad_norm": 0.6233566403388977, "learning_rate": 6.946539968083563e-05, "loss": 0.0671, "step": 8519 }, { "epoch": 0.9236773633998265, "grad_norm": 0.4504123330116272, "learning_rate": 6.94617728129987e-05, "loss": 0.0591, "step": 8520 }, { "epoch": 0.9237857762359063, "grad_norm": 0.5321471691131592, "learning_rate": 6.945814594516176e-05, "loss": 0.0918, "step": 8521 }, { "epoch": 0.9238941890719862, "grad_norm": 0.5699421167373657, "learning_rate": 6.945451907732483e-05, "loss": 0.0669, "step": 8522 }, { "epoch": 0.9240026019080659, "grad_norm": 0.1165224090218544, "learning_rate": 6.94508922094879e-05, "loss": 0.0094, "step": 8523 }, { "epoch": 0.9241110147441457, "grad_norm": 0.11886594444513321, "learning_rate": 6.944726534165096e-05, "loss": 0.0142, "step": 8524 }, { "epoch": 0.9242194275802255, "grad_norm": 0.6676694750785828, "learning_rate": 6.944363847381403e-05, "loss": 0.0597, "step": 8525 }, { "epoch": 0.9243278404163053, "grad_norm": 0.51043701171875, "learning_rate": 6.944001160597708e-05, "loss": 0.0523, "step": 8526 }, { "epoch": 0.9244362532523851, "grad_norm": 0.5281920433044434, "learning_rate": 6.943638473814015e-05, "loss": 0.0363, "step": 8527 }, { "epoch": 0.9245446660884649, "grad_norm": 0.34789010882377625, "learning_rate": 6.943275787030321e-05, "loss": 0.0557, "step": 8528 }, { "epoch": 0.9246530789245446, "grad_norm": 0.43693846464157104, "learning_rate": 6.942913100246627e-05, "loss": 0.036, "step": 8529 }, { "epoch": 0.9247614917606245, "grad_norm": 0.41876521706581116, "learning_rate": 6.942550413462934e-05, "loss": 0.0542, "step": 8530 }, { "epoch": 0.9248699045967043, "grad_norm": 0.37969139218330383, "learning_rate": 6.94218772667924e-05, "loss": 0.0403, "step": 8531 }, { "epoch": 0.924978317432784, "grad_norm": 0.4498459994792938, "learning_rate": 6.941825039895547e-05, "loss": 0.0384, "step": 8532 }, { "epoch": 0.9250867302688638, "grad_norm": 0.14254720509052277, "learning_rate": 6.941462353111853e-05, "loss": 0.0091, "step": 8533 }, { "epoch": 0.9251951431049437, "grad_norm": 0.20808352530002594, "learning_rate": 6.94109966632816e-05, "loss": 0.0479, "step": 8534 }, { "epoch": 0.9253035559410234, "grad_norm": 0.29586806893348694, "learning_rate": 6.940736979544465e-05, "loss": 0.0208, "step": 8535 }, { "epoch": 0.9254119687771032, "grad_norm": 0.25602203607559204, "learning_rate": 6.940374292760771e-05, "loss": 0.0421, "step": 8536 }, { "epoch": 0.925520381613183, "grad_norm": 0.2109437733888626, "learning_rate": 6.940011605977078e-05, "loss": 0.0399, "step": 8537 }, { "epoch": 0.9256287944492628, "grad_norm": 0.34097525477409363, "learning_rate": 6.939648919193384e-05, "loss": 0.0438, "step": 8538 }, { "epoch": 0.9257372072853426, "grad_norm": 0.18180778622627258, "learning_rate": 6.939286232409691e-05, "loss": 0.0238, "step": 8539 }, { "epoch": 0.9258456201214224, "grad_norm": 0.17245902121067047, "learning_rate": 6.938923545625998e-05, "loss": 0.0278, "step": 8540 }, { "epoch": 0.9259540329575021, "grad_norm": 0.19363874197006226, "learning_rate": 6.938560858842305e-05, "loss": 0.0279, "step": 8541 }, { "epoch": 0.926062445793582, "grad_norm": 0.305395245552063, "learning_rate": 6.938198172058611e-05, "loss": 0.0304, "step": 8542 }, { "epoch": 0.9261708586296618, "grad_norm": 0.9525431394577026, "learning_rate": 6.937835485274917e-05, "loss": 0.0559, "step": 8543 }, { "epoch": 0.9262792714657415, "grad_norm": 0.11403235048055649, "learning_rate": 6.937472798491224e-05, "loss": 0.0108, "step": 8544 }, { "epoch": 0.9263876843018214, "grad_norm": 0.290481835603714, "learning_rate": 6.93711011170753e-05, "loss": 0.05, "step": 8545 }, { "epoch": 0.9264960971379012, "grad_norm": 0.43721166253089905, "learning_rate": 6.936747424923837e-05, "loss": 0.0486, "step": 8546 }, { "epoch": 0.9266045099739809, "grad_norm": 0.325173020362854, "learning_rate": 6.936384738140142e-05, "loss": 0.0397, "step": 8547 }, { "epoch": 0.9267129228100607, "grad_norm": 0.19612066447734833, "learning_rate": 6.93602205135645e-05, "loss": 0.024, "step": 8548 }, { "epoch": 0.9268213356461406, "grad_norm": 0.8131573796272278, "learning_rate": 6.935659364572755e-05, "loss": 0.1107, "step": 8549 }, { "epoch": 0.9269297484822203, "grad_norm": 0.5996159315109253, "learning_rate": 6.935296677789062e-05, "loss": 0.089, "step": 8550 }, { "epoch": 0.9270381613183001, "grad_norm": 0.14391957223415375, "learning_rate": 6.934933991005368e-05, "loss": 0.0233, "step": 8551 }, { "epoch": 0.9271465741543798, "grad_norm": 0.06328721344470978, "learning_rate": 6.934571304221674e-05, "loss": 0.009, "step": 8552 }, { "epoch": 0.9272549869904597, "grad_norm": 0.3900749981403351, "learning_rate": 6.934208617437981e-05, "loss": 0.0692, "step": 8553 }, { "epoch": 0.9273633998265395, "grad_norm": 0.3209894895553589, "learning_rate": 6.933845930654286e-05, "loss": 0.0396, "step": 8554 }, { "epoch": 0.9274718126626192, "grad_norm": 0.21198348701000214, "learning_rate": 6.933483243870594e-05, "loss": 0.0181, "step": 8555 }, { "epoch": 0.927580225498699, "grad_norm": 0.17274534702301025, "learning_rate": 6.933120557086899e-05, "loss": 0.0243, "step": 8556 }, { "epoch": 0.9276886383347789, "grad_norm": 0.4032372832298279, "learning_rate": 6.932757870303206e-05, "loss": 0.0564, "step": 8557 }, { "epoch": 0.9277970511708586, "grad_norm": 0.3276500105857849, "learning_rate": 6.932395183519513e-05, "loss": 0.0311, "step": 8558 }, { "epoch": 0.9279054640069384, "grad_norm": 0.1703615039587021, "learning_rate": 6.932032496735819e-05, "loss": 0.0318, "step": 8559 }, { "epoch": 0.9280138768430182, "grad_norm": 0.21789854764938354, "learning_rate": 6.931669809952126e-05, "loss": 0.0361, "step": 8560 }, { "epoch": 0.928122289679098, "grad_norm": 0.3040849566459656, "learning_rate": 6.931307123168432e-05, "loss": 0.0347, "step": 8561 }, { "epoch": 0.9282307025151778, "grad_norm": 0.45903280377388, "learning_rate": 6.930944436384739e-05, "loss": 0.0294, "step": 8562 }, { "epoch": 0.9283391153512576, "grad_norm": 0.2808953821659088, "learning_rate": 6.930581749601045e-05, "loss": 0.0451, "step": 8563 }, { "epoch": 0.9284475281873373, "grad_norm": 0.2572944462299347, "learning_rate": 6.930219062817352e-05, "loss": 0.0413, "step": 8564 }, { "epoch": 0.9285559410234172, "grad_norm": 0.724628746509552, "learning_rate": 6.929856376033658e-05, "loss": 0.0532, "step": 8565 }, { "epoch": 0.928664353859497, "grad_norm": 0.16281066834926605, "learning_rate": 6.929493689249963e-05, "loss": 0.0213, "step": 8566 }, { "epoch": 0.9287727666955767, "grad_norm": 0.6434361338615417, "learning_rate": 6.92913100246627e-05, "loss": 0.0705, "step": 8567 }, { "epoch": 0.9288811795316565, "grad_norm": 0.35201698541641235, "learning_rate": 6.928768315682576e-05, "loss": 0.0411, "step": 8568 }, { "epoch": 0.9289895923677364, "grad_norm": 0.44125017523765564, "learning_rate": 6.928405628898883e-05, "loss": 0.0378, "step": 8569 }, { "epoch": 0.9290980052038161, "grad_norm": 0.40521183609962463, "learning_rate": 6.928042942115189e-05, "loss": 0.0405, "step": 8570 }, { "epoch": 0.9292064180398959, "grad_norm": 0.3409515619277954, "learning_rate": 6.927680255331496e-05, "loss": 0.0448, "step": 8571 }, { "epoch": 0.9293148308759757, "grad_norm": 0.21743057668209076, "learning_rate": 6.927317568547802e-05, "loss": 0.0326, "step": 8572 }, { "epoch": 0.9294232437120555, "grad_norm": 0.24942541122436523, "learning_rate": 6.926954881764109e-05, "loss": 0.0305, "step": 8573 }, { "epoch": 0.9295316565481353, "grad_norm": 0.21337999403476715, "learning_rate": 6.926592194980416e-05, "loss": 0.0236, "step": 8574 }, { "epoch": 0.9296400693842151, "grad_norm": 0.06954275816679001, "learning_rate": 6.926229508196722e-05, "loss": 0.0086, "step": 8575 }, { "epoch": 0.9297484822202949, "grad_norm": 0.42107507586479187, "learning_rate": 6.925866821413029e-05, "loss": 0.0401, "step": 8576 }, { "epoch": 0.9298568950563747, "grad_norm": 0.5969046354293823, "learning_rate": 6.925504134629335e-05, "loss": 0.0602, "step": 8577 }, { "epoch": 0.9299653078924545, "grad_norm": 0.09065394848585129, "learning_rate": 6.925141447845642e-05, "loss": 0.0082, "step": 8578 }, { "epoch": 0.9300737207285342, "grad_norm": 0.42279115319252014, "learning_rate": 6.924778761061947e-05, "loss": 0.0598, "step": 8579 }, { "epoch": 0.9301821335646141, "grad_norm": 0.3133420944213867, "learning_rate": 6.924416074278253e-05, "loss": 0.035, "step": 8580 }, { "epoch": 0.9302905464006939, "grad_norm": 0.384294331073761, "learning_rate": 6.92405338749456e-05, "loss": 0.0621, "step": 8581 }, { "epoch": 0.9303989592367736, "grad_norm": 0.2041803002357483, "learning_rate": 6.923690700710866e-05, "loss": 0.0135, "step": 8582 }, { "epoch": 0.9305073720728534, "grad_norm": 0.13682906329631805, "learning_rate": 6.923328013927173e-05, "loss": 0.0114, "step": 8583 }, { "epoch": 0.9306157849089333, "grad_norm": 0.8747105002403259, "learning_rate": 6.922965327143479e-05, "loss": 0.0658, "step": 8584 }, { "epoch": 0.930724197745013, "grad_norm": 0.23389534652233124, "learning_rate": 6.922602640359786e-05, "loss": 0.028, "step": 8585 }, { "epoch": 0.9308326105810928, "grad_norm": 0.3979817032814026, "learning_rate": 6.922239953576092e-05, "loss": 0.0662, "step": 8586 }, { "epoch": 0.9309410234171726, "grad_norm": 0.17049957811832428, "learning_rate": 6.921877266792399e-05, "loss": 0.0202, "step": 8587 }, { "epoch": 0.9310494362532524, "grad_norm": 0.2122906893491745, "learning_rate": 6.921514580008704e-05, "loss": 0.0342, "step": 8588 }, { "epoch": 0.9311578490893322, "grad_norm": 0.6829455494880676, "learning_rate": 6.92115189322501e-05, "loss": 0.1026, "step": 8589 }, { "epoch": 0.931266261925412, "grad_norm": 0.2943189740180969, "learning_rate": 6.920789206441317e-05, "loss": 0.0273, "step": 8590 }, { "epoch": 0.9313746747614917, "grad_norm": 0.28905922174453735, "learning_rate": 6.920426519657623e-05, "loss": 0.021, "step": 8591 }, { "epoch": 0.9314830875975716, "grad_norm": 0.3179555833339691, "learning_rate": 6.920063832873931e-05, "loss": 0.0248, "step": 8592 }, { "epoch": 0.9315915004336514, "grad_norm": 0.3220468759536743, "learning_rate": 6.919701146090237e-05, "loss": 0.0398, "step": 8593 }, { "epoch": 0.9316999132697311, "grad_norm": 0.3336625099182129, "learning_rate": 6.919338459306543e-05, "loss": 0.0487, "step": 8594 }, { "epoch": 0.9318083261058109, "grad_norm": 0.17236831784248352, "learning_rate": 6.91897577252285e-05, "loss": 0.03, "step": 8595 }, { "epoch": 0.9319167389418908, "grad_norm": 0.2656181752681732, "learning_rate": 6.918613085739156e-05, "loss": 0.0295, "step": 8596 }, { "epoch": 0.9320251517779705, "grad_norm": 0.38658830523490906, "learning_rate": 6.918250398955463e-05, "loss": 0.0293, "step": 8597 }, { "epoch": 0.9321335646140503, "grad_norm": 0.22693975269794464, "learning_rate": 6.917887712171769e-05, "loss": 0.0551, "step": 8598 }, { "epoch": 0.93224197745013, "grad_norm": 0.23047128319740295, "learning_rate": 6.917525025388076e-05, "loss": 0.0228, "step": 8599 }, { "epoch": 0.9323503902862099, "grad_norm": 0.35055211186408997, "learning_rate": 6.917162338604381e-05, "loss": 0.0597, "step": 8600 }, { "epoch": 0.9324588031222897, "grad_norm": 0.1289522498846054, "learning_rate": 6.916799651820688e-05, "loss": 0.0147, "step": 8601 }, { "epoch": 0.9325672159583694, "grad_norm": 0.3778432011604309, "learning_rate": 6.916436965036994e-05, "loss": 0.0562, "step": 8602 }, { "epoch": 0.9326756287944492, "grad_norm": 0.26677432656288147, "learning_rate": 6.9160742782533e-05, "loss": 0.0247, "step": 8603 }, { "epoch": 0.9327840416305291, "grad_norm": 0.31325221061706543, "learning_rate": 6.915711591469607e-05, "loss": 0.0376, "step": 8604 }, { "epoch": 0.9328924544666088, "grad_norm": 0.4098662734031677, "learning_rate": 6.915348904685913e-05, "loss": 0.0258, "step": 8605 }, { "epoch": 0.9330008673026886, "grad_norm": 0.3435814082622528, "learning_rate": 6.91498621790222e-05, "loss": 0.0549, "step": 8606 }, { "epoch": 0.9331092801387685, "grad_norm": 0.578280508518219, "learning_rate": 6.914623531118526e-05, "loss": 0.0328, "step": 8607 }, { "epoch": 0.9332176929748482, "grad_norm": 0.44715940952301025, "learning_rate": 6.914260844334833e-05, "loss": 0.037, "step": 8608 }, { "epoch": 0.933326105810928, "grad_norm": 0.31809285283088684, "learning_rate": 6.91389815755114e-05, "loss": 0.0301, "step": 8609 }, { "epoch": 0.9334345186470078, "grad_norm": 0.29024285078048706, "learning_rate": 6.913535470767445e-05, "loss": 0.0449, "step": 8610 }, { "epoch": 0.9335429314830876, "grad_norm": 0.2465086281299591, "learning_rate": 6.913172783983753e-05, "loss": 0.039, "step": 8611 }, { "epoch": 0.9336513443191674, "grad_norm": 0.27724704146385193, "learning_rate": 6.912810097200058e-05, "loss": 0.0283, "step": 8612 }, { "epoch": 0.9337597571552472, "grad_norm": 0.366102933883667, "learning_rate": 6.912447410416365e-05, "loss": 0.036, "step": 8613 }, { "epoch": 0.9338681699913269, "grad_norm": 0.15542331337928772, "learning_rate": 6.912084723632671e-05, "loss": 0.0247, "step": 8614 }, { "epoch": 0.9339765828274068, "grad_norm": 0.5526222586631775, "learning_rate": 6.911722036848978e-05, "loss": 0.0437, "step": 8615 }, { "epoch": 0.9340849956634866, "grad_norm": 0.27124324440956116, "learning_rate": 6.911359350065284e-05, "loss": 0.0292, "step": 8616 }, { "epoch": 0.9341934084995663, "grad_norm": 0.33338865637779236, "learning_rate": 6.91099666328159e-05, "loss": 0.0325, "step": 8617 }, { "epoch": 0.9343018213356461, "grad_norm": 0.34307438135147095, "learning_rate": 6.910633976497897e-05, "loss": 0.0399, "step": 8618 }, { "epoch": 0.934410234171726, "grad_norm": 0.27152758836746216, "learning_rate": 6.910271289714203e-05, "loss": 0.0339, "step": 8619 }, { "epoch": 0.9345186470078057, "grad_norm": 0.2764996290206909, "learning_rate": 6.90990860293051e-05, "loss": 0.0338, "step": 8620 }, { "epoch": 0.9346270598438855, "grad_norm": 0.31995531916618347, "learning_rate": 6.909545916146815e-05, "loss": 0.0595, "step": 8621 }, { "epoch": 0.9347354726799653, "grad_norm": 0.45372867584228516, "learning_rate": 6.909183229363122e-05, "loss": 0.0379, "step": 8622 }, { "epoch": 0.9348438855160451, "grad_norm": 0.42665621638298035, "learning_rate": 6.908820542579428e-05, "loss": 0.0747, "step": 8623 }, { "epoch": 0.9349522983521249, "grad_norm": 0.3631729781627655, "learning_rate": 6.908457855795735e-05, "loss": 0.0507, "step": 8624 }, { "epoch": 0.9350607111882047, "grad_norm": 0.17646436393260956, "learning_rate": 6.908095169012041e-05, "loss": 0.0233, "step": 8625 }, { "epoch": 0.9351691240242844, "grad_norm": 0.1964511275291443, "learning_rate": 6.907732482228348e-05, "loss": 0.0172, "step": 8626 }, { "epoch": 0.9352775368603643, "grad_norm": 0.414790540933609, "learning_rate": 6.907369795444655e-05, "loss": 0.0449, "step": 8627 }, { "epoch": 0.9353859496964441, "grad_norm": 0.18442043662071228, "learning_rate": 6.907007108660961e-05, "loss": 0.0263, "step": 8628 }, { "epoch": 0.9354943625325238, "grad_norm": 0.261665940284729, "learning_rate": 6.906644421877268e-05, "loss": 0.0442, "step": 8629 }, { "epoch": 0.9356027753686036, "grad_norm": 0.5843579769134521, "learning_rate": 6.906281735093574e-05, "loss": 0.0287, "step": 8630 }, { "epoch": 0.9357111882046835, "grad_norm": 0.2925998866558075, "learning_rate": 6.905919048309881e-05, "loss": 0.0263, "step": 8631 }, { "epoch": 0.9358196010407632, "grad_norm": 0.26907652616500854, "learning_rate": 6.905556361526187e-05, "loss": 0.0279, "step": 8632 }, { "epoch": 0.935928013876843, "grad_norm": 0.28781500458717346, "learning_rate": 6.905193674742492e-05, "loss": 0.0461, "step": 8633 }, { "epoch": 0.9360364267129229, "grad_norm": 0.2106981873512268, "learning_rate": 6.9048309879588e-05, "loss": 0.0231, "step": 8634 }, { "epoch": 0.9361448395490026, "grad_norm": 0.27734941244125366, "learning_rate": 6.904468301175105e-05, "loss": 0.0296, "step": 8635 }, { "epoch": 0.9362532523850824, "grad_norm": 0.27554282546043396, "learning_rate": 6.904105614391412e-05, "loss": 0.0588, "step": 8636 }, { "epoch": 0.9363616652211622, "grad_norm": 0.24190625548362732, "learning_rate": 6.903742927607718e-05, "loss": 0.0256, "step": 8637 }, { "epoch": 0.936470078057242, "grad_norm": 0.4930984079837799, "learning_rate": 6.903380240824025e-05, "loss": 0.0478, "step": 8638 }, { "epoch": 0.9365784908933218, "grad_norm": 0.34532538056373596, "learning_rate": 6.903017554040331e-05, "loss": 0.0268, "step": 8639 }, { "epoch": 0.9366869037294016, "grad_norm": 0.30002355575561523, "learning_rate": 6.902654867256638e-05, "loss": 0.0317, "step": 8640 }, { "epoch": 0.9367953165654813, "grad_norm": 0.3726608455181122, "learning_rate": 6.902292180472944e-05, "loss": 0.0513, "step": 8641 }, { "epoch": 0.9369037294015612, "grad_norm": 0.2963533103466034, "learning_rate": 6.901929493689249e-05, "loss": 0.0889, "step": 8642 }, { "epoch": 0.937012142237641, "grad_norm": 0.42297637462615967, "learning_rate": 6.901566806905558e-05, "loss": 0.04, "step": 8643 }, { "epoch": 0.9371205550737207, "grad_norm": 0.2924095690250397, "learning_rate": 6.901204120121863e-05, "loss": 0.0242, "step": 8644 }, { "epoch": 0.9372289679098005, "grad_norm": 0.26578253507614136, "learning_rate": 6.90084143333817e-05, "loss": 0.0445, "step": 8645 }, { "epoch": 0.9373373807458804, "grad_norm": 0.40864646434783936, "learning_rate": 6.900478746554476e-05, "loss": 0.0662, "step": 8646 }, { "epoch": 0.9374457935819601, "grad_norm": 0.4393713176250458, "learning_rate": 6.900116059770782e-05, "loss": 0.028, "step": 8647 }, { "epoch": 0.9375542064180399, "grad_norm": 0.18935570120811462, "learning_rate": 6.899753372987089e-05, "loss": 0.0211, "step": 8648 }, { "epoch": 0.9376626192541196, "grad_norm": 0.30201980471611023, "learning_rate": 6.899390686203395e-05, "loss": 0.0542, "step": 8649 }, { "epoch": 0.9377710320901995, "grad_norm": 0.4623783230781555, "learning_rate": 6.899027999419702e-05, "loss": 0.0278, "step": 8650 }, { "epoch": 0.9378794449262793, "grad_norm": 0.3435203433036804, "learning_rate": 6.898665312636008e-05, "loss": 0.0571, "step": 8651 }, { "epoch": 0.937987857762359, "grad_norm": 0.36470991373062134, "learning_rate": 6.898302625852315e-05, "loss": 0.0256, "step": 8652 }, { "epoch": 0.9380962705984388, "grad_norm": 0.253609299659729, "learning_rate": 6.89793993906862e-05, "loss": 0.0262, "step": 8653 }, { "epoch": 0.9382046834345187, "grad_norm": 0.2601398527622223, "learning_rate": 6.897577252284928e-05, "loss": 0.0276, "step": 8654 }, { "epoch": 0.9383130962705984, "grad_norm": 0.24173741042613983, "learning_rate": 6.897214565501233e-05, "loss": 0.0329, "step": 8655 }, { "epoch": 0.9384215091066782, "grad_norm": 0.3278713524341583, "learning_rate": 6.896851878717539e-05, "loss": 0.0367, "step": 8656 }, { "epoch": 0.938529921942758, "grad_norm": 0.36384233832359314, "learning_rate": 6.896489191933846e-05, "loss": 0.0286, "step": 8657 }, { "epoch": 0.9386383347788378, "grad_norm": 0.498969703912735, "learning_rate": 6.896126505150152e-05, "loss": 0.0403, "step": 8658 }, { "epoch": 0.9387467476149176, "grad_norm": 0.2550908625125885, "learning_rate": 6.895763818366459e-05, "loss": 0.0254, "step": 8659 }, { "epoch": 0.9388551604509974, "grad_norm": 0.25039681792259216, "learning_rate": 6.895401131582765e-05, "loss": 0.0357, "step": 8660 }, { "epoch": 0.9389635732870771, "grad_norm": 0.41600775718688965, "learning_rate": 6.895038444799072e-05, "loss": 0.0414, "step": 8661 }, { "epoch": 0.939071986123157, "grad_norm": 0.34226009249687195, "learning_rate": 6.894675758015379e-05, "loss": 0.0517, "step": 8662 }, { "epoch": 0.9391803989592368, "grad_norm": 0.8694440722465515, "learning_rate": 6.894313071231685e-05, "loss": 0.0962, "step": 8663 }, { "epoch": 0.9392888117953165, "grad_norm": 0.20600959658622742, "learning_rate": 6.893950384447992e-05, "loss": 0.022, "step": 8664 }, { "epoch": 0.9393972246313964, "grad_norm": 0.12554749846458435, "learning_rate": 6.893587697664297e-05, "loss": 0.0236, "step": 8665 }, { "epoch": 0.9395056374674762, "grad_norm": 0.2158922404050827, "learning_rate": 6.893225010880604e-05, "loss": 0.0515, "step": 8666 }, { "epoch": 0.9396140503035559, "grad_norm": 0.1767231673002243, "learning_rate": 6.89286232409691e-05, "loss": 0.0259, "step": 8667 }, { "epoch": 0.9397224631396357, "grad_norm": 0.35357868671417236, "learning_rate": 6.892499637313217e-05, "loss": 0.0336, "step": 8668 }, { "epoch": 0.9398308759757156, "grad_norm": 0.5042795538902283, "learning_rate": 6.892136950529523e-05, "loss": 0.074, "step": 8669 }, { "epoch": 0.9399392888117953, "grad_norm": 0.531128466129303, "learning_rate": 6.891774263745829e-05, "loss": 0.0541, "step": 8670 }, { "epoch": 0.9400477016478751, "grad_norm": 0.3574438989162445, "learning_rate": 6.891411576962136e-05, "loss": 0.025, "step": 8671 }, { "epoch": 0.9401561144839549, "grad_norm": 0.19234006106853485, "learning_rate": 6.891048890178442e-05, "loss": 0.0252, "step": 8672 }, { "epoch": 0.9402645273200347, "grad_norm": 0.48847055435180664, "learning_rate": 6.890686203394749e-05, "loss": 0.0729, "step": 8673 }, { "epoch": 0.9403729401561145, "grad_norm": 0.10288036614656448, "learning_rate": 6.890323516611054e-05, "loss": 0.0114, "step": 8674 }, { "epoch": 0.9404813529921943, "grad_norm": 0.6496436595916748, "learning_rate": 6.889960829827361e-05, "loss": 0.0501, "step": 8675 }, { "epoch": 0.940589765828274, "grad_norm": 0.3196277320384979, "learning_rate": 6.889598143043667e-05, "loss": 0.0531, "step": 8676 }, { "epoch": 0.9406981786643539, "grad_norm": 0.088812455534935, "learning_rate": 6.889235456259974e-05, "loss": 0.0062, "step": 8677 }, { "epoch": 0.9408065915004337, "grad_norm": 0.32836824655532837, "learning_rate": 6.888872769476281e-05, "loss": 0.0402, "step": 8678 }, { "epoch": 0.9409150043365134, "grad_norm": 0.2462863177061081, "learning_rate": 6.888510082692587e-05, "loss": 0.0175, "step": 8679 }, { "epoch": 0.9410234171725932, "grad_norm": 0.26472002267837524, "learning_rate": 6.888147395908894e-05, "loss": 0.0354, "step": 8680 }, { "epoch": 0.9411318300086731, "grad_norm": 0.17415876686573029, "learning_rate": 6.8877847091252e-05, "loss": 0.03, "step": 8681 }, { "epoch": 0.9412402428447528, "grad_norm": 0.45655786991119385, "learning_rate": 6.887422022341507e-05, "loss": 0.0308, "step": 8682 }, { "epoch": 0.9413486556808326, "grad_norm": 0.46916118264198303, "learning_rate": 6.887059335557813e-05, "loss": 0.0571, "step": 8683 }, { "epoch": 0.9414570685169124, "grad_norm": 0.1713295727968216, "learning_rate": 6.886696648774119e-05, "loss": 0.023, "step": 8684 }, { "epoch": 0.9415654813529922, "grad_norm": 0.26562976837158203, "learning_rate": 6.886333961990426e-05, "loss": 0.0446, "step": 8685 }, { "epoch": 0.941673894189072, "grad_norm": 0.4710153639316559, "learning_rate": 6.885971275206731e-05, "loss": 0.0366, "step": 8686 }, { "epoch": 0.9417823070251518, "grad_norm": 0.37798988819122314, "learning_rate": 6.885608588423038e-05, "loss": 0.0582, "step": 8687 }, { "epoch": 0.9418907198612315, "grad_norm": 0.43419575691223145, "learning_rate": 6.885245901639344e-05, "loss": 0.075, "step": 8688 }, { "epoch": 0.9419991326973114, "grad_norm": 0.33423471450805664, "learning_rate": 6.884883214855651e-05, "loss": 0.0328, "step": 8689 }, { "epoch": 0.9421075455333912, "grad_norm": 0.2612980008125305, "learning_rate": 6.884520528071957e-05, "loss": 0.0575, "step": 8690 }, { "epoch": 0.9422159583694709, "grad_norm": 0.1126309186220169, "learning_rate": 6.884157841288264e-05, "loss": 0.0127, "step": 8691 }, { "epoch": 0.9423243712055508, "grad_norm": 0.16759845614433289, "learning_rate": 6.88379515450457e-05, "loss": 0.0259, "step": 8692 }, { "epoch": 0.9424327840416306, "grad_norm": 0.3084397614002228, "learning_rate": 6.883432467720876e-05, "loss": 0.0411, "step": 8693 }, { "epoch": 0.9425411968777103, "grad_norm": 0.48917514085769653, "learning_rate": 6.883069780937183e-05, "loss": 0.0977, "step": 8694 }, { "epoch": 0.9426496097137901, "grad_norm": 0.236274853348732, "learning_rate": 6.88270709415349e-05, "loss": 0.0486, "step": 8695 }, { "epoch": 0.94275802254987, "grad_norm": 0.2895980477333069, "learning_rate": 6.882344407369797e-05, "loss": 0.0243, "step": 8696 }, { "epoch": 0.9428664353859497, "grad_norm": 0.23061028122901917, "learning_rate": 6.881981720586103e-05, "loss": 0.0465, "step": 8697 }, { "epoch": 0.9429748482220295, "grad_norm": 0.08692857623100281, "learning_rate": 6.881619033802408e-05, "loss": 0.0164, "step": 8698 }, { "epoch": 0.9430832610581092, "grad_norm": 0.24876539409160614, "learning_rate": 6.881256347018715e-05, "loss": 0.0414, "step": 8699 }, { "epoch": 0.9431916738941891, "grad_norm": 0.28399696946144104, "learning_rate": 6.880893660235021e-05, "loss": 0.0432, "step": 8700 }, { "epoch": 0.9433000867302689, "grad_norm": 0.1488759070634842, "learning_rate": 6.880530973451328e-05, "loss": 0.0209, "step": 8701 }, { "epoch": 0.9434084995663486, "grad_norm": 0.3392462730407715, "learning_rate": 6.880168286667634e-05, "loss": 0.0426, "step": 8702 }, { "epoch": 0.9435169124024284, "grad_norm": 0.40193068981170654, "learning_rate": 6.879805599883941e-05, "loss": 0.0466, "step": 8703 }, { "epoch": 0.9436253252385083, "grad_norm": 0.45029327273368835, "learning_rate": 6.879442913100247e-05, "loss": 0.04, "step": 8704 }, { "epoch": 0.943733738074588, "grad_norm": 0.2800440192222595, "learning_rate": 6.879080226316554e-05, "loss": 0.0272, "step": 8705 }, { "epoch": 0.9438421509106678, "grad_norm": 0.5636683106422424, "learning_rate": 6.87871753953286e-05, "loss": 0.0748, "step": 8706 }, { "epoch": 0.9439505637467476, "grad_norm": 0.242878258228302, "learning_rate": 6.878354852749165e-05, "loss": 0.0157, "step": 8707 }, { "epoch": 0.9440589765828274, "grad_norm": 0.34386390447616577, "learning_rate": 6.877992165965472e-05, "loss": 0.0272, "step": 8708 }, { "epoch": 0.9441673894189072, "grad_norm": 0.43182572722435, "learning_rate": 6.877629479181778e-05, "loss": 0.0273, "step": 8709 }, { "epoch": 0.944275802254987, "grad_norm": 0.3301927447319031, "learning_rate": 6.877266792398085e-05, "loss": 0.0713, "step": 8710 }, { "epoch": 0.9443842150910667, "grad_norm": 0.4964463710784912, "learning_rate": 6.876904105614391e-05, "loss": 0.0596, "step": 8711 }, { "epoch": 0.9444926279271466, "grad_norm": 0.7024761438369751, "learning_rate": 6.8765414188307e-05, "loss": 0.0752, "step": 8712 }, { "epoch": 0.9446010407632264, "grad_norm": 0.26784783601760864, "learning_rate": 6.876178732047005e-05, "loss": 0.0196, "step": 8713 }, { "epoch": 0.9447094535993061, "grad_norm": 0.3245520293712616, "learning_rate": 6.875816045263311e-05, "loss": 0.0642, "step": 8714 }, { "epoch": 0.9448178664353859, "grad_norm": 0.16515159606933594, "learning_rate": 6.875453358479618e-05, "loss": 0.0261, "step": 8715 }, { "epoch": 0.9449262792714658, "grad_norm": 0.24609895050525665, "learning_rate": 6.875090671695924e-05, "loss": 0.0415, "step": 8716 }, { "epoch": 0.9450346921075455, "grad_norm": 0.3225051760673523, "learning_rate": 6.874727984912231e-05, "loss": 0.0457, "step": 8717 }, { "epoch": 0.9451431049436253, "grad_norm": 0.166934534907341, "learning_rate": 6.874365298128536e-05, "loss": 0.0355, "step": 8718 }, { "epoch": 0.9452515177797051, "grad_norm": 0.08864831924438477, "learning_rate": 6.874002611344844e-05, "loss": 0.0129, "step": 8719 }, { "epoch": 0.9453599306157849, "grad_norm": 0.2417474240064621, "learning_rate": 6.873639924561149e-05, "loss": 0.0244, "step": 8720 }, { "epoch": 0.9454683434518647, "grad_norm": 0.10082928091287613, "learning_rate": 6.873277237777455e-05, "loss": 0.0081, "step": 8721 }, { "epoch": 0.9455767562879445, "grad_norm": 0.25141990184783936, "learning_rate": 6.872914550993762e-05, "loss": 0.0592, "step": 8722 }, { "epoch": 0.9456851691240243, "grad_norm": 0.4059223234653473, "learning_rate": 6.872551864210068e-05, "loss": 0.0958, "step": 8723 }, { "epoch": 0.9457935819601041, "grad_norm": 0.07906250655651093, "learning_rate": 6.872189177426375e-05, "loss": 0.0106, "step": 8724 }, { "epoch": 0.9459019947961839, "grad_norm": 0.2609533369541168, "learning_rate": 6.87182649064268e-05, "loss": 0.0472, "step": 8725 }, { "epoch": 0.9460104076322636, "grad_norm": 0.2844875156879425, "learning_rate": 6.871463803858988e-05, "loss": 0.0462, "step": 8726 }, { "epoch": 0.9461188204683435, "grad_norm": 0.21783757209777832, "learning_rate": 6.871101117075293e-05, "loss": 0.0248, "step": 8727 }, { "epoch": 0.9462272333044233, "grad_norm": 0.15437407791614532, "learning_rate": 6.8707384302916e-05, "loss": 0.0212, "step": 8728 }, { "epoch": 0.946335646140503, "grad_norm": 0.33628445863723755, "learning_rate": 6.870375743507906e-05, "loss": 0.0737, "step": 8729 }, { "epoch": 0.9464440589765828, "grad_norm": 0.2105012834072113, "learning_rate": 6.870013056724213e-05, "loss": 0.0235, "step": 8730 }, { "epoch": 0.9465524718126627, "grad_norm": 0.2904483675956726, "learning_rate": 6.86965036994052e-05, "loss": 0.0277, "step": 8731 }, { "epoch": 0.9466608846487424, "grad_norm": 0.3527714014053345, "learning_rate": 6.869287683156826e-05, "loss": 0.0401, "step": 8732 }, { "epoch": 0.9467692974848222, "grad_norm": 0.20390109717845917, "learning_rate": 6.868924996373133e-05, "loss": 0.0276, "step": 8733 }, { "epoch": 0.946877710320902, "grad_norm": 0.23532013595104218, "learning_rate": 6.868562309589439e-05, "loss": 0.043, "step": 8734 }, { "epoch": 0.9469861231569818, "grad_norm": 0.27937552332878113, "learning_rate": 6.868199622805746e-05, "loss": 0.0497, "step": 8735 }, { "epoch": 0.9470945359930616, "grad_norm": 0.18809890747070312, "learning_rate": 6.867836936022052e-05, "loss": 0.0356, "step": 8736 }, { "epoch": 0.9472029488291414, "grad_norm": 0.13418184220790863, "learning_rate": 6.867474249238358e-05, "loss": 0.0115, "step": 8737 }, { "epoch": 0.9473113616652211, "grad_norm": 0.405094176530838, "learning_rate": 6.867111562454665e-05, "loss": 0.0295, "step": 8738 }, { "epoch": 0.947419774501301, "grad_norm": 0.4193269908428192, "learning_rate": 6.86674887567097e-05, "loss": 0.0286, "step": 8739 }, { "epoch": 0.9475281873373808, "grad_norm": 0.09749875962734222, "learning_rate": 6.866386188887277e-05, "loss": 0.0109, "step": 8740 }, { "epoch": 0.9476366001734605, "grad_norm": 0.19923505187034607, "learning_rate": 6.866023502103583e-05, "loss": 0.0228, "step": 8741 }, { "epoch": 0.9477450130095403, "grad_norm": 0.2991476356983185, "learning_rate": 6.86566081531989e-05, "loss": 0.012, "step": 8742 }, { "epoch": 0.9478534258456202, "grad_norm": 0.19499701261520386, "learning_rate": 6.865298128536196e-05, "loss": 0.0207, "step": 8743 }, { "epoch": 0.9479618386816999, "grad_norm": 0.13439789414405823, "learning_rate": 6.864935441752503e-05, "loss": 0.0297, "step": 8744 }, { "epoch": 0.9480702515177797, "grad_norm": 0.4281991422176361, "learning_rate": 6.864572754968809e-05, "loss": 0.0486, "step": 8745 }, { "epoch": 0.9481786643538594, "grad_norm": 0.7136232256889343, "learning_rate": 6.864210068185115e-05, "loss": 0.109, "step": 8746 }, { "epoch": 0.9482870771899393, "grad_norm": 0.2562566101551056, "learning_rate": 6.863847381401423e-05, "loss": 0.0243, "step": 8747 }, { "epoch": 0.9483954900260191, "grad_norm": 0.3018079698085785, "learning_rate": 6.863484694617729e-05, "loss": 0.0492, "step": 8748 }, { "epoch": 0.9485039028620988, "grad_norm": 0.24363689124584198, "learning_rate": 6.863122007834036e-05, "loss": 0.0154, "step": 8749 }, { "epoch": 0.9486123156981786, "grad_norm": 0.5201489329338074, "learning_rate": 6.862759321050342e-05, "loss": 0.1053, "step": 8750 }, { "epoch": 0.9487207285342585, "grad_norm": 0.3226419985294342, "learning_rate": 6.862396634266647e-05, "loss": 0.0352, "step": 8751 }, { "epoch": 0.9488291413703382, "grad_norm": 0.2936408519744873, "learning_rate": 6.862033947482954e-05, "loss": 0.0329, "step": 8752 }, { "epoch": 0.948937554206418, "grad_norm": 0.1701357662677765, "learning_rate": 6.86167126069926e-05, "loss": 0.0263, "step": 8753 }, { "epoch": 0.9490459670424979, "grad_norm": 0.39110878109931946, "learning_rate": 6.861308573915567e-05, "loss": 0.0261, "step": 8754 }, { "epoch": 0.9491543798785776, "grad_norm": 0.4418279528617859, "learning_rate": 6.860945887131873e-05, "loss": 0.0629, "step": 8755 }, { "epoch": 0.9492627927146574, "grad_norm": 0.2864217758178711, "learning_rate": 6.86058320034818e-05, "loss": 0.0359, "step": 8756 }, { "epoch": 0.9493712055507372, "grad_norm": 0.7096542119979858, "learning_rate": 6.860220513564486e-05, "loss": 0.059, "step": 8757 }, { "epoch": 0.949479618386817, "grad_norm": 0.30559414625167847, "learning_rate": 6.859857826780793e-05, "loss": 0.0304, "step": 8758 }, { "epoch": 0.9495880312228968, "grad_norm": 0.5686612725257874, "learning_rate": 6.859495139997099e-05, "loss": 0.0507, "step": 8759 }, { "epoch": 0.9496964440589766, "grad_norm": 0.4280737042427063, "learning_rate": 6.859132453213404e-05, "loss": 0.0292, "step": 8760 }, { "epoch": 0.9498048568950563, "grad_norm": 0.3767176568508148, "learning_rate": 6.858769766429711e-05, "loss": 0.0948, "step": 8761 }, { "epoch": 0.9499132697311362, "grad_norm": 0.19638410210609436, "learning_rate": 6.858407079646017e-05, "loss": 0.0271, "step": 8762 }, { "epoch": 0.950021682567216, "grad_norm": 0.09284645318984985, "learning_rate": 6.858044392862324e-05, "loss": 0.0201, "step": 8763 }, { "epoch": 0.9501300954032957, "grad_norm": 0.30662670731544495, "learning_rate": 6.857681706078631e-05, "loss": 0.036, "step": 8764 }, { "epoch": 0.9502385082393755, "grad_norm": 0.4934841990470886, "learning_rate": 6.857319019294937e-05, "loss": 0.0421, "step": 8765 }, { "epoch": 0.9503469210754554, "grad_norm": 0.13816076517105103, "learning_rate": 6.856956332511244e-05, "loss": 0.0175, "step": 8766 }, { "epoch": 0.9504553339115351, "grad_norm": 0.16534768044948578, "learning_rate": 6.85659364572755e-05, "loss": 0.0215, "step": 8767 }, { "epoch": 0.9505637467476149, "grad_norm": 0.22122037410736084, "learning_rate": 6.856230958943857e-05, "loss": 0.0277, "step": 8768 }, { "epoch": 0.9506721595836947, "grad_norm": 0.3052745461463928, "learning_rate": 6.855868272160163e-05, "loss": 0.0364, "step": 8769 }, { "epoch": 0.9507805724197745, "grad_norm": 0.2515302896499634, "learning_rate": 6.85550558537647e-05, "loss": 0.0426, "step": 8770 }, { "epoch": 0.9508889852558543, "grad_norm": 0.3743172883987427, "learning_rate": 6.855142898592776e-05, "loss": 0.0299, "step": 8771 }, { "epoch": 0.9509973980919341, "grad_norm": 0.25547540187835693, "learning_rate": 6.854780211809083e-05, "loss": 0.0295, "step": 8772 }, { "epoch": 0.9511058109280138, "grad_norm": 0.21817614138126373, "learning_rate": 6.854417525025388e-05, "loss": 0.0389, "step": 8773 }, { "epoch": 0.9512142237640937, "grad_norm": 0.34734779596328735, "learning_rate": 6.854054838241694e-05, "loss": 0.0526, "step": 8774 }, { "epoch": 0.9513226366001735, "grad_norm": 0.19315259158611298, "learning_rate": 6.853692151458001e-05, "loss": 0.0207, "step": 8775 }, { "epoch": 0.9514310494362532, "grad_norm": 0.309703528881073, "learning_rate": 6.853329464674307e-05, "loss": 0.0663, "step": 8776 }, { "epoch": 0.951539462272333, "grad_norm": 0.158573180437088, "learning_rate": 6.852966777890614e-05, "loss": 0.026, "step": 8777 }, { "epoch": 0.9516478751084129, "grad_norm": 0.25529539585113525, "learning_rate": 6.85260409110692e-05, "loss": 0.0175, "step": 8778 }, { "epoch": 0.9517562879444926, "grad_norm": 0.36576277017593384, "learning_rate": 6.852241404323227e-05, "loss": 0.0515, "step": 8779 }, { "epoch": 0.9518647007805724, "grad_norm": 0.22533801198005676, "learning_rate": 6.851878717539533e-05, "loss": 0.0249, "step": 8780 }, { "epoch": 0.9519731136166523, "grad_norm": 0.09284713119268417, "learning_rate": 6.85151603075584e-05, "loss": 0.0206, "step": 8781 }, { "epoch": 0.952081526452732, "grad_norm": 0.44394275546073914, "learning_rate": 6.851153343972147e-05, "loss": 0.0404, "step": 8782 }, { "epoch": 0.9521899392888118, "grad_norm": 0.2749615013599396, "learning_rate": 6.850790657188452e-05, "loss": 0.0264, "step": 8783 }, { "epoch": 0.9522983521248916, "grad_norm": 0.20786413550376892, "learning_rate": 6.85042797040476e-05, "loss": 0.0268, "step": 8784 }, { "epoch": 0.9524067649609714, "grad_norm": 0.3471296429634094, "learning_rate": 6.850065283621065e-05, "loss": 0.0414, "step": 8785 }, { "epoch": 0.9525151777970512, "grad_norm": 0.5356845259666443, "learning_rate": 6.849702596837372e-05, "loss": 0.0246, "step": 8786 }, { "epoch": 0.952623590633131, "grad_norm": 0.19496816396713257, "learning_rate": 6.849339910053678e-05, "loss": 0.0148, "step": 8787 }, { "epoch": 0.9527320034692107, "grad_norm": 0.47466331720352173, "learning_rate": 6.848977223269984e-05, "loss": 0.0336, "step": 8788 }, { "epoch": 0.9528404163052906, "grad_norm": 0.22880496084690094, "learning_rate": 6.848614536486291e-05, "loss": 0.0463, "step": 8789 }, { "epoch": 0.9529488291413704, "grad_norm": 0.29763397574424744, "learning_rate": 6.848251849702597e-05, "loss": 0.0655, "step": 8790 }, { "epoch": 0.9530572419774501, "grad_norm": 0.28722497820854187, "learning_rate": 6.847889162918904e-05, "loss": 0.0303, "step": 8791 }, { "epoch": 0.9531656548135299, "grad_norm": 0.25176262855529785, "learning_rate": 6.84752647613521e-05, "loss": 0.0305, "step": 8792 }, { "epoch": 0.9532740676496098, "grad_norm": 0.15735414624214172, "learning_rate": 6.847163789351517e-05, "loss": 0.0137, "step": 8793 }, { "epoch": 0.9533824804856895, "grad_norm": 0.33342841267585754, "learning_rate": 6.846801102567822e-05, "loss": 0.0392, "step": 8794 }, { "epoch": 0.9534908933217693, "grad_norm": 0.23496802151203156, "learning_rate": 6.84643841578413e-05, "loss": 0.0347, "step": 8795 }, { "epoch": 0.953599306157849, "grad_norm": 0.37684863805770874, "learning_rate": 6.846075729000435e-05, "loss": 0.0465, "step": 8796 }, { "epoch": 0.9537077189939289, "grad_norm": 0.23553508520126343, "learning_rate": 6.845713042216741e-05, "loss": 0.0363, "step": 8797 }, { "epoch": 0.9538161318300087, "grad_norm": 0.2070106565952301, "learning_rate": 6.845350355433048e-05, "loss": 0.0294, "step": 8798 }, { "epoch": 0.9539245446660884, "grad_norm": 0.23853082954883575, "learning_rate": 6.844987668649355e-05, "loss": 0.0218, "step": 8799 }, { "epoch": 0.9540329575021682, "grad_norm": 0.5361216068267822, "learning_rate": 6.844624981865662e-05, "loss": 0.0572, "step": 8800 }, { "epoch": 0.9541413703382481, "grad_norm": 0.4648052155971527, "learning_rate": 6.844262295081968e-05, "loss": 0.0388, "step": 8801 }, { "epoch": 0.9542497831743278, "grad_norm": 0.48721975088119507, "learning_rate": 6.843899608298274e-05, "loss": 0.0327, "step": 8802 }, { "epoch": 0.9543581960104076, "grad_norm": 0.2502445876598358, "learning_rate": 6.843536921514581e-05, "loss": 0.0423, "step": 8803 }, { "epoch": 0.9544666088464874, "grad_norm": 0.41584154963493347, "learning_rate": 6.843174234730886e-05, "loss": 0.042, "step": 8804 }, { "epoch": 0.9545750216825672, "grad_norm": 0.14202781021595, "learning_rate": 6.842811547947193e-05, "loss": 0.0239, "step": 8805 }, { "epoch": 0.954683434518647, "grad_norm": 0.15055176615715027, "learning_rate": 6.842448861163499e-05, "loss": 0.0214, "step": 8806 }, { "epoch": 0.9547918473547268, "grad_norm": 0.45419594645500183, "learning_rate": 6.842086174379806e-05, "loss": 0.0696, "step": 8807 }, { "epoch": 0.9549002601908065, "grad_norm": 0.29889822006225586, "learning_rate": 6.841723487596112e-05, "loss": 0.0463, "step": 8808 }, { "epoch": 0.9550086730268864, "grad_norm": 0.366976797580719, "learning_rate": 6.841360800812419e-05, "loss": 0.0414, "step": 8809 }, { "epoch": 0.9551170858629662, "grad_norm": 0.17303895950317383, "learning_rate": 6.840998114028725e-05, "loss": 0.0214, "step": 8810 }, { "epoch": 0.9552254986990459, "grad_norm": 0.2237340807914734, "learning_rate": 6.84063542724503e-05, "loss": 0.0127, "step": 8811 }, { "epoch": 0.9553339115351258, "grad_norm": 0.18123681843280792, "learning_rate": 6.840272740461338e-05, "loss": 0.0247, "step": 8812 }, { "epoch": 0.9554423243712056, "grad_norm": 0.3415771722793579, "learning_rate": 6.839910053677643e-05, "loss": 0.0396, "step": 8813 }, { "epoch": 0.9555507372072853, "grad_norm": 0.34767696261405945, "learning_rate": 6.83954736689395e-05, "loss": 0.0318, "step": 8814 }, { "epoch": 0.9556591500433651, "grad_norm": 0.33737123012542725, "learning_rate": 6.839184680110256e-05, "loss": 0.083, "step": 8815 }, { "epoch": 0.955767562879445, "grad_norm": 0.5781451463699341, "learning_rate": 6.838821993326565e-05, "loss": 0.034, "step": 8816 }, { "epoch": 0.9558759757155247, "grad_norm": 0.3472954034805298, "learning_rate": 6.83845930654287e-05, "loss": 0.0384, "step": 8817 }, { "epoch": 0.9559843885516045, "grad_norm": 0.21248167753219604, "learning_rate": 6.838096619759176e-05, "loss": 0.0125, "step": 8818 }, { "epoch": 0.9560928013876843, "grad_norm": 0.35867100954055786, "learning_rate": 6.837733932975483e-05, "loss": 0.0319, "step": 8819 }, { "epoch": 0.9562012142237641, "grad_norm": 0.3152291476726532, "learning_rate": 6.837371246191789e-05, "loss": 0.0333, "step": 8820 }, { "epoch": 0.9563096270598439, "grad_norm": 0.33742555975914, "learning_rate": 6.837008559408096e-05, "loss": 0.0213, "step": 8821 }, { "epoch": 0.9564180398959237, "grad_norm": 0.280011385679245, "learning_rate": 6.836645872624402e-05, "loss": 0.0242, "step": 8822 }, { "epoch": 0.9565264527320034, "grad_norm": 0.3735247254371643, "learning_rate": 6.836283185840709e-05, "loss": 0.0423, "step": 8823 }, { "epoch": 0.9566348655680833, "grad_norm": 0.3849892020225525, "learning_rate": 6.835920499057015e-05, "loss": 0.0395, "step": 8824 }, { "epoch": 0.9567432784041631, "grad_norm": 0.18205389380455017, "learning_rate": 6.835557812273322e-05, "loss": 0.0128, "step": 8825 }, { "epoch": 0.9568516912402428, "grad_norm": 0.46065184473991394, "learning_rate": 6.835195125489627e-05, "loss": 0.0324, "step": 8826 }, { "epoch": 0.9569601040763226, "grad_norm": 0.2385675609111786, "learning_rate": 6.834832438705933e-05, "loss": 0.0258, "step": 8827 }, { "epoch": 0.9570685169124025, "grad_norm": 0.13993974030017853, "learning_rate": 6.83446975192224e-05, "loss": 0.0232, "step": 8828 }, { "epoch": 0.9571769297484822, "grad_norm": 0.2672877311706543, "learning_rate": 6.834107065138546e-05, "loss": 0.0379, "step": 8829 }, { "epoch": 0.957285342584562, "grad_norm": 0.11633941531181335, "learning_rate": 6.833744378354853e-05, "loss": 0.0079, "step": 8830 }, { "epoch": 0.9573937554206418, "grad_norm": 0.3041675090789795, "learning_rate": 6.833381691571159e-05, "loss": 0.0316, "step": 8831 }, { "epoch": 0.9575021682567216, "grad_norm": 0.7705584764480591, "learning_rate": 6.833019004787466e-05, "loss": 0.0425, "step": 8832 }, { "epoch": 0.9576105810928014, "grad_norm": 0.26741769909858704, "learning_rate": 6.832656318003773e-05, "loss": 0.0601, "step": 8833 }, { "epoch": 0.9577189939288812, "grad_norm": 0.4709000289440155, "learning_rate": 6.832293631220079e-05, "loss": 0.039, "step": 8834 }, { "epoch": 0.9578274067649609, "grad_norm": 0.27619805932044983, "learning_rate": 6.831930944436386e-05, "loss": 0.0201, "step": 8835 }, { "epoch": 0.9579358196010408, "grad_norm": 0.15129554271697998, "learning_rate": 6.831568257652692e-05, "loss": 0.0094, "step": 8836 }, { "epoch": 0.9580442324371206, "grad_norm": 0.3269188702106476, "learning_rate": 6.831205570868999e-05, "loss": 0.0302, "step": 8837 }, { "epoch": 0.9581526452732003, "grad_norm": 0.4247671961784363, "learning_rate": 6.830842884085304e-05, "loss": 0.0241, "step": 8838 }, { "epoch": 0.9582610581092801, "grad_norm": 0.43245214223861694, "learning_rate": 6.830480197301611e-05, "loss": 0.0387, "step": 8839 }, { "epoch": 0.95836947094536, "grad_norm": 0.13976360857486725, "learning_rate": 6.830117510517917e-05, "loss": 0.0168, "step": 8840 }, { "epoch": 0.9584778837814397, "grad_norm": 0.5681300163269043, "learning_rate": 6.829754823734223e-05, "loss": 0.0407, "step": 8841 }, { "epoch": 0.9585862966175195, "grad_norm": 1.0242489576339722, "learning_rate": 6.82939213695053e-05, "loss": 0.0195, "step": 8842 }, { "epoch": 0.9586947094535994, "grad_norm": 0.4772075414657593, "learning_rate": 6.829029450166836e-05, "loss": 0.0634, "step": 8843 }, { "epoch": 0.9588031222896791, "grad_norm": 0.469806432723999, "learning_rate": 6.828666763383143e-05, "loss": 0.0578, "step": 8844 }, { "epoch": 0.9589115351257589, "grad_norm": 0.16798906028270721, "learning_rate": 6.828304076599449e-05, "loss": 0.0251, "step": 8845 }, { "epoch": 0.9590199479618386, "grad_norm": 0.7459988594055176, "learning_rate": 6.827941389815756e-05, "loss": 0.0208, "step": 8846 }, { "epoch": 0.9591283607979185, "grad_norm": 0.29004305601119995, "learning_rate": 6.827578703032061e-05, "loss": 0.0396, "step": 8847 }, { "epoch": 0.9592367736339983, "grad_norm": 0.34233275055885315, "learning_rate": 6.827216016248368e-05, "loss": 0.0326, "step": 8848 }, { "epoch": 0.959345186470078, "grad_norm": 0.28126364946365356, "learning_rate": 6.826853329464674e-05, "loss": 0.0243, "step": 8849 }, { "epoch": 0.9594535993061578, "grad_norm": 0.7667135000228882, "learning_rate": 6.82649064268098e-05, "loss": 0.0998, "step": 8850 }, { "epoch": 0.9595620121422377, "grad_norm": 0.6364787817001343, "learning_rate": 6.826127955897288e-05, "loss": 0.0573, "step": 8851 }, { "epoch": 0.9596704249783174, "grad_norm": 0.5384294390678406, "learning_rate": 6.825765269113594e-05, "loss": 0.0376, "step": 8852 }, { "epoch": 0.9597788378143972, "grad_norm": 0.37745794653892517, "learning_rate": 6.825402582329901e-05, "loss": 0.0249, "step": 8853 }, { "epoch": 0.959887250650477, "grad_norm": 0.3826559782028198, "learning_rate": 6.825039895546207e-05, "loss": 0.0265, "step": 8854 }, { "epoch": 0.9599956634865568, "grad_norm": 0.467867374420166, "learning_rate": 6.824677208762513e-05, "loss": 0.0446, "step": 8855 }, { "epoch": 0.9601040763226366, "grad_norm": 0.365057110786438, "learning_rate": 6.82431452197882e-05, "loss": 0.0397, "step": 8856 }, { "epoch": 0.9602124891587164, "grad_norm": 0.47634583711624146, "learning_rate": 6.823951835195125e-05, "loss": 0.0642, "step": 8857 }, { "epoch": 0.9603209019947961, "grad_norm": 0.27993056178092957, "learning_rate": 6.823589148411433e-05, "loss": 0.0378, "step": 8858 }, { "epoch": 0.960429314830876, "grad_norm": 0.26258474588394165, "learning_rate": 6.823226461627738e-05, "loss": 0.0231, "step": 8859 }, { "epoch": 0.9605377276669558, "grad_norm": 0.6711715459823608, "learning_rate": 6.822863774844045e-05, "loss": 0.0358, "step": 8860 }, { "epoch": 0.9606461405030355, "grad_norm": 0.45726922154426575, "learning_rate": 6.822501088060351e-05, "loss": 0.0384, "step": 8861 }, { "epoch": 0.9607545533391153, "grad_norm": 0.1803128868341446, "learning_rate": 6.822138401276658e-05, "loss": 0.0235, "step": 8862 }, { "epoch": 0.9608629661751952, "grad_norm": 0.15636228024959564, "learning_rate": 6.821775714492964e-05, "loss": 0.0192, "step": 8863 }, { "epoch": 0.9609713790112749, "grad_norm": 0.20925334095954895, "learning_rate": 6.82141302770927e-05, "loss": 0.0177, "step": 8864 }, { "epoch": 0.9610797918473547, "grad_norm": 0.17036090791225433, "learning_rate": 6.821050340925577e-05, "loss": 0.0147, "step": 8865 }, { "epoch": 0.9611882046834345, "grad_norm": 0.31803643703460693, "learning_rate": 6.820687654141883e-05, "loss": 0.043, "step": 8866 }, { "epoch": 0.9612966175195143, "grad_norm": 0.2839731276035309, "learning_rate": 6.82032496735819e-05, "loss": 0.02, "step": 8867 }, { "epoch": 0.9614050303555941, "grad_norm": 0.598720908164978, "learning_rate": 6.819962280574497e-05, "loss": 0.0914, "step": 8868 }, { "epoch": 0.9615134431916739, "grad_norm": 0.4339665472507477, "learning_rate": 6.819599593790802e-05, "loss": 0.0423, "step": 8869 }, { "epoch": 0.9616218560277537, "grad_norm": 0.16725373268127441, "learning_rate": 6.81923690700711e-05, "loss": 0.0278, "step": 8870 }, { "epoch": 0.9617302688638335, "grad_norm": 0.33955278992652893, "learning_rate": 6.818874220223415e-05, "loss": 0.0201, "step": 8871 }, { "epoch": 0.9618386816999133, "grad_norm": 0.20366761088371277, "learning_rate": 6.818511533439722e-05, "loss": 0.0145, "step": 8872 }, { "epoch": 0.961947094535993, "grad_norm": 0.33633866906166077, "learning_rate": 6.818148846656028e-05, "loss": 0.0759, "step": 8873 }, { "epoch": 0.9620555073720729, "grad_norm": 0.553697407245636, "learning_rate": 6.817786159872335e-05, "loss": 0.065, "step": 8874 }, { "epoch": 0.9621639202081527, "grad_norm": 0.7440394759178162, "learning_rate": 6.817423473088641e-05, "loss": 0.0695, "step": 8875 }, { "epoch": 0.9622723330442324, "grad_norm": 0.16036078333854675, "learning_rate": 6.817060786304948e-05, "loss": 0.0167, "step": 8876 }, { "epoch": 0.9623807458803122, "grad_norm": 0.48063698410987854, "learning_rate": 6.816698099521254e-05, "loss": 0.0293, "step": 8877 }, { "epoch": 0.9624891587163921, "grad_norm": 0.16989301145076752, "learning_rate": 6.81633541273756e-05, "loss": 0.019, "step": 8878 }, { "epoch": 0.9625975715524718, "grad_norm": 0.31655600666999817, "learning_rate": 6.815972725953867e-05, "loss": 0.0652, "step": 8879 }, { "epoch": 0.9627059843885516, "grad_norm": 0.23288099467754364, "learning_rate": 6.815610039170172e-05, "loss": 0.0222, "step": 8880 }, { "epoch": 0.9628143972246314, "grad_norm": 0.4431290626525879, "learning_rate": 6.81524735238648e-05, "loss": 0.0545, "step": 8881 }, { "epoch": 0.9629228100607112, "grad_norm": 0.3401435911655426, "learning_rate": 6.814884665602785e-05, "loss": 0.0597, "step": 8882 }, { "epoch": 0.963031222896791, "grad_norm": 0.39198607206344604, "learning_rate": 6.814521978819092e-05, "loss": 0.053, "step": 8883 }, { "epoch": 0.9631396357328708, "grad_norm": 0.2149176001548767, "learning_rate": 6.814159292035398e-05, "loss": 0.0279, "step": 8884 }, { "epoch": 0.9632480485689505, "grad_norm": 0.15078291296958923, "learning_rate": 6.813796605251705e-05, "loss": 0.0182, "step": 8885 }, { "epoch": 0.9633564614050304, "grad_norm": 0.28916701674461365, "learning_rate": 6.813433918468012e-05, "loss": 0.0454, "step": 8886 }, { "epoch": 0.9634648742411102, "grad_norm": 0.13568627834320068, "learning_rate": 6.813071231684318e-05, "loss": 0.0162, "step": 8887 }, { "epoch": 0.9635732870771899, "grad_norm": 0.174917072057724, "learning_rate": 6.812708544900625e-05, "loss": 0.0249, "step": 8888 }, { "epoch": 0.9636816999132697, "grad_norm": 0.18312138319015503, "learning_rate": 6.81234585811693e-05, "loss": 0.0201, "step": 8889 }, { "epoch": 0.9637901127493496, "grad_norm": 0.2935073971748352, "learning_rate": 6.811983171333238e-05, "loss": 0.0244, "step": 8890 }, { "epoch": 0.9638985255854293, "grad_norm": 0.18946701288223267, "learning_rate": 6.811620484549543e-05, "loss": 0.0214, "step": 8891 }, { "epoch": 0.9640069384215091, "grad_norm": 0.2878848910331726, "learning_rate": 6.811257797765849e-05, "loss": 0.0599, "step": 8892 }, { "epoch": 0.9641153512575888, "grad_norm": 0.48078784346580505, "learning_rate": 6.810895110982156e-05, "loss": 0.0304, "step": 8893 }, { "epoch": 0.9642237640936687, "grad_norm": 0.29658424854278564, "learning_rate": 6.810532424198462e-05, "loss": 0.024, "step": 8894 }, { "epoch": 0.9643321769297485, "grad_norm": 0.23439082503318787, "learning_rate": 6.810169737414769e-05, "loss": 0.0231, "step": 8895 }, { "epoch": 0.9644405897658282, "grad_norm": 0.3447830080986023, "learning_rate": 6.809807050631075e-05, "loss": 0.0479, "step": 8896 }, { "epoch": 0.964549002601908, "grad_norm": 0.21775726974010468, "learning_rate": 6.809444363847382e-05, "loss": 0.0425, "step": 8897 }, { "epoch": 0.9646574154379879, "grad_norm": 0.2672657370567322, "learning_rate": 6.809081677063688e-05, "loss": 0.0308, "step": 8898 }, { "epoch": 0.9647658282740676, "grad_norm": 0.4060352146625519, "learning_rate": 6.808718990279995e-05, "loss": 0.0587, "step": 8899 }, { "epoch": 0.9648742411101474, "grad_norm": 0.5489909648895264, "learning_rate": 6.8083563034963e-05, "loss": 0.0553, "step": 8900 }, { "epoch": 0.9649826539462273, "grad_norm": 0.3454727232456207, "learning_rate": 6.807993616712606e-05, "loss": 0.0608, "step": 8901 }, { "epoch": 0.965091066782307, "grad_norm": 0.2988920211791992, "learning_rate": 6.807630929928915e-05, "loss": 0.0328, "step": 8902 }, { "epoch": 0.9651994796183868, "grad_norm": 0.2838676869869232, "learning_rate": 6.80726824314522e-05, "loss": 0.0295, "step": 8903 }, { "epoch": 0.9653078924544666, "grad_norm": 0.09431084245443344, "learning_rate": 6.806905556361527e-05, "loss": 0.0073, "step": 8904 }, { "epoch": 0.9654163052905465, "grad_norm": 0.19344276189804077, "learning_rate": 6.806542869577833e-05, "loss": 0.0266, "step": 8905 }, { "epoch": 0.9655247181266262, "grad_norm": 0.48405173420906067, "learning_rate": 6.806180182794139e-05, "loss": 0.0606, "step": 8906 }, { "epoch": 0.965633130962706, "grad_norm": 0.1367749720811844, "learning_rate": 6.805817496010446e-05, "loss": 0.0255, "step": 8907 }, { "epoch": 0.9657415437987857, "grad_norm": 0.31031563878059387, "learning_rate": 6.805454809226752e-05, "loss": 0.0338, "step": 8908 }, { "epoch": 0.9658499566348656, "grad_norm": 0.42694830894470215, "learning_rate": 6.805092122443059e-05, "loss": 0.0599, "step": 8909 }, { "epoch": 0.9659583694709454, "grad_norm": 0.5256198644638062, "learning_rate": 6.804729435659365e-05, "loss": 0.0937, "step": 8910 }, { "epoch": 0.9660667823070251, "grad_norm": 0.46807488799095154, "learning_rate": 6.804366748875672e-05, "loss": 0.0402, "step": 8911 }, { "epoch": 0.9661751951431049, "grad_norm": 0.16866543889045715, "learning_rate": 6.804004062091977e-05, "loss": 0.0234, "step": 8912 }, { "epoch": 0.9662836079791848, "grad_norm": 0.08131656795740128, "learning_rate": 6.803641375308284e-05, "loss": 0.0116, "step": 8913 }, { "epoch": 0.9663920208152645, "grad_norm": 0.49755001068115234, "learning_rate": 6.80327868852459e-05, "loss": 0.0256, "step": 8914 }, { "epoch": 0.9665004336513443, "grad_norm": 0.2700197398662567, "learning_rate": 6.802916001740896e-05, "loss": 0.0259, "step": 8915 }, { "epoch": 0.9666088464874241, "grad_norm": 0.335829496383667, "learning_rate": 6.802553314957203e-05, "loss": 0.0373, "step": 8916 }, { "epoch": 0.9667172593235039, "grad_norm": 0.40259411931037903, "learning_rate": 6.802190628173509e-05, "loss": 0.0484, "step": 8917 }, { "epoch": 0.9668256721595837, "grad_norm": 0.35808947682380676, "learning_rate": 6.801827941389816e-05, "loss": 0.0376, "step": 8918 }, { "epoch": 0.9669340849956635, "grad_norm": 0.4362283945083618, "learning_rate": 6.801465254606122e-05, "loss": 0.0378, "step": 8919 }, { "epoch": 0.9670424978317432, "grad_norm": 0.4072568714618683, "learning_rate": 6.80110256782243e-05, "loss": 0.0322, "step": 8920 }, { "epoch": 0.9671509106678231, "grad_norm": 0.2822800278663635, "learning_rate": 6.800739881038736e-05, "loss": 0.0636, "step": 8921 }, { "epoch": 0.9672593235039029, "grad_norm": 0.3306816518306732, "learning_rate": 6.800377194255041e-05, "loss": 0.0448, "step": 8922 }, { "epoch": 0.9673677363399826, "grad_norm": 0.3171119689941406, "learning_rate": 6.800014507471349e-05, "loss": 0.0481, "step": 8923 }, { "epoch": 0.9674761491760624, "grad_norm": 0.21855153143405914, "learning_rate": 6.799651820687654e-05, "loss": 0.0371, "step": 8924 }, { "epoch": 0.9675845620121423, "grad_norm": 0.30785638093948364, "learning_rate": 6.799289133903961e-05, "loss": 0.0344, "step": 8925 }, { "epoch": 0.967692974848222, "grad_norm": 0.3371695280075073, "learning_rate": 6.798926447120267e-05, "loss": 0.0514, "step": 8926 }, { "epoch": 0.9678013876843018, "grad_norm": 0.5933191180229187, "learning_rate": 6.798563760336574e-05, "loss": 0.0691, "step": 8927 }, { "epoch": 0.9679098005203816, "grad_norm": 0.21919956803321838, "learning_rate": 6.79820107355288e-05, "loss": 0.0182, "step": 8928 }, { "epoch": 0.9680182133564614, "grad_norm": 0.24352486431598663, "learning_rate": 6.797838386769187e-05, "loss": 0.0302, "step": 8929 }, { "epoch": 0.9681266261925412, "grad_norm": 0.06452076137065887, "learning_rate": 6.797475699985493e-05, "loss": 0.0052, "step": 8930 }, { "epoch": 0.968235039028621, "grad_norm": 0.2924904525279999, "learning_rate": 6.797113013201799e-05, "loss": 0.0424, "step": 8931 }, { "epoch": 0.9683434518647008, "grad_norm": 0.3622613847255707, "learning_rate": 6.796750326418106e-05, "loss": 0.0759, "step": 8932 }, { "epoch": 0.9684518647007806, "grad_norm": 0.2874833047389984, "learning_rate": 6.796387639634411e-05, "loss": 0.0571, "step": 8933 }, { "epoch": 0.9685602775368604, "grad_norm": 0.2692326009273529, "learning_rate": 6.796024952850718e-05, "loss": 0.0339, "step": 8934 }, { "epoch": 0.9686686903729401, "grad_norm": 0.5151327848434448, "learning_rate": 6.795662266067024e-05, "loss": 0.0539, "step": 8935 }, { "epoch": 0.96877710320902, "grad_norm": 0.35934340953826904, "learning_rate": 6.795299579283331e-05, "loss": 0.0561, "step": 8936 }, { "epoch": 0.9688855160450998, "grad_norm": 0.2274145632982254, "learning_rate": 6.794936892499638e-05, "loss": 0.0113, "step": 8937 }, { "epoch": 0.9689939288811795, "grad_norm": 0.153188556432724, "learning_rate": 6.794574205715944e-05, "loss": 0.0111, "step": 8938 }, { "epoch": 0.9691023417172593, "grad_norm": 0.3996763229370117, "learning_rate": 6.794211518932251e-05, "loss": 0.0505, "step": 8939 }, { "epoch": 0.9692107545533392, "grad_norm": 0.45608457922935486, "learning_rate": 6.793848832148557e-05, "loss": 0.0432, "step": 8940 }, { "epoch": 0.9693191673894189, "grad_norm": 0.19157087802886963, "learning_rate": 6.793486145364864e-05, "loss": 0.0387, "step": 8941 }, { "epoch": 0.9694275802254987, "grad_norm": 0.13143213093280792, "learning_rate": 6.79312345858117e-05, "loss": 0.0181, "step": 8942 }, { "epoch": 0.9695359930615784, "grad_norm": 0.4427867531776428, "learning_rate": 6.792760771797477e-05, "loss": 0.0368, "step": 8943 }, { "epoch": 0.9696444058976583, "grad_norm": 0.1356896013021469, "learning_rate": 6.792398085013783e-05, "loss": 0.0234, "step": 8944 }, { "epoch": 0.9697528187337381, "grad_norm": 0.1785522699356079, "learning_rate": 6.792035398230088e-05, "loss": 0.023, "step": 8945 }, { "epoch": 0.9698612315698178, "grad_norm": 0.43630948662757874, "learning_rate": 6.791672711446395e-05, "loss": 0.0578, "step": 8946 }, { "epoch": 0.9699696444058976, "grad_norm": 0.47677403688430786, "learning_rate": 6.791310024662701e-05, "loss": 0.0422, "step": 8947 }, { "epoch": 0.9700780572419775, "grad_norm": 0.4593842923641205, "learning_rate": 6.790947337879008e-05, "loss": 0.0421, "step": 8948 }, { "epoch": 0.9701864700780573, "grad_norm": 0.2210828959941864, "learning_rate": 6.790584651095314e-05, "loss": 0.0233, "step": 8949 }, { "epoch": 0.970294882914137, "grad_norm": 0.6524441242218018, "learning_rate": 6.790221964311621e-05, "loss": 0.0405, "step": 8950 }, { "epoch": 0.9704032957502168, "grad_norm": 0.5765474438667297, "learning_rate": 6.789859277527927e-05, "loss": 0.0319, "step": 8951 }, { "epoch": 0.9705117085862967, "grad_norm": 0.3062908351421356, "learning_rate": 6.789496590744234e-05, "loss": 0.0607, "step": 8952 }, { "epoch": 0.9706201214223764, "grad_norm": 0.3965120315551758, "learning_rate": 6.78913390396054e-05, "loss": 0.0658, "step": 8953 }, { "epoch": 0.9707285342584562, "grad_norm": 0.31562426686286926, "learning_rate": 6.788771217176847e-05, "loss": 0.0504, "step": 8954 }, { "epoch": 0.9708369470945359, "grad_norm": 0.43304184079170227, "learning_rate": 6.788408530393154e-05, "loss": 0.0405, "step": 8955 }, { "epoch": 0.9709453599306158, "grad_norm": 0.29396501183509827, "learning_rate": 6.78804584360946e-05, "loss": 0.0258, "step": 8956 }, { "epoch": 0.9710537727666956, "grad_norm": 0.5361403226852417, "learning_rate": 6.787683156825767e-05, "loss": 0.0303, "step": 8957 }, { "epoch": 0.9711621856027753, "grad_norm": 0.2333678901195526, "learning_rate": 6.787320470042072e-05, "loss": 0.0231, "step": 8958 }, { "epoch": 0.9712705984388552, "grad_norm": 0.21940062940120697, "learning_rate": 6.786957783258378e-05, "loss": 0.0236, "step": 8959 }, { "epoch": 0.971379011274935, "grad_norm": 0.3690195381641388, "learning_rate": 6.786595096474685e-05, "loss": 0.0617, "step": 8960 }, { "epoch": 0.9714874241110147, "grad_norm": 0.1662251055240631, "learning_rate": 6.786232409690991e-05, "loss": 0.0181, "step": 8961 }, { "epoch": 0.9715958369470945, "grad_norm": 0.20502957701683044, "learning_rate": 6.785869722907298e-05, "loss": 0.0266, "step": 8962 }, { "epoch": 0.9717042497831744, "grad_norm": 0.32003724575042725, "learning_rate": 6.785507036123604e-05, "loss": 0.0292, "step": 8963 }, { "epoch": 0.9718126626192541, "grad_norm": 0.22035226225852966, "learning_rate": 6.785144349339911e-05, "loss": 0.0135, "step": 8964 }, { "epoch": 0.9719210754553339, "grad_norm": 0.1960279643535614, "learning_rate": 6.784781662556216e-05, "loss": 0.0142, "step": 8965 }, { "epoch": 0.9720294882914137, "grad_norm": 0.9241666793823242, "learning_rate": 6.784418975772524e-05, "loss": 0.0784, "step": 8966 }, { "epoch": 0.9721379011274935, "grad_norm": 0.763705849647522, "learning_rate": 6.784056288988829e-05, "loss": 0.0776, "step": 8967 }, { "epoch": 0.9722463139635733, "grad_norm": 0.11671851575374603, "learning_rate": 6.783693602205135e-05, "loss": 0.018, "step": 8968 }, { "epoch": 0.9723547267996531, "grad_norm": 0.09910344332456589, "learning_rate": 6.783330915421442e-05, "loss": 0.0094, "step": 8969 }, { "epoch": 0.9724631396357328, "grad_norm": 0.6380448341369629, "learning_rate": 6.782968228637748e-05, "loss": 0.055, "step": 8970 }, { "epoch": 0.9725715524718127, "grad_norm": 0.25762635469436646, "learning_rate": 6.782605541854055e-05, "loss": 0.0469, "step": 8971 }, { "epoch": 0.9726799653078925, "grad_norm": 0.36086323857307434, "learning_rate": 6.782242855070362e-05, "loss": 0.0599, "step": 8972 }, { "epoch": 0.9727883781439722, "grad_norm": 0.26593777537345886, "learning_rate": 6.781880168286668e-05, "loss": 0.0539, "step": 8973 }, { "epoch": 0.972896790980052, "grad_norm": 0.4858396351337433, "learning_rate": 6.781517481502975e-05, "loss": 0.0564, "step": 8974 }, { "epoch": 0.9730052038161319, "grad_norm": 0.35944128036499023, "learning_rate": 6.78115479471928e-05, "loss": 0.0553, "step": 8975 }, { "epoch": 0.9731136166522116, "grad_norm": 0.41235464811325073, "learning_rate": 6.780792107935588e-05, "loss": 0.0556, "step": 8976 }, { "epoch": 0.9732220294882914, "grad_norm": 0.569317102432251, "learning_rate": 6.780429421151893e-05, "loss": 0.0223, "step": 8977 }, { "epoch": 0.9733304423243712, "grad_norm": 0.28881365060806274, "learning_rate": 6.7800667343682e-05, "loss": 0.0186, "step": 8978 }, { "epoch": 0.973438855160451, "grad_norm": 0.12710712850093842, "learning_rate": 6.779704047584506e-05, "loss": 0.0113, "step": 8979 }, { "epoch": 0.9735472679965308, "grad_norm": 0.31141042709350586, "learning_rate": 6.779341360800813e-05, "loss": 0.026, "step": 8980 }, { "epoch": 0.9736556808326106, "grad_norm": 0.25978630781173706, "learning_rate": 6.778978674017119e-05, "loss": 0.0344, "step": 8981 }, { "epoch": 0.9737640936686903, "grad_norm": 0.17087124288082123, "learning_rate": 6.778615987233425e-05, "loss": 0.015, "step": 8982 }, { "epoch": 0.9738725065047702, "grad_norm": 0.24121882021427155, "learning_rate": 6.778253300449732e-05, "loss": 0.0377, "step": 8983 }, { "epoch": 0.97398091934085, "grad_norm": 0.16106177866458893, "learning_rate": 6.777890613666038e-05, "loss": 0.0268, "step": 8984 }, { "epoch": 0.9740893321769297, "grad_norm": 0.6007938981056213, "learning_rate": 6.777527926882345e-05, "loss": 0.0715, "step": 8985 }, { "epoch": 0.9741977450130095, "grad_norm": 0.28747138381004333, "learning_rate": 6.77716524009865e-05, "loss": 0.0521, "step": 8986 }, { "epoch": 0.9743061578490894, "grad_norm": 0.17775921523571014, "learning_rate": 6.776802553314958e-05, "loss": 0.0176, "step": 8987 }, { "epoch": 0.9744145706851691, "grad_norm": 0.1555028110742569, "learning_rate": 6.776439866531263e-05, "loss": 0.0134, "step": 8988 }, { "epoch": 0.9745229835212489, "grad_norm": 0.3260505497455597, "learning_rate": 6.77607717974757e-05, "loss": 0.0406, "step": 8989 }, { "epoch": 0.9746313963573288, "grad_norm": 0.6719079613685608, "learning_rate": 6.775714492963877e-05, "loss": 0.0219, "step": 8990 }, { "epoch": 0.9747398091934085, "grad_norm": 0.3452710509300232, "learning_rate": 6.775351806180183e-05, "loss": 0.0403, "step": 8991 }, { "epoch": 0.9748482220294883, "grad_norm": 0.22616903483867645, "learning_rate": 6.77498911939649e-05, "loss": 0.0357, "step": 8992 }, { "epoch": 0.974956634865568, "grad_norm": 0.3847789168357849, "learning_rate": 6.774626432612796e-05, "loss": 0.025, "step": 8993 }, { "epoch": 0.9750650477016479, "grad_norm": 0.42584747076034546, "learning_rate": 6.774263745829103e-05, "loss": 0.0499, "step": 8994 }, { "epoch": 0.9751734605377277, "grad_norm": 0.5436000227928162, "learning_rate": 6.773901059045409e-05, "loss": 0.0424, "step": 8995 }, { "epoch": 0.9752818733738075, "grad_norm": 0.2597070038318634, "learning_rate": 6.773538372261715e-05, "loss": 0.0317, "step": 8996 }, { "epoch": 0.9753902862098872, "grad_norm": 0.24157388508319855, "learning_rate": 6.773175685478022e-05, "loss": 0.0358, "step": 8997 }, { "epoch": 0.9754986990459671, "grad_norm": 0.595028281211853, "learning_rate": 6.772812998694327e-05, "loss": 0.0609, "step": 8998 }, { "epoch": 0.9756071118820469, "grad_norm": 0.3186508119106293, "learning_rate": 6.772450311910634e-05, "loss": 0.0201, "step": 8999 }, { "epoch": 0.9757155247181266, "grad_norm": 0.4013144075870514, "learning_rate": 6.77208762512694e-05, "loss": 0.0737, "step": 9000 }, { "epoch": 0.9758239375542064, "grad_norm": 0.30863308906555176, "learning_rate": 6.771724938343247e-05, "loss": 0.0333, "step": 9001 }, { "epoch": 0.9759323503902863, "grad_norm": 0.43817296624183655, "learning_rate": 6.771362251559553e-05, "loss": 0.0622, "step": 9002 }, { "epoch": 0.976040763226366, "grad_norm": 0.20429213345050812, "learning_rate": 6.77099956477586e-05, "loss": 0.0314, "step": 9003 }, { "epoch": 0.9761491760624458, "grad_norm": 0.2584207355976105, "learning_rate": 6.770636877992166e-05, "loss": 0.0381, "step": 9004 }, { "epoch": 0.9762575888985255, "grad_norm": 0.5007676482200623, "learning_rate": 6.770274191208472e-05, "loss": 0.0656, "step": 9005 }, { "epoch": 0.9763660017346054, "grad_norm": 0.25188401341438293, "learning_rate": 6.76991150442478e-05, "loss": 0.0351, "step": 9006 }, { "epoch": 0.9764744145706852, "grad_norm": 0.3284643888473511, "learning_rate": 6.769548817641086e-05, "loss": 0.0473, "step": 9007 }, { "epoch": 0.9765828274067649, "grad_norm": 0.37850892543792725, "learning_rate": 6.769186130857393e-05, "loss": 0.0466, "step": 9008 }, { "epoch": 0.9766912402428447, "grad_norm": 0.4504730999469757, "learning_rate": 6.768823444073699e-05, "loss": 0.0328, "step": 9009 }, { "epoch": 0.9767996530789246, "grad_norm": 0.16605974733829498, "learning_rate": 6.768460757290006e-05, "loss": 0.0273, "step": 9010 }, { "epoch": 0.9769080659150043, "grad_norm": 0.27020493149757385, "learning_rate": 6.768098070506311e-05, "loss": 0.0279, "step": 9011 }, { "epoch": 0.9770164787510841, "grad_norm": 0.4246983826160431, "learning_rate": 6.767735383722617e-05, "loss": 0.0381, "step": 9012 }, { "epoch": 0.9771248915871639, "grad_norm": 0.28377583622932434, "learning_rate": 6.767372696938924e-05, "loss": 0.0238, "step": 9013 }, { "epoch": 0.9772333044232437, "grad_norm": 0.5694242119789124, "learning_rate": 6.76701001015523e-05, "loss": 0.0376, "step": 9014 }, { "epoch": 0.9773417172593235, "grad_norm": 0.1952689290046692, "learning_rate": 6.766647323371537e-05, "loss": 0.0157, "step": 9015 }, { "epoch": 0.9774501300954033, "grad_norm": 1.3657580614089966, "learning_rate": 6.766284636587843e-05, "loss": 0.0275, "step": 9016 }, { "epoch": 0.977558542931483, "grad_norm": 0.7350555062294006, "learning_rate": 6.76592194980415e-05, "loss": 0.0288, "step": 9017 }, { "epoch": 0.9776669557675629, "grad_norm": 0.6904537677764893, "learning_rate": 6.765559263020456e-05, "loss": 0.031, "step": 9018 }, { "epoch": 0.9777753686036427, "grad_norm": 0.33351239562034607, "learning_rate": 6.765196576236761e-05, "loss": 0.0464, "step": 9019 }, { "epoch": 0.9778837814397224, "grad_norm": 0.19423964619636536, "learning_rate": 6.764833889453068e-05, "loss": 0.0214, "step": 9020 }, { "epoch": 0.9779921942758023, "grad_norm": 0.35758739709854126, "learning_rate": 6.764471202669374e-05, "loss": 0.0487, "step": 9021 }, { "epoch": 0.9781006071118821, "grad_norm": 0.36489158868789673, "learning_rate": 6.764108515885681e-05, "loss": 0.053, "step": 9022 }, { "epoch": 0.9782090199479618, "grad_norm": 0.4305126667022705, "learning_rate": 6.763745829101988e-05, "loss": 0.0356, "step": 9023 }, { "epoch": 0.9783174327840416, "grad_norm": 2.4768552780151367, "learning_rate": 6.763383142318295e-05, "loss": 0.0235, "step": 9024 }, { "epoch": 0.9784258456201215, "grad_norm": 0.680117666721344, "learning_rate": 6.763020455534601e-05, "loss": 0.0571, "step": 9025 }, { "epoch": 0.9785342584562012, "grad_norm": 0.22769315540790558, "learning_rate": 6.762657768750907e-05, "loss": 0.0249, "step": 9026 }, { "epoch": 0.978642671292281, "grad_norm": 0.5473856925964355, "learning_rate": 6.762295081967214e-05, "loss": 0.0397, "step": 9027 }, { "epoch": 0.9787510841283608, "grad_norm": 0.6261412501335144, "learning_rate": 6.76193239518352e-05, "loss": 0.0787, "step": 9028 }, { "epoch": 0.9788594969644406, "grad_norm": 0.3671247065067291, "learning_rate": 6.761569708399827e-05, "loss": 0.0257, "step": 9029 }, { "epoch": 0.9789679098005204, "grad_norm": 0.1108064278960228, "learning_rate": 6.761207021616132e-05, "loss": 0.0063, "step": 9030 }, { "epoch": 0.9790763226366002, "grad_norm": 0.20601411163806915, "learning_rate": 6.76084433483244e-05, "loss": 0.0157, "step": 9031 }, { "epoch": 0.9791847354726799, "grad_norm": 0.3885669410228729, "learning_rate": 6.760481648048745e-05, "loss": 0.0372, "step": 9032 }, { "epoch": 0.9792931483087598, "grad_norm": 0.27371954917907715, "learning_rate": 6.760118961265052e-05, "loss": 0.015, "step": 9033 }, { "epoch": 0.9794015611448396, "grad_norm": 0.5347564220428467, "learning_rate": 6.759756274481358e-05, "loss": 0.0318, "step": 9034 }, { "epoch": 0.9795099739809193, "grad_norm": 0.27162355184555054, "learning_rate": 6.759393587697664e-05, "loss": 0.0176, "step": 9035 }, { "epoch": 0.9796183868169991, "grad_norm": 0.2152329832315445, "learning_rate": 6.759030900913971e-05, "loss": 0.0193, "step": 9036 }, { "epoch": 0.979726799653079, "grad_norm": 0.3997602164745331, "learning_rate": 6.758668214130277e-05, "loss": 0.0386, "step": 9037 }, { "epoch": 0.9798352124891587, "grad_norm": 0.6495633721351624, "learning_rate": 6.758305527346584e-05, "loss": 0.0348, "step": 9038 }, { "epoch": 0.9799436253252385, "grad_norm": 0.7648553848266602, "learning_rate": 6.75794284056289e-05, "loss": 0.0598, "step": 9039 }, { "epoch": 0.9800520381613183, "grad_norm": 0.5773056745529175, "learning_rate": 6.757580153779197e-05, "loss": 0.0795, "step": 9040 }, { "epoch": 0.9801604509973981, "grad_norm": 0.18385270237922668, "learning_rate": 6.757217466995504e-05, "loss": 0.0305, "step": 9041 }, { "epoch": 0.9802688638334779, "grad_norm": 0.2150455117225647, "learning_rate": 6.75685478021181e-05, "loss": 0.018, "step": 9042 }, { "epoch": 0.9803772766695577, "grad_norm": 0.3992070257663727, "learning_rate": 6.756492093428116e-05, "loss": 0.049, "step": 9043 }, { "epoch": 0.9804856895056374, "grad_norm": 0.2753206491470337, "learning_rate": 6.756129406644422e-05, "loss": 0.0074, "step": 9044 }, { "epoch": 0.9805941023417173, "grad_norm": 0.6686962842941284, "learning_rate": 6.755766719860729e-05, "loss": 0.0815, "step": 9045 }, { "epoch": 0.980702515177797, "grad_norm": 0.4654262661933899, "learning_rate": 6.755404033077035e-05, "loss": 0.0516, "step": 9046 }, { "epoch": 0.9808109280138768, "grad_norm": 0.37764835357666016, "learning_rate": 6.755041346293342e-05, "loss": 0.039, "step": 9047 }, { "epoch": 0.9809193408499567, "grad_norm": 0.7906225919723511, "learning_rate": 6.754678659509648e-05, "loss": 0.0959, "step": 9048 }, { "epoch": 0.9810277536860365, "grad_norm": 0.4254514276981354, "learning_rate": 6.754315972725954e-05, "loss": 0.0549, "step": 9049 }, { "epoch": 0.9811361665221162, "grad_norm": 0.09338802844285965, "learning_rate": 6.753953285942261e-05, "loss": 0.0062, "step": 9050 }, { "epoch": 0.981244579358196, "grad_norm": 0.24337461590766907, "learning_rate": 6.753590599158566e-05, "loss": 0.0249, "step": 9051 }, { "epoch": 0.9813529921942759, "grad_norm": 0.4373590648174286, "learning_rate": 6.753227912374874e-05, "loss": 0.0417, "step": 9052 }, { "epoch": 0.9814614050303556, "grad_norm": 0.12551884353160858, "learning_rate": 6.752865225591179e-05, "loss": 0.0136, "step": 9053 }, { "epoch": 0.9815698178664354, "grad_norm": 0.1318095624446869, "learning_rate": 6.752502538807486e-05, "loss": 0.0117, "step": 9054 }, { "epoch": 0.9816782307025151, "grad_norm": 0.15394365787506104, "learning_rate": 6.752139852023792e-05, "loss": 0.0186, "step": 9055 }, { "epoch": 0.981786643538595, "grad_norm": 0.33228614926338196, "learning_rate": 6.751777165240099e-05, "loss": 0.0213, "step": 9056 }, { "epoch": 0.9818950563746748, "grad_norm": 0.363363116979599, "learning_rate": 6.751414478456405e-05, "loss": 0.043, "step": 9057 }, { "epoch": 0.9820034692107545, "grad_norm": 0.3034701347351074, "learning_rate": 6.751051791672712e-05, "loss": 0.0398, "step": 9058 }, { "epoch": 0.9821118820468343, "grad_norm": 0.22578437626361847, "learning_rate": 6.750689104889019e-05, "loss": 0.0193, "step": 9059 }, { "epoch": 0.9822202948829142, "grad_norm": 0.1268123835325241, "learning_rate": 6.750326418105325e-05, "loss": 0.0143, "step": 9060 }, { "epoch": 0.9823287077189939, "grad_norm": 0.16026748716831207, "learning_rate": 6.749963731321632e-05, "loss": 0.0139, "step": 9061 }, { "epoch": 0.9824371205550737, "grad_norm": 0.2456594854593277, "learning_rate": 6.749601044537938e-05, "loss": 0.037, "step": 9062 }, { "epoch": 0.9825455333911535, "grad_norm": 0.18065069615840912, "learning_rate": 6.749238357754243e-05, "loss": 0.025, "step": 9063 }, { "epoch": 0.9826539462272333, "grad_norm": 0.11125409603118896, "learning_rate": 6.74887567097055e-05, "loss": 0.0127, "step": 9064 }, { "epoch": 0.9827623590633131, "grad_norm": 0.33551880717277527, "learning_rate": 6.748512984186856e-05, "loss": 0.0484, "step": 9065 }, { "epoch": 0.9828707718993929, "grad_norm": 0.2628074884414673, "learning_rate": 6.748150297403163e-05, "loss": 0.0351, "step": 9066 }, { "epoch": 0.9829791847354726, "grad_norm": 0.183634951710701, "learning_rate": 6.747787610619469e-05, "loss": 0.0219, "step": 9067 }, { "epoch": 0.9830875975715525, "grad_norm": 0.49776771664619446, "learning_rate": 6.747424923835776e-05, "loss": 0.0694, "step": 9068 }, { "epoch": 0.9831960104076323, "grad_norm": 0.1013856828212738, "learning_rate": 6.747062237052082e-05, "loss": 0.009, "step": 9069 }, { "epoch": 0.983304423243712, "grad_norm": 0.5439435243606567, "learning_rate": 6.746699550268389e-05, "loss": 0.077, "step": 9070 }, { "epoch": 0.9834128360797918, "grad_norm": 0.3082907795906067, "learning_rate": 6.746336863484695e-05, "loss": 0.0182, "step": 9071 }, { "epoch": 0.9835212489158717, "grad_norm": 0.3297377824783325, "learning_rate": 6.745974176701e-05, "loss": 0.0344, "step": 9072 }, { "epoch": 0.9836296617519514, "grad_norm": 0.3428597152233124, "learning_rate": 6.745611489917307e-05, "loss": 0.0511, "step": 9073 }, { "epoch": 0.9837380745880312, "grad_norm": 0.5203154683113098, "learning_rate": 6.745248803133613e-05, "loss": 0.0835, "step": 9074 }, { "epoch": 0.983846487424111, "grad_norm": 0.3196190297603607, "learning_rate": 6.744886116349922e-05, "loss": 0.0233, "step": 9075 }, { "epoch": 0.9839549002601908, "grad_norm": 0.12204278260469437, "learning_rate": 6.744523429566227e-05, "loss": 0.0112, "step": 9076 }, { "epoch": 0.9840633130962706, "grad_norm": 0.5122092962265015, "learning_rate": 6.744160742782533e-05, "loss": 0.0629, "step": 9077 }, { "epoch": 0.9841717259323504, "grad_norm": 0.4327036142349243, "learning_rate": 6.74379805599884e-05, "loss": 0.0493, "step": 9078 }, { "epoch": 0.9842801387684302, "grad_norm": 0.18489108979701996, "learning_rate": 6.743435369215146e-05, "loss": 0.0113, "step": 9079 }, { "epoch": 0.98438855160451, "grad_norm": 0.3878587484359741, "learning_rate": 6.743072682431453e-05, "loss": 0.0435, "step": 9080 }, { "epoch": 0.9844969644405898, "grad_norm": 0.537608802318573, "learning_rate": 6.742709995647759e-05, "loss": 0.0638, "step": 9081 }, { "epoch": 0.9846053772766695, "grad_norm": 0.5289430618286133, "learning_rate": 6.742347308864066e-05, "loss": 0.0432, "step": 9082 }, { "epoch": 0.9847137901127494, "grad_norm": 0.3289315104484558, "learning_rate": 6.741984622080372e-05, "loss": 0.038, "step": 9083 }, { "epoch": 0.9848222029488292, "grad_norm": 0.22721318900585175, "learning_rate": 6.741621935296679e-05, "loss": 0.0192, "step": 9084 }, { "epoch": 0.9849306157849089, "grad_norm": 0.4751814305782318, "learning_rate": 6.741259248512984e-05, "loss": 0.0731, "step": 9085 }, { "epoch": 0.9850390286209887, "grad_norm": 0.2259502112865448, "learning_rate": 6.74089656172929e-05, "loss": 0.0622, "step": 9086 }, { "epoch": 0.9851474414570686, "grad_norm": 0.11232088506221771, "learning_rate": 6.740533874945597e-05, "loss": 0.0131, "step": 9087 }, { "epoch": 0.9852558542931483, "grad_norm": 0.26376262307167053, "learning_rate": 6.740171188161903e-05, "loss": 0.0272, "step": 9088 }, { "epoch": 0.9853642671292281, "grad_norm": 0.254535973072052, "learning_rate": 6.73980850137821e-05, "loss": 0.0154, "step": 9089 }, { "epoch": 0.9854726799653079, "grad_norm": 0.42004185914993286, "learning_rate": 6.739445814594516e-05, "loss": 0.0471, "step": 9090 }, { "epoch": 0.9855810928013877, "grad_norm": 0.36903008818626404, "learning_rate": 6.739083127810823e-05, "loss": 0.0328, "step": 9091 }, { "epoch": 0.9856895056374675, "grad_norm": 0.6379157304763794, "learning_rate": 6.73872044102713e-05, "loss": 0.0236, "step": 9092 }, { "epoch": 0.9857979184735473, "grad_norm": 0.2249700129032135, "learning_rate": 6.738357754243436e-05, "loss": 0.0163, "step": 9093 }, { "epoch": 0.985906331309627, "grad_norm": 0.3739601969718933, "learning_rate": 6.737995067459743e-05, "loss": 0.0664, "step": 9094 }, { "epoch": 0.9860147441457069, "grad_norm": 0.19811314344406128, "learning_rate": 6.737632380676048e-05, "loss": 0.0152, "step": 9095 }, { "epoch": 0.9861231569817867, "grad_norm": 0.1831386238336563, "learning_rate": 6.737269693892356e-05, "loss": 0.0182, "step": 9096 }, { "epoch": 0.9862315698178664, "grad_norm": 0.4473780393600464, "learning_rate": 6.736907007108661e-05, "loss": 0.0571, "step": 9097 }, { "epoch": 0.9863399826539462, "grad_norm": 0.36564216017723083, "learning_rate": 6.736544320324968e-05, "loss": 0.0461, "step": 9098 }, { "epoch": 0.986448395490026, "grad_norm": 0.17538763582706451, "learning_rate": 6.736181633541274e-05, "loss": 0.0145, "step": 9099 }, { "epoch": 0.9865568083261058, "grad_norm": 0.8411363363265991, "learning_rate": 6.73581894675758e-05, "loss": 0.0729, "step": 9100 }, { "epoch": 0.9866652211621856, "grad_norm": 0.24473391473293304, "learning_rate": 6.735456259973887e-05, "loss": 0.044, "step": 9101 }, { "epoch": 0.9867736339982653, "grad_norm": 0.10340942442417145, "learning_rate": 6.735093573190193e-05, "loss": 0.0133, "step": 9102 }, { "epoch": 0.9868820468343452, "grad_norm": 0.44923073053359985, "learning_rate": 6.7347308864065e-05, "loss": 0.0593, "step": 9103 }, { "epoch": 0.986990459670425, "grad_norm": 0.2645065188407898, "learning_rate": 6.734368199622805e-05, "loss": 0.0693, "step": 9104 }, { "epoch": 0.9870988725065047, "grad_norm": 0.15722286701202393, "learning_rate": 6.734005512839113e-05, "loss": 0.0146, "step": 9105 }, { "epoch": 0.9872072853425846, "grad_norm": 0.3238564729690552, "learning_rate": 6.733642826055418e-05, "loss": 0.0311, "step": 9106 }, { "epoch": 0.9873156981786644, "grad_norm": 0.1689276248216629, "learning_rate": 6.733280139271725e-05, "loss": 0.0139, "step": 9107 }, { "epoch": 0.9874241110147441, "grad_norm": 0.1854236125946045, "learning_rate": 6.732917452488031e-05, "loss": 0.023, "step": 9108 }, { "epoch": 0.9875325238508239, "grad_norm": 0.14174692332744598, "learning_rate": 6.732554765704337e-05, "loss": 0.0185, "step": 9109 }, { "epoch": 0.9876409366869038, "grad_norm": 0.5454586744308472, "learning_rate": 6.732192078920645e-05, "loss": 0.05, "step": 9110 }, { "epoch": 0.9877493495229835, "grad_norm": 0.28015705943107605, "learning_rate": 6.731829392136951e-05, "loss": 0.0422, "step": 9111 }, { "epoch": 0.9878577623590633, "grad_norm": 0.3196266293525696, "learning_rate": 6.731466705353258e-05, "loss": 0.0301, "step": 9112 }, { "epoch": 0.9879661751951431, "grad_norm": 0.2957257628440857, "learning_rate": 6.731104018569564e-05, "loss": 0.0335, "step": 9113 }, { "epoch": 0.988074588031223, "grad_norm": 0.05309990793466568, "learning_rate": 6.730741331785871e-05, "loss": 0.0061, "step": 9114 }, { "epoch": 0.9881830008673027, "grad_norm": 0.5244719386100769, "learning_rate": 6.730378645002177e-05, "loss": 0.0528, "step": 9115 }, { "epoch": 0.9882914137033825, "grad_norm": 0.3319995701313019, "learning_rate": 6.730015958218482e-05, "loss": 0.0744, "step": 9116 }, { "epoch": 0.9883998265394622, "grad_norm": 0.18380212783813477, "learning_rate": 6.72965327143479e-05, "loss": 0.0203, "step": 9117 }, { "epoch": 0.9885082393755421, "grad_norm": 0.35724306106567383, "learning_rate": 6.729290584651095e-05, "loss": 0.0221, "step": 9118 }, { "epoch": 0.9886166522116219, "grad_norm": 0.5652357339859009, "learning_rate": 6.728927897867402e-05, "loss": 0.0537, "step": 9119 }, { "epoch": 0.9887250650477016, "grad_norm": 0.22999289631843567, "learning_rate": 6.728565211083708e-05, "loss": 0.0365, "step": 9120 }, { "epoch": 0.9888334778837814, "grad_norm": 0.3301178216934204, "learning_rate": 6.728202524300015e-05, "loss": 0.0599, "step": 9121 }, { "epoch": 0.9889418907198613, "grad_norm": 0.3603595793247223, "learning_rate": 6.727839837516321e-05, "loss": 0.0547, "step": 9122 }, { "epoch": 0.989050303555941, "grad_norm": 0.40787458419799805, "learning_rate": 6.727477150732628e-05, "loss": 0.0415, "step": 9123 }, { "epoch": 0.9891587163920208, "grad_norm": 0.6103476881980896, "learning_rate": 6.727114463948934e-05, "loss": 0.0476, "step": 9124 }, { "epoch": 0.9892671292281006, "grad_norm": 0.31118980050086975, "learning_rate": 6.72675177716524e-05, "loss": 0.0229, "step": 9125 }, { "epoch": 0.9893755420641804, "grad_norm": 0.32931992411613464, "learning_rate": 6.726389090381547e-05, "loss": 0.0451, "step": 9126 }, { "epoch": 0.9894839549002602, "grad_norm": 0.32883235812187195, "learning_rate": 6.726026403597854e-05, "loss": 0.0314, "step": 9127 }, { "epoch": 0.98959236773634, "grad_norm": 0.5148632526397705, "learning_rate": 6.725663716814161e-05, "loss": 0.044, "step": 9128 }, { "epoch": 0.9897007805724197, "grad_norm": 1.3848955631256104, "learning_rate": 6.725301030030466e-05, "loss": 0.0133, "step": 9129 }, { "epoch": 0.9898091934084996, "grad_norm": 0.31177017092704773, "learning_rate": 6.724938343246772e-05, "loss": 0.0266, "step": 9130 }, { "epoch": 0.9899176062445794, "grad_norm": 0.16001039743423462, "learning_rate": 6.724575656463079e-05, "loss": 0.0299, "step": 9131 }, { "epoch": 0.9900260190806591, "grad_norm": 0.2778678834438324, "learning_rate": 6.724212969679385e-05, "loss": 0.046, "step": 9132 }, { "epoch": 0.9901344319167389, "grad_norm": 0.6046464443206787, "learning_rate": 6.723850282895692e-05, "loss": 0.0684, "step": 9133 }, { "epoch": 0.9902428447528188, "grad_norm": 0.2653820216655731, "learning_rate": 6.723487596111998e-05, "loss": 0.0531, "step": 9134 }, { "epoch": 0.9903512575888985, "grad_norm": 0.27496206760406494, "learning_rate": 6.723124909328305e-05, "loss": 0.0226, "step": 9135 }, { "epoch": 0.9904596704249783, "grad_norm": 0.23357751965522766, "learning_rate": 6.72276222254461e-05, "loss": 0.0255, "step": 9136 }, { "epoch": 0.9905680832610582, "grad_norm": 0.19512823224067688, "learning_rate": 6.722399535760918e-05, "loss": 0.0248, "step": 9137 }, { "epoch": 0.9906764960971379, "grad_norm": 0.4003106951713562, "learning_rate": 6.722036848977223e-05, "loss": 0.03, "step": 9138 }, { "epoch": 0.9907849089332177, "grad_norm": 0.4849165380001068, "learning_rate": 6.721674162193529e-05, "loss": 0.0349, "step": 9139 }, { "epoch": 0.9908933217692975, "grad_norm": 0.2931378185749054, "learning_rate": 6.721311475409836e-05, "loss": 0.0276, "step": 9140 }, { "epoch": 0.9910017346053773, "grad_norm": 0.296080082654953, "learning_rate": 6.720948788626142e-05, "loss": 0.0255, "step": 9141 }, { "epoch": 0.9911101474414571, "grad_norm": 0.49985894560813904, "learning_rate": 6.720586101842449e-05, "loss": 0.0437, "step": 9142 }, { "epoch": 0.9912185602775369, "grad_norm": 0.08690288662910461, "learning_rate": 6.720223415058755e-05, "loss": 0.0094, "step": 9143 }, { "epoch": 0.9913269731136166, "grad_norm": 0.33640381693840027, "learning_rate": 6.719860728275062e-05, "loss": 0.0653, "step": 9144 }, { "epoch": 0.9914353859496965, "grad_norm": 0.27123135328292847, "learning_rate": 6.719498041491369e-05, "loss": 0.0331, "step": 9145 }, { "epoch": 0.9915437987857763, "grad_norm": 0.11968938261270523, "learning_rate": 6.719135354707675e-05, "loss": 0.0172, "step": 9146 }, { "epoch": 0.991652211621856, "grad_norm": 0.3483297824859619, "learning_rate": 6.718772667923982e-05, "loss": 0.0506, "step": 9147 }, { "epoch": 0.9917606244579358, "grad_norm": 0.30438750982284546, "learning_rate": 6.718409981140288e-05, "loss": 0.0562, "step": 9148 }, { "epoch": 0.9918690372940157, "grad_norm": 0.23366060853004456, "learning_rate": 6.718047294356595e-05, "loss": 0.0163, "step": 9149 }, { "epoch": 0.9919774501300954, "grad_norm": 0.12056143581867218, "learning_rate": 6.7176846075729e-05, "loss": 0.0108, "step": 9150 }, { "epoch": 0.9920858629661752, "grad_norm": 0.1304088830947876, "learning_rate": 6.717321920789207e-05, "loss": 0.0204, "step": 9151 }, { "epoch": 0.9921942758022549, "grad_norm": 0.26002997159957886, "learning_rate": 6.716959234005513e-05, "loss": 0.052, "step": 9152 }, { "epoch": 0.9923026886383348, "grad_norm": 0.2536824941635132, "learning_rate": 6.716596547221819e-05, "loss": 0.0683, "step": 9153 }, { "epoch": 0.9924111014744146, "grad_norm": 0.1841413527727127, "learning_rate": 6.716233860438126e-05, "loss": 0.0169, "step": 9154 }, { "epoch": 0.9925195143104943, "grad_norm": 0.36599162220954895, "learning_rate": 6.715871173654432e-05, "loss": 0.0408, "step": 9155 }, { "epoch": 0.9926279271465741, "grad_norm": 0.48096543550491333, "learning_rate": 6.715508486870739e-05, "loss": 0.0499, "step": 9156 }, { "epoch": 0.992736339982654, "grad_norm": 0.14950031042099, "learning_rate": 6.715145800087045e-05, "loss": 0.0328, "step": 9157 }, { "epoch": 0.9928447528187337, "grad_norm": 0.3829144239425659, "learning_rate": 6.714783113303352e-05, "loss": 0.0392, "step": 9158 }, { "epoch": 0.9929531656548135, "grad_norm": 0.427461177110672, "learning_rate": 6.714420426519657e-05, "loss": 0.0554, "step": 9159 }, { "epoch": 0.9930615784908933, "grad_norm": 0.2987295985221863, "learning_rate": 6.714057739735964e-05, "loss": 0.058, "step": 9160 }, { "epoch": 0.9931699913269731, "grad_norm": 0.29299458861351013, "learning_rate": 6.71369505295227e-05, "loss": 0.0286, "step": 9161 }, { "epoch": 0.9932784041630529, "grad_norm": 0.3338589668273926, "learning_rate": 6.713332366168577e-05, "loss": 0.1021, "step": 9162 }, { "epoch": 0.9933868169991327, "grad_norm": 0.41867274045944214, "learning_rate": 6.712969679384884e-05, "loss": 0.0683, "step": 9163 }, { "epoch": 0.9934952298352124, "grad_norm": 0.19165241718292236, "learning_rate": 6.71260699260119e-05, "loss": 0.0379, "step": 9164 }, { "epoch": 0.9936036426712923, "grad_norm": 0.34144964814186096, "learning_rate": 6.712244305817497e-05, "loss": 0.0302, "step": 9165 }, { "epoch": 0.9937120555073721, "grad_norm": 0.202350452542305, "learning_rate": 6.711881619033803e-05, "loss": 0.0267, "step": 9166 }, { "epoch": 0.9938204683434518, "grad_norm": 0.20346184074878693, "learning_rate": 6.711518932250109e-05, "loss": 0.0373, "step": 9167 }, { "epoch": 0.9939288811795317, "grad_norm": 0.34611988067626953, "learning_rate": 6.711156245466416e-05, "loss": 0.0307, "step": 9168 }, { "epoch": 0.9940372940156115, "grad_norm": 0.36059004068374634, "learning_rate": 6.710793558682722e-05, "loss": 0.0594, "step": 9169 }, { "epoch": 0.9941457068516912, "grad_norm": 0.26553693413734436, "learning_rate": 6.710430871899029e-05, "loss": 0.0329, "step": 9170 }, { "epoch": 0.994254119687771, "grad_norm": 0.4785275161266327, "learning_rate": 6.710068185115334e-05, "loss": 0.0467, "step": 9171 }, { "epoch": 0.9943625325238509, "grad_norm": 0.2658349275588989, "learning_rate": 6.709705498331641e-05, "loss": 0.0331, "step": 9172 }, { "epoch": 0.9944709453599306, "grad_norm": 0.32181766629219055, "learning_rate": 6.709342811547947e-05, "loss": 0.0451, "step": 9173 }, { "epoch": 0.9945793581960104, "grad_norm": 0.26721635460853577, "learning_rate": 6.708980124764254e-05, "loss": 0.0338, "step": 9174 }, { "epoch": 0.9946877710320902, "grad_norm": 0.2189682424068451, "learning_rate": 6.70861743798056e-05, "loss": 0.0262, "step": 9175 }, { "epoch": 0.99479618386817, "grad_norm": 0.1619308739900589, "learning_rate": 6.708254751196866e-05, "loss": 0.0182, "step": 9176 }, { "epoch": 0.9949045967042498, "grad_norm": 0.37838706374168396, "learning_rate": 6.707892064413173e-05, "loss": 0.0214, "step": 9177 }, { "epoch": 0.9950130095403296, "grad_norm": 0.35617104172706604, "learning_rate": 6.707529377629479e-05, "loss": 0.0456, "step": 9178 }, { "epoch": 0.9951214223764093, "grad_norm": 0.2152298092842102, "learning_rate": 6.707166690845787e-05, "loss": 0.0331, "step": 9179 }, { "epoch": 0.9952298352124892, "grad_norm": 0.40637698769569397, "learning_rate": 6.706804004062093e-05, "loss": 0.0397, "step": 9180 }, { "epoch": 0.995338248048569, "grad_norm": 0.38366732001304626, "learning_rate": 6.706441317278398e-05, "loss": 0.0496, "step": 9181 }, { "epoch": 0.9954466608846487, "grad_norm": 0.36510035395622253, "learning_rate": 6.706078630494706e-05, "loss": 0.0262, "step": 9182 }, { "epoch": 0.9955550737207285, "grad_norm": 0.46509218215942383, "learning_rate": 6.705715943711011e-05, "loss": 0.0395, "step": 9183 }, { "epoch": 0.9956634865568084, "grad_norm": 0.34880173206329346, "learning_rate": 6.705353256927318e-05, "loss": 0.0405, "step": 9184 }, { "epoch": 0.9957718993928881, "grad_norm": 0.3664461374282837, "learning_rate": 6.704990570143624e-05, "loss": 0.0592, "step": 9185 }, { "epoch": 0.9958803122289679, "grad_norm": 0.27136915922164917, "learning_rate": 6.704627883359931e-05, "loss": 0.0699, "step": 9186 }, { "epoch": 0.9959887250650477, "grad_norm": 0.12537752091884613, "learning_rate": 6.704265196576237e-05, "loss": 0.0164, "step": 9187 }, { "epoch": 0.9960971379011275, "grad_norm": 1.0155209302902222, "learning_rate": 6.703902509792544e-05, "loss": 0.0389, "step": 9188 }, { "epoch": 0.9962055507372073, "grad_norm": 0.3105839490890503, "learning_rate": 6.70353982300885e-05, "loss": 0.0266, "step": 9189 }, { "epoch": 0.996313963573287, "grad_norm": 0.15999151766300201, "learning_rate": 6.703177136225155e-05, "loss": 0.0158, "step": 9190 }, { "epoch": 0.9964223764093668, "grad_norm": 0.3050984740257263, "learning_rate": 6.702814449441463e-05, "loss": 0.039, "step": 9191 }, { "epoch": 0.9965307892454467, "grad_norm": 0.28605344891548157, "learning_rate": 6.702451762657768e-05, "loss": 0.0209, "step": 9192 }, { "epoch": 0.9966392020815265, "grad_norm": 0.16243688762187958, "learning_rate": 6.702089075874075e-05, "loss": 0.0161, "step": 9193 }, { "epoch": 0.9967476149176062, "grad_norm": 0.1414175182580948, "learning_rate": 6.701726389090381e-05, "loss": 0.0277, "step": 9194 }, { "epoch": 0.9968560277536861, "grad_norm": 0.2615200877189636, "learning_rate": 6.701363702306688e-05, "loss": 0.0203, "step": 9195 }, { "epoch": 0.9969644405897659, "grad_norm": 0.26562732458114624, "learning_rate": 6.701001015522995e-05, "loss": 0.0246, "step": 9196 }, { "epoch": 0.9970728534258456, "grad_norm": 0.2738878130912781, "learning_rate": 6.700638328739301e-05, "loss": 0.0153, "step": 9197 }, { "epoch": 0.9971812662619254, "grad_norm": 0.43332380056381226, "learning_rate": 6.700275641955608e-05, "loss": 0.0432, "step": 9198 }, { "epoch": 0.9972896790980053, "grad_norm": 0.6619243025779724, "learning_rate": 6.699912955171914e-05, "loss": 0.0569, "step": 9199 }, { "epoch": 0.997398091934085, "grad_norm": 0.43375420570373535, "learning_rate": 6.699550268388221e-05, "loss": 0.0598, "step": 9200 }, { "epoch": 0.9975065047701648, "grad_norm": 0.3184416592121124, "learning_rate": 6.699187581604527e-05, "loss": 0.039, "step": 9201 }, { "epoch": 0.9976149176062445, "grad_norm": 0.29888716340065, "learning_rate": 6.698824894820834e-05, "loss": 0.0412, "step": 9202 }, { "epoch": 0.9977233304423244, "grad_norm": 0.6579107046127319, "learning_rate": 6.69846220803714e-05, "loss": 0.0564, "step": 9203 }, { "epoch": 0.9978317432784042, "grad_norm": 0.20170535147190094, "learning_rate": 6.698099521253445e-05, "loss": 0.0294, "step": 9204 }, { "epoch": 0.997940156114484, "grad_norm": 0.295149028301239, "learning_rate": 6.697736834469752e-05, "loss": 0.0297, "step": 9205 }, { "epoch": 0.9980485689505637, "grad_norm": 0.4571382403373718, "learning_rate": 6.697374147686058e-05, "loss": 0.0586, "step": 9206 }, { "epoch": 0.9981569817866436, "grad_norm": 0.33137276768684387, "learning_rate": 6.697011460902365e-05, "loss": 0.026, "step": 9207 }, { "epoch": 0.9982653946227233, "grad_norm": 0.4360801875591278, "learning_rate": 6.696648774118671e-05, "loss": 0.0497, "step": 9208 }, { "epoch": 0.9983738074588031, "grad_norm": 0.39160212874412537, "learning_rate": 6.696286087334978e-05, "loss": 0.0289, "step": 9209 }, { "epoch": 0.9984822202948829, "grad_norm": 0.2184346318244934, "learning_rate": 6.695923400551284e-05, "loss": 0.0162, "step": 9210 }, { "epoch": 0.9985906331309627, "grad_norm": 1.8692030906677246, "learning_rate": 6.695560713767591e-05, "loss": 0.0242, "step": 9211 }, { "epoch": 0.9986990459670425, "grad_norm": 0.4108666777610779, "learning_rate": 6.695198026983896e-05, "loss": 0.0488, "step": 9212 }, { "epoch": 0.9988074588031223, "grad_norm": 0.35970473289489746, "learning_rate": 6.694835340200204e-05, "loss": 0.0425, "step": 9213 }, { "epoch": 0.998915871639202, "grad_norm": 0.23019731044769287, "learning_rate": 6.69447265341651e-05, "loss": 0.0261, "step": 9214 }, { "epoch": 0.9990242844752819, "grad_norm": 0.09200383722782135, "learning_rate": 6.694109966632816e-05, "loss": 0.0179, "step": 9215 }, { "epoch": 0.9991326973113617, "grad_norm": 0.5290650725364685, "learning_rate": 6.693747279849123e-05, "loss": 0.0525, "step": 9216 }, { "epoch": 0.9992411101474414, "grad_norm": 0.33128803968429565, "learning_rate": 6.693384593065429e-05, "loss": 0.0214, "step": 9217 }, { "epoch": 0.9993495229835212, "grad_norm": 0.624281108379364, "learning_rate": 6.693021906281736e-05, "loss": 0.058, "step": 9218 }, { "epoch": 0.9994579358196011, "grad_norm": 0.27490609884262085, "learning_rate": 6.692659219498042e-05, "loss": 0.0211, "step": 9219 }, { "epoch": 0.9995663486556808, "grad_norm": 0.5026644468307495, "learning_rate": 6.692296532714348e-05, "loss": 0.0565, "step": 9220 }, { "epoch": 0.9996747614917606, "grad_norm": 0.333249568939209, "learning_rate": 6.691933845930655e-05, "loss": 0.0113, "step": 9221 }, { "epoch": 0.9997831743278404, "grad_norm": 0.2892723083496094, "learning_rate": 6.69157115914696e-05, "loss": 0.0169, "step": 9222 }, { "epoch": 0.9998915871639202, "grad_norm": 0.5279408097267151, "learning_rate": 6.691208472363268e-05, "loss": 0.0169, "step": 9223 }, { "epoch": 1.0, "grad_norm": 0.17905442416667938, "learning_rate": 6.690845785579573e-05, "loss": 0.0113, "step": 9224 }, { "epoch": 1.0001084128360798, "grad_norm": 0.0732870027422905, "learning_rate": 6.69048309879588e-05, "loss": 0.0069, "step": 9225 }, { "epoch": 1.0002168256721595, "grad_norm": 0.6078823208808899, "learning_rate": 6.690120412012186e-05, "loss": 0.031, "step": 9226 }, { "epoch": 1.0003252385082393, "grad_norm": 0.3495068848133087, "learning_rate": 6.689757725228493e-05, "loss": 0.0319, "step": 9227 }, { "epoch": 1.0004336513443193, "grad_norm": 0.11021824926137924, "learning_rate": 6.689395038444799e-05, "loss": 0.008, "step": 9228 }, { "epoch": 1.000542064180399, "grad_norm": 0.2199871987104416, "learning_rate": 6.689032351661105e-05, "loss": 0.0189, "step": 9229 }, { "epoch": 1.0006504770164788, "grad_norm": 0.2226133942604065, "learning_rate": 6.688669664877412e-05, "loss": 0.0194, "step": 9230 }, { "epoch": 1.0007588898525586, "grad_norm": 0.1424850970506668, "learning_rate": 6.688306978093719e-05, "loss": 0.0062, "step": 9231 }, { "epoch": 1.0008673026886383, "grad_norm": 0.1287778615951538, "learning_rate": 6.687944291310026e-05, "loss": 0.0095, "step": 9232 }, { "epoch": 1.000975715524718, "grad_norm": 0.18749378621578217, "learning_rate": 6.687581604526332e-05, "loss": 0.0169, "step": 9233 }, { "epoch": 1.0010841283607979, "grad_norm": 0.11885199695825577, "learning_rate": 6.687218917742638e-05, "loss": 0.0131, "step": 9234 }, { "epoch": 1.0011925411968776, "grad_norm": 0.06443095952272415, "learning_rate": 6.686856230958945e-05, "loss": 0.0023, "step": 9235 }, { "epoch": 1.0013009540329576, "grad_norm": 0.16176556050777435, "learning_rate": 6.68649354417525e-05, "loss": 0.0181, "step": 9236 }, { "epoch": 1.0014093668690374, "grad_norm": 0.1360042840242386, "learning_rate": 6.686130857391557e-05, "loss": 0.0117, "step": 9237 }, { "epoch": 1.0015177797051171, "grad_norm": 0.018401438370347023, "learning_rate": 6.685768170607863e-05, "loss": 0.0013, "step": 9238 }, { "epoch": 1.001626192541197, "grad_norm": 0.23572947084903717, "learning_rate": 6.68540548382417e-05, "loss": 0.0366, "step": 9239 }, { "epoch": 1.0017346053772767, "grad_norm": 0.24277850985527039, "learning_rate": 6.685042797040476e-05, "loss": 0.0073, "step": 9240 }, { "epoch": 1.0018430182133564, "grad_norm": 0.34448501467704773, "learning_rate": 6.684680110256783e-05, "loss": 0.031, "step": 9241 }, { "epoch": 1.0019514310494362, "grad_norm": 0.20990034937858582, "learning_rate": 6.684317423473089e-05, "loss": 0.0185, "step": 9242 }, { "epoch": 1.002059843885516, "grad_norm": 0.20957233011722565, "learning_rate": 6.683954736689395e-05, "loss": 0.0134, "step": 9243 }, { "epoch": 1.002168256721596, "grad_norm": 0.2688913941383362, "learning_rate": 6.683592049905702e-05, "loss": 0.0165, "step": 9244 }, { "epoch": 1.0022766695576757, "grad_norm": 0.7504820227622986, "learning_rate": 6.683229363122007e-05, "loss": 0.0381, "step": 9245 }, { "epoch": 1.0023850823937555, "grad_norm": 0.22339338064193726, "learning_rate": 6.682866676338314e-05, "loss": 0.012, "step": 9246 }, { "epoch": 1.0024934952298352, "grad_norm": 0.3482402265071869, "learning_rate": 6.68250398955462e-05, "loss": 0.0401, "step": 9247 }, { "epoch": 1.002601908065915, "grad_norm": 0.5199922323226929, "learning_rate": 6.682141302770927e-05, "loss": 0.0742, "step": 9248 }, { "epoch": 1.0027103209019947, "grad_norm": 0.23077352344989777, "learning_rate": 6.681778615987234e-05, "loss": 0.0168, "step": 9249 }, { "epoch": 1.0028187337380745, "grad_norm": 0.15103191137313843, "learning_rate": 6.68141592920354e-05, "loss": 0.0126, "step": 9250 }, { "epoch": 1.0029271465741543, "grad_norm": 0.1620297133922577, "learning_rate": 6.681053242419847e-05, "loss": 0.0155, "step": 9251 }, { "epoch": 1.0030355594102343, "grad_norm": 0.07506758719682693, "learning_rate": 6.680690555636153e-05, "loss": 0.0077, "step": 9252 }, { "epoch": 1.003143972246314, "grad_norm": 0.6191442608833313, "learning_rate": 6.68032786885246e-05, "loss": 0.046, "step": 9253 }, { "epoch": 1.0032523850823938, "grad_norm": 0.3612527847290039, "learning_rate": 6.679965182068766e-05, "loss": 0.02, "step": 9254 }, { "epoch": 1.0033607979184735, "grad_norm": 0.20580905675888062, "learning_rate": 6.679602495285073e-05, "loss": 0.0129, "step": 9255 }, { "epoch": 1.0034692107545533, "grad_norm": 0.30203935503959656, "learning_rate": 6.679239808501379e-05, "loss": 0.0423, "step": 9256 }, { "epoch": 1.003577623590633, "grad_norm": 0.5526914596557617, "learning_rate": 6.678877121717684e-05, "loss": 0.0231, "step": 9257 }, { "epoch": 1.0036860364267128, "grad_norm": 0.40000858902931213, "learning_rate": 6.678514434933991e-05, "loss": 0.0349, "step": 9258 }, { "epoch": 1.0037944492627928, "grad_norm": 0.1626068353652954, "learning_rate": 6.678151748150297e-05, "loss": 0.0102, "step": 9259 }, { "epoch": 1.0039028620988726, "grad_norm": 0.19271868467330933, "learning_rate": 6.677789061366604e-05, "loss": 0.0142, "step": 9260 }, { "epoch": 1.0040112749349523, "grad_norm": 0.16930228471755981, "learning_rate": 6.67742637458291e-05, "loss": 0.016, "step": 9261 }, { "epoch": 1.004119687771032, "grad_norm": 0.21286897361278534, "learning_rate": 6.677063687799217e-05, "loss": 0.0207, "step": 9262 }, { "epoch": 1.0042281006071119, "grad_norm": 0.18942880630493164, "learning_rate": 6.676701001015523e-05, "loss": 0.0154, "step": 9263 }, { "epoch": 1.0043365134431916, "grad_norm": 0.23500728607177734, "learning_rate": 6.67633831423183e-05, "loss": 0.0251, "step": 9264 }, { "epoch": 1.0044449262792714, "grad_norm": 0.1680697351694107, "learning_rate": 6.675975627448137e-05, "loss": 0.0173, "step": 9265 }, { "epoch": 1.0045533391153512, "grad_norm": 0.14233455061912537, "learning_rate": 6.675612940664443e-05, "loss": 0.0119, "step": 9266 }, { "epoch": 1.0046617519514311, "grad_norm": 0.23508022725582123, "learning_rate": 6.67525025388075e-05, "loss": 0.0128, "step": 9267 }, { "epoch": 1.004770164787511, "grad_norm": 0.8862161040306091, "learning_rate": 6.674887567097055e-05, "loss": 0.0746, "step": 9268 }, { "epoch": 1.0048785776235907, "grad_norm": 0.6139551997184753, "learning_rate": 6.674524880313363e-05, "loss": 0.033, "step": 9269 }, { "epoch": 1.0049869904596704, "grad_norm": 0.2597910463809967, "learning_rate": 6.674162193529668e-05, "loss": 0.0072, "step": 9270 }, { "epoch": 1.0050954032957502, "grad_norm": 0.15703578293323517, "learning_rate": 6.673799506745974e-05, "loss": 0.0182, "step": 9271 }, { "epoch": 1.00520381613183, "grad_norm": 0.9546087980270386, "learning_rate": 6.673436819962281e-05, "loss": 0.0324, "step": 9272 }, { "epoch": 1.0053122289679097, "grad_norm": 0.24030695855617523, "learning_rate": 6.673074133178587e-05, "loss": 0.004, "step": 9273 }, { "epoch": 1.0054206418039895, "grad_norm": 0.17360363900661469, "learning_rate": 6.672711446394894e-05, "loss": 0.0113, "step": 9274 }, { "epoch": 1.0055290546400695, "grad_norm": 0.2614172697067261, "learning_rate": 6.6723487596112e-05, "loss": 0.0212, "step": 9275 }, { "epoch": 1.0056374674761492, "grad_norm": 0.19861115515232086, "learning_rate": 6.671986072827507e-05, "loss": 0.0142, "step": 9276 }, { "epoch": 1.005745880312229, "grad_norm": 0.29949474334716797, "learning_rate": 6.671623386043812e-05, "loss": 0.0154, "step": 9277 }, { "epoch": 1.0058542931483088, "grad_norm": 0.4777839481830597, "learning_rate": 6.67126069926012e-05, "loss": 0.0305, "step": 9278 }, { "epoch": 1.0059627059843885, "grad_norm": 0.12808792293071747, "learning_rate": 6.670898012476425e-05, "loss": 0.0069, "step": 9279 }, { "epoch": 1.0060711188204683, "grad_norm": 0.6601995229721069, "learning_rate": 6.670535325692731e-05, "loss": 0.0184, "step": 9280 }, { "epoch": 1.006179531656548, "grad_norm": 0.33340564370155334, "learning_rate": 6.670172638909038e-05, "loss": 0.0486, "step": 9281 }, { "epoch": 1.0062879444926278, "grad_norm": 0.12892264127731323, "learning_rate": 6.669809952125344e-05, "loss": 0.0158, "step": 9282 }, { "epoch": 1.0063963573287078, "grad_norm": 0.1801242232322693, "learning_rate": 6.669447265341652e-05, "loss": 0.0148, "step": 9283 }, { "epoch": 1.0065047701647876, "grad_norm": 0.09996093809604645, "learning_rate": 6.669084578557958e-05, "loss": 0.011, "step": 9284 }, { "epoch": 1.0066131830008673, "grad_norm": 0.08704381436109543, "learning_rate": 6.668721891774264e-05, "loss": 0.0055, "step": 9285 }, { "epoch": 1.006721595836947, "grad_norm": 0.40267008543014526, "learning_rate": 6.668359204990571e-05, "loss": 0.0624, "step": 9286 }, { "epoch": 1.0068300086730269, "grad_norm": 0.24978570640087128, "learning_rate": 6.667996518206877e-05, "loss": 0.0182, "step": 9287 }, { "epoch": 1.0069384215091066, "grad_norm": 0.4422915577888489, "learning_rate": 6.667633831423184e-05, "loss": 0.0491, "step": 9288 }, { "epoch": 1.0070468343451864, "grad_norm": 0.4350593686103821, "learning_rate": 6.66727114463949e-05, "loss": 0.0431, "step": 9289 }, { "epoch": 1.0071552471812664, "grad_norm": 0.1615227609872818, "learning_rate": 6.666908457855796e-05, "loss": 0.0145, "step": 9290 }, { "epoch": 1.0072636600173461, "grad_norm": 0.3720538020133972, "learning_rate": 6.666545771072102e-05, "loss": 0.0278, "step": 9291 }, { "epoch": 1.007372072853426, "grad_norm": 0.30933743715286255, "learning_rate": 6.666183084288409e-05, "loss": 0.0505, "step": 9292 }, { "epoch": 1.0074804856895057, "grad_norm": 0.07767612487077713, "learning_rate": 6.665820397504715e-05, "loss": 0.0115, "step": 9293 }, { "epoch": 1.0075888985255854, "grad_norm": 0.10495659708976746, "learning_rate": 6.665457710721021e-05, "loss": 0.0046, "step": 9294 }, { "epoch": 1.0076973113616652, "grad_norm": 0.40044429898262024, "learning_rate": 6.665095023937328e-05, "loss": 0.0263, "step": 9295 }, { "epoch": 1.007805724197745, "grad_norm": 0.21080926060676575, "learning_rate": 6.664732337153634e-05, "loss": 0.0148, "step": 9296 }, { "epoch": 1.0079141370338247, "grad_norm": 0.04684655740857124, "learning_rate": 6.664369650369941e-05, "loss": 0.0017, "step": 9297 }, { "epoch": 1.0080225498699047, "grad_norm": 0.4542241394519806, "learning_rate": 6.664006963586246e-05, "loss": 0.0252, "step": 9298 }, { "epoch": 1.0081309627059845, "grad_norm": 0.17413178086280823, "learning_rate": 6.663644276802554e-05, "loss": 0.0257, "step": 9299 }, { "epoch": 1.0082393755420642, "grad_norm": 0.06291931867599487, "learning_rate": 6.66328159001886e-05, "loss": 0.0042, "step": 9300 }, { "epoch": 1.008347788378144, "grad_norm": 0.08874152600765228, "learning_rate": 6.662918903235166e-05, "loss": 0.0092, "step": 9301 }, { "epoch": 1.0084562012142237, "grad_norm": 0.4166797697544098, "learning_rate": 6.662556216451473e-05, "loss": 0.0314, "step": 9302 }, { "epoch": 1.0085646140503035, "grad_norm": 0.1983240693807602, "learning_rate": 6.662193529667779e-05, "loss": 0.0141, "step": 9303 }, { "epoch": 1.0086730268863833, "grad_norm": 0.2878710627555847, "learning_rate": 6.661830842884086e-05, "loss": 0.0125, "step": 9304 }, { "epoch": 1.008781439722463, "grad_norm": 0.20355910062789917, "learning_rate": 6.661468156100392e-05, "loss": 0.0223, "step": 9305 }, { "epoch": 1.008889852558543, "grad_norm": 0.22292280197143555, "learning_rate": 6.661105469316699e-05, "loss": 0.0201, "step": 9306 }, { "epoch": 1.0089982653946228, "grad_norm": 0.2960481345653534, "learning_rate": 6.660742782533005e-05, "loss": 0.0121, "step": 9307 }, { "epoch": 1.0091066782307025, "grad_norm": 0.19415612518787384, "learning_rate": 6.660380095749312e-05, "loss": 0.0244, "step": 9308 }, { "epoch": 1.0092150910667823, "grad_norm": 0.16570352017879486, "learning_rate": 6.660017408965618e-05, "loss": 0.0289, "step": 9309 }, { "epoch": 1.009323503902862, "grad_norm": 0.43903496861457825, "learning_rate": 6.659654722181923e-05, "loss": 0.026, "step": 9310 }, { "epoch": 1.0094319167389418, "grad_norm": 0.09922769665718079, "learning_rate": 6.65929203539823e-05, "loss": 0.0104, "step": 9311 }, { "epoch": 1.0095403295750216, "grad_norm": 0.043534908443689346, "learning_rate": 6.658929348614536e-05, "loss": 0.0011, "step": 9312 }, { "epoch": 1.0096487424111014, "grad_norm": 0.22223801910877228, "learning_rate": 6.658566661830843e-05, "loss": 0.0117, "step": 9313 }, { "epoch": 1.0097571552471813, "grad_norm": 0.30876535177230835, "learning_rate": 6.658203975047149e-05, "loss": 0.027, "step": 9314 }, { "epoch": 1.0098655680832611, "grad_norm": 0.20680108666419983, "learning_rate": 6.657841288263456e-05, "loss": 0.0199, "step": 9315 }, { "epoch": 1.0099739809193409, "grad_norm": 0.2978736162185669, "learning_rate": 6.657478601479762e-05, "loss": 0.0144, "step": 9316 }, { "epoch": 1.0100823937554206, "grad_norm": 0.11467432975769043, "learning_rate": 6.657115914696069e-05, "loss": 0.0115, "step": 9317 }, { "epoch": 1.0101908065915004, "grad_norm": 0.19903098046779633, "learning_rate": 6.656753227912376e-05, "loss": 0.0193, "step": 9318 }, { "epoch": 1.0102992194275802, "grad_norm": 0.4339867830276489, "learning_rate": 6.656390541128682e-05, "loss": 0.0232, "step": 9319 }, { "epoch": 1.01040763226366, "grad_norm": 0.2925872504711151, "learning_rate": 6.656027854344989e-05, "loss": 0.0267, "step": 9320 }, { "epoch": 1.01051604509974, "grad_norm": 0.2272663116455078, "learning_rate": 6.655665167561295e-05, "loss": 0.0342, "step": 9321 }, { "epoch": 1.0106244579358197, "grad_norm": 0.039808426052331924, "learning_rate": 6.655302480777602e-05, "loss": 0.0021, "step": 9322 }, { "epoch": 1.0107328707718994, "grad_norm": 0.4817880392074585, "learning_rate": 6.654939793993907e-05, "loss": 0.0145, "step": 9323 }, { "epoch": 1.0108412836079792, "grad_norm": 0.21663568913936615, "learning_rate": 6.654577107210213e-05, "loss": 0.0076, "step": 9324 }, { "epoch": 1.010949696444059, "grad_norm": 0.4950638711452484, "learning_rate": 6.65421442042652e-05, "loss": 0.0201, "step": 9325 }, { "epoch": 1.0110581092801387, "grad_norm": 0.37711188197135925, "learning_rate": 6.653851733642826e-05, "loss": 0.0289, "step": 9326 }, { "epoch": 1.0111665221162185, "grad_norm": 0.49859753251075745, "learning_rate": 6.653489046859133e-05, "loss": 0.0332, "step": 9327 }, { "epoch": 1.0112749349522983, "grad_norm": 0.5680445432662964, "learning_rate": 6.653126360075439e-05, "loss": 0.0291, "step": 9328 }, { "epoch": 1.0113833477883782, "grad_norm": 0.06505092978477478, "learning_rate": 6.652763673291746e-05, "loss": 0.0047, "step": 9329 }, { "epoch": 1.011491760624458, "grad_norm": 0.19079484045505524, "learning_rate": 6.652400986508052e-05, "loss": 0.0143, "step": 9330 }, { "epoch": 1.0116001734605378, "grad_norm": 0.479708194732666, "learning_rate": 6.652038299724359e-05, "loss": 0.0351, "step": 9331 }, { "epoch": 1.0117085862966175, "grad_norm": 0.1834098994731903, "learning_rate": 6.651675612940664e-05, "loss": 0.0169, "step": 9332 }, { "epoch": 1.0118169991326973, "grad_norm": 0.02905784174799919, "learning_rate": 6.65131292615697e-05, "loss": 0.0014, "step": 9333 }, { "epoch": 1.011925411968777, "grad_norm": 0.5166054368019104, "learning_rate": 6.650950239373279e-05, "loss": 0.0499, "step": 9334 }, { "epoch": 1.0120338248048568, "grad_norm": 0.3390372395515442, "learning_rate": 6.650587552589584e-05, "loss": 0.0159, "step": 9335 }, { "epoch": 1.0121422376409366, "grad_norm": 0.3277348279953003, "learning_rate": 6.650224865805891e-05, "loss": 0.0087, "step": 9336 }, { "epoch": 1.0122506504770166, "grad_norm": 0.83063805103302, "learning_rate": 6.649862179022197e-05, "loss": 0.0237, "step": 9337 }, { "epoch": 1.0123590633130963, "grad_norm": 0.2727730870246887, "learning_rate": 6.649499492238503e-05, "loss": 0.019, "step": 9338 }, { "epoch": 1.012467476149176, "grad_norm": 1.2059091329574585, "learning_rate": 6.64913680545481e-05, "loss": 0.0498, "step": 9339 }, { "epoch": 1.0125758889852559, "grad_norm": 0.3389684557914734, "learning_rate": 6.648774118671116e-05, "loss": 0.0062, "step": 9340 }, { "epoch": 1.0126843018213356, "grad_norm": 0.39562276005744934, "learning_rate": 6.648411431887423e-05, "loss": 0.0101, "step": 9341 }, { "epoch": 1.0127927146574154, "grad_norm": 0.3421984612941742, "learning_rate": 6.648048745103728e-05, "loss": 0.0094, "step": 9342 }, { "epoch": 1.0129011274934951, "grad_norm": 0.0279091689735651, "learning_rate": 6.647686058320036e-05, "loss": 0.0014, "step": 9343 }, { "epoch": 1.0130095403295751, "grad_norm": 0.2095477432012558, "learning_rate": 6.647323371536341e-05, "loss": 0.0127, "step": 9344 }, { "epoch": 1.013117953165655, "grad_norm": 0.5148013830184937, "learning_rate": 6.646960684752648e-05, "loss": 0.0406, "step": 9345 }, { "epoch": 1.0132263660017347, "grad_norm": 0.5713534951210022, "learning_rate": 6.646597997968954e-05, "loss": 0.0435, "step": 9346 }, { "epoch": 1.0133347788378144, "grad_norm": 0.490468829870224, "learning_rate": 6.64623531118526e-05, "loss": 0.0505, "step": 9347 }, { "epoch": 1.0134431916738942, "grad_norm": 0.47935962677001953, "learning_rate": 6.645872624401567e-05, "loss": 0.0269, "step": 9348 }, { "epoch": 1.013551604509974, "grad_norm": 0.11991070210933685, "learning_rate": 6.645509937617873e-05, "loss": 0.0075, "step": 9349 }, { "epoch": 1.0136600173460537, "grad_norm": 0.333152174949646, "learning_rate": 6.64514725083418e-05, "loss": 0.023, "step": 9350 }, { "epoch": 1.0137684301821335, "grad_norm": 0.38237786293029785, "learning_rate": 6.644784564050486e-05, "loss": 0.0122, "step": 9351 }, { "epoch": 1.0138768430182135, "grad_norm": 0.35875269770622253, "learning_rate": 6.644421877266793e-05, "loss": 0.0456, "step": 9352 }, { "epoch": 1.0139852558542932, "grad_norm": 0.15630558133125305, "learning_rate": 6.6440591904831e-05, "loss": 0.0098, "step": 9353 }, { "epoch": 1.014093668690373, "grad_norm": 0.45971205830574036, "learning_rate": 6.643696503699405e-05, "loss": 0.0893, "step": 9354 }, { "epoch": 1.0142020815264527, "grad_norm": 0.1053619384765625, "learning_rate": 6.643333816915713e-05, "loss": 0.0072, "step": 9355 }, { "epoch": 1.0143104943625325, "grad_norm": 0.21167752146720886, "learning_rate": 6.642971130132018e-05, "loss": 0.0217, "step": 9356 }, { "epoch": 1.0144189071986123, "grad_norm": 0.21139267086982727, "learning_rate": 6.642608443348325e-05, "loss": 0.0514, "step": 9357 }, { "epoch": 1.014527320034692, "grad_norm": 0.11742836236953735, "learning_rate": 6.642245756564631e-05, "loss": 0.0079, "step": 9358 }, { "epoch": 1.0146357328707718, "grad_norm": 0.45519348978996277, "learning_rate": 6.641883069780938e-05, "loss": 0.0144, "step": 9359 }, { "epoch": 1.0147441457068518, "grad_norm": 0.6624757647514343, "learning_rate": 6.641520382997244e-05, "loss": 0.0418, "step": 9360 }, { "epoch": 1.0148525585429315, "grad_norm": 0.2753300070762634, "learning_rate": 6.64115769621355e-05, "loss": 0.0432, "step": 9361 }, { "epoch": 1.0149609713790113, "grad_norm": 0.16991668939590454, "learning_rate": 6.640795009429857e-05, "loss": 0.0088, "step": 9362 }, { "epoch": 1.015069384215091, "grad_norm": 0.4062207043170929, "learning_rate": 6.640432322646162e-05, "loss": 0.0423, "step": 9363 }, { "epoch": 1.0151777970511708, "grad_norm": 0.4806998670101166, "learning_rate": 6.64006963586247e-05, "loss": 0.0249, "step": 9364 }, { "epoch": 1.0152862098872506, "grad_norm": 0.5084930658340454, "learning_rate": 6.639706949078775e-05, "loss": 0.0275, "step": 9365 }, { "epoch": 1.0153946227233304, "grad_norm": 1.8140647411346436, "learning_rate": 6.639344262295082e-05, "loss": 0.0307, "step": 9366 }, { "epoch": 1.0155030355594101, "grad_norm": 0.13268578052520752, "learning_rate": 6.638981575511388e-05, "loss": 0.0067, "step": 9367 }, { "epoch": 1.0156114483954901, "grad_norm": 0.2913324236869812, "learning_rate": 6.638618888727695e-05, "loss": 0.0527, "step": 9368 }, { "epoch": 1.0157198612315699, "grad_norm": 0.4089588224887848, "learning_rate": 6.638256201944002e-05, "loss": 0.0213, "step": 9369 }, { "epoch": 1.0158282740676496, "grad_norm": 0.27850818634033203, "learning_rate": 6.637893515160308e-05, "loss": 0.0237, "step": 9370 }, { "epoch": 1.0159366869037294, "grad_norm": 0.5196137428283691, "learning_rate": 6.637530828376615e-05, "loss": 0.0251, "step": 9371 }, { "epoch": 1.0160450997398092, "grad_norm": 0.251250684261322, "learning_rate": 6.637168141592921e-05, "loss": 0.0084, "step": 9372 }, { "epoch": 1.016153512575889, "grad_norm": 0.060885969549417496, "learning_rate": 6.636805454809228e-05, "loss": 0.0057, "step": 9373 }, { "epoch": 1.0162619254119687, "grad_norm": 0.2688070833683014, "learning_rate": 6.636442768025534e-05, "loss": 0.0134, "step": 9374 }, { "epoch": 1.0163703382480487, "grad_norm": 0.14787247776985168, "learning_rate": 6.63608008124184e-05, "loss": 0.0154, "step": 9375 }, { "epoch": 1.0164787510841284, "grad_norm": 0.27490001916885376, "learning_rate": 6.635717394458146e-05, "loss": 0.0241, "step": 9376 }, { "epoch": 1.0165871639202082, "grad_norm": 0.10688848048448563, "learning_rate": 6.635354707674452e-05, "loss": 0.0076, "step": 9377 }, { "epoch": 1.016695576756288, "grad_norm": 0.18713165819644928, "learning_rate": 6.634992020890759e-05, "loss": 0.0054, "step": 9378 }, { "epoch": 1.0168039895923677, "grad_norm": 0.3901989161968231, "learning_rate": 6.634629334107065e-05, "loss": 0.028, "step": 9379 }, { "epoch": 1.0169124024284475, "grad_norm": 0.49963071942329407, "learning_rate": 6.634266647323372e-05, "loss": 0.0219, "step": 9380 }, { "epoch": 1.0170208152645273, "grad_norm": 0.040974490344524384, "learning_rate": 6.633903960539678e-05, "loss": 0.003, "step": 9381 }, { "epoch": 1.017129228100607, "grad_norm": 0.26904168725013733, "learning_rate": 6.633541273755985e-05, "loss": 0.0459, "step": 9382 }, { "epoch": 1.017237640936687, "grad_norm": 0.43338218331336975, "learning_rate": 6.63317858697229e-05, "loss": 0.0373, "step": 9383 }, { "epoch": 1.0173460537727668, "grad_norm": 0.42119118571281433, "learning_rate": 6.632815900188596e-05, "loss": 0.0331, "step": 9384 }, { "epoch": 1.0174544666088465, "grad_norm": 1.3814064264297485, "learning_rate": 6.632453213404903e-05, "loss": 0.0295, "step": 9385 }, { "epoch": 1.0175628794449263, "grad_norm": 0.41763272881507874, "learning_rate": 6.63209052662121e-05, "loss": 0.0277, "step": 9386 }, { "epoch": 1.017671292281006, "grad_norm": 0.36338183283805847, "learning_rate": 6.631727839837518e-05, "loss": 0.0231, "step": 9387 }, { "epoch": 1.0177797051170858, "grad_norm": 0.24590693414211273, "learning_rate": 6.631365153053823e-05, "loss": 0.0059, "step": 9388 }, { "epoch": 1.0178881179531656, "grad_norm": 0.2684691548347473, "learning_rate": 6.631002466270129e-05, "loss": 0.0467, "step": 9389 }, { "epoch": 1.0179965307892453, "grad_norm": 0.3788217604160309, "learning_rate": 6.630639779486436e-05, "loss": 0.0156, "step": 9390 }, { "epoch": 1.0181049436253253, "grad_norm": 0.6274022459983826, "learning_rate": 6.630277092702742e-05, "loss": 0.0305, "step": 9391 }, { "epoch": 1.018213356461405, "grad_norm": 0.34757956862449646, "learning_rate": 6.629914405919049e-05, "loss": 0.0195, "step": 9392 }, { "epoch": 1.0183217692974849, "grad_norm": 0.22046653926372528, "learning_rate": 6.629551719135355e-05, "loss": 0.0311, "step": 9393 }, { "epoch": 1.0184301821335646, "grad_norm": 0.10345524549484253, "learning_rate": 6.629189032351662e-05, "loss": 0.0095, "step": 9394 }, { "epoch": 1.0185385949696444, "grad_norm": 0.26874640583992004, "learning_rate": 6.628826345567968e-05, "loss": 0.0185, "step": 9395 }, { "epoch": 1.0186470078057241, "grad_norm": 0.0790790468454361, "learning_rate": 6.628463658784275e-05, "loss": 0.0063, "step": 9396 }, { "epoch": 1.018755420641804, "grad_norm": 0.18846456706523895, "learning_rate": 6.62810097200058e-05, "loss": 0.0262, "step": 9397 }, { "epoch": 1.0188638334778837, "grad_norm": 0.5107429623603821, "learning_rate": 6.627738285216886e-05, "loss": 0.0379, "step": 9398 }, { "epoch": 1.0189722463139637, "grad_norm": 0.24876344203948975, "learning_rate": 6.627375598433193e-05, "loss": 0.0235, "step": 9399 }, { "epoch": 1.0190806591500434, "grad_norm": 0.3369363844394684, "learning_rate": 6.627012911649499e-05, "loss": 0.0276, "step": 9400 }, { "epoch": 1.0191890719861232, "grad_norm": 0.44610464572906494, "learning_rate": 6.626650224865806e-05, "loss": 0.0506, "step": 9401 }, { "epoch": 1.019297484822203, "grad_norm": 0.1548318713903427, "learning_rate": 6.626287538082112e-05, "loss": 0.0106, "step": 9402 }, { "epoch": 1.0194058976582827, "grad_norm": 0.16156414151191711, "learning_rate": 6.62592485129842e-05, "loss": 0.0117, "step": 9403 }, { "epoch": 1.0195143104943625, "grad_norm": 0.2851153314113617, "learning_rate": 6.625562164514726e-05, "loss": 0.0296, "step": 9404 }, { "epoch": 1.0196227233304422, "grad_norm": 0.2606176435947418, "learning_rate": 6.625199477731032e-05, "loss": 0.0153, "step": 9405 }, { "epoch": 1.0197311361665222, "grad_norm": 0.4700334668159485, "learning_rate": 6.624836790947339e-05, "loss": 0.0408, "step": 9406 }, { "epoch": 1.019839549002602, "grad_norm": 0.34574615955352783, "learning_rate": 6.624474104163644e-05, "loss": 0.0366, "step": 9407 }, { "epoch": 1.0199479618386817, "grad_norm": 0.03265557810664177, "learning_rate": 6.624111417379952e-05, "loss": 0.0022, "step": 9408 }, { "epoch": 1.0200563746747615, "grad_norm": 0.20744433999061584, "learning_rate": 6.623748730596257e-05, "loss": 0.0163, "step": 9409 }, { "epoch": 1.0201647875108413, "grad_norm": 0.3311403691768646, "learning_rate": 6.623386043812564e-05, "loss": 0.0301, "step": 9410 }, { "epoch": 1.020273200346921, "grad_norm": 0.19645735621452332, "learning_rate": 6.62302335702887e-05, "loss": 0.0171, "step": 9411 }, { "epoch": 1.0203816131830008, "grad_norm": 0.5985245704650879, "learning_rate": 6.622660670245177e-05, "loss": 0.0831, "step": 9412 }, { "epoch": 1.0204900260190806, "grad_norm": 0.5374132394790649, "learning_rate": 6.622297983461483e-05, "loss": 0.0412, "step": 9413 }, { "epoch": 1.0205984388551606, "grad_norm": 0.10941548645496368, "learning_rate": 6.621935296677789e-05, "loss": 0.0046, "step": 9414 }, { "epoch": 1.0207068516912403, "grad_norm": 0.42866450548171997, "learning_rate": 6.621572609894096e-05, "loss": 0.0714, "step": 9415 }, { "epoch": 1.02081526452732, "grad_norm": 0.5790627002716064, "learning_rate": 6.621209923110402e-05, "loss": 0.0581, "step": 9416 }, { "epoch": 1.0209236773633998, "grad_norm": 0.18499553203582764, "learning_rate": 6.620847236326709e-05, "loss": 0.0094, "step": 9417 }, { "epoch": 1.0210320901994796, "grad_norm": 0.13529878854751587, "learning_rate": 6.620484549543014e-05, "loss": 0.0054, "step": 9418 }, { "epoch": 1.0211405030355594, "grad_norm": 0.1435297131538391, "learning_rate": 6.620121862759321e-05, "loss": 0.0223, "step": 9419 }, { "epoch": 1.0212489158716391, "grad_norm": 0.03688652068376541, "learning_rate": 6.619759175975627e-05, "loss": 0.0033, "step": 9420 }, { "epoch": 1.021357328707719, "grad_norm": 0.3791675567626953, "learning_rate": 6.619396489191934e-05, "loss": 0.0336, "step": 9421 }, { "epoch": 1.0214657415437989, "grad_norm": 0.08395209163427353, "learning_rate": 6.619033802408241e-05, "loss": 0.0049, "step": 9422 }, { "epoch": 1.0215741543798786, "grad_norm": 0.24343566596508026, "learning_rate": 6.618671115624547e-05, "loss": 0.0373, "step": 9423 }, { "epoch": 1.0216825672159584, "grad_norm": 0.31454288959503174, "learning_rate": 6.618308428840854e-05, "loss": 0.0216, "step": 9424 }, { "epoch": 1.0217909800520382, "grad_norm": 0.36309850215911865, "learning_rate": 6.61794574205716e-05, "loss": 0.0405, "step": 9425 }, { "epoch": 1.021899392888118, "grad_norm": 0.3921997845172882, "learning_rate": 6.617583055273467e-05, "loss": 0.0619, "step": 9426 }, { "epoch": 1.0220078057241977, "grad_norm": 0.35992196202278137, "learning_rate": 6.617220368489773e-05, "loss": 0.0377, "step": 9427 }, { "epoch": 1.0221162185602775, "grad_norm": 0.39710065722465515, "learning_rate": 6.616857681706078e-05, "loss": 0.0324, "step": 9428 }, { "epoch": 1.0222246313963572, "grad_norm": 0.4354221522808075, "learning_rate": 6.616494994922386e-05, "loss": 0.0353, "step": 9429 }, { "epoch": 1.0223330442324372, "grad_norm": 0.29129964113235474, "learning_rate": 6.616132308138691e-05, "loss": 0.0131, "step": 9430 }, { "epoch": 1.022441457068517, "grad_norm": 0.3801800310611725, "learning_rate": 6.615769621354998e-05, "loss": 0.0173, "step": 9431 }, { "epoch": 1.0225498699045967, "grad_norm": 0.3239506483078003, "learning_rate": 6.615406934571304e-05, "loss": 0.0353, "step": 9432 }, { "epoch": 1.0226582827406765, "grad_norm": 0.33212941884994507, "learning_rate": 6.615044247787611e-05, "loss": 0.0155, "step": 9433 }, { "epoch": 1.0227666955767563, "grad_norm": 0.2244211882352829, "learning_rate": 6.614681561003917e-05, "loss": 0.0257, "step": 9434 }, { "epoch": 1.022875108412836, "grad_norm": 0.17220278084278107, "learning_rate": 6.614318874220224e-05, "loss": 0.0266, "step": 9435 }, { "epoch": 1.0229835212489158, "grad_norm": 0.2707916796207428, "learning_rate": 6.61395618743653e-05, "loss": 0.0318, "step": 9436 }, { "epoch": 1.0230919340849958, "grad_norm": 0.1261947602033615, "learning_rate": 6.613593500652835e-05, "loss": 0.0126, "step": 9437 }, { "epoch": 1.0232003469210755, "grad_norm": 0.34837183356285095, "learning_rate": 6.613230813869144e-05, "loss": 0.0183, "step": 9438 }, { "epoch": 1.0233087597571553, "grad_norm": 0.09244758635759354, "learning_rate": 6.61286812708545e-05, "loss": 0.0098, "step": 9439 }, { "epoch": 1.023417172593235, "grad_norm": 0.1530708372592926, "learning_rate": 6.612505440301757e-05, "loss": 0.0137, "step": 9440 }, { "epoch": 1.0235255854293148, "grad_norm": 0.1989140808582306, "learning_rate": 6.612142753518062e-05, "loss": 0.0141, "step": 9441 }, { "epoch": 1.0236339982653946, "grad_norm": 0.054624997079372406, "learning_rate": 6.611780066734368e-05, "loss": 0.0053, "step": 9442 }, { "epoch": 1.0237424111014743, "grad_norm": 0.24035504460334778, "learning_rate": 6.611417379950675e-05, "loss": 0.017, "step": 9443 }, { "epoch": 1.023850823937554, "grad_norm": 0.3361477851867676, "learning_rate": 6.611054693166981e-05, "loss": 0.0143, "step": 9444 }, { "epoch": 1.023959236773634, "grad_norm": 0.20946179330348969, "learning_rate": 6.610692006383288e-05, "loss": 0.0153, "step": 9445 }, { "epoch": 1.0240676496097139, "grad_norm": 0.11141499876976013, "learning_rate": 6.610329319599594e-05, "loss": 0.0103, "step": 9446 }, { "epoch": 1.0241760624457936, "grad_norm": 0.10054060071706772, "learning_rate": 6.609966632815901e-05, "loss": 0.0175, "step": 9447 }, { "epoch": 1.0242844752818734, "grad_norm": 0.36889368295669556, "learning_rate": 6.609603946032207e-05, "loss": 0.0238, "step": 9448 }, { "epoch": 1.0243928881179531, "grad_norm": 0.5250794887542725, "learning_rate": 6.609241259248514e-05, "loss": 0.0361, "step": 9449 }, { "epoch": 1.024501300954033, "grad_norm": 0.1133662760257721, "learning_rate": 6.60887857246482e-05, "loss": 0.0075, "step": 9450 }, { "epoch": 1.0246097137901127, "grad_norm": 0.6476777791976929, "learning_rate": 6.608515885681125e-05, "loss": 0.0316, "step": 9451 }, { "epoch": 1.0247181266261924, "grad_norm": 0.10093680769205093, "learning_rate": 6.608153198897432e-05, "loss": 0.0058, "step": 9452 }, { "epoch": 1.0248265394622724, "grad_norm": 0.3482140600681305, "learning_rate": 6.607790512113738e-05, "loss": 0.0536, "step": 9453 }, { "epoch": 1.0249349522983522, "grad_norm": 0.5680748224258423, "learning_rate": 6.607427825330045e-05, "loss": 0.028, "step": 9454 }, { "epoch": 1.025043365134432, "grad_norm": 0.46114686131477356, "learning_rate": 6.607065138546352e-05, "loss": 0.0163, "step": 9455 }, { "epoch": 1.0251517779705117, "grad_norm": 0.30733928084373474, "learning_rate": 6.606702451762658e-05, "loss": 0.0165, "step": 9456 }, { "epoch": 1.0252601908065915, "grad_norm": 0.6643882393836975, "learning_rate": 6.606339764978965e-05, "loss": 0.0435, "step": 9457 }, { "epoch": 1.0253686036426712, "grad_norm": 0.4106391370296478, "learning_rate": 6.605977078195271e-05, "loss": 0.0629, "step": 9458 }, { "epoch": 1.025477016478751, "grad_norm": 0.2688944935798645, "learning_rate": 6.605614391411578e-05, "loss": 0.0244, "step": 9459 }, { "epoch": 1.0255854293148308, "grad_norm": 0.0882270410656929, "learning_rate": 6.605251704627884e-05, "loss": 0.0059, "step": 9460 }, { "epoch": 1.0256938421509108, "grad_norm": 0.10138948261737823, "learning_rate": 6.60488901784419e-05, "loss": 0.0045, "step": 9461 }, { "epoch": 1.0258022549869905, "grad_norm": 0.26110684871673584, "learning_rate": 6.604526331060496e-05, "loss": 0.0278, "step": 9462 }, { "epoch": 1.0259106678230703, "grad_norm": 0.134074866771698, "learning_rate": 6.604163644276803e-05, "loss": 0.0063, "step": 9463 }, { "epoch": 1.02601908065915, "grad_norm": 0.11439037322998047, "learning_rate": 6.603800957493109e-05, "loss": 0.0064, "step": 9464 }, { "epoch": 1.0261274934952298, "grad_norm": 0.17741842567920685, "learning_rate": 6.603438270709415e-05, "loss": 0.008, "step": 9465 }, { "epoch": 1.0262359063313096, "grad_norm": 0.3209371566772461, "learning_rate": 6.603075583925722e-05, "loss": 0.0312, "step": 9466 }, { "epoch": 1.0263443191673893, "grad_norm": 0.2871388792991638, "learning_rate": 6.602712897142028e-05, "loss": 0.0565, "step": 9467 }, { "epoch": 1.0264527320034693, "grad_norm": 0.3015412986278534, "learning_rate": 6.602350210358335e-05, "loss": 0.039, "step": 9468 }, { "epoch": 1.026561144839549, "grad_norm": 0.2236144095659256, "learning_rate": 6.60198752357464e-05, "loss": 0.0189, "step": 9469 }, { "epoch": 1.0266695576756288, "grad_norm": 0.4950578808784485, "learning_rate": 6.601624836790948e-05, "loss": 0.0325, "step": 9470 }, { "epoch": 1.0267779705117086, "grad_norm": 0.47897398471832275, "learning_rate": 6.601262150007253e-05, "loss": 0.0201, "step": 9471 }, { "epoch": 1.0268863833477884, "grad_norm": 0.4546453654766083, "learning_rate": 6.60089946322356e-05, "loss": 0.0417, "step": 9472 }, { "epoch": 1.0269947961838681, "grad_norm": 0.36522388458251953, "learning_rate": 6.600536776439868e-05, "loss": 0.0291, "step": 9473 }, { "epoch": 1.027103209019948, "grad_norm": 0.14071007072925568, "learning_rate": 6.600174089656173e-05, "loss": 0.007, "step": 9474 }, { "epoch": 1.0272116218560277, "grad_norm": 0.3722988963127136, "learning_rate": 6.59981140287248e-05, "loss": 0.0395, "step": 9475 }, { "epoch": 1.0273200346921076, "grad_norm": 0.3645186722278595, "learning_rate": 6.599448716088786e-05, "loss": 0.019, "step": 9476 }, { "epoch": 1.0274284475281874, "grad_norm": 0.3759457767009735, "learning_rate": 6.599086029305093e-05, "loss": 0.0462, "step": 9477 }, { "epoch": 1.0275368603642672, "grad_norm": 0.11074624955654144, "learning_rate": 6.598723342521399e-05, "loss": 0.0104, "step": 9478 }, { "epoch": 1.027645273200347, "grad_norm": 0.16439172625541687, "learning_rate": 6.598360655737705e-05, "loss": 0.0115, "step": 9479 }, { "epoch": 1.0277536860364267, "grad_norm": 0.25661569833755493, "learning_rate": 6.597997968954012e-05, "loss": 0.0183, "step": 9480 }, { "epoch": 1.0278620988725065, "grad_norm": 0.11376500129699707, "learning_rate": 6.597635282170318e-05, "loss": 0.0104, "step": 9481 }, { "epoch": 1.0279705117085862, "grad_norm": 0.15322113037109375, "learning_rate": 6.597272595386625e-05, "loss": 0.0099, "step": 9482 }, { "epoch": 1.028078924544666, "grad_norm": 0.11462398618459702, "learning_rate": 6.59690990860293e-05, "loss": 0.014, "step": 9483 }, { "epoch": 1.028187337380746, "grad_norm": 0.3517879247665405, "learning_rate": 6.596547221819237e-05, "loss": 0.0286, "step": 9484 }, { "epoch": 1.0282957502168257, "grad_norm": 0.5334367752075195, "learning_rate": 6.596184535035543e-05, "loss": 0.0143, "step": 9485 }, { "epoch": 1.0284041630529055, "grad_norm": 0.6947367191314697, "learning_rate": 6.59582184825185e-05, "loss": 0.0338, "step": 9486 }, { "epoch": 1.0285125758889853, "grad_norm": 0.13471254706382751, "learning_rate": 6.595459161468156e-05, "loss": 0.0144, "step": 9487 }, { "epoch": 1.028620988725065, "grad_norm": 0.5412716269493103, "learning_rate": 6.595096474684462e-05, "loss": 0.0466, "step": 9488 }, { "epoch": 1.0287294015611448, "grad_norm": 0.30122074484825134, "learning_rate": 6.594733787900769e-05, "loss": 0.0249, "step": 9489 }, { "epoch": 1.0288378143972245, "grad_norm": 0.5805344581604004, "learning_rate": 6.594371101117076e-05, "loss": 0.0788, "step": 9490 }, { "epoch": 1.0289462272333045, "grad_norm": 0.3946979343891144, "learning_rate": 6.594008414333383e-05, "loss": 0.0683, "step": 9491 }, { "epoch": 1.0290546400693843, "grad_norm": 0.37774622440338135, "learning_rate": 6.593645727549689e-05, "loss": 0.0466, "step": 9492 }, { "epoch": 1.029163052905464, "grad_norm": 1.0941725969314575, "learning_rate": 6.593283040765996e-05, "loss": 0.034, "step": 9493 }, { "epoch": 1.0292714657415438, "grad_norm": 0.45746123790740967, "learning_rate": 6.592920353982302e-05, "loss": 0.0199, "step": 9494 }, { "epoch": 1.0293798785776236, "grad_norm": 0.5640988945960999, "learning_rate": 6.592557667198607e-05, "loss": 0.0299, "step": 9495 }, { "epoch": 1.0294882914137033, "grad_norm": 0.10494107007980347, "learning_rate": 6.592194980414914e-05, "loss": 0.0079, "step": 9496 }, { "epoch": 1.0295967042497831, "grad_norm": 0.17646746337413788, "learning_rate": 6.59183229363122e-05, "loss": 0.013, "step": 9497 }, { "epoch": 1.0297051170858629, "grad_norm": 0.3273676335811615, "learning_rate": 6.591469606847527e-05, "loss": 0.0189, "step": 9498 }, { "epoch": 1.0298135299219429, "grad_norm": 0.19992582499980927, "learning_rate": 6.591106920063833e-05, "loss": 0.0156, "step": 9499 }, { "epoch": 1.0299219427580226, "grad_norm": 0.19805295765399933, "learning_rate": 6.59074423328014e-05, "loss": 0.0035, "step": 9500 }, { "epoch": 1.0300303555941024, "grad_norm": 0.29357877373695374, "learning_rate": 6.590381546496446e-05, "loss": 0.0181, "step": 9501 }, { "epoch": 1.0301387684301822, "grad_norm": 0.11039985716342926, "learning_rate": 6.590018859712751e-05, "loss": 0.0082, "step": 9502 }, { "epoch": 1.030247181266262, "grad_norm": 0.19378222525119781, "learning_rate": 6.589656172929059e-05, "loss": 0.0275, "step": 9503 }, { "epoch": 1.0303555941023417, "grad_norm": 0.4799933135509491, "learning_rate": 6.589293486145364e-05, "loss": 0.0339, "step": 9504 }, { "epoch": 1.0304640069384214, "grad_norm": 0.07843973487615585, "learning_rate": 6.588930799361671e-05, "loss": 0.0054, "step": 9505 }, { "epoch": 1.0305724197745012, "grad_norm": 0.09924367070198059, "learning_rate": 6.588568112577977e-05, "loss": 0.0083, "step": 9506 }, { "epoch": 1.0306808326105812, "grad_norm": 0.1849704384803772, "learning_rate": 6.588205425794286e-05, "loss": 0.0127, "step": 9507 }, { "epoch": 1.030789245446661, "grad_norm": 0.8704782128334045, "learning_rate": 6.587842739010591e-05, "loss": 0.0363, "step": 9508 }, { "epoch": 1.0308976582827407, "grad_norm": 0.500392496585846, "learning_rate": 6.587480052226897e-05, "loss": 0.0252, "step": 9509 }, { "epoch": 1.0310060711188205, "grad_norm": 0.42365819215774536, "learning_rate": 6.587117365443204e-05, "loss": 0.0116, "step": 9510 }, { "epoch": 1.0311144839549002, "grad_norm": 0.2868536710739136, "learning_rate": 6.58675467865951e-05, "loss": 0.0315, "step": 9511 }, { "epoch": 1.03122289679098, "grad_norm": 0.243547260761261, "learning_rate": 6.586391991875817e-05, "loss": 0.013, "step": 9512 }, { "epoch": 1.0313313096270598, "grad_norm": 0.15861530601978302, "learning_rate": 6.586029305092123e-05, "loss": 0.0156, "step": 9513 }, { "epoch": 1.0314397224631395, "grad_norm": 0.3420006036758423, "learning_rate": 6.58566661830843e-05, "loss": 0.0276, "step": 9514 }, { "epoch": 1.0315481352992195, "grad_norm": 0.2508895993232727, "learning_rate": 6.585303931524735e-05, "loss": 0.0187, "step": 9515 }, { "epoch": 1.0316565481352993, "grad_norm": 0.28295621275901794, "learning_rate": 6.584941244741043e-05, "loss": 0.0185, "step": 9516 }, { "epoch": 1.031764960971379, "grad_norm": 0.13655109703540802, "learning_rate": 6.584578557957348e-05, "loss": 0.0058, "step": 9517 }, { "epoch": 1.0318733738074588, "grad_norm": 0.03678971529006958, "learning_rate": 6.584215871173654e-05, "loss": 0.0035, "step": 9518 }, { "epoch": 1.0319817866435386, "grad_norm": 0.1545928716659546, "learning_rate": 6.583853184389961e-05, "loss": 0.0127, "step": 9519 }, { "epoch": 1.0320901994796183, "grad_norm": 0.08325203508138657, "learning_rate": 6.583490497606267e-05, "loss": 0.0038, "step": 9520 }, { "epoch": 1.032198612315698, "grad_norm": 0.43636882305145264, "learning_rate": 6.583127810822574e-05, "loss": 0.0365, "step": 9521 }, { "epoch": 1.0323070251517779, "grad_norm": 0.15661822259426117, "learning_rate": 6.58276512403888e-05, "loss": 0.0112, "step": 9522 }, { "epoch": 1.0324154379878578, "grad_norm": 0.1855086386203766, "learning_rate": 6.582402437255187e-05, "loss": 0.011, "step": 9523 }, { "epoch": 1.0325238508239376, "grad_norm": 0.4226362407207489, "learning_rate": 6.582039750471494e-05, "loss": 0.0254, "step": 9524 }, { "epoch": 1.0326322636600174, "grad_norm": 0.12342389672994614, "learning_rate": 6.5816770636878e-05, "loss": 0.0117, "step": 9525 }, { "epoch": 1.0327406764960971, "grad_norm": 0.44331011176109314, "learning_rate": 6.581314376904107e-05, "loss": 0.0423, "step": 9526 }, { "epoch": 1.032849089332177, "grad_norm": 0.22783693671226501, "learning_rate": 6.580951690120412e-05, "loss": 0.0171, "step": 9527 }, { "epoch": 1.0329575021682567, "grad_norm": 0.04566255584359169, "learning_rate": 6.58058900333672e-05, "loss": 0.0028, "step": 9528 }, { "epoch": 1.0330659150043364, "grad_norm": 0.22633586823940277, "learning_rate": 6.580226316553025e-05, "loss": 0.0302, "step": 9529 }, { "epoch": 1.0331743278404164, "grad_norm": 0.28473830223083496, "learning_rate": 6.579863629769332e-05, "loss": 0.0466, "step": 9530 }, { "epoch": 1.0332827406764962, "grad_norm": 0.1397046148777008, "learning_rate": 6.579500942985638e-05, "loss": 0.0069, "step": 9531 }, { "epoch": 1.033391153512576, "grad_norm": 0.30700626969337463, "learning_rate": 6.579138256201944e-05, "loss": 0.0557, "step": 9532 }, { "epoch": 1.0334995663486557, "grad_norm": 0.6069961190223694, "learning_rate": 6.578775569418251e-05, "loss": 0.0299, "step": 9533 }, { "epoch": 1.0336079791847355, "grad_norm": 0.22123217582702637, "learning_rate": 6.578412882634557e-05, "loss": 0.0083, "step": 9534 }, { "epoch": 1.0337163920208152, "grad_norm": 0.7731168270111084, "learning_rate": 6.578050195850864e-05, "loss": 0.0433, "step": 9535 }, { "epoch": 1.033824804856895, "grad_norm": 0.23977257311344147, "learning_rate": 6.57768750906717e-05, "loss": 0.0294, "step": 9536 }, { "epoch": 1.0339332176929747, "grad_norm": 0.06895934790372849, "learning_rate": 6.577324822283477e-05, "loss": 0.0047, "step": 9537 }, { "epoch": 1.0340416305290547, "grad_norm": 0.3843613862991333, "learning_rate": 6.576962135499782e-05, "loss": 0.0564, "step": 9538 }, { "epoch": 1.0341500433651345, "grad_norm": 0.4218132793903351, "learning_rate": 6.57659944871609e-05, "loss": 0.0192, "step": 9539 }, { "epoch": 1.0342584562012143, "grad_norm": 0.034053247421979904, "learning_rate": 6.576236761932395e-05, "loss": 0.0035, "step": 9540 }, { "epoch": 1.034366869037294, "grad_norm": 0.7763137221336365, "learning_rate": 6.575874075148701e-05, "loss": 0.0329, "step": 9541 }, { "epoch": 1.0344752818733738, "grad_norm": 0.5804670453071594, "learning_rate": 6.575511388365009e-05, "loss": 0.0438, "step": 9542 }, { "epoch": 1.0345836947094535, "grad_norm": 0.11540263146162033, "learning_rate": 6.575148701581315e-05, "loss": 0.007, "step": 9543 }, { "epoch": 1.0346921075455333, "grad_norm": 0.4980629086494446, "learning_rate": 6.574786014797622e-05, "loss": 0.0518, "step": 9544 }, { "epoch": 1.034800520381613, "grad_norm": 0.18395531177520752, "learning_rate": 6.574423328013928e-05, "loss": 0.0049, "step": 9545 }, { "epoch": 1.034908933217693, "grad_norm": 0.5896462798118591, "learning_rate": 6.574060641230234e-05, "loss": 0.0271, "step": 9546 }, { "epoch": 1.0350173460537728, "grad_norm": 0.3191685378551483, "learning_rate": 6.57369795444654e-05, "loss": 0.0173, "step": 9547 }, { "epoch": 1.0351257588898526, "grad_norm": 0.15601511299610138, "learning_rate": 6.573335267662846e-05, "loss": 0.005, "step": 9548 }, { "epoch": 1.0352341717259324, "grad_norm": 0.24983595311641693, "learning_rate": 6.572972580879153e-05, "loss": 0.0295, "step": 9549 }, { "epoch": 1.0353425845620121, "grad_norm": 0.2776007056236267, "learning_rate": 6.572609894095459e-05, "loss": 0.0242, "step": 9550 }, { "epoch": 1.0354509973980919, "grad_norm": 0.358669638633728, "learning_rate": 6.572247207311766e-05, "loss": 0.0294, "step": 9551 }, { "epoch": 1.0355594102341716, "grad_norm": 1.5324387550354004, "learning_rate": 6.571884520528072e-05, "loss": 0.0281, "step": 9552 }, { "epoch": 1.0356678230702516, "grad_norm": 0.11251472681760788, "learning_rate": 6.571521833744379e-05, "loss": 0.0082, "step": 9553 }, { "epoch": 1.0357762359063314, "grad_norm": 0.057261284440755844, "learning_rate": 6.571159146960685e-05, "loss": 0.0041, "step": 9554 }, { "epoch": 1.0358846487424112, "grad_norm": 0.383617639541626, "learning_rate": 6.57079646017699e-05, "loss": 0.0197, "step": 9555 }, { "epoch": 1.035993061578491, "grad_norm": 0.37505319714546204, "learning_rate": 6.570433773393298e-05, "loss": 0.0218, "step": 9556 }, { "epoch": 1.0361014744145707, "grad_norm": 0.21497945487499237, "learning_rate": 6.570071086609603e-05, "loss": 0.0242, "step": 9557 }, { "epoch": 1.0362098872506504, "grad_norm": 0.31611326336860657, "learning_rate": 6.56970839982591e-05, "loss": 0.0309, "step": 9558 }, { "epoch": 1.0363183000867302, "grad_norm": 0.1038559228181839, "learning_rate": 6.569345713042218e-05, "loss": 0.0057, "step": 9559 }, { "epoch": 1.03642671292281, "grad_norm": 0.15867938101291656, "learning_rate": 6.568983026258523e-05, "loss": 0.0126, "step": 9560 }, { "epoch": 1.03653512575889, "grad_norm": 0.1606023907661438, "learning_rate": 6.56862033947483e-05, "loss": 0.0084, "step": 9561 }, { "epoch": 1.0366435385949697, "grad_norm": 0.4156319797039032, "learning_rate": 6.568257652691136e-05, "loss": 0.0079, "step": 9562 }, { "epoch": 1.0367519514310495, "grad_norm": 0.22599303722381592, "learning_rate": 6.567894965907443e-05, "loss": 0.0122, "step": 9563 }, { "epoch": 1.0368603642671292, "grad_norm": 0.49372580647468567, "learning_rate": 6.567532279123749e-05, "loss": 0.0356, "step": 9564 }, { "epoch": 1.036968777103209, "grad_norm": 0.3551216423511505, "learning_rate": 6.567169592340056e-05, "loss": 0.0142, "step": 9565 }, { "epoch": 1.0370771899392888, "grad_norm": 0.27045392990112305, "learning_rate": 6.566806905556362e-05, "loss": 0.0209, "step": 9566 }, { "epoch": 1.0371856027753685, "grad_norm": 0.4141976237297058, "learning_rate": 6.566444218772669e-05, "loss": 0.0254, "step": 9567 }, { "epoch": 1.0372940156114483, "grad_norm": 0.05831276252865791, "learning_rate": 6.566081531988975e-05, "loss": 0.0027, "step": 9568 }, { "epoch": 1.0374024284475283, "grad_norm": 0.2969376742839813, "learning_rate": 6.56571884520528e-05, "loss": 0.025, "step": 9569 }, { "epoch": 1.037510841283608, "grad_norm": 0.22483111917972565, "learning_rate": 6.565356158421587e-05, "loss": 0.0044, "step": 9570 }, { "epoch": 1.0376192541196878, "grad_norm": 1.0136620998382568, "learning_rate": 6.564993471637893e-05, "loss": 0.0453, "step": 9571 }, { "epoch": 1.0377276669557676, "grad_norm": 0.292003870010376, "learning_rate": 6.5646307848542e-05, "loss": 0.0097, "step": 9572 }, { "epoch": 1.0378360797918473, "grad_norm": 0.5662567019462585, "learning_rate": 6.564268098070506e-05, "loss": 0.0207, "step": 9573 }, { "epoch": 1.037944492627927, "grad_norm": 0.8665919899940491, "learning_rate": 6.563905411286813e-05, "loss": 0.0463, "step": 9574 }, { "epoch": 1.0380529054640069, "grad_norm": 0.17412632703781128, "learning_rate": 6.563542724503119e-05, "loss": 0.0183, "step": 9575 }, { "epoch": 1.0381613183000868, "grad_norm": 0.5779862999916077, "learning_rate": 6.563180037719426e-05, "loss": 0.014, "step": 9576 }, { "epoch": 1.0382697311361666, "grad_norm": 0.0665341168642044, "learning_rate": 6.562817350935733e-05, "loss": 0.0066, "step": 9577 }, { "epoch": 1.0383781439722464, "grad_norm": 0.10410023480653763, "learning_rate": 6.562454664152039e-05, "loss": 0.0072, "step": 9578 }, { "epoch": 1.0384865568083261, "grad_norm": 0.2287479192018509, "learning_rate": 6.562091977368346e-05, "loss": 0.0136, "step": 9579 }, { "epoch": 1.038594969644406, "grad_norm": 0.41785845160484314, "learning_rate": 6.561729290584651e-05, "loss": 0.0348, "step": 9580 }, { "epoch": 1.0387033824804857, "grad_norm": 0.05632738023996353, "learning_rate": 6.561366603800959e-05, "loss": 0.0026, "step": 9581 }, { "epoch": 1.0388117953165654, "grad_norm": 0.3167613446712494, "learning_rate": 6.561003917017264e-05, "loss": 0.0127, "step": 9582 }, { "epoch": 1.0389202081526452, "grad_norm": 0.17406927049160004, "learning_rate": 6.56064123023357e-05, "loss": 0.0109, "step": 9583 }, { "epoch": 1.0390286209887252, "grad_norm": 0.2863309383392334, "learning_rate": 6.560278543449877e-05, "loss": 0.0205, "step": 9584 }, { "epoch": 1.039137033824805, "grad_norm": 0.4587886333465576, "learning_rate": 6.559915856666183e-05, "loss": 0.0195, "step": 9585 }, { "epoch": 1.0392454466608847, "grad_norm": 0.2594654858112335, "learning_rate": 6.55955316988249e-05, "loss": 0.0222, "step": 9586 }, { "epoch": 1.0393538594969645, "grad_norm": 0.6089381575584412, "learning_rate": 6.559190483098796e-05, "loss": 0.0488, "step": 9587 }, { "epoch": 1.0394622723330442, "grad_norm": 0.40573254227638245, "learning_rate": 6.558827796315103e-05, "loss": 0.0247, "step": 9588 }, { "epoch": 1.039570685169124, "grad_norm": 0.4728049039840698, "learning_rate": 6.558465109531408e-05, "loss": 0.0222, "step": 9589 }, { "epoch": 1.0396790980052038, "grad_norm": 0.14649134874343872, "learning_rate": 6.558102422747716e-05, "loss": 0.0156, "step": 9590 }, { "epoch": 1.0397875108412835, "grad_norm": 0.30889692902565, "learning_rate": 6.557739735964021e-05, "loss": 0.0215, "step": 9591 }, { "epoch": 1.0398959236773635, "grad_norm": 0.11537975817918777, "learning_rate": 6.557377049180327e-05, "loss": 0.0057, "step": 9592 }, { "epoch": 1.0400043365134433, "grad_norm": 0.6968076229095459, "learning_rate": 6.557014362396635e-05, "loss": 0.0633, "step": 9593 }, { "epoch": 1.040112749349523, "grad_norm": 0.4716200828552246, "learning_rate": 6.556651675612941e-05, "loss": 0.0232, "step": 9594 }, { "epoch": 1.0402211621856028, "grad_norm": 0.5916711091995239, "learning_rate": 6.556288988829248e-05, "loss": 0.0254, "step": 9595 }, { "epoch": 1.0403295750216826, "grad_norm": 0.09389752894639969, "learning_rate": 6.555926302045554e-05, "loss": 0.0068, "step": 9596 }, { "epoch": 1.0404379878577623, "grad_norm": 0.5166097283363342, "learning_rate": 6.555563615261861e-05, "loss": 0.0511, "step": 9597 }, { "epoch": 1.040546400693842, "grad_norm": 0.20801134407520294, "learning_rate": 6.555200928478167e-05, "loss": 0.0095, "step": 9598 }, { "epoch": 1.0406548135299218, "grad_norm": 0.7797607183456421, "learning_rate": 6.554838241694473e-05, "loss": 0.0828, "step": 9599 }, { "epoch": 1.0407632263660018, "grad_norm": 0.06644993275403976, "learning_rate": 6.55447555491078e-05, "loss": 0.0022, "step": 9600 }, { "epoch": 1.0408716392020816, "grad_norm": 0.23420989513397217, "learning_rate": 6.554112868127085e-05, "loss": 0.0119, "step": 9601 }, { "epoch": 1.0409800520381614, "grad_norm": 0.09410669654607773, "learning_rate": 6.553750181343393e-05, "loss": 0.01, "step": 9602 }, { "epoch": 1.0410884648742411, "grad_norm": 0.2840609550476074, "learning_rate": 6.553387494559698e-05, "loss": 0.0133, "step": 9603 }, { "epoch": 1.0411968777103209, "grad_norm": 0.20711061358451843, "learning_rate": 6.553024807776005e-05, "loss": 0.0207, "step": 9604 }, { "epoch": 1.0413052905464006, "grad_norm": 0.3887357711791992, "learning_rate": 6.552662120992311e-05, "loss": 0.035, "step": 9605 }, { "epoch": 1.0414137033824804, "grad_norm": 0.3031665086746216, "learning_rate": 6.552299434208617e-05, "loss": 0.023, "step": 9606 }, { "epoch": 1.0415221162185602, "grad_norm": 0.5297395586967468, "learning_rate": 6.551936747424924e-05, "loss": 0.0102, "step": 9607 }, { "epoch": 1.0416305290546402, "grad_norm": 0.19693022966384888, "learning_rate": 6.55157406064123e-05, "loss": 0.0151, "step": 9608 }, { "epoch": 1.04173894189072, "grad_norm": 0.22205528616905212, "learning_rate": 6.551211373857537e-05, "loss": 0.0073, "step": 9609 }, { "epoch": 1.0418473547267997, "grad_norm": 0.23657533526420593, "learning_rate": 6.550848687073842e-05, "loss": 0.0199, "step": 9610 }, { "epoch": 1.0419557675628794, "grad_norm": 0.34648025035858154, "learning_rate": 6.550486000290151e-05, "loss": 0.0216, "step": 9611 }, { "epoch": 1.0420641803989592, "grad_norm": 0.1898803561925888, "learning_rate": 6.550123313506457e-05, "loss": 0.0086, "step": 9612 }, { "epoch": 1.042172593235039, "grad_norm": 0.22346480190753937, "learning_rate": 6.549760626722762e-05, "loss": 0.0222, "step": 9613 }, { "epoch": 1.0422810060711187, "grad_norm": 0.6691129803657532, "learning_rate": 6.54939793993907e-05, "loss": 0.0411, "step": 9614 }, { "epoch": 1.0423894189071987, "grad_norm": 0.21390585601329803, "learning_rate": 6.549035253155375e-05, "loss": 0.0227, "step": 9615 }, { "epoch": 1.0424978317432785, "grad_norm": 0.4978441596031189, "learning_rate": 6.548672566371682e-05, "loss": 0.0448, "step": 9616 }, { "epoch": 1.0426062445793582, "grad_norm": 0.6912100315093994, "learning_rate": 6.548309879587988e-05, "loss": 0.0534, "step": 9617 }, { "epoch": 1.042714657415438, "grad_norm": 0.3696844279766083, "learning_rate": 6.547947192804295e-05, "loss": 0.024, "step": 9618 }, { "epoch": 1.0428230702515178, "grad_norm": 0.14199689030647278, "learning_rate": 6.547584506020601e-05, "loss": 0.0065, "step": 9619 }, { "epoch": 1.0429314830875975, "grad_norm": 0.31375613808631897, "learning_rate": 6.547221819236908e-05, "loss": 0.0262, "step": 9620 }, { "epoch": 1.0430398959236773, "grad_norm": 0.630430281162262, "learning_rate": 6.546859132453214e-05, "loss": 0.0642, "step": 9621 }, { "epoch": 1.043148308759757, "grad_norm": 0.4996170401573181, "learning_rate": 6.54649644566952e-05, "loss": 0.0324, "step": 9622 }, { "epoch": 1.043256721595837, "grad_norm": 0.5143294334411621, "learning_rate": 6.546133758885826e-05, "loss": 0.0572, "step": 9623 }, { "epoch": 1.0433651344319168, "grad_norm": 0.23921051621437073, "learning_rate": 6.545771072102132e-05, "loss": 0.0096, "step": 9624 }, { "epoch": 1.0434735472679966, "grad_norm": 0.2726232707500458, "learning_rate": 6.545408385318439e-05, "loss": 0.0093, "step": 9625 }, { "epoch": 1.0435819601040763, "grad_norm": 0.5382834672927856, "learning_rate": 6.545045698534745e-05, "loss": 0.082, "step": 9626 }, { "epoch": 1.043690372940156, "grad_norm": 0.20619438588619232, "learning_rate": 6.544683011751052e-05, "loss": 0.0121, "step": 9627 }, { "epoch": 1.0437987857762359, "grad_norm": 0.3902222514152527, "learning_rate": 6.544320324967359e-05, "loss": 0.0302, "step": 9628 }, { "epoch": 1.0439071986123156, "grad_norm": 0.36164599657058716, "learning_rate": 6.543957638183665e-05, "loss": 0.0288, "step": 9629 }, { "epoch": 1.0440156114483954, "grad_norm": 0.7217578291893005, "learning_rate": 6.543594951399972e-05, "loss": 0.0205, "step": 9630 }, { "epoch": 1.0441240242844754, "grad_norm": 0.4284552037715912, "learning_rate": 6.543232264616278e-05, "loss": 0.0259, "step": 9631 }, { "epoch": 1.0442324371205551, "grad_norm": 0.49621278047561646, "learning_rate": 6.542869577832585e-05, "loss": 0.0284, "step": 9632 }, { "epoch": 1.044340849956635, "grad_norm": 0.08186179399490356, "learning_rate": 6.54250689104889e-05, "loss": 0.0051, "step": 9633 }, { "epoch": 1.0444492627927147, "grad_norm": 0.05443229526281357, "learning_rate": 6.542144204265198e-05, "loss": 0.0022, "step": 9634 }, { "epoch": 1.0445576756287944, "grad_norm": 0.2273397594690323, "learning_rate": 6.541781517481503e-05, "loss": 0.0397, "step": 9635 }, { "epoch": 1.0446660884648742, "grad_norm": 0.06066218391060829, "learning_rate": 6.541418830697809e-05, "loss": 0.0057, "step": 9636 }, { "epoch": 1.044774501300954, "grad_norm": 0.19624961912631989, "learning_rate": 6.541056143914116e-05, "loss": 0.0125, "step": 9637 }, { "epoch": 1.044882914137034, "grad_norm": 0.15845392644405365, "learning_rate": 6.540693457130422e-05, "loss": 0.0264, "step": 9638 }, { "epoch": 1.0449913269731137, "grad_norm": 0.229770228266716, "learning_rate": 6.540330770346729e-05, "loss": 0.0153, "step": 9639 }, { "epoch": 1.0450997398091935, "grad_norm": 0.36988675594329834, "learning_rate": 6.539968083563035e-05, "loss": 0.0512, "step": 9640 }, { "epoch": 1.0452081526452732, "grad_norm": 0.2203337550163269, "learning_rate": 6.539605396779342e-05, "loss": 0.0142, "step": 9641 }, { "epoch": 1.045316565481353, "grad_norm": 0.16884225606918335, "learning_rate": 6.539242709995648e-05, "loss": 0.012, "step": 9642 }, { "epoch": 1.0454249783174328, "grad_norm": 0.46052035689353943, "learning_rate": 6.538880023211955e-05, "loss": 0.0135, "step": 9643 }, { "epoch": 1.0455333911535125, "grad_norm": 0.4342639744281769, "learning_rate": 6.53851733642826e-05, "loss": 0.0309, "step": 9644 }, { "epoch": 1.0456418039895923, "grad_norm": 0.22907666862010956, "learning_rate": 6.538154649644567e-05, "loss": 0.0151, "step": 9645 }, { "epoch": 1.0457502168256723, "grad_norm": 0.39838141202926636, "learning_rate": 6.537791962860875e-05, "loss": 0.0112, "step": 9646 }, { "epoch": 1.045858629661752, "grad_norm": 0.1338493525981903, "learning_rate": 6.53742927607718e-05, "loss": 0.011, "step": 9647 }, { "epoch": 1.0459670424978318, "grad_norm": 0.6998575329780579, "learning_rate": 6.537066589293487e-05, "loss": 0.0431, "step": 9648 }, { "epoch": 1.0460754553339116, "grad_norm": 0.553726851940155, "learning_rate": 6.536703902509793e-05, "loss": 0.0308, "step": 9649 }, { "epoch": 1.0461838681699913, "grad_norm": 0.13734295964241028, "learning_rate": 6.536341215726099e-05, "loss": 0.0106, "step": 9650 }, { "epoch": 1.046292281006071, "grad_norm": 0.17021673917770386, "learning_rate": 6.535978528942406e-05, "loss": 0.0144, "step": 9651 }, { "epoch": 1.0464006938421508, "grad_norm": 0.25606003403663635, "learning_rate": 6.535615842158712e-05, "loss": 0.0372, "step": 9652 }, { "epoch": 1.0465091066782306, "grad_norm": 0.11252360790967941, "learning_rate": 6.535253155375019e-05, "loss": 0.0091, "step": 9653 }, { "epoch": 1.0466175195143106, "grad_norm": 0.11779748648405075, "learning_rate": 6.534890468591325e-05, "loss": 0.0115, "step": 9654 }, { "epoch": 1.0467259323503904, "grad_norm": 0.22491316497325897, "learning_rate": 6.534527781807632e-05, "loss": 0.0194, "step": 9655 }, { "epoch": 1.0468343451864701, "grad_norm": 0.27369552850723267, "learning_rate": 6.534165095023937e-05, "loss": 0.0178, "step": 9656 }, { "epoch": 1.0469427580225499, "grad_norm": 0.34595805406570435, "learning_rate": 6.533802408240244e-05, "loss": 0.0352, "step": 9657 }, { "epoch": 1.0470511708586296, "grad_norm": 0.6102505922317505, "learning_rate": 6.53343972145655e-05, "loss": 0.0196, "step": 9658 }, { "epoch": 1.0471595836947094, "grad_norm": 0.3399290144443512, "learning_rate": 6.533077034672856e-05, "loss": 0.0689, "step": 9659 }, { "epoch": 1.0472679965307892, "grad_norm": 0.16657286882400513, "learning_rate": 6.532714347889163e-05, "loss": 0.0171, "step": 9660 }, { "epoch": 1.047376409366869, "grad_norm": 0.36694401502609253, "learning_rate": 6.532351661105469e-05, "loss": 0.029, "step": 9661 }, { "epoch": 1.047484822202949, "grad_norm": 0.19914834201335907, "learning_rate": 6.531988974321776e-05, "loss": 0.0132, "step": 9662 }, { "epoch": 1.0475932350390287, "grad_norm": 0.4117288291454315, "learning_rate": 6.531626287538083e-05, "loss": 0.0616, "step": 9663 }, { "epoch": 1.0477016478751084, "grad_norm": 0.2835525572299957, "learning_rate": 6.531263600754389e-05, "loss": 0.0246, "step": 9664 }, { "epoch": 1.0478100607111882, "grad_norm": 0.07457765936851501, "learning_rate": 6.530900913970696e-05, "loss": 0.0071, "step": 9665 }, { "epoch": 1.047918473547268, "grad_norm": 0.4821716248989105, "learning_rate": 6.530538227187001e-05, "loss": 0.0121, "step": 9666 }, { "epoch": 1.0480268863833477, "grad_norm": 0.4805266857147217, "learning_rate": 6.530175540403309e-05, "loss": 0.0225, "step": 9667 }, { "epoch": 1.0481352992194275, "grad_norm": 0.2527020275592804, "learning_rate": 6.529812853619614e-05, "loss": 0.0273, "step": 9668 }, { "epoch": 1.0482437120555073, "grad_norm": 0.5763899087905884, "learning_rate": 6.529450166835921e-05, "loss": 0.0789, "step": 9669 }, { "epoch": 1.0483521248915872, "grad_norm": 0.2822539508342743, "learning_rate": 6.529087480052227e-05, "loss": 0.0214, "step": 9670 }, { "epoch": 1.048460537727667, "grad_norm": 0.2361062616109848, "learning_rate": 6.528724793268534e-05, "loss": 0.0067, "step": 9671 }, { "epoch": 1.0485689505637468, "grad_norm": 0.31591796875, "learning_rate": 6.52836210648484e-05, "loss": 0.0451, "step": 9672 }, { "epoch": 1.0486773633998265, "grad_norm": 0.24330571293830872, "learning_rate": 6.527999419701146e-05, "loss": 0.0541, "step": 9673 }, { "epoch": 1.0487857762359063, "grad_norm": 0.1624186933040619, "learning_rate": 6.527636732917453e-05, "loss": 0.008, "step": 9674 }, { "epoch": 1.048894189071986, "grad_norm": 0.23736348748207092, "learning_rate": 6.527274046133758e-05, "loss": 0.0236, "step": 9675 }, { "epoch": 1.0490026019080658, "grad_norm": 0.6571629643440247, "learning_rate": 6.526911359350066e-05, "loss": 0.0868, "step": 9676 }, { "epoch": 1.0491110147441458, "grad_norm": 0.34840506315231323, "learning_rate": 6.526548672566371e-05, "loss": 0.0401, "step": 9677 }, { "epoch": 1.0492194275802256, "grad_norm": 0.057226475328207016, "learning_rate": 6.526185985782678e-05, "loss": 0.0029, "step": 9678 }, { "epoch": 1.0493278404163053, "grad_norm": 0.052221231162548065, "learning_rate": 6.525823298998984e-05, "loss": 0.0036, "step": 9679 }, { "epoch": 1.049436253252385, "grad_norm": 0.3689563274383545, "learning_rate": 6.525460612215291e-05, "loss": 0.0369, "step": 9680 }, { "epoch": 1.0495446660884649, "grad_norm": 0.17955733835697174, "learning_rate": 6.525097925431598e-05, "loss": 0.016, "step": 9681 }, { "epoch": 1.0496530789245446, "grad_norm": 0.13632555305957794, "learning_rate": 6.524735238647904e-05, "loss": 0.0167, "step": 9682 }, { "epoch": 1.0497614917606244, "grad_norm": 0.23274384438991547, "learning_rate": 6.524372551864211e-05, "loss": 0.0198, "step": 9683 }, { "epoch": 1.0498699045967042, "grad_norm": 0.39970308542251587, "learning_rate": 6.524009865080517e-05, "loss": 0.0235, "step": 9684 }, { "epoch": 1.0499783174327841, "grad_norm": 0.3281114995479584, "learning_rate": 6.523647178296824e-05, "loss": 0.0171, "step": 9685 }, { "epoch": 1.050086730268864, "grad_norm": 0.19392643868923187, "learning_rate": 6.52328449151313e-05, "loss": 0.0087, "step": 9686 }, { "epoch": 1.0501951431049437, "grad_norm": 0.3501998484134674, "learning_rate": 6.522921804729435e-05, "loss": 0.0414, "step": 9687 }, { "epoch": 1.0503035559410234, "grad_norm": 0.1348429173231125, "learning_rate": 6.522559117945742e-05, "loss": 0.0169, "step": 9688 }, { "epoch": 1.0504119687771032, "grad_norm": 0.22281232476234436, "learning_rate": 6.522196431162048e-05, "loss": 0.0202, "step": 9689 }, { "epoch": 1.050520381613183, "grad_norm": 0.059866081923246384, "learning_rate": 6.521833744378355e-05, "loss": 0.0047, "step": 9690 }, { "epoch": 1.0506287944492627, "grad_norm": 0.06309409439563751, "learning_rate": 6.521471057594661e-05, "loss": 0.0035, "step": 9691 }, { "epoch": 1.0507372072853425, "grad_norm": 0.22365961968898773, "learning_rate": 6.521108370810968e-05, "loss": 0.0188, "step": 9692 }, { "epoch": 1.0508456201214225, "grad_norm": 0.39402851462364197, "learning_rate": 6.520745684027274e-05, "loss": 0.0292, "step": 9693 }, { "epoch": 1.0509540329575022, "grad_norm": 0.44179290533065796, "learning_rate": 6.520382997243581e-05, "loss": 0.0573, "step": 9694 }, { "epoch": 1.051062445793582, "grad_norm": 0.5551711320877075, "learning_rate": 6.520020310459887e-05, "loss": 0.0429, "step": 9695 }, { "epoch": 1.0511708586296618, "grad_norm": 0.43623480200767517, "learning_rate": 6.519657623676192e-05, "loss": 0.0379, "step": 9696 }, { "epoch": 1.0512792714657415, "grad_norm": 0.27780479192733765, "learning_rate": 6.519294936892501e-05, "loss": 0.0221, "step": 9697 }, { "epoch": 1.0513876843018213, "grad_norm": 0.14023828506469727, "learning_rate": 6.518932250108807e-05, "loss": 0.0047, "step": 9698 }, { "epoch": 1.051496097137901, "grad_norm": 0.25483182072639465, "learning_rate": 6.518569563325114e-05, "loss": 0.0042, "step": 9699 }, { "epoch": 1.051604509973981, "grad_norm": 0.36311864852905273, "learning_rate": 6.51820687654142e-05, "loss": 0.0154, "step": 9700 }, { "epoch": 1.0517129228100608, "grad_norm": 0.047276753932237625, "learning_rate": 6.517844189757726e-05, "loss": 0.0031, "step": 9701 }, { "epoch": 1.0518213356461406, "grad_norm": 0.36601582169532776, "learning_rate": 6.517481502974032e-05, "loss": 0.0452, "step": 9702 }, { "epoch": 1.0519297484822203, "grad_norm": 0.10474766045808792, "learning_rate": 6.517118816190338e-05, "loss": 0.0079, "step": 9703 }, { "epoch": 1.0520381613183, "grad_norm": 0.48075956106185913, "learning_rate": 6.516756129406645e-05, "loss": 0.0792, "step": 9704 }, { "epoch": 1.0521465741543798, "grad_norm": 0.20347553491592407, "learning_rate": 6.516393442622951e-05, "loss": 0.0228, "step": 9705 }, { "epoch": 1.0522549869904596, "grad_norm": 0.2765829861164093, "learning_rate": 6.516030755839258e-05, "loss": 0.0256, "step": 9706 }, { "epoch": 1.0523633998265394, "grad_norm": 0.12445296347141266, "learning_rate": 6.515668069055564e-05, "loss": 0.0102, "step": 9707 }, { "epoch": 1.0524718126626194, "grad_norm": 0.3410119116306305, "learning_rate": 6.51530538227187e-05, "loss": 0.0155, "step": 9708 }, { "epoch": 1.0525802254986991, "grad_norm": 0.7878500819206238, "learning_rate": 6.514942695488176e-05, "loss": 0.04, "step": 9709 }, { "epoch": 1.0526886383347789, "grad_norm": 0.4858137369155884, "learning_rate": 6.514580008704483e-05, "loss": 0.0494, "step": 9710 }, { "epoch": 1.0527970511708586, "grad_norm": 0.4216688573360443, "learning_rate": 6.514217321920789e-05, "loss": 0.0443, "step": 9711 }, { "epoch": 1.0529054640069384, "grad_norm": 0.07644622027873993, "learning_rate": 6.513854635137095e-05, "loss": 0.0064, "step": 9712 }, { "epoch": 1.0530138768430182, "grad_norm": 0.027459336444735527, "learning_rate": 6.513491948353402e-05, "loss": 0.0018, "step": 9713 }, { "epoch": 1.053122289679098, "grad_norm": 0.07712343335151672, "learning_rate": 6.513129261569709e-05, "loss": 0.0043, "step": 9714 }, { "epoch": 1.0532307025151777, "grad_norm": 0.31300193071365356, "learning_rate": 6.512766574786016e-05, "loss": 0.036, "step": 9715 }, { "epoch": 1.0533391153512577, "grad_norm": 0.439310222864151, "learning_rate": 6.512403888002322e-05, "loss": 0.0695, "step": 9716 }, { "epoch": 1.0534475281873374, "grad_norm": 0.6765933036804199, "learning_rate": 6.512041201218628e-05, "loss": 0.0675, "step": 9717 }, { "epoch": 1.0535559410234172, "grad_norm": 0.44788968563079834, "learning_rate": 6.511678514434935e-05, "loss": 0.0202, "step": 9718 }, { "epoch": 1.053664353859497, "grad_norm": 0.20650741457939148, "learning_rate": 6.51131582765124e-05, "loss": 0.0076, "step": 9719 }, { "epoch": 1.0537727666955767, "grad_norm": 0.20579449832439423, "learning_rate": 6.510953140867548e-05, "loss": 0.0197, "step": 9720 }, { "epoch": 1.0538811795316565, "grad_norm": 0.10836412012577057, "learning_rate": 6.510590454083853e-05, "loss": 0.0088, "step": 9721 }, { "epoch": 1.0539895923677363, "grad_norm": 0.257391095161438, "learning_rate": 6.51022776730016e-05, "loss": 0.0329, "step": 9722 }, { "epoch": 1.0540980052038162, "grad_norm": 0.490480899810791, "learning_rate": 6.509865080516466e-05, "loss": 0.0647, "step": 9723 }, { "epoch": 1.054206418039896, "grad_norm": 0.5007917284965515, "learning_rate": 6.509502393732773e-05, "loss": 0.0365, "step": 9724 }, { "epoch": 1.0543148308759758, "grad_norm": 0.4436313509941101, "learning_rate": 6.509139706949079e-05, "loss": 0.0287, "step": 9725 }, { "epoch": 1.0544232437120555, "grad_norm": 0.23634158074855804, "learning_rate": 6.508777020165385e-05, "loss": 0.0186, "step": 9726 }, { "epoch": 1.0545316565481353, "grad_norm": 0.9025542736053467, "learning_rate": 6.508414333381692e-05, "loss": 0.0573, "step": 9727 }, { "epoch": 1.054640069384215, "grad_norm": 0.31548139452934265, "learning_rate": 6.508051646597998e-05, "loss": 0.0157, "step": 9728 }, { "epoch": 1.0547484822202948, "grad_norm": 0.17856062948703766, "learning_rate": 6.507688959814305e-05, "loss": 0.0324, "step": 9729 }, { "epoch": 1.0548568950563746, "grad_norm": 0.4181350767612457, "learning_rate": 6.50732627303061e-05, "loss": 0.0377, "step": 9730 }, { "epoch": 1.0549653078924546, "grad_norm": 0.155633345246315, "learning_rate": 6.506963586246917e-05, "loss": 0.0104, "step": 9731 }, { "epoch": 1.0550737207285343, "grad_norm": 0.302205353975296, "learning_rate": 6.506600899463225e-05, "loss": 0.0274, "step": 9732 }, { "epoch": 1.055182133564614, "grad_norm": 0.14882728457450867, "learning_rate": 6.50623821267953e-05, "loss": 0.0202, "step": 9733 }, { "epoch": 1.0552905464006939, "grad_norm": 0.2944816052913666, "learning_rate": 6.505875525895837e-05, "loss": 0.0253, "step": 9734 }, { "epoch": 1.0553989592367736, "grad_norm": 0.29984572529792786, "learning_rate": 6.505512839112143e-05, "loss": 0.0177, "step": 9735 }, { "epoch": 1.0555073720728534, "grad_norm": 0.2919718325138092, "learning_rate": 6.50515015232845e-05, "loss": 0.0327, "step": 9736 }, { "epoch": 1.0556157849089332, "grad_norm": 0.24738676846027374, "learning_rate": 6.504787465544756e-05, "loss": 0.0219, "step": 9737 }, { "epoch": 1.055724197745013, "grad_norm": 0.5320285558700562, "learning_rate": 6.504424778761063e-05, "loss": 0.08, "step": 9738 }, { "epoch": 1.055832610581093, "grad_norm": 0.28428444266319275, "learning_rate": 6.504062091977369e-05, "loss": 0.0229, "step": 9739 }, { "epoch": 1.0559410234171727, "grad_norm": 0.4422805905342102, "learning_rate": 6.503699405193674e-05, "loss": 0.0476, "step": 9740 }, { "epoch": 1.0560494362532524, "grad_norm": 0.07865285128355026, "learning_rate": 6.503336718409982e-05, "loss": 0.0053, "step": 9741 }, { "epoch": 1.0561578490893322, "grad_norm": 0.14217230677604675, "learning_rate": 6.502974031626287e-05, "loss": 0.009, "step": 9742 }, { "epoch": 1.056266261925412, "grad_norm": 0.4539218842983246, "learning_rate": 6.502611344842594e-05, "loss": 0.0305, "step": 9743 }, { "epoch": 1.0563746747614917, "grad_norm": 0.10310878604650497, "learning_rate": 6.5022486580589e-05, "loss": 0.0076, "step": 9744 }, { "epoch": 1.0564830875975715, "grad_norm": 0.3144947290420532, "learning_rate": 6.501885971275207e-05, "loss": 0.0239, "step": 9745 }, { "epoch": 1.0565915004336512, "grad_norm": 0.20692501962184906, "learning_rate": 6.501523284491513e-05, "loss": 0.0189, "step": 9746 }, { "epoch": 1.0566999132697312, "grad_norm": 0.23274186253547668, "learning_rate": 6.50116059770782e-05, "loss": 0.0192, "step": 9747 }, { "epoch": 1.056808326105811, "grad_norm": 0.4428690969944, "learning_rate": 6.500797910924126e-05, "loss": 0.0319, "step": 9748 }, { "epoch": 1.0569167389418908, "grad_norm": 0.33020591735839844, "learning_rate": 6.500435224140433e-05, "loss": 0.0641, "step": 9749 }, { "epoch": 1.0570251517779705, "grad_norm": 0.3585101068019867, "learning_rate": 6.50007253735674e-05, "loss": 0.0386, "step": 9750 }, { "epoch": 1.0571335646140503, "grad_norm": 0.33405494689941406, "learning_rate": 6.499709850573046e-05, "loss": 0.0265, "step": 9751 }, { "epoch": 1.05724197745013, "grad_norm": 0.025610022246837616, "learning_rate": 6.499347163789353e-05, "loss": 0.0013, "step": 9752 }, { "epoch": 1.0573503902862098, "grad_norm": 0.17577265202999115, "learning_rate": 6.498984477005658e-05, "loss": 0.0135, "step": 9753 }, { "epoch": 1.0574588031222896, "grad_norm": 0.5367252826690674, "learning_rate": 6.498621790221964e-05, "loss": 0.075, "step": 9754 }, { "epoch": 1.0575672159583696, "grad_norm": 0.41398465633392334, "learning_rate": 6.498259103438271e-05, "loss": 0.0369, "step": 9755 }, { "epoch": 1.0576756287944493, "grad_norm": 0.23350071907043457, "learning_rate": 6.497896416654577e-05, "loss": 0.017, "step": 9756 }, { "epoch": 1.057784041630529, "grad_norm": 0.28312942385673523, "learning_rate": 6.497533729870884e-05, "loss": 0.0284, "step": 9757 }, { "epoch": 1.0578924544666088, "grad_norm": 0.12379775941371918, "learning_rate": 6.49717104308719e-05, "loss": 0.0133, "step": 9758 }, { "epoch": 1.0580008673026886, "grad_norm": 0.42737478017807007, "learning_rate": 6.496808356303497e-05, "loss": 0.0408, "step": 9759 }, { "epoch": 1.0581092801387684, "grad_norm": 0.21803204715251923, "learning_rate": 6.496445669519803e-05, "loss": 0.0174, "step": 9760 }, { "epoch": 1.0582176929748481, "grad_norm": 0.12923865020275116, "learning_rate": 6.49608298273611e-05, "loss": 0.0201, "step": 9761 }, { "epoch": 1.0583261058109281, "grad_norm": 0.0868540108203888, "learning_rate": 6.495720295952415e-05, "loss": 0.0094, "step": 9762 }, { "epoch": 1.0584345186470079, "grad_norm": 0.26653704047203064, "learning_rate": 6.495357609168721e-05, "loss": 0.0131, "step": 9763 }, { "epoch": 1.0585429314830876, "grad_norm": 1.737296462059021, "learning_rate": 6.494994922385028e-05, "loss": 0.0652, "step": 9764 }, { "epoch": 1.0586513443191674, "grad_norm": 0.14104317128658295, "learning_rate": 6.494632235601334e-05, "loss": 0.0125, "step": 9765 }, { "epoch": 1.0587597571552472, "grad_norm": 0.2878091335296631, "learning_rate": 6.494269548817642e-05, "loss": 0.0166, "step": 9766 }, { "epoch": 1.058868169991327, "grad_norm": 0.21358777582645416, "learning_rate": 6.493906862033948e-05, "loss": 0.0093, "step": 9767 }, { "epoch": 1.0589765828274067, "grad_norm": 0.31284210085868835, "learning_rate": 6.493544175250254e-05, "loss": 0.0223, "step": 9768 }, { "epoch": 1.0590849956634865, "grad_norm": 0.34342315793037415, "learning_rate": 6.493181488466561e-05, "loss": 0.0114, "step": 9769 }, { "epoch": 1.0591934084995664, "grad_norm": 0.07997719198465347, "learning_rate": 6.492818801682867e-05, "loss": 0.0084, "step": 9770 }, { "epoch": 1.0593018213356462, "grad_norm": 0.10689250379800797, "learning_rate": 6.492456114899174e-05, "loss": 0.0033, "step": 9771 }, { "epoch": 1.059410234171726, "grad_norm": 0.07173413783311844, "learning_rate": 6.49209342811548e-05, "loss": 0.0035, "step": 9772 }, { "epoch": 1.0595186470078057, "grad_norm": 0.3417961299419403, "learning_rate": 6.491730741331787e-05, "loss": 0.0631, "step": 9773 }, { "epoch": 1.0596270598438855, "grad_norm": 0.20604030787944794, "learning_rate": 6.491368054548092e-05, "loss": 0.0106, "step": 9774 }, { "epoch": 1.0597354726799653, "grad_norm": 1.0238049030303955, "learning_rate": 6.4910053677644e-05, "loss": 0.0323, "step": 9775 }, { "epoch": 1.059843885516045, "grad_norm": 0.6517156362533569, "learning_rate": 6.490642680980705e-05, "loss": 0.0438, "step": 9776 }, { "epoch": 1.0599522983521248, "grad_norm": 0.3853593170642853, "learning_rate": 6.490279994197011e-05, "loss": 0.0222, "step": 9777 }, { "epoch": 1.0600607111882048, "grad_norm": 0.4817078709602356, "learning_rate": 6.489917307413318e-05, "loss": 0.0425, "step": 9778 }, { "epoch": 1.0601691240242845, "grad_norm": 0.2711643874645233, "learning_rate": 6.489554620629624e-05, "loss": 0.0285, "step": 9779 }, { "epoch": 1.0602775368603643, "grad_norm": 0.19709186255931854, "learning_rate": 6.489191933845931e-05, "loss": 0.0161, "step": 9780 }, { "epoch": 1.060385949696444, "grad_norm": 0.2557564973831177, "learning_rate": 6.488829247062237e-05, "loss": 0.015, "step": 9781 }, { "epoch": 1.0604943625325238, "grad_norm": 0.08643113821744919, "learning_rate": 6.488466560278544e-05, "loss": 0.005, "step": 9782 }, { "epoch": 1.0606027753686036, "grad_norm": 0.06269516050815582, "learning_rate": 6.48810387349485e-05, "loss": 0.0037, "step": 9783 }, { "epoch": 1.0607111882046834, "grad_norm": 0.33025258779525757, "learning_rate": 6.487741186711157e-05, "loss": 0.0346, "step": 9784 }, { "epoch": 1.0608196010407633, "grad_norm": 0.2108214795589447, "learning_rate": 6.487378499927464e-05, "loss": 0.0136, "step": 9785 }, { "epoch": 1.060928013876843, "grad_norm": 0.184141606092453, "learning_rate": 6.48701581314377e-05, "loss": 0.0054, "step": 9786 }, { "epoch": 1.0610364267129229, "grad_norm": 0.056671202182769775, "learning_rate": 6.486653126360076e-05, "loss": 0.003, "step": 9787 }, { "epoch": 1.0611448395490026, "grad_norm": 0.110109843313694, "learning_rate": 6.486290439576382e-05, "loss": 0.0041, "step": 9788 }, { "epoch": 1.0612532523850824, "grad_norm": 0.6380323171615601, "learning_rate": 6.485927752792689e-05, "loss": 0.0122, "step": 9789 }, { "epoch": 1.0613616652211622, "grad_norm": 0.2574908435344696, "learning_rate": 6.485565066008995e-05, "loss": 0.0179, "step": 9790 }, { "epoch": 1.061470078057242, "grad_norm": 0.1537146121263504, "learning_rate": 6.485202379225301e-05, "loss": 0.012, "step": 9791 }, { "epoch": 1.0615784908933217, "grad_norm": 0.34692251682281494, "learning_rate": 6.484839692441608e-05, "loss": 0.0275, "step": 9792 }, { "epoch": 1.0616869037294017, "grad_norm": 0.3284478485584259, "learning_rate": 6.484477005657914e-05, "loss": 0.0176, "step": 9793 }, { "epoch": 1.0617953165654814, "grad_norm": 0.2390683889389038, "learning_rate": 6.48411431887422e-05, "loss": 0.0078, "step": 9794 }, { "epoch": 1.0619037294015612, "grad_norm": 0.5258322358131409, "learning_rate": 6.483751632090526e-05, "loss": 0.0266, "step": 9795 }, { "epoch": 1.062012142237641, "grad_norm": 0.27967941761016846, "learning_rate": 6.483388945306833e-05, "loss": 0.0333, "step": 9796 }, { "epoch": 1.0621205550737207, "grad_norm": 0.15213905274868011, "learning_rate": 6.483026258523139e-05, "loss": 0.0073, "step": 9797 }, { "epoch": 1.0622289679098005, "grad_norm": 0.4696086645126343, "learning_rate": 6.482663571739446e-05, "loss": 0.0327, "step": 9798 }, { "epoch": 1.0623373807458802, "grad_norm": 0.343655526638031, "learning_rate": 6.482300884955752e-05, "loss": 0.0133, "step": 9799 }, { "epoch": 1.06244579358196, "grad_norm": 0.3377986550331116, "learning_rate": 6.481938198172058e-05, "loss": 0.0083, "step": 9800 }, { "epoch": 1.06255420641804, "grad_norm": 0.4145689606666565, "learning_rate": 6.481575511388366e-05, "loss": 0.0245, "step": 9801 }, { "epoch": 1.0626626192541198, "grad_norm": 0.3881624639034271, "learning_rate": 6.481212824604672e-05, "loss": 0.0186, "step": 9802 }, { "epoch": 1.0627710320901995, "grad_norm": 0.2610878050327301, "learning_rate": 6.480850137820979e-05, "loss": 0.0286, "step": 9803 }, { "epoch": 1.0628794449262793, "grad_norm": 0.4424939453601837, "learning_rate": 6.480487451037285e-05, "loss": 0.0288, "step": 9804 }, { "epoch": 1.062987857762359, "grad_norm": 0.07271154224872589, "learning_rate": 6.480124764253592e-05, "loss": 0.0051, "step": 9805 }, { "epoch": 1.0630962705984388, "grad_norm": 0.10825426131486893, "learning_rate": 6.479762077469898e-05, "loss": 0.0052, "step": 9806 }, { "epoch": 1.0632046834345186, "grad_norm": 0.10292870551347733, "learning_rate": 6.479399390686203e-05, "loss": 0.006, "step": 9807 }, { "epoch": 1.0633130962705986, "grad_norm": 0.43067243695259094, "learning_rate": 6.47903670390251e-05, "loss": 0.0124, "step": 9808 }, { "epoch": 1.0634215091066783, "grad_norm": 0.1380891352891922, "learning_rate": 6.478674017118816e-05, "loss": 0.0051, "step": 9809 }, { "epoch": 1.063529921942758, "grad_norm": 0.2398487627506256, "learning_rate": 6.478311330335123e-05, "loss": 0.0076, "step": 9810 }, { "epoch": 1.0636383347788378, "grad_norm": 0.3568107783794403, "learning_rate": 6.477948643551429e-05, "loss": 0.0217, "step": 9811 }, { "epoch": 1.0637467476149176, "grad_norm": 0.02894832007586956, "learning_rate": 6.477585956767736e-05, "loss": 0.001, "step": 9812 }, { "epoch": 1.0638551604509974, "grad_norm": 0.27294161915779114, "learning_rate": 6.477223269984042e-05, "loss": 0.0684, "step": 9813 }, { "epoch": 1.0639635732870771, "grad_norm": 0.4633876383304596, "learning_rate": 6.476860583200349e-05, "loss": 0.0469, "step": 9814 }, { "epoch": 1.064071986123157, "grad_norm": 0.1401020884513855, "learning_rate": 6.476497896416655e-05, "loss": 0.0102, "step": 9815 }, { "epoch": 1.0641803989592367, "grad_norm": 0.6595008373260498, "learning_rate": 6.47613520963296e-05, "loss": 0.0602, "step": 9816 }, { "epoch": 1.0642888117953166, "grad_norm": 0.6007850766181946, "learning_rate": 6.475772522849267e-05, "loss": 0.0428, "step": 9817 }, { "epoch": 1.0643972246313964, "grad_norm": 0.6242832541465759, "learning_rate": 6.475409836065574e-05, "loss": 0.0627, "step": 9818 }, { "epoch": 1.0645056374674762, "grad_norm": 0.3812868297100067, "learning_rate": 6.475047149281882e-05, "loss": 0.0157, "step": 9819 }, { "epoch": 1.064614050303556, "grad_norm": 0.034563224762678146, "learning_rate": 6.474684462498187e-05, "loss": 0.0019, "step": 9820 }, { "epoch": 1.0647224631396357, "grad_norm": 0.06710946559906006, "learning_rate": 6.474321775714493e-05, "loss": 0.0033, "step": 9821 }, { "epoch": 1.0648308759757155, "grad_norm": 0.20625461637973785, "learning_rate": 6.4739590889308e-05, "loss": 0.0058, "step": 9822 }, { "epoch": 1.0649392888117952, "grad_norm": 0.21414552628993988, "learning_rate": 6.473596402147106e-05, "loss": 0.016, "step": 9823 }, { "epoch": 1.0650477016478752, "grad_norm": 0.5174596905708313, "learning_rate": 6.473233715363413e-05, "loss": 0.0614, "step": 9824 }, { "epoch": 1.065156114483955, "grad_norm": 0.014360484667122364, "learning_rate": 6.472871028579719e-05, "loss": 0.0009, "step": 9825 }, { "epoch": 1.0652645273200347, "grad_norm": 0.31676068902015686, "learning_rate": 6.472508341796026e-05, "loss": 0.0166, "step": 9826 }, { "epoch": 1.0653729401561145, "grad_norm": 0.23592621088027954, "learning_rate": 6.472145655012331e-05, "loss": 0.0062, "step": 9827 }, { "epoch": 1.0654813529921943, "grad_norm": 0.14994579553604126, "learning_rate": 6.471782968228639e-05, "loss": 0.0107, "step": 9828 }, { "epoch": 1.065589765828274, "grad_norm": 0.2052052766084671, "learning_rate": 6.471420281444944e-05, "loss": 0.0092, "step": 9829 }, { "epoch": 1.0656981786643538, "grad_norm": 0.5449548959732056, "learning_rate": 6.47105759466125e-05, "loss": 0.0429, "step": 9830 }, { "epoch": 1.0658065915004336, "grad_norm": 0.4219750463962555, "learning_rate": 6.470694907877557e-05, "loss": 0.0224, "step": 9831 }, { "epoch": 1.0659150043365135, "grad_norm": 0.2749931216239929, "learning_rate": 6.470332221093863e-05, "loss": 0.0303, "step": 9832 }, { "epoch": 1.0660234171725933, "grad_norm": 0.07205173373222351, "learning_rate": 6.46996953431017e-05, "loss": 0.004, "step": 9833 }, { "epoch": 1.066131830008673, "grad_norm": 0.4439394474029541, "learning_rate": 6.469606847526476e-05, "loss": 0.05, "step": 9834 }, { "epoch": 1.0662402428447528, "grad_norm": 0.27955949306488037, "learning_rate": 6.469244160742783e-05, "loss": 0.0139, "step": 9835 }, { "epoch": 1.0663486556808326, "grad_norm": 0.39340540766716003, "learning_rate": 6.46888147395909e-05, "loss": 0.0194, "step": 9836 }, { "epoch": 1.0664570685169124, "grad_norm": 0.12043057382106781, "learning_rate": 6.468518787175396e-05, "loss": 0.006, "step": 9837 }, { "epoch": 1.0665654813529921, "grad_norm": 0.19529831409454346, "learning_rate": 6.468156100391703e-05, "loss": 0.0086, "step": 9838 }, { "epoch": 1.0666738941890719, "grad_norm": 0.28577250242233276, "learning_rate": 6.467793413608008e-05, "loss": 0.0113, "step": 9839 }, { "epoch": 1.0667823070251519, "grad_norm": 0.4980427026748657, "learning_rate": 6.467430726824316e-05, "loss": 0.0209, "step": 9840 }, { "epoch": 1.0668907198612316, "grad_norm": 0.6372156739234924, "learning_rate": 6.467068040040621e-05, "loss": 0.0243, "step": 9841 }, { "epoch": 1.0669991326973114, "grad_norm": 0.8336536288261414, "learning_rate": 6.466705353256928e-05, "loss": 0.0267, "step": 9842 }, { "epoch": 1.0671075455333912, "grad_norm": 0.1778220534324646, "learning_rate": 6.466342666473234e-05, "loss": 0.0153, "step": 9843 }, { "epoch": 1.067215958369471, "grad_norm": 0.42983001470565796, "learning_rate": 6.46597997968954e-05, "loss": 0.0344, "step": 9844 }, { "epoch": 1.0673243712055507, "grad_norm": 0.2680433690547943, "learning_rate": 6.465617292905847e-05, "loss": 0.0096, "step": 9845 }, { "epoch": 1.0674327840416304, "grad_norm": 0.34274691343307495, "learning_rate": 6.465254606122153e-05, "loss": 0.0451, "step": 9846 }, { "epoch": 1.0675411968777104, "grad_norm": 0.3817528188228607, "learning_rate": 6.46489191933846e-05, "loss": 0.0072, "step": 9847 }, { "epoch": 1.0676496097137902, "grad_norm": 0.27065959572792053, "learning_rate": 6.464529232554765e-05, "loss": 0.0095, "step": 9848 }, { "epoch": 1.06775802254987, "grad_norm": 0.10510580241680145, "learning_rate": 6.464166545771073e-05, "loss": 0.0098, "step": 9849 }, { "epoch": 1.0678664353859497, "grad_norm": 0.09594322741031647, "learning_rate": 6.463803858987378e-05, "loss": 0.0061, "step": 9850 }, { "epoch": 1.0679748482220295, "grad_norm": 0.19555626809597015, "learning_rate": 6.463441172203685e-05, "loss": 0.0103, "step": 9851 }, { "epoch": 1.0680832610581092, "grad_norm": 0.1996634304523468, "learning_rate": 6.463078485419991e-05, "loss": 0.0091, "step": 9852 }, { "epoch": 1.068191673894189, "grad_norm": 0.7934151291847229, "learning_rate": 6.462715798636298e-05, "loss": 0.0445, "step": 9853 }, { "epoch": 1.0683000867302688, "grad_norm": 0.1577211320400238, "learning_rate": 6.462353111852605e-05, "loss": 0.0134, "step": 9854 }, { "epoch": 1.0684084995663488, "grad_norm": 0.1452460139989853, "learning_rate": 6.461990425068911e-05, "loss": 0.0085, "step": 9855 }, { "epoch": 1.0685169124024285, "grad_norm": 0.12908902764320374, "learning_rate": 6.461627738285218e-05, "loss": 0.0117, "step": 9856 }, { "epoch": 1.0686253252385083, "grad_norm": 0.919075608253479, "learning_rate": 6.461265051501524e-05, "loss": 0.06, "step": 9857 }, { "epoch": 1.068733738074588, "grad_norm": 0.19903020560741425, "learning_rate": 6.46090236471783e-05, "loss": 0.0112, "step": 9858 }, { "epoch": 1.0688421509106678, "grad_norm": 0.2491442859172821, "learning_rate": 6.460539677934137e-05, "loss": 0.0126, "step": 9859 }, { "epoch": 1.0689505637467476, "grad_norm": 0.2502278685569763, "learning_rate": 6.460176991150442e-05, "loss": 0.0169, "step": 9860 }, { "epoch": 1.0690589765828273, "grad_norm": 0.16182087361812592, "learning_rate": 6.45981430436675e-05, "loss": 0.0047, "step": 9861 }, { "epoch": 1.069167389418907, "grad_norm": 0.31508776545524597, "learning_rate": 6.459451617583055e-05, "loss": 0.013, "step": 9862 }, { "epoch": 1.069275802254987, "grad_norm": 0.4066592752933502, "learning_rate": 6.459088930799362e-05, "loss": 0.0454, "step": 9863 }, { "epoch": 1.0693842150910668, "grad_norm": 0.11447013169527054, "learning_rate": 6.458726244015668e-05, "loss": 0.0074, "step": 9864 }, { "epoch": 1.0694926279271466, "grad_norm": 0.3795694410800934, "learning_rate": 6.458363557231975e-05, "loss": 0.0221, "step": 9865 }, { "epoch": 1.0696010407632264, "grad_norm": 0.7843456864356995, "learning_rate": 6.458000870448281e-05, "loss": 0.0433, "step": 9866 }, { "epoch": 1.0697094535993061, "grad_norm": 0.34649398922920227, "learning_rate": 6.457638183664587e-05, "loss": 0.0217, "step": 9867 }, { "epoch": 1.069817866435386, "grad_norm": 0.30850377678871155, "learning_rate": 6.457275496880894e-05, "loss": 0.0304, "step": 9868 }, { "epoch": 1.0699262792714657, "grad_norm": 0.6193159222602844, "learning_rate": 6.4569128100972e-05, "loss": 0.0292, "step": 9869 }, { "epoch": 1.0700346921075456, "grad_norm": 0.31445226073265076, "learning_rate": 6.456550123313508e-05, "loss": 0.0168, "step": 9870 }, { "epoch": 1.0701431049436254, "grad_norm": 0.13494236767292023, "learning_rate": 6.456187436529814e-05, "loss": 0.0045, "step": 9871 }, { "epoch": 1.0702515177797052, "grad_norm": 0.30105161666870117, "learning_rate": 6.455824749746119e-05, "loss": 0.0193, "step": 9872 }, { "epoch": 1.070359930615785, "grad_norm": 0.2630878984928131, "learning_rate": 6.455462062962426e-05, "loss": 0.0282, "step": 9873 }, { "epoch": 1.0704683434518647, "grad_norm": 0.25787895917892456, "learning_rate": 6.455099376178732e-05, "loss": 0.02, "step": 9874 }, { "epoch": 1.0705767562879445, "grad_norm": 0.4703613221645355, "learning_rate": 6.454736689395039e-05, "loss": 0.0343, "step": 9875 }, { "epoch": 1.0706851691240242, "grad_norm": 0.2193034589290619, "learning_rate": 6.454374002611345e-05, "loss": 0.0225, "step": 9876 }, { "epoch": 1.070793581960104, "grad_norm": 0.05572213605046272, "learning_rate": 6.454011315827652e-05, "loss": 0.0025, "step": 9877 }, { "epoch": 1.0709019947961838, "grad_norm": 0.28870970010757446, "learning_rate": 6.453648629043958e-05, "loss": 0.0618, "step": 9878 }, { "epoch": 1.0710104076322637, "grad_norm": 0.1945200115442276, "learning_rate": 6.453285942260265e-05, "loss": 0.0302, "step": 9879 }, { "epoch": 1.0711188204683435, "grad_norm": 0.6355773210525513, "learning_rate": 6.45292325547657e-05, "loss": 0.0325, "step": 9880 }, { "epoch": 1.0712272333044233, "grad_norm": 0.06316854059696198, "learning_rate": 6.452560568692876e-05, "loss": 0.0024, "step": 9881 }, { "epoch": 1.071335646140503, "grad_norm": 0.27585238218307495, "learning_rate": 6.452197881909183e-05, "loss": 0.022, "step": 9882 }, { "epoch": 1.0714440589765828, "grad_norm": 0.06193835660815239, "learning_rate": 6.451835195125489e-05, "loss": 0.0026, "step": 9883 }, { "epoch": 1.0715524718126626, "grad_norm": 0.47671419382095337, "learning_rate": 6.451472508341796e-05, "loss": 0.0374, "step": 9884 }, { "epoch": 1.0716608846487423, "grad_norm": 0.6030675172805786, "learning_rate": 6.451109821558102e-05, "loss": 0.0268, "step": 9885 }, { "epoch": 1.0717692974848223, "grad_norm": 0.2562623620033264, "learning_rate": 6.450747134774409e-05, "loss": 0.0189, "step": 9886 }, { "epoch": 1.071877710320902, "grad_norm": 0.268527090549469, "learning_rate": 6.450384447990716e-05, "loss": 0.0116, "step": 9887 }, { "epoch": 1.0719861231569818, "grad_norm": 0.03901873528957367, "learning_rate": 6.450021761207022e-05, "loss": 0.0057, "step": 9888 }, { "epoch": 1.0720945359930616, "grad_norm": 0.07877107709646225, "learning_rate": 6.449659074423329e-05, "loss": 0.006, "step": 9889 }, { "epoch": 1.0722029488291414, "grad_norm": 0.13688302040100098, "learning_rate": 6.449296387639635e-05, "loss": 0.0135, "step": 9890 }, { "epoch": 1.0723113616652211, "grad_norm": 0.15413518249988556, "learning_rate": 6.448933700855942e-05, "loss": 0.0162, "step": 9891 }, { "epoch": 1.0724197745013009, "grad_norm": 0.15115787088871002, "learning_rate": 6.448571014072247e-05, "loss": 0.0124, "step": 9892 }, { "epoch": 1.0725281873373806, "grad_norm": 0.1455148458480835, "learning_rate": 6.448208327288555e-05, "loss": 0.0122, "step": 9893 }, { "epoch": 1.0726366001734606, "grad_norm": 0.5413970947265625, "learning_rate": 6.44784564050486e-05, "loss": 0.0308, "step": 9894 }, { "epoch": 1.0727450130095404, "grad_norm": 1.1734249591827393, "learning_rate": 6.447482953721167e-05, "loss": 0.0653, "step": 9895 }, { "epoch": 1.0728534258456202, "grad_norm": 0.45646679401397705, "learning_rate": 6.447120266937473e-05, "loss": 0.0541, "step": 9896 }, { "epoch": 1.0729618386817, "grad_norm": 0.18868379294872284, "learning_rate": 6.446757580153779e-05, "loss": 0.0069, "step": 9897 }, { "epoch": 1.0730702515177797, "grad_norm": 0.4186886250972748, "learning_rate": 6.446394893370086e-05, "loss": 0.0465, "step": 9898 }, { "epoch": 1.0731786643538594, "grad_norm": 0.06195475906133652, "learning_rate": 6.446032206586392e-05, "loss": 0.0053, "step": 9899 }, { "epoch": 1.0732870771899392, "grad_norm": 0.34593361616134644, "learning_rate": 6.445669519802699e-05, "loss": 0.0342, "step": 9900 }, { "epoch": 1.073395490026019, "grad_norm": 0.558090329170227, "learning_rate": 6.445306833019005e-05, "loss": 0.0169, "step": 9901 }, { "epoch": 1.073503902862099, "grad_norm": 0.05730877071619034, "learning_rate": 6.444944146235312e-05, "loss": 0.0025, "step": 9902 }, { "epoch": 1.0736123156981787, "grad_norm": 0.2225034385919571, "learning_rate": 6.444581459451617e-05, "loss": 0.0407, "step": 9903 }, { "epoch": 1.0737207285342585, "grad_norm": 0.0757293552160263, "learning_rate": 6.444218772667924e-05, "loss": 0.0026, "step": 9904 }, { "epoch": 1.0738291413703382, "grad_norm": 0.44368043541908264, "learning_rate": 6.443856085884232e-05, "loss": 0.0261, "step": 9905 }, { "epoch": 1.073937554206418, "grad_norm": 0.715018093585968, "learning_rate": 6.443493399100537e-05, "loss": 0.0098, "step": 9906 }, { "epoch": 1.0740459670424978, "grad_norm": 0.18354454636573792, "learning_rate": 6.443130712316844e-05, "loss": 0.0093, "step": 9907 }, { "epoch": 1.0741543798785775, "grad_norm": 0.17074327170848846, "learning_rate": 6.44276802553315e-05, "loss": 0.0068, "step": 9908 }, { "epoch": 1.0742627927146575, "grad_norm": 0.5564225912094116, "learning_rate": 6.442405338749457e-05, "loss": 0.0248, "step": 9909 }, { "epoch": 1.0743712055507373, "grad_norm": 0.26614314317703247, "learning_rate": 6.442042651965763e-05, "loss": 0.0148, "step": 9910 }, { "epoch": 1.074479618386817, "grad_norm": 0.7549862265586853, "learning_rate": 6.441679965182069e-05, "loss": 0.0861, "step": 9911 }, { "epoch": 1.0745880312228968, "grad_norm": 0.33191514015197754, "learning_rate": 6.441317278398376e-05, "loss": 0.0135, "step": 9912 }, { "epoch": 1.0746964440589766, "grad_norm": 0.14257694780826569, "learning_rate": 6.440954591614681e-05, "loss": 0.0121, "step": 9913 }, { "epoch": 1.0748048568950563, "grad_norm": 0.2727925181388855, "learning_rate": 6.440591904830989e-05, "loss": 0.0142, "step": 9914 }, { "epoch": 1.074913269731136, "grad_norm": 0.4927476942539215, "learning_rate": 6.440229218047294e-05, "loss": 0.0257, "step": 9915 }, { "epoch": 1.0750216825672159, "grad_norm": 0.24139739573001862, "learning_rate": 6.439866531263601e-05, "loss": 0.0139, "step": 9916 }, { "epoch": 1.0751300954032958, "grad_norm": 0.21562713384628296, "learning_rate": 6.439503844479907e-05, "loss": 0.0092, "step": 9917 }, { "epoch": 1.0752385082393756, "grad_norm": 0.23581472039222717, "learning_rate": 6.439141157696214e-05, "loss": 0.0073, "step": 9918 }, { "epoch": 1.0753469210754554, "grad_norm": 0.3531734049320221, "learning_rate": 6.43877847091252e-05, "loss": 0.0291, "step": 9919 }, { "epoch": 1.0754553339115351, "grad_norm": 0.21424783766269684, "learning_rate": 6.438415784128826e-05, "loss": 0.0113, "step": 9920 }, { "epoch": 1.075563746747615, "grad_norm": 0.16827228665351868, "learning_rate": 6.438053097345133e-05, "loss": 0.0082, "step": 9921 }, { "epoch": 1.0756721595836947, "grad_norm": 0.05203830450773239, "learning_rate": 6.43769041056144e-05, "loss": 0.0097, "step": 9922 }, { "epoch": 1.0757805724197744, "grad_norm": 0.37452515959739685, "learning_rate": 6.437327723777747e-05, "loss": 0.0186, "step": 9923 }, { "epoch": 1.0758889852558542, "grad_norm": 0.6911865472793579, "learning_rate": 6.436965036994053e-05, "loss": 0.0256, "step": 9924 }, { "epoch": 1.0759973980919342, "grad_norm": 0.26747265458106995, "learning_rate": 6.436602350210358e-05, "loss": 0.0115, "step": 9925 }, { "epoch": 1.076105810928014, "grad_norm": 0.16246359050273895, "learning_rate": 6.436239663426665e-05, "loss": 0.0137, "step": 9926 }, { "epoch": 1.0762142237640937, "grad_norm": 0.37386003136634827, "learning_rate": 6.435876976642971e-05, "loss": 0.0435, "step": 9927 }, { "epoch": 1.0763226366001735, "grad_norm": 0.11100441217422485, "learning_rate": 6.435514289859278e-05, "loss": 0.0122, "step": 9928 }, { "epoch": 1.0764310494362532, "grad_norm": 0.2668408751487732, "learning_rate": 6.435151603075584e-05, "loss": 0.0237, "step": 9929 }, { "epoch": 1.076539462272333, "grad_norm": 0.08271052688360214, "learning_rate": 6.434788916291891e-05, "loss": 0.0064, "step": 9930 }, { "epoch": 1.0766478751084128, "grad_norm": 0.33722323179244995, "learning_rate": 6.434426229508197e-05, "loss": 0.0204, "step": 9931 }, { "epoch": 1.0767562879444927, "grad_norm": 0.49184876680374146, "learning_rate": 6.434063542724504e-05, "loss": 0.0237, "step": 9932 }, { "epoch": 1.0768647007805725, "grad_norm": 0.3479284644126892, "learning_rate": 6.43370085594081e-05, "loss": 0.0419, "step": 9933 }, { "epoch": 1.0769731136166523, "grad_norm": 0.027698606252670288, "learning_rate": 6.433338169157115e-05, "loss": 0.0012, "step": 9934 }, { "epoch": 1.077081526452732, "grad_norm": 0.029594235122203827, "learning_rate": 6.432975482373422e-05, "loss": 0.0018, "step": 9935 }, { "epoch": 1.0771899392888118, "grad_norm": 0.598549485206604, "learning_rate": 6.432612795589728e-05, "loss": 0.0153, "step": 9936 }, { "epoch": 1.0772983521248916, "grad_norm": 0.4281620383262634, "learning_rate": 6.432250108806035e-05, "loss": 0.0391, "step": 9937 }, { "epoch": 1.0774067649609713, "grad_norm": 0.4233573079109192, "learning_rate": 6.431887422022341e-05, "loss": 0.031, "step": 9938 }, { "epoch": 1.077515177797051, "grad_norm": 0.9579228162765503, "learning_rate": 6.431524735238648e-05, "loss": 0.0254, "step": 9939 }, { "epoch": 1.0776235906331308, "grad_norm": 0.13751575350761414, "learning_rate": 6.431162048454955e-05, "loss": 0.0041, "step": 9940 }, { "epoch": 1.0777320034692108, "grad_norm": 0.5617097616195679, "learning_rate": 6.430799361671261e-05, "loss": 0.0202, "step": 9941 }, { "epoch": 1.0778404163052906, "grad_norm": 0.4702428877353668, "learning_rate": 6.430436674887568e-05, "loss": 0.0415, "step": 9942 }, { "epoch": 1.0779488291413704, "grad_norm": 0.4472120702266693, "learning_rate": 6.430073988103874e-05, "loss": 0.0236, "step": 9943 }, { "epoch": 1.0780572419774501, "grad_norm": 0.07521204650402069, "learning_rate": 6.429711301320181e-05, "loss": 0.0027, "step": 9944 }, { "epoch": 1.0781656548135299, "grad_norm": 0.15475504100322723, "learning_rate": 6.429348614536487e-05, "loss": 0.0173, "step": 9945 }, { "epoch": 1.0782740676496096, "grad_norm": 0.3935140073299408, "learning_rate": 6.428985927752794e-05, "loss": 0.0373, "step": 9946 }, { "epoch": 1.0783824804856894, "grad_norm": 0.18131552636623383, "learning_rate": 6.4286232409691e-05, "loss": 0.0102, "step": 9947 }, { "epoch": 1.0784908933217694, "grad_norm": 0.37672314047813416, "learning_rate": 6.428260554185405e-05, "loss": 0.038, "step": 9948 }, { "epoch": 1.0785993061578492, "grad_norm": 0.052208613604307175, "learning_rate": 6.427897867401712e-05, "loss": 0.0035, "step": 9949 }, { "epoch": 1.078707718993929, "grad_norm": 0.18462951481342316, "learning_rate": 6.427535180618018e-05, "loss": 0.0286, "step": 9950 }, { "epoch": 1.0788161318300087, "grad_norm": 0.3158790171146393, "learning_rate": 6.427172493834325e-05, "loss": 0.0145, "step": 9951 }, { "epoch": 1.0789245446660884, "grad_norm": 0.2963950037956238, "learning_rate": 6.426809807050631e-05, "loss": 0.0057, "step": 9952 }, { "epoch": 1.0790329575021682, "grad_norm": 0.6645916104316711, "learning_rate": 6.426447120266938e-05, "loss": 0.0262, "step": 9953 }, { "epoch": 1.079141370338248, "grad_norm": 0.5691773891448975, "learning_rate": 6.426084433483244e-05, "loss": 0.0453, "step": 9954 }, { "epoch": 1.0792497831743277, "grad_norm": 0.38162025809288025, "learning_rate": 6.42572174669955e-05, "loss": 0.0253, "step": 9955 }, { "epoch": 1.0793581960104077, "grad_norm": 0.2879716157913208, "learning_rate": 6.425359059915858e-05, "loss": 0.0136, "step": 9956 }, { "epoch": 1.0794666088464875, "grad_norm": 0.49122315645217896, "learning_rate": 6.424996373132163e-05, "loss": 0.0347, "step": 9957 }, { "epoch": 1.0795750216825672, "grad_norm": 0.18883949518203735, "learning_rate": 6.42463368634847e-05, "loss": 0.0143, "step": 9958 }, { "epoch": 1.079683434518647, "grad_norm": 0.2038881778717041, "learning_rate": 6.424270999564776e-05, "loss": 0.0166, "step": 9959 }, { "epoch": 1.0797918473547268, "grad_norm": 0.183350071310997, "learning_rate": 6.423908312781083e-05, "loss": 0.0134, "step": 9960 }, { "epoch": 1.0799002601908065, "grad_norm": 0.2805904448032379, "learning_rate": 6.423545625997389e-05, "loss": 0.006, "step": 9961 }, { "epoch": 1.0800086730268863, "grad_norm": 0.1644899845123291, "learning_rate": 6.423182939213695e-05, "loss": 0.0069, "step": 9962 }, { "epoch": 1.080117085862966, "grad_norm": 0.5611881613731384, "learning_rate": 6.422820252430002e-05, "loss": 0.0273, "step": 9963 }, { "epoch": 1.080225498699046, "grad_norm": 0.199703186750412, "learning_rate": 6.422457565646308e-05, "loss": 0.0126, "step": 9964 }, { "epoch": 1.0803339115351258, "grad_norm": 0.14093823730945587, "learning_rate": 6.422094878862615e-05, "loss": 0.0073, "step": 9965 }, { "epoch": 1.0804423243712056, "grad_norm": 0.3416626453399658, "learning_rate": 6.42173219207892e-05, "loss": 0.0204, "step": 9966 }, { "epoch": 1.0805507372072853, "grad_norm": 0.10818853974342346, "learning_rate": 6.421369505295228e-05, "loss": 0.0114, "step": 9967 }, { "epoch": 1.080659150043365, "grad_norm": 0.07585760205984116, "learning_rate": 6.421006818511533e-05, "loss": 0.0056, "step": 9968 }, { "epoch": 1.0807675628794449, "grad_norm": 0.42179808020591736, "learning_rate": 6.42064413172784e-05, "loss": 0.0429, "step": 9969 }, { "epoch": 1.0808759757155246, "grad_norm": 0.3819943070411682, "learning_rate": 6.420281444944146e-05, "loss": 0.0164, "step": 9970 }, { "epoch": 1.0809843885516046, "grad_norm": 0.25782954692840576, "learning_rate": 6.419918758160452e-05, "loss": 0.0402, "step": 9971 }, { "epoch": 1.0810928013876844, "grad_norm": 0.07672543823719025, "learning_rate": 6.419556071376759e-05, "loss": 0.0049, "step": 9972 }, { "epoch": 1.0812012142237641, "grad_norm": 0.2711561620235443, "learning_rate": 6.419193384593065e-05, "loss": 0.0364, "step": 9973 }, { "epoch": 1.081309627059844, "grad_norm": 0.025864526629447937, "learning_rate": 6.418830697809373e-05, "loss": 0.0013, "step": 9974 }, { "epoch": 1.0814180398959237, "grad_norm": 0.4655689299106598, "learning_rate": 6.418468011025679e-05, "loss": 0.0466, "step": 9975 }, { "epoch": 1.0815264527320034, "grad_norm": 0.4431404173374176, "learning_rate": 6.418105324241985e-05, "loss": 0.0598, "step": 9976 }, { "epoch": 1.0816348655680832, "grad_norm": 8.319951057434082, "learning_rate": 6.417742637458292e-05, "loss": 0.0586, "step": 9977 }, { "epoch": 1.081743278404163, "grad_norm": 0.13751652836799622, "learning_rate": 6.417379950674597e-05, "loss": 0.0063, "step": 9978 }, { "epoch": 1.081851691240243, "grad_norm": 0.13495095074176788, "learning_rate": 6.417017263890905e-05, "loss": 0.0092, "step": 9979 }, { "epoch": 1.0819601040763227, "grad_norm": 0.9873749613761902, "learning_rate": 6.41665457710721e-05, "loss": 0.0625, "step": 9980 }, { "epoch": 1.0820685169124025, "grad_norm": 0.8921065926551819, "learning_rate": 6.416291890323517e-05, "loss": 0.0612, "step": 9981 }, { "epoch": 1.0821769297484822, "grad_norm": 0.21844540536403656, "learning_rate": 6.415929203539823e-05, "loss": 0.0166, "step": 9982 }, { "epoch": 1.082285342584562, "grad_norm": 0.6936382055282593, "learning_rate": 6.41556651675613e-05, "loss": 0.027, "step": 9983 }, { "epoch": 1.0823937554206418, "grad_norm": 0.16416531801223755, "learning_rate": 6.415203829972436e-05, "loss": 0.0171, "step": 9984 }, { "epoch": 1.0825021682567215, "grad_norm": 0.26996660232543945, "learning_rate": 6.414841143188742e-05, "loss": 0.0369, "step": 9985 }, { "epoch": 1.0826105810928013, "grad_norm": 0.3413798213005066, "learning_rate": 6.414478456405049e-05, "loss": 0.0485, "step": 9986 }, { "epoch": 1.0827189939288813, "grad_norm": 0.278227835893631, "learning_rate": 6.414115769621354e-05, "loss": 0.0451, "step": 9987 }, { "epoch": 1.082827406764961, "grad_norm": 0.14797310531139374, "learning_rate": 6.413753082837662e-05, "loss": 0.0107, "step": 9988 }, { "epoch": 1.0829358196010408, "grad_norm": 0.5334630012512207, "learning_rate": 6.413390396053967e-05, "loss": 0.0253, "step": 9989 }, { "epoch": 1.0830442324371206, "grad_norm": 0.2645542621612549, "learning_rate": 6.413027709270274e-05, "loss": 0.024, "step": 9990 }, { "epoch": 1.0831526452732003, "grad_norm": 1.7524389028549194, "learning_rate": 6.412665022486581e-05, "loss": 0.0245, "step": 9991 }, { "epoch": 1.08326105810928, "grad_norm": 0.14880965650081635, "learning_rate": 6.412302335702887e-05, "loss": 0.0046, "step": 9992 }, { "epoch": 1.0833694709453598, "grad_norm": 0.4077512323856354, "learning_rate": 6.411939648919194e-05, "loss": 0.0454, "step": 9993 }, { "epoch": 1.0834778837814398, "grad_norm": 0.5616468191146851, "learning_rate": 6.4115769621355e-05, "loss": 0.0335, "step": 9994 }, { "epoch": 1.0835862966175196, "grad_norm": 0.4368482232093811, "learning_rate": 6.411214275351807e-05, "loss": 0.0161, "step": 9995 }, { "epoch": 1.0836947094535994, "grad_norm": 0.45466694235801697, "learning_rate": 6.410851588568113e-05, "loss": 0.063, "step": 9996 }, { "epoch": 1.0838031222896791, "grad_norm": 0.5124834775924683, "learning_rate": 6.41048890178442e-05, "loss": 0.0193, "step": 9997 }, { "epoch": 1.0839115351257589, "grad_norm": 0.19603633880615234, "learning_rate": 6.410126215000726e-05, "loss": 0.021, "step": 9998 }, { "epoch": 1.0840199479618386, "grad_norm": 0.4944263696670532, "learning_rate": 6.409763528217033e-05, "loss": 0.075, "step": 9999 }, { "epoch": 1.0841283607979184, "grad_norm": 0.16736513376235962, "learning_rate": 6.409400841433338e-05, "loss": 0.0182, "step": 10000 }, { "epoch": 1.0842367736339982, "grad_norm": 0.307000070810318, "learning_rate": 6.409038154649644e-05, "loss": 0.0192, "step": 10001 }, { "epoch": 1.0843451864700782, "grad_norm": 0.5000097751617432, "learning_rate": 6.408675467865951e-05, "loss": 0.0432, "step": 10002 }, { "epoch": 1.084453599306158, "grad_norm": 0.25937166810035706, "learning_rate": 6.408312781082257e-05, "loss": 0.0291, "step": 10003 }, { "epoch": 1.0845620121422377, "grad_norm": 0.29757773876190186, "learning_rate": 6.407950094298564e-05, "loss": 0.0428, "step": 10004 }, { "epoch": 1.0846704249783174, "grad_norm": 0.25110700726509094, "learning_rate": 6.40758740751487e-05, "loss": 0.0163, "step": 10005 }, { "epoch": 1.0847788378143972, "grad_norm": 0.1320500671863556, "learning_rate": 6.407224720731177e-05, "loss": 0.0112, "step": 10006 }, { "epoch": 1.084887250650477, "grad_norm": 0.21558129787445068, "learning_rate": 6.406862033947483e-05, "loss": 0.0356, "step": 10007 }, { "epoch": 1.0849956634865567, "grad_norm": 0.509558916091919, "learning_rate": 6.40649934716379e-05, "loss": 0.0648, "step": 10008 }, { "epoch": 1.0851040763226365, "grad_norm": 1.1023157835006714, "learning_rate": 6.406136660380097e-05, "loss": 0.0574, "step": 10009 }, { "epoch": 1.0852124891587165, "grad_norm": 0.3228279650211334, "learning_rate": 6.405773973596403e-05, "loss": 0.0166, "step": 10010 }, { "epoch": 1.0853209019947962, "grad_norm": 0.05930906534194946, "learning_rate": 6.40541128681271e-05, "loss": 0.0044, "step": 10011 }, { "epoch": 1.085429314830876, "grad_norm": 0.3074333369731903, "learning_rate": 6.405048600029015e-05, "loss": 0.0346, "step": 10012 }, { "epoch": 1.0855377276669558, "grad_norm": 0.07960532605648041, "learning_rate": 6.404685913245322e-05, "loss": 0.0055, "step": 10013 }, { "epoch": 1.0856461405030355, "grad_norm": 0.2732395827770233, "learning_rate": 6.404323226461628e-05, "loss": 0.0355, "step": 10014 }, { "epoch": 1.0857545533391153, "grad_norm": 0.1917271912097931, "learning_rate": 6.403960539677934e-05, "loss": 0.0276, "step": 10015 }, { "epoch": 1.085862966175195, "grad_norm": 0.27661487460136414, "learning_rate": 6.403597852894241e-05, "loss": 0.0264, "step": 10016 }, { "epoch": 1.085971379011275, "grad_norm": 0.22323031723499298, "learning_rate": 6.403235166110547e-05, "loss": 0.0183, "step": 10017 }, { "epoch": 1.0860797918473548, "grad_norm": 0.23779676854610443, "learning_rate": 6.402872479326854e-05, "loss": 0.0171, "step": 10018 }, { "epoch": 1.0861882046834346, "grad_norm": 0.05116945505142212, "learning_rate": 6.40250979254316e-05, "loss": 0.0042, "step": 10019 }, { "epoch": 1.0862966175195143, "grad_norm": 0.21675114333629608, "learning_rate": 6.402147105759467e-05, "loss": 0.0077, "step": 10020 }, { "epoch": 1.086405030355594, "grad_norm": 0.3089196979999542, "learning_rate": 6.401784418975772e-05, "loss": 0.0333, "step": 10021 }, { "epoch": 1.0865134431916739, "grad_norm": 0.1384323388338089, "learning_rate": 6.40142173219208e-05, "loss": 0.0136, "step": 10022 }, { "epoch": 1.0866218560277536, "grad_norm": 0.1086532473564148, "learning_rate": 6.401059045408385e-05, "loss": 0.0202, "step": 10023 }, { "epoch": 1.0867302688638334, "grad_norm": 0.034263648092746735, "learning_rate": 6.400696358624691e-05, "loss": 0.0031, "step": 10024 }, { "epoch": 1.0868386816999132, "grad_norm": 0.05813613533973694, "learning_rate": 6.400333671841e-05, "loss": 0.0048, "step": 10025 }, { "epoch": 1.0869470945359931, "grad_norm": 0.3687255084514618, "learning_rate": 6.399970985057305e-05, "loss": 0.0437, "step": 10026 }, { "epoch": 1.087055507372073, "grad_norm": 0.1578623354434967, "learning_rate": 6.399608298273612e-05, "loss": 0.0196, "step": 10027 }, { "epoch": 1.0871639202081527, "grad_norm": 0.41431179642677307, "learning_rate": 6.399245611489918e-05, "loss": 0.0304, "step": 10028 }, { "epoch": 1.0872723330442324, "grad_norm": 0.14758656919002533, "learning_rate": 6.398882924706224e-05, "loss": 0.0117, "step": 10029 }, { "epoch": 1.0873807458803122, "grad_norm": 0.12949617207050323, "learning_rate": 6.398520237922531e-05, "loss": 0.0144, "step": 10030 }, { "epoch": 1.087489158716392, "grad_norm": 0.13511572778224945, "learning_rate": 6.398157551138837e-05, "loss": 0.0079, "step": 10031 }, { "epoch": 1.0875975715524717, "grad_norm": 0.11995191127061844, "learning_rate": 6.397794864355144e-05, "loss": 0.0116, "step": 10032 }, { "epoch": 1.0877059843885517, "grad_norm": 0.47088032960891724, "learning_rate": 6.39743217757145e-05, "loss": 0.0549, "step": 10033 }, { "epoch": 1.0878143972246315, "grad_norm": 0.302429735660553, "learning_rate": 6.397069490787756e-05, "loss": 0.0168, "step": 10034 }, { "epoch": 1.0879228100607112, "grad_norm": 0.12785367667675018, "learning_rate": 6.396706804004062e-05, "loss": 0.0069, "step": 10035 }, { "epoch": 1.088031222896791, "grad_norm": 0.15298430621623993, "learning_rate": 6.396344117220369e-05, "loss": 0.0063, "step": 10036 }, { "epoch": 1.0881396357328708, "grad_norm": 0.213949516415596, "learning_rate": 6.395981430436675e-05, "loss": 0.0186, "step": 10037 }, { "epoch": 1.0882480485689505, "grad_norm": 0.35286325216293335, "learning_rate": 6.395618743652981e-05, "loss": 0.0414, "step": 10038 }, { "epoch": 1.0883564614050303, "grad_norm": 0.1648302972316742, "learning_rate": 6.395256056869288e-05, "loss": 0.0105, "step": 10039 }, { "epoch": 1.08846487424111, "grad_norm": 0.10511409491300583, "learning_rate": 6.394893370085594e-05, "loss": 0.0043, "step": 10040 }, { "epoch": 1.08857328707719, "grad_norm": 0.133550226688385, "learning_rate": 6.3945306833019e-05, "loss": 0.01, "step": 10041 }, { "epoch": 1.0886816999132698, "grad_norm": 0.3386259078979492, "learning_rate": 6.394167996518206e-05, "loss": 0.087, "step": 10042 }, { "epoch": 1.0887901127493496, "grad_norm": 0.23874764144420624, "learning_rate": 6.393805309734513e-05, "loss": 0.0196, "step": 10043 }, { "epoch": 1.0888985255854293, "grad_norm": 0.3158309757709503, "learning_rate": 6.39344262295082e-05, "loss": 0.0177, "step": 10044 }, { "epoch": 1.089006938421509, "grad_norm": 0.3432324230670929, "learning_rate": 6.393079936167126e-05, "loss": 0.0356, "step": 10045 }, { "epoch": 1.0891153512575888, "grad_norm": 0.14796431362628937, "learning_rate": 6.392717249383433e-05, "loss": 0.0078, "step": 10046 }, { "epoch": 1.0892237640936686, "grad_norm": 0.427420049905777, "learning_rate": 6.392354562599739e-05, "loss": 0.0177, "step": 10047 }, { "epoch": 1.0893321769297484, "grad_norm": 0.37941354513168335, "learning_rate": 6.391991875816046e-05, "loss": 0.0197, "step": 10048 }, { "epoch": 1.0894405897658284, "grad_norm": 0.27997586131095886, "learning_rate": 6.391629189032352e-05, "loss": 0.0126, "step": 10049 }, { "epoch": 1.0895490026019081, "grad_norm": 0.2539987564086914, "learning_rate": 6.391266502248659e-05, "loss": 0.0176, "step": 10050 }, { "epoch": 1.0896574154379879, "grad_norm": 0.3168801963329315, "learning_rate": 6.390903815464965e-05, "loss": 0.0251, "step": 10051 }, { "epoch": 1.0897658282740676, "grad_norm": 0.46537208557128906, "learning_rate": 6.39054112868127e-05, "loss": 0.0172, "step": 10052 }, { "epoch": 1.0898742411101474, "grad_norm": 0.3250625729560852, "learning_rate": 6.390178441897578e-05, "loss": 0.0331, "step": 10053 }, { "epoch": 1.0899826539462272, "grad_norm": 0.40190160274505615, "learning_rate": 6.389815755113883e-05, "loss": 0.0329, "step": 10054 }, { "epoch": 1.090091066782307, "grad_norm": 0.18851737678050995, "learning_rate": 6.38945306833019e-05, "loss": 0.0209, "step": 10055 }, { "epoch": 1.090199479618387, "grad_norm": 0.1906774938106537, "learning_rate": 6.389090381546496e-05, "loss": 0.0033, "step": 10056 }, { "epoch": 1.0903078924544667, "grad_norm": 0.6643574833869934, "learning_rate": 6.388727694762803e-05, "loss": 0.0276, "step": 10057 }, { "epoch": 1.0904163052905465, "grad_norm": 0.08730792999267578, "learning_rate": 6.388365007979109e-05, "loss": 0.0035, "step": 10058 }, { "epoch": 1.0905247181266262, "grad_norm": 0.060899052768945694, "learning_rate": 6.388002321195416e-05, "loss": 0.0045, "step": 10059 }, { "epoch": 1.090633130962706, "grad_norm": 0.19608508050441742, "learning_rate": 6.387639634411723e-05, "loss": 0.0158, "step": 10060 }, { "epoch": 1.0907415437987857, "grad_norm": 0.8627344965934753, "learning_rate": 6.387276947628029e-05, "loss": 0.0544, "step": 10061 }, { "epoch": 1.0908499566348655, "grad_norm": 0.21260878443717957, "learning_rate": 6.386914260844336e-05, "loss": 0.0082, "step": 10062 }, { "epoch": 1.0909583694709453, "grad_norm": 0.8699266910552979, "learning_rate": 6.386551574060642e-05, "loss": 0.0191, "step": 10063 }, { "epoch": 1.0910667823070253, "grad_norm": 0.33917686343193054, "learning_rate": 6.386188887276949e-05, "loss": 0.0109, "step": 10064 }, { "epoch": 1.091175195143105, "grad_norm": 0.14094361662864685, "learning_rate": 6.385826200493254e-05, "loss": 0.008, "step": 10065 }, { "epoch": 1.0912836079791848, "grad_norm": 0.19836200773715973, "learning_rate": 6.38546351370956e-05, "loss": 0.0074, "step": 10066 }, { "epoch": 1.0913920208152645, "grad_norm": 0.07047177851200104, "learning_rate": 6.385100826925867e-05, "loss": 0.0027, "step": 10067 }, { "epoch": 1.0915004336513443, "grad_norm": 1.0950336456298828, "learning_rate": 6.384738140142173e-05, "loss": 0.0199, "step": 10068 }, { "epoch": 1.091608846487424, "grad_norm": 0.17518848180770874, "learning_rate": 6.38437545335848e-05, "loss": 0.0041, "step": 10069 }, { "epoch": 1.0917172593235038, "grad_norm": 0.17215515673160553, "learning_rate": 6.384012766574786e-05, "loss": 0.009, "step": 10070 }, { "epoch": 1.0918256721595836, "grad_norm": 1.0504480600357056, "learning_rate": 6.383650079791093e-05, "loss": 0.0219, "step": 10071 }, { "epoch": 1.0919340849956636, "grad_norm": 0.5805523991584778, "learning_rate": 6.383287393007399e-05, "loss": 0.0138, "step": 10072 }, { "epoch": 1.0920424978317433, "grad_norm": 0.345716267824173, "learning_rate": 6.382924706223706e-05, "loss": 0.0317, "step": 10073 }, { "epoch": 1.092150910667823, "grad_norm": 1.1225942373275757, "learning_rate": 6.382562019440011e-05, "loss": 0.0116, "step": 10074 }, { "epoch": 1.0922593235039029, "grad_norm": 0.36873456835746765, "learning_rate": 6.382199332656317e-05, "loss": 0.0143, "step": 10075 }, { "epoch": 1.0923677363399826, "grad_norm": 0.32964614033699036, "learning_rate": 6.381836645872624e-05, "loss": 0.0128, "step": 10076 }, { "epoch": 1.0924761491760624, "grad_norm": 0.2426622062921524, "learning_rate": 6.381473959088931e-05, "loss": 0.0199, "step": 10077 }, { "epoch": 1.0925845620121422, "grad_norm": 0.14625640213489532, "learning_rate": 6.381111272305238e-05, "loss": 0.0071, "step": 10078 }, { "epoch": 1.0926929748482221, "grad_norm": 0.30787166953086853, "learning_rate": 6.380748585521544e-05, "loss": 0.0231, "step": 10079 }, { "epoch": 1.092801387684302, "grad_norm": 0.8073290586471558, "learning_rate": 6.380385898737851e-05, "loss": 0.0211, "step": 10080 }, { "epoch": 1.0929098005203817, "grad_norm": 0.7942994236946106, "learning_rate": 6.380023211954157e-05, "loss": 0.0247, "step": 10081 }, { "epoch": 1.0930182133564614, "grad_norm": 0.5827321410179138, "learning_rate": 6.379660525170463e-05, "loss": 0.0377, "step": 10082 }, { "epoch": 1.0931266261925412, "grad_norm": 0.1517898291349411, "learning_rate": 6.37929783838677e-05, "loss": 0.002, "step": 10083 }, { "epoch": 1.093235039028621, "grad_norm": 0.01816079206764698, "learning_rate": 6.378935151603076e-05, "loss": 0.0004, "step": 10084 }, { "epoch": 1.0933434518647007, "grad_norm": 0.3339245021343231, "learning_rate": 6.378572464819383e-05, "loss": 0.0381, "step": 10085 }, { "epoch": 1.0934518647007805, "grad_norm": 0.5174158215522766, "learning_rate": 6.378209778035688e-05, "loss": 0.0044, "step": 10086 }, { "epoch": 1.0935602775368602, "grad_norm": 0.17527982592582703, "learning_rate": 6.377847091251996e-05, "loss": 0.0032, "step": 10087 }, { "epoch": 1.0936686903729402, "grad_norm": 1.0443323850631714, "learning_rate": 6.377484404468301e-05, "loss": 0.0356, "step": 10088 }, { "epoch": 1.09377710320902, "grad_norm": 0.1895758956670761, "learning_rate": 6.377121717684607e-05, "loss": 0.007, "step": 10089 }, { "epoch": 1.0938855160450998, "grad_norm": 0.3726498484611511, "learning_rate": 6.376759030900914e-05, "loss": 0.0364, "step": 10090 }, { "epoch": 1.0939939288811795, "grad_norm": 0.08812660723924637, "learning_rate": 6.37639634411722e-05, "loss": 0.0088, "step": 10091 }, { "epoch": 1.0941023417172593, "grad_norm": 0.4265896677970886, "learning_rate": 6.376033657333527e-05, "loss": 0.0246, "step": 10092 }, { "epoch": 1.094210754553339, "grad_norm": 0.025996621698141098, "learning_rate": 6.375670970549833e-05, "loss": 0.0012, "step": 10093 }, { "epoch": 1.0943191673894188, "grad_norm": 0.2602733075618744, "learning_rate": 6.375308283766141e-05, "loss": 0.0382, "step": 10094 }, { "epoch": 1.0944275802254988, "grad_norm": 0.290825217962265, "learning_rate": 6.374945596982447e-05, "loss": 0.006, "step": 10095 }, { "epoch": 1.0945359930615786, "grad_norm": 0.06246113404631615, "learning_rate": 6.374582910198753e-05, "loss": 0.0019, "step": 10096 }, { "epoch": 1.0946444058976583, "grad_norm": 0.5779990553855896, "learning_rate": 6.37422022341506e-05, "loss": 0.0286, "step": 10097 }, { "epoch": 1.094752818733738, "grad_norm": 0.20975393056869507, "learning_rate": 6.373857536631365e-05, "loss": 0.018, "step": 10098 }, { "epoch": 1.0948612315698178, "grad_norm": 0.18587927520275116, "learning_rate": 6.373494849847672e-05, "loss": 0.0034, "step": 10099 }, { "epoch": 1.0949696444058976, "grad_norm": 0.08814095705747604, "learning_rate": 6.373132163063978e-05, "loss": 0.0022, "step": 10100 }, { "epoch": 1.0950780572419774, "grad_norm": 1.1728612184524536, "learning_rate": 6.372769476280285e-05, "loss": 0.0337, "step": 10101 }, { "epoch": 1.0951864700780571, "grad_norm": 0.7446044683456421, "learning_rate": 6.372406789496591e-05, "loss": 0.0537, "step": 10102 }, { "epoch": 1.0952948829141371, "grad_norm": 0.300422728061676, "learning_rate": 6.372044102712898e-05, "loss": 0.0123, "step": 10103 }, { "epoch": 1.0954032957502169, "grad_norm": 0.7720097303390503, "learning_rate": 6.371681415929204e-05, "loss": 0.0272, "step": 10104 }, { "epoch": 1.0955117085862967, "grad_norm": 0.48551467061042786, "learning_rate": 6.37131872914551e-05, "loss": 0.0382, "step": 10105 }, { "epoch": 1.0956201214223764, "grad_norm": 0.1411990225315094, "learning_rate": 6.370956042361817e-05, "loss": 0.008, "step": 10106 }, { "epoch": 1.0957285342584562, "grad_norm": 0.258625328540802, "learning_rate": 6.370593355578122e-05, "loss": 0.0086, "step": 10107 }, { "epoch": 1.095836947094536, "grad_norm": 0.3361755907535553, "learning_rate": 6.37023066879443e-05, "loss": 0.0165, "step": 10108 }, { "epoch": 1.0959453599306157, "grad_norm": 0.12768501043319702, "learning_rate": 6.369867982010735e-05, "loss": 0.0031, "step": 10109 }, { "epoch": 1.0960537727666955, "grad_norm": 0.5747489333152771, "learning_rate": 6.369505295227042e-05, "loss": 0.0577, "step": 10110 }, { "epoch": 1.0961621856027755, "grad_norm": 0.2481241673231125, "learning_rate": 6.369142608443348e-05, "loss": 0.0068, "step": 10111 }, { "epoch": 1.0962705984388552, "grad_norm": 0.01867925003170967, "learning_rate": 6.368779921659655e-05, "loss": 0.0012, "step": 10112 }, { "epoch": 1.096379011274935, "grad_norm": 0.31376349925994873, "learning_rate": 6.368417234875962e-05, "loss": 0.0702, "step": 10113 }, { "epoch": 1.0964874241110147, "grad_norm": 0.22096602618694305, "learning_rate": 6.368054548092268e-05, "loss": 0.0048, "step": 10114 }, { "epoch": 1.0965958369470945, "grad_norm": 0.36404043436050415, "learning_rate": 6.367691861308575e-05, "loss": 0.0426, "step": 10115 }, { "epoch": 1.0967042497831743, "grad_norm": 0.7989422082901001, "learning_rate": 6.367329174524881e-05, "loss": 0.1213, "step": 10116 }, { "epoch": 1.096812662619254, "grad_norm": 0.34134989976882935, "learning_rate": 6.366966487741188e-05, "loss": 0.0893, "step": 10117 }, { "epoch": 1.096921075455334, "grad_norm": 0.45110446214675903, "learning_rate": 6.366603800957494e-05, "loss": 0.0309, "step": 10118 }, { "epoch": 1.0970294882914138, "grad_norm": 0.46711447834968567, "learning_rate": 6.366241114173799e-05, "loss": 0.053, "step": 10119 }, { "epoch": 1.0971379011274935, "grad_norm": 0.3040969669818878, "learning_rate": 6.365878427390106e-05, "loss": 0.0116, "step": 10120 }, { "epoch": 1.0972463139635733, "grad_norm": 0.09349945187568665, "learning_rate": 6.365515740606412e-05, "loss": 0.0022, "step": 10121 }, { "epoch": 1.097354726799653, "grad_norm": 0.6992161870002747, "learning_rate": 6.365153053822719e-05, "loss": 0.0095, "step": 10122 }, { "epoch": 1.0974631396357328, "grad_norm": 0.25803086161613464, "learning_rate": 6.364790367039025e-05, "loss": 0.0094, "step": 10123 }, { "epoch": 1.0975715524718126, "grad_norm": 0.2850036323070526, "learning_rate": 6.364427680255332e-05, "loss": 0.0261, "step": 10124 }, { "epoch": 1.0976799653078924, "grad_norm": 0.41251128911972046, "learning_rate": 6.364064993471638e-05, "loss": 0.0162, "step": 10125 }, { "epoch": 1.0977883781439723, "grad_norm": 0.5215725302696228, "learning_rate": 6.363702306687945e-05, "loss": 0.0302, "step": 10126 }, { "epoch": 1.097896790980052, "grad_norm": 0.6630164980888367, "learning_rate": 6.36333961990425e-05, "loss": 0.0303, "step": 10127 }, { "epoch": 1.0980052038161319, "grad_norm": 0.33455318212509155, "learning_rate": 6.362976933120556e-05, "loss": 0.0531, "step": 10128 }, { "epoch": 1.0981136166522116, "grad_norm": 0.3927175998687744, "learning_rate": 6.362614246336865e-05, "loss": 0.0551, "step": 10129 }, { "epoch": 1.0982220294882914, "grad_norm": 0.23992818593978882, "learning_rate": 6.36225155955317e-05, "loss": 0.0296, "step": 10130 }, { "epoch": 1.0983304423243712, "grad_norm": 0.10012481361627579, "learning_rate": 6.361888872769478e-05, "loss": 0.0091, "step": 10131 }, { "epoch": 1.098438855160451, "grad_norm": 0.38277360796928406, "learning_rate": 6.361526185985783e-05, "loss": 0.0258, "step": 10132 }, { "epoch": 1.0985472679965307, "grad_norm": 0.3363973796367645, "learning_rate": 6.361163499202089e-05, "loss": 0.0318, "step": 10133 }, { "epoch": 1.0986556808326107, "grad_norm": 0.31817495822906494, "learning_rate": 6.360800812418396e-05, "loss": 0.0248, "step": 10134 }, { "epoch": 1.0987640936686904, "grad_norm": 0.26452910900115967, "learning_rate": 6.360438125634702e-05, "loss": 0.0123, "step": 10135 }, { "epoch": 1.0988725065047702, "grad_norm": 0.062352053821086884, "learning_rate": 6.360075438851009e-05, "loss": 0.007, "step": 10136 }, { "epoch": 1.09898091934085, "grad_norm": 0.21525928378105164, "learning_rate": 6.359712752067315e-05, "loss": 0.0164, "step": 10137 }, { "epoch": 1.0990893321769297, "grad_norm": 0.3424285054206848, "learning_rate": 6.359350065283622e-05, "loss": 0.0164, "step": 10138 }, { "epoch": 1.0991977450130095, "grad_norm": 0.21307273209095, "learning_rate": 6.358987378499927e-05, "loss": 0.0136, "step": 10139 }, { "epoch": 1.0993061578490892, "grad_norm": 0.2424747794866562, "learning_rate": 6.358624691716235e-05, "loss": 0.0788, "step": 10140 }, { "epoch": 1.0994145706851692, "grad_norm": 0.2083887755870819, "learning_rate": 6.35826200493254e-05, "loss": 0.0197, "step": 10141 }, { "epoch": 1.099522983521249, "grad_norm": 0.25020796060562134, "learning_rate": 6.357899318148846e-05, "loss": 0.0118, "step": 10142 }, { "epoch": 1.0996313963573288, "grad_norm": 0.23061603307724, "learning_rate": 6.357536631365153e-05, "loss": 0.0283, "step": 10143 }, { "epoch": 1.0997398091934085, "grad_norm": 0.3116014003753662, "learning_rate": 6.357173944581459e-05, "loss": 0.0268, "step": 10144 }, { "epoch": 1.0998482220294883, "grad_norm": 0.6786965727806091, "learning_rate": 6.356811257797766e-05, "loss": 0.0519, "step": 10145 }, { "epoch": 1.099956634865568, "grad_norm": 0.7865932583808899, "learning_rate": 6.356448571014073e-05, "loss": 0.0525, "step": 10146 }, { "epoch": 1.1000650477016478, "grad_norm": 0.3555353283882141, "learning_rate": 6.356085884230379e-05, "loss": 0.0213, "step": 10147 }, { "epoch": 1.1001734605377276, "grad_norm": 0.16005738079547882, "learning_rate": 6.355723197446686e-05, "loss": 0.01, "step": 10148 }, { "epoch": 1.1002818733738076, "grad_norm": 0.7074601650238037, "learning_rate": 6.355360510662992e-05, "loss": 0.0367, "step": 10149 }, { "epoch": 1.1003902862098873, "grad_norm": 0.23847153782844543, "learning_rate": 6.354997823879299e-05, "loss": 0.0185, "step": 10150 }, { "epoch": 1.100498699045967, "grad_norm": 0.37444567680358887, "learning_rate": 6.354635137095604e-05, "loss": 0.0171, "step": 10151 }, { "epoch": 1.1006071118820469, "grad_norm": 0.21038217842578888, "learning_rate": 6.354272450311912e-05, "loss": 0.0238, "step": 10152 }, { "epoch": 1.1007155247181266, "grad_norm": 0.6530075073242188, "learning_rate": 6.353909763528217e-05, "loss": 0.0693, "step": 10153 }, { "epoch": 1.1008239375542064, "grad_norm": 0.20527903735637665, "learning_rate": 6.353547076744524e-05, "loss": 0.0165, "step": 10154 }, { "epoch": 1.1009323503902861, "grad_norm": 0.2343370020389557, "learning_rate": 6.35318438996083e-05, "loss": 0.0237, "step": 10155 }, { "epoch": 1.101040763226366, "grad_norm": 0.26671692728996277, "learning_rate": 6.352821703177136e-05, "loss": 0.0243, "step": 10156 }, { "epoch": 1.101149176062446, "grad_norm": 0.11481800675392151, "learning_rate": 6.352459016393443e-05, "loss": 0.0078, "step": 10157 }, { "epoch": 1.1012575888985257, "grad_norm": 0.1927817016839981, "learning_rate": 6.352096329609749e-05, "loss": 0.0145, "step": 10158 }, { "epoch": 1.1013660017346054, "grad_norm": 0.2775815427303314, "learning_rate": 6.351733642826056e-05, "loss": 0.0289, "step": 10159 }, { "epoch": 1.1014744145706852, "grad_norm": 0.2787628173828125, "learning_rate": 6.351370956042361e-05, "loss": 0.0333, "step": 10160 }, { "epoch": 1.101582827406765, "grad_norm": 0.2512320876121521, "learning_rate": 6.351008269258669e-05, "loss": 0.0187, "step": 10161 }, { "epoch": 1.1016912402428447, "grad_norm": 1.0784363746643066, "learning_rate": 6.350645582474974e-05, "loss": 0.0612, "step": 10162 }, { "epoch": 1.1017996530789245, "grad_norm": 0.19484570622444153, "learning_rate": 6.350282895691281e-05, "loss": 0.0097, "step": 10163 }, { "epoch": 1.1019080659150045, "grad_norm": 0.28729352355003357, "learning_rate": 6.349920208907588e-05, "loss": 0.0301, "step": 10164 }, { "epoch": 1.1020164787510842, "grad_norm": 0.5051394701004028, "learning_rate": 6.349557522123894e-05, "loss": 0.0239, "step": 10165 }, { "epoch": 1.102124891587164, "grad_norm": 0.3344067335128784, "learning_rate": 6.349194835340201e-05, "loss": 0.042, "step": 10166 }, { "epoch": 1.1022333044232437, "grad_norm": 0.1748700737953186, "learning_rate": 6.348832148556507e-05, "loss": 0.0122, "step": 10167 }, { "epoch": 1.1023417172593235, "grad_norm": 0.6593419313430786, "learning_rate": 6.348469461772814e-05, "loss": 0.0293, "step": 10168 }, { "epoch": 1.1024501300954033, "grad_norm": 0.22645226120948792, "learning_rate": 6.34810677498912e-05, "loss": 0.0184, "step": 10169 }, { "epoch": 1.102558542931483, "grad_norm": 0.414470911026001, "learning_rate": 6.347744088205426e-05, "loss": 0.0402, "step": 10170 }, { "epoch": 1.1026669557675628, "grad_norm": 0.3549105226993561, "learning_rate": 6.347381401421733e-05, "loss": 0.0477, "step": 10171 }, { "epoch": 1.1027753686036426, "grad_norm": 0.1673527956008911, "learning_rate": 6.347018714638038e-05, "loss": 0.0151, "step": 10172 }, { "epoch": 1.1028837814397225, "grad_norm": 0.6340863704681396, "learning_rate": 6.346656027854345e-05, "loss": 0.0628, "step": 10173 }, { "epoch": 1.1029921942758023, "grad_norm": 0.33421841263771057, "learning_rate": 6.346293341070651e-05, "loss": 0.0281, "step": 10174 }, { "epoch": 1.103100607111882, "grad_norm": 0.3044464886188507, "learning_rate": 6.345930654286958e-05, "loss": 0.0363, "step": 10175 }, { "epoch": 1.1032090199479618, "grad_norm": 0.5768956542015076, "learning_rate": 6.345567967503264e-05, "loss": 0.0497, "step": 10176 }, { "epoch": 1.1033174327840416, "grad_norm": 0.524469256401062, "learning_rate": 6.345205280719571e-05, "loss": 0.0346, "step": 10177 }, { "epoch": 1.1034258456201214, "grad_norm": 0.34222766757011414, "learning_rate": 6.344842593935877e-05, "loss": 0.0289, "step": 10178 }, { "epoch": 1.1035342584562011, "grad_norm": 0.4201396405696869, "learning_rate": 6.344479907152183e-05, "loss": 0.0213, "step": 10179 }, { "epoch": 1.103642671292281, "grad_norm": 0.3547569811344147, "learning_rate": 6.34411722036849e-05, "loss": 0.0196, "step": 10180 }, { "epoch": 1.1037510841283609, "grad_norm": 0.12357542663812637, "learning_rate": 6.343754533584797e-05, "loss": 0.0081, "step": 10181 }, { "epoch": 1.1038594969644406, "grad_norm": 0.31699997186660767, "learning_rate": 6.343391846801104e-05, "loss": 0.0323, "step": 10182 }, { "epoch": 1.1039679098005204, "grad_norm": 0.41851335763931274, "learning_rate": 6.34302916001741e-05, "loss": 0.0233, "step": 10183 }, { "epoch": 1.1040763226366002, "grad_norm": 0.19511061906814575, "learning_rate": 6.342666473233717e-05, "loss": 0.0121, "step": 10184 }, { "epoch": 1.10418473547268, "grad_norm": 0.231593519449234, "learning_rate": 6.342303786450022e-05, "loss": 0.029, "step": 10185 }, { "epoch": 1.1042931483087597, "grad_norm": 0.11870266497135162, "learning_rate": 6.341941099666328e-05, "loss": 0.0106, "step": 10186 }, { "epoch": 1.1044015611448394, "grad_norm": 0.33307674527168274, "learning_rate": 6.341578412882635e-05, "loss": 0.0112, "step": 10187 }, { "epoch": 1.1045099739809194, "grad_norm": 0.41868358850479126, "learning_rate": 6.341215726098941e-05, "loss": 0.0336, "step": 10188 }, { "epoch": 1.1046183868169992, "grad_norm": 0.44407302141189575, "learning_rate": 6.340853039315248e-05, "loss": 0.0348, "step": 10189 }, { "epoch": 1.104726799653079, "grad_norm": 0.14219403266906738, "learning_rate": 6.340490352531554e-05, "loss": 0.0186, "step": 10190 }, { "epoch": 1.1048352124891587, "grad_norm": 0.6283080577850342, "learning_rate": 6.340127665747861e-05, "loss": 0.0106, "step": 10191 }, { "epoch": 1.1049436253252385, "grad_norm": 0.13473033905029297, "learning_rate": 6.339764978964167e-05, "loss": 0.017, "step": 10192 }, { "epoch": 1.1050520381613183, "grad_norm": 0.2972780168056488, "learning_rate": 6.339402292180474e-05, "loss": 0.0258, "step": 10193 }, { "epoch": 1.105160450997398, "grad_norm": 0.06730814278125763, "learning_rate": 6.33903960539678e-05, "loss": 0.0068, "step": 10194 }, { "epoch": 1.1052688638334778, "grad_norm": 0.3267074525356293, "learning_rate": 6.338676918613085e-05, "loss": 0.0508, "step": 10195 }, { "epoch": 1.1053772766695578, "grad_norm": 0.4892812669277191, "learning_rate": 6.338314231829392e-05, "loss": 0.0861, "step": 10196 }, { "epoch": 1.1054856895056375, "grad_norm": 0.0806453675031662, "learning_rate": 6.337951545045698e-05, "loss": 0.0123, "step": 10197 }, { "epoch": 1.1055941023417173, "grad_norm": 0.3714821934700012, "learning_rate": 6.337588858262006e-05, "loss": 0.0229, "step": 10198 }, { "epoch": 1.105702515177797, "grad_norm": 0.1930965930223465, "learning_rate": 6.337226171478312e-05, "loss": 0.0493, "step": 10199 }, { "epoch": 1.1058109280138768, "grad_norm": 0.24986444413661957, "learning_rate": 6.336863484694618e-05, "loss": 0.027, "step": 10200 }, { "epoch": 1.1059193408499566, "grad_norm": 0.20970578491687775, "learning_rate": 6.336500797910925e-05, "loss": 0.007, "step": 10201 }, { "epoch": 1.1060277536860363, "grad_norm": 0.11918093264102936, "learning_rate": 6.336138111127231e-05, "loss": 0.011, "step": 10202 }, { "epoch": 1.1061361665221163, "grad_norm": 0.1562749743461609, "learning_rate": 6.335775424343538e-05, "loss": 0.0179, "step": 10203 }, { "epoch": 1.106244579358196, "grad_norm": 0.8353657126426697, "learning_rate": 6.335412737559844e-05, "loss": 0.0513, "step": 10204 }, { "epoch": 1.1063529921942759, "grad_norm": 0.18148446083068848, "learning_rate": 6.33505005077615e-05, "loss": 0.0144, "step": 10205 }, { "epoch": 1.1064614050303556, "grad_norm": 0.30223503708839417, "learning_rate": 6.334687363992456e-05, "loss": 0.0434, "step": 10206 }, { "epoch": 1.1065698178664354, "grad_norm": 0.32439976930618286, "learning_rate": 6.334324677208763e-05, "loss": 0.0368, "step": 10207 }, { "epoch": 1.1066782307025151, "grad_norm": 0.1615724116563797, "learning_rate": 6.333961990425069e-05, "loss": 0.0082, "step": 10208 }, { "epoch": 1.106786643538595, "grad_norm": 0.337209552526474, "learning_rate": 6.333599303641375e-05, "loss": 0.0291, "step": 10209 }, { "epoch": 1.1068950563746747, "grad_norm": 0.33062371611595154, "learning_rate": 6.333236616857682e-05, "loss": 0.0308, "step": 10210 }, { "epoch": 1.1070034692107547, "grad_norm": 0.6149241328239441, "learning_rate": 6.332873930073988e-05, "loss": 0.0515, "step": 10211 }, { "epoch": 1.1071118820468344, "grad_norm": 0.20267851650714874, "learning_rate": 6.332511243290295e-05, "loss": 0.0134, "step": 10212 }, { "epoch": 1.1072202948829142, "grad_norm": 0.28985315561294556, "learning_rate": 6.3321485565066e-05, "loss": 0.0162, "step": 10213 }, { "epoch": 1.107328707718994, "grad_norm": 0.6741918325424194, "learning_rate": 6.331785869722908e-05, "loss": 0.0615, "step": 10214 }, { "epoch": 1.1074371205550737, "grad_norm": 0.2942730486392975, "learning_rate": 6.331423182939215e-05, "loss": 0.0224, "step": 10215 }, { "epoch": 1.1075455333911535, "grad_norm": 0.20246338844299316, "learning_rate": 6.33106049615552e-05, "loss": 0.0197, "step": 10216 }, { "epoch": 1.1076539462272332, "grad_norm": 0.4329129755496979, "learning_rate": 6.330697809371828e-05, "loss": 0.0186, "step": 10217 }, { "epoch": 1.107762359063313, "grad_norm": 0.7742851972579956, "learning_rate": 6.330335122588133e-05, "loss": 0.0231, "step": 10218 }, { "epoch": 1.107870771899393, "grad_norm": 0.14818310737609863, "learning_rate": 6.32997243580444e-05, "loss": 0.0109, "step": 10219 }, { "epoch": 1.1079791847354727, "grad_norm": 0.07564400881528854, "learning_rate": 6.329609749020746e-05, "loss": 0.0054, "step": 10220 }, { "epoch": 1.1080875975715525, "grad_norm": 0.25428229570388794, "learning_rate": 6.329247062237053e-05, "loss": 0.0205, "step": 10221 }, { "epoch": 1.1081960104076323, "grad_norm": 0.1546967327594757, "learning_rate": 6.328884375453359e-05, "loss": 0.0078, "step": 10222 }, { "epoch": 1.108304423243712, "grad_norm": 0.42705604434013367, "learning_rate": 6.328521688669665e-05, "loss": 0.0356, "step": 10223 }, { "epoch": 1.1084128360797918, "grad_norm": 0.12154144793748856, "learning_rate": 6.328159001885972e-05, "loss": 0.0149, "step": 10224 }, { "epoch": 1.1085212489158716, "grad_norm": 0.27137646079063416, "learning_rate": 6.327796315102277e-05, "loss": 0.0082, "step": 10225 }, { "epoch": 1.1086296617519515, "grad_norm": 0.60773104429245, "learning_rate": 6.327433628318585e-05, "loss": 0.0416, "step": 10226 }, { "epoch": 1.1087380745880313, "grad_norm": 0.17354077100753784, "learning_rate": 6.32707094153489e-05, "loss": 0.0202, "step": 10227 }, { "epoch": 1.108846487424111, "grad_norm": 0.23534145951271057, "learning_rate": 6.326708254751197e-05, "loss": 0.0134, "step": 10228 }, { "epoch": 1.1089549002601908, "grad_norm": 0.21473358571529388, "learning_rate": 6.326345567967503e-05, "loss": 0.0126, "step": 10229 }, { "epoch": 1.1090633130962706, "grad_norm": 0.29655641317367554, "learning_rate": 6.32598288118381e-05, "loss": 0.0475, "step": 10230 }, { "epoch": 1.1091717259323504, "grad_norm": 0.15837062895298004, "learning_rate": 6.325620194400116e-05, "loss": 0.0105, "step": 10231 }, { "epoch": 1.1092801387684301, "grad_norm": 0.2181999683380127, "learning_rate": 6.325257507616422e-05, "loss": 0.0125, "step": 10232 }, { "epoch": 1.1093885516045099, "grad_norm": 0.6264249682426453, "learning_rate": 6.32489482083273e-05, "loss": 0.0375, "step": 10233 }, { "epoch": 1.1094969644405897, "grad_norm": 0.14766739308834076, "learning_rate": 6.324532134049036e-05, "loss": 0.0103, "step": 10234 }, { "epoch": 1.1096053772766696, "grad_norm": 0.1756710708141327, "learning_rate": 6.324169447265343e-05, "loss": 0.0146, "step": 10235 }, { "epoch": 1.1097137901127494, "grad_norm": 0.09291522949934006, "learning_rate": 6.323806760481649e-05, "loss": 0.0036, "step": 10236 }, { "epoch": 1.1098222029488292, "grad_norm": 0.6492974758148193, "learning_rate": 6.323444073697954e-05, "loss": 0.0559, "step": 10237 }, { "epoch": 1.109930615784909, "grad_norm": 0.27220478653907776, "learning_rate": 6.323081386914261e-05, "loss": 0.0291, "step": 10238 }, { "epoch": 1.1100390286209887, "grad_norm": 0.6588319540023804, "learning_rate": 6.322718700130567e-05, "loss": 0.0394, "step": 10239 }, { "epoch": 1.1101474414570685, "grad_norm": 0.6800788640975952, "learning_rate": 6.322356013346874e-05, "loss": 0.0806, "step": 10240 }, { "epoch": 1.1102558542931482, "grad_norm": 0.17166391015052795, "learning_rate": 6.32199332656318e-05, "loss": 0.0105, "step": 10241 }, { "epoch": 1.1103642671292282, "grad_norm": 0.3613053560256958, "learning_rate": 6.321630639779487e-05, "loss": 0.0253, "step": 10242 }, { "epoch": 1.110472679965308, "grad_norm": 0.2448170781135559, "learning_rate": 6.321267952995793e-05, "loss": 0.0323, "step": 10243 }, { "epoch": 1.1105810928013877, "grad_norm": 0.4852955937385559, "learning_rate": 6.3209052662121e-05, "loss": 0.0443, "step": 10244 }, { "epoch": 1.1106895056374675, "grad_norm": 0.27639105916023254, "learning_rate": 6.320542579428406e-05, "loss": 0.0092, "step": 10245 }, { "epoch": 1.1107979184735473, "grad_norm": 0.2476038932800293, "learning_rate": 6.320179892644711e-05, "loss": 0.0102, "step": 10246 }, { "epoch": 1.110906331309627, "grad_norm": 0.06976521760225296, "learning_rate": 6.319817205861018e-05, "loss": 0.0044, "step": 10247 }, { "epoch": 1.1110147441457068, "grad_norm": 0.31692904233932495, "learning_rate": 6.319454519077324e-05, "loss": 0.0444, "step": 10248 }, { "epoch": 1.1111231569817865, "grad_norm": 0.36412563920021057, "learning_rate": 6.319091832293631e-05, "loss": 0.0302, "step": 10249 }, { "epoch": 1.1112315698178665, "grad_norm": 0.3698088228702545, "learning_rate": 6.318729145509938e-05, "loss": 0.0273, "step": 10250 }, { "epoch": 1.1113399826539463, "grad_norm": 0.1598314344882965, "learning_rate": 6.318366458726244e-05, "loss": 0.0159, "step": 10251 }, { "epoch": 1.111448395490026, "grad_norm": 0.11940030753612518, "learning_rate": 6.318003771942551e-05, "loss": 0.0069, "step": 10252 }, { "epoch": 1.1115568083261058, "grad_norm": 0.07216157764196396, "learning_rate": 6.317641085158857e-05, "loss": 0.0086, "step": 10253 }, { "epoch": 1.1116652211621856, "grad_norm": 0.3598399758338928, "learning_rate": 6.317278398375164e-05, "loss": 0.0253, "step": 10254 }, { "epoch": 1.1117736339982653, "grad_norm": 0.5250319242477417, "learning_rate": 6.31691571159147e-05, "loss": 0.0345, "step": 10255 }, { "epoch": 1.111882046834345, "grad_norm": 0.3049331605434418, "learning_rate": 6.316553024807777e-05, "loss": 0.0286, "step": 10256 }, { "epoch": 1.1119904596704249, "grad_norm": 0.9061501026153564, "learning_rate": 6.316190338024083e-05, "loss": 0.0294, "step": 10257 }, { "epoch": 1.1120988725065049, "grad_norm": 0.20982974767684937, "learning_rate": 6.31582765124039e-05, "loss": 0.0131, "step": 10258 }, { "epoch": 1.1122072853425846, "grad_norm": 0.529334306716919, "learning_rate": 6.315464964456695e-05, "loss": 0.043, "step": 10259 }, { "epoch": 1.1123156981786644, "grad_norm": 0.6324013471603394, "learning_rate": 6.315102277673001e-05, "loss": 0.0199, "step": 10260 }, { "epoch": 1.1124241110147441, "grad_norm": 0.3717574179172516, "learning_rate": 6.314739590889308e-05, "loss": 0.0498, "step": 10261 }, { "epoch": 1.112532523850824, "grad_norm": 0.21268750727176666, "learning_rate": 6.314376904105614e-05, "loss": 0.0227, "step": 10262 }, { "epoch": 1.1126409366869037, "grad_norm": 0.32805225253105164, "learning_rate": 6.314014217321921e-05, "loss": 0.0445, "step": 10263 }, { "epoch": 1.1127493495229834, "grad_norm": 0.12687982618808746, "learning_rate": 6.313651530538227e-05, "loss": 0.0061, "step": 10264 }, { "epoch": 1.1128577623590634, "grad_norm": 0.40231791138648987, "learning_rate": 6.313288843754534e-05, "loss": 0.0228, "step": 10265 }, { "epoch": 1.1129661751951432, "grad_norm": 0.1803501844406128, "learning_rate": 6.31292615697084e-05, "loss": 0.0131, "step": 10266 }, { "epoch": 1.113074588031223, "grad_norm": 0.0787627324461937, "learning_rate": 6.312563470187147e-05, "loss": 0.0055, "step": 10267 }, { "epoch": 1.1131830008673027, "grad_norm": 0.12923163175582886, "learning_rate": 6.312200783403454e-05, "loss": 0.0075, "step": 10268 }, { "epoch": 1.1132914137033825, "grad_norm": 0.32103845477104187, "learning_rate": 6.31183809661976e-05, "loss": 0.0263, "step": 10269 }, { "epoch": 1.1133998265394622, "grad_norm": 0.36799123883247375, "learning_rate": 6.311475409836067e-05, "loss": 0.034, "step": 10270 }, { "epoch": 1.113508239375542, "grad_norm": 0.25217774510383606, "learning_rate": 6.311112723052372e-05, "loss": 0.0147, "step": 10271 }, { "epoch": 1.1136166522116218, "grad_norm": 0.3550906479358673, "learning_rate": 6.31075003626868e-05, "loss": 0.0319, "step": 10272 }, { "epoch": 1.1137250650477017, "grad_norm": 0.25684186816215515, "learning_rate": 6.310387349484985e-05, "loss": 0.0193, "step": 10273 }, { "epoch": 1.1138334778837815, "grad_norm": 0.4257678687572479, "learning_rate": 6.310024662701291e-05, "loss": 0.0297, "step": 10274 }, { "epoch": 1.1139418907198613, "grad_norm": 0.7985143065452576, "learning_rate": 6.309661975917598e-05, "loss": 0.0383, "step": 10275 }, { "epoch": 1.114050303555941, "grad_norm": 0.31509947776794434, "learning_rate": 6.309299289133904e-05, "loss": 0.029, "step": 10276 }, { "epoch": 1.1141587163920208, "grad_norm": 0.03860097751021385, "learning_rate": 6.308936602350211e-05, "loss": 0.0031, "step": 10277 }, { "epoch": 1.1142671292281006, "grad_norm": 0.22989876568317413, "learning_rate": 6.308573915566517e-05, "loss": 0.019, "step": 10278 }, { "epoch": 1.1143755420641803, "grad_norm": 0.4744792878627777, "learning_rate": 6.308211228782824e-05, "loss": 0.023, "step": 10279 }, { "epoch": 1.11448395490026, "grad_norm": 0.3810282349586487, "learning_rate": 6.30784854199913e-05, "loss": 0.0148, "step": 10280 }, { "epoch": 1.11459236773634, "grad_norm": 0.3608677387237549, "learning_rate": 6.307485855215436e-05, "loss": 0.0175, "step": 10281 }, { "epoch": 1.1147007805724198, "grad_norm": 0.24699310958385468, "learning_rate": 6.307123168431742e-05, "loss": 0.0239, "step": 10282 }, { "epoch": 1.1148091934084996, "grad_norm": 0.39790448546409607, "learning_rate": 6.306760481648048e-05, "loss": 0.048, "step": 10283 }, { "epoch": 1.1149176062445794, "grad_norm": 0.1511094719171524, "learning_rate": 6.306397794864355e-05, "loss": 0.0074, "step": 10284 }, { "epoch": 1.1150260190806591, "grad_norm": 0.07286589592695236, "learning_rate": 6.306035108080662e-05, "loss": 0.0035, "step": 10285 }, { "epoch": 1.1151344319167389, "grad_norm": 0.42539089918136597, "learning_rate": 6.305672421296969e-05, "loss": 0.0163, "step": 10286 }, { "epoch": 1.1152428447528187, "grad_norm": 0.2513349652290344, "learning_rate": 6.305309734513275e-05, "loss": 0.0198, "step": 10287 }, { "epoch": 1.1153512575888986, "grad_norm": 0.2422035038471222, "learning_rate": 6.304947047729582e-05, "loss": 0.0123, "step": 10288 }, { "epoch": 1.1154596704249784, "grad_norm": 0.3743007183074951, "learning_rate": 6.304584360945888e-05, "loss": 0.0225, "step": 10289 }, { "epoch": 1.1155680832610582, "grad_norm": 0.6516571044921875, "learning_rate": 6.304221674162193e-05, "loss": 0.0734, "step": 10290 }, { "epoch": 1.115676496097138, "grad_norm": 0.16966493427753448, "learning_rate": 6.3038589873785e-05, "loss": 0.0168, "step": 10291 }, { "epoch": 1.1157849089332177, "grad_norm": 0.7274469137191772, "learning_rate": 6.303496300594806e-05, "loss": 0.0431, "step": 10292 }, { "epoch": 1.1158933217692975, "grad_norm": 0.4885851740837097, "learning_rate": 6.303133613811113e-05, "loss": 0.074, "step": 10293 }, { "epoch": 1.1160017346053772, "grad_norm": 0.25634703040122986, "learning_rate": 6.302770927027419e-05, "loss": 0.0224, "step": 10294 }, { "epoch": 1.116110147441457, "grad_norm": 0.37574467062950134, "learning_rate": 6.302408240243726e-05, "loss": 0.0289, "step": 10295 }, { "epoch": 1.116218560277537, "grad_norm": 0.21487191319465637, "learning_rate": 6.302045553460032e-05, "loss": 0.0324, "step": 10296 }, { "epoch": 1.1163269731136167, "grad_norm": 0.25833287835121155, "learning_rate": 6.301682866676339e-05, "loss": 0.0614, "step": 10297 }, { "epoch": 1.1164353859496965, "grad_norm": 0.6265614032745361, "learning_rate": 6.301320179892645e-05, "loss": 0.0351, "step": 10298 }, { "epoch": 1.1165437987857763, "grad_norm": 0.14241230487823486, "learning_rate": 6.30095749310895e-05, "loss": 0.0172, "step": 10299 }, { "epoch": 1.116652211621856, "grad_norm": 0.4717806875705719, "learning_rate": 6.300594806325258e-05, "loss": 0.0459, "step": 10300 }, { "epoch": 1.1167606244579358, "grad_norm": 0.7005836963653564, "learning_rate": 6.300232119541563e-05, "loss": 0.0738, "step": 10301 }, { "epoch": 1.1168690372940155, "grad_norm": 0.10813872516155243, "learning_rate": 6.299869432757872e-05, "loss": 0.0088, "step": 10302 }, { "epoch": 1.1169774501300953, "grad_norm": 0.1732468456029892, "learning_rate": 6.299506745974177e-05, "loss": 0.0104, "step": 10303 }, { "epoch": 1.1170858629661753, "grad_norm": 0.05084861442446709, "learning_rate": 6.299144059190483e-05, "loss": 0.0039, "step": 10304 }, { "epoch": 1.117194275802255, "grad_norm": 0.24344995617866516, "learning_rate": 6.29878137240679e-05, "loss": 0.024, "step": 10305 }, { "epoch": 1.1173026886383348, "grad_norm": 0.29274946451187134, "learning_rate": 6.298418685623096e-05, "loss": 0.0166, "step": 10306 }, { "epoch": 1.1174111014744146, "grad_norm": 0.45039284229278564, "learning_rate": 6.298055998839403e-05, "loss": 0.0601, "step": 10307 }, { "epoch": 1.1175195143104943, "grad_norm": 0.3445437550544739, "learning_rate": 6.297693312055709e-05, "loss": 0.0281, "step": 10308 }, { "epoch": 1.117627927146574, "grad_norm": 0.19488771259784698, "learning_rate": 6.297330625272016e-05, "loss": 0.021, "step": 10309 }, { "epoch": 1.1177363399826539, "grad_norm": 0.5654057264328003, "learning_rate": 6.296967938488322e-05, "loss": 0.0309, "step": 10310 }, { "epoch": 1.1178447528187339, "grad_norm": 0.24793629348278046, "learning_rate": 6.296605251704629e-05, "loss": 0.0409, "step": 10311 }, { "epoch": 1.1179531656548136, "grad_norm": 0.20547203719615936, "learning_rate": 6.296242564920934e-05, "loss": 0.0103, "step": 10312 }, { "epoch": 1.1180615784908934, "grad_norm": 0.31991609930992126, "learning_rate": 6.29587987813724e-05, "loss": 0.0213, "step": 10313 }, { "epoch": 1.1181699913269731, "grad_norm": 0.07428397238254547, "learning_rate": 6.295517191353547e-05, "loss": 0.0075, "step": 10314 }, { "epoch": 1.118278404163053, "grad_norm": 0.563364565372467, "learning_rate": 6.295154504569853e-05, "loss": 0.0257, "step": 10315 }, { "epoch": 1.1183868169991327, "grad_norm": 0.20835578441619873, "learning_rate": 6.29479181778616e-05, "loss": 0.0124, "step": 10316 }, { "epoch": 1.1184952298352124, "grad_norm": 0.09032239019870758, "learning_rate": 6.294429131002466e-05, "loss": 0.0061, "step": 10317 }, { "epoch": 1.1186036426712922, "grad_norm": 0.03567418083548546, "learning_rate": 6.294066444218773e-05, "loss": 0.0021, "step": 10318 }, { "epoch": 1.118712055507372, "grad_norm": 0.5022011995315552, "learning_rate": 6.29370375743508e-05, "loss": 0.0881, "step": 10319 }, { "epoch": 1.118820468343452, "grad_norm": 0.27856355905532837, "learning_rate": 6.293341070651386e-05, "loss": 0.03, "step": 10320 }, { "epoch": 1.1189288811795317, "grad_norm": 0.2679654657840729, "learning_rate": 6.292978383867693e-05, "loss": 0.0172, "step": 10321 }, { "epoch": 1.1190372940156115, "grad_norm": 0.3449089229106903, "learning_rate": 6.292615697083999e-05, "loss": 0.022, "step": 10322 }, { "epoch": 1.1191457068516912, "grad_norm": 0.3613404631614685, "learning_rate": 6.292253010300306e-05, "loss": 0.021, "step": 10323 }, { "epoch": 1.119254119687771, "grad_norm": 0.2588426470756531, "learning_rate": 6.291890323516611e-05, "loss": 0.0385, "step": 10324 }, { "epoch": 1.1193625325238508, "grad_norm": 0.18536800146102905, "learning_rate": 6.291527636732918e-05, "loss": 0.0072, "step": 10325 }, { "epoch": 1.1194709453599305, "grad_norm": 0.22570419311523438, "learning_rate": 6.291164949949224e-05, "loss": 0.0151, "step": 10326 }, { "epoch": 1.1195793581960105, "grad_norm": 0.21773342788219452, "learning_rate": 6.29080226316553e-05, "loss": 0.0078, "step": 10327 }, { "epoch": 1.1196877710320903, "grad_norm": 0.13691537082195282, "learning_rate": 6.290439576381837e-05, "loss": 0.0143, "step": 10328 }, { "epoch": 1.11979618386817, "grad_norm": 0.5764086246490479, "learning_rate": 6.290076889598143e-05, "loss": 0.0267, "step": 10329 }, { "epoch": 1.1199045967042498, "grad_norm": 0.22902023792266846, "learning_rate": 6.28971420281445e-05, "loss": 0.0197, "step": 10330 }, { "epoch": 1.1200130095403296, "grad_norm": 0.19140486419200897, "learning_rate": 6.289351516030756e-05, "loss": 0.0211, "step": 10331 }, { "epoch": 1.1201214223764093, "grad_norm": 0.5756961107254028, "learning_rate": 6.288988829247063e-05, "loss": 0.0403, "step": 10332 }, { "epoch": 1.120229835212489, "grad_norm": 0.179105743765831, "learning_rate": 6.288626142463368e-05, "loss": 0.0265, "step": 10333 }, { "epoch": 1.1203382480485689, "grad_norm": 0.07344680279493332, "learning_rate": 6.288263455679676e-05, "loss": 0.0058, "step": 10334 }, { "epoch": 1.1204466608846488, "grad_norm": 0.31691041588783264, "learning_rate": 6.287900768895981e-05, "loss": 0.0263, "step": 10335 }, { "epoch": 1.1205550737207286, "grad_norm": 0.4451886713504791, "learning_rate": 6.287538082112288e-05, "loss": 0.0568, "step": 10336 }, { "epoch": 1.1206634865568084, "grad_norm": 0.12023657560348511, "learning_rate": 6.287175395328595e-05, "loss": 0.0113, "step": 10337 }, { "epoch": 1.1207718993928881, "grad_norm": 0.34218931198120117, "learning_rate": 6.286812708544901e-05, "loss": 0.0595, "step": 10338 }, { "epoch": 1.120880312228968, "grad_norm": 0.1669248342514038, "learning_rate": 6.286450021761208e-05, "loss": 0.014, "step": 10339 }, { "epoch": 1.1209887250650477, "grad_norm": 0.13094662129878998, "learning_rate": 6.286087334977514e-05, "loss": 0.005, "step": 10340 }, { "epoch": 1.1210971379011274, "grad_norm": 0.28201010823249817, "learning_rate": 6.28572464819382e-05, "loss": 0.0178, "step": 10341 }, { "epoch": 1.1212055507372072, "grad_norm": 0.3631715774536133, "learning_rate": 6.285361961410127e-05, "loss": 0.0141, "step": 10342 }, { "epoch": 1.1213139635732872, "grad_norm": 0.3819999694824219, "learning_rate": 6.284999274626433e-05, "loss": 0.0384, "step": 10343 }, { "epoch": 1.121422376409367, "grad_norm": 0.3142751455307007, "learning_rate": 6.28463658784274e-05, "loss": 0.0214, "step": 10344 }, { "epoch": 1.1215307892454467, "grad_norm": 0.2747034430503845, "learning_rate": 6.284273901059045e-05, "loss": 0.0065, "step": 10345 }, { "epoch": 1.1216392020815265, "grad_norm": 0.40469807386398315, "learning_rate": 6.283911214275352e-05, "loss": 0.0321, "step": 10346 }, { "epoch": 1.1217476149176062, "grad_norm": 0.46774283051490784, "learning_rate": 6.283548527491658e-05, "loss": 0.03, "step": 10347 }, { "epoch": 1.121856027753686, "grad_norm": 0.04174647852778435, "learning_rate": 6.283185840707965e-05, "loss": 0.0055, "step": 10348 }, { "epoch": 1.1219644405897657, "grad_norm": 0.41426482796669006, "learning_rate": 6.282823153924271e-05, "loss": 0.0344, "step": 10349 }, { "epoch": 1.1220728534258457, "grad_norm": 0.13605144619941711, "learning_rate": 6.282460467140577e-05, "loss": 0.0061, "step": 10350 }, { "epoch": 1.1221812662619255, "grad_norm": 0.1495453417301178, "learning_rate": 6.282097780356884e-05, "loss": 0.0059, "step": 10351 }, { "epoch": 1.1222896790980053, "grad_norm": 0.40096619725227356, "learning_rate": 6.28173509357319e-05, "loss": 0.0224, "step": 10352 }, { "epoch": 1.122398091934085, "grad_norm": 0.44295191764831543, "learning_rate": 6.281372406789497e-05, "loss": 0.037, "step": 10353 }, { "epoch": 1.1225065047701648, "grad_norm": 0.32582008838653564, "learning_rate": 6.281009720005804e-05, "loss": 0.0393, "step": 10354 }, { "epoch": 1.1226149176062445, "grad_norm": 0.5100042819976807, "learning_rate": 6.28064703322211e-05, "loss": 0.0755, "step": 10355 }, { "epoch": 1.1227233304423243, "grad_norm": 0.2675822675228119, "learning_rate": 6.280284346438417e-05, "loss": 0.0416, "step": 10356 }, { "epoch": 1.122831743278404, "grad_norm": 0.5280755162239075, "learning_rate": 6.279921659654722e-05, "loss": 0.0312, "step": 10357 }, { "epoch": 1.122940156114484, "grad_norm": 0.5102770924568176, "learning_rate": 6.27955897287103e-05, "loss": 0.0423, "step": 10358 }, { "epoch": 1.1230485689505638, "grad_norm": 0.24265895783901215, "learning_rate": 6.279196286087335e-05, "loss": 0.0151, "step": 10359 }, { "epoch": 1.1231569817866436, "grad_norm": 0.10971636325120926, "learning_rate": 6.278833599303642e-05, "loss": 0.0099, "step": 10360 }, { "epoch": 1.1232653946227233, "grad_norm": 0.40353158116340637, "learning_rate": 6.278470912519948e-05, "loss": 0.0263, "step": 10361 }, { "epoch": 1.123373807458803, "grad_norm": 0.029332716017961502, "learning_rate": 6.278108225736255e-05, "loss": 0.0016, "step": 10362 }, { "epoch": 1.1234822202948829, "grad_norm": 0.6329399347305298, "learning_rate": 6.277745538952561e-05, "loss": 0.0407, "step": 10363 }, { "epoch": 1.1235906331309626, "grad_norm": 0.3149715065956116, "learning_rate": 6.277382852168866e-05, "loss": 0.017, "step": 10364 }, { "epoch": 1.1236990459670424, "grad_norm": 0.6397053003311157, "learning_rate": 6.277020165385174e-05, "loss": 0.022, "step": 10365 }, { "epoch": 1.1238074588031224, "grad_norm": 0.1894635260105133, "learning_rate": 6.276657478601479e-05, "loss": 0.0073, "step": 10366 }, { "epoch": 1.1239158716392021, "grad_norm": 0.25455334782600403, "learning_rate": 6.276294791817786e-05, "loss": 0.0298, "step": 10367 }, { "epoch": 1.124024284475282, "grad_norm": 0.6917669773101807, "learning_rate": 6.275932105034092e-05, "loss": 0.0351, "step": 10368 }, { "epoch": 1.1241326973113617, "grad_norm": 0.2002905011177063, "learning_rate": 6.275569418250399e-05, "loss": 0.0384, "step": 10369 }, { "epoch": 1.1242411101474414, "grad_norm": 0.3457409739494324, "learning_rate": 6.275206731466705e-05, "loss": 0.0203, "step": 10370 }, { "epoch": 1.1243495229835212, "grad_norm": 0.48988696932792664, "learning_rate": 6.274844044683012e-05, "loss": 0.0468, "step": 10371 }, { "epoch": 1.124457935819601, "grad_norm": 0.10857142508029938, "learning_rate": 6.274481357899319e-05, "loss": 0.0067, "step": 10372 }, { "epoch": 1.124566348655681, "grad_norm": 0.5274525284767151, "learning_rate": 6.274118671115625e-05, "loss": 0.0219, "step": 10373 }, { "epoch": 1.1246747614917607, "grad_norm": 0.09096593409776688, "learning_rate": 6.273755984331932e-05, "loss": 0.0057, "step": 10374 }, { "epoch": 1.1247831743278405, "grad_norm": 0.3645985722541809, "learning_rate": 6.273393297548238e-05, "loss": 0.032, "step": 10375 }, { "epoch": 1.1248915871639202, "grad_norm": 0.19670023024082184, "learning_rate": 6.273030610764545e-05, "loss": 0.0154, "step": 10376 }, { "epoch": 1.125, "grad_norm": 0.37700504064559937, "learning_rate": 6.27266792398085e-05, "loss": 0.0232, "step": 10377 }, { "epoch": 1.1251084128360798, "grad_norm": 0.18639224767684937, "learning_rate": 6.272305237197158e-05, "loss": 0.017, "step": 10378 }, { "epoch": 1.1252168256721595, "grad_norm": 0.7523064613342285, "learning_rate": 6.271942550413463e-05, "loss": 0.051, "step": 10379 }, { "epoch": 1.1253252385082393, "grad_norm": 0.07135703414678574, "learning_rate": 6.271579863629769e-05, "loss": 0.0047, "step": 10380 }, { "epoch": 1.125433651344319, "grad_norm": 0.2791889011859894, "learning_rate": 6.271217176846076e-05, "loss": 0.0272, "step": 10381 }, { "epoch": 1.125542064180399, "grad_norm": 0.3136996328830719, "learning_rate": 6.270854490062382e-05, "loss": 0.0106, "step": 10382 }, { "epoch": 1.1256504770164788, "grad_norm": 0.3455992639064789, "learning_rate": 6.270491803278689e-05, "loss": 0.0838, "step": 10383 }, { "epoch": 1.1257588898525586, "grad_norm": 0.37913551926612854, "learning_rate": 6.270129116494995e-05, "loss": 0.0451, "step": 10384 }, { "epoch": 1.1258673026886383, "grad_norm": 0.6683859825134277, "learning_rate": 6.269766429711302e-05, "loss": 0.0228, "step": 10385 }, { "epoch": 1.125975715524718, "grad_norm": 0.18945349752902985, "learning_rate": 6.269403742927608e-05, "loss": 0.0125, "step": 10386 }, { "epoch": 1.1260841283607979, "grad_norm": 0.2155860811471939, "learning_rate": 6.269041056143913e-05, "loss": 0.0263, "step": 10387 }, { "epoch": 1.1261925411968776, "grad_norm": 0.20770445466041565, "learning_rate": 6.268678369360222e-05, "loss": 0.0137, "step": 10388 }, { "epoch": 1.1263009540329576, "grad_norm": 0.18197959661483765, "learning_rate": 6.268315682576527e-05, "loss": 0.0162, "step": 10389 }, { "epoch": 1.1264093668690374, "grad_norm": 0.21762843430042267, "learning_rate": 6.267952995792835e-05, "loss": 0.0267, "step": 10390 }, { "epoch": 1.1265177797051171, "grad_norm": 0.21435874700546265, "learning_rate": 6.26759030900914e-05, "loss": 0.0188, "step": 10391 }, { "epoch": 1.126626192541197, "grad_norm": 0.18787164986133575, "learning_rate": 6.267227622225447e-05, "loss": 0.0169, "step": 10392 }, { "epoch": 1.1267346053772767, "grad_norm": 0.32835131883621216, "learning_rate": 6.266864935441753e-05, "loss": 0.0521, "step": 10393 }, { "epoch": 1.1268430182133564, "grad_norm": 0.6821100115776062, "learning_rate": 6.266502248658059e-05, "loss": 0.0996, "step": 10394 }, { "epoch": 1.1269514310494362, "grad_norm": 0.30928993225097656, "learning_rate": 6.266139561874366e-05, "loss": 0.0221, "step": 10395 }, { "epoch": 1.1270598438855162, "grad_norm": 0.19961611926555634, "learning_rate": 6.265776875090672e-05, "loss": 0.0216, "step": 10396 }, { "epoch": 1.127168256721596, "grad_norm": 0.08361171185970306, "learning_rate": 6.265414188306979e-05, "loss": 0.0112, "step": 10397 }, { "epoch": 1.1272766695576757, "grad_norm": 0.44213932752609253, "learning_rate": 6.265051501523284e-05, "loss": 0.0426, "step": 10398 }, { "epoch": 1.1273850823937555, "grad_norm": 0.20905935764312744, "learning_rate": 6.264688814739592e-05, "loss": 0.0156, "step": 10399 }, { "epoch": 1.1274934952298352, "grad_norm": 0.13029798865318298, "learning_rate": 6.264326127955897e-05, "loss": 0.0208, "step": 10400 }, { "epoch": 1.127601908065915, "grad_norm": 0.1178261935710907, "learning_rate": 6.263963441172204e-05, "loss": 0.0117, "step": 10401 }, { "epoch": 1.1277103209019947, "grad_norm": 0.35109061002731323, "learning_rate": 6.26360075438851e-05, "loss": 0.0245, "step": 10402 }, { "epoch": 1.1278187337380745, "grad_norm": 0.3088604211807251, "learning_rate": 6.263238067604816e-05, "loss": 0.0606, "step": 10403 }, { "epoch": 1.1279271465741543, "grad_norm": 0.10932672023773193, "learning_rate": 6.262875380821123e-05, "loss": 0.0037, "step": 10404 }, { "epoch": 1.1280355594102343, "grad_norm": 0.343658983707428, "learning_rate": 6.26251269403743e-05, "loss": 0.0144, "step": 10405 }, { "epoch": 1.128143972246314, "grad_norm": 0.3910696506500244, "learning_rate": 6.262150007253737e-05, "loss": 0.0328, "step": 10406 }, { "epoch": 1.1282523850823938, "grad_norm": 0.14490190148353577, "learning_rate": 6.261787320470043e-05, "loss": 0.0083, "step": 10407 }, { "epoch": 1.1283607979184735, "grad_norm": 0.6545136570930481, "learning_rate": 6.261424633686349e-05, "loss": 0.0417, "step": 10408 }, { "epoch": 1.1284692107545533, "grad_norm": 0.4614863395690918, "learning_rate": 6.261061946902656e-05, "loss": 0.0118, "step": 10409 }, { "epoch": 1.128577623590633, "grad_norm": 0.44169872999191284, "learning_rate": 6.260699260118961e-05, "loss": 0.0692, "step": 10410 }, { "epoch": 1.1286860364267128, "grad_norm": 0.11181284487247467, "learning_rate": 6.260336573335268e-05, "loss": 0.0105, "step": 10411 }, { "epoch": 1.1287944492627928, "grad_norm": 0.07375538349151611, "learning_rate": 6.259973886551574e-05, "loss": 0.0098, "step": 10412 }, { "epoch": 1.1289028620988726, "grad_norm": 0.3576388657093048, "learning_rate": 6.259611199767881e-05, "loss": 0.0292, "step": 10413 }, { "epoch": 1.1290112749349523, "grad_norm": 0.15227359533309937, "learning_rate": 6.259248512984187e-05, "loss": 0.0211, "step": 10414 }, { "epoch": 1.129119687771032, "grad_norm": 0.2744569778442383, "learning_rate": 6.258885826200494e-05, "loss": 0.0216, "step": 10415 }, { "epoch": 1.1292281006071119, "grad_norm": 0.2382153421640396, "learning_rate": 6.2585231394168e-05, "loss": 0.019, "step": 10416 }, { "epoch": 1.1293365134431916, "grad_norm": 0.3871980309486389, "learning_rate": 6.258160452633106e-05, "loss": 0.0261, "step": 10417 }, { "epoch": 1.1294449262792714, "grad_norm": 0.1295880675315857, "learning_rate": 6.257797765849413e-05, "loss": 0.0113, "step": 10418 }, { "epoch": 1.1295533391153512, "grad_norm": 0.26439186930656433, "learning_rate": 6.257435079065718e-05, "loss": 0.0242, "step": 10419 }, { "epoch": 1.129661751951431, "grad_norm": 0.1925346404314041, "learning_rate": 6.257072392282025e-05, "loss": 0.0067, "step": 10420 }, { "epoch": 1.129770164787511, "grad_norm": 0.9652021527290344, "learning_rate": 6.256709705498331e-05, "loss": 0.0527, "step": 10421 }, { "epoch": 1.1298785776235907, "grad_norm": 0.5234551429748535, "learning_rate": 6.256347018714638e-05, "loss": 0.029, "step": 10422 }, { "epoch": 1.1299869904596704, "grad_norm": 0.06355813890695572, "learning_rate": 6.255984331930945e-05, "loss": 0.0047, "step": 10423 }, { "epoch": 1.1300954032957502, "grad_norm": 0.25727346539497375, "learning_rate": 6.255621645147251e-05, "loss": 0.0193, "step": 10424 }, { "epoch": 1.13020381613183, "grad_norm": 0.28230488300323486, "learning_rate": 6.255258958363558e-05, "loss": 0.0444, "step": 10425 }, { "epoch": 1.1303122289679097, "grad_norm": 0.3277103900909424, "learning_rate": 6.254896271579864e-05, "loss": 0.0116, "step": 10426 }, { "epoch": 1.1304206418039895, "grad_norm": 0.07373356819152832, "learning_rate": 6.254533584796171e-05, "loss": 0.0075, "step": 10427 }, { "epoch": 1.1305290546400695, "grad_norm": 0.06018185243010521, "learning_rate": 6.254170898012477e-05, "loss": 0.0047, "step": 10428 }, { "epoch": 1.1306374674761492, "grad_norm": 0.12114199250936508, "learning_rate": 6.253808211228784e-05, "loss": 0.0083, "step": 10429 }, { "epoch": 1.130745880312229, "grad_norm": 0.15896180272102356, "learning_rate": 6.25344552444509e-05, "loss": 0.0113, "step": 10430 }, { "epoch": 1.1308542931483088, "grad_norm": 0.17715993523597717, "learning_rate": 6.253082837661395e-05, "loss": 0.0137, "step": 10431 }, { "epoch": 1.1309627059843885, "grad_norm": 0.23708641529083252, "learning_rate": 6.252720150877702e-05, "loss": 0.0191, "step": 10432 }, { "epoch": 1.1310711188204683, "grad_norm": 0.5696302652359009, "learning_rate": 6.252357464094008e-05, "loss": 0.0527, "step": 10433 }, { "epoch": 1.131179531656548, "grad_norm": 0.18272008001804352, "learning_rate": 6.251994777310315e-05, "loss": 0.0218, "step": 10434 }, { "epoch": 1.131287944492628, "grad_norm": 0.35361191630363464, "learning_rate": 6.251632090526621e-05, "loss": 0.0286, "step": 10435 }, { "epoch": 1.1313963573287078, "grad_norm": 0.15833571553230286, "learning_rate": 6.251269403742928e-05, "loss": 0.0246, "step": 10436 }, { "epoch": 1.1315047701647876, "grad_norm": 0.2913338243961334, "learning_rate": 6.250906716959234e-05, "loss": 0.0422, "step": 10437 }, { "epoch": 1.1316131830008673, "grad_norm": 0.16026370227336884, "learning_rate": 6.250544030175541e-05, "loss": 0.015, "step": 10438 }, { "epoch": 1.131721595836947, "grad_norm": 0.03407140076160431, "learning_rate": 6.250181343391847e-05, "loss": 0.0026, "step": 10439 }, { "epoch": 1.1318300086730269, "grad_norm": 0.3354712426662445, "learning_rate": 6.249818656608154e-05, "loss": 0.0436, "step": 10440 }, { "epoch": 1.1319384215091066, "grad_norm": 0.48567190766334534, "learning_rate": 6.249455969824461e-05, "loss": 0.0176, "step": 10441 }, { "epoch": 1.1320468343451864, "grad_norm": 0.45532676577568054, "learning_rate": 6.249093283040766e-05, "loss": 0.0169, "step": 10442 }, { "epoch": 1.1321552471812661, "grad_norm": 0.04071959853172302, "learning_rate": 6.248730596257074e-05, "loss": 0.0021, "step": 10443 }, { "epoch": 1.1322636600173461, "grad_norm": 0.621724009513855, "learning_rate": 6.248367909473379e-05, "loss": 0.0649, "step": 10444 }, { "epoch": 1.132372072853426, "grad_norm": 0.3644142746925354, "learning_rate": 6.248005222689685e-05, "loss": 0.0246, "step": 10445 }, { "epoch": 1.1324804856895057, "grad_norm": 0.4275243580341339, "learning_rate": 6.247642535905992e-05, "loss": 0.0453, "step": 10446 }, { "epoch": 1.1325888985255854, "grad_norm": 0.3239775598049164, "learning_rate": 6.247279849122298e-05, "loss": 0.0211, "step": 10447 }, { "epoch": 1.1326973113616652, "grad_norm": 0.23091664910316467, "learning_rate": 6.246917162338605e-05, "loss": 0.0461, "step": 10448 }, { "epoch": 1.132805724197745, "grad_norm": 0.277742475271225, "learning_rate": 6.246554475554911e-05, "loss": 0.0231, "step": 10449 }, { "epoch": 1.1329141370338247, "grad_norm": 0.3149431347846985, "learning_rate": 6.246191788771218e-05, "loss": 0.0268, "step": 10450 }, { "epoch": 1.1330225498699047, "grad_norm": 0.8952751159667969, "learning_rate": 6.245829101987524e-05, "loss": 0.0642, "step": 10451 }, { "epoch": 1.1331309627059845, "grad_norm": 0.2922840118408203, "learning_rate": 6.24546641520383e-05, "loss": 0.0209, "step": 10452 }, { "epoch": 1.1332393755420642, "grad_norm": 0.08792657405138016, "learning_rate": 6.245103728420136e-05, "loss": 0.0033, "step": 10453 }, { "epoch": 1.133347788378144, "grad_norm": 0.0933837741613388, "learning_rate": 6.244741041636442e-05, "loss": 0.0081, "step": 10454 }, { "epoch": 1.1334562012142237, "grad_norm": 0.3097170293331146, "learning_rate": 6.244378354852749e-05, "loss": 0.0295, "step": 10455 }, { "epoch": 1.1335646140503035, "grad_norm": 0.42660337686538696, "learning_rate": 6.244015668069055e-05, "loss": 0.0494, "step": 10456 }, { "epoch": 1.1336730268863833, "grad_norm": 0.31429195404052734, "learning_rate": 6.243652981285363e-05, "loss": 0.0307, "step": 10457 }, { "epoch": 1.1337814397224633, "grad_norm": 0.21377237141132355, "learning_rate": 6.243290294501669e-05, "loss": 0.0202, "step": 10458 }, { "epoch": 1.133889852558543, "grad_norm": 0.16173505783081055, "learning_rate": 6.242927607717975e-05, "loss": 0.024, "step": 10459 }, { "epoch": 1.1339982653946228, "grad_norm": 0.5872366428375244, "learning_rate": 6.242564920934282e-05, "loss": 0.0385, "step": 10460 }, { "epoch": 1.1341066782307025, "grad_norm": 0.16521833837032318, "learning_rate": 6.242202234150588e-05, "loss": 0.0197, "step": 10461 }, { "epoch": 1.1342150910667823, "grad_norm": 0.31963688135147095, "learning_rate": 6.241839547366895e-05, "loss": 0.0341, "step": 10462 }, { "epoch": 1.134323503902862, "grad_norm": 0.49989160895347595, "learning_rate": 6.2414768605832e-05, "loss": 0.0451, "step": 10463 }, { "epoch": 1.1344319167389418, "grad_norm": 0.4070986807346344, "learning_rate": 6.241114173799508e-05, "loss": 0.0597, "step": 10464 }, { "epoch": 1.1345403295750216, "grad_norm": 0.2618432939052582, "learning_rate": 6.240751487015813e-05, "loss": 0.0296, "step": 10465 }, { "epoch": 1.1346487424111014, "grad_norm": 0.23942779004573822, "learning_rate": 6.24038880023212e-05, "loss": 0.0138, "step": 10466 }, { "epoch": 1.1347571552471813, "grad_norm": 0.4396039545536041, "learning_rate": 6.240026113448426e-05, "loss": 0.0165, "step": 10467 }, { "epoch": 1.1348655680832611, "grad_norm": 0.5287795066833496, "learning_rate": 6.239663426664732e-05, "loss": 0.0466, "step": 10468 }, { "epoch": 1.1349739809193409, "grad_norm": 0.101227305829525, "learning_rate": 6.239300739881039e-05, "loss": 0.0079, "step": 10469 }, { "epoch": 1.1350823937554206, "grad_norm": 0.08798938244581223, "learning_rate": 6.238938053097345e-05, "loss": 0.007, "step": 10470 }, { "epoch": 1.1351908065915004, "grad_norm": 0.11625483632087708, "learning_rate": 6.238575366313652e-05, "loss": 0.0135, "step": 10471 }, { "epoch": 1.1352992194275802, "grad_norm": 0.5406097769737244, "learning_rate": 6.238212679529957e-05, "loss": 0.0466, "step": 10472 }, { "epoch": 1.13540763226366, "grad_norm": 0.18997755646705627, "learning_rate": 6.237849992746265e-05, "loss": 0.0155, "step": 10473 }, { "epoch": 1.13551604509974, "grad_norm": 0.44411754608154297, "learning_rate": 6.23748730596257e-05, "loss": 0.0427, "step": 10474 }, { "epoch": 1.1356244579358197, "grad_norm": 0.3099486529827118, "learning_rate": 6.237124619178877e-05, "loss": 0.0185, "step": 10475 }, { "epoch": 1.1357328707718994, "grad_norm": 0.07597430050373077, "learning_rate": 6.236761932395184e-05, "loss": 0.0038, "step": 10476 }, { "epoch": 1.1358412836079792, "grad_norm": 0.639862596988678, "learning_rate": 6.23639924561149e-05, "loss": 0.033, "step": 10477 }, { "epoch": 1.135949696444059, "grad_norm": 0.2913588881492615, "learning_rate": 6.236036558827797e-05, "loss": 0.0313, "step": 10478 }, { "epoch": 1.1360581092801387, "grad_norm": 0.4957907795906067, "learning_rate": 6.235673872044103e-05, "loss": 0.0805, "step": 10479 }, { "epoch": 1.1361665221162185, "grad_norm": 0.22015073895454407, "learning_rate": 6.23531118526041e-05, "loss": 0.0177, "step": 10480 }, { "epoch": 1.1362749349522985, "grad_norm": 0.21771374344825745, "learning_rate": 6.234948498476716e-05, "loss": 0.017, "step": 10481 }, { "epoch": 1.1363833477883782, "grad_norm": 0.15566150844097137, "learning_rate": 6.234585811693023e-05, "loss": 0.0184, "step": 10482 }, { "epoch": 1.136491760624458, "grad_norm": 0.4550614356994629, "learning_rate": 6.234223124909329e-05, "loss": 0.0308, "step": 10483 }, { "epoch": 1.1366001734605378, "grad_norm": 0.15153174102306366, "learning_rate": 6.233860438125634e-05, "loss": 0.0183, "step": 10484 }, { "epoch": 1.1367085862966175, "grad_norm": 0.1538161188364029, "learning_rate": 6.233497751341941e-05, "loss": 0.0191, "step": 10485 }, { "epoch": 1.1368169991326973, "grad_norm": 0.09120545536279678, "learning_rate": 6.233135064558247e-05, "loss": 0.0072, "step": 10486 }, { "epoch": 1.136925411968777, "grad_norm": 0.22706204652786255, "learning_rate": 6.232772377774554e-05, "loss": 0.0156, "step": 10487 }, { "epoch": 1.1370338248048568, "grad_norm": 0.19765718281269073, "learning_rate": 6.23240969099086e-05, "loss": 0.0152, "step": 10488 }, { "epoch": 1.1371422376409366, "grad_norm": 0.06485424190759659, "learning_rate": 6.232047004207167e-05, "loss": 0.0045, "step": 10489 }, { "epoch": 1.1372506504770166, "grad_norm": 0.27262750267982483, "learning_rate": 6.231684317423473e-05, "loss": 0.0181, "step": 10490 }, { "epoch": 1.1373590633130963, "grad_norm": 0.17700137197971344, "learning_rate": 6.231321630639779e-05, "loss": 0.0369, "step": 10491 }, { "epoch": 1.137467476149176, "grad_norm": 0.06152978539466858, "learning_rate": 6.230958943856087e-05, "loss": 0.003, "step": 10492 }, { "epoch": 1.1375758889852559, "grad_norm": 0.25349554419517517, "learning_rate": 6.230596257072393e-05, "loss": 0.0363, "step": 10493 }, { "epoch": 1.1376843018213356, "grad_norm": 0.5058610439300537, "learning_rate": 6.2302335702887e-05, "loss": 0.0446, "step": 10494 }, { "epoch": 1.1377927146574154, "grad_norm": 0.23845860362052917, "learning_rate": 6.229870883505006e-05, "loss": 0.0106, "step": 10495 }, { "epoch": 1.1379011274934951, "grad_norm": 0.6029142737388611, "learning_rate": 6.229508196721313e-05, "loss": 0.0198, "step": 10496 }, { "epoch": 1.1380095403295751, "grad_norm": 0.2729704678058624, "learning_rate": 6.229145509937618e-05, "loss": 0.0102, "step": 10497 }, { "epoch": 1.138117953165655, "grad_norm": 0.31850749254226685, "learning_rate": 6.228782823153924e-05, "loss": 0.0284, "step": 10498 }, { "epoch": 1.1382263660017347, "grad_norm": 0.19752784073352814, "learning_rate": 6.228420136370231e-05, "loss": 0.0236, "step": 10499 }, { "epoch": 1.1383347788378144, "grad_norm": 0.12215101718902588, "learning_rate": 6.228057449586537e-05, "loss": 0.011, "step": 10500 }, { "epoch": 1.1384431916738942, "grad_norm": 0.5360267162322998, "learning_rate": 6.227694762802844e-05, "loss": 0.0618, "step": 10501 }, { "epoch": 1.138551604509974, "grad_norm": 0.17311814427375793, "learning_rate": 6.22733207601915e-05, "loss": 0.0107, "step": 10502 }, { "epoch": 1.1386600173460537, "grad_norm": 0.34741851687431335, "learning_rate": 6.226969389235457e-05, "loss": 0.0312, "step": 10503 }, { "epoch": 1.1387684301821335, "grad_norm": 0.10388323664665222, "learning_rate": 6.226606702451763e-05, "loss": 0.009, "step": 10504 }, { "epoch": 1.1388768430182132, "grad_norm": 0.43235835433006287, "learning_rate": 6.22624401566807e-05, "loss": 0.0456, "step": 10505 }, { "epoch": 1.1389852558542932, "grad_norm": 0.17396587133407593, "learning_rate": 6.225881328884375e-05, "loss": 0.0083, "step": 10506 }, { "epoch": 1.139093668690373, "grad_norm": 0.4928669035434723, "learning_rate": 6.225518642100681e-05, "loss": 0.0124, "step": 10507 }, { "epoch": 1.1392020815264527, "grad_norm": 0.08937004208564758, "learning_rate": 6.225155955316988e-05, "loss": 0.0084, "step": 10508 }, { "epoch": 1.1393104943625325, "grad_norm": 0.36674174666404724, "learning_rate": 6.224793268533295e-05, "loss": 0.0177, "step": 10509 }, { "epoch": 1.1394189071986123, "grad_norm": 0.1273467093706131, "learning_rate": 6.224430581749602e-05, "loss": 0.0143, "step": 10510 }, { "epoch": 1.139527320034692, "grad_norm": 0.2229447066783905, "learning_rate": 6.224067894965908e-05, "loss": 0.0187, "step": 10511 }, { "epoch": 1.1396357328707718, "grad_norm": 0.7404717803001404, "learning_rate": 6.223705208182214e-05, "loss": 0.0531, "step": 10512 }, { "epoch": 1.1397441457068518, "grad_norm": 0.49241918325424194, "learning_rate": 6.223342521398521e-05, "loss": 0.0622, "step": 10513 }, { "epoch": 1.1398525585429315, "grad_norm": 0.1373569369316101, "learning_rate": 6.222979834614827e-05, "loss": 0.0046, "step": 10514 }, { "epoch": 1.1399609713790113, "grad_norm": 0.07203274965286255, "learning_rate": 6.222617147831134e-05, "loss": 0.0028, "step": 10515 }, { "epoch": 1.140069384215091, "grad_norm": 0.39689144492149353, "learning_rate": 6.22225446104744e-05, "loss": 0.0342, "step": 10516 }, { "epoch": 1.1401777970511708, "grad_norm": 0.23922155797481537, "learning_rate": 6.221891774263747e-05, "loss": 0.0166, "step": 10517 }, { "epoch": 1.1402862098872506, "grad_norm": 0.19788220524787903, "learning_rate": 6.221529087480052e-05, "loss": 0.0127, "step": 10518 }, { "epoch": 1.1403946227233304, "grad_norm": 0.5452765226364136, "learning_rate": 6.22116640069636e-05, "loss": 0.0497, "step": 10519 }, { "epoch": 1.1405030355594103, "grad_norm": 0.28968384861946106, "learning_rate": 6.220803713912665e-05, "loss": 0.0174, "step": 10520 }, { "epoch": 1.1406114483954901, "grad_norm": 0.1457424908876419, "learning_rate": 6.220441027128971e-05, "loss": 0.0105, "step": 10521 }, { "epoch": 1.1407198612315699, "grad_norm": 0.25774335861206055, "learning_rate": 6.220078340345278e-05, "loss": 0.0371, "step": 10522 }, { "epoch": 1.1408282740676496, "grad_norm": 0.11198154091835022, "learning_rate": 6.219715653561584e-05, "loss": 0.0168, "step": 10523 }, { "epoch": 1.1409366869037294, "grad_norm": 0.2589913308620453, "learning_rate": 6.219352966777891e-05, "loss": 0.0299, "step": 10524 }, { "epoch": 1.1410450997398092, "grad_norm": 0.08005809783935547, "learning_rate": 6.218990279994197e-05, "loss": 0.0055, "step": 10525 }, { "epoch": 1.141153512575889, "grad_norm": 0.5832198858261108, "learning_rate": 6.218627593210504e-05, "loss": 0.0425, "step": 10526 }, { "epoch": 1.1412619254119687, "grad_norm": 0.15505759418010712, "learning_rate": 6.218264906426811e-05, "loss": 0.0035, "step": 10527 }, { "epoch": 1.1413703382480485, "grad_norm": 0.06336319446563721, "learning_rate": 6.217902219643116e-05, "loss": 0.0047, "step": 10528 }, { "epoch": 1.1414787510841284, "grad_norm": 0.4092816412448883, "learning_rate": 6.217539532859424e-05, "loss": 0.0209, "step": 10529 }, { "epoch": 1.1415871639202082, "grad_norm": 0.10098306834697723, "learning_rate": 6.217176846075729e-05, "loss": 0.0116, "step": 10530 }, { "epoch": 1.141695576756288, "grad_norm": 0.24708318710327148, "learning_rate": 6.216814159292036e-05, "loss": 0.0147, "step": 10531 }, { "epoch": 1.1418039895923677, "grad_norm": 0.2535569965839386, "learning_rate": 6.216451472508342e-05, "loss": 0.0205, "step": 10532 }, { "epoch": 1.1419124024284475, "grad_norm": 0.13562698662281036, "learning_rate": 6.216088785724649e-05, "loss": 0.0093, "step": 10533 }, { "epoch": 1.1420208152645273, "grad_norm": 0.1911344975233078, "learning_rate": 6.215726098940955e-05, "loss": 0.0119, "step": 10534 }, { "epoch": 1.142129228100607, "grad_norm": 0.33246028423309326, "learning_rate": 6.21536341215726e-05, "loss": 0.0248, "step": 10535 }, { "epoch": 1.142237640936687, "grad_norm": 1.0398575067520142, "learning_rate": 6.215000725373568e-05, "loss": 0.0487, "step": 10536 }, { "epoch": 1.1423460537727668, "grad_norm": 0.060496069490909576, "learning_rate": 6.214638038589873e-05, "loss": 0.0033, "step": 10537 }, { "epoch": 1.1424544666088465, "grad_norm": 0.6174341440200806, "learning_rate": 6.21427535180618e-05, "loss": 0.0142, "step": 10538 }, { "epoch": 1.1425628794449263, "grad_norm": 0.2784239947795868, "learning_rate": 6.213912665022486e-05, "loss": 0.0527, "step": 10539 }, { "epoch": 1.142671292281006, "grad_norm": 0.9919801950454712, "learning_rate": 6.213549978238793e-05, "loss": 0.0369, "step": 10540 }, { "epoch": 1.1427797051170858, "grad_norm": 0.1113794669508934, "learning_rate": 6.213187291455099e-05, "loss": 0.0055, "step": 10541 }, { "epoch": 1.1428881179531656, "grad_norm": 0.35927534103393555, "learning_rate": 6.212824604671406e-05, "loss": 0.0166, "step": 10542 }, { "epoch": 1.1429965307892456, "grad_norm": 0.608966588973999, "learning_rate": 6.212461917887712e-05, "loss": 0.0444, "step": 10543 }, { "epoch": 1.1431049436253253, "grad_norm": 0.07509994506835938, "learning_rate": 6.212099231104019e-05, "loss": 0.0085, "step": 10544 }, { "epoch": 1.143213356461405, "grad_norm": 0.31039538979530334, "learning_rate": 6.211736544320326e-05, "loss": 0.0346, "step": 10545 }, { "epoch": 1.1433217692974849, "grad_norm": 0.1278820037841797, "learning_rate": 6.211373857536632e-05, "loss": 0.008, "step": 10546 }, { "epoch": 1.1434301821335646, "grad_norm": 0.2834998667240143, "learning_rate": 6.211011170752939e-05, "loss": 0.0156, "step": 10547 }, { "epoch": 1.1435385949696444, "grad_norm": 0.4325122833251953, "learning_rate": 6.210648483969245e-05, "loss": 0.0412, "step": 10548 }, { "epoch": 1.1436470078057241, "grad_norm": 0.2785751521587372, "learning_rate": 6.21028579718555e-05, "loss": 0.0235, "step": 10549 }, { "epoch": 1.143755420641804, "grad_norm": 0.3829422891139984, "learning_rate": 6.209923110401857e-05, "loss": 0.0216, "step": 10550 }, { "epoch": 1.1438638334778837, "grad_norm": 0.24447666108608246, "learning_rate": 6.209560423618163e-05, "loss": 0.0295, "step": 10551 }, { "epoch": 1.1439722463139637, "grad_norm": 0.30654364824295044, "learning_rate": 6.20919773683447e-05, "loss": 0.0214, "step": 10552 }, { "epoch": 1.1440806591500434, "grad_norm": 0.2144545167684555, "learning_rate": 6.208835050050776e-05, "loss": 0.0187, "step": 10553 }, { "epoch": 1.1441890719861232, "grad_norm": 0.28927335143089294, "learning_rate": 6.208472363267083e-05, "loss": 0.013, "step": 10554 }, { "epoch": 1.144297484822203, "grad_norm": 0.07567432522773743, "learning_rate": 6.208109676483389e-05, "loss": 0.0048, "step": 10555 }, { "epoch": 1.1444058976582827, "grad_norm": 0.05466800555586815, "learning_rate": 6.207746989699696e-05, "loss": 0.0027, "step": 10556 }, { "epoch": 1.1445143104943625, "grad_norm": 0.3332459032535553, "learning_rate": 6.207384302916002e-05, "loss": 0.0291, "step": 10557 }, { "epoch": 1.1446227233304422, "grad_norm": 0.41335922479629517, "learning_rate": 6.207021616132307e-05, "loss": 0.0286, "step": 10558 }, { "epoch": 1.1447311361665222, "grad_norm": 0.3967713415622711, "learning_rate": 6.206658929348614e-05, "loss": 0.0161, "step": 10559 }, { "epoch": 1.144839549002602, "grad_norm": 0.27465176582336426, "learning_rate": 6.20629624256492e-05, "loss": 0.0286, "step": 10560 }, { "epoch": 1.1449479618386817, "grad_norm": 0.07398399710655212, "learning_rate": 6.205933555781229e-05, "loss": 0.0067, "step": 10561 }, { "epoch": 1.1450563746747615, "grad_norm": 0.04781075194478035, "learning_rate": 6.205570868997534e-05, "loss": 0.005, "step": 10562 }, { "epoch": 1.1451647875108413, "grad_norm": 0.5561419725418091, "learning_rate": 6.205208182213841e-05, "loss": 0.0752, "step": 10563 }, { "epoch": 1.145273200346921, "grad_norm": 0.1442573368549347, "learning_rate": 6.204845495430147e-05, "loss": 0.0138, "step": 10564 }, { "epoch": 1.1453816131830008, "grad_norm": 0.2576906979084015, "learning_rate": 6.204482808646453e-05, "loss": 0.0143, "step": 10565 }, { "epoch": 1.1454900260190806, "grad_norm": 0.23421384394168854, "learning_rate": 6.20412012186276e-05, "loss": 0.0178, "step": 10566 }, { "epoch": 1.1455984388551603, "grad_norm": 0.21432702243328094, "learning_rate": 6.203757435079066e-05, "loss": 0.0227, "step": 10567 }, { "epoch": 1.1457068516912403, "grad_norm": 0.27638694643974304, "learning_rate": 6.203394748295373e-05, "loss": 0.0357, "step": 10568 }, { "epoch": 1.14581526452732, "grad_norm": 0.027728695422410965, "learning_rate": 6.203032061511679e-05, "loss": 0.0024, "step": 10569 }, { "epoch": 1.1459236773633998, "grad_norm": 0.2257746458053589, "learning_rate": 6.202669374727986e-05, "loss": 0.0337, "step": 10570 }, { "epoch": 1.1460320901994796, "grad_norm": 0.3617897033691406, "learning_rate": 6.202306687944291e-05, "loss": 0.013, "step": 10571 }, { "epoch": 1.1461405030355594, "grad_norm": 0.22547917068004608, "learning_rate": 6.201944001160597e-05, "loss": 0.0091, "step": 10572 }, { "epoch": 1.1462489158716391, "grad_norm": 0.4157710373401642, "learning_rate": 6.201581314376904e-05, "loss": 0.0322, "step": 10573 }, { "epoch": 1.146357328707719, "grad_norm": 0.4261000454425812, "learning_rate": 6.20121862759321e-05, "loss": 0.0589, "step": 10574 }, { "epoch": 1.1464657415437989, "grad_norm": 0.8272746801376343, "learning_rate": 6.200855940809517e-05, "loss": 0.0233, "step": 10575 }, { "epoch": 1.1465741543798786, "grad_norm": 0.4031878411769867, "learning_rate": 6.200493254025823e-05, "loss": 0.0265, "step": 10576 }, { "epoch": 1.1466825672159584, "grad_norm": 0.2561779022216797, "learning_rate": 6.20013056724213e-05, "loss": 0.0272, "step": 10577 }, { "epoch": 1.1467909800520382, "grad_norm": 0.28383272886276245, "learning_rate": 6.199767880458437e-05, "loss": 0.0281, "step": 10578 }, { "epoch": 1.146899392888118, "grad_norm": 0.1271139681339264, "learning_rate": 6.199405193674743e-05, "loss": 0.0153, "step": 10579 }, { "epoch": 1.1470078057241977, "grad_norm": 0.26647061109542847, "learning_rate": 6.19904250689105e-05, "loss": 0.0154, "step": 10580 }, { "epoch": 1.1471162185602775, "grad_norm": 0.31399333477020264, "learning_rate": 6.198679820107356e-05, "loss": 0.0196, "step": 10581 }, { "epoch": 1.1472246313963574, "grad_norm": 0.12179718166589737, "learning_rate": 6.198317133323663e-05, "loss": 0.0123, "step": 10582 }, { "epoch": 1.1473330442324372, "grad_norm": 0.4475231468677521, "learning_rate": 6.197954446539968e-05, "loss": 0.0506, "step": 10583 }, { "epoch": 1.147441457068517, "grad_norm": 0.5653029680252075, "learning_rate": 6.197591759756275e-05, "loss": 0.0257, "step": 10584 }, { "epoch": 1.1475498699045967, "grad_norm": 0.1263294816017151, "learning_rate": 6.197229072972581e-05, "loss": 0.0069, "step": 10585 }, { "epoch": 1.1476582827406765, "grad_norm": 0.39753326773643494, "learning_rate": 6.196866386188888e-05, "loss": 0.0163, "step": 10586 }, { "epoch": 1.1477666955767563, "grad_norm": 0.028987834230065346, "learning_rate": 6.196503699405194e-05, "loss": 0.0016, "step": 10587 }, { "epoch": 1.147875108412836, "grad_norm": 0.16952337324619293, "learning_rate": 6.1961410126215e-05, "loss": 0.013, "step": 10588 }, { "epoch": 1.1479835212489158, "grad_norm": 0.21892371773719788, "learning_rate": 6.195778325837807e-05, "loss": 0.0041, "step": 10589 }, { "epoch": 1.1480919340849955, "grad_norm": 0.3564024865627289, "learning_rate": 6.195415639054113e-05, "loss": 0.0191, "step": 10590 }, { "epoch": 1.1482003469210755, "grad_norm": 0.5163295865058899, "learning_rate": 6.19505295227042e-05, "loss": 0.0115, "step": 10591 }, { "epoch": 1.1483087597571553, "grad_norm": 0.20638646185398102, "learning_rate": 6.194690265486725e-05, "loss": 0.0167, "step": 10592 }, { "epoch": 1.148417172593235, "grad_norm": 0.3043556809425354, "learning_rate": 6.194327578703032e-05, "loss": 0.0188, "step": 10593 }, { "epoch": 1.1485255854293148, "grad_norm": 0.28834858536720276, "learning_rate": 6.193964891919338e-05, "loss": 0.008, "step": 10594 }, { "epoch": 1.1486339982653946, "grad_norm": 0.25642871856689453, "learning_rate": 6.193602205135645e-05, "loss": 0.0313, "step": 10595 }, { "epoch": 1.1487424111014743, "grad_norm": 0.31412652134895325, "learning_rate": 6.193239518351952e-05, "loss": 0.0179, "step": 10596 }, { "epoch": 1.148850823937554, "grad_norm": 0.08292818814516068, "learning_rate": 6.192876831568258e-05, "loss": 0.0109, "step": 10597 }, { "epoch": 1.148959236773634, "grad_norm": 0.5878402590751648, "learning_rate": 6.192514144784565e-05, "loss": 0.0875, "step": 10598 }, { "epoch": 1.1490676496097139, "grad_norm": 0.1441003680229187, "learning_rate": 6.192151458000871e-05, "loss": 0.0081, "step": 10599 }, { "epoch": 1.1491760624457936, "grad_norm": 0.18146909773349762, "learning_rate": 6.191788771217178e-05, "loss": 0.008, "step": 10600 }, { "epoch": 1.1492844752818734, "grad_norm": 0.4366976320743561, "learning_rate": 6.191426084433484e-05, "loss": 0.0335, "step": 10601 }, { "epoch": 1.1493928881179531, "grad_norm": 0.22549471259117126, "learning_rate": 6.19106339764979e-05, "loss": 0.0247, "step": 10602 }, { "epoch": 1.149501300954033, "grad_norm": 1.0698552131652832, "learning_rate": 6.190700710866097e-05, "loss": 0.0323, "step": 10603 }, { "epoch": 1.1496097137901127, "grad_norm": 0.6810693740844727, "learning_rate": 6.190338024082402e-05, "loss": 0.0307, "step": 10604 }, { "epoch": 1.1497181266261927, "grad_norm": 0.904473602771759, "learning_rate": 6.18997533729871e-05, "loss": 0.0326, "step": 10605 }, { "epoch": 1.1498265394622724, "grad_norm": 0.43417254090309143, "learning_rate": 6.189612650515015e-05, "loss": 0.0535, "step": 10606 }, { "epoch": 1.1499349522983522, "grad_norm": 0.13265535235404968, "learning_rate": 6.189249963731322e-05, "loss": 0.0104, "step": 10607 }, { "epoch": 1.150043365134432, "grad_norm": 0.07696989923715591, "learning_rate": 6.188887276947628e-05, "loss": 0.0059, "step": 10608 }, { "epoch": 1.1501517779705117, "grad_norm": 0.2800622284412384, "learning_rate": 6.188524590163935e-05, "loss": 0.0214, "step": 10609 }, { "epoch": 1.1502601908065915, "grad_norm": 0.2517385482788086, "learning_rate": 6.188161903380241e-05, "loss": 0.0362, "step": 10610 }, { "epoch": 1.1503686036426712, "grad_norm": 0.08165688067674637, "learning_rate": 6.187799216596546e-05, "loss": 0.0055, "step": 10611 }, { "epoch": 1.150477016478751, "grad_norm": 0.7345150113105774, "learning_rate": 6.187436529812854e-05, "loss": 0.0657, "step": 10612 }, { "epoch": 1.1505854293148308, "grad_norm": 0.4006965160369873, "learning_rate": 6.18707384302916e-05, "loss": 0.0363, "step": 10613 }, { "epoch": 1.1506938421509108, "grad_norm": 0.19799630343914032, "learning_rate": 6.186711156245468e-05, "loss": 0.0108, "step": 10614 }, { "epoch": 1.1508022549869905, "grad_norm": 0.20916029810905457, "learning_rate": 6.186348469461773e-05, "loss": 0.0172, "step": 10615 }, { "epoch": 1.1509106678230703, "grad_norm": 0.30792000889778137, "learning_rate": 6.185985782678079e-05, "loss": 0.0206, "step": 10616 }, { "epoch": 1.15101908065915, "grad_norm": 0.13585607707500458, "learning_rate": 6.185623095894386e-05, "loss": 0.0155, "step": 10617 }, { "epoch": 1.1511274934952298, "grad_norm": 0.12020550668239594, "learning_rate": 6.185260409110692e-05, "loss": 0.0064, "step": 10618 }, { "epoch": 1.1512359063313096, "grad_norm": 0.3209567666053772, "learning_rate": 6.184897722326999e-05, "loss": 0.0738, "step": 10619 }, { "epoch": 1.1513443191673893, "grad_norm": 0.5651852488517761, "learning_rate": 6.184535035543305e-05, "loss": 0.055, "step": 10620 }, { "epoch": 1.1514527320034693, "grad_norm": 0.38067376613616943, "learning_rate": 6.184172348759612e-05, "loss": 0.0176, "step": 10621 }, { "epoch": 1.151561144839549, "grad_norm": 0.3364606499671936, "learning_rate": 6.183809661975918e-05, "loss": 0.0201, "step": 10622 }, { "epoch": 1.1516695576756288, "grad_norm": 0.5190739631652832, "learning_rate": 6.183446975192225e-05, "loss": 0.0304, "step": 10623 }, { "epoch": 1.1517779705117086, "grad_norm": 0.24120451509952545, "learning_rate": 6.18308428840853e-05, "loss": 0.0141, "step": 10624 }, { "epoch": 1.1518863833477884, "grad_norm": 0.2711663842201233, "learning_rate": 6.182721601624836e-05, "loss": 0.0513, "step": 10625 }, { "epoch": 1.1519947961838681, "grad_norm": 0.23065641522407532, "learning_rate": 6.182358914841143e-05, "loss": 0.0141, "step": 10626 }, { "epoch": 1.152103209019948, "grad_norm": 0.2616785764694214, "learning_rate": 6.181996228057449e-05, "loss": 0.0246, "step": 10627 }, { "epoch": 1.1522116218560277, "grad_norm": 0.2699531614780426, "learning_rate": 6.181633541273756e-05, "loss": 0.0221, "step": 10628 }, { "epoch": 1.1523200346921076, "grad_norm": 0.4803770184516907, "learning_rate": 6.181270854490062e-05, "loss": 0.0329, "step": 10629 }, { "epoch": 1.1524284475281874, "grad_norm": 0.09432744979858398, "learning_rate": 6.180908167706369e-05, "loss": 0.0062, "step": 10630 }, { "epoch": 1.1525368603642672, "grad_norm": 0.308733731508255, "learning_rate": 6.180545480922676e-05, "loss": 0.0424, "step": 10631 }, { "epoch": 1.152645273200347, "grad_norm": 0.1512233167886734, "learning_rate": 6.180182794138982e-05, "loss": 0.0181, "step": 10632 }, { "epoch": 1.1527536860364267, "grad_norm": 0.8011527061462402, "learning_rate": 6.179820107355289e-05, "loss": 0.02, "step": 10633 }, { "epoch": 1.1528620988725065, "grad_norm": 0.09917322546243668, "learning_rate": 6.179457420571595e-05, "loss": 0.0041, "step": 10634 }, { "epoch": 1.1529705117085862, "grad_norm": 0.1969919055700302, "learning_rate": 6.179094733787902e-05, "loss": 0.0126, "step": 10635 }, { "epoch": 1.153078924544666, "grad_norm": 0.8533035516738892, "learning_rate": 6.178732047004207e-05, "loss": 0.0334, "step": 10636 }, { "epoch": 1.153187337380746, "grad_norm": 0.18825702369213104, "learning_rate": 6.178369360220515e-05, "loss": 0.0101, "step": 10637 }, { "epoch": 1.1532957502168257, "grad_norm": 0.037048570811748505, "learning_rate": 6.17800667343682e-05, "loss": 0.0028, "step": 10638 }, { "epoch": 1.1534041630529055, "grad_norm": 0.2969827651977539, "learning_rate": 6.177643986653126e-05, "loss": 0.0195, "step": 10639 }, { "epoch": 1.1535125758889853, "grad_norm": 0.2869604527950287, "learning_rate": 6.177281299869433e-05, "loss": 0.0052, "step": 10640 }, { "epoch": 1.153620988725065, "grad_norm": 0.06209433451294899, "learning_rate": 6.176918613085739e-05, "loss": 0.0032, "step": 10641 }, { "epoch": 1.1537294015611448, "grad_norm": 0.3326685130596161, "learning_rate": 6.176555926302046e-05, "loss": 0.0103, "step": 10642 }, { "epoch": 1.1538378143972245, "grad_norm": 0.040402233600616455, "learning_rate": 6.176193239518352e-05, "loss": 0.0013, "step": 10643 }, { "epoch": 1.1539462272333045, "grad_norm": 0.10072314739227295, "learning_rate": 6.175830552734659e-05, "loss": 0.0037, "step": 10644 }, { "epoch": 1.1540546400693843, "grad_norm": 0.6421395540237427, "learning_rate": 6.175467865950964e-05, "loss": 0.0193, "step": 10645 }, { "epoch": 1.154163052905464, "grad_norm": 0.41573038697242737, "learning_rate": 6.175105179167272e-05, "loss": 0.0598, "step": 10646 }, { "epoch": 1.1542714657415438, "grad_norm": 0.49533483386039734, "learning_rate": 6.174742492383579e-05, "loss": 0.0539, "step": 10647 }, { "epoch": 1.1543798785776236, "grad_norm": 0.7932404279708862, "learning_rate": 6.174379805599884e-05, "loss": 0.0595, "step": 10648 }, { "epoch": 1.1544882914137033, "grad_norm": 0.8161742091178894, "learning_rate": 6.174017118816191e-05, "loss": 0.0561, "step": 10649 }, { "epoch": 1.1545967042497831, "grad_norm": 0.2996301054954529, "learning_rate": 6.173654432032497e-05, "loss": 0.0133, "step": 10650 }, { "epoch": 1.1547051170858629, "grad_norm": 0.42958182096481323, "learning_rate": 6.173291745248804e-05, "loss": 0.0608, "step": 10651 }, { "epoch": 1.1548135299219426, "grad_norm": 0.522209882736206, "learning_rate": 6.17292905846511e-05, "loss": 0.063, "step": 10652 }, { "epoch": 1.1549219427580226, "grad_norm": 0.6926705837249756, "learning_rate": 6.172566371681416e-05, "loss": 0.0348, "step": 10653 }, { "epoch": 1.1550303555941024, "grad_norm": 0.2746797204017639, "learning_rate": 6.172203684897723e-05, "loss": 0.02, "step": 10654 }, { "epoch": 1.1551387684301822, "grad_norm": 0.06893208622932434, "learning_rate": 6.171840998114029e-05, "loss": 0.0076, "step": 10655 }, { "epoch": 1.155247181266262, "grad_norm": 0.30010083317756653, "learning_rate": 6.171478311330336e-05, "loss": 0.023, "step": 10656 }, { "epoch": 1.1553555941023417, "grad_norm": 0.20649632811546326, "learning_rate": 6.171115624546641e-05, "loss": 0.0078, "step": 10657 }, { "epoch": 1.1554640069384214, "grad_norm": 0.2212647795677185, "learning_rate": 6.170752937762948e-05, "loss": 0.0214, "step": 10658 }, { "epoch": 1.1555724197745012, "grad_norm": 0.36872634291648865, "learning_rate": 6.170390250979254e-05, "loss": 0.0556, "step": 10659 }, { "epoch": 1.1556808326105812, "grad_norm": 0.5915942192077637, "learning_rate": 6.170027564195561e-05, "loss": 0.0726, "step": 10660 }, { "epoch": 1.155789245446661, "grad_norm": 0.3701423406600952, "learning_rate": 6.169664877411867e-05, "loss": 0.1028, "step": 10661 }, { "epoch": 1.1558976582827407, "grad_norm": 0.3278524875640869, "learning_rate": 6.169302190628173e-05, "loss": 0.0085, "step": 10662 }, { "epoch": 1.1560060711188205, "grad_norm": 0.26779812574386597, "learning_rate": 6.16893950384448e-05, "loss": 0.0269, "step": 10663 }, { "epoch": 1.1561144839549002, "grad_norm": 0.22555099427700043, "learning_rate": 6.168576817060786e-05, "loss": 0.0101, "step": 10664 }, { "epoch": 1.15622289679098, "grad_norm": 0.49967655539512634, "learning_rate": 6.168214130277094e-05, "loss": 0.0792, "step": 10665 }, { "epoch": 1.1563313096270598, "grad_norm": 0.8603308200836182, "learning_rate": 6.1678514434934e-05, "loss": 0.0367, "step": 10666 }, { "epoch": 1.1564397224631398, "grad_norm": 0.45682018995285034, "learning_rate": 6.167488756709707e-05, "loss": 0.0453, "step": 10667 }, { "epoch": 1.1565481352992195, "grad_norm": 0.40554821491241455, "learning_rate": 6.167126069926013e-05, "loss": 0.0206, "step": 10668 }, { "epoch": 1.1566565481352993, "grad_norm": 0.5304154753684998, "learning_rate": 6.166763383142318e-05, "loss": 0.0491, "step": 10669 }, { "epoch": 1.156764960971379, "grad_norm": 0.24042604863643646, "learning_rate": 6.166400696358625e-05, "loss": 0.0115, "step": 10670 }, { "epoch": 1.1568733738074588, "grad_norm": 0.24732597172260284, "learning_rate": 6.166038009574931e-05, "loss": 0.0243, "step": 10671 }, { "epoch": 1.1569817866435386, "grad_norm": 0.2605282962322235, "learning_rate": 6.165675322791238e-05, "loss": 0.0309, "step": 10672 }, { "epoch": 1.1570901994796183, "grad_norm": 0.04615031182765961, "learning_rate": 6.165312636007544e-05, "loss": 0.003, "step": 10673 }, { "epoch": 1.157198612315698, "grad_norm": 0.17662066221237183, "learning_rate": 6.164949949223851e-05, "loss": 0.0287, "step": 10674 }, { "epoch": 1.1573070251517779, "grad_norm": 0.3346020579338074, "learning_rate": 6.164587262440157e-05, "loss": 0.0376, "step": 10675 }, { "epoch": 1.1574154379878578, "grad_norm": 0.4239966869354248, "learning_rate": 6.164224575656462e-05, "loss": 0.0312, "step": 10676 }, { "epoch": 1.1575238508239376, "grad_norm": 0.42601245641708374, "learning_rate": 6.16386188887277e-05, "loss": 0.0358, "step": 10677 }, { "epoch": 1.1576322636600174, "grad_norm": 0.18372191488742828, "learning_rate": 6.163499202089075e-05, "loss": 0.0135, "step": 10678 }, { "epoch": 1.1577406764960971, "grad_norm": 0.2152203470468521, "learning_rate": 6.163136515305382e-05, "loss": 0.0393, "step": 10679 }, { "epoch": 1.157849089332177, "grad_norm": 0.3556971251964569, "learning_rate": 6.162773828521688e-05, "loss": 0.0448, "step": 10680 }, { "epoch": 1.1579575021682567, "grad_norm": 0.08020252734422684, "learning_rate": 6.162411141737995e-05, "loss": 0.0042, "step": 10681 }, { "epoch": 1.1580659150043364, "grad_norm": 0.4539472460746765, "learning_rate": 6.162048454954302e-05, "loss": 0.0272, "step": 10682 }, { "epoch": 1.1581743278404164, "grad_norm": 0.3568081855773926, "learning_rate": 6.161685768170608e-05, "loss": 0.028, "step": 10683 }, { "epoch": 1.1582827406764962, "grad_norm": 0.28330010175704956, "learning_rate": 6.161323081386915e-05, "loss": 0.0269, "step": 10684 }, { "epoch": 1.158391153512576, "grad_norm": 0.35594236850738525, "learning_rate": 6.160960394603221e-05, "loss": 0.0367, "step": 10685 }, { "epoch": 1.1584995663486557, "grad_norm": 0.3549586236476898, "learning_rate": 6.160597707819528e-05, "loss": 0.0473, "step": 10686 }, { "epoch": 1.1586079791847355, "grad_norm": 0.1407424509525299, "learning_rate": 6.160235021035834e-05, "loss": 0.0068, "step": 10687 }, { "epoch": 1.1587163920208152, "grad_norm": 0.24129624664783478, "learning_rate": 6.159872334252141e-05, "loss": 0.0079, "step": 10688 }, { "epoch": 1.158824804856895, "grad_norm": 0.43492591381073, "learning_rate": 6.159509647468447e-05, "loss": 0.0727, "step": 10689 }, { "epoch": 1.158933217692975, "grad_norm": 0.241195946931839, "learning_rate": 6.159146960684754e-05, "loss": 0.0156, "step": 10690 }, { "epoch": 1.1590416305290547, "grad_norm": 0.29735082387924194, "learning_rate": 6.15878427390106e-05, "loss": 0.0248, "step": 10691 }, { "epoch": 1.1591500433651345, "grad_norm": 0.05699089542031288, "learning_rate": 6.158421587117365e-05, "loss": 0.0052, "step": 10692 }, { "epoch": 1.1592584562012143, "grad_norm": 0.17872467637062073, "learning_rate": 6.158058900333672e-05, "loss": 0.0243, "step": 10693 }, { "epoch": 1.159366869037294, "grad_norm": 0.12993837893009186, "learning_rate": 6.157696213549978e-05, "loss": 0.0133, "step": 10694 }, { "epoch": 1.1594752818733738, "grad_norm": 0.1115340143442154, "learning_rate": 6.157333526766285e-05, "loss": 0.0086, "step": 10695 }, { "epoch": 1.1595836947094535, "grad_norm": 0.37937256693840027, "learning_rate": 6.156970839982591e-05, "loss": 0.0319, "step": 10696 }, { "epoch": 1.1596921075455333, "grad_norm": 0.08568882942199707, "learning_rate": 6.156608153198898e-05, "loss": 0.0076, "step": 10697 }, { "epoch": 1.159800520381613, "grad_norm": 0.02674962766468525, "learning_rate": 6.156245466415204e-05, "loss": 0.002, "step": 10698 }, { "epoch": 1.159908933217693, "grad_norm": 0.7433433532714844, "learning_rate": 6.15588277963151e-05, "loss": 0.0174, "step": 10699 }, { "epoch": 1.1600173460537728, "grad_norm": 0.4099312722682953, "learning_rate": 6.155520092847818e-05, "loss": 0.0344, "step": 10700 }, { "epoch": 1.1601257588898526, "grad_norm": 0.25630325078964233, "learning_rate": 6.155157406064123e-05, "loss": 0.0136, "step": 10701 }, { "epoch": 1.1602341717259324, "grad_norm": 0.11562016606330872, "learning_rate": 6.15479471928043e-05, "loss": 0.0159, "step": 10702 }, { "epoch": 1.1603425845620121, "grad_norm": 0.24056194722652435, "learning_rate": 6.154432032496736e-05, "loss": 0.0318, "step": 10703 }, { "epoch": 1.1604509973980919, "grad_norm": 0.044226501137018204, "learning_rate": 6.154069345713043e-05, "loss": 0.0025, "step": 10704 }, { "epoch": 1.1605594102341716, "grad_norm": 0.504046618938446, "learning_rate": 6.153706658929349e-05, "loss": 0.0529, "step": 10705 }, { "epoch": 1.1606678230702516, "grad_norm": 0.44214531779289246, "learning_rate": 6.153343972145655e-05, "loss": 0.0133, "step": 10706 }, { "epoch": 1.1607762359063314, "grad_norm": 0.5437672138214111, "learning_rate": 6.152981285361962e-05, "loss": 0.0698, "step": 10707 }, { "epoch": 1.1608846487424112, "grad_norm": 0.12035117298364639, "learning_rate": 6.152618598578268e-05, "loss": 0.007, "step": 10708 }, { "epoch": 1.160993061578491, "grad_norm": 0.8911815285682678, "learning_rate": 6.152255911794575e-05, "loss": 0.0416, "step": 10709 }, { "epoch": 1.1611014744145707, "grad_norm": 0.1681523621082306, "learning_rate": 6.15189322501088e-05, "loss": 0.0307, "step": 10710 }, { "epoch": 1.1612098872506504, "grad_norm": 0.27199870347976685, "learning_rate": 6.151530538227188e-05, "loss": 0.0141, "step": 10711 }, { "epoch": 1.1613183000867302, "grad_norm": 0.3107839822769165, "learning_rate": 6.151167851443493e-05, "loss": 0.034, "step": 10712 }, { "epoch": 1.16142671292281, "grad_norm": 0.17318914830684662, "learning_rate": 6.1508051646598e-05, "loss": 0.0111, "step": 10713 }, { "epoch": 1.1615351257588897, "grad_norm": 0.238389790058136, "learning_rate": 6.150442477876106e-05, "loss": 0.0163, "step": 10714 }, { "epoch": 1.1616435385949697, "grad_norm": 0.2598877251148224, "learning_rate": 6.150079791092412e-05, "loss": 0.0187, "step": 10715 }, { "epoch": 1.1617519514310495, "grad_norm": 0.3005159795284271, "learning_rate": 6.14971710430872e-05, "loss": 0.0219, "step": 10716 }, { "epoch": 1.1618603642671292, "grad_norm": 0.17676258087158203, "learning_rate": 6.149354417525026e-05, "loss": 0.0236, "step": 10717 }, { "epoch": 1.161968777103209, "grad_norm": 0.13221636414527893, "learning_rate": 6.148991730741333e-05, "loss": 0.0133, "step": 10718 }, { "epoch": 1.1620771899392888, "grad_norm": 0.108761265873909, "learning_rate": 6.148629043957639e-05, "loss": 0.006, "step": 10719 }, { "epoch": 1.1621856027753685, "grad_norm": 0.19556482136249542, "learning_rate": 6.148266357173945e-05, "loss": 0.0341, "step": 10720 }, { "epoch": 1.1622940156114483, "grad_norm": 0.39937353134155273, "learning_rate": 6.147903670390252e-05, "loss": 0.0299, "step": 10721 }, { "epoch": 1.1624024284475283, "grad_norm": 0.32136082649230957, "learning_rate": 6.147540983606557e-05, "loss": 0.0346, "step": 10722 }, { "epoch": 1.162510841283608, "grad_norm": 0.11248846352100372, "learning_rate": 6.147178296822864e-05, "loss": 0.0058, "step": 10723 }, { "epoch": 1.1626192541196878, "grad_norm": 0.1805969625711441, "learning_rate": 6.14681561003917e-05, "loss": 0.0376, "step": 10724 }, { "epoch": 1.1627276669557676, "grad_norm": 0.24655941128730774, "learning_rate": 6.146452923255477e-05, "loss": 0.0113, "step": 10725 }, { "epoch": 1.1628360797918473, "grad_norm": 0.18985183537006378, "learning_rate": 6.146090236471783e-05, "loss": 0.013, "step": 10726 }, { "epoch": 1.162944492627927, "grad_norm": 0.38088199496269226, "learning_rate": 6.14572754968809e-05, "loss": 0.023, "step": 10727 }, { "epoch": 1.1630529054640069, "grad_norm": 0.18479932844638824, "learning_rate": 6.145364862904396e-05, "loss": 0.0106, "step": 10728 }, { "epoch": 1.1631613183000868, "grad_norm": 0.07074850797653198, "learning_rate": 6.145002176120702e-05, "loss": 0.004, "step": 10729 }, { "epoch": 1.1632697311361666, "grad_norm": 0.32908880710601807, "learning_rate": 6.144639489337009e-05, "loss": 0.0357, "step": 10730 }, { "epoch": 1.1633781439722464, "grad_norm": 0.2785753905773163, "learning_rate": 6.144276802553314e-05, "loss": 0.027, "step": 10731 }, { "epoch": 1.1634865568083261, "grad_norm": 0.5235678553581238, "learning_rate": 6.143914115769621e-05, "loss": 0.0653, "step": 10732 }, { "epoch": 1.163594969644406, "grad_norm": 0.3058391809463501, "learning_rate": 6.143551428985927e-05, "loss": 0.0176, "step": 10733 }, { "epoch": 1.1637033824804857, "grad_norm": 0.9633142948150635, "learning_rate": 6.143188742202234e-05, "loss": 0.0556, "step": 10734 }, { "epoch": 1.1638117953165654, "grad_norm": 0.03809322416782379, "learning_rate": 6.142826055418541e-05, "loss": 0.0025, "step": 10735 }, { "epoch": 1.1639202081526452, "grad_norm": 0.30746179819107056, "learning_rate": 6.142463368634847e-05, "loss": 0.0217, "step": 10736 }, { "epoch": 1.164028620988725, "grad_norm": 1.284440040588379, "learning_rate": 6.142100681851154e-05, "loss": 0.0417, "step": 10737 }, { "epoch": 1.164137033824805, "grad_norm": 0.09045501798391342, "learning_rate": 6.14173799506746e-05, "loss": 0.0081, "step": 10738 }, { "epoch": 1.1642454466608847, "grad_norm": 0.24216562509536743, "learning_rate": 6.141375308283767e-05, "loss": 0.0347, "step": 10739 }, { "epoch": 1.1643538594969645, "grad_norm": 0.6398198008537292, "learning_rate": 6.141012621500073e-05, "loss": 0.0341, "step": 10740 }, { "epoch": 1.1644622723330442, "grad_norm": 0.19795802235603333, "learning_rate": 6.14064993471638e-05, "loss": 0.0132, "step": 10741 }, { "epoch": 1.164570685169124, "grad_norm": 0.4077507555484772, "learning_rate": 6.140287247932686e-05, "loss": 0.0351, "step": 10742 }, { "epoch": 1.1646790980052038, "grad_norm": 0.47029006481170654, "learning_rate": 6.139924561148991e-05, "loss": 0.0223, "step": 10743 }, { "epoch": 1.1647875108412835, "grad_norm": 0.5282728672027588, "learning_rate": 6.139561874365298e-05, "loss": 0.0397, "step": 10744 }, { "epoch": 1.1648959236773635, "grad_norm": 0.1524191051721573, "learning_rate": 6.139199187581604e-05, "loss": 0.0206, "step": 10745 }, { "epoch": 1.1650043365134433, "grad_norm": 0.2071402221918106, "learning_rate": 6.138836500797911e-05, "loss": 0.0184, "step": 10746 }, { "epoch": 1.165112749349523, "grad_norm": 0.14343348145484924, "learning_rate": 6.138473814014217e-05, "loss": 0.0123, "step": 10747 }, { "epoch": 1.1652211621856028, "grad_norm": 0.1939758062362671, "learning_rate": 6.138111127230524e-05, "loss": 0.0208, "step": 10748 }, { "epoch": 1.1653295750216826, "grad_norm": 0.2200363427400589, "learning_rate": 6.13774844044683e-05, "loss": 0.0401, "step": 10749 }, { "epoch": 1.1654379878577623, "grad_norm": 0.08159581571817398, "learning_rate": 6.137385753663137e-05, "loss": 0.0048, "step": 10750 }, { "epoch": 1.165546400693842, "grad_norm": 0.15216876566410065, "learning_rate": 6.137023066879444e-05, "loss": 0.0141, "step": 10751 }, { "epoch": 1.165654813529922, "grad_norm": 0.6682757139205933, "learning_rate": 6.13666038009575e-05, "loss": 0.0599, "step": 10752 }, { "epoch": 1.1657632263660018, "grad_norm": 0.5509182214736938, "learning_rate": 6.136297693312057e-05, "loss": 0.038, "step": 10753 }, { "epoch": 1.1658716392020816, "grad_norm": 0.46811166405677795, "learning_rate": 6.135935006528363e-05, "loss": 0.0476, "step": 10754 }, { "epoch": 1.1659800520381614, "grad_norm": 0.02070567011833191, "learning_rate": 6.13557231974467e-05, "loss": 0.0016, "step": 10755 }, { "epoch": 1.1660884648742411, "grad_norm": 0.11650972068309784, "learning_rate": 6.135209632960975e-05, "loss": 0.01, "step": 10756 }, { "epoch": 1.1661968777103209, "grad_norm": 0.38638055324554443, "learning_rate": 6.134846946177281e-05, "loss": 0.0249, "step": 10757 }, { "epoch": 1.1663052905464006, "grad_norm": 0.2705429196357727, "learning_rate": 6.134484259393588e-05, "loss": 0.0264, "step": 10758 }, { "epoch": 1.1664137033824804, "grad_norm": 0.4223899841308594, "learning_rate": 6.134121572609894e-05, "loss": 0.0165, "step": 10759 }, { "epoch": 1.1665221162185602, "grad_norm": 0.11627906560897827, "learning_rate": 6.133758885826201e-05, "loss": 0.0085, "step": 10760 }, { "epoch": 1.1666305290546402, "grad_norm": 0.39256587624549866, "learning_rate": 6.133396199042507e-05, "loss": 0.0154, "step": 10761 }, { "epoch": 1.16673894189072, "grad_norm": 0.2391175925731659, "learning_rate": 6.133033512258814e-05, "loss": 0.0152, "step": 10762 }, { "epoch": 1.1668473547267997, "grad_norm": 0.46596258878707886, "learning_rate": 6.13267082547512e-05, "loss": 0.0175, "step": 10763 }, { "epoch": 1.1669557675628794, "grad_norm": 0.21528106927871704, "learning_rate": 6.132308138691427e-05, "loss": 0.0181, "step": 10764 }, { "epoch": 1.1670641803989592, "grad_norm": 0.4874486029148102, "learning_rate": 6.131945451907732e-05, "loss": 0.0543, "step": 10765 }, { "epoch": 1.167172593235039, "grad_norm": 0.23867738246917725, "learning_rate": 6.131582765124038e-05, "loss": 0.0184, "step": 10766 }, { "epoch": 1.1672810060711187, "grad_norm": 0.16371150314807892, "learning_rate": 6.131220078340345e-05, "loss": 0.0136, "step": 10767 }, { "epoch": 1.1673894189071987, "grad_norm": 0.4697208106517792, "learning_rate": 6.130857391556652e-05, "loss": 0.0256, "step": 10768 }, { "epoch": 1.1674978317432785, "grad_norm": 0.35748055577278137, "learning_rate": 6.13049470477296e-05, "loss": 0.0353, "step": 10769 }, { "epoch": 1.1676062445793582, "grad_norm": 0.3982715308666229, "learning_rate": 6.130132017989265e-05, "loss": 0.0253, "step": 10770 }, { "epoch": 1.167714657415438, "grad_norm": 0.34596318006515503, "learning_rate": 6.129769331205572e-05, "loss": 0.0107, "step": 10771 }, { "epoch": 1.1678230702515178, "grad_norm": 0.4367569386959076, "learning_rate": 6.129406644421878e-05, "loss": 0.0151, "step": 10772 }, { "epoch": 1.1679314830875975, "grad_norm": 0.351195365190506, "learning_rate": 6.129043957638184e-05, "loss": 0.0441, "step": 10773 }, { "epoch": 1.1680398959236773, "grad_norm": 1.1736783981323242, "learning_rate": 6.128681270854491e-05, "loss": 0.0543, "step": 10774 }, { "epoch": 1.168148308759757, "grad_norm": 0.12382756173610687, "learning_rate": 6.128318584070796e-05, "loss": 0.0052, "step": 10775 }, { "epoch": 1.1682567215958368, "grad_norm": 0.33217138051986694, "learning_rate": 6.127955897287104e-05, "loss": 0.0399, "step": 10776 }, { "epoch": 1.1683651344319168, "grad_norm": 0.5295724272727966, "learning_rate": 6.127593210503409e-05, "loss": 0.034, "step": 10777 }, { "epoch": 1.1684735472679966, "grad_norm": 0.16767793893814087, "learning_rate": 6.127230523719716e-05, "loss": 0.0095, "step": 10778 }, { "epoch": 1.1685819601040763, "grad_norm": 0.2628720998764038, "learning_rate": 6.126867836936022e-05, "loss": 0.0096, "step": 10779 }, { "epoch": 1.168690372940156, "grad_norm": 0.1190808042883873, "learning_rate": 6.126505150152329e-05, "loss": 0.0112, "step": 10780 }, { "epoch": 1.1687987857762359, "grad_norm": 0.29830169677734375, "learning_rate": 6.126142463368635e-05, "loss": 0.0079, "step": 10781 }, { "epoch": 1.1689071986123156, "grad_norm": 0.2524240016937256, "learning_rate": 6.12577977658494e-05, "loss": 0.0124, "step": 10782 }, { "epoch": 1.1690156114483954, "grad_norm": 0.32136911153793335, "learning_rate": 6.125417089801248e-05, "loss": 0.0553, "step": 10783 }, { "epoch": 1.1691240242844754, "grad_norm": 0.27492496371269226, "learning_rate": 6.125054403017553e-05, "loss": 0.0292, "step": 10784 }, { "epoch": 1.1692324371205551, "grad_norm": 0.48957884311676025, "learning_rate": 6.12469171623386e-05, "loss": 0.0362, "step": 10785 }, { "epoch": 1.169340849956635, "grad_norm": 0.34346508979797363, "learning_rate": 6.124329029450168e-05, "loss": 0.026, "step": 10786 }, { "epoch": 1.1694492627927147, "grad_norm": 0.3334965705871582, "learning_rate": 6.123966342666473e-05, "loss": 0.0165, "step": 10787 }, { "epoch": 1.1695576756287944, "grad_norm": 0.6242659687995911, "learning_rate": 6.12360365588278e-05, "loss": 0.0189, "step": 10788 }, { "epoch": 1.1696660884648742, "grad_norm": 0.06097845733165741, "learning_rate": 6.123240969099086e-05, "loss": 0.0029, "step": 10789 }, { "epoch": 1.169774501300954, "grad_norm": 0.2759535312652588, "learning_rate": 6.122878282315393e-05, "loss": 0.0471, "step": 10790 }, { "epoch": 1.169882914137034, "grad_norm": 0.38062751293182373, "learning_rate": 6.122515595531699e-05, "loss": 0.0487, "step": 10791 }, { "epoch": 1.1699913269731137, "grad_norm": 0.3134899139404297, "learning_rate": 6.122152908748006e-05, "loss": 0.0203, "step": 10792 }, { "epoch": 1.1700997398091935, "grad_norm": 0.3842780292034149, "learning_rate": 6.121790221964312e-05, "loss": 0.0223, "step": 10793 }, { "epoch": 1.1702081526452732, "grad_norm": 0.28790202736854553, "learning_rate": 6.121427535180619e-05, "loss": 0.0092, "step": 10794 }, { "epoch": 1.170316565481353, "grad_norm": 0.143283873796463, "learning_rate": 6.121064848396925e-05, "loss": 0.0113, "step": 10795 }, { "epoch": 1.1704249783174328, "grad_norm": 0.3664498031139374, "learning_rate": 6.12070216161323e-05, "loss": 0.0402, "step": 10796 }, { "epoch": 1.1705333911535125, "grad_norm": 0.728348970413208, "learning_rate": 6.120339474829537e-05, "loss": 0.0637, "step": 10797 }, { "epoch": 1.1706418039895923, "grad_norm": 0.3254391551017761, "learning_rate": 6.119976788045843e-05, "loss": 0.0234, "step": 10798 }, { "epoch": 1.170750216825672, "grad_norm": 0.16681547462940216, "learning_rate": 6.11961410126215e-05, "loss": 0.0089, "step": 10799 }, { "epoch": 1.170858629661752, "grad_norm": 0.17095625400543213, "learning_rate": 6.119251414478456e-05, "loss": 0.015, "step": 10800 }, { "epoch": 1.1709670424978318, "grad_norm": 0.5584606528282166, "learning_rate": 6.118888727694763e-05, "loss": 0.0169, "step": 10801 }, { "epoch": 1.1710754553339116, "grad_norm": 0.1907382756471634, "learning_rate": 6.118526040911069e-05, "loss": 0.0355, "step": 10802 }, { "epoch": 1.1711838681699913, "grad_norm": 0.2661437690258026, "learning_rate": 6.118163354127376e-05, "loss": 0.0133, "step": 10803 }, { "epoch": 1.171292281006071, "grad_norm": 0.20260001718997955, "learning_rate": 6.117800667343683e-05, "loss": 0.0326, "step": 10804 }, { "epoch": 1.1714006938421508, "grad_norm": 0.07896997779607773, "learning_rate": 6.117437980559989e-05, "loss": 0.0098, "step": 10805 }, { "epoch": 1.1715091066782306, "grad_norm": 0.22214436531066895, "learning_rate": 6.117075293776296e-05, "loss": 0.017, "step": 10806 }, { "epoch": 1.1716175195143106, "grad_norm": 0.20412109792232513, "learning_rate": 6.116712606992602e-05, "loss": 0.0132, "step": 10807 }, { "epoch": 1.1717259323503904, "grad_norm": 0.23442000150680542, "learning_rate": 6.116349920208909e-05, "loss": 0.0101, "step": 10808 }, { "epoch": 1.1718343451864701, "grad_norm": 0.07273970544338226, "learning_rate": 6.115987233425214e-05, "loss": 0.0082, "step": 10809 }, { "epoch": 1.1719427580225499, "grad_norm": 0.20309856534004211, "learning_rate": 6.11562454664152e-05, "loss": 0.0248, "step": 10810 }, { "epoch": 1.1720511708586296, "grad_norm": 0.12231037020683289, "learning_rate": 6.115261859857827e-05, "loss": 0.0094, "step": 10811 }, { "epoch": 1.1721595836947094, "grad_norm": 0.20845608413219452, "learning_rate": 6.114899173074133e-05, "loss": 0.0141, "step": 10812 }, { "epoch": 1.1722679965307892, "grad_norm": 0.20895437896251678, "learning_rate": 6.11453648629044e-05, "loss": 0.0212, "step": 10813 }, { "epoch": 1.1723764093668692, "grad_norm": 0.12773334980010986, "learning_rate": 6.114173799506746e-05, "loss": 0.0066, "step": 10814 }, { "epoch": 1.172484822202949, "grad_norm": 0.6645697355270386, "learning_rate": 6.113811112723053e-05, "loss": 0.0334, "step": 10815 }, { "epoch": 1.1725932350390287, "grad_norm": 0.356010764837265, "learning_rate": 6.113448425939359e-05, "loss": 0.0535, "step": 10816 }, { "epoch": 1.1727016478751084, "grad_norm": 0.1201329380273819, "learning_rate": 6.113085739155666e-05, "loss": 0.0062, "step": 10817 }, { "epoch": 1.1728100607111882, "grad_norm": 0.5534487962722778, "learning_rate": 6.112723052371971e-05, "loss": 0.0504, "step": 10818 }, { "epoch": 1.172918473547268, "grad_norm": 0.5206084847450256, "learning_rate": 6.112360365588277e-05, "loss": 0.0556, "step": 10819 }, { "epoch": 1.1730268863833477, "grad_norm": 0.6351212859153748, "learning_rate": 6.111997678804586e-05, "loss": 0.0536, "step": 10820 }, { "epoch": 1.1731352992194275, "grad_norm": 0.4842783510684967, "learning_rate": 6.111634992020891e-05, "loss": 0.0077, "step": 10821 }, { "epoch": 1.1732437120555073, "grad_norm": 0.05546196177601814, "learning_rate": 6.111272305237198e-05, "loss": 0.0041, "step": 10822 }, { "epoch": 1.1733521248915872, "grad_norm": 0.4103368818759918, "learning_rate": 6.110909618453504e-05, "loss": 0.0133, "step": 10823 }, { "epoch": 1.173460537727667, "grad_norm": 0.19914288818836212, "learning_rate": 6.11054693166981e-05, "loss": 0.0146, "step": 10824 }, { "epoch": 1.1735689505637468, "grad_norm": 0.25092679262161255, "learning_rate": 6.110184244886117e-05, "loss": 0.018, "step": 10825 }, { "epoch": 1.1736773633998265, "grad_norm": 0.4575914740562439, "learning_rate": 6.109821558102423e-05, "loss": 0.042, "step": 10826 }, { "epoch": 1.1737857762359063, "grad_norm": 0.1094200387597084, "learning_rate": 6.10945887131873e-05, "loss": 0.0062, "step": 10827 }, { "epoch": 1.173894189071986, "grad_norm": 0.35380440950393677, "learning_rate": 6.109096184535036e-05, "loss": 0.034, "step": 10828 }, { "epoch": 1.1740026019080658, "grad_norm": 0.15767621994018555, "learning_rate": 6.108733497751343e-05, "loss": 0.0108, "step": 10829 }, { "epoch": 1.1741110147441458, "grad_norm": 0.09371817857027054, "learning_rate": 6.108370810967648e-05, "loss": 0.0046, "step": 10830 }, { "epoch": 1.1742194275802256, "grad_norm": 0.25000256299972534, "learning_rate": 6.108008124183955e-05, "loss": 0.0155, "step": 10831 }, { "epoch": 1.1743278404163053, "grad_norm": 0.4244593381881714, "learning_rate": 6.107645437400261e-05, "loss": 0.0536, "step": 10832 }, { "epoch": 1.174436253252385, "grad_norm": 0.08191267400979996, "learning_rate": 6.107282750616567e-05, "loss": 0.0068, "step": 10833 }, { "epoch": 1.1745446660884649, "grad_norm": 0.46209222078323364, "learning_rate": 6.106920063832874e-05, "loss": 0.0186, "step": 10834 }, { "epoch": 1.1746530789245446, "grad_norm": 0.29165416955947876, "learning_rate": 6.10655737704918e-05, "loss": 0.0268, "step": 10835 }, { "epoch": 1.1747614917606244, "grad_norm": 0.08809764683246613, "learning_rate": 6.106194690265487e-05, "loss": 0.0138, "step": 10836 }, { "epoch": 1.1748699045967044, "grad_norm": 0.18620984256267548, "learning_rate": 6.105832003481794e-05, "loss": 0.0157, "step": 10837 }, { "epoch": 1.1749783174327841, "grad_norm": 0.7130591869354248, "learning_rate": 6.1054693166981e-05, "loss": 0.0345, "step": 10838 }, { "epoch": 1.175086730268864, "grad_norm": 0.31592610478401184, "learning_rate": 6.105106629914407e-05, "loss": 0.0296, "step": 10839 }, { "epoch": 1.1751951431049437, "grad_norm": 0.4363715946674347, "learning_rate": 6.104743943130712e-05, "loss": 0.0322, "step": 10840 }, { "epoch": 1.1753035559410234, "grad_norm": 0.285761296749115, "learning_rate": 6.10438125634702e-05, "loss": 0.0267, "step": 10841 }, { "epoch": 1.1754119687771032, "grad_norm": 0.23709213733673096, "learning_rate": 6.104018569563325e-05, "loss": 0.0179, "step": 10842 }, { "epoch": 1.175520381613183, "grad_norm": 0.09608544409275055, "learning_rate": 6.103655882779632e-05, "loss": 0.0109, "step": 10843 }, { "epoch": 1.1756287944492627, "grad_norm": 0.05153648182749748, "learning_rate": 6.103293195995938e-05, "loss": 0.0072, "step": 10844 }, { "epoch": 1.1757372072853425, "grad_norm": 0.02370915375649929, "learning_rate": 6.1029305092122445e-05, "loss": 0.0018, "step": 10845 }, { "epoch": 1.1758456201214225, "grad_norm": 0.22913484275341034, "learning_rate": 6.102567822428551e-05, "loss": 0.0288, "step": 10846 }, { "epoch": 1.1759540329575022, "grad_norm": 0.5655617713928223, "learning_rate": 6.102205135644857e-05, "loss": 0.0349, "step": 10847 }, { "epoch": 1.176062445793582, "grad_norm": 0.29886654019355774, "learning_rate": 6.101842448861164e-05, "loss": 0.0363, "step": 10848 }, { "epoch": 1.1761708586296618, "grad_norm": 0.5217815041542053, "learning_rate": 6.10147976207747e-05, "loss": 0.0576, "step": 10849 }, { "epoch": 1.1762792714657415, "grad_norm": 0.12832701206207275, "learning_rate": 6.1011170752937766e-05, "loss": 0.0059, "step": 10850 }, { "epoch": 1.1763876843018213, "grad_norm": 0.4801783561706543, "learning_rate": 6.100754388510082e-05, "loss": 0.0244, "step": 10851 }, { "epoch": 1.176496097137901, "grad_norm": 0.11525212228298187, "learning_rate": 6.100391701726389e-05, "loss": 0.0081, "step": 10852 }, { "epoch": 1.176604509973981, "grad_norm": 0.10945798456668854, "learning_rate": 6.100029014942695e-05, "loss": 0.0065, "step": 10853 }, { "epoch": 1.1767129228100608, "grad_norm": 0.1640230417251587, "learning_rate": 6.0996663281590015e-05, "loss": 0.0285, "step": 10854 }, { "epoch": 1.1768213356461406, "grad_norm": 0.09530893713235855, "learning_rate": 6.099303641375309e-05, "loss": 0.0067, "step": 10855 }, { "epoch": 1.1769297484822203, "grad_norm": 0.07973334193229675, "learning_rate": 6.098940954591616e-05, "loss": 0.005, "step": 10856 }, { "epoch": 1.1770381613183, "grad_norm": 0.6685022115707397, "learning_rate": 6.0985782678079214e-05, "loss": 0.029, "step": 10857 }, { "epoch": 1.1771465741543798, "grad_norm": 0.2764836251735687, "learning_rate": 6.098215581024228e-05, "loss": 0.026, "step": 10858 }, { "epoch": 1.1772549869904596, "grad_norm": 0.21454882621765137, "learning_rate": 6.097852894240534e-05, "loss": 0.0177, "step": 10859 }, { "epoch": 1.1773633998265394, "grad_norm": 0.23948346078395844, "learning_rate": 6.097490207456841e-05, "loss": 0.0109, "step": 10860 }, { "epoch": 1.1774718126626191, "grad_norm": 0.22111199796199799, "learning_rate": 6.097127520673147e-05, "loss": 0.0242, "step": 10861 }, { "epoch": 1.1775802254986991, "grad_norm": 0.25477904081344604, "learning_rate": 6.0967648338894535e-05, "loss": 0.0188, "step": 10862 }, { "epoch": 1.1776886383347789, "grad_norm": 0.35298192501068115, "learning_rate": 6.09640214710576e-05, "loss": 0.0408, "step": 10863 }, { "epoch": 1.1777970511708586, "grad_norm": 0.6240109205245972, "learning_rate": 6.096039460322066e-05, "loss": 0.0348, "step": 10864 }, { "epoch": 1.1779054640069384, "grad_norm": 0.10707426071166992, "learning_rate": 6.095676773538373e-05, "loss": 0.0026, "step": 10865 }, { "epoch": 1.1780138768430182, "grad_norm": 0.5542619824409485, "learning_rate": 6.0953140867546785e-05, "loss": 0.0096, "step": 10866 }, { "epoch": 1.178122289679098, "grad_norm": 0.37029123306274414, "learning_rate": 6.094951399970985e-05, "loss": 0.04, "step": 10867 }, { "epoch": 1.1782307025151777, "grad_norm": 0.3286343216896057, "learning_rate": 6.094588713187291e-05, "loss": 0.027, "step": 10868 }, { "epoch": 1.1783391153512577, "grad_norm": 0.12053213268518448, "learning_rate": 6.094226026403598e-05, "loss": 0.007, "step": 10869 }, { "epoch": 1.1784475281873374, "grad_norm": 0.5575470924377441, "learning_rate": 6.093863339619904e-05, "loss": 0.0316, "step": 10870 }, { "epoch": 1.1785559410234172, "grad_norm": 0.6165607571601868, "learning_rate": 6.0935006528362105e-05, "loss": 0.0489, "step": 10871 }, { "epoch": 1.178664353859497, "grad_norm": 0.6122984290122986, "learning_rate": 6.0931379660525176e-05, "loss": 0.0749, "step": 10872 }, { "epoch": 1.1787727666955767, "grad_norm": 0.15410470962524414, "learning_rate": 6.092775279268824e-05, "loss": 0.0092, "step": 10873 }, { "epoch": 1.1788811795316565, "grad_norm": 0.6708109378814697, "learning_rate": 6.0924125924851304e-05, "loss": 0.0603, "step": 10874 }, { "epoch": 1.1789895923677363, "grad_norm": 0.5032920837402344, "learning_rate": 6.092049905701437e-05, "loss": 0.0254, "step": 10875 }, { "epoch": 1.1790980052038162, "grad_norm": 0.394366055727005, "learning_rate": 6.091687218917743e-05, "loss": 0.0471, "step": 10876 }, { "epoch": 1.179206418039896, "grad_norm": 0.1566140055656433, "learning_rate": 6.0913245321340496e-05, "loss": 0.0045, "step": 10877 }, { "epoch": 1.1793148308759758, "grad_norm": 0.3156043291091919, "learning_rate": 6.090961845350356e-05, "loss": 0.0364, "step": 10878 }, { "epoch": 1.1794232437120555, "grad_norm": 0.49750906229019165, "learning_rate": 6.0905991585666625e-05, "loss": 0.0471, "step": 10879 }, { "epoch": 1.1795316565481353, "grad_norm": 0.4112883508205414, "learning_rate": 6.090236471782968e-05, "loss": 0.0128, "step": 10880 }, { "epoch": 1.179640069384215, "grad_norm": 0.10910278558731079, "learning_rate": 6.0898737849992746e-05, "loss": 0.0086, "step": 10881 }, { "epoch": 1.1797484822202948, "grad_norm": 0.47972723841667175, "learning_rate": 6.089511098215581e-05, "loss": 0.0273, "step": 10882 }, { "epoch": 1.1798568950563746, "grad_norm": 0.6041974425315857, "learning_rate": 6.0891484114318874e-05, "loss": 0.0283, "step": 10883 }, { "epoch": 1.1799653078924544, "grad_norm": 0.3782718777656555, "learning_rate": 6.088785724648194e-05, "loss": 0.0366, "step": 10884 }, { "epoch": 1.1800737207285343, "grad_norm": 0.32857921719551086, "learning_rate": 6.0884230378645e-05, "loss": 0.0164, "step": 10885 }, { "epoch": 1.180182133564614, "grad_norm": 0.3899663984775543, "learning_rate": 6.088060351080807e-05, "loss": 0.0237, "step": 10886 }, { "epoch": 1.1802905464006939, "grad_norm": 0.018300412222743034, "learning_rate": 6.087697664297113e-05, "loss": 0.0013, "step": 10887 }, { "epoch": 1.1803989592367736, "grad_norm": 0.7814745903015137, "learning_rate": 6.0873349775134195e-05, "loss": 0.0206, "step": 10888 }, { "epoch": 1.1805073720728534, "grad_norm": 0.2967303395271301, "learning_rate": 6.0869722907297266e-05, "loss": 0.0134, "step": 10889 }, { "epoch": 1.1806157849089332, "grad_norm": 0.19443769752979279, "learning_rate": 6.086609603946033e-05, "loss": 0.0145, "step": 10890 }, { "epoch": 1.180724197745013, "grad_norm": 0.730696439743042, "learning_rate": 6.0862469171623394e-05, "loss": 0.0134, "step": 10891 }, { "epoch": 1.180832610581093, "grad_norm": 0.4500270187854767, "learning_rate": 6.085884230378646e-05, "loss": 0.0477, "step": 10892 }, { "epoch": 1.1809410234171727, "grad_norm": 0.23376725614070892, "learning_rate": 6.085521543594952e-05, "loss": 0.0194, "step": 10893 }, { "epoch": 1.1810494362532524, "grad_norm": 0.22106873989105225, "learning_rate": 6.0851588568112586e-05, "loss": 0.0116, "step": 10894 }, { "epoch": 1.1811578490893322, "grad_norm": 0.1826203465461731, "learning_rate": 6.0847961700275644e-05, "loss": 0.0039, "step": 10895 }, { "epoch": 1.181266261925412, "grad_norm": 0.17852336168289185, "learning_rate": 6.084433483243871e-05, "loss": 0.0116, "step": 10896 }, { "epoch": 1.1813746747614917, "grad_norm": 0.06865184009075165, "learning_rate": 6.084070796460177e-05, "loss": 0.0028, "step": 10897 }, { "epoch": 1.1814830875975715, "grad_norm": 0.5189862847328186, "learning_rate": 6.0837081096764836e-05, "loss": 0.0235, "step": 10898 }, { "epoch": 1.1815915004336515, "grad_norm": 0.8364770412445068, "learning_rate": 6.08334542289279e-05, "loss": 0.0388, "step": 10899 }, { "epoch": 1.1816999132697312, "grad_norm": 0.30599719285964966, "learning_rate": 6.0829827361090964e-05, "loss": 0.0149, "step": 10900 }, { "epoch": 1.181808326105811, "grad_norm": 0.12277281284332275, "learning_rate": 6.082620049325403e-05, "loss": 0.0088, "step": 10901 }, { "epoch": 1.1819167389418908, "grad_norm": 0.2376207411289215, "learning_rate": 6.082257362541709e-05, "loss": 0.0323, "step": 10902 }, { "epoch": 1.1820251517779705, "grad_norm": 0.09026821702718735, "learning_rate": 6.0818946757580156e-05, "loss": 0.005, "step": 10903 }, { "epoch": 1.1821335646140503, "grad_norm": 0.12050437927246094, "learning_rate": 6.0815319889743214e-05, "loss": 0.0124, "step": 10904 }, { "epoch": 1.18224197745013, "grad_norm": 0.3321252763271332, "learning_rate": 6.081169302190628e-05, "loss": 0.06, "step": 10905 }, { "epoch": 1.1823503902862098, "grad_norm": 0.7551049590110779, "learning_rate": 6.0808066154069355e-05, "loss": 0.0547, "step": 10906 }, { "epoch": 1.1824588031222896, "grad_norm": 0.4319460988044739, "learning_rate": 6.080443928623242e-05, "loss": 0.0586, "step": 10907 }, { "epoch": 1.1825672159583696, "grad_norm": 0.15670451521873474, "learning_rate": 6.0800812418395484e-05, "loss": 0.0063, "step": 10908 }, { "epoch": 1.1826756287944493, "grad_norm": 0.2657034695148468, "learning_rate": 6.079718555055854e-05, "loss": 0.0183, "step": 10909 }, { "epoch": 1.182784041630529, "grad_norm": 0.16654731333255768, "learning_rate": 6.0793558682721605e-05, "loss": 0.0194, "step": 10910 }, { "epoch": 1.1828924544666088, "grad_norm": 0.38277357816696167, "learning_rate": 6.078993181488467e-05, "loss": 0.0294, "step": 10911 }, { "epoch": 1.1830008673026886, "grad_norm": 0.6655892133712769, "learning_rate": 6.078630494704773e-05, "loss": 0.0195, "step": 10912 }, { "epoch": 1.1831092801387684, "grad_norm": 0.07569229602813721, "learning_rate": 6.07826780792108e-05, "loss": 0.0039, "step": 10913 }, { "epoch": 1.1832176929748481, "grad_norm": 0.20239154994487762, "learning_rate": 6.077905121137386e-05, "loss": 0.0136, "step": 10914 }, { "epoch": 1.1833261058109281, "grad_norm": 0.09859640151262283, "learning_rate": 6.0775424343536926e-05, "loss": 0.0117, "step": 10915 }, { "epoch": 1.1834345186470079, "grad_norm": 0.41559717059135437, "learning_rate": 6.077179747569999e-05, "loss": 0.008, "step": 10916 }, { "epoch": 1.1835429314830876, "grad_norm": 0.14625369012355804, "learning_rate": 6.0768170607863054e-05, "loss": 0.0133, "step": 10917 }, { "epoch": 1.1836513443191674, "grad_norm": 0.22475136816501617, "learning_rate": 6.076454374002611e-05, "loss": 0.0045, "step": 10918 }, { "epoch": 1.1837597571552472, "grad_norm": 0.280465692281723, "learning_rate": 6.0760916872189175e-05, "loss": 0.0262, "step": 10919 }, { "epoch": 1.183868169991327, "grad_norm": 0.7271826863288879, "learning_rate": 6.075729000435224e-05, "loss": 0.0513, "step": 10920 }, { "epoch": 1.1839765828274067, "grad_norm": 0.3171137273311615, "learning_rate": 6.0753663136515304e-05, "loss": 0.0218, "step": 10921 }, { "epoch": 1.1840849956634865, "grad_norm": 0.3865083158016205, "learning_rate": 6.075003626867837e-05, "loss": 0.0185, "step": 10922 }, { "epoch": 1.1841934084995662, "grad_norm": 0.3236336410045624, "learning_rate": 6.074640940084143e-05, "loss": 0.0247, "step": 10923 }, { "epoch": 1.1843018213356462, "grad_norm": 0.3355090916156769, "learning_rate": 6.07427825330045e-05, "loss": 0.0373, "step": 10924 }, { "epoch": 1.184410234171726, "grad_norm": 0.5237906575202942, "learning_rate": 6.073915566516757e-05, "loss": 0.0504, "step": 10925 }, { "epoch": 1.1845186470078057, "grad_norm": 0.32291179895401, "learning_rate": 6.073552879733063e-05, "loss": 0.0255, "step": 10926 }, { "epoch": 1.1846270598438855, "grad_norm": 0.283848375082016, "learning_rate": 6.0731901929493695e-05, "loss": 0.0391, "step": 10927 }, { "epoch": 1.1847354726799653, "grad_norm": 0.2183794230222702, "learning_rate": 6.072827506165676e-05, "loss": 0.0153, "step": 10928 }, { "epoch": 1.184843885516045, "grad_norm": 0.6160529851913452, "learning_rate": 6.072464819381982e-05, "loss": 0.0234, "step": 10929 }, { "epoch": 