{ "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": 57.150794982910156, "learning_rate": 2.0000000000000002e-07, "loss": 2.3619, "step": 1 }, { "epoch": 0.0002168256721595837, "grad_norm": 59.16875457763672, "learning_rate": 4.0000000000000003e-07, "loss": 2.7161, "step": 2 }, { "epoch": 0.00032523850823937554, "grad_norm": 59.990291595458984, "learning_rate": 6.000000000000001e-07, "loss": 2.8909, "step": 3 }, { "epoch": 0.0004336513443191674, "grad_norm": 51.445213317871094, "learning_rate": 8.000000000000001e-07, "loss": 2.2286, "step": 4 }, { "epoch": 0.0005420641803989592, "grad_norm": 53.74861145019531, "learning_rate": 1.0000000000000002e-06, "loss": 2.8618, "step": 5 }, { "epoch": 0.0006504770164787511, "grad_norm": 57.22797775268555, "learning_rate": 1.2000000000000002e-06, "loss": 2.2517, "step": 6 }, { "epoch": 0.000758889852558543, "grad_norm": 57.10231018066406, "learning_rate": 1.4000000000000001e-06, "loss": 2.8519, "step": 7 }, { "epoch": 0.0008673026886383347, "grad_norm": 69.48921203613281, "learning_rate": 1.6000000000000001e-06, "loss": 4.1166, "step": 8 }, { "epoch": 0.0009757155247181266, "grad_norm": 61.26288604736328, "learning_rate": 1.8000000000000001e-06, "loss": 3.412, "step": 9 }, { "epoch": 0.0010841283607979184, "grad_norm": 54.14117431640625, "learning_rate": 2.0000000000000003e-06, "loss": 2.2459, "step": 10 }, { "epoch": 0.0011925411968777104, "grad_norm": 56.54292678833008, "learning_rate": 2.2e-06, "loss": 2.6444, "step": 11 }, { "epoch": 0.0013009540329575022, "grad_norm": 52.101566314697266, "learning_rate": 2.4000000000000003e-06, "loss": 2.0482, "step": 12 }, { "epoch": 0.001409366869037294, "grad_norm": 50.745853424072266, "learning_rate": 2.6e-06, "loss": 3.0478, "step": 13 }, { "epoch": 0.001517779705117086, "grad_norm": 47.69984817504883, "learning_rate": 2.8000000000000003e-06, "loss": 1.8895, "step": 14 }, { "epoch": 0.0016261925411968777, "grad_norm": 46.83454132080078, "learning_rate": 3e-06, "loss": 1.5868, "step": 15 }, { "epoch": 0.0017346053772766695, "grad_norm": 52.21897506713867, "learning_rate": 3.2000000000000003e-06, "loss": 3.2407, "step": 16 }, { "epoch": 0.0018430182133564615, "grad_norm": 43.224143981933594, "learning_rate": 3.4000000000000005e-06, "loss": 1.9491, "step": 17 }, { "epoch": 0.0019514310494362533, "grad_norm": 155.4527587890625, "learning_rate": 3.6000000000000003e-06, "loss": 1.7262, "step": 18 }, { "epoch": 0.002059843885516045, "grad_norm": 64.4134521484375, "learning_rate": 3.8000000000000005e-06, "loss": 2.0973, "step": 19 }, { "epoch": 0.002168256721595837, "grad_norm": 45.826141357421875, "learning_rate": 4.000000000000001e-06, "loss": 2.1399, "step": 20 }, { "epoch": 0.0022766695576756286, "grad_norm": 35.6070671081543, "learning_rate": 4.2000000000000004e-06, "loss": 0.8614, "step": 21 }, { "epoch": 0.002385082393755421, "grad_norm": 32.088539123535156, "learning_rate": 4.4e-06, "loss": 0.7822, "step": 22 }, { "epoch": 0.0024934952298352126, "grad_norm": 29.70172119140625, "learning_rate": 4.600000000000001e-06, "loss": 0.5557, "step": 23 }, { "epoch": 0.0026019080659150044, "grad_norm": 18.708255767822266, "learning_rate": 4.800000000000001e-06, "loss": 0.3873, "step": 24 }, { "epoch": 0.002710320901994796, "grad_norm": 32.96669006347656, "learning_rate": 5e-06, "loss": 0.8715, "step": 25 }, { "epoch": 0.002818733738074588, "grad_norm": 28.178295135498047, "learning_rate": 5.2e-06, "loss": 0.8427, "step": 26 }, { "epoch": 0.0029271465741543797, "grad_norm": 36.9205436706543, "learning_rate": 5.400000000000001e-06, "loss": 1.652, "step": 27 }, { "epoch": 0.003035559410234172, "grad_norm": 7.869051933288574, "learning_rate": 5.600000000000001e-06, "loss": 0.1751, "step": 28 }, { "epoch": 0.0031439722463139637, "grad_norm": 149.48646545410156, "learning_rate": 5.8e-06, "loss": 0.9414, "step": 29 }, { "epoch": 0.0032523850823937554, "grad_norm": 6.688308238983154, "learning_rate": 6e-06, "loss": 0.1182, "step": 30 }, { "epoch": 0.0033607979184735472, "grad_norm": 14.491106033325195, "learning_rate": 6.200000000000001e-06, "loss": 0.1764, "step": 31 }, { "epoch": 0.003469210754553339, "grad_norm": 13.82690143585205, "learning_rate": 6.4000000000000006e-06, "loss": 0.9307, "step": 32 }, { "epoch": 0.0035776235906331308, "grad_norm": 9.66311264038086, "learning_rate": 6.600000000000001e-06, "loss": 0.512, "step": 33 }, { "epoch": 0.003686036426712923, "grad_norm": 6.834539413452148, "learning_rate": 6.800000000000001e-06, "loss": 0.1162, "step": 34 }, { "epoch": 0.0037944492627927148, "grad_norm": 29.092525482177734, "learning_rate": 7e-06, "loss": 0.2341, "step": 35 }, { "epoch": 0.0039028620988725065, "grad_norm": 21.331005096435547, "learning_rate": 7.2000000000000005e-06, "loss": 0.1279, "step": 36 }, { "epoch": 0.004011274934952298, "grad_norm": 21.174837112426758, "learning_rate": 7.4e-06, "loss": 0.1025, "step": 37 }, { "epoch": 0.00411968777103209, "grad_norm": 10.552841186523438, "learning_rate": 7.600000000000001e-06, "loss": 0.0654, "step": 38 }, { "epoch": 0.004228100607111882, "grad_norm": 10.130504608154297, "learning_rate": 7.800000000000002e-06, "loss": 0.7554, "step": 39 }, { "epoch": 0.004336513443191674, "grad_norm": 32.685733795166016, "learning_rate": 8.000000000000001e-06, "loss": 0.6618, "step": 40 }, { "epoch": 0.004444926279271465, "grad_norm": 14.20527458190918, "learning_rate": 8.2e-06, "loss": 0.1476, "step": 41 }, { "epoch": 0.004553339115351257, "grad_norm": 22.93535804748535, "learning_rate": 8.400000000000001e-06, "loss": 0.1254, "step": 42 }, { "epoch": 0.00466175195143105, "grad_norm": 10.033289909362793, "learning_rate": 8.6e-06, "loss": 0.3569, "step": 43 }, { "epoch": 0.004770164787510842, "grad_norm": 28.221036911010742, "learning_rate": 8.8e-06, "loss": 0.8318, "step": 44 }, { "epoch": 0.004878577623590633, "grad_norm": 9.1015043258667, "learning_rate": 9e-06, "loss": 0.0866, "step": 45 }, { "epoch": 0.004986990459670425, "grad_norm": 15.505698204040527, "learning_rate": 9.200000000000002e-06, "loss": 0.3491, "step": 46 }, { "epoch": 0.005095403295750217, "grad_norm": 11.363885879516602, "learning_rate": 9.4e-06, "loss": 0.1115, "step": 47 }, { "epoch": 0.005203816131830009, "grad_norm": 7.338240146636963, "learning_rate": 9.600000000000001e-06, "loss": 0.3156, "step": 48 }, { "epoch": 0.0053122289679098005, "grad_norm": 9.689183235168457, "learning_rate": 9.800000000000001e-06, "loss": 0.1247, "step": 49 }, { "epoch": 0.005420641803989592, "grad_norm": 7.674718856811523, "learning_rate": 1e-05, "loss": 0.2486, "step": 50 }, { "epoch": 0.005529054640069384, "grad_norm": 7.4485602378845215, "learning_rate": 1.02e-05, "loss": 0.3112, "step": 51 }, { "epoch": 0.005637467476149176, "grad_norm": 8.378499984741211, "learning_rate": 1.04e-05, "loss": 0.0861, "step": 52 }, { "epoch": 0.005745880312228968, "grad_norm": 6.716836452484131, "learning_rate": 1.0600000000000002e-05, "loss": 0.1092, "step": 53 }, { "epoch": 0.005854293148308759, "grad_norm": 10.12954330444336, "learning_rate": 1.0800000000000002e-05, "loss": 0.4116, "step": 54 }, { "epoch": 0.005962705984388552, "grad_norm": 12.715435981750488, "learning_rate": 1.1000000000000001e-05, "loss": 0.3494, "step": 55 }, { "epoch": 0.006071118820468344, "grad_norm": 4.615040302276611, "learning_rate": 1.1200000000000001e-05, "loss": 0.2357, "step": 56 }, { "epoch": 0.0061795316565481356, "grad_norm": 21.645824432373047, "learning_rate": 1.14e-05, "loss": 0.4544, "step": 57 }, { "epoch": 0.006287944492627927, "grad_norm": 30.776884078979492, "learning_rate": 1.16e-05, "loss": 0.4019, "step": 58 }, { "epoch": 0.006396357328707719, "grad_norm": 7.740161418914795, "learning_rate": 1.18e-05, "loss": 0.2946, "step": 59 }, { "epoch": 0.006504770164787511, "grad_norm": 8.730713844299316, "learning_rate": 1.2e-05, "loss": 0.0974, "step": 60 }, { "epoch": 0.006613183000867303, "grad_norm": 4.969794273376465, "learning_rate": 1.22e-05, "loss": 0.0816, "step": 61 }, { "epoch": 0.0067215958369470944, "grad_norm": 6.198244094848633, "learning_rate": 1.2400000000000002e-05, "loss": 0.1684, "step": 62 }, { "epoch": 0.006830008673026886, "grad_norm": 4.78057336807251, "learning_rate": 1.2600000000000001e-05, "loss": 0.0785, "step": 63 }, { "epoch": 0.006938421509106678, "grad_norm": 5.221713542938232, "learning_rate": 1.2800000000000001e-05, "loss": 0.057, "step": 64 }, { "epoch": 0.00704683434518647, "grad_norm": 61.682716369628906, "learning_rate": 1.3000000000000001e-05, "loss": 0.3664, "step": 65 }, { "epoch": 0.0071552471812662615, "grad_norm": 21.07749366760254, "learning_rate": 1.3200000000000002e-05, "loss": 0.1747, "step": 66 }, { "epoch": 0.007263660017346054, "grad_norm": 10.302938461303711, "learning_rate": 1.3400000000000002e-05, "loss": 0.3539, "step": 67 }, { "epoch": 0.007372072853425846, "grad_norm": 5.1282548904418945, "learning_rate": 1.3600000000000002e-05, "loss": 0.0692, "step": 68 }, { "epoch": 0.007480485689505638, "grad_norm": 10.027702331542969, "learning_rate": 1.38e-05, "loss": 0.2008, "step": 69 }, { "epoch": 0.0075888985255854295, "grad_norm": 8.681930541992188, "learning_rate": 1.4e-05, "loss": 0.0812, "step": 70 }, { "epoch": 0.007697311361665221, "grad_norm": 15.262502670288086, "learning_rate": 1.4200000000000001e-05, "loss": 0.1038, "step": 71 }, { "epoch": 0.007805724197745013, "grad_norm": 10.009432792663574, "learning_rate": 1.4400000000000001e-05, "loss": 0.258, "step": 72 }, { "epoch": 0.007914137033824806, "grad_norm": 12.003353118896484, "learning_rate": 1.46e-05, "loss": 0.2482, "step": 73 }, { "epoch": 0.008022549869904597, "grad_norm": 12.08552074432373, "learning_rate": 1.48e-05, "loss": 0.1736, "step": 74 }, { "epoch": 0.00813096270598439, "grad_norm": 7.710125923156738, "learning_rate": 1.5000000000000002e-05, "loss": 0.2453, "step": 75 }, { "epoch": 0.00823937554206418, "grad_norm": 3.9444501399993896, "learning_rate": 1.5200000000000002e-05, "loss": 0.0595, "step": 76 }, { "epoch": 0.008347788378143973, "grad_norm": 23.556873321533203, "learning_rate": 1.54e-05, "loss": 0.1678, "step": 77 }, { "epoch": 0.008456201214223764, "grad_norm": 4.988242149353027, "learning_rate": 1.5600000000000003e-05, "loss": 0.0763, "step": 78 }, { "epoch": 0.008564614050303556, "grad_norm": 3.7871384620666504, "learning_rate": 1.58e-05, "loss": 0.0524, "step": 79 }, { "epoch": 0.008673026886383347, "grad_norm": 7.529158592224121, "learning_rate": 1.6000000000000003e-05, "loss": 0.1269, "step": 80 }, { "epoch": 0.00878143972246314, "grad_norm": 8.613714218139648, "learning_rate": 1.62e-05, "loss": 0.1345, "step": 81 }, { "epoch": 0.00888985255854293, "grad_norm": 7.141444683074951, "learning_rate": 1.64e-05, "loss": 0.0984, "step": 82 }, { "epoch": 0.008998265394622723, "grad_norm": 8.935779571533203, "learning_rate": 1.66e-05, "loss": 0.1203, "step": 83 }, { "epoch": 0.009106678230702514, "grad_norm": 12.946879386901855, "learning_rate": 1.6800000000000002e-05, "loss": 0.0779, "step": 84 }, { "epoch": 0.009215091066782307, "grad_norm": 16.487300872802734, "learning_rate": 1.7e-05, "loss": 0.2744, "step": 85 }, { "epoch": 0.0093235039028621, "grad_norm": 14.098669052124023, "learning_rate": 1.72e-05, "loss": 0.1064, "step": 86 }, { "epoch": 0.00943191673894189, "grad_norm": 10.5635404586792, "learning_rate": 1.7400000000000003e-05, "loss": 0.049, "step": 87 }, { "epoch": 0.009540329575021683, "grad_norm": 14.344745635986328, "learning_rate": 1.76e-05, "loss": 0.1311, "step": 88 }, { "epoch": 0.009648742411101474, "grad_norm": 13.231096267700195, "learning_rate": 1.7800000000000002e-05, "loss": 0.2126, "step": 89 }, { "epoch": 0.009757155247181267, "grad_norm": 12.61841869354248, "learning_rate": 1.8e-05, "loss": 0.0889, "step": 90 }, { "epoch": 0.009865568083261058, "grad_norm": 5.281639099121094, "learning_rate": 1.8200000000000002e-05, "loss": 0.0656, "step": 91 }, { "epoch": 0.00997398091934085, "grad_norm": 5.846783638000488, "learning_rate": 1.8400000000000003e-05, "loss": 0.0669, "step": 92 }, { "epoch": 0.010082393755420641, "grad_norm": 7.0033111572265625, "learning_rate": 1.86e-05, "loss": 0.0708, "step": 93 }, { "epoch": 0.010190806591500434, "grad_norm": 4.681243419647217, "learning_rate": 1.88e-05, "loss": 0.0647, "step": 94 }, { "epoch": 0.010299219427580225, "grad_norm": 4.785581111907959, "learning_rate": 1.9e-05, "loss": 0.115, "step": 95 }, { "epoch": 0.010407632263660017, "grad_norm": 5.340388774871826, "learning_rate": 1.9200000000000003e-05, "loss": 0.0735, "step": 96 }, { "epoch": 0.010516045099739808, "grad_norm": 6.11476993560791, "learning_rate": 1.94e-05, "loss": 0.0924, "step": 97 }, { "epoch": 0.010624457935819601, "grad_norm": 3.6978960037231445, "learning_rate": 1.9600000000000002e-05, "loss": 0.0563, "step": 98 }, { "epoch": 0.010732870771899394, "grad_norm": 6.782694339752197, "learning_rate": 1.98e-05, "loss": 0.0629, "step": 99 }, { "epoch": 0.010841283607979185, "grad_norm": 6.348627090454102, "learning_rate": 2e-05, "loss": 0.0891, "step": 100 }, { "epoch": 0.010949696444058977, "grad_norm": 4.3263840675354, "learning_rate": 1.9999274626432614e-05, "loss": 0.0639, "step": 101 }, { "epoch": 0.011058109280138768, "grad_norm": 4.359460830688477, "learning_rate": 1.9998549252865226e-05, "loss": 0.1048, "step": 102 }, { "epoch": 0.01116652211621856, "grad_norm": 7.1502885818481445, "learning_rate": 1.9997823879297838e-05, "loss": 0.1225, "step": 103 }, { "epoch": 0.011274934952298352, "grad_norm": 6.801154613494873, "learning_rate": 1.9997098505730454e-05, "loss": 0.0624, "step": 104 }, { "epoch": 0.011383347788378144, "grad_norm": 12.252493858337402, "learning_rate": 1.9996373132163066e-05, "loss": 0.0829, "step": 105 }, { "epoch": 0.011491760624457935, "grad_norm": 9.569732666015625, "learning_rate": 1.9995647758595678e-05, "loss": 0.0915, "step": 106 }, { "epoch": 0.011600173460537728, "grad_norm": 6.3287672996521, "learning_rate": 1.999492238502829e-05, "loss": 0.1092, "step": 107 }, { "epoch": 0.011708586296617519, "grad_norm": 8.559252738952637, "learning_rate": 1.9994197011460902e-05, "loss": 0.0921, "step": 108 }, { "epoch": 0.011816999132697311, "grad_norm": 3.028735876083374, "learning_rate": 1.9993471637893518e-05, "loss": 0.0285, "step": 109 }, { "epoch": 0.011925411968777104, "grad_norm": 4.298201084136963, "learning_rate": 1.999274626432613e-05, "loss": 0.0543, "step": 110 }, { "epoch": 0.012033824804856895, "grad_norm": 8.963627815246582, "learning_rate": 1.9992020890758742e-05, "loss": 0.0691, "step": 111 }, { "epoch": 0.012142237640936688, "grad_norm": 15.208434104919434, "learning_rate": 1.9991295517191358e-05, "loss": 0.1211, "step": 112 }, { "epoch": 0.012250650477016478, "grad_norm": 2.675328016281128, "learning_rate": 1.999057014362397e-05, "loss": 0.0403, "step": 113 }, { "epoch": 0.012359063313096271, "grad_norm": 4.500398635864258, "learning_rate": 1.9989844770056582e-05, "loss": 0.0536, "step": 114 }, { "epoch": 0.012467476149176062, "grad_norm": 4.679221153259277, "learning_rate": 1.9989119396489194e-05, "loss": 0.0827, "step": 115 }, { "epoch": 0.012575888985255855, "grad_norm": 1.9377232789993286, "learning_rate": 1.9988394022921806e-05, "loss": 0.0181, "step": 116 }, { "epoch": 0.012684301821335646, "grad_norm": 3.3454489707946777, "learning_rate": 1.9987668649354418e-05, "loss": 0.0468, "step": 117 }, { "epoch": 0.012792714657415438, "grad_norm": 4.785127639770508, "learning_rate": 1.998694327578703e-05, "loss": 0.0861, "step": 118 }, { "epoch": 0.01290112749349523, "grad_norm": 6.856503486633301, "learning_rate": 1.9986217902219646e-05, "loss": 0.1186, "step": 119 }, { "epoch": 0.013009540329575022, "grad_norm": 7.048886775970459, "learning_rate": 1.9985492528652258e-05, "loss": 0.0521, "step": 120 }, { "epoch": 0.013117953165654813, "grad_norm": 12.406437873840332, "learning_rate": 1.998476715508487e-05, "loss": 0.0724, "step": 121 }, { "epoch": 0.013226366001734605, "grad_norm": 5.833383083343506, "learning_rate": 1.9984041781517482e-05, "loss": 0.0796, "step": 122 }, { "epoch": 0.013334778837814398, "grad_norm": 5.889612674713135, "learning_rate": 1.9983316407950094e-05, "loss": 0.0659, "step": 123 }, { "epoch": 0.013443191673894189, "grad_norm": 4.773509502410889, "learning_rate": 1.9982591034382707e-05, "loss": 0.0887, "step": 124 }, { "epoch": 0.013551604509973982, "grad_norm": 8.074460983276367, "learning_rate": 1.9981865660815322e-05, "loss": 0.1175, "step": 125 }, { "epoch": 0.013660017346053772, "grad_norm": 5.160822868347168, "learning_rate": 1.9981140287247934e-05, "loss": 0.1013, "step": 126 }, { "epoch": 0.013768430182133565, "grad_norm": 3.954429864883423, "learning_rate": 1.9980414913680546e-05, "loss": 0.0657, "step": 127 }, { "epoch": 0.013876843018213356, "grad_norm": 6.199354648590088, "learning_rate": 1.9979689540113162e-05, "loss": 0.0887, "step": 128 }, { "epoch": 0.013985255854293149, "grad_norm": 5.278064727783203, "learning_rate": 1.9978964166545774e-05, "loss": 0.0395, "step": 129 }, { "epoch": 0.01409366869037294, "grad_norm": 57.99750518798828, "learning_rate": 1.9978238792978386e-05, "loss": 0.0975, "step": 130 }, { "epoch": 0.014202081526452732, "grad_norm": 5.00956392288208, "learning_rate": 1.9977513419411e-05, "loss": 0.1036, "step": 131 }, { "epoch": 0.014310494362532523, "grad_norm": 7.466270446777344, "learning_rate": 1.997678804584361e-05, "loss": 0.0794, "step": 132 }, { "epoch": 0.014418907198612316, "grad_norm": 7.771742820739746, "learning_rate": 1.9976062672276223e-05, "loss": 0.1888, "step": 133 }, { "epoch": 0.014527320034692108, "grad_norm": 4.39603328704834, "learning_rate": 1.9975337298708838e-05, "loss": 0.0781, "step": 134 }, { "epoch": 0.0146357328707719, "grad_norm": 3.049839973449707, "learning_rate": 1.997461192514145e-05, "loss": 0.0866, "step": 135 }, { "epoch": 0.014744145706851692, "grad_norm": 3.9373066425323486, "learning_rate": 1.9973886551574063e-05, "loss": 0.0934, "step": 136 }, { "epoch": 0.014852558542931483, "grad_norm": 3.2314252853393555, "learning_rate": 1.9973161178006675e-05, "loss": 0.0643, "step": 137 }, { "epoch": 0.014960971379011275, "grad_norm": 9.10379409790039, "learning_rate": 1.9972435804439287e-05, "loss": 0.0805, "step": 138 }, { "epoch": 0.015069384215091066, "grad_norm": 2.0692524909973145, "learning_rate": 1.99717104308719e-05, "loss": 0.0631, "step": 139 }, { "epoch": 0.015177797051170859, "grad_norm": 3.780812978744507, "learning_rate": 1.997098505730451e-05, "loss": 0.1241, "step": 140 }, { "epoch": 0.01528620988725065, "grad_norm": 3.923036813735962, "learning_rate": 1.9970259683737127e-05, "loss": 0.0636, "step": 141 }, { "epoch": 0.015394622723330443, "grad_norm": 4.1278181076049805, "learning_rate": 1.996953431016974e-05, "loss": 0.0726, "step": 142 }, { "epoch": 0.015503035559410233, "grad_norm": 3.6481919288635254, "learning_rate": 1.996880893660235e-05, "loss": 0.0794, "step": 143 }, { "epoch": 0.015611448395490026, "grad_norm": 5.415205001831055, "learning_rate": 1.9968083563034966e-05, "loss": 0.0908, "step": 144 }, { "epoch": 0.015719861231569817, "grad_norm": 10.091710090637207, "learning_rate": 1.996735818946758e-05, "loss": 0.084, "step": 145 }, { "epoch": 0.01582827406764961, "grad_norm": 3.740361452102661, "learning_rate": 1.996663281590019e-05, "loss": 0.0593, "step": 146 }, { "epoch": 0.015936686903729402, "grad_norm": 5.377466201782227, "learning_rate": 1.9965907442332803e-05, "loss": 0.0955, "step": 147 }, { "epoch": 0.016045099739809193, "grad_norm": 3.1083779335021973, "learning_rate": 1.9965182068765415e-05, "loss": 0.0414, "step": 148 }, { "epoch": 0.016153512575888984, "grad_norm": 3.254978656768799, "learning_rate": 1.996445669519803e-05, "loss": 0.086, "step": 149 }, { "epoch": 0.01626192541196878, "grad_norm": 3.5187439918518066, "learning_rate": 1.9963731321630643e-05, "loss": 0.0527, "step": 150 }, { "epoch": 0.01637033824804857, "grad_norm": 6.65952205657959, "learning_rate": 1.9963005948063255e-05, "loss": 0.1023, "step": 151 }, { "epoch": 0.01647875108412836, "grad_norm": 3.7421016693115234, "learning_rate": 1.9962280574495867e-05, "loss": 0.0588, "step": 152 }, { "epoch": 0.01658716392020815, "grad_norm": 3.7461371421813965, "learning_rate": 1.996155520092848e-05, "loss": 0.0412, "step": 153 }, { "epoch": 0.016695576756287946, "grad_norm": 3.9383225440979004, "learning_rate": 1.996082982736109e-05, "loss": 0.0408, "step": 154 }, { "epoch": 0.016803989592367737, "grad_norm": 6.8064165115356445, "learning_rate": 1.9960104453793703e-05, "loss": 0.091, "step": 155 }, { "epoch": 0.016912402428447527, "grad_norm": 8.244215965270996, "learning_rate": 1.995937908022632e-05, "loss": 0.0625, "step": 156 }, { "epoch": 0.01702081526452732, "grad_norm": 2.6720733642578125, "learning_rate": 1.995865370665893e-05, "loss": 0.039, "step": 157 }, { "epoch": 0.017129228100607113, "grad_norm": 4.732871055603027, "learning_rate": 1.9957928333091543e-05, "loss": 0.036, "step": 158 }, { "epoch": 0.017237640936686904, "grad_norm": 4.913729667663574, "learning_rate": 1.9957202959524155e-05, "loss": 0.063, "step": 159 }, { "epoch": 0.017346053772766695, "grad_norm": 5.017018795013428, "learning_rate": 1.9956477585956768e-05, "loss": 0.0554, "step": 160 }, { "epoch": 0.01745446660884649, "grad_norm": 6.174421310424805, "learning_rate": 1.9955752212389383e-05, "loss": 0.0706, "step": 161 }, { "epoch": 0.01756287944492628, "grad_norm": 6.572563648223877, "learning_rate": 1.9955026838821995e-05, "loss": 0.0597, "step": 162 }, { "epoch": 0.01767129228100607, "grad_norm": 3.5350964069366455, "learning_rate": 1.9954301465254607e-05, "loss": 0.0374, "step": 163 }, { "epoch": 0.01777970511708586, "grad_norm": 4.139571666717529, "learning_rate": 1.9953576091687223e-05, "loss": 0.0666, "step": 164 }, { "epoch": 0.017888117953165656, "grad_norm": 1.6026874780654907, "learning_rate": 1.9952850718119835e-05, "loss": 0.0256, "step": 165 }, { "epoch": 0.017996530789245447, "grad_norm": 20.469154357910156, "learning_rate": 1.9952125344552447e-05, "loss": 0.0397, "step": 166 }, { "epoch": 0.018104943625325238, "grad_norm": 4.979197025299072, "learning_rate": 1.995139997098506e-05, "loss": 0.0614, "step": 167 }, { "epoch": 0.01821335646140503, "grad_norm": 20.0063419342041, "learning_rate": 1.995067459741767e-05, "loss": 0.1701, "step": 168 }, { "epoch": 0.018321769297484823, "grad_norm": 11.68804931640625, "learning_rate": 1.9949949223850284e-05, "loss": 0.1776, "step": 169 }, { "epoch": 0.018430182133564614, "grad_norm": 16.497583389282227, "learning_rate": 1.9949223850282896e-05, "loss": 0.0663, "step": 170 }, { "epoch": 0.018538594969644405, "grad_norm": 8.709933280944824, "learning_rate": 1.994849847671551e-05, "loss": 0.0639, "step": 171 }, { "epoch": 0.0186470078057242, "grad_norm": 3.468261957168579, "learning_rate": 1.9947773103148123e-05, "loss": 0.0622, "step": 172 }, { "epoch": 0.01875542064180399, "grad_norm": 3.441314697265625, "learning_rate": 1.9947047729580736e-05, "loss": 0.0298, "step": 173 }, { "epoch": 0.01886383347788378, "grad_norm": 7.033847808837891, "learning_rate": 1.9946322356013348e-05, "loss": 0.1366, "step": 174 }, { "epoch": 0.018972246313963572, "grad_norm": 8.174678802490234, "learning_rate": 1.994559698244596e-05, "loss": 0.0958, "step": 175 }, { "epoch": 0.019080659150043366, "grad_norm": 3.1684985160827637, "learning_rate": 1.9944871608878572e-05, "loss": 0.0562, "step": 176 }, { "epoch": 0.019189071986123157, "grad_norm": 4.6616291999816895, "learning_rate": 1.9944146235311187e-05, "loss": 0.0647, "step": 177 }, { "epoch": 0.019297484822202948, "grad_norm": 7.874721527099609, "learning_rate": 1.99434208617438e-05, "loss": 0.0631, "step": 178 }, { "epoch": 0.01940589765828274, "grad_norm": 2.179142951965332, "learning_rate": 1.9942695488176412e-05, "loss": 0.0353, "step": 179 }, { "epoch": 0.019514310494362534, "grad_norm": 5.5907182693481445, "learning_rate": 1.9941970114609027e-05, "loss": 0.0634, "step": 180 }, { "epoch": 0.019622723330442324, "grad_norm": 4.396827220916748, "learning_rate": 1.994124474104164e-05, "loss": 0.0582, "step": 181 }, { "epoch": 0.019731136166522115, "grad_norm": 4.3426690101623535, "learning_rate": 1.994051936747425e-05, "loss": 0.0973, "step": 182 }, { "epoch": 0.01983954900260191, "grad_norm": 4.41453218460083, "learning_rate": 1.9939793993906864e-05, "loss": 0.0582, "step": 183 }, { "epoch": 0.0199479618386817, "grad_norm": 5.4881181716918945, "learning_rate": 1.9939068620339476e-05, "loss": 0.0873, "step": 184 }, { "epoch": 0.02005637467476149, "grad_norm": 3.6085894107818604, "learning_rate": 1.9938343246772088e-05, "loss": 0.0649, "step": 185 }, { "epoch": 0.020164787510841282, "grad_norm": 4.910674571990967, "learning_rate": 1.9937617873204704e-05, "loss": 0.1088, "step": 186 }, { "epoch": 0.020273200346921077, "grad_norm": 2.6736459732055664, "learning_rate": 1.9936892499637316e-05, "loss": 0.0501, "step": 187 }, { "epoch": 0.020381613183000868, "grad_norm": 2.984663248062134, "learning_rate": 1.9936167126069928e-05, "loss": 0.0475, "step": 188 }, { "epoch": 0.02049002601908066, "grad_norm": 3.905702829360962, "learning_rate": 1.993544175250254e-05, "loss": 0.0773, "step": 189 }, { "epoch": 0.02059843885516045, "grad_norm": 2.531740665435791, "learning_rate": 1.9934716378935152e-05, "loss": 0.0772, "step": 190 }, { "epoch": 0.020706851691240244, "grad_norm": 3.8155481815338135, "learning_rate": 1.9933991005367764e-05, "loss": 0.0443, "step": 191 }, { "epoch": 0.020815264527320035, "grad_norm": 7.006213188171387, "learning_rate": 1.9933265631800376e-05, "loss": 0.0775, "step": 192 }, { "epoch": 0.020923677363399826, "grad_norm": 4.443814277648926, "learning_rate": 1.9932540258232992e-05, "loss": 0.0535, "step": 193 }, { "epoch": 0.021032090199479617, "grad_norm": 3.6865551471710205, "learning_rate": 1.9931814884665604e-05, "loss": 0.0884, "step": 194 }, { "epoch": 0.02114050303555941, "grad_norm": 3.5906333923339844, "learning_rate": 1.9931089511098216e-05, "loss": 0.0527, "step": 195 }, { "epoch": 0.021248915871639202, "grad_norm": 7.352570533752441, "learning_rate": 1.9930364137530832e-05, "loss": 0.0461, "step": 196 }, { "epoch": 0.021357328707718993, "grad_norm": 5.830576419830322, "learning_rate": 1.9929638763963444e-05, "loss": 0.0679, "step": 197 }, { "epoch": 0.021465741543798787, "grad_norm": 5.136837005615234, "learning_rate": 1.9928913390396056e-05, "loss": 0.0519, "step": 198 }, { "epoch": 0.021574154379878578, "grad_norm": 13.402018547058105, "learning_rate": 1.9928188016828668e-05, "loss": 0.1072, "step": 199 }, { "epoch": 0.02168256721595837, "grad_norm": 7.807851314544678, "learning_rate": 1.992746264326128e-05, "loss": 0.0797, "step": 200 }, { "epoch": 0.02179098005203816, "grad_norm": 15.532986640930176, "learning_rate": 1.9926737269693896e-05, "loss": 0.0727, "step": 201 }, { "epoch": 0.021899392888117954, "grad_norm": 3.6494734287261963, "learning_rate": 1.9926011896126508e-05, "loss": 0.0338, "step": 202 }, { "epoch": 0.022007805724197745, "grad_norm": 2.688082456588745, "learning_rate": 1.992528652255912e-05, "loss": 0.0593, "step": 203 }, { "epoch": 0.022116218560277536, "grad_norm": 7.94487190246582, "learning_rate": 1.9924561148991732e-05, "loss": 0.0836, "step": 204 }, { "epoch": 0.022224631396357327, "grad_norm": 5.5981903076171875, "learning_rate": 1.9923835775424344e-05, "loss": 0.0585, "step": 205 }, { "epoch": 0.02233304423243712, "grad_norm": 7.298788070678711, "learning_rate": 1.9923110401856957e-05, "loss": 0.1235, "step": 206 }, { "epoch": 0.022441457068516912, "grad_norm": 6.219573020935059, "learning_rate": 1.992238502828957e-05, "loss": 0.1354, "step": 207 }, { "epoch": 0.022549869904596703, "grad_norm": 3.4363887310028076, "learning_rate": 1.9921659654722184e-05, "loss": 0.0806, "step": 208 }, { "epoch": 0.022658282740676498, "grad_norm": 9.735016822814941, "learning_rate": 1.9920934281154796e-05, "loss": 0.0695, "step": 209 }, { "epoch": 0.02276669557675629, "grad_norm": 1.7484235763549805, "learning_rate": 1.992020890758741e-05, "loss": 0.0277, "step": 210 }, { "epoch": 0.02287510841283608, "grad_norm": 4.7949323654174805, "learning_rate": 1.991948353402002e-05, "loss": 0.0626, "step": 211 }, { "epoch": 0.02298352124891587, "grad_norm": 2.7697622776031494, "learning_rate": 1.9918758160452633e-05, "loss": 0.0425, "step": 212 }, { "epoch": 0.023091934084995665, "grad_norm": 4.308955669403076, "learning_rate": 1.991803278688525e-05, "loss": 0.0568, "step": 213 }, { "epoch": 0.023200346921075456, "grad_norm": 3.626051902770996, "learning_rate": 1.991730741331786e-05, "loss": 0.0631, "step": 214 }, { "epoch": 0.023308759757155247, "grad_norm": 2.2684319019317627, "learning_rate": 1.9916582039750473e-05, "loss": 0.0385, "step": 215 }, { "epoch": 0.023417172593235037, "grad_norm": 3.4415597915649414, "learning_rate": 1.9915856666183088e-05, "loss": 0.0393, "step": 216 }, { "epoch": 0.023525585429314832, "grad_norm": 4.4592180252075195, "learning_rate": 1.99151312926157e-05, "loss": 0.081, "step": 217 }, { "epoch": 0.023633998265394623, "grad_norm": 3.8447790145874023, "learning_rate": 1.9914405919048312e-05, "loss": 0.0826, "step": 218 }, { "epoch": 0.023742411101474414, "grad_norm": 2.5203330516815186, "learning_rate": 1.9913680545480925e-05, "loss": 0.0425, "step": 219 }, { "epoch": 0.023850823937554208, "grad_norm": 9.600034713745117, "learning_rate": 1.9912955171913537e-05, "loss": 0.1071, "step": 220 }, { "epoch": 0.023959236773634, "grad_norm": 5.377425193786621, "learning_rate": 1.991222979834615e-05, "loss": 0.0252, "step": 221 }, { "epoch": 0.02406764960971379, "grad_norm": 5.8447794914245605, "learning_rate": 1.991150442477876e-05, "loss": 0.1114, "step": 222 }, { "epoch": 0.02417606244579358, "grad_norm": 4.683376789093018, "learning_rate": 1.9910779051211377e-05, "loss": 0.069, "step": 223 }, { "epoch": 0.024284475281873375, "grad_norm": 5.5431742668151855, "learning_rate": 1.991005367764399e-05, "loss": 0.0594, "step": 224 }, { "epoch": 0.024392888117953166, "grad_norm": 4.8450751304626465, "learning_rate": 1.99093283040766e-05, "loss": 0.0913, "step": 225 }, { "epoch": 0.024501300954032957, "grad_norm": 3.2571122646331787, "learning_rate": 1.9908602930509213e-05, "loss": 0.0574, "step": 226 }, { "epoch": 0.024609713790112748, "grad_norm": 5.23363733291626, "learning_rate": 1.9907877556941825e-05, "loss": 0.1366, "step": 227 }, { "epoch": 0.024718126626192542, "grad_norm": 2.46696138381958, "learning_rate": 1.9907152183374437e-05, "loss": 0.0241, "step": 228 }, { "epoch": 0.024826539462272333, "grad_norm": 2.6102261543273926, "learning_rate": 1.9906426809807053e-05, "loss": 0.0438, "step": 229 }, { "epoch": 0.024934952298352124, "grad_norm": 2.053680419921875, "learning_rate": 1.9905701436239665e-05, "loss": 0.0466, "step": 230 }, { "epoch": 0.02504336513443192, "grad_norm": 4.854499340057373, "learning_rate": 1.9904976062672277e-05, "loss": 0.1035, "step": 231 }, { "epoch": 0.02515177797051171, "grad_norm": 12.94171142578125, "learning_rate": 1.9904250689104893e-05, "loss": 0.0667, "step": 232 }, { "epoch": 0.0252601908065915, "grad_norm": 3.231529474258423, "learning_rate": 1.9903525315537505e-05, "loss": 0.0406, "step": 233 }, { "epoch": 0.02536860364267129, "grad_norm": 8.283382415771484, "learning_rate": 1.9902799941970117e-05, "loss": 0.0554, "step": 234 }, { "epoch": 0.025477016478751086, "grad_norm": 1.4503451585769653, "learning_rate": 1.990207456840273e-05, "loss": 0.026, "step": 235 }, { "epoch": 0.025585429314830876, "grad_norm": 8.720861434936523, "learning_rate": 1.990134919483534e-05, "loss": 0.0736, "step": 236 }, { "epoch": 0.025693842150910667, "grad_norm": 4.253236770629883, "learning_rate": 1.9900623821267953e-05, "loss": 0.0585, "step": 237 }, { "epoch": 0.02580225498699046, "grad_norm": 4.386358737945557, "learning_rate": 1.989989844770057e-05, "loss": 0.0837, "step": 238 }, { "epoch": 0.025910667823070253, "grad_norm": 12.13973617553711, "learning_rate": 1.989917307413318e-05, "loss": 0.1086, "step": 239 }, { "epoch": 0.026019080659150044, "grad_norm": 8.806114196777344, "learning_rate": 1.9898447700565793e-05, "loss": 0.0853, "step": 240 }, { "epoch": 0.026127493495229834, "grad_norm": 1.9392911195755005, "learning_rate": 1.9897722326998405e-05, "loss": 0.0283, "step": 241 }, { "epoch": 0.026235906331309625, "grad_norm": 12.246511459350586, "learning_rate": 1.9896996953431017e-05, "loss": 0.061, "step": 242 }, { "epoch": 0.02634431916738942, "grad_norm": 4.659475326538086, "learning_rate": 1.989627157986363e-05, "loss": 0.0827, "step": 243 }, { "epoch": 0.02645273200346921, "grad_norm": 18.17683982849121, "learning_rate": 1.9895546206296242e-05, "loss": 0.0427, "step": 244 }, { "epoch": 0.026561144839549, "grad_norm": 2.8831889629364014, "learning_rate": 1.9894820832728857e-05, "loss": 0.0586, "step": 245 }, { "epoch": 0.026669557675628796, "grad_norm": 3.7441036701202393, "learning_rate": 1.989409545916147e-05, "loss": 0.0785, "step": 246 }, { "epoch": 0.026777970511708587, "grad_norm": 3.9006521701812744, "learning_rate": 1.9893370085594085e-05, "loss": 0.0518, "step": 247 }, { "epoch": 0.026886383347788378, "grad_norm": 3.6664905548095703, "learning_rate": 1.9892644712026697e-05, "loss": 0.0435, "step": 248 }, { "epoch": 0.02699479618386817, "grad_norm": 14.854068756103516, "learning_rate": 1.989191933845931e-05, "loss": 0.0709, "step": 249 }, { "epoch": 0.027103209019947963, "grad_norm": 3.5867807865142822, "learning_rate": 1.989119396489192e-05, "loss": 0.0463, "step": 250 }, { "epoch": 0.027211621856027754, "grad_norm": 3.8759021759033203, "learning_rate": 1.9890468591324534e-05, "loss": 0.0479, "step": 251 }, { "epoch": 0.027320034692107545, "grad_norm": 3.7771100997924805, "learning_rate": 1.9889743217757146e-05, "loss": 0.0464, "step": 252 }, { "epoch": 0.027428447528187336, "grad_norm": 6.328597545623779, "learning_rate": 1.988901784418976e-05, "loss": 0.0317, "step": 253 }, { "epoch": 0.02753686036426713, "grad_norm": 42.6296501159668, "learning_rate": 1.9888292470622373e-05, "loss": 0.0967, "step": 254 }, { "epoch": 0.02764527320034692, "grad_norm": 10.1552152633667, "learning_rate": 1.9887567097054986e-05, "loss": 0.0665, "step": 255 }, { "epoch": 0.027753686036426712, "grad_norm": 7.1915364265441895, "learning_rate": 1.9886841723487598e-05, "loss": 0.1072, "step": 256 }, { "epoch": 0.027862098872506506, "grad_norm": 10.965987205505371, "learning_rate": 1.988611634992021e-05, "loss": 0.0474, "step": 257 }, { "epoch": 0.027970511708586297, "grad_norm": 16.371854782104492, "learning_rate": 1.9885390976352822e-05, "loss": 0.0562, "step": 258 }, { "epoch": 0.028078924544666088, "grad_norm": 5.853326797485352, "learning_rate": 1.9884665602785434e-05, "loss": 0.1075, "step": 259 }, { "epoch": 0.02818733738074588, "grad_norm": 19.30817222595215, "learning_rate": 1.988394022921805e-05, "loss": 0.1325, "step": 260 }, { "epoch": 0.028295750216825673, "grad_norm": 5.6454901695251465, "learning_rate": 1.9883214855650662e-05, "loss": 0.0669, "step": 261 }, { "epoch": 0.028404163052905464, "grad_norm": 9.086967468261719, "learning_rate": 1.9882489482083274e-05, "loss": 0.0719, "step": 262 }, { "epoch": 0.028512575888985255, "grad_norm": 55.307029724121094, "learning_rate": 1.9881764108515886e-05, "loss": 0.0576, "step": 263 }, { "epoch": 0.028620988725065046, "grad_norm": 3.4429619312286377, "learning_rate": 1.9881038734948498e-05, "loss": 0.0707, "step": 264 }, { "epoch": 0.02872940156114484, "grad_norm": 7.063963413238525, "learning_rate": 1.9880313361381114e-05, "loss": 0.0877, "step": 265 }, { "epoch": 0.02883781439722463, "grad_norm": 3.5789835453033447, "learning_rate": 1.9879587987813726e-05, "loss": 0.0663, "step": 266 }, { "epoch": 0.028946227233304422, "grad_norm": 3.0446345806121826, "learning_rate": 1.9878862614246338e-05, "loss": 0.0643, "step": 267 }, { "epoch": 0.029054640069384217, "grad_norm": 23.91143035888672, "learning_rate": 1.9878137240678954e-05, "loss": 0.0872, "step": 268 }, { "epoch": 0.029163052905464008, "grad_norm": 9.201111793518066, "learning_rate": 1.9877411867111566e-05, "loss": 0.0471, "step": 269 }, { "epoch": 0.0292714657415438, "grad_norm": 2.6925971508026123, "learning_rate": 1.9876686493544178e-05, "loss": 0.0494, "step": 270 }, { "epoch": 0.02937987857762359, "grad_norm": 3.9950802326202393, "learning_rate": 1.987596111997679e-05, "loss": 0.0317, "step": 271 }, { "epoch": 0.029488291413703384, "grad_norm": 4.800086498260498, "learning_rate": 1.9875235746409402e-05, "loss": 0.0478, "step": 272 }, { "epoch": 0.029596704249783175, "grad_norm": 6.9639129638671875, "learning_rate": 1.9874510372842014e-05, "loss": 0.0886, "step": 273 }, { "epoch": 0.029705117085862966, "grad_norm": 14.149821281433105, "learning_rate": 1.9873784999274626e-05, "loss": 0.0791, "step": 274 }, { "epoch": 0.029813529921942757, "grad_norm": 5.788453578948975, "learning_rate": 1.9873059625707242e-05, "loss": 0.0884, "step": 275 }, { "epoch": 0.02992194275802255, "grad_norm": 3.3581254482269287, "learning_rate": 1.9872334252139854e-05, "loss": 0.0804, "step": 276 }, { "epoch": 0.030030355594102342, "grad_norm": 4.915285587310791, "learning_rate": 1.9871608878572466e-05, "loss": 0.1053, "step": 277 }, { "epoch": 0.030138768430182133, "grad_norm": 17.335182189941406, "learning_rate": 1.987088350500508e-05, "loss": 0.0549, "step": 278 }, { "epoch": 0.030247181266261924, "grad_norm": 3.392001152038574, "learning_rate": 1.987015813143769e-05, "loss": 0.0337, "step": 279 }, { "epoch": 0.030355594102341718, "grad_norm": 14.111245155334473, "learning_rate": 1.9869432757870303e-05, "loss": 0.0691, "step": 280 }, { "epoch": 0.03046400693842151, "grad_norm": 9.920124053955078, "learning_rate": 1.9868707384302918e-05, "loss": 0.0617, "step": 281 }, { "epoch": 0.0305724197745013, "grad_norm": 6.4663286209106445, "learning_rate": 1.986798201073553e-05, "loss": 0.1038, "step": 282 }, { "epoch": 0.030680832610581094, "grad_norm": 10.873244285583496, "learning_rate": 1.9867256637168142e-05, "loss": 