1.1849522983521248, "grad_norm": 0.21891959011554718, "learning_rate": 6.072102132598289e-05, "loss": 0.0094, "step": 10930 }, { "epoch": 1.1850607111882048, "grad_norm": 0.8282943964004517, "learning_rate": 6.071739445814595e-05, "loss": 0.0188, "step": 10931 }, { "epoch": 1.1851691240242845, "grad_norm": 0.1387803703546524, "learning_rate": 6.071376759030901e-05, "loss": 0.014, "step": 10932 }, { "epoch": 1.1852775368603643, "grad_norm": 0.16799429059028625, "learning_rate": 6.071014072247207e-05, "loss": 0.0061, "step": 10933 }, { "epoch": 1.185385949696444, "grad_norm": 0.11381104588508606, "learning_rate": 6.070651385463514e-05, "loss": 0.0056, "step": 10934 }, { "epoch": 1.1854943625325238, "grad_norm": 0.16627779603004456, "learning_rate": 6.07028869867982e-05, "loss": 0.0133, "step": 10935 }, { "epoch": 1.1856027753686036, "grad_norm": 0.2535349726676941, "learning_rate": 6.0699260118961265e-05, "loss": 0.0098, "step": 10936 }, { "epoch": 1.1857111882046834, "grad_norm": 0.40186911821365356, "learning_rate": 6.069563325112433e-05, "loss": 0.0366, "step": 10937 }, { "epoch": 1.1858196010407633, "grad_norm": 0.4324483275413513, "learning_rate": 6.069200638328739e-05, "loss": 0.0336, "step": 10938 }, { "epoch": 1.185928013876843, "grad_norm": 0.19451305270195007, "learning_rate": 6.068837951545046e-05, "loss": 0.0075, "step": 10939 }, { "epoch": 1.1860364267129229, "grad_norm": 0.3616996109485626, "learning_rate": 6.068475264761352e-05, "loss": 0.0396, "step": 10940 }, { "epoch": 1.1861448395490026, "grad_norm": 0.2852098345756531, "learning_rate": 6.068112577977659e-05, "loss": 0.0194, "step": 10941 }, { "epoch": 1.1862532523850824, "grad_norm": 0.06625574827194214, "learning_rate": 6.0677498911939657e-05, "loss": 0.0076, "step": 10942 }, { "epoch": 1.1863616652211622, "grad_norm": 0.14303849637508392, "learning_rate": 6.067387204410272e-05, "loss": 0.0094, "step": 10943 }, { "epoch": 1.186470078057242, "grad_norm": 0.17356248199939728, "learning_rate": 6.0670245176265785e-05, "loss": 0.0201, "step": 10944 }, { "epoch": 1.1865784908933217, "grad_norm": 0.36279353499412537, "learning_rate": 6.066661830842885e-05, "loss": 0.0121, "step": 10945 }, { "epoch": 1.1866869037294014, "grad_norm": 0.2160796821117401, "learning_rate": 6.066299144059191e-05, "loss": 0.0118, "step": 10946 }, { "epoch": 1.1867953165654814, "grad_norm": 0.09702163189649582, "learning_rate": 6.065936457275497e-05, "loss": 0.0027, "step": 10947 }, { "epoch": 1.1869037294015612, "grad_norm": 0.7559143900871277, "learning_rate": 6.0655737704918034e-05, "loss": 0.054, "step": 10948 }, { "epoch": 1.187012142237641, "grad_norm": 0.4044880270957947, "learning_rate": 6.06521108370811e-05, "loss": 0.0101, "step": 10949 }, { "epoch": 1.1871205550737207, "grad_norm": 0.03411959111690521, "learning_rate": 6.064848396924416e-05, "loss": 0.001, "step": 10950 }, { "epoch": 1.1872289679098005, "grad_norm": 0.16722717881202698, "learning_rate": 6.064485710140723e-05, "loss": 0.0134, "step": 10951 }, { "epoch": 1.1873373807458802, "grad_norm": 0.15788115561008453, "learning_rate": 6.064123023357029e-05, "loss": 0.0113, "step": 10952 }, { "epoch": 1.18744579358196, "grad_norm": 0.6703831553459167, "learning_rate": 6.0637603365733355e-05, "loss": 0.0233, "step": 10953 }, { "epoch": 1.18755420641804, "grad_norm": 0.07858318835496902, "learning_rate": 6.063397649789642e-05, "loss": 0.0069, "step": 10954 }, { "epoch": 1.1876626192541198, "grad_norm": 0.17262506484985352, "learning_rate": 6.063034963005948e-05, "loss": 0.007, "step": 10955 }, { "epoch": 1.1877710320901995, "grad_norm": 0.3716699481010437, "learning_rate": 6.062672276222254e-05, "loss": 0.0184, "step": 10956 }, { "epoch": 1.1878794449262793, "grad_norm": 0.7622672915458679, "learning_rate": 6.0623095894385605e-05, "loss": 0.0368, "step": 10957 }, { "epoch": 1.187987857762359, "grad_norm": 0.661274790763855, "learning_rate": 6.061946902654868e-05, "loss": 0.0359, "step": 10958 }, { "epoch": 1.1880962705984388, "grad_norm": 0.3059854805469513, "learning_rate": 6.0615842158711746e-05, "loss": 0.0088, "step": 10959 }, { "epoch": 1.1882046834345186, "grad_norm": 0.15519259870052338, "learning_rate": 6.061221529087481e-05, "loss": 0.0054, "step": 10960 }, { "epoch": 1.1883130962705986, "grad_norm": 0.042434532195329666, "learning_rate": 6.060858842303787e-05, "loss": 0.0023, "step": 10961 }, { "epoch": 1.1884215091066783, "grad_norm": 0.25381648540496826, "learning_rate": 6.060496155520093e-05, "loss": 0.0067, "step": 10962 }, { "epoch": 1.188529921942758, "grad_norm": 0.23643158376216888, "learning_rate": 6.0601334687363996e-05, "loss": 0.0201, "step": 10963 }, { "epoch": 1.1886383347788378, "grad_norm": 0.2897716462612152, "learning_rate": 6.059770781952706e-05, "loss": 0.0117, "step": 10964 }, { "epoch": 1.1887467476149176, "grad_norm": 0.8552771806716919, "learning_rate": 6.0594080951690124e-05, "loss": 0.0397, "step": 10965 }, { "epoch": 1.1888551604509974, "grad_norm": 0.5495986938476562, "learning_rate": 6.059045408385319e-05, "loss": 0.0392, "step": 10966 }, { "epoch": 1.1889635732870771, "grad_norm": 0.2899765968322754, "learning_rate": 6.058682721601625e-05, "loss": 0.0267, "step": 10967 }, { "epoch": 1.189071986123157, "grad_norm": 0.4243474006652832, "learning_rate": 6.0583200348179316e-05, "loss": 0.0313, "step": 10968 }, { "epoch": 1.1891803989592367, "grad_norm": 0.4822659492492676, "learning_rate": 6.057957348034238e-05, "loss": 0.0195, "step": 10969 }, { "epoch": 1.1892888117953166, "grad_norm": 0.17166171967983246, "learning_rate": 6.057594661250544e-05, "loss": 0.0147, "step": 10970 }, { "epoch": 1.1893972246313964, "grad_norm": 0.8376575708389282, "learning_rate": 6.05723197446685e-05, "loss": 0.0489, "step": 10971 }, { "epoch": 1.1895056374674762, "grad_norm": 0.08243618905544281, "learning_rate": 6.0568692876831566e-05, "loss": 0.006, "step": 10972 }, { "epoch": 1.189614050303556, "grad_norm": 0.28286153078079224, "learning_rate": 6.056506600899463e-05, "loss": 0.0151, "step": 10973 }, { "epoch": 1.1897224631396357, "grad_norm": 0.07391517609357834, "learning_rate": 6.0561439141157694e-05, "loss": 0.0068, "step": 10974 }, { "epoch": 1.1898308759757155, "grad_norm": 0.22112753987312317, "learning_rate": 6.055781227332076e-05, "loss": 0.0136, "step": 10975 }, { "epoch": 1.1899392888117952, "grad_norm": 0.3339085578918457, "learning_rate": 6.055418540548383e-05, "loss": 0.0047, "step": 10976 }, { "epoch": 1.1900477016478752, "grad_norm": 0.5050430297851562, "learning_rate": 6.0550558537646893e-05, "loss": 0.0355, "step": 10977 }, { "epoch": 1.190156114483955, "grad_norm": 0.23176456987857819, "learning_rate": 6.054693166980996e-05, "loss": 0.0141, "step": 10978 }, { "epoch": 1.1902645273200347, "grad_norm": 0.5877217650413513, "learning_rate": 6.054330480197302e-05, "loss": 0.0553, "step": 10979 }, { "epoch": 1.1903729401561145, "grad_norm": 0.5296586751937866, "learning_rate": 6.0539677934136086e-05, "loss": 0.0314, "step": 10980 }, { "epoch": 1.1904813529921943, "grad_norm": 0.7014039158821106, "learning_rate": 6.053605106629915e-05, "loss": 0.0636, "step": 10981 }, { "epoch": 1.190589765828274, "grad_norm": 0.13452798128128052, "learning_rate": 6.0532424198462214e-05, "loss": 0.0093, "step": 10982 }, { "epoch": 1.1906981786643538, "grad_norm": 0.2357071042060852, "learning_rate": 6.052879733062528e-05, "loss": 0.0247, "step": 10983 }, { "epoch": 1.1908065915004338, "grad_norm": 0.32370802760124207, "learning_rate": 6.0525170462788335e-05, "loss": 0.0454, "step": 10984 }, { "epoch": 1.1909150043365135, "grad_norm": 0.17404115200042725, "learning_rate": 6.05215435949514e-05, "loss": 0.0211, "step": 10985 }, { "epoch": 1.1910234171725933, "grad_norm": 0.06682012230157852, "learning_rate": 6.0517916727114464e-05, "loss": 0.0073, "step": 10986 }, { "epoch": 1.191131830008673, "grad_norm": 0.5254805684089661, "learning_rate": 6.051428985927753e-05, "loss": 0.0556, "step": 10987 }, { "epoch": 1.1912402428447528, "grad_norm": 0.3789152204990387, "learning_rate": 6.051066299144059e-05, "loss": 0.0263, "step": 10988 }, { "epoch": 1.1913486556808326, "grad_norm": 0.5061303973197937, "learning_rate": 6.0507036123603656e-05, "loss": 0.0236, "step": 10989 }, { "epoch": 1.1914570685169124, "grad_norm": 0.5946060419082642, "learning_rate": 6.050340925576672e-05, "loss": 0.0086, "step": 10990 }, { "epoch": 1.1915654813529921, "grad_norm": 0.4662104547023773, "learning_rate": 6.0499782387929784e-05, "loss": 0.045, "step": 10991 }, { "epoch": 1.1916738941890719, "grad_norm": 0.1567077934741974, "learning_rate": 6.049615552009285e-05, "loss": 0.009, "step": 10992 }, { "epoch": 1.1917823070251519, "grad_norm": 0.3229333460330963, "learning_rate": 6.049252865225592e-05, "loss": 0.0221, "step": 10993 }, { "epoch": 1.1918907198612316, "grad_norm": 0.5480910539627075, "learning_rate": 6.048890178441898e-05, "loss": 0.0403, "step": 10994 }, { "epoch": 1.1919991326973114, "grad_norm": 0.09705239534378052, "learning_rate": 6.048527491658205e-05, "loss": 0.006, "step": 10995 }, { "epoch": 1.1921075455333912, "grad_norm": 0.29558035731315613, "learning_rate": 6.048164804874511e-05, "loss": 0.0136, "step": 10996 }, { "epoch": 1.192215958369471, "grad_norm": 0.3032423257827759, "learning_rate": 6.0478021180908176e-05, "loss": 0.028, "step": 10997 }, { "epoch": 1.1923243712055507, "grad_norm": 0.21107631921768188, "learning_rate": 6.047439431307124e-05, "loss": 0.0179, "step": 10998 }, { "epoch": 1.1924327840416304, "grad_norm": 0.27273231744766235, "learning_rate": 6.04707674452343e-05, "loss": 0.0247, "step": 10999 }, { "epoch": 1.1925411968777104, "grad_norm": 0.6034331321716309, "learning_rate": 6.046714057739736e-05, "loss": 0.0387, "step": 11000 }, { "epoch": 1.1926496097137902, "grad_norm": 0.6469340324401855, "learning_rate": 6.0463513709560425e-05, "loss": 0.0512, "step": 11001 }, { "epoch": 1.19275802254987, "grad_norm": 0.808799684047699, "learning_rate": 6.045988684172349e-05, "loss": 0.0398, "step": 11002 }, { "epoch": 1.1928664353859497, "grad_norm": 0.22428493201732635, "learning_rate": 6.045625997388655e-05, "loss": 0.0228, "step": 11003 }, { "epoch": 1.1929748482220295, "grad_norm": 0.3031150698661804, "learning_rate": 6.045263310604962e-05, "loss": 0.0235, "step": 11004 }, { "epoch": 1.1930832610581092, "grad_norm": 0.4238199293613434, "learning_rate": 6.044900623821268e-05, "loss": 0.0481, "step": 11005 }, { "epoch": 1.193191673894189, "grad_norm": 0.30436569452285767, "learning_rate": 6.0445379370375746e-05, "loss": 0.0175, "step": 11006 }, { "epoch": 1.1933000867302688, "grad_norm": 0.3350875675678253, "learning_rate": 6.044175250253881e-05, "loss": 0.056, "step": 11007 }, { "epoch": 1.1934084995663485, "grad_norm": 0.167430579662323, "learning_rate": 6.043812563470187e-05, "loss": 0.0202, "step": 11008 }, { "epoch": 1.1935169124024285, "grad_norm": 0.6203456521034241, "learning_rate": 6.043449876686493e-05, "loss": 0.0299, "step": 11009 }, { "epoch": 1.1936253252385083, "grad_norm": 0.2431609183549881, "learning_rate": 6.043087189902801e-05, "loss": 0.0225, "step": 11010 }, { "epoch": 1.193733738074588, "grad_norm": 0.39306753873825073, "learning_rate": 6.042724503119107e-05, "loss": 0.0386, "step": 11011 }, { "epoch": 1.1938421509106678, "grad_norm": 0.11044575273990631, "learning_rate": 6.042361816335414e-05, "loss": 0.0079, "step": 11012 }, { "epoch": 1.1939505637467476, "grad_norm": 0.38921576738357544, "learning_rate": 6.0419991295517194e-05, "loss": 0.022, "step": 11013 }, { "epoch": 1.1940589765828273, "grad_norm": 0.2397375851869583, "learning_rate": 6.041636442768026e-05, "loss": 0.0458, "step": 11014 }, { "epoch": 1.194167389418907, "grad_norm": 0.15536381304264069, "learning_rate": 6.041273755984332e-05, "loss": 0.0093, "step": 11015 }, { "epoch": 1.194275802254987, "grad_norm": 0.37700048089027405, "learning_rate": 6.040911069200639e-05, "loss": 0.0404, "step": 11016 }, { "epoch": 1.1943842150910668, "grad_norm": 0.11074396967887878, "learning_rate": 6.040548382416945e-05, "loss": 0.0097, "step": 11017 }, { "epoch": 1.1944926279271466, "grad_norm": 0.2293982058763504, "learning_rate": 6.0401856956332515e-05, "loss": 0.0097, "step": 11018 }, { "epoch": 1.1946010407632264, "grad_norm": 0.8173733949661255, "learning_rate": 6.039823008849558e-05, "loss": 0.0652, "step": 11019 }, { "epoch": 1.1947094535993061, "grad_norm": 0.4601190686225891, "learning_rate": 6.039460322065864e-05, "loss": 0.045, "step": 11020 }, { "epoch": 1.194817866435386, "grad_norm": 0.27400922775268555, "learning_rate": 6.039097635282171e-05, "loss": 0.0306, "step": 11021 }, { "epoch": 1.1949262792714657, "grad_norm": 0.4399823248386383, "learning_rate": 6.0387349484984765e-05, "loss": 0.0462, "step": 11022 }, { "epoch": 1.1950346921075456, "grad_norm": 0.21693627536296844, "learning_rate": 6.038372261714783e-05, "loss": 0.0197, "step": 11023 }, { "epoch": 1.1951431049436254, "grad_norm": 0.32644692063331604, "learning_rate": 6.038009574931089e-05, "loss": 0.0215, "step": 11024 }, { "epoch": 1.1952515177797052, "grad_norm": 0.21747024357318878, "learning_rate": 6.037646888147396e-05, "loss": 0.0153, "step": 11025 }, { "epoch": 1.195359930615785, "grad_norm": 0.20550848543643951, "learning_rate": 6.037284201363702e-05, "loss": 0.0151, "step": 11026 }, { "epoch": 1.1954683434518647, "grad_norm": 0.34488895535469055, "learning_rate": 6.03692151458001e-05, "loss": 0.0134, "step": 11027 }, { "epoch": 1.1955767562879445, "grad_norm": 0.45408278703689575, "learning_rate": 6.0365588277963156e-05, "loss": 0.021, "step": 11028 }, { "epoch": 1.1956851691240242, "grad_norm": 0.4430682957172394, "learning_rate": 6.036196141012622e-05, "loss": 0.0314, "step": 11029 }, { "epoch": 1.195793581960104, "grad_norm": 0.49288755655288696, "learning_rate": 6.0358334542289284e-05, "loss": 0.0518, "step": 11030 }, { "epoch": 1.1959019947961838, "grad_norm": 0.3703306019306183, "learning_rate": 6.035470767445235e-05, "loss": 0.0406, "step": 11031 }, { "epoch": 1.1960104076322637, "grad_norm": 0.19670912623405457, "learning_rate": 6.035108080661541e-05, "loss": 0.0147, "step": 11032 }, { "epoch": 1.1961188204683435, "grad_norm": 0.2847858965396881, "learning_rate": 6.0347453938778477e-05, "loss": 0.0268, "step": 11033 }, { "epoch": 1.1962272333044233, "grad_norm": 0.20093344151973724, "learning_rate": 6.034382707094154e-05, "loss": 0.0151, "step": 11034 }, { "epoch": 1.196335646140503, "grad_norm": 0.21779288351535797, "learning_rate": 6.0340200203104605e-05, "loss": 0.0137, "step": 11035 }, { "epoch": 1.1964440589765828, "grad_norm": 0.3589995503425598, "learning_rate": 6.033657333526766e-05, "loss": 0.0131, "step": 11036 }, { "epoch": 1.1965524718126626, "grad_norm": 0.22744062542915344, "learning_rate": 6.0332946467430726e-05, "loss": 0.0319, "step": 11037 }, { "epoch": 1.1966608846487423, "grad_norm": 0.22625741362571716, "learning_rate": 6.032931959959379e-05, "loss": 0.012, "step": 11038 }, { "epoch": 1.1967692974848223, "grad_norm": 0.24236004054546356, "learning_rate": 6.0325692731756854e-05, "loss": 0.0233, "step": 11039 }, { "epoch": 1.196877710320902, "grad_norm": 0.45213866233825684, "learning_rate": 6.032206586391992e-05, "loss": 0.051, "step": 11040 }, { "epoch": 1.1969861231569818, "grad_norm": 0.399749219417572, "learning_rate": 6.031843899608298e-05, "loss": 0.0228, "step": 11041 }, { "epoch": 1.1970945359930616, "grad_norm": 0.02500615082681179, "learning_rate": 6.031481212824605e-05, "loss": 0.0022, "step": 11042 }, { "epoch": 1.1972029488291414, "grad_norm": 0.40059250593185425, "learning_rate": 6.031118526040911e-05, "loss": 0.0403, "step": 11043 }, { "epoch": 1.1973113616652211, "grad_norm": 0.7319457530975342, "learning_rate": 6.0307558392572175e-05, "loss": 0.0642, "step": 11044 }, { "epoch": 1.1974197745013009, "grad_norm": 0.16907665133476257, "learning_rate": 6.0303931524735246e-05, "loss": 0.0114, "step": 11045 }, { "epoch": 1.1975281873373809, "grad_norm": 0.6102407574653625, "learning_rate": 6.030030465689831e-05, "loss": 0.0298, "step": 11046 }, { "epoch": 1.1976366001734606, "grad_norm": 0.47241073846817017, "learning_rate": 6.0296677789061374e-05, "loss": 0.0237, "step": 11047 }, { "epoch": 1.1977450130095404, "grad_norm": 0.05472034960985184, "learning_rate": 6.029305092122444e-05, "loss": 0.0053, "step": 11048 }, { "epoch": 1.1978534258456202, "grad_norm": 0.11800097674131393, "learning_rate": 6.02894240533875e-05, "loss": 0.0064, "step": 11049 }, { "epoch": 1.1979618386817, "grad_norm": 0.2467850148677826, "learning_rate": 6.0285797185550566e-05, "loss": 0.0105, "step": 11050 }, { "epoch": 1.1980702515177797, "grad_norm": 0.18659591674804688, "learning_rate": 6.0282170317713624e-05, "loss": 0.023, "step": 11051 }, { "epoch": 1.1981786643538594, "grad_norm": 0.0740790143609047, "learning_rate": 6.027854344987669e-05, "loss": 0.005, "step": 11052 }, { "epoch": 1.1982870771899392, "grad_norm": 0.5575767755508423, "learning_rate": 6.027491658203975e-05, "loss": 0.0475, "step": 11053 }, { "epoch": 1.198395490026019, "grad_norm": 0.19689911603927612, "learning_rate": 6.0271289714202816e-05, "loss": 0.0168, "step": 11054 }, { "epoch": 1.198503902862099, "grad_norm": 0.09925101697444916, "learning_rate": 6.026766284636588e-05, "loss": 0.0163, "step": 11055 }, { "epoch": 1.1986123156981787, "grad_norm": 0.11994393169879913, "learning_rate": 6.0264035978528944e-05, "loss": 0.012, "step": 11056 }, { "epoch": 1.1987207285342585, "grad_norm": 0.21506860852241516, "learning_rate": 6.026040911069201e-05, "loss": 0.0157, "step": 11057 }, { "epoch": 1.1988291413703382, "grad_norm": 0.4784945249557495, "learning_rate": 6.025678224285507e-05, "loss": 0.0522, "step": 11058 }, { "epoch": 1.198937554206418, "grad_norm": 0.27810731530189514, "learning_rate": 6.0253155375018136e-05, "loss": 0.0119, "step": 11059 }, { "epoch": 1.1990459670424978, "grad_norm": 0.2058691531419754, "learning_rate": 6.0249528507181194e-05, "loss": 0.0079, "step": 11060 }, { "epoch": 1.1991543798785775, "grad_norm": 0.14057990908622742, "learning_rate": 6.024590163934426e-05, "loss": 0.0132, "step": 11061 }, { "epoch": 1.1992627927146575, "grad_norm": 0.2063267081975937, "learning_rate": 6.0242274771507336e-05, "loss": 0.0175, "step": 11062 }, { "epoch": 1.1993712055507373, "grad_norm": 0.17098745703697205, "learning_rate": 6.02386479036704e-05, "loss": 0.0125, "step": 11063 }, { "epoch": 1.199479618386817, "grad_norm": 0.055443864315748215, "learning_rate": 6.0235021035833464e-05, "loss": 0.0018, "step": 11064 }, { "epoch": 1.1995880312228968, "grad_norm": 0.8765981197357178, "learning_rate": 6.023139416799652e-05, "loss": 0.0557, "step": 11065 }, { "epoch": 1.1996964440589766, "grad_norm": 0.27952393889427185, "learning_rate": 6.0227767300159585e-05, "loss": 0.0188, "step": 11066 }, { "epoch": 1.1998048568950563, "grad_norm": 0.1290396749973297, "learning_rate": 6.022414043232265e-05, "loss": 0.009, "step": 11067 }, { "epoch": 1.199913269731136, "grad_norm": 0.48013901710510254, "learning_rate": 6.0220513564485713e-05, "loss": 0.017, "step": 11068 }, { "epoch": 1.2000216825672159, "grad_norm": 0.35380810499191284, "learning_rate": 6.021688669664878e-05, "loss": 0.029, "step": 11069 }, { "epoch": 1.2001300954032956, "grad_norm": 0.3191160261631012, "learning_rate": 6.021325982881184e-05, "loss": 0.0415, "step": 11070 }, { "epoch": 1.2002385082393756, "grad_norm": 0.19943217933177948, "learning_rate": 6.0209632960974906e-05, "loss": 0.0121, "step": 11071 }, { "epoch": 1.2003469210754554, "grad_norm": 0.1703420728445053, "learning_rate": 6.020600609313797e-05, "loss": 0.0173, "step": 11072 }, { "epoch": 1.2004553339115351, "grad_norm": 0.8047732710838318, "learning_rate": 6.0202379225301034e-05, "loss": 0.0183, "step": 11073 }, { "epoch": 1.200563746747615, "grad_norm": 0.5613587498664856, "learning_rate": 6.019875235746409e-05, "loss": 0.0371, "step": 11074 }, { "epoch": 1.2006721595836947, "grad_norm": 0.6028012633323669, "learning_rate": 6.0195125489627155e-05, "loss": 0.0326, "step": 11075 }, { "epoch": 1.2007805724197744, "grad_norm": 0.3668416142463684, "learning_rate": 6.019149862179022e-05, "loss": 0.0279, "step": 11076 }, { "epoch": 1.2008889852558542, "grad_norm": 0.5033845901489258, "learning_rate": 6.0187871753953284e-05, "loss": 0.0328, "step": 11077 }, { "epoch": 1.2009973980919342, "grad_norm": 0.18825358152389526, "learning_rate": 6.018424488611635e-05, "loss": 0.0184, "step": 11078 }, { "epoch": 1.201105810928014, "grad_norm": 0.8134579062461853, "learning_rate": 6.0180618018279425e-05, "loss": 0.0581, "step": 11079 }, { "epoch": 1.2012142237640937, "grad_norm": 0.39748886227607727, "learning_rate": 6.017699115044248e-05, "loss": 0.0837, "step": 11080 }, { "epoch": 1.2013226366001735, "grad_norm": 0.04663189873099327, "learning_rate": 6.017336428260555e-05, "loss": 0.0026, "step": 11081 }, { "epoch": 1.2014310494362532, "grad_norm": 0.08019008487462997, "learning_rate": 6.016973741476861e-05, "loss": 0.0106, "step": 11082 }, { "epoch": 1.201539462272333, "grad_norm": 0.17194104194641113, "learning_rate": 6.0166110546931675e-05, "loss": 0.0149, "step": 11083 }, { "epoch": 1.2016478751084128, "grad_norm": 0.2538096010684967, "learning_rate": 6.016248367909474e-05, "loss": 0.0175, "step": 11084 }, { "epoch": 1.2017562879444927, "grad_norm": 0.7080484628677368, "learning_rate": 6.01588568112578e-05, "loss": 0.0222, "step": 11085 }, { "epoch": 1.2018647007805725, "grad_norm": 0.16093333065509796, "learning_rate": 6.015522994342087e-05, "loss": 0.0052, "step": 11086 }, { "epoch": 1.2019731136166523, "grad_norm": 0.3611312806606293, "learning_rate": 6.015160307558393e-05, "loss": 0.0493, "step": 11087 }, { "epoch": 1.202081526452732, "grad_norm": 0.3143473267555237, "learning_rate": 6.0147976207746996e-05, "loss": 0.0364, "step": 11088 }, { "epoch": 1.2021899392888118, "grad_norm": 0.2551369369029999, "learning_rate": 6.014434933991005e-05, "loss": 0.0375, "step": 11089 }, { "epoch": 1.2022983521248916, "grad_norm": 0.18893982470035553, "learning_rate": 6.014072247207312e-05, "loss": 0.0209, "step": 11090 }, { "epoch": 1.2024067649609713, "grad_norm": 0.2392355501651764, "learning_rate": 6.013709560423618e-05, "loss": 0.0217, "step": 11091 }, { "epoch": 1.202515177797051, "grad_norm": 0.4564695656299591, "learning_rate": 6.0133468736399245e-05, "loss": 0.0158, "step": 11092 }, { "epoch": 1.2026235906331308, "grad_norm": 0.3993045687675476, "learning_rate": 6.012984186856231e-05, "loss": 0.0178, "step": 11093 }, { "epoch": 1.2027320034692108, "grad_norm": 0.3146633803844452, "learning_rate": 6.012621500072537e-05, "loss": 0.0209, "step": 11094 }, { "epoch": 1.2028404163052906, "grad_norm": 0.3781563639640808, "learning_rate": 6.012258813288844e-05, "loss": 0.0258, "step": 11095 }, { "epoch": 1.2029488291413704, "grad_norm": 0.006960668601095676, "learning_rate": 6.011896126505151e-05, "loss": 0.0008, "step": 11096 }, { "epoch": 1.2030572419774501, "grad_norm": 0.3271784484386444, "learning_rate": 6.011533439721457e-05, "loss": 0.013, "step": 11097 }, { "epoch": 1.2031656548135299, "grad_norm": 0.17914442718029022, "learning_rate": 6.0111707529377637e-05, "loss": 0.0265, "step": 11098 }, { "epoch": 1.2032740676496096, "grad_norm": 0.15158069133758545, "learning_rate": 6.01080806615407e-05, "loss": 0.0065, "step": 11099 }, { "epoch": 1.2033824804856894, "grad_norm": 0.2894028425216675, "learning_rate": 6.0104453793703765e-05, "loss": 0.0291, "step": 11100 }, { "epoch": 1.2034908933217694, "grad_norm": 0.09494257718324661, "learning_rate": 6.010082692586683e-05, "loss": 0.0054, "step": 11101 }, { "epoch": 1.2035993061578492, "grad_norm": 0.2651996612548828, "learning_rate": 6.009720005802989e-05, "loss": 0.0157, "step": 11102 }, { "epoch": 1.203707718993929, "grad_norm": 0.15954431891441345, "learning_rate": 6.009357319019295e-05, "loss": 0.009, "step": 11103 }, { "epoch": 1.2038161318300087, "grad_norm": 0.22045506536960602, "learning_rate": 6.0089946322356014e-05, "loss": 0.0166, "step": 11104 }, { "epoch": 1.2039245446660884, "grad_norm": 0.1146371066570282, "learning_rate": 6.008631945451908e-05, "loss": 0.0133, "step": 11105 }, { "epoch": 1.2040329575021682, "grad_norm": 0.38244593143463135, "learning_rate": 6.008269258668214e-05, "loss": 0.0619, "step": 11106 }, { "epoch": 1.204141370338248, "grad_norm": 0.297760546207428, "learning_rate": 6.007906571884521e-05, "loss": 0.0155, "step": 11107 }, { "epoch": 1.204249783174328, "grad_norm": 0.24030685424804688, "learning_rate": 6.007543885100827e-05, "loss": 0.0354, "step": 11108 }, { "epoch": 1.2043581960104077, "grad_norm": 0.15645799040794373, "learning_rate": 6.0071811983171335e-05, "loss": 0.0078, "step": 11109 }, { "epoch": 1.2044666088464875, "grad_norm": 0.4581066071987152, "learning_rate": 6.00681851153344e-05, "loss": 0.0255, "step": 11110 }, { "epoch": 1.2045750216825672, "grad_norm": 0.1582203060388565, "learning_rate": 6.006455824749746e-05, "loss": 0.0118, "step": 11111 }, { "epoch": 1.204683434518647, "grad_norm": 0.21033377945423126, "learning_rate": 6.006093137966052e-05, "loss": 0.0093, "step": 11112 }, { "epoch": 1.2047918473547268, "grad_norm": 0.5128987431526184, "learning_rate": 6.0057304511823585e-05, "loss": 0.028, "step": 11113 }, { "epoch": 1.2049002601908065, "grad_norm": 0.39697617292404175, "learning_rate": 6.005367764398666e-05, "loss": 0.0236, "step": 11114 }, { "epoch": 1.2050086730268863, "grad_norm": 0.08062652498483658, "learning_rate": 6.0050050776149726e-05, "loss": 0.0071, "step": 11115 }, { "epoch": 1.205117085862966, "grad_norm": 0.6697916984558105, "learning_rate": 6.004642390831279e-05, "loss": 0.0312, "step": 11116 }, { "epoch": 1.205225498699046, "grad_norm": 0.5696442723274231, "learning_rate": 6.004279704047585e-05, "loss": 0.0274, "step": 11117 }, { "epoch": 1.2053339115351258, "grad_norm": 0.03662293404340744, "learning_rate": 6.003917017263891e-05, "loss": 0.003, "step": 11118 }, { "epoch": 1.2054423243712056, "grad_norm": 0.21125128865242004, "learning_rate": 6.0035543304801976e-05, "loss": 0.0214, "step": 11119 }, { "epoch": 1.2055507372072853, "grad_norm": 0.38687458634376526, "learning_rate": 6.003191643696504e-05, "loss": 0.0178, "step": 11120 }, { "epoch": 1.205659150043365, "grad_norm": 0.3962162733078003, "learning_rate": 6.0028289569128104e-05, "loss": 0.0187, "step": 11121 }, { "epoch": 1.2057675628794449, "grad_norm": 0.292776495218277, "learning_rate": 6.002466270129117e-05, "loss": 0.0113, "step": 11122 }, { "epoch": 1.2058759757155246, "grad_norm": 0.061514705419540405, "learning_rate": 6.002103583345423e-05, "loss": 0.0024, "step": 11123 }, { "epoch": 1.2059843885516046, "grad_norm": 0.3405141830444336, "learning_rate": 6.0017408965617297e-05, "loss": 0.0302, "step": 11124 }, { "epoch": 1.2060928013876844, "grad_norm": 0.3418908417224884, "learning_rate": 6.001378209778036e-05, "loss": 0.0401, "step": 11125 }, { "epoch": 1.2062012142237641, "grad_norm": 0.14040280878543854, "learning_rate": 6.001015522994342e-05, "loss": 0.0132, "step": 11126 }, { "epoch": 1.206309627059844, "grad_norm": 0.9149913191795349, "learning_rate": 6.000652836210648e-05, "loss": 0.0391, "step": 11127 }, { "epoch": 1.2064180398959237, "grad_norm": 0.9874237179756165, "learning_rate": 6.0002901494269546e-05, "loss": 0.0363, "step": 11128 }, { "epoch": 1.2065264527320034, "grad_norm": 0.11668366938829422, "learning_rate": 5.999927462643261e-05, "loss": 0.0058, "step": 11129 }, { "epoch": 1.2066348655680832, "grad_norm": 0.10673113912343979, "learning_rate": 5.9995647758595674e-05, "loss": 0.0095, "step": 11130 }, { "epoch": 1.2067432784041632, "grad_norm": 0.1639566719532013, "learning_rate": 5.999202089075875e-05, "loss": 0.0123, "step": 11131 }, { "epoch": 1.206851691240243, "grad_norm": 0.04583308845758438, "learning_rate": 5.998839402292181e-05, "loss": 0.003, "step": 11132 }, { "epoch": 1.2069601040763227, "grad_norm": 0.23661476373672485, "learning_rate": 5.9984767155084873e-05, "loss": 0.0114, "step": 11133 }, { "epoch": 1.2070685169124025, "grad_norm": 0.09808573871850967, "learning_rate": 5.998114028724794e-05, "loss": 0.0039, "step": 11134 }, { "epoch": 1.2071769297484822, "grad_norm": 0.32950299978256226, "learning_rate": 5.9977513419411e-05, "loss": 0.0237, "step": 11135 }, { "epoch": 1.207285342584562, "grad_norm": 0.21742141246795654, "learning_rate": 5.9973886551574066e-05, "loss": 0.0324, "step": 11136 }, { "epoch": 1.2073937554206418, "grad_norm": 0.7295750975608826, "learning_rate": 5.997025968373713e-05, "loss": 0.0305, "step": 11137 }, { "epoch": 1.2075021682567215, "grad_norm": 0.5319674611091614, "learning_rate": 5.9966632815900194e-05, "loss": 0.0316, "step": 11138 }, { "epoch": 1.2076105810928013, "grad_norm": 0.14976492524147034, "learning_rate": 5.996300594806326e-05, "loss": 0.0182, "step": 11139 }, { "epoch": 1.2077189939288813, "grad_norm": 0.10333419591188431, "learning_rate": 5.995937908022632e-05, "loss": 0.0072, "step": 11140 }, { "epoch": 1.207827406764961, "grad_norm": 0.5928612351417542, "learning_rate": 5.995575221238938e-05, "loss": 0.0346, "step": 11141 }, { "epoch": 1.2079358196010408, "grad_norm": 0.9046692848205566, "learning_rate": 5.9952125344552444e-05, "loss": 0.0393, "step": 11142 }, { "epoch": 1.2080442324371206, "grad_norm": 0.2897244989871979, "learning_rate": 5.994849847671551e-05, "loss": 0.0222, "step": 11143 }, { "epoch": 1.2081526452732003, "grad_norm": 0.2443067729473114, "learning_rate": 5.994487160887857e-05, "loss": 0.0453, "step": 11144 }, { "epoch": 1.20826105810928, "grad_norm": 0.08101960271596909, "learning_rate": 5.9941244741041636e-05, "loss": 0.0043, "step": 11145 }, { "epoch": 1.2083694709453598, "grad_norm": 0.5404394865036011, "learning_rate": 5.99376178732047e-05, "loss": 0.0269, "step": 11146 }, { "epoch": 1.2084778837814398, "grad_norm": 0.1606234759092331, "learning_rate": 5.9933991005367764e-05, "loss": 0.0098, "step": 11147 }, { "epoch": 1.2085862966175196, "grad_norm": 0.3334566354751587, "learning_rate": 5.9930364137530835e-05, "loss": 0.0361, "step": 11148 }, { "epoch": 1.2086947094535994, "grad_norm": 0.7833870649337769, "learning_rate": 5.99267372696939e-05, "loss": 0.038, "step": 11149 }, { "epoch": 1.2088031222896791, "grad_norm": 0.3252725303173065, "learning_rate": 5.992311040185696e-05, "loss": 0.0385, "step": 11150 }, { "epoch": 1.2089115351257589, "grad_norm": 0.3608868420124054, "learning_rate": 5.991948353402003e-05, "loss": 0.02, "step": 11151 }, { "epoch": 1.2090199479618386, "grad_norm": 0.08636119961738586, "learning_rate": 5.991585666618309e-05, "loss": 0.0065, "step": 11152 }, { "epoch": 1.2091283607979184, "grad_norm": 0.42338496446609497, "learning_rate": 5.9912229798346156e-05, "loss": 0.0554, "step": 11153 }, { "epoch": 1.2092367736339982, "grad_norm": 0.4845641851425171, "learning_rate": 5.990860293050922e-05, "loss": 0.0233, "step": 11154 }, { "epoch": 1.209345186470078, "grad_norm": 0.3926374614238739, "learning_rate": 5.990497606267228e-05, "loss": 0.024, "step": 11155 }, { "epoch": 1.209453599306158, "grad_norm": 0.31600403785705566, "learning_rate": 5.990134919483534e-05, "loss": 0.0311, "step": 11156 }, { "epoch": 1.2095620121422377, "grad_norm": 0.44039833545684814, "learning_rate": 5.9897722326998405e-05, "loss": 0.0632, "step": 11157 }, { "epoch": 1.2096704249783174, "grad_norm": 0.06654589623212814, "learning_rate": 5.989409545916147e-05, "loss": 0.0024, "step": 11158 }, { "epoch": 1.2097788378143972, "grad_norm": 0.643283486366272, "learning_rate": 5.9890468591324533e-05, "loss": 0.0783, "step": 11159 }, { "epoch": 1.209887250650477, "grad_norm": 0.34519490599632263, "learning_rate": 5.98868417234876e-05, "loss": 0.0325, "step": 11160 }, { "epoch": 1.2099956634865567, "grad_norm": 0.3610798418521881, "learning_rate": 5.988321485565066e-05, "loss": 0.0255, "step": 11161 }, { "epoch": 1.2101040763226365, "grad_norm": 0.3632277250289917, "learning_rate": 5.9879587987813726e-05, "loss": 0.0327, "step": 11162 }, { "epoch": 1.2102124891587165, "grad_norm": 0.3755524456501007, "learning_rate": 5.987596111997679e-05, "loss": 0.0243, "step": 11163 }, { "epoch": 1.2103209019947962, "grad_norm": 1.0142406225204468, "learning_rate": 5.987233425213985e-05, "loss": 0.0276, "step": 11164 }, { "epoch": 1.210429314830876, "grad_norm": 0.17022806406021118, "learning_rate": 5.986870738430291e-05, "loss": 0.0162, "step": 11165 }, { "epoch": 1.2105377276669558, "grad_norm": 0.3404982089996338, "learning_rate": 5.986508051646599e-05, "loss": 0.0229, "step": 11166 }, { "epoch": 1.2106461405030355, "grad_norm": 0.13092797994613647, "learning_rate": 5.986145364862905e-05, "loss": 0.0093, "step": 11167 }, { "epoch": 1.2107545533391153, "grad_norm": 0.4569633901119232, "learning_rate": 5.985782678079212e-05, "loss": 0.0531, "step": 11168 }, { "epoch": 1.210862966175195, "grad_norm": 0.642291784286499, "learning_rate": 5.9854199912955174e-05, "loss": 0.0357, "step": 11169 }, { "epoch": 1.210971379011275, "grad_norm": 0.09751980006694794, "learning_rate": 5.985057304511824e-05, "loss": 0.0099, "step": 11170 }, { "epoch": 1.2110797918473548, "grad_norm": 0.02460041269659996, "learning_rate": 5.98469461772813e-05, "loss": 0.0011, "step": 11171 }, { "epoch": 1.2111882046834346, "grad_norm": 0.2025415450334549, "learning_rate": 5.984331930944437e-05, "loss": 0.0167, "step": 11172 }, { "epoch": 1.2112966175195143, "grad_norm": 0.4348982274532318, "learning_rate": 5.983969244160743e-05, "loss": 0.0334, "step": 11173 }, { "epoch": 1.211405030355594, "grad_norm": 0.1714739203453064, "learning_rate": 5.9836065573770495e-05, "loss": 0.0213, "step": 11174 }, { "epoch": 1.2115134431916739, "grad_norm": 0.06661912053823471, "learning_rate": 5.983243870593356e-05, "loss": 0.0049, "step": 11175 }, { "epoch": 1.2116218560277536, "grad_norm": 0.20038340985774994, "learning_rate": 5.982881183809662e-05, "loss": 0.0241, "step": 11176 }, { "epoch": 1.2117302688638334, "grad_norm": 1.3119105100631714, "learning_rate": 5.982518497025969e-05, "loss": 0.042, "step": 11177 }, { "epoch": 1.2118386816999132, "grad_norm": 0.19930797815322876, "learning_rate": 5.9821558102422745e-05, "loss": 0.008, "step": 11178 }, { "epoch": 1.2119470945359931, "grad_norm": 0.7413843870162964, "learning_rate": 5.981793123458581e-05, "loss": 0.0368, "step": 11179 }, { "epoch": 1.212055507372073, "grad_norm": 0.17011933028697968, "learning_rate": 5.981430436674887e-05, "loss": 0.0055, "step": 11180 }, { "epoch": 1.2121639202081527, "grad_norm": 0.27604469656944275, "learning_rate": 5.981067749891194e-05, "loss": 0.0166, "step": 11181 }, { "epoch": 1.2122723330442324, "grad_norm": 0.5661519765853882, "learning_rate": 5.9807050631075e-05, "loss": 0.0239, "step": 11182 }, { "epoch": 1.2123807458803122, "grad_norm": 0.34915536642074585, "learning_rate": 5.980342376323808e-05, "loss": 0.012, "step": 11183 }, { "epoch": 1.212489158716392, "grad_norm": 0.033065490424633026, "learning_rate": 5.9799796895401136e-05, "loss": 0.0012, "step": 11184 }, { "epoch": 1.2125975715524717, "grad_norm": 0.2592449486255646, "learning_rate": 5.97961700275642e-05, "loss": 0.0199, "step": 11185 }, { "epoch": 1.2127059843885517, "grad_norm": 0.28273335099220276, "learning_rate": 5.9792543159727264e-05, "loss": 0.0421, "step": 11186 }, { "epoch": 1.2128143972246315, "grad_norm": 0.17976361513137817, "learning_rate": 5.978891629189033e-05, "loss": 0.0329, "step": 11187 }, { "epoch": 1.2129228100607112, "grad_norm": 1.131337285041809, "learning_rate": 5.978528942405339e-05, "loss": 0.0638, "step": 11188 }, { "epoch": 1.213031222896791, "grad_norm": 0.339344322681427, "learning_rate": 5.9781662556216457e-05, "loss": 0.019, "step": 11189 }, { "epoch": 1.2131396357328708, "grad_norm": 0.567927360534668, "learning_rate": 5.977803568837952e-05, "loss": 0.0401, "step": 11190 }, { "epoch": 1.2132480485689505, "grad_norm": 0.6472971439361572, "learning_rate": 5.9774408820542585e-05, "loss": 0.0409, "step": 11191 }, { "epoch": 1.2133564614050303, "grad_norm": 0.3278976082801819, "learning_rate": 5.977078195270565e-05, "loss": 0.0192, "step": 11192 }, { "epoch": 1.2134648742411103, "grad_norm": 1.2888426780700684, "learning_rate": 5.9767155084868706e-05, "loss": 0.0498, "step": 11193 }, { "epoch": 1.21357328707719, "grad_norm": 0.10950330644845963, "learning_rate": 5.976352821703177e-05, "loss": 0.0061, "step": 11194 }, { "epoch": 1.2136816999132698, "grad_norm": 0.21333260834217072, "learning_rate": 5.9759901349194834e-05, "loss": 0.0096, "step": 11195 }, { "epoch": 1.2137901127493496, "grad_norm": 0.4081534445285797, "learning_rate": 5.97562744813579e-05, "loss": 0.0154, "step": 11196 }, { "epoch": 1.2138985255854293, "grad_norm": 0.17587199807167053, "learning_rate": 5.975264761352096e-05, "loss": 0.0133, "step": 11197 }, { "epoch": 1.214006938421509, "grad_norm": 0.5464602112770081, "learning_rate": 5.974902074568403e-05, "loss": 0.0846, "step": 11198 }, { "epoch": 1.2141153512575888, "grad_norm": 0.4050174057483673, "learning_rate": 5.974539387784709e-05, "loss": 0.0125, "step": 11199 }, { "epoch": 1.2142237640936686, "grad_norm": 0.606076180934906, "learning_rate": 5.974176701001016e-05, "loss": 0.0272, "step": 11200 }, { "epoch": 1.2143321769297484, "grad_norm": 0.08964681625366211, "learning_rate": 5.9738140142173226e-05, "loss": 0.0118, "step": 11201 }, { "epoch": 1.2144405897658284, "grad_norm": 0.7980536818504333, "learning_rate": 5.973451327433629e-05, "loss": 0.036, "step": 11202 }, { "epoch": 1.2145490026019081, "grad_norm": 0.1300482600927353, "learning_rate": 5.9730886406499354e-05, "loss": 0.0056, "step": 11203 }, { "epoch": 1.2146574154379879, "grad_norm": 0.27554309368133545, "learning_rate": 5.972725953866242e-05, "loss": 0.0261, "step": 11204 }, { "epoch": 1.2147658282740676, "grad_norm": 0.3790702521800995, "learning_rate": 5.972363267082548e-05, "loss": 0.0184, "step": 11205 }, { "epoch": 1.2148742411101474, "grad_norm": 0.2034929096698761, "learning_rate": 5.9720005802988546e-05, "loss": 0.0139, "step": 11206 }, { "epoch": 1.2149826539462272, "grad_norm": 0.1249428391456604, "learning_rate": 5.9716378935151604e-05, "loss": 0.0058, "step": 11207 }, { "epoch": 1.215091066782307, "grad_norm": 0.6434204578399658, "learning_rate": 5.971275206731467e-05, "loss": 0.045, "step": 11208 }, { "epoch": 1.215199479618387, "grad_norm": 0.12713907659053802, "learning_rate": 5.970912519947773e-05, "loss": 0.0067, "step": 11209 }, { "epoch": 1.2153078924544667, "grad_norm": 1.2265046834945679, "learning_rate": 5.9705498331640796e-05, "loss": 0.069, "step": 11210 }, { "epoch": 1.2154163052905465, "grad_norm": 0.08131526410579681, "learning_rate": 5.970187146380386e-05, "loss": 0.0052, "step": 11211 }, { "epoch": 1.2155247181266262, "grad_norm": 0.07950893044471741, "learning_rate": 5.9698244595966924e-05, "loss": 0.0079, "step": 11212 }, { "epoch": 1.215633130962706, "grad_norm": 0.38363102078437805, "learning_rate": 5.969461772812999e-05, "loss": 0.0249, "step": 11213 }, { "epoch": 1.2157415437987857, "grad_norm": 0.049773912876844406, "learning_rate": 5.969099086029305e-05, "loss": 0.0033, "step": 11214 }, { "epoch": 1.2158499566348655, "grad_norm": 0.4350793659687042, "learning_rate": 5.9687363992456117e-05, "loss": 0.0242, "step": 11215 }, { "epoch": 1.2159583694709453, "grad_norm": 0.0615709125995636, "learning_rate": 5.9683737124619174e-05, "loss": 0.0043, "step": 11216 }, { "epoch": 1.216066782307025, "grad_norm": 0.3142652213573456, "learning_rate": 5.968011025678225e-05, "loss": 0.0147, "step": 11217 }, { "epoch": 1.216175195143105, "grad_norm": 0.1349906176328659, "learning_rate": 5.9676483388945316e-05, "loss": 0.0119, "step": 11218 }, { "epoch": 1.2162836079791848, "grad_norm": 0.07999727874994278, "learning_rate": 5.967285652110838e-05, "loss": 0.0064, "step": 11219 }, { "epoch": 1.2163920208152645, "grad_norm": 0.16373105347156525, "learning_rate": 5.9669229653271444e-05, "loss": 0.0205, "step": 11220 }, { "epoch": 1.2165004336513443, "grad_norm": 0.16349203884601593, "learning_rate": 5.96656027854345e-05, "loss": 0.0067, "step": 11221 }, { "epoch": 1.216608846487424, "grad_norm": 0.41207221150398254, "learning_rate": 5.9661975917597565e-05, "loss": 0.0425, "step": 11222 }, { "epoch": 1.2167172593235038, "grad_norm": 0.4827723205089569, "learning_rate": 5.965834904976063e-05, "loss": 0.0251, "step": 11223 }, { "epoch": 1.2168256721595836, "grad_norm": 0.4388124346733093, "learning_rate": 5.9654722181923693e-05, "loss": 0.0585, "step": 11224 }, { "epoch": 1.2169340849956636, "grad_norm": 0.32749143242836, "learning_rate": 5.965109531408676e-05, "loss": 0.0151, "step": 11225 }, { "epoch": 1.2170424978317433, "grad_norm": 0.540225088596344, "learning_rate": 5.964746844624982e-05, "loss": 0.0295, "step": 11226 }, { "epoch": 1.217150910667823, "grad_norm": 0.21174465119838715, "learning_rate": 5.9643841578412886e-05, "loss": 0.021, "step": 11227 }, { "epoch": 1.2172593235039029, "grad_norm": 0.24326428771018982, "learning_rate": 5.964021471057595e-05, "loss": 0.0243, "step": 11228 }, { "epoch": 1.2173677363399826, "grad_norm": 0.29946085810661316, "learning_rate": 5.9636587842739014e-05, "loss": 0.0129, "step": 11229 }, { "epoch": 1.2174761491760624, "grad_norm": 0.5988306999206543, "learning_rate": 5.963296097490207e-05, "loss": 0.0796, "step": 11230 }, { "epoch": 1.2175845620121422, "grad_norm": 0.4638775587081909, "learning_rate": 5.9629334107065135e-05, "loss": 0.0117, "step": 11231 }, { "epoch": 1.2176929748482221, "grad_norm": 0.135441392660141, "learning_rate": 5.96257072392282e-05, "loss": 0.0073, "step": 11232 }, { "epoch": 1.217801387684302, "grad_norm": 0.250658243894577, "learning_rate": 5.9622080371391264e-05, "loss": 0.0335, "step": 11233 }, { "epoch": 1.2179098005203817, "grad_norm": 0.3332378566265106, "learning_rate": 5.961845350355433e-05, "loss": 0.0165, "step": 11234 }, { "epoch": 1.2180182133564614, "grad_norm": 0.3250250518321991, "learning_rate": 5.9614826635717405e-05, "loss": 0.0267, "step": 11235 }, { "epoch": 1.2181266261925412, "grad_norm": 0.14140908420085907, "learning_rate": 5.961119976788046e-05, "loss": 0.0133, "step": 11236 }, { "epoch": 1.218235039028621, "grad_norm": 0.1656959354877472, "learning_rate": 5.960757290004353e-05, "loss": 0.0185, "step": 11237 }, { "epoch": 1.2183434518647007, "grad_norm": 0.3679700493812561, "learning_rate": 5.960394603220659e-05, "loss": 0.0466, "step": 11238 }, { "epoch": 1.2184518647007805, "grad_norm": 0.32746919989585876, "learning_rate": 5.9600319164369655e-05, "loss": 0.0362, "step": 11239 }, { "epoch": 1.2185602775368602, "grad_norm": 0.3129179775714874, "learning_rate": 5.959669229653272e-05, "loss": 0.0197, "step": 11240 }, { "epoch": 1.2186686903729402, "grad_norm": 0.13093939423561096, "learning_rate": 5.959306542869578e-05, "loss": 0.0039, "step": 11241 }, { "epoch": 1.21877710320902, "grad_norm": 0.22856420278549194, "learning_rate": 5.958943856085885e-05, "loss": 0.0084, "step": 11242 }, { "epoch": 1.2188855160450998, "grad_norm": 0.16911420226097107, "learning_rate": 5.958581169302191e-05, "loss": 0.0157, "step": 11243 }, { "epoch": 1.2189939288811795, "grad_norm": 0.8942390084266663, "learning_rate": 5.9582184825184976e-05, "loss": 0.0118, "step": 11244 }, { "epoch": 1.2191023417172593, "grad_norm": 0.24381829798221588, "learning_rate": 5.957855795734803e-05, "loss": 0.0299, "step": 11245 }, { "epoch": 1.219210754553339, "grad_norm": 0.43939992785453796, "learning_rate": 5.95749310895111e-05, "loss": 0.0462, "step": 11246 }, { "epoch": 1.2193191673894188, "grad_norm": 0.7201349139213562, "learning_rate": 5.957130422167416e-05, "loss": 0.0449, "step": 11247 }, { "epoch": 1.2194275802254988, "grad_norm": 0.2527998387813568, "learning_rate": 5.9567677353837225e-05, "loss": 0.0164, "step": 11248 }, { "epoch": 1.2195359930615786, "grad_norm": 0.1219208613038063, "learning_rate": 5.956405048600029e-05, "loss": 0.022, "step": 11249 }, { "epoch": 1.2196444058976583, "grad_norm": 0.28004053235054016, "learning_rate": 5.9560423618163353e-05, "loss": 0.036, "step": 11250 }, { "epoch": 1.219752818733738, "grad_norm": 0.7377181649208069, "learning_rate": 5.955679675032642e-05, "loss": 0.0425, "step": 11251 }, { "epoch": 1.2198612315698178, "grad_norm": 1.103243112564087, "learning_rate": 5.955316988248949e-05, "loss": 0.0988, "step": 11252 }, { "epoch": 1.2199696444058976, "grad_norm": 0.48059552907943726, "learning_rate": 5.954954301465255e-05, "loss": 0.0259, "step": 11253 }, { "epoch": 1.2200780572419774, "grad_norm": 0.3841199576854706, "learning_rate": 5.954591614681562e-05, "loss": 0.039, "step": 11254 }, { "epoch": 1.2201864700780574, "grad_norm": 0.6818476319313049, "learning_rate": 5.954228927897868e-05, "loss": 0.0858, "step": 11255 }, { "epoch": 1.2202948829141371, "grad_norm": 0.1280495524406433, "learning_rate": 5.9538662411141745e-05, "loss": 0.0095, "step": 11256 }, { "epoch": 1.2204032957502169, "grad_norm": 0.27377036213874817, "learning_rate": 5.953503554330481e-05, "loss": 0.0512, "step": 11257 }, { "epoch": 1.2205117085862967, "grad_norm": 0.29328981041908264, "learning_rate": 5.953140867546787e-05, "loss": 0.0272, "step": 11258 }, { "epoch": 1.2206201214223764, "grad_norm": 0.41005104780197144, "learning_rate": 5.952778180763093e-05, "loss": 0.0323, "step": 11259 }, { "epoch": 1.2207285342584562, "grad_norm": 0.5112090706825256, "learning_rate": 5.9524154939793994e-05, "loss": 0.0249, "step": 11260 }, { "epoch": 1.220836947094536, "grad_norm": 0.8896156549453735, "learning_rate": 5.952052807195706e-05, "loss": 0.0285, "step": 11261 }, { "epoch": 1.2209453599306157, "grad_norm": 0.2649916410446167, "learning_rate": 5.951690120412012e-05, "loss": 0.0564, "step": 11262 }, { "epoch": 1.2210537727666955, "grad_norm": 0.12457821518182755, "learning_rate": 5.951327433628319e-05, "loss": 0.0118, "step": 11263 }, { "epoch": 1.2211621856027755, "grad_norm": 0.23574702441692352, "learning_rate": 5.950964746844625e-05, "loss": 0.0348, "step": 11264 }, { "epoch": 1.2212705984388552, "grad_norm": 0.17473973333835602, "learning_rate": 5.9506020600609315e-05, "loss": 0.0061, "step": 11265 }, { "epoch": 1.221379011274935, "grad_norm": 1.0149551630020142, "learning_rate": 5.950239373277238e-05, "loss": 0.0694, "step": 11266 }, { "epoch": 1.2214874241110147, "grad_norm": 0.5092058181762695, "learning_rate": 5.949876686493544e-05, "loss": 0.0266, "step": 11267 }, { "epoch": 1.2215958369470945, "grad_norm": 0.3021254539489746, "learning_rate": 5.94951399970985e-05, "loss": 0.0367, "step": 11268 }, { "epoch": 1.2217042497831743, "grad_norm": 0.20642800629138947, "learning_rate": 5.949151312926158e-05, "loss": 0.0304, "step": 11269 }, { "epoch": 1.221812662619254, "grad_norm": 0.16031275689601898, "learning_rate": 5.948788626142464e-05, "loss": 0.0123, "step": 11270 }, { "epoch": 1.221921075455334, "grad_norm": 0.5682312250137329, "learning_rate": 5.9484259393587706e-05, "loss": 0.0406, "step": 11271 }, { "epoch": 1.2220294882914138, "grad_norm": 0.20228418707847595, "learning_rate": 5.948063252575077e-05, "loss": 0.0163, "step": 11272 }, { "epoch": 1.2221379011274935, "grad_norm": 0.16431492567062378, "learning_rate": 5.9477005657913835e-05, "loss": 0.0102, "step": 11273 }, { "epoch": 1.2222463139635733, "grad_norm": 0.19870567321777344, "learning_rate": 5.947337879007689e-05, "loss": 0.0176, "step": 11274 }, { "epoch": 1.222354726799653, "grad_norm": 0.1329108327627182, "learning_rate": 5.9469751922239956e-05, "loss": 0.024, "step": 11275 }, { "epoch": 1.2224631396357328, "grad_norm": 0.21765555441379547, "learning_rate": 5.946612505440302e-05, "loss": 0.0321, "step": 11276 }, { "epoch": 1.2225715524718126, "grad_norm": 0.21741628646850586, "learning_rate": 5.9462498186566084e-05, "loss": 0.038, "step": 11277 }, { "epoch": 1.2226799653078924, "grad_norm": 0.7680830955505371, "learning_rate": 5.945887131872915e-05, "loss": 0.04, "step": 11278 }, { "epoch": 1.2227883781439723, "grad_norm": 0.3782499134540558, "learning_rate": 5.945524445089221e-05, "loss": 0.0293, "step": 11279 }, { "epoch": 1.222896790980052, "grad_norm": 0.13649877905845642, "learning_rate": 5.9451617583055277e-05, "loss": 0.0098, "step": 11280 }, { "epoch": 1.2230052038161319, "grad_norm": 0.2091059684753418, "learning_rate": 5.944799071521834e-05, "loss": 0.0159, "step": 11281 }, { "epoch": 1.2231136166522116, "grad_norm": 0.03297530114650726, "learning_rate": 5.94443638473814e-05, "loss": 0.0032, "step": 11282 }, { "epoch": 1.2232220294882914, "grad_norm": 0.3342277705669403, "learning_rate": 5.944073697954446e-05, "loss": 0.0174, "step": 11283 }, { "epoch": 1.2233304423243712, "grad_norm": 0.12066476792097092, "learning_rate": 5.9437110111707526e-05, "loss": 0.0066, "step": 11284 }, { "epoch": 1.223438855160451, "grad_norm": 0.3929057717323303, "learning_rate": 5.943348324387059e-05, "loss": 0.0678, "step": 11285 }, { "epoch": 1.2235472679965307, "grad_norm": 0.3667801320552826, "learning_rate": 5.9429856376033654e-05, "loss": 0.0352, "step": 11286 }, { "epoch": 1.2236556808326107, "grad_norm": 0.6019608378410339, "learning_rate": 5.942622950819673e-05, "loss": 0.0626, "step": 11287 }, { "epoch": 1.2237640936686904, "grad_norm": 0.2995245158672333, "learning_rate": 5.942260264035979e-05, "loss": 0.0191, "step": 11288 }, { "epoch": 1.2238725065047702, "grad_norm": 0.31050974130630493, "learning_rate": 5.9418975772522854e-05, "loss": 0.0396, "step": 11289 }, { "epoch": 1.22398091934085, "grad_norm": 0.49990928173065186, "learning_rate": 5.941534890468592e-05, "loss": 0.0398, "step": 11290 }, { "epoch": 1.2240893321769297, "grad_norm": 0.04224356636404991, "learning_rate": 5.941172203684898e-05, "loss": 0.0035, "step": 11291 }, { "epoch": 1.2241977450130095, "grad_norm": 0.6762079000473022, "learning_rate": 5.9408095169012046e-05, "loss": 0.0576, "step": 11292 }, { "epoch": 1.2243061578490892, "grad_norm": 0.2008202075958252, "learning_rate": 5.940446830117511e-05, "loss": 0.015, "step": 11293 }, { "epoch": 1.2244145706851692, "grad_norm": 0.5518162846565247, "learning_rate": 5.9400841433338174e-05, "loss": 0.0177, "step": 11294 }, { "epoch": 1.224522983521249, "grad_norm": 0.07763753086328506, "learning_rate": 5.939721456550124e-05, "loss": 0.0122, "step": 11295 }, { "epoch": 1.2246313963573288, "grad_norm": 0.1313539445400238, "learning_rate": 5.93935876976643e-05, "loss": 0.0139, "step": 11296 }, { "epoch": 1.2247398091934085, "grad_norm": 0.10085040330886841, "learning_rate": 5.938996082982736e-05, "loss": 0.0065, "step": 11297 }, { "epoch": 1.2248482220294883, "grad_norm": 0.37450462579727173, "learning_rate": 5.9386333961990424e-05, "loss": 0.0385, "step": 11298 }, { "epoch": 1.224956634865568, "grad_norm": 0.4217115640640259, "learning_rate": 5.938270709415349e-05, "loss": 0.0475, "step": 11299 }, { "epoch": 1.2250650477016478, "grad_norm": 0.46069571375846863, "learning_rate": 5.937908022631655e-05, "loss": 0.0257, "step": 11300 }, { "epoch": 1.2251734605377276, "grad_norm": 0.5856353640556335, "learning_rate": 5.9375453358479616e-05, "loss": 0.0471, "step": 11301 }, { "epoch": 1.2252818733738073, "grad_norm": 0.6028599143028259, "learning_rate": 5.937182649064268e-05, "loss": 0.0117, "step": 11302 }, { "epoch": 1.2253902862098873, "grad_norm": 0.3202170431613922, "learning_rate": 5.9368199622805744e-05, "loss": 0.017, "step": 11303 }, { "epoch": 1.225498699045967, "grad_norm": 0.1338949203491211, "learning_rate": 5.9364572754968815e-05, "loss": 0.0197, "step": 11304 }, { "epoch": 1.2256071118820469, "grad_norm": 0.35587289929389954, "learning_rate": 5.936094588713188e-05, "loss": 0.0292, "step": 11305 }, { "epoch": 1.2257155247181266, "grad_norm": 0.04320923984050751, "learning_rate": 5.935731901929494e-05, "loss": 0.0031, "step": 11306 }, { "epoch": 1.2258239375542064, "grad_norm": 0.3585285544395447, "learning_rate": 5.935369215145801e-05, "loss": 0.0272, "step": 11307 }, { "epoch": 1.2259323503902861, "grad_norm": 0.3774971663951874, "learning_rate": 5.935006528362107e-05, "loss": 0.0482, "step": 11308 }, { "epoch": 1.226040763226366, "grad_norm": 0.22977112233638763, "learning_rate": 5.9346438415784136e-05, "loss": 0.0083, "step": 11309 }, { "epoch": 1.226149176062446, "grad_norm": 0.23516611754894257, "learning_rate": 5.93428115479472e-05, "loss": 0.0241, "step": 11310 }, { "epoch": 1.2262575888985257, "grad_norm": 0.33026552200317383, "learning_rate": 5.933918468011026e-05, "loss": 0.0244, "step": 11311 }, { "epoch": 1.2263660017346054, "grad_norm": 0.21096448600292206, "learning_rate": 5.933555781227332e-05, "loss": 0.0149, "step": 11312 }, { "epoch": 1.2264744145706852, "grad_norm": 0.35184797644615173, "learning_rate": 5.9331930944436385e-05, "loss": 0.0354, "step": 11313 }, { "epoch": 1.226582827406765, "grad_norm": 0.30040663480758667, "learning_rate": 5.932830407659945e-05, "loss": 0.0222, "step": 11314 }, { "epoch": 1.2266912402428447, "grad_norm": 0.13580653071403503, "learning_rate": 5.9324677208762513e-05, "loss": 0.0075, "step": 11315 }, { "epoch": 1.2267996530789245, "grad_norm": 0.1846986562013626, "learning_rate": 5.932105034092558e-05, "loss": 0.0254, "step": 11316 }, { "epoch": 1.2269080659150045, "grad_norm": 0.414218008518219, "learning_rate": 5.931742347308864e-05, "loss": 0.0122, "step": 11317 }, { "epoch": 1.2270164787510842, "grad_norm": 0.11087576299905777, "learning_rate": 5.9313796605251706e-05, "loss": 0.0084, "step": 11318 }, { "epoch": 1.227124891587164, "grad_norm": 0.2350432574748993, "learning_rate": 5.931016973741477e-05, "loss": 0.0257, "step": 11319 }, { "epoch": 1.2272333044232437, "grad_norm": 0.40808814764022827, "learning_rate": 5.930654286957783e-05, "loss": 0.0282, "step": 11320 }, { "epoch": 1.2273417172593235, "grad_norm": 0.06268011033535004, "learning_rate": 5.9302916001740905e-05, "loss": 0.0033, "step": 11321 }, { "epoch": 1.2274501300954033, "grad_norm": 0.49120262265205383, "learning_rate": 5.929928913390397e-05, "loss": 0.0243, "step": 11322 }, { "epoch": 1.227558542931483, "grad_norm": 0.4684141278266907, "learning_rate": 5.929566226606703e-05, "loss": 0.0611, "step": 11323 }, { "epoch": 1.2276669557675628, "grad_norm": 0.32162413001060486, "learning_rate": 5.92920353982301e-05, "loss": 0.0261, "step": 11324 }, { "epoch": 1.2277753686036426, "grad_norm": 0.11601165682077408, "learning_rate": 5.928840853039316e-05, "loss": 0.005, "step": 11325 }, { "epoch": 1.2278837814397225, "grad_norm": 0.17076489329338074, "learning_rate": 5.928478166255622e-05, "loss": 0.0108, "step": 11326 }, { "epoch": 1.2279921942758023, "grad_norm": 0.41491979360580444, "learning_rate": 5.928115479471928e-05, "loss": 0.0574, "step": 11327 }, { "epoch": 1.228100607111882, "grad_norm": 0.6069610714912415, "learning_rate": 5.927752792688235e-05, "loss": 0.1015, "step": 11328 }, { "epoch": 1.2282090199479618, "grad_norm": 0.3456461429595947, "learning_rate": 5.927390105904541e-05, "loss": 0.0213, "step": 11329 }, { "epoch": 1.2283174327840416, "grad_norm": 0.33462101221084595, "learning_rate": 5.9270274191208475e-05, "loss": 0.0098, "step": 11330 }, { "epoch": 1.2284258456201214, "grad_norm": 0.43447941541671753, "learning_rate": 5.926664732337154e-05, "loss": 0.0213, "step": 11331 }, { "epoch": 1.2285342584562011, "grad_norm": 0.11391758173704147, "learning_rate": 5.92630204555346e-05, "loss": 0.0126, "step": 11332 }, { "epoch": 1.228642671292281, "grad_norm": 0.15030908584594727, "learning_rate": 5.925939358769767e-05, "loss": 0.0157, "step": 11333 }, { "epoch": 1.2287510841283609, "grad_norm": 0.29544058442115784, "learning_rate": 5.9255766719860725e-05, "loss": 0.0274, "step": 11334 }, { "epoch": 1.2288594969644406, "grad_norm": 0.2049311399459839, "learning_rate": 5.925213985202379e-05, "loss": 0.0329, "step": 11335 }, { "epoch": 1.2289679098005204, "grad_norm": 0.39208242297172546, "learning_rate": 5.924851298418685e-05, "loss": 0.0645, "step": 11336 }, { "epoch": 1.2290763226366002, "grad_norm": 0.443327933549881, "learning_rate": 5.924488611634992e-05, "loss": 0.0223, "step": 11337 }, { "epoch": 1.22918473547268, "grad_norm": 0.17700399458408356, "learning_rate": 5.9241259248512995e-05, "loss": 0.0056, "step": 11338 }, { "epoch": 1.2292931483087597, "grad_norm": 0.490675687789917, "learning_rate": 5.923763238067606e-05, "loss": 0.026, "step": 11339 }, { "epoch": 1.2294015611448397, "grad_norm": 0.220392107963562, "learning_rate": 5.9234005512839116e-05, "loss": 0.0096, "step": 11340 }, { "epoch": 1.2295099739809194, "grad_norm": 0.1646120399236679, "learning_rate": 5.923037864500218e-05, "loss": 0.0069, "step": 11341 }, { "epoch": 1.2296183868169992, "grad_norm": 0.3401186764240265, "learning_rate": 5.9226751777165244e-05, "loss": 0.0468, "step": 11342 }, { "epoch": 1.229726799653079, "grad_norm": 0.28217723965644836, "learning_rate": 5.922312490932831e-05, "loss": 0.0198, "step": 11343 }, { "epoch": 1.2298352124891587, "grad_norm": 0.20411986112594604, "learning_rate": 5.921949804149137e-05, "loss": 0.0101, "step": 11344 }, { "epoch": 1.2299436253252385, "grad_norm": 0.1434335857629776, "learning_rate": 5.921587117365444e-05, "loss": 0.0087, "step": 11345 }, { "epoch": 1.2300520381613183, "grad_norm": 0.0664169117808342, "learning_rate": 5.92122443058175e-05, "loss": 0.004, "step": 11346 }, { "epoch": 1.230160450997398, "grad_norm": 0.42773833870887756, "learning_rate": 5.9208617437980565e-05, "loss": 0.0189, "step": 11347 }, { "epoch": 1.2302688638334778, "grad_norm": 0.1541762351989746, "learning_rate": 5.920499057014363e-05, "loss": 0.0131, "step": 11348 }, { "epoch": 1.2303772766695578, "grad_norm": 0.16661779582500458, "learning_rate": 5.9201363702306686e-05, "loss": 0.0149, "step": 11349 }, { "epoch": 1.2304856895056375, "grad_norm": 0.1572912633419037, "learning_rate": 5.919773683446975e-05, "loss": 0.0128, "step": 11350 }, { "epoch": 1.2305941023417173, "grad_norm": 0.08918804675340652, "learning_rate": 5.9194109966632814e-05, "loss": 0.0049, "step": 11351 }, { "epoch": 1.230702515177797, "grad_norm": 0.7059425115585327, "learning_rate": 5.919048309879588e-05, "loss": 0.0628, "step": 11352 }, { "epoch": 1.2308109280138768, "grad_norm": 0.47638261318206787, "learning_rate": 5.918685623095894e-05, "loss": 0.0436, "step": 11353 }, { "epoch": 1.2309193408499566, "grad_norm": 0.17290471494197845, "learning_rate": 5.918322936312201e-05, "loss": 0.0173, "step": 11354 }, { "epoch": 1.2310277536860363, "grad_norm": 0.6459303498268127, "learning_rate": 5.917960249528507e-05, "loss": 0.0449, "step": 11355 }, { "epoch": 1.2311361665221163, "grad_norm": 0.19735680520534515, "learning_rate": 5.917597562744814e-05, "loss": 0.0127, "step": 11356 }, { "epoch": 1.231244579358196, "grad_norm": 0.14320042729377747, "learning_rate": 5.9172348759611206e-05, "loss": 0.005, "step": 11357 }, { "epoch": 1.2313529921942759, "grad_norm": 0.2204110473394394, "learning_rate": 5.916872189177427e-05, "loss": 0.0238, "step": 11358 }, { "epoch": 1.2314614050303556, "grad_norm": 0.5787287950515747, "learning_rate": 5.9165095023937334e-05, "loss": 0.0357, "step": 11359 }, { "epoch": 1.2315698178664354, "grad_norm": 0.16344548761844635, "learning_rate": 5.91614681561004e-05, "loss": 0.021, "step": 11360 }, { "epoch": 1.2316782307025151, "grad_norm": 0.17505759000778198, "learning_rate": 5.915784128826346e-05, "loss": 0.0113, "step": 11361 }, { "epoch": 1.231786643538595, "grad_norm": 0.4204860031604767, "learning_rate": 5.9154214420426526e-05, "loss": 0.0331, "step": 11362 }, { "epoch": 1.2318950563746747, "grad_norm": 0.08667389303445816, "learning_rate": 5.9150587552589584e-05, "loss": 0.0081, "step": 11363 }, { "epoch": 1.2320034692107544, "grad_norm": 0.16230398416519165, "learning_rate": 5.914696068475265e-05, "loss": 0.0099, "step": 11364 }, { "epoch": 1.2321118820468344, "grad_norm": 0.12079015374183655, "learning_rate": 5.914333381691571e-05, "loss": 0.007, "step": 11365 }, { "epoch": 1.2322202948829142, "grad_norm": 0.06706757098436356, "learning_rate": 5.9139706949078776e-05, "loss": 0.0036, "step": 11366 }, { "epoch": 1.232328707718994, "grad_norm": 0.08443135768175125, "learning_rate": 5.913608008124184e-05, "loss": 0.006, "step": 11367 }, { "epoch": 1.2324371205550737, "grad_norm": 0.35531851649284363, "learning_rate": 5.9132453213404904e-05, "loss": 0.0431, "step": 11368 }, { "epoch": 1.2325455333911535, "grad_norm": 0.1472461223602295, "learning_rate": 5.912882634556797e-05, "loss": 0.0064, "step": 11369 }, { "epoch": 1.2326539462272332, "grad_norm": 0.1669243723154068, "learning_rate": 5.912519947773103e-05, "loss": 0.0068, "step": 11370 }, { "epoch": 1.232762359063313, "grad_norm": 0.19715748727321625, "learning_rate": 5.9121572609894097e-05, "loss": 0.0104, "step": 11371 }, { "epoch": 1.232870771899393, "grad_norm": 0.21075625717639923, "learning_rate": 5.9117945742057154e-05, "loss": 0.0255, "step": 11372 }, { "epoch": 1.2329791847354727, "grad_norm": 0.3917035460472107, "learning_rate": 5.911431887422023e-05, "loss": 0.0225, "step": 11373 }, { "epoch": 1.2330875975715525, "grad_norm": 0.06835677474737167, "learning_rate": 5.9110692006383296e-05, "loss": 0.0037, "step": 11374 }, { "epoch": 1.2331960104076323, "grad_norm": 0.6825211644172668, "learning_rate": 5.910706513854636e-05, "loss": 0.0491, "step": 11375 }, { "epoch": 1.233304423243712, "grad_norm": 0.07045560330152512, "learning_rate": 5.9103438270709424e-05, "loss": 0.0066, "step": 11376 }, { "epoch": 1.2334128360797918, "grad_norm": 0.08754471689462662, "learning_rate": 5.909981140287249e-05, "loss": 0.006, "step": 11377 }, { "epoch": 1.2335212489158716, "grad_norm": 0.2802305519580841, "learning_rate": 5.9096184535035545e-05, "loss": 0.0092, "step": 11378 }, { "epoch": 1.2336296617519515, "grad_norm": 0.5318554043769836, "learning_rate": 5.909255766719861e-05, "loss": 0.0262, "step": 11379 }, { "epoch": 1.2337380745880313, "grad_norm": 0.06653600931167603, "learning_rate": 5.9088930799361674e-05, "loss": 0.0032, "step": 11380 }, { "epoch": 1.233846487424111, "grad_norm": 0.24218381941318512, "learning_rate": 5.908530393152474e-05, "loss": 0.0116, "step": 11381 }, { "epoch": 1.2339549002601908, "grad_norm": 0.16062362492084503, "learning_rate": 5.90816770636878e-05, "loss": 0.0187, "step": 11382 }, { "epoch": 1.2340633130962706, "grad_norm": 0.5841581225395203, "learning_rate": 5.9078050195850866e-05, "loss": 0.0641, "step": 11383 }, { "epoch": 1.2341717259323504, "grad_norm": 0.1939353048801422, "learning_rate": 5.907442332801393e-05, "loss": 0.0141, "step": 11384 }, { "epoch": 1.2342801387684301, "grad_norm": 0.24773746728897095, "learning_rate": 5.9070796460176994e-05, "loss": 0.0407, "step": 11385 }, { "epoch": 1.2343885516045099, "grad_norm": 0.3626261055469513, "learning_rate": 5.906716959234005e-05, "loss": 0.0227, "step": 11386 }, { "epoch": 1.2344969644405897, "grad_norm": 0.18248991668224335, "learning_rate": 5.9063542724503116e-05, "loss": 0.0072, "step": 11387 }, { "epoch": 1.2346053772766696, "grad_norm": 0.555553674697876, "learning_rate": 5.905991585666618e-05, "loss": 0.044, "step": 11388 }, { "epoch": 1.2347137901127494, "grad_norm": 0.21443051099777222, "learning_rate": 5.9056288988829244e-05, "loss": 0.0185, "step": 11389 }, { "epoch": 1.2348222029488292, "grad_norm": 0.3670988082885742, "learning_rate": 5.905266212099232e-05, "loss": 0.02, "step": 11390 }, { "epoch": 1.234930615784909, "grad_norm": 0.29863080382347107, "learning_rate": 5.9049035253155385e-05, "loss": 0.0334, "step": 11391 }, { "epoch": 1.2350390286209887, "grad_norm": 0.2959967255592346, "learning_rate": 5.904540838531844e-05, "loss": 0.037, "step": 11392 }, { "epoch": 1.2351474414570685, "grad_norm": 0.47424599528312683, "learning_rate": 5.904178151748151e-05, "loss": 0.0255, "step": 11393 }, { "epoch": 1.2352558542931482, "grad_norm": 0.8092501163482666, "learning_rate": 5.903815464964457e-05, "loss": 0.0187, "step": 11394 }, { "epoch": 1.2353642671292282, "grad_norm": 0.24965816736221313, "learning_rate": 5.9034527781807635e-05, "loss": 0.0148, "step": 11395 }, { "epoch": 1.235472679965308, "grad_norm": 0.494728684425354, "learning_rate": 5.90309009139707e-05, "loss": 0.0308, "step": 11396 }, { "epoch": 1.2355810928013877, "grad_norm": 0.4232160747051239, "learning_rate": 5.902727404613376e-05, "loss": 0.018, "step": 11397 }, { "epoch": 1.2356895056374675, "grad_norm": 0.4235532879829407, "learning_rate": 5.902364717829683e-05, "loss": 0.023, "step": 11398 }, { "epoch": 1.2357979184735473, "grad_norm": 0.39878353476524353, "learning_rate": 5.902002031045989e-05, "loss": 0.0442, "step": 11399 }, { "epoch": 1.235906331309627, "grad_norm": 0.16221150755882263, "learning_rate": 5.9016393442622956e-05, "loss": 0.0156, "step": 11400 }, { "epoch": 1.2360147441457068, "grad_norm": 0.16587641835212708, "learning_rate": 5.901276657478601e-05, "loss": 0.0206, "step": 11401 }, { "epoch": 1.2361231569817868, "grad_norm": 0.49687516689300537, "learning_rate": 5.900913970694908e-05, "loss": 0.052, "step": 11402 }, { "epoch": 1.2362315698178665, "grad_norm": 0.3456786870956421, "learning_rate": 5.900551283911214e-05, "loss": 0.0257, "step": 11403 }, { "epoch": 1.2363399826539463, "grad_norm": 0.3847612142562866, "learning_rate": 5.9001885971275205e-05, "loss": 0.0276, "step": 11404 }, { "epoch": 1.236448395490026, "grad_norm": 0.21402636170387268, "learning_rate": 5.899825910343827e-05, "loss": 0.0199, "step": 11405 }, { "epoch": 1.2365568083261058, "grad_norm": 0.199342280626297, "learning_rate": 5.8994632235601334e-05, "loss": 0.0108, "step": 11406 }, { "epoch": 1.2366652211621856, "grad_norm": 0.22096392512321472, "learning_rate": 5.8991005367764404e-05, "loss": 0.0089, "step": 11407 }, { "epoch": 1.2367736339982653, "grad_norm": 0.3110961616039276, "learning_rate": 5.898737849992747e-05, "loss": 0.0286, "step": 11408 }, { "epoch": 1.236882046834345, "grad_norm": 0.34622523188591003, "learning_rate": 5.898375163209053e-05, "loss": 0.0296, "step": 11409 }, { "epoch": 1.2369904596704249, "grad_norm": 0.2526351511478424, "learning_rate": 5.89801247642536e-05, "loss": 0.0162, "step": 11410 }, { "epoch": 1.2370988725065049, "grad_norm": 0.09007839858531952, "learning_rate": 5.897649789641666e-05, "loss": 0.0064, "step": 11411 }, { "epoch": 1.2372072853425846, "grad_norm": 0.5535373091697693, "learning_rate": 5.8972871028579725e-05, "loss": 0.0369, "step": 11412 }, { "epoch": 1.2373156981786644, "grad_norm": 0.10478748381137848, "learning_rate": 5.896924416074279e-05, "loss": 0.0038, "step": 11413 }, { "epoch": 1.2374241110147441, "grad_norm": 0.17124956846237183, "learning_rate": 5.896561729290585e-05, "loss": 0.0205, "step": 11414 }, { "epoch": 1.237532523850824, "grad_norm": 0.24082285165786743, "learning_rate": 5.896199042506891e-05, "loss": 0.0277, "step": 11415 }, { "epoch": 1.2376409366869037, "grad_norm": 0.33440735936164856, "learning_rate": 5.8958363557231975e-05, "loss": 0.0349, "step": 11416 }, { "epoch": 1.2377493495229834, "grad_norm": 0.06361506134271622, "learning_rate": 5.895473668939504e-05, "loss": 0.0037, "step": 11417 }, { "epoch": 1.2378577623590634, "grad_norm": 0.08202707767486572, "learning_rate": 5.89511098215581e-05, "loss": 0.0083, "step": 11418 }, { "epoch": 1.2379661751951432, "grad_norm": 0.7193726301193237, "learning_rate": 5.894748295372117e-05, "loss": 0.0773, "step": 11419 }, { "epoch": 1.238074588031223, "grad_norm": 0.6527119874954224, "learning_rate": 5.894385608588423e-05, "loss": 0.0525, "step": 11420 }, { "epoch": 1.2381830008673027, "grad_norm": 0.3160642087459564, "learning_rate": 5.8940229218047295e-05, "loss": 0.0241, "step": 11421 }, { "epoch": 1.2382914137033825, "grad_norm": 0.24143017828464508, "learning_rate": 5.893660235021036e-05, "loss": 0.0463, "step": 11422 }, { "epoch": 1.2383998265394622, "grad_norm": 0.5430526733398438, "learning_rate": 5.893297548237342e-05, "loss": 0.0306, "step": 11423 }, { "epoch": 1.238508239375542, "grad_norm": 0.14297685027122498, "learning_rate": 5.892934861453648e-05, "loss": 0.0144, "step": 11424 }, { "epoch": 1.2386166522116218, "grad_norm": 0.3722165822982788, "learning_rate": 5.892572174669956e-05, "loss": 0.0543, "step": 11425 }, { "epoch": 1.2387250650477015, "grad_norm": 0.15695607662200928, "learning_rate": 5.892209487886262e-05, "loss": 0.0158, "step": 11426 }, { "epoch": 1.2388334778837815, "grad_norm": 0.1262882500886917, "learning_rate": 5.8918468011025686e-05, "loss": 0.0113, "step": 11427 }, { "epoch": 1.2389418907198613, "grad_norm": 1.4156732559204102, "learning_rate": 5.891484114318875e-05, "loss": 0.0419, "step": 11428 }, { "epoch": 1.239050303555941, "grad_norm": 0.5525896549224854, "learning_rate": 5.8911214275351815e-05, "loss": 0.0324, "step": 11429 }, { "epoch": 1.2391587163920208, "grad_norm": 0.12470486760139465, "learning_rate": 5.890758740751487e-05, "loss": 0.0031, "step": 11430 }, { "epoch": 1.2392671292281006, "grad_norm": 0.5147790908813477, "learning_rate": 5.8903960539677936e-05, "loss": 0.0251, "step": 11431 }, { "epoch": 1.2393755420641803, "grad_norm": 0.6442490816116333, "learning_rate": 5.8900333671841e-05, "loss": 0.0357, "step": 11432 }, { "epoch": 1.23948395490026, "grad_norm": 0.309363454580307, "learning_rate": 5.8896706804004064e-05, "loss": 0.0225, "step": 11433 }, { "epoch": 1.23959236773634, "grad_norm": 0.8378369212150574, "learning_rate": 5.889307993616713e-05, "loss": 0.0415, "step": 11434 }, { "epoch": 1.2397007805724198, "grad_norm": 0.23243962228298187, "learning_rate": 5.888945306833019e-05, "loss": 0.0172, "step": 11435 }, { "epoch": 1.2398091934084996, "grad_norm": 0.2763577699661255, "learning_rate": 5.888582620049326e-05, "loss": 0.017, "step": 11436 }, { "epoch": 1.2399176062445794, "grad_norm": 0.14338743686676025, "learning_rate": 5.888219933265632e-05, "loss": 0.0105, "step": 11437 }, { "epoch": 1.2400260190806591, "grad_norm": 0.10823959857225418, "learning_rate": 5.8878572464819385e-05, "loss": 0.0091, "step": 11438 }, { "epoch": 1.2401344319167389, "grad_norm": 0.0838814303278923, "learning_rate": 5.887494559698244e-05, "loss": 0.0047, "step": 11439 }, { "epoch": 1.2402428447528187, "grad_norm": 0.05504126474261284, "learning_rate": 5.8871318729145506e-05, "loss": 0.0085, "step": 11440 }, { "epoch": 1.2403512575888986, "grad_norm": 0.36343514919281006, "learning_rate": 5.886769186130857e-05, "loss": 0.0283, "step": 11441 }, { "epoch": 1.2404596704249784, "grad_norm": 0.08811721205711365, "learning_rate": 5.886406499347165e-05, "loss": 0.0073, "step": 11442 }, { "epoch": 1.2405680832610582, "grad_norm": 0.15309393405914307, "learning_rate": 5.886043812563471e-05, "loss": 0.0129, "step": 11443 }, { "epoch": 1.240676496097138, "grad_norm": 0.20676088333129883, "learning_rate": 5.885681125779777e-05, "loss": 0.0156, "step": 11444 }, { "epoch": 1.2407849089332177, "grad_norm": 0.06287895888090134, "learning_rate": 5.8853184389960834e-05, "loss": 0.0068, "step": 11445 }, { "epoch": 1.2408933217692975, "grad_norm": 0.5813807249069214, "learning_rate": 5.88495575221239e-05, "loss": 0.0586, "step": 11446 }, { "epoch": 1.2410017346053772, "grad_norm": 0.18316805362701416, "learning_rate": 5.884593065428696e-05, "loss": 0.0201, "step": 11447 }, { "epoch": 1.241110147441457, "grad_norm": 0.18698470294475555, "learning_rate": 5.8842303786450026e-05, "loss": 0.0122, "step": 11448 }, { "epoch": 1.2412185602775367, "grad_norm": 0.13700087368488312, "learning_rate": 5.883867691861309e-05, "loss": 0.013, "step": 11449 }, { "epoch": 1.2413269731136167, "grad_norm": 0.13742464780807495, "learning_rate": 5.8835050050776154e-05, "loss": 0.0061, "step": 11450 }, { "epoch": 1.2414353859496965, "grad_norm": 0.7181474566459656, "learning_rate": 5.883142318293922e-05, "loss": 0.055, "step": 11451 }, { "epoch": 1.2415437987857763, "grad_norm": 0.4368457496166229, "learning_rate": 5.882779631510228e-05, "loss": 0.0417, "step": 11452 }, { "epoch": 1.241652211621856, "grad_norm": 0.25215572118759155, "learning_rate": 5.882416944726534e-05, "loss": 0.0149, "step": 11453 }, { "epoch": 1.2417606244579358, "grad_norm": 0.3093404769897461, "learning_rate": 5.8820542579428404e-05, "loss": 0.0446, "step": 11454 }, { "epoch": 1.2418690372940155, "grad_norm": 0.14705821871757507, "learning_rate": 5.881691571159147e-05, "loss": 0.0184, "step": 11455 }, { "epoch": 1.2419774501300953, "grad_norm": 0.262520968914032, "learning_rate": 5.881328884375453e-05, "loss": 0.0172, "step": 11456 }, { "epoch": 1.2420858629661753, "grad_norm": 0.3845877945423126, "learning_rate": 5.8809661975917596e-05, "loss": 0.035, "step": 11457 }, { "epoch": 1.242194275802255, "grad_norm": 0.18159590661525726, "learning_rate": 5.880603510808066e-05, "loss": 0.0125, "step": 11458 }, { "epoch": 1.2423026886383348, "grad_norm": 0.4659113585948944, "learning_rate": 5.880240824024373e-05, "loss": 0.0756, "step": 11459 }, { "epoch": 1.2424111014744146, "grad_norm": 0.2889936566352844, "learning_rate": 5.8798781372406795e-05, "loss": 0.0139, "step": 11460 }, { "epoch": 1.2425195143104943, "grad_norm": 0.36299633979797363, "learning_rate": 5.879515450456986e-05, "loss": 0.0321, "step": 11461 }, { "epoch": 1.242627927146574, "grad_norm": 0.4192346930503845, "learning_rate": 5.879152763673292e-05, "loss": 0.0436, "step": 11462 }, { "epoch": 1.2427363399826539, "grad_norm": 0.22250406444072723, "learning_rate": 5.878790076889599e-05, "loss": 0.0132, "step": 11463 }, { "epoch": 1.2428447528187339, "grad_norm": 0.06502280384302139, "learning_rate": 5.878427390105905e-05, "loss": 0.0037, "step": 11464 }, { "epoch": 1.2429531656548136, "grad_norm": 0.04301612824201584, "learning_rate": 5.8780647033222116e-05, "loss": 0.0029, "step": 11465 }, { "epoch": 1.2430615784908934, "grad_norm": 1.0953896045684814, "learning_rate": 5.877702016538518e-05, "loss": 0.0641, "step": 11466 }, { "epoch": 1.2431699913269731, "grad_norm": 0.14623591303825378, "learning_rate": 5.877339329754824e-05, "loss": 0.0133, "step": 11467 }, { "epoch": 1.243278404163053, "grad_norm": 0.4767589271068573, "learning_rate": 5.87697664297113e-05, "loss": 0.0197, "step": 11468 }, { "epoch": 1.2433868169991327, "grad_norm": 0.08633450418710709, "learning_rate": 5.8766139561874365e-05, "loss": 0.0093, "step": 11469 }, { "epoch": 1.2434952298352124, "grad_norm": 0.08153004199266434, "learning_rate": 5.876251269403743e-05, "loss": 0.0068, "step": 11470 }, { "epoch": 1.2436036426712922, "grad_norm": 0.15691961348056793, "learning_rate": 5.8758885826200494e-05, "loss": 0.0121, "step": 11471 }, { "epoch": 1.243712055507372, "grad_norm": 0.32179611921310425, "learning_rate": 5.875525895836356e-05, "loss": 0.032, "step": 11472 }, { "epoch": 1.243820468343452, "grad_norm": 0.09129586070775986, "learning_rate": 5.875163209052662e-05, "loss": 0.005, "step": 11473 }, { "epoch": 1.2439288811795317, "grad_norm": 0.0798460841178894, "learning_rate": 5.8748005222689686e-05, "loss": 0.0086, "step": 11474 }, { "epoch": 1.2440372940156115, "grad_norm": 0.41946953535079956, "learning_rate": 5.874437835485275e-05, "loss": 0.0391, "step": 11475 }, { "epoch": 1.2441457068516912, "grad_norm": 0.5316038131713867, "learning_rate": 5.874075148701581e-05, "loss": 0.0183, "step": 11476 }, { "epoch": 1.244254119687771, "grad_norm": 0.7877255082130432, "learning_rate": 5.8737124619178885e-05, "loss": 0.034, "step": 11477 }, { "epoch": 1.2443625325238508, "grad_norm": 0.07671801745891571, "learning_rate": 5.873349775134195e-05, "loss": 0.0065, "step": 11478 }, { "epoch": 1.2444709453599305, "grad_norm": 0.48771098256111145, "learning_rate": 5.872987088350501e-05, "loss": 0.0325, "step": 11479 }, { "epoch": 1.2445793581960105, "grad_norm": 0.07683831453323364, "learning_rate": 5.872624401566808e-05, "loss": 0.006, "step": 11480 }, { "epoch": 1.2446877710320903, "grad_norm": 0.10583914071321487, "learning_rate": 5.872261714783114e-05, "loss": 0.007, "step": 11481 }, { "epoch": 1.24479618386817, "grad_norm": 0.2036663442850113, "learning_rate": 5.87189902799942e-05, "loss": 0.0152, "step": 11482 }, { "epoch": 1.2449045967042498, "grad_norm": 0.3378361463546753, "learning_rate": 5.871536341215726e-05, "loss": 0.0284, "step": 11483 }, { "epoch": 1.2450130095403296, "grad_norm": 0.20097282528877258, "learning_rate": 5.871173654432033e-05, "loss": 0.0153, "step": 11484 }, { "epoch": 1.2451214223764093, "grad_norm": 0.30140724778175354, "learning_rate": 5.870810967648339e-05, "loss": 0.0263, "step": 11485 }, { "epoch": 1.245229835212489, "grad_norm": 0.7245715856552124, "learning_rate": 5.8704482808646455e-05, "loss": 0.0486, "step": 11486 }, { "epoch": 1.245338248048569, "grad_norm": 0.22981758415699005, "learning_rate": 5.870085594080952e-05, "loss": 0.0173, "step": 11487 }, { "epoch": 1.2454466608846488, "grad_norm": 0.3072170913219452, "learning_rate": 5.869722907297258e-05, "loss": 0.0199, "step": 11488 }, { "epoch": 1.2455550737207286, "grad_norm": 0.15159064531326294, "learning_rate": 5.869360220513565e-05, "loss": 0.0228, "step": 11489 }, { "epoch": 1.2456634865568084, "grad_norm": 0.10424918681383133, "learning_rate": 5.868997533729871e-05, "loss": 0.0057, "step": 11490 }, { "epoch": 1.2457718993928881, "grad_norm": 0.5302345752716064, "learning_rate": 5.868634846946177e-05, "loss": 0.0328, "step": 11491 }, { "epoch": 1.245880312228968, "grad_norm": 0.3213103711605072, "learning_rate": 5.868272160162483e-05, "loss": 0.0133, "step": 11492 }, { "epoch": 1.2459887250650477, "grad_norm": 0.17373152077198029, "learning_rate": 5.86790947337879e-05, "loss": 0.0228, "step": 11493 }, { "epoch": 1.2460971379011274, "grad_norm": 0.2784512937068939, "learning_rate": 5.8675467865950975e-05, "loss": 0.0123, "step": 11494 }, { "epoch": 1.2462055507372072, "grad_norm": 0.31329604983329773, "learning_rate": 5.867184099811404e-05, "loss": 0.0561, "step": 11495 }, { "epoch": 1.2463139635732872, "grad_norm": 0.3004419803619385, "learning_rate": 5.8668214130277096e-05, "loss": 0.0228, "step": 11496 }, { "epoch": 1.246422376409367, "grad_norm": 0.6563018560409546, "learning_rate": 5.866458726244016e-05, "loss": 0.0587, "step": 11497 }, { "epoch": 1.2465307892454467, "grad_norm": 0.33192896842956543, "learning_rate": 5.8660960394603224e-05, "loss": 0.0335, "step": 11498 }, { "epoch": 1.2466392020815265, "grad_norm": 0.20748202502727509, "learning_rate": 5.865733352676629e-05, "loss": 0.0148, "step": 11499 }, { "epoch": 1.2467476149176062, "grad_norm": 0.1716662496328354, "learning_rate": 5.865370665892935e-05, "loss": 0.0164, "step": 11500 }, { "epoch": 1.246856027753686, "grad_norm": 0.023610569536685944, "learning_rate": 5.865007979109242e-05, "loss": 0.0013, "step": 11501 }, { "epoch": 1.2469644405897657, "grad_norm": 0.11357931792736053, "learning_rate": 5.864645292325548e-05, "loss": 0.0157, "step": 11502 }, { "epoch": 1.2470728534258457, "grad_norm": 0.3115115165710449, "learning_rate": 5.8642826055418545e-05, "loss": 0.0142, "step": 11503 }, { "epoch": 1.2471812662619255, "grad_norm": 0.05491800606250763, "learning_rate": 5.863919918758161e-05, "loss": 0.0035, "step": 11504 }, { "epoch": 1.2472896790980053, "grad_norm": 0.12984471023082733, "learning_rate": 5.8635572319744666e-05, "loss": 0.0061, "step": 11505 }, { "epoch": 1.247398091934085, "grad_norm": 0.4413219690322876, "learning_rate": 5.863194545190773e-05, "loss": 0.0272, "step": 11506 }, { "epoch": 1.2475065047701648, "grad_norm": 0.22732608020305634, "learning_rate": 5.8628318584070795e-05, "loss": 0.0151, "step": 11507 }, { "epoch": 1.2476149176062445, "grad_norm": 0.49953988194465637, "learning_rate": 5.862469171623386e-05, "loss": 0.0288, "step": 11508 }, { "epoch": 1.2477233304423243, "grad_norm": 0.0987284928560257, "learning_rate": 5.862106484839692e-05, "loss": 0.0053, "step": 11509 }, { "epoch": 1.247831743278404, "grad_norm": 0.3026294708251953, "learning_rate": 5.861743798055999e-05, "loss": 0.0212, "step": 11510 }, { "epoch": 1.2479401561144838, "grad_norm": 0.35013145208358765, "learning_rate": 5.861381111272306e-05, "loss": 0.0679, "step": 11511 }, { "epoch": 1.2480485689505638, "grad_norm": 0.6961472034454346, "learning_rate": 5.861018424488612e-05, "loss": 0.0273, "step": 11512 }, { "epoch": 1.2481569817866436, "grad_norm": 0.7405171394348145, "learning_rate": 5.8606557377049186e-05, "loss": 0.062, "step": 11513 }, { "epoch": 1.2482653946227233, "grad_norm": 0.14162606000900269, "learning_rate": 5.860293050921225e-05, "loss": 0.014, "step": 11514 }, { "epoch": 1.248373807458803, "grad_norm": 0.19724363088607788, "learning_rate": 5.8599303641375314e-05, "loss": 0.0144, "step": 11515 }, { "epoch": 1.2484822202948829, "grad_norm": 0.1343836486339569, "learning_rate": 5.859567677353838e-05, "loss": 0.0074, "step": 11516 }, { "epoch": 1.2485906331309626, "grad_norm": 0.15672403573989868, "learning_rate": 5.859204990570144e-05, "loss": 0.0084, "step": 11517 }, { "epoch": 1.2486990459670424, "grad_norm": 0.5154720544815063, "learning_rate": 5.8588423037864506e-05, "loss": 0.0513, "step": 11518 }, { "epoch": 1.2488074588031224, "grad_norm": 0.1478710025548935, "learning_rate": 5.8584796170027564e-05, "loss": 0.0226, "step": 11519 }, { "epoch": 1.2489158716392021, "grad_norm": 0.6075013279914856, "learning_rate": 5.858116930219063e-05, "loss": 0.0087, "step": 11520 }, { "epoch": 1.249024284475282, "grad_norm": 0.11665311455726624, "learning_rate": 5.857754243435369e-05, "loss": 0.0088, "step": 11521 }, { "epoch": 1.2491326973113617, "grad_norm": 0.26019591093063354, "learning_rate": 5.8573915566516756e-05, "loss": 0.016, "step": 11522 }, { "epoch": 1.2492411101474414, "grad_norm": 0.8599590063095093, "learning_rate": 5.857028869867982e-05, "loss": 0.0303, "step": 11523 }, { "epoch": 1.2493495229835212, "grad_norm": 0.2348434180021286, "learning_rate": 5.8566661830842884e-05, "loss": 0.0087, "step": 11524 }, { "epoch": 1.249457935819601, "grad_norm": 0.41620320081710815, "learning_rate": 5.856303496300595e-05, "loss": 0.0536, "step": 11525 }, { "epoch": 1.249566348655681, "grad_norm": 0.7128162980079651, "learning_rate": 5.855940809516901e-05, "loss": 0.0384, "step": 11526 }, { "epoch": 1.2496747614917607, "grad_norm": 0.17694289982318878, "learning_rate": 5.855578122733208e-05, "loss": 0.0074, "step": 11527 }, { "epoch": 1.2497831743278405, "grad_norm": 0.560206413269043, "learning_rate": 5.855215435949515e-05, "loss": 0.0892, "step": 11528 }, { "epoch": 1.2498915871639202, "grad_norm": 0.2764417231082916, "learning_rate": 5.854852749165821e-05, "loss": 0.0195, "step": 11529 }, { "epoch": 1.25, "grad_norm": 0.4441162049770355, "learning_rate": 5.8544900623821276e-05, "loss": 0.0363, "step": 11530 }, { "epoch": 1.2501084128360798, "grad_norm": 0.2008340209722519, "learning_rate": 5.854127375598434e-05, "loss": 0.0082, "step": 11531 }, { "epoch": 1.2502168256721595, "grad_norm": 0.3788054585456848, "learning_rate": 5.8537646888147404e-05, "loss": 0.0339, "step": 11532 }, { "epoch": 1.2503252385082395, "grad_norm": 0.09479548782110214, "learning_rate": 5.853402002031047e-05, "loss": 0.0092, "step": 11533 }, { "epoch": 1.250433651344319, "grad_norm": 0.6526015400886536, "learning_rate": 5.8530393152473525e-05, "loss": 0.0194, "step": 11534 }, { "epoch": 1.250542064180399, "grad_norm": 0.28913405537605286, "learning_rate": 5.852676628463659e-05, "loss": 0.0386, "step": 11535 }, { "epoch": 1.2506504770164788, "grad_norm": 0.4671001136302948, "learning_rate": 5.8523139416799654e-05, "loss": 0.0329, "step": 11536 }, { "epoch": 1.2507588898525586, "grad_norm": 0.2907642424106598, "learning_rate": 5.851951254896272e-05, "loss": 0.0253, "step": 11537 }, { "epoch": 1.2508673026886383, "grad_norm": 0.2208724170923233, "learning_rate": 5.851588568112578e-05, "loss": 0.0181, "step": 11538 }, { "epoch": 1.250975715524718, "grad_norm": 0.06520330905914307, "learning_rate": 5.8512258813288846e-05, "loss": 0.0075, "step": 11539 }, { "epoch": 1.2510841283607979, "grad_norm": 0.5477641820907593, "learning_rate": 5.850863194545191e-05, "loss": 0.0344, "step": 11540 }, { "epoch": 1.2511925411968776, "grad_norm": 0.2598611116409302, "learning_rate": 5.8505005077614974e-05, "loss": 0.0159, "step": 11541 }, { "epoch": 1.2513009540329576, "grad_norm": 0.2186604142189026, "learning_rate": 5.850137820977804e-05, "loss": 0.0127, "step": 11542 }, { "epoch": 1.2514093668690374, "grad_norm": 0.18361030519008636, "learning_rate": 5.8497751341941096e-05, "loss": 0.0081, "step": 11543 }, { "epoch": 1.2515177797051171, "grad_norm": 0.3751574158668518, "learning_rate": 5.849412447410416e-05, "loss": 0.0197, "step": 11544 }, { "epoch": 1.251626192541197, "grad_norm": 0.5794646143913269, "learning_rate": 5.8490497606267224e-05, "loss": 0.0782, "step": 11545 }, { "epoch": 1.2517346053772767, "grad_norm": 0.2875455617904663, "learning_rate": 5.84868707384303e-05, "loss": 0.0237, "step": 11546 }, { "epoch": 1.2518430182133564, "grad_norm": 0.7226880192756653, "learning_rate": 5.8483243870593366e-05, "loss": 0.0186, "step": 11547 }, { "epoch": 1.2519514310494362, "grad_norm": 0.6367464661598206, "learning_rate": 5.847961700275642e-05, "loss": 0.0204, "step": 11548 }, { "epoch": 1.2520598438855162, "grad_norm": 0.2673470675945282, "learning_rate": 5.847599013491949e-05, "loss": 0.0107, "step": 11549 }, { "epoch": 1.2521682567215957, "grad_norm": 0.2742372453212738, "learning_rate": 5.847236326708255e-05, "loss": 0.0228, "step": 11550 }, { "epoch": 1.2522766695576757, "grad_norm": 0.2039562165737152, "learning_rate": 5.8468736399245615e-05, "loss": 0.0049, "step": 11551 }, { "epoch": 1.2523850823937555, "grad_norm": 0.47571563720703125, "learning_rate": 5.846510953140868e-05, "loss": 0.0327, "step": 11552 }, { "epoch": 1.2524934952298352, "grad_norm": 0.20161345601081848, "learning_rate": 5.8461482663571743e-05, "loss": 0.0105, "step": 11553 }, { "epoch": 1.252601908065915, "grad_norm": 0.3078593909740448, "learning_rate": 5.845785579573481e-05, "loss": 0.007, "step": 11554 }, { "epoch": 1.2527103209019947, "grad_norm": 0.09849940985441208, "learning_rate": 5.845422892789787e-05, "loss": 0.0049, "step": 11555 }, { "epoch": 1.2528187337380745, "grad_norm": 0.7415028810501099, "learning_rate": 5.8450602060060936e-05, "loss": 0.02, "step": 11556 }, { "epoch": 1.2529271465741543, "grad_norm": 0.2550964653491974, "learning_rate": 5.844697519222399e-05, "loss": 0.0107, "step": 11557 }, { "epoch": 1.2530355594102343, "grad_norm": 0.41742077469825745, "learning_rate": 5.844334832438706e-05, "loss": 0.0291, "step": 11558 }, { "epoch": 1.253143972246314, "grad_norm": 0.10775723308324814, "learning_rate": 5.843972145655012e-05, "loss": 0.0058, "step": 11559 }, { "epoch": 1.2532523850823938, "grad_norm": 0.2998625934123993, "learning_rate": 5.8436094588713185e-05, "loss": 0.0381, "step": 11560 }, { "epoch": 1.2533607979184735, "grad_norm": 0.5141032338142395, "learning_rate": 5.843246772087625e-05, "loss": 0.0461, "step": 11561 }, { "epoch": 1.2534692107545533, "grad_norm": 0.4213206171989441, "learning_rate": 5.8428840853039314e-05, "loss": 0.0316, "step": 11562 }, { "epoch": 1.253577623590633, "grad_norm": 0.3040775954723358, "learning_rate": 5.8425213985202384e-05, "loss": 0.0229, "step": 11563 }, { "epoch": 1.2536860364267128, "grad_norm": 0.4605589807033539, "learning_rate": 5.842158711736545e-05, "loss": 0.0339, "step": 11564 }, { "epoch": 1.2537944492627928, "grad_norm": 0.7693799734115601, "learning_rate": 5.841796024952851e-05, "loss": 0.0154, "step": 11565 }, { "epoch": 1.2539028620988726, "grad_norm": 0.5512450337409973, "learning_rate": 5.841433338169158e-05, "loss": 0.0077, "step": 11566 }, { "epoch": 1.2540112749349523, "grad_norm": 0.290592223405838, "learning_rate": 5.841070651385464e-05, "loss": 0.0125, "step": 11567 }, { "epoch": 1.254119687771032, "grad_norm": 0.4266713559627533, "learning_rate": 5.8407079646017705e-05, "loss": 0.0397, "step": 11568 }, { "epoch": 1.2542281006071119, "grad_norm": 0.09895392507314682, "learning_rate": 5.840345277818077e-05, "loss": 0.0043, "step": 11569 }, { "epoch": 1.2543365134431916, "grad_norm": 0.6615363359451294, "learning_rate": 5.839982591034383e-05, "loss": 0.0467, "step": 11570 }, { "epoch": 1.2544449262792714, "grad_norm": 0.5724539756774902, "learning_rate": 5.839619904250689e-05, "loss": 0.0362, "step": 11571 }, { "epoch": 1.2545533391153514, "grad_norm": 0.23115460574626923, "learning_rate": 5.8392572174669955e-05, "loss": 0.0124, "step": 11572 }, { "epoch": 1.254661751951431, "grad_norm": 0.2514660656452179, "learning_rate": 5.838894530683302e-05, "loss": 0.0172, "step": 11573 }, { "epoch": 1.254770164787511, "grad_norm": 0.31875500082969666, "learning_rate": 5.838531843899608e-05, "loss": 0.0235, "step": 11574 }, { "epoch": 1.2548785776235907, "grad_norm": 0.5145153999328613, "learning_rate": 5.838169157115915e-05, "loss": 0.0255, "step": 11575 }, { "epoch": 1.2549869904596704, "grad_norm": 0.11989371478557587, "learning_rate": 5.837806470332221e-05, "loss": 0.0166, "step": 11576 }, { "epoch": 1.2550954032957502, "grad_norm": 0.4068520963191986, "learning_rate": 5.8374437835485275e-05, "loss": 0.041, "step": 11577 }, { "epoch": 1.25520381613183, "grad_norm": 0.20669540762901306, "learning_rate": 5.837081096764834e-05, "loss": 0.0149, "step": 11578 }, { "epoch": 1.2553122289679097, "grad_norm": 0.33050739765167236, "learning_rate": 5.83671840998114e-05, "loss": 0.0137, "step": 11579 }, { "epoch": 1.2554206418039895, "grad_norm": 0.18988540768623352, "learning_rate": 5.8363557231974474e-05, "loss": 0.0143, "step": 11580 }, { "epoch": 1.2555290546400695, "grad_norm": 0.27805784344673157, "learning_rate": 5.835993036413754e-05, "loss": 0.0305, "step": 11581 }, { "epoch": 1.2556374674761492, "grad_norm": 0.4104885160923004, "learning_rate": 5.83563034963006e-05, "loss": 0.0507, "step": 11582 }, { "epoch": 1.255745880312229, "grad_norm": 0.5297781229019165, "learning_rate": 5.8352676628463667e-05, "loss": 0.0199, "step": 11583 }, { "epoch": 1.2558542931483088, "grad_norm": 0.17216534912586212, "learning_rate": 5.834904976062673e-05, "loss": 0.0044, "step": 11584 }, { "epoch": 1.2559627059843885, "grad_norm": 0.2999384105205536, "learning_rate": 5.8345422892789795e-05, "loss": 0.0086, "step": 11585 }, { "epoch": 1.2560711188204683, "grad_norm": 0.29284194111824036, "learning_rate": 5.834179602495285e-05, "loss": 0.0321, "step": 11586 }, { "epoch": 1.256179531656548, "grad_norm": 0.13907840847969055, "learning_rate": 5.8338169157115916e-05, "loss": 0.0118, "step": 11587 }, { "epoch": 1.256287944492628, "grad_norm": 0.1752367466688156, "learning_rate": 5.833454228927898e-05, "loss": 0.0151, "step": 11588 }, { "epoch": 1.2563963573287076, "grad_norm": 0.09297743439674377, "learning_rate": 5.8330915421442044e-05, "loss": 0.0055, "step": 11589 }, { "epoch": 1.2565047701647876, "grad_norm": 0.2650332450866699, "learning_rate": 5.832728855360511e-05, "loss": 0.0121, "step": 11590 }, { "epoch": 1.2566131830008673, "grad_norm": 0.37532007694244385, "learning_rate": 5.832366168576817e-05, "loss": 0.0447, "step": 11591 }, { "epoch": 1.256721595836947, "grad_norm": 0.1995638608932495, "learning_rate": 5.832003481793124e-05, "loss": 0.0105, "step": 11592 }, { "epoch": 1.2568300086730269, "grad_norm": 1.2122727632522583, "learning_rate": 5.83164079500943e-05, "loss": 0.0413, "step": 11593 }, { "epoch": 1.2569384215091066, "grad_norm": 0.14694467186927795, "learning_rate": 5.8312781082257365e-05, "loss": 0.0154, "step": 11594 }, { "epoch": 1.2570468343451866, "grad_norm": 0.22358767688274384, "learning_rate": 5.830915421442042e-05, "loss": 0.018, "step": 11595 }, { "epoch": 1.2571552471812661, "grad_norm": 0.3530707061290741, "learning_rate": 5.8305527346583486e-05, "loss": 0.0236, "step": 11596 }, { "epoch": 1.2572636600173461, "grad_norm": 0.4920988082885742, "learning_rate": 5.8301900478746564e-05, "loss": 0.0547, "step": 11597 }, { "epoch": 1.257372072853426, "grad_norm": 0.2558691203594208, "learning_rate": 5.829827361090963e-05, "loss": 0.0225, "step": 11598 }, { "epoch": 1.2574804856895057, "grad_norm": 0.21695777773857117, "learning_rate": 5.829464674307269e-05, "loss": 0.0156, "step": 11599 }, { "epoch": 1.2575888985255854, "grad_norm": 0.15246568620204926, "learning_rate": 5.829101987523575e-05, "loss": 0.0065, "step": 11600 }, { "epoch": 1.2576973113616652, "grad_norm": 0.35431528091430664, "learning_rate": 5.8287393007398814e-05, "loss": 0.0297, "step": 11601 }, { "epoch": 1.257805724197745, "grad_norm": 0.1713719218969345, "learning_rate": 5.828376613956188e-05, "loss": 0.0154, "step": 11602 }, { "epoch": 1.2579141370338247, "grad_norm": 0.2650603652000427, "learning_rate": 5.828013927172494e-05, "loss": 0.034, "step": 11603 }, { "epoch": 1.2580225498699047, "grad_norm": 0.19134221971035004, "learning_rate": 5.8276512403888006e-05, "loss": 0.0075, "step": 11604 }, { "epoch": 1.2581309627059845, "grad_norm": 0.22054660320281982, "learning_rate": 5.827288553605107e-05, "loss": 0.0113, "step": 11605 }, { "epoch": 1.2582393755420642, "grad_norm": 0.8356608748435974, "learning_rate": 5.8269258668214134e-05, "loss": 0.0328, "step": 11606 }, { "epoch": 1.258347788378144, "grad_norm": 0.17902372777462006, "learning_rate": 5.82656318003772e-05, "loss": 0.0062, "step": 11607 }, { "epoch": 1.2584562012142237, "grad_norm": 0.27095115184783936, "learning_rate": 5.826200493254026e-05, "loss": 0.0448, "step": 11608 }, { "epoch": 1.2585646140503035, "grad_norm": 0.25239402055740356, "learning_rate": 5.825837806470332e-05, "loss": 0.0268, "step": 11609 }, { "epoch": 1.2586730268863833, "grad_norm": 0.8324661254882812, "learning_rate": 5.8254751196866384e-05, "loss": 0.02, "step": 11610 }, { "epoch": 1.2587814397224633, "grad_norm": 0.4243916869163513, "learning_rate": 5.825112432902945e-05, "loss": 0.0421, "step": 11611 }, { "epoch": 1.2588898525585428, "grad_norm": 0.2795465886592865, "learning_rate": 5.824749746119251e-05, "loss": 0.0339, "step": 11612 }, { "epoch": 1.2589982653946228, "grad_norm": 0.3295671343803406, "learning_rate": 5.8243870593355576e-05, "loss": 0.0231, "step": 11613 }, { "epoch": 1.2591066782307025, "grad_norm": 0.1992173194885254, "learning_rate": 5.824024372551864e-05, "loss": 0.0166, "step": 11614 }, { "epoch": 1.2592150910667823, "grad_norm": 0.2114483267068863, "learning_rate": 5.823661685768171e-05, "loss": 0.0278, "step": 11615 }, { "epoch": 1.259323503902862, "grad_norm": 0.0824337750673294, "learning_rate": 5.8232989989844775e-05, "loss": 0.0041, "step": 11616 }, { "epoch": 1.2594319167389418, "grad_norm": 0.2127380222082138, "learning_rate": 5.822936312200784e-05, "loss": 0.0135, "step": 11617 }, { "epoch": 1.2595403295750216, "grad_norm": 0.23452429473400116, "learning_rate": 5.8225736254170903e-05, "loss": 0.0294, "step": 11618 }, { "epoch": 1.2596487424111014, "grad_norm": 0.09762271493673325, "learning_rate": 5.822210938633397e-05, "loss": 0.0048, "step": 11619 }, { "epoch": 1.2597571552471813, "grad_norm": 0.07628893852233887, "learning_rate": 5.821848251849703e-05, "loss": 0.0083, "step": 11620 }, { "epoch": 1.2598655680832611, "grad_norm": 0.4433712661266327, "learning_rate": 5.8214855650660096e-05, "loss": 0.0177, "step": 11621 }, { "epoch": 1.2599739809193409, "grad_norm": 0.3538372218608856, "learning_rate": 5.821122878282316e-05, "loss": 0.0477, "step": 11622 }, { "epoch": 1.2600823937554206, "grad_norm": 0.3210068941116333, "learning_rate": 5.8207601914986224e-05, "loss": 0.0268, "step": 11623 }, { "epoch": 1.2601908065915004, "grad_norm": 0.5226926803588867, "learning_rate": 5.820397504714928e-05, "loss": 0.0319, "step": 11624 }, { "epoch": 1.2602992194275802, "grad_norm": 0.21538019180297852, "learning_rate": 5.8200348179312345e-05, "loss": 0.0258, "step": 11625 }, { "epoch": 1.26040763226366, "grad_norm": 0.2069346159696579, "learning_rate": 5.819672131147541e-05, "loss": 0.0136, "step": 11626 }, { "epoch": 1.26051604509974, "grad_norm": 0.41044241189956665, "learning_rate": 5.8193094443638474e-05, "loss": 0.0217, "step": 11627 }, { "epoch": 1.2606244579358197, "grad_norm": 0.18852713704109192, "learning_rate": 5.818946757580154e-05, "loss": 0.0084, "step": 11628 }, { "epoch": 1.2607328707718994, "grad_norm": 0.1080058142542839, "learning_rate": 5.81858407079646e-05, "loss": 0.0048, "step": 11629 }, { "epoch": 1.2608412836079792, "grad_norm": 0.3468638062477112, "learning_rate": 5.8182213840127666e-05, "loss": 0.036, "step": 11630 }, { "epoch": 1.260949696444059, "grad_norm": 0.43451622128486633, "learning_rate": 5.817858697229073e-05, "loss": 0.0306, "step": 11631 }, { "epoch": 1.2610581092801387, "grad_norm": 0.2587636411190033, "learning_rate": 5.81749601044538e-05, "loss": 0.0289, "step": 11632 }, { "epoch": 1.2611665221162185, "grad_norm": 0.2144612967967987, "learning_rate": 5.8171333236616865e-05, "loss": 0.0109, "step": 11633 }, { "epoch": 1.2612749349522985, "grad_norm": 0.06528203934431076, "learning_rate": 5.816770636877993e-05, "loss": 0.0027, "step": 11634 }, { "epoch": 1.261383347788378, "grad_norm": 0.5042871832847595, "learning_rate": 5.816407950094299e-05, "loss": 0.0119, "step": 11635 }, { "epoch": 1.261491760624458, "grad_norm": 0.3628268837928772, "learning_rate": 5.816045263310606e-05, "loss": 0.0119, "step": 11636 }, { "epoch": 1.2616001734605378, "grad_norm": 0.8504434823989868, "learning_rate": 5.815682576526912e-05, "loss": 0.0223, "step": 11637 }, { "epoch": 1.2617085862966175, "grad_norm": 0.4446970224380493, "learning_rate": 5.815319889743218e-05, "loss": 0.0837, "step": 11638 }, { "epoch": 1.2618169991326973, "grad_norm": 0.4919340908527374, "learning_rate": 5.814957202959524e-05, "loss": 0.0307, "step": 11639 }, { "epoch": 1.261925411968777, "grad_norm": 1.1944637298583984, "learning_rate": 5.814594516175831e-05, "loss": 0.0555, "step": 11640 }, { "epoch": 1.2620338248048568, "grad_norm": 0.20932650566101074, "learning_rate": 5.814231829392137e-05, "loss": 0.0159, "step": 11641 }, { "epoch": 1.2621422376409366, "grad_norm": 0.4355906844139099, "learning_rate": 5.8138691426084435e-05, "loss": 0.0322, "step": 11642 }, { "epoch": 1.2622506504770166, "grad_norm": 0.6274898648262024, "learning_rate": 5.81350645582475e-05, "loss": 0.0192, "step": 11643 }, { "epoch": 1.2623590633130963, "grad_norm": 0.4797005355358124, "learning_rate": 5.8131437690410563e-05, "loss": 0.0215, "step": 11644 }, { "epoch": 1.262467476149176, "grad_norm": 0.3526361286640167, "learning_rate": 5.812781082257363e-05, "loss": 0.0343, "step": 11645 }, { "epoch": 1.2625758889852559, "grad_norm": 0.5472549200057983, "learning_rate": 5.812418395473669e-05, "loss": 0.0409, "step": 11646 }, { "epoch": 1.2626843018213356, "grad_norm": 0.38439494371414185, "learning_rate": 5.812055708689975e-05, "loss": 0.0452, "step": 11647 }, { "epoch": 1.2627927146574154, "grad_norm": 0.33497145771980286, "learning_rate": 5.811693021906281e-05, "loss": 0.0184, "step": 11648 }, { "epoch": 1.2629011274934951, "grad_norm": 0.46861252188682556, "learning_rate": 5.811330335122589e-05, "loss": 0.0298, "step": 11649 }, { "epoch": 1.2630095403295751, "grad_norm": 0.1599072366952896, "learning_rate": 5.8109676483388955e-05, "loss": 0.0139, "step": 11650 }, { "epoch": 1.263117953165655, "grad_norm": 0.13514922559261322, "learning_rate": 5.810604961555202e-05, "loss": 0.007, "step": 11651 }, { "epoch": 1.2632263660017347, "grad_norm": 0.24646727740764618, "learning_rate": 5.8102422747715076e-05, "loss": 0.022, "step": 11652 }, { "epoch": 1.2633347788378144, "grad_norm": 0.627018928527832, "learning_rate": 5.809879587987814e-05, "loss": 0.0489, "step": 11653 }, { "epoch": 1.2634431916738942, "grad_norm": 1.1954349279403687, "learning_rate": 5.8095169012041204e-05, "loss": 0.025, "step": 11654 }, { "epoch": 1.263551604509974, "grad_norm": 0.14747728407382965, "learning_rate": 5.809154214420427e-05, "loss": 0.0062, "step": 11655 }, { "epoch": 1.2636600173460537, "grad_norm": 0.15636177361011505, "learning_rate": 5.808791527636733e-05, "loss": 0.012, "step": 11656 }, { "epoch": 1.2637684301821337, "grad_norm": 0.23430505394935608, "learning_rate": 5.80842884085304e-05, "loss": 0.0224, "step": 11657 }, { "epoch": 1.2638768430182132, "grad_norm": 0.18389113247394562, "learning_rate": 5.808066154069346e-05, "loss": 0.012, "step": 11658 }, { "epoch": 1.2639852558542932, "grad_norm": 0.7252869606018066, "learning_rate": 5.8077034672856525e-05, "loss": 0.0341, "step": 11659 }, { "epoch": 1.264093668690373, "grad_norm": 0.3083771765232086, "learning_rate": 5.807340780501959e-05, "loss": 0.0257, "step": 11660 }, { "epoch": 1.2642020815264527, "grad_norm": 0.18934696912765503, "learning_rate": 5.8069780937182646e-05, "loss": 0.0084, "step": 11661 }, { "epoch": 1.2643104943625325, "grad_norm": 0.062351975589990616, "learning_rate": 5.806615406934571e-05, "loss": 0.0048, "step": 11662 }, { "epoch": 1.2644189071986123, "grad_norm": 0.3917349874973297, "learning_rate": 5.8062527201508775e-05, "loss": 0.0385, "step": 11663 }, { "epoch": 1.264527320034692, "grad_norm": 0.35214102268218994, "learning_rate": 5.805890033367184e-05, "loss": 0.0217, "step": 11664 }, { "epoch": 1.2646357328707718, "grad_norm": 0.4875592589378357, "learning_rate": 5.80552734658349e-05, "loss": 0.0245, "step": 11665 }, { "epoch": 1.2647441457068518, "grad_norm": 0.11919708549976349, "learning_rate": 5.805164659799797e-05, "loss": 0.0168, "step": 11666 }, { "epoch": 1.2648525585429315, "grad_norm": 0.3542996346950531, "learning_rate": 5.804801973016104e-05, "loss": 0.0287, "step": 11667 }, { "epoch": 1.2649609713790113, "grad_norm": 0.5762345790863037, "learning_rate": 5.80443928623241e-05, "loss": 0.0285, "step": 11668 }, { "epoch": 1.265069384215091, "grad_norm": 0.11305519193410873, "learning_rate": 5.8040765994487166e-05, "loss": 0.0042, "step": 11669 }, { "epoch": 1.2651777970511708, "grad_norm": 0.46496713161468506, "learning_rate": 5.803713912665023e-05, "loss": 0.0238, "step": 11670 }, { "epoch": 1.2652862098872506, "grad_norm": 0.09097305685281754, "learning_rate": 5.8033512258813294e-05, "loss": 0.0067, "step": 11671 }, { "epoch": 1.2653946227233304, "grad_norm": 0.12486754357814789, "learning_rate": 5.802988539097636e-05, "loss": 0.0131, "step": 11672 }, { "epoch": 1.2655030355594103, "grad_norm": 0.18690867722034454, "learning_rate": 5.802625852313942e-05, "loss": 0.0125, "step": 11673 }, { "epoch": 1.26561144839549, "grad_norm": 0.5640289783477783, "learning_rate": 5.8022631655302487e-05, "loss": 0.0348, "step": 11674 }, { "epoch": 1.2657198612315699, "grad_norm": 0.6491755843162537, "learning_rate": 5.801900478746555e-05, "loss": 0.0282, "step": 11675 }, { "epoch": 1.2658282740676496, "grad_norm": 0.2594829499721527, "learning_rate": 5.801537791962861e-05, "loss": 0.0132, "step": 11676 }, { "epoch": 1.2659366869037294, "grad_norm": 0.11972285062074661, "learning_rate": 5.801175105179167e-05, "loss": 0.0123, "step": 11677 }, { "epoch": 1.2660450997398092, "grad_norm": 0.18345700204372406, "learning_rate": 5.8008124183954736e-05, "loss": 0.0176, "step": 11678 }, { "epoch": 1.266153512575889, "grad_norm": 0.2200394719839096, "learning_rate": 5.80044973161178e-05, "loss": 0.0058, "step": 11679 }, { "epoch": 1.266261925411969, "grad_norm": 0.2737250030040741, "learning_rate": 5.8000870448280864e-05, "loss": 0.021, "step": 11680 }, { "epoch": 1.2663703382480485, "grad_norm": 0.12237425148487091, "learning_rate": 5.799724358044393e-05, "loss": 0.0181, "step": 11681 }, { "epoch": 1.2664787510841284, "grad_norm": 0.2631582021713257, "learning_rate": 5.799361671260699e-05, "loss": 0.0302, "step": 11682 }, { "epoch": 1.2665871639202082, "grad_norm": 0.5453513264656067, "learning_rate": 5.798998984477006e-05, "loss": 0.0877, "step": 11683 }, { "epoch": 1.266695576756288, "grad_norm": 0.1204109787940979, "learning_rate": 5.798636297693313e-05, "loss": 0.0092, "step": 11684 }, { "epoch": 1.2668039895923677, "grad_norm": 0.4713704288005829, "learning_rate": 5.798273610909619e-05, "loss": 0.0287, "step": 11685 }, { "epoch": 1.2669124024284475, "grad_norm": 0.46923595666885376, "learning_rate": 5.7979109241259256e-05, "loss": 0.0183, "step": 11686 }, { "epoch": 1.2670208152645273, "grad_norm": 0.3605675995349884, "learning_rate": 5.797548237342232e-05, "loss": 0.019, "step": 11687 }, { "epoch": 1.267129228100607, "grad_norm": 0.18270845711231232, "learning_rate": 5.7971855505585384e-05, "loss": 0.0168, "step": 11688 }, { "epoch": 1.267237640936687, "grad_norm": 0.7698295712471008, "learning_rate": 5.796822863774845e-05, "loss": 0.0371, "step": 11689 }, { "epoch": 1.2673460537727668, "grad_norm": 0.38168954849243164, "learning_rate": 5.7964601769911505e-05, "loss": 0.0118, "step": 11690 }, { "epoch": 1.2674544666088465, "grad_norm": 0.2344413697719574, "learning_rate": 5.796097490207457e-05, "loss": 0.0292, "step": 11691 }, { "epoch": 1.2675628794449263, "grad_norm": 0.40336212515830994, "learning_rate": 5.7957348034237634e-05, "loss": 0.0915, "step": 11692 }, { "epoch": 1.267671292281006, "grad_norm": 0.10499556362628937, "learning_rate": 5.79537211664007e-05, "loss": 0.003, "step": 11693 }, { "epoch": 1.2677797051170858, "grad_norm": 1.1868832111358643, "learning_rate": 5.795009429856376e-05, "loss": 0.1232, "step": 11694 }, { "epoch": 1.2678881179531656, "grad_norm": 0.3691241443157196, "learning_rate": 5.7946467430726826e-05, "loss": 0.0248, "step": 11695 }, { "epoch": 1.2679965307892456, "grad_norm": 0.7434242963790894, "learning_rate": 5.794284056288989e-05, "loss": 0.0139, "step": 11696 }, { "epoch": 1.268104943625325, "grad_norm": 0.08157354593276978, "learning_rate": 5.7939213695052954e-05, "loss": 0.0037, "step": 11697 }, { "epoch": 1.268213356461405, "grad_norm": 0.3128770887851715, "learning_rate": 5.793558682721602e-05, "loss": 0.0229, "step": 11698 }, { "epoch": 1.2683217692974849, "grad_norm": 0.2798350155353546, "learning_rate": 5.7931959959379076e-05, "loss": 0.0174, "step": 11699 }, { "epoch": 1.2684301821335646, "grad_norm": 0.3444887399673462, "learning_rate": 5.792833309154214e-05, "loss": 0.0475, "step": 11700 }, { "epoch": 1.2685385949696444, "grad_norm": 0.5570987462997437, "learning_rate": 5.792470622370522e-05, "loss": 0.0349, "step": 11701 }, { "epoch": 1.2686470078057241, "grad_norm": 0.3163662552833557, "learning_rate": 5.792107935586828e-05, "loss": 0.027, "step": 11702 }, { "epoch": 1.268755420641804, "grad_norm": 0.18175388872623444, "learning_rate": 5.7917452488031346e-05, "loss": 0.0136, "step": 11703 }, { "epoch": 1.2688638334778837, "grad_norm": 0.6610853672027588, "learning_rate": 5.79138256201944e-05, "loss": 0.0757, "step": 11704 }, { "epoch": 1.2689722463139637, "grad_norm": 0.40550586581230164, "learning_rate": 5.791019875235747e-05, "loss": 0.0527, "step": 11705 }, { "epoch": 1.2690806591500434, "grad_norm": 1.117917537689209, "learning_rate": 5.790657188452053e-05, "loss": 0.0443, "step": 11706 }, { "epoch": 1.2691890719861232, "grad_norm": 0.1188923716545105, "learning_rate": 5.7902945016683595e-05, "loss": 0.0096, "step": 11707 }, { "epoch": 1.269297484822203, "grad_norm": 0.3315381705760956, "learning_rate": 5.789931814884666e-05, "loss": 0.0086, "step": 11708 }, { "epoch": 1.2694058976582827, "grad_norm": 0.5957384705543518, "learning_rate": 5.7895691281009723e-05, "loss": 0.0581, "step": 11709 }, { "epoch": 1.2695143104943625, "grad_norm": 0.12129585444927216, "learning_rate": 5.789206441317279e-05, "loss": 0.0074, "step": 11710 }, { "epoch": 1.2696227233304422, "grad_norm": 0.24436679482460022, "learning_rate": 5.788843754533585e-05, "loss": 0.0266, "step": 11711 }, { "epoch": 1.2697311361665222, "grad_norm": 0.2652361989021301, "learning_rate": 5.7884810677498916e-05, "loss": 0.0157, "step": 11712 }, { "epoch": 1.269839549002602, "grad_norm": 0.19002623856067657, "learning_rate": 5.788118380966197e-05, "loss": 0.0104, "step": 11713 }, { "epoch": 1.2699479618386817, "grad_norm": 0.2624953091144562, "learning_rate": 5.787755694182504e-05, "loss": 0.0272, "step": 11714 }, { "epoch": 1.2700563746747615, "grad_norm": 0.2693009078502655, "learning_rate": 5.78739300739881e-05, "loss": 0.0139, "step": 11715 }, { "epoch": 1.2701647875108413, "grad_norm": 0.3035244643688202, "learning_rate": 5.7870303206151165e-05, "loss": 0.0138, "step": 11716 }, { "epoch": 1.270273200346921, "grad_norm": 0.2591654360294342, "learning_rate": 5.786667633831423e-05, "loss": 0.0258, "step": 11717 }, { "epoch": 1.2703816131830008, "grad_norm": 0.514270544052124, "learning_rate": 5.786304947047731e-05, "loss": 0.0461, "step": 11718 }, { "epoch": 1.2704900260190808, "grad_norm": 0.1340692937374115, "learning_rate": 5.7859422602640365e-05, "loss": 0.0046, "step": 11719 }, { "epoch": 1.2705984388551603, "grad_norm": 0.09939900785684586, "learning_rate": 5.785579573480343e-05, "loss": 0.0043, "step": 11720 }, { "epoch": 1.2707068516912403, "grad_norm": 0.09641622751951218, "learning_rate": 5.785216886696649e-05, "loss": 0.0054, "step": 11721 }, { "epoch": 1.27081526452732, "grad_norm": 0.317691445350647, "learning_rate": 5.784854199912956e-05, "loss": 0.0408, "step": 11722 }, { "epoch": 1.2709236773633998, "grad_norm": 0.12758345901966095, "learning_rate": 5.784491513129262e-05, "loss": 0.0093, "step": 11723 }, { "epoch": 1.2710320901994796, "grad_norm": 0.2530488073825836, "learning_rate": 5.7841288263455685e-05, "loss": 0.025, "step": 11724 }, { "epoch": 1.2711405030355594, "grad_norm": 0.04100831598043442, "learning_rate": 5.783766139561875e-05, "loss": 0.0032, "step": 11725 }, { "epoch": 1.2712489158716391, "grad_norm": 0.1944955289363861, "learning_rate": 5.783403452778181e-05, "loss": 0.0144, "step": 11726 }, { "epoch": 1.271357328707719, "grad_norm": 0.19269847869873047, "learning_rate": 5.783040765994488e-05, "loss": 0.0195, "step": 11727 }, { "epoch": 1.2714657415437989, "grad_norm": 0.17086470127105713, "learning_rate": 5.7826780792107935e-05, "loss": 0.0197, "step": 11728 }, { "epoch": 1.2715741543798786, "grad_norm": 0.27097204327583313, "learning_rate": 5.7823153924271e-05, "loss": 0.032, "step": 11729 }, { "epoch": 1.2716825672159584, "grad_norm": 0.29632505774497986, "learning_rate": 5.781952705643406e-05, "loss": 0.0154, "step": 11730 }, { "epoch": 1.2717909800520382, "grad_norm": 0.17001497745513916, "learning_rate": 5.781590018859713e-05, "loss": 0.0075, "step": 11731 }, { "epoch": 1.271899392888118, "grad_norm": 0.15066352486610413, "learning_rate": 5.781227332076019e-05, "loss": 0.0093, "step": 11732 }, { "epoch": 1.2720078057241977, "grad_norm": 0.22282302379608154, "learning_rate": 5.7808646452923255e-05, "loss": 0.0215, "step": 11733 }, { "epoch": 1.2721162185602775, "grad_norm": 0.2437884658575058, "learning_rate": 5.780501958508632e-05, "loss": 0.0166, "step": 11734 }, { "epoch": 1.2722246313963574, "grad_norm": 0.2782266139984131, "learning_rate": 5.7801392717249383e-05, "loss": 0.0418, "step": 11735 }, { "epoch": 1.272333044232437, "grad_norm": 0.21775618195533752, "learning_rate": 5.7797765849412454e-05, "loss": 0.0298, "step": 11736 }, { "epoch": 1.272441457068517, "grad_norm": 0.12563590705394745, "learning_rate": 5.779413898157552e-05, "loss": 0.0078, "step": 11737 }, { "epoch": 1.2725498699045967, "grad_norm": 0.36643925309181213, "learning_rate": 5.779051211373858e-05, "loss": 0.0246, "step": 11738 }, { "epoch": 1.2726582827406765, "grad_norm": 0.11759119480848312, "learning_rate": 5.778688524590165e-05, "loss": 0.0063, "step": 11739 }, { "epoch": 1.2727666955767563, "grad_norm": 0.7006663084030151, "learning_rate": 5.778325837806471e-05, "loss": 0.0437, "step": 11740 }, { "epoch": 1.272875108412836, "grad_norm": 0.4980027973651886, "learning_rate": 5.7779631510227775e-05, "loss": 0.0547, "step": 11741 }, { "epoch": 1.272983521248916, "grad_norm": 0.43192920088768005, "learning_rate": 5.777600464239083e-05, "loss": 0.0563, "step": 11742 }, { "epoch": 1.2730919340849955, "grad_norm": 0.25401487946510315, "learning_rate": 5.7772377774553896e-05, "loss": 0.0192, "step": 11743 }, { "epoch": 1.2732003469210755, "grad_norm": 0.18658263981342316, "learning_rate": 5.776875090671696e-05, "loss": 0.0211, "step": 11744 }, { "epoch": 1.2733087597571553, "grad_norm": 0.28998807072639465, "learning_rate": 5.7765124038880024e-05, "loss": 0.0351, "step": 11745 }, { "epoch": 1.273417172593235, "grad_norm": 0.3284788131713867, "learning_rate": 5.776149717104309e-05, "loss": 0.0204, "step": 11746 }, { "epoch": 1.2735255854293148, "grad_norm": 0.16477812826633453, "learning_rate": 5.775787030320615e-05, "loss": 0.02, "step": 11747 }, { "epoch": 1.2736339982653946, "grad_norm": 0.33579587936401367, "learning_rate": 5.775424343536922e-05, "loss": 0.0189, "step": 11748 }, { "epoch": 1.2737424111014743, "grad_norm": 0.28819867968559265, "learning_rate": 5.775061656753228e-05, "loss": 0.0225, "step": 11749 }, { "epoch": 1.273850823937554, "grad_norm": 0.21262893080711365, "learning_rate": 5.7746989699695345e-05, "loss": 0.027, "step": 11750 }, { "epoch": 1.273959236773634, "grad_norm": 0.14762963354587555, "learning_rate": 5.77433628318584e-05, "loss": 0.0131, "step": 11751 }, { "epoch": 1.2740676496097139, "grad_norm": 0.3597761392593384, "learning_rate": 5.7739735964021466e-05, "loss": 0.0397, "step": 11752 }, { "epoch": 1.2741760624457936, "grad_norm": 0.48285534977912903, "learning_rate": 5.7736109096184544e-05, "loss": 0.0266, "step": 11753 }, { "epoch": 1.2742844752818734, "grad_norm": 0.2123810052871704, "learning_rate": 5.773248222834761e-05, "loss": 0.0124, "step": 11754 }, { "epoch": 1.2743928881179531, "grad_norm": 0.2932767868041992, "learning_rate": 5.772885536051067e-05, "loss": 0.0158, "step": 11755 }, { "epoch": 1.274501300954033, "grad_norm": 0.32319679856300354, "learning_rate": 5.772522849267373e-05, "loss": 0.0213, "step": 11756 }, { "epoch": 1.2746097137901127, "grad_norm": 0.33677244186401367, "learning_rate": 5.7721601624836794e-05, "loss": 0.0205, "step": 11757 }, { "epoch": 1.2747181266261927, "grad_norm": 0.5477257966995239, "learning_rate": 5.771797475699986e-05, "loss": 0.0546, "step": 11758 }, { "epoch": 1.2748265394622722, "grad_norm": 0.1204717829823494, "learning_rate": 5.771434788916292e-05, "loss": 0.0087, "step": 11759 }, { "epoch": 1.2749349522983522, "grad_norm": 0.22929321229457855, "learning_rate": 5.7710721021325986e-05, "loss": 0.0119, "step": 11760 }, { "epoch": 1.275043365134432, "grad_norm": 0.10052405297756195, "learning_rate": 5.770709415348905e-05, "loss": 0.0062, "step": 11761 }, { "epoch": 1.2751517779705117, "grad_norm": 0.38575083017349243, "learning_rate": 5.7703467285652114e-05, "loss": 0.0259, "step": 11762 }, { "epoch": 1.2752601908065915, "grad_norm": 0.10047821700572968, "learning_rate": 5.769984041781518e-05, "loss": 0.0076, "step": 11763 }, { "epoch": 1.2753686036426712, "grad_norm": 0.05607989802956581, "learning_rate": 5.769621354997824e-05, "loss": 0.002, "step": 11764 }, { "epoch": 1.275477016478751, "grad_norm": 0.3701300621032715, "learning_rate": 5.76925866821413e-05, "loss": 0.0617, "step": 11765 }, { "epoch": 1.2755854293148308, "grad_norm": 0.29144296050071716, "learning_rate": 5.7688959814304364e-05, "loss": 0.0266, "step": 11766 }, { "epoch": 1.2756938421509108, "grad_norm": 0.38304099440574646, "learning_rate": 5.768533294646743e-05, "loss": 0.0162, "step": 11767 }, { "epoch": 1.2758022549869905, "grad_norm": 0.22596301138401031, "learning_rate": 5.768170607863049e-05, "loss": 0.0169, "step": 11768 }, { "epoch": 1.2759106678230703, "grad_norm": 0.2174549549818039, "learning_rate": 5.7678079210793556e-05, "loss": 0.024, "step": 11769 }, { "epoch": 1.27601908065915, "grad_norm": 0.3340679407119751, "learning_rate": 5.7674452342956634e-05, "loss": 0.0228, "step": 11770 }, { "epoch": 1.2761274934952298, "grad_norm": 0.12176387757062912, "learning_rate": 5.767082547511969e-05, "loss": 0.0021, "step": 11771 }, { "epoch": 1.2762359063313096, "grad_norm": 0.5236982107162476, "learning_rate": 5.7667198607282755e-05, "loss": 0.0159, "step": 11772 }, { "epoch": 1.2763443191673893, "grad_norm": 0.1644858866930008, "learning_rate": 5.766357173944582e-05, "loss": 0.0058, "step": 11773 }, { "epoch": 1.2764527320034693, "grad_norm": 0.18233846127986908, "learning_rate": 5.7659944871608884e-05, "loss": 0.0104, "step": 11774 }, { "epoch": 1.276561144839549, "grad_norm": 0.5525027513504028, "learning_rate": 5.765631800377195e-05, "loss": 0.0565, "step": 11775 }, { "epoch": 1.2766695576756288, "grad_norm": 0.3438807725906372, "learning_rate": 5.765269113593501e-05, "loss": 0.0178, "step": 11776 }, { "epoch": 1.2767779705117086, "grad_norm": 0.7749746441841125, "learning_rate": 5.7649064268098076e-05, "loss": 0.0914, "step": 11777 }, { "epoch": 1.2768863833477884, "grad_norm": 0.19818547368049622, "learning_rate": 5.764543740026114e-05, "loss": 0.0156, "step": 11778 }, { "epoch": 1.2769947961838681, "grad_norm": 0.6669790148735046, "learning_rate": 5.7641810532424204e-05, "loss": 0.0277, "step": 11779 }, { "epoch": 1.277103209019948, "grad_norm": 0.2712370455265045, "learning_rate": 5.763818366458726e-05, "loss": 0.0268, "step": 11780 }, { "epoch": 1.2772116218560279, "grad_norm": 0.44381237030029297, "learning_rate": 5.7634556796750325e-05, "loss": 0.0447, "step": 11781 }, { "epoch": 1.2773200346921074, "grad_norm": 0.48689591884613037, "learning_rate": 5.763092992891339e-05, "loss": 0.027, "step": 11782 }, { "epoch": 1.2774284475281874, "grad_norm": 0.3926812708377838, "learning_rate": 5.7627303061076454e-05, "loss": 0.0212, "step": 11783 }, { "epoch": 1.2775368603642672, "grad_norm": 0.25303134322166443, "learning_rate": 5.762367619323952e-05, "loss": 0.0214, "step": 11784 }, { "epoch": 1.277645273200347, "grad_norm": 0.36538365483283997, "learning_rate": 5.762004932540258e-05, "loss": 0.0438, "step": 11785 }, { "epoch": 1.2777536860364267, "grad_norm": 0.5566697120666504, "learning_rate": 5.7616422457565646e-05, "loss": 0.0489, "step": 11786 }, { "epoch": 1.2778620988725065, "grad_norm": 0.16122329235076904, "learning_rate": 5.761279558972871e-05, "loss": 0.004, "step": 11787 }, { "epoch": 1.2779705117085862, "grad_norm": 0.3621709644794464, "learning_rate": 5.760916872189178e-05, "loss": 0.0232, "step": 11788 }, { "epoch": 1.278078924544666, "grad_norm": 0.2622627913951874, "learning_rate": 5.7605541854054845e-05, "loss": 0.0216, "step": 11789 }, { "epoch": 1.278187337380746, "grad_norm": 0.41357851028442383, "learning_rate": 5.760191498621791e-05, "loss": 0.0293, "step": 11790 }, { "epoch": 1.2782957502168257, "grad_norm": 0.07023289799690247, "learning_rate": 5.759828811838097e-05, "loss": 0.0074, "step": 11791 }, { "epoch": 1.2784041630529055, "grad_norm": 0.15250913798809052, "learning_rate": 5.759466125054404e-05, "loss": 0.013, "step": 11792 }, { "epoch": 1.2785125758889853, "grad_norm": 0.19244997203350067, "learning_rate": 5.75910343827071e-05, "loss": 0.0068, "step": 11793 }, { "epoch": 1.278620988725065, "grad_norm": 0.49955227971076965, "learning_rate": 5.758740751487016e-05, "loss": 0.0125, "step": 11794 }, { "epoch": 1.2787294015611448, "grad_norm": 0.7168476581573486, "learning_rate": 5.758378064703322e-05, "loss": 0.0496, "step": 11795 }, { "epoch": 1.2788378143972245, "grad_norm": 0.2812376916408539, "learning_rate": 5.758015377919629e-05, "loss": 0.0284, "step": 11796 }, { "epoch": 1.2789462272333045, "grad_norm": 0.15493017435073853, "learning_rate": 5.757652691135935e-05, "loss": 0.0104, "step": 11797 }, { "epoch": 1.2790546400693843, "grad_norm": 0.17439544200897217, "learning_rate": 5.7572900043522415e-05, "loss": 0.0154, "step": 11798 }, { "epoch": 1.279163052905464, "grad_norm": 0.2052180916070938, "learning_rate": 5.756927317568548e-05, "loss": 0.0055, "step": 11799 }, { "epoch": 1.2792714657415438, "grad_norm": 0.4143294394016266, "learning_rate": 5.7565646307848543e-05, "loss": 0.0301, "step": 11800 }, { "epoch": 1.2793798785776236, "grad_norm": 0.29947030544281006, "learning_rate": 5.756201944001161e-05, "loss": 0.0142, "step": 11801 }, { "epoch": 1.2794882914137033, "grad_norm": 0.1154596135020256, "learning_rate": 5.755839257217467e-05, "loss": 0.0083, "step": 11802 }, { "epoch": 1.2795967042497831, "grad_norm": 0.5396146774291992, "learning_rate": 5.755476570433773e-05, "loss": 0.0312, "step": 11803 }, { "epoch": 1.279705117085863, "grad_norm": 0.288608193397522, "learning_rate": 5.755113883650079e-05, "loss": 0.0183, "step": 11804 }, { "epoch": 1.2798135299219426, "grad_norm": 0.3020576536655426, "learning_rate": 5.754751196866387e-05, "loss": 0.0281, "step": 11805 }, { "epoch": 1.2799219427580226, "grad_norm": 0.31535011529922485, "learning_rate": 5.7543885100826935e-05, "loss": 0.0309, "step": 11806 }, { "epoch": 1.2800303555941024, "grad_norm": 0.4865269064903259, "learning_rate": 5.754025823299e-05, "loss": 0.0209, "step": 11807 }, { "epoch": 1.2801387684301822, "grad_norm": 1.1285433769226074, "learning_rate": 5.753663136515306e-05, "loss": 0.0456, "step": 11808 }, { "epoch": 1.280247181266262, "grad_norm": 0.05014348030090332, "learning_rate": 5.753300449731612e-05, "loss": 0.003, "step": 11809 }, { "epoch": 1.2803555941023417, "grad_norm": 0.2679036557674408, "learning_rate": 5.7529377629479185e-05, "loss": 0.0083, "step": 11810 }, { "epoch": 1.2804640069384214, "grad_norm": 0.6036918759346008, "learning_rate": 5.752575076164225e-05, "loss": 0.046, "step": 11811 }, { "epoch": 1.2805724197745012, "grad_norm": 0.22057175636291504, "learning_rate": 5.752212389380531e-05, "loss": 0.0127, "step": 11812 }, { "epoch": 1.2806808326105812, "grad_norm": 0.3193179965019226, "learning_rate": 5.751849702596838e-05, "loss": 0.0309, "step": 11813 }, { "epoch": 1.280789245446661, "grad_norm": 0.16072601079940796, "learning_rate": 5.751487015813144e-05, "loss": 0.0093, "step": 11814 }, { "epoch": 1.2808976582827407, "grad_norm": 0.05544392764568329, "learning_rate": 5.7511243290294505e-05, "loss": 0.0033, "step": 11815 }, { "epoch": 1.2810060711188205, "grad_norm": 0.2883981466293335, "learning_rate": 5.750761642245757e-05, "loss": 0.011, "step": 11816 }, { "epoch": 1.2811144839549002, "grad_norm": 0.5941926836967468, "learning_rate": 5.7503989554620626e-05, "loss": 0.0352, "step": 11817 }, { "epoch": 1.28122289679098, "grad_norm": 0.7230386137962341, "learning_rate": 5.750036268678369e-05, "loss": 0.0525, "step": 11818 }, { "epoch": 1.2813313096270598, "grad_norm": 1.7254552841186523, "learning_rate": 5.7496735818946755e-05, "loss": 0.0384, "step": 11819 }, { "epoch": 1.2814397224631398, "grad_norm": 0.5778511762619019, "learning_rate": 5.749310895110982e-05, "loss": 0.0075, "step": 11820 }, { "epoch": 1.2815481352992193, "grad_norm": 0.49297162890434265, "learning_rate": 5.748948208327288e-05, "loss": 0.0341, "step": 11821 }, { "epoch": 1.2816565481352993, "grad_norm": 0.3643771708011627, "learning_rate": 5.748585521543596e-05, "loss": 0.0403, "step": 11822 }, { "epoch": 1.281764960971379, "grad_norm": 0.02910493314266205, "learning_rate": 5.748222834759902e-05, "loss": 0.0017, "step": 11823 }, { "epoch": 1.2818733738074588, "grad_norm": 0.19407282769680023, "learning_rate": 5.747860147976208e-05, "loss": 0.0084, "step": 11824 }, { "epoch": 1.2819817866435386, "grad_norm": 0.3435899019241333, "learning_rate": 5.7474974611925146e-05, "loss": 0.0088, "step": 11825 }, { "epoch": 1.2820901994796183, "grad_norm": 1.0974936485290527, "learning_rate": 5.747134774408821e-05, "loss": 0.0427, "step": 11826 }, { "epoch": 1.2821986123156983, "grad_norm": 0.2656956613063812, "learning_rate": 5.7467720876251274e-05, "loss": 0.0094, "step": 11827 }, { "epoch": 1.2823070251517779, "grad_norm": 0.38500770926475525, "learning_rate": 5.746409400841434e-05, "loss": 0.0251, "step": 11828 }, { "epoch": 1.2824154379878578, "grad_norm": 0.8231038451194763, "learning_rate": 5.74604671405774e-05, "loss": 0.0693, "step": 11829 }, { "epoch": 1.2825238508239376, "grad_norm": 0.3206973671913147, "learning_rate": 5.745684027274047e-05, "loss": 0.0347, "step": 11830 }, { "epoch": 1.2826322636600174, "grad_norm": 0.3531029522418976, "learning_rate": 5.745321340490353e-05, "loss": 0.0257, "step": 11831 }, { "epoch": 1.2827406764960971, "grad_norm": 0.3649437725543976, "learning_rate": 5.744958653706659e-05, "loss": 0.0229, "step": 11832 }, { "epoch": 1.282849089332177, "grad_norm": 0.1949467808008194, "learning_rate": 5.744595966922965e-05, "loss": 0.0062, "step": 11833 }, { "epoch": 1.2829575021682567, "grad_norm": 0.347684383392334, "learning_rate": 5.7442332801392716e-05, "loss": 0.0144, "step": 11834 }, { "epoch": 1.2830659150043364, "grad_norm": 0.13588136434555054, "learning_rate": 5.743870593355578e-05, "loss": 0.0061, "step": 11835 }, { "epoch": 1.2831743278404164, "grad_norm": 0.34825316071510315, "learning_rate": 5.7435079065718844e-05, "loss": 0.0329, "step": 11836 }, { "epoch": 1.2832827406764962, "grad_norm": 0.09461154788732529, "learning_rate": 5.743145219788191e-05, "loss": 0.0072, "step": 11837 }, { "epoch": 1.283391153512576, "grad_norm": 0.361727774143219, "learning_rate": 5.742782533004497e-05, "loss": 0.0217, "step": 11838 }, { "epoch": 1.2834995663486557, "grad_norm": 0.08778093010187149, "learning_rate": 5.7424198462208044e-05, "loss": 0.0065, "step": 11839 }, { "epoch": 1.2836079791847355, "grad_norm": 0.15192218124866486, "learning_rate": 5.742057159437111e-05, "loss": 0.0106, "step": 11840 }, { "epoch": 1.2837163920208152, "grad_norm": 0.22710293531417847, "learning_rate": 5.741694472653417e-05, "loss": 0.021, "step": 11841 }, { "epoch": 1.283824804856895, "grad_norm": 0.29207053780555725, "learning_rate": 5.7413317858697236e-05, "loss": 0.0211, "step": 11842 }, { "epoch": 1.283933217692975, "grad_norm": 0.5017402768135071, "learning_rate": 5.74096909908603e-05, "loss": 0.0392, "step": 11843 }, { "epoch": 1.2840416305290545, "grad_norm": 0.27244246006011963, "learning_rate": 5.7406064123023364e-05, "loss": 0.027, "step": 11844 }, { "epoch": 1.2841500433651345, "grad_norm": 0.22044846415519714, "learning_rate": 5.740243725518643e-05, "loss": 0.0128, "step": 11845 }, { "epoch": 1.2842584562012143, "grad_norm": 0.29628267884254456, "learning_rate": 5.7398810387349486e-05, "loss": 0.0339, "step": 11846 }, { "epoch": 1.284366869037294, "grad_norm": 0.275361567735672, "learning_rate": 5.739518351951255e-05, "loss": 0.0162, "step": 11847 }, { "epoch": 1.2844752818733738, "grad_norm": 0.39267057180404663, "learning_rate": 5.7391556651675614e-05, "loss": 0.022, "step": 11848 }, { "epoch": 1.2845836947094535, "grad_norm": 0.4177986681461334, "learning_rate": 5.738792978383868e-05, "loss": 0.0553, "step": 11849 }, { "epoch": 1.2846921075455333, "grad_norm": 0.41282179951667786, "learning_rate": 5.738430291600174e-05, "loss": 0.0356, "step": 11850 }, { "epoch": 1.284800520381613, "grad_norm": 0.11931183189153671, "learning_rate": 5.7380676048164806e-05, "loss": 0.0092, "step": 11851 }, { "epoch": 1.284908933217693, "grad_norm": 0.2972579300403595, "learning_rate": 5.737704918032787e-05, "loss": 0.0148, "step": 11852 }, { "epoch": 1.2850173460537728, "grad_norm": 0.3277006149291992, "learning_rate": 5.7373422312490934e-05, "loss": 0.0354, "step": 11853 }, { "epoch": 1.2851257588898526, "grad_norm": 0.42385217547416687, "learning_rate": 5.7369795444654e-05, "loss": 0.0239, "step": 11854 }, { "epoch": 1.2852341717259324, "grad_norm": 0.40025901794433594, "learning_rate": 5.7366168576817056e-05, "loss": 0.0255, "step": 11855 }, { "epoch": 1.2853425845620121, "grad_norm": 0.29293301701545715, "learning_rate": 5.736254170898012e-05, "loss": 0.0227, "step": 11856 }, { "epoch": 1.2854509973980919, "grad_norm": 0.32140257954597473, "learning_rate": 5.73589148411432e-05, "loss": 0.0296, "step": 11857 }, { "epoch": 1.2855594102341716, "grad_norm": 0.16811789572238922, "learning_rate": 5.735528797330626e-05, "loss": 0.0104, "step": 11858 }, { "epoch": 1.2856678230702516, "grad_norm": 0.15022623538970947, "learning_rate": 5.7351661105469326e-05, "loss": 0.0084, "step": 11859 }, { "epoch": 1.2857762359063314, "grad_norm": 0.15269625186920166, "learning_rate": 5.734803423763239e-05, "loss": 0.0146, "step": 11860 }, { "epoch": 1.2858846487424112, "grad_norm": 0.5768656134605408, "learning_rate": 5.734440736979545e-05, "loss": 0.014, "step": 11861 }, { "epoch": 1.285993061578491, "grad_norm": 0.3388875722885132, "learning_rate": 5.734078050195851e-05, "loss": 0.0307, "step": 11862 }, { "epoch": 1.2861014744145707, "grad_norm": 0.22226093709468842, "learning_rate": 5.7337153634121575e-05, "loss": 0.0203, "step": 11863 }, { "epoch": 1.2862098872506504, "grad_norm": 0.15825429558753967, "learning_rate": 5.733352676628464e-05, "loss": 0.0082, "step": 11864 }, { "epoch": 1.2863183000867302, "grad_norm": 0.16465306282043457, "learning_rate": 5.7329899898447704e-05, "loss": 0.014, "step": 11865 }, { "epoch": 1.2864267129228102, "grad_norm": 0.3455354571342468, "learning_rate": 5.732627303061077e-05, "loss": 0.0288, "step": 11866 }, { "epoch": 1.2865351257588897, "grad_norm": 0.07275955379009247, "learning_rate": 5.732264616277383e-05, "loss": 0.0046, "step": 11867 }, { "epoch": 1.2866435385949697, "grad_norm": 0.16291102766990662, "learning_rate": 5.7319019294936896e-05, "loss": 0.0161, "step": 11868 }, { "epoch": 1.2867519514310495, "grad_norm": 0.4102059304714203, "learning_rate": 5.731539242709995e-05, "loss": 0.0239, "step": 11869 }, { "epoch": 1.2868603642671292, "grad_norm": 0.8197019100189209, "learning_rate": 5.731176555926302e-05, "loss": 0.0307, "step": 11870 }, { "epoch": 1.286968777103209, "grad_norm": 0.05300518497824669, "learning_rate": 5.730813869142608e-05, "loss": 0.0033, "step": 11871 }, { "epoch": 1.2870771899392888, "grad_norm": 0.3615443706512451, "learning_rate": 5.7304511823589145e-05, "loss": 0.0282, "step": 11872 }, { "epoch": 1.2871856027753685, "grad_norm": 0.12738633155822754, "learning_rate": 5.730088495575221e-05, "loss": 0.011, "step": 11873 }, { "epoch": 1.2872940156114483, "grad_norm": 0.10769975930452347, "learning_rate": 5.729725808791529e-05, "loss": 0.0114, "step": 11874 }, { "epoch": 1.2874024284475283, "grad_norm": 0.6060208678245544, "learning_rate": 5.7293631220078345e-05, "loss": 0.0272, "step": 11875 }, { "epoch": 1.287510841283608, "grad_norm": 0.23636610805988312, "learning_rate": 5.729000435224141e-05, "loss": 0.0294, "step": 11876 }, { "epoch": 1.2876192541196878, "grad_norm": 0.3553774654865265, "learning_rate": 5.728637748440447e-05, "loss": 0.0124, "step": 11877 }, { "epoch": 1.2877276669557676, "grad_norm": 0.28346729278564453, "learning_rate": 5.728275061656754e-05, "loss": 0.0268, "step": 11878 }, { "epoch": 1.2878360797918473, "grad_norm": 0.22697851061820984, "learning_rate": 5.72791237487306e-05, "loss": 0.028, "step": 11879 }, { "epoch": 1.287944492627927, "grad_norm": 0.34361201524734497, "learning_rate": 5.7275496880893665e-05, "loss": 0.0142, "step": 11880 }, { "epoch": 1.2880529054640069, "grad_norm": 0.20303180813789368, "learning_rate": 5.727187001305673e-05, "loss": 0.0203, "step": 11881 }, { "epoch": 1.2881613183000868, "grad_norm": 0.6233149170875549, "learning_rate": 5.726824314521979e-05, "loss": 0.0418, "step": 11882 }, { "epoch": 1.2882697311361664, "grad_norm": 0.1291319876909256, "learning_rate": 5.726461627738286e-05, "loss": 0.007, "step": 11883 }, { "epoch": 1.2883781439722464, "grad_norm": 0.2561485469341278, "learning_rate": 5.7260989409545915e-05, "loss": 0.0118, "step": 11884 }, { "epoch": 1.2884865568083261, "grad_norm": 0.7793791890144348, "learning_rate": 5.725736254170898e-05, "loss": 0.0501, "step": 11885 }, { "epoch": 1.288594969644406, "grad_norm": 0.19100001454353333, "learning_rate": 5.725373567387204e-05, "loss": 0.0134, "step": 11886 }, { "epoch": 1.2887033824804857, "grad_norm": 0.0654977336525917, "learning_rate": 5.725010880603511e-05, "loss": 0.0052, "step": 11887 }, { "epoch": 1.2888117953165654, "grad_norm": 0.4612754285335541, "learning_rate": 5.724648193819817e-05, "loss": 0.0311, "step": 11888 }, { "epoch": 1.2889202081526454, "grad_norm": 0.2447577863931656, "learning_rate": 5.7242855070361235e-05, "loss": 0.0147, "step": 11889 }, { "epoch": 1.289028620988725, "grad_norm": 0.5860811471939087, "learning_rate": 5.72392282025243e-05, "loss": 0.0373, "step": 11890 }, { "epoch": 1.289137033824805, "grad_norm": 0.26383912563323975, "learning_rate": 5.723560133468737e-05, "loss": 0.0119, "step": 11891 }, { "epoch": 1.2892454466608847, "grad_norm": 0.09371132403612137, "learning_rate": 5.7231974466850434e-05, "loss": 0.0032, "step": 11892 }, { "epoch": 1.2893538594969645, "grad_norm": 0.3033793568611145, "learning_rate": 5.72283475990135e-05, "loss": 0.0467, "step": 11893 }, { "epoch": 1.2894622723330442, "grad_norm": 0.14760085940361023, "learning_rate": 5.722472073117656e-05, "loss": 0.0075, "step": 11894 }, { "epoch": 1.289570685169124, "grad_norm": 0.5642402768135071, "learning_rate": 5.722109386333963e-05, "loss": 0.0635, "step": 11895 }, { "epoch": 1.2896790980052038, "grad_norm": 0.6493581533432007, "learning_rate": 5.721746699550269e-05, "loss": 0.0572, "step": 11896 }, { "epoch": 1.2897875108412835, "grad_norm": 0.39678990840911865, "learning_rate": 5.7213840127665755e-05, "loss": 0.0214, "step": 11897 }, { "epoch": 1.2898959236773635, "grad_norm": 1.3595960140228271, "learning_rate": 5.721021325982881e-05, "loss": 0.0615, "step": 11898 }, { "epoch": 1.2900043365134433, "grad_norm": 0.13673660159111023, "learning_rate": 5.7206586391991876e-05, "loss": 0.0088, "step": 11899 }, { "epoch": 1.290112749349523, "grad_norm": 0.23633654415607452, "learning_rate": 5.720295952415494e-05, "loss": 0.0225, "step": 11900 }, { "epoch": 1.2902211621856028, "grad_norm": 0.2708631455898285, "learning_rate": 5.7199332656318005e-05, "loss": 0.0075, "step": 11901 }, { "epoch": 1.2903295750216826, "grad_norm": 0.3050884008407593, "learning_rate": 5.719570578848107e-05, "loss": 0.0341, "step": 11902 }, { "epoch": 1.2904379878577623, "grad_norm": 0.4548328220844269, "learning_rate": 5.719207892064413e-05, "loss": 0.0413, "step": 11903 }, { "epoch": 1.290546400693842, "grad_norm": 0.1584402173757553, "learning_rate": 5.71884520528072e-05, "loss": 0.0089, "step": 11904 }, { "epoch": 1.290654813529922, "grad_norm": 0.35610079765319824, "learning_rate": 5.718482518497026e-05, "loss": 0.0768, "step": 11905 }, { "epoch": 1.2907632263660016, "grad_norm": 1.0450228452682495, "learning_rate": 5.7181198317133325e-05, "loss": 0.0643, "step": 11906 }, { "epoch": 1.2908716392020816, "grad_norm": 0.38594964146614075, "learning_rate": 5.717757144929638e-05, "loss": 0.0187, "step": 11907 }, { "epoch": 1.2909800520381614, "grad_norm": 0.30406469106674194, "learning_rate": 5.717394458145946e-05, "loss": 0.0257, "step": 11908 }, { "epoch": 1.2910884648742411, "grad_norm": 0.42860502004623413, "learning_rate": 5.7170317713622524e-05, "loss": 0.0384, "step": 11909 }, { "epoch": 1.2911968777103209, "grad_norm": 0.3685663938522339, "learning_rate": 5.716669084578559e-05, "loss": 0.0094, "step": 11910 }, { "epoch": 1.2913052905464006, "grad_norm": 0.3578375279903412, "learning_rate": 5.716306397794865e-05, "loss": 0.0269, "step": 11911 }, { "epoch": 1.2914137033824804, "grad_norm": 0.4625907838344574, "learning_rate": 5.7159437110111716e-05, "loss": 0.0393, "step": 11912 }, { "epoch": 1.2915221162185602, "grad_norm": 0.4099997580051422, "learning_rate": 5.7155810242274774e-05, "loss": 0.0245, "step": 11913 }, { "epoch": 1.2916305290546402, "grad_norm": 0.08361039310693741, "learning_rate": 5.715218337443784e-05, "loss": 0.0051, "step": 11914 }, { "epoch": 1.29173894189072, "grad_norm": 0.013264480978250504, "learning_rate": 5.71485565066009e-05, "loss": 0.0008, "step": 11915 }, { "epoch": 1.2918473547267997, "grad_norm": 0.23708292841911316, "learning_rate": 5.7144929638763966e-05, "loss": 0.0155, "step": 11916 }, { "epoch": 1.2919557675628794, "grad_norm": 0.5143178701400757, "learning_rate": 5.714130277092703e-05, "loss": 0.029, "step": 11917 }, { "epoch": 1.2920641803989592, "grad_norm": 0.09652681648731232, "learning_rate": 5.7137675903090094e-05, "loss": 0.012, "step": 11918 }, { "epoch": 1.292172593235039, "grad_norm": 0.298538476228714, "learning_rate": 5.713404903525316e-05, "loss": 0.0221, "step": 11919 }, { "epoch": 1.2922810060711187, "grad_norm": 0.6648491621017456, "learning_rate": 5.713042216741622e-05, "loss": 0.0307, "step": 11920 }, { "epoch": 1.2923894189071987, "grad_norm": 0.6247923970222473, "learning_rate": 5.712679529957928e-05, "loss": 0.0299, "step": 11921 }, { "epoch": 1.2924978317432785, "grad_norm": 0.10758790373802185, "learning_rate": 5.7123168431742344e-05, "loss": 0.0068, "step": 11922 }, { "epoch": 1.2926062445793582, "grad_norm": 0.24463428556919098, "learning_rate": 5.711954156390541e-05, "loss": 0.0162, "step": 11923 }, { "epoch": 1.292714657415438, "grad_norm": 0.22641071677207947, "learning_rate": 5.711591469606847e-05, "loss": 0.0178, "step": 11924 }, { "epoch": 1.2928230702515178, "grad_norm": 0.6955614686012268, "learning_rate": 5.7112287828231536e-05, "loss": 0.0561, "step": 11925 }, { "epoch": 1.2929314830875975, "grad_norm": 0.20636512339115143, "learning_rate": 5.7108660960394614e-05, "loss": 0.0193, "step": 11926 }, { "epoch": 1.2930398959236773, "grad_norm": 0.10897663235664368, "learning_rate": 5.710503409255767e-05, "loss": 0.0065, "step": 11927 }, { "epoch": 1.2931483087597573, "grad_norm": 0.6046158075332642, "learning_rate": 5.7101407224720735e-05, "loss": 0.0283, "step": 11928 }, { "epoch": 1.2932567215958368, "grad_norm": 0.2239437848329544, "learning_rate": 5.70977803568838e-05, "loss": 0.0115, "step": 11929 }, { "epoch": 1.2933651344319168, "grad_norm": 0.10369594395160675, "learning_rate": 5.7094153489046864e-05, "loss": 0.0052, "step": 11930 }, { "epoch": 1.2934735472679966, "grad_norm": 0.1909942626953125, "learning_rate": 5.709052662120993e-05, "loss": 0.0235, "step": 11931 }, { "epoch": 1.2935819601040763, "grad_norm": 0.12399734556674957, "learning_rate": 5.708689975337299e-05, "loss": 0.0086, "step": 11932 }, { "epoch": 1.293690372940156, "grad_norm": 0.2373342216014862, "learning_rate": 5.7083272885536056e-05, "loss": 0.0135, "step": 11933 }, { "epoch": 1.2937987857762359, "grad_norm": 0.3008796274662018, "learning_rate": 5.707964601769912e-05, "loss": 0.0219, "step": 11934 }, { "epoch": 1.2939071986123156, "grad_norm": 0.03187984973192215, "learning_rate": 5.7076019149862184e-05, "loss": 0.003, "step": 11935 }, { "epoch": 1.2940156114483954, "grad_norm": 0.21874140202999115, "learning_rate": 5.707239228202524e-05, "loss": 0.0158, "step": 11936 }, { "epoch": 1.2941240242844754, "grad_norm": 0.3932831883430481, "learning_rate": 5.7068765414188306e-05, "loss": 0.0299, "step": 11937 }, { "epoch": 1.2942324371205551, "grad_norm": 0.16809743642807007, "learning_rate": 5.706513854635137e-05, "loss": 0.014, "step": 11938 }, { "epoch": 1.294340849956635, "grad_norm": 0.6420126557350159, "learning_rate": 5.7061511678514434e-05, "loss": 0.029, "step": 11939 }, { "epoch": 1.2944492627927147, "grad_norm": 0.038862597197294235, "learning_rate": 5.70578848106775e-05, "loss": 0.0027, "step": 11940 }, { "epoch": 1.2945576756287944, "grad_norm": 0.19458967447280884, "learning_rate": 5.705425794284056e-05, "loss": 0.0095, "step": 11941 }, { "epoch": 1.2946660884648742, "grad_norm": 0.24536357820034027, "learning_rate": 5.7050631075003626e-05, "loss": 0.0133, "step": 11942 }, { "epoch": 1.294774501300954, "grad_norm": 0.2626963257789612, "learning_rate": 5.70470042071667e-05, "loss": 0.0367, "step": 11943 }, { "epoch": 1.294882914137034, "grad_norm": 0.6711513996124268, "learning_rate": 5.704337733932976e-05, "loss": 0.0731, "step": 11944 }, { "epoch": 1.2949913269731137, "grad_norm": 0.25420865416526794, "learning_rate": 5.7039750471492825e-05, "loss": 0.0065, "step": 11945 }, { "epoch": 1.2950997398091935, "grad_norm": 0.08326700329780579, "learning_rate": 5.703612360365589e-05, "loss": 0.0104, "step": 11946 }, { "epoch": 1.2952081526452732, "grad_norm": 0.5037947297096252, "learning_rate": 5.703249673581895e-05, "loss": 0.0575, "step": 11947 }, { "epoch": 1.295316565481353, "grad_norm": 0.22962655127048492, "learning_rate": 5.702886986798202e-05, "loss": 0.0062, "step": 11948 }, { "epoch": 1.2954249783174328, "grad_norm": 0.20516417920589447, "learning_rate": 5.702524300014508e-05, "loss": 0.0094, "step": 11949 }, { "epoch": 1.2955333911535125, "grad_norm": 0.0793779045343399, "learning_rate": 5.702161613230814e-05, "loss": 0.0044, "step": 11950 }, { "epoch": 1.2956418039895925, "grad_norm": 0.10418754070997238, "learning_rate": 5.70179892644712e-05, "loss": 0.0049, "step": 11951 }, { "epoch": 1.295750216825672, "grad_norm": 0.2740260660648346, "learning_rate": 5.701436239663427e-05, "loss": 0.0178, "step": 11952 }, { "epoch": 1.295858629661752, "grad_norm": 0.29597699642181396, "learning_rate": 5.701073552879733e-05, "loss": 0.0224, "step": 11953 }, { "epoch": 1.2959670424978318, "grad_norm": 0.3519432842731476, "learning_rate": 5.7007108660960395e-05, "loss": 0.0474, "step": 11954 }, { "epoch": 1.2960754553339116, "grad_norm": 0.6133596897125244, "learning_rate": 5.700348179312346e-05, "loss": 0.0409, "step": 11955 }, { "epoch": 1.2961838681699913, "grad_norm": 0.21136456727981567, "learning_rate": 5.6999854925286524e-05, "loss": 0.0158, "step": 11956 }, { "epoch": 1.296292281006071, "grad_norm": 0.9151577949523926, "learning_rate": 5.699622805744959e-05, "loss": 0.0279, "step": 11957 }, { "epoch": 1.2964006938421508, "grad_norm": 0.3473403751850128, "learning_rate": 5.699260118961265e-05, "loss": 0.0178, "step": 11958 }, { "epoch": 1.2965091066782306, "grad_norm": 0.10010969638824463, "learning_rate": 5.698897432177571e-05, "loss": 0.0064, "step": 11959 }, { "epoch": 1.2966175195143106, "grad_norm": 0.15476444363594055, "learning_rate": 5.698534745393879e-05, "loss": 0.0306, "step": 11960 }, { "epoch": 1.2967259323503904, "grad_norm": 0.19073349237442017, "learning_rate": 5.698172058610185e-05, "loss": 0.0073, "step": 11961 }, { "epoch": 1.2968343451864701, "grad_norm": 0.32531699538230896, "learning_rate": 5.6978093718264915e-05, "loss": 0.0403, "step": 11962 }, { "epoch": 1.2969427580225499, "grad_norm": 0.48389849066734314, "learning_rate": 5.697446685042798e-05, "loss": 0.0633, "step": 11963 }, { "epoch": 1.2970511708586296, "grad_norm": 0.3551630675792694, "learning_rate": 5.697083998259104e-05, "loss": 0.0114, "step": 11964 }, { "epoch": 1.2971595836947094, "grad_norm": 0.11568370461463928, "learning_rate": 5.69672131147541e-05, "loss": 0.0141, "step": 11965 }, { "epoch": 1.2972679965307892, "grad_norm": 0.5346295833587646, "learning_rate": 5.6963586246917165e-05, "loss": 0.0805, "step": 11966 }, { "epoch": 1.2973764093668692, "grad_norm": 0.17608343064785004, "learning_rate": 5.695995937908023e-05, "loss": 0.0074, "step": 11967 }, { "epoch": 1.2974848222029487, "grad_norm": 0.16267576813697815, "learning_rate": 5.695633251124329e-05, "loss": 0.0148, "step": 11968 }, { "epoch": 1.2975932350390287, "grad_norm": 0.02316313609480858, "learning_rate": 5.695270564340636e-05, "loss": 0.0009, "step": 11969 }, { "epoch": 1.2977016478751084, "grad_norm": 0.49735358357429504, "learning_rate": 5.694907877556942e-05, "loss": 0.033, "step": 11970 }, { "epoch": 1.2978100607111882, "grad_norm": 0.4051618278026581, "learning_rate": 5.6945451907732485e-05, "loss": 0.0227, "step": 11971 }, { "epoch": 1.297918473547268, "grad_norm": 0.5976625084877014, "learning_rate": 5.694182503989555e-05, "loss": 0.0457, "step": 11972 }, { "epoch": 1.2980268863833477, "grad_norm": 0.10677023977041245, "learning_rate": 5.693819817205861e-05, "loss": 0.0058, "step": 11973 }, { "epoch": 1.2981352992194277, "grad_norm": 0.1840280443429947, "learning_rate": 5.693457130422167e-05, "loss": 0.0071, "step": 11974 }, { "epoch": 1.2982437120555073, "grad_norm": 0.1666882187128067, "learning_rate": 5.6930944436384735e-05, "loss": 0.0109, "step": 11975 }, { "epoch": 1.2983521248915872, "grad_norm": 0.42586129903793335, "learning_rate": 5.69273175685478e-05, "loss": 0.0308, "step": 11976 }, { "epoch": 1.298460537727667, "grad_norm": 0.8863441348075867, "learning_rate": 5.692369070071086e-05, "loss": 0.0335, "step": 11977 }, { "epoch": 1.2985689505637468, "grad_norm": 0.4527246356010437, "learning_rate": 5.692006383287394e-05, "loss": 0.0228, "step": 11978 }, { "epoch": 1.2986773633998265, "grad_norm": 0.35009777545928955, "learning_rate": 5.6916436965037e-05, "loss": 0.0149, "step": 11979 }, { "epoch": 1.2987857762359063, "grad_norm": 0.2592957615852356, "learning_rate": 5.691281009720006e-05, "loss": 0.0115, "step": 11980 }, { "epoch": 1.298894189071986, "grad_norm": 0.1727985292673111, "learning_rate": 5.6909183229363126e-05, "loss": 0.0093, "step": 11981 }, { "epoch": 1.2990026019080658, "grad_norm": 0.889973521232605, "learning_rate": 5.690555636152619e-05, "loss": 0.0559, "step": 11982 }, { "epoch": 1.2991110147441458, "grad_norm": 0.7793325781822205, "learning_rate": 5.6901929493689254e-05, "loss": 0.0198, "step": 11983 }, { "epoch": 1.2992194275802256, "grad_norm": 1.1861026287078857, "learning_rate": 5.689830262585232e-05, "loss": 0.0457, "step": 11984 }, { "epoch": 1.2993278404163053, "grad_norm": 0.39266160130500793, "learning_rate": 5.689467575801538e-05, "loss": 0.0345, "step": 11985 }, { "epoch": 1.299436253252385, "grad_norm": 0.3495473265647888, "learning_rate": 5.689104889017845e-05, "loss": 0.0326, "step": 11986 }, { "epoch": 1.2995446660884649, "grad_norm": 0.5283733010292053, "learning_rate": 5.688742202234151e-05, "loss": 0.0361, "step": 11987 }, { "epoch": 1.2996530789245446, "grad_norm": 0.45894020795822144, "learning_rate": 5.688379515450457e-05, "loss": 0.0445, "step": 11988 }, { "epoch": 1.2997614917606244, "grad_norm": 0.34079694747924805, "learning_rate": 5.688016828666763e-05, "loss": 0.0472, "step": 11989 }, { "epoch": 1.2998699045967044, "grad_norm": 0.49698376655578613, "learning_rate": 5.6876541418830696e-05, "loss": 0.031, "step": 11990 }, { "epoch": 1.299978317432784, "grad_norm": 0.13506975769996643, "learning_rate": 5.687291455099376e-05, "loss": 0.0058, "step": 11991 }, { "epoch": 1.300086730268864, "grad_norm": 0.35574913024902344, "learning_rate": 5.6869287683156825e-05, "loss": 0.0607, "step": 11992 }, { "epoch": 1.3001951431049437, "grad_norm": 0.3372373878955841, "learning_rate": 5.686566081531989e-05, "loss": 0.0123, "step": 11993 }, { "epoch": 1.3003035559410234, "grad_norm": 0.5081294775009155, "learning_rate": 5.686203394748295e-05, "loss": 0.0195, "step": 11994 }, { "epoch": 1.3004119687771032, "grad_norm": 0.12935695052146912, "learning_rate": 5.6858407079646024e-05, "loss": 0.012, "step": 11995 }, { "epoch": 1.300520381613183, "grad_norm": 0.1384454220533371, "learning_rate": 5.685478021180909e-05, "loss": 0.0111, "step": 11996 }, { "epoch": 1.3006287944492627, "grad_norm": 0.2555069327354431, "learning_rate": 5.685115334397215e-05, "loss": 0.0253, "step": 11997 }, { "epoch": 1.3007372072853425, "grad_norm": 0.1152535229921341, "learning_rate": 5.6847526476135216e-05, "loss": 0.0068, "step": 11998 }, { "epoch": 1.3008456201214225, "grad_norm": 0.27327704429626465, "learning_rate": 5.684389960829828e-05, "loss": 0.0267, "step": 11999 }, { "epoch": 1.3009540329575022, "grad_norm": 0.34705960750579834, "learning_rate": 5.6840272740461344e-05, "loss": 0.0169, "step": 12000 }, { "epoch": 1.301062445793582, "grad_norm": 0.1844567209482193, "learning_rate": 5.683664587262441e-05, "loss": 0.0238, "step": 12001 }, { "epoch": 1.3011708586296618, "grad_norm": 0.5227825045585632, "learning_rate": 5.6833019004787466e-05, "loss": 0.0539, "step": 12002 }, { "epoch": 1.3012792714657415, "grad_norm": 0.5309001207351685, "learning_rate": 5.682939213695053e-05, "loss": 0.0262, "step": 12003 }, { "epoch": 1.3013876843018213, "grad_norm": 0.12156441062688828, "learning_rate": 5.6825765269113594e-05, "loss": 0.0109, "step": 12004 }, { "epoch": 1.301496097137901, "grad_norm": 0.2880874574184418, "learning_rate": 5.682213840127666e-05, "loss": 0.0174, "step": 12005 }, { "epoch": 1.301604509973981, "grad_norm": 0.2475138157606125, "learning_rate": 5.681851153343972e-05, "loss": 0.0291, "step": 12006 }, { "epoch": 1.3017129228100608, "grad_norm": 0.40009135007858276, "learning_rate": 5.6814884665602786e-05, "loss": 0.0266, "step": 12007 }, { "epoch": 1.3018213356461406, "grad_norm": 0.02819368802011013, "learning_rate": 5.681125779776585e-05, "loss": 0.0031, "step": 12008 }, { "epoch": 1.3019297484822203, "grad_norm": 0.22005170583724976, "learning_rate": 5.6807630929928914e-05, "loss": 0.0082, "step": 12009 }, { "epoch": 1.3020381613183, "grad_norm": 0.37304067611694336, "learning_rate": 5.680400406209198e-05, "loss": 0.0473, "step": 12010 }, { "epoch": 1.3021465741543798, "grad_norm": 0.3480392098426819, "learning_rate": 5.6800377194255036e-05, "loss": 0.0206, "step": 12011 }, { "epoch": 1.3022549869904596, "grad_norm": 0.20478981733322144, "learning_rate": 5.6796750326418113e-05, "loss": 0.0083, "step": 12012 }, { "epoch": 1.3023633998265396, "grad_norm": 0.28147202730178833, "learning_rate": 5.679312345858118e-05, "loss": 0.0309, "step": 12013 }, { "epoch": 1.3024718126626191, "grad_norm": 0.1291205883026123, "learning_rate": 5.678949659074424e-05, "loss": 0.0126, "step": 12014 }, { "epoch": 1.3025802254986991, "grad_norm": 0.4962700605392456, "learning_rate": 5.6785869722907306e-05, "loss": 0.0457, "step": 12015 }, { "epoch": 1.3026886383347789, "grad_norm": 0.3403953015804291, "learning_rate": 5.678224285507037e-05, "loss": 0.0284, "step": 12016 }, { "epoch": 1.3027970511708586, "grad_norm": 0.03093559294939041, "learning_rate": 5.677861598723343e-05, "loss": 0.0032, "step": 12017 }, { "epoch": 1.3029054640069384, "grad_norm": 0.22738173604011536, "learning_rate": 5.677498911939649e-05, "loss": 0.0133, "step": 12018 }, { "epoch": 1.3030138768430182, "grad_norm": 0.0825812965631485, "learning_rate": 5.6771362251559555e-05, "loss": 0.0097, "step": 12019 }, { "epoch": 1.303122289679098, "grad_norm": 0.2232705056667328, "learning_rate": 5.676773538372262e-05, "loss": 0.011, "step": 12020 }, { "epoch": 1.3032307025151777, "grad_norm": 0.10025318711996078, "learning_rate": 5.6764108515885684e-05, "loss": 0.0146, "step": 12021 }, { "epoch": 1.3033391153512577, "grad_norm": 0.16892510652542114, "learning_rate": 5.676048164804875e-05, "loss": 0.0178, "step": 12022 }, { "epoch": 1.3034475281873374, "grad_norm": 0.02669220231473446, "learning_rate": 5.675685478021181e-05, "loss": 0.0016, "step": 12023 }, { "epoch": 1.3035559410234172, "grad_norm": 0.18561674654483795, "learning_rate": 5.6753227912374876e-05, "loss": 0.0158, "step": 12024 }, { "epoch": 1.303664353859497, "grad_norm": 0.26766708493232727, "learning_rate": 5.674960104453794e-05, "loss": 0.0179, "step": 12025 }, { "epoch": 1.3037727666955767, "grad_norm": 0.3026908040046692, "learning_rate": 5.6745974176701e-05, "loss": 0.0238, "step": 12026 }, { "epoch": 1.3038811795316565, "grad_norm": 0.5056848526000977, "learning_rate": 5.674234730886406e-05, "loss": 0.0236, "step": 12027 }, { "epoch": 1.3039895923677363, "grad_norm": 0.07362865656614304, "learning_rate": 5.6738720441027126e-05, "loss": 0.0043, "step": 12028 }, { "epoch": 1.3040980052038162, "grad_norm": 0.47990119457244873, "learning_rate": 5.67350935731902e-05, "loss": 0.0347, "step": 12029 }, { "epoch": 1.3042064180398958, "grad_norm": 0.2870183289051056, "learning_rate": 5.673146670535327e-05, "loss": 0.0222, "step": 12030 }, { "epoch": 1.3043148308759758, "grad_norm": 0.4451252520084381, "learning_rate": 5.6727839837516325e-05, "loss": 0.0735, "step": 12031 }, { "epoch": 1.3044232437120555, "grad_norm": 0.4442691206932068, "learning_rate": 5.672421296967939e-05, "loss": 0.0225, "step": 12032 }, { "epoch": 1.3045316565481353, "grad_norm": 0.3702642023563385, "learning_rate": 5.672058610184245e-05, "loss": 0.0166, "step": 12033 }, { "epoch": 1.304640069384215, "grad_norm": 0.2067805379629135, "learning_rate": 5.671695923400552e-05, "loss": 0.0172, "step": 12034 }, { "epoch": 1.3047484822202948, "grad_norm": 0.47177791595458984, "learning_rate": 5.671333236616858e-05, "loss": 0.0279, "step": 12035 }, { "epoch": 1.3048568950563748, "grad_norm": 0.49192243814468384, "learning_rate": 5.6709705498331645e-05, "loss": 0.0546, "step": 12036 }, { "epoch": 1.3049653078924544, "grad_norm": 0.020149124786257744, "learning_rate": 5.670607863049471e-05, "loss": 0.0013, "step": 12037 }, { "epoch": 1.3050737207285343, "grad_norm": 0.30600133538246155, "learning_rate": 5.670245176265777e-05, "loss": 0.0194, "step": 12038 }, { "epoch": 1.305182133564614, "grad_norm": 0.1962967962026596, "learning_rate": 5.669882489482084e-05, "loss": 0.0174, "step": 12039 }, { "epoch": 1.3052905464006939, "grad_norm": 0.39943885803222656, "learning_rate": 5.6695198026983895e-05, "loss": 0.0713, "step": 12040 }, { "epoch": 1.3053989592367736, "grad_norm": 0.8147252798080444, "learning_rate": 5.669157115914696e-05, "loss": 0.0382, "step": 12041 }, { "epoch": 1.3055073720728534, "grad_norm": 0.3440045118331909, "learning_rate": 5.668794429131002e-05, "loss": 0.0229, "step": 12042 }, { "epoch": 1.3056157849089332, "grad_norm": 0.05432254448533058, "learning_rate": 5.668431742347309e-05, "loss": 0.0036, "step": 12043 }, { "epoch": 1.305724197745013, "grad_norm": 0.6717670559883118, "learning_rate": 5.668069055563615e-05, "loss": 0.0357, "step": 12044 }, { "epoch": 1.305832610581093, "grad_norm": 0.2620905041694641, "learning_rate": 5.6677063687799215e-05, "loss": 0.0351, "step": 12045 }, { "epoch": 1.3059410234171727, "grad_norm": 0.2678307890892029, "learning_rate": 5.667343681996228e-05, "loss": 0.0531, "step": 12046 }, { "epoch": 1.3060494362532524, "grad_norm": 0.08903783559799194, "learning_rate": 5.666980995212535e-05, "loss": 0.0096, "step": 12047 }, { "epoch": 1.3061578490893322, "grad_norm": 0.4838157296180725, "learning_rate": 5.6666183084288414e-05, "loss": 0.0289, "step": 12048 }, { "epoch": 1.306266261925412, "grad_norm": 0.3302449882030487, "learning_rate": 5.666255621645148e-05, "loss": 0.0403, "step": 12049 }, { "epoch": 1.3063746747614917, "grad_norm": 0.22402076423168182, "learning_rate": 5.665892934861454e-05, "loss": 0.0116, "step": 12050 }, { "epoch": 1.3064830875975715, "grad_norm": 0.3699188828468323, "learning_rate": 5.665530248077761e-05, "loss": 0.049, "step": 12051 }, { "epoch": 1.3065915004336515, "grad_norm": 0.34564775228500366, "learning_rate": 5.665167561294067e-05, "loss": 0.0277, "step": 12052 }, { "epoch": 1.306699913269731, "grad_norm": 0.6507998108863831, "learning_rate": 5.6648048745103735e-05, "loss": 0.0255, "step": 12053 }, { "epoch": 1.306808326105811, "grad_norm": 0.09591688215732574, "learning_rate": 5.664442187726679e-05, "loss": 0.0045, "step": 12054 }, { "epoch": 1.3069167389418908, "grad_norm": 0.5499473810195923, "learning_rate": 5.6640795009429856e-05, "loss": 0.0449, "step": 12055 }, { "epoch": 1.3070251517779705, "grad_norm": 0.3700655400753021, "learning_rate": 5.663716814159292e-05, "loss": 0.0372, "step": 12056 }, { "epoch": 1.3071335646140503, "grad_norm": 0.5103635787963867, "learning_rate": 5.6633541273755985e-05, "loss": 0.048, "step": 12057 }, { "epoch": 1.30724197745013, "grad_norm": 0.26333847641944885, "learning_rate": 5.662991440591905e-05, "loss": 0.0097, "step": 12058 }, { "epoch": 1.3073503902862098, "grad_norm": 0.5365592241287231, "learning_rate": 5.662628753808211e-05, "loss": 0.0431, "step": 12059 }, { "epoch": 1.3074588031222896, "grad_norm": 0.26128512620925903, "learning_rate": 5.662266067024518e-05, "loss": 0.0312, "step": 12060 }, { "epoch": 1.3075672159583696, "grad_norm": 0.1094493716955185, "learning_rate": 5.661903380240824e-05, "loss": 0.0151, "step": 12061 }, { "epoch": 1.3076756287944493, "grad_norm": 0.11385276168584824, "learning_rate": 5.6615406934571305e-05, "loss": 0.0118, "step": 12062 }, { "epoch": 1.307784041630529, "grad_norm": 0.09336473047733307, "learning_rate": 5.661178006673436e-05, "loss": 0.0103, "step": 12063 }, { "epoch": 1.3078924544666088, "grad_norm": 0.22928257286548615, "learning_rate": 5.660815319889744e-05, "loss": 0.0261, "step": 12064 }, { "epoch": 1.3080008673026886, "grad_norm": 0.12454507499933243, "learning_rate": 5.6604526331060504e-05, "loss": 0.0079, "step": 12065 }, { "epoch": 1.3081092801387684, "grad_norm": 0.2507763206958771, "learning_rate": 5.660089946322357e-05, "loss": 0.0257, "step": 12066 }, { "epoch": 1.3082176929748481, "grad_norm": 0.10815580934286118, "learning_rate": 5.659727259538663e-05, "loss": 0.0092, "step": 12067 }, { "epoch": 1.3083261058109281, "grad_norm": 0.536492645740509, "learning_rate": 5.6593645727549697e-05, "loss": 0.0271, "step": 12068 }, { "epoch": 1.3084345186470079, "grad_norm": 0.14827628433704376, "learning_rate": 5.6590018859712754e-05, "loss": 0.0099, "step": 12069 }, { "epoch": 1.3085429314830876, "grad_norm": 0.48185157775878906, "learning_rate": 5.658639199187582e-05, "loss": 0.0709, "step": 12070 }, { "epoch": 1.3086513443191674, "grad_norm": 0.06147634983062744, "learning_rate": 5.658276512403888e-05, "loss": 0.0049, "step": 12071 }, { "epoch": 1.3087597571552472, "grad_norm": 0.48429179191589355, "learning_rate": 5.6579138256201946e-05, "loss": 0.0266, "step": 12072 }, { "epoch": 1.308868169991327, "grad_norm": 0.6167802810668945, "learning_rate": 5.657551138836501e-05, "loss": 0.0283, "step": 12073 }, { "epoch": 1.3089765828274067, "grad_norm": 0.3308619260787964, "learning_rate": 5.6571884520528074e-05, "loss": 0.0224, "step": 12074 }, { "epoch": 1.3090849956634867, "grad_norm": 0.17897608876228333, "learning_rate": 5.656825765269114e-05, "loss": 0.0303, "step": 12075 }, { "epoch": 1.3091934084995662, "grad_norm": 0.3977685570716858, "learning_rate": 5.65646307848542e-05, "loss": 0.054, "step": 12076 }, { "epoch": 1.3093018213356462, "grad_norm": 0.14362318813800812, "learning_rate": 5.656100391701727e-05, "loss": 0.0134, "step": 12077 }, { "epoch": 1.309410234171726, "grad_norm": 0.20473845303058624, "learning_rate": 5.6557377049180324e-05, "loss": 0.0341, "step": 12078 }, { "epoch": 1.3095186470078057, "grad_norm": 0.43339648842811584, "learning_rate": 5.655375018134339e-05, "loss": 0.0304, "step": 12079 }, { "epoch": 1.3096270598438855, "grad_norm": 0.03889700025320053, "learning_rate": 5.655012331350645e-05, "loss": 0.0025, "step": 12080 }, { "epoch": 1.3097354726799653, "grad_norm": 0.5104583501815796, "learning_rate": 5.654649644566953e-05, "loss": 0.0421, "step": 12081 }, { "epoch": 1.309843885516045, "grad_norm": 0.3868224620819092, "learning_rate": 5.6542869577832594e-05, "loss": 0.0498, "step": 12082 }, { "epoch": 1.3099522983521248, "grad_norm": 0.1200290098786354, "learning_rate": 5.653924270999565e-05, "loss": 0.0089, "step": 12083 }, { "epoch": 1.3100607111882048, "grad_norm": 0.1050770953297615, "learning_rate": 5.6535615842158715e-05, "loss": 0.0062, "step": 12084 }, { "epoch": 1.3101691240242845, "grad_norm": 0.6092835068702698, "learning_rate": 5.653198897432178e-05, "loss": 0.0964, "step": 12085 }, { "epoch": 1.3102775368603643, "grad_norm": 0.15658724308013916, "learning_rate": 5.6528362106484844e-05, "loss": 0.0166, "step": 12086 }, { "epoch": 1.310385949696444, "grad_norm": 0.3780083656311035, "learning_rate": 5.652473523864791e-05, "loss": 0.0363, "step": 12087 }, { "epoch": 1.3104943625325238, "grad_norm": 0.23924334347248077, "learning_rate": 5.652110837081097e-05, "loss": 0.0205, "step": 12088 }, { "epoch": 1.3106027753686036, "grad_norm": 0.18700966238975525, "learning_rate": 5.6517481502974036e-05, "loss": 0.0401, "step": 12089 }, { "epoch": 1.3107111882046834, "grad_norm": 0.7193470597267151, "learning_rate": 5.65138546351371e-05, "loss": 0.0228, "step": 12090 }, { "epoch": 1.3108196010407633, "grad_norm": 0.44031965732574463, "learning_rate": 5.6510227767300164e-05, "loss": 0.0217, "step": 12091 }, { "epoch": 1.310928013876843, "grad_norm": 0.42624637484550476, "learning_rate": 5.650660089946322e-05, "loss": 0.0166, "step": 12092 }, { "epoch": 1.3110364267129229, "grad_norm": 0.2116471230983734, "learning_rate": 5.6502974031626286e-05, "loss": 0.0155, "step": 12093 }, { "epoch": 1.3111448395490026, "grad_norm": 0.1736718863248825, "learning_rate": 5.649934716378935e-05, "loss": 0.0139, "step": 12094 }, { "epoch": 1.3112532523850824, "grad_norm": 0.3178413510322571, "learning_rate": 5.6495720295952414e-05, "loss": 0.0223, "step": 12095 }, { "epoch": 1.3113616652211622, "grad_norm": 0.4600026607513428, "learning_rate": 5.649209342811548e-05, "loss": 0.0362, "step": 12096 }, { "epoch": 1.311470078057242, "grad_norm": 0.12539832293987274, "learning_rate": 5.648846656027854e-05, "loss": 0.0067, "step": 12097 }, { "epoch": 1.311578490893322, "grad_norm": 0.21684780716896057, "learning_rate": 5.648483969244161e-05, "loss": 0.0197, "step": 12098 }, { "epoch": 1.3116869037294014, "grad_norm": 0.12821348011493683, "learning_rate": 5.648121282460468e-05, "loss": 0.0081, "step": 12099 }, { "epoch": 1.3117953165654814, "grad_norm": 0.42675888538360596, "learning_rate": 5.647758595676774e-05, "loss": 0.0345, "step": 12100 }, { "epoch": 1.3119037294015612, "grad_norm": 0.34642037749290466, "learning_rate": 5.6473959088930805e-05, "loss": 0.028, "step": 12101 }, { "epoch": 1.312012142237641, "grad_norm": 1.1866450309753418, "learning_rate": 5.647033222109387e-05, "loss": 0.0713, "step": 12102 }, { "epoch": 1.3121205550737207, "grad_norm": 0.12672460079193115, "learning_rate": 5.6466705353256933e-05, "loss": 0.006, "step": 12103 }, { "epoch": 1.3122289679098005, "grad_norm": 0.1045265644788742, "learning_rate": 5.646307848542e-05, "loss": 0.006, "step": 12104 }, { "epoch": 1.3123373807458802, "grad_norm": 0.03252212703227997, "learning_rate": 5.645945161758306e-05, "loss": 0.0034, "step": 12105 }, { "epoch": 1.31244579358196, "grad_norm": 0.235101580619812, "learning_rate": 5.645582474974612e-05, "loss": 0.0217, "step": 12106 }, { "epoch": 1.31255420641804, "grad_norm": 0.3632155656814575, "learning_rate": 5.645219788190918e-05, "loss": 0.0319, "step": 12107 }, { "epoch": 1.3126626192541198, "grad_norm": 0.30486857891082764, "learning_rate": 5.644857101407225e-05, "loss": 0.0107, "step": 12108 }, { "epoch": 1.3127710320901995, "grad_norm": 0.2738145887851715, "learning_rate": 5.644494414623531e-05, "loss": 0.0576, "step": 12109 }, { "epoch": 1.3128794449262793, "grad_norm": 0.2052944153547287, "learning_rate": 5.6441317278398375e-05, "loss": 0.0138, "step": 12110 }, { "epoch": 1.312987857762359, "grad_norm": 0.468194842338562, "learning_rate": 5.643769041056144e-05, "loss": 0.0601, "step": 12111 }, { "epoch": 1.3130962705984388, "grad_norm": 0.11674842983484268, "learning_rate": 5.6434063542724504e-05, "loss": 0.0079, "step": 12112 }, { "epoch": 1.3132046834345186, "grad_norm": 0.35781237483024597, "learning_rate": 5.643043667488757e-05, "loss": 0.0386, "step": 12113 }, { "epoch": 1.3133130962705986, "grad_norm": 0.22035814821720123, "learning_rate": 5.642680980705063e-05, "loss": 0.0239, "step": 12114 }, { "epoch": 1.313421509106678, "grad_norm": 0.315667062997818, "learning_rate": 5.642318293921369e-05, "loss": 0.0246, "step": 12115 }, { "epoch": 1.313529921942758, "grad_norm": 0.0912354588508606, "learning_rate": 5.641955607137677e-05, "loss": 0.0119, "step": 12116 }, { "epoch": 1.3136383347788378, "grad_norm": 0.0791173055768013, "learning_rate": 5.641592920353983e-05, "loss": 0.008, "step": 12117 }, { "epoch": 1.3137467476149176, "grad_norm": 0.12522879242897034, "learning_rate": 5.6412302335702895e-05, "loss": 0.0116, "step": 12118 }, { "epoch": 1.3138551604509974, "grad_norm": 0.13936644792556763, "learning_rate": 5.640867546786596e-05, "loss": 0.0103, "step": 12119 }, { "epoch": 1.3139635732870771, "grad_norm": 0.3020187318325043, "learning_rate": 5.640504860002902e-05, "loss": 0.0287, "step": 12120 }, { "epoch": 1.314071986123157, "grad_norm": 0.10209860652685165, "learning_rate": 5.640142173219208e-05, "loss": 0.0063, "step": 12121 }, { "epoch": 1.3141803989592367, "grad_norm": 0.2251698225736618, "learning_rate": 5.6397794864355145e-05, "loss": 0.0222, "step": 12122 }, { "epoch": 1.3142888117953166, "grad_norm": 0.35000020265579224, "learning_rate": 5.639416799651821e-05, "loss": 0.0083, "step": 12123 }, { "epoch": 1.3143972246313964, "grad_norm": 0.26185745000839233, "learning_rate": 5.639054112868127e-05, "loss": 0.0157, "step": 12124 }, { "epoch": 1.3145056374674762, "grad_norm": 0.20386148989200592, "learning_rate": 5.638691426084434e-05, "loss": 0.0102, "step": 12125 }, { "epoch": 1.314614050303556, "grad_norm": 0.44484269618988037, "learning_rate": 5.63832873930074e-05, "loss": 0.0144, "step": 12126 }, { "epoch": 1.3147224631396357, "grad_norm": 0.1368587464094162, "learning_rate": 5.6379660525170465e-05, "loss": 0.0071, "step": 12127 }, { "epoch": 1.3148308759757155, "grad_norm": 0.555603563785553, "learning_rate": 5.637603365733353e-05, "loss": 0.0421, "step": 12128 }, { "epoch": 1.3149392888117952, "grad_norm": 0.18837158381938934, "learning_rate": 5.637240678949659e-05, "loss": 0.0299, "step": 12129 }, { "epoch": 1.3150477016478752, "grad_norm": 0.2106725424528122, "learning_rate": 5.636877992165965e-05, "loss": 0.0066, "step": 12130 }, { "epoch": 1.315156114483955, "grad_norm": 0.09123805910348892, "learning_rate": 5.6365153053822715e-05, "loss": 0.0041, "step": 12131 }, { "epoch": 1.3152645273200347, "grad_norm": 0.11772848665714264, "learning_rate": 5.636152618598578e-05, "loss": 0.0105, "step": 12132 }, { "epoch": 1.3153729401561145, "grad_norm": 0.2952020764350891, "learning_rate": 5.6357899318148857e-05, "loss": 0.0188, "step": 12133 }, { "epoch": 1.3154813529921943, "grad_norm": 0.18199165165424347, "learning_rate": 5.635427245031192e-05, "loss": 0.0068, "step": 12134 }, { "epoch": 1.315589765828274, "grad_norm": 0.35486310720443726, "learning_rate": 5.635064558247498e-05, "loss": 0.0183, "step": 12135 }, { "epoch": 1.3156981786643538, "grad_norm": 0.5937708616256714, "learning_rate": 5.634701871463804e-05, "loss": 0.0363, "step": 12136 }, { "epoch": 1.3158065915004338, "grad_norm": 0.673014223575592, "learning_rate": 5.6343391846801106e-05, "loss": 0.0706, "step": 12137 }, { "epoch": 1.3159150043365133, "grad_norm": 0.12063125520944595, "learning_rate": 5.633976497896417e-05, "loss": 0.0114, "step": 12138 }, { "epoch": 1.3160234171725933, "grad_norm": 0.013977628201246262, "learning_rate": 5.6336138111127234e-05, "loss": 0.0008, "step": 12139 }, { "epoch": 1.316131830008673, "grad_norm": 0.17852753400802612, "learning_rate": 5.63325112432903e-05, "loss": 0.0137, "step": 12140 }, { "epoch": 1.3162402428447528, "grad_norm": 0.2777639925479889, "learning_rate": 5.632888437545336e-05, "loss": 0.0152, "step": 12141 }, { "epoch": 1.3163486556808326, "grad_norm": 0.2579019367694855, "learning_rate": 5.632525750761643e-05, "loss": 0.0176, "step": 12142 }, { "epoch": 1.3164570685169124, "grad_norm": 0.5344899892807007, "learning_rate": 5.632163063977949e-05, "loss": 0.0194, "step": 12143 }, { "epoch": 1.3165654813529921, "grad_norm": 0.3597272038459778, "learning_rate": 5.631800377194255e-05, "loss": 0.0159, "step": 12144 }, { "epoch": 1.3166738941890719, "grad_norm": 0.15308691561222076, "learning_rate": 5.631437690410561e-05, "loss": 0.0127, "step": 12145 }, { "epoch": 1.3167823070251519, "grad_norm": 0.14272621273994446, "learning_rate": 5.6310750036268676e-05, "loss": 0.0078, "step": 12146 }, { "epoch": 1.3168907198612316, "grad_norm": 0.32345426082611084, "learning_rate": 5.630712316843174e-05, "loss": 0.0146, "step": 12147 }, { "epoch": 1.3169991326973114, "grad_norm": 0.22958891093730927, "learning_rate": 5.6303496300594805e-05, "loss": 0.0188, "step": 12148 }, { "epoch": 1.3171075455333912, "grad_norm": 0.35410812497138977, "learning_rate": 5.629986943275787e-05, "loss": 0.0278, "step": 12149 }, { "epoch": 1.317215958369471, "grad_norm": 0.44246232509613037, "learning_rate": 5.629624256492094e-05, "loss": 0.0241, "step": 12150 }, { "epoch": 1.3173243712055507, "grad_norm": 0.24581943452358246, "learning_rate": 5.6292615697084004e-05, "loss": 0.0181, "step": 12151 }, { "epoch": 1.3174327840416304, "grad_norm": 0.1803753674030304, "learning_rate": 5.628898882924707e-05, "loss": 0.0056, "step": 12152 }, { "epoch": 1.3175411968777104, "grad_norm": 0.46456775069236755, "learning_rate": 5.628536196141013e-05, "loss": 0.0316, "step": 12153 }, { "epoch": 1.3176496097137902, "grad_norm": 0.12121604382991791, "learning_rate": 5.6281735093573196e-05, "loss": 0.0033, "step": 12154 }, { "epoch": 1.31775802254987, "grad_norm": 0.3707236349582672, "learning_rate": 5.627810822573626e-05, "loss": 0.0414, "step": 12155 }, { "epoch": 1.3178664353859497, "grad_norm": 0.022942036390304565, "learning_rate": 5.6274481357899324e-05, "loss": 0.0018, "step": 12156 }, { "epoch": 1.3179748482220295, "grad_norm": 0.3494415879249573, "learning_rate": 5.627085449006239e-05, "loss": 0.0152, "step": 12157 }, { "epoch": 1.3180832610581092, "grad_norm": 0.32228708267211914, "learning_rate": 5.626722762222545e-05, "loss": 0.0187, "step": 12158 }, { "epoch": 1.318191673894189, "grad_norm": 0.13085371255874634, "learning_rate": 5.626360075438851e-05, "loss": 0.0063, "step": 12159 }, { "epoch": 1.318300086730269, "grad_norm": 0.298567533493042, "learning_rate": 5.6259973886551574e-05, "loss": 0.0258, "step": 12160 }, { "epoch": 1.3184084995663485, "grad_norm": 0.29537463188171387, "learning_rate": 5.625634701871464e-05, "loss": 0.0447, "step": 12161 }, { "epoch": 1.3185169124024285, "grad_norm": 0.39042049646377563, "learning_rate": 5.62527201508777e-05, "loss": 0.039, "step": 12162 }, { "epoch": 1.3186253252385083, "grad_norm": 0.19283095002174377, "learning_rate": 5.6249093283040766e-05, "loss": 0.0329, "step": 12163 }, { "epoch": 1.318733738074588, "grad_norm": 0.3873671591281891, "learning_rate": 5.624546641520383e-05, "loss": 0.0376, "step": 12164 }, { "epoch": 1.3188421509106678, "grad_norm": 0.4434621334075928, "learning_rate": 5.6241839547366894e-05, "loss": 0.0406, "step": 12165 }, { "epoch": 1.3189505637467476, "grad_norm": 0.42967501282691956, "learning_rate": 5.623821267952996e-05, "loss": 0.0213, "step": 12166 }, { "epoch": 1.3190589765828273, "grad_norm": 0.18511617183685303, "learning_rate": 5.6234585811693016e-05, "loss": 0.0136, "step": 12167 }, { "epoch": 1.319167389418907, "grad_norm": 0.21478788554668427, "learning_rate": 5.6230958943856093e-05, "loss": 0.0066, "step": 12168 }, { "epoch": 1.319275802254987, "grad_norm": 0.2930874824523926, "learning_rate": 5.622733207601916e-05, "loss": 0.0436, "step": 12169 }, { "epoch": 1.3193842150910668, "grad_norm": 0.3103812038898468, "learning_rate": 5.622370520818222e-05, "loss": 0.0226, "step": 12170 }, { "epoch": 1.3194926279271466, "grad_norm": 0.6042952537536621, "learning_rate": 5.6220078340345286e-05, "loss": 0.0313, "step": 12171 }, { "epoch": 1.3196010407632264, "grad_norm": 0.43075838685035706, "learning_rate": 5.621645147250835e-05, "loss": 0.0222, "step": 12172 }, { "epoch": 1.3197094535993061, "grad_norm": 0.4840214252471924, "learning_rate": 5.621282460467141e-05, "loss": 0.0785, "step": 12173 }, { "epoch": 1.319817866435386, "grad_norm": 0.684963047504425, "learning_rate": 5.620919773683447e-05, "loss": 0.0272, "step": 12174 }, { "epoch": 1.3199262792714657, "grad_norm": 0.5545527935028076, "learning_rate": 5.6205570868997535e-05, "loss": 0.1183, "step": 12175 }, { "epoch": 1.3200346921075456, "grad_norm": 0.05585027113556862, "learning_rate": 5.62019440011606e-05, "loss": 0.0053, "step": 12176 }, { "epoch": 1.3201431049436252, "grad_norm": 0.7402557730674744, "learning_rate": 5.6198317133323664e-05, "loss": 0.0593, "step": 12177 }, { "epoch": 1.3202515177797052, "grad_norm": 0.3768298327922821, "learning_rate": 5.619469026548673e-05, "loss": 0.0464, "step": 12178 }, { "epoch": 1.320359930615785, "grad_norm": 0.39278116822242737, "learning_rate": 5.619106339764979e-05, "loss": 0.0397, "step": 12179 }, { "epoch": 1.3204683434518647, "grad_norm": 1.213510513305664, "learning_rate": 5.6187436529812856e-05, "loss": 0.0453, "step": 12180 }, { "epoch": 1.3205767562879445, "grad_norm": 0.26999393105506897, "learning_rate": 5.618380966197592e-05, "loss": 0.0232, "step": 12181 }, { "epoch": 1.3206851691240242, "grad_norm": 0.26279976963996887, "learning_rate": 5.618018279413898e-05, "loss": 0.0156, "step": 12182 }, { "epoch": 1.3207935819601042, "grad_norm": 0.4827944040298462, "learning_rate": 5.617655592630204e-05, "loss": 0.0326, "step": 12183 }, { "epoch": 1.3209019947961838, "grad_norm": 0.09852943569421768, "learning_rate": 5.6172929058465106e-05, "loss": 0.0104, "step": 12184 }, { "epoch": 1.3210104076322637, "grad_norm": 0.3940156400203705, "learning_rate": 5.616930219062818e-05, "loss": 0.025, "step": 12185 }, { "epoch": 1.3211188204683435, "grad_norm": 0.2854669392108917, "learning_rate": 5.616567532279125e-05, "loss": 0.0293, "step": 12186 }, { "epoch": 1.3212272333044233, "grad_norm": 0.2871212363243103, "learning_rate": 5.6162048454954305e-05, "loss": 0.0152, "step": 12187 }, { "epoch": 1.321335646140503, "grad_norm": 0.11391323059797287, "learning_rate": 5.615842158711737e-05, "loss": 0.0131, "step": 12188 }, { "epoch": 1.3214440589765828, "grad_norm": 0.3702511191368103, "learning_rate": 5.615479471928043e-05, "loss": 0.0392, "step": 12189 }, { "epoch": 1.3215524718126626, "grad_norm": 0.6661690473556519, "learning_rate": 5.61511678514435e-05, "loss": 0.0374, "step": 12190 }, { "epoch": 1.3216608846487423, "grad_norm": 0.3473517894744873, "learning_rate": 5.614754098360656e-05, "loss": 0.0219, "step": 12191 }, { "epoch": 1.3217692974848223, "grad_norm": 0.3086492419242859, "learning_rate": 5.6143914115769625e-05, "loss": 0.0358, "step": 12192 }, { "epoch": 1.321877710320902, "grad_norm": 0.24734711647033691, "learning_rate": 5.614028724793269e-05, "loss": 0.0364, "step": 12193 }, { "epoch": 1.3219861231569818, "grad_norm": 0.16709759831428528, "learning_rate": 5.6136660380095753e-05, "loss": 0.0174, "step": 12194 }, { "epoch": 1.3220945359930616, "grad_norm": 0.49679481983184814, "learning_rate": 5.613303351225882e-05, "loss": 0.0161, "step": 12195 }, { "epoch": 1.3222029488291414, "grad_norm": 0.45581868290901184, "learning_rate": 5.6129406644421875e-05, "loss": 0.0635, "step": 12196 }, { "epoch": 1.3223113616652211, "grad_norm": 0.6132321953773499, "learning_rate": 5.612577977658494e-05, "loss": 0.0236, "step": 12197 }, { "epoch": 1.3224197745013009, "grad_norm": 0.5042348504066467, "learning_rate": 5.6122152908748e-05, "loss": 0.0301, "step": 12198 }, { "epoch": 1.3225281873373809, "grad_norm": 0.31497666239738464, "learning_rate": 5.611852604091107e-05, "loss": 0.037, "step": 12199 }, { "epoch": 1.3226366001734604, "grad_norm": 0.20773406326770782, "learning_rate": 5.611489917307413e-05, "loss": 0.0159, "step": 12200 }, { "epoch": 1.3227450130095404, "grad_norm": 0.14420314133167267, "learning_rate": 5.6111272305237195e-05, "loss": 0.0107, "step": 12201 }, { "epoch": 1.3228534258456202, "grad_norm": 0.38636156916618347, "learning_rate": 5.6107645437400266e-05, "loss": 0.0364, "step": 12202 }, { "epoch": 1.3229618386817, "grad_norm": 0.11700478941202164, "learning_rate": 5.610401856956333e-05, "loss": 0.0081, "step": 12203 }, { "epoch": 1.3230702515177797, "grad_norm": 0.17114539444446564, "learning_rate": 5.6100391701726394e-05, "loss": 0.0186, "step": 12204 }, { "epoch": 1.3231786643538594, "grad_norm": 0.0639588013291359, "learning_rate": 5.609676483388946e-05, "loss": 0.0057, "step": 12205 }, { "epoch": 1.3232870771899392, "grad_norm": 0.20804691314697266, "learning_rate": 5.609313796605252e-05, "loss": 0.0217, "step": 12206 }, { "epoch": 1.323395490026019, "grad_norm": 0.23746758699417114, "learning_rate": 5.608951109821559e-05, "loss": 0.0104, "step": 12207 }, { "epoch": 1.323503902862099, "grad_norm": 0.4100251793861389, "learning_rate": 5.608588423037865e-05, "loss": 0.0446, "step": 12208 }, { "epoch": 1.3236123156981787, "grad_norm": 0.4862048327922821, "learning_rate": 5.6082257362541715e-05, "loss": 0.0454, "step": 12209 }, { "epoch": 1.3237207285342585, "grad_norm": 0.4736904799938202, "learning_rate": 5.607863049470478e-05, "loss": 0.0399, "step": 12210 }, { "epoch": 1.3238291413703382, "grad_norm": 0.08349017798900604, "learning_rate": 5.6075003626867836e-05, "loss": 0.0046, "step": 12211 }, { "epoch": 1.323937554206418, "grad_norm": 0.43972086906433105, "learning_rate": 5.60713767590309e-05, "loss": 0.0309, "step": 12212 }, { "epoch": 1.3240459670424978, "grad_norm": 0.38305148482322693, "learning_rate": 5.6067749891193965e-05, "loss": 0.0373, "step": 12213 }, { "epoch": 1.3241543798785775, "grad_norm": 0.49348464608192444, "learning_rate": 5.606412302335703e-05, "loss": 0.0419, "step": 12214 }, { "epoch": 1.3242627927146575, "grad_norm": 0.3436781167984009, "learning_rate": 5.606049615552009e-05, "loss": 0.0362, "step": 12215 }, { "epoch": 1.3243712055507373, "grad_norm": 0.14112971723079681, "learning_rate": 5.605686928768316e-05, "loss": 0.0097, "step": 12216 }, { "epoch": 1.324479618386817, "grad_norm": 0.34596237540245056, "learning_rate": 5.605324241984622e-05, "loss": 0.0351, "step": 12217 }, { "epoch": 1.3245880312228968, "grad_norm": 0.3362559974193573, "learning_rate": 5.6049615552009285e-05, "loss": 0.0292, "step": 12218 }, { "epoch": 1.3246964440589766, "grad_norm": 0.12019281834363937, "learning_rate": 5.6045988684172356e-05, "loss": 0.0153, "step": 12219 }, { "epoch": 1.3248048568950563, "grad_norm": 0.4543495774269104, "learning_rate": 5.604236181633542e-05, "loss": 0.0387, "step": 12220 }, { "epoch": 1.324913269731136, "grad_norm": 0.13923780620098114, "learning_rate": 5.6038734948498484e-05, "loss": 0.0134, "step": 12221 }, { "epoch": 1.325021682567216, "grad_norm": 0.15916544198989868, "learning_rate": 5.603510808066155e-05, "loss": 0.0212, "step": 12222 }, { "epoch": 1.3251300954032956, "grad_norm": 0.08059830218553543, "learning_rate": 5.603148121282461e-05, "loss": 0.0085, "step": 12223 }, { "epoch": 1.3252385082393756, "grad_norm": 0.1454668492078781, "learning_rate": 5.6027854344987677e-05, "loss": 0.0141, "step": 12224 }, { "epoch": 1.3253469210754554, "grad_norm": 0.12990930676460266, "learning_rate": 5.6024227477150734e-05, "loss": 0.0149, "step": 12225 }, { "epoch": 1.3254553339115351, "grad_norm": 0.355217844247818, "learning_rate": 5.60206006093138e-05, "loss": 0.034, "step": 12226 }, { "epoch": 1.325563746747615, "grad_norm": 0.17956668138504028, "learning_rate": 5.601697374147686e-05, "loss": 0.0162, "step": 12227 }, { "epoch": 1.3256721595836947, "grad_norm": 0.2528461813926697, "learning_rate": 5.6013346873639926e-05, "loss": 0.0223, "step": 12228 }, { "epoch": 1.3257805724197744, "grad_norm": 0.49970895051956177, "learning_rate": 5.600972000580299e-05, "loss": 0.0241, "step": 12229 }, { "epoch": 1.3258889852558542, "grad_norm": 0.2617780864238739, "learning_rate": 5.6006093137966054e-05, "loss": 0.0209, "step": 12230 }, { "epoch": 1.3259973980919342, "grad_norm": 0.603455662727356, "learning_rate": 5.600246627012912e-05, "loss": 0.0187, "step": 12231 }, { "epoch": 1.326105810928014, "grad_norm": 0.17041167616844177, "learning_rate": 5.599883940229218e-05, "loss": 0.0143, "step": 12232 }, { "epoch": 1.3262142237640937, "grad_norm": 0.36825111508369446, "learning_rate": 5.599521253445525e-05, "loss": 0.0256, "step": 12233 }, { "epoch": 1.3263226366001735, "grad_norm": 0.23975799977779388, "learning_rate": 5.5991585666618304e-05, "loss": 0.0127, "step": 12234 }, { "epoch": 1.3264310494362532, "grad_norm": 0.2629188597202301, "learning_rate": 5.598795879878137e-05, "loss": 0.0353, "step": 12235 }, { "epoch": 1.326539462272333, "grad_norm": 0.3383443057537079, "learning_rate": 5.598433193094443e-05, "loss": 0.0297, "step": 12236 }, { "epoch": 1.3266478751084128, "grad_norm": 0.4569780230522156, "learning_rate": 5.598070506310751e-05, "loss": 0.0275, "step": 12237 }, { "epoch": 1.3267562879444927, "grad_norm": 0.4067631661891937, "learning_rate": 5.5977078195270574e-05, "loss": 0.0547, "step": 12238 }, { "epoch": 1.3268647007805723, "grad_norm": 0.4056934714317322, "learning_rate": 5.597345132743363e-05, "loss": 0.0318, "step": 12239 }, { "epoch": 1.3269731136166523, "grad_norm": 0.05402279645204544, "learning_rate": 5.5969824459596695e-05, "loss": 0.0032, "step": 12240 }, { "epoch": 1.327081526452732, "grad_norm": 0.638219952583313, "learning_rate": 5.596619759175976e-05, "loss": 0.0359, "step": 12241 }, { "epoch": 1.3271899392888118, "grad_norm": 0.27884286642074585, "learning_rate": 5.5962570723922824e-05, "loss": 0.0254, "step": 12242 }, { "epoch": 1.3272983521248916, "grad_norm": 0.21465902030467987, "learning_rate": 5.595894385608589e-05, "loss": 0.013, "step": 12243 }, { "epoch": 1.3274067649609713, "grad_norm": 0.43251147866249084, "learning_rate": 5.595531698824895e-05, "loss": 0.0218, "step": 12244 }, { "epoch": 1.3275151777970513, "grad_norm": 0.16994482278823853, "learning_rate": 5.5951690120412016e-05, "loss": 0.0055, "step": 12245 }, { "epoch": 1.3276235906331308, "grad_norm": 0.13222891092300415, "learning_rate": 5.594806325257508e-05, "loss": 0.0114, "step": 12246 }, { "epoch": 1.3277320034692108, "grad_norm": 0.02982698753476143, "learning_rate": 5.5944436384738144e-05, "loss": 0.002, "step": 12247 }, { "epoch": 1.3278404163052906, "grad_norm": 0.44941243529319763, "learning_rate": 5.59408095169012e-05, "loss": 0.0277, "step": 12248 }, { "epoch": 1.3279488291413704, "grad_norm": 0.2007380574941635, "learning_rate": 5.5937182649064266e-05, "loss": 0.0173, "step": 12249 }, { "epoch": 1.3280572419774501, "grad_norm": 0.6147595047950745, "learning_rate": 5.593355578122733e-05, "loss": 0.0343, "step": 12250 }, { "epoch": 1.3281656548135299, "grad_norm": 0.1473323255777359, "learning_rate": 5.5929928913390394e-05, "loss": 0.0134, "step": 12251 }, { "epoch": 1.3282740676496096, "grad_norm": 1.542765736579895, "learning_rate": 5.592630204555346e-05, "loss": 0.064, "step": 12252 }, { "epoch": 1.3283824804856894, "grad_norm": 0.41321489214897156, "learning_rate": 5.592267517771652e-05, "loss": 0.0139, "step": 12253 }, { "epoch": 1.3284908933217694, "grad_norm": 0.2920319437980652, "learning_rate": 5.591904830987959e-05, "loss": 0.034, "step": 12254 }, { "epoch": 1.3285993061578492, "grad_norm": 0.23642444610595703, "learning_rate": 5.591542144204266e-05, "loss": 0.0125, "step": 12255 }, { "epoch": 1.328707718993929, "grad_norm": 0.3716852366924286, "learning_rate": 5.591179457420572e-05, "loss": 0.0331, "step": 12256 }, { "epoch": 1.3288161318300087, "grad_norm": 0.3485794961452484, "learning_rate": 5.5908167706368785e-05, "loss": 0.0263, "step": 12257 }, { "epoch": 1.3289245446660884, "grad_norm": 0.6313166618347168, "learning_rate": 5.590454083853185e-05, "loss": 0.0409, "step": 12258 }, { "epoch": 1.3290329575021682, "grad_norm": 0.15008236467838287, "learning_rate": 5.5900913970694913e-05, "loss": 0.0104, "step": 12259 }, { "epoch": 1.329141370338248, "grad_norm": 0.15115593373775482, "learning_rate": 5.589728710285798e-05, "loss": 0.007, "step": 12260 }, { "epoch": 1.329249783174328, "grad_norm": 0.7569429874420166, "learning_rate": 5.589366023502104e-05, "loss": 0.0124, "step": 12261 }, { "epoch": 1.3293581960104075, "grad_norm": 0.18260955810546875, "learning_rate": 5.5890033367184106e-05, "loss": 0.0221, "step": 12262 }, { "epoch": 1.3294666088464875, "grad_norm": 0.08223060518503189, "learning_rate": 5.588640649934716e-05, "loss": 0.0031, "step": 12263 }, { "epoch": 1.3295750216825672, "grad_norm": 0.4509664475917816, "learning_rate": 5.588277963151023e-05, "loss": 0.0309, "step": 12264 }, { "epoch": 1.329683434518647, "grad_norm": 0.42001575231552124, "learning_rate": 5.587915276367329e-05, "loss": 0.0233, "step": 12265 }, { "epoch": 1.3297918473547268, "grad_norm": 0.1474926620721817, "learning_rate": 5.5875525895836355e-05, "loss": 0.009, "step": 12266 }, { "epoch": 1.3299002601908065, "grad_norm": 0.8163942694664001, "learning_rate": 5.587189902799942e-05, "loss": 0.0368, "step": 12267 }, { "epoch": 1.3300086730268863, "grad_norm": 0.6655745506286621, "learning_rate": 5.5868272160162484e-05, "loss": 0.0289, "step": 12268 }, { "epoch": 1.330117085862966, "grad_norm": 0.23494184017181396, "learning_rate": 5.586464529232555e-05, "loss": 0.0193, "step": 12269 }, { "epoch": 1.330225498699046, "grad_norm": 0.1315658539533615, "learning_rate": 5.586101842448861e-05, "loss": 0.0082, "step": 12270 }, { "epoch": 1.3303339115351258, "grad_norm": 0.18841031193733215, "learning_rate": 5.585739155665168e-05, "loss": 0.0075, "step": 12271 }, { "epoch": 1.3304423243712056, "grad_norm": 0.506734311580658, "learning_rate": 5.585376468881475e-05, "loss": 0.0542, "step": 12272 }, { "epoch": 1.3305507372072853, "grad_norm": 0.29543256759643555, "learning_rate": 5.585013782097781e-05, "loss": 0.0155, "step": 12273 }, { "epoch": 1.330659150043365, "grad_norm": 0.030754229053854942, "learning_rate": 5.5846510953140875e-05, "loss": 0.0014, "step": 12274 }, { "epoch": 1.3307675628794449, "grad_norm": 0.10023590177297592, "learning_rate": 5.584288408530394e-05, "loss": 0.0054, "step": 12275 }, { "epoch": 1.3308759757155246, "grad_norm": 0.45701393485069275, "learning_rate": 5.5839257217467e-05, "loss": 0.0405, "step": 12276 }, { "epoch": 1.3309843885516046, "grad_norm": 0.199334055185318, "learning_rate": 5.583563034963006e-05, "loss": 0.0115, "step": 12277 }, { "epoch": 1.3310928013876844, "grad_norm": 0.4916996359825134, "learning_rate": 5.5832003481793125e-05, "loss": 0.0512, "step": 12278 }, { "epoch": 1.3312012142237641, "grad_norm": 0.8321148753166199, "learning_rate": 5.582837661395619e-05, "loss": 0.0157, "step": 12279 }, { "epoch": 1.331309627059844, "grad_norm": 0.21061444282531738, "learning_rate": 5.582474974611925e-05, "loss": 0.0253, "step": 12280 }, { "epoch": 1.3314180398959237, "grad_norm": 0.029191339388489723, "learning_rate": 5.582112287828232e-05, "loss": 0.0013, "step": 12281 }, { "epoch": 1.3315264527320034, "grad_norm": 0.23545239865779877, "learning_rate": 5.581749601044538e-05, "loss": 0.0073, "step": 12282 }, { "epoch": 1.3316348655680832, "grad_norm": 0.12367606163024902, "learning_rate": 5.5813869142608445e-05, "loss": 0.0083, "step": 12283 }, { "epoch": 1.3317432784041632, "grad_norm": 0.22336514294147491, "learning_rate": 5.581024227477151e-05, "loss": 0.0586, "step": 12284 }, { "epoch": 1.3318516912402427, "grad_norm": 0.07436590641736984, "learning_rate": 5.5806615406934573e-05, "loss": 0.0091, "step": 12285 }, { "epoch": 1.3319601040763227, "grad_norm": 0.3344734013080597, "learning_rate": 5.580298853909763e-05, "loss": 0.0173, "step": 12286 }, { "epoch": 1.3320685169124025, "grad_norm": 0.6292176842689514, "learning_rate": 5.5799361671260695e-05, "loss": 0.0218, "step": 12287 }, { "epoch": 1.3321769297484822, "grad_norm": 0.49227023124694824, "learning_rate": 5.579573480342377e-05, "loss": 0.0378, "step": 12288 }, { "epoch": 1.332285342584562, "grad_norm": 0.3947909474372864, "learning_rate": 5.579210793558684e-05, "loss": 0.0539, "step": 12289 }, { "epoch": 1.3323937554206418, "grad_norm": 0.9667146801948547, "learning_rate": 5.57884810677499e-05, "loss": 0.029, "step": 12290 }, { "epoch": 1.3325021682567215, "grad_norm": 0.19203123450279236, "learning_rate": 5.5784854199912965e-05, "loss": 0.0257, "step": 12291 }, { "epoch": 1.3326105810928013, "grad_norm": 0.3206901550292969, "learning_rate": 5.578122733207602e-05, "loss": 0.0271, "step": 12292 }, { "epoch": 1.3327189939288813, "grad_norm": 0.4012281894683838, "learning_rate": 5.5777600464239086e-05, "loss": 0.0134, "step": 12293 }, { "epoch": 1.332827406764961, "grad_norm": 0.40401676297187805, "learning_rate": 5.577397359640215e-05, "loss": 0.03, "step": 12294 }, { "epoch": 1.3329358196010408, "grad_norm": 0.6122289896011353, "learning_rate": 5.5770346728565214e-05, "loss": 0.0166, "step": 12295 }, { "epoch": 1.3330442324371206, "grad_norm": 0.5110373497009277, "learning_rate": 5.576671986072828e-05, "loss": 0.0542, "step": 12296 }, { "epoch": 1.3331526452732003, "grad_norm": 0.39445507526397705, "learning_rate": 5.576309299289134e-05, "loss": 0.0146, "step": 12297 }, { "epoch": 1.33326105810928, "grad_norm": 0.11326418817043304, "learning_rate": 5.575946612505441e-05, "loss": 0.0077, "step": 12298 }, { "epoch": 1.3333694709453598, "grad_norm": 0.21495197713375092, "learning_rate": 5.575583925721747e-05, "loss": 0.0187, "step": 12299 }, { "epoch": 1.3334778837814398, "grad_norm": 0.6151700019836426, "learning_rate": 5.575221238938053e-05, "loss": 0.0333, "step": 12300 }, { "epoch": 1.3335862966175196, "grad_norm": 0.4661928415298462, "learning_rate": 5.574858552154359e-05, "loss": 0.0118, "step": 12301 }, { "epoch": 1.3336947094535994, "grad_norm": 0.17695708572864532, "learning_rate": 5.5744958653706656e-05, "loss": 0.0149, "step": 12302 }, { "epoch": 1.3338031222896791, "grad_norm": 0.5041819214820862, "learning_rate": 5.574133178586972e-05, "loss": 0.0289, "step": 12303 }, { "epoch": 1.3339115351257589, "grad_norm": 0.8376300930976868, "learning_rate": 5.5737704918032785e-05, "loss": 0.01, "step": 12304 }, { "epoch": 1.3340199479618386, "grad_norm": 0.13254904747009277, "learning_rate": 5.573407805019585e-05, "loss": 0.0054, "step": 12305 }, { "epoch": 1.3341283607979184, "grad_norm": 0.10810156911611557, "learning_rate": 5.573045118235892e-05, "loss": 0.0052, "step": 12306 }, { "epoch": 1.3342367736339984, "grad_norm": 0.19752717018127441, "learning_rate": 5.5726824314521984e-05, "loss": 0.008, "step": 12307 }, { "epoch": 1.334345186470078, "grad_norm": 0.31177836656570435, "learning_rate": 5.572319744668505e-05, "loss": 0.0072, "step": 12308 }, { "epoch": 1.334453599306158, "grad_norm": 0.22415538132190704, "learning_rate": 5.571957057884811e-05, "loss": 0.014, "step": 12309 }, { "epoch": 1.3345620121422377, "grad_norm": 0.4987519383430481, "learning_rate": 5.5715943711011176e-05, "loss": 0.048, "step": 12310 }, { "epoch": 1.3346704249783174, "grad_norm": 0.15834780037403107, "learning_rate": 5.571231684317424e-05, "loss": 0.0131, "step": 12311 }, { "epoch": 1.3347788378143972, "grad_norm": 0.013167419470846653, "learning_rate": 5.5708689975337304e-05, "loss": 0.0009, "step": 12312 }, { "epoch": 1.334887250650477, "grad_norm": 0.15469077229499817, "learning_rate": 5.570506310750037e-05, "loss": 0.0156, "step": 12313 }, { "epoch": 1.3349956634865567, "grad_norm": 0.17163847386837006, "learning_rate": 5.570143623966343e-05, "loss": 0.012, "step": 12314 }, { "epoch": 1.3351040763226365, "grad_norm": 0.35372892022132874, "learning_rate": 5.569780937182649e-05, "loss": 0.007, "step": 12315 }, { "epoch": 1.3352124891587165, "grad_norm": 0.06928899139165878, "learning_rate": 5.5694182503989554e-05, "loss": 0.005, "step": 12316 }, { "epoch": 1.3353209019947962, "grad_norm": 0.8321090340614319, "learning_rate": 5.569055563615262e-05, "loss": 0.0263, "step": 12317 }, { "epoch": 1.335429314830876, "grad_norm": 0.964189887046814, "learning_rate": 5.568692876831568e-05, "loss": 0.0452, "step": 12318 }, { "epoch": 1.3355377276669558, "grad_norm": 0.4065014719963074, "learning_rate": 5.5683301900478746e-05, "loss": 0.0274, "step": 12319 }, { "epoch": 1.3356461405030355, "grad_norm": 0.27677658200263977, "learning_rate": 5.567967503264181e-05, "loss": 0.0214, "step": 12320 }, { "epoch": 1.3357545533391153, "grad_norm": 0.15416161715984344, "learning_rate": 5.5676048164804874e-05, "loss": 0.012, "step": 12321 }, { "epoch": 1.335862966175195, "grad_norm": 0.4947289824485779, "learning_rate": 5.567242129696794e-05, "loss": 0.0189, "step": 12322 }, { "epoch": 1.335971379011275, "grad_norm": 0.35326629877090454, "learning_rate": 5.566879442913101e-05, "loss": 0.0153, "step": 12323 }, { "epoch": 1.3360797918473546, "grad_norm": 0.5963765382766724, "learning_rate": 5.5665167561294074e-05, "loss": 0.0319, "step": 12324 }, { "epoch": 1.3361882046834346, "grad_norm": 0.18335101008415222, "learning_rate": 5.566154069345714e-05, "loss": 0.0154, "step": 12325 }, { "epoch": 1.3362966175195143, "grad_norm": 0.47681498527526855, "learning_rate": 5.56579138256202e-05, "loss": 0.0382, "step": 12326 }, { "epoch": 1.336405030355594, "grad_norm": 0.25739917159080505, "learning_rate": 5.5654286957783266e-05, "loss": 0.0286, "step": 12327 }, { "epoch": 1.3365134431916739, "grad_norm": 0.4600265324115753, "learning_rate": 5.565066008994633e-05, "loss": 0.0222, "step": 12328 }, { "epoch": 1.3366218560277536, "grad_norm": 0.17410656809806824, "learning_rate": 5.564703322210939e-05, "loss": 0.0086, "step": 12329 }, { "epoch": 1.3367302688638336, "grad_norm": 0.2959718704223633, "learning_rate": 5.564340635427245e-05, "loss": 0.0193, "step": 12330 }, { "epoch": 1.3368386816999132, "grad_norm": 0.19672247767448425, "learning_rate": 5.5639779486435516e-05, "loss": 0.0134, "step": 12331 }, { "epoch": 1.3369470945359931, "grad_norm": 0.48086169362068176, "learning_rate": 5.563615261859858e-05, "loss": 0.0677, "step": 12332 }, { "epoch": 1.337055507372073, "grad_norm": 0.4379138946533203, "learning_rate": 5.5632525750761644e-05, "loss": 0.0214, "step": 12333 }, { "epoch": 1.3371639202081527, "grad_norm": 0.015324811451137066, "learning_rate": 5.562889888292471e-05, "loss": 0.0008, "step": 12334 }, { "epoch": 1.3372723330442324, "grad_norm": 0.24458374083042145, "learning_rate": 5.562527201508777e-05, "loss": 0.0122, "step": 12335 }, { "epoch": 1.3373807458803122, "grad_norm": 0.4374563694000244, "learning_rate": 5.5621645147250836e-05, "loss": 0.0349, "step": 12336 }, { "epoch": 1.337489158716392, "grad_norm": 0.16035395860671997, "learning_rate": 5.56180182794139e-05, "loss": 0.0084, "step": 12337 }, { "epoch": 1.3375975715524717, "grad_norm": 0.3839770257472992, "learning_rate": 5.561439141157696e-05, "loss": 0.0171, "step": 12338 }, { "epoch": 1.3377059843885517, "grad_norm": 0.0684988871216774, "learning_rate": 5.561076454374002e-05, "loss": 0.0056, "step": 12339 }, { "epoch": 1.3378143972246315, "grad_norm": 0.24364754557609558, "learning_rate": 5.56071376759031e-05, "loss": 0.0159, "step": 12340 }, { "epoch": 1.3379228100607112, "grad_norm": 1.4246294498443604, "learning_rate": 5.560351080806616e-05, "loss": 0.0697, "step": 12341 }, { "epoch": 1.338031222896791, "grad_norm": 0.27907776832580566, "learning_rate": 5.559988394022923e-05, "loss": 0.0285, "step": 12342 }, { "epoch": 1.3381396357328708, "grad_norm": 0.029011966660618782, "learning_rate": 5.559625707239229e-05, "loss": 0.0013, "step": 12343 }, { "epoch": 1.3382480485689505, "grad_norm": 0.2078869640827179, "learning_rate": 5.559263020455535e-05, "loss": 0.0099, "step": 12344 }, { "epoch": 1.3383564614050303, "grad_norm": 0.7112027406692505, "learning_rate": 5.558900333671841e-05, "loss": 0.0236, "step": 12345 }, { "epoch": 1.3384648742411103, "grad_norm": 0.0894809141755104, "learning_rate": 5.558537646888148e-05, "loss": 0.004, "step": 12346 }, { "epoch": 1.3385732870771898, "grad_norm": 0.5264484286308289, "learning_rate": 5.558174960104454e-05, "loss": 0.0178, "step": 12347 }, { "epoch": 1.3386816999132698, "grad_norm": 0.12318727374076843, "learning_rate": 5.5578122733207605e-05, "loss": 0.0076, "step": 12348 }, { "epoch": 1.3387901127493496, "grad_norm": 0.20601016283035278, "learning_rate": 5.557449586537067e-05, "loss": 0.0129, "step": 12349 }, { "epoch": 1.3388985255854293, "grad_norm": 0.15394707024097443, "learning_rate": 5.5570868997533734e-05, "loss": 0.008, "step": 12350 }, { "epoch": 1.339006938421509, "grad_norm": 0.204302117228508, "learning_rate": 5.55672421296968e-05, "loss": 0.0167, "step": 12351 }, { "epoch": 1.3391153512575888, "grad_norm": 0.162128284573555, "learning_rate": 5.5563615261859855e-05, "loss": 0.0061, "step": 12352 }, { "epoch": 1.3392237640936686, "grad_norm": 0.1865505427122116, "learning_rate": 5.555998839402292e-05, "loss": 0.0078, "step": 12353 }, { "epoch": 1.3393321769297484, "grad_norm": 0.18190063536167145, "learning_rate": 5.555636152618598e-05, "loss": 0.0106, "step": 12354 }, { "epoch": 1.3394405897658284, "grad_norm": 0.27976980805397034, "learning_rate": 5.555273465834905e-05, "loss": 0.0183, "step": 12355 }, { "epoch": 1.3395490026019081, "grad_norm": 0.6187976598739624, "learning_rate": 5.554910779051211e-05, "loss": 0.0268, "step": 12356 }, { "epoch": 1.3396574154379879, "grad_norm": 0.05846253037452698, "learning_rate": 5.5545480922675175e-05, "loss": 0.0027, "step": 12357 }, { "epoch": 1.3397658282740676, "grad_norm": 0.1433660238981247, "learning_rate": 5.5541854054838246e-05, "loss": 0.0075, "step": 12358 }, { "epoch": 1.3398742411101474, "grad_norm": 0.4401009678840637, "learning_rate": 5.553822718700131e-05, "loss": 0.0308, "step": 12359 }, { "epoch": 1.3399826539462272, "grad_norm": 0.25359684228897095, "learning_rate": 5.5534600319164375e-05, "loss": 0.0104, "step": 12360 }, { "epoch": 1.340091066782307, "grad_norm": 0.5654672384262085, "learning_rate": 5.553097345132744e-05, "loss": 0.0436, "step": 12361 }, { "epoch": 1.340199479618387, "grad_norm": 0.06453854590654373, "learning_rate": 5.55273465834905e-05, "loss": 0.0028, "step": 12362 }, { "epoch": 1.3403078924544667, "grad_norm": 0.19060026109218597, "learning_rate": 5.552371971565357e-05, "loss": 0.0072, "step": 12363 }, { "epoch": 1.3404163052905465, "grad_norm": 0.2594343423843384, "learning_rate": 5.552009284781663e-05, "loss": 0.0067, "step": 12364 }, { "epoch": 1.3405247181266262, "grad_norm": 0.12507836520671844, "learning_rate": 5.5516465979979695e-05, "loss": 0.0072, "step": 12365 }, { "epoch": 1.340633130962706, "grad_norm": 0.1891610324382782, "learning_rate": 5.551283911214276e-05, "loss": 0.0071, "step": 12366 }, { "epoch": 1.3407415437987857, "grad_norm": 0.23271434009075165, "learning_rate": 5.5509212244305817e-05, "loss": 0.019, "step": 12367 }, { "epoch": 1.3408499566348655, "grad_norm": 0.25281479954719543, "learning_rate": 5.550558537646888e-05, "loss": 0.0099, "step": 12368 }, { "epoch": 1.3409583694709455, "grad_norm": 0.306041419506073, "learning_rate": 5.5501958508631945e-05, "loss": 0.022, "step": 12369 }, { "epoch": 1.341066782307025, "grad_norm": 0.3850405812263489, "learning_rate": 5.549833164079501e-05, "loss": 0.0399, "step": 12370 }, { "epoch": 1.341175195143105, "grad_norm": 0.6003890037536621, "learning_rate": 5.549470477295807e-05, "loss": 0.0295, "step": 12371 }, { "epoch": 1.3412836079791848, "grad_norm": 0.1616150289773941, "learning_rate": 5.549107790512114e-05, "loss": 0.0124, "step": 12372 }, { "epoch": 1.3413920208152645, "grad_norm": 0.35528048872947693, "learning_rate": 5.54874510372842e-05, "loss": 0.0152, "step": 12373 }, { "epoch": 1.3415004336513443, "grad_norm": 0.23277084529399872, "learning_rate": 5.5483824169447265e-05, "loss": 0.014, "step": 12374 }, { "epoch": 1.341608846487424, "grad_norm": 0.21705827116966248, "learning_rate": 5.5480197301610336e-05, "loss": 0.0076, "step": 12375 }, { "epoch": 1.3417172593235038, "grad_norm": 0.6336045861244202, "learning_rate": 5.54765704337734e-05, "loss": 0.0114, "step": 12376 }, { "epoch": 1.3418256721595836, "grad_norm": 0.11276619881391525, "learning_rate": 5.5472943565936464e-05, "loss": 0.0042, "step": 12377 }, { "epoch": 1.3419340849956636, "grad_norm": 0.6215299367904663, "learning_rate": 5.546931669809953e-05, "loss": 0.0105, "step": 12378 }, { "epoch": 1.3420424978317433, "grad_norm": 0.0817003846168518, "learning_rate": 5.546568983026259e-05, "loss": 0.0026, "step": 12379 }, { "epoch": 1.342150910667823, "grad_norm": 0.48930031061172485, "learning_rate": 5.546206296242566e-05, "loss": 0.0358, "step": 12380 }, { "epoch": 1.3422593235039029, "grad_norm": 0.6975247859954834, "learning_rate": 5.5458436094588714e-05, "loss": 0.0309, "step": 12381 }, { "epoch": 1.3423677363399826, "grad_norm": 0.36070993542671204, "learning_rate": 5.545480922675178e-05, "loss": 0.0085, "step": 12382 }, { "epoch": 1.3424761491760624, "grad_norm": 0.23960629105567932, "learning_rate": 5.545118235891484e-05, "loss": 0.0031, "step": 12383 }, { "epoch": 1.3425845620121422, "grad_norm": 0.31347721815109253, "learning_rate": 5.5447555491077906e-05, "loss": 0.0353, "step": 12384 }, { "epoch": 1.3426929748482221, "grad_norm": 0.11428509652614594, "learning_rate": 5.544392862324097e-05, "loss": 0.0113, "step": 12385 }, { "epoch": 1.3428013876843017, "grad_norm": 0.5858684182167053, "learning_rate": 5.5440301755404035e-05, "loss": 0.0275, "step": 12386 }, { "epoch": 1.3429098005203817, "grad_norm": 0.6255422830581665, "learning_rate": 5.54366748875671e-05, "loss": 0.0837, "step": 12387 }, { "epoch": 1.3430182133564614, "grad_norm": 0.7764289975166321, "learning_rate": 5.543304801973016e-05, "loss": 0.0307, "step": 12388 }, { "epoch": 1.3431266261925412, "grad_norm": 0.9587565064430237, "learning_rate": 5.542942115189323e-05, "loss": 0.0445, "step": 12389 }, { "epoch": 1.343235039028621, "grad_norm": 0.34400424361228943, "learning_rate": 5.5425794284056284e-05, "loss": 0.0385, "step": 12390 }, { "epoch": 1.3433434518647007, "grad_norm": 0.16012422740459442, "learning_rate": 5.542216741621935e-05, "loss": 0.0044, "step": 12391 }, { "epoch": 1.3434518647007807, "grad_norm": 0.3109368681907654, "learning_rate": 5.5418540548382426e-05, "loss": 0.026, "step": 12392 }, { "epoch": 1.3435602775368602, "grad_norm": 0.34911537170410156, "learning_rate": 5.541491368054549e-05, "loss": 0.0185, "step": 12393 }, { "epoch": 1.3436686903729402, "grad_norm": 0.4827744662761688, "learning_rate": 5.5411286812708554e-05, "loss": 0.0372, "step": 12394 }, { "epoch": 1.34377710320902, "grad_norm": 0.5379089117050171, "learning_rate": 5.540765994487162e-05, "loss": 0.0507, "step": 12395 }, { "epoch": 1.3438855160450998, "grad_norm": 0.10294961929321289, "learning_rate": 5.5404033077034676e-05, "loss": 0.0056, "step": 12396 }, { "epoch": 1.3439939288811795, "grad_norm": 0.6556851267814636, "learning_rate": 5.540040620919774e-05, "loss": 0.0273, "step": 12397 }, { "epoch": 1.3441023417172593, "grad_norm": 0.05823339521884918, "learning_rate": 5.5396779341360804e-05, "loss": 0.0035, "step": 12398 }, { "epoch": 1.344210754553339, "grad_norm": 0.22032302618026733, "learning_rate": 5.539315247352387e-05, "loss": 0.02, "step": 12399 }, { "epoch": 1.3443191673894188, "grad_norm": 0.43948888778686523, "learning_rate": 5.538952560568693e-05, "loss": 0.0219, "step": 12400 }, { "epoch": 1.3444275802254988, "grad_norm": 0.09448543936014175, "learning_rate": 5.5385898737849996e-05, "loss": 0.011, "step": 12401 }, { "epoch": 1.3445359930615786, "grad_norm": 0.20985189080238342, "learning_rate": 5.538227187001306e-05, "loss": 0.015, "step": 12402 }, { "epoch": 1.3446444058976583, "grad_norm": 0.28723543882369995, "learning_rate": 5.5378645002176124e-05, "loss": 0.0125, "step": 12403 }, { "epoch": 1.344752818733738, "grad_norm": 0.16277971863746643, "learning_rate": 5.537501813433918e-05, "loss": 0.0097, "step": 12404 }, { "epoch": 1.3448612315698178, "grad_norm": 0.15178345143795013, "learning_rate": 5.5371391266502246e-05, "loss": 0.0039, "step": 12405 }, { "epoch": 1.3449696444058976, "grad_norm": 0.08366576582193375, "learning_rate": 5.536776439866531e-05, "loss": 0.0043, "step": 12406 }, { "epoch": 1.3450780572419774, "grad_norm": 0.5469610691070557, "learning_rate": 5.5364137530828374e-05, "loss": 0.019, "step": 12407 }, { "epoch": 1.3451864700780574, "grad_norm": 0.4275522232055664, "learning_rate": 5.536051066299144e-05, "loss": 0.0469, "step": 12408 }, { "epoch": 1.345294882914137, "grad_norm": 0.6442441940307617, "learning_rate": 5.5356883795154516e-05, "loss": 0.0303, "step": 12409 }, { "epoch": 1.3454032957502169, "grad_norm": 0.1776430457830429, "learning_rate": 5.535325692731757e-05, "loss": 0.0076, "step": 12410 }, { "epoch": 1.3455117085862967, "grad_norm": 1.6227015256881714, "learning_rate": 5.534963005948064e-05, "loss": 0.0554, "step": 12411 }, { "epoch": 1.3456201214223764, "grad_norm": 0.3831271231174469, "learning_rate": 5.53460031916437e-05, "loss": 0.0101, "step": 12412 }, { "epoch": 1.3457285342584562, "grad_norm": 1.1107231378555298, "learning_rate": 5.5342376323806765e-05, "loss": 0.0321, "step": 12413 }, { "epoch": 1.345836947094536, "grad_norm": 0.2673652172088623, "learning_rate": 5.533874945596983e-05, "loss": 0.0066, "step": 12414 }, { "epoch": 1.3459453599306157, "grad_norm": 0.4067522883415222, "learning_rate": 5.5335122588132894e-05, "loss": 0.0196, "step": 12415 }, { "epoch": 1.3460537727666955, "grad_norm": 0.032660793513059616, "learning_rate": 5.533149572029596e-05, "loss": 0.0037, "step": 12416 }, { "epoch": 1.3461621856027755, "grad_norm": 0.35410386323928833, "learning_rate": 5.532786885245902e-05, "loss": 0.0099, "step": 12417 }, { "epoch": 1.3462705984388552, "grad_norm": 0.49004456400871277, "learning_rate": 5.5324241984622086e-05, "loss": 0.0714, "step": 12418 }, { "epoch": 1.346379011274935, "grad_norm": 0.3958708345890045, "learning_rate": 5.532061511678514e-05, "loss": 0.0059, "step": 12419 }, { "epoch": 1.3464874241110147, "grad_norm": 0.35186731815338135, "learning_rate": 5.531698824894821e-05, "loss": 0.0129, "step": 12420 }, { "epoch": 1.3465958369470945, "grad_norm": 0.6206365823745728, "learning_rate": 5.531336138111127e-05, "loss": 0.0104, "step": 12421 }, { "epoch": 1.3467042497831743, "grad_norm": 0.4017283618450165, "learning_rate": 5.5309734513274336e-05, "loss": 0.0274, "step": 12422 }, { "epoch": 1.346812662619254, "grad_norm": 0.5993649363517761, "learning_rate": 5.53061076454374e-05, "loss": 0.0405, "step": 12423 }, { "epoch": 1.346921075455334, "grad_norm": 0.19629883766174316, "learning_rate": 5.5302480777600464e-05, "loss": 0.0087, "step": 12424 }, { "epoch": 1.3470294882914138, "grad_norm": 0.3056710958480835, "learning_rate": 5.529885390976353e-05, "loss": 0.0248, "step": 12425 }, { "epoch": 1.3471379011274935, "grad_norm": 0.4378342032432556, "learning_rate": 5.529522704192659e-05, "loss": 0.0254, "step": 12426 }, { "epoch": 1.3472463139635733, "grad_norm": 0.17201049625873566, "learning_rate": 5.529160017408966e-05, "loss": 0.0137, "step": 12427 }, { "epoch": 1.347354726799653, "grad_norm": 0.7695161700248718, "learning_rate": 5.528797330625273e-05, "loss": 0.0583, "step": 12428 }, { "epoch": 1.3474631396357328, "grad_norm": 0.5507811903953552, "learning_rate": 5.528434643841579e-05, "loss": 0.0148, "step": 12429 }, { "epoch": 1.3475715524718126, "grad_norm": 0.07083623111248016, "learning_rate": 5.5280719570578855e-05, "loss": 0.0071, "step": 12430 }, { "epoch": 1.3476799653078926, "grad_norm": 0.25317201018333435, "learning_rate": 5.527709270274192e-05, "loss": 0.0235, "step": 12431 }, { "epoch": 1.3477883781439721, "grad_norm": 0.3081693649291992, "learning_rate": 5.527346583490498e-05, "loss": 0.0151, "step": 12432 }, { "epoch": 1.347896790980052, "grad_norm": 0.29779917001724243, "learning_rate": 5.526983896706804e-05, "loss": 0.0197, "step": 12433 }, { "epoch": 1.3480052038161319, "grad_norm": 1.114243745803833, "learning_rate": 5.5266212099231105e-05, "loss": 0.0642, "step": 12434 }, { "epoch": 1.3481136166522116, "grad_norm": 0.7071415185928345, "learning_rate": 5.526258523139417e-05, "loss": 0.0473, "step": 12435 }, { "epoch": 1.3482220294882914, "grad_norm": 0.4127792716026306, "learning_rate": 5.525895836355723e-05, "loss": 0.0276, "step": 12436 }, { "epoch": 1.3483304423243712, "grad_norm": 0.12889894843101501, "learning_rate": 5.52553314957203e-05, "loss": 0.0086, "step": 12437 }, { "epoch": 1.348438855160451, "grad_norm": 0.5517650246620178, "learning_rate": 5.525170462788336e-05, "loss": 0.053, "step": 12438 }, { "epoch": 1.3485472679965307, "grad_norm": 0.1088985949754715, "learning_rate": 5.5248077760046425e-05, "loss": 0.0069, "step": 12439 }, { "epoch": 1.3486556808326107, "grad_norm": 0.4380507171154022, "learning_rate": 5.524445089220949e-05, "loss": 0.0385, "step": 12440 }, { "epoch": 1.3487640936686904, "grad_norm": 0.37989336252212524, "learning_rate": 5.5240824024372554e-05, "loss": 0.0268, "step": 12441 }, { "epoch": 1.3488725065047702, "grad_norm": 0.37170523405075073, "learning_rate": 5.523719715653561e-05, "loss": 0.0282, "step": 12442 }, { "epoch": 1.34898091934085, "grad_norm": 0.3668992817401886, "learning_rate": 5.5233570288698675e-05, "loss": 0.022, "step": 12443 }, { "epoch": 1.3490893321769297, "grad_norm": 0.10106629133224487, "learning_rate": 5.522994342086175e-05, "loss": 0.0047, "step": 12444 }, { "epoch": 1.3491977450130095, "grad_norm": 0.12490405887365341, "learning_rate": 5.522631655302482e-05, "loss": 0.0041, "step": 12445 }, { "epoch": 1.3493061578490892, "grad_norm": 0.1418175995349884, "learning_rate": 5.522268968518788e-05, "loss": 0.0104, "step": 12446 }, { "epoch": 1.3494145706851692, "grad_norm": 0.30472609400749207, "learning_rate": 5.5219062817350945e-05, "loss": 0.0218, "step": 12447 }, { "epoch": 1.349522983521249, "grad_norm": 0.39322933554649353, "learning_rate": 5.5215435949514e-05, "loss": 0.0663, "step": 12448 }, { "epoch": 1.3496313963573288, "grad_norm": 0.2036990225315094, "learning_rate": 5.5211809081677066e-05, "loss": 0.0113, "step": 12449 }, { "epoch": 1.3497398091934085, "grad_norm": 0.23497207462787628, "learning_rate": 5.520818221384013e-05, "loss": 0.0052, "step": 12450 }, { "epoch": 1.3498482220294883, "grad_norm": 0.18112419545650482, "learning_rate": 5.5204555346003195e-05, "loss": 0.0125, "step": 12451 }, { "epoch": 1.349956634865568, "grad_norm": 0.3313930332660675, "learning_rate": 5.520092847816626e-05, "loss": 0.0139, "step": 12452 }, { "epoch": 1.3500650477016478, "grad_norm": 0.21937771141529083, "learning_rate": 5.519730161032932e-05, "loss": 0.017, "step": 12453 }, { "epoch": 1.3501734605377278, "grad_norm": 0.7464233636856079, "learning_rate": 5.519367474249239e-05, "loss": 0.0533, "step": 12454 }, { "epoch": 1.3502818733738073, "grad_norm": 0.3166821599006653, "learning_rate": 5.519004787465545e-05, "loss": 0.0232, "step": 12455 }, { "epoch": 1.3503902862098873, "grad_norm": 0.667069673538208, "learning_rate": 5.518642100681851e-05, "loss": 0.0275, "step": 12456 }, { "epoch": 1.350498699045967, "grad_norm": 0.09345302730798721, "learning_rate": 5.518279413898157e-05, "loss": 0.004, "step": 12457 }, { "epoch": 1.3506071118820469, "grad_norm": 0.42853304743766785, "learning_rate": 5.5179167271144637e-05, "loss": 0.0098, "step": 12458 }, { "epoch": 1.3507155247181266, "grad_norm": 0.3070312738418579, "learning_rate": 5.51755404033077e-05, "loss": 0.0421, "step": 12459 }, { "epoch": 1.3508239375542064, "grad_norm": 0.8648709654808044, "learning_rate": 5.5171913535470765e-05, "loss": 0.0409, "step": 12460 }, { "epoch": 1.3509323503902861, "grad_norm": 0.27136242389678955, "learning_rate": 5.516828666763384e-05, "loss": 0.0239, "step": 12461 }, { "epoch": 1.351040763226366, "grad_norm": 0.32923781871795654, "learning_rate": 5.51646597997969e-05, "loss": 0.0396, "step": 12462 }, { "epoch": 1.351149176062446, "grad_norm": 0.4761369228363037, "learning_rate": 5.5161032931959964e-05, "loss": 0.0379, "step": 12463 }, { "epoch": 1.3512575888985257, "grad_norm": 0.38243165612220764, "learning_rate": 5.515740606412303e-05, "loss": 0.02, "step": 12464 }, { "epoch": 1.3513660017346054, "grad_norm": 0.2909603714942932, "learning_rate": 5.515377919628609e-05, "loss": 0.0203, "step": 12465 }, { "epoch": 1.3514744145706852, "grad_norm": 0.08900156617164612, "learning_rate": 5.5150152328449156e-05, "loss": 0.0064, "step": 12466 }, { "epoch": 1.351582827406765, "grad_norm": 0.3788560628890991, "learning_rate": 5.514652546061222e-05, "loss": 0.0397, "step": 12467 }, { "epoch": 1.3516912402428447, "grad_norm": 0.30291548371315, "learning_rate": 5.5142898592775284e-05, "loss": 0.0126, "step": 12468 }, { "epoch": 1.3517996530789245, "grad_norm": 0.11590062826871872, "learning_rate": 5.513927172493835e-05, "loss": 0.0061, "step": 12469 }, { "epoch": 1.3519080659150045, "grad_norm": 0.22041618824005127, "learning_rate": 5.513564485710141e-05, "loss": 0.0279, "step": 12470 }, { "epoch": 1.352016478751084, "grad_norm": 0.782823383808136, "learning_rate": 5.513201798926447e-05, "loss": 0.056, "step": 12471 }, { "epoch": 1.352124891587164, "grad_norm": 0.6052126884460449, "learning_rate": 5.5128391121427534e-05, "loss": 0.0393, "step": 12472 }, { "epoch": 1.3522333044232437, "grad_norm": 0.288601279258728, "learning_rate": 5.51247642535906e-05, "loss": 0.0168, "step": 12473 }, { "epoch": 1.3523417172593235, "grad_norm": 0.3613607883453369, "learning_rate": 5.512113738575366e-05, "loss": 0.0327, "step": 12474 }, { "epoch": 1.3524501300954033, "grad_norm": 0.5617159008979797, "learning_rate": 5.5117510517916726e-05, "loss": 0.0258, "step": 12475 }, { "epoch": 1.352558542931483, "grad_norm": 0.28965649008750916, "learning_rate": 5.511388365007979e-05, "loss": 0.0236, "step": 12476 }, { "epoch": 1.352666955767563, "grad_norm": 0.2885388433933258, "learning_rate": 5.5110256782242855e-05, "loss": 0.0107, "step": 12477 }, { "epoch": 1.3527753686036426, "grad_norm": 0.18258138000965118, "learning_rate": 5.510662991440592e-05, "loss": 0.0068, "step": 12478 }, { "epoch": 1.3528837814397225, "grad_norm": 0.17379367351531982, "learning_rate": 5.510300304656899e-05, "loss": 0.0241, "step": 12479 }, { "epoch": 1.3529921942758023, "grad_norm": 0.34339573979377747, "learning_rate": 5.5099376178732054e-05, "loss": 0.031, "step": 12480 }, { "epoch": 1.353100607111882, "grad_norm": 0.04758976772427559, "learning_rate": 5.509574931089512e-05, "loss": 0.0027, "step": 12481 }, { "epoch": 1.3532090199479618, "grad_norm": 0.11340385675430298, "learning_rate": 5.509212244305818e-05, "loss": 0.008, "step": 12482 }, { "epoch": 1.3533174327840416, "grad_norm": 0.23846277594566345, "learning_rate": 5.5088495575221246e-05, "loss": 0.0141, "step": 12483 }, { "epoch": 1.3534258456201214, "grad_norm": 0.3256480097770691, "learning_rate": 5.508486870738431e-05, "loss": 0.019, "step": 12484 }, { "epoch": 1.3535342584562011, "grad_norm": 0.18228885531425476, "learning_rate": 5.508124183954737e-05, "loss": 0.0213, "step": 12485 }, { "epoch": 1.353642671292281, "grad_norm": 0.45395204424858093, "learning_rate": 5.507761497171043e-05, "loss": 0.0433, "step": 12486 }, { "epoch": 1.3537510841283609, "grad_norm": 0.4835364520549774, "learning_rate": 5.5073988103873496e-05, "loss": 0.0837, "step": 12487 }, { "epoch": 1.3538594969644406, "grad_norm": 0.21847115457057953, "learning_rate": 5.507036123603656e-05, "loss": 0.0116, "step": 12488 }, { "epoch": 1.3539679098005204, "grad_norm": 0.10886188596487045, "learning_rate": 5.5066734368199624e-05, "loss": 0.0044, "step": 12489 }, { "epoch": 1.3540763226366002, "grad_norm": 0.24576829373836517, "learning_rate": 5.506310750036269e-05, "loss": 0.0239, "step": 12490 }, { "epoch": 1.35418473547268, "grad_norm": 0.415452778339386, "learning_rate": 5.505948063252575e-05, "loss": 0.0214, "step": 12491 }, { "epoch": 1.3542931483087597, "grad_norm": 0.5650229454040527, "learning_rate": 5.5055853764688816e-05, "loss": 0.0879, "step": 12492 }, { "epoch": 1.3544015611448397, "grad_norm": 0.5214457511901855, "learning_rate": 5.505222689685188e-05, "loss": 0.0911, "step": 12493 }, { "epoch": 1.3545099739809192, "grad_norm": 0.39643335342407227, "learning_rate": 5.504860002901494e-05, "loss": 0.0164, "step": 12494 }, { "epoch": 1.3546183868169992, "grad_norm": 0.26735275983810425, "learning_rate": 5.5044973161178e-05, "loss": 0.0175, "step": 12495 }, { "epoch": 1.354726799653079, "grad_norm": 0.26134777069091797, "learning_rate": 5.504134629334108e-05, "loss": 0.0286, "step": 12496 }, { "epoch": 1.3548352124891587, "grad_norm": 0.46952328085899353, "learning_rate": 5.5037719425504143e-05, "loss": 0.0453, "step": 12497 }, { "epoch": 1.3549436253252385, "grad_norm": 0.2986573576927185, "learning_rate": 5.503409255766721e-05, "loss": 0.0141, "step": 12498 }, { "epoch": 1.3550520381613183, "grad_norm": 1.205327033996582, "learning_rate": 5.503046568983027e-05, "loss": 0.0399, "step": 12499 }, { "epoch": 1.355160450997398, "grad_norm": 0.31863704323768616, "learning_rate": 5.502683882199333e-05, "loss": 0.0309, "step": 12500 }, { "epoch": 1.3552688638334778, "grad_norm": 0.5786599516868591, "learning_rate": 5.502321195415639e-05, "loss": 0.055, "step": 12501 }, { "epoch": 1.3553772766695578, "grad_norm": 0.14520296454429626, "learning_rate": 5.501958508631946e-05, "loss": 0.0065, "step": 12502 }, { "epoch": 1.3554856895056375, "grad_norm": 0.29222413897514343, "learning_rate": 5.501595821848252e-05, "loss": 0.0136, "step": 12503 }, { "epoch": 1.3555941023417173, "grad_norm": 0.2949006259441376, "learning_rate": 5.5012331350645585e-05, "loss": 0.0203, "step": 12504 }, { "epoch": 1.355702515177797, "grad_norm": 0.18181587755680084, "learning_rate": 5.500870448280865e-05, "loss": 0.0107, "step": 12505 }, { "epoch": 1.3558109280138768, "grad_norm": 0.5150316953659058, "learning_rate": 5.5005077614971714e-05, "loss": 0.0735, "step": 12506 }, { "epoch": 1.3559193408499566, "grad_norm": 0.29633480310440063, "learning_rate": 5.500145074713478e-05, "loss": 0.0165, "step": 12507 }, { "epoch": 1.3560277536860363, "grad_norm": 0.3149401545524597, "learning_rate": 5.499782387929784e-05, "loss": 0.0405, "step": 12508 }, { "epoch": 1.3561361665221163, "grad_norm": 0.21140189468860626, "learning_rate": 5.49941970114609e-05, "loss": 0.0208, "step": 12509 }, { "epoch": 1.356244579358196, "grad_norm": 0.3886309862136841, "learning_rate": 5.499057014362396e-05, "loss": 0.0264, "step": 12510 }, { "epoch": 1.3563529921942759, "grad_norm": 0.3288435935974121, "learning_rate": 5.498694327578703e-05, "loss": 0.0405, "step": 12511 }, { "epoch": 1.3564614050303556, "grad_norm": 0.17697854340076447, "learning_rate": 5.498331640795009e-05, "loss": 0.0215, "step": 12512 }, { "epoch": 1.3565698178664354, "grad_norm": 0.2824038565158844, "learning_rate": 5.497968954011317e-05, "loss": 0.022, "step": 12513 }, { "epoch": 1.3566782307025151, "grad_norm": 0.23422794044017792, "learning_rate": 5.4976062672276226e-05, "loss": 0.0163, "step": 12514 }, { "epoch": 1.356786643538595, "grad_norm": 0.5698856711387634, "learning_rate": 5.497243580443929e-05, "loss": 0.0303, "step": 12515 }, { "epoch": 1.356895056374675, "grad_norm": 0.055158328264951706, "learning_rate": 5.4968808936602355e-05, "loss": 0.003, "step": 12516 }, { "epoch": 1.3570034692107544, "grad_norm": 0.08315016329288483, "learning_rate": 5.496518206876542e-05, "loss": 0.0072, "step": 12517 }, { "epoch": 1.3571118820468344, "grad_norm": 0.26076740026474, "learning_rate": 5.496155520092848e-05, "loss": 0.017, "step": 12518 }, { "epoch": 1.3572202948829142, "grad_norm": 0.12448345869779587, "learning_rate": 5.495792833309155e-05, "loss": 0.0117, "step": 12519 }, { "epoch": 1.357328707718994, "grad_norm": 0.12874028086662292, "learning_rate": 5.495430146525461e-05, "loss": 0.0057, "step": 12520 }, { "epoch": 1.3574371205550737, "grad_norm": 0.37189269065856934, "learning_rate": 5.4950674597417675e-05, "loss": 0.026, "step": 12521 }, { "epoch": 1.3575455333911535, "grad_norm": 0.12202882766723633, "learning_rate": 5.494704772958074e-05, "loss": 0.0132, "step": 12522 }, { "epoch": 1.3576539462272332, "grad_norm": 0.1690441370010376, "learning_rate": 5.4943420861743797e-05, "loss": 0.0068, "step": 12523 }, { "epoch": 1.357762359063313, "grad_norm": 0.11854757368564606, "learning_rate": 5.493979399390686e-05, "loss": 0.0058, "step": 12524 }, { "epoch": 1.357870771899393, "grad_norm": 0.17372089624404907, "learning_rate": 5.4936167126069925e-05, "loss": 0.0056, "step": 12525 }, { "epoch": 1.3579791847354727, "grad_norm": 0.5547513365745544, "learning_rate": 5.493254025823299e-05, "loss": 0.0371, "step": 12526 }, { "epoch": 1.3580875975715525, "grad_norm": 0.1344265192747116, "learning_rate": 5.492891339039605e-05, "loss": 0.009, "step": 12527 }, { "epoch": 1.3581960104076323, "grad_norm": 0.29275161027908325, "learning_rate": 5.492528652255912e-05, "loss": 0.023, "step": 12528 }, { "epoch": 1.358304423243712, "grad_norm": 0.9710490703582764, "learning_rate": 5.492165965472218e-05, "loss": 0.034, "step": 12529 }, { "epoch": 1.3584128360797918, "grad_norm": 0.18395093083381653, "learning_rate": 5.491803278688525e-05, "loss": 0.0088, "step": 12530 }, { "epoch": 1.3585212489158716, "grad_norm": 0.08844242244958878, "learning_rate": 5.4914405919048316e-05, "loss": 0.0082, "step": 12531 }, { "epoch": 1.3586296617519515, "grad_norm": 0.05854518339037895, "learning_rate": 5.491077905121138e-05, "loss": 0.0044, "step": 12532 }, { "epoch": 1.358738074588031, "grad_norm": 0.11878269910812378, "learning_rate": 5.4907152183374444e-05, "loss": 0.0072, "step": 12533 }, { "epoch": 1.358846487424111, "grad_norm": 0.5712646245956421, "learning_rate": 5.490352531553751e-05, "loss": 0.024, "step": 12534 }, { "epoch": 1.3589549002601908, "grad_norm": 0.16185356676578522, "learning_rate": 5.489989844770057e-05, "loss": 0.0292, "step": 12535 }, { "epoch": 1.3590633130962706, "grad_norm": 0.08747787028551102, "learning_rate": 5.489627157986364e-05, "loss": 0.0055, "step": 12536 }, { "epoch": 1.3591717259323504, "grad_norm": 0.6513453722000122, "learning_rate": 5.4892644712026694e-05, "loss": 0.0528, "step": 12537 }, { "epoch": 1.3592801387684301, "grad_norm": 0.2778436839580536, "learning_rate": 5.488901784418976e-05, "loss": 0.0301, "step": 12538 }, { "epoch": 1.35938855160451, "grad_norm": 0.3465445339679718, "learning_rate": 5.488539097635282e-05, "loss": 0.0265, "step": 12539 }, { "epoch": 1.3594969644405897, "grad_norm": 0.28212088346481323, "learning_rate": 5.4881764108515886e-05, "loss": 0.0338, "step": 12540 }, { "epoch": 1.3596053772766696, "grad_norm": 0.5102512836456299, "learning_rate": 5.487813724067895e-05, "loss": 0.0084, "step": 12541 }, { "epoch": 1.3597137901127494, "grad_norm": 0.5099467039108276, "learning_rate": 5.4874510372842015e-05, "loss": 0.028, "step": 12542 }, { "epoch": 1.3598222029488292, "grad_norm": 0.19952799379825592, "learning_rate": 5.487088350500508e-05, "loss": 0.0174, "step": 12543 }, { "epoch": 1.359930615784909, "grad_norm": 0.2581043243408203, "learning_rate": 5.486725663716814e-05, "loss": 0.0214, "step": 12544 }, { "epoch": 1.3600390286209887, "grad_norm": 0.16260214149951935, "learning_rate": 5.486362976933121e-05, "loss": 0.0174, "step": 12545 }, { "epoch": 1.3601474414570685, "grad_norm": 0.5279113054275513, "learning_rate": 5.4860002901494264e-05, "loss": 0.0257, "step": 12546 }, { "epoch": 1.3602558542931482, "grad_norm": 0.20032985508441925, "learning_rate": 5.485637603365733e-05, "loss": 0.0175, "step": 12547 }, { "epoch": 1.3603642671292282, "grad_norm": 0.27161306142807007, "learning_rate": 5.4852749165820406e-05, "loss": 0.0182, "step": 12548 }, { "epoch": 1.360472679965308, "grad_norm": 0.5226978659629822, "learning_rate": 5.484912229798347e-05, "loss": 0.0243, "step": 12549 }, { "epoch": 1.3605810928013877, "grad_norm": 0.17884451150894165, "learning_rate": 5.4845495430146534e-05, "loss": 0.0162, "step": 12550 }, { "epoch": 1.3606895056374675, "grad_norm": 0.12366237491369247, "learning_rate": 5.48418685623096e-05, "loss": 0.0112, "step": 12551 }, { "epoch": 1.3607979184735473, "grad_norm": 0.15593066811561584, "learning_rate": 5.4838241694472656e-05, "loss": 0.028, "step": 12552 }, { "epoch": 1.360906331309627, "grad_norm": 0.365204781293869, "learning_rate": 5.483461482663572e-05, "loss": 0.0326, "step": 12553 }, { "epoch": 1.3610147441457068, "grad_norm": 0.9474057555198669, "learning_rate": 5.4830987958798784e-05, "loss": 0.0412, "step": 12554 }, { "epoch": 1.3611231569817868, "grad_norm": 0.3453923463821411, "learning_rate": 5.482736109096185e-05, "loss": 0.0155, "step": 12555 }, { "epoch": 1.3612315698178663, "grad_norm": 0.15597587823867798, "learning_rate": 5.482373422312491e-05, "loss": 0.011, "step": 12556 }, { "epoch": 1.3613399826539463, "grad_norm": 0.19691558182239532, "learning_rate": 5.4820107355287976e-05, "loss": 0.0096, "step": 12557 }, { "epoch": 1.361448395490026, "grad_norm": 0.5900993943214417, "learning_rate": 5.481648048745104e-05, "loss": 0.0077, "step": 12558 }, { "epoch": 1.3615568083261058, "grad_norm": 0.2096913903951645, "learning_rate": 5.4812853619614104e-05, "loss": 0.0185, "step": 12559 }, { "epoch": 1.3616652211621856, "grad_norm": 0.23778769373893738, "learning_rate": 5.480922675177717e-05, "loss": 0.0184, "step": 12560 }, { "epoch": 1.3617736339982653, "grad_norm": 0.4645330309867859, "learning_rate": 5.4805599883940226e-05, "loss": 0.032, "step": 12561 }, { "epoch": 1.361882046834345, "grad_norm": 0.19298090040683746, "learning_rate": 5.480197301610329e-05, "loss": 0.0156, "step": 12562 }, { "epoch": 1.3619904596704249, "grad_norm": 0.43577277660369873, "learning_rate": 5.4798346148266354e-05, "loss": 0.0262, "step": 12563 }, { "epoch": 1.3620988725065049, "grad_norm": 0.0872277244925499, "learning_rate": 5.479471928042942e-05, "loss": 0.0116, "step": 12564 }, { "epoch": 1.3622072853425846, "grad_norm": 0.35998788475990295, "learning_rate": 5.4791092412592496e-05, "loss": 0.0301, "step": 12565 }, { "epoch": 1.3623156981786644, "grad_norm": 0.3861658275127411, "learning_rate": 5.478746554475555e-05, "loss": 0.0271, "step": 12566 }, { "epoch": 1.3624241110147441, "grad_norm": 0.36760830879211426, "learning_rate": 5.478383867691862e-05, "loss": 0.0215, "step": 12567 }, { "epoch": 1.362532523850824, "grad_norm": 0.26009857654571533, "learning_rate": 5.478021180908168e-05, "loss": 0.0122, "step": 12568 }, { "epoch": 1.3626409366869037, "grad_norm": 0.6441898941993713, "learning_rate": 5.4776584941244745e-05, "loss": 0.0387, "step": 12569 }, { "epoch": 1.3627493495229834, "grad_norm": 0.12586133182048798, "learning_rate": 5.477295807340781e-05, "loss": 0.0079, "step": 12570 }, { "epoch": 1.3628577623590634, "grad_norm": 0.2052619606256485, "learning_rate": 5.4769331205570874e-05, "loss": 0.0086, "step": 12571 }, { "epoch": 1.3629661751951432, "grad_norm": 2.1128344535827637, "learning_rate": 5.476570433773394e-05, "loss": 0.0217, "step": 12572 }, { "epoch": 1.363074588031223, "grad_norm": 0.23553214967250824, "learning_rate": 5.4762077469897e-05, "loss": 0.0127, "step": 12573 }, { "epoch": 1.3631830008673027, "grad_norm": 0.8264943957328796, "learning_rate": 5.4758450602060066e-05, "loss": 0.0395, "step": 12574 }, { "epoch": 1.3632914137033825, "grad_norm": 0.3518018126487732, "learning_rate": 5.475482373422312e-05, "loss": 0.013, "step": 12575 }, { "epoch": 1.3633998265394622, "grad_norm": 0.10238824784755707, "learning_rate": 5.475119686638619e-05, "loss": 0.003, "step": 12576 }, { "epoch": 1.363508239375542, "grad_norm": 0.8103001117706299, "learning_rate": 5.474756999854925e-05, "loss": 0.1098, "step": 12577 }, { "epoch": 1.363616652211622, "grad_norm": 0.16100060939788818, "learning_rate": 5.4743943130712316e-05, "loss": 0.0098, "step": 12578 }, { "epoch": 1.3637250650477015, "grad_norm": 0.4997235834598541, "learning_rate": 5.474031626287538e-05, "loss": 0.0386, "step": 12579 }, { "epoch": 1.3638334778837815, "grad_norm": 0.5443381667137146, "learning_rate": 5.4736689395038444e-05, "loss": 0.0599, "step": 12580 }, { "epoch": 1.3639418907198613, "grad_norm": 0.5524847507476807, "learning_rate": 5.473306252720151e-05, "loss": 0.0264, "step": 12581 }, { "epoch": 1.364050303555941, "grad_norm": 0.21141646802425385, "learning_rate": 5.472943565936458e-05, "loss": 0.0098, "step": 12582 }, { "epoch": 1.3641587163920208, "grad_norm": 0.16842259466648102, "learning_rate": 5.472580879152764e-05, "loss": 0.0095, "step": 12583 }, { "epoch": 1.3642671292281006, "grad_norm": 0.06647693365812302, "learning_rate": 5.472218192369071e-05, "loss": 0.0029, "step": 12584 }, { "epoch": 1.3643755420641803, "grad_norm": 0.1942746490240097, "learning_rate": 5.471855505585377e-05, "loss": 0.0332, "step": 12585 }, { "epoch": 1.36448395490026, "grad_norm": 0.08723518252372742, "learning_rate": 5.4714928188016835e-05, "loss": 0.0091, "step": 12586 }, { "epoch": 1.36459236773634, "grad_norm": 0.25404974818229675, "learning_rate": 5.47113013201799e-05, "loss": 0.0215, "step": 12587 }, { "epoch": 1.3647007805724198, "grad_norm": 0.4029097557067871, "learning_rate": 5.4707674452342963e-05, "loss": 0.0161, "step": 12588 }, { "epoch": 1.3648091934084996, "grad_norm": 0.5869114398956299, "learning_rate": 5.470404758450602e-05, "loss": 0.0487, "step": 12589 }, { "epoch": 1.3649176062445794, "grad_norm": 0.14492103457450867, "learning_rate": 5.4700420716669085e-05, "loss": 0.0121, "step": 12590 }, { "epoch": 1.3650260190806591, "grad_norm": 0.43324699997901917, "learning_rate": 5.469679384883215e-05, "loss": 0.0208, "step": 12591 }, { "epoch": 1.3651344319167389, "grad_norm": 0.1785416305065155, "learning_rate": 5.469316698099521e-05, "loss": 0.0108, "step": 12592 }, { "epoch": 1.3652428447528187, "grad_norm": 0.38583484292030334, "learning_rate": 5.468954011315828e-05, "loss": 0.0268, "step": 12593 }, { "epoch": 1.3653512575888986, "grad_norm": 0.31643161177635193, "learning_rate": 5.468591324532134e-05, "loss": 0.0173, "step": 12594 }, { "epoch": 1.3654596704249784, "grad_norm": 0.5002221465110779, "learning_rate": 5.4682286377484405e-05, "loss": 0.0331, "step": 12595 }, { "epoch": 1.3655680832610582, "grad_norm": 0.5415003895759583, "learning_rate": 5.467865950964747e-05, "loss": 0.0446, "step": 12596 }, { "epoch": 1.365676496097138, "grad_norm": 0.22900083661079407, "learning_rate": 5.4675032641810534e-05, "loss": 0.0207, "step": 12597 }, { "epoch": 1.3657849089332177, "grad_norm": 0.02090390771627426, "learning_rate": 5.467140577397359e-05, "loss": 0.0009, "step": 12598 }, { "epoch": 1.3658933217692975, "grad_norm": 0.24325236678123474, "learning_rate": 5.466777890613667e-05, "loss": 0.0172, "step": 12599 }, { "epoch": 1.3660017346053772, "grad_norm": 0.3688880503177643, "learning_rate": 5.466415203829973e-05, "loss": 0.0192, "step": 12600 }, { "epoch": 1.3661101474414572, "grad_norm": 0.21466289460659027, "learning_rate": 5.46605251704628e-05, "loss": 0.0112, "step": 12601 }, { "epoch": 1.3662185602775367, "grad_norm": 0.06644421815872192, "learning_rate": 5.465689830262586e-05, "loss": 0.0041, "step": 12602 }, { "epoch": 1.3663269731136167, "grad_norm": 0.23471905291080475, "learning_rate": 5.4653271434788925e-05, "loss": 0.0256, "step": 12603 }, { "epoch": 1.3664353859496965, "grad_norm": 0.22335122525691986, "learning_rate": 5.464964456695198e-05, "loss": 0.0107, "step": 12604 }, { "epoch": 1.3665437987857763, "grad_norm": 0.4983538091182709, "learning_rate": 5.4646017699115046e-05, "loss": 0.0699, "step": 12605 }, { "epoch": 1.366652211621856, "grad_norm": 0.4050811529159546, "learning_rate": 5.464239083127811e-05, "loss": 0.0136, "step": 12606 }, { "epoch": 1.3667606244579358, "grad_norm": 0.5403530597686768, "learning_rate": 5.4638763963441175e-05, "loss": 0.049, "step": 12607 }, { "epoch": 1.3668690372940155, "grad_norm": 0.8578757047653198, "learning_rate": 5.463513709560424e-05, "loss": 0.0234, "step": 12608 }, { "epoch": 1.3669774501300953, "grad_norm": 0.21008816361427307, "learning_rate": 5.46315102277673e-05, "loss": 0.0136, "step": 12609 }, { "epoch": 1.3670858629661753, "grad_norm": 0.4234195649623871, "learning_rate": 5.462788335993037e-05, "loss": 0.0311, "step": 12610 }, { "epoch": 1.367194275802255, "grad_norm": 0.07104600220918655, "learning_rate": 5.462425649209343e-05, "loss": 0.0046, "step": 12611 }, { "epoch": 1.3673026886383348, "grad_norm": 0.31964290142059326, "learning_rate": 5.4620629624256495e-05, "loss": 0.0209, "step": 12612 }, { "epoch": 1.3674111014744146, "grad_norm": 0.25637856125831604, "learning_rate": 5.461700275641955e-05, "loss": 0.0165, "step": 12613 }, { "epoch": 1.3675195143104943, "grad_norm": 0.1627984493970871, "learning_rate": 5.4613375888582617e-05, "loss": 0.0187, "step": 12614 }, { "epoch": 1.367627927146574, "grad_norm": 0.5868772268295288, "learning_rate": 5.460974902074568e-05, "loss": 0.0464, "step": 12615 }, { "epoch": 1.3677363399826539, "grad_norm": 1.1362757682800293, "learning_rate": 5.4606122152908745e-05, "loss": 0.0459, "step": 12616 }, { "epoch": 1.3678447528187339, "grad_norm": 0.34877750277519226, "learning_rate": 5.460249528507182e-05, "loss": 0.0306, "step": 12617 }, { "epoch": 1.3679531656548134, "grad_norm": 0.5071226358413696, "learning_rate": 5.459886841723488e-05, "loss": 0.0202, "step": 12618 }, { "epoch": 1.3680615784908934, "grad_norm": 0.11792542785406113, "learning_rate": 5.4595241549397944e-05, "loss": 0.0134, "step": 12619 }, { "epoch": 1.3681699913269731, "grad_norm": 1.0257395505905151, "learning_rate": 5.459161468156101e-05, "loss": 0.0499, "step": 12620 }, { "epoch": 1.368278404163053, "grad_norm": 0.2899441421031952, "learning_rate": 5.458798781372407e-05, "loss": 0.0371, "step": 12621 }, { "epoch": 1.3683868169991327, "grad_norm": 0.4459541141986847, "learning_rate": 5.4584360945887136e-05, "loss": 0.0363, "step": 12622 }, { "epoch": 1.3684952298352124, "grad_norm": 0.31771716475486755, "learning_rate": 5.45807340780502e-05, "loss": 0.0377, "step": 12623 }, { "epoch": 1.3686036426712924, "grad_norm": 0.4616518020629883, "learning_rate": 5.4577107210213264e-05, "loss": 0.0102, "step": 12624 }, { "epoch": 1.368712055507372, "grad_norm": 0.6070092916488647, "learning_rate": 5.457348034237633e-05, "loss": 0.0288, "step": 12625 }, { "epoch": 1.368820468343452, "grad_norm": 0.10113231837749481, "learning_rate": 5.456985347453939e-05, "loss": 0.007, "step": 12626 }, { "epoch": 1.3689288811795317, "grad_norm": 0.3552413880825043, "learning_rate": 5.456622660670245e-05, "loss": 0.0316, "step": 12627 }, { "epoch": 1.3690372940156115, "grad_norm": 0.1424650400876999, "learning_rate": 5.4562599738865514e-05, "loss": 0.0235, "step": 12628 }, { "epoch": 1.3691457068516912, "grad_norm": 0.08777153491973877, "learning_rate": 5.455897287102858e-05, "loss": 0.0077, "step": 12629 }, { "epoch": 1.369254119687771, "grad_norm": 0.6069840788841248, "learning_rate": 5.455534600319164e-05, "loss": 0.0226, "step": 12630 }, { "epoch": 1.3693625325238508, "grad_norm": 0.2807922661304474, "learning_rate": 5.4551719135354706e-05, "loss": 0.0171, "step": 12631 }, { "epoch": 1.3694709453599305, "grad_norm": 0.2917855381965637, "learning_rate": 5.454809226751777e-05, "loss": 0.0216, "step": 12632 }, { "epoch": 1.3695793581960105, "grad_norm": 0.11721164733171463, "learning_rate": 5.4544465399680835e-05, "loss": 0.0091, "step": 12633 }, { "epoch": 1.3696877710320903, "grad_norm": 0.8552160859107971, "learning_rate": 5.4540838531843905e-05, "loss": 0.0609, "step": 12634 }, { "epoch": 1.36979618386817, "grad_norm": 0.19028465449810028, "learning_rate": 5.453721166400697e-05, "loss": 0.0112, "step": 12635 }, { "epoch": 1.3699045967042498, "grad_norm": 0.19991528987884521, "learning_rate": 5.4533584796170034e-05, "loss": 0.017, "step": 12636 }, { "epoch": 1.3700130095403296, "grad_norm": 0.19739040732383728, "learning_rate": 5.45299579283331e-05, "loss": 0.0116, "step": 12637 }, { "epoch": 1.3701214223764093, "grad_norm": 0.22265660762786865, "learning_rate": 5.452633106049616e-05, "loss": 0.0272, "step": 12638 }, { "epoch": 1.370229835212489, "grad_norm": 0.3247357904911041, "learning_rate": 5.4522704192659226e-05, "loss": 0.0375, "step": 12639 }, { "epoch": 1.370338248048569, "grad_norm": 0.7578923106193542, "learning_rate": 5.451907732482229e-05, "loss": 0.0276, "step": 12640 }, { "epoch": 1.3704466608846486, "grad_norm": 0.5366566181182861, "learning_rate": 5.451545045698535e-05, "loss": 0.0503, "step": 12641 }, { "epoch": 1.3705550737207286, "grad_norm": 0.4712505638599396, "learning_rate": 5.451182358914841e-05, "loss": 0.0452, "step": 12642 }, { "epoch": 1.3706634865568084, "grad_norm": 0.3412330448627472, "learning_rate": 5.4508196721311476e-05, "loss": 0.0351, "step": 12643 }, { "epoch": 1.3707718993928881, "grad_norm": 0.565691351890564, "learning_rate": 5.450456985347454e-05, "loss": 0.0497, "step": 12644 }, { "epoch": 1.370880312228968, "grad_norm": 0.16764940321445465, "learning_rate": 5.4500942985637604e-05, "loss": 0.0071, "step": 12645 }, { "epoch": 1.3709887250650477, "grad_norm": 0.13019856810569763, "learning_rate": 5.449731611780067e-05, "loss": 0.0041, "step": 12646 }, { "epoch": 1.3710971379011274, "grad_norm": 0.4004473090171814, "learning_rate": 5.449368924996373e-05, "loss": 0.0137, "step": 12647 }, { "epoch": 1.3712055507372072, "grad_norm": 0.3653540313243866, "learning_rate": 5.4490062382126796e-05, "loss": 0.0347, "step": 12648 }, { "epoch": 1.3713139635732872, "grad_norm": 0.4850330352783203, "learning_rate": 5.448643551428986e-05, "loss": 0.0707, "step": 12649 }, { "epoch": 1.371422376409367, "grad_norm": 0.41721466183662415, "learning_rate": 5.448280864645292e-05, "loss": 0.0226, "step": 12650 }, { "epoch": 1.3715307892454467, "grad_norm": 0.328608900308609, "learning_rate": 5.4479181778615995e-05, "loss": 0.0216, "step": 12651 }, { "epoch": 1.3716392020815265, "grad_norm": 0.4351034164428711, "learning_rate": 5.447555491077906e-05, "loss": 0.0456, "step": 12652 }, { "epoch": 1.3717476149176062, "grad_norm": 0.06846949458122253, "learning_rate": 5.4471928042942123e-05, "loss": 0.0073, "step": 12653 }, { "epoch": 1.371856027753686, "grad_norm": 0.3285468518733978, "learning_rate": 5.446830117510519e-05, "loss": 0.0433, "step": 12654 }, { "epoch": 1.3719644405897657, "grad_norm": 0.15766717493534088, "learning_rate": 5.446467430726825e-05, "loss": 0.0108, "step": 12655 }, { "epoch": 1.3720728534258457, "grad_norm": 0.20479200780391693, "learning_rate": 5.446104743943131e-05, "loss": 0.0188, "step": 12656 }, { "epoch": 1.3721812662619255, "grad_norm": 0.14039146900177002, "learning_rate": 5.445742057159437e-05, "loss": 0.0087, "step": 12657 }, { "epoch": 1.3722896790980053, "grad_norm": 0.2811935544013977, "learning_rate": 5.445379370375744e-05, "loss": 0.0371, "step": 12658 }, { "epoch": 1.372398091934085, "grad_norm": 0.21353696286678314, "learning_rate": 5.44501668359205e-05, "loss": 0.0232, "step": 12659 }, { "epoch": 1.3725065047701648, "grad_norm": 0.355176717042923, "learning_rate": 5.4446539968083565e-05, "loss": 0.0147, "step": 12660 }, { "epoch": 1.3726149176062445, "grad_norm": 0.11938157677650452, "learning_rate": 5.444291310024663e-05, "loss": 0.0083, "step": 12661 }, { "epoch": 1.3727233304423243, "grad_norm": 0.0431283637881279, "learning_rate": 5.4439286232409694e-05, "loss": 0.0017, "step": 12662 }, { "epoch": 1.3728317432784043, "grad_norm": 0.4966804087162018, "learning_rate": 5.443565936457276e-05, "loss": 0.0304, "step": 12663 }, { "epoch": 1.3729401561144838, "grad_norm": 0.12578824162483215, "learning_rate": 5.443203249673582e-05, "loss": 0.0206, "step": 12664 }, { "epoch": 1.3730485689505638, "grad_norm": 0.21626992523670197, "learning_rate": 5.442840562889888e-05, "loss": 0.0134, "step": 12665 }, { "epoch": 1.3731569817866436, "grad_norm": 0.2793758511543274, "learning_rate": 5.442477876106194e-05, "loss": 0.0378, "step": 12666 }, { "epoch": 1.3732653946227233, "grad_norm": 0.16279007494449615, "learning_rate": 5.442115189322501e-05, "loss": 0.0099, "step": 12667 }, { "epoch": 1.373373807458803, "grad_norm": 0.2907602787017822, "learning_rate": 5.441752502538807e-05, "loss": 0.0124, "step": 12668 }, { "epoch": 1.3734822202948829, "grad_norm": 0.5453822612762451, "learning_rate": 5.441389815755115e-05, "loss": 0.0433, "step": 12669 }, { "epoch": 1.3735906331309626, "grad_norm": 0.3808872401714325, "learning_rate": 5.4410271289714206e-05, "loss": 0.0299, "step": 12670 }, { "epoch": 1.3736990459670424, "grad_norm": 0.3122072219848633, "learning_rate": 5.440664442187727e-05, "loss": 0.0395, "step": 12671 }, { "epoch": 1.3738074588031224, "grad_norm": 0.08731438964605331, "learning_rate": 5.4403017554040335e-05, "loss": 0.0027, "step": 12672 }, { "epoch": 1.3739158716392021, "grad_norm": 0.2691674828529358, "learning_rate": 5.43993906862034e-05, "loss": 0.0215, "step": 12673 }, { "epoch": 1.374024284475282, "grad_norm": 0.13021233677864075, "learning_rate": 5.439576381836646e-05, "loss": 0.0126, "step": 12674 }, { "epoch": 1.3741326973113617, "grad_norm": 0.2283831685781479, "learning_rate": 5.439213695052953e-05, "loss": 0.0206, "step": 12675 }, { "epoch": 1.3742411101474414, "grad_norm": 0.4543329179286957, "learning_rate": 5.438851008269259e-05, "loss": 0.076, "step": 12676 }, { "epoch": 1.3743495229835212, "grad_norm": 0.4347177743911743, "learning_rate": 5.4384883214855655e-05, "loss": 0.0618, "step": 12677 }, { "epoch": 1.374457935819601, "grad_norm": 0.23709550499916077, "learning_rate": 5.438125634701872e-05, "loss": 0.0498, "step": 12678 }, { "epoch": 1.374566348655681, "grad_norm": 0.6122332215309143, "learning_rate": 5.437762947918178e-05, "loss": 0.047, "step": 12679 }, { "epoch": 1.3746747614917605, "grad_norm": 0.09461686760187149, "learning_rate": 5.437400261134484e-05, "loss": 0.0113, "step": 12680 }, { "epoch": 1.3747831743278405, "grad_norm": 0.2717359662055969, "learning_rate": 5.4370375743507905e-05, "loss": 0.0156, "step": 12681 }, { "epoch": 1.3748915871639202, "grad_norm": 0.2949867248535156, "learning_rate": 5.436674887567097e-05, "loss": 0.0442, "step": 12682 }, { "epoch": 1.375, "grad_norm": 0.5623168349266052, "learning_rate": 5.436312200783403e-05, "loss": 0.0242, "step": 12683 }, { "epoch": 1.3751084128360798, "grad_norm": 0.5624405741691589, "learning_rate": 5.43594951399971e-05, "loss": 0.0523, "step": 12684 }, { "epoch": 1.3752168256721595, "grad_norm": 0.5979024767875671, "learning_rate": 5.435586827216016e-05, "loss": 0.0367, "step": 12685 }, { "epoch": 1.3753252385082395, "grad_norm": 0.0721939280629158, "learning_rate": 5.435224140432323e-05, "loss": 0.0071, "step": 12686 }, { "epoch": 1.375433651344319, "grad_norm": 0.3242858350276947, "learning_rate": 5.4348614536486296e-05, "loss": 0.0135, "step": 12687 }, { "epoch": 1.375542064180399, "grad_norm": 0.35927462577819824, "learning_rate": 5.434498766864936e-05, "loss": 0.0237, "step": 12688 }, { "epoch": 1.3756504770164788, "grad_norm": 0.0673936977982521, "learning_rate": 5.4341360800812424e-05, "loss": 0.0045, "step": 12689 }, { "epoch": 1.3757588898525586, "grad_norm": 0.08780329674482346, "learning_rate": 5.433773393297549e-05, "loss": 0.0111, "step": 12690 }, { "epoch": 1.3758673026886383, "grad_norm": 0.2733408212661743, "learning_rate": 5.433410706513855e-05, "loss": 0.0246, "step": 12691 }, { "epoch": 1.375975715524718, "grad_norm": 0.2414741963148117, "learning_rate": 5.433048019730162e-05, "loss": 0.0129, "step": 12692 }, { "epoch": 1.3760841283607979, "grad_norm": 0.5105483531951904, "learning_rate": 5.432685332946468e-05, "loss": 0.0181, "step": 12693 }, { "epoch": 1.3761925411968776, "grad_norm": 0.3933013081550598, "learning_rate": 5.432322646162774e-05, "loss": 0.0484, "step": 12694 }, { "epoch": 1.3763009540329576, "grad_norm": 0.3190608322620392, "learning_rate": 5.43195995937908e-05, "loss": 0.0218, "step": 12695 }, { "epoch": 1.3764093668690374, "grad_norm": 0.14197298884391785, "learning_rate": 5.4315972725953866e-05, "loss": 0.014, "step": 12696 }, { "epoch": 1.3765177797051171, "grad_norm": 0.0772237628698349, "learning_rate": 5.431234585811693e-05, "loss": 0.0087, "step": 12697 }, { "epoch": 1.376626192541197, "grad_norm": 0.28426703810691833, "learning_rate": 5.4308718990279995e-05, "loss": 0.0279, "step": 12698 }, { "epoch": 1.3767346053772767, "grad_norm": 0.6526366472244263, "learning_rate": 5.430509212244306e-05, "loss": 0.0526, "step": 12699 }, { "epoch": 1.3768430182133564, "grad_norm": 0.24496504664421082, "learning_rate": 5.430146525460612e-05, "loss": 0.0114, "step": 12700 }, { "epoch": 1.3769514310494362, "grad_norm": 0.33237072825431824, "learning_rate": 5.429783838676919e-05, "loss": 0.0819, "step": 12701 }, { "epoch": 1.3770598438855162, "grad_norm": 0.27930012345314026, "learning_rate": 5.4294211518932244e-05, "loss": 0.0054, "step": 12702 }, { "epoch": 1.3771682567215957, "grad_norm": 0.23217399418354034, "learning_rate": 5.429058465109532e-05, "loss": 0.0139, "step": 12703 }, { "epoch": 1.3772766695576757, "grad_norm": 0.3177449703216553, "learning_rate": 5.4286957783258386e-05, "loss": 0.0402, "step": 12704 }, { "epoch": 1.3773850823937555, "grad_norm": 0.3441975712776184, "learning_rate": 5.428333091542145e-05, "loss": 0.0326, "step": 12705 }, { "epoch": 1.3774934952298352, "grad_norm": 0.15707433223724365, "learning_rate": 5.4279704047584514e-05, "loss": 0.0252, "step": 12706 }, { "epoch": 1.377601908065915, "grad_norm": 0.24147391319274902, "learning_rate": 5.427607717974758e-05, "loss": 0.0093, "step": 12707 }, { "epoch": 1.3777103209019947, "grad_norm": 0.43264341354370117, "learning_rate": 5.4272450311910636e-05, "loss": 0.0177, "step": 12708 }, { "epoch": 1.3778187337380745, "grad_norm": 0.12440988421440125, "learning_rate": 5.42688234440737e-05, "loss": 0.0078, "step": 12709 }, { "epoch": 1.3779271465741543, "grad_norm": 0.13632313907146454, "learning_rate": 5.4265196576236764e-05, "loss": 0.0091, "step": 12710 }, { "epoch": 1.3780355594102343, "grad_norm": 0.2555249333381653, "learning_rate": 5.426156970839983e-05, "loss": 0.0231, "step": 12711 }, { "epoch": 1.378143972246314, "grad_norm": 0.4261718988418579, "learning_rate": 5.425794284056289e-05, "loss": 0.0519, "step": 12712 }, { "epoch": 1.3782523850823938, "grad_norm": 0.1919170618057251, "learning_rate": 5.4254315972725956e-05, "loss": 0.0096, "step": 12713 }, { "epoch": 1.3783607979184735, "grad_norm": 1.123105525970459, "learning_rate": 5.425068910488902e-05, "loss": 0.0813, "step": 12714 }, { "epoch": 1.3784692107545533, "grad_norm": 0.22727102041244507, "learning_rate": 5.4247062237052084e-05, "loss": 0.0242, "step": 12715 }, { "epoch": 1.378577623590633, "grad_norm": 0.21366439759731293, "learning_rate": 5.424343536921515e-05, "loss": 0.0233, "step": 12716 }, { "epoch": 1.3786860364267128, "grad_norm": 0.0376497320830822, "learning_rate": 5.4239808501378206e-05, "loss": 0.0019, "step": 12717 }, { "epoch": 1.3787944492627928, "grad_norm": 0.08964318037033081, "learning_rate": 5.423618163354127e-05, "loss": 0.0059, "step": 12718 }, { "epoch": 1.3789028620988726, "grad_norm": 0.30129438638687134, "learning_rate": 5.4232554765704334e-05, "loss": 0.0202, "step": 12719 }, { "epoch": 1.3790112749349523, "grad_norm": 0.15001825988292694, "learning_rate": 5.422892789786741e-05, "loss": 0.0118, "step": 12720 }, { "epoch": 1.379119687771032, "grad_norm": 0.08419723808765411, "learning_rate": 5.4225301030030476e-05, "loss": 0.0033, "step": 12721 }, { "epoch": 1.3792281006071119, "grad_norm": 0.18257305026054382, "learning_rate": 5.422167416219353e-05, "loss": 0.0083, "step": 12722 }, { "epoch": 1.3793365134431916, "grad_norm": 0.2645721137523651, "learning_rate": 5.42180472943566e-05, "loss": 0.0288, "step": 12723 }, { "epoch": 1.3794449262792714, "grad_norm": 0.41313502192497253, "learning_rate": 5.421442042651966e-05, "loss": 0.07, "step": 12724 }, { "epoch": 1.3795533391153514, "grad_norm": 0.25057917833328247, "learning_rate": 5.4210793558682725e-05, "loss": 0.0154, "step": 12725 }, { "epoch": 1.379661751951431, "grad_norm": 0.28530436754226685, "learning_rate": 5.420716669084579e-05, "loss": 0.0372, "step": 12726 }, { "epoch": 1.379770164787511, "grad_norm": 0.15545646846294403, "learning_rate": 5.4203539823008854e-05, "loss": 0.0117, "step": 12727 }, { "epoch": 1.3798785776235907, "grad_norm": 0.4627203047275543, "learning_rate": 5.419991295517192e-05, "loss": 0.0384, "step": 12728 }, { "epoch": 1.3799869904596704, "grad_norm": 0.46890348196029663, "learning_rate": 5.419628608733498e-05, "loss": 0.021, "step": 12729 }, { "epoch": 1.3800954032957502, "grad_norm": 0.12374194711446762, "learning_rate": 5.4192659219498046e-05, "loss": 0.0093, "step": 12730 }, { "epoch": 1.38020381613183, "grad_norm": 0.4854413568973541, "learning_rate": 5.41890323516611e-05, "loss": 0.0375, "step": 12731 }, { "epoch": 1.3803122289679097, "grad_norm": 0.3074842095375061, "learning_rate": 5.418540548382417e-05, "loss": 0.0314, "step": 12732 }, { "epoch": 1.3804206418039895, "grad_norm": 0.253230482339859, "learning_rate": 5.418177861598723e-05, "loss": 0.0323, "step": 12733 }, { "epoch": 1.3805290546400695, "grad_norm": 0.04797014221549034, "learning_rate": 5.4178151748150296e-05, "loss": 0.0037, "step": 12734 }, { "epoch": 1.3806374674761492, "grad_norm": 0.17198114097118378, "learning_rate": 5.417452488031336e-05, "loss": 0.0097, "step": 12735 }, { "epoch": 1.380745880312229, "grad_norm": 0.17553310096263885, "learning_rate": 5.4170898012476424e-05, "loss": 0.0168, "step": 12736 }, { "epoch": 1.3808542931483088, "grad_norm": 0.23891502618789673, "learning_rate": 5.416727114463949e-05, "loss": 0.0139, "step": 12737 }, { "epoch": 1.3809627059843885, "grad_norm": 0.13409017026424408, "learning_rate": 5.416364427680256e-05, "loss": 0.0133, "step": 12738 }, { "epoch": 1.3810711188204683, "grad_norm": 0.23470602929592133, "learning_rate": 5.416001740896562e-05, "loss": 0.0156, "step": 12739 }, { "epoch": 1.381179531656548, "grad_norm": 0.2504102885723114, "learning_rate": 5.415639054112869e-05, "loss": 0.0132, "step": 12740 }, { "epoch": 1.381287944492628, "grad_norm": 0.19027726352214813, "learning_rate": 5.415276367329175e-05, "loss": 0.0154, "step": 12741 }, { "epoch": 1.3813963573287076, "grad_norm": 0.17678681015968323, "learning_rate": 5.4149136805454815e-05, "loss": 0.0191, "step": 12742 }, { "epoch": 1.3815047701647876, "grad_norm": 0.2833021581172943, "learning_rate": 5.414550993761788e-05, "loss": 0.026, "step": 12743 }, { "epoch": 1.3816131830008673, "grad_norm": 0.06096271052956581, "learning_rate": 5.4141883069780943e-05, "loss": 0.0036, "step": 12744 }, { "epoch": 1.381721595836947, "grad_norm": 0.29316824674606323, "learning_rate": 5.413825620194401e-05, "loss": 0.0262, "step": 12745 }, { "epoch": 1.3818300086730269, "grad_norm": 0.21086302399635315, "learning_rate": 5.4134629334107065e-05, "loss": 0.0141, "step": 12746 }, { "epoch": 1.3819384215091066, "grad_norm": 0.5223512649536133, "learning_rate": 5.413100246627013e-05, "loss": 0.0214, "step": 12747 }, { "epoch": 1.3820468343451866, "grad_norm": 0.19965411722660065, "learning_rate": 5.412737559843319e-05, "loss": 0.0181, "step": 12748 }, { "epoch": 1.3821552471812661, "grad_norm": 0.19274428486824036, "learning_rate": 5.412374873059626e-05, "loss": 0.0156, "step": 12749 }, { "epoch": 1.3822636600173461, "grad_norm": 0.15095393359661102, "learning_rate": 5.412012186275932e-05, "loss": 0.0144, "step": 12750 }, { "epoch": 1.382372072853426, "grad_norm": 0.08179327845573425, "learning_rate": 5.4116494994922385e-05, "loss": 0.0082, "step": 12751 }, { "epoch": 1.3824804856895057, "grad_norm": 0.254747599363327, "learning_rate": 5.411286812708545e-05, "loss": 0.0215, "step": 12752 }, { "epoch": 1.3825888985255854, "grad_norm": 1.2592328786849976, "learning_rate": 5.4109241259248514e-05, "loss": 0.0184, "step": 12753 }, { "epoch": 1.3826973113616652, "grad_norm": 0.20717953145503998, "learning_rate": 5.410561439141157e-05, "loss": 0.0138, "step": 12754 }, { "epoch": 1.382805724197745, "grad_norm": 0.4611593782901764, "learning_rate": 5.410198752357465e-05, "loss": 0.0316, "step": 12755 }, { "epoch": 1.3829141370338247, "grad_norm": 0.3010157644748688, "learning_rate": 5.409836065573771e-05, "loss": 0.0087, "step": 12756 }, { "epoch": 1.3830225498699047, "grad_norm": 0.12651923298835754, "learning_rate": 5.409473378790078e-05, "loss": 0.0053, "step": 12757 }, { "epoch": 1.3831309627059845, "grad_norm": 0.1878504902124405, "learning_rate": 5.409110692006384e-05, "loss": 0.0116, "step": 12758 }, { "epoch": 1.3832393755420642, "grad_norm": 0.01758948154747486, "learning_rate": 5.4087480052226905e-05, "loss": 0.0011, "step": 12759 }, { "epoch": 1.383347788378144, "grad_norm": 0.08955375850200653, "learning_rate": 5.408385318438996e-05, "loss": 0.0026, "step": 12760 }, { "epoch": 1.3834562012142237, "grad_norm": 0.3725877106189728, "learning_rate": 5.4080226316553026e-05, "loss": 0.0191, "step": 12761 }, { "epoch": 1.3835646140503035, "grad_norm": 0.26381656527519226, "learning_rate": 5.407659944871609e-05, "loss": 0.0073, "step": 12762 }, { "epoch": 1.3836730268863833, "grad_norm": 1.1066150665283203, "learning_rate": 5.4072972580879155e-05, "loss": 0.0183, "step": 12763 }, { "epoch": 1.3837814397224633, "grad_norm": 0.19267861545085907, "learning_rate": 5.406934571304222e-05, "loss": 0.0073, "step": 12764 }, { "epoch": 1.3838898525585428, "grad_norm": 0.07293453067541122, "learning_rate": 5.406571884520528e-05, "loss": 0.0042, "step": 12765 }, { "epoch": 1.3839982653946228, "grad_norm": 0.35577985644340515, "learning_rate": 5.406209197736835e-05, "loss": 0.0527, "step": 12766 }, { "epoch": 1.3841066782307025, "grad_norm": 0.19457629323005676, "learning_rate": 5.405846510953141e-05, "loss": 0.0118, "step": 12767 }, { "epoch": 1.3842150910667823, "grad_norm": 0.6319605708122253, "learning_rate": 5.4054838241694475e-05, "loss": 0.101, "step": 12768 }, { "epoch": 1.384323503902862, "grad_norm": 0.6083257794380188, "learning_rate": 5.405121137385753e-05, "loss": 0.0498, "step": 12769 }, { "epoch": 1.3844319167389418, "grad_norm": 0.9492250084877014, "learning_rate": 5.40475845060206e-05, "loss": 0.0527, "step": 12770 }, { "epoch": 1.3845403295750216, "grad_norm": 0.039948780089616776, "learning_rate": 5.404395763818366e-05, "loss": 0.0011, "step": 12771 }, { "epoch": 1.3846487424111014, "grad_norm": 0.6607941389083862, "learning_rate": 5.404033077034674e-05, "loss": 0.0535, "step": 12772 }, { "epoch": 1.3847571552471813, "grad_norm": 0.3873600363731384, "learning_rate": 5.40367039025098e-05, "loss": 0.0427, "step": 12773 }, { "epoch": 1.3848655680832611, "grad_norm": 0.465706467628479, "learning_rate": 5.403307703467286e-05, "loss": 0.0284, "step": 12774 }, { "epoch": 1.3849739809193409, "grad_norm": 0.29016900062561035, "learning_rate": 5.4029450166835924e-05, "loss": 0.0206, "step": 12775 }, { "epoch": 1.3850823937554206, "grad_norm": 0.13777410984039307, "learning_rate": 5.402582329899899e-05, "loss": 0.0061, "step": 12776 }, { "epoch": 1.3851908065915004, "grad_norm": 0.6646341681480408, "learning_rate": 5.402219643116205e-05, "loss": 0.0314, "step": 12777 }, { "epoch": 1.3852992194275802, "grad_norm": 0.7163878679275513, "learning_rate": 5.4018569563325116e-05, "loss": 0.0227, "step": 12778 }, { "epoch": 1.38540763226366, "grad_norm": 0.17739520967006683, "learning_rate": 5.401494269548818e-05, "loss": 0.0078, "step": 12779 }, { "epoch": 1.38551604509974, "grad_norm": 0.047758545726537704, "learning_rate": 5.4011315827651244e-05, "loss": 0.0033, "step": 12780 }, { "epoch": 1.3856244579358197, "grad_norm": 0.26400068402290344, "learning_rate": 5.400768895981431e-05, "loss": 0.0205, "step": 12781 }, { "epoch": 1.3857328707718994, "grad_norm": 0.5644921660423279, "learning_rate": 5.400406209197737e-05, "loss": 0.0349, "step": 12782 }, { "epoch": 1.3858412836079792, "grad_norm": 0.23653018474578857, "learning_rate": 5.400043522414043e-05, "loss": 0.0193, "step": 12783 }, { "epoch": 1.385949696444059, "grad_norm": 0.6940923929214478, "learning_rate": 5.3996808356303494e-05, "loss": 0.0658, "step": 12784 }, { "epoch": 1.3860581092801387, "grad_norm": 0.3642328083515167, "learning_rate": 5.399318148846656e-05, "loss": 0.0241, "step": 12785 }, { "epoch": 1.3861665221162185, "grad_norm": 0.14831748604774475, "learning_rate": 5.398955462062962e-05, "loss": 0.004, "step": 12786 }, { "epoch": 1.3862749349522985, "grad_norm": 0.21438929438591003, "learning_rate": 5.3985927752792686e-05, "loss": 0.0291, "step": 12787 }, { "epoch": 1.386383347788378, "grad_norm": 0.2343715876340866, "learning_rate": 5.398230088495575e-05, "loss": 0.03, "step": 12788 }, { "epoch": 1.386491760624458, "grad_norm": 0.2630029022693634, "learning_rate": 5.397867401711882e-05, "loss": 0.0209, "step": 12789 }, { "epoch": 1.3866001734605378, "grad_norm": 0.5991584658622742, "learning_rate": 5.3975047149281886e-05, "loss": 0.0668, "step": 12790 }, { "epoch": 1.3867085862966175, "grad_norm": 0.006668413057923317, "learning_rate": 5.397142028144495e-05, "loss": 0.0005, "step": 12791 }, { "epoch": 1.3868169991326973, "grad_norm": 0.08042676746845245, "learning_rate": 5.3967793413608014e-05, "loss": 0.0028, "step": 12792 }, { "epoch": 1.386925411968777, "grad_norm": 0.25648173689842224, "learning_rate": 5.396416654577108e-05, "loss": 0.0109, "step": 12793 }, { "epoch": 1.3870338248048568, "grad_norm": 0.1826316863298416, "learning_rate": 5.396053967793414e-05, "loss": 0.01, "step": 12794 }, { "epoch": 1.3871422376409366, "grad_norm": 0.6719714403152466, "learning_rate": 5.3956912810097206e-05, "loss": 0.0159, "step": 12795 }, { "epoch": 1.3872506504770166, "grad_norm": 0.7197334170341492, "learning_rate": 5.395328594226027e-05, "loss": 0.0668, "step": 12796 }, { "epoch": 1.3873590633130963, "grad_norm": 0.2880226969718933, "learning_rate": 5.3949659074423334e-05, "loss": 0.0126, "step": 12797 }, { "epoch": 1.387467476149176, "grad_norm": 0.3050473630428314, "learning_rate": 5.394603220658639e-05, "loss": 0.0219, "step": 12798 }, { "epoch": 1.3875758889852559, "grad_norm": 0.1472715139389038, "learning_rate": 5.3942405338749456e-05, "loss": 0.0093, "step": 12799 }, { "epoch": 1.3876843018213356, "grad_norm": 0.17455463111400604, "learning_rate": 5.393877847091252e-05, "loss": 0.0126, "step": 12800 }, { "epoch": 1.3877927146574154, "grad_norm": 0.4426785111427307, "learning_rate": 5.3935151603075584e-05, "loss": 0.0374, "step": 12801 }, { "epoch": 1.3879011274934951, "grad_norm": 0.2670796811580658, "learning_rate": 5.393152473523865e-05, "loss": 0.0207, "step": 12802 }, { "epoch": 1.3880095403295751, "grad_norm": 1.4183906316757202, "learning_rate": 5.392789786740171e-05, "loss": 0.0607, "step": 12803 }, { "epoch": 1.388117953165655, "grad_norm": 0.44713103771209717, "learning_rate": 5.3924270999564776e-05, "loss": 0.0195, "step": 12804 }, { "epoch": 1.3882263660017347, "grad_norm": 0.15982183814048767, "learning_rate": 5.392064413172784e-05, "loss": 0.0104, "step": 12805 }, { "epoch": 1.3883347788378144, "grad_norm": 0.2617640495300293, "learning_rate": 5.39170172638909e-05, "loss": 0.0059, "step": 12806 }, { "epoch": 1.3884431916738942, "grad_norm": 0.4145229160785675, "learning_rate": 5.3913390396053975e-05, "loss": 0.0143, "step": 12807 }, { "epoch": 1.388551604509974, "grad_norm": 0.0858427882194519, "learning_rate": 5.390976352821704e-05, "loss": 0.0082, "step": 12808 }, { "epoch": 1.3886600173460537, "grad_norm": 0.1929968148469925, "learning_rate": 5.3906136660380104e-05, "loss": 0.0217, "step": 12809 }, { "epoch": 1.3887684301821337, "grad_norm": 0.427615761756897, "learning_rate": 5.390250979254317e-05, "loss": 0.02, "step": 12810 }, { "epoch": 1.3888768430182132, "grad_norm": 0.1527344435453415, "learning_rate": 5.389888292470623e-05, "loss": 0.0121, "step": 12811 }, { "epoch": 1.3889852558542932, "grad_norm": 0.11816654354333878, "learning_rate": 5.389525605686929e-05, "loss": 0.0046, "step": 12812 }, { "epoch": 1.389093668690373, "grad_norm": 0.4781285226345062, "learning_rate": 5.389162918903235e-05, "loss": 0.0466, "step": 12813 }, { "epoch": 1.3892020815264527, "grad_norm": 0.34060531854629517, "learning_rate": 5.388800232119542e-05, "loss": 0.0412, "step": 12814 }, { "epoch": 1.3893104943625325, "grad_norm": 0.07691173255443573, "learning_rate": 5.388437545335848e-05, "loss": 0.0035, "step": 12815 }, { "epoch": 1.3894189071986123, "grad_norm": 0.5274964570999146, "learning_rate": 5.3880748585521545e-05, "loss": 0.0294, "step": 12816 }, { "epoch": 1.389527320034692, "grad_norm": 0.6706242561340332, "learning_rate": 5.387712171768461e-05, "loss": 0.0415, "step": 12817 }, { "epoch": 1.3896357328707718, "grad_norm": 0.27034255862236023, "learning_rate": 5.3873494849847674e-05, "loss": 0.0057, "step": 12818 }, { "epoch": 1.3897441457068518, "grad_norm": 0.41918182373046875, "learning_rate": 5.386986798201074e-05, "loss": 0.0362, "step": 12819 }, { "epoch": 1.3898525585429315, "grad_norm": 0.45647233724594116, "learning_rate": 5.38662411141738e-05, "loss": 0.0192, "step": 12820 }, { "epoch": 1.3899609713790113, "grad_norm": 0.031246162950992584, "learning_rate": 5.386261424633686e-05, "loss": 0.0013, "step": 12821 }, { "epoch": 1.390069384215091, "grad_norm": 0.916670560836792, "learning_rate": 5.385898737849992e-05, "loss": 0.0368, "step": 12822 }, { "epoch": 1.3901777970511708, "grad_norm": 0.19234345853328705, "learning_rate": 5.385536051066299e-05, "loss": 0.0064, "step": 12823 }, { "epoch": 1.3902862098872506, "grad_norm": 0.24435247480869293, "learning_rate": 5.3851733642826065e-05, "loss": 0.0122, "step": 12824 }, { "epoch": 1.3903946227233304, "grad_norm": 0.2792252004146576, "learning_rate": 5.384810677498913e-05, "loss": 0.0234, "step": 12825 }, { "epoch": 1.3905030355594103, "grad_norm": 0.34255126118659973, "learning_rate": 5.384447990715219e-05, "loss": 0.017, "step": 12826 }, { "epoch": 1.39061144839549, "grad_norm": 0.1441657841205597, "learning_rate": 5.384085303931525e-05, "loss": 0.0052, "step": 12827 }, { "epoch": 1.3907198612315699, "grad_norm": 0.4091605842113495, "learning_rate": 5.3837226171478315e-05, "loss": 0.0434, "step": 12828 }, { "epoch": 1.3908282740676496, "grad_norm": 0.06866331398487091, "learning_rate": 5.383359930364138e-05, "loss": 0.0015, "step": 12829 }, { "epoch": 1.3909366869037294, "grad_norm": 1.1811031103134155, "learning_rate": 5.382997243580444e-05, "loss": 0.0294, "step": 12830 }, { "epoch": 1.3910450997398092, "grad_norm": 0.4916699528694153, "learning_rate": 5.382634556796751e-05, "loss": 0.0427, "step": 12831 }, { "epoch": 1.391153512575889, "grad_norm": 0.29782867431640625, "learning_rate": 5.382271870013057e-05, "loss": 0.0234, "step": 12832 }, { "epoch": 1.391261925411969, "grad_norm": 0.39742544293403625, "learning_rate": 5.3819091832293635e-05, "loss": 0.0407, "step": 12833 }, { "epoch": 1.3913703382480485, "grad_norm": 0.19333766400814056, "learning_rate": 5.38154649644567e-05, "loss": 0.0152, "step": 12834 }, { "epoch": 1.3914787510841284, "grad_norm": 0.2383442521095276, "learning_rate": 5.381183809661976e-05, "loss": 0.0174, "step": 12835 }, { "epoch": 1.3915871639202082, "grad_norm": 0.10055176913738251, "learning_rate": 5.380821122878282e-05, "loss": 0.0054, "step": 12836 }, { "epoch": 1.391695576756288, "grad_norm": 0.2863456606864929, "learning_rate": 5.3804584360945885e-05, "loss": 0.0151, "step": 12837 }, { "epoch": 1.3918039895923677, "grad_norm": 0.4170888066291809, "learning_rate": 5.380095749310895e-05, "loss": 0.061, "step": 12838 }, { "epoch": 1.3919124024284475, "grad_norm": 0.44729384779930115, "learning_rate": 5.379733062527201e-05, "loss": 0.0232, "step": 12839 }, { "epoch": 1.3920208152645273, "grad_norm": 1.1265997886657715, "learning_rate": 5.379370375743508e-05, "loss": 0.0717, "step": 12840 }, { "epoch": 1.392129228100607, "grad_norm": 0.4085545241832733, "learning_rate": 5.379007688959815e-05, "loss": 0.0307, "step": 12841 }, { "epoch": 1.392237640936687, "grad_norm": 0.32295337319374084, "learning_rate": 5.378645002176121e-05, "loss": 0.0191, "step": 12842 }, { "epoch": 1.3923460537727668, "grad_norm": 0.5797858238220215, "learning_rate": 5.3782823153924276e-05, "loss": 0.0587, "step": 12843 }, { "epoch": 1.3924544666088465, "grad_norm": 0.1473173052072525, "learning_rate": 5.377919628608734e-05, "loss": 0.0078, "step": 12844 }, { "epoch": 1.3925628794449263, "grad_norm": 0.7545580863952637, "learning_rate": 5.3775569418250405e-05, "loss": 0.0388, "step": 12845 }, { "epoch": 1.392671292281006, "grad_norm": 0.22614671289920807, "learning_rate": 5.377194255041347e-05, "loss": 0.0227, "step": 12846 }, { "epoch": 1.3927797051170858, "grad_norm": 0.49407675862312317, "learning_rate": 5.376831568257653e-05, "loss": 0.0622, "step": 12847 }, { "epoch": 1.3928881179531656, "grad_norm": 0.10765185207128525, "learning_rate": 5.37646888147396e-05, "loss": 0.011, "step": 12848 }, { "epoch": 1.3929965307892456, "grad_norm": 0.3606356084346771, "learning_rate": 5.376106194690266e-05, "loss": 0.0168, "step": 12849 }, { "epoch": 1.393104943625325, "grad_norm": 0.404939740896225, "learning_rate": 5.375743507906572e-05, "loss": 0.0173, "step": 12850 }, { "epoch": 1.393213356461405, "grad_norm": 0.4587717652320862, "learning_rate": 5.375380821122878e-05, "loss": 0.0238, "step": 12851 }, { "epoch": 1.3933217692974849, "grad_norm": 0.3335739076137543, "learning_rate": 5.3750181343391846e-05, "loss": 0.0111, "step": 12852 }, { "epoch": 1.3934301821335646, "grad_norm": 0.2729954421520233, "learning_rate": 5.374655447555491e-05, "loss": 0.0066, "step": 12853 }, { "epoch": 1.3935385949696444, "grad_norm": 0.1550508290529251, "learning_rate": 5.3742927607717975e-05, "loss": 0.0039, "step": 12854 }, { "epoch": 1.3936470078057241, "grad_norm": 0.34476861357688904, "learning_rate": 5.373930073988104e-05, "loss": 0.0203, "step": 12855 }, { "epoch": 1.393755420641804, "grad_norm": 0.11159462481737137, "learning_rate": 5.37356738720441e-05, "loss": 0.0071, "step": 12856 }, { "epoch": 1.3938638334778837, "grad_norm": 0.02733108587563038, "learning_rate": 5.373204700420717e-05, "loss": 0.0023, "step": 12857 }, { "epoch": 1.3939722463139637, "grad_norm": 0.2634209096431732, "learning_rate": 5.372842013637023e-05, "loss": 0.0143, "step": 12858 }, { "epoch": 1.3940806591500434, "grad_norm": 0.14277753233909607, "learning_rate": 5.37247932685333e-05, "loss": 0.0122, "step": 12859 }, { "epoch": 1.3941890719861232, "grad_norm": 0.38051843643188477, "learning_rate": 5.3721166400696366e-05, "loss": 0.0384, "step": 12860 }, { "epoch": 1.394297484822203, "grad_norm": 0.5062971711158752, "learning_rate": 5.371753953285943e-05, "loss": 0.0397, "step": 12861 }, { "epoch": 1.3944058976582827, "grad_norm": 0.20371142029762268, "learning_rate": 5.3713912665022494e-05, "loss": 0.0134, "step": 12862 }, { "epoch": 1.3945143104943625, "grad_norm": 0.2433466911315918, "learning_rate": 5.371028579718556e-05, "loss": 0.0279, "step": 12863 }, { "epoch": 1.3946227233304422, "grad_norm": 0.06929977983236313, "learning_rate": 5.3706658929348616e-05, "loss": 0.005, "step": 12864 }, { "epoch": 1.3947311361665222, "grad_norm": 1.0867341756820679, "learning_rate": 5.370303206151168e-05, "loss": 0.0667, "step": 12865 }, { "epoch": 1.394839549002602, "grad_norm": 0.8587682247161865, "learning_rate": 5.3699405193674744e-05, "loss": 0.0704, "step": 12866 }, { "epoch": 1.3949479618386817, "grad_norm": 0.16501881182193756, "learning_rate": 5.369577832583781e-05, "loss": 0.014, "step": 12867 }, { "epoch": 1.3950563746747615, "grad_norm": 1.1718767881393433, "learning_rate": 5.369215145800087e-05, "loss": 0.0342, "step": 12868 }, { "epoch": 1.3951647875108413, "grad_norm": 0.411647230386734, "learning_rate": 5.3688524590163936e-05, "loss": 0.022, "step": 12869 }, { "epoch": 1.395273200346921, "grad_norm": 0.19093044102191925, "learning_rate": 5.3684897722327e-05, "loss": 0.0249, "step": 12870 }, { "epoch": 1.3953816131830008, "grad_norm": 0.44644564390182495, "learning_rate": 5.3681270854490064e-05, "loss": 0.0229, "step": 12871 }, { "epoch": 1.3954900260190808, "grad_norm": 0.2135927677154541, "learning_rate": 5.367764398665313e-05, "loss": 0.0191, "step": 12872 }, { "epoch": 1.3955984388551603, "grad_norm": 0.6561123132705688, "learning_rate": 5.3674017118816186e-05, "loss": 0.0598, "step": 12873 }, { "epoch": 1.3957068516912403, "grad_norm": 0.5278111696243286, "learning_rate": 5.367039025097925e-05, "loss": 0.0129, "step": 12874 }, { "epoch": 1.39581526452732, "grad_norm": 0.5102156400680542, "learning_rate": 5.3666763383142314e-05, "loss": 0.0365, "step": 12875 }, { "epoch": 1.3959236773633998, "grad_norm": 0.3965343236923218, "learning_rate": 5.366313651530539e-05, "loss": 0.0328, "step": 12876 }, { "epoch": 1.3960320901994796, "grad_norm": 0.20279884338378906, "learning_rate": 5.3659509647468456e-05, "loss": 0.0107, "step": 12877 }, { "epoch": 1.3961405030355594, "grad_norm": 1.3397423028945923, "learning_rate": 5.365588277963152e-05, "loss": 0.0262, "step": 12878 }, { "epoch": 1.3962489158716391, "grad_norm": 0.5645922422409058, "learning_rate": 5.365225591179458e-05, "loss": 0.1438, "step": 12879 }, { "epoch": 1.396357328707719, "grad_norm": 0.3138980269432068, "learning_rate": 5.364862904395764e-05, "loss": 0.015, "step": 12880 }, { "epoch": 1.3964657415437989, "grad_norm": 0.14933252334594727, "learning_rate": 5.3645002176120706e-05, "loss": 0.0052, "step": 12881 }, { "epoch": 1.3965741543798786, "grad_norm": 0.41194096207618713, "learning_rate": 5.364137530828377e-05, "loss": 0.038, "step": 12882 }, { "epoch": 1.3966825672159584, "grad_norm": 0.519559919834137, "learning_rate": 5.3637748440446834e-05, "loss": 0.0165, "step": 12883 }, { "epoch": 1.3967909800520382, "grad_norm": 0.1925971508026123, "learning_rate": 5.36341215726099e-05, "loss": 0.0182, "step": 12884 }, { "epoch": 1.396899392888118, "grad_norm": 0.3162689507007599, "learning_rate": 5.363049470477296e-05, "loss": 0.0286, "step": 12885 }, { "epoch": 1.3970078057241977, "grad_norm": 0.6504240036010742, "learning_rate": 5.3626867836936026e-05, "loss": 0.0306, "step": 12886 }, { "epoch": 1.3971162185602775, "grad_norm": 0.44106003642082214, "learning_rate": 5.3623240969099083e-05, "loss": 0.0144, "step": 12887 }, { "epoch": 1.3972246313963574, "grad_norm": 0.20339888334274292, "learning_rate": 5.361961410126215e-05, "loss": 0.018, "step": 12888 }, { "epoch": 1.397333044232437, "grad_norm": 0.32646894454956055, "learning_rate": 5.361598723342521e-05, "loss": 0.0103, "step": 12889 }, { "epoch": 1.397441457068517, "grad_norm": 0.644025444984436, "learning_rate": 5.3612360365588276e-05, "loss": 0.0718, "step": 12890 }, { "epoch": 1.3975498699045967, "grad_norm": 0.28578701615333557, "learning_rate": 5.360873349775134e-05, "loss": 0.0208, "step": 12891 }, { "epoch": 1.3976582827406765, "grad_norm": 0.07062973827123642, "learning_rate": 5.3605106629914404e-05, "loss": 0.0042, "step": 12892 }, { "epoch": 1.3977666955767563, "grad_norm": 0.5555827021598816, "learning_rate": 5.3601479762077475e-05, "loss": 0.0369, "step": 12893 }, { "epoch": 1.397875108412836, "grad_norm": 0.6203746795654297, "learning_rate": 5.359785289424054e-05, "loss": 0.0613, "step": 12894 }, { "epoch": 1.397983521248916, "grad_norm": 0.23935124278068542, "learning_rate": 5.35942260264036e-05, "loss": 0.0134, "step": 12895 }, { "epoch": 1.3980919340849955, "grad_norm": 0.3852565884590149, "learning_rate": 5.359059915856667e-05, "loss": 0.0296, "step": 12896 }, { "epoch": 1.3982003469210755, "grad_norm": 0.18539410829544067, "learning_rate": 5.358697229072973e-05, "loss": 0.019, "step": 12897 }, { "epoch": 1.3983087597571553, "grad_norm": 0.3033115565776825, "learning_rate": 5.3583345422892795e-05, "loss": 0.0257, "step": 12898 }, { "epoch": 1.398417172593235, "grad_norm": 0.36888906359672546, "learning_rate": 5.357971855505586e-05, "loss": 0.0265, "step": 12899 }, { "epoch": 1.3985255854293148, "grad_norm": 0.1357342153787613, "learning_rate": 5.3576091687218924e-05, "loss": 0.0039, "step": 12900 }, { "epoch": 1.3986339982653946, "grad_norm": 0.34552431106567383, "learning_rate": 5.357246481938199e-05, "loss": 0.0424, "step": 12901 }, { "epoch": 1.3987424111014743, "grad_norm": 0.20407503843307495, "learning_rate": 5.3568837951545045e-05, "loss": 0.012, "step": 12902 }, { "epoch": 1.398850823937554, "grad_norm": 0.10702548176050186, "learning_rate": 5.356521108370811e-05, "loss": 0.0154, "step": 12903 }, { "epoch": 1.398959236773634, "grad_norm": 0.06515543162822723, "learning_rate": 5.356158421587117e-05, "loss": 0.0051, "step": 12904 }, { "epoch": 1.3990676496097139, "grad_norm": 0.37212806940078735, "learning_rate": 5.355795734803424e-05, "loss": 0.0379, "step": 12905 }, { "epoch": 1.3991760624457936, "grad_norm": 0.2130606770515442, "learning_rate": 5.35543304801973e-05, "loss": 0.0605, "step": 12906 }, { "epoch": 1.3992844752818734, "grad_norm": 0.41827765107154846, "learning_rate": 5.3550703612360365e-05, "loss": 0.0235, "step": 12907 }, { "epoch": 1.3993928881179531, "grad_norm": 0.2314470112323761, "learning_rate": 5.354707674452343e-05, "loss": 0.0118, "step": 12908 }, { "epoch": 1.399501300954033, "grad_norm": 0.12089177221059799, "learning_rate": 5.3543449876686494e-05, "loss": 0.0118, "step": 12909 }, { "epoch": 1.3996097137901127, "grad_norm": 0.12600575387477875, "learning_rate": 5.3539823008849565e-05, "loss": 0.0114, "step": 12910 }, { "epoch": 1.3997181266261927, "grad_norm": 0.2821230888366699, "learning_rate": 5.353619614101263e-05, "loss": 0.0324, "step": 12911 }, { "epoch": 1.3998265394622722, "grad_norm": 0.44863370060920715, "learning_rate": 5.353256927317569e-05, "loss": 0.0293, "step": 12912 }, { "epoch": 1.3999349522983522, "grad_norm": 0.4704999327659607, "learning_rate": 5.352894240533876e-05, "loss": 0.0157, "step": 12913 }, { "epoch": 1.400043365134432, "grad_norm": 0.3042434751987457, "learning_rate": 5.352531553750182e-05, "loss": 0.0327, "step": 12914 }, { "epoch": 1.4001517779705117, "grad_norm": 0.10624174773693085, "learning_rate": 5.3521688669664885e-05, "loss": 0.009, "step": 12915 }, { "epoch": 1.4002601908065915, "grad_norm": 0.10209698975086212, "learning_rate": 5.351806180182794e-05, "loss": 0.0069, "step": 12916 }, { "epoch": 1.4003686036426712, "grad_norm": 0.420797735452652, "learning_rate": 5.3514434933991007e-05, "loss": 0.0346, "step": 12917 }, { "epoch": 1.400477016478751, "grad_norm": 0.2348572015762329, "learning_rate": 5.351080806615407e-05, "loss": 0.012, "step": 12918 }, { "epoch": 1.4005854293148308, "grad_norm": 0.11380500346422195, "learning_rate": 5.3507181198317135e-05, "loss": 0.0138, "step": 12919 }, { "epoch": 1.4006938421509108, "grad_norm": 0.09432782232761383, "learning_rate": 5.35035543304802e-05, "loss": 0.0082, "step": 12920 }, { "epoch": 1.4008022549869905, "grad_norm": 0.11570489406585693, "learning_rate": 5.349992746264326e-05, "loss": 0.0073, "step": 12921 }, { "epoch": 1.4009106678230703, "grad_norm": 0.09930548071861267, "learning_rate": 5.349630059480633e-05, "loss": 0.0096, "step": 12922 }, { "epoch": 1.40101908065915, "grad_norm": 0.19924329221248627, "learning_rate": 5.349267372696939e-05, "loss": 0.0218, "step": 12923 }, { "epoch": 1.4011274934952298, "grad_norm": 0.10819654166698456, "learning_rate": 5.3489046859132455e-05, "loss": 0.0067, "step": 12924 }, { "epoch": 1.4012359063313096, "grad_norm": 0.3017486333847046, "learning_rate": 5.348541999129551e-05, "loss": 0.0179, "step": 12925 }, { "epoch": 1.4013443191673893, "grad_norm": 0.4343161880970001, "learning_rate": 5.348179312345858e-05, "loss": 0.024, "step": 12926 }, { "epoch": 1.4014527320034693, "grad_norm": 0.0646328553557396, "learning_rate": 5.347816625562164e-05, "loss": 0.0048, "step": 12927 }, { "epoch": 1.401561144839549, "grad_norm": 0.12269652634859085, "learning_rate": 5.347453938778472e-05, "loss": 0.0131, "step": 12928 }, { "epoch": 1.4016695576756288, "grad_norm": 0.21691836416721344, "learning_rate": 5.347091251994778e-05, "loss": 0.0056, "step": 12929 }, { "epoch": 1.4017779705117086, "grad_norm": 0.4082823097705841, "learning_rate": 5.346728565211085e-05, "loss": 0.0342, "step": 12930 }, { "epoch": 1.4018863833477884, "grad_norm": 0.5089136958122253, "learning_rate": 5.3463658784273904e-05, "loss": 0.0201, "step": 12931 }, { "epoch": 1.4019947961838681, "grad_norm": 0.18975205719470978, "learning_rate": 5.346003191643697e-05, "loss": 0.0107, "step": 12932 }, { "epoch": 1.402103209019948, "grad_norm": 0.7104887962341309, "learning_rate": 5.345640504860003e-05, "loss": 0.0792, "step": 12933 }, { "epoch": 1.4022116218560279, "grad_norm": 0.40918558835983276, "learning_rate": 5.3452778180763096e-05, "loss": 0.0316, "step": 12934 }, { "epoch": 1.4023200346921074, "grad_norm": 0.9991421103477478, "learning_rate": 5.344915131292616e-05, "loss": 0.055, "step": 12935 }, { "epoch": 1.4024284475281874, "grad_norm": 0.6803507804870605, "learning_rate": 5.3445524445089225e-05, "loss": 0.0476, "step": 12936 }, { "epoch": 1.4025368603642672, "grad_norm": 0.3877628445625305, "learning_rate": 5.344189757725229e-05, "loss": 0.0197, "step": 12937 }, { "epoch": 1.402645273200347, "grad_norm": 0.4172770380973816, "learning_rate": 5.343827070941535e-05, "loss": 0.0257, "step": 12938 }, { "epoch": 1.4027536860364267, "grad_norm": 0.3641771376132965, "learning_rate": 5.343464384157841e-05, "loss": 0.0169, "step": 12939 }, { "epoch": 1.4028620988725065, "grad_norm": 0.27767106890678406, "learning_rate": 5.3431016973741474e-05, "loss": 0.0171, "step": 12940 }, { "epoch": 1.4029705117085862, "grad_norm": 0.32526299357414246, "learning_rate": 5.342739010590454e-05, "loss": 0.0378, "step": 12941 }, { "epoch": 1.403078924544666, "grad_norm": 0.38540151715278625, "learning_rate": 5.34237632380676e-05, "loss": 0.024, "step": 12942 }, { "epoch": 1.403187337380746, "grad_norm": 0.2556844651699066, "learning_rate": 5.3420136370230667e-05, "loss": 0.0293, "step": 12943 }, { "epoch": 1.4032957502168257, "grad_norm": 0.4113031029701233, "learning_rate": 5.341650950239373e-05, "loss": 0.0245, "step": 12944 }, { "epoch": 1.4034041630529055, "grad_norm": 0.6176387071609497, "learning_rate": 5.34128826345568e-05, "loss": 0.065, "step": 12945 }, { "epoch": 1.4035125758889853, "grad_norm": 0.2795133888721466, "learning_rate": 5.3409255766719866e-05, "loss": 0.0063, "step": 12946 }, { "epoch": 1.403620988725065, "grad_norm": 0.1281915158033371, "learning_rate": 5.340562889888293e-05, "loss": 0.01, "step": 12947 }, { "epoch": 1.4037294015611448, "grad_norm": 0.06146085262298584, "learning_rate": 5.3402002031045994e-05, "loss": 0.003, "step": 12948 }, { "epoch": 1.4038378143972245, "grad_norm": 0.12373367697000504, "learning_rate": 5.339837516320906e-05, "loss": 0.0048, "step": 12949 }, { "epoch": 1.4039462272333045, "grad_norm": 0.1047930121421814, "learning_rate": 5.339474829537212e-05, "loss": 0.0086, "step": 12950 }, { "epoch": 1.4040546400693843, "grad_norm": 0.5408543348312378, "learning_rate": 5.3391121427535186e-05, "loss": 0.0716, "step": 12951 }, { "epoch": 1.404163052905464, "grad_norm": 0.38535815477371216, "learning_rate": 5.338749455969825e-05, "loss": 0.0221, "step": 12952 }, { "epoch": 1.4042714657415438, "grad_norm": 0.5990883708000183, "learning_rate": 5.3383867691861314e-05, "loss": 0.0468, "step": 12953 }, { "epoch": 1.4043798785776236, "grad_norm": 0.4831818640232086, "learning_rate": 5.338024082402437e-05, "loss": 0.0356, "step": 12954 }, { "epoch": 1.4044882914137033, "grad_norm": 0.35344454646110535, "learning_rate": 5.3376613956187436e-05, "loss": 0.024, "step": 12955 }, { "epoch": 1.4045967042497831, "grad_norm": 0.8121089935302734, "learning_rate": 5.33729870883505e-05, "loss": 0.056, "step": 12956 }, { "epoch": 1.404705117085863, "grad_norm": 0.07244604825973511, "learning_rate": 5.3369360220513564e-05, "loss": 0.005, "step": 12957 }, { "epoch": 1.4048135299219426, "grad_norm": 0.18833112716674805, "learning_rate": 5.336573335267663e-05, "loss": 0.0245, "step": 12958 }, { "epoch": 1.4049219427580226, "grad_norm": 0.7658844590187073, "learning_rate": 5.336210648483969e-05, "loss": 0.0454, "step": 12959 }, { "epoch": 1.4050303555941024, "grad_norm": 0.28640618920326233, "learning_rate": 5.3358479617002756e-05, "loss": 0.0159, "step": 12960 }, { "epoch": 1.4051387684301822, "grad_norm": 0.3834720551967621, "learning_rate": 5.335485274916582e-05, "loss": 0.0287, "step": 12961 }, { "epoch": 1.405247181266262, "grad_norm": 0.8218194842338562, "learning_rate": 5.335122588132889e-05, "loss": 0.0806, "step": 12962 }, { "epoch": 1.4053555941023417, "grad_norm": 0.1275855451822281, "learning_rate": 5.3347599013491955e-05, "loss": 0.0104, "step": 12963 }, { "epoch": 1.4054640069384214, "grad_norm": 0.3517224192619324, "learning_rate": 5.334397214565502e-05, "loss": 0.023, "step": 12964 }, { "epoch": 1.4055724197745012, "grad_norm": 0.17502763867378235, "learning_rate": 5.3340345277818084e-05, "loss": 0.0068, "step": 12965 }, { "epoch": 1.4056808326105812, "grad_norm": 0.13244186341762543, "learning_rate": 5.333671840998115e-05, "loss": 0.0057, "step": 12966 }, { "epoch": 1.405789245446661, "grad_norm": 0.36875054240226746, "learning_rate": 5.333309154214421e-05, "loss": 0.0257, "step": 12967 }, { "epoch": 1.4058976582827407, "grad_norm": 0.9421527981758118, "learning_rate": 5.332946467430727e-05, "loss": 0.0595, "step": 12968 }, { "epoch": 1.4060060711188205, "grad_norm": 0.6478481292724609, "learning_rate": 5.332583780647033e-05, "loss": 0.0696, "step": 12969 }, { "epoch": 1.4061144839549002, "grad_norm": 0.1336262822151184, "learning_rate": 5.33222109386334e-05, "loss": 0.0054, "step": 12970 }, { "epoch": 1.40622289679098, "grad_norm": 0.24372538924217224, "learning_rate": 5.331858407079646e-05, "loss": 0.013, "step": 12971 }, { "epoch": 1.4063313096270598, "grad_norm": 0.16136732697486877, "learning_rate": 5.3314957202959526e-05, "loss": 0.0147, "step": 12972 }, { "epoch": 1.4064397224631398, "grad_norm": 0.43488234281539917, "learning_rate": 5.331133033512259e-05, "loss": 0.0166, "step": 12973 }, { "epoch": 1.4065481352992193, "grad_norm": 0.13604003190994263, "learning_rate": 5.3307703467285654e-05, "loss": 0.0079, "step": 12974 }, { "epoch": 1.4066565481352993, "grad_norm": 0.3542327880859375, "learning_rate": 5.330407659944872e-05, "loss": 0.0224, "step": 12975 }, { "epoch": 1.406764960971379, "grad_norm": 0.2562837600708008, "learning_rate": 5.330044973161178e-05, "loss": 0.0413, "step": 12976 }, { "epoch": 1.4068733738074588, "grad_norm": 0.1412433683872223, "learning_rate": 5.329682286377484e-05, "loss": 0.0115, "step": 12977 }, { "epoch": 1.4069817866435386, "grad_norm": 0.1517181694507599, "learning_rate": 5.3293195995937903e-05, "loss": 0.0157, "step": 12978 }, { "epoch": 1.4070901994796183, "grad_norm": 0.2918857932090759, "learning_rate": 5.328956912810097e-05, "loss": 0.0049, "step": 12979 }, { "epoch": 1.4071986123156983, "grad_norm": 0.4263109862804413, "learning_rate": 5.3285942260264045e-05, "loss": 0.0622, "step": 12980 }, { "epoch": 1.4073070251517779, "grad_norm": 0.43451830744743347, "learning_rate": 5.328231539242711e-05, "loss": 0.0383, "step": 12981 }, { "epoch": 1.4074154379878578, "grad_norm": 0.12923182547092438, "learning_rate": 5.327868852459017e-05, "loss": 0.0078, "step": 12982 }, { "epoch": 1.4075238508239376, "grad_norm": 0.25691425800323486, "learning_rate": 5.327506165675323e-05, "loss": 0.027, "step": 12983 }, { "epoch": 1.4076322636600174, "grad_norm": 0.24097391963005066, "learning_rate": 5.3271434788916295e-05, "loss": 0.0282, "step": 12984 }, { "epoch": 1.4077406764960971, "grad_norm": 0.3994257152080536, "learning_rate": 5.326780792107936e-05, "loss": 0.0407, "step": 12985 }, { "epoch": 1.407849089332177, "grad_norm": 0.4136520326137543, "learning_rate": 5.326418105324242e-05, "loss": 0.0208, "step": 12986 }, { "epoch": 1.4079575021682567, "grad_norm": 0.10084924101829529, "learning_rate": 5.326055418540549e-05, "loss": 0.0124, "step": 12987 }, { "epoch": 1.4080659150043364, "grad_norm": 0.1790066361427307, "learning_rate": 5.325692731756855e-05, "loss": 0.0121, "step": 12988 }, { "epoch": 1.4081743278404164, "grad_norm": 0.05344798043370247, "learning_rate": 5.3253300449731615e-05, "loss": 0.0036, "step": 12989 }, { "epoch": 1.4082827406764962, "grad_norm": 0.13583748042583466, "learning_rate": 5.324967358189468e-05, "loss": 0.017, "step": 12990 }, { "epoch": 1.408391153512576, "grad_norm": 0.09920823574066162, "learning_rate": 5.324604671405774e-05, "loss": 0.006, "step": 12991 }, { "epoch": 1.4084995663486557, "grad_norm": 0.07367853075265884, "learning_rate": 5.32424198462208e-05, "loss": 0.0087, "step": 12992 }, { "epoch": 1.4086079791847355, "grad_norm": 0.3938260078430176, "learning_rate": 5.3238792978383865e-05, "loss": 0.0372, "step": 12993 }, { "epoch": 1.4087163920208152, "grad_norm": 0.4411512017250061, "learning_rate": 5.323516611054693e-05, "loss": 0.0273, "step": 12994 }, { "epoch": 1.408824804856895, "grad_norm": 0.47114208340644836, "learning_rate": 5.323153924270999e-05, "loss": 0.0337, "step": 12995 }, { "epoch": 1.408933217692975, "grad_norm": 0.2176007777452469, "learning_rate": 5.322791237487306e-05, "loss": 0.0131, "step": 12996 }, { "epoch": 1.4090416305290545, "grad_norm": 0.34897565841674805, "learning_rate": 5.322428550703613e-05, "loss": 0.022, "step": 12997 }, { "epoch": 1.4091500433651345, "grad_norm": 0.28603145480155945, "learning_rate": 5.322065863919919e-05, "loss": 0.0282, "step": 12998 }, { "epoch": 1.4092584562012143, "grad_norm": 0.1084439754486084, "learning_rate": 5.3217031771362256e-05, "loss": 0.005, "step": 12999 }, { "epoch": 1.409366869037294, "grad_norm": 0.800614595413208, "learning_rate": 5.321340490352532e-05, "loss": 0.0357, "step": 13000 }, { "epoch": 1.4094752818733738, "grad_norm": 0.36591029167175293, "learning_rate": 5.3209778035688385e-05, "loss": 0.0513, "step": 13001 }, { "epoch": 1.4095836947094535, "grad_norm": 0.43548649549484253, "learning_rate": 5.320615116785145e-05, "loss": 0.0461, "step": 13002 }, { "epoch": 1.4096921075455333, "grad_norm": 0.16817940771579742, "learning_rate": 5.320252430001451e-05, "loss": 0.019, "step": 13003 }, { "epoch": 1.409800520381613, "grad_norm": 0.8258457779884338, "learning_rate": 5.319889743217758e-05, "loss": 0.0263, "step": 13004 }, { "epoch": 1.409908933217693, "grad_norm": 0.3349969685077667, "learning_rate": 5.319527056434064e-05, "loss": 0.0199, "step": 13005 }, { "epoch": 1.4100173460537728, "grad_norm": 0.2436937689781189, "learning_rate": 5.31916436965037e-05, "loss": 0.0379, "step": 13006 }, { "epoch": 1.4101257588898526, "grad_norm": 0.3259289562702179, "learning_rate": 5.318801682866676e-05, "loss": 0.0197, "step": 13007 }, { "epoch": 1.4102341717259324, "grad_norm": 0.31959134340286255, "learning_rate": 5.3184389960829827e-05, "loss": 0.0307, "step": 13008 }, { "epoch": 1.4103425845620121, "grad_norm": 0.141441211104393, "learning_rate": 5.318076309299289e-05, "loss": 0.0126, "step": 13009 }, { "epoch": 1.4104509973980919, "grad_norm": 0.30148062109947205, "learning_rate": 5.3177136225155955e-05, "loss": 0.0124, "step": 13010 }, { "epoch": 1.4105594102341716, "grad_norm": 0.1512758731842041, "learning_rate": 5.317350935731902e-05, "loss": 0.0107, "step": 13011 }, { "epoch": 1.4106678230702516, "grad_norm": 0.4642060101032257, "learning_rate": 5.316988248948208e-05, "loss": 0.0358, "step": 13012 }, { "epoch": 1.4107762359063314, "grad_norm": 0.9163885116577148, "learning_rate": 5.316625562164515e-05, "loss": 0.0197, "step": 13013 }, { "epoch": 1.4108846487424112, "grad_norm": 0.2142077386379242, "learning_rate": 5.316262875380822e-05, "loss": 0.0176, "step": 13014 }, { "epoch": 1.410993061578491, "grad_norm": 0.30971628427505493, "learning_rate": 5.315900188597128e-05, "loss": 0.0109, "step": 13015 }, { "epoch": 1.4111014744145707, "grad_norm": 0.17330044507980347, "learning_rate": 5.3155375018134346e-05, "loss": 0.0116, "step": 13016 }, { "epoch": 1.4112098872506504, "grad_norm": 0.10892155766487122, "learning_rate": 5.315174815029741e-05, "loss": 0.0072, "step": 13017 }, { "epoch": 1.4113183000867302, "grad_norm": 0.02485072612762451, "learning_rate": 5.3148121282460474e-05, "loss": 0.0014, "step": 13018 }, { "epoch": 1.4114267129228102, "grad_norm": 0.05377340689301491, "learning_rate": 5.314449441462354e-05, "loss": 0.0033, "step": 13019 }, { "epoch": 1.4115351257588897, "grad_norm": 0.33465898036956787, "learning_rate": 5.3140867546786596e-05, "loss": 0.0181, "step": 13020 }, { "epoch": 1.4116435385949697, "grad_norm": 0.31775277853012085, "learning_rate": 5.313724067894966e-05, "loss": 0.0184, "step": 13021 }, { "epoch": 1.4117519514310495, "grad_norm": 0.21849730610847473, "learning_rate": 5.3133613811112724e-05, "loss": 0.0197, "step": 13022 }, { "epoch": 1.4118603642671292, "grad_norm": 0.7551549077033997, "learning_rate": 5.312998694327579e-05, "loss": 0.0833, "step": 13023 }, { "epoch": 1.411968777103209, "grad_norm": 0.33846768736839294, "learning_rate": 5.312636007543885e-05, "loss": 0.0215, "step": 13024 }, { "epoch": 1.4120771899392888, "grad_norm": 0.42272627353668213, "learning_rate": 5.3122733207601916e-05, "loss": 0.0391, "step": 13025 }, { "epoch": 1.4121856027753685, "grad_norm": 0.06396673619747162, "learning_rate": 5.311910633976498e-05, "loss": 0.0026, "step": 13026 }, { "epoch": 1.4122940156114483, "grad_norm": 0.12244497984647751, "learning_rate": 5.3115479471928045e-05, "loss": 0.0088, "step": 13027 }, { "epoch": 1.4124024284475283, "grad_norm": 0.3249448537826538, "learning_rate": 5.311185260409111e-05, "loss": 0.0327, "step": 13028 }, { "epoch": 1.412510841283608, "grad_norm": 0.15756899118423462, "learning_rate": 5.3108225736254166e-05, "loss": 0.0066, "step": 13029 }, { "epoch": 1.4126192541196878, "grad_norm": 0.08983214199542999, "learning_rate": 5.310459886841723e-05, "loss": 0.0052, "step": 13030 }, { "epoch": 1.4127276669557676, "grad_norm": 0.2491450011730194, "learning_rate": 5.310097200058031e-05, "loss": 0.0133, "step": 13031 }, { "epoch": 1.4128360797918473, "grad_norm": 0.2091214507818222, "learning_rate": 5.309734513274337e-05, "loss": 0.0155, "step": 13032 }, { "epoch": 1.412944492627927, "grad_norm": 0.25206416845321655, "learning_rate": 5.3093718264906436e-05, "loss": 0.0094, "step": 13033 }, { "epoch": 1.4130529054640069, "grad_norm": 0.15804973244667053, "learning_rate": 5.30900913970695e-05, "loss": 0.0107, "step": 13034 }, { "epoch": 1.4131613183000868, "grad_norm": 0.5672937035560608, "learning_rate": 5.308646452923256e-05, "loss": 0.0231, "step": 13035 }, { "epoch": 1.4132697311361664, "grad_norm": 0.6108469367027283, "learning_rate": 5.308283766139562e-05, "loss": 0.0386, "step": 13036 }, { "epoch": 1.4133781439722464, "grad_norm": 0.40040937066078186, "learning_rate": 5.3079210793558686e-05, "loss": 0.023, "step": 13037 }, { "epoch": 1.4134865568083261, "grad_norm": 0.4745422899723053, "learning_rate": 5.307558392572175e-05, "loss": 0.0248, "step": 13038 }, { "epoch": 1.413594969644406, "grad_norm": 0.3491782546043396, "learning_rate": 5.3071957057884814e-05, "loss": 0.0185, "step": 13039 }, { "epoch": 1.4137033824804857, "grad_norm": 0.2722923755645752, "learning_rate": 5.306833019004788e-05, "loss": 0.0622, "step": 13040 }, { "epoch": 1.4138117953165654, "grad_norm": 1.006740927696228, "learning_rate": 5.306470332221094e-05, "loss": 0.0649, "step": 13041 }, { "epoch": 1.4139202081526454, "grad_norm": 0.16257362067699432, "learning_rate": 5.3061076454374006e-05, "loss": 0.0074, "step": 13042 }, { "epoch": 1.414028620988725, "grad_norm": 0.37726613879203796, "learning_rate": 5.305744958653707e-05, "loss": 0.0155, "step": 13043 }, { "epoch": 1.414137033824805, "grad_norm": 0.23619787395000458, "learning_rate": 5.305382271870013e-05, "loss": 0.0166, "step": 13044 }, { "epoch": 1.4142454466608847, "grad_norm": 0.37397557497024536, "learning_rate": 5.305019585086319e-05, "loss": 0.0242, "step": 13045 }, { "epoch": 1.4143538594969645, "grad_norm": 0.3995720446109772, "learning_rate": 5.3046568983026256e-05, "loss": 0.0205, "step": 13046 }, { "epoch": 1.4144622723330442, "grad_norm": 0.7476778626441956, "learning_rate": 5.304294211518932e-05, "loss": 0.0892, "step": 13047 }, { "epoch": 1.414570685169124, "grad_norm": 0.34009426832199097, "learning_rate": 5.3039315247352384e-05, "loss": 0.0103, "step": 13048 }, { "epoch": 1.4146790980052038, "grad_norm": 0.1886848509311676, "learning_rate": 5.3035688379515455e-05, "loss": 0.0126, "step": 13049 }, { "epoch": 1.4147875108412835, "grad_norm": 0.56778484582901, "learning_rate": 5.303206151167852e-05, "loss": 0.0666, "step": 13050 }, { "epoch": 1.4148959236773635, "grad_norm": 0.7790238261222839, "learning_rate": 5.302843464384158e-05, "loss": 0.0341, "step": 13051 }, { "epoch": 1.4150043365134433, "grad_norm": 0.44478869438171387, "learning_rate": 5.302480777600465e-05, "loss": 0.0414, "step": 13052 }, { "epoch": 1.415112749349523, "grad_norm": 0.31953930854797363, "learning_rate": 5.302118090816771e-05, "loss": 0.0554, "step": 13053 }, { "epoch": 1.4152211621856028, "grad_norm": 0.397726446390152, "learning_rate": 5.3017554040330775e-05, "loss": 0.0415, "step": 13054 }, { "epoch": 1.4153295750216826, "grad_norm": 0.2909582853317261, "learning_rate": 5.301392717249384e-05, "loss": 0.0149, "step": 13055 }, { "epoch": 1.4154379878577623, "grad_norm": 0.15415571630001068, "learning_rate": 5.3010300304656904e-05, "loss": 0.0112, "step": 13056 }, { "epoch": 1.415546400693842, "grad_norm": 0.17115125060081482, "learning_rate": 5.300667343681997e-05, "loss": 0.0241, "step": 13057 }, { "epoch": 1.415654813529922, "grad_norm": 0.43739205598831177, "learning_rate": 5.3003046568983025e-05, "loss": 0.0397, "step": 13058 }, { "epoch": 1.4157632263660016, "grad_norm": 0.5228233337402344, "learning_rate": 5.299941970114609e-05, "loss": 0.0313, "step": 13059 }, { "epoch": 1.4158716392020816, "grad_norm": 0.3877914547920227, "learning_rate": 5.299579283330915e-05, "loss": 0.0429, "step": 13060 }, { "epoch": 1.4159800520381614, "grad_norm": 0.2564733326435089, "learning_rate": 5.299216596547222e-05, "loss": 0.0297, "step": 13061 }, { "epoch": 1.4160884648742411, "grad_norm": 0.6848358511924744, "learning_rate": 5.298853909763528e-05, "loss": 0.038, "step": 13062 }, { "epoch": 1.4161968777103209, "grad_norm": 0.48747581243515015, "learning_rate": 5.2984912229798346e-05, "loss": 0.0299, "step": 13063 }, { "epoch": 1.4163052905464006, "grad_norm": 0.5558979511260986, "learning_rate": 5.298128536196141e-05, "loss": 0.0626, "step": 13064 }, { "epoch": 1.4164137033824804, "grad_norm": 0.1968277543783188, "learning_rate": 5.2977658494124474e-05, "loss": 0.0156, "step": 13065 }, { "epoch": 1.4165221162185602, "grad_norm": 0.2748720049858093, "learning_rate": 5.2974031626287545e-05, "loss": 0.0147, "step": 13066 }, { "epoch": 1.4166305290546402, "grad_norm": 0.6401991844177246, "learning_rate": 5.297040475845061e-05, "loss": 0.0119, "step": 13067 }, { "epoch": 1.41673894189072, "grad_norm": 0.3231610655784607, "learning_rate": 5.296677789061367e-05, "loss": 0.0166, "step": 13068 }, { "epoch": 1.4168473547267997, "grad_norm": 0.2521341145038605, "learning_rate": 5.296315102277674e-05, "loss": 0.0396, "step": 13069 }, { "epoch": 1.4169557675628794, "grad_norm": 0.28247687220573425, "learning_rate": 5.29595241549398e-05, "loss": 0.018, "step": 13070 }, { "epoch": 1.4170641803989592, "grad_norm": 0.3806057870388031, "learning_rate": 5.2955897287102865e-05, "loss": 0.0337, "step": 13071 }, { "epoch": 1.417172593235039, "grad_norm": 0.0890340581536293, "learning_rate": 5.295227041926592e-05, "loss": 0.0122, "step": 13072 }, { "epoch": 1.4172810060711187, "grad_norm": 0.2617824375629425, "learning_rate": 5.2948643551428987e-05, "loss": 0.0498, "step": 13073 }, { "epoch": 1.4173894189071987, "grad_norm": 0.5263274908065796, "learning_rate": 5.294501668359205e-05, "loss": 0.0104, "step": 13074 }, { "epoch": 1.4174978317432785, "grad_norm": 0.6958942413330078, "learning_rate": 5.2941389815755115e-05, "loss": 0.0336, "step": 13075 }, { "epoch": 1.4176062445793582, "grad_norm": 0.26493188738822937, "learning_rate": 5.293776294791818e-05, "loss": 0.0277, "step": 13076 }, { "epoch": 1.417714657415438, "grad_norm": 0.2904948890209198, "learning_rate": 5.293413608008124e-05, "loss": 0.0219, "step": 13077 }, { "epoch": 1.4178230702515178, "grad_norm": 0.6319770812988281, "learning_rate": 5.293050921224431e-05, "loss": 0.06, "step": 13078 }, { "epoch": 1.4179314830875975, "grad_norm": 0.2236863672733307, "learning_rate": 5.292688234440737e-05, "loss": 0.0309, "step": 13079 }, { "epoch": 1.4180398959236773, "grad_norm": 0.4488530457019806, "learning_rate": 5.2923255476570435e-05, "loss": 0.0142, "step": 13080 }, { "epoch": 1.4181483087597573, "grad_norm": 0.2416628897190094, "learning_rate": 5.291962860873349e-05, "loss": 0.0209, "step": 13081 }, { "epoch": 1.4182567215958368, "grad_norm": 0.3221261501312256, "learning_rate": 5.291600174089656e-05, "loss": 0.0312, "step": 13082 }, { "epoch": 1.4183651344319168, "grad_norm": 0.34686413407325745, "learning_rate": 5.2912374873059634e-05, "loss": 0.0171, "step": 13083 }, { "epoch": 1.4184735472679966, "grad_norm": 0.6831564903259277, "learning_rate": 5.29087480052227e-05, "loss": 0.0183, "step": 13084 }, { "epoch": 1.4185819601040763, "grad_norm": 0.18571263551712036, "learning_rate": 5.290512113738576e-05, "loss": 0.0189, "step": 13085 }, { "epoch": 1.418690372940156, "grad_norm": 0.21809785068035126, "learning_rate": 5.290149426954883e-05, "loss": 0.0132, "step": 13086 }, { "epoch": 1.4187987857762359, "grad_norm": 0.11254069209098816, "learning_rate": 5.2897867401711884e-05, "loss": 0.0087, "step": 13087 }, { "epoch": 1.4189071986123156, "grad_norm": 0.24576754868030548, "learning_rate": 5.289424053387495e-05, "loss": 0.0158, "step": 13088 }, { "epoch": 1.4190156114483954, "grad_norm": 0.23264822363853455, "learning_rate": 5.289061366603801e-05, "loss": 0.0224, "step": 13089 }, { "epoch": 1.4191240242844754, "grad_norm": 0.26559197902679443, "learning_rate": 5.2886986798201076e-05, "loss": 0.0293, "step": 13090 }, { "epoch": 1.4192324371205551, "grad_norm": 0.1476258933544159, "learning_rate": 5.288335993036414e-05, "loss": 0.0271, "step": 13091 }, { "epoch": 1.419340849956635, "grad_norm": 0.17524273693561554, "learning_rate": 5.2879733062527205e-05, "loss": 0.0097, "step": 13092 }, { "epoch": 1.4194492627927147, "grad_norm": 0.12286711484193802, "learning_rate": 5.287610619469027e-05, "loss": 0.0103, "step": 13093 }, { "epoch": 1.4195576756287944, "grad_norm": 0.11035891622304916, "learning_rate": 5.287247932685333e-05, "loss": 0.0076, "step": 13094 }, { "epoch": 1.4196660884648742, "grad_norm": 0.18131686747074127, "learning_rate": 5.28688524590164e-05, "loss": 0.0109, "step": 13095 }, { "epoch": 1.419774501300954, "grad_norm": 0.16362158954143524, "learning_rate": 5.2865225591179454e-05, "loss": 0.0162, "step": 13096 }, { "epoch": 1.419882914137034, "grad_norm": 0.24254553020000458, "learning_rate": 5.286159872334252e-05, "loss": 0.0198, "step": 13097 }, { "epoch": 1.4199913269731137, "grad_norm": 0.35995951294898987, "learning_rate": 5.285797185550558e-05, "loss": 0.0457, "step": 13098 }, { "epoch": 1.4200997398091935, "grad_norm": 0.11165221780538559, "learning_rate": 5.2854344987668647e-05, "loss": 0.0041, "step": 13099 }, { "epoch": 1.4202081526452732, "grad_norm": 0.10332922637462616, "learning_rate": 5.2850718119831724e-05, "loss": 0.0076, "step": 13100 }, { "epoch": 1.420316565481353, "grad_norm": 0.1362026482820511, "learning_rate": 5.284709125199478e-05, "loss": 0.0072, "step": 13101 }, { "epoch": 1.4204249783174328, "grad_norm": 0.24362967908382416, "learning_rate": 5.2843464384157846e-05, "loss": 0.0329, "step": 13102 }, { "epoch": 1.4205333911535125, "grad_norm": 0.1302005648612976, "learning_rate": 5.283983751632091e-05, "loss": 0.016, "step": 13103 }, { "epoch": 1.4206418039895925, "grad_norm": 0.37397798895835876, "learning_rate": 5.2836210648483974e-05, "loss": 0.0096, "step": 13104 }, { "epoch": 1.420750216825672, "grad_norm": 0.4695297181606293, "learning_rate": 5.283258378064704e-05, "loss": 0.0355, "step": 13105 }, { "epoch": 1.420858629661752, "grad_norm": 0.6023648381233215, "learning_rate": 5.28289569128101e-05, "loss": 0.0625, "step": 13106 }, { "epoch": 1.4209670424978318, "grad_norm": 0.16106538474559784, "learning_rate": 5.2825330044973166e-05, "loss": 0.0073, "step": 13107 }, { "epoch": 1.4210754553339116, "grad_norm": 0.1879245936870575, "learning_rate": 5.282170317713623e-05, "loss": 0.0214, "step": 13108 }, { "epoch": 1.4211838681699913, "grad_norm": 0.3174498379230499, "learning_rate": 5.2818076309299294e-05, "loss": 0.0335, "step": 13109 }, { "epoch": 1.421292281006071, "grad_norm": 0.12626196444034576, "learning_rate": 5.281444944146235e-05, "loss": 0.0035, "step": 13110 }, { "epoch": 1.4214006938421508, "grad_norm": 0.16155897080898285, "learning_rate": 5.2810822573625416e-05, "loss": 0.0093, "step": 13111 }, { "epoch": 1.4215091066782306, "grad_norm": 0.3400628864765167, "learning_rate": 5.280719570578848e-05, "loss": 0.0195, "step": 13112 }, { "epoch": 1.4216175195143106, "grad_norm": 0.720925509929657, "learning_rate": 5.2803568837951544e-05, "loss": 0.0252, "step": 13113 }, { "epoch": 1.4217259323503904, "grad_norm": 0.32998114824295044, "learning_rate": 5.279994197011461e-05, "loss": 0.0185, "step": 13114 }, { "epoch": 1.4218343451864701, "grad_norm": 0.35337182879447937, "learning_rate": 5.279631510227767e-05, "loss": 0.0646, "step": 13115 }, { "epoch": 1.4219427580225499, "grad_norm": 0.4354027211666107, "learning_rate": 5.2792688234440736e-05, "loss": 0.035, "step": 13116 }, { "epoch": 1.4220511708586296, "grad_norm": 0.739544689655304, "learning_rate": 5.27890613666038e-05, "loss": 0.0512, "step": 13117 }, { "epoch": 1.4221595836947094, "grad_norm": 0.23296701908111572, "learning_rate": 5.278543449876687e-05, "loss": 0.0071, "step": 13118 }, { "epoch": 1.4222679965307892, "grad_norm": 0.18492329120635986, "learning_rate": 5.2781807630929935e-05, "loss": 0.0122, "step": 13119 }, { "epoch": 1.4223764093668692, "grad_norm": 0.5075319409370422, "learning_rate": 5.2778180763093e-05, "loss": 0.0454, "step": 13120 }, { "epoch": 1.4224848222029487, "grad_norm": 0.03084600530564785, "learning_rate": 5.2774553895256064e-05, "loss": 0.0015, "step": 13121 }, { "epoch": 1.4225932350390287, "grad_norm": 0.5625961422920227, "learning_rate": 5.277092702741913e-05, "loss": 0.0559, "step": 13122 }, { "epoch": 1.4227016478751084, "grad_norm": 0.0633685514330864, "learning_rate": 5.276730015958219e-05, "loss": 0.0036, "step": 13123 }, { "epoch": 1.4228100607111882, "grad_norm": 0.2618345618247986, "learning_rate": 5.276367329174525e-05, "loss": 0.0305, "step": 13124 }, { "epoch": 1.422918473547268, "grad_norm": 0.1090574860572815, "learning_rate": 5.276004642390831e-05, "loss": 0.0062, "step": 13125 }, { "epoch": 1.4230268863833477, "grad_norm": 0.5270761847496033, "learning_rate": 5.275641955607138e-05, "loss": 0.03, "step": 13126 }, { "epoch": 1.4231352992194277, "grad_norm": 0.29321110248565674, "learning_rate": 5.275279268823444e-05, "loss": 0.0218, "step": 13127 }, { "epoch": 1.4232437120555073, "grad_norm": 0.14486579596996307, "learning_rate": 5.2749165820397506e-05, "loss": 0.005, "step": 13128 }, { "epoch": 1.4233521248915872, "grad_norm": 0.07647547125816345, "learning_rate": 5.274553895256057e-05, "loss": 0.0046, "step": 13129 }, { "epoch": 1.423460537727667, "grad_norm": 0.30863696336746216, "learning_rate": 5.2741912084723634e-05, "loss": 0.0272, "step": 13130 }, { "epoch": 1.4235689505637468, "grad_norm": 0.22110582888126373, "learning_rate": 5.27382852168867e-05, "loss": 0.01, "step": 13131 }, { "epoch": 1.4236773633998265, "grad_norm": 0.09375949949026108, "learning_rate": 5.273465834904976e-05, "loss": 0.0046, "step": 13132 }, { "epoch": 1.4237857762359063, "grad_norm": 0.41543540358543396, "learning_rate": 5.273103148121282e-05, "loss": 0.0484, "step": 13133 }, { "epoch": 1.423894189071986, "grad_norm": 0.3165193200111389, "learning_rate": 5.2727404613375883e-05, "loss": 0.023, "step": 13134 }, { "epoch": 1.4240026019080658, "grad_norm": 0.16575343906879425, "learning_rate": 5.272377774553896e-05, "loss": 0.0116, "step": 13135 }, { "epoch": 1.4241110147441458, "grad_norm": 0.1423223465681076, "learning_rate": 5.2720150877702025e-05, "loss": 0.0122, "step": 13136 }, { "epoch": 1.4242194275802256, "grad_norm": 0.18108366429805756, "learning_rate": 5.271652400986509e-05, "loss": 0.0247, "step": 13137 }, { "epoch": 1.4243278404163053, "grad_norm": 0.6881853938102722, "learning_rate": 5.2712897142028153e-05, "loss": 0.043, "step": 13138 }, { "epoch": 1.424436253252385, "grad_norm": 0.5097973942756653, "learning_rate": 5.270927027419121e-05, "loss": 0.0293, "step": 13139 }, { "epoch": 1.4245446660884649, "grad_norm": 0.4632056653499603, "learning_rate": 5.2705643406354275e-05, "loss": 0.0118, "step": 13140 }, { "epoch": 1.4246530789245446, "grad_norm": 0.23212246596813202, "learning_rate": 5.270201653851734e-05, "loss": 0.0168, "step": 13141 }, { "epoch": 1.4247614917606244, "grad_norm": 0.28523263335227966, "learning_rate": 5.26983896706804e-05, "loss": 0.0259, "step": 13142 }, { "epoch": 1.4248699045967044, "grad_norm": 0.11356202512979507, "learning_rate": 5.269476280284347e-05, "loss": 0.011, "step": 13143 }, { "epoch": 1.424978317432784, "grad_norm": 0.5799917578697205, "learning_rate": 5.269113593500653e-05, "loss": 0.0177, "step": 13144 }, { "epoch": 1.425086730268864, "grad_norm": 0.15792785584926605, "learning_rate": 5.2687509067169595e-05, "loss": 0.008, "step": 13145 }, { "epoch": 1.4251951431049437, "grad_norm": 0.5212429165840149, "learning_rate": 5.268388219933266e-05, "loss": 0.0125, "step": 13146 }, { "epoch": 1.4253035559410234, "grad_norm": 0.5549488663673401, "learning_rate": 5.2680255331495724e-05, "loss": 0.0163, "step": 13147 }, { "epoch": 1.4254119687771032, "grad_norm": 0.3650347888469696, "learning_rate": 5.267662846365878e-05, "loss": 0.0495, "step": 13148 }, { "epoch": 1.425520381613183, "grad_norm": 0.08979831635951996, "learning_rate": 5.2673001595821845e-05, "loss": 0.0039, "step": 13149 }, { "epoch": 1.4256287944492627, "grad_norm": 0.1862143576145172, "learning_rate": 5.266937472798491e-05, "loss": 0.0116, "step": 13150 }, { "epoch": 1.4257372072853425, "grad_norm": 0.14594003558158875, "learning_rate": 5.266574786014797e-05, "loss": 0.0087, "step": 13151 }, { "epoch": 1.4258456201214225, "grad_norm": 0.5505999326705933, "learning_rate": 5.266212099231105e-05, "loss": 0.0335, "step": 13152 }, { "epoch": 1.4259540329575022, "grad_norm": 0.2340489774942398, "learning_rate": 5.265849412447411e-05, "loss": 0.0179, "step": 13153 }, { "epoch": 1.426062445793582, "grad_norm": 0.7127863168716431, "learning_rate": 5.265486725663717e-05, "loss": 0.0274, "step": 13154 }, { "epoch": 1.4261708586296618, "grad_norm": 0.712834894657135, "learning_rate": 5.2651240388800236e-05, "loss": 0.0273, "step": 13155 }, { "epoch": 1.4262792714657415, "grad_norm": 0.18958839774131775, "learning_rate": 5.26476135209633e-05, "loss": 0.0221, "step": 13156 }, { "epoch": 1.4263876843018213, "grad_norm": 0.12729290127754211, "learning_rate": 5.2643986653126365e-05, "loss": 0.0075, "step": 13157 }, { "epoch": 1.426496097137901, "grad_norm": 0.5536417961120605, "learning_rate": 5.264035978528943e-05, "loss": 0.0176, "step": 13158 }, { "epoch": 1.426604509973981, "grad_norm": 0.4274177551269531, "learning_rate": 5.263673291745249e-05, "loss": 0.0333, "step": 13159 }, { "epoch": 1.4267129228100608, "grad_norm": 0.07339891046285629, "learning_rate": 5.263310604961556e-05, "loss": 0.0019, "step": 13160 }, { "epoch": 1.4268213356461406, "grad_norm": 0.23276598751544952, "learning_rate": 5.262947918177862e-05, "loss": 0.0237, "step": 13161 }, { "epoch": 1.4269297484822203, "grad_norm": 0.2776612639427185, "learning_rate": 5.262585231394168e-05, "loss": 0.0219, "step": 13162 }, { "epoch": 1.4270381613183, "grad_norm": 0.5365585088729858, "learning_rate": 5.262222544610474e-05, "loss": 0.0827, "step": 13163 }, { "epoch": 1.4271465741543798, "grad_norm": 0.05500997602939606, "learning_rate": 5.261859857826781e-05, "loss": 0.0028, "step": 13164 }, { "epoch": 1.4272549869904596, "grad_norm": 0.5289893746376038, "learning_rate": 5.261497171043087e-05, "loss": 0.0549, "step": 13165 }, { "epoch": 1.4273633998265396, "grad_norm": 1.1002817153930664, "learning_rate": 5.2611344842593935e-05, "loss": 0.0868, "step": 13166 }, { "epoch": 1.4274718126626191, "grad_norm": 0.5306251049041748, "learning_rate": 5.2607717974757e-05, "loss": 0.0318, "step": 13167 }, { "epoch": 1.4275802254986991, "grad_norm": 0.17228497564792633, "learning_rate": 5.260409110692006e-05, "loss": 0.0134, "step": 13168 }, { "epoch": 1.4276886383347789, "grad_norm": 0.06418590247631073, "learning_rate": 5.260046423908313e-05, "loss": 0.0035, "step": 13169 }, { "epoch": 1.4277970511708586, "grad_norm": 0.318004310131073, "learning_rate": 5.25968373712462e-05, "loss": 0.0102, "step": 13170 }, { "epoch": 1.4279054640069384, "grad_norm": 0.20054952800273895, "learning_rate": 5.259321050340926e-05, "loss": 0.0148, "step": 13171 }, { "epoch": 1.4280138768430182, "grad_norm": 0.2487577199935913, "learning_rate": 5.2589583635572326e-05, "loss": 0.0142, "step": 13172 }, { "epoch": 1.428122289679098, "grad_norm": 0.3106224536895752, "learning_rate": 5.258595676773539e-05, "loss": 0.0151, "step": 13173 }, { "epoch": 1.4282307025151777, "grad_norm": 0.1654895842075348, "learning_rate": 5.2582329899898454e-05, "loss": 0.0088, "step": 13174 }, { "epoch": 1.4283391153512577, "grad_norm": 0.06869856268167496, "learning_rate": 5.257870303206152e-05, "loss": 0.0026, "step": 13175 }, { "epoch": 1.4284475281873374, "grad_norm": 0.30900028347969055, "learning_rate": 5.2575076164224576e-05, "loss": 0.0088, "step": 13176 }, { "epoch": 1.4285559410234172, "grad_norm": 0.021832622587680817, "learning_rate": 5.257144929638764e-05, "loss": 0.001, "step": 13177 }, { "epoch": 1.428664353859497, "grad_norm": 0.30570411682128906, "learning_rate": 5.2567822428550704e-05, "loss": 0.0185, "step": 13178 }, { "epoch": 1.4287727666955767, "grad_norm": 0.04886328801512718, "learning_rate": 5.256419556071377e-05, "loss": 0.0039, "step": 13179 }, { "epoch": 1.4288811795316565, "grad_norm": 0.3053826093673706, "learning_rate": 5.256056869287683e-05, "loss": 0.0126, "step": 13180 }, { "epoch": 1.4289895923677363, "grad_norm": 0.4287624657154083, "learning_rate": 5.2556941825039896e-05, "loss": 0.0196, "step": 13181 }, { "epoch": 1.4290980052038162, "grad_norm": 0.2626650035381317, "learning_rate": 5.255331495720296e-05, "loss": 0.0162, "step": 13182 }, { "epoch": 1.4292064180398958, "grad_norm": 1.012813687324524, "learning_rate": 5.2549688089366025e-05, "loss": 0.0228, "step": 13183 }, { "epoch": 1.4293148308759758, "grad_norm": 0.8894978165626526, "learning_rate": 5.254606122152909e-05, "loss": 0.0628, "step": 13184 }, { "epoch": 1.4294232437120555, "grad_norm": 0.28161704540252686, "learning_rate": 5.2542434353692146e-05, "loss": 0.01, "step": 13185 }, { "epoch": 1.4295316565481353, "grad_norm": 0.18412792682647705, "learning_rate": 5.253880748585521e-05, "loss": 0.0141, "step": 13186 }, { "epoch": 1.429640069384215, "grad_norm": 0.23001593351364136, "learning_rate": 5.253518061801829e-05, "loss": 0.011, "step": 13187 }, { "epoch": 1.4297484822202948, "grad_norm": 0.3074108064174652, "learning_rate": 5.253155375018135e-05, "loss": 0.0124, "step": 13188 }, { "epoch": 1.4298568950563748, "grad_norm": 0.19153855741024017, "learning_rate": 5.2527926882344416e-05, "loss": 0.0102, "step": 13189 }, { "epoch": 1.4299653078924544, "grad_norm": 0.3401869833469391, "learning_rate": 5.252430001450748e-05, "loss": 0.0274, "step": 13190 }, { "epoch": 1.4300737207285343, "grad_norm": 0.09263836592435837, "learning_rate": 5.252067314667054e-05, "loss": 0.0049, "step": 13191 }, { "epoch": 1.430182133564614, "grad_norm": 0.05982242152094841, "learning_rate": 5.25170462788336e-05, "loss": 0.0054, "step": 13192 }, { "epoch": 1.4302905464006939, "grad_norm": 0.03605937212705612, "learning_rate": 5.2513419410996666e-05, "loss": 0.0025, "step": 13193 }, { "epoch": 1.4303989592367736, "grad_norm": 0.5199281573295593, "learning_rate": 5.250979254315973e-05, "loss": 0.0341, "step": 13194 }, { "epoch": 1.4305073720728534, "grad_norm": 0.2626664638519287, "learning_rate": 5.2506165675322794e-05, "loss": 0.0084, "step": 13195 }, { "epoch": 1.4306157849089332, "grad_norm": 0.1085224524140358, "learning_rate": 5.250253880748586e-05, "loss": 0.0055, "step": 13196 }, { "epoch": 1.430724197745013, "grad_norm": 0.16251040995121002, "learning_rate": 5.249891193964892e-05, "loss": 0.006, "step": 13197 }, { "epoch": 1.430832610581093, "grad_norm": 0.21554845571517944, "learning_rate": 5.2495285071811986e-05, "loss": 0.0108, "step": 13198 }, { "epoch": 1.4309410234171727, "grad_norm": 0.21354395151138306, "learning_rate": 5.249165820397505e-05, "loss": 0.0054, "step": 13199 }, { "epoch": 1.4310494362532524, "grad_norm": 0.6058434844017029, "learning_rate": 5.248803133613811e-05, "loss": 0.0537, "step": 13200 }, { "epoch": 1.4311578490893322, "grad_norm": 0.4717138707637787, "learning_rate": 5.248440446830117e-05, "loss": 0.0232, "step": 13201 }, { "epoch": 1.431266261925412, "grad_norm": 0.5652360320091248, "learning_rate": 5.2480777600464236e-05, "loss": 0.0378, "step": 13202 }, { "epoch": 1.4313746747614917, "grad_norm": 1.024874210357666, "learning_rate": 5.24771507326273e-05, "loss": 0.0275, "step": 13203 }, { "epoch": 1.4314830875975715, "grad_norm": 0.4469720125198364, "learning_rate": 5.247352386479038e-05, "loss": 0.0451, "step": 13204 }, { "epoch": 1.4315915004336515, "grad_norm": 0.3845040202140808, "learning_rate": 5.2469896996953435e-05, "loss": 0.0185, "step": 13205 }, { "epoch": 1.431699913269731, "grad_norm": 0.04698564484715462, "learning_rate": 5.24662701291165e-05, "loss": 0.0012, "step": 13206 }, { "epoch": 1.431808326105811, "grad_norm": 0.5158476829528809, "learning_rate": 5.246264326127956e-05, "loss": 0.048, "step": 13207 }, { "epoch": 1.4319167389418908, "grad_norm": 0.04684191197156906, "learning_rate": 5.245901639344263e-05, "loss": 0.0028, "step": 13208 }, { "epoch": 1.4320251517779705, "grad_norm": 0.6030712723731995, "learning_rate": 5.245538952560569e-05, "loss": 0.0698, "step": 13209 }, { "epoch": 1.4321335646140503, "grad_norm": 0.30062708258628845, "learning_rate": 5.2451762657768755e-05, "loss": 0.0093, "step": 13210 }, { "epoch": 1.43224197745013, "grad_norm": 0.6978912949562073, "learning_rate": 5.244813578993182e-05, "loss": 0.0592, "step": 13211 }, { "epoch": 1.4323503902862098, "grad_norm": 0.659473180770874, "learning_rate": 5.2444508922094884e-05, "loss": 0.0646, "step": 13212 }, { "epoch": 1.4324588031222896, "grad_norm": 0.3105602562427521, "learning_rate": 5.244088205425795e-05, "loss": 0.024, "step": 13213 }, { "epoch": 1.4325672159583696, "grad_norm": 0.34842449426651, "learning_rate": 5.2437255186421005e-05, "loss": 0.0288, "step": 13214 }, { "epoch": 1.4326756287944493, "grad_norm": 0.15421177446842194, "learning_rate": 5.243362831858407e-05, "loss": 0.0265, "step": 13215 }, { "epoch": 1.432784041630529, "grad_norm": 0.09704767912626266, "learning_rate": 5.243000145074713e-05, "loss": 0.0038, "step": 13216 }, { "epoch": 1.4328924544666088, "grad_norm": 0.4298778176307678, "learning_rate": 5.24263745829102e-05, "loss": 0.0251, "step": 13217 }, { "epoch": 1.4330008673026886, "grad_norm": 0.46381136775016785, "learning_rate": 5.242274771507326e-05, "loss": 0.0308, "step": 13218 }, { "epoch": 1.4331092801387684, "grad_norm": 0.13653239607810974, "learning_rate": 5.2419120847236326e-05, "loss": 0.0057, "step": 13219 }, { "epoch": 1.4332176929748481, "grad_norm": 0.4248875081539154, "learning_rate": 5.241549397939939e-05, "loss": 0.0326, "step": 13220 }, { "epoch": 1.4333261058109281, "grad_norm": 0.31768175959587097, "learning_rate": 5.241186711156246e-05, "loss": 0.0297, "step": 13221 }, { "epoch": 1.4334345186470079, "grad_norm": 0.6459802985191345, "learning_rate": 5.2408240243725525e-05, "loss": 0.0237, "step": 13222 }, { "epoch": 1.4335429314830876, "grad_norm": 0.16985684633255005, "learning_rate": 5.240461337588859e-05, "loss": 0.0125, "step": 13223 }, { "epoch": 1.4336513443191674, "grad_norm": 0.23224249482154846, "learning_rate": 5.240098650805165e-05, "loss": 0.06, "step": 13224 }, { "epoch": 1.4337597571552472, "grad_norm": 0.4565454125404358, "learning_rate": 5.239735964021472e-05, "loss": 0.0494, "step": 13225 }, { "epoch": 1.433868169991327, "grad_norm": 0.30090346932411194, "learning_rate": 5.239373277237778e-05, "loss": 0.0276, "step": 13226 }, { "epoch": 1.4339765828274067, "grad_norm": 0.3966575860977173, "learning_rate": 5.2390105904540845e-05, "loss": 0.0174, "step": 13227 }, { "epoch": 1.4340849956634867, "grad_norm": 0.11163143068552017, "learning_rate": 5.238647903670391e-05, "loss": 0.0062, "step": 13228 }, { "epoch": 1.4341934084995662, "grad_norm": 0.28484225273132324, "learning_rate": 5.238285216886697e-05, "loss": 0.0264, "step": 13229 }, { "epoch": 1.4343018213356462, "grad_norm": 0.14800065755844116, "learning_rate": 5.237922530103003e-05, "loss": 0.0143, "step": 13230 }, { "epoch": 1.434410234171726, "grad_norm": 0.8203534483909607, "learning_rate": 5.2375598433193095e-05, "loss": 0.015, "step": 13231 }, { "epoch": 1.4345186470078057, "grad_norm": 0.25073838233947754, "learning_rate": 5.237197156535616e-05, "loss": 0.0192, "step": 13232 }, { "epoch": 1.4346270598438855, "grad_norm": 0.12257546931505203, "learning_rate": 5.236834469751922e-05, "loss": 0.0131, "step": 13233 }, { "epoch": 1.4347354726799653, "grad_norm": 0.3344603478908539, "learning_rate": 5.236471782968229e-05, "loss": 0.0212, "step": 13234 }, { "epoch": 1.434843885516045, "grad_norm": 0.1643076092004776, "learning_rate": 5.236109096184535e-05, "loss": 0.0082, "step": 13235 }, { "epoch": 1.4349522983521248, "grad_norm": 0.21933764219284058, "learning_rate": 5.2357464094008415e-05, "loss": 0.0062, "step": 13236 }, { "epoch": 1.4350607111882048, "grad_norm": 0.4465867280960083, "learning_rate": 5.235383722617147e-05, "loss": 0.0132, "step": 13237 }, { "epoch": 1.4351691240242845, "grad_norm": 0.2693493068218231, "learning_rate": 5.235021035833454e-05, "loss": 0.0159, "step": 13238 }, { "epoch": 1.4352775368603643, "grad_norm": 0.507837176322937, "learning_rate": 5.2346583490497614e-05, "loss": 0.022, "step": 13239 }, { "epoch": 1.435385949696444, "grad_norm": 0.16203661262989044, "learning_rate": 5.234295662266068e-05, "loss": 0.0065, "step": 13240 }, { "epoch": 1.4354943625325238, "grad_norm": 0.35611245036125183, "learning_rate": 5.233932975482374e-05, "loss": 0.0325, "step": 13241 }, { "epoch": 1.4356027753686036, "grad_norm": 0.28373217582702637, "learning_rate": 5.233570288698681e-05, "loss": 0.0236, "step": 13242 }, { "epoch": 1.4357111882046834, "grad_norm": 0.7330198287963867, "learning_rate": 5.2332076019149864e-05, "loss": 0.0629, "step": 13243 }, { "epoch": 1.4358196010407633, "grad_norm": 0.3698522448539734, "learning_rate": 5.232844915131293e-05, "loss": 0.0499, "step": 13244 }, { "epoch": 1.435928013876843, "grad_norm": 0.685706377029419, "learning_rate": 5.232482228347599e-05, "loss": 0.0316, "step": 13245 }, { "epoch": 1.4360364267129229, "grad_norm": 0.23503601551055908, "learning_rate": 5.2321195415639056e-05, "loss": 0.0134, "step": 13246 }, { "epoch": 1.4361448395490026, "grad_norm": 0.3870737850666046, "learning_rate": 5.231756854780212e-05, "loss": 0.0404, "step": 13247 }, { "epoch": 1.4362532523850824, "grad_norm": 0.2968829572200775, "learning_rate": 5.2313941679965185e-05, "loss": 0.0263, "step": 13248 }, { "epoch": 1.4363616652211622, "grad_norm": 0.13261112570762634, "learning_rate": 5.231031481212825e-05, "loss": 0.0094, "step": 13249 }, { "epoch": 1.436470078057242, "grad_norm": 0.07426892966032028, "learning_rate": 5.230668794429131e-05, "loss": 0.0066, "step": 13250 }, { "epoch": 1.436578490893322, "grad_norm": 0.45997798442840576, "learning_rate": 5.230306107645438e-05, "loss": 0.0204, "step": 13251 }, { "epoch": 1.4366869037294014, "grad_norm": 0.15967939794063568, "learning_rate": 5.2299434208617434e-05, "loss": 0.007, "step": 13252 }, { "epoch": 1.4367953165654814, "grad_norm": 0.15155450999736786, "learning_rate": 5.22958073407805e-05, "loss": 0.0067, "step": 13253 }, { "epoch": 1.4369037294015612, "grad_norm": 0.37546029686927795, "learning_rate": 5.229218047294356e-05, "loss": 0.0163, "step": 13254 }, { "epoch": 1.437012142237641, "grad_norm": 0.5747086405754089, "learning_rate": 5.228855360510663e-05, "loss": 0.0535, "step": 13255 }, { "epoch": 1.4371205550737207, "grad_norm": 1.8780760765075684, "learning_rate": 5.2284926737269704e-05, "loss": 0.0306, "step": 13256 }, { "epoch": 1.4372289679098005, "grad_norm": 0.3345853090286255, "learning_rate": 5.228129986943276e-05, "loss": 0.031, "step": 13257 }, { "epoch": 1.4373373807458802, "grad_norm": 0.1537674069404602, "learning_rate": 5.2277673001595826e-05, "loss": 0.0099, "step": 13258 }, { "epoch": 1.43744579358196, "grad_norm": 0.19367048144340515, "learning_rate": 5.227404613375889e-05, "loss": 0.027, "step": 13259 }, { "epoch": 1.43755420641804, "grad_norm": 0.7319023013114929, "learning_rate": 5.2270419265921954e-05, "loss": 0.0267, "step": 13260 }, { "epoch": 1.4376626192541198, "grad_norm": 0.1793030947446823, "learning_rate": 5.226679239808502e-05, "loss": 0.0153, "step": 13261 }, { "epoch": 1.4377710320901995, "grad_norm": 0.6055222153663635, "learning_rate": 5.226316553024808e-05, "loss": 0.0602, "step": 13262 }, { "epoch": 1.4378794449262793, "grad_norm": 0.8978057503700256, "learning_rate": 5.2259538662411146e-05, "loss": 0.0431, "step": 13263 }, { "epoch": 1.437987857762359, "grad_norm": 0.10627208650112152, "learning_rate": 5.225591179457421e-05, "loss": 0.0047, "step": 13264 }, { "epoch": 1.4380962705984388, "grad_norm": 0.2562965452671051, "learning_rate": 5.2252284926737274e-05, "loss": 0.0211, "step": 13265 }, { "epoch": 1.4382046834345186, "grad_norm": 0.271900475025177, "learning_rate": 5.224865805890033e-05, "loss": 0.0223, "step": 13266 }, { "epoch": 1.4383130962705986, "grad_norm": 0.39249926805496216, "learning_rate": 5.2245031191063396e-05, "loss": 0.0226, "step": 13267 }, { "epoch": 1.438421509106678, "grad_norm": 0.36349183320999146, "learning_rate": 5.224140432322646e-05, "loss": 0.0494, "step": 13268 }, { "epoch": 1.438529921942758, "grad_norm": 0.2236647754907608, "learning_rate": 5.2237777455389524e-05, "loss": 0.0158, "step": 13269 }, { "epoch": 1.4386383347788378, "grad_norm": 0.37057414650917053, "learning_rate": 5.223415058755259e-05, "loss": 0.0212, "step": 13270 }, { "epoch": 1.4387467476149176, "grad_norm": 0.47611552476882935, "learning_rate": 5.223052371971565e-05, "loss": 0.019, "step": 13271 }, { "epoch": 1.4388551604509974, "grad_norm": 0.28560319542884827, "learning_rate": 5.2226896851878716e-05, "loss": 0.0176, "step": 13272 }, { "epoch": 1.4389635732870771, "grad_norm": 0.17371457815170288, "learning_rate": 5.222326998404179e-05, "loss": 0.0141, "step": 13273 }, { "epoch": 1.439071986123157, "grad_norm": 0.13462668657302856, "learning_rate": 5.221964311620485e-05, "loss": 0.0126, "step": 13274 }, { "epoch": 1.4391803989592367, "grad_norm": 0.49483275413513184, "learning_rate": 5.2216016248367916e-05, "loss": 0.0437, "step": 13275 }, { "epoch": 1.4392888117953166, "grad_norm": 0.48000597953796387, "learning_rate": 5.221238938053098e-05, "loss": 0.0321, "step": 13276 }, { "epoch": 1.4393972246313964, "grad_norm": 0.42118051648139954, "learning_rate": 5.2208762512694044e-05, "loss": 0.0417, "step": 13277 }, { "epoch": 1.4395056374674762, "grad_norm": 0.22947420179843903, "learning_rate": 5.220513564485711e-05, "loss": 0.0164, "step": 13278 }, { "epoch": 1.439614050303556, "grad_norm": 0.09775005280971527, "learning_rate": 5.220150877702017e-05, "loss": 0.008, "step": 13279 }, { "epoch": 1.4397224631396357, "grad_norm": 0.6324146389961243, "learning_rate": 5.2197881909183236e-05, "loss": 0.0377, "step": 13280 }, { "epoch": 1.4398308759757155, "grad_norm": 0.1681380271911621, "learning_rate": 5.219425504134629e-05, "loss": 0.0089, "step": 13281 }, { "epoch": 1.4399392888117952, "grad_norm": 0.23606786131858826, "learning_rate": 5.219062817350936e-05, "loss": 0.0217, "step": 13282 }, { "epoch": 1.4400477016478752, "grad_norm": 0.4419579803943634, "learning_rate": 5.218700130567242e-05, "loss": 0.0279, "step": 13283 }, { "epoch": 1.440156114483955, "grad_norm": 0.18064656853675842, "learning_rate": 5.2183374437835486e-05, "loss": 0.0106, "step": 13284 }, { "epoch": 1.4402645273200347, "grad_norm": 0.2559657096862793, "learning_rate": 5.217974756999855e-05, "loss": 0.0221, "step": 13285 }, { "epoch": 1.4403729401561145, "grad_norm": 0.23526042699813843, "learning_rate": 5.2176120702161614e-05, "loss": 0.0409, "step": 13286 }, { "epoch": 1.4404813529921943, "grad_norm": 0.26286497712135315, "learning_rate": 5.217249383432468e-05, "loss": 0.0113, "step": 13287 }, { "epoch": 1.440589765828274, "grad_norm": 0.37215593457221985, "learning_rate": 5.216886696648774e-05, "loss": 0.0223, "step": 13288 }, { "epoch": 1.4406981786643538, "grad_norm": 0.5615537166595459, "learning_rate": 5.21652400986508e-05, "loss": 0.0817, "step": 13289 }, { "epoch": 1.4408065915004338, "grad_norm": 0.41828399896621704, "learning_rate": 5.216161323081388e-05, "loss": 0.0346, "step": 13290 }, { "epoch": 1.4409150043365133, "grad_norm": 0.32956451177597046, "learning_rate": 5.215798636297694e-05, "loss": 0.0175, "step": 13291 }, { "epoch": 1.4410234171725933, "grad_norm": 0.5109143257141113, "learning_rate": 5.2154359495140005e-05, "loss": 0.0645, "step": 13292 }, { "epoch": 1.441131830008673, "grad_norm": 0.46508994698524475, "learning_rate": 5.215073262730307e-05, "loss": 0.0179, "step": 13293 }, { "epoch": 1.4412402428447528, "grad_norm": 0.0845542773604393, "learning_rate": 5.2147105759466134e-05, "loss": 0.0096, "step": 13294 }, { "epoch": 1.4413486556808326, "grad_norm": 0.3516473174095154, "learning_rate": 5.214347889162919e-05, "loss": 0.0224, "step": 13295 }, { "epoch": 1.4414570685169124, "grad_norm": 0.5925925374031067, "learning_rate": 5.2139852023792255e-05, "loss": 0.0669, "step": 13296 }, { "epoch": 1.4415654813529921, "grad_norm": 0.3079526126384735, "learning_rate": 5.213622515595532e-05, "loss": 0.0915, "step": 13297 }, { "epoch": 1.4416738941890719, "grad_norm": 0.3063575029373169, "learning_rate": 5.213259828811838e-05, "loss": 0.0237, "step": 13298 }, { "epoch": 1.4417823070251519, "grad_norm": 0.09880061447620392, "learning_rate": 5.212897142028145e-05, "loss": 0.0078, "step": 13299 }, { "epoch": 1.4418907198612316, "grad_norm": 0.38597139716148376, "learning_rate": 5.212534455244451e-05, "loss": 0.0226, "step": 13300 }, { "epoch": 1.4419991326973114, "grad_norm": 0.1442762166261673, "learning_rate": 5.2121717684607575e-05, "loss": 0.0077, "step": 13301 }, { "epoch": 1.4421075455333912, "grad_norm": 0.14387118816375732, "learning_rate": 5.211809081677064e-05, "loss": 0.0046, "step": 13302 }, { "epoch": 1.442215958369471, "grad_norm": 0.34561511874198914, "learning_rate": 5.2114463948933704e-05, "loss": 0.0272, "step": 13303 }, { "epoch": 1.4423243712055507, "grad_norm": 0.24864499270915985, "learning_rate": 5.211083708109676e-05, "loss": 0.0178, "step": 13304 }, { "epoch": 1.4424327840416304, "grad_norm": 0.3621777296066284, "learning_rate": 5.2107210213259825e-05, "loss": 0.0219, "step": 13305 }, { "epoch": 1.4425411968777104, "grad_norm": 0.6485814452171326, "learning_rate": 5.210358334542289e-05, "loss": 0.0507, "step": 13306 }, { "epoch": 1.4426496097137902, "grad_norm": 0.14314407110214233, "learning_rate": 5.209995647758595e-05, "loss": 0.0172, "step": 13307 }, { "epoch": 1.44275802254987, "grad_norm": 0.1814870834350586, "learning_rate": 5.209632960974903e-05, "loss": 0.0091, "step": 13308 }, { "epoch": 1.4428664353859497, "grad_norm": 0.4492627680301666, "learning_rate": 5.209270274191209e-05, "loss": 0.038, "step": 13309 }, { "epoch": 1.4429748482220295, "grad_norm": 0.41198673844337463, "learning_rate": 5.208907587407515e-05, "loss": 0.037, "step": 13310 }, { "epoch": 1.4430832610581092, "grad_norm": 0.028393201529979706, "learning_rate": 5.2085449006238217e-05, "loss": 0.0015, "step": 13311 }, { "epoch": 1.443191673894189, "grad_norm": 0.12747722864151, "learning_rate": 5.208182213840128e-05, "loss": 0.0065, "step": 13312 }, { "epoch": 1.443300086730269, "grad_norm": 0.07401273399591446, "learning_rate": 5.2078195270564345e-05, "loss": 0.0054, "step": 13313 }, { "epoch": 1.4434084995663485, "grad_norm": 0.25473752617836, "learning_rate": 5.207456840272741e-05, "loss": 0.0103, "step": 13314 }, { "epoch": 1.4435169124024285, "grad_norm": 0.23564375936985016, "learning_rate": 5.207094153489047e-05, "loss": 0.0074, "step": 13315 }, { "epoch": 1.4436253252385083, "grad_norm": 0.1361839473247528, "learning_rate": 5.206731466705354e-05, "loss": 0.0129, "step": 13316 }, { "epoch": 1.443733738074588, "grad_norm": 0.22717060148715973, "learning_rate": 5.20636877992166e-05, "loss": 0.0177, "step": 13317 }, { "epoch": 1.4438421509106678, "grad_norm": 0.4374808967113495, "learning_rate": 5.206006093137966e-05, "loss": 0.0117, "step": 13318 }, { "epoch": 1.4439505637467476, "grad_norm": 0.12568174302577972, "learning_rate": 5.205643406354272e-05, "loss": 0.0049, "step": 13319 }, { "epoch": 1.4440589765828273, "grad_norm": 0.19195958971977234, "learning_rate": 5.205280719570579e-05, "loss": 0.009, "step": 13320 }, { "epoch": 1.444167389418907, "grad_norm": 0.48694556951522827, "learning_rate": 5.204918032786885e-05, "loss": 0.0534, "step": 13321 }, { "epoch": 1.444275802254987, "grad_norm": 0.09992767125368118, "learning_rate": 5.2045553460031915e-05, "loss": 0.0057, "step": 13322 }, { "epoch": 1.4443842150910668, "grad_norm": 0.2590835690498352, "learning_rate": 5.204192659219498e-05, "loss": 0.0249, "step": 13323 }, { "epoch": 1.4444926279271466, "grad_norm": 0.7577505707740784, "learning_rate": 5.203829972435804e-05, "loss": 0.0498, "step": 13324 }, { "epoch": 1.4446010407632264, "grad_norm": 0.2857058048248291, "learning_rate": 5.2034672856521114e-05, "loss": 0.012, "step": 13325 }, { "epoch": 1.4447094535993061, "grad_norm": 0.5387669205665588, "learning_rate": 5.203104598868418e-05, "loss": 0.0452, "step": 13326 }, { "epoch": 1.444817866435386, "grad_norm": 0.5178661346435547, "learning_rate": 5.202741912084724e-05, "loss": 0.0166, "step": 13327 }, { "epoch": 1.4449262792714657, "grad_norm": 0.26559245586395264, "learning_rate": 5.2023792253010306e-05, "loss": 0.0309, "step": 13328 }, { "epoch": 1.4450346921075456, "grad_norm": 0.4706460237503052, "learning_rate": 5.202016538517337e-05, "loss": 0.0469, "step": 13329 }, { "epoch": 1.4451431049436252, "grad_norm": 0.19334657490253448, "learning_rate": 5.2016538517336435e-05, "loss": 0.0073, "step": 13330 }, { "epoch": 1.4452515177797052, "grad_norm": 0.20865769684314728, "learning_rate": 5.20129116494995e-05, "loss": 0.0071, "step": 13331 }, { "epoch": 1.445359930615785, "grad_norm": 0.11002577096223831, "learning_rate": 5.200928478166256e-05, "loss": 0.0063, "step": 13332 }, { "epoch": 1.4454683434518647, "grad_norm": 0.41749125719070435, "learning_rate": 5.200565791382562e-05, "loss": 0.0081, "step": 13333 }, { "epoch": 1.4455767562879445, "grad_norm": 0.19876106083393097, "learning_rate": 5.2002031045988684e-05, "loss": 0.0092, "step": 13334 }, { "epoch": 1.4456851691240242, "grad_norm": 0.09390123933553696, "learning_rate": 5.199840417815175e-05, "loss": 0.0064, "step": 13335 }, { "epoch": 1.4457935819601042, "grad_norm": 0.6858454346656799, "learning_rate": 5.199477731031481e-05, "loss": 0.0111, "step": 13336 }, { "epoch": 1.4459019947961838, "grad_norm": 0.3740513324737549, "learning_rate": 5.1991150442477876e-05, "loss": 0.0248, "step": 13337 }, { "epoch": 1.4460104076322637, "grad_norm": 0.02216467447578907, "learning_rate": 5.198752357464094e-05, "loss": 0.0012, "step": 13338 }, { "epoch": 1.4461188204683435, "grad_norm": 0.3341980576515198, "learning_rate": 5.1983896706804005e-05, "loss": 0.0119, "step": 13339 }, { "epoch": 1.4462272333044233, "grad_norm": 0.15325741469860077, "learning_rate": 5.198026983896707e-05, "loss": 0.0078, "step": 13340 }, { "epoch": 1.446335646140503, "grad_norm": 0.6096677780151367, "learning_rate": 5.1976642971130126e-05, "loss": 0.0432, "step": 13341 }, { "epoch": 1.4464440589765828, "grad_norm": 0.08186331391334534, "learning_rate": 5.1973016103293204e-05, "loss": 0.0028, "step": 13342 }, { "epoch": 1.4465524718126626, "grad_norm": 0.4706272482872009, "learning_rate": 5.196938923545627e-05, "loss": 0.0327, "step": 13343 }, { "epoch": 1.4466608846487423, "grad_norm": 0.4091936945915222, "learning_rate": 5.196576236761933e-05, "loss": 0.0313, "step": 13344 }, { "epoch": 1.4467692974848223, "grad_norm": 0.40210673213005066, "learning_rate": 5.1962135499782396e-05, "loss": 0.0106, "step": 13345 }, { "epoch": 1.446877710320902, "grad_norm": 0.34608376026153564, "learning_rate": 5.195850863194546e-05, "loss": 0.0269, "step": 13346 }, { "epoch": 1.4469861231569818, "grad_norm": 0.24778437614440918, "learning_rate": 5.195488176410852e-05, "loss": 0.0411, "step": 13347 }, { "epoch": 1.4470945359930616, "grad_norm": 0.5604386329650879, "learning_rate": 5.195125489627158e-05, "loss": 0.0512, "step": 13348 }, { "epoch": 1.4472029488291414, "grad_norm": 0.7243866920471191, "learning_rate": 5.1947628028434646e-05, "loss": 0.0277, "step": 13349 }, { "epoch": 1.4473113616652211, "grad_norm": 0.37749701738357544, "learning_rate": 5.194400116059771e-05, "loss": 0.021, "step": 13350 }, { "epoch": 1.4474197745013009, "grad_norm": 0.2914762496948242, "learning_rate": 5.1940374292760774e-05, "loss": 0.0218, "step": 13351 }, { "epoch": 1.4475281873373809, "grad_norm": 0.33787471055984497, "learning_rate": 5.193674742492384e-05, "loss": 0.0363, "step": 13352 }, { "epoch": 1.4476366001734604, "grad_norm": 0.32678255438804626, "learning_rate": 5.19331205570869e-05, "loss": 0.0276, "step": 13353 }, { "epoch": 1.4477450130095404, "grad_norm": 0.5403574705123901, "learning_rate": 5.1929493689249966e-05, "loss": 0.0389, "step": 13354 }, { "epoch": 1.4478534258456202, "grad_norm": 0.0930873304605484, "learning_rate": 5.192586682141303e-05, "loss": 0.0027, "step": 13355 }, { "epoch": 1.4479618386817, "grad_norm": 0.9174445867538452, "learning_rate": 5.192223995357609e-05, "loss": 0.0216, "step": 13356 }, { "epoch": 1.4480702515177797, "grad_norm": 0.05025334656238556, "learning_rate": 5.191861308573915e-05, "loss": 0.0038, "step": 13357 }, { "epoch": 1.4481786643538594, "grad_norm": 0.08101288229227066, "learning_rate": 5.1914986217902216e-05, "loss": 0.0052, "step": 13358 }, { "epoch": 1.4482870771899392, "grad_norm": 0.12394828349351883, "learning_rate": 5.191135935006528e-05, "loss": 0.0117, "step": 13359 }, { "epoch": 1.448395490026019, "grad_norm": 0.25407734513282776, "learning_rate": 5.190773248222836e-05, "loss": 0.0216, "step": 13360 }, { "epoch": 1.448503902862099, "grad_norm": 0.4655340015888214, "learning_rate": 5.190410561439142e-05, "loss": 0.0586, "step": 13361 }, { "epoch": 1.4486123156981787, "grad_norm": 0.25680211186408997, "learning_rate": 5.190047874655448e-05, "loss": 0.0063, "step": 13362 }, { "epoch": 1.4487207285342585, "grad_norm": 0.22461822628974915, "learning_rate": 5.189685187871754e-05, "loss": 0.0132, "step": 13363 }, { "epoch": 1.4488291413703382, "grad_norm": 0.410895973443985, "learning_rate": 5.189322501088061e-05, "loss": 0.0602, "step": 13364 }, { "epoch": 1.448937554206418, "grad_norm": 0.10101357102394104, "learning_rate": 5.188959814304367e-05, "loss": 0.0053, "step": 13365 }, { "epoch": 1.4490459670424978, "grad_norm": 0.5879252552986145, "learning_rate": 5.1885971275206736e-05, "loss": 0.0146, "step": 13366 }, { "epoch": 1.4491543798785775, "grad_norm": 0.3818717300891876, "learning_rate": 5.18823444073698e-05, "loss": 0.0172, "step": 13367 }, { "epoch": 1.4492627927146575, "grad_norm": 0.25502103567123413, "learning_rate": 5.1878717539532864e-05, "loss": 0.0079, "step": 13368 }, { "epoch": 1.4493712055507373, "grad_norm": 0.5035876631736755, "learning_rate": 5.187509067169593e-05, "loss": 0.0412, "step": 13369 }, { "epoch": 1.449479618386817, "grad_norm": 0.7999685406684875, "learning_rate": 5.1871463803858985e-05, "loss": 0.0061, "step": 13370 }, { "epoch": 1.4495880312228968, "grad_norm": 0.33186933398246765, "learning_rate": 5.186783693602205e-05, "loss": 0.023, "step": 13371 }, { "epoch": 1.4496964440589766, "grad_norm": 0.1263798624277115, "learning_rate": 5.186421006818511e-05, "loss": 0.0052, "step": 13372 }, { "epoch": 1.4498048568950563, "grad_norm": 0.5275840163230896, "learning_rate": 5.186058320034818e-05, "loss": 0.0208, "step": 13373 }, { "epoch": 1.449913269731136, "grad_norm": 0.2832352817058563, "learning_rate": 5.185695633251124e-05, "loss": 0.0333, "step": 13374 }, { "epoch": 1.450021682567216, "grad_norm": 0.2341468185186386, "learning_rate": 5.1853329464674306e-05, "loss": 0.0191, "step": 13375 }, { "epoch": 1.4501300954032956, "grad_norm": 0.06845010817050934, "learning_rate": 5.184970259683737e-05, "loss": 0.0045, "step": 13376 }, { "epoch": 1.4502385082393756, "grad_norm": 0.6379640698432922, "learning_rate": 5.184607572900044e-05, "loss": 0.0325, "step": 13377 }, { "epoch": 1.4503469210754554, "grad_norm": 0.17689378559589386, "learning_rate": 5.1842448861163505e-05, "loss": 0.0229, "step": 13378 }, { "epoch": 1.4504553339115351, "grad_norm": 0.38731008768081665, "learning_rate": 5.183882199332657e-05, "loss": 0.0638, "step": 13379 }, { "epoch": 1.450563746747615, "grad_norm": 0.27319270372390747, "learning_rate": 5.183519512548963e-05, "loss": 0.0278, "step": 13380 }, { "epoch": 1.4506721595836947, "grad_norm": 0.3426084816455841, "learning_rate": 5.18315682576527e-05, "loss": 0.0102, "step": 13381 }, { "epoch": 1.4507805724197744, "grad_norm": 0.23482967913150787, "learning_rate": 5.182794138981576e-05, "loss": 0.0117, "step": 13382 }, { "epoch": 1.4508889852558542, "grad_norm": 0.2836439311504364, "learning_rate": 5.1824314521978825e-05, "loss": 0.0228, "step": 13383 }, { "epoch": 1.4509973980919342, "grad_norm": 0.1801556795835495, "learning_rate": 5.182068765414189e-05, "loss": 0.0128, "step": 13384 }, { "epoch": 1.451105810928014, "grad_norm": 0.8215906023979187, "learning_rate": 5.181706078630495e-05, "loss": 0.0531, "step": 13385 }, { "epoch": 1.4512142237640937, "grad_norm": 0.44446173310279846, "learning_rate": 5.181343391846801e-05, "loss": 0.0368, "step": 13386 }, { "epoch": 1.4513226366001735, "grad_norm": 0.47297829389572144, "learning_rate": 5.1809807050631075e-05, "loss": 0.02, "step": 13387 }, { "epoch": 1.4514310494362532, "grad_norm": 0.39171478152275085, "learning_rate": 5.180618018279414e-05, "loss": 0.0207, "step": 13388 }, { "epoch": 1.451539462272333, "grad_norm": 0.6443740129470825, "learning_rate": 5.18025533149572e-05, "loss": 0.0486, "step": 13389 }, { "epoch": 1.4516478751084128, "grad_norm": 0.17276345193386078, "learning_rate": 5.179892644712027e-05, "loss": 0.0103, "step": 13390 }, { "epoch": 1.4517562879444927, "grad_norm": 0.3903874158859253, "learning_rate": 5.179529957928333e-05, "loss": 0.0216, "step": 13391 }, { "epoch": 1.4518647007805723, "grad_norm": 0.400664359331131, "learning_rate": 5.1791672711446395e-05, "loss": 0.0442, "step": 13392 }, { "epoch": 1.4519731136166523, "grad_norm": 0.3349393606185913, "learning_rate": 5.178804584360946e-05, "loss": 0.0096, "step": 13393 }, { "epoch": 1.452081526452732, "grad_norm": 0.344407320022583, "learning_rate": 5.178441897577253e-05, "loss": 0.0393, "step": 13394 }, { "epoch": 1.4521899392888118, "grad_norm": 0.3301835358142853, "learning_rate": 5.1780792107935595e-05, "loss": 0.0245, "step": 13395 }, { "epoch": 1.4522983521248916, "grad_norm": 0.1273650974035263, "learning_rate": 5.177716524009866e-05, "loss": 0.0078, "step": 13396 }, { "epoch": 1.4524067649609713, "grad_norm": 0.5906363129615784, "learning_rate": 5.177353837226172e-05, "loss": 0.0376, "step": 13397 }, { "epoch": 1.4525151777970513, "grad_norm": 0.15434038639068604, "learning_rate": 5.176991150442479e-05, "loss": 0.0098, "step": 13398 }, { "epoch": 1.4526235906331308, "grad_norm": 0.4307555854320526, "learning_rate": 5.1766284636587844e-05, "loss": 0.029, "step": 13399 }, { "epoch": 1.4527320034692108, "grad_norm": 0.06515669077634811, "learning_rate": 5.176265776875091e-05, "loss": 0.0039, "step": 13400 }, { "epoch": 1.4528404163052906, "grad_norm": 0.3085022568702698, "learning_rate": 5.175903090091397e-05, "loss": 0.0216, "step": 13401 }, { "epoch": 1.4529488291413704, "grad_norm": 0.35837289690971375, "learning_rate": 5.1755404033077037e-05, "loss": 0.0185, "step": 13402 }, { "epoch": 1.4530572419774501, "grad_norm": 0.22114680707454681, "learning_rate": 5.17517771652401e-05, "loss": 0.0105, "step": 13403 }, { "epoch": 1.4531656548135299, "grad_norm": 0.09755078703165054, "learning_rate": 5.1748150297403165e-05, "loss": 0.0074, "step": 13404 }, { "epoch": 1.4532740676496096, "grad_norm": 0.6076834797859192, "learning_rate": 5.174452342956623e-05, "loss": 0.0707, "step": 13405 }, { "epoch": 1.4533824804856894, "grad_norm": 0.30443617701530457, "learning_rate": 5.174089656172929e-05, "loss": 0.0187, "step": 13406 }, { "epoch": 1.4534908933217694, "grad_norm": 0.2928500771522522, "learning_rate": 5.173726969389236e-05, "loss": 0.0816, "step": 13407 }, { "epoch": 1.4535993061578492, "grad_norm": 0.7798283696174622, "learning_rate": 5.1733642826055414e-05, "loss": 0.0952, "step": 13408 }, { "epoch": 1.453707718993929, "grad_norm": 0.16862168908119202, "learning_rate": 5.173001595821848e-05, "loss": 0.0391, "step": 13409 }, { "epoch": 1.4538161318300087, "grad_norm": 0.13337935507297516, "learning_rate": 5.172638909038154e-05, "loss": 0.006, "step": 13410 }, { "epoch": 1.4539245446660884, "grad_norm": 0.07592865824699402, "learning_rate": 5.172276222254462e-05, "loss": 0.0038, "step": 13411 }, { "epoch": 1.4540329575021682, "grad_norm": 0.2221754491329193, "learning_rate": 5.1719135354707684e-05, "loss": 0.0524, "step": 13412 }, { "epoch": 1.454141370338248, "grad_norm": 0.3342292010784149, "learning_rate": 5.171550848687075e-05, "loss": 0.0167, "step": 13413 }, { "epoch": 1.454249783174328, "grad_norm": 0.1976521611213684, "learning_rate": 5.1711881619033806e-05, "loss": 0.0068, "step": 13414 }, { "epoch": 1.4543581960104075, "grad_norm": 0.32844945788383484, "learning_rate": 5.170825475119687e-05, "loss": 0.0272, "step": 13415 }, { "epoch": 1.4544666088464875, "grad_norm": 0.20575930178165436, "learning_rate": 5.1704627883359934e-05, "loss": 0.0113, "step": 13416 }, { "epoch": 1.4545750216825672, "grad_norm": 0.11034994572401047, "learning_rate": 5.1701001015523e-05, "loss": 0.006, "step": 13417 }, { "epoch": 1.454683434518647, "grad_norm": 0.3559282720088959, "learning_rate": 5.169737414768606e-05, "loss": 0.0596, "step": 13418 }, { "epoch": 1.4547918473547268, "grad_norm": 0.5306179523468018, "learning_rate": 5.1693747279849126e-05, "loss": 0.0305, "step": 13419 }, { "epoch": 1.4549002601908065, "grad_norm": 0.37724870443344116, "learning_rate": 5.169012041201219e-05, "loss": 0.0341, "step": 13420 }, { "epoch": 1.4550086730268863, "grad_norm": 0.3674352765083313, "learning_rate": 5.1686493544175255e-05, "loss": 0.0327, "step": 13421 }, { "epoch": 1.455117085862966, "grad_norm": 0.14874623715877533, "learning_rate": 5.168286667633831e-05, "loss": 0.0157, "step": 13422 }, { "epoch": 1.455225498699046, "grad_norm": 0.2589870095252991, "learning_rate": 5.1679239808501376e-05, "loss": 0.0188, "step": 13423 }, { "epoch": 1.4553339115351258, "grad_norm": 0.09461203217506409, "learning_rate": 5.167561294066444e-05, "loss": 0.0052, "step": 13424 }, { "epoch": 1.4554423243712056, "grad_norm": 0.40192127227783203, "learning_rate": 5.1671986072827504e-05, "loss": 0.0611, "step": 13425 }, { "epoch": 1.4555507372072853, "grad_norm": 0.15168291330337524, "learning_rate": 5.166835920499057e-05, "loss": 0.014, "step": 13426 }, { "epoch": 1.455659150043365, "grad_norm": 0.3307635486125946, "learning_rate": 5.166473233715363e-05, "loss": 0.0156, "step": 13427 }, { "epoch": 1.4557675628794449, "grad_norm": 0.8485095500946045, "learning_rate": 5.1661105469316696e-05, "loss": 0.0761, "step": 13428 }, { "epoch": 1.4558759757155246, "grad_norm": 0.15636496245861053, "learning_rate": 5.165747860147977e-05, "loss": 0.0116, "step": 13429 }, { "epoch": 1.4559843885516046, "grad_norm": 0.4266207814216614, "learning_rate": 5.165385173364283e-05, "loss": 0.029, "step": 13430 }, { "epoch": 1.4560928013876844, "grad_norm": 0.17835403978824615, "learning_rate": 5.1650224865805896e-05, "loss": 0.0106, "step": 13431 }, { "epoch": 1.4562012142237641, "grad_norm": 0.24899695813655853, "learning_rate": 5.164659799796896e-05, "loss": 0.0168, "step": 13432 }, { "epoch": 1.456309627059844, "grad_norm": 0.47389715909957886, "learning_rate": 5.1642971130132024e-05, "loss": 0.0291, "step": 13433 }, { "epoch": 1.4564180398959237, "grad_norm": 0.4912714958190918, "learning_rate": 5.163934426229509e-05, "loss": 0.0589, "step": 13434 }, { "epoch": 1.4565264527320034, "grad_norm": 0.4026258885860443, "learning_rate": 5.163571739445815e-05, "loss": 0.0154, "step": 13435 }, { "epoch": 1.4566348655680832, "grad_norm": 0.24115629494190216, "learning_rate": 5.1632090526621216e-05, "loss": 0.0271, "step": 13436 }, { "epoch": 1.4567432784041632, "grad_norm": 0.3312755525112152, "learning_rate": 5.1628463658784273e-05, "loss": 0.0341, "step": 13437 }, { "epoch": 1.4568516912402427, "grad_norm": 0.24749089777469635, "learning_rate": 5.162483679094734e-05, "loss": 0.0167, "step": 13438 }, { "epoch": 1.4569601040763227, "grad_norm": 0.2733791172504425, "learning_rate": 5.16212099231104e-05, "loss": 0.024, "step": 13439 }, { "epoch": 1.4570685169124025, "grad_norm": 0.4575894773006439, "learning_rate": 5.1617583055273466e-05, "loss": 0.0561, "step": 13440 }, { "epoch": 1.4571769297484822, "grad_norm": 0.1984095275402069, "learning_rate": 5.161395618743653e-05, "loss": 0.0172, "step": 13441 }, { "epoch": 1.457285342584562, "grad_norm": 0.2026013880968094, "learning_rate": 5.1610329319599594e-05, "loss": 0.0165, "step": 13442 }, { "epoch": 1.4573937554206418, "grad_norm": 0.5337182879447937, "learning_rate": 5.160670245176266e-05, "loss": 0.0361, "step": 13443 }, { "epoch": 1.4575021682567215, "grad_norm": 0.10314086079597473, "learning_rate": 5.160307558392572e-05, "loss": 0.0048, "step": 13444 }, { "epoch": 1.4576105810928013, "grad_norm": 0.39017245173454285, "learning_rate": 5.1599448716088786e-05, "loss": 0.0408, "step": 13445 }, { "epoch": 1.4577189939288813, "grad_norm": 0.5494146943092346, "learning_rate": 5.159582184825186e-05, "loss": 0.0645, "step": 13446 }, { "epoch": 1.457827406764961, "grad_norm": 0.20036694407463074, "learning_rate": 5.159219498041492e-05, "loss": 0.0169, "step": 13447 }, { "epoch": 1.4579358196010408, "grad_norm": 0.20924213528633118, "learning_rate": 5.1588568112577985e-05, "loss": 0.0236, "step": 13448 }, { "epoch": 1.4580442324371206, "grad_norm": 0.5287615656852722, "learning_rate": 5.158494124474105e-05, "loss": 0.033, "step": 13449 }, { "epoch": 1.4581526452732003, "grad_norm": 0.24318194389343262, "learning_rate": 5.1581314376904114e-05, "loss": 0.019, "step": 13450 }, { "epoch": 1.45826105810928, "grad_norm": 0.6321537494659424, "learning_rate": 5.157768750906717e-05, "loss": 0.033, "step": 13451 }, { "epoch": 1.4583694709453598, "grad_norm": 0.23517312109470367, "learning_rate": 5.1574060641230235e-05, "loss": 0.0178, "step": 13452 }, { "epoch": 1.4584778837814398, "grad_norm": 0.20121562480926514, "learning_rate": 5.15704337733933e-05, "loss": 0.0133, "step": 13453 }, { "epoch": 1.4585862966175196, "grad_norm": 0.02830016240477562, "learning_rate": 5.156680690555636e-05, "loss": 0.0018, "step": 13454 }, { "epoch": 1.4586947094535994, "grad_norm": 0.4309585392475128, "learning_rate": 5.156318003771943e-05, "loss": 0.0362, "step": 13455 }, { "epoch": 1.4588031222896791, "grad_norm": 0.1839328557252884, "learning_rate": 5.155955316988249e-05, "loss": 0.0146, "step": 13456 }, { "epoch": 1.4589115351257589, "grad_norm": 0.2437199205160141, "learning_rate": 5.1555926302045556e-05, "loss": 0.0111, "step": 13457 }, { "epoch": 1.4590199479618386, "grad_norm": 0.5280070900917053, "learning_rate": 5.155229943420862e-05, "loss": 0.0715, "step": 13458 }, { "epoch": 1.4591283607979184, "grad_norm": 0.2526651620864868, "learning_rate": 5.1548672566371684e-05, "loss": 0.0144, "step": 13459 }, { "epoch": 1.4592367736339984, "grad_norm": 0.0981740653514862, "learning_rate": 5.154504569853474e-05, "loss": 0.0113, "step": 13460 }, { "epoch": 1.459345186470078, "grad_norm": 0.10403454303741455, "learning_rate": 5.1541418830697805e-05, "loss": 0.0029, "step": 13461 }, { "epoch": 1.459453599306158, "grad_norm": 0.41808032989501953, "learning_rate": 5.153779196286087e-05, "loss": 0.0295, "step": 13462 }, { "epoch": 1.4595620121422377, "grad_norm": 0.29208850860595703, "learning_rate": 5.153416509502395e-05, "loss": 0.0208, "step": 13463 }, { "epoch": 1.4596704249783174, "grad_norm": 0.5567901134490967, "learning_rate": 5.153053822718701e-05, "loss": 0.0247, "step": 13464 }, { "epoch": 1.4597788378143972, "grad_norm": 0.26942315697669983, "learning_rate": 5.1526911359350075e-05, "loss": 0.022, "step": 13465 }, { "epoch": 1.459887250650477, "grad_norm": 0.39358922839164734, "learning_rate": 5.152328449151313e-05, "loss": 0.053, "step": 13466 }, { "epoch": 1.4599956634865567, "grad_norm": 0.07470051944255829, "learning_rate": 5.1519657623676197e-05, "loss": 0.0081, "step": 13467 }, { "epoch": 1.4601040763226365, "grad_norm": 0.22235526144504547, "learning_rate": 5.151603075583926e-05, "loss": 0.0112, "step": 13468 }, { "epoch": 1.4602124891587165, "grad_norm": 0.4903932809829712, "learning_rate": 5.1512403888002325e-05, "loss": 0.0584, "step": 13469 }, { "epoch": 1.4603209019947962, "grad_norm": 0.2168349176645279, "learning_rate": 5.150877702016539e-05, "loss": 0.0333, "step": 13470 }, { "epoch": 1.460429314830876, "grad_norm": 0.5922860503196716, "learning_rate": 5.150515015232845e-05, "loss": 0.0184, "step": 13471 }, { "epoch": 1.4605377276669558, "grad_norm": 0.25475460290908813, "learning_rate": 5.150152328449152e-05, "loss": 0.0158, "step": 13472 }, { "epoch": 1.4606461405030355, "grad_norm": 0.34331512451171875, "learning_rate": 5.149789641665458e-05, "loss": 0.0261, "step": 13473 }, { "epoch": 1.4607545533391153, "grad_norm": 0.29439157247543335, "learning_rate": 5.149426954881764e-05, "loss": 0.0078, "step": 13474 }, { "epoch": 1.460862966175195, "grad_norm": 0.1643986999988556, "learning_rate": 5.14906426809807e-05, "loss": 0.0088, "step": 13475 }, { "epoch": 1.460971379011275, "grad_norm": 0.1200013980269432, "learning_rate": 5.148701581314377e-05, "loss": 0.0065, "step": 13476 }, { "epoch": 1.4610797918473546, "grad_norm": 0.1376330703496933, "learning_rate": 5.148338894530683e-05, "loss": 0.06, "step": 13477 }, { "epoch": 1.4611882046834346, "grad_norm": 0.09641515463590622, "learning_rate": 5.1479762077469895e-05, "loss": 0.0038, "step": 13478 }, { "epoch": 1.4612966175195143, "grad_norm": 0.7572593092918396, "learning_rate": 5.147613520963296e-05, "loss": 0.0436, "step": 13479 }, { "epoch": 1.461405030355594, "grad_norm": 0.3665655553340912, "learning_rate": 5.147250834179602e-05, "loss": 0.0162, "step": 13480 }, { "epoch": 1.4615134431916739, "grad_norm": 0.3972296416759491, "learning_rate": 5.1468881473959094e-05, "loss": 0.0213, "step": 13481 }, { "epoch": 1.4616218560277536, "grad_norm": 0.4054194390773773, "learning_rate": 5.146525460612216e-05, "loss": 0.0349, "step": 13482 }, { "epoch": 1.4617302688638336, "grad_norm": 0.33768072724342346, "learning_rate": 5.146162773828522e-05, "loss": 0.0293, "step": 13483 }, { "epoch": 1.4618386816999132, "grad_norm": 0.30805057287216187, "learning_rate": 5.1458000870448286e-05, "loss": 0.0263, "step": 13484 }, { "epoch": 1.4619470945359931, "grad_norm": 0.11000169068574905, "learning_rate": 5.145437400261135e-05, "loss": 0.0089, "step": 13485 }, { "epoch": 1.462055507372073, "grad_norm": 0.10329893976449966, "learning_rate": 5.1450747134774415e-05, "loss": 0.0054, "step": 13486 }, { "epoch": 1.4621639202081527, "grad_norm": 0.566037118434906, "learning_rate": 5.144712026693748e-05, "loss": 0.0186, "step": 13487 }, { "epoch": 1.4622723330442324, "grad_norm": 0.0565674789249897, "learning_rate": 5.144349339910054e-05, "loss": 0.0033, "step": 13488 }, { "epoch": 1.4623807458803122, "grad_norm": 0.15827640891075134, "learning_rate": 5.14398665312636e-05, "loss": 0.0172, "step": 13489 }, { "epoch": 1.462489158716392, "grad_norm": 0.19249975681304932, "learning_rate": 5.1436239663426664e-05, "loss": 0.0052, "step": 13490 }, { "epoch": 1.4625975715524717, "grad_norm": 0.18415567278862, "learning_rate": 5.143261279558973e-05, "loss": 0.0089, "step": 13491 }, { "epoch": 1.4627059843885517, "grad_norm": 0.12125144153833389, "learning_rate": 5.142898592775279e-05, "loss": 0.0066, "step": 13492 }, { "epoch": 1.4628143972246315, "grad_norm": 0.10862355679273605, "learning_rate": 5.1425359059915857e-05, "loss": 0.0059, "step": 13493 }, { "epoch": 1.4629228100607112, "grad_norm": 0.5271425843238831, "learning_rate": 5.142173219207892e-05, "loss": 0.0591, "step": 13494 }, { "epoch": 1.463031222896791, "grad_norm": 0.36053764820098877, "learning_rate": 5.1418105324241985e-05, "loss": 0.0159, "step": 13495 }, { "epoch": 1.4631396357328708, "grad_norm": 0.7441746592521667, "learning_rate": 5.141447845640505e-05, "loss": 0.014, "step": 13496 }, { "epoch": 1.4632480485689505, "grad_norm": 0.41629600524902344, "learning_rate": 5.141085158856811e-05, "loss": 0.0323, "step": 13497 }, { "epoch": 1.4633564614050303, "grad_norm": 0.24390211701393127, "learning_rate": 5.1407224720731184e-05, "loss": 0.0179, "step": 13498 }, { "epoch": 1.4634648742411103, "grad_norm": 0.05004759505391121, "learning_rate": 5.140359785289425e-05, "loss": 0.0035, "step": 13499 }, { "epoch": 1.4635732870771898, "grad_norm": 0.14995454251766205, "learning_rate": 5.139997098505731e-05, "loss": 0.0109, "step": 13500 }, { "epoch": 1.4636816999132698, "grad_norm": 0.2750101685523987, "learning_rate": 5.1396344117220376e-05, "loss": 0.0256, "step": 13501 }, { "epoch": 1.4637901127493496, "grad_norm": 0.5448390245437622, "learning_rate": 5.139271724938344e-05, "loss": 0.051, "step": 13502 }, { "epoch": 1.4638985255854293, "grad_norm": 0.1505993902683258, "learning_rate": 5.13890903815465e-05, "loss": 0.0073, "step": 13503 }, { "epoch": 1.464006938421509, "grad_norm": 0.22511595487594604, "learning_rate": 5.138546351370956e-05, "loss": 0.0314, "step": 13504 }, { "epoch": 1.4641153512575888, "grad_norm": 0.34273195266723633, "learning_rate": 5.1381836645872626e-05, "loss": 0.0251, "step": 13505 }, { "epoch": 1.4642237640936686, "grad_norm": 0.9734519124031067, "learning_rate": 5.137820977803569e-05, "loss": 0.0861, "step": 13506 }, { "epoch": 1.4643321769297484, "grad_norm": 0.2157066911458969, "learning_rate": 5.1374582910198754e-05, "loss": 0.0186, "step": 13507 }, { "epoch": 1.4644405897658284, "grad_norm": 0.16396772861480713, "learning_rate": 5.137095604236182e-05, "loss": 0.0551, "step": 13508 }, { "epoch": 1.4645490026019081, "grad_norm": 0.12542220950126648, "learning_rate": 5.136732917452488e-05, "loss": 0.0051, "step": 13509 }, { "epoch": 1.4646574154379879, "grad_norm": 0.2898563742637634, "learning_rate": 5.1363702306687946e-05, "loss": 0.021, "step": 13510 }, { "epoch": 1.4647658282740676, "grad_norm": 0.4314703047275543, "learning_rate": 5.136007543885101e-05, "loss": 0.0085, "step": 13511 }, { "epoch": 1.4648742411101474, "grad_norm": 0.1332511007785797, "learning_rate": 5.135644857101407e-05, "loss": 0.0042, "step": 13512 }, { "epoch": 1.4649826539462272, "grad_norm": 1.5244026184082031, "learning_rate": 5.135282170317713e-05, "loss": 0.0629, "step": 13513 }, { "epoch": 1.465091066782307, "grad_norm": 0.82462477684021, "learning_rate": 5.1349194835340196e-05, "loss": 0.0735, "step": 13514 }, { "epoch": 1.465199479618387, "grad_norm": 0.16121357679367065, "learning_rate": 5.1345567967503274e-05, "loss": 0.0128, "step": 13515 }, { "epoch": 1.4653078924544667, "grad_norm": 0.2598536014556885, "learning_rate": 5.134194109966634e-05, "loss": 0.0102, "step": 13516 }, { "epoch": 1.4654163052905465, "grad_norm": 0.41122856736183167, "learning_rate": 5.13383142318294e-05, "loss": 0.0215, "step": 13517 }, { "epoch": 1.4655247181266262, "grad_norm": 0.6082789301872253, "learning_rate": 5.133468736399246e-05, "loss": 0.0488, "step": 13518 }, { "epoch": 1.465633130962706, "grad_norm": 0.3422655761241913, "learning_rate": 5.133106049615552e-05, "loss": 0.0392, "step": 13519 }, { "epoch": 1.4657415437987857, "grad_norm": 0.7504053115844727, "learning_rate": 5.132743362831859e-05, "loss": 0.033, "step": 13520 }, { "epoch": 1.4658499566348655, "grad_norm": 0.4190962016582489, "learning_rate": 5.132380676048165e-05, "loss": 0.0324, "step": 13521 }, { "epoch": 1.4659583694709455, "grad_norm": 0.07304127514362335, "learning_rate": 5.1320179892644716e-05, "loss": 0.0037, "step": 13522 }, { "epoch": 1.466066782307025, "grad_norm": 0.22517193853855133, "learning_rate": 5.131655302480778e-05, "loss": 0.0054, "step": 13523 }, { "epoch": 1.466175195143105, "grad_norm": 0.14754895865917206, "learning_rate": 5.1312926156970844e-05, "loss": 0.008, "step": 13524 }, { "epoch": 1.4662836079791848, "grad_norm": 0.424998939037323, "learning_rate": 5.130929928913391e-05, "loss": 0.0387, "step": 13525 }, { "epoch": 1.4663920208152645, "grad_norm": 0.18710164725780487, "learning_rate": 5.1305672421296965e-05, "loss": 0.021, "step": 13526 }, { "epoch": 1.4665004336513443, "grad_norm": 0.04178153723478317, "learning_rate": 5.130204555346003e-05, "loss": 0.0021, "step": 13527 }, { "epoch": 1.466608846487424, "grad_norm": 0.3182346224784851, "learning_rate": 5.1298418685623093e-05, "loss": 0.0157, "step": 13528 }, { "epoch": 1.4667172593235038, "grad_norm": 0.4978380501270294, "learning_rate": 5.129479181778616e-05, "loss": 0.0496, "step": 13529 }, { "epoch": 1.4668256721595836, "grad_norm": 0.3285292088985443, "learning_rate": 5.129116494994922e-05, "loss": 0.0172, "step": 13530 }, { "epoch": 1.4669340849956636, "grad_norm": 0.38770896196365356, "learning_rate": 5.1287538082112286e-05, "loss": 0.0103, "step": 13531 }, { "epoch": 1.4670424978317433, "grad_norm": 0.18860799074172974, "learning_rate": 5.128391121427536e-05, "loss": 0.0146, "step": 13532 }, { "epoch": 1.467150910667823, "grad_norm": 0.27961644530296326, "learning_rate": 5.128028434643842e-05, "loss": 0.0135, "step": 13533 }, { "epoch": 1.4672593235039029, "grad_norm": 0.13948076963424683, "learning_rate": 5.1276657478601485e-05, "loss": 0.0297, "step": 13534 }, { "epoch": 1.4673677363399826, "grad_norm": 0.44957640767097473, "learning_rate": 5.127303061076455e-05, "loss": 0.0317, "step": 13535 }, { "epoch": 1.4674761491760624, "grad_norm": 0.2340390533208847, "learning_rate": 5.126940374292761e-05, "loss": 0.0146, "step": 13536 }, { "epoch": 1.4675845620121422, "grad_norm": 0.23584303259849548, "learning_rate": 5.126577687509068e-05, "loss": 0.0081, "step": 13537 }, { "epoch": 1.4676929748482221, "grad_norm": 0.23861834406852722, "learning_rate": 5.126215000725374e-05, "loss": 0.0058, "step": 13538 }, { "epoch": 1.4678013876843017, "grad_norm": 0.35768744349479675, "learning_rate": 5.1258523139416805e-05, "loss": 0.0335, "step": 13539 }, { "epoch": 1.4679098005203817, "grad_norm": 0.17467644810676575, "learning_rate": 5.125489627157987e-05, "loss": 0.0128, "step": 13540 }, { "epoch": 1.4680182133564614, "grad_norm": 0.3323681056499481, "learning_rate": 5.125126940374293e-05, "loss": 0.0584, "step": 13541 }, { "epoch": 1.4681266261925412, "grad_norm": 0.7495136857032776, "learning_rate": 5.124764253590599e-05, "loss": 0.037, "step": 13542 }, { "epoch": 1.468235039028621, "grad_norm": 0.1949152797460556, "learning_rate": 5.1244015668069055e-05, "loss": 0.0097, "step": 13543 }, { "epoch": 1.4683434518647007, "grad_norm": 0.15011484920978546, "learning_rate": 5.124038880023212e-05, "loss": 0.0111, "step": 13544 }, { "epoch": 1.4684518647007807, "grad_norm": 0.39266011118888855, "learning_rate": 5.123676193239518e-05, "loss": 0.042, "step": 13545 }, { "epoch": 1.4685602775368602, "grad_norm": 0.4365762770175934, "learning_rate": 5.123313506455825e-05, "loss": 0.018, "step": 13546 }, { "epoch": 1.4686686903729402, "grad_norm": 0.2100164294242859, "learning_rate": 5.122950819672131e-05, "loss": 0.0145, "step": 13547 }, { "epoch": 1.46877710320902, "grad_norm": 0.43932849168777466, "learning_rate": 5.1225881328884376e-05, "loss": 0.0097, "step": 13548 }, { "epoch": 1.4688855160450998, "grad_norm": 0.1808997541666031, "learning_rate": 5.122225446104744e-05, "loss": 0.0214, "step": 13549 }, { "epoch": 1.4689939288811795, "grad_norm": 0.11134033650159836, "learning_rate": 5.121862759321051e-05, "loss": 0.0087, "step": 13550 }, { "epoch": 1.4691023417172593, "grad_norm": 1.360253095626831, "learning_rate": 5.1215000725373575e-05, "loss": 0.0328, "step": 13551 }, { "epoch": 1.469210754553339, "grad_norm": 0.6402255892753601, "learning_rate": 5.121137385753664e-05, "loss": 0.0474, "step": 13552 }, { "epoch": 1.4693191673894188, "grad_norm": 0.3253662884235382, "learning_rate": 5.12077469896997e-05, "loss": 0.0141, "step": 13553 }, { "epoch": 1.4694275802254988, "grad_norm": 0.32840079069137573, "learning_rate": 5.120412012186277e-05, "loss": 0.0271, "step": 13554 }, { "epoch": 1.4695359930615786, "grad_norm": 0.3709199130535126, "learning_rate": 5.1200493254025824e-05, "loss": 0.0155, "step": 13555 }, { "epoch": 1.4696444058976583, "grad_norm": 0.3125627338886261, "learning_rate": 5.119686638618889e-05, "loss": 0.0136, "step": 13556 }, { "epoch": 1.469752818733738, "grad_norm": 0.16401992738246918, "learning_rate": 5.119323951835195e-05, "loss": 0.0038, "step": 13557 }, { "epoch": 1.4698612315698178, "grad_norm": 0.4531974196434021, "learning_rate": 5.1189612650515017e-05, "loss": 0.0736, "step": 13558 }, { "epoch": 1.4699696444058976, "grad_norm": 0.0915517657995224, "learning_rate": 5.118598578267808e-05, "loss": 0.0069, "step": 13559 }, { "epoch": 1.4700780572419774, "grad_norm": 0.2818731367588043, "learning_rate": 5.1182358914841145e-05, "loss": 0.0127, "step": 13560 }, { "epoch": 1.4701864700780574, "grad_norm": 0.25593650341033936, "learning_rate": 5.117873204700421e-05, "loss": 0.0159, "step": 13561 }, { "epoch": 1.470294882914137, "grad_norm": 0.029830629006028175, "learning_rate": 5.117510517916727e-05, "loss": 0.0026, "step": 13562 }, { "epoch": 1.4704032957502169, "grad_norm": 0.19056323170661926, "learning_rate": 5.117147831133034e-05, "loss": 0.0412, "step": 13563 }, { "epoch": 1.4705117085862967, "grad_norm": 0.3800484836101532, "learning_rate": 5.1167851443493394e-05, "loss": 0.0392, "step": 13564 }, { "epoch": 1.4706201214223764, "grad_norm": 0.2146282196044922, "learning_rate": 5.116422457565646e-05, "loss": 0.0251, "step": 13565 }, { "epoch": 1.4707285342584562, "grad_norm": 0.20354527235031128, "learning_rate": 5.116059770781952e-05, "loss": 0.0116, "step": 13566 }, { "epoch": 1.470836947094536, "grad_norm": 0.20606951415538788, "learning_rate": 5.11569708399826e-05, "loss": 0.013, "step": 13567 }, { "epoch": 1.4709453599306157, "grad_norm": 0.28715941309928894, "learning_rate": 5.1153343972145664e-05, "loss": 0.0164, "step": 13568 }, { "epoch": 1.4710537727666955, "grad_norm": 0.09300091862678528, "learning_rate": 5.114971710430873e-05, "loss": 0.0086, "step": 13569 }, { "epoch": 1.4711621856027755, "grad_norm": 0.47060516476631165, "learning_rate": 5.1146090236471786e-05, "loss": 0.0189, "step": 13570 }, { "epoch": 1.4712705984388552, "grad_norm": 0.18953540921211243, "learning_rate": 5.114246336863485e-05, "loss": 0.0089, "step": 13571 }, { "epoch": 1.471379011274935, "grad_norm": 0.23775310814380646, "learning_rate": 5.1138836500797914e-05, "loss": 0.022, "step": 13572 }, { "epoch": 1.4714874241110147, "grad_norm": 0.073774553835392, "learning_rate": 5.113520963296098e-05, "loss": 0.0044, "step": 13573 }, { "epoch": 1.4715958369470945, "grad_norm": 0.38206785917282104, "learning_rate": 5.113158276512404e-05, "loss": 0.0148, "step": 13574 }, { "epoch": 1.4717042497831743, "grad_norm": 0.6132602691650391, "learning_rate": 5.1127955897287106e-05, "loss": 0.0199, "step": 13575 }, { "epoch": 1.471812662619254, "grad_norm": 0.7859998345375061, "learning_rate": 5.112432902945017e-05, "loss": 0.0326, "step": 13576 }, { "epoch": 1.471921075455334, "grad_norm": 0.2818068861961365, "learning_rate": 5.1120702161613235e-05, "loss": 0.0196, "step": 13577 }, { "epoch": 1.4720294882914138, "grad_norm": 0.5412297248840332, "learning_rate": 5.11170752937763e-05, "loss": 0.0784, "step": 13578 }, { "epoch": 1.4721379011274935, "grad_norm": 0.1593552529811859, "learning_rate": 5.1113448425939356e-05, "loss": 0.0129, "step": 13579 }, { "epoch": 1.4722463139635733, "grad_norm": 0.07823575288057327, "learning_rate": 5.110982155810242e-05, "loss": 0.0088, "step": 13580 }, { "epoch": 1.472354726799653, "grad_norm": 0.03020523488521576, "learning_rate": 5.1106194690265484e-05, "loss": 0.0021, "step": 13581 }, { "epoch": 1.4724631396357328, "grad_norm": 0.08524863421916962, "learning_rate": 5.110256782242855e-05, "loss": 0.0058, "step": 13582 }, { "epoch": 1.4725715524718126, "grad_norm": 0.12967059016227722, "learning_rate": 5.109894095459161e-05, "loss": 0.0116, "step": 13583 }, { "epoch": 1.4726799653078926, "grad_norm": 0.4207919239997864, "learning_rate": 5.109531408675468e-05, "loss": 0.0323, "step": 13584 }, { "epoch": 1.4727883781439721, "grad_norm": 0.36046522855758667, "learning_rate": 5.109168721891775e-05, "loss": 0.0296, "step": 13585 }, { "epoch": 1.472896790980052, "grad_norm": 0.49396389722824097, "learning_rate": 5.108806035108081e-05, "loss": 0.0155, "step": 13586 }, { "epoch": 1.4730052038161319, "grad_norm": 0.2009861320257187, "learning_rate": 5.1084433483243876e-05, "loss": 0.0143, "step": 13587 }, { "epoch": 1.4731136166522116, "grad_norm": 0.34920698404312134, "learning_rate": 5.108080661540694e-05, "loss": 0.0242, "step": 13588 }, { "epoch": 1.4732220294882914, "grad_norm": 0.4848058819770813, "learning_rate": 5.1077179747570004e-05, "loss": 0.0119, "step": 13589 }, { "epoch": 1.4733304423243712, "grad_norm": 0.6124957203865051, "learning_rate": 5.107355287973307e-05, "loss": 0.0557, "step": 13590 }, { "epoch": 1.473438855160451, "grad_norm": 0.1487186700105667, "learning_rate": 5.106992601189613e-05, "loss": 0.0092, "step": 13591 }, { "epoch": 1.4735472679965307, "grad_norm": 0.1913987696170807, "learning_rate": 5.1066299144059196e-05, "loss": 0.0182, "step": 13592 }, { "epoch": 1.4736556808326107, "grad_norm": 0.3048582077026367, "learning_rate": 5.1062672276222253e-05, "loss": 0.0244, "step": 13593 }, { "epoch": 1.4737640936686904, "grad_norm": 0.37604960799217224, "learning_rate": 5.105904540838532e-05, "loss": 0.0247, "step": 13594 }, { "epoch": 1.4738725065047702, "grad_norm": 0.08390282839536667, "learning_rate": 5.105541854054838e-05, "loss": 0.0042, "step": 13595 }, { "epoch": 1.47398091934085, "grad_norm": 0.01850826106965542, "learning_rate": 5.1051791672711446e-05, "loss": 0.0013, "step": 13596 }, { "epoch": 1.4740893321769297, "grad_norm": 1.4067175388336182, "learning_rate": 5.104816480487451e-05, "loss": 0.032, "step": 13597 }, { "epoch": 1.4741977450130095, "grad_norm": 0.407585084438324, "learning_rate": 5.1044537937037574e-05, "loss": 0.0306, "step": 13598 }, { "epoch": 1.4743061578490892, "grad_norm": 0.7592625617980957, "learning_rate": 5.104091106920064e-05, "loss": 0.0365, "step": 13599 }, { "epoch": 1.4744145706851692, "grad_norm": 0.2213350236415863, "learning_rate": 5.10372842013637e-05, "loss": 0.0076, "step": 13600 }, { "epoch": 1.474522983521249, "grad_norm": 0.21007592976093292, "learning_rate": 5.103365733352677e-05, "loss": 0.0281, "step": 13601 }, { "epoch": 1.4746313963573288, "grad_norm": 0.32071590423583984, "learning_rate": 5.103003046568984e-05, "loss": 0.0349, "step": 13602 }, { "epoch": 1.4747398091934085, "grad_norm": 0.3545663058757782, "learning_rate": 5.10264035978529e-05, "loss": 0.0084, "step": 13603 }, { "epoch": 1.4748482220294883, "grad_norm": 0.0912868082523346, "learning_rate": 5.1022776730015965e-05, "loss": 0.0032, "step": 13604 }, { "epoch": 1.474956634865568, "grad_norm": 0.28584155440330505, "learning_rate": 5.101914986217903e-05, "loss": 0.0172, "step": 13605 }, { "epoch": 1.4750650477016478, "grad_norm": 0.0996532142162323, "learning_rate": 5.1015522994342094e-05, "loss": 0.0132, "step": 13606 }, { "epoch": 1.4751734605377278, "grad_norm": 0.5653918981552124, "learning_rate": 5.101189612650515e-05, "loss": 0.0425, "step": 13607 }, { "epoch": 1.4752818733738073, "grad_norm": 0.3601658344268799, "learning_rate": 5.1008269258668215e-05, "loss": 0.0287, "step": 13608 }, { "epoch": 1.4753902862098873, "grad_norm": 0.44537797570228577, "learning_rate": 5.100464239083128e-05, "loss": 0.0156, "step": 13609 }, { "epoch": 1.475498699045967, "grad_norm": 0.13657672703266144, "learning_rate": 5.100101552299434e-05, "loss": 0.0139, "step": 13610 }, { "epoch": 1.4756071118820469, "grad_norm": 0.512143611907959, "learning_rate": 5.099738865515741e-05, "loss": 0.0767, "step": 13611 }, { "epoch": 1.4757155247181266, "grad_norm": 0.018129371106624603, "learning_rate": 5.099376178732047e-05, "loss": 0.0008, "step": 13612 }, { "epoch": 1.4758239375542064, "grad_norm": 0.354849636554718, "learning_rate": 5.0990134919483536e-05, "loss": 0.0184, "step": 13613 }, { "epoch": 1.4759323503902861, "grad_norm": 0.7042337656021118, "learning_rate": 5.09865080516466e-05, "loss": 0.0653, "step": 13614 }, { "epoch": 1.476040763226366, "grad_norm": 0.20824648439884186, "learning_rate": 5.0982881183809664e-05, "loss": 0.0279, "step": 13615 }, { "epoch": 1.476149176062446, "grad_norm": 0.23431843519210815, "learning_rate": 5.097925431597272e-05, "loss": 0.025, "step": 13616 }, { "epoch": 1.4762575888985257, "grad_norm": 0.1818804144859314, "learning_rate": 5.0975627448135785e-05, "loss": 0.0068, "step": 13617 }, { "epoch": 1.4763660017346054, "grad_norm": 0.5027385950088501, "learning_rate": 5.097200058029885e-05, "loss": 0.0512, "step": 13618 }, { "epoch": 1.4764744145706852, "grad_norm": 0.3144082725048065, "learning_rate": 5.096837371246193e-05, "loss": 0.0359, "step": 13619 }, { "epoch": 1.476582827406765, "grad_norm": 0.2501535415649414, "learning_rate": 5.096474684462499e-05, "loss": 0.0101, "step": 13620 }, { "epoch": 1.4766912402428447, "grad_norm": 0.7441010475158691, "learning_rate": 5.0961119976788055e-05, "loss": 0.0289, "step": 13621 }, { "epoch": 1.4767996530789245, "grad_norm": 0.01180132944136858, "learning_rate": 5.095749310895111e-05, "loss": 0.0008, "step": 13622 }, { "epoch": 1.4769080659150045, "grad_norm": 0.3802168369293213, "learning_rate": 5.095386624111418e-05, "loss": 0.0332, "step": 13623 }, { "epoch": 1.477016478751084, "grad_norm": 0.6109721660614014, "learning_rate": 5.095023937327724e-05, "loss": 0.0317, "step": 13624 }, { "epoch": 1.477124891587164, "grad_norm": 0.48009172081947327, "learning_rate": 5.0946612505440305e-05, "loss": 0.0097, "step": 13625 }, { "epoch": 1.4772333044232437, "grad_norm": 0.11219864338636398, "learning_rate": 5.094298563760337e-05, "loss": 0.0068, "step": 13626 }, { "epoch": 1.4773417172593235, "grad_norm": 0.1580335646867752, "learning_rate": 5.093935876976643e-05, "loss": 0.0189, "step": 13627 }, { "epoch": 1.4774501300954033, "grad_norm": 0.3157771825790405, "learning_rate": 5.09357319019295e-05, "loss": 0.0466, "step": 13628 }, { "epoch": 1.477558542931483, "grad_norm": 0.027404554188251495, "learning_rate": 5.093210503409256e-05, "loss": 0.0016, "step": 13629 }, { "epoch": 1.477666955767563, "grad_norm": 0.24819165468215942, "learning_rate": 5.0928478166255625e-05, "loss": 0.0284, "step": 13630 }, { "epoch": 1.4777753686036426, "grad_norm": 0.12479939311742783, "learning_rate": 5.092485129841868e-05, "loss": 0.0046, "step": 13631 }, { "epoch": 1.4778837814397225, "grad_norm": 0.24160565435886383, "learning_rate": 5.092122443058175e-05, "loss": 0.0392, "step": 13632 }, { "epoch": 1.4779921942758023, "grad_norm": 0.4089999198913574, "learning_rate": 5.091759756274481e-05, "loss": 0.0158, "step": 13633 }, { "epoch": 1.478100607111882, "grad_norm": 0.5394616723060608, "learning_rate": 5.0913970694907875e-05, "loss": 0.0235, "step": 13634 }, { "epoch": 1.4782090199479618, "grad_norm": 0.2775457203388214, "learning_rate": 5.091034382707094e-05, "loss": 0.023, "step": 13635 }, { "epoch": 1.4783174327840416, "grad_norm": 0.16327382624149323, "learning_rate": 5.090671695923401e-05, "loss": 0.0323, "step": 13636 }, { "epoch": 1.4784258456201214, "grad_norm": 0.5937915444374084, "learning_rate": 5.0903090091397074e-05, "loss": 0.0327, "step": 13637 }, { "epoch": 1.4785342584562011, "grad_norm": 0.25679782032966614, "learning_rate": 5.089946322356014e-05, "loss": 0.0247, "step": 13638 }, { "epoch": 1.478642671292281, "grad_norm": 0.12792670726776123, "learning_rate": 5.08958363557232e-05, "loss": 0.0064, "step": 13639 }, { "epoch": 1.4787510841283609, "grad_norm": 0.06960596144199371, "learning_rate": 5.0892209487886266e-05, "loss": 0.0029, "step": 13640 }, { "epoch": 1.4788594969644406, "grad_norm": 0.06966380774974823, "learning_rate": 5.088858262004933e-05, "loss": 0.0056, "step": 13641 }, { "epoch": 1.4789679098005204, "grad_norm": 0.30574971437454224, "learning_rate": 5.0884955752212395e-05, "loss": 0.0314, "step": 13642 }, { "epoch": 1.4790763226366002, "grad_norm": 0.14818039536476135, "learning_rate": 5.088132888437546e-05, "loss": 0.0206, "step": 13643 }, { "epoch": 1.47918473547268, "grad_norm": 0.09070871770381927, "learning_rate": 5.087770201653852e-05, "loss": 0.0086, "step": 13644 }, { "epoch": 1.4792931483087597, "grad_norm": 1.060278058052063, "learning_rate": 5.087407514870158e-05, "loss": 0.0398, "step": 13645 }, { "epoch": 1.4794015611448397, "grad_norm": 0.33550456166267395, "learning_rate": 5.0870448280864644e-05, "loss": 0.0293, "step": 13646 }, { "epoch": 1.4795099739809192, "grad_norm": 0.35230788588523865, "learning_rate": 5.086682141302771e-05, "loss": 0.0152, "step": 13647 }, { "epoch": 1.4796183868169992, "grad_norm": 0.2242741584777832, "learning_rate": 5.086319454519077e-05, "loss": 0.0241, "step": 13648 }, { "epoch": 1.479726799653079, "grad_norm": 0.13522398471832275, "learning_rate": 5.0859567677353837e-05, "loss": 0.0087, "step": 13649 }, { "epoch": 1.4798352124891587, "grad_norm": 0.3496687114238739, "learning_rate": 5.08559408095169e-05, "loss": 0.0565, "step": 13650 }, { "epoch": 1.4799436253252385, "grad_norm": 0.22047993540763855, "learning_rate": 5.0852313941679965e-05, "loss": 0.016, "step": 13651 }, { "epoch": 1.4800520381613183, "grad_norm": 0.33319205045700073, "learning_rate": 5.084868707384303e-05, "loss": 0.0138, "step": 13652 }, { "epoch": 1.480160450997398, "grad_norm": 0.14819484949111938, "learning_rate": 5.08450602060061e-05, "loss": 0.0192, "step": 13653 }, { "epoch": 1.4802688638334778, "grad_norm": 0.30572420358657837, "learning_rate": 5.0841433338169164e-05, "loss": 0.0198, "step": 13654 }, { "epoch": 1.4803772766695578, "grad_norm": 0.8793843984603882, "learning_rate": 5.083780647033223e-05, "loss": 0.0231, "step": 13655 }, { "epoch": 1.4804856895056375, "grad_norm": 0.3165290951728821, "learning_rate": 5.083417960249529e-05, "loss": 0.0186, "step": 13656 }, { "epoch": 1.4805941023417173, "grad_norm": 0.4653789699077606, "learning_rate": 5.0830552734658356e-05, "loss": 0.0114, "step": 13657 }, { "epoch": 1.480702515177797, "grad_norm": 1.0238224267959595, "learning_rate": 5.082692586682142e-05, "loss": 0.0663, "step": 13658 }, { "epoch": 1.4808109280138768, "grad_norm": 0.3780401945114136, "learning_rate": 5.082329899898448e-05, "loss": 0.0418, "step": 13659 }, { "epoch": 1.4809193408499566, "grad_norm": 0.22130116820335388, "learning_rate": 5.081967213114754e-05, "loss": 0.0155, "step": 13660 }, { "epoch": 1.4810277536860363, "grad_norm": 0.3935549855232239, "learning_rate": 5.0816045263310606e-05, "loss": 0.0559, "step": 13661 }, { "epoch": 1.4811361665221163, "grad_norm": 0.20015102624893188, "learning_rate": 5.081241839547367e-05, "loss": 0.0381, "step": 13662 }, { "epoch": 1.481244579358196, "grad_norm": 0.2568451464176178, "learning_rate": 5.0808791527636734e-05, "loss": 0.0521, "step": 13663 }, { "epoch": 1.4813529921942759, "grad_norm": 0.4830646216869354, "learning_rate": 5.08051646597998e-05, "loss": 0.0252, "step": 13664 }, { "epoch": 1.4814614050303556, "grad_norm": 0.7184375524520874, "learning_rate": 5.080153779196286e-05, "loss": 0.049, "step": 13665 }, { "epoch": 1.4815698178664354, "grad_norm": 0.3560299873352051, "learning_rate": 5.0797910924125926e-05, "loss": 0.0277, "step": 13666 }, { "epoch": 1.4816782307025151, "grad_norm": 0.2873504161834717, "learning_rate": 5.079428405628899e-05, "loss": 0.0185, "step": 13667 }, { "epoch": 1.481786643538595, "grad_norm": 0.22537493705749512, "learning_rate": 5.079065718845205e-05, "loss": 0.0149, "step": 13668 }, { "epoch": 1.481895056374675, "grad_norm": 0.7090066075325012, "learning_rate": 5.078703032061511e-05, "loss": 0.0609, "step": 13669 }, { "epoch": 1.4820034692107544, "grad_norm": 0.3944646716117859, "learning_rate": 5.0783403452778176e-05, "loss": 0.0288, "step": 13670 }, { "epoch": 1.4821118820468344, "grad_norm": 0.2053227424621582, "learning_rate": 5.0779776584941254e-05, "loss": 0.009, "step": 13671 }, { "epoch": 1.4822202948829142, "grad_norm": 0.5621856451034546, "learning_rate": 5.077614971710432e-05, "loss": 0.0246, "step": 13672 }, { "epoch": 1.482328707718994, "grad_norm": 0.3517876863479614, "learning_rate": 5.077252284926738e-05, "loss": 0.0515, "step": 13673 }, { "epoch": 1.4824371205550737, "grad_norm": 0.10747452825307846, "learning_rate": 5.076889598143044e-05, "loss": 0.0067, "step": 13674 }, { "epoch": 1.4825455333911535, "grad_norm": 0.3773535490036011, "learning_rate": 5.07652691135935e-05, "loss": 0.0476, "step": 13675 }, { "epoch": 1.4826539462272332, "grad_norm": 0.15450993180274963, "learning_rate": 5.076164224575657e-05, "loss": 0.0085, "step": 13676 }, { "epoch": 1.482762359063313, "grad_norm": 0.1317933052778244, "learning_rate": 5.075801537791963e-05, "loss": 0.0079, "step": 13677 }, { "epoch": 1.482870771899393, "grad_norm": 0.26477864384651184, "learning_rate": 5.0754388510082696e-05, "loss": 0.0345, "step": 13678 }, { "epoch": 1.4829791847354727, "grad_norm": 0.29126912355422974, "learning_rate": 5.075076164224576e-05, "loss": 0.0281, "step": 13679 }, { "epoch": 1.4830875975715525, "grad_norm": 0.24426083266735077, "learning_rate": 5.0747134774408824e-05, "loss": 0.0205, "step": 13680 }, { "epoch": 1.4831960104076323, "grad_norm": 0.4302462935447693, "learning_rate": 5.074350790657189e-05, "loss": 0.038, "step": 13681 }, { "epoch": 1.483304423243712, "grad_norm": 0.3724610507488251, "learning_rate": 5.073988103873495e-05, "loss": 0.0164, "step": 13682 }, { "epoch": 1.4834128360797918, "grad_norm": 0.2591601312160492, "learning_rate": 5.073625417089801e-05, "loss": 0.0308, "step": 13683 }, { "epoch": 1.4835212489158716, "grad_norm": 0.2785629630088806, "learning_rate": 5.0732627303061073e-05, "loss": 0.0287, "step": 13684 }, { "epoch": 1.4836296617519515, "grad_norm": 0.3586966097354889, "learning_rate": 5.072900043522414e-05, "loss": 0.0177, "step": 13685 }, { "epoch": 1.483738074588031, "grad_norm": 0.3728211224079132, "learning_rate": 5.07253735673872e-05, "loss": 0.0108, "step": 13686 }, { "epoch": 1.483846487424111, "grad_norm": 0.37486523389816284, "learning_rate": 5.0721746699550266e-05, "loss": 0.0223, "step": 13687 }, { "epoch": 1.4839549002601908, "grad_norm": 0.0698484405875206, "learning_rate": 5.071811983171334e-05, "loss": 0.0024, "step": 13688 }, { "epoch": 1.4840633130962706, "grad_norm": 0.41767826676368713, "learning_rate": 5.07144929638764e-05, "loss": 0.0231, "step": 13689 }, { "epoch": 1.4841717259323504, "grad_norm": 0.38763538002967834, "learning_rate": 5.0710866096039465e-05, "loss": 0.0216, "step": 13690 }, { "epoch": 1.4842801387684301, "grad_norm": 0.03361250087618828, "learning_rate": 5.070723922820253e-05, "loss": 0.0027, "step": 13691 }, { "epoch": 1.48438855160451, "grad_norm": 0.057555895298719406, "learning_rate": 5.070361236036559e-05, "loss": 0.003, "step": 13692 }, { "epoch": 1.4844969644405897, "grad_norm": 0.22759082913398743, "learning_rate": 5.069998549252866e-05, "loss": 0.012, "step": 13693 }, { "epoch": 1.4846053772766696, "grad_norm": 0.638359785079956, "learning_rate": 5.069635862469172e-05, "loss": 0.0247, "step": 13694 }, { "epoch": 1.4847137901127494, "grad_norm": 0.2155371457338333, "learning_rate": 5.0692731756854785e-05, "loss": 0.0186, "step": 13695 }, { "epoch": 1.4848222029488292, "grad_norm": 0.14245642721652985, "learning_rate": 5.068910488901785e-05, "loss": 0.0098, "step": 13696 }, { "epoch": 1.484930615784909, "grad_norm": 0.19308896362781525, "learning_rate": 5.068547802118091e-05, "loss": 0.0197, "step": 13697 }, { "epoch": 1.4850390286209887, "grad_norm": 0.2576218843460083, "learning_rate": 5.068185115334397e-05, "loss": 0.0302, "step": 13698 }, { "epoch": 1.4851474414570685, "grad_norm": 0.46912914514541626, "learning_rate": 5.0678224285507035e-05, "loss": 0.0406, "step": 13699 }, { "epoch": 1.4852558542931482, "grad_norm": 0.24947288632392883, "learning_rate": 5.06745974176701e-05, "loss": 0.0244, "step": 13700 }, { "epoch": 1.4853642671292282, "grad_norm": 0.45726361870765686, "learning_rate": 5.067097054983316e-05, "loss": 0.0721, "step": 13701 }, { "epoch": 1.485472679965308, "grad_norm": 0.44408273696899414, "learning_rate": 5.066734368199623e-05, "loss": 0.0274, "step": 13702 }, { "epoch": 1.4855810928013877, "grad_norm": 0.10238838195800781, "learning_rate": 5.066371681415929e-05, "loss": 0.0045, "step": 13703 }, { "epoch": 1.4856895056374675, "grad_norm": 0.26823845505714417, "learning_rate": 5.0660089946322356e-05, "loss": 0.0234, "step": 13704 }, { "epoch": 1.4857979184735473, "grad_norm": 0.17873522639274597, "learning_rate": 5.0656463078485426e-05, "loss": 0.0086, "step": 13705 }, { "epoch": 1.485906331309627, "grad_norm": 0.508054792881012, "learning_rate": 5.065283621064849e-05, "loss": 0.0544, "step": 13706 }, { "epoch": 1.4860147441457068, "grad_norm": 0.1889670193195343, "learning_rate": 5.0649209342811555e-05, "loss": 0.022, "step": 13707 }, { "epoch": 1.4861231569817868, "grad_norm": 0.2645445168018341, "learning_rate": 5.064558247497462e-05, "loss": 0.0137, "step": 13708 }, { "epoch": 1.4862315698178663, "grad_norm": 0.4072686433792114, "learning_rate": 5.064195560713768e-05, "loss": 0.0426, "step": 13709 }, { "epoch": 1.4863399826539463, "grad_norm": 0.31663382053375244, "learning_rate": 5.063832873930075e-05, "loss": 0.0209, "step": 13710 }, { "epoch": 1.486448395490026, "grad_norm": 0.28808486461639404, "learning_rate": 5.063470187146381e-05, "loss": 0.0341, "step": 13711 }, { "epoch": 1.4865568083261058, "grad_norm": 0.5093058347702026, "learning_rate": 5.063107500362687e-05, "loss": 0.0645, "step": 13712 }, { "epoch": 1.4866652211621856, "grad_norm": 0.2711334526538849, "learning_rate": 5.062744813578993e-05, "loss": 0.0354, "step": 13713 }, { "epoch": 1.4867736339982653, "grad_norm": 0.4088551104068756, "learning_rate": 5.0623821267953e-05, "loss": 0.0352, "step": 13714 }, { "epoch": 1.486882046834345, "grad_norm": 0.5395827889442444, "learning_rate": 5.062019440011606e-05, "loss": 0.0184, "step": 13715 }, { "epoch": 1.4869904596704249, "grad_norm": 0.6393004655838013, "learning_rate": 5.0616567532279125e-05, "loss": 0.0222, "step": 13716 }, { "epoch": 1.4870988725065049, "grad_norm": 0.2548244297504425, "learning_rate": 5.061294066444219e-05, "loss": 0.0266, "step": 13717 }, { "epoch": 1.4872072853425846, "grad_norm": 0.4124760031700134, "learning_rate": 5.060931379660525e-05, "loss": 0.0115, "step": 13718 }, { "epoch": 1.4873156981786644, "grad_norm": 0.2912702262401581, "learning_rate": 5.060568692876832e-05, "loss": 0.0343, "step": 13719 }, { "epoch": 1.4874241110147441, "grad_norm": 0.3598974347114563, "learning_rate": 5.0602060060931375e-05, "loss": 0.0434, "step": 13720 }, { "epoch": 1.487532523850824, "grad_norm": 0.4753813147544861, "learning_rate": 5.059843319309444e-05, "loss": 0.0164, "step": 13721 }, { "epoch": 1.4876409366869037, "grad_norm": 0.2833697199821472, "learning_rate": 5.0594806325257516e-05, "loss": 0.0171, "step": 13722 }, { "epoch": 1.4877493495229834, "grad_norm": 0.5057147145271301, "learning_rate": 5.059117945742058e-05, "loss": 0.0289, "step": 13723 }, { "epoch": 1.4878577623590634, "grad_norm": 0.8333646059036255, "learning_rate": 5.0587552589583644e-05, "loss": 0.0247, "step": 13724 }, { "epoch": 1.4879661751951432, "grad_norm": 0.11554352194070816, "learning_rate": 5.058392572174671e-05, "loss": 0.0076, "step": 13725 }, { "epoch": 1.488074588031223, "grad_norm": 0.3379496932029724, "learning_rate": 5.0580298853909766e-05, "loss": 0.0217, "step": 13726 }, { "epoch": 1.4881830008673027, "grad_norm": 0.15484707057476044, "learning_rate": 5.057667198607283e-05, "loss": 0.0092, "step": 13727 }, { "epoch": 1.4882914137033825, "grad_norm": 0.33701467514038086, "learning_rate": 5.0573045118235894e-05, "loss": 0.0393, "step": 13728 }, { "epoch": 1.4883998265394622, "grad_norm": 0.21008561551570892, "learning_rate": 5.056941825039896e-05, "loss": 0.0164, "step": 13729 }, { "epoch": 1.488508239375542, "grad_norm": 1.1348116397857666, "learning_rate": 5.056579138256202e-05, "loss": 0.029, "step": 13730 }, { "epoch": 1.488616652211622, "grad_norm": 0.3479291498661041, "learning_rate": 5.0562164514725086e-05, "loss": 0.0191, "step": 13731 }, { "epoch": 1.4887250650477015, "grad_norm": 0.18534113466739655, "learning_rate": 5.055853764688815e-05, "loss": 0.0197, "step": 13732 }, { "epoch": 1.4888334778837815, "grad_norm": 0.35380229353904724, "learning_rate": 5.0554910779051215e-05, "loss": 0.0282, "step": 13733 }, { "epoch": 1.4889418907198613, "grad_norm": 0.13114529848098755, "learning_rate": 5.055128391121428e-05, "loss": 0.0047, "step": 13734 }, { "epoch": 1.489050303555941, "grad_norm": 0.12049073725938797, "learning_rate": 5.0547657043377336e-05, "loss": 0.0458, "step": 13735 }, { "epoch": 1.4891587163920208, "grad_norm": 0.7175915241241455, "learning_rate": 5.05440301755404e-05, "loss": 0.0368, "step": 13736 }, { "epoch": 1.4892671292281006, "grad_norm": 0.250973105430603, "learning_rate": 5.0540403307703464e-05, "loss": 0.0237, "step": 13737 }, { "epoch": 1.4893755420641803, "grad_norm": 0.3275166153907776, "learning_rate": 5.053677643986653e-05, "loss": 0.0347, "step": 13738 }, { "epoch": 1.48948395490026, "grad_norm": 0.3214905261993408, "learning_rate": 5.053314957202959e-05, "loss": 0.0398, "step": 13739 }, { "epoch": 1.48959236773634, "grad_norm": 0.2154879868030548, "learning_rate": 5.052952270419266e-05, "loss": 0.0181, "step": 13740 }, { "epoch": 1.4897007805724198, "grad_norm": 0.21172641217708588, "learning_rate": 5.052589583635573e-05, "loss": 0.0244, "step": 13741 }, { "epoch": 1.4898091934084996, "grad_norm": 0.2442108392715454, "learning_rate": 5.052226896851879e-05, "loss": 0.0276, "step": 13742 }, { "epoch": 1.4899176062445794, "grad_norm": 0.37019142508506775, "learning_rate": 5.0518642100681856e-05, "loss": 0.0201, "step": 13743 }, { "epoch": 1.4900260190806591, "grad_norm": 0.3323217034339905, "learning_rate": 5.051501523284492e-05, "loss": 0.0151, "step": 13744 }, { "epoch": 1.4901344319167389, "grad_norm": 0.25240129232406616, "learning_rate": 5.0511388365007984e-05, "loss": 0.013, "step": 13745 }, { "epoch": 1.4902428447528187, "grad_norm": 0.2592679560184479, "learning_rate": 5.050776149717105e-05, "loss": 0.0159, "step": 13746 }, { "epoch": 1.4903512575888986, "grad_norm": 0.3805152475833893, "learning_rate": 5.050413462933411e-05, "loss": 0.0385, "step": 13747 }, { "epoch": 1.4904596704249784, "grad_norm": 0.6409593224525452, "learning_rate": 5.0500507761497176e-05, "loss": 0.0214, "step": 13748 }, { "epoch": 1.4905680832610582, "grad_norm": 0.05796331539750099, "learning_rate": 5.0496880893660234e-05, "loss": 0.0029, "step": 13749 }, { "epoch": 1.490676496097138, "grad_norm": 0.3662923574447632, "learning_rate": 5.04932540258233e-05, "loss": 0.0303, "step": 13750 }, { "epoch": 1.4907849089332177, "grad_norm": 0.2648826837539673, "learning_rate": 5.048962715798636e-05, "loss": 0.0091, "step": 13751 }, { "epoch": 1.4908933217692975, "grad_norm": 0.29185715317726135, "learning_rate": 5.0486000290149426e-05, "loss": 0.0286, "step": 13752 }, { "epoch": 1.4910017346053772, "grad_norm": 0.24098166823387146, "learning_rate": 5.048237342231249e-05, "loss": 0.025, "step": 13753 }, { "epoch": 1.4911101474414572, "grad_norm": 1.37917959690094, "learning_rate": 5.0478746554475554e-05, "loss": 0.0474, "step": 13754 }, { "epoch": 1.4912185602775367, "grad_norm": 0.3542942702770233, "learning_rate": 5.047511968663862e-05, "loss": 0.0334, "step": 13755 }, { "epoch": 1.4913269731136167, "grad_norm": 0.2236185073852539, "learning_rate": 5.047149281880168e-05, "loss": 0.0121, "step": 13756 }, { "epoch": 1.4914353859496965, "grad_norm": 0.27804967761039734, "learning_rate": 5.046786595096475e-05, "loss": 0.0334, "step": 13757 }, { "epoch": 1.4915437987857763, "grad_norm": 0.3481195569038391, "learning_rate": 5.046423908312782e-05, "loss": 0.0122, "step": 13758 }, { "epoch": 1.491652211621856, "grad_norm": 0.28879836201667786, "learning_rate": 5.046061221529088e-05, "loss": 0.0178, "step": 13759 }, { "epoch": 1.4917606244579358, "grad_norm": 0.14615604281425476, "learning_rate": 5.0456985347453945e-05, "loss": 0.0038, "step": 13760 }, { "epoch": 1.4918690372940155, "grad_norm": 0.15281648933887482, "learning_rate": 5.045335847961701e-05, "loss": 0.0075, "step": 13761 }, { "epoch": 1.4919774501300953, "grad_norm": 0.21841496229171753, "learning_rate": 5.0449731611780074e-05, "loss": 0.0206, "step": 13762 }, { "epoch": 1.4920858629661753, "grad_norm": 0.20623821020126343, "learning_rate": 5.044610474394314e-05, "loss": 0.0253, "step": 13763 }, { "epoch": 1.492194275802255, "grad_norm": 0.19152265787124634, "learning_rate": 5.0442477876106195e-05, "loss": 0.0266, "step": 13764 }, { "epoch": 1.4923026886383348, "grad_norm": 0.30570903420448303, "learning_rate": 5.043885100826926e-05, "loss": 0.01, "step": 13765 }, { "epoch": 1.4924111014744146, "grad_norm": 0.3630973696708679, "learning_rate": 5.043522414043232e-05, "loss": 0.0318, "step": 13766 }, { "epoch": 1.4925195143104943, "grad_norm": 0.5736510157585144, "learning_rate": 5.043159727259539e-05, "loss": 0.0668, "step": 13767 }, { "epoch": 1.492627927146574, "grad_norm": 0.054719697684049606, "learning_rate": 5.042797040475845e-05, "loss": 0.0046, "step": 13768 }, { "epoch": 1.4927363399826539, "grad_norm": 0.502293586730957, "learning_rate": 5.0424343536921516e-05, "loss": 0.0258, "step": 13769 }, { "epoch": 1.4928447528187339, "grad_norm": 0.4809841513633728, "learning_rate": 5.042071666908458e-05, "loss": 0.0464, "step": 13770 }, { "epoch": 1.4929531656548134, "grad_norm": 0.20207001268863678, "learning_rate": 5.0417089801247644e-05, "loss": 0.0116, "step": 13771 }, { "epoch": 1.4930615784908934, "grad_norm": 0.12119916826486588, "learning_rate": 5.04134629334107e-05, "loss": 0.0109, "step": 13772 }, { "epoch": 1.4931699913269731, "grad_norm": 0.26309236884117126, "learning_rate": 5.0409836065573765e-05, "loss": 0.0301, "step": 13773 }, { "epoch": 1.493278404163053, "grad_norm": 0.4039737284183502, "learning_rate": 5.040620919773684e-05, "loss": 0.0167, "step": 13774 }, { "epoch": 1.4933868169991327, "grad_norm": 0.07208139449357986, "learning_rate": 5.040258232989991e-05, "loss": 0.0044, "step": 13775 }, { "epoch": 1.4934952298352124, "grad_norm": 0.15053416788578033, "learning_rate": 5.039895546206297e-05, "loss": 0.0143, "step": 13776 }, { "epoch": 1.4936036426712924, "grad_norm": 0.3551269471645355, "learning_rate": 5.0395328594226035e-05, "loss": 0.0362, "step": 13777 }, { "epoch": 1.493712055507372, "grad_norm": 0.31813564896583557, "learning_rate": 5.039170172638909e-05, "loss": 0.059, "step": 13778 }, { "epoch": 1.493820468343452, "grad_norm": 0.42419299483299255, "learning_rate": 5.038807485855216e-05, "loss": 0.0876, "step": 13779 }, { "epoch": 1.4939288811795317, "grad_norm": 0.08679946511983871, "learning_rate": 5.038444799071522e-05, "loss": 0.0072, "step": 13780 }, { "epoch": 1.4940372940156115, "grad_norm": 0.6859447956085205, "learning_rate": 5.0380821122878285e-05, "loss": 0.061, "step": 13781 }, { "epoch": 1.4941457068516912, "grad_norm": 0.978252112865448, "learning_rate": 5.037719425504135e-05, "loss": 0.0189, "step": 13782 }, { "epoch": 1.494254119687771, "grad_norm": 0.34972113370895386, "learning_rate": 5.037356738720441e-05, "loss": 0.0337, "step": 13783 }, { "epoch": 1.4943625325238508, "grad_norm": 0.09993788599967957, "learning_rate": 5.036994051936748e-05, "loss": 0.0051, "step": 13784 }, { "epoch": 1.4944709453599305, "grad_norm": 0.5051190257072449, "learning_rate": 5.036631365153054e-05, "loss": 0.0308, "step": 13785 }, { "epoch": 1.4945793581960105, "grad_norm": 0.09063567966222763, "learning_rate": 5.0362686783693605e-05, "loss": 0.009, "step": 13786 }, { "epoch": 1.4946877710320903, "grad_norm": 0.19022466242313385, "learning_rate": 5.035905991585666e-05, "loss": 0.0244, "step": 13787 }, { "epoch": 1.49479618386817, "grad_norm": 0.22736899554729462, "learning_rate": 5.035543304801973e-05, "loss": 0.0513, "step": 13788 }, { "epoch": 1.4949045967042498, "grad_norm": 0.21688640117645264, "learning_rate": 5.035180618018279e-05, "loss": 0.0185, "step": 13789 }, { "epoch": 1.4950130095403296, "grad_norm": 0.18036824464797974, "learning_rate": 5.0348179312345855e-05, "loss": 0.0102, "step": 13790 }, { "epoch": 1.4951214223764093, "grad_norm": 0.19139684736728668, "learning_rate": 5.034455244450893e-05, "loss": 0.0194, "step": 13791 }, { "epoch": 1.495229835212489, "grad_norm": 0.06692448258399963, "learning_rate": 5.034092557667199e-05, "loss": 0.0042, "step": 13792 }, { "epoch": 1.495338248048569, "grad_norm": 0.10372429341077805, "learning_rate": 5.0337298708835054e-05, "loss": 0.0075, "step": 13793 }, { "epoch": 1.4954466608846486, "grad_norm": 0.22209380567073822, "learning_rate": 5.033367184099812e-05, "loss": 0.0269, "step": 13794 }, { "epoch": 1.4955550737207286, "grad_norm": 0.28499364852905273, "learning_rate": 5.033004497316118e-05, "loss": 0.0051, "step": 13795 }, { "epoch": 1.4956634865568084, "grad_norm": 0.41482383012771606, "learning_rate": 5.0326418105324246e-05, "loss": 0.0303, "step": 13796 }, { "epoch": 1.4957718993928881, "grad_norm": 0.9296926856040955, "learning_rate": 5.032279123748731e-05, "loss": 0.0375, "step": 13797 }, { "epoch": 1.495880312228968, "grad_norm": 0.3652053773403168, "learning_rate": 5.0319164369650375e-05, "loss": 0.0405, "step": 13798 }, { "epoch": 1.4959887250650477, "grad_norm": 0.05261160060763359, "learning_rate": 5.031553750181344e-05, "loss": 0.0045, "step": 13799 }, { "epoch": 1.4960971379011274, "grad_norm": 0.38625892996788025, "learning_rate": 5.03119106339765e-05, "loss": 0.02, "step": 13800 }, { "epoch": 1.4962055507372072, "grad_norm": 0.29397663474082947, "learning_rate": 5.030828376613956e-05, "loss": 0.0155, "step": 13801 }, { "epoch": 1.4963139635732872, "grad_norm": 0.011617216281592846, "learning_rate": 5.0304656898302624e-05, "loss": 0.0006, "step": 13802 }, { "epoch": 1.496422376409367, "grad_norm": 0.04294697195291519, "learning_rate": 5.030103003046569e-05, "loss": 0.0034, "step": 13803 }, { "epoch": 1.4965307892454467, "grad_norm": 0.5212160348892212, "learning_rate": 5.029740316262875e-05, "loss": 0.0451, "step": 13804 }, { "epoch": 1.4966392020815265, "grad_norm": 0.21411344408988953, "learning_rate": 5.029377629479182e-05, "loss": 0.0127, "step": 13805 }, { "epoch": 1.4967476149176062, "grad_norm": 0.40278494358062744, "learning_rate": 5.029014942695488e-05, "loss": 0.0228, "step": 13806 }, { "epoch": 1.496856027753686, "grad_norm": 0.2934308648109436, "learning_rate": 5.0286522559117945e-05, "loss": 0.0549, "step": 13807 }, { "epoch": 1.4969644405897657, "grad_norm": 0.2276456654071808, "learning_rate": 5.028289569128101e-05, "loss": 0.0118, "step": 13808 }, { "epoch": 1.4970728534258457, "grad_norm": 0.24343928694725037, "learning_rate": 5.027926882344408e-05, "loss": 0.0124, "step": 13809 }, { "epoch": 1.4971812662619255, "grad_norm": 0.14151163399219513, "learning_rate": 5.0275641955607144e-05, "loss": 0.0094, "step": 13810 }, { "epoch": 1.4972896790980053, "grad_norm": 0.3196626603603363, "learning_rate": 5.027201508777021e-05, "loss": 0.0217, "step": 13811 }, { "epoch": 1.497398091934085, "grad_norm": 0.1079636886715889, "learning_rate": 5.026838821993327e-05, "loss": 0.0078, "step": 13812 }, { "epoch": 1.4975065047701648, "grad_norm": 0.290417343378067, "learning_rate": 5.0264761352096336e-05, "loss": 0.0161, "step": 13813 }, { "epoch": 1.4976149176062445, "grad_norm": 0.2389984428882599, "learning_rate": 5.02611344842594e-05, "loss": 0.0306, "step": 13814 }, { "epoch": 1.4977233304423243, "grad_norm": 0.0802435353398323, "learning_rate": 5.0257507616422464e-05, "loss": 0.0095, "step": 13815 }, { "epoch": 1.4978317432784043, "grad_norm": 0.34473755955696106, "learning_rate": 5.025388074858552e-05, "loss": 0.0389, "step": 13816 }, { "epoch": 1.4979401561144838, "grad_norm": 0.2281695008277893, "learning_rate": 5.0250253880748586e-05, "loss": 0.0175, "step": 13817 }, { "epoch": 1.4980485689505638, "grad_norm": 0.04371289163827896, "learning_rate": 5.024662701291165e-05, "loss": 0.003, "step": 13818 }, { "epoch": 1.4981569817866436, "grad_norm": 0.34476473927497864, "learning_rate": 5.0243000145074714e-05, "loss": 0.0127, "step": 13819 }, { "epoch": 1.4982653946227233, "grad_norm": 0.06102583929896355, "learning_rate": 5.023937327723778e-05, "loss": 0.0045, "step": 13820 }, { "epoch": 1.498373807458803, "grad_norm": 0.028152599930763245, "learning_rate": 5.023574640940084e-05, "loss": 0.0019, "step": 13821 }, { "epoch": 1.4984822202948829, "grad_norm": 0.14003443717956543, "learning_rate": 5.0232119541563906e-05, "loss": 0.0047, "step": 13822 }, { "epoch": 1.4985906331309626, "grad_norm": 0.39723286032676697, "learning_rate": 5.022849267372697e-05, "loss": 0.0153, "step": 13823 }, { "epoch": 1.4986990459670424, "grad_norm": 0.2362615019083023, "learning_rate": 5.022486580589003e-05, "loss": 0.0151, "step": 13824 }, { "epoch": 1.4988074588031224, "grad_norm": 0.6445430517196655, "learning_rate": 5.022123893805309e-05, "loss": 0.0696, "step": 13825 }, { "epoch": 1.4989158716392021, "grad_norm": 0.5375605225563049, "learning_rate": 5.021761207021617e-05, "loss": 0.0166, "step": 13826 }, { "epoch": 1.499024284475282, "grad_norm": 0.44565680623054504, "learning_rate": 5.0213985202379234e-05, "loss": 0.0234, "step": 13827 }, { "epoch": 1.4991326973113617, "grad_norm": 0.38167470693588257, "learning_rate": 5.02103583345423e-05, "loss": 0.0335, "step": 13828 }, { "epoch": 1.4992411101474414, "grad_norm": 0.152903750538826, "learning_rate": 5.020673146670536e-05, "loss": 0.014, "step": 13829 }, { "epoch": 1.4993495229835212, "grad_norm": 0.3986562490463257, "learning_rate": 5.020310459886842e-05, "loss": 0.0122, "step": 13830 }, { "epoch": 1.499457935819601, "grad_norm": 0.21684980392456055, "learning_rate": 5.0199477731031483e-05, "loss": 0.0165, "step": 13831 }, { "epoch": 1.499566348655681, "grad_norm": 0.3014800250530243, "learning_rate": 5.019585086319455e-05, "loss": 0.0272, "step": 13832 }, { "epoch": 1.4996747614917605, "grad_norm": 0.04730591922998428, "learning_rate": 5.019222399535761e-05, "loss": 0.0049, "step": 13833 }, { "epoch": 1.4997831743278405, "grad_norm": 0.28790774941444397, "learning_rate": 5.0188597127520676e-05, "loss": 0.0099, "step": 13834 }, { "epoch": 1.4998915871639202, "grad_norm": 0.09475813060998917, "learning_rate": 5.018497025968374e-05, "loss": 0.0043, "step": 13835 }, { "epoch": 1.5, "grad_norm": 0.17108304798603058, "learning_rate": 5.0181343391846804e-05, "loss": 0.0109, "step": 13836 }, { "epoch": 1.5001084128360798, "grad_norm": 0.35685721039772034, "learning_rate": 5.017771652400987e-05, "loss": 0.0193, "step": 13837 }, { "epoch": 1.5002168256721595, "grad_norm": 0.23265644907951355, "learning_rate": 5.017408965617293e-05, "loss": 0.0221, "step": 13838 }, { "epoch": 1.5003252385082395, "grad_norm": 0.330231636762619, "learning_rate": 5.017046278833599e-05, "loss": 0.0434, "step": 13839 }, { "epoch": 1.500433651344319, "grad_norm": 0.395175039768219, "learning_rate": 5.0166835920499054e-05, "loss": 0.0229, "step": 13840 }, { "epoch": 1.500542064180399, "grad_norm": 0.31651410460472107, "learning_rate": 5.016320905266212e-05, "loss": 0.0216, "step": 13841 }, { "epoch": 1.5006504770164788, "grad_norm": 0.18921874463558197, "learning_rate": 5.015958218482518e-05, "loss": 0.0075, "step": 13842 }, { "epoch": 1.5007588898525586, "grad_norm": 0.4420757591724396, "learning_rate": 5.015595531698826e-05, "loss": 0.0293, "step": 13843 }, { "epoch": 1.5008673026886383, "grad_norm": 0.4871675968170166, "learning_rate": 5.015232844915132e-05, "loss": 0.0229, "step": 13844 }, { "epoch": 1.500975715524718, "grad_norm": 0.39404696226119995, "learning_rate": 5.014870158131438e-05, "loss": 0.0318, "step": 13845 }, { "epoch": 1.501084128360798, "grad_norm": 0.1911759078502655, "learning_rate": 5.0145074713477445e-05, "loss": 0.0204, "step": 13846 }, { "epoch": 1.5011925411968776, "grad_norm": 0.37953028082847595, "learning_rate": 5.014144784564051e-05, "loss": 0.009, "step": 13847 }, { "epoch": 1.5013009540329576, "grad_norm": 0.4876224994659424, "learning_rate": 5.013782097780357e-05, "loss": 0.0389, "step": 13848 }, { "epoch": 1.5014093668690371, "grad_norm": 0.11045859754085541, "learning_rate": 5.013419410996664e-05, "loss": 0.0055, "step": 13849 }, { "epoch": 1.5015177797051171, "grad_norm": 0.5958200693130493, "learning_rate": 5.01305672421297e-05, "loss": 0.0751, "step": 13850 }, { "epoch": 1.501626192541197, "grad_norm": 0.40385934710502625, "learning_rate": 5.0126940374292765e-05, "loss": 0.0261, "step": 13851 }, { "epoch": 1.5017346053772767, "grad_norm": 0.21852074563503265, "learning_rate": 5.012331350645583e-05, "loss": 0.0075, "step": 13852 }, { "epoch": 1.5018430182133564, "grad_norm": 0.20792293548583984, "learning_rate": 5.011968663861889e-05, "loss": 0.0158, "step": 13853 }, { "epoch": 1.5019514310494362, "grad_norm": 0.6344232559204102, "learning_rate": 5.011605977078195e-05, "loss": 0.0348, "step": 13854 }, { "epoch": 1.5020598438855162, "grad_norm": 0.3593652844429016, "learning_rate": 5.0112432902945015e-05, "loss": 0.0274, "step": 13855 }, { "epoch": 1.5021682567215957, "grad_norm": 0.7457698583602905, "learning_rate": 5.010880603510808e-05, "loss": 0.0634, "step": 13856 }, { "epoch": 1.5022766695576757, "grad_norm": 0.16225439310073853, "learning_rate": 5.010517916727114e-05, "loss": 0.0038, "step": 13857 }, { "epoch": 1.5023850823937555, "grad_norm": 0.2240981012582779, "learning_rate": 5.010155229943421e-05, "loss": 0.0159, "step": 13858 }, { "epoch": 1.5024934952298352, "grad_norm": 0.6583670973777771, "learning_rate": 5.009792543159727e-05, "loss": 0.0561, "step": 13859 }, { "epoch": 1.502601908065915, "grad_norm": 0.3600374162197113, "learning_rate": 5.0094298563760336e-05, "loss": 0.0324, "step": 13860 }, { "epoch": 1.5027103209019947, "grad_norm": 0.018659085035324097, "learning_rate": 5.0090671695923407e-05, "loss": 0.0009, "step": 13861 }, { "epoch": 1.5028187337380747, "grad_norm": 0.44420453906059265, "learning_rate": 5.008704482808647e-05, "loss": 0.0419, "step": 13862 }, { "epoch": 1.5029271465741543, "grad_norm": 0.5265226364135742, "learning_rate": 5.0083417960249535e-05, "loss": 0.0297, "step": 13863 }, { "epoch": 1.5030355594102343, "grad_norm": 0.40797159075737, "learning_rate": 5.00797910924126e-05, "loss": 0.0128, "step": 13864 }, { "epoch": 1.5031439722463138, "grad_norm": 0.20119857788085938, "learning_rate": 5.007616422457566e-05, "loss": 0.0224, "step": 13865 }, { "epoch": 1.5032523850823938, "grad_norm": 0.4168873131275177, "learning_rate": 5.007253735673873e-05, "loss": 0.0246, "step": 13866 }, { "epoch": 1.5033607979184735, "grad_norm": 0.14842262864112854, "learning_rate": 5.006891048890179e-05, "loss": 0.0146, "step": 13867 }, { "epoch": 1.5034692107545533, "grad_norm": 0.1548764705657959, "learning_rate": 5.006528362106485e-05, "loss": 0.026, "step": 13868 }, { "epoch": 1.5035776235906333, "grad_norm": 0.5443940162658691, "learning_rate": 5.006165675322791e-05, "loss": 0.0109, "step": 13869 }, { "epoch": 1.5036860364267128, "grad_norm": 0.16285710036754608, "learning_rate": 5.005802988539098e-05, "loss": 0.0141, "step": 13870 }, { "epoch": 1.5037944492627928, "grad_norm": 0.2416331171989441, "learning_rate": 5.005440301755404e-05, "loss": 0.0142, "step": 13871 }, { "epoch": 1.5039028620988724, "grad_norm": 0.08491260558366776, "learning_rate": 5.0050776149717105e-05, "loss": 0.0047, "step": 13872 }, { "epoch": 1.5040112749349523, "grad_norm": 0.30637750029563904, "learning_rate": 5.004714928188017e-05, "loss": 0.0138, "step": 13873 }, { "epoch": 1.504119687771032, "grad_norm": 0.34583356976509094, "learning_rate": 5.004352241404323e-05, "loss": 0.0219, "step": 13874 }, { "epoch": 1.5042281006071119, "grad_norm": 0.09773733466863632, "learning_rate": 5.00398955462063e-05, "loss": 0.0095, "step": 13875 }, { "epoch": 1.5043365134431916, "grad_norm": 0.19883711636066437, "learning_rate": 5.0036268678369355e-05, "loss": 0.0108, "step": 13876 }, { "epoch": 1.5044449262792714, "grad_norm": 0.3132961690425873, "learning_rate": 5.003264181053242e-05, "loss": 0.0188, "step": 13877 }, { "epoch": 1.5045533391153514, "grad_norm": 0.5108590126037598, "learning_rate": 5.0029014942695496e-05, "loss": 0.0445, "step": 13878 }, { "epoch": 1.504661751951431, "grad_norm": 0.23205901682376862, "learning_rate": 5.002538807485856e-05, "loss": 0.0147, "step": 13879 }, { "epoch": 1.504770164787511, "grad_norm": 0.4014444947242737, "learning_rate": 5.0021761207021625e-05, "loss": 0.0305, "step": 13880 }, { "epoch": 1.5048785776235907, "grad_norm": 0.17086108028888702, "learning_rate": 5.001813433918469e-05, "loss": 0.0189, "step": 13881 }, { "epoch": 1.5049869904596704, "grad_norm": 0.3296429514884949, "learning_rate": 5.0014507471347746e-05, "loss": 0.0494, "step": 13882 }, { "epoch": 1.5050954032957502, "grad_norm": 0.1134624257683754, "learning_rate": 5.001088060351081e-05, "loss": 0.014, "step": 13883 }, { "epoch": 1.50520381613183, "grad_norm": 0.5436873435974121, "learning_rate": 5.0007253735673874e-05, "loss": 0.0367, "step": 13884 }, { "epoch": 1.50531222896791, "grad_norm": 0.06530232727527618, "learning_rate": 5.000362686783694e-05, "loss": 0.0035, "step": 13885 }, { "epoch": 1.5054206418039895, "grad_norm": 0.09163045883178711, "learning_rate": 5e-05, "loss": 0.0074, "step": 13886 }, { "epoch": 1.5055290546400695, "grad_norm": 0.4265252649784088, "learning_rate": 4.9996373132163067e-05, "loss": 0.0276, "step": 13887 }, { "epoch": 1.505637467476149, "grad_norm": 0.09736642986536026, "learning_rate": 4.999274626432613e-05, "loss": 0.0095, "step": 13888 }, { "epoch": 1.505745880312229, "grad_norm": 0.428963840007782, "learning_rate": 4.9989119396489195e-05, "loss": 0.041, "step": 13889 }, { "epoch": 1.5058542931483088, "grad_norm": 0.5397859215736389, "learning_rate": 4.998549252865226e-05, "loss": 0.0422, "step": 13890 }, { "epoch": 1.5059627059843885, "grad_norm": 0.1275709569454193, "learning_rate": 4.998186566081532e-05, "loss": 0.0052, "step": 13891 }, { "epoch": 1.5060711188204683, "grad_norm": 0.7456970810890198, "learning_rate": 4.997823879297839e-05, "loss": 0.0491, "step": 13892 }, { "epoch": 1.506179531656548, "grad_norm": 0.11573462933301926, "learning_rate": 4.997461192514145e-05, "loss": 0.0056, "step": 13893 }, { "epoch": 1.506287944492628, "grad_norm": 0.2881580889225006, "learning_rate": 4.9970985057304515e-05, "loss": 0.0444, "step": 13894 }, { "epoch": 1.5063963573287076, "grad_norm": 0.19624453783035278, "learning_rate": 4.996735818946758e-05, "loss": 0.0136, "step": 13895 }, { "epoch": 1.5065047701647876, "grad_norm": 0.262452095746994, "learning_rate": 4.9963731321630643e-05, "loss": 0.0077, "step": 13896 }, { "epoch": 1.5066131830008673, "grad_norm": 0.41081416606903076, "learning_rate": 4.996010445379371e-05, "loss": 0.0484, "step": 13897 }, { "epoch": 1.506721595836947, "grad_norm": 0.27078384160995483, "learning_rate": 4.9956477585956765e-05, "loss": 0.0209, "step": 13898 }, { "epoch": 1.5068300086730269, "grad_norm": 0.5455141663551331, "learning_rate": 4.9952850718119836e-05, "loss": 0.0244, "step": 13899 }, { "epoch": 1.5069384215091066, "grad_norm": 0.32829537987709045, "learning_rate": 4.99492238502829e-05, "loss": 0.0447, "step": 13900 }, { "epoch": 1.5070468343451866, "grad_norm": 0.7776123285293579, "learning_rate": 4.9945596982445964e-05, "loss": 0.0293, "step": 13901 }, { "epoch": 1.5071552471812661, "grad_norm": 0.2904245853424072, "learning_rate": 4.994197011460903e-05, "loss": 0.0228, "step": 13902 }, { "epoch": 1.5072636600173461, "grad_norm": 0.13094310462474823, "learning_rate": 4.993834324677209e-05, "loss": 0.0094, "step": 13903 }, { "epoch": 1.507372072853426, "grad_norm": 0.3097854256629944, "learning_rate": 4.9934716378935156e-05, "loss": 0.0247, "step": 13904 }, { "epoch": 1.5074804856895057, "grad_norm": 0.2770366370677948, "learning_rate": 4.9931089511098214e-05, "loss": 0.0082, "step": 13905 }, { "epoch": 1.5075888985255854, "grad_norm": 0.13841938972473145, "learning_rate": 4.992746264326128e-05, "loss": 0.0091, "step": 13906 }, { "epoch": 1.5076973113616652, "grad_norm": 0.49993786215782166, "learning_rate": 4.992383577542434e-05, "loss": 0.0556, "step": 13907 }, { "epoch": 1.5078057241977452, "grad_norm": 0.17395126819610596, "learning_rate": 4.992020890758741e-05, "loss": 0.0129, "step": 13908 }, { "epoch": 1.5079141370338247, "grad_norm": 0.3064793348312378, "learning_rate": 4.991658203975048e-05, "loss": 0.0142, "step": 13909 }, { "epoch": 1.5080225498699047, "grad_norm": 0.15257160365581512, "learning_rate": 4.991295517191354e-05, "loss": 0.0122, "step": 13910 }, { "epoch": 1.5081309627059842, "grad_norm": 0.03886968269944191, "learning_rate": 4.9909328304076605e-05, "loss": 0.0044, "step": 13911 }, { "epoch": 1.5082393755420642, "grad_norm": 0.3659505546092987, "learning_rate": 4.990570143623967e-05, "loss": 0.0232, "step": 13912 }, { "epoch": 1.508347788378144, "grad_norm": 0.42105600237846375, "learning_rate": 4.9902074568402726e-05, "loss": 0.0562, "step": 13913 }, { "epoch": 1.5084562012142237, "grad_norm": 0.17761379480361938, "learning_rate": 4.989844770056579e-05, "loss": 0.0041, "step": 13914 }, { "epoch": 1.5085646140503035, "grad_norm": 0.39699792861938477, "learning_rate": 4.9894820832728855e-05, "loss": 0.026, "step": 13915 }, { "epoch": 1.5086730268863833, "grad_norm": 0.22854265570640564, "learning_rate": 4.989119396489192e-05, "loss": 0.0232, "step": 13916 }, { "epoch": 1.5087814397224633, "grad_norm": 0.42977458238601685, "learning_rate": 4.988756709705499e-05, "loss": 0.016, "step": 13917 }, { "epoch": 1.5088898525585428, "grad_norm": 0.16778546571731567, "learning_rate": 4.9883940229218054e-05, "loss": 0.0197, "step": 13918 }, { "epoch": 1.5089982653946228, "grad_norm": 0.35684099793434143, "learning_rate": 4.988031336138112e-05, "loss": 0.0347, "step": 13919 }, { "epoch": 1.5091066782307025, "grad_norm": 0.30408352613449097, "learning_rate": 4.9876686493544175e-05, "loss": 0.0409, "step": 13920 }, { "epoch": 1.5092150910667823, "grad_norm": 0.31330201029777527, "learning_rate": 4.987305962570724e-05, "loss": 0.0185, "step": 13921 }, { "epoch": 1.509323503902862, "grad_norm": 0.9455262422561646, "learning_rate": 4.9869432757870303e-05, "loss": 0.029, "step": 13922 }, { "epoch": 1.5094319167389418, "grad_norm": 0.30188778042793274, "learning_rate": 4.986580589003337e-05, "loss": 0.0207, "step": 13923 }, { "epoch": 1.5095403295750218, "grad_norm": 0.8038282990455627, "learning_rate": 4.986217902219643e-05, "loss": 0.032, "step": 13924 }, { "epoch": 1.5096487424111014, "grad_norm": 0.22189047932624817, "learning_rate": 4.98585521543595e-05, "loss": 0.0274, "step": 13925 }, { "epoch": 1.5097571552471813, "grad_norm": 0.2879008650779724, "learning_rate": 4.9854925286522567e-05, "loss": 0.0077, "step": 13926 }, { "epoch": 1.5098655680832611, "grad_norm": 0.26074638962745667, "learning_rate": 4.9851298418685624e-05, "loss": 0.0338, "step": 13927 }, { "epoch": 1.5099739809193409, "grad_norm": 0.16140034794807434, "learning_rate": 4.984767155084869e-05, "loss": 0.0089, "step": 13928 }, { "epoch": 1.5100823937554206, "grad_norm": 0.5258901715278625, "learning_rate": 4.984404468301175e-05, "loss": 0.048, "step": 13929 }, { "epoch": 1.5101908065915004, "grad_norm": 0.6738485097885132, "learning_rate": 4.9840417815174816e-05, "loss": 0.0118, "step": 13930 }, { "epoch": 1.5102992194275804, "grad_norm": 0.4093397259712219, "learning_rate": 4.983679094733788e-05, "loss": 0.0604, "step": 13931 }, { "epoch": 1.51040763226366, "grad_norm": 0.19972185790538788, "learning_rate": 4.9833164079500944e-05, "loss": 0.0091, "step": 13932 }, { "epoch": 1.51051604509974, "grad_norm": 0.42497578263282776, "learning_rate": 4.982953721166401e-05, "loss": 0.0152, "step": 13933 }, { "epoch": 1.5106244579358195, "grad_norm": 0.07986734807491302, "learning_rate": 4.982591034382707e-05, "loss": 0.0054, "step": 13934 }, { "epoch": 1.5107328707718994, "grad_norm": 0.2470414638519287, "learning_rate": 4.982228347599014e-05, "loss": 0.0472, "step": 13935 }, { "epoch": 1.5108412836079792, "grad_norm": 0.6701948046684265, "learning_rate": 4.98186566081532e-05, "loss": 0.0206, "step": 13936 }, { "epoch": 1.510949696444059, "grad_norm": 0.27887043356895447, "learning_rate": 4.9815029740316265e-05, "loss": 0.0083, "step": 13937 }, { "epoch": 1.5110581092801387, "grad_norm": 0.4671982228755951, "learning_rate": 4.981140287247933e-05, "loss": 0.0126, "step": 13938 }, { "epoch": 1.5111665221162185, "grad_norm": 0.19708774983882904, "learning_rate": 4.980777600464239e-05, "loss": 0.0192, "step": 13939 }, { "epoch": 1.5112749349522985, "grad_norm": 0.6353501677513123, "learning_rate": 4.980414913680546e-05, "loss": 0.0501, "step": 13940 }, { "epoch": 1.511383347788378, "grad_norm": 0.380591481924057, "learning_rate": 4.980052226896852e-05, "loss": 0.0212, "step": 13941 }, { "epoch": 1.511491760624458, "grad_norm": 0.04107436537742615, "learning_rate": 4.9796895401131586e-05, "loss": 0.0026, "step": 13942 }, { "epoch": 1.5116001734605378, "grad_norm": 0.4086317718029022, "learning_rate": 4.979326853329465e-05, "loss": 0.0205, "step": 13943 }, { "epoch": 1.5117085862966175, "grad_norm": 0.03166211396455765, "learning_rate": 4.9789641665457714e-05, "loss": 0.0039, "step": 13944 }, { "epoch": 1.5118169991326973, "grad_norm": 0.22154363989830017, "learning_rate": 4.978601479762078e-05, "loss": 0.0211, "step": 13945 }, { "epoch": 1.511925411968777, "grad_norm": 0.1959402859210968, "learning_rate": 4.978238792978384e-05, "loss": 0.0073, "step": 13946 }, { "epoch": 1.512033824804857, "grad_norm": 0.5425677299499512, "learning_rate": 4.9778761061946906e-05, "loss": 0.0315, "step": 13947 }, { "epoch": 1.5121422376409366, "grad_norm": 1.4149088859558105, "learning_rate": 4.977513419410997e-05, "loss": 0.0617, "step": 13948 }, { "epoch": 1.5122506504770166, "grad_norm": 0.28000274300575256, "learning_rate": 4.9771507326273034e-05, "loss": 0.0191, "step": 13949 }, { "epoch": 1.512359063313096, "grad_norm": 0.1901002675294876, "learning_rate": 4.976788045843609e-05, "loss": 0.0169, "step": 13950 }, { "epoch": 1.512467476149176, "grad_norm": 0.6041348576545715, "learning_rate": 4.976425359059916e-05, "loss": 0.0276, "step": 13951 }, { "epoch": 1.5125758889852559, "grad_norm": 0.4639595150947571, "learning_rate": 4.9760626722762227e-05, "loss": 0.0506, "step": 13952 }, { "epoch": 1.5126843018213356, "grad_norm": 0.15173454582691193, "learning_rate": 4.975699985492529e-05, "loss": 0.0159, "step": 13953 }, { "epoch": 1.5127927146574154, "grad_norm": 0.30972662568092346, "learning_rate": 4.9753372987088355e-05, "loss": 0.0306, "step": 13954 }, { "epoch": 1.5129011274934951, "grad_norm": 0.3614826500415802, "learning_rate": 4.974974611925142e-05, "loss": 0.0361, "step": 13955 }, { "epoch": 1.5130095403295751, "grad_norm": 0.22491849958896637, "learning_rate": 4.974611925141448e-05, "loss": 0.0106, "step": 13956 }, { "epoch": 1.5131179531656547, "grad_norm": 0.22189390659332275, "learning_rate": 4.974249238357754e-05, "loss": 0.0172, "step": 13957 }, { "epoch": 1.5132263660017347, "grad_norm": 0.37280815839767456, "learning_rate": 4.9738865515740604e-05, "loss": 0.0217, "step": 13958 }, { "epoch": 1.5133347788378144, "grad_norm": 0.3869262933731079, "learning_rate": 4.973523864790367e-05, "loss": 0.0248, "step": 13959 }, { "epoch": 1.5134431916738942, "grad_norm": 0.45162609219551086, "learning_rate": 4.973161178006674e-05, "loss": 0.0479, "step": 13960 }, { "epoch": 1.513551604509974, "grad_norm": 0.22839833796024323, "learning_rate": 4.9727984912229804e-05, "loss": 0.0328, "step": 13961 }, { "epoch": 1.5136600173460537, "grad_norm": 0.2717278003692627, "learning_rate": 4.972435804439287e-05, "loss": 0.0398, "step": 13962 }, { "epoch": 1.5137684301821337, "grad_norm": 0.1636069118976593, "learning_rate": 4.972073117655593e-05, "loss": 0.0224, "step": 13963 }, { "epoch": 1.5138768430182132, "grad_norm": 0.20200811326503754, "learning_rate": 4.9717104308718996e-05, "loss": 0.0253, "step": 13964 }, { "epoch": 1.5139852558542932, "grad_norm": 0.33594220876693726, "learning_rate": 4.971347744088205e-05, "loss": 0.0251, "step": 13965 }, { "epoch": 1.514093668690373, "grad_norm": 0.18167753517627716, "learning_rate": 4.970985057304512e-05, "loss": 0.0139, "step": 13966 }, { "epoch": 1.5142020815264527, "grad_norm": 0.021649500355124474, "learning_rate": 4.970622370520818e-05, "loss": 0.0015, "step": 13967 }, { "epoch": 1.5143104943625325, "grad_norm": 0.1128336489200592, "learning_rate": 4.9702596837371245e-05, "loss": 0.0068, "step": 13968 }, { "epoch": 1.5144189071986123, "grad_norm": 0.24975839257240295, "learning_rate": 4.9698969969534316e-05, "loss": 0.0115, "step": 13969 }, { "epoch": 1.5145273200346923, "grad_norm": 0.06872597336769104, "learning_rate": 4.969534310169738e-05, "loss": 0.0038, "step": 13970 }, { "epoch": 1.5146357328707718, "grad_norm": 0.30207717418670654, "learning_rate": 4.9691716233860445e-05, "loss": 0.0475, "step": 13971 }, { "epoch": 1.5147441457068518, "grad_norm": 0.3834053874015808, "learning_rate": 4.96880893660235e-05, "loss": 0.02, "step": 13972 }, { "epoch": 1.5148525585429313, "grad_norm": 0.22775408625602722, "learning_rate": 4.9684462498186566e-05, "loss": 0.0097, "step": 13973 }, { "epoch": 1.5149609713790113, "grad_norm": 0.3415988087654114, "learning_rate": 4.968083563034963e-05, "loss": 0.0384, "step": 13974 }, { "epoch": 1.515069384215091, "grad_norm": 0.388799250125885, "learning_rate": 4.9677208762512694e-05, "loss": 0.0327, "step": 13975 }, { "epoch": 1.5151777970511708, "grad_norm": 0.2097911834716797, "learning_rate": 4.967358189467576e-05, "loss": 0.0157, "step": 13976 }, { "epoch": 1.5152862098872506, "grad_norm": 1.019931435585022, "learning_rate": 4.966995502683883e-05, "loss": 0.0498, "step": 13977 }, { "epoch": 1.5153946227233304, "grad_norm": 0.22880665957927704, "learning_rate": 4.966632815900189e-05, "loss": 0.0125, "step": 13978 }, { "epoch": 1.5155030355594103, "grad_norm": 0.29409947991371155, "learning_rate": 4.966270129116495e-05, "loss": 0.025, "step": 13979 }, { "epoch": 1.51561144839549, "grad_norm": 0.256590336561203, "learning_rate": 4.9659074423328015e-05, "loss": 0.0301, "step": 13980 }, { "epoch": 1.5157198612315699, "grad_norm": 0.19483301043510437, "learning_rate": 4.965544755549108e-05, "loss": 0.0104, "step": 13981 }, { "epoch": 1.5158282740676496, "grad_norm": 0.06763537228107452, "learning_rate": 4.965182068765414e-05, "loss": 0.0056, "step": 13982 }, { "epoch": 1.5159366869037294, "grad_norm": 0.3402501940727234, "learning_rate": 4.964819381981721e-05, "loss": 0.0193, "step": 13983 }, { "epoch": 1.5160450997398092, "grad_norm": 0.6994206309318542, "learning_rate": 4.964456695198027e-05, "loss": 0.0402, "step": 13984 }, { "epoch": 1.516153512575889, "grad_norm": 0.34900933504104614, "learning_rate": 4.9640940084143335e-05, "loss": 0.0394, "step": 13985 }, { "epoch": 1.516261925411969, "grad_norm": 0.6958479285240173, "learning_rate": 4.96373132163064e-05, "loss": 0.0288, "step": 13986 }, { "epoch": 1.5163703382480485, "grad_norm": 0.6638070940971375, "learning_rate": 4.9633686348469463e-05, "loss": 0.1111, "step": 13987 }, { "epoch": 1.5164787510841284, "grad_norm": 0.2504056692123413, "learning_rate": 4.963005948063253e-05, "loss": 0.0174, "step": 13988 }, { "epoch": 1.5165871639202082, "grad_norm": 0.14988908171653748, "learning_rate": 4.962643261279559e-05, "loss": 0.0086, "step": 13989 }, { "epoch": 1.516695576756288, "grad_norm": 0.17920374870300293, "learning_rate": 4.9622805744958656e-05, "loss": 0.0139, "step": 13990 }, { "epoch": 1.5168039895923677, "grad_norm": 0.14229097962379456, "learning_rate": 4.961917887712172e-05, "loss": 0.0129, "step": 13991 }, { "epoch": 1.5169124024284475, "grad_norm": 0.16664841771125793, "learning_rate": 4.9615552009284784e-05, "loss": 0.0152, "step": 13992 }, { "epoch": 1.5170208152645275, "grad_norm": 0.226501926779747, "learning_rate": 4.961192514144785e-05, "loss": 0.0198, "step": 13993 }, { "epoch": 1.517129228100607, "grad_norm": 0.39840832352638245, "learning_rate": 4.960829827361091e-05, "loss": 0.0382, "step": 13994 }, { "epoch": 1.517237640936687, "grad_norm": 0.37838342785835266, "learning_rate": 4.9604671405773976e-05, "loss": 0.027, "step": 13995 }, { "epoch": 1.5173460537727665, "grad_norm": 0.21685896813869476, "learning_rate": 4.960104453793704e-05, "loss": 0.0289, "step": 13996 }, { "epoch": 1.5174544666088465, "grad_norm": 0.471457839012146, "learning_rate": 4.9597417670100105e-05, "loss": 0.0413, "step": 13997 }, { "epoch": 1.5175628794449263, "grad_norm": 0.18281739950180054, "learning_rate": 4.959379080226317e-05, "loss": 0.0119, "step": 13998 }, { "epoch": 1.517671292281006, "grad_norm": 0.32283878326416016, "learning_rate": 4.959016393442623e-05, "loss": 0.007, "step": 13999 }, { "epoch": 1.5177797051170858, "grad_norm": 0.1890493631362915, "learning_rate": 4.95865370665893e-05, "loss": 0.0203, "step": 14000 }, { "epoch": 1.5178881179531656, "grad_norm": 0.2730661630630493, "learning_rate": 4.958291019875236e-05, "loss": 0.0118, "step": 14001 }, { "epoch": 1.5179965307892456, "grad_norm": 0.4369223117828369, "learning_rate": 4.957928333091542e-05, "loss": 0.0546, "step": 14002 }, { "epoch": 1.518104943625325, "grad_norm": 0.14614960551261902, "learning_rate": 4.957565646307849e-05, "loss": 0.0217, "step": 14003 }, { "epoch": 1.518213356461405, "grad_norm": 0.2822895050048828, "learning_rate": 4.957202959524155e-05, "loss": 0.0222, "step": 14004 }, { "epoch": 1.5183217692974849, "grad_norm": 0.1516730785369873, "learning_rate": 4.956840272740462e-05, "loss": 0.0122, "step": 14005 }, { "epoch": 1.5184301821335646, "grad_norm": 0.20401258766651154, "learning_rate": 4.956477585956768e-05, "loss": 0.0104, "step": 14006 }, { "epoch": 1.5185385949696444, "grad_norm": 0.08059319108724594, "learning_rate": 4.9561148991730746e-05, "loss": 0.0083, "step": 14007 }, { "epoch": 1.5186470078057241, "grad_norm": 0.2043047994375229, "learning_rate": 4.955752212389381e-05, "loss": 0.0215, "step": 14008 }, { "epoch": 1.5187554206418041, "grad_norm": 0.264932245016098, "learning_rate": 4.955389525605687e-05, "loss": 0.0167, "step": 14009 }, { "epoch": 1.5188638334778837, "grad_norm": 0.23543579876422882, "learning_rate": 4.955026838821993e-05, "loss": 0.0112, "step": 14010 }, { "epoch": 1.5189722463139637, "grad_norm": 0.09160511940717697, "learning_rate": 4.9546641520382995e-05, "loss": 0.0094, "step": 14011 }, { "epoch": 1.5190806591500432, "grad_norm": 0.046236731112003326, "learning_rate": 4.9543014652546066e-05, "loss": 0.0051, "step": 14012 }, { "epoch": 1.5191890719861232, "grad_norm": 0.12327314913272858, "learning_rate": 4.953938778470913e-05, "loss": 0.0074, "step": 14013 }, { "epoch": 1.519297484822203, "grad_norm": 0.8695070743560791, "learning_rate": 4.9535760916872194e-05, "loss": 0.0853, "step": 14014 }, { "epoch": 1.5194058976582827, "grad_norm": 0.08513779938220978, "learning_rate": 4.953213404903526e-05, "loss": 0.0053, "step": 14015 }, { "epoch": 1.5195143104943627, "grad_norm": 0.32130953669548035, "learning_rate": 4.952850718119832e-05, "loss": 0.0379, "step": 14016 }, { "epoch": 1.5196227233304422, "grad_norm": 0.06349032372236252, "learning_rate": 4.952488031336138e-05, "loss": 0.0067, "step": 14017 }, { "epoch": 1.5197311361665222, "grad_norm": 0.2872560918331146, "learning_rate": 4.9521253445524444e-05, "loss": 0.0177, "step": 14018 }, { "epoch": 1.5198395490026018, "grad_norm": 0.2001810520887375, "learning_rate": 4.951762657768751e-05, "loss": 0.0193, "step": 14019 }, { "epoch": 1.5199479618386817, "grad_norm": 0.2886161804199219, "learning_rate": 4.951399970985058e-05, "loss": 0.0309, "step": 14020 }, { "epoch": 1.5200563746747615, "grad_norm": 0.41221052408218384, "learning_rate": 4.951037284201364e-05, "loss": 0.0216, "step": 14021 }, { "epoch": 1.5201647875108413, "grad_norm": 0.443511962890625, "learning_rate": 4.950674597417671e-05, "loss": 0.0305, "step": 14022 }, { "epoch": 1.520273200346921, "grad_norm": 0.31209611892700195, "learning_rate": 4.950311910633977e-05, "loss": 0.0184, "step": 14023 }, { "epoch": 1.5203816131830008, "grad_norm": 0.3620755076408386, "learning_rate": 4.949949223850283e-05, "loss": 0.0193, "step": 14024 }, { "epoch": 1.5204900260190808, "grad_norm": 0.2452024668455124, "learning_rate": 4.949586537066589e-05, "loss": 0.0313, "step": 14025 }, { "epoch": 1.5205984388551603, "grad_norm": 0.18827863037586212, "learning_rate": 4.949223850282896e-05, "loss": 0.0107, "step": 14026 }, { "epoch": 1.5207068516912403, "grad_norm": 0.5085679292678833, "learning_rate": 4.948861163499202e-05, "loss": 0.0589, "step": 14027 }, { "epoch": 1.52081526452732, "grad_norm": 0.236226886510849, "learning_rate": 4.9484984767155085e-05, "loss": 0.0168, "step": 14028 }, { "epoch": 1.5209236773633998, "grad_norm": 0.31690850853919983, "learning_rate": 4.9481357899318156e-05, "loss": 0.0062, "step": 14029 }, { "epoch": 1.5210320901994796, "grad_norm": 0.3700115978717804, "learning_rate": 4.947773103148122e-05, "loss": 0.0171, "step": 14030 }, { "epoch": 1.5211405030355594, "grad_norm": 0.26806187629699707, "learning_rate": 4.947410416364428e-05, "loss": 0.0406, "step": 14031 }, { "epoch": 1.5212489158716394, "grad_norm": 0.25289320945739746, "learning_rate": 4.947047729580734e-05, "loss": 0.0131, "step": 14032 }, { "epoch": 1.521357328707719, "grad_norm": 0.3827461302280426, "learning_rate": 4.9466850427970406e-05, "loss": 0.0282, "step": 14033 }, { "epoch": 1.5214657415437989, "grad_norm": 0.4630785286426544, "learning_rate": 4.946322356013347e-05, "loss": 0.046, "step": 14034 }, { "epoch": 1.5215741543798784, "grad_norm": 0.14243754744529724, "learning_rate": 4.9459596692296534e-05, "loss": 0.0065, "step": 14035 }, { "epoch": 1.5216825672159584, "grad_norm": 0.9307856559753418, "learning_rate": 4.94559698244596e-05, "loss": 0.018, "step": 14036 }, { "epoch": 1.5217909800520382, "grad_norm": 0.5451241135597229, "learning_rate": 4.945234295662266e-05, "loss": 0.0299, "step": 14037 }, { "epoch": 1.521899392888118, "grad_norm": 0.3275425136089325, "learning_rate": 4.9448716088785726e-05, "loss": 0.0117, "step": 14038 }, { "epoch": 1.5220078057241977, "grad_norm": 0.2884954512119293, "learning_rate": 4.944508922094879e-05, "loss": 0.0136, "step": 14039 }, { "epoch": 1.5221162185602775, "grad_norm": 0.2309017777442932, "learning_rate": 4.9441462353111854e-05, "loss": 0.0103, "step": 14040 }, { "epoch": 1.5222246313963574, "grad_norm": 0.3231171667575836, "learning_rate": 4.943783548527492e-05, "loss": 0.0424, "step": 14041 }, { "epoch": 1.522333044232437, "grad_norm": 0.3799971342086792, "learning_rate": 4.943420861743798e-05, "loss": 0.0284, "step": 14042 }, { "epoch": 1.522441457068517, "grad_norm": 0.3173949420452118, "learning_rate": 4.9430581749601047e-05, "loss": 0.0065, "step": 14043 }, { "epoch": 1.5225498699045967, "grad_norm": 0.32321426272392273, "learning_rate": 4.942695488176411e-05, "loss": 0.0224, "step": 14044 }, { "epoch": 1.5226582827406765, "grad_norm": 0.17725342512130737, "learning_rate": 4.9423328013927175e-05, "loss": 0.005, "step": 14045 }, { "epoch": 1.5227666955767563, "grad_norm": 0.9736925959587097, "learning_rate": 4.941970114609024e-05, "loss": 0.0595, "step": 14046 }, { "epoch": 1.522875108412836, "grad_norm": 0.3579493463039398, "learning_rate": 4.94160742782533e-05, "loss": 0.0288, "step": 14047 }, { "epoch": 1.522983521248916, "grad_norm": 0.26284077763557434, "learning_rate": 4.941244741041637e-05, "loss": 0.0339, "step": 14048 }, { "epoch": 1.5230919340849955, "grad_norm": 0.23750172555446625, "learning_rate": 4.940882054257943e-05, "loss": 0.0086, "step": 14049 }, { "epoch": 1.5232003469210755, "grad_norm": 0.39522773027420044, "learning_rate": 4.9405193674742495e-05, "loss": 0.0081, "step": 14050 }, { "epoch": 1.5233087597571553, "grad_norm": 0.01725705899298191, "learning_rate": 4.940156680690556e-05, "loss": 0.001, "step": 14051 }, { "epoch": 1.523417172593235, "grad_norm": 0.22897031903266907, "learning_rate": 4.9397939939068624e-05, "loss": 0.0211, "step": 14052 }, { "epoch": 1.5235255854293148, "grad_norm": 0.1847984343767166, "learning_rate": 4.939431307123169e-05, "loss": 0.0293, "step": 14053 }, { "epoch": 1.5236339982653946, "grad_norm": 0.09965138137340546, "learning_rate": 4.9390686203394745e-05, "loss": 0.0051, "step": 14054 }, { "epoch": 1.5237424111014746, "grad_norm": 0.11015787720680237, "learning_rate": 4.9387059335557816e-05, "loss": 0.0103, "step": 14055 }, { "epoch": 1.523850823937554, "grad_norm": 0.16221283376216888, "learning_rate": 4.938343246772088e-05, "loss": 0.0117, "step": 14056 }, { "epoch": 1.523959236773634, "grad_norm": 0.3768889307975769, "learning_rate": 4.9379805599883944e-05, "loss": 0.0243, "step": 14057 }, { "epoch": 1.5240676496097136, "grad_norm": 0.5808925628662109, "learning_rate": 4.937617873204701e-05, "loss": 0.0295, "step": 14058 }, { "epoch": 1.5241760624457936, "grad_norm": 0.3042442202568054, "learning_rate": 4.937255186421007e-05, "loss": 0.0133, "step": 14059 }, { "epoch": 1.5242844752818734, "grad_norm": 0.38197702169418335, "learning_rate": 4.9368924996373136e-05, "loss": 0.0146, "step": 14060 }, { "epoch": 1.5243928881179531, "grad_norm": 0.5260334014892578, "learning_rate": 4.9365298128536194e-05, "loss": 0.0197, "step": 14061 }, { "epoch": 1.524501300954033, "grad_norm": 0.1325462907552719, "learning_rate": 4.936167126069926e-05, "loss": 0.0102, "step": 14062 }, { "epoch": 1.5246097137901127, "grad_norm": 0.5338312387466431, "learning_rate": 4.935804439286232e-05, "loss": 0.0329, "step": 14063 }, { "epoch": 1.5247181266261927, "grad_norm": 0.30924031138420105, "learning_rate": 4.935441752502539e-05, "loss": 0.0127, "step": 14064 }, { "epoch": 1.5248265394622722, "grad_norm": 0.3744889497756958, "learning_rate": 4.935079065718846e-05, "loss": 0.0189, "step": 14065 }, { "epoch": 1.5249349522983522, "grad_norm": 0.3189174234867096, "learning_rate": 4.934716378935152e-05, "loss": 0.0316, "step": 14066 }, { "epoch": 1.525043365134432, "grad_norm": 0.33273544907569885, "learning_rate": 4.9343536921514585e-05, "loss": 0.016, "step": 14067 }, { "epoch": 1.5251517779705117, "grad_norm": 0.1631600558757782, "learning_rate": 4.933991005367765e-05, "loss": 0.0086, "step": 14068 }, { "epoch": 1.5252601908065915, "grad_norm": 0.17029248178005219, "learning_rate": 4.9336283185840707e-05, "loss": 0.0048, "step": 14069 }, { "epoch": 1.5253686036426712, "grad_norm": 0.17206059396266937, "learning_rate": 4.933265631800377e-05, "loss": 0.0153, "step": 14070 }, { "epoch": 1.5254770164787512, "grad_norm": 0.016549596562981606, "learning_rate": 4.9329029450166835e-05, "loss": 0.001, "step": 14071 }, { "epoch": 1.5255854293148308, "grad_norm": 0.1824081540107727, "learning_rate": 4.9325402582329906e-05, "loss": 0.0091, "step": 14072 }, { "epoch": 1.5256938421509108, "grad_norm": 0.3518047630786896, "learning_rate": 4.932177571449297e-05, "loss": 0.0162, "step": 14073 }, { "epoch": 1.5258022549869905, "grad_norm": 0.5581148266792297, "learning_rate": 4.9318148846656034e-05, "loss": 0.0546, "step": 14074 }, { "epoch": 1.5259106678230703, "grad_norm": 0.15118078887462616, "learning_rate": 4.93145219788191e-05, "loss": 0.0084, "step": 14075 }, { "epoch": 1.52601908065915, "grad_norm": 0.7236321568489075, "learning_rate": 4.9310895110982155e-05, "loss": 0.0221, "step": 14076 }, { "epoch": 1.5261274934952298, "grad_norm": 0.13526131212711334, "learning_rate": 4.930726824314522e-05, "loss": 0.0034, "step": 14077 }, { "epoch": 1.5262359063313098, "grad_norm": 0.5924608111381531, "learning_rate": 4.9303641375308283e-05, "loss": 0.0448, "step": 14078 }, { "epoch": 1.5263443191673893, "grad_norm": 0.4205310642719269, "learning_rate": 4.930001450747135e-05, "loss": 0.0433, "step": 14079 }, { "epoch": 1.5264527320034693, "grad_norm": 0.14051082730293274, "learning_rate": 4.929638763963441e-05, "loss": 0.0098, "step": 14080 }, { "epoch": 1.5265611448395489, "grad_norm": 0.4190309941768646, "learning_rate": 4.929276077179748e-05, "loss": 0.0192, "step": 14081 }, { "epoch": 1.5266695576756288, "grad_norm": 0.26871174573898315, "learning_rate": 4.928913390396055e-05, "loss": 0.0025, "step": 14082 }, { "epoch": 1.5267779705117086, "grad_norm": 0.5325494408607483, "learning_rate": 4.9285507036123604e-05, "loss": 0.0312, "step": 14083 }, { "epoch": 1.5268863833477884, "grad_norm": 0.5190746784210205, "learning_rate": 4.928188016828667e-05, "loss": 0.0317, "step": 14084 }, { "epoch": 1.5269947961838681, "grad_norm": 0.3286462724208832, "learning_rate": 4.927825330044973e-05, "loss": 0.0103, "step": 14085 }, { "epoch": 1.527103209019948, "grad_norm": 0.23489658534526825, "learning_rate": 4.9274626432612796e-05, "loss": 0.0225, "step": 14086 }, { "epoch": 1.5272116218560279, "grad_norm": 0.32605186104774475, "learning_rate": 4.927099956477586e-05, "loss": 0.0211, "step": 14087 }, { "epoch": 1.5273200346921074, "grad_norm": 0.45353761315345764, "learning_rate": 4.9267372696938925e-05, "loss": 0.0324, "step": 14088 }, { "epoch": 1.5274284475281874, "grad_norm": 0.10188931226730347, "learning_rate": 4.9263745829101995e-05, "loss": 0.0016, "step": 14089 }, { "epoch": 1.5275368603642672, "grad_norm": 0.4455570876598358, "learning_rate": 4.926011896126505e-05, "loss": 0.0105, "step": 14090 }, { "epoch": 1.527645273200347, "grad_norm": 0.12998077273368835, "learning_rate": 4.925649209342812e-05, "loss": 0.0076, "step": 14091 }, { "epoch": 1.5277536860364267, "grad_norm": 0.34095093607902527, "learning_rate": 4.925286522559118e-05, "loss": 0.0186, "step": 14092 }, { "epoch": 1.5278620988725065, "grad_norm": 0.7619838714599609, "learning_rate": 4.9249238357754245e-05, "loss": 0.0788, "step": 14093 }, { "epoch": 1.5279705117085864, "grad_norm": 0.2525511384010315, "learning_rate": 4.924561148991731e-05, "loss": 0.0131, "step": 14094 }, { "epoch": 1.528078924544666, "grad_norm": 0.6385145783424377, "learning_rate": 4.924198462208037e-05, "loss": 0.0221, "step": 14095 }, { "epoch": 1.528187337380746, "grad_norm": 0.03892809897661209, "learning_rate": 4.923835775424344e-05, "loss": 0.0024, "step": 14096 }, { "epoch": 1.5282957502168255, "grad_norm": 0.31011107563972473, "learning_rate": 4.92347308864065e-05, "loss": 0.0227, "step": 14097 }, { "epoch": 1.5284041630529055, "grad_norm": 0.402787983417511, "learning_rate": 4.9231104018569566e-05, "loss": 0.0213, "step": 14098 }, { "epoch": 1.5285125758889853, "grad_norm": 0.32244613766670227, "learning_rate": 4.922747715073263e-05, "loss": 0.0139, "step": 14099 }, { "epoch": 1.528620988725065, "grad_norm": 0.7842459082603455, "learning_rate": 4.9223850282895694e-05, "loss": 0.0472, "step": 14100 }, { "epoch": 1.5287294015611448, "grad_norm": 0.30329784750938416, "learning_rate": 4.922022341505876e-05, "loss": 0.0357, "step": 14101 }, { "epoch": 1.5288378143972245, "grad_norm": 0.2168632298707962, "learning_rate": 4.921659654722182e-05, "loss": 0.0385, "step": 14102 }, { "epoch": 1.5289462272333045, "grad_norm": 0.44712135195732117, "learning_rate": 4.9212969679384886e-05, "loss": 0.0212, "step": 14103 }, { "epoch": 1.529054640069384, "grad_norm": 0.4004531502723694, "learning_rate": 4.920934281154795e-05, "loss": 0.059, "step": 14104 }, { "epoch": 1.529163052905464, "grad_norm": 0.7798322439193726, "learning_rate": 4.9205715943711014e-05, "loss": 0.0706, "step": 14105 }, { "epoch": 1.5292714657415438, "grad_norm": 0.08071456104516983, "learning_rate": 4.920208907587407e-05, "loss": 0.0054, "step": 14106 }, { "epoch": 1.5293798785776236, "grad_norm": 0.4741291105747223, "learning_rate": 4.919846220803714e-05, "loss": 0.0597, "step": 14107 }, { "epoch": 1.5294882914137033, "grad_norm": 0.613944947719574, "learning_rate": 4.919483534020021e-05, "loss": 0.0447, "step": 14108 }, { "epoch": 1.5295967042497831, "grad_norm": 0.6897696256637573, "learning_rate": 4.919120847236327e-05, "loss": 0.0414, "step": 14109 }, { "epoch": 1.529705117085863, "grad_norm": 0.3284660875797272, "learning_rate": 4.9187581604526335e-05, "loss": 0.0259, "step": 14110 }, { "epoch": 1.5298135299219426, "grad_norm": 0.2697812020778656, "learning_rate": 4.91839547366894e-05, "loss": 0.0075, "step": 14111 }, { "epoch": 1.5299219427580226, "grad_norm": 0.20704743266105652, "learning_rate": 4.918032786885246e-05, "loss": 0.0183, "step": 14112 }, { "epoch": 1.5300303555941024, "grad_norm": 0.0764438658952713, "learning_rate": 4.917670100101553e-05, "loss": 0.0058, "step": 14113 }, { "epoch": 1.5301387684301822, "grad_norm": 0.11490405350923538, "learning_rate": 4.9173074133178584e-05, "loss": 0.0064, "step": 14114 }, { "epoch": 1.530247181266262, "grad_norm": 0.06194545701146126, "learning_rate": 4.9169447265341655e-05, "loss": 0.0038, "step": 14115 }, { "epoch": 1.5303555941023417, "grad_norm": 0.3102354407310486, "learning_rate": 4.916582039750472e-05, "loss": 0.0191, "step": 14116 }, { "epoch": 1.5304640069384217, "grad_norm": 0.1965869516134262, "learning_rate": 4.9162193529667784e-05, "loss": 0.0152, "step": 14117 }, { "epoch": 1.5305724197745012, "grad_norm": 0.20974107086658478, "learning_rate": 4.915856666183085e-05, "loss": 0.0117, "step": 14118 }, { "epoch": 1.5306808326105812, "grad_norm": 0.1823868304491043, "learning_rate": 4.915493979399391e-05, "loss": 0.0278, "step": 14119 }, { "epoch": 1.5307892454466607, "grad_norm": 0.12490149587392807, "learning_rate": 4.9151312926156976e-05, "loss": 0.0109, "step": 14120 }, { "epoch": 1.5308976582827407, "grad_norm": 0.13901673257350922, "learning_rate": 4.914768605832003e-05, "loss": 0.0104, "step": 14121 }, { "epoch": 1.5310060711188205, "grad_norm": 0.19064606726169586, "learning_rate": 4.91440591904831e-05, "loss": 0.0191, "step": 14122 }, { "epoch": 1.5311144839549002, "grad_norm": 0.29325854778289795, "learning_rate": 4.914043232264616e-05, "loss": 0.019, "step": 14123 }, { "epoch": 1.53122289679098, "grad_norm": 0.30484887957572937, "learning_rate": 4.913680545480923e-05, "loss": 0.0371, "step": 14124 }, { "epoch": 1.5313313096270598, "grad_norm": 0.12213212251663208, "learning_rate": 4.9133178586972296e-05, "loss": 0.0125, "step": 14125 }, { "epoch": 1.5314397224631398, "grad_norm": 0.20470333099365234, "learning_rate": 4.912955171913536e-05, "loss": 0.0312, "step": 14126 }, { "epoch": 1.5315481352992193, "grad_norm": 0.24623863399028778, "learning_rate": 4.9125924851298425e-05, "loss": 0.0178, "step": 14127 }, { "epoch": 1.5316565481352993, "grad_norm": 1.5583994388580322, "learning_rate": 4.912229798346148e-05, "loss": 0.0655, "step": 14128 }, { "epoch": 1.531764960971379, "grad_norm": 0.11269407719373703, "learning_rate": 4.9118671115624546e-05, "loss": 0.0153, "step": 14129 }, { "epoch": 1.5318733738074588, "grad_norm": 0.2234874665737152, "learning_rate": 4.911504424778761e-05, "loss": 0.0159, "step": 14130 }, { "epoch": 1.5319817866435386, "grad_norm": 0.2410571277141571, "learning_rate": 4.9111417379950674e-05, "loss": 0.0318, "step": 14131 }, { "epoch": 1.5320901994796183, "grad_norm": 0.38191458582878113, "learning_rate": 4.910779051211374e-05, "loss": 0.0415, "step": 14132 }, { "epoch": 1.5321986123156983, "grad_norm": 0.2783202528953552, "learning_rate": 4.910416364427681e-05, "loss": 0.0316, "step": 14133 }, { "epoch": 1.5323070251517779, "grad_norm": 0.20177185535430908, "learning_rate": 4.910053677643987e-05, "loss": 0.0148, "step": 14134 }, { "epoch": 1.5324154379878578, "grad_norm": 0.41819432377815247, "learning_rate": 4.909690990860293e-05, "loss": 0.0266, "step": 14135 }, { "epoch": 1.5325238508239376, "grad_norm": 0.1442306935787201, "learning_rate": 4.9093283040765995e-05, "loss": 0.0054, "step": 14136 }, { "epoch": 1.5326322636600174, "grad_norm": 0.3019821047782898, "learning_rate": 4.908965617292906e-05, "loss": 0.0162, "step": 14137 }, { "epoch": 1.5327406764960971, "grad_norm": 0.17081886529922485, "learning_rate": 4.908602930509212e-05, "loss": 0.0138, "step": 14138 }, { "epoch": 1.532849089332177, "grad_norm": 1.003248929977417, "learning_rate": 4.908240243725519e-05, "loss": 0.0208, "step": 14139 }, { "epoch": 1.5329575021682569, "grad_norm": 0.3575984835624695, "learning_rate": 4.907877556941825e-05, "loss": 0.0312, "step": 14140 }, { "epoch": 1.5330659150043364, "grad_norm": 0.36199361085891724, "learning_rate": 4.907514870158132e-05, "loss": 0.0404, "step": 14141 }, { "epoch": 1.5331743278404164, "grad_norm": 0.38404667377471924, "learning_rate": 4.907152183374438e-05, "loss": 0.0246, "step": 14142 }, { "epoch": 1.533282740676496, "grad_norm": 0.6103993654251099, "learning_rate": 4.9067894965907444e-05, "loss": 0.0215, "step": 14143 }, { "epoch": 1.533391153512576, "grad_norm": 0.1974555104970932, "learning_rate": 4.906426809807051e-05, "loss": 0.0088, "step": 14144 }, { "epoch": 1.5334995663486557, "grad_norm": 0.09675911068916321, "learning_rate": 4.906064123023357e-05, "loss": 0.0083, "step": 14145 }, { "epoch": 1.5336079791847355, "grad_norm": 0.1069258600473404, "learning_rate": 4.9057014362396636e-05, "loss": 0.0057, "step": 14146 }, { "epoch": 1.5337163920208152, "grad_norm": 0.3388003706932068, "learning_rate": 4.90533874945597e-05, "loss": 0.0142, "step": 14147 }, { "epoch": 1.533824804856895, "grad_norm": 0.17291277647018433, "learning_rate": 4.9049760626722764e-05, "loss": 0.0066, "step": 14148 }, { "epoch": 1.533933217692975, "grad_norm": 0.2869206666946411, "learning_rate": 4.904613375888583e-05, "loss": 0.0149, "step": 14149 }, { "epoch": 1.5340416305290545, "grad_norm": 0.21923796832561493, "learning_rate": 4.904250689104889e-05, "loss": 0.0241, "step": 14150 }, { "epoch": 1.5341500433651345, "grad_norm": 0.416329950094223, "learning_rate": 4.9038880023211956e-05, "loss": 0.0342, "step": 14151 }, { "epoch": 1.5342584562012143, "grad_norm": 0.18859343230724335, "learning_rate": 4.903525315537502e-05, "loss": 0.0233, "step": 14152 }, { "epoch": 1.534366869037294, "grad_norm": 0.0992974191904068, "learning_rate": 4.9031626287538085e-05, "loss": 0.0065, "step": 14153 }, { "epoch": 1.5344752818733738, "grad_norm": 0.385593980550766, "learning_rate": 4.902799941970115e-05, "loss": 0.0222, "step": 14154 }, { "epoch": 1.5345836947094535, "grad_norm": 0.19683323800563812, "learning_rate": 4.902437255186421e-05, "loss": 0.0034, "step": 14155 }, { "epoch": 1.5346921075455335, "grad_norm": 0.5190527439117432, "learning_rate": 4.902074568402728e-05, "loss": 0.0149, "step": 14156 }, { "epoch": 1.534800520381613, "grad_norm": 0.17635948956012726, "learning_rate": 4.901711881619034e-05, "loss": 0.011, "step": 14157 }, { "epoch": 1.534908933217693, "grad_norm": 0.28997376561164856, "learning_rate": 4.90134919483534e-05, "loss": 0.0278, "step": 14158 }, { "epoch": 1.5350173460537726, "grad_norm": 0.37030667066574097, "learning_rate": 4.900986508051647e-05, "loss": 0.0322, "step": 14159 }, { "epoch": 1.5351257588898526, "grad_norm": 0.22000618278980255, "learning_rate": 4.900623821267953e-05, "loss": 0.0165, "step": 14160 }, { "epoch": 1.5352341717259324, "grad_norm": 0.2732594907283783, "learning_rate": 4.90026113448426e-05, "loss": 0.0226, "step": 14161 }, { "epoch": 1.5353425845620121, "grad_norm": 0.4517885744571686, "learning_rate": 4.899898447700566e-05, "loss": 0.0161, "step": 14162 }, { "epoch": 1.535450997398092, "grad_norm": 0.6973572373390198, "learning_rate": 4.8995357609168726e-05, "loss": 0.0517, "step": 14163 }, { "epoch": 1.5355594102341716, "grad_norm": 0.08363711833953857, "learning_rate": 4.899173074133179e-05, "loss": 0.0034, "step": 14164 }, { "epoch": 1.5356678230702516, "grad_norm": 0.17265014350414276, "learning_rate": 4.8988103873494854e-05, "loss": 0.0116, "step": 14165 }, { "epoch": 1.5357762359063312, "grad_norm": 0.9226089119911194, "learning_rate": 4.898447700565791e-05, "loss": 0.0483, "step": 14166 }, { "epoch": 1.5358846487424112, "grad_norm": 0.23080626130104065, "learning_rate": 4.898085013782098e-05, "loss": 0.0095, "step": 14167 }, { "epoch": 1.535993061578491, "grad_norm": 0.23232026398181915, "learning_rate": 4.8977223269984046e-05, "loss": 0.0205, "step": 14168 }, { "epoch": 1.5361014744145707, "grad_norm": 0.29016754031181335, "learning_rate": 4.897359640214711e-05, "loss": 0.029, "step": 14169 }, { "epoch": 1.5362098872506504, "grad_norm": 0.29541081190109253, "learning_rate": 4.8969969534310174e-05, "loss": 0.0194, "step": 14170 }, { "epoch": 1.5363183000867302, "grad_norm": 0.38081982731819153, "learning_rate": 4.896634266647324e-05, "loss": 0.016, "step": 14171 }, { "epoch": 1.5364267129228102, "grad_norm": 0.41295358538627625, "learning_rate": 4.89627157986363e-05, "loss": 0.0362, "step": 14172 }, { "epoch": 1.5365351257588897, "grad_norm": 0.07862634211778641, "learning_rate": 4.895908893079936e-05, "loss": 0.0049, "step": 14173 }, { "epoch": 1.5366435385949697, "grad_norm": 0.3383515179157257, "learning_rate": 4.8955462062962424e-05, "loss": 0.0255, "step": 14174 }, { "epoch": 1.5367519514310495, "grad_norm": 0.32199400663375854, "learning_rate": 4.895183519512549e-05, "loss": 0.0081, "step": 14175 }, { "epoch": 1.5368603642671292, "grad_norm": 0.2826126515865326, "learning_rate": 4.894820832728856e-05, "loss": 0.0111, "step": 14176 }, { "epoch": 1.536968777103209, "grad_norm": 0.36004438996315, "learning_rate": 4.894458145945162e-05, "loss": 0.0224, "step": 14177 }, { "epoch": 1.5370771899392888, "grad_norm": 0.12872236967086792, "learning_rate": 4.894095459161469e-05, "loss": 0.0064, "step": 14178 }, { "epoch": 1.5371856027753688, "grad_norm": 0.4248513877391815, "learning_rate": 4.893732772377775e-05, "loss": 0.0261, "step": 14179 }, { "epoch": 1.5372940156114483, "grad_norm": 0.3791310489177704, "learning_rate": 4.893370085594081e-05, "loss": 0.0184, "step": 14180 }, { "epoch": 1.5374024284475283, "grad_norm": 0.2648043930530548, "learning_rate": 4.893007398810387e-05, "loss": 0.0229, "step": 14181 }, { "epoch": 1.5375108412836078, "grad_norm": 0.39530661702156067, "learning_rate": 4.892644712026694e-05, "loss": 0.0132, "step": 14182 }, { "epoch": 1.5376192541196878, "grad_norm": 0.4666728675365448, "learning_rate": 4.892282025243e-05, "loss": 0.019, "step": 14183 }, { "epoch": 1.5377276669557676, "grad_norm": 0.3671933114528656, "learning_rate": 4.8919193384593065e-05, "loss": 0.0189, "step": 14184 }, { "epoch": 1.5378360797918473, "grad_norm": 0.09266585111618042, "learning_rate": 4.8915566516756136e-05, "loss": 0.005, "step": 14185 }, { "epoch": 1.537944492627927, "grad_norm": 0.47021329402923584, "learning_rate": 4.89119396489192e-05, "loss": 0.0237, "step": 14186 }, { "epoch": 1.5380529054640069, "grad_norm": 0.7329664826393127, "learning_rate": 4.890831278108226e-05, "loss": 0.0365, "step": 14187 }, { "epoch": 1.5381613183000868, "grad_norm": 0.5192590355873108, "learning_rate": 4.890468591324532e-05, "loss": 0.0392, "step": 14188 }, { "epoch": 1.5382697311361664, "grad_norm": 0.3258921205997467, "learning_rate": 4.8901059045408386e-05, "loss": 0.0616, "step": 14189 }, { "epoch": 1.5383781439722464, "grad_norm": 0.16826221346855164, "learning_rate": 4.889743217757145e-05, "loss": 0.0081, "step": 14190 }, { "epoch": 1.5384865568083261, "grad_norm": 0.3332054018974304, "learning_rate": 4.8893805309734514e-05, "loss": 0.0074, "step": 14191 }, { "epoch": 1.538594969644406, "grad_norm": 0.13808375597000122, "learning_rate": 4.889017844189758e-05, "loss": 0.0085, "step": 14192 }, { "epoch": 1.5387033824804857, "grad_norm": 0.25704559683799744, "learning_rate": 4.888655157406065e-05, "loss": 0.0349, "step": 14193 }, { "epoch": 1.5388117953165654, "grad_norm": 0.7079044580459595, "learning_rate": 4.8882924706223706e-05, "loss": 0.0369, "step": 14194 }, { "epoch": 1.5389202081526454, "grad_norm": 0.21832461655139923, "learning_rate": 4.887929783838677e-05, "loss": 0.0183, "step": 14195 }, { "epoch": 1.539028620988725, "grad_norm": 0.12355440855026245, "learning_rate": 4.8875670970549834e-05, "loss": 0.0085, "step": 14196 }, { "epoch": 1.539137033824805, "grad_norm": 0.7689732909202576, "learning_rate": 4.88720441027129e-05, "loss": 0.0263, "step": 14197 }, { "epoch": 1.5392454466608847, "grad_norm": 0.5469130277633667, "learning_rate": 4.886841723487596e-05, "loss": 0.0151, "step": 14198 }, { "epoch": 1.5393538594969645, "grad_norm": 0.4284590482711792, "learning_rate": 4.886479036703903e-05, "loss": 0.0605, "step": 14199 }, { "epoch": 1.5394622723330442, "grad_norm": 0.22497400641441345, "learning_rate": 4.886116349920209e-05, "loss": 0.0109, "step": 14200 }, { "epoch": 1.539570685169124, "grad_norm": 0.266339510679245, "learning_rate": 4.8857536631365155e-05, "loss": 0.0089, "step": 14201 }, { "epoch": 1.539679098005204, "grad_norm": 0.4053948223590851, "learning_rate": 4.885390976352822e-05, "loss": 0.0402, "step": 14202 }, { "epoch": 1.5397875108412835, "grad_norm": 0.03947952389717102, "learning_rate": 4.885028289569128e-05, "loss": 0.0026, "step": 14203 }, { "epoch": 1.5398959236773635, "grad_norm": 0.3278539478778839, "learning_rate": 4.884665602785435e-05, "loss": 0.0297, "step": 14204 }, { "epoch": 1.540004336513443, "grad_norm": 0.1723260134458542, "learning_rate": 4.884302916001741e-05, "loss": 0.0064, "step": 14205 }, { "epoch": 1.540112749349523, "grad_norm": 0.06759613752365112, "learning_rate": 4.8839402292180475e-05, "loss": 0.0029, "step": 14206 }, { "epoch": 1.5402211621856028, "grad_norm": 0.6077008247375488, "learning_rate": 4.883577542434354e-05, "loss": 0.0415, "step": 14207 }, { "epoch": 1.5403295750216826, "grad_norm": 0.4266531765460968, "learning_rate": 4.8832148556506604e-05, "loss": 0.0323, "step": 14208 }, { "epoch": 1.5404379878577623, "grad_norm": 0.22945500910282135, "learning_rate": 4.882852168866967e-05, "loss": 0.0439, "step": 14209 }, { "epoch": 1.540546400693842, "grad_norm": 0.1285804957151413, "learning_rate": 4.882489482083273e-05, "loss": 0.004, "step": 14210 }, { "epoch": 1.540654813529922, "grad_norm": 0.356261670589447, "learning_rate": 4.8821267952995796e-05, "loss": 0.0192, "step": 14211 }, { "epoch": 1.5407632263660016, "grad_norm": 0.29546886682510376, "learning_rate": 4.881764108515886e-05, "loss": 0.0134, "step": 14212 }, { "epoch": 1.5408716392020816, "grad_norm": 0.0716450959444046, "learning_rate": 4.8814014217321924e-05, "loss": 0.0039, "step": 14213 }, { "epoch": 1.5409800520381614, "grad_norm": 0.16934868693351746, "learning_rate": 4.881038734948499e-05, "loss": 0.013, "step": 14214 }, { "epoch": 1.5410884648742411, "grad_norm": 0.8102987408638, "learning_rate": 4.880676048164805e-05, "loss": 0.0299, "step": 14215 }, { "epoch": 1.5411968777103209, "grad_norm": 0.24927102029323578, "learning_rate": 4.8803133613811116e-05, "loss": 0.0322, "step": 14216 }, { "epoch": 1.5413052905464006, "grad_norm": 0.2572248876094818, "learning_rate": 4.879950674597418e-05, "loss": 0.0142, "step": 14217 }, { "epoch": 1.5414137033824806, "grad_norm": 0.49695658683776855, "learning_rate": 4.879587987813724e-05, "loss": 0.0599, "step": 14218 }, { "epoch": 1.5415221162185602, "grad_norm": 0.14715105295181274, "learning_rate": 4.879225301030031e-05, "loss": 0.0062, "step": 14219 }, { "epoch": 1.5416305290546402, "grad_norm": 0.1610991507768631, "learning_rate": 4.878862614246337e-05, "loss": 0.0092, "step": 14220 }, { "epoch": 1.54173894189072, "grad_norm": 0.3523007035255432, "learning_rate": 4.878499927462644e-05, "loss": 0.0305, "step": 14221 }, { "epoch": 1.5418473547267997, "grad_norm": 0.4328692555427551, "learning_rate": 4.87813724067895e-05, "loss": 0.0382, "step": 14222 }, { "epoch": 1.5419557675628794, "grad_norm": 0.5377745032310486, "learning_rate": 4.8777745538952565e-05, "loss": 0.0503, "step": 14223 }, { "epoch": 1.5420641803989592, "grad_norm": 0.18284928798675537, "learning_rate": 4.877411867111563e-05, "loss": 0.0487, "step": 14224 }, { "epoch": 1.5421725932350392, "grad_norm": 0.355146586894989, "learning_rate": 4.8770491803278687e-05, "loss": 0.0204, "step": 14225 }, { "epoch": 1.5422810060711187, "grad_norm": 0.5795179605484009, "learning_rate": 4.876686493544175e-05, "loss": 0.0456, "step": 14226 }, { "epoch": 1.5423894189071987, "grad_norm": 0.4113422632217407, "learning_rate": 4.8763238067604815e-05, "loss": 0.027, "step": 14227 }, { "epoch": 1.5424978317432783, "grad_norm": 0.45799529552459717, "learning_rate": 4.8759611199767886e-05, "loss": 0.018, "step": 14228 }, { "epoch": 1.5426062445793582, "grad_norm": 0.24550659954547882, "learning_rate": 4.875598433193095e-05, "loss": 0.0217, "step": 14229 }, { "epoch": 1.542714657415438, "grad_norm": 0.2601926326751709, "learning_rate": 4.8752357464094014e-05, "loss": 0.0134, "step": 14230 }, { "epoch": 1.5428230702515178, "grad_norm": 0.31556597352027893, "learning_rate": 4.874873059625708e-05, "loss": 0.0364, "step": 14231 }, { "epoch": 1.5429314830875975, "grad_norm": 0.40433040261268616, "learning_rate": 4.8745103728420135e-05, "loss": 0.0246, "step": 14232 }, { "epoch": 1.5430398959236773, "grad_norm": 0.5934610366821289, "learning_rate": 4.87414768605832e-05, "loss": 0.1095, "step": 14233 }, { "epoch": 1.5431483087597573, "grad_norm": 0.30693691968917847, "learning_rate": 4.8737849992746264e-05, "loss": 0.0267, "step": 14234 }, { "epoch": 1.5432567215958368, "grad_norm": 0.27059227228164673, "learning_rate": 4.873422312490933e-05, "loss": 0.0185, "step": 14235 }, { "epoch": 1.5433651344319168, "grad_norm": 0.20859763026237488, "learning_rate": 4.87305962570724e-05, "loss": 0.0228, "step": 14236 }, { "epoch": 1.5434735472679966, "grad_norm": 0.3368764817714691, "learning_rate": 4.872696938923546e-05, "loss": 0.0137, "step": 14237 }, { "epoch": 1.5435819601040763, "grad_norm": 0.22396907210350037, "learning_rate": 4.872334252139853e-05, "loss": 0.0116, "step": 14238 }, { "epoch": 1.543690372940156, "grad_norm": 0.31825152039527893, "learning_rate": 4.8719715653561584e-05, "loss": 0.0114, "step": 14239 }, { "epoch": 1.5437987857762359, "grad_norm": 0.5780170559883118, "learning_rate": 4.871608878572465e-05, "loss": 0.0187, "step": 14240 }, { "epoch": 1.5439071986123158, "grad_norm": 0.2943708896636963, "learning_rate": 4.871246191788771e-05, "loss": 0.0117, "step": 14241 }, { "epoch": 1.5440156114483954, "grad_norm": 0.11704523116350174, "learning_rate": 4.8708835050050776e-05, "loss": 0.0078, "step": 14242 }, { "epoch": 1.5441240242844754, "grad_norm": 0.41171425580978394, "learning_rate": 4.870520818221384e-05, "loss": 0.0479, "step": 14243 }, { "epoch": 1.544232437120555, "grad_norm": 0.7222301959991455, "learning_rate": 4.8701581314376905e-05, "loss": 0.0203, "step": 14244 }, { "epoch": 1.544340849956635, "grad_norm": 0.22905448079109192, "learning_rate": 4.8697954446539975e-05, "loss": 0.0125, "step": 14245 }, { "epoch": 1.5444492627927147, "grad_norm": 0.30945494771003723, "learning_rate": 4.869432757870304e-05, "loss": 0.0289, "step": 14246 }, { "epoch": 1.5445576756287944, "grad_norm": 0.07283459603786469, "learning_rate": 4.86907007108661e-05, "loss": 0.0048, "step": 14247 }, { "epoch": 1.5446660884648742, "grad_norm": 0.08447732776403427, "learning_rate": 4.868707384302916e-05, "loss": 0.0047, "step": 14248 }, { "epoch": 1.544774501300954, "grad_norm": 0.26077866554260254, "learning_rate": 4.8683446975192225e-05, "loss": 0.0322, "step": 14249 }, { "epoch": 1.544882914137034, "grad_norm": 0.1315920352935791, "learning_rate": 4.867982010735529e-05, "loss": 0.007, "step": 14250 }, { "epoch": 1.5449913269731135, "grad_norm": 0.15921276807785034, "learning_rate": 4.867619323951835e-05, "loss": 0.0113, "step": 14251 }, { "epoch": 1.5450997398091935, "grad_norm": 0.3566907048225403, "learning_rate": 4.867256637168142e-05, "loss": 0.012, "step": 14252 }, { "epoch": 1.5452081526452732, "grad_norm": 0.2308494746685028, "learning_rate": 4.866893950384448e-05, "loss": 0.0358, "step": 14253 }, { "epoch": 1.545316565481353, "grad_norm": 0.3367787301540375, "learning_rate": 4.8665312636007546e-05, "loss": 0.0356, "step": 14254 }, { "epoch": 1.5454249783174328, "grad_norm": 0.4207119941711426, "learning_rate": 4.866168576817061e-05, "loss": 0.0431, "step": 14255 }, { "epoch": 1.5455333911535125, "grad_norm": 0.3798545002937317, "learning_rate": 4.8658058900333674e-05, "loss": 0.0299, "step": 14256 }, { "epoch": 1.5456418039895925, "grad_norm": 0.4158881902694702, "learning_rate": 4.865443203249674e-05, "loss": 0.0363, "step": 14257 }, { "epoch": 1.545750216825672, "grad_norm": 0.28106701374053955, "learning_rate": 4.86508051646598e-05, "loss": 0.0218, "step": 14258 }, { "epoch": 1.545858629661752, "grad_norm": 0.20954068005084991, "learning_rate": 4.8647178296822866e-05, "loss": 0.0399, "step": 14259 }, { "epoch": 1.5459670424978318, "grad_norm": 0.7444085478782654, "learning_rate": 4.864355142898593e-05, "loss": 0.0323, "step": 14260 }, { "epoch": 1.5460754553339116, "grad_norm": 0.3567039370536804, "learning_rate": 4.8639924561148994e-05, "loss": 0.0214, "step": 14261 }, { "epoch": 1.5461838681699913, "grad_norm": 0.4058631658554077, "learning_rate": 4.863629769331206e-05, "loss": 0.0603, "step": 14262 }, { "epoch": 1.546292281006071, "grad_norm": 0.3058006167411804, "learning_rate": 4.863267082547512e-05, "loss": 0.0126, "step": 14263 }, { "epoch": 1.546400693842151, "grad_norm": 0.38287225365638733, "learning_rate": 4.862904395763819e-05, "loss": 0.0171, "step": 14264 }, { "epoch": 1.5465091066782306, "grad_norm": 0.12275233119726181, "learning_rate": 4.862541708980125e-05, "loss": 0.0095, "step": 14265 }, { "epoch": 1.5466175195143106, "grad_norm": 0.08885951340198517, "learning_rate": 4.8621790221964315e-05, "loss": 0.0056, "step": 14266 }, { "epoch": 1.5467259323503901, "grad_norm": 0.5925626158714294, "learning_rate": 4.861816335412738e-05, "loss": 0.0356, "step": 14267 }, { "epoch": 1.5468343451864701, "grad_norm": 0.16798779368400574, "learning_rate": 4.861453648629044e-05, "loss": 0.0155, "step": 14268 }, { "epoch": 1.5469427580225499, "grad_norm": 0.4190564751625061, "learning_rate": 4.861090961845351e-05, "loss": 0.0162, "step": 14269 }, { "epoch": 1.5470511708586296, "grad_norm": 0.3195575475692749, "learning_rate": 4.8607282750616565e-05, "loss": 0.0063, "step": 14270 }, { "epoch": 1.5471595836947094, "grad_norm": 0.30121973156929016, "learning_rate": 4.8603655882779635e-05, "loss": 0.0252, "step": 14271 }, { "epoch": 1.5472679965307892, "grad_norm": 0.23282065987586975, "learning_rate": 4.86000290149427e-05, "loss": 0.016, "step": 14272 }, { "epoch": 1.5473764093668692, "grad_norm": 0.4224969744682312, "learning_rate": 4.8596402147105764e-05, "loss": 0.0526, "step": 14273 }, { "epoch": 1.5474848222029487, "grad_norm": 0.16093914210796356, "learning_rate": 4.859277527926883e-05, "loss": 0.0145, "step": 14274 }, { "epoch": 1.5475932350390287, "grad_norm": 0.25656071305274963, "learning_rate": 4.858914841143189e-05, "loss": 0.0108, "step": 14275 }, { "epoch": 1.5477016478751084, "grad_norm": 0.2630438804626465, "learning_rate": 4.8585521543594956e-05, "loss": 0.012, "step": 14276 }, { "epoch": 1.5478100607111882, "grad_norm": 0.09318612515926361, "learning_rate": 4.858189467575801e-05, "loss": 0.0045, "step": 14277 }, { "epoch": 1.547918473547268, "grad_norm": 0.19886407256126404, "learning_rate": 4.857826780792108e-05, "loss": 0.025, "step": 14278 }, { "epoch": 1.5480268863833477, "grad_norm": 0.4488184154033661, "learning_rate": 4.857464094008414e-05, "loss": 0.0336, "step": 14279 }, { "epoch": 1.5481352992194277, "grad_norm": 0.1956022083759308, "learning_rate": 4.857101407224721e-05, "loss": 0.0198, "step": 14280 }, { "epoch": 1.5482437120555073, "grad_norm": 0.5343745946884155, "learning_rate": 4.8567387204410276e-05, "loss": 0.0393, "step": 14281 }, { "epoch": 1.5483521248915872, "grad_norm": 0.24682892858982086, "learning_rate": 4.856376033657334e-05, "loss": 0.0242, "step": 14282 }, { "epoch": 1.548460537727667, "grad_norm": 0.1995931714773178, "learning_rate": 4.8560133468736405e-05, "loss": 0.0085, "step": 14283 }, { "epoch": 1.5485689505637468, "grad_norm": 0.6130265593528748, "learning_rate": 4.855650660089946e-05, "loss": 0.012, "step": 14284 }, { "epoch": 1.5486773633998265, "grad_norm": 0.39883658289909363, "learning_rate": 4.8552879733062526e-05, "loss": 0.018, "step": 14285 }, { "epoch": 1.5487857762359063, "grad_norm": 0.2746902108192444, "learning_rate": 4.854925286522559e-05, "loss": 0.0115, "step": 14286 }, { "epoch": 1.5488941890719863, "grad_norm": 0.07011787593364716, "learning_rate": 4.8545625997388654e-05, "loss": 0.0011, "step": 14287 }, { "epoch": 1.5490026019080658, "grad_norm": 0.04471075162291527, "learning_rate": 4.8541999129551725e-05, "loss": 0.0032, "step": 14288 }, { "epoch": 1.5491110147441458, "grad_norm": 0.3077819347381592, "learning_rate": 4.853837226171479e-05, "loss": 0.0622, "step": 14289 }, { "epoch": 1.5492194275802254, "grad_norm": 0.9088491797447205, "learning_rate": 4.8534745393877853e-05, "loss": 0.0137, "step": 14290 }, { "epoch": 1.5493278404163053, "grad_norm": 0.635715663433075, "learning_rate": 4.853111852604091e-05, "loss": 0.0466, "step": 14291 }, { "epoch": 1.549436253252385, "grad_norm": 0.2970966100692749, "learning_rate": 4.8527491658203975e-05, "loss": 0.006, "step": 14292 }, { "epoch": 1.5495446660884649, "grad_norm": 1.2572141885757446, "learning_rate": 4.852386479036704e-05, "loss": 0.0661, "step": 14293 }, { "epoch": 1.5496530789245446, "grad_norm": 0.37643229961395264, "learning_rate": 4.85202379225301e-05, "loss": 0.0686, "step": 14294 }, { "epoch": 1.5497614917606244, "grad_norm": 0.5120288729667664, "learning_rate": 4.851661105469317e-05, "loss": 0.0432, "step": 14295 }, { "epoch": 1.5498699045967044, "grad_norm": 0.3592659831047058, "learning_rate": 4.851298418685623e-05, "loss": 0.021, "step": 14296 }, { "epoch": 1.549978317432784, "grad_norm": 0.45918479561805725, "learning_rate": 4.85093573190193e-05, "loss": 0.0269, "step": 14297 }, { "epoch": 1.550086730268864, "grad_norm": 0.5713006258010864, "learning_rate": 4.8505730451182366e-05, "loss": 0.0073, "step": 14298 }, { "epoch": 1.5501951431049437, "grad_norm": 0.2762765884399414, "learning_rate": 4.8502103583345424e-05, "loss": 0.0174, "step": 14299 }, { "epoch": 1.5503035559410234, "grad_norm": 0.3489520251750946, "learning_rate": 4.849847671550849e-05, "loss": 0.0555, "step": 14300 }, { "epoch": 1.5504119687771032, "grad_norm": 0.2961413264274597, "learning_rate": 4.849484984767155e-05, "loss": 0.0385, "step": 14301 }, { "epoch": 1.550520381613183, "grad_norm": 0.1204766035079956, "learning_rate": 4.8491222979834616e-05, "loss": 0.0052, "step": 14302 }, { "epoch": 1.550628794449263, "grad_norm": 0.21986347436904907, "learning_rate": 4.848759611199768e-05, "loss": 0.0163, "step": 14303 }, { "epoch": 1.5507372072853425, "grad_norm": 0.30021312832832336, "learning_rate": 4.8483969244160744e-05, "loss": 0.0561, "step": 14304 }, { "epoch": 1.5508456201214225, "grad_norm": 0.05241816118359566, "learning_rate": 4.8480342376323815e-05, "loss": 0.0059, "step": 14305 }, { "epoch": 1.550954032957502, "grad_norm": 0.3199070692062378, "learning_rate": 4.847671550848687e-05, "loss": 0.0341, "step": 14306 }, { "epoch": 1.551062445793582, "grad_norm": 0.13288797438144684, "learning_rate": 4.8473088640649936e-05, "loss": 0.0158, "step": 14307 }, { "epoch": 1.5511708586296618, "grad_norm": 0.31272318959236145, "learning_rate": 4.8469461772813e-05, "loss": 0.0232, "step": 14308 }, { "epoch": 1.5512792714657415, "grad_norm": 0.3222551643848419, "learning_rate": 4.8465834904976065e-05, "loss": 0.0156, "step": 14309 }, { "epoch": 1.5513876843018215, "grad_norm": 0.21946470439434052, "learning_rate": 4.846220803713913e-05, "loss": 0.0137, "step": 14310 }, { "epoch": 1.551496097137901, "grad_norm": 0.1694188117980957, "learning_rate": 4.845858116930219e-05, "loss": 0.0188, "step": 14311 }, { "epoch": 1.551604509973981, "grad_norm": 0.34421539306640625, "learning_rate": 4.845495430146526e-05, "loss": 0.0412, "step": 14312 }, { "epoch": 1.5517129228100606, "grad_norm": 0.48916205763816833, "learning_rate": 4.845132743362832e-05, "loss": 0.0696, "step": 14313 }, { "epoch": 1.5518213356461406, "grad_norm": 0.07990273833274841, "learning_rate": 4.8447700565791385e-05, "loss": 0.0037, "step": 14314 }, { "epoch": 1.5519297484822203, "grad_norm": 0.2737704813480377, "learning_rate": 4.844407369795445e-05, "loss": 0.0321, "step": 14315 }, { "epoch": 1.5520381613183, "grad_norm": 0.33886751532554626, "learning_rate": 4.844044683011751e-05, "loss": 0.0204, "step": 14316 }, { "epoch": 1.5521465741543798, "grad_norm": 0.15436819195747375, "learning_rate": 4.843681996228058e-05, "loss": 0.0084, "step": 14317 }, { "epoch": 1.5522549869904596, "grad_norm": 0.15698207914829254, "learning_rate": 4.843319309444364e-05, "loss": 0.0191, "step": 14318 }, { "epoch": 1.5523633998265396, "grad_norm": 0.1507071703672409, "learning_rate": 4.8429566226606706e-05, "loss": 0.0185, "step": 14319 }, { "epoch": 1.5524718126626191, "grad_norm": 0.1284487098455429, "learning_rate": 4.842593935876977e-05, "loss": 0.0101, "step": 14320 }, { "epoch": 1.5525802254986991, "grad_norm": 0.18969236314296722, "learning_rate": 4.8422312490932834e-05, "loss": 0.0162, "step": 14321 }, { "epoch": 1.5526886383347789, "grad_norm": 0.2606620788574219, "learning_rate": 4.841868562309589e-05, "loss": 0.0287, "step": 14322 }, { "epoch": 1.5527970511708586, "grad_norm": 0.10388161242008209, "learning_rate": 4.841505875525896e-05, "loss": 0.0052, "step": 14323 }, { "epoch": 1.5529054640069384, "grad_norm": 0.11376763880252838, "learning_rate": 4.8411431887422026e-05, "loss": 0.0079, "step": 14324 }, { "epoch": 1.5530138768430182, "grad_norm": 0.08202098309993744, "learning_rate": 4.840780501958509e-05, "loss": 0.009, "step": 14325 }, { "epoch": 1.5531222896790982, "grad_norm": 0.07196749001741409, "learning_rate": 4.8404178151748154e-05, "loss": 0.008, "step": 14326 }, { "epoch": 1.5532307025151777, "grad_norm": 0.11865348368883133, "learning_rate": 4.840055128391122e-05, "loss": 0.0076, "step": 14327 }, { "epoch": 1.5533391153512577, "grad_norm": 1.3755141496658325, "learning_rate": 4.839692441607428e-05, "loss": 0.0813, "step": 14328 }, { "epoch": 1.5534475281873372, "grad_norm": 0.18263360857963562, "learning_rate": 4.839329754823734e-05, "loss": 0.0091, "step": 14329 }, { "epoch": 1.5535559410234172, "grad_norm": 0.18004320561885834, "learning_rate": 4.8389670680400404e-05, "loss": 0.0151, "step": 14330 }, { "epoch": 1.553664353859497, "grad_norm": 0.31238853931427, "learning_rate": 4.8386043812563475e-05, "loss": 0.0325, "step": 14331 }, { "epoch": 1.5537727666955767, "grad_norm": 0.4491814076900482, "learning_rate": 4.838241694472654e-05, "loss": 0.0315, "step": 14332 }, { "epoch": 1.5538811795316565, "grad_norm": 0.37984979152679443, "learning_rate": 4.83787900768896e-05, "loss": 0.0386, "step": 14333 }, { "epoch": 1.5539895923677363, "grad_norm": 0.2259201556444168, "learning_rate": 4.837516320905267e-05, "loss": 0.018, "step": 14334 }, { "epoch": 1.5540980052038162, "grad_norm": 0.12458823621273041, "learning_rate": 4.837153634121573e-05, "loss": 0.0138, "step": 14335 }, { "epoch": 1.5542064180398958, "grad_norm": 0.21919207274913788, "learning_rate": 4.836790947337879e-05, "loss": 0.0103, "step": 14336 }, { "epoch": 1.5543148308759758, "grad_norm": 0.35080990195274353, "learning_rate": 4.836428260554185e-05, "loss": 0.0203, "step": 14337 }, { "epoch": 1.5544232437120555, "grad_norm": 0.4475441873073578, "learning_rate": 4.836065573770492e-05, "loss": 0.0093, "step": 14338 }, { "epoch": 1.5545316565481353, "grad_norm": 0.43701139092445374, "learning_rate": 4.835702886986798e-05, "loss": 0.037, "step": 14339 }, { "epoch": 1.554640069384215, "grad_norm": 0.22419902682304382, "learning_rate": 4.835340200203105e-05, "loss": 0.0248, "step": 14340 }, { "epoch": 1.5547484822202948, "grad_norm": 0.7990421652793884, "learning_rate": 4.8349775134194116e-05, "loss": 0.0828, "step": 14341 }, { "epoch": 1.5548568950563748, "grad_norm": 0.41625234484672546, "learning_rate": 4.834614826635718e-05, "loss": 0.035, "step": 14342 }, { "epoch": 1.5549653078924544, "grad_norm": 0.24876631796360016, "learning_rate": 4.834252139852024e-05, "loss": 0.0226, "step": 14343 }, { "epoch": 1.5550737207285343, "grad_norm": 0.16396670043468475, "learning_rate": 4.83388945306833e-05, "loss": 0.0155, "step": 14344 }, { "epoch": 1.555182133564614, "grad_norm": 0.2489834874868393, "learning_rate": 4.8335267662846366e-05, "loss": 0.04, "step": 14345 }, { "epoch": 1.5552905464006939, "grad_norm": 0.0998997837305069, "learning_rate": 4.833164079500943e-05, "loss": 0.0031, "step": 14346 }, { "epoch": 1.5553989592367736, "grad_norm": 0.2571878731250763, "learning_rate": 4.8328013927172494e-05, "loss": 0.0213, "step": 14347 }, { "epoch": 1.5555073720728534, "grad_norm": 0.34087473154067993, "learning_rate": 4.832438705933556e-05, "loss": 0.0112, "step": 14348 }, { "epoch": 1.5556157849089334, "grad_norm": 0.3088362514972687, "learning_rate": 4.832076019149863e-05, "loss": 0.0218, "step": 14349 }, { "epoch": 1.555724197745013, "grad_norm": 0.48567017912864685, "learning_rate": 4.831713332366169e-05, "loss": 0.0451, "step": 14350 }, { "epoch": 1.555832610581093, "grad_norm": 0.2150050699710846, "learning_rate": 4.831350645582475e-05, "loss": 0.0207, "step": 14351 }, { "epoch": 1.5559410234171724, "grad_norm": 0.47936317324638367, "learning_rate": 4.8309879587987814e-05, "loss": 0.0137, "step": 14352 }, { "epoch": 1.5560494362532524, "grad_norm": 0.3959413766860962, "learning_rate": 4.830625272015088e-05, "loss": 0.0255, "step": 14353 }, { "epoch": 1.5561578490893322, "grad_norm": 0.1711713671684265, "learning_rate": 4.830262585231394e-05, "loss": 0.0174, "step": 14354 }, { "epoch": 1.556266261925412, "grad_norm": 0.8680493235588074, "learning_rate": 4.829899898447701e-05, "loss": 0.0295, "step": 14355 }, { "epoch": 1.5563746747614917, "grad_norm": 0.4168882369995117, "learning_rate": 4.829537211664007e-05, "loss": 0.0444, "step": 14356 }, { "epoch": 1.5564830875975715, "grad_norm": 0.38468137383461, "learning_rate": 4.829174524880314e-05, "loss": 0.0386, "step": 14357 }, { "epoch": 1.5565915004336515, "grad_norm": 0.41104933619499207, "learning_rate": 4.82881183809662e-05, "loss": 0.0328, "step": 14358 }, { "epoch": 1.556699913269731, "grad_norm": 0.35285109281539917, "learning_rate": 4.828449151312926e-05, "loss": 0.0247, "step": 14359 }, { "epoch": 1.556808326105811, "grad_norm": 0.4757174849510193, "learning_rate": 4.828086464529233e-05, "loss": 0.0731, "step": 14360 }, { "epoch": 1.5569167389418908, "grad_norm": 0.3238036036491394, "learning_rate": 4.827723777745539e-05, "loss": 0.0166, "step": 14361 }, { "epoch": 1.5570251517779705, "grad_norm": 0.2613467872142792, "learning_rate": 4.8273610909618455e-05, "loss": 0.0087, "step": 14362 }, { "epoch": 1.5571335646140503, "grad_norm": 0.6225875616073608, "learning_rate": 4.826998404178152e-05, "loss": 0.0198, "step": 14363 }, { "epoch": 1.55724197745013, "grad_norm": 0.20247337222099304, "learning_rate": 4.8266357173944584e-05, "loss": 0.0132, "step": 14364 }, { "epoch": 1.55735039028621, "grad_norm": 0.5050885677337646, "learning_rate": 4.826273030610765e-05, "loss": 0.0319, "step": 14365 }, { "epoch": 1.5574588031222896, "grad_norm": 0.24172727763652802, "learning_rate": 4.825910343827071e-05, "loss": 0.0158, "step": 14366 }, { "epoch": 1.5575672159583696, "grad_norm": 0.30885952711105347, "learning_rate": 4.8255476570433776e-05, "loss": 0.0255, "step": 14367 }, { "epoch": 1.557675628794449, "grad_norm": 0.22139354050159454, "learning_rate": 4.825184970259684e-05, "loss": 0.0142, "step": 14368 }, { "epoch": 1.557784041630529, "grad_norm": 0.3480723798274994, "learning_rate": 4.8248222834759904e-05, "loss": 0.0258, "step": 14369 }, { "epoch": 1.5578924544666088, "grad_norm": 0.49446940422058105, "learning_rate": 4.824459596692297e-05, "loss": 0.0239, "step": 14370 }, { "epoch": 1.5580008673026886, "grad_norm": 0.24227970838546753, "learning_rate": 4.824096909908603e-05, "loss": 0.0173, "step": 14371 }, { "epoch": 1.5581092801387686, "grad_norm": 0.1686180830001831, "learning_rate": 4.8237342231249096e-05, "loss": 0.0119, "step": 14372 }, { "epoch": 1.5582176929748481, "grad_norm": 0.25574031472206116, "learning_rate": 4.823371536341216e-05, "loss": 0.0196, "step": 14373 }, { "epoch": 1.5583261058109281, "grad_norm": 0.11789976805448532, "learning_rate": 4.823008849557522e-05, "loss": 0.0069, "step": 14374 }, { "epoch": 1.5584345186470077, "grad_norm": 0.2684650421142578, "learning_rate": 4.822646162773829e-05, "loss": 0.0441, "step": 14375 }, { "epoch": 1.5585429314830876, "grad_norm": 0.14104492962360382, "learning_rate": 4.822283475990135e-05, "loss": 0.019, "step": 14376 }, { "epoch": 1.5586513443191674, "grad_norm": 0.5909505486488342, "learning_rate": 4.821920789206442e-05, "loss": 0.054, "step": 14377 }, { "epoch": 1.5587597571552472, "grad_norm": 0.9425960779190063, "learning_rate": 4.821558102422748e-05, "loss": 0.0293, "step": 14378 }, { "epoch": 1.558868169991327, "grad_norm": 1.814073920249939, "learning_rate": 4.8211954156390545e-05, "loss": 0.0485, "step": 14379 }, { "epoch": 1.5589765828274067, "grad_norm": 0.11458650976419449, "learning_rate": 4.820832728855361e-05, "loss": 0.0158, "step": 14380 }, { "epoch": 1.5590849956634867, "grad_norm": 0.17592544853687286, "learning_rate": 4.820470042071667e-05, "loss": 0.0074, "step": 14381 }, { "epoch": 1.5591934084995662, "grad_norm": 0.3747592866420746, "learning_rate": 4.820107355287973e-05, "loss": 0.0088, "step": 14382 }, { "epoch": 1.5593018213356462, "grad_norm": 0.33584126830101013, "learning_rate": 4.81974466850428e-05, "loss": 0.0219, "step": 14383 }, { "epoch": 1.559410234171726, "grad_norm": 0.20422539114952087, "learning_rate": 4.8193819817205866e-05, "loss": 0.0144, "step": 14384 }, { "epoch": 1.5595186470078057, "grad_norm": 0.2667149603366852, "learning_rate": 4.819019294936893e-05, "loss": 0.0443, "step": 14385 }, { "epoch": 1.5596270598438855, "grad_norm": 0.40607309341430664, "learning_rate": 4.8186566081531994e-05, "loss": 0.007, "step": 14386 }, { "epoch": 1.5597354726799653, "grad_norm": 0.5070035457611084, "learning_rate": 4.818293921369506e-05, "loss": 0.109, "step": 14387 }, { "epoch": 1.5598438855160452, "grad_norm": 0.38425278663635254, "learning_rate": 4.8179312345858115e-05, "loss": 0.0125, "step": 14388 }, { "epoch": 1.5599522983521248, "grad_norm": 0.49591749906539917, "learning_rate": 4.817568547802118e-05, "loss": 0.0393, "step": 14389 }, { "epoch": 1.5600607111882048, "grad_norm": 0.372053861618042, "learning_rate": 4.8172058610184244e-05, "loss": 0.0311, "step": 14390 }, { "epoch": 1.5601691240242843, "grad_norm": 0.5820514559745789, "learning_rate": 4.816843174234731e-05, "loss": 0.0328, "step": 14391 }, { "epoch": 1.5602775368603643, "grad_norm": 0.5011878609657288, "learning_rate": 4.816480487451038e-05, "loss": 0.0309, "step": 14392 }, { "epoch": 1.560385949696444, "grad_norm": 0.4694213271141052, "learning_rate": 4.816117800667344e-05, "loss": 0.0403, "step": 14393 }, { "epoch": 1.5604943625325238, "grad_norm": 0.522555410861969, "learning_rate": 4.815755113883651e-05, "loss": 0.0518, "step": 14394 }, { "epoch": 1.5606027753686036, "grad_norm": 0.08992695808410645, "learning_rate": 4.8153924270999564e-05, "loss": 0.0098, "step": 14395 }, { "epoch": 1.5607111882046834, "grad_norm": 0.42745688557624817, "learning_rate": 4.815029740316263e-05, "loss": 0.0123, "step": 14396 }, { "epoch": 1.5608196010407633, "grad_norm": 0.26528844237327576, "learning_rate": 4.814667053532569e-05, "loss": 0.0367, "step": 14397 }, { "epoch": 1.5609280138768429, "grad_norm": 0.5229685306549072, "learning_rate": 4.8143043667488756e-05, "loss": 0.0235, "step": 14398 }, { "epoch": 1.5610364267129229, "grad_norm": 0.2634964883327484, "learning_rate": 4.813941679965182e-05, "loss": 0.023, "step": 14399 }, { "epoch": 1.5611448395490026, "grad_norm": 0.05534131079912186, "learning_rate": 4.813578993181489e-05, "loss": 0.0024, "step": 14400 }, { "epoch": 1.5612532523850824, "grad_norm": 0.3186909854412079, "learning_rate": 4.8132163063977956e-05, "loss": 0.0423, "step": 14401 }, { "epoch": 1.5613616652211622, "grad_norm": 0.566203773021698, "learning_rate": 4.812853619614102e-05, "loss": 0.0354, "step": 14402 }, { "epoch": 1.561470078057242, "grad_norm": 0.15461450815200806, "learning_rate": 4.812490932830408e-05, "loss": 0.0106, "step": 14403 }, { "epoch": 1.561578490893322, "grad_norm": 0.2941162586212158, "learning_rate": 4.812128246046714e-05, "loss": 0.0293, "step": 14404 }, { "epoch": 1.5616869037294014, "grad_norm": 0.28258365392684937, "learning_rate": 4.8117655592630205e-05, "loss": 0.0403, "step": 14405 }, { "epoch": 1.5617953165654814, "grad_norm": 0.05479099601507187, "learning_rate": 4.811402872479327e-05, "loss": 0.005, "step": 14406 }, { "epoch": 1.5619037294015612, "grad_norm": 0.15841753780841827, "learning_rate": 4.811040185695633e-05, "loss": 0.0289, "step": 14407 }, { "epoch": 1.562012142237641, "grad_norm": 0.6031799912452698, "learning_rate": 4.81067749891194e-05, "loss": 0.038, "step": 14408 }, { "epoch": 1.5621205550737207, "grad_norm": 0.2657932937145233, "learning_rate": 4.810314812128247e-05, "loss": 0.0095, "step": 14409 }, { "epoch": 1.5622289679098005, "grad_norm": 0.5423434972763062, "learning_rate": 4.8099521253445526e-05, "loss": 0.036, "step": 14410 }, { "epoch": 1.5623373807458805, "grad_norm": 0.026263635605573654, "learning_rate": 4.809589438560859e-05, "loss": 0.0031, "step": 14411 }, { "epoch": 1.56244579358196, "grad_norm": 0.2833125591278076, "learning_rate": 4.8092267517771654e-05, "loss": 0.04, "step": 14412 }, { "epoch": 1.56255420641804, "grad_norm": 0.08094768971204758, "learning_rate": 4.808864064993472e-05, "loss": 0.0069, "step": 14413 }, { "epoch": 1.5626626192541195, "grad_norm": 0.2718392312526703, "learning_rate": 4.808501378209778e-05, "loss": 0.0211, "step": 14414 }, { "epoch": 1.5627710320901995, "grad_norm": 0.03966021165251732, "learning_rate": 4.8081386914260846e-05, "loss": 0.0019, "step": 14415 }, { "epoch": 1.5628794449262793, "grad_norm": 0.5073323845863342, "learning_rate": 4.807776004642391e-05, "loss": 0.0276, "step": 14416 }, { "epoch": 1.562987857762359, "grad_norm": 0.29126405715942383, "learning_rate": 4.8074133178586974e-05, "loss": 0.0184, "step": 14417 }, { "epoch": 1.5630962705984388, "grad_norm": 0.14289814233779907, "learning_rate": 4.807050631075004e-05, "loss": 0.0144, "step": 14418 }, { "epoch": 1.5632046834345186, "grad_norm": 0.23556117713451385, "learning_rate": 4.80668794429131e-05, "loss": 0.0118, "step": 14419 }, { "epoch": 1.5633130962705986, "grad_norm": 0.2470722496509552, "learning_rate": 4.806325257507617e-05, "loss": 0.0263, "step": 14420 }, { "epoch": 1.563421509106678, "grad_norm": 0.34345224499702454, "learning_rate": 4.805962570723923e-05, "loss": 0.0205, "step": 14421 }, { "epoch": 1.563529921942758, "grad_norm": 0.12303606420755386, "learning_rate": 4.8055998839402295e-05, "loss": 0.0104, "step": 14422 }, { "epoch": 1.5636383347788378, "grad_norm": 0.4969273805618286, "learning_rate": 4.805237197156536e-05, "loss": 0.0442, "step": 14423 }, { "epoch": 1.5637467476149176, "grad_norm": 0.6003170013427734, "learning_rate": 4.804874510372842e-05, "loss": 0.0545, "step": 14424 }, { "epoch": 1.5638551604509974, "grad_norm": 0.32851043343544006, "learning_rate": 4.804511823589149e-05, "loss": 0.0395, "step": 14425 }, { "epoch": 1.5639635732870771, "grad_norm": 0.4498192369937897, "learning_rate": 4.804149136805455e-05, "loss": 0.0483, "step": 14426 }, { "epoch": 1.5640719861231571, "grad_norm": 0.15514713525772095, "learning_rate": 4.8037864500217615e-05, "loss": 0.021, "step": 14427 }, { "epoch": 1.5641803989592367, "grad_norm": 0.39887019991874695, "learning_rate": 4.803423763238068e-05, "loss": 0.0209, "step": 14428 }, { "epoch": 1.5642888117953166, "grad_norm": 0.20675426721572876, "learning_rate": 4.8030610764543744e-05, "loss": 0.0249, "step": 14429 }, { "epoch": 1.5643972246313964, "grad_norm": 0.25419825315475464, "learning_rate": 4.802698389670681e-05, "loss": 0.0284, "step": 14430 }, { "epoch": 1.5645056374674762, "grad_norm": 0.18944236636161804, "learning_rate": 4.802335702886987e-05, "loss": 0.0101, "step": 14431 }, { "epoch": 1.564614050303556, "grad_norm": 0.2983509600162506, "learning_rate": 4.8019730161032936e-05, "loss": 0.0391, "step": 14432 }, { "epoch": 1.5647224631396357, "grad_norm": 0.325160413980484, "learning_rate": 4.801610329319599e-05, "loss": 0.0212, "step": 14433 }, { "epoch": 1.5648308759757157, "grad_norm": 0.2628079354763031, "learning_rate": 4.801247642535906e-05, "loss": 0.0225, "step": 14434 }, { "epoch": 1.5649392888117952, "grad_norm": 0.07265707850456238, "learning_rate": 4.800884955752213e-05, "loss": 0.004, "step": 14435 }, { "epoch": 1.5650477016478752, "grad_norm": 0.2793140709400177, "learning_rate": 4.800522268968519e-05, "loss": 0.039, "step": 14436 }, { "epoch": 1.5651561144839548, "grad_norm": 0.09055276215076447, "learning_rate": 4.8001595821848257e-05, "loss": 0.0124, "step": 14437 }, { "epoch": 1.5652645273200347, "grad_norm": 0.6444023847579956, "learning_rate": 4.799796895401132e-05, "loss": 0.0271, "step": 14438 }, { "epoch": 1.5653729401561145, "grad_norm": 0.2972630560398102, "learning_rate": 4.7994342086174385e-05, "loss": 0.0303, "step": 14439 }, { "epoch": 1.5654813529921943, "grad_norm": 0.393004834651947, "learning_rate": 4.799071521833744e-05, "loss": 0.0168, "step": 14440 }, { "epoch": 1.565589765828274, "grad_norm": 0.35922837257385254, "learning_rate": 4.7987088350500506e-05, "loss": 0.0184, "step": 14441 }, { "epoch": 1.5656981786643538, "grad_norm": 1.0107924938201904, "learning_rate": 4.798346148266357e-05, "loss": 0.0581, "step": 14442 }, { "epoch": 1.5658065915004338, "grad_norm": 0.39154812693595886, "learning_rate": 4.7979834614826634e-05, "loss": 0.0376, "step": 14443 }, { "epoch": 1.5659150043365133, "grad_norm": 0.18991589546203613, "learning_rate": 4.7976207746989705e-05, "loss": 0.0138, "step": 14444 }, { "epoch": 1.5660234171725933, "grad_norm": 0.11390591412782669, "learning_rate": 4.797258087915277e-05, "loss": 0.009, "step": 14445 }, { "epoch": 1.566131830008673, "grad_norm": 0.24841876327991486, "learning_rate": 4.7968954011315833e-05, "loss": 0.022, "step": 14446 }, { "epoch": 1.5662402428447528, "grad_norm": 0.2371508926153183, "learning_rate": 4.79653271434789e-05, "loss": 0.0219, "step": 14447 }, { "epoch": 1.5663486556808326, "grad_norm": 0.2069695144891739, "learning_rate": 4.7961700275641955e-05, "loss": 0.0117, "step": 14448 }, { "epoch": 1.5664570685169124, "grad_norm": 0.2399846911430359, "learning_rate": 4.795807340780502e-05, "loss": 0.0254, "step": 14449 }, { "epoch": 1.5665654813529923, "grad_norm": 0.4736616313457489, "learning_rate": 4.795444653996808e-05, "loss": 0.0162, "step": 14450 }, { "epoch": 1.5666738941890719, "grad_norm": 0.06206195428967476, "learning_rate": 4.795081967213115e-05, "loss": 0.005, "step": 14451 }, { "epoch": 1.5667823070251519, "grad_norm": 0.28386396169662476, "learning_rate": 4.794719280429422e-05, "loss": 0.0228, "step": 14452 }, { "epoch": 1.5668907198612314, "grad_norm": 0.19553805887699127, "learning_rate": 4.794356593645728e-05, "loss": 0.0273, "step": 14453 }, { "epoch": 1.5669991326973114, "grad_norm": 0.1362690031528473, "learning_rate": 4.7939939068620346e-05, "loss": 0.0081, "step": 14454 }, { "epoch": 1.5671075455333912, "grad_norm": 0.4965503513813019, "learning_rate": 4.7936312200783404e-05, "loss": 0.0277, "step": 14455 }, { "epoch": 1.567215958369471, "grad_norm": 0.04367240518331528, "learning_rate": 4.793268533294647e-05, "loss": 0.0037, "step": 14456 }, { "epoch": 1.567324371205551, "grad_norm": 0.28513824939727783, "learning_rate": 4.792905846510953e-05, "loss": 0.0176, "step": 14457 }, { "epoch": 1.5674327840416304, "grad_norm": 0.14837947487831116, "learning_rate": 4.7925431597272596e-05, "loss": 0.0243, "step": 14458 }, { "epoch": 1.5675411968777104, "grad_norm": 0.22004063427448273, "learning_rate": 4.792180472943566e-05, "loss": 0.0155, "step": 14459 }, { "epoch": 1.56764960971379, "grad_norm": 0.2267882376909256, "learning_rate": 4.7918177861598724e-05, "loss": 0.0155, "step": 14460 }, { "epoch": 1.56775802254987, "grad_norm": 0.5007798075675964, "learning_rate": 4.7914550993761795e-05, "loss": 0.0424, "step": 14461 }, { "epoch": 1.5678664353859497, "grad_norm": 0.23097224533557892, "learning_rate": 4.791092412592485e-05, "loss": 0.0145, "step": 14462 }, { "epoch": 1.5679748482220295, "grad_norm": 0.18253664672374725, "learning_rate": 4.7907297258087916e-05, "loss": 0.0089, "step": 14463 }, { "epoch": 1.5680832610581092, "grad_norm": 0.10324160009622574, "learning_rate": 4.790367039025098e-05, "loss": 0.013, "step": 14464 }, { "epoch": 1.568191673894189, "grad_norm": 0.33996695280075073, "learning_rate": 4.7900043522414045e-05, "loss": 0.0522, "step": 14465 }, { "epoch": 1.568300086730269, "grad_norm": 0.229701966047287, "learning_rate": 4.789641665457711e-05, "loss": 0.0164, "step": 14466 }, { "epoch": 1.5684084995663485, "grad_norm": 0.28219926357269287, "learning_rate": 4.789278978674017e-05, "loss": 0.0273, "step": 14467 }, { "epoch": 1.5685169124024285, "grad_norm": 0.37672746181488037, "learning_rate": 4.788916291890324e-05, "loss": 0.0335, "step": 14468 }, { "epoch": 1.5686253252385083, "grad_norm": 0.3868732154369354, "learning_rate": 4.78855360510663e-05, "loss": 0.0386, "step": 14469 }, { "epoch": 1.568733738074588, "grad_norm": 0.4672487676143646, "learning_rate": 4.7881909183229365e-05, "loss": 0.0466, "step": 14470 }, { "epoch": 1.5688421509106678, "grad_norm": 0.37216681241989136, "learning_rate": 4.787828231539243e-05, "loss": 0.0546, "step": 14471 }, { "epoch": 1.5689505637467476, "grad_norm": 0.1751866191625595, "learning_rate": 4.7874655447555493e-05, "loss": 0.0119, "step": 14472 }, { "epoch": 1.5690589765828276, "grad_norm": 0.42257532477378845, "learning_rate": 4.787102857971856e-05, "loss": 0.0398, "step": 14473 }, { "epoch": 1.569167389418907, "grad_norm": 0.07500217109918594, "learning_rate": 4.786740171188162e-05, "loss": 0.0031, "step": 14474 }, { "epoch": 1.569275802254987, "grad_norm": 0.020185895264148712, "learning_rate": 4.7863774844044686e-05, "loss": 0.0014, "step": 14475 }, { "epoch": 1.5693842150910666, "grad_norm": 0.4978158175945282, "learning_rate": 4.786014797620775e-05, "loss": 0.0564, "step": 14476 }, { "epoch": 1.5694926279271466, "grad_norm": 0.06429732590913773, "learning_rate": 4.7856521108370814e-05, "loss": 0.0043, "step": 14477 }, { "epoch": 1.5696010407632264, "grad_norm": 0.07470114529132843, "learning_rate": 4.785289424053388e-05, "loss": 0.0092, "step": 14478 }, { "epoch": 1.5697094535993061, "grad_norm": 0.36550554633140564, "learning_rate": 4.784926737269694e-05, "loss": 0.024, "step": 14479 }, { "epoch": 1.569817866435386, "grad_norm": 0.3725832402706146, "learning_rate": 4.7845640504860006e-05, "loss": 0.0036, "step": 14480 }, { "epoch": 1.5699262792714657, "grad_norm": 0.2232307642698288, "learning_rate": 4.784201363702307e-05, "loss": 0.0161, "step": 14481 }, { "epoch": 1.5700346921075456, "grad_norm": 0.25527408719062805, "learning_rate": 4.7838386769186134e-05, "loss": 0.0163, "step": 14482 }, { "epoch": 1.5701431049436252, "grad_norm": 0.3952370285987854, "learning_rate": 4.78347599013492e-05, "loss": 0.0172, "step": 14483 }, { "epoch": 1.5702515177797052, "grad_norm": 0.1873122602701187, "learning_rate": 4.783113303351226e-05, "loss": 0.0134, "step": 14484 }, { "epoch": 1.570359930615785, "grad_norm": 0.6370723247528076, "learning_rate": 4.782750616567532e-05, "loss": 0.0393, "step": 14485 }, { "epoch": 1.5704683434518647, "grad_norm": 0.21037492156028748, "learning_rate": 4.7823879297838384e-05, "loss": 0.0091, "step": 14486 }, { "epoch": 1.5705767562879445, "grad_norm": 0.22034454345703125, "learning_rate": 4.7820252430001455e-05, "loss": 0.0209, "step": 14487 }, { "epoch": 1.5706851691240242, "grad_norm": 0.3101862967014313, "learning_rate": 4.781662556216452e-05, "loss": 0.0176, "step": 14488 }, { "epoch": 1.5707935819601042, "grad_norm": 0.4465714991092682, "learning_rate": 4.781299869432758e-05, "loss": 0.0274, "step": 14489 }, { "epoch": 1.5709019947961838, "grad_norm": 0.451453298330307, "learning_rate": 4.780937182649065e-05, "loss": 0.0064, "step": 14490 }, { "epoch": 1.5710104076322637, "grad_norm": 0.08936719596385956, "learning_rate": 4.780574495865371e-05, "loss": 0.0061, "step": 14491 }, { "epoch": 1.5711188204683435, "grad_norm": 0.21958760917186737, "learning_rate": 4.780211809081677e-05, "loss": 0.0202, "step": 14492 }, { "epoch": 1.5712272333044233, "grad_norm": 0.09569903463125229, "learning_rate": 4.779849122297983e-05, "loss": 0.0099, "step": 14493 }, { "epoch": 1.571335646140503, "grad_norm": 0.24800074100494385, "learning_rate": 4.77948643551429e-05, "loss": 0.0116, "step": 14494 }, { "epoch": 1.5714440589765828, "grad_norm": 0.39495450258255005, "learning_rate": 4.779123748730597e-05, "loss": 0.0601, "step": 14495 }, { "epoch": 1.5715524718126628, "grad_norm": 0.19668744504451752, "learning_rate": 4.778761061946903e-05, "loss": 0.012, "step": 14496 }, { "epoch": 1.5716608846487423, "grad_norm": 0.7685051560401917, "learning_rate": 4.7783983751632096e-05, "loss": 0.0643, "step": 14497 }, { "epoch": 1.5717692974848223, "grad_norm": 0.4863896369934082, "learning_rate": 4.778035688379516e-05, "loss": 0.0452, "step": 14498 }, { "epoch": 1.5718777103209018, "grad_norm": 0.14840060472488403, "learning_rate": 4.7776730015958224e-05, "loss": 0.0147, "step": 14499 }, { "epoch": 1.5719861231569818, "grad_norm": 0.26238250732421875, "learning_rate": 4.777310314812128e-05, "loss": 0.0381, "step": 14500 }, { "epoch": 1.5720945359930616, "grad_norm": 0.29456740617752075, "learning_rate": 4.7769476280284346e-05, "loss": 0.0206, "step": 14501 }, { "epoch": 1.5722029488291414, "grad_norm": 0.29670006036758423, "learning_rate": 4.776584941244741e-05, "loss": 0.0203, "step": 14502 }, { "epoch": 1.5723113616652211, "grad_norm": 0.45747387409210205, "learning_rate": 4.7762222544610474e-05, "loss": 0.0319, "step": 14503 }, { "epoch": 1.5724197745013009, "grad_norm": 0.18350239098072052, "learning_rate": 4.7758595676773545e-05, "loss": 0.0149, "step": 14504 }, { "epoch": 1.5725281873373809, "grad_norm": 0.1722545325756073, "learning_rate": 4.775496880893661e-05, "loss": 0.0186, "step": 14505 }, { "epoch": 1.5726366001734604, "grad_norm": 0.2735808789730072, "learning_rate": 4.775134194109967e-05, "loss": 0.0206, "step": 14506 }, { "epoch": 1.5727450130095404, "grad_norm": 0.1596061885356903, "learning_rate": 4.774771507326273e-05, "loss": 0.0166, "step": 14507 }, { "epoch": 1.5728534258456202, "grad_norm": 0.29413843154907227, "learning_rate": 4.7744088205425794e-05, "loss": 0.0068, "step": 14508 }, { "epoch": 1.5729618386817, "grad_norm": 0.30131638050079346, "learning_rate": 4.774046133758886e-05, "loss": 0.0184, "step": 14509 }, { "epoch": 1.5730702515177797, "grad_norm": 0.3254764676094055, "learning_rate": 4.773683446975192e-05, "loss": 0.0265, "step": 14510 }, { "epoch": 1.5731786643538594, "grad_norm": 0.4388393461704254, "learning_rate": 4.773320760191499e-05, "loss": 0.0222, "step": 14511 }, { "epoch": 1.5732870771899394, "grad_norm": 0.1255292445421219, "learning_rate": 4.772958073407805e-05, "loss": 0.0088, "step": 14512 }, { "epoch": 1.573395490026019, "grad_norm": 0.13221070170402527, "learning_rate": 4.772595386624112e-05, "loss": 0.0091, "step": 14513 }, { "epoch": 1.573503902862099, "grad_norm": 0.27825334668159485, "learning_rate": 4.772232699840418e-05, "loss": 0.0307, "step": 14514 }, { "epoch": 1.5736123156981785, "grad_norm": 0.17160899937152863, "learning_rate": 4.771870013056724e-05, "loss": 0.0116, "step": 14515 }, { "epoch": 1.5737207285342585, "grad_norm": 0.189419224858284, "learning_rate": 4.771507326273031e-05, "loss": 0.0215, "step": 14516 }, { "epoch": 1.5738291413703382, "grad_norm": 0.7247928380966187, "learning_rate": 4.771144639489337e-05, "loss": 0.0402, "step": 14517 }, { "epoch": 1.573937554206418, "grad_norm": 0.21541929244995117, "learning_rate": 4.7707819527056435e-05, "loss": 0.0246, "step": 14518 }, { "epoch": 1.574045967042498, "grad_norm": 0.4760742783546448, "learning_rate": 4.77041926592195e-05, "loss": 0.0805, "step": 14519 }, { "epoch": 1.5741543798785775, "grad_norm": 0.03374249115586281, "learning_rate": 4.7700565791382564e-05, "loss": 0.0024, "step": 14520 }, { "epoch": 1.5742627927146575, "grad_norm": 0.4321081042289734, "learning_rate": 4.769693892354563e-05, "loss": 0.0334, "step": 14521 }, { "epoch": 1.574371205550737, "grad_norm": 0.13966552913188934, "learning_rate": 4.769331205570869e-05, "loss": 0.0129, "step": 14522 }, { "epoch": 1.574479618386817, "grad_norm": 0.17625966668128967, "learning_rate": 4.7689685187871756e-05, "loss": 0.0106, "step": 14523 }, { "epoch": 1.5745880312228968, "grad_norm": 0.09576921910047531, "learning_rate": 4.768605832003482e-05, "loss": 0.0065, "step": 14524 }, { "epoch": 1.5746964440589766, "grad_norm": 0.26822617650032043, "learning_rate": 4.7682431452197884e-05, "loss": 0.0247, "step": 14525 }, { "epoch": 1.5748048568950563, "grad_norm": 0.49033674597740173, "learning_rate": 4.767880458436095e-05, "loss": 0.0154, "step": 14526 }, { "epoch": 1.574913269731136, "grad_norm": 0.0838758796453476, "learning_rate": 4.767517771652401e-05, "loss": 0.0048, "step": 14527 }, { "epoch": 1.575021682567216, "grad_norm": 0.15947313606739044, "learning_rate": 4.7671550848687077e-05, "loss": 0.0063, "step": 14528 }, { "epoch": 1.5751300954032956, "grad_norm": 0.38022464513778687, "learning_rate": 4.766792398085014e-05, "loss": 0.0407, "step": 14529 }, { "epoch": 1.5752385082393756, "grad_norm": 0.5308032035827637, "learning_rate": 4.7664297113013205e-05, "loss": 0.0277, "step": 14530 }, { "epoch": 1.5753469210754554, "grad_norm": 0.37086060643196106, "learning_rate": 4.766067024517627e-05, "loss": 0.0296, "step": 14531 }, { "epoch": 1.5754553339115351, "grad_norm": 0.4614492654800415, "learning_rate": 4.765704337733933e-05, "loss": 0.0574, "step": 14532 }, { "epoch": 1.575563746747615, "grad_norm": 0.1997620016336441, "learning_rate": 4.76534165095024e-05, "loss": 0.0097, "step": 14533 }, { "epoch": 1.5756721595836947, "grad_norm": 0.05307130143046379, "learning_rate": 4.764978964166546e-05, "loss": 0.0038, "step": 14534 }, { "epoch": 1.5757805724197746, "grad_norm": 0.47032082080841064, "learning_rate": 4.7646162773828525e-05, "loss": 0.0335, "step": 14535 }, { "epoch": 1.5758889852558542, "grad_norm": 0.6352813243865967, "learning_rate": 4.764253590599159e-05, "loss": 0.0892, "step": 14536 }, { "epoch": 1.5759973980919342, "grad_norm": 0.8298497796058655, "learning_rate": 4.763890903815465e-05, "loss": 0.0413, "step": 14537 }, { "epoch": 1.5761058109280137, "grad_norm": 0.544321596622467, "learning_rate": 4.763528217031771e-05, "loss": 0.035, "step": 14538 }, { "epoch": 1.5762142237640937, "grad_norm": 0.4505211114883423, "learning_rate": 4.763165530248078e-05, "loss": 0.0224, "step": 14539 }, { "epoch": 1.5763226366001735, "grad_norm": 0.26363420486450195, "learning_rate": 4.7628028434643846e-05, "loss": 0.0331, "step": 14540 }, { "epoch": 1.5764310494362532, "grad_norm": 0.04818334802985191, "learning_rate": 4.762440156680691e-05, "loss": 0.001, "step": 14541 }, { "epoch": 1.576539462272333, "grad_norm": 0.35798099637031555, "learning_rate": 4.7620774698969974e-05, "loss": 0.0241, "step": 14542 }, { "epoch": 1.5766478751084128, "grad_norm": 0.11273685097694397, "learning_rate": 4.761714783113304e-05, "loss": 0.0073, "step": 14543 }, { "epoch": 1.5767562879444927, "grad_norm": 0.159731924533844, "learning_rate": 4.7613520963296095e-05, "loss": 0.0106, "step": 14544 }, { "epoch": 1.5768647007805723, "grad_norm": 0.16477200388908386, "learning_rate": 4.760989409545916e-05, "loss": 0.0175, "step": 14545 }, { "epoch": 1.5769731136166523, "grad_norm": 0.7902315855026245, "learning_rate": 4.7606267227622224e-05, "loss": 0.0273, "step": 14546 }, { "epoch": 1.577081526452732, "grad_norm": 0.256433367729187, "learning_rate": 4.7602640359785295e-05, "loss": 0.02, "step": 14547 }, { "epoch": 1.5771899392888118, "grad_norm": 0.16363975405693054, "learning_rate": 4.759901349194836e-05, "loss": 0.0131, "step": 14548 }, { "epoch": 1.5772983521248916, "grad_norm": 0.2350998967885971, "learning_rate": 4.759538662411142e-05, "loss": 0.0151, "step": 14549 }, { "epoch": 1.5774067649609713, "grad_norm": 0.10245131701231003, "learning_rate": 4.759175975627449e-05, "loss": 0.0087, "step": 14550 }, { "epoch": 1.5775151777970513, "grad_norm": 0.06308174133300781, "learning_rate": 4.758813288843755e-05, "loss": 0.0025, "step": 14551 }, { "epoch": 1.5776235906331308, "grad_norm": 0.24672375619411469, "learning_rate": 4.758450602060061e-05, "loss": 0.012, "step": 14552 }, { "epoch": 1.5777320034692108, "grad_norm": 0.3085417151451111, "learning_rate": 4.758087915276367e-05, "loss": 0.0161, "step": 14553 }, { "epoch": 1.5778404163052906, "grad_norm": 0.21545207500457764, "learning_rate": 4.7577252284926736e-05, "loss": 0.0253, "step": 14554 }, { "epoch": 1.5779488291413704, "grad_norm": 0.09191811084747314, "learning_rate": 4.75736254170898e-05, "loss": 0.0058, "step": 14555 }, { "epoch": 1.5780572419774501, "grad_norm": 0.5434587597846985, "learning_rate": 4.756999854925287e-05, "loss": 0.0208, "step": 14556 }, { "epoch": 1.5781656548135299, "grad_norm": 0.24127843976020813, "learning_rate": 4.7566371681415936e-05, "loss": 0.0125, "step": 14557 }, { "epoch": 1.5782740676496099, "grad_norm": 0.5890251994132996, "learning_rate": 4.7562744813579e-05, "loss": 0.0232, "step": 14558 }, { "epoch": 1.5783824804856894, "grad_norm": 0.3348865509033203, "learning_rate": 4.755911794574206e-05, "loss": 0.0288, "step": 14559 }, { "epoch": 1.5784908933217694, "grad_norm": 0.47746625542640686, "learning_rate": 4.755549107790512e-05, "loss": 0.0131, "step": 14560 }, { "epoch": 1.578599306157849, "grad_norm": 0.30000045895576477, "learning_rate": 4.7551864210068185e-05, "loss": 0.0118, "step": 14561 }, { "epoch": 1.578707718993929, "grad_norm": 0.41583186388015747, "learning_rate": 4.754823734223125e-05, "loss": 0.0483, "step": 14562 }, { "epoch": 1.5788161318300087, "grad_norm": 0.14722539484500885, "learning_rate": 4.7544610474394313e-05, "loss": 0.0106, "step": 14563 }, { "epoch": 1.5789245446660884, "grad_norm": 0.816807746887207, "learning_rate": 4.754098360655738e-05, "loss": 0.0398, "step": 14564 }, { "epoch": 1.5790329575021682, "grad_norm": 0.8790367245674133, "learning_rate": 4.753735673872045e-05, "loss": 0.0457, "step": 14565 }, { "epoch": 1.579141370338248, "grad_norm": 0.5899813175201416, "learning_rate": 4.7533729870883506e-05, "loss": 0.0448, "step": 14566 }, { "epoch": 1.579249783174328, "grad_norm": 0.38794010877609253, "learning_rate": 4.753010300304657e-05, "loss": 0.0246, "step": 14567 }, { "epoch": 1.5793581960104075, "grad_norm": 0.09156286716461182, "learning_rate": 4.7526476135209634e-05, "loss": 0.0084, "step": 14568 }, { "epoch": 1.5794666088464875, "grad_norm": 0.23506751656532288, "learning_rate": 4.75228492673727e-05, "loss": 0.0232, "step": 14569 }, { "epoch": 1.5795750216825672, "grad_norm": 0.9008064270019531, "learning_rate": 4.751922239953576e-05, "loss": 0.0532, "step": 14570 }, { "epoch": 1.579683434518647, "grad_norm": 0.1823417693376541, "learning_rate": 4.7515595531698826e-05, "loss": 0.0257, "step": 14571 }, { "epoch": 1.5797918473547268, "grad_norm": 0.3186589181423187, "learning_rate": 4.751196866386189e-05, "loss": 0.0236, "step": 14572 }, { "epoch": 1.5799002601908065, "grad_norm": 0.05025865137577057, "learning_rate": 4.7508341796024954e-05, "loss": 0.0035, "step": 14573 }, { "epoch": 1.5800086730268865, "grad_norm": 0.38789960741996765, "learning_rate": 4.750471492818802e-05, "loss": 0.021, "step": 14574 }, { "epoch": 1.580117085862966, "grad_norm": 0.21725738048553467, "learning_rate": 4.750108806035108e-05, "loss": 0.0178, "step": 14575 }, { "epoch": 1.580225498699046, "grad_norm": 0.10745543986558914, "learning_rate": 4.749746119251415e-05, "loss": 0.0062, "step": 14576 }, { "epoch": 1.5803339115351258, "grad_norm": 0.5659047961235046, "learning_rate": 4.749383432467721e-05, "loss": 0.0438, "step": 14577 }, { "epoch": 1.5804423243712056, "grad_norm": 1.114370584487915, "learning_rate": 4.7490207456840275e-05, "loss": 0.0639, "step": 14578 }, { "epoch": 1.5805507372072853, "grad_norm": 0.2481527179479599, "learning_rate": 4.748658058900334e-05, "loss": 0.0118, "step": 14579 }, { "epoch": 1.580659150043365, "grad_norm": 0.21344546973705292, "learning_rate": 4.74829537211664e-05, "loss": 0.0086, "step": 14580 }, { "epoch": 1.580767562879445, "grad_norm": 0.09539800137281418, "learning_rate": 4.747932685332947e-05, "loss": 0.0078, "step": 14581 }, { "epoch": 1.5808759757155246, "grad_norm": 0.31863775849342346, "learning_rate": 4.747569998549253e-05, "loss": 0.0153, "step": 14582 }, { "epoch": 1.5809843885516046, "grad_norm": 0.09229815751314163, "learning_rate": 4.7472073117655596e-05, "loss": 0.006, "step": 14583 }, { "epoch": 1.5810928013876842, "grad_norm": 0.41184109449386597, "learning_rate": 4.746844624981866e-05, "loss": 0.0267, "step": 14584 }, { "epoch": 1.5812012142237641, "grad_norm": 0.26413318514823914, "learning_rate": 4.7464819381981724e-05, "loss": 0.0193, "step": 14585 }, { "epoch": 1.581309627059844, "grad_norm": 0.19975703954696655, "learning_rate": 4.746119251414479e-05, "loss": 0.0196, "step": 14586 }, { "epoch": 1.5814180398959237, "grad_norm": 0.7266910672187805, "learning_rate": 4.745756564630785e-05, "loss": 0.0814, "step": 14587 }, { "epoch": 1.5815264527320034, "grad_norm": 0.20547613501548767, "learning_rate": 4.7453938778470916e-05, "loss": 0.016, "step": 14588 }, { "epoch": 1.5816348655680832, "grad_norm": 0.13782130181789398, "learning_rate": 4.7450311910633973e-05, "loss": 0.0111, "step": 14589 }, { "epoch": 1.5817432784041632, "grad_norm": 0.03481706976890564, "learning_rate": 4.7446685042797044e-05, "loss": 0.0011, "step": 14590 }, { "epoch": 1.5818516912402427, "grad_norm": 0.33175230026245117, "learning_rate": 4.744305817496011e-05, "loss": 0.0198, "step": 14591 }, { "epoch": 1.5819601040763227, "grad_norm": 0.021510178223252296, "learning_rate": 4.743943130712317e-05, "loss": 0.0013, "step": 14592 }, { "epoch": 1.5820685169124025, "grad_norm": 0.4057615101337433, "learning_rate": 4.7435804439286237e-05, "loss": 0.0291, "step": 14593 }, { "epoch": 1.5821769297484822, "grad_norm": 0.2881619930267334, "learning_rate": 4.74321775714493e-05, "loss": 0.0249, "step": 14594 }, { "epoch": 1.582285342584562, "grad_norm": 1.3596616983413696, "learning_rate": 4.7428550703612365e-05, "loss": 0.0816, "step": 14595 }, { "epoch": 1.5823937554206418, "grad_norm": 0.3232288062572479, "learning_rate": 4.742492383577543e-05, "loss": 0.0164, "step": 14596 }, { "epoch": 1.5825021682567217, "grad_norm": 0.07170556485652924, "learning_rate": 4.7421296967938486e-05, "loss": 0.0046, "step": 14597 }, { "epoch": 1.5826105810928013, "grad_norm": 0.5765684247016907, "learning_rate": 4.741767010010155e-05, "loss": 0.061, "step": 14598 }, { "epoch": 1.5827189939288813, "grad_norm": 0.904475748538971, "learning_rate": 4.741404323226462e-05, "loss": 0.0979, "step": 14599 }, { "epoch": 1.5828274067649608, "grad_norm": 0.11861812323331833, "learning_rate": 4.7410416364427685e-05, "loss": 0.0248, "step": 14600 }, { "epoch": 1.5829358196010408, "grad_norm": 0.34334346652030945, "learning_rate": 4.740678949659075e-05, "loss": 0.0172, "step": 14601 }, { "epoch": 1.5830442324371206, "grad_norm": 0.28205201029777527, "learning_rate": 4.7403162628753814e-05, "loss": 0.0267, "step": 14602 }, { "epoch": 1.5831526452732003, "grad_norm": 0.4066820442676544, "learning_rate": 4.739953576091688e-05, "loss": 0.026, "step": 14603 }, { "epoch": 1.58326105810928, "grad_norm": 0.5028590559959412, "learning_rate": 4.7395908893079935e-05, "loss": 0.0259, "step": 14604 }, { "epoch": 1.5833694709453598, "grad_norm": 0.2538439631462097, "learning_rate": 4.7392282025243e-05, "loss": 0.012, "step": 14605 }, { "epoch": 1.5834778837814398, "grad_norm": 0.21225838363170624, "learning_rate": 4.738865515740606e-05, "loss": 0.0215, "step": 14606 }, { "epoch": 1.5835862966175194, "grad_norm": 0.41438329219818115, "learning_rate": 4.738502828956913e-05, "loss": 0.0123, "step": 14607 }, { "epoch": 1.5836947094535994, "grad_norm": 0.5561299920082092, "learning_rate": 4.73814014217322e-05, "loss": 0.0223, "step": 14608 }, { "epoch": 1.5838031222896791, "grad_norm": 0.3345211148262024, "learning_rate": 4.737777455389526e-05, "loss": 0.0175, "step": 14609 }, { "epoch": 1.5839115351257589, "grad_norm": 0.08813458681106567, "learning_rate": 4.7374147686058326e-05, "loss": 0.0072, "step": 14610 }, { "epoch": 1.5840199479618386, "grad_norm": 0.28671199083328247, "learning_rate": 4.7370520818221384e-05, "loss": 0.0139, "step": 14611 }, { "epoch": 1.5841283607979184, "grad_norm": 0.18875813484191895, "learning_rate": 4.736689395038445e-05, "loss": 0.012, "step": 14612 }, { "epoch": 1.5842367736339984, "grad_norm": 0.3894781768321991, "learning_rate": 4.736326708254751e-05, "loss": 0.0271, "step": 14613 }, { "epoch": 1.584345186470078, "grad_norm": 0.35721367597579956, "learning_rate": 4.7359640214710576e-05, "loss": 0.0617, "step": 14614 }, { "epoch": 1.584453599306158, "grad_norm": 0.11387057602405548, "learning_rate": 4.735601334687364e-05, "loss": 0.0079, "step": 14615 }, { "epoch": 1.5845620121422377, "grad_norm": 0.3843002915382385, "learning_rate": 4.735238647903671e-05, "loss": 0.0353, "step": 14616 }, { "epoch": 1.5846704249783174, "grad_norm": 0.19522453844547272, "learning_rate": 4.7348759611199775e-05, "loss": 0.0243, "step": 14617 }, { "epoch": 1.5847788378143972, "grad_norm": 0.23092080652713776, "learning_rate": 4.734513274336283e-05, "loss": 0.0178, "step": 14618 }, { "epoch": 1.584887250650477, "grad_norm": 0.4258333146572113, "learning_rate": 4.7341505875525897e-05, "loss": 0.0338, "step": 14619 }, { "epoch": 1.584995663486557, "grad_norm": 0.32798367738723755, "learning_rate": 4.733787900768896e-05, "loss": 0.0198, "step": 14620 }, { "epoch": 1.5851040763226365, "grad_norm": 0.46000754833221436, "learning_rate": 4.7334252139852025e-05, "loss": 0.0384, "step": 14621 }, { "epoch": 1.5852124891587165, "grad_norm": 0.2924576699733734, "learning_rate": 4.733062527201509e-05, "loss": 0.0214, "step": 14622 }, { "epoch": 1.585320901994796, "grad_norm": 0.23390311002731323, "learning_rate": 4.732699840417815e-05, "loss": 0.0288, "step": 14623 }, { "epoch": 1.585429314830876, "grad_norm": 0.39118680357933044, "learning_rate": 4.732337153634122e-05, "loss": 0.0371, "step": 14624 }, { "epoch": 1.5855377276669558, "grad_norm": 0.1335793286561966, "learning_rate": 4.731974466850428e-05, "loss": 0.0056, "step": 14625 }, { "epoch": 1.5856461405030355, "grad_norm": 0.5098686814308167, "learning_rate": 4.7316117800667345e-05, "loss": 0.0225, "step": 14626 }, { "epoch": 1.5857545533391153, "grad_norm": 0.1950552612543106, "learning_rate": 4.731249093283041e-05, "loss": 0.0173, "step": 14627 }, { "epoch": 1.585862966175195, "grad_norm": 0.13680389523506165, "learning_rate": 4.7308864064993473e-05, "loss": 0.0084, "step": 14628 }, { "epoch": 1.585971379011275, "grad_norm": 0.4139145016670227, "learning_rate": 4.730523719715654e-05, "loss": 0.0296, "step": 14629 }, { "epoch": 1.5860797918473546, "grad_norm": 0.020595254376530647, "learning_rate": 4.73016103293196e-05, "loss": 0.0015, "step": 14630 }, { "epoch": 1.5861882046834346, "grad_norm": 0.3499215543270111, "learning_rate": 4.7297983461482666e-05, "loss": 0.0137, "step": 14631 }, { "epoch": 1.5862966175195143, "grad_norm": 0.30000951886177063, "learning_rate": 4.729435659364573e-05, "loss": 0.0187, "step": 14632 }, { "epoch": 1.586405030355594, "grad_norm": 0.37696483731269836, "learning_rate": 4.7290729725808794e-05, "loss": 0.0208, "step": 14633 }, { "epoch": 1.5865134431916739, "grad_norm": 0.6324228644371033, "learning_rate": 4.728710285797186e-05, "loss": 0.0349, "step": 14634 }, { "epoch": 1.5866218560277536, "grad_norm": 0.28344711661338806, "learning_rate": 4.728347599013492e-05, "loss": 0.0073, "step": 14635 }, { "epoch": 1.5867302688638336, "grad_norm": 0.40717872977256775, "learning_rate": 4.7279849122297986e-05, "loss": 0.0363, "step": 14636 }, { "epoch": 1.5868386816999132, "grad_norm": 0.7517009973526001, "learning_rate": 4.727622225446105e-05, "loss": 0.0328, "step": 14637 }, { "epoch": 1.5869470945359931, "grad_norm": 0.4776064157485962, "learning_rate": 4.7272595386624115e-05, "loss": 0.0275, "step": 14638 }, { "epoch": 1.587055507372073, "grad_norm": 0.504922091960907, "learning_rate": 4.726896851878718e-05, "loss": 0.0751, "step": 14639 }, { "epoch": 1.5871639202081527, "grad_norm": 0.44332921504974365, "learning_rate": 4.726534165095024e-05, "loss": 0.0355, "step": 14640 }, { "epoch": 1.5872723330442324, "grad_norm": 0.13575460016727448, "learning_rate": 4.72617147831133e-05, "loss": 0.0119, "step": 14641 }, { "epoch": 1.5873807458803122, "grad_norm": 0.33994704484939575, "learning_rate": 4.725808791527637e-05, "loss": 0.041, "step": 14642 }, { "epoch": 1.5874891587163922, "grad_norm": 0.23251447081565857, "learning_rate": 4.7254461047439435e-05, "loss": 0.0125, "step": 14643 }, { "epoch": 1.5875975715524717, "grad_norm": 0.33281025290489197, "learning_rate": 4.72508341796025e-05, "loss": 0.0363, "step": 14644 }, { "epoch": 1.5877059843885517, "grad_norm": 0.3039162755012512, "learning_rate": 4.724720731176556e-05, "loss": 0.0181, "step": 14645 }, { "epoch": 1.5878143972246312, "grad_norm": 0.594876229763031, "learning_rate": 4.724358044392863e-05, "loss": 0.0314, "step": 14646 }, { "epoch": 1.5879228100607112, "grad_norm": 0.42287206649780273, "learning_rate": 4.723995357609169e-05, "loss": 0.0494, "step": 14647 }, { "epoch": 1.588031222896791, "grad_norm": 0.4950776994228363, "learning_rate": 4.7236326708254756e-05, "loss": 0.0381, "step": 14648 }, { "epoch": 1.5881396357328708, "grad_norm": 0.27415451407432556, "learning_rate": 4.723269984041781e-05, "loss": 0.0225, "step": 14649 }, { "epoch": 1.5882480485689505, "grad_norm": 0.033193472772836685, "learning_rate": 4.722907297258088e-05, "loss": 0.0019, "step": 14650 }, { "epoch": 1.5883564614050303, "grad_norm": 0.3924965560436249, "learning_rate": 4.722544610474395e-05, "loss": 0.0755, "step": 14651 }, { "epoch": 1.5884648742411103, "grad_norm": 0.2899896800518036, "learning_rate": 4.722181923690701e-05, "loss": 0.0306, "step": 14652 }, { "epoch": 1.5885732870771898, "grad_norm": 0.3187181055545807, "learning_rate": 4.7218192369070076e-05, "loss": 0.0365, "step": 14653 }, { "epoch": 1.5886816999132698, "grad_norm": 0.49040427803993225, "learning_rate": 4.721456550123314e-05, "loss": 0.0291, "step": 14654 }, { "epoch": 1.5887901127493496, "grad_norm": 0.21825051307678223, "learning_rate": 4.7210938633396204e-05, "loss": 0.0291, "step": 14655 }, { "epoch": 1.5888985255854293, "grad_norm": 0.0858411192893982, "learning_rate": 4.720731176555926e-05, "loss": 0.0084, "step": 14656 }, { "epoch": 1.589006938421509, "grad_norm": 0.22939392924308777, "learning_rate": 4.7203684897722326e-05, "loss": 0.0167, "step": 14657 }, { "epoch": 1.5891153512575888, "grad_norm": 0.1540956199169159, "learning_rate": 4.720005802988539e-05, "loss": 0.0138, "step": 14658 }, { "epoch": 1.5892237640936688, "grad_norm": 0.6196388006210327, "learning_rate": 4.7196431162048454e-05, "loss": 0.0496, "step": 14659 }, { "epoch": 1.5893321769297484, "grad_norm": 0.2974274456501007, "learning_rate": 4.7192804294211525e-05, "loss": 0.029, "step": 14660 }, { "epoch": 1.5894405897658284, "grad_norm": 0.07376693934202194, "learning_rate": 4.718917742637459e-05, "loss": 0.0039, "step": 14661 }, { "epoch": 1.589549002601908, "grad_norm": 0.15271557867527008, "learning_rate": 4.718555055853765e-05, "loss": 0.0139, "step": 14662 }, { "epoch": 1.5896574154379879, "grad_norm": 0.38732391595840454, "learning_rate": 4.718192369070071e-05, "loss": 0.015, "step": 14663 }, { "epoch": 1.5897658282740676, "grad_norm": 0.03505152463912964, "learning_rate": 4.7178296822863775e-05, "loss": 0.0033, "step": 14664 }, { "epoch": 1.5898742411101474, "grad_norm": 0.5370749831199646, "learning_rate": 4.717466995502684e-05, "loss": 0.0481, "step": 14665 }, { "epoch": 1.5899826539462274, "grad_norm": 0.1629495769739151, "learning_rate": 4.71710430871899e-05, "loss": 0.0129, "step": 14666 }, { "epoch": 1.590091066782307, "grad_norm": 0.1776508241891861, "learning_rate": 4.716741621935297e-05, "loss": 0.0388, "step": 14667 }, { "epoch": 1.590199479618387, "grad_norm": 0.5795648097991943, "learning_rate": 4.716378935151604e-05, "loss": 0.0346, "step": 14668 }, { "epoch": 1.5903078924544665, "grad_norm": 0.19400975108146667, "learning_rate": 4.71601624836791e-05, "loss": 0.0139, "step": 14669 }, { "epoch": 1.5904163052905465, "grad_norm": 1.132002830505371, "learning_rate": 4.715653561584216e-05, "loss": 0.0239, "step": 14670 }, { "epoch": 1.5905247181266262, "grad_norm": 0.27369463443756104, "learning_rate": 4.715290874800522e-05, "loss": 0.0323, "step": 14671 }, { "epoch": 1.590633130962706, "grad_norm": 0.18400780856609344, "learning_rate": 4.714928188016829e-05, "loss": 0.018, "step": 14672 }, { "epoch": 1.5907415437987857, "grad_norm": 0.25515925884246826, "learning_rate": 4.714565501233135e-05, "loss": 0.0181, "step": 14673 }, { "epoch": 1.5908499566348655, "grad_norm": 0.26725712418556213, "learning_rate": 4.7142028144494416e-05, "loss": 0.0321, "step": 14674 }, { "epoch": 1.5909583694709455, "grad_norm": 0.2159891426563263, "learning_rate": 4.713840127665748e-05, "loss": 0.0193, "step": 14675 }, { "epoch": 1.591066782307025, "grad_norm": 0.22785614430904388, "learning_rate": 4.7134774408820544e-05, "loss": 0.0234, "step": 14676 }, { "epoch": 1.591175195143105, "grad_norm": 0.5139725208282471, "learning_rate": 4.713114754098361e-05, "loss": 0.0505, "step": 14677 }, { "epoch": 1.5912836079791848, "grad_norm": 0.2527824938297272, "learning_rate": 4.712752067314667e-05, "loss": 0.0378, "step": 14678 }, { "epoch": 1.5913920208152645, "grad_norm": 0.20213568210601807, "learning_rate": 4.7123893805309736e-05, "loss": 0.0088, "step": 14679 }, { "epoch": 1.5915004336513443, "grad_norm": 0.12883836030960083, "learning_rate": 4.71202669374728e-05, "loss": 0.0078, "step": 14680 }, { "epoch": 1.591608846487424, "grad_norm": 0.49458128213882446, "learning_rate": 4.7116640069635864e-05, "loss": 0.0569, "step": 14681 }, { "epoch": 1.591717259323504, "grad_norm": 0.4295879006385803, "learning_rate": 4.711301320179893e-05, "loss": 0.0231, "step": 14682 }, { "epoch": 1.5918256721595836, "grad_norm": 0.40347591042518616, "learning_rate": 4.710938633396199e-05, "loss": 0.024, "step": 14683 }, { "epoch": 1.5919340849956636, "grad_norm": 0.3948805034160614, "learning_rate": 4.7105759466125057e-05, "loss": 0.034, "step": 14684 }, { "epoch": 1.5920424978317431, "grad_norm": 0.836359977722168, "learning_rate": 4.710213259828812e-05, "loss": 0.045, "step": 14685 }, { "epoch": 1.592150910667823, "grad_norm": 0.4797109067440033, "learning_rate": 4.7098505730451185e-05, "loss": 0.0378, "step": 14686 }, { "epoch": 1.5922593235039029, "grad_norm": 0.4725186824798584, "learning_rate": 4.709487886261425e-05, "loss": 0.0353, "step": 14687 }, { "epoch": 1.5923677363399826, "grad_norm": 0.36489760875701904, "learning_rate": 4.709125199477731e-05, "loss": 0.0536, "step": 14688 }, { "epoch": 1.5924761491760624, "grad_norm": 0.2185102254152298, "learning_rate": 4.708762512694038e-05, "loss": 0.0119, "step": 14689 }, { "epoch": 1.5925845620121422, "grad_norm": 0.7614810466766357, "learning_rate": 4.708399825910344e-05, "loss": 0.0507, "step": 14690 }, { "epoch": 1.5926929748482221, "grad_norm": 0.1972731351852417, "learning_rate": 4.7080371391266505e-05, "loss": 0.0327, "step": 14691 }, { "epoch": 1.5928013876843017, "grad_norm": 0.2688463032245636, "learning_rate": 4.707674452342957e-05, "loss": 0.0203, "step": 14692 }, { "epoch": 1.5929098005203817, "grad_norm": 0.5275362133979797, "learning_rate": 4.707311765559263e-05, "loss": 0.0405, "step": 14693 }, { "epoch": 1.5930182133564614, "grad_norm": 0.38647371530532837, "learning_rate": 4.70694907877557e-05, "loss": 0.0442, "step": 14694 }, { "epoch": 1.5931266261925412, "grad_norm": 0.5580289363861084, "learning_rate": 4.706586391991876e-05, "loss": 0.0674, "step": 14695 }, { "epoch": 1.593235039028621, "grad_norm": 0.16480635106563568, "learning_rate": 4.7062237052081826e-05, "loss": 0.013, "step": 14696 }, { "epoch": 1.5933434518647007, "grad_norm": 0.07483012229204178, "learning_rate": 4.705861018424489e-05, "loss": 0.005, "step": 14697 }, { "epoch": 1.5934518647007807, "grad_norm": 0.4406821131706238, "learning_rate": 4.7054983316407954e-05, "loss": 0.0865, "step": 14698 }, { "epoch": 1.5935602775368602, "grad_norm": 0.2413467913866043, "learning_rate": 4.705135644857102e-05, "loss": 0.0376, "step": 14699 }, { "epoch": 1.5936686903729402, "grad_norm": 0.3273029625415802, "learning_rate": 4.704772958073408e-05, "loss": 0.0223, "step": 14700 }, { "epoch": 1.59377710320902, "grad_norm": 0.30262625217437744, "learning_rate": 4.704410271289714e-05, "loss": 0.0352, "step": 14701 }, { "epoch": 1.5938855160450998, "grad_norm": 0.19352880120277405, "learning_rate": 4.7040475845060204e-05, "loss": 0.0333, "step": 14702 }, { "epoch": 1.5939939288811795, "grad_norm": 0.12297724932432175, "learning_rate": 4.7036848977223275e-05, "loss": 0.0102, "step": 14703 }, { "epoch": 1.5941023417172593, "grad_norm": 0.24509835243225098, "learning_rate": 4.703322210938634e-05, "loss": 0.0228, "step": 14704 }, { "epoch": 1.5942107545533393, "grad_norm": 0.29608938097953796, "learning_rate": 4.70295952415494e-05, "loss": 0.0421, "step": 14705 }, { "epoch": 1.5943191673894188, "grad_norm": 0.21108970046043396, "learning_rate": 4.702596837371247e-05, "loss": 0.0065, "step": 14706 }, { "epoch": 1.5944275802254988, "grad_norm": 0.08856988698244095, "learning_rate": 4.702234150587553e-05, "loss": 0.0099, "step": 14707 }, { "epoch": 1.5945359930615783, "grad_norm": 0.07413662225008011, "learning_rate": 4.701871463803859e-05, "loss": 0.0079, "step": 14708 }, { "epoch": 1.5946444058976583, "grad_norm": 0.515688955783844, "learning_rate": 4.701508777020165e-05, "loss": 0.0418, "step": 14709 }, { "epoch": 1.594752818733738, "grad_norm": 0.2021493762731552, "learning_rate": 4.7011460902364717e-05, "loss": 0.0193, "step": 14710 }, { "epoch": 1.5948612315698178, "grad_norm": 0.04889432713389397, "learning_rate": 4.700783403452779e-05, "loss": 0.0025, "step": 14711 }, { "epoch": 1.5949696444058976, "grad_norm": 0.2721955180168152, "learning_rate": 4.700420716669085e-05, "loss": 0.0202, "step": 14712 }, { "epoch": 1.5950780572419774, "grad_norm": 0.1564808040857315, "learning_rate": 4.7000580298853916e-05, "loss": 0.0164, "step": 14713 }, { "epoch": 1.5951864700780574, "grad_norm": 0.11090417206287384, "learning_rate": 4.699695343101698e-05, "loss": 0.0107, "step": 14714 }, { "epoch": 1.595294882914137, "grad_norm": 0.14324985444545746, "learning_rate": 4.699332656318004e-05, "loss": 0.0189, "step": 14715 }, { "epoch": 1.5954032957502169, "grad_norm": 0.36205145716667175, "learning_rate": 4.69896996953431e-05, "loss": 0.0759, "step": 14716 }, { "epoch": 1.5955117085862967, "grad_norm": 0.40193837881088257, "learning_rate": 4.6986072827506165e-05, "loss": 0.0464, "step": 14717 }, { "epoch": 1.5956201214223764, "grad_norm": 0.14488042891025543, "learning_rate": 4.698244595966923e-05, "loss": 0.0098, "step": 14718 }, { "epoch": 1.5957285342584562, "grad_norm": 0.10956131666898727, "learning_rate": 4.6978819091832294e-05, "loss": 0.0098, "step": 14719 }, { "epoch": 1.595836947094536, "grad_norm": 0.43114620447158813, "learning_rate": 4.6975192223995364e-05, "loss": 0.0692, "step": 14720 }, { "epoch": 1.595945359930616, "grad_norm": 0.8351991176605225, "learning_rate": 4.697156535615843e-05, "loss": 0.0342, "step": 14721 }, { "epoch": 1.5960537727666955, "grad_norm": 0.38616451621055603, "learning_rate": 4.6967938488321486e-05, "loss": 0.0121, "step": 14722 }, { "epoch": 1.5961621856027755, "grad_norm": 0.17798243463039398, "learning_rate": 4.696431162048455e-05, "loss": 0.0187, "step": 14723 }, { "epoch": 1.5962705984388552, "grad_norm": 0.10650806874036789, "learning_rate": 4.6960684752647614e-05, "loss": 0.0125, "step": 14724 }, { "epoch": 1.596379011274935, "grad_norm": 0.1636771857738495, "learning_rate": 4.695705788481068e-05, "loss": 0.0106, "step": 14725 }, { "epoch": 1.5964874241110147, "grad_norm": 0.5935080647468567, "learning_rate": 4.695343101697374e-05, "loss": 0.0314, "step": 14726 }, { "epoch": 1.5965958369470945, "grad_norm": 0.25902336835861206, "learning_rate": 4.6949804149136806e-05, "loss": 0.0299, "step": 14727 }, { "epoch": 1.5967042497831745, "grad_norm": 0.08443070948123932, "learning_rate": 4.694617728129987e-05, "loss": 0.0075, "step": 14728 }, { "epoch": 1.596812662619254, "grad_norm": 0.046167585998773575, "learning_rate": 4.6942550413462935e-05, "loss": 0.0032, "step": 14729 }, { "epoch": 1.596921075455334, "grad_norm": 0.24549464881420135, "learning_rate": 4.6938923545626e-05, "loss": 0.0244, "step": 14730 }, { "epoch": 1.5970294882914136, "grad_norm": 0.029058896005153656, "learning_rate": 4.693529667778906e-05, "loss": 0.0014, "step": 14731 }, { "epoch": 1.5971379011274935, "grad_norm": 0.2882100045681, "learning_rate": 4.693166980995213e-05, "loss": 0.0191, "step": 14732 }, { "epoch": 1.5972463139635733, "grad_norm": 0.3127301335334778, "learning_rate": 4.692804294211519e-05, "loss": 0.03, "step": 14733 }, { "epoch": 1.597354726799653, "grad_norm": 0.18602262437343597, "learning_rate": 4.6924416074278255e-05, "loss": 0.0114, "step": 14734 }, { "epoch": 1.5974631396357328, "grad_norm": 0.2581861615180969, "learning_rate": 4.692078920644132e-05, "loss": 0.0189, "step": 14735 }, { "epoch": 1.5975715524718126, "grad_norm": 1.0082981586456299, "learning_rate": 4.691716233860438e-05, "loss": 0.0581, "step": 14736 }, { "epoch": 1.5976799653078926, "grad_norm": 0.7449299097061157, "learning_rate": 4.691353547076745e-05, "loss": 0.0237, "step": 14737 }, { "epoch": 1.5977883781439721, "grad_norm": 0.5912035703659058, "learning_rate": 4.690990860293051e-05, "loss": 0.0379, "step": 14738 }, { "epoch": 1.597896790980052, "grad_norm": 0.6564121246337891, "learning_rate": 4.6906281735093576e-05, "loss": 0.0343, "step": 14739 }, { "epoch": 1.5980052038161319, "grad_norm": 0.26011231541633606, "learning_rate": 4.690265486725664e-05, "loss": 0.0454, "step": 14740 }, { "epoch": 1.5981136166522116, "grad_norm": 0.43317148089408875, "learning_rate": 4.6899027999419704e-05, "loss": 0.0373, "step": 14741 }, { "epoch": 1.5982220294882914, "grad_norm": 0.4146957993507385, "learning_rate": 4.689540113158277e-05, "loss": 0.0501, "step": 14742 }, { "epoch": 1.5983304423243712, "grad_norm": 0.4307241141796112, "learning_rate": 4.689177426374583e-05, "loss": 0.0257, "step": 14743 }, { "epoch": 1.5984388551604511, "grad_norm": 0.5280986428260803, "learning_rate": 4.6888147395908896e-05, "loss": 0.0498, "step": 14744 }, { "epoch": 1.5985472679965307, "grad_norm": 0.21589665114879608, "learning_rate": 4.6884520528071953e-05, "loss": 0.0257, "step": 14745 }, { "epoch": 1.5986556808326107, "grad_norm": 0.15513786673545837, "learning_rate": 4.6880893660235024e-05, "loss": 0.0131, "step": 14746 }, { "epoch": 1.5987640936686902, "grad_norm": 0.26208576560020447, "learning_rate": 4.687726679239809e-05, "loss": 0.0231, "step": 14747 }, { "epoch": 1.5988725065047702, "grad_norm": 0.14056961238384247, "learning_rate": 4.687363992456115e-05, "loss": 0.0163, "step": 14748 }, { "epoch": 1.59898091934085, "grad_norm": 0.19370198249816895, "learning_rate": 4.687001305672422e-05, "loss": 0.0251, "step": 14749 }, { "epoch": 1.5990893321769297, "grad_norm": 0.4447996914386749, "learning_rate": 4.686638618888728e-05, "loss": 0.0611, "step": 14750 }, { "epoch": 1.5991977450130095, "grad_norm": 0.9338819980621338, "learning_rate": 4.6862759321050345e-05, "loss": 0.0552, "step": 14751 }, { "epoch": 1.5993061578490892, "grad_norm": 0.1447344422340393, "learning_rate": 4.685913245321341e-05, "loss": 0.0113, "step": 14752 }, { "epoch": 1.5994145706851692, "grad_norm": 0.32364147901535034, "learning_rate": 4.6855505585376466e-05, "loss": 0.053, "step": 14753 }, { "epoch": 1.5995229835212488, "grad_norm": 0.2903447449207306, "learning_rate": 4.685187871753953e-05, "loss": 0.0093, "step": 14754 }, { "epoch": 1.5996313963573288, "grad_norm": 0.22920919954776764, "learning_rate": 4.68482518497026e-05, "loss": 0.0175, "step": 14755 }, { "epoch": 1.5997398091934085, "grad_norm": 0.7738767266273499, "learning_rate": 4.6844624981865665e-05, "loss": 0.0316, "step": 14756 }, { "epoch": 1.5998482220294883, "grad_norm": 0.24526627361774445, "learning_rate": 4.684099811402873e-05, "loss": 0.0263, "step": 14757 }, { "epoch": 1.599956634865568, "grad_norm": 0.6528075933456421, "learning_rate": 4.6837371246191794e-05, "loss": 0.0373, "step": 14758 }, { "epoch": 1.6000650477016478, "grad_norm": 0.38706034421920776, "learning_rate": 4.683374437835486e-05, "loss": 0.034, "step": 14759 }, { "epoch": 1.6001734605377278, "grad_norm": 0.2034749835729599, "learning_rate": 4.6830117510517915e-05, "loss": 0.0178, "step": 14760 }, { "epoch": 1.6002818733738073, "grad_norm": 0.17350828647613525, "learning_rate": 4.682649064268098e-05, "loss": 0.0115, "step": 14761 }, { "epoch": 1.6003902862098873, "grad_norm": 0.2579394578933716, "learning_rate": 4.682286377484404e-05, "loss": 0.0258, "step": 14762 }, { "epoch": 1.600498699045967, "grad_norm": 0.1898656040430069, "learning_rate": 4.6819236907007114e-05, "loss": 0.0309, "step": 14763 }, { "epoch": 1.6006071118820469, "grad_norm": 0.1335894763469696, "learning_rate": 4.681561003917018e-05, "loss": 0.0095, "step": 14764 }, { "epoch": 1.6007155247181266, "grad_norm": 0.1723490208387375, "learning_rate": 4.681198317133324e-05, "loss": 0.017, "step": 14765 }, { "epoch": 1.6008239375542064, "grad_norm": 0.32303786277770996, "learning_rate": 4.6808356303496306e-05, "loss": 0.035, "step": 14766 }, { "epoch": 1.6009323503902864, "grad_norm": 0.21753625571727753, "learning_rate": 4.6804729435659364e-05, "loss": 0.0177, "step": 14767 }, { "epoch": 1.601040763226366, "grad_norm": 0.1759578138589859, "learning_rate": 4.680110256782243e-05, "loss": 0.0159, "step": 14768 }, { "epoch": 1.601149176062446, "grad_norm": 0.0974426418542862, "learning_rate": 4.679747569998549e-05, "loss": 0.0099, "step": 14769 }, { "epoch": 1.6012575888985254, "grad_norm": 0.5736430883407593, "learning_rate": 4.6793848832148556e-05, "loss": 0.0317, "step": 14770 }, { "epoch": 1.6013660017346054, "grad_norm": 0.1745920479297638, "learning_rate": 4.679022196431162e-05, "loss": 0.0153, "step": 14771 }, { "epoch": 1.6014744145706852, "grad_norm": 0.49990618228912354, "learning_rate": 4.678659509647469e-05, "loss": 0.0383, "step": 14772 }, { "epoch": 1.601582827406765, "grad_norm": 0.3489571809768677, "learning_rate": 4.6782968228637755e-05, "loss": 0.0375, "step": 14773 }, { "epoch": 1.6016912402428447, "grad_norm": 0.3076767027378082, "learning_rate": 4.677934136080081e-05, "loss": 0.0302, "step": 14774 }, { "epoch": 1.6017996530789245, "grad_norm": 0.1479659378528595, "learning_rate": 4.677571449296388e-05, "loss": 0.0204, "step": 14775 }, { "epoch": 1.6019080659150045, "grad_norm": 0.12234755605459213, "learning_rate": 4.677208762512694e-05, "loss": 0.0054, "step": 14776 }, { "epoch": 1.602016478751084, "grad_norm": 0.21076932549476624, "learning_rate": 4.6768460757290005e-05, "loss": 0.0132, "step": 14777 }, { "epoch": 1.602124891587164, "grad_norm": 0.17358526587486267, "learning_rate": 4.676483388945307e-05, "loss": 0.0126, "step": 14778 }, { "epoch": 1.6022333044232437, "grad_norm": 0.12401051819324493, "learning_rate": 4.676120702161613e-05, "loss": 0.0109, "step": 14779 }, { "epoch": 1.6023417172593235, "grad_norm": 0.5228900909423828, "learning_rate": 4.67575801537792e-05, "loss": 0.0447, "step": 14780 }, { "epoch": 1.6024501300954033, "grad_norm": 0.3025560677051544, "learning_rate": 4.675395328594227e-05, "loss": 0.012, "step": 14781 }, { "epoch": 1.602558542931483, "grad_norm": 0.1774042695760727, "learning_rate": 4.6750326418105325e-05, "loss": 0.0173, "step": 14782 }, { "epoch": 1.602666955767563, "grad_norm": 0.12429799139499664, "learning_rate": 4.674669955026839e-05, "loss": 0.0092, "step": 14783 }, { "epoch": 1.6027753686036426, "grad_norm": 0.09606657922267914, "learning_rate": 4.6743072682431454e-05, "loss": 0.0052, "step": 14784 }, { "epoch": 1.6028837814397225, "grad_norm": 0.2876781225204468, "learning_rate": 4.673944581459452e-05, "loss": 0.01, "step": 14785 }, { "epoch": 1.6029921942758023, "grad_norm": 0.3255464732646942, "learning_rate": 4.673581894675758e-05, "loss": 0.0164, "step": 14786 }, { "epoch": 1.603100607111882, "grad_norm": 0.2996829152107239, "learning_rate": 4.6732192078920646e-05, "loss": 0.0163, "step": 14787 }, { "epoch": 1.6032090199479618, "grad_norm": 0.36881500482559204, "learning_rate": 4.672856521108371e-05, "loss": 0.0345, "step": 14788 }, { "epoch": 1.6033174327840416, "grad_norm": 0.4145362675189972, "learning_rate": 4.6724938343246774e-05, "loss": 0.0301, "step": 14789 }, { "epoch": 1.6034258456201216, "grad_norm": 0.25409650802612305, "learning_rate": 4.672131147540984e-05, "loss": 0.0241, "step": 14790 }, { "epoch": 1.6035342584562011, "grad_norm": 0.3368624448776245, "learning_rate": 4.67176846075729e-05, "loss": 0.0381, "step": 14791 }, { "epoch": 1.603642671292281, "grad_norm": 0.38610973954200745, "learning_rate": 4.6714057739735966e-05, "loss": 0.0339, "step": 14792 }, { "epoch": 1.6037510841283606, "grad_norm": 0.31134918332099915, "learning_rate": 4.671043087189903e-05, "loss": 0.0119, "step": 14793 }, { "epoch": 1.6038594969644406, "grad_norm": 0.16897055506706238, "learning_rate": 4.6706804004062095e-05, "loss": 0.0208, "step": 14794 }, { "epoch": 1.6039679098005204, "grad_norm": 0.05063631385564804, "learning_rate": 4.670317713622516e-05, "loss": 0.0021, "step": 14795 }, { "epoch": 1.6040763226366002, "grad_norm": 0.07448220252990723, "learning_rate": 4.669955026838822e-05, "loss": 0.0049, "step": 14796 }, { "epoch": 1.60418473547268, "grad_norm": 0.3601934015750885, "learning_rate": 4.669592340055129e-05, "loss": 0.0188, "step": 14797 }, { "epoch": 1.6042931483087597, "grad_norm": 0.41209232807159424, "learning_rate": 4.669229653271435e-05, "loss": 0.0326, "step": 14798 }, { "epoch": 1.6044015611448397, "grad_norm": 0.3251560628414154, "learning_rate": 4.6688669664877415e-05, "loss": 0.05, "step": 14799 }, { "epoch": 1.6045099739809192, "grad_norm": 0.13284708559513092, "learning_rate": 4.668504279704048e-05, "loss": 0.0095, "step": 14800 }, { "epoch": 1.6046183868169992, "grad_norm": 0.3211849331855774, "learning_rate": 4.668141592920354e-05, "loss": 0.02, "step": 14801 }, { "epoch": 1.604726799653079, "grad_norm": 0.6846868395805359, "learning_rate": 4.667778906136661e-05, "loss": 0.0298, "step": 14802 }, { "epoch": 1.6048352124891587, "grad_norm": 0.12008882313966751, "learning_rate": 4.667416219352967e-05, "loss": 0.0098, "step": 14803 }, { "epoch": 1.6049436253252385, "grad_norm": 0.1930539757013321, "learning_rate": 4.6670535325692736e-05, "loss": 0.014, "step": 14804 }, { "epoch": 1.6050520381613183, "grad_norm": 0.2339845597743988, "learning_rate": 4.666690845785579e-05, "loss": 0.012, "step": 14805 }, { "epoch": 1.6051604509973982, "grad_norm": 0.3808574676513672, "learning_rate": 4.6663281590018864e-05, "loss": 0.0328, "step": 14806 }, { "epoch": 1.6052688638334778, "grad_norm": 0.45146608352661133, "learning_rate": 4.665965472218193e-05, "loss": 0.0362, "step": 14807 }, { "epoch": 1.6053772766695578, "grad_norm": 0.4423640966415405, "learning_rate": 4.665602785434499e-05, "loss": 0.0503, "step": 14808 }, { "epoch": 1.6054856895056373, "grad_norm": 0.2264348417520523, "learning_rate": 4.6652400986508056e-05, "loss": 0.0568, "step": 14809 }, { "epoch": 1.6055941023417173, "grad_norm": 0.20874011516571045, "learning_rate": 4.664877411867112e-05, "loss": 0.0071, "step": 14810 }, { "epoch": 1.605702515177797, "grad_norm": 0.30568209290504456, "learning_rate": 4.6645147250834184e-05, "loss": 0.0271, "step": 14811 }, { "epoch": 1.6058109280138768, "grad_norm": 0.2797524929046631, "learning_rate": 4.664152038299724e-05, "loss": 0.0146, "step": 14812 }, { "epoch": 1.6059193408499568, "grad_norm": 0.227681964635849, "learning_rate": 4.6637893515160306e-05, "loss": 0.0126, "step": 14813 }, { "epoch": 1.6060277536860363, "grad_norm": 0.4648086726665497, "learning_rate": 4.663426664732337e-05, "loss": 0.0163, "step": 14814 }, { "epoch": 1.6061361665221163, "grad_norm": 0.21939007937908173, "learning_rate": 4.663063977948644e-05, "loss": 0.0276, "step": 14815 }, { "epoch": 1.6062445793581959, "grad_norm": 0.1516185998916626, "learning_rate": 4.6627012911649505e-05, "loss": 0.0118, "step": 14816 }, { "epoch": 1.6063529921942759, "grad_norm": 0.10822098702192307, "learning_rate": 4.662338604381257e-05, "loss": 0.0101, "step": 14817 }, { "epoch": 1.6064614050303556, "grad_norm": 0.3556937277317047, "learning_rate": 4.661975917597563e-05, "loss": 0.0255, "step": 14818 }, { "epoch": 1.6065698178664354, "grad_norm": 0.6420767903327942, "learning_rate": 4.661613230813869e-05, "loss": 0.0294, "step": 14819 }, { "epoch": 1.6066782307025151, "grad_norm": 0.1451348513364792, "learning_rate": 4.6612505440301755e-05, "loss": 0.0077, "step": 14820 }, { "epoch": 1.606786643538595, "grad_norm": 0.23000937700271606, "learning_rate": 4.660887857246482e-05, "loss": 0.0162, "step": 14821 }, { "epoch": 1.606895056374675, "grad_norm": 0.3638467490673065, "learning_rate": 4.660525170462788e-05, "loss": 0.0447, "step": 14822 }, { "epoch": 1.6070034692107544, "grad_norm": 0.21442389488220215, "learning_rate": 4.660162483679095e-05, "loss": 0.0415, "step": 14823 }, { "epoch": 1.6071118820468344, "grad_norm": 0.3115204870700836, "learning_rate": 4.659799796895402e-05, "loss": 0.0342, "step": 14824 }, { "epoch": 1.6072202948829142, "grad_norm": 0.21912004053592682, "learning_rate": 4.659437110111708e-05, "loss": 0.0145, "step": 14825 }, { "epoch": 1.607328707718994, "grad_norm": 0.702917754650116, "learning_rate": 4.659074423328014e-05, "loss": 0.0756, "step": 14826 }, { "epoch": 1.6074371205550737, "grad_norm": 0.2842068076133728, "learning_rate": 4.65871173654432e-05, "loss": 0.0252, "step": 14827 }, { "epoch": 1.6075455333911535, "grad_norm": 0.43476906418800354, "learning_rate": 4.658349049760627e-05, "loss": 0.0197, "step": 14828 }, { "epoch": 1.6076539462272335, "grad_norm": 0.13984878361225128, "learning_rate": 4.657986362976933e-05, "loss": 0.0158, "step": 14829 }, { "epoch": 1.607762359063313, "grad_norm": 0.4314601421356201, "learning_rate": 4.6576236761932396e-05, "loss": 0.0351, "step": 14830 }, { "epoch": 1.607870771899393, "grad_norm": 0.18955561518669128, "learning_rate": 4.657260989409546e-05, "loss": 0.0055, "step": 14831 }, { "epoch": 1.6079791847354725, "grad_norm": 0.6682084202766418, "learning_rate": 4.656898302625853e-05, "loss": 0.0859, "step": 14832 }, { "epoch": 1.6080875975715525, "grad_norm": 0.2822227478027344, "learning_rate": 4.6565356158421595e-05, "loss": 0.0115, "step": 14833 }, { "epoch": 1.6081960104076323, "grad_norm": 0.4137432873249054, "learning_rate": 4.656172929058465e-05, "loss": 0.0433, "step": 14834 }, { "epoch": 1.608304423243712, "grad_norm": 0.21764373779296875, "learning_rate": 4.6558102422747716e-05, "loss": 0.0175, "step": 14835 }, { "epoch": 1.6084128360797918, "grad_norm": 0.624523401260376, "learning_rate": 4.655447555491078e-05, "loss": 0.0437, "step": 14836 }, { "epoch": 1.6085212489158716, "grad_norm": 0.1326010376214981, "learning_rate": 4.6550848687073844e-05, "loss": 0.0118, "step": 14837 }, { "epoch": 1.6086296617519515, "grad_norm": 0.3835393190383911, "learning_rate": 4.654722181923691e-05, "loss": 0.0239, "step": 14838 }, { "epoch": 1.608738074588031, "grad_norm": 0.1788053661584854, "learning_rate": 4.654359495139997e-05, "loss": 0.0104, "step": 14839 }, { "epoch": 1.608846487424111, "grad_norm": 0.2620190978050232, "learning_rate": 4.653996808356304e-05, "loss": 0.04, "step": 14840 }, { "epoch": 1.6089549002601908, "grad_norm": 0.2488320916891098, "learning_rate": 4.65363412157261e-05, "loss": 0.016, "step": 14841 }, { "epoch": 1.6090633130962706, "grad_norm": 0.39463087916374207, "learning_rate": 4.6532714347889165e-05, "loss": 0.0127, "step": 14842 }, { "epoch": 1.6091717259323504, "grad_norm": 0.43497011065483093, "learning_rate": 4.652908748005223e-05, "loss": 0.0294, "step": 14843 }, { "epoch": 1.6092801387684301, "grad_norm": 0.24828337132930756, "learning_rate": 4.652546061221529e-05, "loss": 0.0145, "step": 14844 }, { "epoch": 1.60938855160451, "grad_norm": 0.27002575993537903, "learning_rate": 4.652183374437836e-05, "loss": 0.0254, "step": 14845 }, { "epoch": 1.6094969644405897, "grad_norm": 0.23321692645549774, "learning_rate": 4.651820687654142e-05, "loss": 0.0209, "step": 14846 }, { "epoch": 1.6096053772766696, "grad_norm": 0.07801655679941177, "learning_rate": 4.6514580008704485e-05, "loss": 0.0024, "step": 14847 }, { "epoch": 1.6097137901127494, "grad_norm": 0.5126264691352844, "learning_rate": 4.651095314086755e-05, "loss": 0.0287, "step": 14848 }, { "epoch": 1.6098222029488292, "grad_norm": 0.16802847385406494, "learning_rate": 4.6507326273030614e-05, "loss": 0.0129, "step": 14849 }, { "epoch": 1.609930615784909, "grad_norm": 0.053820881992578506, "learning_rate": 4.650369940519368e-05, "loss": 0.0021, "step": 14850 }, { "epoch": 1.6100390286209887, "grad_norm": 0.4343715310096741, "learning_rate": 4.650007253735674e-05, "loss": 0.0344, "step": 14851 }, { "epoch": 1.6101474414570687, "grad_norm": 0.12279964238405228, "learning_rate": 4.6496445669519806e-05, "loss": 0.0105, "step": 14852 }, { "epoch": 1.6102558542931482, "grad_norm": 0.2057075947523117, "learning_rate": 4.649281880168287e-05, "loss": 0.0212, "step": 14853 }, { "epoch": 1.6103642671292282, "grad_norm": 0.23012091219425201, "learning_rate": 4.6489191933845934e-05, "loss": 0.0211, "step": 14854 }, { "epoch": 1.6104726799653077, "grad_norm": 0.1885010451078415, "learning_rate": 4.6485565066009e-05, "loss": 0.0101, "step": 14855 }, { "epoch": 1.6105810928013877, "grad_norm": 0.22062602639198303, "learning_rate": 4.648193819817206e-05, "loss": 0.0243, "step": 14856 }, { "epoch": 1.6106895056374675, "grad_norm": 0.4315732419490814, "learning_rate": 4.647831133033512e-05, "loss": 0.0383, "step": 14857 }, { "epoch": 1.6107979184735473, "grad_norm": 0.084272600710392, "learning_rate": 4.647468446249819e-05, "loss": 0.0067, "step": 14858 }, { "epoch": 1.610906331309627, "grad_norm": 0.13088631629943848, "learning_rate": 4.6471057594661255e-05, "loss": 0.0142, "step": 14859 }, { "epoch": 1.6110147441457068, "grad_norm": 0.5423708558082581, "learning_rate": 4.646743072682432e-05, "loss": 0.0323, "step": 14860 }, { "epoch": 1.6111231569817868, "grad_norm": 0.042135242372751236, "learning_rate": 4.646380385898738e-05, "loss": 0.0028, "step": 14861 }, { "epoch": 1.6112315698178663, "grad_norm": 0.05906867980957031, "learning_rate": 4.646017699115045e-05, "loss": 0.0038, "step": 14862 }, { "epoch": 1.6113399826539463, "grad_norm": 0.16834406554698944, "learning_rate": 4.645655012331351e-05, "loss": 0.015, "step": 14863 }, { "epoch": 1.611448395490026, "grad_norm": 0.5076225399971008, "learning_rate": 4.645292325547657e-05, "loss": 0.0304, "step": 14864 }, { "epoch": 1.6115568083261058, "grad_norm": 0.12011272460222244, "learning_rate": 4.644929638763963e-05, "loss": 0.0129, "step": 14865 }, { "epoch": 1.6116652211621856, "grad_norm": 0.48757651448249817, "learning_rate": 4.64456695198027e-05, "loss": 0.058, "step": 14866 }, { "epoch": 1.6117736339982653, "grad_norm": 0.3716718852519989, "learning_rate": 4.644204265196577e-05, "loss": 0.0287, "step": 14867 }, { "epoch": 1.6118820468343453, "grad_norm": 0.10042772442102432, "learning_rate": 4.643841578412883e-05, "loss": 0.0125, "step": 14868 }, { "epoch": 1.6119904596704249, "grad_norm": 0.3307268023490906, "learning_rate": 4.6434788916291896e-05, "loss": 0.0209, "step": 14869 }, { "epoch": 1.6120988725065049, "grad_norm": 0.11735022813081741, "learning_rate": 4.643116204845496e-05, "loss": 0.0077, "step": 14870 }, { "epoch": 1.6122072853425846, "grad_norm": 0.2666001319885254, "learning_rate": 4.642753518061802e-05, "loss": 0.0137, "step": 14871 }, { "epoch": 1.6123156981786644, "grad_norm": 0.4496389627456665, "learning_rate": 4.642390831278108e-05, "loss": 0.02, "step": 14872 }, { "epoch": 1.6124241110147441, "grad_norm": 0.14522743225097656, "learning_rate": 4.6420281444944145e-05, "loss": 0.0064, "step": 14873 }, { "epoch": 1.612532523850824, "grad_norm": 0.4114629924297333, "learning_rate": 4.641665457710721e-05, "loss": 0.04, "step": 14874 }, { "epoch": 1.612640936686904, "grad_norm": 1.6387221813201904, "learning_rate": 4.6413027709270274e-05, "loss": 0.0309, "step": 14875 }, { "epoch": 1.6127493495229834, "grad_norm": 0.4628521502017975, "learning_rate": 4.6409400841433344e-05, "loss": 0.0555, "step": 14876 }, { "epoch": 1.6128577623590634, "grad_norm": 0.47937846183776855, "learning_rate": 4.640577397359641e-05, "loss": 0.0195, "step": 14877 }, { "epoch": 1.612966175195143, "grad_norm": 0.10396988689899445, "learning_rate": 4.6402147105759466e-05, "loss": 0.0058, "step": 14878 }, { "epoch": 1.613074588031223, "grad_norm": 0.7658771276473999, "learning_rate": 4.639852023792253e-05, "loss": 0.0382, "step": 14879 }, { "epoch": 1.6131830008673027, "grad_norm": 0.24586519598960876, "learning_rate": 4.6394893370085594e-05, "loss": 0.0138, "step": 14880 }, { "epoch": 1.6132914137033825, "grad_norm": 0.05552195757627487, "learning_rate": 4.639126650224866e-05, "loss": 0.0036, "step": 14881 }, { "epoch": 1.6133998265394622, "grad_norm": 0.18877725303173065, "learning_rate": 4.638763963441172e-05, "loss": 0.0153, "step": 14882 }, { "epoch": 1.613508239375542, "grad_norm": 0.23753350973129272, "learning_rate": 4.6384012766574786e-05, "loss": 0.0173, "step": 14883 }, { "epoch": 1.613616652211622, "grad_norm": 0.7121772170066833, "learning_rate": 4.638038589873786e-05, "loss": 0.0609, "step": 14884 }, { "epoch": 1.6137250650477015, "grad_norm": 0.46009889245033264, "learning_rate": 4.637675903090092e-05, "loss": 0.0329, "step": 14885 }, { "epoch": 1.6138334778837815, "grad_norm": 0.45814916491508484, "learning_rate": 4.637313216306398e-05, "loss": 0.036, "step": 14886 }, { "epoch": 1.6139418907198613, "grad_norm": 0.2765478193759918, "learning_rate": 4.636950529522704e-05, "loss": 0.0282, "step": 14887 }, { "epoch": 1.614050303555941, "grad_norm": 0.06041369214653969, "learning_rate": 4.636587842739011e-05, "loss": 0.0057, "step": 14888 }, { "epoch": 1.6141587163920208, "grad_norm": 0.13108770549297333, "learning_rate": 4.636225155955317e-05, "loss": 0.0102, "step": 14889 }, { "epoch": 1.6142671292281006, "grad_norm": 0.17591804265975952, "learning_rate": 4.6358624691716235e-05, "loss": 0.0046, "step": 14890 }, { "epoch": 1.6143755420641805, "grad_norm": 1.0270986557006836, "learning_rate": 4.63549978238793e-05, "loss": 0.0606, "step": 14891 }, { "epoch": 1.61448395490026, "grad_norm": 0.20575842261314392, "learning_rate": 4.635137095604236e-05, "loss": 0.0171, "step": 14892 }, { "epoch": 1.61459236773634, "grad_norm": 0.07533329725265503, "learning_rate": 4.634774408820543e-05, "loss": 0.0127, "step": 14893 }, { "epoch": 1.6147007805724196, "grad_norm": 0.030812222510576248, "learning_rate": 4.634411722036849e-05, "loss": 0.0014, "step": 14894 }, { "epoch": 1.6148091934084996, "grad_norm": 0.14475266635417938, "learning_rate": 4.6340490352531556e-05, "loss": 0.0151, "step": 14895 }, { "epoch": 1.6149176062445794, "grad_norm": 0.6856656074523926, "learning_rate": 4.633686348469462e-05, "loss": 0.0219, "step": 14896 }, { "epoch": 1.6150260190806591, "grad_norm": 0.6036583185195923, "learning_rate": 4.6333236616857684e-05, "loss": 0.0282, "step": 14897 }, { "epoch": 1.6151344319167389, "grad_norm": 0.24524500966072083, "learning_rate": 4.632960974902075e-05, "loss": 0.0163, "step": 14898 }, { "epoch": 1.6152428447528187, "grad_norm": 0.2768019735813141, "learning_rate": 4.632598288118381e-05, "loss": 0.021, "step": 14899 }, { "epoch": 1.6153512575888986, "grad_norm": 0.4350111186504364, "learning_rate": 4.6322356013346876e-05, "loss": 0.0123, "step": 14900 }, { "epoch": 1.6154596704249782, "grad_norm": 0.3424243628978729, "learning_rate": 4.631872914550994e-05, "loss": 0.0503, "step": 14901 }, { "epoch": 1.6155680832610582, "grad_norm": 0.682255744934082, "learning_rate": 4.6315102277673004e-05, "loss": 0.0357, "step": 14902 }, { "epoch": 1.615676496097138, "grad_norm": 1.0943756103515625, "learning_rate": 4.631147540983607e-05, "loss": 0.0687, "step": 14903 }, { "epoch": 1.6157849089332177, "grad_norm": 0.3829042911529541, "learning_rate": 4.630784854199913e-05, "loss": 0.0184, "step": 14904 }, { "epoch": 1.6158933217692975, "grad_norm": 0.1825774610042572, "learning_rate": 4.63042216741622e-05, "loss": 0.01, "step": 14905 }, { "epoch": 1.6160017346053772, "grad_norm": 0.46486032009124756, "learning_rate": 4.630059480632526e-05, "loss": 0.0318, "step": 14906 }, { "epoch": 1.6161101474414572, "grad_norm": 0.34846875071525574, "learning_rate": 4.6296967938488325e-05, "loss": 0.0143, "step": 14907 }, { "epoch": 1.6162185602775367, "grad_norm": 0.47668391466140747, "learning_rate": 4.629334107065139e-05, "loss": 0.0394, "step": 14908 }, { "epoch": 1.6163269731136167, "grad_norm": 0.25471144914627075, "learning_rate": 4.6289714202814446e-05, "loss": 0.0117, "step": 14909 }, { "epoch": 1.6164353859496965, "grad_norm": 0.19193103909492493, "learning_rate": 4.628608733497752e-05, "loss": 0.0085, "step": 14910 }, { "epoch": 1.6165437987857763, "grad_norm": 0.25159546732902527, "learning_rate": 4.628246046714058e-05, "loss": 0.0142, "step": 14911 }, { "epoch": 1.616652211621856, "grad_norm": 0.32764026522636414, "learning_rate": 4.6278833599303645e-05, "loss": 0.0572, "step": 14912 }, { "epoch": 1.6167606244579358, "grad_norm": 0.5921118855476379, "learning_rate": 4.627520673146671e-05, "loss": 0.0714, "step": 14913 }, { "epoch": 1.6168690372940158, "grad_norm": 0.5709491968154907, "learning_rate": 4.6271579863629774e-05, "loss": 0.0187, "step": 14914 }, { "epoch": 1.6169774501300953, "grad_norm": 0.14619076251983643, "learning_rate": 4.626795299579284e-05, "loss": 0.0136, "step": 14915 }, { "epoch": 1.6170858629661753, "grad_norm": 0.1095471978187561, "learning_rate": 4.6264326127955895e-05, "loss": 0.0051, "step": 14916 }, { "epoch": 1.6171942758022548, "grad_norm": 0.2200147807598114, "learning_rate": 4.626069926011896e-05, "loss": 0.0163, "step": 14917 }, { "epoch": 1.6173026886383348, "grad_norm": 0.15895815193653107, "learning_rate": 4.625707239228202e-05, "loss": 0.012, "step": 14918 }, { "epoch": 1.6174111014744146, "grad_norm": 0.29126492142677307, "learning_rate": 4.6253445524445094e-05, "loss": 0.0201, "step": 14919 }, { "epoch": 1.6175195143104943, "grad_norm": 0.17263871431350708, "learning_rate": 4.624981865660816e-05, "loss": 0.0142, "step": 14920 }, { "epoch": 1.617627927146574, "grad_norm": 0.1942504346370697, "learning_rate": 4.624619178877122e-05, "loss": 0.0147, "step": 14921 }, { "epoch": 1.6177363399826539, "grad_norm": 0.22551600635051727, "learning_rate": 4.6242564920934287e-05, "loss": 0.0162, "step": 14922 }, { "epoch": 1.6178447528187339, "grad_norm": 0.3905290365219116, "learning_rate": 4.6238938053097344e-05, "loss": 0.0438, "step": 14923 }, { "epoch": 1.6179531656548134, "grad_norm": 0.31162694096565247, "learning_rate": 4.623531118526041e-05, "loss": 0.0209, "step": 14924 }, { "epoch": 1.6180615784908934, "grad_norm": 0.42741942405700684, "learning_rate": 4.623168431742347e-05, "loss": 0.0346, "step": 14925 }, { "epoch": 1.6181699913269731, "grad_norm": 0.6599777936935425, "learning_rate": 4.6228057449586536e-05, "loss": 0.0348, "step": 14926 }, { "epoch": 1.618278404163053, "grad_norm": 0.25290772318840027, "learning_rate": 4.622443058174961e-05, "loss": 0.0139, "step": 14927 }, { "epoch": 1.6183868169991327, "grad_norm": 0.1905870884656906, "learning_rate": 4.622080371391267e-05, "loss": 0.0181, "step": 14928 }, { "epoch": 1.6184952298352124, "grad_norm": 0.26017218828201294, "learning_rate": 4.6217176846075735e-05, "loss": 0.0153, "step": 14929 }, { "epoch": 1.6186036426712924, "grad_norm": 0.30988311767578125, "learning_rate": 4.621354997823879e-05, "loss": 0.0268, "step": 14930 }, { "epoch": 1.618712055507372, "grad_norm": 0.7569013833999634, "learning_rate": 4.620992311040186e-05, "loss": 0.0506, "step": 14931 }, { "epoch": 1.618820468343452, "grad_norm": 0.17680849134922028, "learning_rate": 4.620629624256492e-05, "loss": 0.0215, "step": 14932 }, { "epoch": 1.6189288811795317, "grad_norm": 0.8537346124649048, "learning_rate": 4.6202669374727985e-05, "loss": 0.0741, "step": 14933 }, { "epoch": 1.6190372940156115, "grad_norm": 0.4818088114261627, "learning_rate": 4.619904250689105e-05, "loss": 0.0155, "step": 14934 }, { "epoch": 1.6191457068516912, "grad_norm": 0.47026702761650085, "learning_rate": 4.619541563905411e-05, "loss": 0.026, "step": 14935 }, { "epoch": 1.619254119687771, "grad_norm": 0.4477614462375641, "learning_rate": 4.6191788771217184e-05, "loss": 0.0337, "step": 14936 }, { "epoch": 1.619362532523851, "grad_norm": 0.08912305533885956, "learning_rate": 4.618816190338025e-05, "loss": 0.0048, "step": 14937 }, { "epoch": 1.6194709453599305, "grad_norm": 0.11228589713573456, "learning_rate": 4.6184535035543305e-05, "loss": 0.0074, "step": 14938 }, { "epoch": 1.6195793581960105, "grad_norm": 0.578609049320221, "learning_rate": 4.618090816770637e-05, "loss": 0.0323, "step": 14939 }, { "epoch": 1.61968777103209, "grad_norm": 0.46783003211021423, "learning_rate": 4.6177281299869434e-05, "loss": 0.024, "step": 14940 }, { "epoch": 1.61979618386817, "grad_norm": 0.16029053926467896, "learning_rate": 4.61736544320325e-05, "loss": 0.0083, "step": 14941 }, { "epoch": 1.6199045967042498, "grad_norm": 0.3855213224887848, "learning_rate": 4.617002756419556e-05, "loss": 0.028, "step": 14942 }, { "epoch": 1.6200130095403296, "grad_norm": 0.44208240509033203, "learning_rate": 4.6166400696358626e-05, "loss": 0.0288, "step": 14943 }, { "epoch": 1.6201214223764093, "grad_norm": 0.3493199646472931, "learning_rate": 4.616277382852169e-05, "loss": 0.0161, "step": 14944 }, { "epoch": 1.620229835212489, "grad_norm": 0.4662051796913147, "learning_rate": 4.6159146960684754e-05, "loss": 0.0331, "step": 14945 }, { "epoch": 1.620338248048569, "grad_norm": 0.3626248836517334, "learning_rate": 4.615552009284782e-05, "loss": 0.0205, "step": 14946 }, { "epoch": 1.6204466608846486, "grad_norm": 0.24033895134925842, "learning_rate": 4.615189322501088e-05, "loss": 0.0219, "step": 14947 }, { "epoch": 1.6205550737207286, "grad_norm": 0.311353474855423, "learning_rate": 4.6148266357173946e-05, "loss": 0.0165, "step": 14948 }, { "epoch": 1.6206634865568084, "grad_norm": 0.17637033760547638, "learning_rate": 4.614463948933701e-05, "loss": 0.0134, "step": 14949 }, { "epoch": 1.6207718993928881, "grad_norm": 0.44095197319984436, "learning_rate": 4.6141012621500075e-05, "loss": 0.0253, "step": 14950 }, { "epoch": 1.620880312228968, "grad_norm": 0.35997915267944336, "learning_rate": 4.613738575366314e-05, "loss": 0.0208, "step": 14951 }, { "epoch": 1.6209887250650477, "grad_norm": 0.38836947083473206, "learning_rate": 4.61337588858262e-05, "loss": 0.0182, "step": 14952 }, { "epoch": 1.6210971379011276, "grad_norm": 0.23788125813007355, "learning_rate": 4.613013201798927e-05, "loss": 0.0591, "step": 14953 }, { "epoch": 1.6212055507372072, "grad_norm": 0.2399986982345581, "learning_rate": 4.612650515015233e-05, "loss": 0.0207, "step": 14954 }, { "epoch": 1.6213139635732872, "grad_norm": 0.11521443724632263, "learning_rate": 4.6122878282315395e-05, "loss": 0.0075, "step": 14955 }, { "epoch": 1.6214223764093667, "grad_norm": 0.2582269310951233, "learning_rate": 4.611925141447846e-05, "loss": 0.0277, "step": 14956 }, { "epoch": 1.6215307892454467, "grad_norm": 0.16231898963451385, "learning_rate": 4.6115624546641523e-05, "loss": 0.0126, "step": 14957 }, { "epoch": 1.6216392020815265, "grad_norm": 0.32884421944618225, "learning_rate": 4.611199767880459e-05, "loss": 0.0295, "step": 14958 }, { "epoch": 1.6217476149176062, "grad_norm": 0.38528913259506226, "learning_rate": 4.610837081096765e-05, "loss": 0.0317, "step": 14959 }, { "epoch": 1.6218560277536862, "grad_norm": 0.4630066156387329, "learning_rate": 4.6104743943130716e-05, "loss": 0.0342, "step": 14960 }, { "epoch": 1.6219644405897657, "grad_norm": 0.32237181067466736, "learning_rate": 4.610111707529377e-05, "loss": 0.0474, "step": 14961 }, { "epoch": 1.6220728534258457, "grad_norm": 0.14773869514465332, "learning_rate": 4.6097490207456844e-05, "loss": 0.0107, "step": 14962 }, { "epoch": 1.6221812662619253, "grad_norm": 0.052002277225255966, "learning_rate": 4.609386333961991e-05, "loss": 0.0041, "step": 14963 }, { "epoch": 1.6222896790980053, "grad_norm": 0.5000280737876892, "learning_rate": 4.609023647178297e-05, "loss": 0.0515, "step": 14964 }, { "epoch": 1.622398091934085, "grad_norm": 0.6475850343704224, "learning_rate": 4.6086609603946036e-05, "loss": 0.0223, "step": 14965 }, { "epoch": 1.6225065047701648, "grad_norm": 0.25660231709480286, "learning_rate": 4.60829827361091e-05, "loss": 0.0264, "step": 14966 }, { "epoch": 1.6226149176062445, "grad_norm": 0.06723209470510483, "learning_rate": 4.6079355868272164e-05, "loss": 0.0048, "step": 14967 }, { "epoch": 1.6227233304423243, "grad_norm": 0.21133001148700714, "learning_rate": 4.607572900043522e-05, "loss": 0.0229, "step": 14968 }, { "epoch": 1.6228317432784043, "grad_norm": 0.16584812104701996, "learning_rate": 4.6072102132598286e-05, "loss": 0.0148, "step": 14969 }, { "epoch": 1.6229401561144838, "grad_norm": 0.15894056856632233, "learning_rate": 4.606847526476135e-05, "loss": 0.0116, "step": 14970 }, { "epoch": 1.6230485689505638, "grad_norm": 0.3751060664653778, "learning_rate": 4.606484839692442e-05, "loss": 0.037, "step": 14971 }, { "epoch": 1.6231569817866436, "grad_norm": 0.6127163767814636, "learning_rate": 4.6061221529087485e-05, "loss": 0.0213, "step": 14972 }, { "epoch": 1.6232653946227233, "grad_norm": 0.5327471494674683, "learning_rate": 4.605759466125055e-05, "loss": 0.0264, "step": 14973 }, { "epoch": 1.623373807458803, "grad_norm": 0.42952367663383484, "learning_rate": 4.605396779341361e-05, "loss": 0.0173, "step": 14974 }, { "epoch": 1.6234822202948829, "grad_norm": 0.4025755524635315, "learning_rate": 4.605034092557667e-05, "loss": 0.0354, "step": 14975 }, { "epoch": 1.6235906331309629, "grad_norm": 0.27456212043762207, "learning_rate": 4.6046714057739735e-05, "loss": 0.0155, "step": 14976 }, { "epoch": 1.6236990459670424, "grad_norm": 0.33670520782470703, "learning_rate": 4.60430871899028e-05, "loss": 0.0309, "step": 14977 }, { "epoch": 1.6238074588031224, "grad_norm": 0.1300262063741684, "learning_rate": 4.603946032206586e-05, "loss": 0.0044, "step": 14978 }, { "epoch": 1.623915871639202, "grad_norm": 0.2225864827632904, "learning_rate": 4.6035833454228934e-05, "loss": 0.0147, "step": 14979 }, { "epoch": 1.624024284475282, "grad_norm": 0.17383186519145966, "learning_rate": 4.6032206586392e-05, "loss": 0.0181, "step": 14980 }, { "epoch": 1.6241326973113617, "grad_norm": 0.08576740324497223, "learning_rate": 4.602857971855506e-05, "loss": 0.0085, "step": 14981 }, { "epoch": 1.6242411101474414, "grad_norm": 0.19383150339126587, "learning_rate": 4.6024952850718126e-05, "loss": 0.0144, "step": 14982 }, { "epoch": 1.6243495229835212, "grad_norm": 0.5090806484222412, "learning_rate": 4.602132598288118e-05, "loss": 0.1058, "step": 14983 }, { "epoch": 1.624457935819601, "grad_norm": 0.5641025900840759, "learning_rate": 4.601769911504425e-05, "loss": 0.0113, "step": 14984 }, { "epoch": 1.624566348655681, "grad_norm": 0.07062377780675888, "learning_rate": 4.601407224720731e-05, "loss": 0.0044, "step": 14985 }, { "epoch": 1.6246747614917605, "grad_norm": 0.3402649760246277, "learning_rate": 4.6010445379370376e-05, "loss": 0.0353, "step": 14986 }, { "epoch": 1.6247831743278405, "grad_norm": 0.25879696011543274, "learning_rate": 4.600681851153344e-05, "loss": 0.0281, "step": 14987 }, { "epoch": 1.6248915871639202, "grad_norm": 0.36756718158721924, "learning_rate": 4.600319164369651e-05, "loss": 0.0276, "step": 14988 }, { "epoch": 1.625, "grad_norm": 0.34935620427131653, "learning_rate": 4.5999564775859575e-05, "loss": 0.0139, "step": 14989 }, { "epoch": 1.6251084128360798, "grad_norm": 0.10579422116279602, "learning_rate": 4.599593790802263e-05, "loss": 0.0105, "step": 14990 }, { "epoch": 1.6252168256721595, "grad_norm": 0.15586557984352112, "learning_rate": 4.5992311040185696e-05, "loss": 0.014, "step": 14991 }, { "epoch": 1.6253252385082395, "grad_norm": 0.2777163088321686, "learning_rate": 4.598868417234876e-05, "loss": 0.0296, "step": 14992 }, { "epoch": 1.625433651344319, "grad_norm": 0.29338744282722473, "learning_rate": 4.5985057304511824e-05, "loss": 0.0132, "step": 14993 }, { "epoch": 1.625542064180399, "grad_norm": 0.3871351480484009, "learning_rate": 4.598143043667489e-05, "loss": 0.0391, "step": 14994 }, { "epoch": 1.6256504770164788, "grad_norm": 0.10124139487743378, "learning_rate": 4.597780356883795e-05, "loss": 0.0071, "step": 14995 }, { "epoch": 1.6257588898525586, "grad_norm": 0.16348791122436523, "learning_rate": 4.5974176701001024e-05, "loss": 0.0064, "step": 14996 }, { "epoch": 1.6258673026886383, "grad_norm": 0.2807394862174988, "learning_rate": 4.597054983316408e-05, "loss": 0.0301, "step": 14997 }, { "epoch": 1.625975715524718, "grad_norm": 0.41638314723968506, "learning_rate": 4.5966922965327145e-05, "loss": 0.0593, "step": 14998 }, { "epoch": 1.626084128360798, "grad_norm": 0.04563359543681145, "learning_rate": 4.596329609749021e-05, "loss": 0.0036, "step": 14999 }, { "epoch": 1.6261925411968776, "grad_norm": 0.15659962594509125, "learning_rate": 4.595966922965327e-05, "loss": 0.0102, "step": 15000 }, { "epoch": 1.6263009540329576, "grad_norm": 0.12303663790225983, "learning_rate": 4.595604236181634e-05, "loss": 0.0068, "step": 15001 }, { "epoch": 1.6264093668690371, "grad_norm": 0.29469263553619385, "learning_rate": 4.59524154939794e-05, "loss": 0.0147, "step": 15002 }, { "epoch": 1.6265177797051171, "grad_norm": 0.6141483187675476, "learning_rate": 4.5948788626142465e-05, "loss": 0.0418, "step": 15003 }, { "epoch": 1.626626192541197, "grad_norm": 0.3288092613220215, "learning_rate": 4.594516175830553e-05, "loss": 0.031, "step": 15004 }, { "epoch": 1.6267346053772767, "grad_norm": 0.38700273633003235, "learning_rate": 4.5941534890468594e-05, "loss": 0.0239, "step": 15005 }, { "epoch": 1.6268430182133564, "grad_norm": 0.35452136397361755, "learning_rate": 4.593790802263166e-05, "loss": 0.0162, "step": 15006 }, { "epoch": 1.6269514310494362, "grad_norm": 0.2743440568447113, "learning_rate": 4.593428115479472e-05, "loss": 0.0263, "step": 15007 }, { "epoch": 1.6270598438855162, "grad_norm": 0.10774027556180954, "learning_rate": 4.5930654286957786e-05, "loss": 0.008, "step": 15008 }, { "epoch": 1.6271682567215957, "grad_norm": 0.2088792324066162, "learning_rate": 4.592702741912085e-05, "loss": 0.016, "step": 15009 }, { "epoch": 1.6272766695576757, "grad_norm": 0.26499491930007935, "learning_rate": 4.5923400551283914e-05, "loss": 0.054, "step": 15010 }, { "epoch": 1.6273850823937555, "grad_norm": 0.12697191536426544, "learning_rate": 4.591977368344698e-05, "loss": 0.0066, "step": 15011 }, { "epoch": 1.6274934952298352, "grad_norm": 0.41363635659217834, "learning_rate": 4.591614681561004e-05, "loss": 0.049, "step": 15012 }, { "epoch": 1.627601908065915, "grad_norm": 0.3485775291919708, "learning_rate": 4.59125199477731e-05, "loss": 0.031, "step": 15013 }, { "epoch": 1.6277103209019947, "grad_norm": 0.3314465582370758, "learning_rate": 4.590889307993617e-05, "loss": 0.0152, "step": 15014 }, { "epoch": 1.6278187337380747, "grad_norm": 0.11751196533441544, "learning_rate": 4.5905266212099235e-05, "loss": 0.0161, "step": 15015 }, { "epoch": 1.6279271465741543, "grad_norm": 0.2412732094526291, "learning_rate": 4.59016393442623e-05, "loss": 0.0346, "step": 15016 }, { "epoch": 1.6280355594102343, "grad_norm": 0.1151905283331871, "learning_rate": 4.589801247642536e-05, "loss": 0.0083, "step": 15017 }, { "epoch": 1.6281439722463138, "grad_norm": 0.9211458563804626, "learning_rate": 4.589438560858843e-05, "loss": 0.0191, "step": 15018 }, { "epoch": 1.6282523850823938, "grad_norm": 0.2139294296503067, "learning_rate": 4.589075874075149e-05, "loss": 0.0191, "step": 15019 }, { "epoch": 1.6283607979184735, "grad_norm": 0.20790955424308777, "learning_rate": 4.588713187291455e-05, "loss": 0.0179, "step": 15020 }, { "epoch": 1.6284692107545533, "grad_norm": 0.34306249022483826, "learning_rate": 4.588350500507761e-05, "loss": 0.0269, "step": 15021 }, { "epoch": 1.6285776235906333, "grad_norm": 0.13464520871639252, "learning_rate": 4.5879878137240683e-05, "loss": 0.0072, "step": 15022 }, { "epoch": 1.6286860364267128, "grad_norm": 0.3935766816139221, "learning_rate": 4.587625126940375e-05, "loss": 0.03, "step": 15023 }, { "epoch": 1.6287944492627928, "grad_norm": 0.23894934356212616, "learning_rate": 4.587262440156681e-05, "loss": 0.0413, "step": 15024 }, { "epoch": 1.6289028620988724, "grad_norm": 0.4596292972564697, "learning_rate": 4.5868997533729876e-05, "loss": 0.0505, "step": 15025 }, { "epoch": 1.6290112749349523, "grad_norm": 0.025236032903194427, "learning_rate": 4.586537066589294e-05, "loss": 0.0015, "step": 15026 }, { "epoch": 1.629119687771032, "grad_norm": 0.20791028439998627, "learning_rate": 4.5861743798056e-05, "loss": 0.0143, "step": 15027 }, { "epoch": 1.6292281006071119, "grad_norm": 0.42731812596321106, "learning_rate": 4.585811693021906e-05, "loss": 0.042, "step": 15028 }, { "epoch": 1.6293365134431916, "grad_norm": 0.28887760639190674, "learning_rate": 4.5854490062382125e-05, "loss": 0.0218, "step": 15029 }, { "epoch": 1.6294449262792714, "grad_norm": 0.2653134763240814, "learning_rate": 4.585086319454519e-05, "loss": 0.0163, "step": 15030 }, { "epoch": 1.6295533391153514, "grad_norm": 0.6109593510627747, "learning_rate": 4.584723632670826e-05, "loss": 0.0446, "step": 15031 }, { "epoch": 1.629661751951431, "grad_norm": 0.29977139830589294, "learning_rate": 4.5843609458871325e-05, "loss": 0.0195, "step": 15032 }, { "epoch": 1.629770164787511, "grad_norm": 0.3177553415298462, "learning_rate": 4.583998259103439e-05, "loss": 0.019, "step": 15033 }, { "epoch": 1.6298785776235907, "grad_norm": 0.17894436419010162, "learning_rate": 4.583635572319745e-05, "loss": 0.0126, "step": 15034 }, { "epoch": 1.6299869904596704, "grad_norm": 0.5922881960868835, "learning_rate": 4.583272885536051e-05, "loss": 0.0453, "step": 15035 }, { "epoch": 1.6300954032957502, "grad_norm": 0.21430206298828125, "learning_rate": 4.5829101987523574e-05, "loss": 0.0129, "step": 15036 }, { "epoch": 1.63020381613183, "grad_norm": 0.2932552099227905, "learning_rate": 4.582547511968664e-05, "loss": 0.0353, "step": 15037 }, { "epoch": 1.63031222896791, "grad_norm": 0.23155976831912994, "learning_rate": 4.58218482518497e-05, "loss": 0.0167, "step": 15038 }, { "epoch": 1.6304206418039895, "grad_norm": 0.36882534623146057, "learning_rate": 4.5818221384012766e-05, "loss": 0.0443, "step": 15039 }, { "epoch": 1.6305290546400695, "grad_norm": 0.4774071276187897, "learning_rate": 4.581459451617584e-05, "loss": 0.044, "step": 15040 }, { "epoch": 1.630637467476149, "grad_norm": 0.5113899111747742, "learning_rate": 4.58109676483389e-05, "loss": 0.0313, "step": 15041 }, { "epoch": 1.630745880312229, "grad_norm": 0.2866593599319458, "learning_rate": 4.580734078050196e-05, "loss": 0.0118, "step": 15042 }, { "epoch": 1.6308542931483088, "grad_norm": 0.0922427624464035, "learning_rate": 4.580371391266502e-05, "loss": 0.0046, "step": 15043 }, { "epoch": 1.6309627059843885, "grad_norm": 0.07975433766841888, "learning_rate": 4.580008704482809e-05, "loss": 0.004, "step": 15044 }, { "epoch": 1.6310711188204683, "grad_norm": 0.43712317943573, "learning_rate": 4.579646017699115e-05, "loss": 0.0196, "step": 15045 }, { "epoch": 1.631179531656548, "grad_norm": 0.615354597568512, "learning_rate": 4.5792833309154215e-05, "loss": 0.0385, "step": 15046 }, { "epoch": 1.631287944492628, "grad_norm": 0.10833761096000671, "learning_rate": 4.578920644131728e-05, "loss": 0.0095, "step": 15047 }, { "epoch": 1.6313963573287076, "grad_norm": 0.38506925106048584, "learning_rate": 4.578557957348035e-05, "loss": 0.0375, "step": 15048 }, { "epoch": 1.6315047701647876, "grad_norm": 0.2680344879627228, "learning_rate": 4.578195270564341e-05, "loss": 0.0272, "step": 15049 }, { "epoch": 1.6316131830008673, "grad_norm": 0.07607005536556244, "learning_rate": 4.577832583780647e-05, "loss": 0.0066, "step": 15050 }, { "epoch": 1.631721595836947, "grad_norm": 0.23648898303508759, "learning_rate": 4.5774698969969536e-05, "loss": 0.0086, "step": 15051 }, { "epoch": 1.6318300086730269, "grad_norm": 0.42018258571624756, "learning_rate": 4.57710721021326e-05, "loss": 0.0281, "step": 15052 }, { "epoch": 1.6319384215091066, "grad_norm": 0.505729615688324, "learning_rate": 4.5767445234295664e-05, "loss": 0.0224, "step": 15053 }, { "epoch": 1.6320468343451866, "grad_norm": 0.43699610233306885, "learning_rate": 4.576381836645873e-05, "loss": 0.0233, "step": 15054 }, { "epoch": 1.6321552471812661, "grad_norm": 0.20756058394908905, "learning_rate": 4.576019149862179e-05, "loss": 0.0107, "step": 15055 }, { "epoch": 1.6322636600173461, "grad_norm": 0.18712693452835083, "learning_rate": 4.5756564630784856e-05, "loss": 0.0166, "step": 15056 }, { "epoch": 1.632372072853426, "grad_norm": 0.22945795953273773, "learning_rate": 4.575293776294792e-05, "loss": 0.0253, "step": 15057 }, { "epoch": 1.6324804856895057, "grad_norm": 0.17093877494335175, "learning_rate": 4.5749310895110984e-05, "loss": 0.0152, "step": 15058 }, { "epoch": 1.6325888985255854, "grad_norm": 0.8181793093681335, "learning_rate": 4.574568402727405e-05, "loss": 0.0782, "step": 15059 }, { "epoch": 1.6326973113616652, "grad_norm": 0.253375381231308, "learning_rate": 4.574205715943711e-05, "loss": 0.0086, "step": 15060 }, { "epoch": 1.6328057241977452, "grad_norm": 0.6136024594306946, "learning_rate": 4.573843029160018e-05, "loss": 0.0365, "step": 15061 }, { "epoch": 1.6329141370338247, "grad_norm": 0.18971309065818787, "learning_rate": 4.573480342376324e-05, "loss": 0.0089, "step": 15062 }, { "epoch": 1.6330225498699047, "grad_norm": 0.3575776219367981, "learning_rate": 4.5731176555926305e-05, "loss": 0.0227, "step": 15063 }, { "epoch": 1.6331309627059842, "grad_norm": 0.08251038938760757, "learning_rate": 4.572754968808937e-05, "loss": 0.0027, "step": 15064 }, { "epoch": 1.6332393755420642, "grad_norm": 0.27500587701797485, "learning_rate": 4.5723922820252426e-05, "loss": 0.0262, "step": 15065 }, { "epoch": 1.633347788378144, "grad_norm": 0.33565059304237366, "learning_rate": 4.57202959524155e-05, "loss": 0.0181, "step": 15066 }, { "epoch": 1.6334562012142237, "grad_norm": 0.5645992755889893, "learning_rate": 4.571666908457856e-05, "loss": 0.0393, "step": 15067 }, { "epoch": 1.6335646140503035, "grad_norm": 0.376102089881897, "learning_rate": 4.5713042216741626e-05, "loss": 0.0277, "step": 15068 }, { "epoch": 1.6336730268863833, "grad_norm": 0.6197220087051392, "learning_rate": 4.570941534890469e-05, "loss": 0.026, "step": 15069 }, { "epoch": 1.6337814397224633, "grad_norm": 0.3617229759693146, "learning_rate": 4.5705788481067754e-05, "loss": 0.0148, "step": 15070 }, { "epoch": 1.6338898525585428, "grad_norm": 0.26673492789268494, "learning_rate": 4.570216161323082e-05, "loss": 0.02, "step": 15071 }, { "epoch": 1.6339982653946228, "grad_norm": 0.35198071599006653, "learning_rate": 4.5698534745393875e-05, "loss": 0.0387, "step": 15072 }, { "epoch": 1.6341066782307025, "grad_norm": 0.11340945214033127, "learning_rate": 4.569490787755694e-05, "loss": 0.013, "step": 15073 }, { "epoch": 1.6342150910667823, "grad_norm": 0.5412535667419434, "learning_rate": 4.569128100972001e-05, "loss": 0.0583, "step": 15074 }, { "epoch": 1.634323503902862, "grad_norm": 0.23431886732578278, "learning_rate": 4.5687654141883074e-05, "loss": 0.0092, "step": 15075 }, { "epoch": 1.6344319167389418, "grad_norm": 0.38891515135765076, "learning_rate": 4.568402727404614e-05, "loss": 0.0382, "step": 15076 }, { "epoch": 1.6345403295750218, "grad_norm": 0.1479424685239792, "learning_rate": 4.56804004062092e-05, "loss": 0.018, "step": 15077 }, { "epoch": 1.6346487424111014, "grad_norm": 0.28544098138809204, "learning_rate": 4.5676773538372267e-05, "loss": 0.0346, "step": 15078 }, { "epoch": 1.6347571552471813, "grad_norm": 0.18811717629432678, "learning_rate": 4.5673146670535324e-05, "loss": 0.016, "step": 15079 }, { "epoch": 1.6348655680832611, "grad_norm": 0.37496599555015564, "learning_rate": 4.566951980269839e-05, "loss": 0.0387, "step": 15080 }, { "epoch": 1.6349739809193409, "grad_norm": 0.22946466505527496, "learning_rate": 4.566589293486145e-05, "loss": 0.013, "step": 15081 }, { "epoch": 1.6350823937554206, "grad_norm": 0.6879671216011047, "learning_rate": 4.5662266067024516e-05, "loss": 0.0166, "step": 15082 }, { "epoch": 1.6351908065915004, "grad_norm": 0.19668257236480713, "learning_rate": 4.565863919918759e-05, "loss": 0.0122, "step": 15083 }, { "epoch": 1.6352992194275804, "grad_norm": 0.27659377455711365, "learning_rate": 4.565501233135065e-05, "loss": 0.0164, "step": 15084 }, { "epoch": 1.63540763226366, "grad_norm": 0.5661901235580444, "learning_rate": 4.5651385463513715e-05, "loss": 0.023, "step": 15085 }, { "epoch": 1.63551604509974, "grad_norm": 0.5252227187156677, "learning_rate": 4.564775859567678e-05, "loss": 0.0315, "step": 15086 }, { "epoch": 1.6356244579358195, "grad_norm": 0.6581758260726929, "learning_rate": 4.564413172783984e-05, "loss": 0.0405, "step": 15087 }, { "epoch": 1.6357328707718994, "grad_norm": 0.37419041991233826, "learning_rate": 4.56405048600029e-05, "loss": 0.0249, "step": 15088 }, { "epoch": 1.6358412836079792, "grad_norm": 0.08412490040063858, "learning_rate": 4.5636877992165965e-05, "loss": 0.0053, "step": 15089 }, { "epoch": 1.635949696444059, "grad_norm": 0.3455710709095001, "learning_rate": 4.563325112432903e-05, "loss": 0.0195, "step": 15090 }, { "epoch": 1.6360581092801387, "grad_norm": 0.1057930663228035, "learning_rate": 4.56296242564921e-05, "loss": 0.0061, "step": 15091 }, { "epoch": 1.6361665221162185, "grad_norm": 0.2192779779434204, "learning_rate": 4.5625997388655164e-05, "loss": 0.0265, "step": 15092 }, { "epoch": 1.6362749349522985, "grad_norm": 0.05363743379712105, "learning_rate": 4.562237052081823e-05, "loss": 0.0021, "step": 15093 }, { "epoch": 1.636383347788378, "grad_norm": 0.2179044932126999, "learning_rate": 4.5618743652981285e-05, "loss": 0.012, "step": 15094 }, { "epoch": 1.636491760624458, "grad_norm": 0.2894279658794403, "learning_rate": 4.561511678514435e-05, "loss": 0.0291, "step": 15095 }, { "epoch": 1.6366001734605378, "grad_norm": 0.018643975257873535, "learning_rate": 4.5611489917307414e-05, "loss": 0.0011, "step": 15096 }, { "epoch": 1.6367085862966175, "grad_norm": 0.3010213375091553, "learning_rate": 4.560786304947048e-05, "loss": 0.0366, "step": 15097 }, { "epoch": 1.6368169991326973, "grad_norm": 1.1781095266342163, "learning_rate": 4.560423618163354e-05, "loss": 0.0358, "step": 15098 }, { "epoch": 1.636925411968777, "grad_norm": 0.3373492658138275, "learning_rate": 4.5600609313796606e-05, "loss": 0.0257, "step": 15099 }, { "epoch": 1.637033824804857, "grad_norm": 0.4703616499900818, "learning_rate": 4.559698244595968e-05, "loss": 0.0148, "step": 15100 }, { "epoch": 1.6371422376409366, "grad_norm": 0.0785907506942749, "learning_rate": 4.5593355578122734e-05, "loss": 0.0078, "step": 15101 }, { "epoch": 1.6372506504770166, "grad_norm": 0.48685988783836365, "learning_rate": 4.55897287102858e-05, "loss": 0.0131, "step": 15102 }, { "epoch": 1.637359063313096, "grad_norm": 0.44408565759658813, "learning_rate": 4.558610184244886e-05, "loss": 0.0254, "step": 15103 }, { "epoch": 1.637467476149176, "grad_norm": 0.17628154158592224, "learning_rate": 4.5582474974611927e-05, "loss": 0.0089, "step": 15104 }, { "epoch": 1.6375758889852559, "grad_norm": 0.6650990843772888, "learning_rate": 4.557884810677499e-05, "loss": 0.0276, "step": 15105 }, { "epoch": 1.6376843018213356, "grad_norm": 0.42481911182403564, "learning_rate": 4.5575221238938055e-05, "loss": 0.0187, "step": 15106 }, { "epoch": 1.6377927146574154, "grad_norm": 0.037151992321014404, "learning_rate": 4.557159437110112e-05, "loss": 0.0016, "step": 15107 }, { "epoch": 1.6379011274934951, "grad_norm": 0.18999069929122925, "learning_rate": 4.556796750326418e-05, "loss": 0.0069, "step": 15108 }, { "epoch": 1.6380095403295751, "grad_norm": 0.3826906085014343, "learning_rate": 4.556434063542725e-05, "loss": 0.0224, "step": 15109 }, { "epoch": 1.6381179531656547, "grad_norm": 0.1875421702861786, "learning_rate": 4.556071376759031e-05, "loss": 0.015, "step": 15110 }, { "epoch": 1.6382263660017347, "grad_norm": 0.1846095621585846, "learning_rate": 4.5557086899753375e-05, "loss": 0.0153, "step": 15111 }, { "epoch": 1.6383347788378144, "grad_norm": 1.5303573608398438, "learning_rate": 4.555346003191644e-05, "loss": 0.026, "step": 15112 }, { "epoch": 1.6384431916738942, "grad_norm": 0.33624207973480225, "learning_rate": 4.5549833164079503e-05, "loss": 0.0517, "step": 15113 }, { "epoch": 1.638551604509974, "grad_norm": 0.23502585291862488, "learning_rate": 4.554620629624257e-05, "loss": 0.0097, "step": 15114 }, { "epoch": 1.6386600173460537, "grad_norm": 0.24570247530937195, "learning_rate": 4.554257942840563e-05, "loss": 0.0231, "step": 15115 }, { "epoch": 1.6387684301821337, "grad_norm": 0.09474168717861176, "learning_rate": 4.5538952560568696e-05, "loss": 0.007, "step": 15116 }, { "epoch": 1.6388768430182132, "grad_norm": 0.06633303314447403, "learning_rate": 4.553532569273176e-05, "loss": 0.0034, "step": 15117 }, { "epoch": 1.6389852558542932, "grad_norm": 0.5623372197151184, "learning_rate": 4.5531698824894824e-05, "loss": 0.0826, "step": 15118 }, { "epoch": 1.639093668690373, "grad_norm": 0.19982729852199554, "learning_rate": 4.552807195705789e-05, "loss": 0.0252, "step": 15119 }, { "epoch": 1.6392020815264527, "grad_norm": 0.23550830781459808, "learning_rate": 4.552444508922095e-05, "loss": 0.0298, "step": 15120 }, { "epoch": 1.6393104943625325, "grad_norm": 0.6257050633430481, "learning_rate": 4.5520818221384016e-05, "loss": 0.0282, "step": 15121 }, { "epoch": 1.6394189071986123, "grad_norm": 0.3878760039806366, "learning_rate": 4.551719135354708e-05, "loss": 0.0761, "step": 15122 }, { "epoch": 1.6395273200346923, "grad_norm": 0.30707207322120667, "learning_rate": 4.5513564485710145e-05, "loss": 0.0165, "step": 15123 }, { "epoch": 1.6396357328707718, "grad_norm": 0.1090986430644989, "learning_rate": 4.55099376178732e-05, "loss": 0.0063, "step": 15124 }, { "epoch": 1.6397441457068518, "grad_norm": 0.32729315757751465, "learning_rate": 4.5506310750036266e-05, "loss": 0.0132, "step": 15125 }, { "epoch": 1.6398525585429313, "grad_norm": 0.6849668025970459, "learning_rate": 4.550268388219934e-05, "loss": 0.0111, "step": 15126 }, { "epoch": 1.6399609713790113, "grad_norm": 0.26090458035469055, "learning_rate": 4.54990570143624e-05, "loss": 0.0197, "step": 15127 }, { "epoch": 1.640069384215091, "grad_norm": 0.2640126049518585, "learning_rate": 4.5495430146525465e-05, "loss": 0.0272, "step": 15128 }, { "epoch": 1.6401777970511708, "grad_norm": 0.3098751902580261, "learning_rate": 4.549180327868853e-05, "loss": 0.0287, "step": 15129 }, { "epoch": 1.6402862098872506, "grad_norm": 0.40946945548057556, "learning_rate": 4.548817641085159e-05, "loss": 0.0348, "step": 15130 }, { "epoch": 1.6403946227233304, "grad_norm": 0.22404305636882782, "learning_rate": 4.548454954301466e-05, "loss": 0.0265, "step": 15131 }, { "epoch": 1.6405030355594103, "grad_norm": 0.5634617209434509, "learning_rate": 4.5480922675177715e-05, "loss": 0.0109, "step": 15132 }, { "epoch": 1.64061144839549, "grad_norm": 0.09939304739236832, "learning_rate": 4.547729580734078e-05, "loss": 0.0092, "step": 15133 }, { "epoch": 1.6407198612315699, "grad_norm": 0.15338410437107086, "learning_rate": 4.547366893950384e-05, "loss": 0.011, "step": 15134 }, { "epoch": 1.6408282740676496, "grad_norm": 0.4447854459285736, "learning_rate": 4.5470042071666914e-05, "loss": 0.067, "step": 15135 }, { "epoch": 1.6409366869037294, "grad_norm": 0.15319669246673584, "learning_rate": 4.546641520382998e-05, "loss": 0.0072, "step": 15136 }, { "epoch": 1.6410450997398092, "grad_norm": 0.1715669184923172, "learning_rate": 4.546278833599304e-05, "loss": 0.0143, "step": 15137 }, { "epoch": 1.641153512575889, "grad_norm": 0.316989004611969, "learning_rate": 4.5459161468156106e-05, "loss": 0.0509, "step": 15138 }, { "epoch": 1.641261925411969, "grad_norm": 0.2589552700519562, "learning_rate": 4.5455534600319163e-05, "loss": 0.0323, "step": 15139 }, { "epoch": 1.6413703382480485, "grad_norm": 0.906112790107727, "learning_rate": 4.545190773248223e-05, "loss": 0.0651, "step": 15140 }, { "epoch": 1.6414787510841284, "grad_norm": 0.8929339051246643, "learning_rate": 4.544828086464529e-05, "loss": 0.0615, "step": 15141 }, { "epoch": 1.6415871639202082, "grad_norm": 0.22462525963783264, "learning_rate": 4.5444653996808356e-05, "loss": 0.0094, "step": 15142 }, { "epoch": 1.641695576756288, "grad_norm": 0.10419126600027084, "learning_rate": 4.544102712897143e-05, "loss": 0.0114, "step": 15143 }, { "epoch": 1.6418039895923677, "grad_norm": 0.6092696189880371, "learning_rate": 4.543740026113449e-05, "loss": 0.0929, "step": 15144 }, { "epoch": 1.6419124024284475, "grad_norm": 0.3345014750957489, "learning_rate": 4.5433773393297555e-05, "loss": 0.0202, "step": 15145 }, { "epoch": 1.6420208152645275, "grad_norm": 0.5130234956741333, "learning_rate": 4.543014652546061e-05, "loss": 0.0194, "step": 15146 }, { "epoch": 1.642129228100607, "grad_norm": 0.2488390952348709, "learning_rate": 4.5426519657623676e-05, "loss": 0.036, "step": 15147 }, { "epoch": 1.642237640936687, "grad_norm": 0.6846168637275696, "learning_rate": 4.542289278978674e-05, "loss": 0.0295, "step": 15148 }, { "epoch": 1.6423460537727665, "grad_norm": 0.720224916934967, "learning_rate": 4.5419265921949804e-05, "loss": 0.051, "step": 15149 }, { "epoch": 1.6424544666088465, "grad_norm": 0.10535364598035812, "learning_rate": 4.541563905411287e-05, "loss": 0.0062, "step": 15150 }, { "epoch": 1.6425628794449263, "grad_norm": 0.2216157466173172, "learning_rate": 4.541201218627593e-05, "loss": 0.0161, "step": 15151 }, { "epoch": 1.642671292281006, "grad_norm": 0.2612256705760956, "learning_rate": 4.5408385318439004e-05, "loss": 0.0419, "step": 15152 }, { "epoch": 1.6427797051170858, "grad_norm": 0.33430272340774536, "learning_rate": 4.540475845060206e-05, "loss": 0.0209, "step": 15153 }, { "epoch": 1.6428881179531656, "grad_norm": 0.2040027529001236, "learning_rate": 4.5401131582765125e-05, "loss": 0.0279, "step": 15154 }, { "epoch": 1.6429965307892456, "grad_norm": 0.43991750478744507, "learning_rate": 4.539750471492819e-05, "loss": 0.0302, "step": 15155 }, { "epoch": 1.643104943625325, "grad_norm": 0.13334117829799652, "learning_rate": 4.539387784709125e-05, "loss": 0.0062, "step": 15156 }, { "epoch": 1.643213356461405, "grad_norm": 0.07639867812395096, "learning_rate": 4.539025097925432e-05, "loss": 0.006, "step": 15157 }, { "epoch": 1.6433217692974849, "grad_norm": 0.38065314292907715, "learning_rate": 4.538662411141738e-05, "loss": 0.0243, "step": 15158 }, { "epoch": 1.6434301821335646, "grad_norm": 0.07422125339508057, "learning_rate": 4.5382997243580446e-05, "loss": 0.0042, "step": 15159 }, { "epoch": 1.6435385949696444, "grad_norm": 0.34853827953338623, "learning_rate": 4.537937037574351e-05, "loss": 0.0308, "step": 15160 }, { "epoch": 1.6436470078057241, "grad_norm": 0.17472918331623077, "learning_rate": 4.5375743507906574e-05, "loss": 0.0115, "step": 15161 }, { "epoch": 1.6437554206418041, "grad_norm": 0.25188595056533813, "learning_rate": 4.537211664006964e-05, "loss": 0.0123, "step": 15162 }, { "epoch": 1.6438638334778837, "grad_norm": 0.4494486451148987, "learning_rate": 4.53684897722327e-05, "loss": 0.0216, "step": 15163 }, { "epoch": 1.6439722463139637, "grad_norm": 0.45156630873680115, "learning_rate": 4.5364862904395766e-05, "loss": 0.0641, "step": 15164 }, { "epoch": 1.6440806591500432, "grad_norm": 0.39130231738090515, "learning_rate": 4.536123603655883e-05, "loss": 0.0492, "step": 15165 }, { "epoch": 1.6441890719861232, "grad_norm": 0.37795931100845337, "learning_rate": 4.5357609168721894e-05, "loss": 0.0359, "step": 15166 }, { "epoch": 1.644297484822203, "grad_norm": 0.590988039970398, "learning_rate": 4.535398230088496e-05, "loss": 0.0342, "step": 15167 }, { "epoch": 1.6444058976582827, "grad_norm": 0.4247991740703583, "learning_rate": 4.535035543304802e-05, "loss": 0.0122, "step": 15168 }, { "epoch": 1.6445143104943627, "grad_norm": 0.25153324007987976, "learning_rate": 4.5346728565211087e-05, "loss": 0.0388, "step": 15169 }, { "epoch": 1.6446227233304422, "grad_norm": 0.18451152741909027, "learning_rate": 4.534310169737415e-05, "loss": 0.0107, "step": 15170 }, { "epoch": 1.6447311361665222, "grad_norm": 0.34178006649017334, "learning_rate": 4.5339474829537215e-05, "loss": 0.0393, "step": 15171 }, { "epoch": 1.6448395490026018, "grad_norm": 0.3024612069129944, "learning_rate": 4.533584796170028e-05, "loss": 0.022, "step": 15172 }, { "epoch": 1.6449479618386817, "grad_norm": 0.34197551012039185, "learning_rate": 4.533222109386334e-05, "loss": 0.0223, "step": 15173 }, { "epoch": 1.6450563746747615, "grad_norm": 0.5556443333625793, "learning_rate": 4.532859422602641e-05, "loss": 0.0397, "step": 15174 }, { "epoch": 1.6451647875108413, "grad_norm": 0.09294982254505157, "learning_rate": 4.532496735818947e-05, "loss": 0.0125, "step": 15175 }, { "epoch": 1.645273200346921, "grad_norm": 0.4156886637210846, "learning_rate": 4.532134049035253e-05, "loss": 0.0528, "step": 15176 }, { "epoch": 1.6453816131830008, "grad_norm": 0.09991083294153214, "learning_rate": 4.531771362251559e-05, "loss": 0.0051, "step": 15177 }, { "epoch": 1.6454900260190808, "grad_norm": 0.058775171637535095, "learning_rate": 4.5314086754678664e-05, "loss": 0.0059, "step": 15178 }, { "epoch": 1.6455984388551603, "grad_norm": 0.15870939195156097, "learning_rate": 4.531045988684173e-05, "loss": 0.0156, "step": 15179 }, { "epoch": 1.6457068516912403, "grad_norm": 0.1186942458152771, "learning_rate": 4.530683301900479e-05, "loss": 0.0128, "step": 15180 }, { "epoch": 1.64581526452732, "grad_norm": 0.10155290365219116, "learning_rate": 4.5303206151167856e-05, "loss": 0.0085, "step": 15181 }, { "epoch": 1.6459236773633998, "grad_norm": 0.4150964915752411, "learning_rate": 4.529957928333092e-05, "loss": 0.0285, "step": 15182 }, { "epoch": 1.6460320901994796, "grad_norm": 0.40668606758117676, "learning_rate": 4.5295952415493984e-05, "loss": 0.0252, "step": 15183 }, { "epoch": 1.6461405030355594, "grad_norm": 0.17216746509075165, "learning_rate": 4.529232554765704e-05, "loss": 0.0097, "step": 15184 }, { "epoch": 1.6462489158716394, "grad_norm": 0.2970157563686371, "learning_rate": 4.5288698679820105e-05, "loss": 0.0355, "step": 15185 }, { "epoch": 1.646357328707719, "grad_norm": 0.3246195614337921, "learning_rate": 4.5285071811983176e-05, "loss": 0.0171, "step": 15186 }, { "epoch": 1.6464657415437989, "grad_norm": 0.5069884061813354, "learning_rate": 4.528144494414624e-05, "loss": 0.0318, "step": 15187 }, { "epoch": 1.6465741543798784, "grad_norm": 0.1728794127702713, "learning_rate": 4.5277818076309305e-05, "loss": 0.0116, "step": 15188 }, { "epoch": 1.6466825672159584, "grad_norm": 0.28070691227912903, "learning_rate": 4.527419120847237e-05, "loss": 0.0278, "step": 15189 }, { "epoch": 1.6467909800520382, "grad_norm": 0.16528014838695526, "learning_rate": 4.527056434063543e-05, "loss": 0.0127, "step": 15190 }, { "epoch": 1.646899392888118, "grad_norm": 0.3200712203979492, "learning_rate": 4.526693747279849e-05, "loss": 0.0284, "step": 15191 }, { "epoch": 1.6470078057241977, "grad_norm": 0.21916607022285461, "learning_rate": 4.5263310604961554e-05, "loss": 0.025, "step": 15192 }, { "epoch": 1.6471162185602775, "grad_norm": 0.4988439977169037, "learning_rate": 4.525968373712462e-05, "loss": 0.0152, "step": 15193 }, { "epoch": 1.6472246313963574, "grad_norm": 0.1647733896970749, "learning_rate": 4.525605686928768e-05, "loss": 0.0147, "step": 15194 }, { "epoch": 1.647333044232437, "grad_norm": 0.23105598986148834, "learning_rate": 4.525243000145075e-05, "loss": 0.019, "step": 15195 }, { "epoch": 1.647441457068517, "grad_norm": 0.29243752360343933, "learning_rate": 4.524880313361382e-05, "loss": 0.0219, "step": 15196 }, { "epoch": 1.6475498699045967, "grad_norm": 0.12195289134979248, "learning_rate": 4.524517626577688e-05, "loss": 0.013, "step": 15197 }, { "epoch": 1.6476582827406765, "grad_norm": 0.38637813925743103, "learning_rate": 4.524154939793994e-05, "loss": 0.0581, "step": 15198 }, { "epoch": 1.6477666955767563, "grad_norm": 0.16296780109405518, "learning_rate": 4.5237922530103e-05, "loss": 0.01, "step": 15199 }, { "epoch": 1.647875108412836, "grad_norm": 0.5053340196609497, "learning_rate": 4.523429566226607e-05, "loss": 0.0539, "step": 15200 }, { "epoch": 1.647983521248916, "grad_norm": 1.2397526502609253, "learning_rate": 4.523066879442913e-05, "loss": 0.0442, "step": 15201 }, { "epoch": 1.6480919340849955, "grad_norm": 0.5521852374076843, "learning_rate": 4.5227041926592195e-05, "loss": 0.0187, "step": 15202 }, { "epoch": 1.6482003469210755, "grad_norm": 0.07253696769475937, "learning_rate": 4.522341505875526e-05, "loss": 0.0073, "step": 15203 }, { "epoch": 1.6483087597571553, "grad_norm": 0.49237748980522156, "learning_rate": 4.521978819091833e-05, "loss": 0.0126, "step": 15204 }, { "epoch": 1.648417172593235, "grad_norm": 0.3924853801727295, "learning_rate": 4.521616132308139e-05, "loss": 0.0509, "step": 15205 }, { "epoch": 1.6485255854293148, "grad_norm": 0.15226635336875916, "learning_rate": 4.521253445524445e-05, "loss": 0.027, "step": 15206 }, { "epoch": 1.6486339982653946, "grad_norm": 0.16810472309589386, "learning_rate": 4.5208907587407516e-05, "loss": 0.0125, "step": 15207 }, { "epoch": 1.6487424111014746, "grad_norm": 0.12486934661865234, "learning_rate": 4.520528071957058e-05, "loss": 0.0092, "step": 15208 }, { "epoch": 1.648850823937554, "grad_norm": 0.5812859535217285, "learning_rate": 4.5201653851733644e-05, "loss": 0.04, "step": 15209 }, { "epoch": 1.648959236773634, "grad_norm": 0.08520323038101196, "learning_rate": 4.519802698389671e-05, "loss": 0.0037, "step": 15210 }, { "epoch": 1.6490676496097136, "grad_norm": 0.12026973068714142, "learning_rate": 4.519440011605977e-05, "loss": 0.0122, "step": 15211 }, { "epoch": 1.6491760624457936, "grad_norm": 0.3207480013370514, "learning_rate": 4.5190773248222836e-05, "loss": 0.0158, "step": 15212 }, { "epoch": 1.6492844752818734, "grad_norm": 0.1747267097234726, "learning_rate": 4.51871463803859e-05, "loss": 0.0292, "step": 15213 }, { "epoch": 1.6493928881179531, "grad_norm": 0.11511347442865372, "learning_rate": 4.5183519512548965e-05, "loss": 0.0088, "step": 15214 }, { "epoch": 1.649501300954033, "grad_norm": 0.2860160171985626, "learning_rate": 4.517989264471203e-05, "loss": 0.0503, "step": 15215 }, { "epoch": 1.6496097137901127, "grad_norm": 0.07409003376960754, "learning_rate": 4.517626577687509e-05, "loss": 0.0025, "step": 15216 }, { "epoch": 1.6497181266261927, "grad_norm": 0.2524983286857605, "learning_rate": 4.517263890903816e-05, "loss": 0.016, "step": 15217 }, { "epoch": 1.6498265394622722, "grad_norm": 0.2522779107093811, "learning_rate": 4.516901204120122e-05, "loss": 0.0222, "step": 15218 }, { "epoch": 1.6499349522983522, "grad_norm": 0.3480606973171234, "learning_rate": 4.5165385173364285e-05, "loss": 0.0296, "step": 15219 }, { "epoch": 1.650043365134432, "grad_norm": 0.3548356592655182, "learning_rate": 4.516175830552735e-05, "loss": 0.0205, "step": 15220 }, { "epoch": 1.6501517779705117, "grad_norm": 0.5809811353683472, "learning_rate": 4.515813143769041e-05, "loss": 0.0459, "step": 15221 }, { "epoch": 1.6502601908065915, "grad_norm": 0.49967625737190247, "learning_rate": 4.515450456985348e-05, "loss": 0.0446, "step": 15222 }, { "epoch": 1.6503686036426712, "grad_norm": 0.5103103518486023, "learning_rate": 4.515087770201654e-05, "loss": 0.0285, "step": 15223 }, { "epoch": 1.6504770164787512, "grad_norm": 0.7448820471763611, "learning_rate": 4.5147250834179606e-05, "loss": 0.0453, "step": 15224 }, { "epoch": 1.6505854293148308, "grad_norm": 0.2934132218360901, "learning_rate": 4.514362396634267e-05, "loss": 0.0156, "step": 15225 }, { "epoch": 1.6506938421509108, "grad_norm": 0.07423242926597595, "learning_rate": 4.5139997098505734e-05, "loss": 0.0042, "step": 15226 }, { "epoch": 1.6508022549869905, "grad_norm": 0.38234245777130127, "learning_rate": 4.51363702306688e-05, "loss": 0.0206, "step": 15227 }, { "epoch": 1.6509106678230703, "grad_norm": 0.008494534529745579, "learning_rate": 4.5132743362831855e-05, "loss": 0.0004, "step": 15228 }, { "epoch": 1.65101908065915, "grad_norm": 0.24051186442375183, "learning_rate": 4.512911649499492e-05, "loss": 0.0128, "step": 15229 }, { "epoch": 1.6511274934952298, "grad_norm": 0.35497069358825684, "learning_rate": 4.512548962715799e-05, "loss": 0.0901, "step": 15230 }, { "epoch": 1.6512359063313098, "grad_norm": 0.3893996477127075, "learning_rate": 4.5121862759321054e-05, "loss": 0.0194, "step": 15231 }, { "epoch": 1.6513443191673893, "grad_norm": 0.21045051515102386, "learning_rate": 4.511823589148412e-05, "loss": 0.0112, "step": 15232 }, { "epoch": 1.6514527320034693, "grad_norm": 0.20983436703681946, "learning_rate": 4.511460902364718e-05, "loss": 0.024, "step": 15233 }, { "epoch": 1.6515611448395489, "grad_norm": 0.4193582832813263, "learning_rate": 4.511098215581025e-05, "loss": 0.0248, "step": 15234 }, { "epoch": 1.6516695576756288, "grad_norm": 0.11249463260173798, "learning_rate": 4.510735528797331e-05, "loss": 0.0144, "step": 15235 }, { "epoch": 1.6517779705117086, "grad_norm": 0.42982056736946106, "learning_rate": 4.510372842013637e-05, "loss": 0.0303, "step": 15236 }, { "epoch": 1.6518863833477884, "grad_norm": 0.10134536772966385, "learning_rate": 4.510010155229943e-05, "loss": 0.0077, "step": 15237 }, { "epoch": 1.6519947961838681, "grad_norm": 0.31507542729377747, "learning_rate": 4.50964746844625e-05, "loss": 0.0167, "step": 15238 }, { "epoch": 1.652103209019948, "grad_norm": 0.08173845708370209, "learning_rate": 4.509284781662557e-05, "loss": 0.0049, "step": 15239 }, { "epoch": 1.6522116218560279, "grad_norm": 1.3714814186096191, "learning_rate": 4.508922094878863e-05, "loss": 0.0601, "step": 15240 }, { "epoch": 1.6523200346921074, "grad_norm": 0.42679110169410706, "learning_rate": 4.5085594080951695e-05, "loss": 0.0301, "step": 15241 }, { "epoch": 1.6524284475281874, "grad_norm": 0.36233019828796387, "learning_rate": 4.508196721311476e-05, "loss": 0.0531, "step": 15242 }, { "epoch": 1.6525368603642672, "grad_norm": 0.6754030585289001, "learning_rate": 4.507834034527782e-05, "loss": 0.0218, "step": 15243 }, { "epoch": 1.652645273200347, "grad_norm": 0.46407443284988403, "learning_rate": 4.507471347744088e-05, "loss": 0.0555, "step": 15244 }, { "epoch": 1.6527536860364267, "grad_norm": 0.4246279001235962, "learning_rate": 4.5071086609603945e-05, "loss": 0.0181, "step": 15245 }, { "epoch": 1.6528620988725065, "grad_norm": 0.4594048261642456, "learning_rate": 4.506745974176701e-05, "loss": 0.0217, "step": 15246 }, { "epoch": 1.6529705117085864, "grad_norm": 0.32269608974456787, "learning_rate": 4.506383287393008e-05, "loss": 0.0182, "step": 15247 }, { "epoch": 1.653078924544666, "grad_norm": 0.2414710819721222, "learning_rate": 4.5060206006093144e-05, "loss": 0.0135, "step": 15248 }, { "epoch": 1.653187337380746, "grad_norm": 0.268826425075531, "learning_rate": 4.505657913825621e-05, "loss": 0.0123, "step": 15249 }, { "epoch": 1.6532957502168255, "grad_norm": 0.183960422873497, "learning_rate": 4.5052952270419266e-05, "loss": 0.0138, "step": 15250 }, { "epoch": 1.6534041630529055, "grad_norm": 0.32511377334594727, "learning_rate": 4.504932540258233e-05, "loss": 0.0403, "step": 15251 }, { "epoch": 1.6535125758889853, "grad_norm": 0.2606298625469208, "learning_rate": 4.5045698534745394e-05, "loss": 0.0085, "step": 15252 }, { "epoch": 1.653620988725065, "grad_norm": 0.21286891400814056, "learning_rate": 4.504207166690846e-05, "loss": 0.0232, "step": 15253 }, { "epoch": 1.6537294015611448, "grad_norm": 0.0443587452173233, "learning_rate": 4.503844479907152e-05, "loss": 0.003, "step": 15254 }, { "epoch": 1.6538378143972245, "grad_norm": 0.10732970386743546, "learning_rate": 4.5034817931234586e-05, "loss": 0.0111, "step": 15255 }, { "epoch": 1.6539462272333045, "grad_norm": 0.21643579006195068, "learning_rate": 4.503119106339766e-05, "loss": 0.0147, "step": 15256 }, { "epoch": 1.654054640069384, "grad_norm": 0.625510573387146, "learning_rate": 4.5027564195560714e-05, "loss": 0.0321, "step": 15257 }, { "epoch": 1.654163052905464, "grad_norm": 0.06063202768564224, "learning_rate": 4.502393732772378e-05, "loss": 0.0033, "step": 15258 }, { "epoch": 1.6542714657415438, "grad_norm": 0.5672829747200012, "learning_rate": 4.502031045988684e-05, "loss": 0.0479, "step": 15259 }, { "epoch": 1.6543798785776236, "grad_norm": 0.42389994859695435, "learning_rate": 4.5016683592049907e-05, "loss": 0.0204, "step": 15260 }, { "epoch": 1.6544882914137033, "grad_norm": 0.18316036462783813, "learning_rate": 4.501305672421297e-05, "loss": 0.0106, "step": 15261 }, { "epoch": 1.6545967042497831, "grad_norm": 0.10517717897891998, "learning_rate": 4.5009429856376035e-05, "loss": 0.0122, "step": 15262 }, { "epoch": 1.654705117085863, "grad_norm": 0.16868473589420319, "learning_rate": 4.50058029885391e-05, "loss": 0.0091, "step": 15263 }, { "epoch": 1.6548135299219426, "grad_norm": 0.23197823762893677, "learning_rate": 4.500217612070216e-05, "loss": 0.0245, "step": 15264 }, { "epoch": 1.6549219427580226, "grad_norm": 0.10754959285259247, "learning_rate": 4.499854925286523e-05, "loss": 0.0115, "step": 15265 }, { "epoch": 1.6550303555941024, "grad_norm": 0.6240234375, "learning_rate": 4.499492238502829e-05, "loss": 0.072, "step": 15266 }, { "epoch": 1.6551387684301822, "grad_norm": 0.10834690183401108, "learning_rate": 4.4991295517191355e-05, "loss": 0.0099, "step": 15267 }, { "epoch": 1.655247181266262, "grad_norm": 0.20059649646282196, "learning_rate": 4.498766864935442e-05, "loss": 0.0175, "step": 15268 }, { "epoch": 1.6553555941023417, "grad_norm": 0.1558246612548828, "learning_rate": 4.4984041781517484e-05, "loss": 0.0094, "step": 15269 }, { "epoch": 1.6554640069384217, "grad_norm": 0.188734769821167, "learning_rate": 4.498041491368055e-05, "loss": 0.0064, "step": 15270 }, { "epoch": 1.6555724197745012, "grad_norm": 0.3709258735179901, "learning_rate": 4.497678804584361e-05, "loss": 0.027, "step": 15271 }, { "epoch": 1.6556808326105812, "grad_norm": 0.47690847516059875, "learning_rate": 4.4973161178006676e-05, "loss": 0.0217, "step": 15272 }, { "epoch": 1.6557892454466607, "grad_norm": 0.3085201680660248, "learning_rate": 4.496953431016974e-05, "loss": 0.0434, "step": 15273 }, { "epoch": 1.6558976582827407, "grad_norm": 2.4053640365600586, "learning_rate": 4.4965907442332804e-05, "loss": 0.025, "step": 15274 }, { "epoch": 1.6560060711188205, "grad_norm": 0.07855120301246643, "learning_rate": 4.496228057449587e-05, "loss": 0.0064, "step": 15275 }, { "epoch": 1.6561144839549002, "grad_norm": 0.49350619316101074, "learning_rate": 4.495865370665893e-05, "loss": 0.0405, "step": 15276 }, { "epoch": 1.65622289679098, "grad_norm": 0.30930453538894653, "learning_rate": 4.4955026838821996e-05, "loss": 0.0145, "step": 15277 }, { "epoch": 1.6563313096270598, "grad_norm": 0.5005086660385132, "learning_rate": 4.495139997098506e-05, "loss": 0.0176, "step": 15278 }, { "epoch": 1.6564397224631398, "grad_norm": 0.021542847156524658, "learning_rate": 4.4947773103148125e-05, "loss": 0.0017, "step": 15279 }, { "epoch": 1.6565481352992193, "grad_norm": 0.05254879593849182, "learning_rate": 4.494414623531118e-05, "loss": 0.0039, "step": 15280 }, { "epoch": 1.6566565481352993, "grad_norm": 0.20190443098545074, "learning_rate": 4.4940519367474246e-05, "loss": 0.023, "step": 15281 }, { "epoch": 1.656764960971379, "grad_norm": 0.32046186923980713, "learning_rate": 4.493689249963732e-05, "loss": 0.0346, "step": 15282 }, { "epoch": 1.6568733738074588, "grad_norm": 0.28367307782173157, "learning_rate": 4.493326563180038e-05, "loss": 0.0233, "step": 15283 }, { "epoch": 1.6569817866435386, "grad_norm": 0.5812732577323914, "learning_rate": 4.4929638763963445e-05, "loss": 0.0532, "step": 15284 }, { "epoch": 1.6570901994796183, "grad_norm": 0.20541808009147644, "learning_rate": 4.492601189612651e-05, "loss": 0.0297, "step": 15285 }, { "epoch": 1.6571986123156983, "grad_norm": 0.581880509853363, "learning_rate": 4.492238502828957e-05, "loss": 0.0324, "step": 15286 }, { "epoch": 1.6573070251517779, "grad_norm": 0.08324727416038513, "learning_rate": 4.491875816045264e-05, "loss": 0.0047, "step": 15287 }, { "epoch": 1.6574154379878578, "grad_norm": 0.09408892691135406, "learning_rate": 4.4915131292615695e-05, "loss": 0.0073, "step": 15288 }, { "epoch": 1.6575238508239376, "grad_norm": 0.13936716318130493, "learning_rate": 4.491150442477876e-05, "loss": 0.0121, "step": 15289 }, { "epoch": 1.6576322636600174, "grad_norm": 0.13892199099063873, "learning_rate": 4.490787755694183e-05, "loss": 0.0091, "step": 15290 }, { "epoch": 1.6577406764960971, "grad_norm": 0.4472629427909851, "learning_rate": 4.4904250689104894e-05, "loss": 0.0179, "step": 15291 }, { "epoch": 1.657849089332177, "grad_norm": 0.1467251032590866, "learning_rate": 4.490062382126796e-05, "loss": 0.0109, "step": 15292 }, { "epoch": 1.6579575021682569, "grad_norm": 0.39333704113960266, "learning_rate": 4.489699695343102e-05, "loss": 0.0315, "step": 15293 }, { "epoch": 1.6580659150043364, "grad_norm": 0.7140470743179321, "learning_rate": 4.4893370085594086e-05, "loss": 0.0165, "step": 15294 }, { "epoch": 1.6581743278404164, "grad_norm": 0.21643653512001038, "learning_rate": 4.4889743217757143e-05, "loss": 0.0089, "step": 15295 }, { "epoch": 1.658282740676496, "grad_norm": 0.462319552898407, "learning_rate": 4.488611634992021e-05, "loss": 0.021, "step": 15296 }, { "epoch": 1.658391153512576, "grad_norm": 0.22423230111598969, "learning_rate": 4.488248948208327e-05, "loss": 0.0142, "step": 15297 }, { "epoch": 1.6584995663486557, "grad_norm": 0.07180973887443542, "learning_rate": 4.4878862614246336e-05, "loss": 0.0047, "step": 15298 }, { "epoch": 1.6586079791847355, "grad_norm": 0.09439234435558319, "learning_rate": 4.487523574640941e-05, "loss": 0.0145, "step": 15299 }, { "epoch": 1.6587163920208152, "grad_norm": 0.41804423928260803, "learning_rate": 4.487160887857247e-05, "loss": 0.0325, "step": 15300 }, { "epoch": 1.658824804856895, "grad_norm": 0.15397858619689941, "learning_rate": 4.4867982010735535e-05, "loss": 0.0125, "step": 15301 }, { "epoch": 1.658933217692975, "grad_norm": 0.5400158166885376, "learning_rate": 4.486435514289859e-05, "loss": 0.0395, "step": 15302 }, { "epoch": 1.6590416305290545, "grad_norm": 0.0901920348405838, "learning_rate": 4.4860728275061656e-05, "loss": 0.0048, "step": 15303 }, { "epoch": 1.6591500433651345, "grad_norm": 0.03131592646241188, "learning_rate": 4.485710140722472e-05, "loss": 0.0013, "step": 15304 }, { "epoch": 1.6592584562012143, "grad_norm": 0.18275746703147888, "learning_rate": 4.4853474539387785e-05, "loss": 0.0186, "step": 15305 }, { "epoch": 1.659366869037294, "grad_norm": 0.43724948167800903, "learning_rate": 4.484984767155085e-05, "loss": 0.0174, "step": 15306 }, { "epoch": 1.6594752818733738, "grad_norm": 0.1739540547132492, "learning_rate": 4.484622080371392e-05, "loss": 0.0113, "step": 15307 }, { "epoch": 1.6595836947094535, "grad_norm": 0.34957972168922424, "learning_rate": 4.4842593935876984e-05, "loss": 0.0085, "step": 15308 }, { "epoch": 1.6596921075455335, "grad_norm": 0.3789554536342621, "learning_rate": 4.483896706804004e-05, "loss": 0.0389, "step": 15309 }, { "epoch": 1.659800520381613, "grad_norm": 0.22611872851848602, "learning_rate": 4.4835340200203105e-05, "loss": 0.016, "step": 15310 }, { "epoch": 1.659908933217693, "grad_norm": 0.5840505957603455, "learning_rate": 4.483171333236617e-05, "loss": 0.0569, "step": 15311 }, { "epoch": 1.6600173460537726, "grad_norm": 0.2110624462366104, "learning_rate": 4.482808646452923e-05, "loss": 0.0103, "step": 15312 }, { "epoch": 1.6601257588898526, "grad_norm": 0.21429936587810516, "learning_rate": 4.48244595966923e-05, "loss": 0.0128, "step": 15313 }, { "epoch": 1.6602341717259324, "grad_norm": 0.3342165946960449, "learning_rate": 4.482083272885536e-05, "loss": 0.0088, "step": 15314 }, { "epoch": 1.6603425845620121, "grad_norm": 0.418606162071228, "learning_rate": 4.4817205861018426e-05, "loss": 0.0284, "step": 15315 }, { "epoch": 1.660450997398092, "grad_norm": 0.5041515827178955, "learning_rate": 4.4813578993181496e-05, "loss": 0.0254, "step": 15316 }, { "epoch": 1.6605594102341716, "grad_norm": 0.10067055374383926, "learning_rate": 4.4809952125344554e-05, "loss": 0.0106, "step": 15317 }, { "epoch": 1.6606678230702516, "grad_norm": 0.06194698065519333, "learning_rate": 4.480632525750762e-05, "loss": 0.0057, "step": 15318 }, { "epoch": 1.6607762359063312, "grad_norm": 0.2596585154533386, "learning_rate": 4.480269838967068e-05, "loss": 0.028, "step": 15319 }, { "epoch": 1.6608846487424112, "grad_norm": 0.01566118374466896, "learning_rate": 4.4799071521833746e-05, "loss": 0.0013, "step": 15320 }, { "epoch": 1.660993061578491, "grad_norm": 0.4178166389465332, "learning_rate": 4.479544465399681e-05, "loss": 0.0442, "step": 15321 }, { "epoch": 1.6611014744145707, "grad_norm": 0.23075135052204132, "learning_rate": 4.4791817786159874e-05, "loss": 0.0311, "step": 15322 }, { "epoch": 1.6612098872506504, "grad_norm": 0.2611112594604492, "learning_rate": 4.478819091832294e-05, "loss": 0.0257, "step": 15323 }, { "epoch": 1.6613183000867302, "grad_norm": 0.22411687672138214, "learning_rate": 4.4784564050486e-05, "loss": 0.0208, "step": 15324 }, { "epoch": 1.6614267129228102, "grad_norm": 0.2602485716342926, "learning_rate": 4.478093718264907e-05, "loss": 0.0101, "step": 15325 }, { "epoch": 1.6615351257588897, "grad_norm": 0.16840048134326935, "learning_rate": 4.477731031481213e-05, "loss": 0.0124, "step": 15326 }, { "epoch": 1.6616435385949697, "grad_norm": 0.21987003087997437, "learning_rate": 4.4773683446975195e-05, "loss": 0.0115, "step": 15327 }, { "epoch": 1.6617519514310495, "grad_norm": 0.5857434272766113, "learning_rate": 4.477005657913826e-05, "loss": 0.0494, "step": 15328 }, { "epoch": 1.6618603642671292, "grad_norm": 0.3254562318325043, "learning_rate": 4.476642971130132e-05, "loss": 0.0398, "step": 15329 }, { "epoch": 1.661968777103209, "grad_norm": 0.0629364550113678, "learning_rate": 4.476280284346439e-05, "loss": 0.0046, "step": 15330 }, { "epoch": 1.6620771899392888, "grad_norm": 0.4128332734107971, "learning_rate": 4.475917597562745e-05, "loss": 0.015, "step": 15331 }, { "epoch": 1.6621856027753688, "grad_norm": 0.6522634625434875, "learning_rate": 4.4755549107790515e-05, "loss": 0.0325, "step": 15332 }, { "epoch": 1.6622940156114483, "grad_norm": 0.6054941415786743, "learning_rate": 4.475192223995358e-05, "loss": 0.0272, "step": 15333 }, { "epoch": 1.6624024284475283, "grad_norm": 0.31402432918548584, "learning_rate": 4.4748295372116644e-05, "loss": 0.0307, "step": 15334 }, { "epoch": 1.6625108412836078, "grad_norm": 0.5075230598449707, "learning_rate": 4.474466850427971e-05, "loss": 0.0354, "step": 15335 }, { "epoch": 1.6626192541196878, "grad_norm": 0.36729100346565247, "learning_rate": 4.474104163644277e-05, "loss": 0.0539, "step": 15336 }, { "epoch": 1.6627276669557676, "grad_norm": 0.2452613115310669, "learning_rate": 4.4737414768605836e-05, "loss": 0.0227, "step": 15337 }, { "epoch": 1.6628360797918473, "grad_norm": 0.5178667902946472, "learning_rate": 4.47337879007689e-05, "loss": 0.0411, "step": 15338 }, { "epoch": 1.662944492627927, "grad_norm": 0.9041081666946411, "learning_rate": 4.4730161032931964e-05, "loss": 0.0309, "step": 15339 }, { "epoch": 1.6630529054640069, "grad_norm": 0.07450022548437119, "learning_rate": 4.472653416509502e-05, "loss": 0.0059, "step": 15340 }, { "epoch": 1.6631613183000868, "grad_norm": 0.6461270451545715, "learning_rate": 4.4722907297258086e-05, "loss": 0.0548, "step": 15341 }, { "epoch": 1.6632697311361664, "grad_norm": 0.35459014773368835, "learning_rate": 4.4719280429421156e-05, "loss": 0.0271, "step": 15342 }, { "epoch": 1.6633781439722464, "grad_norm": 0.26076844334602356, "learning_rate": 4.471565356158422e-05, "loss": 0.0175, "step": 15343 }, { "epoch": 1.6634865568083261, "grad_norm": 0.1168871521949768, "learning_rate": 4.4712026693747285e-05, "loss": 0.0044, "step": 15344 }, { "epoch": 1.663594969644406, "grad_norm": 0.12208829075098038, "learning_rate": 4.470839982591035e-05, "loss": 0.0148, "step": 15345 }, { "epoch": 1.6637033824804857, "grad_norm": 0.08974625915288925, "learning_rate": 4.470477295807341e-05, "loss": 0.008, "step": 15346 }, { "epoch": 1.6638117953165654, "grad_norm": 0.31581807136535645, "learning_rate": 4.470114609023647e-05, "loss": 0.0116, "step": 15347 }, { "epoch": 1.6639202081526454, "grad_norm": 0.0793161392211914, "learning_rate": 4.4697519222399534e-05, "loss": 0.0064, "step": 15348 }, { "epoch": 1.664028620988725, "grad_norm": 0.35977593064308167, "learning_rate": 4.46938923545626e-05, "loss": 0.0205, "step": 15349 }, { "epoch": 1.664137033824805, "grad_norm": 0.3629807233810425, "learning_rate": 4.469026548672566e-05, "loss": 0.0353, "step": 15350 }, { "epoch": 1.6642454466608847, "grad_norm": 0.13142994046211243, "learning_rate": 4.468663861888873e-05, "loss": 0.0072, "step": 15351 }, { "epoch": 1.6643538594969645, "grad_norm": 0.11424772441387177, "learning_rate": 4.46830117510518e-05, "loss": 0.0091, "step": 15352 }, { "epoch": 1.6644622723330442, "grad_norm": 0.7149198651313782, "learning_rate": 4.467938488321486e-05, "loss": 0.0478, "step": 15353 }, { "epoch": 1.664570685169124, "grad_norm": 0.22936224937438965, "learning_rate": 4.467575801537792e-05, "loss": 0.0274, "step": 15354 }, { "epoch": 1.664679098005204, "grad_norm": 0.41219332814216614, "learning_rate": 4.467213114754098e-05, "loss": 0.0176, "step": 15355 }, { "epoch": 1.6647875108412835, "grad_norm": 0.44132545590400696, "learning_rate": 4.466850427970405e-05, "loss": 0.0631, "step": 15356 }, { "epoch": 1.6648959236773635, "grad_norm": 0.7591166496276855, "learning_rate": 4.466487741186711e-05, "loss": 0.0379, "step": 15357 }, { "epoch": 1.665004336513443, "grad_norm": 0.07886320352554321, "learning_rate": 4.4661250544030175e-05, "loss": 0.0064, "step": 15358 }, { "epoch": 1.665112749349523, "grad_norm": 0.09023173898458481, "learning_rate": 4.4657623676193246e-05, "loss": 0.0067, "step": 15359 }, { "epoch": 1.6652211621856028, "grad_norm": 0.4050895869731903, "learning_rate": 4.465399680835631e-05, "loss": 0.0543, "step": 15360 }, { "epoch": 1.6653295750216826, "grad_norm": 0.519588828086853, "learning_rate": 4.465036994051937e-05, "loss": 0.0389, "step": 15361 }, { "epoch": 1.6654379878577623, "grad_norm": 0.1422988623380661, "learning_rate": 4.464674307268243e-05, "loss": 0.008, "step": 15362 }, { "epoch": 1.665546400693842, "grad_norm": 0.647188127040863, "learning_rate": 4.4643116204845496e-05, "loss": 0.0285, "step": 15363 }, { "epoch": 1.665654813529922, "grad_norm": 0.41443687677383423, "learning_rate": 4.463948933700856e-05, "loss": 0.0378, "step": 15364 }, { "epoch": 1.6657632263660016, "grad_norm": 0.4380427598953247, "learning_rate": 4.4635862469171624e-05, "loss": 0.0493, "step": 15365 }, { "epoch": 1.6658716392020816, "grad_norm": 0.11926648765802383, "learning_rate": 4.463223560133469e-05, "loss": 0.0053, "step": 15366 }, { "epoch": 1.6659800520381614, "grad_norm": 0.2661024034023285, "learning_rate": 4.462860873349775e-05, "loss": 0.0353, "step": 15367 }, { "epoch": 1.6660884648742411, "grad_norm": 0.1813504546880722, "learning_rate": 4.462498186566082e-05, "loss": 0.0239, "step": 15368 }, { "epoch": 1.6661968777103209, "grad_norm": 0.30639714002609253, "learning_rate": 4.462135499782388e-05, "loss": 0.0229, "step": 15369 }, { "epoch": 1.6663052905464006, "grad_norm": 0.18141590058803558, "learning_rate": 4.4617728129986945e-05, "loss": 0.0137, "step": 15370 }, { "epoch": 1.6664137033824806, "grad_norm": 0.26628077030181885, "learning_rate": 4.461410126215001e-05, "loss": 0.0152, "step": 15371 }, { "epoch": 1.6665221162185602, "grad_norm": 0.5098579525947571, "learning_rate": 4.461047439431307e-05, "loss": 0.0478, "step": 15372 }, { "epoch": 1.6666305290546402, "grad_norm": 0.1700287163257599, "learning_rate": 4.460684752647614e-05, "loss": 0.0157, "step": 15373 }, { "epoch": 1.66673894189072, "grad_norm": 0.5070639848709106, "learning_rate": 4.46032206586392e-05, "loss": 0.0348, "step": 15374 }, { "epoch": 1.6668473547267997, "grad_norm": 0.3309398889541626, "learning_rate": 4.4599593790802265e-05, "loss": 0.0276, "step": 15375 }, { "epoch": 1.6669557675628794, "grad_norm": 0.130732923746109, "learning_rate": 4.459596692296533e-05, "loss": 0.0157, "step": 15376 }, { "epoch": 1.6670641803989592, "grad_norm": 1.2299985885620117, "learning_rate": 4.459234005512839e-05, "loss": 0.0441, "step": 15377 }, { "epoch": 1.6671725932350392, "grad_norm": 0.178938627243042, "learning_rate": 4.458871318729146e-05, "loss": 0.0146, "step": 15378 }, { "epoch": 1.6672810060711187, "grad_norm": 0.5800272822380066, "learning_rate": 4.458508631945452e-05, "loss": 0.0185, "step": 15379 }, { "epoch": 1.6673894189071987, "grad_norm": 0.3713848292827606, "learning_rate": 4.4581459451617586e-05, "loss": 0.0171, "step": 15380 }, { "epoch": 1.6674978317432783, "grad_norm": 0.2905523180961609, "learning_rate": 4.457783258378065e-05, "loss": 0.0276, "step": 15381 }, { "epoch": 1.6676062445793582, "grad_norm": 0.14683780074119568, "learning_rate": 4.4574205715943714e-05, "loss": 0.0079, "step": 15382 }, { "epoch": 1.667714657415438, "grad_norm": 0.19098173081874847, "learning_rate": 4.457057884810678e-05, "loss": 0.0227, "step": 15383 }, { "epoch": 1.6678230702515178, "grad_norm": 0.40073680877685547, "learning_rate": 4.456695198026984e-05, "loss": 0.0207, "step": 15384 }, { "epoch": 1.6679314830875975, "grad_norm": 0.3202241361141205, "learning_rate": 4.4563325112432906e-05, "loss": 0.0115, "step": 15385 }, { "epoch": 1.6680398959236773, "grad_norm": 0.10734356194734573, "learning_rate": 4.455969824459597e-05, "loss": 0.0069, "step": 15386 }, { "epoch": 1.6681483087597573, "grad_norm": 0.20878049731254578, "learning_rate": 4.4556071376759034e-05, "loss": 0.0201, "step": 15387 }, { "epoch": 1.6682567215958368, "grad_norm": 0.6143689751625061, "learning_rate": 4.45524445089221e-05, "loss": 0.0364, "step": 15388 }, { "epoch": 1.6683651344319168, "grad_norm": 0.4044466018676758, "learning_rate": 4.454881764108516e-05, "loss": 0.0688, "step": 15389 }, { "epoch": 1.6684735472679966, "grad_norm": 0.41129088401794434, "learning_rate": 4.454519077324823e-05, "loss": 0.0451, "step": 15390 }, { "epoch": 1.6685819601040763, "grad_norm": 0.679598331451416, "learning_rate": 4.454156390541129e-05, "loss": 0.0809, "step": 15391 }, { "epoch": 1.668690372940156, "grad_norm": 0.22779369354248047, "learning_rate": 4.453793703757435e-05, "loss": 0.017, "step": 15392 }, { "epoch": 1.6687987857762359, "grad_norm": 0.20638537406921387, "learning_rate": 4.453431016973741e-05, "loss": 0.0123, "step": 15393 }, { "epoch": 1.6689071986123158, "grad_norm": 0.2427484542131424, "learning_rate": 4.453068330190048e-05, "loss": 0.0429, "step": 15394 }, { "epoch": 1.6690156114483954, "grad_norm": 0.22055819630622864, "learning_rate": 4.452705643406355e-05, "loss": 0.0184, "step": 15395 }, { "epoch": 1.6691240242844754, "grad_norm": 0.8234060406684875, "learning_rate": 4.452342956622661e-05, "loss": 0.0619, "step": 15396 }, { "epoch": 1.669232437120555, "grad_norm": 0.32795074582099915, "learning_rate": 4.4519802698389675e-05, "loss": 0.0517, "step": 15397 }, { "epoch": 1.669340849956635, "grad_norm": 0.18862628936767578, "learning_rate": 4.451617583055274e-05, "loss": 0.019, "step": 15398 }, { "epoch": 1.6694492627927147, "grad_norm": 0.32982951402664185, "learning_rate": 4.45125489627158e-05, "loss": 0.0227, "step": 15399 }, { "epoch": 1.6695576756287944, "grad_norm": 0.2510857582092285, "learning_rate": 4.450892209487886e-05, "loss": 0.0257, "step": 15400 }, { "epoch": 1.6696660884648742, "grad_norm": 0.572208821773529, "learning_rate": 4.4505295227041925e-05, "loss": 0.0246, "step": 15401 }, { "epoch": 1.669774501300954, "grad_norm": 0.18463407456874847, "learning_rate": 4.4501668359204996e-05, "loss": 0.0116, "step": 15402 }, { "epoch": 1.669882914137034, "grad_norm": 0.23128554224967957, "learning_rate": 4.449804149136806e-05, "loss": 0.0204, "step": 15403 }, { "epoch": 1.6699913269731135, "grad_norm": 0.4035205841064453, "learning_rate": 4.4494414623531124e-05, "loss": 0.0453, "step": 15404 }, { "epoch": 1.6700997398091935, "grad_norm": 0.2863459885120392, "learning_rate": 4.449078775569419e-05, "loss": 0.025, "step": 15405 }, { "epoch": 1.6702081526452732, "grad_norm": 0.2172311693429947, "learning_rate": 4.4487160887857246e-05, "loss": 0.0158, "step": 15406 }, { "epoch": 1.670316565481353, "grad_norm": 0.17948488891124725, "learning_rate": 4.448353402002031e-05, "loss": 0.0101, "step": 15407 }, { "epoch": 1.6704249783174328, "grad_norm": 0.4399818480014801, "learning_rate": 4.4479907152183374e-05, "loss": 0.0631, "step": 15408 }, { "epoch": 1.6705333911535125, "grad_norm": 0.4758095145225525, "learning_rate": 4.447628028434644e-05, "loss": 0.0421, "step": 15409 }, { "epoch": 1.6706418039895925, "grad_norm": 0.17090445756912231, "learning_rate": 4.44726534165095e-05, "loss": 0.0082, "step": 15410 }, { "epoch": 1.670750216825672, "grad_norm": 0.16373106837272644, "learning_rate": 4.446902654867257e-05, "loss": 0.0226, "step": 15411 }, { "epoch": 1.670858629661752, "grad_norm": 0.10098925232887268, "learning_rate": 4.446539968083564e-05, "loss": 0.0089, "step": 15412 }, { "epoch": 1.6709670424978318, "grad_norm": 0.09881290048360825, "learning_rate": 4.4461772812998694e-05, "loss": 0.0085, "step": 15413 }, { "epoch": 1.6710754553339116, "grad_norm": 0.41797253489494324, "learning_rate": 4.445814594516176e-05, "loss": 0.0273, "step": 15414 }, { "epoch": 1.6711838681699913, "grad_norm": 0.4218585193157196, "learning_rate": 4.445451907732482e-05, "loss": 0.0216, "step": 15415 }, { "epoch": 1.671292281006071, "grad_norm": 0.3921383023262024, "learning_rate": 4.445089220948789e-05, "loss": 0.0439, "step": 15416 }, { "epoch": 1.671400693842151, "grad_norm": 0.8545533418655396, "learning_rate": 4.444726534165095e-05, "loss": 0.0485, "step": 15417 }, { "epoch": 1.6715091066782306, "grad_norm": 0.32144615054130554, "learning_rate": 4.4443638473814015e-05, "loss": 0.0234, "step": 15418 }, { "epoch": 1.6716175195143106, "grad_norm": 0.2197999656200409, "learning_rate": 4.444001160597708e-05, "loss": 0.035, "step": 15419 }, { "epoch": 1.6717259323503901, "grad_norm": 0.12175848335027695, "learning_rate": 4.443638473814015e-05, "loss": 0.0104, "step": 15420 }, { "epoch": 1.6718343451864701, "grad_norm": 0.4496921896934509, "learning_rate": 4.443275787030321e-05, "loss": 0.0408, "step": 15421 }, { "epoch": 1.6719427580225499, "grad_norm": 0.21709927916526794, "learning_rate": 4.442913100246627e-05, "loss": 0.0235, "step": 15422 }, { "epoch": 1.6720511708586296, "grad_norm": 0.07394687831401825, "learning_rate": 4.4425504134629335e-05, "loss": 0.0065, "step": 15423 }, { "epoch": 1.6721595836947094, "grad_norm": 0.1879088282585144, "learning_rate": 4.44218772667924e-05, "loss": 0.0171, "step": 15424 }, { "epoch": 1.6722679965307892, "grad_norm": 0.5368144512176514, "learning_rate": 4.4418250398955464e-05, "loss": 0.0682, "step": 15425 }, { "epoch": 1.6723764093668692, "grad_norm": 0.11185956001281738, "learning_rate": 4.441462353111853e-05, "loss": 0.0177, "step": 15426 }, { "epoch": 1.6724848222029487, "grad_norm": 0.08861545473337173, "learning_rate": 4.441099666328159e-05, "loss": 0.0091, "step": 15427 }, { "epoch": 1.6725932350390287, "grad_norm": 0.03984207287430763, "learning_rate": 4.4407369795444656e-05, "loss": 0.0031, "step": 15428 }, { "epoch": 1.6727016478751084, "grad_norm": 0.2987332344055176, "learning_rate": 4.440374292760772e-05, "loss": 0.024, "step": 15429 }, { "epoch": 1.6728100607111882, "grad_norm": 0.18828855454921722, "learning_rate": 4.4400116059770784e-05, "loss": 0.0334, "step": 15430 }, { "epoch": 1.672918473547268, "grad_norm": 0.1856607049703598, "learning_rate": 4.439648919193385e-05, "loss": 0.0225, "step": 15431 }, { "epoch": 1.6730268863833477, "grad_norm": 0.22898723185062408, "learning_rate": 4.439286232409691e-05, "loss": 0.0133, "step": 15432 }, { "epoch": 1.6731352992194277, "grad_norm": 0.41815078258514404, "learning_rate": 4.4389235456259976e-05, "loss": 0.0622, "step": 15433 }, { "epoch": 1.6732437120555073, "grad_norm": 0.25793731212615967, "learning_rate": 4.438560858842304e-05, "loss": 0.0279, "step": 15434 }, { "epoch": 1.6733521248915872, "grad_norm": 0.20566600561141968, "learning_rate": 4.4381981720586105e-05, "loss": 0.0226, "step": 15435 }, { "epoch": 1.673460537727667, "grad_norm": 0.28430911898612976, "learning_rate": 4.437835485274917e-05, "loss": 0.0228, "step": 15436 }, { "epoch": 1.6735689505637468, "grad_norm": 0.347309947013855, "learning_rate": 4.437472798491223e-05, "loss": 0.0227, "step": 15437 }, { "epoch": 1.6736773633998265, "grad_norm": 0.2534763813018799, "learning_rate": 4.43711011170753e-05, "loss": 0.0112, "step": 15438 }, { "epoch": 1.6737857762359063, "grad_norm": 0.31501439213752747, "learning_rate": 4.436747424923836e-05, "loss": 0.0232, "step": 15439 }, { "epoch": 1.6738941890719863, "grad_norm": 0.6260005235671997, "learning_rate": 4.4363847381401425e-05, "loss": 0.0291, "step": 15440 }, { "epoch": 1.6740026019080658, "grad_norm": 0.24191245436668396, "learning_rate": 4.436022051356449e-05, "loss": 0.0145, "step": 15441 }, { "epoch": 1.6741110147441458, "grad_norm": 0.46599096059799194, "learning_rate": 4.4356593645727553e-05, "loss": 0.0361, "step": 15442 }, { "epoch": 1.6742194275802254, "grad_norm": 0.21068622171878815, "learning_rate": 4.435296677789062e-05, "loss": 0.0156, "step": 15443 }, { "epoch": 1.6743278404163053, "grad_norm": 0.8907602429389954, "learning_rate": 4.4349339910053675e-05, "loss": 0.0585, "step": 15444 }, { "epoch": 1.674436253252385, "grad_norm": 0.015517829917371273, "learning_rate": 4.434571304221674e-05, "loss": 0.0015, "step": 15445 }, { "epoch": 1.6745446660884649, "grad_norm": 0.20810353755950928, "learning_rate": 4.434208617437981e-05, "loss": 0.0175, "step": 15446 }, { "epoch": 1.6746530789245446, "grad_norm": 0.1133483499288559, "learning_rate": 4.4338459306542874e-05, "loss": 0.005, "step": 15447 }, { "epoch": 1.6747614917606244, "grad_norm": 0.22588816285133362, "learning_rate": 4.433483243870594e-05, "loss": 0.0416, "step": 15448 }, { "epoch": 1.6748699045967044, "grad_norm": 0.13758723437786102, "learning_rate": 4.4331205570869e-05, "loss": 0.0165, "step": 15449 }, { "epoch": 1.674978317432784, "grad_norm": 0.2664799690246582, "learning_rate": 4.4327578703032066e-05, "loss": 0.04, "step": 15450 }, { "epoch": 1.675086730268864, "grad_norm": 0.13633817434310913, "learning_rate": 4.4323951835195124e-05, "loss": 0.0105, "step": 15451 }, { "epoch": 1.6751951431049437, "grad_norm": 0.44207799434661865, "learning_rate": 4.432032496735819e-05, "loss": 0.0896, "step": 15452 }, { "epoch": 1.6753035559410234, "grad_norm": 0.38616272807121277, "learning_rate": 4.431669809952125e-05, "loss": 0.0459, "step": 15453 }, { "epoch": 1.6754119687771032, "grad_norm": 0.3754931688308716, "learning_rate": 4.431307123168432e-05, "loss": 0.0463, "step": 15454 }, { "epoch": 1.675520381613183, "grad_norm": 0.21394529938697815, "learning_rate": 4.430944436384739e-05, "loss": 0.0096, "step": 15455 }, { "epoch": 1.675628794449263, "grad_norm": 0.19009099900722504, "learning_rate": 4.430581749601045e-05, "loss": 0.0155, "step": 15456 }, { "epoch": 1.6757372072853425, "grad_norm": 0.16920721530914307, "learning_rate": 4.4302190628173515e-05, "loss": 0.0051, "step": 15457 }, { "epoch": 1.6758456201214225, "grad_norm": 0.1511325240135193, "learning_rate": 4.429856376033657e-05, "loss": 0.0139, "step": 15458 }, { "epoch": 1.675954032957502, "grad_norm": 0.16352923214435577, "learning_rate": 4.4294936892499636e-05, "loss": 0.0214, "step": 15459 }, { "epoch": 1.676062445793582, "grad_norm": 0.2599720358848572, "learning_rate": 4.42913100246627e-05, "loss": 0.0274, "step": 15460 }, { "epoch": 1.6761708586296618, "grad_norm": 0.27040666341781616, "learning_rate": 4.4287683156825765e-05, "loss": 0.0268, "step": 15461 }, { "epoch": 1.6762792714657415, "grad_norm": 0.48898881673812866, "learning_rate": 4.428405628898883e-05, "loss": 0.0286, "step": 15462 }, { "epoch": 1.6763876843018215, "grad_norm": 0.2710225582122803, "learning_rate": 4.42804294211519e-05, "loss": 0.0368, "step": 15463 }, { "epoch": 1.676496097137901, "grad_norm": 0.3478095531463623, "learning_rate": 4.4276802553314964e-05, "loss": 0.0368, "step": 15464 }, { "epoch": 1.676604509973981, "grad_norm": 0.3784315586090088, "learning_rate": 4.427317568547802e-05, "loss": 0.0498, "step": 15465 }, { "epoch": 1.6767129228100606, "grad_norm": 0.14796823263168335, "learning_rate": 4.4269548817641085e-05, "loss": 0.0105, "step": 15466 }, { "epoch": 1.6768213356461406, "grad_norm": 0.28502270579338074, "learning_rate": 4.426592194980415e-05, "loss": 0.0238, "step": 15467 }, { "epoch": 1.6769297484822203, "grad_norm": 0.32650670409202576, "learning_rate": 4.426229508196721e-05, "loss": 0.0276, "step": 15468 }, { "epoch": 1.6770381613183, "grad_norm": 0.25451457500457764, "learning_rate": 4.425866821413028e-05, "loss": 0.0212, "step": 15469 }, { "epoch": 1.6771465741543798, "grad_norm": 0.43443456292152405, "learning_rate": 4.425504134629334e-05, "loss": 0.019, "step": 15470 }, { "epoch": 1.6772549869904596, "grad_norm": 0.32493820786476135, "learning_rate": 4.4251414478456406e-05, "loss": 0.0265, "step": 15471 }, { "epoch": 1.6773633998265396, "grad_norm": 0.2656272351741791, "learning_rate": 4.4247787610619477e-05, "loss": 0.0318, "step": 15472 }, { "epoch": 1.6774718126626191, "grad_norm": 0.37865084409713745, "learning_rate": 4.4244160742782534e-05, "loss": 0.0355, "step": 15473 }, { "epoch": 1.6775802254986991, "grad_norm": 0.2803798317909241, "learning_rate": 4.42405338749456e-05, "loss": 0.0174, "step": 15474 }, { "epoch": 1.6776886383347789, "grad_norm": 0.22647270560264587, "learning_rate": 4.423690700710866e-05, "loss": 0.0286, "step": 15475 }, { "epoch": 1.6777970511708586, "grad_norm": 0.2941763997077942, "learning_rate": 4.4233280139271726e-05, "loss": 0.0274, "step": 15476 }, { "epoch": 1.6779054640069384, "grad_norm": 0.36136215925216675, "learning_rate": 4.422965327143479e-05, "loss": 0.0184, "step": 15477 }, { "epoch": 1.6780138768430182, "grad_norm": 0.38245612382888794, "learning_rate": 4.4226026403597854e-05, "loss": 0.015, "step": 15478 }, { "epoch": 1.6781222896790982, "grad_norm": 0.18531109392642975, "learning_rate": 4.422239953576092e-05, "loss": 0.0084, "step": 15479 }, { "epoch": 1.6782307025151777, "grad_norm": 0.12524177134037018, "learning_rate": 4.421877266792398e-05, "loss": 0.0116, "step": 15480 }, { "epoch": 1.6783391153512577, "grad_norm": 0.4518291652202606, "learning_rate": 4.421514580008705e-05, "loss": 0.0408, "step": 15481 }, { "epoch": 1.6784475281873372, "grad_norm": 0.0965735986828804, "learning_rate": 4.421151893225011e-05, "loss": 0.0096, "step": 15482 }, { "epoch": 1.6785559410234172, "grad_norm": 0.06515245884656906, "learning_rate": 4.4207892064413175e-05, "loss": 0.0052, "step": 15483 }, { "epoch": 1.678664353859497, "grad_norm": 0.10547199100255966, "learning_rate": 4.420426519657624e-05, "loss": 0.0048, "step": 15484 }, { "epoch": 1.6787727666955767, "grad_norm": 0.6148419380187988, "learning_rate": 4.42006383287393e-05, "loss": 0.0248, "step": 15485 }, { "epoch": 1.6788811795316565, "grad_norm": 0.4033851623535156, "learning_rate": 4.419701146090237e-05, "loss": 0.0128, "step": 15486 }, { "epoch": 1.6789895923677363, "grad_norm": 0.7018885612487793, "learning_rate": 4.419338459306543e-05, "loss": 0.0239, "step": 15487 }, { "epoch": 1.6790980052038162, "grad_norm": 0.25431299209594727, "learning_rate": 4.4189757725228495e-05, "loss": 0.0289, "step": 15488 }, { "epoch": 1.6792064180398958, "grad_norm": 0.3158813416957855, "learning_rate": 4.418613085739156e-05, "loss": 0.0221, "step": 15489 }, { "epoch": 1.6793148308759758, "grad_norm": 0.3065476417541504, "learning_rate": 4.4182503989554624e-05, "loss": 0.0111, "step": 15490 }, { "epoch": 1.6794232437120555, "grad_norm": 0.33843329548835754, "learning_rate": 4.417887712171769e-05, "loss": 0.0394, "step": 15491 }, { "epoch": 1.6795316565481353, "grad_norm": 0.15095168352127075, "learning_rate": 4.417525025388075e-05, "loss": 0.008, "step": 15492 }, { "epoch": 1.679640069384215, "grad_norm": 0.10842664539813995, "learning_rate": 4.4171623386043816e-05, "loss": 0.0102, "step": 15493 }, { "epoch": 1.6797484822202948, "grad_norm": 0.44078224897384644, "learning_rate": 4.416799651820688e-05, "loss": 0.0274, "step": 15494 }, { "epoch": 1.6798568950563748, "grad_norm": 0.22426268458366394, "learning_rate": 4.4164369650369944e-05, "loss": 0.0382, "step": 15495 }, { "epoch": 1.6799653078924544, "grad_norm": 0.33699098229408264, "learning_rate": 4.4160742782533e-05, "loss": 0.0465, "step": 15496 }, { "epoch": 1.6800737207285343, "grad_norm": 0.4426887333393097, "learning_rate": 4.415711591469607e-05, "loss": 0.0297, "step": 15497 }, { "epoch": 1.680182133564614, "grad_norm": 0.06861353665590286, "learning_rate": 4.4153489046859136e-05, "loss": 0.0035, "step": 15498 }, { "epoch": 1.6802905464006939, "grad_norm": 0.29098427295684814, "learning_rate": 4.41498621790222e-05, "loss": 0.0466, "step": 15499 }, { "epoch": 1.6803989592367736, "grad_norm": 0.410669207572937, "learning_rate": 4.4146235311185265e-05, "loss": 0.0538, "step": 15500 }, { "epoch": 1.6805073720728534, "grad_norm": 0.3105151653289795, "learning_rate": 4.414260844334833e-05, "loss": 0.0151, "step": 15501 }, { "epoch": 1.6806157849089334, "grad_norm": 0.03773970901966095, "learning_rate": 4.413898157551139e-05, "loss": 0.0031, "step": 15502 }, { "epoch": 1.680724197745013, "grad_norm": 0.15917497873306274, "learning_rate": 4.413535470767445e-05, "loss": 0.0322, "step": 15503 }, { "epoch": 1.680832610581093, "grad_norm": 0.6661949753761292, "learning_rate": 4.4131727839837514e-05, "loss": 0.0342, "step": 15504 }, { "epoch": 1.6809410234171724, "grad_norm": 0.2552779018878937, "learning_rate": 4.412810097200058e-05, "loss": 0.0141, "step": 15505 }, { "epoch": 1.6810494362532524, "grad_norm": 0.32302045822143555, "learning_rate": 4.412447410416365e-05, "loss": 0.0297, "step": 15506 }, { "epoch": 1.6811578490893322, "grad_norm": 0.3816477060317993, "learning_rate": 4.4120847236326713e-05, "loss": 0.0159, "step": 15507 }, { "epoch": 1.681266261925412, "grad_norm": 0.1288911998271942, "learning_rate": 4.411722036848978e-05, "loss": 0.0118, "step": 15508 }, { "epoch": 1.6813746747614917, "grad_norm": 0.2060801237821579, "learning_rate": 4.411359350065284e-05, "loss": 0.0176, "step": 15509 }, { "epoch": 1.6814830875975715, "grad_norm": 0.29854217171669006, "learning_rate": 4.41099666328159e-05, "loss": 0.0156, "step": 15510 }, { "epoch": 1.6815915004336515, "grad_norm": 0.12052249908447266, "learning_rate": 4.410633976497896e-05, "loss": 0.0091, "step": 15511 }, { "epoch": 1.681699913269731, "grad_norm": 0.06880252063274384, "learning_rate": 4.410271289714203e-05, "loss": 0.0041, "step": 15512 }, { "epoch": 1.681808326105811, "grad_norm": 0.11071107536554337, "learning_rate": 4.409908602930509e-05, "loss": 0.0108, "step": 15513 }, { "epoch": 1.6819167389418908, "grad_norm": 0.38395988941192627, "learning_rate": 4.4095459161468155e-05, "loss": 0.0199, "step": 15514 }, { "epoch": 1.6820251517779705, "grad_norm": 0.22219517827033997, "learning_rate": 4.4091832293631226e-05, "loss": 0.0169, "step": 15515 }, { "epoch": 1.6821335646140503, "grad_norm": 0.612748384475708, "learning_rate": 4.408820542579429e-05, "loss": 0.059, "step": 15516 }, { "epoch": 1.68224197745013, "grad_norm": 0.3571707606315613, "learning_rate": 4.4084578557957354e-05, "loss": 0.0322, "step": 15517 }, { "epoch": 1.68235039028621, "grad_norm": 0.11953962594270706, "learning_rate": 4.408095169012041e-05, "loss": 0.0084, "step": 15518 }, { "epoch": 1.6824588031222896, "grad_norm": 0.26514333486557007, "learning_rate": 4.4077324822283476e-05, "loss": 0.0168, "step": 15519 }, { "epoch": 1.6825672159583696, "grad_norm": 0.20032209157943726, "learning_rate": 4.407369795444654e-05, "loss": 0.0145, "step": 15520 }, { "epoch": 1.682675628794449, "grad_norm": 0.458935409784317, "learning_rate": 4.4070071086609604e-05, "loss": 0.035, "step": 15521 }, { "epoch": 1.682784041630529, "grad_norm": 0.30264440178871155, "learning_rate": 4.406644421877267e-05, "loss": 0.0247, "step": 15522 }, { "epoch": 1.6828924544666088, "grad_norm": 0.3688981533050537, "learning_rate": 4.406281735093574e-05, "loss": 0.0225, "step": 15523 }, { "epoch": 1.6830008673026886, "grad_norm": 0.20924818515777588, "learning_rate": 4.40591904830988e-05, "loss": 0.0198, "step": 15524 }, { "epoch": 1.6831092801387686, "grad_norm": 0.13833609223365784, "learning_rate": 4.405556361526186e-05, "loss": 0.0088, "step": 15525 }, { "epoch": 1.6832176929748481, "grad_norm": 0.10532475262880325, "learning_rate": 4.4051936747424925e-05, "loss": 0.0101, "step": 15526 }, { "epoch": 1.6833261058109281, "grad_norm": 0.3267129361629486, "learning_rate": 4.404830987958799e-05, "loss": 0.0305, "step": 15527 }, { "epoch": 1.6834345186470077, "grad_norm": 0.10169622302055359, "learning_rate": 4.404468301175105e-05, "loss": 0.0115, "step": 15528 }, { "epoch": 1.6835429314830876, "grad_norm": 0.4747914969921112, "learning_rate": 4.404105614391412e-05, "loss": 0.0487, "step": 15529 }, { "epoch": 1.6836513443191674, "grad_norm": 0.3368808329105377, "learning_rate": 4.403742927607718e-05, "loss": 0.0252, "step": 15530 }, { "epoch": 1.6837597571552472, "grad_norm": 0.3228149116039276, "learning_rate": 4.4033802408240245e-05, "loss": 0.0313, "step": 15531 }, { "epoch": 1.683868169991327, "grad_norm": 0.15378086268901825, "learning_rate": 4.403017554040331e-05, "loss": 0.0154, "step": 15532 }, { "epoch": 1.6839765828274067, "grad_norm": 1.00437593460083, "learning_rate": 4.4026548672566373e-05, "loss": 0.0219, "step": 15533 }, { "epoch": 1.6840849956634867, "grad_norm": 0.39004430174827576, "learning_rate": 4.402292180472944e-05, "loss": 0.0357, "step": 15534 }, { "epoch": 1.6841934084995662, "grad_norm": 0.5353614687919617, "learning_rate": 4.40192949368925e-05, "loss": 0.0235, "step": 15535 }, { "epoch": 1.6843018213356462, "grad_norm": 0.20610593259334564, "learning_rate": 4.4015668069055566e-05, "loss": 0.0184, "step": 15536 }, { "epoch": 1.684410234171726, "grad_norm": 0.03183535113930702, "learning_rate": 4.401204120121863e-05, "loss": 0.0017, "step": 15537 }, { "epoch": 1.6845186470078057, "grad_norm": 0.19801980257034302, "learning_rate": 4.4008414333381694e-05, "loss": 0.0058, "step": 15538 }, { "epoch": 1.6846270598438855, "grad_norm": 0.4357633888721466, "learning_rate": 4.400478746554476e-05, "loss": 0.0149, "step": 15539 }, { "epoch": 1.6847354726799653, "grad_norm": 0.6201662421226501, "learning_rate": 4.400116059770782e-05, "loss": 0.0375, "step": 15540 }, { "epoch": 1.6848438855160452, "grad_norm": 0.2696853280067444, "learning_rate": 4.3997533729870886e-05, "loss": 0.0203, "step": 15541 }, { "epoch": 1.6849522983521248, "grad_norm": 0.4535067677497864, "learning_rate": 4.399390686203395e-05, "loss": 0.0481, "step": 15542 }, { "epoch": 1.6850607111882048, "grad_norm": 0.1064867228269577, "learning_rate": 4.3990279994197014e-05, "loss": 0.006, "step": 15543 }, { "epoch": 1.6851691240242843, "grad_norm": 0.27523067593574524, "learning_rate": 4.398665312636008e-05, "loss": 0.0105, "step": 15544 }, { "epoch": 1.6852775368603643, "grad_norm": 0.4256581664085388, "learning_rate": 4.398302625852314e-05, "loss": 0.0398, "step": 15545 }, { "epoch": 1.685385949696444, "grad_norm": 0.2616468071937561, "learning_rate": 4.397939939068621e-05, "loss": 0.0099, "step": 15546 }, { "epoch": 1.6854943625325238, "grad_norm": 0.277922660112381, "learning_rate": 4.397577252284927e-05, "loss": 0.0277, "step": 15547 }, { "epoch": 1.6856027753686036, "grad_norm": 0.39857298135757446, "learning_rate": 4.397214565501233e-05, "loss": 0.0169, "step": 15548 }, { "epoch": 1.6857111882046834, "grad_norm": 0.19017158448696136, "learning_rate": 4.39685187871754e-05, "loss": 0.0106, "step": 15549 }, { "epoch": 1.6858196010407633, "grad_norm": 0.7078841924667358, "learning_rate": 4.396489191933846e-05, "loss": 0.0283, "step": 15550 }, { "epoch": 1.6859280138768429, "grad_norm": 0.48665952682495117, "learning_rate": 4.396126505150153e-05, "loss": 0.01, "step": 15551 }, { "epoch": 1.6860364267129229, "grad_norm": 0.46239280700683594, "learning_rate": 4.395763818366459e-05, "loss": 0.0604, "step": 15552 }, { "epoch": 1.6861448395490026, "grad_norm": 0.41291531920433044, "learning_rate": 4.3954011315827656e-05, "loss": 0.0321, "step": 15553 }, { "epoch": 1.6862532523850824, "grad_norm": 0.29615432024002075, "learning_rate": 4.395038444799072e-05, "loss": 0.0201, "step": 15554 }, { "epoch": 1.6863616652211622, "grad_norm": 0.14444862306118011, "learning_rate": 4.394675758015378e-05, "loss": 0.0014, "step": 15555 }, { "epoch": 1.686470078057242, "grad_norm": 0.4375928044319153, "learning_rate": 4.394313071231684e-05, "loss": 0.0455, "step": 15556 }, { "epoch": 1.686578490893322, "grad_norm": 0.1546245962381363, "learning_rate": 4.3939503844479905e-05, "loss": 0.0056, "step": 15557 }, { "epoch": 1.6866869037294014, "grad_norm": 0.28829145431518555, "learning_rate": 4.3935876976642976e-05, "loss": 0.0293, "step": 15558 }, { "epoch": 1.6867953165654814, "grad_norm": 1.0996358394622803, "learning_rate": 4.393225010880604e-05, "loss": 0.0135, "step": 15559 }, { "epoch": 1.6869037294015612, "grad_norm": 0.15136656165122986, "learning_rate": 4.3928623240969104e-05, "loss": 0.0118, "step": 15560 }, { "epoch": 1.687012142237641, "grad_norm": 0.20100238919258118, "learning_rate": 4.392499637313217e-05, "loss": 0.0142, "step": 15561 }, { "epoch": 1.6871205550737207, "grad_norm": 0.8082464337348938, "learning_rate": 4.3921369505295226e-05, "loss": 0.0319, "step": 15562 }, { "epoch": 1.6872289679098005, "grad_norm": 0.42801931500434875, "learning_rate": 4.391774263745829e-05, "loss": 0.0167, "step": 15563 }, { "epoch": 1.6873373807458805, "grad_norm": 0.2629997730255127, "learning_rate": 4.3914115769621354e-05, "loss": 0.0211, "step": 15564 }, { "epoch": 1.68744579358196, "grad_norm": 0.681926965713501, "learning_rate": 4.391048890178442e-05, "loss": 0.0465, "step": 15565 }, { "epoch": 1.68755420641804, "grad_norm": 0.2910138666629791, "learning_rate": 4.390686203394748e-05, "loss": 0.008, "step": 15566 }, { "epoch": 1.6876626192541195, "grad_norm": 0.6244269013404846, "learning_rate": 4.390323516611055e-05, "loss": 0.0197, "step": 15567 }, { "epoch": 1.6877710320901995, "grad_norm": 0.4084286689758301, "learning_rate": 4.389960829827362e-05, "loss": 0.0135, "step": 15568 }, { "epoch": 1.6878794449262793, "grad_norm": 0.5657354593276978, "learning_rate": 4.389598143043668e-05, "loss": 0.0191, "step": 15569 }, { "epoch": 1.687987857762359, "grad_norm": 0.2521076798439026, "learning_rate": 4.389235456259974e-05, "loss": 0.0112, "step": 15570 }, { "epoch": 1.6880962705984388, "grad_norm": 0.087619349360466, "learning_rate": 4.38887276947628e-05, "loss": 0.0046, "step": 15571 }, { "epoch": 1.6882046834345186, "grad_norm": 0.1864563226699829, "learning_rate": 4.388510082692587e-05, "loss": 0.0099, "step": 15572 }, { "epoch": 1.6883130962705986, "grad_norm": 0.35618674755096436, "learning_rate": 4.388147395908893e-05, "loss": 0.0562, "step": 15573 }, { "epoch": 1.688421509106678, "grad_norm": 0.3125320076942444, "learning_rate": 4.3877847091251995e-05, "loss": 0.0471, "step": 15574 }, { "epoch": 1.688529921942758, "grad_norm": 0.364496648311615, "learning_rate": 4.3874220223415066e-05, "loss": 0.0189, "step": 15575 }, { "epoch": 1.6886383347788378, "grad_norm": 0.036897286772727966, "learning_rate": 4.387059335557813e-05, "loss": 0.0021, "step": 15576 }, { "epoch": 1.6887467476149176, "grad_norm": 0.19304999709129333, "learning_rate": 4.386696648774119e-05, "loss": 0.0181, "step": 15577 }, { "epoch": 1.6888551604509974, "grad_norm": 0.2168271392583847, "learning_rate": 4.386333961990425e-05, "loss": 0.0211, "step": 15578 }, { "epoch": 1.6889635732870771, "grad_norm": 0.542078971862793, "learning_rate": 4.3859712752067315e-05, "loss": 0.0485, "step": 15579 }, { "epoch": 1.6890719861231571, "grad_norm": 0.3254661560058594, "learning_rate": 4.385608588423038e-05, "loss": 0.043, "step": 15580 }, { "epoch": 1.6891803989592367, "grad_norm": 0.2699546813964844, "learning_rate": 4.3852459016393444e-05, "loss": 0.0105, "step": 15581 }, { "epoch": 1.6892888117953166, "grad_norm": 0.04894225671887398, "learning_rate": 4.384883214855651e-05, "loss": 0.0024, "step": 15582 }, { "epoch": 1.6893972246313964, "grad_norm": 0.2658379077911377, "learning_rate": 4.384520528071957e-05, "loss": 0.0212, "step": 15583 }, { "epoch": 1.6895056374674762, "grad_norm": 1.224927544593811, "learning_rate": 4.3841578412882636e-05, "loss": 0.1129, "step": 15584 }, { "epoch": 1.689614050303556, "grad_norm": 0.48415037989616394, "learning_rate": 4.38379515450457e-05, "loss": 0.065, "step": 15585 }, { "epoch": 1.6897224631396357, "grad_norm": 0.48887544870376587, "learning_rate": 4.3834324677208764e-05, "loss": 0.0329, "step": 15586 }, { "epoch": 1.6898308759757157, "grad_norm": 0.07213068753480911, "learning_rate": 4.383069780937183e-05, "loss": 0.0044, "step": 15587 }, { "epoch": 1.6899392888117952, "grad_norm": 0.2872772216796875, "learning_rate": 4.382707094153489e-05, "loss": 0.0069, "step": 15588 }, { "epoch": 1.6900477016478752, "grad_norm": 0.19462944567203522, "learning_rate": 4.3823444073697957e-05, "loss": 0.0086, "step": 15589 }, { "epoch": 1.6901561144839548, "grad_norm": 0.3030061423778534, "learning_rate": 4.381981720586102e-05, "loss": 0.0292, "step": 15590 }, { "epoch": 1.6902645273200347, "grad_norm": 0.5277481079101562, "learning_rate": 4.3816190338024085e-05, "loss": 0.0434, "step": 15591 }, { "epoch": 1.6903729401561145, "grad_norm": 0.34935086965560913, "learning_rate": 4.381256347018715e-05, "loss": 0.0148, "step": 15592 }, { "epoch": 1.6904813529921943, "grad_norm": 0.17247717082500458, "learning_rate": 4.380893660235021e-05, "loss": 0.0198, "step": 15593 }, { "epoch": 1.690589765828274, "grad_norm": 0.21574008464813232, "learning_rate": 4.380530973451328e-05, "loss": 0.0067, "step": 15594 }, { "epoch": 1.6906981786643538, "grad_norm": 0.14141784608364105, "learning_rate": 4.380168286667634e-05, "loss": 0.0079, "step": 15595 }, { "epoch": 1.6908065915004338, "grad_norm": 0.30873560905456543, "learning_rate": 4.3798055998839405e-05, "loss": 0.0287, "step": 15596 }, { "epoch": 1.6909150043365133, "grad_norm": 0.1903357058763504, "learning_rate": 4.379442913100247e-05, "loss": 0.0157, "step": 15597 }, { "epoch": 1.6910234171725933, "grad_norm": 0.07542337477207184, "learning_rate": 4.3790802263165533e-05, "loss": 0.007, "step": 15598 }, { "epoch": 1.691131830008673, "grad_norm": 0.2447998970746994, "learning_rate": 4.37871753953286e-05, "loss": 0.017, "step": 15599 }, { "epoch": 1.6912402428447528, "grad_norm": 0.7152003049850464, "learning_rate": 4.3783548527491655e-05, "loss": 0.0517, "step": 15600 }, { "epoch": 1.6913486556808326, "grad_norm": 0.6227346062660217, "learning_rate": 4.3779921659654726e-05, "loss": 0.0177, "step": 15601 }, { "epoch": 1.6914570685169124, "grad_norm": 0.27549034357070923, "learning_rate": 4.377629479181779e-05, "loss": 0.0141, "step": 15602 }, { "epoch": 1.6915654813529923, "grad_norm": 0.2757376730442047, "learning_rate": 4.3772667923980854e-05, "loss": 0.0089, "step": 15603 }, { "epoch": 1.6916738941890719, "grad_norm": 0.6513038277626038, "learning_rate": 4.376904105614392e-05, "loss": 0.0608, "step": 15604 }, { "epoch": 1.6917823070251519, "grad_norm": 0.17027653753757477, "learning_rate": 4.376541418830698e-05, "loss": 0.0151, "step": 15605 }, { "epoch": 1.6918907198612314, "grad_norm": 0.3333185613155365, "learning_rate": 4.3761787320470046e-05, "loss": 0.0167, "step": 15606 }, { "epoch": 1.6919991326973114, "grad_norm": 0.3861425220966339, "learning_rate": 4.3758160452633104e-05, "loss": 0.0347, "step": 15607 }, { "epoch": 1.6921075455333912, "grad_norm": 0.2589627206325531, "learning_rate": 4.375453358479617e-05, "loss": 0.0353, "step": 15608 }, { "epoch": 1.692215958369471, "grad_norm": 0.14674246311187744, "learning_rate": 4.375090671695923e-05, "loss": 0.018, "step": 15609 }, { "epoch": 1.692324371205551, "grad_norm": 0.3889748156070709, "learning_rate": 4.37472798491223e-05, "loss": 0.0173, "step": 15610 }, { "epoch": 1.6924327840416304, "grad_norm": 0.12530840933322906, "learning_rate": 4.374365298128537e-05, "loss": 0.0037, "step": 15611 }, { "epoch": 1.6925411968777104, "grad_norm": 0.37939611077308655, "learning_rate": 4.374002611344843e-05, "loss": 0.0236, "step": 15612 }, { "epoch": 1.69264960971379, "grad_norm": 0.20245689153671265, "learning_rate": 4.3736399245611495e-05, "loss": 0.0206, "step": 15613 }, { "epoch": 1.69275802254987, "grad_norm": 0.20162126421928406, "learning_rate": 4.373277237777455e-05, "loss": 0.0086, "step": 15614 }, { "epoch": 1.6928664353859497, "grad_norm": 0.7061413526535034, "learning_rate": 4.3729145509937616e-05, "loss": 0.0303, "step": 15615 }, { "epoch": 1.6929748482220295, "grad_norm": 0.21027608215808868, "learning_rate": 4.372551864210068e-05, "loss": 0.0233, "step": 15616 }, { "epoch": 1.6930832610581092, "grad_norm": 0.2656490206718445, "learning_rate": 4.3721891774263745e-05, "loss": 0.0385, "step": 15617 }, { "epoch": 1.693191673894189, "grad_norm": 0.2901257872581482, "learning_rate": 4.3718264906426816e-05, "loss": 0.0245, "step": 15618 }, { "epoch": 1.693300086730269, "grad_norm": 0.28342315554618835, "learning_rate": 4.371463803858988e-05, "loss": 0.0249, "step": 15619 }, { "epoch": 1.6934084995663485, "grad_norm": 0.28067663311958313, "learning_rate": 4.3711011170752944e-05, "loss": 0.0286, "step": 15620 }, { "epoch": 1.6935169124024285, "grad_norm": 0.6609609127044678, "learning_rate": 4.370738430291601e-05, "loss": 0.0411, "step": 15621 }, { "epoch": 1.6936253252385083, "grad_norm": 0.2710902988910675, "learning_rate": 4.3703757435079065e-05, "loss": 0.0112, "step": 15622 }, { "epoch": 1.693733738074588, "grad_norm": 0.021974293515086174, "learning_rate": 4.370013056724213e-05, "loss": 0.0012, "step": 15623 }, { "epoch": 1.6938421509106678, "grad_norm": 0.48093652725219727, "learning_rate": 4.3696503699405193e-05, "loss": 0.0231, "step": 15624 }, { "epoch": 1.6939505637467476, "grad_norm": 0.15479356050491333, "learning_rate": 4.369287683156826e-05, "loss": 0.0212, "step": 15625 }, { "epoch": 1.6940589765828276, "grad_norm": 0.40918511152267456, "learning_rate": 4.368924996373132e-05, "loss": 0.0358, "step": 15626 }, { "epoch": 1.694167389418907, "grad_norm": 0.40754467248916626, "learning_rate": 4.368562309589439e-05, "loss": 0.0345, "step": 15627 }, { "epoch": 1.694275802254987, "grad_norm": 0.3579609990119934, "learning_rate": 4.3681996228057457e-05, "loss": 0.0223, "step": 15628 }, { "epoch": 1.6943842150910666, "grad_norm": 0.3127800226211548, "learning_rate": 4.3678369360220514e-05, "loss": 0.0404, "step": 15629 }, { "epoch": 1.6944926279271466, "grad_norm": 0.858625054359436, "learning_rate": 4.367474249238358e-05, "loss": 0.0335, "step": 15630 }, { "epoch": 1.6946010407632264, "grad_norm": 0.1432904452085495, "learning_rate": 4.367111562454664e-05, "loss": 0.0122, "step": 15631 }, { "epoch": 1.6947094535993061, "grad_norm": 0.24882522225379944, "learning_rate": 4.3667488756709706e-05, "loss": 0.0274, "step": 15632 }, { "epoch": 1.694817866435386, "grad_norm": 0.1655498594045639, "learning_rate": 4.366386188887277e-05, "loss": 0.0153, "step": 15633 }, { "epoch": 1.6949262792714657, "grad_norm": 0.12012320756912231, "learning_rate": 4.3660235021035834e-05, "loss": 0.0065, "step": 15634 }, { "epoch": 1.6950346921075456, "grad_norm": 0.4035971164703369, "learning_rate": 4.36566081531989e-05, "loss": 0.0242, "step": 15635 }, { "epoch": 1.6951431049436252, "grad_norm": 0.26036348938941956, "learning_rate": 4.365298128536196e-05, "loss": 0.0158, "step": 15636 }, { "epoch": 1.6952515177797052, "grad_norm": 0.09884040802717209, "learning_rate": 4.364935441752503e-05, "loss": 0.0084, "step": 15637 }, { "epoch": 1.695359930615785, "grad_norm": 0.3204573690891266, "learning_rate": 4.364572754968809e-05, "loss": 0.025, "step": 15638 }, { "epoch": 1.6954683434518647, "grad_norm": 0.230797678232193, "learning_rate": 4.3642100681851155e-05, "loss": 0.0256, "step": 15639 }, { "epoch": 1.6955767562879445, "grad_norm": 0.23775726556777954, "learning_rate": 4.363847381401422e-05, "loss": 0.0197, "step": 15640 }, { "epoch": 1.6956851691240242, "grad_norm": 0.09049312770366669, "learning_rate": 4.363484694617728e-05, "loss": 0.0081, "step": 15641 }, { "epoch": 1.6957935819601042, "grad_norm": 0.2558969259262085, "learning_rate": 4.363122007834035e-05, "loss": 0.0282, "step": 15642 }, { "epoch": 1.6959019947961838, "grad_norm": 0.2051575481891632, "learning_rate": 4.362759321050341e-05, "loss": 0.0211, "step": 15643 }, { "epoch": 1.6960104076322637, "grad_norm": 0.10843249410390854, "learning_rate": 4.3623966342666476e-05, "loss": 0.0089, "step": 15644 }, { "epoch": 1.6961188204683435, "grad_norm": 0.06798864901065826, "learning_rate": 4.362033947482954e-05, "loss": 0.0052, "step": 15645 }, { "epoch": 1.6962272333044233, "grad_norm": 0.1468799114227295, "learning_rate": 4.3616712606992604e-05, "loss": 0.006, "step": 15646 }, { "epoch": 1.696335646140503, "grad_norm": 0.22319777309894562, "learning_rate": 4.361308573915567e-05, "loss": 0.0145, "step": 15647 }, { "epoch": 1.6964440589765828, "grad_norm": 0.027788227424025536, "learning_rate": 4.360945887131873e-05, "loss": 0.0019, "step": 15648 }, { "epoch": 1.6965524718126628, "grad_norm": 0.2717396914958954, "learning_rate": 4.3605832003481796e-05, "loss": 0.011, "step": 15649 }, { "epoch": 1.6966608846487423, "grad_norm": 0.02678760327398777, "learning_rate": 4.360220513564486e-05, "loss": 0.0011, "step": 15650 }, { "epoch": 1.6967692974848223, "grad_norm": 0.23567306995391846, "learning_rate": 4.3598578267807924e-05, "loss": 0.036, "step": 15651 }, { "epoch": 1.6968777103209018, "grad_norm": 0.4668150842189789, "learning_rate": 4.359495139997098e-05, "loss": 0.0676, "step": 15652 }, { "epoch": 1.6969861231569818, "grad_norm": 0.15638481080532074, "learning_rate": 4.359132453213405e-05, "loss": 0.0075, "step": 15653 }, { "epoch": 1.6970945359930616, "grad_norm": 0.553037166595459, "learning_rate": 4.3587697664297117e-05, "loss": 0.0437, "step": 15654 }, { "epoch": 1.6972029488291414, "grad_norm": 0.3159002363681793, "learning_rate": 4.358407079646018e-05, "loss": 0.0263, "step": 15655 }, { "epoch": 1.6973113616652211, "grad_norm": 0.3805217146873474, "learning_rate": 4.3580443928623245e-05, "loss": 0.0345, "step": 15656 }, { "epoch": 1.6974197745013009, "grad_norm": 0.28670403361320496, "learning_rate": 4.357681706078631e-05, "loss": 0.0245, "step": 15657 }, { "epoch": 1.6975281873373809, "grad_norm": 0.35583364963531494, "learning_rate": 4.357319019294937e-05, "loss": 0.0307, "step": 15658 }, { "epoch": 1.6976366001734604, "grad_norm": 0.7701783776283264, "learning_rate": 4.356956332511243e-05, "loss": 0.0703, "step": 15659 }, { "epoch": 1.6977450130095404, "grad_norm": 0.1875278800725937, "learning_rate": 4.3565936457275494e-05, "loss": 0.012, "step": 15660 }, { "epoch": 1.6978534258456202, "grad_norm": 0.005768290255218744, "learning_rate": 4.356230958943856e-05, "loss": 0.0004, "step": 15661 }, { "epoch": 1.6979618386817, "grad_norm": 0.38175293803215027, "learning_rate": 4.355868272160163e-05, "loss": 0.0283, "step": 15662 }, { "epoch": 1.6980702515177797, "grad_norm": 0.19362454116344452, "learning_rate": 4.3555055853764694e-05, "loss": 0.0136, "step": 15663 }, { "epoch": 1.6981786643538594, "grad_norm": 0.5107173919677734, "learning_rate": 4.355142898592776e-05, "loss": 0.0314, "step": 15664 }, { "epoch": 1.6982870771899394, "grad_norm": 0.44227781891822815, "learning_rate": 4.354780211809082e-05, "loss": 0.0247, "step": 15665 }, { "epoch": 1.698395490026019, "grad_norm": 0.4187566637992859, "learning_rate": 4.3544175250253886e-05, "loss": 0.041, "step": 15666 }, { "epoch": 1.698503902862099, "grad_norm": 0.23745545744895935, "learning_rate": 4.354054838241694e-05, "loss": 0.0258, "step": 15667 }, { "epoch": 1.6986123156981785, "grad_norm": 0.4831286072731018, "learning_rate": 4.353692151458001e-05, "loss": 0.0674, "step": 15668 }, { "epoch": 1.6987207285342585, "grad_norm": 0.625627338886261, "learning_rate": 4.353329464674307e-05, "loss": 0.0595, "step": 15669 }, { "epoch": 1.6988291413703382, "grad_norm": 0.8375439047813416, "learning_rate": 4.352966777890614e-05, "loss": 0.027, "step": 15670 }, { "epoch": 1.698937554206418, "grad_norm": 0.5550203919410706, "learning_rate": 4.3526040911069206e-05, "loss": 0.0781, "step": 15671 }, { "epoch": 1.699045967042498, "grad_norm": 0.16307033598423004, "learning_rate": 4.352241404323227e-05, "loss": 0.0081, "step": 15672 }, { "epoch": 1.6991543798785775, "grad_norm": 0.2038692682981491, "learning_rate": 4.3518787175395335e-05, "loss": 0.0269, "step": 15673 }, { "epoch": 1.6992627927146575, "grad_norm": 0.5105304718017578, "learning_rate": 4.351516030755839e-05, "loss": 0.0322, "step": 15674 }, { "epoch": 1.699371205550737, "grad_norm": 0.09459733963012695, "learning_rate": 4.3511533439721456e-05, "loss": 0.0105, "step": 15675 }, { "epoch": 1.699479618386817, "grad_norm": 0.13259786367416382, "learning_rate": 4.350790657188452e-05, "loss": 0.0098, "step": 15676 }, { "epoch": 1.6995880312228968, "grad_norm": 0.2961096167564392, "learning_rate": 4.3504279704047584e-05, "loss": 0.0192, "step": 15677 }, { "epoch": 1.6996964440589766, "grad_norm": 0.26659461855888367, "learning_rate": 4.350065283621065e-05, "loss": 0.0264, "step": 15678 }, { "epoch": 1.6998048568950563, "grad_norm": 0.08318758755922318, "learning_rate": 4.349702596837372e-05, "loss": 0.0079, "step": 15679 }, { "epoch": 1.699913269731136, "grad_norm": 0.4122038781642914, "learning_rate": 4.349339910053678e-05, "loss": 0.0201, "step": 15680 }, { "epoch": 1.700021682567216, "grad_norm": 0.4045594036579132, "learning_rate": 4.348977223269984e-05, "loss": 0.0424, "step": 15681 }, { "epoch": 1.7001300954032956, "grad_norm": 0.13943560421466827, "learning_rate": 4.3486145364862905e-05, "loss": 0.0103, "step": 15682 }, { "epoch": 1.7002385082393756, "grad_norm": 0.22103483974933624, "learning_rate": 4.348251849702597e-05, "loss": 0.0163, "step": 15683 }, { "epoch": 1.7003469210754554, "grad_norm": 0.037883590906858444, "learning_rate": 4.347889162918903e-05, "loss": 0.0019, "step": 15684 }, { "epoch": 1.7004553339115351, "grad_norm": 0.1215423122048378, "learning_rate": 4.34752647613521e-05, "loss": 0.0104, "step": 15685 }, { "epoch": 1.700563746747615, "grad_norm": 0.4600718319416046, "learning_rate": 4.347163789351516e-05, "loss": 0.0589, "step": 15686 }, { "epoch": 1.7006721595836947, "grad_norm": 0.4379611313343048, "learning_rate": 4.346801102567823e-05, "loss": 0.0406, "step": 15687 }, { "epoch": 1.7007805724197746, "grad_norm": 0.20324410498142242, "learning_rate": 4.346438415784129e-05, "loss": 0.0198, "step": 15688 }, { "epoch": 1.7008889852558542, "grad_norm": 0.19058379530906677, "learning_rate": 4.3460757290004353e-05, "loss": 0.011, "step": 15689 }, { "epoch": 1.7009973980919342, "grad_norm": 0.3343375027179718, "learning_rate": 4.345713042216742e-05, "loss": 0.0209, "step": 15690 }, { "epoch": 1.7011058109280137, "grad_norm": 0.2110946774482727, "learning_rate": 4.345350355433048e-05, "loss": 0.0236, "step": 15691 }, { "epoch": 1.7012142237640937, "grad_norm": 0.3484794795513153, "learning_rate": 4.3449876686493546e-05, "loss": 0.0153, "step": 15692 }, { "epoch": 1.7013226366001735, "grad_norm": 0.3930334150791168, "learning_rate": 4.344624981865661e-05, "loss": 0.0261, "step": 15693 }, { "epoch": 1.7014310494362532, "grad_norm": 0.38392719626426697, "learning_rate": 4.3442622950819674e-05, "loss": 0.0591, "step": 15694 }, { "epoch": 1.701539462272333, "grad_norm": 0.612844705581665, "learning_rate": 4.343899608298274e-05, "loss": 0.0543, "step": 15695 }, { "epoch": 1.7016478751084128, "grad_norm": 0.4858139753341675, "learning_rate": 4.34353692151458e-05, "loss": 0.0512, "step": 15696 }, { "epoch": 1.7017562879444927, "grad_norm": 0.15865086019039154, "learning_rate": 4.3431742347308866e-05, "loss": 0.0164, "step": 15697 }, { "epoch": 1.7018647007805723, "grad_norm": 0.2026917040348053, "learning_rate": 4.342811547947193e-05, "loss": 0.0112, "step": 15698 }, { "epoch": 1.7019731136166523, "grad_norm": 0.09332869201898575, "learning_rate": 4.3424488611634995e-05, "loss": 0.0053, "step": 15699 }, { "epoch": 1.702081526452732, "grad_norm": 0.09571392089128494, "learning_rate": 4.342086174379806e-05, "loss": 0.009, "step": 15700 }, { "epoch": 1.7021899392888118, "grad_norm": 0.3111511170864105, "learning_rate": 4.341723487596112e-05, "loss": 0.0352, "step": 15701 }, { "epoch": 1.7022983521248916, "grad_norm": 0.3011283874511719, "learning_rate": 4.341360800812419e-05, "loss": 0.0264, "step": 15702 }, { "epoch": 1.7024067649609713, "grad_norm": 0.16240833699703217, "learning_rate": 4.340998114028725e-05, "loss": 0.0228, "step": 15703 }, { "epoch": 1.7025151777970513, "grad_norm": 0.37778928875923157, "learning_rate": 4.340635427245031e-05, "loss": 0.0196, "step": 15704 }, { "epoch": 1.7026235906331308, "grad_norm": 0.3690347969532013, "learning_rate": 4.340272740461338e-05, "loss": 0.0265, "step": 15705 }, { "epoch": 1.7027320034692108, "grad_norm": 0.2580842673778534, "learning_rate": 4.339910053677644e-05, "loss": 0.0187, "step": 15706 }, { "epoch": 1.7028404163052906, "grad_norm": 0.11845749616622925, "learning_rate": 4.339547366893951e-05, "loss": 0.018, "step": 15707 }, { "epoch": 1.7029488291413704, "grad_norm": 0.2596184313297272, "learning_rate": 4.339184680110257e-05, "loss": 0.0273, "step": 15708 }, { "epoch": 1.7030572419774501, "grad_norm": 0.12215591222047806, "learning_rate": 4.3388219933265636e-05, "loss": 0.009, "step": 15709 }, { "epoch": 1.7031656548135299, "grad_norm": 0.19841855764389038, "learning_rate": 4.33845930654287e-05, "loss": 0.0141, "step": 15710 }, { "epoch": 1.7032740676496099, "grad_norm": 0.12258166819810867, "learning_rate": 4.338096619759176e-05, "loss": 0.009, "step": 15711 }, { "epoch": 1.7033824804856894, "grad_norm": 0.349374920129776, "learning_rate": 4.337733932975482e-05, "loss": 0.0176, "step": 15712 }, { "epoch": 1.7034908933217694, "grad_norm": 0.31255799531936646, "learning_rate": 4.337371246191789e-05, "loss": 0.0533, "step": 15713 }, { "epoch": 1.703599306157849, "grad_norm": 0.09275641292333603, "learning_rate": 4.3370085594080956e-05, "loss": 0.0082, "step": 15714 }, { "epoch": 1.703707718993929, "grad_norm": 0.043422408401966095, "learning_rate": 4.336645872624402e-05, "loss": 0.0028, "step": 15715 }, { "epoch": 1.7038161318300087, "grad_norm": 0.4435616731643677, "learning_rate": 4.3362831858407084e-05, "loss": 0.0235, "step": 15716 }, { "epoch": 1.7039245446660884, "grad_norm": 0.04818570986390114, "learning_rate": 4.335920499057015e-05, "loss": 0.0033, "step": 15717 }, { "epoch": 1.7040329575021682, "grad_norm": 0.20985819399356842, "learning_rate": 