0.0903, "step": 283 }, { "epoch": 0.030789245446660885, "grad_norm": 3.376325845718384, "learning_rate": 1.9866531263600758e-05, "loss": 0.0622, "step": 284 }, { "epoch": 0.030897658282740676, "grad_norm": 6.031406879425049, "learning_rate": 1.986580589003337e-05, "loss": 0.114, "step": 285 }, { "epoch": 0.031006071118820467, "grad_norm": 4.664947986602783, "learning_rate": 1.9865080516465982e-05, "loss": 0.0412, "step": 286 }, { "epoch": 0.03111448395490026, "grad_norm": 7.018142223358154, "learning_rate": 1.9864355142898594e-05, "loss": 0.0759, "step": 287 }, { "epoch": 0.031222896790980052, "grad_norm": 4.504680156707764, "learning_rate": 1.9863629769331207e-05, "loss": 0.0569, "step": 288 }, { "epoch": 0.03133130962705984, "grad_norm": 3.7086575031280518, "learning_rate": 1.986290439576382e-05, "loss": 0.0791, "step": 289 }, { "epoch": 0.031439722463139634, "grad_norm": 2.1956300735473633, "learning_rate": 1.9862179022196434e-05, "loss": 0.0513, "step": 290 }, { "epoch": 0.031548135299219425, "grad_norm": 3.9495577812194824, "learning_rate": 1.9861453648629046e-05, "loss": 0.0623, "step": 291 }, { "epoch": 0.03165654813529922, "grad_norm": 4.02662467956543, "learning_rate": 1.986072827506166e-05, "loss": 0.0304, "step": 292 }, { "epoch": 0.031764960971379014, "grad_norm": 5.850906848907471, "learning_rate": 1.986000290149427e-05, "loss": 0.0771, "step": 293 }, { "epoch": 0.031873373807458805, "grad_norm": 9.47220516204834, "learning_rate": 1.9859277527926883e-05, "loss": 0.0558, "step": 294 }, { "epoch": 0.031981786643538596, "grad_norm": 6.1368279457092285, "learning_rate": 1.9858552154359495e-05, "loss": 0.0857, "step": 295 }, { "epoch": 0.032090199479618386, "grad_norm": 2.3039047718048096, "learning_rate": 1.9857826780792107e-05, "loss": 0.071, "step": 296 }, { "epoch": 0.03219861231569818, "grad_norm": 3.4956159591674805, "learning_rate": 1.9857101407224723e-05, "loss": 0.0597, "step": 297 }, { "epoch": 0.03230702515177797, "grad_norm": 170.1815948486328, "learning_rate": 1.9856376033657335e-05, "loss": 0.0837, "step": 298 }, { "epoch": 0.03241543798785776, "grad_norm": 2.101958990097046, "learning_rate": 1.985565066008995e-05, "loss": 0.0499, "step": 299 }, { "epoch": 0.03252385082393756, "grad_norm": 5.171942710876465, "learning_rate": 1.9854925286522562e-05, "loss": 0.0637, "step": 300 }, { "epoch": 0.03263226366001735, "grad_norm": 5.612536907196045, "learning_rate": 1.9854199912955175e-05, "loss": 0.0375, "step": 301 }, { "epoch": 0.03274067649609714, "grad_norm": 3.1455330848693848, "learning_rate": 1.9853474539387787e-05, "loss": 0.0776, "step": 302 }, { "epoch": 0.03284908933217693, "grad_norm": 2.2144346237182617, "learning_rate": 1.98527491658204e-05, "loss": 0.0213, "step": 303 }, { "epoch": 0.03295750216825672, "grad_norm": 9.37685489654541, "learning_rate": 1.985202379225301e-05, "loss": 0.0663, "step": 304 }, { "epoch": 0.03306591500433651, "grad_norm": 8.454275131225586, "learning_rate": 1.9851298418685627e-05, "loss": 0.0321, "step": 305 }, { "epoch": 0.0331743278404163, "grad_norm": 3.81746506690979, "learning_rate": 1.985057304511824e-05, "loss": 0.0382, "step": 306 }, { "epoch": 0.0332827406764961, "grad_norm": 5.48844575881958, "learning_rate": 1.984984767155085e-05, "loss": 0.0732, "step": 307 }, { "epoch": 0.03339115351257589, "grad_norm": 20.15005111694336, "learning_rate": 1.9849122297983463e-05, "loss": 0.0625, "step": 308 }, { "epoch": 0.03349956634865568, "grad_norm": 3.938211441040039, "learning_rate": 1.9848396924416075e-05, "loss": 0.0418, "step": 309 }, { "epoch": 0.03360797918473547, "grad_norm": 3.8580572605133057, "learning_rate": 1.9847671550848687e-05, "loss": 0.0461, "step": 310 }, { "epoch": 0.033716392020815264, "grad_norm": 8.98977279663086, "learning_rate": 1.98469461772813e-05, "loss": 0.0747, "step": 311 }, { "epoch": 0.033824804856895055, "grad_norm": 9.641656875610352, "learning_rate": 1.9846220803713915e-05, "loss": 0.1041, "step": 312 }, { "epoch": 0.033933217692974846, "grad_norm": 10.642899513244629, "learning_rate": 1.9845495430146527e-05, "loss": 0.0507, "step": 313 }, { "epoch": 0.03404163052905464, "grad_norm": 8.088515281677246, "learning_rate": 1.984477005657914e-05, "loss": 0.1331, "step": 314 }, { "epoch": 0.034150043365134435, "grad_norm": 8.807518005371094, "learning_rate": 1.984404468301175e-05, "loss": 0.0826, "step": 315 }, { "epoch": 0.034258456201214225, "grad_norm": 4.8934326171875, "learning_rate": 1.9843319309444367e-05, "loss": 0.0337, "step": 316 }, { "epoch": 0.034366869037294016, "grad_norm": 33.26197814941406, "learning_rate": 1.984259393587698e-05, "loss": 0.1206, "step": 317 }, { "epoch": 0.03447528187337381, "grad_norm": 4.948579788208008, "learning_rate": 1.984186856230959e-05, "loss": 0.0388, "step": 318 }, { "epoch": 0.0345836947094536, "grad_norm": 4.677768230438232, "learning_rate": 1.9841143188742203e-05, "loss": 0.0508, "step": 319 }, { "epoch": 0.03469210754553339, "grad_norm": 1.509756088256836, "learning_rate": 1.984041781517482e-05, "loss": 0.0327, "step": 320 }, { "epoch": 0.03480052038161318, "grad_norm": 3.588665008544922, "learning_rate": 1.983969244160743e-05, "loss": 0.0424, "step": 321 }, { "epoch": 0.03490893321769298, "grad_norm": 5.868305683135986, "learning_rate": 1.9838967068040043e-05, "loss": 0.0694, "step": 322 }, { "epoch": 0.03501734605377277, "grad_norm": 1.4083083868026733, "learning_rate": 1.9838241694472655e-05, "loss": 0.0275, "step": 323 }, { "epoch": 0.03512575888985256, "grad_norm": 6.810021877288818, "learning_rate": 1.9837516320905267e-05, "loss": 0.0269, "step": 324 }, { "epoch": 0.03523417172593235, "grad_norm": 9.817872047424316, "learning_rate": 1.983679094733788e-05, "loss": 0.0887, "step": 325 }, { "epoch": 0.03534258456201214, "grad_norm": 5.257312297821045, "learning_rate": 1.9836065573770492e-05, "loss": 0.0832, "step": 326 }, { "epoch": 0.03545099739809193, "grad_norm": 2.448499917984009, "learning_rate": 1.9835340200203107e-05, "loss": 0.0648, "step": 327 }, { "epoch": 0.03555941023417172, "grad_norm": 5.345430374145508, "learning_rate": 1.983461482663572e-05, "loss": 0.0838, "step": 328 }, { "epoch": 0.03566782307025152, "grad_norm": 8.000154495239258, "learning_rate": 1.983388945306833e-05, "loss": 0.087, "step": 329 }, { "epoch": 0.03577623590633131, "grad_norm": 6.124927520751953, "learning_rate": 1.9833164079500944e-05, "loss": 0.0548, "step": 330 }, { "epoch": 0.0358846487424111, "grad_norm": 2.438791275024414, "learning_rate": 1.9832438705933556e-05, "loss": 0.024, "step": 331 }, { "epoch": 0.035993061578490894, "grad_norm": 2.835667371749878, "learning_rate": 1.9831713332366168e-05, "loss": 0.0581, "step": 332 }, { "epoch": 0.036101474414570685, "grad_norm": 3.7697689533233643, "learning_rate": 1.9830987958798784e-05, "loss": 0.0695, "step": 333 }, { "epoch": 0.036209887250650476, "grad_norm": 3.7412467002868652, "learning_rate": 1.9830262585231396e-05, "loss": 0.0712, "step": 334 }, { "epoch": 0.03631830008673027, "grad_norm": 107.2884292602539, "learning_rate": 1.9829537211664008e-05, "loss": 0.0468, "step": 335 }, { "epoch": 0.03642671292281006, "grad_norm": 5.656260967254639, "learning_rate": 1.9828811838096623e-05, "loss": 0.067, "step": 336 }, { "epoch": 0.036535125758889855, "grad_norm": 1.4446452856063843, "learning_rate": 1.9828086464529235e-05, "loss": 0.0295, "step": 337 }, { "epoch": 0.036643538594969646, "grad_norm": 4.474142074584961, "learning_rate": 1.9827361090961848e-05, "loss": 0.0757, "step": 338 }, { "epoch": 0.03675195143104944, "grad_norm": 3.2766568660736084, "learning_rate": 1.982663571739446e-05, "loss": 0.0313, "step": 339 }, { "epoch": 0.03686036426712923, "grad_norm": 2.461292028427124, "learning_rate": 1.9825910343827072e-05, "loss": 0.0307, "step": 340 }, { "epoch": 0.03696877710320902, "grad_norm": 7.269660949707031, "learning_rate": 1.9825184970259684e-05, "loss": 0.0939, "step": 341 }, { "epoch": 0.03707718993928881, "grad_norm": 9.260869026184082, "learning_rate": 1.98244595966923e-05, "loss": 0.0262, "step": 342 }, { "epoch": 0.0371856027753686, "grad_norm": 6.417819023132324, "learning_rate": 1.9823734223124912e-05, "loss": 0.1063, "step": 343 }, { "epoch": 0.0372940156114484, "grad_norm": 5.338077545166016, "learning_rate": 1.9823008849557524e-05, "loss": 0.0376, "step": 344 }, { "epoch": 0.03740242844752819, "grad_norm": 5.423496723175049, "learning_rate": 1.9822283475990136e-05, "loss": 0.0793, "step": 345 }, { "epoch": 0.03751084128360798, "grad_norm": 10.714667320251465, "learning_rate": 1.9821558102422748e-05, "loss": 0.091, "step": 346 }, { "epoch": 0.03761925411968777, "grad_norm": 5.531905174255371, "learning_rate": 1.982083272885536e-05, "loss": 0.0579, "step": 347 }, { "epoch": 0.03772766695576756, "grad_norm": 4.275395393371582, "learning_rate": 1.9820107355287976e-05, "loss": 0.0713, "step": 348 }, { "epoch": 0.03783607979184735, "grad_norm": 3.7061495780944824, "learning_rate": 1.9819381981720588e-05, "loss": 0.065, "step": 349 }, { "epoch": 0.037944492627927144, "grad_norm": 2.967381000518799, "learning_rate": 1.98186566081532e-05, "loss": 0.0429, "step": 350 }, { "epoch": 0.038052905464006935, "grad_norm": 5.098580360412598, "learning_rate": 1.9817931234585816e-05, "loss": 0.0557, "step": 351 }, { "epoch": 0.03816131830008673, "grad_norm": 2.673112154006958, "learning_rate": 1.9817205861018428e-05, "loss": 0.0269, "step": 352 }, { "epoch": 0.038269731136166524, "grad_norm": 5.341294765472412, "learning_rate": 1.981648048745104e-05, "loss": 0.0601, "step": 353 }, { "epoch": 0.038378143972246315, "grad_norm": 3.012857437133789, "learning_rate": 1.9815755113883652e-05, "loss": 0.0308, "step": 354 }, { "epoch": 0.038486556808326106, "grad_norm": 4.584516525268555, "learning_rate": 1.9815029740316264e-05, "loss": 0.0749, "step": 355 }, { "epoch": 0.038594969644405897, "grad_norm": 2.5888590812683105, "learning_rate": 1.9814304366748876e-05, "loss": 0.0382, "step": 356 }, { "epoch": 0.03870338248048569, "grad_norm": 6.214788436889648, "learning_rate": 1.9813578993181492e-05, "loss": 0.0725, "step": 357 }, { "epoch": 0.03881179531656548, "grad_norm": 6.987221717834473, "learning_rate": 1.9812853619614104e-05, "loss": 0.1406, "step": 358 }, { "epoch": 0.038920208152645276, "grad_norm": 3.4147772789001465, "learning_rate": 1.9812128246046716e-05, "loss": 0.0486, "step": 359 }, { "epoch": 0.03902862098872507, "grad_norm": 3.715047597885132, "learning_rate": 1.981140287247933e-05, "loss": 0.0358, "step": 360 }, { "epoch": 0.03913703382480486, "grad_norm": 7.043151378631592, "learning_rate": 1.981067749891194e-05, "loss": 0.069, "step": 361 }, { "epoch": 0.03924544666088465, "grad_norm": 3.2895452976226807, "learning_rate": 1.9809952125344553e-05, "loss": 0.0274, "step": 362 }, { "epoch": 0.03935385949696444, "grad_norm": 1.9196919202804565, "learning_rate": 1.9809226751777165e-05, "loss": 0.0524, "step": 363 }, { "epoch": 0.03946227233304423, "grad_norm": 3.510681390762329, "learning_rate": 1.980850137820978e-05, "loss": 0.0637, "step": 364 }, { "epoch": 0.03957068516912402, "grad_norm": 7.1121697425842285, "learning_rate": 1.9807776004642392e-05, "loss": 0.0816, "step": 365 }, { "epoch": 0.03967909800520382, "grad_norm": 3.8429999351501465, "learning_rate": 1.9807050631075005e-05, "loss": 0.0556, "step": 366 }, { "epoch": 0.03978751084128361, "grad_norm": 7.1921210289001465, "learning_rate": 1.9806325257507617e-05, "loss": 0.0933, "step": 367 }, { "epoch": 0.0398959236773634, "grad_norm": 3.2833120822906494, "learning_rate": 1.9805599883940232e-05, "loss": 0.054, "step": 368 }, { "epoch": 0.04000433651344319, "grad_norm": 2.458770275115967, "learning_rate": 1.9804874510372844e-05, "loss": 0.0608, "step": 369 }, { "epoch": 0.04011274934952298, "grad_norm": 3.903201103210449, "learning_rate": 1.9804149136805457e-05, "loss": 0.0275, "step": 370 }, { "epoch": 0.040221162185602774, "grad_norm": 5.030633449554443, "learning_rate": 1.980342376323807e-05, "loss": 0.0428, "step": 371 }, { "epoch": 0.040329575021682565, "grad_norm": 11.119414329528809, "learning_rate": 1.9802698389670684e-05, "loss": 0.0498, "step": 372 }, { "epoch": 0.040437987857762356, "grad_norm": 2.5223538875579834, "learning_rate": 1.9801973016103296e-05, "loss": 0.0487, "step": 373 }, { "epoch": 0.040546400693842154, "grad_norm": 2.412189245223999, "learning_rate": 1.980124764253591e-05, "loss": 0.0262, "step": 374 }, { "epoch": 0.040654813529921945, "grad_norm": 9.408064842224121, "learning_rate": 1.980052226896852e-05, "loss": 0.0984, "step": 375 }, { "epoch": 0.040763226366001735, "grad_norm": 2.651369094848633, "learning_rate": 1.9799796895401133e-05, "loss": 0.0455, "step": 376 }, { "epoch": 0.040871639202081526, "grad_norm": 5.555140972137451, "learning_rate": 1.9799071521833745e-05, "loss": 0.0649, "step": 377 }, { "epoch": 0.04098005203816132, "grad_norm": 5.125089168548584, "learning_rate": 1.9798346148266357e-05, "loss": 0.0532, "step": 378 }, { "epoch": 0.04108846487424111, "grad_norm": 1.8245196342468262, "learning_rate": 1.9797620774698973e-05, "loss": 0.0221, "step": 379 }, { "epoch": 0.0411968777103209, "grad_norm": 4.621825695037842, "learning_rate": 1.9796895401131585e-05, "loss": 0.0801, "step": 380 }, { "epoch": 0.0413052905464007, "grad_norm": 6.772396564483643, "learning_rate": 1.9796170027564197e-05, "loss": 0.0438, "step": 381 }, { "epoch": 0.04141370338248049, "grad_norm": 5.054163932800293, "learning_rate": 1.979544465399681e-05, "loss": 0.0568, "step": 382 }, { "epoch": 0.04152211621856028, "grad_norm": 4.519815921783447, "learning_rate": 1.979471928042942e-05, "loss": 0.0594, "step": 383 }, { "epoch": 0.04163052905464007, "grad_norm": 3.4887712001800537, "learning_rate": 1.9793993906862033e-05, "loss": 0.0309, "step": 384 }, { "epoch": 0.04173894189071986, "grad_norm": 5.960877418518066, "learning_rate": 1.979326853329465e-05, "loss": 0.0547, "step": 385 }, { "epoch": 0.04184735472679965, "grad_norm": 4.736602783203125, "learning_rate": 1.979254315972726e-05, "loss": 0.0403, "step": 386 }, { "epoch": 0.04195576756287944, "grad_norm": 2.9024925231933594, "learning_rate": 1.9791817786159873e-05, "loss": 0.0258, "step": 387 }, { "epoch": 0.04206418039895923, "grad_norm": 2.1538991928100586, "learning_rate": 1.979109241259249e-05, "loss": 0.0209, "step": 388 }, { "epoch": 0.04217259323503903, "grad_norm": 8.763294219970703, "learning_rate": 1.97903670390251e-05, "loss": 0.0718, "step": 389 }, { "epoch": 0.04228100607111882, "grad_norm": 3.4436120986938477, "learning_rate": 1.9789641665457713e-05, "loss": 0.025, "step": 390 }, { "epoch": 0.04238941890719861, "grad_norm": 6.723008155822754, "learning_rate": 1.9788916291890325e-05, "loss": 0.0424, "step": 391 }, { "epoch": 0.042497831743278404, "grad_norm": 4.326019763946533, "learning_rate": 1.9788190918322937e-05, "loss": 0.0292, "step": 392 }, { "epoch": 0.042606244579358195, "grad_norm": 9.598688125610352, "learning_rate": 1.978746554475555e-05, "loss": 0.0751, "step": 393 }, { "epoch": 0.042714657415437986, "grad_norm": 9.677236557006836, "learning_rate": 1.9786740171188165e-05, "loss": 0.1146, "step": 394 }, { "epoch": 0.04282307025151778, "grad_norm": 11.748144149780273, "learning_rate": 1.9786014797620777e-05, "loss": 0.0925, "step": 395 }, { "epoch": 0.042931483087597574, "grad_norm": 9.774069786071777, "learning_rate": 1.978528942405339e-05, "loss": 0.1424, "step": 396 }, { "epoch": 0.043039895923677365, "grad_norm": 8.127826690673828, "learning_rate": 1.9784564050486e-05, "loss": 0.084, "step": 397 }, { "epoch": 0.043148308759757156, "grad_norm": 9.280285835266113, "learning_rate": 1.9783838676918613e-05, "loss": 0.0428, "step": 398 }, { "epoch": 0.04325672159583695, "grad_norm": 4.135801792144775, "learning_rate": 1.9783113303351226e-05, "loss": 0.0646, "step": 399 }, { "epoch": 0.04336513443191674, "grad_norm": 4.154670715332031, "learning_rate": 1.978238792978384e-05, "loss": 0.1007, "step": 400 }, { "epoch": 0.04347354726799653, "grad_norm": 3.1119794845581055, "learning_rate": 1.9781662556216453e-05, "loss": 0.0323, "step": 401 }, { "epoch": 0.04358196010407632, "grad_norm": 1.1979129314422607, "learning_rate": 1.9780937182649065e-05, "loss": 0.0165, "step": 402 }, { "epoch": 0.04369037294015612, "grad_norm": 3.644775390625, "learning_rate": 1.978021180908168e-05, "loss": 0.0545, "step": 403 }, { "epoch": 0.04379878577623591, "grad_norm": 3.353353500366211, "learning_rate": 1.9779486435514293e-05, "loss": 0.0902, "step": 404 }, { "epoch": 0.0439071986123157, "grad_norm": 4.88148832321167, "learning_rate": 1.9778761061946905e-05, "loss": 0.0722, "step": 405 }, { "epoch": 0.04401561144839549, "grad_norm": 2.758671522140503, "learning_rate": 1.9778035688379517e-05, "loss": 0.0601, "step": 406 }, { "epoch": 0.04412402428447528, "grad_norm": 2.5628888607025146, "learning_rate": 1.977731031481213e-05, "loss": 0.0618, "step": 407 }, { "epoch": 0.04423243712055507, "grad_norm": 1.5323463678359985, "learning_rate": 1.9776584941244742e-05, "loss": 0.0257, "step": 408 }, { "epoch": 0.04434084995663486, "grad_norm": 5.035964488983154, "learning_rate": 1.9775859567677357e-05, "loss": 0.0834, "step": 409 }, { "epoch": 0.044449262792714654, "grad_norm": 1.704378604888916, "learning_rate": 1.977513419410997e-05, "loss": 0.0299, "step": 410 }, { "epoch": 0.04455767562879445, "grad_norm": 2.9562039375305176, "learning_rate": 1.977440882054258e-05, "loss": 0.0959, "step": 411 }, { "epoch": 0.04466608846487424, "grad_norm": 5.685033798217773, "learning_rate": 1.9773683446975194e-05, "loss": 0.0681, "step": 412 }, { "epoch": 0.044774501300954034, "grad_norm": 2.814514398574829, "learning_rate": 1.9772958073407806e-05, "loss": 0.0611, "step": 413 }, { "epoch": 0.044882914137033825, "grad_norm": 2.333137273788452, "learning_rate": 1.9772232699840418e-05, "loss": 0.0604, "step": 414 }, { "epoch": 0.044991326973113616, "grad_norm": 2.072645664215088, "learning_rate": 1.977150732627303e-05, "loss": 0.0188, "step": 415 }, { "epoch": 0.045099739809193407, "grad_norm": 5.337463855743408, "learning_rate": 1.9770781952705646e-05, "loss": 0.0976, "step": 416 }, { "epoch": 0.0452081526452732, "grad_norm": 3.384761095046997, "learning_rate": 1.9770056579138258e-05, "loss": 0.0422, "step": 417 }, { "epoch": 0.045316565481352995, "grad_norm": 7.837184906005859, "learning_rate": 1.976933120557087e-05, "loss": 0.0535, "step": 418 }, { "epoch": 0.045424978317432786, "grad_norm": 8.27505874633789, "learning_rate": 1.9768605832003482e-05, "loss": 0.0556, "step": 419 }, { "epoch": 0.04553339115351258, "grad_norm": 3.692570924758911, "learning_rate": 1.9767880458436098e-05, "loss": 0.0756, "step": 420 }, { "epoch": 0.04564180398959237, "grad_norm": 2.7941696643829346, "learning_rate": 1.976715508486871e-05, "loss": 0.0768, "step": 421 }, { "epoch": 0.04575021682567216, "grad_norm": 3.013108015060425, "learning_rate": 1.9766429711301322e-05, "loss": 0.0247, "step": 422 }, { "epoch": 0.04585862966175195, "grad_norm": 7.621976852416992, "learning_rate": 1.9765704337733934e-05, "loss": 0.0696, "step": 423 }, { "epoch": 0.04596704249783174, "grad_norm": 4.811153888702393, "learning_rate": 1.976497896416655e-05, "loss": 0.0765, "step": 424 }, { "epoch": 0.04607545533391154, "grad_norm": 3.4611642360687256, "learning_rate": 1.976425359059916e-05, "loss": 0.0294, "step": 425 }, { "epoch": 0.04618386816999133, "grad_norm": 2.7776541709899902, "learning_rate": 1.9763528217031774e-05, "loss": 0.0425, "step": 426 }, { "epoch": 0.04629228100607112, "grad_norm": 2.638298511505127, "learning_rate": 1.9762802843464386e-05, "loss": 0.0482, "step": 427 }, { "epoch": 0.04640069384215091, "grad_norm": 5.0102858543396, "learning_rate": 1.9762077469896998e-05, "loss": 0.0693, "step": 428 }, { "epoch": 0.0465091066782307, "grad_norm": 6.306221008300781, "learning_rate": 1.976135209632961e-05, "loss": 0.0761, "step": 429 }, { "epoch": 0.04661751951431049, "grad_norm": 2.1465020179748535, "learning_rate": 1.9760626722762222e-05, "loss": 0.0371, "step": 430 }, { "epoch": 0.046725932350390284, "grad_norm": 7.360186576843262, "learning_rate": 1.9759901349194838e-05, "loss": 0.1096, "step": 431 }, { "epoch": 0.046834345186470075, "grad_norm": 8.676478385925293, "learning_rate": 1.975917597562745e-05, "loss": 0.1153, "step": 432 }, { "epoch": 0.04694275802254987, "grad_norm": 13.683759689331055, "learning_rate": 1.9758450602060062e-05, "loss": 0.046, "step": 433 }, { "epoch": 0.047051170858629664, "grad_norm": 7.267298221588135, "learning_rate": 1.9757725228492674e-05, "loss": 0.0574, "step": 434 }, { "epoch": 0.047159583694709455, "grad_norm": 3.0183215141296387, "learning_rate": 1.9756999854925287e-05, "loss": 0.0298, "step": 435 }, { "epoch": 0.047267996530789246, "grad_norm": 5.044855117797852, "learning_rate": 1.97562744813579e-05, "loss": 0.0593, "step": 436 }, { "epoch": 0.047376409366869036, "grad_norm": 2.617769479751587, "learning_rate": 1.9755549107790514e-05, "loss": 0.0345, "step": 437 }, { "epoch": 0.04748482220294883, "grad_norm": 3.7331905364990234, "learning_rate": 1.9754823734223126e-05, "loss": 0.0329, "step": 438 }, { "epoch": 0.04759323503902862, "grad_norm": 6.799117088317871, "learning_rate": 1.975409836065574e-05, "loss": 0.0616, "step": 439 }, { "epoch": 0.047701647875108416, "grad_norm": 2.917686939239502, "learning_rate": 1.9753372987088354e-05, "loss": 0.0543, "step": 440 }, { "epoch": 0.04781006071118821, "grad_norm": 7.316773891448975, "learning_rate": 1.9752647613520966e-05, "loss": 0.1133, "step": 441 }, { "epoch": 0.047918473547268, "grad_norm": 2.055525541305542, "learning_rate": 1.9751922239953578e-05, "loss": 0.0117, "step": 442 }, { "epoch": 0.04802688638334779, "grad_norm": 2.569819450378418, "learning_rate": 1.975119686638619e-05, "loss": 0.0258, "step": 443 }, { "epoch": 0.04813529921942758, "grad_norm": 5.565124034881592, "learning_rate": 1.9750471492818803e-05, "loss": 0.0403, "step": 444 }, { "epoch": 0.04824371205550737, "grad_norm": 7.352942943572998, "learning_rate": 1.9749746119251415e-05, "loss": 0.1041, "step": 445 }, { "epoch": 0.04835212489158716, "grad_norm": 8.025023460388184, "learning_rate": 1.974902074568403e-05, "loss": 0.0733, "step": 446 }, { "epoch": 0.04846053772766695, "grad_norm": 5.223968029022217, "learning_rate": 1.9748295372116642e-05, "loss": 0.0901, "step": 447 }, { "epoch": 0.04856895056374675, "grad_norm": 4.169132709503174, "learning_rate": 1.9747569998549255e-05, "loss": 0.0802, "step": 448 }, { "epoch": 0.04867736339982654, "grad_norm": 9.524415016174316, "learning_rate": 1.9746844624981867e-05, "loss": 0.0699, "step": 449 }, { "epoch": 0.04878577623590633, "grad_norm": 0.5994383096694946, "learning_rate": 1.974611925141448e-05, "loss": 0.0076, "step": 450 }, { "epoch": 0.04889418907198612, "grad_norm": 15.974268913269043, "learning_rate": 1.974539387784709e-05, "loss": 0.0926, "step": 451 }, { "epoch": 0.049002601908065914, "grad_norm": 3.79998517036438, "learning_rate": 1.9744668504279707e-05, "loss": 0.0587, "step": 452 }, { "epoch": 0.049111014744145705, "grad_norm": 3.387145519256592, "learning_rate": 1.974394313071232e-05, "loss": 0.0863, "step": 453 }, { "epoch": 0.049219427580225496, "grad_norm": 5.620550155639648, "learning_rate": 1.974321775714493e-05, "loss": 0.0579, "step": 454 }, { "epoch": 0.049327840416305294, "grad_norm": 5.883560657501221, "learning_rate": 1.9742492383577546e-05, "loss": 0.0715, "step": 455 }, { "epoch": 0.049436253252385085, "grad_norm": 1.3067091703414917, "learning_rate": 1.974176701001016e-05, "loss": 0.0325, "step": 456 }, { "epoch": 0.049544666088464875, "grad_norm": 6.5685625076293945, "learning_rate": 1.974104163644277e-05, "loss": 0.0625, "step": 457 }, { "epoch": 0.049653078924544666, "grad_norm": 6.018320083618164, "learning_rate": 1.9740316262875383e-05, "loss": 0.0608, "step": 458 }, { "epoch": 0.04976149176062446, "grad_norm": 3.3304102420806885, "learning_rate": 1.9739590889307995e-05, "loss": 0.0435, "step": 459 }, { "epoch": 0.04986990459670425, "grad_norm": 5.063429355621338, "learning_rate": 1.9738865515740607e-05, "loss": 0.0555, "step": 460 }, { "epoch": 0.04997831743278404, "grad_norm": 5.78240966796875, "learning_rate": 1.9738140142173223e-05, "loss": 0.0606, "step": 461 }, { "epoch": 0.05008673026886384, "grad_norm": 4.324882507324219, "learning_rate": 1.9737414768605835e-05, "loss": 0.0345, "step": 462 }, { "epoch": 0.05019514310494363, "grad_norm": 6.255583763122559, "learning_rate": 1.9736689395038447e-05, "loss": 0.1117, "step": 463 }, { "epoch": 0.05030355594102342, "grad_norm": 1.147409439086914, "learning_rate": 1.973596402147106e-05, "loss": 0.0358, "step": 464 }, { "epoch": 0.05041196877710321, "grad_norm": 4.4263105392456055, "learning_rate": 1.973523864790367e-05, "loss": 0.0425, "step": 465 }, { "epoch": 0.050520381613183, "grad_norm": 4.60640811920166, "learning_rate": 1.9734513274336283e-05, "loss": 0.0669, "step": 466 }, { "epoch": 0.05062879444926279, "grad_norm": 1.5139434337615967, "learning_rate": 1.9733787900768895e-05, "loss": 0.0139, "step": 467 }, { "epoch": 0.05073720728534258, "grad_norm": 3.9987456798553467, "learning_rate": 1.973306252720151e-05, "loss": 0.0271, "step": 468 }, { "epoch": 0.05084562012142237, "grad_norm": 3.6153066158294678, "learning_rate": 1.9732337153634123e-05, "loss": 0.0387, "step": 469 }, { "epoch": 0.05095403295750217, "grad_norm": 11.281123161315918, "learning_rate": 1.9731611780066735e-05, "loss": 0.1276, "step": 470 }, { "epoch": 0.05106244579358196, "grad_norm": 1.2442491054534912, "learning_rate": 1.9730886406499347e-05, "loss": 0.0173, "step": 471 }, { "epoch": 0.05117085862966175, "grad_norm": 4.804673194885254, "learning_rate": 1.9730161032931963e-05, "loss": 0.0931, "step": 472 }, { "epoch": 0.051279271465741544, "grad_norm": 6.816423416137695, "learning_rate": 1.9729435659364575e-05, "loss": 0.1005, "step": 473 }, { "epoch": 0.051387684301821335, "grad_norm": 3.737102508544922, "learning_rate": 1.9728710285797187e-05, "loss": 0.0478, "step": 474 }, { "epoch": 0.051496097137901126, "grad_norm": 10.502685546875, "learning_rate": 1.97279849122298e-05, "loss": 0.0853, "step": 475 }, { "epoch": 0.05160450997398092, "grad_norm": 4.890417575836182, "learning_rate": 1.9727259538662415e-05, "loss": 0.0221, "step": 476 }, { "epoch": 0.051712922810060714, "grad_norm": 2.127261161804199, "learning_rate": 1.9726534165095027e-05, "loss": 0.0474, "step": 477 }, { "epoch": 0.051821335646140505, "grad_norm": 3.86618709564209, "learning_rate": 1.972580879152764e-05, "loss": 0.0441, "step": 478 }, { "epoch": 0.051929748482220296, "grad_norm": 6.268828868865967, "learning_rate": 1.972508341796025e-05, "loss": 0.1261, "step": 479 }, { "epoch": 0.05203816131830009, "grad_norm": 5.573505878448486, "learning_rate": 1.9724358044392863e-05, "loss": 0.0874, "step": 480 }, { "epoch": 0.05214657415437988, "grad_norm": 2.3709282875061035, "learning_rate": 1.9723632670825476e-05, "loss": 0.0125, "step": 481 }, { "epoch": 0.05225498699045967, "grad_norm": 6.641480445861816, "learning_rate": 1.9722907297258088e-05, "loss": 0.1332, "step": 482 }, { "epoch": 0.05236339982653946, "grad_norm": 6.1242594718933105, "learning_rate": 1.9722181923690703e-05, "loss": 0.0311, "step": 483 }, { "epoch": 0.05247181266261925, "grad_norm": 3.8848891258239746, "learning_rate": 1.9721456550123315e-05, "loss": 0.078, "step": 484 }, { "epoch": 0.05258022549869905, "grad_norm": 4.603445529937744, "learning_rate": 1.9720731176555928e-05, "loss": 0.0468, "step": 485 }, { "epoch": 0.05268863833477884, "grad_norm": 8.149454116821289, "learning_rate": 1.972000580298854e-05, "loss": 0.0937, "step": 486 }, { "epoch": 0.05279705117085863, "grad_norm": 4.274590015411377, "learning_rate": 1.9719280429421152e-05, "loss": 0.0524, "step": 487 }, { "epoch": 0.05290546400693842, "grad_norm": 6.9618425369262695, "learning_rate": 1.9718555055853764e-05, "loss": 0.0796, "step": 488 }, { "epoch": 0.05301387684301821, "grad_norm": 2.2986326217651367, "learning_rate": 1.971782968228638e-05, "loss": 0.0724, "step": 489 }, { "epoch": 0.053122289679098, "grad_norm": 4.903386116027832, "learning_rate": 1.971710430871899e-05, "loss": 0.1104, "step": 490 }, { "epoch": 0.053230702515177794, "grad_norm": 2.072528600692749, "learning_rate": 1.9716378935151604e-05, "loss": 0.0361, "step": 491 }, { "epoch": 0.05333911535125759, "grad_norm": 2.9829540252685547, "learning_rate": 1.971565356158422e-05, "loss": 0.0507, "step": 492 }, { "epoch": 0.05344752818733738, "grad_norm": 1.5802043676376343, "learning_rate": 1.971492818801683e-05, "loss": 0.0452, "step": 493 }, { "epoch": 0.053555941023417174, "grad_norm": 4.537359714508057, "learning_rate": 1.9714202814449444e-05, "loss": 0.0635, "step": 494 }, { "epoch": 0.053664353859496965, "grad_norm": 1.8938134908676147, "learning_rate": 1.9713477440882056e-05, "loss": 0.069, "step": 495 }, { "epoch": 0.053772766695576756, "grad_norm": 1.6262949705123901, "learning_rate": 1.9712752067314668e-05, "loss": 0.0484, "step": 496 }, { "epoch": 0.053881179531656546, "grad_norm": 1.8777375221252441, "learning_rate": 1.971202669374728e-05, "loss": 0.0417, "step": 497 }, { "epoch": 0.05398959236773634, "grad_norm": 3.798391342163086, "learning_rate": 1.9711301320179896e-05, "loss": 0.0458, "step": 498 }, { "epoch": 0.054098005203816135, "grad_norm": 6.934033393859863, "learning_rate": 1.9710575946612508e-05, "loss": 0.0519, "step": 499 }, { "epoch": 0.054206418039895926, "grad_norm": 5.121609687805176, "learning_rate": 1.970985057304512e-05, "loss": 0.0485, "step": 500 }, { "epoch": 0.05431483087597572, "grad_norm": 4.28605318069458, "learning_rate": 1.9709125199477732e-05, "loss": 0.0527, "step": 501 }, { "epoch": 0.05442324371205551, "grad_norm": 3.3651320934295654, "learning_rate": 1.9708399825910344e-05, "loss": 0.0761, "step": 502 }, { "epoch": 0.0545316565481353, "grad_norm": 4.2114410400390625, "learning_rate": 1.9707674452342956e-05, "loss": 0.0499, "step": 503 }, { "epoch": 0.05464006938421509, "grad_norm": 5.572483539581299, "learning_rate": 1.9706949078775572e-05, "loss": 0.0498, "step": 504 }, { "epoch": 0.05474848222029488, "grad_norm": 1.4925713539123535, "learning_rate": 1.9706223705208184e-05, "loss": 0.0149, "step": 505 }, { "epoch": 0.05485689505637467, "grad_norm": 6.853404521942139, "learning_rate": 1.9705498331640796e-05, "loss": 0.0701, "step": 506 }, { "epoch": 0.05496530789245447, "grad_norm": 10.394096374511719, "learning_rate": 1.970477295807341e-05, "loss": 0.0613, "step": 507 }, { "epoch": 0.05507372072853426, "grad_norm": 3.369586706161499, "learning_rate": 1.9704047584506024e-05, "loss": 0.0371, "step": 508 }, { "epoch": 0.05518213356461405, "grad_norm": 6.309421062469482, "learning_rate": 1.9703322210938636e-05, "loss": 0.0556, "step": 509 }, { "epoch": 0.05529054640069384, "grad_norm": 4.110577583312988, "learning_rate": 1.9702596837371248e-05, "loss": 0.0566, "step": 510 }, { "epoch": 0.05539895923677363, "grad_norm": 3.2007029056549072, "learning_rate": 1.970187146380386e-05, "loss": 0.0526, "step": 511 }, { "epoch": 0.055507372072853424, "grad_norm": 2.6445083618164062, "learning_rate": 1.9701146090236472e-05, "loss": 0.0709, "step": 512 }, { "epoch": 0.055615784908933215, "grad_norm": 6.004729270935059, "learning_rate": 1.9700420716669088e-05, "loss": 0.0491, "step": 513 }, { "epoch": 0.05572419774501301, "grad_norm": 2.5099875926971436, "learning_rate": 1.96996953431017e-05, "loss": 0.0299, "step": 514 }, { "epoch": 0.055832610581092804, "grad_norm": 5.836655616760254, "learning_rate": 1.9698969969534312e-05, "loss": 0.0583, "step": 515 }, { "epoch": 0.055941023417172595, "grad_norm": 4.447312831878662, "learning_rate": 1.9698244595966924e-05, "loss": 0.0431, "step": 516 }, { "epoch": 0.056049436253252385, "grad_norm": 3.79913330078125, "learning_rate": 1.9697519222399536e-05, "loss": 0.0581, "step": 517 }, { "epoch": 0.056157849089332176, "grad_norm": 4.628289699554443, "learning_rate": 1.969679384883215e-05, "loss": 0.0504, "step": 518 }, { "epoch": 0.05626626192541197, "grad_norm": 3.628683090209961, "learning_rate": 1.969606847526476e-05, "loss": 0.0491, "step": 519 }, { "epoch": 0.05637467476149176, "grad_norm": 32.523155212402344, "learning_rate": 1.9695343101697376e-05, "loss": 0.0833, "step": 520 }, { "epoch": 0.05648308759757155, "grad_norm": 5.437639236450195, "learning_rate": 1.969461772812999e-05, "loss": 0.1098, "step": 521 }, { "epoch": 0.05659150043365135, "grad_norm": 2.0797617435455322, "learning_rate": 1.96938923545626e-05, "loss": 0.0159, "step": 522 }, { "epoch": 0.05669991326973114, "grad_norm": 5.4442620277404785, "learning_rate": 1.9693166980995213e-05, "loss": 0.074, "step": 523 }, { "epoch": 0.05680832610581093, "grad_norm": 3.8812296390533447, "learning_rate": 1.9692441607427828e-05, "loss": 0.0578, "step": 524 }, { "epoch": 0.05691673894189072, "grad_norm": 4.3080878257751465, "learning_rate": 1.969171623386044e-05, "loss": 0.0528, "step": 525 }, { "epoch": 0.05702515177797051, "grad_norm": 2.624985694885254, "learning_rate": 1.9690990860293053e-05, "loss": 0.026, "step": 526 }, { "epoch": 0.0571335646140503, "grad_norm": 2.349987506866455, "learning_rate": 1.9690265486725665e-05, "loss": 0.0467, "step": 527 }, { "epoch": 0.05724197745013009, "grad_norm": 5.591397285461426, "learning_rate": 1.968954011315828e-05, "loss": 0.0816, "step": 528 }, { "epoch": 0.05735039028620989, "grad_norm": 2.528132200241089, "learning_rate": 1.9688814739590892e-05, "loss": 0.0491, "step": 529 }, { "epoch": 0.05745880312228968, "grad_norm": 1.7353483438491821, "learning_rate": 1.9688089366023505e-05, "loss": 0.0554, "step": 530 }, { "epoch": 0.05756721595836947, "grad_norm": 3.80977463722229, "learning_rate": 1.9687363992456117e-05, "loss": 0.0952, "step": 531 }, { "epoch": 0.05767562879444926, "grad_norm": 4.356201171875, "learning_rate": 1.968663861888873e-05, "loss": 0.0482, "step": 532 }, { "epoch": 0.057784041630529054, "grad_norm": 1.7471568584442139, "learning_rate": 1.968591324532134e-05, "loss": 0.0273, "step": 533 }, { "epoch": 0.057892454466608845, "grad_norm": 2.730456590652466, "learning_rate": 1.9685187871753953e-05, "loss": 0.0359, "step": 534 }, { "epoch": 0.058000867302688636, "grad_norm": 8.531983375549316, "learning_rate": 1.968446249818657e-05, "loss": 0.1297, "step": 535 }, { "epoch": 0.058109280138768434, "grad_norm": 3.639728546142578, "learning_rate": 1.968373712461918e-05, "loss": 0.0609, "step": 536 }, { "epoch": 0.058217692974848224, "grad_norm": 5.149411201477051, "learning_rate": 1.9683011751051793e-05, "loss": 0.0466, "step": 537 }, { "epoch": 0.058326105810928015, "grad_norm": 4.863864421844482, "learning_rate": 1.9682286377484405e-05, "loss": 0.0896, "step": 538 }, { "epoch": 0.058434518647007806, "grad_norm": 2.096083879470825, "learning_rate": 1.9681561003917017e-05, "loss": 0.043, "step": 539 }, { "epoch": 0.0585429314830876, "grad_norm": 3.4719061851501465, "learning_rate": 1.968083563034963e-05, "loss": 0.0446, "step": 540 }, { "epoch": 0.05865134431916739, "grad_norm": 3.5754528045654297, "learning_rate": 1.9680110256782245e-05, "loss": 0.0503, "step": 541 }, { "epoch": 0.05875975715524718, "grad_norm": 7.4635748863220215, "learning_rate": 1.9679384883214857e-05, "loss": 0.0383, "step": 542 }, { "epoch": 0.05886816999132697, "grad_norm": 3.523505210876465, "learning_rate": 1.9678659509647473e-05, "loss": 0.0831, "step": 543 }, { "epoch": 0.05897658282740677, "grad_norm": 4.310359477996826, "learning_rate": 1.9677934136080085e-05, "loss": 0.1123, "step": 544 }, { "epoch": 0.05908499566348656, "grad_norm": 3.585221290588379, "learning_rate": 1.9677208762512697e-05, "loss": 0.0501, "step": 545 }, { "epoch": 0.05919340849956635, "grad_norm": 1.7608178853988647, "learning_rate": 1.967648338894531e-05, "loss": 0.0314, "step": 546 }, { "epoch": 0.05930182133564614, "grad_norm": 1.4747008085250854, "learning_rate": 1.967575801537792e-05, "loss": 0.0209, "step": 547 }, { "epoch": 0.05941023417172593, "grad_norm": 2.551220655441284, "learning_rate": 1.9675032641810533e-05, "loss": 0.0531, "step": 548 }, { "epoch": 0.05951864700780572, "grad_norm": 3.2352406978607178, "learning_rate": 1.9674307268243145e-05, "loss": 0.0537, "step": 549 }, { "epoch": 0.05962705984388551, "grad_norm": 3.283618688583374, "learning_rate": 1.967358189467576e-05, "loss": 0.0467, "step": 550 }, { "epoch": 0.05973547267996531, "grad_norm": 2.257882595062256, "learning_rate": 1.9672856521108373e-05, "loss": 0.0356, "step": 551 }, { "epoch": 0.0598438855160451, "grad_norm": 2.604343891143799, "learning_rate": 1.9672131147540985e-05, "loss": 0.026, "step": 552 }, { "epoch": 0.05995229835212489, "grad_norm": 4.051253795623779, "learning_rate": 1.9671405773973597e-05, "loss": 0.0682, "step": 553 }, { "epoch": 0.060060711188204684, "grad_norm": 3.97097110748291, "learning_rate": 1.967068040040621e-05, "loss": 0.0532, "step": 554 }, { "epoch": 0.060169124024284475, "grad_norm": 6.491000652313232, "learning_rate": 1.966995502683882e-05, "loss": 0.0459, "step": 555 }, { "epoch": 0.060277536860364266, "grad_norm": 2.8348257541656494, "learning_rate": 1.9669229653271437e-05, "loss": 0.0239, "step": 556 }, { "epoch": 0.060385949696444056, "grad_norm": 3.802293539047241, "learning_rate": 1.966850427970405e-05, "loss": 0.0904, "step": 557 }, { "epoch": 0.06049436253252385, "grad_norm": 4.304333686828613, "learning_rate": 1.966777890613666e-05, "loss": 0.08, "step": 558 }, { "epoch": 0.060602775368603645, "grad_norm": 3.207563877105713, "learning_rate": 1.9667053532569277e-05, "loss": 0.0884, "step": 559 }, { "epoch": 0.060711188204683436, "grad_norm": 2.3900439739227295, "learning_rate": 1.966632815900189e-05, "loss": 0.051, "step": 560 }, { "epoch": 0.06081960104076323, "grad_norm": 2.372983455657959, "learning_rate": 1.96656027854345e-05, "loss": 0.0583, "step": 561 }, { "epoch": 0.06092801387684302, "grad_norm": 2.7421178817749023, "learning_rate": 1.9664877411867113e-05, "loss": 0.051, "step": 562 }, { "epoch": 0.06103642671292281, "grad_norm": 2.646981716156006, "learning_rate": 1.9664152038299726e-05, "loss": 0.0388, "step": 563 }, { "epoch": 0.0611448395490026, "grad_norm": 2.744105815887451, "learning_rate": 1.9663426664732338e-05, "loss": 0.0462, "step": 564 }, { "epoch": 0.06125325238508239, "grad_norm": 4.888704299926758, "learning_rate": 1.9662701291164953e-05, "loss": 0.0691, "step": 565 }, { "epoch": 0.06136166522116219, "grad_norm": 7.711500644683838, "learning_rate": 1.9661975917597565e-05, "loss": 0.0421, "step": 566 }, { "epoch": 0.06147007805724198, "grad_norm": 4.723653316497803, "learning_rate": 1.9661250544030178e-05, "loss": 0.1006, "step": 567 }, { "epoch": 0.06157849089332177, "grad_norm": 2.985342264175415, "learning_rate": 1.966052517046279e-05, "loss": 0.0329, "step": 568 }, { "epoch": 0.06168690372940156, "grad_norm": 3.3483784198760986, "learning_rate": 1.9659799796895402e-05, "loss": 0.0746, "step": 569 }, { "epoch": 0.06179531656548135, "grad_norm": 3.1968417167663574, "learning_rate": 1.9659074423328014e-05, "loss": 0.1258, "step": 570 }, { "epoch": 0.06190372940156114, "grad_norm": 4.349743843078613, "learning_rate": 1.9658349049760626e-05, "loss": 0.086, "step": 571 }, { "epoch": 0.062012142237640934, "grad_norm": 2.5741684436798096, "learning_rate": 1.965762367619324e-05, "loss": 0.0422, "step": 572 }, { "epoch": 0.06212055507372073, "grad_norm": 2.871182441711426, "learning_rate": 1.9656898302625854e-05, "loss": 0.0562, "step": 573 }, { "epoch": 0.06222896790980052, "grad_norm": 200.29995727539062, "learning_rate": 1.9656172929058466e-05, "loss": 0.0698, "step": 574 }, { "epoch": 0.062337380745880314, "grad_norm": 2.324767589569092, "learning_rate": 1.9655447555491078e-05, "loss": 0.0639, "step": 575 }, { "epoch": 0.062445793581960105, "grad_norm": 3.198746919631958, "learning_rate": 1.9654722181923694e-05, "loss": 0.058, "step": 576 }, { "epoch": 0.0625542064180399, "grad_norm": 3.399568796157837, "learning_rate": 1.9653996808356306e-05, "loss": 0.0431, "step": 577 }, { "epoch": 0.06266261925411969, "grad_norm": 1.3207125663757324, "learning_rate": 1.9653271434788918e-05, "loss": 0.0312, "step": 578 }, { "epoch": 0.06277103209019948, "grad_norm": 1.4269490242004395, "learning_rate": 1.965254606122153e-05, "loss": 0.032, "step": 579 }, { "epoch": 0.06287944492627927, "grad_norm": 3.001091718673706, "learning_rate": 1.9651820687654146e-05, "loss": 0.0486, "step": 580 }, { "epoch": 0.06298785776235906, "grad_norm": 2.3816397190093994, "learning_rate": 1.9651095314086758e-05, "loss": 0.0513, "step": 581 }, { "epoch": 0.06309627059843885, "grad_norm": 3.422335386276245, "learning_rate": 1.965036994051937e-05, "loss": 0.0428, "step": 582 }, { "epoch": 0.06320468343451864, "grad_norm": 5.5132551193237305, "learning_rate": 1.9649644566951982e-05, "loss": 0.0617, "step": 583 }, { "epoch": 0.06331309627059845, "grad_norm": 23.701677322387695, "learning_rate": 1.9648919193384594e-05, "loss": 0.0427, "step": 584 }, { "epoch": 0.06342150910667824, "grad_norm": 6.793729782104492, "learning_rate": 1.9648193819817206e-05, "loss": 0.103, "step": 585 }, { "epoch": 0.06352992194275803, "grad_norm": 2.5663087368011475, "learning_rate": 1.964746844624982e-05, "loss": 0.0096, "step": 586 }, { "epoch": 0.06363833477883782, "grad_norm": 3.63215708732605, "learning_rate": 1.9646743072682434e-05, "loss": 0.0334, "step": 587 }, { "epoch": 0.06374674761491761, "grad_norm": 2.371110439300537, "learning_rate": 1.9646017699115046e-05, "loss": 0.0249, "step": 588 }, { "epoch": 0.0638551604509974, "grad_norm": 7.108712196350098, "learning_rate": 1.9645292325547658e-05, "loss": 0.0706, "step": 589 }, { "epoch": 0.06396357328707719, "grad_norm": 5.577842712402344, "learning_rate": 1.964456695198027e-05, "loss": 0.0866, "step": 590 }, { "epoch": 0.06407198612315698, "grad_norm": 5.433622360229492, "learning_rate": 1.9643841578412883e-05, "loss": 0.041, "step": 591 }, { "epoch": 0.06418039895923677, "grad_norm": 11.681636810302734, "learning_rate": 1.9643116204845495e-05, "loss": 0.1033, "step": 592 }, { "epoch": 0.06428881179531656, "grad_norm": 7.906980037689209, "learning_rate": 1.964239083127811e-05, "loss": 0.1355, "step": 593 }, { "epoch": 0.06439722463139635, "grad_norm": 3.888061285018921, "learning_rate": 1.9641665457710722e-05, "loss": 0.0277, "step": 594 }, { "epoch": 0.06450563746747615, "grad_norm": 4.829146862030029, "learning_rate": 1.9640940084143338e-05, "loss": 0.0491, "step": 595 }, { "epoch": 0.06461405030355594, "grad_norm": 8.500846862792969, "learning_rate": 1.964021471057595e-05, "loss": 0.0546, "step": 596 }, { "epoch": 0.06472246313963573, "grad_norm": 9.64926528930664, "learning_rate": 1.9639489337008562e-05, "loss": 0.1266, "step": 597 }, { "epoch": 0.06483087597571552, "grad_norm": 2.547597646713257, "learning_rate": 1.9638763963441174e-05, "loss": 0.0261, "step": 598 }, { "epoch": 0.06493928881179532, "grad_norm": 1.596877098083496, "learning_rate": 1.9638038589873786e-05, "loss": 0.0085, "step": 599 }, { "epoch": 0.06504770164787511, "grad_norm": 3.450026273727417, "learning_rate": 1.96373132163064e-05, "loss": 0.0431, "step": 600 }, { "epoch": 0.0651561144839549, "grad_norm": 3.923842430114746, "learning_rate": 1.963658784273901e-05, "loss": 0.1183, "step": 601 }, { "epoch": 0.0652645273200347, "grad_norm": 3.626577138900757, "learning_rate": 1.9635862469171626e-05, "loss": 0.0967, "step": 602 }, { "epoch": 0.06537294015611449, "grad_norm": 3.294724225997925, "learning_rate": 1.963513709560424e-05, "loss": 0.0743, "step": 603 }, { "epoch": 0.06548135299219428, "grad_norm": 3.637589693069458, "learning_rate": 1.963441172203685e-05, "loss": 0.0656, "step": 604 }, { "epoch": 0.06558976582827407, "grad_norm": 5.33973503112793, "learning_rate": 1.9633686348469463e-05, "loss": 0.0726, "step": 605 }, { "epoch": 0.06569817866435386, "grad_norm": 7.680907249450684, "learning_rate": 1.9632960974902075e-05, "loss": 0.1097, "step": 606 }, { "epoch": 0.06580659150043365, "grad_norm": 4.037341594696045, "learning_rate": 1.9632235601334687e-05, "loss": 0.0399, "step": 607 }, { "epoch": 0.06591500433651344, "grad_norm": 1.8878170251846313, "learning_rate": 1.9631510227767303e-05, "loss": 0.0403, "step": 608 }, { "epoch": 0.06602341717259323, "grad_norm": 1.8813974857330322, "learning_rate": 1.9630784854199915e-05, "loss": 0.0433, "step": 609 }, { "epoch": 0.06613183000867302, "grad_norm": 2.0162320137023926, "learning_rate": 1.9630059480632527e-05, "loss": 0.0455, "step": 610 }, { "epoch": 0.06624024284475281, "grad_norm": 3.4538681507110596, "learning_rate": 1.9629334107065142e-05, "loss": 0.0622, "step": 611 }, { "epoch": 0.0663486556808326, "grad_norm": 3.5009891986846924, "learning_rate": 1.9628608733497754e-05, "loss": 0.0683, "step": 612 }, { "epoch": 0.0664570685169124, "grad_norm": 2.186371088027954, "learning_rate": 1.9627883359930367e-05, "loss": 0.0347, "step": 613 }, { "epoch": 0.0665654813529922, "grad_norm": 59.01655197143555, "learning_rate": 1.962715798636298e-05, "loss": 0.064, "step": 614 }, { "epoch": 0.06667389418907199, "grad_norm": 1.8680232763290405, "learning_rate": 1.962643261279559e-05, "loss": 0.0356, "step": 615 }, { "epoch": 0.06678230702515178, "grad_norm": 3.3287734985351562, "learning_rate": 1.9625707239228203e-05, "loss": 0.0619, "step": 616 }, { "epoch": 0.06689071986123157, "grad_norm": 2.2159547805786133, "learning_rate": 1.962498186566082e-05, "loss": 0.0311, "step": 617 }, { "epoch": 0.06699913269731136, "grad_norm": 1.953615427017212, "learning_rate": 1.962425649209343e-05, "loss": 0.0476, "step": 618 }, { "epoch": 0.06710754553339116, "grad_norm": 3.6914849281311035, "learning_rate": 1.9623531118526043e-05, "loss": 0.0688, "step": 619 }, { "epoch": 0.06721595836947095, "grad_norm": 2.9487693309783936, "learning_rate": 1.9622805744958655e-05, "loss": 0.0508, "step": 620 }, { "epoch": 0.06732437120555074, "grad_norm": 2.628812789916992, "learning_rate": 1.9622080371391267e-05, "loss": 0.0249, "step": 621 }, { "epoch": 0.06743278404163053, "grad_norm": 5.874948024749756, "learning_rate": 1.962135499782388e-05, "loss": 0.0503, "step": 622 }, { "epoch": 0.06754119687771032, "grad_norm": 3.161536693572998, "learning_rate": 1.962062962425649e-05, "loss": 0.0249, "step": 623 }, { "epoch": 0.06764960971379011, "grad_norm": 2.91694712638855, "learning_rate": 1.9619904250689107e-05, "loss": 0.066, "step": 624 }, { "epoch": 0.0677580225498699, "grad_norm": 4.106259822845459, "learning_rate": 1.961917887712172e-05, "loss": 0.0604, "step": 625 }, { "epoch": 0.06786643538594969, "grad_norm": 3.7106740474700928, "learning_rate": 1.961845350355433e-05, "loss": 0.0209, "step": 626 }, { "epoch": 0.06797484822202948, "grad_norm": 4.714236259460449, "learning_rate": 1.9617728129986947e-05, "loss": 0.0788, "step": 627 }, { "epoch": 0.06808326105810927, "grad_norm": 5.890994548797607, "learning_rate": 1.961700275641956e-05, "loss": 0.0918, "step": 628 }, { "epoch": 0.06819167389418908, "grad_norm": 3.877474546432495, "learning_rate": 1.961627738285217e-05, "loss": 0.0691, "step": 629 }, { "epoch": 0.06830008673026887, "grad_norm": 2.8131701946258545, "learning_rate": 1.9615552009284783e-05, "loss": 0.0568, "step": 630 }, { "epoch": 0.06840849956634866, "grad_norm": 3.193634510040283, "learning_rate": 1.9614826635717395e-05, "loss": 0.0357, "step": 631 }, { "epoch": 0.06851691240242845, "grad_norm": 2.090355634689331, "learning_rate": 1.961410126215001e-05, "loss": 0.0348, "step": 632 }, { "epoch": 0.06862532523850824, "grad_norm": 7.561713218688965, "learning_rate": 1.9613375888582623e-05, "loss": 0.1246, "step": 633 }, { "epoch": 0.06873373807458803, "grad_norm": 2.9228508472442627, "learning_rate": 1.9612650515015235e-05, "loss": 0.0264, "step": 634 }, { "epoch": 0.06884215091066782, "grad_norm": 8.408668518066406, "learning_rate": 1.9611925141447847e-05, "loss": 0.0579, "step": 635 }, { "epoch": 0.06895056374674761, "grad_norm": 6.449833869934082, "learning_rate": 1.961119976788046e-05, "loss": 0.0407, "step": 636 }, { "epoch": 0.0690589765828274, "grad_norm": 1.160075068473816, "learning_rate": 1.961047439431307e-05, "loss": 0.0214, "step": 637 }, { "epoch": 0.0691673894189072, "grad_norm": 3.1668701171875, "learning_rate": 1.9609749020745684e-05, "loss": 0.0602, "step": 638 }, { "epoch": 0.06927580225498699, "grad_norm": 9.038863182067871, "learning_rate": 1.96090236471783e-05, "loss": 0.0359, "step": 639 }, { "epoch": 0.06938421509106678, "grad_norm": 3.6051650047302246, "learning_rate": 1.960829827361091e-05, "loss": 0.0413, "step": 640 }, { "epoch": 0.06949262792714657, "grad_norm": 2.8582258224487305, "learning_rate": 1.9607572900043524e-05, "loss": 0.0479, "step": 641 }, { "epoch": 0.06960104076322636, "grad_norm": 4.227807998657227, "learning_rate": 1.9606847526476136e-05, "loss": 0.041, "step": 642 }, { "epoch": 0.06970945359930615, "grad_norm": 1.87429940700531, "learning_rate": 1.9606122152908748e-05, "loss": 0.0446, "step": 643 }, { "epoch": 0.06981786643538596, "grad_norm": 4.449476718902588, "learning_rate": 1.960539677934136e-05, "loss": 0.0737, "step": 644 }, { "epoch": 0.06992627927146575, "grad_norm": 2.3853073120117188, "learning_rate": 1.9604671405773976e-05, "loss": 0.0494, "step": 645 }, { "epoch": 0.07003469210754554, "grad_norm": 6.577243328094482, "learning_rate": 1.9603946032206588e-05, "loss": 0.1163, "step": 646 }, { "epoch": 0.07014310494362533, "grad_norm": 9.905420303344727, "learning_rate": 1.9603220658639203e-05, "loss": 0.0824, "step": 647 }, { "epoch": 0.07025151777970512, "grad_norm": 4.698455333709717, "learning_rate": 1.9602495285071815e-05, "loss": 0.0698, "step": 648 }, { "epoch": 0.07035993061578491, "grad_norm": 3.041478395462036, "learning_rate": 1.9601769911504427e-05, "loss": 0.0441, "step": 649 }, { "epoch": 0.0704683434518647, "grad_norm": 1.6452568769454956, "learning_rate": 1.960104453793704e-05, "loss": 0.0187, "step": 650 }, { "epoch": 0.07057675628794449, "grad_norm": 2.0021965503692627, "learning_rate": 1.9600319164369652e-05, "loss": 0.04, "step": 651 }, { "epoch": 0.07068516912402428, "grad_norm": 8.98520565032959, "learning_rate": 1.9599593790802264e-05, "loss": 0.0917, "step": 652 }, { "epoch": 0.07079358196010407, "grad_norm": 1.9804836511611938, "learning_rate": 1.9598868417234876e-05, "loss": 0.031, "step": 653 }, { "epoch": 0.07090199479618386, "grad_norm": 3.0743303298950195, "learning_rate": 1.959814304366749e-05, "loss": 0.0593, "step": 654 }, { "epoch": 0.07101040763226366, "grad_norm": 5.541501045227051, "learning_rate": 1.9597417670100104e-05, "loss": 0.0681, "step": 655 }, { "epoch": 0.07111882046834345, "grad_norm": 1.6542760133743286, "learning_rate": 1.9596692296532716e-05, "loss": 0.0486, "step": 656 }, { "epoch": 0.07122723330442324, "grad_norm": 3.0965757369995117, "learning_rate": 1.9595966922965328e-05, "loss": 0.0585, "step": 657 }, { "epoch": 0.07133564614050304, "grad_norm": 1.6118593215942383, "learning_rate": 1.959524154939794e-05, "loss": 0.0399, "step": 658 }, { "epoch": 0.07144405897658283, "grad_norm": 2.6324992179870605, "learning_rate": 1.9594516175830552e-05, "loss": 0.0395, "step": 659 }, { "epoch": 0.07155247181266262, "grad_norm": 1.3974254131317139, "learning_rate": 1.9593790802263168e-05, "loss": 0.0372, "step": 660 }, { "epoch": 0.07166088464874242, "grad_norm": 3.8112337589263916, "learning_rate": 1.959306542869578e-05, "loss": 0.0381, "step": 661 }, { "epoch": 0.0717692974848222, "grad_norm": 3.6651525497436523, "learning_rate": 1.9592340055128392e-05, "loss": 0.0529, "step": 662 }, { "epoch": 0.071877710320902, "grad_norm": 1.4524255990982056, "learning_rate": 1.9591614681561008e-05, "loss": 0.0347, "step": 663 }, { "epoch": 0.07198612315698179, "grad_norm": 9.386402130126953, "learning_rate": 1.959088930799362e-05, "loss": 0.047, "step": 664 }, { "epoch": 0.07209453599306158, "grad_norm": 5.447416305541992, "learning_rate": 1.9590163934426232e-05, "loss": 0.0591, "step": 665 }, { "epoch": 0.07220294882914137, "grad_norm": 4.788877487182617, "learning_rate": 1.9589438560858844e-05, "loss": 0.0948, "step": 666 }, { "epoch": 0.07231136166522116, "grad_norm": 2.539600133895874, "learning_rate": 1.9588713187291456e-05, "loss": 0.0253, "step": 667 }, { "epoch": 0.07241977450130095, "grad_norm": 4.756677627563477, "learning_rate": 1.958798781372407e-05, "loss": 0.06, "step": 668 }, { "epoch": 0.07252818733738074, "grad_norm": 2.7334864139556885, "learning_rate": 1.9587262440156684e-05, "loss": 0.0279, "step": 669 }, { "epoch": 0.07263660017346053, "grad_norm": 2.4644765853881836, "learning_rate": 1.9586537066589296e-05, "loss": 0.0737, "step": 670 }, { "epoch": 0.07274501300954032, "grad_norm": 10.11903190612793, "learning_rate": 1.9585811693021908e-05, "loss": 0.0505, "step": 671 }, { "epoch": 0.07285342584562012, "grad_norm": 14.488880157470703, "learning_rate": 1.958508631945452e-05, "loss": 0.0474, "step": 672 }, { "epoch": 0.07296183868169992, "grad_norm": 6.2305827140808105, "learning_rate": 1.9584360945887132e-05, "loss": 0.1549, "step": 673 }, { "epoch": 0.07307025151777971, "grad_norm": 0.9940775632858276, "learning_rate": 1.9583635572319745e-05, "loss": 0.0211, "step": 674 }, { "epoch": 0.0731786643538595, "grad_norm": 5.209979057312012, "learning_rate": 1.9582910198752357e-05, "loss": 0.1386, "step": 675 }, { "epoch": 0.07328707718993929, "grad_norm": 5.154874801635742, "learning_rate": 1.9582184825184972e-05, "loss": 0.0766, "step": 676 }, { "epoch": 0.07339549002601908, "grad_norm": 2.8494253158569336, "learning_rate": 1.9581459451617584e-05, "loss": 0.0279, "step": 677 }, { "epoch": 0.07350390286209887, "grad_norm": 4.336530685424805, "learning_rate": 1.9580734078050197e-05, "loss": 0.0691, "step": 678 }, { "epoch": 0.07361231569817867, "grad_norm": 2.5028467178344727, "learning_rate": 1.9580008704482812e-05, "loss": 0.0444, "step": 679 }, { "epoch": 0.07372072853425846, "grad_norm": 2.386195421218872, "learning_rate": 1.9579283330915424e-05, "loss": 0.0447, "step": 680 }, { "epoch": 0.07382914137033825, "grad_norm": 2.9051308631896973, "learning_rate": 1.9578557957348036e-05, "loss": 0.0206, "step": 681 }, { "epoch": 0.07393755420641804, "grad_norm": 2.050185441970825, "learning_rate": 1.957783258378065e-05, "loss": 0.0338, "step": 682 }, { "epoch": 0.07404596704249783, "grad_norm": 2.7176029682159424, "learning_rate": 1.957710721021326e-05, "loss": 0.084, "step": 683 }, { "epoch": 0.07415437987857762, "grad_norm": 5.347723007202148, "learning_rate": 1.9576381836645876e-05, "loss": 0.0712, "step": 684 }, { "epoch": 0.07426279271465741, "grad_norm": 3.3440849781036377, "learning_rate": 1.957565646307849e-05, "loss": 0.0367, "step": 685 }, { "epoch": 0.0743712055507372, "grad_norm": 3.5053160190582275, "learning_rate": 1.95749310895111e-05, "loss": 0.0605, "step": 686 }, { "epoch": 0.07447961838681699, "grad_norm": 4.215742111206055, "learning_rate": 1.9574205715943713e-05, "loss": 0.0601, "step": 687 }, { "epoch": 0.0745880312228968, "grad_norm": 2.844341278076172, "learning_rate": 1.9573480342376325e-05, "loss": 0.0872, "step": 688 }, { "epoch": 0.07469644405897659, "grad_norm": 2.631000280380249, "learning_rate": 1.9572754968808937e-05, "loss": 0.0558, "step": 689 }, { "epoch": 0.07480485689505638, "grad_norm": 2.687541961669922, "learning_rate": 1.957202959524155e-05, "loss": 0.0417, "step": 690 }, { "epoch": 0.07491326973113617, "grad_norm": 6.718609809875488, "learning_rate": 1.9571304221674165e-05, "loss": 0.0681, "step": 691 }, { "epoch": 0.07502168256721596, "grad_norm": 3.563746452331543, "learning_rate": 1.9570578848106777e-05, "loss": 0.0549, "step": 692 }, { "epoch": 0.07513009540329575, "grad_norm": 18.394298553466797, "learning_rate": 1.956985347453939e-05, "loss": 0.0549, "step": 693 }, { "epoch": 0.07523850823937554, "grad_norm": 1.2035529613494873, "learning_rate": 1.9569128100972e-05, "loss": 0.0145, "step": 694 }, { "epoch": 0.07534692107545533, "grad_norm": 2.0338563919067383, "learning_rate": 1.9568402727404613e-05, "loss": 0.0149, "step": 695 }, { "epoch": 0.07545533391153512, "grad_norm": 3.828939914703369, "learning_rate": 1.9567677353837225e-05, "loss": 0.0262, "step": 696 }, { "epoch": 0.07556374674761492, "grad_norm": 2.6827304363250732, "learning_rate": 1.956695198026984e-05, "loss": 0.0512, "step": 697 }, { "epoch": 0.0756721595836947, "grad_norm": 2.7518911361694336, "learning_rate": 1.9566226606702453e-05, "loss": 0.0225, "step": 698 }, { "epoch": 0.0757805724197745, "grad_norm": 3.014322519302368, "learning_rate": 1.956550123313507e-05, "loss": 0.0234, "step": 699 }, { "epoch": 0.07588898525585429, "grad_norm": 1.918426513671875, "learning_rate": 1.956477585956768e-05, "loss": 0.0162, "step": 700 }, { "epoch": 0.07599739809193408, "grad_norm": 8.533122062683105, "learning_rate": 1.9564050486000293e-05, "loss": 0.0854, "step": 701 }, { "epoch": 0.07610581092801387, "grad_norm": 7.353219032287598, "learning_rate": 1.9563325112432905e-05, "loss": 0.027, "step": 702 }, { "epoch": 0.07621422376409367, "grad_norm": 3.9690194129943848, "learning_rate": 1.9562599738865517e-05, "loss": 0.0364, "step": 703 }, { "epoch": 0.07632263660017347, "grad_norm": 10.988621711730957, "learning_rate": 1.956187436529813e-05, "loss": 0.0822, "step": 704 }, { "epoch": 0.07643104943625326, "grad_norm": 5.557504653930664, "learning_rate": 1.956114899173074e-05, "loss": 0.0377, "step": 705 }, { "epoch": 0.07653946227233305, "grad_norm": 5.791282653808594, "learning_rate": 1.9560423618163357e-05, "loss": 0.0279, "step": 706 }, { "epoch": 0.07664787510841284, "grad_norm": 8.127673149108887, "learning_rate": 1.955969824459597e-05, "loss": 0.0501, "step": 707 }, { "epoch": 0.07675628794449263, "grad_norm": 4.467450141906738, "learning_rate": 1.955897287102858e-05, "loss": 0.0572, "step": 708 }, { "epoch": 0.07686470078057242, "grad_norm": 4.788024425506592, "learning_rate": 1.9558247497461193e-05, "loss": 0.1319, "step": 709 }, { "epoch": 0.07697311361665221, "grad_norm": 5.628632545471191, "learning_rate": 1.9557522123893806e-05, "loss": 0.0573, "step": 710 }, { "epoch": 0.077081526452732, "grad_norm": 4.148226737976074, "learning_rate": 1.9556796750326418e-05, "loss": 0.0162, "step": 711 }, { "epoch": 0.07718993928881179, "grad_norm": 4.563852310180664, "learning_rate": 1.9556071376759033e-05, "loss": 0.0327, "step": 712 }, { "epoch": 0.07729835212489158, "grad_norm": 3.549422264099121, "learning_rate": 1.9555346003191645e-05, "loss": 0.0847, "step": 713 }, { "epoch": 0.07740676496097137, "grad_norm": 4.3042802810668945, "learning_rate": 1.9554620629624257e-05, "loss": 0.0226, "step": 714 }, { "epoch": 0.07751517779705117, "grad_norm": 2.785055160522461, "learning_rate": 1.9553895256056873e-05, "loss": 0.0611, "step": 715 }, { "epoch": 0.07762359063313096, "grad_norm": 2.8389837741851807, "learning_rate": 1.9553169882489485e-05, "loss": 0.0173, "step": 716 }, { "epoch": 0.07773200346921076, "grad_norm": 3.3485703468322754, "learning_rate": 1.9552444508922097e-05, "loss": 0.0349, "step": 717 }, { "epoch": 0.07784041630529055, "grad_norm": 6.738200664520264, "learning_rate": 1.955171913535471e-05, "loss": 0.1222, "step": 718 }, { "epoch": 0.07794882914137034, "grad_norm": 7.581382751464844, "learning_rate": 1.955099376178732e-05, "loss": 0.0369, "step": 719 }, { "epoch": 0.07805724197745013, "grad_norm": 4.195493698120117, "learning_rate": 1.9550268388219934e-05, "loss": 0.0649, "step": 720 }, { "epoch": 0.07816565481352993, "grad_norm": 2.419877529144287, "learning_rate": 1.954954301465255e-05, "loss": 0.0552, "step": 721 }, { "epoch": 0.07827406764960972, "grad_norm": 3.4764771461486816, "learning_rate": 1.954881764108516e-05, "loss": 0.0444, "step": 722 }, { "epoch": 0.0783824804856895, "grad_norm": 1.813067078590393, "learning_rate": 1.9548092267517774e-05, "loss": 0.0667, "step": 723 }, { "epoch": 0.0784908933217693, "grad_norm": 4.262580871582031, "learning_rate": 1.9547366893950386e-05, "loss": 0.1131, "step": 724 }, { "epoch": 0.07859930615784909, "grad_norm": 2.7788307666778564, "learning_rate": 1.9546641520382998e-05, "loss": 0.0598, "step": 725 }, { "epoch": 0.07870771899392888, "grad_norm": 2.8912434577941895, "learning_rate": 1.954591614681561e-05, "loss": 0.0443, "step": 726 }, { "epoch": 0.07881613183000867, "grad_norm": 1.4776376485824585, "learning_rate": 1.9545190773248222e-05, "loss": 0.0304, "step": 727 }, { "epoch": 0.07892454466608846, "grad_norm": 2.2209930419921875, "learning_rate": 1.9544465399680838e-05, "loss": 0.081, "step": 728 }, { "epoch": 0.07903295750216825, "grad_norm": 1.2543904781341553, "learning_rate": 1.954374002611345e-05, "loss": 0.0184, "step": 729 }, { "epoch": 0.07914137033824804, "grad_norm": 10.974804878234863, "learning_rate": 1.9543014652546062e-05, "loss": 0.0949, "step": 730 }, { "epoch": 0.07924978317432783, "grad_norm": 3.2001538276672363, "learning_rate": 1.9542289278978677e-05, "loss": 0.0326, "step": 731 }, { "epoch": 0.07935819601040764, "grad_norm": 5.797237396240234, "learning_rate": 1.954156390541129e-05, "loss": 0.039, "step": 732 }, { "epoch": 0.07946660884648743, "grad_norm": 3.0231528282165527, "learning_rate": 1.9540838531843902e-05, "loss": 0.0198, "step": 733 }, { "epoch": 0.07957502168256722, "grad_norm": 11.148136138916016, "learning_rate": 1.9540113158276514e-05, "loss": 0.0612, "step": 734 }, { "epoch": 0.07968343451864701, "grad_norm": 9.366662979125977, "learning_rate": 1.9539387784709126e-05, "loss": 0.0625, "step": 735 }, { "epoch": 0.0797918473547268, "grad_norm": 4.999177932739258, "learning_rate": 1.953866241114174e-05, "loss": 0.1106, "step": 736 }, { "epoch": 0.0799002601908066, "grad_norm": 3.1119775772094727, "learning_rate": 1.9537937037574354e-05, "loss": 0.0612, "step": 737 }, { "epoch": 0.08000867302688638, "grad_norm": 3.8319056034088135, "learning_rate": 1.9537211664006966e-05, "loss": 0.0337, "step": 738 }, { "epoch": 0.08011708586296618, "grad_norm": 3.752347230911255, "learning_rate": 1.9536486290439578e-05, "loss": 0.0549, "step": 739 }, { "epoch": 0.08022549869904597, "grad_norm": 2.003274440765381, "learning_rate": 1.953576091687219e-05, "loss": 0.054, "step": 740 }, { "epoch": 0.08033391153512576, "grad_norm": 6.203457832336426, "learning_rate": 1.9535035543304802e-05, "loss": 0.0424, "step": 741 }, { "epoch": 0.08044232437120555, "grad_norm": 7.752996921539307, "learning_rate": 1.9534310169737414e-05, "loss": 0.0382, "step": 742 }, { "epoch": 0.08055073720728534, "grad_norm": 6.901210308074951, "learning_rate": 1.953358479617003e-05, "loss": 0.0712, "step": 743 }, { "epoch": 0.08065915004336513, "grad_norm": 2.3372488021850586, "learning_rate": 1.9532859422602642e-05, "loss": 0.0258, "step": 744 }, { "epoch": 0.08076756287944492, "grad_norm": 3.2267825603485107, "learning_rate": 1.9532134049035254e-05, "loss": 0.0586, "step": 745 }, { "epoch": 0.08087597571552471, "grad_norm": 3.102271795272827, "learning_rate": 1.9531408675467866e-05, "loss": 0.0235, "step": 746 }, { "epoch": 0.08098438855160452, "grad_norm": 9.77538013458252, "learning_rate": 1.953068330190048e-05, "loss": 0.0356, "step": 747 }, { "epoch": 0.08109280138768431, "grad_norm": 4.605435848236084, "learning_rate": 1.9529957928333094e-05, "loss": 0.0668, "step": 748 }, { "epoch": 0.0812012142237641, "grad_norm": 5.418161392211914, "learning_rate": 1.9529232554765706e-05, "loss": 0.0529, "step": 749 }, { "epoch": 0.08130962705984389, "grad_norm": 10.090349197387695, "learning_rate": 1.952850718119832e-05, "loss": 0.0681, "step": 750 }, { "epoch": 0.08141803989592368, "grad_norm": 6.803849697113037, "learning_rate": 1.9527781807630934e-05, "loss": 0.05, "step": 751 }, { "epoch": 0.08152645273200347, "grad_norm": 3.5309934616088867, "learning_rate": 1.9527056434063546e-05, "loss": 0.0624, "step": 752 }, { "epoch": 0.08163486556808326, "grad_norm": 6.093686580657959, "learning_rate": 1.9526331060496158e-05, "loss": 0.0783, "step": 753 }, { "epoch": 0.08174327840416305, "grad_norm": 3.6027534008026123, "learning_rate": 1.952560568692877e-05, "loss": 0.0115, "step": 754 }, { "epoch": 0.08185169124024284, "grad_norm": 5.483663082122803, "learning_rate": 1.9524880313361382e-05, "loss": 0.0387, "step": 755 }, { "epoch": 0.08196010407632263, "grad_norm": 2.8715968132019043, "learning_rate": 1.9524154939793995e-05, "loss": 0.0579, "step": 756 }, { "epoch": 0.08206851691240243, "grad_norm": 7.9276347160339355, "learning_rate": 1.9523429566226607e-05, "loss": 0.0784, "step": 757 }, { "epoch": 0.08217692974848222, "grad_norm": 3.427947521209717, "learning_rate": 1.9522704192659222e-05, "loss": 0.0357, "step": 758 }, { "epoch": 0.08228534258456201, "grad_norm": 12.16773509979248, "learning_rate": 1.9521978819091834e-05, "loss": 0.0447, "step": 759 }, { "epoch": 0.0823937554206418, "grad_norm": 2.9568984508514404, "learning_rate": 1.9521253445524447e-05, "loss": 0.0173, "step": 760 }, { "epoch": 0.08250216825672159, "grad_norm": 2.1161751747131348, "learning_rate": 1.952052807195706e-05, "loss": 0.072, "step": 761 }, { "epoch": 0.0826105810928014, "grad_norm": 2.0321083068847656, "learning_rate": 1.951980269838967e-05, "loss": 0.0214, "step": 762 }, { "epoch": 0.08271899392888118, "grad_norm": 3.7872097492218018, "learning_rate": 1.9519077324822283e-05, "loss": 0.0139, "step": 763 }, { "epoch": 0.08282740676496098, "grad_norm": 1.3267337083816528, "learning_rate": 1.95183519512549e-05, "loss": 0.0115, "step": 764 }, { "epoch": 0.08293581960104077, "grad_norm": 1.4117015600204468, "learning_rate": 1.951762657768751e-05, "loss": 0.0368, "step": 765 }, { "epoch": 0.08304423243712056, "grad_norm": 6.522543907165527, "learning_rate": 1.9516901204120123e-05, "loss": 0.0897, "step": 766 }, { "epoch": 0.08315264527320035, "grad_norm": 4.2172136306762695, "learning_rate": 1.951617583055274e-05, "loss": 0.0442, "step": 767 }, { "epoch": 0.08326105810928014, "grad_norm": 7.285257339477539, "learning_rate": 1.951545045698535e-05, "loss": 0.0623, "step": 768 }, { "epoch": 0.08336947094535993, "grad_norm": 2.887662887573242, "learning_rate": 1.9514725083417963e-05, "loss": 0.0419, "step": 769 }, { "epoch": 0.08347788378143972, "grad_norm": 3.701214075088501, "learning_rate": 1.9513999709850575e-05, "loss": 0.0269, "step": 770 }, { "epoch": 0.08358629661751951, "grad_norm": 5.561027526855469, "learning_rate": 1.9513274336283187e-05, "loss": 0.0638, "step": 771 }, { "epoch": 0.0836947094535993, "grad_norm": 4.2282328605651855, "learning_rate": 1.95125489627158e-05, "loss": 0.077, "step": 772 }, { "epoch": 0.0838031222896791, "grad_norm": 3.1766786575317383, "learning_rate": 1.9511823589148415e-05, "loss": 0.0544, "step": 773 }, { "epoch": 0.08391153512575888, "grad_norm": 3.1175570487976074, "learning_rate": 1.9511098215581027e-05, "loss": 0.0438, "step": 774 }, { "epoch": 0.08401994796183868, "grad_norm": 1.5555206537246704, "learning_rate": 1.951037284201364e-05, "loss": 0.0195, "step": 775 }, { "epoch": 0.08412836079791847, "grad_norm": 5.114343643188477, "learning_rate": 1.950964746844625e-05, "loss": 0.0738, "step": 776 }, { "epoch": 0.08423677363399827, "grad_norm": 2.408695936203003, "learning_rate": 1.9508922094878863e-05, "loss": 0.0392, "step": 777 }, { "epoch": 0.08434518647007806, "grad_norm": 4.348352909088135, "learning_rate": 1.9508196721311475e-05, "loss": 0.0768, "step": 778 }, { "epoch": 0.08445359930615785, "grad_norm": 2.836061716079712, "learning_rate": 1.9507471347744087e-05, "loss": 0.0384, "step": 779 }, { "epoch": 0.08456201214223764, "grad_norm": 3.3480629920959473, "learning_rate": 1.9506745974176703e-05, "loss": 0.0453, "step": 780 }, { "epoch": 0.08467042497831744, "grad_norm": 5.249707221984863, "learning_rate": 1.9506020600609315e-05, "loss": 0.0189, "step": 781 }, { "epoch": 0.08477883781439723, "grad_norm": 2.8242359161376953, "learning_rate": 1.9505295227041927e-05, "loss": 0.03, "step": 782 }, { "epoch": 0.08488725065047702, "grad_norm": 2.354339599609375, "learning_rate": 1.9504569853474543e-05, "loss": 0.0266, "step": 783 }, { "epoch": 0.08499566348655681, "grad_norm": 4.209244251251221, "learning_rate": 1.9503844479907155e-05, "loss": 0.0843, "step": 784 }, { "epoch": 0.0851040763226366, "grad_norm": 6.297738552093506, "learning_rate": 1.9503119106339767e-05, "loss": 0.1224, "step": 785 }, { "epoch": 0.08521248915871639, "grad_norm": 4.846606731414795, "learning_rate": 1.950239373277238e-05, "loss": 0.0539, "step": 786 }, { "epoch": 0.08532090199479618, "grad_norm": 8.721433639526367, "learning_rate": 1.950166835920499e-05, "loss": 0.0338, "step": 787 }, { "epoch": 0.08542931483087597, "grad_norm": 10.466869354248047, "learning_rate": 1.9500942985637607e-05, "loss": 0.0539, "step": 788 }, { "epoch": 0.08553772766695576, "grad_norm": 4.178712368011475, "learning_rate": 1.950021761207022e-05, "loss": 0.0239, "step": 789 }, { "epoch": 0.08564614050303555, "grad_norm": 22.32151985168457, "learning_rate": 1.949949223850283e-05, "loss": 0.0317, "step": 790 }, { "epoch": 0.08575455333911536, "grad_norm": 2.597494125366211, "learning_rate": 1.9498766864935443e-05, "loss": 0.0509, "step": 791 }, { "epoch": 0.08586296617519515, "grad_norm": 2.050475835800171, "learning_rate": 1.9498041491368055e-05, "loss": 0.0526, "step": 792 }, { "epoch": 0.08597137901127494, "grad_norm": 2.144252061843872, "learning_rate": 1.9497316117800668e-05, "loss": 0.0476, "step": 793 }, { "epoch": 0.08607979184735473, "grad_norm": 3.0752155780792236, "learning_rate": 1.949659074423328e-05, "loss": 0.0677, "step": 794 }, { "epoch": 0.08618820468343452, "grad_norm": 3.5079264640808105, "learning_rate": 1.9495865370665895e-05, "loss": 0.0295, "step": 795 }, { "epoch": 0.08629661751951431, "grad_norm": 2.3550190925598145, "learning_rate": 1.9495139997098507e-05, "loss": 0.054, "step": 796 }, { "epoch": 0.0864050303555941, "grad_norm": 1.5374412536621094, "learning_rate": 1.949441462353112e-05, "loss": 0.0245, "step": 797 }, { "epoch": 0.0865134431916739, "grad_norm": 2.5929198265075684, "learning_rate": 1.9493689249963732e-05, "loss": 0.0437, "step": 798 }, { "epoch": 0.08662185602775369, "grad_norm": 1.936644196510315, "learning_rate": 1.9492963876396344e-05, "loss": 0.0294, "step": 799 }, { "epoch": 0.08673026886383348, "grad_norm": 7.969326972961426, "learning_rate": 1.949223850282896e-05, "loss": 0.0909, "step": 800 }, { "epoch": 0.08683868169991327, "grad_norm": 4.075152397155762, "learning_rate": 1.949151312926157e-05, "loss": 0.0684, "step": 801 }, { "epoch": 0.08694709453599306, "grad_norm": 3.4394586086273193, "learning_rate": 1.9490787755694184e-05, "loss": 0.0186, "step": 802 }, { "epoch": 0.08705550737207285, "grad_norm": 4.25130033493042, "learning_rate": 1.94900623821268e-05, "loss": 0.0407, "step": 803 }, { "epoch": 0.08716392020815264, "grad_norm": 3.1211154460906982, "learning_rate": 1.948933700855941e-05, "loss": 0.0302, "step": 804 }, { "epoch": 0.08727233304423243, "grad_norm": 1.306584358215332, "learning_rate": 1.9488611634992024e-05, "loss": 0.0134, "step": 805 }, { "epoch": 0.08738074588031224, "grad_norm": 2.5777499675750732, "learning_rate": 1.9487886261424636e-05, "loss": 0.0798, "step": 806 }, { "epoch": 0.08748915871639203, "grad_norm": 2.9710445404052734, "learning_rate": 1.9487160887857248e-05, "loss": 0.0436, "step": 807 }, { "epoch": 0.08759757155247182, "grad_norm": 7.196403503417969, "learning_rate": 1.948643551428986e-05, "loss": 0.0945, "step": 808 }, { "epoch": 0.08770598438855161, "grad_norm": 6.128129482269287, "learning_rate": 1.9485710140722472e-05, "loss": 0.0543, "step": 809 }, { "epoch": 0.0878143972246314, "grad_norm": 4.1048688888549805, "learning_rate": 1.9484984767155088e-05, "loss": 0.0379, "step": 810 }, { "epoch": 0.08792281006071119, "grad_norm": 5.294620037078857, "learning_rate": 1.94842593935877e-05, "loss": 0.0196, "step": 811 }, { "epoch": 0.08803122289679098, "grad_norm": 2.6811745166778564, "learning_rate": 1.9483534020020312e-05, "loss": 0.0237, "step": 812 }, { "epoch": 0.08813963573287077, "grad_norm": 2.485431432723999, "learning_rate": 1.9482808646452924e-05, "loss": 0.0196, "step": 813 }, { "epoch": 0.08824804856895056, "grad_norm": 4.948960304260254, "learning_rate": 1.9482083272885536e-05, "loss": 0.0609, "step": 814 }, { "epoch": 0.08835646140503035, "grad_norm": 5.564424991607666, "learning_rate": 1.948135789931815e-05, "loss": 0.0714, "step": 815 }, { "epoch": 0.08846487424111014, "grad_norm": 5.936342239379883, "learning_rate": 1.9480632525750764e-05, "loss": 0.0792, "step": 816 }, { "epoch": 0.08857328707718994, "grad_norm": 4.403844833374023, "learning_rate": 1.9479907152183376e-05, "loss": 0.0542, "step": 817 }, { "epoch": 0.08868169991326973, "grad_norm": 4.872727394104004, "learning_rate": 1.9479181778615988e-05, "loss": 0.062, "step": 818 }, { "epoch": 0.08879011274934952, "grad_norm": 6.880305767059326, "learning_rate": 1.9478456405048604e-05, "loss": 0.0464, "step": 819 }, { "epoch": 0.08889852558542931, "grad_norm": 2.690502882003784, "learning_rate": 1.9477731031481216e-05, "loss": 0.0334, "step": 820 }, { "epoch": 0.08900693842150911, "grad_norm": 1.8893526792526245, "learning_rate": 1.9477005657913828e-05, "loss": 0.0308, "step": 821 }, { "epoch": 0.0891153512575889, "grad_norm": 2.624666929244995, "learning_rate": 1.947628028434644e-05, "loss": 0.0519, "step": 822 }, { "epoch": 0.0892237640936687, "grad_norm": 3.2293307781219482, "learning_rate": 1.9475554910779052e-05, "loss": 0.0602, "step": 823 }, { "epoch": 0.08933217692974849, "grad_norm": 2.775590419769287, "learning_rate": 1.9474829537211664e-05, "loss": 0.0663, "step": 824 }, { "epoch": 0.08944058976582828, "grad_norm": 2.0696823596954346, "learning_rate": 1.947410416364428e-05, "loss": 0.0258, "step": 825 }, { "epoch": 0.08954900260190807, "grad_norm": 6.1522979736328125, "learning_rate": 1.9473378790076892e-05, "loss": 0.0518, "step": 826 }, { "epoch": 0.08965741543798786, "grad_norm": 4.784139633178711, "learning_rate": 1.9472653416509504e-05, "loss": 0.0326, "step": 827 }, { "epoch": 0.08976582827406765, "grad_norm": 2.2920286655426025, "learning_rate": 1.9471928042942116e-05, "loss": 0.0427, "step": 828 }, { "epoch": 0.08987424111014744, "grad_norm": 5.593528747558594, "learning_rate": 1.947120266937473e-05, "loss": 0.1012, "step": 829 }, { "epoch": 0.08998265394622723, "grad_norm": 3.386596441268921, "learning_rate": 1.947047729580734e-05, "loss": 0.0279, "step": 830 }, { "epoch": 0.09009106678230702, "grad_norm": 4.237404823303223, "learning_rate": 1.9469751922239953e-05, "loss": 0.0391, "step": 831 }, { "epoch": 0.09019947961838681, "grad_norm": 2.1987011432647705, "learning_rate": 1.946902654867257e-05, "loss": 0.0254, "step": 832 }, { "epoch": 0.0903078924544666, "grad_norm": 6.0632524490356445, "learning_rate": 1.946830117510518e-05, "loss": 0.0665, "step": 833 }, { "epoch": 0.0904163052905464, "grad_norm": 4.218392848968506, "learning_rate": 1.9467575801537793e-05, "loss": 0.0431, "step": 834 }, { "epoch": 0.09052471812662619, "grad_norm": 2.4300765991210938, "learning_rate": 1.9466850427970408e-05, "loss": 0.0215, "step": 835 }, { "epoch": 0.09063313096270599, "grad_norm": 2.2864792346954346, "learning_rate": 1.946612505440302e-05, "loss": 0.0334, "step": 836 }, { "epoch": 0.09074154379878578, "grad_norm": 5.016496181488037, "learning_rate": 1.9465399680835632e-05, "loss": 0.0454, "step": 837 }, { "epoch": 0.09084995663486557, "grad_norm": 3.930175304412842, "learning_rate": 1.9464674307268245e-05, "loss": 0.0774, "step": 838 }, { "epoch": 0.09095836947094536, "grad_norm": 1.1249840259552002, "learning_rate": 1.9463948933700857e-05, "loss": 0.0239, "step": 839 }, { "epoch": 0.09106678230702515, "grad_norm": 4.084944725036621, "learning_rate": 1.9463223560133472e-05, "loss": 0.0339, "step": 840 }, { "epoch": 0.09117519514310495, "grad_norm": 2.750525712966919, "learning_rate": 1.9462498186566084e-05, "loss": 0.0662, "step": 841 }, { "epoch": 0.09128360797918474, "grad_norm": 7.158928871154785, "learning_rate": 1.9461772812998697e-05, "loss": 