4.335557812273321e-05, "loss": 0.0178, "step": 15718 }, { "epoch": 1.704141370338248, "grad_norm": 0.24609655141830444, "learning_rate": 4.335195125489627e-05, "loss": 0.041, "step": 15719 }, { "epoch": 1.704249783174328, "grad_norm": 0.15674085915088654, "learning_rate": 4.3348324387059334e-05, "loss": 0.0173, "step": 15720 }, { "epoch": 1.7043581960104075, "grad_norm": 0.612869381904602, "learning_rate": 4.33446975192224e-05, "loss": 0.0484, "step": 15721 }, { "epoch": 1.7044666088464875, "grad_norm": 0.6763842701911926, "learning_rate": 4.334107065138547e-05, "loss": 0.024, "step": 15722 }, { "epoch": 1.7045750216825672, "grad_norm": 0.36901384592056274, "learning_rate": 4.333744378354853e-05, "loss": 0.0126, "step": 15723 }, { "epoch": 1.704683434518647, "grad_norm": 0.38659122586250305, "learning_rate": 4.33338169157116e-05, "loss": 0.0391, "step": 15724 }, { "epoch": 1.7047918473547268, "grad_norm": 0.6230427026748657, "learning_rate": 4.333019004787466e-05, "loss": 0.0158, "step": 15725 }, { "epoch": 1.7049002601908065, "grad_norm": 0.13615868985652924, "learning_rate": 4.332656318003772e-05, "loss": 0.009, "step": 15726 }, { "epoch": 1.7050086730268865, "grad_norm": 0.07159395515918732, "learning_rate": 4.332293631220078e-05, "loss": 0.0043, "step": 15727 }, { "epoch": 1.705117085862966, "grad_norm": 0.08962401002645493, "learning_rate": 4.331930944436385e-05, "loss": 0.0063, "step": 15728 }, { "epoch": 1.705225498699046, "grad_norm": 0.1687733232975006, "learning_rate": 4.331568257652691e-05, "loss": 0.0244, "step": 15729 }, { "epoch": 1.7053339115351258, "grad_norm": 0.09431269019842148, "learning_rate": 4.3312055708689975e-05, "loss": 0.0073, "step": 15730 }, { "epoch": 1.7054423243712056, "grad_norm": 0.2285928577184677, "learning_rate": 4.3308428840853046e-05, "loss": 0.0383, "step": 15731 }, { "epoch": 1.7055507372072853, "grad_norm": 0.14867903292179108, "learning_rate": 4.330480197301611e-05, "loss": 0.0135, "step": 15732 }, { "epoch": 1.705659150043365, "grad_norm": 0.49203184247016907, "learning_rate": 4.330117510517917e-05, "loss": 0.0614, "step": 15733 }, { "epoch": 1.705767562879445, "grad_norm": 0.5708432197570801, "learning_rate": 4.329754823734223e-05, "loss": 0.0543, "step": 15734 }, { "epoch": 1.7058759757155246, "grad_norm": 0.06530139595270157, "learning_rate": 4.3293921369505296e-05, "loss": 0.003, "step": 15735 }, { "epoch": 1.7059843885516046, "grad_norm": 0.02785305492579937, "learning_rate": 4.329029450166836e-05, "loss": 0.0014, "step": 15736 }, { "epoch": 1.7060928013876842, "grad_norm": 0.3406427204608917, "learning_rate": 4.3286667633831424e-05, "loss": 0.045, "step": 15737 }, { "epoch": 1.7062012142237641, "grad_norm": 0.3180306553840637, "learning_rate": 4.328304076599449e-05, "loss": 0.0182, "step": 15738 }, { "epoch": 1.706309627059844, "grad_norm": 0.3079701066017151, "learning_rate": 4.327941389815756e-05, "loss": 0.0128, "step": 15739 }, { "epoch": 1.7064180398959237, "grad_norm": 0.3760928511619568, "learning_rate": 4.3275787030320616e-05, "loss": 0.0182, "step": 15740 }, { "epoch": 1.7065264527320034, "grad_norm": 0.30796536803245544, "learning_rate": 4.327216016248368e-05, "loss": 0.022, "step": 15741 }, { "epoch": 1.7066348655680832, "grad_norm": 0.15695926547050476, "learning_rate": 4.3268533294646744e-05, "loss": 0.0068, "step": 15742 }, { "epoch": 1.7067432784041632, "grad_norm": 0.3845485746860504, "learning_rate": 4.326490642680981e-05, "loss": 0.0241, "step": 15743 }, { "epoch": 1.7068516912402427, "grad_norm": 0.43590623140335083, "learning_rate": 4.326127955897287e-05, "loss": 0.024, "step": 15744 }, { "epoch": 1.7069601040763227, "grad_norm": 0.052748292684555054, "learning_rate": 4.3257652691135937e-05, "loss": 0.0023, "step": 15745 }, { "epoch": 1.7070685169124025, "grad_norm": 0.347948282957077, "learning_rate": 4.3254025823299e-05, "loss": 0.0141, "step": 15746 }, { "epoch": 1.7071769297484822, "grad_norm": 0.15960481762886047, "learning_rate": 4.3250398955462065e-05, "loss": 0.0244, "step": 15747 }, { "epoch": 1.707285342584562, "grad_norm": 0.10539225488901138, "learning_rate": 4.324677208762513e-05, "loss": 0.008, "step": 15748 }, { "epoch": 1.7073937554206418, "grad_norm": 0.36667364835739136, "learning_rate": 4.324314521978819e-05, "loss": 0.0176, "step": 15749 }, { "epoch": 1.7075021682567217, "grad_norm": 0.2778635025024414, "learning_rate": 4.323951835195126e-05, "loss": 0.0209, "step": 15750 }, { "epoch": 1.7076105810928013, "grad_norm": 0.7102178931236267, "learning_rate": 4.323589148411432e-05, "loss": 0.0815, "step": 15751 }, { "epoch": 1.7077189939288813, "grad_norm": 0.30185577273368835, "learning_rate": 4.3232264616277385e-05, "loss": 0.0476, "step": 15752 }, { "epoch": 1.7078274067649608, "grad_norm": 0.15919694304466248, "learning_rate": 4.322863774844045e-05, "loss": 0.006, "step": 15753 }, { "epoch": 1.7079358196010408, "grad_norm": 0.18807919323444366, "learning_rate": 4.3225010880603514e-05, "loss": 0.0043, "step": 15754 }, { "epoch": 1.7080442324371206, "grad_norm": 0.4728723466396332, "learning_rate": 4.322138401276658e-05, "loss": 0.0338, "step": 15755 }, { "epoch": 1.7081526452732003, "grad_norm": 0.31317460536956787, "learning_rate": 4.3217757144929635e-05, "loss": 0.0206, "step": 15756 }, { "epoch": 1.70826105810928, "grad_norm": 0.25442370772361755, "learning_rate": 4.3214130277092706e-05, "loss": 0.0249, "step": 15757 }, { "epoch": 1.7083694709453598, "grad_norm": 0.2157474011182785, "learning_rate": 4.321050340925577e-05, "loss": 0.0134, "step": 15758 }, { "epoch": 1.7084778837814398, "grad_norm": 0.1539584994316101, "learning_rate": 4.3206876541418834e-05, "loss": 0.0221, "step": 15759 }, { "epoch": 1.7085862966175194, "grad_norm": 0.17916631698608398, "learning_rate": 4.32032496735819e-05, "loss": 0.0147, "step": 15760 }, { "epoch": 1.7086947094535994, "grad_norm": 0.2050035446882248, "learning_rate": 4.319962280574496e-05, "loss": 0.0149, "step": 15761 }, { "epoch": 1.7088031222896791, "grad_norm": 0.27636754512786865, "learning_rate": 4.3195995937908026e-05, "loss": 0.024, "step": 15762 }, { "epoch": 1.7089115351257589, "grad_norm": 0.48844221234321594, "learning_rate": 4.3192369070071084e-05, "loss": 0.0362, "step": 15763 }, { "epoch": 1.7090199479618386, "grad_norm": 0.2602389454841614, "learning_rate": 4.318874220223415e-05, "loss": 0.0294, "step": 15764 }, { "epoch": 1.7091283607979184, "grad_norm": 0.35248854756355286, "learning_rate": 4.318511533439722e-05, "loss": 0.0499, "step": 15765 }, { "epoch": 1.7092367736339984, "grad_norm": 0.23300288617610931, "learning_rate": 4.318148846656028e-05, "loss": 0.0074, "step": 15766 }, { "epoch": 1.709345186470078, "grad_norm": 0.17339864373207092, "learning_rate": 4.317786159872335e-05, "loss": 0.0094, "step": 15767 }, { "epoch": 1.709453599306158, "grad_norm": 0.23154453933238983, "learning_rate": 4.317423473088641e-05, "loss": 0.0233, "step": 15768 }, { "epoch": 1.7095620121422377, "grad_norm": 0.1886877715587616, "learning_rate": 4.3170607863049475e-05, "loss": 0.0221, "step": 15769 }, { "epoch": 1.7096704249783174, "grad_norm": 0.21473585069179535, "learning_rate": 4.316698099521254e-05, "loss": 0.0194, "step": 15770 }, { "epoch": 1.7097788378143972, "grad_norm": 0.0525059849023819, "learning_rate": 4.3163354127375597e-05, "loss": 0.0036, "step": 15771 }, { "epoch": 1.709887250650477, "grad_norm": 0.14496874809265137, "learning_rate": 4.315972725953866e-05, "loss": 0.0139, "step": 15772 }, { "epoch": 1.709995663486557, "grad_norm": 0.1588352471590042, "learning_rate": 4.3156100391701725e-05, "loss": 0.0119, "step": 15773 }, { "epoch": 1.7101040763226365, "grad_norm": 0.1096348762512207, "learning_rate": 4.3152473523864796e-05, "loss": 0.0085, "step": 15774 }, { "epoch": 1.7102124891587165, "grad_norm": 0.44623640179634094, "learning_rate": 4.314884665602786e-05, "loss": 0.021, "step": 15775 }, { "epoch": 1.710320901994796, "grad_norm": 0.7725031971931458, "learning_rate": 4.3145219788190924e-05, "loss": 0.0458, "step": 15776 }, { "epoch": 1.710429314830876, "grad_norm": 0.03417736664414406, "learning_rate": 4.314159292035399e-05, "loss": 0.0014, "step": 15777 }, { "epoch": 1.7105377276669558, "grad_norm": 0.03640256077051163, "learning_rate": 4.3137966052517045e-05, "loss": 0.0014, "step": 15778 }, { "epoch": 1.7106461405030355, "grad_norm": 0.17702549695968628, "learning_rate": 4.313433918468011e-05, "loss": 0.021, "step": 15779 }, { "epoch": 1.7107545533391153, "grad_norm": 0.32824093103408813, "learning_rate": 4.3130712316843173e-05, "loss": 0.017, "step": 15780 }, { "epoch": 1.710862966175195, "grad_norm": 0.05319412425160408, "learning_rate": 4.312708544900624e-05, "loss": 0.0031, "step": 15781 }, { "epoch": 1.710971379011275, "grad_norm": 0.1313643902540207, "learning_rate": 4.31234585811693e-05, "loss": 0.0051, "step": 15782 }, { "epoch": 1.7110797918473546, "grad_norm": 0.18495501577854156, "learning_rate": 4.311983171333237e-05, "loss": 0.0281, "step": 15783 }, { "epoch": 1.7111882046834346, "grad_norm": 0.07902238517999649, "learning_rate": 4.311620484549544e-05, "loss": 0.0066, "step": 15784 }, { "epoch": 1.7112966175195143, "grad_norm": 0.7749300599098206, "learning_rate": 4.3112577977658494e-05, "loss": 0.0395, "step": 15785 }, { "epoch": 1.711405030355594, "grad_norm": 0.4724506437778473, "learning_rate": 4.310895110982156e-05, "loss": 0.0232, "step": 15786 }, { "epoch": 1.7115134431916739, "grad_norm": 0.10278249531984329, "learning_rate": 4.310532424198462e-05, "loss": 0.0062, "step": 15787 }, { "epoch": 1.7116218560277536, "grad_norm": 0.1628412902355194, "learning_rate": 4.3101697374147686e-05, "loss": 0.0069, "step": 15788 }, { "epoch": 1.7117302688638336, "grad_norm": 0.21595317125320435, "learning_rate": 4.309807050631075e-05, "loss": 0.0121, "step": 15789 }, { "epoch": 1.7118386816999132, "grad_norm": 0.6344152092933655, "learning_rate": 4.3094443638473815e-05, "loss": 0.035, "step": 15790 }, { "epoch": 1.7119470945359931, "grad_norm": 0.4543375074863434, "learning_rate": 4.3090816770636885e-05, "loss": 0.0204, "step": 15791 }, { "epoch": 1.712055507372073, "grad_norm": 0.5367918014526367, "learning_rate": 4.308718990279994e-05, "loss": 0.0345, "step": 15792 }, { "epoch": 1.7121639202081527, "grad_norm": 0.6101149916648865, "learning_rate": 4.308356303496301e-05, "loss": 0.0248, "step": 15793 }, { "epoch": 1.7122723330442324, "grad_norm": 0.5383543372154236, "learning_rate": 4.307993616712607e-05, "loss": 0.0679, "step": 15794 }, { "epoch": 1.7123807458803122, "grad_norm": 0.23939907550811768, "learning_rate": 4.3076309299289135e-05, "loss": 0.0083, "step": 15795 }, { "epoch": 1.7124891587163922, "grad_norm": 0.06517472118139267, "learning_rate": 4.30726824314522e-05, "loss": 0.0028, "step": 15796 }, { "epoch": 1.7125975715524717, "grad_norm": 0.22351668775081635, "learning_rate": 4.306905556361526e-05, "loss": 0.0096, "step": 15797 }, { "epoch": 1.7127059843885517, "grad_norm": 0.3224882483482361, "learning_rate": 4.306542869577833e-05, "loss": 0.0098, "step": 15798 }, { "epoch": 1.7128143972246312, "grad_norm": 0.4052731394767761, "learning_rate": 4.306180182794139e-05, "loss": 0.0139, "step": 15799 }, { "epoch": 1.7129228100607112, "grad_norm": 0.2964823246002197, "learning_rate": 4.3058174960104456e-05, "loss": 0.0185, "step": 15800 }, { "epoch": 1.713031222896791, "grad_norm": 0.3625851273536682, "learning_rate": 4.305454809226752e-05, "loss": 0.039, "step": 15801 }, { "epoch": 1.7131396357328708, "grad_norm": 0.852347195148468, "learning_rate": 4.3050921224430584e-05, "loss": 0.0241, "step": 15802 }, { "epoch": 1.7132480485689505, "grad_norm": 0.48941007256507874, "learning_rate": 4.304729435659365e-05, "loss": 0.0459, "step": 15803 }, { "epoch": 1.7133564614050303, "grad_norm": 1.4858745336532593, "learning_rate": 4.304366748875671e-05, "loss": 0.0437, "step": 15804 }, { "epoch": 1.7134648742411103, "grad_norm": 0.2600640654563904, "learning_rate": 4.3040040620919776e-05, "loss": 0.016, "step": 15805 }, { "epoch": 1.7135732870771898, "grad_norm": 0.07549770921468735, "learning_rate": 4.303641375308284e-05, "loss": 0.0058, "step": 15806 }, { "epoch": 1.7136816999132698, "grad_norm": 0.6127772331237793, "learning_rate": 4.3032786885245904e-05, "loss": 0.0531, "step": 15807 }, { "epoch": 1.7137901127493496, "grad_norm": 0.30518245697021484, "learning_rate": 4.302916001740897e-05, "loss": 0.0133, "step": 15808 }, { "epoch": 1.7138985255854293, "grad_norm": 0.4526020586490631, "learning_rate": 4.302553314957203e-05, "loss": 0.0284, "step": 15809 }, { "epoch": 1.714006938421509, "grad_norm": 0.7302059531211853, "learning_rate": 4.30219062817351e-05, "loss": 0.0156, "step": 15810 }, { "epoch": 1.7141153512575888, "grad_norm": 0.15323589742183685, "learning_rate": 4.301827941389816e-05, "loss": 0.0046, "step": 15811 }, { "epoch": 1.7142237640936688, "grad_norm": 0.18119381368160248, "learning_rate": 4.3014652546061225e-05, "loss": 0.0099, "step": 15812 }, { "epoch": 1.7143321769297484, "grad_norm": 0.254427045583725, "learning_rate": 4.301102567822429e-05, "loss": 0.0203, "step": 15813 }, { "epoch": 1.7144405897658284, "grad_norm": 0.06502872705459595, "learning_rate": 4.300739881038735e-05, "loss": 0.0034, "step": 15814 }, { "epoch": 1.714549002601908, "grad_norm": 0.6894636154174805, "learning_rate": 4.300377194255041e-05, "loss": 0.0679, "step": 15815 }, { "epoch": 1.7146574154379879, "grad_norm": 0.12471593171358109, "learning_rate": 4.3000145074713474e-05, "loss": 0.0049, "step": 15816 }, { "epoch": 1.7147658282740676, "grad_norm": 0.10403747111558914, "learning_rate": 4.2996518206876545e-05, "loss": 0.0083, "step": 15817 }, { "epoch": 1.7148742411101474, "grad_norm": 0.19078484177589417, "learning_rate": 4.299289133903961e-05, "loss": 0.0107, "step": 15818 }, { "epoch": 1.7149826539462274, "grad_norm": 0.2141920030117035, "learning_rate": 4.2989264471202674e-05, "loss": 0.0109, "step": 15819 }, { "epoch": 1.715091066782307, "grad_norm": 0.3557300567626953, "learning_rate": 4.298563760336574e-05, "loss": 0.0334, "step": 15820 }, { "epoch": 1.715199479618387, "grad_norm": 0.3135404884815216, "learning_rate": 4.29820107355288e-05, "loss": 0.0189, "step": 15821 }, { "epoch": 1.7153078924544665, "grad_norm": 0.6330282688140869, "learning_rate": 4.2978383867691866e-05, "loss": 0.0554, "step": 15822 }, { "epoch": 1.7154163052905465, "grad_norm": 0.46567484736442566, "learning_rate": 4.297475699985492e-05, "loss": 0.0147, "step": 15823 }, { "epoch": 1.7155247181266262, "grad_norm": 0.151263028383255, "learning_rate": 4.297113013201799e-05, "loss": 0.0037, "step": 15824 }, { "epoch": 1.715633130962706, "grad_norm": 0.4436873197555542, "learning_rate": 4.296750326418105e-05, "loss": 0.0538, "step": 15825 }, { "epoch": 1.7157415437987857, "grad_norm": 0.31212612986564636, "learning_rate": 4.296387639634412e-05, "loss": 0.0181, "step": 15826 }, { "epoch": 1.7158499566348655, "grad_norm": 0.405695378780365, "learning_rate": 4.2960249528507186e-05, "loss": 0.015, "step": 15827 }, { "epoch": 1.7159583694709455, "grad_norm": 0.697890043258667, "learning_rate": 4.295662266067025e-05, "loss": 0.0316, "step": 15828 }, { "epoch": 1.716066782307025, "grad_norm": 0.09331043064594269, "learning_rate": 4.2952995792833315e-05, "loss": 0.0039, "step": 15829 }, { "epoch": 1.716175195143105, "grad_norm": 0.16684338450431824, "learning_rate": 4.294936892499637e-05, "loss": 0.0084, "step": 15830 }, { "epoch": 1.7162836079791848, "grad_norm": 0.46341419219970703, "learning_rate": 4.2945742057159436e-05, "loss": 0.0286, "step": 15831 }, { "epoch": 1.7163920208152645, "grad_norm": 0.3314788043498993, "learning_rate": 4.29421151893225e-05, "loss": 0.02, "step": 15832 }, { "epoch": 1.7165004336513443, "grad_norm": 0.5484016537666321, "learning_rate": 4.2938488321485564e-05, "loss": 0.0228, "step": 15833 }, { "epoch": 1.716608846487424, "grad_norm": 0.4102276563644409, "learning_rate": 4.2934861453648635e-05, "loss": 0.0258, "step": 15834 }, { "epoch": 1.716717259323504, "grad_norm": 0.2572482228279114, "learning_rate": 4.29312345858117e-05, "loss": 0.0145, "step": 15835 }, { "epoch": 1.7168256721595836, "grad_norm": 0.6547044515609741, "learning_rate": 4.292760771797476e-05, "loss": 0.0357, "step": 15836 }, { "epoch": 1.7169340849956636, "grad_norm": 0.42947131395339966, "learning_rate": 4.292398085013782e-05, "loss": 0.0521, "step": 15837 }, { "epoch": 1.7170424978317431, "grad_norm": 0.20895196497440338, "learning_rate": 4.2920353982300885e-05, "loss": 0.0301, "step": 15838 }, { "epoch": 1.717150910667823, "grad_norm": 0.45096999406814575, "learning_rate": 4.291672711446395e-05, "loss": 0.0285, "step": 15839 }, { "epoch": 1.7172593235039029, "grad_norm": 0.13927926123142242, "learning_rate": 4.291310024662701e-05, "loss": 0.0041, "step": 15840 }, { "epoch": 1.7173677363399826, "grad_norm": 0.7014660239219666, "learning_rate": 4.290947337879008e-05, "loss": 0.0683, "step": 15841 }, { "epoch": 1.7174761491760624, "grad_norm": 0.20155060291290283, "learning_rate": 4.290584651095314e-05, "loss": 0.011, "step": 15842 }, { "epoch": 1.7175845620121422, "grad_norm": 0.26306402683258057, "learning_rate": 4.290221964311621e-05, "loss": 0.0198, "step": 15843 }, { "epoch": 1.7176929748482221, "grad_norm": 0.25407326221466064, "learning_rate": 4.289859277527927e-05, "loss": 0.0339, "step": 15844 }, { "epoch": 1.7178013876843017, "grad_norm": 0.5070744752883911, "learning_rate": 4.2894965907442334e-05, "loss": 0.0402, "step": 15845 }, { "epoch": 1.7179098005203817, "grad_norm": 0.3079254925251007, "learning_rate": 4.28913390396054e-05, "loss": 0.0221, "step": 15846 }, { "epoch": 1.7180182133564614, "grad_norm": 0.3404172956943512, "learning_rate": 4.288771217176846e-05, "loss": 0.0334, "step": 15847 }, { "epoch": 1.7181266261925412, "grad_norm": 0.08514485508203506, "learning_rate": 4.2884085303931526e-05, "loss": 0.0071, "step": 15848 }, { "epoch": 1.718235039028621, "grad_norm": 0.21012650430202484, "learning_rate": 4.288045843609459e-05, "loss": 0.0258, "step": 15849 }, { "epoch": 1.7183434518647007, "grad_norm": 0.20566070079803467, "learning_rate": 4.2876831568257654e-05, "loss": 0.0232, "step": 15850 }, { "epoch": 1.7184518647007807, "grad_norm": 0.08203715831041336, "learning_rate": 4.287320470042072e-05, "loss": 0.0139, "step": 15851 }, { "epoch": 1.7185602775368602, "grad_norm": 0.8089624643325806, "learning_rate": 4.286957783258378e-05, "loss": 0.0452, "step": 15852 }, { "epoch": 1.7186686903729402, "grad_norm": 0.3003951609134674, "learning_rate": 4.2865950964746846e-05, "loss": 0.0159, "step": 15853 }, { "epoch": 1.71877710320902, "grad_norm": 0.4201869070529938, "learning_rate": 4.286232409690991e-05, "loss": 0.0142, "step": 15854 }, { "epoch": 1.7188855160450998, "grad_norm": 0.23632889986038208, "learning_rate": 4.2858697229072975e-05, "loss": 0.0765, "step": 15855 }, { "epoch": 1.7189939288811795, "grad_norm": 0.45512527227401733, "learning_rate": 4.285507036123604e-05, "loss": 0.0299, "step": 15856 }, { "epoch": 1.7191023417172593, "grad_norm": 0.367042601108551, "learning_rate": 4.28514434933991e-05, "loss": 0.0193, "step": 15857 }, { "epoch": 1.7192107545533393, "grad_norm": 0.3581160306930542, "learning_rate": 4.284781662556217e-05, "loss": 0.0368, "step": 15858 }, { "epoch": 1.7193191673894188, "grad_norm": 0.14906927943229675, "learning_rate": 4.284418975772523e-05, "loss": 0.0116, "step": 15859 }, { "epoch": 1.7194275802254988, "grad_norm": 0.2289534956216812, "learning_rate": 4.2840562889888295e-05, "loss": 0.0267, "step": 15860 }, { "epoch": 1.7195359930615783, "grad_norm": 0.44281789660453796, "learning_rate": 4.283693602205136e-05, "loss": 0.0219, "step": 15861 }, { "epoch": 1.7196444058976583, "grad_norm": 0.16992683708667755, "learning_rate": 4.283330915421442e-05, "loss": 0.0116, "step": 15862 }, { "epoch": 1.719752818733738, "grad_norm": 0.2689531743526459, "learning_rate": 4.282968228637749e-05, "loss": 0.01, "step": 15863 }, { "epoch": 1.7198612315698178, "grad_norm": 0.33088254928588867, "learning_rate": 4.282605541854055e-05, "loss": 0.0268, "step": 15864 }, { "epoch": 1.7199696444058976, "grad_norm": 0.1620888113975525, "learning_rate": 4.2822428550703616e-05, "loss": 0.0177, "step": 15865 }, { "epoch": 1.7200780572419774, "grad_norm": 0.28123366832733154, "learning_rate": 4.281880168286668e-05, "loss": 0.0421, "step": 15866 }, { "epoch": 1.7201864700780574, "grad_norm": 0.3316076695919037, "learning_rate": 4.2815174815029744e-05, "loss": 0.0613, "step": 15867 }, { "epoch": 1.720294882914137, "grad_norm": 0.3896149694919586, "learning_rate": 4.28115479471928e-05, "loss": 0.0332, "step": 15868 }, { "epoch": 1.7204032957502169, "grad_norm": 0.13200156390666962, "learning_rate": 4.280792107935587e-05, "loss": 0.0067, "step": 15869 }, { "epoch": 1.7205117085862967, "grad_norm": 0.05392592400312424, "learning_rate": 4.2804294211518936e-05, "loss": 0.0035, "step": 15870 }, { "epoch": 1.7206201214223764, "grad_norm": 0.3757249414920807, "learning_rate": 4.2800667343682e-05, "loss": 0.0368, "step": 15871 }, { "epoch": 1.7207285342584562, "grad_norm": 0.4604334235191345, "learning_rate": 4.2797040475845064e-05, "loss": 0.0266, "step": 15872 }, { "epoch": 1.720836947094536, "grad_norm": 0.5963082909584045, "learning_rate": 4.279341360800813e-05, "loss": 0.0607, "step": 15873 }, { "epoch": 1.720945359930616, "grad_norm": 0.10352315753698349, "learning_rate": 4.278978674017119e-05, "loss": 0.0048, "step": 15874 }, { "epoch": 1.7210537727666955, "grad_norm": 0.29357945919036865, "learning_rate": 4.278615987233425e-05, "loss": 0.0108, "step": 15875 }, { "epoch": 1.7211621856027755, "grad_norm": 0.26145440340042114, "learning_rate": 4.2782533004497314e-05, "loss": 0.0313, "step": 15876 }, { "epoch": 1.7212705984388552, "grad_norm": 0.3474093973636627, "learning_rate": 4.277890613666038e-05, "loss": 0.05, "step": 15877 }, { "epoch": 1.721379011274935, "grad_norm": 0.38487231731414795, "learning_rate": 4.277527926882345e-05, "loss": 0.0252, "step": 15878 }, { "epoch": 1.7214874241110147, "grad_norm": 0.234051913022995, "learning_rate": 4.277165240098651e-05, "loss": 0.0285, "step": 15879 }, { "epoch": 1.7215958369470945, "grad_norm": 0.2222764641046524, "learning_rate": 4.276802553314958e-05, "loss": 0.0319, "step": 15880 }, { "epoch": 1.7217042497831745, "grad_norm": 0.27270346879959106, "learning_rate": 4.276439866531264e-05, "loss": 0.0282, "step": 15881 }, { "epoch": 1.721812662619254, "grad_norm": 0.13747382164001465, "learning_rate": 4.27607717974757e-05, "loss": 0.0052, "step": 15882 }, { "epoch": 1.721921075455334, "grad_norm": 0.1995791345834732, "learning_rate": 4.275714492963876e-05, "loss": 0.0114, "step": 15883 }, { "epoch": 1.7220294882914136, "grad_norm": 0.41125375032424927, "learning_rate": 4.275351806180183e-05, "loss": 0.0096, "step": 15884 }, { "epoch": 1.7221379011274935, "grad_norm": 0.49370816349983215, "learning_rate": 4.274989119396489e-05, "loss": 0.0205, "step": 15885 }, { "epoch": 1.7222463139635733, "grad_norm": 0.29042959213256836, "learning_rate": 4.274626432612796e-05, "loss": 0.0189, "step": 15886 }, { "epoch": 1.722354726799653, "grad_norm": 0.06169931963086128, "learning_rate": 4.2742637458291026e-05, "loss": 0.004, "step": 15887 }, { "epoch": 1.7224631396357328, "grad_norm": 0.28514906764030457, "learning_rate": 4.273901059045409e-05, "loss": 0.0106, "step": 15888 }, { "epoch": 1.7225715524718126, "grad_norm": 0.2183855175971985, "learning_rate": 4.273538372261715e-05, "loss": 0.0214, "step": 15889 }, { "epoch": 1.7226799653078926, "grad_norm": 0.6889684200286865, "learning_rate": 4.273175685478021e-05, "loss": 0.0232, "step": 15890 }, { "epoch": 1.7227883781439721, "grad_norm": 0.2572982609272003, "learning_rate": 4.2728129986943276e-05, "loss": 0.0183, "step": 15891 }, { "epoch": 1.722896790980052, "grad_norm": 0.33859097957611084, "learning_rate": 4.272450311910634e-05, "loss": 0.0251, "step": 15892 }, { "epoch": 1.7230052038161319, "grad_norm": 0.4112875163555145, "learning_rate": 4.2720876251269404e-05, "loss": 0.0308, "step": 15893 }, { "epoch": 1.7231136166522116, "grad_norm": 0.42577579617500305, "learning_rate": 4.271724938343247e-05, "loss": 0.0521, "step": 15894 }, { "epoch": 1.7232220294882914, "grad_norm": 0.3868353068828583, "learning_rate": 4.271362251559554e-05, "loss": 0.028, "step": 15895 }, { "epoch": 1.7233304423243712, "grad_norm": 0.15106746554374695, "learning_rate": 4.2709995647758596e-05, "loss": 0.012, "step": 15896 }, { "epoch": 1.7234388551604511, "grad_norm": 0.3594241440296173, "learning_rate": 4.270636877992166e-05, "loss": 0.048, "step": 15897 }, { "epoch": 1.7235472679965307, "grad_norm": 0.05298831686377525, "learning_rate": 4.2702741912084724e-05, "loss": 0.0075, "step": 15898 }, { "epoch": 1.7236556808326107, "grad_norm": 0.1395510882139206, "learning_rate": 4.269911504424779e-05, "loss": 0.0075, "step": 15899 }, { "epoch": 1.7237640936686902, "grad_norm": 0.06833766400814056, "learning_rate": 4.269548817641085e-05, "loss": 0.0067, "step": 15900 }, { "epoch": 1.7238725065047702, "grad_norm": 0.17720459401607513, "learning_rate": 4.269186130857392e-05, "loss": 0.014, "step": 15901 }, { "epoch": 1.72398091934085, "grad_norm": 0.2600850760936737, "learning_rate": 4.268823444073698e-05, "loss": 0.0197, "step": 15902 }, { "epoch": 1.7240893321769297, "grad_norm": 0.17691729962825775, "learning_rate": 4.268460757290005e-05, "loss": 0.0191, "step": 15903 }, { "epoch": 1.7241977450130095, "grad_norm": 0.25652503967285156, "learning_rate": 4.268098070506311e-05, "loss": 0.0205, "step": 15904 }, { "epoch": 1.7243061578490892, "grad_norm": 0.12971435487270355, "learning_rate": 4.267735383722617e-05, "loss": 0.0085, "step": 15905 }, { "epoch": 1.7244145706851692, "grad_norm": 0.34724536538124084, "learning_rate": 4.267372696938924e-05, "loss": 0.0269, "step": 15906 }, { "epoch": 1.7245229835212488, "grad_norm": 0.3736743628978729, "learning_rate": 4.26701001015523e-05, "loss": 0.0348, "step": 15907 }, { "epoch": 1.7246313963573288, "grad_norm": 0.23854582011699677, "learning_rate": 4.2666473233715365e-05, "loss": 0.0245, "step": 15908 }, { "epoch": 1.7247398091934085, "grad_norm": 0.32235783338546753, "learning_rate": 4.266284636587843e-05, "loss": 0.0214, "step": 15909 }, { "epoch": 1.7248482220294883, "grad_norm": 0.2046772837638855, "learning_rate": 4.2659219498041494e-05, "loss": 0.0355, "step": 15910 }, { "epoch": 1.724956634865568, "grad_norm": 0.2212296724319458, "learning_rate": 4.265559263020456e-05, "loss": 0.0183, "step": 15911 }, { "epoch": 1.7250650477016478, "grad_norm": 0.3337435722351074, "learning_rate": 4.265196576236762e-05, "loss": 0.0291, "step": 15912 }, { "epoch": 1.7251734605377278, "grad_norm": 0.34249401092529297, "learning_rate": 4.2648338894530686e-05, "loss": 0.0116, "step": 15913 }, { "epoch": 1.7252818733738073, "grad_norm": 0.4315585494041443, "learning_rate": 4.264471202669375e-05, "loss": 0.045, "step": 15914 }, { "epoch": 1.7253902862098873, "grad_norm": 0.18158161640167236, "learning_rate": 4.2641085158856814e-05, "loss": 0.013, "step": 15915 }, { "epoch": 1.725498699045967, "grad_norm": 0.167836531996727, "learning_rate": 4.263745829101988e-05, "loss": 0.0069, "step": 15916 }, { "epoch": 1.7256071118820469, "grad_norm": 0.41904184222221375, "learning_rate": 4.263383142318294e-05, "loss": 0.0395, "step": 15917 }, { "epoch": 1.7257155247181266, "grad_norm": 0.3175380825996399, "learning_rate": 4.2630204555346006e-05, "loss": 0.0318, "step": 15918 }, { "epoch": 1.7258239375542064, "grad_norm": 0.20397111773490906, "learning_rate": 4.262657768750907e-05, "loss": 0.0323, "step": 15919 }, { "epoch": 1.7259323503902864, "grad_norm": 0.280813992023468, "learning_rate": 4.262295081967213e-05, "loss": 0.0174, "step": 15920 }, { "epoch": 1.726040763226366, "grad_norm": 0.1677662432193756, "learning_rate": 4.26193239518352e-05, "loss": 0.0111, "step": 15921 }, { "epoch": 1.726149176062446, "grad_norm": 0.15869000554084778, "learning_rate": 4.261569708399826e-05, "loss": 0.0118, "step": 15922 }, { "epoch": 1.7262575888985254, "grad_norm": 0.3618510961532593, "learning_rate": 4.261207021616133e-05, "loss": 0.0392, "step": 15923 }, { "epoch": 1.7263660017346054, "grad_norm": 0.383098840713501, "learning_rate": 4.260844334832439e-05, "loss": 0.0644, "step": 15924 }, { "epoch": 1.7264744145706852, "grad_norm": 0.18704821169376373, "learning_rate": 4.2604816480487455e-05, "loss": 0.0174, "step": 15925 }, { "epoch": 1.726582827406765, "grad_norm": 0.2483035773038864, "learning_rate": 4.260118961265052e-05, "loss": 0.0181, "step": 15926 }, { "epoch": 1.7266912402428447, "grad_norm": 0.7048120498657227, "learning_rate": 4.2597562744813577e-05, "loss": 0.046, "step": 15927 }, { "epoch": 1.7267996530789245, "grad_norm": 0.6714376211166382, "learning_rate": 4.259393587697664e-05, "loss": 0.0278, "step": 15928 }, { "epoch": 1.7269080659150045, "grad_norm": 0.11666588485240936, "learning_rate": 4.259030900913971e-05, "loss": 0.0068, "step": 15929 }, { "epoch": 1.727016478751084, "grad_norm": 0.2936636507511139, "learning_rate": 4.2586682141302776e-05, "loss": 0.0187, "step": 15930 }, { "epoch": 1.727124891587164, "grad_norm": 0.2555483877658844, "learning_rate": 4.258305527346584e-05, "loss": 0.0174, "step": 15931 }, { "epoch": 1.7272333044232437, "grad_norm": 0.1939476877450943, "learning_rate": 4.2579428405628904e-05, "loss": 0.0116, "step": 15932 }, { "epoch": 1.7273417172593235, "grad_norm": 0.5882146954536438, "learning_rate": 4.257580153779197e-05, "loss": 0.0572, "step": 15933 }, { "epoch": 1.7274501300954033, "grad_norm": 0.5158957839012146, "learning_rate": 4.2572174669955025e-05, "loss": 0.0395, "step": 15934 }, { "epoch": 1.727558542931483, "grad_norm": 0.2848737835884094, "learning_rate": 4.256854780211809e-05, "loss": 0.005, "step": 15935 }, { "epoch": 1.727666955767563, "grad_norm": 0.11572940647602081, "learning_rate": 4.2564920934281154e-05, "loss": 0.0034, "step": 15936 }, { "epoch": 1.7277753686036426, "grad_norm": 0.7826778888702393, "learning_rate": 4.256129406644422e-05, "loss": 0.0463, "step": 15937 }, { "epoch": 1.7278837814397225, "grad_norm": 0.38897404074668884, "learning_rate": 4.255766719860729e-05, "loss": 0.0163, "step": 15938 }, { "epoch": 1.7279921942758023, "grad_norm": 0.2625918686389923, "learning_rate": 4.255404033077035e-05, "loss": 0.0211, "step": 15939 }, { "epoch": 1.728100607111882, "grad_norm": 0.09846889972686768, "learning_rate": 4.255041346293342e-05, "loss": 0.0041, "step": 15940 }, { "epoch": 1.7282090199479618, "grad_norm": 0.4695550501346588, "learning_rate": 4.2546786595096474e-05, "loss": 0.0235, "step": 15941 }, { "epoch": 1.7283174327840416, "grad_norm": 0.14001767337322235, "learning_rate": 4.254315972725954e-05, "loss": 0.0115, "step": 15942 }, { "epoch": 1.7284258456201216, "grad_norm": 0.07795755565166473, "learning_rate": 4.25395328594226e-05, "loss": 0.0036, "step": 15943 }, { "epoch": 1.7285342584562011, "grad_norm": 0.39437761902809143, "learning_rate": 4.2535905991585666e-05, "loss": 0.0273, "step": 15944 }, { "epoch": 1.728642671292281, "grad_norm": 0.6050973534584045, "learning_rate": 4.253227912374873e-05, "loss": 0.0726, "step": 15945 }, { "epoch": 1.7287510841283606, "grad_norm": 0.33196431398391724, "learning_rate": 4.2528652255911795e-05, "loss": 0.0369, "step": 15946 }, { "epoch": 1.7288594969644406, "grad_norm": 1.031234622001648, "learning_rate": 4.2525025388074865e-05, "loss": 0.0286, "step": 15947 }, { "epoch": 1.7289679098005204, "grad_norm": 0.4533300995826721, "learning_rate": 4.252139852023792e-05, "loss": 0.0376, "step": 15948 }, { "epoch": 1.7290763226366002, "grad_norm": 0.226840078830719, "learning_rate": 4.251777165240099e-05, "loss": 0.014, "step": 15949 }, { "epoch": 1.72918473547268, "grad_norm": 1.3001091480255127, "learning_rate": 4.251414478456405e-05, "loss": 0.061, "step": 15950 }, { "epoch": 1.7292931483087597, "grad_norm": 0.3393835723400116, "learning_rate": 4.2510517916727115e-05, "loss": 0.021, "step": 15951 }, { "epoch": 1.7294015611448397, "grad_norm": 0.23163437843322754, "learning_rate": 4.250689104889018e-05, "loss": 0.0139, "step": 15952 }, { "epoch": 1.7295099739809192, "grad_norm": 0.395039826631546, "learning_rate": 4.250326418105324e-05, "loss": 0.0458, "step": 15953 }, { "epoch": 1.7296183868169992, "grad_norm": 0.08383955806493759, "learning_rate": 4.249963731321631e-05, "loss": 0.0073, "step": 15954 }, { "epoch": 1.729726799653079, "grad_norm": 0.327983558177948, "learning_rate": 4.249601044537938e-05, "loss": 0.0163, "step": 15955 }, { "epoch": 1.7298352124891587, "grad_norm": 0.7166184186935425, "learning_rate": 4.2492383577542436e-05, "loss": 0.0168, "step": 15956 }, { "epoch": 1.7299436253252385, "grad_norm": 0.13148190081119537, "learning_rate": 4.24887567097055e-05, "loss": 0.0137, "step": 15957 }, { "epoch": 1.7300520381613183, "grad_norm": 0.13669218122959137, "learning_rate": 4.2485129841868564e-05, "loss": 0.0089, "step": 15958 }, { "epoch": 1.7301604509973982, "grad_norm": 0.22900919616222382, "learning_rate": 4.248150297403163e-05, "loss": 0.0145, "step": 15959 }, { "epoch": 1.7302688638334778, "grad_norm": 0.26421284675598145, "learning_rate": 4.247787610619469e-05, "loss": 0.0346, "step": 15960 }, { "epoch": 1.7303772766695578, "grad_norm": 0.47441455721855164, "learning_rate": 4.2474249238357756e-05, "loss": 0.0271, "step": 15961 }, { "epoch": 1.7304856895056373, "grad_norm": 0.8942134976387024, "learning_rate": 4.247062237052082e-05, "loss": 0.0654, "step": 15962 }, { "epoch": 1.7305941023417173, "grad_norm": 0.30230191349983215, "learning_rate": 4.2466995502683884e-05, "loss": 0.014, "step": 15963 }, { "epoch": 1.730702515177797, "grad_norm": 0.5557267069816589, "learning_rate": 4.246336863484695e-05, "loss": 0.0413, "step": 15964 }, { "epoch": 1.7308109280138768, "grad_norm": 0.22538155317306519, "learning_rate": 4.245974176701001e-05, "loss": 0.0048, "step": 15965 }, { "epoch": 1.7309193408499568, "grad_norm": 0.18211036920547485, "learning_rate": 4.245611489917308e-05, "loss": 0.0154, "step": 15966 }, { "epoch": 1.7310277536860363, "grad_norm": 0.16252800822257996, "learning_rate": 4.245248803133614e-05, "loss": 0.0088, "step": 15967 }, { "epoch": 1.7311361665221163, "grad_norm": 0.1948821246623993, "learning_rate": 4.2448861163499205e-05, "loss": 0.0189, "step": 15968 }, { "epoch": 1.7312445793581959, "grad_norm": 0.2079496532678604, "learning_rate": 4.244523429566227e-05, "loss": 0.021, "step": 15969 }, { "epoch": 1.7313529921942759, "grad_norm": 0.38537997007369995, "learning_rate": 4.244160742782533e-05, "loss": 0.0594, "step": 15970 }, { "epoch": 1.7314614050303556, "grad_norm": 0.34203919768333435, "learning_rate": 4.24379805599884e-05, "loss": 0.0173, "step": 15971 }, { "epoch": 1.7315698178664354, "grad_norm": 0.3504927456378937, "learning_rate": 4.2434353692151455e-05, "loss": 0.0378, "step": 15972 }, { "epoch": 1.7316782307025151, "grad_norm": 0.37187573313713074, "learning_rate": 4.2430726824314525e-05, "loss": 0.03, "step": 15973 }, { "epoch": 1.731786643538595, "grad_norm": 0.4266057312488556, "learning_rate": 4.242709995647759e-05, "loss": 0.0538, "step": 15974 }, { "epoch": 1.731895056374675, "grad_norm": 0.405657023191452, "learning_rate": 4.2423473088640654e-05, "loss": 0.0295, "step": 15975 }, { "epoch": 1.7320034692107544, "grad_norm": 0.483864426612854, "learning_rate": 4.241984622080372e-05, "loss": 0.0193, "step": 15976 }, { "epoch": 1.7321118820468344, "grad_norm": 0.33799490332603455, "learning_rate": 4.241621935296678e-05, "loss": 0.0268, "step": 15977 }, { "epoch": 1.7322202948829142, "grad_norm": 0.10421840846538544, "learning_rate": 4.2412592485129846e-05, "loss": 0.0063, "step": 15978 }, { "epoch": 1.732328707718994, "grad_norm": 0.2398473471403122, "learning_rate": 4.24089656172929e-05, "loss": 0.0105, "step": 15979 }, { "epoch": 1.7324371205550737, "grad_norm": 0.14142827689647675, "learning_rate": 4.240533874945597e-05, "loss": 0.0139, "step": 15980 }, { "epoch": 1.7325455333911535, "grad_norm": 0.2622670531272888, "learning_rate": 4.240171188161904e-05, "loss": 0.0341, "step": 15981 }, { "epoch": 1.7326539462272335, "grad_norm": 0.21766825020313263, "learning_rate": 4.23980850137821e-05, "loss": 0.0089, "step": 15982 }, { "epoch": 1.732762359063313, "grad_norm": 0.4416615962982178, "learning_rate": 4.2394458145945166e-05, "loss": 0.0703, "step": 15983 }, { "epoch": 1.732870771899393, "grad_norm": 0.32661300897598267, "learning_rate": 4.239083127810823e-05, "loss": 0.0329, "step": 15984 }, { "epoch": 1.7329791847354725, "grad_norm": 0.157822847366333, "learning_rate": 4.2387204410271295e-05, "loss": 0.0065, "step": 15985 }, { "epoch": 1.7330875975715525, "grad_norm": 0.5353856086730957, "learning_rate": 4.238357754243435e-05, "loss": 0.0603, "step": 15986 }, { "epoch": 1.7331960104076323, "grad_norm": 0.41875210404396057, "learning_rate": 4.2379950674597416e-05, "loss": 0.0177, "step": 15987 }, { "epoch": 1.733304423243712, "grad_norm": 0.3433487117290497, "learning_rate": 4.237632380676048e-05, "loss": 0.0427, "step": 15988 }, { "epoch": 1.7334128360797918, "grad_norm": 0.20207583904266357, "learning_rate": 4.2372696938923544e-05, "loss": 0.035, "step": 15989 }, { "epoch": 1.7335212489158716, "grad_norm": 0.34286826848983765, "learning_rate": 4.2369070071086615e-05, "loss": 0.0215, "step": 15990 }, { "epoch": 1.7336296617519515, "grad_norm": 0.054625965654850006, "learning_rate": 4.236544320324968e-05, "loss": 0.0015, "step": 15991 }, { "epoch": 1.733738074588031, "grad_norm": 0.18880805373191833, "learning_rate": 4.2361816335412743e-05, "loss": 0.026, "step": 15992 }, { "epoch": 1.733846487424111, "grad_norm": 0.1869061142206192, "learning_rate": 4.23581894675758e-05, "loss": 0.0168, "step": 15993 }, { "epoch": 1.7339549002601908, "grad_norm": 0.6338751912117004, "learning_rate": 4.2354562599738865e-05, "loss": 0.0395, "step": 15994 }, { "epoch": 1.7340633130962706, "grad_norm": 0.42517977952957153, "learning_rate": 4.235093573190193e-05, "loss": 0.0349, "step": 15995 }, { "epoch": 1.7341717259323504, "grad_norm": 0.39099395275115967, "learning_rate": 4.234730886406499e-05, "loss": 0.0176, "step": 15996 }, { "epoch": 1.7342801387684301, "grad_norm": 0.36156320571899414, "learning_rate": 4.234368199622806e-05, "loss": 0.0758, "step": 15997 }, { "epoch": 1.73438855160451, "grad_norm": 0.35233697295188904, "learning_rate": 4.234005512839113e-05, "loss": 0.0351, "step": 15998 }, { "epoch": 1.7344969644405897, "grad_norm": 0.36859986186027527, "learning_rate": 4.233642826055419e-05, "loss": 0.0211, "step": 15999 }, { "epoch": 1.7346053772766696, "grad_norm": 0.12743690609931946, "learning_rate": 4.233280139271725e-05, "loss": 0.0143, "step": 16000 }, { "epoch": 1.7347137901127494, "grad_norm": 0.3133273720741272, "learning_rate": 4.2329174524880314e-05, "loss": 0.0369, "step": 16001 }, { "epoch": 1.7348222029488292, "grad_norm": 0.34983065724372864, "learning_rate": 4.232554765704338e-05, "loss": 0.0349, "step": 16002 }, { "epoch": 1.734930615784909, "grad_norm": 1.753492832183838, "learning_rate": 4.232192078920644e-05, "loss": 0.08, "step": 16003 }, { "epoch": 1.7350390286209887, "grad_norm": 0.20478057861328125, "learning_rate": 4.2318293921369506e-05, "loss": 0.02, "step": 16004 }, { "epoch": 1.7351474414570687, "grad_norm": 0.15841932594776154, "learning_rate": 4.231466705353257e-05, "loss": 0.0209, "step": 16005 }, { "epoch": 1.7352558542931482, "grad_norm": 0.25905436277389526, "learning_rate": 4.2311040185695634e-05, "loss": 0.0202, "step": 16006 }, { "epoch": 1.7353642671292282, "grad_norm": 0.21763770282268524, "learning_rate": 4.2307413317858705e-05, "loss": 0.027, "step": 16007 }, { "epoch": 1.7354726799653077, "grad_norm": 0.12826910614967346, "learning_rate": 4.230378645002176e-05, "loss": 0.0088, "step": 16008 }, { "epoch": 1.7355810928013877, "grad_norm": 0.1930946260690689, "learning_rate": 4.2300159582184826e-05, "loss": 0.0165, "step": 16009 }, { "epoch": 1.7356895056374675, "grad_norm": 0.2949894666671753, "learning_rate": 4.229653271434789e-05, "loss": 0.0317, "step": 16010 }, { "epoch": 1.7357979184735473, "grad_norm": 0.23502232134342194, "learning_rate": 4.2292905846510955e-05, "loss": 0.0197, "step": 16011 }, { "epoch": 1.735906331309627, "grad_norm": 0.3040682375431061, "learning_rate": 4.228927897867402e-05, "loss": 0.0477, "step": 16012 }, { "epoch": 1.7360147441457068, "grad_norm": 0.5177572965621948, "learning_rate": 4.228565211083708e-05, "loss": 0.0436, "step": 16013 }, { "epoch": 1.7361231569817868, "grad_norm": 0.3712317645549774, "learning_rate": 4.228202524300015e-05, "loss": 0.0328, "step": 16014 }, { "epoch": 1.7362315698178663, "grad_norm": 0.023364823311567307, "learning_rate": 4.227839837516321e-05, "loss": 0.0025, "step": 16015 }, { "epoch": 1.7363399826539463, "grad_norm": 0.6471971869468689, "learning_rate": 4.2274771507326275e-05, "loss": 0.0553, "step": 16016 }, { "epoch": 1.736448395490026, "grad_norm": 0.38022947311401367, "learning_rate": 4.227114463948934e-05, "loss": 0.023, "step": 16017 }, { "epoch": 1.7365568083261058, "grad_norm": 0.3460233211517334, "learning_rate": 4.22675177716524e-05, "loss": 0.04, "step": 16018 }, { "epoch": 1.7366652211621856, "grad_norm": 0.11571983247995377, "learning_rate": 4.226389090381547e-05, "loss": 0.0277, "step": 16019 }, { "epoch": 1.7367736339982653, "grad_norm": 0.1220792680978775, "learning_rate": 4.226026403597853e-05, "loss": 0.0059, "step": 16020 }, { "epoch": 1.7368820468343453, "grad_norm": 0.4068267345428467, "learning_rate": 4.2256637168141596e-05, "loss": 0.0421, "step": 16021 }, { "epoch": 1.7369904596704249, "grad_norm": 0.16332054138183594, "learning_rate": 4.225301030030466e-05, "loss": 0.0098, "step": 16022 }, { "epoch": 1.7370988725065049, "grad_norm": 0.3646336495876312, "learning_rate": 4.2249383432467724e-05, "loss": 0.011, "step": 16023 }, { "epoch": 1.7372072853425846, "grad_norm": 0.29552021622657776, "learning_rate": 4.224575656463079e-05, "loss": 0.0301, "step": 16024 }, { "epoch": 1.7373156981786644, "grad_norm": 0.23083479702472687, "learning_rate": 4.224212969679385e-05, "loss": 0.0249, "step": 16025 }, { "epoch": 1.7374241110147441, "grad_norm": 0.36902329325675964, "learning_rate": 4.2238502828956916e-05, "loss": 0.0499, "step": 16026 }, { "epoch": 1.737532523850824, "grad_norm": 0.05404912680387497, "learning_rate": 4.223487596111998e-05, "loss": 0.0034, "step": 16027 }, { "epoch": 1.737640936686904, "grad_norm": 0.23139995336532593, "learning_rate": 4.2231249093283044e-05, "loss": 0.0297, "step": 16028 }, { "epoch": 1.7377493495229834, "grad_norm": 0.4355337917804718, "learning_rate": 4.222762222544611e-05, "loss": 0.0567, "step": 16029 }, { "epoch": 1.7378577623590634, "grad_norm": 0.11225979775190353, "learning_rate": 4.222399535760917e-05, "loss": 0.0181, "step": 16030 }, { "epoch": 1.737966175195143, "grad_norm": 0.29034188389778137, "learning_rate": 4.222036848977223e-05, "loss": 0.0294, "step": 16031 }, { "epoch": 1.738074588031223, "grad_norm": 0.2171073853969574, "learning_rate": 4.2216741621935294e-05, "loss": 0.0275, "step": 16032 }, { "epoch": 1.7381830008673027, "grad_norm": 0.09455084800720215, "learning_rate": 4.2213114754098365e-05, "loss": 0.0149, "step": 16033 }, { "epoch": 1.7382914137033825, "grad_norm": 0.07444912940263748, "learning_rate": 4.220948788626143e-05, "loss": 0.0101, "step": 16034 }, { "epoch": 1.7383998265394622, "grad_norm": 1.3835816383361816, "learning_rate": 4.220586101842449e-05, "loss": 0.0685, "step": 16035 }, { "epoch": 1.738508239375542, "grad_norm": 0.07357506453990936, "learning_rate": 4.220223415058756e-05, "loss": 0.0076, "step": 16036 }, { "epoch": 1.738616652211622, "grad_norm": 0.015779556706547737, "learning_rate": 4.219860728275062e-05, "loss": 0.0006, "step": 16037 }, { "epoch": 1.7387250650477015, "grad_norm": 0.23484519124031067, "learning_rate": 4.219498041491368e-05, "loss": 0.0184, "step": 16038 }, { "epoch": 1.7388334778837815, "grad_norm": 0.33731845021247864, "learning_rate": 4.219135354707674e-05, "loss": 0.0185, "step": 16039 }, { "epoch": 1.7389418907198613, "grad_norm": 0.5856398940086365, "learning_rate": 4.218772667923981e-05, "loss": 0.0319, "step": 16040 }, { "epoch": 1.739050303555941, "grad_norm": 0.06725893169641495, "learning_rate": 4.218409981140287e-05, "loss": 0.0034, "step": 16041 }, { "epoch": 1.7391587163920208, "grad_norm": 0.29635709524154663, "learning_rate": 4.218047294356594e-05, "loss": 0.028, "step": 16042 }, { "epoch": 1.7392671292281006, "grad_norm": 0.2115493267774582, "learning_rate": 4.2176846075729006e-05, "loss": 0.0181, "step": 16043 }, { "epoch": 1.7393755420641805, "grad_norm": 0.14490686357021332, "learning_rate": 4.217321920789207e-05, "loss": 0.012, "step": 16044 }, { "epoch": 1.73948395490026, "grad_norm": 0.19518892467021942, "learning_rate": 4.216959234005513e-05, "loss": 0.0224, "step": 16045 }, { "epoch": 1.73959236773634, "grad_norm": 0.29997095465660095, "learning_rate": 4.216596547221819e-05, "loss": 0.0186, "step": 16046 }, { "epoch": 1.7397007805724196, "grad_norm": 0.27907824516296387, "learning_rate": 4.2162338604381256e-05, "loss": 0.0246, "step": 16047 }, { "epoch": 1.7398091934084996, "grad_norm": 0.5515829920768738, "learning_rate": 4.215871173654432e-05, "loss": 0.0592, "step": 16048 }, { "epoch": 1.7399176062445794, "grad_norm": 0.6836092472076416, "learning_rate": 4.2155084868707384e-05, "loss": 0.0343, "step": 16049 }, { "epoch": 1.7400260190806591, "grad_norm": 0.4797031581401825, "learning_rate": 4.2151458000870455e-05, "loss": 0.0289, "step": 16050 }, { "epoch": 1.7401344319167389, "grad_norm": 0.416459321975708, "learning_rate": 4.214783113303352e-05, "loss": 0.0241, "step": 16051 }, { "epoch": 1.7402428447528187, "grad_norm": 0.25807976722717285, "learning_rate": 4.214420426519658e-05, "loss": 0.0214, "step": 16052 }, { "epoch": 1.7403512575888986, "grad_norm": 0.07558754086494446, "learning_rate": 4.214057739735964e-05, "loss": 0.0081, "step": 16053 }, { "epoch": 1.7404596704249782, "grad_norm": 0.14326325058937073, "learning_rate": 4.2136950529522704e-05, "loss": 0.0099, "step": 16054 }, { "epoch": 1.7405680832610582, "grad_norm": 0.2676672339439392, "learning_rate": 4.213332366168577e-05, "loss": 0.0104, "step": 16055 }, { "epoch": 1.740676496097138, "grad_norm": 0.1185244619846344, "learning_rate": 4.212969679384883e-05, "loss": 0.0174, "step": 16056 }, { "epoch": 1.7407849089332177, "grad_norm": 0.1812705397605896, "learning_rate": 4.21260699260119e-05, "loss": 0.0099, "step": 16057 }, { "epoch": 1.7408933217692975, "grad_norm": 0.1841735988855362, "learning_rate": 4.212244305817496e-05, "loss": 0.0116, "step": 16058 }, { "epoch": 1.7410017346053772, "grad_norm": 0.767463743686676, "learning_rate": 4.211881619033803e-05, "loss": 0.049, "step": 16059 }, { "epoch": 1.7411101474414572, "grad_norm": 0.2559269964694977, "learning_rate": 4.211518932250109e-05, "loss": 0.03, "step": 16060 }, { "epoch": 1.7412185602775367, "grad_norm": 0.3396063446998596, "learning_rate": 4.211156245466415e-05, "loss": 0.0603, "step": 16061 }, { "epoch": 1.7413269731136167, "grad_norm": 0.13041998445987701, "learning_rate": 4.210793558682722e-05, "loss": 0.0113, "step": 16062 }, { "epoch": 1.7414353859496965, "grad_norm": 0.29791027307510376, "learning_rate": 4.210430871899028e-05, "loss": 0.0303, "step": 16063 }, { "epoch": 1.7415437987857763, "grad_norm": 0.19853660464286804, "learning_rate": 4.2100681851153345e-05, "loss": 0.0097, "step": 16064 }, { "epoch": 1.741652211621856, "grad_norm": 0.29823699593544006, "learning_rate": 4.209705498331641e-05, "loss": 0.0297, "step": 16065 }, { "epoch": 1.7417606244579358, "grad_norm": 0.17280124127864838, "learning_rate": 4.2093428115479474e-05, "loss": 0.0142, "step": 16066 }, { "epoch": 1.7418690372940158, "grad_norm": 0.0760701596736908, "learning_rate": 4.208980124764254e-05, "loss": 0.0119, "step": 16067 }, { "epoch": 1.7419774501300953, "grad_norm": 0.19648601114749908, "learning_rate": 4.20861743798056e-05, "loss": 0.0152, "step": 16068 }, { "epoch": 1.7420858629661753, "grad_norm": 0.32917848229408264, "learning_rate": 4.2082547511968666e-05, "loss": 0.0244, "step": 16069 }, { "epoch": 1.7421942758022548, "grad_norm": 0.5739530920982361, "learning_rate": 4.207892064413173e-05, "loss": 0.0522, "step": 16070 }, { "epoch": 1.7423026886383348, "grad_norm": 0.17873847484588623, "learning_rate": 4.2075293776294794e-05, "loss": 0.0257, "step": 16071 }, { "epoch": 1.7424111014744146, "grad_norm": 0.2951326072216034, "learning_rate": 4.207166690845786e-05, "loss": 0.0157, "step": 16072 }, { "epoch": 1.7425195143104943, "grad_norm": 0.17087040841579437, "learning_rate": 4.206804004062092e-05, "loss": 0.018, "step": 16073 }, { "epoch": 1.742627927146574, "grad_norm": 0.3196716904640198, "learning_rate": 4.2064413172783986e-05, "loss": 0.0183, "step": 16074 }, { "epoch": 1.7427363399826539, "grad_norm": 0.39413291215896606, "learning_rate": 4.206078630494705e-05, "loss": 0.0326, "step": 16075 }, { "epoch": 1.7428447528187339, "grad_norm": 0.30932220816612244, "learning_rate": 4.2057159437110115e-05, "loss": 0.038, "step": 16076 }, { "epoch": 1.7429531656548134, "grad_norm": 0.07203087210655212, "learning_rate": 4.205353256927318e-05, "loss": 0.0043, "step": 16077 }, { "epoch": 1.7430615784908934, "grad_norm": 0.3061542809009552, "learning_rate": 4.204990570143624e-05, "loss": 0.0316, "step": 16078 }, { "epoch": 1.7431699913269731, "grad_norm": 0.3489949703216553, "learning_rate": 4.204627883359931e-05, "loss": 0.0203, "step": 16079 }, { "epoch": 1.743278404163053, "grad_norm": 0.13049952685832977, "learning_rate": 4.204265196576237e-05, "loss": 0.0091, "step": 16080 }, { "epoch": 1.7433868169991327, "grad_norm": 0.2760947346687317, "learning_rate": 4.2039025097925435e-05, "loss": 0.0177, "step": 16081 }, { "epoch": 1.7434952298352124, "grad_norm": 0.24717912077903748, "learning_rate": 4.20353982300885e-05, "loss": 0.0289, "step": 16082 }, { "epoch": 1.7436036426712924, "grad_norm": 0.3375340700149536, "learning_rate": 4.203177136225156e-05, "loss": 0.0427, "step": 16083 }, { "epoch": 1.743712055507372, "grad_norm": 0.33579936623573303, "learning_rate": 4.202814449441462e-05, "loss": 0.0344, "step": 16084 }, { "epoch": 1.743820468343452, "grad_norm": 0.028990574181079865, "learning_rate": 4.202451762657769e-05, "loss": 0.0013, "step": 16085 }, { "epoch": 1.7439288811795317, "grad_norm": 0.41288986802101135, "learning_rate": 4.2020890758740756e-05, "loss": 0.0139, "step": 16086 }, { "epoch": 1.7440372940156115, "grad_norm": 1.1238253116607666, "learning_rate": 4.201726389090382e-05, "loss": 0.0169, "step": 16087 }, { "epoch": 1.7441457068516912, "grad_norm": 0.33037039637565613, "learning_rate": 4.2013637023066884e-05, "loss": 0.0309, "step": 16088 }, { "epoch": 1.744254119687771, "grad_norm": 0.2134125828742981, "learning_rate": 4.201001015522995e-05, "loss": 0.013, "step": 16089 }, { "epoch": 1.744362532523851, "grad_norm": 0.33970212936401367, "learning_rate": 4.2006383287393005e-05, "loss": 0.0123, "step": 16090 }, { "epoch": 1.7444709453599305, "grad_norm": 0.4138076901435852, "learning_rate": 4.200275641955607e-05, "loss": 0.0293, "step": 16091 }, { "epoch": 1.7445793581960105, "grad_norm": 0.23817910254001617, "learning_rate": 4.1999129551719134e-05, "loss": 0.0131, "step": 16092 }, { "epoch": 1.74468777103209, "grad_norm": 0.4010990262031555, "learning_rate": 4.1995502683882204e-05, "loss": 0.0328, "step": 16093 }, { "epoch": 1.74479618386817, "grad_norm": 0.07639500498771667, "learning_rate": 4.199187581604527e-05, "loss": 0.0081, "step": 16094 }, { "epoch": 1.7449045967042498, "grad_norm": 0.30559056997299194, "learning_rate": 4.198824894820833e-05, "loss": 0.013, "step": 16095 }, { "epoch": 1.7450130095403296, "grad_norm": 0.4220438003540039, "learning_rate": 4.19846220803714e-05, "loss": 0.0298, "step": 16096 }, { "epoch": 1.7451214223764093, "grad_norm": 0.27426275610923767, "learning_rate": 4.1980995212534454e-05, "loss": 0.0166, "step": 16097 }, { "epoch": 1.745229835212489, "grad_norm": 0.2941407561302185, "learning_rate": 4.197736834469752e-05, "loss": 0.022, "step": 16098 }, { "epoch": 1.745338248048569, "grad_norm": 0.26771143078804016, "learning_rate": 4.197374147686058e-05, "loss": 0.0171, "step": 16099 }, { "epoch": 1.7454466608846486, "grad_norm": 0.13297997415065765, "learning_rate": 4.1970114609023646e-05, "loss": 0.0166, "step": 16100 }, { "epoch": 1.7455550737207286, "grad_norm": 0.12370701134204865, "learning_rate": 4.196648774118671e-05, "loss": 0.0028, "step": 16101 }, { "epoch": 1.7456634865568084, "grad_norm": 0.13478025794029236, "learning_rate": 4.196286087334978e-05, "loss": 0.0053, "step": 16102 }, { "epoch": 1.7457718993928881, "grad_norm": 0.1892423927783966, "learning_rate": 4.1959234005512846e-05, "loss": 0.0043, "step": 16103 }, { "epoch": 1.745880312228968, "grad_norm": 0.19249393045902252, "learning_rate": 4.195560713767591e-05, "loss": 0.0157, "step": 16104 }, { "epoch": 1.7459887250650477, "grad_norm": 0.26756978034973145, "learning_rate": 4.195198026983897e-05, "loss": 0.018, "step": 16105 }, { "epoch": 1.7460971379011276, "grad_norm": 0.23566536605358124, "learning_rate": 4.194835340200203e-05, "loss": 0.0174, "step": 16106 }, { "epoch": 1.7462055507372072, "grad_norm": 0.03928259760141373, "learning_rate": 4.1944726534165095e-05, "loss": 0.0015, "step": 16107 }, { "epoch": 1.7463139635732872, "grad_norm": 0.19946347177028656, "learning_rate": 4.194109966632816e-05, "loss": 0.0191, "step": 16108 }, { "epoch": 1.7464223764093667, "grad_norm": 0.16794995963573456, "learning_rate": 4.1937472798491223e-05, "loss": 0.0029, "step": 16109 }, { "epoch": 1.7465307892454467, "grad_norm": 0.3445451557636261, "learning_rate": 4.193384593065429e-05, "loss": 0.0324, "step": 16110 }, { "epoch": 1.7466392020815265, "grad_norm": 0.5847607851028442, "learning_rate": 4.193021906281736e-05, "loss": 0.0196, "step": 16111 }, { "epoch": 1.7467476149176062, "grad_norm": 0.4534061551094055, "learning_rate": 4.1926592194980416e-05, "loss": 0.0633, "step": 16112 }, { "epoch": 1.7468560277536862, "grad_norm": 0.36209067702293396, "learning_rate": 4.192296532714348e-05, "loss": 0.0507, "step": 16113 }, { "epoch": 1.7469644405897657, "grad_norm": 0.30970969796180725, "learning_rate": 4.1919338459306544e-05, "loss": 0.0218, "step": 16114 }, { "epoch": 1.7470728534258457, "grad_norm": 0.6438181400299072, "learning_rate": 4.191571159146961e-05, "loss": 0.0631, "step": 16115 }, { "epoch": 1.7471812662619253, "grad_norm": 0.222379669547081, "learning_rate": 4.191208472363267e-05, "loss": 0.016, "step": 16116 }, { "epoch": 1.7472896790980053, "grad_norm": 0.1499793380498886, "learning_rate": 4.1908457855795736e-05, "loss": 0.0027, "step": 16117 }, { "epoch": 1.747398091934085, "grad_norm": 0.42498642206192017, "learning_rate": 4.19048309879588e-05, "loss": 0.0465, "step": 16118 }, { "epoch": 1.7475065047701648, "grad_norm": 0.10176588594913483, "learning_rate": 4.1901204120121864e-05, "loss": 0.0079, "step": 16119 }, { "epoch": 1.7476149176062445, "grad_norm": 0.3493231236934662, "learning_rate": 4.189757725228493e-05, "loss": 0.0199, "step": 16120 }, { "epoch": 1.7477233304423243, "grad_norm": 0.027354050427675247, "learning_rate": 4.189395038444799e-05, "loss": 0.0008, "step": 16121 }, { "epoch": 1.7478317432784043, "grad_norm": 0.35213375091552734, "learning_rate": 4.189032351661106e-05, "loss": 0.0206, "step": 16122 }, { "epoch": 1.7479401561144838, "grad_norm": 0.21032586693763733, "learning_rate": 4.188669664877412e-05, "loss": 0.0168, "step": 16123 }, { "epoch": 1.7480485689505638, "grad_norm": 0.4517596364021301, "learning_rate": 4.1883069780937185e-05, "loss": 0.0273, "step": 16124 }, { "epoch": 1.7481569817866436, "grad_norm": 0.15279343724250793, "learning_rate": 4.187944291310025e-05, "loss": 0.0087, "step": 16125 }, { "epoch": 1.7482653946227233, "grad_norm": 0.4910203218460083, "learning_rate": 4.187581604526331e-05, "loss": 0.0345, "step": 16126 }, { "epoch": 1.748373807458803, "grad_norm": 0.4726828932762146, "learning_rate": 4.187218917742638e-05, "loss": 0.0685, "step": 16127 }, { "epoch": 1.7484822202948829, "grad_norm": 0.524925172328949, "learning_rate": 4.186856230958944e-05, "loss": 0.0165, "step": 16128 }, { "epoch": 1.7485906331309629, "grad_norm": 0.4042768180370331, "learning_rate": 4.1864935441752505e-05, "loss": 0.0248, "step": 16129 }, { "epoch": 1.7486990459670424, "grad_norm": 0.32944217324256897, "learning_rate": 4.186130857391557e-05, "loss": 0.0152, "step": 16130 }, { "epoch": 1.7488074588031224, "grad_norm": 0.4747861921787262, "learning_rate": 4.1857681706078634e-05, "loss": 0.039, "step": 16131 }, { "epoch": 1.748915871639202, "grad_norm": 0.24762922525405884, "learning_rate": 4.18540548382417e-05, "loss": 0.0159, "step": 16132 }, { "epoch": 1.749024284475282, "grad_norm": 0.5224037170410156, "learning_rate": 4.185042797040476e-05, "loss": 0.0585, "step": 16133 }, { "epoch": 1.7491326973113617, "grad_norm": 0.10484355688095093, "learning_rate": 4.1846801102567826e-05, "loss": 0.0071, "step": 16134 }, { "epoch": 1.7492411101474414, "grad_norm": 0.5501604080200195, "learning_rate": 4.184317423473088e-05, "loss": 0.0359, "step": 16135 }, { "epoch": 1.7493495229835212, "grad_norm": 0.27228349447250366, "learning_rate": 4.183954736689395e-05, "loss": 0.0161, "step": 16136 }, { "epoch": 1.749457935819601, "grad_norm": 0.550237774848938, "learning_rate": 4.183592049905702e-05, "loss": 0.0293, "step": 16137 }, { "epoch": 1.749566348655681, "grad_norm": 0.35360005497932434, "learning_rate": 4.183229363122008e-05, "loss": 0.0257, "step": 16138 }, { "epoch": 1.7496747614917605, "grad_norm": 0.6067412495613098, "learning_rate": 4.1828666763383147e-05, "loss": 0.0352, "step": 16139 }, { "epoch": 1.7497831743278405, "grad_norm": 0.11551126092672348, "learning_rate": 4.182503989554621e-05, "loss": 0.0065, "step": 16140 }, { "epoch": 1.7498915871639202, "grad_norm": 0.21591848134994507, "learning_rate": 4.1821413027709275e-05, "loss": 0.0252, "step": 16141 }, { "epoch": 1.75, "grad_norm": 0.7192476391792297, "learning_rate": 4.181778615987233e-05, "loss": 0.0258, "step": 16142 }, { "epoch": 1.7501084128360798, "grad_norm": 0.07260879129171371, "learning_rate": 4.1814159292035396e-05, "loss": 0.0116, "step": 16143 }, { "epoch": 1.7502168256721595, "grad_norm": 0.15879790484905243, "learning_rate": 4.181053242419846e-05, "loss": 0.0098, "step": 16144 }, { "epoch": 1.7503252385082395, "grad_norm": 0.02593669667840004, "learning_rate": 4.180690555636153e-05, "loss": 0.0016, "step": 16145 }, { "epoch": 1.750433651344319, "grad_norm": 0.3641025722026825, "learning_rate": 4.1803278688524595e-05, "loss": 0.032, "step": 16146 }, { "epoch": 1.750542064180399, "grad_norm": 0.17004601657390594, "learning_rate": 4.179965182068766e-05, "loss": 0.0151, "step": 16147 }, { "epoch": 1.7506504770164788, "grad_norm": 0.41106995940208435, "learning_rate": 4.1796024952850723e-05, "loss": 0.0432, "step": 16148 }, { "epoch": 1.7507588898525586, "grad_norm": 0.12569300830364227, "learning_rate": 4.179239808501378e-05, "loss": 0.0098, "step": 16149 }, { "epoch": 1.7508673026886383, "grad_norm": 0.18153639137744904, "learning_rate": 4.1788771217176845e-05, "loss": 0.016, "step": 16150 }, { "epoch": 1.750975715524718, "grad_norm": 1.829169750213623, "learning_rate": 4.178514434933991e-05, "loss": 0.0492, "step": 16151 }, { "epoch": 1.751084128360798, "grad_norm": 0.6389585137367249, "learning_rate": 4.178151748150297e-05, "loss": 0.0462, "step": 16152 }, { "epoch": 1.7511925411968776, "grad_norm": 0.232137069106102, "learning_rate": 4.177789061366604e-05, "loss": 0.0092, "step": 16153 }, { "epoch": 1.7513009540329576, "grad_norm": 0.2197604477405548, "learning_rate": 4.177426374582911e-05, "loss": 0.0221, "step": 16154 }, { "epoch": 1.7514093668690371, "grad_norm": 0.3506351411342621, "learning_rate": 4.177063687799217e-05, "loss": 0.025, "step": 16155 }, { "epoch": 1.7515177797051171, "grad_norm": 0.8386022448539734, "learning_rate": 4.1767010010155236e-05, "loss": 0.0451, "step": 16156 }, { "epoch": 1.751626192541197, "grad_norm": 0.21591147780418396, "learning_rate": 4.1763383142318294e-05, "loss": 0.0266, "step": 16157 }, { "epoch": 1.7517346053772767, "grad_norm": 0.28596898913383484, "learning_rate": 4.175975627448136e-05, "loss": 0.0345, "step": 16158 }, { "epoch": 1.7518430182133564, "grad_norm": 0.047239918261766434, "learning_rate": 4.175612940664442e-05, "loss": 0.007, "step": 16159 }, { "epoch": 1.7519514310494362, "grad_norm": 0.2846493721008301, "learning_rate": 4.1752502538807486e-05, "loss": 0.0213, "step": 16160 }, { "epoch": 1.7520598438855162, "grad_norm": 0.15004564821720123, "learning_rate": 4.174887567097055e-05, "loss": 0.0065, "step": 16161 }, { "epoch": 1.7521682567215957, "grad_norm": 0.21794071793556213, "learning_rate": 4.1745248803133614e-05, "loss": 0.0129, "step": 16162 }, { "epoch": 1.7522766695576757, "grad_norm": 0.18368279933929443, "learning_rate": 4.1741621935296685e-05, "loss": 0.0081, "step": 16163 }, { "epoch": 1.7523850823937555, "grad_norm": 0.1852950006723404, "learning_rate": 4.173799506745974e-05, "loss": 0.0096, "step": 16164 }, { "epoch": 1.7524934952298352, "grad_norm": 0.08589863777160645, "learning_rate": 4.1734368199622806e-05, "loss": 0.0041, "step": 16165 }, { "epoch": 1.752601908065915, "grad_norm": 0.06173985078930855, "learning_rate": 4.173074133178587e-05, "loss": 0.004, "step": 16166 }, { "epoch": 1.7527103209019947, "grad_norm": 0.06700865179300308, "learning_rate": 4.1727114463948935e-05, "loss": 0.0052, "step": 16167 }, { "epoch": 1.7528187337380747, "grad_norm": 0.1514320969581604, "learning_rate": 4.1723487596112e-05, "loss": 0.0117, "step": 16168 }, { "epoch": 1.7529271465741543, "grad_norm": 0.2249506115913391, "learning_rate": 4.171986072827506e-05, "loss": 0.0236, "step": 16169 }, { "epoch": 1.7530355594102343, "grad_norm": 0.46836355328559875, "learning_rate": 4.171623386043813e-05, "loss": 0.0301, "step": 16170 }, { "epoch": 1.7531439722463138, "grad_norm": 0.22536148130893707, "learning_rate": 4.171260699260119e-05, "loss": 0.0309, "step": 16171 }, { "epoch": 1.7532523850823938, "grad_norm": 0.2206984907388687, "learning_rate": 4.1708980124764255e-05, "loss": 0.0213, "step": 16172 }, { "epoch": 1.7533607979184735, "grad_norm": 1.078031301498413, "learning_rate": 4.170535325692732e-05, "loss": 0.0182, "step": 16173 }, { "epoch": 1.7534692107545533, "grad_norm": 0.5510970950126648, "learning_rate": 4.1701726389090383e-05, "loss": 0.027, "step": 16174 }, { "epoch": 1.7535776235906333, "grad_norm": 0.2599853277206421, "learning_rate": 4.169809952125345e-05, "loss": 0.0272, "step": 16175 }, { "epoch": 1.7536860364267128, "grad_norm": 0.21366094052791595, "learning_rate": 4.169447265341651e-05, "loss": 0.0259, "step": 16176 }, { "epoch": 1.7537944492627928, "grad_norm": 0.4193532466888428, "learning_rate": 4.1690845785579576e-05, "loss": 0.026, "step": 16177 }, { "epoch": 1.7539028620988724, "grad_norm": 0.21021012961864471, "learning_rate": 4.168721891774264e-05, "loss": 0.0125, "step": 16178 }, { "epoch": 1.7540112749349523, "grad_norm": 0.05648142471909523, "learning_rate": 4.1683592049905704e-05, "loss": 0.0032, "step": 16179 }, { "epoch": 1.754119687771032, "grad_norm": 0.6278923749923706, "learning_rate": 4.167996518206877e-05, "loss": 0.0355, "step": 16180 }, { "epoch": 1.7542281006071119, "grad_norm": 0.15901747345924377, "learning_rate": 4.167633831423183e-05, "loss": 0.0072, "step": 16181 }, { "epoch": 1.7543365134431916, "grad_norm": 0.4212300777435303, "learning_rate": 4.1672711446394896e-05, "loss": 0.0501, "step": 16182 }, { "epoch": 1.7544449262792714, "grad_norm": 0.3342169225215912, "learning_rate": 4.166908457855796e-05, "loss": 0.0133, "step": 16183 }, { "epoch": 1.7545533391153514, "grad_norm": 0.36304083466529846, "learning_rate": 4.1665457710721024e-05, "loss": 0.0237, "step": 16184 }, { "epoch": 1.754661751951431, "grad_norm": 0.15548327565193176, "learning_rate": 4.166183084288409e-05, "loss": 0.0132, "step": 16185 }, { "epoch": 1.754770164787511, "grad_norm": 0.10990556329488754, "learning_rate": 4.165820397504715e-05, "loss": 0.0055, "step": 16186 }, { "epoch": 1.7548785776235907, "grad_norm": 0.27303165197372437, "learning_rate": 4.165457710721021e-05, "loss": 0.0221, "step": 16187 }, { "epoch": 1.7549869904596704, "grad_norm": 0.062440305948257446, "learning_rate": 4.165095023937328e-05, "loss": 0.0048, "step": 16188 }, { "epoch": 1.7550954032957502, "grad_norm": 0.15276184678077698, "learning_rate": 4.1647323371536345e-05, "loss": 0.0326, "step": 16189 }, { "epoch": 1.75520381613183, "grad_norm": 0.2746905982494354, "learning_rate": 4.164369650369941e-05, "loss": 0.0166, "step": 16190 }, { "epoch": 1.75531222896791, "grad_norm": 0.15961633622646332, "learning_rate": 4.164006963586247e-05, "loss": 0.0228, "step": 16191 }, { "epoch": 1.7554206418039895, "grad_norm": 0.29070234298706055, "learning_rate": 4.163644276802554e-05, "loss": 0.0197, "step": 16192 }, { "epoch": 1.7555290546400695, "grad_norm": 0.44810888171195984, "learning_rate": 4.16328159001886e-05, "loss": 0.0299, "step": 16193 }, { "epoch": 1.755637467476149, "grad_norm": 0.26389050483703613, "learning_rate": 4.162918903235166e-05, "loss": 0.0237, "step": 16194 }, { "epoch": 1.755745880312229, "grad_norm": 0.2519565224647522, "learning_rate": 4.162556216451472e-05, "loss": 0.0091, "step": 16195 }, { "epoch": 1.7558542931483088, "grad_norm": 0.2405710071325302, "learning_rate": 4.162193529667779e-05, "loss": 0.036, "step": 16196 }, { "epoch": 1.7559627059843885, "grad_norm": 0.08533179759979248, "learning_rate": 4.161830842884086e-05, "loss": 0.0042, "step": 16197 }, { "epoch": 1.7560711188204683, "grad_norm": 0.05871244892477989, "learning_rate": 4.161468156100392e-05, "loss": 0.0021, "step": 16198 }, { "epoch": 1.756179531656548, "grad_norm": 0.4148656725883484, "learning_rate": 4.1611054693166986e-05, "loss": 0.0183, "step": 16199 }, { "epoch": 1.756287944492628, "grad_norm": 0.3944477438926697, "learning_rate": 4.160742782533005e-05, "loss": 0.0213, "step": 16200 }, { "epoch": 1.7563963573287076, "grad_norm": 0.6155033707618713, "learning_rate": 4.1603800957493114e-05, "loss": 0.0701, "step": 16201 }, { "epoch": 1.7565047701647876, "grad_norm": 0.10541139543056488, "learning_rate": 4.160017408965617e-05, "loss": 0.0057, "step": 16202 }, { "epoch": 1.7566131830008673, "grad_norm": 0.020541995763778687, "learning_rate": 4.1596547221819236e-05, "loss": 0.0008, "step": 16203 }, { "epoch": 1.756721595836947, "grad_norm": 0.28878289461135864, "learning_rate": 4.15929203539823e-05, "loss": 0.0178, "step": 16204 }, { "epoch": 1.7568300086730269, "grad_norm": 0.08173111826181412, "learning_rate": 4.1589293486145364e-05, "loss": 0.0076, "step": 16205 }, { "epoch": 1.7569384215091066, "grad_norm": 0.3160014748573303, "learning_rate": 4.1585666618308435e-05, "loss": 0.0089, "step": 16206 }, { "epoch": 1.7570468343451866, "grad_norm": 0.11390196532011032, "learning_rate": 4.15820397504715e-05, "loss": 0.0074, "step": 16207 }, { "epoch": 1.7571552471812661, "grad_norm": 0.1432189792394638, "learning_rate": 4.157841288263456e-05, "loss": 0.0131, "step": 16208 }, { "epoch": 1.7572636600173461, "grad_norm": 0.36285287141799927, "learning_rate": 4.157478601479762e-05, "loss": 0.0279, "step": 16209 }, { "epoch": 1.757372072853426, "grad_norm": 0.13620498776435852, "learning_rate": 4.1571159146960684e-05, "loss": 0.0058, "step": 16210 }, { "epoch": 1.7574804856895057, "grad_norm": 0.5107731223106384, "learning_rate": 4.156753227912375e-05, "loss": 0.0394, "step": 16211 }, { "epoch": 1.7575888985255854, "grad_norm": 0.9414551854133606, "learning_rate": 4.156390541128681e-05, "loss": 0.04, "step": 16212 }, { "epoch": 1.7576973113616652, "grad_norm": 0.4265765845775604, "learning_rate": 4.156027854344988e-05, "loss": 0.0334, "step": 16213 }, { "epoch": 1.7578057241977452, "grad_norm": 0.38053908944129944, "learning_rate": 4.155665167561295e-05, "loss": 0.0178, "step": 16214 }, { "epoch": 1.7579141370338247, "grad_norm": 0.21765346825122833, "learning_rate": 4.155302480777601e-05, "loss": 0.016, "step": 16215 }, { "epoch": 1.7580225498699047, "grad_norm": 0.2740086317062378, "learning_rate": 4.154939793993907e-05, "loss": 0.0192, "step": 16216 }, { "epoch": 1.7581309627059842, "grad_norm": 0.09713181853294373, "learning_rate": 4.154577107210213e-05, "loss": 0.0049, "step": 16217 }, { "epoch": 1.7582393755420642, "grad_norm": 0.38387587666511536, "learning_rate": 4.15421442042652e-05, "loss": 0.0245, "step": 16218 }, { "epoch": 1.758347788378144, "grad_norm": 0.31615379452705383, "learning_rate": 4.153851733642826e-05, "loss": 0.0229, "step": 16219 }, { "epoch": 1.7584562012142237, "grad_norm": 0.5244250893592834, "learning_rate": 4.1534890468591326e-05, "loss": 0.0278, "step": 16220 }, { "epoch": 1.7585646140503035, "grad_norm": 0.339286744594574, "learning_rate": 4.153126360075439e-05, "loss": 0.0173, "step": 16221 }, { "epoch": 1.7586730268863833, "grad_norm": 0.3375965654850006, "learning_rate": 4.1527636732917454e-05, "loss": 0.0152, "step": 16222 }, { "epoch": 1.7587814397224633, "grad_norm": 0.28820356726646423, "learning_rate": 4.152400986508052e-05, "loss": 0.0378, "step": 16223 }, { "epoch": 1.7588898525585428, "grad_norm": 0.285565584897995, "learning_rate": 4.152038299724358e-05, "loss": 0.0107, "step": 16224 }, { "epoch": 1.7589982653946228, "grad_norm": 0.19543001055717468, "learning_rate": 4.1516756129406646e-05, "loss": 0.0088, "step": 16225 }, { "epoch": 1.7591066782307025, "grad_norm": 0.8991948366165161, "learning_rate": 4.151312926156971e-05, "loss": 0.0261, "step": 16226 }, { "epoch": 1.7592150910667823, "grad_norm": 0.395026296377182, "learning_rate": 4.1509502393732774e-05, "loss": 0.0298, "step": 16227 }, { "epoch": 1.759323503902862, "grad_norm": 0.3828030526638031, "learning_rate": 4.150587552589584e-05, "loss": 0.0195, "step": 16228 }, { "epoch": 1.7594319167389418, "grad_norm": 0.7598692774772644, "learning_rate": 4.15022486580589e-05, "loss": 0.0356, "step": 16229 }, { "epoch": 1.7595403295750218, "grad_norm": 0.08648857474327087, "learning_rate": 4.1498621790221967e-05, "loss": 0.0051, "step": 16230 }, { "epoch": 1.7596487424111014, "grad_norm": 0.3905376195907593, "learning_rate": 4.149499492238503e-05, "loss": 0.0356, "step": 16231 }, { "epoch": 1.7597571552471813, "grad_norm": 0.3365510106086731, "learning_rate": 4.1491368054548095e-05, "loss": 0.0632, "step": 16232 }, { "epoch": 1.7598655680832611, "grad_norm": 0.4828578531742096, "learning_rate": 4.148774118671116e-05, "loss": 0.0355, "step": 16233 }, { "epoch": 1.7599739809193409, "grad_norm": 0.36932554841041565, "learning_rate": 4.148411431887422e-05, "loss": 0.027, "step": 16234 }, { "epoch": 1.7600823937554206, "grad_norm": 0.7914566397666931, "learning_rate": 4.148048745103729e-05, "loss": 0.046, "step": 16235 }, { "epoch": 1.7601908065915004, "grad_norm": 0.7703418135643005, "learning_rate": 4.147686058320035e-05, "loss": 0.0283, "step": 16236 }, { "epoch": 1.7602992194275804, "grad_norm": 0.060741301625967026, "learning_rate": 4.1473233715363415e-05, "loss": 0.0067, "step": 16237 }, { "epoch": 1.76040763226366, "grad_norm": 0.26222631335258484, "learning_rate": 4.146960684752648e-05, "loss": 0.0216, "step": 16238 }, { "epoch": 1.76051604509974, "grad_norm": 0.48566651344299316, "learning_rate": 4.146597997968954e-05, "loss": 0.0193, "step": 16239 }, { "epoch": 1.7606244579358195, "grad_norm": 0.036329448223114014, "learning_rate": 4.146235311185261e-05, "loss": 0.0009, "step": 16240 }, { "epoch": 1.7607328707718994, "grad_norm": 0.3583032488822937, "learning_rate": 4.145872624401567e-05, "loss": 0.0175, "step": 16241 }, { "epoch": 1.7608412836079792, "grad_norm": 0.36302438378334045, "learning_rate": 4.1455099376178736e-05, "loss": 0.0151, "step": 16242 }, { "epoch": 1.760949696444059, "grad_norm": 0.1441458761692047, "learning_rate": 4.14514725083418e-05, "loss": 0.0089, "step": 16243 }, { "epoch": 1.7610581092801387, "grad_norm": 0.09548982977867126, "learning_rate": 4.1447845640504864e-05, "loss": 0.0054, "step": 16244 }, { "epoch": 1.7611665221162185, "grad_norm": 0.07249023020267487, "learning_rate": 4.144421877266793e-05, "loss": 0.0073, "step": 16245 }, { "epoch": 1.7612749349522985, "grad_norm": 0.2715713381767273, "learning_rate": 4.1440591904830985e-05, "loss": 0.0106, "step": 16246 }, { "epoch": 1.761383347788378, "grad_norm": 0.23777908086776733, "learning_rate": 4.143696503699405e-05, "loss": 0.0352, "step": 16247 }, { "epoch": 1.761491760624458, "grad_norm": 0.45800352096557617, "learning_rate": 4.1433338169157114e-05, "loss": 0.0202, "step": 16248 }, { "epoch": 1.7616001734605378, "grad_norm": 0.3073173761367798, "learning_rate": 4.1429711301320185e-05, "loss": 0.0174, "step": 16249 }, { "epoch": 1.7617085862966175, "grad_norm": 0.7068946957588196, "learning_rate": 4.142608443348325e-05, "loss": 0.0295, "step": 16250 }, { "epoch": 1.7618169991326973, "grad_norm": 0.26588350534439087, "learning_rate": 4.142245756564631e-05, "loss": 0.0183, "step": 16251 }, { "epoch": 1.761925411968777, "grad_norm": 0.33782774209976196, "learning_rate": 4.141883069780938e-05, "loss": 0.0268, "step": 16252 }, { "epoch": 1.762033824804857, "grad_norm": 0.27743610739707947, "learning_rate": 4.141520382997244e-05, "loss": 0.0102, "step": 16253 }, { "epoch": 1.7621422376409366, "grad_norm": 0.17908842861652374, "learning_rate": 4.14115769621355e-05, "loss": 0.0118, "step": 16254 }, { "epoch": 1.7622506504770166, "grad_norm": 0.47381073236465454, "learning_rate": 4.140795009429856e-05, "loss": 0.0253, "step": 16255 }, { "epoch": 1.762359063313096, "grad_norm": 0.4627314805984497, "learning_rate": 4.1404323226461627e-05, "loss": 0.0145, "step": 16256 }, { "epoch": 1.762467476149176, "grad_norm": 0.42592060565948486, "learning_rate": 4.140069635862469e-05, "loss": 0.0174, "step": 16257 }, { "epoch": 1.7625758889852559, "grad_norm": 0.14772582054138184, "learning_rate": 4.139706949078776e-05, "loss": 0.0051, "step": 16258 }, { "epoch": 1.7626843018213356, "grad_norm": 0.37671053409576416, "learning_rate": 4.1393442622950826e-05, "loss": 0.0604, "step": 16259 }, { "epoch": 1.7627927146574154, "grad_norm": 0.14671802520751953, "learning_rate": 4.138981575511389e-05, "loss": 0.0068, "step": 16260 }, { "epoch": 1.7629011274934951, "grad_norm": 0.44907262921333313, "learning_rate": 4.138618888727695e-05, "loss": 0.0393, "step": 16261 }, { "epoch": 1.7630095403295751, "grad_norm": 0.6234407424926758, "learning_rate": 4.138256201944001e-05, "loss": 0.1029, "step": 16262 }, { "epoch": 1.7631179531656547, "grad_norm": 0.3760308623313904, "learning_rate": 4.1378935151603075e-05, "loss": 0.0243, "step": 16263 }, { "epoch": 1.7632263660017347, "grad_norm": 0.6797264218330383, "learning_rate": 4.137530828376614e-05, "loss": 0.0306, "step": 16264 }, { "epoch": 1.7633347788378144, "grad_norm": 0.09622880816459656, "learning_rate": 4.1371681415929203e-05, "loss": 0.003, "step": 16265 }, { "epoch": 1.7634431916738942, "grad_norm": 0.3744834065437317, "learning_rate": 4.1368054548092274e-05, "loss": 0.0441, "step": 16266 }, { "epoch": 1.763551604509974, "grad_norm": 1.1342756748199463, "learning_rate": 4.136442768025534e-05, "loss": 0.0243, "step": 16267 }, { "epoch": 1.7636600173460537, "grad_norm": 0.3096745014190674, "learning_rate": 4.1360800812418396e-05, "loss": 0.0202, "step": 16268 }, { "epoch": 1.7637684301821337, "grad_norm": 0.39562052488327026, "learning_rate": 4.135717394458146e-05, "loss": 0.0333, "step": 16269 }, { "epoch": 1.7638768430182132, "grad_norm": 0.21695075929164886, "learning_rate": 4.1353547076744524e-05, "loss": 0.024, "step": 16270 }, { "epoch": 1.7639852558542932, "grad_norm": 0.28920280933380127, "learning_rate": 4.134992020890759e-05, "loss": 0.024, "step": 16271 }, { "epoch": 1.764093668690373, "grad_norm": 0.5625963807106018, "learning_rate": 4.134629334107065e-05, "loss": 0.0485, "step": 16272 }, { "epoch": 1.7642020815264527, "grad_norm": 0.225531205534935, "learning_rate": 4.1342666473233716e-05, "loss": 0.0147, "step": 16273 }, { "epoch": 1.7643104943625325, "grad_norm": 0.5888405442237854, "learning_rate": 4.133903960539678e-05, "loss": 0.0677, "step": 16274 }, { "epoch": 1.7644189071986123, "grad_norm": 0.21540281176567078, "learning_rate": 4.1335412737559845e-05, "loss": 0.0189, "step": 16275 }, { "epoch": 1.7645273200346923, "grad_norm": 0.3448605537414551, "learning_rate": 4.133178586972291e-05, "loss": 0.0155, "step": 16276 }, { "epoch": 1.7646357328707718, "grad_norm": 0.5589333772659302, "learning_rate": 4.132815900188597e-05, "loss": 0.0199, "step": 16277 }, { "epoch": 1.7647441457068518, "grad_norm": 0.38930538296699524, "learning_rate": 4.132453213404904e-05, "loss": 0.0218, "step": 16278 }, { "epoch": 1.7648525585429313, "grad_norm": 0.24989433586597443, "learning_rate": 4.13209052662121e-05, "loss": 0.0183, "step": 16279 }, { "epoch": 1.7649609713790113, "grad_norm": 0.29809629917144775, "learning_rate": 4.1317278398375165e-05, "loss": 0.0229, "step": 16280 }, { "epoch": 1.765069384215091, "grad_norm": 0.30638939142227173, "learning_rate": 4.131365153053823e-05, "loss": 0.0335, "step": 16281 }, { "epoch": 1.7651777970511708, "grad_norm": 0.036982834339141846, "learning_rate": 4.131002466270129e-05, "loss": 0.0033, "step": 16282 }, { "epoch": 1.7652862098872506, "grad_norm": 0.7992866039276123, "learning_rate": 4.130639779486436e-05, "loss": 0.0545, "step": 16283 }, { "epoch": 1.7653946227233304, "grad_norm": 0.35245758295059204, "learning_rate": 4.130277092702742e-05, "loss": 0.024, "step": 16284 }, { "epoch": 1.7655030355594103, "grad_norm": 0.8443424105644226, "learning_rate": 4.1299144059190486e-05, "loss": 0.0392, "step": 16285 }, { "epoch": 1.76561144839549, "grad_norm": 0.7009268999099731, "learning_rate": 4.129551719135355e-05, "loss": 0.0308, "step": 16286 }, { "epoch": 1.7657198612315699, "grad_norm": 0.5527660250663757, "learning_rate": 4.1291890323516614e-05, "loss": 0.0226, "step": 16287 }, { "epoch": 1.7658282740676496, "grad_norm": 0.15949410200119019, "learning_rate": 4.128826345567968e-05, "loss": 0.0361, "step": 16288 }, { "epoch": 1.7659366869037294, "grad_norm": 0.2577139735221863, "learning_rate": 4.128463658784274e-05, "loss": 0.0154, "step": 16289 }, { "epoch": 1.7660450997398092, "grad_norm": 0.17777906358242035, "learning_rate": 4.1281009720005806e-05, "loss": 0.0149, "step": 16290 }, { "epoch": 1.766153512575889, "grad_norm": 0.5565987229347229, "learning_rate": 4.1277382852168863e-05, "loss": 0.0495, "step": 16291 }, { "epoch": 1.766261925411969, "grad_norm": 0.6872382760047913, "learning_rate": 4.1273755984331934e-05, "loss": 0.0704, "step": 16292 }, { "epoch": 1.7663703382480485, "grad_norm": 0.06838735938072205, "learning_rate": 4.1270129116495e-05, "loss": 0.0038, "step": 16293 }, { "epoch": 1.7664787510841284, "grad_norm": 0.5039750933647156, "learning_rate": 4.126650224865806e-05, "loss": 0.053, "step": 16294 }, { "epoch": 1.7665871639202082, "grad_norm": 0.13435612618923187, "learning_rate": 4.1262875380821127e-05, "loss": 0.0176, "step": 16295 }, { "epoch": 1.766695576756288, "grad_norm": 0.2772335112094879, "learning_rate": 4.125924851298419e-05, "loss": 0.0194, "step": 16296 }, { "epoch": 1.7668039895923677, "grad_norm": 0.3717644214630127, "learning_rate": 4.1255621645147255e-05, "loss": 0.0288, "step": 16297 }, { "epoch": 1.7669124024284475, "grad_norm": 1.2872748374938965, "learning_rate": 4.125199477731031e-05, "loss": 0.0521, "step": 16298 }, { "epoch": 1.7670208152645275, "grad_norm": 0.12514005601406097, "learning_rate": 4.1248367909473376e-05, "loss": 0.0056, "step": 16299 }, { "epoch": 1.767129228100607, "grad_norm": 0.0496794618666172, "learning_rate": 4.124474104163644e-05, "loss": 0.0034, "step": 16300 }, { "epoch": 1.767237640936687, "grad_norm": 0.15429159998893738, "learning_rate": 4.124111417379951e-05, "loss": 0.0141, "step": 16301 }, { "epoch": 1.7673460537727665, "grad_norm": 0.09335994720458984, "learning_rate": 4.1237487305962575e-05, "loss": 0.0126, "step": 16302 }, { "epoch": 1.7674544666088465, "grad_norm": 0.14039437472820282, "learning_rate": 4.123386043812564e-05, "loss": 0.0113, "step": 16303 }, { "epoch": 1.7675628794449263, "grad_norm": 0.1352689415216446, "learning_rate": 4.1230233570288704e-05, "loss": 0.0126, "step": 16304 }, { "epoch": 1.767671292281006, "grad_norm": 0.06525299698114395, "learning_rate": 4.122660670245177e-05, "loss": 0.0124, "step": 16305 }, { "epoch": 1.7677797051170858, "grad_norm": 0.009527322836220264, "learning_rate": 4.1222979834614825e-05, "loss": 0.0005, "step": 16306 }, { "epoch": 1.7678881179531656, "grad_norm": 0.1530744433403015, "learning_rate": 4.121935296677789e-05, "loss": 0.0131, "step": 16307 }, { "epoch": 1.7679965307892456, "grad_norm": 0.1507968157529831, "learning_rate": 4.121572609894095e-05, "loss": 0.0207, "step": 16308 }, { "epoch": 1.768104943625325, "grad_norm": 0.07875501364469528, "learning_rate": 4.1212099231104024e-05, "loss": 0.0048, "step": 16309 }, { "epoch": 1.768213356461405, "grad_norm": 0.5665362477302551, "learning_rate": 4.120847236326709e-05, "loss": 0.0608, "step": 16310 }, { "epoch": 1.7683217692974849, "grad_norm": 0.20607894659042358, "learning_rate": 4.120484549543015e-05, "loss": 0.0181, "step": 16311 }, { "epoch": 1.7684301821335646, "grad_norm": 0.4152858257293701, "learning_rate": 4.1201218627593216e-05, "loss": 0.0387, "step": 16312 }, { "epoch": 1.7685385949696444, "grad_norm": 0.21772605180740356, "learning_rate": 4.1197591759756274e-05, "loss": 0.0071, "step": 16313 }, { "epoch": 1.7686470078057241, "grad_norm": 0.2261532098054886, "learning_rate": 4.119396489191934e-05, "loss": 0.0168, "step": 16314 }, { "epoch": 1.7687554206418041, "grad_norm": 0.29977601766586304, "learning_rate": 4.11903380240824e-05, "loss": 0.0306, "step": 16315 }, { "epoch": 1.7688638334778837, "grad_norm": 0.21224141120910645, "learning_rate": 4.1186711156245466e-05, "loss": 0.0381, "step": 16316 }, { "epoch": 1.7689722463139637, "grad_norm": 0.3943086564540863, "learning_rate": 4.118308428840853e-05, "loss": 0.0251, "step": 16317 }, { "epoch": 1.7690806591500432, "grad_norm": 0.20152008533477783, "learning_rate": 4.11794574205716e-05, "loss": 0.0172, "step": 16318 }, { "epoch": 1.7691890719861232, "grad_norm": 0.3200722336769104, "learning_rate": 4.1175830552734665e-05, "loss": 0.0218, "step": 16319 }, { "epoch": 1.769297484822203, "grad_norm": 0.115264393389225, "learning_rate": 4.117220368489772e-05, "loss": 0.0079, "step": 16320 }, { "epoch": 1.7694058976582827, "grad_norm": 0.12657983601093292, "learning_rate": 4.1168576817060787e-05, "loss": 0.0049, "step": 16321 }, { "epoch": 1.7695143104943627, "grad_norm": 0.3996013402938843, "learning_rate": 4.116494994922385e-05, "loss": 0.0399, "step": 16322 }, { "epoch": 1.7696227233304422, "grad_norm": 0.3874813914299011, "learning_rate": 4.1161323081386915e-05, "loss": 0.0445, "step": 16323 }, { "epoch": 1.7697311361665222, "grad_norm": 0.32388195395469666, "learning_rate": 4.115769621354998e-05, "loss": 0.0307, "step": 16324 }, { "epoch": 1.7698395490026018, "grad_norm": 0.15198791027069092, "learning_rate": 4.115406934571304e-05, "loss": 0.0134, "step": 16325 }, { "epoch": 1.7699479618386817, "grad_norm": 0.1606980413198471, "learning_rate": 4.115044247787611e-05, "loss": 0.0101, "step": 16326 }, { "epoch": 1.7700563746747615, "grad_norm": 0.3935563266277313, "learning_rate": 4.114681561003917e-05, "loss": 0.03, "step": 16327 }, { "epoch": 1.7701647875108413, "grad_norm": 0.3892815113067627, "learning_rate": 4.1143188742202235e-05, "loss": 0.08, "step": 16328 }, { "epoch": 1.770273200346921, "grad_norm": 0.08465650677680969, "learning_rate": 4.11395618743653e-05, "loss": 0.0168, "step": 16329 }, { "epoch": 1.7703816131830008, "grad_norm": 0.2625415027141571, "learning_rate": 4.1135935006528364e-05, "loss": 0.0241, "step": 16330 }, { "epoch": 1.7704900260190808, "grad_norm": 0.5128301978111267, "learning_rate": 4.113230813869143e-05, "loss": 0.0246, "step": 16331 }, { "epoch": 1.7705984388551603, "grad_norm": 0.37934359908103943, "learning_rate": 4.112868127085449e-05, "loss": 0.0126, "step": 16332 }, { "epoch": 1.7707068516912403, "grad_norm": 0.3777092695236206, "learning_rate": 4.1125054403017556e-05, "loss": 0.0331, "step": 16333 }, { "epoch": 1.77081526452732, "grad_norm": 0.1609169989824295, "learning_rate": 4.112142753518062e-05, "loss": 0.0197, "step": 16334 }, { "epoch": 1.7709236773633998, "grad_norm": 0.12166911363601685, "learning_rate": 4.1117800667343684e-05, "loss": 0.0038, "step": 16335 }, { "epoch": 1.7710320901994796, "grad_norm": 0.6476441621780396, "learning_rate": 4.111417379950675e-05, "loss": 0.0507, "step": 16336 }, { "epoch": 1.7711405030355594, "grad_norm": 0.24882592260837555, "learning_rate": 4.111054693166981e-05, "loss": 0.0256, "step": 16337 }, { "epoch": 1.7712489158716394, "grad_norm": 0.1578785479068756, "learning_rate": 4.1106920063832876e-05, "loss": 0.0113, "step": 16338 }, { "epoch": 1.771357328707719, "grad_norm": 0.07593832910060883, "learning_rate": 4.110329319599594e-05, "loss": 0.0035, "step": 16339 }, { "epoch": 1.7714657415437989, "grad_norm": 0.5418714284896851, "learning_rate": 4.1099666328159005e-05, "loss": 0.1226, "step": 16340 }, { "epoch": 1.7715741543798784, "grad_norm": 0.2436244934797287, "learning_rate": 4.109603946032207e-05, "loss": 0.0231, "step": 16341 }, { "epoch": 1.7716825672159584, "grad_norm": 0.14287759363651276, "learning_rate": 4.109241259248513e-05, "loss": 0.0135, "step": 16342 }, { "epoch": 1.7717909800520382, "grad_norm": 0.20512264966964722, "learning_rate": 4.108878572464819e-05, "loss": 0.0242, "step": 16343 }, { "epoch": 1.771899392888118, "grad_norm": 0.10462633520364761, "learning_rate": 4.108515885681126e-05, "loss": 0.0054, "step": 16344 }, { "epoch": 1.7720078057241977, "grad_norm": 0.056172989308834076, "learning_rate": 4.1081531988974325e-05, "loss": 0.0045, "step": 16345 }, { "epoch": 1.7721162185602775, "grad_norm": 0.2256172001361847, "learning_rate": 4.107790512113739e-05, "loss": 0.0175, "step": 16346 }, { "epoch": 1.7722246313963574, "grad_norm": 0.33613383769989014, "learning_rate": 4.107427825330045e-05, "loss": 0.04, "step": 16347 }, { "epoch": 1.772333044232437, "grad_norm": 0.2165161371231079, "learning_rate": 4.107065138546352e-05, "loss": 0.0223, "step": 16348 }, { "epoch": 1.772441457068517, "grad_norm": 0.2094104140996933, "learning_rate": 4.106702451762658e-05, "loss": 0.0298, "step": 16349 }, { "epoch": 1.7725498699045967, "grad_norm": 0.10809121280908585, "learning_rate": 4.106339764978964e-05, "loss": 0.0079, "step": 16350 }, { "epoch": 1.7726582827406765, "grad_norm": 0.05298265814781189, "learning_rate": 4.10597707819527e-05, "loss": 0.0062, "step": 16351 }, { "epoch": 1.7727666955767563, "grad_norm": 0.11129093915224075, "learning_rate": 4.105614391411577e-05, "loss": 0.0053, "step": 16352 }, { "epoch": 1.772875108412836, "grad_norm": 0.5822787284851074, "learning_rate": 4.105251704627884e-05, "loss": 0.086, "step": 16353 }, { "epoch": 1.772983521248916, "grad_norm": 0.4678027629852295, "learning_rate": 4.10488901784419e-05, "loss": 0.0454, "step": 16354 }, { "epoch": 1.7730919340849955, "grad_norm": 0.2313755452632904, "learning_rate": 4.1045263310604966e-05, "loss": 0.0192, "step": 16355 }, { "epoch": 1.7732003469210755, "grad_norm": 0.5667398571968079, "learning_rate": 4.104163644276803e-05, "loss": 0.0445, "step": 16356 }, { "epoch": 1.7733087597571553, "grad_norm": 0.22224479913711548, "learning_rate": 4.1038009574931094e-05, "loss": 0.02, "step": 16357 }, { "epoch": 1.773417172593235, "grad_norm": 0.7941175699234009, "learning_rate": 4.103438270709415e-05, "loss": 0.0334, "step": 16358 }, { "epoch": 1.7735255854293148, "grad_norm": 0.2942589521408081, "learning_rate": 4.1030755839257216e-05, "loss": 0.0446, "step": 16359 }, { "epoch": 1.7736339982653946, "grad_norm": 0.22666950523853302, "learning_rate": 4.102712897142028e-05, "loss": 0.0156, "step": 16360 }, { "epoch": 1.7737424111014746, "grad_norm": 0.07683713734149933, "learning_rate": 4.102350210358335e-05, "loss": 0.0057, "step": 16361 }, { "epoch": 1.773850823937554, "grad_norm": 0.7360991835594177, "learning_rate": 4.1019875235746415e-05, "loss": 0.0584, "step": 16362 }, { "epoch": 1.773959236773634, "grad_norm": 0.08207883685827255, "learning_rate": 4.101624836790948e-05, "loss": 0.0055, "step": 16363 }, { "epoch": 1.7740676496097136, "grad_norm": 0.1026863232254982, "learning_rate": 4.101262150007254e-05, "loss": 0.007, "step": 16364 }, { "epoch": 1.7741760624457936, "grad_norm": 0.25132569670677185, "learning_rate": 4.10089946322356e-05, "loss": 0.0166, "step": 16365 }, { "epoch": 1.7742844752818734, "grad_norm": 0.3729614317417145, "learning_rate": 4.1005367764398665e-05, "loss": 0.0616, "step": 16366 }, { "epoch": 1.7743928881179531, "grad_norm": 0.1037873774766922, "learning_rate": 4.100174089656173e-05, "loss": 0.0094, "step": 16367 }, { "epoch": 1.774501300954033, "grad_norm": 0.18434301018714905, "learning_rate": 4.099811402872479e-05, "loss": 0.0114, "step": 16368 }, { "epoch": 1.7746097137901127, "grad_norm": 0.3173247277736664, "learning_rate": 4.099448716088786e-05, "loss": 0.0255, "step": 16369 }, { "epoch": 1.7747181266261927, "grad_norm": 0.21924781799316406, "learning_rate": 4.099086029305093e-05, "loss": 0.0147, "step": 16370 }, { "epoch": 1.7748265394622722, "grad_norm": 0.11834361404180527, "learning_rate": 4.098723342521399e-05, "loss": 0.0112, "step": 16371 }, { "epoch": 1.7749349522983522, "grad_norm": 0.07075751572847366, "learning_rate": 4.098360655737705e-05, "loss": 0.009, "step": 16372 }, { "epoch": 1.775043365134432, "grad_norm": 0.14706826210021973, "learning_rate": 4.097997968954011e-05, "loss": 0.0084, "step": 16373 }, { "epoch": 1.7751517779705117, "grad_norm": 0.1928883194923401, "learning_rate": 4.097635282170318e-05, "loss": 0.017, "step": 16374 }, { "epoch": 1.7752601908065915, "grad_norm": 0.12551794946193695, "learning_rate": 4.097272595386624e-05, "loss": 0.0075, "step": 16375 }, { "epoch": 1.7753686036426712, "grad_norm": 0.08415193110704422, "learning_rate": 4.0969099086029306e-05, "loss": 0.0126, "step": 16376 }, { "epoch": 1.7754770164787512, "grad_norm": 0.2638685703277588, "learning_rate": 4.096547221819237e-05, "loss": 0.0386, "step": 16377 }, { "epoch": 1.7755854293148308, "grad_norm": 0.20021581649780273, "learning_rate": 4.0961845350355434e-05, "loss": 0.0278, "step": 16378 }, { "epoch": 1.7756938421509108, "grad_norm": 0.11707642674446106, "learning_rate": 4.09582184825185e-05, "loss": 0.0128, "step": 16379 }, { "epoch": 1.7758022549869905, "grad_norm": 0.1896810084581375, "learning_rate": 4.095459161468156e-05, "loss": 0.0166, "step": 16380 }, { "epoch": 1.7759106678230703, "grad_norm": 0.11566353589296341, "learning_rate": 4.0950964746844626e-05, "loss": 0.0065, "step": 16381 }, { "epoch": 1.77601908065915, "grad_norm": 0.2924099564552307, "learning_rate": 4.094733787900769e-05, "loss": 0.0315, "step": 16382 }, { "epoch": 1.7761274934952298, "grad_norm": 1.020245909690857, "learning_rate": 4.0943711011170754e-05, "loss": 0.0336, "step": 16383 }, { "epoch": 1.7762359063313098, "grad_norm": 0.193955197930336, "learning_rate": 4.094008414333382e-05, "loss": 0.0098, "step": 16384 }, { "epoch": 1.7763443191673893, "grad_norm": 0.27603957056999207, "learning_rate": 4.093645727549688e-05, "loss": 0.0289, "step": 16385 }, { "epoch": 1.7764527320034693, "grad_norm": 0.05876943841576576, "learning_rate": 4.093283040765995e-05, "loss": 0.0037, "step": 16386 }, { "epoch": 1.7765611448395489, "grad_norm": 0.3598862588405609, "learning_rate": 4.092920353982301e-05, "loss": 0.0275, "step": 16387 }, { "epoch": 1.7766695576756288, "grad_norm": 0.3936268985271454, "learning_rate": 4.0925576671986075e-05, "loss": 0.0075, "step": 16388 }, { "epoch": 1.7767779705117086, "grad_norm": 0.5394269227981567, "learning_rate": 4.092194980414914e-05, "loss": 0.0424, "step": 16389 }, { "epoch": 1.7768863833477884, "grad_norm": 0.2343827188014984, "learning_rate": 4.09183229363122e-05, "loss": 0.0225, "step": 16390 }, { "epoch": 1.7769947961838681, "grad_norm": 0.09737219661474228, "learning_rate": 4.091469606847527e-05, "loss": 0.0088, "step": 16391 }, { "epoch": 1.777103209019948, "grad_norm": 0.1530141681432724, "learning_rate": 4.091106920063833e-05, "loss": 0.0077, "step": 16392 }, { "epoch": 1.7772116218560279, "grad_norm": 0.18401192128658295, "learning_rate": 4.0907442332801395e-05, "loss": 0.0152, "step": 16393 }, { "epoch": 1.7773200346921074, "grad_norm": 0.116644948720932, "learning_rate": 4.090381546496446e-05, "loss": 0.0092, "step": 16394 }, { "epoch": 1.7774284475281874, "grad_norm": 0.47095146775245667, "learning_rate": 4.090018859712752e-05, "loss": 0.0249, "step": 16395 }, { "epoch": 1.7775368603642672, "grad_norm": 0.2437497079372406, "learning_rate": 4.089656172929059e-05, "loss": 0.0222, "step": 16396 }, { "epoch": 1.777645273200347, "grad_norm": 0.44485539197921753, "learning_rate": 4.089293486145365e-05, "loss": 0.0172, "step": 16397 }, { "epoch": 1.7777536860364267, "grad_norm": 1.0841749906539917, "learning_rate": 4.0889307993616716e-05, "loss": 0.0119, "step": 16398 }, { "epoch": 1.7778620988725065, "grad_norm": 0.11994177103042603, "learning_rate": 4.088568112577978e-05, "loss": 0.0136, "step": 16399 }, { "epoch": 1.7779705117085864, "grad_norm": 0.05187549442052841, "learning_rate": 4.0882054257942844e-05, "loss": 0.002, "step": 16400 }, { "epoch": 1.778078924544666, "grad_norm": 0.6905632615089417, "learning_rate": 4.087842739010591e-05, "loss": 0.0266, "step": 16401 }, { "epoch": 1.778187337380746, "grad_norm": 0.3450130522251129, "learning_rate": 4.087480052226897e-05, "loss": 0.0089, "step": 16402 }, { "epoch": 1.7782957502168255, "grad_norm": 0.2883599102497101, "learning_rate": 4.087117365443203e-05, "loss": 0.0155, "step": 16403 }, { "epoch": 1.7784041630529055, "grad_norm": 0.2217368483543396, "learning_rate": 4.08675467865951e-05, "loss": 0.0218, "step": 16404 }, { "epoch": 1.7785125758889853, "grad_norm": 0.10976290702819824, "learning_rate": 4.0863919918758165e-05, "loss": 0.0073, "step": 16405 }, { "epoch": 1.778620988725065, "grad_norm": 0.4034128487110138, "learning_rate": 4.086029305092123e-05, "loss": 0.0377, "step": 16406 }, { "epoch": 1.7787294015611448, "grad_norm": 0.07963834702968597, "learning_rate": 4.085666618308429e-05, "loss": 0.0093, "step": 16407 }, { "epoch": 1.7788378143972245, "grad_norm": 0.3250278830528259, "learning_rate": 4.085303931524736e-05, "loss": 0.0285, "step": 16408 }, { "epoch": 1.7789462272333045, "grad_norm": 0.20858357846736908, "learning_rate": 4.084941244741042e-05, "loss": 0.0112, "step": 16409 }, { "epoch": 1.779054640069384, "grad_norm": 0.40314212441444397, "learning_rate": 4.084578557957348e-05, "loss": 0.0407, "step": 16410 }, { "epoch": 1.779163052905464, "grad_norm": 0.18027949333190918, "learning_rate": 4.084215871173654e-05, "loss": 0.0135, "step": 16411 }, { "epoch": 1.7792714657415438, "grad_norm": 0.0804327055811882, "learning_rate": 4.0838531843899607e-05, "loss": 0.0102, "step": 16412 }, { "epoch": 1.7793798785776236, "grad_norm": 0.18500348925590515, "learning_rate": 4.083490497606268e-05, "loss": 0.0258, "step": 16413 }, { "epoch": 1.7794882914137033, "grad_norm": 0.11051376909017563, "learning_rate": 4.083127810822574e-05, "loss": 0.0076, "step": 16414 }, { "epoch": 1.7795967042497831, "grad_norm": 0.17150753736495972, "learning_rate": 4.0827651240388806e-05, "loss": 0.0194, "step": 16415 }, { "epoch": 1.779705117085863, "grad_norm": 0.44028061628341675, "learning_rate": 4.082402437255187e-05, "loss": 0.0381, "step": 16416 }, { "epoch": 1.7798135299219426, "grad_norm": 0.4012584090232849, "learning_rate": 4.082039750471493e-05, "loss": 0.0296, "step": 16417 }, { "epoch": 1.7799219427580226, "grad_norm": 0.06780876964330673, "learning_rate": 4.081677063687799e-05, "loss": 0.0033, "step": 16418 }, { "epoch": 1.7800303555941024, "grad_norm": 0.07851637154817581, "learning_rate": 4.0813143769041055e-05, "loss": 0.0063, "step": 16419 }, { "epoch": 1.7801387684301822, "grad_norm": 0.2464127391576767, "learning_rate": 4.080951690120412e-05, "loss": 0.0301, "step": 16420 }, { "epoch": 1.780247181266262, "grad_norm": 0.17456139624118805, "learning_rate": 4.0805890033367184e-05, "loss": 0.0076, "step": 16421 }, { "epoch": 1.7803555941023417, "grad_norm": 0.40031275153160095, "learning_rate": 4.0802263165530254e-05, "loss": 0.0227, "step": 16422 }, { "epoch": 1.7804640069384217, "grad_norm": 0.16509810090065002, "learning_rate": 4.079863629769332e-05, "loss": 0.0121, "step": 16423 }, { "epoch": 1.7805724197745012, "grad_norm": 0.025855623185634613, "learning_rate": 4.0795009429856376e-05, "loss": 0.001, "step": 16424 }, { "epoch": 1.7806808326105812, "grad_norm": 0.26201745867729187, "learning_rate": 4.079138256201944e-05, "loss": 0.0131, "step": 16425 }, { "epoch": 1.7807892454466607, "grad_norm": 0.2595888078212738, "learning_rate": 4.0787755694182504e-05, "loss": 0.0194, "step": 16426 }, { "epoch": 1.7808976582827407, "grad_norm": 0.14552636444568634, "learning_rate": 4.078412882634557e-05, "loss": 0.0161, "step": 16427 }, { "epoch": 1.7810060711188205, "grad_norm": 0.4649916887283325, "learning_rate": 4.078050195850863e-05, "loss": 0.0185, "step": 16428 }, { "epoch": 1.7811144839549002, "grad_norm": 0.20376132428646088, "learning_rate": 4.0776875090671696e-05, "loss": 0.034, "step": 16429 }, { "epoch": 1.78122289679098, "grad_norm": 0.6050735712051392, "learning_rate": 4.077324822283477e-05, "loss": 0.0351, "step": 16430 }, { "epoch": 1.7813313096270598, "grad_norm": 0.1772879809141159, "learning_rate": 4.0769621354997825e-05, "loss": 0.0184, "step": 16431 }, { "epoch": 1.7814397224631398, "grad_norm": 0.13513706624507904, "learning_rate": 4.076599448716089e-05, "loss": 0.0254, "step": 16432 }, { "epoch": 1.7815481352992193, "grad_norm": 0.31956255435943604, "learning_rate": 4.076236761932395e-05, "loss": 0.0338, "step": 16433 }, { "epoch": 1.7816565481352993, "grad_norm": 0.5657501816749573, "learning_rate": 4.075874075148702e-05, "loss": 0.0166, "step": 16434 }, { "epoch": 1.781764960971379, "grad_norm": 0.15488114953041077, "learning_rate": 4.075511388365008e-05, "loss": 0.02, "step": 16435 }, { "epoch": 1.7818733738074588, "grad_norm": 0.2608163058757782, "learning_rate": 4.0751487015813145e-05, "loss": 0.0207, "step": 16436 }, { "epoch": 1.7819817866435386, "grad_norm": 0.604695737361908, "learning_rate": 4.074786014797621e-05, "loss": 0.0605, "step": 16437 }, { "epoch": 1.7820901994796183, "grad_norm": 0.25395411252975464, "learning_rate": 4.074423328013927e-05, "loss": 0.0316, "step": 16438 }, { "epoch": 1.7821986123156983, "grad_norm": 0.2612467110157013, "learning_rate": 4.074060641230234e-05, "loss": 0.0194, "step": 16439 }, { "epoch": 1.7823070251517779, "grad_norm": 0.084763303399086, "learning_rate": 4.07369795444654e-05, "loss": 0.0054, "step": 16440 }, { "epoch": 1.7824154379878578, "grad_norm": 0.6854812502861023, "learning_rate": 4.0733352676628466e-05, "loss": 0.0384, "step": 16441 }, { "epoch": 1.7825238508239376, "grad_norm": 0.3358289301395416, "learning_rate": 4.072972580879153e-05, "loss": 0.0328, "step": 16442 }, { "epoch": 1.7826322636600174, "grad_norm": 1.0083457231521606, "learning_rate": 4.0726098940954594e-05, "loss": 0.064, "step": 16443 }, { "epoch": 1.7827406764960971, "grad_norm": 0.024099567905068398, "learning_rate": 4.072247207311766e-05, "loss": 0.0014, "step": 16444 }, { "epoch": 1.782849089332177, "grad_norm": 0.5099624395370483, "learning_rate": 4.071884520528072e-05, "loss": 0.0289, "step": 16445 }, { "epoch": 1.7829575021682569, "grad_norm": 0.27223023772239685, "learning_rate": 4.0715218337443786e-05, "loss": 0.0202, "step": 16446 }, { "epoch": 1.7830659150043364, "grad_norm": 0.6226119995117188, "learning_rate": 4.0711591469606843e-05, "loss": 0.0201, "step": 16447 }, { "epoch": 1.7831743278404164, "grad_norm": 0.41048282384872437, "learning_rate": 4.0707964601769914e-05, "loss": 0.044, "step": 16448 }, { "epoch": 1.783282740676496, "grad_norm": 0.4289579689502716, "learning_rate": 4.070433773393298e-05, "loss": 0.0462, "step": 16449 }, { "epoch": 1.783391153512576, "grad_norm": 0.6875882744789124, "learning_rate": 4.070071086609604e-05, "loss": 0.0364, "step": 16450 }, { "epoch": 1.7834995663486557, "grad_norm": 0.08609075099229813, "learning_rate": 4.069708399825911e-05, "loss": 0.0048, "step": 16451 }, { "epoch": 1.7836079791847355, "grad_norm": 0.03804389387369156, "learning_rate": 4.069345713042217e-05, "loss": 0.0021, "step": 16452 }, { "epoch": 1.7837163920208152, "grad_norm": 0.19428618252277374, "learning_rate": 4.0689830262585235e-05, "loss": 0.0143, "step": 16453 }, { "epoch": 1.783824804856895, "grad_norm": 0.24162891507148743, "learning_rate": 4.06862033947483e-05, "loss": 0.0269, "step": 16454 }, { "epoch": 1.783933217692975, "grad_norm": 0.07846648246049881, "learning_rate": 4.0682576526911356e-05, "loss": 0.0061, "step": 16455 }, { "epoch": 1.7840416305290545, "grad_norm": 0.17795966565608978, "learning_rate": 4.067894965907443e-05, "loss": 0.0109, "step": 16456 }, { "epoch": 1.7841500433651345, "grad_norm": 0.3091510534286499, "learning_rate": 4.067532279123749e-05, "loss": 0.012, "step": 16457 }, { "epoch": 1.7842584562012143, "grad_norm": 0.2058553844690323, "learning_rate": 4.0671695923400555e-05, "loss": 0.0224, "step": 16458 }, { "epoch": 1.784366869037294, "grad_norm": 0.22272133827209473, "learning_rate": 4.066806905556362e-05, "loss": 0.0142, "step": 16459 }, { "epoch": 1.7844752818733738, "grad_norm": 0.1907425969839096, "learning_rate": 4.0664442187726684e-05, "loss": 0.0148, "step": 16460 }, { "epoch": 1.7845836947094535, "grad_norm": 0.4453188180923462, "learning_rate": 4.066081531988975e-05, "loss": 0.0365, "step": 16461 }, { "epoch": 1.7846921075455335, "grad_norm": 0.3332178592681885, "learning_rate": 4.0657188452052805e-05, "loss": 0.0137, "step": 16462 }, { "epoch": 1.784800520381613, "grad_norm": 1.194544792175293, "learning_rate": 4.065356158421587e-05, "loss": 0.0233, "step": 16463 }, { "epoch": 1.784908933217693, "grad_norm": 0.11279363185167313, "learning_rate": 4.064993471637893e-05, "loss": 0.0065, "step": 16464 }, { "epoch": 1.7850173460537726, "grad_norm": 0.3702998757362366, "learning_rate": 4.0646307848542004e-05, "loss": 0.0605, "step": 16465 }, { "epoch": 1.7851257588898526, "grad_norm": 0.4531867206096649, "learning_rate": 4.064268098070507e-05, "loss": 0.0698, "step": 16466 }, { "epoch": 1.7852341717259324, "grad_norm": 0.17804113030433655, "learning_rate": 4.063905411286813e-05, "loss": 0.0232, "step": 16467 }, { "epoch": 1.7853425845620121, "grad_norm": 0.7469697594642639, "learning_rate": 4.0635427245031196e-05, "loss": 0.0882, "step": 16468 }, { "epoch": 1.785450997398092, "grad_norm": 0.13241055607795715, "learning_rate": 4.0631800377194254e-05, "loss": 0.0082, "step": 16469 }, { "epoch": 1.7855594102341716, "grad_norm": 0.35025712847709656, "learning_rate": 4.062817350935732e-05, "loss": 0.0484, "step": 16470 }, { "epoch": 1.7856678230702516, "grad_norm": 0.38694024085998535, "learning_rate": 4.062454664152038e-05, "loss": 0.0323, "step": 16471 }, { "epoch": 1.7857762359063312, "grad_norm": 0.351819783449173, "learning_rate": 4.0620919773683446e-05, "loss": 0.0279, "step": 16472 }, { "epoch": 1.7858846487424112, "grad_norm": 0.24719879031181335, "learning_rate": 4.061729290584651e-05, "loss": 0.0219, "step": 16473 }, { "epoch": 1.785993061578491, "grad_norm": 0.4137483537197113, "learning_rate": 4.061366603800958e-05, "loss": 0.0458, "step": 16474 }, { "epoch": 1.7861014744145707, "grad_norm": 0.14902126789093018, "learning_rate": 4.0610039170172645e-05, "loss": 0.0102, "step": 16475 }, { "epoch": 1.7862098872506504, "grad_norm": 0.14323806762695312, "learning_rate": 4.06064123023357e-05, "loss": 0.0103, "step": 16476 }, { "epoch": 1.7863183000867302, "grad_norm": 0.7881564497947693, "learning_rate": 4.060278543449877e-05, "loss": 0.0964, "step": 16477 }, { "epoch": 1.7864267129228102, "grad_norm": 0.055258363485336304, "learning_rate": 4.059915856666183e-05, "loss": 0.0061, "step": 16478 }, { "epoch": 1.7865351257588897, "grad_norm": 0.2065984606742859, "learning_rate": 4.0595531698824895e-05, "loss": 0.0175, "step": 16479 }, { "epoch": 1.7866435385949697, "grad_norm": 0.5167943239212036, "learning_rate": 4.059190483098796e-05, "loss": 0.0545, "step": 16480 }, { "epoch": 1.7867519514310495, "grad_norm": 0.2731074094772339, "learning_rate": 4.058827796315102e-05, "loss": 0.0233, "step": 16481 }, { "epoch": 1.7868603642671292, "grad_norm": 0.1134909987449646, "learning_rate": 4.0584651095314094e-05, "loss": 0.0118, "step": 16482 }, { "epoch": 1.786968777103209, "grad_norm": 0.3969838619232178, "learning_rate": 4.058102422747715e-05, "loss": 0.0653, "step": 16483 }, { "epoch": 1.7870771899392888, "grad_norm": 0.6087436676025391, "learning_rate": 4.0577397359640215e-05, "loss": 0.0751, "step": 16484 }, { "epoch": 1.7871856027753688, "grad_norm": 0.49055612087249756, "learning_rate": 4.057377049180328e-05, "loss": 0.0196, "step": 16485 }, { "epoch": 1.7872940156114483, "grad_norm": 0.3135542571544647, "learning_rate": 4.0570143623966344e-05, "loss": 0.0185, "step": 16486 }, { "epoch": 1.7874024284475283, "grad_norm": 0.3492547571659088, "learning_rate": 4.056651675612941e-05, "loss": 0.0271, "step": 16487 }, { "epoch": 1.7875108412836078, "grad_norm": 0.1601722687482834, "learning_rate": 4.056288988829247e-05, "loss": 0.0056, "step": 16488 }, { "epoch": 1.7876192541196878, "grad_norm": 0.20697246491909027, "learning_rate": 4.0559263020455536e-05, "loss": 0.0134, "step": 16489 }, { "epoch": 1.7877276669557676, "grad_norm": 0.22145384550094604, "learning_rate": 4.05556361526186e-05, "loss": 0.0213, "step": 16490 }, { "epoch": 1.7878360797918473, "grad_norm": 0.31652799248695374, "learning_rate": 4.0552009284781664e-05, "loss": 0.0348, "step": 16491 }, { "epoch": 1.787944492627927, "grad_norm": 0.2795100212097168, "learning_rate": 4.054838241694473e-05, "loss": 0.0226, "step": 16492 }, { "epoch": 1.7880529054640069, "grad_norm": 0.24553737044334412, "learning_rate": 4.054475554910779e-05, "loss": 0.0198, "step": 16493 }, { "epoch": 1.7881613183000868, "grad_norm": 0.19050654768943787, "learning_rate": 4.0541128681270856e-05, "loss": 0.0152, "step": 16494 }, { "epoch": 1.7882697311361664, "grad_norm": 0.09385620802640915, "learning_rate": 4.053750181343392e-05, "loss": 0.0079, "step": 16495 }, { "epoch": 1.7883781439722464, "grad_norm": 0.24239826202392578, "learning_rate": 4.0533874945596985e-05, "loss": 0.019, "step": 16496 }, { "epoch": 1.7884865568083261, "grad_norm": 0.15335646271705627, "learning_rate": 4.053024807776005e-05, "loss": 0.0231, "step": 16497 }, { "epoch": 1.788594969644406, "grad_norm": 0.24136772751808167, "learning_rate": 4.052662120992311e-05, "loss": 0.0294, "step": 16498 }, { "epoch": 1.7887033824804857, "grad_norm": 0.26207560300827026, "learning_rate": 4.052299434208618e-05, "loss": 0.0207, "step": 16499 }, { "epoch": 1.7888117953165654, "grad_norm": 0.3194722831249237, "learning_rate": 4.051936747424924e-05, "loss": 0.0216, "step": 16500 }, { "epoch": 1.7889202081526454, "grad_norm": 0.2625318765640259, "learning_rate": 4.0515740606412305e-05, "loss": 0.0315, "step": 16501 }, { "epoch": 1.789028620988725, "grad_norm": 0.0786290243268013, "learning_rate": 4.051211373857537e-05, "loss": 0.0088, "step": 16502 }, { "epoch": 1.789137033824805, "grad_norm": 0.21986940503120422, "learning_rate": 4.050848687073843e-05, "loss": 0.024, "step": 16503 }, { "epoch": 1.7892454466608847, "grad_norm": 0.0649070143699646, "learning_rate": 4.05048600029015e-05, "loss": 0.0036, "step": 16504 }, { "epoch": 1.7893538594969645, "grad_norm": 0.20544783771038055, "learning_rate": 4.050123313506456e-05, "loss": 0.0244, "step": 16505 }, { "epoch": 1.7894622723330442, "grad_norm": 0.20692329108715057, "learning_rate": 4.0497606267227626e-05, "loss": 0.0067, "step": 16506 }, { "epoch": 1.789570685169124, "grad_norm": 0.1173543855547905, "learning_rate": 4.049397939939068e-05, "loss": 0.011, "step": 16507 }, { "epoch": 1.789679098005204, "grad_norm": 0.30242374539375305, "learning_rate": 4.0490352531553754e-05, "loss": 0.037, "step": 16508 }, { "epoch": 1.7897875108412835, "grad_norm": 0.10002423822879791, "learning_rate": 4.048672566371682e-05, "loss": 0.006, "step": 16509 }, { "epoch": 1.7898959236773635, "grad_norm": 0.2479059398174286, "learning_rate": 4.048309879587988e-05, "loss": 0.0227, "step": 16510 }, { "epoch": 1.790004336513443, "grad_norm": 0.18822504580020905, "learning_rate": 4.0479471928042946e-05, "loss": 0.016, "step": 16511 }, { "epoch": 1.790112749349523, "grad_norm": 0.12710854411125183, "learning_rate": 4.047584506020601e-05, "loss": 0.0092, "step": 16512 }, { "epoch": 1.7902211621856028, "grad_norm": 0.28906747698783875, "learning_rate": 4.0472218192369074e-05, "loss": 0.0313, "step": 16513 }, { "epoch": 1.7903295750216826, "grad_norm": 0.47068923711776733, "learning_rate": 4.046859132453213e-05, "loss": 0.0265, "step": 16514 }, { "epoch": 1.7904379878577623, "grad_norm": 0.29776349663734436, "learning_rate": 4.0464964456695196e-05, "loss": 0.0385, "step": 16515 }, { "epoch": 1.790546400693842, "grad_norm": 0.35247305035591125, "learning_rate": 4.046133758885826e-05, "loss": 0.0209, "step": 16516 }, { "epoch": 1.790654813529922, "grad_norm": 0.14705975353717804, "learning_rate": 4.045771072102133e-05, "loss": 0.0129, "step": 16517 }, { "epoch": 1.7907632263660016, "grad_norm": 0.07901579886674881, "learning_rate": 4.0454083853184395e-05, "loss": 0.0112, "step": 16518 }, { "epoch": 1.7908716392020816, "grad_norm": 0.6002762317657471, "learning_rate": 4.045045698534746e-05, "loss": 0.036, "step": 16519 }, { "epoch": 1.7909800520381614, "grad_norm": 0.33207806944847107, "learning_rate": 4.044683011751052e-05, "loss": 0.0417, "step": 16520 }, { "epoch": 1.7910884648742411, "grad_norm": 0.15373945236206055, "learning_rate": 4.044320324967358e-05, "loss": 0.0138, "step": 16521 }, { "epoch": 1.7911968777103209, "grad_norm": 0.5621877312660217, "learning_rate": 4.0439576381836645e-05, "loss": 0.0228, "step": 16522 }, { "epoch": 1.7913052905464006, "grad_norm": 0.5658290982246399, "learning_rate": 4.043594951399971e-05, "loss": 0.0445, "step": 16523 }, { "epoch": 1.7914137033824806, "grad_norm": 0.39084944128990173, "learning_rate": 4.043232264616277e-05, "loss": 0.0193, "step": 16524 }, { "epoch": 1.7915221162185602, "grad_norm": 0.35047921538352966, "learning_rate": 4.0428695778325844e-05, "loss": 0.0409, "step": 16525 }, { "epoch": 1.7916305290546402, "grad_norm": 0.32822853326797485, "learning_rate": 4.042506891048891e-05, "loss": 0.042, "step": 16526 }, { "epoch": 1.79173894189072, "grad_norm": 0.28385332226753235, "learning_rate": 4.042144204265197e-05, "loss": 0.0341, "step": 16527 }, { "epoch": 1.7918473547267997, "grad_norm": 0.19088205695152283, "learning_rate": 4.041781517481503e-05, "loss": 0.0127, "step": 16528 }, { "epoch": 1.7919557675628794, "grad_norm": 0.20608045160770416, "learning_rate": 4.041418830697809e-05, "loss": 0.0129, "step": 16529 }, { "epoch": 1.7920641803989592, "grad_norm": 0.46773627400398254, "learning_rate": 4.041056143914116e-05, "loss": 0.0263, "step": 16530 }, { "epoch": 1.7921725932350392, "grad_norm": 0.2292139083147049, "learning_rate": 4.040693457130422e-05, "loss": 0.0143, "step": 16531 }, { "epoch": 1.7922810060711187, "grad_norm": 0.23553913831710815, "learning_rate": 4.0403307703467286e-05, "loss": 0.0142, "step": 16532 }, { "epoch": 1.7923894189071987, "grad_norm": 0.9987824559211731, "learning_rate": 4.039968083563035e-05, "loss": 0.0353, "step": 16533 }, { "epoch": 1.7924978317432783, "grad_norm": 0.11723382025957108, "learning_rate": 4.039605396779342e-05, "loss": 0.0127, "step": 16534 }, { "epoch": 1.7926062445793582, "grad_norm": 0.4261508285999298, "learning_rate": 4.0392427099956485e-05, "loss": 0.0428, "step": 16535 }, { "epoch": 1.792714657415438, "grad_norm": 0.3232575058937073, "learning_rate": 4.038880023211954e-05, "loss": 0.0262, "step": 16536 }, { "epoch": 1.7928230702515178, "grad_norm": 0.2995040714740753, "learning_rate": 4.0385173364282606e-05, "loss": 0.0126, "step": 16537 }, { "epoch": 1.7929314830875975, "grad_norm": 0.4478210210800171, "learning_rate": 4.038154649644567e-05, "loss": 0.0276, "step": 16538 }, { "epoch": 1.7930398959236773, "grad_norm": 0.9747508764266968, "learning_rate": 4.0377919628608734e-05, "loss": 0.0304, "step": 16539 }, { "epoch": 1.7931483087597573, "grad_norm": 0.21641886234283447, "learning_rate": 4.03742927607718e-05, "loss": 0.0246, "step": 16540 }, { "epoch": 1.7932567215958368, "grad_norm": 0.3681887090206146, "learning_rate": 4.037066589293486e-05, "loss": 0.0642, "step": 16541 }, { "epoch": 1.7933651344319168, "grad_norm": 0.37716859579086304, "learning_rate": 4.036703902509793e-05, "loss": 0.0421, "step": 16542 }, { "epoch": 1.7934735472679966, "grad_norm": 0.23531575500965118, "learning_rate": 4.036341215726099e-05, "loss": 0.0261, "step": 16543 }, { "epoch": 1.7935819601040763, "grad_norm": 0.41666194796562195, "learning_rate": 4.0359785289424055e-05, "loss": 0.0466, "step": 16544 }, { "epoch": 1.793690372940156, "grad_norm": 0.19648773968219757, "learning_rate": 4.035615842158712e-05, "loss": 0.01, "step": 16545 }, { "epoch": 1.7937987857762359, "grad_norm": 0.1215747594833374, "learning_rate": 4.035253155375018e-05, "loss": 0.0168, "step": 16546 }, { "epoch": 1.7939071986123158, "grad_norm": 0.36153480410575867, "learning_rate": 4.034890468591325e-05, "loss": 0.0513, "step": 16547 }, { "epoch": 1.7940156114483954, "grad_norm": 0.5093486905097961, "learning_rate": 4.034527781807631e-05, "loss": 0.0297, "step": 16548 }, { "epoch": 1.7941240242844754, "grad_norm": 0.2598316967487335, "learning_rate": 4.0341650950239375e-05, "loss": 0.0337, "step": 16549 }, { "epoch": 1.794232437120555, "grad_norm": 1.1322827339172363, "learning_rate": 4.033802408240244e-05, "loss": 0.0794, "step": 16550 }, { "epoch": 1.794340849956635, "grad_norm": 0.14216461777687073, "learning_rate": 4.0334397214565504e-05, "loss": 0.0079, "step": 16551 }, { "epoch": 1.7944492627927147, "grad_norm": 0.31054410338401794, "learning_rate": 4.033077034672857e-05, "loss": 0.028, "step": 16552 }, { "epoch": 1.7945576756287944, "grad_norm": 0.19635549187660217, "learning_rate": 4.032714347889163e-05, "loss": 0.0121, "step": 16553 }, { "epoch": 1.7946660884648742, "grad_norm": 0.15120314061641693, "learning_rate": 4.0323516611054696e-05, "loss": 0.0057, "step": 16554 }, { "epoch": 1.794774501300954, "grad_norm": 0.07320479303598404, "learning_rate": 4.031988974321776e-05, "loss": 0.0045, "step": 16555 }, { "epoch": 1.794882914137034, "grad_norm": 0.31427115201950073, "learning_rate": 4.0316262875380824e-05, "loss": 0.016, "step": 16556 }, { "epoch": 1.7949913269731135, "grad_norm": 0.23360632359981537, "learning_rate": 4.031263600754389e-05, "loss": 0.0134, "step": 16557 }, { "epoch": 1.7950997398091935, "grad_norm": 0.17550891637802124, "learning_rate": 4.030900913970695e-05, "loss": 0.0119, "step": 16558 }, { "epoch": 1.7952081526452732, "grad_norm": 0.19556549191474915, "learning_rate": 4.030538227187001e-05, "loss": 0.0167, "step": 16559 }, { "epoch": 1.795316565481353, "grad_norm": 0.06728808581829071, "learning_rate": 4.030175540403308e-05, "loss": 0.0062, "step": 16560 }, { "epoch": 1.7954249783174328, "grad_norm": 0.5156856775283813, "learning_rate": 4.0298128536196145e-05, "loss": 0.0279, "step": 16561 }, { "epoch": 1.7955333911535125, "grad_norm": 0.5164905786514282, "learning_rate": 4.029450166835921e-05, "loss": 0.0331, "step": 16562 }, { "epoch": 1.7956418039895925, "grad_norm": 0.22572539746761322, "learning_rate": 4.029087480052227e-05, "loss": 0.0087, "step": 16563 }, { "epoch": 1.795750216825672, "grad_norm": 0.2298721820116043, "learning_rate": 4.028724793268534e-05, "loss": 0.0215, "step": 16564 }, { "epoch": 1.795858629661752, "grad_norm": 0.05647287517786026, "learning_rate": 4.02836210648484e-05, "loss": 0.0039, "step": 16565 }, { "epoch": 1.7959670424978318, "grad_norm": 0.411013126373291, "learning_rate": 4.027999419701146e-05, "loss": 0.0587, "step": 16566 }, { "epoch": 1.7960754553339116, "grad_norm": 0.25543731451034546, "learning_rate": 4.027636732917452e-05, "loss": 0.0367, "step": 16567 }, { "epoch": 1.7961838681699913, "grad_norm": 0.3805905282497406, "learning_rate": 4.027274046133759e-05, "loss": 0.0449, "step": 16568 }, { "epoch": 1.796292281006071, "grad_norm": 0.5326171517372131, "learning_rate": 4.026911359350066e-05, "loss": 0.0196, "step": 16569 }, { "epoch": 1.796400693842151, "grad_norm": 0.4035870134830475, "learning_rate": 4.026548672566372e-05, "loss": 0.0609, "step": 16570 }, { "epoch": 1.7965091066782306, "grad_norm": 0.2587232291698456, "learning_rate": 4.0261859857826786e-05, "loss": 0.0142, "step": 16571 }, { "epoch": 1.7966175195143106, "grad_norm": 0.2061544507741928, "learning_rate": 4.025823298998985e-05, "loss": 0.0164, "step": 16572 }, { "epoch": 1.7967259323503901, "grad_norm": 0.6173851490020752, "learning_rate": 4.025460612215291e-05, "loss": 0.0267, "step": 16573 }, { "epoch": 1.7968343451864701, "grad_norm": 0.060293734073638916, "learning_rate": 4.025097925431597e-05, "loss": 0.0055, "step": 16574 }, { "epoch": 1.7969427580225499, "grad_norm": 0.22135964035987854, "learning_rate": 4.0247352386479035e-05, "loss": 0.0197, "step": 16575 }, { "epoch": 1.7970511708586296, "grad_norm": 0.48215222358703613, "learning_rate": 4.02437255186421e-05, "loss": 0.0649, "step": 16576 }, { "epoch": 1.7971595836947094, "grad_norm": 0.30755385756492615, "learning_rate": 4.024009865080517e-05, "loss": 0.0356, "step": 16577 }, { "epoch": 1.7972679965307892, "grad_norm": 0.1654900163412094, "learning_rate": 4.0236471782968234e-05, "loss": 0.012, "step": 16578 }, { "epoch": 1.7973764093668692, "grad_norm": 0.40243104100227356, "learning_rate": 4.02328449151313e-05, "loss": 0.0499, "step": 16579 }, { "epoch": 1.7974848222029487, "grad_norm": 0.6868117451667786, "learning_rate": 4.0229218047294356e-05, "loss": 0.0355, "step": 16580 }, { "epoch": 1.7975932350390287, "grad_norm": 0.23411662876605988, "learning_rate": 4.022559117945742e-05, "loss": 0.0136, "step": 16581 }, { "epoch": 1.7977016478751084, "grad_norm": 0.5809473991394043, "learning_rate": 4.0221964311620484e-05, "loss": 0.0376, "step": 16582 }, { "epoch": 1.7978100607111882, "grad_norm": 0.3865000903606415, "learning_rate": 4.021833744378355e-05, "loss": 0.0435, "step": 16583 }, { "epoch": 1.797918473547268, "grad_norm": 0.3524918854236603, "learning_rate": 4.021471057594661e-05, "loss": 0.0201, "step": 16584 }, { "epoch": 1.7980268863833477, "grad_norm": 0.2350056767463684, "learning_rate": 4.0211083708109676e-05, "loss": 0.0255, "step": 16585 }, { "epoch": 1.7981352992194277, "grad_norm": 0.10584986954927444, "learning_rate": 4.020745684027275e-05, "loss": 0.0085, "step": 16586 }, { "epoch": 1.7982437120555073, "grad_norm": 0.08420724421739578, "learning_rate": 4.020382997243581e-05, "loss": 0.0049, "step": 16587 }, { "epoch": 1.7983521248915872, "grad_norm": 1.1579676866531372, "learning_rate": 4.020020310459887e-05, "loss": 0.021, "step": 16588 }, { "epoch": 1.798460537727667, "grad_norm": 0.3239176273345947, "learning_rate": 4.019657623676193e-05, "loss": 0.0293, "step": 16589 }, { "epoch": 1.7985689505637468, "grad_norm": 0.27067914605140686, "learning_rate": 4.0192949368925e-05, "loss": 0.0178, "step": 16590 }, { "epoch": 1.7986773633998265, "grad_norm": 0.25188228487968445, "learning_rate": 4.018932250108806e-05, "loss": 0.0159, "step": 16591 }, { "epoch": 1.7987857762359063, "grad_norm": 0.2572413384914398, "learning_rate": 4.0185695633251125e-05, "loss": 0.0484, "step": 16592 }, { "epoch": 1.7988941890719863, "grad_norm": 0.4662972390651703, "learning_rate": 4.018206876541419e-05, "loss": 0.0491, "step": 16593 }, { "epoch": 1.7990026019080658, "grad_norm": 0.19925610721111298, "learning_rate": 4.017844189757726e-05, "loss": 0.0129, "step": 16594 }, { "epoch": 1.7991110147441458, "grad_norm": 0.3369899094104767, "learning_rate": 4.017481502974032e-05, "loss": 0.0533, "step": 16595 }, { "epoch": 1.7992194275802254, "grad_norm": 0.3232989013195038, "learning_rate": 4.017118816190338e-05, "loss": 0.0408, "step": 16596 }, { "epoch": 1.7993278404163053, "grad_norm": 0.49494725465774536, "learning_rate": 4.0167561294066446e-05, "loss": 0.0146, "step": 16597 }, { "epoch": 1.799436253252385, "grad_norm": 0.30140194296836853, "learning_rate": 4.016393442622951e-05, "loss": 0.0177, "step": 16598 }, { "epoch": 1.7995446660884649, "grad_norm": 0.45183059573173523, "learning_rate": 4.0160307558392574e-05, "loss": 0.026, "step": 16599 }, { "epoch": 1.7996530789245446, "grad_norm": 0.2919561564922333, "learning_rate": 4.015668069055564e-05, "loss": 0.0241, "step": 16600 }, { "epoch": 1.7997614917606244, "grad_norm": 0.23716388642787933, "learning_rate": 4.01530538227187e-05, "loss": 0.0263, "step": 16601 }, { "epoch": 1.7998699045967044, "grad_norm": 0.6164861917495728, "learning_rate": 4.0149426954881766e-05, "loss": 0.0623, "step": 16602 }, { "epoch": 1.799978317432784, "grad_norm": 0.1844029426574707, "learning_rate": 4.014580008704483e-05, "loss": 0.0159, "step": 16603 }, { "epoch": 1.800086730268864, "grad_norm": 0.1558782160282135, "learning_rate": 4.0142173219207894e-05, "loss": 0.0131, "step": 16604 }, { "epoch": 1.8001951431049437, "grad_norm": 0.2810916602611542, "learning_rate": 4.013854635137096e-05, "loss": 0.0254, "step": 16605 }, { "epoch": 1.8003035559410234, "grad_norm": 0.2967909574508667, "learning_rate": 4.013491948353402e-05, "loss": 0.0164, "step": 16606 }, { "epoch": 1.8004119687771032, "grad_norm": 0.07320219278335571, "learning_rate": 4.013129261569709e-05, "loss": 0.0065, "step": 16607 }, { "epoch": 1.800520381613183, "grad_norm": 0.272594690322876, "learning_rate": 4.012766574786015e-05, "loss": 0.0171, "step": 16608 }, { "epoch": 1.800628794449263, "grad_norm": 0.5559766888618469, "learning_rate": 4.0124038880023215e-05, "loss": 0.0417, "step": 16609 }, { "epoch": 1.8007372072853425, "grad_norm": 0.2762863039970398, "learning_rate": 4.012041201218628e-05, "loss": 0.0325, "step": 16610 }, { "epoch": 1.8008456201214225, "grad_norm": 0.09605888277292252, "learning_rate": 4.0116785144349336e-05, "loss": 0.0067, "step": 16611 }, { "epoch": 1.800954032957502, "grad_norm": 0.17063666880130768, "learning_rate": 4.011315827651241e-05, "loss": 0.014, "step": 16612 }, { "epoch": 1.801062445793582, "grad_norm": 0.3001520037651062, "learning_rate": 4.010953140867547e-05, "loss": 0.0261, "step": 16613 }, { "epoch": 1.8011708586296618, "grad_norm": 0.04227324202656746, "learning_rate": 4.0105904540838535e-05, "loss": 0.0023, "step": 16614 }, { "epoch": 1.8012792714657415, "grad_norm": 0.16916193068027496, "learning_rate": 4.01022776730016e-05, "loss": 0.0078, "step": 16615 }, { "epoch": 1.8013876843018215, "grad_norm": 0.06344034522771835, "learning_rate": 4.0098650805164664e-05, "loss": 0.0056, "step": 16616 }, { "epoch": 1.801496097137901, "grad_norm": 0.13171014189720154, "learning_rate": 4.009502393732773e-05, "loss": 0.0063, "step": 16617 }, { "epoch": 1.801604509973981, "grad_norm": 0.5450962781906128, "learning_rate": 4.0091397069490785e-05, "loss": 0.0452, "step": 16618 }, { "epoch": 1.8017129228100606, "grad_norm": 0.35124722123146057, "learning_rate": 4.008777020165385e-05, "loss": 0.0303, "step": 16619 }, { "epoch": 1.8018213356461406, "grad_norm": 0.2666775584220886, "learning_rate": 4.008414333381692e-05, "loss": 0.0245, "step": 16620 }, { "epoch": 1.8019297484822203, "grad_norm": 0.1210307702422142, "learning_rate": 4.0080516465979984e-05, "loss": 0.0163, "step": 16621 }, { "epoch": 1.8020381613183, "grad_norm": 0.5054937601089478, "learning_rate": 4.007688959814305e-05, "loss": 0.033, "step": 16622 }, { "epoch": 1.8021465741543798, "grad_norm": 0.5109474062919617, "learning_rate": 4.007326273030611e-05, "loss": 0.0348, "step": 16623 }, { "epoch": 1.8022549869904596, "grad_norm": 0.19479714334011078, "learning_rate": 4.0069635862469177e-05, "loss": 0.0125, "step": 16624 }, { "epoch": 1.8023633998265396, "grad_norm": 0.21665261685848236, "learning_rate": 4.0066008994632234e-05, "loss": 0.0123, "step": 16625 }, { "epoch": 1.8024718126626191, "grad_norm": 0.3767600357532501, "learning_rate": 4.00623821267953e-05, "loss": 0.0386, "step": 16626 }, { "epoch": 1.8025802254986991, "grad_norm": 0.5846433043479919, "learning_rate": 4.005875525895836e-05, "loss": 0.027, "step": 16627 }, { "epoch": 1.8026886383347789, "grad_norm": 0.058289535343647, "learning_rate": 4.0055128391121426e-05, "loss": 0.0025, "step": 16628 }, { "epoch": 1.8027970511708586, "grad_norm": 0.27655500173568726, "learning_rate": 4.00515015232845e-05, "loss": 0.0324, "step": 16629 }, { "epoch": 1.8029054640069384, "grad_norm": 0.20827588438987732, "learning_rate": 4.004787465544756e-05, "loss": 0.0203, "step": 16630 }, { "epoch": 1.8030138768430182, "grad_norm": 0.10348949581384659, "learning_rate": 4.0044247787610625e-05, "loss": 0.0058, "step": 16631 }, { "epoch": 1.8031222896790982, "grad_norm": 0.29978621006011963, "learning_rate": 4.004062091977368e-05, "loss": 0.0337, "step": 16632 }, { "epoch": 1.8032307025151777, "grad_norm": 0.25879165530204773, "learning_rate": 4.003699405193675e-05, "loss": 0.034, "step": 16633 }, { "epoch": 1.8033391153512577, "grad_norm": 0.12063654512166977, "learning_rate": 4.003336718409981e-05, "loss": 0.009, "step": 16634 }, { "epoch": 1.8034475281873372, "grad_norm": 0.07873953878879547, "learning_rate": 4.0029740316262875e-05, "loss": 0.0069, "step": 16635 }, { "epoch": 1.8035559410234172, "grad_norm": 0.34676775336265564, "learning_rate": 4.002611344842594e-05, "loss": 0.0214, "step": 16636 }, { "epoch": 1.803664353859497, "grad_norm": 0.23988747596740723, "learning_rate": 4.0022486580589e-05, "loss": 0.0112, "step": 16637 }, { "epoch": 1.8037727666955767, "grad_norm": 0.3172520697116852, "learning_rate": 4.0018859712752074e-05, "loss": 0.0384, "step": 16638 }, { "epoch": 1.8038811795316565, "grad_norm": 0.37945932149887085, "learning_rate": 4.001523284491514e-05, "loss": 0.0518, "step": 16639 }, { "epoch": 1.8039895923677363, "grad_norm": 0.2171405851840973, "learning_rate": 4.0011605977078195e-05, "loss": 0.0251, "step": 16640 }, { "epoch": 1.8040980052038162, "grad_norm": 0.15638895332813263, "learning_rate": 4.000797910924126e-05, "loss": 0.0244, "step": 16641 }, { "epoch": 1.8042064180398958, "grad_norm": 0.2179296910762787, "learning_rate": 4.0004352241404324e-05, "loss": 0.0194, "step": 16642 }, { "epoch": 1.8043148308759758, "grad_norm": 0.7985464334487915, "learning_rate": 4.000072537356739e-05, "loss": 0.03, "step": 16643 }, { "epoch": 1.8044232437120555, "grad_norm": 0.18494975566864014, "learning_rate": 3.999709850573045e-05, "loss": 0.0163, "step": 16644 }, { "epoch": 1.8045316565481353, "grad_norm": 0.032704390585422516, "learning_rate": 3.9993471637893516e-05, "loss": 0.002, "step": 16645 }, { "epoch": 1.804640069384215, "grad_norm": 0.2755989134311676, "learning_rate": 3.998984477005659e-05, "loss": 0.0376, "step": 16646 }, { "epoch": 1.8047484822202948, "grad_norm": 0.6452113389968872, "learning_rate": 3.9986217902219644e-05, "loss": 0.0744, "step": 16647 }, { "epoch": 1.8048568950563748, "grad_norm": 0.08854103088378906, "learning_rate": 3.998259103438271e-05, "loss": 0.0048, "step": 16648 }, { "epoch": 1.8049653078924544, "grad_norm": 0.32939058542251587, "learning_rate": 3.997896416654577e-05, "loss": 0.0087, "step": 16649 }, { "epoch": 1.8050737207285343, "grad_norm": 2.0307865142822266, "learning_rate": 3.9975337298708836e-05, "loss": 0.0467, "step": 16650 }, { "epoch": 1.805182133564614, "grad_norm": 0.4824390113353729, "learning_rate": 3.99717104308719e-05, "loss": 0.0251, "step": 16651 }, { "epoch": 1.8052905464006939, "grad_norm": 0.24458280205726624, "learning_rate": 3.9968083563034965e-05, "loss": 0.0183, "step": 16652 }, { "epoch": 1.8053989592367736, "grad_norm": 0.34719452261924744, "learning_rate": 3.996445669519803e-05, "loss": 0.0414, "step": 16653 }, { "epoch": 1.8055073720728534, "grad_norm": 1.1570382118225098, "learning_rate": 3.996082982736109e-05, "loss": 0.048, "step": 16654 }, { "epoch": 1.8056157849089334, "grad_norm": 0.11030618101358414, "learning_rate": 3.995720295952416e-05, "loss": 0.0073, "step": 16655 }, { "epoch": 1.805724197745013, "grad_norm": 0.5314502120018005, "learning_rate": 3.995357609168722e-05, "loss": 0.0512, "step": 16656 }, { "epoch": 1.805832610581093, "grad_norm": 0.7283213138580322, "learning_rate": 3.9949949223850285e-05, "loss": 0.0323, "step": 16657 }, { "epoch": 1.8059410234171724, "grad_norm": 0.5513988733291626, "learning_rate": 3.994632235601335e-05, "loss": 0.0385, "step": 16658 }, { "epoch": 1.8060494362532524, "grad_norm": 0.2998928129673004, "learning_rate": 3.9942695488176413e-05, "loss": 0.0267, "step": 16659 }, { "epoch": 1.8061578490893322, "grad_norm": 0.06130557879805565, "learning_rate": 3.993906862033948e-05, "loss": 0.0026, "step": 16660 }, { "epoch": 1.806266261925412, "grad_norm": 0.4600975215435028, "learning_rate": 3.993544175250254e-05, "loss": 0.0406, "step": 16661 }, { "epoch": 1.8063746747614917, "grad_norm": 0.17566755414009094, "learning_rate": 3.9931814884665606e-05, "loss": 0.011, "step": 16662 }, { "epoch": 1.8064830875975715, "grad_norm": 0.32059553265571594, "learning_rate": 3.992818801682866e-05, "loss": 0.0436, "step": 16663 }, { "epoch": 1.8065915004336515, "grad_norm": 0.2988574802875519, "learning_rate": 3.9924561148991734e-05, "loss": 0.0259, "step": 16664 }, { "epoch": 1.806699913269731, "grad_norm": 0.21136905252933502, "learning_rate": 3.99209342811548e-05, "loss": 0.0255, "step": 16665 }, { "epoch": 1.806808326105811, "grad_norm": 0.31771713495254517, "learning_rate": 3.991730741331786e-05, "loss": 0.0307, "step": 16666 }, { "epoch": 1.8069167389418908, "grad_norm": 0.3270696997642517, "learning_rate": 3.9913680545480926e-05, "loss": 0.0322, "step": 16667 }, { "epoch": 1.8070251517779705, "grad_norm": 0.2780008912086487, "learning_rate": 3.991005367764399e-05, "loss": 0.0309, "step": 16668 }, { "epoch": 1.8071335646140503, "grad_norm": 0.20186246931552887, "learning_rate": 3.9906426809807054e-05, "loss": 0.0166, "step": 16669 }, { "epoch": 1.80724197745013, "grad_norm": 0.17382116615772247, "learning_rate": 3.990279994197011e-05, "loss": 0.0199, "step": 16670 }, { "epoch": 1.80735039028621, "grad_norm": 0.4374968111515045, "learning_rate": 3.9899173074133176e-05, "loss": 0.0151, "step": 16671 }, { "epoch": 1.8074588031222896, "grad_norm": 0.30412212014198303, "learning_rate": 3.989554620629625e-05, "loss": 0.0272, "step": 16672 }, { "epoch": 1.8075672159583696, "grad_norm": 0.46781986951828003, "learning_rate": 3.989191933845931e-05, "loss": 0.0386, "step": 16673 }, { "epoch": 1.807675628794449, "grad_norm": 0.31757476925849915, "learning_rate": 3.9888292470622375e-05, "loss": 0.0417, "step": 16674 }, { "epoch": 1.807784041630529, "grad_norm": 0.09214551001787186, "learning_rate": 3.988466560278544e-05, "loss": 0.0052, "step": 16675 }, { "epoch": 1.8078924544666088, "grad_norm": 0.5165257453918457, "learning_rate": 3.98810387349485e-05, "loss": 0.0354, "step": 16676 }, { "epoch": 1.8080008673026886, "grad_norm": 0.24627824127674103, "learning_rate": 3.987741186711156e-05, "loss": 0.02, "step": 16677 }, { "epoch": 1.8081092801387686, "grad_norm": 0.13539469242095947, "learning_rate": 3.9873784999274625e-05, "loss": 0.0086, "step": 16678 }, { "epoch": 1.8082176929748481, "grad_norm": 0.01014524046331644, "learning_rate": 3.987015813143769e-05, "loss": 0.0006, "step": 16679 }, { "epoch": 1.8083261058109281, "grad_norm": 0.06511546671390533, "learning_rate": 3.986653126360075e-05, "loss": 0.0044, "step": 16680 }, { "epoch": 1.8084345186470077, "grad_norm": 0.33085897564888, "learning_rate": 3.9862904395763824e-05, "loss": 0.0231, "step": 16681 }, { "epoch": 1.8085429314830876, "grad_norm": 0.040609829127788544, "learning_rate": 3.985927752792689e-05, "loss": 0.0016, "step": 16682 }, { "epoch": 1.8086513443191674, "grad_norm": 0.44667673110961914, "learning_rate": 3.985565066008995e-05, "loss": 0.0184, "step": 16683 }, { "epoch": 1.8087597571552472, "grad_norm": 0.21952039003372192, "learning_rate": 3.985202379225301e-05, "loss": 0.0211, "step": 16684 }, { "epoch": 1.808868169991327, "grad_norm": 0.4213452637195587, "learning_rate": 3.984839692441607e-05, "loss": 0.0527, "step": 16685 }, { "epoch": 1.8089765828274067, "grad_norm": 0.20441921055316925, "learning_rate": 3.984477005657914e-05, "loss": 0.0193, "step": 16686 }, { "epoch": 1.8090849956634867, "grad_norm": 0.13140331208705902, "learning_rate": 3.98411431887422e-05, "loss": 0.0117, "step": 16687 }, { "epoch": 1.8091934084995662, "grad_norm": 0.40358567237854004, "learning_rate": 3.9837516320905266e-05, "loss": 0.0692, "step": 16688 }, { "epoch": 1.8093018213356462, "grad_norm": 0.27640217542648315, "learning_rate": 3.9833889453068337e-05, "loss": 0.0202, "step": 16689 }, { "epoch": 1.809410234171726, "grad_norm": 0.3417545557022095, "learning_rate": 3.98302625852314e-05, "loss": 0.0299, "step": 16690 }, { "epoch": 1.8095186470078057, "grad_norm": 0.4305385947227478, "learning_rate": 3.9826635717394465e-05, "loss": 0.03, "step": 16691 }, { "epoch": 1.8096270598438855, "grad_norm": 0.1145213171839714, "learning_rate": 3.982300884955752e-05, "loss": 0.0058, "step": 16692 }, { "epoch": 1.8097354726799653, "grad_norm": 0.0915602445602417, "learning_rate": 3.9819381981720586e-05, "loss": 0.0045, "step": 16693 }, { "epoch": 1.8098438855160452, "grad_norm": 0.3980706036090851, "learning_rate": 3.981575511388365e-05, "loss": 0.0338, "step": 16694 }, { "epoch": 1.8099522983521248, "grad_norm": 0.2260928750038147, "learning_rate": 3.9812128246046714e-05, "loss": 0.0218, "step": 16695 }, { "epoch": 1.8100607111882048, "grad_norm": 0.2757438123226166, "learning_rate": 3.980850137820978e-05, "loss": 0.0262, "step": 16696 }, { "epoch": 1.8101691240242843, "grad_norm": 0.5261141657829285, "learning_rate": 3.980487451037284e-05, "loss": 0.0303, "step": 16697 }, { "epoch": 1.8102775368603643, "grad_norm": 0.13430236279964447, "learning_rate": 3.9801247642535914e-05, "loss": 0.0095, "step": 16698 }, { "epoch": 1.810385949696444, "grad_norm": 0.056765757501125336, "learning_rate": 3.979762077469897e-05, "loss": 0.0032, "step": 16699 }, { "epoch": 1.8104943625325238, "grad_norm": 0.19783629477024078, "learning_rate": 3.9793993906862035e-05, "loss": 0.0101, "step": 16700 }, { "epoch": 1.8106027753686036, "grad_norm": 0.1939166784286499, "learning_rate": 3.97903670390251e-05, "loss": 0.0246, "step": 16701 }, { "epoch": 1.8107111882046834, "grad_norm": 0.43712282180786133, "learning_rate": 3.978674017118816e-05, "loss": 0.0143, "step": 16702 }, { "epoch": 1.8108196010407633, "grad_norm": 0.14614476263523102, "learning_rate": 3.978311330335123e-05, "loss": 0.0095, "step": 16703 }, { "epoch": 1.8109280138768429, "grad_norm": 0.10844554007053375, "learning_rate": 3.977948643551429e-05, "loss": 0.0077, "step": 16704 }, { "epoch": 1.8110364267129229, "grad_norm": 0.34772732853889465, "learning_rate": 3.9775859567677355e-05, "loss": 0.0169, "step": 16705 }, { "epoch": 1.8111448395490026, "grad_norm": 0.41731178760528564, "learning_rate": 3.977223269984042e-05, "loss": 0.0184, "step": 16706 }, { "epoch": 1.8112532523850824, "grad_norm": 0.24979861080646515, "learning_rate": 3.9768605832003484e-05, "loss": 0.0205, "step": 16707 }, { "epoch": 1.8113616652211622, "grad_norm": 0.20595476031303406, "learning_rate": 3.976497896416655e-05, "loss": 0.019, "step": 16708 }, { "epoch": 1.811470078057242, "grad_norm": 0.08783644437789917, "learning_rate": 3.976135209632961e-05, "loss": 0.0084, "step": 16709 }, { "epoch": 1.811578490893322, "grad_norm": 0.2508951425552368, "learning_rate": 3.9757725228492676e-05, "loss": 0.0284, "step": 16710 }, { "epoch": 1.8116869037294014, "grad_norm": 0.12618689239025116, "learning_rate": 3.975409836065574e-05, "loss": 0.0163, "step": 16711 }, { "epoch": 1.8117953165654814, "grad_norm": 0.0996987521648407, "learning_rate": 3.9750471492818804e-05, "loss": 0.0036, "step": 16712 }, { "epoch": 1.8119037294015612, "grad_norm": 0.2078583836555481, "learning_rate": 3.974684462498187e-05, "loss": 0.0238, "step": 16713 }, { "epoch": 1.812012142237641, "grad_norm": 0.427926242351532, "learning_rate": 3.974321775714493e-05, "loss": 0.0108, "step": 16714 }, { "epoch": 1.8121205550737207, "grad_norm": 0.44841650128364563, "learning_rate": 3.9739590889307997e-05, "loss": 0.019, "step": 16715 }, { "epoch": 1.8122289679098005, "grad_norm": 0.40860721468925476, "learning_rate": 3.973596402147106e-05, "loss": 0.0333, "step": 16716 }, { "epoch": 1.8123373807458805, "grad_norm": 0.2127213478088379, "learning_rate": 3.9732337153634125e-05, "loss": 0.0153, "step": 16717 }, { "epoch": 1.81244579358196, "grad_norm": 0.22697626054286957, "learning_rate": 3.972871028579719e-05, "loss": 0.0169, "step": 16718 }, { "epoch": 1.81255420641804, "grad_norm": 0.1252700537443161, "learning_rate": 3.972508341796025e-05, "loss": 0.0083, "step": 16719 }, { "epoch": 1.8126626192541195, "grad_norm": 0.14359694719314575, "learning_rate": 3.972145655012332e-05, "loss": 0.0117, "step": 16720 }, { "epoch": 1.8127710320901995, "grad_norm": 0.2753995358943939, "learning_rate": 3.971782968228638e-05, "loss": 0.0053, "step": 16721 }, { "epoch": 1.8128794449262793, "grad_norm": 0.5531616806983948, "learning_rate": 3.971420281444944e-05, "loss": 0.0302, "step": 16722 }, { "epoch": 1.812987857762359, "grad_norm": 0.08172830939292908, "learning_rate": 3.97105759466125e-05, "loss": 0.0024, "step": 16723 }, { "epoch": 1.8130962705984388, "grad_norm": 0.4881845712661743, "learning_rate": 3.9706949078775573e-05, "loss": 0.0288, "step": 16724 }, { "epoch": 1.8132046834345186, "grad_norm": 0.12777872383594513, "learning_rate": 3.970332221093864e-05, "loss": 0.0156, "step": 16725 }, { "epoch": 1.8133130962705986, "grad_norm": 0.15779978036880493, "learning_rate": 3.96996953431017e-05, "loss": 0.0059, "step": 16726 }, { "epoch": 1.813421509106678, "grad_norm": 0.6052553057670593, "learning_rate": 3.9696068475264766e-05, "loss": 0.0494, "step": 16727 }, { "epoch": 1.813529921942758, "grad_norm": 0.433937132358551, "learning_rate": 3.969244160742783e-05, "loss": 0.0168, "step": 16728 }, { "epoch": 1.8136383347788378, "grad_norm": 0.2618083655834198, "learning_rate": 3.968881473959089e-05, "loss": 0.0182, "step": 16729 }, { "epoch": 1.8137467476149176, "grad_norm": 0.3085155189037323, "learning_rate": 3.968518787175395e-05, "loss": 0.009, "step": 16730 }, { "epoch": 1.8138551604509974, "grad_norm": 0.2745155990123749, "learning_rate": 3.9681561003917015e-05, "loss": 0.052, "step": 16731 }, { "epoch": 1.8139635732870771, "grad_norm": 0.15579326450824738, "learning_rate": 3.967793413608008e-05, "loss": 0.0042, "step": 16732 }, { "epoch": 1.8140719861231571, "grad_norm": 0.29161790013313293, "learning_rate": 3.967430726824315e-05, "loss": 0.0378, "step": 16733 }, { "epoch": 1.8141803989592367, "grad_norm": 0.07366839051246643, "learning_rate": 3.9670680400406215e-05, "loss": 0.0041, "step": 16734 }, { "epoch": 1.8142888117953166, "grad_norm": 0.23847225308418274, "learning_rate": 3.966705353256928e-05, "loss": 0.0224, "step": 16735 }, { "epoch": 1.8143972246313964, "grad_norm": 0.5301364660263062, "learning_rate": 3.966342666473234e-05, "loss": 0.0111, "step": 16736 }, { "epoch": 1.8145056374674762, "grad_norm": 0.5703951716423035, "learning_rate": 3.96597997968954e-05, "loss": 0.053, "step": 16737 }, { "epoch": 1.814614050303556, "grad_norm": 0.12172641605138779, "learning_rate": 3.9656172929058464e-05, "loss": 0.0089, "step": 16738 }, { "epoch": 1.8147224631396357, "grad_norm": 0.08276782184839249, "learning_rate": 3.965254606122153e-05, "loss": 0.0036, "step": 16739 }, { "epoch": 1.8148308759757157, "grad_norm": 0.12901194393634796, "learning_rate": 3.964891919338459e-05, "loss": 0.0181, "step": 16740 }, { "epoch": 1.8149392888117952, "grad_norm": 0.10474339127540588, "learning_rate": 3.964529232554766e-05, "loss": 0.0049, "step": 16741 }, { "epoch": 1.8150477016478752, "grad_norm": 0.0489133782684803, "learning_rate": 3.964166545771073e-05, "loss": 0.0014, "step": 16742 }, { "epoch": 1.8151561144839548, "grad_norm": 0.29034504294395447, "learning_rate": 3.963803858987379e-05, "loss": 0.008, "step": 16743 }, { "epoch": 1.8152645273200347, "grad_norm": 0.10478394478559494, "learning_rate": 3.963441172203685e-05, "loss": 0.004, "step": 16744 }, { "epoch": 1.8153729401561145, "grad_norm": 0.03860616683959961, "learning_rate": 3.963078485419991e-05, "loss": 0.0016, "step": 16745 }, { "epoch": 1.8154813529921943, "grad_norm": 0.10319705307483673, "learning_rate": 3.962715798636298e-05, "loss": 0.0041, "step": 16746 }, { "epoch": 1.815589765828274, "grad_norm": 0.3748658299446106, "learning_rate": 3.962353111852604e-05, "loss": 0.0275, "step": 16747 }, { "epoch": 1.8156981786643538, "grad_norm": 0.13010895252227783, "learning_rate": 3.9619904250689105e-05, "loss": 0.0079, "step": 16748 }, { "epoch": 1.8158065915004338, "grad_norm": 0.06839603185653687, "learning_rate": 3.961627738285217e-05, "loss": 0.0022, "step": 16749 }, { "epoch": 1.8159150043365133, "grad_norm": 0.379477858543396, "learning_rate": 3.961265051501524e-05, "loss": 0.0233, "step": 16750 }, { "epoch": 1.8160234171725933, "grad_norm": 0.2671823799610138, "learning_rate": 3.96090236471783e-05, "loss": 0.0144, "step": 16751 }, { "epoch": 1.816131830008673, "grad_norm": 0.29449263215065, "learning_rate": 3.960539677934136e-05, "loss": 0.0264, "step": 16752 }, { "epoch": 1.8162402428447528, "grad_norm": 0.2834717631340027, "learning_rate": 3.9601769911504426e-05, "loss": 0.0066, "step": 16753 }, { "epoch": 1.8163486556808326, "grad_norm": 0.12718448042869568, "learning_rate": 3.959814304366749e-05, "loss": 0.0117, "step": 16754 }, { "epoch": 1.8164570685169124, "grad_norm": 0.041575897485017776, "learning_rate": 3.9594516175830554e-05, "loss": 0.0019, "step": 16755 }, { "epoch": 1.8165654813529923, "grad_norm": 0.3078911006450653, "learning_rate": 3.959088930799362e-05, "loss": 0.0127, "step": 16756 }, { "epoch": 1.8166738941890719, "grad_norm": 0.5423623919487, "learning_rate": 3.958726244015668e-05, "loss": 0.024, "step": 16757 }, { "epoch": 1.8167823070251519, "grad_norm": 0.11145256459712982, "learning_rate": 3.9583635572319746e-05, "loss": 0.0083, "step": 16758 }, { "epoch": 1.8168907198612314, "grad_norm": 0.4338279962539673, "learning_rate": 3.958000870448281e-05, "loss": 0.019, "step": 16759 }, { "epoch": 1.8169991326973114, "grad_norm": 0.19274236261844635, "learning_rate": 3.9576381836645874e-05, "loss": 0.0114, "step": 16760 }, { "epoch": 1.8171075455333912, "grad_norm": 0.0914609357714653, "learning_rate": 3.957275496880894e-05, "loss": 0.005, "step": 16761 }, { "epoch": 1.817215958369471, "grad_norm": 0.18131980299949646, "learning_rate": 3.9569128100972e-05, "loss": 0.0162, "step": 16762 }, { "epoch": 1.817324371205551, "grad_norm": 0.45811018347740173, "learning_rate": 3.956550123313507e-05, "loss": 0.0341, "step": 16763 }, { "epoch": 1.8174327840416304, "grad_norm": 0.3545035123825073, "learning_rate": 3.956187436529813e-05, "loss": 0.024, "step": 16764 }, { "epoch": 1.8175411968777104, "grad_norm": 0.5716952085494995, "learning_rate": 3.9558247497461195e-05, "loss": 0.0475, "step": 16765 }, { "epoch": 1.81764960971379, "grad_norm": 0.49014970660209656, "learning_rate": 3.955462062962426e-05, "loss": 0.0263, "step": 16766 }, { "epoch": 1.81775802254987, "grad_norm": 0.16623464226722717, "learning_rate": 3.955099376178732e-05, "loss": 0.0145, "step": 16767 }, { "epoch": 1.8178664353859497, "grad_norm": 0.7258259057998657, "learning_rate": 3.954736689395039e-05, "loss": 0.0406, "step": 16768 }, { "epoch": 1.8179748482220295, "grad_norm": 0.6426705718040466, "learning_rate": 3.954374002611345e-05, "loss": 0.0407, "step": 16769 }, { "epoch": 1.8180832610581092, "grad_norm": 0.07712289690971375, "learning_rate": 3.9540113158276516e-05, "loss": 0.0055, "step": 16770 }, { "epoch": 1.818191673894189, "grad_norm": 0.1544926017522812, "learning_rate": 3.953648629043958e-05, "loss": 0.0157, "step": 16771 }, { "epoch": 1.818300086730269, "grad_norm": 0.577601969242096, "learning_rate": 3.9532859422602644e-05, "loss": 0.0582, "step": 16772 }, { "epoch": 1.8184084995663485, "grad_norm": 0.5120119452476501, "learning_rate": 3.952923255476571e-05, "loss": 0.0576, "step": 16773 }, { "epoch": 1.8185169124024285, "grad_norm": 0.391410231590271, "learning_rate": 3.9525605686928765e-05, "loss": 0.0353, "step": 16774 }, { "epoch": 1.8186253252385083, "grad_norm": 0.1471249759197235, "learning_rate": 3.952197881909183e-05, "loss": 0.011, "step": 16775 }, { "epoch": 1.818733738074588, "grad_norm": 0.18007215857505798, "learning_rate": 3.95183519512549e-05, "loss": 0.0053, "step": 16776 }, { "epoch": 1.8188421509106678, "grad_norm": 0.25645408034324646, "learning_rate": 3.9514725083417964e-05, "loss": 0.0115, "step": 16777 }, { "epoch": 1.8189505637467476, "grad_norm": 0.34256431460380554, "learning_rate": 3.951109821558103e-05, "loss": 0.0259, "step": 16778 }, { "epoch": 1.8190589765828276, "grad_norm": 0.10007838904857635, "learning_rate": 3.950747134774409e-05, "loss": 0.0054, "step": 16779 }, { "epoch": 1.819167389418907, "grad_norm": 0.582176148891449, "learning_rate": 3.9503844479907157e-05, "loss": 0.0296, "step": 16780 }, { "epoch": 1.819275802254987, "grad_norm": 0.21532045304775238, "learning_rate": 3.9500217612070214e-05, "loss": 0.0086, "step": 16781 }, { "epoch": 1.8193842150910666, "grad_norm": 0.17423270642757416, "learning_rate": 3.949659074423328e-05, "loss": 0.0136, "step": 16782 }, { "epoch": 1.8194926279271466, "grad_norm": 0.7569007277488708, "learning_rate": 3.949296387639634e-05, "loss": 0.0588, "step": 16783 }, { "epoch": 1.8196010407632264, "grad_norm": 0.49408677220344543, "learning_rate": 3.948933700855941e-05, "loss": 0.0155, "step": 16784 }, { "epoch": 1.8197094535993061, "grad_norm": 0.02577088214457035, "learning_rate": 3.948571014072248e-05, "loss": 0.0012, "step": 16785 }, { "epoch": 1.819817866435386, "grad_norm": 0.6970524191856384, "learning_rate": 3.948208327288554e-05, "loss": 0.0787, "step": 16786 }, { "epoch": 1.8199262792714657, "grad_norm": 0.4118928909301758, "learning_rate": 3.9478456405048605e-05, "loss": 0.0404, "step": 16787 }, { "epoch": 1.8200346921075456, "grad_norm": 0.2870473563671112, "learning_rate": 3.947482953721167e-05, "loss": 0.0266, "step": 16788 }, { "epoch": 1.8201431049436252, "grad_norm": 0.18867826461791992, "learning_rate": 3.947120266937473e-05, "loss": 0.0117, "step": 16789 }, { "epoch": 1.8202515177797052, "grad_norm": 0.6375380158424377, "learning_rate": 3.946757580153779e-05, "loss": 0.0164, "step": 16790 }, { "epoch": 1.820359930615785, "grad_norm": 0.3540275990962982, "learning_rate": 3.9463948933700855e-05, "loss": 0.0302, "step": 16791 }, { "epoch": 1.8204683434518647, "grad_norm": 0.8982753157615662, "learning_rate": 3.946032206586392e-05, "loss": 0.0262, "step": 16792 }, { "epoch": 1.8205767562879445, "grad_norm": 0.5149053931236267, "learning_rate": 3.945669519802699e-05, "loss": 0.0587, "step": 16793 }, { "epoch": 1.8206851691240242, "grad_norm": 0.1882481575012207, "learning_rate": 3.9453068330190054e-05, "loss": 0.0122, "step": 16794 }, { "epoch": 1.8207935819601042, "grad_norm": 0.24763554334640503, "learning_rate": 3.944944146235312e-05, "loss": 0.0297, "step": 16795 }, { "epoch": 1.8209019947961838, "grad_norm": 0.16656607389450073, "learning_rate": 3.9445814594516175e-05, "loss": 0.0111, "step": 16796 }, { "epoch": 1.8210104076322637, "grad_norm": 0.11493996530771255, "learning_rate": 3.944218772667924e-05, "loss": 0.0092, "step": 16797 }, { "epoch": 1.8211188204683435, "grad_norm": 0.39827942848205566, "learning_rate": 3.9438560858842304e-05, "loss": 0.0228, "step": 16798 }, { "epoch": 1.8212272333044233, "grad_norm": 0.6328428983688354, "learning_rate": 3.943493399100537e-05, "loss": 0.086, "step": 16799 }, { "epoch": 1.821335646140503, "grad_norm": 0.20035071671009064, "learning_rate": 3.943130712316843e-05, "loss": 0.0186, "step": 16800 }, { "epoch": 1.8214440589765828, "grad_norm": 0.23814071714878082, "learning_rate": 3.9427680255331496e-05, "loss": 0.0306, "step": 16801 }, { "epoch": 1.8215524718126628, "grad_norm": 0.16104449331760406, "learning_rate": 3.942405338749457e-05, "loss": 0.0311, "step": 16802 }, { "epoch": 1.8216608846487423, "grad_norm": 0.15213072299957275, "learning_rate": 3.9420426519657624e-05, "loss": 0.0071, "step": 16803 }, { "epoch": 1.8217692974848223, "grad_norm": 0.2683071196079254, "learning_rate": 3.941679965182069e-05, "loss": 0.0229, "step": 16804 }, { "epoch": 1.8218777103209018, "grad_norm": 0.578829824924469, "learning_rate": 3.941317278398375e-05, "loss": 0.0289, "step": 16805 }, { "epoch": 1.8219861231569818, "grad_norm": 0.18909865617752075, "learning_rate": 3.9409545916146817e-05, "loss": 0.0232, "step": 16806 }, { "epoch": 1.8220945359930616, "grad_norm": 0.43807700276374817, "learning_rate": 3.940591904830988e-05, "loss": 0.0128, "step": 16807 }, { "epoch": 1.8222029488291414, "grad_norm": 0.40343809127807617, "learning_rate": 3.9402292180472945e-05, "loss": 0.0258, "step": 16808 }, { "epoch": 1.8223113616652211, "grad_norm": 0.32397666573524475, "learning_rate": 3.939866531263601e-05, "loss": 0.0154, "step": 16809 }, { "epoch": 1.8224197745013009, "grad_norm": 0.22635556757450104, "learning_rate": 3.939503844479907e-05, "loss": 0.0075, "step": 16810 }, { "epoch": 1.8225281873373809, "grad_norm": 0.27923449873924255, "learning_rate": 3.939141157696214e-05, "loss": 0.0339, "step": 16811 }, { "epoch": 1.8226366001734604, "grad_norm": 0.11732693016529083, "learning_rate": 3.93877847091252e-05, "loss": 0.0125, "step": 16812 }, { "epoch": 1.8227450130095404, "grad_norm": 0.36689531803131104, "learning_rate": 3.9384157841288265e-05, "loss": 0.0266, "step": 16813 }, { "epoch": 1.8228534258456202, "grad_norm": 0.14503473043441772, "learning_rate": 3.938053097345133e-05, "loss": 0.0164, "step": 16814 }, { "epoch": 1.8229618386817, "grad_norm": 0.33006975054740906, "learning_rate": 3.9376904105614393e-05, "loss": 0.0168, "step": 16815 }, { "epoch": 1.8230702515177797, "grad_norm": 0.6858761310577393, "learning_rate": 3.937327723777746e-05, "loss": 0.076, "step": 16816 }, { "epoch": 1.8231786643538594, "grad_norm": 0.10287488996982574, "learning_rate": 3.936965036994052e-05, "loss": 0.0083, "step": 16817 }, { "epoch": 1.8232870771899394, "grad_norm": 0.3208649754524231, "learning_rate": 3.9366023502103586e-05, "loss": 0.0373, "step": 16818 }, { "epoch": 1.823395490026019, "grad_norm": 0.1436658799648285, "learning_rate": 3.936239663426665e-05, "loss": 0.0147, "step": 16819 }, { "epoch": 1.823503902862099, "grad_norm": 0.46285805106163025, "learning_rate": 3.9358769766429714e-05, "loss": 0.0729, "step": 16820 }, { "epoch": 1.8236123156981785, "grad_norm": 0.24824385344982147, "learning_rate": 3.935514289859278e-05, "loss": 0.0362, "step": 16821 }, { "epoch": 1.8237207285342585, "grad_norm": 0.15116123855113983, "learning_rate": 3.935151603075584e-05, "loss": 0.0112, "step": 16822 }, { "epoch": 1.8238291413703382, "grad_norm": 0.4479091763496399, "learning_rate": 3.9347889162918906e-05, "loss": 0.0228, "step": 16823 }, { "epoch": 1.823937554206418, "grad_norm": 0.11211470514535904, "learning_rate": 3.934426229508197e-05, "loss": 0.0172, "step": 16824 }, { "epoch": 1.824045967042498, "grad_norm": 0.07414665818214417, "learning_rate": 3.9340635427245035e-05, "loss": 0.0071, "step": 16825 }, { "epoch": 1.8241543798785775, "grad_norm": 0.11089251190423965, "learning_rate": 3.933700855940809e-05, "loss": 0.0082, "step": 16826 }, { "epoch": 1.8242627927146575, "grad_norm": 0.19972257316112518, "learning_rate": 3.9333381691571156e-05, "loss": 0.0199, "step": 16827 }, { "epoch": 1.824371205550737, "grad_norm": 0.11303733289241791, "learning_rate": 3.932975482373423e-05, "loss": 0.0083, "step": 16828 }, { "epoch": 1.824479618386817, "grad_norm": 0.10406023263931274, "learning_rate": 3.932612795589729e-05, "loss": 0.0092, "step": 16829 }, { "epoch": 1.8245880312228968, "grad_norm": 0.33780717849731445, "learning_rate": 3.9322501088060355e-05, "loss": 0.0315, "step": 16830 }, { "epoch": 1.8246964440589766, "grad_norm": 0.19459255039691925, "learning_rate": 3.931887422022342e-05, "loss": 0.0159, "step": 16831 }, { "epoch": 1.8248048568950563, "grad_norm": 0.23330160975456238, "learning_rate": 3.931524735238648e-05, "loss": 0.0113, "step": 16832 }, { "epoch": 1.824913269731136, "grad_norm": 0.14152169227600098, "learning_rate": 3.931162048454954e-05, "loss": 0.0158, "step": 16833 }, { "epoch": 1.825021682567216, "grad_norm": 0.12238092720508575, "learning_rate": 3.9307993616712605e-05, "loss": 0.0113, "step": 16834 }, { "epoch": 1.8251300954032956, "grad_norm": 0.5955954790115356, "learning_rate": 3.930436674887567e-05, "loss": 0.0338, "step": 16835 }, { "epoch": 1.8252385082393756, "grad_norm": 0.5210239291191101, "learning_rate": 3.930073988103874e-05, "loss": 0.0509, "step": 16836 }, { "epoch": 1.8253469210754554, "grad_norm": 0.13769465684890747, "learning_rate": 3.9297113013201804e-05, "loss": 0.0082, "step": 16837 }, { "epoch": 1.8254553339115351, "grad_norm": 0.2804330289363861, "learning_rate": 3.929348614536487e-05, "loss": 0.013, "step": 16838 }, { "epoch": 1.825563746747615, "grad_norm": 0.14026257395744324, "learning_rate": 3.928985927752793e-05, "loss": 0.0121, "step": 16839 }, { "epoch": 1.8256721595836947, "grad_norm": 0.10311644524335861, "learning_rate": 3.9286232409690996e-05, "loss": 0.0092, "step": 16840 }, { "epoch": 1.8257805724197746, "grad_norm": 0.3590647578239441, "learning_rate": 3.9282605541854053e-05, "loss": 0.0162, "step": 16841 }, { "epoch": 1.8258889852558542, "grad_norm": 0.1820966899394989, "learning_rate": 3.927897867401712e-05, "loss": 0.0048, "step": 16842 }, { "epoch": 1.8259973980919342, "grad_norm": 0.2475934624671936, "learning_rate": 3.927535180618018e-05, "loss": 0.0269, "step": 16843 }, { "epoch": 1.8261058109280137, "grad_norm": 0.056315310299396515, "learning_rate": 3.9271724938343246e-05, "loss": 0.0033, "step": 16844 }, { "epoch": 1.8262142237640937, "grad_norm": 0.5874671936035156, "learning_rate": 3.926809807050632e-05, "loss": 0.0573, "step": 16845 }, { "epoch": 1.8263226366001735, "grad_norm": 0.13005994260311127, "learning_rate": 3.926447120266938e-05, "loss": 0.0173, "step": 16846 }, { "epoch": 1.8264310494362532, "grad_norm": 0.4281409680843353, "learning_rate": 3.9260844334832445e-05, "loss": 0.0168, "step": 16847 }, { "epoch": 1.826539462272333, "grad_norm": 0.13601408898830414, "learning_rate": 3.92572174669955e-05, "loss": 0.0093, "step": 16848 }, { "epoch": 1.8266478751084128, "grad_norm": 0.32607555389404297, "learning_rate": 3.9253590599158566e-05, "loss": 0.0279, "step": 16849 }, { "epoch": 1.8267562879444927, "grad_norm": 0.038102030754089355, "learning_rate": 3.924996373132163e-05, "loss": 0.0023, "step": 16850 }, { "epoch": 1.8268647007805723, "grad_norm": 0.05016442760825157, "learning_rate": 3.9246336863484694e-05, "loss": 0.0031, "step": 16851 }, { "epoch": 1.8269731136166523, "grad_norm": 0.40688735246658325, "learning_rate": 3.924270999564776e-05, "loss": 0.0338, "step": 16852 }, { "epoch": 1.827081526452732, "grad_norm": 0.05182474106550217, "learning_rate": 3.923908312781082e-05, "loss": 0.002, "step": 16853 }, { "epoch": 1.8271899392888118, "grad_norm": 0.2259947508573532, "learning_rate": 3.9235456259973894e-05, "loss": 0.0106, "step": 16854 }, { "epoch": 1.8272983521248916, "grad_norm": 0.42652711272239685, "learning_rate": 3.923182939213695e-05, "loss": 0.0259, "step": 16855 }, { "epoch": 1.8274067649609713, "grad_norm": 0.0934097170829773, "learning_rate": 3.9228202524300015e-05, "loss": 0.0051, "step": 16856 }, { "epoch": 1.8275151777970513, "grad_norm": 0.44043928384780884, "learning_rate": 3.922457565646308e-05, "loss": 0.0256, "step": 16857 }, { "epoch": 1.8276235906331308, "grad_norm": 0.45738688111305237, "learning_rate": 3.922094878862614e-05, "loss": 0.0256, "step": 16858 }, { "epoch": 1.8277320034692108, "grad_norm": 0.20454315841197968, "learning_rate": 3.921732192078921e-05, "loss": 0.0195, "step": 16859 }, { "epoch": 1.8278404163052906, "grad_norm": 0.5264817476272583, "learning_rate": 3.921369505295227e-05, "loss": 0.0798, "step": 16860 }, { "epoch": 1.8279488291413704, "grad_norm": 0.33311253786087036, "learning_rate": 3.9210068185115336e-05, "loss": 0.0251, "step": 16861 }, { "epoch": 1.8280572419774501, "grad_norm": 0.6406940221786499, "learning_rate": 3.92064413172784e-05, "loss": 0.0578, "step": 16862 }, { "epoch": 1.8281656548135299, "grad_norm": 0.07051032781600952, "learning_rate": 3.9202814449441464e-05, "loss": 0.0049, "step": 16863 }, { "epoch": 1.8282740676496099, "grad_norm": 0.1336982101202011, "learning_rate": 3.919918758160453e-05, "loss": 0.0066, "step": 16864 }, { "epoch": 1.8283824804856894, "grad_norm": 0.9970080852508545, "learning_rate": 3.919556071376759e-05, "loss": 0.0303, "step": 16865 }, { "epoch": 1.8284908933217694, "grad_norm": 0.13119688630104065, "learning_rate": 3.9191933845930656e-05, "loss": 0.0139, "step": 16866 }, { "epoch": 1.828599306157849, "grad_norm": 0.15820597112178802, "learning_rate": 3.918830697809372e-05, "loss": 0.0119, "step": 16867 }, { "epoch": 1.828707718993929, "grad_norm": 0.14820316433906555, "learning_rate": 3.9184680110256784e-05, "loss": 0.0045, "step": 16868 }, { "epoch": 1.8288161318300087, "grad_norm": 0.08028605580329895, "learning_rate": 3.918105324241985e-05, "loss": 0.0113, "step": 16869 }, { "epoch": 1.8289245446660884, "grad_norm": 0.04554658755660057, "learning_rate": 3.917742637458291e-05, "loss": 0.0033, "step": 16870 }, { "epoch": 1.8290329575021682, "grad_norm": 0.18645071983337402, "learning_rate": 3.9173799506745977e-05, "loss": 0.0237, "step": 16871 }, { "epoch": 1.829141370338248, "grad_norm": 0.3012373447418213, "learning_rate": 3.917017263890904e-05, "loss": 0.0131, "step": 16872 }, { "epoch": 1.829249783174328, "grad_norm": 0.5474960803985596, "learning_rate": 3.9166545771072105e-05, "loss": 0.0281, "step": 16873 }, { "epoch": 1.8293581960104075, "grad_norm": 0.11834892630577087, "learning_rate": 3.916291890323517e-05, "loss": 0.0085, "step": 16874 }, { "epoch": 1.8294666088464875, "grad_norm": 0.056247927248477936, "learning_rate": 3.915929203539823e-05, "loss": 0.0084, "step": 16875 }, { "epoch": 1.8295750216825672, "grad_norm": 0.4789940416812897, "learning_rate": 3.91556651675613e-05, "loss": 0.0269, "step": 16876 }, { "epoch": 1.829683434518647, "grad_norm": 0.15540996193885803, "learning_rate": 3.915203829972436e-05, "loss": 0.0225, "step": 16877 }, { "epoch": 1.8297918473547268, "grad_norm": 0.4956039488315582, "learning_rate": 3.914841143188742e-05, "loss": 0.0188, "step": 16878 }, { "epoch": 1.8299002601908065, "grad_norm": 0.26590588688850403, "learning_rate": 3.914478456405048e-05, "loss": 0.0197, "step": 16879 }, { "epoch": 1.8300086730268865, "grad_norm": 0.43280228972435, "learning_rate": 3.9141157696213554e-05, "loss": 0.0359, "step": 16880 }, { "epoch": 1.830117085862966, "grad_norm": 0.3031655251979828, "learning_rate": 3.913753082837662e-05, "loss": 0.0375, "step": 16881 }, { "epoch": 1.830225498699046, "grad_norm": 0.21508482098579407, "learning_rate": 3.913390396053968e-05, "loss": 0.0202, "step": 16882 }, { "epoch": 1.8303339115351258, "grad_norm": 0.3386666774749756, "learning_rate": 3.9130277092702746e-05, "loss": 0.0135, "step": 16883 }, { "epoch": 1.8304423243712056, "grad_norm": 0.5784828662872314, "learning_rate": 3.912665022486581e-05, "loss": 0.0279, "step": 16884 }, { "epoch": 1.8305507372072853, "grad_norm": 0.36665263772010803, "learning_rate": 3.912302335702887e-05, "loss": 0.0227, "step": 16885 }, { "epoch": 1.830659150043365, "grad_norm": 0.2776011824607849, "learning_rate": 3.911939648919193e-05, "loss": 0.0133, "step": 16886 }, { "epoch": 1.830767562879445, "grad_norm": 0.16670095920562744, "learning_rate": 3.9115769621354995e-05, "loss": 0.0291, "step": 16887 }, { "epoch": 1.8308759757155246, "grad_norm": 0.37181904911994934, "learning_rate": 3.9112142753518066e-05, "loss": 0.0118, "step": 16888 }, { "epoch": 1.8309843885516046, "grad_norm": 0.2828148603439331, "learning_rate": 3.910851588568113e-05, "loss": 0.0264, "step": 16889 }, { "epoch": 1.8310928013876842, "grad_norm": 0.43832534551620483, "learning_rate": 3.9104889017844195e-05, "loss": 0.0277, "step": 16890 }, { "epoch": 1.8312012142237641, "grad_norm": 0.17329005897045135, "learning_rate": 3.910126215000726e-05, "loss": 0.0128, "step": 16891 }, { "epoch": 1.831309627059844, "grad_norm": 0.11325827986001968, "learning_rate": 3.909763528217032e-05, "loss": 0.006, "step": 16892 }, { "epoch": 1.8314180398959237, "grad_norm": 0.07584553956985474, "learning_rate": 3.909400841433338e-05, "loss": 0.0044, "step": 16893 }, { "epoch": 1.8315264527320034, "grad_norm": 0.3393672704696655, "learning_rate": 3.9090381546496444e-05, "loss": 0.0177, "step": 16894 }, { "epoch": 1.8316348655680832, "grad_norm": 0.321664035320282, "learning_rate": 3.908675467865951e-05, "loss": 0.0153, "step": 16895 }, { "epoch": 1.8317432784041632, "grad_norm": 0.1376570165157318, "learning_rate": 3.908312781082257e-05, "loss": 0.0081, "step": 16896 }, { "epoch": 1.8318516912402427, "grad_norm": 0.4337877631187439, "learning_rate": 3.907950094298564e-05, "loss": 0.0458, "step": 16897 }, { "epoch": 1.8319601040763227, "grad_norm": 0.47116032242774963, "learning_rate": 3.907587407514871e-05, "loss": 0.0387, "step": 16898 }, { "epoch": 1.8320685169124025, "grad_norm": 0.5779018998146057, "learning_rate": 3.907224720731177e-05, "loss": 0.0359, "step": 16899 }, { "epoch": 1.8321769297484822, "grad_norm": 0.1318528801202774, "learning_rate": 3.906862033947483e-05, "loss": 0.0045, "step": 16900 }, { "epoch": 1.832285342584562, "grad_norm": 0.352512925863266, "learning_rate": 3.906499347163789e-05, "loss": 0.0338, "step": 16901 }, { "epoch": 1.8323937554206418, "grad_norm": 0.11620448529720306, "learning_rate": 3.906136660380096e-05, "loss": 0.0048, "step": 16902 }, { "epoch": 1.8325021682567217, "grad_norm": 0.317154198884964, "learning_rate": 3.905773973596402e-05, "loss": 0.0174, "step": 16903 }, { "epoch": 1.8326105810928013, "grad_norm": 0.10361205786466599, "learning_rate": 3.9054112868127085e-05, "loss": 0.0094, "step": 16904 }, { "epoch": 1.8327189939288813, "grad_norm": 0.3647449016571045, "learning_rate": 3.9050486000290156e-05, "loss": 0.025, "step": 16905 }, { "epoch": 1.8328274067649608, "grad_norm": 0.45012763142585754, "learning_rate": 3.904685913245322e-05, "loss": 0.0309, "step": 16906 }, { "epoch": 1.8329358196010408, "grad_norm": 0.17418093979358673, "learning_rate": 3.904323226461628e-05, "loss": 0.0226, "step": 16907 }, { "epoch": 1.8330442324371206, "grad_norm": 0.6226233839988708, "learning_rate": 3.903960539677934e-05, "loss": 0.0269, "step": 16908 }, { "epoch": 1.8331526452732003, "grad_norm": 0.4164799749851227, "learning_rate": 3.9035978528942406e-05, "loss": 0.0191, "step": 16909 }, { "epoch": 1.83326105810928, "grad_norm": 0.567233145236969, "learning_rate": 3.903235166110547e-05, "loss": 0.0571, "step": 16910 }, { "epoch": 1.8333694709453598, "grad_norm": 1.1282449960708618, "learning_rate": 3.9028724793268534e-05, "loss": 0.0593, "step": 16911 }, { "epoch": 1.8334778837814398, "grad_norm": 0.24691016972064972, "learning_rate": 3.90250979254316e-05, "loss": 0.0281, "step": 16912 }, { "epoch": 1.8335862966175194, "grad_norm": 0.12825718522071838, "learning_rate": 3.902147105759466e-05, "loss": 0.0047, "step": 16913 }, { "epoch": 1.8336947094535994, "grad_norm": 0.1398690640926361, "learning_rate": 3.9017844189757726e-05, "loss": 0.02, "step": 16914 }, { "epoch": 1.8338031222896791, "grad_norm": 0.21884335577487946, "learning_rate": 3.901421732192079e-05, "loss": 0.0088, "step": 16915 }, { "epoch": 1.8339115351257589, "grad_norm": 0.1933993697166443, "learning_rate": 3.9010590454083855e-05, "loss": 0.0116, "step": 16916 }, { "epoch": 1.8340199479618386, "grad_norm": 0.6797476410865784, "learning_rate": 3.900696358624692e-05, "loss": 0.0503, "step": 16917 }, { "epoch": 1.8341283607979184, "grad_norm": 0.33697330951690674, "learning_rate": 3.900333671840998e-05, "loss": 0.0572, "step": 16918 }, { "epoch": 1.8342367736339984, "grad_norm": 0.33784034848213196, "learning_rate": 3.899970985057305e-05, "loss": 0.0125, "step": 16919 }, { "epoch": 1.834345186470078, "grad_norm": 0.5479280948638916, "learning_rate": 3.899608298273611e-05, "loss": 0.0358, "step": 16920 }, { "epoch": 1.834453599306158, "grad_norm": 0.39394819736480713, "learning_rate": 3.8992456114899175e-05, "loss": 0.0231, "step": 16921 }, { "epoch": 1.8345620121422377, "grad_norm": 0.12141664326190948, "learning_rate": 3.898882924706224e-05, "loss": 0.0151, "step": 16922 }, { "epoch": 1.8346704249783174, "grad_norm": 0.6894132494926453, "learning_rate": 3.89852023792253e-05, "loss": 0.0636, "step": 16923 }, { "epoch": 1.8347788378143972, "grad_norm": 0.06560681760311127, "learning_rate": 3.898157551138837e-05, "loss": 0.01, "step": 16924 }, { "epoch": 1.834887250650477, "grad_norm": 0.08329031616449356, "learning_rate": 3.897794864355143e-05, "loss": 0.0068, "step": 16925 }, { "epoch": 1.834995663486557, "grad_norm": 0.6315045952796936, "learning_rate": 3.8974321775714496e-05, "loss": 0.0316, "step": 16926 }, { "epoch": 1.8351040763226365, "grad_norm": 0.9765238165855408, "learning_rate": 3.897069490787756e-05, "loss": 0.0357, "step": 16927 }, { "epoch": 1.8352124891587165, "grad_norm": 0.14608784019947052, "learning_rate": 3.8967068040040624e-05, "loss": 0.0098, "step": 16928 }, { "epoch": 1.835320901994796, "grad_norm": 0.2797390818595886, "learning_rate": 3.896344117220369e-05, "loss": 0.0145, "step": 16929 }, { "epoch": 1.835429314830876, "grad_norm": 0.5708838105201721, "learning_rate": 3.8959814304366745e-05, "loss": 0.0085, "step": 16930 }, { "epoch": 1.8355377276669558, "grad_norm": 0.10625546425580978, "learning_rate": 3.8956187436529816e-05, "loss": 0.0077, "step": 16931 }, { "epoch": 1.8356461405030355, "grad_norm": 0.3219009041786194, "learning_rate": 3.895256056869288e-05, "loss": 0.013, "step": 16932 }, { "epoch": 1.8357545533391153, "grad_norm": 0.14894212782382965, "learning_rate": 3.8948933700855944e-05, "loss": 0.0103, "step": 16933 }, { "epoch": 1.835862966175195, "grad_norm": 0.21350735425949097, "learning_rate": 3.894530683301901e-05, "loss": 0.0129, "step": 16934 }, { "epoch": 1.835971379011275, "grad_norm": 0.2343835085630417, "learning_rate": 3.894167996518207e-05, "loss": 0.0264, "step": 16935 }, { "epoch": 1.8360797918473546, "grad_norm": 0.4673430919647217, "learning_rate": 3.893805309734514e-05, "loss": 0.0461, "step": 16936 }, { "epoch": 1.8361882046834346, "grad_norm": 0.23846550285816193, "learning_rate": 3.89344262295082e-05, "loss": 0.0206, "step": 16937 }, { "epoch": 1.8362966175195143, "grad_norm": 0.22856535017490387, "learning_rate": 3.893079936167126e-05, "loss": 0.0154, "step": 16938 }, { "epoch": 1.836405030355594, "grad_norm": 0.4063105881214142, "learning_rate": 3.892717249383432e-05, "loss": 0.0465, "step": 16939 }, { "epoch": 1.8365134431916739, "grad_norm": 0.08378846943378448, "learning_rate": 3.892354562599739e-05, "loss": 0.008, "step": 16940 }, { "epoch": 1.8366218560277536, "grad_norm": 0.24247057735919952, "learning_rate": 3.891991875816046e-05, "loss": 0.0399, "step": 16941 }, { "epoch": 1.8367302688638336, "grad_norm": 0.2085721790790558, "learning_rate": 3.891629189032352e-05, "loss": 0.0168, "step": 16942 }, { "epoch": 1.8368386816999132, "grad_norm": 0.5512065291404724, "learning_rate": 3.8912665022486585e-05, "loss": 0.0493, "step": 16943 }, { "epoch": 1.8369470945359931, "grad_norm": 0.07537294179201126, "learning_rate": 3.890903815464965e-05, "loss": 0.0065, "step": 16944 }, { "epoch": 1.837055507372073, "grad_norm": 0.17327308654785156, "learning_rate": 3.890541128681271e-05, "loss": 0.0043, "step": 16945 }, { "epoch": 1.8371639202081527, "grad_norm": 0.47774720191955566, "learning_rate": 3.890178441897577e-05, "loss": 0.0499, "step": 16946 }, { "epoch": 1.8372723330442324, "grad_norm": 0.13521218299865723, "learning_rate": 3.8898157551138835e-05, "loss": 0.009, "step": 16947 }, { "epoch": 1.8373807458803122, "grad_norm": 0.42588064074516296, "learning_rate": 3.88945306833019e-05, "loss": 0.0499, "step": 16948 }, { "epoch": 1.8374891587163922, "grad_norm": 0.09862373769283295, "learning_rate": 3.889090381546497e-05, "loss": 0.0104, "step": 16949 }, { "epoch": 1.8375975715524717, "grad_norm": 0.2894589900970459, "learning_rate": 3.8887276947628034e-05, "loss": 0.0136, "step": 16950 }, { "epoch": 1.8377059843885517, "grad_norm": 0.7059394717216492, "learning_rate": 3.88836500797911e-05, "loss": 0.0665, "step": 16951 }, { "epoch": 1.8378143972246312, "grad_norm": 0.4448823928833008, "learning_rate": 3.8880023211954156e-05, "loss": 0.0547, "step": 16952 }, { "epoch": 1.8379228100607112, "grad_norm": 0.8043590784072876, "learning_rate": 3.887639634411722e-05, "loss": 0.0408, "step": 16953 }, { "epoch": 1.838031222896791, "grad_norm": 0.04303152859210968, "learning_rate": 3.8872769476280284e-05, "loss": 0.0016, "step": 16954 }, { "epoch": 1.8381396357328708, "grad_norm": 0.05151531472802162, "learning_rate": 3.886914260844335e-05, "loss": 0.0027, "step": 16955 }, { "epoch": 1.8382480485689505, "grad_norm": 0.35304656624794006, "learning_rate": 3.886551574060641e-05, "loss": 0.0143, "step": 16956 }, { "epoch": 1.8383564614050303, "grad_norm": 0.21235986053943634, "learning_rate": 3.886188887276948e-05, "loss": 0.0126, "step": 16957 }, { "epoch": 1.8384648742411103, "grad_norm": 0.33286571502685547, "learning_rate": 3.885826200493255e-05, "loss": 0.0122, "step": 16958 }, { "epoch": 1.8385732870771898, "grad_norm": 0.3397631347179413, "learning_rate": 3.8854635137095604e-05, "loss": 0.0215, "step": 16959 }, { "epoch": 1.8386816999132698, "grad_norm": 0.11157844960689545, "learning_rate": 3.885100826925867e-05, "loss": 0.014, "step": 16960 }, { "epoch": 1.8387901127493496, "grad_norm": 0.34036901593208313, "learning_rate": 3.884738140142173e-05, "loss": 0.0201, "step": 16961 }, { "epoch": 1.8388985255854293, "grad_norm": 0.23889024555683136, "learning_rate": 3.8843754533584797e-05, "loss": 0.0061, "step": 16962 }, { "epoch": 1.839006938421509, "grad_norm": 0.4093300700187683, "learning_rate": 3.884012766574786e-05, "loss": 0.0173, "step": 16963 }, { "epoch": 1.8391153512575888, "grad_norm": 0.5784147381782532, "learning_rate": 3.8836500797910925e-05, "loss": 0.0419, "step": 16964 }, { "epoch": 1.8392237640936688, "grad_norm": 0.22434933483600616, "learning_rate": 3.883287393007399e-05, "loss": 0.0258, "step": 16965 }, { "epoch": 1.8393321769297484, "grad_norm": 0.40911853313446045, "learning_rate": 3.882924706223705e-05, "loss": 0.0228, "step": 16966 }, { "epoch": 1.8394405897658284, "grad_norm": 0.30530279874801636, "learning_rate": 3.882562019440012e-05, "loss": 0.021, "step": 16967 }, { "epoch": 1.839549002601908, "grad_norm": 0.5087613463401794, "learning_rate": 3.882199332656318e-05, "loss": 0.0395, "step": 16968 }, { "epoch": 1.8396574154379879, "grad_norm": 0.2348013073205948, "learning_rate": 3.8818366458726245e-05, "loss": 0.0222, "step": 16969 }, { "epoch": 1.8397658282740676, "grad_norm": 0.2210434079170227, "learning_rate": 3.881473959088931e-05, "loss": 0.0216, "step": 16970 }, { "epoch": 1.8398742411101474, "grad_norm": 0.14150136709213257, "learning_rate": 3.8811112723052374e-05, "loss": 0.0178, "step": 16971 }, { "epoch": 1.8399826539462274, "grad_norm": 0.059588536620140076, "learning_rate": 3.880748585521544e-05, "loss": 0.0051, "step": 16972 }, { "epoch": 1.840091066782307, "grad_norm": 0.6177212595939636, "learning_rate": 3.88038589873785e-05, "loss": 0.0315, "step": 16973 }, { "epoch": 1.840199479618387, "grad_norm": 0.6594489812850952, "learning_rate": 3.8800232119541566e-05, "loss": 0.0225, "step": 16974 }, { "epoch": 1.8403078924544665, "grad_norm": 0.10315611958503723, "learning_rate": 3.879660525170463e-05, "loss": 0.0206, "step": 16975 }, { "epoch": 1.8404163052905465, "grad_norm": 0.37239721417427063, "learning_rate": 3.8792978383867694e-05, "loss": 0.0234, "step": 16976 }, { "epoch": 1.8405247181266262, "grad_norm": 0.6501627564430237, "learning_rate": 3.878935151603076e-05, "loss": 0.0291, "step": 16977 }, { "epoch": 1.840633130962706, "grad_norm": 0.6644601225852966, "learning_rate": 3.878572464819382e-05, "loss": 0.0264, "step": 16978 }, { "epoch": 1.8407415437987857, "grad_norm": 0.32011866569519043, "learning_rate": 3.8782097780356886e-05, "loss": 0.013, "step": 16979 }, { "epoch": 1.8408499566348655, "grad_norm": 0.30903562903404236, "learning_rate": 3.877847091251995e-05, "loss": 0.0142, "step": 16980 }, { "epoch": 1.8409583694709455, "grad_norm": 0.3714653551578522, "learning_rate": 3.8774844044683015e-05, "loss": 0.0245, "step": 16981 }, { "epoch": 1.841066782307025, "grad_norm": 0.3085922598838806, "learning_rate": 3.877121717684607e-05, "loss": 0.0237, "step": 16982 }, { "epoch": 1.841175195143105, "grad_norm": 0.2052944004535675, "learning_rate": 3.876759030900914e-05, "loss": 0.0199, "step": 16983 }, { "epoch": 1.8412836079791848, "grad_norm": 0.48733896017074585, "learning_rate": 3.876396344117221e-05, "loss": 0.0255, "step": 16984 }, { "epoch": 1.8413920208152645, "grad_norm": 0.04710415378212929, "learning_rate": 3.876033657333527e-05, "loss": 0.0034, "step": 16985 }, { "epoch": 1.8415004336513443, "grad_norm": 0.7511078119277954, "learning_rate": 3.8756709705498335e-05, "loss": 0.0577, "step": 16986 }, { "epoch": 1.841608846487424, "grad_norm": 0.2782179117202759, "learning_rate": 3.87530828376614e-05, "loss": 0.0095, "step": 16987 }, { "epoch": 1.841717259323504, "grad_norm": 0.042971476912498474, "learning_rate": 3.874945596982446e-05, "loss": 0.0025, "step": 16988 }, { "epoch": 1.8418256721595836, "grad_norm": 0.4401499629020691, "learning_rate": 3.874582910198753e-05, "loss": 0.0293, "step": 16989 }, { "epoch": 1.8419340849956636, "grad_norm": 0.30227789282798767, "learning_rate": 3.8742202234150585e-05, "loss": 0.032, "step": 16990 }, { "epoch": 1.8420424978317431, "grad_norm": 0.28608059883117676, "learning_rate": 3.873857536631365e-05, "loss": 0.0191, "step": 16991 }, { "epoch": 1.842150910667823, "grad_norm": 0.12305755913257599, "learning_rate": 3.873494849847672e-05, "loss": 0.0076, "step": 16992 }, { "epoch": 1.8422593235039029, "grad_norm": 0.419827938079834, "learning_rate": 3.8731321630639784e-05, "loss": 0.0322, "step": 16993 }, { "epoch": 1.8423677363399826, "grad_norm": 0.4267766773700714, "learning_rate": 3.872769476280285e-05, "loss": 0.0387, "step": 16994 }, { "epoch": 1.8424761491760624, "grad_norm": 0.3402411639690399, "learning_rate": 3.872406789496591e-05, "loss": 0.0498, "step": 16995 }, { "epoch": 1.8425845620121422, "grad_norm": 0.28331097960472107, "learning_rate": 3.8720441027128976e-05, "loss": 0.0216, "step": 16996 }, { "epoch": 1.8426929748482221, "grad_norm": 0.4652650058269501, "learning_rate": 3.8716814159292034e-05, "loss": 0.0333, "step": 16997 }, { "epoch": 1.8428013876843017, "grad_norm": 0.28585147857666016, "learning_rate": 3.87131872914551e-05, "loss": 0.0333, "step": 16998 }, { "epoch": 1.8429098005203817, "grad_norm": 0.18550540506839752, "learning_rate": 3.870956042361816e-05, "loss": 0.0098, "step": 16999 }, { "epoch": 1.8430182133564614, "grad_norm": 0.4891132712364197, "learning_rate": 3.870593355578123e-05, "loss": 0.0493, "step": 17000 }, { "epoch": 1.8431266261925412, "grad_norm": 0.3458371162414551, "learning_rate": 3.87023066879443e-05, "loss": 0.0274, "step": 17001 }, { "epoch": 1.843235039028621, "grad_norm": 0.1805337369441986, "learning_rate": 3.869867982010736e-05, "loss": 0.015, "step": 17002 }, { "epoch": 1.8433434518647007, "grad_norm": 0.26099860668182373, "learning_rate": 3.8695052952270425e-05, "loss": 0.0138, "step": 17003 }, { "epoch": 1.8434518647007807, "grad_norm": 0.41969600319862366, "learning_rate": 3.869142608443348e-05, "loss": 0.0319, "step": 17004 }, { "epoch": 1.8435602775368602, "grad_norm": 0.7119556665420532, "learning_rate": 3.8687799216596546e-05, "loss": 0.0167, "step": 17005 }, { "epoch": 1.8436686903729402, "grad_norm": 0.07341433316469193, "learning_rate": 3.868417234875961e-05, "loss": 0.0077, "step": 17006 }, { "epoch": 1.84377710320902, "grad_norm": 0.6231637597084045, "learning_rate": 3.8680545480922675e-05, "loss": 0.0294, "step": 17007 }, { "epoch": 1.8438855160450998, "grad_norm": 0.18367838859558105, "learning_rate": 3.867691861308574e-05, "loss": 0.006, "step": 17008 }, { "epoch": 1.8439939288811795, "grad_norm": 0.14930713176727295, "learning_rate": 3.867329174524881e-05, "loss": 0.0122, "step": 17009 }, { "epoch": 1.8441023417172593, "grad_norm": 0.02726735919713974, "learning_rate": 3.8669664877411874e-05, "loss": 0.0015, "step": 17010 }, { "epoch": 1.8442107545533393, "grad_norm": 0.1257428526878357, "learning_rate": 3.866603800957493e-05, "loss": 0.0132, "step": 17011 }, { "epoch": 1.8443191673894188, "grad_norm": 0.03525874391198158, "learning_rate": 3.8662411141737995e-05, "loss": 0.0015, "step": 17012 }, { "epoch": 1.8444275802254988, "grad_norm": 0.24249115586280823, "learning_rate": 3.865878427390106e-05, "loss": 0.0206, "step": 17013 }, { "epoch": 1.8445359930615783, "grad_norm": 0.4690660536289215, "learning_rate": 3.865515740606412e-05, "loss": 0.0155, "step": 17014 }, { "epoch": 1.8446444058976583, "grad_norm": 0.12209118902683258, "learning_rate": 3.865153053822719e-05, "loss": 0.0056, "step": 17015 }, { "epoch": 1.844752818733738, "grad_norm": 0.6456024646759033, "learning_rate": 3.864790367039025e-05, "loss": 0.0237, "step": 17016 }, { "epoch": 1.8448612315698178, "grad_norm": 0.3327436149120331, "learning_rate": 3.8644276802553316e-05, "loss": 0.0064, "step": 17017 }, { "epoch": 1.8449696444058976, "grad_norm": 0.13584180176258087, "learning_rate": 3.864064993471638e-05, "loss": 0.005, "step": 17018 }, { "epoch": 1.8450780572419774, "grad_norm": 0.19309288263320923, "learning_rate": 3.8637023066879444e-05, "loss": 0.0099, "step": 17019 }, { "epoch": 1.8451864700780574, "grad_norm": 0.6341803669929504, "learning_rate": 3.863339619904251e-05, "loss": 0.0681, "step": 17020 }, { "epoch": 1.845294882914137, "grad_norm": 0.41287538409233093, "learning_rate": 3.862976933120557e-05, "loss": 0.0226, "step": 17021 }, { "epoch": 1.8454032957502169, "grad_norm": 0.22699496150016785, "learning_rate": 3.8626142463368636e-05, "loss": 0.0096, "step": 17022 }, { "epoch": 1.8455117085862967, "grad_norm": 0.17467479407787323, "learning_rate": 3.86225155955317e-05, "loss": 0.0152, "step": 17023 }, { "epoch": 1.8456201214223764, "grad_norm": 0.362618088722229, "learning_rate": 3.8618888727694764e-05, "loss": 0.0157, "step": 17024 }, { "epoch": 1.8457285342584562, "grad_norm": 0.277782678604126, "learning_rate": 3.861526185985783e-05, "loss": 0.0331, "step": 17025 }, { "epoch": 1.845836947094536, "grad_norm": 0.6441827416419983, "learning_rate": 3.861163499202089e-05, "loss": 0.0467, "step": 17026 }, { "epoch": 1.845945359930616, "grad_norm": 0.32174116373062134, "learning_rate": 3.860800812418396e-05, "loss": 0.0209, "step": 17027 }, { "epoch": 1.8460537727666955, "grad_norm": 0.10273218899965286, "learning_rate": 3.860438125634702e-05, "loss": 0.0124, "step": 17028 }, { "epoch": 1.8461621856027755, "grad_norm": 0.3147560656070709, "learning_rate": 3.8600754388510085e-05, "loss": 0.0272, "step": 17029 }, { "epoch": 1.8462705984388552, "grad_norm": 0.28429561853408813, "learning_rate": 3.859712752067315e-05, "loss": 0.0218, "step": 17030 }, { "epoch": 1.846379011274935, "grad_norm": 0.17494159936904907, "learning_rate": 3.859350065283621e-05, "loss": 0.0107, "step": 17031 }, { "epoch": 1.8464874241110147, "grad_norm": 0.03113745152950287, "learning_rate": 3.858987378499928e-05, "loss": 0.002, "step": 17032 }, { "epoch": 1.8465958369470945, "grad_norm": 0.3390742540359497, "learning_rate": 3.858624691716234e-05, "loss": 0.0247, "step": 17033 }, { "epoch": 1.8467042497831745, "grad_norm": 0.1677999347448349, "learning_rate": 3.85826200493254e-05, "loss": 0.0088, "step": 17034 }, { "epoch": 1.846812662619254, "grad_norm": 0.47844916582107544, "learning_rate": 3.857899318148847e-05, "loss": 0.0341, "step": 17035 }, { "epoch": 1.846921075455334, "grad_norm": 1.0794172286987305, "learning_rate": 3.8575366313651534e-05, "loss": 0.0638, "step": 17036 }, { "epoch": 1.8470294882914136, "grad_norm": 0.11406085640192032, "learning_rate": 3.85717394458146e-05, "loss": 0.0119, "step": 17037 }, { "epoch": 1.8471379011274935, "grad_norm": 0.4452165961265564, "learning_rate": 3.856811257797766e-05, "loss": 0.0325, "step": 17038 }, { "epoch": 1.8472463139635733, "grad_norm": 0.16998650133609772, "learning_rate": 3.8564485710140726e-05, "loss": 0.0562, "step": 17039 }, { "epoch": 1.847354726799653, "grad_norm": 0.05767681822180748, "learning_rate": 3.856085884230379e-05, "loss": 0.0016, "step": 17040 }, { "epoch": 1.8474631396357328, "grad_norm": 0.1384941041469574, "learning_rate": 3.8557231974466854e-05, "loss": 0.0087, "step": 17041 }, { "epoch": 1.8475715524718126, "grad_norm": 0.3127438426017761, "learning_rate": 3.855360510662991e-05, "loss": 0.0228, "step": 17042 }, { "epoch": 1.8476799653078926, "grad_norm": 0.2204083502292633, "learning_rate": 3.8549978238792976e-05, "loss": 0.0216, "step": 17043 }, { "epoch": 1.8477883781439721, "grad_norm": 0.10326404869556427, "learning_rate": 3.8546351370956046e-05, "loss": 0.0072, "step": 17044 }, { "epoch": 1.847896790980052, "grad_norm": 0.7201907634735107, "learning_rate": 3.854272450311911e-05, "loss": 0.0087, "step": 17045 }, { "epoch": 1.8480052038161319, "grad_norm": 0.12496568262577057, "learning_rate": 3.8539097635282175e-05, "loss": 0.0041, "step": 17046 }, { "epoch": 1.8481136166522116, "grad_norm": 0.30781090259552, "learning_rate": 3.853547076744524e-05, "loss": 0.027, "step": 17047 }, { "epoch": 1.8482220294882914, "grad_norm": 0.3438965380191803, "learning_rate": 3.85318438996083e-05, "loss": 0.057, "step": 17048 }, { "epoch": 1.8483304423243712, "grad_norm": 0.3188544809818268, "learning_rate": 3.852821703177136e-05, "loss": 0.0221, "step": 17049 }, { "epoch": 1.8484388551604511, "grad_norm": 0.34869566559791565, "learning_rate": 3.8524590163934424e-05, "loss": 0.0111, "step": 17050 }, { "epoch": 1.8485472679965307, "grad_norm": 0.01323853712528944, "learning_rate": 3.852096329609749e-05, "loss": 0.0008, "step": 17051 }, { "epoch": 1.8486556808326107, "grad_norm": 0.22567231953144073, "learning_rate": 3.851733642826056e-05, "loss": 0.0223, "step": 17052 }, { "epoch": 1.8487640936686902, "grad_norm": 0.2627893090248108, "learning_rate": 3.8513709560423623e-05, "loss": 0.0198, "step": 17053 }, { "epoch": 1.8488725065047702, "grad_norm": 0.3061234652996063, "learning_rate": 3.851008269258669e-05, "loss": 0.0708, "step": 17054 }, { "epoch": 1.84898091934085, "grad_norm": 0.23665623366832733, "learning_rate": 3.850645582474975e-05, "loss": 0.0235, "step": 17055 }, { "epoch": 1.8490893321769297, "grad_norm": 0.2418825775384903, "learning_rate": 3.850282895691281e-05, "loss": 0.0568, "step": 17056 }, { "epoch": 1.8491977450130095, "grad_norm": 0.3574891984462738, "learning_rate": 3.849920208907587e-05, "loss": 0.0321, "step": 17057 }, { "epoch": 1.8493061578490892, "grad_norm": 0.22395288944244385, "learning_rate": 3.849557522123894e-05, "loss": 0.008, "step": 17058 }, { "epoch": 1.8494145706851692, "grad_norm": 0.10869802534580231, "learning_rate": 3.8491948353402e-05, "loss": 0.0093, "step": 17059 }, { "epoch": 1.8495229835212488, "grad_norm": 0.24626219272613525, "learning_rate": 3.8488321485565065e-05, "loss": 0.0235, "step": 17060 }, { "epoch": 1.8496313963573288, "grad_norm": 0.1444898098707199, "learning_rate": 3.8484694617728136e-05, "loss": 0.0087, "step": 17061 }, { "epoch": 1.8497398091934085, "grad_norm": 0.1240636557340622, "learning_rate": 3.84810677498912e-05, "loss": 0.0061, "step": 17062 }, { "epoch": 1.8498482220294883, "grad_norm": 0.7234631776809692, "learning_rate": 3.847744088205426e-05, "loss": 0.0665, "step": 17063 }, { "epoch": 1.849956634865568, "grad_norm": 0.2634883522987366, "learning_rate": 3.847381401421732e-05, "loss": 0.0276, "step": 17064 }, { "epoch": 1.8500650477016478, "grad_norm": 0.28925010561943054, "learning_rate": 3.8470187146380386e-05, "loss": 0.0228, "step": 17065 }, { "epoch": 1.8501734605377278, "grad_norm": 0.2614365816116333, "learning_rate": 3.846656027854345e-05, "loss": 0.059, "step": 17066 }, { "epoch": 1.8502818733738073, "grad_norm": 1.2093473672866821, "learning_rate": 3.8462933410706514e-05, "loss": 0.0524, "step": 17067 }, { "epoch": 1.8503902862098873, "grad_norm": 0.3055815100669861, "learning_rate": 3.845930654286958e-05, "loss": 0.0193, "step": 17068 }, { "epoch": 1.850498699045967, "grad_norm": 0.27247998118400574, "learning_rate": 3.845567967503264e-05, "loss": 0.0109, "step": 17069 }, { "epoch": 1.8506071118820469, "grad_norm": 0.12353764474391937, "learning_rate": 3.845205280719571e-05, "loss": 0.0072, "step": 17070 }, { "epoch": 1.8507155247181266, "grad_norm": 0.09136779606342316, "learning_rate": 3.844842593935877e-05, "loss": 0.0042, "step": 17071 }, { "epoch": 1.8508239375542064, "grad_norm": 0.4182257354259491, "learning_rate": 3.8444799071521835e-05, "loss": 0.0971, "step": 17072 }, { "epoch": 1.8509323503902864, "grad_norm": 0.08737203478813171, "learning_rate": 3.84411722036849e-05, "loss": 0.0042, "step": 17073 }, { "epoch": 1.851040763226366, "grad_norm": 0.09318449348211288, "learning_rate": 3.843754533584796e-05, "loss": 0.0064, "step": 17074 }, { "epoch": 1.851149176062446, "grad_norm": 0.2661104202270508, "learning_rate": 3.843391846801103e-05, "loss": 0.0392, "step": 17075 }, { "epoch": 1.8512575888985254, "grad_norm": 0.22410301864147186, "learning_rate": 3.843029160017409e-05, "loss": 0.021, "step": 17076 }, { "epoch": 1.8513660017346054, "grad_norm": 0.17684747278690338, "learning_rate": 3.8426664732337155e-05, "loss": 0.0112, "step": 17077 }, { "epoch": 1.8514744145706852, "grad_norm": 0.23173627257347107, "learning_rate": 3.842303786450022e-05, "loss": 0.0203, "step": 17078 }, { "epoch": 1.851582827406765, "grad_norm": 0.7125820517539978, "learning_rate": 3.841941099666328e-05, "loss": 0.0916, "step": 17079 }, { "epoch": 1.8516912402428447, "grad_norm": 0.48662999272346497, "learning_rate": 3.841578412882635e-05, "loss": 0.0367, "step": 17080 }, { "epoch": 1.8517996530789245, "grad_norm": 0.32799044251441956, "learning_rate": 3.841215726098941e-05, "loss": 0.0273, "step": 17081 }, { "epoch": 1.8519080659150045, "grad_norm": 0.3276412785053253, "learning_rate": 3.8408530393152476e-05, "loss": 0.0468, "step": 17082 }, { "epoch": 1.852016478751084, "grad_norm": 0.16194407641887665, "learning_rate": 3.840490352531554e-05, "loss": 0.0098, "step": 17083 }, { "epoch": 1.852124891587164, "grad_norm": 0.5495631098747253, "learning_rate": 3.8401276657478604e-05, "loss": 0.0326, "step": 17084 }, { "epoch": 1.8522333044232437, "grad_norm": 0.1943565458059311, "learning_rate": 3.839764978964167e-05, "loss": 0.0211, "step": 17085 }, { "epoch": 1.8523417172593235, "grad_norm": 0.5979524254798889, "learning_rate": 3.839402292180473e-05, "loss": 0.0526, "step": 17086 }, { "epoch": 1.8524501300954033, "grad_norm": 0.20699438452720642, "learning_rate": 3.8390396053967796e-05, "loss": 0.0167, "step": 17087 }, { "epoch": 1.852558542931483, "grad_norm": 0.14588689804077148, "learning_rate": 3.838676918613086e-05, "loss": 0.0143, "step": 17088 }, { "epoch": 1.852666955767563, "grad_norm": 0.2953757643699646, "learning_rate": 3.8383142318293924e-05, "loss": 0.0211, "step": 17089 }, { "epoch": 1.8527753686036426, "grad_norm": 0.10145289450883865, "learning_rate": 3.837951545045699e-05, "loss": 0.0178, "step": 17090 }, { "epoch": 1.8528837814397225, "grad_norm": 0.09907930344343185, "learning_rate": 3.837588858262005e-05, "loss": 0.0109, "step": 17091 }, { "epoch": 1.8529921942758023, "grad_norm": 0.8204418420791626, "learning_rate": 3.837226171478312e-05, "loss": 0.024, "step": 17092 }, { "epoch": 1.853100607111882, "grad_norm": 0.27975571155548096, "learning_rate": 3.836863484694618e-05, "loss": 0.0299, "step": 17093 }, { "epoch": 1.8532090199479618, "grad_norm": 0.2063942551612854, "learning_rate": 3.836500797910924e-05, "loss": 0.0242, "step": 17094 }, { "epoch": 1.8533174327840416, "grad_norm": 0.43274229764938354, "learning_rate": 3.836138111127231e-05, "loss": 0.0354, "step": 17095 }, { "epoch": 1.8534258456201216, "grad_norm": 0.24294450879096985, "learning_rate": 3.835775424343537e-05, "loss": 0.02, "step": 17096 }, { "epoch": 1.8535342584562011, "grad_norm": 0.16132999956607819, "learning_rate": 3.835412737559844e-05, "loss": 0.0146, "step": 17097 }, { "epoch": 1.853642671292281, "grad_norm": 0.2566666603088379, "learning_rate": 3.83505005077615e-05, "loss": 0.0174, "step": 17098 }, { "epoch": 1.8537510841283606, "grad_norm": 0.03046601451933384, "learning_rate": 3.8346873639924565e-05, "loss": 0.0023, "step": 17099 }, { "epoch": 1.8538594969644406, "grad_norm": 0.1440083533525467, "learning_rate": 3.834324677208763e-05, "loss": 0.0124, "step": 17100 }, { "epoch": 1.8539679098005204, "grad_norm": 0.2625138461589813, "learning_rate": 3.833961990425069e-05, "loss": 0.0209, "step": 17101 }, { "epoch": 1.8540763226366002, "grad_norm": 1.0032293796539307, "learning_rate": 3.833599303641375e-05, "loss": 0.0586, "step": 17102 }, { "epoch": 1.85418473547268, "grad_norm": 0.14054907858371735, "learning_rate": 3.8332366168576815e-05, "loss": 0.0105, "step": 17103 }, { "epoch": 1.8542931483087597, "grad_norm": 0.49387499690055847, "learning_rate": 3.8328739300739886e-05, "loss": 0.0521, "step": 17104 }, { "epoch": 1.8544015611448397, "grad_norm": 0.1632058322429657, "learning_rate": 3.832511243290295e-05, "loss": 0.0224, "step": 17105 }, { "epoch": 1.8545099739809192, "grad_norm": 0.20049966871738434, "learning_rate": 3.8321485565066014e-05, "loss": 0.0227, "step": 17106 }, { "epoch": 1.8546183868169992, "grad_norm": 0.05384937673807144, "learning_rate": 3.831785869722908e-05, "loss": 0.0023, "step": 17107 }, { "epoch": 1.854726799653079, "grad_norm": 0.4947323203086853, "learning_rate": 3.8314231829392136e-05, "loss": 0.0356, "step": 17108 }, { "epoch": 1.8548352124891587, "grad_norm": 0.4623541533946991, "learning_rate": 3.83106049615552e-05, "loss": 0.0392, "step": 17109 }, { "epoch": 1.8549436253252385, "grad_norm": 0.3442259728908539, "learning_rate": 3.8306978093718264e-05, "loss": 0.0444, "step": 17110 }, { "epoch": 1.8550520381613183, "grad_norm": 0.25306835770606995, "learning_rate": 3.830335122588133e-05, "loss": 0.0217, "step": 17111 }, { "epoch": 1.8551604509973982, "grad_norm": 0.5050566792488098, "learning_rate": 3.829972435804439e-05, "loss": 0.0805, "step": 17112 }, { "epoch": 1.8552688638334778, "grad_norm": 0.20423883199691772, "learning_rate": 3.829609749020746e-05, "loss": 0.0159, "step": 17113 }, { "epoch": 1.8553772766695578, "grad_norm": 0.1880466789007187, "learning_rate": 3.829247062237053e-05, "loss": 0.0426, "step": 17114 }, { "epoch": 1.8554856895056373, "grad_norm": 0.46136945486068726, "learning_rate": 3.8288843754533584e-05, "loss": 0.0468, "step": 17115 }, { "epoch": 1.8555941023417173, "grad_norm": 0.26160749793052673, "learning_rate": 3.828521688669665e-05, "loss": 0.0136, "step": 17116 }, { "epoch": 1.855702515177797, "grad_norm": 0.2851884067058563, "learning_rate": 3.828159001885971e-05, "loss": 0.0301, "step": 17117 }, { "epoch": 1.8558109280138768, "grad_norm": 0.035178396850824356, "learning_rate": 3.827796315102278e-05, "loss": 0.0042, "step": 17118 }, { "epoch": 1.8559193408499568, "grad_norm": 0.18365740776062012, "learning_rate": 3.827433628318584e-05, "loss": 0.012, "step": 17119 }, { "epoch": 1.8560277536860363, "grad_norm": 0.2530440092086792, "learning_rate": 3.8270709415348905e-05, "loss": 0.0196, "step": 17120 }, { "epoch": 1.8561361665221163, "grad_norm": 0.2861773371696472, "learning_rate": 3.8267082547511976e-05, "loss": 0.0184, "step": 17121 }, { "epoch": 1.8562445793581959, "grad_norm": 0.3360894024372101, "learning_rate": 3.826345567967504e-05, "loss": 0.0368, "step": 17122 }, { "epoch": 1.8563529921942759, "grad_norm": 0.7009744644165039, "learning_rate": 3.82598288118381e-05, "loss": 0.0412, "step": 17123 }, { "epoch": 1.8564614050303556, "grad_norm": 0.06279300153255463, "learning_rate": 3.825620194400116e-05, "loss": 0.0045, "step": 17124 }, { "epoch": 1.8565698178664354, "grad_norm": 0.24654904007911682, "learning_rate": 3.8252575076164225e-05, "loss": 0.0203, "step": 17125 }, { "epoch": 1.8566782307025151, "grad_norm": 0.10823940485715866, "learning_rate": 3.824894820832729e-05, "loss": 0.0065, "step": 17126 }, { "epoch": 1.856786643538595, "grad_norm": 0.16327278316020966, "learning_rate": 3.8245321340490354e-05, "loss": 0.0158, "step": 17127 }, { "epoch": 1.856895056374675, "grad_norm": 0.1127859577536583, "learning_rate": 3.824169447265342e-05, "loss": 0.0133, "step": 17128 }, { "epoch": 1.8570034692107544, "grad_norm": 0.02448272705078125, "learning_rate": 3.823806760481648e-05, "loss": 0.002, "step": 17129 }, { "epoch": 1.8571118820468344, "grad_norm": 0.12205442786216736, "learning_rate": 3.8234440736979546e-05, "loss": 0.0068, "step": 17130 }, { "epoch": 1.8572202948829142, "grad_norm": 0.1962658017873764, "learning_rate": 3.823081386914261e-05, "loss": 0.0292, "step": 17131 }, { "epoch": 1.857328707718994, "grad_norm": 0.08554549515247345, "learning_rate": 3.8227187001305674e-05, "loss": 0.0089, "step": 17132 }, { "epoch": 1.8574371205550737, "grad_norm": 0.31666839122772217, "learning_rate": 3.822356013346874e-05, "loss": 0.0347, "step": 17133 }, { "epoch": 1.8575455333911535, "grad_norm": 0.16354867815971375, "learning_rate": 3.82199332656318e-05, "loss": 0.023, "step": 17134 }, { "epoch": 1.8576539462272335, "grad_norm": 0.5795090794563293, "learning_rate": 3.8216306397794866e-05, "loss": 0.0087, "step": 17135 }, { "epoch": 1.857762359063313, "grad_norm": 0.4662829637527466, "learning_rate": 3.821267952995793e-05, "loss": 0.0411, "step": 17136 }, { "epoch": 1.857870771899393, "grad_norm": 0.5510334968566895, "learning_rate": 3.8209052662120995e-05, "loss": 0.0186, "step": 17137 }, { "epoch": 1.8579791847354725, "grad_norm": 0.5114193558692932, "learning_rate": 3.820542579428406e-05, "loss": 0.0278, "step": 17138 }, { "epoch": 1.8580875975715525, "grad_norm": 0.13369062542915344, "learning_rate": 3.820179892644712e-05, "loss": 0.0047, "step": 17139 }, { "epoch": 1.8581960104076323, "grad_norm": 0.3066043257713318, "learning_rate": 3.819817205861019e-05, "loss": 0.0305, "step": 17140 }, { "epoch": 1.858304423243712, "grad_norm": 0.045610591769218445, "learning_rate": 3.819454519077325e-05, "loss": 0.0024, "step": 17141 }, { "epoch": 1.8584128360797918, "grad_norm": 0.16582560539245605, "learning_rate": 3.8190918322936315e-05, "loss": 0.0186, "step": 17142 }, { "epoch": 1.8585212489158716, "grad_norm": 0.33129996061325073, "learning_rate": 3.818729145509938e-05, "loss": 0.03, "step": 17143 }, { "epoch": 1.8586296617519515, "grad_norm": 0.11180268228054047, "learning_rate": 3.8183664587262443e-05, "loss": 0.0129, "step": 17144 }, { "epoch": 1.858738074588031, "grad_norm": 0.490165650844574, "learning_rate": 3.818003771942551e-05, "loss": 0.0245, "step": 17145 }, { "epoch": 1.858846487424111, "grad_norm": 0.3531034290790558, "learning_rate": 3.8176410851588565e-05, "loss": 0.0191, "step": 17146 }, { "epoch": 1.8589549002601908, "grad_norm": 0.23128142952919006, "learning_rate": 3.8172783983751636e-05, "loss": 0.0145, "step": 17147 }, { "epoch": 1.8590633130962706, "grad_norm": 0.31774649024009705, "learning_rate": 3.81691571159147e-05, "loss": 0.0432, "step": 17148 }, { "epoch": 1.8591717259323504, "grad_norm": 0.32991471886634827, "learning_rate": 3.8165530248077764e-05, "loss": 0.0188, "step": 17149 }, { "epoch": 1.8592801387684301, "grad_norm": 0.2427501082420349, "learning_rate": 3.816190338024083e-05, "loss": 0.0157, "step": 17150 }, { "epoch": 1.85938855160451, "grad_norm": 0.20204173028469086, "learning_rate": 3.815827651240389e-05, "loss": 0.0056, "step": 17151 }, { "epoch": 1.8594969644405897, "grad_norm": 0.5861077904701233, "learning_rate": 3.8154649644566956e-05, "loss": 0.0311, "step": 17152 }, { "epoch": 1.8596053772766696, "grad_norm": 0.46562957763671875, "learning_rate": 3.8151022776730014e-05, "loss": 0.0273, "step": 17153 }, { "epoch": 1.8597137901127494, "grad_norm": 0.3282621204853058, "learning_rate": 3.814739590889308e-05, "loss": 0.0142, "step": 17154 }, { "epoch": 1.8598222029488292, "grad_norm": 0.4715195894241333, "learning_rate": 3.814376904105614e-05, "loss": 0.0309, "step": 17155 }, { "epoch": 1.859930615784909, "grad_norm": 0.3301733434200287, "learning_rate": 3.814014217321921e-05, "loss": 0.0155, "step": 17156 }, { "epoch": 1.8600390286209887, "grad_norm": 0.23396678268909454, "learning_rate": 3.813651530538228e-05, "loss": 0.0106, "step": 17157 }, { "epoch": 1.8601474414570687, "grad_norm": 0.24417200684547424, "learning_rate": 3.813288843754534e-05, "loss": 0.0173, "step": 17158 }, { "epoch": 1.8602558542931482, "grad_norm": 0.16632267832756042, "learning_rate": 3.8129261569708405e-05, "loss": 0.0092, "step": 17159 }, { "epoch": 1.8603642671292282, "grad_norm": 0.24149085581302643, "learning_rate": 3.812563470187146e-05, "loss": 0.0292, "step": 17160 }, { "epoch": 1.8604726799653077, "grad_norm": 0.1946777105331421, "learning_rate": 3.8122007834034526e-05, "loss": 0.012, "step": 17161 }, { "epoch": 1.8605810928013877, "grad_norm": 0.37496253848075867, "learning_rate": 3.811838096619759e-05, "loss": 0.027, "step": 17162 }, { "epoch": 1.8606895056374675, "grad_norm": 0.7548540234565735, "learning_rate": 3.8114754098360655e-05, "loss": 0.0513, "step": 17163 }, { "epoch": 1.8607979184735473, "grad_norm": 0.1331607550382614, "learning_rate": 3.811112723052372e-05, "loss": 0.0178, "step": 17164 }, { "epoch": 1.860906331309627, "grad_norm": 0.1637331247329712, "learning_rate": 3.810750036268679e-05, "loss": 0.0203, "step": 17165 }, { "epoch": 1.8610147441457068, "grad_norm": 0.4422706067562103, "learning_rate": 3.8103873494849854e-05, "loss": 0.0455, "step": 17166 }, { "epoch": 1.8611231569817868, "grad_norm": 0.2384800910949707, "learning_rate": 3.810024662701291e-05, "loss": 0.0279, "step": 17167 }, { "epoch": 1.8612315698178663, "grad_norm": 0.147753044962883, "learning_rate": 3.8096619759175975e-05, "loss": 0.0109, "step": 17168 }, { "epoch": 1.8613399826539463, "grad_norm": 0.31073006987571716, "learning_rate": 3.809299289133904e-05, "loss": 0.0322, "step": 17169 }, { "epoch": 1.861448395490026, "grad_norm": 0.16052323579788208, "learning_rate": 3.80893660235021e-05, "loss": 0.0109, "step": 17170 }, { "epoch": 1.8615568083261058, "grad_norm": 0.14222684502601624, "learning_rate": 3.808573915566517e-05, "loss": 0.0047, "step": 17171 }, { "epoch": 1.8616652211621856, "grad_norm": 0.16648361086845398, "learning_rate": 3.808211228782823e-05, "loss": 0.024, "step": 17172 }, { "epoch": 1.8617736339982653, "grad_norm": 0.9316790103912354, "learning_rate": 3.80784854199913e-05, "loss": 0.0177, "step": 17173 }, { "epoch": 1.8618820468343453, "grad_norm": 0.2581319808959961, "learning_rate": 3.8074858552154367e-05, "loss": 0.0287, "step": 17174 }, { "epoch": 1.8619904596704249, "grad_norm": 0.1484869122505188, "learning_rate": 3.8071231684317424e-05, "loss": 0.0054, "step": 17175 }, { "epoch": 1.8620988725065049, "grad_norm": 0.23689508438110352, "learning_rate": 3.806760481648049e-05, "loss": 0.0158, "step": 17176 }, { "epoch": 1.8622072853425846, "grad_norm": 0.4712611734867096, "learning_rate": 3.806397794864355e-05, "loss": 0.0215, "step": 17177 }, { "epoch": 1.8623156981786644, "grad_norm": 0.5378323793411255, "learning_rate": 3.8060351080806616e-05, "loss": 0.0224, "step": 17178 }, { "epoch": 1.8624241110147441, "grad_norm": 0.1485869437456131, "learning_rate": 3.805672421296968e-05, "loss": 0.017, "step": 17179 }, { "epoch": 1.862532523850824, "grad_norm": 0.28550252318382263, "learning_rate": 3.8053097345132744e-05, "loss": 0.009, "step": 17180 }, { "epoch": 1.862640936686904, "grad_norm": 0.25709837675094604, "learning_rate": 3.804947047729581e-05, "loss": 0.0088, "step": 17181 }, { "epoch": 1.8627493495229834, "grad_norm": 0.33564603328704834, "learning_rate": 3.804584360945887e-05, "loss": 0.0254, "step": 17182 }, { "epoch": 1.8628577623590634, "grad_norm": 0.17030394077301025, "learning_rate": 3.804221674162194e-05, "loss": 0.02, "step": 17183 }, { "epoch": 1.862966175195143, "grad_norm": 0.315685898065567, "learning_rate": 3.8038589873785e-05, "loss": 0.0259, "step": 17184 }, { "epoch": 1.863074588031223, "grad_norm": 0.6621593832969666, "learning_rate": 3.8034963005948065e-05, "loss": 0.0435, "step": 17185 }, { "epoch": 1.8631830008673027, "grad_norm": 0.2503606379032135, "learning_rate": 3.803133613811113e-05, "loss": 0.0197, "step": 17186 }, { "epoch": 1.8632914137033825, "grad_norm": 0.4471585750579834, "learning_rate": 3.802770927027419e-05, "loss": 0.0236, "step": 17187 }, { "epoch": 1.8633998265394622, "grad_norm": 0.20578430593013763, "learning_rate": 3.802408240243726e-05, "loss": 0.0136, "step": 17188 }, { "epoch": 1.863508239375542, "grad_norm": 0.22701697051525116, "learning_rate": 3.802045553460032e-05, "loss": 0.0047, "step": 17189 }, { "epoch": 1.863616652211622, "grad_norm": 0.5720329284667969, "learning_rate": 3.8016828666763385e-05, "loss": 0.0255, "step": 17190 }, { "epoch": 1.8637250650477015, "grad_norm": 0.1008603498339653, "learning_rate": 3.801320179892645e-05, "loss": 0.0124, "step": 17191 }, { "epoch": 1.8638334778837815, "grad_norm": 0.387219101190567, "learning_rate": 3.8009574931089514e-05, "loss": 0.0173, "step": 17192 }, { "epoch": 1.8639418907198613, "grad_norm": 0.22567212581634521, "learning_rate": 3.800594806325258e-05, "loss": 0.0115, "step": 17193 }, { "epoch": 1.864050303555941, "grad_norm": 0.22588881850242615, "learning_rate": 3.800232119541564e-05, "loss": 0.0104, "step": 17194 }, { "epoch": 1.8641587163920208, "grad_norm": 0.14507927000522614, "learning_rate": 3.7998694327578706e-05, "loss": 0.025, "step": 17195 }, { "epoch": 1.8642671292281006, "grad_norm": 0.14958049356937408, "learning_rate": 3.799506745974177e-05, "loss": 0.0207, "step": 17196 }, { "epoch": 1.8643755420641805, "grad_norm": 0.8572026491165161, "learning_rate": 3.7991440591904834e-05, "loss": 0.0666, "step": 17197 }, { "epoch": 1.86448395490026, "grad_norm": 0.08597090095281601, "learning_rate": 3.798781372406789e-05, "loss": 0.0016, "step": 17198 }, { "epoch": 1.86459236773634, "grad_norm": 0.08015131950378418, "learning_rate": 3.798418685623096e-05, "loss": 0.0078, "step": 17199 }, { "epoch": 1.8647007805724196, "grad_norm": 0.18150992691516876, "learning_rate": 3.7980559988394027e-05, "loss": 0.0235, "step": 17200 }, { "epoch": 1.8648091934084996, "grad_norm": 0.5266253352165222, "learning_rate": 3.797693312055709e-05, "loss": 0.025, "step": 17201 }, { "epoch": 1.8649176062445794, "grad_norm": 0.13428795337677002, "learning_rate": 3.7973306252720155e-05, "loss": 0.0038, "step": 17202 }, { "epoch": 1.8650260190806591, "grad_norm": 0.2557620704174042, "learning_rate": 3.796967938488322e-05, "loss": 0.0354, "step": 17203 }, { "epoch": 1.8651344319167389, "grad_norm": 0.08889492601156235, "learning_rate": 3.796605251704628e-05, "loss": 0.0043, "step": 17204 }, { "epoch": 1.8652428447528187, "grad_norm": 0.0785517692565918, "learning_rate": 3.796242564920934e-05, "loss": 0.0018, "step": 17205 }, { "epoch": 1.8653512575888986, "grad_norm": 0.48554471135139465, "learning_rate": 3.7958798781372404e-05, "loss": 0.0283, "step": 17206 }, { "epoch": 1.8654596704249782, "grad_norm": 0.7550365924835205, "learning_rate": 3.795517191353547e-05, "loss": 0.0499, "step": 17207 }, { "epoch": 1.8655680832610582, "grad_norm": 0.4875458776950836, "learning_rate": 3.795154504569854e-05, "loss": 0.0764, "step": 17208 }, { "epoch": 1.865676496097138, "grad_norm": 0.08606699109077454, "learning_rate": 3.7947918177861603e-05, "loss": 0.0026, "step": 17209 }, { "epoch": 1.8657849089332177, "grad_norm": 0.34706413745880127, "learning_rate": 3.794429131002467e-05, "loss": 0.0195, "step": 17210 }, { "epoch": 1.8658933217692975, "grad_norm": 0.15056458115577698, "learning_rate": 3.794066444218773e-05, "loss": 0.0078, "step": 17211 }, { "epoch": 1.8660017346053772, "grad_norm": 0.5237996578216553, "learning_rate": 3.793703757435079e-05, "loss": 0.058, "step": 17212 }, { "epoch": 1.8661101474414572, "grad_norm": 0.6356196999549866, "learning_rate": 3.793341070651385e-05, "loss": 0.025, "step": 17213 }, { "epoch": 1.8662185602775367, "grad_norm": 0.4045336842536926, "learning_rate": 3.792978383867692e-05, "loss": 0.0385, "step": 17214 }, { "epoch": 1.8663269731136167, "grad_norm": 0.03733214735984802, "learning_rate": 3.792615697083998e-05, "loss": 0.0015, "step": 17215 }, { "epoch": 1.8664353859496965, "grad_norm": 0.10294004529714584, "learning_rate": 3.792253010300305e-05, "loss": 0.0044, "step": 17216 }, { "epoch": 1.8665437987857763, "grad_norm": 0.6764059066772461, "learning_rate": 3.7918903235166116e-05, "loss": 0.0411, "step": 17217 }, { "epoch": 1.866652211621856, "grad_norm": 0.21972286701202393, "learning_rate": 3.791527636732918e-05, "loss": 0.0243, "step": 17218 }, { "epoch": 1.8667606244579358, "grad_norm": 0.2847893536090851, "learning_rate": 3.791164949949224e-05, "loss": 0.0337, "step": 17219 }, { "epoch": 1.8668690372940158, "grad_norm": 0.40047958493232727, "learning_rate": 3.79080226316553e-05, "loss": 0.0266, "step": 17220 }, { "epoch": 1.8669774501300953, "grad_norm": 0.4219416379928589, "learning_rate": 3.7904395763818366e-05, "loss": 0.0315, "step": 17221 }, { "epoch": 1.8670858629661753, "grad_norm": 0.4961167275905609, "learning_rate": 3.790076889598143e-05, "loss": 0.0171, "step": 17222 }, { "epoch": 1.8671942758022548, "grad_norm": 1.5961652994155884, "learning_rate": 3.7897142028144494e-05, "loss": 0.0358, "step": 17223 }, { "epoch": 1.8673026886383348, "grad_norm": 0.17742182314395905, "learning_rate": 3.789351516030756e-05, "loss": 0.0142, "step": 17224 }, { "epoch": 1.8674111014744146, "grad_norm": 0.4151385724544525, "learning_rate": 3.788988829247063e-05, "loss": 0.0259, "step": 17225 }, { "epoch": 1.8675195143104943, "grad_norm": 0.18593178689479828, "learning_rate": 3.788626142463369e-05, "loss": 0.0082, "step": 17226 }, { "epoch": 1.867627927146574, "grad_norm": 0.24592961370944977, "learning_rate": 3.788263455679675e-05, "loss": 0.0163, "step": 17227 }, { "epoch": 1.8677363399826539, "grad_norm": 0.42832931876182556, "learning_rate": 3.7879007688959815e-05, "loss": 0.0382, "step": 17228 }, { "epoch": 1.8678447528187339, "grad_norm": 0.05460921302437782, "learning_rate": 3.787538082112288e-05, "loss": 0.0047, "step": 17229 }, { "epoch": 1.8679531656548134, "grad_norm": 0.24831774830818176, "learning_rate": 3.787175395328594e-05, "loss": 0.0177, "step": 17230 }, { "epoch": 1.8680615784908934, "grad_norm": 0.36103329062461853, "learning_rate": 3.786812708544901e-05, "loss": 0.0208, "step": 17231 }, { "epoch": 1.8681699913269731, "grad_norm": 0.05910167843103409, "learning_rate": 3.786450021761207e-05, "loss": 0.0038, "step": 17232 }, { "epoch": 1.868278404163053, "grad_norm": 0.20372603833675385, "learning_rate": 3.7860873349775135e-05, "loss": 0.0158, "step": 17233 }, { "epoch": 1.8683868169991327, "grad_norm": 0.5871861577033997, "learning_rate": 3.78572464819382e-05, "loss": 0.0197, "step": 17234 }, { "epoch": 1.8684952298352124, "grad_norm": 0.20645692944526672, "learning_rate": 3.7853619614101263e-05, "loss": 0.0112, "step": 17235 }, { "epoch": 1.8686036426712924, "grad_norm": 0.028339901939034462, "learning_rate": 3.784999274626433e-05, "loss": 0.0022, "step": 17236 }, { "epoch": 1.868712055507372, "grad_norm": 0.5350943803787231, "learning_rate": 3.784636587842739e-05, "loss": 0.0483, "step": 17237 }, { "epoch": 1.868820468343452, "grad_norm": 0.27825120091438293, "learning_rate": 3.7842739010590456e-05, "loss": 0.0263, "step": 17238 }, { "epoch": 1.8689288811795317, "grad_norm": 0.2596844732761383, "learning_rate": 3.783911214275352e-05, "loss": 0.0297, "step": 17239 }, { "epoch": 1.8690372940156115, "grad_norm": 0.02830386534333229, "learning_rate": 3.7835485274916584e-05, "loss": 0.0017, "step": 17240 }, { "epoch": 1.8691457068516912, "grad_norm": 0.5434520840644836, "learning_rate": 3.783185840707965e-05, "loss": 0.0397, "step": 17241 }, { "epoch": 1.869254119687771, "grad_norm": 0.08085209131240845, "learning_rate": 3.782823153924271e-05, "loss": 0.0063, "step": 17242 }, { "epoch": 1.869362532523851, "grad_norm": 0.15866367518901825, "learning_rate": 3.7824604671405776e-05, "loss": 0.0209, "step": 17243 }, { "epoch": 1.8694709453599305, "grad_norm": 0.20368319749832153, "learning_rate": 3.782097780356884e-05, "loss": 0.0134, "step": 17244 }, { "epoch": 1.8695793581960105, "grad_norm": 0.4693523645401001, "learning_rate": 3.7817350935731904e-05, "loss": 0.0162, "step": 17245 }, { "epoch": 1.86968777103209, "grad_norm": 0.20567043125629425, "learning_rate": 3.781372406789497e-05, "loss": 0.0107, "step": 17246 }, { "epoch": 1.86979618386817, "grad_norm": 0.48335129022598267, "learning_rate": 3.781009720005803e-05, "loss": 0.0719, "step": 17247 }, { "epoch": 1.8699045967042498, "grad_norm": 0.11249522119760513, "learning_rate": 3.78064703322211e-05, "loss": 0.0109, "step": 17248 }, { "epoch": 1.8700130095403296, "grad_norm": 0.3057611584663391, "learning_rate": 3.780284346438416e-05, "loss": 0.0212, "step": 17249 }, { "epoch": 1.8701214223764093, "grad_norm": 0.38176438212394714, "learning_rate": 3.779921659654722e-05, "loss": 0.0162, "step": 17250 }, { "epoch": 1.870229835212489, "grad_norm": 0.14463253319263458, "learning_rate": 3.779558972871029e-05, "loss": 0.0043, "step": 17251 }, { "epoch": 1.870338248048569, "grad_norm": 0.4004805088043213, "learning_rate": 3.779196286087335e-05, "loss": 0.0317, "step": 17252 }, { "epoch": 1.8704466608846486, "grad_norm": 0.5395352244377136, "learning_rate": 3.778833599303642e-05, "loss": 0.0525, "step": 17253 }, { "epoch": 1.8705550737207286, "grad_norm": 0.08367689698934555, "learning_rate": 3.778470912519948e-05, "loss": 0.0021, "step": 17254 }, { "epoch": 1.8706634865568084, "grad_norm": 0.5848238468170166, "learning_rate": 3.7781082257362546e-05, "loss": 0.0138, "step": 17255 }, { "epoch": 1.8707718993928881, "grad_norm": 0.4334489107131958, "learning_rate": 3.777745538952561e-05, "loss": 0.0479, "step": 17256 }, { "epoch": 1.870880312228968, "grad_norm": 0.33573347330093384, "learning_rate": 3.777382852168867e-05, "loss": 0.0313, "step": 17257 }, { "epoch": 1.8709887250650477, "grad_norm": 0.14511169493198395, "learning_rate": 3.777020165385173e-05, "loss": 0.0184, "step": 17258 }, { "epoch": 1.8710971379011276, "grad_norm": 0.3294348418712616, "learning_rate": 3.7766574786014795e-05, "loss": 0.0373, "step": 17259 }, { "epoch": 1.8712055507372072, "grad_norm": 0.48710089921951294, "learning_rate": 3.7762947918177866e-05, "loss": 0.0601, "step": 17260 }, { "epoch": 1.8713139635732872, "grad_norm": 0.20223505795001984, "learning_rate": 3.775932105034093e-05, "loss": 0.0132, "step": 17261 }, { "epoch": 1.8714223764093667, "grad_norm": 0.2659092843532562, "learning_rate": 3.7755694182503994e-05, "loss": 0.0165, "step": 17262 }, { "epoch": 1.8715307892454467, "grad_norm": 0.17754195630550385, "learning_rate": 3.775206731466706e-05, "loss": 0.0156, "step": 17263 }, { "epoch": 1.8716392020815265, "grad_norm": 0.4775151014328003, "learning_rate": 3.7748440446830116e-05, "loss": 0.0487, "step": 17264 }, { "epoch": 1.8717476149176062, "grad_norm": 0.059191010892391205, "learning_rate": 3.774481357899318e-05, "loss": 0.0059, "step": 17265 }, { "epoch": 1.8718560277536862, "grad_norm": 0.14160598814487457, "learning_rate": 3.7741186711156244e-05, "loss": 0.0105, "step": 17266 }, { "epoch": 1.8719644405897657, "grad_norm": 0.07277330011129379, "learning_rate": 3.773755984331931e-05, "loss": 0.0037, "step": 17267 }, { "epoch": 1.8720728534258457, "grad_norm": 0.5261088609695435, "learning_rate": 3.773393297548238e-05, "loss": 0.0327, "step": 17268 }, { "epoch": 1.8721812662619253, "grad_norm": 0.2043910026550293, "learning_rate": 3.773030610764544e-05, "loss": 0.0144, "step": 17269 }, { "epoch": 1.8722896790980053, "grad_norm": 0.1317528486251831, "learning_rate": 3.772667923980851e-05, "loss": 0.0128, "step": 17270 }, { "epoch": 1.872398091934085, "grad_norm": 0.5496759414672852, "learning_rate": 3.772305237197157e-05, "loss": 0.0346, "step": 17271 }, { "epoch": 1.8725065047701648, "grad_norm": 0.02906346321105957, "learning_rate": 3.771942550413463e-05, "loss": 0.0015, "step": 17272 }, { "epoch": 1.8726149176062445, "grad_norm": 0.23848211765289307, "learning_rate": 3.771579863629769e-05, "loss": 0.0236, "step": 17273 }, { "epoch": 1.8727233304423243, "grad_norm": 0.3029230833053589, "learning_rate": 3.771217176846076e-05, "loss": 0.018, "step": 17274 }, { "epoch": 1.8728317432784043, "grad_norm": 0.1925761103630066, "learning_rate": 3.770854490062382e-05, "loss": 0.0114, "step": 17275 }, { "epoch": 1.8729401561144838, "grad_norm": 0.11250624060630798, "learning_rate": 3.7704918032786885e-05, "loss": 0.0105, "step": 17276 }, { "epoch": 1.8730485689505638, "grad_norm": 0.4475580155849457, "learning_rate": 3.7701291164949956e-05, "loss": 0.0144, "step": 17277 }, { "epoch": 1.8731569817866436, "grad_norm": 0.16569025814533234, "learning_rate": 3.769766429711302e-05, "loss": 0.0283, "step": 17278 }, { "epoch": 1.8732653946227233, "grad_norm": 0.17641723155975342, "learning_rate": 3.769403742927608e-05, "loss": 0.0121, "step": 17279 }, { "epoch": 1.873373807458803, "grad_norm": 0.16275765001773834, "learning_rate": 3.769041056143914e-05, "loss": 0.0084, "step": 17280 }, { "epoch": 1.8734822202948829, "grad_norm": 0.25519832968711853, "learning_rate": 3.7686783693602205e-05, "loss": 0.0145, "step": 17281 }, { "epoch": 1.8735906331309629, "grad_norm": 0.2991848587989807, "learning_rate": 3.768315682576527e-05, "loss": 0.0128, "step": 17282 }, { "epoch": 1.8736990459670424, "grad_norm": 0.09439102560281754, "learning_rate": 3.7679529957928334e-05, "loss": 0.0059, "step": 17283 }, { "epoch": 1.8738074588031224, "grad_norm": 0.8397454619407654, "learning_rate": 3.76759030900914e-05, "loss": 0.0422, "step": 17284 }, { "epoch": 1.873915871639202, "grad_norm": 0.17022986710071564, "learning_rate": 3.767227622225447e-05, "loss": 0.0067, "step": 17285 }, { "epoch": 1.874024284475282, "grad_norm": 0.013571667484939098, "learning_rate": 3.7668649354417526e-05, "loss": 0.0006, "step": 17286 }, { "epoch": 1.8741326973113617, "grad_norm": 0.503864049911499, "learning_rate": 3.766502248658059e-05, "loss": 0.0346, "step": 17287 }, { "epoch": 1.8742411101474414, "grad_norm": 0.4357435405254364, "learning_rate": 3.7661395618743654e-05, "loss": 0.0127, "step": 17288 }, { "epoch": 1.8743495229835212, "grad_norm": 0.3014495074748993, "learning_rate": 3.765776875090672e-05, "loss": 0.0315, "step": 17289 }, { "epoch": 1.874457935819601, "grad_norm": 0.22822365164756775, "learning_rate": 3.765414188306978e-05, "loss": 0.0131, "step": 17290 }, { "epoch": 1.874566348655681, "grad_norm": 0.4909725487232208, "learning_rate": 3.7650515015232847e-05, "loss": 0.0303, "step": 17291 }, { "epoch": 1.8746747614917605, "grad_norm": 0.38843050599098206, "learning_rate": 3.764688814739591e-05, "loss": 0.0391, "step": 17292 }, { "epoch": 1.8747831743278405, "grad_norm": 0.39825594425201416, "learning_rate": 3.7643261279558975e-05, "loss": 0.0133, "step": 17293 }, { "epoch": 1.8748915871639202, "grad_norm": 0.2690967917442322, "learning_rate": 3.763963441172204e-05, "loss": 0.0213, "step": 17294 }, { "epoch": 1.875, "grad_norm": 0.2372143268585205, "learning_rate": 3.76360075438851e-05, "loss": 0.0217, "step": 17295 }, { "epoch": 1.8751084128360798, "grad_norm": 0.3334646224975586, "learning_rate": 3.763238067604817e-05, "loss": 0.041, "step": 17296 }, { "epoch": 1.8752168256721595, "grad_norm": 0.06362155079841614, "learning_rate": 3.762875380821123e-05, "loss": 0.0051, "step": 17297 }, { "epoch": 1.8753252385082395, "grad_norm": 0.13067829608917236, "learning_rate": 3.7625126940374295e-05, "loss": 0.0049, "step": 17298 }, { "epoch": 1.875433651344319, "grad_norm": 0.06061883270740509, "learning_rate": 3.762150007253736e-05, "loss": 0.0032, "step": 17299 }, { "epoch": 1.875542064180399, "grad_norm": 0.4130927324295044, "learning_rate": 3.7617873204700423e-05, "loss": 0.0472, "step": 17300 }, { "epoch": 1.8756504770164788, "grad_norm": 0.39103588461875916, "learning_rate": 3.761424633686349e-05, "loss": 0.0254, "step": 17301 }, { "epoch": 1.8757588898525586, "grad_norm": 0.8448804616928101, "learning_rate": 3.7610619469026545e-05, "loss": 0.0178, "step": 17302 }, { "epoch": 1.8758673026886383, "grad_norm": 0.42350977659225464, "learning_rate": 3.7606992601189616e-05, "loss": 0.0184, "step": 17303 }, { "epoch": 1.875975715524718, "grad_norm": 0.16385284066200256, "learning_rate": 3.760336573335268e-05, "loss": 0.0086, "step": 17304 }, { "epoch": 1.876084128360798, "grad_norm": 0.7330192923545837, "learning_rate": 3.7599738865515744e-05, "loss": 0.0352, "step": 17305 }, { "epoch": 1.8761925411968776, "grad_norm": 0.03297672048211098, "learning_rate": 3.759611199767881e-05, "loss": 0.0018, "step": 17306 }, { "epoch": 1.8763009540329576, "grad_norm": 0.41355088353157043, "learning_rate": 3.759248512984187e-05, "loss": 0.0198, "step": 17307 }, { "epoch": 1.8764093668690371, "grad_norm": 0.3458738923072815, "learning_rate": 3.7588858262004936e-05, "loss": 0.0444, "step": 17308 }, { "epoch": 1.8765177797051171, "grad_norm": 0.06268715113401413, "learning_rate": 3.7585231394167994e-05, "loss": 0.0024, "step": 17309 }, { "epoch": 1.876626192541197, "grad_norm": 0.2904796898365021, "learning_rate": 3.758160452633106e-05, "loss": 0.0129, "step": 17310 }, { "epoch": 1.8767346053772767, "grad_norm": 0.1573246419429779, "learning_rate": 3.757797765849413e-05, "loss": 0.005, "step": 17311 }, { "epoch": 1.8768430182133564, "grad_norm": 0.1825093924999237, "learning_rate": 3.757435079065719e-05, "loss": 0.0225, "step": 17312 }, { "epoch": 1.8769514310494362, "grad_norm": 0.3571895360946655, "learning_rate": 3.757072392282026e-05, "loss": 0.0274, "step": 17313 }, { "epoch": 1.8770598438855162, "grad_norm": 0.17161919176578522, "learning_rate": 3.756709705498332e-05, "loss": 0.0107, "step": 17314 }, { "epoch": 1.8771682567215957, "grad_norm": 0.1681693196296692, "learning_rate": 3.7563470187146385e-05, "loss": 0.0079, "step": 17315 }, { "epoch": 1.8772766695576757, "grad_norm": 0.1543399542570114, "learning_rate": 3.755984331930944e-05, "loss": 0.0095, "step": 17316 }, { "epoch": 1.8773850823937555, "grad_norm": 0.09113521873950958, "learning_rate": 3.7556216451472506e-05, "loss": 0.0029, "step": 17317 }, { "epoch": 1.8774934952298352, "grad_norm": 0.1889152228832245, "learning_rate": 3.755258958363557e-05, "loss": 0.012, "step": 17318 }, { "epoch": 1.877601908065915, "grad_norm": 0.43323925137519836, "learning_rate": 3.7548962715798635e-05, "loss": 0.0126, "step": 17319 }, { "epoch": 1.8777103209019947, "grad_norm": 0.4224238693714142, "learning_rate": 3.7545335847961706e-05, "loss": 0.0265, "step": 17320 }, { "epoch": 1.8778187337380747, "grad_norm": 0.5509693026542664, "learning_rate": 3.754170898012477e-05, "loss": 0.0301, "step": 17321 }, { "epoch": 1.8779271465741543, "grad_norm": 0.29264476895332336, "learning_rate": 3.7538082112287834e-05, "loss": 0.0092, "step": 17322 }, { "epoch": 1.8780355594102343, "grad_norm": 0.20758365094661713, "learning_rate": 3.75344552444509e-05, "loss": 0.0285, "step": 17323 }, { "epoch": 1.8781439722463138, "grad_norm": 0.14122766256332397, "learning_rate": 3.7530828376613955e-05, "loss": 0.0066, "step": 17324 }, { "epoch": 1.8782523850823938, "grad_norm": 0.1487063616514206, "learning_rate": 3.752720150877702e-05, "loss": 0.0052, "step": 17325 }, { "epoch": 1.8783607979184735, "grad_norm": 0.2892875671386719, "learning_rate": 3.7523574640940083e-05, "loss": 0.0265, "step": 17326 }, { "epoch": 1.8784692107545533, "grad_norm": 0.19391535222530365, "learning_rate": 3.751994777310315e-05, "loss": 0.0145, "step": 17327 }, { "epoch": 1.8785776235906333, "grad_norm": 0.7260000705718994, "learning_rate": 3.751632090526621e-05, "loss": 0.1214, "step": 17328 }, { "epoch": 1.8786860364267128, "grad_norm": 0.3261438310146332, "learning_rate": 3.751269403742928e-05, "loss": 0.0201, "step": 17329 }, { "epoch": 1.8787944492627928, "grad_norm": 0.31509876251220703, "learning_rate": 3.750906716959235e-05, "loss": 0.0142, "step": 17330 }, { "epoch": 1.8789028620988724, "grad_norm": 0.44822803139686584, "learning_rate": 3.7505440301755404e-05, "loss": 0.0133, "step": 17331 }, { "epoch": 1.8790112749349523, "grad_norm": 0.4237363338470459, "learning_rate": 3.750181343391847e-05, "loss": 0.0432, "step": 17332 }, { "epoch": 1.879119687771032, "grad_norm": 0.13861656188964844, "learning_rate": 3.749818656608153e-05, "loss": 0.0082, "step": 17333 }, { "epoch": 1.8792281006071119, "grad_norm": 0.5668742060661316, "learning_rate": 3.7494559698244596e-05, "loss": 0.0118, "step": 17334 }, { "epoch": 1.8793365134431916, "grad_norm": 0.8441087007522583, "learning_rate": 3.749093283040766e-05, "loss": 0.0199, "step": 17335 }, { "epoch": 1.8794449262792714, "grad_norm": 0.29636263847351074, "learning_rate": 3.7487305962570724e-05, "loss": 0.0514, "step": 17336 }, { "epoch": 1.8795533391153514, "grad_norm": 0.19832195341587067, "learning_rate": 3.7483679094733795e-05, "loss": 0.0044, "step": 17337 }, { "epoch": 1.879661751951431, "grad_norm": 0.7137270569801331, "learning_rate": 3.748005222689685e-05, "loss": 0.0311, "step": 17338 }, { "epoch": 1.879770164787511, "grad_norm": 0.3533684015274048, "learning_rate": 3.747642535905992e-05, "loss": 0.0281, "step": 17339 }, { "epoch": 1.8798785776235907, "grad_norm": 0.15018878877162933, "learning_rate": 3.747279849122298e-05, "loss": 0.0045, "step": 17340 }, { "epoch": 1.8799869904596704, "grad_norm": 0.017213035374879837, "learning_rate": 3.7469171623386045e-05, "loss": 0.001, "step": 17341 }, { "epoch": 1.8800954032957502, "grad_norm": 0.0689634308218956, "learning_rate": 3.746554475554911e-05, "loss": 0.0038, "step": 17342 }, { "epoch": 1.88020381613183, "grad_norm": 0.08745719492435455, "learning_rate": 3.746191788771217e-05, "loss": 0.0049, "step": 17343 }, { "epoch": 1.88031222896791, "grad_norm": 0.3600475490093231, "learning_rate": 3.745829101987524e-05, "loss": 0.0512, "step": 17344 }, { "epoch": 1.8804206418039895, "grad_norm": 0.47983652353286743, "learning_rate": 3.74546641520383e-05, "loss": 0.022, "step": 17345 }, { "epoch": 1.8805290546400695, "grad_norm": 0.8252192735671997, "learning_rate": 3.7451037284201366e-05, "loss": 0.0289, "step": 17346 }, { "epoch": 1.880637467476149, "grad_norm": 0.39858487248420715, "learning_rate": 3.744741041636443e-05, "loss": 0.0106, "step": 17347 }, { "epoch": 1.880745880312229, "grad_norm": 0.1886114776134491, "learning_rate": 3.7443783548527494e-05, "loss": 0.0227, "step": 17348 }, { "epoch": 1.8808542931483088, "grad_norm": 0.30619898438453674, "learning_rate": 3.744015668069056e-05, "loss": 0.0178, "step": 17349 }, { "epoch": 1.8809627059843885, "grad_norm": 0.620521068572998, "learning_rate": 3.743652981285362e-05, "loss": 0.0517, "step": 17350 }, { "epoch": 1.8810711188204683, "grad_norm": 0.10260815918445587, "learning_rate": 3.7432902945016686e-05, "loss": 0.0035, "step": 17351 }, { "epoch": 1.881179531656548, "grad_norm": 0.24332846701145172, "learning_rate": 3.742927607717975e-05, "loss": 0.0144, "step": 17352 }, { "epoch": 1.881287944492628, "grad_norm": 0.18328814208507538, "learning_rate": 3.7425649209342814e-05, "loss": 0.0087, "step": 17353 }, { "epoch": 1.8813963573287076, "grad_norm": 0.4431191086769104, "learning_rate": 3.742202234150587e-05, "loss": 0.0651, "step": 17354 }, { "epoch": 1.8815047701647876, "grad_norm": 0.18407650291919708, "learning_rate": 3.741839547366894e-05, "loss": 0.0141, "step": 17355 }, { "epoch": 1.8816131830008673, "grad_norm": 0.2844335734844208, "learning_rate": 3.7414768605832007e-05, "loss": 0.0556, "step": 17356 }, { "epoch": 1.881721595836947, "grad_norm": 0.04766746237874031, "learning_rate": 3.741114173799507e-05, "loss": 0.0019, "step": 17357 }, { "epoch": 1.8818300086730269, "grad_norm": 0.5022897720336914, "learning_rate": 3.7407514870158135e-05, "loss": 0.0156, "step": 17358 }, { "epoch": 1.8819384215091066, "grad_norm": 0.07757983356714249, "learning_rate": 3.74038880023212e-05, "loss": 0.0045, "step": 17359 }, { "epoch": 1.8820468343451866, "grad_norm": 0.2919996976852417, "learning_rate": 3.740026113448426e-05, "loss": 0.0201, "step": 17360 }, { "epoch": 1.8821552471812661, "grad_norm": 0.5378512740135193, "learning_rate": 3.739663426664732e-05, "loss": 0.0624, "step": 17361 }, { "epoch": 1.8822636600173461, "grad_norm": 0.2519703209400177, "learning_rate": 3.7393007398810384e-05, "loss": 0.033, "step": 17362 }, { "epoch": 1.882372072853426, "grad_norm": 0.12682777643203735, "learning_rate": 3.7389380530973455e-05, "loss": 0.0113, "step": 17363 }, { "epoch": 1.8824804856895057, "grad_norm": 0.40305662155151367, "learning_rate": 3.738575366313652e-05, "loss": 0.0598, "step": 17364 }, { "epoch": 1.8825888985255854, "grad_norm": 0.11002078652381897, "learning_rate": 3.7382126795299584e-05, "loss": 0.0044, "step": 17365 }, { "epoch": 1.8826973113616652, "grad_norm": 0.5928676724433899, "learning_rate": 3.737849992746265e-05, "loss": 0.0558, "step": 17366 }, { "epoch": 1.8828057241977452, "grad_norm": 0.0959649458527565, "learning_rate": 3.737487305962571e-05, "loss": 0.007, "step": 17367 }, { "epoch": 1.8829141370338247, "grad_norm": 0.16829578578472137, "learning_rate": 3.737124619178877e-05, "loss": 0.0064, "step": 17368 }, { "epoch": 1.8830225498699047, "grad_norm": 0.2824709415435791, "learning_rate": 3.736761932395183e-05, "loss": 0.0272, "step": 17369 }, { "epoch": 1.8831309627059842, "grad_norm": 0.274333655834198, "learning_rate": 3.73639924561149e-05, "loss": 0.037, "step": 17370 }, { "epoch": 1.8832393755420642, "grad_norm": 0.07760428637266159, "learning_rate": 3.736036558827796e-05, "loss": 0.008, "step": 17371 }, { "epoch": 1.883347788378144, "grad_norm": 0.47568273544311523, "learning_rate": 3.735673872044103e-05, "loss": 0.0416, "step": 17372 }, { "epoch": 1.8834562012142237, "grad_norm": 0.26452669501304626, "learning_rate": 3.7353111852604096e-05, "loss": 0.0339, "step": 17373 }, { "epoch": 1.8835646140503035, "grad_norm": 0.22333310544490814, "learning_rate": 3.734948498476716e-05, "loss": 0.0129, "step": 17374 }, { "epoch": 1.8836730268863833, "grad_norm": 0.3432588279247284, "learning_rate": 3.7345858116930225e-05, "loss": 0.044, "step": 17375 }, { "epoch": 1.8837814397224633, "grad_norm": 0.06752294301986694, "learning_rate": 3.734223124909328e-05, "loss": 0.0047, "step": 17376 }, { "epoch": 1.8838898525585428, "grad_norm": 0.26766982674598694, "learning_rate": 3.7338604381256346e-05, "loss": 0.0216, "step": 17377 }, { "epoch": 1.8839982653946228, "grad_norm": 0.1562613993883133, "learning_rate": 3.733497751341941e-05, "loss": 0.0136, "step": 17378 }, { "epoch": 1.8841066782307025, "grad_norm": 0.22363610565662384, "learning_rate": 3.7331350645582474e-05, "loss": 0.0115, "step": 17379 }, { "epoch": 1.8842150910667823, "grad_norm": 0.28666865825653076, "learning_rate": 3.732772377774554e-05, "loss": 0.0346, "step": 17380 }, { "epoch": 1.884323503902862, "grad_norm": 0.28943657875061035, "learning_rate": 3.732409690990861e-05, "loss": 0.0253, "step": 17381 }, { "epoch": 1.8844319167389418, "grad_norm": 1.2605068683624268, "learning_rate": 3.732047004207167e-05, "loss": 0.0567, "step": 17382 }, { "epoch": 1.8845403295750218, "grad_norm": 0.2899678349494934, "learning_rate": 3.731684317423473e-05, "loss": 0.0068, "step": 17383 }, { "epoch": 1.8846487424111014, "grad_norm": 0.05529259145259857, "learning_rate": 3.7313216306397795e-05, "loss": 0.0038, "step": 17384 }, { "epoch": 1.8847571552471813, "grad_norm": 0.23863530158996582, "learning_rate": 3.730958943856086e-05, "loss": 0.0039, "step": 17385 }, { "epoch": 1.8848655680832611, "grad_norm": 0.4821785092353821, "learning_rate": 3.730596257072392e-05, "loss": 0.0214, "step": 17386 }, { "epoch": 1.8849739809193409, "grad_norm": 0.1569160521030426, "learning_rate": 3.730233570288699e-05, "loss": 0.0092, "step": 17387 }, { "epoch": 1.8850823937554206, "grad_norm": 0.34656018018722534, "learning_rate": 3.729870883505005e-05, "loss": 0.0236, "step": 17388 }, { "epoch": 1.8851908065915004, "grad_norm": 0.6221505999565125, "learning_rate": 3.729508196721312e-05, "loss": 0.029, "step": 17389 }, { "epoch": 1.8852992194275804, "grad_norm": 0.11611853539943695, "learning_rate": 3.729145509937618e-05, "loss": 0.0139, "step": 17390 }, { "epoch": 1.88540763226366, "grad_norm": 0.1146741434931755, "learning_rate": 3.7287828231539243e-05, "loss": 0.014, "step": 17391 }, { "epoch": 1.88551604509974, "grad_norm": 0.16555266082286835, "learning_rate": 3.728420136370231e-05, "loss": 0.0121, "step": 17392 }, { "epoch": 1.8856244579358195, "grad_norm": 0.2551032602787018, "learning_rate": 3.728057449586537e-05, "loss": 0.0262, "step": 17393 }, { "epoch": 1.8857328707718994, "grad_norm": 0.06226015090942383, "learning_rate": 3.7276947628028436e-05, "loss": 0.0036, "step": 17394 }, { "epoch": 1.8858412836079792, "grad_norm": 0.14866113662719727, "learning_rate": 3.72733207601915e-05, "loss": 0.0132, "step": 17395 }, { "epoch": 1.885949696444059, "grad_norm": 0.2709518074989319, "learning_rate": 3.7269693892354564e-05, "loss": 0.0103, "step": 17396 }, { "epoch": 1.8860581092801387, "grad_norm": 0.5078296065330505, "learning_rate": 3.726606702451763e-05, "loss": 0.0611, "step": 17397 }, { "epoch": 1.8861665221162185, "grad_norm": 0.25436171889305115, "learning_rate": 3.726244015668069e-05, "loss": 0.0132, "step": 17398 }, { "epoch": 1.8862749349522985, "grad_norm": 0.2056553065776825, "learning_rate": 3.7258813288843756e-05, "loss": 0.0202, "step": 17399 }, { "epoch": 1.886383347788378, "grad_norm": 0.5479859709739685, "learning_rate": 3.725518642100682e-05, "loss": 0.0406, "step": 17400 }, { "epoch": 1.886491760624458, "grad_norm": 0.13363981246948242, "learning_rate": 3.7251559553169885e-05, "loss": 0.0139, "step": 17401 }, { "epoch": 1.8866001734605378, "grad_norm": 0.3035263419151306, "learning_rate": 3.724793268533295e-05, "loss": 0.0127, "step": 17402 }, { "epoch": 1.8867085862966175, "grad_norm": 0.44351357221603394, "learning_rate": 3.724430581749601e-05, "loss": 0.0517, "step": 17403 }, { "epoch": 1.8868169991326973, "grad_norm": 0.12677955627441406, "learning_rate": 3.724067894965908e-05, "loss": 0.0055, "step": 17404 }, { "epoch": 1.886925411968777, "grad_norm": 0.40414366126060486, "learning_rate": 3.723705208182214e-05, "loss": 0.0345, "step": 17405 }, { "epoch": 1.887033824804857, "grad_norm": 0.5774035453796387, "learning_rate": 3.7233425213985205e-05, "loss": 0.0545, "step": 17406 }, { "epoch": 1.8871422376409366, "grad_norm": 0.4393514394760132, "learning_rate": 3.722979834614827e-05, "loss": 0.0151, "step": 17407 }, { "epoch": 1.8872506504770166, "grad_norm": 0.04000118002295494, "learning_rate": 3.722617147831133e-05, "loss": 0.0043, "step": 17408 }, { "epoch": 1.887359063313096, "grad_norm": 0.3114050328731537, "learning_rate": 3.72225446104744e-05, "loss": 0.0129, "step": 17409 }, { "epoch": 1.887467476149176, "grad_norm": 0.574192464351654, "learning_rate": 3.721891774263746e-05, "loss": 0.0244, "step": 17410 }, { "epoch": 1.8875758889852559, "grad_norm": 0.3070569336414337, "learning_rate": 3.7215290874800526e-05, "loss": 0.0287, "step": 17411 }, { "epoch": 1.8876843018213356, "grad_norm": 0.07348087430000305, "learning_rate": 3.721166400696359e-05, "loss": 0.0049, "step": 17412 }, { "epoch": 1.8877927146574154, "grad_norm": 0.3058561086654663, "learning_rate": 3.720803713912665e-05, "loss": 0.0179, "step": 17413 }, { "epoch": 1.8879011274934951, "grad_norm": 0.47391945123672485, "learning_rate": 3.720441027128971e-05, "loss": 0.0388, "step": 17414 }, { "epoch": 1.8880095403295751, "grad_norm": 0.4205503463745117, "learning_rate": 3.720078340345278e-05, "loss": 0.0298, "step": 17415 }, { "epoch": 1.8881179531656547, "grad_norm": 0.20636342465877533, "learning_rate": 3.7197156535615846e-05, "loss": 0.0061, "step": 17416 }, { "epoch": 1.8882263660017347, "grad_norm": 0.20732718706130981, "learning_rate": 3.719352966777891e-05, "loss": 0.0087, "step": 17417 }, { "epoch": 1.8883347788378144, "grad_norm": 0.19861271977424622, "learning_rate": 3.7189902799941974e-05, "loss": 0.009, "step": 17418 }, { "epoch": 1.8884431916738942, "grad_norm": 0.25597918033599854, "learning_rate": 3.718627593210504e-05, "loss": 0.0214, "step": 17419 }, { "epoch": 1.888551604509974, "grad_norm": 0.20910772681236267, "learning_rate": 3.7182649064268096e-05, "loss": 0.0179, "step": 17420 }, { "epoch": 1.8886600173460537, "grad_norm": 0.029379896819591522, "learning_rate": 3.717902219643116e-05, "loss": 0.0021, "step": 17421 }, { "epoch": 1.8887684301821337, "grad_norm": 0.25739094614982605, "learning_rate": 3.7175395328594224e-05, "loss": 0.0236, "step": 17422 }, { "epoch": 1.8888768430182132, "grad_norm": 0.14580971002578735, "learning_rate": 3.717176846075729e-05, "loss": 0.0123, "step": 17423 }, { "epoch": 1.8889852558542932, "grad_norm": 0.028773317113518715, "learning_rate": 3.716814159292036e-05, "loss": 0.0019, "step": 17424 }, { "epoch": 1.889093668690373, "grad_norm": 0.39399734139442444, "learning_rate": 3.716451472508342e-05, "loss": 0.0278, "step": 17425 }, { "epoch": 1.8892020815264527, "grad_norm": 0.3833235502243042, "learning_rate": 3.716088785724649e-05, "loss": 0.074, "step": 17426 }, { "epoch": 1.8893104943625325, "grad_norm": 0.11950745433568954, "learning_rate": 3.715726098940955e-05, "loss": 0.0051, "step": 17427 }, { "epoch": 1.8894189071986123, "grad_norm": 0.7671478986740112, "learning_rate": 3.715363412157261e-05, "loss": 0.0081, "step": 17428 }, { "epoch": 1.8895273200346923, "grad_norm": 0.2934705317020416, "learning_rate": 3.715000725373567e-05, "loss": 0.0167, "step": 17429 }, { "epoch": 1.8896357328707718, "grad_norm": 0.23598065972328186, "learning_rate": 3.714638038589874e-05, "loss": 0.0157, "step": 17430 }, { "epoch": 1.8897441457068518, "grad_norm": 0.03464808315038681, "learning_rate": 3.71427535180618e-05, "loss": 0.0035, "step": 17431 }, { "epoch": 1.8898525585429313, "grad_norm": 0.27711766958236694, "learning_rate": 3.713912665022487e-05, "loss": 0.0236, "step": 17432 }, { "epoch": 1.8899609713790113, "grad_norm": 0.23520584404468536, "learning_rate": 3.7135499782387936e-05, "loss": 0.0126, "step": 17433 }, { "epoch": 1.890069384215091, "grad_norm": 0.37333881855010986, "learning_rate": 3.7131872914551e-05, "loss": 0.0252, "step": 17434 }, { "epoch": 1.8901777970511708, "grad_norm": 0.4850892424583435, "learning_rate": 3.712824604671406e-05, "loss": 0.0237, "step": 17435 }, { "epoch": 1.8902862098872506, "grad_norm": 0.20958064496517181, "learning_rate": 3.712461917887712e-05, "loss": 0.0287, "step": 17436 }, { "epoch": 1.8903946227233304, "grad_norm": 0.027145983651280403, "learning_rate": 3.7120992311040186e-05, "loss": 0.0014, "step": 17437 }, { "epoch": 1.8905030355594103, "grad_norm": 0.2372352033853531, "learning_rate": 3.711736544320325e-05, "loss": 0.0153, "step": 17438 }, { "epoch": 1.89061144839549, "grad_norm": 0.695888876914978, "learning_rate": 3.7113738575366314e-05, "loss": 0.0645, "step": 17439 }, { "epoch": 1.8907198612315699, "grad_norm": 0.3419210910797119, "learning_rate": 3.711011170752938e-05, "loss": 0.0252, "step": 17440 }, { "epoch": 1.8908282740676496, "grad_norm": 0.32076025009155273, "learning_rate": 3.710648483969245e-05, "loss": 0.0372, "step": 17441 }, { "epoch": 1.8909366869037294, "grad_norm": 0.8578649163246155, "learning_rate": 3.7102857971855506e-05, "loss": 0.0839, "step": 17442 }, { "epoch": 1.8910450997398092, "grad_norm": 0.38530296087265015, "learning_rate": 3.709923110401857e-05, "loss": 0.0433, "step": 17443 }, { "epoch": 1.891153512575889, "grad_norm": 0.12013521045446396, "learning_rate": 3.7095604236181634e-05, "loss": 0.008, "step": 17444 }, { "epoch": 1.891261925411969, "grad_norm": 0.2765122652053833, "learning_rate": 3.70919773683447e-05, "loss": 0.0165, "step": 17445 }, { "epoch": 1.8913703382480485, "grad_norm": 0.09960035234689713, "learning_rate": 3.708835050050776e-05, "loss": 0.0071, "step": 17446 }, { "epoch": 1.8914787510841284, "grad_norm": 0.2791324257850647, "learning_rate": 3.7084723632670827e-05, "loss": 0.0496, "step": 17447 }, { "epoch": 1.8915871639202082, "grad_norm": 0.35846930742263794, "learning_rate": 3.708109676483389e-05, "loss": 0.0391, "step": 17448 }, { "epoch": 1.891695576756288, "grad_norm": 0.5033243298530579, "learning_rate": 3.7077469896996955e-05, "loss": 0.0243, "step": 17449 }, { "epoch": 1.8918039895923677, "grad_norm": 0.6150579452514648, "learning_rate": 3.707384302916002e-05, "loss": 0.0363, "step": 17450 }, { "epoch": 1.8919124024284475, "grad_norm": 0.574787974357605, "learning_rate": 3.707021616132308e-05, "loss": 0.0421, "step": 17451 }, { "epoch": 1.8920208152645275, "grad_norm": 0.0681903138756752, "learning_rate": 3.706658929348615e-05, "loss": 0.0051, "step": 17452 }, { "epoch": 1.892129228100607, "grad_norm": 0.5613524317741394, "learning_rate": 3.706296242564921e-05, "loss": 0.0286, "step": 17453 }, { "epoch": 1.892237640936687, "grad_norm": 0.19540543854236603, "learning_rate": 3.7059335557812275e-05, "loss": 0.0143, "step": 17454 }, { "epoch": 1.8923460537727665, "grad_norm": 0.37897029519081116, "learning_rate": 3.705570868997534e-05, "loss": 0.0243, "step": 17455 }, { "epoch": 1.8924544666088465, "grad_norm": 0.12033072113990784, "learning_rate": 3.7052081822138404e-05, "loss": 0.0131, "step": 17456 }, { "epoch": 1.8925628794449263, "grad_norm": 0.14428496360778809, "learning_rate": 3.704845495430147e-05, "loss": 0.0171, "step": 17457 }, { "epoch": 1.892671292281006, "grad_norm": 0.07554281502962112, "learning_rate": 3.704482808646453e-05, "loss": 0.0061, "step": 17458 }, { "epoch": 1.8927797051170858, "grad_norm": 0.12975482642650604, "learning_rate": 3.7041201218627596e-05, "loss": 0.0053, "step": 17459 }, { "epoch": 1.8928881179531656, "grad_norm": 0.35621705651283264, "learning_rate": 3.703757435079066e-05, "loss": 0.0174, "step": 17460 }, { "epoch": 1.8929965307892456, "grad_norm": 0.17952752113342285, "learning_rate": 3.7033947482953724e-05, "loss": 0.0227, "step": 17461 }, { "epoch": 1.893104943625325, "grad_norm": 0.3986053168773651, "learning_rate": 3.703032061511679e-05, "loss": 0.0588, "step": 17462 }, { "epoch": 1.893213356461405, "grad_norm": 0.7091889977455139, "learning_rate": 3.702669374727985e-05, "loss": 0.0198, "step": 17463 }, { "epoch": 1.8933217692974849, "grad_norm": 0.2938068211078644, "learning_rate": 3.7023066879442916e-05, "loss": 0.0365, "step": 17464 }, { "epoch": 1.8934301821335646, "grad_norm": 0.3499717712402344, "learning_rate": 3.7019440011605974e-05, "loss": 0.0422, "step": 17465 }, { "epoch": 1.8935385949696444, "grad_norm": 0.976920485496521, "learning_rate": 3.701581314376904e-05, "loss": 0.0217, "step": 17466 }, { "epoch": 1.8936470078057241, "grad_norm": 0.07631177455186844, "learning_rate": 3.701218627593211e-05, "loss": 0.0058, "step": 17467 }, { "epoch": 1.8937554206418041, "grad_norm": 0.1380593180656433, "learning_rate": 3.700855940809517e-05, "loss": 0.0097, "step": 17468 }, { "epoch": 1.8938638334778837, "grad_norm": 0.23891592025756836, "learning_rate": 3.700493254025824e-05, "loss": 0.0122, "step": 17469 }, { "epoch": 1.8939722463139637, "grad_norm": 0.30966782569885254, "learning_rate": 3.70013056724213e-05, "loss": 0.0109, "step": 17470 }, { "epoch": 1.8940806591500432, "grad_norm": 0.20464101433753967, "learning_rate": 3.6997678804584365e-05, "loss": 0.0131, "step": 17471 }, { "epoch": 1.8941890719861232, "grad_norm": 0.3197103440761566, "learning_rate": 3.699405193674743e-05, "loss": 0.0181, "step": 17472 }, { "epoch": 1.894297484822203, "grad_norm": 0.2727086544036865, "learning_rate": 3.6990425068910487e-05, "loss": 0.0236, "step": 17473 }, { "epoch": 1.8944058976582827, "grad_norm": 0.33828213810920715, "learning_rate": 3.698679820107355e-05, "loss": 0.0643, "step": 17474 }, { "epoch": 1.8945143104943627, "grad_norm": 0.20444175601005554, "learning_rate": 3.6983171333236615e-05, "loss": 0.0169, "step": 17475 }, { "epoch": 1.8946227233304422, "grad_norm": 0.621374785900116, "learning_rate": 3.6979544465399686e-05, "loss": 0.0365, "step": 17476 }, { "epoch": 1.8947311361665222, "grad_norm": 0.3503977954387665, "learning_rate": 3.697591759756275e-05, "loss": 0.0266, "step": 17477 }, { "epoch": 1.8948395490026018, "grad_norm": 0.5701529383659363, "learning_rate": 3.6972290729725814e-05, "loss": 0.0098, "step": 17478 }, { "epoch": 1.8949479618386817, "grad_norm": 0.08332644402980804, "learning_rate": 3.696866386188888e-05, "loss": 0.0103, "step": 17479 }, { "epoch": 1.8950563746747615, "grad_norm": 0.11723803728818893, "learning_rate": 3.6965036994051935e-05, "loss": 0.0129, "step": 17480 }, { "epoch": 1.8951647875108413, "grad_norm": 0.20629651844501495, "learning_rate": 3.6961410126215e-05, "loss": 0.0125, "step": 17481 }, { "epoch": 1.895273200346921, "grad_norm": 0.1065673902630806, "learning_rate": 3.6957783258378063e-05, "loss": 0.0053, "step": 17482 }, { "epoch": 1.8953816131830008, "grad_norm": 0.7274622321128845, "learning_rate": 3.695415639054113e-05, "loss": 0.044, "step": 17483 }, { "epoch": 1.8954900260190808, "grad_norm": 0.3368411958217621, "learning_rate": 3.69505295227042e-05, "loss": 0.0309, "step": 17484 }, { "epoch": 1.8955984388551603, "grad_norm": 0.678869903087616, "learning_rate": 3.694690265486726e-05, "loss": 0.0297, "step": 17485 }, { "epoch": 1.8957068516912403, "grad_norm": 0.44833406805992126, "learning_rate": 3.694327578703033e-05, "loss": 0.0418, "step": 17486 }, { "epoch": 1.89581526452732, "grad_norm": 0.21908216178417206, "learning_rate": 3.6939648919193384e-05, "loss": 0.0226, "step": 17487 }, { "epoch": 1.8959236773633998, "grad_norm": 0.09155870974063873, "learning_rate": 3.693602205135645e-05, "loss": 0.0077, "step": 17488 }, { "epoch": 1.8960320901994796, "grad_norm": 0.18997174501419067, "learning_rate": 3.693239518351951e-05, "loss": 0.0159, "step": 17489 }, { "epoch": 1.8961405030355594, "grad_norm": 0.26235049962997437, "learning_rate": 3.6928768315682576e-05, "loss": 0.0395, "step": 17490 }, { "epoch": 1.8962489158716394, "grad_norm": 0.6401994824409485, "learning_rate": 3.692514144784564e-05, "loss": 0.0308, "step": 17491 }, { "epoch": 1.896357328707719, "grad_norm": 0.4129992723464966, "learning_rate": 3.6921514580008705e-05, "loss": 0.0307, "step": 17492 }, { "epoch": 1.8964657415437989, "grad_norm": 0.29341667890548706, "learning_rate": 3.6917887712171775e-05, "loss": 0.0204, "step": 17493 }, { "epoch": 1.8965741543798784, "grad_norm": 0.27538472414016724, "learning_rate": 3.691426084433483e-05, "loss": 0.0178, "step": 17494 }, { "epoch": 1.8966825672159584, "grad_norm": 0.3890049457550049, "learning_rate": 3.69106339764979e-05, "loss": 0.0299, "step": 17495 }, { "epoch": 1.8967909800520382, "grad_norm": 0.3893648684024811, "learning_rate": 3.690700710866096e-05, "loss": 0.0196, "step": 17496 }, { "epoch": 1.896899392888118, "grad_norm": 0.1247413158416748, "learning_rate": 3.6903380240824025e-05, "loss": 0.0127, "step": 17497 }, { "epoch": 1.8970078057241977, "grad_norm": 0.01859944313764572, "learning_rate": 3.689975337298709e-05, "loss": 0.0009, "step": 17498 }, { "epoch": 1.8971162185602775, "grad_norm": 0.47413212060928345, "learning_rate": 3.689612650515015e-05, "loss": 0.0154, "step": 17499 }, { "epoch": 1.8972246313963574, "grad_norm": 0.06591970473527908, "learning_rate": 3.689249963731322e-05, "loss": 0.0039, "step": 17500 }, { "epoch": 1.897333044232437, "grad_norm": 0.3667375445365906, "learning_rate": 3.688887276947628e-05, "loss": 0.0128, "step": 17501 }, { "epoch": 1.897441457068517, "grad_norm": 0.01531523559242487, "learning_rate": 3.6885245901639346e-05, "loss": 0.0008, "step": 17502 }, { "epoch": 1.8975498699045967, "grad_norm": 0.3911955952644348, "learning_rate": 3.688161903380241e-05, "loss": 0.0623, "step": 17503 }, { "epoch": 1.8976582827406765, "grad_norm": 0.18144723773002625, "learning_rate": 3.6877992165965474e-05, "loss": 0.008, "step": 17504 }, { "epoch": 1.8977666955767563, "grad_norm": 0.33686572313308716, "learning_rate": 3.687436529812854e-05, "loss": 0.0207, "step": 17505 }, { "epoch": 1.897875108412836, "grad_norm": 0.44439569115638733, "learning_rate": 3.68707384302916e-05, "loss": 0.0416, "step": 17506 }, { "epoch": 1.897983521248916, "grad_norm": 0.13133251667022705, "learning_rate": 3.6867111562454666e-05, "loss": 0.0119, "step": 17507 }, { "epoch": 1.8980919340849955, "grad_norm": 0.16538378596305847, "learning_rate": 3.686348469461773e-05, "loss": 0.0046, "step": 17508 }, { "epoch": 1.8982003469210755, "grad_norm": 0.3790701627731323, "learning_rate": 3.6859857826780794e-05, "loss": 0.0346, "step": 17509 }, { "epoch": 1.8983087597571553, "grad_norm": 0.6911110281944275, "learning_rate": 3.685623095894386e-05, "loss": 0.0496, "step": 17510 }, { "epoch": 1.898417172593235, "grad_norm": 0.8721823692321777, "learning_rate": 3.685260409110692e-05, "loss": 0.0555, "step": 17511 }, { "epoch": 1.8985255854293148, "grad_norm": 0.4976206421852112, "learning_rate": 3.684897722326999e-05, "loss": 0.0423, "step": 17512 }, { "epoch": 1.8986339982653946, "grad_norm": 0.13426457345485687, "learning_rate": 3.684535035543305e-05, "loss": 0.0055, "step": 17513 }, { "epoch": 1.8987424111014746, "grad_norm": 0.32811713218688965, "learning_rate": 3.6841723487596115e-05, "loss": 0.0182, "step": 17514 }, { "epoch": 1.898850823937554, "grad_norm": 0.26768162846565247, "learning_rate": 3.683809661975918e-05, "loss": 0.0229, "step": 17515 }, { "epoch": 1.898959236773634, "grad_norm": 0.39072051644325256, "learning_rate": 3.683446975192224e-05, "loss": 0.0191, "step": 17516 }, { "epoch": 1.8990676496097136, "grad_norm": 0.15296036005020142, "learning_rate": 3.68308428840853e-05, "loss": 0.0105, "step": 17517 }, { "epoch": 1.8991760624457936, "grad_norm": 0.14813074469566345, "learning_rate": 3.6827216016248364e-05, "loss": 0.0121, "step": 17518 }, { "epoch": 1.8992844752818734, "grad_norm": 0.4393932819366455, "learning_rate": 3.6823589148411435e-05, "loss": 0.0202, "step": 17519 }, { "epoch": 1.8993928881179531, "grad_norm": 0.6622102856636047, "learning_rate": 3.68199622805745e-05, "loss": 0.0523, "step": 17520 }, { "epoch": 1.899501300954033, "grad_norm": 0.6510636806488037, "learning_rate": 3.6816335412737564e-05, "loss": 0.0431, "step": 17521 }, { "epoch": 1.8996097137901127, "grad_norm": 0.42302700877189636, "learning_rate": 3.681270854490063e-05, "loss": 0.0427, "step": 17522 }, { "epoch": 1.8997181266261927, "grad_norm": 0.44533979892730713, "learning_rate": 3.680908167706369e-05, "loss": 0.034, "step": 17523 }, { "epoch": 1.8998265394622722, "grad_norm": 0.07145385444164276, "learning_rate": 3.6805454809226756e-05, "loss": 0.0143, "step": 17524 }, { "epoch": 1.8999349522983522, "grad_norm": 0.2823024094104767, "learning_rate": 3.680182794138981e-05, "loss": 0.071, "step": 17525 }, { "epoch": 1.900043365134432, "grad_norm": 0.2839260399341583, "learning_rate": 3.679820107355288e-05, "loss": 0.0111, "step": 17526 }, { "epoch": 1.9001517779705117, "grad_norm": 0.09712380915880203, "learning_rate": 3.679457420571595e-05, "loss": 0.0092, "step": 17527 }, { "epoch": 1.9002601908065915, "grad_norm": 0.3344566524028778, "learning_rate": 3.679094733787901e-05, "loss": 0.0489, "step": 17528 }, { "epoch": 1.9003686036426712, "grad_norm": 0.2600385844707489, "learning_rate": 3.6787320470042076e-05, "loss": 0.0101, "step": 17529 }, { "epoch": 1.9004770164787512, "grad_norm": 0.475355863571167, "learning_rate": 3.678369360220514e-05, "loss": 0.0464, "step": 17530 }, { "epoch": 1.9005854293148308, "grad_norm": 0.02793184481561184, "learning_rate": 3.6780066734368205e-05, "loss": 0.0021, "step": 17531 }, { "epoch": 1.9006938421509108, "grad_norm": 0.1567995697259903, "learning_rate": 3.677643986653126e-05, "loss": 0.0198, "step": 17532 }, { "epoch": 1.9008022549869905, "grad_norm": 0.09857764840126038, "learning_rate": 3.6772812998694326e-05, "loss": 0.0071, "step": 17533 }, { "epoch": 1.9009106678230703, "grad_norm": 0.160963773727417, "learning_rate": 3.676918613085739e-05, "loss": 0.0106, "step": 17534 }, { "epoch": 1.90101908065915, "grad_norm": 0.09798602014780045, "learning_rate": 3.6765559263020454e-05, "loss": 0.0047, "step": 17535 }, { "epoch": 1.9011274934952298, "grad_norm": 0.1186971589922905, "learning_rate": 3.6761932395183525e-05, "loss": 0.0049, "step": 17536 }, { "epoch": 1.9012359063313098, "grad_norm": 0.3059970736503601, "learning_rate": 3.675830552734659e-05, "loss": 0.0459, "step": 17537 }, { "epoch": 1.9013443191673893, "grad_norm": 0.10096029192209244, "learning_rate": 3.675467865950965e-05, "loss": 0.0102, "step": 17538 }, { "epoch": 1.9014527320034693, "grad_norm": 0.10444816946983337, "learning_rate": 3.675105179167271e-05, "loss": 0.006, "step": 17539 }, { "epoch": 1.9015611448395489, "grad_norm": 0.3592434823513031, "learning_rate": 3.6747424923835775e-05, "loss": 0.01, "step": 17540 }, { "epoch": 1.9016695576756288, "grad_norm": 0.3091311752796173, "learning_rate": 3.674379805599884e-05, "loss": 0.0306, "step": 17541 }, { "epoch": 1.9017779705117086, "grad_norm": 0.4923139214515686, "learning_rate": 3.67401711881619e-05, "loss": 0.0357, "step": 17542 }, { "epoch": 1.9018863833477884, "grad_norm": 0.34327220916748047, "learning_rate": 3.673654432032497e-05, "loss": 0.0437, "step": 17543 }, { "epoch": 1.9019947961838681, "grad_norm": 0.5485096573829651, "learning_rate": 3.673291745248803e-05, "loss": 0.0324, "step": 17544 }, { "epoch": 1.902103209019948, "grad_norm": 0.08280451595783234, "learning_rate": 3.67292905846511e-05, "loss": 0.0149, "step": 17545 }, { "epoch": 1.9022116218560279, "grad_norm": 0.14636127650737762, "learning_rate": 3.672566371681416e-05, "loss": 0.0087, "step": 17546 }, { "epoch": 1.9023200346921074, "grad_norm": 0.49776431918144226, "learning_rate": 3.6722036848977224e-05, "loss": 0.0425, "step": 17547 }, { "epoch": 1.9024284475281874, "grad_norm": 0.2606521546840668, "learning_rate": 3.671840998114029e-05, "loss": 0.0113, "step": 17548 }, { "epoch": 1.9025368603642672, "grad_norm": 0.2648662030696869, "learning_rate": 3.671478311330335e-05, "loss": 0.0104, "step": 17549 }, { "epoch": 1.902645273200347, "grad_norm": 0.3014202117919922, "learning_rate": 3.6711156245466416e-05, "loss": 0.0319, "step": 17550 }, { "epoch": 1.9027536860364267, "grad_norm": 0.430823415517807, "learning_rate": 3.670752937762948e-05, "loss": 0.0545, "step": 17551 }, { "epoch": 1.9028620988725065, "grad_norm": 0.04481705650687218, "learning_rate": 3.6703902509792544e-05, "loss": 0.0024, "step": 17552 }, { "epoch": 1.9029705117085864, "grad_norm": 0.3013642430305481, "learning_rate": 3.670027564195561e-05, "loss": 0.0206, "step": 17553 }, { "epoch": 1.903078924544666, "grad_norm": 0.204486683011055, "learning_rate": 3.669664877411867e-05, "loss": 0.0124, "step": 17554 }, { "epoch": 1.903187337380746, "grad_norm": 0.2635554373264313, "learning_rate": 3.6693021906281736e-05, "loss": 0.0104, "step": 17555 }, { "epoch": 1.9032957502168255, "grad_norm": 0.3062429428100586, "learning_rate": 3.66893950384448e-05, "loss": 0.0194, "step": 17556 }, { "epoch": 1.9034041630529055, "grad_norm": 0.4859842360019684, "learning_rate": 3.6685768170607865e-05, "loss": 0.0311, "step": 17557 }, { "epoch": 1.9035125758889853, "grad_norm": 0.1946599781513214, "learning_rate": 3.668214130277093e-05, "loss": 0.0107, "step": 17558 }, { "epoch": 1.903620988725065, "grad_norm": 0.1712275743484497, "learning_rate": 3.667851443493399e-05, "loss": 0.0097, "step": 17559 }, { "epoch": 1.9037294015611448, "grad_norm": 0.3378208875656128, "learning_rate": 3.667488756709706e-05, "loss": 0.0676, "step": 17560 }, { "epoch": 1.9038378143972245, "grad_norm": 0.1945417821407318, "learning_rate": 3.667126069926012e-05, "loss": 0.0153, "step": 17561 }, { "epoch": 1.9039462272333045, "grad_norm": 0.26281702518463135, "learning_rate": 3.6667633831423185e-05, "loss": 0.0166, "step": 17562 }, { "epoch": 1.904054640069384, "grad_norm": 0.33619406819343567, "learning_rate": 3.666400696358625e-05, "loss": 0.0161, "step": 17563 }, { "epoch": 1.904163052905464, "grad_norm": 0.3908732831478119, "learning_rate": 3.666038009574931e-05, "loss": 0.028, "step": 17564 }, { "epoch": 1.9042714657415438, "grad_norm": 0.7073009610176086, "learning_rate": 3.665675322791238e-05, "loss": 0.0689, "step": 17565 }, { "epoch": 1.9043798785776236, "grad_norm": 0.3661779463291168, "learning_rate": 3.665312636007544e-05, "loss": 0.0331, "step": 17566 }, { "epoch": 1.9044882914137033, "grad_norm": 0.24954162538051605, "learning_rate": 3.6649499492238506e-05, "loss": 0.0263, "step": 17567 }, { "epoch": 1.9045967042497831, "grad_norm": 0.26338884234428406, "learning_rate": 3.664587262440157e-05, "loss": 0.0266, "step": 17568 }, { "epoch": 1.904705117085863, "grad_norm": 0.20612230896949768, "learning_rate": 3.664224575656463e-05, "loss": 0.0136, "step": 17569 }, { "epoch": 1.9048135299219426, "grad_norm": 0.38131657242774963, "learning_rate": 3.663861888872769e-05, "loss": 0.0572, "step": 17570 }, { "epoch": 1.9049219427580226, "grad_norm": 0.3760582208633423, "learning_rate": 3.663499202089076e-05, "loss": 0.0438, "step": 17571 }, { "epoch": 1.9050303555941024, "grad_norm": 0.6590770483016968, "learning_rate": 3.6631365153053826e-05, "loss": 0.0497, "step": 17572 }, { "epoch": 1.9051387684301822, "grad_norm": 0.1931842863559723, "learning_rate": 3.662773828521689e-05, "loss": 0.0108, "step": 17573 }, { "epoch": 1.905247181266262, "grad_norm": 0.11256013065576553, "learning_rate": 3.6624111417379954e-05, "loss": 0.0134, "step": 17574 }, { "epoch": 1.9053555941023417, "grad_norm": 0.4227111339569092, "learning_rate": 3.662048454954302e-05, "loss": 0.0261, "step": 17575 }, { "epoch": 1.9054640069384217, "grad_norm": 0.40711089968681335, "learning_rate": 3.661685768170608e-05, "loss": 0.0199, "step": 17576 }, { "epoch": 1.9055724197745012, "grad_norm": 0.1285380870103836, "learning_rate": 3.661323081386914e-05, "loss": 0.0053, "step": 17577 }, { "epoch": 1.9056808326105812, "grad_norm": 0.3088739812374115, "learning_rate": 3.6609603946032204e-05, "loss": 0.0116, "step": 17578 }, { "epoch": 1.9057892454466607, "grad_norm": 0.24249586462974548, "learning_rate": 3.6605977078195275e-05, "loss": 0.0254, "step": 17579 }, { "epoch": 1.9058976582827407, "grad_norm": 0.2530665099620819, "learning_rate": 3.660235021035834e-05, "loss": 0.0188, "step": 17580 }, { "epoch": 1.9060060711188205, "grad_norm": 0.11397404968738556, "learning_rate": 3.65987233425214e-05, "loss": 0.0126, "step": 17581 }, { "epoch": 1.9061144839549002, "grad_norm": 0.29408717155456543, "learning_rate": 3.659509647468447e-05, "loss": 0.0147, "step": 17582 }, { "epoch": 1.90622289679098, "grad_norm": 0.6073276400566101, "learning_rate": 3.659146960684753e-05, "loss": 0.0489, "step": 17583 }, { "epoch": 1.9063313096270598, "grad_norm": 0.3238656520843506, "learning_rate": 3.658784273901059e-05, "loss": 0.0228, "step": 17584 }, { "epoch": 1.9064397224631398, "grad_norm": 0.4550282061100006, "learning_rate": 3.658421587117365e-05, "loss": 0.0092, "step": 17585 }, { "epoch": 1.9065481352992193, "grad_norm": 0.8561038374900818, "learning_rate": 3.658058900333672e-05, "loss": 0.0633, "step": 17586 }, { "epoch": 1.9066565481352993, "grad_norm": 0.33921146392822266, "learning_rate": 3.657696213549978e-05, "loss": 0.0187, "step": 17587 }, { "epoch": 1.906764960971379, "grad_norm": 0.5801106691360474, "learning_rate": 3.657333526766285e-05, "loss": 0.0734, "step": 17588 }, { "epoch": 1.9068733738074588, "grad_norm": 0.6772412657737732, "learning_rate": 3.6569708399825916e-05, "loss": 0.0253, "step": 17589 }, { "epoch": 1.9069817866435386, "grad_norm": 0.6991814970970154, "learning_rate": 3.656608153198898e-05, "loss": 0.0769, "step": 17590 }, { "epoch": 1.9070901994796183, "grad_norm": 0.21749182045459747, "learning_rate": 3.656245466415204e-05, "loss": 0.0152, "step": 17591 }, { "epoch": 1.9071986123156983, "grad_norm": 0.13483206927776337, "learning_rate": 3.65588277963151e-05, "loss": 0.0096, "step": 17592 }, { "epoch": 1.9073070251517779, "grad_norm": 0.03620819374918938, "learning_rate": 3.6555200928478166e-05, "loss": 0.002, "step": 17593 }, { "epoch": 1.9074154379878578, "grad_norm": 0.20578868687152863, "learning_rate": 3.655157406064123e-05, "loss": 0.0244, "step": 17594 }, { "epoch": 1.9075238508239376, "grad_norm": 0.13232304155826569, "learning_rate": 3.6547947192804294e-05, "loss": 0.0131, "step": 17595 }, { "epoch": 1.9076322636600174, "grad_norm": 0.27904629707336426, "learning_rate": 3.6544320324967365e-05, "loss": 0.0217, "step": 17596 }, { "epoch": 1.9077406764960971, "grad_norm": 0.06722723692655563, "learning_rate": 3.654069345713043e-05, "loss": 0.0033, "step": 17597 }, { "epoch": 1.907849089332177, "grad_norm": 0.08732295781373978, "learning_rate": 3.6537066589293486e-05, "loss": 0.0073, "step": 17598 }, { "epoch": 1.9079575021682569, "grad_norm": 0.5643523335456848, "learning_rate": 3.653343972145655e-05, "loss": 0.0266, "step": 17599 }, { "epoch": 1.9080659150043364, "grad_norm": 0.1416211724281311, "learning_rate": 3.6529812853619614e-05, "loss": 0.0114, "step": 17600 }, { "epoch": 1.9081743278404164, "grad_norm": 0.07609901577234268, "learning_rate": 3.652618598578268e-05, "loss": 0.0083, "step": 17601 }, { "epoch": 1.908282740676496, "grad_norm": 0.15218999981880188, "learning_rate": 3.652255911794574e-05, "loss": 0.0125, "step": 17602 }, { "epoch": 1.908391153512576, "grad_norm": 0.42388081550598145, "learning_rate": 3.651893225010881e-05, "loss": 0.0342, "step": 17603 }, { "epoch": 1.9084995663486557, "grad_norm": 0.2981821894645691, "learning_rate": 3.651530538227187e-05, "loss": 0.019, "step": 17604 }, { "epoch": 1.9086079791847355, "grad_norm": 0.20404234528541565, "learning_rate": 3.651167851443494e-05, "loss": 0.0127, "step": 17605 }, { "epoch": 1.9087163920208152, "grad_norm": 1.0107996463775635, "learning_rate": 3.6508051646598e-05, "loss": 0.0815, "step": 17606 }, { "epoch": 1.908824804856895, "grad_norm": 0.08046919852495193, "learning_rate": 3.650442477876106e-05, "loss": 0.0047, "step": 17607 }, { "epoch": 1.908933217692975, "grad_norm": 0.29175451397895813, "learning_rate": 3.650079791092413e-05, "loss": 0.0071, "step": 17608 }, { "epoch": 1.9090416305290545, "grad_norm": 0.1338047832250595, "learning_rate": 3.649717104308719e-05, "loss": 0.0138, "step": 17609 }, { "epoch": 1.9091500433651345, "grad_norm": 0.2731681168079376, "learning_rate": 3.6493544175250255e-05, "loss": 0.0231, "step": 17610 }, { "epoch": 1.9092584562012143, "grad_norm": 0.15807883441448212, "learning_rate": 3.648991730741332e-05, "loss": 0.016, "step": 17611 }, { "epoch": 1.909366869037294, "grad_norm": 0.11493819952011108, "learning_rate": 3.6486290439576384e-05, "loss": 0.0084, "step": 17612 }, { "epoch": 1.9094752818733738, "grad_norm": 0.36209240555763245, "learning_rate": 3.648266357173945e-05, "loss": 0.032, "step": 17613 }, { "epoch": 1.9095836947094535, "grad_norm": 0.6882893443107605, "learning_rate": 3.647903670390251e-05, "loss": 0.0605, "step": 17614 }, { "epoch": 1.9096921075455335, "grad_norm": 0.5817270278930664, "learning_rate": 3.6475409836065576e-05, "loss": 0.049, "step": 17615 }, { "epoch": 1.909800520381613, "grad_norm": 0.25028473138809204, "learning_rate": 3.647178296822864e-05, "loss": 0.0161, "step": 17616 }, { "epoch": 1.909908933217693, "grad_norm": 0.599592387676239, "learning_rate": 3.6468156100391704e-05, "loss": 0.0583, "step": 17617 }, { "epoch": 1.9100173460537726, "grad_norm": 0.0772780179977417, "learning_rate": 3.646452923255477e-05, "loss": 0.0106, "step": 17618 }, { "epoch": 1.9101257588898526, "grad_norm": 0.4415279030799866, "learning_rate": 3.646090236471783e-05, "loss": 0.0254, "step": 17619 }, { "epoch": 1.9102341717259324, "grad_norm": 0.10304926335811615, "learning_rate": 3.6457275496880896e-05, "loss": 0.0039, "step": 17620 }, { "epoch": 1.9103425845620121, "grad_norm": 0.10129610449075699, "learning_rate": 3.645364862904396e-05, "loss": 0.0089, "step": 17621 }, { "epoch": 1.910450997398092, "grad_norm": 0.652520477771759, "learning_rate": 3.6450021761207025e-05, "loss": 0.0555, "step": 17622 }, { "epoch": 1.9105594102341716, "grad_norm": 0.2145688831806183, "learning_rate": 3.644639489337009e-05, "loss": 0.0106, "step": 17623 }, { "epoch": 1.9106678230702516, "grad_norm": 0.41137516498565674, "learning_rate": 3.644276802553315e-05, "loss": 0.0381, "step": 17624 }, { "epoch": 1.9107762359063312, "grad_norm": 0.5078140497207642, "learning_rate": 3.643914115769622e-05, "loss": 0.033, "step": 17625 }, { "epoch": 1.9108846487424112, "grad_norm": 0.3349344730377197, "learning_rate": 3.643551428985928e-05, "loss": 0.0318, "step": 17626 }, { "epoch": 1.910993061578491, "grad_norm": 0.07751034200191498, "learning_rate": 3.6431887422022345e-05, "loss": 0.0056, "step": 17627 }, { "epoch": 1.9111014744145707, "grad_norm": 0.28453561663627625, "learning_rate": 3.642826055418541e-05, "loss": 0.0247, "step": 17628 }, { "epoch": 1.9112098872506504, "grad_norm": 0.14964762330055237, "learning_rate": 3.6424633686348467e-05, "loss": 0.0096, "step": 17629 }, { "epoch": 1.9113183000867302, "grad_norm": 0.10218265652656555, "learning_rate": 3.642100681851153e-05, "loss": 0.006, "step": 17630 }, { "epoch": 1.9114267129228102, "grad_norm": 0.18349163234233856, "learning_rate": 3.64173799506746e-05, "loss": 0.0193, "step": 17631 }, { "epoch": 1.9115351257588897, "grad_norm": 0.19548913836479187, "learning_rate": 3.6413753082837666e-05, "loss": 0.0122, "step": 17632 }, { "epoch": 1.9116435385949697, "grad_norm": 0.3725823760032654, "learning_rate": 3.641012621500073e-05, "loss": 0.0176, "step": 17633 }, { "epoch": 1.9117519514310495, "grad_norm": 0.16939768195152283, "learning_rate": 3.6406499347163794e-05, "loss": 0.0137, "step": 17634 }, { "epoch": 1.9118603642671292, "grad_norm": 0.13442707061767578, "learning_rate": 3.640287247932686e-05, "loss": 0.0054, "step": 17635 }, { "epoch": 1.911968777103209, "grad_norm": 0.5734000205993652, "learning_rate": 3.6399245611489915e-05, "loss": 0.028, "step": 17636 }, { "epoch": 1.9120771899392888, "grad_norm": 1.0734684467315674, "learning_rate": 3.639561874365298e-05, "loss": 0.0545, "step": 17637 }, { "epoch": 1.9121856027753688, "grad_norm": 0.25971171259880066, "learning_rate": 3.6391991875816044e-05, "loss": 0.031, "step": 17638 }, { "epoch": 1.9122940156114483, "grad_norm": 0.1948702037334442, "learning_rate": 3.638836500797911e-05, "loss": 0.0107, "step": 17639 }, { "epoch": 1.9124024284475283, "grad_norm": 0.40334975719451904, "learning_rate": 3.638473814014218e-05, "loss": 0.0398, "step": 17640 }, { "epoch": 1.9125108412836078, "grad_norm": 0.21488098800182343, "learning_rate": 3.638111127230524e-05, "loss": 0.033, "step": 17641 }, { "epoch": 1.9126192541196878, "grad_norm": 0.5735954642295837, "learning_rate": 3.637748440446831e-05, "loss": 0.0284, "step": 17642 }, { "epoch": 1.9127276669557676, "grad_norm": 0.8990463018417358, "learning_rate": 3.6373857536631364e-05, "loss": 0.0323, "step": 17643 }, { "epoch": 1.9128360797918473, "grad_norm": 0.17215131223201752, "learning_rate": 3.637023066879443e-05, "loss": 0.0111, "step": 17644 }, { "epoch": 1.912944492627927, "grad_norm": 0.2678224444389343, "learning_rate": 3.636660380095749e-05, "loss": 0.0215, "step": 17645 }, { "epoch": 1.9130529054640069, "grad_norm": 0.20196771621704102, "learning_rate": 3.6362976933120556e-05, "loss": 0.0199, "step": 17646 }, { "epoch": 1.9131613183000868, "grad_norm": 0.5917120575904846, "learning_rate": 3.635935006528362e-05, "loss": 0.108, "step": 17647 }, { "epoch": 1.9132697311361664, "grad_norm": 0.1662321239709854, "learning_rate": 3.635572319744669e-05, "loss": 0.0173, "step": 17648 }, { "epoch": 1.9133781439722464, "grad_norm": 0.3195350468158722, "learning_rate": 3.6352096329609755e-05, "loss": 0.0256, "step": 17649 }, { "epoch": 1.9134865568083261, "grad_norm": 0.17689745128154755, "learning_rate": 3.634846946177281e-05, "loss": 0.011, "step": 17650 }, { "epoch": 1.913594969644406, "grad_norm": 0.33606693148612976, "learning_rate": 3.634484259393588e-05, "loss": 0.0276, "step": 17651 }, { "epoch": 1.9137033824804857, "grad_norm": 0.12542608380317688, "learning_rate": 3.634121572609894e-05, "loss": 0.0051, "step": 17652 }, { "epoch": 1.9138117953165654, "grad_norm": 0.1596822440624237, "learning_rate": 3.6337588858262005e-05, "loss": 0.0115, "step": 17653 }, { "epoch": 1.9139202081526454, "grad_norm": 0.23377357423305511, "learning_rate": 3.633396199042507e-05, "loss": 0.0177, "step": 17654 }, { "epoch": 1.914028620988725, "grad_norm": 0.12384834885597229, "learning_rate": 3.633033512258813e-05, "loss": 0.0041, "step": 17655 }, { "epoch": 1.914137033824805, "grad_norm": 0.5118054151535034, "learning_rate": 3.63267082547512e-05, "loss": 0.0573, "step": 17656 }, { "epoch": 1.9142454466608847, "grad_norm": 0.23774926364421844, "learning_rate": 3.632308138691427e-05, "loss": 0.0539, "step": 17657 }, { "epoch": 1.9143538594969645, "grad_norm": 0.13746513426303864, "learning_rate": 3.6319454519077326e-05, "loss": 0.0143, "step": 17658 }, { "epoch": 1.9144622723330442, "grad_norm": 0.07294177263975143, "learning_rate": 3.631582765124039e-05, "loss": 0.0036, "step": 17659 }, { "epoch": 1.914570685169124, "grad_norm": 0.20960979163646698, "learning_rate": 3.6312200783403454e-05, "loss": 0.0329, "step": 17660 }, { "epoch": 1.914679098005204, "grad_norm": 0.3826886713504791, "learning_rate": 3.630857391556652e-05, "loss": 0.0332, "step": 17661 }, { "epoch": 1.9147875108412835, "grad_norm": 0.43407461047172546, "learning_rate": 3.630494704772958e-05, "loss": 0.0286, "step": 17662 }, { "epoch": 1.9148959236773635, "grad_norm": 0.16556404531002045, "learning_rate": 3.6301320179892646e-05, "loss": 0.0072, "step": 17663 }, { "epoch": 1.915004336513443, "grad_norm": 0.30233192443847656, "learning_rate": 3.629769331205571e-05, "loss": 0.016, "step": 17664 }, { "epoch": 1.915112749349523, "grad_norm": 0.8951980471611023, "learning_rate": 3.6294066444218774e-05, "loss": 0.0613, "step": 17665 }, { "epoch": 1.9152211621856028, "grad_norm": 0.6564478278160095, "learning_rate": 3.629043957638184e-05, "loss": 0.0921, "step": 17666 }, { "epoch": 1.9153295750216826, "grad_norm": 0.5965380072593689, "learning_rate": 3.62868127085449e-05, "loss": 0.02, "step": 17667 }, { "epoch": 1.9154379878577623, "grad_norm": 0.25794467329978943, "learning_rate": 3.628318584070797e-05, "loss": 0.0254, "step": 17668 }, { "epoch": 1.915546400693842, "grad_norm": 0.16298364102840424, "learning_rate": 3.627955897287103e-05, "loss": 0.0175, "step": 17669 }, { "epoch": 1.915654813529922, "grad_norm": 0.7299712300300598, "learning_rate": 3.6275932105034095e-05, "loss": 0.0205, "step": 17670 }, { "epoch": 1.9157632263660016, "grad_norm": 0.2556162476539612, "learning_rate": 3.627230523719716e-05, "loss": 0.0331, "step": 17671 }, { "epoch": 1.9158716392020816, "grad_norm": 0.17768308520317078, "learning_rate": 3.626867836936022e-05, "loss": 0.0057, "step": 17672 }, { "epoch": 1.9159800520381614, "grad_norm": 0.3794786036014557, "learning_rate": 3.626505150152329e-05, "loss": 0.0434, "step": 17673 }, { "epoch": 1.9160884648742411, "grad_norm": 0.3403984010219574, "learning_rate": 3.626142463368635e-05, "loss": 0.0358, "step": 17674 }, { "epoch": 1.9161968777103209, "grad_norm": 1.0194728374481201, "learning_rate": 3.6257797765849415e-05, "loss": 0.0549, "step": 17675 }, { "epoch": 1.9163052905464006, "grad_norm": 0.36065003275871277, "learning_rate": 3.625417089801248e-05, "loss": 0.0089, "step": 17676 }, { "epoch": 1.9164137033824806, "grad_norm": 0.37504521012306213, "learning_rate": 3.6250544030175544e-05, "loss": 0.0379, "step": 17677 }, { "epoch": 1.9165221162185602, "grad_norm": 0.2870925962924957, "learning_rate": 3.624691716233861e-05, "loss": 0.033, "step": 17678 }, { "epoch": 1.9166305290546402, "grad_norm": 0.3511714041233063, "learning_rate": 3.624329029450167e-05, "loss": 0.0166, "step": 17679 }, { "epoch": 1.91673894189072, "grad_norm": 0.16503407061100006, "learning_rate": 3.6239663426664736e-05, "loss": 0.0101, "step": 17680 }, { "epoch": 1.9168473547267997, "grad_norm": 0.3145829737186432, "learning_rate": 3.623603655882779e-05, "loss": 0.021, "step": 17681 }, { "epoch": 1.9169557675628794, "grad_norm": 0.27141594886779785, "learning_rate": 3.623240969099086e-05, "loss": 0.0215, "step": 17682 }, { "epoch": 1.9170641803989592, "grad_norm": 0.45384347438812256, "learning_rate": 3.622878282315393e-05, "loss": 0.0288, "step": 17683 }, { "epoch": 1.9171725932350392, "grad_norm": 0.16474740207195282, "learning_rate": 3.622515595531699e-05, "loss": 0.0189, "step": 17684 }, { "epoch": 1.9172810060711187, "grad_norm": 0.20572634041309357, "learning_rate": 3.6221529087480056e-05, "loss": 0.0165, "step": 17685 }, { "epoch": 1.9173894189071987, "grad_norm": 0.19266538321971893, "learning_rate": 3.621790221964312e-05, "loss": 0.0182, "step": 17686 }, { "epoch": 1.9174978317432783, "grad_norm": 0.03931727632880211, "learning_rate": 3.6214275351806185e-05, "loss": 0.0039, "step": 17687 }, { "epoch": 1.9176062445793582, "grad_norm": 0.10192891210317612, "learning_rate": 3.621064848396924e-05, "loss": 0.0074, "step": 17688 }, { "epoch": 1.917714657415438, "grad_norm": 0.16375842690467834, "learning_rate": 3.6207021616132306e-05, "loss": 0.005, "step": 17689 }, { "epoch": 1.9178230702515178, "grad_norm": 0.4836276173591614, "learning_rate": 3.620339474829537e-05, "loss": 0.0392, "step": 17690 }, { "epoch": 1.9179314830875975, "grad_norm": 0.5722072720527649, "learning_rate": 3.619976788045844e-05, "loss": 0.0832, "step": 17691 }, { "epoch": 1.9180398959236773, "grad_norm": 0.199371799826622, "learning_rate": 3.6196141012621505e-05, "loss": 0.0131, "step": 17692 }, { "epoch": 1.9181483087597573, "grad_norm": 0.35058632493019104, "learning_rate": 3.619251414478457e-05, "loss": 0.0133, "step": 17693 }, { "epoch": 1.9182567215958368, "grad_norm": 0.18818436563014984, "learning_rate": 3.6188887276947633e-05, "loss": 0.0145, "step": 17694 }, { "epoch": 1.9183651344319168, "grad_norm": 0.49059543013572693, "learning_rate": 3.618526040911069e-05, "loss": 0.0457, "step": 17695 }, { "epoch": 1.9184735472679966, "grad_norm": 0.38712501525878906, "learning_rate": 3.6181633541273755e-05, "loss": 0.0736, "step": 17696 }, { "epoch": 1.9185819601040763, "grad_norm": 0.14264056086540222, "learning_rate": 3.617800667343682e-05, "loss": 0.0053, "step": 17697 }, { "epoch": 1.918690372940156, "grad_norm": 0.14300471544265747, "learning_rate": 3.617437980559988e-05, "loss": 0.0132, "step": 17698 }, { "epoch": 1.9187987857762359, "grad_norm": 0.3776351511478424, "learning_rate": 3.617075293776295e-05, "loss": 0.0415, "step": 17699 }, { "epoch": 1.9189071986123158, "grad_norm": 0.31480810046195984, "learning_rate": 3.616712606992602e-05, "loss": 0.0309, "step": 17700 }, { "epoch": 1.9190156114483954, "grad_norm": 0.2274540215730667, "learning_rate": 3.616349920208908e-05, "loss": 0.0134, "step": 17701 }, { "epoch": 1.9191240242844754, "grad_norm": 0.4219222366809845, "learning_rate": 3.615987233425214e-05, "loss": 0.0463, "step": 17702 }, { "epoch": 1.919232437120555, "grad_norm": 0.1115489974617958, "learning_rate": 3.6156245466415204e-05, "loss": 0.0079, "step": 17703 }, { "epoch": 1.919340849956635, "grad_norm": 0.2682945132255554, "learning_rate": 3.615261859857827e-05, "loss": 0.0322, "step": 17704 }, { "epoch": 1.9194492627927147, "grad_norm": 0.3241261839866638, "learning_rate": 3.614899173074133e-05, "loss": 0.0177, "step": 17705 }, { "epoch": 1.9195576756287944, "grad_norm": 0.41023287177085876, "learning_rate": 3.6145364862904396e-05, "loss": 0.1173, "step": 17706 }, { "epoch": 1.9196660884648742, "grad_norm": 0.7238110303878784, "learning_rate": 3.614173799506746e-05, "loss": 0.0395, "step": 17707 }, { "epoch": 1.919774501300954, "grad_norm": 0.2826400101184845, "learning_rate": 3.6138111127230524e-05, "loss": 0.041, "step": 17708 }, { "epoch": 1.919882914137034, "grad_norm": 0.5669704079627991, "learning_rate": 3.6134484259393595e-05, "loss": 0.0434, "step": 17709 }, { "epoch": 1.9199913269731135, "grad_norm": 0.22986319661140442, "learning_rate": 3.613085739155665e-05, "loss": 0.011, "step": 17710 }, { "epoch": 1.9200997398091935, "grad_norm": 0.16270653903484344, "learning_rate": 3.6127230523719716e-05, "loss": 0.0061, "step": 17711 }, { "epoch": 1.9202081526452732, "grad_norm": 0.23635777831077576, "learning_rate": 3.612360365588278e-05, "loss": 0.0244, "step": 17712 }, { "epoch": 1.920316565481353, "grad_norm": 0.3334432542324066, "learning_rate": 3.6119976788045845e-05, "loss": 0.0414, "step": 17713 }, { "epoch": 1.9204249783174328, "grad_norm": 0.3747796416282654, "learning_rate": 3.611634992020891e-05, "loss": 0.0521, "step": 17714 }, { "epoch": 1.9205333911535125, "grad_norm": 0.11309348046779633, "learning_rate": 3.611272305237197e-05, "loss": 0.0087, "step": 17715 }, { "epoch": 1.9206418039895925, "grad_norm": 0.5978844165802002, "learning_rate": 3.610909618453504e-05, "loss": 0.0496, "step": 17716 }, { "epoch": 1.920750216825672, "grad_norm": 0.21286740899085999, "learning_rate": 3.61054693166981e-05, "loss": 0.029, "step": 17717 }, { "epoch": 1.920858629661752, "grad_norm": 0.32270827889442444, "learning_rate": 3.6101842448861165e-05, "loss": 0.0347, "step": 17718 }, { "epoch": 1.9209670424978318, "grad_norm": 0.1325012445449829, "learning_rate": 3.609821558102423e-05, "loss": 0.0133, "step": 17719 }, { "epoch": 1.9210754553339116, "grad_norm": 0.07683905959129333, "learning_rate": 3.6094588713187293e-05, "loss": 0.0049, "step": 17720 }, { "epoch": 1.9211838681699913, "grad_norm": 0.4777466952800751, "learning_rate": 3.609096184535036e-05, "loss": 0.0411, "step": 17721 }, { "epoch": 1.921292281006071, "grad_norm": 0.3888419568538666, "learning_rate": 3.608733497751342e-05, "loss": 0.0464, "step": 17722 }, { "epoch": 1.921400693842151, "grad_norm": 0.18992333114147186, "learning_rate": 3.6083708109676486e-05, "loss": 0.0148, "step": 17723 }, { "epoch": 1.9215091066782306, "grad_norm": 1.122143030166626, "learning_rate": 3.608008124183955e-05, "loss": 0.0311, "step": 17724 }, { "epoch": 1.9216175195143106, "grad_norm": 0.19327928125858307, "learning_rate": 3.6076454374002614e-05, "loss": 0.0063, "step": 17725 }, { "epoch": 1.9217259323503901, "grad_norm": 0.2237372249364853, "learning_rate": 3.607282750616568e-05, "loss": 0.0199, "step": 17726 }, { "epoch": 1.9218343451864701, "grad_norm": 0.21626757085323334, "learning_rate": 3.606920063832874e-05, "loss": 0.0197, "step": 17727 }, { "epoch": 1.9219427580225499, "grad_norm": 0.13619175553321838, "learning_rate": 3.6065573770491806e-05, "loss": 0.0195, "step": 17728 }, { "epoch": 1.9220511708586296, "grad_norm": 0.13528913259506226, "learning_rate": 3.606194690265487e-05, "loss": 0.0262, "step": 17729 }, { "epoch": 1.9221595836947094, "grad_norm": 0.46904757618904114, "learning_rate": 3.6058320034817934e-05, "loss": 0.0339, "step": 17730 }, { "epoch": 1.9222679965307892, "grad_norm": 0.19531115889549255, "learning_rate": 3.6054693166981e-05, "loss": 0.0198, "step": 17731 }, { "epoch": 1.9223764093668692, "grad_norm": 0.19188310205936432, "learning_rate": 3.605106629914406e-05, "loss": 0.0118, "step": 17732 }, { "epoch": 1.9224848222029487, "grad_norm": 0.38385117053985596, "learning_rate": 3.604743943130712e-05, "loss": 0.0597, "step": 17733 }, { "epoch": 1.9225932350390287, "grad_norm": 0.5838009715080261, "learning_rate": 3.6043812563470184e-05, "loss": 0.0447, "step": 17734 }, { "epoch": 1.9227016478751084, "grad_norm": 0.3360239863395691, "learning_rate": 3.6040185695633255e-05, "loss": 0.0201, "step": 17735 }, { "epoch": 1.9228100607111882, "grad_norm": 0.49756917357444763, "learning_rate": 3.603655882779632e-05, "loss": 0.0374, "step": 17736 }, { "epoch": 1.922918473547268, "grad_norm": 0.29483482241630554, "learning_rate": 3.603293195995938e-05, "loss": 0.0256, "step": 17737 }, { "epoch": 1.9230268863833477, "grad_norm": 0.25358590483665466, "learning_rate": 3.602930509212245e-05, "loss": 0.0349, "step": 17738 }, { "epoch": 1.9231352992194277, "grad_norm": 0.19145236909389496, "learning_rate": 3.602567822428551e-05, "loss": 0.0169, "step": 17739 }, { "epoch": 1.9232437120555073, "grad_norm": 0.3147064745426178, "learning_rate": 3.602205135644857e-05, "loss": 0.0442, "step": 17740 }, { "epoch": 1.9233521248915872, "grad_norm": 0.20994439721107483, "learning_rate": 3.601842448861163e-05, "loss": 0.0125, "step": 17741 }, { "epoch": 1.923460537727667, "grad_norm": 0.10555710643529892, "learning_rate": 3.60147976207747e-05, "loss": 0.0078, "step": 17742 }, { "epoch": 1.9235689505637468, "grad_norm": 0.30371880531311035, "learning_rate": 3.601117075293777e-05, "loss": 0.0463, "step": 17743 }, { "epoch": 1.9236773633998265, "grad_norm": 0.5713117122650146, "learning_rate": 3.600754388510083e-05, "loss": 0.096, "step": 17744 }, { "epoch": 1.9237857762359063, "grad_norm": 0.16949060559272766, "learning_rate": 3.6003917017263896e-05, "loss": 0.0212, "step": 17745 }, { "epoch": 1.9238941890719863, "grad_norm": 0.3272553086280823, "learning_rate": 3.600029014942696e-05, "loss": 0.0323, "step": 17746 }, { "epoch": 1.9240026019080658, "grad_norm": 0.20764365792274475, "learning_rate": 3.599666328159002e-05, "loss": 0.0277, "step": 17747 }, { "epoch": 1.9241110147441458, "grad_norm": 0.3755286633968353, "learning_rate": 3.599303641375308e-05, "loss": 0.0348, "step": 17748 }, { "epoch": 1.9242194275802254, "grad_norm": 0.13032345473766327, "learning_rate": 3.5989409545916146e-05, "loss": 0.0082, "step": 17749 }, { "epoch": 1.9243278404163053, "grad_norm": 0.17076130211353302, "learning_rate": 3.598578267807921e-05, "loss": 0.0268, "step": 17750 }, { "epoch": 1.924436253252385, "grad_norm": 0.05259056016802788, "learning_rate": 3.5982155810242274e-05, "loss": 0.0071, "step": 17751 }, { "epoch": 1.9245446660884649, "grad_norm": 0.3131055235862732, "learning_rate": 3.5978528942405345e-05, "loss": 0.0184, "step": 17752 }, { "epoch": 1.9246530789245446, "grad_norm": 0.25169119238853455, "learning_rate": 3.597490207456841e-05, "loss": 0.0221, "step": 17753 }, { "epoch": 1.9247614917606244, "grad_norm": 0.1440248340368271, "learning_rate": 3.5971275206731466e-05, "loss": 0.0138, "step": 17754 }, { "epoch": 1.9248699045967044, "grad_norm": 0.2590298354625702, "learning_rate": 3.596764833889453e-05, "loss": 0.0305, "step": 17755 }, { "epoch": 1.924978317432784, "grad_norm": 0.19170206785202026, "learning_rate": 3.5964021471057594e-05, "loss": 0.0189, "step": 17756 }, { "epoch": 1.925086730268864, "grad_norm": 0.15056782960891724, "learning_rate": 3.596039460322066e-05, "loss": 0.0244, "step": 17757 }, { "epoch": 1.9251951431049437, "grad_norm": 0.3128053545951843, "learning_rate": 3.595676773538372e-05, "loss": 0.0447, "step": 17758 }, { "epoch": 1.9253035559410234, "grad_norm": 0.3251541256904602, "learning_rate": 3.595314086754679e-05, "loss": 0.0353, "step": 17759 }, { "epoch": 1.9254119687771032, "grad_norm": 0.304521381855011, "learning_rate": 3.594951399970985e-05, "loss": 0.017, "step": 17760 }, { "epoch": 1.925520381613183, "grad_norm": 0.3619040250778198, "learning_rate": 3.594588713187292e-05, "loss": 0.0195, "step": 17761 }, { "epoch": 1.925628794449263, "grad_norm": 0.07176385819911957, "learning_rate": 3.594226026403598e-05, "loss": 0.0055, "step": 17762 }, { "epoch": 1.9257372072853425, "grad_norm": 0.15825247764587402, "learning_rate": 3.593863339619904e-05, "loss": 0.0196, "step": 17763 }, { "epoch": 1.9258456201214225, "grad_norm": 0.2537519931793213, "learning_rate": 3.593500652836211e-05, "loss": 0.0214, "step": 17764 }, { "epoch": 1.925954032957502, "grad_norm": 0.48677098751068115, "learning_rate": 3.593137966052517e-05, "loss": 0.028, "step": 17765 }, { "epoch": 1.926062445793582, "grad_norm": 0.1718205064535141, "learning_rate": 3.5927752792688235e-05, "loss": 0.0263, "step": 17766 }, { "epoch": 1.9261708586296618, "grad_norm": 0.21668346226215363, "learning_rate": 3.59241259248513e-05, "loss": 0.028, "step": 17767 }, { "epoch": 1.9262792714657415, "grad_norm": 0.1026102751493454, "learning_rate": 3.5920499057014364e-05, "loss": 0.0075, "step": 17768 }, { "epoch": 1.9263876843018215, "grad_norm": 0.22693705558776855, "learning_rate": 3.591687218917743e-05, "loss": 0.0285, "step": 17769 }, { "epoch": 1.926496097137901, "grad_norm": 0.2647886276245117, "learning_rate": 3.591324532134049e-05, "loss": 0.0254, "step": 17770 }, { "epoch": 1.926604509973981, "grad_norm": 0.7080113291740417, "learning_rate": 3.5909618453503556e-05, "loss": 0.0444, "step": 17771 }, { "epoch": 1.9267129228100606, "grad_norm": 0.29062145948410034, "learning_rate": 3.590599158566662e-05, "loss": 0.0367, "step": 17772 }, { "epoch": 1.9268213356461406, "grad_norm": 0.12150848656892776, "learning_rate": 3.5902364717829684e-05, "loss": 0.0131, "step": 17773 }, { "epoch": 1.9269297484822203, "grad_norm": 0.5577600002288818, "learning_rate": 3.589873784999275e-05, "loss": 0.032, "step": 17774 }, { "epoch": 1.9270381613183, "grad_norm": 0.3718123137950897, "learning_rate": 3.589511098215581e-05, "loss": 0.0169, "step": 17775 }, { "epoch": 1.9271465741543798, "grad_norm": 0.18419751524925232, "learning_rate": 3.5891484114318876e-05, "loss": 0.0126, "step": 17776 }, { "epoch": 1.9272549869904596, "grad_norm": 0.435935914516449, "learning_rate": 3.588785724648194e-05, "loss": 0.0305, "step": 17777 }, { "epoch": 1.9273633998265396, "grad_norm": 0.14368456602096558, "learning_rate": 3.5884230378645005e-05, "loss": 0.0109, "step": 17778 }, { "epoch": 1.9274718126626191, "grad_norm": 0.09478957951068878, "learning_rate": 3.588060351080807e-05, "loss": 0.0062, "step": 17779 }, { "epoch": 1.9275802254986991, "grad_norm": 0.14702002704143524, "learning_rate": 3.587697664297113e-05, "loss": 0.0071, "step": 17780 }, { "epoch": 1.9276886383347789, "grad_norm": 0.1466512382030487, "learning_rate": 3.58733497751342e-05, "loss": 0.0033, "step": 17781 }, { "epoch": 1.9277970511708586, "grad_norm": 0.3964667022228241, "learning_rate": 3.586972290729726e-05, "loss": 0.0449, "step": 17782 }, { "epoch": 1.9279054640069384, "grad_norm": 0.4519418776035309, "learning_rate": 3.5866096039460325e-05, "loss": 0.0491, "step": 17783 }, { "epoch": 1.9280138768430182, "grad_norm": 0.20138241350650787, "learning_rate": 3.586246917162339e-05, "loss": 0.0177, "step": 17784 }, { "epoch": 1.9281222896790982, "grad_norm": 0.19466562569141388, "learning_rate": 3.585884230378645e-05, "loss": 0.0181, "step": 17785 }, { "epoch": 1.9282307025151777, "grad_norm": 0.0747990608215332, "learning_rate": 3.585521543594952e-05, "loss": 0.0034, "step": 17786 }, { "epoch": 1.9283391153512577, "grad_norm": 0.2332250326871872, "learning_rate": 3.585158856811258e-05, "loss": 0.0396, "step": 17787 }, { "epoch": 1.9284475281873372, "grad_norm": 0.3400624692440033, "learning_rate": 3.5847961700275646e-05, "loss": 0.016, "step": 17788 }, { "epoch": 1.9285559410234172, "grad_norm": 0.58684903383255, "learning_rate": 3.584433483243871e-05, "loss": 0.0393, "step": 17789 }, { "epoch": 1.928664353859497, "grad_norm": 0.34257835149765015, "learning_rate": 3.5840707964601774e-05, "loss": 0.0147, "step": 17790 }, { "epoch": 1.9287727666955767, "grad_norm": 0.11376725882291794, "learning_rate": 3.583708109676484e-05, "loss": 0.0153, "step": 17791 }, { "epoch": 1.9288811795316565, "grad_norm": 0.12143739312887192, "learning_rate": 3.5833454228927895e-05, "loss": 0.0077, "step": 17792 }, { "epoch": 1.9289895923677363, "grad_norm": 0.27388083934783936, "learning_rate": 3.582982736109096e-05, "loss": 0.0292, "step": 17793 }, { "epoch": 1.9290980052038162, "grad_norm": 0.157830610871315, "learning_rate": 3.5826200493254024e-05, "loss": 0.0097, "step": 17794 }, { "epoch": 1.9292064180398958, "grad_norm": 0.16226781904697418, "learning_rate": 3.5822573625417094e-05, "loss": 0.0147, "step": 17795 }, { "epoch": 1.9293148308759758, "grad_norm": 0.10565702617168427, "learning_rate": 3.581894675758016e-05, "loss": 0.0119, "step": 17796 }, { "epoch": 1.9294232437120555, "grad_norm": 0.24343131482601166, "learning_rate": 3.581531988974322e-05, "loss": 0.0205, "step": 17797 }, { "epoch": 1.9295316565481353, "grad_norm": 0.2175632268190384, "learning_rate": 3.581169302190629e-05, "loss": 0.0153, "step": 17798 }, { "epoch": 1.929640069384215, "grad_norm": 0.21960113942623138, "learning_rate": 3.5808066154069344e-05, "loss": 0.0095, "step": 17799 }, { "epoch": 1.9297484822202948, "grad_norm": 0.28081610798835754, "learning_rate": 3.580443928623241e-05, "loss": 0.0499, "step": 17800 }, { "epoch": 1.9298568950563748, "grad_norm": 0.1671210676431656, "learning_rate": 3.580081241839547e-05, "loss": 0.0087, "step": 17801 }, { "epoch": 1.9299653078924544, "grad_norm": 0.6245725154876709, "learning_rate": 3.5797185550558536e-05, "loss": 0.0119, "step": 17802 }, { "epoch": 1.9300737207285343, "grad_norm": 0.7047401070594788, "learning_rate": 3.57935586827216e-05, "loss": 0.0404, "step": 17803 }, { "epoch": 1.930182133564614, "grad_norm": 0.5621724724769592, "learning_rate": 3.578993181488467e-05, "loss": 0.0261, "step": 17804 }, { "epoch": 1.9302905464006939, "grad_norm": 0.6562891602516174, "learning_rate": 3.5786304947047736e-05, "loss": 0.0218, "step": 17805 }, { "epoch": 1.9303989592367736, "grad_norm": 0.3544449806213379, "learning_rate": 3.57826780792108e-05, "loss": 0.0192, "step": 17806 }, { "epoch": 1.9305073720728534, "grad_norm": 0.3847654461860657, "learning_rate": 3.577905121137386e-05, "loss": 0.0069, "step": 17807 }, { "epoch": 1.9306157849089334, "grad_norm": 0.46643683314323425, "learning_rate": 3.577542434353692e-05, "loss": 0.0361, "step": 17808 }, { "epoch": 1.930724197745013, "grad_norm": 0.06730366498231888, "learning_rate": 3.5771797475699985e-05, "loss": 0.0024, "step": 17809 }, { "epoch": 1.930832610581093, "grad_norm": 0.20348156988620758, "learning_rate": 3.576817060786305e-05, "loss": 0.0094, "step": 17810 }, { "epoch": 1.9309410234171724, "grad_norm": 0.6815780401229858, "learning_rate": 3.5764543740026113e-05, "loss": 0.0334, "step": 17811 }, { "epoch": 1.9310494362532524, "grad_norm": 0.16929610073566437, "learning_rate": 3.5760916872189184e-05, "loss": 0.0117, "step": 17812 }, { "epoch": 1.9311578490893322, "grad_norm": 0.5589072704315186, "learning_rate": 3.575729000435225e-05, "loss": 0.0206, "step": 17813 }, { "epoch": 1.931266261925412, "grad_norm": 0.33643341064453125, "learning_rate": 3.5753663136515306e-05, "loss": 0.025, "step": 17814 }, { "epoch": 1.9313746747614917, "grad_norm": 0.23999452590942383, "learning_rate": 3.575003626867837e-05, "loss": 0.0196, "step": 17815 }, { "epoch": 1.9314830875975715, "grad_norm": 0.5711586475372314, "learning_rate": 3.5746409400841434e-05, "loss": 0.0344, "step": 17816 }, { "epoch": 1.9315915004336515, "grad_norm": 0.40418559312820435, "learning_rate": 3.57427825330045e-05, "loss": 0.0529, "step": 17817 }, { "epoch": 1.931699913269731, "grad_norm": 0.3551748991012573, "learning_rate": 3.573915566516756e-05, "loss": 0.0645, "step": 17818 }, { "epoch": 1.931808326105811, "grad_norm": 0.6883854866027832, "learning_rate": 3.5735528797330626e-05, "loss": 0.0332, "step": 17819 }, { "epoch": 1.9319167389418908, "grad_norm": 0.2812161147594452, "learning_rate": 3.573190192949369e-05, "loss": 0.0084, "step": 17820 }, { "epoch": 1.9320251517779705, "grad_norm": 0.22618642449378967, "learning_rate": 3.5728275061656754e-05, "loss": 0.0151, "step": 17821 }, { "epoch": 1.9321335646140503, "grad_norm": 0.22511005401611328, "learning_rate": 3.572464819381982e-05, "loss": 0.018, "step": 17822 }, { "epoch": 1.93224197745013, "grad_norm": 0.4657496213912964, "learning_rate": 3.572102132598288e-05, "loss": 0.056, "step": 17823 }, { "epoch": 1.93235039028621, "grad_norm": 0.22094953060150146, "learning_rate": 3.571739445814595e-05, "loss": 0.0357, "step": 17824 }, { "epoch": 1.9324588031222896, "grad_norm": 0.27672073245048523, "learning_rate": 3.571376759030901e-05, "loss": 0.0332, "step": 17825 }, { "epoch": 1.9325672159583696, "grad_norm": 0.24356117844581604, "learning_rate": 3.5710140722472075e-05, "loss": 0.0327, "step": 17826 }, { "epoch": 1.932675628794449, "grad_norm": 0.3526766002178192, "learning_rate": 3.570651385463514e-05, "loss": 0.0212, "step": 17827 }, { "epoch": 1.932784041630529, "grad_norm": 0.31155648827552795, "learning_rate": 3.57028869867982e-05, "loss": 0.0378, "step": 17828 }, { "epoch": 1.9328924544666088, "grad_norm": 0.2422623485326767, "learning_rate": 3.569926011896127e-05, "loss": 0.0262, "step": 17829 }, { "epoch": 1.9330008673026886, "grad_norm": 0.2599552869796753, "learning_rate": 3.569563325112433e-05, "loss": 0.0214, "step": 17830 }, { "epoch": 1.9331092801387686, "grad_norm": 0.2519659399986267, "learning_rate": 3.5692006383287395e-05, "loss": 0.0053, "step": 17831 }, { "epoch": 1.9332176929748481, "grad_norm": 0.5422288775444031, "learning_rate": 3.568837951545046e-05, "loss": 0.0233, "step": 17832 }, { "epoch": 1.9333261058109281, "grad_norm": 0.37226101756095886, "learning_rate": 3.5684752647613524e-05, "loss": 0.0321, "step": 17833 }, { "epoch": 1.9334345186470077, "grad_norm": 0.4798102378845215, "learning_rate": 3.568112577977659e-05, "loss": 0.0157, "step": 17834 }, { "epoch": 1.9335429314830876, "grad_norm": 0.23797619342803955, "learning_rate": 3.567749891193965e-05, "loss": 0.0298, "step": 17835 }, { "epoch": 1.9336513443191674, "grad_norm": 0.3033655285835266, "learning_rate": 3.5673872044102716e-05, "loss": 0.0566, "step": 17836 }, { "epoch": 1.9337597571552472, "grad_norm": 0.48119181394577026, "learning_rate": 3.567024517626577e-05, "loss": 0.0169, "step": 17837 }, { "epoch": 1.933868169991327, "grad_norm": 0.3139524459838867, "learning_rate": 3.5666618308428844e-05, "loss": 0.0357, "step": 17838 }, { "epoch": 1.9339765828274067, "grad_norm": 0.18718725442886353, "learning_rate": 3.566299144059191e-05, "loss": 0.02, "step": 17839 }, { "epoch": 1.9340849956634867, "grad_norm": 0.1757311224937439, "learning_rate": 3.565936457275497e-05, "loss": 0.0224, "step": 17840 }, { "epoch": 1.9341934084995662, "grad_norm": 0.19293126463890076, "learning_rate": 3.5655737704918037e-05, "loss": 0.028, "step": 17841 }, { "epoch": 1.9343018213356462, "grad_norm": 0.09368447959423065, "learning_rate": 3.56521108370811e-05, "loss": 0.0033, "step": 17842 }, { "epoch": 1.934410234171726, "grad_norm": 0.1920975297689438, "learning_rate": 3.5648483969244165e-05, "loss": 0.0129, "step": 17843 }, { "epoch": 1.9345186470078057, "grad_norm": 0.22367563843727112, "learning_rate": 3.564485710140722e-05, "loss": 0.0268, "step": 17844 }, { "epoch": 1.9346270598438855, "grad_norm": 0.5955338478088379, "learning_rate": 3.5641230233570286e-05, "loss": 0.0255, "step": 17845 }, { "epoch": 1.9347354726799653, "grad_norm": 0.18131805956363678, "learning_rate": 3.563760336573335e-05, "loss": 0.0082, "step": 17846 }, { "epoch": 1.9348438855160452, "grad_norm": 0.09971942007541656, "learning_rate": 3.563397649789642e-05, "loss": 0.0098, "step": 17847 }, { "epoch": 1.9349522983521248, "grad_norm": 0.1362304985523224, "learning_rate": 3.5630349630059485e-05, "loss": 0.0188, "step": 17848 }, { "epoch": 1.9350607111882048, "grad_norm": 0.8382359147071838, "learning_rate": 3.562672276222255e-05, "loss": 0.0243, "step": 17849 }, { "epoch": 1.9351691240242843, "grad_norm": 0.5500849485397339, "learning_rate": 3.5623095894385613e-05, "loss": 0.0371, "step": 17850 }, { "epoch": 1.9352775368603643, "grad_norm": 0.12660333514213562, "learning_rate": 3.561946902654867e-05, "loss": 0.0073, "step": 17851 }, { "epoch": 1.935385949696444, "grad_norm": 0.2530958950519562, "learning_rate": 3.5615842158711735e-05, "loss": 0.0284, "step": 17852 }, { "epoch": 1.9354943625325238, "grad_norm": 0.5552671551704407, "learning_rate": 3.56122152908748e-05, "loss": 0.0244, "step": 17853 }, { "epoch": 1.9356027753686036, "grad_norm": 0.6423521637916565, "learning_rate": 3.560858842303786e-05, "loss": 0.1179, "step": 17854 }, { "epoch": 1.9357111882046834, "grad_norm": 0.1672683209180832, "learning_rate": 3.560496155520093e-05, "loss": 0.0091, "step": 17855 }, { "epoch": 1.9358196010407633, "grad_norm": 0.47089463472366333, "learning_rate": 3.5601334687364e-05, "loss": 0.083, "step": 17856 }, { "epoch": 1.9359280138768429, "grad_norm": 0.1835063397884369, "learning_rate": 3.559770781952706e-05, "loss": 0.0156, "step": 17857 }, { "epoch": 1.9360364267129229, "grad_norm": 0.10668645054101944, "learning_rate": 3.5594080951690126e-05, "loss": 0.0094, "step": 17858 }, { "epoch": 1.9361448395490026, "grad_norm": 0.2504855990409851, "learning_rate": 3.5590454083853184e-05, "loss": 0.0272, "step": 17859 }, { "epoch": 1.9362532523850824, "grad_norm": 0.24937598407268524, "learning_rate": 3.558682721601625e-05, "loss": 0.0145, "step": 17860 }, { "epoch": 1.9363616652211622, "grad_norm": 0.08464299887418747, "learning_rate": 3.558320034817931e-05, "loss": 0.0087, "step": 17861 }, { "epoch": 1.936470078057242, "grad_norm": 0.0856596976518631, "learning_rate": 3.5579573480342376e-05, "loss": 0.0136, "step": 17862 }, { "epoch": 1.936578490893322, "grad_norm": 0.23798838257789612, "learning_rate": 3.557594661250544e-05, "loss": 0.0218, "step": 17863 }, { "epoch": 1.9366869037294014, "grad_norm": 0.22736328840255737, "learning_rate": 3.557231974466851e-05, "loss": 0.0103, "step": 17864 }, { "epoch": 1.9367953165654814, "grad_norm": 0.15396445989608765, "learning_rate": 3.5568692876831575e-05, "loss": 0.0068, "step": 17865 }, { "epoch": 1.9369037294015612, "grad_norm": 0.44021904468536377, "learning_rate": 3.556506600899463e-05, "loss": 0.0342, "step": 17866 }, { "epoch": 1.937012142237641, "grad_norm": 0.43844202160835266, "learning_rate": 3.5561439141157697e-05, "loss": 0.0122, "step": 17867 }, { "epoch": 1.9371205550737207, "grad_norm": 0.10742313414812088, "learning_rate": 3.555781227332076e-05, "loss": 0.0076, "step": 17868 }, { "epoch": 1.9372289679098005, "grad_norm": 0.5104467868804932, "learning_rate": 3.5554185405483825e-05, "loss": 0.0196, "step": 17869 }, { "epoch": 1.9373373807458805, "grad_norm": 0.1010444164276123, "learning_rate": 3.555055853764689e-05, "loss": 0.0069, "step": 17870 }, { "epoch": 1.93744579358196, "grad_norm": 0.1529260277748108, "learning_rate": 3.554693166980995e-05, "loss": 0.0178, "step": 17871 }, { "epoch": 1.93755420641804, "grad_norm": 0.37520381808280945, "learning_rate": 3.554330480197302e-05, "loss": 0.0179, "step": 17872 }, { "epoch": 1.9376626192541195, "grad_norm": 0.37004998326301575, "learning_rate": 3.553967793413608e-05, "loss": 0.0203, "step": 17873 }, { "epoch": 1.9377710320901995, "grad_norm": 0.25683343410491943, "learning_rate": 3.5536051066299145e-05, "loss": 0.0111, "step": 17874 }, { "epoch": 1.9378794449262793, "grad_norm": 0.09690585732460022, "learning_rate": 3.553242419846221e-05, "loss": 0.0054, "step": 17875 }, { "epoch": 1.937987857762359, "grad_norm": 0.16168548166751862, "learning_rate": 3.5528797330625273e-05, "loss": 0.0165, "step": 17876 }, { "epoch": 1.9380962705984388, "grad_norm": 0.0950360894203186, "learning_rate": 3.552517046278834e-05, "loss": 0.0063, "step": 17877 }, { "epoch": 1.9382046834345186, "grad_norm": 0.34255337715148926, "learning_rate": 3.55215435949514e-05, "loss": 0.0269, "step": 17878 }, { "epoch": 1.9383130962705986, "grad_norm": 0.5291417241096497, "learning_rate": 3.5517916727114466e-05, "loss": 0.0484, "step": 17879 }, { "epoch": 1.938421509106678, "grad_norm": 0.1467948853969574, "learning_rate": 3.551428985927753e-05, "loss": 0.022, "step": 17880 }, { "epoch": 1.938529921942758, "grad_norm": 0.04371589794754982, "learning_rate": 3.5510662991440594e-05, "loss": 0.0059, "step": 17881 }, { "epoch": 1.9386383347788378, "grad_norm": 0.4402106702327728, "learning_rate": 3.550703612360366e-05, "loss": 0.0361, "step": 17882 }, { "epoch": 1.9387467476149176, "grad_norm": 0.2265613079071045, "learning_rate": 3.550340925576672e-05, "loss": 0.0158, "step": 17883 }, { "epoch": 1.9388551604509974, "grad_norm": 0.01895718462765217, "learning_rate": 3.5499782387929786e-05, "loss": 0.0015, "step": 17884 }, { "epoch": 1.9389635732870771, "grad_norm": 0.21191266179084778, "learning_rate": 3.549615552009285e-05, "loss": 0.0287, "step": 17885 }, { "epoch": 1.9390719861231571, "grad_norm": 0.18859604001045227, "learning_rate": 3.5492528652255915e-05, "loss": 0.0118, "step": 17886 }, { "epoch": 1.9391803989592367, "grad_norm": 0.21574920415878296, "learning_rate": 3.548890178441898e-05, "loss": 0.0298, "step": 17887 }, { "epoch": 1.9392888117953166, "grad_norm": 0.28872668743133545, "learning_rate": 3.548527491658204e-05, "loss": 0.0249, "step": 17888 }, { "epoch": 1.9393972246313964, "grad_norm": 0.46392565965652466, "learning_rate": 3.54816480487451e-05, "loss": 0.0336, "step": 17889 }, { "epoch": 1.9395056374674762, "grad_norm": 0.29912659525871277, "learning_rate": 3.547802118090817e-05, "loss": 0.0411, "step": 17890 }, { "epoch": 1.939614050303556, "grad_norm": 0.38559627532958984, "learning_rate": 3.5474394313071235e-05, "loss": 0.0469, "step": 17891 }, { "epoch": 1.9397224631396357, "grad_norm": 0.016675259917974472, "learning_rate": 3.54707674452343e-05, "loss": 0.0011, "step": 17892 }, { "epoch": 1.9398308759757157, "grad_norm": 0.3398454785346985, "learning_rate": 3.546714057739736e-05, "loss": 0.0326, "step": 17893 }, { "epoch": 1.9399392888117952, "grad_norm": 0.14604021608829498, "learning_rate": 3.546351370956043e-05, "loss": 0.0128, "step": 17894 }, { "epoch": 1.9400477016478752, "grad_norm": 0.2446536421775818, "learning_rate": 3.545988684172349e-05, "loss": 0.0149, "step": 17895 }, { "epoch": 1.9401561144839548, "grad_norm": 0.08033875375986099, "learning_rate": 3.545625997388655e-05, "loss": 0.0066, "step": 17896 }, { "epoch": 1.9402645273200347, "grad_norm": 0.14643383026123047, "learning_rate": 3.545263310604961e-05, "loss": 0.014, "step": 17897 }, { "epoch": 1.9403729401561145, "grad_norm": 0.1722831279039383, "learning_rate": 3.544900623821268e-05, "loss": 0.0111, "step": 17898 }, { "epoch": 1.9404813529921943, "grad_norm": 0.16626891493797302, "learning_rate": 3.544537937037575e-05, "loss": 0.0133, "step": 17899 }, { "epoch": 1.940589765828274, "grad_norm": 0.20604674518108368, "learning_rate": 3.544175250253881e-05, "loss": 0.0265, "step": 17900 }, { "epoch": 1.9406981786643538, "grad_norm": 0.2332707792520523, "learning_rate": 3.5438125634701876e-05, "loss": 0.0218, "step": 17901 }, { "epoch": 1.9408065915004338, "grad_norm": 0.326698362827301, "learning_rate": 3.543449876686494e-05, "loss": 0.0539, "step": 17902 }, { "epoch": 1.9409150043365133, "grad_norm": 0.39439651370048523, "learning_rate": 3.5430871899028e-05, "loss": 0.0325, "step": 17903 }, { "epoch": 1.9410234171725933, "grad_norm": 0.33524978160858154, "learning_rate": 3.542724503119106e-05, "loss": 0.0177, "step": 17904 }, { "epoch": 1.941131830008673, "grad_norm": 0.28533056378364563, "learning_rate": 3.5423618163354126e-05, "loss": 0.0089, "step": 17905 }, { "epoch": 1.9412402428447528, "grad_norm": 0.7308453321456909, "learning_rate": 3.541999129551719e-05, "loss": 0.0202, "step": 17906 }, { "epoch": 1.9413486556808326, "grad_norm": 0.17856931686401367, "learning_rate": 3.541636442768026e-05, "loss": 0.0137, "step": 17907 }, { "epoch": 1.9414570685169124, "grad_norm": 0.20579250156879425, "learning_rate": 3.5412737559843325e-05, "loss": 0.0069, "step": 17908 }, { "epoch": 1.9415654813529923, "grad_norm": 0.2930542826652527, "learning_rate": 3.540911069200639e-05, "loss": 0.0303, "step": 17909 }, { "epoch": 1.9416738941890719, "grad_norm": 0.343147337436676, "learning_rate": 3.540548382416945e-05, "loss": 0.0187, "step": 17910 }, { "epoch": 1.9417823070251519, "grad_norm": 0.3837733864784241, "learning_rate": 3.540185695633251e-05, "loss": 0.0124, "step": 17911 }, { "epoch": 1.9418907198612314, "grad_norm": 0.20834940671920776, "learning_rate": 3.5398230088495574e-05, "loss": 0.0198, "step": 17912 }, { "epoch": 1.9419991326973114, "grad_norm": 0.24622327089309692, "learning_rate": 3.539460322065864e-05, "loss": 0.024, "step": 17913 }, { "epoch": 1.9421075455333912, "grad_norm": 0.5704571604728699, "learning_rate": 3.53909763528217e-05, "loss": 0.0073, "step": 17914 }, { "epoch": 1.942215958369471, "grad_norm": 0.32091185450553894, "learning_rate": 3.538734948498477e-05, "loss": 0.0328, "step": 17915 }, { "epoch": 1.942324371205551, "grad_norm": 0.14845967292785645, "learning_rate": 3.538372261714784e-05, "loss": 0.0062, "step": 17916 }, { "epoch": 1.9424327840416304, "grad_norm": 0.5205281972885132, "learning_rate": 3.53800957493109e-05, "loss": 0.0414, "step": 17917 }, { "epoch": 1.9425411968777104, "grad_norm": 0.2170548290014267, "learning_rate": 3.537646888147396e-05, "loss": 0.0151, "step": 17918 }, { "epoch": 1.94264960971379, "grad_norm": 0.2981375455856323, "learning_rate": 3.537284201363702e-05, "loss": 0.0157, "step": 17919 }, { "epoch": 1.94275802254987, "grad_norm": 0.34069639444351196, "learning_rate": 3.536921514580009e-05, "loss": 0.0571, "step": 17920 }, { "epoch": 1.9428664353859497, "grad_norm": 0.0963456928730011, "learning_rate": 3.536558827796315e-05, "loss": 0.006, "step": 17921 }, { "epoch": 1.9429748482220295, "grad_norm": 0.5856128334999084, "learning_rate": 3.5361961410126216e-05, "loss": 0.0396, "step": 17922 }, { "epoch": 1.9430832610581092, "grad_norm": 0.05881983041763306, "learning_rate": 3.535833454228928e-05, "loss": 0.0034, "step": 17923 }, { "epoch": 1.943191673894189, "grad_norm": 0.0729755163192749, "learning_rate": 3.5354707674452344e-05, "loss": 0.0039, "step": 17924 }, { "epoch": 1.943300086730269, "grad_norm": 0.2660179138183594, "learning_rate": 3.535108080661541e-05, "loss": 0.0144, "step": 17925 }, { "epoch": 1.9434084995663485, "grad_norm": 0.42345961928367615, "learning_rate": 3.534745393877847e-05, "loss": 0.0156, "step": 17926 }, { "epoch": 1.9435169124024285, "grad_norm": 0.3132723569869995, "learning_rate": 3.5343827070941536e-05, "loss": 0.0175, "step": 17927 }, { "epoch": 1.9436253252385083, "grad_norm": 0.29757797718048096, "learning_rate": 3.53402002031046e-05, "loss": 0.022, "step": 17928 }, { "epoch": 1.943733738074588, "grad_norm": 0.4059496223926544, "learning_rate": 3.5336573335267664e-05, "loss": 0.0295, "step": 17929 }, { "epoch": 1.9438421509106678, "grad_norm": 0.05323558673262596, "learning_rate": 3.533294646743073e-05, "loss": 0.0035, "step": 17930 }, { "epoch": 1.9439505637467476, "grad_norm": 0.3901371657848358, "learning_rate": 3.532931959959379e-05, "loss": 0.0195, "step": 17931 }, { "epoch": 1.9440589765828276, "grad_norm": 0.22401167452335358, "learning_rate": 3.5325692731756857e-05, "loss": 0.0172, "step": 17932 }, { "epoch": 1.944167389418907, "grad_norm": 0.17357318103313446, "learning_rate": 3.532206586391992e-05, "loss": 0.0211, "step": 17933 }, { "epoch": 1.944275802254987, "grad_norm": 0.23127587139606476, "learning_rate": 3.5318438996082985e-05, "loss": 0.0152, "step": 17934 }, { "epoch": 1.9443842150910666, "grad_norm": 0.12412513047456741, "learning_rate": 3.531481212824605e-05, "loss": 0.0117, "step": 17935 }, { "epoch": 1.9444926279271466, "grad_norm": 0.29514241218566895, "learning_rate": 3.531118526040911e-05, "loss": 0.0295, "step": 17936 }, { "epoch": 1.9446010407632264, "grad_norm": 0.3788785934448242, "learning_rate": 3.530755839257218e-05, "loss": 0.0177, "step": 17937 }, { "epoch": 1.9447094535993061, "grad_norm": 0.4465462863445282, "learning_rate": 3.530393152473524e-05, "loss": 0.0271, "step": 17938 }, { "epoch": 1.944817866435386, "grad_norm": 0.7892173528671265, "learning_rate": 3.5300304656898305e-05, "loss": 0.0165, "step": 17939 }, { "epoch": 1.9449262792714657, "grad_norm": 0.6543638110160828, "learning_rate": 3.529667778906137e-05, "loss": 0.0978, "step": 17940 }, { "epoch": 1.9450346921075456, "grad_norm": 0.9498644471168518, "learning_rate": 3.529305092122443e-05, "loss": 0.0402, "step": 17941 }, { "epoch": 1.9451431049436252, "grad_norm": 0.38094422221183777, "learning_rate": 3.52894240533875e-05, "loss": 0.0171, "step": 17942 }, { "epoch": 1.9452515177797052, "grad_norm": 0.25082600116729736, "learning_rate": 3.528579718555056e-05, "loss": 0.0281, "step": 17943 }, { "epoch": 1.945359930615785, "grad_norm": 0.14640206098556519, "learning_rate": 3.5282170317713626e-05, "loss": 0.0084, "step": 17944 }, { "epoch": 1.9454683434518647, "grad_norm": 0.2200016975402832, "learning_rate": 3.527854344987669e-05, "loss": 0.0192, "step": 17945 }, { "epoch": 1.9455767562879445, "grad_norm": 0.1981624811887741, "learning_rate": 3.5274916582039754e-05, "loss": 0.0226, "step": 17946 }, { "epoch": 1.9456851691240242, "grad_norm": 0.6915402412414551, "learning_rate": 3.527128971420282e-05, "loss": 0.0487, "step": 17947 }, { "epoch": 1.9457935819601042, "grad_norm": 0.10140787810087204, "learning_rate": 3.5267662846365875e-05, "loss": 0.0069, "step": 17948 }, { "epoch": 1.9459019947961838, "grad_norm": 0.3327118754386902, "learning_rate": 3.526403597852894e-05, "loss": 0.0323, "step": 17949 }, { "epoch": 1.9460104076322637, "grad_norm": 0.26898908615112305, "learning_rate": 3.5260409110692004e-05, "loss": 0.0381, "step": 17950 }, { "epoch": 1.9461188204683435, "grad_norm": 0.2649421691894531, "learning_rate": 3.5256782242855075e-05, "loss": 0.0121, "step": 17951 }, { "epoch": 1.9462272333044233, "grad_norm": 0.25922176241874695, "learning_rate": 3.525315537501814e-05, "loss": 0.0127, "step": 17952 }, { "epoch": 1.946335646140503, "grad_norm": 0.4685761630535126, "learning_rate": 3.52495285071812e-05, "loss": 0.0176, "step": 17953 }, { "epoch": 1.9464440589765828, "grad_norm": 0.41432657837867737, "learning_rate": 3.524590163934427e-05, "loss": 0.0155, "step": 17954 }, { "epoch": 1.9465524718126628, "grad_norm": 0.18177242577075958, "learning_rate": 3.524227477150733e-05, "loss": 0.0215, "step": 17955 }, { "epoch": 1.9466608846487423, "grad_norm": 0.03976050019264221, "learning_rate": 3.523864790367039e-05, "loss": 0.0022, "step": 17956 }, { "epoch": 1.9467692974848223, "grad_norm": 0.9175809621810913, "learning_rate": 3.523502103583345e-05, "loss": 0.0695, "step": 17957 }, { "epoch": 1.9468777103209018, "grad_norm": 0.15585799515247345, "learning_rate": 3.5231394167996517e-05, "loss": 0.0236, "step": 17958 }, { "epoch": 1.9469861231569818, "grad_norm": 0.7073931097984314, "learning_rate": 3.522776730015959e-05, "loss": 0.0348, "step": 17959 }, { "epoch": 1.9470945359930616, "grad_norm": 0.059143826365470886, "learning_rate": 3.522414043232265e-05, "loss": 0.0043, "step": 17960 }, { "epoch": 1.9472029488291414, "grad_norm": 0.19965362548828125, "learning_rate": 3.5220513564485716e-05, "loss": 0.0174, "step": 17961 }, { "epoch": 1.9473113616652211, "grad_norm": 0.50702303647995, "learning_rate": 3.521688669664878e-05, "loss": 0.0128, "step": 17962 }, { "epoch": 1.9474197745013009, "grad_norm": 0.11064401268959045, "learning_rate": 3.521325982881184e-05, "loss": 0.0091, "step": 17963 }, { "epoch": 1.9475281873373809, "grad_norm": 0.3896629214286804, "learning_rate": 3.52096329609749e-05, "loss": 0.0766, "step": 17964 }, { "epoch": 1.9476366001734604, "grad_norm": 0.2672920227050781, "learning_rate": 3.5206006093137965e-05, "loss": 0.0246, "step": 17965 }, { "epoch": 1.9477450130095404, "grad_norm": 0.3221117854118347, "learning_rate": 3.520237922530103e-05, "loss": 0.0149, "step": 17966 }, { "epoch": 1.9478534258456202, "grad_norm": 0.9633017778396606, "learning_rate": 3.5198752357464093e-05, "loss": 0.0364, "step": 17967 }, { "epoch": 1.9479618386817, "grad_norm": 0.09804534167051315, "learning_rate": 3.5195125489627164e-05, "loss": 0.0057, "step": 17968 }, { "epoch": 1.9480702515177797, "grad_norm": 0.36647123098373413, "learning_rate": 3.519149862179023e-05, "loss": 0.0651, "step": 17969 }, { "epoch": 1.9481786643538594, "grad_norm": 0.4024413824081421, "learning_rate": 3.5187871753953286e-05, "loss": 0.0442, "step": 17970 }, { "epoch": 1.9482870771899394, "grad_norm": 0.04566218703985214, "learning_rate": 3.518424488611635e-05, "loss": 0.0027, "step": 17971 }, { "epoch": 1.948395490026019, "grad_norm": 0.18065574765205383, "learning_rate": 3.5180618018279414e-05, "loss": 0.009, "step": 17972 }, { "epoch": 1.948503902862099, "grad_norm": 0.2658451497554779, "learning_rate": 3.517699115044248e-05, "loss": 0.0208, "step": 17973 }, { "epoch": 1.9486123156981785, "grad_norm": 0.2284041792154312, "learning_rate": 3.517336428260554e-05, "loss": 0.0176, "step": 17974 }, { "epoch": 1.9487207285342585, "grad_norm": 0.33948880434036255, "learning_rate": 3.5169737414768606e-05, "loss": 0.0428, "step": 17975 }, { "epoch": 1.9488291413703382, "grad_norm": 0.2598833441734314, "learning_rate": 3.516611054693167e-05, "loss": 0.017, "step": 17976 }, { "epoch": 1.948937554206418, "grad_norm": 0.29657885432243347, "learning_rate": 3.5162483679094735e-05, "loss": 0.0437, "step": 17977 }, { "epoch": 1.949045967042498, "grad_norm": 0.4354487359523773, "learning_rate": 3.51588568112578e-05, "loss": 0.0162, "step": 17978 }, { "epoch": 1.9491543798785775, "grad_norm": 0.20035044848918915, "learning_rate": 3.515522994342086e-05, "loss": 0.01, "step": 17979 }, { "epoch": 1.9492627927146575, "grad_norm": 0.09173539280891418, "learning_rate": 3.515160307558393e-05, "loss": 0.0076, "step": 17980 }, { "epoch": 1.949371205550737, "grad_norm": 0.46176329255104065, "learning_rate": 3.514797620774699e-05, "loss": 0.0209, "step": 17981 }, { "epoch": 1.949479618386817, "grad_norm": 0.6017420887947083, "learning_rate": 3.5144349339910055e-05, "loss": 0.0376, "step": 17982 }, { "epoch": 1.9495880312228968, "grad_norm": 0.14920207858085632, "learning_rate": 3.514072247207312e-05, "loss": 0.0089, "step": 17983 }, { "epoch": 1.9496964440589766, "grad_norm": 0.4174264669418335, "learning_rate": 3.513709560423618e-05, "loss": 0.0234, "step": 17984 }, { "epoch": 1.9498048568950563, "grad_norm": 0.16475233435630798, "learning_rate": 3.513346873639925e-05, "loss": 0.0052, "step": 17985 }, { "epoch": 1.949913269731136, "grad_norm": 0.3061111569404602, "learning_rate": 3.512984186856231e-05, "loss": 0.0144, "step": 17986 }, { "epoch": 1.950021682567216, "grad_norm": 0.5396195650100708, "learning_rate": 3.5126215000725376e-05, "loss": 0.0216, "step": 17987 }, { "epoch": 1.9501300954032956, "grad_norm": 0.33297160267829895, "learning_rate": 3.512258813288844e-05, "loss": 0.0142, "step": 17988 }, { "epoch": 1.9502385082393756, "grad_norm": 0.1957148015499115, "learning_rate": 3.5118961265051504e-05, "loss": 0.0129, "step": 17989 }, { "epoch": 1.9503469210754554, "grad_norm": 0.14027827978134155, "learning_rate": 3.511533439721457e-05, "loss": 0.0085, "step": 17990 }, { "epoch": 1.9504553339115351, "grad_norm": 0.33406487107276917, "learning_rate": 3.511170752937763e-05, "loss": 0.0154, "step": 17991 }, { "epoch": 1.950563746747615, "grad_norm": 0.12762901186943054, "learning_rate": 3.5108080661540696e-05, "loss": 0.0076, "step": 17992 }, { "epoch": 1.9506721595836947, "grad_norm": 0.2058998942375183, "learning_rate": 3.5104453793703753e-05, "loss": 0.01, "step": 17993 }, { "epoch": 1.9507805724197746, "grad_norm": 0.10902518779039383, "learning_rate": 3.5100826925866824e-05, "loss": 0.0085, "step": 17994 }, { "epoch": 1.9508889852558542, "grad_norm": 0.4221484661102295, "learning_rate": 3.509720005802989e-05, "loss": 0.0344, "step": 17995 }, { "epoch": 1.9509973980919342, "grad_norm": 0.18250064551830292, "learning_rate": 3.509357319019295e-05, "loss": 0.0216, "step": 17996 }, { "epoch": 1.9511058109280137, "grad_norm": 0.21676845848560333, "learning_rate": 3.508994632235602e-05, "loss": 0.0143, "step": 17997 }, { "epoch": 1.9512142237640937, "grad_norm": 0.5802922248840332, "learning_rate": 3.508631945451908e-05, "loss": 0.0246, "step": 17998 }, { "epoch": 1.9513226366001735, "grad_norm": 1.3024166822433472, "learning_rate": 3.5082692586682145e-05, "loss": 0.0306, "step": 17999 }, { "epoch": 1.9514310494362532, "grad_norm": 0.3958039879798889, "learning_rate": 3.50790657188452e-05, "loss": 0.0657, "step": 18000 }, { "epoch": 1.951539462272333, "grad_norm": 0.5807530879974365, "learning_rate": 3.5075438851008266e-05, "loss": 0.0179, "step": 18001 }, { "epoch": 1.9516478751084128, "grad_norm": 0.09244966506958008, "learning_rate": 3.507181198317134e-05, "loss": 0.0087, "step": 18002 }, { "epoch": 1.9517562879444927, "grad_norm": 0.37546107172966003, "learning_rate": 3.50681851153344e-05, "loss": 0.0186, "step": 18003 }, { "epoch": 1.9518647007805723, "grad_norm": 0.5455124378204346, "learning_rate": 3.5064558247497465e-05, "loss": 0.0244, "step": 18004 }, { "epoch": 1.9519731136166523, "grad_norm": 0.23579247295856476, "learning_rate": 3.506093137966053e-05, "loss": 0.018, "step": 18005 }, { "epoch": 1.952081526452732, "grad_norm": 0.1684151589870453, "learning_rate": 3.5057304511823594e-05, "loss": 0.0112, "step": 18006 }, { "epoch": 1.9521899392888118, "grad_norm": 1.0004994869232178, "learning_rate": 3.505367764398666e-05, "loss": 0.0395, "step": 18007 }, { "epoch": 1.9522983521248916, "grad_norm": 0.4167932868003845, "learning_rate": 3.5050050776149715e-05, "loss": 0.024, "step": 18008 }, { "epoch": 1.9524067649609713, "grad_norm": 0.29119592905044556, "learning_rate": 3.504642390831278e-05, "loss": 0.0384, "step": 18009 }, { "epoch": 1.9525151777970513, "grad_norm": 0.16424503922462463, "learning_rate": 3.504279704047584e-05, "loss": 0.0092, "step": 18010 }, { "epoch": 1.9526235906331308, "grad_norm": 0.36056697368621826, "learning_rate": 3.5039170172638914e-05, "loss": 0.0321, "step": 18011 }, { "epoch": 1.9527320034692108, "grad_norm": 0.1512211710214615, "learning_rate": 3.503554330480198e-05, "loss": 0.0058, "step": 18012 }, { "epoch": 1.9528404163052906, "grad_norm": 0.13878792524337769, "learning_rate": 3.503191643696504e-05, "loss": 0.0099, "step": 18013 }, { "epoch": 1.9529488291413704, "grad_norm": 0.7293976545333862, "learning_rate": 3.5028289569128106e-05, "loss": 0.0667, "step": 18014 }, { "epoch": 1.9530572419774501, "grad_norm": 0.23638825118541718, "learning_rate": 3.5024662701291164e-05, "loss": 0.0107, "step": 18015 }, { "epoch": 1.9531656548135299, "grad_norm": 0.020581506192684174, "learning_rate": 3.502103583345423e-05, "loss": 0.0009, "step": 18016 }, { "epoch": 1.9532740676496099, "grad_norm": 0.1589319109916687, "learning_rate": 3.501740896561729e-05, "loss": 0.0112, "step": 18017 }, { "epoch": 1.9533824804856894, "grad_norm": 1.3422154188156128, "learning_rate": 3.5013782097780356e-05, "loss": 0.0352, "step": 18018 }, { "epoch": 1.9534908933217694, "grad_norm": 0.16163195669651031, "learning_rate": 3.501015522994342e-05, "loss": 0.0148, "step": 18019 }, { "epoch": 1.953599306157849, "grad_norm": 0.6149461269378662, "learning_rate": 3.500652836210649e-05, "loss": 0.0125, "step": 18020 }, { "epoch": 1.953707718993929, "grad_norm": 0.1959712654352188, "learning_rate": 3.5002901494269555e-05, "loss": 0.014, "step": 18021 }, { "epoch": 1.9538161318300087, "grad_norm": 0.4907843768596649, "learning_rate": 3.499927462643261e-05, "loss": 0.0421, "step": 18022 }, { "epoch": 1.9539245446660884, "grad_norm": 0.559040367603302, "learning_rate": 3.4995647758595677e-05, "loss": 0.0618, "step": 18023 }, { "epoch": 1.9540329575021682, "grad_norm": 0.2744952440261841, "learning_rate": 3.499202089075874e-05, "loss": 0.0353, "step": 18024 }, { "epoch": 1.954141370338248, "grad_norm": 0.1759941428899765, "learning_rate": 3.4988394022921805e-05, "loss": 0.0084, "step": 18025 }, { "epoch": 1.954249783174328, "grad_norm": 0.43632134795188904, "learning_rate": 3.498476715508487e-05, "loss": 0.0255, "step": 18026 }, { "epoch": 1.9543581960104075, "grad_norm": 0.10811944305896759, "learning_rate": 3.498114028724793e-05, "loss": 0.0078, "step": 18027 }, { "epoch": 1.9544666088464875, "grad_norm": 0.38810625672340393, "learning_rate": 3.4977513419411004e-05, "loss": 0.0422, "step": 18028 }, { "epoch": 1.9545750216825672, "grad_norm": 0.252089262008667, "learning_rate": 3.497388655157406e-05, "loss": 0.0142, "step": 18029 }, { "epoch": 1.954683434518647, "grad_norm": 0.5974622964859009, "learning_rate": 3.4970259683737125e-05, "loss": 0.017, "step": 18030 }, { "epoch": 1.9547918473547268, "grad_norm": 0.20980125665664673, "learning_rate": 3.496663281590019e-05, "loss": 0.0197, "step": 18031 }, { "epoch": 1.9549002601908065, "grad_norm": 0.18198195099830627, "learning_rate": 3.4963005948063254e-05, "loss": 0.0055, "step": 18032 }, { "epoch": 1.9550086730268865, "grad_norm": 0.7629634737968445, "learning_rate": 3.495937908022632e-05, "loss": 0.0464, "step": 18033 }, { "epoch": 1.955117085862966, "grad_norm": 0.44118720293045044, "learning_rate": 3.495575221238938e-05, "loss": 0.0347, "step": 18034 }, { "epoch": 1.955225498699046, "grad_norm": 0.5683943629264832, "learning_rate": 3.4952125344552446e-05, "loss": 0.0334, "step": 18035 }, { "epoch": 1.9553339115351258, "grad_norm": 0.6399344205856323, "learning_rate": 3.494849847671551e-05, "loss": 0.0461, "step": 18036 }, { "epoch": 1.9554423243712056, "grad_norm": 0.3127158284187317, "learning_rate": 3.4944871608878574e-05, "loss": 0.0155, "step": 18037 }, { "epoch": 1.9555507372072853, "grad_norm": 0.05228430777788162, "learning_rate": 3.494124474104164e-05, "loss": 0.0016, "step": 18038 }, { "epoch": 1.955659150043365, "grad_norm": 0.056939367204904556, "learning_rate": 3.49376178732047e-05, "loss": 0.005, "step": 18039 }, { "epoch": 1.955767562879445, "grad_norm": 0.6150699257850647, "learning_rate": 3.4933991005367766e-05, "loss": 0.0319, "step": 18040 }, { "epoch": 1.9558759757155246, "grad_norm": 0.32713115215301514, "learning_rate": 3.493036413753083e-05, "loss": 0.0432, "step": 18041 }, { "epoch": 1.9559843885516046, "grad_norm": 0.31233835220336914, "learning_rate": 3.4926737269693895e-05, "loss": 0.0166, "step": 18042 }, { "epoch": 1.9560928013876842, "grad_norm": 0.28876230120658875, "learning_rate": 3.492311040185696e-05, "loss": 0.0201, "step": 18043 }, { "epoch": 1.9562012142237641, "grad_norm": 0.2819922864437103, "learning_rate": 3.491948353402002e-05, "loss": 0.0128, "step": 18044 }, { "epoch": 1.956309627059844, "grad_norm": 0.2081754207611084, "learning_rate": 3.491585666618308e-05, "loss": 0.019, "step": 18045 }, { "epoch": 1.9564180398959237, "grad_norm": 0.5066971182823181, "learning_rate": 3.491222979834615e-05, "loss": 0.0317, "step": 18046 }, { "epoch": 1.9565264527320034, "grad_norm": 0.4258561134338379, "learning_rate": 3.4908602930509215e-05, "loss": 0.0437, "step": 18047 }, { "epoch": 1.9566348655680832, "grad_norm": 0.16105954349040985, "learning_rate": 3.490497606267228e-05, "loss": 0.0067, "step": 18048 }, { "epoch": 1.9567432784041632, "grad_norm": 0.14158469438552856, "learning_rate": 3.490134919483534e-05, "loss": 0.0131, "step": 18049 }, { "epoch": 1.9568516912402427, "grad_norm": 0.4180658459663391, "learning_rate": 3.489772232699841e-05, "loss": 0.0315, "step": 18050 }, { "epoch": 1.9569601040763227, "grad_norm": 0.1678055077791214, "learning_rate": 3.489409545916147e-05, "loss": 0.0125, "step": 18051 }, { "epoch": 1.9570685169124025, "grad_norm": 0.24015900492668152, "learning_rate": 3.489046859132453e-05, "loss": 0.0116, "step": 18052 }, { "epoch": 1.9571769297484822, "grad_norm": 0.08869631588459015, "learning_rate": 3.488684172348759e-05, "loss": 0.015, "step": 18053 }, { "epoch": 1.957285342584562, "grad_norm": 0.031275901943445206, "learning_rate": 3.4883214855650664e-05, "loss": 0.0022, "step": 18054 }, { "epoch": 1.9573937554206418, "grad_norm": 0.46203669905662537, "learning_rate": 3.487958798781373e-05, "loss": 0.0271, "step": 18055 }, { "epoch": 1.9575021682567217, "grad_norm": 0.322418212890625, "learning_rate": 3.487596111997679e-05, "loss": 0.0497, "step": 18056 }, { "epoch": 1.9576105810928013, "grad_norm": 0.48228123784065247, "learning_rate": 3.4872334252139856e-05, "loss": 0.0312, "step": 18057 }, { "epoch": 1.9577189939288813, "grad_norm": 0.2920283377170563, "learning_rate": 3.486870738430292e-05, "loss": 0.0355, "step": 18058 }, { "epoch": 1.9578274067649608, "grad_norm": 0.5267195701599121, "learning_rate": 3.4865080516465984e-05, "loss": 0.0185, "step": 18059 }, { "epoch": 1.9579358196010408, "grad_norm": 0.1823534369468689, "learning_rate": 3.486145364862904e-05, "loss": 0.0124, "step": 18060 }, { "epoch": 1.9580442324371206, "grad_norm": 0.12940114736557007, "learning_rate": 3.4857826780792106e-05, "loss": 0.01, "step": 18061 }, { "epoch": 1.9581526452732003, "grad_norm": 0.3624146580696106, "learning_rate": 3.485419991295517e-05, "loss": 0.0613, "step": 18062 }, { "epoch": 1.95826105810928, "grad_norm": 0.16164088249206543, "learning_rate": 3.485057304511824e-05, "loss": 0.0149, "step": 18063 }, { "epoch": 1.9583694709453598, "grad_norm": 0.1756267249584198, "learning_rate": 3.4846946177281305e-05, "loss": 0.0084, "step": 18064 }, { "epoch": 1.9584778837814398, "grad_norm": 0.36464858055114746, "learning_rate": 3.484331930944437e-05, "loss": 0.0224, "step": 18065 }, { "epoch": 1.9585862966175194, "grad_norm": 0.0908917710185051, "learning_rate": 3.483969244160743e-05, "loss": 0.0069, "step": 18066 }, { "epoch": 1.9586947094535994, "grad_norm": 0.1351131796836853, "learning_rate": 3.483606557377049e-05, "loss": 0.0147, "step": 18067 }, { "epoch": 1.9588031222896791, "grad_norm": 0.27988478541374207, "learning_rate": 3.4832438705933555e-05, "loss": 0.0121, "step": 18068 }, { "epoch": 1.9589115351257589, "grad_norm": 0.23178958892822266, "learning_rate": 3.482881183809662e-05, "loss": 0.0152, "step": 18069 }, { "epoch": 1.9590199479618386, "grad_norm": 0.13252392411231995, "learning_rate": 3.482518497025968e-05, "loss": 0.0119, "step": 18070 }, { "epoch": 1.9591283607979184, "grad_norm": 0.23246417939662933, "learning_rate": 3.482155810242275e-05, "loss": 0.0099, "step": 18071 }, { "epoch": 1.9592367736339984, "grad_norm": 0.03716409578919411, "learning_rate": 3.481793123458582e-05, "loss": 0.003, "step": 18072 }, { "epoch": 1.959345186470078, "grad_norm": 0.1404065042734146, "learning_rate": 3.481430436674888e-05, "loss": 0.0091, "step": 18073 }, { "epoch": 1.959453599306158, "grad_norm": 0.05279981717467308, "learning_rate": 3.481067749891194e-05, "loss": 0.0038, "step": 18074 }, { "epoch": 1.9595620121422377, "grad_norm": 0.5719907283782959, "learning_rate": 3.4807050631075e-05, "loss": 0.1171, "step": 18075 }, { "epoch": 1.9596704249783174, "grad_norm": 0.16789063811302185, "learning_rate": 3.480342376323807e-05, "loss": 0.0206, "step": 18076 }, { "epoch": 1.9597788378143972, "grad_norm": 0.34049364924430847, "learning_rate": 3.479979689540113e-05, "loss": 0.0245, "step": 18077 }, { "epoch": 1.959887250650477, "grad_norm": 0.19284333288669586, "learning_rate": 3.4796170027564196e-05, "loss": 0.007, "step": 18078 }, { "epoch": 1.959995663486557, "grad_norm": 0.28610309958457947, "learning_rate": 3.479254315972726e-05, "loss": 0.0074, "step": 18079 }, { "epoch": 1.9601040763226365, "grad_norm": 0.09794466197490692, "learning_rate": 3.478891629189033e-05, "loss": 0.006, "step": 18080 }, { "epoch": 1.9602124891587165, "grad_norm": 0.8347213268280029, "learning_rate": 3.478528942405339e-05, "loss": 0.0375, "step": 18081 }, { "epoch": 1.960320901994796, "grad_norm": 0.08025315403938293, "learning_rate": 3.478166255621645e-05, "loss": 0.0073, "step": 18082 }, { "epoch": 1.960429314830876, "grad_norm": 0.35260772705078125, "learning_rate": 3.4778035688379516e-05, "loss": 0.037, "step": 18083 }, { "epoch": 1.9605377276669558, "grad_norm": 0.3127249777317047, "learning_rate": 3.477440882054258e-05, "loss": 0.0358, "step": 18084 }, { "epoch": 1.9606461405030355, "grad_norm": 0.09664583951234818, "learning_rate": 3.4770781952705644e-05, "loss": 0.0094, "step": 18085 }, { "epoch": 1.9607545533391153, "grad_norm": 0.28697553277015686, "learning_rate": 3.476715508486871e-05, "loss": 0.0226, "step": 18086 }, { "epoch": 1.960862966175195, "grad_norm": 0.3330938518047333, "learning_rate": 3.476352821703177e-05, "loss": 0.0199, "step": 18087 }, { "epoch": 1.960971379011275, "grad_norm": 0.31463706493377686, "learning_rate": 3.475990134919484e-05, "loss": 0.0347, "step": 18088 }, { "epoch": 1.9610797918473546, "grad_norm": 0.10582945495843887, "learning_rate": 3.47562744813579e-05, "loss": 0.0046, "step": 18089 }, { "epoch": 1.9611882046834346, "grad_norm": 0.15798239409923553, "learning_rate": 3.4752647613520965e-05, "loss": 0.0066, "step": 18090 }, { "epoch": 1.9612966175195143, "grad_norm": 0.3249526917934418, "learning_rate": 3.474902074568403e-05, "loss": 0.0088, "step": 18091 }, { "epoch": 1.961405030355594, "grad_norm": 0.6106975674629211, "learning_rate": 3.474539387784709e-05, "loss": 0.0149, "step": 18092 }, { "epoch": 1.9615134431916739, "grad_norm": 0.24662566184997559, "learning_rate": 3.474176701001016e-05, "loss": 0.0105, "step": 18093 }, { "epoch": 1.9616218560277536, "grad_norm": 0.36206111311912537, "learning_rate": 3.473814014217322e-05, "loss": 0.0236, "step": 18094 }, { "epoch": 1.9617302688638336, "grad_norm": 0.45935189723968506, "learning_rate": 3.4734513274336285e-05, "loss": 0.0646, "step": 18095 }, { "epoch": 1.9618386816999132, "grad_norm": 0.38376083970069885, "learning_rate": 3.473088640649935e-05, "loss": 0.0168, "step": 18096 }, { "epoch": 1.9619470945359931, "grad_norm": 0.2500099539756775, "learning_rate": 3.4727259538662414e-05, "loss": 0.0245, "step": 18097 }, { "epoch": 1.962055507372073, "grad_norm": 0.6075060367584229, "learning_rate": 3.472363267082548e-05, "loss": 0.0436, "step": 18098 }, { "epoch": 1.9621639202081527, "grad_norm": 0.033615462481975555, "learning_rate": 3.472000580298854e-05, "loss": 0.0021, "step": 18099 }, { "epoch": 1.9622723330442324, "grad_norm": 0.18203526735305786, "learning_rate": 3.4716378935151606e-05, "loss": 0.0146, "step": 18100 }, { "epoch": 1.9623807458803122, "grad_norm": 0.2282760888338089, "learning_rate": 3.471275206731467e-05, "loss": 0.0094, "step": 18101 }, { "epoch": 1.9624891587163922, "grad_norm": 0.3146471083164215, "learning_rate": 3.4709125199477734e-05, "loss": 0.0142, "step": 18102 }, { "epoch": 1.9625975715524717, "grad_norm": 0.631327211856842, "learning_rate": 3.47054983316408e-05, "loss": 0.0297, "step": 18103 }, { "epoch": 1.9627059843885517, "grad_norm": 0.11509621888399124, "learning_rate": 3.4701871463803856e-05, "loss": 0.0103, "step": 18104 }, { "epoch": 1.9628143972246312, "grad_norm": 0.2913076877593994, "learning_rate": 3.469824459596692e-05, "loss": 0.047, "step": 18105 }, { "epoch": 1.9629228100607112, "grad_norm": 0.2152356654405594, "learning_rate": 3.469461772812999e-05, "loss": 0.0161, "step": 18106 }, { "epoch": 1.963031222896791, "grad_norm": 0.049324411898851395, "learning_rate": 3.4690990860293055e-05, "loss": 0.0036, "step": 18107 }, { "epoch": 1.9631396357328708, "grad_norm": 0.32443946599960327, "learning_rate": 3.468736399245612e-05, "loss": 0.0222, "step": 18108 }, { "epoch": 1.9632480485689505, "grad_norm": 0.14122702181339264, "learning_rate": 3.468373712461918e-05, "loss": 0.0047, "step": 18109 }, { "epoch": 1.9633564614050303, "grad_norm": 0.11214371770620346, "learning_rate": 3.468011025678225e-05, "loss": 0.0119, "step": 18110 }, { "epoch": 1.9634648742411103, "grad_norm": 0.33531925082206726, "learning_rate": 3.467648338894531e-05, "loss": 0.0169, "step": 18111 }, { "epoch": 