0.0666, "step": 842 }, { "epoch": 0.09139202081526453, "grad_norm": 5.887353897094727, "learning_rate": 1.946104743943131e-05, "loss": 0.0521, "step": 843 }, { "epoch": 0.09150043365134432, "grad_norm": 7.903385639190674, "learning_rate": 1.946032206586392e-05, "loss": 0.0429, "step": 844 }, { "epoch": 0.09160884648742411, "grad_norm": 2.3328089714050293, "learning_rate": 1.9459596692296533e-05, "loss": 0.0265, "step": 845 }, { "epoch": 0.0917172593235039, "grad_norm": 2.6469457149505615, "learning_rate": 1.9458871318729145e-05, "loss": 0.0711, "step": 846 }, { "epoch": 0.09182567215958369, "grad_norm": 7.08705472946167, "learning_rate": 1.945814594516176e-05, "loss": 0.0689, "step": 847 }, { "epoch": 0.09193408499566348, "grad_norm": 3.3031203746795654, "learning_rate": 1.9457420571594373e-05, "loss": 0.0339, "step": 848 }, { "epoch": 0.09204249783174327, "grad_norm": 3.240690231323242, "learning_rate": 1.9456695198026985e-05, "loss": 0.0277, "step": 849 }, { "epoch": 0.09215091066782308, "grad_norm": 8.260613441467285, "learning_rate": 1.9455969824459597e-05, "loss": 0.092, "step": 850 }, { "epoch": 0.09225932350390287, "grad_norm": 4.072226047515869, "learning_rate": 1.945524445089221e-05, "loss": 0.0426, "step": 851 }, { "epoch": 0.09236773633998266, "grad_norm": 2.708096742630005, "learning_rate": 1.9454519077324825e-05, "loss": 0.036, "step": 852 }, { "epoch": 0.09247614917606245, "grad_norm": 3.855125904083252, "learning_rate": 1.9453793703757437e-05, "loss": 0.1204, "step": 853 }, { "epoch": 0.09258456201214224, "grad_norm": 2.942384719848633, "learning_rate": 1.945306833019005e-05, "loss": 0.0317, "step": 854 }, { "epoch": 0.09269297484822203, "grad_norm": 3.4443674087524414, "learning_rate": 1.9452342956622665e-05, "loss": 0.0889, "step": 855 }, { "epoch": 0.09280138768430182, "grad_norm": 4.126702785491943, "learning_rate": 1.9451617583055277e-05, "loss": 0.0583, "step": 856 }, { "epoch": 0.09290980052038161, "grad_norm": 1.837633728981018, "learning_rate": 1.945089220948789e-05, "loss": 0.0256, "step": 857 }, { "epoch": 0.0930182133564614, "grad_norm": 3.95297908782959, "learning_rate": 1.94501668359205e-05, "loss": 0.0746, "step": 858 }, { "epoch": 0.0931266261925412, "grad_norm": 2.722135305404663, "learning_rate": 1.9449441462353113e-05, "loss": 0.0465, "step": 859 }, { "epoch": 0.09323503902862099, "grad_norm": 5.034002304077148, "learning_rate": 1.9448716088785725e-05, "loss": 0.0615, "step": 860 }, { "epoch": 0.09334345186470078, "grad_norm": 4.5056867599487305, "learning_rate": 1.9447990715218337e-05, "loss": 0.0384, "step": 861 }, { "epoch": 0.09345186470078057, "grad_norm": 2.252169370651245, "learning_rate": 1.9447265341650953e-05, "loss": 0.0369, "step": 862 }, { "epoch": 0.09356027753686036, "grad_norm": 2.785912275314331, "learning_rate": 1.9446539968083565e-05, "loss": 0.072, "step": 863 }, { "epoch": 0.09366869037294015, "grad_norm": 2.261704206466675, "learning_rate": 1.9445814594516177e-05, "loss": 0.02, "step": 864 }, { "epoch": 0.09377710320901995, "grad_norm": 0.9894945621490479, "learning_rate": 1.944508922094879e-05, "loss": 0.0204, "step": 865 }, { "epoch": 0.09388551604509975, "grad_norm": 3.30781888961792, "learning_rate": 1.94443638473814e-05, "loss": 0.0427, "step": 866 }, { "epoch": 0.09399392888117954, "grad_norm": 2.8767144680023193, "learning_rate": 1.9443638473814014e-05, "loss": 0.0433, "step": 867 }, { "epoch": 0.09410234171725933, "grad_norm": 1.9042267799377441, "learning_rate": 1.944291310024663e-05, "loss": 0.014, "step": 868 }, { "epoch": 0.09421075455333912, "grad_norm": 6.084083557128906, "learning_rate": 1.944218772667924e-05, "loss": 0.0977, "step": 869 }, { "epoch": 0.09431916738941891, "grad_norm": 2.3878490924835205, "learning_rate": 1.9441462353111853e-05, "loss": 0.0369, "step": 870 }, { "epoch": 0.0944275802254987, "grad_norm": 6.110289096832275, "learning_rate": 1.944073697954447e-05, "loss": 0.0313, "step": 871 }, { "epoch": 0.09453599306157849, "grad_norm": 0.9022179841995239, "learning_rate": 1.944001160597708e-05, "loss": 0.0153, "step": 872 }, { "epoch": 0.09464440589765828, "grad_norm": 2.7285642623901367, "learning_rate": 1.9439286232409693e-05, "loss": 0.0318, "step": 873 }, { "epoch": 0.09475281873373807, "grad_norm": 2.9747815132141113, "learning_rate": 1.9438560858842305e-05, "loss": 0.0361, "step": 874 }, { "epoch": 0.09486123156981786, "grad_norm": 4.3245086669921875, "learning_rate": 1.9437835485274918e-05, "loss": 0.0603, "step": 875 }, { "epoch": 0.09496964440589765, "grad_norm": 2.495152473449707, "learning_rate": 1.943711011170753e-05, "loss": 0.0289, "step": 876 }, { "epoch": 0.09507805724197745, "grad_norm": 1.9128282070159912, "learning_rate": 1.9436384738140145e-05, "loss": 0.0237, "step": 877 }, { "epoch": 0.09518647007805724, "grad_norm": 5.109134197235107, "learning_rate": 1.9435659364572757e-05, "loss": 0.0611, "step": 878 }, { "epoch": 0.09529488291413703, "grad_norm": 2.1022770404815674, "learning_rate": 1.943493399100537e-05, "loss": 0.034, "step": 879 }, { "epoch": 0.09540329575021683, "grad_norm": 16.06694984436035, "learning_rate": 1.9434208617437982e-05, "loss": 0.1397, "step": 880 }, { "epoch": 0.09551170858629662, "grad_norm": 4.573064804077148, "learning_rate": 1.9433483243870594e-05, "loss": 0.0861, "step": 881 }, { "epoch": 0.09562012142237641, "grad_norm": 10.64177131652832, "learning_rate": 1.9432757870303206e-05, "loss": 0.1144, "step": 882 }, { "epoch": 0.0957285342584562, "grad_norm": 4.74035120010376, "learning_rate": 1.9432032496735818e-05, "loss": 0.0725, "step": 883 }, { "epoch": 0.095836947094536, "grad_norm": 5.673234939575195, "learning_rate": 1.9431307123168434e-05, "loss": 0.0401, "step": 884 }, { "epoch": 0.09594535993061579, "grad_norm": 7.192576885223389, "learning_rate": 1.9430581749601046e-05, "loss": 0.0265, "step": 885 }, { "epoch": 0.09605377276669558, "grad_norm": 3.279780387878418, "learning_rate": 1.9429856376033658e-05, "loss": 0.0238, "step": 886 }, { "epoch": 0.09616218560277537, "grad_norm": 9.844951629638672, "learning_rate": 1.9429131002466273e-05, "loss": 0.1348, "step": 887 }, { "epoch": 0.09627059843885516, "grad_norm": 5.554569244384766, "learning_rate": 1.9428405628898886e-05, "loss": 0.096, "step": 888 }, { "epoch": 0.09637901127493495, "grad_norm": 3.6295995712280273, "learning_rate": 1.9427680255331498e-05, "loss": 0.064, "step": 889 }, { "epoch": 0.09648742411101474, "grad_norm": 5.648554801940918, "learning_rate": 1.942695488176411e-05, "loss": 0.0432, "step": 890 }, { "epoch": 0.09659583694709453, "grad_norm": 3.9648802280426025, "learning_rate": 1.9426229508196722e-05, "loss": 0.0768, "step": 891 }, { "epoch": 0.09670424978317432, "grad_norm": 1.8938851356506348, "learning_rate": 1.9425504134629338e-05, "loss": 0.0124, "step": 892 }, { "epoch": 0.09681266261925411, "grad_norm": 7.8632941246032715, "learning_rate": 1.942477876106195e-05, "loss": 0.0334, "step": 893 }, { "epoch": 0.0969210754553339, "grad_norm": 4.8264360427856445, "learning_rate": 1.9424053387494562e-05, "loss": 0.0288, "step": 894 }, { "epoch": 0.09702948829141371, "grad_norm": 1.8001034259796143, "learning_rate": 1.9423328013927174e-05, "loss": 0.0166, "step": 895 }, { "epoch": 0.0971379011274935, "grad_norm": 2.623150587081909, "learning_rate": 1.9422602640359786e-05, "loss": 0.0289, "step": 896 }, { "epoch": 0.09724631396357329, "grad_norm": 4.615998268127441, "learning_rate": 1.9421877266792398e-05, "loss": 0.0572, "step": 897 }, { "epoch": 0.09735472679965308, "grad_norm": 3.9828710556030273, "learning_rate": 1.942115189322501e-05, "loss": 0.0503, "step": 898 }, { "epoch": 0.09746313963573287, "grad_norm": 2.7297587394714355, "learning_rate": 1.9420426519657626e-05, "loss": 0.0393, "step": 899 }, { "epoch": 0.09757155247181266, "grad_norm": 6.7991943359375, "learning_rate": 1.9419701146090238e-05, "loss": 0.1481, "step": 900 }, { "epoch": 0.09767996530789246, "grad_norm": 6.709141731262207, "learning_rate": 1.941897577252285e-05, "loss": 0.0693, "step": 901 }, { "epoch": 0.09778837814397225, "grad_norm": 3.301853895187378, "learning_rate": 1.9418250398955462e-05, "loss": 0.0337, "step": 902 }, { "epoch": 0.09789679098005204, "grad_norm": 10.837639808654785, "learning_rate": 1.9417525025388075e-05, "loss": 0.0402, "step": 903 }, { "epoch": 0.09800520381613183, "grad_norm": 11.711947441101074, "learning_rate": 1.941679965182069e-05, "loss": 0.0543, "step": 904 }, { "epoch": 0.09811361665221162, "grad_norm": 6.632266044616699, "learning_rate": 1.9416074278253302e-05, "loss": 0.0865, "step": 905 }, { "epoch": 0.09822202948829141, "grad_norm": 4.5187530517578125, "learning_rate": 1.9415348904685914e-05, "loss": 0.0623, "step": 906 }, { "epoch": 0.0983304423243712, "grad_norm": 5.382956504821777, "learning_rate": 1.941462353111853e-05, "loss": 0.0807, "step": 907 }, { "epoch": 0.09843885516045099, "grad_norm": 3.8207318782806396, "learning_rate": 1.9413898157551142e-05, "loss": 0.051, "step": 908 }, { "epoch": 0.09854726799653078, "grad_norm": 1.7093274593353271, "learning_rate": 1.9413172783983754e-05, "loss": 0.0299, "step": 909 }, { "epoch": 0.09865568083261059, "grad_norm": 2.629997491836548, "learning_rate": 1.9412447410416366e-05, "loss": 0.0412, "step": 910 }, { "epoch": 0.09876409366869038, "grad_norm": 2.52372670173645, "learning_rate": 1.941172203684898e-05, "loss": 0.0452, "step": 911 }, { "epoch": 0.09887250650477017, "grad_norm": 2.8381447792053223, "learning_rate": 1.941099666328159e-05, "loss": 0.0567, "step": 912 }, { "epoch": 0.09898091934084996, "grad_norm": 5.571340560913086, "learning_rate": 1.9410271289714203e-05, "loss": 0.0728, "step": 913 }, { "epoch": 0.09908933217692975, "grad_norm": 3.7457363605499268, "learning_rate": 1.9409545916146818e-05, "loss": 0.0478, "step": 914 }, { "epoch": 0.09919774501300954, "grad_norm": 5.680902481079102, "learning_rate": 1.940882054257943e-05, "loss": 0.0627, "step": 915 }, { "epoch": 0.09930615784908933, "grad_norm": 6.845755577087402, "learning_rate": 1.9408095169012043e-05, "loss": 0.0701, "step": 916 }, { "epoch": 0.09941457068516912, "grad_norm": 5.28903865814209, "learning_rate": 1.9407369795444655e-05, "loss": 0.0348, "step": 917 }, { "epoch": 0.09952298352124891, "grad_norm": 3.4323174953460693, "learning_rate": 1.9406644421877267e-05, "loss": 0.0415, "step": 918 }, { "epoch": 0.0996313963573287, "grad_norm": 3.9965567588806152, "learning_rate": 1.940591904830988e-05, "loss": 0.0409, "step": 919 }, { "epoch": 0.0997398091934085, "grad_norm": 4.845130443572998, "learning_rate": 1.9405193674742495e-05, "loss": 0.0395, "step": 920 }, { "epoch": 0.09984822202948829, "grad_norm": 1.5713557004928589, "learning_rate": 1.9404468301175107e-05, "loss": 0.0188, "step": 921 }, { "epoch": 0.09995663486556808, "grad_norm": 4.149486064910889, "learning_rate": 1.940374292760772e-05, "loss": 0.0588, "step": 922 }, { "epoch": 0.10006504770164787, "grad_norm": 2.8854572772979736, "learning_rate": 1.9403017554040334e-05, "loss": 0.0312, "step": 923 }, { "epoch": 0.10017346053772767, "grad_norm": 4.027515888214111, "learning_rate": 1.9402292180472947e-05, "loss": 0.0438, "step": 924 }, { "epoch": 0.10028187337380746, "grad_norm": 5.885611057281494, "learning_rate": 1.940156680690556e-05, "loss": 0.0375, "step": 925 }, { "epoch": 0.10039028620988726, "grad_norm": 9.912297248840332, "learning_rate": 1.940084143333817e-05, "loss": 0.0604, "step": 926 }, { "epoch": 0.10049869904596705, "grad_norm": 9.894631385803223, "learning_rate": 1.9400116059770783e-05, "loss": 0.122, "step": 927 }, { "epoch": 0.10060711188204684, "grad_norm": 26.33985710144043, "learning_rate": 1.9399390686203395e-05, "loss": 0.0721, "step": 928 }, { "epoch": 0.10071552471812663, "grad_norm": 6.342766284942627, "learning_rate": 1.939866531263601e-05, "loss": 0.0987, "step": 929 }, { "epoch": 0.10082393755420642, "grad_norm": 3.6526436805725098, "learning_rate": 1.9397939939068623e-05, "loss": 0.0521, "step": 930 }, { "epoch": 0.10093235039028621, "grad_norm": 3.168010950088501, "learning_rate": 1.9397214565501235e-05, "loss": 0.079, "step": 931 }, { "epoch": 0.101040763226366, "grad_norm": 1.92548406124115, "learning_rate": 1.9396489191933847e-05, "loss": 0.0216, "step": 932 }, { "epoch": 0.10114917606244579, "grad_norm": 3.108233690261841, "learning_rate": 1.939576381836646e-05, "loss": 0.0508, "step": 933 }, { "epoch": 0.10125758889852558, "grad_norm": 6.930770397186279, "learning_rate": 1.939503844479907e-05, "loss": 0.098, "step": 934 }, { "epoch": 0.10136600173460537, "grad_norm": 3.681476593017578, "learning_rate": 1.9394313071231687e-05, "loss": 0.064, "step": 935 }, { "epoch": 0.10147441457068516, "grad_norm": 1.7169609069824219, "learning_rate": 1.93935876976643e-05, "loss": 0.0176, "step": 936 }, { "epoch": 0.10158282740676496, "grad_norm": 2.954493284225464, "learning_rate": 1.939286232409691e-05, "loss": 0.0661, "step": 937 }, { "epoch": 0.10169124024284475, "grad_norm": 2.99491286277771, "learning_rate": 1.9392136950529527e-05, "loss": 0.0446, "step": 938 }, { "epoch": 0.10179965307892455, "grad_norm": 4.382067680358887, "learning_rate": 1.939141157696214e-05, "loss": 0.064, "step": 939 }, { "epoch": 0.10190806591500434, "grad_norm": 7.659345626831055, "learning_rate": 1.939068620339475e-05, "loss": 0.0741, "step": 940 }, { "epoch": 0.10201647875108413, "grad_norm": 2.07871675491333, "learning_rate": 1.9389960829827363e-05, "loss": 0.0211, "step": 941 }, { "epoch": 0.10212489158716392, "grad_norm": 9.69133186340332, "learning_rate": 1.9389235456259975e-05, "loss": 0.0784, "step": 942 }, { "epoch": 0.10223330442324371, "grad_norm": 4.981085300445557, "learning_rate": 1.9388510082692587e-05, "loss": 0.0415, "step": 943 }, { "epoch": 0.1023417172593235, "grad_norm": 3.998870611190796, "learning_rate": 1.9387784709125203e-05, "loss": 0.0495, "step": 944 }, { "epoch": 0.1024501300954033, "grad_norm": 3.048024892807007, "learning_rate": 1.9387059335557815e-05, "loss": 0.0238, "step": 945 }, { "epoch": 0.10255854293148309, "grad_norm": 2.7445292472839355, "learning_rate": 1.9386333961990427e-05, "loss": 0.0409, "step": 946 }, { "epoch": 0.10266695576756288, "grad_norm": 2.0206730365753174, "learning_rate": 1.938560858842304e-05, "loss": 0.0539, "step": 947 }, { "epoch": 0.10277536860364267, "grad_norm": 2.45066237449646, "learning_rate": 1.938488321485565e-05, "loss": 0.0152, "step": 948 }, { "epoch": 0.10288378143972246, "grad_norm": 3.919203042984009, "learning_rate": 1.9384157841288264e-05, "loss": 0.0485, "step": 949 }, { "epoch": 0.10299219427580225, "grad_norm": 6.257627010345459, "learning_rate": 1.9383432467720876e-05, "loss": 0.0328, "step": 950 }, { "epoch": 0.10310060711188204, "grad_norm": 2.6437480449676514, "learning_rate": 1.938270709415349e-05, "loss": 0.0404, "step": 951 }, { "epoch": 0.10320901994796183, "grad_norm": 2.4085679054260254, "learning_rate": 1.9381981720586103e-05, "loss": 0.0295, "step": 952 }, { "epoch": 0.10331743278404162, "grad_norm": 4.051645755767822, "learning_rate": 1.9381256347018716e-05, "loss": 0.059, "step": 953 }, { "epoch": 0.10342584562012143, "grad_norm": 2.3031208515167236, "learning_rate": 1.9380530973451328e-05, "loss": 0.0322, "step": 954 }, { "epoch": 0.10353425845620122, "grad_norm": 5.206526756286621, "learning_rate": 1.937980559988394e-05, "loss": 0.033, "step": 955 }, { "epoch": 0.10364267129228101, "grad_norm": 7.631002902984619, "learning_rate": 1.9379080226316555e-05, "loss": 0.0958, "step": 956 }, { "epoch": 0.1037510841283608, "grad_norm": 7.548944473266602, "learning_rate": 1.9378354852749168e-05, "loss": 0.0595, "step": 957 }, { "epoch": 0.10385949696444059, "grad_norm": 4.5409135818481445, "learning_rate": 1.937762947918178e-05, "loss": 0.0849, "step": 958 }, { "epoch": 0.10396790980052038, "grad_norm": 2.8107078075408936, "learning_rate": 1.9376904105614395e-05, "loss": 0.0347, "step": 959 }, { "epoch": 0.10407632263660017, "grad_norm": 1.69484543800354, "learning_rate": 1.9376178732047007e-05, "loss": 0.0194, "step": 960 }, { "epoch": 0.10418473547267997, "grad_norm": 1.9799071550369263, "learning_rate": 1.937545335847962e-05, "loss": 0.0361, "step": 961 }, { "epoch": 0.10429314830875976, "grad_norm": 2.8099043369293213, "learning_rate": 1.937472798491223e-05, "loss": 0.0372, "step": 962 }, { "epoch": 0.10440156114483955, "grad_norm": 0.5516229867935181, "learning_rate": 1.9374002611344844e-05, "loss": 0.0061, "step": 963 }, { "epoch": 0.10450997398091934, "grad_norm": 4.2889933586120605, "learning_rate": 1.9373277237777456e-05, "loss": 0.0553, "step": 964 }, { "epoch": 0.10461838681699913, "grad_norm": 5.557774543762207, "learning_rate": 1.9372551864210068e-05, "loss": 0.0448, "step": 965 }, { "epoch": 0.10472679965307892, "grad_norm": 6.727304458618164, "learning_rate": 1.9371826490642684e-05, "loss": 0.053, "step": 966 }, { "epoch": 0.10483521248915871, "grad_norm": 3.0608901977539062, "learning_rate": 1.9371101117075296e-05, "loss": 0.0275, "step": 967 }, { "epoch": 0.1049436253252385, "grad_norm": 4.460418224334717, "learning_rate": 1.9370375743507908e-05, "loss": 0.0576, "step": 968 }, { "epoch": 0.1050520381613183, "grad_norm": 1.1777561902999878, "learning_rate": 1.936965036994052e-05, "loss": 0.0261, "step": 969 }, { "epoch": 0.1051604509973981, "grad_norm": 2.77899169921875, "learning_rate": 1.9368924996373132e-05, "loss": 0.0345, "step": 970 }, { "epoch": 0.10526886383347789, "grad_norm": 4.0367889404296875, "learning_rate": 1.9368199622805744e-05, "loss": 0.0378, "step": 971 }, { "epoch": 0.10537727666955768, "grad_norm": 10.892387390136719, "learning_rate": 1.936747424923836e-05, "loss": 0.0761, "step": 972 }, { "epoch": 0.10548568950563747, "grad_norm": 3.297447681427002, "learning_rate": 1.9366748875670972e-05, "loss": 0.0589, "step": 973 }, { "epoch": 0.10559410234171726, "grad_norm": 18.084257125854492, "learning_rate": 1.9366023502103584e-05, "loss": 0.0983, "step": 974 }, { "epoch": 0.10570251517779705, "grad_norm": 2.9690489768981934, "learning_rate": 1.93652981285362e-05, "loss": 0.0297, "step": 975 }, { "epoch": 0.10581092801387684, "grad_norm": 4.528113842010498, "learning_rate": 1.9364572754968812e-05, "loss": 0.0094, "step": 976 }, { "epoch": 0.10591934084995663, "grad_norm": 16.445838928222656, "learning_rate": 1.9363847381401424e-05, "loss": 0.1151, "step": 977 }, { "epoch": 0.10602775368603642, "grad_norm": 3.6172068119049072, "learning_rate": 1.9363122007834036e-05, "loss": 0.0326, "step": 978 }, { "epoch": 0.10613616652211622, "grad_norm": 4.9492011070251465, "learning_rate": 1.9362396634266648e-05, "loss": 0.0601, "step": 979 }, { "epoch": 0.106244579358196, "grad_norm": 2.6434972286224365, "learning_rate": 1.936167126069926e-05, "loss": 0.0248, "step": 980 }, { "epoch": 0.1063529921942758, "grad_norm": 2.2966673374176025, "learning_rate": 1.9360945887131876e-05, "loss": 0.0451, "step": 981 }, { "epoch": 0.10646140503035559, "grad_norm": 4.004544258117676, "learning_rate": 1.9360220513564488e-05, "loss": 0.0486, "step": 982 }, { "epoch": 0.10656981786643538, "grad_norm": 2.1232287883758545, "learning_rate": 1.93594951399971e-05, "loss": 0.0161, "step": 983 }, { "epoch": 0.10667823070251518, "grad_norm": 1.0392482280731201, "learning_rate": 1.9358769766429712e-05, "loss": 0.0085, "step": 984 }, { "epoch": 0.10678664353859497, "grad_norm": 4.162996768951416, "learning_rate": 1.9358044392862325e-05, "loss": 0.0767, "step": 985 }, { "epoch": 0.10689505637467477, "grad_norm": 4.3984551429748535, "learning_rate": 1.9357319019294937e-05, "loss": 0.0838, "step": 986 }, { "epoch": 0.10700346921075456, "grad_norm": 5.447037696838379, "learning_rate": 1.9356593645727552e-05, "loss": 0.0231, "step": 987 }, { "epoch": 0.10711188204683435, "grad_norm": 1.4274616241455078, "learning_rate": 1.9355868272160164e-05, "loss": 0.0166, "step": 988 }, { "epoch": 0.10722029488291414, "grad_norm": 10.249421119689941, "learning_rate": 1.9355142898592776e-05, "loss": 0.0518, "step": 989 }, { "epoch": 0.10732870771899393, "grad_norm": 2.47817063331604, "learning_rate": 1.9354417525025392e-05, "loss": 0.0328, "step": 990 }, { "epoch": 0.10743712055507372, "grad_norm": 3.008467674255371, "learning_rate": 1.9353692151458004e-05, "loss": 0.0171, "step": 991 }, { "epoch": 0.10754553339115351, "grad_norm": 4.006641387939453, "learning_rate": 1.9352966777890616e-05, "loss": 0.0474, "step": 992 }, { "epoch": 0.1076539462272333, "grad_norm": 3.6877782344818115, "learning_rate": 1.935224140432323e-05, "loss": 0.0298, "step": 993 }, { "epoch": 0.10776235906331309, "grad_norm": 4.536433696746826, "learning_rate": 1.935151603075584e-05, "loss": 0.0847, "step": 994 }, { "epoch": 0.10787077189939288, "grad_norm": 8.372797012329102, "learning_rate": 1.9350790657188453e-05, "loss": 0.0758, "step": 995 }, { "epoch": 0.10797918473547267, "grad_norm": 1.181340217590332, "learning_rate": 1.9350065283621068e-05, "loss": 0.0259, "step": 996 }, { "epoch": 0.10808759757155247, "grad_norm": 3.212491273880005, "learning_rate": 1.934933991005368e-05, "loss": 0.0495, "step": 997 }, { "epoch": 0.10819601040763227, "grad_norm": 2.8238728046417236, "learning_rate": 1.9348614536486293e-05, "loss": 0.0668, "step": 998 }, { "epoch": 0.10830442324371206, "grad_norm": 7.3045549392700195, "learning_rate": 1.9347889162918905e-05, "loss": 0.0559, "step": 999 }, { "epoch": 0.10841283607979185, "grad_norm": 6.315626621246338, "learning_rate": 1.9347163789351517e-05, "loss": 0.0546, "step": 1000 }, { "epoch": 0.10852124891587164, "grad_norm": 2.8727989196777344, "learning_rate": 1.934643841578413e-05, "loss": 0.0572, "step": 1001 }, { "epoch": 0.10862966175195143, "grad_norm": 3.2851154804229736, "learning_rate": 1.934571304221674e-05, "loss": 0.0275, "step": 1002 }, { "epoch": 0.10873807458803122, "grad_norm": 2.156942844390869, "learning_rate": 1.9344987668649357e-05, "loss": 0.0323, "step": 1003 }, { "epoch": 0.10884648742411102, "grad_norm": 2.9692845344543457, "learning_rate": 1.934426229508197e-05, "loss": 0.0689, "step": 1004 }, { "epoch": 0.1089549002601908, "grad_norm": 2.6225502490997314, "learning_rate": 1.934353692151458e-05, "loss": 0.0158, "step": 1005 }, { "epoch": 0.1090633130962706, "grad_norm": 1.7410624027252197, "learning_rate": 1.9342811547947193e-05, "loss": 0.0204, "step": 1006 }, { "epoch": 0.10917172593235039, "grad_norm": 4.736367702484131, "learning_rate": 1.934208617437981e-05, "loss": 0.0769, "step": 1007 }, { "epoch": 0.10928013876843018, "grad_norm": 3.5612106323242188, "learning_rate": 1.934136080081242e-05, "loss": 0.0283, "step": 1008 }, { "epoch": 0.10938855160450997, "grad_norm": 1.9974383115768433, "learning_rate": 1.9340635427245033e-05, "loss": 0.0159, "step": 1009 }, { "epoch": 0.10949696444058976, "grad_norm": 5.5498833656311035, "learning_rate": 1.9339910053677645e-05, "loss": 0.066, "step": 1010 }, { "epoch": 0.10960537727666955, "grad_norm": 1.5772004127502441, "learning_rate": 1.933918468011026e-05, "loss": 0.0311, "step": 1011 }, { "epoch": 0.10971379011274934, "grad_norm": 1.9001479148864746, "learning_rate": 1.9338459306542873e-05, "loss": 0.0222, "step": 1012 }, { "epoch": 0.10982220294882915, "grad_norm": 6.2408599853515625, "learning_rate": 1.9337733932975485e-05, "loss": 0.0509, "step": 1013 }, { "epoch": 0.10993061578490894, "grad_norm": 5.997000217437744, "learning_rate": 1.9337008559408097e-05, "loss": 0.0437, "step": 1014 }, { "epoch": 0.11003902862098873, "grad_norm": 5.023983001708984, "learning_rate": 1.933628318584071e-05, "loss": 0.0563, "step": 1015 }, { "epoch": 0.11014744145706852, "grad_norm": 3.514404296875, "learning_rate": 1.933555781227332e-05, "loss": 0.0526, "step": 1016 }, { "epoch": 0.11025585429314831, "grad_norm": 5.5685906410217285, "learning_rate": 1.9334832438705933e-05, "loss": 0.0549, "step": 1017 }, { "epoch": 0.1103642671292281, "grad_norm": 3.942329168319702, "learning_rate": 1.933410706513855e-05, "loss": 0.0418, "step": 1018 }, { "epoch": 0.1104726799653079, "grad_norm": 2.5972647666931152, "learning_rate": 1.933338169157116e-05, "loss": 0.0393, "step": 1019 }, { "epoch": 0.11058109280138768, "grad_norm": 5.024824142456055, "learning_rate": 1.9332656318003773e-05, "loss": 0.0903, "step": 1020 }, { "epoch": 0.11068950563746748, "grad_norm": 4.055258274078369, "learning_rate": 1.9331930944436385e-05, "loss": 0.0576, "step": 1021 }, { "epoch": 0.11079791847354727, "grad_norm": 1.0334407091140747, "learning_rate": 1.9331205570868998e-05, "loss": 0.0217, "step": 1022 }, { "epoch": 0.11090633130962706, "grad_norm": 2.911144495010376, "learning_rate": 1.933048019730161e-05, "loss": 0.0654, "step": 1023 }, { "epoch": 0.11101474414570685, "grad_norm": 2.859400987625122, "learning_rate": 1.9329754823734225e-05, "loss": 0.0356, "step": 1024 }, { "epoch": 0.11112315698178664, "grad_norm": 4.0686516761779785, "learning_rate": 1.9329029450166837e-05, "loss": 0.0393, "step": 1025 }, { "epoch": 0.11123156981786643, "grad_norm": 2.4703938961029053, "learning_rate": 1.932830407659945e-05, "loss": 0.0424, "step": 1026 }, { "epoch": 0.11133998265394622, "grad_norm": 1.588588833808899, "learning_rate": 1.9327578703032065e-05, "loss": 0.0306, "step": 1027 }, { "epoch": 0.11144839549002603, "grad_norm": 1.799839973449707, "learning_rate": 1.9326853329464677e-05, "loss": 0.0269, "step": 1028 }, { "epoch": 0.11155680832610582, "grad_norm": 1.8187702894210815, "learning_rate": 1.932612795589729e-05, "loss": 0.032, "step": 1029 }, { "epoch": 0.11166522116218561, "grad_norm": 3.2693564891815186, "learning_rate": 1.93254025823299e-05, "loss": 0.0395, "step": 1030 }, { "epoch": 0.1117736339982654, "grad_norm": 4.0324788093566895, "learning_rate": 1.9324677208762514e-05, "loss": 0.0615, "step": 1031 }, { "epoch": 0.11188204683434519, "grad_norm": 2.944573163986206, "learning_rate": 1.9323951835195126e-05, "loss": 0.0653, "step": 1032 }, { "epoch": 0.11199045967042498, "grad_norm": 3.648163318634033, "learning_rate": 1.932322646162774e-05, "loss": 0.0317, "step": 1033 }, { "epoch": 0.11209887250650477, "grad_norm": 4.361079216003418, "learning_rate": 1.9322501088060353e-05, "loss": 0.0737, "step": 1034 }, { "epoch": 0.11220728534258456, "grad_norm": 3.5698628425598145, "learning_rate": 1.9321775714492966e-05, "loss": 0.0836, "step": 1035 }, { "epoch": 0.11231569817866435, "grad_norm": 6.867100238800049, "learning_rate": 1.9321050340925578e-05, "loss": 0.0682, "step": 1036 }, { "epoch": 0.11242411101474414, "grad_norm": 4.285452842712402, "learning_rate": 1.932032496735819e-05, "loss": 0.053, "step": 1037 }, { "epoch": 0.11253252385082393, "grad_norm": 3.947122573852539, "learning_rate": 1.9319599593790802e-05, "loss": 0.0474, "step": 1038 }, { "epoch": 0.11264093668690373, "grad_norm": 2.4991767406463623, "learning_rate": 1.9318874220223418e-05, "loss": 0.0243, "step": 1039 }, { "epoch": 0.11274934952298352, "grad_norm": 3.5815367698669434, "learning_rate": 1.931814884665603e-05, "loss": 0.039, "step": 1040 }, { "epoch": 0.11285776235906331, "grad_norm": 7.264245986938477, "learning_rate": 1.9317423473088642e-05, "loss": 0.0909, "step": 1041 }, { "epoch": 0.1129661751951431, "grad_norm": 3.9370789527893066, "learning_rate": 1.9316698099521257e-05, "loss": 0.0357, "step": 1042 }, { "epoch": 0.1130745880312229, "grad_norm": 2.014901876449585, "learning_rate": 1.931597272595387e-05, "loss": 0.0414, "step": 1043 }, { "epoch": 0.1131830008673027, "grad_norm": 3.62974214553833, "learning_rate": 1.931524735238648e-05, "loss": 0.0783, "step": 1044 }, { "epoch": 0.11329141370338248, "grad_norm": 1.7013665437698364, "learning_rate": 1.9314521978819094e-05, "loss": 0.0342, "step": 1045 }, { "epoch": 0.11339982653946228, "grad_norm": 3.1625094413757324, "learning_rate": 1.9313796605251706e-05, "loss": 0.0559, "step": 1046 }, { "epoch": 0.11350823937554207, "grad_norm": 2.6888036727905273, "learning_rate": 1.9313071231684318e-05, "loss": 0.0577, "step": 1047 }, { "epoch": 0.11361665221162186, "grad_norm": 6.508138179779053, "learning_rate": 1.9312345858116934e-05, "loss": 0.071, "step": 1048 }, { "epoch": 0.11372506504770165, "grad_norm": 6.889479637145996, "learning_rate": 1.9311620484549546e-05, "loss": 0.0433, "step": 1049 }, { "epoch": 0.11383347788378144, "grad_norm": 5.9738874435424805, "learning_rate": 1.9310895110982158e-05, "loss": 0.0451, "step": 1050 }, { "epoch": 0.11394189071986123, "grad_norm": 1.835567593574524, "learning_rate": 1.931016973741477e-05, "loss": 0.0426, "step": 1051 }, { "epoch": 0.11405030355594102, "grad_norm": 3.3752517700195312, "learning_rate": 1.9309444363847382e-05, "loss": 0.0392, "step": 1052 }, { "epoch": 0.11415871639202081, "grad_norm": 3.3394083976745605, "learning_rate": 1.9308718990279994e-05, "loss": 0.045, "step": 1053 }, { "epoch": 0.1142671292281006, "grad_norm": 3.055692434310913, "learning_rate": 1.9307993616712606e-05, "loss": 0.0803, "step": 1054 }, { "epoch": 0.1143755420641804, "grad_norm": 3.366401433944702, "learning_rate": 1.9307268243145222e-05, "loss": 0.0544, "step": 1055 }, { "epoch": 0.11448395490026018, "grad_norm": 4.179691791534424, "learning_rate": 1.9306542869577834e-05, "loss": 0.0422, "step": 1056 }, { "epoch": 0.11459236773633999, "grad_norm": 4.466884613037109, "learning_rate": 1.9305817496010446e-05, "loss": 0.0269, "step": 1057 }, { "epoch": 0.11470078057241978, "grad_norm": 4.7068891525268555, "learning_rate": 1.930509212244306e-05, "loss": 0.0517, "step": 1058 }, { "epoch": 0.11480919340849957, "grad_norm": 3.970276355743408, "learning_rate": 1.9304366748875674e-05, "loss": 0.0951, "step": 1059 }, { "epoch": 0.11491760624457936, "grad_norm": 0.8652684688568115, "learning_rate": 1.9303641375308286e-05, "loss": 0.0062, "step": 1060 }, { "epoch": 0.11502601908065915, "grad_norm": 1.9369392395019531, "learning_rate": 1.9302916001740898e-05, "loss": 0.0486, "step": 1061 }, { "epoch": 0.11513443191673894, "grad_norm": 2.541285753250122, "learning_rate": 1.930219062817351e-05, "loss": 0.0548, "step": 1062 }, { "epoch": 0.11524284475281873, "grad_norm": 3.536526918411255, "learning_rate": 1.9301465254606126e-05, "loss": 0.0346, "step": 1063 }, { "epoch": 0.11535125758889853, "grad_norm": 23.606826782226562, "learning_rate": 1.9300739881038738e-05, "loss": 0.064, "step": 1064 }, { "epoch": 0.11545967042497832, "grad_norm": 5.904280185699463, "learning_rate": 1.930001450747135e-05, "loss": 0.1052, "step": 1065 }, { "epoch": 0.11556808326105811, "grad_norm": 4.764672756195068, "learning_rate": 1.9299289133903962e-05, "loss": 0.0956, "step": 1066 }, { "epoch": 0.1156764960971379, "grad_norm": 4.991963863372803, "learning_rate": 1.9298563760336574e-05, "loss": 0.0625, "step": 1067 }, { "epoch": 0.11578490893321769, "grad_norm": 6.016916275024414, "learning_rate": 1.9297838386769187e-05, "loss": 0.0637, "step": 1068 }, { "epoch": 0.11589332176929748, "grad_norm": 7.226493835449219, "learning_rate": 1.92971130132018e-05, "loss": 0.071, "step": 1069 }, { "epoch": 0.11600173460537727, "grad_norm": 1.1319265365600586, "learning_rate": 1.9296387639634414e-05, "loss": 0.0145, "step": 1070 }, { "epoch": 0.11611014744145706, "grad_norm": 3.3728537559509277, "learning_rate": 1.9295662266067026e-05, "loss": 0.0389, "step": 1071 }, { "epoch": 0.11621856027753687, "grad_norm": 2.944307804107666, "learning_rate": 1.929493689249964e-05, "loss": 0.071, "step": 1072 }, { "epoch": 0.11632697311361666, "grad_norm": 4.3437418937683105, "learning_rate": 1.929421151893225e-05, "loss": 0.0498, "step": 1073 }, { "epoch": 0.11643538594969645, "grad_norm": 4.058696746826172, "learning_rate": 1.9293486145364863e-05, "loss": 0.0495, "step": 1074 }, { "epoch": 0.11654379878577624, "grad_norm": 5.968066692352295, "learning_rate": 1.9292760771797475e-05, "loss": 0.0281, "step": 1075 }, { "epoch": 0.11665221162185603, "grad_norm": 2.6207613945007324, "learning_rate": 1.929203539823009e-05, "loss": 0.0396, "step": 1076 }, { "epoch": 0.11676062445793582, "grad_norm": 2.251448154449463, "learning_rate": 1.9291310024662703e-05, "loss": 0.02, "step": 1077 }, { "epoch": 0.11686903729401561, "grad_norm": 1.183027982711792, "learning_rate": 1.9290584651095315e-05, "loss": 0.0172, "step": 1078 }, { "epoch": 0.1169774501300954, "grad_norm": 1.5833767652511597, "learning_rate": 1.928985927752793e-05, "loss": 0.0481, "step": 1079 }, { "epoch": 0.1170858629661752, "grad_norm": 4.027816295623779, "learning_rate": 1.9289133903960543e-05, "loss": 0.0584, "step": 1080 }, { "epoch": 0.11719427580225499, "grad_norm": 3.0380747318267822, "learning_rate": 1.9288408530393155e-05, "loss": 0.0212, "step": 1081 }, { "epoch": 0.11730268863833478, "grad_norm": 2.854672431945801, "learning_rate": 1.9287683156825767e-05, "loss": 0.0368, "step": 1082 }, { "epoch": 0.11741110147441457, "grad_norm": 4.447218894958496, "learning_rate": 1.928695778325838e-05, "loss": 0.0652, "step": 1083 }, { "epoch": 0.11751951431049436, "grad_norm": 5.960728645324707, "learning_rate": 1.928623240969099e-05, "loss": 0.0764, "step": 1084 }, { "epoch": 0.11762792714657415, "grad_norm": 1.4092861413955688, "learning_rate": 1.9285507036123607e-05, "loss": 0.0255, "step": 1085 }, { "epoch": 0.11773633998265394, "grad_norm": 4.048739433288574, "learning_rate": 1.928478166255622e-05, "loss": 0.07, "step": 1086 }, { "epoch": 0.11784475281873374, "grad_norm": 3.2452545166015625, "learning_rate": 1.928405628898883e-05, "loss": 0.0432, "step": 1087 }, { "epoch": 0.11795316565481354, "grad_norm": 5.995724201202393, "learning_rate": 1.9283330915421443e-05, "loss": 0.0454, "step": 1088 }, { "epoch": 0.11806157849089333, "grad_norm": 2.860337734222412, "learning_rate": 1.9282605541854055e-05, "loss": 0.0236, "step": 1089 }, { "epoch": 0.11816999132697312, "grad_norm": 0.8466657400131226, "learning_rate": 1.9281880168286667e-05, "loss": 0.0183, "step": 1090 }, { "epoch": 0.11827840416305291, "grad_norm": 6.596236228942871, "learning_rate": 1.9281154794719283e-05, "loss": 0.0601, "step": 1091 }, { "epoch": 0.1183868169991327, "grad_norm": 5.504456520080566, "learning_rate": 1.9280429421151895e-05, "loss": 0.0647, "step": 1092 }, { "epoch": 0.11849522983521249, "grad_norm": 9.244415283203125, "learning_rate": 1.9279704047584507e-05, "loss": 0.0613, "step": 1093 }, { "epoch": 0.11860364267129228, "grad_norm": 5.415821552276611, "learning_rate": 1.9278978674017123e-05, "loss": 0.0688, "step": 1094 }, { "epoch": 0.11871205550737207, "grad_norm": 3.328974485397339, "learning_rate": 1.9278253300449735e-05, "loss": 0.0357, "step": 1095 }, { "epoch": 0.11882046834345186, "grad_norm": 5.373629093170166, "learning_rate": 1.9277527926882347e-05, "loss": 0.0553, "step": 1096 }, { "epoch": 0.11892888117953165, "grad_norm": 2.994518756866455, "learning_rate": 1.927680255331496e-05, "loss": 0.0213, "step": 1097 }, { "epoch": 0.11903729401561144, "grad_norm": 5.483613967895508, "learning_rate": 1.927607717974757e-05, "loss": 0.0778, "step": 1098 }, { "epoch": 0.11914570685169124, "grad_norm": 3.0589277744293213, "learning_rate": 1.9275351806180183e-05, "loss": 0.0482, "step": 1099 }, { "epoch": 0.11925411968777103, "grad_norm": 4.606000900268555, "learning_rate": 1.92746264326128e-05, "loss": 0.0599, "step": 1100 }, { "epoch": 0.11936253252385082, "grad_norm": 2.3170621395111084, "learning_rate": 1.927390105904541e-05, "loss": 0.0495, "step": 1101 }, { "epoch": 0.11947094535993062, "grad_norm": 1.7834346294403076, "learning_rate": 1.9273175685478023e-05, "loss": 0.0389, "step": 1102 }, { "epoch": 0.11957935819601041, "grad_norm": 2.1121456623077393, "learning_rate": 1.9272450311910635e-05, "loss": 0.0285, "step": 1103 }, { "epoch": 0.1196877710320902, "grad_norm": 2.7881906032562256, "learning_rate": 1.9271724938343248e-05, "loss": 0.0396, "step": 1104 }, { "epoch": 0.11979618386817, "grad_norm": 2.3865692615509033, "learning_rate": 1.927099956477586e-05, "loss": 0.0528, "step": 1105 }, { "epoch": 0.11990459670424979, "grad_norm": 6.020346641540527, "learning_rate": 1.9270274191208472e-05, "loss": 0.0453, "step": 1106 }, { "epoch": 0.12001300954032958, "grad_norm": 7.437219619750977, "learning_rate": 1.9269548817641087e-05, "loss": 0.0586, "step": 1107 }, { "epoch": 0.12012142237640937, "grad_norm": 1.579870343208313, "learning_rate": 1.92688234440737e-05, "loss": 0.0261, "step": 1108 }, { "epoch": 0.12022983521248916, "grad_norm": 2.082535982131958, "learning_rate": 1.926809807050631e-05, "loss": 0.0668, "step": 1109 }, { "epoch": 0.12033824804856895, "grad_norm": 1.8020262718200684, "learning_rate": 1.9267372696938924e-05, "loss": 0.0264, "step": 1110 }, { "epoch": 0.12044666088464874, "grad_norm": 1.170836091041565, "learning_rate": 1.926664732337154e-05, "loss": 0.032, "step": 1111 }, { "epoch": 0.12055507372072853, "grad_norm": 3.501936197280884, "learning_rate": 1.926592194980415e-05, "loss": 0.0614, "step": 1112 }, { "epoch": 0.12066348655680832, "grad_norm": 7.5757832527160645, "learning_rate": 1.9265196576236764e-05, "loss": 0.0796, "step": 1113 }, { "epoch": 0.12077189939288811, "grad_norm": 4.569133758544922, "learning_rate": 1.9264471202669376e-05, "loss": 0.0524, "step": 1114 }, { "epoch": 0.1208803122289679, "grad_norm": 10.534955978393555, "learning_rate": 1.926374582910199e-05, "loss": 0.0327, "step": 1115 }, { "epoch": 0.1209887250650477, "grad_norm": 4.449105739593506, "learning_rate": 1.9263020455534603e-05, "loss": 0.0942, "step": 1116 }, { "epoch": 0.1210971379011275, "grad_norm": 2.1609609127044678, "learning_rate": 1.9262295081967216e-05, "loss": 0.0279, "step": 1117 }, { "epoch": 0.12120555073720729, "grad_norm": 1.7409405708312988, "learning_rate": 1.9261569708399828e-05, "loss": 0.0529, "step": 1118 }, { "epoch": 0.12131396357328708, "grad_norm": 2.2967281341552734, "learning_rate": 1.926084433483244e-05, "loss": 0.0385, "step": 1119 }, { "epoch": 0.12142237640936687, "grad_norm": 5.816412925720215, "learning_rate": 1.9260118961265052e-05, "loss": 0.0485, "step": 1120 }, { "epoch": 0.12153078924544666, "grad_norm": 4.406930923461914, "learning_rate": 1.9259393587697664e-05, "loss": 0.0621, "step": 1121 }, { "epoch": 0.12163920208152645, "grad_norm": 4.63575553894043, "learning_rate": 1.925866821413028e-05, "loss": 0.0633, "step": 1122 }, { "epoch": 0.12174761491760624, "grad_norm": 16.90717124938965, "learning_rate": 1.9257942840562892e-05, "loss": 0.0347, "step": 1123 }, { "epoch": 0.12185602775368604, "grad_norm": 3.9032716751098633, "learning_rate": 1.9257217466995504e-05, "loss": 0.0542, "step": 1124 }, { "epoch": 0.12196444058976583, "grad_norm": 2.8389832973480225, "learning_rate": 1.9256492093428116e-05, "loss": 0.0502, "step": 1125 }, { "epoch": 0.12207285342584562, "grad_norm": 2.5625410079956055, "learning_rate": 1.9255766719860728e-05, "loss": 0.0424, "step": 1126 }, { "epoch": 0.12218126626192541, "grad_norm": 8.242104530334473, "learning_rate": 1.925504134629334e-05, "loss": 0.1151, "step": 1127 }, { "epoch": 0.1222896790980052, "grad_norm": 2.235353469848633, "learning_rate": 1.9254315972725956e-05, "loss": 0.031, "step": 1128 }, { "epoch": 0.12239809193408499, "grad_norm": 7.541988372802734, "learning_rate": 1.9253590599158568e-05, "loss": 0.0433, "step": 1129 }, { "epoch": 0.12250650477016478, "grad_norm": 3.4970531463623047, "learning_rate": 1.9252865225591184e-05, "loss": 0.0447, "step": 1130 }, { "epoch": 0.12261491760624459, "grad_norm": 1.354711890220642, "learning_rate": 1.9252139852023796e-05, "loss": 0.0367, "step": 1131 }, { "epoch": 0.12272333044232438, "grad_norm": 1.8160450458526611, "learning_rate": 1.9251414478456408e-05, "loss": 0.0281, "step": 1132 }, { "epoch": 0.12283174327840417, "grad_norm": 2.1734607219696045, "learning_rate": 1.925068910488902e-05, "loss": 0.0184, "step": 1133 }, { "epoch": 0.12294015611448396, "grad_norm": 3.931589126586914, "learning_rate": 1.9249963731321632e-05, "loss": 0.0303, "step": 1134 }, { "epoch": 0.12304856895056375, "grad_norm": 2.3039729595184326, "learning_rate": 1.9249238357754244e-05, "loss": 0.0489, "step": 1135 }, { "epoch": 0.12315698178664354, "grad_norm": 3.399628162384033, "learning_rate": 1.9248512984186856e-05, "loss": 0.0654, "step": 1136 }, { "epoch": 0.12326539462272333, "grad_norm": 5.775940895080566, "learning_rate": 1.9247787610619472e-05, "loss": 0.027, "step": 1137 }, { "epoch": 0.12337380745880312, "grad_norm": 1.8424224853515625, "learning_rate": 1.9247062237052084e-05, "loss": 0.0198, "step": 1138 }, { "epoch": 0.12348222029488291, "grad_norm": 5.8622822761535645, "learning_rate": 1.9246336863484696e-05, "loss": 0.0812, "step": 1139 }, { "epoch": 0.1235906331309627, "grad_norm": 4.230886459350586, "learning_rate": 1.924561148991731e-05, "loss": 0.0315, "step": 1140 }, { "epoch": 0.1236990459670425, "grad_norm": 0.8386187553405762, "learning_rate": 1.924488611634992e-05, "loss": 0.0182, "step": 1141 }, { "epoch": 0.12380745880312229, "grad_norm": 4.419314384460449, "learning_rate": 1.9244160742782533e-05, "loss": 0.0302, "step": 1142 }, { "epoch": 0.12391587163920208, "grad_norm": 3.4525835514068604, "learning_rate": 1.9243435369215148e-05, "loss": 0.0485, "step": 1143 }, { "epoch": 0.12402428447528187, "grad_norm": 9.753133773803711, "learning_rate": 1.924270999564776e-05, "loss": 0.0937, "step": 1144 }, { "epoch": 0.12413269731136166, "grad_norm": 3.947042226791382, "learning_rate": 1.9241984622080372e-05, "loss": 0.0217, "step": 1145 }, { "epoch": 0.12424111014744146, "grad_norm": 10.208592414855957, "learning_rate": 1.9241259248512988e-05, "loss": 0.1327, "step": 1146 }, { "epoch": 0.12434952298352125, "grad_norm": 7.569797039031982, "learning_rate": 1.92405338749456e-05, "loss": 0.0819, "step": 1147 }, { "epoch": 0.12445793581960105, "grad_norm": 3.3138656616210938, "learning_rate": 1.9239808501378212e-05, "loss": 0.018, "step": 1148 }, { "epoch": 0.12456634865568084, "grad_norm": 1.298483967781067, "learning_rate": 1.9239083127810824e-05, "loss": 0.0077, "step": 1149 }, { "epoch": 0.12467476149176063, "grad_norm": 2.76349139213562, "learning_rate": 1.9238357754243437e-05, "loss": 0.0206, "step": 1150 }, { "epoch": 0.12478317432784042, "grad_norm": 1.3342387676239014, "learning_rate": 1.923763238067605e-05, "loss": 0.0196, "step": 1151 }, { "epoch": 0.12489158716392021, "grad_norm": 8.840266227722168, "learning_rate": 1.9236907007108664e-05, "loss": 0.0707, "step": 1152 }, { "epoch": 0.125, "grad_norm": 3.6214404106140137, "learning_rate": 1.9236181633541276e-05, "loss": 0.0807, "step": 1153 }, { "epoch": 0.1251084128360798, "grad_norm": 2.848018169403076, "learning_rate": 1.923545625997389e-05, "loss": 0.0637, "step": 1154 }, { "epoch": 0.12521682567215958, "grad_norm": 1.217905879020691, "learning_rate": 1.92347308864065e-05, "loss": 0.0268, "step": 1155 }, { "epoch": 0.12532523850823937, "grad_norm": 3.9987759590148926, "learning_rate": 1.9234005512839113e-05, "loss": 0.0533, "step": 1156 }, { "epoch": 0.12543365134431916, "grad_norm": 1.8355499505996704, "learning_rate": 1.9233280139271725e-05, "loss": 0.0358, "step": 1157 }, { "epoch": 0.12554206418039895, "grad_norm": 8.62895393371582, "learning_rate": 1.9232554765704337e-05, "loss": 0.0593, "step": 1158 }, { "epoch": 0.12565047701647875, "grad_norm": 10.064172744750977, "learning_rate": 1.9231829392136953e-05, "loss": 0.02, "step": 1159 }, { "epoch": 0.12575888985255854, "grad_norm": 7.585537433624268, "learning_rate": 1.9231104018569565e-05, "loss": 0.0961, "step": 1160 }, { "epoch": 0.12586730268863833, "grad_norm": 4.459357738494873, "learning_rate": 1.9230378645002177e-05, "loss": 0.036, "step": 1161 }, { "epoch": 0.12597571552471812, "grad_norm": 3.714052438735962, "learning_rate": 1.922965327143479e-05, "loss": 0.0147, "step": 1162 }, { "epoch": 0.1260841283607979, "grad_norm": 5.57831335067749, "learning_rate": 1.9228927897867405e-05, "loss": 0.0597, "step": 1163 }, { "epoch": 0.1261925411968777, "grad_norm": 3.8941125869750977, "learning_rate": 1.9228202524300017e-05, "loss": 0.0379, "step": 1164 }, { "epoch": 0.1263009540329575, "grad_norm": 2.306638240814209, "learning_rate": 1.922747715073263e-05, "loss": 0.0194, "step": 1165 }, { "epoch": 0.12640936686903728, "grad_norm": 4.569910526275635, "learning_rate": 1.922675177716524e-05, "loss": 0.0669, "step": 1166 }, { "epoch": 0.12651777970511707, "grad_norm": 3.347764253616333, "learning_rate": 1.9226026403597857e-05, "loss": 0.0408, "step": 1167 }, { "epoch": 0.1266261925411969, "grad_norm": 3.9203906059265137, "learning_rate": 1.922530103003047e-05, "loss": 0.0541, "step": 1168 }, { "epoch": 0.12673460537727668, "grad_norm": 2.8156092166900635, "learning_rate": 1.922457565646308e-05, "loss": 0.0329, "step": 1169 }, { "epoch": 0.12684301821335647, "grad_norm": 3.584484577178955, "learning_rate": 1.9223850282895693e-05, "loss": 0.0585, "step": 1170 }, { "epoch": 0.12695143104943626, "grad_norm": 2.0831005573272705, "learning_rate": 1.9223124909328305e-05, "loss": 0.0285, "step": 1171 }, { "epoch": 0.12705984388551606, "grad_norm": 4.939266681671143, "learning_rate": 1.9222399535760917e-05, "loss": 0.027, "step": 1172 }, { "epoch": 0.12716825672159585, "grad_norm": 1.9977176189422607, "learning_rate": 1.922167416219353e-05, "loss": 0.0349, "step": 1173 }, { "epoch": 0.12727666955767564, "grad_norm": 4.450558662414551, "learning_rate": 1.9220948788626145e-05, "loss": 0.0418, "step": 1174 }, { "epoch": 0.12738508239375543, "grad_norm": 1.5738935470581055, "learning_rate": 1.9220223415058757e-05, "loss": 0.0109, "step": 1175 }, { "epoch": 0.12749349522983522, "grad_norm": 4.013413429260254, "learning_rate": 1.921949804149137e-05, "loss": 0.0763, "step": 1176 }, { "epoch": 0.127601908065915, "grad_norm": 3.448164224624634, "learning_rate": 1.921877266792398e-05, "loss": 0.0768, "step": 1177 }, { "epoch": 0.1277103209019948, "grad_norm": 3.4096570014953613, "learning_rate": 1.9218047294356594e-05, "loss": 0.0288, "step": 1178 }, { "epoch": 0.1278187337380746, "grad_norm": 5.576076507568359, "learning_rate": 1.9217321920789206e-05, "loss": 0.0454, "step": 1179 }, { "epoch": 0.12792714657415438, "grad_norm": 4.393715858459473, "learning_rate": 1.921659654722182e-05, "loss": 0.0312, "step": 1180 }, { "epoch": 0.12803555941023417, "grad_norm": 11.054884910583496, "learning_rate": 1.9215871173654433e-05, "loss": 0.06, "step": 1181 }, { "epoch": 0.12814397224631396, "grad_norm": 4.296885013580322, "learning_rate": 1.921514580008705e-05, "loss": 0.0562, "step": 1182 }, { "epoch": 0.12825238508239376, "grad_norm": 2.8200554847717285, "learning_rate": 1.921442042651966e-05, "loss": 0.0478, "step": 1183 }, { "epoch": 0.12836079791847355, "grad_norm": 2.2711334228515625, "learning_rate": 1.9213695052952273e-05, "loss": 0.0613, "step": 1184 }, { "epoch": 0.12846921075455334, "grad_norm": 7.597480297088623, "learning_rate": 1.9212969679384885e-05, "loss": 0.0594, "step": 1185 }, { "epoch": 0.12857762359063313, "grad_norm": 5.389342308044434, "learning_rate": 1.9212244305817497e-05, "loss": 0.0972, "step": 1186 }, { "epoch": 0.12868603642671292, "grad_norm": 4.004721164703369, "learning_rate": 1.921151893225011e-05, "loss": 0.0308, "step": 1187 }, { "epoch": 0.1287944492627927, "grad_norm": 2.669168710708618, "learning_rate": 1.9210793558682722e-05, "loss": 0.0305, "step": 1188 }, { "epoch": 0.1289028620988725, "grad_norm": 3.514125108718872, "learning_rate": 1.9210068185115337e-05, "loss": 0.0625, "step": 1189 }, { "epoch": 0.1290112749349523, "grad_norm": 1.4406317472457886, "learning_rate": 1.920934281154795e-05, "loss": 0.0186, "step": 1190 }, { "epoch": 0.12911968777103208, "grad_norm": 2.402221202850342, "learning_rate": 1.920861743798056e-05, "loss": 0.0352, "step": 1191 }, { "epoch": 0.12922810060711187, "grad_norm": 2.093083620071411, "learning_rate": 1.9207892064413174e-05, "loss": 0.0653, "step": 1192 }, { "epoch": 0.12933651344319166, "grad_norm": 9.45799732208252, "learning_rate": 1.9207166690845786e-05, "loss": 0.0681, "step": 1193 }, { "epoch": 0.12944492627927146, "grad_norm": 3.914604902267456, "learning_rate": 1.9206441317278398e-05, "loss": 0.0494, "step": 1194 }, { "epoch": 0.12955333911535125, "grad_norm": 3.4606213569641113, "learning_rate": 1.9205715943711014e-05, "loss": 0.0438, "step": 1195 }, { "epoch": 0.12966175195143104, "grad_norm": 3.6632673740386963, "learning_rate": 1.9204990570143626e-05, "loss": 0.0521, "step": 1196 }, { "epoch": 0.12977016478751083, "grad_norm": 2.7598891258239746, "learning_rate": 1.9204265196576238e-05, "loss": 0.054, "step": 1197 }, { "epoch": 0.12987857762359065, "grad_norm": 2.0078623294830322, "learning_rate": 1.9203539823008853e-05, "loss": 0.0333, "step": 1198 }, { "epoch": 0.12998699045967044, "grad_norm": 2.103442668914795, "learning_rate": 1.9202814449441466e-05, "loss": 0.0278, "step": 1199 }, { "epoch": 0.13009540329575023, "grad_norm": 2.1172313690185547, "learning_rate": 1.9202089075874078e-05, "loss": 0.0194, "step": 1200 }, { "epoch": 0.13020381613183002, "grad_norm": 8.887829780578613, "learning_rate": 1.920136370230669e-05, "loss": 0.1277, "step": 1201 }, { "epoch": 0.1303122289679098, "grad_norm": 1.6509336233139038, "learning_rate": 1.9200638328739302e-05, "loss": 0.032, "step": 1202 }, { "epoch": 0.1304206418039896, "grad_norm": 8.903374671936035, "learning_rate": 1.9199912955171914e-05, "loss": 0.0734, "step": 1203 }, { "epoch": 0.1305290546400694, "grad_norm": 6.450918674468994, "learning_rate": 1.919918758160453e-05, "loss": 0.0486, "step": 1204 }, { "epoch": 0.13063746747614918, "grad_norm": 2.8837990760803223, "learning_rate": 1.9198462208037142e-05, "loss": 0.0541, "step": 1205 }, { "epoch": 0.13074588031222897, "grad_norm": 4.057215213775635, "learning_rate": 1.9197736834469754e-05, "loss": 0.0454, "step": 1206 }, { "epoch": 0.13085429314830876, "grad_norm": 2.121990203857422, "learning_rate": 1.9197011460902366e-05, "loss": 0.0312, "step": 1207 }, { "epoch": 0.13096270598438856, "grad_norm": 1.98967444896698, "learning_rate": 1.9196286087334978e-05, "loss": 0.0374, "step": 1208 }, { "epoch": 0.13107111882046835, "grad_norm": 3.0927000045776367, "learning_rate": 1.919556071376759e-05, "loss": 0.028, "step": 1209 }, { "epoch": 0.13117953165654814, "grad_norm": 8.945293426513672, "learning_rate": 1.9194835340200202e-05, "loss": 0.1318, "step": 1210 }, { "epoch": 0.13128794449262793, "grad_norm": 3.721538543701172, "learning_rate": 1.9194109966632818e-05, "loss": 0.059, "step": 1211 }, { "epoch": 0.13139635732870772, "grad_norm": 4.9199347496032715, "learning_rate": 1.919338459306543e-05, "loss": 0.0576, "step": 1212 }, { "epoch": 0.1315047701647875, "grad_norm": 2.2870566844940186, "learning_rate": 1.9192659219498042e-05, "loss": 0.0583, "step": 1213 }, { "epoch": 0.1316131830008673, "grad_norm": 1.663770079612732, "learning_rate": 1.9191933845930654e-05, "loss": 0.0098, "step": 1214 }, { "epoch": 0.1317215958369471, "grad_norm": 1.9058054685592651, "learning_rate": 1.919120847236327e-05, "loss": 0.0168, "step": 1215 }, { "epoch": 0.13183000867302688, "grad_norm": 7.665785312652588, "learning_rate": 1.9190483098795882e-05, "loss": 0.0943, "step": 1216 }, { "epoch": 0.13193842150910667, "grad_norm": 3.545429229736328, "learning_rate": 1.9189757725228494e-05, "loss": 0.033, "step": 1217 }, { "epoch": 0.13204683434518646, "grad_norm": 18.907554626464844, "learning_rate": 1.9189032351661106e-05, "loss": 0.0648, "step": 1218 }, { "epoch": 0.13215524718126626, "grad_norm": 19.52834129333496, "learning_rate": 1.9188306978093722e-05, "loss": 0.0688, "step": 1219 }, { "epoch": 0.13226366001734605, "grad_norm": 3.4156417846679688, "learning_rate": 1.9187581604526334e-05, "loss": 0.0356, "step": 1220 }, { "epoch": 0.13237207285342584, "grad_norm": 2.088742733001709, "learning_rate": 1.9186856230958946e-05, "loss": 0.0202, "step": 1221 }, { "epoch": 0.13248048568950563, "grad_norm": 3.4974536895751953, "learning_rate": 1.918613085739156e-05, "loss": 0.0698, "step": 1222 }, { "epoch": 0.13258889852558542, "grad_norm": 7.816551685333252, "learning_rate": 1.918540548382417e-05, "loss": 0.0718, "step": 1223 }, { "epoch": 0.1326973113616652, "grad_norm": 4.824656963348389, "learning_rate": 1.9184680110256783e-05, "loss": 0.0954, "step": 1224 }, { "epoch": 0.132805724197745, "grad_norm": 5.53004264831543, "learning_rate": 1.9183954736689395e-05, "loss": 0.0616, "step": 1225 }, { "epoch": 0.1329141370338248, "grad_norm": 1.4525220394134521, "learning_rate": 1.918322936312201e-05, "loss": 0.0236, "step": 1226 }, { "epoch": 0.1330225498699046, "grad_norm": 2.851175308227539, "learning_rate": 1.9182503989554622e-05, "loss": 0.0211, "step": 1227 }, { "epoch": 0.1331309627059844, "grad_norm": 12.310846328735352, "learning_rate": 1.9181778615987235e-05, "loss": 0.0708, "step": 1228 }, { "epoch": 0.1332393755420642, "grad_norm": 7.336771488189697, "learning_rate": 1.9181053242419847e-05, "loss": 0.0587, "step": 1229 }, { "epoch": 0.13334778837814398, "grad_norm": 4.415979862213135, "learning_rate": 1.918032786885246e-05, "loss": 0.0829, "step": 1230 }, { "epoch": 0.13345620121422377, "grad_norm": 9.758378982543945, "learning_rate": 1.917960249528507e-05, "loss": 0.0593, "step": 1231 }, { "epoch": 0.13356461405030357, "grad_norm": 7.299767017364502, "learning_rate": 1.9178877121717687e-05, "loss": 0.0259, "step": 1232 }, { "epoch": 0.13367302688638336, "grad_norm": 2.2551581859588623, "learning_rate": 1.91781517481503e-05, "loss": 0.0466, "step": 1233 }, { "epoch": 0.13378143972246315, "grad_norm": 6.570895195007324, "learning_rate": 1.9177426374582914e-05, "loss": 0.0679, "step": 1234 }, { "epoch": 0.13388985255854294, "grad_norm": 3.91605281829834, "learning_rate": 1.9176701001015526e-05, "loss": 0.1063, "step": 1235 }, { "epoch": 0.13399826539462273, "grad_norm": 3.386834144592285, "learning_rate": 1.917597562744814e-05, "loss": 0.0293, "step": 1236 }, { "epoch": 0.13410667823070252, "grad_norm": 1.0158891677856445, "learning_rate": 1.917525025388075e-05, "loss": 0.0382, "step": 1237 }, { "epoch": 0.1342150910667823, "grad_norm": 2.439605474472046, "learning_rate": 1.9174524880313363e-05, "loss": 0.0193, "step": 1238 }, { "epoch": 0.1343235039028621, "grad_norm": 5.837812423706055, "learning_rate": 1.9173799506745975e-05, "loss": 0.0381, "step": 1239 }, { "epoch": 0.1344319167389419, "grad_norm": 2.584083318710327, "learning_rate": 1.9173074133178587e-05, "loss": 0.0366, "step": 1240 }, { "epoch": 0.13454032957502168, "grad_norm": 2.913602590560913, "learning_rate": 1.9172348759611203e-05, "loss": 0.0622, "step": 1241 }, { "epoch": 0.13464874241110147, "grad_norm": 4.550890922546387, "learning_rate": 1.9171623386043815e-05, "loss": 0.0322, "step": 1242 }, { "epoch": 0.13475715524718127, "grad_norm": 2.3694815635681152, "learning_rate": 1.9170898012476427e-05, "loss": 0.0291, "step": 1243 }, { "epoch": 0.13486556808326106, "grad_norm": 1.5211050510406494, "learning_rate": 1.917017263890904e-05, "loss": 0.0207, "step": 1244 }, { "epoch": 0.13497398091934085, "grad_norm": 0.8549856543540955, "learning_rate": 1.916944726534165e-05, "loss": 0.0131, "step": 1245 }, { "epoch": 0.13508239375542064, "grad_norm": 4.153370380401611, "learning_rate": 1.9168721891774263e-05, "loss": 0.05, "step": 1246 }, { "epoch": 0.13519080659150043, "grad_norm": 4.2538018226623535, "learning_rate": 1.916799651820688e-05, "loss": 0.0261, "step": 1247 }, { "epoch": 0.13529921942758022, "grad_norm": 3.1326849460601807, "learning_rate": 1.916727114463949e-05, "loss": 0.053, "step": 1248 }, { "epoch": 0.13540763226366, "grad_norm": 6.884396553039551, "learning_rate": 1.9166545771072103e-05, "loss": 0.0511, "step": 1249 }, { "epoch": 0.1355160450997398, "grad_norm": 4.385169506072998, "learning_rate": 1.916582039750472e-05, "loss": 0.0603, "step": 1250 }, { "epoch": 0.1356244579358196, "grad_norm": 2.732043981552124, "learning_rate": 1.916509502393733e-05, "loss": 0.034, "step": 1251 }, { "epoch": 0.13573287077189938, "grad_norm": 2.6808526515960693, "learning_rate": 1.9164369650369943e-05, "loss": 0.038, "step": 1252 }, { "epoch": 0.13584128360797917, "grad_norm": 5.436823844909668, "learning_rate": 1.9163644276802555e-05, "loss": 0.0857, "step": 1253 }, { "epoch": 0.13594969644405897, "grad_norm": 2.881376028060913, "learning_rate": 1.9162918903235167e-05, "loss": 0.0364, "step": 1254 }, { "epoch": 0.13605810928013876, "grad_norm": 5.708237171173096, "learning_rate": 1.916219352966778e-05, "loss": 0.0182, "step": 1255 }, { "epoch": 0.13616652211621855, "grad_norm": 5.147614002227783, "learning_rate": 1.9161468156100395e-05, "loss": 0.0802, "step": 1256 }, { "epoch": 0.13627493495229837, "grad_norm": 36.53892517089844, "learning_rate": 1.9160742782533007e-05, "loss": 0.056, "step": 1257 }, { "epoch": 0.13638334778837816, "grad_norm": 3.195683240890503, "learning_rate": 1.916001740896562e-05, "loss": 0.0384, "step": 1258 }, { "epoch": 0.13649176062445795, "grad_norm": 3.981727361679077, "learning_rate": 1.915929203539823e-05, "loss": 0.0916, "step": 1259 }, { "epoch": 0.13660017346053774, "grad_norm": 6.221698760986328, "learning_rate": 1.9158566661830844e-05, "loss": 0.0972, "step": 1260 }, { "epoch": 0.13670858629661753, "grad_norm": 9.651379585266113, "learning_rate": 1.9157841288263456e-05, "loss": 0.0471, "step": 1261 }, { "epoch": 0.13681699913269732, "grad_norm": 3.821098804473877, "learning_rate": 1.9157115914696068e-05, "loss": 0.0511, "step": 1262 }, { "epoch": 0.1369254119687771, "grad_norm": 2.459247589111328, "learning_rate": 1.9156390541128683e-05, "loss": 0.0354, "step": 1263 }, { "epoch": 0.1370338248048569, "grad_norm": 3.0561373233795166, "learning_rate": 1.9155665167561295e-05, "loss": 0.0602, "step": 1264 }, { "epoch": 0.1371422376409367, "grad_norm": 2.36387300491333, "learning_rate": 1.9154939793993908e-05, "loss": 0.0219, "step": 1265 }, { "epoch": 0.13725065047701648, "grad_norm": 3.186690330505371, "learning_rate": 1.915421442042652e-05, "loss": 0.0753, "step": 1266 }, { "epoch": 0.13735906331309627, "grad_norm": 2.263157367706299, "learning_rate": 1.9153489046859135e-05, "loss": 0.025, "step": 1267 }, { "epoch": 0.13746747614917607, "grad_norm": 5.596763610839844, "learning_rate": 1.9152763673291747e-05, "loss": 0.0547, "step": 1268 }, { "epoch": 0.13757588898525586, "grad_norm": 7.394713401794434, "learning_rate": 1.915203829972436e-05, "loss": 0.1054, "step": 1269 }, { "epoch": 0.13768430182133565, "grad_norm": 2.2678916454315186, "learning_rate": 1.9151312926156972e-05, "loss": 0.0352, "step": 1270 }, { "epoch": 0.13779271465741544, "grad_norm": 2.771493434906006, "learning_rate": 1.9150587552589587e-05, "loss": 0.0521, "step": 1271 }, { "epoch": 0.13790112749349523, "grad_norm": 2.2982516288757324, "learning_rate": 1.91498621790222e-05, "loss": 0.0355, "step": 1272 }, { "epoch": 0.13800954032957502, "grad_norm": 2.0687367916107178, "learning_rate": 1.914913680545481e-05, "loss": 0.0332, "step": 1273 }, { "epoch": 0.1381179531656548, "grad_norm": 2.398897409439087, "learning_rate": 1.9148411431887424e-05, "loss": 0.0326, "step": 1274 }, { "epoch": 0.1382263660017346, "grad_norm": 2.0471372604370117, "learning_rate": 1.9147686058320036e-05, "loss": 0.0287, "step": 1275 }, { "epoch": 0.1383347788378144, "grad_norm": 1.810795783996582, "learning_rate": 1.9146960684752648e-05, "loss": 0.0412, "step": 1276 }, { "epoch": 0.13844319167389418, "grad_norm": 5.257486343383789, "learning_rate": 1.914623531118526e-05, "loss": 0.0588, "step": 1277 }, { "epoch": 0.13855160450997397, "grad_norm": 0.8870477676391602, "learning_rate": 1.9145509937617876e-05, "loss": 0.0251, "step": 1278 }, { "epoch": 0.13866001734605377, "grad_norm": 2.560333013534546, "learning_rate": 1.9144784564050488e-05, "loss": 0.025, "step": 1279 }, { "epoch": 0.13876843018213356, "grad_norm": 2.977498769760132, "learning_rate": 1.91440591904831e-05, "loss": 0.0493, "step": 1280 }, { "epoch": 0.13887684301821335, "grad_norm": 1.3191267251968384, "learning_rate": 1.9143333816915712e-05, "loss": 0.0322, "step": 1281 }, { "epoch": 0.13898525585429314, "grad_norm": 10.610727310180664, "learning_rate": 1.9142608443348324e-05, "loss": 0.0607, "step": 1282 }, { "epoch": 0.13909366869037293, "grad_norm": 3.0712146759033203, "learning_rate": 1.9141883069780936e-05, "loss": 0.0269, "step": 1283 }, { "epoch": 0.13920208152645272, "grad_norm": 0.9590259194374084, "learning_rate": 1.9141157696213552e-05, "loss": 0.0064, "step": 1284 }, { "epoch": 0.1393104943625325, "grad_norm": 2.5050888061523438, "learning_rate": 1.9140432322646164e-05, "loss": 0.0319, "step": 1285 }, { "epoch": 0.1394189071986123, "grad_norm": 3.8101646900177, "learning_rate": 1.913970694907878e-05, "loss": 0.0982, "step": 1286 }, { "epoch": 0.13952732003469212, "grad_norm": 3.402686595916748, "learning_rate": 1.9138981575511392e-05, "loss": 0.0631, "step": 1287 }, { "epoch": 0.1396357328707719, "grad_norm": 1.9755038022994995, "learning_rate": 1.9138256201944004e-05, "loss": 0.027, "step": 1288 }, { "epoch": 0.1397441457068517, "grad_norm": 6.46635627746582, "learning_rate": 1.9137530828376616e-05, "loss": 0.0672, "step": 1289 }, { "epoch": 0.1398525585429315, "grad_norm": 6.0683746337890625, "learning_rate": 1.9136805454809228e-05, "loss": 0.1465, "step": 1290 }, { "epoch": 0.13996097137901128, "grad_norm": 3.5179128646850586, "learning_rate": 1.913608008124184e-05, "loss": 0.0321, "step": 1291 }, { "epoch": 0.14006938421509108, "grad_norm": 5.485747814178467, "learning_rate": 1.9135354707674452e-05, "loss": 0.0652, "step": 1292 }, { "epoch": 0.14017779705117087, "grad_norm": 7.0771613121032715, "learning_rate": 1.9134629334107068e-05, "loss": 0.1348, "step": 1293 }, { "epoch": 0.14028620988725066, "grad_norm": 1.611924409866333, "learning_rate": 1.913390396053968e-05, "loss": 0.0237, "step": 1294 }, { "epoch": 0.14039462272333045, "grad_norm": 5.149456977844238, "learning_rate": 1.9133178586972292e-05, "loss": 0.032, "step": 1295 }, { "epoch": 0.14050303555941024, "grad_norm": 6.291048526763916, "learning_rate": 1.9132453213404904e-05, "loss": 0.1058, "step": 1296 }, { "epoch": 0.14061144839549003, "grad_norm": 5.7909040451049805, "learning_rate": 1.9131727839837517e-05, "loss": 0.062, "step": 1297 }, { "epoch": 0.14071986123156982, "grad_norm": 12.452860832214355, "learning_rate": 1.913100246627013e-05, "loss": 0.0608, "step": 1298 }, { "epoch": 0.1408282740676496, "grad_norm": 5.489335060119629, "learning_rate": 1.9130277092702744e-05, "loss": 0.0676, "step": 1299 }, { "epoch": 0.1409366869037294, "grad_norm": 0.8968993425369263, "learning_rate": 1.9129551719135356e-05, "loss": 0.0187, "step": 1300 }, { "epoch": 0.1410450997398092, "grad_norm": 6.906027317047119, "learning_rate": 1.912882634556797e-05, "loss": 0.0494, "step": 1301 }, { "epoch": 0.14115351257588898, "grad_norm": 1.3072197437286377, "learning_rate": 1.9128100972000584e-05, "loss": 0.0244, "step": 1302 }, { "epoch": 0.14126192541196878, "grad_norm": 5.167081832885742, "learning_rate": 1.9127375598433196e-05, "loss": 0.0271, "step": 1303 }, { "epoch": 0.14137033824804857, "grad_norm": 3.680100917816162, "learning_rate": 1.912665022486581e-05, "loss": 0.0509, "step": 1304 }, { "epoch": 0.14147875108412836, "grad_norm": 8.463040351867676, "learning_rate": 1.912592485129842e-05, "loss": 0.0829, "step": 1305 }, { "epoch": 0.14158716392020815, "grad_norm": 3.531355619430542, "learning_rate": 1.9125199477731033e-05, "loss": 0.0512, "step": 1306 }, { "epoch": 0.14169557675628794, "grad_norm": 11.488526344299316, "learning_rate": 1.9124474104163645e-05, "loss": 0.0234, "step": 1307 }, { "epoch": 0.14180398959236773, "grad_norm": 4.253864765167236, "learning_rate": 1.912374873059626e-05, "loss": 0.0834, "step": 1308 }, { "epoch": 0.14191240242844752, "grad_norm": 0.507516622543335, "learning_rate": 1.9123023357028872e-05, "loss": 0.0091, "step": 1309 }, { "epoch": 0.1420208152645273, "grad_norm": 2.2670392990112305, "learning_rate": 1.9122297983461485e-05, "loss": 0.0404, "step": 1310 }, { "epoch": 0.1421292281006071, "grad_norm": 1.2955973148345947, "learning_rate": 1.9121572609894097e-05, "loss": 0.0258, "step": 1311 }, { "epoch": 0.1422376409366869, "grad_norm": 1.4095816612243652, "learning_rate": 1.912084723632671e-05, "loss": 0.0138, "step": 1312 }, { "epoch": 0.14234605377276668, "grad_norm": 2.918628454208374, "learning_rate": 1.912012186275932e-05, "loss": 0.0266, "step": 1313 }, { "epoch": 0.14245446660884648, "grad_norm": 1.5921732187271118, "learning_rate": 1.9119396489191933e-05, "loss": 0.021, "step": 1314 }, { "epoch": 0.14256287944492627, "grad_norm": 3.776434898376465, "learning_rate": 1.911867111562455e-05, "loss": 0.0824, "step": 1315 }, { "epoch": 0.14267129228100608, "grad_norm": 0.9290714859962463, "learning_rate": 1.911794574205716e-05, "loss": 0.0227, "step": 1316 }, { "epoch": 0.14277970511708588, "grad_norm": 6.571296691894531, "learning_rate": 1.9117220368489773e-05, "loss": 0.0775, "step": 1317 }, { "epoch": 0.14288811795316567, "grad_norm": 5.78933572769165, "learning_rate": 1.911649499492239e-05, "loss": 0.0546, "step": 1318 }, { "epoch": 0.14299653078924546, "grad_norm": 4.875555515289307, "learning_rate": 1.9115769621355e-05, "loss": 0.0408, "step": 1319 }, { "epoch": 0.14310494362532525, "grad_norm": 3.8488211631774902, "learning_rate": 1.9115044247787613e-05, "loss": 0.08, "step": 1320 }, { "epoch": 0.14321335646140504, "grad_norm": 1.904443621635437, "learning_rate": 1.9114318874220225e-05, "loss": 0.0264, "step": 1321 }, { "epoch": 0.14332176929748483, "grad_norm": 4.940797328948975, "learning_rate": 1.9113593500652837e-05, "loss": 0.1269, "step": 1322 }, { "epoch": 0.14343018213356462, "grad_norm": 6.931077480316162, "learning_rate": 1.9112868127085453e-05, "loss": 0.1028, "step": 1323 }, { "epoch": 0.1435385949696444, "grad_norm": 0.5539878606796265, "learning_rate": 1.9112142753518065e-05, "loss": 0.0053, "step": 1324 }, { "epoch": 0.1436470078057242, "grad_norm": 3.3218047618865967, "learning_rate": 1.9111417379950677e-05, "loss": 0.0327, "step": 1325 }, { "epoch": 0.143755420641804, "grad_norm": 1.6470084190368652, "learning_rate": 1.911069200638329e-05, "loss": 0.0227, "step": 1326 }, { "epoch": 0.14386383347788378, "grad_norm": 4.072000026702881, "learning_rate": 1.91099666328159e-05, "loss": 0.0954, "step": 1327 }, { "epoch": 0.14397224631396358, "grad_norm": 6.917088508605957, "learning_rate": 1.9109241259248513e-05, "loss": 0.1217, "step": 1328 }, { "epoch": 0.14408065915004337, "grad_norm": 6.24390983581543, "learning_rate": 1.9108515885681125e-05, "loss": 0.0373, "step": 1329 }, { "epoch": 0.14418907198612316, "grad_norm": 2.6872828006744385, "learning_rate": 1.910779051211374e-05, "loss": 0.0446, "step": 1330 }, { "epoch": 0.14429748482220295, "grad_norm": 2.2917144298553467, "learning_rate": 1.9107065138546353e-05, "loss": 0.0507, "step": 1331 }, { "epoch": 0.14440589765828274, "grad_norm": 1.4356447458267212, "learning_rate": 1.9106339764978965e-05, "loss": 0.0392, "step": 1332 }, { "epoch": 0.14451431049436253, "grad_norm": 5.2131781578063965, "learning_rate": 1.9105614391411577e-05, "loss": 0.0952, "step": 1333 }, { "epoch": 0.14462272333044232, "grad_norm": 4.9826507568359375, "learning_rate": 1.910488901784419e-05, "loss": 0.0411, "step": 1334 }, { "epoch": 0.1447311361665221, "grad_norm": 2.603278636932373, "learning_rate": 1.9104163644276802e-05, "loss": 0.0246, "step": 1335 }, { "epoch": 0.1448395490026019, "grad_norm": 4.757152080535889, "learning_rate": 1.9103438270709417e-05, "loss": 0.0642, "step": 1336 }, { "epoch": 0.1449479618386817, "grad_norm": 3.1797256469726562, "learning_rate": 1.910271289714203e-05, "loss": 0.0403, "step": 1337 }, { "epoch": 0.14505637467476148, "grad_norm": 2.4967145919799805, "learning_rate": 1.9101987523574645e-05, "loss": 0.0296, "step": 1338 }, { "epoch": 0.14516478751084128, "grad_norm": 2.625363826751709, "learning_rate": 1.9101262150007257e-05, "loss": 0.0579, "step": 1339 }, { "epoch": 0.14527320034692107, "grad_norm": 2.940911054611206, "learning_rate": 1.910053677643987e-05, "loss": 0.0611, "step": 1340 }, { "epoch": 0.14538161318300086, "grad_norm": 3.173529624938965, "learning_rate": 1.909981140287248e-05, "loss": 0.0638, "step": 1341 }, { "epoch": 0.14549002601908065, "grad_norm": 12.963750839233398, "learning_rate": 1.9099086029305093e-05, "loss": 0.0906, "step": 1342 }, { "epoch": 0.14559843885516044, "grad_norm": 2.9435830116271973, "learning_rate": 1.9098360655737706e-05, "loss": 0.068, "step": 1343 }, { "epoch": 0.14570685169124023, "grad_norm": 1.8355516195297241, "learning_rate": 1.9097635282170318e-05, "loss": 0.0445, "step": 1344 }, { "epoch": 0.14581526452732002, "grad_norm": 2.778642416000366, "learning_rate": 1.9096909908602933e-05, "loss": 0.098, "step": 1345 }, { "epoch": 0.14592367736339984, "grad_norm": 1.322336196899414, "learning_rate": 1.9096184535035545e-05, "loss": 0.0425, "step": 1346 }, { "epoch": 0.14603209019947963, "grad_norm": 3.8535614013671875, "learning_rate": 1.9095459161468158e-05, "loss": 0.0262, "step": 1347 }, { "epoch": 0.14614050303555942, "grad_norm": 4.63035249710083, "learning_rate": 1.909473378790077e-05, "loss": 0.0687, "step": 1348 }, { "epoch": 0.1462489158716392, "grad_norm": 3.5339505672454834, "learning_rate": 1.9094008414333382e-05, "loss": 0.0551, "step": 1349 }, { "epoch": 0.146357328707719, "grad_norm": 2.34502911567688, "learning_rate": 1.9093283040765994e-05, "loss": 0.0538, "step": 1350 }, { "epoch": 0.1464657415437988, "grad_norm": 6.704124927520752, "learning_rate": 1.909255766719861e-05, "loss": 0.0593, "step": 1351 }, { "epoch": 0.14657415437987859, "grad_norm": 1.2608609199523926, "learning_rate": 1.9091832293631222e-05, "loss": 0.0191, "step": 1352 }, { "epoch": 0.14668256721595838, "grad_norm": 4.192587852478027, "learning_rate": 1.9091106920063834e-05, "loss": 0.0434, "step": 1353 }, { "epoch": 0.14679098005203817, "grad_norm": 3.872663974761963, "learning_rate": 1.909038154649645e-05, "loss": 0.119, "step": 1354 }, { "epoch": 0.14689939288811796, "grad_norm": 3.8649837970733643, "learning_rate": 1.908965617292906e-05, "loss": 0.0571, "step": 1355 }, { "epoch": 0.14700780572419775, "grad_norm": 3.0332186222076416, "learning_rate": 1.9088930799361674e-05, "loss": 0.0802, "step": 1356 }, { "epoch": 0.14711621856027754, "grad_norm": 4.954740524291992, "learning_rate": 1.9088205425794286e-05, "loss": 0.0785, "step": 1357 }, { "epoch": 0.14722463139635733, "grad_norm": 1.431950569152832, "learning_rate": 1.9087480052226898e-05, "loss": 0.014, "step": 1358 }, { "epoch": 0.14733304423243712, "grad_norm": 2.2343928813934326, "learning_rate": 1.908675467865951e-05, "loss": 0.0631, "step": 1359 }, { "epoch": 0.1474414570685169, "grad_norm": 1.8303436040878296, "learning_rate": 1.9086029305092126e-05, "loss": 0.0504, "step": 1360 }, { "epoch": 0.1475498699045967, "grad_norm": 2.579394817352295, "learning_rate": 1.9085303931524738e-05, "loss": 0.0799, "step": 1361 }, { "epoch": 0.1476582827406765, "grad_norm": 0.7952420115470886, "learning_rate": 1.908457855795735e-05, "loss": 0.009, "step": 1362 }, { "epoch": 0.14776669557675629, "grad_norm": 8.474079132080078, "learning_rate": 1.9083853184389962e-05, "loss": 0.0612, "step": 1363 }, { "epoch": 0.14787510841283608, "grad_norm": 1.2540382146835327, "learning_rate": 1.9083127810822574e-05, "loss": 0.0139, "step": 1364 }, { "epoch": 0.14798352124891587, "grad_norm": 2.4870564937591553, "learning_rate": 1.9082402437255186e-05, "loss": 0.0402, "step": 1365 }, { "epoch": 0.14809193408499566, "grad_norm": 4.926992416381836, "learning_rate": 1.90816770636878e-05, "loss": 0.0523, "step": 1366 }, { "epoch": 0.14820034692107545, "grad_norm": 1.7655718326568604, "learning_rate": 1.9080951690120414e-05, "loss": 0.0527, "step": 1367 }, { "epoch": 0.14830875975715524, "grad_norm": 2.1654579639434814, "learning_rate": 1.9080226316553026e-05, "loss": 0.0396, "step": 1368 }, { "epoch": 0.14841717259323503, "grad_norm": 6.198690891265869, "learning_rate": 1.9079500942985638e-05, "loss": 0.0712, "step": 1369 }, { "epoch": 0.14852558542931482, "grad_norm": 4.8393659591674805, "learning_rate": 1.9078775569418254e-05, "loss": 0.0536, "step": 1370 }, { "epoch": 0.1486339982653946, "grad_norm": 2.1558830738067627, "learning_rate": 1.9078050195850866e-05, "loss": 0.0772, "step": 1371 }, { "epoch": 0.1487424111014744, "grad_norm": 1.5405118465423584, "learning_rate": 1.9077324822283478e-05, "loss": 0.0657, "step": 1372 }, { "epoch": 0.1488508239375542, "grad_norm": 2.6464338302612305, "learning_rate": 1.907659944871609e-05, "loss": 0.0222, "step": 1373 }, { "epoch": 0.14895923677363399, "grad_norm": 3.7339909076690674, "learning_rate": 1.9075874075148702e-05, "loss": 0.05, "step": 1374 }, { "epoch": 0.1490676496097138, "grad_norm": 8.265535354614258, "learning_rate": 1.9075148701581318e-05, "loss": 0.0824, "step": 1375 }, { "epoch": 0.1491760624457936, "grad_norm": 5.7868757247924805, "learning_rate": 1.907442332801393e-05, "loss": 0.0607, "step": 1376 }, { "epoch": 0.14928447528187339, "grad_norm": 5.0863423347473145, "learning_rate": 1.9073697954446542e-05, "loss": 0.0454, "step": 1377 }, { "epoch": 0.14939288811795318, "grad_norm": 4.364648818969727, "learning_rate": 1.9072972580879154e-05, "loss": 0.0528, "step": 1378 }, { "epoch": 0.14950130095403297, "grad_norm": 7.509159088134766, "learning_rate": 1.9072247207311767e-05, "loss": 0.0763, "step": 1379 }, { "epoch": 0.14960971379011276, "grad_norm": 3.4646801948547363, "learning_rate": 1.907152183374438e-05, "loss": 0.0223, "step": 1380 }, { "epoch": 0.14971812662619255, "grad_norm": 4.009344577789307, "learning_rate": 1.907079646017699e-05, "loss": 0.0511, "step": 1381 }, { "epoch": 0.14982653946227234, "grad_norm": 1.0901401042938232, "learning_rate": 1.9070071086609606e-05, "loss": 0.0218, "step": 1382 }, { "epoch": 0.14993495229835213, "grad_norm": 2.5534934997558594, "learning_rate": 1.906934571304222e-05, "loss": 0.0434, "step": 1383 }, { "epoch": 0.15004336513443192, "grad_norm": 4.483579635620117, "learning_rate": 1.906862033947483e-05, "loss": 0.0933, "step": 1384 }, { "epoch": 0.1501517779705117, "grad_norm": 2.448882579803467, "learning_rate": 1.9067894965907443e-05, "loss": 0.0498, "step": 1385 }, { "epoch": 0.1502601908065915, "grad_norm": 1.4382247924804688, "learning_rate": 1.9067169592340055e-05, "loss": 0.025, "step": 1386 }, { "epoch": 0.1503686036426713, "grad_norm": 3.7975175380706787, "learning_rate": 1.9066444218772667e-05, "loss": 0.0678, "step": 1387 }, { "epoch": 0.15047701647875109, "grad_norm": 6.865841388702393, "learning_rate": 1.9065718845205283e-05, "loss": 0.0564, "step": 1388 }, { "epoch": 0.15058542931483088, "grad_norm": 3.269357919692993, "learning_rate": 1.9064993471637895e-05, "loss": 0.0932, "step": 1389 }, { "epoch": 0.15069384215091067, "grad_norm": 1.4064887762069702, "learning_rate": 1.906426809807051e-05, "loss": 0.0319, "step": 1390 }, { "epoch": 0.15080225498699046, "grad_norm": 1.8426434993743896, "learning_rate": 1.9063542724503122e-05, "loss": 0.0326, "step": 1391 }, { "epoch": 0.15091066782307025, "grad_norm": 8.604515075683594, "learning_rate": 1.9062817350935735e-05, "loss": 0.033, "step": 1392 }, { "epoch": 0.15101908065915004, "grad_norm": 4.1906962394714355, "learning_rate": 1.9062091977368347e-05, "loss": 0.0942, "step": 1393 }, { "epoch": 0.15112749349522983, "grad_norm": 1.7389769554138184, "learning_rate": 1.906136660380096e-05, "loss": 0.0395, "step": 1394 }, { "epoch": 0.15123590633130962, "grad_norm": 3.668720006942749, "learning_rate": 1.906064123023357e-05, "loss": 0.0383, "step": 1395 }, { "epoch": 0.1513443191673894, "grad_norm": 3.9928879737854004, "learning_rate": 1.9059915856666183e-05, "loss": 0.0352, "step": 1396 }, { "epoch": 0.1514527320034692, "grad_norm": 2.351640462875366, "learning_rate": 1.90591904830988e-05, "loss": 0.0316, "step": 1397 }, { "epoch": 0.151561144839549, "grad_norm": 3.7497057914733887, "learning_rate": 1.905846510953141e-05, "loss": 0.0458, "step": 1398 }, { "epoch": 0.15166955767562879, "grad_norm": 3.9514079093933105, "learning_rate": 1.9057739735964023e-05, "loss": 0.0586, "step": 1399 }, { "epoch": 