1.9635732870771898, "grad_norm": 0.41178905963897705, "learning_rate": 3.467285652110837e-05, "loss": 0.0165, "step": 18112 }, { "epoch": 1.9636816999132698, "grad_norm": 0.3123442828655243, "learning_rate": 3.466922965327143e-05, "loss": 0.0323, "step": 18113 }, { "epoch": 1.9637901127493496, "grad_norm": 0.40661296248435974, "learning_rate": 3.4665602785434497e-05, "loss": 0.0229, "step": 18114 }, { "epoch": 1.9638985255854293, "grad_norm": 0.1365908980369568, "learning_rate": 3.466197591759757e-05, "loss": 0.0217, "step": 18115 }, { "epoch": 1.964006938421509, "grad_norm": 0.18089836835861206, "learning_rate": 3.465834904976063e-05, "loss": 0.01, "step": 18116 }, { "epoch": 1.9641153512575888, "grad_norm": 0.23331090807914734, "learning_rate": 3.4654722181923696e-05, "loss": 0.0059, "step": 18117 }, { "epoch": 1.9642237640936688, "grad_norm": 0.2567060887813568, "learning_rate": 3.465109531408676e-05, "loss": 0.0201, "step": 18118 }, { "epoch": 1.9643321769297484, "grad_norm": 0.13811400532722473, "learning_rate": 3.464746844624982e-05, "loss": 0.0044, "step": 18119 }, { "epoch": 1.9644405897658284, "grad_norm": 0.2675662636756897, "learning_rate": 3.464384157841288e-05, "loss": 0.0225, "step": 18120 }, { "epoch": 1.964549002601908, "grad_norm": 0.15625663101673126, "learning_rate": 3.4640214710575945e-05, "loss": 0.01, "step": 18121 }, { "epoch": 1.9646574154379879, "grad_norm": 0.28698787093162537, "learning_rate": 3.463658784273901e-05, "loss": 0.0439, "step": 18122 }, { "epoch": 1.9647658282740676, "grad_norm": 0.20168673992156982, "learning_rate": 3.463296097490208e-05, "loss": 0.0232, "step": 18123 }, { "epoch": 1.9648742411101474, "grad_norm": 0.24618588387966156, "learning_rate": 3.4629334107065144e-05, "loss": 0.0298, "step": 18124 }, { "epoch": 1.9649826539462274, "grad_norm": 0.23156194388866425, "learning_rate": 3.462570723922821e-05, "loss": 0.0099, "step": 18125 }, { "epoch": 1.965091066782307, "grad_norm": 0.6472756862640381, "learning_rate": 3.4622080371391266e-05, "loss": 0.0501, "step": 18126 }, { "epoch": 1.965199479618387, "grad_norm": 0.20283560454845428, "learning_rate": 3.461845350355433e-05, "loss": 0.0201, "step": 18127 }, { "epoch": 1.9653078924544665, "grad_norm": 0.17879749834537506, "learning_rate": 3.4614826635717394e-05, "loss": 0.0132, "step": 18128 }, { "epoch": 1.9654163052905465, "grad_norm": 0.15263454616069794, "learning_rate": 3.461119976788046e-05, "loss": 0.0094, "step": 18129 }, { "epoch": 1.9655247181266262, "grad_norm": 0.41423502564430237, "learning_rate": 3.460757290004352e-05, "loss": 0.0483, "step": 18130 }, { "epoch": 1.965633130962706, "grad_norm": 0.38392722606658936, "learning_rate": 3.4603946032206586e-05, "loss": 0.0493, "step": 18131 }, { "epoch": 1.9657415437987857, "grad_norm": 0.26911452412605286, "learning_rate": 3.460031916436966e-05, "loss": 0.0058, "step": 18132 }, { "epoch": 1.9658499566348655, "grad_norm": 0.47063595056533813, "learning_rate": 3.4596692296532715e-05, "loss": 0.0364, "step": 18133 }, { "epoch": 1.9659583694709455, "grad_norm": 0.2746301293373108, "learning_rate": 3.459306542869578e-05, "loss": 0.0107, "step": 18134 }, { "epoch": 1.966066782307025, "grad_norm": 0.2757658362388611, "learning_rate": 3.458943856085884e-05, "loss": 0.0111, "step": 18135 }, { "epoch": 1.966175195143105, "grad_norm": 0.41213923692703247, "learning_rate": 3.458581169302191e-05, "loss": 0.0278, "step": 18136 }, { "epoch": 1.9662836079791848, "grad_norm": 0.40713855624198914, "learning_rate": 3.458218482518497e-05, "loss": 0.0174, "step": 18137 }, { "epoch": 1.9663920208152645, "grad_norm": 0.1856483370065689, "learning_rate": 3.4578557957348035e-05, "loss": 0.0139, "step": 18138 }, { "epoch": 1.9665004336513443, "grad_norm": 0.2797282636165619, "learning_rate": 3.45749310895111e-05, "loss": 0.0119, "step": 18139 }, { "epoch": 1.966608846487424, "grad_norm": 0.0736088901758194, "learning_rate": 3.457130422167416e-05, "loss": 0.0056, "step": 18140 }, { "epoch": 1.966717259323504, "grad_norm": 0.18605034053325653, "learning_rate": 3.456767735383723e-05, "loss": 0.0079, "step": 18141 }, { "epoch": 1.9668256721595836, "grad_norm": 0.3502441346645355, "learning_rate": 3.456405048600029e-05, "loss": 0.0273, "step": 18142 }, { "epoch": 1.9669340849956636, "grad_norm": 0.09776149690151215, "learning_rate": 3.4560423618163356e-05, "loss": 0.0063, "step": 18143 }, { "epoch": 1.9670424978317431, "grad_norm": 0.2760366201400757, "learning_rate": 3.455679675032642e-05, "loss": 0.0375, "step": 18144 }, { "epoch": 1.967150910667823, "grad_norm": 0.1310238093137741, "learning_rate": 3.4553169882489484e-05, "loss": 0.0083, "step": 18145 }, { "epoch": 1.9672593235039029, "grad_norm": 0.3322928547859192, "learning_rate": 3.454954301465255e-05, "loss": 0.0259, "step": 18146 }, { "epoch": 1.9673677363399826, "grad_norm": 0.26482269167900085, "learning_rate": 3.454591614681561e-05, "loss": 0.0167, "step": 18147 }, { "epoch": 1.9674761491760624, "grad_norm": 0.027627913281321526, "learning_rate": 3.4542289278978676e-05, "loss": 0.0023, "step": 18148 }, { "epoch": 1.9675845620121422, "grad_norm": 0.4260930120944977, "learning_rate": 3.453866241114174e-05, "loss": 0.0195, "step": 18149 }, { "epoch": 1.9676929748482221, "grad_norm": 0.3518095910549164, "learning_rate": 3.4535035543304804e-05, "loss": 0.028, "step": 18150 }, { "epoch": 1.9678013876843017, "grad_norm": 0.3535339832305908, "learning_rate": 3.453140867546787e-05, "loss": 0.0315, "step": 18151 }, { "epoch": 1.9679098005203817, "grad_norm": 0.28594356775283813, "learning_rate": 3.452778180763093e-05, "loss": 0.0175, "step": 18152 }, { "epoch": 1.9680182133564614, "grad_norm": 0.0898812860250473, "learning_rate": 3.4524154939794e-05, "loss": 0.0049, "step": 18153 }, { "epoch": 1.9681266261925412, "grad_norm": 0.18882718682289124, "learning_rate": 3.452052807195706e-05, "loss": 0.007, "step": 18154 }, { "epoch": 1.968235039028621, "grad_norm": 0.19508646428585052, "learning_rate": 3.4516901204120125e-05, "loss": 0.0058, "step": 18155 }, { "epoch": 1.9683434518647007, "grad_norm": 0.37230563163757324, "learning_rate": 3.451327433628319e-05, "loss": 0.0212, "step": 18156 }, { "epoch": 1.9684518647007807, "grad_norm": 0.2684052884578705, "learning_rate": 3.4509647468446246e-05, "loss": 0.0247, "step": 18157 }, { "epoch": 1.9685602775368602, "grad_norm": 0.037547241896390915, "learning_rate": 3.450602060060932e-05, "loss": 0.0025, "step": 18158 }, { "epoch": 1.9686686903729402, "grad_norm": 0.39684420824050903, "learning_rate": 3.450239373277238e-05, "loss": 0.0396, "step": 18159 }, { "epoch": 1.96877710320902, "grad_norm": 0.19586768746376038, "learning_rate": 3.4498766864935445e-05, "loss": 0.0085, "step": 18160 }, { "epoch": 1.9688855160450998, "grad_norm": 0.4898887276649475, "learning_rate": 3.449513999709851e-05, "loss": 0.023, "step": 18161 }, { "epoch": 1.9689939288811795, "grad_norm": 0.38761165738105774, "learning_rate": 3.4491513129261574e-05, "loss": 0.0343, "step": 18162 }, { "epoch": 1.9691023417172593, "grad_norm": 0.21699607372283936, "learning_rate": 3.448788626142464e-05, "loss": 0.0123, "step": 18163 }, { "epoch": 1.9692107545533393, "grad_norm": 0.12991023063659668, "learning_rate": 3.4484259393587695e-05, "loss": 0.0033, "step": 18164 }, { "epoch": 1.9693191673894188, "grad_norm": 0.09901297092437744, "learning_rate": 3.448063252575076e-05, "loss": 0.0034, "step": 18165 }, { "epoch": 1.9694275802254988, "grad_norm": 0.35562586784362793, "learning_rate": 3.447700565791382e-05, "loss": 0.0226, "step": 18166 }, { "epoch": 1.9695359930615783, "grad_norm": 0.44102075695991516, "learning_rate": 3.4473378790076894e-05, "loss": 0.0447, "step": 18167 }, { "epoch": 1.9696444058976583, "grad_norm": 0.23449651896953583, "learning_rate": 3.446975192223996e-05, "loss": 0.0236, "step": 18168 }, { "epoch": 1.969752818733738, "grad_norm": 0.43643486499786377, "learning_rate": 3.446612505440302e-05, "loss": 0.0529, "step": 18169 }, { "epoch": 1.9698612315698178, "grad_norm": 0.007923784665763378, "learning_rate": 3.4462498186566086e-05, "loss": 0.0007, "step": 18170 }, { "epoch": 1.9699696444058976, "grad_norm": 0.12875452637672424, "learning_rate": 3.4458871318729144e-05, "loss": 0.0078, "step": 18171 }, { "epoch": 1.9700780572419774, "grad_norm": 0.3925309181213379, "learning_rate": 3.445524445089221e-05, "loss": 0.0286, "step": 18172 }, { "epoch": 1.9701864700780574, "grad_norm": 0.16449852287769318, "learning_rate": 3.445161758305527e-05, "loss": 0.0069, "step": 18173 }, { "epoch": 1.970294882914137, "grad_norm": 0.09894625842571259, "learning_rate": 3.4447990715218336e-05, "loss": 0.0048, "step": 18174 }, { "epoch": 1.9704032957502169, "grad_norm": 0.3320346772670746, "learning_rate": 3.444436384738141e-05, "loss": 0.0169, "step": 18175 }, { "epoch": 1.9705117085862967, "grad_norm": 0.21435558795928955, "learning_rate": 3.444073697954447e-05, "loss": 0.0211, "step": 18176 }, { "epoch": 1.9706201214223764, "grad_norm": 0.15624341368675232, "learning_rate": 3.4437110111707535e-05, "loss": 0.0065, "step": 18177 }, { "epoch": 1.9707285342584562, "grad_norm": 0.20727503299713135, "learning_rate": 3.443348324387059e-05, "loss": 0.0222, "step": 18178 }, { "epoch": 1.970836947094536, "grad_norm": 0.21224094927310944, "learning_rate": 3.442985637603366e-05, "loss": 0.0276, "step": 18179 }, { "epoch": 1.970945359930616, "grad_norm": 0.34777653217315674, "learning_rate": 3.442622950819672e-05, "loss": 0.0223, "step": 18180 }, { "epoch": 1.9710537727666955, "grad_norm": 0.15074753761291504, "learning_rate": 3.4422602640359785e-05, "loss": 0.0061, "step": 18181 }, { "epoch": 1.9711621856027755, "grad_norm": 0.6351958513259888, "learning_rate": 3.441897577252285e-05, "loss": 0.0452, "step": 18182 }, { "epoch": 1.9712705984388552, "grad_norm": 0.43833598494529724, "learning_rate": 3.441534890468591e-05, "loss": 0.0272, "step": 18183 }, { "epoch": 1.971379011274935, "grad_norm": 0.7216944098472595, "learning_rate": 3.4411722036848984e-05, "loss": 0.0328, "step": 18184 }, { "epoch": 1.9714874241110147, "grad_norm": 0.5739669799804688, "learning_rate": 3.440809516901204e-05, "loss": 0.0251, "step": 18185 }, { "epoch": 1.9715958369470945, "grad_norm": 0.5471044778823853, "learning_rate": 3.4404468301175105e-05, "loss": 0.0415, "step": 18186 }, { "epoch": 1.9717042497831745, "grad_norm": 0.11049075424671173, "learning_rate": 3.440084143333817e-05, "loss": 0.0104, "step": 18187 }, { "epoch": 1.971812662619254, "grad_norm": 0.1508290022611618, "learning_rate": 3.4397214565501234e-05, "loss": 0.0085, "step": 18188 }, { "epoch": 1.971921075455334, "grad_norm": 0.49615582823753357, "learning_rate": 3.43935876976643e-05, "loss": 0.058, "step": 18189 }, { "epoch": 1.9720294882914136, "grad_norm": 0.17562031745910645, "learning_rate": 3.438996082982736e-05, "loss": 0.0188, "step": 18190 }, { "epoch": 1.9721379011274935, "grad_norm": 0.32504698634147644, "learning_rate": 3.4386333961990426e-05, "loss": 0.0103, "step": 18191 }, { "epoch": 1.9722463139635733, "grad_norm": 0.04498681426048279, "learning_rate": 3.43827070941535e-05, "loss": 0.0022, "step": 18192 }, { "epoch": 1.972354726799653, "grad_norm": 0.570600152015686, "learning_rate": 3.4379080226316554e-05, "loss": 0.0386, "step": 18193 }, { "epoch": 1.9724631396357328, "grad_norm": 0.3690933883190155, "learning_rate": 3.437545335847962e-05, "loss": 0.0142, "step": 18194 }, { "epoch": 1.9725715524718126, "grad_norm": 0.04699675738811493, "learning_rate": 3.437182649064268e-05, "loss": 0.0029, "step": 18195 }, { "epoch": 1.9726799653078926, "grad_norm": 0.36665448546409607, "learning_rate": 3.4368199622805746e-05, "loss": 0.0434, "step": 18196 }, { "epoch": 1.9727883781439721, "grad_norm": 0.10022545605897903, "learning_rate": 3.436457275496881e-05, "loss": 0.0112, "step": 18197 }, { "epoch": 1.972896790980052, "grad_norm": 0.16686084866523743, "learning_rate": 3.4360945887131875e-05, "loss": 0.0133, "step": 18198 }, { "epoch": 1.9730052038161319, "grad_norm": 0.479827880859375, "learning_rate": 3.435731901929494e-05, "loss": 0.0177, "step": 18199 }, { "epoch": 1.9731136166522116, "grad_norm": 0.41634148359298706, "learning_rate": 3.4353692151458e-05, "loss": 0.0258, "step": 18200 }, { "epoch": 1.9732220294882914, "grad_norm": 0.266673743724823, "learning_rate": 3.435006528362107e-05, "loss": 0.0395, "step": 18201 }, { "epoch": 1.9733304423243712, "grad_norm": 0.3024066984653473, "learning_rate": 3.434643841578413e-05, "loss": 0.0226, "step": 18202 }, { "epoch": 1.9734388551604511, "grad_norm": 0.034575458616018295, "learning_rate": 3.4342811547947195e-05, "loss": 0.0017, "step": 18203 }, { "epoch": 1.9735472679965307, "grad_norm": 0.35508105158805847, "learning_rate": 3.433918468011026e-05, "loss": 0.0374, "step": 18204 }, { "epoch": 1.9736556808326107, "grad_norm": 0.025507722049951553, "learning_rate": 3.433555781227332e-05, "loss": 0.0017, "step": 18205 }, { "epoch": 1.9737640936686902, "grad_norm": 0.4201156198978424, "learning_rate": 3.433193094443639e-05, "loss": 0.0265, "step": 18206 }, { "epoch": 1.9738725065047702, "grad_norm": 0.34719419479370117, "learning_rate": 3.432830407659945e-05, "loss": 0.0301, "step": 18207 }, { "epoch": 1.97398091934085, "grad_norm": 0.5910968780517578, "learning_rate": 3.4324677208762516e-05, "loss": 0.021, "step": 18208 }, { "epoch": 1.9740893321769297, "grad_norm": 0.5857619643211365, "learning_rate": 3.432105034092557e-05, "loss": 0.0227, "step": 18209 }, { "epoch": 1.9741977450130095, "grad_norm": 0.10306449979543686, "learning_rate": 3.4317423473088644e-05, "loss": 0.0035, "step": 18210 }, { "epoch": 1.9743061578490892, "grad_norm": 0.2661355137825012, "learning_rate": 3.431379660525171e-05, "loss": 0.0518, "step": 18211 }, { "epoch": 1.9744145706851692, "grad_norm": 0.6572985649108887, "learning_rate": 3.431016973741477e-05, "loss": 0.036, "step": 18212 }, { "epoch": 1.9745229835212488, "grad_norm": 0.5463226437568665, "learning_rate": 3.4306542869577836e-05, "loss": 0.0374, "step": 18213 }, { "epoch": 1.9746313963573288, "grad_norm": 0.5436978936195374, "learning_rate": 3.43029160017409e-05, "loss": 0.0277, "step": 18214 }, { "epoch": 1.9747398091934085, "grad_norm": 0.42436984181404114, "learning_rate": 3.4299289133903964e-05, "loss": 0.0161, "step": 18215 }, { "epoch": 1.9748482220294883, "grad_norm": 0.31794804334640503, "learning_rate": 3.429566226606702e-05, "loss": 0.0143, "step": 18216 }, { "epoch": 1.974956634865568, "grad_norm": 0.5260637402534485, "learning_rate": 3.4292035398230086e-05, "loss": 0.0497, "step": 18217 }, { "epoch": 1.9750650477016478, "grad_norm": 0.5602203011512756, "learning_rate": 3.428840853039316e-05, "loss": 0.0445, "step": 18218 }, { "epoch": 1.9751734605377278, "grad_norm": 0.02001717872917652, "learning_rate": 3.428478166255622e-05, "loss": 0.0015, "step": 18219 }, { "epoch": 1.9752818733738073, "grad_norm": 0.2013360857963562, "learning_rate": 3.4281154794719285e-05, "loss": 0.0095, "step": 18220 }, { "epoch": 1.9753902862098873, "grad_norm": 0.2979792356491089, "learning_rate": 3.427752792688235e-05, "loss": 0.0132, "step": 18221 }, { "epoch": 1.975498699045967, "grad_norm": 0.5201578140258789, "learning_rate": 3.427390105904541e-05, "loss": 0.053, "step": 18222 }, { "epoch": 1.9756071118820469, "grad_norm": 0.14752987027168274, "learning_rate": 3.427027419120847e-05, "loss": 0.006, "step": 18223 }, { "epoch": 1.9757155247181266, "grad_norm": 0.5405149459838867, "learning_rate": 3.4266647323371535e-05, "loss": 0.0236, "step": 18224 }, { "epoch": 1.9758239375542064, "grad_norm": 1.021347165107727, "learning_rate": 3.42630204555346e-05, "loss": 0.0156, "step": 18225 }, { "epoch": 1.9759323503902864, "grad_norm": 0.31757500767707825, "learning_rate": 3.425939358769766e-05, "loss": 0.0366, "step": 18226 }, { "epoch": 1.976040763226366, "grad_norm": 0.2568012773990631, "learning_rate": 3.4255766719860734e-05, "loss": 0.0101, "step": 18227 }, { "epoch": 1.976149176062446, "grad_norm": 0.23416483402252197, "learning_rate": 3.42521398520238e-05, "loss": 0.0251, "step": 18228 }, { "epoch": 1.9762575888985254, "grad_norm": 0.14212800562381744, "learning_rate": 3.424851298418686e-05, "loss": 0.0045, "step": 18229 }, { "epoch": 1.9763660017346054, "grad_norm": 0.21352458000183105, "learning_rate": 3.424488611634992e-05, "loss": 0.0172, "step": 18230 }, { "epoch": 1.9764744145706852, "grad_norm": 0.10706482082605362, "learning_rate": 3.424125924851298e-05, "loss": 0.0054, "step": 18231 }, { "epoch": 1.976582827406765, "grad_norm": 0.22380170226097107, "learning_rate": 3.423763238067605e-05, "loss": 0.0112, "step": 18232 }, { "epoch": 1.9766912402428447, "grad_norm": 0.25933754444122314, "learning_rate": 3.423400551283911e-05, "loss": 0.0899, "step": 18233 }, { "epoch": 1.9767996530789245, "grad_norm": 0.21167634427547455, "learning_rate": 3.4230378645002176e-05, "loss": 0.0129, "step": 18234 }, { "epoch": 1.9769080659150045, "grad_norm": 0.09747377038002014, "learning_rate": 3.422675177716524e-05, "loss": 0.0078, "step": 18235 }, { "epoch": 1.977016478751084, "grad_norm": 0.4687906503677368, "learning_rate": 3.422312490932831e-05, "loss": 0.0323, "step": 18236 }, { "epoch": 1.977124891587164, "grad_norm": 0.1274549514055252, "learning_rate": 3.421949804149137e-05, "loss": 0.0093, "step": 18237 }, { "epoch": 1.9772333044232437, "grad_norm": 0.6698305606842041, "learning_rate": 3.421587117365443e-05, "loss": 0.0306, "step": 18238 }, { "epoch": 1.9773417172593235, "grad_norm": 0.299827516078949, "learning_rate": 3.4212244305817496e-05, "loss": 0.0106, "step": 18239 }, { "epoch": 1.9774501300954033, "grad_norm": 0.1578783243894577, "learning_rate": 3.420861743798056e-05, "loss": 0.0041, "step": 18240 }, { "epoch": 1.977558542931483, "grad_norm": 0.09714178740978241, "learning_rate": 3.4204990570143624e-05, "loss": 0.0089, "step": 18241 }, { "epoch": 1.977666955767563, "grad_norm": 0.5668211579322815, "learning_rate": 3.420136370230669e-05, "loss": 0.0349, "step": 18242 }, { "epoch": 1.9777753686036426, "grad_norm": 0.49993205070495605, "learning_rate": 3.419773683446975e-05, "loss": 0.0286, "step": 18243 }, { "epoch": 1.9778837814397225, "grad_norm": 0.09565512090921402, "learning_rate": 3.4194109966632823e-05, "loss": 0.0075, "step": 18244 }, { "epoch": 1.9779921942758023, "grad_norm": 1.0165281295776367, "learning_rate": 3.419048309879588e-05, "loss": 0.0164, "step": 18245 }, { "epoch": 1.978100607111882, "grad_norm": 0.47434762120246887, "learning_rate": 3.4186856230958945e-05, "loss": 0.0254, "step": 18246 }, { "epoch": 1.9782090199479618, "grad_norm": 0.2908584773540497, "learning_rate": 3.418322936312201e-05, "loss": 0.0084, "step": 18247 }, { "epoch": 1.9783174327840416, "grad_norm": 0.26704269647598267, "learning_rate": 3.417960249528507e-05, "loss": 0.0109, "step": 18248 }, { "epoch": 1.9784258456201216, "grad_norm": 0.3295659124851227, "learning_rate": 3.417597562744814e-05, "loss": 0.0405, "step": 18249 }, { "epoch": 1.9785342584562011, "grad_norm": 0.10155724734067917, "learning_rate": 3.41723487596112e-05, "loss": 0.0046, "step": 18250 }, { "epoch": 1.978642671292281, "grad_norm": 0.5132068991661072, "learning_rate": 3.4168721891774265e-05, "loss": 0.0206, "step": 18251 }, { "epoch": 1.9787510841283606, "grad_norm": 0.14604242146015167, "learning_rate": 3.416509502393733e-05, "loss": 0.0064, "step": 18252 }, { "epoch": 1.9788594969644406, "grad_norm": 0.03388093784451485, "learning_rate": 3.4161468156100394e-05, "loss": 0.0018, "step": 18253 }, { "epoch": 1.9789679098005204, "grad_norm": 0.6165720224380493, "learning_rate": 3.415784128826346e-05, "loss": 0.0954, "step": 18254 }, { "epoch": 1.9790763226366002, "grad_norm": 0.32435405254364014, "learning_rate": 3.415421442042652e-05, "loss": 0.0344, "step": 18255 }, { "epoch": 1.97918473547268, "grad_norm": 0.1568147987127304, "learning_rate": 3.4150587552589586e-05, "loss": 0.0288, "step": 18256 }, { "epoch": 1.9792931483087597, "grad_norm": 0.5166603326797485, "learning_rate": 3.414696068475265e-05, "loss": 0.0316, "step": 18257 }, { "epoch": 1.9794015611448397, "grad_norm": 0.4227866232395172, "learning_rate": 3.4143333816915714e-05, "loss": 0.0096, "step": 18258 }, { "epoch": 1.9795099739809192, "grad_norm": 0.6699614524841309, "learning_rate": 3.413970694907878e-05, "loss": 0.0317, "step": 18259 }, { "epoch": 1.9796183868169992, "grad_norm": 0.10989491641521454, "learning_rate": 3.413608008124184e-05, "loss": 0.0043, "step": 18260 }, { "epoch": 1.979726799653079, "grad_norm": 0.19671137630939484, "learning_rate": 3.41324532134049e-05, "loss": 0.0358, "step": 18261 }, { "epoch": 1.9798352124891587, "grad_norm": 0.31989285349845886, "learning_rate": 3.412882634556797e-05, "loss": 0.0222, "step": 18262 }, { "epoch": 1.9799436253252385, "grad_norm": 0.0787743479013443, "learning_rate": 3.4125199477731035e-05, "loss": 0.0025, "step": 18263 }, { "epoch": 1.9800520381613183, "grad_norm": 0.1820041984319687, "learning_rate": 3.41215726098941e-05, "loss": 0.0177, "step": 18264 }, { "epoch": 1.9801604509973982, "grad_norm": 0.06022098660469055, "learning_rate": 3.411794574205716e-05, "loss": 0.0047, "step": 18265 }, { "epoch": 1.9802688638334778, "grad_norm": 0.739500880241394, "learning_rate": 3.411431887422023e-05, "loss": 0.0312, "step": 18266 }, { "epoch": 1.9803772766695578, "grad_norm": 0.3642283082008362, "learning_rate": 3.411069200638329e-05, "loss": 0.0361, "step": 18267 }, { "epoch": 1.9804856895056373, "grad_norm": 0.547404944896698, "learning_rate": 3.410706513854635e-05, "loss": 0.021, "step": 18268 }, { "epoch": 1.9805941023417173, "grad_norm": 0.39698776602745056, "learning_rate": 3.410343827070941e-05, "loss": 0.0192, "step": 18269 }, { "epoch": 1.980702515177797, "grad_norm": 0.32892340421676636, "learning_rate": 3.4099811402872483e-05, "loss": 0.0182, "step": 18270 }, { "epoch": 1.9808109280138768, "grad_norm": 0.5297157168388367, "learning_rate": 3.409618453503555e-05, "loss": 0.0164, "step": 18271 }, { "epoch": 1.9809193408499568, "grad_norm": 0.06010344624519348, "learning_rate": 3.409255766719861e-05, "loss": 0.0023, "step": 18272 }, { "epoch": 1.9810277536860363, "grad_norm": 0.3418269455432892, "learning_rate": 3.4088930799361676e-05, "loss": 0.0067, "step": 18273 }, { "epoch": 1.9811361665221163, "grad_norm": 0.18213912844657898, "learning_rate": 3.408530393152474e-05, "loss": 0.0137, "step": 18274 }, { "epoch": 1.9812445793581959, "grad_norm": 0.42709365487098694, "learning_rate": 3.40816770636878e-05, "loss": 0.0292, "step": 18275 }, { "epoch": 1.9813529921942759, "grad_norm": 0.25118115544319153, "learning_rate": 3.407805019585086e-05, "loss": 0.0136, "step": 18276 }, { "epoch": 1.9814614050303556, "grad_norm": 0.009074204601347446, "learning_rate": 3.4074423328013925e-05, "loss": 0.0006, "step": 18277 }, { "epoch": 1.9815698178664354, "grad_norm": 0.3290278911590576, "learning_rate": 3.407079646017699e-05, "loss": 0.0176, "step": 18278 }, { "epoch": 1.9816782307025151, "grad_norm": 0.1658206284046173, "learning_rate": 3.406716959234006e-05, "loss": 0.0051, "step": 18279 }, { "epoch": 1.981786643538595, "grad_norm": 0.2491174042224884, "learning_rate": 3.4063542724503124e-05, "loss": 0.0143, "step": 18280 }, { "epoch": 1.981895056374675, "grad_norm": 0.42666077613830566, "learning_rate": 3.405991585666619e-05, "loss": 0.0214, "step": 18281 }, { "epoch": 1.9820034692107544, "grad_norm": 0.3557695150375366, "learning_rate": 3.4056288988829246e-05, "loss": 0.0359, "step": 18282 }, { "epoch": 1.9821118820468344, "grad_norm": 0.205272376537323, "learning_rate": 3.405266212099231e-05, "loss": 0.0089, "step": 18283 }, { "epoch": 1.9822202948829142, "grad_norm": 0.47130686044692993, "learning_rate": 3.4049035253155374e-05, "loss": 0.021, "step": 18284 }, { "epoch": 1.982328707718994, "grad_norm": 0.7539918422698975, "learning_rate": 3.404540838531844e-05, "loss": 0.0197, "step": 18285 }, { "epoch": 1.9824371205550737, "grad_norm": 0.3480517566204071, "learning_rate": 3.40417815174815e-05, "loss": 0.0515, "step": 18286 }, { "epoch": 1.9825455333911535, "grad_norm": 0.33863237500190735, "learning_rate": 3.403815464964457e-05, "loss": 0.0104, "step": 18287 }, { "epoch": 1.9826539462272335, "grad_norm": 0.30483442544937134, "learning_rate": 3.403452778180764e-05, "loss": 0.0257, "step": 18288 }, { "epoch": 1.982762359063313, "grad_norm": 0.8200328350067139, "learning_rate": 3.4030900913970695e-05, "loss": 0.0331, "step": 18289 }, { "epoch": 1.982870771899393, "grad_norm": 0.37818413972854614, "learning_rate": 3.402727404613376e-05, "loss": 0.042, "step": 18290 }, { "epoch": 1.9829791847354725, "grad_norm": 0.09554973244667053, "learning_rate": 3.402364717829682e-05, "loss": 0.0071, "step": 18291 }, { "epoch": 1.9830875975715525, "grad_norm": 0.19352921843528748, "learning_rate": 3.402002031045989e-05, "loss": 0.023, "step": 18292 }, { "epoch": 1.9831960104076323, "grad_norm": 0.09727716445922852, "learning_rate": 3.401639344262295e-05, "loss": 0.0055, "step": 18293 }, { "epoch": 1.983304423243712, "grad_norm": 0.11476528644561768, "learning_rate": 3.4012766574786015e-05, "loss": 0.0053, "step": 18294 }, { "epoch": 1.9834128360797918, "grad_norm": 0.4497016966342926, "learning_rate": 3.400913970694908e-05, "loss": 0.0438, "step": 18295 }, { "epoch": 1.9835212489158716, "grad_norm": 0.35097044706344604, "learning_rate": 3.400551283911215e-05, "loss": 0.0517, "step": 18296 }, { "epoch": 1.9836296617519515, "grad_norm": 0.2165139615535736, "learning_rate": 3.400188597127521e-05, "loss": 0.0167, "step": 18297 }, { "epoch": 1.983738074588031, "grad_norm": 0.16172172129154205, "learning_rate": 3.399825910343827e-05, "loss": 0.0129, "step": 18298 }, { "epoch": 1.983846487424111, "grad_norm": 0.20296518504619598, "learning_rate": 3.3994632235601336e-05, "loss": 0.0317, "step": 18299 }, { "epoch": 1.9839549002601908, "grad_norm": 0.32071322202682495, "learning_rate": 3.39910053677644e-05, "loss": 0.0355, "step": 18300 }, { "epoch": 1.9840633130962706, "grad_norm": 0.27769896388053894, "learning_rate": 3.3987378499927464e-05, "loss": 0.0225, "step": 18301 }, { "epoch": 1.9841717259323504, "grad_norm": 0.5903927087783813, "learning_rate": 3.398375163209053e-05, "loss": 0.032, "step": 18302 }, { "epoch": 1.9842801387684301, "grad_norm": 0.6038026213645935, "learning_rate": 3.398012476425359e-05, "loss": 0.0831, "step": 18303 }, { "epoch": 1.98438855160451, "grad_norm": 0.3944522440433502, "learning_rate": 3.3976497896416656e-05, "loss": 0.0482, "step": 18304 }, { "epoch": 1.9844969644405897, "grad_norm": 0.27622532844543457, "learning_rate": 3.397287102857972e-05, "loss": 0.021, "step": 18305 }, { "epoch": 1.9846053772766696, "grad_norm": 0.6349733471870422, "learning_rate": 3.3969244160742784e-05, "loss": 0.0398, "step": 18306 }, { "epoch": 1.9847137901127494, "grad_norm": 0.2038789540529251, "learning_rate": 3.396561729290585e-05, "loss": 0.0272, "step": 18307 }, { "epoch": 1.9848222029488292, "grad_norm": 0.3049021363258362, "learning_rate": 3.396199042506891e-05, "loss": 0.0178, "step": 18308 }, { "epoch": 1.984930615784909, "grad_norm": 0.09665058553218842, "learning_rate": 3.395836355723198e-05, "loss": 0.0065, "step": 18309 }, { "epoch": 1.9850390286209887, "grad_norm": 0.15226781368255615, "learning_rate": 3.395473668939504e-05, "loss": 0.0083, "step": 18310 }, { "epoch": 1.9851474414570687, "grad_norm": 0.08101841062307358, "learning_rate": 3.3951109821558105e-05, "loss": 0.0065, "step": 18311 }, { "epoch": 1.9852558542931482, "grad_norm": 0.20765531063079834, "learning_rate": 3.394748295372117e-05, "loss": 0.0272, "step": 18312 }, { "epoch": 1.9853642671292282, "grad_norm": 0.4430813789367676, "learning_rate": 3.394385608588423e-05, "loss": 0.0467, "step": 18313 }, { "epoch": 1.9854726799653077, "grad_norm": 0.369577556848526, "learning_rate": 3.39402292180473e-05, "loss": 0.029, "step": 18314 }, { "epoch": 1.9855810928013877, "grad_norm": 0.19595032930374146, "learning_rate": 3.393660235021036e-05, "loss": 0.0224, "step": 18315 }, { "epoch": 1.9856895056374675, "grad_norm": 0.40141382813453674, "learning_rate": 3.3932975482373425e-05, "loss": 0.0278, "step": 18316 }, { "epoch": 1.9857979184735473, "grad_norm": 0.27284902334213257, "learning_rate": 3.392934861453649e-05, "loss": 0.0219, "step": 18317 }, { "epoch": 1.985906331309627, "grad_norm": 0.06570786237716675, "learning_rate": 3.3925721746699554e-05, "loss": 0.004, "step": 18318 }, { "epoch": 1.9860147441457068, "grad_norm": 0.36180225014686584, "learning_rate": 3.392209487886262e-05, "loss": 0.0517, "step": 18319 }, { "epoch": 1.9861231569817868, "grad_norm": 0.3120826482772827, "learning_rate": 3.3918468011025675e-05, "loss": 0.0128, "step": 18320 }, { "epoch": 1.9862315698178663, "grad_norm": 0.2575778067111969, "learning_rate": 3.391484114318874e-05, "loss": 0.0218, "step": 18321 }, { "epoch": 1.9863399826539463, "grad_norm": 0.1975460946559906, "learning_rate": 3.391121427535181e-05, "loss": 0.0235, "step": 18322 }, { "epoch": 1.986448395490026, "grad_norm": 0.10237597674131393, "learning_rate": 3.3907587407514874e-05, "loss": 0.0069, "step": 18323 }, { "epoch": 1.9865568083261058, "grad_norm": 0.4869581162929535, "learning_rate": 3.390396053967794e-05, "loss": 0.0379, "step": 18324 }, { "epoch": 1.9866652211621856, "grad_norm": 0.12774448096752167, "learning_rate": 3.3900333671841e-05, "loss": 0.0054, "step": 18325 }, { "epoch": 1.9867736339982653, "grad_norm": 0.22827251255512238, "learning_rate": 3.3896706804004067e-05, "loss": 0.0133, "step": 18326 }, { "epoch": 1.9868820468343453, "grad_norm": 0.23609070479869843, "learning_rate": 3.3893079936167124e-05, "loss": 0.0098, "step": 18327 }, { "epoch": 1.9869904596704249, "grad_norm": 0.28791576623916626, "learning_rate": 3.388945306833019e-05, "loss": 0.0324, "step": 18328 }, { "epoch": 1.9870988725065049, "grad_norm": 0.055514898151159286, "learning_rate": 3.388582620049325e-05, "loss": 0.0027, "step": 18329 }, { "epoch": 1.9872072853425846, "grad_norm": 0.06952733546495438, "learning_rate": 3.3882199332656316e-05, "loss": 0.0053, "step": 18330 }, { "epoch": 1.9873156981786644, "grad_norm": 0.3744919002056122, "learning_rate": 3.387857246481939e-05, "loss": 0.0272, "step": 18331 }, { "epoch": 1.9874241110147441, "grad_norm": 0.43717652559280396, "learning_rate": 3.387494559698245e-05, "loss": 0.058, "step": 18332 }, { "epoch": 1.987532523850824, "grad_norm": 0.7355182766914368, "learning_rate": 3.3871318729145515e-05, "loss": 0.0636, "step": 18333 }, { "epoch": 1.987640936686904, "grad_norm": 0.3301489055156708, "learning_rate": 3.386769186130857e-05, "loss": 0.0221, "step": 18334 }, { "epoch": 1.9877493495229834, "grad_norm": 0.26757991313934326, "learning_rate": 3.386406499347164e-05, "loss": 0.0211, "step": 18335 }, { "epoch": 1.9878577623590634, "grad_norm": 0.20549722015857697, "learning_rate": 3.38604381256347e-05, "loss": 0.0082, "step": 18336 }, { "epoch": 1.987966175195143, "grad_norm": 0.3791678845882416, "learning_rate": 3.3856811257797765e-05, "loss": 0.03, "step": 18337 }, { "epoch": 1.988074588031223, "grad_norm": 0.345289945602417, "learning_rate": 3.385318438996083e-05, "loss": 0.0183, "step": 18338 }, { "epoch": 1.9881830008673027, "grad_norm": 0.1582217663526535, "learning_rate": 3.38495575221239e-05, "loss": 0.0173, "step": 18339 }, { "epoch": 1.9882914137033825, "grad_norm": 0.08841009438037872, "learning_rate": 3.3845930654286964e-05, "loss": 0.0069, "step": 18340 }, { "epoch": 1.9883998265394622, "grad_norm": 0.05176493898034096, "learning_rate": 3.384230378645003e-05, "loss": 0.0046, "step": 18341 }, { "epoch": 1.988508239375542, "grad_norm": 0.17720043659210205, "learning_rate": 3.3838676918613085e-05, "loss": 0.0092, "step": 18342 }, { "epoch": 1.988616652211622, "grad_norm": 0.14381255209445953, "learning_rate": 3.383505005077615e-05, "loss": 0.0166, "step": 18343 }, { "epoch": 1.9887250650477015, "grad_norm": 0.19541172683238983, "learning_rate": 3.3831423182939214e-05, "loss": 0.0182, "step": 18344 }, { "epoch": 1.9888334778837815, "grad_norm": 0.29782336950302124, "learning_rate": 3.382779631510228e-05, "loss": 0.0088, "step": 18345 }, { "epoch": 1.9889418907198613, "grad_norm": 0.2400396466255188, "learning_rate": 3.382416944726534e-05, "loss": 0.0196, "step": 18346 }, { "epoch": 1.989050303555941, "grad_norm": 0.07614895701408386, "learning_rate": 3.3820542579428406e-05, "loss": 0.0038, "step": 18347 }, { "epoch": 1.9891587163920208, "grad_norm": 0.29446929693222046, "learning_rate": 3.381691571159148e-05, "loss": 0.0182, "step": 18348 }, { "epoch": 1.9892671292281006, "grad_norm": 0.17957550287246704, "learning_rate": 3.3813288843754534e-05, "loss": 0.0179, "step": 18349 }, { "epoch": 1.9893755420641805, "grad_norm": 0.7513321042060852, "learning_rate": 3.38096619759176e-05, "loss": 0.0265, "step": 18350 }, { "epoch": 1.98948395490026, "grad_norm": 0.5383992791175842, "learning_rate": 3.380603510808066e-05, "loss": 0.0403, "step": 18351 }, { "epoch": 1.98959236773634, "grad_norm": 0.10200928151607513, "learning_rate": 3.3802408240243726e-05, "loss": 0.0068, "step": 18352 }, { "epoch": 1.9897007805724196, "grad_norm": 0.33500537276268005, "learning_rate": 3.379878137240679e-05, "loss": 0.0203, "step": 18353 }, { "epoch": 1.9898091934084996, "grad_norm": 0.24003110826015472, "learning_rate": 3.3795154504569855e-05, "loss": 0.0165, "step": 18354 }, { "epoch": 1.9899176062445794, "grad_norm": 0.3596959412097931, "learning_rate": 3.379152763673292e-05, "loss": 0.0271, "step": 18355 }, { "epoch": 1.9900260190806591, "grad_norm": 0.2373732626438141, "learning_rate": 3.378790076889598e-05, "loss": 0.029, "step": 18356 }, { "epoch": 1.9901344319167389, "grad_norm": 0.26110026240348816, "learning_rate": 3.378427390105905e-05, "loss": 0.0164, "step": 18357 }, { "epoch": 1.9902428447528187, "grad_norm": 0.018814291805028915, "learning_rate": 3.378064703322211e-05, "loss": 0.0009, "step": 18358 }, { "epoch": 1.9903512575888986, "grad_norm": 0.5884260535240173, "learning_rate": 3.3777020165385175e-05, "loss": 0.0268, "step": 18359 }, { "epoch": 1.9904596704249782, "grad_norm": 0.4342050552368164, "learning_rate": 3.377339329754824e-05, "loss": 0.028, "step": 18360 }, { "epoch": 1.9905680832610582, "grad_norm": 0.19835583865642548, "learning_rate": 3.3769766429711303e-05, "loss": 0.018, "step": 18361 }, { "epoch": 1.990676496097138, "grad_norm": 0.08151987195014954, "learning_rate": 3.376613956187437e-05, "loss": 0.003, "step": 18362 }, { "epoch": 1.9907849089332177, "grad_norm": 0.1885734349489212, "learning_rate": 3.376251269403743e-05, "loss": 0.0174, "step": 18363 }, { "epoch": 1.9908933217692975, "grad_norm": 0.24005751311779022, "learning_rate": 3.3758885826200496e-05, "loss": 0.0207, "step": 18364 }, { "epoch": 1.9910017346053772, "grad_norm": 0.3040619194507599, "learning_rate": 3.375525895836356e-05, "loss": 0.0168, "step": 18365 }, { "epoch": 1.9911101474414572, "grad_norm": 0.18895909190177917, "learning_rate": 3.3751632090526624e-05, "loss": 0.0236, "step": 18366 }, { "epoch": 1.9912185602775367, "grad_norm": 0.16528508067131042, "learning_rate": 3.374800522268969e-05, "loss": 0.0209, "step": 18367 }, { "epoch": 1.9913269731136167, "grad_norm": 0.4022747576236725, "learning_rate": 3.374437835485275e-05, "loss": 0.0094, "step": 18368 }, { "epoch": 1.9914353859496965, "grad_norm": 0.2782110571861267, "learning_rate": 3.3740751487015816e-05, "loss": 0.0241, "step": 18369 }, { "epoch": 1.9915437987857763, "grad_norm": 0.2592637836933136, "learning_rate": 3.373712461917888e-05, "loss": 0.0199, "step": 18370 }, { "epoch": 1.991652211621856, "grad_norm": 0.31187471747398376, "learning_rate": 3.3733497751341944e-05, "loss": 0.0424, "step": 18371 }, { "epoch": 1.9917606244579358, "grad_norm": 0.6041263937950134, "learning_rate": 3.3729870883505e-05, "loss": 0.0233, "step": 18372 }, { "epoch": 1.9918690372940158, "grad_norm": 0.22743897140026093, "learning_rate": 3.3726244015668066e-05, "loss": 0.021, "step": 18373 }, { "epoch": 1.9919774501300953, "grad_norm": 0.46587786078453064, "learning_rate": 3.372261714783114e-05, "loss": 0.0393, "step": 18374 }, { "epoch": 1.9920858629661753, "grad_norm": 0.32035231590270996, "learning_rate": 3.37189902799942e-05, "loss": 0.0275, "step": 18375 }, { "epoch": 1.9921942758022548, "grad_norm": 0.258626788854599, "learning_rate": 3.3715363412157265e-05, "loss": 0.025, "step": 18376 }, { "epoch": 1.9923026886383348, "grad_norm": 0.13877655565738678, "learning_rate": 3.371173654432033e-05, "loss": 0.0113, "step": 18377 }, { "epoch": 1.9924111014744146, "grad_norm": 0.2763576805591583, "learning_rate": 3.370810967648339e-05, "loss": 0.0281, "step": 18378 }, { "epoch": 1.9925195143104943, "grad_norm": 0.22571247816085815, "learning_rate": 3.370448280864645e-05, "loss": 0.024, "step": 18379 }, { "epoch": 1.992627927146574, "grad_norm": 0.18480059504508972, "learning_rate": 3.3700855940809515e-05, "loss": 0.0163, "step": 18380 }, { "epoch": 1.9927363399826539, "grad_norm": 0.5122673511505127, "learning_rate": 3.369722907297258e-05, "loss": 0.0258, "step": 18381 }, { "epoch": 1.9928447528187339, "grad_norm": 0.2227293848991394, "learning_rate": 3.369360220513565e-05, "loss": 0.0116, "step": 18382 }, { "epoch": 1.9929531656548134, "grad_norm": 0.35787779092788696, "learning_rate": 3.3689975337298714e-05, "loss": 0.0318, "step": 18383 }, { "epoch": 1.9930615784908934, "grad_norm": 0.29300543665885925, "learning_rate": 3.368634846946178e-05, "loss": 0.0214, "step": 18384 }, { "epoch": 1.9931699913269731, "grad_norm": 0.2728540003299713, "learning_rate": 3.368272160162484e-05, "loss": 0.018, "step": 18385 }, { "epoch": 1.993278404163053, "grad_norm": 0.08009552955627441, "learning_rate": 3.36790947337879e-05, "loss": 0.0044, "step": 18386 }, { "epoch": 1.9933868169991327, "grad_norm": 0.7692989110946655, "learning_rate": 3.367546786595096e-05, "loss": 0.041, "step": 18387 }, { "epoch": 1.9934952298352124, "grad_norm": 0.2678688168525696, "learning_rate": 3.367184099811403e-05, "loss": 0.0175, "step": 18388 }, { "epoch": 1.9936036426712924, "grad_norm": 0.14375387132167816, "learning_rate": 3.366821413027709e-05, "loss": 0.0103, "step": 18389 }, { "epoch": 1.993712055507372, "grad_norm": 0.7989799380302429, "learning_rate": 3.3664587262440156e-05, "loss": 0.0378, "step": 18390 }, { "epoch": 1.993820468343452, "grad_norm": 0.17577645182609558, "learning_rate": 3.3660960394603227e-05, "loss": 0.023, "step": 18391 }, { "epoch": 1.9939288811795317, "grad_norm": 0.12504169344902039, "learning_rate": 3.365733352676629e-05, "loss": 0.0069, "step": 18392 }, { "epoch": 1.9940372940156115, "grad_norm": 0.15783433616161346, "learning_rate": 3.3653706658929355e-05, "loss": 0.0043, "step": 18393 }, { "epoch": 1.9941457068516912, "grad_norm": 0.043443404138088226, "learning_rate": 3.365007979109241e-05, "loss": 0.0022, "step": 18394 }, { "epoch": 1.994254119687771, "grad_norm": 0.15227369964122772, "learning_rate": 3.3646452923255476e-05, "loss": 0.014, "step": 18395 }, { "epoch": 1.994362532523851, "grad_norm": 0.5006371140480042, "learning_rate": 3.364282605541854e-05, "loss": 0.0289, "step": 18396 }, { "epoch": 1.9944709453599305, "grad_norm": 0.3244014084339142, "learning_rate": 3.3639199187581604e-05, "loss": 0.019, "step": 18397 }, { "epoch": 1.9945793581960105, "grad_norm": 0.24677477777004242, "learning_rate": 3.363557231974467e-05, "loss": 0.0091, "step": 18398 }, { "epoch": 1.99468777103209, "grad_norm": 0.155538871884346, "learning_rate": 3.363194545190773e-05, "loss": 0.0115, "step": 18399 }, { "epoch": 1.99479618386817, "grad_norm": 0.5160179734230042, "learning_rate": 3.3628318584070804e-05, "loss": 0.0428, "step": 18400 }, { "epoch": 1.9949045967042498, "grad_norm": 0.19616834819316864, "learning_rate": 3.362469171623386e-05, "loss": 0.0148, "step": 18401 }, { "epoch": 1.9950130095403296, "grad_norm": 0.276792049407959, "learning_rate": 3.3621064848396925e-05, "loss": 0.0106, "step": 18402 }, { "epoch": 1.9951214223764093, "grad_norm": 0.3899284601211548, "learning_rate": 3.361743798055999e-05, "loss": 0.0203, "step": 18403 }, { "epoch": 1.995229835212489, "grad_norm": 0.18961592018604279, "learning_rate": 3.361381111272305e-05, "loss": 0.0141, "step": 18404 }, { "epoch": 1.995338248048569, "grad_norm": 0.4656829237937927, "learning_rate": 3.361018424488612e-05, "loss": 0.0429, "step": 18405 }, { "epoch": 1.9954466608846486, "grad_norm": 0.427015483379364, "learning_rate": 3.360655737704918e-05, "loss": 0.0528, "step": 18406 }, { "epoch": 1.9955550737207286, "grad_norm": 0.09933114796876907, "learning_rate": 3.3602930509212245e-05, "loss": 0.008, "step": 18407 }, { "epoch": 1.9956634865568084, "grad_norm": 0.42188912630081177, "learning_rate": 3.359930364137531e-05, "loss": 0.0536, "step": 18408 }, { "epoch": 1.9957718993928881, "grad_norm": 0.18131816387176514, "learning_rate": 3.3595676773538374e-05, "loss": 0.0086, "step": 18409 }, { "epoch": 1.995880312228968, "grad_norm": 0.21262861788272858, "learning_rate": 3.359204990570144e-05, "loss": 0.0137, "step": 18410 }, { "epoch": 1.9959887250650477, "grad_norm": 0.294856458902359, "learning_rate": 3.35884230378645e-05, "loss": 0.0249, "step": 18411 }, { "epoch": 1.9960971379011276, "grad_norm": 0.5025202035903931, "learning_rate": 3.3584796170027566e-05, "loss": 0.0245, "step": 18412 }, { "epoch": 1.9962055507372072, "grad_norm": 0.18760591745376587, "learning_rate": 3.358116930219063e-05, "loss": 0.0185, "step": 18413 }, { "epoch": 1.9963139635732872, "grad_norm": 0.4597384035587311, "learning_rate": 3.3577542434353694e-05, "loss": 0.0229, "step": 18414 }, { "epoch": 1.9964223764093667, "grad_norm": 0.8509889245033264, "learning_rate": 3.357391556651676e-05, "loss": 0.0651, "step": 18415 }, { "epoch": 1.9965307892454467, "grad_norm": 0.07114943116903305, "learning_rate": 3.357028869867982e-05, "loss": 0.004, "step": 18416 }, { "epoch": 1.9966392020815265, "grad_norm": 0.5029305815696716, "learning_rate": 3.3566661830842887e-05, "loss": 0.0877, "step": 18417 }, { "epoch": 1.9967476149176062, "grad_norm": 0.4415932893753052, "learning_rate": 3.356303496300595e-05, "loss": 0.0185, "step": 18418 }, { "epoch": 1.9968560277536862, "grad_norm": 0.3147784471511841, "learning_rate": 3.3559408095169015e-05, "loss": 0.0304, "step": 18419 }, { "epoch": 1.9969644405897657, "grad_norm": 0.17444907128810883, "learning_rate": 3.355578122733208e-05, "loss": 0.0159, "step": 18420 }, { "epoch": 1.9970728534258457, "grad_norm": 0.22488513588905334, "learning_rate": 3.355215435949514e-05, "loss": 0.025, "step": 18421 }, { "epoch": 1.9971812662619253, "grad_norm": 0.11184507608413696, "learning_rate": 3.354852749165821e-05, "loss": 0.0107, "step": 18422 }, { "epoch": 1.9972896790980053, "grad_norm": 0.3164368271827698, "learning_rate": 3.354490062382127e-05, "loss": 0.0238, "step": 18423 }, { "epoch": 1.997398091934085, "grad_norm": 0.17712539434432983, "learning_rate": 3.354127375598433e-05, "loss": 0.013, "step": 18424 }, { "epoch": 1.9975065047701648, "grad_norm": 0.24682408571243286, "learning_rate": 3.353764688814739e-05, "loss": 0.0369, "step": 18425 }, { "epoch": 1.9976149176062445, "grad_norm": 0.25939488410949707, "learning_rate": 3.3534020020310463e-05, "loss": 0.0112, "step": 18426 }, { "epoch": 1.9977233304423243, "grad_norm": 0.1352865993976593, "learning_rate": 3.353039315247353e-05, "loss": 0.0065, "step": 18427 }, { "epoch": 1.9978317432784043, "grad_norm": 0.22841273248195648, "learning_rate": 3.352676628463659e-05, "loss": 0.0185, "step": 18428 }, { "epoch": 1.9979401561144838, "grad_norm": 0.12163017690181732, "learning_rate": 3.3523139416799656e-05, "loss": 0.0047, "step": 18429 }, { "epoch": 1.9980485689505638, "grad_norm": 0.5114683508872986, "learning_rate": 3.351951254896272e-05, "loss": 0.0409, "step": 18430 }, { "epoch": 1.9981569817866436, "grad_norm": 0.3170413374900818, "learning_rate": 3.351588568112578e-05, "loss": 0.0354, "step": 18431 }, { "epoch": 1.9982653946227233, "grad_norm": 0.3766387104988098, "learning_rate": 3.351225881328884e-05, "loss": 0.0231, "step": 18432 }, { "epoch": 1.998373807458803, "grad_norm": 0.18670649826526642, "learning_rate": 3.3508631945451905e-05, "loss": 0.0311, "step": 18433 }, { "epoch": 1.9984822202948829, "grad_norm": 0.476663738489151, "learning_rate": 3.3505005077614976e-05, "loss": 0.021, "step": 18434 }, { "epoch": 1.9985906331309629, "grad_norm": 0.21407663822174072, "learning_rate": 3.350137820977804e-05, "loss": 0.0122, "step": 18435 }, { "epoch": 1.9986990459670424, "grad_norm": 0.40308472514152527, "learning_rate": 3.3497751341941105e-05, "loss": 0.0357, "step": 18436 }, { "epoch": 1.9988074588031224, "grad_norm": 0.06229107081890106, "learning_rate": 3.349412447410417e-05, "loss": 0.005, "step": 18437 }, { "epoch": 1.998915871639202, "grad_norm": 0.12850557267665863, "learning_rate": 3.3490497606267226e-05, "loss": 0.0071, "step": 18438 }, { "epoch": 1.999024284475282, "grad_norm": 0.2401999533176422, "learning_rate": 3.348687073843029e-05, "loss": 0.01, "step": 18439 }, { "epoch": 1.9991326973113617, "grad_norm": 0.08507757633924484, "learning_rate": 3.3483243870593354e-05, "loss": 0.0064, "step": 18440 }, { "epoch": 1.9992411101474414, "grad_norm": 0.25249001383781433, "learning_rate": 3.347961700275642e-05, "loss": 0.0369, "step": 18441 }, { "epoch": 1.9993495229835212, "grad_norm": 0.16802050173282623, "learning_rate": 3.347599013491948e-05, "loss": 0.0265, "step": 18442 }, { "epoch": 1.999457935819601, "grad_norm": 0.2922319769859314, "learning_rate": 3.347236326708255e-05, "loss": 0.0227, "step": 18443 }, { "epoch": 1.999566348655681, "grad_norm": 0.5181487798690796, "learning_rate": 3.346873639924562e-05, "loss": 0.0279, "step": 18444 }, { "epoch": 1.9996747614917605, "grad_norm": 0.4591347873210907, "learning_rate": 3.346510953140868e-05, "loss": 0.036, "step": 18445 }, { "epoch": 1.9997831743278405, "grad_norm": 0.19345144927501678, "learning_rate": 3.346148266357174e-05, "loss": 0.0191, "step": 18446 }, { "epoch": 1.9998915871639202, "grad_norm": 0.2389986217021942, "learning_rate": 3.34578557957348e-05, "loss": 0.022, "step": 18447 }, { "epoch": 2.0, "grad_norm": 0.2766694128513336, "learning_rate": 3.345422892789787e-05, "loss": 0.0304, "step": 18448 }, { "epoch": 2.00010841283608, "grad_norm": 0.06038445234298706, "learning_rate": 3.345060206006093e-05, "loss": 0.0039, "step": 18449 }, { "epoch": 2.0002168256721595, "grad_norm": 0.17353543639183044, "learning_rate": 3.3446975192223995e-05, "loss": 0.0198, "step": 18450 }, { "epoch": 2.0003252385082395, "grad_norm": 0.08195535093545914, "learning_rate": 3.344334832438706e-05, "loss": 0.013, "step": 18451 }, { "epoch": 2.000433651344319, "grad_norm": 0.059756599366664886, "learning_rate": 3.343972145655013e-05, "loss": 0.0039, "step": 18452 }, { "epoch": 2.000542064180399, "grad_norm": 0.23425720632076263, "learning_rate": 3.343609458871319e-05, "loss": 0.0235, "step": 18453 }, { "epoch": 2.0006504770164786, "grad_norm": 0.17236502468585968, "learning_rate": 3.343246772087625e-05, "loss": 0.0092, "step": 18454 }, { "epoch": 2.0007588898525586, "grad_norm": 0.0401528999209404, "learning_rate": 3.3428840853039316e-05, "loss": 0.0031, "step": 18455 }, { "epoch": 2.0008673026886385, "grad_norm": 0.09281701594591141, "learning_rate": 3.342521398520238e-05, "loss": 0.006, "step": 18456 }, { "epoch": 2.000975715524718, "grad_norm": 0.09738173335790634, "learning_rate": 3.3421587117365444e-05, "loss": 0.0047, "step": 18457 }, { "epoch": 2.001084128360798, "grad_norm": 0.16477292776107788, "learning_rate": 3.341796024952851e-05, "loss": 0.0087, "step": 18458 }, { "epoch": 2.0011925411968776, "grad_norm": 0.18617013096809387, "learning_rate": 3.341433338169157e-05, "loss": 0.0068, "step": 18459 }, { "epoch": 2.0013009540329576, "grad_norm": 0.38992127776145935, "learning_rate": 3.3410706513854636e-05, "loss": 0.0264, "step": 18460 }, { "epoch": 2.001409366869037, "grad_norm": 0.07511984556913376, "learning_rate": 3.34070796460177e-05, "loss": 0.0053, "step": 18461 }, { "epoch": 2.001517779705117, "grad_norm": 0.14939677715301514, "learning_rate": 3.3403452778180764e-05, "loss": 0.0078, "step": 18462 }, { "epoch": 2.0016261925411967, "grad_norm": 0.7285293340682983, "learning_rate": 3.339982591034383e-05, "loss": 0.0432, "step": 18463 }, { "epoch": 2.0017346053772767, "grad_norm": 0.1949400007724762, "learning_rate": 3.339619904250689e-05, "loss": 0.0262, "step": 18464 }, { "epoch": 2.0018430182133566, "grad_norm": 0.0625271275639534, "learning_rate": 3.339257217466996e-05, "loss": 0.0024, "step": 18465 }, { "epoch": 2.001951431049436, "grad_norm": 0.0808129608631134, "learning_rate": 3.338894530683302e-05, "loss": 0.0026, "step": 18466 }, { "epoch": 2.002059843885516, "grad_norm": 0.22016479074954987, "learning_rate": 3.3385318438996085e-05, "loss": 0.0079, "step": 18467 }, { "epoch": 2.0021682567215957, "grad_norm": 0.24599739909172058, "learning_rate": 3.338169157115915e-05, "loss": 0.0093, "step": 18468 }, { "epoch": 2.0022766695576757, "grad_norm": 0.0440729483962059, "learning_rate": 3.337806470332221e-05, "loss": 0.0012, "step": 18469 }, { "epoch": 2.0023850823937552, "grad_norm": 0.08593850582838058, "learning_rate": 3.337443783548528e-05, "loss": 0.0042, "step": 18470 }, { "epoch": 2.002493495229835, "grad_norm": 0.29534128308296204, "learning_rate": 3.337081096764834e-05, "loss": 0.0157, "step": 18471 }, { "epoch": 2.002601908065915, "grad_norm": 0.25365155935287476, "learning_rate": 3.3367184099811406e-05, "loss": 0.0098, "step": 18472 }, { "epoch": 2.0027103209019947, "grad_norm": 0.15606002509593964, "learning_rate": 3.336355723197447e-05, "loss": 0.0034, "step": 18473 }, { "epoch": 2.0028187337380747, "grad_norm": 0.10940512269735336, "learning_rate": 3.3359930364137534e-05, "loss": 0.0084, "step": 18474 }, { "epoch": 2.0029271465741543, "grad_norm": 0.027254007756710052, "learning_rate": 3.33563034963006e-05, "loss": 0.0022, "step": 18475 }, { "epoch": 2.0030355594102343, "grad_norm": 0.09253470599651337, "learning_rate": 3.3352676628463655e-05, "loss": 0.003, "step": 18476 }, { "epoch": 2.003143972246314, "grad_norm": 0.11543616652488708, "learning_rate": 3.334904976062672e-05, "loss": 0.0055, "step": 18477 }, { "epoch": 2.003252385082394, "grad_norm": 0.13701969385147095, "learning_rate": 3.334542289278979e-05, "loss": 0.0046, "step": 18478 }, { "epoch": 2.0033607979184738, "grad_norm": 0.3501553237438202, "learning_rate": 3.3341796024952854e-05, "loss": 0.0271, "step": 18479 }, { "epoch": 2.0034692107545533, "grad_norm": 0.17141512036323547, "learning_rate": 3.333816915711592e-05, "loss": 0.0149, "step": 18480 }, { "epoch": 2.0035776235906333, "grad_norm": 0.12259020656347275, "learning_rate": 3.333454228927898e-05, "loss": 0.0083, "step": 18481 }, { "epoch": 2.003686036426713, "grad_norm": 0.10773944109678268, "learning_rate": 3.3330915421442047e-05, "loss": 0.0102, "step": 18482 }, { "epoch": 2.003794449262793, "grad_norm": 0.08726733922958374, "learning_rate": 3.3327288553605104e-05, "loss": 0.005, "step": 18483 }, { "epoch": 2.0039028620988724, "grad_norm": 0.10530463606119156, "learning_rate": 3.332366168576817e-05, "loss": 0.0072, "step": 18484 }, { "epoch": 2.0040112749349523, "grad_norm": 0.017119551077485085, "learning_rate": 3.332003481793123e-05, "loss": 0.0008, "step": 18485 }, { "epoch": 2.004119687771032, "grad_norm": 0.12933757901191711, "learning_rate": 3.33164079500943e-05, "loss": 0.0065, "step": 18486 }, { "epoch": 2.004228100607112, "grad_norm": 0.18385405838489532, "learning_rate": 3.331278108225737e-05, "loss": 0.0132, "step": 18487 }, { "epoch": 2.004336513443192, "grad_norm": 0.32405054569244385, "learning_rate": 3.330915421442043e-05, "loss": 0.017, "step": 18488 }, { "epoch": 2.0044449262792714, "grad_norm": 0.04886303097009659, "learning_rate": 3.3305527346583495e-05, "loss": 0.0016, "step": 18489 }, { "epoch": 2.0045533391153514, "grad_norm": 0.0646393820643425, "learning_rate": 3.330190047874656e-05, "loss": 0.0029, "step": 18490 }, { "epoch": 2.004661751951431, "grad_norm": 0.44613853096961975, "learning_rate": 3.329827361090962e-05, "loss": 0.0225, "step": 18491 }, { "epoch": 2.004770164787511, "grad_norm": 0.08106285333633423, "learning_rate": 3.329464674307268e-05, "loss": 0.0049, "step": 18492 }, { "epoch": 2.0048785776235905, "grad_norm": 0.47650447487831116, "learning_rate": 3.3291019875235745e-05, "loss": 0.0207, "step": 18493 }, { "epoch": 2.0049869904596704, "grad_norm": 0.6319919228553772, "learning_rate": 3.328739300739881e-05, "loss": 0.0407, "step": 18494 }, { "epoch": 2.0050954032957504, "grad_norm": 0.006692942697554827, "learning_rate": 3.328376613956188e-05, "loss": 0.0006, "step": 18495 }, { "epoch": 2.00520381613183, "grad_norm": 0.19221697747707367, "learning_rate": 3.3280139271724944e-05, "loss": 0.0125, "step": 18496 }, { "epoch": 2.00531222896791, "grad_norm": 0.18173439800739288, "learning_rate": 3.327651240388801e-05, "loss": 0.0112, "step": 18497 }, { "epoch": 2.0054206418039895, "grad_norm": 0.08233343064785004, "learning_rate": 3.3272885536051065e-05, "loss": 0.0032, "step": 18498 }, { "epoch": 2.0055290546400695, "grad_norm": 0.7199202179908752, "learning_rate": 3.326925866821413e-05, "loss": 0.0265, "step": 18499 }, { "epoch": 2.005637467476149, "grad_norm": 0.46729832887649536, "learning_rate": 3.3265631800377194e-05, "loss": 0.0147, "step": 18500 }, { "epoch": 2.005745880312229, "grad_norm": 0.5155790448188782, "learning_rate": 3.326200493254026e-05, "loss": 0.0231, "step": 18501 }, { "epoch": 2.0058542931483085, "grad_norm": 0.10330893099308014, "learning_rate": 3.325837806470332e-05, "loss": 0.0096, "step": 18502 }, { "epoch": 2.0059627059843885, "grad_norm": 0.04821530729532242, "learning_rate": 3.325475119686639e-05, "loss": 0.0026, "step": 18503 }, { "epoch": 2.0060711188204685, "grad_norm": 0.04945116862654686, "learning_rate": 3.325112432902946e-05, "loss": 0.0029, "step": 18504 }, { "epoch": 2.006179531656548, "grad_norm": 0.243442103266716, "learning_rate": 3.3247497461192514e-05, "loss": 0.033, "step": 18505 }, { "epoch": 2.006287944492628, "grad_norm": 0.03779000788927078, "learning_rate": 3.324387059335558e-05, "loss": 0.001, "step": 18506 }, { "epoch": 2.0063963573287076, "grad_norm": 0.06478804349899292, "learning_rate": 3.324024372551864e-05, "loss": 0.0019, "step": 18507 }, { "epoch": 2.0065047701647876, "grad_norm": 0.07118672132492065, "learning_rate": 3.3236616857681707e-05, "loss": 0.0046, "step": 18508 }, { "epoch": 2.006613183000867, "grad_norm": 0.14162328839302063, "learning_rate": 3.323298998984477e-05, "loss": 0.0061, "step": 18509 }, { "epoch": 2.006721595836947, "grad_norm": 0.472299724817276, "learning_rate": 3.3229363122007835e-05, "loss": 0.0319, "step": 18510 }, { "epoch": 2.006830008673027, "grad_norm": 0.07191315293312073, "learning_rate": 3.32257362541709e-05, "loss": 0.0023, "step": 18511 }, { "epoch": 2.0069384215091066, "grad_norm": 0.1023993119597435, "learning_rate": 3.322210938633396e-05, "loss": 0.0028, "step": 18512 }, { "epoch": 2.0070468343451866, "grad_norm": 0.1765691637992859, "learning_rate": 3.321848251849703e-05, "loss": 0.0057, "step": 18513 }, { "epoch": 2.007155247181266, "grad_norm": 0.05367935448884964, "learning_rate": 3.321485565066009e-05, "loss": 0.0017, "step": 18514 }, { "epoch": 2.007263660017346, "grad_norm": 0.23666909337043762, "learning_rate": 3.3211228782823155e-05, "loss": 0.0096, "step": 18515 }, { "epoch": 2.0073720728534257, "grad_norm": 0.12926724553108215, "learning_rate": 3.320760191498622e-05, "loss": 0.0154, "step": 18516 }, { "epoch": 2.0074804856895057, "grad_norm": 0.3814084529876709, "learning_rate": 3.3203975047149283e-05, "loss": 0.0502, "step": 18517 }, { "epoch": 2.0075888985255856, "grad_norm": 0.04914742335677147, "learning_rate": 3.320034817931235e-05, "loss": 0.0038, "step": 18518 }, { "epoch": 2.007697311361665, "grad_norm": 0.05151907354593277, "learning_rate": 3.319672131147541e-05, "loss": 0.0032, "step": 18519 }, { "epoch": 2.007805724197745, "grad_norm": 0.4060712456703186, "learning_rate": 3.3193094443638476e-05, "loss": 0.0135, "step": 18520 }, { "epoch": 2.0079141370338247, "grad_norm": 0.3899535536766052, "learning_rate": 3.318946757580154e-05, "loss": 0.0251, "step": 18521 }, { "epoch": 2.0080225498699047, "grad_norm": 0.29567232728004456, "learning_rate": 3.3185840707964604e-05, "loss": 0.0063, "step": 18522 }, { "epoch": 2.0081309627059842, "grad_norm": 0.1698606014251709, "learning_rate": 3.318221384012767e-05, "loss": 0.0152, "step": 18523 }, { "epoch": 2.008239375542064, "grad_norm": 0.02139529399573803, "learning_rate": 3.317858697229073e-05, "loss": 0.0009, "step": 18524 }, { "epoch": 2.0083477883781438, "grad_norm": 0.3058786392211914, "learning_rate": 3.3174960104453796e-05, "loss": 0.0199, "step": 18525 }, { "epoch": 2.0084562012142237, "grad_norm": 0.07060880959033966, "learning_rate": 3.317133323661686e-05, "loss": 0.0029, "step": 18526 }, { "epoch": 2.0085646140503037, "grad_norm": 0.216874897480011, "learning_rate": 3.3167706368779925e-05, "loss": 0.0194, "step": 18527 }, { "epoch": 2.0086730268863833, "grad_norm": 0.09671475738286972, "learning_rate": 3.316407950094298e-05, "loss": 0.0125, "step": 18528 }, { "epoch": 2.0087814397224633, "grad_norm": 0.11448747664690018, "learning_rate": 3.316045263310605e-05, "loss": 0.0065, "step": 18529 }, { "epoch": 2.008889852558543, "grad_norm": 0.06121969223022461, "learning_rate": 3.315682576526912e-05, "loss": 0.0027, "step": 18530 }, { "epoch": 2.008998265394623, "grad_norm": 0.17535719275474548, "learning_rate": 3.315319889743218e-05, "loss": 0.0046, "step": 18531 }, { "epoch": 2.0091066782307023, "grad_norm": 0.21330459415912628, "learning_rate": 3.3149572029595245e-05, "loss": 0.0117, "step": 18532 }, { "epoch": 2.0092150910667823, "grad_norm": 0.3316175639629364, "learning_rate": 3.314594516175831e-05, "loss": 0.0388, "step": 18533 }, { "epoch": 2.0093235039028623, "grad_norm": 0.18016469478607178, "learning_rate": 3.314231829392137e-05, "loss": 0.0143, "step": 18534 }, { "epoch": 2.009431916738942, "grad_norm": 0.006592886988073587, "learning_rate": 3.313869142608443e-05, "loss": 0.0004, "step": 18535 }, { "epoch": 2.009540329575022, "grad_norm": 0.1335146427154541, "learning_rate": 3.3135064558247495e-05, "loss": 0.0117, "step": 18536 }, { "epoch": 2.0096487424111014, "grad_norm": 0.04012858867645264, "learning_rate": 3.313143769041056e-05, "loss": 0.0018, "step": 18537 }, { "epoch": 2.0097571552471813, "grad_norm": 0.17068862915039062, "learning_rate": 3.312781082257363e-05, "loss": 0.0132, "step": 18538 }, { "epoch": 2.009865568083261, "grad_norm": 0.11992821097373962, "learning_rate": 3.3124183954736694e-05, "loss": 0.0069, "step": 18539 }, { "epoch": 2.009973980919341, "grad_norm": 0.47702786326408386, "learning_rate": 3.312055708689976e-05, "loss": 0.0207, "step": 18540 }, { "epoch": 2.010082393755421, "grad_norm": 0.23038671910762787, "learning_rate": 3.311693021906282e-05, "loss": 0.02, "step": 18541 }, { "epoch": 2.0101908065915004, "grad_norm": 0.16509807109832764, "learning_rate": 3.3113303351225886e-05, "loss": 0.0198, "step": 18542 }, { "epoch": 2.0102992194275804, "grad_norm": 0.12668615579605103, "learning_rate": 3.3109676483388943e-05, "loss": 0.0091, "step": 18543 }, { "epoch": 2.01040763226366, "grad_norm": 0.08947659283876419, "learning_rate": 3.310604961555201e-05, "loss": 0.007, "step": 18544 }, { "epoch": 2.01051604509974, "grad_norm": 0.3688923418521881, "learning_rate": 3.310242274771507e-05, "loss": 0.0309, "step": 18545 }, { "epoch": 2.0106244579358195, "grad_norm": 0.3124043345451355, "learning_rate": 3.3098795879878136e-05, "loss": 0.0077, "step": 18546 }, { "epoch": 2.0107328707718994, "grad_norm": 0.0263985488563776, "learning_rate": 3.309516901204121e-05, "loss": 0.0013, "step": 18547 }, { "epoch": 2.010841283607979, "grad_norm": 0.19759376347064972, "learning_rate": 3.309154214420427e-05, "loss": 0.003, "step": 18548 }, { "epoch": 2.010949696444059, "grad_norm": 0.40341827273368835, "learning_rate": 3.3087915276367335e-05, "loss": 0.0482, "step": 18549 }, { "epoch": 2.011058109280139, "grad_norm": 0.04702698439359665, "learning_rate": 3.308428840853039e-05, "loss": 0.0015, "step": 18550 }, { "epoch": 2.0111665221162185, "grad_norm": 0.3241540491580963, "learning_rate": 3.3080661540693456e-05, "loss": 0.0084, "step": 18551 }, { "epoch": 2.0112749349522985, "grad_norm": 0.276679664850235, "learning_rate": 3.307703467285652e-05, "loss": 0.0123, "step": 18552 }, { "epoch": 2.011383347788378, "grad_norm": 0.03844515606760979, "learning_rate": 3.3073407805019585e-05, "loss": 0.0025, "step": 18553 }, { "epoch": 2.011491760624458, "grad_norm": 0.07188230007886887, "learning_rate": 3.306978093718265e-05, "loss": 0.0037, "step": 18554 }, { "epoch": 2.0116001734605375, "grad_norm": 0.3819196820259094, "learning_rate": 3.306615406934572e-05, "loss": 0.0159, "step": 18555 }, { "epoch": 2.0117085862966175, "grad_norm": 0.02416691929101944, "learning_rate": 3.3062527201508784e-05, "loss": 0.0013, "step": 18556 }, { "epoch": 2.0118169991326975, "grad_norm": 0.4618435502052307, "learning_rate": 3.305890033367184e-05, "loss": 0.0289, "step": 18557 }, { "epoch": 2.011925411968777, "grad_norm": 0.31717729568481445, "learning_rate": 3.3055273465834905e-05, "loss": 0.0065, "step": 18558 }, { "epoch": 2.012033824804857, "grad_norm": 0.213802769780159, "learning_rate": 3.305164659799797e-05, "loss": 0.0254, "step": 18559 }, { "epoch": 2.0121422376409366, "grad_norm": 0.20214219391345978, "learning_rate": 3.304801973016103e-05, "loss": 0.0115, "step": 18560 }, { "epoch": 2.0122506504770166, "grad_norm": 0.017057808116078377, "learning_rate": 3.30443928623241e-05, "loss": 0.0011, "step": 18561 }, { "epoch": 2.012359063313096, "grad_norm": 0.15848508477210999, "learning_rate": 3.304076599448716e-05, "loss": 0.0098, "step": 18562 }, { "epoch": 2.012467476149176, "grad_norm": 0.18903090059757233, "learning_rate": 3.3037139126650226e-05, "loss": 0.01, "step": 18563 }, { "epoch": 2.0125758889852556, "grad_norm": 0.34729933738708496, "learning_rate": 3.303351225881329e-05, "loss": 0.02, "step": 18564 }, { "epoch": 2.0126843018213356, "grad_norm": 0.2743389308452606, "learning_rate": 3.3029885390976354e-05, "loss": 0.049, "step": 18565 }, { "epoch": 2.0127927146574156, "grad_norm": 0.5085231065750122, "learning_rate": 3.302625852313942e-05, "loss": 0.0416, "step": 18566 }, { "epoch": 2.012901127493495, "grad_norm": 0.06291292607784271, "learning_rate": 3.302263165530248e-05, "loss": 0.0068, "step": 18567 }, { "epoch": 2.013009540329575, "grad_norm": 0.11316132545471191, "learning_rate": 3.3019004787465546e-05, "loss": 0.003, "step": 18568 }, { "epoch": 2.0131179531656547, "grad_norm": 0.021794695407152176, "learning_rate": 3.301537791962861e-05, "loss": 0.0009, "step": 18569 }, { "epoch": 2.0132263660017347, "grad_norm": 0.04838208854198456, "learning_rate": 3.3011751051791674e-05, "loss": 0.0021, "step": 18570 }, { "epoch": 2.013334778837814, "grad_norm": 0.6614845991134644, "learning_rate": 3.300812418395474e-05, "loss": 0.0219, "step": 18571 }, { "epoch": 2.013443191673894, "grad_norm": 0.31261178851127625, "learning_rate": 3.30044973161178e-05, "loss": 0.0181, "step": 18572 }, { "epoch": 2.013551604509974, "grad_norm": 0.2501368522644043, "learning_rate": 3.3000870448280867e-05, "loss": 0.0059, "step": 18573 }, { "epoch": 2.0136600173460537, "grad_norm": 0.4228506088256836, "learning_rate": 3.299724358044393e-05, "loss": 0.0192, "step": 18574 }, { "epoch": 2.0137684301821337, "grad_norm": 0.08817946165800095, "learning_rate": 3.2993616712606995e-05, "loss": 0.0021, "step": 18575 }, { "epoch": 2.0138768430182132, "grad_norm": 0.08666329830884933, "learning_rate": 3.298998984477006e-05, "loss": 0.0041, "step": 18576 }, { "epoch": 2.013985255854293, "grad_norm": 0.3271419405937195, "learning_rate": 3.298636297693312e-05, "loss": 0.0201, "step": 18577 }, { "epoch": 2.0140936686903728, "grad_norm": 0.3446890115737915, "learning_rate": 3.298273610909619e-05, "loss": 0.0226, "step": 18578 }, { "epoch": 2.0142020815264527, "grad_norm": 0.12403976917266846, "learning_rate": 3.297910924125925e-05, "loss": 0.0051, "step": 18579 }, { "epoch": 2.0143104943625327, "grad_norm": 0.13691194355487823, "learning_rate": 3.297548237342231e-05, "loss": 0.0102, "step": 18580 }, { "epoch": 2.0144189071986123, "grad_norm": 0.060935866087675095, "learning_rate": 3.297185550558538e-05, "loss": 0.0015, "step": 18581 }, { "epoch": 2.0145273200346923, "grad_norm": 0.034754812717437744, "learning_rate": 3.2968228637748444e-05, "loss": 0.0018, "step": 18582 }, { "epoch": 2.014635732870772, "grad_norm": 0.020970789715647697, "learning_rate": 3.296460176991151e-05, "loss": 0.0008, "step": 18583 }, { "epoch": 2.014744145706852, "grad_norm": 0.18883870542049408, "learning_rate": 3.296097490207457e-05, "loss": 0.0093, "step": 18584 }, { "epoch": 2.0148525585429313, "grad_norm": 0.2712763845920563, "learning_rate": 3.2957348034237636e-05, "loss": 0.0199, "step": 18585 }, { "epoch": 2.0149609713790113, "grad_norm": 0.25068777799606323, "learning_rate": 3.29537211664007e-05, "loss": 0.0059, "step": 18586 }, { "epoch": 2.015069384215091, "grad_norm": 0.7053316831588745, "learning_rate": 3.295009429856376e-05, "loss": 0.0449, "step": 18587 }, { "epoch": 2.015177797051171, "grad_norm": 0.6180164217948914, "learning_rate": 3.294646743072682e-05, "loss": 0.0528, "step": 18588 }, { "epoch": 2.015286209887251, "grad_norm": 0.2689445912837982, "learning_rate": 3.2942840562889886e-05, "loss": 0.0391, "step": 18589 }, { "epoch": 2.0153946227233304, "grad_norm": 0.04642666503787041, "learning_rate": 3.2939213695052956e-05, "loss": 0.0018, "step": 18590 }, { "epoch": 2.0155030355594103, "grad_norm": 0.23855218291282654, "learning_rate": 3.293558682721602e-05, "loss": 0.0032, "step": 18591 }, { "epoch": 2.01561144839549, "grad_norm": 0.5267186760902405, "learning_rate": 3.2931959959379085e-05, "loss": 0.0216, "step": 18592 }, { "epoch": 2.01571986123157, "grad_norm": 0.08791270852088928, "learning_rate": 3.292833309154215e-05, "loss": 0.0078, "step": 18593 }, { "epoch": 2.0158282740676494, "grad_norm": 0.18111751973628998, "learning_rate": 3.292470622370521e-05, "loss": 0.015, "step": 18594 }, { "epoch": 2.0159366869037294, "grad_norm": 0.1725200116634369, "learning_rate": 3.292107935586827e-05, "loss": 0.0039, "step": 18595 }, { "epoch": 2.0160450997398094, "grad_norm": 0.09313630312681198, "learning_rate": 3.2917452488031334e-05, "loss": 0.0072, "step": 18596 }, { "epoch": 2.016153512575889, "grad_norm": 0.2025742381811142, "learning_rate": 3.29138256201944e-05, "loss": 0.0056, "step": 18597 }, { "epoch": 2.016261925411969, "grad_norm": 0.07328173518180847, "learning_rate": 3.291019875235747e-05, "loss": 0.0022, "step": 18598 }, { "epoch": 2.0163703382480485, "grad_norm": 0.2058146446943283, "learning_rate": 3.290657188452053e-05, "loss": 0.0153, "step": 18599 }, { "epoch": 2.0164787510841284, "grad_norm": 0.1483505815267563, "learning_rate": 3.29029450166836e-05, "loss": 0.0071, "step": 18600 }, { "epoch": 2.016587163920208, "grad_norm": 0.12780599296092987, "learning_rate": 3.289931814884666e-05, "loss": 0.0069, "step": 18601 }, { "epoch": 2.016695576756288, "grad_norm": 0.2271605134010315, "learning_rate": 3.289569128100972e-05, "loss": 0.0328, "step": 18602 }, { "epoch": 2.016803989592368, "grad_norm": 0.1165897473692894, "learning_rate": 3.289206441317278e-05, "loss": 0.0054, "step": 18603 }, { "epoch": 2.0169124024284475, "grad_norm": 0.10341901332139969, "learning_rate": 3.288843754533585e-05, "loss": 0.0057, "step": 18604 }, { "epoch": 2.0170208152645275, "grad_norm": 0.03723657503724098, "learning_rate": 3.288481067749891e-05, "loss": 0.0014, "step": 18605 }, { "epoch": 2.017129228100607, "grad_norm": 0.8754670023918152, "learning_rate": 3.2881183809661975e-05, "loss": 0.0136, "step": 18606 }, { "epoch": 2.017237640936687, "grad_norm": 0.2915992736816406, "learning_rate": 3.2877556941825046e-05, "loss": 0.0122, "step": 18607 }, { "epoch": 2.0173460537727665, "grad_norm": 0.06000570207834244, "learning_rate": 3.287393007398811e-05, "loss": 0.0014, "step": 18608 }, { "epoch": 2.0174544666088465, "grad_norm": 0.294069766998291, "learning_rate": 3.287030320615117e-05, "loss": 0.0181, "step": 18609 }, { "epoch": 2.017562879444926, "grad_norm": 0.08518525213003159, "learning_rate": 3.286667633831423e-05, "loss": 0.0047, "step": 18610 }, { "epoch": 2.017671292281006, "grad_norm": 0.09337987005710602, "learning_rate": 3.2863049470477296e-05, "loss": 0.0041, "step": 18611 }, { "epoch": 2.017779705117086, "grad_norm": 0.13478122651576996, "learning_rate": 3.285942260264036e-05, "loss": 0.0056, "step": 18612 }, { "epoch": 2.0178881179531656, "grad_norm": 0.5060659050941467, "learning_rate": 3.2855795734803424e-05, "loss": 0.0432, "step": 18613 }, { "epoch": 2.0179965307892456, "grad_norm": 0.10256342589855194, "learning_rate": 3.285216886696649e-05, "loss": 0.0061, "step": 18614 }, { "epoch": 2.018104943625325, "grad_norm": 0.2348618358373642, "learning_rate": 3.284854199912955e-05, "loss": 0.0185, "step": 18615 }, { "epoch": 2.018213356461405, "grad_norm": 0.08016673475503922, "learning_rate": 3.2844915131292616e-05, "loss": 0.0056, "step": 18616 }, { "epoch": 2.0183217692974846, "grad_norm": 0.10888997465372086, "learning_rate": 3.284128826345568e-05, "loss": 0.0063, "step": 18617 }, { "epoch": 2.0184301821335646, "grad_norm": 0.24879798293113708, "learning_rate": 3.2837661395618745e-05, "loss": 0.0172, "step": 18618 }, { "epoch": 2.0185385949696446, "grad_norm": 0.17727933824062347, "learning_rate": 3.283403452778181e-05, "loss": 0.0098, "step": 18619 }, { "epoch": 2.018647007805724, "grad_norm": 0.20482133328914642, "learning_rate": 3.283040765994487e-05, "loss": 0.0095, "step": 18620 }, { "epoch": 2.018755420641804, "grad_norm": 0.0903138667345047, "learning_rate": 3.282678079210794e-05, "loss": 0.0077, "step": 18621 }, { "epoch": 2.0188638334778837, "grad_norm": 0.3888425827026367, "learning_rate": 3.2823153924271e-05, "loss": 0.0103, "step": 18622 }, { "epoch": 2.0189722463139637, "grad_norm": 0.24493859708309174, "learning_rate": 3.2819527056434065e-05, "loss": 0.0144, "step": 18623 }, { "epoch": 2.019080659150043, "grad_norm": 0.5070556402206421, "learning_rate": 3.281590018859713e-05, "loss": 0.0302, "step": 18624 }, { "epoch": 2.019189071986123, "grad_norm": 0.3979529142379761, "learning_rate": 3.281227332076019e-05, "loss": 0.0072, "step": 18625 }, { "epoch": 2.0192974848222027, "grad_norm": 0.48458126187324524, "learning_rate": 3.280864645292326e-05, "loss": 0.0118, "step": 18626 }, { "epoch": 2.0194058976582827, "grad_norm": 0.2555078864097595, "learning_rate": 3.280501958508632e-05, "loss": 0.0054, "step": 18627 }, { "epoch": 2.0195143104943627, "grad_norm": 0.29094845056533813, "learning_rate": 3.2801392717249386e-05, "loss": 0.0193, "step": 18628 }, { "epoch": 2.0196227233304422, "grad_norm": 0.2329322248697281, "learning_rate": 3.279776584941245e-05, "loss": 0.0154, "step": 18629 }, { "epoch": 2.0197311361665222, "grad_norm": 0.12901440262794495, "learning_rate": 3.2794138981575514e-05, "loss": 0.0076, "step": 18630 }, { "epoch": 2.0198395490026018, "grad_norm": 0.12964105606079102, "learning_rate": 3.279051211373858e-05, "loss": 0.0043, "step": 18631 }, { "epoch": 2.0199479618386817, "grad_norm": 0.08082586526870728, "learning_rate": 3.2786885245901635e-05, "loss": 0.003, "step": 18632 }, { "epoch": 2.0200563746747613, "grad_norm": 0.229654923081398, "learning_rate": 3.2783258378064706e-05, "loss": 0.0053, "step": 18633 }, { "epoch": 2.0201647875108413, "grad_norm": 0.38303881883621216, "learning_rate": 3.277963151022777e-05, "loss": 0.0158, "step": 18634 }, { "epoch": 2.0202732003469213, "grad_norm": 0.02514888159930706, "learning_rate": 3.2776004642390834e-05, "loss": 0.0011, "step": 18635 }, { "epoch": 2.020381613183001, "grad_norm": 0.37518516182899475, "learning_rate": 3.27723777745539e-05, "loss": 0.0072, "step": 18636 }, { "epoch": 2.020490026019081, "grad_norm": 0.3010469377040863, "learning_rate": 3.276875090671696e-05, "loss": 0.0167, "step": 18637 }, { "epoch": 2.0205984388551603, "grad_norm": 0.18021869659423828, "learning_rate": 3.276512403888003e-05, "loss": 0.0081, "step": 18638 }, { "epoch": 2.0207068516912403, "grad_norm": 0.22811073064804077, "learning_rate": 3.2761497171043084e-05, "loss": 0.0078, "step": 18639 }, { "epoch": 2.02081526452732, "grad_norm": 0.7239392995834351, "learning_rate": 3.275787030320615e-05, "loss": 0.0272, "step": 18640 }, { "epoch": 2.0209236773634, "grad_norm": 0.06251546740531921, "learning_rate": 3.275424343536921e-05, "loss": 0.0036, "step": 18641 }, { "epoch": 2.02103209019948, "grad_norm": 0.626270592212677, "learning_rate": 3.275061656753228e-05, "loss": 0.0522, "step": 18642 }, { "epoch": 2.0211405030355594, "grad_norm": 0.07422395795583725, "learning_rate": 3.274698969969535e-05, "loss": 0.0024, "step": 18643 }, { "epoch": 2.0212489158716394, "grad_norm": 0.33009546995162964, "learning_rate": 3.274336283185841e-05, "loss": 0.0279, "step": 18644 }, { "epoch": 2.021357328707719, "grad_norm": 0.016731947660446167, "learning_rate": 3.2739735964021475e-05, "loss": 0.0011, "step": 18645 }, { "epoch": 2.021465741543799, "grad_norm": 0.07928450405597687, "learning_rate": 3.273610909618454e-05, "loss": 0.0025, "step": 18646 }, { "epoch": 2.0215741543798784, "grad_norm": 0.050199951976537704, "learning_rate": 3.27324822283476e-05, "loss": 0.0013, "step": 18647 }, { "epoch": 2.0216825672159584, "grad_norm": 0.10483904182910919, "learning_rate": 3.272885536051066e-05, "loss": 0.003, "step": 18648 }, { "epoch": 2.021790980052038, "grad_norm": 0.18434907495975494, "learning_rate": 3.2725228492673725e-05, "loss": 0.0046, "step": 18649 }, { "epoch": 2.021899392888118, "grad_norm": 0.08244209736585617, "learning_rate": 3.2721601624836796e-05, "loss": 0.0044, "step": 18650 }, { "epoch": 2.022007805724198, "grad_norm": 0.13741768896579742, "learning_rate": 3.271797475699986e-05, "loss": 0.0087, "step": 18651 }, { "epoch": 2.0221162185602775, "grad_norm": 0.04285309463739395, "learning_rate": 3.2714347889162924e-05, "loss": 0.0023, "step": 18652 }, { "epoch": 2.0222246313963574, "grad_norm": 0.019190367311239243, "learning_rate": 3.271072102132599e-05, "loss": 0.001, "step": 18653 }, { "epoch": 2.022333044232437, "grad_norm": 0.4297703802585602, "learning_rate": 3.2707094153489046e-05, "loss": 0.0193, "step": 18654 }, { "epoch": 2.022441457068517, "grad_norm": 0.20637191832065582, "learning_rate": 3.270346728565211e-05, "loss": 0.0313, "step": 18655 }, { "epoch": 2.0225498699045965, "grad_norm": 0.14333099126815796, "learning_rate": 3.2699840417815174e-05, "loss": 0.0074, "step": 18656 }, { "epoch": 2.0226582827406765, "grad_norm": 0.0009189508273266256, "learning_rate": 3.269621354997824e-05, "loss": 0.0001, "step": 18657 }, { "epoch": 2.0227666955767565, "grad_norm": 0.1667500138282776, "learning_rate": 3.26925866821413e-05, "loss": 0.0035, "step": 18658 }, { "epoch": 2.022875108412836, "grad_norm": 0.11238647997379303, "learning_rate": 3.268895981430437e-05, "loss": 0.0053, "step": 18659 }, { "epoch": 2.022983521248916, "grad_norm": 0.13091637194156647, "learning_rate": 3.268533294646744e-05, "loss": 0.01, "step": 18660 }, { "epoch": 2.0230919340849955, "grad_norm": 0.2056054025888443, "learning_rate": 3.2681706078630494e-05, "loss": 0.0062, "step": 18661 }, { "epoch": 2.0232003469210755, "grad_norm": 0.10934139043092728, "learning_rate": 3.267807921079356e-05, "loss": 0.011, "step": 18662 }, { "epoch": 2.023308759757155, "grad_norm": 0.19499753415584564, "learning_rate": 3.267445234295662e-05, "loss": 0.005, "step": 18663 }, { "epoch": 2.023417172593235, "grad_norm": 0.4000171422958374, "learning_rate": 3.267082547511969e-05, "loss": 0.0154, "step": 18664 }, { "epoch": 2.023525585429315, "grad_norm": 0.29381775856018066, "learning_rate": 3.266719860728275e-05, "loss": 0.0149, "step": 18665 }, { "epoch": 2.0236339982653946, "grad_norm": 0.668935239315033, "learning_rate": 3.2663571739445815e-05, "loss": 0.02, "step": 18666 }, { "epoch": 2.0237424111014746, "grad_norm": 0.18388192355632782, "learning_rate": 3.265994487160888e-05, "loss": 0.0064, "step": 18667 }, { "epoch": 2.023850823937554, "grad_norm": 0.13165631890296936, "learning_rate": 3.265631800377194e-05, "loss": 0.0067, "step": 18668 }, { "epoch": 2.023959236773634, "grad_norm": 0.43000638484954834, "learning_rate": 3.265269113593501e-05, "loss": 0.0178, "step": 18669 }, { "epoch": 2.0240676496097136, "grad_norm": 0.05655582994222641, "learning_rate": 3.264906426809807e-05, "loss": 0.0026, "step": 18670 }, { "epoch": 2.0241760624457936, "grad_norm": 0.2636546492576599, "learning_rate": 3.2645437400261135e-05, "loss": 0.0262, "step": 18671 }, { "epoch": 2.024284475281873, "grad_norm": 0.11990836262702942, "learning_rate": 3.26418105324242e-05, "loss": 0.0035, "step": 18672 }, { "epoch": 2.024392888117953, "grad_norm": 0.2808198034763336, "learning_rate": 3.2638183664587264e-05, "loss": 0.008, "step": 18673 }, { "epoch": 2.024501300954033, "grad_norm": 0.029758276417851448, "learning_rate": 3.263455679675033e-05, "loss": 0.0012, "step": 18674 }, { "epoch": 2.0246097137901127, "grad_norm": 0.0211519505828619, "learning_rate": 3.263092992891339e-05, "loss": 0.0007, "step": 18675 }, { "epoch": 2.0247181266261927, "grad_norm": 0.14153560996055603, "learning_rate": 3.2627303061076456e-05, "loss": 0.0051, "step": 18676 }, { "epoch": 2.024826539462272, "grad_norm": 0.43770676851272583, "learning_rate": 3.262367619323952e-05, "loss": 0.0131, "step": 18677 }, { "epoch": 2.024934952298352, "grad_norm": 0.6831501126289368, "learning_rate": 3.2620049325402584e-05, "loss": 0.0601, "step": 18678 }, { "epoch": 2.0250433651344317, "grad_norm": 0.04267507791519165, "learning_rate": 3.261642245756565e-05, "loss": 0.0006, "step": 18679 }, { "epoch": 2.0251517779705117, "grad_norm": 0.5446438789367676, "learning_rate": 3.261279558972871e-05, "loss": 0.0207, "step": 18680 }, { "epoch": 2.0252601908065917, "grad_norm": 0.09549997001886368, "learning_rate": 3.2609168721891776e-05, "loss": 0.0043, "step": 18681 }, { "epoch": 2.0253686036426712, "grad_norm": 0.11486112326383591, "learning_rate": 3.260554185405484e-05, "loss": 0.0034, "step": 18682 }, { "epoch": 2.0254770164787512, "grad_norm": 0.05934321880340576, "learning_rate": 3.2601914986217905e-05, "loss": 0.0021, "step": 18683 }, { "epoch": 2.0255854293148308, "grad_norm": 0.27473214268684387, "learning_rate": 3.259828811838096e-05, "loss": 0.0125, "step": 18684 }, { "epoch": 2.0256938421509108, "grad_norm": 0.07146963477134705, "learning_rate": 3.259466125054403e-05, "loss": 0.0027, "step": 18685 }, { "epoch": 2.0258022549869903, "grad_norm": 0.10341061651706696, "learning_rate": 3.25910343827071e-05, "loss": 0.0041, "step": 18686 }, { "epoch": 2.0259106678230703, "grad_norm": 0.04331531003117561, "learning_rate": 3.258740751487016e-05, "loss": 0.0018, "step": 18687 }, { "epoch": 2.0260190806591503, "grad_norm": 0.20261451601982117, "learning_rate": 3.2583780647033225e-05, "loss": 0.0253, "step": 18688 }, { "epoch": 2.02612749349523, "grad_norm": 0.7761660814285278, "learning_rate": 3.258015377919629e-05, "loss": 0.0544, "step": 18689 }, { "epoch": 2.02623590633131, "grad_norm": 0.1284826099872589, "learning_rate": 3.257652691135935e-05, "loss": 0.0136, "step": 18690 }, { "epoch": 2.0263443191673893, "grad_norm": 0.45542630553245544, "learning_rate": 3.257290004352242e-05, "loss": 0.0076, "step": 18691 }, { "epoch": 2.0264527320034693, "grad_norm": 0.7654567956924438, "learning_rate": 3.2569273175685475e-05, "loss": 0.0165, "step": 18692 }, { "epoch": 2.026561144839549, "grad_norm": 0.05864578112959862, "learning_rate": 3.2565646307848546e-05, "loss": 0.0039, "step": 18693 }, { "epoch": 2.026669557675629, "grad_norm": 0.14717520773410797, "learning_rate": 3.256201944001161e-05, "loss": 0.0023, "step": 18694 }, { "epoch": 2.0267779705117084, "grad_norm": 0.16753830015659332, "learning_rate": 3.2558392572174674e-05, "loss": 0.0121, "step": 18695 }, { "epoch": 2.0268863833477884, "grad_norm": 0.14035604894161224, "learning_rate": 3.255476570433774e-05, "loss": 0.0097, "step": 18696 }, { "epoch": 2.0269947961838684, "grad_norm": 0.3071437180042267, "learning_rate": 3.25511388365008e-05, "loss": 0.0102, "step": 18697 }, { "epoch": 2.027103209019948, "grad_norm": 1.245635747909546, "learning_rate": 3.2547511968663866e-05, "loss": 0.0313, "step": 18698 }, { "epoch": 2.027211621856028, "grad_norm": 0.09642757475376129, "learning_rate": 3.2543885100826924e-05, "loss": 0.0028, "step": 18699 }, { "epoch": 2.0273200346921074, "grad_norm": 0.03357429429888725, "learning_rate": 3.254025823298999e-05, "loss": 0.0016, "step": 18700 }, { "epoch": 2.0274284475281874, "grad_norm": 0.4233146011829376, "learning_rate": 3.253663136515305e-05, "loss": 0.0235, "step": 18701 }, { "epoch": 2.027536860364267, "grad_norm": 0.1267985999584198, "learning_rate": 3.253300449731612e-05, "loss": 0.0047, "step": 18702 }, { "epoch": 2.027645273200347, "grad_norm": 0.06831089407205582, "learning_rate": 3.252937762947919e-05, "loss": 0.0025, "step": 18703 }, { "epoch": 2.027753686036427, "grad_norm": 0.44213786721229553, "learning_rate": 3.252575076164225e-05, "loss": 0.0087, "step": 18704 }, { "epoch": 2.0278620988725065, "grad_norm": 0.778359591960907, "learning_rate": 3.2522123893805315e-05, "loss": 0.0367, "step": 18705 }, { "epoch": 2.0279705117085864, "grad_norm": 0.4309968054294586, "learning_rate": 3.251849702596837e-05, "loss": 0.0099, "step": 18706 }, { "epoch": 2.028078924544666, "grad_norm": 0.21182458102703094, "learning_rate": 3.2514870158131436e-05, "loss": 0.0032, "step": 18707 }, { "epoch": 2.028187337380746, "grad_norm": 0.02069532684981823, "learning_rate": 3.25112432902945e-05, "loss": 0.0008, "step": 18708 }, { "epoch": 2.0282957502168255, "grad_norm": 0.3382418155670166, "learning_rate": 3.2507616422457565e-05, "loss": 0.0577, "step": 18709 }, { "epoch": 2.0284041630529055, "grad_norm": 0.12242097407579422, "learning_rate": 3.250398955462063e-05, "loss": 0.0038, "step": 18710 }, { "epoch": 2.028512575888985, "grad_norm": 0.01813608594238758, "learning_rate": 3.25003626867837e-05, "loss": 0.0008, "step": 18711 }, { "epoch": 2.028620988725065, "grad_norm": 0.3478396236896515, "learning_rate": 3.2496735818946764e-05, "loss": 0.0133, "step": 18712 }, { "epoch": 2.028729401561145, "grad_norm": 0.290500670671463, "learning_rate": 3.249310895110982e-05, "loss": 0.0038, "step": 18713 }, { "epoch": 2.0288378143972245, "grad_norm": 0.02010934241116047, "learning_rate": 3.2489482083272885e-05, "loss": 0.0007, "step": 18714 }, { "epoch": 2.0289462272333045, "grad_norm": 0.616459310054779, "learning_rate": 3.248585521543595e-05, "loss": 0.0562, "step": 18715 }, { "epoch": 2.029054640069384, "grad_norm": 0.11712510883808136, "learning_rate": 3.248222834759901e-05, "loss": 0.0035, "step": 18716 }, { "epoch": 2.029163052905464, "grad_norm": 1.051455020904541, "learning_rate": 3.247860147976208e-05, "loss": 0.0372, "step": 18717 }, { "epoch": 2.0292714657415436, "grad_norm": 0.21627463400363922, "learning_rate": 3.247497461192514e-05, "loss": 0.0121, "step": 18718 }, { "epoch": 2.0293798785776236, "grad_norm": 0.538940966129303, "learning_rate": 3.247134774408821e-05, "loss": 0.0123, "step": 18719 }, { "epoch": 2.0294882914137036, "grad_norm": 0.11543121933937073, "learning_rate": 3.246772087625127e-05, "loss": 0.0025, "step": 18720 }, { "epoch": 2.029596704249783, "grad_norm": 0.10328670591115952, "learning_rate": 3.2464094008414334e-05, "loss": 0.0031, "step": 18721 }, { "epoch": 2.029705117085863, "grad_norm": 0.013299732469022274, "learning_rate": 3.24604671405774e-05, "loss": 0.0007, "step": 18722 }, { "epoch": 2.0298135299219426, "grad_norm": 0.10380150377750397, "learning_rate": 3.245684027274046e-05, "loss": 0.0021, "step": 18723 }, { "epoch": 2.0299219427580226, "grad_norm": 0.25830891728401184, "learning_rate": 3.2453213404903526e-05, "loss": 0.0132, "step": 18724 }, { "epoch": 2.030030355594102, "grad_norm": 0.36275604367256165, "learning_rate": 3.244958653706659e-05, "loss": 0.0103, "step": 18725 }, { "epoch": 2.030138768430182, "grad_norm": 0.3648751974105835, "learning_rate": 3.2445959669229654e-05, "loss": 0.0198, "step": 18726 }, { "epoch": 2.030247181266262, "grad_norm": 0.27389541268348694, "learning_rate": 3.244233280139272e-05, "loss": 0.015, "step": 18727 }, { "epoch": 2.0303555941023417, "grad_norm": 0.27674633264541626, "learning_rate": 3.243870593355578e-05, "loss": 0.0159, "step": 18728 }, { "epoch": 2.0304640069384217, "grad_norm": 0.2244367152452469, "learning_rate": 3.243507906571885e-05, "loss": 0.0121, "step": 18729 }, { "epoch": 2.030572419774501, "grad_norm": 0.027060536667704582, "learning_rate": 3.243145219788191e-05, "loss": 0.0009, "step": 18730 }, { "epoch": 2.030680832610581, "grad_norm": 0.5236746668815613, "learning_rate": 3.2427825330044975e-05, "loss": 0.0353, "step": 18731 }, { "epoch": 2.0307892454466607, "grad_norm": 0.8703826665878296, "learning_rate": 3.242419846220804e-05, "loss": 0.0184, "step": 18732 }, { "epoch": 2.0308976582827407, "grad_norm": 0.2593085467815399, "learning_rate": 3.24205715943711e-05, "loss": 0.0114, "step": 18733 }, { "epoch": 2.0310060711188203, "grad_norm": 0.1579066812992096, "learning_rate": 3.241694472653417e-05, "loss": 0.0045, "step": 18734 }, { "epoch": 2.0311144839549002, "grad_norm": 0.1438855081796646, "learning_rate": 3.241331785869723e-05, "loss": 0.008, "step": 18735 }, { "epoch": 2.0312228967909802, "grad_norm": 0.37895140051841736, "learning_rate": 3.240969099086029e-05, "loss": 0.0411, "step": 18736 }, { "epoch": 2.0313313096270598, "grad_norm": 0.3615497052669525, "learning_rate": 3.240606412302336e-05, "loss": 0.0235, "step": 18737 }, { "epoch": 2.0314397224631398, "grad_norm": 0.26030418276786804, "learning_rate": 3.2402437255186424e-05, "loss": 0.0056, "step": 18738 }, { "epoch": 2.0315481352992193, "grad_norm": 0.03045736625790596, "learning_rate": 3.239881038734949e-05, "loss": 0.0009, "step": 18739 }, { "epoch": 2.0316565481352993, "grad_norm": 0.07400547713041306, "learning_rate": 3.239518351951255e-05, "loss": 0.0029, "step": 18740 }, { "epoch": 2.031764960971379, "grad_norm": 0.04138167202472687, "learning_rate": 3.2391556651675616e-05, "loss": 0.0014, "step": 18741 }, { "epoch": 2.031873373807459, "grad_norm": 0.04664801061153412, "learning_rate": 3.238792978383868e-05, "loss": 0.0016, "step": 18742 }, { "epoch": 2.031981786643539, "grad_norm": 0.5857093334197998, "learning_rate": 3.2384302916001744e-05, "loss": 0.0123, "step": 18743 }, { "epoch": 2.0320901994796183, "grad_norm": 0.5488927364349365, "learning_rate": 3.23806760481648e-05, "loss": 0.0138, "step": 18744 }, { "epoch": 2.0321986123156983, "grad_norm": 0.27515631914138794, "learning_rate": 3.237704918032787e-05, "loss": 0.0047, "step": 18745 }, { "epoch": 2.032307025151778, "grad_norm": 0.3014337718486786, "learning_rate": 3.2373422312490936e-05, "loss": 0.0399, "step": 18746 }, { "epoch": 2.032415437987858, "grad_norm": 0.9305770993232727, "learning_rate": 3.2369795444654e-05, "loss": 0.0147, "step": 18747 }, { "epoch": 2.0325238508239374, "grad_norm": 0.3260987401008606, "learning_rate": 3.2366168576817065e-05, "loss": 0.0037, "step": 18748 }, { "epoch": 2.0326322636600174, "grad_norm": 0.3259705901145935, "learning_rate": 3.236254170898013e-05, "loss": 0.0148, "step": 18749 }, { "epoch": 2.0327406764960974, "grad_norm": 0.356620192527771, "learning_rate": 3.235891484114319e-05, "loss": 0.0044, "step": 18750 }, { "epoch": 2.032849089332177, "grad_norm": 0.8681148886680603, "learning_rate": 3.235528797330625e-05, "loss": 0.0427, "step": 18751 }, { "epoch": 2.032957502168257, "grad_norm": 0.12560495734214783, "learning_rate": 3.2351661105469314e-05, "loss": 0.0088, "step": 18752 }, { "epoch": 2.0330659150043364, "grad_norm": 1.596985936164856, "learning_rate": 3.234803423763238e-05, "loss": 0.0216, "step": 18753 }, { "epoch": 2.0331743278404164, "grad_norm": 0.3778775930404663, "learning_rate": 3.234440736979545e-05, "loss": 0.0097, "step": 18754 }, { "epoch": 2.033282740676496, "grad_norm": 0.1785956174135208, "learning_rate": 3.2340780501958513e-05, "loss": 0.0084, "step": 18755 }, { "epoch": 2.033391153512576, "grad_norm": 0.5154134631156921, "learning_rate": 3.233715363412158e-05, "loss": 0.0119, "step": 18756 }, { "epoch": 2.0334995663486555, "grad_norm": 0.1959385722875595, "learning_rate": 3.233352676628464e-05, "loss": 0.0083, "step": 18757 }, { "epoch": 2.0336079791847355, "grad_norm": 0.1507204920053482, "learning_rate": 3.23298998984477e-05, "loss": 0.0036, "step": 18758 }, { "epoch": 2.0337163920208154, "grad_norm": 0.016162719577550888, "learning_rate": 3.232627303061076e-05, "loss": 0.0008, "step": 18759 }, { "epoch": 2.033824804856895, "grad_norm": 0.25419721007347107, "learning_rate": 3.232264616277383e-05, "loss": 0.0096, "step": 18760 }, { "epoch": 2.033933217692975, "grad_norm": 0.3931259214878082, "learning_rate": 3.231901929493689e-05, "loss": 0.0224, "step": 18761 }, { "epoch": 2.0340416305290545, "grad_norm": 0.26105743646621704, "learning_rate": 3.2315392427099955e-05, "loss": 0.0162, "step": 18762 }, { "epoch": 2.0341500433651345, "grad_norm": 0.2661619484424591, "learning_rate": 3.2311765559263026e-05, "loss": 0.0081, "step": 18763 }, { "epoch": 2.034258456201214, "grad_norm": 0.15241371095180511, "learning_rate": 3.230813869142609e-05, "loss": 0.0071, "step": 18764 }, { "epoch": 2.034366869037294, "grad_norm": 0.5082676410675049, "learning_rate": 3.230451182358915e-05, "loss": 0.014, "step": 18765 }, { "epoch": 2.034475281873374, "grad_norm": 0.11369289457798004, "learning_rate": 3.230088495575221e-05, "loss": 0.0053, "step": 18766 }, { "epoch": 2.0345836947094535, "grad_norm": 0.4588376581668854, "learning_rate": 3.2297258087915276e-05, "loss": 0.0439, "step": 18767 }, { "epoch": 2.0346921075455335, "grad_norm": 0.007626429665833712, "learning_rate": 3.229363122007834e-05, "loss": 0.0004, "step": 18768 }, { "epoch": 2.034800520381613, "grad_norm": 0.47828173637390137, "learning_rate": 3.2290004352241404e-05, "loss": 0.0339, "step": 18769 }, { "epoch": 2.034908933217693, "grad_norm": 0.24952095746994019, "learning_rate": 3.228637748440447e-05, "loss": 0.0148, "step": 18770 }, { "epoch": 2.0350173460537726, "grad_norm": 0.1642792969942093, "learning_rate": 3.228275061656754e-05, "loss": 0.0076, "step": 18771 }, { "epoch": 2.0351257588898526, "grad_norm": 0.3433755040168762, "learning_rate": 3.2279123748730596e-05, "loss": 0.0218, "step": 18772 }, { "epoch": 2.035234171725932, "grad_norm": 0.38463783264160156, "learning_rate": 3.227549688089366e-05, "loss": 0.0489, "step": 18773 }, { "epoch": 2.035342584562012, "grad_norm": 0.34429916739463806, "learning_rate": 3.2271870013056725e-05, "loss": 0.0143, "step": 18774 }, { "epoch": 2.035450997398092, "grad_norm": 0.09267941117286682, "learning_rate": 3.226824314521979e-05, "loss": 0.0033, "step": 18775 }, { "epoch": 2.0355594102341716, "grad_norm": 0.01679554581642151, "learning_rate": 3.226461627738285e-05, "loss": 0.0008, "step": 18776 }, { "epoch": 2.0356678230702516, "grad_norm": 0.03881499916315079, "learning_rate": 3.226098940954592e-05, "loss": 0.0014, "step": 18777 }, { "epoch": 2.035776235906331, "grad_norm": 0.012896263971924782, "learning_rate": 3.225736254170898e-05, "loss": 0.0005, "step": 18778 }, { "epoch": 2.035884648742411, "grad_norm": 0.13329483568668365, "learning_rate": 3.2253735673872045e-05, "loss": 0.0036, "step": 18779 }, { "epoch": 2.0359930615784907, "grad_norm": 0.015888525173068047, "learning_rate": 3.225010880603511e-05, "loss": 0.0009, "step": 18780 }, { "epoch": 2.0361014744145707, "grad_norm": 0.14282645285129547, "learning_rate": 3.224648193819817e-05, "loss": 0.009, "step": 18781 }, { "epoch": 2.0362098872506507, "grad_norm": 0.39710378646850586, "learning_rate": 3.224285507036124e-05, "loss": 0.0207, "step": 18782 }, { "epoch": 2.03631830008673, "grad_norm": 0.17619365453720093, "learning_rate": 3.22392282025243e-05, "loss": 0.018, "step": 18783 }, { "epoch": 2.03642671292281, "grad_norm": 0.12210440635681152, "learning_rate": 3.2235601334687366e-05, "loss": 0.0045, "step": 18784 }, { "epoch": 2.0365351257588897, "grad_norm": 0.36737093329429626, "learning_rate": 3.223197446685043e-05, "loss": 0.023, "step": 18785 }, { "epoch": 2.0366435385949697, "grad_norm": 0.0927489846944809, "learning_rate": 3.2228347599013494e-05, "loss": 0.0032, "step": 18786 }, { "epoch": 2.0367519514310493, "grad_norm": 0.18613915145397186, "learning_rate": 3.222472073117656e-05, "loss": 0.0204, "step": 18787 }, { "epoch": 2.0368603642671292, "grad_norm": 0.05818060413002968, "learning_rate": 3.222109386333962e-05, "loss": 0.0019, "step": 18788 }, { "epoch": 2.0369687771032092, "grad_norm": 0.58335942029953, "learning_rate": 3.2217466995502686e-05, "loss": 0.0142, "step": 18789 }, { "epoch": 2.0370771899392888, "grad_norm": 0.4041042923927307, "learning_rate": 3.221384012766575e-05, "loss": 0.0205, "step": 18790 }, { "epoch": 2.0371856027753688, "grad_norm": 0.07680247724056244, "learning_rate": 3.2210213259828814e-05, "loss": 0.0018, "step": 18791 }, { "epoch": 2.0372940156114483, "grad_norm": 0.13350172340869904, "learning_rate": 3.220658639199188e-05, "loss": 0.01, "step": 18792 }, { "epoch": 2.0374024284475283, "grad_norm": 0.0858035683631897, "learning_rate": 3.220295952415494e-05, "loss": 0.0055, "step": 18793 }, { "epoch": 2.037510841283608, "grad_norm": 0.30811846256256104, "learning_rate": 3.219933265631801e-05, "loss": 0.0135, "step": 18794 }, { "epoch": 2.037619254119688, "grad_norm": 0.4269574284553528, "learning_rate": 3.219570578848107e-05, "loss": 0.008, "step": 18795 }, { "epoch": 2.0377276669557673, "grad_norm": 0.013500829227268696, "learning_rate": 3.219207892064413e-05, "loss": 0.0011, "step": 18796 }, { "epoch": 2.0378360797918473, "grad_norm": 0.0982949510216713, "learning_rate": 3.21884520528072e-05, "loss": 0.0145, "step": 18797 }, { "epoch": 2.0379444926279273, "grad_norm": 0.06753239035606384, "learning_rate": 3.218482518497026e-05, "loss": 0.0029, "step": 18798 }, { "epoch": 2.038052905464007, "grad_norm": 0.32083871960639954, "learning_rate": 3.218119831713333e-05, "loss": 0.0297, "step": 18799 }, { "epoch": 2.038161318300087, "grad_norm": 0.4868074059486389, "learning_rate": 3.217757144929639e-05, "loss": 0.0384, "step": 18800 }, { "epoch": 2.0382697311361664, "grad_norm": 0.009612719528377056, "learning_rate": 3.2173944581459455e-05, "loss": 0.0004, "step": 18801 }, { "epoch": 2.0383781439722464, "grad_norm": 0.38322892785072327, "learning_rate": 3.217031771362252e-05, "loss": 0.0125, "step": 18802 }, { "epoch": 2.038486556808326, "grad_norm": 0.3073200285434723, "learning_rate": 3.216669084578558e-05, "loss": 0.0123, "step": 18803 }, { "epoch": 2.038594969644406, "grad_norm": 0.06660638749599457, "learning_rate": 3.216306397794864e-05, "loss": 0.0023, "step": 18804 }, { "epoch": 2.038703382480486, "grad_norm": 0.21909931302070618, "learning_rate": 3.2159437110111705e-05, "loss": 0.0046, "step": 18805 }, { "epoch": 2.0388117953165654, "grad_norm": 0.05752824246883392, "learning_rate": 3.2155810242274776e-05, "loss": 0.0016, "step": 18806 }, { "epoch": 2.0389202081526454, "grad_norm": 0.26285281777381897, "learning_rate": 3.215218337443784e-05, "loss": 0.0089, "step": 18807 }, { "epoch": 2.039028620988725, "grad_norm": 0.43699780106544495, "learning_rate": 3.2148556506600904e-05, "loss": 0.0243, "step": 18808 }, { "epoch": 2.039137033824805, "grad_norm": 1.7018980979919434, "learning_rate": 3.214492963876397e-05, "loss": 0.0496, "step": 18809 }, { "epoch": 2.0392454466608845, "grad_norm": 0.09941211342811584, "learning_rate": 3.2141302770927026e-05, "loss": 0.0042, "step": 18810 }, { "epoch": 2.0393538594969645, "grad_norm": 0.26380982995033264, "learning_rate": 3.213767590309009e-05, "loss": 0.0374, "step": 18811 }, { "epoch": 2.0394622723330444, "grad_norm": 0.09568411111831665, "learning_rate": 3.2134049035253154e-05, "loss": 0.0022, "step": 18812 }, { "epoch": 2.039570685169124, "grad_norm": 0.09223924577236176, "learning_rate": 3.213042216741622e-05, "loss": 0.0058, "step": 18813 }, { "epoch": 2.039679098005204, "grad_norm": 0.05691694840788841, "learning_rate": 3.212679529957929e-05, "loss": 0.0019, "step": 18814 }, { "epoch": 2.0397875108412835, "grad_norm": 0.18776531517505646, "learning_rate": 3.212316843174235e-05, "loss": 0.0081, "step": 18815 }, { "epoch": 2.0398959236773635, "grad_norm": 0.3909391760826111, "learning_rate": 3.211954156390542e-05, "loss": 0.0135, "step": 18816 }, { "epoch": 2.040004336513443, "grad_norm": 0.16250266134738922, "learning_rate": 3.2115914696068474e-05, "loss": 0.0069, "step": 18817 }, { "epoch": 2.040112749349523, "grad_norm": 0.20402474701404572, "learning_rate": 3.211228782823154e-05, "loss": 0.0241, "step": 18818 }, { "epoch": 2.0402211621856026, "grad_norm": 0.0057461922988295555, "learning_rate": 3.21086609603946e-05, "loss": 0.0002, "step": 18819 }, { "epoch": 2.0403295750216826, "grad_norm": 0.3832765519618988, "learning_rate": 3.210503409255767e-05, "loss": 0.0057, "step": 18820 }, { "epoch": 2.0404379878577625, "grad_norm": 0.17140208184719086, "learning_rate": 3.210140722472073e-05, "loss": 0.0122, "step": 18821 }, { "epoch": 2.040546400693842, "grad_norm": 0.1332557648420334, "learning_rate": 3.2097780356883795e-05, "loss": 0.0099, "step": 18822 }, { "epoch": 2.040654813529922, "grad_norm": 0.15169280767440796, "learning_rate": 3.2094153489046866e-05, "loss": 0.0074, "step": 18823 }, { "epoch": 2.0407632263660016, "grad_norm": 0.006402851082384586, "learning_rate": 3.209052662120992e-05, "loss": 0.0002, "step": 18824 }, { "epoch": 2.0408716392020816, "grad_norm": 0.5949395895004272, "learning_rate": 3.208689975337299e-05, "loss": 0.0716, "step": 18825 }, { "epoch": 2.040980052038161, "grad_norm": 0.018689000979065895, "learning_rate": 3.208327288553605e-05, "loss": 0.0005, "step": 18826 }, { "epoch": 2.041088464874241, "grad_norm": 0.8736394047737122, "learning_rate": 3.2079646017699115e-05, "loss": 0.027, "step": 18827 }, { "epoch": 2.041196877710321, "grad_norm": 0.34571725130081177, "learning_rate": 3.207601914986218e-05, "loss": 0.0139, "step": 18828 }, { "epoch": 2.0413052905464006, "grad_norm": 0.3833322823047638, "learning_rate": 3.2072392282025244e-05, "loss": 0.0183, "step": 18829 }, { "epoch": 2.0414137033824806, "grad_norm": 0.06217781826853752, "learning_rate": 3.206876541418831e-05, "loss": 0.003, "step": 18830 }, { "epoch": 2.04152211621856, "grad_norm": 0.19251276552677155, "learning_rate": 3.206513854635137e-05, "loss": 0.0041, "step": 18831 }, { "epoch": 2.04163052905464, "grad_norm": 0.3287268579006195, "learning_rate": 3.2061511678514436e-05, "loss": 0.0136, "step": 18832 }, { "epoch": 2.0417389418907197, "grad_norm": 0.3350420892238617, "learning_rate": 3.20578848106775e-05, "loss": 0.0171, "step": 18833 }, { "epoch": 2.0418473547267997, "grad_norm": 0.36020779609680176, "learning_rate": 3.2054257942840564e-05, "loss": 0.0189, "step": 18834 }, { "epoch": 2.0419557675628797, "grad_norm": 2.211329460144043, "learning_rate": 3.205063107500363e-05, "loss": 0.0192, "step": 18835 }, { "epoch": 2.042064180398959, "grad_norm": 1.0155584812164307, "learning_rate": 3.204700420716669e-05, "loss": 0.0265, "step": 18836 }, { "epoch": 2.042172593235039, "grad_norm": 0.024629903957247734, "learning_rate": 3.2043377339329756e-05, "loss": 0.0006, "step": 18837 }, { "epoch": 2.0422810060711187, "grad_norm": 0.09226828813552856, "learning_rate": 3.203975047149282e-05, "loss": 0.0078, "step": 18838 }, { "epoch": 2.0423894189071987, "grad_norm": 0.5813199281692505, "learning_rate": 3.2036123603655885e-05, "loss": 0.0154, "step": 18839 }, { "epoch": 2.0424978317432783, "grad_norm": 0.09440486878156662, "learning_rate": 3.203249673581895e-05, "loss": 0.003, "step": 18840 }, { "epoch": 2.0426062445793582, "grad_norm": 0.048720043152570724, "learning_rate": 3.202886986798201e-05, "loss": 0.0034, "step": 18841 }, { "epoch": 2.042714657415438, "grad_norm": 0.08484942466020584, "learning_rate": 3.202524300014508e-05, "loss": 0.0053, "step": 18842 }, { "epoch": 2.0428230702515178, "grad_norm": 0.23811952769756317, "learning_rate": 3.202161613230814e-05, "loss": 0.0102, "step": 18843 }, { "epoch": 2.0429314830875978, "grad_norm": 0.04612459987401962, "learning_rate": 3.2017989264471205e-05, "loss": 0.0025, "step": 18844 }, { "epoch": 2.0430398959236773, "grad_norm": 0.1556217521429062, "learning_rate": 3.201436239663427e-05, "loss": 0.0041, "step": 18845 }, { "epoch": 2.0431483087597573, "grad_norm": 0.02027890272438526, "learning_rate": 3.2010735528797333e-05, "loss": 0.0008, "step": 18846 }, { "epoch": 2.043256721595837, "grad_norm": 0.07330784201622009, "learning_rate": 3.20071086609604e-05, "loss": 0.0032, "step": 18847 }, { "epoch": 2.043365134431917, "grad_norm": 1.3342154026031494, "learning_rate": 3.2003481793123455e-05, "loss": 0.0851, "step": 18848 }, { "epoch": 2.0434735472679963, "grad_norm": 0.16462017595767975, "learning_rate": 3.1999854925286526e-05, "loss": 0.0036, "step": 18849 }, { "epoch": 2.0435819601040763, "grad_norm": 0.4878266155719757, "learning_rate": 3.199622805744959e-05, "loss": 0.0258, "step": 18850 }, { "epoch": 2.0436903729401563, "grad_norm": 0.31397518515586853, "learning_rate": 3.1992601189612654e-05, "loss": 0.023, "step": 18851 }, { "epoch": 2.043798785776236, "grad_norm": 0.2325313538312912, "learning_rate": 3.198897432177572e-05, "loss": 0.0071, "step": 18852 }, { "epoch": 2.043907198612316, "grad_norm": 0.09697356820106506, "learning_rate": 3.198534745393878e-05, "loss": 0.0023, "step": 18853 }, { "epoch": 2.0440156114483954, "grad_norm": 0.17337392270565033, "learning_rate": 3.1981720586101846e-05, "loss": 0.0053, "step": 18854 }, { "epoch": 2.0441240242844754, "grad_norm": 0.13551250100135803, "learning_rate": 3.1978093718264904e-05, "loss": 0.0033, "step": 18855 }, { "epoch": 2.044232437120555, "grad_norm": 0.171996608376503, "learning_rate": 3.197446685042797e-05, "loss": 0.009, "step": 18856 }, { "epoch": 2.044340849956635, "grad_norm": 0.2338186353445053, "learning_rate": 3.197083998259103e-05, "loss": 0.009, "step": 18857 }, { "epoch": 2.0444492627927144, "grad_norm": 1.3145726919174194, "learning_rate": 3.19672131147541e-05, "loss": 0.0422, "step": 18858 }, { "epoch": 2.0445576756287944, "grad_norm": 0.19993098080158234, "learning_rate": 3.196358624691717e-05, "loss": 0.0068, "step": 18859 }, { "epoch": 2.0446660884648744, "grad_norm": 0.032210614532232285, "learning_rate": 3.195995937908023e-05, "loss": 0.0014, "step": 18860 }, { "epoch": 2.044774501300954, "grad_norm": 0.04836287721991539, "learning_rate": 3.1956332511243295e-05, "loss": 0.0018, "step": 18861 }, { "epoch": 2.044882914137034, "grad_norm": 0.1406579166650772, "learning_rate": 3.195270564340635e-05, "loss": 0.0015, "step": 18862 }, { "epoch": 2.0449913269731135, "grad_norm": 0.17106181383132935, "learning_rate": 3.1949078775569416e-05, "loss": 0.0137, "step": 18863 }, { "epoch": 2.0450997398091935, "grad_norm": 0.1448395848274231, "learning_rate": 3.194545190773248e-05, "loss": 0.0173, "step": 18864 }, { "epoch": 2.045208152645273, "grad_norm": 0.37481924891471863, "learning_rate": 3.1941825039895545e-05, "loss": 0.0131, "step": 18865 }, { "epoch": 2.045316565481353, "grad_norm": 0.4407820403575897, "learning_rate": 3.1938198172058616e-05, "loss": 0.029, "step": 18866 }, { "epoch": 2.045424978317433, "grad_norm": 0.058128632605075836, "learning_rate": 3.193457130422168e-05, "loss": 0.0028, "step": 18867 }, { "epoch": 2.0455333911535125, "grad_norm": 0.12208499759435654, "learning_rate": 3.1930944436384744e-05, "loss": 0.0072, "step": 18868 }, { "epoch": 2.0456418039895925, "grad_norm": 0.1332550346851349, "learning_rate": 3.19273175685478e-05, "loss": 0.0079, "step": 18869 }, { "epoch": 2.045750216825672, "grad_norm": 0.09277956187725067, "learning_rate": 3.1923690700710865e-05, "loss": 0.0015, "step": 18870 }, { "epoch": 2.045858629661752, "grad_norm": 0.28738468885421753, "learning_rate": 3.192006383287393e-05, "loss": 0.0272, "step": 18871 }, { "epoch": 2.0459670424978316, "grad_norm": 0.14618420600891113, "learning_rate": 3.191643696503699e-05, "loss": 0.0218, "step": 18872 }, { "epoch": 2.0460754553339116, "grad_norm": 0.15333117544651031, "learning_rate": 3.191281009720006e-05, "loss": 0.0107, "step": 18873 }, { "epoch": 2.0461838681699915, "grad_norm": 0.29341670870780945, "learning_rate": 3.190918322936312e-05, "loss": 0.0059, "step": 18874 }, { "epoch": 2.046292281006071, "grad_norm": 0.25048571825027466, "learning_rate": 3.190555636152619e-05, "loss": 0.0119, "step": 18875 }, { "epoch": 2.046400693842151, "grad_norm": 0.14912942051887512, "learning_rate": 3.1901929493689257e-05, "loss": 0.0064, "step": 18876 }, { "epoch": 2.0465091066782306, "grad_norm": 0.3825646936893463, "learning_rate": 3.1898302625852314e-05, "loss": 0.018, "step": 18877 }, { "epoch": 2.0466175195143106, "grad_norm": 0.39027029275894165, "learning_rate": 3.189467575801538e-05, "loss": 0.0195, "step": 18878 }, { "epoch": 2.04672593235039, "grad_norm": 0.3380383551120758, "learning_rate": 3.189104889017844e-05, "loss": 0.0161, "step": 18879 }, { "epoch": 2.04683434518647, "grad_norm": 0.28889498114585876, "learning_rate": 3.1887422022341506e-05, "loss": 0.0075, "step": 18880 }, { "epoch": 2.0469427580225497, "grad_norm": 0.1927078515291214, "learning_rate": 3.188379515450457e-05, "loss": 0.009, "step": 18881 }, { "epoch": 2.0470511708586296, "grad_norm": 0.38827863335609436, "learning_rate": 3.1880168286667634e-05, "loss": 0.0451, "step": 18882 }, { "epoch": 2.0471595836947096, "grad_norm": 0.030492886900901794, "learning_rate": 3.1876541418830705e-05, "loss": 0.0012, "step": 18883 }, { "epoch": 2.047267996530789, "grad_norm": 0.21929188072681427, "learning_rate": 3.187291455099376e-05, "loss": 0.0136, "step": 18884 }, { "epoch": 2.047376409366869, "grad_norm": 0.1473194658756256, "learning_rate": 3.186928768315683e-05, "loss": 0.0135, "step": 18885 }, { "epoch": 2.0474848222029487, "grad_norm": 0.16183841228485107, "learning_rate": 3.186566081531989e-05, "loss": 0.0062, "step": 18886 }, { "epoch": 2.0475932350390287, "grad_norm": 0.399125337600708, "learning_rate": 3.1862033947482955e-05, "loss": 0.0181, "step": 18887 }, { "epoch": 2.047701647875108, "grad_norm": 0.08383280783891678, "learning_rate": 3.185840707964602e-05, "loss": 0.0026, "step": 18888 }, { "epoch": 2.047810060711188, "grad_norm": 0.1476723849773407, "learning_rate": 3.185478021180908e-05, "loss": 0.004, "step": 18889 }, { "epoch": 2.047918473547268, "grad_norm": 0.6650351881980896, "learning_rate": 3.185115334397215e-05, "loss": 0.0136, "step": 18890 }, { "epoch": 2.0480268863833477, "grad_norm": 0.10492320358753204, "learning_rate": 3.184752647613521e-05, "loss": 0.0038, "step": 18891 }, { "epoch": 2.0481352992194277, "grad_norm": 0.8874378800392151, "learning_rate": 3.1843899608298275e-05, "loss": 0.0311, "step": 18892 }, { "epoch": 2.0482437120555073, "grad_norm": 0.24254122376441956, "learning_rate": 3.184027274046134e-05, "loss": 0.0145, "step": 18893 }, { "epoch": 2.0483521248915872, "grad_norm": 0.08552351593971252, "learning_rate": 3.1836645872624404e-05, "loss": 0.0035, "step": 18894 }, { "epoch": 2.048460537727667, "grad_norm": 0.1571345329284668, "learning_rate": 3.183301900478747e-05, "loss": 0.0041, "step": 18895 }, { "epoch": 2.0485689505637468, "grad_norm": 0.014783589169383049, "learning_rate": 3.182939213695053e-05, "loss": 0.0007, "step": 18896 }, { "epoch": 2.0486773633998268, "grad_norm": 0.05668625608086586, "learning_rate": 3.1825765269113596e-05, "loss": 0.0021, "step": 18897 }, { "epoch": 2.0487857762359063, "grad_norm": 0.0196865051984787, "learning_rate": 3.182213840127666e-05, "loss": 0.0009, "step": 18898 }, { "epoch": 2.0488941890719863, "grad_norm": 0.12121431529521942, "learning_rate": 3.1818511533439724e-05, "loss": 0.0019, "step": 18899 }, { "epoch": 2.049002601908066, "grad_norm": 1.2226872444152832, "learning_rate": 3.181488466560278e-05, "loss": 0.0346, "step": 18900 }, { "epoch": 2.049111014744146, "grad_norm": 0.17181874811649323, "learning_rate": 3.181125779776585e-05, "loss": 0.0216, "step": 18901 }, { "epoch": 2.0492194275802254, "grad_norm": 0.07300236076116562, "learning_rate": 3.1807630929928917e-05, "loss": 0.003, "step": 18902 }, { "epoch": 2.0493278404163053, "grad_norm": 0.07916311174631119, "learning_rate": 3.180400406209198e-05, "loss": 0.003, "step": 18903 }, { "epoch": 2.049436253252385, "grad_norm": 0.4002538025379181, "learning_rate": 3.1800377194255045e-05, "loss": 0.0092, "step": 18904 }, { "epoch": 2.049544666088465, "grad_norm": 0.154643714427948, "learning_rate": 3.179675032641811e-05, "loss": 0.0029, "step": 18905 }, { "epoch": 2.049653078924545, "grad_norm": 0.005370952654629946, "learning_rate": 3.179312345858117e-05, "loss": 0.0003, "step": 18906 }, { "epoch": 2.0497614917606244, "grad_norm": 0.032008785754442215, "learning_rate": 3.178949659074423e-05, "loss": 0.0012, "step": 18907 }, { "epoch": 2.0498699045967044, "grad_norm": 0.014583437703549862, "learning_rate": 3.1785869722907294e-05, "loss": 0.0009, "step": 18908 }, { "epoch": 2.049978317432784, "grad_norm": 0.16648271679878235, "learning_rate": 3.1782242855070365e-05, "loss": 0.0029, "step": 18909 }, { "epoch": 2.050086730268864, "grad_norm": 0.10988244414329529, "learning_rate": 3.177861598723343e-05, "loss": 0.0021, "step": 18910 }, { "epoch": 2.0501951431049434, "grad_norm": 0.02018510177731514, "learning_rate": 3.1774989119396493e-05, "loss": 0.0008, "step": 18911 }, { "epoch": 2.0503035559410234, "grad_norm": 0.3436075448989868, "learning_rate": 3.177136225155956e-05, "loss": 0.006, "step": 18912 }, { "epoch": 2.0504119687771034, "grad_norm": 0.7952069640159607, "learning_rate": 3.176773538372262e-05, "loss": 0.0501, "step": 18913 }, { "epoch": 2.050520381613183, "grad_norm": 0.12046986073255539, "learning_rate": 3.176410851588568e-05, "loss": 0.0045, "step": 18914 }, { "epoch": 2.050628794449263, "grad_norm": 0.18160679936408997, "learning_rate": 3.176048164804874e-05, "loss": 0.0116, "step": 18915 }, { "epoch": 2.0507372072853425, "grad_norm": 0.5306549668312073, "learning_rate": 3.175685478021181e-05, "loss": 0.0493, "step": 18916 }, { "epoch": 2.0508456201214225, "grad_norm": 0.5228347182273865, "learning_rate": 3.175322791237487e-05, "loss": 0.0204, "step": 18917 }, { "epoch": 2.050954032957502, "grad_norm": 0.5237078070640564, "learning_rate": 3.174960104453794e-05, "loss": 0.0056, "step": 18918 }, { "epoch": 2.051062445793582, "grad_norm": 0.5673643350601196, "learning_rate": 3.1745974176701006e-05, "loss": 0.0136, "step": 18919 }, { "epoch": 2.0511708586296615, "grad_norm": 0.1231713593006134, "learning_rate": 3.174234730886407e-05, "loss": 0.0042, "step": 18920 }, { "epoch": 2.0512792714657415, "grad_norm": 0.20945775508880615, "learning_rate": 3.173872044102713e-05, "loss": 0.0252, "step": 18921 }, { "epoch": 2.0513876843018215, "grad_norm": 0.06458526104688644, "learning_rate": 3.173509357319019e-05, "loss": 0.0032, "step": 18922 }, { "epoch": 2.051496097137901, "grad_norm": 1.5472980737686157, "learning_rate": 3.1731466705353256e-05, "loss": 0.024, "step": 18923 }, { "epoch": 2.051604509973981, "grad_norm": 0.20368023216724396, "learning_rate": 3.172783983751632e-05, "loss": 0.0149, "step": 18924 }, { "epoch": 2.0517129228100606, "grad_norm": 0.1744769811630249, "learning_rate": 3.1724212969679384e-05, "loss": 0.0031, "step": 18925 }, { "epoch": 2.0518213356461406, "grad_norm": 0.5409833192825317, "learning_rate": 3.172058610184245e-05, "loss": 0.0101, "step": 18926 }, { "epoch": 2.05192974848222, "grad_norm": 0.22174842655658722, "learning_rate": 3.171695923400552e-05, "loss": 0.014, "step": 18927 }, { "epoch": 2.0520381613183, "grad_norm": 1.327747106552124, "learning_rate": 3.171333236616858e-05, "loss": 0.0209, "step": 18928 }, { "epoch": 2.05214657415438, "grad_norm": 0.18250280618667603, "learning_rate": 3.170970549833164e-05, "loss": 0.0056, "step": 18929 }, { "epoch": 2.0522549869904596, "grad_norm": 0.281464546918869, "learning_rate": 3.1706078630494705e-05, "loss": 0.0046, "step": 18930 }, { "epoch": 2.0523633998265396, "grad_norm": 0.04529425501823425, "learning_rate": 3.170245176265777e-05, "loss": 0.0015, "step": 18931 }, { "epoch": 2.052471812662619, "grad_norm": 0.1645939201116562, "learning_rate": 3.169882489482083e-05, "loss": 0.0124, "step": 18932 }, { "epoch": 2.052580225498699, "grad_norm": 0.1090959683060646, "learning_rate": 3.16951980269839e-05, "loss": 0.0026, "step": 18933 }, { "epoch": 2.0526886383347787, "grad_norm": 0.21448370814323425, "learning_rate": 3.169157115914696e-05, "loss": 0.0042, "step": 18934 }, { "epoch": 2.0527970511708586, "grad_norm": 0.2533312141895294, "learning_rate": 3.168794429131003e-05, "loss": 0.0049, "step": 18935 }, { "epoch": 2.0529054640069386, "grad_norm": 0.32014667987823486, "learning_rate": 3.168431742347309e-05, "loss": 0.0133, "step": 18936 }, { "epoch": 2.053013876843018, "grad_norm": 0.023750493302941322, "learning_rate": 3.1680690555636153e-05, "loss": 0.0008, "step": 18937 }, { "epoch": 2.053122289679098, "grad_norm": 0.06968525052070618, "learning_rate": 3.167706368779922e-05, "loss": 0.0017, "step": 18938 }, { "epoch": 2.0532307025151777, "grad_norm": 0.06418093293905258, "learning_rate": 3.167343681996228e-05, "loss": 0.0029, "step": 18939 }, { "epoch": 2.0533391153512577, "grad_norm": 0.46637311577796936, "learning_rate": 3.1669809952125346e-05, "loss": 0.0113, "step": 18940 }, { "epoch": 2.0534475281873372, "grad_norm": 0.11510252952575684, "learning_rate": 3.166618308428841e-05, "loss": 0.0026, "step": 18941 }, { "epoch": 2.053555941023417, "grad_norm": 0.17626133561134338, "learning_rate": 3.1662556216451474e-05, "loss": 0.0058, "step": 18942 }, { "epoch": 2.0536643538594967, "grad_norm": 1.3289299011230469, "learning_rate": 3.165892934861454e-05, "loss": 0.0325, "step": 18943 }, { "epoch": 2.0537727666955767, "grad_norm": 0.14818565547466278, "learning_rate": 3.16553024807776e-05, "loss": 0.0076, "step": 18944 }, { "epoch": 2.0538811795316567, "grad_norm": 0.5152555108070374, "learning_rate": 3.1651675612940666e-05, "loss": 0.0308, "step": 18945 }, { "epoch": 2.0539895923677363, "grad_norm": 0.02158859372138977, "learning_rate": 3.164804874510373e-05, "loss": 0.0006, "step": 18946 }, { "epoch": 2.0540980052038162, "grad_norm": 0.2423202097415924, "learning_rate": 3.1644421877266794e-05, "loss": 0.0119, "step": 18947 }, { "epoch": 2.054206418039896, "grad_norm": 0.017047956585884094, "learning_rate": 3.164079500942986e-05, "loss": 0.0005, "step": 18948 }, { "epoch": 2.0543148308759758, "grad_norm": 0.23649422824382782, "learning_rate": 3.163716814159292e-05, "loss": 0.0139, "step": 18949 }, { "epoch": 2.0544232437120553, "grad_norm": 0.32828986644744873, "learning_rate": 3.163354127375599e-05, "loss": 0.0453, "step": 18950 }, { "epoch": 2.0545316565481353, "grad_norm": 0.017759187147021294, "learning_rate": 3.162991440591905e-05, "loss": 0.0004, "step": 18951 }, { "epoch": 2.0546400693842153, "grad_norm": 0.3596729338169098, "learning_rate": 3.162628753808211e-05, "loss": 0.0151, "step": 18952 }, { "epoch": 2.054748482220295, "grad_norm": 0.05268967151641846, "learning_rate": 3.162266067024518e-05, "loss": 0.0021, "step": 18953 }, { "epoch": 2.054856895056375, "grad_norm": 0.08029509335756302, "learning_rate": 3.161903380240824e-05, "loss": 0.0033, "step": 18954 }, { "epoch": 2.0549653078924544, "grad_norm": 0.13288433849811554, "learning_rate": 3.161540693457131e-05, "loss": 0.0021, "step": 18955 }, { "epoch": 2.0550737207285343, "grad_norm": 0.5631069540977478, "learning_rate": 3.161178006673437e-05, "loss": 0.0069, "step": 18956 }, { "epoch": 2.055182133564614, "grad_norm": 0.06962867826223373, "learning_rate": 3.1608153198897436e-05, "loss": 0.0018, "step": 18957 }, { "epoch": 2.055290546400694, "grad_norm": 0.02897876687347889, "learning_rate": 3.16045263310605e-05, "loss": 0.0008, "step": 18958 }, { "epoch": 2.055398959236774, "grad_norm": 0.007587254513055086, "learning_rate": 3.160089946322356e-05, "loss": 0.0002, "step": 18959 }, { "epoch": 2.0555073720728534, "grad_norm": 0.2912980616092682, "learning_rate": 3.159727259538662e-05, "loss": 0.0449, "step": 18960 }, { "epoch": 2.0556157849089334, "grad_norm": 0.09537617862224579, "learning_rate": 3.159364572754969e-05, "loss": 0.0014, "step": 18961 }, { "epoch": 2.055724197745013, "grad_norm": 0.13571366667747498, "learning_rate": 3.1590018859712756e-05, "loss": 0.0033, "step": 18962 }, { "epoch": 2.055832610581093, "grad_norm": 0.5959500670433044, "learning_rate": 3.158639199187582e-05, "loss": 0.0095, "step": 18963 }, { "epoch": 2.0559410234171724, "grad_norm": 0.040442075580358505, "learning_rate": 3.1582765124038884e-05, "loss": 0.0045, "step": 18964 }, { "epoch": 2.0560494362532524, "grad_norm": 0.11357929557561874, "learning_rate": 3.157913825620195e-05, "loss": 0.0042, "step": 18965 }, { "epoch": 2.056157849089332, "grad_norm": 0.1866024136543274, "learning_rate": 3.1575511388365006e-05, "loss": 0.0129, "step": 18966 }, { "epoch": 2.056266261925412, "grad_norm": 0.11488837003707886, "learning_rate": 3.157188452052807e-05, "loss": 0.0199, "step": 18967 }, { "epoch": 2.056374674761492, "grad_norm": 0.030297858640551567, "learning_rate": 3.1568257652691134e-05, "loss": 0.0008, "step": 18968 }, { "epoch": 2.0564830875975715, "grad_norm": 0.06747103482484818, "learning_rate": 3.15646307848542e-05, "loss": 0.0038, "step": 18969 }, { "epoch": 2.0565915004336515, "grad_norm": 0.04026716202497482, "learning_rate": 3.156100391701727e-05, "loss": 0.0022, "step": 18970 }, { "epoch": 2.056699913269731, "grad_norm": 0.28078147768974304, "learning_rate": 3.155737704918033e-05, "loss": 0.0121, "step": 18971 }, { "epoch": 2.056808326105811, "grad_norm": 0.08818759769201279, "learning_rate": 3.15537501813434e-05, "loss": 0.0024, "step": 18972 }, { "epoch": 2.0569167389418905, "grad_norm": 0.05051581934094429, "learning_rate": 3.1550123313506454e-05, "loss": 0.0036, "step": 18973 }, { "epoch": 2.0570251517779705, "grad_norm": 0.19955496490001678, "learning_rate": 3.154649644566952e-05, "loss": 0.0038, "step": 18974 }, { "epoch": 2.0571335646140505, "grad_norm": 0.2079816609621048, "learning_rate": 3.154286957783258e-05, "loss": 0.0094, "step": 18975 }, { "epoch": 2.05724197745013, "grad_norm": 0.2402070313692093, "learning_rate": 3.153924270999565e-05, "loss": 0.0432, "step": 18976 }, { "epoch": 2.05735039028621, "grad_norm": 0.08783865720033646, "learning_rate": 3.153561584215871e-05, "loss": 0.0035, "step": 18977 }, { "epoch": 2.0574588031222896, "grad_norm": 0.255966454744339, "learning_rate": 3.1531988974321775e-05, "loss": 0.0227, "step": 18978 }, { "epoch": 2.0575672159583696, "grad_norm": 0.47732627391815186, "learning_rate": 3.1528362106484846e-05, "loss": 0.0254, "step": 18979 }, { "epoch": 2.057675628794449, "grad_norm": 0.16049544513225555, "learning_rate": 3.152473523864791e-05, "loss": 0.0076, "step": 18980 }, { "epoch": 2.057784041630529, "grad_norm": 0.48897191882133484, "learning_rate": 3.152110837081097e-05, "loss": 0.0092, "step": 18981 }, { "epoch": 2.057892454466609, "grad_norm": 0.2783471345901489, "learning_rate": 3.151748150297403e-05, "loss": 0.0159, "step": 18982 }, { "epoch": 2.0580008673026886, "grad_norm": 0.39767250418663025, "learning_rate": 3.1513854635137095e-05, "loss": 0.0192, "step": 18983 }, { "epoch": 2.0581092801387686, "grad_norm": 0.028486555442214012, "learning_rate": 3.151022776730016e-05, "loss": 0.0007, "step": 18984 }, { "epoch": 2.058217692974848, "grad_norm": 0.1031767800450325, "learning_rate": 3.1506600899463224e-05, "loss": 0.004, "step": 18985 }, { "epoch": 2.058326105810928, "grad_norm": 0.2918100357055664, "learning_rate": 3.150297403162629e-05, "loss": 0.0083, "step": 18986 }, { "epoch": 2.0584345186470077, "grad_norm": 0.044731155037879944, "learning_rate": 3.149934716378936e-05, "loss": 0.001, "step": 18987 }, { "epoch": 2.0585429314830876, "grad_norm": 0.03814701363444328, "learning_rate": 3.1495720295952416e-05, "loss": 0.0018, "step": 18988 }, { "epoch": 2.058651344319167, "grad_norm": 0.11165960878133774, "learning_rate": 3.149209342811548e-05, "loss": 0.0033, "step": 18989 }, { "epoch": 2.058759757155247, "grad_norm": 0.12486186623573303, "learning_rate": 3.1488466560278544e-05, "loss": 0.0021, "step": 18990 }, { "epoch": 2.058868169991327, "grad_norm": 0.13218232989311218, "learning_rate": 3.148483969244161e-05, "loss": 0.0025, "step": 18991 }, { "epoch": 2.0589765828274067, "grad_norm": 0.06167975068092346, "learning_rate": 3.148121282460467e-05, "loss": 0.0032, "step": 18992 }, { "epoch": 2.0590849956634867, "grad_norm": 0.49549177289009094, "learning_rate": 3.1477585956767737e-05, "loss": 0.028, "step": 18993 }, { "epoch": 2.0591934084995662, "grad_norm": 0.36983662843704224, "learning_rate": 3.14739590889308e-05, "loss": 0.0226, "step": 18994 }, { "epoch": 2.059301821335646, "grad_norm": 0.11741405725479126, "learning_rate": 3.1470332221093865e-05, "loss": 0.0044, "step": 18995 }, { "epoch": 2.0594102341717258, "grad_norm": 0.13564951717853546, "learning_rate": 3.146670535325693e-05, "loss": 0.0073, "step": 18996 }, { "epoch": 2.0595186470078057, "grad_norm": 0.6953256726264954, "learning_rate": 3.146307848541999e-05, "loss": 0.0163, "step": 18997 }, { "epoch": 2.0596270598438857, "grad_norm": 0.04246196523308754, "learning_rate": 3.145945161758306e-05, "loss": 0.0011, "step": 18998 }, { "epoch": 2.0597354726799653, "grad_norm": 0.05437842383980751, "learning_rate": 3.145582474974612e-05, "loss": 0.0028, "step": 18999 }, { "epoch": 2.0598438855160452, "grad_norm": 0.6507350206375122, "learning_rate": 3.1452197881909185e-05, "loss": 0.0219, "step": 19000 }, { "epoch": 2.059952298352125, "grad_norm": 0.11948409676551819, "learning_rate": 3.144857101407225e-05, "loss": 0.0048, "step": 19001 }, { "epoch": 2.0600607111882048, "grad_norm": 0.5012694001197815, "learning_rate": 3.1444944146235313e-05, "loss": 0.0347, "step": 19002 }, { "epoch": 2.0601691240242843, "grad_norm": 0.224204421043396, "learning_rate": 3.144131727839838e-05, "loss": 0.014, "step": 19003 }, { "epoch": 2.0602775368603643, "grad_norm": 0.04725182056427002, "learning_rate": 3.143769041056144e-05, "loss": 0.0017, "step": 19004 }, { "epoch": 2.060385949696444, "grad_norm": 0.4737296402454376, "learning_rate": 3.1434063542724506e-05, "loss": 0.0124, "step": 19005 }, { "epoch": 2.060494362532524, "grad_norm": 0.28940892219543457, "learning_rate": 3.143043667488757e-05, "loss": 0.0092, "step": 19006 }, { "epoch": 2.060602775368604, "grad_norm": 0.6289536952972412, "learning_rate": 3.1426809807050634e-05, "loss": 0.0423, "step": 19007 }, { "epoch": 2.0607111882046834, "grad_norm": 0.4232667088508606, "learning_rate": 3.14231829392137e-05, "loss": 0.0122, "step": 19008 }, { "epoch": 2.0608196010407633, "grad_norm": 0.12901495397090912, "learning_rate": 3.141955607137676e-05, "loss": 0.0011, "step": 19009 }, { "epoch": 2.060928013876843, "grad_norm": 0.3784874975681305, "learning_rate": 3.1415929203539826e-05, "loss": 0.0076, "step": 19010 }, { "epoch": 2.061036426712923, "grad_norm": 0.041531212627887726, "learning_rate": 3.1412302335702884e-05, "loss": 0.0012, "step": 19011 }, { "epoch": 2.0611448395490024, "grad_norm": 0.7476817965507507, "learning_rate": 3.140867546786595e-05, "loss": 0.0149, "step": 19012 }, { "epoch": 2.0612532523850824, "grad_norm": 0.11500242352485657, "learning_rate": 3.140504860002902e-05, "loss": 0.0062, "step": 19013 }, { "epoch": 2.0613616652211624, "grad_norm": 0.15036888420581818, "learning_rate": 3.140142173219208e-05, "loss": 0.0102, "step": 19014 }, { "epoch": 2.061470078057242, "grad_norm": 0.3199011981487274, "learning_rate": 3.139779486435515e-05, "loss": 0.0188, "step": 19015 }, { "epoch": 2.061578490893322, "grad_norm": 0.2927398383617401, "learning_rate": 3.139416799651821e-05, "loss": 0.0153, "step": 19016 }, { "epoch": 2.0616869037294014, "grad_norm": 0.34226319193840027, "learning_rate": 3.1390541128681275e-05, "loss": 0.0227, "step": 19017 }, { "epoch": 2.0617953165654814, "grad_norm": 0.1604640930891037, "learning_rate": 3.138691426084433e-05, "loss": 0.009, "step": 19018 }, { "epoch": 2.061903729401561, "grad_norm": 0.25846245884895325, "learning_rate": 3.1383287393007396e-05, "loss": 0.011, "step": 19019 }, { "epoch": 2.062012142237641, "grad_norm": 0.44609156250953674, "learning_rate": 3.137966052517046e-05, "loss": 0.0298, "step": 19020 }, { "epoch": 2.062120555073721, "grad_norm": 0.07619615644216537, "learning_rate": 3.1376033657333525e-05, "loss": 0.0046, "step": 19021 }, { "epoch": 2.0622289679098005, "grad_norm": 0.5681552886962891, "learning_rate": 3.1372406789496596e-05, "loss": 0.0383, "step": 19022 }, { "epoch": 2.0623373807458805, "grad_norm": 0.2726024389266968, "learning_rate": 3.136877992165966e-05, "loss": 0.0175, "step": 19023 }, { "epoch": 2.06244579358196, "grad_norm": 0.08237221837043762, "learning_rate": 3.1365153053822724e-05, "loss": 0.0029, "step": 19024 }, { "epoch": 2.06255420641804, "grad_norm": 0.41843533515930176, "learning_rate": 3.136152618598579e-05, "loss": 0.0161, "step": 19025 }, { "epoch": 2.0626626192541195, "grad_norm": 0.39585795998573303, "learning_rate": 3.1357899318148845e-05, "loss": 0.0198, "step": 19026 }, { "epoch": 2.0627710320901995, "grad_norm": 0.10567224025726318, "learning_rate": 3.135427245031191e-05, "loss": 0.005, "step": 19027 }, { "epoch": 2.062879444926279, "grad_norm": 0.018482089042663574, "learning_rate": 3.1350645582474973e-05, "loss": 0.0005, "step": 19028 }, { "epoch": 2.062987857762359, "grad_norm": 0.803825855255127, "learning_rate": 3.134701871463804e-05, "loss": 0.0072, "step": 19029 }, { "epoch": 2.063096270598439, "grad_norm": 0.38803717494010925, "learning_rate": 3.134339184680111e-05, "loss": 0.0205, "step": 19030 }, { "epoch": 2.0632046834345186, "grad_norm": 0.5481421947479248, "learning_rate": 3.133976497896417e-05, "loss": 0.0172, "step": 19031 }, { "epoch": 2.0633130962705986, "grad_norm": 0.17730091512203217, "learning_rate": 3.133613811112724e-05, "loss": 0.0202, "step": 19032 }, { "epoch": 2.063421509106678, "grad_norm": 0.45739153027534485, "learning_rate": 3.1332511243290294e-05, "loss": 0.0215, "step": 19033 }, { "epoch": 2.063529921942758, "grad_norm": 0.28666234016418457, "learning_rate": 3.132888437545336e-05, "loss": 0.0321, "step": 19034 }, { "epoch": 2.0636383347788376, "grad_norm": 0.20470190048217773, "learning_rate": 3.132525750761642e-05, "loss": 0.0093, "step": 19035 }, { "epoch": 2.0637467476149176, "grad_norm": 0.058378204703330994, "learning_rate": 3.1321630639779486e-05, "loss": 0.0034, "step": 19036 }, { "epoch": 2.0638551604509976, "grad_norm": 0.10123511403799057, "learning_rate": 3.131800377194255e-05, "loss": 0.0024, "step": 19037 }, { "epoch": 2.063963573287077, "grad_norm": 0.07810567319393158, "learning_rate": 3.1314376904105614e-05, "loss": 0.0044, "step": 19038 }, { "epoch": 2.064071986123157, "grad_norm": 0.2620231509208679, "learning_rate": 3.1310750036268685e-05, "loss": 0.0052, "step": 19039 }, { "epoch": 2.0641803989592367, "grad_norm": 0.025640128180384636, "learning_rate": 3.130712316843174e-05, "loss": 0.0007, "step": 19040 }, { "epoch": 2.0642888117953166, "grad_norm": 0.03282734379172325, "learning_rate": 3.130349630059481e-05, "loss": 0.0018, "step": 19041 }, { "epoch": 2.064397224631396, "grad_norm": 0.38304805755615234, "learning_rate": 3.129986943275787e-05, "loss": 0.0181, "step": 19042 }, { "epoch": 2.064505637467476, "grad_norm": 0.2170393317937851, "learning_rate": 3.1296242564920935e-05, "loss": 0.0191, "step": 19043 }, { "epoch": 2.0646140503035557, "grad_norm": 0.18715034425258636, "learning_rate": 3.1292615697084e-05, "loss": 0.0093, "step": 19044 }, { "epoch": 2.0647224631396357, "grad_norm": 0.032609425485134125, "learning_rate": 3.128898882924706e-05, "loss": 0.0011, "step": 19045 }, { "epoch": 2.0648308759757157, "grad_norm": 0.17312973737716675, "learning_rate": 3.128536196141013e-05, "loss": 0.0113, "step": 19046 }, { "epoch": 2.0649392888117952, "grad_norm": 0.016110224649310112, "learning_rate": 3.128173509357319e-05, "loss": 0.0005, "step": 19047 }, { "epoch": 2.065047701647875, "grad_norm": 0.2070401906967163, "learning_rate": 3.1278108225736256e-05, "loss": 0.0176, "step": 19048 }, { "epoch": 2.0651561144839548, "grad_norm": 0.37303993105888367, "learning_rate": 3.127448135789932e-05, "loss": 0.0177, "step": 19049 }, { "epoch": 2.0652645273200347, "grad_norm": 0.05091094970703125, "learning_rate": 3.1270854490062384e-05, "loss": 0.0017, "step": 19050 }, { "epoch": 2.0653729401561143, "grad_norm": 0.9539963603019714, "learning_rate": 3.126722762222545e-05, "loss": 0.0389, "step": 19051 }, { "epoch": 2.0654813529921943, "grad_norm": 0.11771805584430695, "learning_rate": 3.126360075438851e-05, "loss": 0.0039, "step": 19052 }, { "epoch": 2.0655897658282742, "grad_norm": 0.049142178148031235, "learning_rate": 3.1259973886551576e-05, "loss": 0.0016, "step": 19053 }, { "epoch": 2.065698178664354, "grad_norm": 0.3602350950241089, "learning_rate": 3.125634701871464e-05, "loss": 0.0114, "step": 19054 }, { "epoch": 2.0658065915004338, "grad_norm": 0.16239455342292786, "learning_rate": 3.1252720150877704e-05, "loss": 0.0089, "step": 19055 }, { "epoch": 2.0659150043365133, "grad_norm": 0.004090330097824335, "learning_rate": 3.124909328304077e-05, "loss": 0.0003, "step": 19056 }, { "epoch": 2.0660234171725933, "grad_norm": 0.06136186420917511, "learning_rate": 3.124546641520383e-05, "loss": 0.0028, "step": 19057 }, { "epoch": 2.066131830008673, "grad_norm": 0.9096423387527466, "learning_rate": 3.1241839547366897e-05, "loss": 0.0189, "step": 19058 }, { "epoch": 2.066240242844753, "grad_norm": 0.31967946887016296, "learning_rate": 3.123821267952996e-05, "loss": 0.0196, "step": 19059 }, { "epoch": 2.066348655680833, "grad_norm": 0.0746348574757576, "learning_rate": 3.1234585811693025e-05, "loss": 0.0014, "step": 19060 }, { "epoch": 2.0664570685169124, "grad_norm": 0.437886506319046, "learning_rate": 3.123095894385609e-05, "loss": 0.0135, "step": 19061 }, { "epoch": 2.0665654813529923, "grad_norm": 0.10457199811935425, "learning_rate": 3.122733207601915e-05, "loss": 0.0013, "step": 19062 }, { "epoch": 2.066673894189072, "grad_norm": 0.18989133834838867, "learning_rate": 3.122370520818221e-05, "loss": 0.0111, "step": 19063 }, { "epoch": 2.066782307025152, "grad_norm": 0.3596336245536804, "learning_rate": 3.1220078340345274e-05, "loss": 0.0693, "step": 19064 }, { "epoch": 2.0668907198612314, "grad_norm": 0.23664245009422302, "learning_rate": 3.1216451472508345e-05, "loss": 0.0072, "step": 19065 }, { "epoch": 2.0669991326973114, "grad_norm": 0.027301784604787827, "learning_rate": 3.121282460467141e-05, "loss": 0.0009, "step": 19066 }, { "epoch": 2.067107545533391, "grad_norm": 0.03289651498198509, "learning_rate": 3.1209197736834474e-05, "loss": 0.0013, "step": 19067 }, { "epoch": 2.067215958369471, "grad_norm": 0.13502825796604156, "learning_rate": 3.120557086899754e-05, "loss": 0.0109, "step": 19068 }, { "epoch": 2.067324371205551, "grad_norm": 0.1017226055264473, "learning_rate": 3.12019440011606e-05, "loss": 0.0054, "step": 19069 }, { "epoch": 2.0674327840416304, "grad_norm": 0.2564317286014557, "learning_rate": 3.119831713332366e-05, "loss": 0.0067, "step": 19070 }, { "epoch": 2.0675411968777104, "grad_norm": 0.33744150400161743, "learning_rate": 3.119469026548672e-05, "loss": 0.0084, "step": 19071 }, { "epoch": 2.06764960971379, "grad_norm": 0.113357312977314, "learning_rate": 3.119106339764979e-05, "loss": 0.005, "step": 19072 }, { "epoch": 2.06775802254987, "grad_norm": 0.513304591178894, "learning_rate": 3.118743652981285e-05, "loss": 0.0166, "step": 19073 }, { "epoch": 2.0678664353859495, "grad_norm": 0.14314813911914825, "learning_rate": 3.118380966197592e-05, "loss": 0.0225, "step": 19074 }, { "epoch": 2.0679748482220295, "grad_norm": 0.42210468649864197, "learning_rate": 3.1180182794138986e-05, "loss": 0.0178, "step": 19075 }, { "epoch": 2.0680832610581095, "grad_norm": 0.5732184648513794, "learning_rate": 3.117655592630205e-05, "loss": 0.0311, "step": 19076 }, { "epoch": 2.068191673894189, "grad_norm": 0.6075601577758789, "learning_rate": 3.1172929058465115e-05, "loss": 0.0253, "step": 19077 }, { "epoch": 2.068300086730269, "grad_norm": 0.04537966102361679, "learning_rate": 3.116930219062817e-05, "loss": 0.0021, "step": 19078 }, { "epoch": 2.0684084995663485, "grad_norm": 0.22682437300682068, "learning_rate": 3.1165675322791236e-05, "loss": 0.0049, "step": 19079 }, { "epoch": 2.0685169124024285, "grad_norm": 0.9483124613761902, "learning_rate": 3.11620484549543e-05, "loss": 0.0274, "step": 19080 }, { "epoch": 2.068625325238508, "grad_norm": 0.068853460252285, "learning_rate": 3.1158421587117364e-05, "loss": 0.0025, "step": 19081 }, { "epoch": 2.068733738074588, "grad_norm": 0.6654665470123291, "learning_rate": 3.1154794719280435e-05, "loss": 0.011, "step": 19082 }, { "epoch": 2.068842150910668, "grad_norm": 0.042714402079582214, "learning_rate": 3.11511678514435e-05, "loss": 0.0017, "step": 19083 }, { "epoch": 2.0689505637467476, "grad_norm": 0.3681472837924957, "learning_rate": 3.114754098360656e-05, "loss": 0.0423, "step": 19084 }, { "epoch": 2.0690589765828276, "grad_norm": 0.21091994643211365, "learning_rate": 3.114391411576962e-05, "loss": 0.006, "step": 19085 }, { "epoch": 2.069167389418907, "grad_norm": 0.08564717322587967, "learning_rate": 3.1140287247932685e-05, "loss": 0.0089, "step": 19086 }, { "epoch": 2.069275802254987, "grad_norm": 0.05007481947541237, "learning_rate": 3.113666038009575e-05, "loss": 0.002, "step": 19087 }, { "epoch": 2.0693842150910666, "grad_norm": 0.32004430890083313, "learning_rate": 3.113303351225881e-05, "loss": 0.018, "step": 19088 }, { "epoch": 2.0694926279271466, "grad_norm": 0.1622927188873291, "learning_rate": 3.112940664442188e-05, "loss": 0.009, "step": 19089 }, { "epoch": 2.069601040763226, "grad_norm": 0.07896365970373154, "learning_rate": 3.112577977658494e-05, "loss": 0.0021, "step": 19090 }, { "epoch": 2.069709453599306, "grad_norm": 0.22126920521259308, "learning_rate": 3.112215290874801e-05, "loss": 0.0161, "step": 19091 }, { "epoch": 2.069817866435386, "grad_norm": 0.050579361617565155, "learning_rate": 3.111852604091107e-05, "loss": 0.0044, "step": 19092 }, { "epoch": 2.0699262792714657, "grad_norm": 0.46171125769615173, "learning_rate": 3.1114899173074133e-05, "loss": 0.0203, "step": 19093 }, { "epoch": 2.0700346921075456, "grad_norm": 0.0641867145895958, "learning_rate": 3.11112723052372e-05, "loss": 0.0025, "step": 19094 }, { "epoch": 2.070143104943625, "grad_norm": 0.12698054313659668, "learning_rate": 3.110764543740026e-05, "loss": 0.0042, "step": 19095 }, { "epoch": 2.070251517779705, "grad_norm": 0.18860290944576263, "learning_rate": 3.1104018569563326e-05, "loss": 0.0065, "step": 19096 }, { "epoch": 2.0703599306157847, "grad_norm": 0.38759925961494446, "learning_rate": 3.110039170172639e-05, "loss": 0.0189, "step": 19097 }, { "epoch": 2.0704683434518647, "grad_norm": 0.04376931115984917, "learning_rate": 3.1096764833889454e-05, "loss": 0.0025, "step": 19098 }, { "epoch": 2.0705767562879447, "grad_norm": 0.2293529063463211, "learning_rate": 3.109313796605252e-05, "loss": 0.0063, "step": 19099 }, { "epoch": 2.0706851691240242, "grad_norm": 0.08291628211736679, "learning_rate": 3.108951109821558e-05, "loss": 0.0042, "step": 19100 }, { "epoch": 2.070793581960104, "grad_norm": 0.2770484387874603, "learning_rate": 3.1085884230378646e-05, "loss": 0.0106, "step": 19101 }, { "epoch": 2.0709019947961838, "grad_norm": 0.0651158019900322, "learning_rate": 3.108225736254171e-05, "loss": 0.0014, "step": 19102 }, { "epoch": 2.0710104076322637, "grad_norm": 0.5721139907836914, "learning_rate": 3.1078630494704775e-05, "loss": 0.0189, "step": 19103 }, { "epoch": 2.0711188204683433, "grad_norm": 0.2765270471572876, "learning_rate": 3.107500362686784e-05, "loss": 0.0145, "step": 19104 }, { "epoch": 2.0712272333044233, "grad_norm": 0.09543820470571518, "learning_rate": 3.10713767590309e-05, "loss": 0.0019, "step": 19105 }, { "epoch": 2.0713356461405033, "grad_norm": 0.05532454699277878, "learning_rate": 3.106774989119397e-05, "loss": 0.0024, "step": 19106 }, { "epoch": 2.071444058976583, "grad_norm": 0.27497902512550354, "learning_rate": 3.106412302335703e-05, "loss": 0.0113, "step": 19107 }, { "epoch": 2.0715524718126628, "grad_norm": 1.2888370752334595, "learning_rate": 3.1060496155520095e-05, "loss": 0.042, "step": 19108 }, { "epoch": 2.0716608846487423, "grad_norm": 1.2140417098999023, "learning_rate": 3.105686928768316e-05, "loss": 0.0217, "step": 19109 }, { "epoch": 2.0717692974848223, "grad_norm": 0.06824146211147308, "learning_rate": 3.105324241984622e-05, "loss": 0.0058, "step": 19110 }, { "epoch": 2.071877710320902, "grad_norm": 0.6490814089775085, "learning_rate": 3.104961555200929e-05, "loss": 0.0499, "step": 19111 }, { "epoch": 2.071986123156982, "grad_norm": 0.09271205216646194, "learning_rate": 3.104598868417235e-05, "loss": 0.0026, "step": 19112 }, { "epoch": 2.0720945359930614, "grad_norm": 0.02009536884725094, "learning_rate": 3.1042361816335416e-05, "loss": 0.0008, "step": 19113 }, { "epoch": 2.0722029488291414, "grad_norm": 0.46419721841812134, "learning_rate": 3.103873494849848e-05, "loss": 0.0323, "step": 19114 }, { "epoch": 2.0723113616652213, "grad_norm": 0.25556349754333496, "learning_rate": 3.103510808066154e-05, "loss": 0.0106, "step": 19115 }, { "epoch": 2.072419774501301, "grad_norm": 0.6117364168167114, "learning_rate": 3.10314812128246e-05, "loss": 0.0248, "step": 19116 }, { "epoch": 2.072528187337381, "grad_norm": 0.26563751697540283, "learning_rate": 3.102785434498767e-05, "loss": 0.0368, "step": 19117 }, { "epoch": 2.0726366001734604, "grad_norm": 0.1539796143770218, "learning_rate": 3.1024227477150736e-05, "loss": 0.0092, "step": 19118 }, { "epoch": 2.0727450130095404, "grad_norm": 0.283194363117218, "learning_rate": 3.10206006093138e-05, "loss": 0.0049, "step": 19119 }, { "epoch": 2.07285342584562, "grad_norm": 0.34492480754852295, "learning_rate": 3.1016973741476864e-05, "loss": 0.0135, "step": 19120 }, { "epoch": 2.0729618386817, "grad_norm": 0.18742242455482483, "learning_rate": 3.101334687363993e-05, "loss": 0.0038, "step": 19121 }, { "epoch": 2.07307025151778, "grad_norm": 0.8632857799530029, "learning_rate": 3.1009720005802986e-05, "loss": 0.077, "step": 19122 }, { "epoch": 2.0731786643538594, "grad_norm": 0.2708989381790161, "learning_rate": 3.100609313796605e-05, "loss": 0.005, "step": 19123 }, { "epoch": 2.0732870771899394, "grad_norm": 0.35209378600120544, "learning_rate": 3.1002466270129114e-05, "loss": 0.014, "step": 19124 }, { "epoch": 2.073395490026019, "grad_norm": 0.5156334638595581, "learning_rate": 3.0998839402292185e-05, "loss": 0.0232, "step": 19125 }, { "epoch": 2.073503902862099, "grad_norm": 0.05312040448188782, "learning_rate": 3.099521253445525e-05, "loss": 0.0038, "step": 19126 }, { "epoch": 2.0736123156981785, "grad_norm": 0.14966320991516113, "learning_rate": 3.099158566661831e-05, "loss": 0.0672, "step": 19127 }, { "epoch": 2.0737207285342585, "grad_norm": 0.1629396229982376, "learning_rate": 3.098795879878138e-05, "loss": 0.0069, "step": 19128 }, { "epoch": 2.0738291413703385, "grad_norm": 0.20172888040542603, "learning_rate": 3.098433193094444e-05, "loss": 0.0043, "step": 19129 }, { "epoch": 2.073937554206418, "grad_norm": 0.12365494668483734, "learning_rate": 3.09807050631075e-05, "loss": 0.008, "step": 19130 }, { "epoch": 2.074045967042498, "grad_norm": 0.0934762954711914, "learning_rate": 3.097707819527056e-05, "loss": 0.0031, "step": 19131 }, { "epoch": 2.0741543798785775, "grad_norm": 0.39575737714767456, "learning_rate": 3.097345132743363e-05, "loss": 0.0174, "step": 19132 }, { "epoch": 2.0742627927146575, "grad_norm": 0.37200331687927246, "learning_rate": 3.096982445959669e-05, "loss": 0.0388, "step": 19133 }, { "epoch": 2.074371205550737, "grad_norm": 0.412841260433197, "learning_rate": 3.096619759175976e-05, "loss": 0.0114, "step": 19134 }, { "epoch": 2.074479618386817, "grad_norm": 0.32551392912864685, "learning_rate": 3.0962570723922826e-05, "loss": 0.0176, "step": 19135 }, { "epoch": 2.0745880312228966, "grad_norm": 0.4107619524002075, "learning_rate": 3.095894385608589e-05, "loss": 0.0304, "step": 19136 }, { "epoch": 2.0746964440589766, "grad_norm": 0.008635983802378178, "learning_rate": 3.095531698824895e-05, "loss": 0.0004, "step": 19137 }, { "epoch": 2.0748048568950566, "grad_norm": 0.1535593718290329, "learning_rate": 3.095169012041201e-05, "loss": 0.0042, "step": 19138 }, { "epoch": 2.074913269731136, "grad_norm": 0.13016675412654877, "learning_rate": 3.0948063252575076e-05, "loss": 0.0071, "step": 19139 }, { "epoch": 2.075021682567216, "grad_norm": 0.1239510178565979, "learning_rate": 3.094443638473814e-05, "loss": 0.0115, "step": 19140 }, { "epoch": 2.0751300954032956, "grad_norm": 0.14493529498577118, "learning_rate": 3.0940809516901204e-05, "loss": 0.0062, "step": 19141 }, { "epoch": 2.0752385082393756, "grad_norm": 0.0664249062538147, "learning_rate": 3.093718264906427e-05, "loss": 0.0017, "step": 19142 }, { "epoch": 2.075346921075455, "grad_norm": 0.17406700551509857, "learning_rate": 3.093355578122734e-05, "loss": 0.0082, "step": 19143 }, { "epoch": 2.075455333911535, "grad_norm": 0.7285809516906738, "learning_rate": 3.0929928913390396e-05, "loss": 0.0187, "step": 19144 }, { "epoch": 2.075563746747615, "grad_norm": 0.8849684596061707, "learning_rate": 3.092630204555346e-05, "loss": 0.0301, "step": 19145 }, { "epoch": 2.0756721595836947, "grad_norm": 0.4709831774234772, "learning_rate": 3.0922675177716524e-05, "loss": 0.0262, "step": 19146 }, { "epoch": 2.0757805724197746, "grad_norm": 0.6440481543540955, "learning_rate": 3.091904830987959e-05, "loss": 0.0068, "step": 19147 }, { "epoch": 2.075888985255854, "grad_norm": 0.02843373641371727, "learning_rate": 3.091542144204265e-05, "loss": 0.0012, "step": 19148 }, { "epoch": 2.075997398091934, "grad_norm": 0.3919008672237396, "learning_rate": 3.0911794574205717e-05, "loss": 0.0125, "step": 19149 }, { "epoch": 2.0761058109280137, "grad_norm": 0.1815161556005478, "learning_rate": 3.090816770636878e-05, "loss": 0.0031, "step": 19150 }, { "epoch": 2.0762142237640937, "grad_norm": 0.34086012840270996, "learning_rate": 3.0904540838531845e-05, "loss": 0.0045, "step": 19151 }, { "epoch": 2.0763226366001737, "grad_norm": 0.2698257863521576, "learning_rate": 3.090091397069491e-05, "loss": 0.0216, "step": 19152 }, { "epoch": 2.0764310494362532, "grad_norm": 0.20524552464485168, "learning_rate": 3.089728710285797e-05, "loss": 0.011, "step": 19153 }, { "epoch": 2.076539462272333, "grad_norm": 0.0829937607049942, "learning_rate": 3.089366023502104e-05, "loss": 0.0027, "step": 19154 }, { "epoch": 2.0766478751084128, "grad_norm": 0.022320792078971863, "learning_rate": 3.08900333671841e-05, "loss": 0.0006, "step": 19155 }, { "epoch": 2.0767562879444927, "grad_norm": 0.08633539825677872, "learning_rate": 3.0886406499347165e-05, "loss": 0.0048, "step": 19156 }, { "epoch": 2.0768647007805723, "grad_norm": 0.44707557559013367, "learning_rate": 3.088277963151023e-05, "loss": 0.026, "step": 19157 }, { "epoch": 2.0769731136166523, "grad_norm": 0.08101548999547958, "learning_rate": 3.0879152763673294e-05, "loss": 0.0027, "step": 19158 }, { "epoch": 2.077081526452732, "grad_norm": 0.20078444480895996, "learning_rate": 3.087552589583636e-05, "loss": 0.0075, "step": 19159 }, { "epoch": 2.077189939288812, "grad_norm": 0.06476140767335892, "learning_rate": 3.087189902799942e-05, "loss": 0.0027, "step": 19160 }, { "epoch": 2.077298352124892, "grad_norm": 0.05546550452709198, "learning_rate": 3.0868272160162486e-05, "loss": 0.0018, "step": 19161 }, { "epoch": 2.0774067649609713, "grad_norm": 0.6418185830116272, "learning_rate": 3.086464529232555e-05, "loss": 0.0434, "step": 19162 }, { "epoch": 2.0775151777970513, "grad_norm": 0.400517076253891, "learning_rate": 3.0861018424488614e-05, "loss": 0.0035, "step": 19163 }, { "epoch": 2.077623590633131, "grad_norm": 0.2071353942155838, "learning_rate": 3.085739155665168e-05, "loss": 0.0154, "step": 19164 }, { "epoch": 2.077732003469211, "grad_norm": 0.3283311724662781, "learning_rate": 3.085376468881474e-05, "loss": 0.0249, "step": 19165 }, { "epoch": 2.0778404163052904, "grad_norm": 0.2861040234565735, "learning_rate": 3.0850137820977806e-05, "loss": 0.0122, "step": 19166 }, { "epoch": 2.0779488291413704, "grad_norm": 0.3268243372440338, "learning_rate": 3.0846510953140864e-05, "loss": 0.02, "step": 19167 }, { "epoch": 2.0780572419774503, "grad_norm": 0.12577280402183533, "learning_rate": 3.084288408530393e-05, "loss": 0.007, "step": 19168 }, { "epoch": 2.07816565481353, "grad_norm": 0.07959456741809845, "learning_rate": 3.0839257217467e-05, "loss": 0.0021, "step": 19169 }, { "epoch": 2.07827406764961, "grad_norm": 0.015904797241091728, "learning_rate": 3.083563034963006e-05, "loss": 0.0005, "step": 19170 }, { "epoch": 2.0783824804856894, "grad_norm": 0.057382166385650635, "learning_rate": 3.083200348179313e-05, "loss": 0.0024, "step": 19171 }, { "epoch": 2.0784908933217694, "grad_norm": 0.483988881111145, "learning_rate": 3.082837661395619e-05, "loss": 0.0089, "step": 19172 }, { "epoch": 2.078599306157849, "grad_norm": 0.4376314580440521, "learning_rate": 3.0824749746119255e-05, "loss": 0.055, "step": 19173 }, { "epoch": 2.078707718993929, "grad_norm": 0.2578575909137726, "learning_rate": 3.082112287828231e-05, "loss": 0.011, "step": 19174 }, { "epoch": 2.0788161318300085, "grad_norm": 0.05700714886188507, "learning_rate": 3.0817496010445377e-05, "loss": 0.0022, "step": 19175 }, { "epoch": 2.0789245446660884, "grad_norm": 0.4914894700050354, "learning_rate": 3.081386914260844e-05, "loss": 0.018, "step": 19176 }, { "epoch": 2.0790329575021684, "grad_norm": 0.25390753149986267, "learning_rate": 3.081024227477151e-05, "loss": 0.0103, "step": 19177 }, { "epoch": 2.079141370338248, "grad_norm": 0.10024009644985199, "learning_rate": 3.0806615406934576e-05, "loss": 0.0032, "step": 19178 }, { "epoch": 2.079249783174328, "grad_norm": 0.2567968964576721, "learning_rate": 3.080298853909764e-05, "loss": 0.0058, "step": 19179 }, { "epoch": 2.0793581960104075, "grad_norm": 0.5099648237228394, "learning_rate": 3.0799361671260704e-05, "loss": 0.0151, "step": 19180 }, { "epoch": 2.0794666088464875, "grad_norm": 0.22643707692623138, "learning_rate": 3.079573480342377e-05, "loss": 0.0106, "step": 19181 }, { "epoch": 2.079575021682567, "grad_norm": 0.1426687240600586, "learning_rate": 3.0792107935586825e-05, "loss": 0.011, "step": 19182 }, { "epoch": 2.079683434518647, "grad_norm": 0.37731581926345825, "learning_rate": 3.078848106774989e-05, "loss": 0.0127, "step": 19183 }, { "epoch": 2.079791847354727, "grad_norm": 0.27969232201576233, "learning_rate": 3.0784854199912953e-05, "loss": 0.0085, "step": 19184 }, { "epoch": 2.0799002601908065, "grad_norm": 0.15823650360107422, "learning_rate": 3.078122733207602e-05, "loss": 0.0079, "step": 19185 }, { "epoch": 2.0800086730268865, "grad_norm": 0.6863232254981995, "learning_rate": 3.077760046423909e-05, "loss": 0.0188, "step": 19186 }, { "epoch": 2.080117085862966, "grad_norm": 0.0985613539814949, "learning_rate": 3.077397359640215e-05, "loss": 0.0042, "step": 19187 }, { "epoch": 2.080225498699046, "grad_norm": 0.37918397784233093, "learning_rate": 3.077034672856522e-05, "loss": 0.0272, "step": 19188 }, { "epoch": 2.0803339115351256, "grad_norm": 2.2308223247528076, "learning_rate": 3.0766719860728274e-05, "loss": 0.0352, "step": 19189 }, { "epoch": 2.0804423243712056, "grad_norm": 0.009519235230982304, "learning_rate": 3.076309299289134e-05, "loss": 0.0004, "step": 19190 }, { "epoch": 2.080550737207285, "grad_norm": 0.7469504475593567, "learning_rate": 3.07594661250544e-05, "loss": 0.0127, "step": 19191 }, { "epoch": 2.080659150043365, "grad_norm": 0.4843592941761017, "learning_rate": 3.0755839257217466e-05, "loss": 0.0114, "step": 19192 }, { "epoch": 2.080767562879445, "grad_norm": 0.16885049641132355, "learning_rate": 3.075221238938053e-05, "loss": 0.0034, "step": 19193 }, { "epoch": 2.0808759757155246, "grad_norm": 0.2142779678106308, "learning_rate": 3.07485855215436e-05, "loss": 0.0041, "step": 19194 }, { "epoch": 2.0809843885516046, "grad_norm": 0.08412124961614609, "learning_rate": 3.0744958653706665e-05, "loss": 0.0032, "step": 19195 }, { "epoch": 2.081092801387684, "grad_norm": 0.12360456585884094, "learning_rate": 3.074133178586972e-05, "loss": 0.0033, "step": 19196 }, { "epoch": 2.081201214223764, "grad_norm": 0.16752393543720245, "learning_rate": 3.073770491803279e-05, "loss": 0.0071, "step": 19197 }, { "epoch": 2.0813096270598437, "grad_norm": 0.12978163361549377, "learning_rate": 3.073407805019585e-05, "loss": 0.0568, "step": 19198 }, { "epoch": 2.0814180398959237, "grad_norm": 0.229604110121727, "learning_rate": 3.0730451182358915e-05, "loss": 0.0201, "step": 19199 }, { "epoch": 2.0815264527320037, "grad_norm": 1.2975919246673584, "learning_rate": 3.072682431452198e-05, "loss": 0.0211, "step": 19200 }, { "epoch": 2.081634865568083, "grad_norm": 0.09284891933202744, "learning_rate": 3.072319744668504e-05, "loss": 0.0017, "step": 19201 }, { "epoch": 2.081743278404163, "grad_norm": 0.1570577621459961, "learning_rate": 3.071957057884811e-05, "loss": 0.0053, "step": 19202 }, { "epoch": 2.0818516912402427, "grad_norm": 0.22183220088481903, "learning_rate": 3.071594371101117e-05, "loss": 0.0029, "step": 19203 }, { "epoch": 2.0819601040763227, "grad_norm": 0.06654176115989685, "learning_rate": 3.0712316843174236e-05, "loss": 0.0028, "step": 19204 }, { "epoch": 2.0820685169124022, "grad_norm": 0.203951895236969, "learning_rate": 3.07086899753373e-05, "loss": 0.009, "step": 19205 }, { "epoch": 2.0821769297484822, "grad_norm": 0.17629967629909515, "learning_rate": 3.0705063107500364e-05, "loss": 0.0097, "step": 19206 }, { "epoch": 2.082285342584562, "grad_norm": 0.18690714240074158, "learning_rate": 3.070143623966343e-05, "loss": 0.0087, "step": 19207 }, { "epoch": 2.0823937554206418, "grad_norm": 0.00967979896813631, "learning_rate": 3.069780937182649e-05, "loss": 0.0005, "step": 19208 }, { "epoch": 2.0825021682567217, "grad_norm": 0.3675827085971832, "learning_rate": 3.0694182503989556e-05, "loss": 0.0147, "step": 19209 }, { "epoch": 2.0826105810928013, "grad_norm": 0.10727719962596893, "learning_rate": 3.069055563615262e-05, "loss": 0.0072, "step": 19210 }, { "epoch": 2.0827189939288813, "grad_norm": 0.18568336963653564, "learning_rate": 3.0686928768315684e-05, "loss": 0.0065, "step": 19211 }, { "epoch": 2.082827406764961, "grad_norm": 0.4694843590259552, "learning_rate": 3.068330190047875e-05, "loss": 0.0303, "step": 19212 }, { "epoch": 2.082935819601041, "grad_norm": 0.25289493799209595, "learning_rate": 3.067967503264181e-05, "loss": 0.0116, "step": 19213 }, { "epoch": 2.0830442324371203, "grad_norm": 0.22451956570148468, "learning_rate": 3.067604816480488e-05, "loss": 0.0121, "step": 19214 }, { "epoch": 2.0831526452732003, "grad_norm": 0.03911159560084343, "learning_rate": 3.067242129696794e-05, "loss": 0.0027, "step": 19215 }, { "epoch": 2.0832610581092803, "grad_norm": 0.47841379046440125, "learning_rate": 3.0668794429131005e-05, "loss": 0.0152, "step": 19216 }, { "epoch": 2.08336947094536, "grad_norm": 0.2437903732061386, "learning_rate": 3.066516756129407e-05, "loss": 0.0106, "step": 19217 }, { "epoch": 2.08347788378144, "grad_norm": 0.007414733525365591, "learning_rate": 3.066154069345713e-05, "loss": 0.0002, "step": 19218 }, { "epoch": 2.0835862966175194, "grad_norm": 0.08569814264774323, "learning_rate": 3.065791382562019e-05, "loss": 0.0043, "step": 19219 }, { "epoch": 2.0836947094535994, "grad_norm": 0.072870634496212, "learning_rate": 3.065428695778326e-05, "loss": 0.0024, "step": 19220 }, { "epoch": 2.083803122289679, "grad_norm": 0.5890456438064575, "learning_rate": 3.0650660089946325e-05, "loss": 0.0194, "step": 19221 }, { "epoch": 2.083911535125759, "grad_norm": 0.059195615351200104, "learning_rate": 3.064703322210939e-05, "loss": 0.0031, "step": 19222 }, { "epoch": 2.084019947961839, "grad_norm": 0.2081974595785141, "learning_rate": 3.0643406354272454e-05, "loss": 0.015, "step": 19223 }, { "epoch": 2.0841283607979184, "grad_norm": 0.07032988220453262, "learning_rate": 3.063977948643552e-05, "loss": 0.0032, "step": 19224 }, { "epoch": 2.0842367736339984, "grad_norm": 0.26460254192352295, "learning_rate": 3.063615261859858e-05, "loss": 0.0265, "step": 19225 }, { "epoch": 2.084345186470078, "grad_norm": 0.49707111716270447, "learning_rate": 3.0632525750761646e-05, "loss": 0.0065, "step": 19226 }, { "epoch": 2.084453599306158, "grad_norm": 0.46920907497406006, "learning_rate": 3.06288988829247e-05, "loss": 0.0173, "step": 19227 }, { "epoch": 2.0845620121422375, "grad_norm": 0.004590631928294897, "learning_rate": 3.062527201508777e-05, "loss": 0.0003, "step": 19228 }, { "epoch": 2.0846704249783174, "grad_norm": 0.7918078303337097, "learning_rate": 3.062164514725084e-05, "loss": 0.0365, "step": 19229 }, { "epoch": 2.0847788378143974, "grad_norm": 0.03899264708161354, "learning_rate": 3.06180182794139e-05, "loss": 0.0021, "step": 19230 }, { "epoch": 2.084887250650477, "grad_norm": 0.7568074464797974, "learning_rate": 3.0614391411576966e-05, "loss": 0.0594, "step": 19231 }, { "epoch": 2.084995663486557, "grad_norm": 0.43001705408096313, "learning_rate": 3.061076454374003e-05, "loss": 0.0173, "step": 19232 }, { "epoch": 2.0851040763226365, "grad_norm": 0.13658227026462555, "learning_rate": 3.0607137675903095e-05, "loss": 0.008, "step": 19233 }, { "epoch": 2.0852124891587165, "grad_norm": 0.20449276268482208, "learning_rate": 3.060351080806615e-05, "loss": 0.0084, "step": 19234 }, { "epoch": 2.085320901994796, "grad_norm": 0.009925587102770805, "learning_rate": 3.0599883940229216e-05, "loss": 0.0003, "step": 19235 }, { "epoch": 2.085429314830876, "grad_norm": 0.7961086630821228, "learning_rate": 3.059625707239228e-05, "loss": 0.0864, "step": 19236 }, { "epoch": 2.0855377276669556, "grad_norm": 0.4010922610759735, "learning_rate": 3.0592630204555344e-05, "loss": 0.0222, "step": 19237 }, { "epoch": 2.0856461405030355, "grad_norm": 0.4251542091369629, "learning_rate": 3.0589003336718415e-05, "loss": 0.0186, "step": 19238 }, { "epoch": 2.0857545533391155, "grad_norm": 0.31891998648643494, "learning_rate": 3.058537646888148e-05, "loss": 0.031, "step": 19239 }, { "epoch": 2.085862966175195, "grad_norm": 0.14001356065273285, "learning_rate": 3.058174960104454e-05, "loss": 0.0054, "step": 19240 }, { "epoch": 2.085971379011275, "grad_norm": 0.43697476387023926, "learning_rate": 3.05781227332076e-05, "loss": 0.0255, "step": 19241 }, { "epoch": 2.0860797918473546, "grad_norm": 0.15500931441783905, "learning_rate": 3.0574495865370665e-05, "loss": 0.0124, "step": 19242 }, { "epoch": 2.0861882046834346, "grad_norm": 0.18494191765785217, "learning_rate": 3.057086899753373e-05, "loss": 0.0132, "step": 19243 }, { "epoch": 2.086296617519514, "grad_norm": 0.5260168313980103, "learning_rate": 3.056724212969679e-05, "loss": 0.0097, "step": 19244 }, { "epoch": 2.086405030355594, "grad_norm": 0.020247988402843475, "learning_rate": 3.056361526185986e-05, "loss": 0.001, "step": 19245 }, { "epoch": 2.086513443191674, "grad_norm": 0.048238158226013184, "learning_rate": 3.055998839402293e-05, "loss": 0.0024, "step": 19246 }, { "epoch": 2.0866218560277536, "grad_norm": 0.03523024171590805, "learning_rate": 3.055636152618599e-05, "loss": 0.0018, "step": 19247 }, { "epoch": 2.0867302688638336, "grad_norm": 0.06067211925983429, "learning_rate": 3.055273465834905e-05, "loss": 0.0019, "step": 19248 }, { "epoch": 2.086838681699913, "grad_norm": 0.11041948199272156, "learning_rate": 3.0549107790512114e-05, "loss": 0.0044, "step": 19249 }, { "epoch": 2.086947094535993, "grad_norm": 0.9436168074607849, "learning_rate": 3.054548092267518e-05, "loss": 0.02, "step": 19250 }, { "epoch": 2.0870555073720727, "grad_norm": 0.0413273423910141, "learning_rate": 3.054185405483824e-05, "loss": 0.0011, "step": 19251 }, { "epoch": 2.0871639202081527, "grad_norm": 0.5011230707168579, "learning_rate": 3.0538227187001306e-05, "loss": 0.0162, "step": 19252 }, { "epoch": 2.0872723330442327, "grad_norm": 0.1507548689842224, "learning_rate": 3.053460031916437e-05, "loss": 0.004, "step": 19253 }, { "epoch": 2.087380745880312, "grad_norm": 0.10617195814847946, "learning_rate": 3.0530973451327434e-05, "loss": 0.0033, "step": 19254 }, { "epoch": 2.087489158716392, "grad_norm": 0.12058689445257187, "learning_rate": 3.05273465834905e-05, "loss": 0.0037, "step": 19255 }, { "epoch": 2.0875975715524717, "grad_norm": 0.23163694143295288, "learning_rate": 3.052371971565356e-05, "loss": 0.0144, "step": 19256 }, { "epoch": 2.0877059843885517, "grad_norm": 1.4540479183197021, "learning_rate": 3.0520092847816626e-05, "loss": 0.0426, "step": 19257 }, { "epoch": 2.0878143972246312, "grad_norm": 0.7791006565093994, "learning_rate": 3.051646597997969e-05, "loss": 0.0123, "step": 19258 }, { "epoch": 2.0879228100607112, "grad_norm": 0.01583593524992466, "learning_rate": 3.0512839112142755e-05, "loss": 0.0008, "step": 19259 }, { "epoch": 2.0880312228967908, "grad_norm": 0.23619870841503143, "learning_rate": 3.050921224430582e-05, "loss": 0.0083, "step": 19260 }, { "epoch": 2.0881396357328708, "grad_norm": 0.42598235607147217, "learning_rate": 3.0505585376468883e-05, "loss": 0.0288, "step": 19261 }, { "epoch": 2.0882480485689507, "grad_norm": 0.24772466719150543, "learning_rate": 3.0501958508631944e-05, "loss": 0.0284, "step": 19262 }, { "epoch": 2.0883564614050303, "grad_norm": 0.15780070424079895, "learning_rate": 3.0498331640795008e-05, "loss": 0.007, "step": 19263 }, { "epoch": 2.0884648742411103, "grad_norm": 0.34194645285606384, "learning_rate": 3.049470477295808e-05, "loss": 0.0099, "step": 19264 }, { "epoch": 2.08857328707719, "grad_norm": 1.3592175245285034, "learning_rate": 3.049107790512114e-05, "loss": 0.0156, "step": 19265 }, { "epoch": 2.08868169991327, "grad_norm": 0.8134543299674988, "learning_rate": 3.0487451037284203e-05, "loss": 0.0341, "step": 19266 }, { "epoch": 2.0887901127493493, "grad_norm": 0.18879687786102295, "learning_rate": 3.0483824169447267e-05, "loss": 0.0066, "step": 19267 }, { "epoch": 2.0888985255854293, "grad_norm": 0.17614081501960754, "learning_rate": 3.048019730161033e-05, "loss": 0.0086, "step": 19268 }, { "epoch": 2.0890069384215093, "grad_norm": 0.06610678881406784, "learning_rate": 3.0476570433773392e-05, "loss": 0.002, "step": 19269 }, { "epoch": 2.089115351257589, "grad_norm": 0.08662351220846176, "learning_rate": 3.0472943565936456e-05, "loss": 0.0037, "step": 19270 }, { "epoch": 2.089223764093669, "grad_norm": 0.31722450256347656, "learning_rate": 3.046931669809952e-05, "loss": 0.0132, "step": 19271 }, { "epoch": 2.0893321769297484, "grad_norm": 0.9060353636741638, "learning_rate": 3.0465689830262588e-05, "loss": 0.0174, "step": 19272 }, { "epoch": 2.0894405897658284, "grad_norm": 0.6152048707008362, "learning_rate": 3.0462062962425652e-05, "loss": 0.0128, "step": 19273 }, { "epoch": 2.089549002601908, "grad_norm": 0.32653725147247314, "learning_rate": 3.0458436094588716e-05, "loss": 0.0192, "step": 19274 }, { "epoch": 2.089657415437988, "grad_norm": 0.033827949315309525, "learning_rate": 3.045480922675178e-05, "loss": 0.001, "step": 19275 }, { "epoch": 2.089765828274068, "grad_norm": 0.005051515530794859, "learning_rate": 3.045118235891484e-05, "loss": 0.0002, "step": 19276 }, { "epoch": 2.0898742411101474, "grad_norm": 0.3863898813724518, "learning_rate": 3.0447555491077905e-05, "loss": 0.0285, "step": 19277 }, { "epoch": 2.0899826539462274, "grad_norm": 0.602298378944397, "learning_rate": 3.044392862324097e-05, "loss": 0.0172, "step": 19278 }, { "epoch": 2.090091066782307, "grad_norm": 0.4386432468891144, "learning_rate": 3.0440301755404033e-05, "loss": 0.0307, "step": 19279 }, { "epoch": 2.090199479618387, "grad_norm": 0.1665373593568802, "learning_rate": 3.0436674887567097e-05, "loss": 0.0058, "step": 19280 }, { "epoch": 2.0903078924544665, "grad_norm": 0.05041861534118652, "learning_rate": 3.0433048019730165e-05, "loss": 0.0019, "step": 19281 }, { "epoch": 2.0904163052905465, "grad_norm": 0.39343690872192383, "learning_rate": 3.042942115189323e-05, "loss": 0.0458, "step": 19282 }, { "epoch": 2.090524718126626, "grad_norm": 0.08826886117458344, "learning_rate": 3.0425794284056293e-05, "loss": 0.0047, "step": 19283 }, { "epoch": 2.090633130962706, "grad_norm": 0.023973237723112106, "learning_rate": 3.0422167416219354e-05, "loss": 0.0008, "step": 19284 }, { "epoch": 2.090741543798786, "grad_norm": 0.4715981185436249, "learning_rate": 3.0418540548382418e-05, "loss": 0.0277, "step": 19285 }, { "epoch": 2.0908499566348655, "grad_norm": 0.3948509395122528, "learning_rate": 3.0414913680545482e-05, "loss": 0.0266, "step": 19286 }, { "epoch": 2.0909583694709455, "grad_norm": 0.7218407392501831, "learning_rate": 3.0411286812708546e-05, "loss": 0.0147, "step": 19287 }, { "epoch": 2.091066782307025, "grad_norm": 0.47657909989356995, "learning_rate": 3.0407659944871607e-05, "loss": 0.0191, "step": 19288 }, { "epoch": 2.091175195143105, "grad_norm": 0.12872016429901123, "learning_rate": 3.0404033077034678e-05, "loss": 0.0048, "step": 19289 }, { "epoch": 2.0912836079791846, "grad_norm": 0.10011904686689377, "learning_rate": 3.0400406209197742e-05, "loss": 0.0014, "step": 19290 }, { "epoch": 2.0913920208152645, "grad_norm": 0.08939608186483383, "learning_rate": 3.0396779341360803e-05, "loss": 0.006, "step": 19291 }, { "epoch": 2.0915004336513445, "grad_norm": 0.413465678691864, "learning_rate": 3.0393152473523867e-05, "loss": 0.0448, "step": 19292 }, { "epoch": 2.091608846487424, "grad_norm": 0.026331493631005287, "learning_rate": 3.038952560568693e-05, "loss": 0.0009, "step": 19293 }, { "epoch": 2.091717259323504, "grad_norm": 0.6627519726753235, "learning_rate": 3.0385898737849995e-05, "loss": 0.0303, "step": 19294 }, { "epoch": 2.0918256721595836, "grad_norm": 0.22046242654323578, "learning_rate": 3.0382271870013056e-05, "loss": 0.0355, "step": 19295 }, { "epoch": 2.0919340849956636, "grad_norm": 0.5624960064888, "learning_rate": 3.037864500217612e-05, "loss": 0.0333, "step": 19296 }, { "epoch": 2.092042497831743, "grad_norm": 0.24581976234912872, "learning_rate": 3.0375018134339184e-05, "loss": 0.019, "step": 19297 }, { "epoch": 2.092150910667823, "grad_norm": 0.5009682178497314, "learning_rate": 3.037139126650225e-05, "loss": 0.0207, "step": 19298 }, { "epoch": 2.092259323503903, "grad_norm": 0.4238545596599579, "learning_rate": 3.0367764398665315e-05, "loss": 0.023, "step": 19299 }, { "epoch": 2.0923677363399826, "grad_norm": 0.2871512770652771, "learning_rate": 3.036413753082838e-05, "loss": 0.0276, "step": 19300 }, { "epoch": 2.0924761491760626, "grad_norm": 0.3479864299297333, "learning_rate": 3.0360510662991444e-05, "loss": 0.0305, "step": 19301 }, { "epoch": 2.092584562012142, "grad_norm": 0.06504354625940323, "learning_rate": 3.0356883795154504e-05, "loss": 0.0012, "step": 19302 }, { "epoch": 2.092692974848222, "grad_norm": 0.3432495594024658, "learning_rate": 3.035325692731757e-05, "loss": 0.0384, "step": 19303 }, { "epoch": 2.0928013876843017, "grad_norm": 0.2860909402370453, "learning_rate": 3.0349630059480633e-05, "loss": 0.0269, "step": 19304 }, { "epoch": 2.0929098005203817, "grad_norm": 0.11786840856075287, "learning_rate": 3.0346003191643697e-05, "loss": 0.0062, "step": 19305 }, { "epoch": 2.093018213356461, "grad_norm": 0.011300504207611084, "learning_rate": 3.034237632380676e-05, "loss": 0.0007, "step": 19306 }, { "epoch": 2.093126626192541, "grad_norm": 0.08125802129507065, "learning_rate": 3.0338749455969828e-05, "loss": 0.0023, "step": 19307 }, { "epoch": 2.093235039028621, "grad_norm": 0.015544652938842773, "learning_rate": 3.0335122588132892e-05, "loss": 0.0005, "step": 19308 }, { "epoch": 2.0933434518647007, "grad_norm": 0.42824187874794006, "learning_rate": 3.0331495720295956e-05, "loss": 0.0372, "step": 19309 }, { "epoch": 2.0934518647007807, "grad_norm": 0.21691282093524933, "learning_rate": 3.0327868852459017e-05, "loss": 0.0067, "step": 19310 }, { "epoch": 2.0935602775368602, "grad_norm": 0.28503280878067017, "learning_rate": 3.032424198462208e-05, "loss": 0.0423, "step": 19311 }, { "epoch": 2.0936686903729402, "grad_norm": 0.30788326263427734, "learning_rate": 3.0320615116785145e-05, "loss": 0.0135, "step": 19312 }, { "epoch": 2.0937771032090198, "grad_norm": 0.17952805757522583, "learning_rate": 3.031698824894821e-05, "loss": 0.0041, "step": 19313 }, { "epoch": 2.0938855160450998, "grad_norm": 0.05282435566186905, "learning_rate": 3.031336138111127e-05, "loss": 0.0027, "step": 19314 }, { "epoch": 2.0939939288811797, "grad_norm": 0.09351501613855362, "learning_rate": 3.030973451327434e-05, "loss": 0.0029, "step": 19315 }, { "epoch": 2.0941023417172593, "grad_norm": 0.3872138559818268, "learning_rate": 3.0306107645437405e-05, "loss": 0.0078, "step": 19316 }, { "epoch": 2.0942107545533393, "grad_norm": 0.0866725817322731, "learning_rate": 3.0302480777600466e-05, "loss": 0.0048, "step": 19317 }, { "epoch": 2.094319167389419, "grad_norm": 0.04703013598918915, "learning_rate": 3.029885390976353e-05, "loss": 0.0018, "step": 19318 }, { "epoch": 2.094427580225499, "grad_norm": 0.06131306290626526, "learning_rate": 3.0295227041926594e-05, "loss": 0.0027, "step": 19319 }, { "epoch": 2.0945359930615783, "grad_norm": 0.19470146298408508, "learning_rate": 3.0291600174089658e-05, "loss": 0.0207, "step": 19320 }, { "epoch": 2.0946444058976583, "grad_norm": 0.008762297220528126, "learning_rate": 3.028797330625272e-05, "loss": 0.0005, "step": 19321 }, { "epoch": 2.094752818733738, "grad_norm": 0.04068969562649727, "learning_rate": 3.0284346438415783e-05, "loss": 0.0027, "step": 19322 }, { "epoch": 2.094861231569818, "grad_norm": 0.19583992660045624, "learning_rate": 3.0280719570578847e-05, "loss": 0.0161, "step": 19323 }, { "epoch": 2.094969644405898, "grad_norm": 0.3383992314338684, "learning_rate": 3.0277092702741915e-05, "loss": 0.0111, "step": 19324 }, { "epoch": 2.0950780572419774, "grad_norm": 0.05252508074045181, "learning_rate": 3.027346583490498e-05, "loss": 0.0027, "step": 19325 }, { "epoch": 2.0951864700780574, "grad_norm": 0.19158458709716797, "learning_rate": 3.0269838967068043e-05, "loss": 0.0151, "step": 19326 }, { "epoch": 2.095294882914137, "grad_norm": 0.19247028231620789, "learning_rate": 3.0266212099231107e-05, "loss": 0.0084, "step": 19327 }, { "epoch": 2.095403295750217, "grad_norm": 0.300951212644577, "learning_rate": 3.0262585231394168e-05, "loss": 0.0095, "step": 19328 }, { "epoch": 2.0955117085862964, "grad_norm": 0.33738401532173157, "learning_rate": 3.0258958363557232e-05, "loss": 0.0093, "step": 19329 }, { "epoch": 2.0956201214223764, "grad_norm": 0.11082067340612411, "learning_rate": 3.0255331495720296e-05, "loss": 0.0064, "step": 19330 }, { "epoch": 2.0957285342584564, "grad_norm": 0.09534765034914017, "learning_rate": 3.025170462788336e-05, "loss": 0.0047, "step": 19331 }, { "epoch": 2.095836947094536, "grad_norm": 0.018377412110567093, "learning_rate": 3.0248077760046424e-05, "loss": 0.0008, "step": 19332 }, { "epoch": 2.095945359930616, "grad_norm": 0.23912149667739868, "learning_rate": 3.024445089220949e-05, "loss": 0.0387, "step": 19333 }, { "epoch": 2.0960537727666955, "grad_norm": 0.20912547409534454, "learning_rate": 3.0240824024372556e-05, "loss": 0.0147, "step": 19334 }, { "epoch": 2.0961621856027755, "grad_norm": 0.11668436974287033, "learning_rate": 3.023719715653562e-05, "loss": 0.0062, "step": 19335 }, { "epoch": 2.096270598438855, "grad_norm": 0.173745259642601, "learning_rate": 3.023357028869868e-05, "loss": 0.0111, "step": 19336 }, { "epoch": 2.096379011274935, "grad_norm": 0.16296595335006714, "learning_rate": 3.0229943420861745e-05, "loss": 0.0178, "step": 19337 }, { "epoch": 2.0964874241110145, "grad_norm": 0.2222719043493271, "learning_rate": 3.022631655302481e-05, "loss": 0.0093, "step": 19338 }, { "epoch": 2.0965958369470945, "grad_norm": 0.40443792939186096, "learning_rate": 3.0222689685187873e-05, "loss": 0.0609, "step": 19339 }, { "epoch": 2.0967042497831745, "grad_norm": 0.1257236748933792, "learning_rate": 3.0219062817350934e-05, "loss": 0.0063, "step": 19340 }, { "epoch": 2.096812662619254, "grad_norm": 0.28519806265830994, "learning_rate": 3.0215435949514004e-05, "loss": 0.0182, "step": 19341 }, { "epoch": 2.096921075455334, "grad_norm": 0.32682010531425476, "learning_rate": 3.021180908167707e-05, "loss": 0.0146, "step": 19342 }, { "epoch": 2.0970294882914136, "grad_norm": 0.8463889360427856, "learning_rate": 3.020818221384013e-05, "loss": 0.0173, "step": 19343 }, { "epoch": 2.0971379011274935, "grad_norm": 0.05124668404459953, "learning_rate": 3.0204555346003193e-05, "loss": 0.0035, "step": 19344 }, { "epoch": 2.097246313963573, "grad_norm": 0.11000502109527588, "learning_rate": 3.0200928478166257e-05, "loss": 0.009, "step": 19345 }, { "epoch": 2.097354726799653, "grad_norm": 0.3624473810195923, "learning_rate": 3.019730161032932e-05, "loss": 0.0209, "step": 19346 }, { "epoch": 2.097463139635733, "grad_norm": 0.021867286413908005, "learning_rate": 3.0193674742492382e-05, "loss": 0.0009, "step": 19347 }, { "epoch": 2.0975715524718126, "grad_norm": 0.23098544776439667, "learning_rate": 3.0190047874655446e-05, "loss": 0.0103, "step": 19348 }, { "epoch": 2.0976799653078926, "grad_norm": 0.12598010897636414, "learning_rate": 3.018642100681851e-05, "loss": 0.0081, "step": 19349 }, { "epoch": 2.097788378143972, "grad_norm": 0.10258784890174866, "learning_rate": 3.0182794138981578e-05, "loss": 0.0055, "step": 19350 }, { "epoch": 2.097896790980052, "grad_norm": 0.12694157660007477, "learning_rate": 3.0179167271144642e-05, "loss": 0.0028, "step": 19351 }, { "epoch": 2.0980052038161316, "grad_norm": 0.3393573462963104, "learning_rate": 3.0175540403307706e-05, "loss": 0.0075, "step": 19352 }, { "epoch": 2.0981136166522116, "grad_norm": 0.09386581927537918, "learning_rate": 3.017191353547077e-05, "loss": 0.0054, "step": 19353 }, { "epoch": 2.0982220294882916, "grad_norm": 0.29835763573646545, "learning_rate": 3.016828666763383e-05, "loss": 0.015, "step": 19354 }, { "epoch": 2.098330442324371, "grad_norm": 0.41198989748954773, "learning_rate": 3.0164659799796895e-05, "loss": 0.0242, "step": 19355 }, { "epoch": 2.098438855160451, "grad_norm": 0.1537097543478012, "learning_rate": 3.016103293195996e-05, "loss": 0.0082, "step": 19356 }, { "epoch": 2.0985472679965307, "grad_norm": 0.391883909702301, "learning_rate": 3.0157406064123023e-05, "loss": 0.0146, "step": 19357 }, { "epoch": 2.0986556808326107, "grad_norm": 0.10598345100879669, "learning_rate": 3.0153779196286087e-05, "loss": 0.0031, "step": 19358 }, { "epoch": 2.09876409366869, "grad_norm": 0.1861560344696045, "learning_rate": 3.0150152328449155e-05, "loss": 0.0047, "step": 19359 }, { "epoch": 2.09887250650477, "grad_norm": 0.050734326243400574, "learning_rate": 3.014652546061222e-05, "loss": 0.0013, "step": 19360 }, { "epoch": 2.0989809193408497, "grad_norm": 0.0376712940633297, "learning_rate": 3.0142898592775283e-05, "loss": 0.0013, "step": 19361 }, { "epoch": 2.0990893321769297, "grad_norm": 0.2696501314640045, "learning_rate": 3.0139271724938344e-05, "loss": 0.0088, "step": 19362 }, { "epoch": 2.0991977450130097, "grad_norm": 0.032941099256277084, "learning_rate": 3.0135644857101408e-05, "loss": 0.0009, "step": 19363 }, { "epoch": 2.0993061578490892, "grad_norm": 0.2496790587902069, "learning_rate": 3.0132017989264472e-05, "loss": 0.0085, "step": 19364 }, { "epoch": 2.0994145706851692, "grad_norm": 0.7893526554107666, "learning_rate": 3.0128391121427536e-05, "loss": 0.0251, "step": 19365 }, { "epoch": 2.0995229835212488, "grad_norm": 0.18630164861679077, "learning_rate": 3.0124764253590597e-05, "loss": 0.0077, "step": 19366 }, { "epoch": 2.0996313963573288, "grad_norm": 0.07195071876049042, "learning_rate": 3.0121137385753668e-05, "loss": 0.0031, "step": 19367 }, { "epoch": 2.0997398091934083, "grad_norm": 0.706720769405365, "learning_rate": 3.0117510517916732e-05, "loss": 0.0182, "step": 19368 }, { "epoch": 2.0998482220294883, "grad_norm": 0.356166809797287, "learning_rate": 3.0113883650079793e-05, "loss": 0.0075, "step": 19369 }, { "epoch": 2.0999566348655683, "grad_norm": 0.45139873027801514, "learning_rate": 3.0110256782242857e-05, "loss": 0.0136, "step": 19370 }, { "epoch": 2.100065047701648, "grad_norm": 0.017087826505303383, "learning_rate": 3.010662991440592e-05, "loss": 0.0007, "step": 19371 }, { "epoch": 2.100173460537728, "grad_norm": 1.0682367086410522, "learning_rate": 3.0103003046568985e-05, "loss": 0.0327, "step": 19372 }, { "epoch": 2.1002818733738073, "grad_norm": 0.018885135650634766, "learning_rate": 3.0099376178732046e-05, "loss": 0.001, "step": 19373 }, { "epoch": 2.1003902862098873, "grad_norm": 0.16225138306617737, "learning_rate": 3.009574931089511e-05, "loss": 0.0076, "step": 19374 }, { "epoch": 2.100498699045967, "grad_norm": 0.21798613667488098, "learning_rate": 3.0092122443058174e-05, "loss": 0.0074, "step": 19375 }, { "epoch": 2.100607111882047, "grad_norm": 0.288813054561615, "learning_rate": 3.008849557522124e-05, "loss": 0.0056, "step": 19376 }, { "epoch": 2.100715524718127, "grad_norm": 0.0984637662768364, "learning_rate": 3.0084868707384305e-05, "loss": 0.0013, "step": 19377 }, { "epoch": 2.1008239375542064, "grad_norm": 0.02620062604546547, "learning_rate": 3.008124183954737e-05, "loss": 0.0006, "step": 19378 }, { "epoch": 2.1009323503902864, "grad_norm": 0.0920504778623581, "learning_rate": 3.0077614971710434e-05, "loss": 0.0022, "step": 19379 }, { "epoch": 2.101040763226366, "grad_norm": 0.3934493660926819, "learning_rate": 3.0073988103873498e-05, "loss": 0.0593, "step": 19380 }, { "epoch": 2.101149176062446, "grad_norm": 0.48727014660835266, "learning_rate": 3.007036123603656e-05, "loss": 0.0373, "step": 19381 }, { "epoch": 2.1012575888985254, "grad_norm": 0.27879124879837036, "learning_rate": 3.0066734368199623e-05, "loss": 0.0114, "step": 19382 }, { "epoch": 2.1013660017346054, "grad_norm": 0.045127298682928085, "learning_rate": 3.0063107500362687e-05, "loss": 0.0018, "step": 19383 }, { "epoch": 2.101474414570685, "grad_norm": 0.3525896668434143, "learning_rate": 3.0059480632525754e-05, "loss": 0.0193, "step": 19384 }, { "epoch": 2.101582827406765, "grad_norm": 0.7852990627288818, "learning_rate": 3.0055853764688818e-05, "loss": 0.0324, "step": 19385 }, { "epoch": 2.101691240242845, "grad_norm": 0.6316555142402649, "learning_rate": 3.0052226896851882e-05, "loss": 0.0127, "step": 19386 }, { "epoch": 2.1017996530789245, "grad_norm": 0.03412403538823128, "learning_rate": 3.0048600029014946e-05, "loss": 0.0014, "step": 19387 }, { "epoch": 2.1019080659150045, "grad_norm": 0.04483181983232498, "learning_rate": 3.0044973161178007e-05, "loss": 0.0014, "step": 19388 }, { "epoch": 2.102016478751084, "grad_norm": 0.08256559073925018, "learning_rate": 3.004134629334107e-05, "loss": 0.0101, "step": 19389 }, { "epoch": 2.102124891587164, "grad_norm": 0.41413694620132446, "learning_rate": 3.0037719425504135e-05, "loss": 0.0156, "step": 19390 }, { "epoch": 2.1022333044232435, "grad_norm": 0.021186133846640587, "learning_rate": 3.00340925576672e-05, "loss": 0.0008, "step": 19391 }, { "epoch": 2.1023417172593235, "grad_norm": 0.26241421699523926, "learning_rate": 3.003046568983026e-05, "loss": 0.0393, "step": 19392 }, { "epoch": 2.1024501300954035, "grad_norm": 0.7593815326690674, "learning_rate": 3.002683882199333e-05, "loss": 0.0295, "step": 19393 }, { "epoch": 2.102558542931483, "grad_norm": 0.054096098989248276, "learning_rate": 3.0023211954156395e-05, "loss": 0.0027, "step": 19394 }, { "epoch": 2.102666955767563, "grad_norm": 0.06055089831352234, "learning_rate": 3.0019585086319456e-05, "loss": 0.0026, "step": 19395 }, { "epoch": 2.1027753686036426, "grad_norm": 0.1269424855709076, "learning_rate": 3.001595821848252e-05, "loss": 0.0049, "step": 19396 }, { "epoch": 2.1028837814397225, "grad_norm": 0.30738845467567444, "learning_rate": 3.0012331350645584e-05, "loss": 0.0078, "step": 19397 }, { "epoch": 2.102992194275802, "grad_norm": 0.01913570612668991, "learning_rate": 3.0008704482808648e-05, "loss": 0.0007, "step": 19398 }, { "epoch": 2.103100607111882, "grad_norm": 0.06642471253871918, "learning_rate": 3.000507761497171e-05, "loss": 0.0019, "step": 19399 }, { "epoch": 2.103209019947962, "grad_norm": 0.0015226865652948618, "learning_rate": 3.0001450747134773e-05, "loss": 0.0001, "step": 19400 }, { "epoch": 2.1033174327840416, "grad_norm": 1.1344376802444458, "learning_rate": 2.9997823879297837e-05, "loss": 0.0536, "step": 19401 }, { "epoch": 2.1034258456201216, "grad_norm": 0.11098356544971466, "learning_rate": 2.9994197011460905e-05, "loss": 0.0022, "step": 19402 }, { "epoch": 2.103534258456201, "grad_norm": 0.05135548859834671, "learning_rate": 2.999057014362397e-05, "loss": 0.001, "step": 19403 }, { "epoch": 2.103642671292281, "grad_norm": 0.20398861169815063, "learning_rate": 2.9986943275787033e-05, "loss": 0.0084, "step": 19404 }, { "epoch": 2.1037510841283606, "grad_norm": 0.026975153014063835, "learning_rate": 2.9983316407950097e-05, "loss": 0.0014, "step": 19405 }, { "epoch": 2.1038594969644406, "grad_norm": 0.1671818047761917, "learning_rate": 2.997968954011316e-05, "loss": 0.012, "step": 19406 }, { "epoch": 2.10396790980052, "grad_norm": 0.022302625700831413, "learning_rate": 2.9976062672276222e-05, "loss": 0.0005, "step": 19407 }, { "epoch": 2.1040763226366, "grad_norm": 0.017408017069101334, "learning_rate": 2.9972435804439286e-05, "loss": 0.0008, "step": 19408 }, { "epoch": 2.10418473547268, "grad_norm": 0.36048606038093567, "learning_rate": 2.996880893660235e-05, "loss": 0.0316, "step": 19409 }, { "epoch": 2.1042931483087597, "grad_norm": 0.0574212484061718, "learning_rate": 2.9965182068765418e-05, "loss": 0.0019, "step": 19410 }, { "epoch": 2.1044015611448397, "grad_norm": 0.6058133840560913, "learning_rate": 2.996155520092848e-05, "loss": 0.0412, "step": 19411 }, { "epoch": 2.104509973980919, "grad_norm": 0.010517880320549011, "learning_rate": 2.9957928333091546e-05, "loss": 0.0003, "step": 19412 }, { "epoch": 2.104618386816999, "grad_norm": 0.18545740842819214, "learning_rate": 2.995430146525461e-05, "loss": 0.0096, "step": 19413 }, { "epoch": 2.1047267996530787, "grad_norm": 0.3654257655143738, "learning_rate": 2.995067459741767e-05, "loss": 0.0092, "step": 19414 }, { "epoch": 2.1048352124891587, "grad_norm": 0.0425407774746418, "learning_rate": 2.9947047729580735e-05, "loss": 0.001, "step": 19415 }, { "epoch": 2.1049436253252387, "grad_norm": 0.2873050570487976, "learning_rate": 2.99434208617438e-05, "loss": 0.0205, "step": 19416 }, { "epoch": 2.1050520381613183, "grad_norm": 0.6847046613693237, "learning_rate": 2.9939793993906863e-05, "loss": 0.0112, "step": 19417 }, { "epoch": 2.1051604509973982, "grad_norm": 0.28964418172836304, "learning_rate": 2.9936167126069924e-05, "loss": 0.0304, "step": 19418 }, { "epoch": 2.1052688638334778, "grad_norm": 0.17888930439949036, "learning_rate": 2.9932540258232994e-05, "loss": 0.0223, "step": 19419 }, { "epoch": 2.1053772766695578, "grad_norm": 0.30619916319847107, "learning_rate": 2.992891339039606e-05, "loss": 0.0086, "step": 19420 }, { "epoch": 2.1054856895056373, "grad_norm": 0.22211085259914398, "learning_rate": 2.992528652255912e-05, "loss": 0.0105, "step": 19421 }, { "epoch": 2.1055941023417173, "grad_norm": 0.327374666929245, "learning_rate": 2.9921659654722183e-05, "loss": 0.011, "step": 19422 }, { "epoch": 2.1057025151777973, "grad_norm": 0.09729939699172974, "learning_rate": 2.9918032786885248e-05, "loss": 0.0037, "step": 19423 }, { "epoch": 2.105810928013877, "grad_norm": 0.6308396458625793, "learning_rate": 2.991440591904831e-05, "loss": 0.0385, "step": 19424 }, { "epoch": 2.105919340849957, "grad_norm": 0.20908409357070923, "learning_rate": 2.9910779051211372e-05, "loss": 0.0107, "step": 19425 }, { "epoch": 2.1060277536860363, "grad_norm": 0.06004876270890236, "learning_rate": 2.9907152183374436e-05, "loss": 0.0045, "step": 19426 }, { "epoch": 2.1061361665221163, "grad_norm": 0.34802740812301636, "learning_rate": 2.99035253155375e-05, "loss": 0.0446, "step": 19427 }, { "epoch": 2.106244579358196, "grad_norm": 0.2551008462905884, "learning_rate": 2.9899898447700568e-05, "loss": 0.0071, "step": 19428 }, { "epoch": 2.106352992194276, "grad_norm": 0.010050755925476551, "learning_rate": 2.9896271579863632e-05, "loss": 0.0002, "step": 19429 }, { "epoch": 2.1064614050303554, "grad_norm": 0.4946863353252411, "learning_rate": 2.9892644712026696e-05, "loss": 0.0198, "step": 19430 }, { "epoch": 2.1065698178664354, "grad_norm": 0.5601319670677185, "learning_rate": 2.988901784418976e-05, "loss": 0.0107, "step": 19431 }, { "epoch": 2.1066782307025154, "grad_norm": 0.05559967830777168, "learning_rate": 2.9885390976352824e-05, "loss": 0.0033, "step": 19432 }, { "epoch": 2.106786643538595, "grad_norm": 0.038836684077978134, "learning_rate": 2.9881764108515885e-05, "loss": 0.0019, "step": 19433 }, { "epoch": 2.106895056374675, "grad_norm": 0.02478106878697872, "learning_rate": 2.987813724067895e-05, "loss": 0.0011, "step": 19434 }, { "epoch": 2.1070034692107544, "grad_norm": 0.13700756430625916, "learning_rate": 2.9874510372842013e-05, "loss": 0.0069, "step": 19435 }, { "epoch": 2.1071118820468344, "grad_norm": 0.5864073634147644, "learning_rate": 2.987088350500508e-05, "loss": 0.0045, "step": 19436 }, { "epoch": 2.107220294882914, "grad_norm": 0.1779802292585373, "learning_rate": 2.9867256637168145e-05, "loss": 0.0046, "step": 19437 }, { "epoch": 2.107328707718994, "grad_norm": 0.12631569802761078, "learning_rate": 2.986362976933121e-05, "loss": 0.0077, "step": 19438 }, { "epoch": 2.107437120555074, "grad_norm": 0.05185403674840927, "learning_rate": 2.9860002901494273e-05, "loss": 0.003, "step": 19439 }, { "epoch": 2.1075455333911535, "grad_norm": 0.11681708693504333, "learning_rate": 2.9856376033657334e-05, "loss": 0.005, "step": 19440 }, { "epoch": 2.1076539462272335, "grad_norm": 0.08574952185153961, "learning_rate": 2.9852749165820398e-05, "loss": 0.0033, "step": 19441 }, { "epoch": 2.107762359063313, "grad_norm": 0.22151130437850952, "learning_rate": 2.9849122297983462e-05, "loss": 0.0116, "step": 19442 }, { "epoch": 2.107870771899393, "grad_norm": 0.04448322951793671, "learning_rate": 2.9845495430146526e-05, "loss": 0.0016, "step": 19443 }, { "epoch": 2.1079791847354725, "grad_norm": 0.5861104726791382, "learning_rate": 2.9841868562309587e-05, "loss": 0.0067, "step": 19444 }, { "epoch": 2.1080875975715525, "grad_norm": 0.22810372710227966, "learning_rate": 2.9838241694472658e-05, "loss": 0.0153, "step": 19445 }, { "epoch": 2.1081960104076325, "grad_norm": 0.1818462610244751, "learning_rate": 2.9834614826635722e-05, "loss": 0.0121, "step": 19446 }, { "epoch": 2.108304423243712, "grad_norm": 0.10117796063423157, "learning_rate": 2.9830987958798783e-05, "loss": 0.004, "step": 19447 }, { "epoch": 2.108412836079792, "grad_norm": 0.3351619243621826, "learning_rate": 2.9827361090961847e-05, "loss": 0.0131, "step": 19448 }, { "epoch": 2.1085212489158716, "grad_norm": 0.07247935235500336, "learning_rate": 2.982373422312491e-05, "loss": 0.0024, "step": 19449 }, { "epoch": 2.1086296617519515, "grad_norm": 0.08533736318349838, "learning_rate": 2.9820107355287975e-05, "loss": 0.0025, "step": 19450 }, { "epoch": 2.108738074588031, "grad_norm": 0.02421988546848297, "learning_rate": 2.9816480487451036e-05, "loss": 0.0011, "step": 19451 }, { "epoch": 2.108846487424111, "grad_norm": 0.0959606021642685, "learning_rate": 2.98128536196141e-05, "loss": 0.0022, "step": 19452 }, { "epoch": 2.1089549002601906, "grad_norm": 0.2455490380525589, "learning_rate": 2.9809226751777164e-05, "loss": 0.0086, "step": 19453 }, { "epoch": 2.1090633130962706, "grad_norm": 0.07946522533893585, "learning_rate": 2.980559988394023e-05, "loss": 0.0019, "step": 19454 }, { "epoch": 2.1091717259323506, "grad_norm": 0.29457512497901917, "learning_rate": 2.9801973016103295e-05, "loss": 0.0178, "step": 19455 }, { "epoch": 2.10928013876843, "grad_norm": 0.5954134464263916, "learning_rate": 2.979834614826636e-05, "loss": 0.0486, "step": 19456 }, { "epoch": 2.10938855160451, "grad_norm": 0.014836609363555908, "learning_rate": 2.9794719280429424e-05, "loss": 0.0007, "step": 19457 }, { "epoch": 2.1094969644405897, "grad_norm": 0.2697697877883911, "learning_rate": 2.9791092412592488e-05, "loss": 0.0407, "step": 19458 }, { "epoch": 2.1096053772766696, "grad_norm": 0.0834064930677414, "learning_rate": 2.978746554475555e-05, "loss": 0.003, "step": 19459 }, { "epoch": 2.109713790112749, "grad_norm": 0.113616444170475, "learning_rate": 2.9783838676918613e-05, "loss": 0.0042, "step": 19460 }, { "epoch": 2.109822202948829, "grad_norm": 0.5616153478622437, "learning_rate": 2.9780211809081677e-05, "loss": 0.0251, "step": 19461 }, { "epoch": 2.109930615784909, "grad_norm": 0.6000633835792542, "learning_rate": 2.9776584941244744e-05, "loss": 0.0105, "step": 19462 }, { "epoch": 2.1100390286209887, "grad_norm": 0.01639443077147007, "learning_rate": 2.977295807340781e-05, "loss": 0.0005, "step": 19463 }, { "epoch": 2.1101474414570687, "grad_norm": 0.07008283585309982, "learning_rate": 2.9769331205570872e-05, "loss": 0.0027, "step": 19464 }, { "epoch": 2.110255854293148, "grad_norm": 0.18650759756565094, "learning_rate": 2.9765704337733937e-05, "loss": 0.0077, "step": 19465 }, { "epoch": 2.110364267129228, "grad_norm": 0.16039516031742096, "learning_rate": 2.9762077469896997e-05, "loss": 0.0074, "step": 19466 }, { "epoch": 2.1104726799653077, "grad_norm": 0.4150090515613556, "learning_rate": 2.975845060206006e-05, "loss": 0.0306, "step": 19467 }, { "epoch": 2.1105810928013877, "grad_norm": 0.25500187277793884, "learning_rate": 2.9754823734223125e-05, "loss": 0.009, "step": 19468 }, { "epoch": 2.1106895056374673, "grad_norm": 0.11858168989419937, "learning_rate": 2.975119686638619e-05, "loss": 0.0056, "step": 19469 }, { "epoch": 2.1107979184735473, "grad_norm": 0.5853834748268127, "learning_rate": 2.974756999854925e-05, "loss": 0.0139, "step": 19470 }, { "epoch": 2.1109063313096272, "grad_norm": 0.09323892742395401, "learning_rate": 2.974394313071232e-05, "loss": 0.0026, "step": 19471 }, { "epoch": 2.111014744145707, "grad_norm": 0.22645588219165802, "learning_rate": 2.9740316262875385e-05, "loss": 0.0216, "step": 19472 }, { "epoch": 2.1111231569817868, "grad_norm": 0.169172003865242, "learning_rate": 2.9736689395038446e-05, "loss": 0.0037, "step": 19473 }, { "epoch": 2.1112315698178663, "grad_norm": 0.01813490130007267, "learning_rate": 2.973306252720151e-05, "loss": 0.0005, "step": 19474 }, { "epoch": 2.1113399826539463, "grad_norm": 0.3298380374908447, "learning_rate": 2.9729435659364574e-05, "loss": 0.0127, "step": 19475 }, { "epoch": 2.111448395490026, "grad_norm": 0.13248558342456818, "learning_rate": 2.9725808791527638e-05, "loss": 0.0029, "step": 19476 }, { "epoch": 2.111556808326106, "grad_norm": 0.07201328873634338, "learning_rate": 2.97221819236907e-05, "loss": 0.0039, "step": 19477 }, { "epoch": 2.111665221162186, "grad_norm": 0.021504253149032593, "learning_rate": 2.9718555055853763e-05, "loss": 0.0006, "step": 19478 }, { "epoch": 2.1117736339982653, "grad_norm": 0.6321849822998047, "learning_rate": 2.9714928188016827e-05, "loss": 0.0203, "step": 19479 }, { "epoch": 2.1118820468343453, "grad_norm": 0.4494834840297699, "learning_rate": 2.9711301320179895e-05, "loss": 0.0249, "step": 19480 }, { "epoch": 2.111990459670425, "grad_norm": 0.2142605185508728, "learning_rate": 2.970767445234296e-05, "loss": 0.0089, "step": 19481 }, { "epoch": 2.112098872506505, "grad_norm": 0.27671515941619873, "learning_rate": 2.9704047584506023e-05, "loss": 0.0179, "step": 19482 }, { "epoch": 2.1122072853425844, "grad_norm": 0.14969070255756378, "learning_rate": 2.9700420716669087e-05, "loss": 0.0102, "step": 19483 }, { "epoch": 2.1123156981786644, "grad_norm": 0.25843295454978943, "learning_rate": 2.969679384883215e-05, "loss": 0.0099, "step": 19484 }, { "epoch": 2.112424111014744, "grad_norm": 0.1095641702413559, "learning_rate": 2.9693166980995212e-05, "loss": 0.003, "step": 19485 }, { "epoch": 2.112532523850824, "grad_norm": 0.49813538789749146, "learning_rate": 2.9689540113158276e-05, "loss": 0.0224, "step": 19486 }, { "epoch": 2.112640936686904, "grad_norm": 0.9768075346946716, "learning_rate": 2.968591324532134e-05, "loss": 0.014, "step": 19487 }, { "epoch": 2.1127493495229834, "grad_norm": 0.4574405550956726, "learning_rate": 2.9682286377484408e-05, "loss": 0.0451, "step": 19488 }, { "epoch": 2.1128577623590634, "grad_norm": 0.8092764019966125, "learning_rate": 2.967865950964747e-05, "loss": 0.0231, "step": 19489 }, { "epoch": 2.112966175195143, "grad_norm": 0.006308979820460081, "learning_rate": 2.9675032641810536e-05, "loss": 0.0001, "step": 19490 }, { "epoch": 2.113074588031223, "grad_norm": 0.21875950694084167, "learning_rate": 2.96714057739736e-05, "loss": 0.0102, "step": 19491 }, { "epoch": 2.1131830008673025, "grad_norm": 0.2146606743335724, "learning_rate": 2.966777890613666e-05, "loss": 0.0045, "step": 19492 }, { "epoch": 2.1132914137033825, "grad_norm": 0.37941351532936096, "learning_rate": 2.9664152038299725e-05, "loss": 0.0129, "step": 19493 }, { "epoch": 2.1133998265394625, "grad_norm": 0.023922398686408997, "learning_rate": 2.966052517046279e-05, "loss": 0.001, "step": 19494 }, { "epoch": 2.113508239375542, "grad_norm": 0.054845962673425674, "learning_rate": 2.9656898302625853e-05, "loss": 0.004, "step": 19495 }, { "epoch": 2.113616652211622, "grad_norm": 0.4724596440792084, "learning_rate": 2.9653271434788914e-05, "loss": 0.0277, "step": 19496 }, { "epoch": 2.1137250650477015, "grad_norm": 0.024988481774926186, "learning_rate": 2.9649644566951985e-05, "loss": 0.0012, "step": 19497 }, { "epoch": 2.1138334778837815, "grad_norm": 0.07520914822816849, "learning_rate": 2.964601769911505e-05, "loss": 0.0011, "step": 19498 }, { "epoch": 2.113941890719861, "grad_norm": 0.25237226486206055, "learning_rate": 2.964239083127811e-05, "loss": 0.016, "step": 19499 }, { "epoch": 2.114050303555941, "grad_norm": 0.24645160138607025, "learning_rate": 2.9638763963441173e-05, "loss": 0.007, "step": 19500 }, { "epoch": 2.114158716392021, "grad_norm": 0.182042196393013, "learning_rate": 2.9635137095604238e-05, "loss": 0.0264, "step": 19501 }, { "epoch": 2.1142671292281006, "grad_norm": 0.6918849945068359, "learning_rate": 2.96315102277673e-05, "loss": 0.0447, "step": 19502 }, { "epoch": 2.1143755420641805, "grad_norm": 0.04287317767739296, "learning_rate": 2.9627883359930362e-05, "loss": 0.0012, "step": 19503 }, { "epoch": 2.11448395490026, "grad_norm": 0.06884721666574478, "learning_rate": 2.9624256492093426e-05, "loss": 0.0025, "step": 19504 }, { "epoch": 2.11459236773634, "grad_norm": 0.07415907084941864, "learning_rate": 2.9620629624256497e-05, "loss": 0.0028, "step": 19505 }, { "epoch": 2.1147007805724196, "grad_norm": 0.01782127283513546, "learning_rate": 2.9617002756419558e-05, "loss": 0.0009, "step": 19506 }, { "epoch": 2.1148091934084996, "grad_norm": 0.20190252363681793, "learning_rate": 2.9613375888582622e-05, "loss": 0.0144, "step": 19507 }, { "epoch": 2.114917606244579, "grad_norm": 0.30579614639282227, "learning_rate": 2.9609749020745686e-05, "loss": 0.0069, "step": 19508 }, { "epoch": 2.115026019080659, "grad_norm": 0.034102343022823334, "learning_rate": 2.960612215290875e-05, "loss": 0.001, "step": 19509 }, { "epoch": 2.115134431916739, "grad_norm": 0.08368091285228729, "learning_rate": 2.9602495285071814e-05, "loss": 0.0019, "step": 19510 }, { "epoch": 2.1152428447528187, "grad_norm": 0.2620549499988556, "learning_rate": 2.9598868417234875e-05, "loss": 0.0038, "step": 19511 }, { "epoch": 2.1153512575888986, "grad_norm": 0.062192484736442566, "learning_rate": 2.959524154939794e-05, "loss": 0.0034, "step": 19512 }, { "epoch": 2.115459670424978, "grad_norm": 0.10824277997016907, "learning_rate": 2.9591614681561003e-05, "loss": 0.0027, "step": 19513 }, { "epoch": 2.115568083261058, "grad_norm": 0.05453374236822128, "learning_rate": 2.958798781372407e-05, "loss": 0.0029, "step": 19514 }, { "epoch": 2.1156764960971377, "grad_norm": 0.45805564522743225, "learning_rate": 2.9584360945887135e-05, "loss": 0.0279, "step": 19515 }, { "epoch": 2.1157849089332177, "grad_norm": 0.03981664776802063, "learning_rate": 2.95807340780502e-05, "loss": 0.0016, "step": 19516 }, { "epoch": 2.1158933217692977, "grad_norm": 0.21940863132476807, "learning_rate": 2.9577107210213263e-05, "loss": 0.0082, "step": 19517 }, { "epoch": 2.116001734605377, "grad_norm": 0.07750770449638367, "learning_rate": 2.9573480342376324e-05, "loss": 0.0036, "step": 19518 }, { "epoch": 2.116110147441457, "grad_norm": 0.18054993450641632, "learning_rate": 2.9569853474539388e-05, "loss": 0.007, "step": 19519 }, { "epoch": 2.1162185602775367, "grad_norm": 0.27735161781311035, "learning_rate": 2.9566226606702452e-05, "loss": 0.0333, "step": 19520 }, { "epoch": 2.1163269731136167, "grad_norm": 0.3049995005130768, "learning_rate": 2.9562599738865516e-05, "loss": 0.004, "step": 19521 }, { "epoch": 2.1164353859496963, "grad_norm": 0.1194593533873558, "learning_rate": 2.9558972871028577e-05, "loss": 0.0032, "step": 19522 }, { "epoch": 2.1165437987857763, "grad_norm": 0.40098047256469727, "learning_rate": 2.9555346003191648e-05, "loss": 0.0229, "step": 19523 }, { "epoch": 2.1166522116218562, "grad_norm": 0.3918851315975189, "learning_rate": 2.9551719135354712e-05, "loss": 0.0076, "step": 19524 }, { "epoch": 2.116760624457936, "grad_norm": 0.1543273627758026, "learning_rate": 2.9548092267517773e-05, "loss": 0.0054, "step": 19525 }, { "epoch": 2.1168690372940158, "grad_norm": 0.13756640255451202, "learning_rate": 2.9544465399680837e-05, "loss": 0.0079, "step": 19526 }, { "epoch": 2.1169774501300953, "grad_norm": 0.08421692252159119, "learning_rate": 2.95408385318439e-05, "loss": 0.0022, "step": 19527 }, { "epoch": 2.1170858629661753, "grad_norm": 0.436318039894104, "learning_rate": 2.9537211664006965e-05, "loss": 0.021, "step": 19528 }, { "epoch": 2.117194275802255, "grad_norm": 0.3852590322494507, "learning_rate": 2.9533584796170026e-05, "loss": 0.0254, "step": 19529 }, { "epoch": 2.117302688638335, "grad_norm": 0.16993479430675507, "learning_rate": 2.952995792833309e-05, "loss": 0.0073, "step": 19530 }, { "epoch": 2.1174111014744144, "grad_norm": 0.009423209354281425, "learning_rate": 2.952633106049616e-05, "loss": 0.0003, "step": 19531 }, { "epoch": 2.1175195143104943, "grad_norm": 0.0012334943749010563, "learning_rate": 2.952270419265922e-05, "loss": 0.0001, "step": 19532 }, { "epoch": 2.1176279271465743, "grad_norm": 0.25333598256111145, "learning_rate": 2.9519077324822286e-05, "loss": 0.0057, "step": 19533 }, { "epoch": 2.117736339982654, "grad_norm": 0.16746649146080017, "learning_rate": 2.951545045698535e-05, "loss": 0.0037, "step": 19534 }, { "epoch": 2.117844752818734, "grad_norm": 1.1339188814163208, "learning_rate": 2.9511823589148414e-05, "loss": 0.0225, "step": 19535 }, { "epoch": 2.1179531656548134, "grad_norm": 0.49849605560302734, "learning_rate": 2.9508196721311478e-05, "loss": 0.0208, "step": 19536 }, { "epoch": 2.1180615784908934, "grad_norm": 0.020051704719662666, "learning_rate": 2.950456985347454e-05, "loss": 0.0009, "step": 19537 }, { "epoch": 2.118169991326973, "grad_norm": 0.18973083794116974, "learning_rate": 2.9500942985637603e-05, "loss": 0.0053, "step": 19538 }, { "epoch": 2.118278404163053, "grad_norm": 0.5685594081878662, "learning_rate": 2.9497316117800667e-05, "loss": 0.0359, "step": 19539 }, { "epoch": 2.118386816999133, "grad_norm": 0.23800688982009888, "learning_rate": 2.9493689249963734e-05, "loss": 0.0094, "step": 19540 }, { "epoch": 2.1184952298352124, "grad_norm": 0.06906338036060333, "learning_rate": 2.94900623821268e-05, "loss": 0.0011, "step": 19541 }, { "epoch": 2.1186036426712924, "grad_norm": 0.23612529039382935, "learning_rate": 2.9486435514289862e-05, "loss": 0.0063, "step": 19542 }, { "epoch": 2.118712055507372, "grad_norm": 0.5668684840202332, "learning_rate": 2.9482808646452927e-05, "loss": 0.0086, "step": 19543 }, { "epoch": 2.118820468343452, "grad_norm": 0.42139849066734314, "learning_rate": 2.9479181778615987e-05, "loss": 0.0127, "step": 19544 }, { "epoch": 2.1189288811795315, "grad_norm": 0.03134675323963165, "learning_rate": 2.947555491077905e-05, "loss": 0.0013, "step": 19545 }, { "epoch": 2.1190372940156115, "grad_norm": 0.21984080970287323, "learning_rate": 2.9471928042942115e-05, "loss": 0.0063, "step": 19546 }, { "epoch": 2.1191457068516915, "grad_norm": 0.05323709547519684, "learning_rate": 2.946830117510518e-05, "loss": 0.0023, "step": 19547 }, { "epoch": 2.119254119687771, "grad_norm": 0.18772660195827484, "learning_rate": 2.946467430726824e-05, "loss": 0.008, "step": 19548 }, { "epoch": 2.119362532523851, "grad_norm": 0.14772364497184753, "learning_rate": 2.946104743943131e-05, "loss": 0.0163, "step": 19549 }, { "epoch": 2.1194709453599305, "grad_norm": 0.10488588362932205, "learning_rate": 2.9457420571594375e-05, "loss": 0.0017, "step": 19550 }, { "epoch": 2.1195793581960105, "grad_norm": 0.34405064582824707, "learning_rate": 2.9453793703757436e-05, "loss": 0.0093, "step": 19551 }, { "epoch": 2.11968777103209, "grad_norm": 0.059062764048576355, "learning_rate": 2.94501668359205e-05, "loss": 0.0041, "step": 19552 }, { "epoch": 2.11979618386817, "grad_norm": 0.22532565891742706, "learning_rate": 2.9446539968083564e-05, "loss": 0.0071, "step": 19553 }, { "epoch": 2.1199045967042496, "grad_norm": 0.30754444003105164, "learning_rate": 2.944291310024663e-05, "loss": 0.0345, "step": 19554 }, { "epoch": 2.1200130095403296, "grad_norm": 1.1261539459228516, "learning_rate": 2.9439286232409692e-05, "loss": 0.0188, "step": 19555 }, { "epoch": 2.1201214223764095, "grad_norm": 1.030835747718811, "learning_rate": 2.9435659364572753e-05, "loss": 0.0145, "step": 19556 }, { "epoch": 2.120229835212489, "grad_norm": 0.16143111884593964, "learning_rate": 2.9432032496735824e-05, "loss": 0.007, "step": 19557 }, { "epoch": 2.120338248048569, "grad_norm": 0.3264698088169098, "learning_rate": 2.9428405628898885e-05, "loss": 0.0134, "step": 19558 }, { "epoch": 2.1204466608846486, "grad_norm": 0.692852258682251, "learning_rate": 2.942477876106195e-05, "loss": 0.0141, "step": 19559 }, { "epoch": 2.1205550737207286, "grad_norm": 0.11570455878973007, "learning_rate": 2.9421151893225013e-05, "loss": 0.0087, "step": 19560 }, { "epoch": 2.120663486556808, "grad_norm": 0.07721903175115585, "learning_rate": 2.9417525025388077e-05, "loss": 0.0016, "step": 19561 }, { "epoch": 2.120771899392888, "grad_norm": 0.03834002465009689, "learning_rate": 2.941389815755114e-05, "loss": 0.0013, "step": 19562 }, { "epoch": 2.120880312228968, "grad_norm": 0.030109722167253494, "learning_rate": 2.9410271289714202e-05, "loss": 0.0013, "step": 19563 }, { "epoch": 2.1209887250650477, "grad_norm": 0.04420950263738632, "learning_rate": 2.9406644421877266e-05, "loss": 0.0011, "step": 19564 }, { "epoch": 2.1210971379011276, "grad_norm": 0.4227501451969147, "learning_rate": 2.940301755404033e-05, "loss": 0.0199, "step": 19565 }, { "epoch": 2.121205550737207, "grad_norm": 0.21245485544204712, "learning_rate": 2.9399390686203398e-05, "loss": 0.0094, "step": 19566 }, { "epoch": 2.121313963573287, "grad_norm": 0.29343312978744507, "learning_rate": 2.939576381836646e-05, "loss": 0.0204, "step": 19567 }, { "epoch": 2.1214223764093667, "grad_norm": 0.04414821416139603, "learning_rate": 2.9392136950529526e-05, "loss": 0.0017, "step": 19568 }, { "epoch": 2.1215307892454467, "grad_norm": 0.0632801502943039, "learning_rate": 2.938851008269259e-05, "loss": 0.0029, "step": 19569 }, { "epoch": 2.1216392020815267, "grad_norm": 0.6726939678192139, "learning_rate": 2.938488321485565e-05, "loss": 0.0533, "step": 19570 }, { "epoch": 2.121747614917606, "grad_norm": 0.08243368566036224, "learning_rate": 2.9381256347018715e-05, "loss": 0.0069, "step": 19571 }, { "epoch": 2.121856027753686, "grad_norm": 0.03232554718852043, "learning_rate": 2.937762947918178e-05, "loss": 0.0009, "step": 19572 }, { "epoch": 2.1219644405897657, "grad_norm": 0.2467309832572937, "learning_rate": 2.9374002611344843e-05, "loss": 0.0126, "step": 19573 }, { "epoch": 2.1220728534258457, "grad_norm": 0.0508660189807415, "learning_rate": 2.9370375743507904e-05, "loss": 0.0021, "step": 19574 }, { "epoch": 2.1221812662619253, "grad_norm": 0.17441096901893616, "learning_rate": 2.9366748875670975e-05, "loss": 0.0108, "step": 19575 }, { "epoch": 2.1222896790980053, "grad_norm": 0.26809123158454895, "learning_rate": 2.936312200783404e-05, "loss": 0.0037, "step": 19576 }, { "epoch": 2.122398091934085, "grad_norm": 0.14266882836818695, "learning_rate": 2.93594951399971e-05, "loss": 0.0022, "step": 19577 }, { "epoch": 2.122506504770165, "grad_norm": 0.17604169249534607, "learning_rate": 2.9355868272160163e-05, "loss": 0.0025, "step": 19578 }, { "epoch": 2.1226149176062448, "grad_norm": 0.033220838755369186, "learning_rate": 2.9352241404323228e-05, "loss": 0.0012, "step": 19579 }, { "epoch": 2.1227233304423243, "grad_norm": 0.20559200644493103, "learning_rate": 2.934861453648629e-05, "loss": 0.0041, "step": 19580 }, { "epoch": 2.1228317432784043, "grad_norm": 0.10819020122289658, "learning_rate": 2.9344987668649356e-05, "loss": 0.0042, "step": 19581 }, { "epoch": 2.122940156114484, "grad_norm": 0.0027315346524119377, "learning_rate": 2.9341360800812416e-05, "loss": 0.0001, "step": 19582 }, { "epoch": 2.123048568950564, "grad_norm": 0.2925894558429718, "learning_rate": 2.9337733932975487e-05, "loss": 0.003, "step": 19583 }, { "epoch": 2.1231569817866434, "grad_norm": 0.12831221520900726, "learning_rate": 2.9334107065138548e-05, "loss": 0.0073, "step": 19584 }, { "epoch": 2.1232653946227233, "grad_norm": 0.09201085567474365, "learning_rate": 2.9330480197301612e-05, "loss": 0.0033, "step": 19585 }, { "epoch": 2.1233738074588033, "grad_norm": 0.30330824851989746, "learning_rate": 2.9326853329464676e-05, "loss": 0.001, "step": 19586 }, { "epoch": 2.123482220294883, "grad_norm": 0.2522023916244507, "learning_rate": 2.932322646162774e-05, "loss": 0.0042, "step": 19587 }, { "epoch": 2.123590633130963, "grad_norm": 0.4506736695766449, "learning_rate": 2.9319599593790805e-05, "loss": 0.0083, "step": 19588 }, { "epoch": 2.1236990459670424, "grad_norm": 1.0260899066925049, "learning_rate": 2.9315972725953865e-05, "loss": 0.0075, "step": 19589 }, { "epoch": 2.1238074588031224, "grad_norm": 0.17546887695789337, "learning_rate": 2.931234585811693e-05, "loss": 0.0042, "step": 19590 }, { "epoch": 2.123915871639202, "grad_norm": 0.4399581849575043, "learning_rate": 2.9308718990279993e-05, "loss": 0.0285, "step": 19591 }, { "epoch": 2.124024284475282, "grad_norm": 0.14149412512779236, "learning_rate": 2.930509212244306e-05, "loss": 0.0057, "step": 19592 }, { "epoch": 2.1241326973113615, "grad_norm": 0.07810509949922562, "learning_rate": 2.9301465254606125e-05, "loss": 0.0022, "step": 19593 }, { "epoch": 2.1242411101474414, "grad_norm": 0.035713084042072296, "learning_rate": 2.929783838676919e-05, "loss": 0.0007, "step": 19594 }, { "epoch": 2.1243495229835214, "grad_norm": 0.8514319658279419, "learning_rate": 2.9294211518932253e-05, "loss": 0.0148, "step": 19595 }, { "epoch": 2.124457935819601, "grad_norm": 0.2192745953798294, "learning_rate": 2.9290584651095314e-05, "loss": 0.0238, "step": 19596 }, { "epoch": 2.124566348655681, "grad_norm": 0.2091880738735199, "learning_rate": 2.9286957783258378e-05, "loss": 0.0058, "step": 19597 }, { "epoch": 2.1246747614917605, "grad_norm": 0.08505495637655258, "learning_rate": 2.9283330915421442e-05, "loss": 0.0018, "step": 19598 }, { "epoch": 2.1247831743278405, "grad_norm": 0.2833898067474365, "learning_rate": 2.9279704047584506e-05, "loss": 0.0059, "step": 19599 }, { "epoch": 2.12489158716392, "grad_norm": 0.38583120703697205, "learning_rate": 2.9276077179747574e-05, "loss": 0.0441, "step": 19600 }, { "epoch": 2.125, "grad_norm": 0.028279872611165047, "learning_rate": 2.9272450311910638e-05, "loss": 0.0007, "step": 19601 }, { "epoch": 2.12510841283608, "grad_norm": 0.35193532705307007, "learning_rate": 2.9268823444073702e-05, "loss": 0.0215, "step": 19602 }, { "epoch": 2.1252168256721595, "grad_norm": 0.29543769359588623, "learning_rate": 2.9265196576236763e-05, "loss": 0.0021, "step": 19603 }, { "epoch": 2.1253252385082395, "grad_norm": 0.3368113934993744, "learning_rate": 2.9261569708399827e-05, "loss": 0.0098, "step": 19604 }, { "epoch": 2.125433651344319, "grad_norm": 0.18113823235034943, "learning_rate": 2.925794284056289e-05, "loss": 0.0051, "step": 19605 }, { "epoch": 2.125542064180399, "grad_norm": 0.39473414421081543, "learning_rate": 2.9254315972725955e-05, "loss": 0.0189, "step": 19606 }, { "epoch": 2.1256504770164786, "grad_norm": 0.3648361265659332, "learning_rate": 2.925068910488902e-05, "loss": 0.0077, "step": 19607 }, { "epoch": 2.1257588898525586, "grad_norm": 0.3181924521923065, "learning_rate": 2.924706223705208e-05, "loss": 0.0332, "step": 19608 }, { "epoch": 2.125867302688638, "grad_norm": 0.056560542434453964, "learning_rate": 2.924343536921515e-05, "loss": 0.0032, "step": 19609 }, { "epoch": 2.125975715524718, "grad_norm": 0.44361215829849243, "learning_rate": 2.923980850137821e-05, "loss": 0.0099, "step": 19610 }, { "epoch": 2.126084128360798, "grad_norm": 0.24182885885238647, "learning_rate": 2.9236181633541276e-05, "loss": 0.0171, "step": 19611 }, { "epoch": 2.1261925411968776, "grad_norm": 0.6075482964515686, "learning_rate": 2.923255476570434e-05, "loss": 0.0445, "step": 19612 }, { "epoch": 2.1263009540329576, "grad_norm": 0.3222310543060303, "learning_rate": 2.9228927897867404e-05, "loss": 0.0139, "step": 19613 }, { "epoch": 2.126409366869037, "grad_norm": 0.0695456936955452, "learning_rate": 2.9225301030030468e-05, "loss": 0.0017, "step": 19614 }, { "epoch": 2.126517779705117, "grad_norm": 0.5908879637718201, "learning_rate": 2.922167416219353e-05, "loss": 0.0189, "step": 19615 }, { "epoch": 2.126626192541197, "grad_norm": 0.032977256923913956, "learning_rate": 2.9218047294356593e-05, "loss": 0.001, "step": 19616 }, { "epoch": 2.1267346053772767, "grad_norm": 0.30510368943214417, "learning_rate": 2.9214420426519657e-05, "loss": 0.0386, "step": 19617 }, { "epoch": 2.1268430182133566, "grad_norm": 0.22163768112659454, "learning_rate": 2.9210793558682724e-05, "loss": 0.01, "step": 19618 }, { "epoch": 2.126951431049436, "grad_norm": 0.05778849124908447, "learning_rate": 2.920716669084579e-05, "loss": 0.0027, "step": 19619 }, { "epoch": 2.127059843885516, "grad_norm": 0.16649176180362701, "learning_rate": 2.9203539823008852e-05, "loss": 0.0054, "step": 19620 }, { "epoch": 2.1271682567215957, "grad_norm": 0.43763867020606995, "learning_rate": 2.9199912955171917e-05, "loss": 0.0283, "step": 19621 }, { "epoch": 2.1272766695576757, "grad_norm": 0.26969656348228455, "learning_rate": 2.9196286087334977e-05, "loss": 0.0121, "step": 19622 }, { "epoch": 2.1273850823937552, "grad_norm": 0.2489357888698578, "learning_rate": 2.919265921949804e-05, "loss": 0.0079, "step": 19623 }, { "epoch": 2.127493495229835, "grad_norm": 0.14616148173809052, "learning_rate": 2.9189032351661106e-05, "loss": 0.0042, "step": 19624 }, { "epoch": 2.127601908065915, "grad_norm": 0.15684647858142853, "learning_rate": 2.918540548382417e-05, "loss": 0.0073, "step": 19625 }, { "epoch": 2.1277103209019947, "grad_norm": 0.7691785097122192, "learning_rate": 2.9181778615987237e-05, "loss": 0.0172, "step": 19626 }, { "epoch": 2.1278187337380747, "grad_norm": 0.2983497083187103, "learning_rate": 2.91781517481503e-05, "loss": 0.0126, "step": 19627 }, { "epoch": 2.1279271465741543, "grad_norm": 0.14911402761936188, "learning_rate": 2.9174524880313365e-05, "loss": 0.0052, "step": 19628 }, { "epoch": 2.1280355594102343, "grad_norm": 0.7084394097328186, "learning_rate": 2.9170898012476426e-05, "loss": 0.0088, "step": 19629 }, { "epoch": 2.128143972246314, "grad_norm": 0.05379252880811691, "learning_rate": 2.916727114463949e-05, "loss": 0.0026, "step": 19630 }, { "epoch": 2.128252385082394, "grad_norm": 0.2560676634311676, "learning_rate": 2.9163644276802554e-05, "loss": 0.0207, "step": 19631 }, { "epoch": 2.1283607979184733, "grad_norm": 0.009684489108622074, "learning_rate": 2.916001740896562e-05, "loss": 0.0004, "step": 19632 }, { "epoch": 2.1284692107545533, "grad_norm": 0.2971743941307068, "learning_rate": 2.9156390541128682e-05, "loss": 0.004, "step": 19633 }, { "epoch": 2.1285776235906333, "grad_norm": 0.2904013395309448, "learning_rate": 2.9152763673291743e-05, "loss": 0.0215, "step": 19634 }, { "epoch": 2.128686036426713, "grad_norm": 0.11635518819093704, "learning_rate": 2.9149136805454814e-05, "loss": 0.0128, "step": 19635 }, { "epoch": 2.128794449262793, "grad_norm": 0.3762355446815491, "learning_rate": 2.9145509937617875e-05, "loss": 0.0127, "step": 19636 }, { "epoch": 2.1289028620988724, "grad_norm": 1.3296804428100586, "learning_rate": 2.914188306978094e-05, "loss": 0.0186, "step": 19637 }, { "epoch": 2.1290112749349523, "grad_norm": 0.8120493292808533, "learning_rate": 2.9138256201944003e-05, "loss": 0.0152, "step": 19638 }, { "epoch": 2.129119687771032, "grad_norm": 0.18142953515052795, "learning_rate": 2.9134629334107067e-05, "loss": 0.0056, "step": 19639 }, { "epoch": 2.129228100607112, "grad_norm": 0.35390421748161316, "learning_rate": 2.913100246627013e-05, "loss": 0.0627, "step": 19640 }, { "epoch": 2.129336513443192, "grad_norm": 0.08741144835948944, "learning_rate": 2.9127375598433192e-05, "loss": 0.0016, "step": 19641 }, { "epoch": 2.1294449262792714, "grad_norm": 0.3470548391342163, "learning_rate": 2.9123748730596256e-05, "loss": 0.0232, "step": 19642 }, { "epoch": 2.1295533391153514, "grad_norm": 0.3622577488422394, "learning_rate": 2.912012186275932e-05, "loss": 0.0173, "step": 19643 }, { "epoch": 2.129661751951431, "grad_norm": 0.14012156426906586, "learning_rate": 2.9116494994922388e-05, "loss": 0.0105, "step": 19644 }, { "epoch": 2.129770164787511, "grad_norm": 0.00735832192003727, "learning_rate": 2.9112868127085452e-05, "loss": 0.0004, "step": 19645 }, { "epoch": 2.1298785776235905, "grad_norm": 0.42535680532455444, "learning_rate": 2.9109241259248516e-05, "loss": 0.031, "step": 19646 }, { "epoch": 2.1299869904596704, "grad_norm": 0.01762431301176548, "learning_rate": 2.910561439141158e-05, "loss": 0.0005, "step": 19647 }, { "epoch": 2.1300954032957504, "grad_norm": 0.7214445471763611, "learning_rate": 2.910198752357464e-05, "loss": 0.0695, "step": 19648 }, { "epoch": 2.13020381613183, "grad_norm": 0.10819759964942932, "learning_rate": 2.9098360655737705e-05, "loss": 0.0101, "step": 19649 }, { "epoch": 2.13031222896791, "grad_norm": 0.025900155305862427, "learning_rate": 2.909473378790077e-05, "loss": 0.0009, "step": 19650 }, { "epoch": 2.1304206418039895, "grad_norm": 0.13618843257427216, "learning_rate": 2.9091106920063833e-05, "loss": 0.0019, "step": 19651 }, { "epoch": 2.1305290546400695, "grad_norm": 0.012333321385085583, "learning_rate": 2.90874800522269e-05, "loss": 0.0002, "step": 19652 }, { "epoch": 2.130637467476149, "grad_norm": 0.02726898528635502, "learning_rate": 2.9083853184389965e-05, "loss": 0.0016, "step": 19653 }, { "epoch": 2.130745880312229, "grad_norm": 0.2608853578567505, "learning_rate": 2.908022631655303e-05, "loss": 0.0115, "step": 19654 }, { "epoch": 2.1308542931483085, "grad_norm": 0.1978541910648346, "learning_rate": 2.907659944871609e-05, "loss": 0.0105, "step": 19655 }, { "epoch": 2.1309627059843885, "grad_norm": 0.023885900154709816, "learning_rate": 2.9072972580879153e-05, "loss": 0.0009, "step": 19656 }, { "epoch": 2.1310711188204685, "grad_norm": 0.5370809435844421, "learning_rate": 2.9069345713042218e-05, "loss": 0.0215, "step": 19657 }, { "epoch": 2.131179531656548, "grad_norm": 0.5544674396514893, "learning_rate": 2.9065718845205282e-05, "loss": 0.0145, "step": 19658 }, { "epoch": 2.131287944492628, "grad_norm": 0.02597721852362156, "learning_rate": 2.9062091977368346e-05, "loss": 0.0006, "step": 19659 }, { "epoch": 2.1313963573287076, "grad_norm": 0.03734906017780304, "learning_rate": 2.9058465109531407e-05, "loss": 0.0016, "step": 19660 }, { "epoch": 2.1315047701647876, "grad_norm": 0.1321205049753189, "learning_rate": 2.9054838241694477e-05, "loss": 0.0037, "step": 19661 }, { "epoch": 2.131613183000867, "grad_norm": 0.1301378309726715, "learning_rate": 2.9051211373857538e-05, "loss": 0.0083, "step": 19662 }, { "epoch": 2.131721595836947, "grad_norm": 0.5427965521812439, "learning_rate": 2.9047584506020602e-05, "loss": 0.0122, "step": 19663 }, { "epoch": 2.131830008673027, "grad_norm": 0.030834397301077843, "learning_rate": 2.9043957638183666e-05, "loss": 0.0011, "step": 19664 }, { "epoch": 2.1319384215091066, "grad_norm": 0.34965553879737854, "learning_rate": 2.904033077034673e-05, "loss": 0.0338, "step": 19665 }, { "epoch": 2.1320468343451866, "grad_norm": 0.03643196448683739, "learning_rate": 2.9036703902509795e-05, "loss": 0.0009, "step": 19666 }, { "epoch": 2.132155247181266, "grad_norm": 0.12780074775218964, "learning_rate": 2.9033077034672855e-05, "loss": 0.007, "step": 19667 }, { "epoch": 2.132263660017346, "grad_norm": 0.5097417831420898, "learning_rate": 2.902945016683592e-05, "loss": 0.0266, "step": 19668 }, { "epoch": 2.1323720728534257, "grad_norm": 0.10821370780467987, "learning_rate": 2.9025823298998983e-05, "loss": 0.0038, "step": 19669 }, { "epoch": 2.1324804856895057, "grad_norm": 0.36361220479011536, "learning_rate": 2.902219643116205e-05, "loss": 0.0125, "step": 19670 }, { "epoch": 2.1325888985255856, "grad_norm": 0.48363426327705383, "learning_rate": 2.9018569563325115e-05, "loss": 0.0366, "step": 19671 }, { "epoch": 2.132697311361665, "grad_norm": 0.5022659301757812, "learning_rate": 2.901494269548818e-05, "loss": 0.023, "step": 19672 }, { "epoch": 2.132805724197745, "grad_norm": 0.7725353837013245, "learning_rate": 2.9011315827651243e-05, "loss": 0.0783, "step": 19673 }, { "epoch": 2.1329141370338247, "grad_norm": 0.3437465727329254, "learning_rate": 2.9007688959814304e-05, "loss": 0.0162, "step": 19674 }, { "epoch": 2.1330225498699047, "grad_norm": 0.2132890671491623, "learning_rate": 2.9004062091977368e-05, "loss": 0.0102, "step": 19675 }, { "epoch": 2.1331309627059842, "grad_norm": 0.12217406183481216, "learning_rate": 2.9000435224140432e-05, "loss": 0.0044, "step": 19676 }, { "epoch": 2.133239375542064, "grad_norm": 0.7955842614173889, "learning_rate": 2.8996808356303496e-05, "loss": 0.012, "step": 19677 }, { "epoch": 2.1333477883781438, "grad_norm": 0.5799275636672974, "learning_rate": 2.8993181488466564e-05, "loss": 0.0347, "step": 19678 }, { "epoch": 2.1334562012142237, "grad_norm": 0.19221259653568268, "learning_rate": 2.8989554620629628e-05, "loss": 0.0148, "step": 19679 }, { "epoch": 2.1335646140503037, "grad_norm": 0.45539969205856323, "learning_rate": 2.8985927752792692e-05, "loss": 0.013, "step": 19680 }, { "epoch": 2.1336730268863833, "grad_norm": 0.25117772817611694, "learning_rate": 2.8982300884955753e-05, "loss": 0.0096, "step": 19681 }, { "epoch": 2.1337814397224633, "grad_norm": 0.4164513647556305, "learning_rate": 2.8978674017118817e-05, "loss": 0.0282, "step": 19682 }, { "epoch": 2.133889852558543, "grad_norm": 0.09868612885475159, "learning_rate": 2.897504714928188e-05, "loss": 0.0015, "step": 19683 }, { "epoch": 2.133998265394623, "grad_norm": 0.12783555686473846, "learning_rate": 2.8971420281444945e-05, "loss": 0.0088, "step": 19684 }, { "epoch": 2.1341066782307023, "grad_norm": 0.062341589480638504, "learning_rate": 2.896779341360801e-05, "loss": 0.0041, "step": 19685 }, { "epoch": 2.1342150910667823, "grad_norm": 0.9115880727767944, "learning_rate": 2.896416654577107e-05, "loss": 0.0333, "step": 19686 }, { "epoch": 2.1343235039028623, "grad_norm": 0.108636274933815, "learning_rate": 2.896053967793414e-05, "loss": 0.0037, "step": 19687 }, { "epoch": 2.134431916738942, "grad_norm": 0.04707226902246475, "learning_rate": 2.89569128100972e-05, "loss": 0.0024, "step": 19688 }, { "epoch": 2.134540329575022, "grad_norm": 0.26575013995170593, "learning_rate": 2.8953285942260266e-05, "loss": 0.0083, "step": 19689 }, { "epoch": 2.1346487424111014, "grad_norm": 0.34725460410118103, "learning_rate": 2.894965907442333e-05, "loss": 0.0103, "step": 19690 }, { "epoch": 2.1347571552471813, "grad_norm": 0.02905023656785488, "learning_rate": 2.8946032206586394e-05, "loss": 0.0016, "step": 19691 }, { "epoch": 2.134865568083261, "grad_norm": 0.046874236315488815, "learning_rate": 2.8942405338749458e-05, "loss": 0.0018, "step": 19692 }, { "epoch": 2.134973980919341, "grad_norm": 0.33627933263778687, "learning_rate": 2.893877847091252e-05, "loss": 0.0055, "step": 19693 }, { "epoch": 2.135082393755421, "grad_norm": 0.12068265676498413, "learning_rate": 2.8935151603075583e-05, "loss": 0.0101, "step": 19694 }, { "epoch": 2.1351908065915004, "grad_norm": 0.7013906240463257, "learning_rate": 2.8931524735238654e-05, "loss": 0.032, "step": 19695 }, { "epoch": 2.1352992194275804, "grad_norm": 0.0787893533706665, "learning_rate": 2.8927897867401714e-05, "loss": 0.0028, "step": 19696 }, { "epoch": 2.13540763226366, "grad_norm": 0.14682748913764954, "learning_rate": 2.892427099956478e-05, "loss": 0.0046, "step": 19697 }, { "epoch": 2.13551604509974, "grad_norm": 0.02182028256356716, "learning_rate": 2.8920644131727843e-05, "loss": 0.0004, "step": 19698 }, { "epoch": 2.1356244579358195, "grad_norm": 0.04008963331580162, "learning_rate": 2.8917017263890907e-05, "loss": 0.0023, "step": 19699 }, { "epoch": 2.1357328707718994, "grad_norm": 0.11137339472770691, "learning_rate": 2.8913390396053967e-05, "loss": 0.0043, "step": 19700 }, { "epoch": 2.135841283607979, "grad_norm": 0.11367831379175186, "learning_rate": 2.890976352821703e-05, "loss": 0.0037, "step": 19701 }, { "epoch": 2.135949696444059, "grad_norm": 0.4233419895172119, "learning_rate": 2.8906136660380096e-05, "loss": 0.0433, "step": 19702 }, { "epoch": 2.136058109280139, "grad_norm": 0.01007490698248148, "learning_rate": 2.890250979254316e-05, "loss": 0.0004, "step": 19703 }, { "epoch": 2.1361665221162185, "grad_norm": 0.3721123933792114, "learning_rate": 2.8898882924706227e-05, "loss": 0.0213, "step": 19704 }, { "epoch": 2.1362749349522985, "grad_norm": 0.6599748134613037, "learning_rate": 2.889525605686929e-05, "loss": 0.0416, "step": 19705 }, { "epoch": 2.136383347788378, "grad_norm": 0.1985107958316803, "learning_rate": 2.8891629189032355e-05, "loss": 0.0043, "step": 19706 }, { "epoch": 2.136491760624458, "grad_norm": 0.5315186977386475, "learning_rate": 2.8888002321195416e-05, "loss": 0.0233, "step": 19707 }, { "epoch": 2.1366001734605375, "grad_norm": 0.24313895404338837, "learning_rate": 2.888437545335848e-05, "loss": 0.01, "step": 19708 }, { "epoch": 2.1367085862966175, "grad_norm": 0.5363327264785767, "learning_rate": 2.8880748585521544e-05, "loss": 0.029, "step": 19709 }, { "epoch": 2.1368169991326975, "grad_norm": 0.06696970015764236, "learning_rate": 2.887712171768461e-05, "loss": 0.0034, "step": 19710 }, { "epoch": 2.136925411968777, "grad_norm": 0.1375177800655365, "learning_rate": 2.8873494849847672e-05, "loss": 0.0054, "step": 19711 }, { "epoch": 2.137033824804857, "grad_norm": 0.338117778301239, "learning_rate": 2.8869867982010733e-05, "loss": 0.0043, "step": 19712 }, { "epoch": 2.1371422376409366, "grad_norm": 0.058297548443078995, "learning_rate": 2.8866241114173804e-05, "loss": 0.001, "step": 19713 }, { "epoch": 2.1372506504770166, "grad_norm": 0.09387915581464767, "learning_rate": 2.8862614246336865e-05, "loss": 0.0027, "step": 19714 }, { "epoch": 2.137359063313096, "grad_norm": 0.44925519824028015, "learning_rate": 2.885898737849993e-05, "loss": 0.0227, "step": 19715 }, { "epoch": 2.137467476149176, "grad_norm": 0.18892081081867218, "learning_rate": 2.8855360510662993e-05, "loss": 0.0065, "step": 19716 }, { "epoch": 2.137575888985256, "grad_norm": 0.4037776589393616, "learning_rate": 2.8851733642826057e-05, "loss": 0.0153, "step": 19717 }, { "epoch": 2.1376843018213356, "grad_norm": 0.09861689060926437, "learning_rate": 2.884810677498912e-05, "loss": 0.0057, "step": 19718 }, { "epoch": 2.1377927146574156, "grad_norm": 0.28327587246894836, "learning_rate": 2.8844479907152182e-05, "loss": 0.0232, "step": 19719 }, { "epoch": 2.137901127493495, "grad_norm": 0.6938585638999939, "learning_rate": 2.8840853039315246e-05, "loss": 0.0334, "step": 19720 }, { "epoch": 2.138009540329575, "grad_norm": 0.5608080625534058, "learning_rate": 2.8837226171478317e-05, "loss": 0.0391, "step": 19721 }, { "epoch": 2.1381179531656547, "grad_norm": 0.04834356531500816, "learning_rate": 2.8833599303641378e-05, "loss": 0.0017, "step": 19722 }, { "epoch": 2.1382263660017347, "grad_norm": 0.4614926278591156, "learning_rate": 2.8829972435804442e-05, "loss": 0.0128, "step": 19723 }, { "epoch": 2.138334778837814, "grad_norm": 0.368956595659256, "learning_rate": 2.8826345567967506e-05, "loss": 0.0148, "step": 19724 }, { "epoch": 2.138443191673894, "grad_norm": 0.2694704830646515, "learning_rate": 2.882271870013057e-05, "loss": 0.0221, "step": 19725 }, { "epoch": 2.138551604509974, "grad_norm": 0.2038300335407257, "learning_rate": 2.881909183229363e-05, "loss": 0.0074, "step": 19726 }, { "epoch": 2.1386600173460537, "grad_norm": 0.04656725004315376, "learning_rate": 2.8815464964456695e-05, "loss": 0.0017, "step": 19727 }, { "epoch": 2.1387684301821337, "grad_norm": 0.45812249183654785, "learning_rate": 2.881183809661976e-05, "loss": 0.0117, "step": 19728 }, { "epoch": 2.1388768430182132, "grad_norm": 0.1826857328414917, "learning_rate": 2.8808211228782823e-05, "loss": 0.0054, "step": 19729 }, { "epoch": 2.138985255854293, "grad_norm": 0.17565305531024933, "learning_rate": 2.880458436094589e-05, "loss": 0.0059, "step": 19730 }, { "epoch": 2.1390936686903728, "grad_norm": 0.07905510812997818, "learning_rate": 2.8800957493108955e-05, "loss": 0.0079, "step": 19731 }, { "epoch": 2.1392020815264527, "grad_norm": 0.10713163018226624, "learning_rate": 2.879733062527202e-05, "loss": 0.0023, "step": 19732 }, { "epoch": 2.1393104943625323, "grad_norm": 0.020916886627674103, "learning_rate": 2.879370375743508e-05, "loss": 0.0007, "step": 19733 }, { "epoch": 2.1394189071986123, "grad_norm": 0.37344053387641907, "learning_rate": 2.8790076889598144e-05, "loss": 0.0093, "step": 19734 }, { "epoch": 2.1395273200346923, "grad_norm": 0.059648074209690094, "learning_rate": 2.8786450021761208e-05, "loss": 0.0035, "step": 19735 }, { "epoch": 2.139635732870772, "grad_norm": 0.3055761754512787, "learning_rate": 2.8782823153924272e-05, "loss": 0.0056, "step": 19736 }, { "epoch": 2.139744145706852, "grad_norm": 0.0045026070438325405, "learning_rate": 2.8779196286087336e-05, "loss": 0.0002, "step": 19737 }, { "epoch": 2.1398525585429313, "grad_norm": 0.005251294933259487, "learning_rate": 2.8775569418250397e-05, "loss": 0.0001, "step": 19738 }, { "epoch": 2.1399609713790113, "grad_norm": 0.12639950215816498, "learning_rate": 2.8771942550413467e-05, "loss": 0.0033, "step": 19739 }, { "epoch": 2.1400693842150913, "grad_norm": 0.3486785590648651, "learning_rate": 2.876831568257653e-05, "loss": 0.0182, "step": 19740 }, { "epoch": 2.140177797051171, "grad_norm": 0.4031227231025696, "learning_rate": 2.8764688814739592e-05, "loss": 0.0206, "step": 19741 }, { "epoch": 2.140286209887251, "grad_norm": 0.11390207707881927, "learning_rate": 2.8761061946902656e-05, "loss": 0.0038, "step": 19742 }, { "epoch": 2.1403946227233304, "grad_norm": 0.08330009132623672, "learning_rate": 2.875743507906572e-05, "loss": 0.0048, "step": 19743 }, { "epoch": 2.1405030355594103, "grad_norm": 1.3648635149002075, "learning_rate": 2.8753808211228785e-05, "loss": 0.0109, "step": 19744 }, { "epoch": 2.14061144839549, "grad_norm": 0.15555571019649506, "learning_rate": 2.8750181343391845e-05, "loss": 0.0037, "step": 19745 }, { "epoch": 2.14071986123157, "grad_norm": 0.4471617639064789, "learning_rate": 2.874655447555491e-05, "loss": 0.0066, "step": 19746 }, { "epoch": 2.1408282740676494, "grad_norm": 0.0550469309091568, "learning_rate": 2.874292760771798e-05, "loss": 0.0017, "step": 19747 }, { "epoch": 2.1409366869037294, "grad_norm": 0.3296780586242676, "learning_rate": 2.873930073988104e-05, "loss": 0.0126, "step": 19748 }, { "epoch": 2.1410450997398094, "grad_norm": 0.280306339263916, "learning_rate": 2.8735673872044105e-05, "loss": 0.0374, "step": 19749 }, { "epoch": 2.141153512575889, "grad_norm": 0.3491913378238678, "learning_rate": 2.873204700420717e-05, "loss": 0.0338, "step": 19750 }, { "epoch": 2.141261925411969, "grad_norm": 0.6277536749839783, "learning_rate": 2.8728420136370233e-05, "loss": 0.021, "step": 19751 }, { "epoch": 2.1413703382480485, "grad_norm": 0.1766171157360077, "learning_rate": 2.8724793268533294e-05, "loss": 0.0113, "step": 19752 }, { "epoch": 2.1414787510841284, "grad_norm": 0.02402045950293541, "learning_rate": 2.8721166400696358e-05, "loss": 0.0014, "step": 19753 }, { "epoch": 2.141587163920208, "grad_norm": 0.4948946535587311, "learning_rate": 2.8717539532859422e-05, "loss": 0.0102, "step": 19754 }, { "epoch": 2.141695576756288, "grad_norm": 0.15363799035549164, "learning_rate": 2.8713912665022486e-05, "loss": 0.0026, "step": 19755 }, { "epoch": 2.1418039895923675, "grad_norm": 0.3914048373699188, "learning_rate": 2.8710285797185554e-05, "loss": 0.0248, "step": 19756 }, { "epoch": 2.1419124024284475, "grad_norm": 0.1865537464618683, "learning_rate": 2.8706658929348618e-05, "loss": 0.007, "step": 19757 }, { "epoch": 2.1420208152645275, "grad_norm": 0.05617239698767662, "learning_rate": 2.8703032061511682e-05, "loss": 0.0022, "step": 19758 }, { "epoch": 2.142129228100607, "grad_norm": 0.0077605717815458775, "learning_rate": 2.8699405193674743e-05, "loss": 0.0003, "step": 19759 }, { "epoch": 2.142237640936687, "grad_norm": 0.3043464720249176, "learning_rate": 2.8695778325837807e-05, "loss": 0.0081, "step": 19760 }, { "epoch": 2.1423460537727665, "grad_norm": 0.11236158758401871, "learning_rate": 2.869215145800087e-05, "loss": 0.0014, "step": 19761 }, { "epoch": 2.1424544666088465, "grad_norm": 0.4138520359992981, "learning_rate": 2.8688524590163935e-05, "loss": 0.0137, "step": 19762 }, { "epoch": 2.142562879444926, "grad_norm": 0.15211892127990723, "learning_rate": 2.8684897722327e-05, "loss": 0.0031, "step": 19763 }, { "epoch": 2.142671292281006, "grad_norm": 0.2445128858089447, "learning_rate": 2.868127085449006e-05, "loss": 0.0087, "step": 19764 }, { "epoch": 2.142779705117086, "grad_norm": 0.6648997068405151, "learning_rate": 2.867764398665313e-05, "loss": 0.0269, "step": 19765 }, { "epoch": 2.1428881179531656, "grad_norm": 0.3217805027961731, "learning_rate": 2.8674017118816195e-05, "loss": 0.0256, "step": 19766 }, { "epoch": 2.1429965307892456, "grad_norm": 0.14049778878688812, "learning_rate": 2.8670390250979256e-05, "loss": 0.0076, "step": 19767 }, { "epoch": 2.143104943625325, "grad_norm": 0.09753194451332092, "learning_rate": 2.866676338314232e-05, "loss": 0.004, "step": 19768 }, { "epoch": 2.143213356461405, "grad_norm": 0.007623621728271246, "learning_rate": 2.8663136515305384e-05, "loss": 0.0004, "step": 19769 }, { "epoch": 2.1433217692974846, "grad_norm": 0.037709791213274, "learning_rate": 2.8659509647468448e-05, "loss": 0.0012, "step": 19770 }, { "epoch": 2.1434301821335646, "grad_norm": 0.08213063329458237, "learning_rate": 2.865588277963151e-05, "loss": 0.002, "step": 19771 }, { "epoch": 2.1435385949696446, "grad_norm": 0.021854370832443237, "learning_rate": 2.8652255911794573e-05, "loss": 0.0011, "step": 19772 }, { "epoch": 2.143647007805724, "grad_norm": 0.2275741994380951, "learning_rate": 2.8648629043957644e-05, "loss": 0.0048, "step": 19773 }, { "epoch": 2.143755420641804, "grad_norm": 0.21674844622612, "learning_rate": 2.8645002176120704e-05, "loss": 0.0105, "step": 19774 }, { "epoch": 2.1438638334778837, "grad_norm": 0.34723392128944397, "learning_rate": 2.864137530828377e-05, "loss": 0.0066, "step": 19775 }, { "epoch": 2.1439722463139637, "grad_norm": 0.7790408134460449, "learning_rate": 2.8637748440446833e-05, "loss": 0.0447, "step": 19776 }, { "epoch": 2.144080659150043, "grad_norm": 0.21463848650455475, "learning_rate": 2.8634121572609897e-05, "loss": 0.0045, "step": 19777 }, { "epoch": 2.144189071986123, "grad_norm": 0.03360318765044212, "learning_rate": 2.8630494704772957e-05, "loss": 0.0008, "step": 19778 }, { "epoch": 2.1442974848222027, "grad_norm": 0.5309805274009705, "learning_rate": 2.862686783693602e-05, "loss": 0.0491, "step": 19779 }, { "epoch": 2.1444058976582827, "grad_norm": 0.02770155295729637, "learning_rate": 2.8623240969099086e-05, "loss": 0.0011, "step": 19780 }, { "epoch": 2.1445143104943627, "grad_norm": 0.13038049638271332, "learning_rate": 2.861961410126215e-05, "loss": 0.0043, "step": 19781 }, { "epoch": 2.1446227233304422, "grad_norm": 0.38056427240371704, "learning_rate": 2.8615987233425217e-05, "loss": 0.0066, "step": 19782 }, { "epoch": 2.1447311361665222, "grad_norm": 0.4241929352283478, "learning_rate": 2.861236036558828e-05, "loss": 0.016, "step": 19783 }, { "epoch": 2.1448395490026018, "grad_norm": 0.5726903676986694, "learning_rate": 2.8608733497751345e-05, "loss": 0.0138, "step": 19784 }, { "epoch": 2.1449479618386817, "grad_norm": 0.2566996216773987, "learning_rate": 2.8605106629914406e-05, "loss": 0.0101, "step": 19785 }, { "epoch": 2.1450563746747613, "grad_norm": 0.052581485360860825, "learning_rate": 2.860147976207747e-05, "loss": 0.0018, "step": 19786 }, { "epoch": 2.1451647875108413, "grad_norm": 0.14301583170890808, "learning_rate": 2.8597852894240534e-05, "loss": 0.0111, "step": 19787 }, { "epoch": 2.1452732003469213, "grad_norm": 0.4737410545349121, "learning_rate": 2.85942260264036e-05, "loss": 0.0059, "step": 19788 }, { "epoch": 2.145381613183001, "grad_norm": 0.20228904485702515, "learning_rate": 2.8590599158566663e-05, "loss": 0.0096, "step": 19789 }, { "epoch": 2.145490026019081, "grad_norm": 0.04018872231245041, "learning_rate": 2.858697229072973e-05, "loss": 0.001, "step": 19790 }, { "epoch": 2.1455984388551603, "grad_norm": 0.18245583772659302, "learning_rate": 2.8583345422892794e-05, "loss": 0.0221, "step": 19791 }, { "epoch": 2.1457068516912403, "grad_norm": 0.13346076011657715, "learning_rate": 2.8579718555055858e-05, "loss": 0.0074, "step": 19792 }, { "epoch": 2.14581526452732, "grad_norm": 0.17884491384029388, "learning_rate": 2.857609168721892e-05, "loss": 0.0051, "step": 19793 }, { "epoch": 2.1459236773634, "grad_norm": 0.13369064033031464, "learning_rate": 2.8572464819381983e-05, "loss": 0.0021, "step": 19794 }, { "epoch": 2.14603209019948, "grad_norm": 0.4117506146430969, "learning_rate": 2.8568837951545047e-05, "loss": 0.018, "step": 19795 }, { "epoch": 2.1461405030355594, "grad_norm": 0.05201709270477295, "learning_rate": 2.856521108370811e-05, "loss": 0.0029, "step": 19796 }, { "epoch": 2.1462489158716394, "grad_norm": 0.3756210505962372, "learning_rate": 2.8561584215871172e-05, "loss": 0.0234, "step": 19797 }, { "epoch": 2.146357328707719, "grad_norm": 0.44771891832351685, "learning_rate": 2.8557957348034236e-05, "loss": 0.0106, "step": 19798 }, { "epoch": 2.146465741543799, "grad_norm": 0.4363272786140442, "learning_rate": 2.8554330480197307e-05, "loss": 0.0076, "step": 19799 }, { "epoch": 2.1465741543798784, "grad_norm": 0.10535752028226852, "learning_rate": 2.8550703612360368e-05, "loss": 0.0037, "step": 19800 }, { "epoch": 2.1466825672159584, "grad_norm": 0.39787524938583374, "learning_rate": 2.8547076744523432e-05, "loss": 0.0273, "step": 19801 }, { "epoch": 2.146790980052038, "grad_norm": 0.3672882914543152, "learning_rate": 2.8543449876686496e-05, "loss": 0.0125, "step": 19802 }, { "epoch": 2.146899392888118, "grad_norm": 0.032280225306749344, "learning_rate": 2.853982300884956e-05, "loss": 0.001, "step": 19803 }, { "epoch": 2.147007805724198, "grad_norm": 0.009266696870326996, "learning_rate": 2.853619614101262e-05, "loss": 0.0003, "step": 19804 }, { "epoch": 2.1471162185602775, "grad_norm": 0.6083958148956299, "learning_rate": 2.8532569273175685e-05, "loss": 0.0438, "step": 19805 }, { "epoch": 2.1472246313963574, "grad_norm": 0.3913974463939667, "learning_rate": 2.852894240533875e-05, "loss": 0.0117, "step": 19806 }, { "epoch": 2.147333044232437, "grad_norm": 0.15604518353939056, "learning_rate": 2.8525315537501813e-05, "loss": 0.0047, "step": 19807 }, { "epoch": 2.147441457068517, "grad_norm": 0.16488860547542572, "learning_rate": 2.852168866966488e-05, "loss": 0.0045, "step": 19808 }, { "epoch": 2.1475498699045965, "grad_norm": 0.3259908854961395, "learning_rate": 2.8518061801827945e-05, "loss": 0.0202, "step": 19809 }, { "epoch": 2.1476582827406765, "grad_norm": 0.13416872918605804, "learning_rate": 2.851443493399101e-05, "loss": 0.0021, "step": 19810 }, { "epoch": 2.1477666955767565, "grad_norm": 0.10452253371477127, "learning_rate": 2.851080806615407e-05, "loss": 0.0038, "step": 19811 }, { "epoch": 2.147875108412836, "grad_norm": 0.26034724712371826, "learning_rate": 2.8507181198317134e-05, "loss": 0.0136, "step": 19812 }, { "epoch": 2.147983521248916, "grad_norm": 1.3848375082015991, "learning_rate": 2.8503554330480198e-05, "loss": 0.0287, "step": 19813 }, { "epoch": 2.1480919340849955, "grad_norm": 0.052568044513463974, "learning_rate": 2.8499927462643262e-05, "loss": 0.0016, "step": 19814 }, { "epoch": 2.1482003469210755, "grad_norm": 0.42910948395729065, "learning_rate": 2.8496300594806326e-05, "loss": 0.0127, "step": 19815 }, { "epoch": 2.148308759757155, "grad_norm": 0.20970292389392853, "learning_rate": 2.8492673726969393e-05, "loss": 0.0088, "step": 19816 }, { "epoch": 2.148417172593235, "grad_norm": 0.10885448753833771, "learning_rate": 2.8489046859132457e-05, "loss": 0.003, "step": 19817 }, { "epoch": 2.148525585429315, "grad_norm": 0.08912988752126694, "learning_rate": 2.848541999129552e-05, "loss": 0.0045, "step": 19818 }, { "epoch": 2.1486339982653946, "grad_norm": 0.3608379364013672, "learning_rate": 2.8481793123458582e-05, "loss": 0.0078, "step": 19819 }, { "epoch": 2.1487424111014746, "grad_norm": 0.20527294278144836, "learning_rate": 2.8478166255621646e-05, "loss": 0.0123, "step": 19820 }, { "epoch": 2.148850823937554, "grad_norm": 2.8338382244110107, "learning_rate": 2.847453938778471e-05, "loss": 0.0283, "step": 19821 }, { "epoch": 2.148959236773634, "grad_norm": 0.8526871204376221, "learning_rate": 2.8470912519947775e-05, "loss": 0.0241, "step": 19822 }, { "epoch": 2.1490676496097136, "grad_norm": 0.98990398645401, "learning_rate": 2.8467285652110835e-05, "loss": 0.0255, "step": 19823 }, { "epoch": 2.1491760624457936, "grad_norm": 0.3745395243167877, "learning_rate": 2.84636587842739e-05, "loss": 0.0106, "step": 19824 }, { "epoch": 2.149284475281873, "grad_norm": 0.008205831050872803, "learning_rate": 2.846003191643697e-05, "loss": 0.0004, "step": 19825 }, { "epoch": 2.149392888117953, "grad_norm": 0.10439944267272949, "learning_rate": 2.845640504860003e-05, "loss": 0.0042, "step": 19826 }, { "epoch": 2.149501300954033, "grad_norm": 0.28059306740760803, "learning_rate": 2.8452778180763095e-05, "loss": 0.0149, "step": 19827 }, { "epoch": 2.1496097137901127, "grad_norm": 0.6366961598396301, "learning_rate": 2.844915131292616e-05, "loss": 0.06, "step": 19828 }, { "epoch": 2.1497181266261927, "grad_norm": 0.05154072865843773, "learning_rate": 2.8445524445089223e-05, "loss": 0.0016, "step": 19829 }, { "epoch": 2.149826539462272, "grad_norm": 0.2116784155368805, "learning_rate": 2.8441897577252284e-05, "loss": 0.0116, "step": 19830 }, { "epoch": 2.149934952298352, "grad_norm": 0.020998576655983925, "learning_rate": 2.8438270709415348e-05, "loss": 0.0006, "step": 19831 }, { "epoch": 2.1500433651344317, "grad_norm": 0.41963115334510803, "learning_rate": 2.8434643841578412e-05, "loss": 0.0137, "step": 19832 }, { "epoch": 2.1501517779705117, "grad_norm": 0.023892955854535103, "learning_rate": 2.8431016973741476e-05, "loss": 0.0008, "step": 19833 }, { "epoch": 2.1502601908065917, "grad_norm": 1.2680431604385376, "learning_rate": 2.8427390105904544e-05, "loss": 0.0156, "step": 19834 }, { "epoch": 2.1503686036426712, "grad_norm": 0.003673360450193286, "learning_rate": 2.8423763238067608e-05, "loss": 0.0001, "step": 19835 }, { "epoch": 2.1504770164787512, "grad_norm": 0.742850661277771, "learning_rate": 2.8420136370230672e-05, "loss": 0.0111, "step": 19836 }, { "epoch": 2.1505854293148308, "grad_norm": 0.1915694773197174, "learning_rate": 2.8416509502393733e-05, "loss": 0.0078, "step": 19837 }, { "epoch": 2.1506938421509108, "grad_norm": 0.18028859794139862, "learning_rate": 2.8412882634556797e-05, "loss": 0.0052, "step": 19838 }, { "epoch": 2.1508022549869903, "grad_norm": 0.06951835006475449, "learning_rate": 2.840925576671986e-05, "loss": 0.0016, "step": 19839 }, { "epoch": 2.1509106678230703, "grad_norm": 0.07176310569047928, "learning_rate": 2.8405628898882925e-05, "loss": 0.0039, "step": 19840 }, { "epoch": 2.1510190806591503, "grad_norm": 0.2194337248802185, "learning_rate": 2.840200203104599e-05, "loss": 0.0092, "step": 19841 }, { "epoch": 2.15112749349523, "grad_norm": 0.5853887796401978, "learning_rate": 2.8398375163209057e-05, "loss": 0.0238, "step": 19842 }, { "epoch": 2.15123590633131, "grad_norm": 0.01690443977713585, "learning_rate": 2.839474829537212e-05, "loss": 0.0009, "step": 19843 }, { "epoch": 2.1513443191673893, "grad_norm": 0.05284154415130615, "learning_rate": 2.8391121427535185e-05, "loss": 0.0022, "step": 19844 }, { "epoch": 2.1514527320034693, "grad_norm": 0.6446244120597839, "learning_rate": 2.8387494559698246e-05, "loss": 0.0229, "step": 19845 }, { "epoch": 2.151561144839549, "grad_norm": 0.04677470400929451, "learning_rate": 2.838386769186131e-05, "loss": 0.0021, "step": 19846 }, { "epoch": 2.151669557675629, "grad_norm": 0.5297915935516357, "learning_rate": 2.8380240824024374e-05, "loss": 0.0653, "step": 19847 }, { "epoch": 2.1517779705117084, "grad_norm": 0.28213798999786377, "learning_rate": 2.8376613956187438e-05, "loss": 0.01, "step": 19848 }, { "epoch": 2.1518863833477884, "grad_norm": 0.027948234230279922, "learning_rate": 2.83729870883505e-05, "loss": 0.0008, "step": 19849 }, { "epoch": 2.1519947961838684, "grad_norm": 0.25202393531799316, "learning_rate": 2.8369360220513563e-05, "loss": 0.022, "step": 19850 }, { "epoch": 2.152103209019948, "grad_norm": 0.2887209355831146, "learning_rate": 2.8365733352676634e-05, "loss": 0.0159, "step": 19851 }, { "epoch": 2.152211621856028, "grad_norm": 0.2763582468032837, "learning_rate": 2.8362106484839694e-05, "loss": 0.0166, "step": 19852 }, { "epoch": 2.1523200346921074, "grad_norm": 0.17904502153396606, "learning_rate": 2.835847961700276e-05, "loss": 0.0183, "step": 19853 }, { "epoch": 2.1524284475281874, "grad_norm": 0.29696381092071533, "learning_rate": 2.8354852749165823e-05, "loss": 0.0061, "step": 19854 }, { "epoch": 2.152536860364267, "grad_norm": 0.4770019054412842, "learning_rate": 2.8351225881328887e-05, "loss": 0.03, "step": 19855 }, { "epoch": 2.152645273200347, "grad_norm": 0.48364585638046265, "learning_rate": 2.8347599013491947e-05, "loss": 0.0222, "step": 19856 }, { "epoch": 2.152753686036427, "grad_norm": 0.2768019139766693, "learning_rate": 2.834397214565501e-05, "loss": 0.0086, "step": 19857 }, { "epoch": 2.1528620988725065, "grad_norm": 0.1830015927553177, "learning_rate": 2.8340345277818076e-05, "loss": 0.0075, "step": 19858 }, { "epoch": 2.1529705117085864, "grad_norm": 0.18571357429027557, "learning_rate": 2.833671840998114e-05, "loss": 0.009, "step": 19859 }, { "epoch": 2.153078924544666, "grad_norm": 0.4508056044578552, "learning_rate": 2.8333091542144207e-05, "loss": 0.0285, "step": 19860 }, { "epoch": 2.153187337380746, "grad_norm": 0.30407389998435974, "learning_rate": 2.832946467430727e-05, "loss": 0.0362, "step": 19861 }, { "epoch": 2.1532957502168255, "grad_norm": 0.2314799427986145, "learning_rate": 2.8325837806470335e-05, "loss": 0.0152, "step": 19862 }, { "epoch": 2.1534041630529055, "grad_norm": 0.07447698712348938, "learning_rate": 2.8322210938633396e-05, "loss": 0.0011, "step": 19863 }, { "epoch": 2.1535125758889855, "grad_norm": 1.1006735563278198, "learning_rate": 2.831858407079646e-05, "loss": 0.0364, "step": 19864 }, { "epoch": 2.153620988725065, "grad_norm": 0.42659029364585876, "learning_rate": 2.8314957202959524e-05, "loss": 0.0133, "step": 19865 }, { "epoch": 2.153729401561145, "grad_norm": 0.19723235070705414, "learning_rate": 2.831133033512259e-05, "loss": 0.0076, "step": 19866 }, { "epoch": 2.1538378143972245, "grad_norm": 0.37442848086357117, "learning_rate": 2.8307703467285653e-05, "loss": 0.0756, "step": 19867 }, { "epoch": 2.1539462272333045, "grad_norm": 0.6657845377922058, "learning_rate": 2.830407659944872e-05, "loss": 0.0166, "step": 19868 }, { "epoch": 2.154054640069384, "grad_norm": 0.19566498696804047, "learning_rate": 2.8300449731611784e-05, "loss": 0.0096, "step": 19869 }, { "epoch": 2.154163052905464, "grad_norm": 0.05509301647543907, "learning_rate": 2.8296822863774848e-05, "loss": 0.0021, "step": 19870 }, { "epoch": 2.1542714657415436, "grad_norm": 0.17315126955509186, "learning_rate": 2.829319599593791e-05, "loss": 0.0073, "step": 19871 }, { "epoch": 2.1543798785776236, "grad_norm": 0.3728533089160919, "learning_rate": 2.8289569128100973e-05, "loss": 0.0166, "step": 19872 }, { "epoch": 2.1544882914137036, "grad_norm": 0.06829632073640823, "learning_rate": 2.8285942260264037e-05, "loss": 0.0043, "step": 19873 }, { "epoch": 2.154596704249783, "grad_norm": 0.7266457080841064, "learning_rate": 2.82823153924271e-05, "loss": 0.028, "step": 19874 }, { "epoch": 2.154705117085863, "grad_norm": 0.2134793996810913, "learning_rate": 2.8278688524590162e-05, "loss": 0.0045, "step": 19875 }, { "epoch": 2.1548135299219426, "grad_norm": 0.3133321702480316, "learning_rate": 2.8275061656753226e-05, "loss": 0.0071, "step": 19876 }, { "epoch": 2.1549219427580226, "grad_norm": 0.12350417673587799, "learning_rate": 2.8271434788916297e-05, "loss": 0.0033, "step": 19877 }, { "epoch": 2.155030355594102, "grad_norm": 0.32272759079933167, "learning_rate": 2.8267807921079358e-05, "loss": 0.0506, "step": 19878 }, { "epoch": 2.155138768430182, "grad_norm": 0.16260968148708344, "learning_rate": 2.8264181053242422e-05, "loss": 0.0078, "step": 19879 }, { "epoch": 2.1552471812662617, "grad_norm": 0.30350029468536377, "learning_rate": 2.8260554185405486e-05, "loss": 0.0113, "step": 19880 }, { "epoch": 2.1553555941023417, "grad_norm": 0.3612561821937561, "learning_rate": 2.825692731756855e-05, "loss": 0.0105, "step": 19881 }, { "epoch": 2.1554640069384217, "grad_norm": 0.10504119843244553, "learning_rate": 2.825330044973161e-05, "loss": 0.0021, "step": 19882 }, { "epoch": 2.155572419774501, "grad_norm": 0.5114923715591431, "learning_rate": 2.8249673581894675e-05, "loss": 0.0204, "step": 19883 }, { "epoch": 2.155680832610581, "grad_norm": 0.5542790293693542, "learning_rate": 2.824604671405774e-05, "loss": 0.0252, "step": 19884 }, { "epoch": 2.1557892454466607, "grad_norm": 0.33327051997184753, "learning_rate": 2.8242419846220806e-05, "loss": 0.0263, "step": 19885 }, { "epoch": 2.1558976582827407, "grad_norm": 0.24312599003314972, "learning_rate": 2.823879297838387e-05, "loss": 0.005, "step": 19886 }, { "epoch": 2.1560060711188207, "grad_norm": 0.4445796310901642, "learning_rate": 2.8235166110546935e-05, "loss": 0.019, "step": 19887 }, { "epoch": 2.1561144839549002, "grad_norm": 0.2551735043525696, "learning_rate": 2.823153924271e-05, "loss": 0.0176, "step": 19888 }, { "epoch": 2.1562228967909802, "grad_norm": 0.17517393827438354, "learning_rate": 2.822791237487306e-05, "loss": 0.0098, "step": 19889 }, { "epoch": 2.1563313096270598, "grad_norm": 0.3382321000099182, "learning_rate": 2.8224285507036124e-05, "loss": 0.054, "step": 19890 }, { "epoch": 2.1564397224631398, "grad_norm": 0.46605634689331055, "learning_rate": 2.8220658639199188e-05, "loss": 0.0357, "step": 19891 }, { "epoch": 2.1565481352992193, "grad_norm": 0.8365363478660583, "learning_rate": 2.8217031771362252e-05, "loss": 0.0187, "step": 19892 }, { "epoch": 2.1566565481352993, "grad_norm": 0.14181141555309296, "learning_rate": 2.8213404903525316e-05, "loss": 0.0051, "step": 19893 }, { "epoch": 2.156764960971379, "grad_norm": 0.23108907043933868, "learning_rate": 2.8209778035688383e-05, "loss": 0.0077, "step": 19894 }, { "epoch": 2.156873373807459, "grad_norm": 0.24213488399982452, "learning_rate": 2.8206151167851448e-05, "loss": 0.0081, "step": 19895 }, { "epoch": 2.156981786643539, "grad_norm": 0.1277211606502533, "learning_rate": 2.820252430001451e-05, "loss": 0.0065, "step": 19896 }, { "epoch": 2.1570901994796183, "grad_norm": 0.1451846808195114, "learning_rate": 2.8198897432177572e-05, "loss": 0.015, "step": 19897 }, { "epoch": 2.1571986123156983, "grad_norm": 0.1956653743982315, "learning_rate": 2.8195270564340636e-05, "loss": 0.0059, "step": 19898 }, { "epoch": 2.157307025151778, "grad_norm": 0.156326025724411, "learning_rate": 2.81916436965037e-05, "loss": 0.0203, "step": 19899 }, { "epoch": 2.157415437987858, "grad_norm": 0.1137174740433693, "learning_rate": 2.8188016828666765e-05, "loss": 0.0074, "step": 19900 }, { "epoch": 2.1575238508239374, "grad_norm": 0.2755041718482971, "learning_rate": 2.8184389960829825e-05, "loss": 0.0229, "step": 19901 }, { "epoch": 2.1576322636600174, "grad_norm": 0.3252244293689728, "learning_rate": 2.818076309299289e-05, "loss": 0.0327, "step": 19902 }, { "epoch": 2.157740676496097, "grad_norm": 0.6780464053153992, "learning_rate": 2.817713622515596e-05, "loss": 0.0271, "step": 19903 }, { "epoch": 2.157849089332177, "grad_norm": 0.24649979174137115, "learning_rate": 2.817350935731902e-05, "loss": 0.0076, "step": 19904 }, { "epoch": 2.157957502168257, "grad_norm": 0.02567661739885807, "learning_rate": 2.8169882489482085e-05, "loss": 0.0008, "step": 19905 }, { "epoch": 2.1580659150043364, "grad_norm": 0.13826459646224976, "learning_rate": 2.816625562164515e-05, "loss": 0.0021, "step": 19906 }, { "epoch": 2.1581743278404164, "grad_norm": 0.6704341769218445, "learning_rate": 2.8162628753808213e-05, "loss": 0.0343, "step": 19907 }, { "epoch": 2.158282740676496, "grad_norm": 0.33475640416145325, "learning_rate": 2.8159001885971274e-05, "loss": 0.0323, "step": 19908 }, { "epoch": 2.158391153512576, "grad_norm": 0.0948113426566124, "learning_rate": 2.8155375018134338e-05, "loss": 0.004, "step": 19909 }, { "epoch": 2.1584995663486555, "grad_norm": 0.5381160974502563, "learning_rate": 2.8151748150297402e-05, "loss": 0.0112, "step": 19910 }, { "epoch": 2.1586079791847355, "grad_norm": 0.5852712988853455, "learning_rate": 2.814812128246047e-05, "loss": 0.0201, "step": 19911 }, { "epoch": 2.1587163920208154, "grad_norm": 0.07763999700546265, "learning_rate": 2.8144494414623534e-05, "loss": 0.0022, "step": 19912 }, { "epoch": 2.158824804856895, "grad_norm": 0.5337122082710266, "learning_rate": 2.8140867546786598e-05, "loss": 0.0302, "step": 19913 }, { "epoch": 2.158933217692975, "grad_norm": 0.2094637155532837, "learning_rate": 2.8137240678949662e-05, "loss": 0.01, "step": 19914 }, { "epoch": 2.1590416305290545, "grad_norm": 0.05231409892439842, "learning_rate": 2.8133613811112726e-05, "loss": 0.0009, "step": 19915 }, { "epoch": 2.1591500433651345, "grad_norm": 0.1469845175743103, "learning_rate": 2.8129986943275787e-05, "loss": 0.007, "step": 19916 }, { "epoch": 2.159258456201214, "grad_norm": 0.16009508073329926, "learning_rate": 2.812636007543885e-05, "loss": 0.0168, "step": 19917 }, { "epoch": 2.159366869037294, "grad_norm": 0.00985963735729456, "learning_rate": 2.8122733207601915e-05, "loss": 0.0003, "step": 19918 }, { "epoch": 2.159475281873374, "grad_norm": 0.3648321032524109, "learning_rate": 2.811910633976498e-05, "loss": 0.0111, "step": 19919 }, { "epoch": 2.1595836947094535, "grad_norm": 0.18573933839797974, "learning_rate": 2.8115479471928047e-05, "loss": 0.0088, "step": 19920 }, { "epoch": 2.1596921075455335, "grad_norm": 0.18147999048233032, "learning_rate": 2.811185260409111e-05, "loss": 0.0045, "step": 19921 }, { "epoch": 2.159800520381613, "grad_norm": 0.08437569439411163, "learning_rate": 2.8108225736254175e-05, "loss": 0.0061, "step": 19922 }, { "epoch": 2.159908933217693, "grad_norm": 0.05043500289320946, "learning_rate": 2.8104598868417236e-05, "loss": 0.0011, "step": 19923 }, { "epoch": 2.1600173460537726, "grad_norm": 0.4720612168312073, "learning_rate": 2.81009720005803e-05, "loss": 0.0065, "step": 19924 }, { "epoch": 2.1601257588898526, "grad_norm": 0.2898027002811432, "learning_rate": 2.8097345132743364e-05, "loss": 0.015, "step": 19925 }, { "epoch": 2.160234171725932, "grad_norm": 0.026346884667873383, "learning_rate": 2.8093718264906428e-05, "loss": 0.001, "step": 19926 }, { "epoch": 2.160342584562012, "grad_norm": 0.05474528670310974, "learning_rate": 2.809009139706949e-05, "loss": 0.0013, "step": 19927 }, { "epoch": 2.160450997398092, "grad_norm": 0.11136192828416824, "learning_rate": 2.8086464529232553e-05, "loss": 0.0035, "step": 19928 }, { "epoch": 2.1605594102341716, "grad_norm": 1.5023589134216309, "learning_rate": 2.8082837661395624e-05, "loss": 0.0244, "step": 19929 }, { "epoch": 2.1606678230702516, "grad_norm": 0.32409366965293884, "learning_rate": 2.8079210793558684e-05, "loss": 0.0113, "step": 19930 }, { "epoch": 2.160776235906331, "grad_norm": 0.25893157720565796, "learning_rate": 2.807558392572175e-05, "loss": 0.0139, "step": 19931 }, { "epoch": 2.160884648742411, "grad_norm": 0.627827525138855, "learning_rate": 2.8071957057884813e-05, "loss": 0.0114, "step": 19932 }, { "epoch": 2.1609930615784907, "grad_norm": 0.23528572916984558, "learning_rate": 2.8068330190047877e-05, "loss": 0.0078, "step": 19933 }, { "epoch": 2.1611014744145707, "grad_norm": 0.22724293172359467, "learning_rate": 2.8064703322210937e-05, "loss": 0.0086, "step": 19934 }, { "epoch": 2.1612098872506507, "grad_norm": 0.012469225563108921, "learning_rate": 2.8061076454374e-05, "loss": 0.0005, "step": 19935 }, { "epoch": 2.16131830008673, "grad_norm": 0.3034713864326477, "learning_rate": 2.8057449586537066e-05, "loss": 0.0158, "step": 19936 }, { "epoch": 2.16142671292281, "grad_norm": 0.14703980088233948, "learning_rate": 2.8053822718700133e-05, "loss": 0.0075, "step": 19937 }, { "epoch": 2.1615351257588897, "grad_norm": 0.4075981676578522, "learning_rate": 2.8050195850863197e-05, "loss": 0.0529, "step": 19938 }, { "epoch": 2.1616435385949697, "grad_norm": 0.12430527806282043, "learning_rate": 2.804656898302626e-05, "loss": 0.0077, "step": 19939 }, { "epoch": 2.1617519514310493, "grad_norm": 0.2664439082145691, "learning_rate": 2.8042942115189325e-05, "loss": 0.01, "step": 19940 }, { "epoch": 2.1618603642671292, "grad_norm": 0.16491781175136566, "learning_rate": 2.803931524735239e-05, "loss": 0.0107, "step": 19941 }, { "epoch": 2.1619687771032092, "grad_norm": 0.6883934736251831, "learning_rate": 2.803568837951545e-05, "loss": 0.0297, "step": 19942 }, { "epoch": 2.1620771899392888, "grad_norm": 0.5646257996559143, "learning_rate": 2.8032061511678514e-05, "loss": 0.028, "step": 19943 }, { "epoch": 2.1621856027753688, "grad_norm": 0.18854288756847382, "learning_rate": 2.802843464384158e-05, "loss": 0.0066, "step": 19944 }, { "epoch": 2.1622940156114483, "grad_norm": 0.14011043310165405, "learning_rate": 2.8024807776004643e-05, "loss": 0.0069, "step": 19945 }, { "epoch": 2.1624024284475283, "grad_norm": 0.46879202127456665, "learning_rate": 2.802118090816771e-05, "loss": 0.0175, "step": 19946 }, { "epoch": 2.162510841283608, "grad_norm": 0.032360151410102844, "learning_rate": 2.8017554040330774e-05, "loss": 0.0014, "step": 19947 }, { "epoch": 2.162619254119688, "grad_norm": 0.24689383804798126, "learning_rate": 2.8013927172493838e-05, "loss": 0.0139, "step": 19948 }, { "epoch": 2.1627276669557673, "grad_norm": 0.07492700964212418, "learning_rate": 2.80103003046569e-05, "loss": 0.0034, "step": 19949 }, { "epoch": 2.1628360797918473, "grad_norm": 0.49635225534439087, "learning_rate": 2.8006673436819963e-05, "loss": 0.0352, "step": 19950 }, { "epoch": 2.1629444926279273, "grad_norm": 0.17893743515014648, "learning_rate": 2.8003046568983027e-05, "loss": 0.0065, "step": 19951 }, { "epoch": 2.163052905464007, "grad_norm": 0.04252991825342178, "learning_rate": 2.799941970114609e-05, "loss": 0.0031, "step": 19952 }, { "epoch": 2.163161318300087, "grad_norm": 0.24151387810707092, "learning_rate": 2.7995792833309152e-05, "loss": 0.0131, "step": 19953 }, { "epoch": 2.1632697311361664, "grad_norm": 0.2027982473373413, "learning_rate": 2.7992165965472216e-05, "loss": 0.0284, "step": 19954 }, { "epoch": 2.1633781439722464, "grad_norm": 0.34030091762542725, "learning_rate": 2.7988539097635287e-05, "loss": 0.0137, "step": 19955 }, { "epoch": 2.163486556808326, "grad_norm": 0.008485047146677971, "learning_rate": 2.7984912229798348e-05, "loss": 0.0004, "step": 19956 }, { "epoch": 2.163594969644406, "grad_norm": 0.560066282749176, "learning_rate": 2.7981285361961412e-05, "loss": 0.0172, "step": 19957 }, { "epoch": 2.163703382480486, "grad_norm": 0.009812576696276665, "learning_rate": 2.7977658494124476e-05, "loss": 0.0005, "step": 19958 }, { "epoch": 2.1638117953165654, "grad_norm": 0.525850772857666, "learning_rate": 2.797403162628754e-05, "loss": 0.0211, "step": 19959 }, { "epoch": 2.1639202081526454, "grad_norm": 0.16523748636245728, "learning_rate": 2.79704047584506e-05, "loss": 0.0145, "step": 19960 }, { "epoch": 2.164028620988725, "grad_norm": 0.19198475778102875, "learning_rate": 2.7966777890613665e-05, "loss": 0.0041, "step": 19961 }, { "epoch": 2.164137033824805, "grad_norm": 0.10394898802042007, "learning_rate": 2.796315102277673e-05, "loss": 0.0032, "step": 19962 }, { "epoch": 2.1642454466608845, "grad_norm": 0.039666276425123215, "learning_rate": 2.7959524154939796e-05, "loss": 0.0011, "step": 19963 }, { "epoch": 2.1643538594969645, "grad_norm": 0.5543839931488037, "learning_rate": 2.795589728710286e-05, "loss": 0.0273, "step": 19964 }, { "epoch": 2.1644622723330444, "grad_norm": 0.16681323945522308, "learning_rate": 2.7952270419265925e-05, "loss": 0.0041, "step": 19965 }, { "epoch": 2.164570685169124, "grad_norm": 0.3193115293979645, "learning_rate": 2.794864355142899e-05, "loss": 0.0151, "step": 19966 }, { "epoch": 2.164679098005204, "grad_norm": 0.17628727853298187, "learning_rate": 2.7945016683592053e-05, "loss": 0.0157, "step": 19967 }, { "epoch": 2.1647875108412835, "grad_norm": 0.09914545714855194, "learning_rate": 2.7941389815755114e-05, "loss": 0.0043, "step": 19968 }, { "epoch": 2.1648959236773635, "grad_norm": 0.7410712242126465, "learning_rate": 2.7937762947918178e-05, "loss": 0.038, "step": 19969 }, { "epoch": 2.165004336513443, "grad_norm": 0.3549085557460785, "learning_rate": 2.7934136080081242e-05, "loss": 0.0159, "step": 19970 }, { "epoch": 2.165112749349523, "grad_norm": 0.4710460603237152, "learning_rate": 2.7930509212244306e-05, "loss": 0.0164, "step": 19971 }, { "epoch": 2.1652211621856026, "grad_norm": 0.24504192173480988, "learning_rate": 2.7926882344407373e-05, "loss": 0.0085, "step": 19972 }, { "epoch": 2.1653295750216826, "grad_norm": 0.16472095251083374, "learning_rate": 2.7923255476570438e-05, "loss": 0.0063, "step": 19973 }, { "epoch": 2.1654379878577625, "grad_norm": 0.47991272807121277, "learning_rate": 2.79196286087335e-05, "loss": 0.0095, "step": 19974 }, { "epoch": 2.165546400693842, "grad_norm": 0.35755234956741333, "learning_rate": 2.7916001740896562e-05, "loss": 0.0224, "step": 19975 }, { "epoch": 2.165654813529922, "grad_norm": 1.0587867498397827, "learning_rate": 2.7912374873059626e-05, "loss": 0.0197, "step": 19976 }, { "epoch": 2.1657632263660016, "grad_norm": 0.13459105789661407, "learning_rate": 2.790874800522269e-05, "loss": 0.0055, "step": 19977 }, { "epoch": 2.1658716392020816, "grad_norm": 0.3463399112224579, "learning_rate": 2.7905121137385755e-05, "loss": 0.0074, "step": 19978 }, { "epoch": 2.165980052038161, "grad_norm": 0.12900064885616302, "learning_rate": 2.7901494269548815e-05, "loss": 0.0088, "step": 19979 }, { "epoch": 2.166088464874241, "grad_norm": 0.28447768092155457, "learning_rate": 2.7897867401711886e-05, "loss": 0.0452, "step": 19980 }, { "epoch": 2.166196877710321, "grad_norm": 0.171989306807518, "learning_rate": 2.789424053387495e-05, "loss": 0.0029, "step": 19981 }, { "epoch": 2.1663052905464006, "grad_norm": 0.08094844967126846, "learning_rate": 2.789061366603801e-05, "loss": 0.0035, "step": 19982 }, { "epoch": 2.1664137033824806, "grad_norm": 1.3213757276535034, "learning_rate": 2.7886986798201075e-05, "loss": 0.0505, "step": 19983 }, { "epoch": 2.16652211621856, "grad_norm": 0.936379075050354, "learning_rate": 2.788335993036414e-05, "loss": 0.0309, "step": 19984 }, { "epoch": 2.16663052905464, "grad_norm": 0.04735448956489563, "learning_rate": 2.7879733062527203e-05, "loss": 0.0031, "step": 19985 }, { "epoch": 2.1667389418907197, "grad_norm": 0.24096165597438812, "learning_rate": 2.7876106194690264e-05, "loss": 0.0117, "step": 19986 }, { "epoch": 2.1668473547267997, "grad_norm": 1.2021863460540771, "learning_rate": 2.7872479326853328e-05, "loss": 0.0207, "step": 19987 }, { "epoch": 2.1669557675628797, "grad_norm": 0.6897015571594238, "learning_rate": 2.7868852459016392e-05, "loss": 0.0352, "step": 19988 }, { "epoch": 2.167064180398959, "grad_norm": 0.551975429058075, "learning_rate": 2.786522559117946e-05, "loss": 0.0534, "step": 19989 }, { "epoch": 2.167172593235039, "grad_norm": 0.3057928681373596, "learning_rate": 2.7861598723342524e-05, "loss": 0.0221, "step": 19990 }, { "epoch": 2.1672810060711187, "grad_norm": 0.2448939085006714, "learning_rate": 2.7857971855505588e-05, "loss": 0.0061, "step": 19991 }, { "epoch": 2.1673894189071987, "grad_norm": 0.2809481918811798, "learning_rate": 2.7854344987668652e-05, "loss": 0.0107, "step": 19992 }, { "epoch": 2.1674978317432783, "grad_norm": 0.13759903609752655, "learning_rate": 2.7850718119831716e-05, "loss": 0.0036, "step": 19993 }, { "epoch": 2.1676062445793582, "grad_norm": 0.2356802076101303, "learning_rate": 2.7847091251994777e-05, "loss": 0.0267, "step": 19994 }, { "epoch": 2.167714657415438, "grad_norm": 0.2765938937664032, "learning_rate": 2.784346438415784e-05, "loss": 0.0151, "step": 19995 }, { "epoch": 2.1678230702515178, "grad_norm": 0.24332299828529358, "learning_rate": 2.7839837516320905e-05, "loss": 0.0073, "step": 19996 }, { "epoch": 2.1679314830875978, "grad_norm": 0.30163243412971497, "learning_rate": 2.783621064848397e-05, "loss": 0.0092, "step": 19997 }, { "epoch": 2.1680398959236773, "grad_norm": 0.5125442743301392, "learning_rate": 2.7832583780647037e-05, "loss": 0.0223, "step": 19998 }, { "epoch": 2.1681483087597573, "grad_norm": 0.23117758333683014, "learning_rate": 2.78289569128101e-05, "loss": 0.0369, "step": 19999 }, { "epoch": 2.168256721595837, "grad_norm": 0.5635876655578613, "learning_rate": 2.7825330044973165e-05, "loss": 0.0322, "step": 20000 }, { "epoch": 2.168365134431917, "grad_norm": 0.07901908457279205, "learning_rate": 2.7821703177136226e-05, "loss": 0.0038, "step": 20001 }, { "epoch": 2.1684735472679963, "grad_norm": 0.21597591042518616, "learning_rate": 2.781807630929929e-05, "loss": 0.0168, "step": 20002 }, { "epoch": 2.1685819601040763, "grad_norm": 0.7668721079826355, "learning_rate": 2.7814449441462354e-05, "loss": 0.0354, "step": 20003 }, { "epoch": 2.1686903729401563, "grad_norm": 0.059829577803611755, "learning_rate": 2.7810822573625418e-05, "loss": 0.0019, "step": 20004 }, { "epoch": 2.168798785776236, "grad_norm": 0.5949057340621948, "learning_rate": 2.780719570578848e-05, "loss": 0.0182, "step": 20005 }, { "epoch": 2.168907198612316, "grad_norm": 0.036181140691041946, "learning_rate": 2.780356883795155e-05, "loss": 0.0027, "step": 20006 }, { "epoch": 2.1690156114483954, "grad_norm": 0.16528455913066864, "learning_rate": 2.7799941970114614e-05, "loss": 0.0033, "step": 20007 }, { "epoch": 2.1691240242844754, "grad_norm": 0.10364045202732086, "learning_rate": 2.7796315102277674e-05, "loss": 0.0028, "step": 20008 }, { "epoch": 2.169232437120555, "grad_norm": 0.5872246026992798, "learning_rate": 2.779268823444074e-05, "loss": 0.024, "step": 20009 }, { "epoch": 2.169340849956635, "grad_norm": 0.48217353224754333, "learning_rate": 2.7789061366603803e-05, "loss": 0.033, "step": 20010 }, { "epoch": 2.169449262792715, "grad_norm": 0.5909720063209534, "learning_rate": 2.7785434498766867e-05, "loss": 0.026, "step": 20011 }, { "epoch": 2.1695576756287944, "grad_norm": 0.10803855210542679, "learning_rate": 2.7781807630929927e-05, "loss": 0.0029, "step": 20012 }, { "epoch": 2.1696660884648744, "grad_norm": 0.24143189191818237, "learning_rate": 2.777818076309299e-05, "loss": 0.0106, "step": 20013 }, { "epoch": 2.169774501300954, "grad_norm": 0.12114968150854111, "learning_rate": 2.7774553895256056e-05, "loss": 0.006, "step": 20014 }, { "epoch": 2.169882914137034, "grad_norm": 0.08027128875255585, "learning_rate": 2.7770927027419123e-05, "loss": 0.0047, "step": 20015 }, { "epoch": 2.1699913269731135, "grad_norm": 0.09818895906209946, "learning_rate": 2.7767300159582187e-05, "loss": 0.0041, "step": 20016 }, { "epoch": 2.1700997398091935, "grad_norm": 0.03511907532811165, "learning_rate": 2.776367329174525e-05, "loss": 0.0017, "step": 20017 }, { "epoch": 2.170208152645273, "grad_norm": 0.054277453571558, "learning_rate": 2.7760046423908315e-05, "loss": 0.0033, "step": 20018 }, { "epoch": 2.170316565481353, "grad_norm": 0.7359291911125183, "learning_rate": 2.775641955607138e-05, "loss": 0.0393, "step": 20019 }, { "epoch": 2.170424978317433, "grad_norm": 0.3330562114715576, "learning_rate": 2.775279268823444e-05, "loss": 0.0098, "step": 20020 }, { "epoch": 2.1705333911535125, "grad_norm": 0.27217745780944824, "learning_rate": 2.7749165820397504e-05, "loss": 0.0088, "step": 20021 }, { "epoch": 2.1706418039895925, "grad_norm": 0.3952724039554596, "learning_rate": 2.774553895256057e-05, "loss": 0.0165, "step": 20022 }, { "epoch": 2.170750216825672, "grad_norm": 0.4959860146045685, "learning_rate": 2.7741912084723633e-05, "loss": 0.0093, "step": 20023 }, { "epoch": 2.170858629661752, "grad_norm": 0.12810489535331726, "learning_rate": 2.77382852168867e-05, "loss": 0.0072, "step": 20024 }, { "epoch": 2.1709670424978316, "grad_norm": 0.03925089165568352, "learning_rate": 2.7734658349049764e-05, "loss": 0.0011, "step": 20025 }, { "epoch": 2.1710754553339116, "grad_norm": 0.32461124658584595, "learning_rate": 2.773103148121283e-05, "loss": 0.0124, "step": 20026 }, { "epoch": 2.171183868169991, "grad_norm": 0.0023873329628258944, "learning_rate": 2.772740461337589e-05, "loss": 0.0001, "step": 20027 }, { "epoch": 2.171292281006071, "grad_norm": 0.12232998013496399, "learning_rate": 2.7723777745538953e-05, "loss": 0.0044, "step": 20028 }, { "epoch": 2.171400693842151, "grad_norm": 0.7395641207695007, "learning_rate": 2.7720150877702017e-05, "loss": 0.0456, "step": 20029 }, { "epoch": 2.1715091066782306, "grad_norm": 0.18097831308841705, "learning_rate": 2.771652400986508e-05, "loss": 0.0036, "step": 20030 }, { "epoch": 2.1716175195143106, "grad_norm": 0.05053621530532837, "learning_rate": 2.7712897142028142e-05, "loss": 0.0025, "step": 20031 }, { "epoch": 2.17172593235039, "grad_norm": 0.1475697159767151, "learning_rate": 2.7709270274191213e-05, "loss": 0.0115, "step": 20032 }, { "epoch": 2.17183434518647, "grad_norm": 0.6486405730247498, "learning_rate": 2.7705643406354277e-05, "loss": 0.029, "step": 20033 }, { "epoch": 2.17194275802255, "grad_norm": 0.7821561694145203, "learning_rate": 2.7702016538517338e-05, "loss": 0.0505, "step": 20034 }, { "epoch": 2.1720511708586296, "grad_norm": 0.04546774551272392, "learning_rate": 2.7698389670680402e-05, "loss": 0.0012, "step": 20035 }, { "epoch": 2.1721595836947096, "grad_norm": 0.06735944002866745, "learning_rate": 2.7694762802843466e-05, "loss": 0.0009, "step": 20036 }, { "epoch": 2.172267996530789, "grad_norm": 0.08629970252513885, "learning_rate": 2.769113593500653e-05, "loss": 0.0042, "step": 20037 }, { "epoch": 2.172376409366869, "grad_norm": 0.3467884957790375, "learning_rate": 2.768750906716959e-05, "loss": 0.0163, "step": 20038 }, { "epoch": 2.1724848222029487, "grad_norm": 0.5547547936439514, "learning_rate": 2.7683882199332655e-05, "loss": 0.0278, "step": 20039 }, { "epoch": 2.1725932350390287, "grad_norm": 0.8265954256057739, "learning_rate": 2.768025533149572e-05, "loss": 0.0249, "step": 20040 }, { "epoch": 2.172701647875108, "grad_norm": 0.5005296468734741, "learning_rate": 2.7676628463658787e-05, "loss": 0.0167, "step": 20041 }, { "epoch": 2.172810060711188, "grad_norm": 0.1323697417974472, "learning_rate": 2.767300159582185e-05, "loss": 0.003, "step": 20042 }, { "epoch": 2.172918473547268, "grad_norm": 0.0875769779086113, "learning_rate": 2.7669374727984915e-05, "loss": 0.0016, "step": 20043 }, { "epoch": 2.1730268863833477, "grad_norm": 0.19762420654296875, "learning_rate": 2.766574786014798e-05, "loss": 0.0061, "step": 20044 }, { "epoch": 2.1731352992194277, "grad_norm": 0.48088201880455017, "learning_rate": 2.7662120992311043e-05, "loss": 0.0095, "step": 20045 }, { "epoch": 2.1732437120555073, "grad_norm": 0.38323381543159485, "learning_rate": 2.7658494124474104e-05, "loss": 0.0118, "step": 20046 }, { "epoch": 2.1733521248915872, "grad_norm": 0.08084606379270554, "learning_rate": 2.7654867256637168e-05, "loss": 0.002, "step": 20047 }, { "epoch": 2.173460537727667, "grad_norm": 0.05971945822238922, "learning_rate": 2.7651240388800232e-05, "loss": 0.0032, "step": 20048 }, { "epoch": 2.1735689505637468, "grad_norm": 0.17450155317783356, "learning_rate": 2.7647613520963296e-05, "loss": 0.0077, "step": 20049 }, { "epoch": 2.1736773633998263, "grad_norm": 0.3474022150039673, "learning_rate": 2.7643986653126363e-05, "loss": 0.0106, "step": 20050 }, { "epoch": 2.1737857762359063, "grad_norm": 0.09813440591096878, "learning_rate": 2.7640359785289428e-05, "loss": 0.004, "step": 20051 }, { "epoch": 2.1738941890719863, "grad_norm": 0.1998991221189499, "learning_rate": 2.763673291745249e-05, "loss": 0.0088, "step": 20052 }, { "epoch": 2.174002601908066, "grad_norm": 0.09865362197160721, "learning_rate": 2.7633106049615552e-05, "loss": 0.0033, "step": 20053 }, { "epoch": 2.174111014744146, "grad_norm": 0.02256091684103012, "learning_rate": 2.7629479181778616e-05, "loss": 0.0012, "step": 20054 }, { "epoch": 2.1742194275802254, "grad_norm": 0.295035183429718, "learning_rate": 2.762585231394168e-05, "loss": 0.0068, "step": 20055 }, { "epoch": 2.1743278404163053, "grad_norm": 0.03071301057934761, "learning_rate": 2.7622225446104745e-05, "loss": 0.0015, "step": 20056 }, { "epoch": 2.174436253252385, "grad_norm": 0.14934946596622467, "learning_rate": 2.7618598578267805e-05, "loss": 0.0053, "step": 20057 }, { "epoch": 2.174544666088465, "grad_norm": 0.3507674038410187, "learning_rate": 2.7614971710430876e-05, "loss": 0.0304, "step": 20058 }, { "epoch": 2.174653078924545, "grad_norm": 0.15191751718521118, "learning_rate": 2.761134484259394e-05, "loss": 0.0069, "step": 20059 }, { "epoch": 2.1747614917606244, "grad_norm": 0.4014205038547516, "learning_rate": 2.7607717974757e-05, "loss": 0.0119, "step": 20060 }, { "epoch": 2.1748699045967044, "grad_norm": 0.08743064850568771, "learning_rate": 2.7604091106920065e-05, "loss": 0.0021, "step": 20061 }, { "epoch": 2.174978317432784, "grad_norm": 0.22780101001262665, "learning_rate": 2.760046423908313e-05, "loss": 0.0123, "step": 20062 }, { "epoch": 2.175086730268864, "grad_norm": 0.1915084719657898, "learning_rate": 2.7596837371246193e-05, "loss": 0.0058, "step": 20063 }, { "epoch": 2.1751951431049434, "grad_norm": 0.5768685340881348, "learning_rate": 2.7593210503409254e-05, "loss": 0.0533, "step": 20064 }, { "epoch": 2.1753035559410234, "grad_norm": 0.2523672580718994, "learning_rate": 2.7589583635572318e-05, "loss": 0.0234, "step": 20065 }, { "epoch": 2.1754119687771034, "grad_norm": 0.09142595529556274, "learning_rate": 2.7585956767735382e-05, "loss": 0.0059, "step": 20066 }, { "epoch": 2.175520381613183, "grad_norm": 0.5278452634811401, "learning_rate": 2.758232989989845e-05, "loss": 0.0167, "step": 20067 }, { "epoch": 2.175628794449263, "grad_norm": 0.23851294815540314, "learning_rate": 2.7578703032061514e-05, "loss": 0.0045, "step": 20068 }, { "epoch": 2.1757372072853425, "grad_norm": 0.23243945837020874, "learning_rate": 2.7575076164224578e-05, "loss": 0.0193, "step": 20069 }, { "epoch": 2.1758456201214225, "grad_norm": 0.8773288130760193, "learning_rate": 2.7571449296387642e-05, "loss": 0.051, "step": 20070 }, { "epoch": 2.175954032957502, "grad_norm": 0.03563256934285164, "learning_rate": 2.7567822428550706e-05, "loss": 0.0011, "step": 20071 }, { "epoch": 2.176062445793582, "grad_norm": 0.04694271460175514, "learning_rate": 2.7564195560713767e-05, "loss": 0.0018, "step": 20072 }, { "epoch": 2.1761708586296615, "grad_norm": 0.4483264088630676, "learning_rate": 2.756056869287683e-05, "loss": 0.0073, "step": 20073 }, { "epoch": 2.1762792714657415, "grad_norm": 0.09371057897806168, "learning_rate": 2.7556941825039895e-05, "loss": 0.003, "step": 20074 }, { "epoch": 2.1763876843018215, "grad_norm": 0.6942010521888733, "learning_rate": 2.755331495720296e-05, "loss": 0.0462, "step": 20075 }, { "epoch": 2.176496097137901, "grad_norm": 0.6972586512565613, "learning_rate": 2.7549688089366027e-05, "loss": 0.0287, "step": 20076 }, { "epoch": 2.176604509973981, "grad_norm": 0.08754787594079971, "learning_rate": 2.754606122152909e-05, "loss": 0.0017, "step": 20077 }, { "epoch": 2.1767129228100606, "grad_norm": 0.24815234541893005, "learning_rate": 2.7542434353692155e-05, "loss": 0.0071, "step": 20078 }, { "epoch": 2.1768213356461406, "grad_norm": 0.23193208873271942, "learning_rate": 2.7538807485855216e-05, "loss": 0.0122, "step": 20079 }, { "epoch": 2.17692974848222, "grad_norm": 0.10988558083772659, "learning_rate": 2.753518061801828e-05, "loss": 0.0109, "step": 20080 }, { "epoch": 2.1770381613183, "grad_norm": 0.09666268527507782, "learning_rate": 2.7531553750181344e-05, "loss": 0.0024, "step": 20081 }, { "epoch": 2.17714657415438, "grad_norm": 0.16468854248523712, "learning_rate": 2.7527926882344408e-05, "loss": 0.011, "step": 20082 }, { "epoch": 2.1772549869904596, "grad_norm": 0.027437930926680565, "learning_rate": 2.752430001450747e-05, "loss": 0.0008, "step": 20083 }, { "epoch": 2.1773633998265396, "grad_norm": 0.4589044153690338, "learning_rate": 2.752067314667054e-05, "loss": 0.0315, "step": 20084 }, { "epoch": 2.177471812662619, "grad_norm": 0.06993892043828964, "learning_rate": 2.7517046278833604e-05, "loss": 0.0031, "step": 20085 }, { "epoch": 2.177580225498699, "grad_norm": 0.947026252746582, "learning_rate": 2.7513419410996664e-05, "loss": 0.0098, "step": 20086 }, { "epoch": 2.1776886383347787, "grad_norm": 0.12681147456169128, "learning_rate": 2.750979254315973e-05, "loss": 0.0069, "step": 20087 }, { "epoch": 2.1777970511708586, "grad_norm": 0.10593177378177643, "learning_rate": 2.7506165675322793e-05, "loss": 0.0053, "step": 20088 }, { "epoch": 2.1779054640069386, "grad_norm": 0.0697854682803154, "learning_rate": 2.7502538807485857e-05, "loss": 0.0037, "step": 20089 }, { "epoch": 2.178013876843018, "grad_norm": 0.035650867968797684, "learning_rate": 2.749891193964892e-05, "loss": 0.0029, "step": 20090 }, { "epoch": 2.178122289679098, "grad_norm": 0.391721248626709, "learning_rate": 2.749528507181198e-05, "loss": 0.0087, "step": 20091 }, { "epoch": 2.1782307025151777, "grad_norm": 0.1281684935092926, "learning_rate": 2.7491658203975046e-05, "loss": 0.0034, "step": 20092 }, { "epoch": 2.1783391153512577, "grad_norm": 0.07409696280956268, "learning_rate": 2.7488031336138113e-05, "loss": 0.0024, "step": 20093 }, { "epoch": 2.1784475281873372, "grad_norm": 0.05856196954846382, "learning_rate": 2.7484404468301177e-05, "loss": 0.0018, "step": 20094 }, { "epoch": 2.178555941023417, "grad_norm": 0.07827600836753845, "learning_rate": 2.748077760046424e-05, "loss": 0.0028, "step": 20095 }, { "epoch": 2.1786643538594967, "grad_norm": 0.04295435920357704, "learning_rate": 2.7477150732627306e-05, "loss": 0.0018, "step": 20096 }, { "epoch": 2.1787727666955767, "grad_norm": 0.12336158752441406, "learning_rate": 2.747352386479037e-05, "loss": 0.0029, "step": 20097 }, { "epoch": 2.1788811795316567, "grad_norm": 0.21384954452514648, "learning_rate": 2.746989699695343e-05, "loss": 0.0111, "step": 20098 }, { "epoch": 2.1789895923677363, "grad_norm": 0.13974089920520782, "learning_rate": 2.7466270129116494e-05, "loss": 0.0092, "step": 20099 }, { "epoch": 2.1790980052038162, "grad_norm": 0.15753895044326782, "learning_rate": 2.746264326127956e-05, "loss": 0.004, "step": 20100 }, { "epoch": 2.179206418039896, "grad_norm": 0.517997145652771, "learning_rate": 2.7459016393442626e-05, "loss": 0.0284, "step": 20101 }, { "epoch": 2.1793148308759758, "grad_norm": 0.26089227199554443, "learning_rate": 2.745538952560569e-05, "loss": 0.0155, "step": 20102 }, { "epoch": 2.1794232437120553, "grad_norm": 0.3822813332080841, "learning_rate": 2.7451762657768754e-05, "loss": 0.0078, "step": 20103 }, { "epoch": 2.1795316565481353, "grad_norm": 0.2854234278202057, "learning_rate": 2.744813578993182e-05, "loss": 0.0114, "step": 20104 }, { "epoch": 2.1796400693842153, "grad_norm": 0.26055189967155457, "learning_rate": 2.744450892209488e-05, "loss": 0.0071, "step": 20105 }, { "epoch": 2.179748482220295, "grad_norm": 0.2515030801296234, "learning_rate": 2.7440882054257943e-05, "loss": 0.0057, "step": 20106 }, { "epoch": 2.179856895056375, "grad_norm": 0.482950359582901, "learning_rate": 2.7437255186421007e-05, "loss": 0.0311, "step": 20107 }, { "epoch": 2.1799653078924544, "grad_norm": 0.1588926464319229, "learning_rate": 2.743362831858407e-05, "loss": 0.0137, "step": 20108 }, { "epoch": 2.1800737207285343, "grad_norm": 0.19222034513950348, "learning_rate": 2.7430001450747132e-05, "loss": 0.0035, "step": 20109 }, { "epoch": 2.180182133564614, "grad_norm": 1.0744966268539429, "learning_rate": 2.7426374582910203e-05, "loss": 0.0379, "step": 20110 }, { "epoch": 2.180290546400694, "grad_norm": 0.0556117445230484, "learning_rate": 2.7422747715073267e-05, "loss": 0.0013, "step": 20111 }, { "epoch": 2.180398959236774, "grad_norm": 0.21827346086502075, "learning_rate": 2.7419120847236328e-05, "loss": 0.0073, "step": 20112 }, { "epoch": 2.1805073720728534, "grad_norm": 0.541539192199707, "learning_rate": 2.7415493979399392e-05, "loss": 0.0824, "step": 20113 }, { "epoch": 2.1806157849089334, "grad_norm": 0.5176228284835815, "learning_rate": 2.7411867111562456e-05, "loss": 0.0339, "step": 20114 }, { "epoch": 2.180724197745013, "grad_norm": 0.03184801712632179, "learning_rate": 2.740824024372552e-05, "loss": 0.0016, "step": 20115 }, { "epoch": 2.180832610581093, "grad_norm": 0.038876134902238846, "learning_rate": 2.7404613375888584e-05, "loss": 0.0013, "step": 20116 }, { "epoch": 2.1809410234171724, "grad_norm": 0.015221749432384968, "learning_rate": 2.7400986508051645e-05, "loss": 0.0005, "step": 20117 }, { "epoch": 2.1810494362532524, "grad_norm": 0.6739238500595093, "learning_rate": 2.739735964021471e-05, "loss": 0.0303, "step": 20118 }, { "epoch": 2.181157849089332, "grad_norm": 0.10142781585454941, "learning_rate": 2.7393732772377777e-05, "loss": 0.0025, "step": 20119 }, { "epoch": 2.181266261925412, "grad_norm": 0.019116804003715515, "learning_rate": 2.739010590454084e-05, "loss": 0.0005, "step": 20120 }, { "epoch": 2.181374674761492, "grad_norm": 0.35354679822921753, "learning_rate": 2.7386479036703905e-05, "loss": 0.0094, "step": 20121 }, { "epoch": 2.1814830875975715, "grad_norm": 0.20755435526371002, "learning_rate": 2.738285216886697e-05, "loss": 0.0101, "step": 20122 }, { "epoch": 2.1815915004336515, "grad_norm": 0.3637346029281616, "learning_rate": 2.7379225301030033e-05, "loss": 0.0258, "step": 20123 }, { "epoch": 2.181699913269731, "grad_norm": 0.45970067381858826, "learning_rate": 2.7375598433193094e-05, "loss": 0.0253, "step": 20124 }, { "epoch": 2.181808326105811, "grad_norm": 0.27484628558158875, "learning_rate": 2.7371971565356158e-05, "loss": 0.009, "step": 20125 }, { "epoch": 2.1819167389418905, "grad_norm": 0.3076968789100647, "learning_rate": 2.7368344697519222e-05, "loss": 0.0249, "step": 20126 }, { "epoch": 2.1820251517779705, "grad_norm": 0.21315206587314606, "learning_rate": 2.736471782968229e-05, "loss": 0.0118, "step": 20127 }, { "epoch": 2.1821335646140505, "grad_norm": 0.3946366608142853, "learning_rate": 2.7361090961845353e-05, "loss": 0.0257, "step": 20128 }, { "epoch": 2.18224197745013, "grad_norm": 0.6035423278808594, "learning_rate": 2.7357464094008418e-05, "loss": 0.0317, "step": 20129 }, { "epoch": 2.18235039028621, "grad_norm": 0.20205752551555634, "learning_rate": 2.7353837226171482e-05, "loss": 0.0131, "step": 20130 }, { "epoch": 2.1824588031222896, "grad_norm": 0.28476086258888245, "learning_rate": 2.7350210358334542e-05, "loss": 0.0182, "step": 20131 }, { "epoch": 2.1825672159583696, "grad_norm": 0.03956092149019241, "learning_rate": 2.7346583490497607e-05, "loss": 0.0031, "step": 20132 }, { "epoch": 2.182675628794449, "grad_norm": 0.0142410509288311, "learning_rate": 2.734295662266067e-05, "loss": 0.0005, "step": 20133 }, { "epoch": 2.182784041630529, "grad_norm": 1.244152307510376, "learning_rate": 2.7339329754823735e-05, "loss": 0.0142, "step": 20134 }, { "epoch": 2.182892454466609, "grad_norm": 0.0294363871216774, "learning_rate": 2.7335702886986795e-05, "loss": 0.0014, "step": 20135 }, { "epoch": 2.1830008673026886, "grad_norm": 0.022176289930939674, "learning_rate": 2.7332076019149866e-05, "loss": 0.0007, "step": 20136 }, { "epoch": 2.1831092801387686, "grad_norm": 0.29799360036849976, "learning_rate": 2.732844915131293e-05, "loss": 0.0268, "step": 20137 }, { "epoch": 2.183217692974848, "grad_norm": 0.1814149022102356, "learning_rate": 2.732482228347599e-05, "loss": 0.0069, "step": 20138 }, { "epoch": 2.183326105810928, "grad_norm": 0.07830435037612915, "learning_rate": 2.7321195415639055e-05, "loss": 0.0013, "step": 20139 }, { "epoch": 2.1834345186470077, "grad_norm": 0.030570028349757195, "learning_rate": 2.731756854780212e-05, "loss": 0.0014, "step": 20140 }, { "epoch": 2.1835429314830876, "grad_norm": 0.23155073821544647, "learning_rate": 2.7313941679965183e-05, "loss": 0.0059, "step": 20141 }, { "epoch": 2.183651344319167, "grad_norm": 0.1454671174287796, "learning_rate": 2.7310314812128248e-05, "loss": 0.004, "step": 20142 }, { "epoch": 2.183759757155247, "grad_norm": 0.17966851592063904, "learning_rate": 2.7306687944291308e-05, "loss": 0.0068, "step": 20143 }, { "epoch": 2.183868169991327, "grad_norm": 0.12655818462371826, "learning_rate": 2.7303061076454372e-05, "loss": 0.0041, "step": 20144 }, { "epoch": 2.1839765828274067, "grad_norm": 0.3528444766998291, "learning_rate": 2.729943420861744e-05, "loss": 0.0296, "step": 20145 }, { "epoch": 2.1840849956634867, "grad_norm": 0.5408018827438354, "learning_rate": 2.7295807340780504e-05, "loss": 0.029, "step": 20146 }, { "epoch": 2.1841934084995662, "grad_norm": 0.10304787755012512, "learning_rate": 2.7292180472943568e-05, "loss": 0.0054, "step": 20147 }, { "epoch": 2.184301821335646, "grad_norm": 0.4601851999759674, "learning_rate": 2.7288553605106632e-05, "loss": 0.0359, "step": 20148 }, { "epoch": 2.1844102341717258, "grad_norm": 0.13482597470283508, "learning_rate": 2.7284926737269696e-05, "loss": 0.0044, "step": 20149 }, { "epoch": 2.1845186470078057, "grad_norm": 0.06435734033584595, "learning_rate": 2.7281299869432757e-05, "loss": 0.002, "step": 20150 }, { "epoch": 2.1846270598438857, "grad_norm": 0.30611109733581543, "learning_rate": 2.727767300159582e-05, "loss": 0.0245, "step": 20151 }, { "epoch": 2.1847354726799653, "grad_norm": 0.5315962433815002, "learning_rate": 2.7274046133758885e-05, "loss": 0.0453, "step": 20152 }, { "epoch": 2.1848438855160452, "grad_norm": 0.0910719484090805, "learning_rate": 2.7270419265921953e-05, "loss": 0.004, "step": 20153 }, { "epoch": 2.184952298352125, "grad_norm": 0.185548797249794, "learning_rate": 2.7266792398085017e-05, "loss": 0.0043, "step": 20154 }, { "epoch": 2.1850607111882048, "grad_norm": 0.7082991003990173, "learning_rate": 2.726316553024808e-05, "loss": 0.0137, "step": 20155 }, { "epoch": 2.1851691240242843, "grad_norm": 0.007663064636290073, "learning_rate": 2.7259538662411145e-05, "loss": 0.0004, "step": 20156 }, { "epoch": 2.1852775368603643, "grad_norm": 0.27735745906829834, "learning_rate": 2.7255911794574206e-05, "loss": 0.0125, "step": 20157 }, { "epoch": 2.1853859496964443, "grad_norm": 0.31621053814888, "learning_rate": 2.725228492673727e-05, "loss": 0.0155, "step": 20158 }, { "epoch": 2.185494362532524, "grad_norm": 0.19957801699638367, "learning_rate": 2.7248658058900334e-05, "loss": 0.0067, "step": 20159 }, { "epoch": 2.185602775368604, "grad_norm": 0.10690611600875854, "learning_rate": 2.7245031191063398e-05, "loss": 0.0018, "step": 20160 }, { "epoch": 2.1857111882046834, "grad_norm": 0.167607381939888, "learning_rate": 2.724140432322646e-05, "loss": 0.0071, "step": 20161 }, { "epoch": 2.1858196010407633, "grad_norm": 0.18870426714420319, "learning_rate": 2.723777745538953e-05, "loss": 0.0121, "step": 20162 }, { "epoch": 2.185928013876843, "grad_norm": 0.29292789101600647, "learning_rate": 2.7234150587552594e-05, "loss": 0.0167, "step": 20163 }, { "epoch": 2.186036426712923, "grad_norm": 0.4790742099285126, "learning_rate": 2.7230523719715654e-05, "loss": 0.0094, "step": 20164 }, { "epoch": 2.1861448395490024, "grad_norm": 1.3975169658660889, "learning_rate": 2.722689685187872e-05, "loss": 0.0199, "step": 20165 }, { "epoch": 2.1862532523850824, "grad_norm": 0.5935145020484924, "learning_rate": 2.7223269984041783e-05, "loss": 0.0364, "step": 20166 }, { "epoch": 2.1863616652211624, "grad_norm": 0.23082703351974487, "learning_rate": 2.7219643116204847e-05, "loss": 0.0063, "step": 20167 }, { "epoch": 2.186470078057242, "grad_norm": 0.20341598987579346, "learning_rate": 2.721601624836791e-05, "loss": 0.0178, "step": 20168 }, { "epoch": 2.186578490893322, "grad_norm": 0.017256567254662514, "learning_rate": 2.721238938053097e-05, "loss": 0.0009, "step": 20169 }, { "epoch": 2.1866869037294014, "grad_norm": 0.5023469924926758, "learning_rate": 2.7208762512694036e-05, "loss": 0.0149, "step": 20170 }, { "epoch": 2.1867953165654814, "grad_norm": 0.2624396085739136, "learning_rate": 2.7205135644857103e-05, "loss": 0.0329, "step": 20171 }, { "epoch": 2.186903729401561, "grad_norm": 0.15497319400310516, "learning_rate": 2.7201508777020167e-05, "loss": 0.0096, "step": 20172 }, { "epoch": 2.187012142237641, "grad_norm": 0.10147274285554886, "learning_rate": 2.719788190918323e-05, "loss": 0.0034, "step": 20173 }, { "epoch": 2.1871205550737205, "grad_norm": 0.3014751374721527, "learning_rate": 2.7194255041346296e-05, "loss": 0.0064, "step": 20174 }, { "epoch": 2.1872289679098005, "grad_norm": 0.0766424834728241, "learning_rate": 2.719062817350936e-05, "loss": 0.0032, "step": 20175 }, { "epoch": 2.1873373807458805, "grad_norm": 0.0412665419280529, "learning_rate": 2.718700130567242e-05, "loss": 0.0018, "step": 20176 }, { "epoch": 2.18744579358196, "grad_norm": 0.13669750094413757, "learning_rate": 2.7183374437835484e-05, "loss": 0.0072, "step": 20177 }, { "epoch": 2.18755420641804, "grad_norm": 0.15954329073429108, "learning_rate": 2.717974756999855e-05, "loss": 0.0037, "step": 20178 }, { "epoch": 2.1876626192541195, "grad_norm": 0.06947756558656693, "learning_rate": 2.7176120702161616e-05, "loss": 0.0016, "step": 20179 }, { "epoch": 2.1877710320901995, "grad_norm": 0.09311943501234055, "learning_rate": 2.717249383432468e-05, "loss": 0.005, "step": 20180 }, { "epoch": 2.1878794449262795, "grad_norm": 0.024295229464769363, "learning_rate": 2.7168866966487744e-05, "loss": 0.001, "step": 20181 }, { "epoch": 2.187987857762359, "grad_norm": 0.3867889940738678, "learning_rate": 2.716524009865081e-05, "loss": 0.02, "step": 20182 }, { "epoch": 2.188096270598439, "grad_norm": 0.30291932821273804, "learning_rate": 2.716161323081387e-05, "loss": 0.0029, "step": 20183 }, { "epoch": 2.1882046834345186, "grad_norm": 0.20316816866397858, "learning_rate": 2.7157986362976933e-05, "loss": 0.011, "step": 20184 }, { "epoch": 2.1883130962705986, "grad_norm": 0.02483072504401207, "learning_rate": 2.7154359495139997e-05, "loss": 0.0005, "step": 20185 }, { "epoch": 2.188421509106678, "grad_norm": 0.009149466641247272, "learning_rate": 2.715073262730306e-05, "loss": 0.0006, "step": 20186 }, { "epoch": 2.188529921942758, "grad_norm": 0.08598104119300842, "learning_rate": 2.7147105759466122e-05, "loss": 0.0042, "step": 20187 }, { "epoch": 2.1886383347788376, "grad_norm": 0.6479744911193848, "learning_rate": 2.7143478891629193e-05, "loss": 0.0142, "step": 20188 }, { "epoch": 2.1887467476149176, "grad_norm": 0.1752554029226303, "learning_rate": 2.7139852023792257e-05, "loss": 0.0072, "step": 20189 }, { "epoch": 2.1888551604509976, "grad_norm": 0.28184717893600464, "learning_rate": 2.7136225155955318e-05, "loss": 0.0186, "step": 20190 }, { "epoch": 2.188963573287077, "grad_norm": 0.15604740381240845, "learning_rate": 2.7132598288118382e-05, "loss": 0.0099, "step": 20191 }, { "epoch": 2.189071986123157, "grad_norm": 0.143233984708786, "learning_rate": 2.7128971420281446e-05, "loss": 0.006, "step": 20192 }, { "epoch": 2.1891803989592367, "grad_norm": 0.17090195417404175, "learning_rate": 2.712534455244451e-05, "loss": 0.0027, "step": 20193 }, { "epoch": 2.1892888117953166, "grad_norm": 0.1630031317472458, "learning_rate": 2.7121717684607574e-05, "loss": 0.0036, "step": 20194 }, { "epoch": 2.189397224631396, "grad_norm": 0.04637341573834419, "learning_rate": 2.7118090816770635e-05, "loss": 0.0023, "step": 20195 }, { "epoch": 2.189505637467476, "grad_norm": 0.05049823597073555, "learning_rate": 2.7114463948933706e-05, "loss": 0.0028, "step": 20196 }, { "epoch": 2.1896140503035557, "grad_norm": 0.10517243295907974, "learning_rate": 2.7110837081096767e-05, "loss": 0.0032, "step": 20197 }, { "epoch": 2.1897224631396357, "grad_norm": 0.23247306048870087, "learning_rate": 2.710721021325983e-05, "loss": 0.0054, "step": 20198 }, { "epoch": 2.1898308759757157, "grad_norm": 0.2907208204269409, "learning_rate": 2.7103583345422895e-05, "loss": 0.0123, "step": 20199 }, { "epoch": 2.1899392888117952, "grad_norm": 0.24673502147197723, "learning_rate": 2.709995647758596e-05, "loss": 0.0433, "step": 20200 }, { "epoch": 2.190047701647875, "grad_norm": 0.052613042294979095, "learning_rate": 2.7096329609749023e-05, "loss": 0.0036, "step": 20201 }, { "epoch": 2.1901561144839548, "grad_norm": 0.3927035629749298, "learning_rate": 2.7092702741912084e-05, "loss": 0.0105, "step": 20202 }, { "epoch": 2.1902645273200347, "grad_norm": 0.03451991453766823, "learning_rate": 2.7089075874075148e-05, "loss": 0.001, "step": 20203 }, { "epoch": 2.1903729401561143, "grad_norm": 0.14878804981708527, "learning_rate": 2.7085449006238212e-05, "loss": 0.0069, "step": 20204 }, { "epoch": 2.1904813529921943, "grad_norm": 0.28686094284057617, "learning_rate": 2.708182213840128e-05, "loss": 0.0146, "step": 20205 }, { "epoch": 2.1905897658282742, "grad_norm": 0.5141535997390747, "learning_rate": 2.7078195270564344e-05, "loss": 0.0194, "step": 20206 }, { "epoch": 2.190698178664354, "grad_norm": 0.2472430318593979, "learning_rate": 2.7074568402727408e-05, "loss": 0.0174, "step": 20207 }, { "epoch": 2.1908065915004338, "grad_norm": 0.1449122577905655, "learning_rate": 2.7070941534890472e-05, "loss": 0.0091, "step": 20208 }, { "epoch": 2.1909150043365133, "grad_norm": 0.1455947905778885, "learning_rate": 2.7067314667053532e-05, "loss": 0.0029, "step": 20209 }, { "epoch": 2.1910234171725933, "grad_norm": 0.1071050763130188, "learning_rate": 2.7063687799216597e-05, "loss": 0.0055, "step": 20210 }, { "epoch": 2.191131830008673, "grad_norm": 0.041248589754104614, "learning_rate": 2.706006093137966e-05, "loss": 0.0014, "step": 20211 }, { "epoch": 2.191240242844753, "grad_norm": 0.46479764580726624, "learning_rate": 2.7056434063542725e-05, "loss": 0.0194, "step": 20212 }, { "epoch": 2.191348655680833, "grad_norm": 0.3366667330265045, "learning_rate": 2.7052807195705785e-05, "loss": 0.0229, "step": 20213 }, { "epoch": 2.1914570685169124, "grad_norm": 0.42718416452407837, "learning_rate": 2.7049180327868856e-05, "loss": 0.0224, "step": 20214 }, { "epoch": 2.1915654813529923, "grad_norm": 0.4394599199295044, "learning_rate": 2.704555346003192e-05, "loss": 0.0086, "step": 20215 }, { "epoch": 2.191673894189072, "grad_norm": 0.3153631091117859, "learning_rate": 2.704192659219498e-05, "loss": 0.0124, "step": 20216 }, { "epoch": 2.191782307025152, "grad_norm": 0.46231019496917725, "learning_rate": 2.7038299724358045e-05, "loss": 0.0289, "step": 20217 }, { "epoch": 2.1918907198612314, "grad_norm": 0.2673262357711792, "learning_rate": 2.703467285652111e-05, "loss": 0.0071, "step": 20218 }, { "epoch": 2.1919991326973114, "grad_norm": 0.17999401688575745, "learning_rate": 2.7031045988684174e-05, "loss": 0.0296, "step": 20219 }, { "epoch": 2.192107545533391, "grad_norm": 0.4827684164047241, "learning_rate": 2.7027419120847238e-05, "loss": 0.0216, "step": 20220 }, { "epoch": 2.192215958369471, "grad_norm": 0.21426834166049957, "learning_rate": 2.70237922530103e-05, "loss": 0.0023, "step": 20221 }, { "epoch": 2.192324371205551, "grad_norm": 0.2292153537273407, "learning_rate": 2.702016538517337e-05, "loss": 0.0063, "step": 20222 }, { "epoch": 2.1924327840416304, "grad_norm": 0.1399233341217041, "learning_rate": 2.701653851733643e-05, "loss": 0.0049, "step": 20223 }, { "epoch": 2.1925411968777104, "grad_norm": 0.2470935583114624, "learning_rate": 2.7012911649499494e-05, "loss": 0.0117, "step": 20224 }, { "epoch": 2.19264960971379, "grad_norm": 0.12258882075548172, "learning_rate": 2.7009284781662558e-05, "loss": 0.0038, "step": 20225 }, { "epoch": 2.19275802254987, "grad_norm": 0.4911910891532898, "learning_rate": 2.7005657913825622e-05, "loss": 0.0341, "step": 20226 }, { "epoch": 2.1928664353859495, "grad_norm": 0.4260755479335785, "learning_rate": 2.7002031045988686e-05, "loss": 0.0471, "step": 20227 }, { "epoch": 2.1929748482220295, "grad_norm": 0.5893017053604126, "learning_rate": 2.6998404178151747e-05, "loss": 0.0094, "step": 20228 }, { "epoch": 2.1930832610581095, "grad_norm": 0.3568941056728363, "learning_rate": 2.699477731031481e-05, "loss": 0.0097, "step": 20229 }, { "epoch": 2.193191673894189, "grad_norm": 0.14509692788124084, "learning_rate": 2.6991150442477875e-05, "loss": 0.0038, "step": 20230 }, { "epoch": 2.193300086730269, "grad_norm": 0.07391110807657242, "learning_rate": 2.6987523574640943e-05, "loss": 0.0012, "step": 20231 }, { "epoch": 2.1934084995663485, "grad_norm": 0.022591667249798775, "learning_rate": 2.6983896706804007e-05, "loss": 0.001, "step": 20232 }, { "epoch": 2.1935169124024285, "grad_norm": 0.10084041208028793, "learning_rate": 2.698026983896707e-05, "loss": 0.0038, "step": 20233 }, { "epoch": 2.193625325238508, "grad_norm": 0.12829215824604034, "learning_rate": 2.6976642971130135e-05, "loss": 0.0044, "step": 20234 }, { "epoch": 2.193733738074588, "grad_norm": 0.2190525084733963, "learning_rate": 2.6973016103293196e-05, "loss": 0.0186, "step": 20235 }, { "epoch": 2.193842150910668, "grad_norm": 0.09237239509820938, "learning_rate": 2.696938923545626e-05, "loss": 0.0032, "step": 20236 }, { "epoch": 2.1939505637467476, "grad_norm": 0.2080995738506317, "learning_rate": 2.6965762367619324e-05, "loss": 0.0054, "step": 20237 }, { "epoch": 2.1940589765828276, "grad_norm": 0.09026369452476501, "learning_rate": 2.6962135499782388e-05, "loss": 0.0106, "step": 20238 }, { "epoch": 2.194167389418907, "grad_norm": 0.4674038290977478, "learning_rate": 2.695850863194545e-05, "loss": 0.042, "step": 20239 }, { "epoch": 2.194275802254987, "grad_norm": 0.47870707511901855, "learning_rate": 2.695488176410852e-05, "loss": 0.0323, "step": 20240 }, { "epoch": 2.1943842150910666, "grad_norm": 0.1148705929517746, "learning_rate": 2.6951254896271584e-05, "loss": 0.0055, "step": 20241 }, { "epoch": 2.1944926279271466, "grad_norm": 0.7875736951828003, "learning_rate": 2.6947628028434645e-05, "loss": 0.0246, "step": 20242 }, { "epoch": 2.194601040763226, "grad_norm": 0.5109798312187195, "learning_rate": 2.694400116059771e-05, "loss": 0.0332, "step": 20243 }, { "epoch": 2.194709453599306, "grad_norm": 0.15286745131015778, "learning_rate": 2.6940374292760773e-05, "loss": 0.0063, "step": 20244 }, { "epoch": 2.194817866435386, "grad_norm": 0.3486166298389435, "learning_rate": 2.6936747424923837e-05, "loss": 0.0154, "step": 20245 }, { "epoch": 2.1949262792714657, "grad_norm": 0.3338913321495056, "learning_rate": 2.69331205570869e-05, "loss": 0.0107, "step": 20246 }, { "epoch": 2.1950346921075456, "grad_norm": 0.18675552308559418, "learning_rate": 2.692949368924996e-05, "loss": 0.0058, "step": 20247 }, { "epoch": 2.195143104943625, "grad_norm": 0.10826613008975983, "learning_rate": 2.6925866821413033e-05, "loss": 0.0075, "step": 20248 }, { "epoch": 2.195251517779705, "grad_norm": 0.07707344740629196, "learning_rate": 2.6922239953576097e-05, "loss": 0.001, "step": 20249 }, { "epoch": 2.1953599306157847, "grad_norm": 1.008783221244812, "learning_rate": 2.6918613085739157e-05, "loss": 0.0365, "step": 20250 }, { "epoch": 2.1954683434518647, "grad_norm": 0.7519934177398682, "learning_rate": 2.691498621790222e-05, "loss": 0.0187, "step": 20251 }, { "epoch": 2.1955767562879447, "grad_norm": 0.05040065944194794, "learning_rate": 2.6911359350065286e-05, "loss": 0.0014, "step": 20252 }, { "epoch": 2.1956851691240242, "grad_norm": 0.15024790167808533, "learning_rate": 2.690773248222835e-05, "loss": 0.0202, "step": 20253 }, { "epoch": 2.195793581960104, "grad_norm": 0.3397265374660492, "learning_rate": 2.690410561439141e-05, "loss": 0.0093, "step": 20254 }, { "epoch": 2.1959019947961838, "grad_norm": 0.31309959292411804, "learning_rate": 2.6900478746554475e-05, "loss": 0.0103, "step": 20255 }, { "epoch": 2.1960104076322637, "grad_norm": 0.8874375224113464, "learning_rate": 2.689685187871754e-05, "loss": 0.0431, "step": 20256 }, { "epoch": 2.1961188204683433, "grad_norm": 0.013455485925078392, "learning_rate": 2.6893225010880606e-05, "loss": 0.0005, "step": 20257 }, { "epoch": 2.1962272333044233, "grad_norm": 0.047668565064668655, "learning_rate": 2.688959814304367e-05, "loss": 0.0025, "step": 20258 }, { "epoch": 2.1963356461405033, "grad_norm": 0.17547903954982758, "learning_rate": 2.6885971275206734e-05, "loss": 0.0051, "step": 20259 }, { "epoch": 2.196444058976583, "grad_norm": 0.10909546166658401, "learning_rate": 2.68823444073698e-05, "loss": 0.0028, "step": 20260 }, { "epoch": 2.1965524718126628, "grad_norm": 0.12482701987028122, "learning_rate": 2.687871753953286e-05, "loss": 0.001, "step": 20261 }, { "epoch": 2.1966608846487423, "grad_norm": 0.40256211161613464, "learning_rate": 2.6875090671695923e-05, "loss": 0.0227, "step": 20262 }, { "epoch": 2.1967692974848223, "grad_norm": 0.25933754444122314, "learning_rate": 2.6871463803858987e-05, "loss": 0.0093, "step": 20263 }, { "epoch": 2.196877710320902, "grad_norm": 0.31910306215286255, "learning_rate": 2.686783693602205e-05, "loss": 0.0338, "step": 20264 }, { "epoch": 2.196986123156982, "grad_norm": 1.144392490386963, "learning_rate": 2.6864210068185116e-05, "loss": 0.0158, "step": 20265 }, { "epoch": 2.1970945359930614, "grad_norm": 0.7763740420341492, "learning_rate": 2.6860583200348183e-05, "loss": 0.0288, "step": 20266 }, { "epoch": 2.1972029488291414, "grad_norm": 0.16920630633831024, "learning_rate": 2.6856956332511247e-05, "loss": 0.0052, "step": 20267 }, { "epoch": 2.1973113616652213, "grad_norm": 0.0016589784063398838, "learning_rate": 2.6853329464674308e-05, "loss": 0.0001, "step": 20268 }, { "epoch": 2.197419774501301, "grad_norm": 0.30418485403060913, "learning_rate": 2.6849702596837372e-05, "loss": 0.0316, "step": 20269 }, { "epoch": 2.197528187337381, "grad_norm": 0.09971851855516434, "learning_rate": 2.6846075729000436e-05, "loss": 0.0029, "step": 20270 }, { "epoch": 2.1976366001734604, "grad_norm": 0.015591775998473167, "learning_rate": 2.68424488611635e-05, "loss": 0.0014, "step": 20271 }, { "epoch": 2.1977450130095404, "grad_norm": 0.7693046927452087, "learning_rate": 2.6838821993326564e-05, "loss": 0.017, "step": 20272 }, { "epoch": 2.19785342584562, "grad_norm": 0.29250282049179077, "learning_rate": 2.6835195125489625e-05, "loss": 0.0136, "step": 20273 }, { "epoch": 2.1979618386817, "grad_norm": 0.4041667878627777, "learning_rate": 2.6831568257652696e-05, "loss": 0.0172, "step": 20274 }, { "epoch": 2.19807025151778, "grad_norm": 0.03103492595255375, "learning_rate": 2.682794138981576e-05, "loss": 0.0017, "step": 20275 }, { "epoch": 2.1981786643538594, "grad_norm": 0.1303878277540207, "learning_rate": 2.682431452197882e-05, "loss": 0.0036, "step": 20276 }, { "epoch": 2.1982870771899394, "grad_norm": 0.3090706467628479, "learning_rate": 2.6820687654141885e-05, "loss": 0.0045, "step": 20277 }, { "epoch": 2.198395490026019, "grad_norm": 1.7211968898773193, "learning_rate": 2.681706078630495e-05, "loss": 0.0449, "step": 20278 }, { "epoch": 2.198503902862099, "grad_norm": 0.44942420721054077, "learning_rate": 2.6813433918468013e-05, "loss": 0.0263, "step": 20279 }, { "epoch": 2.1986123156981785, "grad_norm": 0.21932858228683472, "learning_rate": 2.6809807050631074e-05, "loss": 0.005, "step": 20280 }, { "epoch": 2.1987207285342585, "grad_norm": 0.4201655089855194, "learning_rate": 2.6806180182794138e-05, "loss": 0.0232, "step": 20281 }, { "epoch": 2.1988291413703385, "grad_norm": 0.21275769174098969, "learning_rate": 2.6802553314957202e-05, "loss": 0.0104, "step": 20282 }, { "epoch": 2.198937554206418, "grad_norm": 0.319120854139328, "learning_rate": 2.679892644712027e-05, "loss": 0.01, "step": 20283 }, { "epoch": 2.199045967042498, "grad_norm": 0.11337641626596451, "learning_rate": 2.6795299579283334e-05, "loss": 0.006, "step": 20284 }, { "epoch": 2.1991543798785775, "grad_norm": 0.09186289459466934, "learning_rate": 2.6791672711446398e-05, "loss": 0.0039, "step": 20285 }, { "epoch": 2.1992627927146575, "grad_norm": 0.11652157455682755, "learning_rate": 2.6788045843609462e-05, "loss": 0.0092, "step": 20286 }, { "epoch": 2.199371205550737, "grad_norm": 0.04534243047237396, "learning_rate": 2.6784418975772522e-05, "loss": 0.0005, "step": 20287 }, { "epoch": 2.199479618386817, "grad_norm": 0.15385538339614868, "learning_rate": 2.6780792107935587e-05, "loss": 0.009, "step": 20288 }, { "epoch": 2.1995880312228966, "grad_norm": 0.9766150712966919, "learning_rate": 2.677716524009865e-05, "loss": 0.0198, "step": 20289 }, { "epoch": 2.1996964440589766, "grad_norm": 0.017537062987685204, "learning_rate": 2.6773538372261715e-05, "loss": 0.0007, "step": 20290 }, { "epoch": 2.1998048568950566, "grad_norm": 0.15250521898269653, "learning_rate": 2.6769911504424782e-05, "loss": 0.0036, "step": 20291 }, { "epoch": 2.199913269731136, "grad_norm": 0.4311025142669678, "learning_rate": 2.6766284636587846e-05, "loss": 0.0226, "step": 20292 }, { "epoch": 2.200021682567216, "grad_norm": 0.2787629961967468, "learning_rate": 2.676265776875091e-05, "loss": 0.018, "step": 20293 }, { "epoch": 2.2001300954032956, "grad_norm": 0.3168792724609375, "learning_rate": 2.675903090091397e-05, "loss": 0.011, "step": 20294 }, { "epoch": 2.2002385082393756, "grad_norm": 0.3270113468170166, "learning_rate": 2.6755404033077035e-05, "loss": 0.0067, "step": 20295 }, { "epoch": 2.200346921075455, "grad_norm": 0.01859758235514164, "learning_rate": 2.67517771652401e-05, "loss": 0.0008, "step": 20296 }, { "epoch": 2.200455333911535, "grad_norm": 0.47992730140686035, "learning_rate": 2.6748150297403164e-05, "loss": 0.0148, "step": 20297 }, { "epoch": 2.200563746747615, "grad_norm": 0.5666335821151733, "learning_rate": 2.6744523429566228e-05, "loss": 0.0303, "step": 20298 }, { "epoch": 2.2006721595836947, "grad_norm": 0.3217761516571045, "learning_rate": 2.674089656172929e-05, "loss": 0.0109, "step": 20299 }, { "epoch": 2.2007805724197746, "grad_norm": 0.793358325958252, "learning_rate": 2.673726969389236e-05, "loss": 0.0175, "step": 20300 }, { "epoch": 2.200888985255854, "grad_norm": 0.3162403404712677, "learning_rate": 2.6733642826055423e-05, "loss": 0.0144, "step": 20301 }, { "epoch": 2.200997398091934, "grad_norm": 0.0509801022708416, "learning_rate": 2.6730015958218484e-05, "loss": 0.0008, "step": 20302 }, { "epoch": 2.2011058109280137, "grad_norm": 0.11284984648227692, "learning_rate": 2.6726389090381548e-05, "loss": 0.0033, "step": 20303 }, { "epoch": 2.2012142237640937, "grad_norm": 0.384907066822052, "learning_rate": 2.6722762222544612e-05, "loss": 0.0343, "step": 20304 }, { "epoch": 2.2013226366001737, "grad_norm": 0.07200159877538681, "learning_rate": 2.6719135354707676e-05, "loss": 0.0015, "step": 20305 }, { "epoch": 2.2014310494362532, "grad_norm": 0.2146218866109848, "learning_rate": 2.6715508486870737e-05, "loss": 0.004, "step": 20306 }, { "epoch": 2.201539462272333, "grad_norm": 0.1462472528219223, "learning_rate": 2.67118816190338e-05, "loss": 0.0057, "step": 20307 }, { "epoch": 2.2016478751084128, "grad_norm": 0.25283902883529663, "learning_rate": 2.6708254751196865e-05, "loss": 0.003, "step": 20308 }, { "epoch": 2.2017562879444927, "grad_norm": 0.21612001955509186, "learning_rate": 2.6704627883359933e-05, "loss": 0.0044, "step": 20309 }, { "epoch": 2.2018647007805723, "grad_norm": 0.22896835207939148, "learning_rate": 2.6701001015522997e-05, "loss": 0.009, "step": 20310 }, { "epoch": 2.2019731136166523, "grad_norm": 0.03439106419682503, "learning_rate": 2.669737414768606e-05, "loss": 0.0016, "step": 20311 }, { "epoch": 2.202081526452732, "grad_norm": 0.31369632482528687, "learning_rate": 2.6693747279849125e-05, "loss": 0.0264, "step": 20312 }, { "epoch": 2.202189939288812, "grad_norm": 0.5850998163223267, "learning_rate": 2.6690120412012186e-05, "loss": 0.0493, "step": 20313 }, { "epoch": 2.202298352124892, "grad_norm": 0.4178762435913086, "learning_rate": 2.668649354417525e-05, "loss": 0.0182, "step": 20314 }, { "epoch": 2.2024067649609713, "grad_norm": 0.3323060870170593, "learning_rate": 2.6682866676338314e-05, "loss": 0.0145, "step": 20315 }, { "epoch": 2.2025151777970513, "grad_norm": 0.049269482493400574, "learning_rate": 2.6679239808501378e-05, "loss": 0.0038, "step": 20316 }, { "epoch": 2.202623590633131, "grad_norm": 0.5932125449180603, "learning_rate": 2.6675612940664446e-05, "loss": 0.0258, "step": 20317 }, { "epoch": 2.202732003469211, "grad_norm": 0.4436738193035126, "learning_rate": 2.667198607282751e-05, "loss": 0.0138, "step": 20318 }, { "epoch": 2.2028404163052904, "grad_norm": 0.6724504232406616, "learning_rate": 2.6668359204990574e-05, "loss": 0.039, "step": 20319 }, { "epoch": 2.2029488291413704, "grad_norm": 0.011537323705852032, "learning_rate": 2.6664732337153635e-05, "loss": 0.0004, "step": 20320 }, { "epoch": 2.20305724197745, "grad_norm": 0.17665229737758636, "learning_rate": 2.66611054693167e-05, "loss": 0.0066, "step": 20321 }, { "epoch": 2.20316565481353, "grad_norm": 0.10087208449840546, "learning_rate": 2.6657478601479763e-05, "loss": 0.0021, "step": 20322 }, { "epoch": 2.20327406764961, "grad_norm": 0.1441391557455063, "learning_rate": 2.6653851733642827e-05, "loss": 0.0028, "step": 20323 }, { "epoch": 2.2033824804856894, "grad_norm": 0.18530923128128052, "learning_rate": 2.665022486580589e-05, "loss": 0.007, "step": 20324 }, { "epoch": 2.2034908933217694, "grad_norm": 0.20802362263202667, "learning_rate": 2.6646597997968952e-05, "loss": 0.0135, "step": 20325 }, { "epoch": 2.203599306157849, "grad_norm": 0.454031378030777, "learning_rate": 2.6642971130132023e-05, "loss": 0.0227, "step": 20326 }, { "epoch": 2.203707718993929, "grad_norm": 0.4622000455856323, "learning_rate": 2.6639344262295087e-05, "loss": 0.0052, "step": 20327 }, { "epoch": 2.203816131830009, "grad_norm": 0.1923576146364212, "learning_rate": 2.6635717394458147e-05, "loss": 0.0027, "step": 20328 }, { "epoch": 2.2039245446660884, "grad_norm": 0.2656182646751404, "learning_rate": 2.663209052662121e-05, "loss": 0.0102, "step": 20329 }, { "epoch": 2.2040329575021684, "grad_norm": 0.2697799801826477, "learning_rate": 2.6628463658784276e-05, "loss": 0.0249, "step": 20330 }, { "epoch": 2.204141370338248, "grad_norm": 0.06997840106487274, "learning_rate": 2.662483679094734e-05, "loss": 0.0027, "step": 20331 }, { "epoch": 2.204249783174328, "grad_norm": 1.0614218711853027, "learning_rate": 2.66212099231104e-05, "loss": 0.0398, "step": 20332 }, { "epoch": 2.2043581960104075, "grad_norm": 0.005003216676414013, "learning_rate": 2.6617583055273465e-05, "loss": 0.0002, "step": 20333 }, { "epoch": 2.2044666088464875, "grad_norm": 0.04547610506415367, "learning_rate": 2.661395618743653e-05, "loss": 0.0011, "step": 20334 }, { "epoch": 2.204575021682567, "grad_norm": 0.040504004806280136, "learning_rate": 2.6610329319599596e-05, "loss": 0.0022, "step": 20335 }, { "epoch": 2.204683434518647, "grad_norm": 0.31542056798934937, "learning_rate": 2.660670245176266e-05, "loss": 0.0259, "step": 20336 }, { "epoch": 2.204791847354727, "grad_norm": 0.8861528038978577, "learning_rate": 2.6603075583925724e-05, "loss": 0.0586, "step": 20337 }, { "epoch": 2.2049002601908065, "grad_norm": 0.07964538037776947, "learning_rate": 2.659944871608879e-05, "loss": 0.0035, "step": 20338 }, { "epoch": 2.2050086730268865, "grad_norm": 0.37403979897499084, "learning_rate": 2.659582184825185e-05, "loss": 0.0316, "step": 20339 }, { "epoch": 2.205117085862966, "grad_norm": 0.5734421610832214, "learning_rate": 2.6592194980414913e-05, "loss": 0.0168, "step": 20340 }, { "epoch": 2.205225498699046, "grad_norm": 0.3109769821166992, "learning_rate": 2.6588568112577977e-05, "loss": 0.0129, "step": 20341 }, { "epoch": 2.2053339115351256, "grad_norm": 0.6526346802711487, "learning_rate": 2.658494124474104e-05, "loss": 0.0503, "step": 20342 }, { "epoch": 2.2054423243712056, "grad_norm": 0.0708538070321083, "learning_rate": 2.658131437690411e-05, "loss": 0.0017, "step": 20343 }, { "epoch": 2.205550737207285, "grad_norm": 0.18476465344429016, "learning_rate": 2.6577687509067173e-05, "loss": 0.0112, "step": 20344 }, { "epoch": 2.205659150043365, "grad_norm": 0.22026076912879944, "learning_rate": 2.6574060641230237e-05, "loss": 0.0054, "step": 20345 }, { "epoch": 2.205767562879445, "grad_norm": 0.2357274293899536, "learning_rate": 2.6570433773393298e-05, "loss": 0.0202, "step": 20346 }, { "epoch": 2.2058759757155246, "grad_norm": 0.0078588193282485, "learning_rate": 2.6566806905556362e-05, "loss": 0.0003, "step": 20347 }, { "epoch": 2.2059843885516046, "grad_norm": 0.05169616639614105, "learning_rate": 2.6563180037719426e-05, "loss": 0.0026, "step": 20348 }, { "epoch": 2.206092801387684, "grad_norm": 0.04418068751692772, "learning_rate": 2.655955316988249e-05, "loss": 0.0009, "step": 20349 }, { "epoch": 2.206201214223764, "grad_norm": 0.5484185218811035, "learning_rate": 2.6555926302045554e-05, "loss": 0.039, "step": 20350 }, { "epoch": 2.2063096270598437, "grad_norm": 0.025809984654188156, "learning_rate": 2.6552299434208615e-05, "loss": 0.0008, "step": 20351 }, { "epoch": 2.2064180398959237, "grad_norm": 0.03750743344426155, "learning_rate": 2.6548672566371686e-05, "loss": 0.0013, "step": 20352 }, { "epoch": 2.2065264527320037, "grad_norm": 0.19865958392620087, "learning_rate": 2.654504569853475e-05, "loss": 0.0091, "step": 20353 }, { "epoch": 2.206634865568083, "grad_norm": 0.4851151704788208, "learning_rate": 2.654141883069781e-05, "loss": 0.0263, "step": 20354 }, { "epoch": 2.206743278404163, "grad_norm": 0.032688163220882416, "learning_rate": 2.6537791962860875e-05, "loss": 0.0011, "step": 20355 }, { "epoch": 2.2068516912402427, "grad_norm": 0.0371289998292923, "learning_rate": 2.653416509502394e-05, "loss": 0.0013, "step": 20356 }, { "epoch": 2.2069601040763227, "grad_norm": 0.17205487191677094, "learning_rate": 2.6530538227187003e-05, "loss": 0.0044, "step": 20357 }, { "epoch": 2.2070685169124022, "grad_norm": 0.5446401238441467, "learning_rate": 2.6526911359350064e-05, "loss": 0.0199, "step": 20358 }, { "epoch": 2.2071769297484822, "grad_norm": 0.5398255586624146, "learning_rate": 2.6523284491513128e-05, "loss": 0.0186, "step": 20359 }, { "epoch": 2.207285342584562, "grad_norm": 0.021754518151283264, "learning_rate": 2.6519657623676192e-05, "loss": 0.0006, "step": 20360 }, { "epoch": 2.2073937554206418, "grad_norm": 0.06631404906511307, "learning_rate": 2.651603075583926e-05, "loss": 0.0035, "step": 20361 }, { "epoch": 2.2075021682567217, "grad_norm": 0.2505146563053131, "learning_rate": 2.6512403888002324e-05, "loss": 0.0381, "step": 20362 }, { "epoch": 2.2076105810928013, "grad_norm": 0.5458799004554749, "learning_rate": 2.6508777020165388e-05, "loss": 0.033, "step": 20363 }, { "epoch": 2.2077189939288813, "grad_norm": 1.0479624271392822, "learning_rate": 2.6505150152328452e-05, "loss": 0.0101, "step": 20364 }, { "epoch": 2.207827406764961, "grad_norm": 0.18003644049167633, "learning_rate": 2.6501523284491513e-05, "loss": 0.0042, "step": 20365 }, { "epoch": 2.207935819601041, "grad_norm": 0.24468547105789185, "learning_rate": 2.6497896416654577e-05, "loss": 0.011, "step": 20366 }, { "epoch": 2.2080442324371203, "grad_norm": 0.9479695558547974, "learning_rate": 2.649426954881764e-05, "loss": 0.0261, "step": 20367 }, { "epoch": 2.2081526452732003, "grad_norm": 0.2520923614501953, "learning_rate": 2.6490642680980705e-05, "loss": 0.0106, "step": 20368 }, { "epoch": 2.2082610581092803, "grad_norm": 0.37901338934898376, "learning_rate": 2.6487015813143772e-05, "loss": 0.012, "step": 20369 }, { "epoch": 2.20836947094536, "grad_norm": 0.5971913933753967, "learning_rate": 2.6483388945306836e-05, "loss": 0.0135, "step": 20370 }, { "epoch": 2.20847788378144, "grad_norm": 0.0496055893599987, "learning_rate": 2.64797620774699e-05, "loss": 0.0031, "step": 20371 }, { "epoch": 2.2085862966175194, "grad_norm": 0.06225229799747467, "learning_rate": 2.647613520963296e-05, "loss": 0.0025, "step": 20372 }, { "epoch": 2.2086947094535994, "grad_norm": 0.33104410767555237, "learning_rate": 2.6472508341796025e-05, "loss": 0.0352, "step": 20373 }, { "epoch": 2.208803122289679, "grad_norm": 0.337323933839798, "learning_rate": 2.646888147395909e-05, "loss": 0.0147, "step": 20374 }, { "epoch": 2.208911535125759, "grad_norm": 0.838523268699646, "learning_rate": 2.6465254606122154e-05, "loss": 0.0311, "step": 20375 }, { "epoch": 2.209019947961839, "grad_norm": 0.07756441831588745, "learning_rate": 2.6461627738285218e-05, "loss": 0.0033, "step": 20376 }, { "epoch": 2.2091283607979184, "grad_norm": 0.17946496605873108, "learning_rate": 2.645800087044828e-05, "loss": 0.0137, "step": 20377 }, { "epoch": 2.2092367736339984, "grad_norm": 0.07618163526058197, "learning_rate": 2.645437400261135e-05, "loss": 0.0077, "step": 20378 }, { "epoch": 2.209345186470078, "grad_norm": 0.17508842051029205, "learning_rate": 2.6450747134774413e-05, "loss": 0.0046, "step": 20379 }, { "epoch": 2.209453599306158, "grad_norm": 0.1915258765220642, "learning_rate": 2.6447120266937474e-05, "loss": 0.0087, "step": 20380 }, { "epoch": 2.2095620121422375, "grad_norm": 0.2851095497608185, "learning_rate": 2.6443493399100538e-05, "loss": 0.0095, "step": 20381 }, { "epoch": 2.2096704249783174, "grad_norm": 0.11312881112098694, "learning_rate": 2.6439866531263602e-05, "loss": 0.0067, "step": 20382 }, { "epoch": 2.2097788378143974, "grad_norm": 0.5247384905815125, "learning_rate": 2.6436239663426666e-05, "loss": 0.0208, "step": 20383 }, { "epoch": 2.209887250650477, "grad_norm": 0.4186636805534363, "learning_rate": 2.6432612795589727e-05, "loss": 0.0271, "step": 20384 }, { "epoch": 2.209995663486557, "grad_norm": 0.01197507418692112, "learning_rate": 2.642898592775279e-05, "loss": 0.0002, "step": 20385 }, { "epoch": 2.2101040763226365, "grad_norm": 0.16723257303237915, "learning_rate": 2.6425359059915862e-05, "loss": 0.0036, "step": 20386 }, { "epoch": 2.2102124891587165, "grad_norm": 0.058083243668079376, "learning_rate": 2.6421732192078923e-05, "loss": 0.0024, "step": 20387 }, { "epoch": 2.210320901994796, "grad_norm": 0.13466840982437134, "learning_rate": 2.6418105324241987e-05, "loss": 0.0036, "step": 20388 }, { "epoch": 2.210429314830876, "grad_norm": 0.5253906846046448, "learning_rate": 2.641447845640505e-05, "loss": 0.0206, "step": 20389 }, { "epoch": 2.2105377276669556, "grad_norm": 0.5016104578971863, "learning_rate": 2.6410851588568115e-05, "loss": 0.0093, "step": 20390 }, { "epoch": 2.2106461405030355, "grad_norm": 0.18870563805103302, "learning_rate": 2.6407224720731176e-05, "loss": 0.0195, "step": 20391 }, { "epoch": 2.2107545533391155, "grad_norm": 1.0968351364135742, "learning_rate": 2.640359785289424e-05, "loss": 0.0287, "step": 20392 }, { "epoch": 2.210862966175195, "grad_norm": 0.12247232347726822, "learning_rate": 2.6399970985057304e-05, "loss": 0.0026, "step": 20393 }, { "epoch": 2.210971379011275, "grad_norm": 0.04621126875281334, "learning_rate": 2.6396344117220368e-05, "loss": 0.0027, "step": 20394 }, { "epoch": 2.2110797918473546, "grad_norm": 0.03583477437496185, "learning_rate": 2.6392717249383436e-05, "loss": 0.0017, "step": 20395 }, { "epoch": 2.2111882046834346, "grad_norm": 0.5076167583465576, "learning_rate": 2.63890903815465e-05, "loss": 0.0098, "step": 20396 }, { "epoch": 2.211296617519514, "grad_norm": 0.08242014795541763, "learning_rate": 2.6385463513709564e-05, "loss": 0.0024, "step": 20397 }, { "epoch": 2.211405030355594, "grad_norm": 0.028292959555983543, "learning_rate": 2.6381836645872625e-05, "loss": 0.0014, "step": 20398 }, { "epoch": 2.211513443191674, "grad_norm": 0.00732934195548296, "learning_rate": 2.637820977803569e-05, "loss": 0.0003, "step": 20399 }, { "epoch": 2.2116218560277536, "grad_norm": 0.36114242672920227, "learning_rate": 2.6374582910198753e-05, "loss": 0.0215, "step": 20400 }, { "epoch": 2.2117302688638336, "grad_norm": 0.3386140465736389, "learning_rate": 2.6370956042361817e-05, "loss": 0.0173, "step": 20401 }, { "epoch": 2.211838681699913, "grad_norm": 0.20541618764400482, "learning_rate": 2.636732917452488e-05, "loss": 0.0033, "step": 20402 }, { "epoch": 2.211947094535993, "grad_norm": 0.6116564869880676, "learning_rate": 2.6363702306687942e-05, "loss": 0.0165, "step": 20403 }, { "epoch": 2.2120555073720727, "grad_norm": 0.4289875030517578, "learning_rate": 2.6360075438851013e-05, "loss": 0.0307, "step": 20404 }, { "epoch": 2.2121639202081527, "grad_norm": 0.29785671830177307, "learning_rate": 2.6356448571014077e-05, "loss": 0.0568, "step": 20405 }, { "epoch": 2.2122723330442327, "grad_norm": 0.021036555990576744, "learning_rate": 2.6352821703177137e-05, "loss": 0.0007, "step": 20406 }, { "epoch": 2.212380745880312, "grad_norm": 0.44853997230529785, "learning_rate": 2.63491948353402e-05, "loss": 0.0479, "step": 20407 }, { "epoch": 2.212489158716392, "grad_norm": 0.14461612701416016, "learning_rate": 2.6345567967503266e-05, "loss": 0.0056, "step": 20408 }, { "epoch": 2.2125975715524717, "grad_norm": 0.2382323443889618, "learning_rate": 2.634194109966633e-05, "loss": 0.0078, "step": 20409 }, { "epoch": 2.2127059843885517, "grad_norm": 0.04937906190752983, "learning_rate": 2.633831423182939e-05, "loss": 0.0025, "step": 20410 }, { "epoch": 2.2128143972246312, "grad_norm": 0.08640727400779724, "learning_rate": 2.6334687363992455e-05, "loss": 0.0041, "step": 20411 }, { "epoch": 2.2129228100607112, "grad_norm": 0.55048668384552, "learning_rate": 2.6331060496155525e-05, "loss": 0.0235, "step": 20412 }, { "epoch": 2.2130312228967908, "grad_norm": 0.16916456818580627, "learning_rate": 2.6327433628318586e-05, "loss": 0.0075, "step": 20413 }, { "epoch": 2.2131396357328708, "grad_norm": 0.1030150055885315, "learning_rate": 2.632380676048165e-05, "loss": 0.0047, "step": 20414 }, { "epoch": 2.2132480485689507, "grad_norm": 0.38621485233306885, "learning_rate": 2.6320179892644714e-05, "loss": 0.0204, "step": 20415 }, { "epoch": 2.2133564614050303, "grad_norm": 0.20584093034267426, "learning_rate": 2.631655302480778e-05, "loss": 0.0128, "step": 20416 }, { "epoch": 2.2134648742411103, "grad_norm": 0.1897944211959839, "learning_rate": 2.631292615697084e-05, "loss": 0.0066, "step": 20417 }, { "epoch": 2.21357328707719, "grad_norm": 0.10154253989458084, "learning_rate": 2.6309299289133903e-05, "loss": 0.002, "step": 20418 }, { "epoch": 2.21368169991327, "grad_norm": 0.1098906472325325, "learning_rate": 2.6305672421296967e-05, "loss": 0.0027, "step": 20419 }, { "epoch": 2.2137901127493493, "grad_norm": 1.0399160385131836, "learning_rate": 2.630204555346003e-05, "loss": 0.0324, "step": 20420 }, { "epoch": 2.2138985255854293, "grad_norm": 0.5543344020843506, "learning_rate": 2.62984186856231e-05, "loss": 0.0416, "step": 20421 }, { "epoch": 2.2140069384215093, "grad_norm": 0.19123704731464386, "learning_rate": 2.6294791817786163e-05, "loss": 0.0141, "step": 20422 }, { "epoch": 2.214115351257589, "grad_norm": 0.22452929615974426, "learning_rate": 2.6291164949949227e-05, "loss": 0.0172, "step": 20423 }, { "epoch": 2.214223764093669, "grad_norm": 0.06867078691720963, "learning_rate": 2.6287538082112288e-05, "loss": 0.0021, "step": 20424 }, { "epoch": 2.2143321769297484, "grad_norm": 0.18484167754650116, "learning_rate": 2.6283911214275352e-05, "loss": 0.0163, "step": 20425 }, { "epoch": 2.2144405897658284, "grad_norm": 0.2479550540447235, "learning_rate": 2.6280284346438416e-05, "loss": 0.0064, "step": 20426 }, { "epoch": 2.214549002601908, "grad_norm": 0.2672143876552582, "learning_rate": 2.627665747860148e-05, "loss": 0.0213, "step": 20427 }, { "epoch": 2.214657415437988, "grad_norm": 0.3092128336429596, "learning_rate": 2.6273030610764544e-05, "loss": 0.0338, "step": 20428 }, { "epoch": 2.214765828274068, "grad_norm": 0.1793333739042282, "learning_rate": 2.6269403742927605e-05, "loss": 0.0215, "step": 20429 }, { "epoch": 2.2148742411101474, "grad_norm": 0.5462740063667297, "learning_rate": 2.6265776875090676e-05, "loss": 0.0114, "step": 20430 }, { "epoch": 2.2149826539462274, "grad_norm": 0.3960001766681671, "learning_rate": 2.626215000725374e-05, "loss": 0.041, "step": 20431 }, { "epoch": 2.215091066782307, "grad_norm": 0.33777520060539246, "learning_rate": 2.62585231394168e-05, "loss": 0.0595, "step": 20432 }, { "epoch": 2.215199479618387, "grad_norm": 0.21093399822711945, "learning_rate": 2.6254896271579865e-05, "loss": 0.0077, "step": 20433 }, { "epoch": 2.2153078924544665, "grad_norm": 0.05877342075109482, "learning_rate": 2.625126940374293e-05, "loss": 0.0014, "step": 20434 }, { "epoch": 2.2154163052905465, "grad_norm": 0.31422150135040283, "learning_rate": 2.6247642535905993e-05, "loss": 0.0123, "step": 20435 }, { "epoch": 2.215524718126626, "grad_norm": 0.08330630511045456, "learning_rate": 2.6244015668069054e-05, "loss": 0.0034, "step": 20436 }, { "epoch": 2.215633130962706, "grad_norm": 0.2325442135334015, "learning_rate": 2.6240388800232118e-05, "loss": 0.0095, "step": 20437 }, { "epoch": 2.215741543798786, "grad_norm": 0.12490047514438629, "learning_rate": 2.623676193239519e-05, "loss": 0.0076, "step": 20438 }, { "epoch": 2.2158499566348655, "grad_norm": 0.9596431851387024, "learning_rate": 2.623313506455825e-05, "loss": 0.0207, "step": 20439 }, { "epoch": 2.2159583694709455, "grad_norm": 0.26120322942733765, "learning_rate": 2.6229508196721314e-05, "loss": 0.0219, "step": 20440 }, { "epoch": 2.216066782307025, "grad_norm": 0.19937284290790558, "learning_rate": 2.6225881328884378e-05, "loss": 0.0131, "step": 20441 }, { "epoch": 2.216175195143105, "grad_norm": 0.3822804391384125, "learning_rate": 2.6222254461047442e-05, "loss": 0.0211, "step": 20442 }, { "epoch": 2.2162836079791846, "grad_norm": 0.08891277760267258, "learning_rate": 2.6218627593210503e-05, "loss": 0.0016, "step": 20443 }, { "epoch": 2.2163920208152645, "grad_norm": 0.05723361298441887, "learning_rate": 2.6215000725373567e-05, "loss": 0.0025, "step": 20444 }, { "epoch": 2.2165004336513445, "grad_norm": 0.264003723859787, "learning_rate": 2.621137385753663e-05, "loss": 0.0315, "step": 20445 }, { "epoch": 2.216608846487424, "grad_norm": 0.03212973102927208, "learning_rate": 2.6207746989699695e-05, "loss": 0.0012, "step": 20446 }, { "epoch": 2.216717259323504, "grad_norm": 0.15312515199184418, "learning_rate": 2.6204120121862762e-05, "loss": 0.0066, "step": 20447 }, { "epoch": 2.2168256721595836, "grad_norm": 0.2547338306903839, "learning_rate": 2.6200493254025826e-05, "loss": 0.0076, "step": 20448 }, { "epoch": 2.2169340849956636, "grad_norm": 0.9128046631813049, "learning_rate": 2.619686638618889e-05, "loss": 0.0438, "step": 20449 }, { "epoch": 2.217042497831743, "grad_norm": 0.15040387213230133, "learning_rate": 2.6193239518351955e-05, "loss": 0.0082, "step": 20450 }, { "epoch": 2.217150910667823, "grad_norm": 0.3043955862522125, "learning_rate": 2.6189612650515015e-05, "loss": 0.0133, "step": 20451 }, { "epoch": 2.217259323503903, "grad_norm": 0.332500159740448, "learning_rate": 2.618598578267808e-05, "loss": 0.0119, "step": 20452 }, { "epoch": 2.2173677363399826, "grad_norm": 0.663490355014801, "learning_rate": 2.6182358914841144e-05, "loss": 0.022, "step": 20453 }, { "epoch": 2.2174761491760626, "grad_norm": 0.2624356746673584, "learning_rate": 2.6178732047004208e-05, "loss": 0.0145, "step": 20454 }, { "epoch": 2.217584562012142, "grad_norm": 0.14058290421962738, "learning_rate": 2.617510517916727e-05, "loss": 0.0036, "step": 20455 }, { "epoch": 2.217692974848222, "grad_norm": 0.15889421105384827, "learning_rate": 2.617147831133034e-05, "loss": 0.0057, "step": 20456 }, { "epoch": 2.2178013876843017, "grad_norm": 0.4072410762310028, "learning_rate": 2.6167851443493403e-05, "loss": 0.0085, "step": 20457 }, { "epoch": 2.2179098005203817, "grad_norm": 0.21126163005828857, "learning_rate": 2.6164224575656464e-05, "loss": 0.0146, "step": 20458 }, { "epoch": 2.218018213356461, "grad_norm": 1.0043097734451294, "learning_rate": 2.6160597707819528e-05, "loss": 0.0178, "step": 20459 }, { "epoch": 2.218126626192541, "grad_norm": 1.239183783531189, "learning_rate": 2.6156970839982592e-05, "loss": 0.0081, "step": 20460 }, { "epoch": 2.218235039028621, "grad_norm": 0.16463014483451843, "learning_rate": 2.6153343972145656e-05, "loss": 0.0329, "step": 20461 }, { "epoch": 2.2183434518647007, "grad_norm": 0.2119622826576233, "learning_rate": 2.6149717104308717e-05, "loss": 0.0117, "step": 20462 }, { "epoch": 2.2184518647007807, "grad_norm": 0.11824488639831543, "learning_rate": 2.614609023647178e-05, "loss": 0.0061, "step": 20463 }, { "epoch": 2.2185602775368602, "grad_norm": 0.12435683608055115, "learning_rate": 2.6142463368634852e-05, "loss": 0.007, "step": 20464 }, { "epoch": 2.2186686903729402, "grad_norm": 0.2259911745786667, "learning_rate": 2.6138836500797913e-05, "loss": 0.0156, "step": 20465 }, { "epoch": 2.2187771032090198, "grad_norm": 0.23476096987724304, "learning_rate": 2.6135209632960977e-05, "loss": 0.0128, "step": 20466 }, { "epoch": 2.2188855160450998, "grad_norm": 0.6489803791046143, "learning_rate": 2.613158276512404e-05, "loss": 0.0247, "step": 20467 }, { "epoch": 2.2189939288811793, "grad_norm": 0.07958425581455231, "learning_rate": 2.6127955897287105e-05, "loss": 0.0029, "step": 20468 }, { "epoch": 2.2191023417172593, "grad_norm": 0.20381315052509308, "learning_rate": 2.6124329029450166e-05, "loss": 0.0085, "step": 20469 }, { "epoch": 2.2192107545533393, "grad_norm": 0.024917636066675186, "learning_rate": 2.612070216161323e-05, "loss": 0.0009, "step": 20470 }, { "epoch": 2.219319167389419, "grad_norm": 0.37684422731399536, "learning_rate": 2.6117075293776294e-05, "loss": 0.0158, "step": 20471 }, { "epoch": 2.219427580225499, "grad_norm": 0.05379609391093254, "learning_rate": 2.6113448425939358e-05, "loss": 0.0025, "step": 20472 }, { "epoch": 2.2195359930615783, "grad_norm": 0.3293430805206299, "learning_rate": 2.6109821558102426e-05, "loss": 0.0141, "step": 20473 }, { "epoch": 2.2196444058976583, "grad_norm": 0.20725777745246887, "learning_rate": 2.610619469026549e-05, "loss": 0.0051, "step": 20474 }, { "epoch": 2.2197528187337383, "grad_norm": 0.08223430067300797, "learning_rate": 2.6102567822428554e-05, "loss": 0.0025, "step": 20475 }, { "epoch": 2.219861231569818, "grad_norm": 0.4275589883327484, "learning_rate": 2.6098940954591618e-05, "loss": 0.0282, "step": 20476 }, { "epoch": 2.219969644405898, "grad_norm": 0.30915793776512146, "learning_rate": 2.609531408675468e-05, "loss": 0.0158, "step": 20477 }, { "epoch": 2.2200780572419774, "grad_norm": 0.35402777791023254, "learning_rate": 2.6091687218917743e-05, "loss": 0.016, "step": 20478 }, { "epoch": 2.2201864700780574, "grad_norm": 0.13620123267173767, "learning_rate": 2.6088060351080807e-05, "loss": 0.0031, "step": 20479 }, { "epoch": 2.220294882914137, "grad_norm": 0.20208333432674408, "learning_rate": 2.608443348324387e-05, "loss": 0.0054, "step": 20480 }, { "epoch": 2.220403295750217, "grad_norm": 0.3153170943260193, "learning_rate": 2.608080661540694e-05, "loss": 0.0116, "step": 20481 }, { "epoch": 2.2205117085862964, "grad_norm": 0.07075671851634979, "learning_rate": 2.6077179747570003e-05, "loss": 0.0035, "step": 20482 }, { "epoch": 2.2206201214223764, "grad_norm": 0.1229262575507164, "learning_rate": 2.6073552879733067e-05, "loss": 0.0095, "step": 20483 }, { "epoch": 2.2207285342584564, "grad_norm": 0.15333032608032227, "learning_rate": 2.6069926011896127e-05, "loss": 0.014, "step": 20484 }, { "epoch": 2.220836947094536, "grad_norm": 0.34478625655174255, "learning_rate": 2.606629914405919e-05, "loss": 0.039, "step": 20485 }, { "epoch": 2.220945359930616, "grad_norm": 0.36893579363822937, "learning_rate": 2.6062672276222256e-05, "loss": 0.0267, "step": 20486 }, { "epoch": 2.2210537727666955, "grad_norm": 0.06448099762201309, "learning_rate": 2.605904540838532e-05, "loss": 0.0031, "step": 20487 }, { "epoch": 2.2211621856027755, "grad_norm": 0.29257357120513916, "learning_rate": 2.605541854054838e-05, "loss": 0.0251, "step": 20488 }, { "epoch": 2.221270598438855, "grad_norm": 0.273454874753952, "learning_rate": 2.6051791672711445e-05, "loss": 0.01, "step": 20489 }, { "epoch": 2.221379011274935, "grad_norm": 1.1567457914352417, "learning_rate": 2.6048164804874515e-05, "loss": 0.0228, "step": 20490 }, { "epoch": 2.2214874241110145, "grad_norm": 0.08936028182506561, "learning_rate": 2.6044537937037576e-05, "loss": 0.0039, "step": 20491 }, { "epoch": 2.2215958369470945, "grad_norm": 0.12007641047239304, "learning_rate": 2.604091106920064e-05, "loss": 0.0039, "step": 20492 }, { "epoch": 2.2217042497831745, "grad_norm": 0.024414028972387314, "learning_rate": 2.6037284201363704e-05, "loss": 0.0013, "step": 20493 }, { "epoch": 2.221812662619254, "grad_norm": 0.30709362030029297, "learning_rate": 2.603365733352677e-05, "loss": 0.0221, "step": 20494 }, { "epoch": 2.221921075455334, "grad_norm": 0.35519903898239136, "learning_rate": 2.603003046568983e-05, "loss": 0.0171, "step": 20495 }, { "epoch": 2.2220294882914136, "grad_norm": 0.15743589401245117, "learning_rate": 2.6026403597852893e-05, "loss": 0.0179, "step": 20496 }, { "epoch": 2.2221379011274935, "grad_norm": 0.2969105839729309, "learning_rate": 2.6022776730015957e-05, "loss": 0.008, "step": 20497 }, { "epoch": 2.222246313963573, "grad_norm": 1.7324012517929077, "learning_rate": 2.601914986217902e-05, "loss": 0.0496, "step": 20498 }, { "epoch": 2.222354726799653, "grad_norm": 0.47045207023620605, "learning_rate": 2.601552299434209e-05, "loss": 0.006, "step": 20499 }, { "epoch": 2.222463139635733, "grad_norm": 0.13988757133483887, "learning_rate": 2.6011896126505153e-05, "loss": 0.0097, "step": 20500 }, { "epoch": 2.2225715524718126, "grad_norm": 0.39119160175323486, "learning_rate": 2.6008269258668217e-05, "loss": 0.0134, "step": 20501 }, { "epoch": 2.2226799653078926, "grad_norm": 1.387442946434021, "learning_rate": 2.600464239083128e-05, "loss": 0.0467, "step": 20502 }, { "epoch": 2.222788378143972, "grad_norm": 0.2573024332523346, "learning_rate": 