0.15177797051170858, "grad_norm": 4.267385959625244, "learning_rate": 1.9057014362396635e-05, "loss": 0.092, "step": 1400 }, { "epoch": 0.15188638334778837, "grad_norm": 6.046668529510498, "learning_rate": 1.9056288988829247e-05, "loss": 0.0534, "step": 1401 }, { "epoch": 0.15199479618386816, "grad_norm": 5.229403495788574, "learning_rate": 1.905556361526186e-05, "loss": 0.0416, "step": 1402 }, { "epoch": 0.15210320901994795, "grad_norm": 1.9791476726531982, "learning_rate": 1.9054838241694475e-05, "loss": 0.0253, "step": 1403 }, { "epoch": 0.15221162185602774, "grad_norm": 1.3622289896011353, "learning_rate": 1.9054112868127087e-05, "loss": 0.0404, "step": 1404 }, { "epoch": 0.15232003469210756, "grad_norm": 4.706960678100586, "learning_rate": 1.90533874945597e-05, "loss": 0.0652, "step": 1405 }, { "epoch": 0.15242844752818735, "grad_norm": 5.189091682434082, "learning_rate": 1.9052662120992315e-05, "loss": 0.0506, "step": 1406 }, { "epoch": 0.15253686036426714, "grad_norm": 5.274691581726074, "learning_rate": 1.9051936747424927e-05, "loss": 0.0868, "step": 1407 }, { "epoch": 0.15264527320034693, "grad_norm": 2.072005033493042, "learning_rate": 1.905121137385754e-05, "loss": 0.031, "step": 1408 }, { "epoch": 0.15275368603642672, "grad_norm": 1.3224989175796509, "learning_rate": 1.905048600029015e-05, "loss": 0.0408, "step": 1409 }, { "epoch": 0.1528620988725065, "grad_norm": 1.6099603176116943, "learning_rate": 1.9049760626722763e-05, "loss": 0.0284, "step": 1410 }, { "epoch": 0.1529705117085863, "grad_norm": 3.7451839447021484, "learning_rate": 1.9049035253155375e-05, "loss": 0.0751, "step": 1411 }, { "epoch": 0.1530789245446661, "grad_norm": 3.5911951065063477, "learning_rate": 1.904830987958799e-05, "loss": 0.0639, "step": 1412 }, { "epoch": 0.15318733738074589, "grad_norm": 3.4534854888916016, "learning_rate": 1.9047584506020603e-05, "loss": 0.0424, "step": 1413 }, { "epoch": 0.15329575021682568, "grad_norm": 0.8451389670372009, "learning_rate": 1.9046859132453215e-05, "loss": 0.0274, "step": 1414 }, { "epoch": 0.15340416305290547, "grad_norm": 2.427379608154297, "learning_rate": 1.9046133758885827e-05, "loss": 0.0371, "step": 1415 }, { "epoch": 0.15351257588898526, "grad_norm": 3.4890849590301514, "learning_rate": 1.904540838531844e-05, "loss": 0.0152, "step": 1416 }, { "epoch": 0.15362098872506505, "grad_norm": 4.909672260284424, "learning_rate": 1.904468301175105e-05, "loss": 0.0611, "step": 1417 }, { "epoch": 0.15372940156114484, "grad_norm": 3.0705180168151855, "learning_rate": 1.9043957638183664e-05, "loss": 0.0441, "step": 1418 }, { "epoch": 0.15383781439722463, "grad_norm": 1.7144695520401, "learning_rate": 1.904323226461628e-05, "loss": 0.0492, "step": 1419 }, { "epoch": 0.15394622723330442, "grad_norm": 3.164097309112549, "learning_rate": 1.904250689104889e-05, "loss": 0.0789, "step": 1420 }, { "epoch": 0.1540546400693842, "grad_norm": 1.121982216835022, "learning_rate": 1.9041781517481504e-05, "loss": 0.0066, "step": 1421 }, { "epoch": 0.154163052905464, "grad_norm": 3.0081751346588135, "learning_rate": 1.904105614391412e-05, "loss": 0.0213, "step": 1422 }, { "epoch": 0.1542714657415438, "grad_norm": 13.539392471313477, "learning_rate": 1.904033077034673e-05, "loss": 0.0644, "step": 1423 }, { "epoch": 0.15437987857762359, "grad_norm": 2.822376251220703, "learning_rate": 1.9039605396779343e-05, "loss": 0.0453, "step": 1424 }, { "epoch": 0.15448829141370338, "grad_norm": 9.320452690124512, "learning_rate": 1.9038880023211956e-05, "loss": 0.0239, "step": 1425 }, { "epoch": 0.15459670424978317, "grad_norm": 2.29229736328125, "learning_rate": 1.9038154649644568e-05, "loss": 0.0463, "step": 1426 }, { "epoch": 0.15470511708586296, "grad_norm": 3.1911444664001465, "learning_rate": 1.9037429276077183e-05, "loss": 0.0581, "step": 1427 }, { "epoch": 0.15481352992194275, "grad_norm": 4.257416248321533, "learning_rate": 1.9036703902509795e-05, "loss": 0.0619, "step": 1428 }, { "epoch": 0.15492194275802254, "grad_norm": 3.5913503170013428, "learning_rate": 1.9035978528942408e-05, "loss": 0.0419, "step": 1429 }, { "epoch": 0.15503035559410233, "grad_norm": 6.37202262878418, "learning_rate": 1.903525315537502e-05, "loss": 0.0392, "step": 1430 }, { "epoch": 0.15513876843018212, "grad_norm": 4.784931659698486, "learning_rate": 1.9034527781807632e-05, "loss": 0.0339, "step": 1431 }, { "epoch": 0.1552471812662619, "grad_norm": 8.56386947631836, "learning_rate": 1.9033802408240244e-05, "loss": 0.0251, "step": 1432 }, { "epoch": 0.1553555941023417, "grad_norm": 2.9795119762420654, "learning_rate": 1.9033077034672856e-05, "loss": 0.0322, "step": 1433 }, { "epoch": 0.15546400693842152, "grad_norm": 2.63767147064209, "learning_rate": 1.903235166110547e-05, "loss": 0.049, "step": 1434 }, { "epoch": 0.1555724197745013, "grad_norm": 3.261509656906128, "learning_rate": 1.9031626287538084e-05, "loss": 0.0304, "step": 1435 }, { "epoch": 0.1556808326105811, "grad_norm": 4.237035274505615, "learning_rate": 1.9030900913970696e-05, "loss": 0.065, "step": 1436 }, { "epoch": 0.1557892454466609, "grad_norm": 4.505300521850586, "learning_rate": 1.9030175540403308e-05, "loss": 0.0349, "step": 1437 }, { "epoch": 0.1558976582827407, "grad_norm": 3.4144582748413086, "learning_rate": 1.902945016683592e-05, "loss": 0.0726, "step": 1438 }, { "epoch": 0.15600607111882048, "grad_norm": 6.949901103973389, "learning_rate": 1.9028724793268536e-05, "loss": 0.075, "step": 1439 }, { "epoch": 0.15611448395490027, "grad_norm": 5.896381855010986, "learning_rate": 1.9027999419701148e-05, "loss": 0.0419, "step": 1440 }, { "epoch": 0.15622289679098006, "grad_norm": 4.596411228179932, "learning_rate": 1.902727404613376e-05, "loss": 0.0299, "step": 1441 }, { "epoch": 0.15633130962705985, "grad_norm": 2.5799100399017334, "learning_rate": 1.9026548672566376e-05, "loss": 0.0771, "step": 1442 }, { "epoch": 0.15643972246313964, "grad_norm": 4.134519577026367, "learning_rate": 1.9025823298998988e-05, "loss": 0.0544, "step": 1443 }, { "epoch": 0.15654813529921943, "grad_norm": 13.16629409790039, "learning_rate": 1.90250979254316e-05, "loss": 0.0648, "step": 1444 }, { "epoch": 0.15665654813529922, "grad_norm": 2.9229018688201904, "learning_rate": 1.9024372551864212e-05, "loss": 0.0635, "step": 1445 }, { "epoch": 0.156764960971379, "grad_norm": 11.201581954956055, "learning_rate": 1.9023647178296824e-05, "loss": 0.0278, "step": 1446 }, { "epoch": 0.1568733738074588, "grad_norm": 2.8623557090759277, "learning_rate": 1.9022921804729436e-05, "loss": 0.0879, "step": 1447 }, { "epoch": 0.1569817866435386, "grad_norm": 3.3052797317504883, "learning_rate": 1.902219643116205e-05, "loss": 0.047, "step": 1448 }, { "epoch": 0.1570901994796184, "grad_norm": 1.8266209363937378, "learning_rate": 1.9021471057594664e-05, "loss": 0.0322, "step": 1449 }, { "epoch": 0.15719861231569818, "grad_norm": 3.3682548999786377, "learning_rate": 1.9020745684027276e-05, "loss": 0.0605, "step": 1450 }, { "epoch": 0.15730702515177797, "grad_norm": 1.7285605669021606, "learning_rate": 1.9020020310459888e-05, "loss": 0.036, "step": 1451 }, { "epoch": 0.15741543798785776, "grad_norm": 3.803764581680298, "learning_rate": 1.90192949368925e-05, "loss": 0.0556, "step": 1452 }, { "epoch": 0.15752385082393755, "grad_norm": 3.4797234535217285, "learning_rate": 1.9018569563325113e-05, "loss": 0.0423, "step": 1453 }, { "epoch": 0.15763226366001734, "grad_norm": 1.56516432762146, "learning_rate": 1.9017844189757725e-05, "loss": 0.0384, "step": 1454 }, { "epoch": 0.15774067649609713, "grad_norm": 2.2497055530548096, "learning_rate": 1.901711881619034e-05, "loss": 0.0439, "step": 1455 }, { "epoch": 0.15784908933217692, "grad_norm": 8.50703239440918, "learning_rate": 1.9016393442622952e-05, "loss": 0.05, "step": 1456 }, { "epoch": 0.1579575021682567, "grad_norm": 3.5913050174713135, "learning_rate": 1.9015668069055565e-05, "loss": 0.0397, "step": 1457 }, { "epoch": 0.1580659150043365, "grad_norm": 1.3636397123336792, "learning_rate": 1.901494269548818e-05, "loss": 0.0173, "step": 1458 }, { "epoch": 0.1581743278404163, "grad_norm": 2.6381967067718506, "learning_rate": 1.9014217321920792e-05, "loss": 0.0348, "step": 1459 }, { "epoch": 0.1582827406764961, "grad_norm": 1.763604998588562, "learning_rate": 1.9013491948353404e-05, "loss": 0.0379, "step": 1460 }, { "epoch": 0.15839115351257588, "grad_norm": 2.7311553955078125, "learning_rate": 1.9012766574786016e-05, "loss": 0.0759, "step": 1461 }, { "epoch": 0.15849956634865567, "grad_norm": 3.9640729427337646, "learning_rate": 1.901204120121863e-05, "loss": 0.0777, "step": 1462 }, { "epoch": 0.15860797918473546, "grad_norm": 1.7732070684432983, "learning_rate": 1.901131582765124e-05, "loss": 0.0319, "step": 1463 }, { "epoch": 0.15871639202081528, "grad_norm": 3.259857416152954, "learning_rate": 1.9010590454083856e-05, "loss": 0.0404, "step": 1464 }, { "epoch": 0.15882480485689507, "grad_norm": 3.2646498680114746, "learning_rate": 1.900986508051647e-05, "loss": 0.0249, "step": 1465 }, { "epoch": 0.15893321769297486, "grad_norm": 4.078449726104736, "learning_rate": 1.900913970694908e-05, "loss": 0.035, "step": 1466 }, { "epoch": 0.15904163052905465, "grad_norm": 1.1673407554626465, "learning_rate": 1.9008414333381693e-05, "loss": 0.0314, "step": 1467 }, { "epoch": 0.15915004336513444, "grad_norm": 4.430547714233398, "learning_rate": 1.9007688959814305e-05, "loss": 0.0544, "step": 1468 }, { "epoch": 0.15925845620121423, "grad_norm": 1.7951265573501587, "learning_rate": 1.9006963586246917e-05, "loss": 0.0412, "step": 1469 }, { "epoch": 0.15936686903729402, "grad_norm": 0.8044027090072632, "learning_rate": 1.900623821267953e-05, "loss": 0.0243, "step": 1470 }, { "epoch": 0.15947528187337381, "grad_norm": 1.2637789249420166, "learning_rate": 1.9005512839112145e-05, "loss": 0.0492, "step": 1471 }, { "epoch": 0.1595836947094536, "grad_norm": 6.071672439575195, "learning_rate": 1.9004787465544757e-05, "loss": 0.0851, "step": 1472 }, { "epoch": 0.1596921075455334, "grad_norm": 1.6540604829788208, "learning_rate": 1.900406209197737e-05, "loss": 0.0065, "step": 1473 }, { "epoch": 0.1598005203816132, "grad_norm": 1.5154149532318115, "learning_rate": 1.9003336718409985e-05, "loss": 0.0271, "step": 1474 }, { "epoch": 0.15990893321769298, "grad_norm": 7.661324501037598, "learning_rate": 1.9002611344842597e-05, "loss": 0.0536, "step": 1475 }, { "epoch": 0.16001734605377277, "grad_norm": 6.9063191413879395, "learning_rate": 1.900188597127521e-05, "loss": 0.0549, "step": 1476 }, { "epoch": 0.16012575888985256, "grad_norm": 9.424605369567871, "learning_rate": 1.900116059770782e-05, "loss": 0.0787, "step": 1477 }, { "epoch": 0.16023417172593235, "grad_norm": 4.37427282333374, "learning_rate": 1.9000435224140433e-05, "loss": 0.0509, "step": 1478 }, { "epoch": 0.16034258456201214, "grad_norm": 2.929408073425293, "learning_rate": 1.899970985057305e-05, "loss": 0.0473, "step": 1479 }, { "epoch": 0.16045099739809193, "grad_norm": 3.7617056369781494, "learning_rate": 1.899898447700566e-05, "loss": 0.0391, "step": 1480 }, { "epoch": 0.16055941023417172, "grad_norm": 3.8176121711730957, "learning_rate": 1.8998259103438273e-05, "loss": 0.0696, "step": 1481 }, { "epoch": 0.16066782307025151, "grad_norm": 6.677292823791504, "learning_rate": 1.8997533729870885e-05, "loss": 0.0633, "step": 1482 }, { "epoch": 0.1607762359063313, "grad_norm": 2.9845030307769775, "learning_rate": 1.8996808356303497e-05, "loss": 0.0465, "step": 1483 }, { "epoch": 0.1608846487424111, "grad_norm": 2.3970468044281006, "learning_rate": 1.899608298273611e-05, "loss": 0.0937, "step": 1484 }, { "epoch": 0.1609930615784909, "grad_norm": 4.429291725158691, "learning_rate": 1.899535760916872e-05, "loss": 0.0803, "step": 1485 }, { "epoch": 0.16110147441457068, "grad_norm": 6.014004230499268, "learning_rate": 1.8994632235601337e-05, "loss": 0.1175, "step": 1486 }, { "epoch": 0.16120988725065047, "grad_norm": 4.537809371948242, "learning_rate": 1.899390686203395e-05, "loss": 0.0939, "step": 1487 }, { "epoch": 0.16131830008673026, "grad_norm": 1.805601954460144, "learning_rate": 1.899318148846656e-05, "loss": 0.0422, "step": 1488 }, { "epoch": 0.16142671292281005, "grad_norm": 5.594374179840088, "learning_rate": 1.8992456114899173e-05, "loss": 0.0677, "step": 1489 }, { "epoch": 0.16153512575888984, "grad_norm": 1.9872925281524658, "learning_rate": 1.8991730741331786e-05, "loss": 0.0789, "step": 1490 }, { "epoch": 0.16164353859496963, "grad_norm": 9.475882530212402, "learning_rate": 1.89910053677644e-05, "loss": 0.0557, "step": 1491 }, { "epoch": 0.16175195143104942, "grad_norm": 9.296744346618652, "learning_rate": 1.8990279994197013e-05, "loss": 0.0398, "step": 1492 }, { "epoch": 0.16186036426712924, "grad_norm": 1.8843454122543335, "learning_rate": 1.8989554620629625e-05, "loss": 0.0289, "step": 1493 }, { "epoch": 0.16196877710320903, "grad_norm": 3.705742359161377, "learning_rate": 1.898882924706224e-05, "loss": 0.0767, "step": 1494 }, { "epoch": 0.16207718993928882, "grad_norm": 4.488726615905762, "learning_rate": 1.8988103873494853e-05, "loss": 0.0825, "step": 1495 }, { "epoch": 0.16218560277536861, "grad_norm": 3.5412325859069824, "learning_rate": 1.8987378499927465e-05, "loss": 0.0435, "step": 1496 }, { "epoch": 0.1622940156114484, "grad_norm": 2.4925620555877686, "learning_rate": 1.8986653126360077e-05, "loss": 0.0692, "step": 1497 }, { "epoch": 0.1624024284475282, "grad_norm": 2.120558500289917, "learning_rate": 1.898592775279269e-05, "loss": 0.0407, "step": 1498 }, { "epoch": 0.162510841283608, "grad_norm": 2.227409839630127, "learning_rate": 1.89852023792253e-05, "loss": 0.0553, "step": 1499 }, { "epoch": 0.16261925411968778, "grad_norm": 4.7017436027526855, "learning_rate": 1.8984477005657914e-05, "loss": 0.069, "step": 1500 }, { "epoch": 0.16272766695576757, "grad_norm": 2.329498291015625, "learning_rate": 1.898375163209053e-05, "loss": 0.0538, "step": 1501 }, { "epoch": 0.16283607979184736, "grad_norm": 1.2555930614471436, "learning_rate": 1.898302625852314e-05, "loss": 0.0449, "step": 1502 }, { "epoch": 0.16294449262792715, "grad_norm": 2.0895612239837646, "learning_rate": 1.8982300884955754e-05, "loss": 0.0383, "step": 1503 }, { "epoch": 0.16305290546400694, "grad_norm": 3.2115399837493896, "learning_rate": 1.8981575511388366e-05, "loss": 0.041, "step": 1504 }, { "epoch": 0.16316131830008673, "grad_norm": 6.8263092041015625, "learning_rate": 1.8980850137820978e-05, "loss": 0.0885, "step": 1505 }, { "epoch": 0.16326973113616652, "grad_norm": 2.756488561630249, "learning_rate": 1.898012476425359e-05, "loss": 0.025, "step": 1506 }, { "epoch": 0.16337814397224631, "grad_norm": 2.101292371749878, "learning_rate": 1.8979399390686206e-05, "loss": 0.0661, "step": 1507 }, { "epoch": 0.1634865568083261, "grad_norm": 2.732633113861084, "learning_rate": 1.8978674017118818e-05, "loss": 0.051, "step": 1508 }, { "epoch": 0.1635949696444059, "grad_norm": 3.5260939598083496, "learning_rate": 1.897794864355143e-05, "loss": 0.0791, "step": 1509 }, { "epoch": 0.1637033824804857, "grad_norm": 5.4847307205200195, "learning_rate": 1.8977223269984045e-05, "loss": 0.0365, "step": 1510 }, { "epoch": 0.16381179531656548, "grad_norm": 4.862500190734863, "learning_rate": 1.8976497896416658e-05, "loss": 0.0458, "step": 1511 }, { "epoch": 0.16392020815264527, "grad_norm": 3.390568733215332, "learning_rate": 1.897577252284927e-05, "loss": 0.0285, "step": 1512 }, { "epoch": 0.16402862098872506, "grad_norm": 7.670914173126221, "learning_rate": 1.8975047149281882e-05, "loss": 0.0624, "step": 1513 }, { "epoch": 0.16413703382480485, "grad_norm": 2.2439026832580566, "learning_rate": 1.8974321775714494e-05, "loss": 0.0322, "step": 1514 }, { "epoch": 0.16424544666088464, "grad_norm": 3.394962787628174, "learning_rate": 1.8973596402147106e-05, "loss": 0.0593, "step": 1515 }, { "epoch": 0.16435385949696443, "grad_norm": 4.1995625495910645, "learning_rate": 1.897287102857972e-05, "loss": 0.0584, "step": 1516 }, { "epoch": 0.16446227233304422, "grad_norm": 4.487072467803955, "learning_rate": 1.8972145655012334e-05, "loss": 0.0641, "step": 1517 }, { "epoch": 0.16457068516912401, "grad_norm": 1.0569872856140137, "learning_rate": 1.8971420281444946e-05, "loss": 0.0115, "step": 1518 }, { "epoch": 0.1646790980052038, "grad_norm": 0.5218547582626343, "learning_rate": 1.8970694907877558e-05, "loss": 0.0064, "step": 1519 }, { "epoch": 0.1647875108412836, "grad_norm": 1.5296101570129395, "learning_rate": 1.896996953431017e-05, "loss": 0.0319, "step": 1520 }, { "epoch": 0.1648959236773634, "grad_norm": 2.9178764820098877, "learning_rate": 1.8969244160742782e-05, "loss": 0.04, "step": 1521 }, { "epoch": 0.16500433651344318, "grad_norm": 2.5294408798217773, "learning_rate": 1.8968518787175394e-05, "loss": 0.0492, "step": 1522 }, { "epoch": 0.165112749349523, "grad_norm": 2.631291151046753, "learning_rate": 1.896779341360801e-05, "loss": 0.0598, "step": 1523 }, { "epoch": 0.1652211621856028, "grad_norm": 3.4038445949554443, "learning_rate": 1.8967068040040622e-05, "loss": 0.0682, "step": 1524 }, { "epoch": 0.16532957502168258, "grad_norm": 2.0142180919647217, "learning_rate": 1.8966342666473234e-05, "loss": 0.023, "step": 1525 }, { "epoch": 0.16543798785776237, "grad_norm": 5.278994083404541, "learning_rate": 1.896561729290585e-05, "loss": 0.0195, "step": 1526 }, { "epoch": 0.16554640069384216, "grad_norm": 5.81040096282959, "learning_rate": 1.8964891919338462e-05, "loss": 0.0706, "step": 1527 }, { "epoch": 0.16565481352992195, "grad_norm": 3.895327568054199, "learning_rate": 1.8964166545771074e-05, "loss": 0.0625, "step": 1528 }, { "epoch": 0.16576322636600174, "grad_norm": 1.8968441486358643, "learning_rate": 1.8963441172203686e-05, "loss": 0.0112, "step": 1529 }, { "epoch": 0.16587163920208153, "grad_norm": 4.757837295532227, "learning_rate": 1.89627157986363e-05, "loss": 0.0249, "step": 1530 }, { "epoch": 0.16598005203816132, "grad_norm": 1.5856717824935913, "learning_rate": 1.8961990425068914e-05, "loss": 0.0311, "step": 1531 }, { "epoch": 0.16608846487424112, "grad_norm": 2.681685447692871, "learning_rate": 1.8961265051501526e-05, "loss": 0.0315, "step": 1532 }, { "epoch": 0.1661968777103209, "grad_norm": 7.963295936584473, "learning_rate": 1.8960539677934138e-05, "loss": 0.1081, "step": 1533 }, { "epoch": 0.1663052905464007, "grad_norm": 3.5902507305145264, "learning_rate": 1.895981430436675e-05, "loss": 0.0296, "step": 1534 }, { "epoch": 0.1664137033824805, "grad_norm": 2.8729965686798096, "learning_rate": 1.8959088930799363e-05, "loss": 0.0677, "step": 1535 }, { "epoch": 0.16652211621856028, "grad_norm": 5.115811347961426, "learning_rate": 1.8958363557231975e-05, "loss": 0.0504, "step": 1536 }, { "epoch": 0.16663052905464007, "grad_norm": 4.320963382720947, "learning_rate": 1.8957638183664587e-05, "loss": 0.079, "step": 1537 }, { "epoch": 0.16673894189071986, "grad_norm": 1.2365803718566895, "learning_rate": 1.8956912810097202e-05, "loss": 0.0117, "step": 1538 }, { "epoch": 0.16684735472679965, "grad_norm": 2.0738964080810547, "learning_rate": 1.8956187436529814e-05, "loss": 0.0332, "step": 1539 }, { "epoch": 0.16695576756287944, "grad_norm": 3.67879056930542, "learning_rate": 1.8955462062962427e-05, "loss": 0.0445, "step": 1540 }, { "epoch": 0.16706418039895923, "grad_norm": 4.318144798278809, "learning_rate": 1.895473668939504e-05, "loss": 0.0822, "step": 1541 }, { "epoch": 0.16717259323503902, "grad_norm": 5.453789234161377, "learning_rate": 1.895401131582765e-05, "loss": 0.0514, "step": 1542 }, { "epoch": 0.16728100607111882, "grad_norm": 2.1080965995788574, "learning_rate": 1.8953285942260266e-05, "loss": 0.0576, "step": 1543 }, { "epoch": 0.1673894189071986, "grad_norm": 1.4324147701263428, "learning_rate": 1.895256056869288e-05, "loss": 0.0187, "step": 1544 }, { "epoch": 0.1674978317432784, "grad_norm": 4.453859329223633, "learning_rate": 1.895183519512549e-05, "loss": 0.0315, "step": 1545 }, { "epoch": 0.1676062445793582, "grad_norm": 2.826490879058838, "learning_rate": 1.8951109821558106e-05, "loss": 0.0492, "step": 1546 }, { "epoch": 0.16771465741543798, "grad_norm": 2.4576807022094727, "learning_rate": 1.895038444799072e-05, "loss": 0.0792, "step": 1547 }, { "epoch": 0.16782307025151777, "grad_norm": 5.968149185180664, "learning_rate": 1.894965907442333e-05, "loss": 0.0503, "step": 1548 }, { "epoch": 0.16793148308759756, "grad_norm": 4.821911811828613, "learning_rate": 1.8948933700855943e-05, "loss": 0.0174, "step": 1549 }, { "epoch": 0.16803989592367735, "grad_norm": 3.3098955154418945, "learning_rate": 1.8948208327288555e-05, "loss": 0.0198, "step": 1550 }, { "epoch": 0.16814830875975714, "grad_norm": 7.464756011962891, "learning_rate": 1.8947482953721167e-05, "loss": 0.0273, "step": 1551 }, { "epoch": 0.16825672159583693, "grad_norm": 0.8502285480499268, "learning_rate": 1.894675758015378e-05, "loss": 0.0218, "step": 1552 }, { "epoch": 0.16836513443191675, "grad_norm": 2.790886878967285, "learning_rate": 1.8946032206586395e-05, "loss": 0.0452, "step": 1553 }, { "epoch": 0.16847354726799654, "grad_norm": 5.181031227111816, "learning_rate": 1.8945306833019007e-05, "loss": 0.0576, "step": 1554 }, { "epoch": 0.16858196010407633, "grad_norm": 4.863188743591309, "learning_rate": 1.894458145945162e-05, "loss": 0.0619, "step": 1555 }, { "epoch": 0.16869037294015612, "grad_norm": 5.030806064605713, "learning_rate": 1.894385608588423e-05, "loss": 0.0847, "step": 1556 }, { "epoch": 0.16879878577623592, "grad_norm": 2.8656654357910156, "learning_rate": 1.8943130712316843e-05, "loss": 0.0364, "step": 1557 }, { "epoch": 0.1689071986123157, "grad_norm": 1.3484383821487427, "learning_rate": 1.8942405338749455e-05, "loss": 0.0334, "step": 1558 }, { "epoch": 0.1690156114483955, "grad_norm": 9.93315315246582, "learning_rate": 1.894167996518207e-05, "loss": 0.1107, "step": 1559 }, { "epoch": 0.1691240242844753, "grad_norm": 2.9157307147979736, "learning_rate": 1.8940954591614683e-05, "loss": 0.0479, "step": 1560 }, { "epoch": 0.16923243712055508, "grad_norm": 3.7372398376464844, "learning_rate": 1.8940229218047295e-05, "loss": 0.0754, "step": 1561 }, { "epoch": 0.16934084995663487, "grad_norm": 1.31802499294281, "learning_rate": 1.893950384447991e-05, "loss": 0.0116, "step": 1562 }, { "epoch": 0.16944926279271466, "grad_norm": 3.729741096496582, "learning_rate": 1.8938778470912523e-05, "loss": 0.055, "step": 1563 }, { "epoch": 0.16955767562879445, "grad_norm": 4.936605453491211, "learning_rate": 1.8938053097345135e-05, "loss": 0.0741, "step": 1564 }, { "epoch": 0.16966608846487424, "grad_norm": 1.8494702577590942, "learning_rate": 1.8937327723777747e-05, "loss": 0.024, "step": 1565 }, { "epoch": 0.16977450130095403, "grad_norm": 1.961987018585205, "learning_rate": 1.893660235021036e-05, "loss": 0.0546, "step": 1566 }, { "epoch": 0.16988291413703382, "grad_norm": 3.431089162826538, "learning_rate": 1.893587697664297e-05, "loss": 0.0405, "step": 1567 }, { "epoch": 0.16999132697311362, "grad_norm": 2.0468738079071045, "learning_rate": 1.8935151603075587e-05, "loss": 0.0635, "step": 1568 }, { "epoch": 0.1700997398091934, "grad_norm": 2.47259521484375, "learning_rate": 1.89344262295082e-05, "loss": 0.0452, "step": 1569 }, { "epoch": 0.1702081526452732, "grad_norm": 3.909184455871582, "learning_rate": 1.893370085594081e-05, "loss": 0.0486, "step": 1570 }, { "epoch": 0.170316565481353, "grad_norm": 2.9811952114105225, "learning_rate": 1.8932975482373423e-05, "loss": 0.0274, "step": 1571 }, { "epoch": 0.17042497831743278, "grad_norm": 2.116734504699707, "learning_rate": 1.8932250108806036e-05, "loss": 0.0378, "step": 1572 }, { "epoch": 0.17053339115351257, "grad_norm": 4.104889392852783, "learning_rate": 1.8931524735238648e-05, "loss": 0.0725, "step": 1573 }, { "epoch": 0.17064180398959236, "grad_norm": 4.514908313751221, "learning_rate": 1.8930799361671263e-05, "loss": 0.0625, "step": 1574 }, { "epoch": 0.17075021682567215, "grad_norm": 2.9042246341705322, "learning_rate": 1.8930073988103875e-05, "loss": 0.0481, "step": 1575 }, { "epoch": 0.17085862966175194, "grad_norm": 2.0155646800994873, "learning_rate": 1.8929348614536488e-05, "loss": 0.0221, "step": 1576 }, { "epoch": 0.17096704249783173, "grad_norm": 6.1094231605529785, "learning_rate": 1.89286232409691e-05, "loss": 0.1582, "step": 1577 }, { "epoch": 0.17107545533391152, "grad_norm": 1.4451045989990234, "learning_rate": 1.8927897867401715e-05, "loss": 0.0292, "step": 1578 }, { "epoch": 0.17118386816999132, "grad_norm": 2.622486114501953, "learning_rate": 1.8927172493834327e-05, "loss": 0.0419, "step": 1579 }, { "epoch": 0.1712922810060711, "grad_norm": 0.9187362194061279, "learning_rate": 1.892644712026694e-05, "loss": 0.0157, "step": 1580 }, { "epoch": 0.1714006938421509, "grad_norm": 3.7956149578094482, "learning_rate": 1.892572174669955e-05, "loss": 0.0313, "step": 1581 }, { "epoch": 0.17150910667823072, "grad_norm": 2.745793104171753, "learning_rate": 1.8924996373132164e-05, "loss": 0.0325, "step": 1582 }, { "epoch": 0.1716175195143105, "grad_norm": 1.8621249198913574, "learning_rate": 1.892427099956478e-05, "loss": 0.0289, "step": 1583 }, { "epoch": 0.1717259323503903, "grad_norm": 2.897221565246582, "learning_rate": 1.892354562599739e-05, "loss": 0.0748, "step": 1584 }, { "epoch": 0.1718343451864701, "grad_norm": 1.70366370677948, "learning_rate": 1.8922820252430004e-05, "loss": 0.0328, "step": 1585 }, { "epoch": 0.17194275802254988, "grad_norm": 3.506347894668579, "learning_rate": 1.8922094878862616e-05, "loss": 0.0217, "step": 1586 }, { "epoch": 0.17205117085862967, "grad_norm": 7.0737833976745605, "learning_rate": 1.8921369505295228e-05, "loss": 0.0383, "step": 1587 }, { "epoch": 0.17215958369470946, "grad_norm": 2.7156851291656494, "learning_rate": 1.892064413172784e-05, "loss": 0.0645, "step": 1588 }, { "epoch": 0.17226799653078925, "grad_norm": 0.8619655966758728, "learning_rate": 1.8919918758160452e-05, "loss": 0.0185, "step": 1589 }, { "epoch": 0.17237640936686904, "grad_norm": 3.5216896533966064, "learning_rate": 1.8919193384593068e-05, "loss": 0.056, "step": 1590 }, { "epoch": 0.17248482220294883, "grad_norm": 3.409641742706299, "learning_rate": 1.891846801102568e-05, "loss": 0.0426, "step": 1591 }, { "epoch": 0.17259323503902863, "grad_norm": 10.313385963439941, "learning_rate": 1.8917742637458292e-05, "loss": 0.0545, "step": 1592 }, { "epoch": 0.17270164787510842, "grad_norm": 3.8169374465942383, "learning_rate": 1.8917017263890904e-05, "loss": 0.0799, "step": 1593 }, { "epoch": 0.1728100607111882, "grad_norm": 2.92805552482605, "learning_rate": 1.8916291890323516e-05, "loss": 0.0501, "step": 1594 }, { "epoch": 0.172918473547268, "grad_norm": 5.617769718170166, "learning_rate": 1.8915566516756132e-05, "loss": 0.0482, "step": 1595 }, { "epoch": 0.1730268863833478, "grad_norm": 3.168116807937622, "learning_rate": 1.8914841143188744e-05, "loss": 0.0261, "step": 1596 }, { "epoch": 0.17313529921942758, "grad_norm": 5.254677772521973, "learning_rate": 1.8914115769621356e-05, "loss": 0.0847, "step": 1597 }, { "epoch": 0.17324371205550737, "grad_norm": 0.5149936676025391, "learning_rate": 1.891339039605397e-05, "loss": 0.0057, "step": 1598 }, { "epoch": 0.17335212489158716, "grad_norm": 1.907289743423462, "learning_rate": 1.8912665022486584e-05, "loss": 0.0245, "step": 1599 }, { "epoch": 0.17346053772766695, "grad_norm": 2.8973097801208496, "learning_rate": 1.8911939648919196e-05, "loss": 0.0469, "step": 1600 }, { "epoch": 0.17356895056374674, "grad_norm": 1.6570935249328613, "learning_rate": 1.8911214275351808e-05, "loss": 0.0323, "step": 1601 }, { "epoch": 0.17367736339982653, "grad_norm": 1.1381151676177979, "learning_rate": 1.891048890178442e-05, "loss": 0.0115, "step": 1602 }, { "epoch": 0.17378577623590633, "grad_norm": 2.3629262447357178, "learning_rate": 1.8909763528217032e-05, "loss": 0.0461, "step": 1603 }, { "epoch": 0.17389418907198612, "grad_norm": 2.939028263092041, "learning_rate": 1.8909038154649644e-05, "loss": 0.0198, "step": 1604 }, { "epoch": 0.1740026019080659, "grad_norm": 4.455678939819336, "learning_rate": 1.890831278108226e-05, "loss": 0.0396, "step": 1605 }, { "epoch": 0.1741110147441457, "grad_norm": 12.555522918701172, "learning_rate": 1.8907587407514872e-05, "loss": 0.0783, "step": 1606 }, { "epoch": 0.1742194275802255, "grad_norm": 5.511351108551025, "learning_rate": 1.8906862033947484e-05, "loss": 0.0305, "step": 1607 }, { "epoch": 0.17432784041630528, "grad_norm": 4.1229071617126465, "learning_rate": 1.8906136660380096e-05, "loss": 0.0336, "step": 1608 }, { "epoch": 0.17443625325238507, "grad_norm": 3.8895089626312256, "learning_rate": 1.890541128681271e-05, "loss": 0.0517, "step": 1609 }, { "epoch": 0.17454466608846486, "grad_norm": 8.188236236572266, "learning_rate": 1.890468591324532e-05, "loss": 0.0892, "step": 1610 }, { "epoch": 0.17465307892454465, "grad_norm": 4.510853290557861, "learning_rate": 1.8903960539677936e-05, "loss": 0.0714, "step": 1611 }, { "epoch": 0.17476149176062447, "grad_norm": 1.281721591949463, "learning_rate": 1.890323516611055e-05, "loss": 0.005, "step": 1612 }, { "epoch": 0.17486990459670426, "grad_norm": 1.2551952600479126, "learning_rate": 1.890250979254316e-05, "loss": 0.0162, "step": 1613 }, { "epoch": 0.17497831743278405, "grad_norm": 0.7441969513893127, "learning_rate": 1.8901784418975776e-05, "loss": 0.0105, "step": 1614 }, { "epoch": 0.17508673026886384, "grad_norm": 6.399380207061768, "learning_rate": 1.8901059045408388e-05, "loss": 0.0803, "step": 1615 }, { "epoch": 0.17519514310494363, "grad_norm": 1.7269190549850464, "learning_rate": 1.8900333671841e-05, "loss": 0.0322, "step": 1616 }, { "epoch": 0.17530355594102343, "grad_norm": 1.6448591947555542, "learning_rate": 1.8899608298273612e-05, "loss": 0.017, "step": 1617 }, { "epoch": 0.17541196877710322, "grad_norm": 6.6541314125061035, "learning_rate": 1.8898882924706225e-05, "loss": 0.0482, "step": 1618 }, { "epoch": 0.175520381613183, "grad_norm": 2.704582452774048, "learning_rate": 1.8898157551138837e-05, "loss": 0.0763, "step": 1619 }, { "epoch": 0.1756287944492628, "grad_norm": 5.296087265014648, "learning_rate": 1.8897432177571452e-05, "loss": 0.0993, "step": 1620 }, { "epoch": 0.1757372072853426, "grad_norm": 3.1600918769836426, "learning_rate": 1.8896706804004064e-05, "loss": 0.0424, "step": 1621 }, { "epoch": 0.17584562012142238, "grad_norm": 3.997498035430908, "learning_rate": 1.8895981430436677e-05, "loss": 0.0551, "step": 1622 }, { "epoch": 0.17595403295750217, "grad_norm": 3.3871896266937256, "learning_rate": 1.889525605686929e-05, "loss": 0.0565, "step": 1623 }, { "epoch": 0.17606244579358196, "grad_norm": 5.400032043457031, "learning_rate": 1.88945306833019e-05, "loss": 0.0724, "step": 1624 }, { "epoch": 0.17617085862966175, "grad_norm": 1.611013650894165, "learning_rate": 1.8893805309734513e-05, "loss": 0.0215, "step": 1625 }, { "epoch": 0.17627927146574154, "grad_norm": 1.4600011110305786, "learning_rate": 1.889307993616713e-05, "loss": 0.0253, "step": 1626 }, { "epoch": 0.17638768430182133, "grad_norm": 6.369128227233887, "learning_rate": 1.889235456259974e-05, "loss": 0.1102, "step": 1627 }, { "epoch": 0.17649609713790113, "grad_norm": 5.361586570739746, "learning_rate": 1.8891629189032353e-05, "loss": 0.0461, "step": 1628 }, { "epoch": 0.17660450997398092, "grad_norm": 7.446223258972168, "learning_rate": 1.889090381546497e-05, "loss": 0.0615, "step": 1629 }, { "epoch": 0.1767129228100607, "grad_norm": 3.0896265506744385, "learning_rate": 1.889017844189758e-05, "loss": 0.0266, "step": 1630 }, { "epoch": 0.1768213356461405, "grad_norm": 7.039620876312256, "learning_rate": 1.8889453068330193e-05, "loss": 0.0699, "step": 1631 }, { "epoch": 0.1769297484822203, "grad_norm": 0.9471395015716553, "learning_rate": 1.8888727694762805e-05, "loss": 0.011, "step": 1632 }, { "epoch": 0.17703816131830008, "grad_norm": 1.4048537015914917, "learning_rate": 1.8888002321195417e-05, "loss": 0.0189, "step": 1633 }, { "epoch": 0.17714657415437987, "grad_norm": 5.450591564178467, "learning_rate": 1.888727694762803e-05, "loss": 0.0289, "step": 1634 }, { "epoch": 0.17725498699045966, "grad_norm": 2.1313657760620117, "learning_rate": 1.8886551574060645e-05, "loss": 0.0396, "step": 1635 }, { "epoch": 0.17736339982653945, "grad_norm": 3.09202241897583, "learning_rate": 1.8885826200493257e-05, "loss": 0.0379, "step": 1636 }, { "epoch": 0.17747181266261924, "grad_norm": 1.675308108329773, "learning_rate": 1.888510082692587e-05, "loss": 0.0243, "step": 1637 }, { "epoch": 0.17758022549869903, "grad_norm": 0.6644577383995056, "learning_rate": 1.888437545335848e-05, "loss": 0.0232, "step": 1638 }, { "epoch": 0.17768863833477883, "grad_norm": 2.7463035583496094, "learning_rate": 1.8883650079791093e-05, "loss": 0.0275, "step": 1639 }, { "epoch": 0.17779705117085862, "grad_norm": 6.322324752807617, "learning_rate": 1.8882924706223705e-05, "loss": 0.0537, "step": 1640 }, { "epoch": 0.17790546400693844, "grad_norm": 4.021119594573975, "learning_rate": 1.8882199332656317e-05, "loss": 0.0609, "step": 1641 }, { "epoch": 0.17801387684301823, "grad_norm": 2.887024164199829, "learning_rate": 1.8881473959088933e-05, "loss": 0.0284, "step": 1642 }, { "epoch": 0.17812228967909802, "grad_norm": 0.6210408806800842, "learning_rate": 1.8880748585521545e-05, "loss": 0.0041, "step": 1643 }, { "epoch": 0.1782307025151778, "grad_norm": 1.6946988105773926, "learning_rate": 1.8880023211954157e-05, "loss": 0.0139, "step": 1644 }, { "epoch": 0.1783391153512576, "grad_norm": 3.3758907318115234, "learning_rate": 1.887929783838677e-05, "loss": 0.0286, "step": 1645 }, { "epoch": 0.1784475281873374, "grad_norm": 2.5002388954162598, "learning_rate": 1.887857246481938e-05, "loss": 0.0105, "step": 1646 }, { "epoch": 0.17855594102341718, "grad_norm": 8.138860702514648, "learning_rate": 1.8877847091251997e-05, "loss": 0.0779, "step": 1647 }, { "epoch": 0.17866435385949697, "grad_norm": 5.734771251678467, "learning_rate": 1.887712171768461e-05, "loss": 0.1201, "step": 1648 }, { "epoch": 0.17877276669557676, "grad_norm": 3.791304111480713, "learning_rate": 1.887639634411722e-05, "loss": 0.0355, "step": 1649 }, { "epoch": 0.17888117953165655, "grad_norm": 5.317184925079346, "learning_rate": 1.8875670970549837e-05, "loss": 0.0663, "step": 1650 }, { "epoch": 0.17898959236773634, "grad_norm": 4.652544021606445, "learning_rate": 1.887494559698245e-05, "loss": 0.0363, "step": 1651 }, { "epoch": 0.17909800520381614, "grad_norm": 2.5250160694122314, "learning_rate": 1.887422022341506e-05, "loss": 0.0113, "step": 1652 }, { "epoch": 0.17920641803989593, "grad_norm": 3.337904453277588, "learning_rate": 1.8873494849847673e-05, "loss": 0.0238, "step": 1653 }, { "epoch": 0.17931483087597572, "grad_norm": 8.656427383422852, "learning_rate": 1.8872769476280286e-05, "loss": 0.0346, "step": 1654 }, { "epoch": 0.1794232437120555, "grad_norm": 5.356227874755859, "learning_rate": 1.8872044102712898e-05, "loss": 0.0354, "step": 1655 }, { "epoch": 0.1795316565481353, "grad_norm": 5.172378063201904, "learning_rate": 1.887131872914551e-05, "loss": 0.0316, "step": 1656 }, { "epoch": 0.1796400693842151, "grad_norm": 1.5454739332199097, "learning_rate": 1.8870593355578125e-05, "loss": 0.0144, "step": 1657 }, { "epoch": 0.17974848222029488, "grad_norm": 5.321766376495361, "learning_rate": 1.8869867982010737e-05, "loss": 0.0653, "step": 1658 }, { "epoch": 0.17985689505637467, "grad_norm": 6.512908458709717, "learning_rate": 1.886914260844335e-05, "loss": 0.142, "step": 1659 }, { "epoch": 0.17996530789245446, "grad_norm": 10.018754959106445, "learning_rate": 1.8868417234875962e-05, "loss": 0.0738, "step": 1660 }, { "epoch": 0.18007372072853425, "grad_norm": 0.8955700993537903, "learning_rate": 1.8867691861308574e-05, "loss": 0.014, "step": 1661 }, { "epoch": 0.18018213356461404, "grad_norm": 4.828739166259766, "learning_rate": 1.8866966487741186e-05, "loss": 0.0583, "step": 1662 }, { "epoch": 0.18029054640069384, "grad_norm": 1.473697304725647, "learning_rate": 1.88662411141738e-05, "loss": 0.0493, "step": 1663 }, { "epoch": 0.18039895923677363, "grad_norm": 3.1593263149261475, "learning_rate": 1.8865515740606414e-05, "loss": 0.0528, "step": 1664 }, { "epoch": 0.18050737207285342, "grad_norm": 3.7692530155181885, "learning_rate": 1.8864790367039026e-05, "loss": 0.0392, "step": 1665 }, { "epoch": 0.1806157849089332, "grad_norm": 3.972344160079956, "learning_rate": 1.886406499347164e-05, "loss": 0.0672, "step": 1666 }, { "epoch": 0.180724197745013, "grad_norm": 1.0967549085617065, "learning_rate": 1.8863339619904254e-05, "loss": 0.0106, "step": 1667 }, { "epoch": 0.1808326105810928, "grad_norm": 3.7517549991607666, "learning_rate": 1.8862614246336866e-05, "loss": 0.0241, "step": 1668 }, { "epoch": 0.18094102341717258, "grad_norm": 4.305379390716553, "learning_rate": 1.8861888872769478e-05, "loss": 0.0883, "step": 1669 }, { "epoch": 0.18104943625325237, "grad_norm": 4.868812084197998, "learning_rate": 1.886116349920209e-05, "loss": 0.0654, "step": 1670 }, { "epoch": 0.1811578490893322, "grad_norm": 3.31890606880188, "learning_rate": 1.8860438125634702e-05, "loss": 0.0367, "step": 1671 }, { "epoch": 0.18126626192541198, "grad_norm": 2.6490845680236816, "learning_rate": 1.8859712752067318e-05, "loss": 0.0298, "step": 1672 }, { "epoch": 0.18137467476149177, "grad_norm": 1.9236161708831787, "learning_rate": 1.885898737849993e-05, "loss": 0.0429, "step": 1673 }, { "epoch": 0.18148308759757156, "grad_norm": 4.125512599945068, "learning_rate": 1.8858262004932542e-05, "loss": 0.036, "step": 1674 }, { "epoch": 0.18159150043365135, "grad_norm": 7.552781581878662, "learning_rate": 1.8857536631365154e-05, "loss": 0.0642, "step": 1675 }, { "epoch": 0.18169991326973114, "grad_norm": 4.521940231323242, "learning_rate": 1.8856811257797766e-05, "loss": 0.0355, "step": 1676 }, { "epoch": 0.18180832610581094, "grad_norm": 4.496877670288086, "learning_rate": 1.885608588423038e-05, "loss": 0.0587, "step": 1677 }, { "epoch": 