2.6001015522994342e-05, "loss": 0.0056, "step": 20503 }, { "epoch": 2.222896790980052, "grad_norm": 0.2302689403295517, "learning_rate": 2.5997388655157406e-05, "loss": 0.0049, "step": 20504 }, { "epoch": 2.2230052038161316, "grad_norm": 0.0374494269490242, "learning_rate": 2.599376178732047e-05, "loss": 0.0019, "step": 20505 }, { "epoch": 2.2231136166522116, "grad_norm": 0.288585364818573, "learning_rate": 2.5990134919483534e-05, "loss": 0.0109, "step": 20506 }, { "epoch": 2.2232220294882916, "grad_norm": 0.06288176774978638, "learning_rate": 2.5986508051646602e-05, "loss": 0.0023, "step": 20507 }, { "epoch": 2.223330442324371, "grad_norm": 0.2701221704483032, "learning_rate": 2.5982881183809666e-05, "loss": 0.0123, "step": 20508 }, { "epoch": 2.223438855160451, "grad_norm": 0.5511100888252258, "learning_rate": 2.597925431597273e-05, "loss": 0.05, "step": 20509 }, { "epoch": 2.2235472679965307, "grad_norm": 0.9313886165618896, "learning_rate": 2.597562744813579e-05, "loss": 0.0168, "step": 20510 }, { "epoch": 2.2236556808326107, "grad_norm": 0.08373485505580902, "learning_rate": 2.5972000580298855e-05, "loss": 0.0053, "step": 20511 }, { "epoch": 2.22376409366869, "grad_norm": 0.03578402101993561, "learning_rate": 2.596837371246192e-05, "loss": 0.0017, "step": 20512 }, { "epoch": 2.22387250650477, "grad_norm": 0.06562213599681854, "learning_rate": 2.5964746844624983e-05, "loss": 0.0023, "step": 20513 }, { "epoch": 2.2239809193408497, "grad_norm": 0.24406081438064575, "learning_rate": 2.5961119976788044e-05, "loss": 0.0166, "step": 20514 }, { "epoch": 2.2240893321769297, "grad_norm": 0.34064483642578125, "learning_rate": 2.5957493108951108e-05, "loss": 0.02, "step": 20515 }, { "epoch": 2.2241977450130097, "grad_norm": 0.04047362133860588, "learning_rate": 2.595386624111418e-05, "loss": 0.002, "step": 20516 }, { "epoch": 2.2243061578490892, "grad_norm": 0.2970673739910126, "learning_rate": 2.595023937327724e-05, "loss": 0.0222, "step": 20517 }, { "epoch": 2.2244145706851692, "grad_norm": 0.13157980144023895, "learning_rate": 2.5946612505440304e-05, "loss": 0.0037, "step": 20518 }, { "epoch": 2.2245229835212488, "grad_norm": 0.6665973663330078, "learning_rate": 2.5942985637603368e-05, "loss": 0.0198, "step": 20519 }, { "epoch": 2.2246313963573288, "grad_norm": 0.4033890664577484, "learning_rate": 2.5939358769766432e-05, "loss": 0.0208, "step": 20520 }, { "epoch": 2.2247398091934083, "grad_norm": 0.08100636303424835, "learning_rate": 2.5935731901929493e-05, "loss": 0.0036, "step": 20521 }, { "epoch": 2.2248482220294883, "grad_norm": 2.603133201599121, "learning_rate": 2.5932105034092557e-05, "loss": 0.0288, "step": 20522 }, { "epoch": 2.2249566348655683, "grad_norm": 0.3090175688266754, "learning_rate": 2.592847816625562e-05, "loss": 0.008, "step": 20523 }, { "epoch": 2.225065047701648, "grad_norm": 0.1073075607419014, "learning_rate": 2.5924851298418685e-05, "loss": 0.0048, "step": 20524 }, { "epoch": 2.225173460537728, "grad_norm": 0.3271026909351349, "learning_rate": 2.5921224430581752e-05, "loss": 0.0149, "step": 20525 }, { "epoch": 2.2252818733738073, "grad_norm": 0.052412353456020355, "learning_rate": 2.5917597562744816e-05, "loss": 0.0018, "step": 20526 }, { "epoch": 2.2253902862098873, "grad_norm": 0.0345749594271183, "learning_rate": 2.591397069490788e-05, "loss": 0.0018, "step": 20527 }, { "epoch": 2.225498699045967, "grad_norm": 0.2294090986251831, "learning_rate": 2.5910343827070945e-05, "loss": 0.0062, "step": 20528 }, { "epoch": 2.225607111882047, "grad_norm": 0.1297631561756134, "learning_rate": 2.5906716959234005e-05, "loss": 0.0081, "step": 20529 }, { "epoch": 2.225715524718127, "grad_norm": 0.1328798085451126, "learning_rate": 2.590309009139707e-05, "loss": 0.0028, "step": 20530 }, { "epoch": 2.2258239375542064, "grad_norm": 0.00466580456122756, "learning_rate": 2.5899463223560134e-05, "loss": 0.0002, "step": 20531 }, { "epoch": 2.2259323503902864, "grad_norm": 0.12162816524505615, "learning_rate": 2.5895836355723198e-05, "loss": 0.0027, "step": 20532 }, { "epoch": 2.226040763226366, "grad_norm": 0.028595926240086555, "learning_rate": 2.5892209487886265e-05, "loss": 0.0012, "step": 20533 }, { "epoch": 2.226149176062446, "grad_norm": 0.010865728370845318, "learning_rate": 2.588858262004933e-05, "loss": 0.0005, "step": 20534 }, { "epoch": 2.2262575888985254, "grad_norm": 0.09441427141427994, "learning_rate": 2.5884955752212393e-05, "loss": 0.0041, "step": 20535 }, { "epoch": 2.2263660017346054, "grad_norm": 0.5203025341033936, "learning_rate": 2.5881328884375454e-05, "loss": 0.0213, "step": 20536 }, { "epoch": 2.226474414570685, "grad_norm": 0.04420589655637741, "learning_rate": 2.5877702016538518e-05, "loss": 0.0011, "step": 20537 }, { "epoch": 2.226582827406765, "grad_norm": 0.9138628244400024, "learning_rate": 2.5874075148701582e-05, "loss": 0.0229, "step": 20538 }, { "epoch": 2.226691240242845, "grad_norm": 0.9925790429115295, "learning_rate": 2.5870448280864646e-05, "loss": 0.0648, "step": 20539 }, { "epoch": 2.2267996530789245, "grad_norm": 0.2150135189294815, "learning_rate": 2.5866821413027707e-05, "loss": 0.0182, "step": 20540 }, { "epoch": 2.2269080659150045, "grad_norm": 0.5428144335746765, "learning_rate": 2.586319454519077e-05, "loss": 0.0142, "step": 20541 }, { "epoch": 2.227016478751084, "grad_norm": 0.39508628845214844, "learning_rate": 2.5859567677353842e-05, "loss": 0.0368, "step": 20542 }, { "epoch": 2.227124891587164, "grad_norm": 0.4310813546180725, "learning_rate": 2.5855940809516903e-05, "loss": 0.0234, "step": 20543 }, { "epoch": 2.2272333044232435, "grad_norm": 0.5197627544403076, "learning_rate": 2.5852313941679967e-05, "loss": 0.0188, "step": 20544 }, { "epoch": 2.2273417172593235, "grad_norm": 0.04666389524936676, "learning_rate": 2.584868707384303e-05, "loss": 0.0034, "step": 20545 }, { "epoch": 2.2274501300954035, "grad_norm": 0.20051251351833344, "learning_rate": 2.5845060206006095e-05, "loss": 0.0066, "step": 20546 }, { "epoch": 2.227558542931483, "grad_norm": 0.5337197184562683, "learning_rate": 2.5841433338169156e-05, "loss": 0.0284, "step": 20547 }, { "epoch": 2.227666955767563, "grad_norm": 0.2622639238834381, "learning_rate": 2.583780647033222e-05, "loss": 0.0189, "step": 20548 }, { "epoch": 2.2277753686036426, "grad_norm": 0.5475643873214722, "learning_rate": 2.5834179602495284e-05, "loss": 0.0257, "step": 20549 }, { "epoch": 2.2278837814397225, "grad_norm": 0.027227748185396194, "learning_rate": 2.5830552734658348e-05, "loss": 0.0004, "step": 20550 }, { "epoch": 2.227992194275802, "grad_norm": 0.5236145853996277, "learning_rate": 2.5826925866821416e-05, "loss": 0.036, "step": 20551 }, { "epoch": 2.228100607111882, "grad_norm": 0.327388197183609, "learning_rate": 2.582329899898448e-05, "loss": 0.0072, "step": 20552 }, { "epoch": 2.228209019947962, "grad_norm": 0.4713682234287262, "learning_rate": 2.5819672131147544e-05, "loss": 0.0317, "step": 20553 }, { "epoch": 2.2283174327840416, "grad_norm": 0.3194127380847931, "learning_rate": 2.5816045263310608e-05, "loss": 0.0056, "step": 20554 }, { "epoch": 2.2284258456201216, "grad_norm": 0.34273824095726013, "learning_rate": 2.581241839547367e-05, "loss": 0.0118, "step": 20555 }, { "epoch": 2.228534258456201, "grad_norm": 0.4726010859012604, "learning_rate": 2.5808791527636733e-05, "loss": 0.0479, "step": 20556 }, { "epoch": 2.228642671292281, "grad_norm": 0.09693196415901184, "learning_rate": 2.5805164659799797e-05, "loss": 0.0016, "step": 20557 }, { "epoch": 2.2287510841283606, "grad_norm": 0.14731912314891815, "learning_rate": 2.580153779196286e-05, "loss": 0.0059, "step": 20558 }, { "epoch": 2.2288594969644406, "grad_norm": 0.021784288808703423, "learning_rate": 2.579791092412593e-05, "loss": 0.0007, "step": 20559 }, { "epoch": 2.22896790980052, "grad_norm": 0.3179944157600403, "learning_rate": 2.5794284056288993e-05, "loss": 0.0156, "step": 20560 }, { "epoch": 2.2290763226366, "grad_norm": 0.06556690484285355, "learning_rate": 2.5790657188452057e-05, "loss": 0.0036, "step": 20561 }, { "epoch": 2.22918473547268, "grad_norm": 0.6137803792953491, "learning_rate": 2.5787030320615118e-05, "loss": 0.0166, "step": 20562 }, { "epoch": 2.2292931483087597, "grad_norm": 0.07950280606746674, "learning_rate": 2.578340345277818e-05, "loss": 0.0037, "step": 20563 }, { "epoch": 2.2294015611448397, "grad_norm": 0.03244227543473244, "learning_rate": 2.5779776584941246e-05, "loss": 0.0013, "step": 20564 }, { "epoch": 2.229509973980919, "grad_norm": 0.1692754030227661, "learning_rate": 2.577614971710431e-05, "loss": 0.0068, "step": 20565 }, { "epoch": 2.229618386816999, "grad_norm": 0.05840495601296425, "learning_rate": 2.577252284926737e-05, "loss": 0.0023, "step": 20566 }, { "epoch": 2.2297267996530787, "grad_norm": 0.16374200582504272, "learning_rate": 2.5768895981430435e-05, "loss": 0.004, "step": 20567 }, { "epoch": 2.2298352124891587, "grad_norm": 0.0444830060005188, "learning_rate": 2.5765269113593506e-05, "loss": 0.0012, "step": 20568 }, { "epoch": 2.2299436253252387, "grad_norm": 0.16508591175079346, "learning_rate": 2.5761642245756566e-05, "loss": 0.0066, "step": 20569 }, { "epoch": 2.2300520381613183, "grad_norm": 0.323503702878952, "learning_rate": 2.575801537791963e-05, "loss": 0.0168, "step": 20570 }, { "epoch": 2.2301604509973982, "grad_norm": 0.14474987983703613, "learning_rate": 2.5754388510082694e-05, "loss": 0.0053, "step": 20571 }, { "epoch": 2.2302688638334778, "grad_norm": 0.010287064127624035, "learning_rate": 2.575076164224576e-05, "loss": 0.0006, "step": 20572 }, { "epoch": 2.2303772766695578, "grad_norm": 0.08623089641332626, "learning_rate": 2.574713477440882e-05, "loss": 0.0037, "step": 20573 }, { "epoch": 2.2304856895056373, "grad_norm": 0.8984767198562622, "learning_rate": 2.5743507906571883e-05, "loss": 0.0267, "step": 20574 }, { "epoch": 2.2305941023417173, "grad_norm": 0.09418883919715881, "learning_rate": 2.5739881038734947e-05, "loss": 0.0043, "step": 20575 }, { "epoch": 2.2307025151777973, "grad_norm": 0.04362454265356064, "learning_rate": 2.573625417089801e-05, "loss": 0.0017, "step": 20576 }, { "epoch": 2.230810928013877, "grad_norm": 0.09809631109237671, "learning_rate": 2.573262730306108e-05, "loss": 0.0044, "step": 20577 }, { "epoch": 2.230919340849957, "grad_norm": 0.0958312526345253, "learning_rate": 2.5729000435224143e-05, "loss": 0.0053, "step": 20578 }, { "epoch": 2.2310277536860363, "grad_norm": 0.5990625023841858, "learning_rate": 2.5725373567387207e-05, "loss": 0.0362, "step": 20579 }, { "epoch": 2.2311361665221163, "grad_norm": 0.13520634174346924, "learning_rate": 2.572174669955027e-05, "loss": 0.0037, "step": 20580 }, { "epoch": 2.231244579358196, "grad_norm": 0.24183844029903412, "learning_rate": 2.5718119831713332e-05, "loss": 0.0128, "step": 20581 }, { "epoch": 2.231352992194276, "grad_norm": 0.09941611438989639, "learning_rate": 2.5714492963876396e-05, "loss": 0.0025, "step": 20582 }, { "epoch": 2.2314614050303554, "grad_norm": 0.33978956937789917, "learning_rate": 2.571086609603946e-05, "loss": 0.005, "step": 20583 }, { "epoch": 2.2315698178664354, "grad_norm": 0.13478277623653412, "learning_rate": 2.5707239228202524e-05, "loss": 0.0038, "step": 20584 }, { "epoch": 2.2316782307025154, "grad_norm": 0.019253868609666824, "learning_rate": 2.5703612360365592e-05, "loss": 0.0009, "step": 20585 }, { "epoch": 2.231786643538595, "grad_norm": 0.09631126374006271, "learning_rate": 2.5699985492528656e-05, "loss": 0.0019, "step": 20586 }, { "epoch": 2.231895056374675, "grad_norm": 0.011782714165747166, "learning_rate": 2.569635862469172e-05, "loss": 0.0004, "step": 20587 }, { "epoch": 2.2320034692107544, "grad_norm": 0.19771210849285126, "learning_rate": 2.569273175685478e-05, "loss": 0.0494, "step": 20588 }, { "epoch": 2.2321118820468344, "grad_norm": 0.27134931087493896, "learning_rate": 2.5689104889017845e-05, "loss": 0.0209, "step": 20589 }, { "epoch": 2.232220294882914, "grad_norm": 0.012644149363040924, "learning_rate": 2.568547802118091e-05, "loss": 0.0008, "step": 20590 }, { "epoch": 2.232328707718994, "grad_norm": 0.43245169520378113, "learning_rate": 2.5681851153343973e-05, "loss": 0.0159, "step": 20591 }, { "epoch": 2.232437120555074, "grad_norm": 0.10930324345827103, "learning_rate": 2.5678224285507034e-05, "loss": 0.004, "step": 20592 }, { "epoch": 2.2325455333911535, "grad_norm": 0.5819087624549866, "learning_rate": 2.5674597417670098e-05, "loss": 0.0108, "step": 20593 }, { "epoch": 2.2326539462272335, "grad_norm": 0.29140231013298035, "learning_rate": 2.567097054983317e-05, "loss": 0.0148, "step": 20594 }, { "epoch": 2.232762359063313, "grad_norm": 0.2989339232444763, "learning_rate": 2.566734368199623e-05, "loss": 0.0075, "step": 20595 }, { "epoch": 2.232870771899393, "grad_norm": 0.9007887840270996, "learning_rate": 2.5663716814159294e-05, "loss": 0.008, "step": 20596 }, { "epoch": 2.2329791847354725, "grad_norm": 0.11410082876682281, "learning_rate": 2.5660089946322358e-05, "loss": 0.0033, "step": 20597 }, { "epoch": 2.2330875975715525, "grad_norm": 0.05025266483426094, "learning_rate": 2.5656463078485422e-05, "loss": 0.0015, "step": 20598 }, { "epoch": 2.2331960104076325, "grad_norm": 0.13933102786540985, "learning_rate": 2.5652836210648483e-05, "loss": 0.0102, "step": 20599 }, { "epoch": 2.233304423243712, "grad_norm": 0.23189161717891693, "learning_rate": 2.5649209342811547e-05, "loss": 0.0055, "step": 20600 }, { "epoch": 2.233412836079792, "grad_norm": 1.651162028312683, "learning_rate": 2.564558247497461e-05, "loss": 0.0628, "step": 20601 }, { "epoch": 2.2335212489158716, "grad_norm": 0.04176893085241318, "learning_rate": 2.564195560713768e-05, "loss": 0.0013, "step": 20602 }, { "epoch": 2.2336296617519515, "grad_norm": 0.3473730981349945, "learning_rate": 2.5638328739300742e-05, "loss": 0.0177, "step": 20603 }, { "epoch": 2.233738074588031, "grad_norm": 0.28623253107070923, "learning_rate": 2.5634701871463807e-05, "loss": 0.0094, "step": 20604 }, { "epoch": 2.233846487424111, "grad_norm": 0.10508446395397186, "learning_rate": 2.563107500362687e-05, "loss": 0.0044, "step": 20605 }, { "epoch": 2.2339549002601906, "grad_norm": 0.6517322659492493, "learning_rate": 2.5627448135789935e-05, "loss": 0.017, "step": 20606 }, { "epoch": 2.2340633130962706, "grad_norm": 0.3252972364425659, "learning_rate": 2.5623821267952995e-05, "loss": 0.0193, "step": 20607 }, { "epoch": 2.2341717259323506, "grad_norm": 0.09636809676885605, "learning_rate": 2.562019440011606e-05, "loss": 0.0022, "step": 20608 }, { "epoch": 2.23428013876843, "grad_norm": 0.40637877583503723, "learning_rate": 2.5616567532279124e-05, "loss": 0.0056, "step": 20609 }, { "epoch": 2.23438855160451, "grad_norm": 0.26541805267333984, "learning_rate": 2.5612940664442188e-05, "loss": 0.0055, "step": 20610 }, { "epoch": 2.2344969644405897, "grad_norm": 0.8386604189872742, "learning_rate": 2.5609313796605255e-05, "loss": 0.0164, "step": 20611 }, { "epoch": 2.2346053772766696, "grad_norm": 0.3146897852420807, "learning_rate": 2.560568692876832e-05, "loss": 0.0108, "step": 20612 }, { "epoch": 2.234713790112749, "grad_norm": 1.378635287284851, "learning_rate": 2.5602060060931383e-05, "loss": 0.0395, "step": 20613 }, { "epoch": 2.234822202948829, "grad_norm": 0.43302375078201294, "learning_rate": 2.5598433193094444e-05, "loss": 0.0207, "step": 20614 }, { "epoch": 2.2349306157849087, "grad_norm": 0.7477960586547852, "learning_rate": 2.5594806325257508e-05, "loss": 0.0322, "step": 20615 }, { "epoch": 2.2350390286209887, "grad_norm": 0.4923234283924103, "learning_rate": 2.5591179457420572e-05, "loss": 0.0058, "step": 20616 }, { "epoch": 2.2351474414570687, "grad_norm": 0.1746530383825302, "learning_rate": 2.5587552589583637e-05, "loss": 0.0061, "step": 20617 }, { "epoch": 2.235255854293148, "grad_norm": 0.493795245885849, "learning_rate": 2.5583925721746697e-05, "loss": 0.0228, "step": 20618 }, { "epoch": 2.235364267129228, "grad_norm": 0.4481219947338104, "learning_rate": 2.558029885390976e-05, "loss": 0.0194, "step": 20619 }, { "epoch": 2.2354726799653077, "grad_norm": 0.09101750701665878, "learning_rate": 2.5576671986072832e-05, "loss": 0.0036, "step": 20620 }, { "epoch": 2.2355810928013877, "grad_norm": 1.9529320001602173, "learning_rate": 2.5573045118235893e-05, "loss": 0.0271, "step": 20621 }, { "epoch": 2.2356895056374677, "grad_norm": 0.15408556163311005, "learning_rate": 2.5569418250398957e-05, "loss": 0.0022, "step": 20622 }, { "epoch": 2.2357979184735473, "grad_norm": 0.4137675166130066, "learning_rate": 2.556579138256202e-05, "loss": 0.0123, "step": 20623 }, { "epoch": 2.2359063313096272, "grad_norm": 0.09441574662923813, "learning_rate": 2.5562164514725085e-05, "loss": 0.0032, "step": 20624 }, { "epoch": 2.236014744145707, "grad_norm": 0.5157192945480347, "learning_rate": 2.555853764688815e-05, "loss": 0.0434, "step": 20625 }, { "epoch": 2.2361231569817868, "grad_norm": 0.16177625954151154, "learning_rate": 2.555491077905121e-05, "loss": 0.0077, "step": 20626 }, { "epoch": 2.2362315698178663, "grad_norm": 0.04592723771929741, "learning_rate": 2.5551283911214274e-05, "loss": 0.0025, "step": 20627 }, { "epoch": 2.2363399826539463, "grad_norm": 0.14265914261341095, "learning_rate": 2.554765704337734e-05, "loss": 0.0052, "step": 20628 }, { "epoch": 2.236448395490026, "grad_norm": 0.0357060506939888, "learning_rate": 2.5544030175540406e-05, "loss": 0.002, "step": 20629 }, { "epoch": 2.236556808326106, "grad_norm": 0.14469008147716522, "learning_rate": 2.554040330770347e-05, "loss": 0.0029, "step": 20630 }, { "epoch": 2.236665221162186, "grad_norm": 0.3044574558734894, "learning_rate": 2.5536776439866534e-05, "loss": 0.0213, "step": 20631 }, { "epoch": 2.2367736339982653, "grad_norm": 0.21984539926052094, "learning_rate": 2.5533149572029598e-05, "loss": 0.0137, "step": 20632 }, { "epoch": 2.2368820468343453, "grad_norm": 0.2098478078842163, "learning_rate": 2.552952270419266e-05, "loss": 0.0084, "step": 20633 }, { "epoch": 2.236990459670425, "grad_norm": 0.06679418683052063, "learning_rate": 2.5525895836355723e-05, "loss": 0.0024, "step": 20634 }, { "epoch": 2.237098872506505, "grad_norm": 0.48413094878196716, "learning_rate": 2.5522268968518787e-05, "loss": 0.0269, "step": 20635 }, { "epoch": 2.2372072853425844, "grad_norm": 0.11844391375780106, "learning_rate": 2.551864210068185e-05, "loss": 0.0047, "step": 20636 }, { "epoch": 2.2373156981786644, "grad_norm": 0.16121675074100494, "learning_rate": 2.551501523284492e-05, "loss": 0.0055, "step": 20637 }, { "epoch": 2.237424111014744, "grad_norm": 0.4517497718334198, "learning_rate": 2.5511388365007983e-05, "loss": 0.0151, "step": 20638 }, { "epoch": 2.237532523850824, "grad_norm": 0.10118542611598969, "learning_rate": 2.5507761497171047e-05, "loss": 0.0014, "step": 20639 }, { "epoch": 2.237640936686904, "grad_norm": 0.28228408098220825, "learning_rate": 2.5504134629334108e-05, "loss": 0.0146, "step": 20640 }, { "epoch": 2.2377493495229834, "grad_norm": 0.033889371901750565, "learning_rate": 2.550050776149717e-05, "loss": 0.0018, "step": 20641 }, { "epoch": 2.2378577623590634, "grad_norm": 0.08121437579393387, "learning_rate": 2.5496880893660236e-05, "loss": 0.0023, "step": 20642 }, { "epoch": 2.237966175195143, "grad_norm": 0.5973477363586426, "learning_rate": 2.54932540258233e-05, "loss": 0.004, "step": 20643 }, { "epoch": 2.238074588031223, "grad_norm": 0.11249377578496933, "learning_rate": 2.548962715798636e-05, "loss": 0.003, "step": 20644 }, { "epoch": 2.2381830008673025, "grad_norm": 0.35074958205223083, "learning_rate": 2.5486000290149425e-05, "loss": 0.0037, "step": 20645 }, { "epoch": 2.2382914137033825, "grad_norm": 0.1686282902956009, "learning_rate": 2.5482373422312496e-05, "loss": 0.0057, "step": 20646 }, { "epoch": 2.2383998265394625, "grad_norm": 0.2620636522769928, "learning_rate": 2.5478746554475556e-05, "loss": 0.0118, "step": 20647 }, { "epoch": 2.238508239375542, "grad_norm": 0.49356067180633545, "learning_rate": 2.547511968663862e-05, "loss": 0.0201, "step": 20648 }, { "epoch": 2.238616652211622, "grad_norm": 0.32719331979751587, "learning_rate": 2.5471492818801684e-05, "loss": 0.0086, "step": 20649 }, { "epoch": 2.2387250650477015, "grad_norm": 0.0616631843149662, "learning_rate": 2.546786595096475e-05, "loss": 0.002, "step": 20650 }, { "epoch": 2.2388334778837815, "grad_norm": 0.329374760389328, "learning_rate": 2.5464239083127813e-05, "loss": 0.0136, "step": 20651 }, { "epoch": 2.238941890719861, "grad_norm": 0.09656049311161041, "learning_rate": 2.5460612215290873e-05, "loss": 0.0079, "step": 20652 }, { "epoch": 2.239050303555941, "grad_norm": 0.28986379504203796, "learning_rate": 2.5456985347453938e-05, "loss": 0.0086, "step": 20653 }, { "epoch": 2.239158716392021, "grad_norm": 0.035513464361429214, "learning_rate": 2.5453358479617005e-05, "loss": 0.001, "step": 20654 }, { "epoch": 2.2392671292281006, "grad_norm": 0.07227522879838943, "learning_rate": 2.544973161178007e-05, "loss": 0.0033, "step": 20655 }, { "epoch": 2.2393755420641805, "grad_norm": 0.3119616210460663, "learning_rate": 2.5446104743943133e-05, "loss": 0.0372, "step": 20656 }, { "epoch": 2.23948395490026, "grad_norm": 0.005665392614901066, "learning_rate": 2.5442477876106197e-05, "loss": 0.0003, "step": 20657 }, { "epoch": 2.23959236773634, "grad_norm": 0.3291650414466858, "learning_rate": 2.543885100826926e-05, "loss": 0.0142, "step": 20658 }, { "epoch": 2.2397007805724196, "grad_norm": 0.4900696575641632, "learning_rate": 2.5435224140432322e-05, "loss": 0.0325, "step": 20659 }, { "epoch": 2.2398091934084996, "grad_norm": 0.004256607964634895, "learning_rate": 2.5431597272595386e-05, "loss": 0.0002, "step": 20660 }, { "epoch": 2.239917606244579, "grad_norm": 0.06512494385242462, "learning_rate": 2.542797040475845e-05, "loss": 0.002, "step": 20661 }, { "epoch": 2.240026019080659, "grad_norm": 0.36073794960975647, "learning_rate": 2.5424343536921514e-05, "loss": 0.0112, "step": 20662 }, { "epoch": 2.240134431916739, "grad_norm": 0.019349241629242897, "learning_rate": 2.5420716669084582e-05, "loss": 0.0007, "step": 20663 }, { "epoch": 2.2402428447528187, "grad_norm": 0.20332837104797363, "learning_rate": 2.5417089801247646e-05, "loss": 0.0107, "step": 20664 }, { "epoch": 2.2403512575888986, "grad_norm": 0.9577022790908813, "learning_rate": 2.541346293341071e-05, "loss": 0.0306, "step": 20665 }, { "epoch": 2.240459670424978, "grad_norm": 0.4676945209503174, "learning_rate": 2.540983606557377e-05, "loss": 0.0367, "step": 20666 }, { "epoch": 2.240568083261058, "grad_norm": 0.16091258823871613, "learning_rate": 2.5406209197736835e-05, "loss": 0.014, "step": 20667 }, { "epoch": 2.2406764960971377, "grad_norm": 2.09100079536438, "learning_rate": 2.54025823298999e-05, "loss": 0.059, "step": 20668 }, { "epoch": 2.2407849089332177, "grad_norm": 0.012019691057503223, "learning_rate": 2.5398955462062963e-05, "loss": 0.0006, "step": 20669 }, { "epoch": 2.2408933217692977, "grad_norm": 0.08496547490358353, "learning_rate": 2.5395328594226024e-05, "loss": 0.0018, "step": 20670 }, { "epoch": 2.241001734605377, "grad_norm": 0.048525433987379074, "learning_rate": 2.5391701726389088e-05, "loss": 0.0022, "step": 20671 }, { "epoch": 2.241110147441457, "grad_norm": 0.008726806379854679, "learning_rate": 2.538807485855216e-05, "loss": 0.0004, "step": 20672 }, { "epoch": 2.2412185602775367, "grad_norm": 0.14458267390727997, "learning_rate": 2.538444799071522e-05, "loss": 0.0018, "step": 20673 }, { "epoch": 2.2413269731136167, "grad_norm": 0.2450709342956543, "learning_rate": 2.5380821122878284e-05, "loss": 0.0224, "step": 20674 }, { "epoch": 2.2414353859496963, "grad_norm": 0.40354013442993164, "learning_rate": 2.5377194255041348e-05, "loss": 0.0439, "step": 20675 }, { "epoch": 2.2415437987857763, "grad_norm": 0.3240302801132202, "learning_rate": 2.5373567387204412e-05, "loss": 0.0123, "step": 20676 }, { "epoch": 2.2416522116218562, "grad_norm": 0.31950095295906067, "learning_rate": 2.5369940519367476e-05, "loss": 0.008, "step": 20677 }, { "epoch": 2.241760624457936, "grad_norm": 0.53046715259552, "learning_rate": 2.5366313651530537e-05, "loss": 0.0173, "step": 20678 }, { "epoch": 2.2418690372940158, "grad_norm": 0.44943660497665405, "learning_rate": 2.53626867836936e-05, "loss": 0.0274, "step": 20679 }, { "epoch": 2.2419774501300953, "grad_norm": 0.020980309695005417, "learning_rate": 2.535905991585667e-05, "loss": 0.0004, "step": 20680 }, { "epoch": 2.2420858629661753, "grad_norm": 0.17073142528533936, "learning_rate": 2.5355433048019732e-05, "loss": 0.0076, "step": 20681 }, { "epoch": 2.242194275802255, "grad_norm": 0.3525930941104889, "learning_rate": 2.5351806180182797e-05, "loss": 0.0373, "step": 20682 }, { "epoch": 2.242302688638335, "grad_norm": 0.30971524119377136, "learning_rate": 2.534817931234586e-05, "loss": 0.0272, "step": 20683 }, { "epoch": 2.2424111014744144, "grad_norm": 0.19433391094207764, "learning_rate": 2.5344552444508925e-05, "loss": 0.018, "step": 20684 }, { "epoch": 2.2425195143104943, "grad_norm": 0.4261694848537445, "learning_rate": 2.5340925576671985e-05, "loss": 0.017, "step": 20685 }, { "epoch": 2.2426279271465743, "grad_norm": 0.5104659795761108, "learning_rate": 2.533729870883505e-05, "loss": 0.0237, "step": 20686 }, { "epoch": 2.242736339982654, "grad_norm": 0.6495062708854675, "learning_rate": 2.5333671840998114e-05, "loss": 0.0159, "step": 20687 }, { "epoch": 2.242844752818734, "grad_norm": 0.08941724896430969, "learning_rate": 2.5330044973161178e-05, "loss": 0.0027, "step": 20688 }, { "epoch": 2.2429531656548134, "grad_norm": 0.14497482776641846, "learning_rate": 2.5326418105324245e-05, "loss": 0.0042, "step": 20689 }, { "epoch": 2.2430615784908934, "grad_norm": 0.24714499711990356, "learning_rate": 2.532279123748731e-05, "loss": 0.0099, "step": 20690 }, { "epoch": 2.243169991326973, "grad_norm": 0.9376620650291443, "learning_rate": 2.5319164369650374e-05, "loss": 0.0093, "step": 20691 }, { "epoch": 2.243278404163053, "grad_norm": 0.2699986696243286, "learning_rate": 2.5315537501813434e-05, "loss": 0.0085, "step": 20692 }, { "epoch": 2.243386816999133, "grad_norm": 0.058790866285562515, "learning_rate": 2.53119106339765e-05, "loss": 0.0016, "step": 20693 }, { "epoch": 2.2434952298352124, "grad_norm": 1.1300915479660034, "learning_rate": 2.5308283766139562e-05, "loss": 0.0515, "step": 20694 }, { "epoch": 2.2436036426712924, "grad_norm": 0.23694059252738953, "learning_rate": 2.5304656898302627e-05, "loss": 0.0272, "step": 20695 }, { "epoch": 2.243712055507372, "grad_norm": 0.6037888526916504, "learning_rate": 2.5301030030465687e-05, "loss": 0.0448, "step": 20696 }, { "epoch": 2.243820468343452, "grad_norm": 0.2133927196264267, "learning_rate": 2.5297403162628758e-05, "loss": 0.0144, "step": 20697 }, { "epoch": 2.2439288811795315, "grad_norm": 0.027397451922297478, "learning_rate": 2.5293776294791822e-05, "loss": 0.0015, "step": 20698 }, { "epoch": 2.2440372940156115, "grad_norm": 0.39306050539016724, "learning_rate": 2.5290149426954883e-05, "loss": 0.0192, "step": 20699 }, { "epoch": 2.2441457068516915, "grad_norm": 0.09248573333024979, "learning_rate": 2.5286522559117947e-05, "loss": 0.0031, "step": 20700 }, { "epoch": 2.244254119687771, "grad_norm": 0.2329544574022293, "learning_rate": 2.528289569128101e-05, "loss": 0.022, "step": 20701 }, { "epoch": 2.244362532523851, "grad_norm": 0.6485735177993774, "learning_rate": 2.5279268823444075e-05, "loss": 0.0136, "step": 20702 }, { "epoch": 2.2444709453599305, "grad_norm": 0.09658420830965042, "learning_rate": 2.527564195560714e-05, "loss": 0.0064, "step": 20703 }, { "epoch": 2.2445793581960105, "grad_norm": 0.0721011608839035, "learning_rate": 2.52720150877702e-05, "loss": 0.0023, "step": 20704 }, { "epoch": 2.24468777103209, "grad_norm": 0.3884609639644623, "learning_rate": 2.5268388219933264e-05, "loss": 0.0175, "step": 20705 }, { "epoch": 2.24479618386817, "grad_norm": 0.2765452265739441, "learning_rate": 2.526476135209633e-05, "loss": 0.009, "step": 20706 }, { "epoch": 2.2449045967042496, "grad_norm": 0.4666326344013214, "learning_rate": 2.5261134484259396e-05, "loss": 0.0153, "step": 20707 }, { "epoch": 2.2450130095403296, "grad_norm": 0.2374478429555893, "learning_rate": 2.525750761642246e-05, "loss": 0.0095, "step": 20708 }, { "epoch": 2.2451214223764095, "grad_norm": 0.10452738404273987, "learning_rate": 2.5253880748585524e-05, "loss": 0.0042, "step": 20709 }, { "epoch": 2.245229835212489, "grad_norm": 0.3944002389907837, "learning_rate": 2.5250253880748588e-05, "loss": 0.0118, "step": 20710 }, { "epoch": 2.245338248048569, "grad_norm": 0.07701262086629868, "learning_rate": 2.524662701291165e-05, "loss": 0.0023, "step": 20711 }, { "epoch": 2.2454466608846486, "grad_norm": 0.03997766226530075, "learning_rate": 2.5243000145074713e-05, "loss": 0.0017, "step": 20712 }, { "epoch": 2.2455550737207286, "grad_norm": 0.00628134747967124, "learning_rate": 2.5239373277237777e-05, "loss": 0.0004, "step": 20713 }, { "epoch": 2.245663486556808, "grad_norm": 0.030198944732546806, "learning_rate": 2.523574640940084e-05, "loss": 0.0018, "step": 20714 }, { "epoch": 2.245771899392888, "grad_norm": 0.06089436635375023, "learning_rate": 2.523211954156391e-05, "loss": 0.0024, "step": 20715 }, { "epoch": 2.245880312228968, "grad_norm": 0.10717763751745224, "learning_rate": 2.5228492673726973e-05, "loss": 0.0018, "step": 20716 }, { "epoch": 2.2459887250650477, "grad_norm": 0.8557115197181702, "learning_rate": 2.5224865805890037e-05, "loss": 0.0243, "step": 20717 }, { "epoch": 2.2460971379011276, "grad_norm": 0.6019756197929382, "learning_rate": 2.5221238938053098e-05, "loss": 0.01, "step": 20718 }, { "epoch": 2.246205550737207, "grad_norm": 0.16232594847679138, "learning_rate": 2.521761207021616e-05, "loss": 0.0136, "step": 20719 }, { "epoch": 2.246313963573287, "grad_norm": 0.6511714458465576, "learning_rate": 2.5213985202379226e-05, "loss": 0.0356, "step": 20720 }, { "epoch": 2.2464223764093667, "grad_norm": 0.20270690321922302, "learning_rate": 2.521035833454229e-05, "loss": 0.0055, "step": 20721 }, { "epoch": 2.2465307892454467, "grad_norm": 0.19417399168014526, "learning_rate": 2.520673146670535e-05, "loss": 0.009, "step": 20722 }, { "epoch": 2.2466392020815267, "grad_norm": 0.10552126169204712, "learning_rate": 2.520310459886842e-05, "loss": 0.0026, "step": 20723 }, { "epoch": 2.246747614917606, "grad_norm": 0.28432831168174744, "learning_rate": 2.5199477731031486e-05, "loss": 0.0118, "step": 20724 }, { "epoch": 2.246856027753686, "grad_norm": 0.24909813702106476, "learning_rate": 2.5195850863194546e-05, "loss": 0.0095, "step": 20725 }, { "epoch": 2.2469644405897657, "grad_norm": 0.08161352574825287, "learning_rate": 2.519222399535761e-05, "loss": 0.003, "step": 20726 }, { "epoch": 2.2470728534258457, "grad_norm": 0.43759045004844666, "learning_rate": 2.5188597127520675e-05, "loss": 0.0606, "step": 20727 }, { "epoch": 2.2471812662619253, "grad_norm": 0.18864116072654724, "learning_rate": 2.518497025968374e-05, "loss": 0.0055, "step": 20728 }, { "epoch": 2.2472896790980053, "grad_norm": 0.22403503954410553, "learning_rate": 2.5181343391846803e-05, "loss": 0.0091, "step": 20729 }, { "epoch": 2.247398091934085, "grad_norm": 0.5348201990127563, "learning_rate": 2.5177716524009863e-05, "loss": 0.0221, "step": 20730 }, { "epoch": 2.247506504770165, "grad_norm": 0.38158750534057617, "learning_rate": 2.5174089656172928e-05, "loss": 0.0111, "step": 20731 }, { "epoch": 2.2476149176062448, "grad_norm": 0.8222165703773499, "learning_rate": 2.5170462788335995e-05, "loss": 0.0293, "step": 20732 }, { "epoch": 2.2477233304423243, "grad_norm": 0.28323590755462646, "learning_rate": 2.516683592049906e-05, "loss": 0.0123, "step": 20733 }, { "epoch": 2.2478317432784043, "grad_norm": 0.2552342712879181, "learning_rate": 2.5163209052662123e-05, "loss": 0.0123, "step": 20734 }, { "epoch": 2.247940156114484, "grad_norm": 0.19634440541267395, "learning_rate": 2.5159582184825187e-05, "loss": 0.0106, "step": 20735 }, { "epoch": 2.248048568950564, "grad_norm": 0.515792965888977, "learning_rate": 2.515595531698825e-05, "loss": 0.0627, "step": 20736 }, { "epoch": 2.2481569817866434, "grad_norm": 0.12921246886253357, "learning_rate": 2.5152328449151312e-05, "loss": 0.0055, "step": 20737 }, { "epoch": 2.2482653946227233, "grad_norm": 0.4799380898475647, "learning_rate": 2.5148701581314376e-05, "loss": 0.0181, "step": 20738 }, { "epoch": 2.248373807458803, "grad_norm": 0.026493769139051437, "learning_rate": 2.514507471347744e-05, "loss": 0.0013, "step": 20739 }, { "epoch": 2.248482220294883, "grad_norm": 0.414421021938324, "learning_rate": 2.5141447845640504e-05, "loss": 0.0083, "step": 20740 }, { "epoch": 2.248590633130963, "grad_norm": 0.4320922791957855, "learning_rate": 2.5137820977803572e-05, "loss": 0.0258, "step": 20741 }, { "epoch": 2.2486990459670424, "grad_norm": 0.2015598714351654, "learning_rate": 2.5134194109966636e-05, "loss": 0.0069, "step": 20742 }, { "epoch": 2.2488074588031224, "grad_norm": 0.09981398284435272, "learning_rate": 2.51305672421297e-05, "loss": 0.0047, "step": 20743 }, { "epoch": 2.248915871639202, "grad_norm": 0.18461601436138153, "learning_rate": 2.512694037429276e-05, "loss": 0.0089, "step": 20744 }, { "epoch": 2.249024284475282, "grad_norm": 0.13940612971782684, "learning_rate": 2.5123313506455825e-05, "loss": 0.0035, "step": 20745 }, { "epoch": 2.249132697311362, "grad_norm": 0.6174472570419312, "learning_rate": 2.511968663861889e-05, "loss": 0.0163, "step": 20746 }, { "epoch": 2.2492411101474414, "grad_norm": 0.32039979100227356, "learning_rate": 2.5116059770781953e-05, "loss": 0.0097, "step": 20747 }, { "epoch": 2.2493495229835214, "grad_norm": 0.4086063504219055, "learning_rate": 2.5112432902945014e-05, "loss": 0.011, "step": 20748 }, { "epoch": 2.249457935819601, "grad_norm": 0.6688022613525391, "learning_rate": 2.5108806035108085e-05, "loss": 0.0123, "step": 20749 }, { "epoch": 2.249566348655681, "grad_norm": 0.22819627821445465, "learning_rate": 2.510517916727115e-05, "loss": 0.0184, "step": 20750 }, { "epoch": 2.2496747614917605, "grad_norm": 0.17440924048423767, "learning_rate": 2.510155229943421e-05, "loss": 0.0075, "step": 20751 }, { "epoch": 2.2497831743278405, "grad_norm": 0.07733175158500671, "learning_rate": 2.5097925431597274e-05, "loss": 0.0023, "step": 20752 }, { "epoch": 2.24989158716392, "grad_norm": 0.036850396543741226, "learning_rate": 2.5094298563760338e-05, "loss": 0.0012, "step": 20753 }, { "epoch": 2.25, "grad_norm": 0.522826075553894, "learning_rate": 2.5090671695923402e-05, "loss": 0.0377, "step": 20754 }, { "epoch": 2.25010841283608, "grad_norm": 0.1910332292318344, "learning_rate": 2.5087044828086466e-05, "loss": 0.0278, "step": 20755 }, { "epoch": 2.2502168256721595, "grad_norm": 0.22771622240543365, "learning_rate": 2.5083417960249527e-05, "loss": 0.005, "step": 20756 }, { "epoch": 2.2503252385082395, "grad_norm": 0.11088798940181732, "learning_rate": 2.507979109241259e-05, "loss": 0.0033, "step": 20757 }, { "epoch": 2.250433651344319, "grad_norm": 0.31214776635169983, "learning_rate": 2.507616422457566e-05, "loss": 0.0024, "step": 20758 }, { "epoch": 2.250542064180399, "grad_norm": 0.26898935437202454, "learning_rate": 2.5072537356738722e-05, "loss": 0.0217, "step": 20759 }, { "epoch": 2.2506504770164786, "grad_norm": 0.12064699083566666, "learning_rate": 2.5068910488901787e-05, "loss": 0.0041, "step": 20760 }, { "epoch": 2.2507588898525586, "grad_norm": 0.16970694065093994, "learning_rate": 2.506528362106485e-05, "loss": 0.0098, "step": 20761 }, { "epoch": 2.250867302688638, "grad_norm": 0.14307977259159088, "learning_rate": 2.5061656753227915e-05, "loss": 0.0115, "step": 20762 }, { "epoch": 2.250975715524718, "grad_norm": 0.5340742468833923, "learning_rate": 2.5058029885390976e-05, "loss": 0.0329, "step": 20763 }, { "epoch": 2.251084128360798, "grad_norm": 0.35319823026657104, "learning_rate": 2.505440301755404e-05, "loss": 0.0092, "step": 20764 }, { "epoch": 2.2511925411968776, "grad_norm": 0.13176220655441284, "learning_rate": 2.5050776149717104e-05, "loss": 0.0056, "step": 20765 }, { "epoch": 2.2513009540329576, "grad_norm": 0.05831707641482353, "learning_rate": 2.5047149281880168e-05, "loss": 0.0036, "step": 20766 }, { "epoch": 2.251409366869037, "grad_norm": 0.08795486390590668, "learning_rate": 2.5043522414043235e-05, "loss": 0.0045, "step": 20767 }, { "epoch": 2.251517779705117, "grad_norm": 0.18181955814361572, "learning_rate": 2.50398955462063e-05, "loss": 0.0459, "step": 20768 }, { "epoch": 2.251626192541197, "grad_norm": 0.1028175950050354, "learning_rate": 2.5036268678369364e-05, "loss": 0.0033, "step": 20769 }, { "epoch": 2.2517346053772767, "grad_norm": 0.07686345279216766, "learning_rate": 2.5032641810532424e-05, "loss": 0.001, "step": 20770 }, { "epoch": 2.2518430182133566, "grad_norm": 0.8407502174377441, "learning_rate": 2.502901494269549e-05, "loss": 0.0165, "step": 20771 }, { "epoch": 2.251951431049436, "grad_norm": 0.1808386594057083, "learning_rate": 2.5025388074858552e-05, "loss": 0.0121, "step": 20772 }, { "epoch": 2.252059843885516, "grad_norm": 1.980563759803772, "learning_rate": 2.5021761207021617e-05, "loss": 0.0353, "step": 20773 }, { "epoch": 2.2521682567215957, "grad_norm": 0.2822168469429016, "learning_rate": 2.5018134339184677e-05, "loss": 0.0036, "step": 20774 }, { "epoch": 2.2522766695576757, "grad_norm": 0.04781848192214966, "learning_rate": 2.5014507471347748e-05, "loss": 0.0015, "step": 20775 }, { "epoch": 2.2523850823937552, "grad_norm": 0.22887735068798065, "learning_rate": 2.5010880603510812e-05, "loss": 0.0078, "step": 20776 }, { "epoch": 2.252493495229835, "grad_norm": 0.019810875877738, "learning_rate": 2.5007253735673873e-05, "loss": 0.0007, "step": 20777 }, { "epoch": 2.252601908065915, "grad_norm": 0.43622827529907227, "learning_rate": 2.5003626867836937e-05, "loss": 0.033, "step": 20778 }, { "epoch": 2.2527103209019947, "grad_norm": 0.2100599855184555, "learning_rate": 2.5e-05, "loss": 0.047, "step": 20779 }, { "epoch": 2.2528187337380747, "grad_norm": 0.1348097324371338, "learning_rate": 2.4996373132163065e-05, "loss": 0.0051, "step": 20780 }, { "epoch": 2.2529271465741543, "grad_norm": 0.02814588136970997, "learning_rate": 2.499274626432613e-05, "loss": 0.0011, "step": 20781 }, { "epoch": 2.2530355594102343, "grad_norm": 0.02566603198647499, "learning_rate": 2.4989119396489194e-05, "loss": 0.0011, "step": 20782 }, { "epoch": 2.253143972246314, "grad_norm": 0.24259622395038605, "learning_rate": 2.4985492528652258e-05, "loss": 0.0073, "step": 20783 }, { "epoch": 2.253252385082394, "grad_norm": 0.888046145439148, "learning_rate": 2.4981865660815322e-05, "loss": 0.0096, "step": 20784 }, { "epoch": 2.2533607979184733, "grad_norm": 0.23939365148544312, "learning_rate": 2.4978238792978382e-05, "loss": 0.0075, "step": 20785 }, { "epoch": 2.2534692107545533, "grad_norm": 0.09696883708238602, "learning_rate": 2.497461192514145e-05, "loss": 0.001, "step": 20786 }, { "epoch": 2.2535776235906333, "grad_norm": 0.0896734818816185, "learning_rate": 2.4970985057304514e-05, "loss": 0.0024, "step": 20787 }, { "epoch": 2.253686036426713, "grad_norm": 0.02126256376504898, "learning_rate": 2.4967358189467578e-05, "loss": 0.0006, "step": 20788 }, { "epoch": 2.253794449262793, "grad_norm": 0.08710619062185287, "learning_rate": 2.496373132163064e-05, "loss": 0.0014, "step": 20789 }, { "epoch": 2.2539028620988724, "grad_norm": 0.2974337935447693, "learning_rate": 2.4960104453793706e-05, "loss": 0.0182, "step": 20790 }, { "epoch": 2.2540112749349523, "grad_norm": 0.556472659111023, "learning_rate": 2.495647758595677e-05, "loss": 0.017, "step": 20791 }, { "epoch": 2.2541196877710323, "grad_norm": 0.21358317136764526, "learning_rate": 2.4952850718119835e-05, "loss": 0.0063, "step": 20792 }, { "epoch": 2.254228100607112, "grad_norm": 0.2106279581785202, "learning_rate": 2.4949223850282895e-05, "loss": 0.0072, "step": 20793 }, { "epoch": 2.254336513443192, "grad_norm": 0.17948289215564728, "learning_rate": 2.494559698244596e-05, "loss": 0.0067, "step": 20794 }, { "epoch": 2.2544449262792714, "grad_norm": 0.45715516805648804, "learning_rate": 2.4941970114609027e-05, "loss": 0.0152, "step": 20795 }, { "epoch": 2.2545533391153514, "grad_norm": 0.009120122529566288, "learning_rate": 2.4938343246772088e-05, "loss": 0.0003, "step": 20796 }, { "epoch": 2.254661751951431, "grad_norm": 0.5137813687324524, "learning_rate": 2.4934716378935152e-05, "loss": 0.0221, "step": 20797 }, { "epoch": 2.254770164787511, "grad_norm": 0.039515309035778046, "learning_rate": 2.4931089511098216e-05, "loss": 0.0014, "step": 20798 }, { "epoch": 2.2548785776235905, "grad_norm": 0.33209455013275146, "learning_rate": 2.4927462643261283e-05, "loss": 0.0051, "step": 20799 }, { "epoch": 2.2549869904596704, "grad_norm": 0.33651629090309143, "learning_rate": 2.4923835775424344e-05, "loss": 0.0152, "step": 20800 }, { "epoch": 2.2550954032957504, "grad_norm": 0.45494577288627625, "learning_rate": 2.4920208907587408e-05, "loss": 0.0338, "step": 20801 }, { "epoch": 2.25520381613183, "grad_norm": 0.2717268168926239, "learning_rate": 2.4916582039750472e-05, "loss": 0.0355, "step": 20802 }, { "epoch": 2.25531222896791, "grad_norm": 0.15106454491615295, "learning_rate": 2.4912955171913536e-05, "loss": 0.0069, "step": 20803 }, { "epoch": 2.2554206418039895, "grad_norm": 0.2491343766450882, "learning_rate": 2.49093283040766e-05, "loss": 0.0084, "step": 20804 }, { "epoch": 2.2555290546400695, "grad_norm": 0.3176606595516205, "learning_rate": 2.4905701436239665e-05, "loss": 0.0265, "step": 20805 }, { "epoch": 2.255637467476149, "grad_norm": 0.24783194065093994, "learning_rate": 2.490207456840273e-05, "loss": 0.0041, "step": 20806 }, { "epoch": 2.255745880312229, "grad_norm": 0.5521543025970459, "learning_rate": 2.4898447700565793e-05, "loss": 0.0369, "step": 20807 }, { "epoch": 2.2558542931483085, "grad_norm": 1.026973843574524, "learning_rate": 2.4894820832728857e-05, "loss": 0.0113, "step": 20808 }, { "epoch": 2.2559627059843885, "grad_norm": 0.4774925112724304, "learning_rate": 2.489119396489192e-05, "loss": 0.0115, "step": 20809 }, { "epoch": 2.2560711188204685, "grad_norm": 0.08713331818580627, "learning_rate": 2.4887567097054985e-05, "loss": 0.0041, "step": 20810 }, { "epoch": 2.256179531656548, "grad_norm": 0.021268213167786598, "learning_rate": 2.4883940229218046e-05, "loss": 0.0011, "step": 20811 }, { "epoch": 2.256287944492628, "grad_norm": 0.10787219554185867, "learning_rate": 2.4880313361381113e-05, "loss": 0.0048, "step": 20812 }, { "epoch": 2.2563963573287076, "grad_norm": 0.23874206840991974, "learning_rate": 2.4876686493544177e-05, "loss": 0.0093, "step": 20813 }, { "epoch": 2.2565047701647876, "grad_norm": 0.5041338801383972, "learning_rate": 2.487305962570724e-05, "loss": 0.0234, "step": 20814 }, { "epoch": 2.2566131830008676, "grad_norm": 0.02721540443599224, "learning_rate": 2.4869432757870302e-05, "loss": 0.0011, "step": 20815 }, { "epoch": 2.256721595836947, "grad_norm": 0.11136846989393234, "learning_rate": 2.486580589003337e-05, "loss": 0.0066, "step": 20816 }, { "epoch": 2.256830008673027, "grad_norm": 0.39618703722953796, "learning_rate": 2.4862179022196434e-05, "loss": 0.0312, "step": 20817 }, { "epoch": 2.2569384215091066, "grad_norm": 0.2418733686208725, "learning_rate": 2.4858552154359498e-05, "loss": 0.0105, "step": 20818 }, { "epoch": 2.2570468343451866, "grad_norm": 0.0322749949991703, "learning_rate": 2.485492528652256e-05, "loss": 0.0012, "step": 20819 }, { "epoch": 2.257155247181266, "grad_norm": 0.12087931483983994, "learning_rate": 2.4851298418685623e-05, "loss": 0.0089, "step": 20820 }, { "epoch": 2.257263660017346, "grad_norm": 0.16236524283885956, "learning_rate": 2.484767155084869e-05, "loss": 0.0138, "step": 20821 }, { "epoch": 2.2573720728534257, "grad_norm": 0.7203881144523621, "learning_rate": 2.484404468301175e-05, "loss": 0.0442, "step": 20822 }, { "epoch": 2.2574804856895057, "grad_norm": 0.3058406114578247, "learning_rate": 2.4840417815174815e-05, "loss": 0.0151, "step": 20823 }, { "epoch": 2.2575888985255856, "grad_norm": 0.23371632397174835, "learning_rate": 2.483679094733788e-05, "loss": 0.006, "step": 20824 }, { "epoch": 2.257697311361665, "grad_norm": 0.4294237792491913, "learning_rate": 2.4833164079500947e-05, "loss": 0.0299, "step": 20825 }, { "epoch": 2.257805724197745, "grad_norm": 0.28592661023139954, "learning_rate": 2.4829537211664007e-05, "loss": 0.0085, "step": 20826 }, { "epoch": 2.2579141370338247, "grad_norm": 0.02304144576191902, "learning_rate": 2.482591034382707e-05, "loss": 0.0006, "step": 20827 }, { "epoch": 2.2580225498699047, "grad_norm": 0.44675755500793457, "learning_rate": 2.4822283475990136e-05, "loss": 0.0105, "step": 20828 }, { "epoch": 2.2581309627059842, "grad_norm": 0.4249299466609955, "learning_rate": 2.48186566081532e-05, "loss": 0.0136, "step": 20829 }, { "epoch": 2.258239375542064, "grad_norm": 0.34917065501213074, "learning_rate": 2.4815029740316264e-05, "loss": 0.0198, "step": 20830 }, { "epoch": 2.2583477883781438, "grad_norm": 0.4899526834487915, "learning_rate": 2.4811402872479328e-05, "loss": 0.0212, "step": 20831 }, { "epoch": 2.2584562012142237, "grad_norm": 0.23859116435050964, "learning_rate": 2.4807776004642392e-05, "loss": 0.0158, "step": 20832 }, { "epoch": 2.2585646140503037, "grad_norm": 0.26164528727531433, "learning_rate": 2.4804149136805456e-05, "loss": 0.0195, "step": 20833 }, { "epoch": 2.2586730268863833, "grad_norm": 0.24475227296352386, "learning_rate": 2.480052226896852e-05, "loss": 0.0211, "step": 20834 }, { "epoch": 2.2587814397224633, "grad_norm": 0.0463869646191597, "learning_rate": 2.4796895401131584e-05, "loss": 0.0019, "step": 20835 }, { "epoch": 2.258889852558543, "grad_norm": 0.03886668384075165, "learning_rate": 2.479326853329465e-05, "loss": 0.0028, "step": 20836 }, { "epoch": 2.258998265394623, "grad_norm": 0.7681169509887695, "learning_rate": 2.478964166545771e-05, "loss": 0.0212, "step": 20837 }, { "epoch": 2.2591066782307023, "grad_norm": 0.02718483842909336, "learning_rate": 2.4786014797620777e-05, "loss": 0.0009, "step": 20838 }, { "epoch": 2.2592150910667823, "grad_norm": 0.6610274910926819, "learning_rate": 2.478238792978384e-05, "loss": 0.0267, "step": 20839 }, { "epoch": 2.259323503902862, "grad_norm": 0.19675800204277039, "learning_rate": 2.4778761061946905e-05, "loss": 0.0064, "step": 20840 }, { "epoch": 2.259431916738942, "grad_norm": 0.4172794818878174, "learning_rate": 2.4775134194109966e-05, "loss": 0.0088, "step": 20841 }, { "epoch": 2.259540329575022, "grad_norm": 0.18575870990753174, "learning_rate": 2.4771507326273033e-05, "loss": 0.0061, "step": 20842 }, { "epoch": 2.2596487424111014, "grad_norm": 0.01639615371823311, "learning_rate": 2.4767880458436097e-05, "loss": 0.0004, "step": 20843 }, { "epoch": 2.2597571552471813, "grad_norm": 0.11647704988718033, "learning_rate": 2.476425359059916e-05, "loss": 0.0079, "step": 20844 }, { "epoch": 2.259865568083261, "grad_norm": 0.40530532598495483, "learning_rate": 2.4760626722762222e-05, "loss": 0.0185, "step": 20845 }, { "epoch": 2.259973980919341, "grad_norm": 0.3477421700954437, "learning_rate": 2.475699985492529e-05, "loss": 0.0214, "step": 20846 }, { "epoch": 2.260082393755421, "grad_norm": 0.07050589472055435, "learning_rate": 2.4753372987088354e-05, "loss": 0.0037, "step": 20847 }, { "epoch": 2.2601908065915004, "grad_norm": 0.0452730730175972, "learning_rate": 2.4749746119251414e-05, "loss": 0.0015, "step": 20848 }, { "epoch": 2.2602992194275804, "grad_norm": 0.732265830039978, "learning_rate": 2.474611925141448e-05, "loss": 0.0371, "step": 20849 }, { "epoch": 2.26040763226366, "grad_norm": 0.03365051746368408, "learning_rate": 2.4742492383577542e-05, "loss": 0.0011, "step": 20850 }, { "epoch": 2.26051604509974, "grad_norm": 0.1332390010356903, "learning_rate": 2.473886551574061e-05, "loss": 0.0053, "step": 20851 }, { "epoch": 2.2606244579358195, "grad_norm": 0.11246483027935028, "learning_rate": 2.473523864790367e-05, "loss": 0.0039, "step": 20852 }, { "epoch": 2.2607328707718994, "grad_norm": 0.2565196752548218, "learning_rate": 2.4731611780066735e-05, "loss": 0.008, "step": 20853 }, { "epoch": 2.260841283607979, "grad_norm": 0.26335322856903076, "learning_rate": 2.47279849122298e-05, "loss": 0.0109, "step": 20854 }, { "epoch": 2.260949696444059, "grad_norm": 0.24347375333309174, "learning_rate": 2.4724358044392863e-05, "loss": 0.0145, "step": 20855 }, { "epoch": 2.261058109280139, "grad_norm": 0.8647462129592896, "learning_rate": 2.4720731176555927e-05, "loss": 0.0217, "step": 20856 }, { "epoch": 2.2611665221162185, "grad_norm": 0.05808773636817932, "learning_rate": 2.471710430871899e-05, "loss": 0.0023, "step": 20857 }, { "epoch": 2.2612749349522985, "grad_norm": 0.1507454812526703, "learning_rate": 2.4713477440882055e-05, "loss": 0.0029, "step": 20858 }, { "epoch": 2.261383347788378, "grad_norm": 0.6147405505180359, "learning_rate": 2.470985057304512e-05, "loss": 0.0367, "step": 20859 }, { "epoch": 2.261491760624458, "grad_norm": 0.01688275672495365, "learning_rate": 2.4706223705208184e-05, "loss": 0.0008, "step": 20860 }, { "epoch": 2.2616001734605375, "grad_norm": 0.09697515517473221, "learning_rate": 2.4702596837371248e-05, "loss": 0.0038, "step": 20861 }, { "epoch": 2.2617085862966175, "grad_norm": 0.02905716374516487, "learning_rate": 2.4698969969534312e-05, "loss": 0.001, "step": 20862 }, { "epoch": 2.261816999132697, "grad_norm": 0.20290417969226837, "learning_rate": 2.4695343101697372e-05, "loss": 0.0159, "step": 20863 }, { "epoch": 2.261925411968777, "grad_norm": 0.7088044285774231, "learning_rate": 2.469171623386044e-05, "loss": 0.0344, "step": 20864 }, { "epoch": 2.262033824804857, "grad_norm": 0.6287381052970886, "learning_rate": 2.4688089366023504e-05, "loss": 0.0528, "step": 20865 }, { "epoch": 2.2621422376409366, "grad_norm": 0.21850596368312836, "learning_rate": 2.4684462498186568e-05, "loss": 0.014, "step": 20866 }, { "epoch": 2.2622506504770166, "grad_norm": 0.11664903908967972, "learning_rate": 2.468083563034963e-05, "loss": 0.007, "step": 20867 }, { "epoch": 2.262359063313096, "grad_norm": 0.17478731274604797, "learning_rate": 2.4677208762512696e-05, "loss": 0.0048, "step": 20868 }, { "epoch": 2.262467476149176, "grad_norm": 0.17020271718502045, "learning_rate": 2.467358189467576e-05, "loss": 0.0055, "step": 20869 }, { "epoch": 2.262575888985256, "grad_norm": 0.0035745035856962204, "learning_rate": 2.4669955026838825e-05, "loss": 0.0001, "step": 20870 }, { "epoch": 2.2626843018213356, "grad_norm": 0.06184342876076698, "learning_rate": 2.4666328159001885e-05, "loss": 0.0017, "step": 20871 }, { "epoch": 2.2627927146574156, "grad_norm": 0.27434152364730835, "learning_rate": 2.4662701291164953e-05, "loss": 0.0092, "step": 20872 }, { "epoch": 2.262901127493495, "grad_norm": 0.10500746965408325, "learning_rate": 2.4659074423328017e-05, "loss": 0.0023, "step": 20873 }, { "epoch": 2.263009540329575, "grad_norm": 0.4327632486820221, "learning_rate": 2.4655447555491078e-05, "loss": 0.0072, "step": 20874 }, { "epoch": 2.2631179531656547, "grad_norm": 0.09802286326885223, "learning_rate": 2.4651820687654142e-05, "loss": 0.0022, "step": 20875 }, { "epoch": 2.2632263660017347, "grad_norm": 0.16144013404846191, "learning_rate": 2.4648193819817206e-05, "loss": 0.005, "step": 20876 }, { "epoch": 2.263334778837814, "grad_norm": 0.276444673538208, "learning_rate": 2.4644566951980273e-05, "loss": 0.0077, "step": 20877 }, { "epoch": 2.263443191673894, "grad_norm": 0.06783103942871094, "learning_rate": 2.4640940084143334e-05, "loss": 0.0054, "step": 20878 }, { "epoch": 2.263551604509974, "grad_norm": 0.19441528618335724, "learning_rate": 2.4637313216306398e-05, "loss": 0.0108, "step": 20879 }, { "epoch": 2.2636600173460537, "grad_norm": 0.11620030552148819, "learning_rate": 2.4633686348469462e-05, "loss": 0.0058, "step": 20880 }, { "epoch": 2.2637684301821337, "grad_norm": 0.7566347122192383, "learning_rate": 2.4630059480632526e-05, "loss": 0.0361, "step": 20881 }, { "epoch": 2.2638768430182132, "grad_norm": 0.013037195429205894, "learning_rate": 2.462643261279559e-05, "loss": 0.0005, "step": 20882 }, { "epoch": 2.263985255854293, "grad_norm": 1.2009038925170898, "learning_rate": 2.4622805744958655e-05, "loss": 0.0237, "step": 20883 }, { "epoch": 2.2640936686903728, "grad_norm": 0.8777422904968262, "learning_rate": 2.461917887712172e-05, "loss": 0.0218, "step": 20884 }, { "epoch": 2.2642020815264527, "grad_norm": 0.35764187574386597, "learning_rate": 2.4615552009284783e-05, "loss": 0.0344, "step": 20885 }, { "epoch": 2.2643104943625323, "grad_norm": 0.08632554113864899, "learning_rate": 2.4611925141447847e-05, "loss": 0.0029, "step": 20886 }, { "epoch": 2.2644189071986123, "grad_norm": 0.19073674082756042, "learning_rate": 2.460829827361091e-05, "loss": 0.0123, "step": 20887 }, { "epoch": 2.2645273200346923, "grad_norm": 0.41270020604133606, "learning_rate": 2.4604671405773975e-05, "loss": 0.0102, "step": 20888 }, { "epoch": 2.264635732870772, "grad_norm": 0.4547208547592163, "learning_rate": 2.4601044537937036e-05, "loss": 0.0264, "step": 20889 }, { "epoch": 2.264744145706852, "grad_norm": 0.20470409095287323, "learning_rate": 2.4597417670100103e-05, "loss": 0.0046, "step": 20890 }, { "epoch": 2.2648525585429313, "grad_norm": 0.5926874279975891, "learning_rate": 2.4593790802263167e-05, "loss": 0.0098, "step": 20891 }, { "epoch": 2.2649609713790113, "grad_norm": 0.4626474380493164, "learning_rate": 2.459016393442623e-05, "loss": 0.0191, "step": 20892 }, { "epoch": 2.2650693842150913, "grad_norm": 0.1299559324979782, "learning_rate": 2.4586537066589292e-05, "loss": 0.0022, "step": 20893 }, { "epoch": 2.265177797051171, "grad_norm": 0.16366824507713318, "learning_rate": 2.458291019875236e-05, "loss": 0.0026, "step": 20894 }, { "epoch": 2.265286209887251, "grad_norm": 0.25816285610198975, "learning_rate": 2.4579283330915424e-05, "loss": 0.0078, "step": 20895 }, { "epoch": 2.2653946227233304, "grad_norm": 0.13192497193813324, "learning_rate": 2.4575656463078488e-05, "loss": 0.0018, "step": 20896 }, { "epoch": 2.2655030355594103, "grad_norm": 0.27597904205322266, "learning_rate": 2.457202959524155e-05, "loss": 0.0107, "step": 20897 }, { "epoch": 2.26561144839549, "grad_norm": 0.009921815246343613, "learning_rate": 2.4568402727404616e-05, "loss": 0.0003, "step": 20898 }, { "epoch": 2.26571986123157, "grad_norm": 0.022645512595772743, "learning_rate": 2.456477585956768e-05, "loss": 0.0006, "step": 20899 }, { "epoch": 2.2658282740676494, "grad_norm": 0.29283133149147034, "learning_rate": 2.456114899173074e-05, "loss": 0.0075, "step": 20900 }, { "epoch": 2.2659366869037294, "grad_norm": 0.17464667558670044, "learning_rate": 2.4557522123893805e-05, "loss": 0.0137, "step": 20901 }, { "epoch": 2.2660450997398094, "grad_norm": 0.7817360162734985, "learning_rate": 2.455389525605687e-05, "loss": 0.0111, "step": 20902 }, { "epoch": 2.266153512575889, "grad_norm": 0.265836238861084, "learning_rate": 2.4550268388219937e-05, "loss": 0.0153, "step": 20903 }, { "epoch": 2.266261925411969, "grad_norm": 0.011576279066503048, "learning_rate": 2.4546641520382997e-05, "loss": 0.0002, "step": 20904 }, { "epoch": 2.2663703382480485, "grad_norm": 0.5531864166259766, "learning_rate": 2.454301465254606e-05, "loss": 0.0228, "step": 20905 }, { "epoch": 2.2664787510841284, "grad_norm": 0.0682983547449112, "learning_rate": 2.4539387784709126e-05, "loss": 0.0024, "step": 20906 }, { "epoch": 2.266587163920208, "grad_norm": 0.07614824920892715, "learning_rate": 2.453576091687219e-05, "loss": 0.003, "step": 20907 }, { "epoch": 2.266695576756288, "grad_norm": 0.09665251523256302, "learning_rate": 2.4532134049035254e-05, "loss": 0.0029, "step": 20908 }, { "epoch": 2.2668039895923675, "grad_norm": 0.06923861801624298, "learning_rate": 2.4528507181198318e-05, "loss": 0.003, "step": 20909 }, { "epoch": 2.2669124024284475, "grad_norm": 0.20973186194896698, "learning_rate": 2.4524880313361382e-05, "loss": 0.0056, "step": 20910 }, { "epoch": 2.2670208152645275, "grad_norm": 0.07444905489683151, "learning_rate": 2.4521253445524446e-05, "loss": 0.002, "step": 20911 }, { "epoch": 2.267129228100607, "grad_norm": 0.010848809964954853, "learning_rate": 2.451762657768751e-05, "loss": 0.0002, "step": 20912 }, { "epoch": 2.267237640936687, "grad_norm": 0.04644732549786568, "learning_rate": 2.4513999709850574e-05, "loss": 0.0022, "step": 20913 }, { "epoch": 2.2673460537727665, "grad_norm": 0.06606346368789673, "learning_rate": 2.451037284201364e-05, "loss": 0.0025, "step": 20914 }, { "epoch": 2.2674544666088465, "grad_norm": 0.05246062949299812, "learning_rate": 2.45067459741767e-05, "loss": 0.0017, "step": 20915 }, { "epoch": 2.2675628794449265, "grad_norm": 0.13715872168540955, "learning_rate": 2.4503119106339767e-05, "loss": 0.0054, "step": 20916 }, { "epoch": 2.267671292281006, "grad_norm": 0.36914241313934326, "learning_rate": 2.449949223850283e-05, "loss": 0.0106, "step": 20917 }, { "epoch": 2.267779705117086, "grad_norm": 0.014606151729822159, "learning_rate": 2.4495865370665895e-05, "loss": 0.0007, "step": 20918 }, { "epoch": 2.2678881179531656, "grad_norm": 0.03157338500022888, "learning_rate": 2.4492238502828956e-05, "loss": 0.0012, "step": 20919 }, { "epoch": 2.2679965307892456, "grad_norm": 0.21096527576446533, "learning_rate": 2.4488611634992023e-05, "loss": 0.0163, "step": 20920 }, { "epoch": 2.268104943625325, "grad_norm": 0.47766730189323425, "learning_rate": 2.4484984767155087e-05, "loss": 0.0423, "step": 20921 }, { "epoch": 2.268213356461405, "grad_norm": 0.22429652512073517, "learning_rate": 2.448135789931815e-05, "loss": 0.0046, "step": 20922 }, { "epoch": 2.2683217692974846, "grad_norm": 0.2662055492401123, "learning_rate": 2.4477731031481212e-05, "loss": 0.0172, "step": 20923 }, { "epoch": 2.2684301821335646, "grad_norm": 0.023360999301075935, "learning_rate": 2.447410416364428e-05, "loss": 0.0004, "step": 20924 }, { "epoch": 2.2685385949696446, "grad_norm": 0.6556142568588257, "learning_rate": 2.4470477295807344e-05, "loss": 0.0179, "step": 20925 }, { "epoch": 2.268647007805724, "grad_norm": 0.15600791573524475, "learning_rate": 2.4466850427970404e-05, "loss": 0.0043, "step": 20926 }, { "epoch": 2.268755420641804, "grad_norm": 0.6460934281349182, "learning_rate": 2.446322356013347e-05, "loss": 0.0166, "step": 20927 }, { "epoch": 2.2688638334778837, "grad_norm": 0.23721586167812347, "learning_rate": 2.4459596692296533e-05, "loss": 0.0171, "step": 20928 }, { "epoch": 2.2689722463139637, "grad_norm": 0.45001116394996643, "learning_rate": 2.44559698244596e-05, "loss": 0.011, "step": 20929 }, { "epoch": 2.269080659150043, "grad_norm": 0.13132250308990479, "learning_rate": 2.445234295662266e-05, "loss": 0.0039, "step": 20930 }, { "epoch": 2.269189071986123, "grad_norm": 0.05131402984261513, "learning_rate": 2.4448716088785725e-05, "loss": 0.0029, "step": 20931 }, { "epoch": 2.2692974848222027, "grad_norm": 0.5244303345680237, "learning_rate": 2.444508922094879e-05, "loss": 0.0099, "step": 20932 }, { "epoch": 2.2694058976582827, "grad_norm": 0.05784386023879051, "learning_rate": 2.4441462353111853e-05, "loss": 0.0016, "step": 20933 }, { "epoch": 2.2695143104943627, "grad_norm": 0.3724193274974823, "learning_rate": 2.4437835485274917e-05, "loss": 0.0196, "step": 20934 }, { "epoch": 2.2696227233304422, "grad_norm": 0.4498310983181, "learning_rate": 2.443420861743798e-05, "loss": 0.0132, "step": 20935 }, { "epoch": 2.2697311361665222, "grad_norm": 0.0908106341958046, "learning_rate": 2.4430581749601045e-05, "loss": 0.0023, "step": 20936 }, { "epoch": 2.2698395490026018, "grad_norm": 0.7297777533531189, "learning_rate": 2.442695488176411e-05, "loss": 0.0293, "step": 20937 }, { "epoch": 2.2699479618386817, "grad_norm": 0.4929225444793701, "learning_rate": 2.4423328013927174e-05, "loss": 0.0411, "step": 20938 }, { "epoch": 2.2700563746747617, "grad_norm": 0.7196268439292908, "learning_rate": 2.4419701146090238e-05, "loss": 0.0207, "step": 20939 }, { "epoch": 2.2701647875108413, "grad_norm": 0.15503579378128052, "learning_rate": 2.4416074278253302e-05, "loss": 0.0056, "step": 20940 }, { "epoch": 2.2702732003469213, "grad_norm": 0.9864652752876282, "learning_rate": 2.4412447410416366e-05, "loss": 0.0771, "step": 20941 }, { "epoch": 2.270381613183001, "grad_norm": 0.3767068386077881, "learning_rate": 2.440882054257943e-05, "loss": 0.0296, "step": 20942 }, { "epoch": 2.270490026019081, "grad_norm": 0.3361334204673767, "learning_rate": 2.4405193674742494e-05, "loss": 0.0253, "step": 20943 }, { "epoch": 2.2705984388551603, "grad_norm": 0.11302629113197327, "learning_rate": 2.4401566806905558e-05, "loss": 0.0024, "step": 20944 }, { "epoch": 2.2707068516912403, "grad_norm": 0.08600515872240067, "learning_rate": 2.439793993906862e-05, "loss": 0.0055, "step": 20945 }, { "epoch": 2.27081526452732, "grad_norm": 0.04493909329175949, "learning_rate": 2.4394313071231686e-05, "loss": 0.0009, "step": 20946 }, { "epoch": 2.2709236773634, "grad_norm": 0.0455460399389267, "learning_rate": 2.439068620339475e-05, "loss": 0.0025, "step": 20947 }, { "epoch": 2.27103209019948, "grad_norm": 0.7627886533737183, "learning_rate": 2.4387059335557815e-05, "loss": 0.0136, "step": 20948 }, { "epoch": 2.2711405030355594, "grad_norm": 0.03341567516326904, "learning_rate": 2.4383432467720875e-05, "loss": 0.0016, "step": 20949 }, { "epoch": 2.2712489158716394, "grad_norm": 0.27954596281051636, "learning_rate": 2.4379805599883943e-05, "loss": 0.0095, "step": 20950 }, { "epoch": 2.271357328707719, "grad_norm": 0.08736982941627502, "learning_rate": 2.4376178732047007e-05, "loss": 0.0022, "step": 20951 }, { "epoch": 2.271465741543799, "grad_norm": 0.13821814954280853, "learning_rate": 2.4372551864210068e-05, "loss": 0.0086, "step": 20952 }, { "epoch": 2.2715741543798784, "grad_norm": 0.29089465737342834, "learning_rate": 2.4368924996373132e-05, "loss": 0.0107, "step": 20953 }, { "epoch": 2.2716825672159584, "grad_norm": 0.16756360232830048, "learning_rate": 2.43652981285362e-05, "loss": 0.0068, "step": 20954 }, { "epoch": 2.271790980052038, "grad_norm": 0.7056508660316467, "learning_rate": 2.4361671260699263e-05, "loss": 0.0239, "step": 20955 }, { "epoch": 2.271899392888118, "grad_norm": 0.8290842175483704, "learning_rate": 2.4358044392862324e-05, "loss": 0.0247, "step": 20956 }, { "epoch": 2.272007805724198, "grad_norm": 0.7785975337028503, "learning_rate": 2.4354417525025388e-05, "loss": 0.0075, "step": 20957 }, { "epoch": 2.2721162185602775, "grad_norm": 0.09436548501253128, "learning_rate": 2.4350790657188452e-05, "loss": 0.0033, "step": 20958 }, { "epoch": 2.2722246313963574, "grad_norm": 0.7031676173210144, "learning_rate": 2.434716378935152e-05, "loss": 0.0261, "step": 20959 }, { "epoch": 2.272333044232437, "grad_norm": 0.027937185019254684, "learning_rate": 2.434353692151458e-05, "loss": 0.0015, "step": 20960 }, { "epoch": 2.272441457068517, "grad_norm": 0.5364357233047485, "learning_rate": 2.4339910053677645e-05, "loss": 0.028, "step": 20961 }, { "epoch": 2.272549869904597, "grad_norm": 0.34598270058631897, "learning_rate": 2.433628318584071e-05, "loss": 0.0104, "step": 20962 }, { "epoch": 2.2726582827406765, "grad_norm": 0.23331555724143982, "learning_rate": 2.4332656318003773e-05, "loss": 0.0095, "step": 20963 }, { "epoch": 2.2727666955767565, "grad_norm": 0.026955556124448776, "learning_rate": 2.4329029450166837e-05, "loss": 0.001, "step": 20964 }, { "epoch": 2.272875108412836, "grad_norm": 0.47510063648223877, "learning_rate": 2.43254025823299e-05, "loss": 0.005, "step": 20965 }, { "epoch": 2.272983521248916, "grad_norm": 0.015689633786678314, "learning_rate": 2.4321775714492965e-05, "loss": 0.0008, "step": 20966 }, { "epoch": 2.2730919340849955, "grad_norm": 1.735357642173767, "learning_rate": 2.431814884665603e-05, "loss": 0.0386, "step": 20967 }, { "epoch": 2.2732003469210755, "grad_norm": 0.4621245265007019, "learning_rate": 2.4314521978819093e-05, "loss": 0.0073, "step": 20968 }, { "epoch": 2.273308759757155, "grad_norm": 0.11469026654958725, "learning_rate": 2.4310895110982157e-05, "loss": 0.0027, "step": 20969 }, { "epoch": 2.273417172593235, "grad_norm": 0.31105634570121765, "learning_rate": 2.430726824314522e-05, "loss": 0.0094, "step": 20970 }, { "epoch": 2.273525585429315, "grad_norm": 0.09672147780656815, "learning_rate": 2.4303641375308282e-05, "loss": 0.0064, "step": 20971 }, { "epoch": 2.2736339982653946, "grad_norm": 0.2627466320991516, "learning_rate": 2.430001450747135e-05, "loss": 0.0093, "step": 20972 }, { "epoch": 2.2737424111014746, "grad_norm": 0.15983714163303375, "learning_rate": 2.4296387639634414e-05, "loss": 0.0043, "step": 20973 }, { "epoch": 2.273850823937554, "grad_norm": 0.48230499029159546, "learning_rate": 2.4292760771797478e-05, "loss": 0.0094, "step": 20974 }, { "epoch": 2.273959236773634, "grad_norm": 0.1696319878101349, "learning_rate": 2.428913390396054e-05, "loss": 0.0059, "step": 20975 }, { "epoch": 2.2740676496097136, "grad_norm": 0.23293228447437286, "learning_rate": 2.4285507036123606e-05, "loss": 0.0113, "step": 20976 }, { "epoch": 2.2741760624457936, "grad_norm": 0.3175424039363861, "learning_rate": 2.428188016828667e-05, "loss": 0.0083, "step": 20977 }, { "epoch": 2.274284475281873, "grad_norm": 0.07092534750699997, "learning_rate": 2.427825330044973e-05, "loss": 0.0029, "step": 20978 }, { "epoch": 2.274392888117953, "grad_norm": 0.03521927073597908, "learning_rate": 2.4274626432612795e-05, "loss": 0.0007, "step": 20979 }, { "epoch": 2.274501300954033, "grad_norm": 0.328946053981781, "learning_rate": 2.4270999564775863e-05, "loss": 0.0123, "step": 20980 }, { "epoch": 2.2746097137901127, "grad_norm": 0.020721185952425003, "learning_rate": 2.4267372696938927e-05, "loss": 0.0013, "step": 20981 }, { "epoch": 2.2747181266261927, "grad_norm": 0.5294219851493835, "learning_rate": 2.4263745829101987e-05, "loss": 0.0241, "step": 20982 }, { "epoch": 2.274826539462272, "grad_norm": 0.2294122874736786, "learning_rate": 2.426011896126505e-05, "loss": 0.0059, "step": 20983 }, { "epoch": 2.274934952298352, "grad_norm": 1.0142184495925903, "learning_rate": 2.4256492093428116e-05, "loss": 0.0638, "step": 20984 }, { "epoch": 2.2750433651344317, "grad_norm": 0.44561731815338135, "learning_rate": 2.4252865225591183e-05, "loss": 0.04, "step": 20985 }, { "epoch": 2.2751517779705117, "grad_norm": 0.49333086609840393, "learning_rate": 2.4249238357754244e-05, "loss": 0.0141, "step": 20986 }, { "epoch": 2.2752601908065913, "grad_norm": 0.6755169034004211, "learning_rate": 2.4245611489917308e-05, "loss": 0.0414, "step": 20987 }, { "epoch": 2.2753686036426712, "grad_norm": 0.22823411226272583, "learning_rate": 2.4241984622080372e-05, "loss": 0.0166, "step": 20988 }, { "epoch": 2.2754770164787512, "grad_norm": 0.17469963431358337, "learning_rate": 2.4238357754243436e-05, "loss": 0.0085, "step": 20989 }, { "epoch": 2.2755854293148308, "grad_norm": 0.012266688980162144, "learning_rate": 2.42347308864065e-05, "loss": 0.0004, "step": 20990 }, { "epoch": 2.2756938421509108, "grad_norm": 0.28670474886894226, "learning_rate": 2.4231104018569564e-05, "loss": 0.0067, "step": 20991 }, { "epoch": 2.2758022549869903, "grad_norm": 0.363198846578598, "learning_rate": 2.422747715073263e-05, "loss": 0.0188, "step": 20992 }, { "epoch": 2.2759106678230703, "grad_norm": 0.4114953577518463, "learning_rate": 2.4223850282895693e-05, "loss": 0.0271, "step": 20993 }, { "epoch": 2.2760190806591503, "grad_norm": 0.5931246280670166, "learning_rate": 2.4220223415058757e-05, "loss": 0.0169, "step": 20994 }, { "epoch": 2.27612749349523, "grad_norm": 0.7786319255828857, "learning_rate": 2.421659654722182e-05, "loss": 0.0294, "step": 20995 }, { "epoch": 2.27623590633131, "grad_norm": 0.08070899546146393, "learning_rate": 2.4212969679384885e-05, "loss": 0.0028, "step": 20996 }, { "epoch": 2.2763443191673893, "grad_norm": 0.44339144229888916, "learning_rate": 2.4209342811547946e-05, "loss": 0.0834, "step": 20997 }, { "epoch": 2.2764527320034693, "grad_norm": 0.14320853352546692, "learning_rate": 2.4205715943711013e-05, "loss": 0.0018, "step": 20998 }, { "epoch": 2.276561144839549, "grad_norm": 0.8046587109565735, "learning_rate": 2.4202089075874077e-05, "loss": 0.013, "step": 20999 }, { "epoch": 2.276669557675629, "grad_norm": 0.024489186704158783, "learning_rate": 2.419846220803714e-05, "loss": 0.0009, "step": 21000 }, { "epoch": 2.2767779705117084, "grad_norm": 0.03707531467080116, "learning_rate": 2.4194835340200202e-05, "loss": 0.0009, "step": 21001 }, { "epoch": 2.2768863833477884, "grad_norm": 0.06689997017383575, "learning_rate": 2.419120847236327e-05, "loss": 0.0026, "step": 21002 }, { "epoch": 2.2769947961838684, "grad_norm": 0.2754250168800354, "learning_rate": 2.4187581604526334e-05, "loss": 0.0133, "step": 21003 }, { "epoch": 2.277103209019948, "grad_norm": 0.12714123725891113, "learning_rate": 2.4183954736689394e-05, "loss": 0.0027, "step": 21004 }, { "epoch": 2.277211621856028, "grad_norm": 0.17556414008140564, "learning_rate": 2.418032786885246e-05, "loss": 0.0031, "step": 21005 }, { "epoch": 2.2773200346921074, "grad_norm": 0.16042378544807434, "learning_rate": 2.4176701001015526e-05, "loss": 0.0076, "step": 21006 }, { "epoch": 2.2774284475281874, "grad_norm": 0.5007526278495789, "learning_rate": 2.417307413317859e-05, "loss": 0.0236, "step": 21007 }, { "epoch": 2.277536860364267, "grad_norm": 0.22508832812309265, "learning_rate": 2.416944726534165e-05, "loss": 0.005, "step": 21008 }, { "epoch": 2.277645273200347, "grad_norm": 0.14020583033561707, "learning_rate": 2.4165820397504715e-05, "loss": 0.0092, "step": 21009 }, { "epoch": 2.2777536860364265, "grad_norm": 0.3663398325443268, "learning_rate": 2.416219352966778e-05, "loss": 0.0087, "step": 21010 }, { "epoch": 2.2778620988725065, "grad_norm": 0.08613841235637665, "learning_rate": 2.4158566661830846e-05, "loss": 0.0047, "step": 21011 }, { "epoch": 2.2779705117085864, "grad_norm": 0.6261516213417053, "learning_rate": 2.4154939793993907e-05, "loss": 0.0201, "step": 21012 }, { "epoch": 2.278078924544666, "grad_norm": 0.016480019316077232, "learning_rate": 2.415131292615697e-05, "loss": 0.0007, "step": 21013 }, { "epoch": 2.278187337380746, "grad_norm": 0.07006333768367767, "learning_rate": 2.4147686058320035e-05, "loss": 0.0016, "step": 21014 }, { "epoch": 2.2782957502168255, "grad_norm": 0.19838884472846985, "learning_rate": 2.41440591904831e-05, "loss": 0.0047, "step": 21015 }, { "epoch": 2.2784041630529055, "grad_norm": 0.06355523318052292, "learning_rate": 2.4140432322646164e-05, "loss": 0.0022, "step": 21016 }, { "epoch": 2.2785125758889855, "grad_norm": 0.6464172005653381, "learning_rate": 2.4136805454809228e-05, "loss": 0.0634, "step": 21017 }, { "epoch": 2.278620988725065, "grad_norm": 0.8716325163841248, "learning_rate": 2.4133178586972292e-05, "loss": 0.012, "step": 21018 }, { "epoch": 2.278729401561145, "grad_norm": 0.22538544237613678, "learning_rate": 2.4129551719135356e-05, "loss": 0.0076, "step": 21019 }, { "epoch": 2.2788378143972245, "grad_norm": 0.8522082567214966, "learning_rate": 2.412592485129842e-05, "loss": 0.048, "step": 21020 }, { "epoch": 2.2789462272333045, "grad_norm": 0.15603876113891602, "learning_rate": 2.4122297983461484e-05, "loss": 0.0045, "step": 21021 }, { "epoch": 2.279054640069384, "grad_norm": 0.4262273907661438, "learning_rate": 2.4118671115624548e-05, "loss": 0.0202, "step": 21022 }, { "epoch": 2.279163052905464, "grad_norm": 0.081180140376091, "learning_rate": 2.411504424778761e-05, "loss": 0.0014, "step": 21023 }, { "epoch": 2.2792714657415436, "grad_norm": 0.3186872601509094, "learning_rate": 2.4111417379950676e-05, "loss": 0.0114, "step": 21024 }, { "epoch": 2.2793798785776236, "grad_norm": 0.36256566643714905, "learning_rate": 2.410779051211374e-05, "loss": 0.0047, "step": 21025 }, { "epoch": 2.2794882914137036, "grad_norm": 0.15052156150341034, "learning_rate": 2.4104163644276805e-05, "loss": 0.0026, "step": 21026 }, { "epoch": 2.279596704249783, "grad_norm": 0.4104742407798767, "learning_rate": 2.4100536776439865e-05, "loss": 0.0152, "step": 21027 }, { "epoch": 2.279705117085863, "grad_norm": 0.39326852560043335, "learning_rate": 2.4096909908602933e-05, "loss": 0.0365, "step": 21028 }, { "epoch": 2.2798135299219426, "grad_norm": 0.07167940586805344, "learning_rate": 2.4093283040765997e-05, "loss": 0.0015, "step": 21029 }, { "epoch": 2.2799219427580226, "grad_norm": 0.32051265239715576, "learning_rate": 2.4089656172929058e-05, "loss": 0.0096, "step": 21030 }, { "epoch": 2.280030355594102, "grad_norm": 0.05122016742825508, "learning_rate": 2.4086029305092122e-05, "loss": 0.0014, "step": 21031 }, { "epoch": 2.280138768430182, "grad_norm": 0.30012011528015137, "learning_rate": 2.408240243725519e-05, "loss": 0.022, "step": 21032 }, { "epoch": 2.2802471812662617, "grad_norm": 0.2867344915866852, "learning_rate": 2.4078775569418253e-05, "loss": 0.0098, "step": 21033 }, { "epoch": 2.2803555941023417, "grad_norm": 0.19112519919872284, "learning_rate": 2.4075148701581314e-05, "loss": 0.0075, "step": 21034 }, { "epoch": 2.2804640069384217, "grad_norm": 0.4719691276550293, "learning_rate": 2.4071521833744378e-05, "loss": 0.0164, "step": 21035 }, { "epoch": 2.280572419774501, "grad_norm": 0.5014389753341675, "learning_rate": 2.4067894965907446e-05, "loss": 0.0544, "step": 21036 }, { "epoch": 2.280680832610581, "grad_norm": 0.8408514261245728, "learning_rate": 2.406426809807051e-05, "loss": 0.0291, "step": 21037 }, { "epoch": 2.2807892454466607, "grad_norm": 0.2739812731742859, "learning_rate": 2.406064123023357e-05, "loss": 0.006, "step": 21038 }, { "epoch": 2.2808976582827407, "grad_norm": 0.2761092782020569, "learning_rate": 2.4057014362396635e-05, "loss": 0.015, "step": 21039 }, { "epoch": 2.2810060711188207, "grad_norm": 0.2900695204734802, "learning_rate": 2.40533874945597e-05, "loss": 0.0151, "step": 21040 }, { "epoch": 2.2811144839549002, "grad_norm": 0.4598482549190521, "learning_rate": 2.4049760626722763e-05, "loss": 0.0867, "step": 21041 }, { "epoch": 2.2812228967909802, "grad_norm": 0.07526609301567078, "learning_rate": 2.4046133758885827e-05, "loss": 0.0019, "step": 21042 }, { "epoch": 2.2813313096270598, "grad_norm": 0.3701874613761902, "learning_rate": 2.404250689104889e-05, "loss": 0.0135, "step": 21043 }, { "epoch": 2.2814397224631398, "grad_norm": 0.05057772621512413, "learning_rate": 2.4038880023211955e-05, "loss": 0.0011, "step": 21044 }, { "epoch": 2.2815481352992193, "grad_norm": 0.36110997200012207, "learning_rate": 2.403525315537502e-05, "loss": 0.0083, "step": 21045 }, { "epoch": 2.2816565481352993, "grad_norm": 0.37894222140312195, "learning_rate": 2.4031626287538083e-05, "loss": 0.0204, "step": 21046 }, { "epoch": 2.281764960971379, "grad_norm": 0.3029083013534546, "learning_rate": 2.4027999419701147e-05, "loss": 0.015, "step": 21047 }, { "epoch": 2.281873373807459, "grad_norm": 0.4745171070098877, "learning_rate": 2.402437255186421e-05, "loss": 0.0081, "step": 21048 }, { "epoch": 2.281981786643539, "grad_norm": 0.5904983282089233, "learning_rate": 2.4020745684027276e-05, "loss": 0.0572, "step": 21049 }, { "epoch": 2.2820901994796183, "grad_norm": 0.027864228934049606, "learning_rate": 2.401711881619034e-05, "loss": 0.0008, "step": 21050 }, { "epoch": 2.2821986123156983, "grad_norm": 0.6616306900978088, "learning_rate": 2.4013491948353404e-05, "loss": 0.0159, "step": 21051 }, { "epoch": 2.282307025151778, "grad_norm": 0.13997423648834229, "learning_rate": 2.4009865080516468e-05, "loss": 0.0047, "step": 21052 }, { "epoch": 2.282415437987858, "grad_norm": 0.06749094277620316, "learning_rate": 2.400623821267953e-05, "loss": 0.0017, "step": 21053 }, { "epoch": 2.2825238508239374, "grad_norm": 1.0532373189926147, "learning_rate": 2.4002611344842596e-05, "loss": 0.058, "step": 21054 }, { "epoch": 2.2826322636600174, "grad_norm": 0.0283553097397089, "learning_rate": 2.399898447700566e-05, "loss": 0.0011, "step": 21055 }, { "epoch": 2.282740676496097, "grad_norm": 0.052653227001428604, "learning_rate": 2.399535760916872e-05, "loss": 0.0022, "step": 21056 }, { "epoch": 2.282849089332177, "grad_norm": 0.16377823054790497, "learning_rate": 2.3991730741331785e-05, "loss": 0.0063, "step": 21057 }, { "epoch": 2.282957502168257, "grad_norm": 0.03128773346543312, "learning_rate": 2.3988103873494853e-05, "loss": 0.0014, "step": 21058 }, { "epoch": 2.2830659150043364, "grad_norm": 0.16581933200359344, "learning_rate": 2.3984477005657917e-05, "loss": 0.0076, "step": 21059 }, { "epoch": 2.2831743278404164, "grad_norm": 0.015050487592816353, "learning_rate": 2.3980850137820977e-05, "loss": 0.0008, "step": 21060 }, { "epoch": 2.283282740676496, "grad_norm": 0.22096218168735504, "learning_rate": 2.397722326998404e-05, "loss": 0.0096, "step": 21061 }, { "epoch": 2.283391153512576, "grad_norm": 0.24294576048851013, "learning_rate": 2.397359640214711e-05, "loss": 0.0105, "step": 21062 }, { "epoch": 2.283499566348656, "grad_norm": 0.2231406569480896, "learning_rate": 2.3969969534310173e-05, "loss": 0.0229, "step": 21063 }, { "epoch": 2.2836079791847355, "grad_norm": 0.9899451732635498, "learning_rate": 2.3966342666473234e-05, "loss": 0.0069, "step": 21064 }, { "epoch": 2.2837163920208154, "grad_norm": 0.06564615666866302, "learning_rate": 2.3962715798636298e-05, "loss": 0.0019, "step": 21065 }, { "epoch": 2.283824804856895, "grad_norm": 0.2468237578868866, "learning_rate": 2.3959088930799362e-05, "loss": 0.0141, "step": 21066 }, { "epoch": 2.283933217692975, "grad_norm": 0.06616946309804916, "learning_rate": 2.3955462062962426e-05, "loss": 0.0024, "step": 21067 }, { "epoch": 2.2840416305290545, "grad_norm": 0.038993630558252335, "learning_rate": 2.395183519512549e-05, "loss": 0.0017, "step": 21068 }, { "epoch": 2.2841500433651345, "grad_norm": 0.049928370863199234, "learning_rate": 2.3948208327288554e-05, "loss": 0.0019, "step": 21069 }, { "epoch": 2.284258456201214, "grad_norm": 0.440035343170166, "learning_rate": 2.394458145945162e-05, "loss": 0.0364, "step": 21070 }, { "epoch": 2.284366869037294, "grad_norm": 0.018039081245660782, "learning_rate": 2.3940954591614683e-05, "loss": 0.0008, "step": 21071 }, { "epoch": 2.284475281873374, "grad_norm": 0.07134753465652466, "learning_rate": 2.3937327723777747e-05, "loss": 0.0014, "step": 21072 }, { "epoch": 2.2845836947094535, "grad_norm": 0.15327979624271393, "learning_rate": 2.393370085594081e-05, "loss": 0.008, "step": 21073 }, { "epoch": 2.2846921075455335, "grad_norm": 0.25815197825431824, "learning_rate": 2.3930073988103875e-05, "loss": 0.0067, "step": 21074 }, { "epoch": 2.284800520381613, "grad_norm": 0.5222631692886353, "learning_rate": 2.392644712026694e-05, "loss": 0.0455, "step": 21075 }, { "epoch": 2.284908933217693, "grad_norm": 0.1897786408662796, "learning_rate": 2.3922820252430003e-05, "loss": 0.0034, "step": 21076 }, { "epoch": 2.2850173460537726, "grad_norm": 0.5454238057136536, "learning_rate": 2.3919193384593067e-05, "loss": 0.0122, "step": 21077 }, { "epoch": 2.2851257588898526, "grad_norm": 0.24217116832733154, "learning_rate": 2.391556651675613e-05, "loss": 0.007, "step": 21078 }, { "epoch": 2.285234171725932, "grad_norm": 0.38970157504081726, "learning_rate": 2.3911939648919192e-05, "loss": 0.0316, "step": 21079 }, { "epoch": 2.285342584562012, "grad_norm": 0.048823196440935135, "learning_rate": 2.390831278108226e-05, "loss": 0.0017, "step": 21080 }, { "epoch": 2.285450997398092, "grad_norm": 0.19582319259643555, "learning_rate": 2.3904685913245324e-05, "loss": 0.0084, "step": 21081 }, { "epoch": 2.2855594102341716, "grad_norm": 0.21805450320243835, "learning_rate": 2.3901059045408384e-05, "loss": 0.0189, "step": 21082 }, { "epoch": 2.2856678230702516, "grad_norm": 0.3126150369644165, "learning_rate": 2.389743217757145e-05, "loss": 0.009, "step": 21083 }, { "epoch": 2.285776235906331, "grad_norm": 0.5704996585845947, "learning_rate": 2.3893805309734516e-05, "loss": 0.0292, "step": 21084 }, { "epoch": 2.285884648742411, "grad_norm": 0.18868619203567505, "learning_rate": 2.389017844189758e-05, "loss": 0.0033, "step": 21085 }, { "epoch": 2.285993061578491, "grad_norm": 0.017801115289330482, "learning_rate": 2.388655157406064e-05, "loss": 0.0008, "step": 21086 }, { "epoch": 2.2861014744145707, "grad_norm": 0.2445869743824005, "learning_rate": 2.3882924706223705e-05, "loss": 0.0157, "step": 21087 }, { "epoch": 2.2862098872506507, "grad_norm": 0.646140456199646, "learning_rate": 2.3879297838386772e-05, "loss": 0.0597, "step": 21088 }, { "epoch": 2.28631830008673, "grad_norm": 0.3946373164653778, "learning_rate": 2.3875670970549837e-05, "loss": 0.0249, "step": 21089 }, { "epoch": 2.28642671292281, "grad_norm": 0.10316520184278488, "learning_rate": 2.3872044102712897e-05, "loss": 0.0071, "step": 21090 }, { "epoch": 2.2865351257588897, "grad_norm": 0.19851773977279663, "learning_rate": 2.386841723487596e-05, "loss": 0.0076, "step": 21091 }, { "epoch": 2.2866435385949697, "grad_norm": 0.0284042339771986, "learning_rate": 2.3864790367039025e-05, "loss": 0.001, "step": 21092 }, { "epoch": 2.2867519514310493, "grad_norm": 0.46758508682250977, "learning_rate": 2.386116349920209e-05, "loss": 0.0077, "step": 21093 }, { "epoch": 2.2868603642671292, "grad_norm": 0.14084482192993164, "learning_rate": 2.3857536631365154e-05, "loss": 0.0038, "step": 21094 }, { "epoch": 2.2869687771032092, "grad_norm": 0.08044816553592682, "learning_rate": 2.3853909763528218e-05, "loss": 0.0043, "step": 21095 }, { "epoch": 2.2870771899392888, "grad_norm": 0.14025671780109406, "learning_rate": 2.3850282895691282e-05, "loss": 0.0038, "step": 21096 }, { "epoch": 2.2871856027753688, "grad_norm": 0.096467524766922, "learning_rate": 2.3846656027854346e-05, "loss": 0.0043, "step": 21097 }, { "epoch": 2.2872940156114483, "grad_norm": 0.3619433343410492, "learning_rate": 2.384302916001741e-05, "loss": 0.0094, "step": 21098 }, { "epoch": 2.2874024284475283, "grad_norm": 0.2161736786365509, "learning_rate": 2.3839402292180474e-05, "loss": 0.011, "step": 21099 }, { "epoch": 2.287510841283608, "grad_norm": 0.029463429003953934, "learning_rate": 2.3835775424343538e-05, "loss": 0.0006, "step": 21100 }, { "epoch": 2.287619254119688, "grad_norm": 0.46790680289268494, "learning_rate": 2.3832148556506602e-05, "loss": 0.0172, "step": 21101 }, { "epoch": 2.2877276669557673, "grad_norm": 0.38218188285827637, "learning_rate": 2.3828521688669666e-05, "loss": 0.024, "step": 21102 }, { "epoch": 2.2878360797918473, "grad_norm": 0.10742480307817459, "learning_rate": 2.382489482083273e-05, "loss": 0.0044, "step": 21103 }, { "epoch": 2.2879444926279273, "grad_norm": 0.2280297577381134, "learning_rate": 2.3821267952995795e-05, "loss": 0.0065, "step": 21104 }, { "epoch": 2.288052905464007, "grad_norm": 0.08164010941982269, "learning_rate": 2.3817641085158855e-05, "loss": 0.003, "step": 21105 }, { "epoch": 2.288161318300087, "grad_norm": 0.15641574561595917, "learning_rate": 2.3814014217321923e-05, "loss": 0.0054, "step": 21106 }, { "epoch": 2.2882697311361664, "grad_norm": 0.5256544351577759, "learning_rate": 2.3810387349484987e-05, "loss": 0.0177, "step": 21107 }, { "epoch": 2.2883781439722464, "grad_norm": 0.47092223167419434, "learning_rate": 2.3806760481648048e-05, "loss": 0.0374, "step": 21108 }, { "epoch": 2.2884865568083264, "grad_norm": 0.5252938270568848, "learning_rate": 2.3803133613811112e-05, "loss": 0.0062, "step": 21109 }, { "epoch": 2.288594969644406, "grad_norm": 0.39641308784484863, "learning_rate": 2.379950674597418e-05, "loss": 0.028, "step": 21110 }, { "epoch": 2.288703382480486, "grad_norm": 0.19146430492401123, "learning_rate": 2.3795879878137243e-05, "loss": 0.0086, "step": 21111 }, { "epoch": 2.2888117953165654, "grad_norm": 0.19207173585891724, "learning_rate": 2.3792253010300304e-05, "loss": 0.0118, "step": 21112 }, { "epoch": 2.2889202081526454, "grad_norm": 0.407754123210907, "learning_rate": 2.3788626142463368e-05, "loss": 0.0263, "step": 21113 }, { "epoch": 2.289028620988725, "grad_norm": 0.017470698803663254, "learning_rate": 2.3784999274626436e-05, "loss": 0.0006, "step": 21114 }, { "epoch": 2.289137033824805, "grad_norm": 0.44433456659317017, "learning_rate": 2.37813724067895e-05, "loss": 0.0137, "step": 21115 }, { "epoch": 2.2892454466608845, "grad_norm": 0.9361205697059631, "learning_rate": 2.377774553895256e-05, "loss": 0.0253, "step": 21116 }, { "epoch": 2.2893538594969645, "grad_norm": 0.08278854936361313, "learning_rate": 2.3774118671115625e-05, "loss": 0.004, "step": 21117 }, { "epoch": 2.2894622723330444, "grad_norm": 0.08526475727558136, "learning_rate": 2.377049180327869e-05, "loss": 0.0032, "step": 21118 }, { "epoch": 2.289570685169124, "grad_norm": 0.796617865562439, "learning_rate": 2.3766864935441753e-05, "loss": 0.0208, "step": 21119 }, { "epoch": 2.289679098005204, "grad_norm": 0.06368191540241241, "learning_rate": 2.3763238067604817e-05, "loss": 0.0015, "step": 21120 }, { "epoch": 2.2897875108412835, "grad_norm": 0.16176681220531464, "learning_rate": 2.375961119976788e-05, "loss": 0.0049, "step": 21121 }, { "epoch": 2.2898959236773635, "grad_norm": 0.03919529542326927, "learning_rate": 2.3755984331930945e-05, "loss": 0.001, "step": 21122 }, { "epoch": 2.290004336513443, "grad_norm": 0.28356656432151794, "learning_rate": 2.375235746409401e-05, "loss": 0.0109, "step": 21123 }, { "epoch": 2.290112749349523, "grad_norm": 0.14192414283752441, "learning_rate": 2.3748730596257073e-05, "loss": 0.0055, "step": 21124 }, { "epoch": 2.2902211621856026, "grad_norm": 0.1077616959810257, "learning_rate": 2.3745103728420138e-05, "loss": 0.0048, "step": 21125 }, { "epoch": 2.2903295750216826, "grad_norm": 0.13779006898403168, "learning_rate": 2.37414768605832e-05, "loss": 0.0028, "step": 21126 }, { "epoch": 2.2904379878577625, "grad_norm": 0.09653101116418839, "learning_rate": 2.3737849992746266e-05, "loss": 0.0027, "step": 21127 }, { "epoch": 2.290546400693842, "grad_norm": 0.0786709263920784, "learning_rate": 2.373422312490933e-05, "loss": 0.0033, "step": 21128 }, { "epoch": 2.290654813529922, "grad_norm": 0.17522680759429932, "learning_rate": 2.3730596257072394e-05, "loss": 0.0077, "step": 21129 }, { "epoch": 2.2907632263660016, "grad_norm": 0.4080250561237335, "learning_rate": 2.3726969389235458e-05, "loss": 0.0505, "step": 21130 }, { "epoch": 2.2908716392020816, "grad_norm": 0.15801864862442017, "learning_rate": 2.3723342521398522e-05, "loss": 0.0094, "step": 21131 }, { "epoch": 2.290980052038161, "grad_norm": 0.06430748105049133, "learning_rate": 2.3719715653561586e-05, "loss": 0.0015, "step": 21132 }, { "epoch": 2.291088464874241, "grad_norm": 0.41488736867904663, "learning_rate": 2.371608878572465e-05, "loss": 0.0483, "step": 21133 }, { "epoch": 2.2911968777103207, "grad_norm": 0.41403982043266296, "learning_rate": 2.3712461917887714e-05, "loss": 0.0103, "step": 21134 }, { "epoch": 2.2913052905464006, "grad_norm": 0.07232562452554703, "learning_rate": 2.3708835050050775e-05, "loss": 0.0024, "step": 21135 }, { "epoch": 2.2914137033824806, "grad_norm": 0.056506842374801636, "learning_rate": 2.3705208182213843e-05, "loss": 0.003, "step": 21136 }, { "epoch": 2.29152211621856, "grad_norm": 0.05267344415187836, "learning_rate": 2.3701581314376907e-05, "loss": 0.0014, "step": 21137 }, { "epoch": 2.29163052905464, "grad_norm": 0.03665214031934738, "learning_rate": 2.3697954446539967e-05, "loss": 0.0009, "step": 21138 }, { "epoch": 2.2917389418907197, "grad_norm": 0.12823890149593353, "learning_rate": 2.369432757870303e-05, "loss": 0.007, "step": 21139 }, { "epoch": 2.2918473547267997, "grad_norm": 0.23937182128429413, "learning_rate": 2.36907007108661e-05, "loss": 0.01, "step": 21140 }, { "epoch": 2.2919557675628797, "grad_norm": 0.6152009963989258, "learning_rate": 2.3687073843029163e-05, "loss": 0.012, "step": 21141 }, { "epoch": 2.292064180398959, "grad_norm": 0.3305315375328064, "learning_rate": 2.3683446975192224e-05, "loss": 0.0127, "step": 21142 }, { "epoch": 2.292172593235039, "grad_norm": 0.4070303738117218, "learning_rate": 2.3679820107355288e-05, "loss": 0.0346, "step": 21143 }, { "epoch": 2.2922810060711187, "grad_norm": 0.013800747692584991, "learning_rate": 2.3676193239518356e-05, "loss": 0.0006, "step": 21144 }, { "epoch": 2.2923894189071987, "grad_norm": 0.1782960593700409, "learning_rate": 2.3672566371681416e-05, "loss": 0.0189, "step": 21145 }, { "epoch": 2.2924978317432783, "grad_norm": 0.09606306999921799, "learning_rate": 2.366893950384448e-05, "loss": 0.0026, "step": 21146 }, { "epoch": 2.2926062445793582, "grad_norm": 0.29592281579971313, "learning_rate": 2.3665312636007544e-05, "loss": 0.009, "step": 21147 }, { "epoch": 2.292714657415438, "grad_norm": 0.2051515430212021, "learning_rate": 2.366168576817061e-05, "loss": 0.0053, "step": 21148 }, { "epoch": 2.2928230702515178, "grad_norm": 0.2908532917499542, "learning_rate": 2.3658058900333673e-05, "loss": 0.0084, "step": 21149 }, { "epoch": 2.2929314830875978, "grad_norm": 0.10413758456707001, "learning_rate": 2.3654432032496737e-05, "loss": 0.0031, "step": 21150 }, { "epoch": 2.2930398959236773, "grad_norm": 0.3995596170425415, "learning_rate": 2.36508051646598e-05, "loss": 0.0139, "step": 21151 }, { "epoch": 2.2931483087597573, "grad_norm": 0.19478604197502136, "learning_rate": 2.3647178296822865e-05, "loss": 0.0099, "step": 21152 }, { "epoch": 2.293256721595837, "grad_norm": 0.11818322539329529, "learning_rate": 2.364355142898593e-05, "loss": 0.0041, "step": 21153 }, { "epoch": 2.293365134431917, "grad_norm": 0.7334927320480347, "learning_rate": 2.3639924561148993e-05, "loss": 0.037, "step": 21154 }, { "epoch": 2.2934735472679963, "grad_norm": 0.009848827496170998, "learning_rate": 2.3636297693312057e-05, "loss": 0.0005, "step": 21155 }, { "epoch": 2.2935819601040763, "grad_norm": 0.38674941658973694, "learning_rate": 2.363267082547512e-05, "loss": 0.0292, "step": 21156 }, { "epoch": 2.293690372940156, "grad_norm": 0.2882811725139618, "learning_rate": 2.3629043957638185e-05, "loss": 0.0074, "step": 21157 }, { "epoch": 2.293798785776236, "grad_norm": 0.02262142114341259, "learning_rate": 2.362541708980125e-05, "loss": 0.0004, "step": 21158 }, { "epoch": 2.293907198612316, "grad_norm": 0.616718053817749, "learning_rate": 2.3621790221964314e-05, "loss": 0.0305, "step": 21159 }, { "epoch": 2.2940156114483954, "grad_norm": 0.2520439028739929, "learning_rate": 2.3618163354127378e-05, "loss": 0.0423, "step": 21160 }, { "epoch": 2.2941240242844754, "grad_norm": 0.1711507886648178, "learning_rate": 2.361453648629044e-05, "loss": 0.016, "step": 21161 }, { "epoch": 2.294232437120555, "grad_norm": 0.1283523440361023, "learning_rate": 2.3610909618453506e-05, "loss": 0.0067, "step": 21162 }, { "epoch": 2.294340849956635, "grad_norm": 0.5640002489089966, "learning_rate": 2.360728275061657e-05, "loss": 0.0117, "step": 21163 }, { "epoch": 2.294449262792715, "grad_norm": 0.0991310328245163, "learning_rate": 2.360365588277963e-05, "loss": 0.0071, "step": 21164 }, { "epoch": 2.2945576756287944, "grad_norm": 0.03655029088258743, "learning_rate": 2.3600029014942695e-05, "loss": 0.0009, "step": 21165 }, { "epoch": 2.2946660884648744, "grad_norm": 0.5190181136131287, "learning_rate": 2.3596402147105762e-05, "loss": 0.0125, "step": 21166 }, { "epoch": 2.294774501300954, "grad_norm": 0.21401363611221313, "learning_rate": 2.3592775279268827e-05, "loss": 0.0076, "step": 21167 }, { "epoch": 2.294882914137034, "grad_norm": 0.5255025625228882, "learning_rate": 2.3589148411431887e-05, "loss": 0.012, "step": 21168 }, { "epoch": 2.2949913269731135, "grad_norm": 0.5589965581893921, "learning_rate": 2.358552154359495e-05, "loss": 0.0247, "step": 21169 }, { "epoch": 2.2950997398091935, "grad_norm": 0.3185420036315918, "learning_rate": 2.358189467575802e-05, "loss": 0.0542, "step": 21170 }, { "epoch": 2.295208152645273, "grad_norm": 0.6408085227012634, "learning_rate": 2.357826780792108e-05, "loss": 0.0367, "step": 21171 }, { "epoch": 2.295316565481353, "grad_norm": 0.18944697082042694, "learning_rate": 2.3574640940084144e-05, "loss": 0.0132, "step": 21172 }, { "epoch": 2.295424978317433, "grad_norm": 0.4657552242279053, "learning_rate": 2.3571014072247208e-05, "loss": 0.0419, "step": 21173 }, { "epoch": 2.2955333911535125, "grad_norm": 0.33513540029525757, "learning_rate": 2.3567387204410272e-05, "loss": 0.0069, "step": 21174 }, { "epoch": 2.2956418039895925, "grad_norm": 0.0983128771185875, "learning_rate": 2.3563760336573336e-05, "loss": 0.0017, "step": 21175 }, { "epoch": 2.295750216825672, "grad_norm": 0.2778994143009186, "learning_rate": 2.35601334687364e-05, "loss": 0.0096, "step": 21176 }, { "epoch": 2.295858629661752, "grad_norm": 0.09138496965169907, "learning_rate": 2.3556506600899464e-05, "loss": 0.0029, "step": 21177 }, { "epoch": 2.2959670424978316, "grad_norm": 0.1874014139175415, "learning_rate": 2.3552879733062528e-05, "loss": 0.0029, "step": 21178 }, { "epoch": 2.2960754553339116, "grad_norm": 0.5745845437049866, "learning_rate": 2.3549252865225592e-05, "loss": 0.0303, "step": 21179 }, { "epoch": 2.296183868169991, "grad_norm": 0.4624638855457306, "learning_rate": 2.3545625997388657e-05, "loss": 0.008, "step": 21180 }, { "epoch": 2.296292281006071, "grad_norm": 0.17645522952079773, "learning_rate": 2.354199912955172e-05, "loss": 0.0065, "step": 21181 }, { "epoch": 2.296400693842151, "grad_norm": 0.354830801486969, "learning_rate": 2.3538372261714785e-05, "loss": 0.0194, "step": 21182 }, { "epoch": 2.2965091066782306, "grad_norm": 0.17208246886730194, "learning_rate": 2.353474539387785e-05, "loss": 0.0022, "step": 21183 }, { "epoch": 2.2966175195143106, "grad_norm": 0.0803866982460022, "learning_rate": 2.3531118526040913e-05, "loss": 0.0062, "step": 21184 }, { "epoch": 2.29672593235039, "grad_norm": 0.04704241827130318, "learning_rate": 2.3527491658203977e-05, "loss": 0.0027, "step": 21185 }, { "epoch": 2.29683434518647, "grad_norm": 0.050512418150901794, "learning_rate": 2.352386479036704e-05, "loss": 0.0017, "step": 21186 }, { "epoch": 2.29694275802255, "grad_norm": 0.29713937640190125, "learning_rate": 2.3520237922530102e-05, "loss": 0.0069, "step": 21187 }, { "epoch": 2.2970511708586296, "grad_norm": 0.15609154105186462, "learning_rate": 2.351661105469317e-05, "loss": 0.0026, "step": 21188 }, { "epoch": 2.2971595836947096, "grad_norm": 0.2085496485233307, "learning_rate": 2.3512984186856233e-05, "loss": 0.0046, "step": 21189 }, { "epoch": 2.297267996530789, "grad_norm": 0.11280825734138489, "learning_rate": 2.3509357319019294e-05, "loss": 0.0027, "step": 21190 }, { "epoch": 2.297376409366869, "grad_norm": 0.49431928992271423, "learning_rate": 2.3505730451182358e-05, "loss": 0.0248, "step": 21191 }, { "epoch": 2.2974848222029487, "grad_norm": 1.0906801223754883, "learning_rate": 2.3502103583345426e-05, "loss": 0.0477, "step": 21192 }, { "epoch": 2.2975932350390287, "grad_norm": 0.1001160740852356, "learning_rate": 2.349847671550849e-05, "loss": 0.0044, "step": 21193 }, { "epoch": 2.297701647875108, "grad_norm": 0.2051565945148468, "learning_rate": 2.349484984767155e-05, "loss": 0.0041, "step": 21194 }, { "epoch": 2.297810060711188, "grad_norm": 0.1092207208275795, "learning_rate": 2.3491222979834615e-05, "loss": 0.0049, "step": 21195 }, { "epoch": 2.297918473547268, "grad_norm": 0.18097448348999023, "learning_rate": 2.3487596111997682e-05, "loss": 0.0036, "step": 21196 }, { "epoch": 2.2980268863833477, "grad_norm": 0.09170278906822205, "learning_rate": 2.3483969244160743e-05, "loss": 0.0023, "step": 21197 }, { "epoch": 2.2981352992194277, "grad_norm": 0.0780690535902977, "learning_rate": 2.3480342376323807e-05, "loss": 0.003, "step": 21198 }, { "epoch": 2.2982437120555073, "grad_norm": 0.06634145230054855, "learning_rate": 2.347671550848687e-05, "loss": 0.0037, "step": 21199 }, { "epoch": 2.2983521248915872, "grad_norm": 0.2525270879268646, "learning_rate": 2.3473088640649935e-05, "loss": 0.0097, "step": 21200 }, { "epoch": 2.298460537727667, "grad_norm": 0.10882757604122162, "learning_rate": 2.3469461772813e-05, "loss": 0.0028, "step": 21201 }, { "epoch": 2.2985689505637468, "grad_norm": 0.3499370217323303, "learning_rate": 2.3465834904976063e-05, "loss": 0.0166, "step": 21202 }, { "epoch": 2.2986773633998263, "grad_norm": 0.07039009034633636, "learning_rate": 2.3462208037139128e-05, "loss": 0.0048, "step": 21203 }, { "epoch": 2.2987857762359063, "grad_norm": 0.12859739363193512, "learning_rate": 2.345858116930219e-05, "loss": 0.0062, "step": 21204 }, { "epoch": 2.2988941890719863, "grad_norm": 0.019190339371562004, "learning_rate": 2.3454954301465256e-05, "loss": 0.0009, "step": 21205 }, { "epoch": 2.299002601908066, "grad_norm": 0.6634999513626099, "learning_rate": 2.345132743362832e-05, "loss": 0.019, "step": 21206 }, { "epoch": 2.299111014744146, "grad_norm": 0.24103862047195435, "learning_rate": 2.3447700565791384e-05, "loss": 0.04, "step": 21207 }, { "epoch": 2.2992194275802254, "grad_norm": 0.2627263069152832, "learning_rate": 2.3444073697954448e-05, "loss": 0.0051, "step": 21208 }, { "epoch": 2.2993278404163053, "grad_norm": 0.12343274801969528, "learning_rate": 2.3440446830117512e-05, "loss": 0.0044, "step": 21209 }, { "epoch": 2.2994362532523853, "grad_norm": 0.031923722475767136, "learning_rate": 2.3436819962280576e-05, "loss": 0.0012, "step": 21210 }, { "epoch": 2.299544666088465, "grad_norm": 0.6427035927772522, "learning_rate": 2.343319309444364e-05, "loss": 0.0461, "step": 21211 }, { "epoch": 2.299653078924545, "grad_norm": 0.14703530073165894, "learning_rate": 2.3429566226606704e-05, "loss": 0.0037, "step": 21212 }, { "epoch": 2.2997614917606244, "grad_norm": 0.2758888304233551, "learning_rate": 2.3425939358769765e-05, "loss": 0.0137, "step": 21213 }, { "epoch": 2.2998699045967044, "grad_norm": 0.1456880122423172, "learning_rate": 2.3422312490932833e-05, "loss": 0.0098, "step": 21214 }, { "epoch": 2.299978317432784, "grad_norm": 0.09188881516456604, "learning_rate": 2.3418685623095897e-05, "loss": 0.0037, "step": 21215 }, { "epoch": 2.300086730268864, "grad_norm": 0.05071069672703743, "learning_rate": 2.3415058755258958e-05, "loss": 0.0009, "step": 21216 }, { "epoch": 2.3001951431049434, "grad_norm": 0.38292399048805237, "learning_rate": 2.341143188742202e-05, "loss": 0.0137, "step": 21217 }, { "epoch": 2.3003035559410234, "grad_norm": 0.25339508056640625, "learning_rate": 2.340780501958509e-05, "loss": 0.0127, "step": 21218 }, { "epoch": 2.3004119687771034, "grad_norm": 0.11810626089572906, "learning_rate": 2.3404178151748153e-05, "loss": 0.0079, "step": 21219 }, { "epoch": 2.300520381613183, "grad_norm": 0.36724913120269775, "learning_rate": 2.3400551283911214e-05, "loss": 0.0107, "step": 21220 }, { "epoch": 2.300628794449263, "grad_norm": 0.04490507021546364, "learning_rate": 2.3396924416074278e-05, "loss": 0.0018, "step": 21221 }, { "epoch": 2.3007372072853425, "grad_norm": 0.2899623215198517, "learning_rate": 2.3393297548237346e-05, "loss": 0.0023, "step": 21222 }, { "epoch": 2.3008456201214225, "grad_norm": 0.06343154609203339, "learning_rate": 2.3389670680400406e-05, "loss": 0.0026, "step": 21223 }, { "epoch": 2.300954032957502, "grad_norm": 0.04101072624325752, "learning_rate": 2.338604381256347e-05, "loss": 0.0016, "step": 21224 }, { "epoch": 2.301062445793582, "grad_norm": 0.04477531462907791, "learning_rate": 2.3382416944726534e-05, "loss": 0.001, "step": 21225 }, { "epoch": 2.3011708586296615, "grad_norm": 0.9488735198974609, "learning_rate": 2.33787900768896e-05, "loss": 0.0134, "step": 21226 }, { "epoch": 2.3012792714657415, "grad_norm": 0.47543466091156006, "learning_rate": 2.3375163209052663e-05, "loss": 0.027, "step": 21227 }, { "epoch": 2.3013876843018215, "grad_norm": 0.15642575919628143, "learning_rate": 2.3371536341215727e-05, "loss": 0.0076, "step": 21228 }, { "epoch": 2.301496097137901, "grad_norm": 0.9653103351593018, "learning_rate": 2.336790947337879e-05, "loss": 0.0074, "step": 21229 }, { "epoch": 2.301604509973981, "grad_norm": 0.1154787465929985, "learning_rate": 2.3364282605541855e-05, "loss": 0.0062, "step": 21230 }, { "epoch": 2.3017129228100606, "grad_norm": 0.223423570394516, "learning_rate": 2.336065573770492e-05, "loss": 0.0106, "step": 21231 }, { "epoch": 2.3018213356461406, "grad_norm": 0.09639803320169449, "learning_rate": 2.3357028869867983e-05, "loss": 0.0016, "step": 21232 }, { "epoch": 2.3019297484822205, "grad_norm": 0.8752520680427551, "learning_rate": 2.3353402002031047e-05, "loss": 0.0517, "step": 21233 }, { "epoch": 2.3020381613183, "grad_norm": 0.3657379448413849, "learning_rate": 2.334977513419411e-05, "loss": 0.006, "step": 21234 }, { "epoch": 2.30214657415438, "grad_norm": 0.2226465344429016, "learning_rate": 2.3346148266357176e-05, "loss": 0.0085, "step": 21235 }, { "epoch": 2.3022549869904596, "grad_norm": 0.5393344759941101, "learning_rate": 2.334252139852024e-05, "loss": 0.0177, "step": 21236 }, { "epoch": 2.3023633998265396, "grad_norm": 0.04338093474507332, "learning_rate": 2.3338894530683304e-05, "loss": 0.0008, "step": 21237 }, { "epoch": 2.302471812662619, "grad_norm": 0.08377570658922195, "learning_rate": 2.3335267662846368e-05, "loss": 0.0028, "step": 21238 }, { "epoch": 2.302580225498699, "grad_norm": 0.23354938626289368, "learning_rate": 2.3331640795009432e-05, "loss": 0.009, "step": 21239 }, { "epoch": 2.3026886383347787, "grad_norm": 0.22892728447914124, "learning_rate": 2.3328013927172496e-05, "loss": 0.0427, "step": 21240 }, { "epoch": 2.3027970511708586, "grad_norm": 0.22142207622528076, "learning_rate": 2.332438705933556e-05, "loss": 0.0472, "step": 21241 }, { "epoch": 2.3029054640069386, "grad_norm": 0.07374557107686996, "learning_rate": 2.332076019149862e-05, "loss": 0.003, "step": 21242 }, { "epoch": 2.303013876843018, "grad_norm": 0.021653957664966583, "learning_rate": 2.3317133323661685e-05, "loss": 0.0008, "step": 21243 }, { "epoch": 2.303122289679098, "grad_norm": 0.11771643161773682, "learning_rate": 2.3313506455824752e-05, "loss": 0.0047, "step": 21244 }, { "epoch": 2.3032307025151777, "grad_norm": 0.14329661428928375, "learning_rate": 2.3309879587987817e-05, "loss": 0.0049, "step": 21245 }, { "epoch": 2.3033391153512577, "grad_norm": 0.0779193565249443, "learning_rate": 2.3306252720150877e-05, "loss": 0.0025, "step": 21246 }, { "epoch": 2.3034475281873372, "grad_norm": 1.0611926317214966, "learning_rate": 2.330262585231394e-05, "loss": 0.0486, "step": 21247 }, { "epoch": 2.303555941023417, "grad_norm": 0.19918596744537354, "learning_rate": 2.329899898447701e-05, "loss": 0.0054, "step": 21248 }, { "epoch": 2.3036643538594967, "grad_norm": 0.04410098120570183, "learning_rate": 2.329537211664007e-05, "loss": 0.0022, "step": 21249 }, { "epoch": 2.3037727666955767, "grad_norm": 0.41771984100341797, "learning_rate": 2.3291745248803134e-05, "loss": 0.0164, "step": 21250 }, { "epoch": 2.3038811795316567, "grad_norm": 0.15952861309051514, "learning_rate": 2.3288118380966198e-05, "loss": 0.0097, "step": 21251 }, { "epoch": 2.3039895923677363, "grad_norm": 0.03380357101559639, "learning_rate": 2.3284491513129265e-05, "loss": 0.0006, "step": 21252 }, { "epoch": 2.3040980052038162, "grad_norm": 0.2107371836900711, "learning_rate": 2.3280864645292326e-05, "loss": 0.0038, "step": 21253 }, { "epoch": 2.304206418039896, "grad_norm": 0.6440184712409973, "learning_rate": 2.327723777745539e-05, "loss": 0.0554, "step": 21254 }, { "epoch": 2.3043148308759758, "grad_norm": 0.0851806029677391, "learning_rate": 2.3273610909618454e-05, "loss": 0.0028, "step": 21255 }, { "epoch": 2.3044232437120553, "grad_norm": 0.3835139274597168, "learning_rate": 2.326998404178152e-05, "loss": 0.011, "step": 21256 }, { "epoch": 2.3045316565481353, "grad_norm": 0.11967403441667557, "learning_rate": 2.3266357173944582e-05, "loss": 0.0041, "step": 21257 }, { "epoch": 2.3046400693842153, "grad_norm": 0.025734109804034233, "learning_rate": 2.3262730306107647e-05, "loss": 0.0007, "step": 21258 }, { "epoch": 2.304748482220295, "grad_norm": 0.14934265613555908, "learning_rate": 2.325910343827071e-05, "loss": 0.0041, "step": 21259 }, { "epoch": 2.304856895056375, "grad_norm": 0.5907986164093018, "learning_rate": 2.3255476570433775e-05, "loss": 0.02, "step": 21260 }, { "epoch": 2.3049653078924544, "grad_norm": 0.5130912661552429, "learning_rate": 2.325184970259684e-05, "loss": 0.0363, "step": 21261 }, { "epoch": 2.3050737207285343, "grad_norm": 0.5536184906959534, "learning_rate": 2.3248222834759903e-05, "loss": 0.029, "step": 21262 }, { "epoch": 2.305182133564614, "grad_norm": 0.0868314802646637, "learning_rate": 2.3244595966922967e-05, "loss": 0.0051, "step": 21263 }, { "epoch": 2.305290546400694, "grad_norm": 0.8509822487831116, "learning_rate": 2.324096909908603e-05, "loss": 0.029, "step": 21264 }, { "epoch": 2.305398959236774, "grad_norm": 0.14200210571289062, "learning_rate": 2.3237342231249095e-05, "loss": 0.0098, "step": 21265 }, { "epoch": 2.3055073720728534, "grad_norm": 0.083480603992939, "learning_rate": 2.323371536341216e-05, "loss": 0.0021, "step": 21266 }, { "epoch": 2.3056157849089334, "grad_norm": 0.235914945602417, "learning_rate": 2.3230088495575223e-05, "loss": 0.0059, "step": 21267 }, { "epoch": 2.305724197745013, "grad_norm": 0.07556386291980743, "learning_rate": 2.3226461627738284e-05, "loss": 0.0018, "step": 21268 }, { "epoch": 2.305832610581093, "grad_norm": 0.1543133705854416, "learning_rate": 2.322283475990135e-05, "loss": 0.0052, "step": 21269 }, { "epoch": 2.3059410234171724, "grad_norm": 1.0944077968597412, "learning_rate": 2.3219207892064416e-05, "loss": 0.0222, "step": 21270 }, { "epoch": 2.3060494362532524, "grad_norm": 0.26512858271598816, "learning_rate": 2.321558102422748e-05, "loss": 0.0132, "step": 21271 }, { "epoch": 2.306157849089332, "grad_norm": 0.05546777322888374, "learning_rate": 2.321195415639054e-05, "loss": 0.002, "step": 21272 }, { "epoch": 2.306266261925412, "grad_norm": 0.603508472442627, "learning_rate": 2.3208327288553605e-05, "loss": 0.013, "step": 21273 }, { "epoch": 2.306374674761492, "grad_norm": 0.09436092525720596, "learning_rate": 2.3204700420716672e-05, "loss": 0.0059, "step": 21274 }, { "epoch": 2.3064830875975715, "grad_norm": 0.430202841758728, "learning_rate": 2.3201073552879733e-05, "loss": 0.0394, "step": 21275 }, { "epoch": 2.3065915004336515, "grad_norm": 0.2524537444114685, "learning_rate": 2.3197446685042797e-05, "loss": 0.0372, "step": 21276 }, { "epoch": 2.306699913269731, "grad_norm": 0.05777370184659958, "learning_rate": 2.319381981720586e-05, "loss": 0.0016, "step": 21277 }, { "epoch": 2.306808326105811, "grad_norm": 0.289930522441864, "learning_rate": 2.319019294936893e-05, "loss": 0.0082, "step": 21278 }, { "epoch": 2.3069167389418905, "grad_norm": 0.02794354222714901, "learning_rate": 2.318656608153199e-05, "loss": 0.0014, "step": 21279 }, { "epoch": 2.3070251517779705, "grad_norm": 0.056028448045253754, "learning_rate": 2.3182939213695053e-05, "loss": 0.002, "step": 21280 }, { "epoch": 2.30713356461405, "grad_norm": 0.6380273699760437, "learning_rate": 2.3179312345858118e-05, "loss": 0.02, "step": 21281 }, { "epoch": 2.30724197745013, "grad_norm": 0.70786052942276, "learning_rate": 2.317568547802118e-05, "loss": 0.0226, "step": 21282 }, { "epoch": 2.30735039028621, "grad_norm": 0.14380356669425964, "learning_rate": 2.3172058610184246e-05, "loss": 0.0035, "step": 21283 }, { "epoch": 2.3074588031222896, "grad_norm": 0.49502089619636536, "learning_rate": 2.316843174234731e-05, "loss": 0.0139, "step": 21284 }, { "epoch": 2.3075672159583696, "grad_norm": 0.31940600275993347, "learning_rate": 2.3164804874510374e-05, "loss": 0.0224, "step": 21285 }, { "epoch": 2.307675628794449, "grad_norm": 0.14960874617099762, "learning_rate": 2.3161178006673438e-05, "loss": 0.0075, "step": 21286 }, { "epoch": 2.307784041630529, "grad_norm": 0.016359275206923485, "learning_rate": 2.3157551138836502e-05, "loss": 0.0009, "step": 21287 }, { "epoch": 2.307892454466609, "grad_norm": 0.24869437515735626, "learning_rate": 2.3153924270999566e-05, "loss": 0.0104, "step": 21288 }, { "epoch": 2.3080008673026886, "grad_norm": 0.14590191841125488, "learning_rate": 2.315029740316263e-05, "loss": 0.0133, "step": 21289 }, { "epoch": 2.3081092801387686, "grad_norm": 0.7570198774337769, "learning_rate": 2.3146670535325695e-05, "loss": 0.0114, "step": 21290 }, { "epoch": 2.308217692974848, "grad_norm": 0.05739245563745499, "learning_rate": 2.314304366748876e-05, "loss": 0.0017, "step": 21291 }, { "epoch": 2.308326105810928, "grad_norm": 0.43752190470695496, "learning_rate": 2.3139416799651823e-05, "loss": 0.0262, "step": 21292 }, { "epoch": 2.3084345186470077, "grad_norm": 0.3883049488067627, "learning_rate": 2.3135789931814887e-05, "loss": 0.0185, "step": 21293 }, { "epoch": 2.3085429314830876, "grad_norm": 0.42677098512649536, "learning_rate": 2.3132163063977948e-05, "loss": 0.0239, "step": 21294 }, { "epoch": 2.308651344319167, "grad_norm": 0.04287644848227501, "learning_rate": 2.312853619614101e-05, "loss": 0.0018, "step": 21295 }, { "epoch": 2.308759757155247, "grad_norm": 0.2974546253681183, "learning_rate": 2.312490932830408e-05, "loss": 0.0114, "step": 21296 }, { "epoch": 2.308868169991327, "grad_norm": 0.5498813986778259, "learning_rate": 2.3121282460467143e-05, "loss": 0.0062, "step": 21297 }, { "epoch": 2.3089765828274067, "grad_norm": 0.20326553285121918, "learning_rate": 2.3117655592630204e-05, "loss": 0.008, "step": 21298 }, { "epoch": 2.3090849956634867, "grad_norm": 0.6973010301589966, "learning_rate": 2.3114028724793268e-05, "loss": 0.0239, "step": 21299 }, { "epoch": 2.3091934084995662, "grad_norm": 0.17231757938861847, "learning_rate": 2.3110401856956336e-05, "loss": 0.0015, "step": 21300 }, { "epoch": 2.309301821335646, "grad_norm": 0.40789341926574707, "learning_rate": 2.3106774989119396e-05, "loss": 0.0137, "step": 21301 }, { "epoch": 2.3094102341717258, "grad_norm": 0.2048458307981491, "learning_rate": 2.310314812128246e-05, "loss": 0.0259, "step": 21302 }, { "epoch": 2.3095186470078057, "grad_norm": 0.08937451243400574, "learning_rate": 2.3099521253445524e-05, "loss": 0.0041, "step": 21303 }, { "epoch": 2.3096270598438853, "grad_norm": 0.4549413025379181, "learning_rate": 2.3095894385608592e-05, "loss": 0.0356, "step": 21304 }, { "epoch": 2.3097354726799653, "grad_norm": 0.10901233553886414, "learning_rate": 2.3092267517771653e-05, "loss": 0.0058, "step": 21305 }, { "epoch": 2.3098438855160452, "grad_norm": 0.4807925224304199, "learning_rate": 2.3088640649934717e-05, "loss": 0.0092, "step": 21306 }, { "epoch": 2.309952298352125, "grad_norm": 0.19297419488430023, "learning_rate": 2.308501378209778e-05, "loss": 0.0133, "step": 21307 }, { "epoch": 2.3100607111882048, "grad_norm": 0.07286485284566879, "learning_rate": 2.3081386914260845e-05, "loss": 0.0022, "step": 21308 }, { "epoch": 2.3101691240242843, "grad_norm": 0.20948906242847443, "learning_rate": 2.307776004642391e-05, "loss": 0.0047, "step": 21309 }, { "epoch": 2.3102775368603643, "grad_norm": 0.12888211011886597, "learning_rate": 2.3074133178586973e-05, "loss": 0.0027, "step": 21310 }, { "epoch": 2.3103859496964443, "grad_norm": 0.17250202596187592, "learning_rate": 2.3070506310750037e-05, "loss": 0.0055, "step": 21311 }, { "epoch": 2.310494362532524, "grad_norm": 0.0554269440472126, "learning_rate": 2.30668794429131e-05, "loss": 0.0022, "step": 21312 }, { "epoch": 2.310602775368604, "grad_norm": 0.1271011233329773, "learning_rate": 2.3063252575076166e-05, "loss": 0.0044, "step": 21313 }, { "epoch": 2.3107111882046834, "grad_norm": 0.2177571803331375, "learning_rate": 2.305962570723923e-05, "loss": 0.0053, "step": 21314 }, { "epoch": 2.3108196010407633, "grad_norm": 0.8582581877708435, "learning_rate": 2.3055998839402294e-05, "loss": 0.0361, "step": 21315 }, { "epoch": 2.310928013876843, "grad_norm": 0.48260629177093506, "learning_rate": 2.3052371971565358e-05, "loss": 0.041, "step": 21316 }, { "epoch": 2.311036426712923, "grad_norm": 0.030521510168910027, "learning_rate": 2.3048745103728422e-05, "loss": 0.0018, "step": 21317 }, { "epoch": 2.3111448395490024, "grad_norm": 0.5604192018508911, "learning_rate": 2.3045118235891486e-05, "loss": 0.0281, "step": 21318 }, { "epoch": 2.3112532523850824, "grad_norm": 0.04975352808833122, "learning_rate": 2.304149136805455e-05, "loss": 0.0026, "step": 21319 }, { "epoch": 2.3113616652211624, "grad_norm": 0.22641094028949738, "learning_rate": 2.303786450021761e-05, "loss": 0.0108, "step": 21320 }, { "epoch": 2.311470078057242, "grad_norm": 0.016105197370052338, "learning_rate": 2.3034237632380675e-05, "loss": 0.0004, "step": 21321 }, { "epoch": 2.311578490893322, "grad_norm": 0.2813414931297302, "learning_rate": 2.3030610764543742e-05, "loss": 0.0084, "step": 21322 }, { "epoch": 2.3116869037294014, "grad_norm": 0.16627222299575806, "learning_rate": 2.3026983896706807e-05, "loss": 0.0065, "step": 21323 }, { "epoch": 2.3117953165654814, "grad_norm": 0.15373395383358002, "learning_rate": 2.3023357028869867e-05, "loss": 0.0045, "step": 21324 }, { "epoch": 2.311903729401561, "grad_norm": 0.08669885247945786, "learning_rate": 2.301973016103293e-05, "loss": 0.0019, "step": 21325 }, { "epoch": 2.312012142237641, "grad_norm": 0.4085577428340912, "learning_rate": 2.3016103293196e-05, "loss": 0.054, "step": 21326 }, { "epoch": 2.3121205550737205, "grad_norm": 0.2428598552942276, "learning_rate": 2.3012476425359063e-05, "loss": 0.0159, "step": 21327 }, { "epoch": 2.3122289679098005, "grad_norm": 0.06831461191177368, "learning_rate": 2.3008849557522124e-05, "loss": 0.0021, "step": 21328 }, { "epoch": 2.3123373807458805, "grad_norm": 0.7272937893867493, "learning_rate": 2.3005222689685188e-05, "loss": 0.0157, "step": 21329 }, { "epoch": 2.31244579358196, "grad_norm": 0.3783957064151764, "learning_rate": 2.3001595821848255e-05, "loss": 0.0305, "step": 21330 }, { "epoch": 2.31255420641804, "grad_norm": 0.9134128093719482, "learning_rate": 2.2997968954011316e-05, "loss": 0.0535, "step": 21331 }, { "epoch": 2.3126626192541195, "grad_norm": 0.23863142728805542, "learning_rate": 2.299434208617438e-05, "loss": 0.0049, "step": 21332 }, { "epoch": 2.3127710320901995, "grad_norm": 0.04070305824279785, "learning_rate": 2.2990715218337444e-05, "loss": 0.001, "step": 21333 }, { "epoch": 2.3128794449262795, "grad_norm": 0.45296090841293335, "learning_rate": 2.2987088350500512e-05, "loss": 0.0127, "step": 21334 }, { "epoch": 2.312987857762359, "grad_norm": 1.5146554708480835, "learning_rate": 2.2983461482663572e-05, "loss": 0.0311, "step": 21335 }, { "epoch": 2.313096270598439, "grad_norm": 0.015926038846373558, "learning_rate": 2.2979834614826637e-05, "loss": 0.0005, "step": 21336 }, { "epoch": 2.3132046834345186, "grad_norm": 1.1259169578552246, "learning_rate": 2.29762077469897e-05, "loss": 0.0131, "step": 21337 }, { "epoch": 2.3133130962705986, "grad_norm": 0.3024592399597168, "learning_rate": 2.2972580879152765e-05, "loss": 0.0123, "step": 21338 }, { "epoch": 2.313421509106678, "grad_norm": 0.27829885482788086, "learning_rate": 2.296895401131583e-05, "loss": 0.0196, "step": 21339 }, { "epoch": 2.313529921942758, "grad_norm": 0.06275855004787445, "learning_rate": 2.2965327143478893e-05, "loss": 0.0029, "step": 21340 }, { "epoch": 2.3136383347788376, "grad_norm": 0.1906346082687378, "learning_rate": 2.2961700275641957e-05, "loss": 0.0072, "step": 21341 }, { "epoch": 2.3137467476149176, "grad_norm": 0.025412961840629578, "learning_rate": 2.295807340780502e-05, "loss": 0.0011, "step": 21342 }, { "epoch": 2.3138551604509976, "grad_norm": 0.10657743364572525, "learning_rate": 2.2954446539968085e-05, "loss": 0.0032, "step": 21343 }, { "epoch": 2.313963573287077, "grad_norm": 0.04830821603536606, "learning_rate": 2.295081967213115e-05, "loss": 0.001, "step": 21344 }, { "epoch": 2.314071986123157, "grad_norm": 0.09221100062131882, "learning_rate": 2.2947192804294214e-05, "loss": 0.0026, "step": 21345 }, { "epoch": 2.3141803989592367, "grad_norm": 0.21518781781196594, "learning_rate": 2.2943565936457274e-05, "loss": 0.0148, "step": 21346 }, { "epoch": 2.3142888117953166, "grad_norm": 0.08000915497541428, "learning_rate": 2.2939939068620342e-05, "loss": 0.0033, "step": 21347 }, { "epoch": 2.314397224631396, "grad_norm": 0.04852323606610298, "learning_rate": 2.2936312200783406e-05, "loss": 0.0019, "step": 21348 }, { "epoch": 2.314505637467476, "grad_norm": 0.019097773358225822, "learning_rate": 2.293268533294647e-05, "loss": 0.0007, "step": 21349 }, { "epoch": 2.3146140503035557, "grad_norm": 0.1127798780798912, "learning_rate": 2.292905846510953e-05, "loss": 0.0037, "step": 21350 }, { "epoch": 2.3147224631396357, "grad_norm": 0.22803384065628052, "learning_rate": 2.2925431597272595e-05, "loss": 0.0054, "step": 21351 }, { "epoch": 2.3148308759757157, "grad_norm": 0.16405561566352844, "learning_rate": 2.2921804729435662e-05, "loss": 0.0047, "step": 21352 }, { "epoch": 2.3149392888117952, "grad_norm": 0.03067697025835514, "learning_rate": 2.2918177861598726e-05, "loss": 0.0009, "step": 21353 }, { "epoch": 2.315047701647875, "grad_norm": 0.3645455241203308, "learning_rate": 2.2914550993761787e-05, "loss": 0.0084, "step": 21354 }, { "epoch": 2.3151561144839548, "grad_norm": 0.5631335377693176, "learning_rate": 2.291092412592485e-05, "loss": 0.0161, "step": 21355 }, { "epoch": 2.3152645273200347, "grad_norm": 0.3451363146305084, "learning_rate": 2.290729725808792e-05, "loss": 0.0196, "step": 21356 }, { "epoch": 2.3153729401561147, "grad_norm": 0.7495381832122803, "learning_rate": 2.290367039025098e-05, "loss": 0.0084, "step": 21357 }, { "epoch": 2.3154813529921943, "grad_norm": 0.18342657387256622, "learning_rate": 2.2900043522414044e-05, "loss": 0.0016, "step": 21358 }, { "epoch": 2.3155897658282742, "grad_norm": 0.26360777020454407, "learning_rate": 2.2896416654577108e-05, "loss": 0.0082, "step": 21359 }, { "epoch": 2.315698178664354, "grad_norm": 0.5583156943321228, "learning_rate": 2.2892789786740175e-05, "loss": 0.042, "step": 21360 }, { "epoch": 2.3158065915004338, "grad_norm": 0.3674461245536804, "learning_rate": 2.2889162918903236e-05, "loss": 0.0126, "step": 21361 }, { "epoch": 2.3159150043365133, "grad_norm": 0.3156927227973938, "learning_rate": 2.28855360510663e-05, "loss": 0.0161, "step": 21362 }, { "epoch": 2.3160234171725933, "grad_norm": 0.2757118344306946, "learning_rate": 2.2881909183229364e-05, "loss": 0.0103, "step": 21363 }, { "epoch": 2.316131830008673, "grad_norm": 0.3278678059577942, "learning_rate": 2.2878282315392428e-05, "loss": 0.0172, "step": 21364 }, { "epoch": 2.316240242844753, "grad_norm": 0.003558847587555647, "learning_rate": 2.2874655447555492e-05, "loss": 0.0001, "step": 21365 }, { "epoch": 2.316348655680833, "grad_norm": 0.45061302185058594, "learning_rate": 2.2871028579718556e-05, "loss": 0.0433, "step": 21366 }, { "epoch": 2.3164570685169124, "grad_norm": 0.022328870370984077, "learning_rate": 2.286740171188162e-05, "loss": 0.0007, "step": 21367 }, { "epoch": 2.3165654813529923, "grad_norm": 0.46817654371261597, "learning_rate": 2.2863774844044685e-05, "loss": 0.0197, "step": 21368 }, { "epoch": 2.316673894189072, "grad_norm": 0.2538590431213379, "learning_rate": 2.286014797620775e-05, "loss": 0.0249, "step": 21369 }, { "epoch": 2.316782307025152, "grad_norm": 0.2355223447084427, "learning_rate": 2.2856521108370813e-05, "loss": 0.0106, "step": 21370 }, { "epoch": 2.3168907198612314, "grad_norm": 0.08649566024541855, "learning_rate": 2.2852894240533877e-05, "loss": 0.0018, "step": 21371 }, { "epoch": 2.3169991326973114, "grad_norm": 0.5090131163597107, "learning_rate": 2.2849267372696938e-05, "loss": 0.011, "step": 21372 }, { "epoch": 2.317107545533391, "grad_norm": 0.04627629369497299, "learning_rate": 2.2845640504860005e-05, "loss": 0.0013, "step": 21373 }, { "epoch": 2.317215958369471, "grad_norm": 0.17196479439735413, "learning_rate": 2.284201363702307e-05, "loss": 0.0047, "step": 21374 }, { "epoch": 2.317324371205551, "grad_norm": 0.029393184930086136, "learning_rate": 2.2838386769186133e-05, "loss": 0.0017, "step": 21375 }, { "epoch": 2.3174327840416304, "grad_norm": 0.2113654911518097, "learning_rate": 2.2834759901349194e-05, "loss": 0.0112, "step": 21376 }, { "epoch": 2.3175411968777104, "grad_norm": 0.09751223772764206, "learning_rate": 2.2831133033512258e-05, "loss": 0.0012, "step": 21377 }, { "epoch": 2.31764960971379, "grad_norm": 0.4742652475833893, "learning_rate": 2.2827506165675326e-05, "loss": 0.0073, "step": 21378 }, { "epoch": 2.31775802254987, "grad_norm": 0.09961172938346863, "learning_rate": 2.282387929783839e-05, "loss": 0.0051, "step": 21379 }, { "epoch": 2.31786643538595, "grad_norm": 0.02458449825644493, "learning_rate": 2.282025243000145e-05, "loss": 0.0008, "step": 21380 }, { "epoch": 2.3179748482220295, "grad_norm": 0.3562353551387787, "learning_rate": 2.2816625562164515e-05, "loss": 0.0108, "step": 21381 }, { "epoch": 2.3180832610581095, "grad_norm": 1.1214821338653564, "learning_rate": 2.2812998694327582e-05, "loss": 0.0164, "step": 21382 }, { "epoch": 2.318191673894189, "grad_norm": 0.15868914127349854, "learning_rate": 2.2809371826490643e-05, "loss": 0.0554, "step": 21383 }, { "epoch": 2.318300086730269, "grad_norm": 0.18598125874996185, "learning_rate": 2.2805744958653707e-05, "loss": 0.0086, "step": 21384 }, { "epoch": 2.3184084995663485, "grad_norm": 0.2333780825138092, "learning_rate": 2.280211809081677e-05, "loss": 0.0243, "step": 21385 }, { "epoch": 2.3185169124024285, "grad_norm": 0.09314814954996109, "learning_rate": 2.279849122297984e-05, "loss": 0.0043, "step": 21386 }, { "epoch": 2.318625325238508, "grad_norm": 0.627099871635437, "learning_rate": 2.27948643551429e-05, "loss": 0.0078, "step": 21387 }, { "epoch": 2.318733738074588, "grad_norm": 0.6408283114433289, "learning_rate": 2.2791237487305963e-05, "loss": 0.0191, "step": 21388 }, { "epoch": 2.318842150910668, "grad_norm": 0.6736798286437988, "learning_rate": 2.2787610619469027e-05, "loss": 0.0364, "step": 21389 }, { "epoch": 2.3189505637467476, "grad_norm": 0.19599096477031708, "learning_rate": 2.278398375163209e-05, "loss": 0.0083, "step": 21390 }, { "epoch": 2.3190589765828276, "grad_norm": 0.040056295692920685, "learning_rate": 2.2780356883795156e-05, "loss": 0.0014, "step": 21391 }, { "epoch": 2.319167389418907, "grad_norm": 0.5634153485298157, "learning_rate": 2.277673001595822e-05, "loss": 0.0283, "step": 21392 }, { "epoch": 2.319275802254987, "grad_norm": 0.0322139672935009, "learning_rate": 2.2773103148121284e-05, "loss": 0.0011, "step": 21393 }, { "epoch": 2.3193842150910666, "grad_norm": 0.20233437418937683, "learning_rate": 2.2769476280284348e-05, "loss": 0.0077, "step": 21394 }, { "epoch": 2.3194926279271466, "grad_norm": 0.309600830078125, "learning_rate": 2.2765849412447412e-05, "loss": 0.0287, "step": 21395 }, { "epoch": 2.319601040763226, "grad_norm": 0.06763008236885071, "learning_rate": 2.2762222544610476e-05, "loss": 0.0024, "step": 21396 }, { "epoch": 2.319709453599306, "grad_norm": 0.49561169743537903, "learning_rate": 2.275859567677354e-05, "loss": 0.0119, "step": 21397 }, { "epoch": 2.319817866435386, "grad_norm": 0.18620187044143677, "learning_rate": 2.27549688089366e-05, "loss": 0.011, "step": 21398 }, { "epoch": 2.3199262792714657, "grad_norm": 0.10974155366420746, "learning_rate": 2.275134194109967e-05, "loss": 0.0074, "step": 21399 }, { "epoch": 2.3200346921075456, "grad_norm": 0.34159332513809204, "learning_rate": 2.2747715073262733e-05, "loss": 0.0079, "step": 21400 }, { "epoch": 2.320143104943625, "grad_norm": 0.2067742496728897, "learning_rate": 2.2744088205425797e-05, "loss": 0.0066, "step": 21401 }, { "epoch": 2.320251517779705, "grad_norm": 0.02947753667831421, "learning_rate": 2.2740461337588857e-05, "loss": 0.0019, "step": 21402 }, { "epoch": 2.3203599306157847, "grad_norm": 0.34178534150123596, "learning_rate": 2.273683446975192e-05, "loss": 0.0261, "step": 21403 }, { "epoch": 2.3204683434518647, "grad_norm": 0.03543788567185402, "learning_rate": 2.273320760191499e-05, "loss": 0.0009, "step": 21404 }, { "epoch": 2.3205767562879447, "grad_norm": 0.3760812282562256, "learning_rate": 2.2729580734078053e-05, "loss": 0.0422, "step": 21405 }, { "epoch": 2.3206851691240242, "grad_norm": 0.32099419832229614, "learning_rate": 2.2725953866241114e-05, "loss": 0.0151, "step": 21406 }, { "epoch": 2.320793581960104, "grad_norm": 0.2500982880592346, "learning_rate": 2.2722326998404178e-05, "loss": 0.024, "step": 21407 }, { "epoch": 2.3209019947961838, "grad_norm": 0.16042393445968628, "learning_rate": 2.2718700130567245e-05, "loss": 0.0155, "step": 21408 }, { "epoch": 2.3210104076322637, "grad_norm": 0.4106130003929138, "learning_rate": 2.2715073262730306e-05, "loss": 0.0143, "step": 21409 }, { "epoch": 2.3211188204683433, "grad_norm": 0.5873095989227295, "learning_rate": 2.271144639489337e-05, "loss": 0.0171, "step": 21410 }, { "epoch": 2.3212272333044233, "grad_norm": 0.07893194258213043, "learning_rate": 2.2707819527056434e-05, "loss": 0.0033, "step": 21411 }, { "epoch": 2.3213356461405033, "grad_norm": 0.1666736751794815, "learning_rate": 2.2704192659219502e-05, "loss": 0.005, "step": 21412 }, { "epoch": 2.321444058976583, "grad_norm": 0.10069736838340759, "learning_rate": 2.2700565791382563e-05, "loss": 0.0025, "step": 21413 }, { "epoch": 2.3215524718126628, "grad_norm": 0.07887028157711029, "learning_rate": 2.2696938923545627e-05, "loss": 0.0033, "step": 21414 }, { "epoch": 2.3216608846487423, "grad_norm": 0.7130598425865173, "learning_rate": 2.269331205570869e-05, "loss": 0.0509, "step": 21415 }, { "epoch": 2.3217692974848223, "grad_norm": 0.25470679998397827, "learning_rate": 2.2689685187871755e-05, "loss": 0.006, "step": 21416 }, { "epoch": 2.321877710320902, "grad_norm": 0.12548255920410156, "learning_rate": 2.268605832003482e-05, "loss": 0.0072, "step": 21417 }, { "epoch": 2.321986123156982, "grad_norm": 0.0900622010231018, "learning_rate": 2.2682431452197883e-05, "loss": 0.0103, "step": 21418 }, { "epoch": 2.3220945359930614, "grad_norm": 0.033428601920604706, "learning_rate": 2.2678804584360947e-05, "loss": 0.0016, "step": 21419 }, { "epoch": 2.3222029488291414, "grad_norm": 0.2757578492164612, "learning_rate": 2.267517771652401e-05, "loss": 0.0147, "step": 21420 }, { "epoch": 2.3223113616652213, "grad_norm": 0.08113999664783478, "learning_rate": 2.2671550848687075e-05, "loss": 0.0029, "step": 21421 }, { "epoch": 2.322419774501301, "grad_norm": 1.1155825853347778, "learning_rate": 2.266792398085014e-05, "loss": 0.0709, "step": 21422 }, { "epoch": 2.322528187337381, "grad_norm": 0.16825492680072784, "learning_rate": 2.2664297113013204e-05, "loss": 0.0104, "step": 21423 }, { "epoch": 2.3226366001734604, "grad_norm": 0.035937532782554626, "learning_rate": 2.2660670245176264e-05, "loss": 0.0021, "step": 21424 }, { "epoch": 2.3227450130095404, "grad_norm": 0.23187381029129028, "learning_rate": 2.2657043377339332e-05, "loss": 0.0197, "step": 21425 }, { "epoch": 2.32285342584562, "grad_norm": 0.38326483964920044, "learning_rate": 2.2653416509502396e-05, "loss": 0.0102, "step": 21426 }, { "epoch": 2.3229618386817, "grad_norm": 0.24670810997486115, "learning_rate": 2.264978964166546e-05, "loss": 0.0144, "step": 21427 }, { "epoch": 2.3230702515177795, "grad_norm": 0.23341184854507446, "learning_rate": 2.264616277382852e-05, "loss": 0.0097, "step": 21428 }, { "epoch": 2.3231786643538594, "grad_norm": 0.4157334864139557, "learning_rate": 2.2642535905991588e-05, "loss": 0.011, "step": 21429 }, { "epoch": 2.3232870771899394, "grad_norm": 0.11696621030569077, "learning_rate": 2.2638909038154652e-05, "loss": 0.0033, "step": 21430 }, { "epoch": 2.323395490026019, "grad_norm": 0.24425120651721954, "learning_rate": 2.2635282170317716e-05, "loss": 0.0059, "step": 21431 }, { "epoch": 2.323503902862099, "grad_norm": 0.4295918047428131, "learning_rate": 2.2631655302480777e-05, "loss": 0.0237, "step": 21432 }, { "epoch": 2.3236123156981785, "grad_norm": 0.08729303628206253, "learning_rate": 2.262802843464384e-05, "loss": 0.0044, "step": 21433 }, { "epoch": 2.3237207285342585, "grad_norm": 0.36519870162010193, "learning_rate": 2.262440156680691e-05, "loss": 0.0086, "step": 21434 }, { "epoch": 2.3238291413703385, "grad_norm": 0.6440272331237793, "learning_rate": 2.262077469896997e-05, "loss": 0.0167, "step": 21435 }, { "epoch": 2.323937554206418, "grad_norm": 0.8426991701126099, "learning_rate": 2.2617147831133034e-05, "loss": 0.0993, "step": 21436 }, { "epoch": 2.324045967042498, "grad_norm": 0.22501584887504578, "learning_rate": 2.2613520963296098e-05, "loss": 0.0169, "step": 21437 }, { "epoch": 2.3241543798785775, "grad_norm": 0.013164815492928028, "learning_rate": 2.2609894095459165e-05, "loss": 0.0003, "step": 21438 }, { "epoch": 2.3242627927146575, "grad_norm": 0.2351577877998352, "learning_rate": 2.2606267227622226e-05, "loss": 0.0094, "step": 21439 }, { "epoch": 2.324371205550737, "grad_norm": 0.14364612102508545, "learning_rate": 2.260264035978529e-05, "loss": 0.0117, "step": 21440 }, { "epoch": 2.324479618386817, "grad_norm": 0.02561958134174347, "learning_rate": 2.2599013491948354e-05, "loss": 0.0011, "step": 21441 }, { "epoch": 2.3245880312228966, "grad_norm": 0.05965960770845413, "learning_rate": 2.2595386624111418e-05, "loss": 0.0028, "step": 21442 }, { "epoch": 2.3246964440589766, "grad_norm": 0.28286704421043396, "learning_rate": 2.2591759756274482e-05, "loss": 0.0123, "step": 21443 }, { "epoch": 2.3248048568950566, "grad_norm": 0.0254037044942379, "learning_rate": 2.2588132888437546e-05, "loss": 0.0008, "step": 21444 }, { "epoch": 2.324913269731136, "grad_norm": 0.12245184928178787, "learning_rate": 2.258450602060061e-05, "loss": 0.0023, "step": 21445 }, { "epoch": 2.325021682567216, "grad_norm": 0.11600754410028458, "learning_rate": 2.2580879152763675e-05, "loss": 0.0024, "step": 21446 }, { "epoch": 2.3251300954032956, "grad_norm": 0.042550016194581985, "learning_rate": 2.257725228492674e-05, "loss": 0.0019, "step": 21447 }, { "epoch": 2.3252385082393756, "grad_norm": 0.37032032012939453, "learning_rate": 2.2573625417089803e-05, "loss": 0.009, "step": 21448 }, { "epoch": 2.325346921075455, "grad_norm": 0.8211894631385803, "learning_rate": 2.2569998549252867e-05, "loss": 0.0292, "step": 21449 }, { "epoch": 2.325455333911535, "grad_norm": 0.15441009402275085, "learning_rate": 2.2566371681415928e-05, "loss": 0.0048, "step": 21450 }, { "epoch": 2.3255637467476147, "grad_norm": 0.6664011478424072, "learning_rate": 2.2562744813578995e-05, "loss": 0.0232, "step": 21451 }, { "epoch": 2.3256721595836947, "grad_norm": 0.3714180886745453, "learning_rate": 2.255911794574206e-05, "loss": 0.0231, "step": 21452 }, { "epoch": 2.3257805724197746, "grad_norm": 0.22924408316612244, "learning_rate": 2.2555491077905123e-05, "loss": 0.0065, "step": 21453 }, { "epoch": 2.325888985255854, "grad_norm": 0.12241699546575546, "learning_rate": 2.2551864210068184e-05, "loss": 0.0017, "step": 21454 }, { "epoch": 2.325997398091934, "grad_norm": 0.401670902967453, "learning_rate": 2.254823734223125e-05, "loss": 0.0213, "step": 21455 }, { "epoch": 2.3261058109280137, "grad_norm": 0.19270217418670654, "learning_rate": 2.2544610474394316e-05, "loss": 0.0094, "step": 21456 }, { "epoch": 2.3262142237640937, "grad_norm": 0.03962382674217224, "learning_rate": 2.254098360655738e-05, "loss": 0.0018, "step": 21457 }, { "epoch": 2.3263226366001737, "grad_norm": 0.24213212728500366, "learning_rate": 2.253735673872044e-05, "loss": 0.008, "step": 21458 }, { "epoch": 2.3264310494362532, "grad_norm": 0.13835136592388153, "learning_rate": 2.2533729870883505e-05, "loss": 0.0072, "step": 21459 }, { "epoch": 2.326539462272333, "grad_norm": 0.4819132089614868, "learning_rate": 2.2530103003046572e-05, "loss": 0.0185, "step": 21460 }, { "epoch": 2.3266478751084128, "grad_norm": 0.3210914134979248, "learning_rate": 2.2526476135209633e-05, "loss": 0.0197, "step": 21461 }, { "epoch": 2.3267562879444927, "grad_norm": 0.07948805391788483, "learning_rate": 2.2522849267372697e-05, "loss": 0.0021, "step": 21462 }, { "epoch": 2.3268647007805723, "grad_norm": 0.3670153021812439, "learning_rate": 2.251922239953576e-05, "loss": 0.0135, "step": 21463 }, { "epoch": 2.3269731136166523, "grad_norm": 0.291859894990921, "learning_rate": 2.251559553169883e-05, "loss": 0.0096, "step": 21464 }, { "epoch": 2.327081526452732, "grad_norm": 0.19784541428089142, "learning_rate": 2.251196866386189e-05, "loss": 0.0162, "step": 21465 }, { "epoch": 2.327189939288812, "grad_norm": 0.7528513073921204, "learning_rate": 2.2508341796024953e-05, "loss": 0.0632, "step": 21466 }, { "epoch": 2.327298352124892, "grad_norm": 0.3091149926185608, "learning_rate": 2.2504714928188017e-05, "loss": 0.0059, "step": 21467 }, { "epoch": 2.3274067649609713, "grad_norm": 0.1495925486087799, "learning_rate": 2.250108806035108e-05, "loss": 0.0082, "step": 21468 }, { "epoch": 2.3275151777970513, "grad_norm": 0.1570836454629898, "learning_rate": 2.2497461192514146e-05, "loss": 0.0109, "step": 21469 }, { "epoch": 2.327623590633131, "grad_norm": 0.2929331064224243, "learning_rate": 2.249383432467721e-05, "loss": 0.0256, "step": 21470 }, { "epoch": 2.327732003469211, "grad_norm": 0.010525296442210674, "learning_rate": 2.2490207456840274e-05, "loss": 0.0004, "step": 21471 }, { "epoch": 2.3278404163052904, "grad_norm": 0.34221315383911133, "learning_rate": 2.2486580589003338e-05, "loss": 0.0065, "step": 21472 }, { "epoch": 2.3279488291413704, "grad_norm": 0.02131068892776966, "learning_rate": 2.2482953721166402e-05, "loss": 0.0006, "step": 21473 }, { "epoch": 2.32805724197745, "grad_norm": 0.1787450611591339, "learning_rate": 2.2479326853329466e-05, "loss": 0.0136, "step": 21474 }, { "epoch": 2.32816565481353, "grad_norm": 0.13099130988121033, "learning_rate": 2.247569998549253e-05, "loss": 0.0019, "step": 21475 }, { "epoch": 2.32827406764961, "grad_norm": 0.3091093897819519, "learning_rate": 2.247207311765559e-05, "loss": 0.0184, "step": 21476 }, { "epoch": 2.3283824804856894, "grad_norm": 0.0028834675904363394, "learning_rate": 2.246844624981866e-05, "loss": 0.0001, "step": 21477 }, { "epoch": 2.3284908933217694, "grad_norm": 0.08485205471515656, "learning_rate": 2.2464819381981723e-05, "loss": 0.0021, "step": 21478 }, { "epoch": 2.328599306157849, "grad_norm": 0.08157916367053986, "learning_rate": 2.2461192514144787e-05, "loss": 0.0036, "step": 21479 }, { "epoch": 2.328707718993929, "grad_norm": 0.22616952657699585, "learning_rate": 2.2457565646307847e-05, "loss": 0.0077, "step": 21480 }, { "epoch": 2.328816131830009, "grad_norm": 0.2131870687007904, "learning_rate": 2.2453938778470915e-05, "loss": 0.0108, "step": 21481 }, { "epoch": 2.3289245446660884, "grad_norm": 0.36084017157554626, "learning_rate": 2.245031191063398e-05, "loss": 0.0031, "step": 21482 }, { "epoch": 2.3290329575021684, "grad_norm": 0.5414370894432068, "learning_rate": 2.2446685042797043e-05, "loss": 0.015, "step": 21483 }, { "epoch": 2.329141370338248, "grad_norm": 0.36587658524513245, "learning_rate": 2.2443058174960104e-05, "loss": 0.0164, "step": 21484 }, { "epoch": 2.329249783174328, "grad_norm": 0.32379457354545593, "learning_rate": 2.2439431307123168e-05, "loss": 0.0098, "step": 21485 }, { "epoch": 2.3293581960104075, "grad_norm": 0.10726682841777802, "learning_rate": 2.2435804439286235e-05, "loss": 0.0035, "step": 21486 }, { "epoch": 2.3294666088464875, "grad_norm": 0.34190860390663147, "learning_rate": 2.2432177571449296e-05, "loss": 0.0179, "step": 21487 }, { "epoch": 2.329575021682567, "grad_norm": 0.1443691998720169, "learning_rate": 2.242855070361236e-05, "loss": 0.0078, "step": 21488 }, { "epoch": 2.329683434518647, "grad_norm": 0.19681867957115173, "learning_rate": 2.2424923835775424e-05, "loss": 0.0072, "step": 21489 }, { "epoch": 2.329791847354727, "grad_norm": 0.025078369304537773, "learning_rate": 2.2421296967938492e-05, "loss": 0.0009, "step": 21490 }, { "epoch": 2.3299002601908065, "grad_norm": 0.0014552491484209895, "learning_rate": 2.2417670100101553e-05, "loss": 0.0001, "step": 21491 }, { "epoch": 2.3300086730268865, "grad_norm": 0.0340154655277729, "learning_rate": 2.2414043232264617e-05, "loss": 0.001, "step": 21492 }, { "epoch": 2.330117085862966, "grad_norm": 0.3974072337150574, "learning_rate": 2.241041636442768e-05, "loss": 0.0124, "step": 21493 }, { "epoch": 2.330225498699046, "grad_norm": 0.29335129261016846, "learning_rate": 2.2406789496590748e-05, "loss": 0.0335, "step": 21494 }, { "epoch": 2.3303339115351256, "grad_norm": 0.21931160986423492, "learning_rate": 2.240316262875381e-05, "loss": 0.0083, "step": 21495 }, { "epoch": 2.3304423243712056, "grad_norm": 0.1876162737607956, "learning_rate": 2.2399535760916873e-05, "loss": 0.0013, "step": 21496 }, { "epoch": 2.330550737207285, "grad_norm": 0.24588914215564728, "learning_rate": 2.2395908893079937e-05, "loss": 0.0075, "step": 21497 }, { "epoch": 2.330659150043365, "grad_norm": 0.18223460018634796, "learning_rate": 2.2392282025243e-05, "loss": 0.0101, "step": 21498 }, { "epoch": 2.330767562879445, "grad_norm": 0.0817522257566452, "learning_rate": 2.2388655157406065e-05, "loss": 0.0039, "step": 21499 }, { "epoch": 2.3308759757155246, "grad_norm": 0.6052305102348328, "learning_rate": 2.238502828956913e-05, "loss": 0.0123, "step": 21500 }, { "epoch": 2.3309843885516046, "grad_norm": 0.6446964144706726, "learning_rate": 2.2381401421732194e-05, "loss": 0.024, "step": 21501 }, { "epoch": 2.331092801387684, "grad_norm": 1.138173222541809, "learning_rate": 2.2377774553895258e-05, "loss": 0.0235, "step": 21502 }, { "epoch": 2.331201214223764, "grad_norm": 0.25838565826416016, "learning_rate": 2.2374147686058322e-05, "loss": 0.0211, "step": 21503 }, { "epoch": 2.331309627059844, "grad_norm": 0.3539731800556183, "learning_rate": 2.2370520818221386e-05, "loss": 0.0037, "step": 21504 }, { "epoch": 2.3314180398959237, "grad_norm": 0.33238285779953003, "learning_rate": 2.236689395038445e-05, "loss": 0.0097, "step": 21505 }, { "epoch": 2.3315264527320037, "grad_norm": 0.3595082759857178, "learning_rate": 2.236326708254751e-05, "loss": 0.0186, "step": 21506 }, { "epoch": 2.331634865568083, "grad_norm": 0.02971828728914261, "learning_rate": 2.2359640214710578e-05, "loss": 0.001, "step": 21507 }, { "epoch": 2.331743278404163, "grad_norm": 0.4106092154979706, "learning_rate": 2.2356013346873642e-05, "loss": 0.0317, "step": 21508 }, { "epoch": 2.3318516912402427, "grad_norm": 0.3312353491783142, "learning_rate": 2.2352386479036706e-05, "loss": 0.0095, "step": 21509 }, { "epoch": 2.3319601040763227, "grad_norm": 0.2145569920539856, "learning_rate": 2.2348759611199767e-05, "loss": 0.0144, "step": 21510 }, { "epoch": 2.3320685169124022, "grad_norm": 1.0504404306411743, "learning_rate": 2.234513274336283e-05, "loss": 0.0666, "step": 21511 }, { "epoch": 2.3321769297484822, "grad_norm": 0.19424493610858917, "learning_rate": 2.23415058755259e-05, "loss": 0.0076, "step": 21512 }, { "epoch": 2.332285342584562, "grad_norm": 0.16764988005161285, "learning_rate": 2.233787900768896e-05, "loss": 0.0092, "step": 21513 }, { "epoch": 2.3323937554206418, "grad_norm": 0.11733746528625488, "learning_rate": 2.2334252139852024e-05, "loss": 0.0045, "step": 21514 }, { "epoch": 2.3325021682567217, "grad_norm": 0.3946448266506195, "learning_rate": 2.2330625272015088e-05, "loss": 0.0225, "step": 21515 }, { "epoch": 2.3326105810928013, "grad_norm": 0.468873530626297, "learning_rate": 2.2326998404178155e-05, "loss": 0.0079, "step": 21516 }, { "epoch": 2.3327189939288813, "grad_norm": 1.434896469116211, "learning_rate": 2.2323371536341216e-05, "loss": 0.0264, "step": 21517 }, { "epoch": 2.332827406764961, "grad_norm": 0.2476894110441208, "learning_rate": 2.231974466850428e-05, "loss": 0.0118, "step": 21518 }, { "epoch": 2.332935819601041, "grad_norm": 0.0035963570699095726, "learning_rate": 2.2316117800667344e-05, "loss": 0.0001, "step": 21519 }, { "epoch": 2.3330442324371203, "grad_norm": 0.45630964636802673, "learning_rate": 2.231249093283041e-05, "loss": 0.0208, "step": 21520 }, { "epoch": 2.3331526452732003, "grad_norm": 1.5303916931152344, "learning_rate": 2.2308864064993472e-05, "loss": 0.0351, "step": 21521 }, { "epoch": 2.3332610581092803, "grad_norm": 0.45385679602622986, "learning_rate": 2.2305237197156536e-05, "loss": 0.0188, "step": 21522 }, { "epoch": 2.33336947094536, "grad_norm": 0.6048917174339294, "learning_rate": 2.23016103293196e-05, "loss": 0.006, "step": 21523 }, { "epoch": 2.33347788378144, "grad_norm": 0.4919056296348572, "learning_rate": 2.2297983461482665e-05, "loss": 0.049, "step": 21524 }, { "epoch": 2.3335862966175194, "grad_norm": 0.24925851821899414, "learning_rate": 2.229435659364573e-05, "loss": 0.0093, "step": 21525 }, { "epoch": 2.3336947094535994, "grad_norm": 0.011092386208474636, "learning_rate": 2.2290729725808793e-05, "loss": 0.0003, "step": 21526 }, { "epoch": 2.3338031222896793, "grad_norm": 0.22895067930221558, "learning_rate": 2.2287102857971857e-05, "loss": 0.0079, "step": 21527 }, { "epoch": 2.333911535125759, "grad_norm": 0.028456997126340866, "learning_rate": 2.228347599013492e-05, "loss": 0.0009, "step": 21528 }, { "epoch": 2.334019947961839, "grad_norm": 0.1371828019618988, "learning_rate": 2.2279849122297985e-05, "loss": 0.0062, "step": 21529 }, { "epoch": 2.3341283607979184, "grad_norm": 0.32109686732292175, "learning_rate": 2.227622225446105e-05, "loss": 0.0043, "step": 21530 }, { "epoch": 2.3342367736339984, "grad_norm": 0.03930029645562172, "learning_rate": 2.2272595386624113e-05, "loss": 0.0027, "step": 21531 }, { "epoch": 2.334345186470078, "grad_norm": 0.9356905817985535, "learning_rate": 2.2268968518787174e-05, "loss": 0.0205, "step": 21532 }, { "epoch": 2.334453599306158, "grad_norm": 0.35345524549484253, "learning_rate": 2.226534165095024e-05, "loss": 0.0185, "step": 21533 }, { "epoch": 2.3345620121422375, "grad_norm": 0.09783186763525009, "learning_rate": 2.2261714783113306e-05, "loss": 0.0028, "step": 21534 }, { "epoch": 2.3346704249783174, "grad_norm": 0.1508801430463791, "learning_rate": 2.225808791527637e-05, "loss": 0.0193, "step": 21535 }, { "epoch": 2.3347788378143974, "grad_norm": 0.2331497073173523, "learning_rate": 2.225446104743943e-05, "loss": 0.0031, "step": 21536 }, { "epoch": 2.334887250650477, "grad_norm": 0.023455148562788963, "learning_rate": 2.2250834179602498e-05, "loss": 0.001, "step": 21537 }, { "epoch": 2.334995663486557, "grad_norm": 0.3231172561645508, "learning_rate": 2.2247207311765562e-05, "loss": 0.0114, "step": 21538 }, { "epoch": 2.3351040763226365, "grad_norm": 0.23700016736984253, "learning_rate": 2.2243580443928623e-05, "loss": 0.0424, "step": 21539 }, { "epoch": 2.3352124891587165, "grad_norm": 0.04099327325820923, "learning_rate": 2.2239953576091687e-05, "loss": 0.0021, "step": 21540 }, { "epoch": 2.335320901994796, "grad_norm": 0.0563892163336277, "learning_rate": 2.223632670825475e-05, "loss": 0.0032, "step": 21541 }, { "epoch": 2.335429314830876, "grad_norm": 0.32141727209091187, "learning_rate": 2.223269984041782e-05, "loss": 0.0073, "step": 21542 }, { "epoch": 2.3355377276669556, "grad_norm": 0.27440890669822693, "learning_rate": 2.222907297258088e-05, "loss": 0.0144, "step": 21543 }, { "epoch": 2.3356461405030355, "grad_norm": 0.16082608699798584, "learning_rate": 2.2225446104743943e-05, "loss": 0.0064, "step": 21544 }, { "epoch": 2.3357545533391155, "grad_norm": 0.36664828658103943, "learning_rate": 2.2221819236907007e-05, "loss": 0.0285, "step": 21545 }, { "epoch": 2.335862966175195, "grad_norm": 0.2512025833129883, "learning_rate": 2.2218192369070075e-05, "loss": 0.0114, "step": 21546 }, { "epoch": 2.335971379011275, "grad_norm": 0.7323904633522034, "learning_rate": 2.2214565501233136e-05, "loss": 0.0245, "step": 21547 }, { "epoch": 2.3360797918473546, "grad_norm": 0.02056986466050148, "learning_rate": 2.22109386333962e-05, "loss": 0.0006, "step": 21548 }, { "epoch": 2.3361882046834346, "grad_norm": 0.19560573995113373, "learning_rate": 2.2207311765559264e-05, "loss": 0.0085, "step": 21549 }, { "epoch": 2.336296617519514, "grad_norm": 0.024181846529245377, "learning_rate": 2.2203684897722328e-05, "loss": 0.0015, "step": 21550 }, { "epoch": 2.336405030355594, "grad_norm": 0.22179186344146729, "learning_rate": 2.2200058029885392e-05, "loss": 0.0077, "step": 21551 }, { "epoch": 2.3365134431916736, "grad_norm": 1.4734539985656738, "learning_rate": 2.2196431162048456e-05, "loss": 0.0276, "step": 21552 }, { "epoch": 2.3366218560277536, "grad_norm": 0.423594206571579, "learning_rate": 2.219280429421152e-05, "loss": 0.0075, "step": 21553 }, { "epoch": 2.3367302688638336, "grad_norm": 0.024431802332401276, "learning_rate": 2.2189177426374584e-05, "loss": 0.0013, "step": 21554 }, { "epoch": 2.336838681699913, "grad_norm": 0.5902124643325806, "learning_rate": 2.218555055853765e-05, "loss": 0.0339, "step": 21555 }, { "epoch": 2.336947094535993, "grad_norm": 0.6920711398124695, "learning_rate": 2.2181923690700713e-05, "loss": 0.0225, "step": 21556 }, { "epoch": 2.3370555073720727, "grad_norm": 0.4197554886341095, "learning_rate": 2.2178296822863777e-05, "loss": 0.0145, "step": 21557 }, { "epoch": 2.3371639202081527, "grad_norm": 0.24913297593593597, "learning_rate": 2.2174669955026837e-05, "loss": 0.012, "step": 21558 }, { "epoch": 2.3372723330442327, "grad_norm": 0.24860461056232452, "learning_rate": 2.2171043087189905e-05, "loss": 0.0168, "step": 21559 }, { "epoch": 2.337380745880312, "grad_norm": 0.2694483995437622, "learning_rate": 2.216741621935297e-05, "loss": 0.0069, "step": 21560 }, { "epoch": 2.337489158716392, "grad_norm": 0.22382310032844543, "learning_rate": 2.2163789351516033e-05, "loss": 0.0077, "step": 21561 }, { "epoch": 2.3375975715524717, "grad_norm": 0.528704822063446, "learning_rate": 2.2160162483679094e-05, "loss": 0.017, "step": 21562 }, { "epoch": 2.3377059843885517, "grad_norm": 0.12326207756996155, "learning_rate": 2.215653561584216e-05, "loss": 0.004, "step": 21563 }, { "epoch": 2.3378143972246312, "grad_norm": 0.11865855008363724, "learning_rate": 2.2152908748005225e-05, "loss": 0.0039, "step": 21564 }, { "epoch": 2.3379228100607112, "grad_norm": 0.23605211079120636, "learning_rate": 2.2149281880168286e-05, "loss": 0.0049, "step": 21565 }, { "epoch": 2.3380312228967908, "grad_norm": 0.1839771866798401, "learning_rate": 2.214565501233135e-05, "loss": 0.0041, "step": 21566 }, { "epoch": 2.3381396357328708, "grad_norm": 0.14403577148914337, "learning_rate": 2.2142028144494414e-05, "loss": 0.0079, "step": 21567 }, { "epoch": 2.3382480485689507, "grad_norm": 0.22602106630802155, "learning_rate": 2.2138401276657482e-05, "loss": 0.0095, "step": 21568 }, { "epoch": 2.3383564614050303, "grad_norm": 0.28543776273727417, "learning_rate": 2.2134774408820543e-05, "loss": 0.005, "step": 21569 }, { "epoch": 2.3384648742411103, "grad_norm": 0.12404311448335648, "learning_rate": 2.2131147540983607e-05, "loss": 0.0031, "step": 21570 }, { "epoch": 2.33857328707719, "grad_norm": 0.05501605197787285, "learning_rate": 2.212752067314667e-05, "loss": 0.0038, "step": 21571 }, { "epoch": 2.33868169991327, "grad_norm": 0.3142530918121338, "learning_rate": 2.2123893805309738e-05, "loss": 0.0145, "step": 21572 }, { "epoch": 2.3387901127493493, "grad_norm": 0.08944924175739288, "learning_rate": 2.21202669374728e-05, "loss": 0.0044, "step": 21573 }, { "epoch": 2.3388985255854293, "grad_norm": 0.3208363354206085, "learning_rate": 2.2116640069635863e-05, "loss": 0.0258, "step": 21574 }, { "epoch": 2.339006938421509, "grad_norm": 2.0313661098480225, "learning_rate": 2.2113013201798927e-05, "loss": 0.0267, "step": 21575 }, { "epoch": 2.339115351257589, "grad_norm": 0.3605486750602722, "learning_rate": 2.210938633396199e-05, "loss": 0.0244, "step": 21576 }, { "epoch": 2.339223764093669, "grad_norm": 0.49660900235176086, "learning_rate": 2.2105759466125055e-05, "loss": 0.0444, "step": 21577 }, { "epoch": 2.3393321769297484, "grad_norm": 0.07549742609262466, "learning_rate": 2.210213259828812e-05, "loss": 0.0019, "step": 21578 }, { "epoch": 2.3394405897658284, "grad_norm": 0.08713794499635696, "learning_rate": 2.2098505730451184e-05, "loss": 0.0041, "step": 21579 }, { "epoch": 2.339549002601908, "grad_norm": 0.4658828675746918, "learning_rate": 2.2094878862614248e-05, "loss": 0.0117, "step": 21580 }, { "epoch": 2.339657415437988, "grad_norm": 0.6521451473236084, "learning_rate": 2.2091251994777312e-05, "loss": 0.0779, "step": 21581 }, { "epoch": 2.339765828274068, "grad_norm": 0.2787415087223053, "learning_rate": 2.2087625126940376e-05, "loss": 0.0109, "step": 21582 }, { "epoch": 2.3398742411101474, "grad_norm": 0.0675138384103775, "learning_rate": 2.208399825910344e-05, "loss": 0.002, "step": 21583 }, { "epoch": 2.3399826539462274, "grad_norm": 0.2540389597415924, "learning_rate": 2.20803713912665e-05, "loss": 0.0147, "step": 21584 }, { "epoch": 2.340091066782307, "grad_norm": 0.30888307094573975, "learning_rate": 2.2076744523429568e-05, "loss": 0.014, "step": 21585 }, { "epoch": 2.340199479618387, "grad_norm": 0.09376300126314163, "learning_rate": 2.2073117655592632e-05, "loss": 0.0024, "step": 21586 }, { "epoch": 2.3403078924544665, "grad_norm": 0.07584085315465927, "learning_rate": 2.2069490787755696e-05, "loss": 0.0051, "step": 21587 }, { "epoch": 2.3404163052905465, "grad_norm": 0.05756469815969467, "learning_rate": 2.2065863919918757e-05, "loss": 0.0038, "step": 21588 }, { "epoch": 2.340524718126626, "grad_norm": 0.24154138565063477, "learning_rate": 2.2062237052081825e-05, "loss": 0.0459, "step": 21589 }, { "epoch": 2.340633130962706, "grad_norm": 0.04896991327404976, "learning_rate": 2.205861018424489e-05, "loss": 0.0019, "step": 21590 }, { "epoch": 2.340741543798786, "grad_norm": 0.42510250210762024, "learning_rate": 2.205498331640795e-05, "loss": 0.0477, "step": 21591 }, { "epoch": 2.3408499566348655, "grad_norm": 0.1929902583360672, "learning_rate": 2.2051356448571014e-05, "loss": 0.0086, "step": 21592 }, { "epoch": 2.3409583694709455, "grad_norm": 0.7737979888916016, "learning_rate": 2.2047729580734078e-05, "loss": 0.0377, "step": 21593 }, { "epoch": 2.341066782307025, "grad_norm": 0.10944432020187378, "learning_rate": 2.2044102712897145e-05, "loss": 0.009, "step": 21594 }, { "epoch": 2.341175195143105, "grad_norm": 0.04354925826191902, "learning_rate": 2.2040475845060206e-05, "loss": 0.0013, "step": 21595 }, { "epoch": 2.3412836079791846, "grad_norm": 0.08478797227144241, "learning_rate": 2.203684897722327e-05, "loss": 0.0038, "step": 21596 }, { "epoch": 2.3413920208152645, "grad_norm": 0.754458487033844, "learning_rate": 2.2033222109386334e-05, "loss": 0.046, "step": 21597 }, { "epoch": 2.341500433651344, "grad_norm": 0.2524363100528717, "learning_rate": 2.20295952415494e-05, "loss": 0.008, "step": 21598 }, { "epoch": 2.341608846487424, "grad_norm": 0.27946314215660095, "learning_rate": 2.2025968373712462e-05, "loss": 0.0094, "step": 21599 }, { "epoch": 2.341717259323504, "grad_norm": 0.21779075264930725, "learning_rate": 2.2022341505875526e-05, "loss": 0.0102, "step": 21600 }, { "epoch": 2.3418256721595836, "grad_norm": 0.13149702548980713, "learning_rate": 2.201871463803859e-05, "loss": 0.0029, "step": 21601 }, { "epoch": 2.3419340849956636, "grad_norm": 0.22298972308635712, "learning_rate": 2.2015087770201655e-05, "loss": 0.026, "step": 21602 }, { "epoch": 2.342042497831743, "grad_norm": 0.4584830403327942, "learning_rate": 2.201146090236472e-05, "loss": 0.0275, "step": 21603 }, { "epoch": 2.342150910667823, "grad_norm": 0.2180413156747818, "learning_rate": 2.2007834034527783e-05, "loss": 0.0186, "step": 21604 }, { "epoch": 2.342259323503903, "grad_norm": 0.05735940486192703, "learning_rate": 2.2004207166690847e-05, "loss": 0.0018, "step": 21605 }, { "epoch": 2.3423677363399826, "grad_norm": 0.42143094539642334, "learning_rate": 2.200058029885391e-05, "loss": 0.011, "step": 21606 }, { "epoch": 2.3424761491760626, "grad_norm": 0.0763929933309555, "learning_rate": 2.1996953431016975e-05, "loss": 0.0067, "step": 21607 }, { "epoch": 2.342584562012142, "grad_norm": 0.4664498567581177, "learning_rate": 2.199332656318004e-05, "loss": 0.0189, "step": 21608 }, { "epoch": 2.342692974848222, "grad_norm": 0.40463146567344666, "learning_rate": 2.1989699695343103e-05, "loss": 0.0163, "step": 21609 }, { "epoch": 2.3428013876843017, "grad_norm": 0.2577608823776245, "learning_rate": 2.1986072827506164e-05, "loss": 0.0093, "step": 21610 }, { "epoch": 2.3429098005203817, "grad_norm": 0.20154234766960144, "learning_rate": 2.198244595966923e-05, "loss": 0.0497, "step": 21611 }, { "epoch": 2.343018213356461, "grad_norm": 0.12947210669517517, "learning_rate": 2.1978819091832296e-05, "loss": 0.0026, "step": 21612 }, { "epoch": 2.343126626192541, "grad_norm": 0.6458091139793396, "learning_rate": 2.197519222399536e-05, "loss": 0.0119, "step": 21613 }, { "epoch": 2.343235039028621, "grad_norm": 0.2593913972377777, "learning_rate": 2.197156535615842e-05, "loss": 0.0142, "step": 21614 }, { "epoch": 2.3433434518647007, "grad_norm": 0.24034947156906128, "learning_rate": 2.1967938488321488e-05, "loss": 0.0133, "step": 21615 }, { "epoch": 2.3434518647007807, "grad_norm": 0.0819428563117981, "learning_rate": 2.1964311620484552e-05, "loss": 0.0038, "step": 21616 }, { "epoch": 2.3435602775368602, "grad_norm": 0.1401345431804657, "learning_rate": 2.1960684752647613e-05, "loss": 0.0018, "step": 21617 }, { "epoch": 2.3436686903729402, "grad_norm": 0.07589729875326157, "learning_rate": 2.1957057884810677e-05, "loss": 0.0035, "step": 21618 }, { "epoch": 2.3437771032090198, "grad_norm": 0.24588130414485931, "learning_rate": 2.195343101697374e-05, "loss": 0.0113, "step": 21619 }, { "epoch": 2.3438855160450998, "grad_norm": 0.04782566428184509, "learning_rate": 2.194980414913681e-05, "loss": 0.0027, "step": 21620 }, { "epoch": 2.3439939288811793, "grad_norm": 0.21734775602817535, "learning_rate": 2.194617728129987e-05, "loss": 0.0209, "step": 21621 }, { "epoch": 2.3441023417172593, "grad_norm": 0.660586953163147, "learning_rate": 2.1942550413462933e-05, "loss": 0.0213, "step": 21622 }, { "epoch": 2.3442107545533393, "grad_norm": 0.07469896227121353, "learning_rate": 2.1938923545625997e-05, "loss": 0.002, "step": 21623 }, { "epoch": 2.344319167389419, "grad_norm": 0.3918987810611725, "learning_rate": 2.1935296677789065e-05, "loss": 0.0165, "step": 21624 }, { "epoch": 2.344427580225499, "grad_norm": 0.18994955718517303, "learning_rate": 2.1931669809952126e-05, "loss": 0.0065, "step": 21625 }, { "epoch": 2.3445359930615783, "grad_norm": 0.19313138723373413, "learning_rate": 2.192804294211519e-05, "loss": 0.01, "step": 21626 }, { "epoch": 2.3446444058976583, "grad_norm": 0.03774604573845863, "learning_rate": 2.1924416074278254e-05, "loss": 0.001, "step": 21627 }, { "epoch": 2.3447528187337383, "grad_norm": 0.1854581981897354, "learning_rate": 2.1920789206441318e-05, "loss": 0.0162, "step": 21628 }, { "epoch": 2.344861231569818, "grad_norm": 0.43673479557037354, "learning_rate": 2.1917162338604382e-05, "loss": 0.0362, "step": 21629 }, { "epoch": 2.344969644405898, "grad_norm": 0.16947980225086212, "learning_rate": 2.1913535470767446e-05, "loss": 0.0082, "step": 21630 }, { "epoch": 2.3450780572419774, "grad_norm": 0.05617791414260864, "learning_rate": 2.190990860293051e-05, "loss": 0.0014, "step": 21631 }, { "epoch": 2.3451864700780574, "grad_norm": 0.3601355254650116, "learning_rate": 2.1906281735093574e-05, "loss": 0.0204, "step": 21632 }, { "epoch": 2.345294882914137, "grad_norm": 0.5297051072120667, "learning_rate": 2.190265486725664e-05, "loss": 0.0067, "step": 21633 }, { "epoch": 2.345403295750217, "grad_norm": 0.3349968194961548, "learning_rate": 2.1899027999419703e-05, "loss": 0.0105, "step": 21634 }, { "epoch": 2.3455117085862964, "grad_norm": 0.5417985916137695, "learning_rate": 2.1895401131582767e-05, "loss": 0.0172, "step": 21635 }, { "epoch": 2.3456201214223764, "grad_norm": 0.1219545379281044, "learning_rate": 2.1891774263745827e-05, "loss": 0.0063, "step": 21636 }, { "epoch": 2.3457285342584564, "grad_norm": 0.06491053849458694, "learning_rate": 2.1888147395908895e-05, "loss": 0.0021, "step": 21637 }, { "epoch": 2.345836947094536, "grad_norm": 0.2952733337879181, "learning_rate": 2.188452052807196e-05, "loss": 0.0216, "step": 21638 }, { "epoch": 2.345945359930616, "grad_norm": 0.33201050758361816, "learning_rate": 2.1880893660235023e-05, "loss": 0.0149, "step": 21639 }, { "epoch": 2.3460537727666955, "grad_norm": 0.07537253946065903, "learning_rate": 2.1877266792398084e-05, "loss": 0.0024, "step": 21640 }, { "epoch": 2.3461621856027755, "grad_norm": 0.09363221377134323, "learning_rate": 2.187363992456115e-05, "loss": 0.0079, "step": 21641 }, { "epoch": 2.346270598438855, "grad_norm": 0.0877758041024208, "learning_rate": 2.1870013056724215e-05, "loss": 0.0021, "step": 21642 }, { "epoch": 2.346379011274935, "grad_norm": 0.3329819440841675, "learning_rate": 2.1866386188887276e-05, "loss": 0.0107, "step": 21643 }, { "epoch": 2.3464874241110145, "grad_norm": 0.13266544044017792, "learning_rate": 2.186275932105034e-05, "loss": 0.0085, "step": 21644 }, { "epoch": 2.3465958369470945, "grad_norm": 0.23863781988620758, "learning_rate": 2.1859132453213408e-05, "loss": 0.0148, "step": 21645 }, { "epoch": 2.3467042497831745, "grad_norm": 0.09158080816268921, "learning_rate": 2.1855505585376472e-05, "loss": 0.0038, "step": 21646 }, { "epoch": 2.346812662619254, "grad_norm": 0.3441501259803772, "learning_rate": 2.1851878717539533e-05, "loss": 0.0264, "step": 21647 }, { "epoch": 2.346921075455334, "grad_norm": 0.16393424570560455, "learning_rate": 2.1848251849702597e-05, "loss": 0.0082, "step": 21648 }, { "epoch": 2.3470294882914136, "grad_norm": 0.04492081701755524, "learning_rate": 2.184462498186566e-05, "loss": 0.0011, "step": 21649 }, { "epoch": 2.3471379011274935, "grad_norm": 0.5653244256973267, "learning_rate": 2.1840998114028728e-05, "loss": 0.0149, "step": 21650 }, { "epoch": 2.3472463139635735, "grad_norm": 0.12303367257118225, "learning_rate": 2.183737124619179e-05, "loss": 0.0043, "step": 21651 }, { "epoch": 2.347354726799653, "grad_norm": 0.05032329261302948, "learning_rate": 2.1833744378354853e-05, "loss": 0.0015, "step": 21652 }, { "epoch": 2.347463139635733, "grad_norm": 0.6621467471122742, "learning_rate": 2.1830117510517917e-05, "loss": 0.0205, "step": 21653 }, { "epoch": 2.3475715524718126, "grad_norm": 0.08648882806301117, "learning_rate": 2.182649064268098e-05, "loss": 0.0016, "step": 21654 }, { "epoch": 2.3476799653078926, "grad_norm": 0.6192938685417175, "learning_rate": 2.1822863774844045e-05, "loss": 0.0282, "step": 21655 }, { "epoch": 2.347788378143972, "grad_norm": 0.22817955911159515, "learning_rate": 2.181923690700711e-05, "loss": 0.0052, "step": 21656 }, { "epoch": 2.347896790980052, "grad_norm": 0.25809305906295776, "learning_rate": 2.1815610039170174e-05, "loss": 0.0119, "step": 21657 }, { "epoch": 2.3480052038161316, "grad_norm": 0.43906885385513306, "learning_rate": 2.1811983171333238e-05, "loss": 0.0175, "step": 21658 }, { "epoch": 2.3481136166522116, "grad_norm": 0.013383598066866398, "learning_rate": 2.1808356303496302e-05, "loss": 0.0006, "step": 21659 }, { "epoch": 2.3482220294882916, "grad_norm": 0.20111384987831116, "learning_rate": 2.1804729435659366e-05, "loss": 0.0064, "step": 21660 }, { "epoch": 2.348330442324371, "grad_norm": 0.08609101176261902, "learning_rate": 2.180110256782243e-05, "loss": 0.0014, "step": 21661 }, { "epoch": 2.348438855160451, "grad_norm": 0.5221149325370789, "learning_rate": 2.179747569998549e-05, "loss": 0.0101, "step": 21662 }, { "epoch": 2.3485472679965307, "grad_norm": 0.1335216909646988, "learning_rate": 2.1793848832148558e-05, "loss": 0.0058, "step": 21663 }, { "epoch": 2.3486556808326107, "grad_norm": 0.0963730439543724, "learning_rate": 2.1790221964311622e-05, "loss": 0.0074, "step": 21664 }, { "epoch": 2.34876409366869, "grad_norm": 0.7636513113975525, "learning_rate": 2.1786595096474686e-05, "loss": 0.0161, "step": 21665 }, { "epoch": 2.34887250650477, "grad_norm": 0.5670487880706787, "learning_rate": 2.1782968228637747e-05, "loss": 0.0112, "step": 21666 }, { "epoch": 2.3489809193408497, "grad_norm": 0.5921527147293091, "learning_rate": 2.1779341360800815e-05, "loss": 0.0088, "step": 21667 }, { "epoch": 2.3490893321769297, "grad_norm": 0.3564683496952057, "learning_rate": 2.177571449296388e-05, "loss": 0.0478, "step": 21668 }, { "epoch": 2.3491977450130097, "grad_norm": 0.10806147009134293, "learning_rate": 2.1772087625126943e-05, "loss": 0.0083, "step": 21669 }, { "epoch": 2.3493061578490892, "grad_norm": 0.17234742641448975, "learning_rate": 2.1768460757290004e-05, "loss": 0.0079, "step": 21670 }, { "epoch": 2.3494145706851692, "grad_norm": 0.1754116415977478, "learning_rate": 2.176483388945307e-05, "loss": 0.0045, "step": 21671 }, { "epoch": 2.3495229835212488, "grad_norm": 0.08943270146846771, "learning_rate": 2.1761207021616135e-05, "loss": 0.003, "step": 21672 }, { "epoch": 2.3496313963573288, "grad_norm": 0.21065165102481842, "learning_rate": 2.1757580153779196e-05, "loss": 0.0073, "step": 21673 }, { "epoch": 2.3497398091934087, "grad_norm": 0.2309858202934265, "learning_rate": 2.175395328594226e-05, "loss": 0.0038, "step": 21674 }, { "epoch": 2.3498482220294883, "grad_norm": 0.04155096411705017, "learning_rate": 2.1750326418105324e-05, "loss": 0.0008, "step": 21675 }, { "epoch": 2.3499566348655683, "grad_norm": 0.12230045348405838, "learning_rate": 2.174669955026839e-05, "loss": 0.0097, "step": 21676 }, { "epoch": 2.350065047701648, "grad_norm": 0.13738510012626648, "learning_rate": 2.1743072682431452e-05, "loss": 0.0024, "step": 21677 }, { "epoch": 2.350173460537728, "grad_norm": 0.7201036214828491, "learning_rate": 2.1739445814594516e-05, "loss": 0.033, "step": 21678 }, { "epoch": 2.3502818733738073, "grad_norm": 0.34119686484336853, "learning_rate": 2.173581894675758e-05, "loss": 0.0168, "step": 21679 }, { "epoch": 2.3503902862098873, "grad_norm": 0.3256940245628357, "learning_rate": 2.1732192078920645e-05, "loss": 0.0361, "step": 21680 }, { "epoch": 2.350498699045967, "grad_norm": 0.4510301351547241, "learning_rate": 2.172856521108371e-05, "loss": 0.0361, "step": 21681 }, { "epoch": 2.350607111882047, "grad_norm": 0.37242552638053894, "learning_rate": 2.1724938343246773e-05, "loss": 0.0106, "step": 21682 }, { "epoch": 2.350715524718127, "grad_norm": 0.06668786704540253, "learning_rate": 2.1721311475409837e-05, "loss": 0.0016, "step": 21683 }, { "epoch": 2.3508239375542064, "grad_norm": 0.22818611562252045, "learning_rate": 2.17176846075729e-05, "loss": 0.0079, "step": 21684 }, { "epoch": 2.3509323503902864, "grad_norm": 0.3242584764957428, "learning_rate": 2.1714057739735965e-05, "loss": 0.0158, "step": 21685 }, { "epoch": 2.351040763226366, "grad_norm": 0.026747366413474083, "learning_rate": 2.171043087189903e-05, "loss": 0.0008, "step": 21686 }, { "epoch": 2.351149176062446, "grad_norm": 0.08427401632070541, "learning_rate": 2.1706804004062093e-05, "loss": 0.0052, "step": 21687 }, { "epoch": 2.3512575888985254, "grad_norm": 0.4528485834598541, "learning_rate": 2.1703177136225154e-05, "loss": 0.0037, "step": 21688 }, { "epoch": 2.3513660017346054, "grad_norm": 0.19995924830436707, "learning_rate": 2.169955026838822e-05, "loss": 0.0031, "step": 21689 }, { "epoch": 2.351474414570685, "grad_norm": 0.3560730814933777, "learning_rate": 2.1695923400551286e-05, "loss": 0.0097, "step": 21690 }, { "epoch": 2.351582827406765, "grad_norm": 0.5879050493240356, "learning_rate": 2.169229653271435e-05, "loss": 0.0141, "step": 21691 }, { "epoch": 2.351691240242845, "grad_norm": 0.18379664421081543, "learning_rate": 2.168866966487741e-05, "loss": 0.0042, "step": 21692 }, { "epoch": 2.3517996530789245, "grad_norm": 0.12625160813331604, "learning_rate": 2.1685042797040478e-05, "loss": 0.006, "step": 21693 }, { "epoch": 2.3519080659150045, "grad_norm": 0.4281415343284607, "learning_rate": 2.1681415929203542e-05, "loss": 0.0217, "step": 21694 }, { "epoch": 2.352016478751084, "grad_norm": 0.04873506352305412, "learning_rate": 2.1677789061366606e-05, "loss": 0.0018, "step": 21695 }, { "epoch": 2.352124891587164, "grad_norm": 0.9591158628463745, "learning_rate": 2.1674162193529667e-05, "loss": 0.015, "step": 21696 }, { "epoch": 2.3522333044232435, "grad_norm": 0.06412741541862488, "learning_rate": 2.1670535325692734e-05, "loss": 0.0027, "step": 21697 }, { "epoch": 2.3523417172593235, "grad_norm": 0.2491501122713089, "learning_rate": 2.16669084578558e-05, "loss": 0.0419, "step": 21698 }, { "epoch": 2.352450130095403, "grad_norm": 0.1085418090224266, "learning_rate": 2.166328159001886e-05, "loss": 0.009, "step": 21699 }, { "epoch": 2.352558542931483, "grad_norm": 0.4027385711669922, "learning_rate": 2.1659654722181923e-05, "loss": 0.0397, "step": 21700 }, { "epoch": 2.352666955767563, "grad_norm": 0.3127419054508209, "learning_rate": 2.1656027854344987e-05, "loss": 0.0056, "step": 21701 }, { "epoch": 2.3527753686036426, "grad_norm": 0.4836040437221527, "learning_rate": 2.1652400986508055e-05, "loss": 0.0104, "step": 21702 }, { "epoch": 2.3528837814397225, "grad_norm": 0.49531808495521545, "learning_rate": 2.1648774118671116e-05, "loss": 0.032, "step": 21703 }, { "epoch": 2.352992194275802, "grad_norm": 0.18430761992931366, "learning_rate": 2.164514725083418e-05, "loss": 0.0244, "step": 21704 }, { "epoch": 2.353100607111882, "grad_norm": 0.577063262462616, "learning_rate": 2.1641520382997244e-05, "loss": 0.0109, "step": 21705 }, { "epoch": 2.353209019947962, "grad_norm": 0.24055305123329163, "learning_rate": 2.1637893515160308e-05, "loss": 0.0043, "step": 21706 }, { "epoch": 2.3533174327840416, "grad_norm": 0.18297022581100464, "learning_rate": 2.1634266647323372e-05, "loss": 0.0071, "step": 21707 }, { "epoch": 2.3534258456201216, "grad_norm": 0.6557499766349792, "learning_rate": 2.1630639779486436e-05, "loss": 0.0365, "step": 21708 }, { "epoch": 2.353534258456201, "grad_norm": 0.15002533793449402, "learning_rate": 2.16270129116495e-05, "loss": 0.0112, "step": 21709 }, { "epoch": 2.353642671292281, "grad_norm": 0.7030441761016846, "learning_rate": 2.1623386043812564e-05, "loss": 0.0059, "step": 21710 }, { "epoch": 2.3537510841283606, "grad_norm": 0.062310706824064255, "learning_rate": 2.161975917597563e-05, "loss": 0.0047, "step": 21711 }, { "epoch": 2.3538594969644406, "grad_norm": 0.08341635018587112, "learning_rate": 2.1616132308138693e-05, "loss": 0.0026, "step": 21712 }, { "epoch": 2.35396790980052, "grad_norm": 0.4399484395980835, "learning_rate": 2.1612505440301757e-05, "loss": 0.0171, "step": 21713 }, { "epoch": 2.3540763226366, "grad_norm": 0.18947921693325043, "learning_rate": 2.1608878572464817e-05, "loss": 0.0101, "step": 21714 }, { "epoch": 2.35418473547268, "grad_norm": 0.05634130910038948, "learning_rate": 2.1605251704627885e-05, "loss": 0.0033, "step": 21715 }, { "epoch": 2.3542931483087597, "grad_norm": 0.16298630833625793, "learning_rate": 2.160162483679095e-05, "loss": 0.015, "step": 21716 }, { "epoch": 2.3544015611448397, "grad_norm": 0.7926329374313354, "learning_rate": 2.1597997968954013e-05, "loss": 0.0074, "step": 21717 }, { "epoch": 2.354509973980919, "grad_norm": 0.09186424314975739, "learning_rate": 2.1594371101117074e-05, "loss": 0.0069, "step": 21718 }, { "epoch": 2.354618386816999, "grad_norm": 0.1421530693769455, "learning_rate": 2.159074423328014e-05, "loss": 0.0159, "step": 21719 }, { "epoch": 2.3547267996530787, "grad_norm": 0.1741170585155487, "learning_rate": 2.1587117365443205e-05, "loss": 0.007, "step": 21720 }, { "epoch": 2.3548352124891587, "grad_norm": 0.0429743155837059, "learning_rate": 2.158349049760627e-05, "loss": 0.0026, "step": 21721 }, { "epoch": 2.3549436253252383, "grad_norm": 0.014853385277092457, "learning_rate": 2.157986362976933e-05, "loss": 0.0006, "step": 21722 }, { "epoch": 2.3550520381613183, "grad_norm": 0.05383583903312683, "learning_rate": 2.1576236761932398e-05, "loss": 0.0019, "step": 21723 }, { "epoch": 2.3551604509973982, "grad_norm": 0.25183066725730896, "learning_rate": 2.1572609894095462e-05, "loss": 0.0109, "step": 21724 }, { "epoch": 2.3552688638334778, "grad_norm": 0.02456754446029663, "learning_rate": 2.1568983026258523e-05, "loss": 0.0007, "step": 21725 }, { "epoch": 2.3553772766695578, "grad_norm": 0.22772181034088135, "learning_rate": 2.1565356158421587e-05, "loss": 0.0094, "step": 21726 }, { "epoch": 2.3554856895056373, "grad_norm": 0.3955250680446625, "learning_rate": 2.156172929058465e-05, "loss": 0.0062, "step": 21727 }, { "epoch": 2.3555941023417173, "grad_norm": 0.046103015542030334, "learning_rate": 2.155810242274772e-05, "loss": 0.0017, "step": 21728 }, { "epoch": 2.3557025151777973, "grad_norm": 0.647631824016571, "learning_rate": 2.155447555491078e-05, "loss": 0.046, "step": 21729 }, { "epoch": 2.355810928013877, "grad_norm": 0.25226956605911255, "learning_rate": 2.1550848687073843e-05, "loss": 0.0207, "step": 21730 }, { "epoch": 2.355919340849957, "grad_norm": 0.46524539589881897, "learning_rate": 2.1547221819236907e-05, "loss": 0.0188, "step": 21731 }, { "epoch": 2.3560277536860363, "grad_norm": 0.4105837345123291, "learning_rate": 2.154359495139997e-05, "loss": 0.0148, "step": 21732 }, { "epoch": 2.3561361665221163, "grad_norm": 0.024677699431777, "learning_rate": 2.1539968083563035e-05, "loss": 0.0006, "step": 21733 }, { "epoch": 2.356244579358196, "grad_norm": 0.04854535311460495, "learning_rate": 2.15363412157261e-05, "loss": 0.0018, "step": 21734 }, { "epoch": 2.356352992194276, "grad_norm": 0.12775525450706482, "learning_rate": 2.1532714347889164e-05, "loss": 0.002, "step": 21735 }, { "epoch": 2.3564614050303554, "grad_norm": 0.37202683091163635, "learning_rate": 2.1529087480052228e-05, "loss": 0.0197, "step": 21736 }, { "epoch": 2.3565698178664354, "grad_norm": 0.36494070291519165, "learning_rate": 2.1525460612215292e-05, "loss": 0.0379, "step": 21737 }, { "epoch": 2.3566782307025154, "grad_norm": 0.3541432321071625, "learning_rate": 2.1521833744378356e-05, "loss": 0.014, "step": 21738 }, { "epoch": 2.356786643538595, "grad_norm": 0.11566077172756195, "learning_rate": 2.151820687654142e-05, "loss": 0.0023, "step": 21739 }, { "epoch": 2.356895056374675, "grad_norm": 0.20350980758666992, "learning_rate": 2.1514580008704484e-05, "loss": 0.0028, "step": 21740 }, { "epoch": 2.3570034692107544, "grad_norm": 0.0031376765109598637, "learning_rate": 2.151095314086755e-05, "loss": 0.0002, "step": 21741 }, { "epoch": 2.3571118820468344, "grad_norm": 0.2811061143875122, "learning_rate": 2.1507326273030612e-05, "loss": 0.0165, "step": 21742 }, { "epoch": 2.357220294882914, "grad_norm": 0.6075887680053711, "learning_rate": 2.1503699405193677e-05, "loss": 0.0309, "step": 21743 }, { "epoch": 2.357328707718994, "grad_norm": 0.5455458760261536, "learning_rate": 2.1500072537356737e-05, "loss": 0.0214, "step": 21744 }, { "epoch": 2.3574371205550735, "grad_norm": 1.1952996253967285, "learning_rate": 2.1496445669519805e-05, "loss": 0.031, "step": 21745 }, { "epoch": 2.3575455333911535, "grad_norm": 0.425828218460083, "learning_rate": 2.149281880168287e-05, "loss": 0.0162, "step": 21746 }, { "epoch": 2.3576539462272335, "grad_norm": 1.2739596366882324, "learning_rate": 2.1489191933845933e-05, "loss": 0.0361, "step": 21747 }, { "epoch": 2.357762359063313, "grad_norm": 0.24841178953647614, "learning_rate": 2.1485565066008994e-05, "loss": 0.0054, "step": 21748 }, { "epoch": 2.357870771899393, "grad_norm": 0.02942705899477005, "learning_rate": 2.148193819817206e-05, "loss": 0.0017, "step": 21749 }, { "epoch": 2.3579791847354725, "grad_norm": 0.3747456669807434, "learning_rate": 2.1478311330335125e-05, "loss": 0.008, "step": 21750 }, { "epoch": 2.3580875975715525, "grad_norm": 0.07527102530002594, "learning_rate": 2.1474684462498186e-05, "loss": 0.0023, "step": 21751 }, { "epoch": 2.3581960104076325, "grad_norm": 0.1829969733953476, "learning_rate": 2.147105759466125e-05, "loss": 0.0266, "step": 21752 }, { "epoch": 2.358304423243712, "grad_norm": 0.06411745399236679, "learning_rate": 2.1467430726824318e-05, "loss": 0.0016, "step": 21753 }, { "epoch": 2.358412836079792, "grad_norm": 0.01236570905894041, "learning_rate": 2.146380385898738e-05, "loss": 0.0004, "step": 21754 }, { "epoch": 2.3585212489158716, "grad_norm": 0.19159868359565735, "learning_rate": 2.1460176991150442e-05, "loss": 0.0108, "step": 21755 }, { "epoch": 2.3586296617519515, "grad_norm": 0.03335253894329071, "learning_rate": 2.1456550123313507e-05, "loss": 0.0018, "step": 21756 }, { "epoch": 2.358738074588031, "grad_norm": 0.18909673392772675, "learning_rate": 2.145292325547657e-05, "loss": 0.0038, "step": 21757 }, { "epoch": 2.358846487424111, "grad_norm": 0.0249034333974123, "learning_rate": 2.1449296387639635e-05, "loss": 0.0009, "step": 21758 }, { "epoch": 2.3589549002601906, "grad_norm": 0.3769148588180542, "learning_rate": 2.14456695198027e-05, "loss": 0.0157, "step": 21759 }, { "epoch": 2.3590633130962706, "grad_norm": 0.24087639153003693, "learning_rate": 2.1442042651965763e-05, "loss": 0.0088, "step": 21760 }, { "epoch": 2.3591717259323506, "grad_norm": 0.44056180119514465, "learning_rate": 2.1438415784128827e-05, "loss": 0.0186, "step": 21761 }, { "epoch": 2.35928013876843, "grad_norm": 0.06747200340032578, "learning_rate": 2.143478891629189e-05, "loss": 0.0025, "step": 21762 }, { "epoch": 2.35938855160451, "grad_norm": 0.06891320645809174, "learning_rate": 2.1431162048454955e-05, "loss": 0.006, "step": 21763 }, { "epoch": 2.3594969644405897, "grad_norm": 0.013334178365767002, "learning_rate": 2.142753518061802e-05, "loss": 0.0004, "step": 21764 }, { "epoch": 2.3596053772766696, "grad_norm": 0.30665817856788635, "learning_rate": 2.1423908312781083e-05, "loss": 0.0043, "step": 21765 }, { "epoch": 2.359713790112749, "grad_norm": 0.1311628371477127, "learning_rate": 2.1420281444944148e-05, "loss": 0.0034, "step": 21766 }, { "epoch": 2.359822202948829, "grad_norm": 0.22633381187915802, "learning_rate": 2.141665457710721e-05, "loss": 0.0079, "step": 21767 }, { "epoch": 2.3599306157849087, "grad_norm": 0.07716207951307297, "learning_rate": 2.1413027709270276e-05, "loss": 0.0019, "step": 21768 }, { "epoch": 2.3600390286209887, "grad_norm": 0.05439882352948189, "learning_rate": 2.140940084143334e-05, "loss": 0.002, "step": 21769 }, { "epoch": 2.3601474414570687, "grad_norm": 0.008486775681376457, "learning_rate": 2.14057739735964e-05, "loss": 0.0003, "step": 21770 }, { "epoch": 2.360255854293148, "grad_norm": 0.11215853691101074, "learning_rate": 2.1402147105759468e-05, "loss": 0.0045, "step": 21771 }, { "epoch": 2.360364267129228, "grad_norm": 0.12139536440372467, "learning_rate": 2.1398520237922532e-05, "loss": 0.0043, "step": 21772 }, { "epoch": 2.3604726799653077, "grad_norm": 0.7157633900642395, "learning_rate": 2.1394893370085596e-05, "loss": 0.0605, "step": 21773 }, { "epoch": 2.3605810928013877, "grad_norm": 0.32551583647727966, "learning_rate": 2.1391266502248657e-05, "loss": 0.0097, "step": 21774 }, { "epoch": 2.3606895056374677, "grad_norm": 0.5899133682250977, "learning_rate": 2.1387639634411724e-05, "loss": 0.0229, "step": 21775 }, { "epoch": 2.3607979184735473, "grad_norm": 0.13309043645858765, "learning_rate": 2.138401276657479e-05, "loss": 0.0026, "step": 21776 }, { "epoch": 2.3609063313096272, "grad_norm": 0.10275395959615707, "learning_rate": 2.138038589873785e-05, "loss": 0.0093, "step": 21777 }, { "epoch": 2.361014744145707, "grad_norm": 0.18525128066539764, "learning_rate": 2.1376759030900913e-05, "loss": 0.0099, "step": 21778 }, { "epoch": 2.3611231569817868, "grad_norm": 0.06401802599430084, "learning_rate": 2.137313216306398e-05, "loss": 0.0016, "step": 21779 }, { "epoch": 2.3612315698178663, "grad_norm": 0.27773186564445496, "learning_rate": 2.1369505295227045e-05, "loss": 0.0102, "step": 21780 }, { "epoch": 2.3613399826539463, "grad_norm": 0.1390756517648697, "learning_rate": 2.1365878427390106e-05, "loss": 0.0085, "step": 21781 }, { "epoch": 2.361448395490026, "grad_norm": 0.02646363154053688, "learning_rate": 2.136225155955317e-05, "loss": 0.0012, "step": 21782 }, { "epoch": 2.361556808326106, "grad_norm": 0.08398012071847916, "learning_rate": 2.1358624691716234e-05, "loss": 0.0032, "step": 21783 }, { "epoch": 2.361665221162186, "grad_norm": 0.3063156008720398, "learning_rate": 2.1354997823879298e-05, "loss": 0.0293, "step": 21784 }, { "epoch": 2.3617736339982653, "grad_norm": 0.33835428953170776, "learning_rate": 2.1351370956042362e-05, "loss": 0.0058, "step": 21785 }, { "epoch": 2.3618820468343453, "grad_norm": 1.0557096004486084, "learning_rate": 2.1347744088205426e-05, "loss": 0.0214, "step": 21786 }, { "epoch": 2.361990459670425, "grad_norm": 0.08063042163848877, "learning_rate": 2.134411722036849e-05, "loss": 0.0013, "step": 21787 }, { "epoch": 2.362098872506505, "grad_norm": 0.38710498809814453, "learning_rate": 2.1340490352531554e-05, "loss": 0.0144, "step": 21788 }, { "epoch": 2.3622072853425844, "grad_norm": 0.19190730154514313, "learning_rate": 2.133686348469462e-05, "loss": 0.0107, "step": 21789 }, { "epoch": 2.3623156981786644, "grad_norm": 0.27350908517837524, "learning_rate": 2.1333236616857683e-05, "loss": 0.0075, "step": 21790 }, { "epoch": 2.362424111014744, "grad_norm": 0.25873610377311707, "learning_rate": 2.1329609749020747e-05, "loss": 0.0195, "step": 21791 }, { "epoch": 2.362532523850824, "grad_norm": 0.2940765917301178, "learning_rate": 2.132598288118381e-05, "loss": 0.007, "step": 21792 }, { "epoch": 2.362640936686904, "grad_norm": 0.377023845911026, "learning_rate": 2.1322356013346875e-05, "loss": 0.0192, "step": 21793 }, { "epoch": 2.3627493495229834, "grad_norm": 0.017746232450008392, "learning_rate": 2.131872914550994e-05, "loss": 0.0003, "step": 21794 }, { "epoch": 2.3628577623590634, "grad_norm": 0.30400800704956055, "learning_rate": 2.1315102277673003e-05, "loss": 0.0103, "step": 21795 }, { "epoch": 2.362966175195143, "grad_norm": 0.03956077992916107, "learning_rate": 2.1311475409836064e-05, "loss": 0.0029, "step": 21796 }, { "epoch": 2.363074588031223, "grad_norm": 0.4183410704135895, "learning_rate": 2.130784854199913e-05, "loss": 0.0092, "step": 21797 }, { "epoch": 2.363183000867303, "grad_norm": 0.09232469648122787, "learning_rate": 2.1304221674162196e-05, "loss": 0.0037, "step": 21798 }, { "epoch": 2.3632914137033825, "grad_norm": 0.1586833894252777, "learning_rate": 2.130059480632526e-05, "loss": 0.0065, "step": 21799 }, { "epoch": 2.3633998265394625, "grad_norm": 0.40809381008148193, "learning_rate": 2.129696793848832e-05, "loss": 0.0509, "step": 21800 }, { "epoch": 2.363508239375542, "grad_norm": 0.22082529962062836, "learning_rate": 2.1293341070651388e-05, "loss": 0.0153, "step": 21801 }, { "epoch": 2.363616652211622, "grad_norm": 0.33949869871139526, "learning_rate": 2.1289714202814452e-05, "loss": 0.0104, "step": 21802 }, { "epoch": 2.3637250650477015, "grad_norm": 0.4759007692337036, "learning_rate": 2.1286087334977513e-05, "loss": 0.0156, "step": 21803 }, { "epoch": 2.3638334778837815, "grad_norm": 0.4956602156162262, "learning_rate": 2.1282460467140577e-05, "loss": 0.0079, "step": 21804 }, { "epoch": 2.363941890719861, "grad_norm": 0.017482614144682884, "learning_rate": 2.1278833599303644e-05, "loss": 0.0013, "step": 21805 }, { "epoch": 2.364050303555941, "grad_norm": 0.07806036621332169, "learning_rate": 2.127520673146671e-05, "loss": 0.0041, "step": 21806 }, { "epoch": 2.364158716392021, "grad_norm": 0.6664482355117798, "learning_rate": 2.127157986362977e-05, "loss": 0.008, "step": 21807 }, { "epoch": 2.3642671292281006, "grad_norm": 0.4577501118183136, "learning_rate": 2.1267952995792833e-05, "loss": 0.0111, "step": 21808 }, { "epoch": 2.3643755420641805, "grad_norm": 0.014001202769577503, "learning_rate": 2.1264326127955897e-05, "loss": 0.0006, "step": 21809 }, { "epoch": 2.36448395490026, "grad_norm": 0.0036827954463660717, "learning_rate": 2.126069926011896e-05, "loss": 0.0002, "step": 21810 }, { "epoch": 2.36459236773634, "grad_norm": 0.308401882648468, "learning_rate": 2.1257072392282026e-05, "loss": 0.0203, "step": 21811 }, { "epoch": 2.3647007805724196, "grad_norm": 0.21437779068946838, "learning_rate": 2.125344552444509e-05, "loss": 0.0131, "step": 21812 }, { "epoch": 2.3648091934084996, "grad_norm": 0.28407806158065796, "learning_rate": 2.1249818656608154e-05, "loss": 0.0139, "step": 21813 }, { "epoch": 2.364917606244579, "grad_norm": 0.040464162826538086, "learning_rate": 2.1246191788771218e-05, "loss": 0.0024, "step": 21814 }, { "epoch": 2.365026019080659, "grad_norm": 0.038629237562417984, "learning_rate": 2.1242564920934282e-05, "loss": 0.0019, "step": 21815 }, { "epoch": 2.365134431916739, "grad_norm": 0.04392495006322861, "learning_rate": 2.1238938053097346e-05, "loss": 0.0015, "step": 21816 }, { "epoch": 2.3652428447528187, "grad_norm": 0.26567164063453674, "learning_rate": 2.123531118526041e-05, "loss": 0.0151, "step": 21817 }, { "epoch": 2.3653512575888986, "grad_norm": 0.24568015336990356, "learning_rate": 2.1231684317423474e-05, "loss": 0.0186, "step": 21818 }, { "epoch": 2.365459670424978, "grad_norm": 0.29886898398399353, "learning_rate": 2.122805744958654e-05, "loss": 0.014, "step": 21819 }, { "epoch": 2.365568083261058, "grad_norm": 0.3380885124206543, "learning_rate": 2.1224430581749602e-05, "loss": 0.0026, "step": 21820 }, { "epoch": 2.365676496097138, "grad_norm": 0.1651502102613449, "learning_rate": 2.1220803713912667e-05, "loss": 0.0181, "step": 21821 }, { "epoch": 2.3657849089332177, "grad_norm": 0.7928196787834167, "learning_rate": 2.1217176846075727e-05, "loss": 0.0198, "step": 21822 }, { "epoch": 2.3658933217692977, "grad_norm": 0.2856834828853607, "learning_rate": 2.1213549978238795e-05, "loss": 0.0188, "step": 21823 }, { "epoch": 2.366001734605377, "grad_norm": 0.12185011804103851, "learning_rate": 2.120992311040186e-05, "loss": 0.0024, "step": 21824 }, { "epoch": 2.366110147441457, "grad_norm": 0.14527559280395508, "learning_rate": 2.1206296242564923e-05, "loss": 0.0082, "step": 21825 }, { "epoch": 2.3662185602775367, "grad_norm": 0.12363912165164948, "learning_rate": 2.1202669374727984e-05, "loss": 0.0035, "step": 21826 }, { "epoch": 2.3663269731136167, "grad_norm": 0.028311606496572495, "learning_rate": 2.119904250689105e-05, "loss": 0.0006, "step": 21827 }, { "epoch": 2.3664353859496963, "grad_norm": 0.1416434794664383, "learning_rate": 2.1195415639054115e-05, "loss": 0.0092, "step": 21828 }, { "epoch": 2.3665437987857763, "grad_norm": 0.08802976459264755, "learning_rate": 2.1191788771217176e-05, "loss": 0.0068, "step": 21829 }, { "epoch": 2.3666522116218562, "grad_norm": 0.32084786891937256, "learning_rate": 2.118816190338024e-05, "loss": 0.0064, "step": 21830 }, { "epoch": 2.366760624457936, "grad_norm": 0.5255085229873657, "learning_rate": 2.1184535035543308e-05, "loss": 0.0255, "step": 21831 }, { "epoch": 2.3668690372940158, "grad_norm": 0.3831707835197449, "learning_rate": 2.1180908167706372e-05, "loss": 0.0254, "step": 21832 }, { "epoch": 2.3669774501300953, "grad_norm": 1.1530624628067017, "learning_rate": 2.1177281299869432e-05, "loss": 0.0443, "step": 21833 }, { "epoch": 2.3670858629661753, "grad_norm": 0.9484485983848572, "learning_rate": 2.1173654432032497e-05, "loss": 0.0339, "step": 21834 }, { "epoch": 2.367194275802255, "grad_norm": 0.07114075869321823, "learning_rate": 2.1170027564195564e-05, "loss": 0.0013, "step": 21835 }, { "epoch": 2.367302688638335, "grad_norm": 0.7387588620185852, "learning_rate": 2.1166400696358625e-05, "loss": 0.0204, "step": 21836 }, { "epoch": 2.3674111014744144, "grad_norm": 0.5109208822250366, "learning_rate": 2.116277382852169e-05, "loss": 0.0199, "step": 21837 }, { "epoch": 2.3675195143104943, "grad_norm": 0.2163952738046646, "learning_rate": 2.1159146960684753e-05, "loss": 0.0169, "step": 21838 }, { "epoch": 2.3676279271465743, "grad_norm": 0.07022111117839813, "learning_rate": 2.1155520092847817e-05, "loss": 0.0037, "step": 21839 }, { "epoch": 2.367736339982654, "grad_norm": 0.4253091514110565, "learning_rate": 2.115189322501088e-05, "loss": 0.0103, "step": 21840 }, { "epoch": 2.367844752818734, "grad_norm": 0.025669803842902184, "learning_rate": 2.1148266357173945e-05, "loss": 0.0012, "step": 21841 }, { "epoch": 2.3679531656548134, "grad_norm": 0.3989272117614746, "learning_rate": 2.114463948933701e-05, "loss": 0.0247, "step": 21842 }, { "epoch": 2.3680615784908934, "grad_norm": 0.041683778166770935, "learning_rate": 2.1141012621500073e-05, "loss": 0.0024, "step": 21843 }, { "epoch": 2.368169991326973, "grad_norm": 0.05057619884610176, "learning_rate": 2.1137385753663138e-05, "loss": 0.002, "step": 21844 }, { "epoch": 2.368278404163053, "grad_norm": 0.19451303780078888, "learning_rate": 2.11337588858262e-05, "loss": 0.0076, "step": 21845 }, { "epoch": 2.3683868169991324, "grad_norm": 0.2729867994785309, "learning_rate": 2.1130132017989266e-05, "loss": 0.0214, "step": 21846 }, { "epoch": 2.3684952298352124, "grad_norm": 0.6891353130340576, "learning_rate": 2.112650515015233e-05, "loss": 0.0065, "step": 21847 }, { "epoch": 2.3686036426712924, "grad_norm": 0.44570815563201904, "learning_rate": 2.1122878282315394e-05, "loss": 0.0071, "step": 21848 }, { "epoch": 2.368712055507372, "grad_norm": 0.36314985156059265, "learning_rate": 2.1119251414478458e-05, "loss": 0.0331, "step": 21849 }, { "epoch": 2.368820468343452, "grad_norm": 0.28482702374458313, "learning_rate": 2.1115624546641522e-05, "loss": 0.0081, "step": 21850 }, { "epoch": 2.3689288811795315, "grad_norm": 0.9398401975631714, "learning_rate": 2.1111997678804586e-05, "loss": 0.0118, "step": 21851 }, { "epoch": 2.3690372940156115, "grad_norm": 0.5973455905914307, "learning_rate": 2.1108370810967647e-05, "loss": 0.032, "step": 21852 }, { "epoch": 2.3691457068516915, "grad_norm": 0.25059229135513306, "learning_rate": 2.1104743943130715e-05, "loss": 0.023, "step": 21853 }, { "epoch": 2.369254119687771, "grad_norm": 0.4857873320579529, "learning_rate": 2.110111707529378e-05, "loss": 0.0219, "step": 21854 }, { "epoch": 2.369362532523851, "grad_norm": 0.4408752918243408, "learning_rate": 2.109749020745684e-05, "loss": 0.0187, "step": 21855 }, { "epoch": 2.3694709453599305, "grad_norm": 0.06594621390104294, "learning_rate": 2.1093863339619903e-05, "loss": 0.002, "step": 21856 }, { "epoch": 2.3695793581960105, "grad_norm": 0.1107528805732727, "learning_rate": 2.109023647178297e-05, "loss": 0.0019, "step": 21857 }, { "epoch": 2.36968777103209, "grad_norm": 0.04458941891789436, "learning_rate": 2.1086609603946035e-05, "loss": 0.0011, "step": 21858 }, { "epoch": 2.36979618386817, "grad_norm": 0.03794051706790924, "learning_rate": 2.1082982736109096e-05, "loss": 0.0012, "step": 21859 }, { "epoch": 2.3699045967042496, "grad_norm": 0.49074822664260864, "learning_rate": 2.107935586827216e-05, "loss": 0.0211, "step": 21860 }, { "epoch": 2.3700130095403296, "grad_norm": 0.3442274034023285, "learning_rate": 2.1075729000435227e-05, "loss": 0.0116, "step": 21861 }, { "epoch": 2.3701214223764095, "grad_norm": 0.6154758334159851, "learning_rate": 2.107210213259829e-05, "loss": 0.0316, "step": 21862 }, { "epoch": 2.370229835212489, "grad_norm": 0.11706939339637756, "learning_rate": 2.1068475264761352e-05, "loss": 0.0036, "step": 21863 }, { "epoch": 2.370338248048569, "grad_norm": 0.2606600224971771, "learning_rate": 2.1064848396924416e-05, "loss": 0.0031, "step": 21864 }, { "epoch": 2.3704466608846486, "grad_norm": 0.07824672758579254, "learning_rate": 2.106122152908748e-05, "loss": 0.0029, "step": 21865 }, { "epoch": 2.3705550737207286, "grad_norm": 0.5411006808280945, "learning_rate": 2.1057594661250545e-05, "loss": 0.0275, "step": 21866 }, { "epoch": 2.370663486556808, "grad_norm": 0.05982557311654091, "learning_rate": 2.105396779341361e-05, "loss": 0.0008, "step": 21867 }, { "epoch": 2.370771899392888, "grad_norm": 0.027625873684883118, "learning_rate": 2.1050340925576673e-05, "loss": 0.0008, "step": 21868 }, { "epoch": 2.3708803122289677, "grad_norm": 0.04722397029399872, "learning_rate": 2.1046714057739737e-05, "loss": 0.0012, "step": 21869 }, { "epoch": 2.3709887250650477, "grad_norm": 0.0638367235660553, "learning_rate": 2.10430871899028e-05, "loss": 0.0026, "step": 21870 }, { "epoch": 2.3710971379011276, "grad_norm": 0.039885006844997406, "learning_rate": 2.1039460322065865e-05, "loss": 0.0017, "step": 21871 }, { "epoch": 2.371205550737207, "grad_norm": 0.025452932342886925, "learning_rate": 2.103583345422893e-05, "loss": 0.0008, "step": 21872 }, { "epoch": 2.371313963573287, "grad_norm": 0.016358280554413795, "learning_rate": 2.1032206586391993e-05, "loss": 0.0005, "step": 21873 }, { "epoch": 2.3714223764093667, "grad_norm": 0.26329588890075684, "learning_rate": 2.1028579718555057e-05, "loss": 0.0115, "step": 21874 }, { "epoch": 2.3715307892454467, "grad_norm": 1.2829320430755615, "learning_rate": 2.102495285071812e-05, "loss": 0.0155, "step": 21875 }, { "epoch": 2.3716392020815267, "grad_norm": 0.04224204644560814, "learning_rate": 2.1021325982881186e-05, "loss": 0.0015, "step": 21876 }, { "epoch": 2.371747614917606, "grad_norm": 0.2506833076477051, "learning_rate": 2.101769911504425e-05, "loss": 0.0036, "step": 21877 }, { "epoch": 2.371856027753686, "grad_norm": 0.05359607934951782, "learning_rate": 2.101407224720731e-05, "loss": 0.0015, "step": 21878 }, { "epoch": 2.3719644405897657, "grad_norm": 0.32023051381111145, "learning_rate": 2.1010445379370378e-05, "loss": 0.0252, "step": 21879 }, { "epoch": 2.3720728534258457, "grad_norm": 0.5112828016281128, "learning_rate": 2.1006818511533442e-05, "loss": 0.0373, "step": 21880 }, { "epoch": 2.3721812662619253, "grad_norm": 0.32184380292892456, "learning_rate": 2.1003191643696503e-05, "loss": 0.0125, "step": 21881 }, { "epoch": 2.3722896790980053, "grad_norm": 0.43854886293411255, "learning_rate": 2.0999564775859567e-05, "loss": 0.0114, "step": 21882 }, { "epoch": 2.372398091934085, "grad_norm": 0.01031789556145668, "learning_rate": 2.0995937908022634e-05, "loss": 0.0003, "step": 21883 }, { "epoch": 2.372506504770165, "grad_norm": 0.12809757888317108, "learning_rate": 2.09923110401857e-05, "loss": 0.0038, "step": 21884 }, { "epoch": 2.3726149176062448, "grad_norm": 0.2621270716190338, "learning_rate": 2.098868417234876e-05, "loss": 0.0045, "step": 21885 }, { "epoch": 2.3727233304423243, "grad_norm": 0.8807539343833923, "learning_rate": 2.0985057304511823e-05, "loss": 0.0209, "step": 21886 }, { "epoch": 2.3728317432784043, "grad_norm": 0.21625466644763947, "learning_rate": 2.098143043667489e-05, "loss": 0.0157, "step": 21887 }, { "epoch": 2.372940156114484, "grad_norm": 0.33454659581184387, "learning_rate": 2.0977803568837955e-05, "loss": 0.0502, "step": 21888 }, { "epoch": 2.373048568950564, "grad_norm": 0.3151574730873108, "learning_rate": 2.0974176701001016e-05, "loss": 0.0349, "step": 21889 }, { "epoch": 2.3731569817866434, "grad_norm": 0.051048923283815384, "learning_rate": 2.097054983316408e-05, "loss": 0.003, "step": 21890 }, { "epoch": 2.3732653946227233, "grad_norm": 0.9106405973434448, "learning_rate": 2.0966922965327144e-05, "loss": 0.0503, "step": 21891 }, { "epoch": 2.373373807458803, "grad_norm": 0.026377031579613686, "learning_rate": 2.0963296097490208e-05, "loss": 0.0009, "step": 21892 }, { "epoch": 2.373482220294883, "grad_norm": 0.32839280366897583, "learning_rate": 2.0959669229653272e-05, "loss": 0.0131, "step": 21893 }, { "epoch": 2.373590633130963, "grad_norm": 0.12248660624027252, "learning_rate": 2.0956042361816336e-05, "loss": 0.0052, "step": 21894 }, { "epoch": 2.3736990459670424, "grad_norm": 0.32509884238243103, "learning_rate": 2.09524154939794e-05, "loss": 0.0081, "step": 21895 }, { "epoch": 2.3738074588031224, "grad_norm": 0.03546901419758797, "learning_rate": 2.0948788626142464e-05, "loss": 0.0021, "step": 21896 }, { "epoch": 2.373915871639202, "grad_norm": 0.18260148167610168, "learning_rate": 2.094516175830553e-05, "loss": 0.0065, "step": 21897 }, { "epoch": 2.374024284475282, "grad_norm": 0.0840575248003006, "learning_rate": 2.0941534890468592e-05, "loss": 0.0039, "step": 21898 }, { "epoch": 2.374132697311362, "grad_norm": 0.48656368255615234, "learning_rate": 2.0937908022631657e-05, "loss": 0.0191, "step": 21899 }, { "epoch": 2.3742411101474414, "grad_norm": 0.16368553042411804, "learning_rate": 2.093428115479472e-05, "loss": 0.0076, "step": 21900 }, { "epoch": 2.3743495229835214, "grad_norm": 0.33200040459632874, "learning_rate": 2.0930654286957785e-05, "loss": 0.0288, "step": 21901 }, { "epoch": 2.374457935819601, "grad_norm": 0.6891906261444092, "learning_rate": 2.092702741912085e-05, "loss": 0.0183, "step": 21902 }, { "epoch": 2.374566348655681, "grad_norm": 0.4170505106449127, "learning_rate": 2.0923400551283913e-05, "loss": 0.0061, "step": 21903 }, { "epoch": 2.3746747614917605, "grad_norm": 0.17609813809394836, "learning_rate": 2.0919773683446974e-05, "loss": 0.0049, "step": 21904 }, { "epoch": 2.3747831743278405, "grad_norm": 0.36204463243484497, "learning_rate": 2.091614681561004e-05, "loss": 0.0217, "step": 21905 }, { "epoch": 2.37489158716392, "grad_norm": 0.023164784535765648, "learning_rate": 2.0912519947773105e-05, "loss": 0.0007, "step": 21906 }, { "epoch": 2.375, "grad_norm": 0.09147393703460693, "learning_rate": 2.0908893079936166e-05, "loss": 0.0075, "step": 21907 }, { "epoch": 2.37510841283608, "grad_norm": 0.24553918838500977, "learning_rate": 2.090526621209923e-05, "loss": 0.0078, "step": 21908 }, { "epoch": 2.3752168256721595, "grad_norm": 0.008050776086747646, "learning_rate": 2.0901639344262298e-05, "loss": 0.0003, "step": 21909 }, { "epoch": 2.3753252385082395, "grad_norm": 0.0029821048956364393, "learning_rate": 2.0898012476425362e-05, "loss": 0.0002, "step": 21910 }, { "epoch": 2.375433651344319, "grad_norm": 0.0235946886241436, "learning_rate": 2.0894385608588422e-05, "loss": 0.0009, "step": 21911 }, { "epoch": 2.375542064180399, "grad_norm": 0.36681291460990906, "learning_rate": 2.0890758740751487e-05, "loss": 0.0083, "step": 21912 }, { "epoch": 2.3756504770164786, "grad_norm": 0.5710471272468567, "learning_rate": 2.0887131872914554e-05, "loss": 0.0481, "step": 21913 }, { "epoch": 2.3757588898525586, "grad_norm": 0.4055030345916748, "learning_rate": 2.0883505005077618e-05, "loss": 0.0103, "step": 21914 }, { "epoch": 2.375867302688638, "grad_norm": 0.2978963851928711, "learning_rate": 2.087987813724068e-05, "loss": 0.0184, "step": 21915 }, { "epoch": 2.375975715524718, "grad_norm": 0.4673668444156647, "learning_rate": 2.0876251269403743e-05, "loss": 0.0056, "step": 21916 }, { "epoch": 2.376084128360798, "grad_norm": 0.27464550733566284, "learning_rate": 2.0872624401566807e-05, "loss": 0.0329, "step": 21917 }, { "epoch": 2.3761925411968776, "grad_norm": 0.28501439094543457, "learning_rate": 2.086899753372987e-05, "loss": 0.0125, "step": 21918 }, { "epoch": 2.3763009540329576, "grad_norm": 0.224089577794075, "learning_rate": 2.0865370665892935e-05, "loss": 0.0184, "step": 21919 }, { "epoch": 2.376409366869037, "grad_norm": 0.05799476429820061, "learning_rate": 2.0861743798056e-05, "loss": 0.0039, "step": 21920 }, { "epoch": 2.376517779705117, "grad_norm": 0.10079358518123627, "learning_rate": 2.0858116930219064e-05, "loss": 0.0027, "step": 21921 }, { "epoch": 2.376626192541197, "grad_norm": 0.030141044408082962, "learning_rate": 2.0854490062382128e-05, "loss": 0.0013, "step": 21922 }, { "epoch": 2.3767346053772767, "grad_norm": 0.07851611822843552, "learning_rate": 2.0850863194545192e-05, "loss": 0.0022, "step": 21923 }, { "epoch": 2.3768430182133566, "grad_norm": 0.15996576845645905, "learning_rate": 2.0847236326708256e-05, "loss": 0.0024, "step": 21924 }, { "epoch": 2.376951431049436, "grad_norm": 1.639017939567566, "learning_rate": 2.084360945887132e-05, "loss": 0.0201, "step": 21925 }, { "epoch": 2.377059843885516, "grad_norm": 0.34055954217910767, "learning_rate": 2.0839982591034384e-05, "loss": 0.0191, "step": 21926 }, { "epoch": 2.3771682567215957, "grad_norm": 0.27797889709472656, "learning_rate": 2.0836355723197448e-05, "loss": 0.0068, "step": 21927 }, { "epoch": 2.3772766695576757, "grad_norm": 0.35334450006484985, "learning_rate": 2.0832728855360512e-05, "loss": 0.0162, "step": 21928 }, { "epoch": 2.3773850823937552, "grad_norm": 0.012788660824298859, "learning_rate": 2.0829101987523576e-05, "loss": 0.0003, "step": 21929 }, { "epoch": 2.377493495229835, "grad_norm": 0.1100567877292633, "learning_rate": 2.082547511968664e-05, "loss": 0.005, "step": 21930 }, { "epoch": 2.377601908065915, "grad_norm": 0.07438690960407257, "learning_rate": 2.0821848251849705e-05, "loss": 0.0021, "step": 21931 }, { "epoch": 2.3777103209019947, "grad_norm": 0.18819649517536163, "learning_rate": 2.081822138401277e-05, "loss": 0.02, "step": 21932 }, { "epoch": 2.3778187337380747, "grad_norm": 0.023707294836640358, "learning_rate": 2.081459451617583e-05, "loss": 0.0007, "step": 21933 }, { "epoch": 2.3779271465741543, "grad_norm": 0.32789847254753113, "learning_rate": 2.0810967648338893e-05, "loss": 0.0099, "step": 21934 }, { "epoch": 2.3780355594102343, "grad_norm": 0.27004823088645935, "learning_rate": 2.080734078050196e-05, "loss": 0.0102, "step": 21935 }, { "epoch": 2.378143972246314, "grad_norm": 0.6784853935241699, "learning_rate": 2.0803713912665025e-05, "loss": 0.0261, "step": 21936 }, { "epoch": 2.378252385082394, "grad_norm": 0.022621307522058487, "learning_rate": 2.0800087044828086e-05, "loss": 0.0014, "step": 21937 }, { "epoch": 2.3783607979184733, "grad_norm": 0.24435684084892273, "learning_rate": 2.079646017699115e-05, "loss": 0.0083, "step": 21938 }, { "epoch": 2.3784692107545533, "grad_norm": 0.3681144416332245, "learning_rate": 2.0792833309154217e-05, "loss": 0.0334, "step": 21939 }, { "epoch": 2.3785776235906333, "grad_norm": 0.4063379466533661, "learning_rate": 2.078920644131728e-05, "loss": 0.0107, "step": 21940 }, { "epoch": 2.378686036426713, "grad_norm": 0.13401663303375244, "learning_rate": 2.0785579573480342e-05, "loss": 0.0047, "step": 21941 }, { "epoch": 2.378794449262793, "grad_norm": 0.04311502352356911, "learning_rate": 2.0781952705643406e-05, "loss": 0.0027, "step": 21942 }, { "epoch": 2.3789028620988724, "grad_norm": 0.004408066626638174, "learning_rate": 2.0778325837806474e-05, "loss": 0.0002, "step": 21943 }, { "epoch": 2.3790112749349523, "grad_norm": 0.4566093385219574, "learning_rate": 2.0774698969969535e-05, "loss": 0.0397, "step": 21944 }, { "epoch": 2.3791196877710323, "grad_norm": 0.12058542668819427, "learning_rate": 2.07710721021326e-05, "loss": 0.0069, "step": 21945 }, { "epoch": 2.379228100607112, "grad_norm": 0.05673754960298538, "learning_rate": 2.0767445234295663e-05, "loss": 0.0012, "step": 21946 }, { "epoch": 2.379336513443192, "grad_norm": 0.2568706274032593, "learning_rate": 2.0763818366458727e-05, "loss": 0.0146, "step": 21947 }, { "epoch": 2.3794449262792714, "grad_norm": 0.6387216448783875, "learning_rate": 2.076019149862179e-05, "loss": 0.0516, "step": 21948 }, { "epoch": 2.3795533391153514, "grad_norm": 0.06782805174589157, "learning_rate": 2.0756564630784855e-05, "loss": 0.0067, "step": 21949 }, { "epoch": 2.379661751951431, "grad_norm": 0.13507556915283203, "learning_rate": 2.075293776294792e-05, "loss": 0.008, "step": 21950 }, { "epoch": 2.379770164787511, "grad_norm": 0.1490907222032547, "learning_rate": 2.0749310895110983e-05, "loss": 0.0036, "step": 21951 }, { "epoch": 2.3798785776235905, "grad_norm": 0.08131896704435349, "learning_rate": 2.0745684027274047e-05, "loss": 0.0019, "step": 21952 }, { "epoch": 2.3799869904596704, "grad_norm": 0.4863413870334625, "learning_rate": 2.074205715943711e-05, "loss": 0.0115, "step": 21953 }, { "epoch": 2.3800954032957504, "grad_norm": 0.24225126206874847, "learning_rate": 2.0738430291600176e-05, "loss": 0.0037, "step": 21954 }, { "epoch": 2.38020381613183, "grad_norm": 0.3511522114276886, "learning_rate": 2.073480342376324e-05, "loss": 0.0055, "step": 21955 }, { "epoch": 2.38031222896791, "grad_norm": 0.19447962939739227, "learning_rate": 2.0731176555926304e-05, "loss": 0.0037, "step": 21956 }, { "epoch": 2.3804206418039895, "grad_norm": 0.013702819123864174, "learning_rate": 2.0727549688089368e-05, "loss": 0.0008, "step": 21957 }, { "epoch": 2.3805290546400695, "grad_norm": 0.18060003221035004, "learning_rate": 2.0723922820252432e-05, "loss": 0.018, "step": 21958 }, { "epoch": 2.380637467476149, "grad_norm": 0.2276695817708969, "learning_rate": 2.0720295952415493e-05, "loss": 0.0102, "step": 21959 }, { "epoch": 2.380745880312229, "grad_norm": 0.12458807975053787, "learning_rate": 2.0716669084578557e-05, "loss": 0.0046, "step": 21960 }, { "epoch": 2.3808542931483085, "grad_norm": 0.5984693765640259, "learning_rate": 2.0713042216741624e-05, "loss": 0.0099, "step": 21961 }, { "epoch": 2.3809627059843885, "grad_norm": 0.5351478457450867, "learning_rate": 2.070941534890469e-05, "loss": 0.0277, "step": 21962 }, { "epoch": 2.3810711188204685, "grad_norm": 0.34439951181411743, "learning_rate": 2.070578848106775e-05, "loss": 0.0259, "step": 21963 }, { "epoch": 2.381179531656548, "grad_norm": 0.16836969554424286, "learning_rate": 2.0702161613230813e-05, "loss": 0.0047, "step": 21964 }, { "epoch": 2.381287944492628, "grad_norm": 0.04531442001461983, "learning_rate": 2.069853474539388e-05, "loss": 0.0027, "step": 21965 }, { "epoch": 2.3813963573287076, "grad_norm": 0.31563639640808105, "learning_rate": 2.0694907877556945e-05, "loss": 0.0193, "step": 21966 }, { "epoch": 2.3815047701647876, "grad_norm": 0.0017949964385479689, "learning_rate": 2.0691281009720006e-05, "loss": 0.0001, "step": 21967 }, { "epoch": 2.3816131830008676, "grad_norm": 0.574523389339447, "learning_rate": 2.068765414188307e-05, "loss": 0.0798, "step": 21968 }, { "epoch": 2.381721595836947, "grad_norm": 0.3348400592803955, "learning_rate": 2.0684027274046137e-05, "loss": 0.0222, "step": 21969 }, { "epoch": 2.381830008673027, "grad_norm": 0.10997524112462997, "learning_rate": 2.0680400406209198e-05, "loss": 0.002, "step": 21970 }, { "epoch": 2.3819384215091066, "grad_norm": 0.00901925377547741, "learning_rate": 2.0676773538372262e-05, "loss": 0.0004, "step": 21971 }, { "epoch": 2.3820468343451866, "grad_norm": 0.1756443828344345, "learning_rate": 2.0673146670535326e-05, "loss": 0.0116, "step": 21972 }, { "epoch": 2.382155247181266, "grad_norm": 0.6930745840072632, "learning_rate": 2.066951980269839e-05, "loss": 0.0511, "step": 21973 }, { "epoch": 2.382263660017346, "grad_norm": 0.12090811878442764, "learning_rate": 2.0665892934861454e-05, "loss": 0.0062, "step": 21974 }, { "epoch": 2.3823720728534257, "grad_norm": 0.08506327867507935, "learning_rate": 2.066226606702452e-05, "loss": 0.0036, "step": 21975 }, { "epoch": 2.3824804856895057, "grad_norm": 0.19933508336544037, "learning_rate": 2.0658639199187583e-05, "loss": 0.0075, "step": 21976 }, { "epoch": 2.3825888985255856, "grad_norm": 0.4690454602241516, "learning_rate": 2.0655012331350647e-05, "loss": 0.0375, "step": 21977 }, { "epoch": 2.382697311361665, "grad_norm": 0.005439559929072857, "learning_rate": 2.065138546351371e-05, "loss": 0.0001, "step": 21978 }, { "epoch": 2.382805724197745, "grad_norm": 0.00671608280390501, "learning_rate": 2.0647758595676775e-05, "loss": 0.0003, "step": 21979 }, { "epoch": 2.3829141370338247, "grad_norm": 0.1847034990787506, "learning_rate": 2.064413172783984e-05, "loss": 0.0098, "step": 21980 }, { "epoch": 2.3830225498699047, "grad_norm": 0.27794456481933594, "learning_rate": 2.0640504860002903e-05, "loss": 0.008, "step": 21981 }, { "epoch": 2.3831309627059842, "grad_norm": 0.1661183387041092, "learning_rate": 2.0636877992165967e-05, "loss": 0.0028, "step": 21982 }, { "epoch": 2.383239375542064, "grad_norm": 0.1238623782992363, "learning_rate": 2.063325112432903e-05, "loss": 0.0046, "step": 21983 }, { "epoch": 2.3833477883781438, "grad_norm": 0.02628328837454319, "learning_rate": 2.0629624256492095e-05, "loss": 0.0012, "step": 21984 }, { "epoch": 2.3834562012142237, "grad_norm": 0.33925703167915344, "learning_rate": 2.0625997388655156e-05, "loss": 0.009, "step": 21985 }, { "epoch": 2.3835646140503037, "grad_norm": 1.5341140031814575, "learning_rate": 2.062237052081822e-05, "loss": 0.0112, "step": 21986 }, { "epoch": 2.3836730268863833, "grad_norm": 0.9426352977752686, "learning_rate": 2.0618743652981288e-05, "loss": 0.0495, "step": 21987 }, { "epoch": 2.3837814397224633, "grad_norm": 0.22581835091114044, "learning_rate": 2.0615116785144352e-05, "loss": 0.0115, "step": 21988 }, { "epoch": 2.383889852558543, "grad_norm": 0.05617407336831093, "learning_rate": 2.0611489917307412e-05, "loss": 0.0019, "step": 21989 }, { "epoch": 2.383998265394623, "grad_norm": 0.3605363070964813, "learning_rate": 2.0607863049470477e-05, "loss": 0.0213, "step": 21990 }, { "epoch": 2.3841066782307023, "grad_norm": 0.17828653752803802, "learning_rate": 2.0604236181633544e-05, "loss": 0.0102, "step": 21991 }, { "epoch": 2.3842150910667823, "grad_norm": 0.017998874187469482, "learning_rate": 2.0600609313796608e-05, "loss": 0.0009, "step": 21992 }, { "epoch": 2.384323503902862, "grad_norm": 0.326246052980423, "learning_rate": 2.059698244595967e-05, "loss": 0.0191, "step": 21993 }, { "epoch": 2.384431916738942, "grad_norm": 0.0028136298060417175, "learning_rate": 2.0593355578122733e-05, "loss": 0.0001, "step": 21994 }, { "epoch": 2.384540329575022, "grad_norm": 0.18876489996910095, "learning_rate": 2.05897287102858e-05, "loss": 0.0086, "step": 21995 }, { "epoch": 2.3846487424111014, "grad_norm": 0.10340934246778488, "learning_rate": 2.058610184244886e-05, "loss": 0.0019, "step": 21996 }, { "epoch": 2.3847571552471813, "grad_norm": 0.16827556490898132, "learning_rate": 2.0582474974611925e-05, "loss": 0.0199, "step": 21997 }, { "epoch": 2.384865568083261, "grad_norm": 0.19920530915260315, "learning_rate": 2.057884810677499e-05, "loss": 0.0053, "step": 21998 }, { "epoch": 2.384973980919341, "grad_norm": 0.447405070066452, "learning_rate": 2.0575221238938054e-05, "loss": 0.0437, "step": 21999 }, { "epoch": 2.385082393755421, "grad_norm": 0.32431480288505554, "learning_rate": 2.0571594371101118e-05, "loss": 0.0322, "step": 22000 }, { "epoch": 2.3851908065915004, "grad_norm": 0.018350979313254356, "learning_rate": 2.0567967503264182e-05, "loss": 0.0007, "step": 22001 }, { "epoch": 2.3852992194275804, "grad_norm": 0.0477888286113739, "learning_rate": 2.0564340635427246e-05, "loss": 0.0026, "step": 22002 }, { "epoch": 2.38540763226366, "grad_norm": 0.2599429190158844, "learning_rate": 2.056071376759031e-05, "loss": 0.0084, "step": 22003 }, { "epoch": 2.38551604509974, "grad_norm": 0.14835041761398315, "learning_rate": 2.0557086899753374e-05, "loss": 0.0043, "step": 22004 }, { "epoch": 2.3856244579358195, "grad_norm": 0.4803420305252075, "learning_rate": 2.0553460031916438e-05, "loss": 0.0366, "step": 22005 }, { "epoch": 2.3857328707718994, "grad_norm": 0.750335693359375, "learning_rate": 2.0549833164079502e-05, "loss": 0.0589, "step": 22006 }, { "epoch": 2.385841283607979, "grad_norm": 0.17698949575424194, "learning_rate": 2.0546206296242566e-05, "loss": 0.0029, "step": 22007 }, { "epoch": 2.385949696444059, "grad_norm": 0.091469906270504, "learning_rate": 2.054257942840563e-05, "loss": 0.0021, "step": 22008 }, { "epoch": 2.386058109280139, "grad_norm": 0.079957976937294, "learning_rate": 2.0538952560568695e-05, "loss": 0.0034, "step": 22009 }, { "epoch": 2.3861665221162185, "grad_norm": 0.029040588065981865, "learning_rate": 2.053532569273176e-05, "loss": 0.0012, "step": 22010 }, { "epoch": 2.3862749349522985, "grad_norm": 0.09217824041843414, "learning_rate": 2.053169882489482e-05, "loss": 0.0051, "step": 22011 }, { "epoch": 2.386383347788378, "grad_norm": 0.44557443261146545, "learning_rate": 2.0528071957057884e-05, "loss": 0.0297, "step": 22012 }, { "epoch": 2.386491760624458, "grad_norm": 0.1460614651441574, "learning_rate": 2.052444508922095e-05, "loss": 0.0054, "step": 22013 }, { "epoch": 2.3866001734605375, "grad_norm": 0.016710665076971054, "learning_rate": 2.0520818221384015e-05, "loss": 0.0012, "step": 22014 }, { "epoch": 2.3867085862966175, "grad_norm": 0.09976696968078613, "learning_rate": 2.0517191353547076e-05, "loss": 0.0087, "step": 22015 }, { "epoch": 2.386816999132697, "grad_norm": 0.5088410377502441, "learning_rate": 2.051356448571014e-05, "loss": 0.008, "step": 22016 }, { "epoch": 2.386925411968777, "grad_norm": 0.07900069653987885, "learning_rate": 2.0509937617873207e-05, "loss": 0.001, "step": 22017 }, { "epoch": 2.387033824804857, "grad_norm": 0.07875801622867584, "learning_rate": 2.050631075003627e-05, "loss": 0.0067, "step": 22018 }, { "epoch": 2.3871422376409366, "grad_norm": 0.02524634636938572, "learning_rate": 2.0502683882199332e-05, "loss": 0.0009, "step": 22019 }, { "epoch": 2.3872506504770166, "grad_norm": 0.22625510394573212, "learning_rate": 2.0499057014362396e-05, "loss": 0.0046, "step": 22020 }, { "epoch": 2.387359063313096, "grad_norm": 0.13687290251255035, "learning_rate": 2.0495430146525464e-05, "loss": 0.0063, "step": 22021 }, { "epoch": 2.387467476149176, "grad_norm": 0.036860786378383636, "learning_rate": 2.0491803278688525e-05, "loss": 0.0014, "step": 22022 }, { "epoch": 2.387575888985256, "grad_norm": 0.4815143048763275, "learning_rate": 2.048817641085159e-05, "loss": 0.0358, "step": 22023 }, { "epoch": 2.3876843018213356, "grad_norm": 0.017285538837313652, "learning_rate": 2.0484549543014653e-05, "loss": 0.0006, "step": 22024 }, { "epoch": 2.3877927146574156, "grad_norm": 0.37205764651298523, "learning_rate": 2.0480922675177717e-05, "loss": 0.0072, "step": 22025 }, { "epoch": 2.387901127493495, "grad_norm": 0.023009387776255608, "learning_rate": 2.047729580734078e-05, "loss": 0.0013, "step": 22026 }, { "epoch": 2.388009540329575, "grad_norm": 0.4463103115558624, "learning_rate": 2.0473668939503845e-05, "loss": 0.0151, "step": 22027 }, { "epoch": 2.3881179531656547, "grad_norm": 0.08690556138753891, "learning_rate": 2.047004207166691e-05, "loss": 0.0021, "step": 22028 }, { "epoch": 2.3882263660017347, "grad_norm": 0.4356878697872162, "learning_rate": 2.0466415203829973e-05, "loss": 0.0185, "step": 22029 }, { "epoch": 2.388334778837814, "grad_norm": 0.3475307822227478, "learning_rate": 2.0462788335993037e-05, "loss": 0.0179, "step": 22030 }, { "epoch": 2.388443191673894, "grad_norm": 0.13898195326328278, "learning_rate": 2.04591614681561e-05, "loss": 0.0068, "step": 22031 }, { "epoch": 2.388551604509974, "grad_norm": 0.8151077032089233, "learning_rate": 2.0455534600319166e-05, "loss": 0.0294, "step": 22032 }, { "epoch": 2.3886600173460537, "grad_norm": 0.05281190201640129, "learning_rate": 2.045190773248223e-05, "loss": 0.0023, "step": 22033 }, { "epoch": 2.3887684301821337, "grad_norm": 0.3390999734401703, "learning_rate": 2.0448280864645294e-05, "loss": 0.0316, "step": 22034 }, { "epoch": 2.3888768430182132, "grad_norm": 0.26752394437789917, "learning_rate": 2.0444653996808358e-05, "loss": 0.0122, "step": 22035 }, { "epoch": 2.388985255854293, "grad_norm": 0.306253045797348, "learning_rate": 2.0441027128971422e-05, "loss": 0.0246, "step": 22036 }, { "epoch": 2.3890936686903728, "grad_norm": 0.15802574157714844, "learning_rate": 2.0437400261134486e-05, "loss": 0.0088, "step": 22037 }, { "epoch": 2.3892020815264527, "grad_norm": 0.3901515305042267, "learning_rate": 2.043377339329755e-05, "loss": 0.0199, "step": 22038 }, { "epoch": 2.3893104943625323, "grad_norm": 0.3103525936603546, "learning_rate": 2.0430146525460614e-05, "loss": 0.0142, "step": 22039 }, { "epoch": 2.3894189071986123, "grad_norm": 0.013165104202926159, "learning_rate": 2.042651965762368e-05, "loss": 0.0006, "step": 22040 }, { "epoch": 2.3895273200346923, "grad_norm": 0.04698243737220764, "learning_rate": 2.042289278978674e-05, "loss": 0.0015, "step": 22041 }, { "epoch": 2.389635732870772, "grad_norm": 0.07310065627098083, "learning_rate": 2.0419265921949803e-05, "loss": 0.0049, "step": 22042 }, { "epoch": 2.389744145706852, "grad_norm": 0.049909621477127075, "learning_rate": 2.041563905411287e-05, "loss": 0.0014, "step": 22043 }, { "epoch": 2.3898525585429313, "grad_norm": 0.2270817756652832, "learning_rate": 2.0412012186275935e-05, "loss": 0.0378, "step": 22044 }, { "epoch": 2.3899609713790113, "grad_norm": 0.6340302228927612, "learning_rate": 2.0408385318438996e-05, "loss": 0.0383, "step": 22045 }, { "epoch": 2.3900693842150913, "grad_norm": 0.10026947408914566, "learning_rate": 2.040475845060206e-05, "loss": 0.0101, "step": 22046 }, { "epoch": 2.390177797051171, "grad_norm": 0.10544144362211227, "learning_rate": 2.0401131582765127e-05, "loss": 0.0052, "step": 22047 }, { "epoch": 2.390286209887251, "grad_norm": 0.13682322204113007, "learning_rate": 2.0397504714928188e-05, "loss": 0.0091, "step": 22048 }, { "epoch": 2.3903946227233304, "grad_norm": 0.7762598395347595, "learning_rate": 2.0393877847091252e-05, "loss": 0.0101, "step": 22049 }, { "epoch": 2.3905030355594103, "grad_norm": 0.225009948015213, "learning_rate": 2.0390250979254316e-05, "loss": 0.0105, "step": 22050 }, { "epoch": 2.39061144839549, "grad_norm": 0.15664233267307281, "learning_rate": 2.0386624111417384e-05, "loss": 0.01, "step": 22051 }, { "epoch": 2.39071986123157, "grad_norm": 0.08712290227413177, "learning_rate": 2.0382997243580444e-05, "loss": 0.0068, "step": 22052 }, { "epoch": 2.3908282740676494, "grad_norm": 0.6376786231994629, "learning_rate": 2.037937037574351e-05, "loss": 0.0118, "step": 22053 }, { "epoch": 2.3909366869037294, "grad_norm": 0.046508900821208954, "learning_rate": 2.0375743507906573e-05, "loss": 0.0019, "step": 22054 }, { "epoch": 2.3910450997398094, "grad_norm": 0.382405161857605, "learning_rate": 2.0372116640069637e-05, "loss": 0.025, "step": 22055 }, { "epoch": 2.391153512575889, "grad_norm": 0.11627539992332458, "learning_rate": 2.03684897722327e-05, "loss": 0.0055, "step": 22056 }, { "epoch": 2.391261925411969, "grad_norm": 0.49322912096977234, "learning_rate": 2.0364862904395765e-05, "loss": 0.0139, "step": 22057 }, { "epoch": 2.3913703382480485, "grad_norm": 0.38787421584129333, "learning_rate": 2.036123603655883e-05, "loss": 0.0219, "step": 22058 }, { "epoch": 2.3914787510841284, "grad_norm": 0.14599795639514923, "learning_rate": 2.0357609168721893e-05, "loss": 0.0058, "step": 22059 }, { "epoch": 2.391587163920208, "grad_norm": 0.5580705404281616, "learning_rate": 2.0353982300884957e-05, "loss": 0.0191, "step": 22060 }, { "epoch": 2.391695576756288, "grad_norm": 0.16304153203964233, "learning_rate": 2.035035543304802e-05, "loss": 0.0049, "step": 22061 }, { "epoch": 2.3918039895923675, "grad_norm": 0.21397048234939575, "learning_rate": 2.0346728565211085e-05, "loss": 0.0029, "step": 22062 }, { "epoch": 2.3919124024284475, "grad_norm": 0.3983433246612549, "learning_rate": 2.034310169737415e-05, "loss": 0.0316, "step": 22063 }, { "epoch": 2.3920208152645275, "grad_norm": 0.4418089985847473, "learning_rate": 2.0339474829537214e-05, "loss": 0.0197, "step": 22064 }, { "epoch": 2.392129228100607, "grad_norm": 0.097810298204422, "learning_rate": 2.0335847961700278e-05, "loss": 0.0024, "step": 22065 }, { "epoch": 2.392237640936687, "grad_norm": 0.40386635065078735, "learning_rate": 2.0332221093863342e-05, "loss": 0.0224, "step": 22066 }, { "epoch": 2.3923460537727665, "grad_norm": 0.12473202496767044, "learning_rate": 2.0328594226026403e-05, "loss": 0.0027, "step": 22067 }, { "epoch": 2.3924544666088465, "grad_norm": 0.28133484721183777, "learning_rate": 2.0324967358189467e-05, "loss": 0.0065, "step": 22068 }, { "epoch": 2.3925628794449265, "grad_norm": 0.15515874326229095, "learning_rate": 2.0321340490352534e-05, "loss": 0.011, "step": 22069 }, { "epoch": 2.392671292281006, "grad_norm": 0.05675395950675011, "learning_rate": 2.0317713622515598e-05, "loss": 0.0022, "step": 22070 }, { "epoch": 2.392779705117086, "grad_norm": 0.1003284677863121, "learning_rate": 2.031408675467866e-05, "loss": 0.002, "step": 22071 }, { "epoch": 2.3928881179531656, "grad_norm": 0.3846568763256073, "learning_rate": 2.0310459886841723e-05, "loss": 0.0311, "step": 22072 }, { "epoch": 2.3929965307892456, "grad_norm": 2.134253978729248, "learning_rate": 2.030683301900479e-05, "loss": 0.0454, "step": 22073 }, { "epoch": 2.393104943625325, "grad_norm": 0.37739235162734985, "learning_rate": 2.030320615116785e-05, "loss": 0.0176, "step": 22074 }, { "epoch": 2.393213356461405, "grad_norm": 0.11696259677410126, "learning_rate": 2.0299579283330915e-05, "loss": 0.0055, "step": 22075 }, { "epoch": 2.3933217692974846, "grad_norm": 0.14084649085998535, "learning_rate": 2.029595241549398e-05, "loss": 0.0063, "step": 22076 }, { "epoch": 2.3934301821335646, "grad_norm": 0.29614102840423584, "learning_rate": 2.0292325547657047e-05, "loss": 0.0227, "step": 22077 }, { "epoch": 2.3935385949696446, "grad_norm": 0.13854847848415375, "learning_rate": 2.0288698679820108e-05, "loss": 0.0333, "step": 22078 }, { "epoch": 2.393647007805724, "grad_norm": 0.05668973550200462, "learning_rate": 2.0285071811983172e-05, "loss": 0.0012, "step": 22079 }, { "epoch": 2.393755420641804, "grad_norm": 0.39156922698020935, "learning_rate": 2.0281444944146236e-05, "loss": 0.0234, "step": 22080 }, { "epoch": 2.3938638334778837, "grad_norm": 0.19498902559280396, "learning_rate": 2.02778180763093e-05, "loss": 0.0187, "step": 22081 }, { "epoch": 2.3939722463139637, "grad_norm": 0.453147828578949, "learning_rate": 2.0274191208472364e-05, "loss": 0.0219, "step": 22082 }, { "epoch": 2.394080659150043, "grad_norm": 0.026663366705179214, "learning_rate": 2.0270564340635428e-05, "loss": 0.0015, "step": 22083 }, { "epoch": 2.394189071986123, "grad_norm": 0.11833525449037552, "learning_rate": 2.0266937472798492e-05, "loss": 0.0028, "step": 22084 }, { "epoch": 2.3942974848222027, "grad_norm": 0.15719279646873474, "learning_rate": 2.0263310604961556e-05, "loss": 0.0041, "step": 22085 }, { "epoch": 2.3944058976582827, "grad_norm": 0.01950611174106598, "learning_rate": 2.025968373712462e-05, "loss": 0.0008, "step": 22086 }, { "epoch": 2.3945143104943627, "grad_norm": 0.10383200645446777, "learning_rate": 2.0256056869287685e-05, "loss": 0.0021, "step": 22087 }, { "epoch": 2.3946227233304422, "grad_norm": 0.0047638448886573315, "learning_rate": 2.025243000145075e-05, "loss": 0.0003, "step": 22088 }, { "epoch": 2.3947311361665222, "grad_norm": 0.009024052880704403, "learning_rate": 2.0248803133613813e-05, "loss": 0.0003, "step": 22089 }, { "epoch": 2.3948395490026018, "grad_norm": 0.04065267741680145, "learning_rate": 2.0245176265776877e-05, "loss": 0.0021, "step": 22090 }, { "epoch": 2.3949479618386817, "grad_norm": 0.42607563734054565, "learning_rate": 2.024154939793994e-05, "loss": 0.0287, "step": 22091 }, { "epoch": 2.3950563746747617, "grad_norm": 0.3402930200099945, "learning_rate": 2.0237922530103005e-05, "loss": 0.0072, "step": 22092 }, { "epoch": 2.3951647875108413, "grad_norm": 0.16955991089344025, "learning_rate": 2.0234295662266066e-05, "loss": 0.009, "step": 22093 }, { "epoch": 2.3952732003469213, "grad_norm": 0.04210831597447395, "learning_rate": 2.023066879442913e-05, "loss": 0.0025, "step": 22094 }, { "epoch": 2.395381613183001, "grad_norm": 0.0438174307346344, "learning_rate": 2.0227041926592197e-05, "loss": 0.0017, "step": 22095 }, { "epoch": 2.395490026019081, "grad_norm": 0.0717536211013794, "learning_rate": 2.022341505875526e-05, "loss": 0.0048, "step": 22096 }, { "epoch": 2.3955984388551603, "grad_norm": 0.013741196133196354, "learning_rate": 2.0219788190918322e-05, "loss": 0.0007, "step": 22097 }, { "epoch": 2.3957068516912403, "grad_norm": 0.06662794947624207, "learning_rate": 2.0216161323081386e-05, "loss": 0.004, "step": 22098 }, { "epoch": 2.39581526452732, "grad_norm": 0.3831084072589874, "learning_rate": 2.0212534455244454e-05, "loss": 0.0094, "step": 22099 }, { "epoch": 2.3959236773634, "grad_norm": 0.061199724674224854, "learning_rate": 2.0208907587407515e-05, "loss": 0.0017, "step": 22100 }, { "epoch": 2.39603209019948, "grad_norm": 0.0211555827409029, "learning_rate": 2.020528071957058e-05, "loss": 0.0008, "step": 22101 }, { "epoch": 2.3961405030355594, "grad_norm": 0.21709170937538147, "learning_rate": 2.0201653851733643e-05, "loss": 0.007, "step": 22102 }, { "epoch": 2.3962489158716394, "grad_norm": 0.25243785977363586, "learning_rate": 2.019802698389671e-05, "loss": 0.0164, "step": 22103 }, { "epoch": 2.396357328707719, "grad_norm": 0.0791945680975914, "learning_rate": 2.019440011605977e-05, "loss": 0.0018, "step": 22104 }, { "epoch": 2.396465741543799, "grad_norm": 0.15498873591423035, "learning_rate": 2.0190773248222835e-05, "loss": 0.0121, "step": 22105 }, { "epoch": 2.3965741543798784, "grad_norm": 0.47723492980003357, "learning_rate": 2.01871463803859e-05, "loss": 0.0159, "step": 22106 }, { "epoch": 2.3966825672159584, "grad_norm": 0.25245776772499084, "learning_rate": 2.0183519512548963e-05, "loss": 0.028, "step": 22107 }, { "epoch": 2.396790980052038, "grad_norm": 0.6448187828063965, "learning_rate": 2.0179892644712027e-05, "loss": 0.0323, "step": 22108 }, { "epoch": 2.396899392888118, "grad_norm": 0.018105516210198402, "learning_rate": 2.017626577687509e-05, "loss": 0.0008, "step": 22109 }, { "epoch": 2.397007805724198, "grad_norm": 0.38474780321121216, "learning_rate": 2.0172638909038156e-05, "loss": 0.0171, "step": 22110 }, { "epoch": 2.3971162185602775, "grad_norm": 0.036197394132614136, "learning_rate": 2.016901204120122e-05, "loss": 0.0031, "step": 22111 }, { "epoch": 2.3972246313963574, "grad_norm": 0.0030938356649130583, "learning_rate": 2.0165385173364284e-05, "loss": 0.0001, "step": 22112 }, { "epoch": 2.397333044232437, "grad_norm": 0.12297525256872177, "learning_rate": 2.0161758305527348e-05, "loss": 0.0017, "step": 22113 }, { "epoch": 2.397441457068517, "grad_norm": 0.2863951027393341, "learning_rate": 2.0158131437690412e-05, "loss": 0.0204, "step": 22114 }, { "epoch": 2.397549869904597, "grad_norm": 0.1709936112165451, "learning_rate": 2.0154504569853476e-05, "loss": 0.0068, "step": 22115 }, { "epoch": 2.3976582827406765, "grad_norm": 0.006402973085641861, "learning_rate": 2.015087770201654e-05, "loss": 0.0003, "step": 22116 }, { "epoch": 2.3977666955767565, "grad_norm": 0.016366977244615555, "learning_rate": 2.0147250834179604e-05, "loss": 0.0008, "step": 22117 }, { "epoch": 2.397875108412836, "grad_norm": 0.2163642793893814, "learning_rate": 2.014362396634267e-05, "loss": 0.0024, "step": 22118 }, { "epoch": 2.397983521248916, "grad_norm": 0.7091227173805237, "learning_rate": 2.013999709850573e-05, "loss": 0.0377, "step": 22119 }, { "epoch": 2.3980919340849955, "grad_norm": 0.5205673575401306, "learning_rate": 2.0136370230668793e-05, "loss": 0.028, "step": 22120 }, { "epoch": 2.3982003469210755, "grad_norm": 0.06239496171474457, "learning_rate": 2.013274336283186e-05, "loss": 0.0055, "step": 22121 }, { "epoch": 2.398308759757155, "grad_norm": 0.23279525339603424, "learning_rate": 2.0129116494994925e-05, "loss": 0.0083, "step": 22122 }, { "epoch": 2.398417172593235, "grad_norm": 0.43013817071914673, "learning_rate": 2.0125489627157986e-05, "loss": 0.007, "step": 22123 }, { "epoch": 2.398525585429315, "grad_norm": 0.15949510037899017, "learning_rate": 2.012186275932105e-05, "loss": 0.0055, "step": 22124 }, { "epoch": 2.3986339982653946, "grad_norm": 0.12400440871715546, "learning_rate": 2.0118235891484117e-05, "loss": 0.0031, "step": 22125 }, { "epoch": 2.3987424111014746, "grad_norm": 0.07164543122053146, "learning_rate": 2.0114609023647178e-05, "loss": 0.0041, "step": 22126 }, { "epoch": 2.398850823937554, "grad_norm": 0.1678069531917572, "learning_rate": 2.0110982155810242e-05, "loss": 0.0057, "step": 22127 }, { "epoch": 2.398959236773634, "grad_norm": 1.327380895614624, "learning_rate": 2.0107355287973306e-05, "loss": 0.0201, "step": 22128 }, { "epoch": 2.3990676496097136, "grad_norm": 0.31050851941108704, "learning_rate": 2.0103728420136374e-05, "loss": 0.0053, "step": 22129 }, { "epoch": 2.3991760624457936, "grad_norm": 0.4430831968784332, "learning_rate": 2.0100101552299434e-05, "loss": 0.0396, "step": 22130 }, { "epoch": 2.399284475281873, "grad_norm": 0.05003448203206062, "learning_rate": 2.00964746844625e-05, "loss": 0.0006, "step": 22131 }, { "epoch": 2.399392888117953, "grad_norm": 0.06101791560649872, "learning_rate": 2.0092847816625563e-05, "loss": 0.0022, "step": 22132 }, { "epoch": 2.399501300954033, "grad_norm": 0.2207874059677124, "learning_rate": 2.008922094878863e-05, "loss": 0.0069, "step": 22133 }, { "epoch": 2.3996097137901127, "grad_norm": 0.14849159121513367, "learning_rate": 2.008559408095169e-05, "loss": 0.0101, "step": 22134 }, { "epoch": 2.3997181266261927, "grad_norm": 0.2888718247413635, "learning_rate": 2.0081967213114755e-05, "loss": 0.0072, "step": 22135 }, { "epoch": 2.399826539462272, "grad_norm": 0.26971879601478577, "learning_rate": 2.007834034527782e-05, "loss": 0.0119, "step": 22136 }, { "epoch": 2.399934952298352, "grad_norm": 0.5437084436416626, "learning_rate": 2.0074713477440883e-05, "loss": 0.0209, "step": 22137 }, { "epoch": 2.4000433651344317, "grad_norm": 0.03335900977253914, "learning_rate": 2.0071086609603947e-05, "loss": 0.0012, "step": 22138 }, { "epoch": 2.4001517779705117, "grad_norm": 0.23327618837356567, "learning_rate": 2.006745974176701e-05, "loss": 0.0046, "step": 22139 }, { "epoch": 2.4002601908065913, "grad_norm": 0.059677451848983765, "learning_rate": 2.0063832873930075e-05, "loss": 0.002, "step": 22140 }, { "epoch": 2.4003686036426712, "grad_norm": 0.27882128953933716, "learning_rate": 2.006020600609314e-05, "loss": 0.0195, "step": 22141 }, { "epoch": 2.4004770164787512, "grad_norm": 0.02098613604903221, "learning_rate": 2.0056579138256204e-05, "loss": 0.0011, "step": 22142 }, { "epoch": 2.4005854293148308, "grad_norm": 0.1269848644733429, "learning_rate": 2.0052952270419268e-05, "loss": 0.0048, "step": 22143 }, { "epoch": 2.4006938421509108, "grad_norm": 0.042971204966306686, "learning_rate": 2.0049325402582332e-05, "loss": 0.0015, "step": 22144 }, { "epoch": 2.4008022549869903, "grad_norm": 0.008767359890043736, "learning_rate": 2.0045698534745393e-05, "loss": 0.0003, "step": 22145 }, { "epoch": 2.4009106678230703, "grad_norm": 0.0972987562417984, "learning_rate": 2.004207166690846e-05, "loss": 0.0045, "step": 22146 }, { "epoch": 2.4010190806591503, "grad_norm": 0.3428449332714081, "learning_rate": 2.0038444799071524e-05, "loss": 0.0185, "step": 22147 }, { "epoch": 2.40112749349523, "grad_norm": 0.01922505348920822, "learning_rate": 2.0034817931234588e-05, "loss": 0.0006, "step": 22148 }, { "epoch": 2.40123590633131, "grad_norm": 0.07120776176452637, "learning_rate": 2.003119106339765e-05, "loss": 0.0025, "step": 22149 }, { "epoch": 2.4013443191673893, "grad_norm": 0.1005803793668747, "learning_rate": 2.0027564195560713e-05, "loss": 0.0071, "step": 22150 }, { "epoch": 2.4014527320034693, "grad_norm": 0.0461789071559906, "learning_rate": 2.002393732772378e-05, "loss": 0.001, "step": 22151 }, { "epoch": 2.401561144839549, "grad_norm": 0.26230573654174805, "learning_rate": 2.002031045988684e-05, "loss": 0.0057, "step": 22152 }, { "epoch": 2.401669557675629, "grad_norm": 0.09916950017213821, "learning_rate": 2.0016683592049905e-05, "loss": 0.005, "step": 22153 }, { "epoch": 2.4017779705117084, "grad_norm": 0.15279802680015564, "learning_rate": 2.001305672421297e-05, "loss": 0.0065, "step": 22154 }, { "epoch": 2.4018863833477884, "grad_norm": 0.08335483819246292, "learning_rate": 2.0009429856376037e-05, "loss": 0.004, "step": 22155 }, { "epoch": 2.4019947961838684, "grad_norm": 0.10098076611757278, "learning_rate": 2.0005802988539098e-05, "loss": 0.0041, "step": 22156 }, { "epoch": 2.402103209019948, "grad_norm": 0.3756150007247925, "learning_rate": 2.0002176120702162e-05, "loss": 0.0203, "step": 22157 }, { "epoch": 2.402211621856028, "grad_norm": 0.12604467570781708, "learning_rate": 1.9998549252865226e-05, "loss": 0.0057, "step": 22158 }, { "epoch": 2.4023200346921074, "grad_norm": 0.315388023853302, "learning_rate": 1.9994922385028293e-05, "loss": 0.0054, "step": 22159 }, { "epoch": 2.4024284475281874, "grad_norm": 0.25478988885879517, "learning_rate": 1.9991295517191354e-05, "loss": 0.0072, "step": 22160 }, { "epoch": 2.402536860364267, "grad_norm": 0.02211907133460045, "learning_rate": 1.9987668649354418e-05, "loss": 0.0009, "step": 22161 }, { "epoch": 2.402645273200347, "grad_norm": 0.2699792981147766, "learning_rate": 1.9984041781517482e-05, "loss": 0.0133, "step": 22162 }, { "epoch": 2.4027536860364265, "grad_norm": 0.22988785803318024, "learning_rate": 1.9980414913680546e-05, "loss": 0.0043, "step": 22163 }, { "epoch": 2.4028620988725065, "grad_norm": 0.075852170586586, "learning_rate": 1.997678804584361e-05, "loss": 0.0015, "step": 22164 }, { "epoch": 2.4029705117085864, "grad_norm": 0.24418748915195465, "learning_rate": 1.9973161178006675e-05, "loss": 0.0166, "step": 22165 }, { "epoch": 2.403078924544666, "grad_norm": 0.032678425312042236, "learning_rate": 1.996953431016974e-05, "loss": 0.001, "step": 22166 }, { "epoch": 2.403187337380746, "grad_norm": 0.12942945957183838, "learning_rate": 1.9965907442332803e-05, "loss": 0.0061, "step": 22167 }, { "epoch": 2.4032957502168255, "grad_norm": 0.13559721410274506, "learning_rate": 1.9962280574495867e-05, "loss": 0.0064, "step": 22168 }, { "epoch": 2.4034041630529055, "grad_norm": 0.16448982059955597, "learning_rate": 1.995865370665893e-05, "loss": 0.0113, "step": 22169 }, { "epoch": 2.4035125758889855, "grad_norm": 0.5196808576583862, "learning_rate": 1.9955026838821995e-05, "loss": 0.021, "step": 22170 }, { "epoch": 2.403620988725065, "grad_norm": 0.062347449362277985, "learning_rate": 1.9951399970985056e-05, "loss": 0.0018, "step": 22171 }, { "epoch": 2.403729401561145, "grad_norm": 0.266899049282074, "learning_rate": 1.9947773103148123e-05, "loss": 0.0127, "step": 22172 }, { "epoch": 2.4038378143972245, "grad_norm": 0.04728478193283081, "learning_rate": 1.9944146235311187e-05, "loss": 0.0034, "step": 22173 }, { "epoch": 2.4039462272333045, "grad_norm": 0.29368236660957336, "learning_rate": 1.994051936747425e-05, "loss": 0.0147, "step": 22174 }, { "epoch": 2.404054640069384, "grad_norm": 0.5688653588294983, "learning_rate": 1.9936892499637312e-05, "loss": 0.0126, "step": 22175 }, { "epoch": 2.404163052905464, "grad_norm": 0.051969319581985474, "learning_rate": 1.9933265631800376e-05, "loss": 0.0016, "step": 22176 }, { "epoch": 2.4042714657415436, "grad_norm": 0.13483084738254547, "learning_rate": 1.9929638763963444e-05, "loss": 0.0046, "step": 22177 }, { "epoch": 2.4043798785776236, "grad_norm": 0.3094255030155182, "learning_rate": 1.9926011896126505e-05, "loss": 0.0114, "step": 22178 }, { "epoch": 2.4044882914137036, "grad_norm": 0.20699994266033173, "learning_rate": 1.992238502828957e-05, "loss": 0.009, "step": 22179 }, { "epoch": 2.404596704249783, "grad_norm": 0.7894415259361267, "learning_rate": 1.9918758160452633e-05, "loss": 0.0278, "step": 22180 }, { "epoch": 2.404705117085863, "grad_norm": 0.25375568866729736, "learning_rate": 1.99151312926157e-05, "loss": 0.0184, "step": 22181 }, { "epoch": 2.4048135299219426, "grad_norm": 0.23945695161819458, "learning_rate": 1.991150442477876e-05, "loss": 0.0295, "step": 22182 }, { "epoch": 2.4049219427580226, "grad_norm": 0.9971113204956055, "learning_rate": 1.9907877556941825e-05, "loss": 0.032, "step": 22183 }, { "epoch": 2.405030355594102, "grad_norm": 0.12741966545581818, "learning_rate": 1.990425068910489e-05, "loss": 0.0058, "step": 22184 }, { "epoch": 2.405138768430182, "grad_norm": 0.8384946584701538, "learning_rate": 1.9900623821267957e-05, "loss": 0.0185, "step": 22185 }, { "epoch": 2.4052471812662617, "grad_norm": 0.7987326383590698, "learning_rate": 1.9896996953431017e-05, "loss": 0.0277, "step": 22186 }, { "epoch": 2.4053555941023417, "grad_norm": 0.5154522657394409, "learning_rate": 1.989337008559408e-05, "loss": 0.0092, "step": 22187 }, { "epoch": 2.4054640069384217, "grad_norm": 1.5167620182037354, "learning_rate": 1.9889743217757146e-05, "loss": 0.0159, "step": 22188 }, { "epoch": 2.405572419774501, "grad_norm": 0.16023817658424377, "learning_rate": 1.988611634992021e-05, "loss": 0.0049, "step": 22189 }, { "epoch": 2.405680832610581, "grad_norm": 0.26258814334869385, "learning_rate": 1.9882489482083274e-05, "loss": 0.0047, "step": 22190 }, { "epoch": 2.4057892454466607, "grad_norm": 0.210764542222023, "learning_rate": 1.9878862614246338e-05, "loss": 0.005, "step": 22191 }, { "epoch": 2.4058976582827407, "grad_norm": 0.35925036668777466, "learning_rate": 1.9875235746409402e-05, "loss": 0.0113, "step": 22192 }, { "epoch": 2.4060060711188207, "grad_norm": 0.12890562415122986, "learning_rate": 1.9871608878572466e-05, "loss": 0.0026, "step": 22193 }, { "epoch": 2.4061144839549002, "grad_norm": 0.6616328954696655, "learning_rate": 1.986798201073553e-05, "loss": 0.0186, "step": 22194 }, { "epoch": 2.4062228967909802, "grad_norm": 0.318050354719162, "learning_rate": 1.9864355142898594e-05, "loss": 0.0104, "step": 22195 }, { "epoch": 2.4063313096270598, "grad_norm": 0.025497620925307274, "learning_rate": 1.986072827506166e-05, "loss": 0.0012, "step": 22196 }, { "epoch": 2.4064397224631398, "grad_norm": 0.22410379350185394, "learning_rate": 1.985710140722472e-05, "loss": 0.0116, "step": 22197 }, { "epoch": 2.4065481352992193, "grad_norm": 0.09357256442308426, "learning_rate": 1.9853474539387787e-05, "loss": 0.001, "step": 22198 }, { "epoch": 2.4066565481352993, "grad_norm": 0.007366664241999388, "learning_rate": 1.984984767155085e-05, "loss": 0.0003, "step": 22199 }, { "epoch": 2.406764960971379, "grad_norm": 0.041610680520534515, "learning_rate": 1.9846220803713915e-05, "loss": 0.0011, "step": 22200 }, { "epoch": 2.406873373807459, "grad_norm": 0.13461673259735107, "learning_rate": 1.9842593935876976e-05, "loss": 0.0119, "step": 22201 }, { "epoch": 2.406981786643539, "grad_norm": 0.11395198106765747, "learning_rate": 1.983896706804004e-05, "loss": 0.0038, "step": 22202 }, { "epoch": 2.4070901994796183, "grad_norm": 0.6076453924179077, "learning_rate": 1.9835340200203107e-05, "loss": 0.0427, "step": 22203 }, { "epoch": 2.4071986123156983, "grad_norm": 0.24479016661643982, "learning_rate": 1.983171333236617e-05, "loss": 0.0193, "step": 22204 }, { "epoch": 2.407307025151778, "grad_norm": 0.16266635060310364, "learning_rate": 1.9828086464529232e-05, "loss": 0.006, "step": 22205 }, { "epoch": 2.407415437987858, "grad_norm": 0.34300607442855835, "learning_rate": 1.9824459596692296e-05, "loss": 0.0099, "step": 22206 }, { "epoch": 2.4075238508239374, "grad_norm": 0.7879237532615662, "learning_rate": 1.9820832728855364e-05, "loss": 0.0244, "step": 22207 }, { "epoch": 2.4076322636600174, "grad_norm": 0.25219854712486267, "learning_rate": 1.9817205861018424e-05, "loss": 0.0096, "step": 22208 }, { "epoch": 2.407740676496097, "grad_norm": 0.03215053305029869, "learning_rate": 1.981357899318149e-05, "loss": 0.0009, "step": 22209 }, { "epoch": 2.407849089332177, "grad_norm": 0.20014958083629608, "learning_rate": 1.9809952125344553e-05, "loss": 0.0106, "step": 22210 }, { "epoch": 2.407957502168257, "grad_norm": 0.6179288625717163, "learning_rate": 1.980632525750762e-05, "loss": 0.018, "step": 22211 }, { "epoch": 2.4080659150043364, "grad_norm": 0.1881912499666214, "learning_rate": 1.980269838967068e-05, "loss": 0.0141, "step": 22212 }, { "epoch": 2.4081743278404164, "grad_norm": 0.6897785067558289, "learning_rate": 1.9799071521833745e-05, "loss": 0.033, "step": 22213 }, { "epoch": 2.408282740676496, "grad_norm": 0.26546144485473633, "learning_rate": 1.979544465399681e-05, "loss": 0.0158, "step": 22214 }, { "epoch": 2.408391153512576, "grad_norm": 0.14979995787143707, "learning_rate": 1.9791817786159873e-05, "loss": 0.0119, "step": 22215 }, { "epoch": 2.408499566348656, "grad_norm": 0.5649095177650452, "learning_rate": 1.9788190918322937e-05, "loss": 0.0293, "step": 22216 }, { "epoch": 2.4086079791847355, "grad_norm": 0.008569609373807907, "learning_rate": 1.9784564050486e-05, "loss": 0.0003, "step": 22217 }, { "epoch": 2.4087163920208154, "grad_norm": 0.2022348940372467, "learning_rate": 1.9780937182649065e-05, "loss": 0.0049, "step": 22218 }, { "epoch": 2.408824804856895, "grad_norm": 0.02393081970512867, "learning_rate": 1.977731031481213e-05, "loss": 0.0009, "step": 22219 }, { "epoch": 2.408933217692975, "grad_norm": 0.05926434323191643, "learning_rate": 1.9773683446975194e-05, "loss": 0.003, "step": 22220 }, { "epoch": 2.4090416305290545, "grad_norm": 0.04076635465025902, "learning_rate": 1.9770056579138258e-05, "loss": 0.0008, "step": 22221 }, { "epoch": 2.4091500433651345, "grad_norm": 0.30755192041397095, "learning_rate": 1.9766429711301322e-05, "loss": 0.0221, "step": 22222 }, { "epoch": 2.409258456201214, "grad_norm": 0.4825984537601471, "learning_rate": 1.9762802843464383e-05, "loss": 0.0517, "step": 22223 }, { "epoch": 2.409366869037294, "grad_norm": 0.3468611240386963, "learning_rate": 1.975917597562745e-05, "loss": 0.0177, "step": 22224 }, { "epoch": 2.409475281873374, "grad_norm": 0.32448920607566833, "learning_rate": 1.9755549107790514e-05, "loss": 0.0174, "step": 22225 }, { "epoch": 2.4095836947094535, "grad_norm": 0.2208189219236374, "learning_rate": 1.9751922239953578e-05, "loss": 0.0111, "step": 22226 }, { "epoch": 2.4096921075455335, "grad_norm": 0.048531483858823776, "learning_rate": 1.974829537211664e-05, "loss": 0.001, "step": 22227 }, { "epoch": 2.409800520381613, "grad_norm": 0.01988508366048336, "learning_rate": 1.9744668504279707e-05, "loss": 0.0007, "step": 22228 }, { "epoch": 2.409908933217693, "grad_norm": 0.10065032541751862, "learning_rate": 1.974104163644277e-05, "loss": 0.0019, "step": 22229 }, { "epoch": 2.4100173460537726, "grad_norm": 0.21235017478466034, "learning_rate": 1.9737414768605835e-05, "loss": 0.005, "step": 22230 }, { "epoch": 2.4101257588898526, "grad_norm": 0.3615131080150604, "learning_rate": 1.9733787900768895e-05, "loss": 0.0141, "step": 22231 }, { "epoch": 2.410234171725932, "grad_norm": 0.431247353553772, "learning_rate": 1.973016103293196e-05, "loss": 0.013, "step": 22232 }, { "epoch": 2.410342584562012, "grad_norm": 0.47349250316619873, "learning_rate": 1.9726534165095027e-05, "loss": 0.0154, "step": 22233 }, { "epoch": 2.410450997398092, "grad_norm": 0.0152767114341259, "learning_rate": 1.9722907297258088e-05, "loss": 0.0007, "step": 22234 }, { "epoch": 2.4105594102341716, "grad_norm": 0.3789989650249481, "learning_rate": 1.9719280429421152e-05, "loss": 0.0136, "step": 22235 }, { "epoch": 2.4106678230702516, "grad_norm": 0.42302098870277405, "learning_rate": 1.9715653561584216e-05, "loss": 0.012, "step": 22236 }, { "epoch": 2.410776235906331, "grad_norm": 0.7474116086959839, "learning_rate": 1.9712026693747283e-05, "loss": 0.0756, "step": 22237 }, { "epoch": 2.410884648742411, "grad_norm": 0.269169420003891, "learning_rate": 1.9708399825910344e-05, "loss": 0.0092, "step": 22238 }, { "epoch": 2.410993061578491, "grad_norm": 0.10266099870204926, "learning_rate": 1.9704772958073408e-05, "loss": 0.0055, "step": 22239 }, { "epoch": 2.4111014744145707, "grad_norm": 0.620410680770874, "learning_rate": 1.9701146090236472e-05, "loss": 0.0475, "step": 22240 }, { "epoch": 2.4112098872506507, "grad_norm": 0.37364375591278076, "learning_rate": 1.9697519222399536e-05, "loss": 0.0452, "step": 22241 }, { "epoch": 2.41131830008673, "grad_norm": 0.15416115522384644, "learning_rate": 1.96938923545626e-05, "loss": 0.0071, "step": 22242 }, { "epoch": 2.41142671292281, "grad_norm": 0.23506636917591095, "learning_rate": 1.9690265486725665e-05, "loss": 0.013, "step": 22243 }, { "epoch": 2.4115351257588897, "grad_norm": 0.026185350492596626, "learning_rate": 1.968663861888873e-05, "loss": 0.0006, "step": 22244 }, { "epoch": 2.4116435385949697, "grad_norm": 0.061806969344615936, "learning_rate": 1.9683011751051793e-05, "loss": 0.0032, "step": 22245 }, { "epoch": 2.4117519514310493, "grad_norm": 0.16166777908802032, "learning_rate": 1.9679384883214857e-05, "loss": 0.0027, "step": 22246 }, { "epoch": 2.4118603642671292, "grad_norm": 0.2894691824913025, "learning_rate": 1.967575801537792e-05, "loss": 0.0205, "step": 22247 }, { "epoch": 2.4119687771032092, "grad_norm": 0.12018419802188873, "learning_rate": 1.9672131147540985e-05, "loss": 0.0028, "step": 22248 }, { "epoch": 2.4120771899392888, "grad_norm": 0.07602852582931519, "learning_rate": 1.9668504279704046e-05, "loss": 0.0013, "step": 22249 }, { "epoch": 2.4121856027753688, "grad_norm": 0.9298140406608582, "learning_rate": 1.9664877411867113e-05, "loss": 0.0196, "step": 22250 }, { "epoch": 2.4122940156114483, "grad_norm": 0.24200229346752167, "learning_rate": 1.9661250544030178e-05, "loss": 0.0227, "step": 22251 }, { "epoch": 2.4124024284475283, "grad_norm": 0.20102505385875702, "learning_rate": 1.965762367619324e-05, "loss": 0.0048, "step": 22252 }, { "epoch": 2.412510841283608, "grad_norm": 0.2885849177837372, "learning_rate": 1.9653996808356302e-05, "loss": 0.0152, "step": 22253 }, { "epoch": 2.412619254119688, "grad_norm": 0.024490375071763992, "learning_rate": 1.965036994051937e-05, "loss": 0.0005, "step": 22254 }, { "epoch": 2.4127276669557673, "grad_norm": 0.6557495594024658, "learning_rate": 1.9646743072682434e-05, "loss": 0.0259, "step": 22255 }, { "epoch": 2.4128360797918473, "grad_norm": 0.07689481973648071, "learning_rate": 1.9643116204845498e-05, "loss": 0.0024, "step": 22256 }, { "epoch": 2.4129444926279273, "grad_norm": 0.4180212914943695, "learning_rate": 1.963948933700856e-05, "loss": 0.022, "step": 22257 }, { "epoch": 2.413052905464007, "grad_norm": 0.23486527800559998, "learning_rate": 1.9635862469171623e-05, "loss": 0.0108, "step": 22258 }, { "epoch": 2.413161318300087, "grad_norm": 0.06167329102754593, "learning_rate": 1.963223560133469e-05, "loss": 0.0013, "step": 22259 }, { "epoch": 2.4132697311361664, "grad_norm": 0.561768114566803, "learning_rate": 1.962860873349775e-05, "loss": 0.0116, "step": 22260 }, { "epoch": 2.4133781439722464, "grad_norm": 0.038709044456481934, "learning_rate": 1.9624981865660815e-05, "loss": 0.0014, "step": 22261 }, { "epoch": 2.4134865568083264, "grad_norm": 0.03400809317827225, "learning_rate": 1.962135499782388e-05, "loss": 0.001, "step": 22262 }, { "epoch": 2.413594969644406, "grad_norm": 0.5588842034339905, "learning_rate": 1.9617728129986947e-05, "loss": 0.026, "step": 22263 }, { "epoch": 2.413703382480486, "grad_norm": 0.18700197339057922, "learning_rate": 1.9614101262150008e-05, "loss": 0.0057, "step": 22264 }, { "epoch": 2.4138117953165654, "grad_norm": 0.5659370422363281, "learning_rate": 1.961047439431307e-05, "loss": 0.0059, "step": 22265 }, { "epoch": 2.4139202081526454, "grad_norm": 0.07811512053012848, "learning_rate": 1.9606847526476136e-05, "loss": 0.0031, "step": 22266 }, { "epoch": 2.414028620988725, "grad_norm": 0.4338991045951843, "learning_rate": 1.96032206586392e-05, "loss": 0.0074, "step": 22267 }, { "epoch": 2.414137033824805, "grad_norm": 0.03284785896539688, "learning_rate": 1.9599593790802264e-05, "loss": 0.0012, "step": 22268 }, { "epoch": 2.4142454466608845, "grad_norm": 0.5512046217918396, "learning_rate": 1.9595966922965328e-05, "loss": 0.016, "step": 22269 }, { "epoch": 2.4143538594969645, "grad_norm": 0.07706846296787262, "learning_rate": 1.9592340055128392e-05, "loss": 0.002, "step": 22270 }, { "epoch": 2.4144622723330444, "grad_norm": 0.35447922348976135, "learning_rate": 1.9588713187291456e-05, "loss": 0.0192, "step": 22271 }, { "epoch": 2.414570685169124, "grad_norm": 0.7225720882415771, "learning_rate": 1.958508631945452e-05, "loss": 0.0404, "step": 22272 }, { "epoch": 2.414679098005204, "grad_norm": 0.24377386271953583, "learning_rate": 1.9581459451617584e-05, "loss": 0.0107, "step": 22273 }, { "epoch": 2.4147875108412835, "grad_norm": 0.9296860694885254, "learning_rate": 1.957783258378065e-05, "loss": 0.0881, "step": 22274 }, { "epoch": 2.4148959236773635, "grad_norm": 0.004582986701279879, "learning_rate": 1.957420571594371e-05, "loss": 0.0003, "step": 22275 }, { "epoch": 2.415004336513443, "grad_norm": 0.016098489984869957, "learning_rate": 1.9570578848106777e-05, "loss": 0.0006, "step": 22276 }, { "epoch": 2.415112749349523, "grad_norm": 0.14342236518859863, "learning_rate": 1.956695198026984e-05, "loss": 0.007, "step": 22277 }, { "epoch": 2.4152211621856026, "grad_norm": 0.42926591634750366, "learning_rate": 1.9563325112432905e-05, "loss": 0.0104, "step": 22278 }, { "epoch": 2.4153295750216826, "grad_norm": 0.30632296204566956, "learning_rate": 1.9559698244595966e-05, "loss": 0.0186, "step": 22279 }, { "epoch": 2.4154379878577625, "grad_norm": 0.0857710987329483, "learning_rate": 1.9556071376759033e-05, "loss": 0.0021, "step": 22280 }, { "epoch": 2.415546400693842, "grad_norm": 1.1724908351898193, "learning_rate": 1.9552444508922097e-05, "loss": 0.0116, "step": 22281 }, { "epoch": 2.415654813529922, "grad_norm": 0.22249065339565277, "learning_rate": 1.954881764108516e-05, "loss": 0.0079, "step": 22282 }, { "epoch": 2.4157632263660016, "grad_norm": 0.5858030319213867, "learning_rate": 1.9545190773248222e-05, "loss": 0.0242, "step": 22283 }, { "epoch": 2.4158716392020816, "grad_norm": 0.140059232711792, "learning_rate": 1.9541563905411286e-05, "loss": 0.0051, "step": 22284 }, { "epoch": 2.415980052038161, "grad_norm": 0.15846799314022064, "learning_rate": 1.9537937037574354e-05, "loss": 0.002, "step": 22285 }, { "epoch": 2.416088464874241, "grad_norm": 0.3699762225151062, "learning_rate": 1.9534310169737414e-05, "loss": 0.0298, "step": 22286 }, { "epoch": 2.4161968777103207, "grad_norm": 0.2819546163082123, "learning_rate": 1.953068330190048e-05, "loss": 0.013, "step": 22287 }, { "epoch": 2.4163052905464006, "grad_norm": 0.34403669834136963, "learning_rate": 1.9527056434063543e-05, "loss": 0.0366, "step": 22288 }, { "epoch": 2.4164137033824806, "grad_norm": 0.1338491439819336, "learning_rate": 1.952342956622661e-05, "loss": 0.004, "step": 22289 }, { "epoch": 2.41652211621856, "grad_norm": 0.04895315319299698, "learning_rate": 1.951980269838967e-05, "loss": 0.002, "step": 22290 }, { "epoch": 2.41663052905464, "grad_norm": 0.4628581702709198, "learning_rate": 1.9516175830552735e-05, "loss": 0.0249, "step": 22291 }, { "epoch": 2.4167389418907197, "grad_norm": 0.11528247594833374, "learning_rate": 1.95125489627158e-05, "loss": 0.0019, "step": 22292 }, { "epoch": 2.4168473547267997, "grad_norm": 0.31402599811553955, "learning_rate": 1.9508922094878863e-05, "loss": 0.0165, "step": 22293 }, { "epoch": 2.4169557675628797, "grad_norm": 0.5506410598754883, "learning_rate": 1.9505295227041927e-05, "loss": 0.017, "step": 22294 }, { "epoch": 2.417064180398959, "grad_norm": 0.03940678760409355, "learning_rate": 1.950166835920499e-05, "loss": 0.0011, "step": 22295 }, { "epoch": 2.417172593235039, "grad_norm": 0.3405865430831909, "learning_rate": 1.9498041491368055e-05, "loss": 0.0084, "step": 22296 }, { "epoch": 2.4172810060711187, "grad_norm": 0.929758608341217, "learning_rate": 1.949441462353112e-05, "loss": 0.0726, "step": 22297 }, { "epoch": 2.4173894189071987, "grad_norm": 0.3355571925640106, "learning_rate": 1.9490787755694184e-05, "loss": 0.0147, "step": 22298 }, { "epoch": 2.4174978317432783, "grad_norm": 0.057300642132759094, "learning_rate": 1.9487160887857248e-05, "loss": 0.0037, "step": 22299 }, { "epoch": 2.4176062445793582, "grad_norm": 0.1481972634792328, "learning_rate": 1.9483534020020312e-05, "loss": 0.0066, "step": 22300 }, { "epoch": 2.417714657415438, "grad_norm": 0.023906145244836807, "learning_rate": 1.9479907152183373e-05, "loss": 0.0017, "step": 22301 }, { "epoch": 2.4178230702515178, "grad_norm": 0.2308051586151123, "learning_rate": 1.947628028434644e-05, "loss": 0.0048, "step": 22302 }, { "epoch": 2.4179314830875978, "grad_norm": 0.013473758473992348, "learning_rate": 1.9472653416509504e-05, "loss": 0.0003, "step": 22303 }, { "epoch": 2.4180398959236773, "grad_norm": 0.05467718094587326, "learning_rate": 1.946902654867257e-05, "loss": 0.0017, "step": 22304 }, { "epoch": 2.4181483087597573, "grad_norm": 0.09550271928310394, "learning_rate": 1.946539968083563e-05, "loss": 0.0024, "step": 22305 }, { "epoch": 2.418256721595837, "grad_norm": 0.3976123034954071, "learning_rate": 1.9461772812998697e-05, "loss": 0.0068, "step": 22306 }, { "epoch": 2.418365134431917, "grad_norm": 0.009107436053454876, "learning_rate": 1.945814594516176e-05, "loss": 0.0003, "step": 22307 }, { "epoch": 2.4184735472679963, "grad_norm": 0.395796537399292, "learning_rate": 1.9454519077324825e-05, "loss": 0.026, "step": 22308 }, { "epoch": 2.4185819601040763, "grad_norm": 0.43981343507766724, "learning_rate": 1.9450892209487885e-05, "loss": 0.0099, "step": 22309 }, { "epoch": 2.418690372940156, "grad_norm": 0.5258937478065491, "learning_rate": 1.944726534165095e-05, "loss": 0.0189, "step": 22310 }, { "epoch": 2.418798785776236, "grad_norm": 0.03815248981118202, "learning_rate": 1.9443638473814017e-05, "loss": 0.0013, "step": 22311 }, { "epoch": 2.418907198612316, "grad_norm": 1.4079556465148926, "learning_rate": 1.9440011605977078e-05, "loss": 0.0894, "step": 22312 }, { "epoch": 2.4190156114483954, "grad_norm": 0.40963122248649597, "learning_rate": 1.9436384738140142e-05, "loss": 0.031, "step": 22313 }, { "epoch": 2.4191240242844754, "grad_norm": 0.3090563118457794, "learning_rate": 1.9432757870303206e-05, "loss": 0.0126, "step": 22314 }, { "epoch": 2.419232437120555, "grad_norm": 0.411765456199646, "learning_rate": 1.9429131002466273e-05, "loss": 0.0239, "step": 22315 }, { "epoch": 2.419340849956635, "grad_norm": 0.17026036977767944, "learning_rate": 1.9425504134629334e-05, "loss": 0.016, "step": 22316 }, { "epoch": 2.419449262792715, "grad_norm": 0.22202499210834503, "learning_rate": 1.9421877266792398e-05, "loss": 0.0099, "step": 22317 }, { "epoch": 2.4195576756287944, "grad_norm": 0.5404980778694153, "learning_rate": 1.9418250398955462e-05, "loss": 0.0349, "step": 22318 }, { "epoch": 2.4196660884648744, "grad_norm": 0.5541621446609497, "learning_rate": 1.9414623531118527e-05, "loss": 0.0062, "step": 22319 }, { "epoch": 2.419774501300954, "grad_norm": 0.3026607930660248, "learning_rate": 1.941099666328159e-05, "loss": 0.022, "step": 22320 }, { "epoch": 2.419882914137034, "grad_norm": 0.5329273343086243, "learning_rate": 1.9407369795444655e-05, "loss": 0.0127, "step": 22321 }, { "epoch": 2.4199913269731135, "grad_norm": 0.015864146873354912, "learning_rate": 1.940374292760772e-05, "loss": 0.0004, "step": 22322 }, { "epoch": 2.4200997398091935, "grad_norm": 0.20576678216457367, "learning_rate": 1.9400116059770783e-05, "loss": 0.0107, "step": 22323 }, { "epoch": 2.420208152645273, "grad_norm": 0.29925745725631714, "learning_rate": 1.9396489191933847e-05, "loss": 0.0138, "step": 22324 }, { "epoch": 2.420316565481353, "grad_norm": 0.42845189571380615, "learning_rate": 1.939286232409691e-05, "loss": 0.0147, "step": 22325 }, { "epoch": 2.420424978317433, "grad_norm": 0.07180602848529816, "learning_rate": 1.9389235456259975e-05, "loss": 0.003, "step": 22326 }, { "epoch": 2.4205333911535125, "grad_norm": 0.11884623765945435, "learning_rate": 1.9385608588423036e-05, "loss": 0.0021, "step": 22327 }, { "epoch": 2.4206418039895925, "grad_norm": 0.22013899683952332, "learning_rate": 1.9381981720586103e-05, "loss": 0.006, "step": 22328 }, { "epoch": 2.420750216825672, "grad_norm": 0.7169013023376465, "learning_rate": 1.9378354852749168e-05, "loss": 0.0094, "step": 22329 }, { "epoch": 2.420858629661752, "grad_norm": 0.08354472368955612, "learning_rate": 1.937472798491223e-05, "loss": 0.0025, "step": 22330 }, { "epoch": 2.4209670424978316, "grad_norm": 0.19395823776721954, "learning_rate": 1.9371101117075292e-05, "loss": 0.0223, "step": 22331 }, { "epoch": 2.4210754553339116, "grad_norm": 0.1448187381029129, "learning_rate": 1.936747424923836e-05, "loss": 0.006, "step": 22332 }, { "epoch": 2.421183868169991, "grad_norm": 0.027333548292517662, "learning_rate": 1.9363847381401424e-05, "loss": 0.0004, "step": 22333 }, { "epoch": 2.421292281006071, "grad_norm": 0.43841424584388733, "learning_rate": 1.9360220513564488e-05, "loss": 0.0125, "step": 22334 }, { "epoch": 2.421400693842151, "grad_norm": 0.18139822781085968, "learning_rate": 1.935659364572755e-05, "loss": 0.0048, "step": 22335 }, { "epoch": 2.4215091066782306, "grad_norm": 0.05672474950551987, "learning_rate": 1.9352966777890616e-05, "loss": 0.0011, "step": 22336 }, { "epoch": 2.4216175195143106, "grad_norm": 0.5036770105361938, "learning_rate": 1.934933991005368e-05, "loss": 0.0291, "step": 22337 }, { "epoch": 2.42172593235039, "grad_norm": 0.04968767240643501, "learning_rate": 1.934571304221674e-05, "loss": 0.0021, "step": 22338 }, { "epoch": 2.42183434518647, "grad_norm": 0.06512080878019333, "learning_rate": 1.9342086174379805e-05, "loss": 0.0041, "step": 22339 }, { "epoch": 2.42194275802255, "grad_norm": 0.21563397347927094, "learning_rate": 1.933845930654287e-05, "loss": 0.0033, "step": 22340 }, { "epoch": 2.4220511708586296, "grad_norm": 0.3400299549102783, "learning_rate": 1.9334832438705937e-05, "loss": 0.0114, "step": 22341 }, { "epoch": 2.4221595836947096, "grad_norm": 0.34519413113594055, "learning_rate": 1.9331205570868998e-05, "loss": 0.0237, "step": 22342 }, { "epoch": 2.422267996530789, "grad_norm": 0.030679970979690552, "learning_rate": 1.932757870303206e-05, "loss": 0.0015, "step": 22343 }, { "epoch": 2.422376409366869, "grad_norm": 0.3981316387653351, "learning_rate": 1.9323951835195126e-05, "loss": 0.02, "step": 22344 }, { "epoch": 2.4224848222029487, "grad_norm": 0.14815855026245117, "learning_rate": 1.932032496735819e-05, "loss": 0.0073, "step": 22345 }, { "epoch": 2.4225932350390287, "grad_norm": 0.21507133543491364, "learning_rate": 1.9316698099521254e-05, "loss": 0.0031, "step": 22346 }, { "epoch": 2.422701647875108, "grad_norm": 0.15371572971343994, "learning_rate": 1.9313071231684318e-05, "loss": 0.0058, "step": 22347 }, { "epoch": 2.422810060711188, "grad_norm": 0.03463813289999962, "learning_rate": 1.9309444363847382e-05, "loss": 0.0015, "step": 22348 }, { "epoch": 2.422918473547268, "grad_norm": 0.4310118556022644, "learning_rate": 1.9305817496010446e-05, "loss": 0.0132, "step": 22349 }, { "epoch": 2.4230268863833477, "grad_norm": 0.3571814000606537, "learning_rate": 1.930219062817351e-05, "loss": 0.0098, "step": 22350 }, { "epoch": 2.4231352992194277, "grad_norm": 0.04190029576420784, "learning_rate": 1.9298563760336574e-05, "loss": 0.0015, "step": 22351 }, { "epoch": 2.4232437120555073, "grad_norm": 0.04676158353686333, "learning_rate": 1.929493689249964e-05, "loss": 0.001, "step": 22352 }, { "epoch": 2.4233521248915872, "grad_norm": 0.3553076982498169, "learning_rate": 1.92913100246627e-05, "loss": 0.0067, "step": 22353 }, { "epoch": 2.423460537727667, "grad_norm": 1.6237120628356934, "learning_rate": 1.9287683156825767e-05, "loss": 0.0203, "step": 22354 }, { "epoch": 2.4235689505637468, "grad_norm": 0.6776833534240723, "learning_rate": 1.928405628898883e-05, "loss": 0.0118, "step": 22355 }, { "epoch": 2.4236773633998263, "grad_norm": 0.4938538670539856, "learning_rate": 1.9280429421151895e-05, "loss": 0.0064, "step": 22356 }, { "epoch": 2.4237857762359063, "grad_norm": 0.15290236473083496, "learning_rate": 1.9276802553314956e-05, "loss": 0.0028, "step": 22357 }, { "epoch": 2.4238941890719863, "grad_norm": 0.28431084752082825, "learning_rate": 1.9273175685478023e-05, "loss": 0.0092, "step": 22358 }, { "epoch": 2.424002601908066, "grad_norm": 0.050758518278598785, "learning_rate": 1.9269548817641087e-05, "loss": 0.001, "step": 22359 }, { "epoch": 2.424111014744146, "grad_norm": 0.28400829434394836, "learning_rate": 1.926592194980415e-05, "loss": 0.006, "step": 22360 }, { "epoch": 2.4242194275802254, "grad_norm": 0.27901723980903625, "learning_rate": 1.9262295081967212e-05, "loss": 0.0217, "step": 22361 }, { "epoch": 2.4243278404163053, "grad_norm": 0.3380624055862427, "learning_rate": 1.925866821413028e-05, "loss": 0.0101, "step": 22362 }, { "epoch": 2.4244362532523853, "grad_norm": 0.19682824611663818, "learning_rate": 1.9255041346293344e-05, "loss": 0.0135, "step": 22363 }, { "epoch": 2.424544666088465, "grad_norm": 0.2255931943655014, "learning_rate": 1.9251414478456404e-05, "loss": 0.019, "step": 22364 }, { "epoch": 2.424653078924545, "grad_norm": 0.4254080653190613, "learning_rate": 1.924778761061947e-05, "loss": 0.0661, "step": 22365 }, { "epoch": 2.4247614917606244, "grad_norm": 0.32041555643081665, "learning_rate": 1.9244160742782533e-05, "loss": 0.017, "step": 22366 }, { "epoch": 2.4248699045967044, "grad_norm": 0.3499475419521332, "learning_rate": 1.92405338749456e-05, "loss": 0.017, "step": 22367 }, { "epoch": 2.424978317432784, "grad_norm": 0.10530610382556915, "learning_rate": 1.923690700710866e-05, "loss": 0.0132, "step": 22368 }, { "epoch": 2.425086730268864, "grad_norm": 0.7873397469520569, "learning_rate": 1.9233280139271725e-05, "loss": 0.022, "step": 22369 }, { "epoch": 2.4251951431049434, "grad_norm": 0.05349385365843773, "learning_rate": 1.922965327143479e-05, "loss": 0.0012, "step": 22370 }, { "epoch": 2.4253035559410234, "grad_norm": 0.048888951539993286, "learning_rate": 1.9226026403597857e-05, "loss": 0.0015, "step": 22371 }, { "epoch": 2.4254119687771034, "grad_norm": 0.043052785098552704, "learning_rate": 1.9222399535760917e-05, "loss": 0.0006, "step": 22372 }, { "epoch": 2.425520381613183, "grad_norm": 0.11572155356407166, "learning_rate": 1.921877266792398e-05, "loss": 0.0076, "step": 22373 }, { "epoch": 2.425628794449263, "grad_norm": 0.24801161885261536, "learning_rate": 1.9215145800087046e-05, "loss": 0.02, "step": 22374 }, { "epoch": 2.4257372072853425, "grad_norm": 0.4680229127407074, "learning_rate": 1.921151893225011e-05, "loss": 0.0252, "step": 22375 }, { "epoch": 2.4258456201214225, "grad_norm": 0.08100324124097824, "learning_rate": 1.9207892064413174e-05, "loss": 0.0031, "step": 22376 }, { "epoch": 2.425954032957502, "grad_norm": 0.3422389328479767, "learning_rate": 1.9204265196576238e-05, "loss": 0.0196, "step": 22377 }, { "epoch": 2.426062445793582, "grad_norm": 0.48194172978401184, "learning_rate": 1.9200638328739302e-05, "loss": 0.0364, "step": 22378 }, { "epoch": 2.4261708586296615, "grad_norm": 0.03961923345923424, "learning_rate": 1.9197011460902366e-05, "loss": 0.0027, "step": 22379 }, { "epoch": 2.4262792714657415, "grad_norm": 0.3762153387069702, "learning_rate": 1.919338459306543e-05, "loss": 0.0135, "step": 22380 }, { "epoch": 2.4263876843018215, "grad_norm": 0.19800272583961487, "learning_rate": 1.9189757725228494e-05, "loss": 0.0075, "step": 22381 }, { "epoch": 2.426496097137901, "grad_norm": 1.0743567943572998, "learning_rate": 1.918613085739156e-05, "loss": 0.0236, "step": 22382 }, { "epoch": 2.426604509973981, "grad_norm": 0.32624396681785583, "learning_rate": 1.918250398955462e-05, "loss": 0.011, "step": 22383 }, { "epoch": 2.4267129228100606, "grad_norm": 0.35863953828811646, "learning_rate": 1.9178877121717687e-05, "loss": 0.0188, "step": 22384 }, { "epoch": 2.4268213356461406, "grad_norm": 0.6843384504318237, "learning_rate": 1.917525025388075e-05, "loss": 0.0237, "step": 22385 }, { "epoch": 2.4269297484822205, "grad_norm": 0.46105846762657166, "learning_rate": 1.9171623386043815e-05, "loss": 0.0172, "step": 22386 }, { "epoch": 2.4270381613183, "grad_norm": 0.19442997872829437, "learning_rate": 1.9167996518206875e-05, "loss": 0.01, "step": 22387 }, { "epoch": 2.42714657415438, "grad_norm": 0.2125992625951767, "learning_rate": 1.9164369650369943e-05, "loss": 0.0035, "step": 22388 }, { "epoch": 2.4272549869904596, "grad_norm": 0.0709022656083107, "learning_rate": 1.9160742782533007e-05, "loss": 0.0031, "step": 22389 }, { "epoch": 2.4273633998265396, "grad_norm": 0.04252518713474274, "learning_rate": 1.9157115914696068e-05, "loss": 0.0017, "step": 22390 }, { "epoch": 2.427471812662619, "grad_norm": 0.22513243556022644, "learning_rate": 1.9153489046859132e-05, "loss": 0.0089, "step": 22391 }, { "epoch": 2.427580225498699, "grad_norm": 0.29896244406700134, "learning_rate": 1.9149862179022196e-05, "loss": 0.0226, "step": 22392 }, { "epoch": 2.4276886383347787, "grad_norm": 0.19209282100200653, "learning_rate": 1.9146235311185264e-05, "loss": 0.0057, "step": 22393 }, { "epoch": 2.4277970511708586, "grad_norm": 0.18383896350860596, "learning_rate": 1.9142608443348324e-05, "loss": 0.0212, "step": 22394 }, { "epoch": 2.4279054640069386, "grad_norm": 0.36791080236434937, "learning_rate": 1.913898157551139e-05, "loss": 0.0161, "step": 22395 }, { "epoch": 2.428013876843018, "grad_norm": 0.06927280873060226, "learning_rate": 1.9135354707674452e-05, "loss": 0.0023, "step": 22396 }, { "epoch": 2.428122289679098, "grad_norm": 0.35806313157081604, "learning_rate": 1.913172783983752e-05, "loss": 0.0124, "step": 22397 }, { "epoch": 2.4282307025151777, "grad_norm": 0.4023003578186035, "learning_rate": 1.912810097200058e-05, "loss": 0.0079, "step": 22398 }, { "epoch": 2.4283391153512577, "grad_norm": 0.682968020439148, "learning_rate": 1.9124474104163645e-05, "loss": 0.0381, "step": 22399 }, { "epoch": 2.4284475281873372, "grad_norm": 0.08255541324615479, "learning_rate": 1.912084723632671e-05, "loss": 0.0047, "step": 22400 }, { "epoch": 2.428555941023417, "grad_norm": 0.0330754779279232, "learning_rate": 1.9117220368489773e-05, "loss": 0.0011, "step": 22401 }, { "epoch": 2.4286643538594967, "grad_norm": 0.10858996212482452, "learning_rate": 1.9113593500652837e-05, "loss": 0.0038, "step": 22402 }, { "epoch": 2.4287727666955767, "grad_norm": 0.606782078742981, "learning_rate": 1.91099666328159e-05, "loss": 0.0193, "step": 22403 }, { "epoch": 2.4288811795316567, "grad_norm": 0.9814237952232361, "learning_rate": 1.9106339764978965e-05, "loss": 0.023, "step": 22404 }, { "epoch": 2.4289895923677363, "grad_norm": 0.007121768780052662, "learning_rate": 1.910271289714203e-05, "loss": 0.0003, "step": 22405 }, { "epoch": 2.4290980052038162, "grad_norm": 0.44130071997642517, "learning_rate": 1.9099086029305093e-05, "loss": 0.0153, "step": 22406 }, { "epoch": 2.429206418039896, "grad_norm": 0.0449138805270195, "learning_rate": 1.9095459161468158e-05, "loss": 0.0019, "step": 22407 }, { "epoch": 2.4293148308759758, "grad_norm": 0.15890823304653168, "learning_rate": 1.9091832293631222e-05, "loss": 0.0043, "step": 22408 }, { "epoch": 2.4294232437120553, "grad_norm": 0.09583571553230286, "learning_rate": 1.9088205425794282e-05, "loss": 0.0022, "step": 22409 }, { "epoch": 2.4295316565481353, "grad_norm": 0.24204851686954498, "learning_rate": 1.908457855795735e-05, "loss": 0.0032, "step": 22410 }, { "epoch": 2.4296400693842153, "grad_norm": 0.06794213503599167, "learning_rate": 1.9080951690120414e-05, "loss": 0.0015, "step": 22411 }, { "epoch": 2.429748482220295, "grad_norm": 0.5272440314292908, "learning_rate": 1.9077324822283478e-05, "loss": 0.0322, "step": 22412 }, { "epoch": 2.429856895056375, "grad_norm": 0.5028864145278931, "learning_rate": 1.907369795444654e-05, "loss": 0.0098, "step": 22413 }, { "epoch": 2.4299653078924544, "grad_norm": 0.3541831970214844, "learning_rate": 1.9070071086609606e-05, "loss": 0.0175, "step": 22414 }, { "epoch": 2.4300737207285343, "grad_norm": 0.802454948425293, "learning_rate": 1.906644421877267e-05, "loss": 0.0326, "step": 22415 }, { "epoch": 2.430182133564614, "grad_norm": 0.03158295899629593, "learning_rate": 1.906281735093573e-05, "loss": 0.0009, "step": 22416 }, { "epoch": 2.430290546400694, "grad_norm": 0.4330398738384247, "learning_rate": 1.9059190483098795e-05, "loss": 0.018, "step": 22417 }, { "epoch": 2.430398959236774, "grad_norm": 0.08499117195606232, "learning_rate": 1.905556361526186e-05, "loss": 0.004, "step": 22418 }, { "epoch": 2.4305073720728534, "grad_norm": 0.18306481838226318, "learning_rate": 1.9051936747424927e-05, "loss": 0.0039, "step": 22419 }, { "epoch": 2.4306157849089334, "grad_norm": 0.2807021141052246, "learning_rate": 1.9048309879587988e-05, "loss": 0.0097, "step": 22420 }, { "epoch": 2.430724197745013, "grad_norm": 0.4810023009777069, "learning_rate": 1.904468301175105e-05, "loss": 0.0225, "step": 22421 }, { "epoch": 2.430832610581093, "grad_norm": 0.26975545287132263, "learning_rate": 1.9041056143914116e-05, "loss": 0.0166, "step": 22422 }, { "epoch": 2.4309410234171724, "grad_norm": 0.23816603422164917, "learning_rate": 1.9037429276077183e-05, "loss": 0.0233, "step": 22423 }, { "epoch": 2.4310494362532524, "grad_norm": 0.9153109192848206, "learning_rate": 1.9033802408240244e-05, "loss": 0.0674, "step": 22424 }, { "epoch": 2.431157849089332, "grad_norm": 0.6407825946807861, "learning_rate": 1.9030175540403308e-05, "loss": 0.0467, "step": 22425 }, { "epoch": 2.431266261925412, "grad_norm": 0.23943693935871124, "learning_rate": 1.9026548672566372e-05, "loss": 0.0125, "step": 22426 }, { "epoch": 2.431374674761492, "grad_norm": 0.05452825501561165, "learning_rate": 1.9022921804729436e-05, "loss": 0.0012, "step": 22427 }, { "epoch": 2.4314830875975715, "grad_norm": 0.16427965462207794, "learning_rate": 1.90192949368925e-05, "loss": 0.0067, "step": 22428 }, { "epoch": 2.4315915004336515, "grad_norm": 0.19458501040935516, "learning_rate": 1.9015668069055565e-05, "loss": 0.0032, "step": 22429 }, { "epoch": 2.431699913269731, "grad_norm": 0.2905532717704773, "learning_rate": 1.901204120121863e-05, "loss": 0.0082, "step": 22430 }, { "epoch": 2.431808326105811, "grad_norm": 0.013001888059079647, "learning_rate": 1.9008414333381693e-05, "loss": 0.0005, "step": 22431 }, { "epoch": 2.4319167389418905, "grad_norm": 0.15647481381893158, "learning_rate": 1.9004787465544757e-05, "loss": 0.0116, "step": 22432 }, { "epoch": 2.4320251517779705, "grad_norm": 0.13622431457042694, "learning_rate": 1.900116059770782e-05, "loss": 0.0069, "step": 22433 }, { "epoch": 2.43213356461405, "grad_norm": 0.4811602532863617, "learning_rate": 1.8997533729870885e-05, "loss": 0.0342, "step": 22434 }, { "epoch": 2.43224197745013, "grad_norm": 0.09496606141328812, "learning_rate": 1.8993906862033946e-05, "loss": 0.0016, "step": 22435 }, { "epoch": 2.43235039028621, "grad_norm": 0.24426327645778656, "learning_rate": 1.8990279994197013e-05, "loss": 0.0085, "step": 22436 }, { "epoch": 2.4324588031222896, "grad_norm": 0.12364772707223892, "learning_rate": 1.8986653126360077e-05, "loss": 0.0036, "step": 22437 }, { "epoch": 2.4325672159583696, "grad_norm": 0.44805002212524414, "learning_rate": 1.898302625852314e-05, "loss": 0.0151, "step": 22438 }, { "epoch": 2.432675628794449, "grad_norm": 0.3003665506839752, "learning_rate": 1.8979399390686202e-05, "loss": 0.0137, "step": 22439 }, { "epoch": 2.432784041630529, "grad_norm": 0.692666232585907, "learning_rate": 1.897577252284927e-05, "loss": 0.0375, "step": 22440 }, { "epoch": 2.432892454466609, "grad_norm": 0.18349412083625793, "learning_rate": 1.8972145655012334e-05, "loss": 0.005, "step": 22441 }, { "epoch": 2.4330008673026886, "grad_norm": 0.3370670676231384, "learning_rate": 1.8968518787175394e-05, "loss": 0.0106, "step": 22442 }, { "epoch": 2.4331092801387686, "grad_norm": 0.158386692404747, "learning_rate": 1.896489191933846e-05, "loss": 0.0083, "step": 22443 }, { "epoch": 2.433217692974848, "grad_norm": 0.47495073080062866, "learning_rate": 1.8961265051501526e-05, "loss": 0.0189, "step": 22444 }, { "epoch": 2.433326105810928, "grad_norm": 0.1354014128446579, "learning_rate": 1.895763818366459e-05, "loss": 0.0032, "step": 22445 }, { "epoch": 2.4334345186470077, "grad_norm": 0.298458456993103, "learning_rate": 1.895401131582765e-05, "loss": 0.0169, "step": 22446 }, { "epoch": 2.4335429314830876, "grad_norm": 0.3305337131023407, "learning_rate": 1.8950384447990715e-05, "loss": 0.0173, "step": 22447 }, { "epoch": 2.433651344319167, "grad_norm": 0.41777175664901733, "learning_rate": 1.894675758015378e-05, "loss": 0.0403, "step": 22448 }, { "epoch": 2.433759757155247, "grad_norm": 0.518334150314331, "learning_rate": 1.8943130712316847e-05, "loss": 0.053, "step": 22449 }, { "epoch": 2.433868169991327, "grad_norm": 0.11279036104679108, "learning_rate": 1.8939503844479907e-05, "loss": 0.0044, "step": 22450 }, { "epoch": 2.4339765828274067, "grad_norm": 0.06814210116863251, "learning_rate": 1.893587697664297e-05, "loss": 0.0017, "step": 22451 }, { "epoch": 2.4340849956634867, "grad_norm": 0.01037073228508234, "learning_rate": 1.8932250108806036e-05, "loss": 0.0003, "step": 22452 }, { "epoch": 2.4341934084995662, "grad_norm": 0.12114449590444565, "learning_rate": 1.89286232409691e-05, "loss": 0.0056, "step": 22453 }, { "epoch": 2.434301821335646, "grad_norm": 0.4302559792995453, "learning_rate": 1.8924996373132164e-05, "loss": 0.017, "step": 22454 }, { "epoch": 2.4344102341717258, "grad_norm": 0.03422718495130539, "learning_rate": 1.8921369505295228e-05, "loss": 0.0014, "step": 22455 }, { "epoch": 2.4345186470078057, "grad_norm": 0.16312909126281738, "learning_rate": 1.8917742637458292e-05, "loss": 0.0061, "step": 22456 }, { "epoch": 2.4346270598438853, "grad_norm": 0.1425856500864029, "learning_rate": 1.8914115769621356e-05, "loss": 0.0073, "step": 22457 }, { "epoch": 2.4347354726799653, "grad_norm": 0.2507503628730774, "learning_rate": 1.891048890178442e-05, "loss": 0.0104, "step": 22458 }, { "epoch": 2.4348438855160452, "grad_norm": 0.5314831137657166, "learning_rate": 1.8906862033947484e-05, "loss": 0.0144, "step": 22459 }, { "epoch": 2.434952298352125, "grad_norm": 0.1227702796459198, "learning_rate": 1.890323516611055e-05, "loss": 0.0065, "step": 22460 }, { "epoch": 2.4350607111882048, "grad_norm": 0.07721756398677826, "learning_rate": 1.889960829827361e-05, "loss": 0.0066, "step": 22461 }, { "epoch": 2.4351691240242843, "grad_norm": 0.1736232042312622, "learning_rate": 1.8895981430436677e-05, "loss": 0.0047, "step": 22462 }, { "epoch": 2.4352775368603643, "grad_norm": 0.6157433986663818, "learning_rate": 1.889235456259974e-05, "loss": 0.0148, "step": 22463 }, { "epoch": 2.4353859496964443, "grad_norm": 0.11102001368999481, "learning_rate": 1.8888727694762805e-05, "loss": 0.0051, "step": 22464 }, { "epoch": 2.435494362532524, "grad_norm": 0.052805181592702866, "learning_rate": 1.8885100826925866e-05, "loss": 0.0021, "step": 22465 }, { "epoch": 2.435602775368604, "grad_norm": 0.3710218667984009, "learning_rate": 1.8881473959088933e-05, "loss": 0.0133, "step": 22466 }, { "epoch": 2.4357111882046834, "grad_norm": 0.2631613612174988, "learning_rate": 1.8877847091251997e-05, "loss": 0.0215, "step": 22467 }, { "epoch": 2.4358196010407633, "grad_norm": 0.013147303834557533, "learning_rate": 1.8874220223415058e-05, "loss": 0.0009, "step": 22468 }, { "epoch": 2.435928013876843, "grad_norm": 0.30755847692489624, "learning_rate": 1.8870593355578122e-05, "loss": 0.0633, "step": 22469 }, { "epoch": 2.436036426712923, "grad_norm": 0.08647464960813522, "learning_rate": 1.886696648774119e-05, "loss": 0.0025, "step": 22470 }, { "epoch": 2.4361448395490024, "grad_norm": 0.22402013838291168, "learning_rate": 1.8863339619904254e-05, "loss": 0.0054, "step": 22471 }, { "epoch": 2.4362532523850824, "grad_norm": 0.5645180344581604, "learning_rate": 1.8859712752067314e-05, "loss": 0.0125, "step": 22472 }, { "epoch": 2.4363616652211624, "grad_norm": 0.27247121930122375, "learning_rate": 1.885608588423038e-05, "loss": 0.0207, "step": 22473 }, { "epoch": 2.436470078057242, "grad_norm": 0.8206548690795898, "learning_rate": 1.8852459016393442e-05, "loss": 0.0151, "step": 22474 }, { "epoch": 2.436578490893322, "grad_norm": 0.08631675690412521, "learning_rate": 1.884883214855651e-05, "loss": 0.0034, "step": 22475 }, { "epoch": 2.4366869037294014, "grad_norm": 0.28558188676834106, "learning_rate": 1.884520528071957e-05, "loss": 0.007, "step": 22476 }, { "epoch": 2.4367953165654814, "grad_norm": 0.2992844581604004, "learning_rate": 1.8841578412882635e-05, "loss": 0.0283, "step": 22477 }, { "epoch": 2.436903729401561, "grad_norm": 0.39049583673477173, "learning_rate": 1.88379515450457e-05, "loss": 0.0354, "step": 22478 }, { "epoch": 2.437012142237641, "grad_norm": 0.5609043836593628, "learning_rate": 1.8834324677208763e-05, "loss": 0.0296, "step": 22479 }, { "epoch": 2.4371205550737205, "grad_norm": 0.025686070322990417, "learning_rate": 1.8830697809371827e-05, "loss": 0.0014, "step": 22480 }, { "epoch": 2.4372289679098005, "grad_norm": 0.4009736180305481, "learning_rate": 1.882707094153489e-05, "loss": 0.0073, "step": 22481 }, { "epoch": 2.4373373807458805, "grad_norm": 0.31510472297668457, "learning_rate": 1.8823444073697955e-05, "loss": 0.0132, "step": 22482 }, { "epoch": 2.43744579358196, "grad_norm": 0.02266431786119938, "learning_rate": 1.881981720586102e-05, "loss": 0.0013, "step": 22483 }, { "epoch": 2.43755420641804, "grad_norm": 0.04460814967751503, "learning_rate": 1.8816190338024084e-05, "loss": 0.0008, "step": 22484 }, { "epoch": 2.4376626192541195, "grad_norm": 0.16400916874408722, "learning_rate": 1.8812563470187148e-05, "loss": 0.0057, "step": 22485 }, { "epoch": 2.4377710320901995, "grad_norm": 0.03794652968645096, "learning_rate": 1.8808936602350212e-05, "loss": 0.0009, "step": 22486 }, { "epoch": 2.4378794449262795, "grad_norm": 0.06750157475471497, "learning_rate": 1.8805309734513272e-05, "loss": 0.0015, "step": 22487 }, { "epoch": 2.437987857762359, "grad_norm": 0.3454088866710663, "learning_rate": 1.880168286667634e-05, "loss": 0.0072, "step": 22488 }, { "epoch": 2.438096270598439, "grad_norm": 0.5748313665390015, "learning_rate": 1.8798055998839404e-05, "loss": 0.0147, "step": 22489 }, { "epoch": 2.4382046834345186, "grad_norm": 0.8446692228317261, "learning_rate": 1.8794429131002468e-05, "loss": 0.0452, "step": 22490 }, { "epoch": 2.4383130962705986, "grad_norm": 0.13802562654018402, "learning_rate": 1.879080226316553e-05, "loss": 0.006, "step": 22491 }, { "epoch": 2.438421509106678, "grad_norm": 0.3328436017036438, "learning_rate": 1.8787175395328596e-05, "loss": 0.0145, "step": 22492 }, { "epoch": 2.438529921942758, "grad_norm": 0.08248448371887207, "learning_rate": 1.878354852749166e-05, "loss": 0.0024, "step": 22493 }, { "epoch": 2.4386383347788376, "grad_norm": 0.33661943674087524, "learning_rate": 1.877992165965472e-05, "loss": 0.0099, "step": 22494 }, { "epoch": 2.4387467476149176, "grad_norm": 0.053406745195388794, "learning_rate": 1.8776294791817785e-05, "loss": 0.0023, "step": 22495 }, { "epoch": 2.4388551604509976, "grad_norm": 0.0449434369802475, "learning_rate": 1.8772667923980853e-05, "loss": 0.0009, "step": 22496 }, { "epoch": 2.438963573287077, "grad_norm": 0.4638935625553131, "learning_rate": 1.8769041056143917e-05, "loss": 0.0185, "step": 22497 }, { "epoch": 2.439071986123157, "grad_norm": 0.19681881368160248, "learning_rate": 1.8765414188306978e-05, "loss": 0.004, "step": 22498 }, { "epoch": 2.4391803989592367, "grad_norm": 0.045569051057100296, "learning_rate": 1.8761787320470042e-05, "loss": 0.0024, "step": 22499 }, { "epoch": 2.4392888117953166, "grad_norm": 0.6984700560569763, "learning_rate": 1.8758160452633106e-05, "loss": 0.0293, "step": 22500 }, { "epoch": 2.439397224631396, "grad_norm": 0.18341940641403198, "learning_rate": 1.8754533584796173e-05, "loss": 0.0201, "step": 22501 }, { "epoch": 2.439505637467476, "grad_norm": 0.0862254649400711, "learning_rate": 1.8750906716959234e-05, "loss": 0.0077, "step": 22502 }, { "epoch": 2.4396140503035557, "grad_norm": 0.06674156337976456, "learning_rate": 1.8747279849122298e-05, "loss": 0.0032, "step": 22503 }, { "epoch": 2.4397224631396357, "grad_norm": 0.3736957013607025, "learning_rate": 1.8743652981285362e-05, "loss": 0.0089, "step": 22504 }, { "epoch": 2.4398308759757157, "grad_norm": 0.18431754410266876, "learning_rate": 1.8740026113448426e-05, "loss": 0.0049, "step": 22505 }, { "epoch": 2.4399392888117952, "grad_norm": 0.023836100473999977, "learning_rate": 1.873639924561149e-05, "loss": 0.0008, "step": 22506 }, { "epoch": 2.440047701647875, "grad_norm": 0.10267618298530579, "learning_rate": 1.8732772377774555e-05, "loss": 0.0015, "step": 22507 }, { "epoch": 2.4401561144839548, "grad_norm": 0.06642840057611465, "learning_rate": 1.872914550993762e-05, "loss": 0.0019, "step": 22508 }, { "epoch": 2.4402645273200347, "grad_norm": 0.2168828248977661, "learning_rate": 1.8725518642100683e-05, "loss": 0.0058, "step": 22509 }, { "epoch": 2.4403729401561147, "grad_norm": 0.7405616641044617, "learning_rate": 1.8721891774263747e-05, "loss": 0.0377, "step": 22510 }, { "epoch": 2.4404813529921943, "grad_norm": 0.2544853389263153, "learning_rate": 1.871826490642681e-05, "loss": 0.0056, "step": 22511 }, { "epoch": 2.4405897658282742, "grad_norm": 0.3729017972946167, "learning_rate": 1.8714638038589875e-05, "loss": 0.0079, "step": 22512 }, { "epoch": 2.440698178664354, "grad_norm": 0.2889224588871002, "learning_rate": 1.8711011170752936e-05, "loss": 0.024, "step": 22513 }, { "epoch": 2.4408065915004338, "grad_norm": 0.023918408900499344, "learning_rate": 1.8707384302916003e-05, "loss": 0.0009, "step": 22514 }, { "epoch": 2.4409150043365133, "grad_norm": 0.4861137270927429, "learning_rate": 1.8703757435079067e-05, "loss": 0.0254, "step": 22515 }, { "epoch": 2.4410234171725933, "grad_norm": 0.07201509177684784, "learning_rate": 1.870013056724213e-05, "loss": 0.0015, "step": 22516 }, { "epoch": 2.441131830008673, "grad_norm": 0.21906989812850952, "learning_rate": 1.8696503699405192e-05, "loss": 0.0065, "step": 22517 }, { "epoch": 2.441240242844753, "grad_norm": 0.023361429572105408, "learning_rate": 1.869287683156826e-05, "loss": 0.0008, "step": 22518 }, { "epoch": 2.441348655680833, "grad_norm": 0.5451588034629822, "learning_rate": 1.8689249963731324e-05, "loss": 0.0307, "step": 22519 }, { "epoch": 2.4414570685169124, "grad_norm": 0.30489400029182434, "learning_rate": 1.8685623095894385e-05, "loss": 0.0184, "step": 22520 }, { "epoch": 2.4415654813529923, "grad_norm": 0.03881627321243286, "learning_rate": 1.868199622805745e-05, "loss": 0.0009, "step": 22521 }, { "epoch": 2.441673894189072, "grad_norm": 0.020754583179950714, "learning_rate": 1.8678369360220516e-05, "loss": 0.0008, "step": 22522 }, { "epoch": 2.441782307025152, "grad_norm": 0.2844342887401581, "learning_rate": 1.867474249238358e-05, "loss": 0.0069, "step": 22523 }, { "epoch": 2.4418907198612314, "grad_norm": 1.4764665365219116, "learning_rate": 1.867111562454664e-05, "loss": 0.016, "step": 22524 }, { "epoch": 2.4419991326973114, "grad_norm": 0.0488557368516922, "learning_rate": 1.8667488756709705e-05, "loss": 0.0017, "step": 22525 }, { "epoch": 2.442107545533391, "grad_norm": 0.4487963914871216, "learning_rate": 1.866386188887277e-05, "loss": 0.0215, "step": 22526 }, { "epoch": 2.442215958369471, "grad_norm": 0.21239177882671356, "learning_rate": 1.8660235021035837e-05, "loss": 0.0032, "step": 22527 }, { "epoch": 2.442324371205551, "grad_norm": 0.29841315746307373, "learning_rate": 1.8656608153198897e-05, "loss": 0.0163, "step": 22528 }, { "epoch": 2.4424327840416304, "grad_norm": 0.3876361846923828, "learning_rate": 1.865298128536196e-05, "loss": 0.0158, "step": 22529 }, { "epoch": 2.4425411968777104, "grad_norm": 0.6968163251876831, "learning_rate": 1.8649354417525026e-05, "loss": 0.0538, "step": 22530 }, { "epoch": 2.44264960971379, "grad_norm": 0.1353745460510254, "learning_rate": 1.864572754968809e-05, "loss": 0.0052, "step": 22531 }, { "epoch": 2.44275802254987, "grad_norm": 0.0695176050066948, "learning_rate": 1.8642100681851154e-05, "loss": 0.0036, "step": 22532 }, { "epoch": 2.44286643538595, "grad_norm": 0.49879005551338196, "learning_rate": 1.8638473814014218e-05, "loss": 0.0285, "step": 22533 }, { "epoch": 2.4429748482220295, "grad_norm": 0.6263419985771179, "learning_rate": 1.8634846946177282e-05, "loss": 0.0268, "step": 22534 }, { "epoch": 2.4430832610581095, "grad_norm": 0.3995843231678009, "learning_rate": 1.8631220078340346e-05, "loss": 0.0255, "step": 22535 }, { "epoch": 2.443191673894189, "grad_norm": 0.03118862770497799, "learning_rate": 1.862759321050341e-05, "loss": 0.0015, "step": 22536 }, { "epoch": 2.443300086730269, "grad_norm": 0.019843649119138718, "learning_rate": 1.8623966342666474e-05, "loss": 0.0009, "step": 22537 }, { "epoch": 2.4434084995663485, "grad_norm": 0.2963521480560303, "learning_rate": 1.862033947482954e-05, "loss": 0.028, "step": 22538 }, { "epoch": 2.4435169124024285, "grad_norm": 2.4594390392303467, "learning_rate": 1.8616712606992603e-05, "loss": 0.0176, "step": 22539 }, { "epoch": 2.443625325238508, "grad_norm": 0.24274495244026184, "learning_rate": 1.8613085739155667e-05, "loss": 0.0109, "step": 22540 }, { "epoch": 2.443733738074588, "grad_norm": 0.3028285801410675, "learning_rate": 1.860945887131873e-05, "loss": 0.0136, "step": 22541 }, { "epoch": 2.443842150910668, "grad_norm": 0.414546936750412, "learning_rate": 1.8605832003481795e-05, "loss": 0.0272, "step": 22542 }, { "epoch": 2.4439505637467476, "grad_norm": 0.31544581055641174, "learning_rate": 1.8602205135644856e-05, "loss": 0.0092, "step": 22543 }, { "epoch": 2.4440589765828276, "grad_norm": 0.04471280798316002, "learning_rate": 1.8598578267807923e-05, "loss": 0.0014, "step": 22544 }, { "epoch": 2.444167389418907, "grad_norm": 0.02146095409989357, "learning_rate": 1.8594951399970987e-05, "loss": 0.0012, "step": 22545 }, { "epoch": 2.444275802254987, "grad_norm": 0.04020777717232704, "learning_rate": 1.8591324532134048e-05, "loss": 0.0013, "step": 22546 }, { "epoch": 2.4443842150910666, "grad_norm": 0.06381653249263763, "learning_rate": 1.8587697664297112e-05, "loss": 0.0012, "step": 22547 }, { "epoch": 2.4444926279271466, "grad_norm": 0.06723971664905548, "learning_rate": 1.858407079646018e-05, "loss": 0.003, "step": 22548 }, { "epoch": 2.444601040763226, "grad_norm": 0.020572874695062637, "learning_rate": 1.8580443928623244e-05, "loss": 0.0008, "step": 22549 }, { "epoch": 2.444709453599306, "grad_norm": 0.16797295212745667, "learning_rate": 1.8576817060786304e-05, "loss": 0.0057, "step": 22550 }, { "epoch": 2.444817866435386, "grad_norm": 0.11588191241025925, "learning_rate": 1.857319019294937e-05, "loss": 0.0051, "step": 22551 }, { "epoch": 2.4449262792714657, "grad_norm": 0.07566792517900467, "learning_rate": 1.8569563325112436e-05, "loss": 0.0021, "step": 22552 }, { "epoch": 2.4450346921075456, "grad_norm": 0.43539705872535706, "learning_rate": 1.85659364572755e-05, "loss": 0.0253, "step": 22553 }, { "epoch": 2.445143104943625, "grad_norm": 0.47008636593818665, "learning_rate": 1.856230958943856e-05, "loss": 0.0094, "step": 22554 }, { "epoch": 2.445251517779705, "grad_norm": 0.26858338713645935, "learning_rate": 1.8558682721601625e-05, "loss": 0.036, "step": 22555 }, { "epoch": 2.4453599306157847, "grad_norm": 0.12094222009181976, "learning_rate": 1.855505585376469e-05, "loss": 0.003, "step": 22556 }, { "epoch": 2.4454683434518647, "grad_norm": 0.292815625667572, "learning_rate": 1.8551428985927753e-05, "loss": 0.0043, "step": 22557 }, { "epoch": 2.4455767562879447, "grad_norm": 0.2285940796136856, "learning_rate": 1.8547802118090817e-05, "loss": 0.0287, "step": 22558 }, { "epoch": 2.4456851691240242, "grad_norm": 0.08151058852672577, "learning_rate": 1.854417525025388e-05, "loss": 0.0033, "step": 22559 }, { "epoch": 2.445793581960104, "grad_norm": 0.26088035106658936, "learning_rate": 1.8540548382416945e-05, "loss": 0.0071, "step": 22560 }, { "epoch": 2.4459019947961838, "grad_norm": 0.12770628929138184, "learning_rate": 1.853692151458001e-05, "loss": 0.0064, "step": 22561 }, { "epoch": 2.4460104076322637, "grad_norm": 0.40978559851646423, "learning_rate": 1.8533294646743074e-05, "loss": 0.0214, "step": 22562 }, { "epoch": 2.4461188204683433, "grad_norm": 0.13204900920391083, "learning_rate": 1.8529667778906138e-05, "loss": 0.0048, "step": 22563 }, { "epoch": 2.4462272333044233, "grad_norm": 0.09241076558828354, "learning_rate": 1.8526040911069202e-05, "loss": 0.0059, "step": 22564 }, { "epoch": 2.4463356461405033, "grad_norm": 0.3199125826358795, "learning_rate": 1.8522414043232266e-05, "loss": 0.0079, "step": 22565 }, { "epoch": 2.446444058976583, "grad_norm": 0.42470303177833557, "learning_rate": 1.851878717539533e-05, "loss": 0.0159, "step": 22566 }, { "epoch": 2.4465524718126628, "grad_norm": 0.11225871741771698, "learning_rate": 1.8515160307558394e-05, "loss": 0.0033, "step": 22567 }, { "epoch": 2.4466608846487423, "grad_norm": 0.27363884449005127, "learning_rate": 1.8511533439721458e-05, "loss": 0.0115, "step": 22568 }, { "epoch": 2.4467692974848223, "grad_norm": 0.6126505732536316, "learning_rate": 1.850790657188452e-05, "loss": 0.0162, "step": 22569 }, { "epoch": 2.446877710320902, "grad_norm": 0.24848905205726624, "learning_rate": 1.8504279704047586e-05, "loss": 0.0078, "step": 22570 }, { "epoch": 2.446986123156982, "grad_norm": 0.1250009983778, "learning_rate": 1.850065283621065e-05, "loss": 0.0046, "step": 22571 }, { "epoch": 2.4470945359930614, "grad_norm": 0.5004531741142273, "learning_rate": 1.8497025968373715e-05, "loss": 0.0495, "step": 22572 }, { "epoch": 2.4472029488291414, "grad_norm": 0.09599778801202774, "learning_rate": 1.8493399100536775e-05, "loss": 0.0023, "step": 22573 }, { "epoch": 2.4473113616652213, "grad_norm": 0.05587185546755791, "learning_rate": 1.8489772232699843e-05, "loss": 0.0031, "step": 22574 }, { "epoch": 2.447419774501301, "grad_norm": 0.10801661014556885, "learning_rate": 1.8486145364862907e-05, "loss": 0.0048, "step": 22575 }, { "epoch": 2.447528187337381, "grad_norm": 0.37744322419166565, "learning_rate": 1.8482518497025968e-05, "loss": 0.0284, "step": 22576 }, { "epoch": 2.4476366001734604, "grad_norm": 0.37810197472572327, "learning_rate": 1.8478891629189032e-05, "loss": 0.0411, "step": 22577 }, { "epoch": 2.4477450130095404, "grad_norm": 0.2591671049594879, "learning_rate": 1.84752647613521e-05, "loss": 0.0175, "step": 22578 }, { "epoch": 2.44785342584562, "grad_norm": 0.29059356451034546, "learning_rate": 1.8471637893515163e-05, "loss": 0.0118, "step": 22579 }, { "epoch": 2.4479618386817, "grad_norm": 0.26028990745544434, "learning_rate": 1.8468011025678224e-05, "loss": 0.0335, "step": 22580 }, { "epoch": 2.4480702515177795, "grad_norm": 0.08340941369533539, "learning_rate": 1.8464384157841288e-05, "loss": 0.0027, "step": 22581 }, { "epoch": 2.4481786643538594, "grad_norm": 0.10643038153648376, "learning_rate": 1.8460757290004352e-05, "loss": 0.0078, "step": 22582 }, { "epoch": 2.4482870771899394, "grad_norm": 1.264926791191101, "learning_rate": 1.8457130422167416e-05, "loss": 0.028, "step": 22583 }, { "epoch": 2.448395490026019, "grad_norm": 0.7916272282600403, "learning_rate": 1.845350355433048e-05, "loss": 0.037, "step": 22584 }, { "epoch": 2.448503902862099, "grad_norm": 0.13961198925971985, "learning_rate": 1.8449876686493545e-05, "loss": 0.0025, "step": 22585 }, { "epoch": 2.4486123156981785, "grad_norm": 0.2529444992542267, "learning_rate": 1.844624981865661e-05, "loss": 0.0117, "step": 22586 }, { "epoch": 2.4487207285342585, "grad_norm": 0.06091795489192009, "learning_rate": 1.8442622950819673e-05, "loss": 0.0015, "step": 22587 }, { "epoch": 2.4488291413703385, "grad_norm": 0.2186698317527771, "learning_rate": 1.8438996082982737e-05, "loss": 0.015, "step": 22588 }, { "epoch": 2.448937554206418, "grad_norm": 0.15851663053035736, "learning_rate": 1.84353692151458e-05, "loss": 0.0049, "step": 22589 }, { "epoch": 2.449045967042498, "grad_norm": 0.267348051071167, "learning_rate": 1.8431742347308865e-05, "loss": 0.0096, "step": 22590 }, { "epoch": 2.4491543798785775, "grad_norm": 0.5948613882064819, "learning_rate": 1.842811547947193e-05, "loss": 0.0044, "step": 22591 }, { "epoch": 2.4492627927146575, "grad_norm": 0.3862023949623108, "learning_rate": 1.8424488611634993e-05, "loss": 0.0276, "step": 22592 }, { "epoch": 2.449371205550737, "grad_norm": 0.2564553916454315, "learning_rate": 1.8420861743798057e-05, "loss": 0.0087, "step": 22593 }, { "epoch": 2.449479618386817, "grad_norm": 0.3828364312648773, "learning_rate": 1.841723487596112e-05, "loss": 0.0499, "step": 22594 }, { "epoch": 2.4495880312228966, "grad_norm": 0.785597026348114, "learning_rate": 1.8413608008124182e-05, "loss": 0.0423, "step": 22595 }, { "epoch": 2.4496964440589766, "grad_norm": 1.516911506652832, "learning_rate": 1.840998114028725e-05, "loss": 0.0242, "step": 22596 }, { "epoch": 2.4498048568950566, "grad_norm": 0.1188722774386406, "learning_rate": 1.8406354272450314e-05, "loss": 0.0039, "step": 22597 }, { "epoch": 2.449913269731136, "grad_norm": 0.25709930062294006, "learning_rate": 1.8402727404613378e-05, "loss": 0.0136, "step": 22598 }, { "epoch": 2.450021682567216, "grad_norm": 0.013042761012911797, "learning_rate": 1.839910053677644e-05, "loss": 0.0006, "step": 22599 }, { "epoch": 2.4501300954032956, "grad_norm": 0.2558773458003998, "learning_rate": 1.8395473668939506e-05, "loss": 0.0148, "step": 22600 }, { "epoch": 2.4502385082393756, "grad_norm": 0.1600954830646515, "learning_rate": 1.839184680110257e-05, "loss": 0.014, "step": 22601 }, { "epoch": 2.450346921075455, "grad_norm": 0.14257393777370453, "learning_rate": 1.838821993326563e-05, "loss": 0.0087, "step": 22602 }, { "epoch": 2.450455333911535, "grad_norm": 0.44016537070274353, "learning_rate": 1.8384593065428695e-05, "loss": 0.0125, "step": 22603 }, { "epoch": 2.4505637467476147, "grad_norm": 0.19014765322208405, "learning_rate": 1.8380966197591763e-05, "loss": 0.0052, "step": 22604 }, { "epoch": 2.4506721595836947, "grad_norm": 0.0569966584444046, "learning_rate": 1.8377339329754827e-05, "loss": 0.0013, "step": 22605 }, { "epoch": 2.4507805724197746, "grad_norm": 0.09509970992803574, "learning_rate": 1.8373712461917887e-05, "loss": 0.0045, "step": 22606 }, { "epoch": 2.450888985255854, "grad_norm": 0.19836163520812988, "learning_rate": 1.837008559408095e-05, "loss": 0.0146, "step": 22607 }, { "epoch": 2.450997398091934, "grad_norm": 0.32737037539482117, "learning_rate": 1.8366458726244016e-05, "loss": 0.0094, "step": 22608 }, { "epoch": 2.4511058109280137, "grad_norm": 0.06675009429454803, "learning_rate": 1.836283185840708e-05, "loss": 0.0033, "step": 22609 }, { "epoch": 2.4512142237640937, "grad_norm": 0.2561923861503601, "learning_rate": 1.8359204990570144e-05, "loss": 0.0116, "step": 22610 }, { "epoch": 2.4513226366001737, "grad_norm": 0.10648063570261002, "learning_rate": 1.8355578122733208e-05, "loss": 0.0028, "step": 22611 }, { "epoch": 2.4514310494362532, "grad_norm": 0.02430950477719307, "learning_rate": 1.8351951254896272e-05, "loss": 0.0008, "step": 22612 }, { "epoch": 2.451539462272333, "grad_norm": 0.3120865821838379, "learning_rate": 1.8348324387059336e-05, "loss": 0.0193, "step": 22613 }, { "epoch": 2.4516478751084128, "grad_norm": 0.5205713510513306, "learning_rate": 1.83446975192224e-05, "loss": 0.0659, "step": 22614 }, { "epoch": 2.4517562879444927, "grad_norm": 0.2997397482395172, "learning_rate": 1.8341070651385464e-05, "loss": 0.0127, "step": 22615 }, { "epoch": 2.4518647007805723, "grad_norm": 0.14625242352485657, "learning_rate": 1.833744378354853e-05, "loss": 0.0053, "step": 22616 }, { "epoch": 2.4519731136166523, "grad_norm": 0.414804071187973, "learning_rate": 1.8333816915711593e-05, "loss": 0.0249, "step": 22617 }, { "epoch": 2.452081526452732, "grad_norm": 0.11743324995040894, "learning_rate": 1.8330190047874657e-05, "loss": 0.0027, "step": 22618 }, { "epoch": 2.452189939288812, "grad_norm": 0.4846581220626831, "learning_rate": 1.832656318003772e-05, "loss": 0.0418, "step": 22619 }, { "epoch": 2.452298352124892, "grad_norm": 0.3391304612159729, "learning_rate": 1.8322936312200785e-05, "loss": 0.0248, "step": 22620 }, { "epoch": 2.4524067649609713, "grad_norm": 0.6100304126739502, "learning_rate": 1.8319309444363846e-05, "loss": 0.0362, "step": 22621 }, { "epoch": 2.4525151777970513, "grad_norm": 0.05406125262379646, "learning_rate": 1.8315682576526913e-05, "loss": 0.002, "step": 22622 }, { "epoch": 2.452623590633131, "grad_norm": 0.0560615211725235, "learning_rate": 1.8312055708689977e-05, "loss": 0.0031, "step": 22623 }, { "epoch": 2.452732003469211, "grad_norm": 0.711448073387146, "learning_rate": 1.830842884085304e-05, "loss": 0.0434, "step": 22624 }, { "epoch": 2.4528404163052904, "grad_norm": 0.35768604278564453, "learning_rate": 1.8304801973016102e-05, "loss": 0.0269, "step": 22625 }, { "epoch": 2.4529488291413704, "grad_norm": 0.09129073470830917, "learning_rate": 1.830117510517917e-05, "loss": 0.002, "step": 22626 }, { "epoch": 2.45305724197745, "grad_norm": 0.9152354001998901, "learning_rate": 1.8297548237342234e-05, "loss": 0.0405, "step": 22627 }, { "epoch": 2.45316565481353, "grad_norm": 0.2007777839899063, "learning_rate": 1.8293921369505294e-05, "loss": 0.015, "step": 22628 }, { "epoch": 2.45327406764961, "grad_norm": 0.3083088994026184, "learning_rate": 1.829029450166836e-05, "loss": 0.0062, "step": 22629 }, { "epoch": 2.4533824804856894, "grad_norm": 0.05840444564819336, "learning_rate": 1.8286667633831426e-05, "loss": 0.003, "step": 22630 }, { "epoch": 2.4534908933217694, "grad_norm": 0.20532357692718506, "learning_rate": 1.828304076599449e-05, "loss": 0.008, "step": 22631 }, { "epoch": 2.453599306157849, "grad_norm": 0.11572214215993881, "learning_rate": 1.827941389815755e-05, "loss": 0.008, "step": 22632 }, { "epoch": 2.453707718993929, "grad_norm": 0.2736824154853821, "learning_rate": 1.8275787030320615e-05, "loss": 0.0134, "step": 22633 }, { "epoch": 2.453816131830009, "grad_norm": 0.41566720604896545, "learning_rate": 1.8272160162483682e-05, "loss": 0.0387, "step": 22634 }, { "epoch": 2.4539245446660884, "grad_norm": 0.4884481132030487, "learning_rate": 1.8268533294646743e-05, "loss": 0.0185, "step": 22635 }, { "epoch": 2.4540329575021684, "grad_norm": 0.07495149970054626, "learning_rate": 1.8264906426809807e-05, "loss": 0.0023, "step": 22636 }, { "epoch": 2.454141370338248, "grad_norm": 0.24427549540996552, "learning_rate": 1.826127955897287e-05, "loss": 0.0392, "step": 22637 }, { "epoch": 2.454249783174328, "grad_norm": 0.04021228104829788, "learning_rate": 1.8257652691135935e-05, "loss": 0.0015, "step": 22638 }, { "epoch": 2.4543581960104075, "grad_norm": 0.29897332191467285, "learning_rate": 1.8254025823299e-05, "loss": 0.0184, "step": 22639 }, { "epoch": 2.4544666088464875, "grad_norm": 0.3784756660461426, "learning_rate": 1.8250398955462064e-05, "loss": 0.0182, "step": 22640 }, { "epoch": 2.454575021682567, "grad_norm": 0.37926027178764343, "learning_rate": 1.8246772087625128e-05, "loss": 0.0268, "step": 22641 }, { "epoch": 2.454683434518647, "grad_norm": 0.6086623072624207, "learning_rate": 1.8243145219788192e-05, "loss": 0.0388, "step": 22642 }, { "epoch": 2.454791847354727, "grad_norm": 0.2529933750629425, "learning_rate": 1.8239518351951256e-05, "loss": 0.0075, "step": 22643 }, { "epoch": 2.4549002601908065, "grad_norm": 0.186903715133667, "learning_rate": 1.823589148411432e-05, "loss": 0.0105, "step": 22644 }, { "epoch": 2.4550086730268865, "grad_norm": 0.02750028297305107, "learning_rate": 1.8232264616277384e-05, "loss": 0.0011, "step": 22645 }, { "epoch": 2.455117085862966, "grad_norm": 0.04136986657977104, "learning_rate": 1.8228637748440448e-05, "loss": 0.002, "step": 22646 }, { "epoch": 2.455225498699046, "grad_norm": 0.6878626942634583, "learning_rate": 1.8225010880603512e-05, "loss": 0.0162, "step": 22647 }, { "epoch": 2.4553339115351256, "grad_norm": 0.12485439330339432, "learning_rate": 1.8221384012766576e-05, "loss": 0.0031, "step": 22648 }, { "epoch": 2.4554423243712056, "grad_norm": 0.09164302051067352, "learning_rate": 1.821775714492964e-05, "loss": 0.0033, "step": 22649 }, { "epoch": 2.455550737207285, "grad_norm": 0.3834995925426483, "learning_rate": 1.8214130277092705e-05, "loss": 0.0074, "step": 22650 }, { "epoch": 2.455659150043365, "grad_norm": 0.7035269737243652, "learning_rate": 1.8210503409255765e-05, "loss": 0.0422, "step": 22651 }, { "epoch": 2.455767562879445, "grad_norm": 0.023904964327812195, "learning_rate": 1.8206876541418833e-05, "loss": 0.0012, "step": 22652 }, { "epoch": 2.4558759757155246, "grad_norm": 0.30918121337890625, "learning_rate": 1.8203249673581897e-05, "loss": 0.0082, "step": 22653 }, { "epoch": 2.4559843885516046, "grad_norm": 0.240866020321846, "learning_rate": 1.8199622805744958e-05, "loss": 0.0287, "step": 22654 }, { "epoch": 2.456092801387684, "grad_norm": 0.19109748303890228, "learning_rate": 1.8195995937908022e-05, "loss": 0.0182, "step": 22655 }, { "epoch": 2.456201214223764, "grad_norm": 0.07203960418701172, "learning_rate": 1.819236907007109e-05, "loss": 0.0041, "step": 22656 }, { "epoch": 2.456309627059844, "grad_norm": 0.12841351330280304, "learning_rate": 1.8188742202234153e-05, "loss": 0.0043, "step": 22657 }, { "epoch": 2.4564180398959237, "grad_norm": 0.12607501447200775, "learning_rate": 1.8185115334397214e-05, "loss": 0.004, "step": 22658 }, { "epoch": 2.4565264527320037, "grad_norm": 0.4978419542312622, "learning_rate": 1.8181488466560278e-05, "loss": 0.0201, "step": 22659 }, { "epoch": 2.456634865568083, "grad_norm": 0.3094406723976135, "learning_rate": 1.8177861598723346e-05, "loss": 0.0166, "step": 22660 }, { "epoch": 2.456743278404163, "grad_norm": 0.48773622512817383, "learning_rate": 1.8174234730886406e-05, "loss": 0.0148, "step": 22661 }, { "epoch": 2.4568516912402427, "grad_norm": 0.4011152982711792, "learning_rate": 1.817060786304947e-05, "loss": 0.0307, "step": 22662 }, { "epoch": 2.4569601040763227, "grad_norm": 0.6075466871261597, "learning_rate": 1.8166980995212535e-05, "loss": 0.0496, "step": 22663 }, { "epoch": 2.4570685169124022, "grad_norm": 0.8514379262924194, "learning_rate": 1.81633541273756e-05, "loss": 0.0531, "step": 22664 }, { "epoch": 2.4571769297484822, "grad_norm": 0.4539283812046051, "learning_rate": 1.8159727259538663e-05, "loss": 0.0292, "step": 22665 }, { "epoch": 2.457285342584562, "grad_norm": 0.18163800239562988, "learning_rate": 1.8156100391701727e-05, "loss": 0.0075, "step": 22666 }, { "epoch": 2.4573937554206418, "grad_norm": 0.42405828833580017, "learning_rate": 1.815247352386479e-05, "loss": 0.0146, "step": 22667 }, { "epoch": 2.4575021682567217, "grad_norm": 0.13168549537658691, "learning_rate": 1.8148846656027855e-05, "loss": 0.0043, "step": 22668 }, { "epoch": 2.4576105810928013, "grad_norm": 0.18025025725364685, "learning_rate": 1.814521978819092e-05, "loss": 0.0089, "step": 22669 }, { "epoch": 2.4577189939288813, "grad_norm": 0.021348008885979652, "learning_rate": 1.8141592920353983e-05, "loss": 0.0015, "step": 22670 }, { "epoch": 2.457827406764961, "grad_norm": 0.24600820243358612, "learning_rate": 1.8137966052517047e-05, "loss": 0.0118, "step": 22671 }, { "epoch": 2.457935819601041, "grad_norm": 0.28854694962501526, "learning_rate": 1.813433918468011e-05, "loss": 0.0531, "step": 22672 }, { "epoch": 2.4580442324371203, "grad_norm": 1.2684993743896484, "learning_rate": 1.8130712316843176e-05, "loss": 0.0204, "step": 22673 }, { "epoch": 2.4581526452732003, "grad_norm": 0.10882370918989182, "learning_rate": 1.812708544900624e-05, "loss": 0.004, "step": 22674 }, { "epoch": 2.4582610581092803, "grad_norm": 0.2640223503112793, "learning_rate": 1.8123458581169304e-05, "loss": 0.028, "step": 22675 }, { "epoch": 2.45836947094536, "grad_norm": 0.0933675691485405, "learning_rate": 1.8119831713332368e-05, "loss": 0.0043, "step": 22676 }, { "epoch": 2.45847788378144, "grad_norm": 0.4320003390312195, "learning_rate": 1.811620484549543e-05, "loss": 0.0101, "step": 22677 }, { "epoch": 2.4585862966175194, "grad_norm": 0.15299856662750244, "learning_rate": 1.8112577977658496e-05, "loss": 0.0067, "step": 22678 }, { "epoch": 2.4586947094535994, "grad_norm": 0.6536733508110046, "learning_rate": 1.810895110982156e-05, "loss": 0.0275, "step": 22679 }, { "epoch": 2.4588031222896793, "grad_norm": 0.24917741119861603, "learning_rate": 1.810532424198462e-05, "loss": 0.0391, "step": 22680 }, { "epoch": 2.458911535125759, "grad_norm": 0.3548935353755951, "learning_rate": 1.8101697374147685e-05, "loss": 0.0099, "step": 22681 }, { "epoch": 2.459019947961839, "grad_norm": 0.14273490011692047, "learning_rate": 1.8098070506310753e-05, "loss": 0.0061, "step": 22682 }, { "epoch": 2.4591283607979184, "grad_norm": 0.32906848192214966, "learning_rate": 1.8094443638473817e-05, "loss": 0.0225, "step": 22683 }, { "epoch": 2.4592367736339984, "grad_norm": 0.12706252932548523, "learning_rate": 1.8090816770636877e-05, "loss": 0.0058, "step": 22684 }, { "epoch": 2.459345186470078, "grad_norm": 1.0446950197219849, "learning_rate": 1.808718990279994e-05, "loss": 0.0159, "step": 22685 }, { "epoch": 2.459453599306158, "grad_norm": 0.15697279572486877, "learning_rate": 1.808356303496301e-05, "loss": 0.0067, "step": 22686 }, { "epoch": 2.4595620121422375, "grad_norm": 0.46633753180503845, "learning_rate": 1.807993616712607e-05, "loss": 0.0313, "step": 22687 }, { "epoch": 2.4596704249783174, "grad_norm": 0.18442358076572418, "learning_rate": 1.8076309299289134e-05, "loss": 0.0122, "step": 22688 }, { "epoch": 2.4597788378143974, "grad_norm": 0.2839449942111969, "learning_rate": 1.8072682431452198e-05, "loss": 0.011, "step": 22689 }, { "epoch": 2.459887250650477, "grad_norm": 0.20606204867362976, "learning_rate": 1.8069055563615262e-05, "loss": 0.0038, "step": 22690 }, { "epoch": 2.459995663486557, "grad_norm": 0.06163540109992027, "learning_rate": 1.8065428695778326e-05, "loss": 0.0023, "step": 22691 }, { "epoch": 2.4601040763226365, "grad_norm": 0.24620544910430908, "learning_rate": 1.806180182794139e-05, "loss": 0.0099, "step": 22692 }, { "epoch": 2.4602124891587165, "grad_norm": 0.2904827892780304, "learning_rate": 1.8058174960104454e-05, "loss": 0.0244, "step": 22693 }, { "epoch": 2.460320901994796, "grad_norm": 0.3374767303466797, "learning_rate": 1.805454809226752e-05, "loss": 0.0153, "step": 22694 }, { "epoch": 2.460429314830876, "grad_norm": 0.17888975143432617, "learning_rate": 1.8050921224430583e-05, "loss": 0.0167, "step": 22695 }, { "epoch": 2.4605377276669556, "grad_norm": 0.3038838803768158, "learning_rate": 1.8047294356593647e-05, "loss": 0.0159, "step": 22696 }, { "epoch": 2.4606461405030355, "grad_norm": 0.060701433569192886, "learning_rate": 1.804366748875671e-05, "loss": 0.0022, "step": 22697 }, { "epoch": 2.4607545533391155, "grad_norm": 0.0023416196927428246, "learning_rate": 1.8040040620919775e-05, "loss": 0.0002, "step": 22698 }, { "epoch": 2.460862966175195, "grad_norm": 0.04983219504356384, "learning_rate": 1.803641375308284e-05, "loss": 0.0019, "step": 22699 }, { "epoch": 2.460971379011275, "grad_norm": 0.16368605196475983, "learning_rate": 1.8032786885245903e-05, "loss": 0.0044, "step": 22700 }, { "epoch": 2.4610797918473546, "grad_norm": 0.49457696080207825, "learning_rate": 1.8029160017408967e-05, "loss": 0.0388, "step": 22701 }, { "epoch": 2.4611882046834346, "grad_norm": 0.1627136468887329, "learning_rate": 1.802553314957203e-05, "loss": 0.0098, "step": 22702 }, { "epoch": 2.461296617519514, "grad_norm": 0.20893564820289612, "learning_rate": 1.8021906281735092e-05, "loss": 0.0296, "step": 22703 }, { "epoch": 2.461405030355594, "grad_norm": 0.09347951412200928, "learning_rate": 1.801827941389816e-05, "loss": 0.0032, "step": 22704 }, { "epoch": 2.4615134431916736, "grad_norm": 0.10265252739191055, "learning_rate": 1.8014652546061224e-05, "loss": 0.0025, "step": 22705 }, { "epoch": 2.4616218560277536, "grad_norm": 0.01870592124760151, "learning_rate": 1.8011025678224284e-05, "loss": 0.0009, "step": 22706 }, { "epoch": 2.4617302688638336, "grad_norm": 0.08224500715732574, "learning_rate": 1.800739881038735e-05, "loss": 0.0026, "step": 22707 }, { "epoch": 2.461838681699913, "grad_norm": 0.15267205238342285, "learning_rate": 1.8003771942550416e-05, "loss": 0.0079, "step": 22708 }, { "epoch": 2.461947094535993, "grad_norm": 0.16288529336452484, "learning_rate": 1.800014507471348e-05, "loss": 0.0131, "step": 22709 }, { "epoch": 2.4620555073720727, "grad_norm": 0.04592116177082062, "learning_rate": 1.799651820687654e-05, "loss": 0.0022, "step": 22710 }, { "epoch": 2.4621639202081527, "grad_norm": 0.1912573277950287, "learning_rate": 1.7992891339039605e-05, "loss": 0.0086, "step": 22711 }, { "epoch": 2.4622723330442327, "grad_norm": 0.08117291331291199, "learning_rate": 1.7989264471202672e-05, "loss": 0.0029, "step": 22712 }, { "epoch": 2.462380745880312, "grad_norm": 0.4729783833026886, "learning_rate": 1.7985637603365733e-05, "loss": 0.0398, "step": 22713 }, { "epoch": 2.462489158716392, "grad_norm": 0.06423906981945038, "learning_rate": 1.7982010735528797e-05, "loss": 0.0023, "step": 22714 }, { "epoch": 2.4625975715524717, "grad_norm": 0.3025718927383423, "learning_rate": 1.797838386769186e-05, "loss": 0.0078, "step": 22715 }, { "epoch": 2.4627059843885517, "grad_norm": 0.07669378817081451, "learning_rate": 1.7974756999854925e-05, "loss": 0.0073, "step": 22716 }, { "epoch": 2.4628143972246312, "grad_norm": 0.5089635848999023, "learning_rate": 1.797113013201799e-05, "loss": 0.0148, "step": 22717 }, { "epoch": 2.4629228100607112, "grad_norm": 0.053078655153512955, "learning_rate": 1.7967503264181054e-05, "loss": 0.0042, "step": 22718 }, { "epoch": 2.4630312228967908, "grad_norm": 0.2890462279319763, "learning_rate": 1.7963876396344118e-05, "loss": 0.02, "step": 22719 }, { "epoch": 2.4631396357328708, "grad_norm": 0.6746135354042053, "learning_rate": 1.7960249528507182e-05, "loss": 0.0221, "step": 22720 }, { "epoch": 2.4632480485689507, "grad_norm": 0.3261265158653259, "learning_rate": 1.7956622660670246e-05, "loss": 0.0129, "step": 22721 }, { "epoch": 2.4633564614050303, "grad_norm": 0.13484878838062286, "learning_rate": 1.795299579283331e-05, "loss": 0.0029, "step": 22722 }, { "epoch": 2.4634648742411103, "grad_norm": 0.376932293176651, "learning_rate": 1.7949368924996374e-05, "loss": 0.0296, "step": 22723 }, { "epoch": 2.46357328707719, "grad_norm": 0.050240907818078995, "learning_rate": 1.7945742057159438e-05, "loss": 0.0031, "step": 22724 }, { "epoch": 2.46368169991327, "grad_norm": 0.16211983561515808, "learning_rate": 1.7942115189322502e-05, "loss": 0.0071, "step": 22725 }, { "epoch": 2.4637901127493493, "grad_norm": 0.05401553958654404, "learning_rate": 1.7938488321485566e-05, "loss": 0.0019, "step": 22726 }, { "epoch": 2.4638985255854293, "grad_norm": 0.14917409420013428, "learning_rate": 1.793486145364863e-05, "loss": 0.007, "step": 22727 }, { "epoch": 2.464006938421509, "grad_norm": 0.0303595382720232, "learning_rate": 1.7931234585811695e-05, "loss": 0.0009, "step": 22728 }, { "epoch": 2.464115351257589, "grad_norm": 0.052950698882341385, "learning_rate": 1.792760771797476e-05, "loss": 0.0032, "step": 22729 }, { "epoch": 2.464223764093669, "grad_norm": 0.164017915725708, "learning_rate": 1.7923980850137823e-05, "loss": 0.006, "step": 22730 }, { "epoch": 2.4643321769297484, "grad_norm": 0.056801751255989075, "learning_rate": 1.7920353982300887e-05, "loss": 0.0037, "step": 22731 }, { "epoch": 2.4644405897658284, "grad_norm": 0.12113955616950989, "learning_rate": 1.7916727114463948e-05, "loss": 0.0074, "step": 22732 }, { "epoch": 2.464549002601908, "grad_norm": 0.2235167771577835, "learning_rate": 1.7913100246627012e-05, "loss": 0.0051, "step": 22733 }, { "epoch": 2.464657415437988, "grad_norm": 0.08962727338075638, "learning_rate": 1.790947337879008e-05, "loss": 0.0071, "step": 22734 }, { "epoch": 2.464765828274068, "grad_norm": 0.07521592825651169, "learning_rate": 1.7905846510953143e-05, "loss": 0.004, "step": 22735 }, { "epoch": 2.4648742411101474, "grad_norm": 0.060321759432554245, "learning_rate": 1.7902219643116204e-05, "loss": 0.0029, "step": 22736 }, { "epoch": 2.4649826539462274, "grad_norm": 0.2507387697696686, "learning_rate": 1.7898592775279268e-05, "loss": 0.0553, "step": 22737 }, { "epoch": 2.465091066782307, "grad_norm": 0.0699893981218338, "learning_rate": 1.7894965907442336e-05, "loss": 0.0032, "step": 22738 }, { "epoch": 2.465199479618387, "grad_norm": 0.37379586696624756, "learning_rate": 1.78913390396054e-05, "loss": 0.0064, "step": 22739 }, { "epoch": 2.4653078924544665, "grad_norm": 0.027760297060012817, "learning_rate": 1.788771217176846e-05, "loss": 0.0017, "step": 22740 }, { "epoch": 2.4654163052905465, "grad_norm": 0.16955088078975677, "learning_rate": 1.7884085303931525e-05, "loss": 0.0078, "step": 22741 }, { "epoch": 2.465524718126626, "grad_norm": 0.21548713743686676, "learning_rate": 1.7880458436094592e-05, "loss": 0.0156, "step": 22742 }, { "epoch": 2.465633130962706, "grad_norm": 0.07751300185918808, "learning_rate": 1.7876831568257653e-05, "loss": 0.0036, "step": 22743 }, { "epoch": 2.465741543798786, "grad_norm": 0.07509622722864151, "learning_rate": 1.7873204700420717e-05, "loss": 0.0031, "step": 22744 }, { "epoch": 2.4658499566348655, "grad_norm": 0.02105109766125679, "learning_rate": 1.786957783258378e-05, "loss": 0.001, "step": 22745 }, { "epoch": 2.4659583694709455, "grad_norm": 0.11927168816328049, "learning_rate": 1.7865950964746845e-05, "loss": 0.0097, "step": 22746 }, { "epoch": 2.466066782307025, "grad_norm": 0.23615027964115143, "learning_rate": 1.786232409690991e-05, "loss": 0.0153, "step": 22747 }, { "epoch": 2.466175195143105, "grad_norm": 0.0849103033542633, "learning_rate": 1.7858697229072973e-05, "loss": 0.003, "step": 22748 }, { "epoch": 2.4662836079791846, "grad_norm": 0.1361808329820633, "learning_rate": 1.7855070361236037e-05, "loss": 0.0075, "step": 22749 }, { "epoch": 2.4663920208152645, "grad_norm": 0.5224087834358215, "learning_rate": 1.78514434933991e-05, "loss": 0.0158, "step": 22750 }, { "epoch": 2.466500433651344, "grad_norm": 0.16332510113716125, "learning_rate": 1.7847816625562166e-05, "loss": 0.0066, "step": 22751 }, { "epoch": 2.466608846487424, "grad_norm": 0.1618436872959137, "learning_rate": 1.784418975772523e-05, "loss": 0.0068, "step": 22752 }, { "epoch": 2.466717259323504, "grad_norm": 0.08431743830442429, "learning_rate": 1.7840562889888294e-05, "loss": 0.0054, "step": 22753 }, { "epoch": 2.4668256721595836, "grad_norm": 0.28481608629226685, "learning_rate": 1.7836936022051358e-05, "loss": 0.0434, "step": 22754 }, { "epoch": 2.4669340849956636, "grad_norm": 0.21039347350597382, "learning_rate": 1.7833309154214422e-05, "loss": 0.006, "step": 22755 }, { "epoch": 2.467042497831743, "grad_norm": 0.16721022129058838, "learning_rate": 1.7829682286377486e-05, "loss": 0.0088, "step": 22756 }, { "epoch": 2.467150910667823, "grad_norm": 0.2571236193180084, "learning_rate": 1.782605541854055e-05, "loss": 0.0093, "step": 22757 }, { "epoch": 2.467259323503903, "grad_norm": 0.11395612359046936, "learning_rate": 1.782242855070361e-05, "loss": 0.0018, "step": 22758 }, { "epoch": 2.4673677363399826, "grad_norm": 0.13062933087348938, "learning_rate": 1.7818801682866675e-05, "loss": 0.0105, "step": 22759 }, { "epoch": 2.4674761491760626, "grad_norm": 0.45015236735343933, "learning_rate": 1.7815174815029743e-05, "loss": 0.018, "step": 22760 }, { "epoch": 2.467584562012142, "grad_norm": 0.027188031002879143, "learning_rate": 1.7811547947192807e-05, "loss": 0.0009, "step": 22761 }, { "epoch": 2.467692974848222, "grad_norm": 0.21135756373405457, "learning_rate": 1.7807921079355867e-05, "loss": 0.0089, "step": 22762 }, { "epoch": 2.4678013876843017, "grad_norm": 0.25081944465637207, "learning_rate": 1.780429421151893e-05, "loss": 0.002, "step": 22763 }, { "epoch": 2.4679098005203817, "grad_norm": 0.31636565923690796, "learning_rate": 1.7800667343682e-05, "loss": 0.0057, "step": 22764 }, { "epoch": 2.468018213356461, "grad_norm": 0.09331128746271133, "learning_rate": 1.7797040475845063e-05, "loss": 0.004, "step": 22765 }, { "epoch": 2.468126626192541, "grad_norm": 0.3661139905452728, "learning_rate": 1.7793413608008124e-05, "loss": 0.0385, "step": 22766 }, { "epoch": 2.468235039028621, "grad_norm": 0.33622536063194275, "learning_rate": 1.7789786740171188e-05, "loss": 0.0214, "step": 22767 }, { "epoch": 2.4683434518647007, "grad_norm": 0.4262198507785797, "learning_rate": 1.7786159872334255e-05, "loss": 0.0193, "step": 22768 }, { "epoch": 2.4684518647007807, "grad_norm": 0.20657584071159363, "learning_rate": 1.7782533004497316e-05, "loss": 0.0123, "step": 22769 }, { "epoch": 2.4685602775368602, "grad_norm": 0.5364599227905273, "learning_rate": 1.777890613666038e-05, "loss": 0.0089, "step": 22770 }, { "epoch": 2.4686686903729402, "grad_norm": 0.1944526582956314, "learning_rate": 1.7775279268823444e-05, "loss": 0.0033, "step": 22771 }, { "epoch": 2.4687771032090198, "grad_norm": 0.7529550790786743, "learning_rate": 1.777165240098651e-05, "loss": 0.0204, "step": 22772 }, { "epoch": 2.4688855160450998, "grad_norm": 0.9261488914489746, "learning_rate": 1.7768025533149573e-05, "loss": 0.0229, "step": 22773 }, { "epoch": 2.4689939288811793, "grad_norm": 0.20720897614955902, "learning_rate": 1.7764398665312637e-05, "loss": 0.0055, "step": 22774 }, { "epoch": 2.4691023417172593, "grad_norm": 0.15972746908664703, "learning_rate": 1.77607717974757e-05, "loss": 0.0039, "step": 22775 }, { "epoch": 2.4692107545533393, "grad_norm": 0.6415539383888245, "learning_rate": 1.7757144929638765e-05, "loss": 0.0737, "step": 22776 }, { "epoch": 2.469319167389419, "grad_norm": 0.28526800870895386, "learning_rate": 1.775351806180183e-05, "loss": 0.0342, "step": 22777 }, { "epoch": 2.469427580225499, "grad_norm": 0.3835124671459198, "learning_rate": 1.7749891193964893e-05, "loss": 0.0129, "step": 22778 }, { "epoch": 2.4695359930615783, "grad_norm": 0.23499757051467896, "learning_rate": 1.7746264326127957e-05, "loss": 0.0056, "step": 22779 }, { "epoch": 2.4696444058976583, "grad_norm": 0.1640121042728424, "learning_rate": 1.774263745829102e-05, "loss": 0.0066, "step": 22780 }, { "epoch": 2.4697528187337383, "grad_norm": 0.012382245622575283, "learning_rate": 1.7739010590454085e-05, "loss": 0.0003, "step": 22781 }, { "epoch": 2.469861231569818, "grad_norm": 0.3275541365146637, "learning_rate": 1.773538372261715e-05, "loss": 0.0095, "step": 22782 }, { "epoch": 2.469969644405898, "grad_norm": 0.30155259370803833, "learning_rate": 1.7731756854780214e-05, "loss": 0.0093, "step": 22783 }, { "epoch": 2.4700780572419774, "grad_norm": 0.07070165872573853, "learning_rate": 1.7728129986943274e-05, "loss": 0.0075, "step": 22784 }, { "epoch": 2.4701864700780574, "grad_norm": 0.45168137550354004, "learning_rate": 1.772450311910634e-05, "loss": 0.0205, "step": 22785 }, { "epoch": 2.470294882914137, "grad_norm": 0.361773282289505, "learning_rate": 1.7720876251269406e-05, "loss": 0.0195, "step": 22786 }, { "epoch": 2.470403295750217, "grad_norm": 0.0408918634057045, "learning_rate": 1.771724938343247e-05, "loss": 0.0024, "step": 22787 }, { "epoch": 2.4705117085862964, "grad_norm": 0.1001177504658699, "learning_rate": 1.771362251559553e-05, "loss": 0.0038, "step": 22788 }, { "epoch": 2.4706201214223764, "grad_norm": 0.20501981675624847, "learning_rate": 1.7709995647758595e-05, "loss": 0.0165, "step": 22789 }, { "epoch": 2.4707285342584564, "grad_norm": 1.438302993774414, "learning_rate": 1.7706368779921662e-05, "loss": 0.02, "step": 22790 }, { "epoch": 2.470836947094536, "grad_norm": 0.26583266258239746, "learning_rate": 1.7702741912084727e-05, "loss": 0.0066, "step": 22791 }, { "epoch": 2.470945359930616, "grad_norm": 0.02455035224556923, "learning_rate": 1.7699115044247787e-05, "loss": 0.0007, "step": 22792 }, { "epoch": 2.4710537727666955, "grad_norm": 0.08691895753145218, "learning_rate": 1.769548817641085e-05, "loss": 0.0072, "step": 22793 }, { "epoch": 2.4711621856027755, "grad_norm": 0.20225191116333008, "learning_rate": 1.769186130857392e-05, "loss": 0.0092, "step": 22794 }, { "epoch": 2.471270598438855, "grad_norm": 0.11862362176179886, "learning_rate": 1.768823444073698e-05, "loss": 0.0064, "step": 22795 }, { "epoch": 2.471379011274935, "grad_norm": 0.007027067709714174, "learning_rate": 1.7684607572900044e-05, "loss": 0.0003, "step": 22796 }, { "epoch": 2.4714874241110145, "grad_norm": 0.18016977608203888, "learning_rate": 1.7680980705063108e-05, "loss": 0.0034, "step": 22797 }, { "epoch": 2.4715958369470945, "grad_norm": 0.17587438225746155, "learning_rate": 1.7677353837226172e-05, "loss": 0.006, "step": 22798 }, { "epoch": 2.4717042497831745, "grad_norm": 0.40685391426086426, "learning_rate": 1.7673726969389236e-05, "loss": 0.0258, "step": 22799 }, { "epoch": 2.471812662619254, "grad_norm": 0.20004130899906158, "learning_rate": 1.76701001015523e-05, "loss": 0.0032, "step": 22800 }, { "epoch": 2.471921075455334, "grad_norm": 0.037266574800014496, "learning_rate": 1.7666473233715364e-05, "loss": 0.0014, "step": 22801 }, { "epoch": 2.4720294882914136, "grad_norm": 0.21092188358306885, "learning_rate": 1.7662846365878428e-05, "loss": 0.0058, "step": 22802 }, { "epoch": 2.4721379011274935, "grad_norm": 0.3598325252532959, "learning_rate": 1.7659219498041492e-05, "loss": 0.0444, "step": 22803 }, { "epoch": 2.4722463139635735, "grad_norm": 0.11879956722259521, "learning_rate": 1.7655592630204556e-05, "loss": 0.0014, "step": 22804 }, { "epoch": 2.472354726799653, "grad_norm": 0.013101222924888134, "learning_rate": 1.765196576236762e-05, "loss": 0.0005, "step": 22805 }, { "epoch": 2.472463139635733, "grad_norm": 0.5099951028823853, "learning_rate": 1.7648338894530685e-05, "loss": 0.0383, "step": 22806 }, { "epoch": 2.4725715524718126, "grad_norm": 0.08061216026544571, "learning_rate": 1.764471202669375e-05, "loss": 0.0038, "step": 22807 }, { "epoch": 2.4726799653078926, "grad_norm": 0.2331630438566208, "learning_rate": 1.7641085158856813e-05, "loss": 0.0267, "step": 22808 }, { "epoch": 2.472788378143972, "grad_norm": 0.28601348400115967, "learning_rate": 1.7637458291019877e-05, "loss": 0.0113, "step": 22809 }, { "epoch": 2.472896790980052, "grad_norm": 0.04617460072040558, "learning_rate": 1.7633831423182938e-05, "loss": 0.0022, "step": 22810 }, { "epoch": 2.4730052038161316, "grad_norm": 0.32200631499290466, "learning_rate": 1.7630204555346002e-05, "loss": 0.0168, "step": 22811 }, { "epoch": 2.4731136166522116, "grad_norm": 0.004977768287062645, "learning_rate": 1.762657768750907e-05, "loss": 0.0003, "step": 22812 }, { "epoch": 2.4732220294882916, "grad_norm": 0.04171619564294815, "learning_rate": 1.7622950819672133e-05, "loss": 0.0022, "step": 22813 }, { "epoch": 2.473330442324371, "grad_norm": 0.06143800914287567, "learning_rate": 1.7619323951835194e-05, "loss": 0.0028, "step": 22814 }, { "epoch": 2.473438855160451, "grad_norm": 0.025912484154105186, "learning_rate": 1.7615697083998258e-05, "loss": 0.0006, "step": 22815 }, { "epoch": 2.4735472679965307, "grad_norm": 0.11763051152229309, "learning_rate": 1.7612070216161326e-05, "loss": 0.0041, "step": 22816 }, { "epoch": 2.4736556808326107, "grad_norm": 0.09466946125030518, "learning_rate": 1.760844334832439e-05, "loss": 0.0075, "step": 22817 }, { "epoch": 2.47376409366869, "grad_norm": 0.23009884357452393, "learning_rate": 1.760481648048745e-05, "loss": 0.0013, "step": 22818 }, { "epoch": 2.47387250650477, "grad_norm": 0.08009390532970428, "learning_rate": 1.7601189612650515e-05, "loss": 0.0041, "step": 22819 }, { "epoch": 2.4739809193408497, "grad_norm": 0.5004438757896423, "learning_rate": 1.7597562744813582e-05, "loss": 0.0192, "step": 22820 }, { "epoch": 2.4740893321769297, "grad_norm": 0.19733896851539612, "learning_rate": 1.7593935876976643e-05, "loss": 0.0083, "step": 22821 }, { "epoch": 2.4741977450130097, "grad_norm": 1.0665874481201172, "learning_rate": 1.7590309009139707e-05, "loss": 0.0332, "step": 22822 }, { "epoch": 2.4743061578490892, "grad_norm": 0.21096695959568024, "learning_rate": 1.758668214130277e-05, "loss": 0.0111, "step": 22823 }, { "epoch": 2.4744145706851692, "grad_norm": 0.5264006853103638, "learning_rate": 1.7583055273465835e-05, "loss": 0.0511, "step": 22824 }, { "epoch": 2.4745229835212488, "grad_norm": 0.1573418378829956, "learning_rate": 1.75794284056289e-05, "loss": 0.0208, "step": 22825 }, { "epoch": 2.4746313963573288, "grad_norm": 0.7779071927070618, "learning_rate": 1.7575801537791963e-05, "loss": 0.0225, "step": 22826 }, { "epoch": 2.4747398091934087, "grad_norm": 0.005619532894343138, "learning_rate": 1.7572174669955028e-05, "loss": 0.0002, "step": 22827 }, { "epoch": 2.4748482220294883, "grad_norm": 0.05298645794391632, "learning_rate": 1.756854780211809e-05, "loss": 0.0023, "step": 22828 }, { "epoch": 2.4749566348655683, "grad_norm": 0.02094622142612934, "learning_rate": 1.7564920934281156e-05, "loss": 0.0009, "step": 22829 }, { "epoch": 2.475065047701648, "grad_norm": 0.08666977286338806, "learning_rate": 1.756129406644422e-05, "loss": 0.0032, "step": 22830 }, { "epoch": 2.475173460537728, "grad_norm": 0.05891605466604233, "learning_rate": 1.7557667198607284e-05, "loss": 0.0025, "step": 22831 }, { "epoch": 2.4752818733738073, "grad_norm": 0.048769768327474594, "learning_rate": 1.7554040330770348e-05, "loss": 0.0027, "step": 22832 }, { "epoch": 2.4753902862098873, "grad_norm": 0.014527015388011932, "learning_rate": 1.7550413462933412e-05, "loss": 0.0007, "step": 22833 }, { "epoch": 2.475498699045967, "grad_norm": 0.17237108945846558, "learning_rate": 1.7546786595096476e-05, "loss": 0.0049, "step": 22834 }, { "epoch": 2.475607111882047, "grad_norm": 0.06833886355161667, "learning_rate": 1.754315972725954e-05, "loss": 0.0042, "step": 22835 }, { "epoch": 2.475715524718127, "grad_norm": 0.5246057510375977, "learning_rate": 1.75395328594226e-05, "loss": 0.0464, "step": 22836 }, { "epoch": 2.4758239375542064, "grad_norm": 0.01748717576265335, "learning_rate": 1.753590599158567e-05, "loss": 0.0002, "step": 22837 }, { "epoch": 2.4759323503902864, "grad_norm": 0.12218271940946579, "learning_rate": 1.7532279123748733e-05, "loss": 0.0031, "step": 22838 }, { "epoch": 2.476040763226366, "grad_norm": 0.18384651839733124, "learning_rate": 1.7528652255911797e-05, "loss": 0.0122, "step": 22839 }, { "epoch": 2.476149176062446, "grad_norm": 0.17351363599300385, "learning_rate": 1.7525025388074857e-05, "loss": 0.0087, "step": 22840 }, { "epoch": 2.4762575888985254, "grad_norm": 0.14039796590805054, "learning_rate": 1.752139852023792e-05, "loss": 0.0037, "step": 22841 }, { "epoch": 2.4763660017346054, "grad_norm": 0.5149449706077576, "learning_rate": 1.751777165240099e-05, "loss": 0.0397, "step": 22842 }, { "epoch": 2.476474414570685, "grad_norm": 0.05872190743684769, "learning_rate": 1.7514144784564053e-05, "loss": 0.0019, "step": 22843 }, { "epoch": 2.476582827406765, "grad_norm": 0.14733299612998962, "learning_rate": 1.7510517916727114e-05, "loss": 0.0047, "step": 22844 }, { "epoch": 2.476691240242845, "grad_norm": 0.600688636302948, "learning_rate": 1.7506891048890178e-05, "loss": 0.0076, "step": 22845 }, { "epoch": 2.4767996530789245, "grad_norm": 0.010618382133543491, "learning_rate": 1.7503264181053246e-05, "loss": 0.0005, "step": 22846 }, { "epoch": 2.4769080659150045, "grad_norm": 0.7270671725273132, "learning_rate": 1.7499637313216306e-05, "loss": 0.0289, "step": 22847 }, { "epoch": 2.477016478751084, "grad_norm": 0.03508497029542923, "learning_rate": 1.749601044537937e-05, "loss": 0.0018, "step": 22848 }, { "epoch": 2.477124891587164, "grad_norm": 0.0284888818860054, "learning_rate": 1.7492383577542434e-05, "loss": 0.001, "step": 22849 }, { "epoch": 2.4772333044232435, "grad_norm": 0.10678502917289734, "learning_rate": 1.7488756709705502e-05, "loss": 0.0091, "step": 22850 }, { "epoch": 2.4773417172593235, "grad_norm": 0.22911414504051208, "learning_rate": 1.7485129841868563e-05, "loss": 0.01, "step": 22851 }, { "epoch": 2.477450130095403, "grad_norm": 0.005662217736244202, "learning_rate": 1.7481502974031627e-05, "loss": 0.0003, "step": 22852 }, { "epoch": 2.477558542931483, "grad_norm": 0.050391845405101776, "learning_rate": 1.747787610619469e-05, "loss": 0.0014, "step": 22853 }, { "epoch": 2.477666955767563, "grad_norm": 0.23625901341438293, "learning_rate": 1.7474249238357755e-05, "loss": 0.0072, "step": 22854 }, { "epoch": 2.4777753686036426, "grad_norm": 0.15512403845787048, "learning_rate": 1.747062237052082e-05, "loss": 0.0196, "step": 22855 }, { "epoch": 2.4778837814397225, "grad_norm": 0.4946348965167999, "learning_rate": 1.7466995502683883e-05, "loss": 0.0355, "step": 22856 }, { "epoch": 2.477992194275802, "grad_norm": 0.5010631084442139, "learning_rate": 1.7463368634846947e-05, "loss": 0.0106, "step": 22857 }, { "epoch": 2.478100607111882, "grad_norm": 0.26937949657440186, "learning_rate": 1.745974176701001e-05, "loss": 0.0077, "step": 22858 }, { "epoch": 2.478209019947962, "grad_norm": 0.0286610946059227, "learning_rate": 1.7456114899173075e-05, "loss": 0.0004, "step": 22859 }, { "epoch": 2.4783174327840416, "grad_norm": 0.5291430354118347, "learning_rate": 1.745248803133614e-05, "loss": 0.0086, "step": 22860 }, { "epoch": 2.4784258456201216, "grad_norm": 0.6095461845397949, "learning_rate": 1.7448861163499204e-05, "loss": 0.0367, "step": 22861 }, { "epoch": 2.478534258456201, "grad_norm": 0.31705498695373535, "learning_rate": 1.7445234295662264e-05, "loss": 0.0521, "step": 22862 }, { "epoch": 2.478642671292281, "grad_norm": 0.3100045919418335, "learning_rate": 1.7441607427825332e-05, "loss": 0.0172, "step": 22863 }, { "epoch": 2.4787510841283606, "grad_norm": 0.2867496609687805, "learning_rate": 1.7437980559988396e-05, "loss": 0.0163, "step": 22864 }, { "epoch": 2.4788594969644406, "grad_norm": 0.16319642961025238, "learning_rate": 1.743435369215146e-05, "loss": 0.0046, "step": 22865 }, { "epoch": 2.47896790980052, "grad_norm": 0.2107553780078888, "learning_rate": 1.743072682431452e-05, "loss": 0.0119, "step": 22866 }, { "epoch": 2.4790763226366, "grad_norm": 0.4038124084472656, "learning_rate": 1.7427099956477585e-05, "loss": 0.0142, "step": 22867 }, { "epoch": 2.47918473547268, "grad_norm": 0.14673855900764465, "learning_rate": 1.7423473088640652e-05, "loss": 0.0059, "step": 22868 }, { "epoch": 2.4792931483087597, "grad_norm": 0.30127862095832825, "learning_rate": 1.7419846220803717e-05, "loss": 0.009, "step": 22869 }, { "epoch": 2.4794015611448397, "grad_norm": 1.4149786233901978, "learning_rate": 1.7416219352966777e-05, "loss": 0.0182, "step": 22870 }, { "epoch": 2.479509973980919, "grad_norm": 0.4864886403083801, "learning_rate": 1.741259248512984e-05, "loss": 0.0195, "step": 22871 }, { "epoch": 2.479618386816999, "grad_norm": 0.29114702343940735, "learning_rate": 1.740896561729291e-05, "loss": 0.0208, "step": 22872 }, { "epoch": 2.4797267996530787, "grad_norm": 0.41348540782928467, "learning_rate": 1.740533874945597e-05, "loss": 0.0323, "step": 22873 }, { "epoch": 2.4798352124891587, "grad_norm": 0.1288217157125473, "learning_rate": 1.7401711881619034e-05, "loss": 0.0054, "step": 22874 }, { "epoch": 2.4799436253252383, "grad_norm": 0.9052203893661499, "learning_rate": 1.7398085013782098e-05, "loss": 0.0286, "step": 22875 }, { "epoch": 2.4800520381613183, "grad_norm": 0.6908016204833984, "learning_rate": 1.7394458145945165e-05, "loss": 0.0382, "step": 22876 }, { "epoch": 2.4801604509973982, "grad_norm": 0.08682242035865784, "learning_rate": 1.7390831278108226e-05, "loss": 0.003, "step": 22877 }, { "epoch": 2.4802688638334778, "grad_norm": 0.470598042011261, "learning_rate": 1.738720441027129e-05, "loss": 0.0457, "step": 22878 }, { "epoch": 2.4803772766695578, "grad_norm": 0.02213933691382408, "learning_rate": 1.7383577542434354e-05, "loss": 0.0011, "step": 22879 }, { "epoch": 2.4804856895056373, "grad_norm": 0.5321665406227112, "learning_rate": 1.737995067459742e-05, "loss": 0.0135, "step": 22880 }, { "epoch": 2.4805941023417173, "grad_norm": 0.5891757011413574, "learning_rate": 1.7376323806760482e-05, "loss": 0.0154, "step": 22881 }, { "epoch": 2.4807025151777973, "grad_norm": 0.38307926058769226, "learning_rate": 1.7372696938923547e-05, "loss": 0.0232, "step": 22882 }, { "epoch": 2.480810928013877, "grad_norm": 0.10252974182367325, "learning_rate": 1.736907007108661e-05, "loss": 0.0039, "step": 22883 }, { "epoch": 2.480919340849957, "grad_norm": 0.14482735097408295, "learning_rate": 1.7365443203249675e-05, "loss": 0.0037, "step": 22884 }, { "epoch": 2.4810277536860363, "grad_norm": 0.5489217638969421, "learning_rate": 1.736181633541274e-05, "loss": 0.0129, "step": 22885 }, { "epoch": 2.4811361665221163, "grad_norm": 0.034969039261341095, "learning_rate": 1.7358189467575803e-05, "loss": 0.0015, "step": 22886 }, { "epoch": 2.481244579358196, "grad_norm": 0.14414331316947937, "learning_rate": 1.7354562599738867e-05, "loss": 0.0072, "step": 22887 }, { "epoch": 2.481352992194276, "grad_norm": 0.07426956295967102, "learning_rate": 1.7350935731901928e-05, "loss": 0.0046, "step": 22888 }, { "epoch": 2.4814614050303554, "grad_norm": 0.0136848920956254, "learning_rate": 1.7347308864064995e-05, "loss": 0.0006, "step": 22889 }, { "epoch": 2.4815698178664354, "grad_norm": 0.050258226692676544, "learning_rate": 1.734368199622806e-05, "loss": 0.0017, "step": 22890 }, { "epoch": 2.4816782307025154, "grad_norm": 0.17259155213832855, "learning_rate": 1.7340055128391123e-05, "loss": 0.015, "step": 22891 }, { "epoch": 2.481786643538595, "grad_norm": 0.018933534622192383, "learning_rate": 1.7336428260554184e-05, "loss": 0.0006, "step": 22892 }, { "epoch": 2.481895056374675, "grad_norm": 0.47956976294517517, "learning_rate": 1.7332801392717248e-05, "loss": 0.0124, "step": 22893 }, { "epoch": 2.4820034692107544, "grad_norm": 0.04555768147110939, "learning_rate": 1.7329174524880316e-05, "loss": 0.0033, "step": 22894 }, { "epoch": 2.4821118820468344, "grad_norm": 0.06642400473356247, "learning_rate": 1.732554765704338e-05, "loss": 0.0016, "step": 22895 }, { "epoch": 2.482220294882914, "grad_norm": 0.05270777642726898, "learning_rate": 1.732192078920644e-05, "loss": 0.0029, "step": 22896 }, { "epoch": 2.482328707718994, "grad_norm": 0.11134673655033112, "learning_rate": 1.7318293921369505e-05, "loss": 0.0029, "step": 22897 }, { "epoch": 2.4824371205550735, "grad_norm": 0.4075678586959839, "learning_rate": 1.7314667053532572e-05, "loss": 0.0147, "step": 22898 }, { "epoch": 2.4825455333911535, "grad_norm": 0.19992218911647797, "learning_rate": 1.7311040185695633e-05, "loss": 0.0074, "step": 22899 }, { "epoch": 2.4826539462272335, "grad_norm": 0.09404238313436508, "learning_rate": 1.7307413317858697e-05, "loss": 0.0044, "step": 22900 }, { "epoch": 2.482762359063313, "grad_norm": 0.11768823862075806, "learning_rate": 1.730378645002176e-05, "loss": 0.003, "step": 22901 }, { "epoch": 2.482870771899393, "grad_norm": 0.27902746200561523, "learning_rate": 1.730015958218483e-05, "loss": 0.0159, "step": 22902 }, { "epoch": 2.4829791847354725, "grad_norm": 0.5315399765968323, "learning_rate": 1.729653271434789e-05, "loss": 0.0073, "step": 22903 }, { "epoch": 2.4830875975715525, "grad_norm": 0.2696318030357361, "learning_rate": 1.7292905846510953e-05, "loss": 0.0238, "step": 22904 }, { "epoch": 2.4831960104076325, "grad_norm": 0.06516165286302567, "learning_rate": 1.7289278978674018e-05, "loss": 0.0024, "step": 22905 }, { "epoch": 2.483304423243712, "grad_norm": 0.16314266622066498, "learning_rate": 1.728565211083708e-05, "loss": 0.012, "step": 22906 }, { "epoch": 2.483412836079792, "grad_norm": 0.03024725802242756, "learning_rate": 1.7282025243000146e-05, "loss": 0.0013, "step": 22907 }, { "epoch": 2.4835212489158716, "grad_norm": 0.8839693069458008, "learning_rate": 1.727839837516321e-05, "loss": 0.0316, "step": 22908 }, { "epoch": 2.4836296617519515, "grad_norm": 0.5089051723480225, "learning_rate": 1.7274771507326274e-05, "loss": 0.0358, "step": 22909 }, { "epoch": 2.483738074588031, "grad_norm": 0.049038924276828766, "learning_rate": 1.7271144639489338e-05, "loss": 0.0016, "step": 22910 }, { "epoch": 2.483846487424111, "grad_norm": 0.1336706280708313, "learning_rate": 1.7267517771652402e-05, "loss": 0.0075, "step": 22911 }, { "epoch": 2.4839549002601906, "grad_norm": 0.35769492387771606, "learning_rate": 1.7263890903815466e-05, "loss": 0.0148, "step": 22912 }, { "epoch": 2.4840633130962706, "grad_norm": 0.36360403895378113, "learning_rate": 1.726026403597853e-05, "loss": 0.0244, "step": 22913 }, { "epoch": 2.4841717259323506, "grad_norm": 0.1627500057220459, "learning_rate": 1.7256637168141594e-05, "loss": 0.0072, "step": 22914 }, { "epoch": 2.48428013876843, "grad_norm": 0.016461756080389023, "learning_rate": 1.725301030030466e-05, "loss": 0.0009, "step": 22915 }, { "epoch": 2.48438855160451, "grad_norm": 0.17337749898433685, "learning_rate": 1.7249383432467723e-05, "loss": 0.0055, "step": 22916 }, { "epoch": 2.4844969644405897, "grad_norm": 0.08323980867862701, "learning_rate": 1.7245756564630787e-05, "loss": 0.0041, "step": 22917 }, { "epoch": 2.4846053772766696, "grad_norm": 0.283197283744812, "learning_rate": 1.7242129696793848e-05, "loss": 0.0068, "step": 22918 }, { "epoch": 2.484713790112749, "grad_norm": 0.35849529504776, "learning_rate": 1.723850282895691e-05, "loss": 0.0075, "step": 22919 }, { "epoch": 2.484822202948829, "grad_norm": 0.011363341473042965, "learning_rate": 1.723487596111998e-05, "loss": 0.0003, "step": 22920 }, { "epoch": 2.4849306157849087, "grad_norm": 0.12465684860944748, "learning_rate": 1.7231249093283043e-05, "loss": 0.02, "step": 22921 }, { "epoch": 2.4850390286209887, "grad_norm": 0.015258913859724998, "learning_rate": 1.7227622225446104e-05, "loss": 0.0007, "step": 22922 }, { "epoch": 2.4851474414570687, "grad_norm": 0.2641959488391876, "learning_rate": 1.7223995357609168e-05, "loss": 0.0163, "step": 22923 }, { "epoch": 2.485255854293148, "grad_norm": 0.03191094845533371, "learning_rate": 1.7220368489772236e-05, "loss": 0.0007, "step": 22924 }, { "epoch": 2.485364267129228, "grad_norm": 0.1422004997730255, "learning_rate": 1.7216741621935296e-05, "loss": 0.0067, "step": 22925 }, { "epoch": 2.4854726799653077, "grad_norm": 0.13296376168727875, "learning_rate": 1.721311475409836e-05, "loss": 0.0035, "step": 22926 }, { "epoch": 2.4855810928013877, "grad_norm": 0.28916019201278687, "learning_rate": 1.7209487886261424e-05, "loss": 0.0247, "step": 22927 }, { "epoch": 2.4856895056374677, "grad_norm": 0.03694775700569153, "learning_rate": 1.7205861018424492e-05, "loss": 0.0016, "step": 22928 }, { "epoch": 2.4857979184735473, "grad_norm": 0.01120119821280241, "learning_rate": 1.7202234150587553e-05, "loss": 0.0003, "step": 22929 }, { "epoch": 2.4859063313096272, "grad_norm": 0.07905221730470657, "learning_rate": 1.7198607282750617e-05, "loss": 0.0016, "step": 22930 }, { "epoch": 2.486014744145707, "grad_norm": 0.0342172272503376, "learning_rate": 1.719498041491368e-05, "loss": 0.0009, "step": 22931 }, { "epoch": 2.4861231569817868, "grad_norm": 0.18436501920223236, "learning_rate": 1.719135354707675e-05, "loss": 0.0053, "step": 22932 }, { "epoch": 2.4862315698178663, "grad_norm": 0.03335234522819519, "learning_rate": 1.718772667923981e-05, "loss": 0.0019, "step": 22933 }, { "epoch": 2.4863399826539463, "grad_norm": 0.06542611122131348, "learning_rate": 1.7184099811402873e-05, "loss": 0.005, "step": 22934 }, { "epoch": 2.486448395490026, "grad_norm": 0.12553350627422333, "learning_rate": 1.7180472943565937e-05, "loss": 0.0067, "step": 22935 }, { "epoch": 2.486556808326106, "grad_norm": 0.5226551294326782, "learning_rate": 1.7176846075729e-05, "loss": 0.0111, "step": 22936 }, { "epoch": 2.486665221162186, "grad_norm": 0.2745679020881653, "learning_rate": 1.7173219207892066e-05, "loss": 0.0236, "step": 22937 }, { "epoch": 2.4867736339982653, "grad_norm": 0.13899946212768555, "learning_rate": 1.716959234005513e-05, "loss": 0.006, "step": 22938 }, { "epoch": 2.4868820468343453, "grad_norm": 0.1097935363650322, "learning_rate": 1.7165965472218194e-05, "loss": 0.0049, "step": 22939 }, { "epoch": 2.486990459670425, "grad_norm": 0.3678241968154907, "learning_rate": 1.7162338604381258e-05, "loss": 0.0288, "step": 22940 }, { "epoch": 2.487098872506505, "grad_norm": 0.11780698597431183, "learning_rate": 1.7158711736544322e-05, "loss": 0.0098, "step": 22941 }, { "epoch": 2.4872072853425844, "grad_norm": 0.7409178614616394, "learning_rate": 1.7155084868707386e-05, "loss": 0.0547, "step": 22942 }, { "epoch": 2.4873156981786644, "grad_norm": 0.2507471442222595, "learning_rate": 1.715145800087045e-05, "loss": 0.0145, "step": 22943 }, { "epoch": 2.487424111014744, "grad_norm": 0.03770846128463745, "learning_rate": 1.714783113303351e-05, "loss": 0.0008, "step": 22944 }, { "epoch": 2.487532523850824, "grad_norm": 0.09385951608419418, "learning_rate": 1.714420426519658e-05, "loss": 0.0021, "step": 22945 }, { "epoch": 2.487640936686904, "grad_norm": 0.09139050543308258, "learning_rate": 1.7140577397359642e-05, "loss": 0.0025, "step": 22946 }, { "epoch": 2.4877493495229834, "grad_norm": 0.32955265045166016, "learning_rate": 1.7136950529522707e-05, "loss": 0.0083, "step": 22947 }, { "epoch": 2.4878577623590634, "grad_norm": 0.057345274835824966, "learning_rate": 1.7133323661685767e-05, "loss": 0.0019, "step": 22948 }, { "epoch": 2.487966175195143, "grad_norm": 0.1784893125295639, "learning_rate": 1.712969679384883e-05, "loss": 0.0029, "step": 22949 }, { "epoch": 2.488074588031223, "grad_norm": 0.056260980665683746, "learning_rate": 1.71260699260119e-05, "loss": 0.0016, "step": 22950 }, { "epoch": 2.488183000867303, "grad_norm": 0.12946531176567078, "learning_rate": 1.712244305817496e-05, "loss": 0.0025, "step": 22951 }, { "epoch": 2.4882914137033825, "grad_norm": 0.4288652241230011, "learning_rate": 1.7118816190338024e-05, "loss": 0.023, "step": 22952 }, { "epoch": 2.4883998265394625, "grad_norm": 0.2502591907978058, "learning_rate": 1.7115189322501088e-05, "loss": 0.0067, "step": 22953 }, { "epoch": 2.488508239375542, "grad_norm": 0.10676165670156479, "learning_rate": 1.7111562454664155e-05, "loss": 0.0048, "step": 22954 }, { "epoch": 2.488616652211622, "grad_norm": 0.16379910707473755, "learning_rate": 1.7107935586827216e-05, "loss": 0.0085, "step": 22955 }, { "epoch": 2.4887250650477015, "grad_norm": 0.5551300048828125, "learning_rate": 1.710430871899028e-05, "loss": 0.0358, "step": 22956 }, { "epoch": 2.4888334778837815, "grad_norm": 0.5768399238586426, "learning_rate": 1.7100681851153344e-05, "loss": 0.0144, "step": 22957 }, { "epoch": 2.488941890719861, "grad_norm": 0.006585698574781418, "learning_rate": 1.7097054983316412e-05, "loss": 0.0002, "step": 22958 }, { "epoch": 2.489050303555941, "grad_norm": 0.0036985219921916723, "learning_rate": 1.7093428115479472e-05, "loss": 0.0002, "step": 22959 }, { "epoch": 2.489158716392021, "grad_norm": 0.6779606938362122, "learning_rate": 1.7089801247642537e-05, "loss": 0.0128, "step": 22960 }, { "epoch": 2.4892671292281006, "grad_norm": 0.3172156512737274, "learning_rate": 1.70861743798056e-05, "loss": 0.0165, "step": 22961 }, { "epoch": 2.4893755420641805, "grad_norm": 0.19893725216388702, "learning_rate": 1.7082547511968665e-05, "loss": 0.0109, "step": 22962 }, { "epoch": 2.48948395490026, "grad_norm": 1.187977910041809, "learning_rate": 1.707892064413173e-05, "loss": 0.0292, "step": 22963 }, { "epoch": 2.48959236773634, "grad_norm": 0.21255987882614136, "learning_rate": 1.7075293776294793e-05, "loss": 0.0059, "step": 22964 }, { "epoch": 2.4897007805724196, "grad_norm": 0.1734425574541092, "learning_rate": 1.7071666908457857e-05, "loss": 0.006, "step": 22965 }, { "epoch": 2.4898091934084996, "grad_norm": 0.06528456509113312, "learning_rate": 1.706804004062092e-05, "loss": 0.005, "step": 22966 }, { "epoch": 2.489917606244579, "grad_norm": 0.018225671723484993, "learning_rate": 1.7064413172783985e-05, "loss": 0.0007, "step": 22967 }, { "epoch": 2.490026019080659, "grad_norm": 0.014040235430002213, "learning_rate": 1.706078630494705e-05, "loss": 0.0004, "step": 22968 }, { "epoch": 2.490134431916739, "grad_norm": 0.1599138379096985, "learning_rate": 1.7057159437110114e-05, "loss": 0.0023, "step": 22969 }, { "epoch": 2.4902428447528187, "grad_norm": 0.07256465405225754, "learning_rate": 1.7053532569273174e-05, "loss": 0.003, "step": 22970 }, { "epoch": 2.4903512575888986, "grad_norm": 0.04479508101940155, "learning_rate": 1.7049905701436242e-05, "loss": 0.0026, "step": 22971 }, { "epoch": 2.490459670424978, "grad_norm": 0.7147378921508789, "learning_rate": 1.7046278833599306e-05, "loss": 0.0292, "step": 22972 }, { "epoch": 2.490568083261058, "grad_norm": 0.09848253428936005, "learning_rate": 1.704265196576237e-05, "loss": 0.0055, "step": 22973 }, { "epoch": 2.490676496097138, "grad_norm": 0.4598461389541626, "learning_rate": 1.703902509792543e-05, "loss": 0.0333, "step": 22974 }, { "epoch": 2.4907849089332177, "grad_norm": 0.4215727150440216, "learning_rate": 1.7035398230088495e-05, "loss": 0.0223, "step": 22975 }, { "epoch": 2.4908933217692977, "grad_norm": 0.44084975123405457, "learning_rate": 1.7031771362251562e-05, "loss": 0.0268, "step": 22976 }, { "epoch": 2.491001734605377, "grad_norm": 0.07484698295593262, "learning_rate": 1.7028144494414623e-05, "loss": 0.0025, "step": 22977 }, { "epoch": 2.491110147441457, "grad_norm": 0.3851065933704376, "learning_rate": 1.7024517626577687e-05, "loss": 0.0167, "step": 22978 }, { "epoch": 2.4912185602775367, "grad_norm": 0.7785789370536804, "learning_rate": 1.702089075874075e-05, "loss": 0.0138, "step": 22979 }, { "epoch": 2.4913269731136167, "grad_norm": 0.6820226311683655, "learning_rate": 1.701726389090382e-05, "loss": 0.0477, "step": 22980 }, { "epoch": 2.4914353859496963, "grad_norm": 0.4932788908481598, "learning_rate": 1.701363702306688e-05, "loss": 0.0092, "step": 22981 }, { "epoch": 2.4915437987857763, "grad_norm": 0.17175619304180145, "learning_rate": 1.7010010155229943e-05, "loss": 0.0086, "step": 22982 }, { "epoch": 2.4916522116218562, "grad_norm": 0.08175290375947952, "learning_rate": 1.7006383287393008e-05, "loss": 0.0055, "step": 22983 }, { "epoch": 2.491760624457936, "grad_norm": 0.1266147643327713, "learning_rate": 1.7002756419556075e-05, "loss": 0.0062, "step": 22984 }, { "epoch": 2.4918690372940158, "grad_norm": 0.4683764576911926, "learning_rate": 1.6999129551719136e-05, "loss": 0.0219, "step": 22985 }, { "epoch": 2.4919774501300953, "grad_norm": 0.25771987438201904, "learning_rate": 1.69955026838822e-05, "loss": 0.0068, "step": 22986 }, { "epoch": 2.4920858629661753, "grad_norm": 0.22687366604804993, "learning_rate": 1.6991875816045264e-05, "loss": 0.0082, "step": 22987 }, { "epoch": 2.492194275802255, "grad_norm": 0.23904363811016083, "learning_rate": 1.6988248948208328e-05, "loss": 0.007, "step": 22988 }, { "epoch": 2.492302688638335, "grad_norm": 0.04967387020587921, "learning_rate": 1.6984622080371392e-05, "loss": 0.0022, "step": 22989 }, { "epoch": 2.4924111014744144, "grad_norm": 0.05424977466464043, "learning_rate": 1.6980995212534456e-05, "loss": 0.0027, "step": 22990 }, { "epoch": 2.4925195143104943, "grad_norm": 0.22329756617546082, "learning_rate": 1.697736834469752e-05, "loss": 0.0083, "step": 22991 }, { "epoch": 2.4926279271465743, "grad_norm": 0.028833169490098953, "learning_rate": 1.6973741476860585e-05, "loss": 0.0014, "step": 22992 }, { "epoch": 2.492736339982654, "grad_norm": 0.37206748127937317, "learning_rate": 1.697011460902365e-05, "loss": 0.0163, "step": 22993 }, { "epoch": 2.492844752818734, "grad_norm": 0.4543530344963074, "learning_rate": 1.6966487741186713e-05, "loss": 0.0245, "step": 22994 }, { "epoch": 2.4929531656548134, "grad_norm": 0.023374872282147408, "learning_rate": 1.6962860873349777e-05, "loss": 0.001, "step": 22995 }, { "epoch": 2.4930615784908934, "grad_norm": 0.5033283829689026, "learning_rate": 1.6959234005512838e-05, "loss": 0.0073, "step": 22996 }, { "epoch": 2.493169991326973, "grad_norm": 0.18788452446460724, "learning_rate": 1.6955607137675905e-05, "loss": 0.0042, "step": 22997 }, { "epoch": 2.493278404163053, "grad_norm": 0.5818939805030823, "learning_rate": 1.695198026983897e-05, "loss": 0.0189, "step": 22998 }, { "epoch": 2.4933868169991324, "grad_norm": 0.12205447256565094, "learning_rate": 1.6948353402002033e-05, "loss": 0.0054, "step": 22999 }, { "epoch": 2.4934952298352124, "grad_norm": 0.288786381483078, "learning_rate": 1.6944726534165094e-05, "loss": 0.0084, "step": 23000 }, { "epoch": 2.4936036426712924, "grad_norm": 0.1865539848804474, "learning_rate": 1.6941099666328158e-05, "loss": 0.0054, "step": 23001 }, { "epoch": 2.493712055507372, "grad_norm": 0.4717243015766144, "learning_rate": 1.6937472798491226e-05, "loss": 0.0448, "step": 23002 }, { "epoch": 2.493820468343452, "grad_norm": 0.20443691313266754, "learning_rate": 1.6933845930654286e-05, "loss": 0.0116, "step": 23003 }, { "epoch": 2.4939288811795315, "grad_norm": 0.6660239100456238, "learning_rate": 1.693021906281735e-05, "loss": 0.0087, "step": 23004 }, { "epoch": 2.4940372940156115, "grad_norm": 0.3442312479019165, "learning_rate": 1.6926592194980415e-05, "loss": 0.0239, "step": 23005 }, { "epoch": 2.4941457068516915, "grad_norm": 0.2500593066215515, "learning_rate": 1.6922965327143482e-05, "loss": 0.0123, "step": 23006 }, { "epoch": 2.494254119687771, "grad_norm": 0.012722363695502281, "learning_rate": 1.6919338459306543e-05, "loss": 0.0006, "step": 23007 }, { "epoch": 2.494362532523851, "grad_norm": 0.019918039441108704, "learning_rate": 1.6915711591469607e-05, "loss": 0.0006, "step": 23008 }, { "epoch": 2.4944709453599305, "grad_norm": 0.12333992123603821, "learning_rate": 1.691208472363267e-05, "loss": 0.0119, "step": 23009 }, { "epoch": 2.4945793581960105, "grad_norm": 0.039962783455848694, "learning_rate": 1.690845785579574e-05, "loss": 0.0017, "step": 23010 }, { "epoch": 2.49468777103209, "grad_norm": 0.24011752009391785, "learning_rate": 1.69048309879588e-05, "loss": 0.0212, "step": 23011 }, { "epoch": 2.49479618386817, "grad_norm": 0.4050386846065521, "learning_rate": 1.6901204120121863e-05, "loss": 0.0168, "step": 23012 }, { "epoch": 2.4949045967042496, "grad_norm": 0.0862322673201561, "learning_rate": 1.6897577252284927e-05, "loss": 0.0022, "step": 23013 }, { "epoch": 2.4950130095403296, "grad_norm": 0.04632658138871193, "learning_rate": 1.689395038444799e-05, "loss": 0.0024, "step": 23014 }, { "epoch": 2.4951214223764095, "grad_norm": 0.08350548148155212, "learning_rate": 1.6890323516611056e-05, "loss": 0.0029, "step": 23015 }, { "epoch": 2.495229835212489, "grad_norm": 0.7735429406166077, "learning_rate": 1.688669664877412e-05, "loss": 0.0363, "step": 23016 }, { "epoch": 2.495338248048569, "grad_norm": 1.6869208812713623, "learning_rate": 1.6883069780937184e-05, "loss": 0.045, "step": 23017 }, { "epoch": 2.4954466608846486, "grad_norm": 0.16349193453788757, "learning_rate": 1.6879442913100248e-05, "loss": 0.0033, "step": 23018 }, { "epoch": 2.4955550737207286, "grad_norm": 0.5315712690353394, "learning_rate": 1.6875816045263312e-05, "loss": 0.0494, "step": 23019 }, { "epoch": 2.495663486556808, "grad_norm": 0.18466781079769135, "learning_rate": 1.6872189177426376e-05, "loss": 0.0045, "step": 23020 }, { "epoch": 2.495771899392888, "grad_norm": 0.08969329297542572, "learning_rate": 1.686856230958944e-05, "loss": 0.006, "step": 23021 }, { "epoch": 2.4958803122289677, "grad_norm": 0.02790103293955326, "learning_rate": 1.68649354417525e-05, "loss": 0.0005, "step": 23022 }, { "epoch": 2.4959887250650477, "grad_norm": 0.4387020766735077, "learning_rate": 1.686130857391557e-05, "loss": 0.0186, "step": 23023 }, { "epoch": 2.4960971379011276, "grad_norm": 0.07737917453050613, "learning_rate": 1.6857681706078633e-05, "loss": 0.0036, "step": 23024 }, { "epoch": 2.496205550737207, "grad_norm": 0.13189749419689178, "learning_rate": 1.6854054838241697e-05, "loss": 0.0103, "step": 23025 }, { "epoch": 2.496313963573287, "grad_norm": 0.12085679918527603, "learning_rate": 1.6850427970404757e-05, "loss": 0.0073, "step": 23026 }, { "epoch": 2.4964223764093667, "grad_norm": 0.07824511080980301, "learning_rate": 1.6846801102567825e-05, "loss": 0.0024, "step": 23027 }, { "epoch": 2.4965307892454467, "grad_norm": 0.02398492395877838, "learning_rate": 1.684317423473089e-05, "loss": 0.0007, "step": 23028 }, { "epoch": 2.4966392020815267, "grad_norm": 0.12184034287929535, "learning_rate": 1.683954736689395e-05, "loss": 0.0042, "step": 23029 }, { "epoch": 2.496747614917606, "grad_norm": 0.04987740516662598, "learning_rate": 1.6835920499057014e-05, "loss": 0.0017, "step": 23030 }, { "epoch": 2.496856027753686, "grad_norm": 0.05110112205147743, "learning_rate": 1.6832293631220078e-05, "loss": 0.0038, "step": 23031 }, { "epoch": 2.4969644405897657, "grad_norm": 0.2089807242155075, "learning_rate": 1.6828666763383145e-05, "loss": 0.005, "step": 23032 }, { "epoch": 2.4970728534258457, "grad_norm": 0.13244618475437164, "learning_rate": 1.6825039895546206e-05, "loss": 0.0031, "step": 23033 }, { "epoch": 2.4971812662619253, "grad_norm": 0.20924530923366547, "learning_rate": 1.682141302770927e-05, "loss": 0.0046, "step": 23034 }, { "epoch": 2.4972896790980053, "grad_norm": 0.15398456156253815, "learning_rate": 1.6817786159872334e-05, "loss": 0.0063, "step": 23035 }, { "epoch": 2.497398091934085, "grad_norm": 0.03317990526556969, "learning_rate": 1.6814159292035402e-05, "loss": 0.0013, "step": 23036 }, { "epoch": 2.497506504770165, "grad_norm": 1.1060059070587158, "learning_rate": 1.6810532424198462e-05, "loss": 0.0402, "step": 23037 }, { "epoch": 2.4976149176062448, "grad_norm": 0.5764657258987427, "learning_rate": 1.6806905556361527e-05, "loss": 0.0187, "step": 23038 }, { "epoch": 2.4977233304423243, "grad_norm": 0.06297232210636139, "learning_rate": 1.680327868852459e-05, "loss": 0.0016, "step": 23039 }, { "epoch": 2.4978317432784043, "grad_norm": 0.22675269842147827, "learning_rate": 1.6799651820687655e-05, "loss": 0.0041, "step": 23040 }, { "epoch": 2.497940156114484, "grad_norm": 0.23554924130439758, "learning_rate": 1.679602495285072e-05, "loss": 0.0077, "step": 23041 }, { "epoch": 2.498048568950564, "grad_norm": 0.43167248368263245, "learning_rate": 1.6792398085013783e-05, "loss": 0.0107, "step": 23042 }, { "epoch": 2.4981569817866434, "grad_norm": 0.5811240673065186, "learning_rate": 1.6788771217176847e-05, "loss": 0.0248, "step": 23043 }, { "epoch": 2.4982653946227233, "grad_norm": 0.3704950511455536, "learning_rate": 1.678514434933991e-05, "loss": 0.0079, "step": 23044 }, { "epoch": 2.498373807458803, "grad_norm": 0.5035800337791443, "learning_rate": 1.6781517481502975e-05, "loss": 0.031, "step": 23045 }, { "epoch": 2.498482220294883, "grad_norm": 0.2071518748998642, "learning_rate": 1.677789061366604e-05, "loss": 0.0166, "step": 23046 }, { "epoch": 2.498590633130963, "grad_norm": 0.30832982063293457, "learning_rate": 1.6774263745829104e-05, "loss": 0.0083, "step": 23047 }, { "epoch": 2.4986990459670424, "grad_norm": 1.329906702041626, "learning_rate": 1.6770636877992164e-05, "loss": 0.0135, "step": 23048 }, { "epoch": 2.4988074588031224, "grad_norm": 0.06554878503084183, "learning_rate": 1.6767010010155232e-05, "loss": 0.0018, "step": 23049 }, { "epoch": 2.498915871639202, "grad_norm": 0.07677700370550156, "learning_rate": 1.6763383142318296e-05, "loss": 0.0061, "step": 23050 }, { "epoch": 2.499024284475282, "grad_norm": 0.25584498047828674, "learning_rate": 1.675975627448136e-05, "loss": 0.0143, "step": 23051 }, { "epoch": 2.499132697311362, "grad_norm": 0.08139250427484512, "learning_rate": 1.675612940664442e-05, "loss": 0.0051, "step": 23052 }, { "epoch": 2.4992411101474414, "grad_norm": 0.3803671896457672, "learning_rate": 1.6752502538807488e-05, "loss": 0.0068, "step": 23053 }, { "epoch": 2.4993495229835214, "grad_norm": 0.05394848808646202, "learning_rate": 1.6748875670970552e-05, "loss": 0.0032, "step": 23054 }, { "epoch": 2.499457935819601, "grad_norm": 0.22973304986953735, "learning_rate": 1.6745248803133613e-05, "loss": 0.0087, "step": 23055 }, { "epoch": 2.499566348655681, "grad_norm": 0.17072561383247375, "learning_rate": 1.6741621935296677e-05, "loss": 0.0086, "step": 23056 }, { "epoch": 2.4996747614917605, "grad_norm": 0.14202198386192322, "learning_rate": 1.673799506745974e-05, "loss": 0.0035, "step": 23057 }, { "epoch": 2.4997831743278405, "grad_norm": 0.3907772898674011, "learning_rate": 1.673436819962281e-05, "loss": 0.0211, "step": 23058 }, { "epoch": 2.49989158716392, "grad_norm": 0.4557224214076996, "learning_rate": 1.673074133178587e-05, "loss": 0.0097, "step": 23059 }, { "epoch": 2.5, "grad_norm": 0.24661485850811005, "learning_rate": 1.6727114463948934e-05, "loss": 0.01, "step": 23060 }, { "epoch": 2.50010841283608, "grad_norm": 0.5532169938087463, "learning_rate": 1.6723487596111998e-05, "loss": 0.0283, "step": 23061 }, { "epoch": 2.5002168256721595, "grad_norm": 1.1084017753601074, "learning_rate": 1.6719860728275065e-05, "loss": 0.0496, "step": 23062 }, { "epoch": 2.5003252385082395, "grad_norm": 0.16798315942287445, "learning_rate": 1.6716233860438126e-05, "loss": 0.0092, "step": 23063 }, { "epoch": 2.500433651344319, "grad_norm": 0.18566718697547913, "learning_rate": 1.671260699260119e-05, "loss": 0.0075, "step": 23064 }, { "epoch": 2.500542064180399, "grad_norm": 0.29909923672676086, "learning_rate": 1.6708980124764254e-05, "loss": 0.0399, "step": 23065 }, { "epoch": 2.500650477016479, "grad_norm": 0.30596625804901123, "learning_rate": 1.6705353256927318e-05, "loss": 0.0066, "step": 23066 }, { "epoch": 2.5007588898525586, "grad_norm": 0.08837446570396423, "learning_rate": 1.6701726389090382e-05, "loss": 0.004, "step": 23067 }, { "epoch": 2.500867302688638, "grad_norm": 0.05479996278882027, "learning_rate": 1.6698099521253446e-05, "loss": 0.0015, "step": 23068 }, { "epoch": 2.500975715524718, "grad_norm": 0.31566616892814636, "learning_rate": 1.669447265341651e-05, "loss": 0.0249, "step": 23069 }, { "epoch": 2.501084128360798, "grad_norm": 0.3021043539047241, "learning_rate": 1.6690845785579575e-05, "loss": 0.0075, "step": 23070 }, { "epoch": 2.5011925411968776, "grad_norm": 0.13078227639198303, "learning_rate": 1.668721891774264e-05, "loss": 0.0048, "step": 23071 }, { "epoch": 2.5013009540329576, "grad_norm": 0.6641668081283569, "learning_rate": 1.6683592049905703e-05, "loss": 0.0176, "step": 23072 }, { "epoch": 2.501409366869037, "grad_norm": 0.04137709364295006, "learning_rate": 1.6679965182068767e-05, "loss": 0.0025, "step": 23073 }, { "epoch": 2.501517779705117, "grad_norm": 0.10363436490297318, "learning_rate": 1.6676338314231828e-05, "loss": 0.0053, "step": 23074 }, { "epoch": 2.501626192541197, "grad_norm": 0.09657301753759384, "learning_rate": 1.6672711446394895e-05, "loss": 0.004, "step": 23075 }, { "epoch": 2.5017346053772767, "grad_norm": 0.2995036244392395, "learning_rate": 1.666908457855796e-05, "loss": 0.0052, "step": 23076 }, { "epoch": 2.501843018213356, "grad_norm": 0.038241125643253326, "learning_rate": 1.6665457710721023e-05, "loss": 0.0015, "step": 23077 }, { "epoch": 2.501951431049436, "grad_norm": 0.0997464582324028, "learning_rate": 1.6661830842884084e-05, "loss": 0.0086, "step": 23078 }, { "epoch": 2.502059843885516, "grad_norm": 0.8453040719032288, "learning_rate": 1.665820397504715e-05, "loss": 0.0208, "step": 23079 }, { "epoch": 2.5021682567215957, "grad_norm": 0.04222584515810013, "learning_rate": 1.6654577107210216e-05, "loss": 0.0016, "step": 23080 }, { "epoch": 2.5022766695576757, "grad_norm": 0.5182273387908936, "learning_rate": 1.665095023937328e-05, "loss": 0.0122, "step": 23081 }, { "epoch": 2.5023850823937552, "grad_norm": 0.36121293902397156, "learning_rate": 1.664732337153634e-05, "loss": 0.0137, "step": 23082 }, { "epoch": 2.502493495229835, "grad_norm": 0.16575711965560913, "learning_rate": 1.6643696503699405e-05, "loss": 0.0205, "step": 23083 }, { "epoch": 2.502601908065915, "grad_norm": 0.17595450580120087, "learning_rate": 1.6640069635862472e-05, "loss": 0.006, "step": 23084 }, { "epoch": 2.5027103209019947, "grad_norm": 0.5458499193191528, "learning_rate": 1.6636442768025533e-05, "loss": 0.0043, "step": 23085 }, { "epoch": 2.5028187337380747, "grad_norm": 0.034461408853530884, "learning_rate": 1.6632815900188597e-05, "loss": 0.0012, "step": 23086 }, { "epoch": 2.5029271465741543, "grad_norm": 0.18715640902519226, "learning_rate": 1.662918903235166e-05, "loss": 0.0127, "step": 23087 }, { "epoch": 2.5030355594102343, "grad_norm": 0.7837252616882324, "learning_rate": 1.662556216451473e-05, "loss": 0.0169, "step": 23088 }, { "epoch": 2.503143972246314, "grad_norm": 0.15495899319648743, "learning_rate": 1.662193529667779e-05, "loss": 0.0039, "step": 23089 }, { "epoch": 2.503252385082394, "grad_norm": 0.6213909983634949, "learning_rate": 1.6618308428840853e-05, "loss": 0.0247, "step": 23090 }, { "epoch": 2.5033607979184733, "grad_norm": 0.37744519114494324, "learning_rate": 1.6614681561003917e-05, "loss": 0.0255, "step": 23091 }, { "epoch": 2.5034692107545533, "grad_norm": 0.10086793452501297, "learning_rate": 1.661105469316698e-05, "loss": 0.0041, "step": 23092 }, { "epoch": 2.5035776235906333, "grad_norm": 0.04261600226163864, "learning_rate": 1.6607427825330046e-05, "loss": 0.0029, "step": 23093 }, { "epoch": 2.503686036426713, "grad_norm": 0.5135709643363953, "learning_rate": 1.660380095749311e-05, "loss": 0.0111, "step": 23094 }, { "epoch": 2.503794449262793, "grad_norm": 0.1824594885110855, "learning_rate": 1.6600174089656174e-05, "loss": 0.0098, "step": 23095 }, { "epoch": 2.5039028620988724, "grad_norm": 0.467975914478302, "learning_rate": 1.6596547221819238e-05, "loss": 0.0154, "step": 23096 }, { "epoch": 2.5040112749349523, "grad_norm": 0.12875713407993317, "learning_rate": 1.6592920353982302e-05, "loss": 0.0046, "step": 23097 }, { "epoch": 2.5041196877710323, "grad_norm": 0.5655761361122131, "learning_rate": 1.6589293486145366e-05, "loss": 0.0137, "step": 23098 }, { "epoch": 2.504228100607112, "grad_norm": 0.703046441078186, "learning_rate": 1.658566661830843e-05, "loss": 0.0367, "step": 23099 }, { "epoch": 2.5043365134431914, "grad_norm": 0.003454432589933276, "learning_rate": 1.658203975047149e-05, "loss": 0.0001, "step": 23100 }, { "epoch": 2.5044449262792714, "grad_norm": 0.7195692658424377, "learning_rate": 1.657841288263456e-05, "loss": 0.0206, "step": 23101 }, { "epoch": 2.5045533391153514, "grad_norm": 0.1731950044631958, "learning_rate": 1.6574786014797623e-05, "loss": 0.0038, "step": 23102 }, { "epoch": 2.504661751951431, "grad_norm": 0.11589300632476807, "learning_rate": 1.6571159146960687e-05, "loss": 0.0032, "step": 23103 }, { "epoch": 2.504770164787511, "grad_norm": 0.023452792316675186, "learning_rate": 1.6567532279123747e-05, "loss": 0.0007, "step": 23104 }, { "epoch": 2.5048785776235905, "grad_norm": 0.05007253587245941, "learning_rate": 1.6563905411286815e-05, "loss": 0.0017, "step": 23105 }, { "epoch": 2.5049869904596704, "grad_norm": 0.2006697952747345, "learning_rate": 1.656027854344988e-05, "loss": 0.0047, "step": 23106 }, { "epoch": 2.5050954032957504, "grad_norm": 0.1586412787437439, "learning_rate": 1.6556651675612943e-05, "loss": 0.0028, "step": 23107 }, { "epoch": 2.50520381613183, "grad_norm": 0.4377419948577881, "learning_rate": 1.6553024807776004e-05, "loss": 0.0268, "step": 23108 }, { "epoch": 2.50531222896791, "grad_norm": 0.4577236473560333, "learning_rate": 1.6549397939939068e-05, "loss": 0.0124, "step": 23109 }, { "epoch": 2.5054206418039895, "grad_norm": 1.4064252376556396, "learning_rate": 1.6545771072102135e-05, "loss": 0.0446, "step": 23110 }, { "epoch": 2.5055290546400695, "grad_norm": 0.2636089324951172, "learning_rate": 1.6542144204265196e-05, "loss": 0.0136, "step": 23111 }, { "epoch": 2.505637467476149, "grad_norm": 0.09070076048374176, "learning_rate": 1.653851733642826e-05, "loss": 0.006, "step": 23112 }, { "epoch": 2.505745880312229, "grad_norm": 0.10528072714805603, "learning_rate": 1.6534890468591324e-05, "loss": 0.0027, "step": 23113 }, { "epoch": 2.5058542931483085, "grad_norm": 0.062169935554265976, "learning_rate": 1.6531263600754392e-05, "loss": 0.0024, "step": 23114 }, { "epoch": 2.5059627059843885, "grad_norm": 0.08146460354328156, "learning_rate": 1.6527636732917453e-05, "loss": 0.0061, "step": 23115 }, { "epoch": 2.5060711188204685, "grad_norm": 0.3910146653652191, "learning_rate": 1.6524009865080517e-05, "loss": 0.016, "step": 23116 }, { "epoch": 2.506179531656548, "grad_norm": 0.019820760935544968, "learning_rate": 1.652038299724358e-05, "loss": 0.0009, "step": 23117 }, { "epoch": 2.506287944492628, "grad_norm": 0.5068745613098145, "learning_rate": 1.6516756129406645e-05, "loss": 0.0238, "step": 23118 }, { "epoch": 2.5063963573287076, "grad_norm": 0.029063580557703972, "learning_rate": 1.651312926156971e-05, "loss": 0.0009, "step": 23119 }, { "epoch": 2.5065047701647876, "grad_norm": 0.10066881775856018, "learning_rate": 1.6509502393732773e-05, "loss": 0.0041, "step": 23120 }, { "epoch": 2.5066131830008676, "grad_norm": 0.2463212013244629, "learning_rate": 1.6505875525895837e-05, "loss": 0.0043, "step": 23121 }, { "epoch": 2.506721595836947, "grad_norm": 0.3305889368057251, "learning_rate": 1.65022486580589e-05, "loss": 0.0201, "step": 23122 }, { "epoch": 2.5068300086730266, "grad_norm": 0.023446356877684593, "learning_rate": 1.6498621790221965e-05, "loss": 0.0005, "step": 23123 }, { "epoch": 2.5069384215091066, "grad_norm": 0.3575919568538666, "learning_rate": 1.649499492238503e-05, "loss": 0.013, "step": 23124 }, { "epoch": 2.5070468343451866, "grad_norm": 0.15833187103271484, "learning_rate": 1.6491368054548094e-05, "loss": 0.0077, "step": 23125 }, { "epoch": 2.507155247181266, "grad_norm": 0.39249879121780396, "learning_rate": 1.6487741186711154e-05, "loss": 0.0194, "step": 23126 }, { "epoch": 2.507263660017346, "grad_norm": 0.5711762309074402, "learning_rate": 1.6484114318874222e-05, "loss": 0.04, "step": 23127 }, { "epoch": 2.5073720728534257, "grad_norm": 0.1568945348262787, "learning_rate": 1.6480487451037286e-05, "loss": 0.0021, "step": 23128 }, { "epoch": 2.5074804856895057, "grad_norm": 0.38276201486587524, "learning_rate": 1.647686058320035e-05, "loss": 0.0212, "step": 23129 }, { "epoch": 2.5075888985255856, "grad_norm": 0.1012115553021431, "learning_rate": 1.647323371536341e-05, "loss": 0.0025, "step": 23130 }, { "epoch": 2.507697311361665, "grad_norm": 0.030938783660531044, "learning_rate": 1.6469606847526478e-05, "loss": 0.0015, "step": 23131 }, { "epoch": 2.507805724197745, "grad_norm": 0.020362619310617447, "learning_rate": 1.6465979979689542e-05, "loss": 0.0007, "step": 23132 }, { "epoch": 2.5079141370338247, "grad_norm": 0.19109836220741272, "learning_rate": 1.6462353111852606e-05, "loss": 0.0074, "step": 23133 }, { "epoch": 2.5080225498699047, "grad_norm": 0.11069212108850479, "learning_rate": 1.6458726244015667e-05, "loss": 0.0067, "step": 23134 }, { "epoch": 2.5081309627059842, "grad_norm": 0.5347039699554443, "learning_rate": 1.6455099376178735e-05, "loss": 0.0105, "step": 23135 }, { "epoch": 2.508239375542064, "grad_norm": 0.279217004776001, "learning_rate": 1.64514725083418e-05, "loss": 0.0088, "step": 23136 }, { "epoch": 2.5083477883781438, "grad_norm": 0.23488245904445648, "learning_rate": 1.644784564050486e-05, "loss": 0.0105, "step": 23137 }, { "epoch": 2.5084562012142237, "grad_norm": 0.2745215594768524, "learning_rate": 1.6444218772667924e-05, "loss": 0.0109, "step": 23138 }, { "epoch": 2.5085646140503037, "grad_norm": 0.15413331985473633, "learning_rate": 1.6440591904830988e-05, "loss": 0.0065, "step": 23139 }, { "epoch": 2.5086730268863833, "grad_norm": 0.021056203171610832, "learning_rate": 1.6436965036994055e-05, "loss": 0.001, "step": 23140 }, { "epoch": 2.5087814397224633, "grad_norm": 0.6664252281188965, "learning_rate": 1.6433338169157116e-05, "loss": 0.0224, "step": 23141 }, { "epoch": 2.508889852558543, "grad_norm": 0.38762161135673523, "learning_rate": 1.642971130132018e-05, "loss": 0.0197, "step": 23142 }, { "epoch": 2.508998265394623, "grad_norm": 0.21367478370666504, "learning_rate": 1.6426084433483244e-05, "loss": 0.0077, "step": 23143 }, { "epoch": 2.5091066782307028, "grad_norm": 0.02838440239429474, "learning_rate": 1.6422457565646308e-05, "loss": 0.0016, "step": 23144 }, { "epoch": 2.5092150910667823, "grad_norm": 0.38399192690849304, "learning_rate": 1.6418830697809372e-05, "loss": 0.018, "step": 23145 }, { "epoch": 2.509323503902862, "grad_norm": 0.01259525679051876, "learning_rate": 1.6415203829972436e-05, "loss": 0.0006, "step": 23146 }, { "epoch": 2.509431916738942, "grad_norm": 0.09378872811794281, "learning_rate": 1.64115769621355e-05, "loss": 0.0017, "step": 23147 }, { "epoch": 2.509540329575022, "grad_norm": 0.45868387818336487, "learning_rate": 1.6407950094298565e-05, "loss": 0.011, "step": 23148 }, { "epoch": 2.5096487424111014, "grad_norm": 0.44103002548217773, "learning_rate": 1.640432322646163e-05, "loss": 0.0152, "step": 23149 }, { "epoch": 2.5097571552471813, "grad_norm": 0.06384371966123581, "learning_rate": 1.6400696358624693e-05, "loss": 0.0028, "step": 23150 }, { "epoch": 2.509865568083261, "grad_norm": 0.12266995012760162, "learning_rate": 1.6397069490787757e-05, "loss": 0.0041, "step": 23151 }, { "epoch": 2.509973980919341, "grad_norm": 0.09607254713773727, "learning_rate": 1.6393442622950818e-05, "loss": 0.0051, "step": 23152 }, { "epoch": 2.510082393755421, "grad_norm": 0.3959726393222809, "learning_rate": 1.6389815755113885e-05, "loss": 0.0123, "step": 23153 }, { "epoch": 2.5101908065915004, "grad_norm": 1.1053192615509033, "learning_rate": 1.638618888727695e-05, "loss": 0.0232, "step": 23154 }, { "epoch": 2.5102992194275804, "grad_norm": 0.7596005797386169, "learning_rate": 1.6382562019440013e-05, "loss": 0.0097, "step": 23155 }, { "epoch": 2.51040763226366, "grad_norm": 1.0707744359970093, "learning_rate": 1.6378935151603074e-05, "loss": 0.0248, "step": 23156 }, { "epoch": 2.51051604509974, "grad_norm": 0.15295548737049103, "learning_rate": 1.637530828376614e-05, "loss": 0.0103, "step": 23157 }, { "epoch": 2.5106244579358195, "grad_norm": 0.5419741272926331, "learning_rate": 1.6371681415929206e-05, "loss": 0.0086, "step": 23158 }, { "epoch": 2.5107328707718994, "grad_norm": 0.3137054443359375, "learning_rate": 1.636805454809227e-05, "loss": 0.0234, "step": 23159 }, { "epoch": 2.510841283607979, "grad_norm": 0.22585457563400269, "learning_rate": 1.636442768025533e-05, "loss": 0.0045, "step": 23160 }, { "epoch": 2.510949696444059, "grad_norm": 0.24719539284706116, "learning_rate": 1.6360800812418398e-05, "loss": 0.0108, "step": 23161 }, { "epoch": 2.511058109280139, "grad_norm": 0.17837931215763092, "learning_rate": 1.6357173944581462e-05, "loss": 0.0054, "step": 23162 }, { "epoch": 2.5111665221162185, "grad_norm": 0.1708821952342987, "learning_rate": 1.6353547076744523e-05, "loss": 0.0017, "step": 23163 }, { "epoch": 2.5112749349522985, "grad_norm": 0.3638252019882202, "learning_rate": 1.6349920208907587e-05, "loss": 0.0047, "step": 23164 }, { "epoch": 2.511383347788378, "grad_norm": 0.14819025993347168, "learning_rate": 1.634629334107065e-05, "loss": 0.0059, "step": 23165 }, { "epoch": 2.511491760624458, "grad_norm": 0.025895075872540474, "learning_rate": 1.634266647323372e-05, "loss": 0.0007, "step": 23166 }, { "epoch": 2.511600173460538, "grad_norm": 0.025191888213157654, "learning_rate": 1.633903960539678e-05, "loss": 0.0007, "step": 23167 }, { "epoch": 2.5117085862966175, "grad_norm": 0.18462342023849487, "learning_rate": 1.6335412737559843e-05, "loss": 0.0059, "step": 23168 }, { "epoch": 2.511816999132697, "grad_norm": 0.45892488956451416, "learning_rate": 1.6331785869722907e-05, "loss": 0.017, "step": 23169 }, { "epoch": 2.511925411968777, "grad_norm": 0.04678056389093399, "learning_rate": 1.632815900188597e-05, "loss": 0.0021, "step": 23170 }, { "epoch": 2.512033824804857, "grad_norm": 0.17495915293693542, "learning_rate": 1.6324532134049036e-05, "loss": 0.0034, "step": 23171 }, { "epoch": 2.5121422376409366, "grad_norm": 0.5147162079811096, "learning_rate": 1.63209052662121e-05, "loss": 0.02, "step": 23172 }, { "epoch": 2.5122506504770166, "grad_norm": 0.03776179626584053, "learning_rate": 1.6317278398375164e-05, "loss": 0.0025, "step": 23173 }, { "epoch": 2.512359063313096, "grad_norm": 0.23995822668075562, "learning_rate": 1.6313651530538228e-05, "loss": 0.0158, "step": 23174 }, { "epoch": 2.512467476149176, "grad_norm": 0.1383472979068756, "learning_rate": 1.6310024662701292e-05, "loss": 0.0042, "step": 23175 }, { "epoch": 2.512575888985256, "grad_norm": 0.12182096391916275, "learning_rate": 1.6306397794864356e-05, "loss": 0.0044, "step": 23176 }, { "epoch": 2.5126843018213356, "grad_norm": 0.22178691625595093, "learning_rate": 1.630277092702742e-05, "loss": 0.0052, "step": 23177 }, { "epoch": 2.512792714657415, "grad_norm": 0.22910210490226746, "learning_rate": 1.629914405919048e-05, "loss": 0.0056, "step": 23178 }, { "epoch": 2.512901127493495, "grad_norm": 0.017086366191506386, "learning_rate": 1.629551719135355e-05, "loss": 0.0002, "step": 23179 }, { "epoch": 2.513009540329575, "grad_norm": 0.10584461688995361, "learning_rate": 1.6291890323516613e-05, "loss": 0.0061, "step": 23180 }, { "epoch": 2.5131179531656547, "grad_norm": 0.25218549370765686, "learning_rate": 1.6288263455679677e-05, "loss": 0.0125, "step": 23181 }, { "epoch": 2.5132263660017347, "grad_norm": 0.15593545138835907, "learning_rate": 1.6284636587842737e-05, "loss": 0.0063, "step": 23182 }, { "epoch": 2.513334778837814, "grad_norm": 0.09184374660253525, "learning_rate": 1.6281009720005805e-05, "loss": 0.0036, "step": 23183 }, { "epoch": 2.513443191673894, "grad_norm": 0.13887125253677368, "learning_rate": 1.627738285216887e-05, "loss": 0.0035, "step": 23184 }, { "epoch": 2.513551604509974, "grad_norm": 0.15227466821670532, "learning_rate": 1.6273755984331933e-05, "loss": 0.003, "step": 23185 }, { "epoch": 2.5136600173460537, "grad_norm": 0.0881679430603981, "learning_rate": 1.6270129116494994e-05, "loss": 0.0051, "step": 23186 }, { "epoch": 2.5137684301821337, "grad_norm": 0.8291332125663757, "learning_rate": 1.626650224865806e-05, "loss": 0.0188, "step": 23187 }, { "epoch": 2.5138768430182132, "grad_norm": 0.0634671002626419, "learning_rate": 1.6262875380821125e-05, "loss": 0.0019, "step": 23188 }, { "epoch": 2.513985255854293, "grad_norm": 0.9335357546806335, "learning_rate": 1.6259248512984186e-05, "loss": 0.0316, "step": 23189 }, { "epoch": 2.514093668690373, "grad_norm": 1.9136507511138916, "learning_rate": 1.625562164514725e-05, "loss": 0.0383, "step": 23190 }, { "epoch": 2.5142020815264527, "grad_norm": 0.3625932037830353, "learning_rate": 1.6251994777310314e-05, "loss": 0.0059, "step": 23191 }, { "epoch": 2.5143104943625323, "grad_norm": 0.24262572824954987, "learning_rate": 1.6248367909473382e-05, "loss": 0.0143, "step": 23192 }, { "epoch": 2.5144189071986123, "grad_norm": 0.09226365387439728, "learning_rate": 1.6244741041636443e-05, "loss": 0.0024, "step": 23193 }, { "epoch": 2.5145273200346923, "grad_norm": 0.32037118077278137, "learning_rate": 1.6241114173799507e-05, "loss": 0.0071, "step": 23194 }, { "epoch": 2.514635732870772, "grad_norm": 0.44275468587875366, "learning_rate": 1.623748730596257e-05, "loss": 0.037, "step": 23195 }, { "epoch": 2.514744145706852, "grad_norm": 0.1364677995443344, "learning_rate": 1.6233860438125635e-05, "loss": 0.0075, "step": 23196 }, { "epoch": 2.5148525585429313, "grad_norm": 0.3554658889770508, "learning_rate": 1.62302335702887e-05, "loss": 0.0194, "step": 23197 }, { "epoch": 2.5149609713790113, "grad_norm": 1.3698337078094482, "learning_rate": 1.6226606702451763e-05, "loss": 0.0224, "step": 23198 }, { "epoch": 2.5150693842150913, "grad_norm": 0.47189339995384216, "learning_rate": 1.6222979834614827e-05, "loss": 0.0233, "step": 23199 }, { "epoch": 2.515177797051171, "grad_norm": 0.21666426956653595, "learning_rate": 1.621935296677789e-05, "loss": 0.0057, "step": 23200 }, { "epoch": 2.5152862098872504, "grad_norm": 0.487396240234375, "learning_rate": 1.6215726098940955e-05, "loss": 0.0234, "step": 23201 }, { "epoch": 2.5153946227233304, "grad_norm": 0.1332075148820877, "learning_rate": 1.621209923110402e-05, "loss": 0.005, "step": 23202 }, { "epoch": 2.5155030355594103, "grad_norm": 0.2500932514667511, "learning_rate": 1.6208472363267084e-05, "loss": 0.0124, "step": 23203 }, { "epoch": 2.51561144839549, "grad_norm": 0.08704373240470886, "learning_rate": 1.6204845495430144e-05, "loss": 0.0019, "step": 23204 }, { "epoch": 2.51571986123157, "grad_norm": 0.5264275074005127, "learning_rate": 1.6201218627593212e-05, "loss": 0.018, "step": 23205 }, { "epoch": 2.5158282740676494, "grad_norm": 0.03536974638700485, "learning_rate": 1.6197591759756276e-05, "loss": 0.001, "step": 23206 }, { "epoch": 2.5159366869037294, "grad_norm": 0.02163458801805973, "learning_rate": 1.619396489191934e-05, "loss": 0.0006, "step": 23207 }, { "epoch": 2.5160450997398094, "grad_norm": 0.16956916451454163, "learning_rate": 1.61903380240824e-05, "loss": 0.0056, "step": 23208 }, { "epoch": 2.516153512575889, "grad_norm": 0.016115132719278336, "learning_rate": 1.6186711156245468e-05, "loss": 0.001, "step": 23209 }, { "epoch": 2.516261925411969, "grad_norm": 0.12139025330543518, "learning_rate": 1.6183084288408532e-05, "loss": 0.0018, "step": 23210 }, { "epoch": 2.5163703382480485, "grad_norm": 0.3880782723426819, "learning_rate": 1.6179457420571596e-05, "loss": 0.0113, "step": 23211 }, { "epoch": 2.5164787510841284, "grad_norm": 0.2786714732646942, "learning_rate": 1.6175830552734657e-05, "loss": 0.0592, "step": 23212 }, { "epoch": 2.5165871639202084, "grad_norm": 0.10789816081523895, "learning_rate": 1.6172203684897725e-05, "loss": 0.0032, "step": 23213 }, { "epoch": 2.516695576756288, "grad_norm": 0.44049593806266785, "learning_rate": 1.616857681706079e-05, "loss": 0.0268, "step": 23214 }, { "epoch": 2.5168039895923675, "grad_norm": 0.17893247306346893, "learning_rate": 1.616494994922385e-05, "loss": 0.0087, "step": 23215 }, { "epoch": 2.5169124024284475, "grad_norm": 1.4329063892364502, "learning_rate": 1.6161323081386914e-05, "loss": 0.0389, "step": 23216 }, { "epoch": 2.5170208152645275, "grad_norm": 0.20195430517196655, "learning_rate": 1.6157696213549978e-05, "loss": 0.0067, "step": 23217 }, { "epoch": 2.517129228100607, "grad_norm": 0.1462588906288147, "learning_rate": 1.6154069345713045e-05, "loss": 0.0042, "step": 23218 }, { "epoch": 2.517237640936687, "grad_norm": 0.45476529002189636, "learning_rate": 1.6150442477876106e-05, "loss": 0.0444, "step": 23219 }, { "epoch": 2.5173460537727665, "grad_norm": 0.01951124146580696, "learning_rate": 1.614681561003917e-05, "loss": 0.001, "step": 23220 }, { "epoch": 2.5174544666088465, "grad_norm": 0.08894813805818558, "learning_rate": 1.6143188742202234e-05, "loss": 0.006, "step": 23221 }, { "epoch": 2.5175628794449265, "grad_norm": 0.02168564684689045, "learning_rate": 1.6139561874365298e-05, "loss": 0.0008, "step": 23222 }, { "epoch": 2.517671292281006, "grad_norm": 0.2966763377189636, "learning_rate": 1.6135935006528362e-05, "loss": 0.0088, "step": 23223 }, { "epoch": 2.5177797051170856, "grad_norm": 0.30151546001434326, "learning_rate": 1.6132308138691426e-05, "loss": 0.0176, "step": 23224 }, { "epoch": 2.5178881179531656, "grad_norm": 0.6546090841293335, "learning_rate": 1.612868127085449e-05, "loss": 0.02, "step": 23225 }, { "epoch": 2.5179965307892456, "grad_norm": 0.6841340661048889, "learning_rate": 1.6125054403017555e-05, "loss": 0.0158, "step": 23226 }, { "epoch": 2.518104943625325, "grad_norm": 0.24079899489879608, "learning_rate": 1.612142753518062e-05, "loss": 0.03, "step": 23227 }, { "epoch": 2.518213356461405, "grad_norm": 0.4019692540168762, "learning_rate": 1.6117800667343683e-05, "loss": 0.0143, "step": 23228 }, { "epoch": 2.5183217692974846, "grad_norm": 0.06358475983142853, "learning_rate": 1.6114173799506747e-05, "loss": 0.0019, "step": 23229 }, { "epoch": 2.5184301821335646, "grad_norm": 0.1232016310095787, "learning_rate": 1.611054693166981e-05, "loss": 0.0069, "step": 23230 }, { "epoch": 2.5185385949696446, "grad_norm": 0.30661460757255554, "learning_rate": 1.6106920063832875e-05, "loss": 0.0063, "step": 23231 }, { "epoch": 2.518647007805724, "grad_norm": 0.028686903417110443, "learning_rate": 1.610329319599594e-05, "loss": 0.0005, "step": 23232 }, { "epoch": 2.518755420641804, "grad_norm": 0.4288885295391083, "learning_rate": 1.6099666328159003e-05, "loss": 0.0136, "step": 23233 }, { "epoch": 2.5188638334778837, "grad_norm": 0.1260562688112259, "learning_rate": 1.6096039460322064e-05, "loss": 0.0043, "step": 23234 }, { "epoch": 2.5189722463139637, "grad_norm": 0.1664419174194336, "learning_rate": 1.609241259248513e-05, "loss": 0.0032, "step": 23235 }, { "epoch": 2.519080659150043, "grad_norm": 0.00830740388482809, "learning_rate": 1.6088785724648196e-05, "loss": 0.0003, "step": 23236 }, { "epoch": 2.519189071986123, "grad_norm": 0.15982690453529358, "learning_rate": 1.608515885681126e-05, "loss": 0.0118, "step": 23237 }, { "epoch": 2.5192974848222027, "grad_norm": 0.12367314100265503, "learning_rate": 1.608153198897432e-05, "loss": 0.0011, "step": 23238 }, { "epoch": 2.5194058976582827, "grad_norm": 0.37028297781944275, "learning_rate": 1.6077905121137388e-05, "loss": 0.0281, "step": 23239 }, { "epoch": 2.5195143104943627, "grad_norm": 0.15162019431591034, "learning_rate": 1.6074278253300452e-05, "loss": 0.012, "step": 23240 }, { "epoch": 2.5196227233304422, "grad_norm": 0.03320207819342613, "learning_rate": 1.6070651385463513e-05, "loss": 0.001, "step": 23241 }, { "epoch": 2.5197311361665222, "grad_norm": 0.14403392374515533, "learning_rate": 1.6067024517626577e-05, "loss": 0.0034, "step": 23242 }, { "epoch": 2.5198395490026018, "grad_norm": 0.3809419870376587, "learning_rate": 1.6063397649789644e-05, "loss": 0.0107, "step": 23243 }, { "epoch": 2.5199479618386817, "grad_norm": 0.14571958780288696, "learning_rate": 1.605977078195271e-05, "loss": 0.0013, "step": 23244 }, { "epoch": 2.5200563746747617, "grad_norm": 0.2142760008573532, "learning_rate": 1.605614391411577e-05, "loss": 0.05, "step": 23245 }, { "epoch": 2.5201647875108413, "grad_norm": 0.25937607884407043, "learning_rate": 1.6052517046278833e-05, "loss": 0.0221, "step": 23246 }, { "epoch": 2.520273200346921, "grad_norm": 0.11447194218635559, "learning_rate": 1.6048890178441897e-05, "loss": 0.008, "step": 23247 }, { "epoch": 2.520381613183001, "grad_norm": 0.08958912640810013, "learning_rate": 1.604526331060496e-05, "loss": 0.0022, "step": 23248 }, { "epoch": 2.520490026019081, "grad_norm": 0.007946720346808434, "learning_rate": 1.6041636442768026e-05, "loss": 0.0003, "step": 23249 }, { "epoch": 2.5205984388551603, "grad_norm": 0.5285903215408325, "learning_rate": 1.603800957493109e-05, "loss": 0.0261, "step": 23250 }, { "epoch": 2.5207068516912403, "grad_norm": 0.2940058708190918, "learning_rate": 1.6034382707094154e-05, "loss": 0.0049, "step": 23251 }, { "epoch": 2.52081526452732, "grad_norm": 0.06968862563371658, "learning_rate": 1.6030755839257218e-05, "loss": 0.0017, "step": 23252 }, { "epoch": 2.5209236773634, "grad_norm": 0.055867668241262436, "learning_rate": 1.6027128971420282e-05, "loss": 0.0033, "step": 23253 }, { "epoch": 2.52103209019948, "grad_norm": 0.3566065728664398, "learning_rate": 1.6023502103583346e-05, "loss": 0.0091, "step": 23254 }, { "epoch": 2.5211405030355594, "grad_norm": 0.002076550154015422, "learning_rate": 1.601987523574641e-05, "loss": 0.0001, "step": 23255 }, { "epoch": 2.5212489158716394, "grad_norm": 0.052805639803409576, "learning_rate": 1.6016248367909474e-05, "loss": 0.0015, "step": 23256 }, { "epoch": 2.521357328707719, "grad_norm": 0.06141383945941925, "learning_rate": 1.601262150007254e-05, "loss": 0.0037, "step": 23257 }, { "epoch": 2.521465741543799, "grad_norm": 0.5560550093650818, "learning_rate": 1.6008994632235603e-05, "loss": 0.0305, "step": 23258 }, { "epoch": 2.5215741543798784, "grad_norm": 0.02734651416540146, "learning_rate": 1.6005367764398667e-05, "loss": 0.001, "step": 23259 }, { "epoch": 2.5216825672159584, "grad_norm": 0.07414039224386215, "learning_rate": 1.6001740896561727e-05, "loss": 0.0026, "step": 23260 }, { "epoch": 2.521790980052038, "grad_norm": 0.09509257972240448, "learning_rate": 1.5998114028724795e-05, "loss": 0.0047, "step": 23261 }, { "epoch": 2.521899392888118, "grad_norm": 0.06693628430366516, "learning_rate": 1.599448716088786e-05, "loss": 0.0021, "step": 23262 }, { "epoch": 2.522007805724198, "grad_norm": 0.15853670239448547, "learning_rate": 1.5990860293050923e-05, "loss": 0.0066, "step": 23263 }, { "epoch": 2.5221162185602775, "grad_norm": 0.30240556597709656, "learning_rate": 1.5987233425213984e-05, "loss": 0.0319, "step": 23264 }, { "epoch": 2.5222246313963574, "grad_norm": 0.031011324375867844, "learning_rate": 1.598360655737705e-05, "loss": 0.0005, "step": 23265 }, { "epoch": 2.522333044232437, "grad_norm": 0.07441212236881256, "learning_rate": 1.5979979689540115e-05, "loss": 0.0026, "step": 23266 }, { "epoch": 2.522441457068517, "grad_norm": 0.34483101963996887, "learning_rate": 1.5976352821703176e-05, "loss": 0.016, "step": 23267 }, { "epoch": 2.522549869904597, "grad_norm": 0.9453844428062439, "learning_rate": 1.597272595386624e-05, "loss": 0.0381, "step": 23268 }, { "epoch": 2.5226582827406765, "grad_norm": 0.8463857173919678, "learning_rate": 1.5969099086029308e-05, "loss": 0.0426, "step": 23269 }, { "epoch": 2.522766695576756, "grad_norm": 0.20150913298130035, "learning_rate": 1.5965472218192372e-05, "loss": 0.0232, "step": 23270 }, { "epoch": 2.522875108412836, "grad_norm": 0.20013149082660675, "learning_rate": 1.5961845350355433e-05, "loss": 0.0026, "step": 23271 }, { "epoch": 2.522983521248916, "grad_norm": 0.015006804838776588, "learning_rate": 1.5958218482518497e-05, "loss": 0.0007, "step": 23272 }, { "epoch": 2.5230919340849955, "grad_norm": 0.04881008714437485, "learning_rate": 1.595459161468156e-05, "loss": 0.0029, "step": 23273 }, { "epoch": 2.5232003469210755, "grad_norm": 0.5434913635253906, "learning_rate": 1.5950964746844628e-05, "loss": 0.0353, "step": 23274 }, { "epoch": 2.523308759757155, "grad_norm": 0.4339800477027893, "learning_rate": 1.594733787900769e-05, "loss": 0.0113, "step": 23275 }, { "epoch": 2.523417172593235, "grad_norm": 0.5658507943153381, "learning_rate": 1.5943711011170753e-05, "loss": 0.0206, "step": 23276 }, { "epoch": 2.523525585429315, "grad_norm": 0.7158030867576599, "learning_rate": 1.5940084143333817e-05, "loss": 0.0178, "step": 23277 }, { "epoch": 2.5236339982653946, "grad_norm": 0.31337544322013855, "learning_rate": 1.593645727549688e-05, "loss": 0.0264, "step": 23278 }, { "epoch": 2.5237424111014746, "grad_norm": 0.1233946830034256, "learning_rate": 1.5932830407659945e-05, "loss": 0.0035, "step": 23279 }, { "epoch": 2.523850823937554, "grad_norm": 0.6428635716438293, "learning_rate": 1.592920353982301e-05, "loss": 0.0318, "step": 23280 }, { "epoch": 2.523959236773634, "grad_norm": 0.4661742150783539, "learning_rate": 1.5925576671986074e-05, "loss": 0.023, "step": 23281 }, { "epoch": 2.5240676496097136, "grad_norm": 0.04945961758494377, "learning_rate": 1.5921949804149138e-05, "loss": 0.0009, "step": 23282 }, { "epoch": 2.5241760624457936, "grad_norm": 0.2026444524526596, "learning_rate": 1.5918322936312202e-05, "loss": 0.0082, "step": 23283 }, { "epoch": 2.524284475281873, "grad_norm": 0.2233152985572815, "learning_rate": 1.5914696068475266e-05, "loss": 0.0101, "step": 23284 }, { "epoch": 2.524392888117953, "grad_norm": 0.49913308024406433, "learning_rate": 1.591106920063833e-05, "loss": 0.0181, "step": 23285 }, { "epoch": 2.524501300954033, "grad_norm": 0.04960649833083153, "learning_rate": 1.590744233280139e-05, "loss": 0.0027, "step": 23286 }, { "epoch": 2.5246097137901127, "grad_norm": 0.0799703299999237, "learning_rate": 1.5903815464964458e-05, "loss": 0.0016, "step": 23287 }, { "epoch": 2.5247181266261927, "grad_norm": 0.0272109005600214, "learning_rate": 1.5900188597127522e-05, "loss": 0.0009, "step": 23288 }, { "epoch": 2.524826539462272, "grad_norm": 0.1456470787525177, "learning_rate": 1.5896561729290586e-05, "loss": 0.0038, "step": 23289 }, { "epoch": 2.524934952298352, "grad_norm": 0.36289602518081665, "learning_rate": 1.5892934861453647e-05, "loss": 0.0071, "step": 23290 }, { "epoch": 2.525043365134432, "grad_norm": 0.08102882653474808, "learning_rate": 1.5889307993616715e-05, "loss": 0.0055, "step": 23291 }, { "epoch": 2.5251517779705117, "grad_norm": 0.1317683905363083, "learning_rate": 1.588568112577978e-05, "loss": 0.0054, "step": 23292 }, { "epoch": 2.5252601908065913, "grad_norm": 0.02903061918914318, "learning_rate": 1.588205425794284e-05, "loss": 0.0014, "step": 23293 }, { "epoch": 2.5253686036426712, "grad_norm": 0.5985320210456848, "learning_rate": 1.5878427390105904e-05, "loss": 0.0251, "step": 23294 }, { "epoch": 2.5254770164787512, "grad_norm": 0.5925419926643372, "learning_rate": 1.587480052226897e-05, "loss": 0.0255, "step": 23295 }, { "epoch": 2.5255854293148308, "grad_norm": 0.016453679651021957, "learning_rate": 1.5871173654432035e-05, "loss": 0.0007, "step": 23296 }, { "epoch": 2.5256938421509108, "grad_norm": 0.22322498261928558, "learning_rate": 1.5867546786595096e-05, "loss": 0.0086, "step": 23297 }, { "epoch": 2.5258022549869903, "grad_norm": 0.05426068231463432, "learning_rate": 1.586391991875816e-05, "loss": 0.0012, "step": 23298 }, { "epoch": 2.5259106678230703, "grad_norm": 0.10382247716188431, "learning_rate": 1.5860293050921224e-05, "loss": 0.0024, "step": 23299 }, { "epoch": 2.5260190806591503, "grad_norm": 0.1161779910326004, "learning_rate": 1.585666618308429e-05, "loss": 0.0032, "step": 23300 }, { "epoch": 2.52612749349523, "grad_norm": 0.4115673005580902, "learning_rate": 1.5853039315247352e-05, "loss": 0.016, "step": 23301 }, { "epoch": 2.52623590633131, "grad_norm": 0.051200222223997116, "learning_rate": 1.5849412447410416e-05, "loss": 0.0011, "step": 23302 }, { "epoch": 2.5263443191673893, "grad_norm": 0.2841026484966278, "learning_rate": 1.584578557957348e-05, "loss": 0.0135, "step": 23303 }, { "epoch": 2.5264527320034693, "grad_norm": 0.011283042840659618, "learning_rate": 1.5842158711736545e-05, "loss": 0.0004, "step": 23304 }, { "epoch": 2.526561144839549, "grad_norm": 0.0650472640991211, "learning_rate": 1.583853184389961e-05, "loss": 0.0008, "step": 23305 }, { "epoch": 2.526669557675629, "grad_norm": 0.04643440991640091, "learning_rate": 1.5834904976062673e-05, "loss": 0.0018, "step": 23306 }, { "epoch": 2.5267779705117084, "grad_norm": 0.11855518817901611, "learning_rate": 1.5831278108225737e-05, "loss": 0.0048, "step": 23307 }, { "epoch": 2.5268863833477884, "grad_norm": 0.12645204365253448, "learning_rate": 1.58276512403888e-05, "loss": 0.0034, "step": 23308 }, { "epoch": 2.5269947961838684, "grad_norm": 0.0475655198097229, "learning_rate": 1.5824024372551865e-05, "loss": 0.0009, "step": 23309 }, { "epoch": 2.527103209019948, "grad_norm": 0.02244686707854271, "learning_rate": 1.582039750471493e-05, "loss": 0.0013, "step": 23310 }, { "epoch": 2.527211621856028, "grad_norm": 0.9598869681358337, "learning_rate": 1.5816770636877993e-05, "loss": 0.0735, "step": 23311 }, { "epoch": 2.5273200346921074, "grad_norm": 0.028607213869690895, "learning_rate": 1.5813143769041054e-05, "loss": 0.0018, "step": 23312 }, { "epoch": 2.5274284475281874, "grad_norm": 0.20133593678474426, "learning_rate": 1.580951690120412e-05, "loss": 0.0327, "step": 23313 }, { "epoch": 2.5275368603642674, "grad_norm": 0.2211133986711502, "learning_rate": 1.5805890033367186e-05, "loss": 0.0053, "step": 23314 }, { "epoch": 2.527645273200347, "grad_norm": 0.1675684154033661, "learning_rate": 1.580226316553025e-05, "loss": 0.0026, "step": 23315 }, { "epoch": 2.5277536860364265, "grad_norm": 0.3795546591281891, "learning_rate": 1.579863629769331e-05, "loss": 0.017, "step": 23316 }, { "epoch": 2.5278620988725065, "grad_norm": 0.1512807011604309, "learning_rate": 1.5795009429856378e-05, "loss": 0.0027, "step": 23317 }, { "epoch": 2.5279705117085864, "grad_norm": 0.024053266271948814, "learning_rate": 1.5791382562019442e-05, "loss": 0.0006, "step": 23318 }, { "epoch": 2.528078924544666, "grad_norm": 0.1599905788898468, "learning_rate": 1.5787755694182503e-05, "loss": 0.0065, "step": 23319 }, { "epoch": 2.528187337380746, "grad_norm": 0.0289156511425972, "learning_rate": 1.5784128826345567e-05, "loss": 0.0004, "step": 23320 }, { "epoch": 2.5282957502168255, "grad_norm": 0.4477823078632355, "learning_rate": 1.5780501958508634e-05, "loss": 0.0077, "step": 23321 }, { "epoch": 2.5284041630529055, "grad_norm": 0.23095539212226868, "learning_rate": 1.57768750906717e-05, "loss": 0.0034, "step": 23322 }, { "epoch": 2.5285125758889855, "grad_norm": 0.22152051329612732, "learning_rate": 1.577324822283476e-05, "loss": 0.014, "step": 23323 }, { "epoch": 2.528620988725065, "grad_norm": 0.7777047157287598, "learning_rate": 1.5769621354997823e-05, "loss": 0.0108, "step": 23324 }, { "epoch": 2.5287294015611446, "grad_norm": 0.5478969812393188, "learning_rate": 1.5765994487160887e-05, "loss": 0.0096, "step": 23325 }, { "epoch": 2.5288378143972245, "grad_norm": 0.4280600845813751, "learning_rate": 1.5762367619323955e-05, "loss": 0.0179, "step": 23326 }, { "epoch": 2.5289462272333045, "grad_norm": 0.29276853799819946, "learning_rate": 1.5758740751487016e-05, "loss": 0.0068, "step": 23327 }, { "epoch": 2.529054640069384, "grad_norm": 0.08191211521625519, "learning_rate": 1.575511388365008e-05, "loss": 0.0009, "step": 23328 }, { "epoch": 2.529163052905464, "grad_norm": 0.28002363443374634, "learning_rate": 1.5751487015813144e-05, "loss": 0.0109, "step": 23329 }, { "epoch": 2.5292714657415436, "grad_norm": 0.01873757690191269, "learning_rate": 1.5747860147976208e-05, "loss": 0.0006, "step": 23330 }, { "epoch": 2.5293798785776236, "grad_norm": 0.07446684688329697, "learning_rate": 1.5744233280139272e-05, "loss": 0.0024, "step": 23331 }, { "epoch": 2.5294882914137036, "grad_norm": 0.019081443548202515, "learning_rate": 1.5740606412302336e-05, "loss": 0.0008, "step": 23332 }, { "epoch": 2.529596704249783, "grad_norm": 0.4379786550998688, "learning_rate": 1.57369795444654e-05, "loss": 0.0327, "step": 23333 }, { "epoch": 2.529705117085863, "grad_norm": 0.23870393633842468, "learning_rate": 1.5733352676628464e-05, "loss": 0.0057, "step": 23334 }, { "epoch": 2.5298135299219426, "grad_norm": 0.3737412393093109, "learning_rate": 1.572972580879153e-05, "loss": 0.006, "step": 23335 }, { "epoch": 2.5299219427580226, "grad_norm": 0.23456864058971405, "learning_rate": 1.5726098940954593e-05, "loss": 0.0027, "step": 23336 }, { "epoch": 2.5300303555941026, "grad_norm": 0.21575412154197693, "learning_rate": 1.5722472073117657e-05, "loss": 0.0107, "step": 23337 }, { "epoch": 2.530138768430182, "grad_norm": 0.6081791520118713, "learning_rate": 1.571884520528072e-05, "loss": 0.0152, "step": 23338 }, { "epoch": 2.5302471812662617, "grad_norm": 0.009540209546685219, "learning_rate": 1.5715218337443785e-05, "loss": 0.0004, "step": 23339 }, { "epoch": 2.5303555941023417, "grad_norm": 0.26505765318870544, "learning_rate": 1.571159146960685e-05, "loss": 0.0105, "step": 23340 }, { "epoch": 2.5304640069384217, "grad_norm": 0.2840334475040436, "learning_rate": 1.5707964601769913e-05, "loss": 0.0079, "step": 23341 }, { "epoch": 2.530572419774501, "grad_norm": 0.8461410403251648, "learning_rate": 1.5704337733932974e-05, "loss": 0.0372, "step": 23342 }, { "epoch": 2.530680832610581, "grad_norm": 0.5994271636009216, "learning_rate": 1.570071086609604e-05, "loss": 0.0182, "step": 23343 }, { "epoch": 2.5307892454466607, "grad_norm": 1.2013899087905884, "learning_rate": 1.5697083998259105e-05, "loss": 0.0084, "step": 23344 }, { "epoch": 2.5308976582827407, "grad_norm": 0.030550764873623848, "learning_rate": 1.5693457130422166e-05, "loss": 0.0006, "step": 23345 }, { "epoch": 2.5310060711188207, "grad_norm": 0.0666256695985794, "learning_rate": 1.568983026258523e-05, "loss": 0.0023, "step": 23346 }, { "epoch": 2.5311144839549002, "grad_norm": 0.21505329012870789, "learning_rate": 1.5686203394748298e-05, "loss": 0.0075, "step": 23347 }, { "epoch": 2.53122289679098, "grad_norm": 0.38947588205337524, "learning_rate": 1.5682576526911362e-05, "loss": 0.0339, "step": 23348 }, { "epoch": 2.5313313096270598, "grad_norm": 0.0878186747431755, "learning_rate": 1.5678949659074423e-05, "loss": 0.0032, "step": 23349 }, { "epoch": 2.5314397224631398, "grad_norm": 0.8904933929443359, "learning_rate": 1.5675322791237487e-05, "loss": 0.0112, "step": 23350 }, { "epoch": 2.5315481352992193, "grad_norm": 0.27642419934272766, "learning_rate": 1.5671695923400554e-05, "loss": 0.0106, "step": 23351 }, { "epoch": 2.5316565481352993, "grad_norm": 0.290088951587677, "learning_rate": 1.566806905556362e-05, "loss": 0.0081, "step": 23352 }, { "epoch": 2.531764960971379, "grad_norm": 0.32963305711746216, "learning_rate": 1.566444218772668e-05, "loss": 0.0141, "step": 23353 }, { "epoch": 2.531873373807459, "grad_norm": 0.2376202791929245, "learning_rate": 1.5660815319889743e-05, "loss": 0.0045, "step": 23354 }, { "epoch": 2.531981786643539, "grad_norm": 0.004899779800325632, "learning_rate": 1.5657188452052807e-05, "loss": 0.0001, "step": 23355 }, { "epoch": 2.5320901994796183, "grad_norm": 0.6510080695152283, "learning_rate": 1.565356158421587e-05, "loss": 0.0263, "step": 23356 }, { "epoch": 2.5321986123156983, "grad_norm": 0.18372327089309692, "learning_rate": 1.5649934716378935e-05, "loss": 0.0071, "step": 23357 }, { "epoch": 2.532307025151778, "grad_norm": 1.403581976890564, "learning_rate": 1.5646307848542e-05, "loss": 0.0529, "step": 23358 }, { "epoch": 2.532415437987858, "grad_norm": 0.2146555632352829, "learning_rate": 1.5642680980705064e-05, "loss": 0.0067, "step": 23359 }, { "epoch": 2.532523850823938, "grad_norm": 0.12126027047634125, "learning_rate": 1.5639054112868128e-05, "loss": 0.002, "step": 23360 }, { "epoch": 2.5326322636600174, "grad_norm": 0.4387345016002655, "learning_rate": 1.5635427245031192e-05, "loss": 0.0544, "step": 23361 }, { "epoch": 2.532740676496097, "grad_norm": 0.0957263931632042, "learning_rate": 1.5631800377194256e-05, "loss": 0.0041, "step": 23362 }, { "epoch": 2.532849089332177, "grad_norm": 0.5197587609291077, "learning_rate": 1.562817350935732e-05, "loss": 0.0208, "step": 23363 }, { "epoch": 2.532957502168257, "grad_norm": 0.0718155950307846, "learning_rate": 1.5624546641520384e-05, "loss": 0.0016, "step": 23364 }, { "epoch": 2.5330659150043364, "grad_norm": 0.25734323263168335, "learning_rate": 1.5620919773683448e-05, "loss": 0.0175, "step": 23365 }, { "epoch": 2.5331743278404164, "grad_norm": 0.344462126493454, "learning_rate": 1.5617292905846512e-05, "loss": 0.0074, "step": 23366 }, { "epoch": 2.533282740676496, "grad_norm": 0.11172907054424286, "learning_rate": 1.5613666038009577e-05, "loss": 0.0037, "step": 23367 }, { "epoch": 2.533391153512576, "grad_norm": 0.0014516919618472457, "learning_rate": 1.5610039170172637e-05, "loss": 0.0001, "step": 23368 }, { "epoch": 2.533499566348656, "grad_norm": 0.2678016126155853, "learning_rate": 1.5606412302335705e-05, "loss": 0.0095, "step": 23369 }, { "epoch": 2.5336079791847355, "grad_norm": 0.29372289776802063, "learning_rate": 1.560278543449877e-05, "loss": 0.0231, "step": 23370 }, { "epoch": 2.533716392020815, "grad_norm": 0.416070818901062, "learning_rate": 1.559915856666183e-05, "loss": 0.0125, "step": 23371 }, { "epoch": 2.533824804856895, "grad_norm": 0.2413017898797989, "learning_rate": 1.5595531698824894e-05, "loss": 0.0088, "step": 23372 }, { "epoch": 2.533933217692975, "grad_norm": 0.008693465031683445, "learning_rate": 1.559190483098796e-05, "loss": 0.0003, "step": 23373 }, { "epoch": 2.5340416305290545, "grad_norm": 0.5353596806526184, "learning_rate": 1.5588277963151025e-05, "loss": 0.034, "step": 23374 }, { "epoch": 2.5341500433651345, "grad_norm": 0.02338857203722, "learning_rate": 1.5584651095314086e-05, "loss": 0.0005, "step": 23375 }, { "epoch": 2.534258456201214, "grad_norm": 0.18934616446495056, "learning_rate": 1.558102422747715e-05, "loss": 0.0068, "step": 23376 }, { "epoch": 2.534366869037294, "grad_norm": 0.05252350866794586, "learning_rate": 1.5577397359640218e-05, "loss": 0.0034, "step": 23377 }, { "epoch": 2.534475281873374, "grad_norm": 0.91739821434021, "learning_rate": 1.557377049180328e-05, "loss": 0.0135, "step": 23378 }, { "epoch": 2.5345836947094535, "grad_norm": 0.5096384882926941, "learning_rate": 1.5570143623966342e-05, "loss": 0.0139, "step": 23379 }, { "epoch": 2.5346921075455335, "grad_norm": 0.2798953056335449, "learning_rate": 1.5566516756129406e-05, "loss": 0.0169, "step": 23380 }, { "epoch": 2.534800520381613, "grad_norm": 0.7482418417930603, "learning_rate": 1.556288988829247e-05, "loss": 0.0188, "step": 23381 }, { "epoch": 2.534908933217693, "grad_norm": 0.24364033341407776, "learning_rate": 1.5559263020455535e-05, "loss": 0.0317, "step": 23382 }, { "epoch": 2.5350173460537726, "grad_norm": 0.1741611361503601, "learning_rate": 1.55556361526186e-05, "loss": 0.0126, "step": 23383 }, { "epoch": 2.5351257588898526, "grad_norm": 0.007132337894290686, "learning_rate": 1.5552009284781663e-05, "loss": 0.0002, "step": 23384 }, { "epoch": 2.535234171725932, "grad_norm": 0.07838258892297745, "learning_rate": 1.5548382416944727e-05, "loss": 0.0033, "step": 23385 }, { "epoch": 2.535342584562012, "grad_norm": 0.025911344215273857, "learning_rate": 1.554475554910779e-05, "loss": 0.001, "step": 23386 }, { "epoch": 2.535450997398092, "grad_norm": 0.019618090242147446, "learning_rate": 1.5541128681270855e-05, "loss": 0.0004, "step": 23387 }, { "epoch": 2.5355594102341716, "grad_norm": 0.33378803730010986, "learning_rate": 1.553750181343392e-05, "loss": 0.0095, "step": 23388 }, { "epoch": 2.5356678230702516, "grad_norm": 0.13997283577919006, "learning_rate": 1.5533874945596983e-05, "loss": 0.0067, "step": 23389 }, { "epoch": 2.535776235906331, "grad_norm": 0.0741022452712059, "learning_rate": 1.5530248077760048e-05, "loss": 0.0027, "step": 23390 }, { "epoch": 2.535884648742411, "grad_norm": 0.06628310680389404, "learning_rate": 1.552662120992311e-05, "loss": 0.0022, "step": 23391 }, { "epoch": 2.535993061578491, "grad_norm": 0.5292840600013733, "learning_rate": 1.5522994342086176e-05, "loss": 0.0354, "step": 23392 }, { "epoch": 2.5361014744145707, "grad_norm": 1.288217544555664, "learning_rate": 1.551936747424924e-05, "loss": 0.0298, "step": 23393 }, { "epoch": 2.53620988725065, "grad_norm": 0.015918763354420662, "learning_rate": 1.55157406064123e-05, "loss": 0.0005, "step": 23394 }, { "epoch": 2.53631830008673, "grad_norm": 0.3245273530483246, "learning_rate": 1.5512113738575368e-05, "loss": 0.0178, "step": 23395 }, { "epoch": 2.53642671292281, "grad_norm": 0.1321428120136261, "learning_rate": 1.5508486870738432e-05, "loss": 0.0054, "step": 23396 }, { "epoch": 2.5365351257588897, "grad_norm": 0.28304076194763184, "learning_rate": 1.5504860002901493e-05, "loss": 0.0071, "step": 23397 }, { "epoch": 2.5366435385949697, "grad_norm": 0.20115533471107483, "learning_rate": 1.5501233135064557e-05, "loss": 0.0079, "step": 23398 }, { "epoch": 2.5367519514310493, "grad_norm": 0.042425017803907394, "learning_rate": 1.5497606267227624e-05, "loss": 0.0018, "step": 23399 }, { "epoch": 2.5368603642671292, "grad_norm": 0.30347418785095215, "learning_rate": 1.549397939939069e-05, "loss": 0.0243, "step": 23400 }, { "epoch": 2.5369687771032092, "grad_norm": 0.6206626892089844, "learning_rate": 1.549035253155375e-05, "loss": 0.0157, "step": 23401 }, { "epoch": 2.5370771899392888, "grad_norm": 0.42159324884414673, "learning_rate": 1.5486725663716813e-05, "loss": 0.0082, "step": 23402 }, { "epoch": 2.5371856027753688, "grad_norm": 0.22353024780750275, "learning_rate": 1.548309879587988e-05, "loss": 0.0078, "step": 23403 }, { "epoch": 2.5372940156114483, "grad_norm": 0.08684898912906647, "learning_rate": 1.5479471928042945e-05, "loss": 0.0029, "step": 23404 }, { "epoch": 2.5374024284475283, "grad_norm": 0.0954294353723526, "learning_rate": 1.5475845060206006e-05, "loss": 0.0015, "step": 23405 }, { "epoch": 2.537510841283608, "grad_norm": 0.06260640174150467, "learning_rate": 1.547221819236907e-05, "loss": 0.0018, "step": 23406 }, { "epoch": 2.537619254119688, "grad_norm": 0.17575255036354065, "learning_rate": 1.5468591324532134e-05, "loss": 0.0098, "step": 23407 }, { "epoch": 2.5377276669557673, "grad_norm": 1.3409184217453003, "learning_rate": 1.5464964456695198e-05, "loss": 0.0242, "step": 23408 }, { "epoch": 2.5378360797918473, "grad_norm": 0.244790181517601, "learning_rate": 1.5461337588858262e-05, "loss": 0.0141, "step": 23409 }, { "epoch": 2.5379444926279273, "grad_norm": 0.0623084120452404, "learning_rate": 1.5457710721021326e-05, "loss": 0.0019, "step": 23410 }, { "epoch": 2.538052905464007, "grad_norm": 0.21016433835029602, "learning_rate": 1.545408385318439e-05, "loss": 0.0044, "step": 23411 }, { "epoch": 2.538161318300087, "grad_norm": 0.3493812382221222, "learning_rate": 1.5450456985347454e-05, "loss": 0.0094, "step": 23412 }, { "epoch": 2.5382697311361664, "grad_norm": 0.09911726415157318, "learning_rate": 1.544683011751052e-05, "loss": 0.0009, "step": 23413 }, { "epoch": 2.5383781439722464, "grad_norm": 0.5152121782302856, "learning_rate": 1.5443203249673583e-05, "loss": 0.043, "step": 23414 }, { "epoch": 2.5384865568083264, "grad_norm": 0.008667162619531155, "learning_rate": 1.5439576381836647e-05, "loss": 0.0004, "step": 23415 }, { "epoch": 2.538594969644406, "grad_norm": 0.535169243812561, "learning_rate": 1.543594951399971e-05, "loss": 0.0186, "step": 23416 }, { "epoch": 2.5387033824804854, "grad_norm": 0.06050395593047142, "learning_rate": 1.5432322646162775e-05, "loss": 0.001, "step": 23417 }, { "epoch": 2.5388117953165654, "grad_norm": 0.4793457090854645, "learning_rate": 1.542869577832584e-05, "loss": 0.0155, "step": 23418 }, { "epoch": 2.5389202081526454, "grad_norm": 0.06181543320417404, "learning_rate": 1.5425068910488903e-05, "loss": 0.0019, "step": 23419 }, { "epoch": 2.539028620988725, "grad_norm": 0.14784210920333862, "learning_rate": 1.5421442042651964e-05, "loss": 0.0055, "step": 23420 }, { "epoch": 2.539137033824805, "grad_norm": 0.3128909766674042, "learning_rate": 1.541781517481503e-05, "loss": 0.0292, "step": 23421 }, { "epoch": 2.5392454466608845, "grad_norm": 0.03602016344666481, "learning_rate": 1.5414188306978096e-05, "loss": 0.0011, "step": 23422 }, { "epoch": 2.5393538594969645, "grad_norm": 0.03742838650941849, "learning_rate": 1.5410561439141156e-05, "loss": 0.0012, "step": 23423 }, { "epoch": 2.5394622723330444, "grad_norm": 0.49832314252853394, "learning_rate": 1.540693457130422e-05, "loss": 0.0202, "step": 23424 }, { "epoch": 2.539570685169124, "grad_norm": 1.2278907299041748, "learning_rate": 1.5403307703467288e-05, "loss": 0.0203, "step": 23425 }, { "epoch": 2.539679098005204, "grad_norm": 0.00836259312927723, "learning_rate": 1.5399680835630352e-05, "loss": 0.0002, "step": 23426 }, { "epoch": 2.5397875108412835, "grad_norm": 0.30716899037361145, "learning_rate": 1.5396053967793413e-05, "loss": 0.0123, "step": 23427 }, { "epoch": 2.5398959236773635, "grad_norm": 0.25804129242897034, "learning_rate": 1.5392427099956477e-05, "loss": 0.0066, "step": 23428 }, { "epoch": 2.540004336513443, "grad_norm": 0.06175188347697258, "learning_rate": 1.5388800232119544e-05, "loss": 0.0027, "step": 23429 }, { "epoch": 2.540112749349523, "grad_norm": 0.37226420640945435, "learning_rate": 1.538517336428261e-05, "loss": 0.019, "step": 23430 }, { "epoch": 2.5402211621856026, "grad_norm": 0.3352978825569153, "learning_rate": 1.538154649644567e-05, "loss": 0.0167, "step": 23431 }, { "epoch": 2.5403295750216826, "grad_norm": 0.4969772696495056, "learning_rate": 1.5377919628608733e-05, "loss": 0.0115, "step": 23432 }, { "epoch": 2.5404379878577625, "grad_norm": 0.14179396629333496, "learning_rate": 1.53742927607718e-05, "loss": 0.0038, "step": 23433 }, { "epoch": 2.540546400693842, "grad_norm": 0.23965585231781006, "learning_rate": 1.537066589293486e-05, "loss": 0.0144, "step": 23434 }, { "epoch": 2.540654813529922, "grad_norm": 0.18362803757190704, "learning_rate": 1.5367039025097925e-05, "loss": 0.0052, "step": 23435 }, { "epoch": 2.5407632263660016, "grad_norm": 0.32210537791252136, "learning_rate": 1.536341215726099e-05, "loss": 0.0124, "step": 23436 }, { "epoch": 2.5408716392020816, "grad_norm": 0.2460847645998001, "learning_rate": 1.5359785289424054e-05, "loss": 0.0025, "step": 23437 }, { "epoch": 2.5409800520381616, "grad_norm": 0.3614455759525299, "learning_rate": 1.5356158421587118e-05, "loss": 0.0334, "step": 23438 }, { "epoch": 2.541088464874241, "grad_norm": 0.1050322949886322, "learning_rate": 1.5352531553750182e-05, "loss": 0.0031, "step": 23439 }, { "epoch": 2.5411968777103207, "grad_norm": 0.12892897427082062, "learning_rate": 1.5348904685913246e-05, "loss": 0.0047, "step": 23440 }, { "epoch": 2.5413052905464006, "grad_norm": 0.02661052718758583, "learning_rate": 1.534527781807631e-05, "loss": 0.0008, "step": 23441 }, { "epoch": 2.5414137033824806, "grad_norm": 0.7970420122146606, "learning_rate": 1.5341650950239374e-05, "loss": 0.0164, "step": 23442 }, { "epoch": 2.54152211621856, "grad_norm": 0.2188979834318161, "learning_rate": 1.533802408240244e-05, "loss": 0.0117, "step": 23443 }, { "epoch": 2.54163052905464, "grad_norm": 0.11159373074769974, "learning_rate": 1.5334397214565502e-05, "loss": 0.0044, "step": 23444 }, { "epoch": 2.5417389418907197, "grad_norm": 0.27520519495010376, "learning_rate": 1.5330770346728567e-05, "loss": 0.0114, "step": 23445 }, { "epoch": 2.5418473547267997, "grad_norm": 0.22404725849628448, "learning_rate": 1.532714347889163e-05, "loss": 0.0109, "step": 23446 }, { "epoch": 2.5419557675628797, "grad_norm": 0.28528061509132385, "learning_rate": 1.5323516611054695e-05, "loss": 0.0197, "step": 23447 }, { "epoch": 2.542064180398959, "grad_norm": 0.20750658214092255, "learning_rate": 1.531988974321776e-05, "loss": 0.0061, "step": 23448 }, { "epoch": 2.542172593235039, "grad_norm": 0.037786852568387985, "learning_rate": 1.5316262875380823e-05, "loss": 0.0022, "step": 23449 }, { "epoch": 2.5422810060711187, "grad_norm": 0.5012169480323792, "learning_rate": 1.5312636007543884e-05, "loss": 0.0156, "step": 23450 }, { "epoch": 2.5423894189071987, "grad_norm": 0.7907159924507141, "learning_rate": 1.530900913970695e-05, "loss": 0.0258, "step": 23451 }, { "epoch": 2.5424978317432783, "grad_norm": 0.5814344882965088, "learning_rate": 1.5305382271870015e-05, "loss": 0.0094, "step": 23452 }, { "epoch": 2.5426062445793582, "grad_norm": 0.047494273632764816, "learning_rate": 1.5301755404033076e-05, "loss": 0.0014, "step": 23453 }, { "epoch": 2.542714657415438, "grad_norm": 0.24854478240013123, "learning_rate": 1.529812853619614e-05, "loss": 0.0263, "step": 23454 }, { "epoch": 2.5428230702515178, "grad_norm": 0.17490950226783752, "learning_rate": 1.5294501668359208e-05, "loss": 0.005, "step": 23455 }, { "epoch": 2.5429314830875978, "grad_norm": 0.1870368868112564, "learning_rate": 1.529087480052227e-05, "loss": 0.0067, "step": 23456 }, { "epoch": 2.5430398959236773, "grad_norm": 0.6336843371391296, "learning_rate": 1.5287247932685332e-05, "loss": 0.0263, "step": 23457 }, { "epoch": 2.5431483087597573, "grad_norm": 0.2482505887746811, "learning_rate": 1.5283621064848397e-05, "loss": 0.0042, "step": 23458 }, { "epoch": 2.543256721595837, "grad_norm": 0.3861711621284485, "learning_rate": 1.5279994197011464e-05, "loss": 0.0237, "step": 23459 }, { "epoch": 2.543365134431917, "grad_norm": 0.2393985390663147, "learning_rate": 1.5276367329174525e-05, "loss": 0.0037, "step": 23460 }, { "epoch": 2.543473547267997, "grad_norm": 0.18327175080776215, "learning_rate": 1.527274046133759e-05, "loss": 0.0072, "step": 23461 }, { "epoch": 2.5435819601040763, "grad_norm": 0.3013463616371155, "learning_rate": 1.5269113593500653e-05, "loss": 0.0117, "step": 23462 }, { "epoch": 2.543690372940156, "grad_norm": 0.01968376338481903, "learning_rate": 1.5265486725663717e-05, "loss": 0.0007, "step": 23463 }, { "epoch": 2.543798785776236, "grad_norm": 0.13387739658355713, "learning_rate": 1.526185985782678e-05, "loss": 0.0044, "step": 23464 }, { "epoch": 2.543907198612316, "grad_norm": 0.4885692298412323, "learning_rate": 1.5258232989989845e-05, "loss": 0.0179, "step": 23465 }, { "epoch": 2.5440156114483954, "grad_norm": 0.389094740152359, "learning_rate": 1.525460612215291e-05, "loss": 0.006, "step": 23466 }, { "epoch": 2.5441240242844754, "grad_norm": 1.6349211931228638, "learning_rate": 1.5250979254315972e-05, "loss": 0.0878, "step": 23467 }, { "epoch": 2.544232437120555, "grad_norm": 0.03823406621813774, "learning_rate": 1.524735238647904e-05, "loss": 0.0012, "step": 23468 }, { "epoch": 2.544340849956635, "grad_norm": 0.570917010307312, "learning_rate": 1.5243725518642102e-05, "loss": 0.0353, "step": 23469 }, { "epoch": 2.544449262792715, "grad_norm": 0.27186307311058044, "learning_rate": 1.5240098650805166e-05, "loss": 0.0073, "step": 23470 }, { "epoch": 2.5445576756287944, "grad_norm": 0.2805120348930359, "learning_rate": 1.5236471782968228e-05, "loss": 0.0043, "step": 23471 }, { "epoch": 2.544666088464874, "grad_norm": 0.06979408860206604, "learning_rate": 1.5232844915131294e-05, "loss": 0.0054, "step": 23472 }, { "epoch": 2.544774501300954, "grad_norm": 0.20424553751945496, "learning_rate": 1.5229218047294358e-05, "loss": 0.0085, "step": 23473 }, { "epoch": 2.544882914137034, "grad_norm": 0.02184894308447838, "learning_rate": 1.522559117945742e-05, "loss": 0.001, "step": 23474 }, { "epoch": 2.5449913269731135, "grad_norm": 0.036458831280469894, "learning_rate": 1.5221964311620485e-05, "loss": 0.001, "step": 23475 }, { "epoch": 2.5450997398091935, "grad_norm": 1.015704870223999, "learning_rate": 1.5218337443783549e-05, "loss": 0.0308, "step": 23476 }, { "epoch": 2.545208152645273, "grad_norm": 0.3143274486064911, "learning_rate": 1.5214710575946615e-05, "loss": 0.0136, "step": 23477 }, { "epoch": 2.545316565481353, "grad_norm": 0.3237030506134033, "learning_rate": 1.5211083708109677e-05, "loss": 0.0074, "step": 23478 }, { "epoch": 2.545424978317433, "grad_norm": 2.091128349304199, "learning_rate": 1.5207456840272741e-05, "loss": 0.028, "step": 23479 }, { "epoch": 2.5455333911535125, "grad_norm": 0.2643117904663086, "learning_rate": 1.5203829972435803e-05, "loss": 0.0221, "step": 23480 }, { "epoch": 2.5456418039895925, "grad_norm": 0.1523558795452118, "learning_rate": 1.5200203104598871e-05, "loss": 0.0023, "step": 23481 }, { "epoch": 2.545750216825672, "grad_norm": 0.09418534487485886, "learning_rate": 1.5196576236761933e-05, "loss": 0.0027, "step": 23482 }, { "epoch": 2.545858629661752, "grad_norm": 0.2659491300582886, "learning_rate": 1.5192949368924997e-05, "loss": 0.0136, "step": 23483 }, { "epoch": 2.545967042497832, "grad_norm": 0.2168029546737671, "learning_rate": 1.518932250108806e-05, "loss": 0.0056, "step": 23484 }, { "epoch": 2.5460754553339116, "grad_norm": 0.09883074462413788, "learning_rate": 1.5185695633251126e-05, "loss": 0.0014, "step": 23485 }, { "epoch": 2.546183868169991, "grad_norm": 0.08751505613327026, "learning_rate": 1.518206876541419e-05, "loss": 0.0011, "step": 23486 }, { "epoch": 2.546292281006071, "grad_norm": 0.20453310012817383, "learning_rate": 1.5178441897577252e-05, "loss": 0.0105, "step": 23487 }, { "epoch": 2.546400693842151, "grad_norm": 0.1596715748310089, "learning_rate": 1.5174815029740316e-05, "loss": 0.006, "step": 23488 }, { "epoch": 2.5465091066782306, "grad_norm": 0.13039687275886536, "learning_rate": 1.517118816190338e-05, "loss": 0.0036, "step": 23489 }, { "epoch": 2.5466175195143106, "grad_norm": 0.16579502820968628, "learning_rate": 1.5167561294066446e-05, "loss": 0.0068, "step": 23490 }, { "epoch": 2.54672593235039, "grad_norm": 0.10308530181646347, "learning_rate": 1.5163934426229509e-05, "loss": 0.0041, "step": 23491 }, { "epoch": 2.54683434518647, "grad_norm": 0.19555121660232544, "learning_rate": 1.5160307558392573e-05, "loss": 0.0082, "step": 23492 }, { "epoch": 2.54694275802255, "grad_norm": 0.7297589182853699, "learning_rate": 1.5156680690555635e-05, "loss": 0.015, "step": 23493 }, { "epoch": 2.5470511708586296, "grad_norm": 0.27259939908981323, "learning_rate": 1.5153053822718703e-05, "loss": 0.0117, "step": 23494 }, { "epoch": 2.547159583694709, "grad_norm": 0.7559054493904114, "learning_rate": 1.5149426954881765e-05, "loss": 0.0343, "step": 23495 }, { "epoch": 2.547267996530789, "grad_norm": 0.2745044231414795, "learning_rate": 1.5145800087044829e-05, "loss": 0.012, "step": 23496 }, { "epoch": 2.547376409366869, "grad_norm": 0.28030312061309814, "learning_rate": 1.5142173219207892e-05, "loss": 0.01, "step": 23497 }, { "epoch": 2.5474848222029487, "grad_norm": 0.05620432645082474, "learning_rate": 1.5138546351370957e-05, "loss": 0.0017, "step": 23498 }, { "epoch": 2.5475932350390287, "grad_norm": 0.28250330686569214, "learning_rate": 1.5134919483534021e-05, "loss": 0.0242, "step": 23499 }, { "epoch": 2.547701647875108, "grad_norm": 0.5561184883117676, "learning_rate": 1.5131292615697084e-05, "loss": 0.0117, "step": 23500 }, { "epoch": 2.547810060711188, "grad_norm": 0.18822695314884186, "learning_rate": 1.5127665747860148e-05, "loss": 0.0062, "step": 23501 }, { "epoch": 2.547918473547268, "grad_norm": 0.18112613260746002, "learning_rate": 1.5124038880023212e-05, "loss": 0.0043, "step": 23502 }, { "epoch": 2.5480268863833477, "grad_norm": 0.4470226764678955, "learning_rate": 1.5120412012186278e-05, "loss": 0.0245, "step": 23503 }, { "epoch": 2.5481352992194277, "grad_norm": 0.4242689907550812, "learning_rate": 1.511678514434934e-05, "loss": 0.0164, "step": 23504 }, { "epoch": 2.5482437120555073, "grad_norm": 0.4637989401817322, "learning_rate": 1.5113158276512404e-05, "loss": 0.042, "step": 23505 }, { "epoch": 2.5483521248915872, "grad_norm": 0.03235954791307449, "learning_rate": 1.5109531408675467e-05, "loss": 0.0014, "step": 23506 }, { "epoch": 2.5484605377276672, "grad_norm": 0.09666503965854645, "learning_rate": 1.5105904540838534e-05, "loss": 0.0033, "step": 23507 }, { "epoch": 2.5485689505637468, "grad_norm": 0.9600180387496948, "learning_rate": 1.5102277673001597e-05, "loss": 0.0301, "step": 23508 }, { "epoch": 2.5486773633998263, "grad_norm": 0.019146475940942764, "learning_rate": 1.509865080516466e-05, "loss": 0.0009, "step": 23509 }, { "epoch": 2.5487857762359063, "grad_norm": 0.3419698476791382, "learning_rate": 1.5095023937327723e-05, "loss": 0.0116, "step": 23510 }, { "epoch": 2.5488941890719863, "grad_norm": 0.04061976447701454, "learning_rate": 1.5091397069490789e-05, "loss": 0.0019, "step": 23511 }, { "epoch": 2.549002601908066, "grad_norm": 0.05270599573850632, "learning_rate": 1.5087770201653853e-05, "loss": 0.0009, "step": 23512 }, { "epoch": 2.549111014744146, "grad_norm": 0.09794996678829193, "learning_rate": 1.5084143333816916e-05, "loss": 0.0026, "step": 23513 }, { "epoch": 2.5492194275802254, "grad_norm": 0.06639016419649124, "learning_rate": 1.508051646597998e-05, "loss": 0.0015, "step": 23514 }, { "epoch": 2.5493278404163053, "grad_norm": 0.0987878069281578, "learning_rate": 1.5076889598143044e-05, "loss": 0.005, "step": 23515 }, { "epoch": 2.5494362532523853, "grad_norm": 0.07152825593948364, "learning_rate": 1.507326273030611e-05, "loss": 0.0029, "step": 23516 }, { "epoch": 2.549544666088465, "grad_norm": 0.16828350722789764, "learning_rate": 1.5069635862469172e-05, "loss": 0.0027, "step": 23517 }, { "epoch": 2.5496530789245444, "grad_norm": 0.9036481380462646, "learning_rate": 1.5066008994632236e-05, "loss": 0.0345, "step": 23518 }, { "epoch": 2.5497614917606244, "grad_norm": 0.40196698904037476, "learning_rate": 1.5062382126795298e-05, "loss": 0.0242, "step": 23519 }, { "epoch": 2.5498699045967044, "grad_norm": 0.38841962814331055, "learning_rate": 1.5058755258958366e-05, "loss": 0.0139, "step": 23520 }, { "epoch": 2.549978317432784, "grad_norm": 0.03245001286268234, "learning_rate": 1.5055128391121428e-05, "loss": 0.0009, "step": 23521 }, { "epoch": 2.550086730268864, "grad_norm": 0.09438664466142654, "learning_rate": 1.5051501523284492e-05, "loss": 0.0008, "step": 23522 }, { "epoch": 2.5501951431049434, "grad_norm": 0.015489969402551651, "learning_rate": 1.5047874655447555e-05, "loss": 0.0003, "step": 23523 }, { "epoch": 2.5503035559410234, "grad_norm": 0.3994394540786743, "learning_rate": 1.504424778761062e-05, "loss": 0.0144, "step": 23524 }, { "epoch": 2.5504119687771034, "grad_norm": 0.27861809730529785, "learning_rate": 1.5040620919773685e-05, "loss": 0.0158, "step": 23525 }, { "epoch": 2.550520381613183, "grad_norm": 0.311434268951416, "learning_rate": 1.5036994051936749e-05, "loss": 0.0087, "step": 23526 }, { "epoch": 2.550628794449263, "grad_norm": 0.10131490230560303, "learning_rate": 1.5033367184099811e-05, "loss": 0.003, "step": 23527 }, { "epoch": 2.5507372072853425, "grad_norm": 0.9228306412696838, "learning_rate": 1.5029740316262877e-05, "loss": 0.0154, "step": 23528 }, { "epoch": 2.5508456201214225, "grad_norm": 0.338665634393692, "learning_rate": 1.5026113448425941e-05, "loss": 0.0078, "step": 23529 }, { "epoch": 2.550954032957502, "grad_norm": 0.12541525065898895, "learning_rate": 1.5022486580589004e-05, "loss": 0.0065, "step": 23530 }, { "epoch": 2.551062445793582, "grad_norm": 0.21206901967525482, "learning_rate": 1.5018859712752068e-05, "loss": 0.0119, "step": 23531 }, { "epoch": 2.5511708586296615, "grad_norm": 0.11552118510007858, "learning_rate": 1.501523284491513e-05, "loss": 0.0025, "step": 23532 }, { "epoch": 2.5512792714657415, "grad_norm": 0.1260966807603836, "learning_rate": 1.5011605977078198e-05, "loss": 0.0057, "step": 23533 }, { "epoch": 2.5513876843018215, "grad_norm": 0.031246673315763474, "learning_rate": 1.500797910924126e-05, "loss": 0.0006, "step": 23534 }, { "epoch": 2.551496097137901, "grad_norm": 0.2081829160451889, "learning_rate": 1.5004352241404324e-05, "loss": 0.008, "step": 23535 }, { "epoch": 2.551604509973981, "grad_norm": 0.04020465537905693, "learning_rate": 1.5000725373567387e-05, "loss": 0.0012, "step": 23536 }, { "epoch": 2.5517129228100606, "grad_norm": 0.3532480299472809, "learning_rate": 1.4997098505730452e-05, "loss": 0.0132, "step": 23537 }, { "epoch": 2.5518213356461406, "grad_norm": 0.10094507038593292, "learning_rate": 1.4993471637893516e-05, "loss": 0.0037, "step": 23538 }, { "epoch": 2.5519297484822205, "grad_norm": 0.3330426812171936, "learning_rate": 1.498984477005658e-05, "loss": 0.038, "step": 23539 }, { "epoch": 2.5520381613183, "grad_norm": 0.0922248363494873, "learning_rate": 1.4986217902219643e-05, "loss": 0.0028, "step": 23540 }, { "epoch": 2.5521465741543796, "grad_norm": 1.773315191268921, "learning_rate": 1.4982591034382709e-05, "loss": 0.0342, "step": 23541 }, { "epoch": 2.5522549869904596, "grad_norm": 0.34565621614456177, "learning_rate": 1.4978964166545773e-05, "loss": 0.0132, "step": 23542 }, { "epoch": 2.5523633998265396, "grad_norm": 0.23422764241695404, "learning_rate": 1.4975337298708835e-05, "loss": 0.0148, "step": 23543 }, { "epoch": 2.552471812662619, "grad_norm": 0.05609515681862831, "learning_rate": 1.49717104308719e-05, "loss": 0.0016, "step": 23544 }, { "epoch": 2.552580225498699, "grad_norm": 0.03358065336942673, "learning_rate": 1.4968083563034962e-05, "loss": 0.0012, "step": 23545 }, { "epoch": 2.5526886383347787, "grad_norm": 0.17358717322349548, "learning_rate": 1.496445669519803e-05, "loss": 0.0055, "step": 23546 }, { "epoch": 2.5527970511708586, "grad_norm": 0.792437732219696, "learning_rate": 1.4960829827361092e-05, "loss": 0.0095, "step": 23547 }, { "epoch": 2.5529054640069386, "grad_norm": 0.12641359865665436, "learning_rate": 1.4957202959524156e-05, "loss": 0.003, "step": 23548 }, { "epoch": 2.553013876843018, "grad_norm": 2.706202268600464, "learning_rate": 1.4953576091687218e-05, "loss": 0.0645, "step": 23549 }, { "epoch": 2.553122289679098, "grad_norm": 0.09033519774675369, "learning_rate": 1.4949949223850284e-05, "loss": 0.0026, "step": 23550 }, { "epoch": 2.5532307025151777, "grad_norm": 0.24725674092769623, "learning_rate": 1.4946322356013348e-05, "loss": 0.0114, "step": 23551 }, { "epoch": 2.5533391153512577, "grad_norm": 0.286171019077301, "learning_rate": 1.4942695488176412e-05, "loss": 0.0064, "step": 23552 }, { "epoch": 2.5534475281873372, "grad_norm": 0.2534068822860718, "learning_rate": 1.4939068620339475e-05, "loss": 0.0179, "step": 23553 }, { "epoch": 2.553555941023417, "grad_norm": 0.007956307381391525, "learning_rate": 1.493544175250254e-05, "loss": 0.0002, "step": 23554 }, { "epoch": 2.5536643538594967, "grad_norm": 0.252794474363327, "learning_rate": 1.4931814884665605e-05, "loss": 0.0068, "step": 23555 }, { "epoch": 2.5537727666955767, "grad_norm": 0.46114808320999146, "learning_rate": 1.4928188016828667e-05, "loss": 0.0288, "step": 23556 }, { "epoch": 2.5538811795316567, "grad_norm": 0.13910746574401855, "learning_rate": 1.4924561148991731e-05, "loss": 0.0045, "step": 23557 }, { "epoch": 2.5539895923677363, "grad_norm": 0.0118170864880085, "learning_rate": 1.4920934281154793e-05, "loss": 0.0003, "step": 23558 }, { "epoch": 2.5540980052038162, "grad_norm": 0.23538033664226532, "learning_rate": 1.4917307413317861e-05, "loss": 0.0108, "step": 23559 }, { "epoch": 2.554206418039896, "grad_norm": 0.32961758971214294, "learning_rate": 1.4913680545480923e-05, "loss": 0.0162, "step": 23560 }, { "epoch": 2.5543148308759758, "grad_norm": 0.09584356099367142, "learning_rate": 1.4910053677643987e-05, "loss": 0.0022, "step": 23561 }, { "epoch": 2.5544232437120558, "grad_norm": 0.7315700054168701, "learning_rate": 1.490642680980705e-05, "loss": 0.0282, "step": 23562 }, { "epoch": 2.5545316565481353, "grad_norm": 0.07265391945838928, "learning_rate": 1.4902799941970116e-05, "loss": 0.0035, "step": 23563 }, { "epoch": 2.554640069384215, "grad_norm": 0.016534602269530296, "learning_rate": 1.489917307413318e-05, "loss": 0.0009, "step": 23564 }, { "epoch": 2.554748482220295, "grad_norm": 0.8722814321517944, "learning_rate": 1.4895546206296244e-05, "loss": 0.0634, "step": 23565 }, { "epoch": 2.554856895056375, "grad_norm": 0.007899895310401917, "learning_rate": 1.4891919338459306e-05, "loss": 0.0004, "step": 23566 }, { "epoch": 2.5549653078924544, "grad_norm": 0.2308926284313202, "learning_rate": 1.4888292470622372e-05, "loss": 0.0083, "step": 23567 }, { "epoch": 2.5550737207285343, "grad_norm": 0.2970424294471741, "learning_rate": 1.4884665602785436e-05, "loss": 0.0054, "step": 23568 }, { "epoch": 2.555182133564614, "grad_norm": 0.8133546113967896, "learning_rate": 1.4881038734948499e-05, "loss": 0.0705, "step": 23569 }, { "epoch": 2.555290546400694, "grad_norm": 0.7444315552711487, "learning_rate": 1.4877411867111563e-05, "loss": 0.0245, "step": 23570 }, { "epoch": 2.555398959236774, "grad_norm": 0.2010936439037323, "learning_rate": 1.4873784999274625e-05, "loss": 0.0125, "step": 23571 }, { "epoch": 2.5555073720728534, "grad_norm": 0.3200032413005829, "learning_rate": 1.4870158131437693e-05, "loss": 0.0099, "step": 23572 }, { "epoch": 2.5556157849089334, "grad_norm": 0.23057296872138977, "learning_rate": 1.4866531263600755e-05, "loss": 0.0056, "step": 23573 }, { "epoch": 2.555724197745013, "grad_norm": 0.4267900288105011, "learning_rate": 1.4862904395763819e-05, "loss": 0.0078, "step": 23574 }, { "epoch": 2.555832610581093, "grad_norm": 0.5596920251846313, "learning_rate": 1.4859277527926882e-05, "loss": 0.0159, "step": 23575 }, { "epoch": 2.5559410234171724, "grad_norm": 0.11261603236198425, "learning_rate": 1.4855650660089947e-05, "loss": 0.0045, "step": 23576 }, { "epoch": 2.5560494362532524, "grad_norm": 0.2025311142206192, "learning_rate": 1.4852023792253011e-05, "loss": 0.0047, "step": 23577 }, { "epoch": 2.556157849089332, "grad_norm": 0.11249435693025589, "learning_rate": 1.4848396924416076e-05, "loss": 0.0029, "step": 23578 }, { "epoch": 2.556266261925412, "grad_norm": 0.09329406917095184, "learning_rate": 1.4844770056579138e-05, "loss": 0.0053, "step": 23579 }, { "epoch": 2.556374674761492, "grad_norm": 0.1768946498632431, "learning_rate": 1.4841143188742204e-05, "loss": 0.0135, "step": 23580 }, { "epoch": 2.5564830875975715, "grad_norm": 0.04817601293325424, "learning_rate": 1.4837516320905268e-05, "loss": 0.0014, "step": 23581 }, { "epoch": 2.5565915004336515, "grad_norm": 0.23536144196987152, "learning_rate": 1.483388945306833e-05, "loss": 0.006, "step": 23582 }, { "epoch": 2.556699913269731, "grad_norm": 0.6193796992301941, "learning_rate": 1.4830262585231394e-05, "loss": 0.053, "step": 23583 }, { "epoch": 2.556808326105811, "grad_norm": 0.14830367267131805, "learning_rate": 1.4826635717394457e-05, "loss": 0.0047, "step": 23584 }, { "epoch": 2.556916738941891, "grad_norm": 0.3454485535621643, "learning_rate": 1.4823008849557524e-05, "loss": 0.0253, "step": 23585 }, { "epoch": 2.5570251517779705, "grad_norm": 0.03319458290934563, "learning_rate": 1.4819381981720587e-05, "loss": 0.0025, "step": 23586 }, { "epoch": 2.55713356461405, "grad_norm": 0.525915265083313, "learning_rate": 1.481575511388365e-05, "loss": 0.0201, "step": 23587 }, { "epoch": 2.55724197745013, "grad_norm": 0.3598097264766693, "learning_rate": 1.4812128246046713e-05, "loss": 0.0202, "step": 23588 }, { "epoch": 2.55735039028621, "grad_norm": 0.7284365892410278, "learning_rate": 1.4808501378209779e-05, "loss": 0.0105, "step": 23589 }, { "epoch": 2.5574588031222896, "grad_norm": 0.04132778197526932, "learning_rate": 1.4804874510372843e-05, "loss": 0.0021, "step": 23590 }, { "epoch": 2.5575672159583696, "grad_norm": 0.13301508128643036, "learning_rate": 1.4801247642535907e-05, "loss": 0.0063, "step": 23591 }, { "epoch": 2.557675628794449, "grad_norm": 0.27448251843452454, "learning_rate": 1.479762077469897e-05, "loss": 0.0279, "step": 23592 }, { "epoch": 2.557784041630529, "grad_norm": 0.03408564254641533, "learning_rate": 1.4793993906862035e-05, "loss": 0.0009, "step": 23593 }, { "epoch": 2.557892454466609, "grad_norm": 0.80808025598526, "learning_rate": 1.47903670390251e-05, "loss": 0.0178, "step": 23594 }, { "epoch": 2.5580008673026886, "grad_norm": 0.45621317625045776, "learning_rate": 1.4786740171188162e-05, "loss": 0.0061, "step": 23595 }, { "epoch": 2.5581092801387686, "grad_norm": 0.25621476769447327, "learning_rate": 1.4783113303351226e-05, "loss": 0.0086, "step": 23596 }, { "epoch": 2.558217692974848, "grad_norm": 0.602728009223938, "learning_rate": 1.4779486435514288e-05, "loss": 0.0058, "step": 23597 }, { "epoch": 2.558326105810928, "grad_norm": 0.2131596803665161, "learning_rate": 1.4775859567677356e-05, "loss": 0.0079, "step": 23598 }, { "epoch": 2.5584345186470077, "grad_norm": 0.38763874769210815, "learning_rate": 1.4772232699840418e-05, "loss": 0.0419, "step": 23599 }, { "epoch": 2.5585429314830876, "grad_norm": 0.4821307361125946, "learning_rate": 1.4768605832003482e-05, "loss": 0.0402, "step": 23600 }, { "epoch": 2.558651344319167, "grad_norm": 0.00634965393692255, "learning_rate": 1.4764978964166545e-05, "loss": 0.0003, "step": 23601 }, { "epoch": 2.558759757155247, "grad_norm": 0.0801568329334259, "learning_rate": 1.476135209632961e-05, "loss": 0.0043, "step": 23602 }, { "epoch": 2.558868169991327, "grad_norm": 0.06842633336782455, "learning_rate": 1.4757725228492675e-05, "loss": 0.0029, "step": 23603 }, { "epoch": 2.5589765828274067, "grad_norm": 0.6260518431663513, "learning_rate": 1.4754098360655739e-05, "loss": 0.0084, "step": 23604 }, { "epoch": 2.5590849956634867, "grad_norm": 0.0313928946852684, "learning_rate": 1.4750471492818801e-05, "loss": 0.0015, "step": 23605 }, { "epoch": 2.5591934084995662, "grad_norm": 0.11045894771814346, "learning_rate": 1.4746844624981867e-05, "loss": 0.0097, "step": 23606 }, { "epoch": 2.559301821335646, "grad_norm": 0.002104961546137929, "learning_rate": 1.4743217757144931e-05, "loss": 0.0001, "step": 23607 }, { "epoch": 2.559410234171726, "grad_norm": 0.7110666632652283, "learning_rate": 1.4739590889307994e-05, "loss": 0.0381, "step": 23608 }, { "epoch": 2.5595186470078057, "grad_norm": 0.07138856500387192, "learning_rate": 1.4735964021471058e-05, "loss": 0.0028, "step": 23609 }, { "epoch": 2.5596270598438853, "grad_norm": 0.04688474163413048, "learning_rate": 1.473233715363412e-05, "loss": 0.0015, "step": 23610 }, { "epoch": 2.5597354726799653, "grad_norm": 0.15217041969299316, "learning_rate": 1.4728710285797188e-05, "loss": 0.0059, "step": 23611 }, { "epoch": 2.5598438855160452, "grad_norm": 0.3333805799484253, "learning_rate": 1.472508341796025e-05, "loss": 0.0023, "step": 23612 }, { "epoch": 2.559952298352125, "grad_norm": 0.04914478212594986, "learning_rate": 1.4721456550123314e-05, "loss": 0.0019, "step": 23613 }, { "epoch": 2.5600607111882048, "grad_norm": 0.12957893311977386, "learning_rate": 1.4717829682286377e-05, "loss": 0.0039, "step": 23614 }, { "epoch": 2.5601691240242843, "grad_norm": 0.008542319759726524, "learning_rate": 1.4714202814449442e-05, "loss": 0.0003, "step": 23615 }, { "epoch": 2.5602775368603643, "grad_norm": 0.31780728697776794, "learning_rate": 1.4710575946612506e-05, "loss": 0.0105, "step": 23616 }, { "epoch": 2.5603859496964443, "grad_norm": 0.07268258184194565, "learning_rate": 1.470694907877557e-05, "loss": 0.0013, "step": 23617 }, { "epoch": 2.560494362532524, "grad_norm": 0.16222448647022247, "learning_rate": 1.4703322210938633e-05, "loss": 0.0028, "step": 23618 }, { "epoch": 2.5606027753686034, "grad_norm": 0.2780648171901703, "learning_rate": 1.4699695343101699e-05, "loss": 0.0072, "step": 23619 }, { "epoch": 2.5607111882046834, "grad_norm": 0.563738226890564, "learning_rate": 1.4696068475264763e-05, "loss": 0.0103, "step": 23620 }, { "epoch": 2.5608196010407633, "grad_norm": 0.33428531885147095, "learning_rate": 1.4692441607427825e-05, "loss": 0.0177, "step": 23621 }, { "epoch": 2.560928013876843, "grad_norm": 0.306153267621994, "learning_rate": 1.468881473959089e-05, "loss": 0.0738, "step": 23622 }, { "epoch": 2.561036426712923, "grad_norm": 1.1148422956466675, "learning_rate": 1.4685187871753952e-05, "loss": 0.0163, "step": 23623 }, { "epoch": 2.5611448395490024, "grad_norm": 0.052175793796777725, "learning_rate": 1.468156100391702e-05, "loss": 0.0014, "step": 23624 }, { "epoch": 2.5612532523850824, "grad_norm": 0.3297058343887329, "learning_rate": 1.4677934136080082e-05, "loss": 0.0124, "step": 23625 }, { "epoch": 2.5613616652211624, "grad_norm": 0.11173022538423538, "learning_rate": 1.4674307268243146e-05, "loss": 0.0128, "step": 23626 }, { "epoch": 2.561470078057242, "grad_norm": 0.010182510130107403, "learning_rate": 1.4670680400406208e-05, "loss": 0.0002, "step": 23627 }, { "epoch": 2.561578490893322, "grad_norm": 0.46224522590637207, "learning_rate": 1.4667053532569274e-05, "loss": 0.0231, "step": 23628 }, { "epoch": 2.5616869037294014, "grad_norm": 0.24327918887138367, "learning_rate": 1.4663426664732338e-05, "loss": 0.0064, "step": 23629 }, { "epoch": 2.5617953165654814, "grad_norm": 0.44072550535202026, "learning_rate": 1.4659799796895402e-05, "loss": 0.0337, "step": 23630 }, { "epoch": 2.5619037294015614, "grad_norm": 0.33739498257637024, "learning_rate": 1.4656172929058465e-05, "loss": 0.0266, "step": 23631 }, { "epoch": 2.562012142237641, "grad_norm": 0.03375609964132309, "learning_rate": 1.465254606122153e-05, "loss": 0.0019, "step": 23632 }, { "epoch": 2.5621205550737205, "grad_norm": 0.33477330207824707, "learning_rate": 1.4648919193384595e-05, "loss": 0.0231, "step": 23633 }, { "epoch": 2.5622289679098005, "grad_norm": 0.649075984954834, "learning_rate": 1.4645292325547657e-05, "loss": 0.0431, "step": 23634 }, { "epoch": 2.5623373807458805, "grad_norm": 0.16700004041194916, "learning_rate": 1.4641665457710721e-05, "loss": 0.0076, "step": 23635 }, { "epoch": 2.56244579358196, "grad_norm": 0.15763865411281586, "learning_rate": 1.4638038589873787e-05, "loss": 0.0028, "step": 23636 }, { "epoch": 2.56255420641804, "grad_norm": 0.1444905400276184, "learning_rate": 1.4634411722036851e-05, "loss": 0.0037, "step": 23637 }, { "epoch": 2.5626626192541195, "grad_norm": 0.4544379711151123, "learning_rate": 1.4630784854199913e-05, "loss": 0.01, "step": 23638 }, { "epoch": 2.5627710320901995, "grad_norm": 0.05433208867907524, "learning_rate": 1.4627157986362978e-05, "loss": 0.0022, "step": 23639 }, { "epoch": 2.5628794449262795, "grad_norm": 0.5164932608604431, "learning_rate": 1.462353111852604e-05, "loss": 0.0355, "step": 23640 }, { "epoch": 2.562987857762359, "grad_norm": 0.1779394894838333, "learning_rate": 1.4619904250689106e-05, "loss": 0.0042, "step": 23641 }, { "epoch": 2.5630962705984386, "grad_norm": 0.3706801235675812, "learning_rate": 1.461627738285217e-05, "loss": 0.015, "step": 23642 }, { "epoch": 2.5632046834345186, "grad_norm": 0.4341681897640228, "learning_rate": 1.4612650515015234e-05, "loss": 0.0237, "step": 23643 }, { "epoch": 2.5633130962705986, "grad_norm": 0.05478677153587341, "learning_rate": 1.4609023647178296e-05, "loss": 0.0022, "step": 23644 }, { "epoch": 2.563421509106678, "grad_norm": 0.06054489687085152, "learning_rate": 1.4605396779341362e-05, "loss": 0.0025, "step": 23645 }, { "epoch": 2.563529921942758, "grad_norm": 0.07332322001457214, "learning_rate": 1.4601769911504426e-05, "loss": 0.0029, "step": 23646 }, { "epoch": 2.5636383347788376, "grad_norm": 0.05295491963624954, "learning_rate": 1.4598143043667489e-05, "loss": 0.0026, "step": 23647 }, { "epoch": 2.5637467476149176, "grad_norm": 0.1376011073589325, "learning_rate": 1.4594516175830553e-05, "loss": 0.0077, "step": 23648 }, { "epoch": 2.5638551604509976, "grad_norm": 0.41627058386802673, "learning_rate": 1.4590889307993619e-05, "loss": 0.0126, "step": 23649 }, { "epoch": 2.563963573287077, "grad_norm": 0.25982600450515747, "learning_rate": 1.4587262440156683e-05, "loss": 0.0219, "step": 23650 }, { "epoch": 2.564071986123157, "grad_norm": 0.13169997930526733, "learning_rate": 1.4583635572319745e-05, "loss": 0.0036, "step": 23651 }, { "epoch": 2.5641803989592367, "grad_norm": 0.14788977801799774, "learning_rate": 1.458000870448281e-05, "loss": 0.0047, "step": 23652 }, { "epoch": 2.5642888117953166, "grad_norm": 0.16917967796325684, "learning_rate": 1.4576381836645872e-05, "loss": 0.0072, "step": 23653 }, { "epoch": 2.5643972246313966, "grad_norm": 0.16269974410533905, "learning_rate": 1.4572754968808937e-05, "loss": 0.0086, "step": 23654 }, { "epoch": 2.564505637467476, "grad_norm": 0.356587678194046, "learning_rate": 1.4569128100972001e-05, "loss": 0.0076, "step": 23655 }, { "epoch": 2.5646140503035557, "grad_norm": 0.007092293817549944, "learning_rate": 1.4565501233135066e-05, "loss": 0.0003, "step": 23656 }, { "epoch": 2.5647224631396357, "grad_norm": 0.027828089892864227, "learning_rate": 1.4561874365298128e-05, "loss": 0.0008, "step": 23657 }, { "epoch": 2.5648308759757157, "grad_norm": 0.26997390389442444, "learning_rate": 1.4558247497461194e-05, "loss": 0.0059, "step": 23658 }, { "epoch": 2.5649392888117952, "grad_norm": 0.2782917618751526, "learning_rate": 1.4554620629624258e-05, "loss": 0.0436, "step": 23659 }, { "epoch": 2.565047701647875, "grad_norm": 0.008621926419436932, "learning_rate": 1.455099376178732e-05, "loss": 0.0003, "step": 23660 }, { "epoch": 2.5651561144839548, "grad_norm": 1.0855776071548462, "learning_rate": 1.4547366893950384e-05, "loss": 0.0495, "step": 23661 }, { "epoch": 2.5652645273200347, "grad_norm": 0.17971602082252502, "learning_rate": 1.454374002611345e-05, "loss": 0.0041, "step": 23662 }, { "epoch": 2.5653729401561147, "grad_norm": 0.16827431321144104, "learning_rate": 1.4540113158276514e-05, "loss": 0.0116, "step": 23663 }, { "epoch": 2.5654813529921943, "grad_norm": 0.7375296950340271, "learning_rate": 1.4536486290439577e-05, "loss": 0.0495, "step": 23664 }, { "epoch": 2.565589765828274, "grad_norm": 0.3499232530593872, "learning_rate": 1.4532859422602641e-05, "loss": 0.0392, "step": 23665 }, { "epoch": 2.565698178664354, "grad_norm": 0.22169406712055206, "learning_rate": 1.4529232554765703e-05, "loss": 0.0094, "step": 23666 }, { "epoch": 2.5658065915004338, "grad_norm": 0.3318246901035309, "learning_rate": 1.4525605686928769e-05, "loss": 0.0145, "step": 23667 }, { "epoch": 2.5659150043365133, "grad_norm": 0.3075641393661499, "learning_rate": 1.4521978819091833e-05, "loss": 0.0227, "step": 23668 }, { "epoch": 2.5660234171725933, "grad_norm": 0.2302781194448471, "learning_rate": 1.4518351951254897e-05, "loss": 0.0237, "step": 23669 }, { "epoch": 2.566131830008673, "grad_norm": 0.393742173910141, "learning_rate": 1.451472508341796e-05, "loss": 0.055, "step": 23670 }, { "epoch": 2.566240242844753, "grad_norm": 0.0681810975074768, "learning_rate": 1.4511098215581025e-05, "loss": 0.0046, "step": 23671 }, { "epoch": 2.566348655680833, "grad_norm": 0.2597894072532654, "learning_rate": 1.450747134774409e-05, "loss": 0.007, "step": 23672 }, { "epoch": 2.5664570685169124, "grad_norm": 0.014945873990654945, "learning_rate": 1.4503844479907152e-05, "loss": 0.0006, "step": 23673 }, { "epoch": 2.5665654813529923, "grad_norm": 0.016444280743598938, "learning_rate": 1.4500217612070216e-05, "loss": 0.0006, "step": 23674 }, { "epoch": 2.566673894189072, "grad_norm": 0.17422086000442505, "learning_rate": 1.4496590744233282e-05, "loss": 0.0064, "step": 23675 }, { "epoch": 2.566782307025152, "grad_norm": 0.190760537981987, "learning_rate": 1.4492963876396346e-05, "loss": 0.0062, "step": 23676 }, { "epoch": 2.5668907198612314, "grad_norm": 0.13363997638225555, "learning_rate": 1.4489337008559408e-05, "loss": 0.0043, "step": 23677 }, { "epoch": 2.5669991326973114, "grad_norm": 0.2791021764278412, "learning_rate": 1.4485710140722473e-05, "loss": 0.0114, "step": 23678 }, { "epoch": 2.567107545533391, "grad_norm": 0.29731395840644836, "learning_rate": 1.4482083272885535e-05, "loss": 0.0243, "step": 23679 }, { "epoch": 2.567215958369471, "grad_norm": 0.1122979000210762, "learning_rate": 1.44784564050486e-05, "loss": 0.0055, "step": 23680 }, { "epoch": 2.567324371205551, "grad_norm": 0.03941349685192108, "learning_rate": 1.4474829537211665e-05, "loss": 0.0018, "step": 23681 }, { "epoch": 2.5674327840416304, "grad_norm": 0.05658110976219177, "learning_rate": 1.4471202669374729e-05, "loss": 0.0017, "step": 23682 }, { "epoch": 2.5675411968777104, "grad_norm": 0.06538267433643341, "learning_rate": 1.4467575801537791e-05, "loss": 0.0026, "step": 23683 }, { "epoch": 2.56764960971379, "grad_norm": 0.24341613054275513, "learning_rate": 1.4463948933700857e-05, "loss": 0.0103, "step": 23684 }, { "epoch": 2.56775802254987, "grad_norm": 0.6069218516349792, "learning_rate": 1.4460322065863921e-05, "loss": 0.0141, "step": 23685 }, { "epoch": 2.56786643538595, "grad_norm": 0.10611648112535477, "learning_rate": 1.4456695198026984e-05, "loss": 0.0026, "step": 23686 }, { "epoch": 2.5679748482220295, "grad_norm": 0.0988689661026001, "learning_rate": 1.4453068330190048e-05, "loss": 0.0061, "step": 23687 }, { "epoch": 2.568083261058109, "grad_norm": 0.4666038155555725, "learning_rate": 1.4449441462353114e-05, "loss": 0.0151, "step": 23688 }, { "epoch": 2.568191673894189, "grad_norm": 0.037803567945957184, "learning_rate": 1.4445814594516178e-05, "loss": 0.0016, "step": 23689 }, { "epoch": 2.568300086730269, "grad_norm": 0.4366854727268219, "learning_rate": 1.444218772667924e-05, "loss": 0.0105, "step": 23690 }, { "epoch": 2.5684084995663485, "grad_norm": 0.19052249193191528, "learning_rate": 1.4438560858842304e-05, "loss": 0.0069, "step": 23691 }, { "epoch": 2.5685169124024285, "grad_norm": 0.33358514308929443, "learning_rate": 1.4434933991005367e-05, "loss": 0.0289, "step": 23692 }, { "epoch": 2.568625325238508, "grad_norm": 0.17621469497680664, "learning_rate": 1.4431307123168432e-05, "loss": 0.0052, "step": 23693 }, { "epoch": 2.568733738074588, "grad_norm": 0.04010797291994095, "learning_rate": 1.4427680255331497e-05, "loss": 0.0018, "step": 23694 }, { "epoch": 2.568842150910668, "grad_norm": 0.2505079209804535, "learning_rate": 1.442405338749456e-05, "loss": 0.0107, "step": 23695 }, { "epoch": 2.5689505637467476, "grad_norm": 0.20576444268226624, "learning_rate": 1.4420426519657623e-05, "loss": 0.0108, "step": 23696 }, { "epoch": 2.5690589765828276, "grad_norm": 0.2000615894794464, "learning_rate": 1.4416799651820689e-05, "loss": 0.012, "step": 23697 }, { "epoch": 2.569167389418907, "grad_norm": 0.2244492620229721, "learning_rate": 1.4413172783983753e-05, "loss": 0.021, "step": 23698 }, { "epoch": 2.569275802254987, "grad_norm": 0.3431118428707123, "learning_rate": 1.4409545916146815e-05, "loss": 0.0071, "step": 23699 }, { "epoch": 2.5693842150910666, "grad_norm": 0.047259047627449036, "learning_rate": 1.440591904830988e-05, "loss": 0.0024, "step": 23700 }, { "epoch": 2.5694926279271466, "grad_norm": 0.303735613822937, "learning_rate": 1.4402292180472945e-05, "loss": 0.0098, "step": 23701 }, { "epoch": 2.569601040763226, "grad_norm": 0.1977057009935379, "learning_rate": 1.439866531263601e-05, "loss": 0.0067, "step": 23702 }, { "epoch": 2.569709453599306, "grad_norm": 0.4957166910171509, "learning_rate": 1.4395038444799072e-05, "loss": 0.0223, "step": 23703 }, { "epoch": 2.569817866435386, "grad_norm": 0.17677690088748932, "learning_rate": 1.4391411576962136e-05, "loss": 0.0052, "step": 23704 }, { "epoch": 2.5699262792714657, "grad_norm": 0.653282642364502, "learning_rate": 1.4387784709125198e-05, "loss": 0.0721, "step": 23705 }, { "epoch": 2.5700346921075456, "grad_norm": 0.0672612339258194, "learning_rate": 1.4384157841288266e-05, "loss": 0.002, "step": 23706 }, { "epoch": 2.570143104943625, "grad_norm": 0.28661754727363586, "learning_rate": 1.4380530973451328e-05, "loss": 0.0075, "step": 23707 }, { "epoch": 2.570251517779705, "grad_norm": 0.35549846291542053, "learning_rate": 1.4376904105614392e-05, "loss": 0.008, "step": 23708 }, { "epoch": 2.570359930615785, "grad_norm": 0.08315572142601013, "learning_rate": 1.4373277237777455e-05, "loss": 0.0029, "step": 23709 }, { "epoch": 2.5704683434518647, "grad_norm": 0.45395949482917786, "learning_rate": 1.436965036994052e-05, "loss": 0.0283, "step": 23710 }, { "epoch": 2.5705767562879442, "grad_norm": 0.019073812291026115, "learning_rate": 1.4366023502103585e-05, "loss": 0.0009, "step": 23711 }, { "epoch": 2.5706851691240242, "grad_norm": 0.5342745780944824, "learning_rate": 1.4362396634266647e-05, "loss": 0.0388, "step": 23712 }, { "epoch": 2.570793581960104, "grad_norm": 0.27053534984588623, "learning_rate": 1.4358769766429711e-05, "loss": 0.0082, "step": 23713 }, { "epoch": 2.5709019947961838, "grad_norm": 0.0574057437479496, "learning_rate": 1.4355142898592777e-05, "loss": 0.0058, "step": 23714 }, { "epoch": 2.5710104076322637, "grad_norm": 0.07263461500406265, "learning_rate": 1.4351516030755841e-05, "loss": 0.0028, "step": 23715 }, { "epoch": 2.5711188204683433, "grad_norm": 0.29540976881980896, "learning_rate": 1.4347889162918903e-05, "loss": 0.0155, "step": 23716 }, { "epoch": 2.5712272333044233, "grad_norm": 0.2358170747756958, "learning_rate": 1.4344262295081968e-05, "loss": 0.0192, "step": 23717 }, { "epoch": 2.5713356461405033, "grad_norm": 0.03196338191628456, "learning_rate": 1.434063542724503e-05, "loss": 0.0013, "step": 23718 }, { "epoch": 2.571444058976583, "grad_norm": 0.021567195653915405, "learning_rate": 1.4337008559408097e-05, "loss": 0.0005, "step": 23719 }, { "epoch": 2.5715524718126628, "grad_norm": 0.3918647766113281, "learning_rate": 1.433338169157116e-05, "loss": 0.0102, "step": 23720 }, { "epoch": 2.5716608846487423, "grad_norm": 0.07390424609184265, "learning_rate": 1.4329754823734224e-05, "loss": 0.0015, "step": 23721 }, { "epoch": 2.5717692974848223, "grad_norm": 0.21763578057289124, "learning_rate": 1.4326127955897286e-05, "loss": 0.0168, "step": 23722 }, { "epoch": 2.571877710320902, "grad_norm": 0.1144028902053833, "learning_rate": 1.4322501088060352e-05, "loss": 0.0046, "step": 23723 }, { "epoch": 2.571986123156982, "grad_norm": 0.13968446850776672, "learning_rate": 1.4318874220223416e-05, "loss": 0.0063, "step": 23724 }, { "epoch": 2.5720945359930614, "grad_norm": 0.21349747478961945, "learning_rate": 1.4315247352386479e-05, "loss": 0.0068, "step": 23725 }, { "epoch": 2.5722029488291414, "grad_norm": 0.4318760633468628, "learning_rate": 1.4311620484549543e-05, "loss": 0.0187, "step": 23726 }, { "epoch": 2.5723113616652213, "grad_norm": 0.11504865437746048, "learning_rate": 1.4307993616712609e-05, "loss": 0.0038, "step": 23727 }, { "epoch": 2.572419774501301, "grad_norm": 0.1519317775964737, "learning_rate": 1.4304366748875673e-05, "loss": 0.004, "step": 23728 }, { "epoch": 2.572528187337381, "grad_norm": 0.0013680423144251108, "learning_rate": 1.4300739881038735e-05, "loss": 0.0001, "step": 23729 }, { "epoch": 2.5726366001734604, "grad_norm": 0.2642452120780945, "learning_rate": 1.42971130132018e-05, "loss": 0.0272, "step": 23730 }, { "epoch": 2.5727450130095404, "grad_norm": 0.4349948465824127, "learning_rate": 1.4293486145364865e-05, "loss": 0.03, "step": 23731 }, { "epoch": 2.5728534258456204, "grad_norm": 0.8076443672180176, "learning_rate": 1.4289859277527929e-05, "loss": 0.0903, "step": 23732 }, { "epoch": 2.5729618386817, "grad_norm": 0.6744415163993835, "learning_rate": 1.4286232409690992e-05, "loss": 0.0204, "step": 23733 }, { "epoch": 2.5730702515177795, "grad_norm": 0.45907190442085266, "learning_rate": 1.4282605541854056e-05, "loss": 0.0186, "step": 23734 }, { "epoch": 2.5731786643538594, "grad_norm": 0.0003366768651176244, "learning_rate": 1.4278978674017118e-05, "loss": 0.0, "step": 23735 }, { "epoch": 2.5732870771899394, "grad_norm": 0.07530497759580612, "learning_rate": 1.4275351806180184e-05, "loss": 0.0034, "step": 23736 }, { "epoch": 2.573395490026019, "grad_norm": 0.15975451469421387, "learning_rate": 1.4271724938343248e-05, "loss": 0.0043, "step": 23737 }, { "epoch": 2.573503902862099, "grad_norm": 0.1960310935974121, "learning_rate": 1.426809807050631e-05, "loss": 0.0061, "step": 23738 }, { "epoch": 2.5736123156981785, "grad_norm": 0.039787519723176956, "learning_rate": 1.4264471202669374e-05, "loss": 0.0008, "step": 23739 }, { "epoch": 2.5737207285342585, "grad_norm": 0.17713604867458344, "learning_rate": 1.426084433483244e-05, "loss": 0.0029, "step": 23740 }, { "epoch": 2.5738291413703385, "grad_norm": 0.39384710788726807, "learning_rate": 1.4257217466995504e-05, "loss": 0.0139, "step": 23741 }, { "epoch": 2.573937554206418, "grad_norm": 0.23679746687412262, "learning_rate": 1.4253590599158567e-05, "loss": 0.0036, "step": 23742 }, { "epoch": 2.574045967042498, "grad_norm": 0.23684586584568024, "learning_rate": 1.4249963731321631e-05, "loss": 0.0128, "step": 23743 }, { "epoch": 2.5741543798785775, "grad_norm": 0.18863840401172638, "learning_rate": 1.4246336863484697e-05, "loss": 0.0141, "step": 23744 }, { "epoch": 2.5742627927146575, "grad_norm": 0.21692651510238647, "learning_rate": 1.424270999564776e-05, "loss": 0.0152, "step": 23745 }, { "epoch": 2.574371205550737, "grad_norm": 0.04846267029643059, "learning_rate": 1.4239083127810823e-05, "loss": 0.0024, "step": 23746 }, { "epoch": 2.574479618386817, "grad_norm": 0.36860740184783936, "learning_rate": 1.4235456259973887e-05, "loss": 0.0127, "step": 23747 }, { "epoch": 2.5745880312228966, "grad_norm": 0.29957282543182373, "learning_rate": 1.423182939213695e-05, "loss": 0.0079, "step": 23748 }, { "epoch": 2.5746964440589766, "grad_norm": 0.27149438858032227, "learning_rate": 1.4228202524300016e-05, "loss": 0.0147, "step": 23749 }, { "epoch": 2.5748048568950566, "grad_norm": 0.007849999703466892, "learning_rate": 1.422457565646308e-05, "loss": 0.0003, "step": 23750 }, { "epoch": 2.574913269731136, "grad_norm": 0.07396005839109421, "learning_rate": 1.4220948788626142e-05, "loss": 0.0033, "step": 23751 }, { "epoch": 2.575021682567216, "grad_norm": 0.7118238210678101, "learning_rate": 1.4217321920789206e-05, "loss": 0.0224, "step": 23752 }, { "epoch": 2.5751300954032956, "grad_norm": 0.0804838165640831, "learning_rate": 1.4213695052952272e-05, "loss": 0.0024, "step": 23753 }, { "epoch": 2.5752385082393756, "grad_norm": 0.06705974042415619, "learning_rate": 1.4210068185115336e-05, "loss": 0.0022, "step": 23754 }, { "epoch": 2.5753469210754556, "grad_norm": 0.16080473363399506, "learning_rate": 1.4206441317278398e-05, "loss": 0.0033, "step": 23755 }, { "epoch": 2.575455333911535, "grad_norm": 0.2491999715566635, "learning_rate": 1.4202814449441463e-05, "loss": 0.0132, "step": 23756 }, { "epoch": 2.5755637467476147, "grad_norm": 1.248144507408142, "learning_rate": 1.4199187581604528e-05, "loss": 0.016, "step": 23757 }, { "epoch": 2.5756721595836947, "grad_norm": 0.24270997941493988, "learning_rate": 1.4195560713767592e-05, "loss": 0.0117, "step": 23758 }, { "epoch": 2.5757805724197746, "grad_norm": 0.14829093217849731, "learning_rate": 1.4191933845930655e-05, "loss": 0.0059, "step": 23759 }, { "epoch": 2.575888985255854, "grad_norm": 0.04159635305404663, "learning_rate": 1.4188306978093719e-05, "loss": 0.0016, "step": 23760 }, { "epoch": 2.575997398091934, "grad_norm": 0.2174486368894577, "learning_rate": 1.4184680110256781e-05, "loss": 0.0027, "step": 23761 }, { "epoch": 2.5761058109280137, "grad_norm": 0.4001103937625885, "learning_rate": 1.4181053242419847e-05, "loss": 0.0315, "step": 23762 }, { "epoch": 2.5762142237640937, "grad_norm": 0.9500319957733154, "learning_rate": 1.4177426374582911e-05, "loss": 0.0437, "step": 23763 }, { "epoch": 2.5763226366001737, "grad_norm": 0.801923930644989, "learning_rate": 1.4173799506745974e-05, "loss": 0.0321, "step": 23764 }, { "epoch": 2.5764310494362532, "grad_norm": 0.1442846655845642, "learning_rate": 1.4170172638909038e-05, "loss": 0.0062, "step": 23765 }, { "epoch": 2.5765394622723328, "grad_norm": 0.008016291074454784, "learning_rate": 1.4166545771072104e-05, "loss": 0.0005, "step": 23766 }, { "epoch": 2.5766478751084128, "grad_norm": 0.4641664922237396, "learning_rate": 1.4162918903235168e-05, "loss": 0.0282, "step": 23767 }, { "epoch": 2.5767562879444927, "grad_norm": 0.31027543544769287, "learning_rate": 1.415929203539823e-05, "loss": 0.0222, "step": 23768 }, { "epoch": 2.5768647007805723, "grad_norm": 0.5015925765037537, "learning_rate": 1.4155665167561294e-05, "loss": 0.0067, "step": 23769 }, { "epoch": 2.5769731136166523, "grad_norm": 0.017387419939041138, "learning_rate": 1.415203829972436e-05, "loss": 0.0007, "step": 23770 }, { "epoch": 2.577081526452732, "grad_norm": 0.08730659633874893, "learning_rate": 1.4148411431887424e-05, "loss": 0.0019, "step": 23771 }, { "epoch": 2.577189939288812, "grad_norm": 0.13568174839019775, "learning_rate": 1.4144784564050487e-05, "loss": 0.0512, "step": 23772 }, { "epoch": 2.577298352124892, "grad_norm": 0.21465030312538147, "learning_rate": 1.414115769621355e-05, "loss": 0.0135, "step": 23773 }, { "epoch": 2.5774067649609713, "grad_norm": 0.0972193256020546, "learning_rate": 1.4137530828376613e-05, "loss": 0.007, "step": 23774 }, { "epoch": 2.5775151777970513, "grad_norm": 0.17709848284721375, "learning_rate": 1.4133903960539679e-05, "loss": 0.0047, "step": 23775 }, { "epoch": 2.577623590633131, "grad_norm": 0.17090080678462982, "learning_rate": 1.4130277092702743e-05, "loss": 0.0117, "step": 23776 }, { "epoch": 2.577732003469211, "grad_norm": 0.644858181476593, "learning_rate": 1.4126650224865805e-05, "loss": 0.0185, "step": 23777 }, { "epoch": 2.577840416305291, "grad_norm": 0.16151203215122223, "learning_rate": 1.412302335702887e-05, "loss": 0.0066, "step": 23778 }, { "epoch": 2.5779488291413704, "grad_norm": 0.6255480051040649, "learning_rate": 1.4119396489191935e-05, "loss": 0.0155, "step": 23779 }, { "epoch": 2.57805724197745, "grad_norm": 0.29836419224739075, "learning_rate": 1.4115769621355e-05, "loss": 0.0123, "step": 23780 }, { "epoch": 2.57816565481353, "grad_norm": 0.008977541700005531, "learning_rate": 1.4112142753518062e-05, "loss": 0.0004, "step": 23781 }, { "epoch": 2.57827406764961, "grad_norm": 0.013386547565460205, "learning_rate": 1.4108515885681126e-05, "loss": 0.0003, "step": 23782 }, { "epoch": 2.5783824804856894, "grad_norm": 0.33111441135406494, "learning_rate": 1.4104889017844192e-05, "loss": 0.0176, "step": 23783 }, { "epoch": 2.5784908933217694, "grad_norm": 0.047615885734558105, "learning_rate": 1.4101262150007256e-05, "loss": 0.0012, "step": 23784 }, { "epoch": 2.578599306157849, "grad_norm": 0.0020547262392938137, "learning_rate": 1.4097635282170318e-05, "loss": 0.0001, "step": 23785 }, { "epoch": 2.578707718993929, "grad_norm": 0.1161179393529892, "learning_rate": 1.4094008414333382e-05, "loss": 0.0033, "step": 23786 }, { "epoch": 2.578816131830009, "grad_norm": 0.07612530142068863, "learning_rate": 1.4090381546496445e-05, "loss": 0.0021, "step": 23787 }, { "epoch": 2.5789245446660884, "grad_norm": 0.3100142478942871, "learning_rate": 1.408675467865951e-05, "loss": 0.0053, "step": 23788 }, { "epoch": 2.579032957502168, "grad_norm": 0.5045644044876099, "learning_rate": 1.4083127810822575e-05, "loss": 0.0155, "step": 23789 }, { "epoch": 2.579141370338248, "grad_norm": 0.018516631796956062, "learning_rate": 1.4079500942985637e-05, "loss": 0.0005, "step": 23790 }, { "epoch": 2.579249783174328, "grad_norm": 0.3596976399421692, "learning_rate": 1.4075874075148701e-05, "loss": 0.0064, "step": 23791 }, { "epoch": 2.5793581960104075, "grad_norm": 0.14594800770282745, "learning_rate": 1.4072247207311767e-05, "loss": 0.0054, "step": 23792 }, { "epoch": 2.5794666088464875, "grad_norm": 0.11123230308294296, "learning_rate": 1.4068620339474831e-05, "loss": 0.0034, "step": 23793 }, { "epoch": 2.579575021682567, "grad_norm": 1.0411823987960815, "learning_rate": 1.4064993471637893e-05, "loss": 0.007, "step": 23794 }, { "epoch": 2.579683434518647, "grad_norm": 0.3605840802192688, "learning_rate": 1.4061366603800958e-05, "loss": 0.0084, "step": 23795 }, { "epoch": 2.579791847354727, "grad_norm": 0.15342113375663757, "learning_rate": 1.4057739735964023e-05, "loss": 0.0041, "step": 23796 }, { "epoch": 2.5799002601908065, "grad_norm": 0.04921441525220871, "learning_rate": 1.4054112868127087e-05, "loss": 0.0017, "step": 23797 }, { "epoch": 2.5800086730268865, "grad_norm": 0.0980360209941864, "learning_rate": 1.405048600029015e-05, "loss": 0.0025, "step": 23798 }, { "epoch": 2.580117085862966, "grad_norm": 0.11273305863142014, "learning_rate": 1.4046859132453214e-05, "loss": 0.0031, "step": 23799 }, { "epoch": 2.580225498699046, "grad_norm": 0.30744707584381104, "learning_rate": 1.4043232264616276e-05, "loss": 0.0098, "step": 23800 }, { "epoch": 2.580333911535126, "grad_norm": 0.5606127381324768, "learning_rate": 1.4039605396779342e-05, "loss": 0.008, "step": 23801 }, { "epoch": 2.5804423243712056, "grad_norm": 0.3817525804042816, "learning_rate": 1.4035978528942406e-05, "loss": 0.0328, "step": 23802 }, { "epoch": 2.580550737207285, "grad_norm": 0.2238970845937729, "learning_rate": 1.4032351661105469e-05, "loss": 0.0076, "step": 23803 }, { "epoch": 2.580659150043365, "grad_norm": 0.6183623671531677, "learning_rate": 1.4028724793268533e-05, "loss": 0.02, "step": 23804 }, { "epoch": 2.580767562879445, "grad_norm": 0.3222830891609192, "learning_rate": 1.4025097925431599e-05, "loss": 0.0181, "step": 23805 }, { "epoch": 2.5808759757155246, "grad_norm": 0.179033562541008, "learning_rate": 1.4021471057594663e-05, "loss": 0.0071, "step": 23806 }, { "epoch": 2.5809843885516046, "grad_norm": 0.17016485333442688, "learning_rate": 1.4017844189757725e-05, "loss": 0.008, "step": 23807 }, { "epoch": 2.581092801387684, "grad_norm": 0.5941089391708374, "learning_rate": 1.401421732192079e-05, "loss": 0.0276, "step": 23808 }, { "epoch": 2.581201214223764, "grad_norm": 0.4352600872516632, "learning_rate": 1.4010590454083855e-05, "loss": 0.0149, "step": 23809 }, { "epoch": 2.581309627059844, "grad_norm": 0.4511319696903229, "learning_rate": 1.4006963586246919e-05, "loss": 0.0111, "step": 23810 }, { "epoch": 2.5814180398959237, "grad_norm": 0.8210345506668091, "learning_rate": 1.4003336718409982e-05, "loss": 0.0103, "step": 23811 }, { "epoch": 2.581526452732003, "grad_norm": 0.05475064367055893, "learning_rate": 1.3999709850573046e-05, "loss": 0.0021, "step": 23812 }, { "epoch": 2.581634865568083, "grad_norm": 0.33422622084617615, "learning_rate": 1.3996082982736108e-05, "loss": 0.0188, "step": 23813 }, { "epoch": 2.581743278404163, "grad_norm": 0.02749072015285492, "learning_rate": 1.3992456114899174e-05, "loss": 0.0008, "step": 23814 }, { "epoch": 2.5818516912402427, "grad_norm": 0.18034668266773224, "learning_rate": 1.3988829247062238e-05, "loss": 0.0064, "step": 23815 }, { "epoch": 2.5819601040763227, "grad_norm": 1.4634475708007812, "learning_rate": 1.39852023792253e-05, "loss": 0.0199, "step": 23816 }, { "epoch": 2.5820685169124022, "grad_norm": 0.29428401589393616, "learning_rate": 1.3981575511388364e-05, "loss": 0.0095, "step": 23817 }, { "epoch": 2.5821769297484822, "grad_norm": 0.17958766222000122, "learning_rate": 1.397794864355143e-05, "loss": 0.0083, "step": 23818 }, { "epoch": 2.582285342584562, "grad_norm": 0.2435998171567917, "learning_rate": 1.3974321775714494e-05, "loss": 0.0044, "step": 23819 }, { "epoch": 2.5823937554206418, "grad_norm": 0.20599660277366638, "learning_rate": 1.3970694907877557e-05, "loss": 0.0053, "step": 23820 }, { "epoch": 2.5825021682567217, "grad_norm": 0.24100621044635773, "learning_rate": 1.3967068040040621e-05, "loss": 0.0045, "step": 23821 }, { "epoch": 2.5826105810928013, "grad_norm": 0.6487931609153748, "learning_rate": 1.3963441172203687e-05, "loss": 0.0195, "step": 23822 }, { "epoch": 2.5827189939288813, "grad_norm": 0.19276179373264313, "learning_rate": 1.395981430436675e-05, "loss": 0.0098, "step": 23823 }, { "epoch": 2.582827406764961, "grad_norm": 0.2885727286338806, "learning_rate": 1.3956187436529813e-05, "loss": 0.0134, "step": 23824 }, { "epoch": 2.582935819601041, "grad_norm": 0.1453784555196762, "learning_rate": 1.3952560568692877e-05, "loss": 0.0026, "step": 23825 }, { "epoch": 2.5830442324371203, "grad_norm": 0.07019221037626266, "learning_rate": 1.3948933700855943e-05, "loss": 0.004, "step": 23826 }, { "epoch": 2.5831526452732003, "grad_norm": 0.0939292162656784, "learning_rate": 1.3945306833019006e-05, "loss": 0.0019, "step": 23827 }, { "epoch": 2.5832610581092803, "grad_norm": 0.26141557097435, "learning_rate": 1.394167996518207e-05, "loss": 0.0145, "step": 23828 }, { "epoch": 2.58336947094536, "grad_norm": 0.2614768445491791, "learning_rate": 1.3938053097345132e-05, "loss": 0.0062, "step": 23829 }, { "epoch": 2.58347788378144, "grad_norm": 0.31071728467941284, "learning_rate": 1.3934426229508196e-05, "loss": 0.0101, "step": 23830 }, { "epoch": 2.5835862966175194, "grad_norm": 0.23615746200084686, "learning_rate": 1.3930799361671262e-05, "loss": 0.0064, "step": 23831 }, { "epoch": 2.5836947094535994, "grad_norm": 0.26311784982681274, "learning_rate": 1.3927172493834326e-05, "loss": 0.0259, "step": 23832 }, { "epoch": 2.5838031222896793, "grad_norm": 0.23364746570587158, "learning_rate": 1.3923545625997388e-05, "loss": 0.0269, "step": 23833 }, { "epoch": 2.583911535125759, "grad_norm": 0.56456059217453, "learning_rate": 1.3919918758160453e-05, "loss": 0.0183, "step": 23834 }, { "epoch": 2.5840199479618384, "grad_norm": 0.04037989675998688, "learning_rate": 1.3916291890323518e-05, "loss": 0.0017, "step": 23835 }, { "epoch": 2.5841283607979184, "grad_norm": 0.03386178985238075, "learning_rate": 1.3912665022486582e-05, "loss": 0.0009, "step": 23836 }, { "epoch": 2.5842367736339984, "grad_norm": 0.37873542308807373, "learning_rate": 1.3909038154649645e-05, "loss": 0.0216, "step": 23837 }, { "epoch": 2.584345186470078, "grad_norm": 0.014921490103006363, "learning_rate": 1.3905411286812709e-05, "loss": 0.0004, "step": 23838 }, { "epoch": 2.584453599306158, "grad_norm": 0.4418288469314575, "learning_rate": 1.3901784418975775e-05, "loss": 0.0143, "step": 23839 }, { "epoch": 2.5845620121422375, "grad_norm": 0.5723410248756409, "learning_rate": 1.3898157551138837e-05, "loss": 0.0553, "step": 23840 }, { "epoch": 2.5846704249783174, "grad_norm": 0.5305019617080688, "learning_rate": 1.3894530683301901e-05, "loss": 0.0143, "step": 23841 }, { "epoch": 2.5847788378143974, "grad_norm": 0.015943674370646477, "learning_rate": 1.3890903815464964e-05, "loss": 0.0005, "step": 23842 }, { "epoch": 2.584887250650477, "grad_norm": 0.11600498855113983, "learning_rate": 1.3887276947628028e-05, "loss": 0.0024, "step": 23843 }, { "epoch": 2.584995663486557, "grad_norm": 0.2210267335176468, "learning_rate": 1.3883650079791094e-05, "loss": 0.0132, "step": 23844 }, { "epoch": 2.5851040763226365, "grad_norm": 0.14595411717891693, "learning_rate": 1.3880023211954158e-05, "loss": 0.0068, "step": 23845 }, { "epoch": 2.5852124891587165, "grad_norm": 0.03402750939130783, "learning_rate": 1.387639634411722e-05, "loss": 0.001, "step": 23846 }, { "epoch": 2.585320901994796, "grad_norm": 0.11862139403820038, "learning_rate": 1.3872769476280284e-05, "loss": 0.0039, "step": 23847 }, { "epoch": 2.585429314830876, "grad_norm": 0.017138348892331123, "learning_rate": 1.386914260844335e-05, "loss": 0.0011, "step": 23848 }, { "epoch": 2.5855377276669556, "grad_norm": 0.016700511798262596, "learning_rate": 1.3865515740606414e-05, "loss": 0.0004, "step": 23849 }, { "epoch": 2.5856461405030355, "grad_norm": 0.34983059763908386, "learning_rate": 1.3861888872769477e-05, "loss": 0.0151, "step": 23850 }, { "epoch": 2.5857545533391155, "grad_norm": 0.026946743950247765, "learning_rate": 1.385826200493254e-05, "loss": 0.0009, "step": 23851 }, { "epoch": 2.585862966175195, "grad_norm": 0.6108907461166382, "learning_rate": 1.3854635137095606e-05, "loss": 0.012, "step": 23852 }, { "epoch": 2.585971379011275, "grad_norm": 0.2594529092311859, "learning_rate": 1.3851008269258669e-05, "loss": 0.0082, "step": 23853 }, { "epoch": 2.5860797918473546, "grad_norm": 0.17106463015079498, "learning_rate": 1.3847381401421733e-05, "loss": 0.0114, "step": 23854 }, { "epoch": 2.5861882046834346, "grad_norm": 0.10862597078084946, "learning_rate": 1.3843754533584795e-05, "loss": 0.0024, "step": 23855 }, { "epoch": 2.5862966175195146, "grad_norm": 0.0897291973233223, "learning_rate": 1.384012766574786e-05, "loss": 0.0022, "step": 23856 }, { "epoch": 2.586405030355594, "grad_norm": 0.030264871194958687, "learning_rate": 1.3836500797910925e-05, "loss": 0.0009, "step": 23857 }, { "epoch": 2.5865134431916736, "grad_norm": 0.11076536774635315, "learning_rate": 1.383287393007399e-05, "loss": 0.0097, "step": 23858 }, { "epoch": 2.5866218560277536, "grad_norm": 0.1894054114818573, "learning_rate": 1.3829247062237052e-05, "loss": 0.0063, "step": 23859 }, { "epoch": 2.5867302688638336, "grad_norm": 0.5465701818466187, "learning_rate": 1.3825620194400116e-05, "loss": 0.0139, "step": 23860 }, { "epoch": 2.586838681699913, "grad_norm": 0.018504494801163673, "learning_rate": 1.3821993326563182e-05, "loss": 0.0005, "step": 23861 }, { "epoch": 2.586947094535993, "grad_norm": 0.41407182812690735, "learning_rate": 1.3818366458726246e-05, "loss": 0.0127, "step": 23862 }, { "epoch": 2.5870555073720727, "grad_norm": 0.17830757796764374, "learning_rate": 1.3814739590889308e-05, "loss": 0.0051, "step": 23863 }, { "epoch": 2.5871639202081527, "grad_norm": 0.1885230988264084, "learning_rate": 1.3811112723052372e-05, "loss": 0.006, "step": 23864 }, { "epoch": 2.5872723330442327, "grad_norm": 0.6777321696281433, "learning_rate": 1.3807485855215438e-05, "loss": 0.0302, "step": 23865 }, { "epoch": 2.587380745880312, "grad_norm": 0.03352963924407959, "learning_rate": 1.38038589873785e-05, "loss": 0.0017, "step": 23866 }, { "epoch": 2.587489158716392, "grad_norm": 0.33073940873146057, "learning_rate": 1.3800232119541565e-05, "loss": 0.0222, "step": 23867 }, { "epoch": 2.5875975715524717, "grad_norm": 0.8848483562469482, "learning_rate": 1.3796605251704627e-05, "loss": 0.0466, "step": 23868 }, { "epoch": 2.5877059843885517, "grad_norm": 0.6612239480018616, "learning_rate": 1.3792978383867691e-05, "loss": 0.0458, "step": 23869 }, { "epoch": 2.5878143972246312, "grad_norm": 0.3963782489299774, "learning_rate": 1.3789351516030757e-05, "loss": 0.0195, "step": 23870 }, { "epoch": 2.5879228100607112, "grad_norm": 0.3876580595970154, "learning_rate": 1.3785724648193821e-05, "loss": 0.026, "step": 23871 }, { "epoch": 2.5880312228967908, "grad_norm": 0.3197879195213318, "learning_rate": 1.3782097780356883e-05, "loss": 0.008, "step": 23872 }, { "epoch": 2.5881396357328708, "grad_norm": 0.18346132338047028, "learning_rate": 1.3778470912519948e-05, "loss": 0.0068, "step": 23873 }, { "epoch": 2.5882480485689507, "grad_norm": 0.06076611205935478, "learning_rate": 1.3774844044683013e-05, "loss": 0.0024, "step": 23874 }, { "epoch": 2.5883564614050303, "grad_norm": 0.11039009690284729, "learning_rate": 1.3771217176846078e-05, "loss": 0.006, "step": 23875 }, { "epoch": 2.5884648742411103, "grad_norm": 0.5450807213783264, "learning_rate": 1.376759030900914e-05, "loss": 0.0154, "step": 23876 }, { "epoch": 2.58857328707719, "grad_norm": 0.22725938260555267, "learning_rate": 1.3763963441172204e-05, "loss": 0.0147, "step": 23877 }, { "epoch": 2.58868169991327, "grad_norm": 0.5411317348480225, "learning_rate": 1.376033657333527e-05, "loss": 0.0107, "step": 23878 }, { "epoch": 2.58879011274935, "grad_norm": 0.21731604635715485, "learning_rate": 1.3756709705498332e-05, "loss": 0.0069, "step": 23879 }, { "epoch": 2.5888985255854293, "grad_norm": 0.23796215653419495, "learning_rate": 1.3753082837661396e-05, "loss": 0.0084, "step": 23880 }, { "epoch": 2.589006938421509, "grad_norm": 0.4647291302680969, "learning_rate": 1.374945596982446e-05, "loss": 0.0145, "step": 23881 }, { "epoch": 2.589115351257589, "grad_norm": 0.023632202297449112, "learning_rate": 1.3745829101987523e-05, "loss": 0.0011, "step": 23882 }, { "epoch": 2.589223764093669, "grad_norm": 0.43425482511520386, "learning_rate": 1.3742202234150589e-05, "loss": 0.0673, "step": 23883 }, { "epoch": 2.5893321769297484, "grad_norm": 0.054404366761446, "learning_rate": 1.3738575366313653e-05, "loss": 0.0032, "step": 23884 }, { "epoch": 2.5894405897658284, "grad_norm": 0.6139491200447083, "learning_rate": 1.3734948498476715e-05, "loss": 0.0142, "step": 23885 }, { "epoch": 2.589549002601908, "grad_norm": 0.7434055209159851, "learning_rate": 1.373132163063978e-05, "loss": 0.0276, "step": 23886 }, { "epoch": 2.589657415437988, "grad_norm": 0.06577274203300476, "learning_rate": 1.3727694762802845e-05, "loss": 0.0039, "step": 23887 }, { "epoch": 2.589765828274068, "grad_norm": 0.4075237512588501, "learning_rate": 1.372406789496591e-05, "loss": 0.0408, "step": 23888 }, { "epoch": 2.5898742411101474, "grad_norm": 0.12258860468864441, "learning_rate": 1.3720441027128972e-05, "loss": 0.0106, "step": 23889 }, { "epoch": 2.5899826539462274, "grad_norm": 0.013304152525961399, "learning_rate": 1.3716814159292036e-05, "loss": 0.0004, "step": 23890 }, { "epoch": 2.590091066782307, "grad_norm": 0.15770290791988373, "learning_rate": 1.3713187291455101e-05, "loss": 0.0041, "step": 23891 }, { "epoch": 2.590199479618387, "grad_norm": 0.1489821970462799, "learning_rate": 1.3709560423618164e-05, "loss": 0.0049, "step": 23892 }, { "epoch": 2.5903078924544665, "grad_norm": 0.1631118208169937, "learning_rate": 1.3705933555781228e-05, "loss": 0.0057, "step": 23893 }, { "epoch": 2.5904163052905465, "grad_norm": 0.04940279573202133, "learning_rate": 1.3702306687944292e-05, "loss": 0.0021, "step": 23894 }, { "epoch": 2.590524718126626, "grad_norm": 0.36209794878959656, "learning_rate": 1.3698679820107355e-05, "loss": 0.0265, "step": 23895 }, { "epoch": 2.590633130962706, "grad_norm": 0.12696757912635803, "learning_rate": 1.369505295227042e-05, "loss": 0.0062, "step": 23896 }, { "epoch": 2.590741543798786, "grad_norm": 0.12037524580955505, "learning_rate": 1.3691426084433484e-05, "loss": 0.0039, "step": 23897 }, { "epoch": 2.5908499566348655, "grad_norm": 0.14705944061279297, "learning_rate": 1.3687799216596547e-05, "loss": 0.0027, "step": 23898 }, { "epoch": 2.5909583694709455, "grad_norm": 0.09756441414356232, "learning_rate": 1.3684172348759611e-05, "loss": 0.0029, "step": 23899 }, { "epoch": 2.591066782307025, "grad_norm": 0.8086994290351868, "learning_rate": 1.3680545480922677e-05, "loss": 0.0264, "step": 23900 }, { "epoch": 2.591175195143105, "grad_norm": 0.08972927927970886, "learning_rate": 1.3676918613085741e-05, "loss": 0.0025, "step": 23901 }, { "epoch": 2.591283607979185, "grad_norm": 0.49588602781295776, "learning_rate": 1.3673291745248803e-05, "loss": 0.0189, "step": 23902 }, { "epoch": 2.5913920208152645, "grad_norm": 0.3735921084880829, "learning_rate": 1.3669664877411867e-05, "loss": 0.0194, "step": 23903 }, { "epoch": 2.591500433651344, "grad_norm": 0.15307852625846863, "learning_rate": 1.3666038009574933e-05, "loss": 0.0078, "step": 23904 }, { "epoch": 2.591608846487424, "grad_norm": 0.1654791682958603, "learning_rate": 1.3662411141737996e-05, "loss": 0.0078, "step": 23905 }, { "epoch": 2.591717259323504, "grad_norm": 0.22268293797969818, "learning_rate": 1.365878427390106e-05, "loss": 0.0069, "step": 23906 }, { "epoch": 2.5918256721595836, "grad_norm": 0.39397305250167847, "learning_rate": 1.3655157406064124e-05, "loss": 0.0254, "step": 23907 }, { "epoch": 2.5919340849956636, "grad_norm": 0.3395386040210724, "learning_rate": 1.3651530538227186e-05, "loss": 0.0084, "step": 23908 }, { "epoch": 2.592042497831743, "grad_norm": 0.2868722975254059, "learning_rate": 1.3647903670390252e-05, "loss": 0.0195, "step": 23909 }, { "epoch": 2.592150910667823, "grad_norm": 0.015321282669901848, "learning_rate": 1.3644276802553316e-05, "loss": 0.0008, "step": 23910 }, { "epoch": 2.592259323503903, "grad_norm": 0.5079232454299927, "learning_rate": 1.3640649934716379e-05, "loss": 0.0369, "step": 23911 }, { "epoch": 2.5923677363399826, "grad_norm": 0.4992334246635437, "learning_rate": 1.3637023066879443e-05, "loss": 0.0546, "step": 23912 }, { "epoch": 2.592476149176062, "grad_norm": 0.30014631152153015, "learning_rate": 1.3633396199042508e-05, "loss": 0.0039, "step": 23913 }, { "epoch": 2.592584562012142, "grad_norm": 0.0597672276198864, "learning_rate": 1.3629769331205573e-05, "loss": 0.002, "step": 23914 }, { "epoch": 2.592692974848222, "grad_norm": 0.31187868118286133, "learning_rate": 1.3626142463368635e-05, "loss": 0.0146, "step": 23915 }, { "epoch": 2.5928013876843017, "grad_norm": 0.395920068025589, "learning_rate": 1.3622515595531699e-05, "loss": 0.055, "step": 23916 }, { "epoch": 2.5929098005203817, "grad_norm": 0.056353457272052765, "learning_rate": 1.3618888727694765e-05, "loss": 0.001, "step": 23917 }, { "epoch": 2.593018213356461, "grad_norm": 0.12910111248493195, "learning_rate": 1.3615261859857827e-05, "loss": 0.0078, "step": 23918 }, { "epoch": 2.593126626192541, "grad_norm": 0.21762387454509735, "learning_rate": 1.3611634992020891e-05, "loss": 0.0086, "step": 23919 }, { "epoch": 2.593235039028621, "grad_norm": 0.32253497838974, "learning_rate": 1.3608008124183955e-05, "loss": 0.0201, "step": 23920 }, { "epoch": 2.5933434518647007, "grad_norm": 0.6398048400878906, "learning_rate": 1.3604381256347018e-05, "loss": 0.033, "step": 23921 }, { "epoch": 2.5934518647007807, "grad_norm": 0.13111282885074615, "learning_rate": 1.3600754388510084e-05, "loss": 0.0053, "step": 23922 }, { "epoch": 2.5935602775368602, "grad_norm": 0.1726614236831665, "learning_rate": 1.3597127520673148e-05, "loss": 0.0056, "step": 23923 }, { "epoch": 2.5936686903729402, "grad_norm": 0.1856471300125122, "learning_rate": 1.359350065283621e-05, "loss": 0.014, "step": 23924 }, { "epoch": 2.59377710320902, "grad_norm": 0.5164024233818054, "learning_rate": 1.3589873784999274e-05, "loss": 0.0113, "step": 23925 }, { "epoch": 2.5938855160450998, "grad_norm": 0.7422977089881897, "learning_rate": 1.358624691716234e-05, "loss": 0.0584, "step": 23926 }, { "epoch": 2.5939939288811793, "grad_norm": 0.1470259577035904, "learning_rate": 1.3582620049325404e-05, "loss": 0.0095, "step": 23927 }, { "epoch": 2.5941023417172593, "grad_norm": 0.4759635031223297, "learning_rate": 1.3578993181488467e-05, "loss": 0.0266, "step": 23928 }, { "epoch": 2.5942107545533393, "grad_norm": 0.12655934691429138, "learning_rate": 1.357536631365153e-05, "loss": 0.0053, "step": 23929 }, { "epoch": 2.594319167389419, "grad_norm": 0.10873007029294968, "learning_rate": 1.3571739445814597e-05, "loss": 0.0043, "step": 23930 }, { "epoch": 2.594427580225499, "grad_norm": 0.02691095694899559, "learning_rate": 1.3568112577977659e-05, "loss": 0.0011, "step": 23931 }, { "epoch": 2.5945359930615783, "grad_norm": 0.572110652923584, "learning_rate": 1.3564485710140723e-05, "loss": 0.0266, "step": 23932 }, { "epoch": 2.5946444058976583, "grad_norm": 0.18879741430282593, "learning_rate": 1.3560858842303787e-05, "loss": 0.0083, "step": 23933 }, { "epoch": 2.5947528187337383, "grad_norm": 0.18675032258033752, "learning_rate": 1.3557231974466853e-05, "loss": 0.0054, "step": 23934 }, { "epoch": 2.594861231569818, "grad_norm": 0.014376387931406498, "learning_rate": 1.3553605106629915e-05, "loss": 0.0006, "step": 23935 }, { "epoch": 2.5949696444058974, "grad_norm": 0.6952183842658997, "learning_rate": 1.354997823879298e-05, "loss": 0.0321, "step": 23936 }, { "epoch": 2.5950780572419774, "grad_norm": 0.21785078942775726, "learning_rate": 1.3546351370956042e-05, "loss": 0.0235, "step": 23937 }, { "epoch": 2.5951864700780574, "grad_norm": 0.009434306994080544, "learning_rate": 1.3542724503119106e-05, "loss": 0.0004, "step": 23938 }, { "epoch": 2.595294882914137, "grad_norm": 0.08431539684534073, "learning_rate": 1.3539097635282172e-05, "loss": 0.0014, "step": 23939 }, { "epoch": 2.595403295750217, "grad_norm": 0.09145476669073105, "learning_rate": 1.3535470767445236e-05, "loss": 0.0023, "step": 23940 }, { "epoch": 2.5955117085862964, "grad_norm": 0.23279525339603424, "learning_rate": 1.3531843899608298e-05, "loss": 0.0117, "step": 23941 }, { "epoch": 2.5956201214223764, "grad_norm": 0.03839661180973053, "learning_rate": 1.3528217031771362e-05, "loss": 0.0014, "step": 23942 }, { "epoch": 2.5957285342584564, "grad_norm": 0.27187228202819824, "learning_rate": 1.3524590163934428e-05, "loss": 0.0184, "step": 23943 }, { "epoch": 2.595836947094536, "grad_norm": 0.19126306474208832, "learning_rate": 1.352096329609749e-05, "loss": 0.0105, "step": 23944 }, { "epoch": 2.595945359930616, "grad_norm": 0.40347418189048767, "learning_rate": 1.3517336428260555e-05, "loss": 0.0268, "step": 23945 }, { "epoch": 2.5960537727666955, "grad_norm": 0.009984714910387993, "learning_rate": 1.3513709560423619e-05, "loss": 0.0005, "step": 23946 }, { "epoch": 2.5961621856027755, "grad_norm": 0.016280658543109894, "learning_rate": 1.3510082692586685e-05, "loss": 0.0005, "step": 23947 }, { "epoch": 2.5962705984388554, "grad_norm": 0.1344170719385147, "learning_rate": 1.3506455824749747e-05, "loss": 0.0105, "step": 23948 }, { "epoch": 2.596379011274935, "grad_norm": 0.6798088550567627, "learning_rate": 1.3502828956912811e-05, "loss": 0.02, "step": 23949 }, { "epoch": 2.5964874241110145, "grad_norm": 0.21392838656902313, "learning_rate": 1.3499202089075874e-05, "loss": 0.0047, "step": 23950 }, { "epoch": 2.5965958369470945, "grad_norm": 0.19316396117210388, "learning_rate": 1.3495575221238938e-05, "loss": 0.0045, "step": 23951 }, { "epoch": 2.5967042497831745, "grad_norm": 0.10063201934099197, "learning_rate": 1.3491948353402003e-05, "loss": 0.0038, "step": 23952 }, { "epoch": 2.596812662619254, "grad_norm": 0.2061205804347992, "learning_rate": 1.3488321485565068e-05, "loss": 0.0107, "step": 23953 }, { "epoch": 2.596921075455334, "grad_norm": 0.4967925250530243, "learning_rate": 1.348469461772813e-05, "loss": 0.0202, "step": 23954 }, { "epoch": 2.5970294882914136, "grad_norm": 0.6851656436920166, "learning_rate": 1.3481067749891194e-05, "loss": 0.0182, "step": 23955 }, { "epoch": 2.5971379011274935, "grad_norm": 0.2293076515197754, "learning_rate": 1.347744088205426e-05, "loss": 0.0229, "step": 23956 }, { "epoch": 2.5972463139635735, "grad_norm": 0.047569140791893005, "learning_rate": 1.3473814014217322e-05, "loss": 0.0023, "step": 23957 }, { "epoch": 2.597354726799653, "grad_norm": 0.6059607863426208, "learning_rate": 1.3470187146380386e-05, "loss": 0.0121, "step": 23958 }, { "epoch": 2.5974631396357326, "grad_norm": 0.03380446881055832, "learning_rate": 1.346656027854345e-05, "loss": 0.0011, "step": 23959 }, { "epoch": 2.5975715524718126, "grad_norm": 0.06518231332302094, "learning_rate": 1.3462933410706516e-05, "loss": 0.0012, "step": 23960 }, { "epoch": 2.5976799653078926, "grad_norm": 0.12490937113761902, "learning_rate": 1.3459306542869579e-05, "loss": 0.0038, "step": 23961 }, { "epoch": 2.597788378143972, "grad_norm": 0.7789212465286255, "learning_rate": 1.3455679675032643e-05, "loss": 0.0077, "step": 23962 }, { "epoch": 2.597896790980052, "grad_norm": 0.008425794541835785, "learning_rate": 1.3452052807195705e-05, "loss": 0.0003, "step": 23963 }, { "epoch": 2.5980052038161316, "grad_norm": 0.04358787462115288, "learning_rate": 1.344842593935877e-05, "loss": 0.0016, "step": 23964 }, { "epoch": 2.5981136166522116, "grad_norm": 0.3890528976917267, "learning_rate": 1.3444799071521835e-05, "loss": 0.0099, "step": 23965 }, { "epoch": 2.5982220294882916, "grad_norm": 0.4602483808994293, "learning_rate": 1.34411722036849e-05, "loss": 0.0067, "step": 23966 }, { "epoch": 2.598330442324371, "grad_norm": 0.3956831097602844, "learning_rate": 1.3437545335847962e-05, "loss": 0.0093, "step": 23967 }, { "epoch": 2.598438855160451, "grad_norm": 0.24850444495677948, "learning_rate": 1.3433918468011026e-05, "loss": 0.0204, "step": 23968 }, { "epoch": 2.5985472679965307, "grad_norm": 0.2310355007648468, "learning_rate": 1.3430291600174092e-05, "loss": 0.0133, "step": 23969 }, { "epoch": 2.5986556808326107, "grad_norm": 0.411678284406662, "learning_rate": 1.3426664732337154e-05, "loss": 0.0246, "step": 23970 }, { "epoch": 2.59876409366869, "grad_norm": 0.2602326571941376, "learning_rate": 1.3423037864500218e-05, "loss": 0.014, "step": 23971 }, { "epoch": 2.59887250650477, "grad_norm": 0.04545233026146889, "learning_rate": 1.3419410996663282e-05, "loss": 0.0025, "step": 23972 }, { "epoch": 2.5989809193408497, "grad_norm": 0.5619577169418335, "learning_rate": 1.3415784128826348e-05, "loss": 0.0255, "step": 23973 }, { "epoch": 2.5990893321769297, "grad_norm": 0.22258128225803375, "learning_rate": 1.341215726098941e-05, "loss": 0.0121, "step": 23974 }, { "epoch": 2.5991977450130097, "grad_norm": 0.16828438639640808, "learning_rate": 1.3408530393152474e-05, "loss": 0.0077, "step": 23975 }, { "epoch": 2.5993061578490892, "grad_norm": 0.35026291012763977, "learning_rate": 1.3404903525315537e-05, "loss": 0.0171, "step": 23976 }, { "epoch": 2.5994145706851692, "grad_norm": 0.3450620472431183, "learning_rate": 1.3401276657478601e-05, "loss": 0.0059, "step": 23977 }, { "epoch": 2.5995229835212488, "grad_norm": 0.4980635941028595, "learning_rate": 1.3397649789641667e-05, "loss": 0.0331, "step": 23978 }, { "epoch": 2.5996313963573288, "grad_norm": 0.06660082936286926, "learning_rate": 1.3394022921804731e-05, "loss": 0.0015, "step": 23979 }, { "epoch": 2.5997398091934087, "grad_norm": 0.3166588842868805, "learning_rate": 1.3390396053967793e-05, "loss": 0.0191, "step": 23980 }, { "epoch": 2.5998482220294883, "grad_norm": 0.11129062622785568, "learning_rate": 1.3386769186130857e-05, "loss": 0.0048, "step": 23981 }, { "epoch": 2.599956634865568, "grad_norm": 0.0026050712913274765, "learning_rate": 1.3383142318293923e-05, "loss": 0.0001, "step": 23982 }, { "epoch": 2.600065047701648, "grad_norm": 0.02499811165034771, "learning_rate": 1.3379515450456986e-05, "loss": 0.0009, "step": 23983 }, { "epoch": 2.600173460537728, "grad_norm": 0.09370657801628113, "learning_rate": 1.337588858262005e-05, "loss": 0.0012, "step": 23984 }, { "epoch": 2.6002818733738073, "grad_norm": 0.38371601700782776, "learning_rate": 1.3372261714783114e-05, "loss": 0.0066, "step": 23985 }, { "epoch": 2.6003902862098873, "grad_norm": 0.20634916424751282, "learning_rate": 1.336863484694618e-05, "loss": 0.0049, "step": 23986 }, { "epoch": 2.600498699045967, "grad_norm": 0.11828683316707611, "learning_rate": 1.3365007979109242e-05, "loss": 0.008, "step": 23987 }, { "epoch": 2.600607111882047, "grad_norm": 0.19016744196414948, "learning_rate": 1.3361381111272306e-05, "loss": 0.0068, "step": 23988 }, { "epoch": 2.600715524718127, "grad_norm": 0.44991788268089294, "learning_rate": 1.3357754243435369e-05, "loss": 0.0315, "step": 23989 }, { "epoch": 2.6008239375542064, "grad_norm": 0.03528974950313568, "learning_rate": 1.3354127375598433e-05, "loss": 0.001, "step": 23990 }, { "epoch": 2.6009323503902864, "grad_norm": 0.1590517908334732, "learning_rate": 1.3350500507761498e-05, "loss": 0.0088, "step": 23991 }, { "epoch": 2.601040763226366, "grad_norm": 0.01107184961438179, "learning_rate": 1.3346873639924563e-05, "loss": 0.0003, "step": 23992 }, { "epoch": 2.601149176062446, "grad_norm": 0.2147321254014969, "learning_rate": 1.3343246772087625e-05, "loss": 0.0163, "step": 23993 }, { "epoch": 2.6012575888985254, "grad_norm": 0.02246036008000374, "learning_rate": 1.3339619904250689e-05, "loss": 0.0008, "step": 23994 }, { "epoch": 2.6013660017346054, "grad_norm": 0.09543959051370621, "learning_rate": 1.3335993036413755e-05, "loss": 0.0025, "step": 23995 }, { "epoch": 2.601474414570685, "grad_norm": 0.1554953157901764, "learning_rate": 1.3332366168576817e-05, "loss": 0.0066, "step": 23996 }, { "epoch": 2.601582827406765, "grad_norm": 0.9672561883926392, "learning_rate": 1.3328739300739881e-05, "loss": 0.012, "step": 23997 }, { "epoch": 2.601691240242845, "grad_norm": 0.13981643319129944, "learning_rate": 1.3325112432902945e-05, "loss": 0.0046, "step": 23998 }, { "epoch": 2.6017996530789245, "grad_norm": 0.02348262630403042, "learning_rate": 1.3321485565066011e-05, "loss": 0.0009, "step": 23999 }, { "epoch": 2.6019080659150045, "grad_norm": 0.015751926228404045, "learning_rate": 1.3317858697229074e-05, "loss": 0.0002, "step": 24000 }, { "epoch": 2.602016478751084, "grad_norm": 0.10289165377616882, "learning_rate": 1.3314231829392138e-05, "loss": 0.0024, "step": 24001 }, { "epoch": 2.602124891587164, "grad_norm": 0.022018052637577057, "learning_rate": 1.33106049615552e-05, "loss": 0.0008, "step": 24002 }, { "epoch": 2.602233304423244, "grad_norm": 0.309540718793869, "learning_rate": 1.3306978093718264e-05, "loss": 0.0145, "step": 24003 }, { "epoch": 2.6023417172593235, "grad_norm": 0.0666707381606102, "learning_rate": 1.330335122588133e-05, "loss": 0.0022, "step": 24004 }, { "epoch": 2.602450130095403, "grad_norm": 0.9251570105552673, "learning_rate": 1.3299724358044394e-05, "loss": 0.0187, "step": 24005 }, { "epoch": 2.602558542931483, "grad_norm": 0.3972999155521393, "learning_rate": 1.3296097490207457e-05, "loss": 0.038, "step": 24006 }, { "epoch": 2.602666955767563, "grad_norm": 0.39334383606910706, "learning_rate": 1.329247062237052e-05, "loss": 0.0104, "step": 24007 }, { "epoch": 2.6027753686036426, "grad_norm": 0.08422704041004181, "learning_rate": 1.3288843754533587e-05, "loss": 0.0031, "step": 24008 }, { "epoch": 2.6028837814397225, "grad_norm": 0.18223848938941956, "learning_rate": 1.3285216886696649e-05, "loss": 0.0123, "step": 24009 }, { "epoch": 2.602992194275802, "grad_norm": 0.7465641498565674, "learning_rate": 1.3281590018859713e-05, "loss": 0.0197, "step": 24010 }, { "epoch": 2.603100607111882, "grad_norm": 0.49200016260147095, "learning_rate": 1.3277963151022777e-05, "loss": 0.0108, "step": 24011 }, { "epoch": 2.603209019947962, "grad_norm": 0.4997461438179016, "learning_rate": 1.3274336283185843e-05, "loss": 0.024, "step": 24012 }, { "epoch": 2.6033174327840416, "grad_norm": 0.4846801161766052, "learning_rate": 1.3270709415348905e-05, "loss": 0.0144, "step": 24013 }, { "epoch": 2.6034258456201216, "grad_norm": 0.059265509247779846, "learning_rate": 1.326708254751197e-05, "loss": 0.003, "step": 24014 }, { "epoch": 2.603534258456201, "grad_norm": 0.015779336914420128, "learning_rate": 1.3263455679675032e-05, "loss": 0.0006, "step": 24015 }, { "epoch": 2.603642671292281, "grad_norm": 0.06765997409820557, "learning_rate": 1.3259828811838096e-05, "loss": 0.0027, "step": 24016 }, { "epoch": 2.6037510841283606, "grad_norm": 0.06482914835214615, "learning_rate": 1.3256201944001162e-05, "loss": 0.0018, "step": 24017 }, { "epoch": 2.6038594969644406, "grad_norm": 0.059905365109443665, "learning_rate": 1.3252575076164226e-05, "loss": 0.0024, "step": 24018 }, { "epoch": 2.60396790980052, "grad_norm": 0.8202674388885498, "learning_rate": 1.3248948208327288e-05, "loss": 0.0244, "step": 24019 }, { "epoch": 2.6040763226366, "grad_norm": 0.27958497405052185, "learning_rate": 1.3245321340490352e-05, "loss": 0.0025, "step": 24020 }, { "epoch": 2.60418473547268, "grad_norm": 0.04582122713327408, "learning_rate": 1.3241694472653418e-05, "loss": 0.001, "step": 24021 }, { "epoch": 2.6042931483087597, "grad_norm": 0.6143049001693726, "learning_rate": 1.323806760481648e-05, "loss": 0.021, "step": 24022 }, { "epoch": 2.6044015611448397, "grad_norm": 0.01973331905901432, "learning_rate": 1.3234440736979545e-05, "loss": 0.0005, "step": 24023 }, { "epoch": 2.604509973980919, "grad_norm": 0.5769267082214355, "learning_rate": 1.3230813869142609e-05, "loss": 0.0289, "step": 24024 }, { "epoch": 2.604618386816999, "grad_norm": 0.276274174451828, "learning_rate": 1.3227187001305675e-05, "loss": 0.0122, "step": 24025 }, { "epoch": 2.604726799653079, "grad_norm": 0.36149856448173523, "learning_rate": 1.3223560133468737e-05, "loss": 0.0384, "step": 24026 }, { "epoch": 2.6048352124891587, "grad_norm": 0.08762425929307938, "learning_rate": 1.3219933265631801e-05, "loss": 0.0027, "step": 24027 }, { "epoch": 2.6049436253252383, "grad_norm": 0.0018650508718565106, "learning_rate": 1.3216306397794864e-05, "loss": 0.0001, "step": 24028 }, { "epoch": 2.6050520381613183, "grad_norm": 0.32152116298675537, "learning_rate": 1.3212679529957931e-05, "loss": 0.0067, "step": 24029 }, { "epoch": 2.6051604509973982, "grad_norm": 0.29629695415496826, "learning_rate": 1.3209052662120993e-05, "loss": 0.0078, "step": 24030 }, { "epoch": 2.6052688638334778, "grad_norm": 0.4659564793109894, "learning_rate": 1.3205425794284058e-05, "loss": 0.0189, "step": 24031 }, { "epoch": 2.6053772766695578, "grad_norm": 0.19969774782657623, "learning_rate": 1.320179892644712e-05, "loss": 0.0095, "step": 24032 }, { "epoch": 2.6054856895056373, "grad_norm": 0.42648953199386597, "learning_rate": 1.3198172058610184e-05, "loss": 0.0127, "step": 24033 }, { "epoch": 2.6055941023417173, "grad_norm": 0.11022477596998215, "learning_rate": 1.319454519077325e-05, "loss": 0.0025, "step": 24034 }, { "epoch": 2.6057025151777973, "grad_norm": 0.06444721668958664, "learning_rate": 1.3190918322936312e-05, "loss": 0.0023, "step": 24035 }, { "epoch": 2.605810928013877, "grad_norm": 0.21592694520950317, "learning_rate": 1.3187291455099376e-05, "loss": 0.0051, "step": 24036 }, { "epoch": 2.605919340849957, "grad_norm": 0.10118419677019119, "learning_rate": 1.318366458726244e-05, "loss": 0.0049, "step": 24037 }, { "epoch": 2.6060277536860363, "grad_norm": 0.3248506486415863, "learning_rate": 1.3180037719425506e-05, "loss": 0.0305, "step": 24038 }, { "epoch": 2.6061361665221163, "grad_norm": 0.26416629552841187, "learning_rate": 1.3176410851588569e-05, "loss": 0.0047, "step": 24039 }, { "epoch": 2.606244579358196, "grad_norm": 0.23535284399986267, "learning_rate": 1.3172783983751633e-05, "loss": 0.0096, "step": 24040 }, { "epoch": 2.606352992194276, "grad_norm": 0.18370957672595978, "learning_rate": 1.3169157115914695e-05, "loss": 0.0064, "step": 24041 }, { "epoch": 2.6064614050303554, "grad_norm": 0.02644202671945095, "learning_rate": 1.3165530248077763e-05, "loss": 0.001, "step": 24042 }, { "epoch": 2.6065698178664354, "grad_norm": 0.20157356560230255, "learning_rate": 1.3161903380240825e-05, "loss": 0.0292, "step": 24043 }, { "epoch": 2.6066782307025154, "grad_norm": 0.07034890353679657, "learning_rate": 1.315827651240389e-05, "loss": 0.0013, "step": 24044 }, { "epoch": 2.606786643538595, "grad_norm": 0.21106283366680145, "learning_rate": 1.3154649644566952e-05, "loss": 0.0078, "step": 24045 }, { "epoch": 2.606895056374675, "grad_norm": 0.01794479414820671, "learning_rate": 1.3151022776730016e-05, "loss": 0.0007, "step": 24046 }, { "epoch": 2.6070034692107544, "grad_norm": 0.17623670399188995, "learning_rate": 1.3147395908893082e-05, "loss": 0.0056, "step": 24047 }, { "epoch": 2.6071118820468344, "grad_norm": 0.35874104499816895, "learning_rate": 1.3143769041056144e-05, "loss": 0.0114, "step": 24048 }, { "epoch": 2.6072202948829144, "grad_norm": 0.1337415724992752, "learning_rate": 1.3140142173219208e-05, "loss": 0.0042, "step": 24049 }, { "epoch": 2.607328707718994, "grad_norm": 0.558815062046051, "learning_rate": 1.3136515305382272e-05, "loss": 0.0262, "step": 24050 }, { "epoch": 2.6074371205550735, "grad_norm": 0.1353435516357422, "learning_rate": 1.3132888437545338e-05, "loss": 0.0074, "step": 24051 }, { "epoch": 2.6075455333911535, "grad_norm": 0.0924317017197609, "learning_rate": 1.31292615697084e-05, "loss": 0.0014, "step": 24052 }, { "epoch": 2.6076539462272335, "grad_norm": 0.02901453897356987, "learning_rate": 1.3125634701871464e-05, "loss": 0.0018, "step": 24053 }, { "epoch": 2.607762359063313, "grad_norm": 0.1915954053401947, "learning_rate": 1.3122007834034527e-05, "loss": 0.0079, "step": 24054 }, { "epoch": 2.607870771899393, "grad_norm": 0.06684299558401108, "learning_rate": 1.3118380966197594e-05, "loss": 0.002, "step": 24055 }, { "epoch": 2.6079791847354725, "grad_norm": 0.022361116483807564, "learning_rate": 1.3114754098360657e-05, "loss": 0.0009, "step": 24056 }, { "epoch": 2.6080875975715525, "grad_norm": 0.3118404448032379, "learning_rate": 1.3111127230523721e-05, "loss": 0.0161, "step": 24057 }, { "epoch": 2.6081960104076325, "grad_norm": 0.25268658995628357, "learning_rate": 1.3107500362686783e-05, "loss": 0.007, "step": 24058 }, { "epoch": 2.608304423243712, "grad_norm": 0.054708924144506454, "learning_rate": 1.3103873494849847e-05, "loss": 0.001, "step": 24059 }, { "epoch": 2.6084128360797916, "grad_norm": 0.3605562746524811, "learning_rate": 1.3100246627012913e-05, "loss": 0.0104, "step": 24060 }, { "epoch": 2.6085212489158716, "grad_norm": 0.04808741807937622, "learning_rate": 1.3096619759175977e-05, "loss": 0.0016, "step": 24061 }, { "epoch": 2.6086296617519515, "grad_norm": 0.32586151361465454, "learning_rate": 1.309299289133904e-05, "loss": 0.022, "step": 24062 }, { "epoch": 2.608738074588031, "grad_norm": 0.3389526307582855, "learning_rate": 1.3089366023502104e-05, "loss": 0.0207, "step": 24063 }, { "epoch": 2.608846487424111, "grad_norm": 0.2645746171474457, "learning_rate": 1.308573915566517e-05, "loss": 0.0082, "step": 24064 }, { "epoch": 2.6089549002601906, "grad_norm": 0.040750570595264435, "learning_rate": 1.3082112287828232e-05, "loss": 0.0012, "step": 24065 }, { "epoch": 2.6090633130962706, "grad_norm": 0.3541122078895569, "learning_rate": 1.3078485419991296e-05, "loss": 0.0061, "step": 24066 }, { "epoch": 2.6091717259323506, "grad_norm": 0.19668051600456238, "learning_rate": 1.3074858552154359e-05, "loss": 0.0036, "step": 24067 }, { "epoch": 2.60928013876843, "grad_norm": 1.1837894916534424, "learning_rate": 1.3071231684317426e-05, "loss": 0.0273, "step": 24068 }, { "epoch": 2.60938855160451, "grad_norm": 0.010927073657512665, "learning_rate": 1.3067604816480488e-05, "loss": 0.0006, "step": 24069 }, { "epoch": 2.6094969644405897, "grad_norm": 0.06434771418571472, "learning_rate": 1.3063977948643553e-05, "loss": 0.0014, "step": 24070 }, { "epoch": 2.6096053772766696, "grad_norm": 0.1786927580833435, "learning_rate": 1.3060351080806615e-05, "loss": 0.0073, "step": 24071 }, { "epoch": 2.6097137901127496, "grad_norm": 0.015906954184174538, "learning_rate": 1.3056724212969679e-05, "loss": 0.0007, "step": 24072 }, { "epoch": 2.609822202948829, "grad_norm": 0.5089921355247498, "learning_rate": 1.3053097345132745e-05, "loss": 0.0109, "step": 24073 }, { "epoch": 2.6099306157849087, "grad_norm": 0.5193244814872742, "learning_rate": 1.3049470477295809e-05, "loss": 0.0183, "step": 24074 }, { "epoch": 2.6100390286209887, "grad_norm": 0.10426349192857742, "learning_rate": 1.3045843609458871e-05, "loss": 0.0029, "step": 24075 }, { "epoch": 2.6101474414570687, "grad_norm": 0.021388497203588486, "learning_rate": 1.3042216741621936e-05, "loss": 0.0014, "step": 24076 }, { "epoch": 2.610255854293148, "grad_norm": 0.6344500780105591, "learning_rate": 1.3038589873785001e-05, "loss": 0.0234, "step": 24077 }, { "epoch": 2.610364267129228, "grad_norm": 0.4677828550338745, "learning_rate": 1.3034963005948064e-05, "loss": 0.0138, "step": 24078 }, { "epoch": 2.6104726799653077, "grad_norm": 0.3088289499282837, "learning_rate": 1.3031336138111128e-05, "loss": 0.0282, "step": 24079 }, { "epoch": 2.6105810928013877, "grad_norm": 0.5633017420768738, "learning_rate": 1.302770927027419e-05, "loss": 0.0149, "step": 24080 }, { "epoch": 2.6106895056374677, "grad_norm": 0.2860667407512665, "learning_rate": 1.3024082402437258e-05, "loss": 0.0081, "step": 24081 }, { "epoch": 2.6107979184735473, "grad_norm": 0.10598216950893402, "learning_rate": 1.302045553460032e-05, "loss": 0.0018, "step": 24082 }, { "epoch": 2.610906331309627, "grad_norm": 0.5636740922927856, "learning_rate": 1.3016828666763384e-05, "loss": 0.0413, "step": 24083 }, { "epoch": 2.611014744145707, "grad_norm": 0.7375160455703735, "learning_rate": 1.3013201798926447e-05, "loss": 0.013, "step": 24084 }, { "epoch": 2.6111231569817868, "grad_norm": 0.13861405849456787, "learning_rate": 1.300957493108951e-05, "loss": 0.0044, "step": 24085 }, { "epoch": 2.6112315698178663, "grad_norm": 0.35284239053726196, "learning_rate": 1.3005948063252577e-05, "loss": 0.0288, "step": 24086 }, { "epoch": 2.6113399826539463, "grad_norm": 0.20866531133651733, "learning_rate": 1.300232119541564e-05, "loss": 0.0066, "step": 24087 }, { "epoch": 2.611448395490026, "grad_norm": 0.11406560242176056, "learning_rate": 1.2998694327578703e-05, "loss": 0.004, "step": 24088 }, { "epoch": 2.611556808326106, "grad_norm": 0.20369325578212738, "learning_rate": 1.2995067459741767e-05, "loss": 0.0134, "step": 24089 }, { "epoch": 2.611665221162186, "grad_norm": 0.4080032408237457, "learning_rate": 1.2991440591904833e-05, "loss": 0.0092, "step": 24090 }, { "epoch": 2.6117736339982653, "grad_norm": 1.158424735069275, "learning_rate": 1.2987813724067895e-05, "loss": 0.0554, "step": 24091 }, { "epoch": 2.6118820468343453, "grad_norm": 0.1135537326335907, "learning_rate": 1.298418685623096e-05, "loss": 0.0025, "step": 24092 }, { "epoch": 2.611990459670425, "grad_norm": 0.007083368021994829, "learning_rate": 1.2980559988394022e-05, "loss": 0.0005, "step": 24093 }, { "epoch": 2.612098872506505, "grad_norm": 0.4624654948711395, "learning_rate": 1.297693312055709e-05, "loss": 0.0259, "step": 24094 }, { "epoch": 2.612207285342585, "grad_norm": 0.2748902440071106, "learning_rate": 1.2973306252720152e-05, "loss": 0.0077, "step": 24095 }, { "epoch": 2.6123156981786644, "grad_norm": 0.39385026693344116, "learning_rate": 1.2969679384883216e-05, "loss": 0.0079, "step": 24096 }, { "epoch": 2.612424111014744, "grad_norm": 0.3795577585697174, "learning_rate": 1.2966052517046278e-05, "loss": 0.0151, "step": 24097 }, { "epoch": 2.612532523850824, "grad_norm": 0.1565268635749817, "learning_rate": 1.2962425649209342e-05, "loss": 0.0057, "step": 24098 }, { "epoch": 2.612640936686904, "grad_norm": 0.2015109807252884, "learning_rate": 1.2958798781372408e-05, "loss": 0.0093, "step": 24099 }, { "epoch": 2.6127493495229834, "grad_norm": 0.039048757404088974, "learning_rate": 1.2955171913535472e-05, "loss": 0.0016, "step": 24100 }, { "epoch": 2.6128577623590634, "grad_norm": 0.14922481775283813, "learning_rate": 1.2951545045698535e-05, "loss": 0.0021, "step": 24101 }, { "epoch": 2.612966175195143, "grad_norm": 0.09127446264028549, "learning_rate": 1.2947918177861599e-05, "loss": 0.0047, "step": 24102 }, { "epoch": 2.613074588031223, "grad_norm": 0.020023485645651817, "learning_rate": 1.2944291310024665e-05, "loss": 0.0007, "step": 24103 }, { "epoch": 2.613183000867303, "grad_norm": 0.9014489650726318, "learning_rate": 1.2940664442187727e-05, "loss": 0.0582, "step": 24104 }, { "epoch": 2.6132914137033825, "grad_norm": 0.08225574344396591, "learning_rate": 1.2937037574350791e-05, "loss": 0.0029, "step": 24105 }, { "epoch": 2.613399826539462, "grad_norm": 0.05318531394004822, "learning_rate": 1.2933410706513854e-05, "loss": 0.0007, "step": 24106 }, { "epoch": 2.613508239375542, "grad_norm": 0.2932349145412445, "learning_rate": 1.2929783838676921e-05, "loss": 0.0091, "step": 24107 }, { "epoch": 2.613616652211622, "grad_norm": 0.1321103721857071, "learning_rate": 1.2926156970839983e-05, "loss": 0.0023, "step": 24108 }, { "epoch": 2.6137250650477015, "grad_norm": 1.0777431726455688, "learning_rate": 1.2922530103003048e-05, "loss": 0.0175, "step": 24109 }, { "epoch": 2.6138334778837815, "grad_norm": 0.020178139209747314, "learning_rate": 1.291890323516611e-05, "loss": 0.0005, "step": 24110 }, { "epoch": 2.613941890719861, "grad_norm": 0.04598291590809822, "learning_rate": 1.2915276367329174e-05, "loss": 0.0023, "step": 24111 }, { "epoch": 2.614050303555941, "grad_norm": 0.02906941808760166, "learning_rate": 1.291164949949224e-05, "loss": 0.0009, "step": 24112 }, { "epoch": 2.614158716392021, "grad_norm": 0.02997991442680359, "learning_rate": 1.2908022631655304e-05, "loss": 0.0009, "step": 24113 }, { "epoch": 2.6142671292281006, "grad_norm": 0.47035014629364014, "learning_rate": 1.2904395763818366e-05, "loss": 0.0133, "step": 24114 }, { "epoch": 2.6143755420641805, "grad_norm": 0.22033606469631195, "learning_rate": 1.290076889598143e-05, "loss": 0.0039, "step": 24115 }, { "epoch": 2.61448395490026, "grad_norm": 0.27487248182296753, "learning_rate": 1.2897142028144496e-05, "loss": 0.0299, "step": 24116 }, { "epoch": 2.61459236773634, "grad_norm": 0.055790647864341736, "learning_rate": 1.2893515160307559e-05, "loss": 0.0028, "step": 24117 }, { "epoch": 2.6147007805724196, "grad_norm": 0.05665529891848564, "learning_rate": 1.2889888292470623e-05, "loss": 0.0013, "step": 24118 }, { "epoch": 2.6148091934084996, "grad_norm": 0.7495352625846863, "learning_rate": 1.2886261424633685e-05, "loss": 0.0559, "step": 24119 }, { "epoch": 2.614917606244579, "grad_norm": 0.39915770292282104, "learning_rate": 1.2882634556796753e-05, "loss": 0.0346, "step": 24120 }, { "epoch": 2.615026019080659, "grad_norm": 0.03720659390091896, "learning_rate": 1.2879007688959815e-05, "loss": 0.0013, "step": 24121 }, { "epoch": 2.615134431916739, "grad_norm": 0.11730312556028366, "learning_rate": 1.287538082112288e-05, "loss": 0.0021, "step": 24122 }, { "epoch": 2.6152428447528187, "grad_norm": 0.10856456309556961, "learning_rate": 1.2871753953285942e-05, "loss": 0.0034, "step": 24123 }, { "epoch": 2.6153512575888986, "grad_norm": 0.180314838886261, "learning_rate": 1.2868127085449006e-05, "loss": 0.0122, "step": 24124 }, { "epoch": 2.615459670424978, "grad_norm": 0.19076012074947357, "learning_rate": 1.2864500217612072e-05, "loss": 0.0047, "step": 24125 }, { "epoch": 2.615568083261058, "grad_norm": 0.2223639339208603, "learning_rate": 1.2860873349775136e-05, "loss": 0.0259, "step": 24126 }, { "epoch": 2.615676496097138, "grad_norm": 0.2382490634918213, "learning_rate": 1.2857246481938198e-05, "loss": 0.0091, "step": 24127 }, { "epoch": 2.6157849089332177, "grad_norm": 0.3718857765197754, "learning_rate": 1.2853619614101262e-05, "loss": 0.017, "step": 24128 }, { "epoch": 2.6158933217692972, "grad_norm": 0.03167646750807762, "learning_rate": 1.2849992746264328e-05, "loss": 0.0011, "step": 24129 }, { "epoch": 2.616001734605377, "grad_norm": 0.369290292263031, "learning_rate": 1.284636587842739e-05, "loss": 0.0094, "step": 24130 }, { "epoch": 2.616110147441457, "grad_norm": 0.35397201776504517, "learning_rate": 1.2842739010590455e-05, "loss": 0.0174, "step": 24131 }, { "epoch": 2.6162185602775367, "grad_norm": 0.0765848383307457, "learning_rate": 1.2839112142753517e-05, "loss": 0.0022, "step": 24132 }, { "epoch": 2.6163269731136167, "grad_norm": 0.609891951084137, "learning_rate": 1.2835485274916584e-05, "loss": 0.02, "step": 24133 }, { "epoch": 2.6164353859496963, "grad_norm": 0.22036655247211456, "learning_rate": 1.2831858407079647e-05, "loss": 0.0058, "step": 24134 }, { "epoch": 2.6165437987857763, "grad_norm": 0.26396334171295166, "learning_rate": 1.2828231539242711e-05, "loss": 0.0106, "step": 24135 }, { "epoch": 2.6166522116218562, "grad_norm": 0.5813115835189819, "learning_rate": 1.2824604671405773e-05, "loss": 0.0048, "step": 24136 }, { "epoch": 2.616760624457936, "grad_norm": 0.5367153882980347, "learning_rate": 1.282097780356884e-05, "loss": 0.0614, "step": 24137 }, { "epoch": 2.6168690372940158, "grad_norm": 0.26954251527786255, "learning_rate": 1.2817350935731903e-05, "loss": 0.0109, "step": 24138 }, { "epoch": 2.6169774501300953, "grad_norm": 1.2367072105407715, "learning_rate": 1.2813724067894967e-05, "loss": 0.0162, "step": 24139 }, { "epoch": 2.6170858629661753, "grad_norm": 2.1014082431793213, "learning_rate": 1.281009720005803e-05, "loss": 0.0276, "step": 24140 }, { "epoch": 2.617194275802255, "grad_norm": 0.2530899941921234, "learning_rate": 1.2806470332221094e-05, "loss": 0.006, "step": 24141 }, { "epoch": 2.617302688638335, "grad_norm": 0.7558504939079285, "learning_rate": 1.280284346438416e-05, "loss": 0.0462, "step": 24142 }, { "epoch": 2.6174111014744144, "grad_norm": 0.07500448822975159, "learning_rate": 1.2799216596547222e-05, "loss": 0.0019, "step": 24143 }, { "epoch": 2.6175195143104943, "grad_norm": 0.0453970804810524, "learning_rate": 1.2795589728710286e-05, "loss": 0.0011, "step": 24144 }, { "epoch": 2.6176279271465743, "grad_norm": 0.31149914860725403, "learning_rate": 1.2791962860873349e-05, "loss": 0.0164, "step": 24145 }, { "epoch": 2.617736339982654, "grad_norm": 0.35986363887786865, "learning_rate": 1.2788335993036416e-05, "loss": 0.0041, "step": 24146 }, { "epoch": 2.617844752818734, "grad_norm": 0.5180646777153015, "learning_rate": 1.2784709125199479e-05, "loss": 0.051, "step": 24147 }, { "epoch": 2.6179531656548134, "grad_norm": 0.40494048595428467, "learning_rate": 1.2781082257362543e-05, "loss": 0.0197, "step": 24148 }, { "epoch": 2.6180615784908934, "grad_norm": 0.5414285659790039, "learning_rate": 1.2777455389525605e-05, "loss": 0.0229, "step": 24149 }, { "epoch": 2.6181699913269734, "grad_norm": 0.007616163697093725, "learning_rate": 1.277382852168867e-05, "loss": 0.0003, "step": 24150 }, { "epoch": 2.618278404163053, "grad_norm": 0.710097074508667, "learning_rate": 1.2770201653851735e-05, "loss": 0.0247, "step": 24151 }, { "epoch": 2.6183868169991324, "grad_norm": 0.16046226024627686, "learning_rate": 1.2766574786014799e-05, "loss": 0.0033, "step": 24152 }, { "epoch": 2.6184952298352124, "grad_norm": 0.23853690922260284, "learning_rate": 1.2762947918177861e-05, "loss": 0.008, "step": 24153 }, { "epoch": 2.6186036426712924, "grad_norm": 0.1403616964817047, "learning_rate": 1.2759321050340926e-05, "loss": 0.0052, "step": 24154 }, { "epoch": 2.618712055507372, "grad_norm": 0.23902131617069244, "learning_rate": 1.2755694182503991e-05, "loss": 0.0126, "step": 24155 }, { "epoch": 2.618820468343452, "grad_norm": 0.313058078289032, "learning_rate": 1.2752067314667054e-05, "loss": 0.0023, "step": 24156 }, { "epoch": 2.6189288811795315, "grad_norm": 0.16643375158309937, "learning_rate": 1.2748440446830118e-05, "loss": 0.0112, "step": 24157 }, { "epoch": 2.6190372940156115, "grad_norm": 0.07431981712579727, "learning_rate": 1.274481357899318e-05, "loss": 0.0019, "step": 24158 }, { "epoch": 2.6191457068516915, "grad_norm": 0.46951910853385925, "learning_rate": 1.2741186711156248e-05, "loss": 0.0125, "step": 24159 }, { "epoch": 2.619254119687771, "grad_norm": 0.3099778890609741, "learning_rate": 1.273755984331931e-05, "loss": 0.0144, "step": 24160 }, { "epoch": 2.619362532523851, "grad_norm": 0.27489951252937317, "learning_rate": 1.2733932975482374e-05, "loss": 0.0036, "step": 24161 }, { "epoch": 2.6194709453599305, "grad_norm": 0.13619285821914673, "learning_rate": 1.2730306107645437e-05, "loss": 0.0027, "step": 24162 }, { "epoch": 2.6195793581960105, "grad_norm": 0.21666009724140167, "learning_rate": 1.2726679239808503e-05, "loss": 0.0191, "step": 24163 }, { "epoch": 2.61968777103209, "grad_norm": 0.24713772535324097, "learning_rate": 1.2723052371971567e-05, "loss": 0.0099, "step": 24164 }, { "epoch": 2.61979618386817, "grad_norm": 0.0036130377557128668, "learning_rate": 1.271942550413463e-05, "loss": 0.0002, "step": 24165 }, { "epoch": 2.6199045967042496, "grad_norm": 0.7280780076980591, "learning_rate": 1.2715798636297693e-05, "loss": 0.0162, "step": 24166 }, { "epoch": 2.6200130095403296, "grad_norm": 0.3174953758716583, "learning_rate": 1.2712171768460757e-05, "loss": 0.0123, "step": 24167 }, { "epoch": 2.6201214223764095, "grad_norm": 0.0037333269137889147, "learning_rate": 1.2708544900623823e-05, "loss": 0.0002, "step": 24168 }, { "epoch": 2.620229835212489, "grad_norm": 0.49975529313087463, "learning_rate": 1.2704918032786885e-05, "loss": 0.0295, "step": 24169 }, { "epoch": 2.620338248048569, "grad_norm": 0.6154452562332153, "learning_rate": 1.270129116494995e-05, "loss": 0.0199, "step": 24170 }, { "epoch": 2.6204466608846486, "grad_norm": 0.1951661854982376, "learning_rate": 1.2697664297113012e-05, "loss": 0.0043, "step": 24171 }, { "epoch": 2.6205550737207286, "grad_norm": 0.07743636518716812, "learning_rate": 1.269403742927608e-05, "loss": 0.005, "step": 24172 }, { "epoch": 2.6206634865568086, "grad_norm": 0.22466549277305603, "learning_rate": 1.2690410561439142e-05, "loss": 0.0242, "step": 24173 }, { "epoch": 2.620771899392888, "grad_norm": 0.27474167943000793, "learning_rate": 1.2686783693602206e-05, "loss": 0.0119, "step": 24174 }, { "epoch": 2.6208803122289677, "grad_norm": 0.7714352011680603, "learning_rate": 1.2683156825765268e-05, "loss": 0.0291, "step": 24175 }, { "epoch": 2.6209887250650477, "grad_norm": 0.11436442285776138, "learning_rate": 1.2679529957928334e-05, "loss": 0.004, "step": 24176 }, { "epoch": 2.6210971379011276, "grad_norm": 0.2783268690109253, "learning_rate": 1.2675903090091398e-05, "loss": 0.0243, "step": 24177 }, { "epoch": 2.621205550737207, "grad_norm": 0.009309789165854454, "learning_rate": 1.2672276222254462e-05, "loss": 0.0004, "step": 24178 }, { "epoch": 2.621313963573287, "grad_norm": 0.6748118996620178, "learning_rate": 1.2668649354417525e-05, "loss": 0.0442, "step": 24179 }, { "epoch": 2.6214223764093667, "grad_norm": 0.11694177240133286, "learning_rate": 1.2665022486580589e-05, "loss": 0.0031, "step": 24180 }, { "epoch": 2.6215307892454467, "grad_norm": 0.12030746042728424, "learning_rate": 1.2661395618743655e-05, "loss": 0.0046, "step": 24181 }, { "epoch": 2.6216392020815267, "grad_norm": 0.4639107584953308, "learning_rate": 1.2657768750906717e-05, "loss": 0.0241, "step": 24182 }, { "epoch": 2.621747614917606, "grad_norm": 0.020556384697556496, "learning_rate": 1.2654141883069781e-05, "loss": 0.0007, "step": 24183 }, { "epoch": 2.621856027753686, "grad_norm": 0.2717675268650055, "learning_rate": 1.2650515015232844e-05, "loss": 0.0299, "step": 24184 }, { "epoch": 2.6219644405897657, "grad_norm": 0.1916658878326416, "learning_rate": 1.2646888147395911e-05, "loss": 0.0077, "step": 24185 }, { "epoch": 2.6220728534258457, "grad_norm": 0.2693249583244324, "learning_rate": 1.2643261279558974e-05, "loss": 0.0173, "step": 24186 }, { "epoch": 2.6221812662619253, "grad_norm": 0.06424935162067413, "learning_rate": 1.2639634411722038e-05, "loss": 0.0016, "step": 24187 }, { "epoch": 2.6222896790980053, "grad_norm": 0.5490413308143616, "learning_rate": 1.26360075438851e-05, "loss": 0.0327, "step": 24188 }, { "epoch": 2.622398091934085, "grad_norm": 0.11485438048839569, "learning_rate": 1.2632380676048166e-05, "loss": 0.0033, "step": 24189 }, { "epoch": 2.622506504770165, "grad_norm": 0.2269626408815384, "learning_rate": 1.262875380821123e-05, "loss": 0.0057, "step": 24190 }, { "epoch": 2.6226149176062448, "grad_norm": 0.3710063695907593, "learning_rate": 1.2625126940374294e-05, "loss": 0.0068, "step": 24191 }, { "epoch": 2.6227233304423243, "grad_norm": 0.028469450771808624, "learning_rate": 1.2621500072537356e-05, "loss": 0.001, "step": 24192 }, { "epoch": 2.6228317432784043, "grad_norm": 0.21923036873340607, "learning_rate": 1.261787320470042e-05, "loss": 0.0129, "step": 24193 }, { "epoch": 2.622940156114484, "grad_norm": 0.4883020520210266, "learning_rate": 1.2614246336863486e-05, "loss": 0.0218, "step": 24194 }, { "epoch": 2.623048568950564, "grad_norm": 0.4267228841781616, "learning_rate": 1.2610619469026549e-05, "loss": 0.0138, "step": 24195 }, { "epoch": 2.623156981786644, "grad_norm": 0.5201172232627869, "learning_rate": 1.2606992601189613e-05, "loss": 0.0047, "step": 24196 }, { "epoch": 2.6232653946227233, "grad_norm": 0.05474989861249924, "learning_rate": 1.2603365733352675e-05, "loss": 0.0008, "step": 24197 }, { "epoch": 2.623373807458803, "grad_norm": 0.6310835480690002, "learning_rate": 1.2599738865515743e-05, "loss": 0.029, "step": 24198 }, { "epoch": 2.623482220294883, "grad_norm": 0.027452781796455383, "learning_rate": 1.2596111997678805e-05, "loss": 0.0011, "step": 24199 }, { "epoch": 2.623590633130963, "grad_norm": 0.08770691603422165, "learning_rate": 1.259248512984187e-05, "loss": 0.0035, "step": 24200 }, { "epoch": 2.6236990459670424, "grad_norm": 0.1491255909204483, "learning_rate": 1.2588858262004932e-05, "loss": 0.0067, "step": 24201 }, { "epoch": 2.6238074588031224, "grad_norm": 0.3517906665802002, "learning_rate": 1.2585231394167998e-05, "loss": 0.0264, "step": 24202 }, { "epoch": 2.623915871639202, "grad_norm": 0.05991298332810402, "learning_rate": 1.2581604526331062e-05, "loss": 0.0042, "step": 24203 }, { "epoch": 2.624024284475282, "grad_norm": 0.21797052025794983, "learning_rate": 1.2577977658494126e-05, "loss": 0.0057, "step": 24204 }, { "epoch": 2.624132697311362, "grad_norm": 0.02518143318593502, "learning_rate": 1.2574350790657188e-05, "loss": 0.0011, "step": 24205 }, { "epoch": 2.6242411101474414, "grad_norm": 0.36976873874664307, "learning_rate": 1.2570723922820252e-05, "loss": 0.0131, "step": 24206 }, { "epoch": 2.624349522983521, "grad_norm": 0.49434468150138855, "learning_rate": 1.2567097054983318e-05, "loss": 0.0055, "step": 24207 }, { "epoch": 2.624457935819601, "grad_norm": 0.03163067251443863, "learning_rate": 1.256347018714638e-05, "loss": 0.0008, "step": 24208 }, { "epoch": 2.624566348655681, "grad_norm": 0.2852790355682373, "learning_rate": 1.2559843319309445e-05, "loss": 0.0139, "step": 24209 }, { "epoch": 2.6246747614917605, "grad_norm": 0.11814501881599426, "learning_rate": 1.2556216451472507e-05, "loss": 0.0021, "step": 24210 }, { "epoch": 2.6247831743278405, "grad_norm": 0.1230512335896492, "learning_rate": 1.2552589583635574e-05, "loss": 0.0045, "step": 24211 }, { "epoch": 2.62489158716392, "grad_norm": 0.15140806138515472, "learning_rate": 1.2548962715798637e-05, "loss": 0.007, "step": 24212 }, { "epoch": 2.625, "grad_norm": 0.07934936136007309, "learning_rate": 1.2545335847961701e-05, "loss": 0.0043, "step": 24213 }, { "epoch": 2.62510841283608, "grad_norm": 0.21526135504245758, "learning_rate": 1.2541708980124763e-05, "loss": 0.0026, "step": 24214 }, { "epoch": 2.6252168256721595, "grad_norm": 0.3922690153121948, "learning_rate": 1.253808211228783e-05, "loss": 0.0101, "step": 24215 }, { "epoch": 2.6253252385082395, "grad_norm": 0.5859732031822205, "learning_rate": 1.2534455244450893e-05, "loss": 0.0063, "step": 24216 }, { "epoch": 2.625433651344319, "grad_norm": 0.061567842960357666, "learning_rate": 1.2530828376613957e-05, "loss": 0.0026, "step": 24217 }, { "epoch": 2.625542064180399, "grad_norm": 0.2846832573413849, "learning_rate": 1.252720150877702e-05, "loss": 0.0143, "step": 24218 }, { "epoch": 2.625650477016479, "grad_norm": 0.16361014544963837, "learning_rate": 1.2523574640940084e-05, "loss": 0.009, "step": 24219 }, { "epoch": 2.6257588898525586, "grad_norm": 0.08993978053331375, "learning_rate": 1.251994777310315e-05, "loss": 0.0034, "step": 24220 }, { "epoch": 2.625867302688638, "grad_norm": 0.2095518410205841, "learning_rate": 1.2516320905266212e-05, "loss": 0.0069, "step": 24221 }, { "epoch": 2.625975715524718, "grad_norm": 0.5737338066101074, "learning_rate": 1.2512694037429276e-05, "loss": 0.0151, "step": 24222 }, { "epoch": 2.626084128360798, "grad_norm": 0.5427811741828918, "learning_rate": 1.2509067169592339e-05, "loss": 0.0081, "step": 24223 }, { "epoch": 2.6261925411968776, "grad_norm": 0.3051753342151642, "learning_rate": 1.2505440301755406e-05, "loss": 0.0194, "step": 24224 }, { "epoch": 2.6263009540329576, "grad_norm": 0.045906912535429, "learning_rate": 1.2501813433918469e-05, "loss": 0.002, "step": 24225 }, { "epoch": 2.626409366869037, "grad_norm": 0.0554354265332222, "learning_rate": 1.2498186566081533e-05, "loss": 0.002, "step": 24226 }, { "epoch": 2.626517779705117, "grad_norm": 0.47322651743888855, "learning_rate": 1.2494559698244597e-05, "loss": 0.009, "step": 24227 }, { "epoch": 2.626626192541197, "grad_norm": 0.6149033904075623, "learning_rate": 1.2490932830407661e-05, "loss": 0.0158, "step": 24228 }, { "epoch": 2.6267346053772767, "grad_norm": 0.530764639377594, "learning_rate": 1.2487305962570725e-05, "loss": 0.0044, "step": 24229 }, { "epoch": 2.626843018213356, "grad_norm": 0.14026181399822235, "learning_rate": 1.2483679094733789e-05, "loss": 0.0044, "step": 24230 }, { "epoch": 2.626951431049436, "grad_norm": 0.09818123281002045, "learning_rate": 1.2480052226896853e-05, "loss": 0.0042, "step": 24231 }, { "epoch": 2.627059843885516, "grad_norm": 0.11977532505989075, "learning_rate": 1.2476425359059917e-05, "loss": 0.0059, "step": 24232 }, { "epoch": 2.6271682567215957, "grad_norm": 0.07943400740623474, "learning_rate": 1.247279849122298e-05, "loss": 0.0032, "step": 24233 }, { "epoch": 2.6272766695576757, "grad_norm": 0.4919225573539734, "learning_rate": 1.2469171623386044e-05, "loss": 0.0261, "step": 24234 }, { "epoch": 2.6273850823937552, "grad_norm": 0.239597350358963, "learning_rate": 1.2465544755549108e-05, "loss": 0.0094, "step": 24235 }, { "epoch": 2.627493495229835, "grad_norm": 0.28328871726989746, "learning_rate": 1.2461917887712172e-05, "loss": 0.0098, "step": 24236 }, { "epoch": 2.627601908065915, "grad_norm": 0.28717026114463806, "learning_rate": 1.2458291019875236e-05, "loss": 0.0083, "step": 24237 }, { "epoch": 2.6277103209019947, "grad_norm": 0.6662259697914124, "learning_rate": 1.24546641520383e-05, "loss": 0.0074, "step": 24238 }, { "epoch": 2.6278187337380747, "grad_norm": 0.12078679352998734, "learning_rate": 1.2451037284201364e-05, "loss": 0.0033, "step": 24239 }, { "epoch": 2.6279271465741543, "grad_norm": 0.45335090160369873, "learning_rate": 1.2447410416364428e-05, "loss": 0.0173, "step": 24240 }, { "epoch": 2.6280355594102343, "grad_norm": 1.731752872467041, "learning_rate": 1.2443783548527493e-05, "loss": 0.0236, "step": 24241 }, { "epoch": 2.628143972246314, "grad_norm": 0.3318996727466583, "learning_rate": 1.2440156680690557e-05, "loss": 0.016, "step": 24242 }, { "epoch": 2.628252385082394, "grad_norm": 0.27788662910461426, "learning_rate": 1.243652981285362e-05, "loss": 0.0077, "step": 24243 }, { "epoch": 2.6283607979184733, "grad_norm": 0.19210122525691986, "learning_rate": 1.2432902945016685e-05, "loss": 0.0065, "step": 24244 }, { "epoch": 2.6284692107545533, "grad_norm": 0.2745826244354248, "learning_rate": 1.2429276077179749e-05, "loss": 0.0143, "step": 24245 }, { "epoch": 2.6285776235906333, "grad_norm": 0.052326273173093796, "learning_rate": 1.2425649209342811e-05, "loss": 0.0015, "step": 24246 }, { "epoch": 2.628686036426713, "grad_norm": 0.054941121488809586, "learning_rate": 1.2422022341505875e-05, "loss": 0.0037, "step": 24247 }, { "epoch": 2.628794449262793, "grad_norm": 0.05482811480760574, "learning_rate": 1.241839547366894e-05, "loss": 0.0022, "step": 24248 }, { "epoch": 2.6289028620988724, "grad_norm": 0.10437436401844025, "learning_rate": 1.2414768605832004e-05, "loss": 0.0012, "step": 24249 }, { "epoch": 2.6290112749349523, "grad_norm": 0.4258829653263092, "learning_rate": 1.2411141737995068e-05, "loss": 0.015, "step": 24250 }, { "epoch": 2.6291196877710323, "grad_norm": 0.2375132143497467, "learning_rate": 1.2407514870158132e-05, "loss": 0.0167, "step": 24251 }, { "epoch": 2.629228100607112, "grad_norm": 0.013311898335814476, "learning_rate": 1.2403888002321196e-05, "loss": 0.0012, "step": 24252 }, { "epoch": 2.6293365134431914, "grad_norm": 0.10816046595573425, "learning_rate": 1.240026113448426e-05, "loss": 0.0016, "step": 24253 }, { "epoch": 2.6294449262792714, "grad_norm": 0.33076658844947815, "learning_rate": 1.2396634266647324e-05, "loss": 0.0105, "step": 24254 }, { "epoch": 2.6295533391153514, "grad_norm": 0.13306427001953125, "learning_rate": 1.2393007398810388e-05, "loss": 0.0037, "step": 24255 }, { "epoch": 2.629661751951431, "grad_norm": 0.0864630937576294, "learning_rate": 1.2389380530973452e-05, "loss": 0.0063, "step": 24256 }, { "epoch": 2.629770164787511, "grad_norm": 0.059148047119379044, "learning_rate": 1.2385753663136517e-05, "loss": 0.0023, "step": 24257 }, { "epoch": 2.6298785776235905, "grad_norm": 0.912697434425354, "learning_rate": 1.238212679529958e-05, "loss": 0.0293, "step": 24258 }, { "epoch": 2.6299869904596704, "grad_norm": 0.17125852406024933, "learning_rate": 1.2378499927462645e-05, "loss": 0.0034, "step": 24259 }, { "epoch": 2.6300954032957504, "grad_norm": 0.02963990718126297, "learning_rate": 1.2374873059625707e-05, "loss": 0.0012, "step": 24260 }, { "epoch": 2.63020381613183, "grad_norm": 0.2773638665676117, "learning_rate": 1.2371246191788771e-05, "loss": 0.0135, "step": 24261 }, { "epoch": 2.63031222896791, "grad_norm": 0.30743423104286194, "learning_rate": 1.2367619323951835e-05, "loss": 0.0053, "step": 24262 }, { "epoch": 2.6304206418039895, "grad_norm": 0.38864341378211975, "learning_rate": 1.23639924561149e-05, "loss": 0.0293, "step": 24263 }, { "epoch": 2.6305290546400695, "grad_norm": 0.07901522517204285, "learning_rate": 1.2360365588277964e-05, "loss": 0.0024, "step": 24264 }, { "epoch": 2.630637467476149, "grad_norm": 0.007009692955762148, "learning_rate": 1.2356738720441028e-05, "loss": 0.0003, "step": 24265 }, { "epoch": 2.630745880312229, "grad_norm": 0.05052917078137398, "learning_rate": 1.2353111852604092e-05, "loss": 0.0012, "step": 24266 }, { "epoch": 2.6308542931483085, "grad_norm": 0.2138267606496811, "learning_rate": 1.2349484984767156e-05, "loss": 0.026, "step": 24267 }, { "epoch": 2.6309627059843885, "grad_norm": 0.23964515328407288, "learning_rate": 1.234585811693022e-05, "loss": 0.0166, "step": 24268 }, { "epoch": 2.6310711188204685, "grad_norm": 0.012693529017269611, "learning_rate": 1.2342231249093284e-05, "loss": 0.0006, "step": 24269 }, { "epoch": 2.631179531656548, "grad_norm": 0.4323958456516266, "learning_rate": 1.2338604381256348e-05, "loss": 0.0236, "step": 24270 }, { "epoch": 2.631287944492628, "grad_norm": 0.05827486515045166, "learning_rate": 1.2334977513419412e-05, "loss": 0.0011, "step": 24271 }, { "epoch": 2.6313963573287076, "grad_norm": 0.04705863818526268, "learning_rate": 1.2331350645582476e-05, "loss": 0.0014, "step": 24272 }, { "epoch": 2.6315047701647876, "grad_norm": 0.3646966814994812, "learning_rate": 1.2327723777745539e-05, "loss": 0.0067, "step": 24273 }, { "epoch": 2.6316131830008676, "grad_norm": 0.1829620897769928, "learning_rate": 1.2324096909908603e-05, "loss": 0.006, "step": 24274 }, { "epoch": 2.631721595836947, "grad_norm": 0.4484454393386841, "learning_rate": 1.2320470042071667e-05, "loss": 0.0222, "step": 24275 }, { "epoch": 2.6318300086730266, "grad_norm": 0.2928803861141205, "learning_rate": 1.2316843174234731e-05, "loss": 0.0069, "step": 24276 }, { "epoch": 2.6319384215091066, "grad_norm": 0.2802458703517914, "learning_rate": 1.2313216306397795e-05, "loss": 0.0087, "step": 24277 }, { "epoch": 2.6320468343451866, "grad_norm": 0.8756179213523865, "learning_rate": 1.230958943856086e-05, "loss": 0.0331, "step": 24278 }, { "epoch": 2.632155247181266, "grad_norm": 0.08956480771303177, "learning_rate": 1.2305962570723923e-05, "loss": 0.0031, "step": 24279 }, { "epoch": 2.632263660017346, "grad_norm": 0.02890467830002308, "learning_rate": 1.2302335702886988e-05, "loss": 0.0012, "step": 24280 }, { "epoch": 2.6323720728534257, "grad_norm": 0.1162102222442627, "learning_rate": 1.2298708835050052e-05, "loss": 0.0042, "step": 24281 }, { "epoch": 2.6324804856895057, "grad_norm": 0.3502788543701172, "learning_rate": 1.2295081967213116e-05, "loss": 0.008, "step": 24282 }, { "epoch": 2.6325888985255856, "grad_norm": 0.7161483764648438, "learning_rate": 1.229145509937618e-05, "loss": 0.0172, "step": 24283 }, { "epoch": 2.632697311361665, "grad_norm": 0.17124627530574799, "learning_rate": 1.2287828231539244e-05, "loss": 0.0058, "step": 24284 }, { "epoch": 2.632805724197745, "grad_norm": 0.13107971847057343, "learning_rate": 1.2284201363702308e-05, "loss": 0.0058, "step": 24285 }, { "epoch": 2.6329141370338247, "grad_norm": 0.021057510748505592, "learning_rate": 1.228057449586537e-05, "loss": 0.0006, "step": 24286 }, { "epoch": 2.6330225498699047, "grad_norm": 0.22734352946281433, "learning_rate": 1.2276947628028435e-05, "loss": 0.0042, "step": 24287 }, { "epoch": 2.6331309627059842, "grad_norm": 0.2924121022224426, "learning_rate": 1.2273320760191499e-05, "loss": 0.0073, "step": 24288 }, { "epoch": 2.633239375542064, "grad_norm": 0.26939263939857483, "learning_rate": 1.2269693892354563e-05, "loss": 0.0066, "step": 24289 }, { "epoch": 2.6333477883781438, "grad_norm": 0.6784183382987976, "learning_rate": 1.2266067024517627e-05, "loss": 0.0259, "step": 24290 }, { "epoch": 2.6334562012142237, "grad_norm": 0.22612540423870087, "learning_rate": 1.2262440156680691e-05, "loss": 0.0096, "step": 24291 }, { "epoch": 2.6335646140503037, "grad_norm": 0.352591872215271, "learning_rate": 1.2258813288843755e-05, "loss": 0.028, "step": 24292 }, { "epoch": 2.6336730268863833, "grad_norm": 0.2768988311290741, "learning_rate": 1.225518642100682e-05, "loss": 0.0063, "step": 24293 }, { "epoch": 2.6337814397224633, "grad_norm": 0.34808728098869324, "learning_rate": 1.2251559553169883e-05, "loss": 0.0195, "step": 24294 }, { "epoch": 2.633889852558543, "grad_norm": 0.4261743128299713, "learning_rate": 1.2247932685332947e-05, "loss": 0.0672, "step": 24295 }, { "epoch": 2.633998265394623, "grad_norm": 0.17475375533103943, "learning_rate": 1.2244305817496012e-05, "loss": 0.0028, "step": 24296 }, { "epoch": 2.6341066782307028, "grad_norm": 0.5355966687202454, "learning_rate": 1.2240678949659076e-05, "loss": 0.0413, "step": 24297 }, { "epoch": 2.6342150910667823, "grad_norm": 0.5469224452972412, "learning_rate": 1.223705208182214e-05, "loss": 0.0104, "step": 24298 }, { "epoch": 2.634323503902862, "grad_norm": 0.4956668019294739, "learning_rate": 1.2233425213985202e-05, "loss": 0.0399, "step": 24299 }, { "epoch": 2.634431916738942, "grad_norm": 0.2785385847091675, "learning_rate": 1.2229798346148266e-05, "loss": 0.0141, "step": 24300 }, { "epoch": 2.634540329575022, "grad_norm": 0.4670298993587494, "learning_rate": 1.222617147831133e-05, "loss": 0.0196, "step": 24301 }, { "epoch": 2.6346487424111014, "grad_norm": 0.002360390964895487, "learning_rate": 1.2222544610474394e-05, "loss": 0.0001, "step": 24302 }, { "epoch": 2.6347571552471813, "grad_norm": 0.1823752522468567, "learning_rate": 1.2218917742637459e-05, "loss": 0.0055, "step": 24303 }, { "epoch": 2.634865568083261, "grad_norm": 0.2187662422657013, "learning_rate": 1.2215290874800523e-05, "loss": 0.0103, "step": 24304 }, { "epoch": 2.634973980919341, "grad_norm": 0.747526228427887, "learning_rate": 1.2211664006963587e-05, "loss": 0.0187, "step": 24305 }, { "epoch": 2.635082393755421, "grad_norm": 0.2645169198513031, "learning_rate": 1.2208037139126651e-05, "loss": 0.0279, "step": 24306 }, { "epoch": 2.6351908065915004, "grad_norm": 0.1405518501996994, "learning_rate": 1.2204410271289715e-05, "loss": 0.005, "step": 24307 }, { "epoch": 2.6352992194275804, "grad_norm": 0.07081759721040726, "learning_rate": 1.2200783403452779e-05, "loss": 0.0053, "step": 24308 }, { "epoch": 2.63540763226366, "grad_norm": 0.5222303867340088, "learning_rate": 1.2197156535615843e-05, "loss": 0.024, "step": 24309 }, { "epoch": 2.63551604509974, "grad_norm": 0.23757341504096985, "learning_rate": 1.2193529667778907e-05, "loss": 0.0053, "step": 24310 }, { "epoch": 2.6356244579358195, "grad_norm": 0.09501243382692337, "learning_rate": 1.2189902799941971e-05, "loss": 0.0038, "step": 24311 }, { "epoch": 2.6357328707718994, "grad_norm": 0.26856184005737305, "learning_rate": 1.2186275932105034e-05, "loss": 0.0115, "step": 24312 }, { "epoch": 2.635841283607979, "grad_norm": 0.015525045804679394, "learning_rate": 1.21826490642681e-05, "loss": 0.001, "step": 24313 }, { "epoch": 2.635949696444059, "grad_norm": 0.23984849452972412, "learning_rate": 1.2179022196431162e-05, "loss": 0.0125, "step": 24314 }, { "epoch": 2.636058109280139, "grad_norm": 0.026718253269791603, "learning_rate": 1.2175395328594226e-05, "loss": 0.001, "step": 24315 }, { "epoch": 2.6361665221162185, "grad_norm": 0.10082849115133286, "learning_rate": 1.217176846075729e-05, "loss": 0.0023, "step": 24316 }, { "epoch": 2.6362749349522985, "grad_norm": 0.040494728833436966, "learning_rate": 1.2168141592920354e-05, "loss": 0.0009, "step": 24317 }, { "epoch": 2.636383347788378, "grad_norm": 0.7850908041000366, "learning_rate": 1.2164514725083418e-05, "loss": 0.0546, "step": 24318 }, { "epoch": 2.636491760624458, "grad_norm": 0.20899957418441772, "learning_rate": 1.2160887857246483e-05, "loss": 0.0068, "step": 24319 }, { "epoch": 2.636600173460538, "grad_norm": 0.573191225528717, "learning_rate": 1.2157260989409547e-05, "loss": 0.0115, "step": 24320 }, { "epoch": 2.6367085862966175, "grad_norm": 0.2686340808868408, "learning_rate": 1.215363412157261e-05, "loss": 0.0058, "step": 24321 }, { "epoch": 2.636816999132697, "grad_norm": 0.10316578298807144, "learning_rate": 1.2150007253735675e-05, "loss": 0.0021, "step": 24322 }, { "epoch": 2.636925411968777, "grad_norm": 0.2966938614845276, "learning_rate": 1.2146380385898739e-05, "loss": 0.0079, "step": 24323 }, { "epoch": 2.637033824804857, "grad_norm": 0.06670884788036346, "learning_rate": 1.2142753518061803e-05, "loss": 0.0023, "step": 24324 }, { "epoch": 2.6371422376409366, "grad_norm": 0.16287894546985626, "learning_rate": 1.2139126650224866e-05, "loss": 0.0054, "step": 24325 }, { "epoch": 2.6372506504770166, "grad_norm": 0.2691844701766968, "learning_rate": 1.2135499782387931e-05, "loss": 0.011, "step": 24326 }, { "epoch": 2.637359063313096, "grad_norm": 0.20793019235134125, "learning_rate": 1.2131872914550994e-05, "loss": 0.0075, "step": 24327 }, { "epoch": 2.637467476149176, "grad_norm": 0.4170450270175934, "learning_rate": 1.2128246046714058e-05, "loss": 0.0079, "step": 24328 }, { "epoch": 2.637575888985256, "grad_norm": 0.28797927498817444, "learning_rate": 1.2124619178877122e-05, "loss": 0.0085, "step": 24329 }, { "epoch": 2.6376843018213356, "grad_norm": 0.3026721775531769, "learning_rate": 1.2120992311040186e-05, "loss": 0.0203, "step": 24330 }, { "epoch": 2.637792714657415, "grad_norm": 0.4943004846572876, "learning_rate": 1.211736544320325e-05, "loss": 0.019, "step": 24331 }, { "epoch": 2.637901127493495, "grad_norm": 0.10154515504837036, "learning_rate": 1.2113738575366314e-05, "loss": 0.0014, "step": 24332 }, { "epoch": 2.638009540329575, "grad_norm": 0.48039501905441284, "learning_rate": 1.2110111707529378e-05, "loss": 0.0127, "step": 24333 }, { "epoch": 2.6381179531656547, "grad_norm": 0.3951919674873352, "learning_rate": 1.2106484839692442e-05, "loss": 0.0164, "step": 24334 }, { "epoch": 2.6382263660017347, "grad_norm": 0.15808160603046417, "learning_rate": 1.2102857971855507e-05, "loss": 0.0019, "step": 24335 }, { "epoch": 2.638334778837814, "grad_norm": 0.07148158550262451, "learning_rate": 1.209923110401857e-05, "loss": 0.0018, "step": 24336 }, { "epoch": 2.638443191673894, "grad_norm": 0.3593054711818695, "learning_rate": 1.2095604236181635e-05, "loss": 0.0095, "step": 24337 }, { "epoch": 2.638551604509974, "grad_norm": 0.05055947229266167, "learning_rate": 1.2091977368344697e-05, "loss": 0.0015, "step": 24338 }, { "epoch": 2.6386600173460537, "grad_norm": 0.4924376904964447, "learning_rate": 1.2088350500507763e-05, "loss": 0.0072, "step": 24339 }, { "epoch": 2.6387684301821337, "grad_norm": 0.18909414112567902, "learning_rate": 1.2084723632670825e-05, "loss": 0.0235, "step": 24340 }, { "epoch": 2.6388768430182132, "grad_norm": 0.7044569253921509, "learning_rate": 1.208109676483389e-05, "loss": 0.0284, "step": 24341 }, { "epoch": 2.638985255854293, "grad_norm": 0.24783141911029816, "learning_rate": 1.2077469896996954e-05, "loss": 0.0076, "step": 24342 }, { "epoch": 2.639093668690373, "grad_norm": 0.4678783118724823, "learning_rate": 1.2073843029160018e-05, "loss": 0.044, "step": 24343 }, { "epoch": 2.6392020815264527, "grad_norm": 0.009106104262173176, "learning_rate": 1.2070216161323082e-05, "loss": 0.0005, "step": 24344 }, { "epoch": 2.6393104943625323, "grad_norm": 0.18765248358249664, "learning_rate": 1.2066589293486146e-05, "loss": 0.0021, "step": 24345 }, { "epoch": 2.6394189071986123, "grad_norm": 0.19915245473384857, "learning_rate": 1.206296242564921e-05, "loss": 0.0099, "step": 24346 }, { "epoch": 2.6395273200346923, "grad_norm": 0.3279624283313751, "learning_rate": 1.2059335557812274e-05, "loss": 0.0094, "step": 24347 }, { "epoch": 2.639635732870772, "grad_norm": 0.017605146393179893, "learning_rate": 1.2055708689975338e-05, "loss": 0.0006, "step": 24348 }, { "epoch": 2.639744145706852, "grad_norm": 0.08239897340536118, "learning_rate": 1.2052081822138402e-05, "loss": 0.0033, "step": 24349 }, { "epoch": 2.6398525585429313, "grad_norm": 0.21774809062480927, "learning_rate": 1.2048454954301466e-05, "loss": 0.0055, "step": 24350 }, { "epoch": 2.6399609713790113, "grad_norm": 0.3807052969932556, "learning_rate": 1.2044828086464529e-05, "loss": 0.0231, "step": 24351 }, { "epoch": 2.6400693842150913, "grad_norm": 0.0809030756354332, "learning_rate": 1.2041201218627595e-05, "loss": 0.0026, "step": 24352 }, { "epoch": 2.640177797051171, "grad_norm": 0.07046784460544586, "learning_rate": 1.2037574350790657e-05, "loss": 0.0034, "step": 24353 }, { "epoch": 2.6402862098872504, "grad_norm": 0.3669710159301758, "learning_rate": 1.2033947482953723e-05, "loss": 0.0049, "step": 24354 }, { "epoch": 2.6403946227233304, "grad_norm": 0.12420357018709183, "learning_rate": 1.2030320615116785e-05, "loss": 0.002, "step": 24355 }, { "epoch": 2.6405030355594103, "grad_norm": 0.6265481114387512, "learning_rate": 1.202669374727985e-05, "loss": 0.0138, "step": 24356 }, { "epoch": 2.64061144839549, "grad_norm": 0.03663777932524681, "learning_rate": 1.2023066879442913e-05, "loss": 0.0005, "step": 24357 }, { "epoch": 2.64071986123157, "grad_norm": 0.09291127324104309, "learning_rate": 1.2019440011605978e-05, "loss": 0.0041, "step": 24358 }, { "epoch": 2.6408282740676494, "grad_norm": 0.4147184193134308, "learning_rate": 1.2015813143769042e-05, "loss": 0.0203, "step": 24359 }, { "epoch": 2.6409366869037294, "grad_norm": 0.37264585494995117, "learning_rate": 1.2012186275932106e-05, "loss": 0.0096, "step": 24360 }, { "epoch": 2.6410450997398094, "grad_norm": 0.28817927837371826, "learning_rate": 1.200855940809517e-05, "loss": 0.0156, "step": 24361 }, { "epoch": 2.641153512575889, "grad_norm": 0.44403138756752014, "learning_rate": 1.2004932540258234e-05, "loss": 0.0095, "step": 24362 }, { "epoch": 2.641261925411969, "grad_norm": 0.022121626883745193, "learning_rate": 1.2001305672421298e-05, "loss": 0.0006, "step": 24363 }, { "epoch": 2.6413703382480485, "grad_norm": 0.12840773165225983, "learning_rate": 1.199767880458436e-05, "loss": 0.0022, "step": 24364 }, { "epoch": 2.6414787510841284, "grad_norm": 0.4555426239967346, "learning_rate": 1.1994051936747426e-05, "loss": 0.0194, "step": 24365 }, { "epoch": 2.6415871639202084, "grad_norm": 0.3415161967277527, "learning_rate": 1.1990425068910489e-05, "loss": 0.0054, "step": 24366 }, { "epoch": 2.641695576756288, "grad_norm": 0.06823313236236572, "learning_rate": 1.1986798201073555e-05, "loss": 0.0016, "step": 24367 }, { "epoch": 2.6418039895923675, "grad_norm": 0.292145311832428, "learning_rate": 1.1983171333236617e-05, "loss": 0.0069, "step": 24368 }, { "epoch": 2.6419124024284475, "grad_norm": 0.026540368795394897, "learning_rate": 1.1979544465399681e-05, "loss": 0.0009, "step": 24369 }, { "epoch": 2.6420208152645275, "grad_norm": 0.46488529443740845, "learning_rate": 1.1975917597562745e-05, "loss": 0.0028, "step": 24370 }, { "epoch": 2.642129228100607, "grad_norm": 0.10544238984584808, "learning_rate": 1.197229072972581e-05, "loss": 0.0027, "step": 24371 }, { "epoch": 2.642237640936687, "grad_norm": 0.45572835206985474, "learning_rate": 1.1968663861888873e-05, "loss": 0.0142, "step": 24372 }, { "epoch": 2.6423460537727665, "grad_norm": 0.056439489126205444, "learning_rate": 1.1965036994051937e-05, "loss": 0.0026, "step": 24373 }, { "epoch": 2.6424544666088465, "grad_norm": 0.17390137910842896, "learning_rate": 1.1961410126215002e-05, "loss": 0.0037, "step": 24374 }, { "epoch": 2.6425628794449265, "grad_norm": 0.09315530955791473, "learning_rate": 1.1957783258378066e-05, "loss": 0.0058, "step": 24375 }, { "epoch": 2.642671292281006, "grad_norm": 0.2738155722618103, "learning_rate": 1.195415639054113e-05, "loss": 0.0157, "step": 24376 }, { "epoch": 2.6427797051170856, "grad_norm": 0.9426441788673401, "learning_rate": 1.1950529522704192e-05, "loss": 0.0568, "step": 24377 }, { "epoch": 2.6428881179531656, "grad_norm": 0.416206955909729, "learning_rate": 1.1946902654867258e-05, "loss": 0.0118, "step": 24378 }, { "epoch": 2.6429965307892456, "grad_norm": 0.16580627858638763, "learning_rate": 1.194327578703032e-05, "loss": 0.0096, "step": 24379 }, { "epoch": 2.643104943625325, "grad_norm": 0.3346889019012451, "learning_rate": 1.1939648919193386e-05, "loss": 0.0137, "step": 24380 }, { "epoch": 2.643213356461405, "grad_norm": 0.5001726746559143, "learning_rate": 1.1936022051356449e-05, "loss": 0.0414, "step": 24381 }, { "epoch": 2.6433217692974846, "grad_norm": 0.512121856212616, "learning_rate": 1.1932395183519513e-05, "loss": 0.0264, "step": 24382 }, { "epoch": 2.6434301821335646, "grad_norm": 0.008235208690166473, "learning_rate": 1.1928768315682577e-05, "loss": 0.0003, "step": 24383 }, { "epoch": 2.6435385949696446, "grad_norm": 0.354129433631897, "learning_rate": 1.1925141447845641e-05, "loss": 0.0169, "step": 24384 }, { "epoch": 2.643647007805724, "grad_norm": 0.34078818559646606, "learning_rate": 1.1921514580008705e-05, "loss": 0.0047, "step": 24385 }, { "epoch": 2.643755420641804, "grad_norm": 0.08879704773426056, "learning_rate": 1.1917887712171769e-05, "loss": 0.0032, "step": 24386 }, { "epoch": 2.6438638334778837, "grad_norm": 0.04034773260354996, "learning_rate": 1.1914260844334833e-05, "loss": 0.0014, "step": 24387 }, { "epoch": 2.6439722463139637, "grad_norm": 0.14584963023662567, "learning_rate": 1.1910633976497897e-05, "loss": 0.005, "step": 24388 }, { "epoch": 2.644080659150043, "grad_norm": 0.24248093366622925, "learning_rate": 1.1907007108660961e-05, "loss": 0.0095, "step": 24389 }, { "epoch": 2.644189071986123, "grad_norm": 0.15806815028190613, "learning_rate": 1.1903380240824024e-05, "loss": 0.007, "step": 24390 }, { "epoch": 2.6442974848222027, "grad_norm": 0.7002266645431519, "learning_rate": 1.189975337298709e-05, "loss": 0.0625, "step": 24391 }, { "epoch": 2.6444058976582827, "grad_norm": 0.05757802724838257, "learning_rate": 1.1896126505150152e-05, "loss": 0.002, "step": 24392 }, { "epoch": 2.6445143104943627, "grad_norm": 0.08684460073709488, "learning_rate": 1.1892499637313218e-05, "loss": 0.0036, "step": 24393 }, { "epoch": 2.6446227233304422, "grad_norm": 0.08435644954442978, "learning_rate": 1.188887276947628e-05, "loss": 0.0025, "step": 24394 }, { "epoch": 2.6447311361665222, "grad_norm": 0.036481935530900955, "learning_rate": 1.1885245901639344e-05, "loss": 0.0015, "step": 24395 }, { "epoch": 2.6448395490026018, "grad_norm": 0.02308863028883934, "learning_rate": 1.1881619033802408e-05, "loss": 0.0007, "step": 24396 }, { "epoch": 2.6449479618386817, "grad_norm": 0.012126950547099113, "learning_rate": 1.1877992165965473e-05, "loss": 0.0007, "step": 24397 }, { "epoch": 2.6450563746747617, "grad_norm": 0.02362884394824505, "learning_rate": 1.1874365298128537e-05, "loss": 0.0006, "step": 24398 }, { "epoch": 2.6451647875108413, "grad_norm": 0.008365122601389885, "learning_rate": 1.18707384302916e-05, "loss": 0.0002, "step": 24399 }, { "epoch": 2.645273200346921, "grad_norm": 0.23851963877677917, "learning_rate": 1.1867111562454665e-05, "loss": 0.0138, "step": 24400 }, { "epoch": 2.645381613183001, "grad_norm": 0.024041369557380676, "learning_rate": 1.1863484694617729e-05, "loss": 0.001, "step": 24401 }, { "epoch": 2.645490026019081, "grad_norm": 0.8901177048683167, "learning_rate": 1.1859857826780793e-05, "loss": 0.0144, "step": 24402 }, { "epoch": 2.6455984388551603, "grad_norm": 0.08799216896295547, "learning_rate": 1.1856230958943857e-05, "loss": 0.0022, "step": 24403 }, { "epoch": 2.6457068516912403, "grad_norm": 0.18189162015914917, "learning_rate": 1.1852604091106921e-05, "loss": 0.0027, "step": 24404 }, { "epoch": 2.64581526452732, "grad_norm": 0.18793262541294098, "learning_rate": 1.1848977223269984e-05, "loss": 0.0027, "step": 24405 }, { "epoch": 2.6459236773634, "grad_norm": 0.19498443603515625, "learning_rate": 1.184535035543305e-05, "loss": 0.0067, "step": 24406 }, { "epoch": 2.64603209019948, "grad_norm": 0.13026435673236847, "learning_rate": 1.1841723487596112e-05, "loss": 0.0043, "step": 24407 }, { "epoch": 2.6461405030355594, "grad_norm": 0.0033029792830348015, "learning_rate": 1.1838096619759178e-05, "loss": 0.0002, "step": 24408 }, { "epoch": 2.6462489158716394, "grad_norm": 0.248934805393219, "learning_rate": 1.183446975192224e-05, "loss": 0.0197, "step": 24409 }, { "epoch": 2.646357328707719, "grad_norm": 0.326131671667099, "learning_rate": 1.1830842884085304e-05, "loss": 0.0111, "step": 24410 }, { "epoch": 2.646465741543799, "grad_norm": 0.08173923194408417, "learning_rate": 1.1827216016248368e-05, "loss": 0.0043, "step": 24411 }, { "epoch": 2.6465741543798784, "grad_norm": 0.4116476774215698, "learning_rate": 1.1823589148411432e-05, "loss": 0.0319, "step": 24412 }, { "epoch": 2.6466825672159584, "grad_norm": 0.34347909688949585, "learning_rate": 1.1819962280574497e-05, "loss": 0.027, "step": 24413 }, { "epoch": 2.646790980052038, "grad_norm": 0.21019718050956726, "learning_rate": 1.181633541273756e-05, "loss": 0.0066, "step": 24414 }, { "epoch": 2.646899392888118, "grad_norm": 0.06785981357097626, "learning_rate": 1.1812708544900625e-05, "loss": 0.0023, "step": 24415 }, { "epoch": 2.647007805724198, "grad_norm": 0.09950189292430878, "learning_rate": 1.1809081677063689e-05, "loss": 0.0019, "step": 24416 }, { "epoch": 2.6471162185602775, "grad_norm": 0.024480313062667847, "learning_rate": 1.1805454809226753e-05, "loss": 0.0008, "step": 24417 }, { "epoch": 2.6472246313963574, "grad_norm": 0.04825459420681, "learning_rate": 1.1801827941389815e-05, "loss": 0.0015, "step": 24418 }, { "epoch": 2.647333044232437, "grad_norm": 0.6584050059318542, "learning_rate": 1.1798201073552881e-05, "loss": 0.0226, "step": 24419 }, { "epoch": 2.647441457068517, "grad_norm": 0.002992876572534442, "learning_rate": 1.1794574205715944e-05, "loss": 0.0001, "step": 24420 }, { "epoch": 2.647549869904597, "grad_norm": 0.10632070153951645, "learning_rate": 1.179094733787901e-05, "loss": 0.0022, "step": 24421 }, { "epoch": 2.6476582827406765, "grad_norm": 0.5856199860572815, "learning_rate": 1.1787320470042072e-05, "loss": 0.0266, "step": 24422 }, { "epoch": 2.647766695576756, "grad_norm": 0.926779568195343, "learning_rate": 1.1783693602205136e-05, "loss": 0.0396, "step": 24423 }, { "epoch": 2.647875108412836, "grad_norm": 0.588840126991272, "learning_rate": 1.17800667343682e-05, "loss": 0.0136, "step": 24424 }, { "epoch": 2.647983521248916, "grad_norm": 0.18823933601379395, "learning_rate": 1.1776439866531264e-05, "loss": 0.0025, "step": 24425 }, { "epoch": 2.6480919340849955, "grad_norm": 0.13736139237880707, "learning_rate": 1.1772812998694328e-05, "loss": 0.0029, "step": 24426 }, { "epoch": 2.6482003469210755, "grad_norm": 0.12165020406246185, "learning_rate": 1.1769186130857392e-05, "loss": 0.0097, "step": 24427 }, { "epoch": 2.648308759757155, "grad_norm": 0.04982815310359001, "learning_rate": 1.1765559263020456e-05, "loss": 0.0015, "step": 24428 }, { "epoch": 2.648417172593235, "grad_norm": 0.5295215249061584, "learning_rate": 1.176193239518352e-05, "loss": 0.0378, "step": 24429 }, { "epoch": 2.648525585429315, "grad_norm": 0.16462114453315735, "learning_rate": 1.1758305527346585e-05, "loss": 0.0041, "step": 24430 }, { "epoch": 2.6486339982653946, "grad_norm": 0.025482596829533577, "learning_rate": 1.1754678659509647e-05, "loss": 0.0009, "step": 24431 }, { "epoch": 2.6487424111014746, "grad_norm": 2.1419177055358887, "learning_rate": 1.1751051791672713e-05, "loss": 0.0192, "step": 24432 }, { "epoch": 2.648850823937554, "grad_norm": 0.013129083439707756, "learning_rate": 1.1747424923835775e-05, "loss": 0.0003, "step": 24433 }, { "epoch": 2.648959236773634, "grad_norm": 0.29928210377693176, "learning_rate": 1.1743798055998841e-05, "loss": 0.0156, "step": 24434 }, { "epoch": 2.6490676496097136, "grad_norm": 0.6304391026496887, "learning_rate": 1.1740171188161904e-05, "loss": 0.0203, "step": 24435 }, { "epoch": 2.6491760624457936, "grad_norm": 0.03409792110323906, "learning_rate": 1.1736544320324968e-05, "loss": 0.0019, "step": 24436 }, { "epoch": 2.649284475281873, "grad_norm": 0.027217242866754532, "learning_rate": 1.1732917452488032e-05, "loss": 0.0008, "step": 24437 }, { "epoch": 2.649392888117953, "grad_norm": 0.1722770780324936, "learning_rate": 1.1729290584651096e-05, "loss": 0.0041, "step": 24438 }, { "epoch": 2.649501300954033, "grad_norm": 0.24204500019550323, "learning_rate": 1.172566371681416e-05, "loss": 0.0021, "step": 24439 }, { "epoch": 2.6496097137901127, "grad_norm": 0.035665519535541534, "learning_rate": 1.1722036848977224e-05, "loss": 0.0005, "step": 24440 }, { "epoch": 2.6497181266261927, "grad_norm": 0.16373036801815033, "learning_rate": 1.1718409981140288e-05, "loss": 0.0044, "step": 24441 }, { "epoch": 2.649826539462272, "grad_norm": 0.016161715611815453, "learning_rate": 1.1714783113303352e-05, "loss": 0.0009, "step": 24442 }, { "epoch": 2.649934952298352, "grad_norm": 0.08845756202936172, "learning_rate": 1.1711156245466416e-05, "loss": 0.0043, "step": 24443 }, { "epoch": 2.650043365134432, "grad_norm": 0.17040270566940308, "learning_rate": 1.1707529377629479e-05, "loss": 0.0069, "step": 24444 }, { "epoch": 2.6501517779705117, "grad_norm": 0.7726210355758667, "learning_rate": 1.1703902509792545e-05, "loss": 0.0467, "step": 24445 }, { "epoch": 2.6502601908065913, "grad_norm": 0.4560994803905487, "learning_rate": 1.1700275641955607e-05, "loss": 0.0072, "step": 24446 }, { "epoch": 2.6503686036426712, "grad_norm": 0.1748550981283188, "learning_rate": 1.1696648774118673e-05, "loss": 0.0089, "step": 24447 }, { "epoch": 2.6504770164787512, "grad_norm": 0.3947104513645172, "learning_rate": 1.1693021906281735e-05, "loss": 0.0245, "step": 24448 }, { "epoch": 2.6505854293148308, "grad_norm": 0.13739420473575592, "learning_rate": 1.16893950384448e-05, "loss": 0.0015, "step": 24449 }, { "epoch": 2.6506938421509108, "grad_norm": 0.7290895581245422, "learning_rate": 1.1685768170607863e-05, "loss": 0.0104, "step": 24450 }, { "epoch": 2.6508022549869903, "grad_norm": 0.3030123710632324, "learning_rate": 1.1682141302770927e-05, "loss": 0.0247, "step": 24451 }, { "epoch": 2.6509106678230703, "grad_norm": 0.23834700882434845, "learning_rate": 1.1678514434933992e-05, "loss": 0.0116, "step": 24452 }, { "epoch": 2.6510190806591503, "grad_norm": 0.09141667932271957, "learning_rate": 1.1674887567097056e-05, "loss": 0.0042, "step": 24453 }, { "epoch": 2.65112749349523, "grad_norm": 0.47013744711875916, "learning_rate": 1.167126069926012e-05, "loss": 0.0187, "step": 24454 }, { "epoch": 2.65123590633131, "grad_norm": 0.11208392679691315, "learning_rate": 1.1667633831423184e-05, "loss": 0.0019, "step": 24455 }, { "epoch": 2.6513443191673893, "grad_norm": 0.4756484031677246, "learning_rate": 1.1664006963586248e-05, "loss": 0.028, "step": 24456 }, { "epoch": 2.6514527320034693, "grad_norm": 0.17853105068206787, "learning_rate": 1.166038009574931e-05, "loss": 0.0049, "step": 24457 }, { "epoch": 2.651561144839549, "grad_norm": 0.36328092217445374, "learning_rate": 1.1656753227912376e-05, "loss": 0.0045, "step": 24458 }, { "epoch": 2.651669557675629, "grad_norm": 0.9411813020706177, "learning_rate": 1.1653126360075439e-05, "loss": 0.0205, "step": 24459 }, { "epoch": 2.6517779705117084, "grad_norm": 0.8983807563781738, "learning_rate": 1.1649499492238504e-05, "loss": 0.0104, "step": 24460 }, { "epoch": 2.6518863833477884, "grad_norm": 0.2748269736766815, "learning_rate": 1.1645872624401567e-05, "loss": 0.0147, "step": 24461 }, { "epoch": 2.6519947961838684, "grad_norm": 0.40418457984924316, "learning_rate": 1.1642245756564633e-05, "loss": 0.0173, "step": 24462 }, { "epoch": 2.652103209019948, "grad_norm": 0.3481309413909912, "learning_rate": 1.1638618888727695e-05, "loss": 0.0068, "step": 24463 }, { "epoch": 2.652211621856028, "grad_norm": 0.08045873790979385, "learning_rate": 1.163499202089076e-05, "loss": 0.003, "step": 24464 }, { "epoch": 2.6523200346921074, "grad_norm": 0.030711157247424126, "learning_rate": 1.1631365153053823e-05, "loss": 0.0011, "step": 24465 }, { "epoch": 2.6524284475281874, "grad_norm": 0.05420391261577606, "learning_rate": 1.1627738285216887e-05, "loss": 0.0019, "step": 24466 }, { "epoch": 2.6525368603642674, "grad_norm": 0.14158381521701813, "learning_rate": 1.1624111417379951e-05, "loss": 0.0089, "step": 24467 }, { "epoch": 2.652645273200347, "grad_norm": 0.3953673243522644, "learning_rate": 1.1620484549543016e-05, "loss": 0.0158, "step": 24468 }, { "epoch": 2.6527536860364265, "grad_norm": 0.0560775026679039, "learning_rate": 1.161685768170608e-05, "loss": 0.0011, "step": 24469 }, { "epoch": 2.6528620988725065, "grad_norm": 0.5411193370819092, "learning_rate": 1.1613230813869142e-05, "loss": 0.0413, "step": 24470 }, { "epoch": 2.6529705117085864, "grad_norm": 0.005506465211510658, "learning_rate": 1.1609603946032208e-05, "loss": 0.0002, "step": 24471 }, { "epoch": 2.653078924544666, "grad_norm": 0.007339069619774818, "learning_rate": 1.160597707819527e-05, "loss": 0.0002, "step": 24472 }, { "epoch": 2.653187337380746, "grad_norm": 0.26976314187049866, "learning_rate": 1.1602350210358336e-05, "loss": 0.0114, "step": 24473 }, { "epoch": 2.6532957502168255, "grad_norm": 1.167872667312622, "learning_rate": 1.1598723342521399e-05, "loss": 0.0089, "step": 24474 }, { "epoch": 2.6534041630529055, "grad_norm": 0.0016307518817484379, "learning_rate": 1.1595096474684464e-05, "loss": 0.0001, "step": 24475 }, { "epoch": 2.6535125758889855, "grad_norm": 0.1477421075105667, "learning_rate": 1.1591469606847527e-05, "loss": 0.0068, "step": 24476 }, { "epoch": 2.653620988725065, "grad_norm": 0.17506301403045654, "learning_rate": 1.158784273901059e-05, "loss": 0.0065, "step": 24477 }, { "epoch": 2.6537294015611446, "grad_norm": 0.07063562422990799, "learning_rate": 1.1584215871173655e-05, "loss": 0.0021, "step": 24478 }, { "epoch": 2.6538378143972245, "grad_norm": 0.23243317008018494, "learning_rate": 1.1580589003336719e-05, "loss": 0.0111, "step": 24479 }, { "epoch": 2.6539462272333045, "grad_norm": 0.20646607875823975, "learning_rate": 1.1576962135499783e-05, "loss": 0.0058, "step": 24480 }, { "epoch": 2.654054640069384, "grad_norm": 0.9950678944587708, "learning_rate": 1.1573335267662847e-05, "loss": 0.0234, "step": 24481 }, { "epoch": 2.654163052905464, "grad_norm": 0.11994901299476624, "learning_rate": 1.1569708399825911e-05, "loss": 0.0048, "step": 24482 }, { "epoch": 2.6542714657415436, "grad_norm": 0.0013891651760786772, "learning_rate": 1.1566081531988974e-05, "loss": 0.0001, "step": 24483 }, { "epoch": 2.6543798785776236, "grad_norm": 0.3551316261291504, "learning_rate": 1.156245466415204e-05, "loss": 0.0094, "step": 24484 }, { "epoch": 2.6544882914137036, "grad_norm": 0.2906198501586914, "learning_rate": 1.1558827796315102e-05, "loss": 0.0091, "step": 24485 }, { "epoch": 2.654596704249783, "grad_norm": 0.3172932267189026, "learning_rate": 1.1555200928478168e-05, "loss": 0.023, "step": 24486 }, { "epoch": 2.654705117085863, "grad_norm": 1.2541509866714478, "learning_rate": 1.155157406064123e-05, "loss": 0.0189, "step": 24487 }, { "epoch": 2.6548135299219426, "grad_norm": 0.6394608616828918, "learning_rate": 1.1547947192804296e-05, "loss": 0.0271, "step": 24488 }, { "epoch": 2.6549219427580226, "grad_norm": 0.027856333181262016, "learning_rate": 1.1544320324967358e-05, "loss": 0.0008, "step": 24489 }, { "epoch": 2.6550303555941026, "grad_norm": 0.06155106797814369, "learning_rate": 1.1540693457130423e-05, "loss": 0.0028, "step": 24490 }, { "epoch": 2.655138768430182, "grad_norm": 0.2508755922317505, "learning_rate": 1.1537066589293487e-05, "loss": 0.003, "step": 24491 }, { "epoch": 2.6552471812662617, "grad_norm": 0.25024014711380005, "learning_rate": 1.153343972145655e-05, "loss": 0.0045, "step": 24492 }, { "epoch": 2.6553555941023417, "grad_norm": 0.00502878800034523, "learning_rate": 1.1529812853619615e-05, "loss": 0.0002, "step": 24493 }, { "epoch": 2.6554640069384217, "grad_norm": 0.10610396414995193, "learning_rate": 1.1526185985782679e-05, "loss": 0.0045, "step": 24494 }, { "epoch": 2.655572419774501, "grad_norm": 0.2630273103713989, "learning_rate": 1.1522559117945743e-05, "loss": 0.0086, "step": 24495 }, { "epoch": 2.655680832610581, "grad_norm": 0.091191865503788, "learning_rate": 1.1518932250108805e-05, "loss": 0.0029, "step": 24496 }, { "epoch": 2.6557892454466607, "grad_norm": 0.33305928111076355, "learning_rate": 1.1515305382271871e-05, "loss": 0.0099, "step": 24497 }, { "epoch": 2.6558976582827407, "grad_norm": 0.015621080994606018, "learning_rate": 1.1511678514434934e-05, "loss": 0.0002, "step": 24498 }, { "epoch": 2.6560060711188207, "grad_norm": 0.03345659002661705, "learning_rate": 1.1508051646598e-05, "loss": 0.001, "step": 24499 }, { "epoch": 2.6561144839549002, "grad_norm": 0.21162107586860657, "learning_rate": 1.1504424778761062e-05, "loss": 0.0218, "step": 24500 }, { "epoch": 2.65622289679098, "grad_norm": 0.7090880870819092, "learning_rate": 1.1500797910924128e-05, "loss": 0.0148, "step": 24501 }, { "epoch": 2.6563313096270598, "grad_norm": 0.23958756029605865, "learning_rate": 1.149717104308719e-05, "loss": 0.0115, "step": 24502 }, { "epoch": 2.6564397224631398, "grad_norm": 0.10172172635793686, "learning_rate": 1.1493544175250256e-05, "loss": 0.004, "step": 24503 }, { "epoch": 2.6565481352992193, "grad_norm": 0.11978550255298615, "learning_rate": 1.1489917307413318e-05, "loss": 0.0029, "step": 24504 }, { "epoch": 2.6566565481352993, "grad_norm": 0.11900217086076736, "learning_rate": 1.1486290439576382e-05, "loss": 0.0042, "step": 24505 }, { "epoch": 2.656764960971379, "grad_norm": 0.0978192538022995, "learning_rate": 1.1482663571739446e-05, "loss": 0.0017, "step": 24506 }, { "epoch": 2.656873373807459, "grad_norm": 0.42005249857902527, "learning_rate": 1.147903670390251e-05, "loss": 0.025, "step": 24507 }, { "epoch": 2.656981786643539, "grad_norm": 0.18306046724319458, "learning_rate": 1.1475409836065575e-05, "loss": 0.0061, "step": 24508 }, { "epoch": 2.6570901994796183, "grad_norm": 0.12329166382551193, "learning_rate": 1.1471782968228637e-05, "loss": 0.0026, "step": 24509 }, { "epoch": 2.6571986123156983, "grad_norm": 0.04065527766942978, "learning_rate": 1.1468156100391703e-05, "loss": 0.0009, "step": 24510 }, { "epoch": 2.657307025151778, "grad_norm": 0.3896717131137848, "learning_rate": 1.1464529232554765e-05, "loss": 0.0215, "step": 24511 }, { "epoch": 2.657415437987858, "grad_norm": 0.19650942087173462, "learning_rate": 1.1460902364717831e-05, "loss": 0.0054, "step": 24512 }, { "epoch": 2.657523850823938, "grad_norm": 0.3454135060310364, "learning_rate": 1.1457275496880894e-05, "loss": 0.0192, "step": 24513 }, { "epoch": 2.6576322636600174, "grad_norm": 0.20949827134609222, "learning_rate": 1.145364862904396e-05, "loss": 0.0046, "step": 24514 }, { "epoch": 2.657740676496097, "grad_norm": 0.4883800148963928, "learning_rate": 1.1450021761207022e-05, "loss": 0.016, "step": 24515 }, { "epoch": 2.657849089332177, "grad_norm": 0.07226426154375076, "learning_rate": 1.1446394893370088e-05, "loss": 0.0045, "step": 24516 }, { "epoch": 2.657957502168257, "grad_norm": 0.1683238297700882, "learning_rate": 1.144276802553315e-05, "loss": 0.0053, "step": 24517 }, { "epoch": 2.6580659150043364, "grad_norm": 0.2321484535932541, "learning_rate": 1.1439141157696214e-05, "loss": 0.0076, "step": 24518 }, { "epoch": 2.6581743278404164, "grad_norm": 0.002361508086323738, "learning_rate": 1.1435514289859278e-05, "loss": 0.0001, "step": 24519 }, { "epoch": 2.658282740676496, "grad_norm": 0.08132611960172653, "learning_rate": 1.1431887422022342e-05, "loss": 0.003, "step": 24520 }, { "epoch": 2.658391153512576, "grad_norm": 0.3685363531112671, "learning_rate": 1.1428260554185406e-05, "loss": 0.0043, "step": 24521 }, { "epoch": 2.658499566348656, "grad_norm": 0.27264124155044556, "learning_rate": 1.1424633686348469e-05, "loss": 0.0058, "step": 24522 }, { "epoch": 2.6586079791847355, "grad_norm": 0.007678070105612278, "learning_rate": 1.1421006818511535e-05, "loss": 0.0002, "step": 24523 }, { "epoch": 2.658716392020815, "grad_norm": 0.20457226037979126, "learning_rate": 1.1417379950674597e-05, "loss": 0.0055, "step": 24524 }, { "epoch": 2.658824804856895, "grad_norm": 0.7888684868812561, "learning_rate": 1.1413753082837663e-05, "loss": 0.0175, "step": 24525 }, { "epoch": 2.658933217692975, "grad_norm": 0.180334210395813, "learning_rate": 1.1410126215000725e-05, "loss": 0.0083, "step": 24526 }, { "epoch": 2.6590416305290545, "grad_norm": 0.054217629134655, "learning_rate": 1.1406499347163791e-05, "loss": 0.0009, "step": 24527 }, { "epoch": 2.6591500433651345, "grad_norm": 0.24494241178035736, "learning_rate": 1.1402872479326853e-05, "loss": 0.0051, "step": 24528 }, { "epoch": 2.659258456201214, "grad_norm": 0.2162473052740097, "learning_rate": 1.139924561148992e-05, "loss": 0.0137, "step": 24529 }, { "epoch": 2.659366869037294, "grad_norm": 0.0527154840528965, "learning_rate": 1.1395618743652982e-05, "loss": 0.0018, "step": 24530 }, { "epoch": 2.659475281873374, "grad_norm": 0.3884638249874115, "learning_rate": 1.1391991875816046e-05, "loss": 0.0208, "step": 24531 }, { "epoch": 2.6595836947094535, "grad_norm": 0.3267013430595398, "learning_rate": 1.138836500797911e-05, "loss": 0.0245, "step": 24532 }, { "epoch": 2.6596921075455335, "grad_norm": 0.048034921288490295, "learning_rate": 1.1384738140142174e-05, "loss": 0.0011, "step": 24533 }, { "epoch": 2.659800520381613, "grad_norm": 0.1540808081626892, "learning_rate": 1.1381111272305238e-05, "loss": 0.0052, "step": 24534 }, { "epoch": 2.659908933217693, "grad_norm": 0.019423872232437134, "learning_rate": 1.13774844044683e-05, "loss": 0.0007, "step": 24535 }, { "epoch": 2.6600173460537726, "grad_norm": 0.2646021246910095, "learning_rate": 1.1373857536631366e-05, "loss": 0.0079, "step": 24536 }, { "epoch": 2.6601257588898526, "grad_norm": 0.3464758098125458, "learning_rate": 1.1370230668794429e-05, "loss": 0.0064, "step": 24537 }, { "epoch": 2.660234171725932, "grad_norm": 0.01038351096212864, "learning_rate": 1.1366603800957494e-05, "loss": 0.0005, "step": 24538 }, { "epoch": 2.660342584562012, "grad_norm": 0.13909053802490234, "learning_rate": 1.1362976933120557e-05, "loss": 0.002, "step": 24539 }, { "epoch": 2.660450997398092, "grad_norm": 0.42104488611221313, "learning_rate": 1.1359350065283623e-05, "loss": 0.0085, "step": 24540 }, { "epoch": 2.6605594102341716, "grad_norm": 0.4863753318786621, "learning_rate": 1.1355723197446685e-05, "loss": 0.0089, "step": 24541 }, { "epoch": 2.6606678230702516, "grad_norm": 0.30469051003456116, "learning_rate": 1.1352096329609751e-05, "loss": 0.0111, "step": 24542 }, { "epoch": 2.660776235906331, "grad_norm": 0.2438877373933792, "learning_rate": 1.1348469461772813e-05, "loss": 0.0097, "step": 24543 }, { "epoch": 2.660884648742411, "grad_norm": 0.00912232231348753, "learning_rate": 1.1344842593935877e-05, "loss": 0.0003, "step": 24544 }, { "epoch": 2.660993061578491, "grad_norm": 0.3190235495567322, "learning_rate": 1.1341215726098942e-05, "loss": 0.0067, "step": 24545 }, { "epoch": 2.6611014744145707, "grad_norm": 0.34746575355529785, "learning_rate": 1.1337588858262006e-05, "loss": 0.0317, "step": 24546 }, { "epoch": 2.66120988725065, "grad_norm": 0.05251516401767731, "learning_rate": 1.133396199042507e-05, "loss": 0.0027, "step": 24547 }, { "epoch": 2.66131830008673, "grad_norm": 0.30085617303848267, "learning_rate": 1.1330335122588132e-05, "loss": 0.038, "step": 24548 }, { "epoch": 2.66142671292281, "grad_norm": 0.05008244514465332, "learning_rate": 1.1326708254751198e-05, "loss": 0.0015, "step": 24549 }, { "epoch": 2.6615351257588897, "grad_norm": 0.7139992117881775, "learning_rate": 1.132308138691426e-05, "loss": 0.0384, "step": 24550 }, { "epoch": 2.6616435385949697, "grad_norm": 0.2758225202560425, "learning_rate": 1.1319454519077326e-05, "loss": 0.0062, "step": 24551 }, { "epoch": 2.6617519514310493, "grad_norm": 0.36705684661865234, "learning_rate": 1.1315827651240389e-05, "loss": 0.0094, "step": 24552 }, { "epoch": 2.6618603642671292, "grad_norm": 0.23231244087219238, "learning_rate": 1.1312200783403454e-05, "loss": 0.0051, "step": 24553 }, { "epoch": 2.6619687771032092, "grad_norm": 0.005131023935973644, "learning_rate": 1.1308573915566517e-05, "loss": 0.0003, "step": 24554 }, { "epoch": 2.6620771899392888, "grad_norm": 0.2006380558013916, "learning_rate": 1.1304947047729583e-05, "loss": 0.0051, "step": 24555 }, { "epoch": 2.6621856027753688, "grad_norm": 0.10121537744998932, "learning_rate": 1.1301320179892645e-05, "loss": 0.0046, "step": 24556 }, { "epoch": 2.6622940156114483, "grad_norm": 0.131924569606781, "learning_rate": 1.1297693312055709e-05, "loss": 0.0048, "step": 24557 }, { "epoch": 2.6624024284475283, "grad_norm": 0.170563206076622, "learning_rate": 1.1294066444218773e-05, "loss": 0.0075, "step": 24558 }, { "epoch": 2.662510841283608, "grad_norm": 0.2082175463438034, "learning_rate": 1.1290439576381837e-05, "loss": 0.0094, "step": 24559 }, { "epoch": 2.662619254119688, "grad_norm": 0.0932091772556305, "learning_rate": 1.1286812708544901e-05, "loss": 0.0026, "step": 24560 }, { "epoch": 2.6627276669557673, "grad_norm": 0.37721213698387146, "learning_rate": 1.1283185840707964e-05, "loss": 0.0244, "step": 24561 }, { "epoch": 2.6628360797918473, "grad_norm": 0.1586003303527832, "learning_rate": 1.127955897287103e-05, "loss": 0.003, "step": 24562 }, { "epoch": 2.6629444926279273, "grad_norm": 0.3423826992511749, "learning_rate": 1.1275932105034092e-05, "loss": 0.0099, "step": 24563 }, { "epoch": 2.663052905464007, "grad_norm": 0.02859671227633953, "learning_rate": 1.1272305237197158e-05, "loss": 0.0011, "step": 24564 }, { "epoch": 2.663161318300087, "grad_norm": 0.29399722814559937, "learning_rate": 1.126867836936022e-05, "loss": 0.0108, "step": 24565 }, { "epoch": 2.6632697311361664, "grad_norm": 1.1264609098434448, "learning_rate": 1.1265051501523286e-05, "loss": 0.057, "step": 24566 }, { "epoch": 2.6633781439722464, "grad_norm": 0.1339339315891266, "learning_rate": 1.1261424633686348e-05, "loss": 0.005, "step": 24567 }, { "epoch": 2.6634865568083264, "grad_norm": 0.5363168120384216, "learning_rate": 1.1257797765849414e-05, "loss": 0.0162, "step": 24568 }, { "epoch": 2.663594969644406, "grad_norm": 0.006536179222166538, "learning_rate": 1.1254170898012477e-05, "loss": 0.0003, "step": 24569 }, { "epoch": 2.6637033824804854, "grad_norm": 0.0881006270647049, "learning_rate": 1.125054403017554e-05, "loss": 0.002, "step": 24570 }, { "epoch": 2.6638117953165654, "grad_norm": 0.015318399295210838, "learning_rate": 1.1246917162338605e-05, "loss": 0.0004, "step": 24571 }, { "epoch": 2.6639202081526454, "grad_norm": 0.48571258783340454, "learning_rate": 1.1243290294501669e-05, "loss": 0.0288, "step": 24572 }, { "epoch": 2.664028620988725, "grad_norm": 0.0428212471306324, "learning_rate": 1.1239663426664733e-05, "loss": 0.0021, "step": 24573 }, { "epoch": 2.664137033824805, "grad_norm": 0.14094235002994537, "learning_rate": 1.1236036558827795e-05, "loss": 0.0035, "step": 24574 }, { "epoch": 2.6642454466608845, "grad_norm": 0.025683196261525154, "learning_rate": 1.1232409690990861e-05, "loss": 0.0011, "step": 24575 }, { "epoch": 2.6643538594969645, "grad_norm": 0.5901579260826111, "learning_rate": 1.1228782823153924e-05, "loss": 0.0021, "step": 24576 }, { "epoch": 2.6644622723330444, "grad_norm": 0.21009743213653564, "learning_rate": 1.122515595531699e-05, "loss": 0.0095, "step": 24577 }, { "epoch": 2.664570685169124, "grad_norm": 0.07757120579481125, "learning_rate": 1.1221529087480052e-05, "loss": 0.0035, "step": 24578 }, { "epoch": 2.664679098005204, "grad_norm": 0.03992731496691704, "learning_rate": 1.1217902219643118e-05, "loss": 0.0012, "step": 24579 }, { "epoch": 2.6647875108412835, "grad_norm": 0.10152532160282135, "learning_rate": 1.121427535180618e-05, "loss": 0.0065, "step": 24580 }, { "epoch": 2.6648959236773635, "grad_norm": 0.03246917203068733, "learning_rate": 1.1210648483969246e-05, "loss": 0.001, "step": 24581 }, { "epoch": 2.665004336513443, "grad_norm": 0.27202489972114563, "learning_rate": 1.1207021616132308e-05, "loss": 0.0304, "step": 24582 }, { "epoch": 2.665112749349523, "grad_norm": 0.12410921603441238, "learning_rate": 1.1203394748295374e-05, "loss": 0.0031, "step": 24583 }, { "epoch": 2.6652211621856026, "grad_norm": 0.6566013693809509, "learning_rate": 1.1199767880458437e-05, "loss": 0.0431, "step": 24584 }, { "epoch": 2.6653295750216826, "grad_norm": 0.6689053773880005, "learning_rate": 1.11961410126215e-05, "loss": 0.0101, "step": 24585 }, { "epoch": 2.6654379878577625, "grad_norm": 0.31584450602531433, "learning_rate": 1.1192514144784565e-05, "loss": 0.0059, "step": 24586 }, { "epoch": 2.665546400693842, "grad_norm": 0.032363951206207275, "learning_rate": 1.1188887276947629e-05, "loss": 0.0016, "step": 24587 }, { "epoch": 2.665654813529922, "grad_norm": 0.18155096471309662, "learning_rate": 1.1185260409110693e-05, "loss": 0.0049, "step": 24588 }, { "epoch": 2.6657632263660016, "grad_norm": 0.06043308228254318, "learning_rate": 1.1181633541273755e-05, "loss": 0.0028, "step": 24589 }, { "epoch": 2.6658716392020816, "grad_norm": 0.0696878433227539, "learning_rate": 1.1178006673436821e-05, "loss": 0.0021, "step": 24590 }, { "epoch": 2.6659800520381616, "grad_norm": 0.18430307507514954, "learning_rate": 1.1174379805599884e-05, "loss": 0.0042, "step": 24591 }, { "epoch": 2.666088464874241, "grad_norm": 0.4304974675178528, "learning_rate": 1.117075293776295e-05, "loss": 0.0113, "step": 24592 }, { "epoch": 2.6661968777103207, "grad_norm": 0.01373264379799366, "learning_rate": 1.1167126069926012e-05, "loss": 0.0004, "step": 24593 }, { "epoch": 2.6663052905464006, "grad_norm": 0.7109944224357605, "learning_rate": 1.1163499202089078e-05, "loss": 0.018, "step": 24594 }, { "epoch": 2.6664137033824806, "grad_norm": 0.0005021812976337969, "learning_rate": 1.115987233425214e-05, "loss": 0.0, "step": 24595 }, { "epoch": 2.66652211621856, "grad_norm": 0.034429192543029785, "learning_rate": 1.1156245466415206e-05, "loss": 0.0013, "step": 24596 }, { "epoch": 2.66663052905464, "grad_norm": 0.037137679755687714, "learning_rate": 1.1152618598578268e-05, "loss": 0.0021, "step": 24597 }, { "epoch": 2.6667389418907197, "grad_norm": 0.40755194425582886, "learning_rate": 1.1148991730741332e-05, "loss": 0.0225, "step": 24598 }, { "epoch": 2.6668473547267997, "grad_norm": 0.28457391262054443, "learning_rate": 1.1145364862904396e-05, "loss": 0.0091, "step": 24599 }, { "epoch": 2.6669557675628797, "grad_norm": 0.21721385419368744, "learning_rate": 1.114173799506746e-05, "loss": 0.0101, "step": 24600 }, { "epoch": 2.667064180398959, "grad_norm": 0.32133743166923523, "learning_rate": 1.1138111127230525e-05, "loss": 0.0087, "step": 24601 }, { "epoch": 2.667172593235039, "grad_norm": 0.6366978883743286, "learning_rate": 1.1134484259393587e-05, "loss": 0.0185, "step": 24602 }, { "epoch": 2.6672810060711187, "grad_norm": 0.46238693594932556, "learning_rate": 1.1130857391556653e-05, "loss": 0.0054, "step": 24603 }, { "epoch": 2.6673894189071987, "grad_norm": 0.11422207951545715, "learning_rate": 1.1127230523719715e-05, "loss": 0.0019, "step": 24604 }, { "epoch": 2.6674978317432783, "grad_norm": 0.09753943979740143, "learning_rate": 1.1123603655882781e-05, "loss": 0.0031, "step": 24605 }, { "epoch": 2.6676062445793582, "grad_norm": 0.4218962490558624, "learning_rate": 1.1119976788045843e-05, "loss": 0.0109, "step": 24606 }, { "epoch": 2.667714657415438, "grad_norm": 0.3553309738636017, "learning_rate": 1.111634992020891e-05, "loss": 0.0218, "step": 24607 }, { "epoch": 2.6678230702515178, "grad_norm": 0.6323025226593018, "learning_rate": 1.1112723052371972e-05, "loss": 0.0179, "step": 24608 }, { "epoch": 2.6679314830875978, "grad_norm": 0.45110785961151123, "learning_rate": 1.1109096184535037e-05, "loss": 0.0295, "step": 24609 }, { "epoch": 2.6680398959236773, "grad_norm": 0.09255235642194748, "learning_rate": 1.11054693166981e-05, "loss": 0.0035, "step": 24610 }, { "epoch": 2.6681483087597573, "grad_norm": 0.16642266511917114, "learning_rate": 1.1101842448861164e-05, "loss": 0.0021, "step": 24611 }, { "epoch": 2.668256721595837, "grad_norm": 0.49116384983062744, "learning_rate": 1.1098215581024228e-05, "loss": 0.021, "step": 24612 }, { "epoch": 2.668365134431917, "grad_norm": 0.057111307978630066, "learning_rate": 1.1094588713187292e-05, "loss": 0.0015, "step": 24613 }, { "epoch": 2.668473547267997, "grad_norm": 0.02515004388988018, "learning_rate": 1.1090961845350356e-05, "loss": 0.0007, "step": 24614 }, { "epoch": 2.6685819601040763, "grad_norm": 0.16194532811641693, "learning_rate": 1.1087334977513419e-05, "loss": 0.0079, "step": 24615 }, { "epoch": 2.668690372940156, "grad_norm": 0.9165415167808533, "learning_rate": 1.1083708109676485e-05, "loss": 0.0241, "step": 24616 }, { "epoch": 2.668798785776236, "grad_norm": 0.7250927090644836, "learning_rate": 1.1080081241839547e-05, "loss": 0.0135, "step": 24617 }, { "epoch": 2.668907198612316, "grad_norm": 0.1056106835603714, "learning_rate": 1.1076454374002613e-05, "loss": 0.0023, "step": 24618 }, { "epoch": 2.6690156114483954, "grad_norm": 0.06589653342962265, "learning_rate": 1.1072827506165675e-05, "loss": 0.0034, "step": 24619 }, { "epoch": 2.6691240242844754, "grad_norm": 0.3047848343849182, "learning_rate": 1.1069200638328741e-05, "loss": 0.0096, "step": 24620 }, { "epoch": 2.669232437120555, "grad_norm": 0.026268312707543373, "learning_rate": 1.1065573770491803e-05, "loss": 0.0008, "step": 24621 }, { "epoch": 2.669340849956635, "grad_norm": 0.022808071225881577, "learning_rate": 1.1061946902654869e-05, "loss": 0.0008, "step": 24622 }, { "epoch": 2.669449262792715, "grad_norm": 0.0495765246450901, "learning_rate": 1.1058320034817932e-05, "loss": 0.0009, "step": 24623 }, { "epoch": 2.6695576756287944, "grad_norm": 0.19182275235652924, "learning_rate": 1.1054693166980996e-05, "loss": 0.0099, "step": 24624 }, { "epoch": 2.669666088464874, "grad_norm": 0.2832716405391693, "learning_rate": 1.105106629914406e-05, "loss": 0.0065, "step": 24625 }, { "epoch": 2.669774501300954, "grad_norm": 0.02006353624165058, "learning_rate": 1.1047439431307124e-05, "loss": 0.0008, "step": 24626 }, { "epoch": 2.669882914137034, "grad_norm": 0.02166622504591942, "learning_rate": 1.1043812563470188e-05, "loss": 0.0006, "step": 24627 }, { "epoch": 2.6699913269731135, "grad_norm": 0.12877458333969116, "learning_rate": 1.104018569563325e-05, "loss": 0.0044, "step": 24628 }, { "epoch": 2.6700997398091935, "grad_norm": 0.20266786217689514, "learning_rate": 1.1036558827796316e-05, "loss": 0.0075, "step": 24629 }, { "epoch": 2.670208152645273, "grad_norm": 0.01991160400211811, "learning_rate": 1.1032931959959379e-05, "loss": 0.0011, "step": 24630 }, { "epoch": 2.670316565481353, "grad_norm": 0.2618618905544281, "learning_rate": 1.1029305092122444e-05, "loss": 0.0164, "step": 24631 }, { "epoch": 2.670424978317433, "grad_norm": 1.1133737564086914, "learning_rate": 1.1025678224285507e-05, "loss": 0.0411, "step": 24632 }, { "epoch": 2.6705333911535125, "grad_norm": 0.17982333898544312, "learning_rate": 1.1022051356448573e-05, "loss": 0.0096, "step": 24633 }, { "epoch": 2.6706418039895925, "grad_norm": 0.3336167335510254, "learning_rate": 1.1018424488611635e-05, "loss": 0.0339, "step": 24634 }, { "epoch": 2.670750216825672, "grad_norm": 0.16527925431728363, "learning_rate": 1.10147976207747e-05, "loss": 0.0042, "step": 24635 }, { "epoch": 2.670858629661752, "grad_norm": 0.5659811496734619, "learning_rate": 1.1011170752937763e-05, "loss": 0.0193, "step": 24636 }, { "epoch": 2.670967042497832, "grad_norm": 0.017629535868763924, "learning_rate": 1.1007543885100827e-05, "loss": 0.0004, "step": 24637 }, { "epoch": 2.6710754553339116, "grad_norm": 0.22245904803276062, "learning_rate": 1.1003917017263891e-05, "loss": 0.0292, "step": 24638 }, { "epoch": 2.671183868169991, "grad_norm": 0.4754694104194641, "learning_rate": 1.1000290149426956e-05, "loss": 0.0111, "step": 24639 }, { "epoch": 2.671292281006071, "grad_norm": 0.008982932195067406, "learning_rate": 1.099666328159002e-05, "loss": 0.0004, "step": 24640 }, { "epoch": 2.671400693842151, "grad_norm": 0.47386887669563293, "learning_rate": 1.0993036413753082e-05, "loss": 0.0368, "step": 24641 }, { "epoch": 2.6715091066782306, "grad_norm": 0.19520965218544006, "learning_rate": 1.0989409545916148e-05, "loss": 0.0064, "step": 24642 }, { "epoch": 2.6716175195143106, "grad_norm": 1.0266358852386475, "learning_rate": 1.098578267807921e-05, "loss": 0.0228, "step": 24643 }, { "epoch": 2.67172593235039, "grad_norm": 0.03239065781235695, "learning_rate": 1.0982155810242276e-05, "loss": 0.0011, "step": 24644 }, { "epoch": 2.67183434518647, "grad_norm": 0.031126633286476135, "learning_rate": 1.0978528942405338e-05, "loss": 0.0007, "step": 24645 }, { "epoch": 2.67194275802255, "grad_norm": 0.27344977855682373, "learning_rate": 1.0974902074568404e-05, "loss": 0.0136, "step": 24646 }, { "epoch": 2.6720511708586296, "grad_norm": 0.05664873495697975, "learning_rate": 1.0971275206731467e-05, "loss": 0.0014, "step": 24647 }, { "epoch": 2.672159583694709, "grad_norm": 0.12724345922470093, "learning_rate": 1.0967648338894532e-05, "loss": 0.0038, "step": 24648 }, { "epoch": 2.672267996530789, "grad_norm": 0.1535516381263733, "learning_rate": 1.0964021471057595e-05, "loss": 0.0117, "step": 24649 }, { "epoch": 2.672376409366869, "grad_norm": 0.08577253669500351, "learning_rate": 1.0960394603220659e-05, "loss": 0.003, "step": 24650 }, { "epoch": 2.6724848222029487, "grad_norm": 0.028697846457362175, "learning_rate": 1.0956767735383723e-05, "loss": 0.0012, "step": 24651 }, { "epoch": 2.6725932350390287, "grad_norm": 0.4119138717651367, "learning_rate": 1.0953140867546787e-05, "loss": 0.0091, "step": 24652 }, { "epoch": 2.672701647875108, "grad_norm": 0.06371291726827621, "learning_rate": 1.0949513999709851e-05, "loss": 0.0019, "step": 24653 }, { "epoch": 2.672810060711188, "grad_norm": 0.07806721329689026, "learning_rate": 1.0945887131872914e-05, "loss": 0.0037, "step": 24654 }, { "epoch": 2.672918473547268, "grad_norm": 0.6507779359817505, "learning_rate": 1.094226026403598e-05, "loss": 0.0154, "step": 24655 }, { "epoch": 2.6730268863833477, "grad_norm": 0.09150074422359467, "learning_rate": 1.0938633396199042e-05, "loss": 0.0037, "step": 24656 }, { "epoch": 2.6731352992194277, "grad_norm": 0.018848204985260963, "learning_rate": 1.0935006528362108e-05, "loss": 0.0006, "step": 24657 }, { "epoch": 2.6732437120555073, "grad_norm": 0.15699538588523865, "learning_rate": 1.093137966052517e-05, "loss": 0.0031, "step": 24658 }, { "epoch": 2.6733521248915872, "grad_norm": 0.21732929348945618, "learning_rate": 1.0927752792688236e-05, "loss": 0.0062, "step": 24659 }, { "epoch": 2.6734605377276672, "grad_norm": 0.12678754329681396, "learning_rate": 1.0924125924851298e-05, "loss": 0.012, "step": 24660 }, { "epoch": 2.6735689505637468, "grad_norm": 0.05866123363375664, "learning_rate": 1.0920499057014364e-05, "loss": 0.0008, "step": 24661 }, { "epoch": 2.6736773633998263, "grad_norm": 0.3123863935470581, "learning_rate": 1.0916872189177427e-05, "loss": 0.0083, "step": 24662 }, { "epoch": 2.6737857762359063, "grad_norm": 0.21518802642822266, "learning_rate": 1.091324532134049e-05, "loss": 0.0033, "step": 24663 }, { "epoch": 2.6738941890719863, "grad_norm": 0.3169326186180115, "learning_rate": 1.0909618453503555e-05, "loss": 0.0318, "step": 24664 }, { "epoch": 2.674002601908066, "grad_norm": 0.027215389534831047, "learning_rate": 1.0905991585666619e-05, "loss": 0.0008, "step": 24665 }, { "epoch": 2.674111014744146, "grad_norm": 0.03317982703447342, "learning_rate": 1.0902364717829683e-05, "loss": 0.0019, "step": 24666 }, { "epoch": 2.6742194275802254, "grad_norm": 0.136710524559021, "learning_rate": 1.0898737849992745e-05, "loss": 0.0024, "step": 24667 }, { "epoch": 2.6743278404163053, "grad_norm": 0.45553845167160034, "learning_rate": 1.0895110982155811e-05, "loss": 0.0095, "step": 24668 }, { "epoch": 2.6744362532523853, "grad_norm": 0.10372424870729446, "learning_rate": 1.0891484114318874e-05, "loss": 0.0034, "step": 24669 }, { "epoch": 2.674544666088465, "grad_norm": 0.03432851284742355, "learning_rate": 1.088785724648194e-05, "loss": 0.0016, "step": 24670 }, { "epoch": 2.6746530789245444, "grad_norm": 0.1425633430480957, "learning_rate": 1.0884230378645002e-05, "loss": 0.0215, "step": 24671 }, { "epoch": 2.6747614917606244, "grad_norm": 0.6898622512817383, "learning_rate": 1.0880603510808068e-05, "loss": 0.005, "step": 24672 }, { "epoch": 2.6748699045967044, "grad_norm": 0.025103796273469925, "learning_rate": 1.087697664297113e-05, "loss": 0.001, "step": 24673 }, { "epoch": 2.674978317432784, "grad_norm": 0.22001436352729797, "learning_rate": 1.0873349775134196e-05, "loss": 0.0225, "step": 24674 }, { "epoch": 2.675086730268864, "grad_norm": 0.022142181172966957, "learning_rate": 1.0869722907297258e-05, "loss": 0.0006, "step": 24675 }, { "epoch": 2.6751951431049434, "grad_norm": 0.22290223836898804, "learning_rate": 1.0866096039460322e-05, "loss": 0.0096, "step": 24676 }, { "epoch": 2.6753035559410234, "grad_norm": 0.1911383718252182, "learning_rate": 1.0862469171623386e-05, "loss": 0.0113, "step": 24677 }, { "epoch": 2.6754119687771034, "grad_norm": 0.6228031516075134, "learning_rate": 1.085884230378645e-05, "loss": 0.027, "step": 24678 }, { "epoch": 2.675520381613183, "grad_norm": 0.48164474964141846, "learning_rate": 1.0855215435949515e-05, "loss": 0.0238, "step": 24679 }, { "epoch": 2.675628794449263, "grad_norm": 0.3465045094490051, "learning_rate": 1.0851588568112577e-05, "loss": 0.0076, "step": 24680 }, { "epoch": 2.6757372072853425, "grad_norm": 0.14140866696834564, "learning_rate": 1.0847961700275643e-05, "loss": 0.0044, "step": 24681 }, { "epoch": 2.6758456201214225, "grad_norm": 0.04288231581449509, "learning_rate": 1.0844334832438705e-05, "loss": 0.0028, "step": 24682 }, { "epoch": 2.675954032957502, "grad_norm": 1.2002636194229126, "learning_rate": 1.0840707964601771e-05, "loss": 0.0117, "step": 24683 }, { "epoch": 2.676062445793582, "grad_norm": 0.02245161309838295, "learning_rate": 1.0837081096764833e-05, "loss": 0.0008, "step": 24684 }, { "epoch": 2.6761708586296615, "grad_norm": 0.04041930288076401, "learning_rate": 1.08334542289279e-05, "loss": 0.0018, "step": 24685 }, { "epoch": 2.6762792714657415, "grad_norm": 0.3404718041419983, "learning_rate": 1.0829827361090962e-05, "loss": 0.0121, "step": 24686 }, { "epoch": 2.6763876843018215, "grad_norm": 0.07847652584314346, "learning_rate": 1.0826200493254027e-05, "loss": 0.0045, "step": 24687 }, { "epoch": 2.676496097137901, "grad_norm": 0.05412865802645683, "learning_rate": 1.082257362541709e-05, "loss": 0.0017, "step": 24688 }, { "epoch": 2.676604509973981, "grad_norm": 0.14333079755306244, "learning_rate": 1.0818946757580154e-05, "loss": 0.0028, "step": 24689 }, { "epoch": 2.6767129228100606, "grad_norm": 0.6492358446121216, "learning_rate": 1.0815319889743218e-05, "loss": 0.0344, "step": 24690 }, { "epoch": 2.6768213356461406, "grad_norm": 0.7298753261566162, "learning_rate": 1.0811693021906282e-05, "loss": 0.0463, "step": 24691 }, { "epoch": 2.6769297484822205, "grad_norm": 0.044834915548563004, "learning_rate": 1.0808066154069346e-05, "loss": 0.0017, "step": 24692 }, { "epoch": 2.6770381613183, "grad_norm": 0.27669262886047363, "learning_rate": 1.0804439286232409e-05, "loss": 0.0098, "step": 24693 }, { "epoch": 2.6771465741543796, "grad_norm": 0.0533253476023674, "learning_rate": 1.0800812418395475e-05, "loss": 0.0015, "step": 24694 }, { "epoch": 2.6772549869904596, "grad_norm": 0.010697704739868641, "learning_rate": 1.0797185550558537e-05, "loss": 0.0004, "step": 24695 }, { "epoch": 2.6773633998265396, "grad_norm": 0.04421784356236458, "learning_rate": 1.0793558682721603e-05, "loss": 0.0017, "step": 24696 }, { "epoch": 2.677471812662619, "grad_norm": 0.1882835179567337, "learning_rate": 1.0789931814884665e-05, "loss": 0.0022, "step": 24697 }, { "epoch": 2.677580225498699, "grad_norm": 0.041104599833488464, "learning_rate": 1.0786304947047731e-05, "loss": 0.0019, "step": 24698 }, { "epoch": 2.6776886383347787, "grad_norm": 0.23412321507930756, "learning_rate": 1.0782678079210793e-05, "loss": 0.0074, "step": 24699 }, { "epoch": 2.6777970511708586, "grad_norm": 0.02684859000146389, "learning_rate": 1.077905121137386e-05, "loss": 0.0018, "step": 24700 }, { "epoch": 2.6779054640069386, "grad_norm": 0.1986325979232788, "learning_rate": 1.0775424343536922e-05, "loss": 0.0299, "step": 24701 }, { "epoch": 2.678013876843018, "grad_norm": 0.40509724617004395, "learning_rate": 1.0771797475699986e-05, "loss": 0.0276, "step": 24702 }, { "epoch": 2.678122289679098, "grad_norm": 0.15238356590270996, "learning_rate": 1.076817060786305e-05, "loss": 0.0032, "step": 24703 }, { "epoch": 2.6782307025151777, "grad_norm": 0.7087475657463074, "learning_rate": 1.0764543740026114e-05, "loss": 0.0244, "step": 24704 }, { "epoch": 2.6783391153512577, "grad_norm": 0.43330806493759155, "learning_rate": 1.0760916872189178e-05, "loss": 0.0364, "step": 24705 }, { "epoch": 2.6784475281873372, "grad_norm": 0.0376439206302166, "learning_rate": 1.0757290004352242e-05, "loss": 0.0014, "step": 24706 }, { "epoch": 2.678555941023417, "grad_norm": 0.2553424835205078, "learning_rate": 1.0753663136515306e-05, "loss": 0.0275, "step": 24707 }, { "epoch": 2.6786643538594967, "grad_norm": 0.30654919147491455, "learning_rate": 1.0750036268678369e-05, "loss": 0.0094, "step": 24708 }, { "epoch": 2.6787727666955767, "grad_norm": 0.19554615020751953, "learning_rate": 1.0746409400841434e-05, "loss": 0.0072, "step": 24709 }, { "epoch": 2.6788811795316567, "grad_norm": 0.3952511250972748, "learning_rate": 1.0742782533004497e-05, "loss": 0.0416, "step": 24710 }, { "epoch": 2.6789895923677363, "grad_norm": 0.30534446239471436, "learning_rate": 1.0739155665167563e-05, "loss": 0.0172, "step": 24711 }, { "epoch": 2.6790980052038162, "grad_norm": 0.05377117544412613, "learning_rate": 1.0735528797330625e-05, "loss": 0.0018, "step": 24712 }, { "epoch": 2.679206418039896, "grad_norm": 0.10605897009372711, "learning_rate": 1.073190192949369e-05, "loss": 0.0042, "step": 24713 }, { "epoch": 2.6793148308759758, "grad_norm": 0.2898287773132324, "learning_rate": 1.0728275061656753e-05, "loss": 0.0065, "step": 24714 }, { "epoch": 2.6794232437120558, "grad_norm": 0.09754633158445358, "learning_rate": 1.0724648193819817e-05, "loss": 0.0014, "step": 24715 }, { "epoch": 2.6795316565481353, "grad_norm": 0.03556609898805618, "learning_rate": 1.0721021325982881e-05, "loss": 0.0017, "step": 24716 }, { "epoch": 2.679640069384215, "grad_norm": 0.396084189414978, "learning_rate": 1.0717394458145946e-05, "loss": 0.0186, "step": 24717 }, { "epoch": 2.679748482220295, "grad_norm": 0.13475488126277924, "learning_rate": 1.071376759030901e-05, "loss": 0.0031, "step": 24718 }, { "epoch": 2.679856895056375, "grad_norm": 0.36596134305000305, "learning_rate": 1.0710140722472074e-05, "loss": 0.0149, "step": 24719 }, { "epoch": 2.6799653078924544, "grad_norm": 0.1650235652923584, "learning_rate": 1.0706513854635138e-05, "loss": 0.004, "step": 24720 }, { "epoch": 2.6800737207285343, "grad_norm": 0.7312959432601929, "learning_rate": 1.07028869867982e-05, "loss": 0.0173, "step": 24721 }, { "epoch": 2.680182133564614, "grad_norm": 0.32282140851020813, "learning_rate": 1.0699260118961266e-05, "loss": 0.0413, "step": 24722 }, { "epoch": 2.680290546400694, "grad_norm": 0.09990152716636658, "learning_rate": 1.0695633251124329e-05, "loss": 0.0025, "step": 24723 }, { "epoch": 2.680398959236774, "grad_norm": 0.03529422730207443, "learning_rate": 1.0692006383287394e-05, "loss": 0.0019, "step": 24724 }, { "epoch": 2.6805073720728534, "grad_norm": 0.06681540608406067, "learning_rate": 1.0688379515450457e-05, "loss": 0.0031, "step": 24725 }, { "epoch": 2.6806157849089334, "grad_norm": 0.44182732701301575, "learning_rate": 1.0684752647613523e-05, "loss": 0.0489, "step": 24726 }, { "epoch": 2.680724197745013, "grad_norm": 0.10000045597553253, "learning_rate": 1.0681125779776585e-05, "loss": 0.0052, "step": 24727 }, { "epoch": 2.680832610581093, "grad_norm": 0.01474437303841114, "learning_rate": 1.0677498911939649e-05, "loss": 0.0006, "step": 24728 }, { "epoch": 2.6809410234171724, "grad_norm": 1.1965786218643188, "learning_rate": 1.0673872044102713e-05, "loss": 0.0295, "step": 24729 }, { "epoch": 2.6810494362532524, "grad_norm": 0.536025881767273, "learning_rate": 1.0670245176265777e-05, "loss": 0.0176, "step": 24730 }, { "epoch": 2.681157849089332, "grad_norm": 0.04963372275233269, "learning_rate": 1.0666618308428841e-05, "loss": 0.0018, "step": 24731 }, { "epoch": 2.681266261925412, "grad_norm": 0.07388631254434586, "learning_rate": 1.0662991440591905e-05, "loss": 0.0015, "step": 24732 }, { "epoch": 2.681374674761492, "grad_norm": 0.17256808280944824, "learning_rate": 1.065936457275497e-05, "loss": 0.003, "step": 24733 }, { "epoch": 2.6814830875975715, "grad_norm": 0.36480987071990967, "learning_rate": 1.0655737704918032e-05, "loss": 0.0307, "step": 24734 }, { "epoch": 2.6815915004336515, "grad_norm": 0.25355029106140137, "learning_rate": 1.0652110837081098e-05, "loss": 0.0054, "step": 24735 }, { "epoch": 2.681699913269731, "grad_norm": 0.6438297629356384, "learning_rate": 1.064848396924416e-05, "loss": 0.0319, "step": 24736 }, { "epoch": 2.681808326105811, "grad_norm": 0.10692254453897476, "learning_rate": 1.0644857101407226e-05, "loss": 0.0029, "step": 24737 }, { "epoch": 2.681916738941891, "grad_norm": 0.0019987442065030336, "learning_rate": 1.0641230233570288e-05, "loss": 0.0001, "step": 24738 }, { "epoch": 2.6820251517779705, "grad_norm": 0.004142913967370987, "learning_rate": 1.0637603365733354e-05, "loss": 0.0001, "step": 24739 }, { "epoch": 2.68213356461405, "grad_norm": 0.49977144598960876, "learning_rate": 1.0633976497896417e-05, "loss": 0.0191, "step": 24740 }, { "epoch": 2.68224197745013, "grad_norm": 0.12045430392026901, "learning_rate": 1.063034963005948e-05, "loss": 0.0029, "step": 24741 }, { "epoch": 2.68235039028621, "grad_norm": 0.057020775973796844, "learning_rate": 1.0626722762222545e-05, "loss": 0.0032, "step": 24742 }, { "epoch": 2.6824588031222896, "grad_norm": 0.2765911817550659, "learning_rate": 1.0623095894385609e-05, "loss": 0.0079, "step": 24743 }, { "epoch": 2.6825672159583696, "grad_norm": 0.022680876776576042, "learning_rate": 1.0619469026548673e-05, "loss": 0.0005, "step": 24744 }, { "epoch": 2.682675628794449, "grad_norm": 0.7963927984237671, "learning_rate": 1.0615842158711737e-05, "loss": 0.0538, "step": 24745 }, { "epoch": 2.682784041630529, "grad_norm": 0.02541131153702736, "learning_rate": 1.0612215290874801e-05, "loss": 0.0013, "step": 24746 }, { "epoch": 2.682892454466609, "grad_norm": 0.01113484613597393, "learning_rate": 1.0608588423037864e-05, "loss": 0.0005, "step": 24747 }, { "epoch": 2.6830008673026886, "grad_norm": 0.1340111941099167, "learning_rate": 1.060496155520093e-05, "loss": 0.0039, "step": 24748 }, { "epoch": 2.6831092801387686, "grad_norm": 0.14739485085010529, "learning_rate": 1.0601334687363992e-05, "loss": 0.002, "step": 24749 }, { "epoch": 2.683217692974848, "grad_norm": 0.8956540822982788, "learning_rate": 1.0597707819527058e-05, "loss": 0.0374, "step": 24750 }, { "epoch": 2.683326105810928, "grad_norm": 0.01977674849331379, "learning_rate": 1.059408095169012e-05, "loss": 0.0009, "step": 24751 }, { "epoch": 2.6834345186470077, "grad_norm": 0.017406003549695015, "learning_rate": 1.0590454083853186e-05, "loss": 0.0006, "step": 24752 }, { "epoch": 2.6835429314830876, "grad_norm": 0.33946895599365234, "learning_rate": 1.0586827216016248e-05, "loss": 0.0052, "step": 24753 }, { "epoch": 2.683651344319167, "grad_norm": 0.20420798659324646, "learning_rate": 1.0583200348179312e-05, "loss": 0.0072, "step": 24754 }, { "epoch": 2.683759757155247, "grad_norm": 0.10869387537240982, "learning_rate": 1.0579573480342376e-05, "loss": 0.0034, "step": 24755 }, { "epoch": 2.683868169991327, "grad_norm": 0.03577856719493866, "learning_rate": 1.057594661250544e-05, "loss": 0.0011, "step": 24756 }, { "epoch": 2.6839765828274067, "grad_norm": 0.9685298204421997, "learning_rate": 1.0572319744668505e-05, "loss": 0.0431, "step": 24757 }, { "epoch": 2.6840849956634867, "grad_norm": 0.2170974463224411, "learning_rate": 1.0568692876831569e-05, "loss": 0.0196, "step": 24758 }, { "epoch": 2.6841934084995662, "grad_norm": 0.07075049728155136, "learning_rate": 1.0565066008994633e-05, "loss": 0.0029, "step": 24759 }, { "epoch": 2.684301821335646, "grad_norm": 0.46522945165634155, "learning_rate": 1.0561439141157697e-05, "loss": 0.008, "step": 24760 }, { "epoch": 2.684410234171726, "grad_norm": 0.06776938587427139, "learning_rate": 1.0557812273320761e-05, "loss": 0.0035, "step": 24761 }, { "epoch": 2.6845186470078057, "grad_norm": 0.35423243045806885, "learning_rate": 1.0554185405483824e-05, "loss": 0.0147, "step": 24762 }, { "epoch": 2.6846270598438853, "grad_norm": 0.17058716714382172, "learning_rate": 1.055055853764689e-05, "loss": 0.0195, "step": 24763 }, { "epoch": 2.6847354726799653, "grad_norm": 0.30091315507888794, "learning_rate": 1.0546931669809952e-05, "loss": 0.0076, "step": 24764 }, { "epoch": 2.6848438855160452, "grad_norm": 0.4958970248699188, "learning_rate": 1.0543304801973018e-05, "loss": 0.027, "step": 24765 }, { "epoch": 2.684952298352125, "grad_norm": 0.4775470793247223, "learning_rate": 1.053967793413608e-05, "loss": 0.0162, "step": 24766 }, { "epoch": 2.6850607111882048, "grad_norm": 0.2167055457830429, "learning_rate": 1.0536051066299146e-05, "loss": 0.0116, "step": 24767 }, { "epoch": 2.6851691240242843, "grad_norm": 0.08847537636756897, "learning_rate": 1.0532424198462208e-05, "loss": 0.0028, "step": 24768 }, { "epoch": 2.6852775368603643, "grad_norm": 0.3041974604129791, "learning_rate": 1.0528797330625272e-05, "loss": 0.0136, "step": 24769 }, { "epoch": 2.6853859496964443, "grad_norm": 0.6310842633247375, "learning_rate": 1.0525170462788336e-05, "loss": 0.0585, "step": 24770 }, { "epoch": 2.685494362532524, "grad_norm": 0.11503284424543381, "learning_rate": 1.05215435949514e-05, "loss": 0.0032, "step": 24771 }, { "epoch": 2.6856027753686034, "grad_norm": 0.3123563826084137, "learning_rate": 1.0517916727114465e-05, "loss": 0.0178, "step": 24772 }, { "epoch": 2.6857111882046834, "grad_norm": 0.516790509223938, "learning_rate": 1.0514289859277529e-05, "loss": 0.0123, "step": 24773 }, { "epoch": 2.6858196010407633, "grad_norm": 0.39421674609184265, "learning_rate": 1.0510662991440593e-05, "loss": 0.0096, "step": 24774 }, { "epoch": 2.685928013876843, "grad_norm": 0.12513108551502228, "learning_rate": 1.0507036123603655e-05, "loss": 0.0061, "step": 24775 }, { "epoch": 2.686036426712923, "grad_norm": 0.05499718710780144, "learning_rate": 1.0503409255766721e-05, "loss": 0.0024, "step": 24776 }, { "epoch": 2.6861448395490024, "grad_norm": 0.04504622519016266, "learning_rate": 1.0499782387929783e-05, "loss": 0.0012, "step": 24777 }, { "epoch": 2.6862532523850824, "grad_norm": 0.21739593148231506, "learning_rate": 1.049615552009285e-05, "loss": 0.0141, "step": 24778 }, { "epoch": 2.6863616652211624, "grad_norm": 0.4876972734928131, "learning_rate": 1.0492528652255912e-05, "loss": 0.0121, "step": 24779 }, { "epoch": 2.686470078057242, "grad_norm": 0.1900065541267395, "learning_rate": 1.0488901784418977e-05, "loss": 0.0103, "step": 24780 }, { "epoch": 2.686578490893322, "grad_norm": 0.09665103256702423, "learning_rate": 1.048527491658204e-05, "loss": 0.0024, "step": 24781 }, { "epoch": 2.6866869037294014, "grad_norm": 0.20972509682178497, "learning_rate": 1.0481648048745104e-05, "loss": 0.016, "step": 24782 }, { "epoch": 2.6867953165654814, "grad_norm": 0.6455304622650146, "learning_rate": 1.0478021180908168e-05, "loss": 0.0266, "step": 24783 }, { "epoch": 2.6869037294015614, "grad_norm": 0.23610718548297882, "learning_rate": 1.0474394313071232e-05, "loss": 0.0142, "step": 24784 }, { "epoch": 2.687012142237641, "grad_norm": 0.17168880999088287, "learning_rate": 1.0470767445234296e-05, "loss": 0.0034, "step": 24785 }, { "epoch": 2.6871205550737205, "grad_norm": 0.371316134929657, "learning_rate": 1.046714057739736e-05, "loss": 0.0108, "step": 24786 }, { "epoch": 2.6872289679098005, "grad_norm": 0.21006883680820465, "learning_rate": 1.0463513709560424e-05, "loss": 0.0154, "step": 24787 }, { "epoch": 2.6873373807458805, "grad_norm": 0.01016711350530386, "learning_rate": 1.0459886841723487e-05, "loss": 0.0006, "step": 24788 }, { "epoch": 2.68744579358196, "grad_norm": 0.20600692927837372, "learning_rate": 1.0456259973886553e-05, "loss": 0.0075, "step": 24789 }, { "epoch": 2.68755420641804, "grad_norm": 0.1590295433998108, "learning_rate": 1.0452633106049615e-05, "loss": 0.0022, "step": 24790 }, { "epoch": 2.6876626192541195, "grad_norm": 0.01682285964488983, "learning_rate": 1.0449006238212681e-05, "loss": 0.0003, "step": 24791 }, { "epoch": 2.6877710320901995, "grad_norm": 0.010989517904818058, "learning_rate": 1.0445379370375743e-05, "loss": 0.0004, "step": 24792 }, { "epoch": 2.6878794449262795, "grad_norm": 0.2418571263551712, "learning_rate": 1.0441752502538809e-05, "loss": 0.013, "step": 24793 }, { "epoch": 2.687987857762359, "grad_norm": 0.09002356231212616, "learning_rate": 1.0438125634701871e-05, "loss": 0.0034, "step": 24794 }, { "epoch": 2.6880962705984386, "grad_norm": 0.0666629821062088, "learning_rate": 1.0434498766864936e-05, "loss": 0.0024, "step": 24795 }, { "epoch": 2.6882046834345186, "grad_norm": 0.07365454733371735, "learning_rate": 1.0430871899028e-05, "loss": 0.0016, "step": 24796 }, { "epoch": 2.6883130962705986, "grad_norm": 0.1509372740983963, "learning_rate": 1.0427245031191064e-05, "loss": 0.0032, "step": 24797 }, { "epoch": 2.688421509106678, "grad_norm": 0.03404250741004944, "learning_rate": 1.0423618163354128e-05, "loss": 0.0012, "step": 24798 }, { "epoch": 2.688529921942758, "grad_norm": 0.4575647711753845, "learning_rate": 1.0419991295517192e-05, "loss": 0.0098, "step": 24799 }, { "epoch": 2.6886383347788376, "grad_norm": 0.1578475534915924, "learning_rate": 1.0416364427680256e-05, "loss": 0.0073, "step": 24800 }, { "epoch": 2.6887467476149176, "grad_norm": 0.4309719502925873, "learning_rate": 1.041273755984332e-05, "loss": 0.0052, "step": 24801 }, { "epoch": 2.6888551604509976, "grad_norm": 0.2198134809732437, "learning_rate": 1.0409110692006384e-05, "loss": 0.0191, "step": 24802 }, { "epoch": 2.688963573287077, "grad_norm": 0.3370872139930725, "learning_rate": 1.0405483824169447e-05, "loss": 0.0109, "step": 24803 }, { "epoch": 2.689071986123157, "grad_norm": 0.047555163502693176, "learning_rate": 1.0401856956332513e-05, "loss": 0.0008, "step": 24804 }, { "epoch": 2.6891803989592367, "grad_norm": 0.015899399295449257, "learning_rate": 1.0398230088495575e-05, "loss": 0.001, "step": 24805 }, { "epoch": 2.6892888117953166, "grad_norm": 0.027456097304821014, "learning_rate": 1.039460322065864e-05, "loss": 0.0019, "step": 24806 }, { "epoch": 2.6893972246313966, "grad_norm": 0.33773019909858704, "learning_rate": 1.0390976352821703e-05, "loss": 0.0114, "step": 24807 }, { "epoch": 2.689505637467476, "grad_norm": 0.1829083412885666, "learning_rate": 1.0387349484984767e-05, "loss": 0.01, "step": 24808 }, { "epoch": 2.6896140503035557, "grad_norm": 0.9616237282752991, "learning_rate": 1.0383722617147831e-05, "loss": 0.0245, "step": 24809 }, { "epoch": 2.6897224631396357, "grad_norm": 1.024528980255127, "learning_rate": 1.0380095749310895e-05, "loss": 0.0061, "step": 24810 }, { "epoch": 2.6898308759757157, "grad_norm": 0.17101137340068817, "learning_rate": 1.037646888147396e-05, "loss": 0.0045, "step": 24811 }, { "epoch": 2.6899392888117952, "grad_norm": 0.019047532230615616, "learning_rate": 1.0372842013637024e-05, "loss": 0.0007, "step": 24812 }, { "epoch": 2.690047701647875, "grad_norm": 0.39584091305732727, "learning_rate": 1.0369215145800088e-05, "loss": 0.0101, "step": 24813 }, { "epoch": 2.6901561144839548, "grad_norm": 0.3590252101421356, "learning_rate": 1.0365588277963152e-05, "loss": 0.0121, "step": 24814 }, { "epoch": 2.6902645273200347, "grad_norm": 0.6320313215255737, "learning_rate": 1.0361961410126216e-05, "loss": 0.0061, "step": 24815 }, { "epoch": 2.6903729401561147, "grad_norm": 0.11467999219894409, "learning_rate": 1.0358334542289278e-05, "loss": 0.0033, "step": 24816 }, { "epoch": 2.6904813529921943, "grad_norm": 0.5336560010910034, "learning_rate": 1.0354707674452344e-05, "loss": 0.0097, "step": 24817 }, { "epoch": 2.690589765828274, "grad_norm": 1.3196154832839966, "learning_rate": 1.0351080806615407e-05, "loss": 0.0365, "step": 24818 }, { "epoch": 2.690698178664354, "grad_norm": 0.32722121477127075, "learning_rate": 1.0347453938778472e-05, "loss": 0.021, "step": 24819 }, { "epoch": 2.6908065915004338, "grad_norm": 0.33365631103515625, "learning_rate": 1.0343827070941535e-05, "loss": 0.0119, "step": 24820 }, { "epoch": 2.6909150043365133, "grad_norm": 0.1560981273651123, "learning_rate": 1.0340200203104599e-05, "loss": 0.009, "step": 24821 }, { "epoch": 2.6910234171725933, "grad_norm": 0.018549567088484764, "learning_rate": 1.0336573335267663e-05, "loss": 0.0005, "step": 24822 }, { "epoch": 2.691131830008673, "grad_norm": 0.1175684779882431, "learning_rate": 1.0332946467430727e-05, "loss": 0.002, "step": 24823 }, { "epoch": 2.691240242844753, "grad_norm": 0.5999309420585632, "learning_rate": 1.0329319599593791e-05, "loss": 0.0064, "step": 24824 }, { "epoch": 2.691348655680833, "grad_norm": 0.0741117000579834, "learning_rate": 1.0325692731756855e-05, "loss": 0.0031, "step": 24825 }, { "epoch": 2.6914570685169124, "grad_norm": 0.07018940150737762, "learning_rate": 1.032206586391992e-05, "loss": 0.0014, "step": 24826 }, { "epoch": 2.6915654813529923, "grad_norm": 0.15803003311157227, "learning_rate": 1.0318438996082984e-05, "loss": 0.002, "step": 24827 }, { "epoch": 2.691673894189072, "grad_norm": 0.2852788269519806, "learning_rate": 1.0314812128246048e-05, "loss": 0.0187, "step": 24828 }, { "epoch": 2.691782307025152, "grad_norm": 0.014839651994407177, "learning_rate": 1.031118526040911e-05, "loss": 0.0004, "step": 24829 }, { "epoch": 2.6918907198612314, "grad_norm": 0.06007016822695732, "learning_rate": 1.0307558392572176e-05, "loss": 0.0031, "step": 24830 }, { "epoch": 2.6919991326973114, "grad_norm": 0.3216661214828491, "learning_rate": 1.0303931524735238e-05, "loss": 0.0226, "step": 24831 }, { "epoch": 2.692107545533391, "grad_norm": 0.5154679417610168, "learning_rate": 1.0300304656898304e-05, "loss": 0.0092, "step": 24832 }, { "epoch": 2.692215958369471, "grad_norm": 0.010489855892956257, "learning_rate": 1.0296677789061367e-05, "loss": 0.0004, "step": 24833 }, { "epoch": 2.692324371205551, "grad_norm": 3.4701642990112305, "learning_rate": 1.029305092122443e-05, "loss": 0.0141, "step": 24834 }, { "epoch": 2.6924327840416304, "grad_norm": 0.21209308505058289, "learning_rate": 1.0289424053387495e-05, "loss": 0.0081, "step": 24835 }, { "epoch": 2.6925411968777104, "grad_norm": 0.25912976264953613, "learning_rate": 1.0285797185550559e-05, "loss": 0.0087, "step": 24836 }, { "epoch": 2.69264960971379, "grad_norm": 0.03617394343018532, "learning_rate": 1.0282170317713623e-05, "loss": 0.0013, "step": 24837 }, { "epoch": 2.69275802254987, "grad_norm": 0.810269296169281, "learning_rate": 1.0278543449876687e-05, "loss": 0.055, "step": 24838 }, { "epoch": 2.69286643538595, "grad_norm": 0.13321998715400696, "learning_rate": 1.0274916582039751e-05, "loss": 0.0049, "step": 24839 }, { "epoch": 2.6929748482220295, "grad_norm": 0.4633880853652954, "learning_rate": 1.0271289714202815e-05, "loss": 0.0366, "step": 24840 }, { "epoch": 2.693083261058109, "grad_norm": 0.021174421533942223, "learning_rate": 1.026766284636588e-05, "loss": 0.0007, "step": 24841 }, { "epoch": 2.693191673894189, "grad_norm": 0.06349356472492218, "learning_rate": 1.0264035978528942e-05, "loss": 0.0017, "step": 24842 }, { "epoch": 2.693300086730269, "grad_norm": 0.23859845101833344, "learning_rate": 1.0260409110692008e-05, "loss": 0.0056, "step": 24843 }, { "epoch": 2.6934084995663485, "grad_norm": 0.01499002892524004, "learning_rate": 1.025678224285507e-05, "loss": 0.0007, "step": 24844 }, { "epoch": 2.6935169124024285, "grad_norm": 0.028989216312766075, "learning_rate": 1.0253155375018136e-05, "loss": 0.0013, "step": 24845 }, { "epoch": 2.693625325238508, "grad_norm": 0.6625309586524963, "learning_rate": 1.0249528507181198e-05, "loss": 0.062, "step": 24846 }, { "epoch": 2.693733738074588, "grad_norm": 0.6023298501968384, "learning_rate": 1.0245901639344262e-05, "loss": 0.0139, "step": 24847 }, { "epoch": 2.693842150910668, "grad_norm": 0.6158992052078247, "learning_rate": 1.0242274771507326e-05, "loss": 0.0048, "step": 24848 }, { "epoch": 2.6939505637467476, "grad_norm": 0.07070279866456985, "learning_rate": 1.023864790367039e-05, "loss": 0.0029, "step": 24849 }, { "epoch": 2.6940589765828276, "grad_norm": 0.028075316920876503, "learning_rate": 1.0235021035833455e-05, "loss": 0.001, "step": 24850 }, { "epoch": 2.694167389418907, "grad_norm": 0.14074747264385223, "learning_rate": 1.0231394167996519e-05, "loss": 0.0046, "step": 24851 }, { "epoch": 2.694275802254987, "grad_norm": 0.38755151629447937, "learning_rate": 1.0227767300159583e-05, "loss": 0.0117, "step": 24852 }, { "epoch": 2.6943842150910666, "grad_norm": 0.12160531431436539, "learning_rate": 1.0224140432322647e-05, "loss": 0.0037, "step": 24853 }, { "epoch": 2.6944926279271466, "grad_norm": 0.5171259045600891, "learning_rate": 1.0220513564485711e-05, "loss": 0.019, "step": 24854 }, { "epoch": 2.694601040763226, "grad_norm": 0.630591869354248, "learning_rate": 1.0216886696648775e-05, "loss": 0.0134, "step": 24855 }, { "epoch": 2.694709453599306, "grad_norm": 0.3966572880744934, "learning_rate": 1.021325982881184e-05, "loss": 0.0106, "step": 24856 }, { "epoch": 2.694817866435386, "grad_norm": 0.29761242866516113, "learning_rate": 1.0209632960974902e-05, "loss": 0.0151, "step": 24857 }, { "epoch": 2.6949262792714657, "grad_norm": 0.2780522406101227, "learning_rate": 1.0206006093137967e-05, "loss": 0.0126, "step": 24858 }, { "epoch": 2.6950346921075456, "grad_norm": 0.3703601360321045, "learning_rate": 1.020237922530103e-05, "loss": 0.027, "step": 24859 }, { "epoch": 2.695143104943625, "grad_norm": 0.49838560819625854, "learning_rate": 1.0198752357464094e-05, "loss": 0.029, "step": 24860 }, { "epoch": 2.695251517779705, "grad_norm": 0.09143758565187454, "learning_rate": 1.0195125489627158e-05, "loss": 0.0037, "step": 24861 }, { "epoch": 2.695359930615785, "grad_norm": 0.2729853093624115, "learning_rate": 1.0191498621790222e-05, "loss": 0.0139, "step": 24862 }, { "epoch": 2.6954683434518647, "grad_norm": 0.160059854388237, "learning_rate": 1.0187871753953286e-05, "loss": 0.0079, "step": 24863 }, { "epoch": 2.6955767562879442, "grad_norm": 0.14022697508335114, "learning_rate": 1.018424488611635e-05, "loss": 0.0054, "step": 24864 }, { "epoch": 2.6956851691240242, "grad_norm": 0.025378922000527382, "learning_rate": 1.0180618018279414e-05, "loss": 0.0012, "step": 24865 }, { "epoch": 2.695793581960104, "grad_norm": 1.3638591766357422, "learning_rate": 1.0176991150442479e-05, "loss": 0.0296, "step": 24866 }, { "epoch": 2.6959019947961838, "grad_norm": 0.087345190346241, "learning_rate": 1.0173364282605543e-05, "loss": 0.0054, "step": 24867 }, { "epoch": 2.6960104076322637, "grad_norm": 0.1410331428050995, "learning_rate": 1.0169737414768607e-05, "loss": 0.0053, "step": 24868 }, { "epoch": 2.6961188204683433, "grad_norm": 0.847376823425293, "learning_rate": 1.0166110546931671e-05, "loss": 0.0151, "step": 24869 }, { "epoch": 2.6962272333044233, "grad_norm": 0.3146536350250244, "learning_rate": 1.0162483679094733e-05, "loss": 0.0118, "step": 24870 }, { "epoch": 2.6963356461405033, "grad_norm": 0.07850154489278793, "learning_rate": 1.0158856811257799e-05, "loss": 0.0055, "step": 24871 }, { "epoch": 2.696444058976583, "grad_norm": 0.40820378065109253, "learning_rate": 1.0155229943420862e-05, "loss": 0.0056, "step": 24872 }, { "epoch": 2.6965524718126628, "grad_norm": 0.2315705120563507, "learning_rate": 1.0151603075583926e-05, "loss": 0.0123, "step": 24873 }, { "epoch": 2.6966608846487423, "grad_norm": 0.30752867460250854, "learning_rate": 1.014797620774699e-05, "loss": 0.021, "step": 24874 }, { "epoch": 2.6967692974848223, "grad_norm": 0.1572759747505188, "learning_rate": 1.0144349339910054e-05, "loss": 0.0025, "step": 24875 }, { "epoch": 2.696877710320902, "grad_norm": 0.06129913032054901, "learning_rate": 1.0140722472073118e-05, "loss": 0.0017, "step": 24876 }, { "epoch": 2.696986123156982, "grad_norm": 0.21109552681446075, "learning_rate": 1.0137095604236182e-05, "loss": 0.0116, "step": 24877 }, { "epoch": 2.6970945359930614, "grad_norm": 0.28557705879211426, "learning_rate": 1.0133468736399246e-05, "loss": 0.0294, "step": 24878 }, { "epoch": 2.6972029488291414, "grad_norm": 0.2702373266220093, "learning_rate": 1.012984186856231e-05, "loss": 0.0086, "step": 24879 }, { "epoch": 2.6973113616652213, "grad_norm": 0.10995937883853912, "learning_rate": 1.0126215000725374e-05, "loss": 0.0018, "step": 24880 }, { "epoch": 2.697419774501301, "grad_norm": 0.16789980232715607, "learning_rate": 1.0122588132888438e-05, "loss": 0.0095, "step": 24881 }, { "epoch": 2.697528187337381, "grad_norm": 0.07958489656448364, "learning_rate": 1.0118961265051503e-05, "loss": 0.0026, "step": 24882 }, { "epoch": 2.6976366001734604, "grad_norm": 0.0487847700715065, "learning_rate": 1.0115334397214565e-05, "loss": 0.002, "step": 24883 }, { "epoch": 2.6977450130095404, "grad_norm": 0.04400648921728134, "learning_rate": 1.011170752937763e-05, "loss": 0.001, "step": 24884 }, { "epoch": 2.6978534258456204, "grad_norm": 1.0073535442352295, "learning_rate": 1.0108080661540693e-05, "loss": 0.0158, "step": 24885 }, { "epoch": 2.6979618386817, "grad_norm": 0.12215782701969147, "learning_rate": 1.0104453793703757e-05, "loss": 0.0049, "step": 24886 }, { "epoch": 2.6980702515177795, "grad_norm": 0.04256097227334976, "learning_rate": 1.0100826925866821e-05, "loss": 0.0012, "step": 24887 }, { "epoch": 2.6981786643538594, "grad_norm": 0.24371378123760223, "learning_rate": 1.0097200058029886e-05, "loss": 0.0029, "step": 24888 }, { "epoch": 2.6982870771899394, "grad_norm": 0.5747612714767456, "learning_rate": 1.009357319019295e-05, "loss": 0.0171, "step": 24889 }, { "epoch": 2.698395490026019, "grad_norm": 0.11091244220733643, "learning_rate": 1.0089946322356014e-05, "loss": 0.0033, "step": 24890 }, { "epoch": 2.698503902862099, "grad_norm": 0.3963001072406769, "learning_rate": 1.0086319454519078e-05, "loss": 0.0144, "step": 24891 }, { "epoch": 2.6986123156981785, "grad_norm": 0.024063840508461, "learning_rate": 1.0082692586682142e-05, "loss": 0.0005, "step": 24892 }, { "epoch": 2.6987207285342585, "grad_norm": 0.1786816567182541, "learning_rate": 1.0079065718845206e-05, "loss": 0.0056, "step": 24893 }, { "epoch": 2.6988291413703385, "grad_norm": 0.23207537829875946, "learning_rate": 1.007543885100827e-05, "loss": 0.0128, "step": 24894 }, { "epoch": 2.698937554206418, "grad_norm": 0.2699648439884186, "learning_rate": 1.0071811983171334e-05, "loss": 0.0549, "step": 24895 }, { "epoch": 2.699045967042498, "grad_norm": 0.09888678044080734, "learning_rate": 1.0068185115334397e-05, "loss": 0.0035, "step": 24896 }, { "epoch": 2.6991543798785775, "grad_norm": 0.1485946625471115, "learning_rate": 1.0064558247497462e-05, "loss": 0.0031, "step": 24897 }, { "epoch": 2.6992627927146575, "grad_norm": 0.11399861425161362, "learning_rate": 1.0060931379660525e-05, "loss": 0.0039, "step": 24898 }, { "epoch": 2.699371205550737, "grad_norm": 0.5346825122833252, "learning_rate": 1.0057304511823589e-05, "loss": 0.0507, "step": 24899 }, { "epoch": 2.699479618386817, "grad_norm": 0.004743837285786867, "learning_rate": 1.0053677643986653e-05, "loss": 0.0001, "step": 24900 }, { "epoch": 2.6995880312228966, "grad_norm": 1.153895616531372, "learning_rate": 1.0050050776149717e-05, "loss": 0.0387, "step": 24901 }, { "epoch": 2.6996964440589766, "grad_norm": 0.262442022562027, "learning_rate": 1.0046423908312781e-05, "loss": 0.0048, "step": 24902 }, { "epoch": 2.6998048568950566, "grad_norm": 0.0766642689704895, "learning_rate": 1.0042797040475845e-05, "loss": 0.0027, "step": 24903 }, { "epoch": 2.699913269731136, "grad_norm": 0.07901016622781754, "learning_rate": 1.003917017263891e-05, "loss": 0.0055, "step": 24904 }, { "epoch": 2.700021682567216, "grad_norm": 0.03090631775557995, "learning_rate": 1.0035543304801974e-05, "loss": 0.0012, "step": 24905 }, { "epoch": 2.7001300954032956, "grad_norm": 0.140573650598526, "learning_rate": 1.0031916436965038e-05, "loss": 0.0079, "step": 24906 }, { "epoch": 2.7002385082393756, "grad_norm": 0.4316219687461853, "learning_rate": 1.0028289569128102e-05, "loss": 0.0139, "step": 24907 }, { "epoch": 2.7003469210754556, "grad_norm": 0.14923930168151855, "learning_rate": 1.0024662701291166e-05, "loss": 0.0055, "step": 24908 }, { "epoch": 2.700455333911535, "grad_norm": 0.3723445236682892, "learning_rate": 1.002103583345423e-05, "loss": 0.0398, "step": 24909 }, { "epoch": 2.7005637467476147, "grad_norm": 0.2385961413383484, "learning_rate": 1.0017408965617294e-05, "loss": 0.0082, "step": 24910 }, { "epoch": 2.7006721595836947, "grad_norm": 0.38944369554519653, "learning_rate": 1.0013782097780357e-05, "loss": 0.0112, "step": 24911 }, { "epoch": 2.7007805724197746, "grad_norm": 0.4143291115760803, "learning_rate": 1.001015522994342e-05, "loss": 0.0203, "step": 24912 }, { "epoch": 2.700888985255854, "grad_norm": 0.033474378287792206, "learning_rate": 1.0006528362106485e-05, "loss": 0.001, "step": 24913 }, { "epoch": 2.700997398091934, "grad_norm": 0.29688823223114014, "learning_rate": 1.0002901494269549e-05, "loss": 0.013, "step": 24914 }, { "epoch": 2.7011058109280137, "grad_norm": 0.1112644299864769, "learning_rate": 9.999274626432613e-06, "loss": 0.0023, "step": 24915 }, { "epoch": 2.7012142237640937, "grad_norm": 0.01278412714600563, "learning_rate": 9.995647758595677e-06, "loss": 0.0005, "step": 24916 }, { "epoch": 2.7013226366001737, "grad_norm": 0.0925745815038681, "learning_rate": 9.992020890758741e-06, "loss": 0.0047, "step": 24917 }, { "epoch": 2.7014310494362532, "grad_norm": 0.0385882630944252, "learning_rate": 9.988394022921805e-06, "loss": 0.0017, "step": 24918 }, { "epoch": 2.7015394622723328, "grad_norm": 0.37702837586402893, "learning_rate": 9.98476715508487e-06, "loss": 0.0096, "step": 24919 }, { "epoch": 2.7016478751084128, "grad_norm": 0.11779254674911499, "learning_rate": 9.981140287247933e-06, "loss": 0.0086, "step": 24920 }, { "epoch": 2.7017562879444927, "grad_norm": 0.06473109871149063, "learning_rate": 9.977513419410998e-06, "loss": 0.0018, "step": 24921 }, { "epoch": 2.7018647007805723, "grad_norm": 0.12538956105709076, "learning_rate": 9.973886551574062e-06, "loss": 0.0021, "step": 24922 }, { "epoch": 2.7019731136166523, "grad_norm": 0.007300298660993576, "learning_rate": 9.970259683737126e-06, "loss": 0.0002, "step": 24923 }, { "epoch": 2.702081526452732, "grad_norm": 0.14254333078861237, "learning_rate": 9.966632815900188e-06, "loss": 0.0036, "step": 24924 }, { "epoch": 2.702189939288812, "grad_norm": 0.36667606234550476, "learning_rate": 9.963005948063252e-06, "loss": 0.0217, "step": 24925 }, { "epoch": 2.702298352124892, "grad_norm": 0.017828689888119698, "learning_rate": 9.959379080226316e-06, "loss": 0.0003, "step": 24926 }, { "epoch": 2.7024067649609713, "grad_norm": 0.1054994985461235, "learning_rate": 9.95575221238938e-06, "loss": 0.0058, "step": 24927 }, { "epoch": 2.7025151777970513, "grad_norm": 0.01271122507750988, "learning_rate": 9.952125344552445e-06, "loss": 0.0005, "step": 24928 }, { "epoch": 2.702623590633131, "grad_norm": 0.048549361526966095, "learning_rate": 9.948498476715509e-06, "loss": 0.0013, "step": 24929 }, { "epoch": 2.702732003469211, "grad_norm": 0.6712545156478882, "learning_rate": 9.944871608878573e-06, "loss": 0.0202, "step": 24930 }, { "epoch": 2.702840416305291, "grad_norm": 0.30991843342781067, "learning_rate": 9.941244741041637e-06, "loss": 0.0206, "step": 24931 }, { "epoch": 2.7029488291413704, "grad_norm": 0.01692647486925125, "learning_rate": 9.937617873204701e-06, "loss": 0.0008, "step": 24932 }, { "epoch": 2.70305724197745, "grad_norm": 0.04271753877401352, "learning_rate": 9.933991005367765e-06, "loss": 0.0045, "step": 24933 }, { "epoch": 2.70316565481353, "grad_norm": 0.6781840920448303, "learning_rate": 9.93036413753083e-06, "loss": 0.0184, "step": 24934 }, { "epoch": 2.70327406764961, "grad_norm": 0.13484667241573334, "learning_rate": 9.926737269693893e-06, "loss": 0.0049, "step": 24935 }, { "epoch": 2.7033824804856894, "grad_norm": 0.07668924331665039, "learning_rate": 9.923110401856957e-06, "loss": 0.0016, "step": 24936 }, { "epoch": 2.7034908933217694, "grad_norm": 0.1868005245923996, "learning_rate": 9.91948353402002e-06, "loss": 0.0032, "step": 24937 }, { "epoch": 2.703599306157849, "grad_norm": 0.0009667320409789681, "learning_rate": 9.915856666183086e-06, "loss": 0.0, "step": 24938 }, { "epoch": 2.703707718993929, "grad_norm": 0.44900476932525635, "learning_rate": 9.912229798346148e-06, "loss": 0.0058, "step": 24939 }, { "epoch": 2.703816131830009, "grad_norm": 0.04508165270090103, "learning_rate": 9.908602930509212e-06, "loss": 0.0016, "step": 24940 }, { "epoch": 2.7039245446660884, "grad_norm": 0.5388070344924927, "learning_rate": 9.904976062672276e-06, "loss": 0.0334, "step": 24941 }, { "epoch": 2.704032957502168, "grad_norm": 0.2907605767250061, "learning_rate": 9.90134919483534e-06, "loss": 0.0039, "step": 24942 }, { "epoch": 2.704141370338248, "grad_norm": 0.18687498569488525, "learning_rate": 9.897722326998405e-06, "loss": 0.0022, "step": 24943 }, { "epoch": 2.704249783174328, "grad_norm": 0.012353921309113503, "learning_rate": 9.894095459161469e-06, "loss": 0.0004, "step": 24944 }, { "epoch": 2.7043581960104075, "grad_norm": 0.13275399804115295, "learning_rate": 9.890468591324533e-06, "loss": 0.0054, "step": 24945 }, { "epoch": 2.7044666088464875, "grad_norm": 0.18392375111579895, "learning_rate": 9.886841723487597e-06, "loss": 0.0107, "step": 24946 }, { "epoch": 2.704575021682567, "grad_norm": 0.1088273748755455, "learning_rate": 9.883214855650661e-06, "loss": 0.0035, "step": 24947 }, { "epoch": 2.704683434518647, "grad_norm": 0.07135391235351562, "learning_rate": 9.879587987813725e-06, "loss": 0.0012, "step": 24948 }, { "epoch": 2.704791847354727, "grad_norm": 0.5726509690284729, "learning_rate": 9.875961119976789e-06, "loss": 0.0197, "step": 24949 }, { "epoch": 2.7049002601908065, "grad_norm": 0.42905259132385254, "learning_rate": 9.872334252139853e-06, "loss": 0.0156, "step": 24950 }, { "epoch": 2.7050086730268865, "grad_norm": 0.016757972538471222, "learning_rate": 9.868707384302917e-06, "loss": 0.0003, "step": 24951 }, { "epoch": 2.705117085862966, "grad_norm": 0.2484089881181717, "learning_rate": 9.86508051646598e-06, "loss": 0.0045, "step": 24952 }, { "epoch": 2.705225498699046, "grad_norm": 0.1771782785654068, "learning_rate": 9.861453648629044e-06, "loss": 0.0107, "step": 24953 }, { "epoch": 2.705333911535126, "grad_norm": 0.16229505836963654, "learning_rate": 9.857826780792108e-06, "loss": 0.0043, "step": 24954 }, { "epoch": 2.7054423243712056, "grad_norm": 0.057302240282297134, "learning_rate": 9.854199912955172e-06, "loss": 0.0018, "step": 24955 }, { "epoch": 2.705550737207285, "grad_norm": 0.38796889781951904, "learning_rate": 9.850573045118236e-06, "loss": 0.0095, "step": 24956 }, { "epoch": 2.705659150043365, "grad_norm": 0.18048280477523804, "learning_rate": 9.8469461772813e-06, "loss": 0.0093, "step": 24957 }, { "epoch": 2.705767562879445, "grad_norm": 0.2226238250732422, "learning_rate": 9.843319309444364e-06, "loss": 0.0067, "step": 24958 }, { "epoch": 2.7058759757155246, "grad_norm": 0.06350935250520706, "learning_rate": 9.839692441607429e-06, "loss": 0.0021, "step": 24959 }, { "epoch": 2.7059843885516046, "grad_norm": 1.1544744968414307, "learning_rate": 9.836065573770493e-06, "loss": 0.035, "step": 24960 }, { "epoch": 2.706092801387684, "grad_norm": 0.5754572749137878, "learning_rate": 9.832438705933557e-06, "loss": 0.0313, "step": 24961 }, { "epoch": 2.706201214223764, "grad_norm": 0.3272056579589844, "learning_rate": 9.82881183809662e-06, "loss": 0.0214, "step": 24962 }, { "epoch": 2.706309627059844, "grad_norm": 0.23944658041000366, "learning_rate": 9.825184970259685e-06, "loss": 0.0062, "step": 24963 }, { "epoch": 2.7064180398959237, "grad_norm": 0.03728903457522392, "learning_rate": 9.821558102422749e-06, "loss": 0.0008, "step": 24964 }, { "epoch": 2.706526452732003, "grad_norm": 0.07134287059307098, "learning_rate": 9.817931234585811e-06, "loss": 0.0023, "step": 24965 }, { "epoch": 2.706634865568083, "grad_norm": 0.7877167463302612, "learning_rate": 9.814304366748876e-06, "loss": 0.0402, "step": 24966 }, { "epoch": 2.706743278404163, "grad_norm": 0.2015967071056366, "learning_rate": 9.81067749891194e-06, "loss": 0.0049, "step": 24967 }, { "epoch": 2.7068516912402427, "grad_norm": 0.40893736481666565, "learning_rate": 9.807050631075004e-06, "loss": 0.0206, "step": 24968 }, { "epoch": 2.7069601040763227, "grad_norm": 0.08341654390096664, "learning_rate": 9.803423763238068e-06, "loss": 0.0019, "step": 24969 }, { "epoch": 2.7070685169124022, "grad_norm": 0.9742446541786194, "learning_rate": 9.799796895401132e-06, "loss": 0.0217, "step": 24970 }, { "epoch": 2.7071769297484822, "grad_norm": 0.20496243238449097, "learning_rate": 9.796170027564196e-06, "loss": 0.0197, "step": 24971 }, { "epoch": 2.707285342584562, "grad_norm": 0.37965109944343567, "learning_rate": 9.79254315972726e-06, "loss": 0.0097, "step": 24972 }, { "epoch": 2.7073937554206418, "grad_norm": 0.05139191076159477, "learning_rate": 9.788916291890324e-06, "loss": 0.0011, "step": 24973 }, { "epoch": 2.7075021682567217, "grad_norm": 0.0017826971597969532, "learning_rate": 9.785289424053388e-06, "loss": 0.0001, "step": 24974 }, { "epoch": 2.7076105810928013, "grad_norm": 0.1922415941953659, "learning_rate": 9.781662556216452e-06, "loss": 0.0239, "step": 24975 }, { "epoch": 2.7077189939288813, "grad_norm": 0.02574317157268524, "learning_rate": 9.778035688379517e-06, "loss": 0.001, "step": 24976 }, { "epoch": 2.707827406764961, "grad_norm": 0.2697688937187195, "learning_rate": 9.77440882054258e-06, "loss": 0.0101, "step": 24977 }, { "epoch": 2.707935819601041, "grad_norm": 0.23103450238704681, "learning_rate": 9.770781952705643e-06, "loss": 0.0064, "step": 24978 }, { "epoch": 2.7080442324371203, "grad_norm": 0.009638706222176552, "learning_rate": 9.767155084868707e-06, "loss": 0.0005, "step": 24979 }, { "epoch": 2.7081526452732003, "grad_norm": 0.4425007700920105, "learning_rate": 9.763528217031771e-06, "loss": 0.0115, "step": 24980 }, { "epoch": 2.7082610581092803, "grad_norm": 0.04181494936347008, "learning_rate": 9.759901349194835e-06, "loss": 0.0009, "step": 24981 }, { "epoch": 2.70836947094536, "grad_norm": 0.6896070837974548, "learning_rate": 9.7562744813579e-06, "loss": 0.0087, "step": 24982 }, { "epoch": 2.70847788378144, "grad_norm": 0.13016951084136963, "learning_rate": 9.752647613520964e-06, "loss": 0.0055, "step": 24983 }, { "epoch": 2.7085862966175194, "grad_norm": 0.035989079624414444, "learning_rate": 9.749020745684028e-06, "loss": 0.0014, "step": 24984 }, { "epoch": 2.7086947094535994, "grad_norm": 0.8060900568962097, "learning_rate": 9.745393877847092e-06, "loss": 0.0111, "step": 24985 }, { "epoch": 2.7088031222896793, "grad_norm": 0.20604270696640015, "learning_rate": 9.741767010010156e-06, "loss": 0.0134, "step": 24986 }, { "epoch": 2.708911535125759, "grad_norm": 0.4906022548675537, "learning_rate": 9.73814014217322e-06, "loss": 0.0214, "step": 24987 }, { "epoch": 2.7090199479618384, "grad_norm": 0.5470768213272095, "learning_rate": 9.734513274336284e-06, "loss": 0.008, "step": 24988 }, { "epoch": 2.7091283607979184, "grad_norm": 0.19553345441818237, "learning_rate": 9.730886406499348e-06, "loss": 0.0058, "step": 24989 }, { "epoch": 2.7092367736339984, "grad_norm": 0.03749927505850792, "learning_rate": 9.727259538662412e-06, "loss": 0.0011, "step": 24990 }, { "epoch": 2.709345186470078, "grad_norm": 0.11967775970697403, "learning_rate": 9.723632670825475e-06, "loss": 0.0027, "step": 24991 }, { "epoch": 2.709453599306158, "grad_norm": 0.7431172132492065, "learning_rate": 9.720005802988539e-06, "loss": 0.0611, "step": 24992 }, { "epoch": 2.7095620121422375, "grad_norm": 0.040352605283260345, "learning_rate": 9.716378935151603e-06, "loss": 0.0009, "step": 24993 }, { "epoch": 2.7096704249783174, "grad_norm": 0.1910463571548462, "learning_rate": 9.712752067314667e-06, "loss": 0.0051, "step": 24994 }, { "epoch": 2.7097788378143974, "grad_norm": 0.02162211947143078, "learning_rate": 9.709125199477731e-06, "loss": 0.0009, "step": 24995 }, { "epoch": 2.709887250650477, "grad_norm": 0.26905593276023865, "learning_rate": 9.705498331640795e-06, "loss": 0.0209, "step": 24996 }, { "epoch": 2.709995663486557, "grad_norm": 0.15871933102607727, "learning_rate": 9.70187146380386e-06, "loss": 0.0078, "step": 24997 }, { "epoch": 2.7101040763226365, "grad_norm": 0.41247642040252686, "learning_rate": 9.698244595966924e-06, "loss": 0.0102, "step": 24998 }, { "epoch": 2.7102124891587165, "grad_norm": 0.244564026594162, "learning_rate": 9.694617728129988e-06, "loss": 0.0058, "step": 24999 }, { "epoch": 2.710320901994796, "grad_norm": 0.1463056206703186, "learning_rate": 9.690990860293052e-06, "loss": 0.0033, "step": 25000 }, { "epoch": 2.710429314830876, "grad_norm": 0.8308271765708923, "learning_rate": 9.687363992456116e-06, "loss": 0.0259, "step": 25001 }, { "epoch": 2.7105377276669556, "grad_norm": 0.15213122963905334, "learning_rate": 9.68373712461918e-06, "loss": 0.0024, "step": 25002 }, { "epoch": 2.7106461405030355, "grad_norm": 0.1313008964061737, "learning_rate": 9.680110256782244e-06, "loss": 0.0059, "step": 25003 }, { "epoch": 2.7107545533391155, "grad_norm": 0.3649630844593048, "learning_rate": 9.676483388945308e-06, "loss": 0.0166, "step": 25004 }, { "epoch": 2.710862966175195, "grad_norm": 0.5197329521179199, "learning_rate": 9.67285652110837e-06, "loss": 0.0263, "step": 25005 }, { "epoch": 2.710971379011275, "grad_norm": 0.027596157044172287, "learning_rate": 9.669229653271435e-06, "loss": 0.0014, "step": 25006 }, { "epoch": 2.7110797918473546, "grad_norm": 0.2598634660243988, "learning_rate": 9.665602785434499e-06, "loss": 0.0173, "step": 25007 }, { "epoch": 2.7111882046834346, "grad_norm": 0.2334776669740677, "learning_rate": 9.661975917597563e-06, "loss": 0.0044, "step": 25008 }, { "epoch": 2.7112966175195146, "grad_norm": 0.1987866759300232, "learning_rate": 9.658349049760627e-06, "loss": 0.0087, "step": 25009 }, { "epoch": 2.711405030355594, "grad_norm": 0.6797810196876526, "learning_rate": 9.654722181923691e-06, "loss": 0.0328, "step": 25010 }, { "epoch": 2.7115134431916736, "grad_norm": 0.16635873913764954, "learning_rate": 9.651095314086755e-06, "loss": 0.0123, "step": 25011 }, { "epoch": 2.7116218560277536, "grad_norm": 0.025115367025136948, "learning_rate": 9.64746844624982e-06, "loss": 0.0008, "step": 25012 }, { "epoch": 2.7117302688638336, "grad_norm": 0.43432125449180603, "learning_rate": 9.643841578412883e-06, "loss": 0.0107, "step": 25013 }, { "epoch": 2.711838681699913, "grad_norm": 0.09094492346048355, "learning_rate": 9.640214710575948e-06, "loss": 0.002, "step": 25014 }, { "epoch": 2.711947094535993, "grad_norm": 0.46727579832077026, "learning_rate": 9.636587842739012e-06, "loss": 0.0339, "step": 25015 }, { "epoch": 2.7120555073720727, "grad_norm": 0.683101236820221, "learning_rate": 9.632960974902076e-06, "loss": 0.0163, "step": 25016 }, { "epoch": 2.7121639202081527, "grad_norm": 0.011679991148412228, "learning_rate": 9.62933410706514e-06, "loss": 0.0004, "step": 25017 }, { "epoch": 2.7122723330442327, "grad_norm": 0.5868610739707947, "learning_rate": 9.625707239228202e-06, "loss": 0.0086, "step": 25018 }, { "epoch": 2.712380745880312, "grad_norm": 0.5823597311973572, "learning_rate": 9.622080371391266e-06, "loss": 0.0256, "step": 25019 }, { "epoch": 2.712489158716392, "grad_norm": 0.37724706530570984, "learning_rate": 9.61845350355433e-06, "loss": 0.0131, "step": 25020 }, { "epoch": 2.7125975715524717, "grad_norm": 0.133225217461586, "learning_rate": 9.614826635717395e-06, "loss": 0.0051, "step": 25021 }, { "epoch": 2.7127059843885517, "grad_norm": 0.7130805253982544, "learning_rate": 9.611199767880459e-06, "loss": 0.0287, "step": 25022 }, { "epoch": 2.7128143972246312, "grad_norm": 0.3205452859401703, "learning_rate": 9.607572900043523e-06, "loss": 0.0074, "step": 25023 }, { "epoch": 2.7129228100607112, "grad_norm": 0.5155733227729797, "learning_rate": 9.603946032206587e-06, "loss": 0.0429, "step": 25024 }, { "epoch": 2.7130312228967908, "grad_norm": 0.21417170763015747, "learning_rate": 9.600319164369651e-06, "loss": 0.0091, "step": 25025 }, { "epoch": 2.7131396357328708, "grad_norm": 0.38647082448005676, "learning_rate": 9.596692296532715e-06, "loss": 0.0218, "step": 25026 }, { "epoch": 2.7132480485689507, "grad_norm": 0.07295946031808853, "learning_rate": 9.59306542869578e-06, "loss": 0.0022, "step": 25027 }, { "epoch": 2.7133564614050303, "grad_norm": 0.34313368797302246, "learning_rate": 9.589438560858843e-06, "loss": 0.0083, "step": 25028 }, { "epoch": 2.7134648742411103, "grad_norm": 0.022946268320083618, "learning_rate": 9.585811693021907e-06, "loss": 0.001, "step": 25029 }, { "epoch": 2.71357328707719, "grad_norm": 0.24769577383995056, "learning_rate": 9.582184825184971e-06, "loss": 0.004, "step": 25030 }, { "epoch": 2.71368169991327, "grad_norm": 0.43131136894226074, "learning_rate": 9.578557957348034e-06, "loss": 0.0185, "step": 25031 }, { "epoch": 2.71379011274935, "grad_norm": 0.04112663492560387, "learning_rate": 9.574931089511098e-06, "loss": 0.0008, "step": 25032 }, { "epoch": 2.7138985255854293, "grad_norm": 0.027974288910627365, "learning_rate": 9.571304221674162e-06, "loss": 0.0012, "step": 25033 }, { "epoch": 2.714006938421509, "grad_norm": 0.1835310012102127, "learning_rate": 9.567677353837226e-06, "loss": 0.0234, "step": 25034 }, { "epoch": 2.714115351257589, "grad_norm": 0.308075487613678, "learning_rate": 9.56405048600029e-06, "loss": 0.0074, "step": 25035 }, { "epoch": 2.714223764093669, "grad_norm": 0.09098398685455322, "learning_rate": 9.560423618163354e-06, "loss": 0.0035, "step": 25036 }, { "epoch": 2.7143321769297484, "grad_norm": 0.1154792532324791, "learning_rate": 9.556796750326419e-06, "loss": 0.0031, "step": 25037 }, { "epoch": 2.7144405897658284, "grad_norm": 0.3888178765773773, "learning_rate": 9.553169882489483e-06, "loss": 0.0081, "step": 25038 }, { "epoch": 2.714549002601908, "grad_norm": 0.16371624171733856, "learning_rate": 9.549543014652547e-06, "loss": 0.0084, "step": 25039 }, { "epoch": 2.714657415437988, "grad_norm": 0.4084357023239136, "learning_rate": 9.545916146815611e-06, "loss": 0.0461, "step": 25040 }, { "epoch": 2.714765828274068, "grad_norm": 0.27106180787086487, "learning_rate": 9.542289278978675e-06, "loss": 0.0084, "step": 25041 }, { "epoch": 2.7148742411101474, "grad_norm": 0.22111085057258606, "learning_rate": 9.538662411141739e-06, "loss": 0.0058, "step": 25042 }, { "epoch": 2.7149826539462274, "grad_norm": 0.24346813559532166, "learning_rate": 9.535035543304803e-06, "loss": 0.0163, "step": 25043 }, { "epoch": 2.715091066782307, "grad_norm": 0.07251361757516861, "learning_rate": 9.531408675467866e-06, "loss": 0.0027, "step": 25044 }, { "epoch": 2.715199479618387, "grad_norm": 0.12757064402103424, "learning_rate": 9.52778180763093e-06, "loss": 0.0043, "step": 25045 }, { "epoch": 2.7153078924544665, "grad_norm": 0.2866775393486023, "learning_rate": 9.524154939793994e-06, "loss": 0.0118, "step": 25046 }, { "epoch": 2.7154163052905465, "grad_norm": 0.21507559716701508, "learning_rate": 9.520528071957058e-06, "loss": 0.0022, "step": 25047 }, { "epoch": 2.715524718126626, "grad_norm": 0.22831736505031586, "learning_rate": 9.516901204120122e-06, "loss": 0.0157, "step": 25048 }, { "epoch": 2.715633130962706, "grad_norm": 0.03980375826358795, "learning_rate": 9.513274336283186e-06, "loss": 0.0019, "step": 25049 }, { "epoch": 2.715741543798786, "grad_norm": 0.19699004292488098, "learning_rate": 9.50964746844625e-06, "loss": 0.0062, "step": 25050 }, { "epoch": 2.7158499566348655, "grad_norm": 0.3141443431377411, "learning_rate": 9.506020600609314e-06, "loss": 0.0294, "step": 25051 }, { "epoch": 2.7159583694709455, "grad_norm": 0.34536856412887573, "learning_rate": 9.502393732772378e-06, "loss": 0.0076, "step": 25052 }, { "epoch": 2.716066782307025, "grad_norm": 0.46378692984580994, "learning_rate": 9.498766864935443e-06, "loss": 0.0122, "step": 25053 }, { "epoch": 2.716175195143105, "grad_norm": 0.22706983983516693, "learning_rate": 9.495139997098507e-06, "loss": 0.0073, "step": 25054 }, { "epoch": 2.716283607979185, "grad_norm": 0.6221592426300049, "learning_rate": 9.49151312926157e-06, "loss": 0.0244, "step": 25055 }, { "epoch": 2.7163920208152645, "grad_norm": 0.6780825257301331, "learning_rate": 9.487886261424635e-06, "loss": 0.0221, "step": 25056 }, { "epoch": 2.716500433651344, "grad_norm": 0.0704621821641922, "learning_rate": 9.484259393587697e-06, "loss": 0.0014, "step": 25057 }, { "epoch": 2.716608846487424, "grad_norm": 0.0754556879401207, "learning_rate": 9.480632525750763e-06, "loss": 0.0045, "step": 25058 }, { "epoch": 2.716717259323504, "grad_norm": 0.14394406974315643, "learning_rate": 9.477005657913825e-06, "loss": 0.002, "step": 25059 }, { "epoch": 2.7168256721595836, "grad_norm": 0.1411111205816269, "learning_rate": 9.47337879007689e-06, "loss": 0.0144, "step": 25060 }, { "epoch": 2.7169340849956636, "grad_norm": 0.1679987609386444, "learning_rate": 9.469751922239954e-06, "loss": 0.01, "step": 25061 }, { "epoch": 2.717042497831743, "grad_norm": 0.17306271195411682, "learning_rate": 9.466125054403018e-06, "loss": 0.004, "step": 25062 }, { "epoch": 2.717150910667823, "grad_norm": 0.36850494146347046, "learning_rate": 9.462498186566082e-06, "loss": 0.0198, "step": 25063 }, { "epoch": 2.717259323503903, "grad_norm": 0.046791333705186844, "learning_rate": 9.458871318729146e-06, "loss": 0.0014, "step": 25064 }, { "epoch": 2.7173677363399826, "grad_norm": 0.09178820252418518, "learning_rate": 9.45524445089221e-06, "loss": 0.0022, "step": 25065 }, { "epoch": 2.717476149176062, "grad_norm": 1.2722738981246948, "learning_rate": 9.451617583055274e-06, "loss": 0.0357, "step": 25066 }, { "epoch": 2.717584562012142, "grad_norm": 0.7176387906074524, "learning_rate": 9.447990715218338e-06, "loss": 0.0196, "step": 25067 }, { "epoch": 2.717692974848222, "grad_norm": 0.08232401311397552, "learning_rate": 9.444363847381402e-06, "loss": 0.0045, "step": 25068 }, { "epoch": 2.7178013876843017, "grad_norm": 0.15130743384361267, "learning_rate": 9.440736979544467e-06, "loss": 0.0039, "step": 25069 }, { "epoch": 2.7179098005203817, "grad_norm": 0.01253002230077982, "learning_rate": 9.437110111707529e-06, "loss": 0.0006, "step": 25070 }, { "epoch": 2.718018213356461, "grad_norm": 0.4448099434375763, "learning_rate": 9.433483243870595e-06, "loss": 0.0078, "step": 25071 }, { "epoch": 2.718126626192541, "grad_norm": 0.19906431436538696, "learning_rate": 9.429856376033657e-06, "loss": 0.0044, "step": 25072 }, { "epoch": 2.718235039028621, "grad_norm": 0.341501384973526, "learning_rate": 9.426229508196721e-06, "loss": 0.0138, "step": 25073 }, { "epoch": 2.7183434518647007, "grad_norm": 0.04039527475833893, "learning_rate": 9.422602640359785e-06, "loss": 0.0012, "step": 25074 }, { "epoch": 2.7184518647007807, "grad_norm": 0.3119129240512848, "learning_rate": 9.41897577252285e-06, "loss": 0.0193, "step": 25075 }, { "epoch": 2.7185602775368602, "grad_norm": 0.05528373643755913, "learning_rate": 9.415348904685914e-06, "loss": 0.001, "step": 25076 }, { "epoch": 2.7186686903729402, "grad_norm": 0.05651574581861496, "learning_rate": 9.411722036848978e-06, "loss": 0.0016, "step": 25077 }, { "epoch": 2.71877710320902, "grad_norm": 0.14086079597473145, "learning_rate": 9.408095169012042e-06, "loss": 0.005, "step": 25078 }, { "epoch": 2.7188855160450998, "grad_norm": 0.12488584220409393, "learning_rate": 9.404468301175106e-06, "loss": 0.0047, "step": 25079 }, { "epoch": 2.7189939288811793, "grad_norm": 0.6505849957466125, "learning_rate": 9.40084143333817e-06, "loss": 0.0181, "step": 25080 }, { "epoch": 2.7191023417172593, "grad_norm": 0.5855522751808167, "learning_rate": 9.397214565501234e-06, "loss": 0.0146, "step": 25081 }, { "epoch": 2.7192107545533393, "grad_norm": 0.012380960397422314, "learning_rate": 9.393587697664298e-06, "loss": 0.0001, "step": 25082 }, { "epoch": 2.719319167389419, "grad_norm": 0.5181304812431335, "learning_rate": 9.38996082982736e-06, "loss": 0.0183, "step": 25083 }, { "epoch": 2.719427580225499, "grad_norm": 0.3429332673549652, "learning_rate": 9.386333961990426e-06, "loss": 0.0152, "step": 25084 }, { "epoch": 2.7195359930615783, "grad_norm": 0.8226736187934875, "learning_rate": 9.382707094153489e-06, "loss": 0.0178, "step": 25085 }, { "epoch": 2.7196444058976583, "grad_norm": 0.06744194775819778, "learning_rate": 9.379080226316553e-06, "loss": 0.0026, "step": 25086 }, { "epoch": 2.7197528187337383, "grad_norm": 0.2880370020866394, "learning_rate": 9.375453358479617e-06, "loss": 0.0075, "step": 25087 }, { "epoch": 2.719861231569818, "grad_norm": 1.1163755655288696, "learning_rate": 9.371826490642681e-06, "loss": 0.0336, "step": 25088 }, { "epoch": 2.7199696444058974, "grad_norm": 0.8369730114936829, "learning_rate": 9.368199622805745e-06, "loss": 0.0166, "step": 25089 }, { "epoch": 2.7200780572419774, "grad_norm": 0.22511444985866547, "learning_rate": 9.36457275496881e-06, "loss": 0.0124, "step": 25090 }, { "epoch": 2.7201864700780574, "grad_norm": 0.14486917853355408, "learning_rate": 9.360945887131873e-06, "loss": 0.0106, "step": 25091 }, { "epoch": 2.720294882914137, "grad_norm": 0.5835540294647217, "learning_rate": 9.357319019294938e-06, "loss": 0.0246, "step": 25092 }, { "epoch": 2.720403295750217, "grad_norm": 0.09523485600948334, "learning_rate": 9.353692151458002e-06, "loss": 0.0044, "step": 25093 }, { "epoch": 2.7205117085862964, "grad_norm": 0.7050431966781616, "learning_rate": 9.350065283621066e-06, "loss": 0.0275, "step": 25094 }, { "epoch": 2.7206201214223764, "grad_norm": 0.04324037954211235, "learning_rate": 9.34643841578413e-06, "loss": 0.0015, "step": 25095 }, { "epoch": 2.7207285342584564, "grad_norm": 0.23485113680362701, "learning_rate": 9.342811547947192e-06, "loss": 0.0059, "step": 25096 }, { "epoch": 2.720836947094536, "grad_norm": 0.629476010799408, "learning_rate": 9.339184680110258e-06, "loss": 0.0321, "step": 25097 }, { "epoch": 2.720945359930616, "grad_norm": 0.1262139528989792, "learning_rate": 9.33555781227332e-06, "loss": 0.0121, "step": 25098 }, { "epoch": 2.7210537727666955, "grad_norm": 0.06967511028051376, "learning_rate": 9.331930944436385e-06, "loss": 0.0037, "step": 25099 }, { "epoch": 2.7211621856027755, "grad_norm": 0.09769978374242783, "learning_rate": 9.328304076599449e-06, "loss": 0.0029, "step": 25100 }, { "epoch": 2.7212705984388554, "grad_norm": 0.15400563180446625, "learning_rate": 9.324677208762513e-06, "loss": 0.0041, "step": 25101 }, { "epoch": 2.721379011274935, "grad_norm": 0.21256078779697418, "learning_rate": 9.321050340925577e-06, "loss": 0.0144, "step": 25102 }, { "epoch": 2.7214874241110145, "grad_norm": 0.8670041561126709, "learning_rate": 9.317423473088641e-06, "loss": 0.0133, "step": 25103 }, { "epoch": 2.7215958369470945, "grad_norm": 0.5290213823318481, "learning_rate": 9.313796605251705e-06, "loss": 0.0267, "step": 25104 }, { "epoch": 2.7217042497831745, "grad_norm": 0.8116108775138855, "learning_rate": 9.31016973741477e-06, "loss": 0.0191, "step": 25105 }, { "epoch": 2.721812662619254, "grad_norm": 0.04966522008180618, "learning_rate": 9.306542869577833e-06, "loss": 0.0014, "step": 25106 }, { "epoch": 2.721921075455334, "grad_norm": 0.007273877039551735, "learning_rate": 9.302916001740897e-06, "loss": 0.0004, "step": 25107 }, { "epoch": 2.7220294882914136, "grad_norm": 0.500658392906189, "learning_rate": 9.299289133903962e-06, "loss": 0.035, "step": 25108 }, { "epoch": 2.7221379011274935, "grad_norm": 0.1578916609287262, "learning_rate": 9.295662266067024e-06, "loss": 0.004, "step": 25109 }, { "epoch": 2.7222463139635735, "grad_norm": 0.7947498559951782, "learning_rate": 9.29203539823009e-06, "loss": 0.0432, "step": 25110 }, { "epoch": 2.722354726799653, "grad_norm": 0.003322927514091134, "learning_rate": 9.288408530393152e-06, "loss": 0.0002, "step": 25111 }, { "epoch": 2.7224631396357326, "grad_norm": 0.10149095952510834, "learning_rate": 9.284781662556218e-06, "loss": 0.0042, "step": 25112 }, { "epoch": 2.7225715524718126, "grad_norm": 0.1305629014968872, "learning_rate": 9.28115479471928e-06, "loss": 0.0033, "step": 25113 }, { "epoch": 2.7226799653078926, "grad_norm": 0.10532025247812271, "learning_rate": 9.277527926882344e-06, "loss": 0.0102, "step": 25114 }, { "epoch": 2.722788378143972, "grad_norm": 0.33076420426368713, "learning_rate": 9.273901059045409e-06, "loss": 0.0239, "step": 25115 }, { "epoch": 2.722896790980052, "grad_norm": 0.14701975882053375, "learning_rate": 9.270274191208473e-06, "loss": 0.0108, "step": 25116 }, { "epoch": 2.7230052038161316, "grad_norm": 0.5070822834968567, "learning_rate": 9.266647323371537e-06, "loss": 0.0181, "step": 25117 }, { "epoch": 2.7231136166522116, "grad_norm": 0.42822492122650146, "learning_rate": 9.263020455534601e-06, "loss": 0.0075, "step": 25118 }, { "epoch": 2.7232220294882916, "grad_norm": 0.17369629442691803, "learning_rate": 9.259393587697665e-06, "loss": 0.01, "step": 25119 }, { "epoch": 2.723330442324371, "grad_norm": 0.4072871506214142, "learning_rate": 9.255766719860729e-06, "loss": 0.0689, "step": 25120 }, { "epoch": 2.723438855160451, "grad_norm": 0.2576722204685211, "learning_rate": 9.252139852023793e-06, "loss": 0.0096, "step": 25121 }, { "epoch": 2.7235472679965307, "grad_norm": 0.07769332081079483, "learning_rate": 9.248512984186857e-06, "loss": 0.0025, "step": 25122 }, { "epoch": 2.7236556808326107, "grad_norm": 0.014249333180487156, "learning_rate": 9.244886116349921e-06, "loss": 0.0003, "step": 25123 }, { "epoch": 2.72376409366869, "grad_norm": 0.3895241320133209, "learning_rate": 9.241259248512984e-06, "loss": 0.0109, "step": 25124 }, { "epoch": 2.72387250650477, "grad_norm": 0.07947612553834915, "learning_rate": 9.23763238067605e-06, "loss": 0.0046, "step": 25125 }, { "epoch": 2.7239809193408497, "grad_norm": 0.6378334760665894, "learning_rate": 9.234005512839112e-06, "loss": 0.0384, "step": 25126 }, { "epoch": 2.7240893321769297, "grad_norm": 0.26789361238479614, "learning_rate": 9.230378645002176e-06, "loss": 0.013, "step": 25127 }, { "epoch": 2.7241977450130097, "grad_norm": 0.17222826182842255, "learning_rate": 9.22675177716524e-06, "loss": 0.0055, "step": 25128 }, { "epoch": 2.7243061578490892, "grad_norm": 0.03423543646931648, "learning_rate": 9.223124909328304e-06, "loss": 0.0012, "step": 25129 }, { "epoch": 2.7244145706851692, "grad_norm": 0.28108111023902893, "learning_rate": 9.219498041491368e-06, "loss": 0.0058, "step": 25130 }, { "epoch": 2.7245229835212488, "grad_norm": 0.13084076344966888, "learning_rate": 9.215871173654433e-06, "loss": 0.0097, "step": 25131 }, { "epoch": 2.7246313963573288, "grad_norm": 0.10548652708530426, "learning_rate": 9.212244305817497e-06, "loss": 0.0091, "step": 25132 }, { "epoch": 2.7247398091934087, "grad_norm": 0.13381114602088928, "learning_rate": 9.20861743798056e-06, "loss": 0.0037, "step": 25133 }, { "epoch": 2.7248482220294883, "grad_norm": 0.03677919879555702, "learning_rate": 9.204990570143625e-06, "loss": 0.0023, "step": 25134 }, { "epoch": 2.724956634865568, "grad_norm": 0.3830868601799011, "learning_rate": 9.201363702306689e-06, "loss": 0.0062, "step": 25135 }, { "epoch": 2.725065047701648, "grad_norm": 0.31428828835487366, "learning_rate": 9.197736834469753e-06, "loss": 0.0053, "step": 25136 }, { "epoch": 2.725173460537728, "grad_norm": 0.00434316648170352, "learning_rate": 9.194109966632815e-06, "loss": 0.0001, "step": 25137 }, { "epoch": 2.7252818733738073, "grad_norm": 0.06635894626379013, "learning_rate": 9.190483098795881e-06, "loss": 0.0027, "step": 25138 }, { "epoch": 2.7253902862098873, "grad_norm": 0.12977276742458344, "learning_rate": 9.186856230958944e-06, "loss": 0.0054, "step": 25139 }, { "epoch": 2.725498699045967, "grad_norm": 0.672354519367218, "learning_rate": 9.183229363122008e-06, "loss": 0.0192, "step": 25140 }, { "epoch": 2.725607111882047, "grad_norm": 0.1398499459028244, "learning_rate": 9.179602495285072e-06, "loss": 0.0151, "step": 25141 }, { "epoch": 2.725715524718127, "grad_norm": 0.5691012740135193, "learning_rate": 9.175975627448136e-06, "loss": 0.0095, "step": 25142 }, { "epoch": 2.7258239375542064, "grad_norm": 0.05728555843234062, "learning_rate": 9.1723487596112e-06, "loss": 0.0017, "step": 25143 }, { "epoch": 2.7259323503902864, "grad_norm": 0.8324936032295227, "learning_rate": 9.168721891774264e-06, "loss": 0.0133, "step": 25144 }, { "epoch": 2.726040763226366, "grad_norm": 0.40401965379714966, "learning_rate": 9.165095023937328e-06, "loss": 0.0114, "step": 25145 }, { "epoch": 2.726149176062446, "grad_norm": 0.06426563113927841, "learning_rate": 9.161468156100392e-06, "loss": 0.0011, "step": 25146 }, { "epoch": 2.7262575888985254, "grad_norm": 0.3486958146095276, "learning_rate": 9.157841288263457e-06, "loss": 0.0209, "step": 25147 }, { "epoch": 2.7263660017346054, "grad_norm": 0.5247217416763306, "learning_rate": 9.15421442042652e-06, "loss": 0.0167, "step": 25148 }, { "epoch": 2.726474414570685, "grad_norm": 0.017997412011027336, "learning_rate": 9.150587552589585e-06, "loss": 0.0011, "step": 25149 }, { "epoch": 2.726582827406765, "grad_norm": 0.44264379143714905, "learning_rate": 9.146960684752647e-06, "loss": 0.023, "step": 25150 }, { "epoch": 2.726691240242845, "grad_norm": 0.08521728962659836, "learning_rate": 9.143333816915713e-06, "loss": 0.0026, "step": 25151 }, { "epoch": 2.7267996530789245, "grad_norm": 0.05623113736510277, "learning_rate": 9.139706949078775e-06, "loss": 0.0013, "step": 25152 }, { "epoch": 2.7269080659150045, "grad_norm": 0.3849354088306427, "learning_rate": 9.136080081241841e-06, "loss": 0.0155, "step": 25153 }, { "epoch": 2.727016478751084, "grad_norm": 0.5817087292671204, "learning_rate": 9.132453213404904e-06, "loss": 0.0196, "step": 25154 }, { "epoch": 2.727124891587164, "grad_norm": 0.4798714220523834, "learning_rate": 9.128826345567968e-06, "loss": 0.0356, "step": 25155 }, { "epoch": 2.727233304423244, "grad_norm": 0.13638333976268768, "learning_rate": 9.125199477731032e-06, "loss": 0.0041, "step": 25156 }, { "epoch": 2.7273417172593235, "grad_norm": 0.1289420872926712, "learning_rate": 9.121572609894096e-06, "loss": 0.0062, "step": 25157 }, { "epoch": 2.727450130095403, "grad_norm": 0.10144529491662979, "learning_rate": 9.11794574205716e-06, "loss": 0.0048, "step": 25158 }, { "epoch": 2.727558542931483, "grad_norm": 0.07889509946107864, "learning_rate": 9.114318874220224e-06, "loss": 0.003, "step": 25159 }, { "epoch": 2.727666955767563, "grad_norm": 0.38012251257896423, "learning_rate": 9.110692006383288e-06, "loss": 0.0034, "step": 25160 }, { "epoch": 2.7277753686036426, "grad_norm": 0.06015585735440254, "learning_rate": 9.107065138546352e-06, "loss": 0.0021, "step": 25161 }, { "epoch": 2.7278837814397225, "grad_norm": 0.2267133742570877, "learning_rate": 9.103438270709416e-06, "loss": 0.0083, "step": 25162 }, { "epoch": 2.727992194275802, "grad_norm": 0.08814336359500885, "learning_rate": 9.099811402872479e-06, "loss": 0.0019, "step": 25163 }, { "epoch": 2.728100607111882, "grad_norm": 0.09792319685220718, "learning_rate": 9.096184535035545e-06, "loss": 0.0025, "step": 25164 }, { "epoch": 2.728209019947962, "grad_norm": 0.005042390897870064, "learning_rate": 9.092557667198607e-06, "loss": 0.0002, "step": 25165 }, { "epoch": 2.7283174327840416, "grad_norm": 0.05147905275225639, "learning_rate": 9.088930799361673e-06, "loss": 0.003, "step": 25166 }, { "epoch": 2.7284258456201216, "grad_norm": 0.16449187695980072, "learning_rate": 9.085303931524735e-06, "loss": 0.0035, "step": 25167 }, { "epoch": 2.728534258456201, "grad_norm": 0.4887688159942627, "learning_rate": 9.0816770636878e-06, "loss": 0.0145, "step": 25168 }, { "epoch": 2.728642671292281, "grad_norm": 0.04648695886135101, "learning_rate": 9.078050195850863e-06, "loss": 0.0013, "step": 25169 }, { "epoch": 2.7287510841283606, "grad_norm": 0.28141993284225464, "learning_rate": 9.074423328013928e-06, "loss": 0.004, "step": 25170 }, { "epoch": 2.7288594969644406, "grad_norm": 0.22110578417778015, "learning_rate": 9.070796460176992e-06, "loss": 0.0111, "step": 25171 }, { "epoch": 2.72896790980052, "grad_norm": 0.1225641742348671, "learning_rate": 9.067169592340056e-06, "loss": 0.0038, "step": 25172 }, { "epoch": 2.7290763226366, "grad_norm": 0.2090422809123993, "learning_rate": 9.06354272450312e-06, "loss": 0.0052, "step": 25173 }, { "epoch": 2.72918473547268, "grad_norm": 0.5808700919151306, "learning_rate": 9.059915856666184e-06, "loss": 0.0202, "step": 25174 }, { "epoch": 2.7292931483087597, "grad_norm": 0.6418222188949585, "learning_rate": 9.056288988829248e-06, "loss": 0.0063, "step": 25175 }, { "epoch": 2.7294015611448397, "grad_norm": 0.3416087329387665, "learning_rate": 9.05266212099231e-06, "loss": 0.0092, "step": 25176 }, { "epoch": 2.729509973980919, "grad_norm": 0.8814167976379395, "learning_rate": 9.049035253155376e-06, "loss": 0.0261, "step": 25177 }, { "epoch": 2.729618386816999, "grad_norm": 0.40792763233184814, "learning_rate": 9.045408385318439e-06, "loss": 0.0353, "step": 25178 }, { "epoch": 2.729726799653079, "grad_norm": 0.09786952286958694, "learning_rate": 9.041781517481505e-06, "loss": 0.0037, "step": 25179 }, { "epoch": 2.7298352124891587, "grad_norm": 0.020147046074271202, "learning_rate": 9.038154649644567e-06, "loss": 0.0011, "step": 25180 }, { "epoch": 2.7299436253252383, "grad_norm": 0.00587802566587925, "learning_rate": 9.034527781807631e-06, "loss": 0.0003, "step": 25181 }, { "epoch": 2.7300520381613183, "grad_norm": 0.013750316575169563, "learning_rate": 9.030900913970695e-06, "loss": 0.001, "step": 25182 }, { "epoch": 2.7301604509973982, "grad_norm": 0.07346375286579132, "learning_rate": 9.02727404613376e-06, "loss": 0.0033, "step": 25183 }, { "epoch": 2.7302688638334778, "grad_norm": 0.028424996882677078, "learning_rate": 9.023647178296823e-06, "loss": 0.0009, "step": 25184 }, { "epoch": 2.7303772766695578, "grad_norm": 0.0055931247770786285, "learning_rate": 9.020020310459887e-06, "loss": 0.0002, "step": 25185 }, { "epoch": 2.7304856895056373, "grad_norm": 0.02782849781215191, "learning_rate": 9.016393442622952e-06, "loss": 0.001, "step": 25186 }, { "epoch": 2.7305941023417173, "grad_norm": 0.1200508177280426, "learning_rate": 9.012766574786016e-06, "loss": 0.0172, "step": 25187 }, { "epoch": 2.7307025151777973, "grad_norm": 0.13134273886680603, "learning_rate": 9.00913970694908e-06, "loss": 0.0051, "step": 25188 }, { "epoch": 2.730810928013877, "grad_norm": 0.04209139198064804, "learning_rate": 9.005512839112142e-06, "loss": 0.0021, "step": 25189 }, { "epoch": 2.730919340849957, "grad_norm": 0.09892184287309647, "learning_rate": 9.001885971275208e-06, "loss": 0.0025, "step": 25190 }, { "epoch": 2.7310277536860363, "grad_norm": 0.043637294322252274, "learning_rate": 8.99825910343827e-06, "loss": 0.0025, "step": 25191 }, { "epoch": 2.7311361665221163, "grad_norm": 0.11037000268697739, "learning_rate": 8.994632235601336e-06, "loss": 0.0012, "step": 25192 }, { "epoch": 2.731244579358196, "grad_norm": 0.4994478225708008, "learning_rate": 8.991005367764399e-06, "loss": 0.0347, "step": 25193 }, { "epoch": 2.731352992194276, "grad_norm": 0.7649836540222168, "learning_rate": 8.987378499927463e-06, "loss": 0.0131, "step": 25194 }, { "epoch": 2.7314614050303554, "grad_norm": 0.5108914375305176, "learning_rate": 8.983751632090527e-06, "loss": 0.015, "step": 25195 }, { "epoch": 2.7315698178664354, "grad_norm": 0.2474510371685028, "learning_rate": 8.980124764253591e-06, "loss": 0.0096, "step": 25196 }, { "epoch": 2.7316782307025154, "grad_norm": 1.3225547075271606, "learning_rate": 8.976497896416655e-06, "loss": 0.0515, "step": 25197 }, { "epoch": 2.731786643538595, "grad_norm": 0.14086006581783295, "learning_rate": 8.972871028579719e-06, "loss": 0.0019, "step": 25198 }, { "epoch": 2.731895056374675, "grad_norm": 0.336251437664032, "learning_rate": 8.969244160742783e-06, "loss": 0.02, "step": 25199 }, { "epoch": 2.7320034692107544, "grad_norm": 0.0958172082901001, "learning_rate": 8.965617292905847e-06, "loss": 0.0029, "step": 25200 }, { "epoch": 2.7321118820468344, "grad_norm": 0.14080439507961273, "learning_rate": 8.961990425068911e-06, "loss": 0.0047, "step": 25201 }, { "epoch": 2.7322202948829144, "grad_norm": 0.05333348736166954, "learning_rate": 8.958363557231974e-06, "loss": 0.0012, "step": 25202 }, { "epoch": 2.732328707718994, "grad_norm": 0.17726503312587738, "learning_rate": 8.95473668939504e-06, "loss": 0.0066, "step": 25203 }, { "epoch": 2.7324371205550735, "grad_norm": 0.13125573098659515, "learning_rate": 8.951109821558102e-06, "loss": 0.0038, "step": 25204 }, { "epoch": 2.7325455333911535, "grad_norm": 0.34419164061546326, "learning_rate": 8.947482953721168e-06, "loss": 0.0081, "step": 25205 }, { "epoch": 2.7326539462272335, "grad_norm": 0.6407275795936584, "learning_rate": 8.94385608588423e-06, "loss": 0.0509, "step": 25206 }, { "epoch": 2.732762359063313, "grad_norm": 0.022663353011012077, "learning_rate": 8.940229218047296e-06, "loss": 0.0008, "step": 25207 }, { "epoch": 2.732870771899393, "grad_norm": 0.6565684676170349, "learning_rate": 8.936602350210358e-06, "loss": 0.0142, "step": 25208 }, { "epoch": 2.7329791847354725, "grad_norm": 0.03961638733744621, "learning_rate": 8.932975482373423e-06, "loss": 0.0009, "step": 25209 }, { "epoch": 2.7330875975715525, "grad_norm": 0.048471663147211075, "learning_rate": 8.929348614536487e-06, "loss": 0.0016, "step": 25210 }, { "epoch": 2.7331960104076325, "grad_norm": 0.19754287600517273, "learning_rate": 8.92572174669955e-06, "loss": 0.0123, "step": 25211 }, { "epoch": 2.733304423243712, "grad_norm": 0.02262309566140175, "learning_rate": 8.922094878862615e-06, "loss": 0.001, "step": 25212 }, { "epoch": 2.7334128360797916, "grad_norm": 0.04951820150017738, "learning_rate": 8.918468011025679e-06, "loss": 0.0021, "step": 25213 }, { "epoch": 2.7335212489158716, "grad_norm": 0.005205790977925062, "learning_rate": 8.914841143188743e-06, "loss": 0.0002, "step": 25214 }, { "epoch": 2.7336296617519515, "grad_norm": 0.37911084294319153, "learning_rate": 8.911214275351806e-06, "loss": 0.0178, "step": 25215 }, { "epoch": 2.733738074588031, "grad_norm": 0.2723504304885864, "learning_rate": 8.907587407514871e-06, "loss": 0.008, "step": 25216 }, { "epoch": 2.733846487424111, "grad_norm": 0.6536766290664673, "learning_rate": 8.903960539677934e-06, "loss": 0.0313, "step": 25217 }, { "epoch": 2.7339549002601906, "grad_norm": 0.6421968936920166, "learning_rate": 8.900333671841e-06, "loss": 0.0319, "step": 25218 }, { "epoch": 2.7340633130962706, "grad_norm": 0.15685908496379852, "learning_rate": 8.896706804004062e-06, "loss": 0.0045, "step": 25219 }, { "epoch": 2.7341717259323506, "grad_norm": 0.01486979890614748, "learning_rate": 8.893079936167128e-06, "loss": 0.0007, "step": 25220 }, { "epoch": 2.73428013876843, "grad_norm": 0.660857617855072, "learning_rate": 8.88945306833019e-06, "loss": 0.0176, "step": 25221 }, { "epoch": 2.73438855160451, "grad_norm": 0.5527271628379822, "learning_rate": 8.885826200493254e-06, "loss": 0.0329, "step": 25222 }, { "epoch": 2.7344969644405897, "grad_norm": 0.06955079734325409, "learning_rate": 8.882199332656318e-06, "loss": 0.0092, "step": 25223 }, { "epoch": 2.7346053772766696, "grad_norm": 0.06564673036336899, "learning_rate": 8.878572464819382e-06, "loss": 0.0025, "step": 25224 }, { "epoch": 2.7347137901127496, "grad_norm": 0.052582647651433945, "learning_rate": 8.874945596982447e-06, "loss": 0.0037, "step": 25225 }, { "epoch": 2.734822202948829, "grad_norm": 0.10882037878036499, "learning_rate": 8.87131872914551e-06, "loss": 0.003, "step": 25226 }, { "epoch": 2.7349306157849087, "grad_norm": 0.017904918640851974, "learning_rate": 8.867691861308575e-06, "loss": 0.0006, "step": 25227 }, { "epoch": 2.7350390286209887, "grad_norm": 0.07255540043115616, "learning_rate": 8.864064993471637e-06, "loss": 0.002, "step": 25228 }, { "epoch": 2.7351474414570687, "grad_norm": 0.07860427349805832, "learning_rate": 8.860438125634703e-06, "loss": 0.0017, "step": 25229 }, { "epoch": 2.735255854293148, "grad_norm": 0.21639111638069153, "learning_rate": 8.856811257797765e-06, "loss": 0.0085, "step": 25230 }, { "epoch": 2.735364267129228, "grad_norm": 0.01428751926869154, "learning_rate": 8.853184389960831e-06, "loss": 0.0007, "step": 25231 }, { "epoch": 2.7354726799653077, "grad_norm": 0.18684662878513336, "learning_rate": 8.849557522123894e-06, "loss": 0.004, "step": 25232 }, { "epoch": 2.7355810928013877, "grad_norm": 0.36222416162490845, "learning_rate": 8.84593065428696e-06, "loss": 0.0191, "step": 25233 }, { "epoch": 2.7356895056374677, "grad_norm": 0.049372151494026184, "learning_rate": 8.842303786450022e-06, "loss": 0.0015, "step": 25234 }, { "epoch": 2.7357979184735473, "grad_norm": 0.2607831358909607, "learning_rate": 8.838676918613086e-06, "loss": 0.005, "step": 25235 }, { "epoch": 2.735906331309627, "grad_norm": 0.07348933815956116, "learning_rate": 8.83505005077615e-06, "loss": 0.0025, "step": 25236 }, { "epoch": 2.736014744145707, "grad_norm": 0.1606011986732483, "learning_rate": 8.831423182939214e-06, "loss": 0.0024, "step": 25237 }, { "epoch": 2.7361231569817868, "grad_norm": 0.6635103225708008, "learning_rate": 8.827796315102278e-06, "loss": 0.0237, "step": 25238 }, { "epoch": 2.7362315698178663, "grad_norm": 0.74142986536026, "learning_rate": 8.824169447265342e-06, "loss": 0.0124, "step": 25239 }, { "epoch": 2.7363399826539463, "grad_norm": 0.026970122009515762, "learning_rate": 8.820542579428406e-06, "loss": 0.0011, "step": 25240 }, { "epoch": 2.736448395490026, "grad_norm": 0.19776497781276703, "learning_rate": 8.816915711591469e-06, "loss": 0.0052, "step": 25241 }, { "epoch": 2.736556808326106, "grad_norm": 0.33493149280548096, "learning_rate": 8.813288843754535e-06, "loss": 0.013, "step": 25242 }, { "epoch": 2.736665221162186, "grad_norm": 0.32688042521476746, "learning_rate": 8.809661975917597e-06, "loss": 0.0055, "step": 25243 }, { "epoch": 2.7367736339982653, "grad_norm": 0.4433665871620178, "learning_rate": 8.806035108080663e-06, "loss": 0.0086, "step": 25244 }, { "epoch": 2.7368820468343453, "grad_norm": 0.3565356135368347, "learning_rate": 8.802408240243725e-06, "loss": 0.009, "step": 25245 }, { "epoch": 2.736990459670425, "grad_norm": 0.11053590476512909, "learning_rate": 8.798781372406791e-06, "loss": 0.0018, "step": 25246 }, { "epoch": 2.737098872506505, "grad_norm": 0.054194375872612, "learning_rate": 8.795154504569853e-06, "loss": 0.001, "step": 25247 }, { "epoch": 2.737207285342585, "grad_norm": 0.5369780659675598, "learning_rate": 8.791527636732918e-06, "loss": 0.0113, "step": 25248 }, { "epoch": 2.7373156981786644, "grad_norm": 0.21508866548538208, "learning_rate": 8.787900768895982e-06, "loss": 0.0049, "step": 25249 }, { "epoch": 2.737424111014744, "grad_norm": 0.0583292581140995, "learning_rate": 8.784273901059046e-06, "loss": 0.0011, "step": 25250 }, { "epoch": 2.737532523850824, "grad_norm": 0.6399322748184204, "learning_rate": 8.78064703322211e-06, "loss": 0.0103, "step": 25251 }, { "epoch": 2.737640936686904, "grad_norm": 0.5497017502784729, "learning_rate": 8.777020165385174e-06, "loss": 0.0677, "step": 25252 }, { "epoch": 2.7377493495229834, "grad_norm": 0.10970282554626465, "learning_rate": 8.773393297548238e-06, "loss": 0.0056, "step": 25253 }, { "epoch": 2.7378577623590634, "grad_norm": 0.0356314554810524, "learning_rate": 8.7697664297113e-06, "loss": 0.0026, "step": 25254 }, { "epoch": 2.737966175195143, "grad_norm": 0.013109339401125908, "learning_rate": 8.766139561874366e-06, "loss": 0.0004, "step": 25255 }, { "epoch": 2.738074588031223, "grad_norm": 0.02318984642624855, "learning_rate": 8.762512694037429e-06, "loss": 0.0008, "step": 25256 }, { "epoch": 2.738183000867303, "grad_norm": 0.16538521647453308, "learning_rate": 8.758885826200495e-06, "loss": 0.008, "step": 25257 }, { "epoch": 2.7382914137033825, "grad_norm": 0.20477762818336487, "learning_rate": 8.755258958363557e-06, "loss": 0.0118, "step": 25258 }, { "epoch": 2.738399826539462, "grad_norm": 0.5476346015930176, "learning_rate": 8.751632090526623e-06, "loss": 0.0191, "step": 25259 }, { "epoch": 2.738508239375542, "grad_norm": 0.502692699432373, "learning_rate": 8.748005222689685e-06, "loss": 0.0235, "step": 25260 }, { "epoch": 2.738616652211622, "grad_norm": 0.05887381359934807, "learning_rate": 8.744378354852751e-06, "loss": 0.0009, "step": 25261 }, { "epoch": 2.7387250650477015, "grad_norm": 0.1422102004289627, "learning_rate": 8.740751487015813e-06, "loss": 0.0039, "step": 25262 }, { "epoch": 2.7388334778837815, "grad_norm": 0.10489442944526672, "learning_rate": 8.737124619178877e-06, "loss": 0.0037, "step": 25263 }, { "epoch": 2.738941890719861, "grad_norm": 0.06909450143575668, "learning_rate": 8.733497751341942e-06, "loss": 0.0056, "step": 25264 }, { "epoch": 2.739050303555941, "grad_norm": 0.08510144799947739, "learning_rate": 8.729870883505006e-06, "loss": 0.0016, "step": 25265 }, { "epoch": 2.739158716392021, "grad_norm": 0.21210068464279175, "learning_rate": 8.72624401566807e-06, "loss": 0.0025, "step": 25266 }, { "epoch": 2.7392671292281006, "grad_norm": 0.050249189138412476, "learning_rate": 8.722617147831132e-06, "loss": 0.0009, "step": 25267 }, { "epoch": 2.7393755420641805, "grad_norm": 0.49504831433296204, "learning_rate": 8.718990279994198e-06, "loss": 0.033, "step": 25268 }, { "epoch": 2.73948395490026, "grad_norm": 0.19983789324760437, "learning_rate": 8.71536341215726e-06, "loss": 0.0081, "step": 25269 }, { "epoch": 2.73959236773634, "grad_norm": 0.12437651306390762, "learning_rate": 8.711736544320326e-06, "loss": 0.0033, "step": 25270 }, { "epoch": 2.7397007805724196, "grad_norm": 0.03662273287773132, "learning_rate": 8.708109676483389e-06, "loss": 0.0007, "step": 25271 }, { "epoch": 2.7398091934084996, "grad_norm": 0.45880189538002014, "learning_rate": 8.704482808646454e-06, "loss": 0.0199, "step": 25272 }, { "epoch": 2.739917606244579, "grad_norm": 0.12181639671325684, "learning_rate": 8.700855940809517e-06, "loss": 0.0051, "step": 25273 }, { "epoch": 2.740026019080659, "grad_norm": 0.05119018629193306, "learning_rate": 8.697229072972583e-06, "loss": 0.0017, "step": 25274 }, { "epoch": 2.740134431916739, "grad_norm": 0.41398748755455017, "learning_rate": 8.693602205135645e-06, "loss": 0.023, "step": 25275 }, { "epoch": 2.7402428447528187, "grad_norm": 0.25423261523246765, "learning_rate": 8.68997533729871e-06, "loss": 0.0285, "step": 25276 }, { "epoch": 2.7403512575888986, "grad_norm": 0.4541343152523041, "learning_rate": 8.686348469461773e-06, "loss": 0.0419, "step": 25277 }, { "epoch": 2.740459670424978, "grad_norm": 0.2376175969839096, "learning_rate": 8.682721601624837e-06, "loss": 0.008, "step": 25278 }, { "epoch": 2.740568083261058, "grad_norm": 0.007328800391405821, "learning_rate": 8.679094733787901e-06, "loss": 0.0003, "step": 25279 }, { "epoch": 2.740676496097138, "grad_norm": 0.020005498081445694, "learning_rate": 8.675467865950964e-06, "loss": 0.0008, "step": 25280 }, { "epoch": 2.7407849089332177, "grad_norm": 0.05810319632291794, "learning_rate": 