0.18191673894189073, "grad_norm": 5.7501726150512695, "learning_rate": 1.8855360510662994e-05, "loss": 0.0533, "step": 1678 }, { "epoch": 0.18202515177797052, "grad_norm": 1.2683775424957275, "learning_rate": 1.8854635137095606e-05, "loss": 0.0083, "step": 1679 }, { "epoch": 0.1821335646140503, "grad_norm": 0.6529747843742371, "learning_rate": 1.8853909763528218e-05, "loss": 0.0137, "step": 1680 }, { "epoch": 0.1822419774501301, "grad_norm": 2.8199589252471924, "learning_rate": 1.8853184389960834e-05, "loss": 0.0374, "step": 1681 }, { "epoch": 0.1823503902862099, "grad_norm": 1.2602980136871338, "learning_rate": 1.8852459016393446e-05, "loss": 0.0339, "step": 1682 }, { "epoch": 0.18245880312228968, "grad_norm": 3.038363218307495, "learning_rate": 1.8851733642826058e-05, "loss": 0.0498, "step": 1683 }, { "epoch": 0.18256721595836947, "grad_norm": 2.5433733463287354, "learning_rate": 1.885100826925867e-05, "loss": 0.0681, "step": 1684 }, { "epoch": 0.18267562879444926, "grad_norm": 1.728600263595581, "learning_rate": 1.8850282895691282e-05, "loss": 0.0436, "step": 1685 }, { "epoch": 0.18278404163052905, "grad_norm": 3.624324083328247, "learning_rate": 1.8849557522123894e-05, "loss": 0.0483, "step": 1686 }, { "epoch": 0.18289245446660884, "grad_norm": 2.5839145183563232, "learning_rate": 1.884883214855651e-05, "loss": 0.0459, "step": 1687 }, { "epoch": 0.18300086730268864, "grad_norm": 3.4046337604522705, "learning_rate": 1.8848106774989122e-05, "loss": 0.0274, "step": 1688 }, { "epoch": 0.18310928013876843, "grad_norm": 3.441505193710327, "learning_rate": 1.8847381401421734e-05, "loss": 0.0316, "step": 1689 }, { "epoch": 0.18321769297484822, "grad_norm": 2.488413095474243, "learning_rate": 1.8846656027854346e-05, "loss": 0.0264, "step": 1690 }, { "epoch": 0.183326105810928, "grad_norm": 2.9953160285949707, "learning_rate": 1.884593065428696e-05, "loss": 0.0454, "step": 1691 }, { "epoch": 0.1834345186470078, "grad_norm": 1.0529086589813232, "learning_rate": 1.884520528071957e-05, "loss": 0.0167, "step": 1692 }, { "epoch": 0.1835429314830876, "grad_norm": 3.657735586166382, "learning_rate": 1.8844479907152183e-05, "loss": 0.0462, "step": 1693 }, { "epoch": 0.18365134431916738, "grad_norm": 1.456354022026062, "learning_rate": 1.88437545335848e-05, "loss": 0.0208, "step": 1694 }, { "epoch": 0.18375975715524717, "grad_norm": 2.1619229316711426, "learning_rate": 1.884302916001741e-05, "loss": 0.0375, "step": 1695 }, { "epoch": 0.18386816999132696, "grad_norm": 1.6743431091308594, "learning_rate": 1.8842303786450023e-05, "loss": 0.0123, "step": 1696 }, { "epoch": 0.18397658282740675, "grad_norm": 1.953519344329834, "learning_rate": 1.8841578412882635e-05, "loss": 0.0252, "step": 1697 }, { "epoch": 0.18408499566348654, "grad_norm": 0.9062800407409668, "learning_rate": 1.8840853039315247e-05, "loss": 0.0127, "step": 1698 }, { "epoch": 0.18419340849956634, "grad_norm": 2.8923449516296387, "learning_rate": 1.8840127665747862e-05, "loss": 0.0162, "step": 1699 }, { "epoch": 0.18430182133564615, "grad_norm": 3.49638032913208, "learning_rate": 1.8839402292180475e-05, "loss": 0.0435, "step": 1700 }, { "epoch": 0.18441023417172595, "grad_norm": 3.973360538482666, "learning_rate": 1.8838676918613087e-05, "loss": 0.0766, "step": 1701 }, { "epoch": 0.18451864700780574, "grad_norm": 0.3396821916103363, "learning_rate": 1.8837951545045702e-05, "loss": 0.0019, "step": 1702 }, { "epoch": 0.18462705984388553, "grad_norm": 0.9565585255622864, "learning_rate": 1.8837226171478314e-05, "loss": 0.0038, "step": 1703 }, { "epoch": 0.18473547267996532, "grad_norm": 5.762755393981934, "learning_rate": 1.8836500797910927e-05, "loss": 0.0471, "step": 1704 }, { "epoch": 0.1848438855160451, "grad_norm": 4.6389851570129395, "learning_rate": 1.883577542434354e-05, "loss": 0.0878, "step": 1705 }, { "epoch": 0.1849522983521249, "grad_norm": 5.039500713348389, "learning_rate": 1.883505005077615e-05, "loss": 0.1082, "step": 1706 }, { "epoch": 0.1850607111882047, "grad_norm": 8.960662841796875, "learning_rate": 1.8834324677208763e-05, "loss": 0.011, "step": 1707 }, { "epoch": 0.18516912402428448, "grad_norm": 3.873474597930908, "learning_rate": 1.8833599303641375e-05, "loss": 0.0386, "step": 1708 }, { "epoch": 0.18527753686036427, "grad_norm": 7.126178741455078, "learning_rate": 1.883287393007399e-05, "loss": 0.1105, "step": 1709 }, { "epoch": 0.18538594969644406, "grad_norm": 5.120516777038574, "learning_rate": 1.8832148556506603e-05, "loss": 0.058, "step": 1710 }, { "epoch": 0.18549436253252385, "grad_norm": 3.7303919792175293, "learning_rate": 1.8831423182939215e-05, "loss": 0.029, "step": 1711 }, { "epoch": 0.18560277536860365, "grad_norm": 6.057851791381836, "learning_rate": 1.8830697809371827e-05, "loss": 0.0248, "step": 1712 }, { "epoch": 0.18571118820468344, "grad_norm": 4.2531843185424805, "learning_rate": 1.882997243580444e-05, "loss": 0.0515, "step": 1713 }, { "epoch": 0.18581960104076323, "grad_norm": 5.374675750732422, "learning_rate": 1.882924706223705e-05, "loss": 0.0495, "step": 1714 }, { "epoch": 0.18592801387684302, "grad_norm": 1.6805189847946167, "learning_rate": 1.8828521688669667e-05, "loss": 0.0362, "step": 1715 }, { "epoch": 0.1860364267129228, "grad_norm": 2.5934762954711914, "learning_rate": 1.882779631510228e-05, "loss": 0.023, "step": 1716 }, { "epoch": 0.1861448395490026, "grad_norm": 5.113025188446045, "learning_rate": 1.882707094153489e-05, "loss": 0.1124, "step": 1717 }, { "epoch": 0.1862532523850824, "grad_norm": 4.765270709991455, "learning_rate": 1.8826345567967507e-05, "loss": 0.0323, "step": 1718 }, { "epoch": 0.18636166522116218, "grad_norm": 1.9401779174804688, "learning_rate": 1.882562019440012e-05, "loss": 0.0351, "step": 1719 }, { "epoch": 0.18647007805724197, "grad_norm": 2.739037036895752, "learning_rate": 1.882489482083273e-05, "loss": 0.0404, "step": 1720 }, { "epoch": 0.18657849089332176, "grad_norm": 8.07030963897705, "learning_rate": 1.8824169447265343e-05, "loss": 0.0976, "step": 1721 }, { "epoch": 0.18668690372940155, "grad_norm": 1.6786787509918213, "learning_rate": 1.8823444073697955e-05, "loss": 0.0089, "step": 1722 }, { "epoch": 0.18679531656548135, "grad_norm": 2.2984533309936523, "learning_rate": 1.8822718700130567e-05, "loss": 0.0467, "step": 1723 }, { "epoch": 0.18690372940156114, "grad_norm": 2.128157615661621, "learning_rate": 1.8821993326563183e-05, "loss": 0.0266, "step": 1724 }, { "epoch": 0.18701214223764093, "grad_norm": 3.75980544090271, "learning_rate": 1.8821267952995795e-05, "loss": 0.0695, "step": 1725 }, { "epoch": 0.18712055507372072, "grad_norm": 2.4706525802612305, "learning_rate": 1.8820542579428407e-05, "loss": 0.019, "step": 1726 }, { "epoch": 0.1872289679098005, "grad_norm": 2.905276298522949, "learning_rate": 1.881981720586102e-05, "loss": 0.0136, "step": 1727 }, { "epoch": 0.1873373807458803, "grad_norm": 2.5303237438201904, "learning_rate": 1.881909183229363e-05, "loss": 0.0356, "step": 1728 }, { "epoch": 0.1874457935819601, "grad_norm": 3.756929636001587, "learning_rate": 1.8818366458726244e-05, "loss": 0.025, "step": 1729 }, { "epoch": 0.1875542064180399, "grad_norm": 1.5494335889816284, "learning_rate": 1.881764108515886e-05, "loss": 0.039, "step": 1730 }, { "epoch": 0.1876626192541197, "grad_norm": 3.5698626041412354, "learning_rate": 1.881691571159147e-05, "loss": 0.0624, "step": 1731 }, { "epoch": 0.1877710320901995, "grad_norm": 2.599508047103882, "learning_rate": 1.8816190338024084e-05, "loss": 0.0155, "step": 1732 }, { "epoch": 0.18787944492627928, "grad_norm": 3.5373220443725586, "learning_rate": 1.88154649644567e-05, "loss": 0.026, "step": 1733 }, { "epoch": 0.18798785776235907, "grad_norm": 1.171321153640747, "learning_rate": 1.881473959088931e-05, "loss": 0.0198, "step": 1734 }, { "epoch": 0.18809627059843886, "grad_norm": 5.500809192657471, "learning_rate": 1.8814014217321923e-05, "loss": 0.0948, "step": 1735 }, { "epoch": 0.18820468343451865, "grad_norm": 2.174288034439087, "learning_rate": 1.8813288843754535e-05, "loss": 0.0282, "step": 1736 }, { "epoch": 0.18831309627059845, "grad_norm": 1.8637984991073608, "learning_rate": 1.8812563470187148e-05, "loss": 0.0502, "step": 1737 }, { "epoch": 0.18842150910667824, "grad_norm": 1.270059585571289, "learning_rate": 1.881183809661976e-05, "loss": 0.0334, "step": 1738 }, { "epoch": 0.18852992194275803, "grad_norm": 1.4935879707336426, "learning_rate": 1.8811112723052375e-05, "loss": 0.0267, "step": 1739 }, { "epoch": 0.18863833477883782, "grad_norm": 2.689511775970459, "learning_rate": 1.8810387349484987e-05, "loss": 0.0324, "step": 1740 }, { "epoch": 0.1887467476149176, "grad_norm": 6.184523582458496, "learning_rate": 1.88096619759176e-05, "loss": 0.0606, "step": 1741 }, { "epoch": 0.1888551604509974, "grad_norm": 9.309029579162598, "learning_rate": 1.8808936602350212e-05, "loss": 0.0554, "step": 1742 }, { "epoch": 0.1889635732870772, "grad_norm": 3.4542183876037598, "learning_rate": 1.8808211228782824e-05, "loss": 0.0533, "step": 1743 }, { "epoch": 0.18907198612315698, "grad_norm": 5.597772598266602, "learning_rate": 1.8807485855215436e-05, "loss": 0.069, "step": 1744 }, { "epoch": 0.18918039895923677, "grad_norm": 2.111151695251465, "learning_rate": 1.8806760481648048e-05, "loss": 0.041, "step": 1745 }, { "epoch": 0.18928881179531656, "grad_norm": 3.2351884841918945, "learning_rate": 1.8806035108080664e-05, "loss": 0.0352, "step": 1746 }, { "epoch": 0.18939722463139635, "grad_norm": 4.0817341804504395, "learning_rate": 1.8805309734513276e-05, "loss": 0.075, "step": 1747 }, { "epoch": 0.18950563746747615, "grad_norm": 2.0487570762634277, "learning_rate": 1.8804584360945888e-05, "loss": 0.0432, "step": 1748 }, { "epoch": 0.18961405030355594, "grad_norm": 4.6601386070251465, "learning_rate": 1.88038589873785e-05, "loss": 0.0524, "step": 1749 }, { "epoch": 0.18972246313963573, "grad_norm": 4.216650485992432, "learning_rate": 1.8803133613811116e-05, "loss": 0.0502, "step": 1750 }, { "epoch": 0.18983087597571552, "grad_norm": 1.657463550567627, "learning_rate": 1.8802408240243728e-05, "loss": 0.0216, "step": 1751 }, { "epoch": 0.1899392888117953, "grad_norm": 3.163404941558838, "learning_rate": 1.880168286667634e-05, "loss": 0.0362, "step": 1752 }, { "epoch": 0.1900477016478751, "grad_norm": 3.3968119621276855, "learning_rate": 1.8800957493108952e-05, "loss": 0.0471, "step": 1753 }, { "epoch": 0.1901561144839549, "grad_norm": 3.881002902984619, "learning_rate": 1.8800232119541568e-05, "loss": 0.0346, "step": 1754 }, { "epoch": 0.19026452732003468, "grad_norm": 1.4144635200500488, "learning_rate": 1.879950674597418e-05, "loss": 0.0477, "step": 1755 }, { "epoch": 0.19037294015611447, "grad_norm": 2.9111294746398926, "learning_rate": 1.8798781372406792e-05, "loss": 0.043, "step": 1756 }, { "epoch": 0.19048135299219426, "grad_norm": 4.747547149658203, "learning_rate": 1.8798055998839404e-05, "loss": 0.0917, "step": 1757 }, { "epoch": 0.19058976582827405, "grad_norm": 6.539524555206299, "learning_rate": 1.8797330625272016e-05, "loss": 0.0274, "step": 1758 }, { "epoch": 0.19069817866435385, "grad_norm": 2.457893133163452, "learning_rate": 1.879660525170463e-05, "loss": 0.0403, "step": 1759 }, { "epoch": 0.19080659150043366, "grad_norm": 3.20430588722229, "learning_rate": 1.879587987813724e-05, "loss": 0.0445, "step": 1760 }, { "epoch": 0.19091500433651346, "grad_norm": 1.8889365196228027, "learning_rate": 1.8795154504569856e-05, "loss": 0.0425, "step": 1761 }, { "epoch": 0.19102341717259325, "grad_norm": 2.9654033184051514, "learning_rate": 1.8794429131002468e-05, "loss": 0.0424, "step": 1762 }, { "epoch": 0.19113183000867304, "grad_norm": 2.9464879035949707, "learning_rate": 1.879370375743508e-05, "loss": 0.0764, "step": 1763 }, { "epoch": 0.19124024284475283, "grad_norm": 2.68556809425354, "learning_rate": 1.8792978383867692e-05, "loss": 0.0534, "step": 1764 }, { "epoch": 0.19134865568083262, "grad_norm": 1.256510615348816, "learning_rate": 1.8792253010300305e-05, "loss": 0.0275, "step": 1765 }, { "epoch": 0.1914570685169124, "grad_norm": 4.069701671600342, "learning_rate": 1.8791527636732917e-05, "loss": 0.0781, "step": 1766 }, { "epoch": 0.1915654813529922, "grad_norm": 1.271476149559021, "learning_rate": 1.8790802263165532e-05, "loss": 0.0258, "step": 1767 }, { "epoch": 0.191673894189072, "grad_norm": 3.4687252044677734, "learning_rate": 1.8790076889598144e-05, "loss": 0.0672, "step": 1768 }, { "epoch": 0.19178230702515178, "grad_norm": 3.4007506370544434, "learning_rate": 1.878935151603076e-05, "loss": 0.0664, "step": 1769 }, { "epoch": 0.19189071986123157, "grad_norm": 2.4307336807250977, "learning_rate": 1.8788626142463372e-05, "loss": 0.0149, "step": 1770 }, { "epoch": 0.19199913269731136, "grad_norm": 3.1595728397369385, "learning_rate": 1.8787900768895984e-05, "loss": 0.0533, "step": 1771 }, { "epoch": 0.19210754553339116, "grad_norm": 3.5696446895599365, "learning_rate": 1.8787175395328596e-05, "loss": 0.0823, "step": 1772 }, { "epoch": 0.19221595836947095, "grad_norm": 1.3878051042556763, "learning_rate": 1.878645002176121e-05, "loss": 0.0284, "step": 1773 }, { "epoch": 0.19232437120555074, "grad_norm": 3.2927491664886475, "learning_rate": 1.878572464819382e-05, "loss": 0.0843, "step": 1774 }, { "epoch": 0.19243278404163053, "grad_norm": 4.831797122955322, "learning_rate": 1.8784999274626433e-05, "loss": 0.0461, "step": 1775 }, { "epoch": 0.19254119687771032, "grad_norm": 2.338535785675049, "learning_rate": 1.878427390105905e-05, "loss": 0.0456, "step": 1776 }, { "epoch": 0.1926496097137901, "grad_norm": 3.594252347946167, "learning_rate": 1.878354852749166e-05, "loss": 0.0444, "step": 1777 }, { "epoch": 0.1927580225498699, "grad_norm": 6.168079376220703, "learning_rate": 1.8782823153924273e-05, "loss": 0.0454, "step": 1778 }, { "epoch": 0.1928664353859497, "grad_norm": 4.732415199279785, "learning_rate": 1.8782097780356885e-05, "loss": 0.0607, "step": 1779 }, { "epoch": 0.19297484822202948, "grad_norm": 1.37123441696167, "learning_rate": 1.8781372406789497e-05, "loss": 0.0209, "step": 1780 }, { "epoch": 0.19308326105810927, "grad_norm": 2.4975171089172363, "learning_rate": 1.878064703322211e-05, "loss": 0.0465, "step": 1781 }, { "epoch": 0.19319167389418906, "grad_norm": 2.0887317657470703, "learning_rate": 1.8779921659654725e-05, "loss": 0.0295, "step": 1782 }, { "epoch": 0.19330008673026886, "grad_norm": 3.6916728019714355, "learning_rate": 1.8779196286087337e-05, "loss": 0.0371, "step": 1783 }, { "epoch": 0.19340849956634865, "grad_norm": 0.9713116884231567, "learning_rate": 1.877847091251995e-05, "loss": 0.0245, "step": 1784 }, { "epoch": 0.19351691240242844, "grad_norm": 2.0479559898376465, "learning_rate": 1.8777745538952564e-05, "loss": 0.0453, "step": 1785 }, { "epoch": 0.19362532523850823, "grad_norm": 3.615816831588745, "learning_rate": 1.8777020165385177e-05, "loss": 0.0514, "step": 1786 }, { "epoch": 0.19373373807458802, "grad_norm": 4.149629592895508, "learning_rate": 1.877629479181779e-05, "loss": 0.0509, "step": 1787 }, { "epoch": 0.1938421509106678, "grad_norm": 2.5714848041534424, "learning_rate": 1.87755694182504e-05, "loss": 0.0217, "step": 1788 }, { "epoch": 0.19395056374674763, "grad_norm": 4.823827266693115, "learning_rate": 1.8774844044683013e-05, "loss": 0.0627, "step": 1789 }, { "epoch": 0.19405897658282742, "grad_norm": 2.4648993015289307, "learning_rate": 1.8774118671115625e-05, "loss": 0.0303, "step": 1790 }, { "epoch": 0.1941673894189072, "grad_norm": 3.398707628250122, "learning_rate": 1.877339329754824e-05, "loss": 0.0667, "step": 1791 }, { "epoch": 0.194275802254987, "grad_norm": 10.004213333129883, "learning_rate": 1.8772667923980853e-05, "loss": 0.035, "step": 1792 }, { "epoch": 0.1943842150910668, "grad_norm": 2.8902697563171387, "learning_rate": 1.8771942550413465e-05, "loss": 0.052, "step": 1793 }, { "epoch": 0.19449262792714658, "grad_norm": 5.1142802238464355, "learning_rate": 1.8771217176846077e-05, "loss": 0.0826, "step": 1794 }, { "epoch": 0.19460104076322637, "grad_norm": 11.185250282287598, "learning_rate": 1.877049180327869e-05, "loss": 0.0858, "step": 1795 }, { "epoch": 0.19470945359930616, "grad_norm": 2.663217067718506, "learning_rate": 1.87697664297113e-05, "loss": 0.0154, "step": 1796 }, { "epoch": 0.19481786643538596, "grad_norm": 0.9897289872169495, "learning_rate": 1.8769041056143913e-05, "loss": 0.0129, "step": 1797 }, { "epoch": 0.19492627927146575, "grad_norm": 2.3298797607421875, "learning_rate": 1.876831568257653e-05, "loss": 0.0388, "step": 1798 }, { "epoch": 0.19503469210754554, "grad_norm": 9.4794282913208, "learning_rate": 1.876759030900914e-05, "loss": 0.1952, "step": 1799 }, { "epoch": 0.19514310494362533, "grad_norm": 4.3696722984313965, "learning_rate": 1.8766864935441753e-05, "loss": 0.0777, "step": 1800 }, { "epoch": 0.19525151777970512, "grad_norm": 3.095144748687744, "learning_rate": 1.8766139561874365e-05, "loss": 0.0702, "step": 1801 }, { "epoch": 0.1953599306157849, "grad_norm": 3.7234649658203125, "learning_rate": 1.876541418830698e-05, "loss": 0.0416, "step": 1802 }, { "epoch": 0.1954683434518647, "grad_norm": 4.201516628265381, "learning_rate": 1.8764688814739593e-05, "loss": 0.0663, "step": 1803 }, { "epoch": 0.1955767562879445, "grad_norm": 1.1575225591659546, "learning_rate": 1.8763963441172205e-05, "loss": 0.0233, "step": 1804 }, { "epoch": 0.19568516912402428, "grad_norm": 2.0900442600250244, "learning_rate": 1.8763238067604817e-05, "loss": 0.0485, "step": 1805 }, { "epoch": 0.19579358196010407, "grad_norm": 1.2497777938842773, "learning_rate": 1.8762512694037433e-05, "loss": 0.0142, "step": 1806 }, { "epoch": 0.19590199479618386, "grad_norm": 4.069040298461914, "learning_rate": 1.8761787320470045e-05, "loss": 0.047, "step": 1807 }, { "epoch": 0.19601040763226366, "grad_norm": 3.8930745124816895, "learning_rate": 1.8761061946902657e-05, "loss": 0.0448, "step": 1808 }, { "epoch": 0.19611882046834345, "grad_norm": 1.594922423362732, "learning_rate": 1.876033657333527e-05, "loss": 0.029, "step": 1809 }, { "epoch": 0.19622723330442324, "grad_norm": 5.28524923324585, "learning_rate": 1.875961119976788e-05, "loss": 0.0467, "step": 1810 }, { "epoch": 0.19633564614050303, "grad_norm": 7.054874420166016, "learning_rate": 1.8758885826200494e-05, "loss": 0.1332, "step": 1811 }, { "epoch": 0.19644405897658282, "grad_norm": 1.0667582750320435, "learning_rate": 1.8758160452633106e-05, "loss": 0.0101, "step": 1812 }, { "epoch": 0.1965524718126626, "grad_norm": 3.430934429168701, "learning_rate": 1.875743507906572e-05, "loss": 0.0424, "step": 1813 }, { "epoch": 0.1966608846487424, "grad_norm": 2.3212263584136963, "learning_rate": 1.8756709705498333e-05, "loss": 0.031, "step": 1814 }, { "epoch": 0.1967692974848222, "grad_norm": 1.3087822198867798, "learning_rate": 1.8755984331930946e-05, "loss": 0.0243, "step": 1815 }, { "epoch": 0.19687771032090198, "grad_norm": 1.9394686222076416, "learning_rate": 1.8755258958363558e-05, "loss": 0.0294, "step": 1816 }, { "epoch": 0.19698612315698177, "grad_norm": 1.8840099573135376, "learning_rate": 1.875453358479617e-05, "loss": 0.0107, "step": 1817 }, { "epoch": 0.19709453599306156, "grad_norm": 1.7105519771575928, "learning_rate": 1.8753808211228782e-05, "loss": 0.0327, "step": 1818 }, { "epoch": 0.19720294882914138, "grad_norm": 7.11010217666626, "learning_rate": 1.8753082837661398e-05, "loss": 0.1103, "step": 1819 }, { "epoch": 0.19731136166522117, "grad_norm": 2.0721638202667236, "learning_rate": 1.875235746409401e-05, "loss": 0.049, "step": 1820 }, { "epoch": 0.19741977450130097, "grad_norm": 16.90634536743164, "learning_rate": 1.8751632090526625e-05, "loss": 0.0654, "step": 1821 }, { "epoch": 0.19752818733738076, "grad_norm": 2.045233726501465, "learning_rate": 1.8750906716959237e-05, "loss": 0.0383, "step": 1822 }, { "epoch": 0.19763660017346055, "grad_norm": 2.7606468200683594, "learning_rate": 1.875018134339185e-05, "loss": 0.0352, "step": 1823 }, { "epoch": 0.19774501300954034, "grad_norm": 5.2819061279296875, "learning_rate": 1.8749455969824462e-05, "loss": 0.0324, "step": 1824 }, { "epoch": 0.19785342584562013, "grad_norm": 7.188791751861572, "learning_rate": 1.8748730596257074e-05, "loss": 0.0518, "step": 1825 }, { "epoch": 0.19796183868169992, "grad_norm": 5.6657209396362305, "learning_rate": 1.8748005222689686e-05, "loss": 0.0535, "step": 1826 }, { "epoch": 0.1980702515177797, "grad_norm": 2.9019484519958496, "learning_rate": 1.8747279849122298e-05, "loss": 0.076, "step": 1827 }, { "epoch": 0.1981786643538595, "grad_norm": 4.921034812927246, "learning_rate": 1.8746554475554914e-05, "loss": 0.0402, "step": 1828 }, { "epoch": 0.1982870771899393, "grad_norm": 4.765851974487305, "learning_rate": 1.8745829101987526e-05, "loss": 0.0581, "step": 1829 }, { "epoch": 0.19839549002601908, "grad_norm": 2.605135440826416, "learning_rate": 1.8745103728420138e-05, "loss": 0.0543, "step": 1830 }, { "epoch": 0.19850390286209887, "grad_norm": 3.7803080081939697, "learning_rate": 1.874437835485275e-05, "loss": 0.0443, "step": 1831 }, { "epoch": 0.19861231569817867, "grad_norm": 1.9878581762313843, "learning_rate": 1.8743652981285362e-05, "loss": 0.0347, "step": 1832 }, { "epoch": 0.19872072853425846, "grad_norm": 1.9453648328781128, "learning_rate": 1.8742927607717974e-05, "loss": 0.0409, "step": 1833 }, { "epoch": 0.19882914137033825, "grad_norm": 1.3350743055343628, "learning_rate": 1.874220223415059e-05, "loss": 0.015, "step": 1834 }, { "epoch": 0.19893755420641804, "grad_norm": 1.0316818952560425, "learning_rate": 1.8741476860583202e-05, "loss": 0.0153, "step": 1835 }, { "epoch": 0.19904596704249783, "grad_norm": 4.847692489624023, "learning_rate": 1.8740751487015814e-05, "loss": 0.034, "step": 1836 }, { "epoch": 0.19915437987857762, "grad_norm": 1.4470031261444092, "learning_rate": 1.874002611344843e-05, "loss": 0.0194, "step": 1837 }, { "epoch": 0.1992627927146574, "grad_norm": 4.125117778778076, "learning_rate": 1.8739300739881042e-05, "loss": 0.0495, "step": 1838 }, { "epoch": 0.1993712055507372, "grad_norm": 2.4185397624969482, "learning_rate": 1.8738575366313654e-05, "loss": 0.0423, "step": 1839 }, { "epoch": 0.199479618386817, "grad_norm": 5.176458358764648, "learning_rate": 1.8737849992746266e-05, "loss": 0.0295, "step": 1840 }, { "epoch": 0.19958803122289678, "grad_norm": 1.9741768836975098, "learning_rate": 1.8737124619178878e-05, "loss": 0.0431, "step": 1841 }, { "epoch": 0.19969644405897657, "grad_norm": 3.6994824409484863, "learning_rate": 1.873639924561149e-05, "loss": 0.0553, "step": 1842 }, { "epoch": 0.19980485689505637, "grad_norm": 3.7285959720611572, "learning_rate": 1.8735673872044106e-05, "loss": 0.0543, "step": 1843 }, { "epoch": 0.19991326973113616, "grad_norm": 9.05383014678955, "learning_rate": 1.8734948498476718e-05, "loss": 0.0323, "step": 1844 }, { "epoch": 0.20002168256721595, "grad_norm": 4.029322624206543, "learning_rate": 1.873422312490933e-05, "loss": 0.0285, "step": 1845 }, { "epoch": 0.20013009540329574, "grad_norm": 0.7580206990242004, "learning_rate": 1.8733497751341942e-05, "loss": 0.0134, "step": 1846 }, { "epoch": 0.20023850823937553, "grad_norm": 4.610371112823486, "learning_rate": 1.8732772377774555e-05, "loss": 0.0943, "step": 1847 }, { "epoch": 0.20034692107545535, "grad_norm": 1.3792623281478882, "learning_rate": 1.8732047004207167e-05, "loss": 0.0299, "step": 1848 }, { "epoch": 0.20045533391153514, "grad_norm": 4.2092390060424805, "learning_rate": 1.873132163063978e-05, "loss": 0.0763, "step": 1849 }, { "epoch": 0.20056374674761493, "grad_norm": 4.065605163574219, "learning_rate": 1.8730596257072394e-05, "loss": 0.0239, "step": 1850 }, { "epoch": 0.20067215958369472, "grad_norm": 11.844099998474121, "learning_rate": 1.8729870883505007e-05, "loss": 0.1245, "step": 1851 }, { "epoch": 0.2007805724197745, "grad_norm": 2.0847575664520264, "learning_rate": 1.872914550993762e-05, "loss": 0.0163, "step": 1852 }, { "epoch": 0.2008889852558543, "grad_norm": 1.7159353494644165, "learning_rate": 1.872842013637023e-05, "loss": 0.0287, "step": 1853 }, { "epoch": 0.2009973980919341, "grad_norm": 2.9598846435546875, "learning_rate": 1.8727694762802846e-05, "loss": 0.0288, "step": 1854 }, { "epoch": 0.20110581092801388, "grad_norm": 2.40850830078125, "learning_rate": 1.872696938923546e-05, "loss": 0.061, "step": 1855 }, { "epoch": 0.20121422376409367, "grad_norm": 6.244211673736572, "learning_rate": 1.872624401566807e-05, "loss": 0.0743, "step": 1856 }, { "epoch": 0.20132263660017347, "grad_norm": 6.104053497314453, "learning_rate": 1.8725518642100683e-05, "loss": 0.0551, "step": 1857 }, { "epoch": 0.20143104943625326, "grad_norm": 4.7868876457214355, "learning_rate": 1.8724793268533298e-05, "loss": 0.0213, "step": 1858 }, { "epoch": 0.20153946227233305, "grad_norm": 4.9834489822387695, "learning_rate": 1.872406789496591e-05, "loss": 0.0895, "step": 1859 }, { "epoch": 0.20164787510841284, "grad_norm": 4.263760089874268, "learning_rate": 1.8723342521398523e-05, "loss": 0.0361, "step": 1860 }, { "epoch": 0.20175628794449263, "grad_norm": 11.211773872375488, "learning_rate": 1.8722617147831135e-05, "loss": 0.1032, "step": 1861 }, { "epoch": 0.20186470078057242, "grad_norm": 3.6368887424468994, "learning_rate": 1.8721891774263747e-05, "loss": 0.0549, "step": 1862 }, { "epoch": 0.2019731136166522, "grad_norm": 2.4606916904449463, "learning_rate": 1.872116640069636e-05, "loss": 0.0386, "step": 1863 }, { "epoch": 0.202081526452732, "grad_norm": 1.3627846240997314, "learning_rate": 1.872044102712897e-05, "loss": 0.0343, "step": 1864 }, { "epoch": 0.2021899392888118, "grad_norm": 1.0970255136489868, "learning_rate": 1.8719715653561587e-05, "loss": 0.0178, "step": 1865 }, { "epoch": 0.20229835212489158, "grad_norm": 3.8560900688171387, "learning_rate": 1.87189902799942e-05, "loss": 0.0616, "step": 1866 }, { "epoch": 0.20240676496097137, "grad_norm": 3.291414260864258, "learning_rate": 1.871826490642681e-05, "loss": 0.0322, "step": 1867 }, { "epoch": 0.20251517779705117, "grad_norm": 1.149593472480774, "learning_rate": 1.8717539532859423e-05, "loss": 0.027, "step": 1868 }, { "epoch": 0.20262359063313096, "grad_norm": 1.6921229362487793, "learning_rate": 1.8716814159292035e-05, "loss": 0.0184, "step": 1869 }, { "epoch": 0.20273200346921075, "grad_norm": 0.4794982075691223, "learning_rate": 1.8716088785724647e-05, "loss": 0.006, "step": 1870 }, { "epoch": 0.20284041630529054, "grad_norm": 0.6537759900093079, "learning_rate": 1.8715363412157263e-05, "loss": 0.0231, "step": 1871 }, { "epoch": 0.20294882914137033, "grad_norm": 4.249161243438721, "learning_rate": 1.8714638038589875e-05, "loss": 0.0753, "step": 1872 }, { "epoch": 0.20305724197745012, "grad_norm": 5.235248565673828, "learning_rate": 1.871391266502249e-05, "loss": 0.0582, "step": 1873 }, { "epoch": 0.2031656548135299, "grad_norm": 6.286915302276611, "learning_rate": 1.8713187291455103e-05, "loss": 0.0214, "step": 1874 }, { "epoch": 0.2032740676496097, "grad_norm": 9.014761924743652, "learning_rate": 1.8712461917887715e-05, "loss": 0.0565, "step": 1875 }, { "epoch": 0.2033824804856895, "grad_norm": 4.281393051147461, "learning_rate": 1.8711736544320327e-05, "loss": 0.0265, "step": 1876 }, { "epoch": 0.20349089332176928, "grad_norm": 7.424221515655518, "learning_rate": 1.871101117075294e-05, "loss": 0.0223, "step": 1877 }, { "epoch": 0.2035993061578491, "grad_norm": 6.082618236541748, "learning_rate": 1.871028579718555e-05, "loss": 0.0448, "step": 1878 }, { "epoch": 0.2037077189939289, "grad_norm": 3.4363393783569336, "learning_rate": 1.8709560423618163e-05, "loss": 0.0292, "step": 1879 }, { "epoch": 0.20381613183000868, "grad_norm": 5.163367748260498, "learning_rate": 1.870883505005078e-05, "loss": 0.0447, "step": 1880 }, { "epoch": 0.20392454466608848, "grad_norm": 4.071750164031982, "learning_rate": 1.870810967648339e-05, "loss": 0.034, "step": 1881 }, { "epoch": 0.20403295750216827, "grad_norm": 4.280087471008301, "learning_rate": 1.8707384302916003e-05, "loss": 0.0402, "step": 1882 }, { "epoch": 0.20414137033824806, "grad_norm": 2.542794942855835, "learning_rate": 1.8706658929348615e-05, "loss": 0.0445, "step": 1883 }, { "epoch": 0.20424978317432785, "grad_norm": 3.081498622894287, "learning_rate": 1.8705933555781228e-05, "loss": 0.0236, "step": 1884 }, { "epoch": 0.20435819601040764, "grad_norm": 7.395126819610596, "learning_rate": 1.870520818221384e-05, "loss": 0.0965, "step": 1885 }, { "epoch": 0.20446660884648743, "grad_norm": 6.149028301239014, "learning_rate": 1.8704482808646455e-05, "loss": 0.0656, "step": 1886 }, { "epoch": 0.20457502168256722, "grad_norm": 3.5014052391052246, "learning_rate": 1.8703757435079067e-05, "loss": 0.0414, "step": 1887 }, { "epoch": 0.204683434518647, "grad_norm": 3.915926933288574, "learning_rate": 1.870303206151168e-05, "loss": 0.0383, "step": 1888 }, { "epoch": 0.2047918473547268, "grad_norm": 3.1064977645874023, "learning_rate": 1.8702306687944295e-05, "loss": 0.0248, "step": 1889 }, { "epoch": 0.2049002601908066, "grad_norm": 6.849834442138672, "learning_rate": 1.8701581314376907e-05, "loss": 0.0488, "step": 1890 }, { "epoch": 0.20500867302688638, "grad_norm": 4.392830848693848, "learning_rate": 1.870085594080952e-05, "loss": 0.0656, "step": 1891 }, { "epoch": 0.20511708586296618, "grad_norm": 9.576290130615234, "learning_rate": 1.870013056724213e-05, "loss": 0.0538, "step": 1892 }, { "epoch": 0.20522549869904597, "grad_norm": 12.15164852142334, "learning_rate": 1.8699405193674744e-05, "loss": 0.0827, "step": 1893 }, { "epoch": 0.20533391153512576, "grad_norm": 3.5373830795288086, "learning_rate": 1.8698679820107356e-05, "loss": 0.0711, "step": 1894 }, { "epoch": 0.20544232437120555, "grad_norm": 3.314443349838257, "learning_rate": 1.869795444653997e-05, "loss": 0.0388, "step": 1895 }, { "epoch": 0.20555073720728534, "grad_norm": 5.140676498413086, "learning_rate": 1.8697229072972583e-05, "loss": 0.0301, "step": 1896 }, { "epoch": 0.20565915004336513, "grad_norm": 2.5582520961761475, "learning_rate": 1.8696503699405196e-05, "loss": 0.0543, "step": 1897 }, { "epoch": 0.20576756287944492, "grad_norm": 1.7593055963516235, "learning_rate": 1.8695778325837808e-05, "loss": 0.0418, "step": 1898 }, { "epoch": 0.2058759757155247, "grad_norm": 3.910550594329834, "learning_rate": 1.869505295227042e-05, "loss": 0.0584, "step": 1899 }, { "epoch": 0.2059843885516045, "grad_norm": 1.638284683227539, "learning_rate": 1.8694327578703032e-05, "loss": 0.0277, "step": 1900 }, { "epoch": 0.2060928013876843, "grad_norm": 2.716043472290039, "learning_rate": 1.8693602205135644e-05, "loss": 0.0706, "step": 1901 }, { "epoch": 0.20620121422376408, "grad_norm": 2.8654251098632812, "learning_rate": 1.869287683156826e-05, "loss": 0.0317, "step": 1902 }, { "epoch": 0.20630962705984388, "grad_norm": 2.0290579795837402, "learning_rate": 1.8692151458000872e-05, "loss": 0.058, "step": 1903 }, { "epoch": 0.20641803989592367, "grad_norm": 3.8135578632354736, "learning_rate": 1.8691426084433484e-05, "loss": 0.0314, "step": 1904 }, { "epoch": 0.20652645273200346, "grad_norm": 1.52046537399292, "learning_rate": 1.8690700710866096e-05, "loss": 0.0163, "step": 1905 }, { "epoch": 0.20663486556808325, "grad_norm": 2.30281400680542, "learning_rate": 1.868997533729871e-05, "loss": 0.0587, "step": 1906 }, { "epoch": 0.20674327840416307, "grad_norm": 0.4157661497592926, "learning_rate": 1.8689249963731324e-05, "loss": 0.0148, "step": 1907 }, { "epoch": 0.20685169124024286, "grad_norm": 1.8604278564453125, "learning_rate": 1.8688524590163936e-05, "loss": 0.0384, "step": 1908 }, { "epoch": 0.20696010407632265, "grad_norm": 3.034778118133545, "learning_rate": 1.8687799216596548e-05, "loss": 0.0426, "step": 1909 }, { "epoch": 0.20706851691240244, "grad_norm": 3.5760505199432373, "learning_rate": 1.8687073843029164e-05, "loss": 0.0535, "step": 1910 }, { "epoch": 0.20717692974848223, "grad_norm": 1.2786681652069092, "learning_rate": 1.8686348469461776e-05, "loss": 0.0118, "step": 1911 }, { "epoch": 0.20728534258456202, "grad_norm": 2.6568751335144043, "learning_rate": 1.8685623095894388e-05, "loss": 0.0192, "step": 1912 }, { "epoch": 0.2073937554206418, "grad_norm": 3.4205095767974854, "learning_rate": 1.8684897722327e-05, "loss": 0.0404, "step": 1913 }, { "epoch": 0.2075021682567216, "grad_norm": 2.894912004470825, "learning_rate": 1.8684172348759612e-05, "loss": 0.0402, "step": 1914 }, { "epoch": 0.2076105810928014, "grad_norm": 1.8814215660095215, "learning_rate": 1.8683446975192224e-05, "loss": 0.0312, "step": 1915 }, { "epoch": 0.20771899392888118, "grad_norm": 2.6859564781188965, "learning_rate": 1.8682721601624836e-05, "loss": 0.0483, "step": 1916 }, { "epoch": 0.20782740676496098, "grad_norm": 2.629734992980957, "learning_rate": 1.8681996228057452e-05, "loss": 0.0396, "step": 1917 }, { "epoch": 0.20793581960104077, "grad_norm": 1.7746551036834717, "learning_rate": 1.8681270854490064e-05, "loss": 0.0104, "step": 1918 }, { "epoch": 0.20804423243712056, "grad_norm": 3.546943426132202, "learning_rate": 1.8680545480922676e-05, "loss": 0.0115, "step": 1919 }, { "epoch": 0.20815264527320035, "grad_norm": 6.344888687133789, "learning_rate": 1.867982010735529e-05, "loss": 0.0237, "step": 1920 }, { "epoch": 0.20826105810928014, "grad_norm": 5.017126083374023, "learning_rate": 1.86790947337879e-05, "loss": 0.0466, "step": 1921 }, { "epoch": 0.20836947094535993, "grad_norm": 5.580167293548584, "learning_rate": 1.8678369360220513e-05, "loss": 0.0553, "step": 1922 }, { "epoch": 0.20847788378143972, "grad_norm": 2.66015887260437, "learning_rate": 1.8677643986653128e-05, "loss": 0.0451, "step": 1923 }, { "epoch": 0.2085862966175195, "grad_norm": 5.061932563781738, "learning_rate": 1.867691861308574e-05, "loss": 0.0252, "step": 1924 }, { "epoch": 0.2086947094535993, "grad_norm": 5.548668384552002, "learning_rate": 1.8676193239518356e-05, "loss": 0.035, "step": 1925 }, { "epoch": 0.2088031222896791, "grad_norm": 0.8943248391151428, "learning_rate": 1.8675467865950968e-05, "loss": 0.0175, "step": 1926 }, { "epoch": 0.20891153512575888, "grad_norm": 5.461342811584473, "learning_rate": 1.867474249238358e-05, "loss": 0.0935, "step": 1927 }, { "epoch": 0.20901994796183868, "grad_norm": 1.645192265510559, "learning_rate": 1.8674017118816192e-05, "loss": 0.0513, "step": 1928 }, { "epoch": 0.20912836079791847, "grad_norm": 2.5944290161132812, "learning_rate": 1.8673291745248805e-05, "loss": 0.0367, "step": 1929 }, { "epoch": 0.20923677363399826, "grad_norm": 9.972146034240723, "learning_rate": 1.8672566371681417e-05, "loss": 0.0468, "step": 1930 }, { "epoch": 0.20934518647007805, "grad_norm": 2.4978153705596924, "learning_rate": 1.867184099811403e-05, "loss": 0.0393, "step": 1931 }, { "epoch": 0.20945359930615784, "grad_norm": 4.726276874542236, "learning_rate": 1.8671115624546644e-05, "loss": 0.0636, "step": 1932 }, { "epoch": 0.20956201214223763, "grad_norm": 2.855945587158203, "learning_rate": 1.8670390250979256e-05, "loss": 0.0249, "step": 1933 }, { "epoch": 0.20967042497831742, "grad_norm": 2.931718111038208, "learning_rate": 1.866966487741187e-05, "loss": 0.0559, "step": 1934 }, { "epoch": 0.2097788378143972, "grad_norm": 0.4301213026046753, "learning_rate": 1.866893950384448e-05, "loss": 0.0122, "step": 1935 }, { "epoch": 0.209887250650477, "grad_norm": 4.286283016204834, "learning_rate": 1.8668214130277093e-05, "loss": 0.0299, "step": 1936 }, { "epoch": 0.20999566348655682, "grad_norm": 1.8554677963256836, "learning_rate": 1.8667488756709705e-05, "loss": 0.0294, "step": 1937 }, { "epoch": 0.2101040763226366, "grad_norm": 5.123551845550537, "learning_rate": 1.866676338314232e-05, "loss": 0.0442, "step": 1938 }, { "epoch": 0.2102124891587164, "grad_norm": 2.298487424850464, "learning_rate": 1.8666038009574933e-05, "loss": 0.0434, "step": 1939 }, { "epoch": 0.2103209019947962, "grad_norm": 3.6581997871398926, "learning_rate": 1.8665312636007545e-05, "loss": 0.0636, "step": 1940 }, { "epoch": 0.21042931483087599, "grad_norm": 2.8374500274658203, "learning_rate": 1.866458726244016e-05, "loss": 0.0293, "step": 1941 }, { "epoch": 0.21053772766695578, "grad_norm": 7.696929454803467, "learning_rate": 1.8663861888872773e-05, "loss": 0.0605, "step": 1942 }, { "epoch": 0.21064614050303557, "grad_norm": 2.90761399269104, "learning_rate": 1.8663136515305385e-05, "loss": 0.0692, "step": 1943 }, { "epoch": 0.21075455333911536, "grad_norm": 3.904369831085205, "learning_rate": 1.8662411141737997e-05, "loss": 0.1115, "step": 1944 }, { "epoch": 0.21086296617519515, "grad_norm": 1.5237563848495483, "learning_rate": 1.866168576817061e-05, "loss": 0.0528, "step": 1945 }, { "epoch": 0.21097137901127494, "grad_norm": 0.7372379899024963, "learning_rate": 1.866096039460322e-05, "loss": 0.0159, "step": 1946 }, { "epoch": 0.21107979184735473, "grad_norm": 3.1505513191223145, "learning_rate": 1.8660235021035837e-05, "loss": 0.0619, "step": 1947 }, { "epoch": 0.21118820468343452, "grad_norm": 1.562857985496521, "learning_rate": 1.865950964746845e-05, "loss": 0.0304, "step": 1948 }, { "epoch": 0.2112966175195143, "grad_norm": 1.4528892040252686, "learning_rate": 1.865878427390106e-05, "loss": 0.0388, "step": 1949 }, { "epoch": 0.2114050303555941, "grad_norm": 2.631692886352539, "learning_rate": 1.8658058900333673e-05, "loss": 0.0728, "step": 1950 }, { "epoch": 0.2115134431916739, "grad_norm": 1.404726505279541, "learning_rate": 1.8657333526766285e-05, "loss": 0.0442, "step": 1951 }, { "epoch": 0.21162185602775369, "grad_norm": 4.718901634216309, "learning_rate": 1.8656608153198897e-05, "loss": 0.0545, "step": 1952 }, { "epoch": 0.21173026886383348, "grad_norm": 2.296643018722534, "learning_rate": 1.865588277963151e-05, "loss": 0.035, "step": 1953 }, { "epoch": 0.21183868169991327, "grad_norm": 1.305454134941101, "learning_rate": 1.8655157406064125e-05, "loss": 0.0253, "step": 1954 }, { "epoch": 0.21194709453599306, "grad_norm": 0.88303142786026, "learning_rate": 1.8654432032496737e-05, "loss": 0.0193, "step": 1955 }, { "epoch": 