8.67184099811403e-06, "loss": 0.0013, "step": 25281 }, { "epoch": 2.7408933217692972, "grad_norm": 0.04118172079324722, "learning_rate": 8.668214130277092e-06, "loss": 0.0025, "step": 25282 }, { "epoch": 2.741001734605377, "grad_norm": 0.16829858720302582, "learning_rate": 8.664587262440158e-06, "loss": 0.0059, "step": 25283 }, { "epoch": 2.741110147441457, "grad_norm": 0.05066194012761116, "learning_rate": 8.66096039460322e-06, "loss": 0.0026, "step": 25284 }, { "epoch": 2.7412185602775367, "grad_norm": 0.37923091650009155, "learning_rate": 8.657333526766286e-06, "loss": 0.0105, "step": 25285 }, { "epoch": 2.7413269731136167, "grad_norm": 0.6891000270843506, "learning_rate": 8.653706658929349e-06, "loss": 0.0272, "step": 25286 }, { "epoch": 2.7414353859496963, "grad_norm": 0.09194482117891312, "learning_rate": 8.650079791092414e-06, "loss": 0.002, "step": 25287 }, { "epoch": 2.7415437987857763, "grad_norm": 0.005320721305906773, "learning_rate": 8.646452923255477e-06, "loss": 0.0002, "step": 25288 }, { "epoch": 2.7416522116218562, "grad_norm": 0.03365785628557205, "learning_rate": 8.64282605541854e-06, "loss": 0.0006, "step": 25289 }, { "epoch": 2.741760624457936, "grad_norm": 0.42986994981765747, "learning_rate": 8.639199187581605e-06, "loss": 0.0106, "step": 25290 }, { "epoch": 2.7418690372940158, "grad_norm": 0.009097973816096783, "learning_rate": 8.635572319744669e-06, "loss": 0.0003, "step": 25291 }, { "epoch": 2.7419774501300953, "grad_norm": 0.0716981515288353, "learning_rate": 8.631945451907733e-06, "loss": 0.0021, "step": 25292 }, { "epoch": 2.7420858629661753, "grad_norm": 0.002477112924680114, "learning_rate": 8.628318584070797e-06, "loss": 0.0001, "step": 25293 }, { "epoch": 2.742194275802255, "grad_norm": 0.1888778805732727, "learning_rate": 8.624691716233861e-06, "loss": 0.0039, "step": 25294 }, { "epoch": 2.742302688638335, "grad_norm": 0.01632937230169773, "learning_rate": 8.621064848396924e-06, "loss": 0.0005, "step": 25295 }, { "epoch": 2.7424111014744144, "grad_norm": 0.3699072599411011, "learning_rate": 8.61743798055999e-06, "loss": 0.0113, "step": 25296 }, { "epoch": 2.7425195143104943, "grad_norm": 0.6290966868400574, "learning_rate": 8.613811112723052e-06, "loss": 0.0248, "step": 25297 }, { "epoch": 2.7426279271465743, "grad_norm": 0.3235389292240143, "learning_rate": 8.610184244886118e-06, "loss": 0.0078, "step": 25298 }, { "epoch": 2.742736339982654, "grad_norm": 0.052772946655750275, "learning_rate": 8.60655737704918e-06, "loss": 0.0018, "step": 25299 }, { "epoch": 2.742844752818734, "grad_norm": 0.8287891149520874, "learning_rate": 8.602930509212246e-06, "loss": 0.0217, "step": 25300 }, { "epoch": 2.7429531656548134, "grad_norm": 0.0812661424279213, "learning_rate": 8.599303641375308e-06, "loss": 0.0021, "step": 25301 }, { "epoch": 2.7430615784908934, "grad_norm": 0.02158442698419094, "learning_rate": 8.595676773538374e-06, "loss": 0.0004, "step": 25302 }, { "epoch": 2.7431699913269734, "grad_norm": 0.0029234648682177067, "learning_rate": 8.592049905701437e-06, "loss": 0.0001, "step": 25303 }, { "epoch": 2.743278404163053, "grad_norm": 0.06000927835702896, "learning_rate": 8.5884230378645e-06, "loss": 0.0017, "step": 25304 }, { "epoch": 2.7433868169991324, "grad_norm": 0.24174025654792786, "learning_rate": 8.584796170027565e-06, "loss": 0.0065, "step": 25305 }, { "epoch": 2.7434952298352124, "grad_norm": 0.19100995361804962, "learning_rate": 8.581169302190629e-06, "loss": 0.0042, "step": 25306 }, { "epoch": 2.7436036426712924, "grad_norm": 0.1346786469221115, "learning_rate": 8.577542434353693e-06, "loss": 0.0054, "step": 25307 }, { "epoch": 2.743712055507372, "grad_norm": 0.6438578963279724, "learning_rate": 8.573915566516755e-06, "loss": 0.0494, "step": 25308 }, { "epoch": 2.743820468343452, "grad_norm": 0.22380228340625763, "learning_rate": 8.570288698679821e-06, "loss": 0.0026, "step": 25309 }, { "epoch": 2.7439288811795315, "grad_norm": 0.07758905738592148, "learning_rate": 8.566661830842884e-06, "loss": 0.0045, "step": 25310 }, { "epoch": 2.7440372940156115, "grad_norm": 0.37655025720596313, "learning_rate": 8.56303496300595e-06, "loss": 0.0261, "step": 25311 }, { "epoch": 2.7441457068516915, "grad_norm": 0.37092113494873047, "learning_rate": 8.559408095169012e-06, "loss": 0.0093, "step": 25312 }, { "epoch": 2.744254119687771, "grad_norm": 0.12801708281040192, "learning_rate": 8.555781227332078e-06, "loss": 0.0057, "step": 25313 }, { "epoch": 2.744362532523851, "grad_norm": 0.22888189554214478, "learning_rate": 8.55215435949514e-06, "loss": 0.0059, "step": 25314 }, { "epoch": 2.7444709453599305, "grad_norm": 0.567191481590271, "learning_rate": 8.548527491658206e-06, "loss": 0.0159, "step": 25315 }, { "epoch": 2.7445793581960105, "grad_norm": 0.08586257696151733, "learning_rate": 8.544900623821268e-06, "loss": 0.002, "step": 25316 }, { "epoch": 2.74468777103209, "grad_norm": 0.28914564847946167, "learning_rate": 8.541273755984332e-06, "loss": 0.0056, "step": 25317 }, { "epoch": 2.74479618386817, "grad_norm": 0.0027779063675552607, "learning_rate": 8.537646888147396e-06, "loss": 0.0001, "step": 25318 }, { "epoch": 2.7449045967042496, "grad_norm": 0.19709959626197815, "learning_rate": 8.53402002031046e-06, "loss": 0.0078, "step": 25319 }, { "epoch": 2.7450130095403296, "grad_norm": 0.1326499879360199, "learning_rate": 8.530393152473525e-06, "loss": 0.0062, "step": 25320 }, { "epoch": 2.7451214223764095, "grad_norm": 0.1903143972158432, "learning_rate": 8.526766284636587e-06, "loss": 0.0066, "step": 25321 }, { "epoch": 2.745229835212489, "grad_norm": 0.2946547269821167, "learning_rate": 8.523139416799653e-06, "loss": 0.0081, "step": 25322 }, { "epoch": 2.745338248048569, "grad_norm": 0.25241437554359436, "learning_rate": 8.519512548962715e-06, "loss": 0.0185, "step": 25323 }, { "epoch": 2.7454466608846486, "grad_norm": 0.5626646876335144, "learning_rate": 8.515885681125781e-06, "loss": 0.0051, "step": 25324 }, { "epoch": 2.7455550737207286, "grad_norm": 0.15364453196525574, "learning_rate": 8.512258813288844e-06, "loss": 0.0022, "step": 25325 }, { "epoch": 2.7456634865568086, "grad_norm": 0.004554495215415955, "learning_rate": 8.50863194545191e-06, "loss": 0.0001, "step": 25326 }, { "epoch": 2.745771899392888, "grad_norm": 0.3019258677959442, "learning_rate": 8.505005077614972e-06, "loss": 0.0087, "step": 25327 }, { "epoch": 2.7458803122289677, "grad_norm": 0.23210972547531128, "learning_rate": 8.501378209778038e-06, "loss": 0.0042, "step": 25328 }, { "epoch": 2.7459887250650477, "grad_norm": 0.033130936324596405, "learning_rate": 8.4977513419411e-06, "loss": 0.0009, "step": 25329 }, { "epoch": 2.7460971379011276, "grad_norm": 0.03517840430140495, "learning_rate": 8.494124474104164e-06, "loss": 0.0011, "step": 25330 }, { "epoch": 2.746205550737207, "grad_norm": 0.0024173678830266, "learning_rate": 8.490497606267228e-06, "loss": 0.0001, "step": 25331 }, { "epoch": 2.746313963573287, "grad_norm": 0.2594146728515625, "learning_rate": 8.486870738430292e-06, "loss": 0.0179, "step": 25332 }, { "epoch": 2.7464223764093667, "grad_norm": 0.08009152114391327, "learning_rate": 8.483243870593356e-06, "loss": 0.0035, "step": 25333 }, { "epoch": 2.7465307892454467, "grad_norm": 0.3254483938217163, "learning_rate": 8.479617002756419e-06, "loss": 0.0045, "step": 25334 }, { "epoch": 2.7466392020815267, "grad_norm": 0.03859478607773781, "learning_rate": 8.475990134919485e-06, "loss": 0.0016, "step": 25335 }, { "epoch": 2.746747614917606, "grad_norm": 0.10405120998620987, "learning_rate": 8.472363267082547e-06, "loss": 0.0019, "step": 25336 }, { "epoch": 2.746856027753686, "grad_norm": 0.4408813416957855, "learning_rate": 8.468736399245613e-06, "loss": 0.0217, "step": 25337 }, { "epoch": 2.7469644405897657, "grad_norm": 0.0234964769333601, "learning_rate": 8.465109531408675e-06, "loss": 0.001, "step": 25338 }, { "epoch": 2.7470728534258457, "grad_norm": 0.030130727216601372, "learning_rate": 8.461482663571741e-06, "loss": 0.0004, "step": 25339 }, { "epoch": 2.7471812662619253, "grad_norm": 0.5314936637878418, "learning_rate": 8.457855795734803e-06, "loss": 0.0184, "step": 25340 }, { "epoch": 2.7472896790980053, "grad_norm": 0.01407370064407587, "learning_rate": 8.45422892789787e-06, "loss": 0.0004, "step": 25341 }, { "epoch": 2.747398091934085, "grad_norm": 0.7326737642288208, "learning_rate": 8.450602060060932e-06, "loss": 0.0091, "step": 25342 }, { "epoch": 2.747506504770165, "grad_norm": 0.13498756289482117, "learning_rate": 8.446975192223996e-06, "loss": 0.0019, "step": 25343 }, { "epoch": 2.7476149176062448, "grad_norm": 0.049927130341529846, "learning_rate": 8.44334832438706e-06, "loss": 0.002, "step": 25344 }, { "epoch": 2.7477233304423243, "grad_norm": 0.009114758111536503, "learning_rate": 8.439721456550124e-06, "loss": 0.0004, "step": 25345 }, { "epoch": 2.7478317432784043, "grad_norm": 0.0386880561709404, "learning_rate": 8.436094588713188e-06, "loss": 0.0011, "step": 25346 }, { "epoch": 2.747940156114484, "grad_norm": 0.047698725014925, "learning_rate": 8.43246772087625e-06, "loss": 0.0008, "step": 25347 }, { "epoch": 2.748048568950564, "grad_norm": 0.09868432581424713, "learning_rate": 8.428840853039316e-06, "loss": 0.0016, "step": 25348 }, { "epoch": 2.748156981786644, "grad_norm": 0.026218686252832413, "learning_rate": 8.425213985202379e-06, "loss": 0.0008, "step": 25349 }, { "epoch": 2.7482653946227233, "grad_norm": 0.48052898049354553, "learning_rate": 8.421587117365444e-06, "loss": 0.0145, "step": 25350 }, { "epoch": 2.748373807458803, "grad_norm": 0.18433882296085358, "learning_rate": 8.417960249528507e-06, "loss": 0.0083, "step": 25351 }, { "epoch": 2.748482220294883, "grad_norm": 0.18402346968650818, "learning_rate": 8.414333381691573e-06, "loss": 0.0156, "step": 25352 }, { "epoch": 2.748590633130963, "grad_norm": 0.17024295032024384, "learning_rate": 8.410706513854635e-06, "loss": 0.0064, "step": 25353 }, { "epoch": 2.7486990459670424, "grad_norm": 0.05604886636137962, "learning_rate": 8.407079646017701e-06, "loss": 0.0038, "step": 25354 }, { "epoch": 2.7488074588031224, "grad_norm": 0.5721002817153931, "learning_rate": 8.403452778180763e-06, "loss": 0.0193, "step": 25355 }, { "epoch": 2.748915871639202, "grad_norm": 0.14485004544258118, "learning_rate": 8.399825910343827e-06, "loss": 0.0088, "step": 25356 }, { "epoch": 2.749024284475282, "grad_norm": 0.4659962058067322, "learning_rate": 8.396199042506892e-06, "loss": 0.0192, "step": 25357 }, { "epoch": 2.749132697311362, "grad_norm": 0.009110277518630028, "learning_rate": 8.392572174669956e-06, "loss": 0.0002, "step": 25358 }, { "epoch": 2.7492411101474414, "grad_norm": 0.3251863420009613, "learning_rate": 8.38894530683302e-06, "loss": 0.0078, "step": 25359 }, { "epoch": 2.749349522983521, "grad_norm": 0.38674405217170715, "learning_rate": 8.385318438996082e-06, "loss": 0.021, "step": 25360 }, { "epoch": 2.749457935819601, "grad_norm": 0.010769244283437729, "learning_rate": 8.381691571159148e-06, "loss": 0.0004, "step": 25361 }, { "epoch": 2.749566348655681, "grad_norm": 0.2093512862920761, "learning_rate": 8.37806470332221e-06, "loss": 0.0052, "step": 25362 }, { "epoch": 2.7496747614917605, "grad_norm": 0.07437516003847122, "learning_rate": 8.374437835485276e-06, "loss": 0.0022, "step": 25363 }, { "epoch": 2.7497831743278405, "grad_norm": 0.04318968951702118, "learning_rate": 8.370810967648339e-06, "loss": 0.0013, "step": 25364 }, { "epoch": 2.74989158716392, "grad_norm": 0.42443615198135376, "learning_rate": 8.367184099811404e-06, "loss": 0.0099, "step": 25365 }, { "epoch": 2.75, "grad_norm": 0.46390700340270996, "learning_rate": 8.363557231974467e-06, "loss": 0.0071, "step": 25366 }, { "epoch": 2.75010841283608, "grad_norm": 0.25429999828338623, "learning_rate": 8.359930364137533e-06, "loss": 0.0102, "step": 25367 }, { "epoch": 2.7502168256721595, "grad_norm": 0.01654152199625969, "learning_rate": 8.356303496300595e-06, "loss": 0.0004, "step": 25368 }, { "epoch": 2.7503252385082395, "grad_norm": 0.060636576265096664, "learning_rate": 8.352676628463659e-06, "loss": 0.002, "step": 25369 }, { "epoch": 2.750433651344319, "grad_norm": 0.2244206815958023, "learning_rate": 8.349049760626723e-06, "loss": 0.0089, "step": 25370 }, { "epoch": 2.750542064180399, "grad_norm": 0.24700994789600372, "learning_rate": 8.345422892789787e-06, "loss": 0.0081, "step": 25371 }, { "epoch": 2.750650477016479, "grad_norm": 0.12465742230415344, "learning_rate": 8.341796024952851e-06, "loss": 0.0063, "step": 25372 }, { "epoch": 2.7507588898525586, "grad_norm": 0.5183313488960266, "learning_rate": 8.338169157115914e-06, "loss": 0.0337, "step": 25373 }, { "epoch": 2.750867302688638, "grad_norm": 0.3757792115211487, "learning_rate": 8.33454228927898e-06, "loss": 0.0106, "step": 25374 }, { "epoch": 2.750975715524718, "grad_norm": 0.043408919125795364, "learning_rate": 8.330915421442042e-06, "loss": 0.0011, "step": 25375 }, { "epoch": 2.751084128360798, "grad_norm": 0.5015031695365906, "learning_rate": 8.327288553605108e-06, "loss": 0.0101, "step": 25376 }, { "epoch": 2.7511925411968776, "grad_norm": 0.05557313561439514, "learning_rate": 8.32366168576817e-06, "loss": 0.0019, "step": 25377 }, { "epoch": 2.7513009540329576, "grad_norm": 0.04138147085905075, "learning_rate": 8.320034817931236e-06, "loss": 0.001, "step": 25378 }, { "epoch": 2.751409366869037, "grad_norm": 0.13066422939300537, "learning_rate": 8.316407950094298e-06, "loss": 0.0104, "step": 25379 }, { "epoch": 2.751517779705117, "grad_norm": 0.1953285038471222, "learning_rate": 8.312781082257364e-06, "loss": 0.0156, "step": 25380 }, { "epoch": 2.751626192541197, "grad_norm": 0.03517993912100792, "learning_rate": 8.309154214420427e-06, "loss": 0.0013, "step": 25381 }, { "epoch": 2.7517346053772767, "grad_norm": 0.35071685910224915, "learning_rate": 8.30552734658349e-06, "loss": 0.0175, "step": 25382 }, { "epoch": 2.751843018213356, "grad_norm": 0.26753032207489014, "learning_rate": 8.301900478746555e-06, "loss": 0.0259, "step": 25383 }, { "epoch": 2.751951431049436, "grad_norm": 0.20245401561260223, "learning_rate": 8.298273610909619e-06, "loss": 0.0116, "step": 25384 }, { "epoch": 2.752059843885516, "grad_norm": 0.2670312225818634, "learning_rate": 8.294646743072683e-06, "loss": 0.0055, "step": 25385 }, { "epoch": 2.7521682567215957, "grad_norm": 0.37135860323905945, "learning_rate": 8.291019875235745e-06, "loss": 0.0075, "step": 25386 }, { "epoch": 2.7522766695576757, "grad_norm": 0.20078152418136597, "learning_rate": 8.287393007398811e-06, "loss": 0.0044, "step": 25387 }, { "epoch": 2.7523850823937552, "grad_norm": 0.23487171530723572, "learning_rate": 8.283766139561874e-06, "loss": 0.0047, "step": 25388 }, { "epoch": 2.752493495229835, "grad_norm": 0.11044184118509293, "learning_rate": 8.28013927172494e-06, "loss": 0.0019, "step": 25389 }, { "epoch": 2.752601908065915, "grad_norm": 0.6081058382987976, "learning_rate": 8.276512403888002e-06, "loss": 0.0206, "step": 25390 }, { "epoch": 2.7527103209019947, "grad_norm": 0.12588496506214142, "learning_rate": 8.272885536051068e-06, "loss": 0.004, "step": 25391 }, { "epoch": 2.7528187337380747, "grad_norm": 0.40639832615852356, "learning_rate": 8.26925866821413e-06, "loss": 0.0322, "step": 25392 }, { "epoch": 2.7529271465741543, "grad_norm": 1.0135353803634644, "learning_rate": 8.265631800377196e-06, "loss": 0.0273, "step": 25393 }, { "epoch": 2.7530355594102343, "grad_norm": 0.6693283915519714, "learning_rate": 8.262004932540258e-06, "loss": 0.0201, "step": 25394 }, { "epoch": 2.753143972246314, "grad_norm": 0.8323928117752075, "learning_rate": 8.258378064703322e-06, "loss": 0.0505, "step": 25395 }, { "epoch": 2.753252385082394, "grad_norm": 0.03186406195163727, "learning_rate": 8.254751196866387e-06, "loss": 0.0006, "step": 25396 }, { "epoch": 2.7533607979184733, "grad_norm": 0.1104969009757042, "learning_rate": 8.25112432902945e-06, "loss": 0.0052, "step": 25397 }, { "epoch": 2.7534692107545533, "grad_norm": 0.2874401807785034, "learning_rate": 8.247497461192515e-06, "loss": 0.0097, "step": 25398 }, { "epoch": 2.7535776235906333, "grad_norm": 0.0283906739205122, "learning_rate": 8.243870593355577e-06, "loss": 0.0007, "step": 25399 }, { "epoch": 2.753686036426713, "grad_norm": 0.09812413901090622, "learning_rate": 8.240243725518643e-06, "loss": 0.0025, "step": 25400 }, { "epoch": 2.753794449262793, "grad_norm": 0.46067872643470764, "learning_rate": 8.236616857681705e-06, "loss": 0.0092, "step": 25401 }, { "epoch": 2.7539028620988724, "grad_norm": 0.06338628381490707, "learning_rate": 8.232989989844771e-06, "loss": 0.0018, "step": 25402 }, { "epoch": 2.7540112749349523, "grad_norm": 0.3482819199562073, "learning_rate": 8.229363122007834e-06, "loss": 0.0152, "step": 25403 }, { "epoch": 2.7541196877710323, "grad_norm": 0.06562897562980652, "learning_rate": 8.2257362541709e-06, "loss": 0.0017, "step": 25404 }, { "epoch": 2.754228100607112, "grad_norm": 0.1266961395740509, "learning_rate": 8.222109386333962e-06, "loss": 0.0049, "step": 25405 }, { "epoch": 2.7543365134431914, "grad_norm": 0.1247154176235199, "learning_rate": 8.218482518497028e-06, "loss": 0.0043, "step": 25406 }, { "epoch": 2.7544449262792714, "grad_norm": 0.05837244540452957, "learning_rate": 8.21485565066009e-06, "loss": 0.0045, "step": 25407 }, { "epoch": 2.7545533391153514, "grad_norm": 0.025627275928854942, "learning_rate": 8.211228782823154e-06, "loss": 0.0012, "step": 25408 }, { "epoch": 2.754661751951431, "grad_norm": 0.40666893124580383, "learning_rate": 8.207601914986218e-06, "loss": 0.0162, "step": 25409 }, { "epoch": 2.754770164787511, "grad_norm": 0.36309096217155457, "learning_rate": 8.203975047149282e-06, "loss": 0.0223, "step": 25410 }, { "epoch": 2.7548785776235905, "grad_norm": 0.06344898790121078, "learning_rate": 8.200348179312346e-06, "loss": 0.0024, "step": 25411 }, { "epoch": 2.7549869904596704, "grad_norm": 0.06843984872102737, "learning_rate": 8.196721311475409e-06, "loss": 0.0018, "step": 25412 }, { "epoch": 2.7550954032957504, "grad_norm": 0.06292769312858582, "learning_rate": 8.193094443638475e-06, "loss": 0.0017, "step": 25413 }, { "epoch": 2.75520381613183, "grad_norm": 0.16151787340641022, "learning_rate": 8.189467575801537e-06, "loss": 0.0024, "step": 25414 }, { "epoch": 2.75531222896791, "grad_norm": 0.6983410120010376, "learning_rate": 8.185840707964603e-06, "loss": 0.0127, "step": 25415 }, { "epoch": 2.7554206418039895, "grad_norm": 0.27867552638053894, "learning_rate": 8.182213840127665e-06, "loss": 0.015, "step": 25416 }, { "epoch": 2.7555290546400695, "grad_norm": 0.13318690657615662, "learning_rate": 8.178586972290731e-06, "loss": 0.0091, "step": 25417 }, { "epoch": 2.755637467476149, "grad_norm": 0.10874345898628235, "learning_rate": 8.174960104453793e-06, "loss": 0.0022, "step": 25418 }, { "epoch": 2.755745880312229, "grad_norm": 0.0005777213955298066, "learning_rate": 8.17133323661686e-06, "loss": 0.0, "step": 25419 }, { "epoch": 2.7558542931483085, "grad_norm": 1.4815016984939575, "learning_rate": 8.167706368779922e-06, "loss": 0.0214, "step": 25420 }, { "epoch": 2.7559627059843885, "grad_norm": 1.865895390510559, "learning_rate": 8.164079500942986e-06, "loss": 0.0287, "step": 25421 }, { "epoch": 2.7560711188204685, "grad_norm": 0.14340630173683167, "learning_rate": 8.16045263310605e-06, "loss": 0.0051, "step": 25422 }, { "epoch": 2.756179531656548, "grad_norm": 0.4922105073928833, "learning_rate": 8.156825765269114e-06, "loss": 0.0127, "step": 25423 }, { "epoch": 2.756287944492628, "grad_norm": 0.5375343561172485, "learning_rate": 8.153198897432178e-06, "loss": 0.0208, "step": 25424 }, { "epoch": 2.7563963573287076, "grad_norm": 0.08460375666618347, "learning_rate": 8.14957202959524e-06, "loss": 0.0027, "step": 25425 }, { "epoch": 2.7565047701647876, "grad_norm": 0.009863439947366714, "learning_rate": 8.145945161758306e-06, "loss": 0.0003, "step": 25426 }, { "epoch": 2.7566131830008676, "grad_norm": 0.3389238119125366, "learning_rate": 8.142318293921369e-06, "loss": 0.0134, "step": 25427 }, { "epoch": 2.756721595836947, "grad_norm": 0.8531919717788696, "learning_rate": 8.138691426084434e-06, "loss": 0.0312, "step": 25428 }, { "epoch": 2.7568300086730266, "grad_norm": 0.417400985956192, "learning_rate": 8.135064558247497e-06, "loss": 0.0175, "step": 25429 }, { "epoch": 2.7569384215091066, "grad_norm": 0.25635793805122375, "learning_rate": 8.131437690410563e-06, "loss": 0.0099, "step": 25430 }, { "epoch": 2.7570468343451866, "grad_norm": 0.4667787551879883, "learning_rate": 8.127810822573625e-06, "loss": 0.0426, "step": 25431 }, { "epoch": 2.757155247181266, "grad_norm": 0.20295396447181702, "learning_rate": 8.124183954736691e-06, "loss": 0.0119, "step": 25432 }, { "epoch": 2.757263660017346, "grad_norm": 0.46043533086776733, "learning_rate": 8.120557086899753e-06, "loss": 0.0246, "step": 25433 }, { "epoch": 2.7573720728534257, "grad_norm": 0.2853856384754181, "learning_rate": 8.116930219062817e-06, "loss": 0.0071, "step": 25434 }, { "epoch": 2.7574804856895057, "grad_norm": 0.18763142824172974, "learning_rate": 8.113303351225882e-06, "loss": 0.0077, "step": 25435 }, { "epoch": 2.7575888985255856, "grad_norm": 0.13747525215148926, "learning_rate": 8.109676483388946e-06, "loss": 0.0032, "step": 25436 }, { "epoch": 2.757697311361665, "grad_norm": 0.7032822966575623, "learning_rate": 8.10604961555201e-06, "loss": 0.0169, "step": 25437 }, { "epoch": 2.757805724197745, "grad_norm": 0.028145702555775642, "learning_rate": 8.102422747715072e-06, "loss": 0.0015, "step": 25438 }, { "epoch": 2.7579141370338247, "grad_norm": 0.6675248742103577, "learning_rate": 8.098795879878138e-06, "loss": 0.0509, "step": 25439 }, { "epoch": 2.7580225498699047, "grad_norm": 0.45288553833961487, "learning_rate": 8.0951690120412e-06, "loss": 0.0105, "step": 25440 }, { "epoch": 2.7581309627059842, "grad_norm": 0.2135772705078125, "learning_rate": 8.091542144204266e-06, "loss": 0.0094, "step": 25441 }, { "epoch": 2.758239375542064, "grad_norm": 0.02153155766427517, "learning_rate": 8.087915276367329e-06, "loss": 0.0009, "step": 25442 }, { "epoch": 2.7583477883781438, "grad_norm": 0.17441540956497192, "learning_rate": 8.084288408530394e-06, "loss": 0.0073, "step": 25443 }, { "epoch": 2.7584562012142237, "grad_norm": 0.11471952497959137, "learning_rate": 8.080661540693457e-06, "loss": 0.0042, "step": 25444 }, { "epoch": 2.7585646140503037, "grad_norm": 0.4471246898174286, "learning_rate": 8.077034672856523e-06, "loss": 0.0154, "step": 25445 }, { "epoch": 2.7586730268863833, "grad_norm": 1.1131527423858643, "learning_rate": 8.073407805019585e-06, "loss": 0.0349, "step": 25446 }, { "epoch": 2.7587814397224633, "grad_norm": 0.14573226869106293, "learning_rate": 8.069780937182649e-06, "loss": 0.0039, "step": 25447 }, { "epoch": 2.758889852558543, "grad_norm": 0.16697275638580322, "learning_rate": 8.066154069345713e-06, "loss": 0.0026, "step": 25448 }, { "epoch": 2.758998265394623, "grad_norm": 0.013579036109149456, "learning_rate": 8.062527201508777e-06, "loss": 0.0002, "step": 25449 }, { "epoch": 2.7591066782307028, "grad_norm": 0.0802825316786766, "learning_rate": 8.058900333671841e-06, "loss": 0.0039, "step": 25450 }, { "epoch": 2.7592150910667823, "grad_norm": 0.3498421907424927, "learning_rate": 8.055273465834906e-06, "loss": 0.0077, "step": 25451 }, { "epoch": 2.759323503902862, "grad_norm": 0.3585348427295685, "learning_rate": 8.05164659799797e-06, "loss": 0.0221, "step": 25452 }, { "epoch": 2.759431916738942, "grad_norm": 0.08633267879486084, "learning_rate": 8.048019730161032e-06, "loss": 0.0018, "step": 25453 }, { "epoch": 2.759540329575022, "grad_norm": 0.9583837389945984, "learning_rate": 8.044392862324098e-06, "loss": 0.045, "step": 25454 }, { "epoch": 2.7596487424111014, "grad_norm": 0.0884544774889946, "learning_rate": 8.04076599448716e-06, "loss": 0.0023, "step": 25455 }, { "epoch": 2.7597571552471813, "grad_norm": 0.27777424454689026, "learning_rate": 8.037139126650226e-06, "loss": 0.0053, "step": 25456 }, { "epoch": 2.759865568083261, "grad_norm": 0.011257817037403584, "learning_rate": 8.033512258813288e-06, "loss": 0.0004, "step": 25457 }, { "epoch": 2.759973980919341, "grad_norm": 0.175065815448761, "learning_rate": 8.029885390976354e-06, "loss": 0.0036, "step": 25458 }, { "epoch": 2.760082393755421, "grad_norm": 0.03932482749223709, "learning_rate": 8.026258523139417e-06, "loss": 0.0014, "step": 25459 }, { "epoch": 2.7601908065915004, "grad_norm": 0.014970450662076473, "learning_rate": 8.02263165530248e-06, "loss": 0.0004, "step": 25460 }, { "epoch": 2.7602992194275804, "grad_norm": 0.8897171020507812, "learning_rate": 8.019004787465545e-06, "loss": 0.0221, "step": 25461 }, { "epoch": 2.76040763226366, "grad_norm": 0.12827979028224945, "learning_rate": 8.015377919628609e-06, "loss": 0.0032, "step": 25462 }, { "epoch": 2.76051604509974, "grad_norm": 0.09111529588699341, "learning_rate": 8.011751051791673e-06, "loss": 0.0018, "step": 25463 }, { "epoch": 2.7606244579358195, "grad_norm": 0.5167951583862305, "learning_rate": 8.008124183954737e-06, "loss": 0.0355, "step": 25464 }, { "epoch": 2.7607328707718994, "grad_norm": 0.0860108733177185, "learning_rate": 8.004497316117801e-06, "loss": 0.0031, "step": 25465 }, { "epoch": 2.760841283607979, "grad_norm": 0.47522905468940735, "learning_rate": 8.000870448280864e-06, "loss": 0.0277, "step": 25466 }, { "epoch": 2.760949696444059, "grad_norm": 0.16367541253566742, "learning_rate": 7.99724358044393e-06, "loss": 0.0098, "step": 25467 }, { "epoch": 2.761058109280139, "grad_norm": 0.1816336065530777, "learning_rate": 7.993616712606992e-06, "loss": 0.0113, "step": 25468 }, { "epoch": 2.7611665221162185, "grad_norm": 0.3772253096103668, "learning_rate": 7.989989844770058e-06, "loss": 0.0115, "step": 25469 }, { "epoch": 2.7612749349522985, "grad_norm": 0.0996190533041954, "learning_rate": 7.98636297693312e-06, "loss": 0.0039, "step": 25470 }, { "epoch": 2.761383347788378, "grad_norm": 0.33276456594467163, "learning_rate": 7.982736109096186e-06, "loss": 0.0105, "step": 25471 }, { "epoch": 2.761491760624458, "grad_norm": 0.16934603452682495, "learning_rate": 7.979109241259248e-06, "loss": 0.0078, "step": 25472 }, { "epoch": 2.761600173460538, "grad_norm": 0.6401922106742859, "learning_rate": 7.975482373422314e-06, "loss": 0.0139, "step": 25473 }, { "epoch": 2.7617085862966175, "grad_norm": 0.28709450364112854, "learning_rate": 7.971855505585377e-06, "loss": 0.0515, "step": 25474 }, { "epoch": 2.761816999132697, "grad_norm": 0.14594919979572296, "learning_rate": 7.96822863774844e-06, "loss": 0.005, "step": 25475 }, { "epoch": 2.761925411968777, "grad_norm": 0.12637761235237122, "learning_rate": 7.964601769911505e-06, "loss": 0.0087, "step": 25476 }, { "epoch": 2.762033824804857, "grad_norm": 0.4172908067703247, "learning_rate": 7.960974902074569e-06, "loss": 0.0158, "step": 25477 }, { "epoch": 2.7621422376409366, "grad_norm": 0.14510634541511536, "learning_rate": 7.957348034237633e-06, "loss": 0.0054, "step": 25478 }, { "epoch": 2.7622506504770166, "grad_norm": 0.01061554066836834, "learning_rate": 7.953721166400695e-06, "loss": 0.0004, "step": 25479 }, { "epoch": 2.762359063313096, "grad_norm": 0.33252447843551636, "learning_rate": 7.950094298563761e-06, "loss": 0.0169, "step": 25480 }, { "epoch": 2.762467476149176, "grad_norm": 0.0949581190943718, "learning_rate": 7.946467430726824e-06, "loss": 0.003, "step": 25481 }, { "epoch": 2.762575888985256, "grad_norm": 0.10759378224611282, "learning_rate": 7.94284056288989e-06, "loss": 0.0047, "step": 25482 }, { "epoch": 2.7626843018213356, "grad_norm": 0.27488431334495544, "learning_rate": 7.939213695052952e-06, "loss": 0.017, "step": 25483 }, { "epoch": 2.762792714657415, "grad_norm": 0.3765949308872223, "learning_rate": 7.935586827216018e-06, "loss": 0.0113, "step": 25484 }, { "epoch": 2.762901127493495, "grad_norm": 0.1900867074728012, "learning_rate": 7.93195995937908e-06, "loss": 0.0072, "step": 25485 }, { "epoch": 2.763009540329575, "grad_norm": 0.7573415040969849, "learning_rate": 7.928333091542146e-06, "loss": 0.0233, "step": 25486 }, { "epoch": 2.7631179531656547, "grad_norm": 0.053010400384664536, "learning_rate": 7.924706223705208e-06, "loss": 0.0019, "step": 25487 }, { "epoch": 2.7632263660017347, "grad_norm": 0.3894694149494171, "learning_rate": 7.921079355868272e-06, "loss": 0.0185, "step": 25488 }, { "epoch": 2.763334778837814, "grad_norm": 0.6668796539306641, "learning_rate": 7.917452488031336e-06, "loss": 0.0257, "step": 25489 }, { "epoch": 2.763443191673894, "grad_norm": 0.26754650473594666, "learning_rate": 7.9138256201944e-06, "loss": 0.0096, "step": 25490 }, { "epoch": 2.763551604509974, "grad_norm": 0.10013530403375626, "learning_rate": 7.910198752357465e-06, "loss": 0.0017, "step": 25491 }, { "epoch": 2.7636600173460537, "grad_norm": 0.4053761959075928, "learning_rate": 7.906571884520527e-06, "loss": 0.0236, "step": 25492 }, { "epoch": 2.7637684301821337, "grad_norm": 0.5514742732048035, "learning_rate": 7.902945016683593e-06, "loss": 0.0178, "step": 25493 }, { "epoch": 2.7638768430182132, "grad_norm": 0.4520929157733917, "learning_rate": 7.899318148846655e-06, "loss": 0.0071, "step": 25494 }, { "epoch": 2.763985255854293, "grad_norm": 0.5065397620201111, "learning_rate": 7.895691281009721e-06, "loss": 0.0187, "step": 25495 }, { "epoch": 2.764093668690373, "grad_norm": 0.32919034361839294, "learning_rate": 7.892064413172783e-06, "loss": 0.0063, "step": 25496 }, { "epoch": 2.7642020815264527, "grad_norm": 0.3190292418003082, "learning_rate": 7.88843754533585e-06, "loss": 0.039, "step": 25497 }, { "epoch": 2.7643104943625323, "grad_norm": 0.027104375883936882, "learning_rate": 7.884810677498912e-06, "loss": 0.0013, "step": 25498 }, { "epoch": 2.7644189071986123, "grad_norm": 0.06017564609646797, "learning_rate": 7.881183809661977e-06, "loss": 0.0025, "step": 25499 }, { "epoch": 2.7645273200346923, "grad_norm": 0.021190643310546875, "learning_rate": 7.87755694182504e-06, "loss": 0.0005, "step": 25500 }, { "epoch": 2.764635732870772, "grad_norm": 0.3135448694229126, "learning_rate": 7.873930073988104e-06, "loss": 0.0489, "step": 25501 }, { "epoch": 2.764744145706852, "grad_norm": 0.5341702103614807, "learning_rate": 7.870303206151168e-06, "loss": 0.0118, "step": 25502 }, { "epoch": 2.7648525585429313, "grad_norm": 0.24162405729293823, "learning_rate": 7.866676338314232e-06, "loss": 0.0071, "step": 25503 }, { "epoch": 2.7649609713790113, "grad_norm": 0.35758429765701294, "learning_rate": 7.863049470477296e-06, "loss": 0.0103, "step": 25504 }, { "epoch": 2.7650693842150913, "grad_norm": 0.09439630806446075, "learning_rate": 7.85942260264036e-06, "loss": 0.0039, "step": 25505 }, { "epoch": 2.765177797051171, "grad_norm": 0.4433755874633789, "learning_rate": 7.855795734803425e-06, "loss": 0.0112, "step": 25506 }, { "epoch": 2.7652862098872504, "grad_norm": 0.08986688405275345, "learning_rate": 7.852168866966487e-06, "loss": 0.0036, "step": 25507 }, { "epoch": 2.7653946227233304, "grad_norm": 0.31004783511161804, "learning_rate": 7.848541999129553e-06, "loss": 0.0094, "step": 25508 }, { "epoch": 2.7655030355594103, "grad_norm": 0.4429796636104584, "learning_rate": 7.844915131292615e-06, "loss": 0.0164, "step": 25509 }, { "epoch": 2.76561144839549, "grad_norm": 0.11816715449094772, "learning_rate": 7.841288263455681e-06, "loss": 0.0062, "step": 25510 }, { "epoch": 2.76571986123157, "grad_norm": 0.004806828685104847, "learning_rate": 7.837661395618743e-06, "loss": 0.0003, "step": 25511 }, { "epoch": 2.7658282740676494, "grad_norm": 0.16960053145885468, "learning_rate": 7.83403452778181e-06, "loss": 0.0023, "step": 25512 }, { "epoch": 2.7659366869037294, "grad_norm": 0.3081686794757843, "learning_rate": 7.830407659944872e-06, "loss": 0.0171, "step": 25513 }, { "epoch": 2.7660450997398094, "grad_norm": 0.04944249987602234, "learning_rate": 7.826780792107936e-06, "loss": 0.002, "step": 25514 }, { "epoch": 2.766153512575889, "grad_norm": 0.9733138084411621, "learning_rate": 7.823153924271e-06, "loss": 0.0323, "step": 25515 }, { "epoch": 2.766261925411969, "grad_norm": 0.3086920380592346, "learning_rate": 7.819527056434064e-06, "loss": 0.0035, "step": 25516 }, { "epoch": 2.7663703382480485, "grad_norm": 0.7014166116714478, "learning_rate": 7.815900188597128e-06, "loss": 0.0396, "step": 25517 }, { "epoch": 2.7664787510841284, "grad_norm": 0.4577993154525757, "learning_rate": 7.812273320760192e-06, "loss": 0.006, "step": 25518 }, { "epoch": 2.7665871639202084, "grad_norm": 0.45394638180732727, "learning_rate": 7.808646452923256e-06, "loss": 0.0339, "step": 25519 }, { "epoch": 2.766695576756288, "grad_norm": 0.06824857741594315, "learning_rate": 7.805019585086319e-06, "loss": 0.0023, "step": 25520 }, { "epoch": 2.7668039895923675, "grad_norm": 0.33371713757514954, "learning_rate": 7.801392717249384e-06, "loss": 0.0188, "step": 25521 }, { "epoch": 2.7669124024284475, "grad_norm": 0.07138640433549881, "learning_rate": 7.797765849412447e-06, "loss": 0.0039, "step": 25522 }, { "epoch": 2.7670208152645275, "grad_norm": 0.18912765383720398, "learning_rate": 7.794138981575513e-06, "loss": 0.0044, "step": 25523 }, { "epoch": 2.767129228100607, "grad_norm": 0.5943118333816528, "learning_rate": 7.790512113738575e-06, "loss": 0.0391, "step": 25524 }, { "epoch": 2.767237640936687, "grad_norm": 0.4166707992553711, "learning_rate": 7.78688524590164e-06, "loss": 0.0079, "step": 25525 }, { "epoch": 2.7673460537727665, "grad_norm": 0.2575004994869232, "learning_rate": 7.783258378064703e-06, "loss": 0.0117, "step": 25526 }, { "epoch": 2.7674544666088465, "grad_norm": 0.3159855008125305, "learning_rate": 7.779631510227767e-06, "loss": 0.0082, "step": 25527 }, { "epoch": 2.7675628794449265, "grad_norm": 1.0165764093399048, "learning_rate": 7.776004642390831e-06, "loss": 0.0202, "step": 25528 }, { "epoch": 2.767671292281006, "grad_norm": 0.3024688959121704, "learning_rate": 7.772377774553896e-06, "loss": 0.0143, "step": 25529 }, { "epoch": 2.7677797051170856, "grad_norm": 0.017618747428059578, "learning_rate": 7.76875090671696e-06, "loss": 0.0004, "step": 25530 }, { "epoch": 2.7678881179531656, "grad_norm": 0.787907600402832, "learning_rate": 7.765124038880024e-06, "loss": 0.0277, "step": 25531 }, { "epoch": 2.7679965307892456, "grad_norm": 0.43566998839378357, "learning_rate": 7.761497171043088e-06, "loss": 0.0116, "step": 25532 }, { "epoch": 2.768104943625325, "grad_norm": 0.14865007996559143, "learning_rate": 7.75787030320615e-06, "loss": 0.0099, "step": 25533 }, { "epoch": 2.768213356461405, "grad_norm": 0.43054836988449097, "learning_rate": 7.754243435369216e-06, "loss": 0.0091, "step": 25534 }, { "epoch": 2.7683217692974846, "grad_norm": 0.11645526438951492, "learning_rate": 7.750616567532278e-06, "loss": 0.0035, "step": 25535 }, { "epoch": 2.7684301821335646, "grad_norm": 0.03162102401256561, "learning_rate": 7.746989699695344e-06, "loss": 0.0006, "step": 25536 }, { "epoch": 2.7685385949696446, "grad_norm": 0.002068346831947565, "learning_rate": 7.743362831858407e-06, "loss": 0.0001, "step": 25537 }, { "epoch": 2.768647007805724, "grad_norm": 0.17341439425945282, "learning_rate": 7.739735964021473e-06, "loss": 0.0064, "step": 25538 }, { "epoch": 2.768755420641804, "grad_norm": 0.4456026554107666, "learning_rate": 7.736109096184535e-06, "loss": 0.0295, "step": 25539 }, { "epoch": 2.7688638334778837, "grad_norm": 0.17157413065433502, "learning_rate": 7.732482228347599e-06, "loss": 0.0104, "step": 25540 }, { "epoch": 2.7689722463139637, "grad_norm": 0.4455009996891022, "learning_rate": 7.728855360510663e-06, "loss": 0.0177, "step": 25541 }, { "epoch": 2.769080659150043, "grad_norm": 0.46665143966674805, "learning_rate": 7.725228492673727e-06, "loss": 0.0318, "step": 25542 }, { "epoch": 2.769189071986123, "grad_norm": 0.17431022226810455, "learning_rate": 7.721601624836791e-06, "loss": 0.0071, "step": 25543 }, { "epoch": 2.7692974848222027, "grad_norm": 1.9470609426498413, "learning_rate": 7.717974756999855e-06, "loss": 0.0187, "step": 25544 }, { "epoch": 2.7694058976582827, "grad_norm": 0.535302460193634, "learning_rate": 7.71434788916292e-06, "loss": 0.0063, "step": 25545 }, { "epoch": 2.7695143104943627, "grad_norm": 0.4092554748058319, "learning_rate": 7.710721021325982e-06, "loss": 0.0185, "step": 25546 }, { "epoch": 2.7696227233304422, "grad_norm": 0.2171017974615097, "learning_rate": 7.707094153489048e-06, "loss": 0.0029, "step": 25547 }, { "epoch": 2.7697311361665222, "grad_norm": 0.16628456115722656, "learning_rate": 7.70346728565211e-06, "loss": 0.0082, "step": 25548 }, { "epoch": 2.7698395490026018, "grad_norm": 0.4403412640094757, "learning_rate": 7.699840417815176e-06, "loss": 0.0104, "step": 25549 }, { "epoch": 2.7699479618386817, "grad_norm": 0.1715703308582306, "learning_rate": 7.696213549978238e-06, "loss": 0.0049, "step": 25550 }, { "epoch": 2.7700563746747617, "grad_norm": 0.0856977254152298, "learning_rate": 7.692586682141304e-06, "loss": 0.0025, "step": 25551 }, { "epoch": 2.7701647875108413, "grad_norm": 0.5758081674575806, "learning_rate": 7.688959814304367e-06, "loss": 0.0265, "step": 25552 }, { "epoch": 2.770273200346921, "grad_norm": 0.06318548321723938, "learning_rate": 7.68533294646743e-06, "loss": 0.0009, "step": 25553 }, { "epoch": 2.770381613183001, "grad_norm": 0.09768863767385483, "learning_rate": 7.681706078630495e-06, "loss": 0.0034, "step": 25554 }, { "epoch": 2.770490026019081, "grad_norm": 0.38685157895088196, "learning_rate": 7.678079210793559e-06, "loss": 0.0131, "step": 25555 }, { "epoch": 2.7705984388551603, "grad_norm": 0.06038337200880051, "learning_rate": 7.674452342956623e-06, "loss": 0.0008, "step": 25556 }, { "epoch": 2.7707068516912403, "grad_norm": 0.32103872299194336, "learning_rate": 7.670825475119687e-06, "loss": 0.0121, "step": 25557 }, { "epoch": 2.77081526452732, "grad_norm": 0.2600281536579132, "learning_rate": 7.667198607282751e-06, "loss": 0.0244, "step": 25558 }, { "epoch": 2.7709236773634, "grad_norm": 0.11396069079637527, "learning_rate": 7.663571739445815e-06, "loss": 0.0025, "step": 25559 }, { "epoch": 2.77103209019948, "grad_norm": 0.49645695090293884, "learning_rate": 7.65994487160888e-06, "loss": 0.0276, "step": 25560 }, { "epoch": 2.7711405030355594, "grad_norm": 0.3951297104358673, "learning_rate": 7.656318003771942e-06, "loss": 0.0098, "step": 25561 }, { "epoch": 2.7712489158716394, "grad_norm": 0.006598444655537605, "learning_rate": 7.652691135935008e-06, "loss": 0.0002, "step": 25562 }, { "epoch": 2.771357328707719, "grad_norm": 0.2805139124393463, "learning_rate": 7.64906426809807e-06, "loss": 0.0105, "step": 25563 }, { "epoch": 2.771465741543799, "grad_norm": 0.21793316304683685, "learning_rate": 7.645437400261136e-06, "loss": 0.0064, "step": 25564 }, { "epoch": 2.7715741543798784, "grad_norm": 0.4762653708457947, "learning_rate": 7.641810532424198e-06, "loss": 0.0231, "step": 25565 }, { "epoch": 2.7716825672159584, "grad_norm": 0.1681995540857315, "learning_rate": 7.638183664587262e-06, "loss": 0.0095, "step": 25566 }, { "epoch": 2.771790980052038, "grad_norm": 0.04248739406466484, "learning_rate": 7.634556796750326e-06, "loss": 0.0011, "step": 25567 }, { "epoch": 2.771899392888118, "grad_norm": 0.06371374428272247, "learning_rate": 7.63092992891339e-06, "loss": 0.0008, "step": 25568 }, { "epoch": 2.772007805724198, "grad_norm": 0.0459190271794796, "learning_rate": 7.627303061076455e-06, "loss": 0.0011, "step": 25569 }, { "epoch": 2.7721162185602775, "grad_norm": 0.20707100629806519, "learning_rate": 7.62367619323952e-06, "loss": 0.004, "step": 25570 }, { "epoch": 2.7722246313963574, "grad_norm": 0.2030859738588333, "learning_rate": 7.620049325402583e-06, "loss": 0.0034, "step": 25571 }, { "epoch": 2.772333044232437, "grad_norm": 0.08499995619058609, "learning_rate": 7.616422457565647e-06, "loss": 0.0039, "step": 25572 }, { "epoch": 2.772441457068517, "grad_norm": 0.023568905889987946, "learning_rate": 7.61279558972871e-06, "loss": 0.0003, "step": 25573 }, { "epoch": 2.772549869904597, "grad_norm": 0.15617205202579498, "learning_rate": 7.609168721891774e-06, "loss": 0.0038, "step": 25574 }, { "epoch": 2.7726582827406765, "grad_norm": 0.30860182642936707, "learning_rate": 7.6055418540548385e-06, "loss": 0.025, "step": 25575 }, { "epoch": 2.772766695576756, "grad_norm": 0.12778161466121674, "learning_rate": 7.601914986217902e-06, "loss": 0.0068, "step": 25576 }, { "epoch": 2.772875108412836, "grad_norm": 0.16494488716125488, "learning_rate": 7.598288118380967e-06, "loss": 0.0043, "step": 25577 }, { "epoch": 2.772983521248916, "grad_norm": 0.31685250997543335, "learning_rate": 7.59466125054403e-06, "loss": 0.0143, "step": 25578 }, { "epoch": 2.7730919340849955, "grad_norm": 0.06357922405004501, "learning_rate": 7.591034382707095e-06, "loss": 0.0016, "step": 25579 }, { "epoch": 2.7732003469210755, "grad_norm": 0.12425339221954346, "learning_rate": 7.587407514870158e-06, "loss": 0.0061, "step": 25580 }, { "epoch": 2.773308759757155, "grad_norm": 0.4995340406894684, "learning_rate": 7.583780647033223e-06, "loss": 0.0148, "step": 25581 }, { "epoch": 2.773417172593235, "grad_norm": 0.338986337184906, "learning_rate": 7.580153779196286e-06, "loss": 0.0122, "step": 25582 }, { "epoch": 2.773525585429315, "grad_norm": 0.6951065063476562, "learning_rate": 7.576526911359351e-06, "loss": 0.0632, "step": 25583 }, { "epoch": 2.7736339982653946, "grad_norm": 0.5033257603645325, "learning_rate": 7.5729000435224146e-06, "loss": 0.0282, "step": 25584 }, { "epoch": 2.7737424111014746, "grad_norm": 0.4479058086872101, "learning_rate": 7.569273175685479e-06, "loss": 0.0336, "step": 25585 }, { "epoch": 2.773850823937554, "grad_norm": 0.21957720816135406, "learning_rate": 7.565646307848542e-06, "loss": 0.0081, "step": 25586 }, { "epoch": 2.773959236773634, "grad_norm": 0.039358124136924744, "learning_rate": 7.562019440011606e-06, "loss": 0.0017, "step": 25587 }, { "epoch": 2.7740676496097136, "grad_norm": 0.2068265825510025, "learning_rate": 7.55839257217467e-06, "loss": 0.0189, "step": 25588 }, { "epoch": 2.7741760624457936, "grad_norm": 0.007985188625752926, "learning_rate": 7.554765704337733e-06, "loss": 0.0002, "step": 25589 }, { "epoch": 2.774284475281873, "grad_norm": 0.06843847036361694, "learning_rate": 7.551138836500798e-06, "loss": 0.0038, "step": 25590 }, { "epoch": 2.774392888117953, "grad_norm": 0.3214041292667389, "learning_rate": 7.547511968663862e-06, "loss": 0.0207, "step": 25591 }, { "epoch": 2.774501300954033, "grad_norm": 0.02561839297413826, "learning_rate": 7.5438851008269266e-06, "loss": 0.0009, "step": 25592 }, { "epoch": 2.7746097137901127, "grad_norm": 0.6022772789001465, "learning_rate": 7.54025823298999e-06, "loss": 0.0451, "step": 25593 }, { "epoch": 2.7747181266261927, "grad_norm": 0.0045685009099543095, "learning_rate": 7.536631365153055e-06, "loss": 0.0002, "step": 25594 }, { "epoch": 2.774826539462272, "grad_norm": 0.49313414096832275, "learning_rate": 7.533004497316118e-06, "loss": 0.0155, "step": 25595 }, { "epoch": 2.774934952298352, "grad_norm": 0.015638502314686775, "learning_rate": 7.529377629479183e-06, "loss": 0.0005, "step": 25596 }, { "epoch": 2.775043365134432, "grad_norm": 0.24323105812072754, "learning_rate": 7.525750761642246e-06, "loss": 0.0025, "step": 25597 }, { "epoch": 2.7751517779705117, "grad_norm": 0.5707228779792786, "learning_rate": 7.52212389380531e-06, "loss": 0.0237, "step": 25598 }, { "epoch": 2.7752601908065913, "grad_norm": 0.1295861005783081, "learning_rate": 7.5184970259683744e-06, "loss": 0.0063, "step": 25599 }, { "epoch": 2.7753686036426712, "grad_norm": 0.06552168726921082, "learning_rate": 7.5148701581314385e-06, "loss": 0.0023, "step": 25600 }, { "epoch": 2.7754770164787512, "grad_norm": 0.18421471118927002, "learning_rate": 7.511243290294502e-06, "loss": 0.0042, "step": 25601 }, { "epoch": 2.7755854293148308, "grad_norm": 0.1638912558555603, "learning_rate": 7.507616422457565e-06, "loss": 0.011, "step": 25602 }, { "epoch": 2.7756938421509108, "grad_norm": 0.6132904887199402, "learning_rate": 7.50398955462063e-06, "loss": 0.0116, "step": 25603 }, { "epoch": 2.7758022549869903, "grad_norm": 0.1649697870016098, "learning_rate": 7.500362686783693e-06, "loss": 0.005, "step": 25604 }, { "epoch": 2.7759106678230703, "grad_norm": 0.26867732405662537, "learning_rate": 7.496735818946758e-06, "loss": 0.0133, "step": 25605 }, { "epoch": 2.7760190806591503, "grad_norm": 0.04886358976364136, "learning_rate": 7.4931089511098215e-06, "loss": 0.0012, "step": 25606 }, { "epoch": 2.77612749349523, "grad_norm": 0.2743353247642517, "learning_rate": 7.4894820832728864e-06, "loss": 0.0083, "step": 25607 }, { "epoch": 2.77623590633131, "grad_norm": 0.8480798602104187, "learning_rate": 7.48585521543595e-06, "loss": 0.0177, "step": 25608 }, { "epoch": 2.7763443191673893, "grad_norm": 0.2875537574291229, "learning_rate": 7.482228347599015e-06, "loss": 0.0123, "step": 25609 }, { "epoch": 2.7764527320034693, "grad_norm": 0.14967095851898193, "learning_rate": 7.478601479762078e-06, "loss": 0.0085, "step": 25610 }, { "epoch": 2.776561144839549, "grad_norm": 0.025727543979883194, "learning_rate": 7.474974611925142e-06, "loss": 0.001, "step": 25611 }, { "epoch": 2.776669557675629, "grad_norm": 0.0028800179716199636, "learning_rate": 7.471347744088206e-06, "loss": 0.0001, "step": 25612 }, { "epoch": 2.7767779705117084, "grad_norm": 0.5847757458686829, "learning_rate": 7.46772087625127e-06, "loss": 0.0231, "step": 25613 }, { "epoch": 2.7768863833477884, "grad_norm": 0.1366773098707199, "learning_rate": 7.4640940084143335e-06, "loss": 0.0069, "step": 25614 }, { "epoch": 2.7769947961838684, "grad_norm": 0.12710468471050262, "learning_rate": 7.460467140577397e-06, "loss": 0.0018, "step": 25615 }, { "epoch": 2.777103209019948, "grad_norm": 0.028848707675933838, "learning_rate": 7.456840272740462e-06, "loss": 0.0008, "step": 25616 }, { "epoch": 2.777211621856028, "grad_norm": 0.19300439953804016, "learning_rate": 7.453213404903525e-06, "loss": 0.0078, "step": 25617 }, { "epoch": 2.7773200346921074, "grad_norm": 0.6806313991546631, "learning_rate": 7.44958653706659e-06, "loss": 0.0315, "step": 25618 }, { "epoch": 2.7774284475281874, "grad_norm": 0.04556131735444069, "learning_rate": 7.445959669229653e-06, "loss": 0.0014, "step": 25619 }, { "epoch": 2.7775368603642674, "grad_norm": 0.27875441312789917, "learning_rate": 7.442332801392718e-06, "loss": 0.0041, "step": 25620 }, { "epoch": 2.777645273200347, "grad_norm": 0.20764517784118652, "learning_rate": 7.438705933555781e-06, "loss": 0.0064, "step": 25621 }, { "epoch": 2.7777536860364265, "grad_norm": 0.18311525881290436, "learning_rate": 7.435079065718846e-06, "loss": 0.0161, "step": 25622 }, { "epoch": 2.7778620988725065, "grad_norm": 0.4169103801250458, "learning_rate": 7.4314521978819096e-06, "loss": 0.0197, "step": 25623 }, { "epoch": 2.7779705117085864, "grad_norm": 0.35086655616760254, "learning_rate": 7.427825330044974e-06, "loss": 0.0061, "step": 25624 }, { "epoch": 2.778078924544666, "grad_norm": 0.4046837091445923, "learning_rate": 7.424198462208038e-06, "loss": 0.0209, "step": 25625 }, { "epoch": 2.778187337380746, "grad_norm": 0.021625014021992683, "learning_rate": 7.420571594371102e-06, "loss": 0.0006, "step": 25626 }, { "epoch": 2.7782957502168255, "grad_norm": 0.21710605919361115, "learning_rate": 7.416944726534165e-06, "loss": 0.0168, "step": 25627 }, { "epoch": 2.7784041630529055, "grad_norm": 0.09132242202758789, "learning_rate": 7.413317858697228e-06, "loss": 0.0044, "step": 25628 }, { "epoch": 2.7785125758889855, "grad_norm": 0.9596697688102722, "learning_rate": 7.409690990860293e-06, "loss": 0.0301, "step": 25629 }, { "epoch": 2.778620988725065, "grad_norm": 0.38333892822265625, "learning_rate": 7.406064123023357e-06, "loss": 0.0104, "step": 25630 }, { "epoch": 2.7787294015611446, "grad_norm": 0.19715139269828796, "learning_rate": 7.4024372551864216e-06, "loss": 0.0093, "step": 25631 }, { "epoch": 2.7788378143972245, "grad_norm": 0.20026423037052155, "learning_rate": 7.398810387349485e-06, "loss": 0.0084, "step": 25632 }, { "epoch": 2.7789462272333045, "grad_norm": 0.29792991280555725, "learning_rate": 7.39518351951255e-06, "loss": 0.031, "step": 25633 }, { "epoch": 2.779054640069384, "grad_norm": 0.5728687644004822, "learning_rate": 7.391556651675613e-06, "loss": 0.0529, "step": 25634 }, { "epoch": 2.779163052905464, "grad_norm": 0.6721164584159851, "learning_rate": 7.387929783838678e-06, "loss": 0.0223, "step": 25635 }, { "epoch": 2.7792714657415436, "grad_norm": 0.3986703157424927, "learning_rate": 7.384302916001741e-06, "loss": 0.0594, "step": 25636 }, { "epoch": 2.7793798785776236, "grad_norm": 0.18608458340168, "learning_rate": 7.380676048164805e-06, "loss": 0.0051, "step": 25637 }, { "epoch": 2.7794882914137036, "grad_norm": 0.03414109721779823, "learning_rate": 7.3770491803278695e-06, "loss": 0.0009, "step": 25638 }, { "epoch": 2.779596704249783, "grad_norm": 0.0403050072491169, "learning_rate": 7.3734223124909336e-06, "loss": 0.0015, "step": 25639 }, { "epoch": 2.779705117085863, "grad_norm": 0.07876552641391754, "learning_rate": 7.369795444653997e-06, "loss": 0.0029, "step": 25640 }, { "epoch": 2.7798135299219426, "grad_norm": 0.07816990464925766, "learning_rate": 7.36616857681706e-06, "loss": 0.0019, "step": 25641 }, { "epoch": 2.7799219427580226, "grad_norm": 0.5144844651222229, "learning_rate": 7.362541708980125e-06, "loss": 0.0427, "step": 25642 }, { "epoch": 2.7800303555941026, "grad_norm": 0.020473238080739975, "learning_rate": 7.358914841143188e-06, "loss": 0.0009, "step": 25643 }, { "epoch": 2.780138768430182, "grad_norm": 0.05986803025007248, "learning_rate": 7.355287973306253e-06, "loss": 0.0006, "step": 25644 }, { "epoch": 2.7802471812662617, "grad_norm": 0.17761585116386414, "learning_rate": 7.3516611054693165e-06, "loss": 0.0031, "step": 25645 }, { "epoch": 2.7803555941023417, "grad_norm": 0.33347827196121216, "learning_rate": 7.3480342376323814e-06, "loss": 0.0105, "step": 25646 }, { "epoch": 2.7804640069384217, "grad_norm": 0.7152073383331299, "learning_rate": 7.344407369795445e-06, "loss": 0.0091, "step": 25647 }, { "epoch": 2.780572419774501, "grad_norm": 0.39765310287475586, "learning_rate": 7.34078050195851e-06, "loss": 0.0098, "step": 25648 }, { "epoch": 2.780680832610581, "grad_norm": 0.5008273124694824, "learning_rate": 7.337153634121573e-06, "loss": 0.0322, "step": 25649 }, { "epoch": 2.7807892454466607, "grad_norm": 0.05815810710191727, "learning_rate": 7.333526766284637e-06, "loss": 0.0012, "step": 25650 }, { "epoch": 2.7808976582827407, "grad_norm": 0.5419452786445618, "learning_rate": 7.329899898447701e-06, "loss": 0.0222, "step": 25651 }, { "epoch": 2.7810060711188207, "grad_norm": 0.332062304019928, "learning_rate": 7.326273030610765e-06, "loss": 0.0113, "step": 25652 }, { "epoch": 2.7811144839549002, "grad_norm": 0.5663121342658997, "learning_rate": 7.3226461627738285e-06, "loss": 0.0232, "step": 25653 }, { "epoch": 2.78122289679098, "grad_norm": 0.7637326717376709, "learning_rate": 7.3190192949368934e-06, "loss": 0.0305, "step": 25654 }, { "epoch": 2.7813313096270598, "grad_norm": 0.27542710304260254, "learning_rate": 7.315392427099957e-06, "loss": 0.0129, "step": 25655 }, { "epoch": 2.7814397224631398, "grad_norm": 0.13676053285598755, "learning_rate": 7.31176555926302e-06, "loss": 0.0066, "step": 25656 }, { "epoch": 2.7815481352992193, "grad_norm": 0.42859315872192383, "learning_rate": 7.308138691426085e-06, "loss": 0.015, "step": 25657 }, { "epoch": 2.7816565481352993, "grad_norm": 0.3654547929763794, "learning_rate": 7.304511823589148e-06, "loss": 0.0066, "step": 25658 }, { "epoch": 2.781764960971379, "grad_norm": 0.5920817255973816, "learning_rate": 7.300884955752213e-06, "loss": 0.0189, "step": 25659 }, { "epoch": 2.781873373807459, "grad_norm": 0.006265748757869005, "learning_rate": 7.297258087915276e-06, "loss": 0.0003, "step": 25660 }, { "epoch": 2.781981786643539, "grad_norm": 0.5476061701774597, "learning_rate": 7.293631220078341e-06, "loss": 0.023, "step": 25661 }, { "epoch": 2.7820901994796183, "grad_norm": 0.11232728511095047, "learning_rate": 7.290004352241405e-06, "loss": 0.0014, "step": 25662 }, { "epoch": 2.7821986123156983, "grad_norm": 0.3388698101043701, "learning_rate": 7.286377484404469e-06, "loss": 0.0094, "step": 25663 }, { "epoch": 2.782307025151778, "grad_norm": 0.6080227494239807, "learning_rate": 7.282750616567533e-06, "loss": 0.0169, "step": 25664 }, { "epoch": 2.782415437987858, "grad_norm": 0.39745277166366577, "learning_rate": 7.279123748730597e-06, "loss": 0.0177, "step": 25665 }, { "epoch": 2.782523850823938, "grad_norm": 0.19676825404167175, "learning_rate": 7.27549688089366e-06, "loss": 0.008, "step": 25666 }, { "epoch": 2.7826322636600174, "grad_norm": 1.2519134283065796, "learning_rate": 7.271870013056725e-06, "loss": 0.0096, "step": 25667 }, { "epoch": 2.782740676496097, "grad_norm": 0.005370257887989283, "learning_rate": 7.268243145219788e-06, "loss": 0.0002, "step": 25668 }, { "epoch": 2.782849089332177, "grad_norm": 0.10339274257421494, "learning_rate": 7.264616277382852e-06, "loss": 0.0022, "step": 25669 }, { "epoch": 2.782957502168257, "grad_norm": 0.11301480978727341, "learning_rate": 7.260989409545917e-06, "loss": 0.0057, "step": 25670 }, { "epoch": 2.7830659150043364, "grad_norm": 0.28918445110321045, "learning_rate": 7.25736254170898e-06, "loss": 0.0141, "step": 25671 }, { "epoch": 2.7831743278404164, "grad_norm": 0.0199101772159338, "learning_rate": 7.253735673872045e-06, "loss": 0.0005, "step": 25672 }, { "epoch": 2.783282740676496, "grad_norm": 0.3439948260784149, "learning_rate": 7.250108806035108e-06, "loss": 0.0279, "step": 25673 }, { "epoch": 2.783391153512576, "grad_norm": 0.12629632651805878, "learning_rate": 7.246481938198173e-06, "loss": 0.0057, "step": 25674 }, { "epoch": 2.783499566348656, "grad_norm": 0.24103619158267975, "learning_rate": 7.242855070361236e-06, "loss": 0.0164, "step": 25675 }, { "epoch": 2.7836079791847355, "grad_norm": 0.030362091958522797, "learning_rate": 7.2392282025243e-06, "loss": 0.0013, "step": 25676 }, { "epoch": 2.783716392020815, "grad_norm": 0.014319460839033127, "learning_rate": 7.2356013346873645e-06, "loss": 0.0005, "step": 25677 }, { "epoch": 2.783824804856895, "grad_norm": 1.05342435836792, "learning_rate": 7.231974466850429e-06, "loss": 0.0276, "step": 25678 }, { "epoch": 2.783933217692975, "grad_norm": 0.07713748514652252, "learning_rate": 7.228347599013492e-06, "loss": 0.007, "step": 25679 }, { "epoch": 2.7840416305290545, "grad_norm": 0.17379453778266907, "learning_rate": 7.224720731176557e-06, "loss": 0.0051, "step": 25680 }, { "epoch": 2.7841500433651345, "grad_norm": 0.036787860095500946, "learning_rate": 7.22109386333962e-06, "loss": 0.0009, "step": 25681 }, { "epoch": 2.784258456201214, "grad_norm": 0.3040308952331543, "learning_rate": 7.217466995502683e-06, "loss": 0.0065, "step": 25682 }, { "epoch": 2.784366869037294, "grad_norm": 0.05786246806383133, "learning_rate": 7.213840127665748e-06, "loss": 0.0025, "step": 25683 }, { "epoch": 2.784475281873374, "grad_norm": 0.47976139187812805, "learning_rate": 7.2102132598288115e-06, "loss": 0.0087, "step": 25684 }, { "epoch": 2.7845836947094535, "grad_norm": 0.4995425045490265, "learning_rate": 7.2065863919918765e-06, "loss": 0.0551, "step": 25685 }, { "epoch": 2.7846921075455335, "grad_norm": 0.09981177002191544, "learning_rate": 7.20295952415494e-06, "loss": 0.0041, "step": 25686 }, { "epoch": 2.784800520381613, "grad_norm": 1.2458094358444214, "learning_rate": 7.199332656318005e-06, "loss": 0.0181, "step": 25687 }, { "epoch": 2.784908933217693, "grad_norm": 0.19627894461154938, "learning_rate": 7.195705788481068e-06, "loss": 0.0071, "step": 25688 }, { "epoch": 2.7850173460537726, "grad_norm": 0.15736635029315948, "learning_rate": 7.192078920644133e-06, "loss": 0.0072, "step": 25689 }, { "epoch": 2.7851257588898526, "grad_norm": 0.5571046471595764, "learning_rate": 7.188452052807196e-06, "loss": 0.0462, "step": 25690 }, { "epoch": 2.785234171725932, "grad_norm": 0.03283014893531799, "learning_rate": 7.18482518497026e-06, "loss": 0.0012, "step": 25691 }, { "epoch": 2.785342584562012, "grad_norm": 0.5829312205314636, "learning_rate": 7.1811983171333235e-06, "loss": 0.0118, "step": 25692 }, { "epoch": 2.785450997398092, "grad_norm": 0.18416641652584076, "learning_rate": 7.1775714492963885e-06, "loss": 0.0066, "step": 25693 }, { "epoch": 2.7855594102341716, "grad_norm": 0.26841896772384644, "learning_rate": 7.173944581459452e-06, "loss": 0.0032, "step": 25694 }, { "epoch": 2.7856678230702516, "grad_norm": 0.42396193742752075, "learning_rate": 7.170317713622515e-06, "loss": 0.0311, "step": 25695 }, { "epoch": 2.785776235906331, "grad_norm": 0.3708990216255188, "learning_rate": 7.16669084578558e-06, "loss": 0.0192, "step": 25696 }, { "epoch": 2.785884648742411, "grad_norm": 0.003947828430682421, "learning_rate": 7.163063977948643e-06, "loss": 0.0001, "step": 25697 }, { "epoch": 2.785993061578491, "grad_norm": 1.7359576225280762, "learning_rate": 7.159437110111708e-06, "loss": 0.0323, "step": 25698 }, { "epoch": 2.7861014744145707, "grad_norm": 0.02506963163614273, "learning_rate": 7.155810242274771e-06, "loss": 0.0004, "step": 25699 }, { "epoch": 2.78620988725065, "grad_norm": 0.10037415474653244, "learning_rate": 7.152183374437836e-06, "loss": 0.0026, "step": 25700 }, { "epoch": 2.78631830008673, "grad_norm": 0.19727692008018494, "learning_rate": 7.1485565066009e-06, "loss": 0.0097, "step": 25701 }, { "epoch": 2.78642671292281, "grad_norm": 0.07074473053216934, "learning_rate": 7.1449296387639646e-06, "loss": 0.0027, "step": 25702 }, { "epoch": 2.7865351257588897, "grad_norm": 0.6931443810462952, "learning_rate": 7.141302770927028e-06, "loss": 0.0252, "step": 25703 }, { "epoch": 2.7866435385949697, "grad_norm": 0.3450548052787781, "learning_rate": 7.137675903090092e-06, "loss": 0.0178, "step": 25704 }, { "epoch": 2.7867519514310493, "grad_norm": 0.06950698792934418, "learning_rate": 7.134049035253155e-06, "loss": 0.0023, "step": 25705 }, { "epoch": 2.7868603642671292, "grad_norm": 0.028518570587038994, "learning_rate": 7.13042216741622e-06, "loss": 0.002, "step": 25706 }, { "epoch": 2.7869687771032092, "grad_norm": 0.015608563087880611, "learning_rate": 7.126795299579283e-06, "loss": 0.0005, "step": 25707 }, { "epoch": 2.7870771899392888, "grad_norm": 0.016530944034457207, "learning_rate": 7.123168431742348e-06, "loss": 0.0006, "step": 25708 }, { "epoch": 2.7871856027753688, "grad_norm": 0.25247490406036377, "learning_rate": 7.119541563905412e-06, "loss": 0.0465, "step": 25709 }, { "epoch": 2.7872940156114483, "grad_norm": 0.5436049699783325, "learning_rate": 7.115914696068475e-06, "loss": 0.0292, "step": 25710 }, { "epoch": 2.7874024284475283, "grad_norm": 0.46133920550346375, "learning_rate": 7.11228782823154e-06, "loss": 0.0313, "step": 25711 }, { "epoch": 2.787510841283608, "grad_norm": 0.2428770661354065, "learning_rate": 7.108660960394603e-06, "loss": 0.0238, "step": 25712 }, { "epoch": 2.787619254119688, "grad_norm": 0.021088240668177605, "learning_rate": 7.105034092557668e-06, "loss": 0.0008, "step": 25713 }, { "epoch": 2.7877276669557673, "grad_norm": 0.06151896342635155, "learning_rate": 7.101407224720731e-06, "loss": 0.0017, "step": 25714 }, { "epoch": 2.7878360797918473, "grad_norm": 0.0436994731426239, "learning_rate": 7.097780356883796e-06, "loss": 0.0016, "step": 25715 }, { "epoch": 2.7879444926279273, "grad_norm": 0.06751265376806259, "learning_rate": 7.0941534890468595e-06, "loss": 0.0046, "step": 25716 }, { "epoch": 2.788052905464007, "grad_norm": 0.2087819129228592, "learning_rate": 7.090526621209924e-06, "loss": 0.0101, "step": 25717 }, { "epoch": 2.788161318300087, "grad_norm": 0.0474126860499382, "learning_rate": 7.086899753372987e-06, "loss": 0.001, "step": 25718 }, { "epoch": 2.7882697311361664, "grad_norm": 0.8202502131462097, "learning_rate": 7.083272885536052e-06, "loss": 0.0509, "step": 25719 }, { "epoch": 2.7883781439722464, "grad_norm": 0.3370038568973541, "learning_rate": 7.079646017699115e-06, "loss": 0.0112, "step": 25720 }, { "epoch": 2.7884865568083264, "grad_norm": 0.6126031875610352, "learning_rate": 7.07601914986218e-06, "loss": 0.0131, "step": 25721 }, { "epoch": 2.788594969644406, "grad_norm": 0.25456658005714417, "learning_rate": 7.072392282025243e-06, "loss": 0.0045, "step": 25722 }, { "epoch": 2.7887033824804854, "grad_norm": 0.017743831500411034, "learning_rate": 7.0687654141883065e-06, "loss": 0.0008, "step": 25723 }, { "epoch": 2.7888117953165654, "grad_norm": 0.43332037329673767, "learning_rate": 7.0651385463513715e-06, "loss": 0.0736, "step": 25724 }, { "epoch": 2.7889202081526454, "grad_norm": 0.19773834943771362, "learning_rate": 7.061511678514435e-06, "loss": 0.0109, "step": 25725 }, { "epoch": 2.789028620988725, "grad_norm": 0.2199353575706482, "learning_rate": 7.0578848106775e-06, "loss": 0.0083, "step": 25726 }, { "epoch": 2.789137033824805, "grad_norm": 0.050437990576028824, "learning_rate": 7.054257942840563e-06, "loss": 0.002, "step": 25727 }, { "epoch": 2.7892454466608845, "grad_norm": 0.07845814526081085, "learning_rate": 7.050631075003628e-06, "loss": 0.0008, "step": 25728 }, { "epoch": 2.7893538594969645, "grad_norm": 0.1352112889289856, "learning_rate": 7.047004207166691e-06, "loss": 0.0236, "step": 25729 }, { "epoch": 2.7894622723330444, "grad_norm": 0.057094793766736984, "learning_rate": 7.043377339329755e-06, "loss": 0.0008, "step": 25730 }, { "epoch": 2.789570685169124, "grad_norm": 0.2736172080039978, "learning_rate": 7.0397504714928185e-06, "loss": 0.025, "step": 25731 }, { "epoch": 2.789679098005204, "grad_norm": 0.03755683824419975, "learning_rate": 7.0361236036558835e-06, "loss": 0.0015, "step": 25732 }, { "epoch": 2.7897875108412835, "grad_norm": 0.18474583327770233, "learning_rate": 7.032496735818947e-06, "loss": 0.0084, "step": 25733 }, { "epoch": 2.7898959236773635, "grad_norm": 0.11877470463514328, "learning_rate": 7.028869867982012e-06, "loss": 0.0072, "step": 25734 }, { "epoch": 2.790004336513443, "grad_norm": 0.10130979865789413, "learning_rate": 7.025243000145075e-06, "loss": 0.0022, "step": 25735 }, { "epoch": 2.790112749349523, "grad_norm": 0.2783637046813965, "learning_rate": 7.021616132308138e-06, "loss": 0.0138, "step": 25736 }, { "epoch": 2.7902211621856026, "grad_norm": 0.3871459662914276, "learning_rate": 7.017989264471203e-06, "loss": 0.0054, "step": 25737 }, { "epoch": 2.7903295750216826, "grad_norm": 0.14669480919837952, "learning_rate": 7.014362396634266e-06, "loss": 0.0038, "step": 25738 }, { "epoch": 2.7904379878577625, "grad_norm": 0.7701541185379028, "learning_rate": 7.010735528797331e-06, "loss": 0.038, "step": 25739 }, { "epoch": 2.790546400693842, "grad_norm": 0.07078231871128082, "learning_rate": 7.007108660960395e-06, "loss": 0.005, "step": 25740 }, { "epoch": 2.790654813529922, "grad_norm": 0.22026926279067993, "learning_rate": 7.0034817931234596e-06, "loss": 0.0136, "step": 25741 }, { "epoch": 2.7907632263660016, "grad_norm": 0.08402137458324432, "learning_rate": 6.999854925286523e-06, "loss": 0.0029, "step": 25742 }, { "epoch": 2.7908716392020816, "grad_norm": 0.41031768918037415, "learning_rate": 6.996228057449587e-06, "loss": 0.0291, "step": 25743 }, { "epoch": 2.7909800520381616, "grad_norm": 0.17615023255348206, "learning_rate": 6.99260118961265e-06, "loss": 0.0083, "step": 25744 }, { "epoch": 2.791088464874241, "grad_norm": 0.02703823335468769, "learning_rate": 6.988974321775715e-06, "loss": 0.0004, "step": 25745 }, { "epoch": 2.7911968777103207, "grad_norm": 0.01219170168042183, "learning_rate": 6.985347453938778e-06, "loss": 0.0003, "step": 25746 }, { "epoch": 2.7913052905464006, "grad_norm": 0.22963422536849976, "learning_rate": 6.981720586101843e-06, "loss": 0.0139, "step": 25747 }, { "epoch": 2.7914137033824806, "grad_norm": 0.08635087311267853, "learning_rate": 6.978093718264907e-06, "loss": 0.0018, "step": 25748 }, { "epoch": 2.79152211621856, "grad_norm": 0.10406351089477539, "learning_rate": 6.9744668504279716e-06, "loss": 0.0028, "step": 25749 }, { "epoch": 2.79163052905464, "grad_norm": 0.0400237999856472, "learning_rate": 6.970839982591035e-06, "loss": 0.0008, "step": 25750 }, { "epoch": 2.7917389418907197, "grad_norm": 0.20389124751091003, "learning_rate": 6.967213114754098e-06, "loss": 0.0083, "step": 25751 }, { "epoch": 2.7918473547267997, "grad_norm": 0.36619246006011963, "learning_rate": 6.963586246917163e-06, "loss": 0.0176, "step": 25752 }, { "epoch": 2.7919557675628797, "grad_norm": 0.49723705649375916, "learning_rate": 6.959959379080226e-06, "loss": 0.0149, "step": 25753 }, { "epoch": 2.792064180398959, "grad_norm": 0.44050970673561096, "learning_rate": 6.956332511243291e-06, "loss": 0.0109, "step": 25754 }, { "epoch": 2.792172593235039, "grad_norm": 0.5280941724777222, "learning_rate": 6.9527056434063545e-06, "loss": 0.0071, "step": 25755 }, { "epoch": 2.7922810060711187, "grad_norm": 0.1445249319076538, "learning_rate": 6.949078775569419e-06, "loss": 0.0045, "step": 25756 }, { "epoch": 2.7923894189071987, "grad_norm": 0.5757155418395996, "learning_rate": 6.945451907732482e-06, "loss": 0.0145, "step": 25757 }, { "epoch": 2.7924978317432783, "grad_norm": 0.2514324188232422, "learning_rate": 6.941825039895547e-06, "loss": 0.0071, "step": 25758 }, { "epoch": 2.7926062445793582, "grad_norm": 0.038970302790403366, "learning_rate": 6.93819817205861e-06, "loss": 0.0012, "step": 25759 }, { "epoch": 2.792714657415438, "grad_norm": 0.47304558753967285, "learning_rate": 6.934571304221675e-06, "loss": 0.0119, "step": 25760 }, { "epoch": 2.7928230702515178, "grad_norm": 0.09902982413768768, "learning_rate": 6.930944436384738e-06, "loss": 0.0035, "step": 25761 }, { "epoch": 2.7929314830875978, "grad_norm": 0.6223418712615967, "learning_rate": 6.927317568547803e-06, "loss": 0.0384, "step": 25762 }, { "epoch": 2.7930398959236773, "grad_norm": 0.12595179677009583, "learning_rate": 6.9236907007108665e-06, "loss": 0.0043, "step": 25763 }, { "epoch": 2.7931483087597573, "grad_norm": 0.21087713539600372, "learning_rate": 6.92006383287393e-06, "loss": 0.0038, "step": 25764 }, { "epoch": 2.793256721595837, "grad_norm": 0.3490905463695526, "learning_rate": 6.916436965036995e-06, "loss": 0.0056, "step": 25765 }, { "epoch": 2.793365134431917, "grad_norm": 0.624508261680603, "learning_rate": 6.912810097200058e-06, "loss": 0.0608, "step": 25766 }, { "epoch": 2.793473547267997, "grad_norm": 0.01230335421860218, "learning_rate": 6.909183229363123e-06, "loss": 0.0006, "step": 25767 }, { "epoch": 2.7935819601040763, "grad_norm": 0.4233675003051758, "learning_rate": 6.905556361526186e-06, "loss": 0.0114, "step": 25768 }, { "epoch": 2.793690372940156, "grad_norm": 0.18905647099018097, "learning_rate": 6.90192949368925e-06, "loss": 0.0229, "step": 25769 }, { "epoch": 2.793798785776236, "grad_norm": 0.12281285971403122, "learning_rate": 6.8983026258523135e-06, "loss": 0.0031, "step": 25770 }, { "epoch": 2.793907198612316, "grad_norm": 0.2480817437171936, "learning_rate": 6.8946757580153785e-06, "loss": 0.01, "step": 25771 }, { "epoch": 2.7940156114483954, "grad_norm": 0.2978716790676117, "learning_rate": 6.891048890178442e-06, "loss": 0.0142, "step": 25772 }, { "epoch": 2.7941240242844754, "grad_norm": 0.3524934649467468, "learning_rate": 6.887422022341507e-06, "loss": 0.0313, "step": 25773 }, { "epoch": 2.794232437120555, "grad_norm": 0.24407291412353516, "learning_rate": 6.88379515450457e-06, "loss": 0.0133, "step": 25774 }, { "epoch": 2.794340849956635, "grad_norm": 0.2799128592014313, "learning_rate": 6.880168286667635e-06, "loss": 0.009, "step": 25775 }, { "epoch": 2.794449262792715, "grad_norm": 0.26892659068107605, "learning_rate": 6.876541418830698e-06, "loss": 0.0068, "step": 25776 }, { "epoch": 2.7945576756287944, "grad_norm": 0.02489987388253212, "learning_rate": 6.8729145509937614e-06, "loss": 0.001, "step": 25777 }, { "epoch": 2.794666088464874, "grad_norm": 0.11375366896390915, "learning_rate": 6.869287683156826e-06, "loss": 0.0015, "step": 25778 }, { "epoch": 2.794774501300954, "grad_norm": 0.16915534436702728, "learning_rate": 6.86566081531989e-06, "loss": 0.0081, "step": 25779 }, { "epoch": 2.794882914137034, "grad_norm": 0.1493707299232483, "learning_rate": 6.862033947482955e-06, "loss": 0.0047, "step": 25780 }, { "epoch": 2.7949913269731135, "grad_norm": 0.49674978852272034, "learning_rate": 6.858407079646018e-06, "loss": 0.035, "step": 25781 }, { "epoch": 2.7950997398091935, "grad_norm": 0.03680987283587456, "learning_rate": 6.854780211809082e-06, "loss": 0.0012, "step": 25782 }, { "epoch": 2.795208152645273, "grad_norm": 0.05341963842511177, "learning_rate": 6.851153343972146e-06, "loss": 0.0018, "step": 25783 }, { "epoch": 2.795316565481353, "grad_norm": 0.26997092366218567, "learning_rate": 6.84752647613521e-06, "loss": 0.0113, "step": 25784 }, { "epoch": 2.795424978317433, "grad_norm": 0.2790471613407135, "learning_rate": 6.843899608298273e-06, "loss": 0.0593, "step": 25785 }, { "epoch": 2.7955333911535125, "grad_norm": 0.14611463248729706, "learning_rate": 6.840272740461338e-06, "loss": 0.0033, "step": 25786 }, { "epoch": 2.7956418039895925, "grad_norm": 0.16766640543937683, "learning_rate": 6.836645872624402e-06, "loss": 0.0059, "step": 25787 }, { "epoch": 2.795750216825672, "grad_norm": 0.14659735560417175, "learning_rate": 6.833019004787467e-06, "loss": 0.0049, "step": 25788 }, { "epoch": 2.795858629661752, "grad_norm": 0.7064833641052246, "learning_rate": 6.82939213695053e-06, "loss": 0.0208, "step": 25789 }, { "epoch": 2.795967042497832, "grad_norm": 0.09353364259004593, "learning_rate": 6.825765269113593e-06, "loss": 0.0026, "step": 25790 }, { "epoch": 2.7960754553339116, "grad_norm": 0.019451815634965897, "learning_rate": 6.822138401276658e-06, "loss": 0.0005, "step": 25791 }, { "epoch": 2.796183868169991, "grad_norm": 0.22293627262115479, "learning_rate": 6.818511533439721e-06, "loss": 0.0092, "step": 25792 }, { "epoch": 2.796292281006071, "grad_norm": 0.18710337579250336, "learning_rate": 6.814884665602786e-06, "loss": 0.0191, "step": 25793 }, { "epoch": 2.796400693842151, "grad_norm": 0.20583124458789825, "learning_rate": 6.8112577977658495e-06, "loss": 0.0072, "step": 25794 }, { "epoch": 2.7965091066782306, "grad_norm": 0.054098714143037796, "learning_rate": 6.807630929928914e-06, "loss": 0.0017, "step": 25795 }, { "epoch": 2.7966175195143106, "grad_norm": 0.00824024062603712, "learning_rate": 6.804004062091978e-06, "loss": 0.0003, "step": 25796 }, { "epoch": 2.79672593235039, "grad_norm": 0.002838611137121916, "learning_rate": 6.800377194255042e-06, "loss": 0.0001, "step": 25797 }, { "epoch": 2.79683434518647, "grad_norm": 0.23219609260559082, "learning_rate": 6.796750326418105e-06, "loss": 0.0108, "step": 25798 }, { "epoch": 2.79694275802255, "grad_norm": 0.18906569480895996, "learning_rate": 6.79312345858117e-06, "loss": 0.0088, "step": 25799 }, { "epoch": 2.7970511708586296, "grad_norm": 0.1294875144958496, "learning_rate": 6.789496590744233e-06, "loss": 0.0039, "step": 25800 }, { "epoch": 2.797159583694709, "grad_norm": 0.05340808629989624, "learning_rate": 6.785869722907298e-06, "loss": 0.001, "step": 25801 }, { "epoch": 2.797267996530789, "grad_norm": 0.05793498083949089, "learning_rate": 6.7822428550703615e-06, "loss": 0.0019, "step": 25802 }, { "epoch": 2.797376409366869, "grad_norm": 0.044097673147916794, "learning_rate": 6.7786159872334265e-06, "loss": 0.0018, "step": 25803 }, { "epoch": 2.7974848222029487, "grad_norm": 0.02256731688976288, "learning_rate": 6.77498911939649e-06, "loss": 0.0005, "step": 25804 }, { "epoch": 2.7975932350390287, "grad_norm": 0.3754621148109436, "learning_rate": 6.771362251559553e-06, "loss": 0.0234, "step": 25805 }, { "epoch": 2.797701647875108, "grad_norm": 0.5887630581855774, "learning_rate": 6.767735383722618e-06, "loss": 0.0125, "step": 25806 }, { "epoch": 2.797810060711188, "grad_norm": 0.05315375328063965, "learning_rate": 6.764108515885681e-06, "loss": 0.0012, "step": 25807 }, { "epoch": 2.797918473547268, "grad_norm": 0.18337653577327728, "learning_rate": 6.760481648048745e-06, "loss": 0.0086, "step": 25808 }, { "epoch": 2.7980268863833477, "grad_norm": 0.36029842495918274, "learning_rate": 6.756854780211809e-06, "loss": 0.0198, "step": 25809 }, { "epoch": 2.7981352992194277, "grad_norm": 0.2992214262485504, "learning_rate": 6.7532279123748735e-06, "loss": 0.0059, "step": 25810 }, { "epoch": 2.7982437120555073, "grad_norm": 0.007240333128720522, "learning_rate": 6.749601044537937e-06, "loss": 0.0005, "step": 25811 }, { "epoch": 2.7983521248915872, "grad_norm": 0.07723325490951538, "learning_rate": 6.745974176701002e-06, "loss": 0.0024, "step": 25812 }, { "epoch": 2.7984605377276672, "grad_norm": 0.6686266660690308, "learning_rate": 6.742347308864065e-06, "loss": 0.0256, "step": 25813 }, { "epoch": 2.7985689505637468, "grad_norm": 1.0686651468276978, "learning_rate": 6.73872044102713e-06, "loss": 0.0388, "step": 25814 }, { "epoch": 2.7986773633998263, "grad_norm": 0.2373836636543274, "learning_rate": 6.735093573190193e-06, "loss": 0.0148, "step": 25815 }, { "epoch": 2.7987857762359063, "grad_norm": 0.6027774214744568, "learning_rate": 6.731466705353258e-06, "loss": 0.0262, "step": 25816 }, { "epoch": 2.7988941890719863, "grad_norm": 0.054266348481178284, "learning_rate": 6.727839837516321e-06, "loss": 0.0031, "step": 25817 }, { "epoch": 2.799002601908066, "grad_norm": 0.15764093399047852, "learning_rate": 6.724212969679385e-06, "loss": 0.0027, "step": 25818 }, { "epoch": 2.799111014744146, "grad_norm": 0.1638413965702057, "learning_rate": 6.72058610184245e-06, "loss": 0.0114, "step": 25819 }, { "epoch": 2.7992194275802254, "grad_norm": 0.1788637787103653, "learning_rate": 6.716959234005513e-06, "loss": 0.007, "step": 25820 }, { "epoch": 2.7993278404163053, "grad_norm": 0.42030462622642517, "learning_rate": 6.713332366168577e-06, "loss": 0.0221, "step": 25821 }, { "epoch": 2.7994362532523853, "grad_norm": 0.5277817845344543, "learning_rate": 6.709705498331641e-06, "loss": 0.045, "step": 25822 }, { "epoch": 2.799544666088465, "grad_norm": 0.04608068987727165, "learning_rate": 6.706078630494705e-06, "loss": 0.0018, "step": 25823 }, { "epoch": 2.7996530789245444, "grad_norm": 0.01855475641787052, "learning_rate": 6.7024517626577684e-06, "loss": 0.0009, "step": 25824 }, { "epoch": 2.7997614917606244, "grad_norm": 0.4023604094982147, "learning_rate": 6.698824894820833e-06, "loss": 0.0058, "step": 25825 }, { "epoch": 2.7998699045967044, "grad_norm": 0.05233156308531761, "learning_rate": 6.695198026983897e-06, "loss": 0.0029, "step": 25826 }, { "epoch": 2.799978317432784, "grad_norm": 0.03866267949342728, "learning_rate": 6.691571159146962e-06, "loss": 0.0019, "step": 25827 }, { "epoch": 2.800086730268864, "grad_norm": 0.019175328314304352, "learning_rate": 6.687944291310025e-06, "loss": 0.0007, "step": 25828 }, { "epoch": 2.8001951431049434, "grad_norm": 0.12028539925813675, "learning_rate": 6.68431742347309e-06, "loss": 0.0051, "step": 25829 }, { "epoch": 2.8003035559410234, "grad_norm": 0.13950666785240173, "learning_rate": 6.680690555636153e-06, "loss": 0.0044, "step": 25830 }, { "epoch": 2.8004119687771034, "grad_norm": 0.12068232148885727, "learning_rate": 6.677063687799216e-06, "loss": 0.0045, "step": 25831 }, { "epoch": 2.800520381613183, "grad_norm": 0.6452484130859375, "learning_rate": 6.673436819962281e-06, "loss": 0.0227, "step": 25832 }, { "epoch": 2.800628794449263, "grad_norm": 0.1983538269996643, "learning_rate": 6.6698099521253445e-06, "loss": 0.0102, "step": 25833 }, { "epoch": 2.8007372072853425, "grad_norm": 0.4605609178543091, "learning_rate": 6.666183084288409e-06, "loss": 0.0433, "step": 25834 }, { "epoch": 2.8008456201214225, "grad_norm": 0.06218094378709793, "learning_rate": 6.662556216451473e-06, "loss": 0.0037, "step": 25835 }, { "epoch": 2.800954032957502, "grad_norm": 0.11187535524368286, "learning_rate": 6.658929348614537e-06, "loss": 0.0047, "step": 25836 }, { "epoch": 2.801062445793582, "grad_norm": 0.2605136036872864, "learning_rate": 6.6553024807776e-06, "loss": 0.0452, "step": 25837 }, { "epoch": 2.8011708586296615, "grad_norm": 0.5547738075256348, "learning_rate": 6.651675612940665e-06, "loss": 0.0247, "step": 25838 }, { "epoch": 2.8012792714657415, "grad_norm": 0.3161798119544983, "learning_rate": 6.648048745103728e-06, "loss": 0.0032, "step": 25839 }, { "epoch": 2.8013876843018215, "grad_norm": 0.13423867523670197, "learning_rate": 6.644421877266793e-06, "loss": 0.006, "step": 25840 }, { "epoch": 2.801496097137901, "grad_norm": 0.16745781898498535, "learning_rate": 6.6407950094298565e-06, "loss": 0.0091, "step": 25841 }, { "epoch": 2.801604509973981, "grad_norm": 0.07165201008319855, "learning_rate": 6.6371681415929215e-06, "loss": 0.0023, "step": 25842 }, { "epoch": 2.8017129228100606, "grad_norm": 0.052362676709890366, "learning_rate": 6.633541273755985e-06, "loss": 0.0023, "step": 25843 }, { "epoch": 2.8018213356461406, "grad_norm": 0.379268079996109, "learning_rate": 6.629914405919048e-06, "loss": 0.0105, "step": 25844 }, { "epoch": 2.8019297484822205, "grad_norm": 0.679625391960144, "learning_rate": 6.626287538082113e-06, "loss": 0.0073, "step": 25845 }, { "epoch": 2.8020381613183, "grad_norm": 0.028506606817245483, "learning_rate": 6.622660670245176e-06, "loss": 0.0006, "step": 25846 }, { "epoch": 2.8021465741543796, "grad_norm": 0.295123428106308, "learning_rate": 6.61903380240824e-06, "loss": 0.0146, "step": 25847 }, { "epoch": 2.8022549869904596, "grad_norm": 0.20141714811325073, "learning_rate": 6.615406934571304e-06, "loss": 0.017, "step": 25848 }, { "epoch": 2.8023633998265396, "grad_norm": 1.4318727254867554, "learning_rate": 6.6117800667343685e-06, "loss": 0.0203, "step": 25849 }, { "epoch": 2.802471812662619, "grad_norm": 0.07067222893238068, "learning_rate": 6.608153198897432e-06, "loss": 0.0025, "step": 25850 }, { "epoch": 2.802580225498699, "grad_norm": 0.13915106654167175, "learning_rate": 6.604526331060497e-06, "loss": 0.0109, "step": 25851 }, { "epoch": 2.8026886383347787, "grad_norm": 0.42166125774383545, "learning_rate": 6.60089946322356e-06, "loss": 0.0336, "step": 25852 }, { "epoch": 2.8027970511708586, "grad_norm": 0.015294468030333519, "learning_rate": 6.597272595386625e-06, "loss": 0.0004, "step": 25853 }, { "epoch": 2.8029054640069386, "grad_norm": 0.4489971697330475, "learning_rate": 6.593645727549688e-06, "loss": 0.0256, "step": 25854 }, { "epoch": 2.803013876843018, "grad_norm": 0.08550719916820526, "learning_rate": 6.590018859712753e-06, "loss": 0.0037, "step": 25855 }, { "epoch": 2.803122289679098, "grad_norm": 0.3706221282482147, "learning_rate": 6.586391991875816e-06, "loss": 0.0091, "step": 25856 }, { "epoch": 2.8032307025151777, "grad_norm": 0.030814338475465775, "learning_rate": 6.582765124038881e-06, "loss": 0.0012, "step": 25857 }, { "epoch": 2.8033391153512577, "grad_norm": 0.23188796639442444, "learning_rate": 6.579138256201945e-06, "loss": 0.0174, "step": 25858 }, { "epoch": 2.8034475281873372, "grad_norm": 0.05604196712374687, "learning_rate": 6.575511388365008e-06, "loss": 0.0022, "step": 25859 }, { "epoch": 2.803555941023417, "grad_norm": 0.004714311100542545, "learning_rate": 6.571884520528072e-06, "loss": 0.0002, "step": 25860 }, { "epoch": 2.8036643538594967, "grad_norm": 0.09110435843467712, "learning_rate": 6.568257652691136e-06, "loss": 0.0019, "step": 25861 }, { "epoch": 2.8037727666955767, "grad_norm": 0.2707395851612091, "learning_rate": 6.5646307848542e-06, "loss": 0.0212, "step": 25862 }, { "epoch": 2.8038811795316567, "grad_norm": 0.2930934727191925, "learning_rate": 6.5610039170172635e-06, "loss": 0.0108, "step": 25863 }, { "epoch": 2.8039895923677363, "grad_norm": 0.4126232862472534, "learning_rate": 6.557377049180328e-06, "loss": 0.0072, "step": 25864 }, { "epoch": 2.8040980052038162, "grad_norm": 0.003740283427760005, "learning_rate": 6.553750181343392e-06, "loss": 0.0001, "step": 25865 }, { "epoch": 2.804206418039896, "grad_norm": 0.11298093944787979, "learning_rate": 6.550123313506457e-06, "loss": 0.0027, "step": 25866 }, { "epoch": 2.8043148308759758, "grad_norm": 0.22530575096607208, "learning_rate": 6.54649644566952e-06, "loss": 0.0078, "step": 25867 }, { "epoch": 2.8044232437120558, "grad_norm": 0.25120580196380615, "learning_rate": 6.542869577832585e-06, "loss": 0.0237, "step": 25868 }, { "epoch": 2.8045316565481353, "grad_norm": 0.0636233389377594, "learning_rate": 6.539242709995648e-06, "loss": 0.0018, "step": 25869 }, { "epoch": 2.804640069384215, "grad_norm": 0.2653357684612274, "learning_rate": 6.535615842158713e-06, "loss": 0.0134, "step": 25870 }, { "epoch": 2.804748482220295, "grad_norm": 0.2734096944332123, "learning_rate": 6.531988974321776e-06, "loss": 0.0055, "step": 25871 }, { "epoch": 2.804856895056375, "grad_norm": 0.05973652005195618, "learning_rate": 6.5283621064848396e-06, "loss": 0.002, "step": 25872 }, { "epoch": 2.8049653078924544, "grad_norm": 0.19424313306808472, "learning_rate": 6.5247352386479045e-06, "loss": 0.0036, "step": 25873 }, { "epoch": 2.8050737207285343, "grad_norm": 0.3754294514656067, "learning_rate": 6.521108370810968e-06, "loss": 0.0272, "step": 25874 }, { "epoch": 2.805182133564614, "grad_norm": 0.15405774116516113, "learning_rate": 6.517481502974032e-06, "loss": 0.0052, "step": 25875 }, { "epoch": 2.805290546400694, "grad_norm": 0.6706696152687073, "learning_rate": 6.513854635137095e-06, "loss": 0.0139, "step": 25876 }, { "epoch": 2.805398959236774, "grad_norm": 0.05168876796960831, "learning_rate": 6.51022776730016e-06, "loss": 0.001, "step": 25877 }, { "epoch": 2.8055073720728534, "grad_norm": 0.02816765382885933, "learning_rate": 6.506600899463223e-06, "loss": 0.001, "step": 25878 }, { "epoch": 2.8056157849089334, "grad_norm": 0.6717432141304016, "learning_rate": 6.502974031626288e-06, "loss": 0.0163, "step": 25879 }, { "epoch": 2.805724197745013, "grad_norm": 0.02181178890168667, "learning_rate": 6.4993471637893515e-06, "loss": 0.001, "step": 25880 }, { "epoch": 2.805832610581093, "grad_norm": 0.03176309913396835, "learning_rate": 6.4957202959524165e-06, "loss": 0.0012, "step": 25881 }, { "epoch": 2.8059410234171724, "grad_norm": 0.4482181966304779, "learning_rate": 6.49209342811548e-06, "loss": 0.0234, "step": 25882 }, { "epoch": 2.8060494362532524, "grad_norm": 1.1628245115280151, "learning_rate": 6.488466560278545e-06, "loss": 0.0231, "step": 25883 }, { "epoch": 2.806157849089332, "grad_norm": 0.11039596050977707, "learning_rate": 6.484839692441608e-06, "loss": 0.0031, "step": 25884 }, { "epoch": 2.806266261925412, "grad_norm": 0.0745917335152626, "learning_rate": 6.481212824604671e-06, "loss": 0.0016, "step": 25885 }, { "epoch": 2.806374674761492, "grad_norm": 0.07649863511323929, "learning_rate": 6.477585956767736e-06, "loss": 0.0021, "step": 25886 }, { "epoch": 2.8064830875975715, "grad_norm": 0.021806711331009865, "learning_rate": 6.4739590889307994e-06, "loss": 0.0009, "step": 25887 }, { "epoch": 2.8065915004336515, "grad_norm": 0.14233361184597015, "learning_rate": 6.4703322210938635e-06, "loss": 0.0038, "step": 25888 }, { "epoch": 2.806699913269731, "grad_norm": 0.29494839906692505, "learning_rate": 6.466705353256927e-06, "loss": 0.0115, "step": 25889 }, { "epoch": 2.806808326105811, "grad_norm": 0.07111767679452896, "learning_rate": 6.463078485419992e-06, "loss": 0.0023, "step": 25890 }, { "epoch": 2.806916738941891, "grad_norm": 0.08961372822523117, "learning_rate": 6.459451617583055e-06, "loss": 0.0023, "step": 25891 }, { "epoch": 2.8070251517779705, "grad_norm": 0.577144205570221, "learning_rate": 6.45582474974612e-06, "loss": 0.0241, "step": 25892 }, { "epoch": 2.80713356461405, "grad_norm": 0.061721399426460266, "learning_rate": 6.452197881909183e-06, "loss": 0.0028, "step": 25893 }, { "epoch": 2.80724197745013, "grad_norm": 0.49603280425071716, "learning_rate": 6.448571014072248e-06, "loss": 0.0227, "step": 25894 }, { "epoch": 2.80735039028621, "grad_norm": 0.24931398034095764, "learning_rate": 6.4449441462353114e-06, "loss": 0.0118, "step": 25895 }, { "epoch": 2.8074588031222896, "grad_norm": 0.029830677434802055, "learning_rate": 6.441317278398376e-06, "loss": 0.0013, "step": 25896 }, { "epoch": 2.8075672159583696, "grad_norm": 0.09223991632461548, "learning_rate": 6.43769041056144e-06, "loss": 0.0048, "step": 25897 }, { "epoch": 2.807675628794449, "grad_norm": 0.27118971943855286, "learning_rate": 6.434063542724503e-06, "loss": 0.0181, "step": 25898 }, { "epoch": 2.807784041630529, "grad_norm": 0.06041555479168892, "learning_rate": 6.430436674887568e-06, "loss": 0.0025, "step": 25899 }, { "epoch": 2.807892454466609, "grad_norm": 0.8888362050056458, "learning_rate": 6.426809807050631e-06, "loss": 0.0148, "step": 25900 }, { "epoch": 2.8080008673026886, "grad_norm": 0.09217444062232971, "learning_rate": 6.423182939213695e-06, "loss": 0.0023, "step": 25901 }, { "epoch": 2.8081092801387686, "grad_norm": 0.28281912207603455, "learning_rate": 6.4195560713767585e-06, "loss": 0.0062, "step": 25902 }, { "epoch": 2.808217692974848, "grad_norm": 0.1350410282611847, "learning_rate": 6.415929203539823e-06, "loss": 0.0059, "step": 25903 }, { "epoch": 2.808326105810928, "grad_norm": 0.09387747943401337, "learning_rate": 6.412302335702887e-06, "loss": 0.003, "step": 25904 }, { "epoch": 2.8084345186470077, "grad_norm": 0.027760127559304237, "learning_rate": 6.408675467865952e-06, "loss": 0.0013, "step": 25905 }, { "epoch": 2.8085429314830876, "grad_norm": 0.5019604563713074, "learning_rate": 6.405048600029015e-06, "loss": 0.0493, "step": 25906 }, { "epoch": 2.808651344319167, "grad_norm": 0.20545469224452972, "learning_rate": 6.40142173219208e-06, "loss": 0.0095, "step": 25907 }, { "epoch": 2.808759757155247, "grad_norm": 0.062790647149086, "learning_rate": 6.397794864355143e-06, "loss": 0.0023, "step": 25908 }, { "epoch": 2.808868169991327, "grad_norm": 0.13641788065433502, "learning_rate": 6.394167996518208e-06, "loss": 0.0528, "step": 25909 }, { "epoch": 2.8089765828274067, "grad_norm": 0.11379081755876541, "learning_rate": 6.390541128681271e-06, "loss": 0.0041, "step": 25910 }, { "epoch": 2.8090849956634867, "grad_norm": 0.8303281664848328, "learning_rate": 6.386914260844335e-06, "loss": 0.0397, "step": 25911 }, { "epoch": 2.8091934084995662, "grad_norm": 0.2808750867843628, "learning_rate": 6.3832873930073995e-06, "loss": 0.0183, "step": 25912 }, { "epoch": 2.809301821335646, "grad_norm": 0.9812910556793213, "learning_rate": 6.379660525170463e-06, "loss": 0.0174, "step": 25913 }, { "epoch": 2.809410234171726, "grad_norm": 0.38221976161003113, "learning_rate": 6.376033657333527e-06, "loss": 0.002, "step": 25914 }, { "epoch": 2.8095186470078057, "grad_norm": 0.3075401484966278, "learning_rate": 6.37240678949659e-06, "loss": 0.0051, "step": 25915 }, { "epoch": 2.8096270598438853, "grad_norm": 0.04299464821815491, "learning_rate": 6.368779921659655e-06, "loss": 0.0013, "step": 25916 }, { "epoch": 2.8097354726799653, "grad_norm": 2.0543267726898193, "learning_rate": 6.365153053822718e-06, "loss": 0.016, "step": 25917 }, { "epoch": 2.8098438855160452, "grad_norm": 0.24877673387527466, "learning_rate": 6.361526185985783e-06, "loss": 0.0195, "step": 25918 }, { "epoch": 2.809952298352125, "grad_norm": 3.8839974403381348, "learning_rate": 6.3578993181488466e-06, "loss": 0.0288, "step": 25919 }, { "epoch": 2.8100607111882048, "grad_norm": 1.2062193155288696, "learning_rate": 6.3542724503119115e-06, "loss": 0.0147, "step": 25920 }, { "epoch": 2.8101691240242843, "grad_norm": 0.34124284982681274, "learning_rate": 6.350645582474975e-06, "loss": 0.0131, "step": 25921 }, { "epoch": 2.8102775368603643, "grad_norm": 0.12139873206615448, "learning_rate": 6.34701871463804e-06, "loss": 0.0078, "step": 25922 }, { "epoch": 2.8103859496964443, "grad_norm": 0.16014069318771362, "learning_rate": 6.343391846801103e-06, "loss": 0.0055, "step": 25923 }, { "epoch": 2.810494362532524, "grad_norm": 0.012572226114571095, "learning_rate": 6.339764978964167e-06, "loss": 0.0005, "step": 25924 }, { "epoch": 2.8106027753686034, "grad_norm": 0.5708527565002441, "learning_rate": 6.336138111127231e-06, "loss": 0.0357, "step": 25925 }, { "epoch": 2.8107111882046834, "grad_norm": 0.2241598218679428, "learning_rate": 6.3325112432902945e-06, "loss": 0.0145, "step": 25926 }, { "epoch": 2.8108196010407633, "grad_norm": 0.09331457316875458, "learning_rate": 6.3288843754533586e-06, "loss": 0.0019, "step": 25927 }, { "epoch": 2.810928013876843, "grad_norm": 0.33331653475761414, "learning_rate": 6.325257507616422e-06, "loss": 0.0415, "step": 25928 }, { "epoch": 2.811036426712923, "grad_norm": 0.5911080837249756, "learning_rate": 6.321630639779487e-06, "loss": 0.0456, "step": 25929 }, { "epoch": 2.8111448395490024, "grad_norm": 0.04744995757937431, "learning_rate": 6.31800377194255e-06, "loss": 0.0017, "step": 25930 }, { "epoch": 2.8112532523850824, "grad_norm": 0.15357081592082977, "learning_rate": 6.314376904105615e-06, "loss": 0.009, "step": 25931 }, { "epoch": 2.8113616652211624, "grad_norm": 0.06895767152309418, "learning_rate": 6.310750036268678e-06, "loss": 0.0021, "step": 25932 }, { "epoch": 2.811470078057242, "grad_norm": 1.5073262453079224, "learning_rate": 6.307123168431743e-06, "loss": 0.0509, "step": 25933 }, { "epoch": 2.811578490893322, "grad_norm": 0.23491054773330688, "learning_rate": 6.3034963005948064e-06, "loss": 0.0046, "step": 25934 }, { "epoch": 2.8116869037294014, "grad_norm": 0.17575833201408386, "learning_rate": 6.299869432757871e-06, "loss": 0.003, "step": 25935 }, { "epoch": 2.8117953165654814, "grad_norm": 0.12894931435585022, "learning_rate": 6.296242564920935e-06, "loss": 0.0033, "step": 25936 }, { "epoch": 2.8119037294015614, "grad_norm": 0.17820371687412262, "learning_rate": 6.292615697083999e-06, "loss": 0.0027, "step": 25937 }, { "epoch": 2.812012142237641, "grad_norm": 0.06566059589385986, "learning_rate": 6.288988829247063e-06, "loss": 0.0019, "step": 25938 }, { "epoch": 2.8121205550737205, "grad_norm": 0.395733118057251, "learning_rate": 6.285361961410126e-06, "loss": 0.0142, "step": 25939 }, { "epoch": 2.8122289679098005, "grad_norm": 0.5209813117980957, "learning_rate": 6.28173509357319e-06, "loss": 0.0195, "step": 25940 }, { "epoch": 2.8123373807458805, "grad_norm": 0.016320159658789635, "learning_rate": 6.2781082257362535e-06, "loss": 0.0003, "step": 25941 }, { "epoch": 2.81244579358196, "grad_norm": 0.06114736199378967, "learning_rate": 6.2744813578993184e-06, "loss": 0.0028, "step": 25942 }, { "epoch": 2.81255420641804, "grad_norm": 0.05805659666657448, "learning_rate": 6.270854490062382e-06, "loss": 0.002, "step": 25943 }, { "epoch": 2.8126626192541195, "grad_norm": 0.15888330340385437, "learning_rate": 6.267227622225447e-06, "loss": 0.0082, "step": 25944 }, { "epoch": 2.8127710320901995, "grad_norm": 0.3597123324871063, "learning_rate": 6.26360075438851e-06, "loss": 0.004, "step": 25945 }, { "epoch": 2.8128794449262795, "grad_norm": 0.2456710785627365, "learning_rate": 6.259973886551575e-06, "loss": 0.0077, "step": 25946 }, { "epoch": 2.812987857762359, "grad_norm": 0.23937758803367615, "learning_rate": 6.256347018714638e-06, "loss": 0.0059, "step": 25947 }, { "epoch": 2.8130962705984386, "grad_norm": 0.14490413665771484, "learning_rate": 6.252720150877703e-06, "loss": 0.0068, "step": 25948 }, { "epoch": 2.8132046834345186, "grad_norm": 0.31253522634506226, "learning_rate": 6.249093283040766e-06, "loss": 0.0262, "step": 25949 }, { "epoch": 2.8133130962705986, "grad_norm": 0.2679647207260132, "learning_rate": 6.2454664152038304e-06, "loss": 0.0372, "step": 25950 }, { "epoch": 2.813421509106678, "grad_norm": 0.16446012258529663, "learning_rate": 6.2418395473668945e-06, "loss": 0.0031, "step": 25951 }, { "epoch": 2.813529921942758, "grad_norm": 0.3799724578857422, "learning_rate": 6.238212679529959e-06, "loss": 0.0186, "step": 25952 }, { "epoch": 2.8136383347788376, "grad_norm": 0.04146279767155647, "learning_rate": 6.234585811693022e-06, "loss": 0.0025, "step": 25953 }, { "epoch": 2.8137467476149176, "grad_norm": 0.0579582042992115, "learning_rate": 6.230958943856086e-06, "loss": 0.0029, "step": 25954 }, { "epoch": 2.8138551604509976, "grad_norm": 0.20618748664855957, "learning_rate": 6.22733207601915e-06, "loss": 0.003, "step": 25955 }, { "epoch": 2.813963573287077, "grad_norm": 0.2888427674770355, "learning_rate": 6.223705208182214e-06, "loss": 0.0143, "step": 25956 }, { "epoch": 2.814071986123157, "grad_norm": 0.06927136331796646, "learning_rate": 6.220078340345278e-06, "loss": 0.0026, "step": 25957 }, { "epoch": 2.8141803989592367, "grad_norm": 0.24007770419120789, "learning_rate": 6.2164514725083424e-06, "loss": 0.0139, "step": 25958 }, { "epoch": 2.8142888117953166, "grad_norm": 0.0898553878068924, "learning_rate": 6.212824604671406e-06, "loss": 0.0018, "step": 25959 }, { "epoch": 2.8143972246313966, "grad_norm": 0.10760781168937683, "learning_rate": 6.20919773683447e-06, "loss": 0.0031, "step": 25960 }, { "epoch": 2.814505637467476, "grad_norm": 0.4271145164966583, "learning_rate": 6.205570868997534e-06, "loss": 0.0366, "step": 25961 }, { "epoch": 2.8146140503035557, "grad_norm": 1.2014561891555786, "learning_rate": 6.201944001160598e-06, "loss": 0.0586, "step": 25962 }, { "epoch": 2.8147224631396357, "grad_norm": 1.1650606393814087, "learning_rate": 6.198317133323662e-06, "loss": 0.0313, "step": 25963 }, { "epoch": 2.8148308759757157, "grad_norm": 0.07480345666408539, "learning_rate": 6.194690265486726e-06, "loss": 0.0014, "step": 25964 }, { "epoch": 2.8149392888117952, "grad_norm": 0.38920289278030396, "learning_rate": 6.19106339764979e-06, "loss": 0.013, "step": 25965 }, { "epoch": 2.815047701647875, "grad_norm": 0.43278688192367554, "learning_rate": 6.1874365298128536e-06, "loss": 0.047, "step": 25966 }, { "epoch": 2.8151561144839548, "grad_norm": 0.2140701413154602, "learning_rate": 6.183809661975918e-06, "loss": 0.0042, "step": 25967 }, { "epoch": 2.8152645273200347, "grad_norm": 0.29353654384613037, "learning_rate": 6.180182794138982e-06, "loss": 0.0235, "step": 25968 }, { "epoch": 2.8153729401561147, "grad_norm": 0.23700156807899475, "learning_rate": 6.176555926302046e-06, "loss": 0.01, "step": 25969 }, { "epoch": 2.8154813529921943, "grad_norm": 0.2978646457195282, "learning_rate": 6.17292905846511e-06, "loss": 0.0297, "step": 25970 }, { "epoch": 2.815589765828274, "grad_norm": 0.18641752004623413, "learning_rate": 6.169302190628174e-06, "loss": 0.0085, "step": 25971 }, { "epoch": 2.815698178664354, "grad_norm": 0.16656646132469177, "learning_rate": 6.165675322791238e-06, "loss": 0.0075, "step": 25972 }, { "epoch": 2.8158065915004338, "grad_norm": 0.5608525276184082, "learning_rate": 6.1620484549543015e-06, "loss": 0.021, "step": 25973 }, { "epoch": 2.8159150043365133, "grad_norm": 0.17052800953388214, "learning_rate": 6.1584215871173656e-06, "loss": 0.0042, "step": 25974 }, { "epoch": 2.8160234171725933, "grad_norm": 0.15584127604961395, "learning_rate": 6.15479471928043e-06, "loss": 0.0074, "step": 25975 }, { "epoch": 2.816131830008673, "grad_norm": 0.0832461416721344, "learning_rate": 6.151167851443494e-06, "loss": 0.002, "step": 25976 }, { "epoch": 2.816240242844753, "grad_norm": 0.008070490323007107, "learning_rate": 6.147540983606558e-06, "loss": 0.0003, "step": 25977 }, { "epoch": 2.816348655680833, "grad_norm": 0.6212633848190308, "learning_rate": 6.143914115769622e-06, "loss": 0.0136, "step": 25978 }, { "epoch": 2.8164570685169124, "grad_norm": 0.042040809988975525, "learning_rate": 6.140287247932685e-06, "loss": 0.001, "step": 25979 }, { "epoch": 2.8165654813529923, "grad_norm": 0.09468460828065872, "learning_rate": 6.136660380095749e-06, "loss": 0.0026, "step": 25980 }, { "epoch": 2.816673894189072, "grad_norm": 0.15021638572216034, "learning_rate": 6.1330335122588135e-06, "loss": 0.0137, "step": 25981 }, { "epoch": 2.816782307025152, "grad_norm": 0.24562042951583862, "learning_rate": 6.1294066444218776e-06, "loss": 0.0178, "step": 25982 }, { "epoch": 2.8168907198612314, "grad_norm": 0.34412816166877747, "learning_rate": 6.125779776584942e-06, "loss": 0.0092, "step": 25983 }, { "epoch": 2.8169991326973114, "grad_norm": 0.25622114539146423, "learning_rate": 6.122152908748006e-06, "loss": 0.0093, "step": 25984 }, { "epoch": 2.817107545533391, "grad_norm": 0.08237013965845108, "learning_rate": 6.11852604091107e-06, "loss": 0.003, "step": 25985 }, { "epoch": 2.817215958369471, "grad_norm": 0.21737945079803467, "learning_rate": 6.114899173074133e-06, "loss": 0.0071, "step": 25986 }, { "epoch": 2.817324371205551, "grad_norm": 0.21364730596542358, "learning_rate": 6.111272305237197e-06, "loss": 0.0068, "step": 25987 }, { "epoch": 2.8174327840416304, "grad_norm": 0.18254752457141876, "learning_rate": 6.107645437400261e-06, "loss": 0.0079, "step": 25988 }, { "epoch": 2.8175411968777104, "grad_norm": 0.1596994400024414, "learning_rate": 6.1040185695633254e-06, "loss": 0.0038, "step": 25989 }, { "epoch": 2.81764960971379, "grad_norm": 0.5184817314147949, "learning_rate": 6.1003917017263896e-06, "loss": 0.039, "step": 25990 }, { "epoch": 2.81775802254987, "grad_norm": 0.14424045383930206, "learning_rate": 6.096764833889454e-06, "loss": 0.0055, "step": 25991 }, { "epoch": 2.81786643538595, "grad_norm": 0.16585467755794525, "learning_rate": 6.093137966052517e-06, "loss": 0.0067, "step": 25992 }, { "epoch": 2.8179748482220295, "grad_norm": 0.16979412734508514, "learning_rate": 6.089511098215581e-06, "loss": 0.0061, "step": 25993 }, { "epoch": 2.818083261058109, "grad_norm": 0.09295177459716797, "learning_rate": 6.085884230378645e-06, "loss": 0.0038, "step": 25994 }, { "epoch": 2.818191673894189, "grad_norm": 0.15988709032535553, "learning_rate": 6.082257362541709e-06, "loss": 0.0033, "step": 25995 }, { "epoch": 2.818300086730269, "grad_norm": 0.3720681071281433, "learning_rate": 6.078630494704773e-06, "loss": 0.0221, "step": 25996 }, { "epoch": 2.8184084995663485, "grad_norm": 0.06404709815979004, "learning_rate": 6.0750036268678374e-06, "loss": 0.0021, "step": 25997 }, { "epoch": 2.8185169124024285, "grad_norm": 0.07763046026229858, "learning_rate": 6.0713767590309015e-06, "loss": 0.002, "step": 25998 }, { "epoch": 2.818625325238508, "grad_norm": 0.4447040855884552, "learning_rate": 6.067749891193966e-06, "loss": 0.055, "step": 25999 }, { "epoch": 2.818733738074588, "grad_norm": 0.07427924871444702, "learning_rate": 6.064123023357029e-06, "loss": 0.0036, "step": 26000 }, { "epoch": 2.818842150910668, "grad_norm": 0.10214904695749283, "learning_rate": 6.060496155520093e-06, "loss": 0.0054, "step": 26001 }, { "epoch": 2.8189505637467476, "grad_norm": 0.13940808176994324, "learning_rate": 6.056869287683157e-06, "loss": 0.0052, "step": 26002 }, { "epoch": 2.8190589765828276, "grad_norm": 0.340719074010849, "learning_rate": 6.053242419846221e-06, "loss": 0.0153, "step": 26003 }, { "epoch": 2.819167389418907, "grad_norm": 0.18581923842430115, "learning_rate": 6.049615552009285e-06, "loss": 0.017, "step": 26004 }, { "epoch": 2.819275802254987, "grad_norm": 0.13254056870937347, "learning_rate": 6.045988684172349e-06, "loss": 0.0058, "step": 26005 }, { "epoch": 2.8193842150910666, "grad_norm": 0.15142862498760223, "learning_rate": 6.042361816335413e-06, "loss": 0.0102, "step": 26006 }, { "epoch": 2.8194926279271466, "grad_norm": 0.35201191902160645, "learning_rate": 6.038734948498477e-06, "loss": 0.0169, "step": 26007 }, { "epoch": 2.819601040763226, "grad_norm": 0.2750207781791687, "learning_rate": 6.035108080661541e-06, "loss": 0.0418, "step": 26008 }, { "epoch": 2.819709453599306, "grad_norm": 0.4474894404411316, "learning_rate": 6.031481212824605e-06, "loss": 0.0341, "step": 26009 }, { "epoch": 2.819817866435386, "grad_norm": 0.3272250294685364, "learning_rate": 6.027854344987669e-06, "loss": 0.0085, "step": 26010 }, { "epoch": 2.8199262792714657, "grad_norm": 0.4946572184562683, "learning_rate": 6.024227477150733e-06, "loss": 0.0443, "step": 26011 }, { "epoch": 2.8200346921075456, "grad_norm": 0.5066923499107361, "learning_rate": 6.020600609313797e-06, "loss": 0.0422, "step": 26012 }, { "epoch": 2.820143104943625, "grad_norm": 0.23417259752750397, "learning_rate": 6.0169737414768614e-06, "loss": 0.0296, "step": 26013 }, { "epoch": 2.820251517779705, "grad_norm": 0.401385635137558, "learning_rate": 6.013346873639925e-06, "loss": 0.0161, "step": 26014 }, { "epoch": 2.820359930615785, "grad_norm": 0.1584821343421936, "learning_rate": 6.009720005802989e-06, "loss": 0.0082, "step": 26015 }, { "epoch": 2.8204683434518647, "grad_norm": 0.07036881148815155, "learning_rate": 6.006093137966053e-06, "loss": 0.0022, "step": 26016 }, { "epoch": 2.8205767562879442, "grad_norm": 0.3514431416988373, "learning_rate": 6.002466270129117e-06, "loss": 0.0205, "step": 26017 }, { "epoch": 2.8206851691240242, "grad_norm": 0.2361535280942917, "learning_rate": 5.99883940229218e-06, "loss": 0.0095, "step": 26018 }, { "epoch": 2.820793581960104, "grad_norm": 0.03886920213699341, "learning_rate": 5.995212534455244e-06, "loss": 0.0012, "step": 26019 }, { "epoch": 2.8209019947961838, "grad_norm": 0.06612224131822586, "learning_rate": 5.9915856666183085e-06, "loss": 0.0022, "step": 26020 }, { "epoch": 2.8210104076322637, "grad_norm": 0.11101578176021576, "learning_rate": 5.9879587987813726e-06, "loss": 0.0042, "step": 26021 }, { "epoch": 2.8211188204683433, "grad_norm": 0.017510976642370224, "learning_rate": 5.984331930944437e-06, "loss": 0.0007, "step": 26022 }, { "epoch": 2.8212272333044233, "grad_norm": 0.07295913994312286, "learning_rate": 5.980705063107501e-06, "loss": 0.0036, "step": 26023 }, { "epoch": 2.8213356461405033, "grad_norm": 0.2569240927696228, "learning_rate": 5.977078195270565e-06, "loss": 0.004, "step": 26024 }, { "epoch": 2.821444058976583, "grad_norm": 0.29445144534111023, "learning_rate": 5.973451327433629e-06, "loss": 0.0105, "step": 26025 }, { "epoch": 2.8215524718126628, "grad_norm": 0.009427841752767563, "learning_rate": 5.969824459596693e-06, "loss": 0.0003, "step": 26026 }, { "epoch": 2.8216608846487423, "grad_norm": 0.1016194075345993, "learning_rate": 5.966197591759756e-06, "loss": 0.0026, "step": 26027 }, { "epoch": 2.8217692974848223, "grad_norm": 0.49223610758781433, "learning_rate": 5.9625707239228205e-06, "loss": 0.0249, "step": 26028 }, { "epoch": 2.821877710320902, "grad_norm": 0.0038387568201869726, "learning_rate": 5.9589438560858846e-06, "loss": 0.0002, "step": 26029 }, { "epoch": 2.821986123156982, "grad_norm": 0.23636087775230408, "learning_rate": 5.955316988248949e-06, "loss": 0.0124, "step": 26030 }, { "epoch": 2.8220945359930614, "grad_norm": 0.10834304243326187, "learning_rate": 5.951690120412012e-06, "loss": 0.0027, "step": 26031 }, { "epoch": 2.8222029488291414, "grad_norm": 0.5931714177131653, "learning_rate": 5.948063252575076e-06, "loss": 0.0175, "step": 26032 }, { "epoch": 2.8223113616652213, "grad_norm": 0.21422462165355682, "learning_rate": 5.94443638473814e-06, "loss": 0.0069, "step": 26033 }, { "epoch": 2.822419774501301, "grad_norm": 0.40803638100624084, "learning_rate": 5.940809516901204e-06, "loss": 0.0218, "step": 26034 }, { "epoch": 2.822528187337381, "grad_norm": 1.60585618019104, "learning_rate": 5.937182649064268e-06, "loss": 0.0874, "step": 26035 }, { "epoch": 2.8226366001734604, "grad_norm": 0.011956591159105301, "learning_rate": 5.9335557812273325e-06, "loss": 0.0005, "step": 26036 }, { "epoch": 2.8227450130095404, "grad_norm": 0.022430678829550743, "learning_rate": 5.9299289133903966e-06, "loss": 0.0007, "step": 26037 }, { "epoch": 2.8228534258456204, "grad_norm": 0.35269081592559814, "learning_rate": 5.926302045553461e-06, "loss": 0.0153, "step": 26038 }, { "epoch": 2.8229618386817, "grad_norm": 0.022098634392023087, "learning_rate": 5.922675177716525e-06, "loss": 0.001, "step": 26039 }, { "epoch": 2.8230702515177795, "grad_norm": 0.08971825242042542, "learning_rate": 5.919048309879589e-06, "loss": 0.0036, "step": 26040 }, { "epoch": 2.8231786643538594, "grad_norm": 0.20723606646060944, "learning_rate": 5.915421442042652e-06, "loss": 0.0079, "step": 26041 }, { "epoch": 2.8232870771899394, "grad_norm": 0.2319401055574417, "learning_rate": 5.911794574205716e-06, "loss": 0.011, "step": 26042 }, { "epoch": 2.823395490026019, "grad_norm": 0.12842677533626556, "learning_rate": 5.90816770636878e-06, "loss": 0.0062, "step": 26043 }, { "epoch": 2.823503902862099, "grad_norm": 0.028866948559880257, "learning_rate": 5.9045408385318445e-06, "loss": 0.0011, "step": 26044 }, { "epoch": 2.8236123156981785, "grad_norm": 0.07839353382587433, "learning_rate": 5.900913970694908e-06, "loss": 0.0039, "step": 26045 }, { "epoch": 2.8237207285342585, "grad_norm": 0.07375327497720718, "learning_rate": 5.897287102857972e-06, "loss": 0.0051, "step": 26046 }, { "epoch": 2.8238291413703385, "grad_norm": 0.44641226530075073, "learning_rate": 5.893660235021036e-06, "loss": 0.0243, "step": 26047 }, { "epoch": 2.823937554206418, "grad_norm": 0.634028434753418, "learning_rate": 5.8900333671841e-06, "loss": 0.0296, "step": 26048 }, { "epoch": 2.824045967042498, "grad_norm": 0.0037068671081215143, "learning_rate": 5.886406499347164e-06, "loss": 0.0002, "step": 26049 }, { "epoch": 2.8241543798785775, "grad_norm": 0.18875987827777863, "learning_rate": 5.882779631510228e-06, "loss": 0.0098, "step": 26050 }, { "epoch": 2.8242627927146575, "grad_norm": 0.1957278698682785, "learning_rate": 5.879152763673292e-06, "loss": 0.0037, "step": 26051 }, { "epoch": 2.824371205550737, "grad_norm": 0.4204651713371277, "learning_rate": 5.8755258958363564e-06, "loss": 0.0155, "step": 26052 }, { "epoch": 2.824479618386817, "grad_norm": 0.06553365290164948, "learning_rate": 5.8718990279994205e-06, "loss": 0.0008, "step": 26053 }, { "epoch": 2.8245880312228966, "grad_norm": 0.45043107867240906, "learning_rate": 5.868272160162484e-06, "loss": 0.01, "step": 26054 }, { "epoch": 2.8246964440589766, "grad_norm": 0.8261581063270569, "learning_rate": 5.864645292325548e-06, "loss": 0.0232, "step": 26055 }, { "epoch": 2.8248048568950566, "grad_norm": 0.3277738392353058, "learning_rate": 5.861018424488612e-06, "loss": 0.0044, "step": 26056 }, { "epoch": 2.824913269731136, "grad_norm": 0.2595895230770111, "learning_rate": 5.857391556651676e-06, "loss": 0.0085, "step": 26057 }, { "epoch": 2.825021682567216, "grad_norm": 0.07046253979206085, "learning_rate": 5.853764688814739e-06, "loss": 0.0019, "step": 26058 }, { "epoch": 2.8251300954032956, "grad_norm": 0.4018004834651947, "learning_rate": 5.8501378209778035e-06, "loss": 0.0206, "step": 26059 }, { "epoch": 2.8252385082393756, "grad_norm": 0.13352495431900024, "learning_rate": 5.846510953140868e-06, "loss": 0.0052, "step": 26060 }, { "epoch": 2.8253469210754556, "grad_norm": 0.3877624571323395, "learning_rate": 5.842884085303932e-06, "loss": 0.0215, "step": 26061 }, { "epoch": 2.825455333911535, "grad_norm": 0.29285675287246704, "learning_rate": 5.839257217466996e-06, "loss": 0.0157, "step": 26062 }, { "epoch": 2.8255637467476147, "grad_norm": 0.0619705393910408, "learning_rate": 5.83563034963006e-06, "loss": 0.0024, "step": 26063 }, { "epoch": 2.8256721595836947, "grad_norm": 0.10189400613307953, "learning_rate": 5.832003481793124e-06, "loss": 0.0045, "step": 26064 }, { "epoch": 2.8257805724197746, "grad_norm": 0.32315412163734436, "learning_rate": 5.828376613956188e-06, "loss": 0.0136, "step": 26065 }, { "epoch": 2.825888985255854, "grad_norm": 0.6145509481430054, "learning_rate": 5.824749746119252e-06, "loss": 0.0194, "step": 26066 }, { "epoch": 2.825997398091934, "grad_norm": 0.30869027972221375, "learning_rate": 5.821122878282316e-06, "loss": 0.0145, "step": 26067 }, { "epoch": 2.8261058109280137, "grad_norm": 0.25862717628479004, "learning_rate": 5.81749601044538e-06, "loss": 0.014, "step": 26068 }, { "epoch": 2.8262142237640937, "grad_norm": 0.33118656277656555, "learning_rate": 5.813869142608444e-06, "loss": 0.016, "step": 26069 }, { "epoch": 2.8263226366001737, "grad_norm": 0.2285832017660141, "learning_rate": 5.810242274771508e-06, "loss": 0.0203, "step": 26070 }, { "epoch": 2.8264310494362532, "grad_norm": 0.02907983399927616, "learning_rate": 5.806615406934571e-06, "loss": 0.0006, "step": 26071 }, { "epoch": 2.8265394622723328, "grad_norm": 0.09317328780889511, "learning_rate": 5.802988539097635e-06, "loss": 0.003, "step": 26072 }, { "epoch": 2.8266478751084128, "grad_norm": 0.08494895696640015, "learning_rate": 5.799361671260699e-06, "loss": 0.0044, "step": 26073 }, { "epoch": 2.8267562879444927, "grad_norm": 0.4220934510231018, "learning_rate": 5.795734803423763e-06, "loss": 0.0418, "step": 26074 }, { "epoch": 2.8268647007805723, "grad_norm": 0.25051918625831604, "learning_rate": 5.7921079355868275e-06, "loss": 0.0078, "step": 26075 }, { "epoch": 2.8269731136166523, "grad_norm": 0.1292569786310196, "learning_rate": 5.788481067749892e-06, "loss": 0.0032, "step": 26076 }, { "epoch": 2.827081526452732, "grad_norm": 0.2886243164539337, "learning_rate": 5.784854199912956e-06, "loss": 0.0077, "step": 26077 }, { "epoch": 2.827189939288812, "grad_norm": 0.012450393289327621, "learning_rate": 5.78122733207602e-06, "loss": 0.0005, "step": 26078 }, { "epoch": 2.827298352124892, "grad_norm": 0.6392906308174133, "learning_rate": 5.777600464239084e-06, "loss": 0.0333, "step": 26079 }, { "epoch": 2.8274067649609713, "grad_norm": 0.9155651330947876, "learning_rate": 5.773973596402148e-06, "loss": 0.0216, "step": 26080 }, { "epoch": 2.8275151777970513, "grad_norm": 0.34595346450805664, "learning_rate": 5.770346728565211e-06, "loss": 0.0136, "step": 26081 }, { "epoch": 2.827623590633131, "grad_norm": 0.38733381032943726, "learning_rate": 5.766719860728275e-06, "loss": 0.0138, "step": 26082 }, { "epoch": 2.827732003469211, "grad_norm": 0.3233235776424408, "learning_rate": 5.7630929928913395e-06, "loss": 0.0153, "step": 26083 }, { "epoch": 2.827840416305291, "grad_norm": 0.13617780804634094, "learning_rate": 5.759466125054403e-06, "loss": 0.0043, "step": 26084 }, { "epoch": 2.8279488291413704, "grad_norm": 0.21814075112342834, "learning_rate": 5.755839257217467e-06, "loss": 0.0191, "step": 26085 }, { "epoch": 2.82805724197745, "grad_norm": 0.07462179660797119, "learning_rate": 5.752212389380531e-06, "loss": 0.005, "step": 26086 }, { "epoch": 2.82816565481353, "grad_norm": 0.3333975374698639, "learning_rate": 5.748585521543595e-06, "loss": 0.0075, "step": 26087 }, { "epoch": 2.82827406764961, "grad_norm": 0.10336615145206451, "learning_rate": 5.744958653706659e-06, "loss": 0.0053, "step": 26088 }, { "epoch": 2.8283824804856894, "grad_norm": 0.3101027309894562, "learning_rate": 5.741331785869723e-06, "loss": 0.0106, "step": 26089 }, { "epoch": 2.8284908933217694, "grad_norm": 0.6427663564682007, "learning_rate": 5.737704918032787e-06, "loss": 0.0074, "step": 26090 }, { "epoch": 2.828599306157849, "grad_norm": 0.16052033007144928, "learning_rate": 5.7340780501958515e-06, "loss": 0.0059, "step": 26091 }, { "epoch": 2.828707718993929, "grad_norm": 0.3362211287021637, "learning_rate": 5.7304511823589156e-06, "loss": 0.0453, "step": 26092 }, { "epoch": 2.828816131830009, "grad_norm": 0.06973949074745178, "learning_rate": 5.72682431452198e-06, "loss": 0.0039, "step": 26093 }, { "epoch": 2.8289245446660884, "grad_norm": 0.8161518573760986, "learning_rate": 5.723197446685044e-06, "loss": 0.025, "step": 26094 }, { "epoch": 2.829032957502168, "grad_norm": 0.2537006139755249, "learning_rate": 5.719570578848107e-06, "loss": 0.0142, "step": 26095 }, { "epoch": 2.829141370338248, "grad_norm": 0.44634777307510376, "learning_rate": 5.715943711011171e-06, "loss": 0.0111, "step": 26096 }, { "epoch": 2.829249783174328, "grad_norm": 0.04116188362240791, "learning_rate": 5.712316843174234e-06, "loss": 0.0018, "step": 26097 }, { "epoch": 2.8293581960104075, "grad_norm": 0.03325962647795677, "learning_rate": 5.7086899753372985e-06, "loss": 0.0017, "step": 26098 }, { "epoch": 2.8294666088464875, "grad_norm": 0.0468672476708889, "learning_rate": 5.705063107500363e-06, "loss": 0.0018, "step": 26099 }, { "epoch": 2.829575021682567, "grad_norm": 0.24547764658927917, "learning_rate": 5.701436239663427e-06, "loss": 0.0051, "step": 26100 }, { "epoch": 2.829683434518647, "grad_norm": 0.11405301839113235, "learning_rate": 5.697809371826491e-06, "loss": 0.0738, "step": 26101 }, { "epoch": 2.829791847354727, "grad_norm": 1.307559847831726, "learning_rate": 5.694182503989555e-06, "loss": 0.0543, "step": 26102 }, { "epoch": 2.8299002601908065, "grad_norm": 0.19331975281238556, "learning_rate": 5.690555636152619e-06, "loss": 0.0109, "step": 26103 }, { "epoch": 2.8300086730268865, "grad_norm": 0.3454751670360565, "learning_rate": 5.686928768315683e-06, "loss": 0.0053, "step": 26104 }, { "epoch": 2.830117085862966, "grad_norm": 0.6474696397781372, "learning_rate": 5.683301900478747e-06, "loss": 0.0229, "step": 26105 }, { "epoch": 2.830225498699046, "grad_norm": 0.5072271823883057, "learning_rate": 5.679675032641811e-06, "loss": 0.0281, "step": 26106 }, { "epoch": 2.830333911535126, "grad_norm": 0.39637529850006104, "learning_rate": 5.6760481648048754e-06, "loss": 0.009, "step": 26107 }, { "epoch": 2.8304423243712056, "grad_norm": 0.22432439029216766, "learning_rate": 5.672421296967939e-06, "loss": 0.0123, "step": 26108 }, { "epoch": 2.830550737207285, "grad_norm": 0.059331264346838, "learning_rate": 5.668794429131003e-06, "loss": 0.0024, "step": 26109 }, { "epoch": 2.830659150043365, "grad_norm": 0.6931152939796448, "learning_rate": 5.665167561294066e-06, "loss": 0.0289, "step": 26110 }, { "epoch": 2.830767562879445, "grad_norm": 0.2466350495815277, "learning_rate": 5.66154069345713e-06, "loss": 0.0108, "step": 26111 }, { "epoch": 2.8308759757155246, "grad_norm": 0.09127943217754364, "learning_rate": 5.657913825620194e-06, "loss": 0.0035, "step": 26112 }, { "epoch": 2.8309843885516046, "grad_norm": 0.1545710265636444, "learning_rate": 5.654286957783258e-06, "loss": 0.0022, "step": 26113 }, { "epoch": 2.831092801387684, "grad_norm": 0.5632163882255554, "learning_rate": 5.6506600899463225e-06, "loss": 0.0238, "step": 26114 }, { "epoch": 2.831201214223764, "grad_norm": 0.003448242088779807, "learning_rate": 5.647033222109387e-06, "loss": 0.0001, "step": 26115 }, { "epoch": 2.831309627059844, "grad_norm": 0.15498927235603333, "learning_rate": 5.643406354272451e-06, "loss": 0.0044, "step": 26116 }, { "epoch": 2.8314180398959237, "grad_norm": 0.14275436103343964, "learning_rate": 5.639779486435515e-06, "loss": 0.0594, "step": 26117 }, { "epoch": 2.831526452732003, "grad_norm": 0.12280430644750595, "learning_rate": 5.636152618598579e-06, "loss": 0.0057, "step": 26118 }, { "epoch": 2.831634865568083, "grad_norm": 0.37120920419692993, "learning_rate": 5.632525750761643e-06, "loss": 0.0157, "step": 26119 }, { "epoch": 2.831743278404163, "grad_norm": 0.4665619730949402, "learning_rate": 5.628898882924707e-06, "loss": 0.0173, "step": 26120 }, { "epoch": 2.8318516912402427, "grad_norm": 0.617780327796936, "learning_rate": 5.62527201508777e-06, "loss": 0.0482, "step": 26121 }, { "epoch": 2.8319601040763227, "grad_norm": 0.11751459538936615, "learning_rate": 5.6216451472508345e-06, "loss": 0.0034, "step": 26122 }, { "epoch": 2.8320685169124022, "grad_norm": 0.1751476228237152, "learning_rate": 5.618018279413898e-06, "loss": 0.0074, "step": 26123 }, { "epoch": 2.8321769297484822, "grad_norm": 0.051566556096076965, "learning_rate": 5.614391411576962e-06, "loss": 0.0026, "step": 26124 }, { "epoch": 2.832285342584562, "grad_norm": 0.6639743447303772, "learning_rate": 5.610764543740026e-06, "loss": 0.0198, "step": 26125 }, { "epoch": 2.8323937554206418, "grad_norm": 0.3032630383968353, "learning_rate": 5.60713767590309e-06, "loss": 0.009, "step": 26126 }, { "epoch": 2.8325021682567217, "grad_norm": 0.11847765743732452, "learning_rate": 5.603510808066154e-06, "loss": 0.0076, "step": 26127 }, { "epoch": 2.8326105810928013, "grad_norm": 0.20301023125648499, "learning_rate": 5.599883940229218e-06, "loss": 0.0035, "step": 26128 }, { "epoch": 2.8327189939288813, "grad_norm": 0.23091243207454681, "learning_rate": 5.596257072392282e-06, "loss": 0.0112, "step": 26129 }, { "epoch": 2.832827406764961, "grad_norm": 0.7605233192443848, "learning_rate": 5.5926302045553465e-06, "loss": 0.0361, "step": 26130 }, { "epoch": 2.832935819601041, "grad_norm": 0.24847476184368134, "learning_rate": 5.589003336718411e-06, "loss": 0.0065, "step": 26131 }, { "epoch": 2.8330442324371203, "grad_norm": 0.4043782949447632, "learning_rate": 5.585376468881475e-06, "loss": 0.0222, "step": 26132 }, { "epoch": 2.8331526452732003, "grad_norm": 0.3290174603462219, "learning_rate": 5.581749601044539e-06, "loss": 0.0169, "step": 26133 }, { "epoch": 2.8332610581092803, "grad_norm": 0.8678651452064514, "learning_rate": 5.578122733207603e-06, "loss": 0.0185, "step": 26134 }, { "epoch": 2.83336947094536, "grad_norm": 0.18718384206295013, "learning_rate": 5.574495865370666e-06, "loss": 0.0096, "step": 26135 }, { "epoch": 2.83347788378144, "grad_norm": 0.01638893038034439, "learning_rate": 5.57086899753373e-06, "loss": 0.0007, "step": 26136 }, { "epoch": 2.8335862966175194, "grad_norm": 0.047527704387903214, "learning_rate": 5.5672421296967935e-06, "loss": 0.0036, "step": 26137 }, { "epoch": 2.8336947094535994, "grad_norm": 0.05006347969174385, "learning_rate": 5.563615261859858e-06, "loss": 0.0013, "step": 26138 }, { "epoch": 2.8338031222896793, "grad_norm": 0.697839617729187, "learning_rate": 5.559988394022922e-06, "loss": 0.017, "step": 26139 }, { "epoch": 2.833911535125759, "grad_norm": 0.22678875923156738, "learning_rate": 5.556361526185986e-06, "loss": 0.0097, "step": 26140 }, { "epoch": 2.8340199479618384, "grad_norm": 0.3824573755264282, "learning_rate": 5.55273465834905e-06, "loss": 0.0318, "step": 26141 }, { "epoch": 2.8341283607979184, "grad_norm": 0.4670046269893646, "learning_rate": 5.549107790512114e-06, "loss": 0.016, "step": 26142 }, { "epoch": 2.8342367736339984, "grad_norm": 0.10850292444229126, "learning_rate": 5.545480922675178e-06, "loss": 0.0037, "step": 26143 }, { "epoch": 2.834345186470078, "grad_norm": 0.024531308561563492, "learning_rate": 5.541854054838242e-06, "loss": 0.0008, "step": 26144 }, { "epoch": 2.834453599306158, "grad_norm": 0.3694719672203064, "learning_rate": 5.538227187001306e-06, "loss": 0.0136, "step": 26145 }, { "epoch": 2.8345620121422375, "grad_norm": 0.17817135155200958, "learning_rate": 5.5346003191643705e-06, "loss": 0.0101, "step": 26146 }, { "epoch": 2.8346704249783174, "grad_norm": 0.2171272337436676, "learning_rate": 5.5309734513274346e-06, "loss": 0.0148, "step": 26147 }, { "epoch": 2.8347788378143974, "grad_norm": 0.1275685876607895, "learning_rate": 5.527346583490498e-06, "loss": 0.0041, "step": 26148 }, { "epoch": 2.834887250650477, "grad_norm": 0.21936902403831482, "learning_rate": 5.523719715653562e-06, "loss": 0.0108, "step": 26149 }, { "epoch": 2.834995663486557, "grad_norm": 0.38836562633514404, "learning_rate": 5.520092847816625e-06, "loss": 0.021, "step": 26150 }, { "epoch": 2.8351040763226365, "grad_norm": 0.013184101320803165, "learning_rate": 5.516465979979689e-06, "loss": 0.0007, "step": 26151 }, { "epoch": 2.8352124891587165, "grad_norm": 0.121375672519207, "learning_rate": 5.512839112142753e-06, "loss": 0.0044, "step": 26152 }, { "epoch": 2.835320901994796, "grad_norm": 0.19705374538898468, "learning_rate": 5.5092122443058175e-06, "loss": 0.0163, "step": 26153 }, { "epoch": 2.835429314830876, "grad_norm": 0.35862499475479126, "learning_rate": 5.505585376468882e-06, "loss": 0.0154, "step": 26154 }, { "epoch": 2.8355377276669556, "grad_norm": 0.32755815982818604, "learning_rate": 5.501958508631946e-06, "loss": 0.011, "step": 26155 }, { "epoch": 2.8356461405030355, "grad_norm": 0.14148390293121338, "learning_rate": 5.49833164079501e-06, "loss": 0.0027, "step": 26156 }, { "epoch": 2.8357545533391155, "grad_norm": 0.006857555825263262, "learning_rate": 5.494704772958074e-06, "loss": 0.0005, "step": 26157 }, { "epoch": 2.835862966175195, "grad_norm": 0.044564396142959595, "learning_rate": 5.491077905121138e-06, "loss": 0.0014, "step": 26158 }, { "epoch": 2.835971379011275, "grad_norm": 0.04443054273724556, "learning_rate": 5.487451037284202e-06, "loss": 0.0028, "step": 26159 }, { "epoch": 2.8360797918473546, "grad_norm": 0.5222647786140442, "learning_rate": 5.483824169447266e-06, "loss": 0.0267, "step": 26160 }, { "epoch": 2.8361882046834346, "grad_norm": 0.2069280594587326, "learning_rate": 5.4801973016103295e-06, "loss": 0.0079, "step": 26161 }, { "epoch": 2.8362966175195146, "grad_norm": 0.6249726414680481, "learning_rate": 5.476570433773394e-06, "loss": 0.0198, "step": 26162 }, { "epoch": 2.836405030355594, "grad_norm": 0.052283477038145065, "learning_rate": 5.472943565936457e-06, "loss": 0.0027, "step": 26163 }, { "epoch": 2.8365134431916736, "grad_norm": 0.14309293031692505, "learning_rate": 5.469316698099521e-06, "loss": 0.007, "step": 26164 }, { "epoch": 2.8366218560277536, "grad_norm": 0.36472755670547485, "learning_rate": 5.465689830262585e-06, "loss": 0.0286, "step": 26165 }, { "epoch": 2.8367302688638336, "grad_norm": 0.1688878834247589, "learning_rate": 5.462062962425649e-06, "loss": 0.0036, "step": 26166 }, { "epoch": 2.836838681699913, "grad_norm": 0.13428543508052826, "learning_rate": 5.458436094588713e-06, "loss": 0.0083, "step": 26167 }, { "epoch": 2.836947094535993, "grad_norm": 0.12024354189634323, "learning_rate": 5.454809226751777e-06, "loss": 0.0021, "step": 26168 }, { "epoch": 2.8370555073720727, "grad_norm": 0.06613154709339142, "learning_rate": 5.4511823589148415e-06, "loss": 0.0016, "step": 26169 }, { "epoch": 2.8371639202081527, "grad_norm": 0.256673663854599, "learning_rate": 5.447555491077906e-06, "loss": 0.0038, "step": 26170 }, { "epoch": 2.8372723330442327, "grad_norm": 0.3526802062988281, "learning_rate": 5.44392862324097e-06, "loss": 0.0167, "step": 26171 }, { "epoch": 2.837380745880312, "grad_norm": 0.06343986839056015, "learning_rate": 5.440301755404034e-06, "loss": 0.0015, "step": 26172 }, { "epoch": 2.837489158716392, "grad_norm": 0.2998158633708954, "learning_rate": 5.436674887567098e-06, "loss": 0.0166, "step": 26173 }, { "epoch": 2.8375975715524717, "grad_norm": 0.05269750580191612, "learning_rate": 5.433048019730161e-06, "loss": 0.0012, "step": 26174 }, { "epoch": 2.8377059843885517, "grad_norm": 0.40715333819389343, "learning_rate": 5.429421151893225e-06, "loss": 0.011, "step": 26175 }, { "epoch": 2.8378143972246312, "grad_norm": 0.195270374417305, "learning_rate": 5.4257942840562885e-06, "loss": 0.0043, "step": 26176 }, { "epoch": 2.8379228100607112, "grad_norm": 0.03689098730683327, "learning_rate": 5.422167416219353e-06, "loss": 0.0012, "step": 26177 }, { "epoch": 2.8380312228967908, "grad_norm": 0.0776306688785553, "learning_rate": 5.418540548382417e-06, "loss": 0.0021, "step": 26178 }, { "epoch": 2.8381396357328708, "grad_norm": 0.40970638394355774, "learning_rate": 5.414913680545481e-06, "loss": 0.0192, "step": 26179 }, { "epoch": 2.8382480485689507, "grad_norm": 0.25484463572502136, "learning_rate": 5.411286812708545e-06, "loss": 0.0101, "step": 26180 }, { "epoch": 2.8383564614050303, "grad_norm": 0.2153930515050888, "learning_rate": 5.407659944871609e-06, "loss": 0.0061, "step": 26181 }, { "epoch": 2.8384648742411103, "grad_norm": 0.23294249176979065, "learning_rate": 5.404033077034673e-06, "loss": 0.0119, "step": 26182 }, { "epoch": 2.83857328707719, "grad_norm": 0.07581159472465515, "learning_rate": 5.400406209197737e-06, "loss": 0.0024, "step": 26183 }, { "epoch": 2.83868169991327, "grad_norm": 0.1855873465538025, "learning_rate": 5.396779341360801e-06, "loss": 0.0056, "step": 26184 }, { "epoch": 2.83879011274935, "grad_norm": 0.45599648356437683, "learning_rate": 5.3931524735238655e-06, "loss": 0.0341, "step": 26185 }, { "epoch": 2.8388985255854293, "grad_norm": 0.062293890863657, "learning_rate": 5.38952560568693e-06, "loss": 0.0022, "step": 26186 }, { "epoch": 2.839006938421509, "grad_norm": 0.06598888337612152, "learning_rate": 5.385898737849993e-06, "loss": 0.0018, "step": 26187 }, { "epoch": 2.839115351257589, "grad_norm": 0.08241978287696838, "learning_rate": 5.382271870013057e-06, "loss": 0.0022, "step": 26188 }, { "epoch": 2.839223764093669, "grad_norm": 0.050445154309272766, "learning_rate": 5.378645002176121e-06, "loss": 0.0037, "step": 26189 }, { "epoch": 2.8393321769297484, "grad_norm": 0.061592720448970795, "learning_rate": 5.375018134339184e-06, "loss": 0.0014, "step": 26190 }, { "epoch": 2.8394405897658284, "grad_norm": 0.20128977298736572, "learning_rate": 5.371391266502248e-06, "loss": 0.0176, "step": 26191 }, { "epoch": 2.839549002601908, "grad_norm": 0.08920685201883316, "learning_rate": 5.3677643986653125e-06, "loss": 0.0012, "step": 26192 }, { "epoch": 2.839657415437988, "grad_norm": 0.24536536633968353, "learning_rate": 5.364137530828377e-06, "loss": 0.0127, "step": 26193 }, { "epoch": 2.839765828274068, "grad_norm": 0.9365809559822083, "learning_rate": 5.360510662991441e-06, "loss": 0.024, "step": 26194 }, { "epoch": 2.8398742411101474, "grad_norm": 0.44604936242103577, "learning_rate": 5.356883795154505e-06, "loss": 0.0244, "step": 26195 }, { "epoch": 2.8399826539462274, "grad_norm": 0.07339726388454437, "learning_rate": 5.353256927317569e-06, "loss": 0.0039, "step": 26196 }, { "epoch": 2.840091066782307, "grad_norm": 0.5272530317306519, "learning_rate": 5.349630059480633e-06, "loss": 0.0331, "step": 26197 }, { "epoch": 2.840199479618387, "grad_norm": 0.10188938677310944, "learning_rate": 5.346003191643697e-06, "loss": 0.0021, "step": 26198 }, { "epoch": 2.8403078924544665, "grad_norm": 0.13431410491466522, "learning_rate": 5.342376323806761e-06, "loss": 0.0038, "step": 26199 }, { "epoch": 2.8404163052905465, "grad_norm": 0.29474031925201416, "learning_rate": 5.3387494559698245e-06, "loss": 0.009, "step": 26200 }, { "epoch": 2.840524718126626, "grad_norm": 0.15262886881828308, "learning_rate": 5.335122588132889e-06, "loss": 0.0058, "step": 26201 }, { "epoch": 2.840633130962706, "grad_norm": 0.8116877675056458, "learning_rate": 5.331495720295953e-06, "loss": 0.0465, "step": 26202 }, { "epoch": 2.840741543798786, "grad_norm": 0.3509962260723114, "learning_rate": 5.327868852459016e-06, "loss": 0.0124, "step": 26203 }, { "epoch": 2.8408499566348655, "grad_norm": 0.005122718866914511, "learning_rate": 5.32424198462208e-06, "loss": 0.0002, "step": 26204 }, { "epoch": 2.8409583694709455, "grad_norm": 0.13744451105594635, "learning_rate": 5.320615116785144e-06, "loss": 0.004, "step": 26205 }, { "epoch": 2.841066782307025, "grad_norm": 0.1548549234867096, "learning_rate": 5.316988248948208e-06, "loss": 0.004, "step": 26206 }, { "epoch": 2.841175195143105, "grad_norm": 0.24720169603824615, "learning_rate": 5.313361381111272e-06, "loss": 0.0143, "step": 26207 }, { "epoch": 2.841283607979185, "grad_norm": 0.16258323192596436, "learning_rate": 5.3097345132743365e-06, "loss": 0.0064, "step": 26208 }, { "epoch": 2.8413920208152645, "grad_norm": 0.520325243473053, "learning_rate": 5.306107645437401e-06, "loss": 0.018, "step": 26209 }, { "epoch": 2.841500433651344, "grad_norm": 0.14227192103862762, "learning_rate": 5.302480777600465e-06, "loss": 0.0025, "step": 26210 }, { "epoch": 2.841608846487424, "grad_norm": 0.07204481214284897, "learning_rate": 5.298853909763529e-06, "loss": 0.0045, "step": 26211 }, { "epoch": 2.841717259323504, "grad_norm": 0.2650201618671417, "learning_rate": 5.295227041926593e-06, "loss": 0.0072, "step": 26212 }, { "epoch": 2.8418256721595836, "grad_norm": 0.04218243435025215, "learning_rate": 5.291600174089656e-06, "loss": 0.0025, "step": 26213 }, { "epoch": 2.8419340849956636, "grad_norm": 0.11228033155202866, "learning_rate": 5.28797330625272e-06, "loss": 0.0046, "step": 26214 }, { "epoch": 2.842042497831743, "grad_norm": 0.624141275882721, "learning_rate": 5.284346438415784e-06, "loss": 0.0518, "step": 26215 }, { "epoch": 2.842150910667823, "grad_norm": 0.3674488365650177, "learning_rate": 5.2807195705788485e-06, "loss": 0.0276, "step": 26216 }, { "epoch": 2.842259323503903, "grad_norm": 0.22252213954925537, "learning_rate": 5.277092702741912e-06, "loss": 0.0063, "step": 26217 }, { "epoch": 2.8423677363399826, "grad_norm": 0.11035474389791489, "learning_rate": 5.273465834904976e-06, "loss": 0.0024, "step": 26218 }, { "epoch": 2.842476149176062, "grad_norm": 0.05229153111577034, "learning_rate": 5.26983896706804e-06, "loss": 0.0017, "step": 26219 }, { "epoch": 2.842584562012142, "grad_norm": 0.25044286251068115, "learning_rate": 5.266212099231104e-06, "loss": 0.013, "step": 26220 }, { "epoch": 2.842692974848222, "grad_norm": 0.06920814514160156, "learning_rate": 5.262585231394168e-06, "loss": 0.0028, "step": 26221 }, { "epoch": 2.8428013876843017, "grad_norm": 0.15338920056819916, "learning_rate": 5.258958363557232e-06, "loss": 0.0032, "step": 26222 }, { "epoch": 2.8429098005203817, "grad_norm": 0.33080318570137024, "learning_rate": 5.255331495720296e-06, "loss": 0.0057, "step": 26223 }, { "epoch": 2.843018213356461, "grad_norm": 0.351199209690094, "learning_rate": 5.2517046278833605e-06, "loss": 0.0235, "step": 26224 }, { "epoch": 2.843126626192541, "grad_norm": 0.3578704297542572, "learning_rate": 5.248077760046425e-06, "loss": 0.0038, "step": 26225 }, { "epoch": 2.843235039028621, "grad_norm": 0.3838939070701599, "learning_rate": 5.244450892209489e-06, "loss": 0.0248, "step": 26226 }, { "epoch": 2.8433434518647007, "grad_norm": 0.10365187376737595, "learning_rate": 5.240824024372552e-06, "loss": 0.0028, "step": 26227 }, { "epoch": 2.8434518647007807, "grad_norm": 0.07388972491025925, "learning_rate": 5.237197156535616e-06, "loss": 0.0015, "step": 26228 }, { "epoch": 2.8435602775368602, "grad_norm": 0.1581181138753891, "learning_rate": 5.23357028869868e-06, "loss": 0.0035, "step": 26229 }, { "epoch": 2.8436686903729402, "grad_norm": 0.2111949771642685, "learning_rate": 5.2299434208617434e-06, "loss": 0.006, "step": 26230 }, { "epoch": 2.84377710320902, "grad_norm": 0.06341686844825745, "learning_rate": 5.2263165530248075e-06, "loss": 0.002, "step": 26231 }, { "epoch": 2.8438855160450998, "grad_norm": 0.01695217564702034, "learning_rate": 5.222689685187872e-06, "loss": 0.0007, "step": 26232 }, { "epoch": 2.8439939288811793, "grad_norm": 0.453837126493454, "learning_rate": 5.219062817350936e-06, "loss": 0.0086, "step": 26233 }, { "epoch": 2.8441023417172593, "grad_norm": 0.7352877855300903, "learning_rate": 5.215435949514e-06, "loss": 0.0184, "step": 26234 }, { "epoch": 2.8442107545533393, "grad_norm": 0.05946328118443489, "learning_rate": 5.211809081677064e-06, "loss": 0.0018, "step": 26235 }, { "epoch": 2.844319167389419, "grad_norm": 0.2675095796585083, "learning_rate": 5.208182213840128e-06, "loss": 0.0053, "step": 26236 }, { "epoch": 2.844427580225499, "grad_norm": 0.03052656725049019, "learning_rate": 5.204555346003192e-06, "loss": 0.0009, "step": 26237 }, { "epoch": 2.8445359930615783, "grad_norm": 1.5399855375289917, "learning_rate": 5.200928478166256e-06, "loss": 0.0296, "step": 26238 }, { "epoch": 2.8446444058976583, "grad_norm": 0.03413156792521477, "learning_rate": 5.19730161032932e-06, "loss": 0.0024, "step": 26239 }, { "epoch": 2.8447528187337383, "grad_norm": 0.4141591489315033, "learning_rate": 5.193674742492384e-06, "loss": 0.0354, "step": 26240 }, { "epoch": 2.844861231569818, "grad_norm": 0.288778692483902, "learning_rate": 5.190047874655448e-06, "loss": 0.0135, "step": 26241 }, { "epoch": 2.8449696444058974, "grad_norm": 0.2095767855644226, "learning_rate": 5.186421006818512e-06, "loss": 0.0075, "step": 26242 }, { "epoch": 2.8450780572419774, "grad_norm": 0.301479309797287, "learning_rate": 5.182794138981576e-06, "loss": 0.0096, "step": 26243 }, { "epoch": 2.8451864700780574, "grad_norm": 0.09926864504814148, "learning_rate": 5.179167271144639e-06, "loss": 0.0022, "step": 26244 }, { "epoch": 2.845294882914137, "grad_norm": 0.08507244288921356, "learning_rate": 5.175540403307703e-06, "loss": 0.0011, "step": 26245 }, { "epoch": 2.845403295750217, "grad_norm": 0.08931761980056763, "learning_rate": 5.171913535470767e-06, "loss": 0.0037, "step": 26246 }, { "epoch": 2.8455117085862964, "grad_norm": 0.3832938075065613, "learning_rate": 5.1682866676338315e-06, "loss": 0.016, "step": 26247 }, { "epoch": 2.8456201214223764, "grad_norm": 0.30706164240837097, "learning_rate": 5.164659799796896e-06, "loss": 0.0169, "step": 26248 }, { "epoch": 2.8457285342584564, "grad_norm": 0.05541360005736351, "learning_rate": 5.16103293195996e-06, "loss": 0.0026, "step": 26249 }, { "epoch": 2.845836947094536, "grad_norm": 0.1983805149793625, "learning_rate": 5.157406064123024e-06, "loss": 0.0079, "step": 26250 }, { "epoch": 2.845945359930616, "grad_norm": 0.5414854288101196, "learning_rate": 5.153779196286088e-06, "loss": 0.0214, "step": 26251 }, { "epoch": 2.8460537727666955, "grad_norm": 0.20365417003631592, "learning_rate": 5.150152328449152e-06, "loss": 0.0076, "step": 26252 }, { "epoch": 2.8461621856027755, "grad_norm": 0.515979528427124, "learning_rate": 5.146525460612215e-06, "loss": 0.0075, "step": 26253 }, { "epoch": 2.8462705984388554, "grad_norm": 0.6243401765823364, "learning_rate": 5.142898592775279e-06, "loss": 0.0066, "step": 26254 }, { "epoch": 2.846379011274935, "grad_norm": 1.1086505651474, "learning_rate": 5.1392717249383435e-06, "loss": 0.0831, "step": 26255 }, { "epoch": 2.8464874241110145, "grad_norm": 0.3560929596424103, "learning_rate": 5.135644857101408e-06, "loss": 0.0115, "step": 26256 }, { "epoch": 2.8465958369470945, "grad_norm": 1.014679193496704, "learning_rate": 5.132017989264471e-06, "loss": 0.0276, "step": 26257 }, { "epoch": 2.8467042497831745, "grad_norm": 0.08910616487264633, "learning_rate": 5.128391121427535e-06, "loss": 0.0062, "step": 26258 }, { "epoch": 2.846812662619254, "grad_norm": 0.9337719082832336, "learning_rate": 5.124764253590599e-06, "loss": 0.0137, "step": 26259 }, { "epoch": 2.846921075455334, "grad_norm": 0.21764948964118958, "learning_rate": 5.121137385753663e-06, "loss": 0.0149, "step": 26260 }, { "epoch": 2.8470294882914136, "grad_norm": 0.1278868168592453, "learning_rate": 5.117510517916727e-06, "loss": 0.0022, "step": 26261 }, { "epoch": 2.8471379011274935, "grad_norm": 1.1775803565979004, "learning_rate": 5.113883650079791e-06, "loss": 0.0458, "step": 26262 }, { "epoch": 2.8472463139635735, "grad_norm": 0.19346459209918976, "learning_rate": 5.1102567822428555e-06, "loss": 0.0056, "step": 26263 }, { "epoch": 2.847354726799653, "grad_norm": 0.14127692580223083, "learning_rate": 5.10662991440592e-06, "loss": 0.0158, "step": 26264 }, { "epoch": 2.8474631396357326, "grad_norm": 0.1665896326303482, "learning_rate": 5.103003046568984e-06, "loss": 0.0098, "step": 26265 }, { "epoch": 2.8475715524718126, "grad_norm": 0.1790287047624588, "learning_rate": 5.099376178732047e-06, "loss": 0.0051, "step": 26266 }, { "epoch": 2.8476799653078926, "grad_norm": 0.16255982220172882, "learning_rate": 5.095749310895111e-06, "loss": 0.009, "step": 26267 }, { "epoch": 2.847788378143972, "grad_norm": 0.3702239394187927, "learning_rate": 5.092122443058175e-06, "loss": 0.0042, "step": 26268 }, { "epoch": 2.847896790980052, "grad_norm": 0.10264705121517181, "learning_rate": 5.088495575221239e-06, "loss": 0.0009, "step": 26269 }, { "epoch": 2.8480052038161316, "grad_norm": 0.4560513198375702, "learning_rate": 5.084868707384303e-06, "loss": 0.0162, "step": 26270 }, { "epoch": 2.8481136166522116, "grad_norm": 0.08846200257539749, "learning_rate": 5.081241839547367e-06, "loss": 0.0021, "step": 26271 }, { "epoch": 2.8482220294882916, "grad_norm": 0.16520194709300995, "learning_rate": 5.077614971710431e-06, "loss": 0.0023, "step": 26272 }, { "epoch": 2.848330442324371, "grad_norm": 0.5112224221229553, "learning_rate": 5.073988103873495e-06, "loss": 0.0538, "step": 26273 }, { "epoch": 2.848438855160451, "grad_norm": 0.020258650183677673, "learning_rate": 5.070361236036559e-06, "loss": 0.0005, "step": 26274 }, { "epoch": 2.8485472679965307, "grad_norm": 0.23313935101032257, "learning_rate": 5.066734368199623e-06, "loss": 0.01, "step": 26275 }, { "epoch": 2.8486556808326107, "grad_norm": 0.6731151342391968, "learning_rate": 5.063107500362687e-06, "loss": 0.0215, "step": 26276 }, { "epoch": 2.84876409366869, "grad_norm": 0.3003558814525604, "learning_rate": 5.059480632525751e-06, "loss": 0.0176, "step": 26277 }, { "epoch": 2.84887250650477, "grad_norm": 0.07138436287641525, "learning_rate": 5.055853764688815e-06, "loss": 0.0018, "step": 26278 }, { "epoch": 2.8489809193408497, "grad_norm": 0.6241644024848938, "learning_rate": 5.052226896851879e-06, "loss": 0.0131, "step": 26279 }, { "epoch": 2.8490893321769297, "grad_norm": 0.22829249501228333, "learning_rate": 5.048600029014943e-06, "loss": 0.0086, "step": 26280 }, { "epoch": 2.8491977450130097, "grad_norm": 0.3210376799106598, "learning_rate": 5.044973161178007e-06, "loss": 0.014, "step": 26281 }, { "epoch": 2.8493061578490892, "grad_norm": 0.1291036158800125, "learning_rate": 5.041346293341071e-06, "loss": 0.0065, "step": 26282 }, { "epoch": 2.8494145706851692, "grad_norm": 0.019043173640966415, "learning_rate": 5.037719425504135e-06, "loss": 0.0008, "step": 26283 }, { "epoch": 2.8495229835212488, "grad_norm": 0.02038268931210041, "learning_rate": 5.034092557667198e-06, "loss": 0.0012, "step": 26284 }, { "epoch": 2.8496313963573288, "grad_norm": 0.17303667962551117, "learning_rate": 5.0304656898302624e-06, "loss": 0.0096, "step": 26285 }, { "epoch": 2.8497398091934087, "grad_norm": 0.24758002161979675, "learning_rate": 5.0268388219933265e-06, "loss": 0.0052, "step": 26286 }, { "epoch": 2.8498482220294883, "grad_norm": 0.42896372079849243, "learning_rate": 5.023211954156391e-06, "loss": 0.0113, "step": 26287 }, { "epoch": 2.849956634865568, "grad_norm": 0.12439046055078506, "learning_rate": 5.019585086319455e-06, "loss": 0.0135, "step": 26288 }, { "epoch": 2.850065047701648, "grad_norm": 0.5241265892982483, "learning_rate": 5.015958218482519e-06, "loss": 0.0154, "step": 26289 }, { "epoch": 2.850173460537728, "grad_norm": 0.2590121328830719, "learning_rate": 5.012331350645583e-06, "loss": 0.0088, "step": 26290 }, { "epoch": 2.8502818733738073, "grad_norm": 0.15902358293533325, "learning_rate": 5.008704482808647e-06, "loss": 0.0036, "step": 26291 }, { "epoch": 2.8503902862098873, "grad_norm": 0.22633451223373413, "learning_rate": 5.00507761497171e-06, "loss": 0.0127, "step": 26292 }, { "epoch": 2.850498699045967, "grad_norm": 0.1770586371421814, "learning_rate": 5.0014507471347744e-06, "loss": 0.0103, "step": 26293 }, { "epoch": 2.850607111882047, "grad_norm": 0.21279789507389069, "learning_rate": 4.9978238792978385e-06, "loss": 0.0068, "step": 26294 }, { "epoch": 2.850715524718127, "grad_norm": 0.06449954956769943, "learning_rate": 4.994197011460903e-06, "loss": 0.0016, "step": 26295 }, { "epoch": 2.8508239375542064, "grad_norm": 0.049397341907024384, "learning_rate": 4.990570143623967e-06, "loss": 0.0021, "step": 26296 }, { "epoch": 2.8509323503902864, "grad_norm": 1.0787298679351807, "learning_rate": 4.986943275787031e-06, "loss": 0.0315, "step": 26297 }, { "epoch": 2.851040763226366, "grad_norm": 0.2515639662742615, "learning_rate": 4.983316407950094e-06, "loss": 0.0045, "step": 26298 }, { "epoch": 2.851149176062446, "grad_norm": 0.2281826287508011, "learning_rate": 4.979689540113158e-06, "loss": 0.0202, "step": 26299 }, { "epoch": 2.8512575888985254, "grad_norm": 0.08432447165250778, "learning_rate": 4.976062672276222e-06, "loss": 0.0027, "step": 26300 }, { "epoch": 2.8513660017346054, "grad_norm": 0.15877477824687958, "learning_rate": 4.972435804439286e-06, "loss": 0.0106, "step": 26301 }, { "epoch": 2.851474414570685, "grad_norm": 0.04409167915582657, "learning_rate": 4.9688089366023505e-06, "loss": 0.0015, "step": 26302 }, { "epoch": 2.851582827406765, "grad_norm": 0.32312285900115967, "learning_rate": 4.965182068765415e-06, "loss": 0.0143, "step": 26303 }, { "epoch": 2.851691240242845, "grad_norm": 0.13020610809326172, "learning_rate": 4.961555200928479e-06, "loss": 0.0059, "step": 26304 }, { "epoch": 2.8517996530789245, "grad_norm": 0.3607533872127533, "learning_rate": 4.957928333091543e-06, "loss": 0.0252, "step": 26305 }, { "epoch": 2.8519080659150045, "grad_norm": 0.15851791203022003, "learning_rate": 4.954301465254606e-06, "loss": 0.004, "step": 26306 }, { "epoch": 2.852016478751084, "grad_norm": 0.2751329243183136, "learning_rate": 4.95067459741767e-06, "loss": 0.0094, "step": 26307 }, { "epoch": 2.852124891587164, "grad_norm": 0.3058631420135498, "learning_rate": 4.947047729580734e-06, "loss": 0.024, "step": 26308 }, { "epoch": 2.852233304423244, "grad_norm": 0.06715685874223709, "learning_rate": 4.943420861743798e-06, "loss": 0.0023, "step": 26309 }, { "epoch": 2.8523417172593235, "grad_norm": 0.20986928045749664, "learning_rate": 4.9397939939068625e-06, "loss": 0.0069, "step": 26310 }, { "epoch": 2.852450130095403, "grad_norm": 0.6137868762016296, "learning_rate": 4.936167126069927e-06, "loss": 0.0145, "step": 26311 }, { "epoch": 2.852558542931483, "grad_norm": 0.04433555155992508, "learning_rate": 4.93254025823299e-06, "loss": 0.0023, "step": 26312 }, { "epoch": 2.852666955767563, "grad_norm": 0.10260391235351562, "learning_rate": 4.928913390396054e-06, "loss": 0.0032, "step": 26313 }, { "epoch": 2.8527753686036426, "grad_norm": 0.12348484247922897, "learning_rate": 4.925286522559118e-06, "loss": 0.0033, "step": 26314 }, { "epoch": 2.8528837814397225, "grad_norm": 0.21554987132549286, "learning_rate": 4.921659654722182e-06, "loss": 0.0107, "step": 26315 }, { "epoch": 2.852992194275802, "grad_norm": 0.19359730184078217, "learning_rate": 4.918032786885246e-06, "loss": 0.0111, "step": 26316 }, { "epoch": 2.853100607111882, "grad_norm": 0.22198078036308289, "learning_rate": 4.91440591904831e-06, "loss": 0.0039, "step": 26317 }, { "epoch": 2.853209019947962, "grad_norm": 0.062496934086084366, "learning_rate": 4.9107790512113745e-06, "loss": 0.0016, "step": 26318 }, { "epoch": 2.8533174327840416, "grad_norm": 0.14257948100566864, "learning_rate": 4.907152183374438e-06, "loss": 0.0065, "step": 26319 }, { "epoch": 2.8534258456201216, "grad_norm": 0.10009976476430893, "learning_rate": 4.903525315537502e-06, "loss": 0.0045, "step": 26320 }, { "epoch": 2.853534258456201, "grad_norm": 0.5264227986335754, "learning_rate": 4.899898447700566e-06, "loss": 0.0178, "step": 26321 }, { "epoch": 2.853642671292281, "grad_norm": 0.4161526560783386, "learning_rate": 4.89627157986363e-06, "loss": 0.0469, "step": 26322 }, { "epoch": 2.8537510841283606, "grad_norm": 0.1735866516828537, "learning_rate": 4.892644712026694e-06, "loss": 0.0043, "step": 26323 }, { "epoch": 2.8538594969644406, "grad_norm": 0.11552571505308151, "learning_rate": 4.889017844189758e-06, "loss": 0.0044, "step": 26324 }, { "epoch": 2.85396790980052, "grad_norm": 0.1215367242693901, "learning_rate": 4.8853909763528216e-06, "loss": 0.0067, "step": 26325 }, { "epoch": 2.8540763226366, "grad_norm": 0.00512929679825902, "learning_rate": 4.881764108515886e-06, "loss": 0.0002, "step": 26326 }, { "epoch": 2.85418473547268, "grad_norm": 0.5794405937194824, "learning_rate": 4.87813724067895e-06, "loss": 0.0542, "step": 26327 }, { "epoch": 2.8542931483087597, "grad_norm": 0.11420798301696777, "learning_rate": 4.874510372842014e-06, "loss": 0.005, "step": 26328 }, { "epoch": 2.8544015611448397, "grad_norm": 0.030105724930763245, "learning_rate": 4.870883505005078e-06, "loss": 0.0014, "step": 26329 }, { "epoch": 2.854509973980919, "grad_norm": 0.06399211287498474, "learning_rate": 4.867256637168142e-06, "loss": 0.0017, "step": 26330 }, { "epoch": 2.854618386816999, "grad_norm": 0.02638021484017372, "learning_rate": 4.863629769331206e-06, "loss": 0.0007, "step": 26331 }, { "epoch": 2.854726799653079, "grad_norm": 0.33760350942611694, "learning_rate": 4.8600029014942694e-06, "loss": 0.0076, "step": 26332 }, { "epoch": 2.8548352124891587, "grad_norm": 0.3785528540611267, "learning_rate": 4.8563760336573335e-06, "loss": 0.0131, "step": 26333 }, { "epoch": 2.8549436253252383, "grad_norm": 0.20445352792739868, "learning_rate": 4.852749165820398e-06, "loss": 0.006, "step": 26334 }, { "epoch": 2.8550520381613183, "grad_norm": 0.3681899607181549, "learning_rate": 4.849122297983462e-06, "loss": 0.0154, "step": 26335 }, { "epoch": 2.8551604509973982, "grad_norm": 0.11255461722612381, "learning_rate": 4.845495430146526e-06, "loss": 0.0023, "step": 26336 }, { "epoch": 2.8552688638334778, "grad_norm": 0.05817298963665962, "learning_rate": 4.84186856230959e-06, "loss": 0.0016, "step": 26337 }, { "epoch": 2.8553772766695578, "grad_norm": 0.09692100435495377, "learning_rate": 4.838241694472654e-06, "loss": 0.0054, "step": 26338 }, { "epoch": 2.8554856895056373, "grad_norm": 0.18967927992343903, "learning_rate": 4.834614826635717e-06, "loss": 0.0031, "step": 26339 }, { "epoch": 2.8555941023417173, "grad_norm": 0.03687290474772453, "learning_rate": 4.8309879587987814e-06, "loss": 0.0009, "step": 26340 }, { "epoch": 2.8557025151777973, "grad_norm": 0.007551872171461582, "learning_rate": 4.8273610909618455e-06, "loss": 0.0003, "step": 26341 }, { "epoch": 2.855810928013877, "grad_norm": 0.000951021327637136, "learning_rate": 4.82373422312491e-06, "loss": 0.0001, "step": 26342 }, { "epoch": 2.855919340849957, "grad_norm": 0.26651719212532043, "learning_rate": 4.820107355287974e-06, "loss": 0.0091, "step": 26343 }, { "epoch": 2.8560277536860363, "grad_norm": 0.44822606444358826, "learning_rate": 4.816480487451038e-06, "loss": 0.0075, "step": 26344 }, { "epoch": 2.8561361665221163, "grad_norm": 0.2967243194580078, "learning_rate": 4.812853619614101e-06, "loss": 0.0106, "step": 26345 }, { "epoch": 2.856244579358196, "grad_norm": 0.4938386380672455, "learning_rate": 4.809226751777165e-06, "loss": 0.0245, "step": 26346 }, { "epoch": 2.856352992194276, "grad_norm": 0.06200317293405533, "learning_rate": 4.805599883940229e-06, "loss": 0.0037, "step": 26347 }, { "epoch": 2.8564614050303554, "grad_norm": 0.21645799279212952, "learning_rate": 4.8019730161032934e-06, "loss": 0.0057, "step": 26348 }, { "epoch": 2.8565698178664354, "grad_norm": 0.4572121500968933, "learning_rate": 4.7983461482663575e-06, "loss": 0.0113, "step": 26349 }, { "epoch": 2.8566782307025154, "grad_norm": 0.2435360997915268, "learning_rate": 4.794719280429422e-06, "loss": 0.0022, "step": 26350 }, { "epoch": 2.856786643538595, "grad_norm": 0.26418405771255493, "learning_rate": 4.791092412592486e-06, "loss": 0.0057, "step": 26351 }, { "epoch": 2.856895056374675, "grad_norm": 0.3865765333175659, "learning_rate": 4.787465544755549e-06, "loss": 0.0187, "step": 26352 }, { "epoch": 2.8570034692107544, "grad_norm": 0.23961326479911804, "learning_rate": 4.783838676918613e-06, "loss": 0.0037, "step": 26353 }, { "epoch": 2.8571118820468344, "grad_norm": 0.27760881185531616, "learning_rate": 4.780211809081677e-06, "loss": 0.0176, "step": 26354 }, { "epoch": 2.8572202948829144, "grad_norm": 0.39399099349975586, "learning_rate": 4.776584941244741e-06, "loss": 0.0095, "step": 26355 }, { "epoch": 2.857328707718994, "grad_norm": 0.19827552139759064, "learning_rate": 4.7729580734078054e-06, "loss": 0.0028, "step": 26356 }, { "epoch": 2.8574371205550735, "grad_norm": 0.1577875316143036, "learning_rate": 4.7693312055708695e-06, "loss": 0.0055, "step": 26357 }, { "epoch": 2.8575455333911535, "grad_norm": 0.2376677542924881, "learning_rate": 4.765704337733933e-06, "loss": 0.0115, "step": 26358 }, { "epoch": 2.8576539462272335, "grad_norm": 0.042973391711711884, "learning_rate": 4.762077469896997e-06, "loss": 0.0018, "step": 26359 }, { "epoch": 2.857762359063313, "grad_norm": 0.10699529200792313, "learning_rate": 4.758450602060061e-06, "loss": 0.0026, "step": 26360 }, { "epoch": 2.857870771899393, "grad_norm": 0.10848388075828552, "learning_rate": 4.754823734223125e-06, "loss": 0.0129, "step": 26361 }, { "epoch": 2.8579791847354725, "grad_norm": 0.0049741268157958984, "learning_rate": 4.751196866386189e-06, "loss": 0.0002, "step": 26362 }, { "epoch": 2.8580875975715525, "grad_norm": 0.3255486488342285, "learning_rate": 4.747569998549253e-06, "loss": 0.0062, "step": 26363 }, { "epoch": 2.8581960104076325, "grad_norm": 0.32648006081581116, "learning_rate": 4.743943130712317e-06, "loss": 0.0111, "step": 26364 }, { "epoch": 2.858304423243712, "grad_norm": 0.23614223301410675, "learning_rate": 4.7403162628753815e-06, "loss": 0.005, "step": 26365 }, { "epoch": 2.8584128360797916, "grad_norm": 0.11962026357650757, "learning_rate": 4.736689395038445e-06, "loss": 0.0026, "step": 26366 }, { "epoch": 2.8585212489158716, "grad_norm": 0.15916961431503296, "learning_rate": 4.733062527201509e-06, "loss": 0.0039, "step": 26367 }, { "epoch": 2.8586296617519515, "grad_norm": 0.3574700653553009, "learning_rate": 4.729435659364573e-06, "loss": 0.008, "step": 26368 }, { "epoch": 2.858738074588031, "grad_norm": 0.3595740795135498, "learning_rate": 4.725808791527637e-06, "loss": 0.0279, "step": 26369 }, { "epoch": 2.858846487424111, "grad_norm": 0.04156280681490898, "learning_rate": 4.722181923690701e-06, "loss": 0.0016, "step": 26370 }, { "epoch": 2.8589549002601906, "grad_norm": 0.6367396712303162, "learning_rate": 4.7185550558537645e-06, "loss": 0.0091, "step": 26371 }, { "epoch": 2.8590633130962706, "grad_norm": 0.07627832889556885, "learning_rate": 4.7149281880168286e-06, "loss": 0.0029, "step": 26372 }, { "epoch": 2.8591717259323506, "grad_norm": 0.10426957905292511, "learning_rate": 4.711301320179893e-06, "loss": 0.0033, "step": 26373 }, { "epoch": 2.85928013876843, "grad_norm": 0.3758677542209625, "learning_rate": 4.707674452342957e-06, "loss": 0.0043, "step": 26374 }, { "epoch": 2.85938855160451, "grad_norm": 0.4416901767253876, "learning_rate": 4.704047584506021e-06, "loss": 0.0419, "step": 26375 }, { "epoch": 2.8594969644405897, "grad_norm": 0.18509331345558167, "learning_rate": 4.700420716669085e-06, "loss": 0.0051, "step": 26376 }, { "epoch": 2.8596053772766696, "grad_norm": 0.4229997396469116, "learning_rate": 4.696793848832149e-06, "loss": 0.0189, "step": 26377 }, { "epoch": 2.8597137901127496, "grad_norm": 0.04345241189002991, "learning_rate": 4.693166980995213e-06, "loss": 0.0018, "step": 26378 }, { "epoch": 2.859822202948829, "grad_norm": 0.005862313322722912, "learning_rate": 4.6895401131582765e-06, "loss": 0.0002, "step": 26379 }, { "epoch": 2.8599306157849087, "grad_norm": 0.2357911467552185, "learning_rate": 4.6859132453213406e-06, "loss": 0.0059, "step": 26380 }, { "epoch": 2.8600390286209887, "grad_norm": 0.08297745883464813, "learning_rate": 4.682286377484405e-06, "loss": 0.0028, "step": 26381 }, { "epoch": 2.8601474414570687, "grad_norm": 0.2925245463848114, "learning_rate": 4.678659509647469e-06, "loss": 0.0047, "step": 26382 }, { "epoch": 2.860255854293148, "grad_norm": 0.2356266975402832, "learning_rate": 4.675032641810533e-06, "loss": 0.0121, "step": 26383 }, { "epoch": 2.860364267129228, "grad_norm": 0.8315449953079224, "learning_rate": 4.671405773973596e-06, "loss": 0.0364, "step": 26384 }, { "epoch": 2.8604726799653077, "grad_norm": 0.033163830637931824, "learning_rate": 4.66777890613666e-06, "loss": 0.0014, "step": 26385 }, { "epoch": 2.8605810928013877, "grad_norm": 0.0037032237742096186, "learning_rate": 4.664152038299724e-06, "loss": 0.0002, "step": 26386 }, { "epoch": 2.8606895056374677, "grad_norm": 1.04709792137146, "learning_rate": 4.6605251704627884e-06, "loss": 0.0051, "step": 26387 }, { "epoch": 2.8607979184735473, "grad_norm": 0.42211857438087463, "learning_rate": 4.6568983026258526e-06, "loss": 0.0113, "step": 26388 }, { "epoch": 2.860906331309627, "grad_norm": 0.17498177289962769, "learning_rate": 4.653271434788917e-06, "loss": 0.0128, "step": 26389 }, { "epoch": 2.861014744145707, "grad_norm": 0.1873674839735031, "learning_rate": 4.649644566951981e-06, "loss": 0.0033, "step": 26390 }, { "epoch": 2.8611231569817868, "grad_norm": 1.104207158088684, "learning_rate": 4.646017699115045e-06, "loss": 0.0084, "step": 26391 }, { "epoch": 2.8612315698178663, "grad_norm": 0.459433376789093, "learning_rate": 4.642390831278109e-06, "loss": 0.0155, "step": 26392 }, { "epoch": 2.8613399826539463, "grad_norm": 0.2789182662963867, "learning_rate": 4.638763963441172e-06, "loss": 0.0071, "step": 26393 }, { "epoch": 2.861448395490026, "grad_norm": 0.18441897630691528, "learning_rate": 4.635137095604236e-06, "loss": 0.0066, "step": 26394 }, { "epoch": 2.861556808326106, "grad_norm": 0.03784739226102829, "learning_rate": 4.6315102277673004e-06, "loss": 0.0015, "step": 26395 }, { "epoch": 2.861665221162186, "grad_norm": 0.1928882896900177, "learning_rate": 4.6278833599303645e-06, "loss": 0.0103, "step": 26396 }, { "epoch": 2.8617736339982653, "grad_norm": 0.07135850936174393, "learning_rate": 4.624256492093429e-06, "loss": 0.0068, "step": 26397 }, { "epoch": 2.8618820468343453, "grad_norm": 0.1890621930360794, "learning_rate": 4.620629624256492e-06, "loss": 0.0041, "step": 26398 }, { "epoch": 2.861990459670425, "grad_norm": 0.5786687731742859, "learning_rate": 4.617002756419556e-06, "loss": 0.025, "step": 26399 }, { "epoch": 2.862098872506505, "grad_norm": 0.02733904868364334, "learning_rate": 4.61337588858262e-06, "loss": 0.0011, "step": 26400 }, { "epoch": 2.862207285342585, "grad_norm": 0.08959732949733734, "learning_rate": 4.609749020745684e-06, "loss": 0.0021, "step": 26401 }, { "epoch": 2.8623156981786644, "grad_norm": 0.3945440649986267, "learning_rate": 4.606122152908748e-06, "loss": 0.0104, "step": 26402 }, { "epoch": 2.862424111014744, "grad_norm": 0.007671266328543425, "learning_rate": 4.6024952850718124e-06, "loss": 0.0002, "step": 26403 }, { "epoch": 2.862532523850824, "grad_norm": 0.4187735319137573, "learning_rate": 4.5988684172348765e-06, "loss": 0.0334, "step": 26404 }, { "epoch": 2.862640936686904, "grad_norm": 0.006280922330915928, "learning_rate": 4.595241549397941e-06, "loss": 0.0002, "step": 26405 }, { "epoch": 2.8627493495229834, "grad_norm": 0.2898940443992615, "learning_rate": 4.591614681561004e-06, "loss": 0.0141, "step": 26406 }, { "epoch": 2.8628577623590634, "grad_norm": 0.09947916120290756, "learning_rate": 4.587987813724068e-06, "loss": 0.0036, "step": 26407 }, { "epoch": 2.862966175195143, "grad_norm": 0.13565956056118011, "learning_rate": 4.584360945887132e-06, "loss": 0.0042, "step": 26408 }, { "epoch": 2.863074588031223, "grad_norm": 0.6327357292175293, "learning_rate": 4.580734078050196e-06, "loss": 0.0305, "step": 26409 }, { "epoch": 2.863183000867303, "grad_norm": 0.2102048695087433, "learning_rate": 4.57710721021326e-06, "loss": 0.0061, "step": 26410 }, { "epoch": 2.8632914137033825, "grad_norm": 0.20894047617912292, "learning_rate": 4.573480342376324e-06, "loss": 0.0126, "step": 26411 }, { "epoch": 2.863399826539462, "grad_norm": 0.04477998986840248, "learning_rate": 4.569853474539388e-06, "loss": 0.0009, "step": 26412 }, { "epoch": 2.863508239375542, "grad_norm": 0.23746734857559204, "learning_rate": 4.566226606702452e-06, "loss": 0.0141, "step": 26413 }, { "epoch": 2.863616652211622, "grad_norm": 0.3240020275115967, "learning_rate": 4.562599738865516e-06, "loss": 0.0118, "step": 26414 }, { "epoch": 2.8637250650477015, "grad_norm": 0.11444826424121857, "learning_rate": 4.55897287102858e-06, "loss": 0.0061, "step": 26415 }, { "epoch": 2.8638334778837815, "grad_norm": 0.21261310577392578, "learning_rate": 4.555346003191644e-06, "loss": 0.0059, "step": 26416 }, { "epoch": 2.863941890719861, "grad_norm": 0.036756906658411026, "learning_rate": 4.551719135354708e-06, "loss": 0.001, "step": 26417 }, { "epoch": 2.864050303555941, "grad_norm": 0.3196620047092438, "learning_rate": 4.548092267517772e-06, "loss": 0.005, "step": 26418 }, { "epoch": 2.864158716392021, "grad_norm": 0.3595045208930969, "learning_rate": 4.544465399680836e-06, "loss": 0.0432, "step": 26419 }, { "epoch": 2.8642671292281006, "grad_norm": 0.34482118487358093, "learning_rate": 4.5408385318439e-06, "loss": 0.0258, "step": 26420 }, { "epoch": 2.8643755420641805, "grad_norm": 0.19398532807826996, "learning_rate": 4.537211664006964e-06, "loss": 0.012, "step": 26421 }, { "epoch": 2.86448395490026, "grad_norm": 0.45988672971725464, "learning_rate": 4.533584796170028e-06, "loss": 0.028, "step": 26422 }, { "epoch": 2.86459236773634, "grad_norm": 0.8707817196846008, "learning_rate": 4.529957928333092e-06, "loss": 0.0255, "step": 26423 }, { "epoch": 2.8647007805724196, "grad_norm": 0.38314417004585266, "learning_rate": 4.526331060496155e-06, "loss": 0.0158, "step": 26424 }, { "epoch": 2.8648091934084996, "grad_norm": 0.10156438499689102, "learning_rate": 4.522704192659219e-06, "loss": 0.0044, "step": 26425 }, { "epoch": 2.864917606244579, "grad_norm": 0.16390475630760193, "learning_rate": 4.5190773248222835e-06, "loss": 0.0066, "step": 26426 }, { "epoch": 2.865026019080659, "grad_norm": 1.2417727708816528, "learning_rate": 4.5154504569853476e-06, "loss": 0.0297, "step": 26427 }, { "epoch": 2.865134431916739, "grad_norm": 0.4569658637046814, "learning_rate": 4.511823589148412e-06, "loss": 0.008, "step": 26428 }, { "epoch": 2.8652428447528187, "grad_norm": 0.21749211847782135, "learning_rate": 4.508196721311476e-06, "loss": 0.0087, "step": 26429 }, { "epoch": 2.8653512575888986, "grad_norm": 0.03336706385016441, "learning_rate": 4.50456985347454e-06, "loss": 0.0015, "step": 26430 }, { "epoch": 2.865459670424978, "grad_norm": 0.16887971758842468, "learning_rate": 4.500942985637604e-06, "loss": 0.0035, "step": 26431 }, { "epoch": 2.865568083261058, "grad_norm": 0.08326515555381775, "learning_rate": 4.497316117800668e-06, "loss": 0.0014, "step": 26432 }, { "epoch": 2.865676496097138, "grad_norm": 0.1010260283946991, "learning_rate": 4.493689249963731e-06, "loss": 0.0043, "step": 26433 }, { "epoch": 2.8657849089332177, "grad_norm": 0.13578838109970093, "learning_rate": 4.4900623821267955e-06, "loss": 0.0096, "step": 26434 }, { "epoch": 2.8658933217692972, "grad_norm": 0.16494448482990265, "learning_rate": 4.4864355142898596e-06, "loss": 0.0039, "step": 26435 }, { "epoch": 2.866001734605377, "grad_norm": 0.1331080049276352, "learning_rate": 4.482808646452924e-06, "loss": 0.0072, "step": 26436 }, { "epoch": 2.866110147441457, "grad_norm": 0.11402565240859985, "learning_rate": 4.479181778615987e-06, "loss": 0.0111, "step": 26437 }, { "epoch": 2.8662185602775367, "grad_norm": 0.685797929763794, "learning_rate": 4.475554910779051e-06, "loss": 0.0356, "step": 26438 }, { "epoch": 2.8663269731136167, "grad_norm": 0.07673295587301254, "learning_rate": 4.471928042942115e-06, "loss": 0.0027, "step": 26439 }, { "epoch": 2.8664353859496963, "grad_norm": 0.2461184710264206, "learning_rate": 4.468301175105179e-06, "loss": 0.0102, "step": 26440 }, { "epoch": 2.8665437987857763, "grad_norm": 0.013829629868268967, "learning_rate": 4.464674307268243e-06, "loss": 0.0005, "step": 26441 }, { "epoch": 2.8666522116218562, "grad_norm": 0.053907718509435654, "learning_rate": 4.4610474394313075e-06, "loss": 0.0023, "step": 26442 }, { "epoch": 2.866760624457936, "grad_norm": 0.016838109120726585, "learning_rate": 4.4574205715943716e-06, "loss": 0.0006, "step": 26443 }, { "epoch": 2.8668690372940158, "grad_norm": 0.017464209347963333, "learning_rate": 4.453793703757436e-06, "loss": 0.0009, "step": 26444 }, { "epoch": 2.8669774501300953, "grad_norm": 0.24498476088047028, "learning_rate": 4.4501668359205e-06, "loss": 0.0083, "step": 26445 }, { "epoch": 2.8670858629661753, "grad_norm": 0.18181337416172028, "learning_rate": 4.446539968083564e-06, "loss": 0.0017, "step": 26446 }, { "epoch": 2.867194275802255, "grad_norm": 0.5000318884849548, "learning_rate": 4.442913100246627e-06, "loss": 0.0217, "step": 26447 }, { "epoch": 2.867302688638335, "grad_norm": 0.1069079115986824, "learning_rate": 4.439286232409691e-06, "loss": 0.0032, "step": 26448 }, { "epoch": 2.8674111014744144, "grad_norm": 0.16467291116714478, "learning_rate": 4.435659364572755e-06, "loss": 0.0068, "step": 26449 }, { "epoch": 2.8675195143104943, "grad_norm": 0.12103388458490372, "learning_rate": 4.432032496735819e-06, "loss": 0.004, "step": 26450 }, { "epoch": 2.8676279271465743, "grad_norm": 0.23784855008125305, "learning_rate": 4.428405628898883e-06, "loss": 0.0113, "step": 26451 }, { "epoch": 2.867736339982654, "grad_norm": 0.06178809702396393, "learning_rate": 4.424778761061947e-06, "loss": 0.002, "step": 26452 }, { "epoch": 2.867844752818734, "grad_norm": 0.32528021931648254, "learning_rate": 4.421151893225011e-06, "loss": 0.0235, "step": 26453 }, { "epoch": 2.8679531656548134, "grad_norm": 0.18471293151378632, "learning_rate": 4.417525025388075e-06, "loss": 0.004, "step": 26454 }, { "epoch": 2.8680615784908934, "grad_norm": 0.2716190218925476, "learning_rate": 4.413898157551139e-06, "loss": 0.0198, "step": 26455 }, { "epoch": 2.8681699913269734, "grad_norm": 0.31685441732406616, "learning_rate": 4.410271289714203e-06, "loss": 0.0128, "step": 26456 }, { "epoch": 2.868278404163053, "grad_norm": 0.406192809343338, "learning_rate": 4.406644421877267e-06, "loss": 0.0055, "step": 26457 }, { "epoch": 2.8683868169991324, "grad_norm": 0.0012341841356828809, "learning_rate": 4.4030175540403314e-06, "loss": 0.0001, "step": 26458 }, { "epoch": 2.8684952298352124, "grad_norm": 0.15545165538787842, "learning_rate": 4.3993906862033955e-06, "loss": 0.0497, "step": 26459 }, { "epoch": 2.8686036426712924, "grad_norm": 0.14702367782592773, "learning_rate": 4.395763818366459e-06, "loss": 0.0062, "step": 26460 }, { "epoch": 2.868712055507372, "grad_norm": 0.645697295665741, "learning_rate": 4.392136950529523e-06, "loss": 0.0154, "step": 26461 }, { "epoch": 2.868820468343452, "grad_norm": 0.056744229048490524, "learning_rate": 4.388510082692587e-06, "loss": 0.0014, "step": 26462 }, { "epoch": 2.8689288811795315, "grad_norm": 0.4372522532939911, "learning_rate": 4.38488321485565e-06, "loss": 0.0405, "step": 26463 }, { "epoch": 2.8690372940156115, "grad_norm": 0.0666666328907013, "learning_rate": 4.381256347018714e-06, "loss": 0.003, "step": 26464 }, { "epoch": 2.8691457068516915, "grad_norm": 0.7117469310760498, "learning_rate": 4.3776294791817785e-06, "loss": 0.0354, "step": 26465 }, { "epoch": 2.869254119687771, "grad_norm": 0.15652741491794586, "learning_rate": 4.374002611344843e-06, "loss": 0.0065, "step": 26466 }, { "epoch": 2.869362532523851, "grad_norm": 0.07098501920700073, "learning_rate": 4.370375743507907e-06, "loss": 0.0019, "step": 26467 }, { "epoch": 2.8694709453599305, "grad_norm": 0.019774556159973145, "learning_rate": 4.366748875670971e-06, "loss": 0.0005, "step": 26468 }, { "epoch": 2.8695793581960105, "grad_norm": 0.1620386391878128, "learning_rate": 4.363122007834035e-06, "loss": 0.0028, "step": 26469 }, { "epoch": 2.86968777103209, "grad_norm": 1.7757625579833984, "learning_rate": 4.359495139997099e-06, "loss": 0.0209, "step": 26470 }, { "epoch": 2.86979618386817, "grad_norm": 0.8046036958694458, "learning_rate": 4.355868272160163e-06, "loss": 0.0464, "step": 26471 }, { "epoch": 2.8699045967042496, "grad_norm": 0.35758376121520996, "learning_rate": 4.352241404323227e-06, "loss": 0.0271, "step": 26472 }, { "epoch": 2.8700130095403296, "grad_norm": 0.3123522996902466, "learning_rate": 4.348614536486291e-06, "loss": 0.0093, "step": 26473 }, { "epoch": 2.8701214223764095, "grad_norm": 0.10033993422985077, "learning_rate": 4.344987668649355e-06, "loss": 0.0042, "step": 26474 }, { "epoch": 2.870229835212489, "grad_norm": 0.13475532829761505, "learning_rate": 4.341360800812419e-06, "loss": 0.0027, "step": 26475 }, { "epoch": 2.870338248048569, "grad_norm": 0.3401046395301819, "learning_rate": 4.337733932975482e-06, "loss": 0.027, "step": 26476 }, { "epoch": 2.8704466608846486, "grad_norm": 0.5352150797843933, "learning_rate": 4.334107065138546e-06, "loss": 0.0135, "step": 26477 }, { "epoch": 2.8705550737207286, "grad_norm": 0.16812464594841003, "learning_rate": 4.33048019730161e-06, "loss": 0.0098, "step": 26478 }, { "epoch": 2.8706634865568086, "grad_norm": 0.17817988991737366, "learning_rate": 4.326853329464674e-06, "loss": 0.0104, "step": 26479 }, { "epoch": 2.870771899392888, "grad_norm": 0.05831344798207283, "learning_rate": 4.323226461627738e-06, "loss": 0.0014, "step": 26480 }, { "epoch": 2.8708803122289677, "grad_norm": 0.43649783730506897, "learning_rate": 4.3195995937908025e-06, "loss": 0.0166, "step": 26481 }, { "epoch": 2.8709887250650477, "grad_norm": 0.21358689665794373, "learning_rate": 4.3159727259538666e-06, "loss": 0.0199, "step": 26482 }, { "epoch": 2.8710971379011276, "grad_norm": 0.10797823220491409, "learning_rate": 4.312345858116931e-06, "loss": 0.002, "step": 26483 }, { "epoch": 2.871205550737207, "grad_norm": 0.0565427765250206, "learning_rate": 4.308718990279995e-06, "loss": 0.0014, "step": 26484 }, { "epoch": 2.871313963573287, "grad_norm": 0.4867881238460541, "learning_rate": 4.305092122443059e-06, "loss": 0.0255, "step": 26485 }, { "epoch": 2.8714223764093667, "grad_norm": 0.256414532661438, "learning_rate": 4.301465254606123e-06, "loss": 0.0083, "step": 26486 }, { "epoch": 2.8715307892454467, "grad_norm": 0.16508924961090088, "learning_rate": 4.297838386769187e-06, "loss": 0.0033, "step": 26487 }, { "epoch": 2.8716392020815267, "grad_norm": 0.23375600576400757, "learning_rate": 4.29421151893225e-06, "loss": 0.0102, "step": 26488 }, { "epoch": 2.871747614917606, "grad_norm": 0.32375195622444153, "learning_rate": 4.2905846510953145e-06, "loss": 0.0093, "step": 26489 }, { "epoch": 2.871856027753686, "grad_norm": 0.03824548423290253, "learning_rate": 4.286957783258378e-06, "loss": 0.0013, "step": 26490 }, { "epoch": 2.8719644405897657, "grad_norm": 0.14086085557937622, "learning_rate": 4.283330915421442e-06, "loss": 0.0153, "step": 26491 }, { "epoch": 2.8720728534258457, "grad_norm": 0.0533638522028923, "learning_rate": 4.279704047584506e-06, "loss": 0.0014, "step": 26492 }, { "epoch": 2.8721812662619253, "grad_norm": 0.37522047758102417, "learning_rate": 4.27607717974757e-06, "loss": 0.0067, "step": 26493 }, { "epoch": 2.8722896790980053, "grad_norm": 0.03824478015303612, "learning_rate": 4.272450311910634e-06, "loss": 0.0008, "step": 26494 }, { "epoch": 2.872398091934085, "grad_norm": 0.15365269780158997, "learning_rate": 4.268823444073698e-06, "loss": 0.0138, "step": 26495 }, { "epoch": 2.872506504770165, "grad_norm": 0.11242156475782394, "learning_rate": 4.265196576236762e-06, "loss": 0.0035, "step": 26496 }, { "epoch": 2.8726149176062448, "grad_norm": 0.8693903088569641, "learning_rate": 4.2615697083998265e-06, "loss": 0.0297, "step": 26497 }, { "epoch": 2.8727233304423243, "grad_norm": 0.4061163067817688, "learning_rate": 4.2579428405628906e-06, "loss": 0.0101, "step": 26498 }, { "epoch": 2.8728317432784043, "grad_norm": 0.55023193359375, "learning_rate": 4.254315972725955e-06, "loss": 0.0555, "step": 26499 }, { "epoch": 2.872940156114484, "grad_norm": 0.4609758257865906, "learning_rate": 4.250689104889019e-06, "loss": 0.0338, "step": 26500 }, { "epoch": 2.873048568950564, "grad_norm": 0.04196538031101227, "learning_rate": 4.247062237052082e-06, "loss": 0.002, "step": 26501 }, { "epoch": 2.873156981786644, "grad_norm": 0.32567596435546875, "learning_rate": 4.243435369215146e-06, "loss": 0.0168, "step": 26502 }, { "epoch": 2.8732653946227233, "grad_norm": 0.1674070805311203, "learning_rate": 4.239808501378209e-06, "loss": 0.0119, "step": 26503 }, { "epoch": 2.873373807458803, "grad_norm": 0.22634616494178772, "learning_rate": 4.2361816335412735e-06, "loss": 0.0254, "step": 26504 }, { "epoch": 2.873482220294883, "grad_norm": 0.3676835894584656, "learning_rate": 4.232554765704338e-06, "loss": 0.0099, "step": 26505 }, { "epoch": 2.873590633130963, "grad_norm": 0.11335155367851257, "learning_rate": 4.228927897867402e-06, "loss": 0.0021, "step": 26506 }, { "epoch": 2.8736990459670424, "grad_norm": 0.19044756889343262, "learning_rate": 4.225301030030466e-06, "loss": 0.0038, "step": 26507 }, { "epoch": 2.8738074588031224, "grad_norm": 0.020032497122883797, "learning_rate": 4.22167416219353e-06, "loss": 0.0006, "step": 26508 }, { "epoch": 2.873915871639202, "grad_norm": 0.09449711441993713, "learning_rate": 4.218047294356594e-06, "loss": 0.0035, "step": 26509 }, { "epoch": 2.874024284475282, "grad_norm": 0.028824185952544212, "learning_rate": 4.214420426519658e-06, "loss": 0.001, "step": 26510 }, { "epoch": 2.874132697311362, "grad_norm": 0.1521598845720291, "learning_rate": 4.210793558682722e-06, "loss": 0.0085, "step": 26511 }, { "epoch": 2.8742411101474414, "grad_norm": 0.11252488940954208, "learning_rate": 4.207166690845786e-06, "loss": 0.0048, "step": 26512 }, { "epoch": 2.874349522983521, "grad_norm": 0.36301541328430176, "learning_rate": 4.2035398230088504e-06, "loss": 0.0171, "step": 26513 }, { "epoch": 2.874457935819601, "grad_norm": 0.027292512357234955, "learning_rate": 4.199912955171914e-06, "loss": 0.0009, "step": 26514 }, { "epoch": 2.874566348655681, "grad_norm": 0.19698725640773773, "learning_rate": 4.196286087334978e-06, "loss": 0.0169, "step": 26515 }, { "epoch": 2.8746747614917605, "grad_norm": 0.3415163457393646, "learning_rate": 4.192659219498041e-06, "loss": 0.0032, "step": 26516 }, { "epoch": 2.8747831743278405, "grad_norm": 0.1563342958688736, "learning_rate": 4.189032351661105e-06, "loss": 0.0078, "step": 26517 }, { "epoch": 2.87489158716392, "grad_norm": 0.20107023417949677, "learning_rate": 4.185405483824169e-06, "loss": 0.0032, "step": 26518 }, { "epoch": 2.875, "grad_norm": 0.2179519385099411, "learning_rate": 4.181778615987233e-06, "loss": 0.0434, "step": 26519 }, { "epoch": 2.87510841283608, "grad_norm": 0.32567456364631653, "learning_rate": 4.1781517481502975e-06, "loss": 0.0089, "step": 26520 }, { "epoch": 2.8752168256721595, "grad_norm": 0.030590975657105446, "learning_rate": 4.174524880313362e-06, "loss": 0.0005, "step": 26521 }, { "epoch": 2.8753252385082395, "grad_norm": 0.15078744292259216, "learning_rate": 4.170898012476426e-06, "loss": 0.0038, "step": 26522 }, { "epoch": 2.875433651344319, "grad_norm": 0.3587553799152374, "learning_rate": 4.16727114463949e-06, "loss": 0.031, "step": 26523 }, { "epoch": 2.875542064180399, "grad_norm": 0.0026548216119408607, "learning_rate": 4.163644276802554e-06, "loss": 0.0001, "step": 26524 }, { "epoch": 2.875650477016479, "grad_norm": 0.20058295130729675, "learning_rate": 4.160017408965618e-06, "loss": 0.0033, "step": 26525 }, { "epoch": 2.8757588898525586, "grad_norm": 0.39883434772491455, "learning_rate": 4.156390541128682e-06, "loss": 0.0594, "step": 26526 }, { "epoch": 2.875867302688638, "grad_norm": 0.31216204166412354, "learning_rate": 4.152763673291745e-06, "loss": 0.0042, "step": 26527 }, { "epoch": 2.875975715524718, "grad_norm": 0.3700256943702698, "learning_rate": 4.1491368054548095e-06, "loss": 0.0084, "step": 26528 }, { "epoch": 2.876084128360798, "grad_norm": 0.27925196290016174, "learning_rate": 4.145509937617873e-06, "loss": 0.0065, "step": 26529 }, { "epoch": 2.8761925411968776, "grad_norm": 0.19852448999881744, "learning_rate": 4.141883069780937e-06, "loss": 0.0046, "step": 26530 }, { "epoch": 2.8763009540329576, "grad_norm": 0.2719900608062744, "learning_rate": 4.138256201944001e-06, "loss": 0.0125, "step": 26531 }, { "epoch": 2.876409366869037, "grad_norm": 0.5523450374603271, "learning_rate": 4.134629334107065e-06, "loss": 0.0102, "step": 26532 }, { "epoch": 2.876517779705117, "grad_norm": 0.17212702333927155, "learning_rate": 4.131002466270129e-06, "loss": 0.0078, "step": 26533 }, { "epoch": 2.876626192541197, "grad_norm": 0.6090685725212097, "learning_rate": 4.127375598433193e-06, "loss": 0.0234, "step": 26534 }, { "epoch": 2.8767346053772767, "grad_norm": 0.48762452602386475, "learning_rate": 4.123748730596257e-06, "loss": 0.022, "step": 26535 }, { "epoch": 2.876843018213356, "grad_norm": 0.30101606249809265, "learning_rate": 4.1201218627593215e-06, "loss": 0.0073, "step": 26536 }, { "epoch": 2.876951431049436, "grad_norm": 0.20235538482666016, "learning_rate": 4.1164949949223856e-06, "loss": 0.0056, "step": 26537 }, { "epoch": 2.877059843885516, "grad_norm": 0.44745272397994995, "learning_rate": 4.11286812708545e-06, "loss": 0.0203, "step": 26538 }, { "epoch": 2.8771682567215957, "grad_norm": 0.06015089899301529, "learning_rate": 4.109241259248514e-06, "loss": 0.0025, "step": 26539 }, { "epoch": 2.8772766695576757, "grad_norm": 0.07880505174398422, "learning_rate": 4.105614391411577e-06, "loss": 0.0024, "step": 26540 }, { "epoch": 2.8773850823937552, "grad_norm": 0.11066824197769165, "learning_rate": 4.101987523574641e-06, "loss": 0.0054, "step": 26541 }, { "epoch": 2.877493495229835, "grad_norm": 0.537844181060791, "learning_rate": 4.098360655737704e-06, "loss": 0.0175, "step": 26542 }, { "epoch": 2.877601908065915, "grad_norm": 0.1665395349264145, "learning_rate": 4.0947337879007685e-06, "loss": 0.0055, "step": 26543 }, { "epoch": 2.8777103209019947, "grad_norm": 0.28430578112602234, "learning_rate": 4.091106920063833e-06, "loss": 0.0224, "step": 26544 }, { "epoch": 2.8778187337380747, "grad_norm": 0.017868688330054283, "learning_rate": 4.087480052226897e-06, "loss": 0.0009, "step": 26545 }, { "epoch": 2.8779271465741543, "grad_norm": 0.2271568775177002, "learning_rate": 4.083853184389961e-06, "loss": 0.0106, "step": 26546 }, { "epoch": 2.8780355594102343, "grad_norm": 0.15610907971858978, "learning_rate": 4.080226316553025e-06, "loss": 0.0048, "step": 26547 }, { "epoch": 2.878143972246314, "grad_norm": 0.04084246978163719, "learning_rate": 4.076599448716089e-06, "loss": 0.0018, "step": 26548 }, { "epoch": 2.878252385082394, "grad_norm": 0.0062080686911940575, "learning_rate": 4.072972580879153e-06, "loss": 0.0002, "step": 26549 }, { "epoch": 2.8783607979184733, "grad_norm": 0.254801481962204, "learning_rate": 4.069345713042217e-06, "loss": 0.0038, "step": 26550 }, { "epoch": 2.8784692107545533, "grad_norm": 0.14210863411426544, "learning_rate": 4.065718845205281e-06, "loss": 0.0073, "step": 26551 }, { "epoch": 2.8785776235906333, "grad_norm": 0.01637694612145424, "learning_rate": 4.0620919773683455e-06, "loss": 0.0004, "step": 26552 }, { "epoch": 2.878686036426713, "grad_norm": 0.2127731293439865, "learning_rate": 4.058465109531409e-06, "loss": 0.009, "step": 26553 }, { "epoch": 2.878794449262793, "grad_norm": 0.6770634651184082, "learning_rate": 4.054838241694473e-06, "loss": 0.0658, "step": 26554 }, { "epoch": 2.8789028620988724, "grad_norm": 0.22603274881839752, "learning_rate": 4.051211373857536e-06, "loss": 0.0084, "step": 26555 }, { "epoch": 2.8790112749349523, "grad_norm": 0.4792407155036926, "learning_rate": 4.0475845060206e-06, "loss": 0.0367, "step": 26556 }, { "epoch": 2.8791196877710323, "grad_norm": 0.06295127421617508, "learning_rate": 4.043957638183664e-06, "loss": 0.0014, "step": 26557 }, { "epoch": 2.879228100607112, "grad_norm": 0.643528401851654, "learning_rate": 4.040330770346728e-06, "loss": 0.0094, "step": 26558 }, { "epoch": 2.8793365134431914, "grad_norm": 0.016165869310498238, "learning_rate": 4.0367039025097925e-06, "loss": 0.0004, "step": 26559 }, { "epoch": 2.8794449262792714, "grad_norm": 0.13841204345226288, "learning_rate": 4.033077034672857e-06, "loss": 0.0039, "step": 26560 }, { "epoch": 2.8795533391153514, "grad_norm": 0.26583123207092285, "learning_rate": 4.029450166835921e-06, "loss": 0.0177, "step": 26561 }, { "epoch": 2.879661751951431, "grad_norm": 0.4515591859817505, "learning_rate": 4.025823298998985e-06, "loss": 0.0182, "step": 26562 }, { "epoch": 2.879770164787511, "grad_norm": 0.16282109916210175, "learning_rate": 4.022196431162049e-06, "loss": 0.0045, "step": 26563 }, { "epoch": 2.8798785776235905, "grad_norm": 0.4221813678741455, "learning_rate": 4.018569563325113e-06, "loss": 0.0176, "step": 26564 }, { "epoch": 2.8799869904596704, "grad_norm": 0.26200661063194275, "learning_rate": 4.014942695488177e-06, "loss": 0.0057, "step": 26565 }, { "epoch": 2.8800954032957504, "grad_norm": 0.023730967193841934, "learning_rate": 4.01131582765124e-06, "loss": 0.0003, "step": 26566 }, { "epoch": 2.88020381613183, "grad_norm": 0.10112445801496506, "learning_rate": 4.0076889598143045e-06, "loss": 0.0042, "step": 26567 }, { "epoch": 2.88031222896791, "grad_norm": 0.07112760841846466, "learning_rate": 4.004062091977369e-06, "loss": 0.0032, "step": 26568 }, { "epoch": 2.8804206418039895, "grad_norm": 0.07443796843290329, "learning_rate": 4.000435224140432e-06, "loss": 0.002, "step": 26569 }, { "epoch": 2.8805290546400695, "grad_norm": 0.40597832202911377, "learning_rate": 3.996808356303496e-06, "loss": 0.0195, "step": 26570 }, { "epoch": 2.880637467476149, "grad_norm": 0.3449859619140625, "learning_rate": 3.99318148846656e-06, "loss": 0.0201, "step": 26571 }, { "epoch": 2.880745880312229, "grad_norm": 0.4885784685611725, "learning_rate": 3.989554620629624e-06, "loss": 0.0129, "step": 26572 }, { "epoch": 2.8808542931483085, "grad_norm": 0.13784189522266388, "learning_rate": 3.985927752792688e-06, "loss": 0.0064, "step": 26573 }, { "epoch": 2.8809627059843885, "grad_norm": 0.4629564881324768, "learning_rate": 3.982300884955752e-06, "loss": 0.0187, "step": 26574 }, { "epoch": 2.8810711188204685, "grad_norm": 0.033136337995529175, "learning_rate": 3.9786740171188165e-06, "loss": 0.0008, "step": 26575 }, { "epoch": 2.881179531656548, "grad_norm": 0.35764428973197937, "learning_rate": 3.975047149281881e-06, "loss": 0.0224, "step": 26576 }, { "epoch": 2.881287944492628, "grad_norm": 0.28005945682525635, "learning_rate": 3.971420281444945e-06, "loss": 0.0116, "step": 26577 }, { "epoch": 2.8813963573287076, "grad_norm": 0.3660005033016205, "learning_rate": 3.967793413608009e-06, "loss": 0.0043, "step": 26578 }, { "epoch": 2.8815047701647876, "grad_norm": 0.3955487310886383, "learning_rate": 3.964166545771073e-06, "loss": 0.0112, "step": 26579 }, { "epoch": 2.8816131830008676, "grad_norm": 0.10259187966585159, "learning_rate": 3.960539677934136e-06, "loss": 0.0015, "step": 26580 }, { "epoch": 2.881721595836947, "grad_norm": 0.04573582857847214, "learning_rate": 3.9569128100972e-06, "loss": 0.0019, "step": 26581 }, { "epoch": 2.8818300086730266, "grad_norm": 0.42445141077041626, "learning_rate": 3.9532859422602635e-06, "loss": 0.0099, "step": 26582 }, { "epoch": 2.8819384215091066, "grad_norm": 0.1281522959470749, "learning_rate": 3.949659074423328e-06, "loss": 0.0047, "step": 26583 }, { "epoch": 2.8820468343451866, "grad_norm": 0.04752945527434349, "learning_rate": 3.946032206586392e-06, "loss": 0.0009, "step": 26584 }, { "epoch": 2.882155247181266, "grad_norm": 0.35892021656036377, "learning_rate": 3.942405338749456e-06, "loss": 0.0178, "step": 26585 }, { "epoch": 2.882263660017346, "grad_norm": 0.29237034916877747, "learning_rate": 3.93877847091252e-06, "loss": 0.0119, "step": 26586 }, { "epoch": 2.8823720728534257, "grad_norm": 0.9998631477355957, "learning_rate": 3.935151603075584e-06, "loss": 0.0139, "step": 26587 }, { "epoch": 2.8824804856895057, "grad_norm": 0.5024252533912659, "learning_rate": 3.931524735238648e-06, "loss": 0.0323, "step": 26588 }, { "epoch": 2.8825888985255856, "grad_norm": 0.5339805483818054, "learning_rate": 3.927897867401712e-06, "loss": 0.0434, "step": 26589 }, { "epoch": 2.882697311361665, "grad_norm": 0.04459188133478165, "learning_rate": 3.924270999564776e-06, "loss": 0.0016, "step": 26590 }, { "epoch": 2.882805724197745, "grad_norm": 0.11842811852693558, "learning_rate": 3.9206441317278405e-06, "loss": 0.0034, "step": 26591 }, { "epoch": 2.8829141370338247, "grad_norm": 0.060768354684114456, "learning_rate": 3.917017263890905e-06, "loss": 0.0014, "step": 26592 }, { "epoch": 2.8830225498699047, "grad_norm": 0.014447192661464214, "learning_rate": 3.913390396053968e-06, "loss": 0.0004, "step": 26593 }, { "epoch": 2.8831309627059842, "grad_norm": 0.02398042380809784, "learning_rate": 3.909763528217032e-06, "loss": 0.0011, "step": 26594 }, { "epoch": 2.883239375542064, "grad_norm": 0.2170247882604599, "learning_rate": 3.906136660380096e-06, "loss": 0.0075, "step": 26595 }, { "epoch": 2.8833477883781438, "grad_norm": 0.3542724549770355, "learning_rate": 3.902509792543159e-06, "loss": 0.0079, "step": 26596 }, { "epoch": 2.8834562012142237, "grad_norm": 0.18869571387767792, "learning_rate": 3.898882924706223e-06, "loss": 0.0132, "step": 26597 }, { "epoch": 2.8835646140503037, "grad_norm": 0.14809726178646088, "learning_rate": 3.8952560568692875e-06, "loss": 0.0042, "step": 26598 }, { "epoch": 2.8836730268863833, "grad_norm": 0.002546598669141531, "learning_rate": 3.891629189032352e-06, "loss": 0.0001, "step": 26599 }, { "epoch": 2.8837814397224633, "grad_norm": 0.18808384239673615, "learning_rate": 3.888002321195416e-06, "loss": 0.0067, "step": 26600 }, { "epoch": 2.883889852558543, "grad_norm": 0.13170433044433594, "learning_rate": 3.88437545335848e-06, "loss": 0.004, "step": 26601 }, { "epoch": 2.883998265394623, "grad_norm": 0.5960659980773926, "learning_rate": 3.880748585521544e-06, "loss": 0.0353, "step": 26602 }, { "epoch": 2.8841066782307028, "grad_norm": 0.14729386568069458, "learning_rate": 3.877121717684608e-06, "loss": 0.0036, "step": 26603 }, { "epoch": 2.8842150910667823, "grad_norm": 0.13657249510288239, "learning_rate": 3.873494849847672e-06, "loss": 0.0028, "step": 26604 }, { "epoch": 2.884323503902862, "grad_norm": 0.19188576936721802, "learning_rate": 3.869867982010736e-06, "loss": 0.0049, "step": 26605 }, { "epoch": 2.884431916738942, "grad_norm": 0.905070960521698, "learning_rate": 3.8662411141737995e-06, "loss": 0.0148, "step": 26606 }, { "epoch": 2.884540329575022, "grad_norm": 0.13471084833145142, "learning_rate": 3.862614246336864e-06, "loss": 0.005, "step": 26607 }, { "epoch": 2.8846487424111014, "grad_norm": 0.31359153985977173, "learning_rate": 3.858987378499928e-06, "loss": 0.0061, "step": 26608 }, { "epoch": 2.8847571552471813, "grad_norm": 0.15520530939102173, "learning_rate": 3.855360510662991e-06, "loss": 0.011, "step": 26609 }, { "epoch": 2.884865568083261, "grad_norm": 0.06434861570596695, "learning_rate": 3.851733642826055e-06, "loss": 0.0032, "step": 26610 }, { "epoch": 2.884973980919341, "grad_norm": 0.295117050409317, "learning_rate": 3.848106774989119e-06, "loss": 0.0125, "step": 26611 }, { "epoch": 2.885082393755421, "grad_norm": 0.1525033712387085, "learning_rate": 3.844479907152183e-06, "loss": 0.0075, "step": 26612 }, { "epoch": 2.8851908065915004, "grad_norm": 0.005896999035030603, "learning_rate": 3.840853039315247e-06, "loss": 0.0002, "step": 26613 }, { "epoch": 2.8852992194275804, "grad_norm": 0.45803597569465637, "learning_rate": 3.8372261714783115e-06, "loss": 0.0083, "step": 26614 }, { "epoch": 2.88540763226366, "grad_norm": 0.26463037729263306, "learning_rate": 3.833599303641376e-06, "loss": 0.0092, "step": 26615 }, { "epoch": 2.88551604509974, "grad_norm": 1.5319217443466187, "learning_rate": 3.82997243580444e-06, "loss": 0.03, "step": 26616 }, { "epoch": 2.8856244579358195, "grad_norm": 0.5991218686103821, "learning_rate": 3.826345567967504e-06, "loss": 0.0245, "step": 26617 }, { "epoch": 2.8857328707718994, "grad_norm": 0.316794216632843, "learning_rate": 3.822718700130568e-06, "loss": 0.0078, "step": 26618 }, { "epoch": 2.885841283607979, "grad_norm": 0.4469429850578308, "learning_rate": 3.819091832293631e-06, "loss": 0.0187, "step": 26619 }, { "epoch": 2.885949696444059, "grad_norm": 0.049045223742723465, "learning_rate": 3.815464964456695e-06, "loss": 0.0008, "step": 26620 }, { "epoch": 2.886058109280139, "grad_norm": 0.46846961975097656, "learning_rate": 3.81183809661976e-06, "loss": 0.0192, "step": 26621 }, { "epoch": 2.8861665221162185, "grad_norm": 0.23838821053504944, "learning_rate": 3.8082112287828235e-06, "loss": 0.0067, "step": 26622 }, { "epoch": 2.8862749349522985, "grad_norm": 0.22829145193099976, "learning_rate": 3.804584360945887e-06, "loss": 0.0081, "step": 26623 }, { "epoch": 2.886383347788378, "grad_norm": 0.34516310691833496, "learning_rate": 3.800957493108951e-06, "loss": 0.0179, "step": 26624 }, { "epoch": 2.886491760624458, "grad_norm": 0.2321377694606781, "learning_rate": 3.797330625272015e-06, "loss": 0.0051, "step": 26625 }, { "epoch": 2.886600173460538, "grad_norm": 0.14357851445674896, "learning_rate": 3.793703757435079e-06, "loss": 0.0052, "step": 26626 }, { "epoch": 2.8867085862966175, "grad_norm": 0.10130764544010162, "learning_rate": 3.790076889598143e-06, "loss": 0.0043, "step": 26627 }, { "epoch": 2.886816999132697, "grad_norm": 0.21379214525222778, "learning_rate": 3.7864500217612073e-06, "loss": 0.0143, "step": 26628 }, { "epoch": 2.886925411968777, "grad_norm": 0.30759352445602417, "learning_rate": 3.782823153924271e-06, "loss": 0.0056, "step": 26629 }, { "epoch": 2.887033824804857, "grad_norm": 0.06840921193361282, "learning_rate": 3.779196286087335e-06, "loss": 0.0019, "step": 26630 }, { "epoch": 2.8871422376409366, "grad_norm": 0.30573102831840515, "learning_rate": 3.775569418250399e-06, "loss": 0.0083, "step": 26631 }, { "epoch": 2.8872506504770166, "grad_norm": 0.27347850799560547, "learning_rate": 3.7719425504134633e-06, "loss": 0.0091, "step": 26632 }, { "epoch": 2.887359063313096, "grad_norm": 0.14557522535324097, "learning_rate": 3.7683156825765274e-06, "loss": 0.0061, "step": 26633 }, { "epoch": 2.887467476149176, "grad_norm": 0.46600207686424255, "learning_rate": 3.7646888147395915e-06, "loss": 0.0187, "step": 26634 }, { "epoch": 2.887575888985256, "grad_norm": 0.11028841882944107, "learning_rate": 3.761061946902655e-06, "loss": 0.0025, "step": 26635 }, { "epoch": 2.8876843018213356, "grad_norm": 0.5997940301895142, "learning_rate": 3.7574350790657193e-06, "loss": 0.0209, "step": 26636 }, { "epoch": 2.887792714657415, "grad_norm": 0.04085668548941612, "learning_rate": 3.7538082112287825e-06, "loss": 0.0019, "step": 26637 }, { "epoch": 2.887901127493495, "grad_norm": 0.3215525150299072, "learning_rate": 3.7501813433918466e-06, "loss": 0.0114, "step": 26638 }, { "epoch": 2.888009540329575, "grad_norm": 0.026551522314548492, "learning_rate": 3.7465544755549107e-06, "loss": 0.0014, "step": 26639 }, { "epoch": 2.8881179531656547, "grad_norm": 0.4520193040370941, "learning_rate": 3.742927607717975e-06, "loss": 0.0186, "step": 26640 }, { "epoch": 2.8882263660017347, "grad_norm": 0.6772012114524841, "learning_rate": 3.739300739881039e-06, "loss": 0.0248, "step": 26641 }, { "epoch": 2.888334778837814, "grad_norm": 0.04329332709312439, "learning_rate": 3.735673872044103e-06, "loss": 0.0019, "step": 26642 }, { "epoch": 2.888443191673894, "grad_norm": 0.2520647346973419, "learning_rate": 3.7320470042071667e-06, "loss": 0.0138, "step": 26643 }, { "epoch": 2.888551604509974, "grad_norm": 0.02307090535759926, "learning_rate": 3.728420136370231e-06, "loss": 0.0008, "step": 26644 }, { "epoch": 2.8886600173460537, "grad_norm": 0.27458977699279785, "learning_rate": 3.724793268533295e-06, "loss": 0.0143, "step": 26645 }, { "epoch": 2.8887684301821337, "grad_norm": 0.037260256707668304, "learning_rate": 3.721166400696359e-06, "loss": 0.0014, "step": 26646 }, { "epoch": 2.8888768430182132, "grad_norm": 0.09846467524766922, "learning_rate": 3.717539532859423e-06, "loss": 0.0046, "step": 26647 }, { "epoch": 2.888985255854293, "grad_norm": 0.5525621771812439, "learning_rate": 3.713912665022487e-06, "loss": 0.0245, "step": 26648 }, { "epoch": 2.889093668690373, "grad_norm": 0.2868046164512634, "learning_rate": 3.710285797185551e-06, "loss": 0.0058, "step": 26649 }, { "epoch": 2.8892020815264527, "grad_norm": 0.10946479439735413, "learning_rate": 3.706658929348614e-06, "loss": 0.0043, "step": 26650 }, { "epoch": 2.8893104943625323, "grad_norm": 0.021424060687422752, "learning_rate": 3.7030320615116783e-06, "loss": 0.0012, "step": 26651 }, { "epoch": 2.8894189071986123, "grad_norm": 0.24890592694282532, "learning_rate": 3.6994051936747424e-06, "loss": 0.0231, "step": 26652 }, { "epoch": 2.8895273200346923, "grad_norm": 0.06789553910493851, "learning_rate": 3.6957783258378065e-06, "loss": 0.0026, "step": 26653 }, { "epoch": 2.889635732870772, "grad_norm": 0.5304775238037109, "learning_rate": 3.6921514580008706e-06, "loss": 0.0335, "step": 26654 }, { "epoch": 2.889744145706852, "grad_norm": 0.09950969368219376, "learning_rate": 3.6885245901639347e-06, "loss": 0.003, "step": 26655 }, { "epoch": 2.8898525585429313, "grad_norm": 0.25432702898979187, "learning_rate": 3.6848977223269984e-06, "loss": 0.0158, "step": 26656 }, { "epoch": 2.8899609713790113, "grad_norm": 0.7455123662948608, "learning_rate": 3.6812708544900625e-06, "loss": 0.0125, "step": 26657 }, { "epoch": 2.8900693842150913, "grad_norm": 0.5846641063690186, "learning_rate": 3.6776439866531266e-06, "loss": 0.0291, "step": 26658 }, { "epoch": 2.890177797051171, "grad_norm": 0.08204476535320282, "learning_rate": 3.6740171188161907e-06, "loss": 0.0043, "step": 26659 }, { "epoch": 2.8902862098872504, "grad_norm": 1.028001308441162, "learning_rate": 3.670390250979255e-06, "loss": 0.0273, "step": 26660 }, { "epoch": 2.8903946227233304, "grad_norm": 0.4101198613643646, "learning_rate": 3.6667633831423185e-06, "loss": 0.0145, "step": 26661 }, { "epoch": 2.8905030355594103, "grad_norm": 0.12192638218402863, "learning_rate": 3.6631365153053826e-06, "loss": 0.0026, "step": 26662 }, { "epoch": 2.89061144839549, "grad_norm": 0.2603270709514618, "learning_rate": 3.6595096474684467e-06, "loss": 0.0085, "step": 26663 }, { "epoch": 2.89071986123157, "grad_norm": 0.5702664256095886, "learning_rate": 3.65588277963151e-06, "loss": 0.0331, "step": 26664 }, { "epoch": 2.8908282740676494, "grad_norm": 0.20073416829109192, "learning_rate": 3.652255911794574e-06, "loss": 0.004, "step": 26665 }, { "epoch": 2.8909366869037294, "grad_norm": 0.30001088976860046, "learning_rate": 3.648629043957638e-06, "loss": 0.0145, "step": 26666 }, { "epoch": 2.8910450997398094, "grad_norm": 0.7866137623786926, "learning_rate": 3.6450021761207023e-06, "loss": 0.0302, "step": 26667 }, { "epoch": 2.891153512575889, "grad_norm": 0.14016355574131012, "learning_rate": 3.6413753082837664e-06, "loss": 0.003, "step": 26668 }, { "epoch": 2.891261925411969, "grad_norm": 0.14673557877540588, "learning_rate": 3.63774844044683e-06, "loss": 0.0045, "step": 26669 }, { "epoch": 2.8913703382480485, "grad_norm": 0.29000595211982727, "learning_rate": 3.634121572609894e-06, "loss": 0.006, "step": 26670 }, { "epoch": 2.8914787510841284, "grad_norm": 0.11437948048114777, "learning_rate": 3.6304947047729583e-06, "loss": 0.004, "step": 26671 }, { "epoch": 2.8915871639202084, "grad_norm": 0.012906131334602833, "learning_rate": 3.6268678369360224e-06, "loss": 0.0005, "step": 26672 }, { "epoch": 2.891695576756288, "grad_norm": 0.47628024220466614, "learning_rate": 3.6232409690990865e-06, "loss": 0.0122, "step": 26673 }, { "epoch": 2.8918039895923675, "grad_norm": 0.3389463722705841, "learning_rate": 3.61961410126215e-06, "loss": 0.0135, "step": 26674 }, { "epoch": 2.8919124024284475, "grad_norm": 0.0979042574763298, "learning_rate": 3.6159872334252143e-06, "loss": 0.0082, "step": 26675 }, { "epoch": 2.8920208152645275, "grad_norm": 0.28389325737953186, "learning_rate": 3.6123603655882784e-06, "loss": 0.0165, "step": 26676 }, { "epoch": 2.892129228100607, "grad_norm": 0.037683501839637756, "learning_rate": 3.6087334977513417e-06, "loss": 0.001, "step": 26677 }, { "epoch": 2.892237640936687, "grad_norm": 0.021385669708251953, "learning_rate": 3.6051066299144058e-06, "loss": 0.0008, "step": 26678 }, { "epoch": 2.8923460537727665, "grad_norm": 0.1105828806757927, "learning_rate": 3.60147976207747e-06, "loss": 0.0073, "step": 26679 }, { "epoch": 2.8924544666088465, "grad_norm": 0.25753480195999146, "learning_rate": 3.597852894240534e-06, "loss": 0.0128, "step": 26680 }, { "epoch": 2.8925628794449265, "grad_norm": 0.07374598830938339, "learning_rate": 3.594226026403598e-06, "loss": 0.0026, "step": 26681 }, { "epoch": 2.892671292281006, "grad_norm": 0.06502015143632889, "learning_rate": 3.5905991585666618e-06, "loss": 0.0028, "step": 26682 }, { "epoch": 2.8927797051170856, "grad_norm": 0.09537963569164276, "learning_rate": 3.586972290729726e-06, "loss": 0.0055, "step": 26683 }, { "epoch": 2.8928881179531656, "grad_norm": 0.18022289872169495, "learning_rate": 3.58334542289279e-06, "loss": 0.0149, "step": 26684 }, { "epoch": 2.8929965307892456, "grad_norm": 0.16595318913459778, "learning_rate": 3.579718555055854e-06, "loss": 0.0039, "step": 26685 }, { "epoch": 2.893104943625325, "grad_norm": 0.19411197304725647, "learning_rate": 3.576091687218918e-06, "loss": 0.0033, "step": 26686 }, { "epoch": 2.893213356461405, "grad_norm": 0.09600167721509933, "learning_rate": 3.5724648193819823e-06, "loss": 0.0018, "step": 26687 }, { "epoch": 2.8933217692974846, "grad_norm": 0.19563712179660797, "learning_rate": 3.568837951545046e-06, "loss": 0.045, "step": 26688 }, { "epoch": 2.8934301821335646, "grad_norm": 0.07289344817399979, "learning_rate": 3.56521108370811e-06, "loss": 0.0027, "step": 26689 }, { "epoch": 2.8935385949696446, "grad_norm": 0.09913992136716843, "learning_rate": 3.561584215871174e-06, "loss": 0.0016, "step": 26690 }, { "epoch": 2.893647007805724, "grad_norm": 0.020114107057452202, "learning_rate": 3.5579573480342374e-06, "loss": 0.0008, "step": 26691 }, { "epoch": 2.893755420641804, "grad_norm": 0.2257937639951706, "learning_rate": 3.5543304801973015e-06, "loss": 0.0091, "step": 26692 }, { "epoch": 2.8938638334778837, "grad_norm": 0.15305693447589874, "learning_rate": 3.5507036123603656e-06, "loss": 0.0109, "step": 26693 }, { "epoch": 2.8939722463139637, "grad_norm": 0.3131386339664459, "learning_rate": 3.5470767445234297e-06, "loss": 0.0365, "step": 26694 }, { "epoch": 2.894080659150043, "grad_norm": 0.16707704961299896, "learning_rate": 3.5434498766864934e-06, "loss": 0.0041, "step": 26695 }, { "epoch": 2.894189071986123, "grad_norm": 0.08168575912714005, "learning_rate": 3.5398230088495575e-06, "loss": 0.0027, "step": 26696 }, { "epoch": 2.8942974848222027, "grad_norm": 0.20598113536834717, "learning_rate": 3.5361961410126216e-06, "loss": 0.0059, "step": 26697 }, { "epoch": 2.8944058976582827, "grad_norm": 0.043759822845458984, "learning_rate": 3.5325692731756857e-06, "loss": 0.0023, "step": 26698 }, { "epoch": 2.8945143104943627, "grad_norm": 0.03668353706598282, "learning_rate": 3.52894240533875e-06, "loss": 0.0016, "step": 26699 }, { "epoch": 2.8946227233304422, "grad_norm": 0.21364009380340576, "learning_rate": 3.525315537501814e-06, "loss": 0.0049, "step": 26700 }, { "epoch": 2.8947311361665222, "grad_norm": 0.3563852608203888, "learning_rate": 3.5216886696648776e-06, "loss": 0.0288, "step": 26701 }, { "epoch": 2.8948395490026018, "grad_norm": 0.1160755306482315, "learning_rate": 3.5180618018279417e-06, "loss": 0.0019, "step": 26702 }, { "epoch": 2.8949479618386817, "grad_norm": 0.6005909442901611, "learning_rate": 3.514434933991006e-06, "loss": 0.059, "step": 26703 }, { "epoch": 2.8950563746747617, "grad_norm": 0.2742411494255066, "learning_rate": 3.510808066154069e-06, "loss": 0.014, "step": 26704 }, { "epoch": 2.8951647875108413, "grad_norm": 0.12199003994464874, "learning_rate": 3.507181198317133e-06, "loss": 0.0053, "step": 26705 }, { "epoch": 2.895273200346921, "grad_norm": 0.2916029095649719, "learning_rate": 3.5035543304801973e-06, "loss": 0.0039, "step": 26706 }, { "epoch": 2.895381613183001, "grad_norm": 0.1872485876083374, "learning_rate": 3.4999274626432614e-06, "loss": 0.0043, "step": 26707 }, { "epoch": 2.895490026019081, "grad_norm": 0.4431307315826416, "learning_rate": 3.496300594806325e-06, "loss": 0.0246, "step": 26708 }, { "epoch": 2.8955984388551603, "grad_norm": 0.12323419749736786, "learning_rate": 3.492673726969389e-06, "loss": 0.0028, "step": 26709 }, { "epoch": 2.8957068516912403, "grad_norm": 0.12036960572004318, "learning_rate": 3.4890468591324533e-06, "loss": 0.0077, "step": 26710 }, { "epoch": 2.89581526452732, "grad_norm": 0.7562093138694763, "learning_rate": 3.4854199912955174e-06, "loss": 0.0732, "step": 26711 }, { "epoch": 2.8959236773634, "grad_norm": 0.019258510321378708, "learning_rate": 3.4817931234585815e-06, "loss": 0.0003, "step": 26712 }, { "epoch": 2.89603209019948, "grad_norm": 0.4284766614437103, "learning_rate": 3.4781662556216456e-06, "loss": 0.0043, "step": 26713 }, { "epoch": 2.8961405030355594, "grad_norm": 0.5656800866127014, "learning_rate": 3.4745393877847093e-06, "loss": 0.015, "step": 26714 }, { "epoch": 2.8962489158716394, "grad_norm": 0.021169370040297508, "learning_rate": 3.4709125199477734e-06, "loss": 0.0009, "step": 26715 }, { "epoch": 2.896357328707719, "grad_norm": 0.017451532185077667, "learning_rate": 3.4672856521108375e-06, "loss": 0.0007, "step": 26716 }, { "epoch": 2.896465741543799, "grad_norm": 0.3013531267642975, "learning_rate": 3.4636587842739016e-06, "loss": 0.0353, "step": 26717 }, { "epoch": 2.8965741543798784, "grad_norm": 0.10919179767370224, "learning_rate": 3.460031916436965e-06, "loss": 0.0029, "step": 26718 }, { "epoch": 2.8966825672159584, "grad_norm": 0.08904863148927689, "learning_rate": 3.456405048600029e-06, "loss": 0.0075, "step": 26719 }, { "epoch": 2.896790980052038, "grad_norm": 0.3016156852245331, "learning_rate": 3.452778180763093e-06, "loss": 0.0114, "step": 26720 }, { "epoch": 2.896899392888118, "grad_norm": 0.06391920149326324, "learning_rate": 3.4491513129261568e-06, "loss": 0.0023, "step": 26721 }, { "epoch": 2.897007805724198, "grad_norm": 0.3477064073085785, "learning_rate": 3.445524445089221e-06, "loss": 0.0091, "step": 26722 }, { "epoch": 2.8971162185602775, "grad_norm": 0.343923956155777, "learning_rate": 3.441897577252285e-06, "loss": 0.0376, "step": 26723 }, { "epoch": 2.8972246313963574, "grad_norm": 0.07738195359706879, "learning_rate": 3.438270709415349e-06, "loss": 0.0016, "step": 26724 }, { "epoch": 2.897333044232437, "grad_norm": 0.2951239049434662, "learning_rate": 3.434643841578413e-06, "loss": 0.0116, "step": 26725 }, { "epoch": 2.897441457068517, "grad_norm": 0.38046795129776, "learning_rate": 3.4310169737414773e-06, "loss": 0.0269, "step": 26726 }, { "epoch": 2.897549869904597, "grad_norm": 0.40805211663246155, "learning_rate": 3.427390105904541e-06, "loss": 0.0056, "step": 26727 }, { "epoch": 2.8976582827406765, "grad_norm": 0.09054847061634064, "learning_rate": 3.423763238067605e-06, "loss": 0.0052, "step": 26728 }, { "epoch": 2.897766695576756, "grad_norm": 0.10008547455072403, "learning_rate": 3.420136370230669e-06, "loss": 0.0057, "step": 26729 }, { "epoch": 2.897875108412836, "grad_norm": 0.022013595327734947, "learning_rate": 3.4165095023937333e-06, "loss": 0.001, "step": 26730 }, { "epoch": 2.897983521248916, "grad_norm": 0.3068338632583618, "learning_rate": 3.4128826345567965e-06, "loss": 0.0273, "step": 26731 }, { "epoch": 2.8980919340849955, "grad_norm": 0.40351518988609314, "learning_rate": 3.4092557667198607e-06, "loss": 0.0205, "step": 26732 }, { "epoch": 2.8982003469210755, "grad_norm": 0.19794932007789612, "learning_rate": 3.4056288988829248e-06, "loss": 0.0027, "step": 26733 }, { "epoch": 2.898308759757155, "grad_norm": 0.6674168705940247, "learning_rate": 3.402002031045989e-06, "loss": 0.0215, "step": 26734 }, { "epoch": 2.898417172593235, "grad_norm": 0.06701244413852692, "learning_rate": 3.3983751632090525e-06, "loss": 0.0018, "step": 26735 }, { "epoch": 2.898525585429315, "grad_norm": 0.2961958050727844, "learning_rate": 3.3947482953721167e-06, "loss": 0.0068, "step": 26736 }, { "epoch": 2.8986339982653946, "grad_norm": 0.09733636677265167, "learning_rate": 3.3911214275351808e-06, "loss": 0.0072, "step": 26737 }, { "epoch": 2.8987424111014746, "grad_norm": 0.5733073353767395, "learning_rate": 3.387494559698245e-06, "loss": 0.0158, "step": 26738 }, { "epoch": 2.898850823937554, "grad_norm": 0.1997862607240677, "learning_rate": 3.383867691861309e-06, "loss": 0.0106, "step": 26739 }, { "epoch": 2.898959236773634, "grad_norm": 0.42312875390052795, "learning_rate": 3.3802408240243726e-06, "loss": 0.0148, "step": 26740 }, { "epoch": 2.8990676496097136, "grad_norm": 0.23827986419200897, "learning_rate": 3.3766139561874368e-06, "loss": 0.0044, "step": 26741 }, { "epoch": 2.8991760624457936, "grad_norm": 0.6088412404060364, "learning_rate": 3.372987088350501e-06, "loss": 0.0308, "step": 26742 }, { "epoch": 2.899284475281873, "grad_norm": 0.23553583025932312, "learning_rate": 3.369360220513565e-06, "loss": 0.0208, "step": 26743 }, { "epoch": 2.899392888117953, "grad_norm": 0.031042249873280525, "learning_rate": 3.365733352676629e-06, "loss": 0.0021, "step": 26744 }, { "epoch": 2.899501300954033, "grad_norm": 0.218124657869339, "learning_rate": 3.3621064848396923e-06, "loss": 0.0135, "step": 26745 }, { "epoch": 2.8996097137901127, "grad_norm": 0.10942486673593521, "learning_rate": 3.3584796170027564e-06, "loss": 0.0024, "step": 26746 }, { "epoch": 2.8997181266261927, "grad_norm": 0.04881789907813072, "learning_rate": 3.3548527491658205e-06, "loss": 0.003, "step": 26747 }, { "epoch": 2.899826539462272, "grad_norm": 0.06684819608926773, "learning_rate": 3.3512258813288842e-06, "loss": 0.0025, "step": 26748 }, { "epoch": 2.899934952298352, "grad_norm": 0.03828782960772514, "learning_rate": 3.3475990134919483e-06, "loss": 0.002, "step": 26749 }, { "epoch": 2.900043365134432, "grad_norm": 0.09221260994672775, "learning_rate": 3.3439721456550124e-06, "loss": 0.0017, "step": 26750 }, { "epoch": 2.9001517779705117, "grad_norm": 0.35889557003974915, "learning_rate": 3.3403452778180765e-06, "loss": 0.0206, "step": 26751 }, { "epoch": 2.9002601908065913, "grad_norm": 0.29999855160713196, "learning_rate": 3.3367184099811406e-06, "loss": 0.0276, "step": 26752 }, { "epoch": 2.9003686036426712, "grad_norm": 0.15380653738975525, "learning_rate": 3.3330915421442043e-06, "loss": 0.0028, "step": 26753 }, { "epoch": 2.9004770164787512, "grad_norm": 0.05201709643006325, "learning_rate": 3.3294646743072684e-06, "loss": 0.0015, "step": 26754 }, { "epoch": 2.9005854293148308, "grad_norm": 0.09040804207324982, "learning_rate": 3.3258378064703325e-06, "loss": 0.0092, "step": 26755 }, { "epoch": 2.9006938421509108, "grad_norm": 0.038820527493953705, "learning_rate": 3.3222109386333966e-06, "loss": 0.0012, "step": 26756 }, { "epoch": 2.9008022549869903, "grad_norm": 0.5412798523902893, "learning_rate": 3.3185840707964607e-06, "loss": 0.0204, "step": 26757 }, { "epoch": 2.9009106678230703, "grad_norm": 0.500661313533783, "learning_rate": 3.314957202959524e-06, "loss": 0.0544, "step": 26758 }, { "epoch": 2.9010190806591503, "grad_norm": 0.47127318382263184, "learning_rate": 3.311330335122588e-06, "loss": 0.0121, "step": 26759 }, { "epoch": 2.90112749349523, "grad_norm": 1.212078332901001, "learning_rate": 3.307703467285652e-06, "loss": 0.0144, "step": 26760 }, { "epoch": 2.90123590633131, "grad_norm": 0.42180025577545166, "learning_rate": 3.304076599448716e-06, "loss": 0.0282, "step": 26761 }, { "epoch": 2.9013443191673893, "grad_norm": 0.10611245036125183, "learning_rate": 3.30044973161178e-06, "loss": 0.0039, "step": 26762 }, { "epoch": 2.9014527320034693, "grad_norm": 0.14640390872955322, "learning_rate": 3.296822863774844e-06, "loss": 0.0016, "step": 26763 }, { "epoch": 2.901561144839549, "grad_norm": 0.4293256402015686, "learning_rate": 3.293195995937908e-06, "loss": 0.0213, "step": 26764 }, { "epoch": 2.901669557675629, "grad_norm": 0.2075101137161255, "learning_rate": 3.2895691281009723e-06, "loss": 0.0165, "step": 26765 }, { "epoch": 2.9017779705117084, "grad_norm": 0.012405920773744583, "learning_rate": 3.285942260264036e-06, "loss": 0.0002, "step": 26766 }, { "epoch": 2.9018863833477884, "grad_norm": 0.26575300097465515, "learning_rate": 3.2823153924271e-06, "loss": 0.0089, "step": 26767 }, { "epoch": 2.9019947961838684, "grad_norm": 0.02631305530667305, "learning_rate": 3.278688524590164e-06, "loss": 0.0018, "step": 26768 }, { "epoch": 2.902103209019948, "grad_norm": 0.12938013672828674, "learning_rate": 3.2750616567532283e-06, "loss": 0.0063, "step": 26769 }, { "epoch": 2.902211621856028, "grad_norm": 0.20367486774921417, "learning_rate": 3.2714347889162924e-06, "loss": 0.0069, "step": 26770 }, { "epoch": 2.9023200346921074, "grad_norm": 0.01981309987604618, "learning_rate": 3.2678079210793565e-06, "loss": 0.0003, "step": 26771 }, { "epoch": 2.9024284475281874, "grad_norm": 0.23274411261081696, "learning_rate": 3.2641810532424198e-06, "loss": 0.0069, "step": 26772 }, { "epoch": 2.9025368603642674, "grad_norm": 0.08137962222099304, "learning_rate": 3.260554185405484e-06, "loss": 0.0029, "step": 26773 }, { "epoch": 2.902645273200347, "grad_norm": 0.030133254826068878, "learning_rate": 3.2569273175685476e-06, "loss": 0.0012, "step": 26774 }, { "epoch": 2.9027536860364265, "grad_norm": 0.8286929726600647, "learning_rate": 3.2533004497316117e-06, "loss": 0.0158, "step": 26775 }, { "epoch": 2.9028620988725065, "grad_norm": 0.3202131986618042, "learning_rate": 3.2496735818946758e-06, "loss": 0.0272, "step": 26776 }, { "epoch": 2.9029705117085864, "grad_norm": 0.42904219031333923, "learning_rate": 3.24604671405774e-06, "loss": 0.0104, "step": 26777 }, { "epoch": 2.903078924544666, "grad_norm": 0.22722816467285156, "learning_rate": 3.242419846220804e-06, "loss": 0.0066, "step": 26778 }, { "epoch": 2.903187337380746, "grad_norm": 0.10690025985240936, "learning_rate": 3.238792978383868e-06, "loss": 0.0014, "step": 26779 }, { "epoch": 2.9032957502168255, "grad_norm": 0.14433681964874268, "learning_rate": 3.2351661105469318e-06, "loss": 0.0081, "step": 26780 }, { "epoch": 2.9034041630529055, "grad_norm": 1.3138387203216553, "learning_rate": 3.231539242709996e-06, "loss": 0.0054, "step": 26781 }, { "epoch": 2.9035125758889855, "grad_norm": 0.199420765042305, "learning_rate": 3.22791237487306e-06, "loss": 0.0037, "step": 26782 }, { "epoch": 2.903620988725065, "grad_norm": 0.032477784901857376, "learning_rate": 3.224285507036124e-06, "loss": 0.0009, "step": 26783 }, { "epoch": 2.9037294015611446, "grad_norm": 0.5529232025146484, "learning_rate": 3.220658639199188e-06, "loss": 0.0441, "step": 26784 }, { "epoch": 2.9038378143972245, "grad_norm": 0.40673208236694336, "learning_rate": 3.2170317713622514e-06, "loss": 0.0174, "step": 26785 }, { "epoch": 2.9039462272333045, "grad_norm": 0.07029543817043304, "learning_rate": 3.2134049035253156e-06, "loss": 0.0011, "step": 26786 }, { "epoch": 2.904054640069384, "grad_norm": 1.1109106540679932, "learning_rate": 3.2097780356883792e-06, "loss": 0.0268, "step": 26787 }, { "epoch": 2.904163052905464, "grad_norm": 0.037506356835365295, "learning_rate": 3.2061511678514433e-06, "loss": 0.0012, "step": 26788 }, { "epoch": 2.9042714657415436, "grad_norm": 0.1734418272972107, "learning_rate": 3.2025243000145074e-06, "loss": 0.0081, "step": 26789 }, { "epoch": 2.9043798785776236, "grad_norm": 0.3354155719280243, "learning_rate": 3.1988974321775715e-06, "loss": 0.0156, "step": 26790 }, { "epoch": 2.9044882914137036, "grad_norm": 0.43936195969581604, "learning_rate": 3.1952705643406357e-06, "loss": 0.012, "step": 26791 }, { "epoch": 2.904596704249783, "grad_norm": 0.28525859117507935, "learning_rate": 3.1916436965036998e-06, "loss": 0.0178, "step": 26792 }, { "epoch": 2.904705117085863, "grad_norm": 0.11422095447778702, "learning_rate": 3.1880168286667634e-06, "loss": 0.0031, "step": 26793 }, { "epoch": 2.9048135299219426, "grad_norm": 0.3383813202381134, "learning_rate": 3.1843899608298275e-06, "loss": 0.0098, "step": 26794 }, { "epoch": 2.9049219427580226, "grad_norm": 0.3930410146713257, "learning_rate": 3.1807630929928917e-06, "loss": 0.0104, "step": 26795 }, { "epoch": 2.9050303555941026, "grad_norm": 0.7749707102775574, "learning_rate": 3.1771362251559558e-06, "loss": 0.0302, "step": 26796 }, { "epoch": 2.905138768430182, "grad_norm": 0.0929717868566513, "learning_rate": 3.17350935731902e-06, "loss": 0.0021, "step": 26797 }, { "epoch": 2.9052471812662617, "grad_norm": 0.3377801477909088, "learning_rate": 3.1698824894820835e-06, "loss": 0.0144, "step": 26798 }, { "epoch": 2.9053555941023417, "grad_norm": 0.09546060115098953, "learning_rate": 3.1662556216451472e-06, "loss": 0.0041, "step": 26799 }, { "epoch": 2.9054640069384217, "grad_norm": 0.012622918002307415, "learning_rate": 3.162628753808211e-06, "loss": 0.0004, "step": 26800 }, { "epoch": 2.905572419774501, "grad_norm": 0.2640180289745331, "learning_rate": 3.159001885971275e-06, "loss": 0.0148, "step": 26801 }, { "epoch": 2.905680832610581, "grad_norm": 0.21851500868797302, "learning_rate": 3.155375018134339e-06, "loss": 0.0051, "step": 26802 }, { "epoch": 2.9057892454466607, "grad_norm": 0.23935644328594208, "learning_rate": 3.1517481502974032e-06, "loss": 0.0094, "step": 26803 }, { "epoch": 2.9058976582827407, "grad_norm": 0.11602140218019485, "learning_rate": 3.1481212824604673e-06, "loss": 0.0047, "step": 26804 }, { "epoch": 2.9060060711188207, "grad_norm": 0.07892443984746933, "learning_rate": 3.1444944146235314e-06, "loss": 0.0008, "step": 26805 }, { "epoch": 2.9061144839549002, "grad_norm": 0.526361346244812, "learning_rate": 3.140867546786595e-06, "loss": 0.0098, "step": 26806 }, { "epoch": 2.90622289679098, "grad_norm": 0.18840782344341278, "learning_rate": 3.1372406789496592e-06, "loss": 0.0019, "step": 26807 }, { "epoch": 2.9063313096270598, "grad_norm": 0.1445152461528778, "learning_rate": 3.1336138111127233e-06, "loss": 0.0058, "step": 26808 }, { "epoch": 2.9064397224631398, "grad_norm": 0.009581135585904121, "learning_rate": 3.1299869432757874e-06, "loss": 0.0002, "step": 26809 }, { "epoch": 2.9065481352992193, "grad_norm": 0.07237162441015244, "learning_rate": 3.1263600754388515e-06, "loss": 0.0019, "step": 26810 }, { "epoch": 2.9066565481352993, "grad_norm": 0.26967620849609375, "learning_rate": 3.1227332076019152e-06, "loss": 0.0166, "step": 26811 }, { "epoch": 2.906764960971379, "grad_norm": 0.4160161018371582, "learning_rate": 3.1191063397649793e-06, "loss": 0.0124, "step": 26812 }, { "epoch": 2.906873373807459, "grad_norm": 0.4108083248138428, "learning_rate": 3.115479471928043e-06, "loss": 0.0263, "step": 26813 }, { "epoch": 2.906981786643539, "grad_norm": 0.39092281460762024, "learning_rate": 3.111852604091107e-06, "loss": 0.0066, "step": 26814 }, { "epoch": 2.9070901994796183, "grad_norm": 0.22130313515663147, "learning_rate": 3.1082257362541712e-06, "loss": 0.0047, "step": 26815 }, { "epoch": 2.9071986123156983, "grad_norm": 0.05192117393016815, "learning_rate": 3.104598868417235e-06, "loss": 0.0031, "step": 26816 }, { "epoch": 2.907307025151778, "grad_norm": 0.35049840807914734, "learning_rate": 3.100972000580299e-06, "loss": 0.0495, "step": 26817 }, { "epoch": 2.907415437987858, "grad_norm": 0.011506636627018452, "learning_rate": 3.097345132743363e-06, "loss": 0.0006, "step": 26818 }, { "epoch": 2.907523850823938, "grad_norm": 0.2470596730709076, "learning_rate": 3.0937182649064268e-06, "loss": 0.0191, "step": 26819 }, { "epoch": 2.9076322636600174, "grad_norm": 0.08912234008312225, "learning_rate": 3.090091397069491e-06, "loss": 0.0032, "step": 26820 }, { "epoch": 2.907740676496097, "grad_norm": 0.29406046867370605, "learning_rate": 3.086464529232555e-06, "loss": 0.0071, "step": 26821 }, { "epoch": 2.907849089332177, "grad_norm": 0.3206673562526703, "learning_rate": 3.082837661395619e-06, "loss": 0.0119, "step": 26822 }, { "epoch": 2.907957502168257, "grad_norm": 0.450431764125824, "learning_rate": 3.0792107935586828e-06, "loss": 0.022, "step": 26823 }, { "epoch": 2.9080659150043364, "grad_norm": 0.050666067749261856, "learning_rate": 3.075583925721747e-06, "loss": 0.0022, "step": 26824 }, { "epoch": 2.9081743278404164, "grad_norm": 0.23392841219902039, "learning_rate": 3.071957057884811e-06, "loss": 0.0131, "step": 26825 }, { "epoch": 2.908282740676496, "grad_norm": 0.2189619541168213, "learning_rate": 3.0683301900478747e-06, "loss": 0.0081, "step": 26826 }, { "epoch": 2.908391153512576, "grad_norm": 0.38797828555107117, "learning_rate": 3.0647033222109388e-06, "loss": 0.0306, "step": 26827 }, { "epoch": 2.908499566348656, "grad_norm": 0.17316794395446777, "learning_rate": 3.061076454374003e-06, "loss": 0.0089, "step": 26828 }, { "epoch": 2.9086079791847355, "grad_norm": 0.5344799160957336, "learning_rate": 3.0574495865370666e-06, "loss": 0.0155, "step": 26829 }, { "epoch": 2.908716392020815, "grad_norm": 0.11093878746032715, "learning_rate": 3.0538227187001307e-06, "loss": 0.0029, "step": 26830 }, { "epoch": 2.908824804856895, "grad_norm": 0.27902960777282715, "learning_rate": 3.0501958508631948e-06, "loss": 0.0076, "step": 26831 }, { "epoch": 2.908933217692975, "grad_norm": 0.08160023391246796, "learning_rate": 3.0465689830262585e-06, "loss": 0.0034, "step": 26832 }, { "epoch": 2.9090416305290545, "grad_norm": 0.23993928730487823, "learning_rate": 3.0429421151893226e-06, "loss": 0.0095, "step": 26833 }, { "epoch": 2.9091500433651345, "grad_norm": 0.0799642875790596, "learning_rate": 3.0393152473523867e-06, "loss": 0.0033, "step": 26834 }, { "epoch": 2.909258456201214, "grad_norm": 0.641628623008728, "learning_rate": 3.0356883795154508e-06, "loss": 0.0051, "step": 26835 }, { "epoch": 2.909366869037294, "grad_norm": 0.166531503200531, "learning_rate": 3.0320615116785145e-06, "loss": 0.0061, "step": 26836 }, { "epoch": 2.909475281873374, "grad_norm": 0.005329940468072891, "learning_rate": 3.0284346438415786e-06, "loss": 0.0001, "step": 26837 }, { "epoch": 2.9095836947094535, "grad_norm": 0.02393016777932644, "learning_rate": 3.0248077760046427e-06, "loss": 0.0008, "step": 26838 }, { "epoch": 2.9096921075455335, "grad_norm": 0.03438663110136986, "learning_rate": 3.0211809081677063e-06, "loss": 0.0008, "step": 26839 }, { "epoch": 2.909800520381613, "grad_norm": 0.3028174638748169, "learning_rate": 3.0175540403307705e-06, "loss": 0.0129, "step": 26840 }, { "epoch": 2.909908933217693, "grad_norm": 0.37351009249687195, "learning_rate": 3.0139271724938346e-06, "loss": 0.0123, "step": 26841 }, { "epoch": 2.9100173460537726, "grad_norm": 0.20296820998191833, "learning_rate": 3.0103003046568987e-06, "loss": 0.0083, "step": 26842 }, { "epoch": 2.9101257588898526, "grad_norm": 0.9046611189842224, "learning_rate": 3.0066734368199623e-06, "loss": 0.0814, "step": 26843 }, { "epoch": 2.910234171725932, "grad_norm": 0.5967562794685364, "learning_rate": 3.0030465689830264e-06, "loss": 0.0193, "step": 26844 }, { "epoch": 2.910342584562012, "grad_norm": 0.7880851030349731, "learning_rate": 2.99941970114609e-06, "loss": 0.0177, "step": 26845 }, { "epoch": 2.910450997398092, "grad_norm": 0.1559860110282898, "learning_rate": 2.9957928333091542e-06, "loss": 0.0034, "step": 26846 }, { "epoch": 2.9105594102341716, "grad_norm": 0.38359197974205017, "learning_rate": 2.9921659654722183e-06, "loss": 0.0395, "step": 26847 }, { "epoch": 2.9106678230702516, "grad_norm": 0.3021227717399597, "learning_rate": 2.9885390976352824e-06, "loss": 0.0305, "step": 26848 }, { "epoch": 2.910776235906331, "grad_norm": 0.3455236554145813, "learning_rate": 2.9849122297983465e-06, "loss": 0.023, "step": 26849 }, { "epoch": 2.910884648742411, "grad_norm": 0.30977341532707214, "learning_rate": 2.9812853619614102e-06, "loss": 0.0045, "step": 26850 }, { "epoch": 2.910993061578491, "grad_norm": 0.23466132581233978, "learning_rate": 2.9776584941244743e-06, "loss": 0.0092, "step": 26851 }, { "epoch": 2.9111014744145707, "grad_norm": 0.0075409566052258015, "learning_rate": 2.974031626287538e-06, "loss": 0.0002, "step": 26852 }, { "epoch": 2.91120988725065, "grad_norm": 0.04342003911733627, "learning_rate": 2.970404758450602e-06, "loss": 0.0013, "step": 26853 }, { "epoch": 2.91131830008673, "grad_norm": 1.091870903968811, "learning_rate": 2.9667778906136662e-06, "loss": 0.044, "step": 26854 }, { "epoch": 2.91142671292281, "grad_norm": 0.5582923293113708, "learning_rate": 2.9631510227767303e-06, "loss": 0.0156, "step": 26855 }, { "epoch": 2.9115351257588897, "grad_norm": 0.4882892966270447, "learning_rate": 2.9595241549397944e-06, "loss": 0.009, "step": 26856 }, { "epoch": 2.9116435385949697, "grad_norm": 0.020624877884984016, "learning_rate": 2.955897287102858e-06, "loss": 0.0004, "step": 26857 }, { "epoch": 2.9117519514310493, "grad_norm": 0.19630281627178192, "learning_rate": 2.9522704192659222e-06, "loss": 0.0218, "step": 26858 }, { "epoch": 2.9118603642671292, "grad_norm": 0.4232795238494873, "learning_rate": 2.948643551428986e-06, "loss": 0.0337, "step": 26859 }, { "epoch": 2.9119687771032092, "grad_norm": 0.0782526433467865, "learning_rate": 2.94501668359205e-06, "loss": 0.0032, "step": 26860 }, { "epoch": 2.9120771899392888, "grad_norm": 0.12255788594484329, "learning_rate": 2.941389815755114e-06, "loss": 0.0048, "step": 26861 }, { "epoch": 2.9121856027753688, "grad_norm": 0.10860855132341385, "learning_rate": 2.9377629479181782e-06, "loss": 0.0097, "step": 26862 }, { "epoch": 2.9122940156114483, "grad_norm": 0.44313380122184753, "learning_rate": 2.934136080081242e-06, "loss": 0.0169, "step": 26863 }, { "epoch": 2.9124024284475283, "grad_norm": 0.30589547753334045, "learning_rate": 2.930509212244306e-06, "loss": 0.0051, "step": 26864 }, { "epoch": 2.912510841283608, "grad_norm": 0.15247493982315063, "learning_rate": 2.9268823444073697e-06, "loss": 0.0166, "step": 26865 }, { "epoch": 2.912619254119688, "grad_norm": 0.5060873031616211, "learning_rate": 2.923255476570434e-06, "loss": 0.0173, "step": 26866 }, { "epoch": 2.9127276669557673, "grad_norm": 0.19043287634849548, "learning_rate": 2.919628608733498e-06, "loss": 0.0072, "step": 26867 }, { "epoch": 2.9128360797918473, "grad_norm": 0.30467644333839417, "learning_rate": 2.916001740896562e-06, "loss": 0.0172, "step": 26868 }, { "epoch": 2.9129444926279273, "grad_norm": 0.09231214970350266, "learning_rate": 2.912374873059626e-06, "loss": 0.0019, "step": 26869 }, { "epoch": 2.913052905464007, "grad_norm": 0.11526475846767426, "learning_rate": 2.90874800522269e-06, "loss": 0.0044, "step": 26870 }, { "epoch": 2.913161318300087, "grad_norm": 0.4683895409107208, "learning_rate": 2.905121137385754e-06, "loss": 0.0094, "step": 26871 }, { "epoch": 2.9132697311361664, "grad_norm": 0.1024043932557106, "learning_rate": 2.9014942695488176e-06, "loss": 0.0042, "step": 26872 }, { "epoch": 2.9133781439722464, "grad_norm": 0.26426926255226135, "learning_rate": 2.8978674017118817e-06, "loss": 0.0216, "step": 26873 }, { "epoch": 2.9134865568083264, "grad_norm": 0.44611048698425293, "learning_rate": 2.894240533874946e-06, "loss": 0.0475, "step": 26874 }, { "epoch": 2.913594969644406, "grad_norm": 0.8203545808792114, "learning_rate": 2.89061366603801e-06, "loss": 0.0249, "step": 26875 }, { "epoch": 2.9137033824804854, "grad_norm": 0.06843376904726028, "learning_rate": 2.886986798201074e-06, "loss": 0.0022, "step": 26876 }, { "epoch": 2.9138117953165654, "grad_norm": 1.3892967700958252, "learning_rate": 2.8833599303641377e-06, "loss": 0.0888, "step": 26877 }, { "epoch": 2.9139202081526454, "grad_norm": 0.15370233356952667, "learning_rate": 2.8797330625272014e-06, "loss": 0.0103, "step": 26878 }, { "epoch": 2.914028620988725, "grad_norm": 0.6180523633956909, "learning_rate": 2.8761061946902655e-06, "loss": 0.0188, "step": 26879 }, { "epoch": 2.914137033824805, "grad_norm": 0.4413053095340729, "learning_rate": 2.8724793268533296e-06, "loss": 0.0229, "step": 26880 }, { "epoch": 2.9142454466608845, "grad_norm": 0.021617084741592407, "learning_rate": 2.8688524590163937e-06, "loss": 0.0008, "step": 26881 }, { "epoch": 2.9143538594969645, "grad_norm": 0.16451385617256165, "learning_rate": 2.8652255911794578e-06, "loss": 0.0061, "step": 26882 }, { "epoch": 2.9144622723330444, "grad_norm": 0.3162957429885864, "learning_rate": 2.861598723342522e-06, "loss": 0.012, "step": 26883 }, { "epoch": 2.914570685169124, "grad_norm": 0.15416985750198364, "learning_rate": 2.8579718555055856e-06, "loss": 0.0047, "step": 26884 }, { "epoch": 2.914679098005204, "grad_norm": 0.08938705176115036, "learning_rate": 2.8543449876686493e-06, "loss": 0.0034, "step": 26885 }, { "epoch": 2.9147875108412835, "grad_norm": 0.3627640902996063, "learning_rate": 2.8507181198317134e-06, "loss": 0.011, "step": 26886 }, { "epoch": 2.9148959236773635, "grad_norm": 0.018364127725362778, "learning_rate": 2.8470912519947775e-06, "loss": 0.0005, "step": 26887 }, { "epoch": 2.915004336513443, "grad_norm": 0.0849330946803093, "learning_rate": 2.8434643841578416e-06, "loss": 0.0055, "step": 26888 }, { "epoch": 2.915112749349523, "grad_norm": 0.5162132978439331, "learning_rate": 2.8398375163209057e-06, "loss": 0.0184, "step": 26889 }, { "epoch": 2.9152211621856026, "grad_norm": 0.5503566861152649, "learning_rate": 2.8362106484839694e-06, "loss": 0.0244, "step": 26890 }, { "epoch": 2.9153295750216826, "grad_norm": 0.3710215985774994, "learning_rate": 2.832583780647033e-06, "loss": 0.0093, "step": 26891 }, { "epoch": 2.9154379878577625, "grad_norm": 0.15236276388168335, "learning_rate": 2.828956912810097e-06, "loss": 0.0072, "step": 26892 }, { "epoch": 2.915546400693842, "grad_norm": 0.36297500133514404, "learning_rate": 2.8253300449731612e-06, "loss": 0.01, "step": 26893 }, { "epoch": 2.915654813529922, "grad_norm": 0.14675365388393402, "learning_rate": 2.8217031771362253e-06, "loss": 0.0037, "step": 26894 }, { "epoch": 2.9157632263660016, "grad_norm": 0.18795859813690186, "learning_rate": 2.8180763092992895e-06, "loss": 0.0049, "step": 26895 }, { "epoch": 2.9158716392020816, "grad_norm": 0.0006078079459257424, "learning_rate": 2.8144494414623536e-06, "loss": 0.0, "step": 26896 }, { "epoch": 2.9159800520381616, "grad_norm": 0.023870345205068588, "learning_rate": 2.8108225736254172e-06, "loss": 0.0007, "step": 26897 }, { "epoch": 2.916088464874241, "grad_norm": 0.5072842240333557, "learning_rate": 2.807195705788481e-06, "loss": 0.0094, "step": 26898 }, { "epoch": 2.9161968777103207, "grad_norm": 0.19260650873184204, "learning_rate": 2.803568837951545e-06, "loss": 0.004, "step": 26899 }, { "epoch": 2.9163052905464006, "grad_norm": 0.059281472116708755, "learning_rate": 2.799941970114609e-06, "loss": 0.0034, "step": 26900 }, { "epoch": 2.9164137033824806, "grad_norm": 0.27934640645980835, "learning_rate": 2.7963151022776732e-06, "loss": 0.0209, "step": 26901 }, { "epoch": 2.91652211621856, "grad_norm": 0.8463630676269531, "learning_rate": 2.7926882344407373e-06, "loss": 0.0363, "step": 26902 }, { "epoch": 2.91663052905464, "grad_norm": 0.4324823319911957, "learning_rate": 2.7890613666038014e-06, "loss": 0.0232, "step": 26903 }, { "epoch": 2.9167389418907197, "grad_norm": 0.03930259123444557, "learning_rate": 2.785434498766865e-06, "loss": 0.002, "step": 26904 }, { "epoch": 2.9168473547267997, "grad_norm": 0.17127446830272675, "learning_rate": 2.781807630929929e-06, "loss": 0.0061, "step": 26905 }, { "epoch": 2.9169557675628797, "grad_norm": 0.017995288595557213, "learning_rate": 2.778180763092993e-06, "loss": 0.0009, "step": 26906 }, { "epoch": 2.917064180398959, "grad_norm": 0.4170098304748535, "learning_rate": 2.774553895256057e-06, "loss": 0.0177, "step": 26907 }, { "epoch": 2.917172593235039, "grad_norm": 0.22159965336322784, "learning_rate": 2.770927027419121e-06, "loss": 0.0144, "step": 26908 }, { "epoch": 2.9172810060711187, "grad_norm": 0.24438779056072235, "learning_rate": 2.7673001595821852e-06, "loss": 0.0061, "step": 26909 }, { "epoch": 2.9173894189071987, "grad_norm": 0.23286698758602142, "learning_rate": 2.763673291745249e-06, "loss": 0.004, "step": 26910 }, { "epoch": 2.9174978317432783, "grad_norm": 0.3862392008304596, "learning_rate": 2.7600464239083126e-06, "loss": 0.0077, "step": 26911 }, { "epoch": 2.9176062445793582, "grad_norm": 0.22984451055526733, "learning_rate": 2.7564195560713767e-06, "loss": 0.0074, "step": 26912 }, { "epoch": 2.917714657415438, "grad_norm": 0.16523835062980652, "learning_rate": 2.752792688234441e-06, "loss": 0.0043, "step": 26913 }, { "epoch": 2.9178230702515178, "grad_norm": 0.011675462126731873, "learning_rate": 2.749165820397505e-06, "loss": 0.0005, "step": 26914 }, { "epoch": 2.9179314830875978, "grad_norm": 0.17331278324127197, "learning_rate": 2.745538952560569e-06, "loss": 0.0012, "step": 26915 }, { "epoch": 2.9180398959236773, "grad_norm": 0.13268502056598663, "learning_rate": 2.741912084723633e-06, "loss": 0.0051, "step": 26916 }, { "epoch": 2.9181483087597573, "grad_norm": 0.08435418456792831, "learning_rate": 2.738285216886697e-06, "loss": 0.0011, "step": 26917 }, { "epoch": 2.918256721595837, "grad_norm": 0.04846735671162605, "learning_rate": 2.7346583490497605e-06, "loss": 0.0015, "step": 26918 }, { "epoch": 2.918365134431917, "grad_norm": 0.07489223033189774, "learning_rate": 2.7310314812128246e-06, "loss": 0.0024, "step": 26919 }, { "epoch": 2.918473547267997, "grad_norm": 0.2688296139240265, "learning_rate": 2.7274046133758887e-06, "loss": 0.014, "step": 26920 }, { "epoch": 2.9185819601040763, "grad_norm": 0.143405482172966, "learning_rate": 2.723777745538953e-06, "loss": 0.0021, "step": 26921 }, { "epoch": 2.918690372940156, "grad_norm": 0.028144633397459984, "learning_rate": 2.720150877702017e-06, "loss": 0.0008, "step": 26922 }, { "epoch": 2.918798785776236, "grad_norm": 0.1460426151752472, "learning_rate": 2.7165240098650806e-06, "loss": 0.0028, "step": 26923 }, { "epoch": 2.918907198612316, "grad_norm": 0.06496277451515198, "learning_rate": 2.7128971420281443e-06, "loss": 0.0013, "step": 26924 }, { "epoch": 2.9190156114483954, "grad_norm": 0.27698707580566406, "learning_rate": 2.7092702741912084e-06, "loss": 0.0051, "step": 26925 }, { "epoch": 2.9191240242844754, "grad_norm": 0.20809224247932434, "learning_rate": 2.7056434063542725e-06, "loss": 0.0073, "step": 26926 }, { "epoch": 2.919232437120555, "grad_norm": 0.10365085303783417, "learning_rate": 2.7020165385173366e-06, "loss": 0.0013, "step": 26927 }, { "epoch": 2.919340849956635, "grad_norm": 0.0787157490849495, "learning_rate": 2.6983896706804007e-06, "loss": 0.0031, "step": 26928 }, { "epoch": 2.919449262792715, "grad_norm": 0.1031641811132431, "learning_rate": 2.694762802843465e-06, "loss": 0.0025, "step": 26929 }, { "epoch": 2.9195576756287944, "grad_norm": 0.2457694113254547, "learning_rate": 2.6911359350065285e-06, "loss": 0.0151, "step": 26930 }, { "epoch": 2.919666088464874, "grad_norm": 0.12131080031394958, "learning_rate": 2.687509067169592e-06, "loss": 0.002, "step": 26931 }, { "epoch": 2.919774501300954, "grad_norm": 0.0593213215470314, "learning_rate": 2.6838821993326563e-06, "loss": 0.0017, "step": 26932 }, { "epoch": 2.919882914137034, "grad_norm": 0.218566432595253, "learning_rate": 2.6802553314957204e-06, "loss": 0.0081, "step": 26933 }, { "epoch": 2.9199913269731135, "grad_norm": 0.030308129265904427, "learning_rate": 2.6766284636587845e-06, "loss": 0.0012, "step": 26934 }, { "epoch": 2.9200997398091935, "grad_norm": 0.0769207701086998, "learning_rate": 2.6730015958218486e-06, "loss": 0.0016, "step": 26935 }, { "epoch": 2.920208152645273, "grad_norm": 0.2869926393032074, "learning_rate": 2.6693747279849123e-06, "loss": 0.018, "step": 26936 }, { "epoch": 2.920316565481353, "grad_norm": 0.26111721992492676, "learning_rate": 2.6657478601479764e-06, "loss": 0.0103, "step": 26937 }, { "epoch": 2.920424978317433, "grad_norm": 0.06270158290863037, "learning_rate": 2.66212099231104e-06, "loss": 0.002, "step": 26938 }, { "epoch": 2.9205333911535125, "grad_norm": 0.30999231338500977, "learning_rate": 2.658494124474104e-06, "loss": 0.0066, "step": 26939 }, { "epoch": 2.9206418039895925, "grad_norm": 0.47414636611938477, "learning_rate": 2.6548672566371683e-06, "loss": 0.0349, "step": 26940 }, { "epoch": 2.920750216825672, "grad_norm": 0.3837592303752899, "learning_rate": 2.6512403888002324e-06, "loss": 0.0266, "step": 26941 }, { "epoch": 2.920858629661752, "grad_norm": 0.3391658365726471, "learning_rate": 2.6476135209632965e-06, "loss": 0.0135, "step": 26942 }, { "epoch": 2.920967042497832, "grad_norm": 0.05991638824343681, "learning_rate": 2.64398665312636e-06, "loss": 0.0023, "step": 26943 }, { "epoch": 2.9210754553339116, "grad_norm": 0.09809510409832001, "learning_rate": 2.6403597852894243e-06, "loss": 0.0029, "step": 26944 }, { "epoch": 2.921183868169991, "grad_norm": 0.2675640285015106, "learning_rate": 2.636732917452488e-06, "loss": 0.004, "step": 26945 }, { "epoch": 2.921292281006071, "grad_norm": 0.008110033348202705, "learning_rate": 2.633106049615552e-06, "loss": 0.0003, "step": 26946 }, { "epoch": 2.921400693842151, "grad_norm": 0.03578396514058113, "learning_rate": 2.629479181778616e-06, "loss": 0.0009, "step": 26947 }, { "epoch": 2.9215091066782306, "grad_norm": 1.6490191221237183, "learning_rate": 2.6258523139416802e-06, "loss": 0.0398, "step": 26948 }, { "epoch": 2.9216175195143106, "grad_norm": 0.46972233057022095, "learning_rate": 2.6222254461047444e-06, "loss": 0.0181, "step": 26949 }, { "epoch": 2.92172593235039, "grad_norm": 0.27591100335121155, "learning_rate": 2.618598578267808e-06, "loss": 0.0223, "step": 26950 }, { "epoch": 2.92183434518647, "grad_norm": 0.4640396535396576, "learning_rate": 2.6149717104308717e-06, "loss": 0.011, "step": 26951 }, { "epoch": 2.92194275802255, "grad_norm": 0.011744949035346508, "learning_rate": 2.611344842593936e-06, "loss": 0.0004, "step": 26952 }, { "epoch": 2.9220511708586296, "grad_norm": 0.17483076453208923, "learning_rate": 2.607717974757e-06, "loss": 0.0039, "step": 26953 }, { "epoch": 2.922159583694709, "grad_norm": 0.18401743471622467, "learning_rate": 2.604091106920064e-06, "loss": 0.0246, "step": 26954 }, { "epoch": 2.922267996530789, "grad_norm": 0.06666938960552216, "learning_rate": 2.600464239083128e-06, "loss": 0.0047, "step": 26955 }, { "epoch": 2.922376409366869, "grad_norm": 0.5878305435180664, "learning_rate": 2.596837371246192e-06, "loss": 0.0309, "step": 26956 }, { "epoch": 2.9224848222029487, "grad_norm": 0.31539398431777954, "learning_rate": 2.593210503409256e-06, "loss": 0.0071, "step": 26957 }, { "epoch": 2.9225932350390287, "grad_norm": 0.17221374809741974, "learning_rate": 2.5895836355723196e-06, "loss": 0.0055, "step": 26958 }, { "epoch": 2.922701647875108, "grad_norm": 0.17941692471504211, "learning_rate": 2.5859567677353837e-06, "loss": 0.0038, "step": 26959 }, { "epoch": 2.922810060711188, "grad_norm": 0.18540410697460175, "learning_rate": 2.582329899898448e-06, "loss": 0.0096, "step": 26960 }, { "epoch": 2.922918473547268, "grad_norm": 0.23055711388587952, "learning_rate": 2.578703032061512e-06, "loss": 0.0054, "step": 26961 }, { "epoch": 2.9230268863833477, "grad_norm": 0.12045839428901672, "learning_rate": 2.575076164224576e-06, "loss": 0.0035, "step": 26962 }, { "epoch": 2.9231352992194277, "grad_norm": 0.06035109981894493, "learning_rate": 2.5714492963876397e-06, "loss": 0.0019, "step": 26963 }, { "epoch": 2.9232437120555073, "grad_norm": 0.0063812993466854095, "learning_rate": 2.567822428550704e-06, "loss": 0.0003, "step": 26964 }, { "epoch": 2.9233521248915872, "grad_norm": 0.1305014044046402, "learning_rate": 2.5641955607137675e-06, "loss": 0.0041, "step": 26965 }, { "epoch": 2.9234605377276672, "grad_norm": 0.5605446100234985, "learning_rate": 2.5605686928768316e-06, "loss": 0.0047, "step": 26966 }, { "epoch": 2.9235689505637468, "grad_norm": 0.056955527514219284, "learning_rate": 2.5569418250398957e-06, "loss": 0.0019, "step": 26967 }, { "epoch": 2.9236773633998263, "grad_norm": 0.13335856795310974, "learning_rate": 2.55331495720296e-06, "loss": 0.0025, "step": 26968 }, { "epoch": 2.9237857762359063, "grad_norm": 0.7048362493515015, "learning_rate": 2.5496880893660235e-06, "loss": 0.0362, "step": 26969 }, { "epoch": 2.9238941890719863, "grad_norm": 0.28902193903923035, "learning_rate": 2.5460612215290876e-06, "loss": 0.0269, "step": 26970 }, { "epoch": 2.924002601908066, "grad_norm": 0.34422725439071655, "learning_rate": 2.5424343536921517e-06, "loss": 0.0153, "step": 26971 }, { "epoch": 2.924111014744146, "grad_norm": 0.22786638140678406, "learning_rate": 2.5388074858552154e-06, "loss": 0.0115, "step": 26972 }, { "epoch": 2.9242194275802254, "grad_norm": 0.2523210942745209, "learning_rate": 2.5351806180182795e-06, "loss": 0.0098, "step": 26973 }, { "epoch": 2.9243278404163053, "grad_norm": 0.04707362875342369, "learning_rate": 2.5315537501813436e-06, "loss": 0.0029, "step": 26974 }, { "epoch": 2.9244362532523853, "grad_norm": 0.07268938422203064, "learning_rate": 2.5279268823444077e-06, "loss": 0.0015, "step": 26975 }, { "epoch": 2.924544666088465, "grad_norm": 0.2653670310974121, "learning_rate": 2.5243000145074714e-06, "loss": 0.0122, "step": 26976 }, { "epoch": 2.9246530789245444, "grad_norm": 0.05809495598077774, "learning_rate": 2.5206731466705355e-06, "loss": 0.0025, "step": 26977 }, { "epoch": 2.9247614917606244, "grad_norm": 0.0896076038479805, "learning_rate": 2.517046278833599e-06, "loss": 0.0014, "step": 26978 }, { "epoch": 2.9248699045967044, "grad_norm": 0.3007358908653259, "learning_rate": 2.5134194109966633e-06, "loss": 0.0249, "step": 26979 }, { "epoch": 2.924978317432784, "grad_norm": 0.24620479345321655, "learning_rate": 2.5097925431597274e-06, "loss": 0.009, "step": 26980 }, { "epoch": 2.925086730268864, "grad_norm": 0.3232211470603943, "learning_rate": 2.5061656753227915e-06, "loss": 0.0183, "step": 26981 }, { "epoch": 2.9251951431049434, "grad_norm": 0.10523737967014313, "learning_rate": 2.502538807485855e-06, "loss": 0.0052, "step": 26982 }, { "epoch": 2.9253035559410234, "grad_norm": 0.5286332964897156, "learning_rate": 2.4989119396489193e-06, "loss": 0.0268, "step": 26983 }, { "epoch": 2.9254119687771034, "grad_norm": 2.0776445865631104, "learning_rate": 2.4952850718119834e-06, "loss": 0.0486, "step": 26984 }, { "epoch": 2.925520381613183, "grad_norm": 0.0539921410381794, "learning_rate": 2.491658203975047e-06, "loss": 0.0016, "step": 26985 }, { "epoch": 2.925628794449263, "grad_norm": 0.0018896377878263593, "learning_rate": 2.488031336138111e-06, "loss": 0.0001, "step": 26986 }, { "epoch": 2.9257372072853425, "grad_norm": 0.15019185841083527, "learning_rate": 2.4844044683011753e-06, "loss": 0.0038, "step": 26987 }, { "epoch": 2.9258456201214225, "grad_norm": 0.4791730046272278, "learning_rate": 2.4807776004642394e-06, "loss": 0.0346, "step": 26988 }, { "epoch": 2.925954032957502, "grad_norm": 0.05006438493728638, "learning_rate": 2.477150732627303e-06, "loss": 0.0018, "step": 26989 }, { "epoch": 2.926062445793582, "grad_norm": 0.05218677595257759, "learning_rate": 2.473523864790367e-06, "loss": 0.0008, "step": 26990 }, { "epoch": 2.9261708586296615, "grad_norm": 0.04889858886599541, "learning_rate": 2.4698969969534313e-06, "loss": 0.0017, "step": 26991 }, { "epoch": 2.9262792714657415, "grad_norm": 0.04292302951216698, "learning_rate": 2.466270129116495e-06, "loss": 0.0018, "step": 26992 }, { "epoch": 2.9263876843018215, "grad_norm": 0.26412761211395264, "learning_rate": 2.462643261279559e-06, "loss": 0.0127, "step": 26993 }, { "epoch": 2.926496097137901, "grad_norm": 0.055532343685626984, "learning_rate": 2.459016393442623e-06, "loss": 0.0023, "step": 26994 }, { "epoch": 2.926604509973981, "grad_norm": 0.6008001565933228, "learning_rate": 2.4553895256056873e-06, "loss": 0.0242, "step": 26995 }, { "epoch": 2.9267129228100606, "grad_norm": 0.08564899116754532, "learning_rate": 2.451762657768751e-06, "loss": 0.0012, "step": 26996 }, { "epoch": 2.9268213356461406, "grad_norm": 0.043165434151887894, "learning_rate": 2.448135789931815e-06, "loss": 0.0015, "step": 26997 }, { "epoch": 2.9269297484822205, "grad_norm": 0.46524280309677124, "learning_rate": 2.444508922094879e-06, "loss": 0.0158, "step": 26998 }, { "epoch": 2.9270381613183, "grad_norm": 0.44278401136398315, "learning_rate": 2.440882054257943e-06, "loss": 0.0154, "step": 26999 }, { "epoch": 2.9271465741543796, "grad_norm": 0.18715617060661316, "learning_rate": 2.437255186421007e-06, "loss": 0.0029, "step": 27000 }, { "epoch": 2.9272549869904596, "grad_norm": 0.5333412885665894, "learning_rate": 2.433628318584071e-06, "loss": 0.0145, "step": 27001 }, { "epoch": 2.9273633998265396, "grad_norm": 0.17949597537517548, "learning_rate": 2.4300014507471347e-06, "loss": 0.0042, "step": 27002 }, { "epoch": 2.927471812662619, "grad_norm": 0.2624751627445221, "learning_rate": 2.426374582910199e-06, "loss": 0.0086, "step": 27003 }, { "epoch": 2.927580225498699, "grad_norm": 0.5019612908363342, "learning_rate": 2.422747715073263e-06, "loss": 0.0272, "step": 27004 }, { "epoch": 2.9276886383347787, "grad_norm": 0.6425331234931946, "learning_rate": 2.419120847236327e-06, "loss": 0.0074, "step": 27005 }, { "epoch": 2.9277970511708586, "grad_norm": 0.394558846950531, "learning_rate": 2.4154939793993907e-06, "loss": 0.0126, "step": 27006 }, { "epoch": 2.9279054640069386, "grad_norm": 0.2413591742515564, "learning_rate": 2.411867111562455e-06, "loss": 0.0068, "step": 27007 }, { "epoch": 2.928013876843018, "grad_norm": 0.13025793433189392, "learning_rate": 2.408240243725519e-06, "loss": 0.0055, "step": 27008 }, { "epoch": 2.928122289679098, "grad_norm": 0.5232930779457092, "learning_rate": 2.4046133758885826e-06, "loss": 0.015, "step": 27009 }, { "epoch": 2.9282307025151777, "grad_norm": 0.043606892228126526, "learning_rate": 2.4009865080516467e-06, "loss": 0.0013, "step": 27010 }, { "epoch": 2.9283391153512577, "grad_norm": 0.016179058700799942, "learning_rate": 2.397359640214711e-06, "loss": 0.0006, "step": 27011 }, { "epoch": 2.9284475281873372, "grad_norm": 0.27193108201026917, "learning_rate": 2.3937327723777745e-06, "loss": 0.0055, "step": 27012 }, { "epoch": 2.928555941023417, "grad_norm": 0.22720858454704285, "learning_rate": 2.3901059045408386e-06, "loss": 0.0091, "step": 27013 }, { "epoch": 2.9286643538594967, "grad_norm": 0.4084378480911255, "learning_rate": 2.3864790367039027e-06, "loss": 0.0119, "step": 27014 }, { "epoch": 2.9287727666955767, "grad_norm": 1.4904615879058838, "learning_rate": 2.3828521688669664e-06, "loss": 0.0272, "step": 27015 }, { "epoch": 2.9288811795316567, "grad_norm": 0.613554060459137, "learning_rate": 2.3792253010300305e-06, "loss": 0.0253, "step": 27016 }, { "epoch": 2.9289895923677363, "grad_norm": 0.7993758916854858, "learning_rate": 2.3755984331930946e-06, "loss": 0.038, "step": 27017 }, { "epoch": 2.9290980052038162, "grad_norm": 0.23987045884132385, "learning_rate": 2.3719715653561587e-06, "loss": 0.0036, "step": 27018 }, { "epoch": 2.929206418039896, "grad_norm": 0.14490655064582825, "learning_rate": 2.3683446975192224e-06, "loss": 0.004, "step": 27019 }, { "epoch": 2.9293148308759758, "grad_norm": 0.17447493970394135, "learning_rate": 2.3647178296822865e-06, "loss": 0.0047, "step": 27020 }, { "epoch": 2.9294232437120558, "grad_norm": 0.5957038998603821, "learning_rate": 2.3610909618453506e-06, "loss": 0.0325, "step": 27021 }, { "epoch": 2.9295316565481353, "grad_norm": 0.6457841992378235, "learning_rate": 2.3574640940084143e-06, "loss": 0.0183, "step": 27022 }, { "epoch": 2.929640069384215, "grad_norm": 0.06676426529884338, "learning_rate": 2.3538372261714784e-06, "loss": 0.0009, "step": 27023 }, { "epoch": 2.929748482220295, "grad_norm": 0.7509095072746277, "learning_rate": 2.3502103583345425e-06, "loss": 0.0402, "step": 27024 }, { "epoch": 2.929856895056375, "grad_norm": 0.16677938401699066, "learning_rate": 2.3465834904976066e-06, "loss": 0.006, "step": 27025 }, { "epoch": 2.9299653078924544, "grad_norm": 0.3659834563732147, "learning_rate": 2.3429566226606703e-06, "loss": 0.0104, "step": 27026 }, { "epoch": 2.9300737207285343, "grad_norm": 0.5933224558830261, "learning_rate": 2.3393297548237344e-06, "loss": 0.0425, "step": 27027 }, { "epoch": 2.930182133564614, "grad_norm": 0.19359081983566284, "learning_rate": 2.335702886986798e-06, "loss": 0.0046, "step": 27028 }, { "epoch": 2.930290546400694, "grad_norm": 0.2158060371875763, "learning_rate": 2.332076019149862e-06, "loss": 0.0115, "step": 27029 }, { "epoch": 2.930398959236774, "grad_norm": 0.21626198291778564, "learning_rate": 2.3284491513129263e-06, "loss": 0.0152, "step": 27030 }, { "epoch": 2.9305073720728534, "grad_norm": 0.08657363802194595, "learning_rate": 2.3248222834759904e-06, "loss": 0.0023, "step": 27031 }, { "epoch": 2.9306157849089334, "grad_norm": 0.6545466184616089, "learning_rate": 2.3211954156390545e-06, "loss": 0.0173, "step": 27032 }, { "epoch": 2.930724197745013, "grad_norm": 0.2297113984823227, "learning_rate": 2.317568547802118e-06, "loss": 0.0073, "step": 27033 }, { "epoch": 2.930832610581093, "grad_norm": 0.1873278170824051, "learning_rate": 2.3139416799651823e-06, "loss": 0.0045, "step": 27034 }, { "epoch": 2.9309410234171724, "grad_norm": 0.27301904559135437, "learning_rate": 2.310314812128246e-06, "loss": 0.0135, "step": 27035 }, { "epoch": 2.9310494362532524, "grad_norm": 0.13885243237018585, "learning_rate": 2.30668794429131e-06, "loss": 0.006, "step": 27036 }, { "epoch": 2.931157849089332, "grad_norm": 0.2712904214859009, "learning_rate": 2.303061076454374e-06, "loss": 0.0082, "step": 27037 }, { "epoch": 2.931266261925412, "grad_norm": 0.245233416557312, "learning_rate": 2.2994342086174383e-06, "loss": 0.0057, "step": 27038 }, { "epoch": 2.931374674761492, "grad_norm": 0.11341529339551926, "learning_rate": 2.295807340780502e-06, "loss": 0.0039, "step": 27039 }, { "epoch": 2.9314830875975715, "grad_norm": 0.1261885166168213, "learning_rate": 2.292180472943566e-06, "loss": 0.0038, "step": 27040 }, { "epoch": 2.9315915004336515, "grad_norm": 0.03651398792862892, "learning_rate": 2.28855360510663e-06, "loss": 0.0006, "step": 27041 }, { "epoch": 2.931699913269731, "grad_norm": 0.17955803871154785, "learning_rate": 2.284926737269694e-06, "loss": 0.0069, "step": 27042 }, { "epoch": 2.931808326105811, "grad_norm": 0.04190392792224884, "learning_rate": 2.281299869432758e-06, "loss": 0.0006, "step": 27043 }, { "epoch": 2.931916738941891, "grad_norm": 0.054528187960386276, "learning_rate": 2.277673001595822e-06, "loss": 0.0019, "step": 27044 }, { "epoch": 2.9320251517779705, "grad_norm": 0.1918078362941742, "learning_rate": 2.274046133758886e-06, "loss": 0.0024, "step": 27045 }, { "epoch": 2.93213356461405, "grad_norm": 0.0773109570145607, "learning_rate": 2.27041926592195e-06, "loss": 0.0016, "step": 27046 }, { "epoch": 2.93224197745013, "grad_norm": 0.09334615617990494, "learning_rate": 2.266792398085014e-06, "loss": 0.0049, "step": 27047 }, { "epoch": 2.93235039028621, "grad_norm": 0.013589494861662388, "learning_rate": 2.2631655302480776e-06, "loss": 0.0005, "step": 27048 }, { "epoch": 2.9324588031222896, "grad_norm": 0.09428440034389496, "learning_rate": 2.2595386624111417e-06, "loss": 0.0019, "step": 27049 }, { "epoch": 2.9325672159583696, "grad_norm": 0.025603272020816803, "learning_rate": 2.255911794574206e-06, "loss": 0.0006, "step": 27050 }, { "epoch": 2.932675628794449, "grad_norm": 0.12793783843517303, "learning_rate": 2.25228492673727e-06, "loss": 0.0045, "step": 27051 }, { "epoch": 2.932784041630529, "grad_norm": 0.6783414483070374, "learning_rate": 2.248658058900334e-06, "loss": 0.0153, "step": 27052 }, { "epoch": 2.932892454466609, "grad_norm": 0.10081004351377487, "learning_rate": 2.2450311910633977e-06, "loss": 0.0012, "step": 27053 }, { "epoch": 2.9330008673026886, "grad_norm": 0.11879368871450424, "learning_rate": 2.241404323226462e-06, "loss": 0.0036, "step": 27054 }, { "epoch": 2.9331092801387686, "grad_norm": 0.9179801940917969, "learning_rate": 2.2377774553895255e-06, "loss": 0.012, "step": 27055 }, { "epoch": 2.933217692974848, "grad_norm": 0.0982750803232193, "learning_rate": 2.2341505875525896e-06, "loss": 0.0028, "step": 27056 }, { "epoch": 2.933326105810928, "grad_norm": 0.6981083154678345, "learning_rate": 2.2305237197156537e-06, "loss": 0.037, "step": 27057 }, { "epoch": 2.9334345186470077, "grad_norm": 0.14417682588100433, "learning_rate": 2.226896851878718e-06, "loss": 0.0091, "step": 27058 }, { "epoch": 2.9335429314830876, "grad_norm": 0.16743779182434082, "learning_rate": 2.223269984041782e-06, "loss": 0.0032, "step": 27059 }, { "epoch": 2.933651344319167, "grad_norm": 0.3432527482509613, "learning_rate": 2.2196431162048456e-06, "loss": 0.0202, "step": 27060 }, { "epoch": 2.933759757155247, "grad_norm": 0.0914430022239685, "learning_rate": 2.2160162483679093e-06, "loss": 0.0023, "step": 27061 }, { "epoch": 2.933868169991327, "grad_norm": 0.29660025238990784, "learning_rate": 2.2123893805309734e-06, "loss": 0.0079, "step": 27062 }, { "epoch": 2.9339765828274067, "grad_norm": 0.006374336779117584, "learning_rate": 2.2087625126940375e-06, "loss": 0.0003, "step": 27063 }, { "epoch": 2.9340849956634867, "grad_norm": 0.21158716082572937, "learning_rate": 2.2051356448571016e-06, "loss": 0.0102, "step": 27064 }, { "epoch": 2.9341934084995662, "grad_norm": 0.6711576581001282, "learning_rate": 2.2015087770201657e-06, "loss": 0.0103, "step": 27065 }, { "epoch": 2.934301821335646, "grad_norm": 0.37261679768562317, "learning_rate": 2.1978819091832294e-06, "loss": 0.0041, "step": 27066 }, { "epoch": 2.934410234171726, "grad_norm": 0.08034593611955643, "learning_rate": 2.1942550413462935e-06, "loss": 0.0021, "step": 27067 }, { "epoch": 2.9345186470078057, "grad_norm": 0.2485874891281128, "learning_rate": 2.190628173509357e-06, "loss": 0.0116, "step": 27068 }, { "epoch": 2.9346270598438853, "grad_norm": 0.21448904275894165, "learning_rate": 2.1870013056724213e-06, "loss": 0.0075, "step": 27069 }, { "epoch": 2.9347354726799653, "grad_norm": 0.1953466385602951, "learning_rate": 2.1833744378354854e-06, "loss": 0.011, "step": 27070 }, { "epoch": 2.9348438855160452, "grad_norm": 0.34414049983024597, "learning_rate": 2.1797475699985495e-06, "loss": 0.006, "step": 27071 }, { "epoch": 2.934952298352125, "grad_norm": 1.06207275390625, "learning_rate": 2.1761207021616136e-06, "loss": 0.0198, "step": 27072 }, { "epoch": 2.9350607111882048, "grad_norm": 0.15254555642604828, "learning_rate": 2.1724938343246773e-06, "loss": 0.01, "step": 27073 }, { "epoch": 2.9351691240242843, "grad_norm": 0.07170450687408447, "learning_rate": 2.168866966487741e-06, "loss": 0.0035, "step": 27074 }, { "epoch": 2.9352775368603643, "grad_norm": 0.5080083608627319, "learning_rate": 2.165240098650805e-06, "loss": 0.0355, "step": 27075 }, { "epoch": 2.9353859496964443, "grad_norm": 0.17877382040023804, "learning_rate": 2.161613230813869e-06, "loss": 0.0157, "step": 27076 }, { "epoch": 2.935494362532524, "grad_norm": 0.06341256201267242, "learning_rate": 2.1579863629769333e-06, "loss": 0.0022, "step": 27077 }, { "epoch": 2.9356027753686034, "grad_norm": 0.20580701529979706, "learning_rate": 2.1543594951399974e-06, "loss": 0.0095, "step": 27078 }, { "epoch": 2.9357111882046834, "grad_norm": 0.335415244102478, "learning_rate": 2.1507326273030615e-06, "loss": 0.0191, "step": 27079 }, { "epoch": 2.9358196010407633, "grad_norm": 0.17467959225177765, "learning_rate": 2.147105759466125e-06, "loss": 0.0078, "step": 27080 }, { "epoch": 2.935928013876843, "grad_norm": 0.6908308267593384, "learning_rate": 2.143478891629189e-06, "loss": 0.0379, "step": 27081 }, { "epoch": 2.936036426712923, "grad_norm": 0.013465547934174538, "learning_rate": 2.139852023792253e-06, "loss": 0.0004, "step": 27082 }, { "epoch": 2.9361448395490024, "grad_norm": 0.290541410446167, "learning_rate": 2.136225155955317e-06, "loss": 0.0065, "step": 27083 }, { "epoch": 2.9362532523850824, "grad_norm": 0.1719875931739807, "learning_rate": 2.132598288118381e-06, "loss": 0.0049, "step": 27084 }, { "epoch": 2.9363616652211624, "grad_norm": 0.48784375190734863, "learning_rate": 2.1289714202814453e-06, "loss": 0.0078, "step": 27085 }, { "epoch": 2.936470078057242, "grad_norm": 0.9903083443641663, "learning_rate": 2.1253445524445094e-06, "loss": 0.0175, "step": 27086 }, { "epoch": 2.936578490893322, "grad_norm": 0.04975248873233795, "learning_rate": 2.121717684607573e-06, "loss": 0.003, "step": 27087 }, { "epoch": 2.9366869037294014, "grad_norm": 0.05096400901675224, "learning_rate": 2.1180908167706367e-06, "loss": 0.0014, "step": 27088 }, { "epoch": 2.9367953165654814, "grad_norm": 0.05664130672812462, "learning_rate": 2.114463948933701e-06, "loss": 0.0009, "step": 27089 }, { "epoch": 2.9369037294015614, "grad_norm": 0.3993097245693207, "learning_rate": 2.110837081096765e-06, "loss": 0.0076, "step": 27090 }, { "epoch": 2.937012142237641, "grad_norm": 0.45267340540885925, "learning_rate": 2.107210213259829e-06, "loss": 0.0214, "step": 27091 }, { "epoch": 2.9371205550737205, "grad_norm": 0.051204729825258255, "learning_rate": 2.103583345422893e-06, "loss": 0.0022, "step": 27092 }, { "epoch": 2.9372289679098005, "grad_norm": 0.05855122581124306, "learning_rate": 2.099956477585957e-06, "loss": 0.0015, "step": 27093 }, { "epoch": 2.9373373807458805, "grad_norm": 0.6796669960021973, "learning_rate": 2.0963296097490205e-06, "loss": 0.0137, "step": 27094 }, { "epoch": 2.93744579358196, "grad_norm": 0.11678621917963028, "learning_rate": 2.0927027419120846e-06, "loss": 0.0033, "step": 27095 }, { "epoch": 2.93755420641804, "grad_norm": 0.043750911951065063, "learning_rate": 2.0890758740751487e-06, "loss": 0.001, "step": 27096 }, { "epoch": 2.9376626192541195, "grad_norm": 0.6074960827827454, "learning_rate": 2.085449006238213e-06, "loss": 0.0265, "step": 27097 }, { "epoch": 2.9377710320901995, "grad_norm": 0.2249925434589386, "learning_rate": 2.081822138401277e-06, "loss": 0.0079, "step": 27098 }, { "epoch": 2.9378794449262795, "grad_norm": 0.45162156224250793, "learning_rate": 2.078195270564341e-06, "loss": 0.0108, "step": 27099 }, { "epoch": 2.937987857762359, "grad_norm": 0.123451828956604, "learning_rate": 2.0745684027274047e-06, "loss": 0.0053, "step": 27100 }, { "epoch": 2.9380962705984386, "grad_norm": 0.027811532840132713, "learning_rate": 2.0709415348904684e-06, "loss": 0.0007, "step": 27101 }, { "epoch": 2.9382046834345186, "grad_norm": 0.42685750126838684, "learning_rate": 2.0673146670535325e-06, "loss": 0.0109, "step": 27102 }, { "epoch": 2.9383130962705986, "grad_norm": 0.8062107563018799, "learning_rate": 2.0636877992165966e-06, "loss": 0.015, "step": 27103 }, { "epoch": 2.938421509106678, "grad_norm": 0.17341136932373047, "learning_rate": 2.0600609313796607e-06, "loss": 0.0057, "step": 27104 }, { "epoch": 2.938529921942758, "grad_norm": 0.02147076278924942, "learning_rate": 2.056434063542725e-06, "loss": 0.0007, "step": 27105 }, { "epoch": 2.9386383347788376, "grad_norm": 0.23688708245754242, "learning_rate": 2.0528071957057885e-06, "loss": 0.0132, "step": 27106 }, { "epoch": 2.9387467476149176, "grad_norm": 0.42524200677871704, "learning_rate": 2.049180327868852e-06, "loss": 0.0352, "step": 27107 }, { "epoch": 2.9388551604509976, "grad_norm": 0.5223076939582825, "learning_rate": 2.0455534600319163e-06, "loss": 0.0276, "step": 27108 }, { "epoch": 2.938963573287077, "grad_norm": 0.13908828794956207, "learning_rate": 2.0419265921949804e-06, "loss": 0.0069, "step": 27109 }, { "epoch": 2.939071986123157, "grad_norm": 0.04560648277401924, "learning_rate": 2.0382997243580445e-06, "loss": 0.0017, "step": 27110 }, { "epoch": 2.9391803989592367, "grad_norm": 1.1566106081008911, "learning_rate": 2.0346728565211086e-06, "loss": 0.0498, "step": 27111 }, { "epoch": 2.9392888117953166, "grad_norm": 0.2978690564632416, "learning_rate": 2.0310459886841727e-06, "loss": 0.0036, "step": 27112 }, { "epoch": 2.9393972246313966, "grad_norm": 0.24381160736083984, "learning_rate": 2.0274191208472364e-06, "loss": 0.0104, "step": 27113 }, { "epoch": 2.939505637467476, "grad_norm": 0.04856814071536064, "learning_rate": 2.0237922530103e-06, "loss": 0.0014, "step": 27114 }, { "epoch": 2.9396140503035557, "grad_norm": 0.14154694974422455, "learning_rate": 2.020165385173364e-06, "loss": 0.0079, "step": 27115 }, { "epoch": 2.9397224631396357, "grad_norm": 0.01129171997308731, "learning_rate": 2.0165385173364283e-06, "loss": 0.0004, "step": 27116 }, { "epoch": 2.9398308759757157, "grad_norm": 0.34478136897087097, "learning_rate": 2.0129116494994924e-06, "loss": 0.0117, "step": 27117 }, { "epoch": 2.9399392888117952, "grad_norm": 0.40779826045036316, "learning_rate": 2.0092847816625565e-06, "loss": 0.0253, "step": 27118 }, { "epoch": 2.940047701647875, "grad_norm": 0.8019106984138489, "learning_rate": 2.00565791382562e-06, "loss": 0.0319, "step": 27119 }, { "epoch": 2.9401561144839548, "grad_norm": 0.2330586463212967, "learning_rate": 2.0020310459886843e-06, "loss": 0.0059, "step": 27120 }, { "epoch": 2.9402645273200347, "grad_norm": 0.15050667524337769, "learning_rate": 1.998404178151748e-06, "loss": 0.0052, "step": 27121 }, { "epoch": 2.9403729401561147, "grad_norm": 0.08467765897512436, "learning_rate": 1.994777310314812e-06, "loss": 0.0027, "step": 27122 }, { "epoch": 2.9404813529921943, "grad_norm": 0.5304864048957825, "learning_rate": 1.991150442477876e-06, "loss": 0.0214, "step": 27123 }, { "epoch": 2.940589765828274, "grad_norm": 0.03376775607466698, "learning_rate": 1.9875235746409403e-06, "loss": 0.0006, "step": 27124 }, { "epoch": 2.940698178664354, "grad_norm": 0.030004125088453293, "learning_rate": 1.9838967068040044e-06, "loss": 0.0013, "step": 27125 }, { "epoch": 2.9408065915004338, "grad_norm": 0.02029966376721859, "learning_rate": 1.980269838967068e-06, "loss": 0.0007, "step": 27126 }, { "epoch": 2.9409150043365133, "grad_norm": 0.07459286600351334, "learning_rate": 1.9766429711301318e-06, "loss": 0.0021, "step": 27127 }, { "epoch": 2.9410234171725933, "grad_norm": 0.23867477476596832, "learning_rate": 1.973016103293196e-06, "loss": 0.0028, "step": 27128 }, { "epoch": 2.941131830008673, "grad_norm": 0.10567045956850052, "learning_rate": 1.96938923545626e-06, "loss": 0.005, "step": 27129 }, { "epoch": 2.941240242844753, "grad_norm": 0.5220819711685181, "learning_rate": 1.965762367619324e-06, "loss": 0.0333, "step": 27130 }, { "epoch": 2.941348655680833, "grad_norm": 0.024770088493824005, "learning_rate": 1.962135499782388e-06, "loss": 0.0005, "step": 27131 }, { "epoch": 2.9414570685169124, "grad_norm": 0.4738270044326782, "learning_rate": 1.9585086319454523e-06, "loss": 0.0275, "step": 27132 }, { "epoch": 2.9415654813529923, "grad_norm": 0.08558323979377747, "learning_rate": 1.954881764108516e-06, "loss": 0.0069, "step": 27133 }, { "epoch": 2.941673894189072, "grad_norm": 0.6398273706436157, "learning_rate": 1.9512548962715797e-06, "loss": 0.0185, "step": 27134 }, { "epoch": 2.941782307025152, "grad_norm": 0.1856488734483719, "learning_rate": 1.9476280284346438e-06, "loss": 0.0083, "step": 27135 }, { "epoch": 2.9418907198612314, "grad_norm": 0.1058722659945488, "learning_rate": 1.944001160597708e-06, "loss": 0.0034, "step": 27136 }, { "epoch": 2.9419991326973114, "grad_norm": 0.00497696828097105, "learning_rate": 1.940374292760772e-06, "loss": 0.0002, "step": 27137 }, { "epoch": 2.942107545533391, "grad_norm": 0.07695899903774261, "learning_rate": 1.936747424923836e-06, "loss": 0.0021, "step": 27138 }, { "epoch": 2.942215958369471, "grad_norm": 0.22852851450443268, "learning_rate": 1.9331205570868998e-06, "loss": 0.0056, "step": 27139 }, { "epoch": 2.942324371205551, "grad_norm": 0.07133394479751587, "learning_rate": 1.929493689249964e-06, "loss": 0.0023, "step": 27140 }, { "epoch": 2.9424327840416304, "grad_norm": 0.7455148100852966, "learning_rate": 1.9258668214130275e-06, "loss": 0.0305, "step": 27141 }, { "epoch": 2.9425411968777104, "grad_norm": 0.10365469753742218, "learning_rate": 1.9222399535760916e-06, "loss": 0.0019, "step": 27142 }, { "epoch": 2.94264960971379, "grad_norm": 0.01877663843333721, "learning_rate": 1.9186130857391558e-06, "loss": 0.0006, "step": 27143 }, { "epoch": 2.94275802254987, "grad_norm": 0.14308352768421173, "learning_rate": 1.91498621790222e-06, "loss": 0.0018, "step": 27144 }, { "epoch": 2.94286643538595, "grad_norm": 0.410430371761322, "learning_rate": 1.911359350065284e-06, "loss": 0.0107, "step": 27145 }, { "epoch": 2.9429748482220295, "grad_norm": 0.1600872278213501, "learning_rate": 1.9077324822283476e-06, "loss": 0.0052, "step": 27146 }, { "epoch": 2.943083261058109, "grad_norm": 0.5963397026062012, "learning_rate": 1.9041056143914117e-06, "loss": 0.0409, "step": 27147 }, { "epoch": 2.943191673894189, "grad_norm": 0.09438954293727875, "learning_rate": 1.9004787465544754e-06, "loss": 0.0016, "step": 27148 }, { "epoch": 2.943300086730269, "grad_norm": 0.3323172628879547, "learning_rate": 1.8968518787175395e-06, "loss": 0.0224, "step": 27149 }, { "epoch": 2.9434084995663485, "grad_norm": 0.17778320610523224, "learning_rate": 1.8932250108806036e-06, "loss": 0.0073, "step": 27150 }, { "epoch": 2.9435169124024285, "grad_norm": 0.23579555749893188, "learning_rate": 1.8895981430436675e-06, "loss": 0.0232, "step": 27151 }, { "epoch": 2.943625325238508, "grad_norm": 0.1671725958585739, "learning_rate": 1.8859712752067316e-06, "loss": 0.0066, "step": 27152 }, { "epoch": 2.943733738074588, "grad_norm": 0.5645357966423035, "learning_rate": 1.8823444073697957e-06, "loss": 0.0193, "step": 27153 }, { "epoch": 2.943842150910668, "grad_norm": 0.38669830560684204, "learning_rate": 1.8787175395328596e-06, "loss": 0.0184, "step": 27154 }, { "epoch": 2.9439505637467476, "grad_norm": 0.36713340878486633, "learning_rate": 1.8750906716959233e-06, "loss": 0.01, "step": 27155 }, { "epoch": 2.9440589765828276, "grad_norm": 0.02421564981341362, "learning_rate": 1.8714638038589874e-06, "loss": 0.0011, "step": 27156 }, { "epoch": 2.944167389418907, "grad_norm": 1.2906641960144043, "learning_rate": 1.8678369360220515e-06, "loss": 0.0163, "step": 27157 }, { "epoch": 2.944275802254987, "grad_norm": 0.0958399698138237, "learning_rate": 1.8642100681851154e-06, "loss": 0.0041, "step": 27158 }, { "epoch": 2.9443842150910666, "grad_norm": 0.011394387111067772, "learning_rate": 1.8605832003481795e-06, "loss": 0.0005, "step": 27159 }, { "epoch": 2.9444926279271466, "grad_norm": 0.13022200763225555, "learning_rate": 1.8569563325112434e-06, "loss": 0.0024, "step": 27160 }, { "epoch": 2.944601040763226, "grad_norm": 0.23077499866485596, "learning_rate": 1.853329464674307e-06, "loss": 0.0116, "step": 27161 }, { "epoch": 2.944709453599306, "grad_norm": 0.18895813822746277, "learning_rate": 1.8497025968373712e-06, "loss": 0.0066, "step": 27162 }, { "epoch": 2.944817866435386, "grad_norm": 0.1175924688577652, "learning_rate": 1.8460757290004353e-06, "loss": 0.0065, "step": 27163 }, { "epoch": 2.9449262792714657, "grad_norm": 0.2811516225337982, "learning_rate": 1.8424488611634992e-06, "loss": 0.0111, "step": 27164 }, { "epoch": 2.9450346921075456, "grad_norm": 0.29361167550086975, "learning_rate": 1.8388219933265633e-06, "loss": 0.0035, "step": 27165 }, { "epoch": 2.945143104943625, "grad_norm": 0.40273648500442505, "learning_rate": 1.8351951254896274e-06, "loss": 0.0191, "step": 27166 }, { "epoch": 2.945251517779705, "grad_norm": 0.09091491997241974, "learning_rate": 1.8315682576526913e-06, "loss": 0.0033, "step": 27167 }, { "epoch": 2.945359930615785, "grad_norm": 0.6695307493209839, "learning_rate": 1.827941389815755e-06, "loss": 0.0272, "step": 27168 }, { "epoch": 2.9454683434518647, "grad_norm": 0.251944363117218, "learning_rate": 1.824314521978819e-06, "loss": 0.0047, "step": 27169 }, { "epoch": 2.9455767562879442, "grad_norm": 0.30388274788856506, "learning_rate": 1.8206876541418832e-06, "loss": 0.0099, "step": 27170 }, { "epoch": 2.9456851691240242, "grad_norm": 0.2593817412853241, "learning_rate": 1.817060786304947e-06, "loss": 0.0145, "step": 27171 }, { "epoch": 2.945793581960104, "grad_norm": 0.10504504293203354, "learning_rate": 1.8134339184680112e-06, "loss": 0.0036, "step": 27172 }, { "epoch": 2.9459019947961838, "grad_norm": 0.0834280401468277, "learning_rate": 1.809807050631075e-06, "loss": 0.0034, "step": 27173 }, { "epoch": 2.9460104076322637, "grad_norm": 1.0441974401474, "learning_rate": 1.8061801827941392e-06, "loss": 0.0119, "step": 27174 }, { "epoch": 2.9461188204683433, "grad_norm": 1.0687878131866455, "learning_rate": 1.8025533149572029e-06, "loss": 0.0272, "step": 27175 }, { "epoch": 2.9462272333044233, "grad_norm": 0.03841376304626465, "learning_rate": 1.798926447120267e-06, "loss": 0.0012, "step": 27176 }, { "epoch": 2.9463356461405033, "grad_norm": 0.38878190517425537, "learning_rate": 1.7952995792833309e-06, "loss": 0.0097, "step": 27177 }, { "epoch": 2.946444058976583, "grad_norm": 0.591892421245575, "learning_rate": 1.791672711446395e-06, "loss": 0.0197, "step": 27178 }, { "epoch": 2.9465524718126628, "grad_norm": 0.2759312391281128, "learning_rate": 1.788045843609459e-06, "loss": 0.0082, "step": 27179 }, { "epoch": 2.9466608846487423, "grad_norm": 0.51516193151474, "learning_rate": 1.784418975772523e-06, "loss": 0.012, "step": 27180 }, { "epoch": 2.9467692974848223, "grad_norm": 0.09270429611206055, "learning_rate": 1.780792107935587e-06, "loss": 0.0011, "step": 27181 }, { "epoch": 2.946877710320902, "grad_norm": 0.5044288635253906, "learning_rate": 1.7771652400986508e-06, "loss": 0.0229, "step": 27182 }, { "epoch": 2.946986123156982, "grad_norm": 0.23048825562000275, "learning_rate": 1.7735383722617149e-06, "loss": 0.0171, "step": 27183 }, { "epoch": 2.9470945359930614, "grad_norm": 0.07871027290821075, "learning_rate": 1.7699115044247788e-06, "loss": 0.0021, "step": 27184 }, { "epoch": 2.9472029488291414, "grad_norm": 0.09750145673751831, "learning_rate": 1.7662846365878429e-06, "loss": 0.0026, "step": 27185 }, { "epoch": 2.9473113616652213, "grad_norm": 1.34622061252594, "learning_rate": 1.762657768750907e-06, "loss": 0.0236, "step": 27186 }, { "epoch": 2.947419774501301, "grad_norm": 0.1277388632297516, "learning_rate": 1.7590309009139709e-06, "loss": 0.0089, "step": 27187 }, { "epoch": 2.947528187337381, "grad_norm": 0.19663049280643463, "learning_rate": 1.7554040330770346e-06, "loss": 0.0069, "step": 27188 }, { "epoch": 2.9476366001734604, "grad_norm": 0.27811622619628906, "learning_rate": 1.7517771652400987e-06, "loss": 0.0377, "step": 27189 }, { "epoch": 2.9477450130095404, "grad_norm": 0.4779227077960968, "learning_rate": 1.7481502974031625e-06, "loss": 0.0259, "step": 27190 }, { "epoch": 2.9478534258456204, "grad_norm": 0.2518920302391052, "learning_rate": 1.7445234295662267e-06, "loss": 0.0093, "step": 27191 }, { "epoch": 2.9479618386817, "grad_norm": 0.15114913880825043, "learning_rate": 1.7408965617292908e-06, "loss": 0.0031, "step": 27192 }, { "epoch": 2.9480702515177795, "grad_norm": 0.2970113456249237, "learning_rate": 1.7372696938923547e-06, "loss": 0.0113, "step": 27193 }, { "epoch": 2.9481786643538594, "grad_norm": 0.019025659188628197, "learning_rate": 1.7336428260554188e-06, "loss": 0.0004, "step": 27194 }, { "epoch": 2.9482870771899394, "grad_norm": 0.3378843069076538, "learning_rate": 1.7300159582184824e-06, "loss": 0.0073, "step": 27195 }, { "epoch": 2.948395490026019, "grad_norm": 0.16530300676822662, "learning_rate": 1.7263890903815465e-06, "loss": 0.0027, "step": 27196 }, { "epoch": 2.948503902862099, "grad_norm": 0.1297134906053543, "learning_rate": 1.7227622225446104e-06, "loss": 0.0033, "step": 27197 }, { "epoch": 2.9486123156981785, "grad_norm": 0.010465242899954319, "learning_rate": 1.7191353547076745e-06, "loss": 0.0006, "step": 27198 }, { "epoch": 2.9487207285342585, "grad_norm": 0.28965136408805847, "learning_rate": 1.7155084868707386e-06, "loss": 0.019, "step": 27199 }, { "epoch": 2.9488291413703385, "grad_norm": 0.08775326609611511, "learning_rate": 1.7118816190338025e-06, "loss": 0.0049, "step": 27200 }, { "epoch": 2.948937554206418, "grad_norm": 0.023980166763067245, "learning_rate": 1.7082547511968666e-06, "loss": 0.0009, "step": 27201 }, { "epoch": 2.949045967042498, "grad_norm": 0.15906693041324615, "learning_rate": 1.7046278833599303e-06, "loss": 0.0027, "step": 27202 }, { "epoch": 2.9491543798785775, "grad_norm": 0.6008946299552917, "learning_rate": 1.7010010155229944e-06, "loss": 0.0168, "step": 27203 }, { "epoch": 2.9492627927146575, "grad_norm": 0.08468622714281082, "learning_rate": 1.6973741476860583e-06, "loss": 0.0063, "step": 27204 }, { "epoch": 2.949371205550737, "grad_norm": 0.08283323794603348, "learning_rate": 1.6937472798491224e-06, "loss": 0.002, "step": 27205 }, { "epoch": 2.949479618386817, "grad_norm": 0.8997433185577393, "learning_rate": 1.6901204120121863e-06, "loss": 0.0334, "step": 27206 }, { "epoch": 2.9495880312228966, "grad_norm": 0.047190334647893906, "learning_rate": 1.6864935441752504e-06, "loss": 0.0021, "step": 27207 }, { "epoch": 2.9496964440589766, "grad_norm": 0.43842047452926636, "learning_rate": 1.6828666763383145e-06, "loss": 0.0177, "step": 27208 }, { "epoch": 2.9498048568950566, "grad_norm": 1.1833186149597168, "learning_rate": 1.6792398085013782e-06, "loss": 0.0115, "step": 27209 }, { "epoch": 2.949913269731136, "grad_norm": 0.18301381170749664, "learning_rate": 1.6756129406644421e-06, "loss": 0.0044, "step": 27210 }, { "epoch": 2.950021682567216, "grad_norm": 0.16769443452358246, "learning_rate": 1.6719860728275062e-06, "loss": 0.0029, "step": 27211 }, { "epoch": 2.9501300954032956, "grad_norm": 0.22960154712200165, "learning_rate": 1.6683592049905703e-06, "loss": 0.0149, "step": 27212 }, { "epoch": 2.9502385082393756, "grad_norm": 0.2669218182563782, "learning_rate": 1.6647323371536342e-06, "loss": 0.0115, "step": 27213 }, { "epoch": 2.9503469210754556, "grad_norm": 0.3790699541568756, "learning_rate": 1.6611054693166983e-06, "loss": 0.0153, "step": 27214 }, { "epoch": 2.950455333911535, "grad_norm": 0.10666069388389587, "learning_rate": 1.657478601479762e-06, "loss": 0.0027, "step": 27215 }, { "epoch": 2.9505637467476147, "grad_norm": 0.1566365659236908, "learning_rate": 1.653851733642826e-06, "loss": 0.0065, "step": 27216 }, { "epoch": 2.9506721595836947, "grad_norm": 1.2477245330810547, "learning_rate": 1.65022486580589e-06, "loss": 0.0233, "step": 27217 }, { "epoch": 2.9507805724197746, "grad_norm": 0.0731799378991127, "learning_rate": 1.646597997968954e-06, "loss": 0.0031, "step": 27218 }, { "epoch": 2.950888985255854, "grad_norm": 0.10084179043769836, "learning_rate": 1.642971130132018e-06, "loss": 0.0043, "step": 27219 }, { "epoch": 2.950997398091934, "grad_norm": 0.11797641962766647, "learning_rate": 1.639344262295082e-06, "loss": 0.0065, "step": 27220 }, { "epoch": 2.9511058109280137, "grad_norm": 0.9161997437477112, "learning_rate": 1.6357173944581462e-06, "loss": 0.0303, "step": 27221 }, { "epoch": 2.9512142237640937, "grad_norm": 0.3659031391143799, "learning_rate": 1.6320905266212099e-06, "loss": 0.0239, "step": 27222 }, { "epoch": 2.9513226366001737, "grad_norm": 0.02096499502658844, "learning_rate": 1.6284636587842738e-06, "loss": 0.0007, "step": 27223 }, { "epoch": 2.9514310494362532, "grad_norm": 0.13263504207134247, "learning_rate": 1.6248367909473379e-06, "loss": 0.0028, "step": 27224 }, { "epoch": 2.9515394622723328, "grad_norm": 0.3549148440361023, "learning_rate": 1.621209923110402e-06, "loss": 0.0356, "step": 27225 }, { "epoch": 2.9516478751084128, "grad_norm": 0.026554595679044724, "learning_rate": 1.6175830552734659e-06, "loss": 0.0005, "step": 27226 }, { "epoch": 2.9517562879444927, "grad_norm": 0.3428588807582855, "learning_rate": 1.61395618743653e-06, "loss": 0.0125, "step": 27227 }, { "epoch": 2.9518647007805723, "grad_norm": 0.0664813295006752, "learning_rate": 1.610329319599594e-06, "loss": 0.0017, "step": 27228 }, { "epoch": 2.9519731136166523, "grad_norm": 0.05231619253754616, "learning_rate": 1.6067024517626578e-06, "loss": 0.001, "step": 27229 }, { "epoch": 2.952081526452732, "grad_norm": 0.38511309027671814, "learning_rate": 1.6030755839257217e-06, "loss": 0.0231, "step": 27230 }, { "epoch": 2.952189939288812, "grad_norm": 0.21902316808700562, "learning_rate": 1.5994487160887858e-06, "loss": 0.0212, "step": 27231 }, { "epoch": 2.952298352124892, "grad_norm": 0.038534149527549744, "learning_rate": 1.5958218482518499e-06, "loss": 0.0018, "step": 27232 }, { "epoch": 2.9524067649609713, "grad_norm": 0.010786288417875767, "learning_rate": 1.5921949804149138e-06, "loss": 0.0002, "step": 27233 }, { "epoch": 2.9525151777970513, "grad_norm": 0.2568308711051941, "learning_rate": 1.5885681125779779e-06, "loss": 0.0131, "step": 27234 }, { "epoch": 2.952623590633131, "grad_norm": 0.49421587586402893, "learning_rate": 1.5849412447410418e-06, "loss": 0.0184, "step": 27235 }, { "epoch": 2.952732003469211, "grad_norm": 0.25352033972740173, "learning_rate": 1.5813143769041055e-06, "loss": 0.0078, "step": 27236 }, { "epoch": 2.952840416305291, "grad_norm": 0.1338663548231125, "learning_rate": 1.5776875090671696e-06, "loss": 0.0034, "step": 27237 }, { "epoch": 2.9529488291413704, "grad_norm": 0.010463283397257328, "learning_rate": 1.5740606412302337e-06, "loss": 0.0004, "step": 27238 }, { "epoch": 2.95305724197745, "grad_norm": 0.025637703016400337, "learning_rate": 1.5704337733932976e-06, "loss": 0.0016, "step": 27239 }, { "epoch": 2.95316565481353, "grad_norm": 0.4385586678981781, "learning_rate": 1.5668069055563617e-06, "loss": 0.0081, "step": 27240 }, { "epoch": 2.95327406764961, "grad_norm": 0.06634768098592758, "learning_rate": 1.5631800377194258e-06, "loss": 0.0034, "step": 27241 }, { "epoch": 2.9533824804856894, "grad_norm": 0.2726915776729584, "learning_rate": 1.5595531698824897e-06, "loss": 0.0021, "step": 27242 }, { "epoch": 2.9534908933217694, "grad_norm": 0.2407081127166748, "learning_rate": 1.5559263020455536e-06, "loss": 0.0074, "step": 27243 }, { "epoch": 2.953599306157849, "grad_norm": 0.3193092942237854, "learning_rate": 1.5522994342086174e-06, "loss": 0.0008, "step": 27244 }, { "epoch": 2.953707718993929, "grad_norm": 0.07939758896827698, "learning_rate": 1.5486725663716816e-06, "loss": 0.0028, "step": 27245 }, { "epoch": 2.953816131830009, "grad_norm": 0.15582157671451569, "learning_rate": 1.5450456985347454e-06, "loss": 0.0046, "step": 27246 }, { "epoch": 2.9539245446660884, "grad_norm": 0.041702695190906525, "learning_rate": 1.5414188306978096e-06, "loss": 0.0012, "step": 27247 }, { "epoch": 2.954032957502168, "grad_norm": 0.41647353768348694, "learning_rate": 1.5377919628608734e-06, "loss": 0.0171, "step": 27248 }, { "epoch": 2.954141370338248, "grad_norm": 0.2268003225326538, "learning_rate": 1.5341650950239373e-06, "loss": 0.0034, "step": 27249 }, { "epoch": 2.954249783174328, "grad_norm": 0.2257259339094162, "learning_rate": 1.5305382271870014e-06, "loss": 0.0083, "step": 27250 }, { "epoch": 2.9543581960104075, "grad_norm": 0.01904500648379326, "learning_rate": 1.5269113593500653e-06, "loss": 0.0005, "step": 27251 }, { "epoch": 2.9544666088464875, "grad_norm": 0.02861502766609192, "learning_rate": 1.5232844915131292e-06, "loss": 0.0007, "step": 27252 }, { "epoch": 2.954575021682567, "grad_norm": 0.18946591019630432, "learning_rate": 1.5196576236761933e-06, "loss": 0.0059, "step": 27253 }, { "epoch": 2.954683434518647, "grad_norm": 0.016255026683211327, "learning_rate": 1.5160307558392572e-06, "loss": 0.0007, "step": 27254 }, { "epoch": 2.954791847354727, "grad_norm": 0.11080655455589294, "learning_rate": 1.5124038880023213e-06, "loss": 0.0044, "step": 27255 }, { "epoch": 2.9549002601908065, "grad_norm": 0.4173935353755951, "learning_rate": 1.5087770201653852e-06, "loss": 0.0105, "step": 27256 }, { "epoch": 2.9550086730268865, "grad_norm": 0.4342551529407501, "learning_rate": 1.5051501523284493e-06, "loss": 0.0245, "step": 27257 }, { "epoch": 2.955117085862966, "grad_norm": 0.1210242509841919, "learning_rate": 1.5015232844915132e-06, "loss": 0.0025, "step": 27258 }, { "epoch": 2.955225498699046, "grad_norm": 0.32560020685195923, "learning_rate": 1.4978964166545771e-06, "loss": 0.0095, "step": 27259 }, { "epoch": 2.955333911535126, "grad_norm": 0.3059634864330292, "learning_rate": 1.4942695488176412e-06, "loss": 0.0317, "step": 27260 }, { "epoch": 2.9554423243712056, "grad_norm": 0.013504276052117348, "learning_rate": 1.4906426809807051e-06, "loss": 0.0006, "step": 27261 }, { "epoch": 2.955550737207285, "grad_norm": 0.16168834269046783, "learning_rate": 1.487015813143769e-06, "loss": 0.0049, "step": 27262 }, { "epoch": 2.955659150043365, "grad_norm": 0.159016415476799, "learning_rate": 1.4833889453068331e-06, "loss": 0.0023, "step": 27263 }, { "epoch": 2.955767562879445, "grad_norm": 0.03908640146255493, "learning_rate": 1.4797620774698972e-06, "loss": 0.0017, "step": 27264 }, { "epoch": 2.9558759757155246, "grad_norm": 0.03311764821410179, "learning_rate": 1.4761352096329611e-06, "loss": 0.0011, "step": 27265 }, { "epoch": 2.9559843885516046, "grad_norm": 0.26832500100135803, "learning_rate": 1.472508341796025e-06, "loss": 0.009, "step": 27266 }, { "epoch": 2.956092801387684, "grad_norm": 0.0825762152671814, "learning_rate": 1.4688814739590891e-06, "loss": 0.0027, "step": 27267 }, { "epoch": 2.956201214223764, "grad_norm": 0.18468859791755676, "learning_rate": 1.465254606122153e-06, "loss": 0.0113, "step": 27268 }, { "epoch": 2.956309627059844, "grad_norm": 0.01661319099366665, "learning_rate": 1.461627738285217e-06, "loss": 0.0008, "step": 27269 }, { "epoch": 2.9564180398959237, "grad_norm": 0.2472582310438156, "learning_rate": 1.458000870448281e-06, "loss": 0.016, "step": 27270 }, { "epoch": 2.956526452732003, "grad_norm": 0.09157047420740128, "learning_rate": 1.454374002611345e-06, "loss": 0.0069, "step": 27271 }, { "epoch": 2.956634865568083, "grad_norm": 0.6468042135238647, "learning_rate": 1.4507471347744088e-06, "loss": 0.0176, "step": 27272 }, { "epoch": 2.956743278404163, "grad_norm": 0.006518915295600891, "learning_rate": 1.447120266937473e-06, "loss": 0.0004, "step": 27273 }, { "epoch": 2.9568516912402427, "grad_norm": 0.8401753902435303, "learning_rate": 1.443493399100537e-06, "loss": 0.0784, "step": 27274 }, { "epoch": 2.9569601040763227, "grad_norm": 0.17080554366111755, "learning_rate": 1.4398665312636007e-06, "loss": 0.0084, "step": 27275 }, { "epoch": 2.9570685169124022, "grad_norm": 0.3111277222633362, "learning_rate": 1.4362396634266648e-06, "loss": 0.0085, "step": 27276 }, { "epoch": 2.9571769297484822, "grad_norm": 0.2903002202510834, "learning_rate": 1.4326127955897289e-06, "loss": 0.0153, "step": 27277 }, { "epoch": 2.957285342584562, "grad_norm": 0.05693913251161575, "learning_rate": 1.4289859277527928e-06, "loss": 0.0033, "step": 27278 }, { "epoch": 2.9573937554206418, "grad_norm": 0.4287244379520416, "learning_rate": 1.4253590599158567e-06, "loss": 0.0455, "step": 27279 }, { "epoch": 2.9575021682567217, "grad_norm": 0.223907470703125, "learning_rate": 1.4217321920789208e-06, "loss": 0.0044, "step": 27280 }, { "epoch": 2.9576105810928013, "grad_norm": 0.2652001976966858, "learning_rate": 1.4181053242419847e-06, "loss": 0.0112, "step": 27281 }, { "epoch": 2.9577189939288813, "grad_norm": 0.03968208655714989, "learning_rate": 1.4144784564050486e-06, "loss": 0.001, "step": 27282 }, { "epoch": 2.957827406764961, "grad_norm": 0.29128003120422363, "learning_rate": 1.4108515885681127e-06, "loss": 0.0118, "step": 27283 }, { "epoch": 2.957935819601041, "grad_norm": 0.27479270100593567, "learning_rate": 1.4072247207311768e-06, "loss": 0.0042, "step": 27284 }, { "epoch": 2.9580442324371203, "grad_norm": 0.3873399794101715, "learning_rate": 1.4035978528942405e-06, "loss": 0.0058, "step": 27285 }, { "epoch": 2.9581526452732003, "grad_norm": 0.4993467330932617, "learning_rate": 1.3999709850573046e-06, "loss": 0.0101, "step": 27286 }, { "epoch": 2.9582610581092803, "grad_norm": 0.04683474451303482, "learning_rate": 1.3963441172203687e-06, "loss": 0.0012, "step": 27287 }, { "epoch": 2.95836947094536, "grad_norm": 0.1548195630311966, "learning_rate": 1.3927172493834326e-06, "loss": 0.0061, "step": 27288 }, { "epoch": 2.95847788378144, "grad_norm": 0.41584697365760803, "learning_rate": 1.3890903815464965e-06, "loss": 0.0134, "step": 27289 }, { "epoch": 2.9585862966175194, "grad_norm": 0.3705047070980072, "learning_rate": 1.3854635137095606e-06, "loss": 0.0219, "step": 27290 }, { "epoch": 2.9586947094535994, "grad_norm": 0.1822313666343689, "learning_rate": 1.3818366458726245e-06, "loss": 0.003, "step": 27291 }, { "epoch": 2.9588031222896793, "grad_norm": 0.04535568132996559, "learning_rate": 1.3782097780356883e-06, "loss": 0.0013, "step": 27292 }, { "epoch": 2.958911535125759, "grad_norm": 0.31635183095932007, "learning_rate": 1.3745829101987525e-06, "loss": 0.0062, "step": 27293 }, { "epoch": 2.9590199479618384, "grad_norm": 0.19585125148296356, "learning_rate": 1.3709560423618166e-06, "loss": 0.0049, "step": 27294 }, { "epoch": 2.9591283607979184, "grad_norm": 0.10675062984228134, "learning_rate": 1.3673291745248802e-06, "loss": 0.0018, "step": 27295 }, { "epoch": 2.9592367736339984, "grad_norm": 0.7196853756904602, "learning_rate": 1.3637023066879443e-06, "loss": 0.0225, "step": 27296 }, { "epoch": 2.959345186470078, "grad_norm": 0.00830007717013359, "learning_rate": 1.3600754388510085e-06, "loss": 0.0005, "step": 27297 }, { "epoch": 2.959453599306158, "grad_norm": 0.42419949173927307, "learning_rate": 1.3564485710140721e-06, "loss": 0.0081, "step": 27298 }, { "epoch": 2.9595620121422375, "grad_norm": 0.12879174947738647, "learning_rate": 1.3528217031771362e-06, "loss": 0.0066, "step": 27299 }, { "epoch": 2.9596704249783174, "grad_norm": 0.13822109997272491, "learning_rate": 1.3491948353402003e-06, "loss": 0.006, "step": 27300 }, { "epoch": 2.9597788378143974, "grad_norm": 0.35484984517097473, "learning_rate": 1.3455679675032642e-06, "loss": 0.0039, "step": 27301 }, { "epoch": 2.959887250650477, "grad_norm": 0.4939919114112854, "learning_rate": 1.3419410996663281e-06, "loss": 0.0267, "step": 27302 }, { "epoch": 2.959995663486557, "grad_norm": 0.07630804181098938, "learning_rate": 1.3383142318293922e-06, "loss": 0.0016, "step": 27303 }, { "epoch": 2.9601040763226365, "grad_norm": 0.5079255104064941, "learning_rate": 1.3346873639924561e-06, "loss": 0.0172, "step": 27304 }, { "epoch": 2.9602124891587165, "grad_norm": 0.28443753719329834, "learning_rate": 1.33106049615552e-06, "loss": 0.0106, "step": 27305 }, { "epoch": 2.960320901994796, "grad_norm": 0.29648706316947937, "learning_rate": 1.3274336283185841e-06, "loss": 0.0145, "step": 27306 }, { "epoch": 2.960429314830876, "grad_norm": 0.27408885955810547, "learning_rate": 1.3238067604816482e-06, "loss": 0.007, "step": 27307 }, { "epoch": 2.9605377276669556, "grad_norm": 0.5131554007530212, "learning_rate": 1.3201798926447121e-06, "loss": 0.0238, "step": 27308 }, { "epoch": 2.9606461405030355, "grad_norm": 0.09344600141048431, "learning_rate": 1.316553024807776e-06, "loss": 0.0035, "step": 27309 }, { "epoch": 2.9607545533391155, "grad_norm": 0.3116630017757416, "learning_rate": 1.3129261569708401e-06, "loss": 0.0122, "step": 27310 }, { "epoch": 2.960862966175195, "grad_norm": 0.15834417939186096, "learning_rate": 1.309299289133904e-06, "loss": 0.011, "step": 27311 }, { "epoch": 2.960971379011275, "grad_norm": 0.3030621409416199, "learning_rate": 1.305672421296968e-06, "loss": 0.0119, "step": 27312 }, { "epoch": 2.9610797918473546, "grad_norm": 0.4303765296936035, "learning_rate": 1.302045553460032e-06, "loss": 0.0197, "step": 27313 }, { "epoch": 2.9611882046834346, "grad_norm": 0.04257897660136223, "learning_rate": 1.298418685623096e-06, "loss": 0.0019, "step": 27314 }, { "epoch": 2.9612966175195146, "grad_norm": 0.14755114912986755, "learning_rate": 1.2947918177861598e-06, "loss": 0.0056, "step": 27315 }, { "epoch": 2.961405030355594, "grad_norm": 0.013669239357113838, "learning_rate": 1.291164949949224e-06, "loss": 0.0006, "step": 27316 }, { "epoch": 2.9615134431916736, "grad_norm": 0.4323631823062897, "learning_rate": 1.287538082112288e-06, "loss": 0.0058, "step": 27317 }, { "epoch": 2.9616218560277536, "grad_norm": 0.7294409275054932, "learning_rate": 1.283911214275352e-06, "loss": 0.0194, "step": 27318 }, { "epoch": 2.9617302688638336, "grad_norm": 0.15775470435619354, "learning_rate": 1.2802843464384158e-06, "loss": 0.012, "step": 27319 }, { "epoch": 2.961838681699913, "grad_norm": 0.7162168622016907, "learning_rate": 1.27665747860148e-06, "loss": 0.0231, "step": 27320 }, { "epoch": 2.961947094535993, "grad_norm": 0.7105095982551575, "learning_rate": 1.2730306107645438e-06, "loss": 0.0336, "step": 27321 }, { "epoch": 2.9620555073720727, "grad_norm": 0.26824626326560974, "learning_rate": 1.2694037429276077e-06, "loss": 0.029, "step": 27322 }, { "epoch": 2.9621639202081527, "grad_norm": 0.07208257168531418, "learning_rate": 1.2657768750906718e-06, "loss": 0.0014, "step": 27323 }, { "epoch": 2.9622723330442327, "grad_norm": 0.024652952328324318, "learning_rate": 1.2621500072537357e-06, "loss": 0.0008, "step": 27324 }, { "epoch": 2.962380745880312, "grad_norm": 0.31446391344070435, "learning_rate": 1.2585231394167996e-06, "loss": 0.0065, "step": 27325 }, { "epoch": 2.962489158716392, "grad_norm": 0.3741227388381958, "learning_rate": 1.2548962715798637e-06, "loss": 0.0267, "step": 27326 }, { "epoch": 2.9625975715524717, "grad_norm": 0.020856380462646484, "learning_rate": 1.2512694037429276e-06, "loss": 0.0007, "step": 27327 }, { "epoch": 2.9627059843885517, "grad_norm": 0.1293351948261261, "learning_rate": 1.2476425359059917e-06, "loss": 0.0061, "step": 27328 }, { "epoch": 2.9628143972246312, "grad_norm": 1.578442931175232, "learning_rate": 1.2440156680690556e-06, "loss": 0.0249, "step": 27329 }, { "epoch": 2.9629228100607112, "grad_norm": 0.25154685974121094, "learning_rate": 1.2403888002321197e-06, "loss": 0.0034, "step": 27330 }, { "epoch": 2.9630312228967908, "grad_norm": 0.10017546266317368, "learning_rate": 1.2367619323951836e-06, "loss": 0.0026, "step": 27331 }, { "epoch": 2.9631396357328708, "grad_norm": 0.022011635825037956, "learning_rate": 1.2331350645582475e-06, "loss": 0.001, "step": 27332 }, { "epoch": 2.9632480485689507, "grad_norm": 0.5394918322563171, "learning_rate": 1.2295081967213116e-06, "loss": 0.0152, "step": 27333 }, { "epoch": 2.9633564614050303, "grad_norm": 0.1929914504289627, "learning_rate": 1.2258813288843755e-06, "loss": 0.0052, "step": 27334 }, { "epoch": 2.9634648742411103, "grad_norm": 0.15478289127349854, "learning_rate": 1.2222544610474396e-06, "loss": 0.0029, "step": 27335 }, { "epoch": 2.96357328707719, "grad_norm": 0.2670459747314453, "learning_rate": 1.2186275932105035e-06, "loss": 0.0283, "step": 27336 }, { "epoch": 2.96368169991327, "grad_norm": 0.06791310012340546, "learning_rate": 1.2150007253735674e-06, "loss": 0.0016, "step": 27337 }, { "epoch": 2.96379011274935, "grad_norm": 0.5798225402832031, "learning_rate": 1.2113738575366315e-06, "loss": 0.0181, "step": 27338 }, { "epoch": 2.9638985255854293, "grad_norm": 0.043265510350465775, "learning_rate": 1.2077469896996954e-06, "loss": 0.0011, "step": 27339 }, { "epoch": 2.964006938421509, "grad_norm": 0.19585321843624115, "learning_rate": 1.2041201218627595e-06, "loss": 0.0134, "step": 27340 }, { "epoch": 2.964115351257589, "grad_norm": 0.10811874270439148, "learning_rate": 1.2004932540258234e-06, "loss": 0.003, "step": 27341 }, { "epoch": 2.964223764093669, "grad_norm": 0.07078178226947784, "learning_rate": 1.1968663861888873e-06, "loss": 0.0026, "step": 27342 }, { "epoch": 2.9643321769297484, "grad_norm": 0.03456612303853035, "learning_rate": 1.1932395183519514e-06, "loss": 0.0022, "step": 27343 }, { "epoch": 2.9644405897658284, "grad_norm": 0.17935426533222198, "learning_rate": 1.1896126505150152e-06, "loss": 0.0066, "step": 27344 }, { "epoch": 2.964549002601908, "grad_norm": 1.2349064350128174, "learning_rate": 1.1859857826780794e-06, "loss": 0.0304, "step": 27345 }, { "epoch": 2.964657415437988, "grad_norm": 0.04999767988920212, "learning_rate": 1.1823589148411432e-06, "loss": 0.0021, "step": 27346 }, { "epoch": 2.964765828274068, "grad_norm": 0.09309349209070206, "learning_rate": 1.1787320470042071e-06, "loss": 0.0042, "step": 27347 }, { "epoch": 2.9648742411101474, "grad_norm": 0.020481763407588005, "learning_rate": 1.1751051791672712e-06, "loss": 0.0008, "step": 27348 }, { "epoch": 2.9649826539462274, "grad_norm": 0.031532805413007736, "learning_rate": 1.1714783113303351e-06, "loss": 0.0007, "step": 27349 }, { "epoch": 2.965091066782307, "grad_norm": 0.008131907321512699, "learning_rate": 1.167851443493399e-06, "loss": 0.0002, "step": 27350 }, { "epoch": 2.965199479618387, "grad_norm": 0.08944297581911087, "learning_rate": 1.1642245756564631e-06, "loss": 0.0012, "step": 27351 }, { "epoch": 2.9653078924544665, "grad_norm": 0.04473688080906868, "learning_rate": 1.1605977078195272e-06, "loss": 0.0016, "step": 27352 }, { "epoch": 2.9654163052905465, "grad_norm": 0.06310074031352997, "learning_rate": 1.1569708399825911e-06, "loss": 0.0011, "step": 27353 }, { "epoch": 2.965524718126626, "grad_norm": 0.04884692281484604, "learning_rate": 1.153343972145655e-06, "loss": 0.002, "step": 27354 }, { "epoch": 2.965633130962706, "grad_norm": 0.05784223973751068, "learning_rate": 1.1497171043087191e-06, "loss": 0.0034, "step": 27355 }, { "epoch": 2.965741543798786, "grad_norm": 0.8429482579231262, "learning_rate": 1.146090236471783e-06, "loss": 0.0209, "step": 27356 }, { "epoch": 2.9658499566348655, "grad_norm": 0.9124667048454285, "learning_rate": 1.142463368634847e-06, "loss": 0.036, "step": 27357 }, { "epoch": 2.9659583694709455, "grad_norm": 0.027579473331570625, "learning_rate": 1.138836500797911e-06, "loss": 0.0005, "step": 27358 }, { "epoch": 2.966066782307025, "grad_norm": 0.15532995760440826, "learning_rate": 1.135209632960975e-06, "loss": 0.0071, "step": 27359 }, { "epoch": 2.966175195143105, "grad_norm": 0.09269075840711594, "learning_rate": 1.1315827651240388e-06, "loss": 0.0017, "step": 27360 }, { "epoch": 2.966283607979185, "grad_norm": 0.20245154201984406, "learning_rate": 1.127955897287103e-06, "loss": 0.0241, "step": 27361 }, { "epoch": 2.9663920208152645, "grad_norm": 0.07960420846939087, "learning_rate": 1.124329029450167e-06, "loss": 0.0045, "step": 27362 }, { "epoch": 2.966500433651344, "grad_norm": 0.36624476313591003, "learning_rate": 1.120702161613231e-06, "loss": 0.0189, "step": 27363 }, { "epoch": 2.966608846487424, "grad_norm": 0.3634895086288452, "learning_rate": 1.1170752937762948e-06, "loss": 0.0211, "step": 27364 }, { "epoch": 2.966717259323504, "grad_norm": 1.4841054677963257, "learning_rate": 1.113448425939359e-06, "loss": 0.0133, "step": 27365 }, { "epoch": 2.9668256721595836, "grad_norm": 0.007956192828714848, "learning_rate": 1.1098215581024228e-06, "loss": 0.0004, "step": 27366 }, { "epoch": 2.9669340849956636, "grad_norm": 1.43861722946167, "learning_rate": 1.1061946902654867e-06, "loss": 0.0381, "step": 27367 }, { "epoch": 2.967042497831743, "grad_norm": 0.4073296785354614, "learning_rate": 1.1025678224285508e-06, "loss": 0.0272, "step": 27368 }, { "epoch": 2.967150910667823, "grad_norm": 0.21401949226856232, "learning_rate": 1.0989409545916147e-06, "loss": 0.0049, "step": 27369 }, { "epoch": 2.967259323503903, "grad_norm": 0.003880556672811508, "learning_rate": 1.0953140867546786e-06, "loss": 0.0001, "step": 27370 }, { "epoch": 2.9673677363399826, "grad_norm": 0.7682885527610779, "learning_rate": 1.0916872189177427e-06, "loss": 0.0098, "step": 27371 }, { "epoch": 2.967476149176062, "grad_norm": 0.2967657148838043, "learning_rate": 1.0880603510808068e-06, "loss": 0.0072, "step": 27372 }, { "epoch": 2.967584562012142, "grad_norm": 0.5818121433258057, "learning_rate": 1.0844334832438705e-06, "loss": 0.0155, "step": 27373 }, { "epoch": 2.967692974848222, "grad_norm": 0.07402730733156204, "learning_rate": 1.0808066154069346e-06, "loss": 0.0028, "step": 27374 }, { "epoch": 2.9678013876843017, "grad_norm": 0.9666144251823425, "learning_rate": 1.0771797475699987e-06, "loss": 0.0574, "step": 27375 }, { "epoch": 2.9679098005203817, "grad_norm": 0.3321648836135864, "learning_rate": 1.0735528797330626e-06, "loss": 0.0045, "step": 27376 }, { "epoch": 2.968018213356461, "grad_norm": 0.0400611087679863, "learning_rate": 1.0699260118961265e-06, "loss": 0.0009, "step": 27377 }, { "epoch": 2.968126626192541, "grad_norm": 0.07679173350334167, "learning_rate": 1.0662991440591906e-06, "loss": 0.006, "step": 27378 }, { "epoch": 2.968235039028621, "grad_norm": 0.0070570469833910465, "learning_rate": 1.0626722762222547e-06, "loss": 0.0004, "step": 27379 }, { "epoch": 2.9683434518647007, "grad_norm": 0.18407253921031952, "learning_rate": 1.0590454083853184e-06, "loss": 0.0075, "step": 27380 }, { "epoch": 2.9684518647007807, "grad_norm": 0.44521796703338623, "learning_rate": 1.0554185405483825e-06, "loss": 0.0209, "step": 27381 }, { "epoch": 2.9685602775368602, "grad_norm": 0.1229880228638649, "learning_rate": 1.0517916727114466e-06, "loss": 0.0117, "step": 27382 }, { "epoch": 2.9686686903729402, "grad_norm": 0.12495172768831253, "learning_rate": 1.0481648048745103e-06, "loss": 0.0042, "step": 27383 }, { "epoch": 2.96877710320902, "grad_norm": 0.4601232409477234, "learning_rate": 1.0445379370375744e-06, "loss": 0.0157, "step": 27384 }, { "epoch": 2.9688855160450998, "grad_norm": 0.7516085505485535, "learning_rate": 1.0409110692006385e-06, "loss": 0.0327, "step": 27385 }, { "epoch": 2.9689939288811793, "grad_norm": 0.1073184609413147, "learning_rate": 1.0372842013637024e-06, "loss": 0.0046, "step": 27386 }, { "epoch": 2.9691023417172593, "grad_norm": 0.40961557626724243, "learning_rate": 1.0336573335267663e-06, "loss": 0.0065, "step": 27387 }, { "epoch": 2.9692107545533393, "grad_norm": 0.15571972727775574, "learning_rate": 1.0300304656898304e-06, "loss": 0.0041, "step": 27388 }, { "epoch": 2.969319167389419, "grad_norm": 0.61713707447052, "learning_rate": 1.0264035978528943e-06, "loss": 0.0252, "step": 27389 }, { "epoch": 2.969427580225499, "grad_norm": 0.4472348392009735, "learning_rate": 1.0227767300159582e-06, "loss": 0.0354, "step": 27390 }, { "epoch": 2.9695359930615783, "grad_norm": 0.09206485003232956, "learning_rate": 1.0191498621790223e-06, "loss": 0.003, "step": 27391 }, { "epoch": 2.9696444058976583, "grad_norm": 0.5865218043327332, "learning_rate": 1.0155229943420864e-06, "loss": 0.032, "step": 27392 }, { "epoch": 2.9697528187337383, "grad_norm": 0.8192828297615051, "learning_rate": 1.01189612650515e-06, "loss": 0.0192, "step": 27393 }, { "epoch": 2.969861231569818, "grad_norm": 0.4618258774280548, "learning_rate": 1.0082692586682142e-06, "loss": 0.0076, "step": 27394 }, { "epoch": 2.9699696444058974, "grad_norm": 0.2800365388393402, "learning_rate": 1.0046423908312783e-06, "loss": 0.0109, "step": 27395 }, { "epoch": 2.9700780572419774, "grad_norm": 0.4669666588306427, "learning_rate": 1.0010155229943421e-06, "loss": 0.0173, "step": 27396 }, { "epoch": 2.9701864700780574, "grad_norm": 0.9763092994689941, "learning_rate": 9.97388655157406e-07, "loss": 0.0467, "step": 27397 }, { "epoch": 2.970294882914137, "grad_norm": 0.17145784199237823, "learning_rate": 9.937617873204701e-07, "loss": 0.0016, "step": 27398 }, { "epoch": 2.970403295750217, "grad_norm": 0.3364754319190979, "learning_rate": 9.90134919483534e-07, "loss": 0.0201, "step": 27399 }, { "epoch": 2.9705117085862964, "grad_norm": 0.4710831344127655, "learning_rate": 9.86508051646598e-07, "loss": 0.0301, "step": 27400 }, { "epoch": 2.9706201214223764, "grad_norm": 0.5742149353027344, "learning_rate": 9.82881183809662e-07, "loss": 0.008, "step": 27401 }, { "epoch": 2.9707285342584564, "grad_norm": 0.057288069278001785, "learning_rate": 9.792543159727261e-07, "loss": 0.0014, "step": 27402 }, { "epoch": 2.970836947094536, "grad_norm": 0.15221649408340454, "learning_rate": 9.756274481357898e-07, "loss": 0.0088, "step": 27403 }, { "epoch": 2.970945359930616, "grad_norm": 0.20503664016723633, "learning_rate": 9.72000580298854e-07, "loss": 0.0085, "step": 27404 }, { "epoch": 2.9710537727666955, "grad_norm": 0.07545040547847748, "learning_rate": 9.68373712461918e-07, "loss": 0.0031, "step": 27405 }, { "epoch": 2.9711621856027755, "grad_norm": 0.09462549537420273, "learning_rate": 9.64746844624982e-07, "loss": 0.0021, "step": 27406 }, { "epoch": 2.9712705984388554, "grad_norm": 1.246430516242981, "learning_rate": 9.611199767880458e-07, "loss": 0.0385, "step": 27407 }, { "epoch": 2.971379011274935, "grad_norm": 0.13709549605846405, "learning_rate": 9.5749310895111e-07, "loss": 0.0035, "step": 27408 }, { "epoch": 2.9714874241110145, "grad_norm": 0.05620526894927025, "learning_rate": 9.538662411141738e-07, "loss": 0.0024, "step": 27409 }, { "epoch": 2.9715958369470945, "grad_norm": 0.34010303020477295, "learning_rate": 9.502393732772377e-07, "loss": 0.0073, "step": 27410 }, { "epoch": 2.9717042497831745, "grad_norm": 0.31446143984794617, "learning_rate": 9.466125054403018e-07, "loss": 0.0121, "step": 27411 }, { "epoch": 2.971812662619254, "grad_norm": 0.22201742231845856, "learning_rate": 9.429856376033658e-07, "loss": 0.0134, "step": 27412 }, { "epoch": 2.971921075455334, "grad_norm": 0.12107221782207489, "learning_rate": 9.393587697664298e-07, "loss": 0.0026, "step": 27413 }, { "epoch": 2.9720294882914136, "grad_norm": 0.27341657876968384, "learning_rate": 9.357319019294937e-07, "loss": 0.0113, "step": 27414 }, { "epoch": 2.9721379011274935, "grad_norm": 0.56048983335495, "learning_rate": 9.321050340925577e-07, "loss": 0.0139, "step": 27415 }, { "epoch": 2.9722463139635735, "grad_norm": 0.4051955044269562, "learning_rate": 9.284781662556217e-07, "loss": 0.0227, "step": 27416 }, { "epoch": 2.972354726799653, "grad_norm": 0.6241340041160583, "learning_rate": 9.248512984186856e-07, "loss": 0.0142, "step": 27417 }, { "epoch": 2.9724631396357326, "grad_norm": 0.1031327024102211, "learning_rate": 9.212244305817496e-07, "loss": 0.0047, "step": 27418 }, { "epoch": 2.9725715524718126, "grad_norm": 0.1108965277671814, "learning_rate": 9.175975627448137e-07, "loss": 0.0052, "step": 27419 }, { "epoch": 2.9726799653078926, "grad_norm": 0.15438152849674225, "learning_rate": 9.139706949078775e-07, "loss": 0.0057, "step": 27420 }, { "epoch": 2.972788378143972, "grad_norm": 0.08255418390035629, "learning_rate": 9.103438270709416e-07, "loss": 0.0027, "step": 27421 }, { "epoch": 2.972896790980052, "grad_norm": 0.012861003167927265, "learning_rate": 9.067169592340056e-07, "loss": 0.0004, "step": 27422 }, { "epoch": 2.9730052038161316, "grad_norm": 0.2910100817680359, "learning_rate": 9.030900913970696e-07, "loss": 0.0085, "step": 27423 }, { "epoch": 2.9731136166522116, "grad_norm": 0.17082032561302185, "learning_rate": 8.994632235601335e-07, "loss": 0.0048, "step": 27424 }, { "epoch": 2.9732220294882916, "grad_norm": 0.3874531686306, "learning_rate": 8.958363557231975e-07, "loss": 0.0141, "step": 27425 }, { "epoch": 2.973330442324371, "grad_norm": 0.46352237462997437, "learning_rate": 8.922094878862615e-07, "loss": 0.0034, "step": 27426 }, { "epoch": 2.973438855160451, "grad_norm": 0.04038894176483154, "learning_rate": 8.885826200493254e-07, "loss": 0.0015, "step": 27427 }, { "epoch": 2.9735472679965307, "grad_norm": 0.21701262891292572, "learning_rate": 8.849557522123894e-07, "loss": 0.0085, "step": 27428 }, { "epoch": 2.9736556808326107, "grad_norm": 0.06863825023174286, "learning_rate": 8.813288843754535e-07, "loss": 0.0015, "step": 27429 }, { "epoch": 2.97376409366869, "grad_norm": 0.33542582392692566, "learning_rate": 8.777020165385173e-07, "loss": 0.0091, "step": 27430 }, { "epoch": 2.97387250650477, "grad_norm": 0.38644301891326904, "learning_rate": 8.740751487015813e-07, "loss": 0.0146, "step": 27431 }, { "epoch": 2.9739809193408497, "grad_norm": 0.579369306564331, "learning_rate": 8.704482808646454e-07, "loss": 0.0115, "step": 27432 }, { "epoch": 2.9740893321769297, "grad_norm": 0.22495141625404358, "learning_rate": 8.668214130277094e-07, "loss": 0.0057, "step": 27433 }, { "epoch": 2.9741977450130097, "grad_norm": 0.0585058256983757, "learning_rate": 8.631945451907733e-07, "loss": 0.0021, "step": 27434 }, { "epoch": 2.9743061578490892, "grad_norm": 0.8332356214523315, "learning_rate": 8.595676773538373e-07, "loss": 0.0183, "step": 27435 }, { "epoch": 2.9744145706851692, "grad_norm": 0.03779049590229988, "learning_rate": 8.559408095169013e-07, "loss": 0.0011, "step": 27436 }, { "epoch": 2.9745229835212488, "grad_norm": 1.2397115230560303, "learning_rate": 8.523139416799652e-07, "loss": 0.032, "step": 27437 }, { "epoch": 2.9746313963573288, "grad_norm": 0.011872000060975552, "learning_rate": 8.486870738430292e-07, "loss": 0.0007, "step": 27438 }, { "epoch": 2.9747398091934087, "grad_norm": 0.020363755524158478, "learning_rate": 8.450602060060932e-07, "loss": 0.0006, "step": 27439 }, { "epoch": 2.9748482220294883, "grad_norm": 0.10695764422416687, "learning_rate": 8.414333381691573e-07, "loss": 0.0025, "step": 27440 }, { "epoch": 2.974956634865568, "grad_norm": 0.14873266220092773, "learning_rate": 8.378064703322211e-07, "loss": 0.0037, "step": 27441 }, { "epoch": 2.975065047701648, "grad_norm": 0.006959856487810612, "learning_rate": 8.341796024952852e-07, "loss": 0.0002, "step": 27442 }, { "epoch": 2.975173460537728, "grad_norm": 0.1353529840707779, "learning_rate": 8.305527346583492e-07, "loss": 0.0089, "step": 27443 }, { "epoch": 2.9752818733738073, "grad_norm": 0.03029697760939598, "learning_rate": 8.26925866821413e-07, "loss": 0.0006, "step": 27444 }, { "epoch": 2.9753902862098873, "grad_norm": 0.1748109757900238, "learning_rate": 8.23298998984477e-07, "loss": 0.0063, "step": 27445 }, { "epoch": 2.975498699045967, "grad_norm": 0.26738694310188293, "learning_rate": 8.19672131147541e-07, "loss": 0.0102, "step": 27446 }, { "epoch": 2.975607111882047, "grad_norm": 0.6424230933189392, "learning_rate": 8.160452633106049e-07, "loss": 0.0294, "step": 27447 }, { "epoch": 2.975715524718127, "grad_norm": 0.35697969794273376, "learning_rate": 8.124183954736689e-07, "loss": 0.0455, "step": 27448 }, { "epoch": 2.9758239375542064, "grad_norm": 0.04452721029520035, "learning_rate": 8.087915276367329e-07, "loss": 0.0011, "step": 27449 }, { "epoch": 2.9759323503902864, "grad_norm": 0.028823239728808403, "learning_rate": 8.05164659799797e-07, "loss": 0.0007, "step": 27450 }, { "epoch": 2.976040763226366, "grad_norm": 0.6490762829780579, "learning_rate": 8.015377919628608e-07, "loss": 0.0194, "step": 27451 }, { "epoch": 2.976149176062446, "grad_norm": 0.4141819477081299, "learning_rate": 7.979109241259249e-07, "loss": 0.0163, "step": 27452 }, { "epoch": 2.9762575888985254, "grad_norm": 0.38165852427482605, "learning_rate": 7.942840562889889e-07, "loss": 0.0031, "step": 27453 }, { "epoch": 2.9763660017346054, "grad_norm": 0.2734130322933197, "learning_rate": 7.906571884520527e-07, "loss": 0.0283, "step": 27454 }, { "epoch": 2.976474414570685, "grad_norm": 0.2932601571083069, "learning_rate": 7.870303206151168e-07, "loss": 0.0048, "step": 27455 }, { "epoch": 2.976582827406765, "grad_norm": 0.02993524633347988, "learning_rate": 7.834034527781808e-07, "loss": 0.0012, "step": 27456 }, { "epoch": 2.976691240242845, "grad_norm": 0.624642550945282, "learning_rate": 7.797765849412448e-07, "loss": 0.0143, "step": 27457 }, { "epoch": 2.9767996530789245, "grad_norm": 0.17151248455047607, "learning_rate": 7.761497171043087e-07, "loss": 0.011, "step": 27458 }, { "epoch": 2.9769080659150045, "grad_norm": 0.610086977481842, "learning_rate": 7.725228492673727e-07, "loss": 0.0176, "step": 27459 }, { "epoch": 2.977016478751084, "grad_norm": 0.2768871784210205, "learning_rate": 7.688959814304367e-07, "loss": 0.0045, "step": 27460 }, { "epoch": 2.977124891587164, "grad_norm": 0.648600161075592, "learning_rate": 7.652691135935007e-07, "loss": 0.0241, "step": 27461 }, { "epoch": 2.977233304423244, "grad_norm": 0.010738784447312355, "learning_rate": 7.616422457565646e-07, "loss": 0.0005, "step": 27462 }, { "epoch": 2.9773417172593235, "grad_norm": 0.4640492796897888, "learning_rate": 7.580153779196286e-07, "loss": 0.0145, "step": 27463 }, { "epoch": 2.977450130095403, "grad_norm": 0.1256556659936905, "learning_rate": 7.543885100826926e-07, "loss": 0.003, "step": 27464 }, { "epoch": 2.977558542931483, "grad_norm": 1.2585132122039795, "learning_rate": 7.507616422457566e-07, "loss": 0.0586, "step": 27465 }, { "epoch": 2.977666955767563, "grad_norm": 0.21739041805267334, "learning_rate": 7.471347744088206e-07, "loss": 0.0073, "step": 27466 }, { "epoch": 2.9777753686036426, "grad_norm": 0.06639911234378815, "learning_rate": 7.435079065718845e-07, "loss": 0.002, "step": 27467 }, { "epoch": 2.9778837814397225, "grad_norm": 0.10394369065761566, "learning_rate": 7.398810387349486e-07, "loss": 0.0055, "step": 27468 }, { "epoch": 2.977992194275802, "grad_norm": 0.35403501987457275, "learning_rate": 7.362541708980125e-07, "loss": 0.0057, "step": 27469 }, { "epoch": 2.978100607111882, "grad_norm": 0.11486223340034485, "learning_rate": 7.326273030610765e-07, "loss": 0.0025, "step": 27470 }, { "epoch": 2.978209019947962, "grad_norm": 0.2471044957637787, "learning_rate": 7.290004352241405e-07, "loss": 0.0085, "step": 27471 }, { "epoch": 2.9783174327840416, "grad_norm": 0.44429683685302734, "learning_rate": 7.253735673872044e-07, "loss": 0.0217, "step": 27472 }, { "epoch": 2.9784258456201216, "grad_norm": 0.5244717001914978, "learning_rate": 7.217466995502685e-07, "loss": 0.0154, "step": 27473 }, { "epoch": 2.978534258456201, "grad_norm": 0.036735497415065765, "learning_rate": 7.181198317133324e-07, "loss": 0.0012, "step": 27474 }, { "epoch": 2.978642671292281, "grad_norm": 0.06224239245057106, "learning_rate": 7.144929638763964e-07, "loss": 0.0026, "step": 27475 }, { "epoch": 2.9787510841283606, "grad_norm": 0.3303505778312683, "learning_rate": 7.108660960394604e-07, "loss": 0.0111, "step": 27476 }, { "epoch": 2.9788594969644406, "grad_norm": 0.5166433453559875, "learning_rate": 7.072392282025243e-07, "loss": 0.0039, "step": 27477 }, { "epoch": 2.97896790980052, "grad_norm": 0.23667098581790924, "learning_rate": 7.036123603655884e-07, "loss": 0.0064, "step": 27478 }, { "epoch": 2.9790763226366, "grad_norm": 0.13564087450504303, "learning_rate": 6.999854925286523e-07, "loss": 0.0026, "step": 27479 }, { "epoch": 2.97918473547268, "grad_norm": 0.09405141323804855, "learning_rate": 6.963586246917163e-07, "loss": 0.0025, "step": 27480 }, { "epoch": 2.9792931483087597, "grad_norm": 0.1682424694299698, "learning_rate": 6.927317568547803e-07, "loss": 0.0057, "step": 27481 }, { "epoch": 2.9794015611448397, "grad_norm": 0.06586576253175735, "learning_rate": 6.891048890178442e-07, "loss": 0.0027, "step": 27482 }, { "epoch": 2.979509973980919, "grad_norm": 0.2299811691045761, "learning_rate": 6.854780211809083e-07, "loss": 0.0163, "step": 27483 }, { "epoch": 2.979618386816999, "grad_norm": 0.4601118266582489, "learning_rate": 6.818511533439722e-07, "loss": 0.0181, "step": 27484 }, { "epoch": 2.979726799653079, "grad_norm": 0.429184228181839, "learning_rate": 6.782242855070361e-07, "loss": 0.0519, "step": 27485 }, { "epoch": 2.9798352124891587, "grad_norm": 0.31110700964927673, "learning_rate": 6.745974176701002e-07, "loss": 0.0216, "step": 27486 }, { "epoch": 2.9799436253252383, "grad_norm": 0.06950041651725769, "learning_rate": 6.709705498331641e-07, "loss": 0.0028, "step": 27487 }, { "epoch": 2.9800520381613183, "grad_norm": 0.9086229205131531, "learning_rate": 6.673436819962281e-07, "loss": 0.068, "step": 27488 }, { "epoch": 2.9801604509973982, "grad_norm": 0.07664806395769119, "learning_rate": 6.637168141592921e-07, "loss": 0.0045, "step": 27489 }, { "epoch": 2.9802688638334778, "grad_norm": 0.19596083462238312, "learning_rate": 6.600899463223561e-07, "loss": 0.0176, "step": 27490 }, { "epoch": 2.9803772766695578, "grad_norm": 0.42681884765625, "learning_rate": 6.564630784854201e-07, "loss": 0.0073, "step": 27491 }, { "epoch": 2.9804856895056373, "grad_norm": 0.46076804399490356, "learning_rate": 6.52836210648484e-07, "loss": 0.015, "step": 27492 }, { "epoch": 2.9805941023417173, "grad_norm": 0.14714030921459198, "learning_rate": 6.49209342811548e-07, "loss": 0.0073, "step": 27493 }, { "epoch": 2.9807025151777973, "grad_norm": 0.286894828081131, "learning_rate": 6.45582474974612e-07, "loss": 0.0148, "step": 27494 }, { "epoch": 2.980810928013877, "grad_norm": 0.008170685730874538, "learning_rate": 6.41955607137676e-07, "loss": 0.0004, "step": 27495 }, { "epoch": 2.980919340849957, "grad_norm": 0.3946040868759155, "learning_rate": 6.3832873930074e-07, "loss": 0.0225, "step": 27496 }, { "epoch": 2.9810277536860363, "grad_norm": 0.15954498946666718, "learning_rate": 6.347018714638038e-07, "loss": 0.0023, "step": 27497 }, { "epoch": 2.9811361665221163, "grad_norm": 0.28237298130989075, "learning_rate": 6.310750036268678e-07, "loss": 0.0238, "step": 27498 }, { "epoch": 2.981244579358196, "grad_norm": 0.03315046429634094, "learning_rate": 6.274481357899318e-07, "loss": 0.0014, "step": 27499 }, { "epoch": 2.981352992194276, "grad_norm": 0.17799833416938782, "learning_rate": 6.238212679529958e-07, "loss": 0.0059, "step": 27500 }, { "epoch": 2.9814614050303554, "grad_norm": 0.7039925456047058, "learning_rate": 6.201944001160598e-07, "loss": 0.0426, "step": 27501 }, { "epoch": 2.9815698178664354, "grad_norm": 0.015571041032671928, "learning_rate": 6.165675322791237e-07, "loss": 0.0004, "step": 27502 }, { "epoch": 2.9816782307025154, "grad_norm": 0.23461690545082092, "learning_rate": 6.129406644421877e-07, "loss": 0.0131, "step": 27503 }, { "epoch": 2.981786643538595, "grad_norm": 0.15187840163707733, "learning_rate": 6.093137966052517e-07, "loss": 0.0099, "step": 27504 }, { "epoch": 2.981895056374675, "grad_norm": 0.9624627232551575, "learning_rate": 6.056869287683157e-07, "loss": 0.033, "step": 27505 }, { "epoch": 2.9820034692107544, "grad_norm": 0.29973024129867554, "learning_rate": 6.020600609313797e-07, "loss": 0.0113, "step": 27506 }, { "epoch": 2.9821118820468344, "grad_norm": 0.3360414206981659, "learning_rate": 5.984331930944436e-07, "loss": 0.0107, "step": 27507 }, { "epoch": 2.9822202948829144, "grad_norm": 0.01010508555918932, "learning_rate": 5.948063252575076e-07, "loss": 0.0004, "step": 27508 }, { "epoch": 2.982328707718994, "grad_norm": 0.5914275646209717, "learning_rate": 5.911794574205716e-07, "loss": 0.0266, "step": 27509 }, { "epoch": 2.9824371205550735, "grad_norm": 0.7058467864990234, "learning_rate": 5.875525895836356e-07, "loss": 0.0287, "step": 27510 }, { "epoch": 2.9825455333911535, "grad_norm": 0.6272668242454529, "learning_rate": 5.839257217466995e-07, "loss": 0.0085, "step": 27511 }, { "epoch": 2.9826539462272335, "grad_norm": 1.1080498695373535, "learning_rate": 5.802988539097636e-07, "loss": 0.0199, "step": 27512 }, { "epoch": 2.982762359063313, "grad_norm": 0.35537365078926086, "learning_rate": 5.766719860728275e-07, "loss": 0.0127, "step": 27513 }, { "epoch": 2.982870771899393, "grad_norm": 0.06148688495159149, "learning_rate": 5.730451182358915e-07, "loss": 0.0016, "step": 27514 }, { "epoch": 2.9829791847354725, "grad_norm": 0.7169224619865417, "learning_rate": 5.694182503989555e-07, "loss": 0.0136, "step": 27515 }, { "epoch": 2.9830875975715525, "grad_norm": 0.3297152817249298, "learning_rate": 5.657913825620194e-07, "loss": 0.0416, "step": 27516 }, { "epoch": 2.9831960104076325, "grad_norm": 0.34356793761253357, "learning_rate": 5.621645147250835e-07, "loss": 0.017, "step": 27517 }, { "epoch": 2.983304423243712, "grad_norm": 0.08278651535511017, "learning_rate": 5.585376468881474e-07, "loss": 0.0029, "step": 27518 }, { "epoch": 2.9834128360797916, "grad_norm": 0.5562822222709656, "learning_rate": 5.549107790512114e-07, "loss": 0.0129, "step": 27519 }, { "epoch": 2.9835212489158716, "grad_norm": 0.02035330794751644, "learning_rate": 5.512839112142754e-07, "loss": 0.0007, "step": 27520 }, { "epoch": 2.9836296617519515, "grad_norm": 0.1483653485774994, "learning_rate": 5.476570433773393e-07, "loss": 0.0054, "step": 27521 }, { "epoch": 2.983738074588031, "grad_norm": 0.04959665983915329, "learning_rate": 5.440301755404034e-07, "loss": 0.0022, "step": 27522 }, { "epoch": 2.983846487424111, "grad_norm": 0.5873525142669678, "learning_rate": 5.404033077034673e-07, "loss": 0.0283, "step": 27523 }, { "epoch": 2.9839549002601906, "grad_norm": 0.0021806678269058466, "learning_rate": 5.367764398665313e-07, "loss": 0.0001, "step": 27524 }, { "epoch": 2.9840633130962706, "grad_norm": 0.12561443448066711, "learning_rate": 5.331495720295953e-07, "loss": 0.0034, "step": 27525 }, { "epoch": 2.9841717259323506, "grad_norm": 0.09751656651496887, "learning_rate": 5.295227041926592e-07, "loss": 0.0038, "step": 27526 }, { "epoch": 2.98428013876843, "grad_norm": 0.16861754655838013, "learning_rate": 5.258958363557233e-07, "loss": 0.0082, "step": 27527 }, { "epoch": 2.98438855160451, "grad_norm": 1.021716594696045, "learning_rate": 5.222689685187872e-07, "loss": 0.0351, "step": 27528 }, { "epoch": 2.9844969644405897, "grad_norm": 0.05736188590526581, "learning_rate": 5.186421006818512e-07, "loss": 0.0014, "step": 27529 }, { "epoch": 2.9846053772766696, "grad_norm": 1.9935749769210815, "learning_rate": 5.150152328449152e-07, "loss": 0.0202, "step": 27530 }, { "epoch": 2.9847137901127496, "grad_norm": 0.4683420956134796, "learning_rate": 5.113883650079791e-07, "loss": 0.0161, "step": 27531 }, { "epoch": 2.984822202948829, "grad_norm": 1.2773622274398804, "learning_rate": 5.077614971710432e-07, "loss": 0.0688, "step": 27532 }, { "epoch": 2.9849306157849087, "grad_norm": 0.44720661640167236, "learning_rate": 5.041346293341071e-07, "loss": 0.0247, "step": 27533 }, { "epoch": 2.9850390286209887, "grad_norm": 0.1748523712158203, "learning_rate": 5.005077614971711e-07, "loss": 0.0115, "step": 27534 }, { "epoch": 2.9851474414570687, "grad_norm": 0.37546780705451965, "learning_rate": 4.968808936602351e-07, "loss": 0.0051, "step": 27535 }, { "epoch": 2.985255854293148, "grad_norm": 0.028543172404170036, "learning_rate": 4.93254025823299e-07, "loss": 0.0006, "step": 27536 }, { "epoch": 2.985364267129228, "grad_norm": 0.6413180232048035, "learning_rate": 4.896271579863631e-07, "loss": 0.0133, "step": 27537 }, { "epoch": 2.9854726799653077, "grad_norm": 0.3713192641735077, "learning_rate": 4.86000290149427e-07, "loss": 0.0458, "step": 27538 }, { "epoch": 2.9855810928013877, "grad_norm": 1.1620351076126099, "learning_rate": 4.82373422312491e-07, "loss": 0.0367, "step": 27539 }, { "epoch": 2.9856895056374677, "grad_norm": 0.6935722827911377, "learning_rate": 4.78746554475555e-07, "loss": 0.0281, "step": 27540 }, { "epoch": 2.9857979184735473, "grad_norm": 0.3151262700557709, "learning_rate": 4.7511968663861886e-07, "loss": 0.0176, "step": 27541 }, { "epoch": 2.985906331309627, "grad_norm": 1.1338361501693726, "learning_rate": 4.714928188016829e-07, "loss": 0.0336, "step": 27542 }, { "epoch": 2.986014744145707, "grad_norm": 0.03587858006358147, "learning_rate": 4.6786595096474686e-07, "loss": 0.0014, "step": 27543 }, { "epoch": 2.9861231569817868, "grad_norm": 0.015606004744768143, "learning_rate": 4.6423908312781085e-07, "loss": 0.0007, "step": 27544 }, { "epoch": 2.9862315698178663, "grad_norm": 1.0434123277664185, "learning_rate": 4.606122152908748e-07, "loss": 0.0358, "step": 27545 }, { "epoch": 2.9863399826539463, "grad_norm": 0.7403131127357483, "learning_rate": 4.5698534745393875e-07, "loss": 0.0205, "step": 27546 }, { "epoch": 2.986448395490026, "grad_norm": 0.5530480146408081, "learning_rate": 4.533584796170028e-07, "loss": 0.019, "step": 27547 }, { "epoch": 2.986556808326106, "grad_norm": 0.6266639232635498, "learning_rate": 4.4973161178006675e-07, "loss": 0.0301, "step": 27548 }, { "epoch": 2.986665221162186, "grad_norm": 0.17189836502075195, "learning_rate": 4.4610474394313075e-07, "loss": 0.0087, "step": 27549 }, { "epoch": 2.9867736339982653, "grad_norm": 0.0642176941037178, "learning_rate": 4.424778761061947e-07, "loss": 0.005, "step": 27550 }, { "epoch": 2.9868820468343453, "grad_norm": 0.2996918857097626, "learning_rate": 4.3885100826925864e-07, "loss": 0.0128, "step": 27551 }, { "epoch": 2.986990459670425, "grad_norm": 0.033483803272247314, "learning_rate": 4.352241404323227e-07, "loss": 0.0008, "step": 27552 }, { "epoch": 2.987098872506505, "grad_norm": 0.03988485038280487, "learning_rate": 4.3159727259538664e-07, "loss": 0.0008, "step": 27553 }, { "epoch": 2.987207285342585, "grad_norm": 0.0461481511592865, "learning_rate": 4.2797040475845064e-07, "loss": 0.0012, "step": 27554 }, { "epoch": 2.9873156981786644, "grad_norm": 0.40110644698143005, "learning_rate": 4.243435369215146e-07, "loss": 0.0202, "step": 27555 }, { "epoch": 2.987424111014744, "grad_norm": 0.16838990151882172, "learning_rate": 4.2071666908457863e-07, "loss": 0.0165, "step": 27556 }, { "epoch": 2.987532523850824, "grad_norm": 0.0634392648935318, "learning_rate": 4.170898012476426e-07, "loss": 0.001, "step": 27557 }, { "epoch": 2.987640936686904, "grad_norm": 0.048763707280159, "learning_rate": 4.134629334107065e-07, "loss": 0.0025, "step": 27558 }, { "epoch": 2.9877493495229834, "grad_norm": 0.30282139778137207, "learning_rate": 4.098360655737705e-07, "loss": 0.0091, "step": 27559 }, { "epoch": 2.9878577623590634, "grad_norm": 0.09607642889022827, "learning_rate": 4.0620919773683447e-07, "loss": 0.0018, "step": 27560 }, { "epoch": 2.987966175195143, "grad_norm": 0.1481587141752243, "learning_rate": 4.025823298998985e-07, "loss": 0.001, "step": 27561 }, { "epoch": 2.988074588031223, "grad_norm": 0.027628827840089798, "learning_rate": 3.9895546206296247e-07, "loss": 0.0004, "step": 27562 }, { "epoch": 2.988183000867303, "grad_norm": 0.07790400087833405, "learning_rate": 3.9532859422602636e-07, "loss": 0.0016, "step": 27563 }, { "epoch": 2.9882914137033825, "grad_norm": 0.1230144202709198, "learning_rate": 3.917017263890904e-07, "loss": 0.0104, "step": 27564 }, { "epoch": 2.988399826539462, "grad_norm": 0.533863365650177, "learning_rate": 3.8807485855215436e-07, "loss": 0.0161, "step": 27565 }, { "epoch": 2.988508239375542, "grad_norm": 0.10329627990722656, "learning_rate": 3.8444799071521836e-07, "loss": 0.0032, "step": 27566 }, { "epoch": 2.988616652211622, "grad_norm": 0.6317532658576965, "learning_rate": 3.808211228782823e-07, "loss": 0.0429, "step": 27567 }, { "epoch": 2.9887250650477015, "grad_norm": 0.15409813821315765, "learning_rate": 3.771942550413463e-07, "loss": 0.0034, "step": 27568 }, { "epoch": 2.9888334778837815, "grad_norm": 0.017513083294034004, "learning_rate": 3.735673872044103e-07, "loss": 0.0005, "step": 27569 }, { "epoch": 2.988941890719861, "grad_norm": 0.203131303191185, "learning_rate": 3.699405193674743e-07, "loss": 0.0051, "step": 27570 }, { "epoch": 2.989050303555941, "grad_norm": 0.007399025373160839, "learning_rate": 3.6631365153053825e-07, "loss": 0.0002, "step": 27571 }, { "epoch": 2.989158716392021, "grad_norm": 0.02060064487159252, "learning_rate": 3.626867836936022e-07, "loss": 0.0013, "step": 27572 }, { "epoch": 2.9892671292281006, "grad_norm": 0.43553051352500916, "learning_rate": 3.590599158566662e-07, "loss": 0.0197, "step": 27573 }, { "epoch": 2.9893755420641805, "grad_norm": 0.08346153050661087, "learning_rate": 3.554330480197302e-07, "loss": 0.0056, "step": 27574 }, { "epoch": 2.98948395490026, "grad_norm": 0.8903997540473938, "learning_rate": 3.518061801827942e-07, "loss": 0.0193, "step": 27575 }, { "epoch": 2.98959236773634, "grad_norm": 0.2956666350364685, "learning_rate": 3.4817931234585814e-07, "loss": 0.0064, "step": 27576 }, { "epoch": 2.9897007805724196, "grad_norm": 0.06717220693826675, "learning_rate": 3.445524445089221e-07, "loss": 0.0026, "step": 27577 }, { "epoch": 2.9898091934084996, "grad_norm": 0.009800978004932404, "learning_rate": 3.409255766719861e-07, "loss": 0.0004, "step": 27578 }, { "epoch": 2.989917606244579, "grad_norm": 0.5940839052200317, "learning_rate": 3.372987088350501e-07, "loss": 0.0153, "step": 27579 }, { "epoch": 2.990026019080659, "grad_norm": 0.4005361497402191, "learning_rate": 3.3367184099811403e-07, "loss": 0.0231, "step": 27580 }, { "epoch": 2.990134431916739, "grad_norm": 0.09844835847616196, "learning_rate": 3.3004497316117803e-07, "loss": 0.0046, "step": 27581 }, { "epoch": 2.9902428447528187, "grad_norm": 0.05583963543176651, "learning_rate": 3.26418105324242e-07, "loss": 0.0021, "step": 27582 }, { "epoch": 2.9903512575888986, "grad_norm": 0.07771594077348709, "learning_rate": 3.22791237487306e-07, "loss": 0.0034, "step": 27583 }, { "epoch": 2.990459670424978, "grad_norm": 0.058548811823129654, "learning_rate": 3.1916436965037e-07, "loss": 0.0037, "step": 27584 }, { "epoch": 2.990568083261058, "grad_norm": 0.16886182129383087, "learning_rate": 3.155375018134339e-07, "loss": 0.0046, "step": 27585 }, { "epoch": 2.990676496097138, "grad_norm": 0.0983131006360054, "learning_rate": 3.119106339764979e-07, "loss": 0.0042, "step": 27586 }, { "epoch": 2.9907849089332177, "grad_norm": 0.22632381319999695, "learning_rate": 3.0828376613956187e-07, "loss": 0.0072, "step": 27587 }, { "epoch": 2.9908933217692972, "grad_norm": 0.6589479446411133, "learning_rate": 3.0465689830262587e-07, "loss": 0.012, "step": 27588 }, { "epoch": 2.991001734605377, "grad_norm": 0.8017088174819946, "learning_rate": 3.0103003046568987e-07, "loss": 0.0279, "step": 27589 }, { "epoch": 2.991110147441457, "grad_norm": 0.15725868940353394, "learning_rate": 2.974031626287538e-07, "loss": 0.0051, "step": 27590 }, { "epoch": 2.9912185602775367, "grad_norm": 0.5554371476173401, "learning_rate": 2.937762947918178e-07, "loss": 0.035, "step": 27591 }, { "epoch": 2.9913269731136167, "grad_norm": 0.024219699203968048, "learning_rate": 2.901494269548818e-07, "loss": 0.0007, "step": 27592 }, { "epoch": 2.9914353859496963, "grad_norm": 0.019555479288101196, "learning_rate": 2.8652255911794576e-07, "loss": 0.0007, "step": 27593 }, { "epoch": 2.9915437987857763, "grad_norm": 0.09132435917854309, "learning_rate": 2.828956912810097e-07, "loss": 0.003, "step": 27594 }, { "epoch": 2.9916522116218562, "grad_norm": 0.16245554387569427, "learning_rate": 2.792688234440737e-07, "loss": 0.0055, "step": 27595 }, { "epoch": 2.991760624457936, "grad_norm": 0.07527179270982742, "learning_rate": 2.756419556071377e-07, "loss": 0.0022, "step": 27596 }, { "epoch": 2.9918690372940158, "grad_norm": 0.3356591761112213, "learning_rate": 2.720150877702017e-07, "loss": 0.0265, "step": 27597 }, { "epoch": 2.9919774501300953, "grad_norm": 0.06245657056570053, "learning_rate": 2.6838821993326565e-07, "loss": 0.0014, "step": 27598 }, { "epoch": 2.9920858629661753, "grad_norm": 0.1920478343963623, "learning_rate": 2.647613520963296e-07, "loss": 0.015, "step": 27599 }, { "epoch": 2.992194275802255, "grad_norm": 0.12098287045955658, "learning_rate": 2.611344842593936e-07, "loss": 0.0044, "step": 27600 }, { "epoch": 2.992302688638335, "grad_norm": 0.1473841369152069, "learning_rate": 2.575076164224576e-07, "loss": 0.0042, "step": 27601 }, { "epoch": 2.9924111014744144, "grad_norm": 0.006451491732150316, "learning_rate": 2.538807485855216e-07, "loss": 0.0002, "step": 27602 }, { "epoch": 2.9925195143104943, "grad_norm": 0.6110760569572449, "learning_rate": 2.5025388074858554e-07, "loss": 0.03, "step": 27603 }, { "epoch": 2.9926279271465743, "grad_norm": 0.3332903981208801, "learning_rate": 2.466270129116495e-07, "loss": 0.0371, "step": 27604 }, { "epoch": 2.992736339982654, "grad_norm": 0.4727643132209778, "learning_rate": 2.430001450747135e-07, "loss": 0.0267, "step": 27605 }, { "epoch": 2.992844752818734, "grad_norm": 0.25018036365509033, "learning_rate": 2.393732772377775e-07, "loss": 0.0089, "step": 27606 }, { "epoch": 2.9929531656548134, "grad_norm": 0.3326977491378784, "learning_rate": 2.3574640940084145e-07, "loss": 0.0064, "step": 27607 }, { "epoch": 2.9930615784908934, "grad_norm": 0.40432310104370117, "learning_rate": 2.3211954156390543e-07, "loss": 0.0167, "step": 27608 }, { "epoch": 2.9931699913269734, "grad_norm": 0.612736165523529, "learning_rate": 2.2849267372696937e-07, "loss": 0.0276, "step": 27609 }, { "epoch": 2.993278404163053, "grad_norm": 0.1597273200750351, "learning_rate": 2.2486580589003337e-07, "loss": 0.0031, "step": 27610 }, { "epoch": 2.9933868169991324, "grad_norm": 0.6600261926651001, "learning_rate": 2.2123893805309735e-07, "loss": 0.0301, "step": 27611 }, { "epoch": 2.9934952298352124, "grad_norm": 0.15431982278823853, "learning_rate": 2.1761207021616134e-07, "loss": 0.0086, "step": 27612 }, { "epoch": 2.9936036426712924, "grad_norm": 0.2821410000324249, "learning_rate": 2.1398520237922532e-07, "loss": 0.0297, "step": 27613 }, { "epoch": 2.993712055507372, "grad_norm": 0.005579422228038311, "learning_rate": 2.1035833454228932e-07, "loss": 0.0001, "step": 27614 }, { "epoch": 2.993820468343452, "grad_norm": 0.017437513917684555, "learning_rate": 2.0673146670535326e-07, "loss": 0.0006, "step": 27615 }, { "epoch": 2.9939288811795315, "grad_norm": 0.1883324235677719, "learning_rate": 2.0310459886841724e-07, "loss": 0.0107, "step": 27616 }, { "epoch": 2.9940372940156115, "grad_norm": 0.4188523590564728, "learning_rate": 1.9947773103148123e-07, "loss": 0.0175, "step": 27617 }, { "epoch": 2.9941457068516915, "grad_norm": 0.6699482798576355, "learning_rate": 1.958508631945452e-07, "loss": 0.0868, "step": 27618 }, { "epoch": 2.994254119687771, "grad_norm": 0.059027813374996185, "learning_rate": 1.9222399535760918e-07, "loss": 0.002, "step": 27619 }, { "epoch": 2.994362532523851, "grad_norm": 0.3719128966331482, "learning_rate": 1.8859712752067315e-07, "loss": 0.0163, "step": 27620 }, { "epoch": 2.9944709453599305, "grad_norm": 0.5359962582588196, "learning_rate": 1.8497025968373715e-07, "loss": 0.0114, "step": 27621 }, { "epoch": 2.9945793581960105, "grad_norm": 1.1345618963241577, "learning_rate": 1.813433918468011e-07, "loss": 0.0226, "step": 27622 }, { "epoch": 2.99468777103209, "grad_norm": 0.9460872411727905, "learning_rate": 1.777165240098651e-07, "loss": 0.0161, "step": 27623 }, { "epoch": 2.99479618386817, "grad_norm": 0.01220521330833435, "learning_rate": 1.7408965617292907e-07, "loss": 0.0005, "step": 27624 }, { "epoch": 2.9949045967042496, "grad_norm": 0.6258031725883484, "learning_rate": 1.7046278833599304e-07, "loss": 0.0397, "step": 27625 }, { "epoch": 2.9950130095403296, "grad_norm": 0.07884293794631958, "learning_rate": 1.6683592049905702e-07, "loss": 0.0023, "step": 27626 }, { "epoch": 2.9951214223764095, "grad_norm": 0.1865757405757904, "learning_rate": 1.63209052662121e-07, "loss": 0.0079, "step": 27627 }, { "epoch": 2.995229835212489, "grad_norm": 0.41399478912353516, "learning_rate": 1.59582184825185e-07, "loss": 0.0064, "step": 27628 }, { "epoch": 2.995338248048569, "grad_norm": 0.01785346306860447, "learning_rate": 1.5595531698824896e-07, "loss": 0.0006, "step": 27629 }, { "epoch": 2.9954466608846486, "grad_norm": 0.06929989904165268, "learning_rate": 1.5232844915131293e-07, "loss": 0.002, "step": 27630 }, { "epoch": 2.9955550737207286, "grad_norm": 0.20495456457138062, "learning_rate": 1.487015813143769e-07, "loss": 0.0131, "step": 27631 }, { "epoch": 2.9956634865568086, "grad_norm": 0.20525658130645752, "learning_rate": 1.450747134774409e-07, "loss": 0.0241, "step": 27632 }, { "epoch": 2.995771899392888, "grad_norm": 0.7662085294723511, "learning_rate": 1.4144784564050485e-07, "loss": 0.0749, "step": 27633 }, { "epoch": 2.9958803122289677, "grad_norm": 0.5704798102378845, "learning_rate": 1.3782097780356885e-07, "loss": 0.0201, "step": 27634 }, { "epoch": 2.9959887250650477, "grad_norm": 0.07760833948850632, "learning_rate": 1.3419410996663282e-07, "loss": 0.0039, "step": 27635 }, { "epoch": 2.9960971379011276, "grad_norm": 0.05911081284284592, "learning_rate": 1.305672421296968e-07, "loss": 0.0022, "step": 27636 }, { "epoch": 2.996205550737207, "grad_norm": 0.4955950677394867, "learning_rate": 1.269403742927608e-07, "loss": 0.0586, "step": 27637 }, { "epoch": 2.996313963573287, "grad_norm": 0.2743597626686096, "learning_rate": 1.2331350645582474e-07, "loss": 0.0149, "step": 27638 }, { "epoch": 2.9964223764093667, "grad_norm": 0.006071415729820728, "learning_rate": 1.1968663861888874e-07, "loss": 0.0002, "step": 27639 }, { "epoch": 2.9965307892454467, "grad_norm": 0.11086331307888031, "learning_rate": 1.1605977078195271e-07, "loss": 0.0052, "step": 27640 }, { "epoch": 2.9966392020815267, "grad_norm": 1.2721658945083618, "learning_rate": 1.1243290294501669e-07, "loss": 0.0292, "step": 27641 }, { "epoch": 2.996747614917606, "grad_norm": 0.24592599272727966, "learning_rate": 1.0880603510808067e-07, "loss": 0.0144, "step": 27642 }, { "epoch": 2.996856027753686, "grad_norm": 0.3026806116104126, "learning_rate": 1.0517916727114466e-07, "loss": 0.0156, "step": 27643 }, { "epoch": 2.9969644405897657, "grad_norm": 0.055637408047914505, "learning_rate": 1.0155229943420862e-07, "loss": 0.0011, "step": 27644 }, { "epoch": 2.9970728534258457, "grad_norm": 0.2608259618282318, "learning_rate": 9.79254315972726e-08, "loss": 0.0108, "step": 27645 }, { "epoch": 2.9971812662619253, "grad_norm": 0.19064922630786896, "learning_rate": 9.429856376033658e-08, "loss": 0.0128, "step": 27646 }, { "epoch": 2.9972896790980053, "grad_norm": 0.05801335349678993, "learning_rate": 9.067169592340055e-08, "loss": 0.0022, "step": 27647 }, { "epoch": 2.997398091934085, "grad_norm": 0.15303568542003632, "learning_rate": 8.704482808646454e-08, "loss": 0.005, "step": 27648 }, { "epoch": 2.997506504770165, "grad_norm": 0.380320280790329, "learning_rate": 8.341796024952851e-08, "loss": 0.0082, "step": 27649 }, { "epoch": 2.9976149176062448, "grad_norm": 0.054780829697847366, "learning_rate": 7.97910924125925e-08, "loss": 0.0014, "step": 27650 }, { "epoch": 2.9977233304423243, "grad_norm": 0.03390486165881157, "learning_rate": 7.616422457565647e-08, "loss": 0.0013, "step": 27651 }, { "epoch": 2.9978317432784043, "grad_norm": 0.4922010004520416, "learning_rate": 7.253735673872045e-08, "loss": 0.04, "step": 27652 }, { "epoch": 2.997940156114484, "grad_norm": 0.4482513666152954, "learning_rate": 6.891048890178443e-08, "loss": 0.0126, "step": 27653 }, { "epoch": 2.998048568950564, "grad_norm": 0.7189289927482605, "learning_rate": 6.52836210648484e-08, "loss": 0.0414, "step": 27654 }, { "epoch": 2.998156981786644, "grad_norm": 0.4415978789329529, "learning_rate": 6.165675322791237e-08, "loss": 0.0547, "step": 27655 }, { "epoch": 2.9982653946227233, "grad_norm": 0.08458178490400314, "learning_rate": 5.802988539097636e-08, "loss": 0.0021, "step": 27656 }, { "epoch": 2.998373807458803, "grad_norm": 0.7715836763381958, "learning_rate": 5.4403017554040336e-08, "loss": 0.0209, "step": 27657 }, { "epoch": 2.998482220294883, "grad_norm": 0.8543118238449097, "learning_rate": 5.077614971710431e-08, "loss": 0.0501, "step": 27658 }, { "epoch": 2.998590633130963, "grad_norm": 0.3826597332954407, "learning_rate": 4.714928188016829e-08, "loss": 0.0283, "step": 27659 }, { "epoch": 2.9986990459670424, "grad_norm": 0.1643172800540924, "learning_rate": 4.352241404323227e-08, "loss": 0.0061, "step": 27660 }, { "epoch": 2.9988074588031224, "grad_norm": 0.11992722004652023, "learning_rate": 3.989554620629625e-08, "loss": 0.004, "step": 27661 }, { "epoch": 2.998915871639202, "grad_norm": 1.0668443441390991, "learning_rate": 3.6268678369360226e-08, "loss": 0.0607, "step": 27662 }, { "epoch": 2.999024284475282, "grad_norm": 0.15698309242725372, "learning_rate": 3.26418105324242e-08, "loss": 0.0073, "step": 27663 }, { "epoch": 2.999132697311362, "grad_norm": 0.11465127021074295, "learning_rate": 2.901494269548818e-08, "loss": 0.0034, "step": 27664 }, { "epoch": 2.9992411101474414, "grad_norm": 0.39433804154396057, "learning_rate": 2.5388074858552154e-08, "loss": 0.0228, "step": 27665 }, { "epoch": 2.999349522983521, "grad_norm": 0.04121667146682739, "learning_rate": 2.1761207021616134e-08, "loss": 0.0015, "step": 27666 }, { "epoch": 2.999457935819601, "grad_norm": 0.30822187662124634, "learning_rate": 1.8134339184680113e-08, "loss": 0.0046, "step": 27667 }, { "epoch": 2.999566348655681, "grad_norm": 0.0032224145252257586, "learning_rate": 1.450747134774409e-08, "loss": 0.0001, "step": 27668 }, { "epoch": 2.9996747614917605, "grad_norm": 0.02294725552201271, "learning_rate": 1.0880603510808067e-08, "loss": 0.0011, "step": 27669 }, { "epoch": 2.9997831743278405, "grad_norm": 0.3940373659133911, "learning_rate": 7.253735673872045e-09, "loss": 0.0181, "step": 27670 }, { "epoch": 2.99989158716392, "grad_norm": 0.013100064359605312, "learning_rate": 3.6268678369360223e-09, "loss": 0.0003, "step": 27671 }, { "epoch": 3.0, "grad_norm": 0.04788263514637947, "learning_rate": 0.0, "loss": 0.0016, "step": 27672 }, { "epoch": 3.0, "step": 27672, "total_flos": 3.556142009271976e+18, "train_loss": 0.030263915256900305, "train_runtime": 5200.1946, "train_samples_per_second": 85.139, "train_steps_per_second": 5.321 } ], "logging_steps": 1.0, "max_steps": 27672, "num_input_tokens_seen": 0, "num_train_epochs": 3, "save_steps": 10000, "stateful_callbacks": { "TrainerControl": { "args": { "should_epoch_stop": false, "should_evaluate": false, "should_log": false, "should_save": true, "should_training_stop": true }, "attributes": {} } }, "total_flos": 3.556142009271976e+18, "train_batch_size": 2, "trial_name": null, "trial_params": null }