0.21205550737207285, "grad_norm": 6.764994144439697, "learning_rate": 1.865370665892935e-05, "loss": 0.0375, "step": 1956 }, { "epoch": 0.21216392020815264, "grad_norm": 2.6388790607452393, "learning_rate": 1.865298128536196e-05, "loss": 0.0481, "step": 1957 }, { "epoch": 0.21227233304423243, "grad_norm": 5.436862945556641, "learning_rate": 1.8652255911794577e-05, "loss": 0.0232, "step": 1958 }, { "epoch": 0.21238074588031222, "grad_norm": 2.8281404972076416, "learning_rate": 1.865153053822719e-05, "loss": 0.0273, "step": 1959 }, { "epoch": 0.212489158716392, "grad_norm": 3.041167974472046, "learning_rate": 1.86508051646598e-05, "loss": 0.0706, "step": 1960 }, { "epoch": 0.2125975715524718, "grad_norm": 1.460571050643921, "learning_rate": 1.8650079791092413e-05, "loss": 0.0239, "step": 1961 }, { "epoch": 0.2127059843885516, "grad_norm": 3.49908709526062, "learning_rate": 1.864935441752503e-05, "loss": 0.0328, "step": 1962 }, { "epoch": 0.21281439722463139, "grad_norm": 5.716815948486328, "learning_rate": 1.864862904395764e-05, "loss": 0.1057, "step": 1963 }, { "epoch": 0.21292281006071118, "grad_norm": 6.099899768829346, "learning_rate": 1.8647903670390253e-05, "loss": 0.045, "step": 1964 }, { "epoch": 0.21303122289679097, "grad_norm": 2.5405941009521484, "learning_rate": 1.8647178296822865e-05, "loss": 0.0239, "step": 1965 }, { "epoch": 0.21313963573287076, "grad_norm": 2.8335931301116943, "learning_rate": 1.8646452923255478e-05, "loss": 0.0613, "step": 1966 }, { "epoch": 0.21324804856895058, "grad_norm": 3.2391598224639893, "learning_rate": 1.864572754968809e-05, "loss": 0.0322, "step": 1967 }, { "epoch": 0.21335646140503037, "grad_norm": 3.078348159790039, "learning_rate": 1.8645002176120702e-05, "loss": 0.0668, "step": 1968 }, { "epoch": 0.21346487424111016, "grad_norm": 2.2072694301605225, "learning_rate": 1.8644276802553317e-05, "loss": 0.0285, "step": 1969 }, { "epoch": 0.21357328707718995, "grad_norm": 4.996213436126709, "learning_rate": 1.864355142898593e-05, "loss": 0.0483, "step": 1970 }, { "epoch": 0.21368169991326974, "grad_norm": 4.570842742919922, "learning_rate": 1.864282605541854e-05, "loss": 0.0111, "step": 1971 }, { "epoch": 0.21379011274934953, "grad_norm": 3.4045798778533936, "learning_rate": 1.8642100681851154e-05, "loss": 0.0767, "step": 1972 }, { "epoch": 0.21389852558542932, "grad_norm": 5.723826885223389, "learning_rate": 1.8641375308283766e-05, "loss": 0.0502, "step": 1973 }, { "epoch": 0.2140069384215091, "grad_norm": 3.45723032951355, "learning_rate": 1.8640649934716378e-05, "loss": 0.037, "step": 1974 }, { "epoch": 0.2141153512575889, "grad_norm": 4.042830944061279, "learning_rate": 1.8639924561148994e-05, "loss": 0.035, "step": 1975 }, { "epoch": 0.2142237640936687, "grad_norm": 1.6292234659194946, "learning_rate": 1.8639199187581606e-05, "loss": 0.0288, "step": 1976 }, { "epoch": 0.21433217692974849, "grad_norm": 8.189017295837402, "learning_rate": 1.863847381401422e-05, "loss": 0.06, "step": 1977 }, { "epoch": 0.21444058976582828, "grad_norm": 7.678196907043457, "learning_rate": 1.8637748440446833e-05, "loss": 0.0748, "step": 1978 }, { "epoch": 0.21454900260190807, "grad_norm": 2.1545324325561523, "learning_rate": 1.8637023066879446e-05, "loss": 0.0308, "step": 1979 }, { "epoch": 0.21465741543798786, "grad_norm": 5.143594741821289, "learning_rate": 1.8636297693312058e-05, "loss": 0.0288, "step": 1980 }, { "epoch": 0.21476582827406765, "grad_norm": 4.155313491821289, "learning_rate": 1.863557231974467e-05, "loss": 0.0623, "step": 1981 }, { "epoch": 0.21487424111014744, "grad_norm": 2.9132087230682373, "learning_rate": 1.8634846946177282e-05, "loss": 0.0483, "step": 1982 }, { "epoch": 0.21498265394622723, "grad_norm": 5.384251117706299, "learning_rate": 1.8634121572609894e-05, "loss": 0.0769, "step": 1983 }, { "epoch": 0.21509106678230702, "grad_norm": 5.284180164337158, "learning_rate": 1.863339619904251e-05, "loss": 0.0351, "step": 1984 }, { "epoch": 0.2151994796183868, "grad_norm": 1.9795001745224, "learning_rate": 1.8632670825475122e-05, "loss": 0.0286, "step": 1985 }, { "epoch": 0.2153078924544666, "grad_norm": 3.6244258880615234, "learning_rate": 1.8631945451907734e-05, "loss": 0.083, "step": 1986 }, { "epoch": 0.2154163052905464, "grad_norm": 2.759110927581787, "learning_rate": 1.8631220078340346e-05, "loss": 0.0493, "step": 1987 }, { "epoch": 0.21552471812662619, "grad_norm": 2.3940837383270264, "learning_rate": 1.8630494704772958e-05, "loss": 0.0258, "step": 1988 }, { "epoch": 0.21563313096270598, "grad_norm": 3.364384412765503, "learning_rate": 1.862976933120557e-05, "loss": 0.0357, "step": 1989 }, { "epoch": 0.21574154379878577, "grad_norm": 2.8774030208587646, "learning_rate": 1.8629043957638186e-05, "loss": 0.0525, "step": 1990 }, { "epoch": 0.21584995663486556, "grad_norm": 2.247885227203369, "learning_rate": 1.8628318584070798e-05, "loss": 0.0192, "step": 1991 }, { "epoch": 0.21595836947094535, "grad_norm": 1.1869213581085205, "learning_rate": 1.862759321050341e-05, "loss": 0.032, "step": 1992 }, { "epoch": 0.21606678230702514, "grad_norm": 2.6712942123413086, "learning_rate": 1.8626867836936026e-05, "loss": 0.0386, "step": 1993 }, { "epoch": 0.21617519514310493, "grad_norm": 1.4412792921066284, "learning_rate": 1.8626142463368638e-05, "loss": 0.0436, "step": 1994 }, { "epoch": 0.21628360797918472, "grad_norm": 2.30116605758667, "learning_rate": 1.862541708980125e-05, "loss": 0.0224, "step": 1995 }, { "epoch": 0.21639202081526454, "grad_norm": 1.9316219091415405, "learning_rate": 1.8624691716233862e-05, "loss": 0.0092, "step": 1996 }, { "epoch": 0.21650043365134433, "grad_norm": 8.752148628234863, "learning_rate": 1.8623966342666474e-05, "loss": 0.0418, "step": 1997 }, { "epoch": 0.21660884648742412, "grad_norm": 4.391937732696533, "learning_rate": 1.8623240969099086e-05, "loss": 0.0512, "step": 1998 }, { "epoch": 0.2167172593235039, "grad_norm": 2.757997989654541, "learning_rate": 1.8622515595531702e-05, "loss": 0.0197, "step": 1999 }, { "epoch": 0.2168256721595837, "grad_norm": 4.2663164138793945, "learning_rate": 1.8621790221964314e-05, "loss": 0.0551, "step": 2000 }, { "epoch": 0.2169340849956635, "grad_norm": 10.376988410949707, "learning_rate": 1.8621064848396926e-05, "loss": 0.1354, "step": 2001 }, { "epoch": 0.2170424978317433, "grad_norm": 0.8230305910110474, "learning_rate": 1.862033947482954e-05, "loss": 0.0083, "step": 2002 }, { "epoch": 0.21715091066782308, "grad_norm": 5.454364776611328, "learning_rate": 1.861961410126215e-05, "loss": 0.0549, "step": 2003 }, { "epoch": 0.21725932350390287, "grad_norm": 12.435718536376953, "learning_rate": 1.8618888727694763e-05, "loss": 0.0935, "step": 2004 }, { "epoch": 0.21736773633998266, "grad_norm": 8.55462646484375, "learning_rate": 1.8618163354127375e-05, "loss": 0.1081, "step": 2005 }, { "epoch": 0.21747614917606245, "grad_norm": 2.76648211479187, "learning_rate": 1.861743798055999e-05, "loss": 0.019, "step": 2006 }, { "epoch": 0.21758456201214224, "grad_norm": 7.607486724853516, "learning_rate": 1.8616712606992603e-05, "loss": 0.0407, "step": 2007 }, { "epoch": 0.21769297484822203, "grad_norm": 1.3832521438598633, "learning_rate": 1.8615987233425215e-05, "loss": 0.0056, "step": 2008 }, { "epoch": 0.21780138768430182, "grad_norm": 4.243964672088623, "learning_rate": 1.861526185985783e-05, "loss": 0.085, "step": 2009 }, { "epoch": 0.2179098005203816, "grad_norm": 3.72666072845459, "learning_rate": 1.8614536486290442e-05, "loss": 0.0807, "step": 2010 }, { "epoch": 0.2180182133564614, "grad_norm": 2.2045540809631348, "learning_rate": 1.8613811112723054e-05, "loss": 0.0331, "step": 2011 }, { "epoch": 0.2181266261925412, "grad_norm": 1.533197045326233, "learning_rate": 1.8613085739155667e-05, "loss": 0.0304, "step": 2012 }, { "epoch": 0.218235039028621, "grad_norm": 3.764500856399536, "learning_rate": 1.861236036558828e-05, "loss": 0.0385, "step": 2013 }, { "epoch": 0.21834345186470078, "grad_norm": 6.463815689086914, "learning_rate": 1.8611634992020894e-05, "loss": 0.0446, "step": 2014 }, { "epoch": 0.21845186470078057, "grad_norm": 4.745038032531738, "learning_rate": 1.8610909618453506e-05, "loss": 0.0844, "step": 2015 }, { "epoch": 0.21856027753686036, "grad_norm": 3.600796699523926, "learning_rate": 1.861018424488612e-05, "loss": 0.0272, "step": 2016 }, { "epoch": 0.21866869037294015, "grad_norm": 6.4561028480529785, "learning_rate": 1.860945887131873e-05, "loss": 0.0312, "step": 2017 }, { "epoch": 0.21877710320901994, "grad_norm": 1.43626070022583, "learning_rate": 1.8608733497751343e-05, "loss": 0.0162, "step": 2018 }, { "epoch": 0.21888551604509973, "grad_norm": 7.704798221588135, "learning_rate": 1.8608008124183955e-05, "loss": 0.0926, "step": 2019 }, { "epoch": 0.21899392888117952, "grad_norm": 3.1802337169647217, "learning_rate": 1.8607282750616567e-05, "loss": 0.033, "step": 2020 }, { "epoch": 0.2191023417172593, "grad_norm": 5.571354389190674, "learning_rate": 1.8606557377049183e-05, "loss": 0.0609, "step": 2021 }, { "epoch": 0.2192107545533391, "grad_norm": 3.835280656814575, "learning_rate": 1.8605832003481795e-05, "loss": 0.1042, "step": 2022 }, { "epoch": 0.2193191673894189, "grad_norm": 1.8684300184249878, "learning_rate": 1.8605106629914407e-05, "loss": 0.0514, "step": 2023 }, { "epoch": 0.2194275802254987, "grad_norm": 3.5838959217071533, "learning_rate": 1.860438125634702e-05, "loss": 0.0747, "step": 2024 }, { "epoch": 0.21953599306157848, "grad_norm": 4.5568623542785645, "learning_rate": 1.860365588277963e-05, "loss": 0.0177, "step": 2025 }, { "epoch": 0.2196444058976583, "grad_norm": 4.506129264831543, "learning_rate": 1.8602930509212243e-05, "loss": 0.0315, "step": 2026 }, { "epoch": 0.2197528187337381, "grad_norm": 1.7641839981079102, "learning_rate": 1.860220513564486e-05, "loss": 0.0232, "step": 2027 }, { "epoch": 0.21986123156981788, "grad_norm": 2.2906386852264404, "learning_rate": 1.860147976207747e-05, "loss": 0.0311, "step": 2028 }, { "epoch": 0.21996964440589767, "grad_norm": 5.365222454071045, "learning_rate": 1.8600754388510087e-05, "loss": 0.0417, "step": 2029 }, { "epoch": 0.22007805724197746, "grad_norm": 3.9442365169525146, "learning_rate": 1.86000290149427e-05, "loss": 0.0778, "step": 2030 }, { "epoch": 0.22018647007805725, "grad_norm": 6.816285133361816, "learning_rate": 1.859930364137531e-05, "loss": 0.0473, "step": 2031 }, { "epoch": 0.22029488291413704, "grad_norm": 2.550457000732422, "learning_rate": 1.8598578267807923e-05, "loss": 0.0159, "step": 2032 }, { "epoch": 0.22040329575021683, "grad_norm": 2.4390032291412354, "learning_rate": 1.8597852894240535e-05, "loss": 0.0392, "step": 2033 }, { "epoch": 0.22051170858629662, "grad_norm": 2.0388622283935547, "learning_rate": 1.8597127520673147e-05, "loss": 0.0306, "step": 2034 }, { "epoch": 0.22062012142237641, "grad_norm": 3.4650070667266846, "learning_rate": 1.859640214710576e-05, "loss": 0.0512, "step": 2035 }, { "epoch": 0.2207285342584562, "grad_norm": 2.0178494453430176, "learning_rate": 1.8595676773538375e-05, "loss": 0.0336, "step": 2036 }, { "epoch": 0.220836947094536, "grad_norm": 1.37201988697052, "learning_rate": 1.8594951399970987e-05, "loss": 0.0231, "step": 2037 }, { "epoch": 0.2209453599306158, "grad_norm": 2.6714706420898438, "learning_rate": 1.85942260264036e-05, "loss": 0.0303, "step": 2038 }, { "epoch": 0.22105377276669558, "grad_norm": 5.108695983886719, "learning_rate": 1.859350065283621e-05, "loss": 0.0507, "step": 2039 }, { "epoch": 0.22116218560277537, "grad_norm": 1.5161573886871338, "learning_rate": 1.8592775279268824e-05, "loss": 0.0162, "step": 2040 }, { "epoch": 0.22127059843885516, "grad_norm": 2.4456608295440674, "learning_rate": 1.8592049905701436e-05, "loss": 0.0432, "step": 2041 }, { "epoch": 0.22137901127493495, "grad_norm": 2.9961752891540527, "learning_rate": 1.859132453213405e-05, "loss": 0.0414, "step": 2042 }, { "epoch": 0.22148742411101474, "grad_norm": 1.6956781148910522, "learning_rate": 1.8590599158566663e-05, "loss": 0.0246, "step": 2043 }, { "epoch": 0.22159583694709453, "grad_norm": 4.466977119445801, "learning_rate": 1.8589873784999276e-05, "loss": 0.048, "step": 2044 }, { "epoch": 0.22170424978317432, "grad_norm": 2.3768792152404785, "learning_rate": 1.858914841143189e-05, "loss": 0.0642, "step": 2045 }, { "epoch": 0.22181266261925411, "grad_norm": 3.457324504852295, "learning_rate": 1.8588423037864503e-05, "loss": 0.0458, "step": 2046 }, { "epoch": 0.2219210754553339, "grad_norm": 3.4797682762145996, "learning_rate": 1.8587697664297115e-05, "loss": 0.0594, "step": 2047 }, { "epoch": 0.2220294882914137, "grad_norm": 3.3729019165039062, "learning_rate": 1.8586972290729728e-05, "loss": 0.0252, "step": 2048 }, { "epoch": 0.2221379011274935, "grad_norm": 5.625702857971191, "learning_rate": 1.858624691716234e-05, "loss": 0.0449, "step": 2049 }, { "epoch": 0.22224631396357328, "grad_norm": 3.6980791091918945, "learning_rate": 1.8585521543594952e-05, "loss": 0.0255, "step": 2050 }, { "epoch": 0.22235472679965307, "grad_norm": 2.80441951751709, "learning_rate": 1.8584796170027567e-05, "loss": 0.0429, "step": 2051 }, { "epoch": 0.22246313963573286, "grad_norm": 2.221177339553833, "learning_rate": 1.858407079646018e-05, "loss": 0.0417, "step": 2052 }, { "epoch": 0.22257155247181265, "grad_norm": 4.930531024932861, "learning_rate": 1.858334542289279e-05, "loss": 0.0343, "step": 2053 }, { "epoch": 0.22267996530789244, "grad_norm": 2.7757508754730225, "learning_rate": 1.8582620049325404e-05, "loss": 0.0799, "step": 2054 }, { "epoch": 0.22278837814397226, "grad_norm": 2.5565848350524902, "learning_rate": 1.8581894675758016e-05, "loss": 0.0374, "step": 2055 }, { "epoch": 0.22289679098005205, "grad_norm": 1.3138514757156372, "learning_rate": 1.8581169302190628e-05, "loss": 0.0115, "step": 2056 }, { "epoch": 0.22300520381613184, "grad_norm": 5.2762556076049805, "learning_rate": 1.858044392862324e-05, "loss": 0.0382, "step": 2057 }, { "epoch": 0.22311361665221163, "grad_norm": 7.803668022155762, "learning_rate": 1.8579718555055856e-05, "loss": 0.0306, "step": 2058 }, { "epoch": 0.22322202948829142, "grad_norm": 3.5136587619781494, "learning_rate": 1.8578993181488468e-05, "loss": 0.0169, "step": 2059 }, { "epoch": 0.22333044232437121, "grad_norm": 1.7283376455307007, "learning_rate": 1.857826780792108e-05, "loss": 0.0355, "step": 2060 }, { "epoch": 0.223438855160451, "grad_norm": 2.160999298095703, "learning_rate": 1.8577542434353696e-05, "loss": 0.0362, "step": 2061 }, { "epoch": 0.2235472679965308, "grad_norm": 3.5223448276519775, "learning_rate": 1.8576817060786308e-05, "loss": 0.083, "step": 2062 }, { "epoch": 0.2236556808326106, "grad_norm": 1.5266669988632202, "learning_rate": 1.857609168721892e-05, "loss": 0.0299, "step": 2063 }, { "epoch": 0.22376409366869038, "grad_norm": 1.065332055091858, "learning_rate": 1.8575366313651532e-05, "loss": 0.0252, "step": 2064 }, { "epoch": 0.22387250650477017, "grad_norm": 1.8727484941482544, "learning_rate": 1.8574640940084144e-05, "loss": 0.0166, "step": 2065 }, { "epoch": 0.22398091934084996, "grad_norm": 4.677763938903809, "learning_rate": 1.857391556651676e-05, "loss": 0.0339, "step": 2066 }, { "epoch": 0.22408933217692975, "grad_norm": 2.8796446323394775, "learning_rate": 1.8573190192949372e-05, "loss": 0.0343, "step": 2067 }, { "epoch": 0.22419774501300954, "grad_norm": 2.7613320350646973, "learning_rate": 1.8572464819381984e-05, "loss": 0.0269, "step": 2068 }, { "epoch": 0.22430615784908933, "grad_norm": 12.932377815246582, "learning_rate": 1.8571739445814596e-05, "loss": 0.1301, "step": 2069 }, { "epoch": 0.22441457068516912, "grad_norm": 12.157830238342285, "learning_rate": 1.8571014072247208e-05, "loss": 0.0619, "step": 2070 }, { "epoch": 0.22452298352124891, "grad_norm": 3.720839023590088, "learning_rate": 1.857028869867982e-05, "loss": 0.0598, "step": 2071 }, { "epoch": 0.2246313963573287, "grad_norm": 3.7272355556488037, "learning_rate": 1.8569563325112433e-05, "loss": 0.0252, "step": 2072 }, { "epoch": 0.2247398091934085, "grad_norm": 3.481440544128418, "learning_rate": 1.8568837951545048e-05, "loss": 0.0368, "step": 2073 }, { "epoch": 0.2248482220294883, "grad_norm": 3.3832387924194336, "learning_rate": 1.856811257797766e-05, "loss": 0.0537, "step": 2074 }, { "epoch": 0.22495663486556808, "grad_norm": 4.154917240142822, "learning_rate": 1.8567387204410272e-05, "loss": 0.0636, "step": 2075 }, { "epoch": 0.22506504770164787, "grad_norm": 4.8399481773376465, "learning_rate": 1.8566661830842884e-05, "loss": 0.0597, "step": 2076 }, { "epoch": 0.22517346053772766, "grad_norm": 2.196871519088745, "learning_rate": 1.8565936457275497e-05, "loss": 0.0297, "step": 2077 }, { "epoch": 0.22528187337380745, "grad_norm": 3.5722484588623047, "learning_rate": 1.856521108370811e-05, "loss": 0.0522, "step": 2078 }, { "epoch": 0.22539028620988724, "grad_norm": 3.4415993690490723, "learning_rate": 1.8564485710140724e-05, "loss": 0.0558, "step": 2079 }, { "epoch": 0.22549869904596703, "grad_norm": 4.060909748077393, "learning_rate": 1.8563760336573336e-05, "loss": 0.1178, "step": 2080 }, { "epoch": 0.22560711188204682, "grad_norm": 2.8827085494995117, "learning_rate": 1.8563034963005952e-05, "loss": 0.1006, "step": 2081 }, { "epoch": 0.22571552471812661, "grad_norm": 0.3476030230522156, "learning_rate": 1.8562309589438564e-05, "loss": 0.0034, "step": 2082 }, { "epoch": 0.2258239375542064, "grad_norm": 2.243060827255249, "learning_rate": 1.8561584215871176e-05, "loss": 0.0397, "step": 2083 }, { "epoch": 0.2259323503902862, "grad_norm": 1.24293053150177, "learning_rate": 1.856085884230379e-05, "loss": 0.0144, "step": 2084 }, { "epoch": 0.22604076322636601, "grad_norm": 3.950688362121582, "learning_rate": 1.85601334687364e-05, "loss": 0.0527, "step": 2085 }, { "epoch": 0.2261491760624458, "grad_norm": 0.7265391945838928, "learning_rate": 1.8559408095169013e-05, "loss": 0.0271, "step": 2086 }, { "epoch": 0.2262575888985256, "grad_norm": 4.998872756958008, "learning_rate": 1.8558682721601625e-05, "loss": 0.1126, "step": 2087 }, { "epoch": 0.2263660017346054, "grad_norm": 2.468989610671997, "learning_rate": 1.855795734803424e-05, "loss": 0.0503, "step": 2088 }, { "epoch": 0.22647441457068518, "grad_norm": 1.4757988452911377, "learning_rate": 1.8557231974466852e-05, "loss": 0.0163, "step": 2089 }, { "epoch": 0.22658282740676497, "grad_norm": 3.479743003845215, "learning_rate": 1.8556506600899465e-05, "loss": 0.0333, "step": 2090 }, { "epoch": 0.22669124024284476, "grad_norm": 1.070991039276123, "learning_rate": 1.8555781227332077e-05, "loss": 0.0222, "step": 2091 }, { "epoch": 0.22679965307892455, "grad_norm": 3.464447021484375, "learning_rate": 1.855505585376469e-05, "loss": 0.0804, "step": 2092 }, { "epoch": 0.22690806591500434, "grad_norm": 2.8617517948150635, "learning_rate": 1.85543304801973e-05, "loss": 0.0259, "step": 2093 }, { "epoch": 0.22701647875108413, "grad_norm": 5.801333904266357, "learning_rate": 1.8553605106629917e-05, "loss": 0.0999, "step": 2094 }, { "epoch": 0.22712489158716392, "grad_norm": 1.167912244796753, "learning_rate": 1.855287973306253e-05, "loss": 0.0185, "step": 2095 }, { "epoch": 0.22723330442324371, "grad_norm": 5.88765811920166, "learning_rate": 1.855215435949514e-05, "loss": 0.0462, "step": 2096 }, { "epoch": 0.2273417172593235, "grad_norm": 2.772599220275879, "learning_rate": 1.8551428985927756e-05, "loss": 0.0381, "step": 2097 }, { "epoch": 0.2274501300954033, "grad_norm": 17.722431182861328, "learning_rate": 1.855070361236037e-05, "loss": 0.1229, "step": 2098 }, { "epoch": 0.2275585429314831, "grad_norm": 7.550802707672119, "learning_rate": 1.854997823879298e-05, "loss": 0.0759, "step": 2099 }, { "epoch": 0.22766695576756288, "grad_norm": 8.753244400024414, "learning_rate": 1.8549252865225593e-05, "loss": 0.0586, "step": 2100 }, { "epoch": 0.22777536860364267, "grad_norm": 3.7986648082733154, "learning_rate": 1.8548527491658205e-05, "loss": 0.0319, "step": 2101 }, { "epoch": 0.22788378143972246, "grad_norm": 5.641029357910156, "learning_rate": 1.8547802118090817e-05, "loss": 0.0286, "step": 2102 }, { "epoch": 0.22799219427580225, "grad_norm": 4.546044826507568, "learning_rate": 1.8547076744523433e-05, "loss": 0.0773, "step": 2103 }, { "epoch": 0.22810060711188204, "grad_norm": 5.328761100769043, "learning_rate": 1.8546351370956045e-05, "loss": 0.0319, "step": 2104 }, { "epoch": 0.22820901994796183, "grad_norm": 5.621862888336182, "learning_rate": 1.8545625997388657e-05, "loss": 0.0316, "step": 2105 }, { "epoch": 0.22831743278404162, "grad_norm": 2.6781628131866455, "learning_rate": 1.854490062382127e-05, "loss": 0.0325, "step": 2106 }, { "epoch": 0.22842584562012141, "grad_norm": 8.858444213867188, "learning_rate": 1.854417525025388e-05, "loss": 0.0791, "step": 2107 }, { "epoch": 0.2285342584562012, "grad_norm": 2.913025140762329, "learning_rate": 1.8543449876686493e-05, "loss": 0.0429, "step": 2108 }, { "epoch": 0.228642671292281, "grad_norm": 2.9093704223632812, "learning_rate": 1.8542724503119106e-05, "loss": 0.0578, "step": 2109 }, { "epoch": 0.2287510841283608, "grad_norm": 1.5569632053375244, "learning_rate": 1.854199912955172e-05, "loss": 0.0337, "step": 2110 }, { "epoch": 0.22885949696444058, "grad_norm": 3.9003169536590576, "learning_rate": 1.8541273755984333e-05, "loss": 0.0859, "step": 2111 }, { "epoch": 0.22896790980052037, "grad_norm": 2.120272159576416, "learning_rate": 1.8540548382416945e-05, "loss": 0.0675, "step": 2112 }, { "epoch": 0.22907632263660016, "grad_norm": 3.5751917362213135, "learning_rate": 1.853982300884956e-05, "loss": 0.0284, "step": 2113 }, { "epoch": 0.22918473547267998, "grad_norm": 2.751748561859131, "learning_rate": 1.8539097635282173e-05, "loss": 0.0405, "step": 2114 }, { "epoch": 0.22929314830875977, "grad_norm": 4.085898399353027, "learning_rate": 1.8538372261714785e-05, "loss": 0.0458, "step": 2115 }, { "epoch": 0.22940156114483956, "grad_norm": 1.7627034187316895, "learning_rate": 1.8537646888147397e-05, "loss": 0.0439, "step": 2116 }, { "epoch": 0.22950997398091935, "grad_norm": 5.053165435791016, "learning_rate": 1.853692151458001e-05, "loss": 0.1137, "step": 2117 }, { "epoch": 0.22961838681699914, "grad_norm": 0.5215350985527039, "learning_rate": 1.8536196141012625e-05, "loss": 0.0085, "step": 2118 }, { "epoch": 0.22972679965307893, "grad_norm": 1.7100955247879028, "learning_rate": 1.8535470767445237e-05, "loss": 0.0218, "step": 2119 }, { "epoch": 0.22983521248915872, "grad_norm": 3.7407519817352295, "learning_rate": 1.853474539387785e-05, "loss": 0.0636, "step": 2120 }, { "epoch": 0.22994362532523852, "grad_norm": 3.9875926971435547, "learning_rate": 1.853402002031046e-05, "loss": 0.067, "step": 2121 }, { "epoch": 0.2300520381613183, "grad_norm": 4.835668087005615, "learning_rate": 1.8533294646743074e-05, "loss": 0.0563, "step": 2122 }, { "epoch": 0.2301604509973981, "grad_norm": 3.6783812046051025, "learning_rate": 1.8532569273175686e-05, "loss": 0.0513, "step": 2123 }, { "epoch": 0.2302688638334779, "grad_norm": 2.2716434001922607, "learning_rate": 1.8531843899608298e-05, "loss": 0.0397, "step": 2124 }, { "epoch": 0.23037727666955768, "grad_norm": 5.942279815673828, "learning_rate": 1.8531118526040913e-05, "loss": 0.0637, "step": 2125 }, { "epoch": 0.23048568950563747, "grad_norm": 2.8923628330230713, "learning_rate": 1.8530393152473526e-05, "loss": 0.0282, "step": 2126 }, { "epoch": 0.23059410234171726, "grad_norm": 15.509276390075684, "learning_rate": 1.8529667778906138e-05, "loss": 0.0282, "step": 2127 }, { "epoch": 0.23070251517779705, "grad_norm": 4.3234686851501465, "learning_rate": 1.852894240533875e-05, "loss": 0.0875, "step": 2128 }, { "epoch": 0.23081092801387684, "grad_norm": 2.610443353652954, "learning_rate": 1.8528217031771362e-05, "loss": 0.0143, "step": 2129 }, { "epoch": 0.23091934084995663, "grad_norm": 4.49663782119751, "learning_rate": 1.8527491658203977e-05, "loss": 0.0492, "step": 2130 }, { "epoch": 0.23102775368603642, "grad_norm": 5.111809730529785, "learning_rate": 1.852676628463659e-05, "loss": 0.0278, "step": 2131 }, { "epoch": 0.23113616652211622, "grad_norm": 5.723284721374512, "learning_rate": 1.8526040911069202e-05, "loss": 0.0687, "step": 2132 }, { "epoch": 0.231244579358196, "grad_norm": 2.4856889247894287, "learning_rate": 1.8525315537501817e-05, "loss": 0.0223, "step": 2133 }, { "epoch": 0.2313529921942758, "grad_norm": 2.7637875080108643, "learning_rate": 1.852459016393443e-05, "loss": 0.0282, "step": 2134 }, { "epoch": 0.2314614050303556, "grad_norm": 3.0925018787384033, "learning_rate": 1.852386479036704e-05, "loss": 0.0168, "step": 2135 }, { "epoch": 0.23156981786643538, "grad_norm": 6.77753210067749, "learning_rate": 1.8523139416799654e-05, "loss": 0.0752, "step": 2136 }, { "epoch": 0.23167823070251517, "grad_norm": 1.9882392883300781, "learning_rate": 1.8522414043232266e-05, "loss": 0.0328, "step": 2137 }, { "epoch": 0.23178664353859496, "grad_norm": 4.42117977142334, "learning_rate": 1.8521688669664878e-05, "loss": 0.0348, "step": 2138 }, { "epoch": 0.23189505637467475, "grad_norm": 4.879851341247559, "learning_rate": 1.852096329609749e-05, "loss": 0.0674, "step": 2139 }, { "epoch": 0.23200346921075454, "grad_norm": 1.3506556749343872, "learning_rate": 1.8520237922530106e-05, "loss": 0.0228, "step": 2140 }, { "epoch": 0.23211188204683433, "grad_norm": 3.3138418197631836, "learning_rate": 1.8519512548962718e-05, "loss": 0.065, "step": 2141 }, { "epoch": 0.23222029488291412, "grad_norm": 2.1845178604125977, "learning_rate": 1.851878717539533e-05, "loss": 0.0443, "step": 2142 }, { "epoch": 0.23232870771899392, "grad_norm": 1.877223014831543, "learning_rate": 1.8518061801827942e-05, "loss": 0.0439, "step": 2143 }, { "epoch": 0.23243712055507373, "grad_norm": 1.5509802103042603, "learning_rate": 1.8517336428260554e-05, "loss": 0.0271, "step": 2144 }, { "epoch": 0.23254553339115352, "grad_norm": 1.949537754058838, "learning_rate": 1.8516611054693166e-05, "loss": 0.0358, "step": 2145 }, { "epoch": 0.23265394622723332, "grad_norm": 2.9190587997436523, "learning_rate": 1.8515885681125782e-05, "loss": 0.0453, "step": 2146 }, { "epoch": 0.2327623590633131, "grad_norm": 6.025536060333252, "learning_rate": 1.8515160307558394e-05, "loss": 0.062, "step": 2147 }, { "epoch": 0.2328707718993929, "grad_norm": 0.8982529640197754, "learning_rate": 1.8514434933991006e-05, "loss": 0.0093, "step": 2148 }, { "epoch": 0.2329791847354727, "grad_norm": 3.6219534873962402, "learning_rate": 1.8513709560423622e-05, "loss": 0.0577, "step": 2149 }, { "epoch": 0.23308759757155248, "grad_norm": 2.7941670417785645, "learning_rate": 1.8512984186856234e-05, "loss": 0.0256, "step": 2150 }, { "epoch": 0.23319601040763227, "grad_norm": 2.1307103633880615, "learning_rate": 1.8512258813288846e-05, "loss": 0.0404, "step": 2151 }, { "epoch": 0.23330442324371206, "grad_norm": 0.34133240580558777, "learning_rate": 1.8511533439721458e-05, "loss": 0.0092, "step": 2152 }, { "epoch": 0.23341283607979185, "grad_norm": 0.4956394135951996, "learning_rate": 1.851080806615407e-05, "loss": 0.0049, "step": 2153 }, { "epoch": 0.23352124891587164, "grad_norm": 4.0048747062683105, "learning_rate": 1.8510082692586682e-05, "loss": 0.0362, "step": 2154 }, { "epoch": 0.23362966175195143, "grad_norm": 6.632531642913818, "learning_rate": 1.8509357319019298e-05, "loss": 0.0686, "step": 2155 }, { "epoch": 0.23373807458803122, "grad_norm": 0.9986993074417114, "learning_rate": 1.850863194545191e-05, "loss": 0.013, "step": 2156 }, { "epoch": 0.23384648742411102, "grad_norm": 3.8015921115875244, "learning_rate": 1.8507906571884522e-05, "loss": 0.0258, "step": 2157 }, { "epoch": 0.2339549002601908, "grad_norm": 5.205173492431641, "learning_rate": 1.8507181198317134e-05, "loss": 0.035, "step": 2158 }, { "epoch": 0.2340633130962706, "grad_norm": 2.069166898727417, "learning_rate": 1.8506455824749747e-05, "loss": 0.0139, "step": 2159 }, { "epoch": 0.2341717259323504, "grad_norm": 4.969053268432617, "learning_rate": 1.850573045118236e-05, "loss": 0.0284, "step": 2160 }, { "epoch": 0.23428013876843018, "grad_norm": 4.030860900878906, "learning_rate": 1.850500507761497e-05, "loss": 0.0574, "step": 2161 }, { "epoch": 0.23438855160450997, "grad_norm": 12.875411033630371, "learning_rate": 1.8504279704047586e-05, "loss": 0.0802, "step": 2162 }, { "epoch": 0.23449696444058976, "grad_norm": 5.8605546951293945, "learning_rate": 1.85035543304802e-05, "loss": 0.0707, "step": 2163 }, { "epoch": 0.23460537727666955, "grad_norm": 4.8646368980407715, "learning_rate": 1.850282895691281e-05, "loss": 0.076, "step": 2164 }, { "epoch": 0.23471379011274934, "grad_norm": 8.501477241516113, "learning_rate": 1.8502103583345426e-05, "loss": 0.0556, "step": 2165 }, { "epoch": 0.23482220294882913, "grad_norm": 4.191046714782715, "learning_rate": 1.850137820977804e-05, "loss": 0.0629, "step": 2166 }, { "epoch": 0.23493061578490892, "grad_norm": 2.665210247039795, "learning_rate": 1.850065283621065e-05, "loss": 0.0259, "step": 2167 }, { "epoch": 0.23503902862098872, "grad_norm": 7.699797630310059, "learning_rate": 1.8499927462643263e-05, "loss": 0.097, "step": 2168 }, { "epoch": 0.2351474414570685, "grad_norm": 3.4575424194335938, "learning_rate": 1.8499202089075875e-05, "loss": 0.0648, "step": 2169 }, { "epoch": 0.2352558542931483, "grad_norm": 3.8955934047698975, "learning_rate": 1.849847671550849e-05, "loss": 0.0206, "step": 2170 }, { "epoch": 0.2353642671292281, "grad_norm": 3.307621955871582, "learning_rate": 1.8497751341941102e-05, "loss": 0.039, "step": 2171 }, { "epoch": 0.23547267996530788, "grad_norm": 6.6789164543151855, "learning_rate": 1.8497025968373715e-05, "loss": 0.059, "step": 2172 }, { "epoch": 0.2355810928013877, "grad_norm": 0.7097261548042297, "learning_rate": 1.8496300594806327e-05, "loss": 0.0254, "step": 2173 }, { "epoch": 0.2356895056374675, "grad_norm": 4.309938430786133, "learning_rate": 1.849557522123894e-05, "loss": 0.0346, "step": 2174 }, { "epoch": 0.23579791847354728, "grad_norm": 2.258169174194336, "learning_rate": 1.849484984767155e-05, "loss": 0.021, "step": 2175 }, { "epoch": 0.23590633130962707, "grad_norm": 2.3712358474731445, "learning_rate": 1.8494124474104163e-05, "loss": 0.0295, "step": 2176 }, { "epoch": 0.23601474414570686, "grad_norm": 5.480093002319336, "learning_rate": 1.849339910053678e-05, "loss": 0.0636, "step": 2177 }, { "epoch": 0.23612315698178665, "grad_norm": 4.472269535064697, "learning_rate": 1.849267372696939e-05, "loss": 0.0672, "step": 2178 }, { "epoch": 0.23623156981786644, "grad_norm": 4.345576286315918, "learning_rate": 1.8491948353402003e-05, "loss": 0.053, "step": 2179 }, { "epoch": 0.23633998265394623, "grad_norm": 3.567464828491211, "learning_rate": 1.8491222979834615e-05, "loss": 0.0642, "step": 2180 }, { "epoch": 0.23644839549002603, "grad_norm": 2.6532084941864014, "learning_rate": 1.8490497606267227e-05, "loss": 0.0507, "step": 2181 }, { "epoch": 0.23655680832610582, "grad_norm": 3.666071891784668, "learning_rate": 1.8489772232699843e-05, "loss": 0.0378, "step": 2182 }, { "epoch": 0.2366652211621856, "grad_norm": 1.9446992874145508, "learning_rate": 1.8489046859132455e-05, "loss": 0.0247, "step": 2183 }, { "epoch": 0.2367736339982654, "grad_norm": 2.431840181350708, "learning_rate": 1.8488321485565067e-05, "loss": 0.0362, "step": 2184 }, { "epoch": 0.2368820468343452, "grad_norm": 2.2068326473236084, "learning_rate": 1.8487596111997683e-05, "loss": 0.041, "step": 2185 }, { "epoch": 0.23699045967042498, "grad_norm": 1.8767603635787964, "learning_rate": 1.8486870738430295e-05, "loss": 0.0235, "step": 2186 }, { "epoch": 0.23709887250650477, "grad_norm": 2.081414222717285, "learning_rate": 1.8486145364862907e-05, "loss": 0.0686, "step": 2187 }, { "epoch": 0.23720728534258456, "grad_norm": 2.7908482551574707, "learning_rate": 1.848541999129552e-05, "loss": 0.0429, "step": 2188 }, { "epoch": 0.23731569817866435, "grad_norm": 1.9445381164550781, "learning_rate": 1.848469461772813e-05, "loss": 0.0248, "step": 2189 }, { "epoch": 0.23742411101474414, "grad_norm": 2.395372152328491, "learning_rate": 1.8483969244160743e-05, "loss": 0.0423, "step": 2190 }, { "epoch": 0.23753252385082393, "grad_norm": 5.361285209655762, "learning_rate": 1.8483243870593355e-05, "loss": 0.0685, "step": 2191 }, { "epoch": 0.23764093668690373, "grad_norm": 3.01960825920105, "learning_rate": 1.848251849702597e-05, "loss": 0.0419, "step": 2192 }, { "epoch": 0.23774934952298352, "grad_norm": 9.202164649963379, "learning_rate": 1.8481793123458583e-05, "loss": 0.0435, "step": 2193 }, { "epoch": 0.2378577623590633, "grad_norm": 2.9778575897216797, "learning_rate": 1.8481067749891195e-05, "loss": 0.0332, "step": 2194 }, { "epoch": 0.2379661751951431, "grad_norm": 3.9009039402008057, "learning_rate": 1.8480342376323807e-05, "loss": 0.0789, "step": 2195 }, { "epoch": 0.2380745880312229, "grad_norm": 4.695461273193359, "learning_rate": 1.847961700275642e-05, "loss": 0.0594, "step": 2196 }, { "epoch": 0.23818300086730268, "grad_norm": 1.1401792764663696, "learning_rate": 1.8478891629189032e-05, "loss": 0.0219, "step": 2197 }, { "epoch": 0.23829141370338247, "grad_norm": 4.923417568206787, "learning_rate": 1.8478166255621647e-05, "loss": 0.0283, "step": 2198 }, { "epoch": 0.23839982653946226, "grad_norm": 3.1239607334136963, "learning_rate": 1.847744088205426e-05, "loss": 0.0206, "step": 2199 }, { "epoch": 0.23850823937554205, "grad_norm": 3.807370185852051, "learning_rate": 1.847671550848687e-05, "loss": 0.0492, "step": 2200 }, { "epoch": 0.23861665221162184, "grad_norm": 7.616703033447266, "learning_rate": 1.8475990134919487e-05, "loss": 0.1128, "step": 2201 }, { "epoch": 0.23872506504770163, "grad_norm": 2.811962366104126, "learning_rate": 1.84752647613521e-05, "loss": 0.021, "step": 2202 }, { "epoch": 0.23883347788378145, "grad_norm": 2.5918426513671875, "learning_rate": 1.847453938778471e-05, "loss": 0.0252, "step": 2203 }, { "epoch": 0.23894189071986124, "grad_norm": 0.43371033668518066, "learning_rate": 1.8473814014217324e-05, "loss": 0.0037, "step": 2204 }, { "epoch": 0.23905030355594103, "grad_norm": 3.764815092086792, "learning_rate": 1.8473088640649936e-05, "loss": 0.055, "step": 2205 }, { "epoch": 0.23915871639202083, "grad_norm": 1.822660207748413, "learning_rate": 1.8472363267082548e-05, "loss": 0.0353, "step": 2206 }, { "epoch": 0.23926712922810062, "grad_norm": 6.946071147918701, "learning_rate": 1.8471637893515163e-05, "loss": 0.061, "step": 2207 }, { "epoch": 0.2393755420641804, "grad_norm": 3.511577606201172, "learning_rate": 1.8470912519947775e-05, "loss": 0.0257, "step": 2208 }, { "epoch": 0.2394839549002602, "grad_norm": 2.5702109336853027, "learning_rate": 1.8470187146380388e-05, "loss": 0.0089, "step": 2209 }, { "epoch": 0.23959236773634, "grad_norm": 1.3210867643356323, "learning_rate": 1.8469461772813e-05, "loss": 0.0132, "step": 2210 }, { "epoch": 0.23970078057241978, "grad_norm": 0.6997748613357544, "learning_rate": 1.8468736399245612e-05, "loss": 0.0063, "step": 2211 }, { "epoch": 0.23980919340849957, "grad_norm": 3.581684112548828, "learning_rate": 1.8468011025678224e-05, "loss": 0.0491, "step": 2212 }, { "epoch": 0.23991760624457936, "grad_norm": 1.4010196924209595, "learning_rate": 1.846728565211084e-05, "loss": 0.0108, "step": 2213 }, { "epoch": 0.24002601908065915, "grad_norm": 1.5381355285644531, "learning_rate": 1.8466560278543452e-05, "loss": 0.016, "step": 2214 }, { "epoch": 0.24013443191673894, "grad_norm": 4.634088039398193, "learning_rate": 1.8465834904976064e-05, "loss": 0.0328, "step": 2215 }, { "epoch": 0.24024284475281873, "grad_norm": 8.336968421936035, "learning_rate": 1.8465109531408676e-05, "loss": 0.0632, "step": 2216 }, { "epoch": 0.24035125758889853, "grad_norm": 3.773860216140747, "learning_rate": 1.846438415784129e-05, "loss": 0.0762, "step": 2217 }, { "epoch": 0.24045967042497832, "grad_norm": 2.861882448196411, "learning_rate": 1.8463658784273904e-05, "loss": 0.0442, "step": 2218 }, { "epoch": 0.2405680832610581, "grad_norm": 8.191266059875488, "learning_rate": 1.8462933410706516e-05, "loss": 0.0785, "step": 2219 }, { "epoch": 0.2406764960971379, "grad_norm": 2.4908738136291504, "learning_rate": 1.8462208037139128e-05, "loss": 0.0182, "step": 2220 }, { "epoch": 0.2407849089332177, "grad_norm": 6.417301654815674, "learning_rate": 1.846148266357174e-05, "loss": 0.0898, "step": 2221 }, { "epoch": 0.24089332176929748, "grad_norm": 1.6050708293914795, "learning_rate": 1.8460757290004356e-05, "loss": 0.014, "step": 2222 }, { "epoch": 0.24100173460537727, "grad_norm": 2.4688150882720947, "learning_rate": 1.8460031916436968e-05, "loss": 0.0472, "step": 2223 }, { "epoch": 0.24111014744145706, "grad_norm": 4.332587242126465, "learning_rate": 1.845930654286958e-05, "loss": 0.0789, "step": 2224 }, { "epoch": 0.24121856027753685, "grad_norm": 1.020340085029602, "learning_rate": 1.8458581169302192e-05, "loss": 0.0145, "step": 2225 }, { "epoch": 0.24132697311361664, "grad_norm": 2.936638355255127, "learning_rate": 1.8457855795734804e-05, "loss": 0.1288, "step": 2226 }, { "epoch": 0.24143538594969643, "grad_norm": 4.468433856964111, "learning_rate": 1.8457130422167416e-05, "loss": 0.0482, "step": 2227 }, { "epoch": 0.24154379878577623, "grad_norm": 4.49636173248291, "learning_rate": 1.845640504860003e-05, "loss": 0.02, "step": 2228 }, { "epoch": 0.24165221162185602, "grad_norm": 5.120630264282227, "learning_rate": 1.8455679675032644e-05, "loss": 0.0536, "step": 2229 }, { "epoch": 0.2417606244579358, "grad_norm": 1.3519819974899292, "learning_rate": 1.8454954301465256e-05, "loss": 0.0165, "step": 2230 }, { "epoch": 0.2418690372940156, "grad_norm": 0.41720065474510193, "learning_rate": 1.845422892789787e-05, "loss": 0.0032, "step": 2231 }, { "epoch": 0.2419774501300954, "grad_norm": 1.3706642389297485, "learning_rate": 1.845350355433048e-05, "loss": 0.0262, "step": 2232 }, { "epoch": 0.2420858629661752, "grad_norm": 1.8307055234909058, "learning_rate": 1.8452778180763093e-05, "loss": 0.0442, "step": 2233 }, { "epoch": 0.242194275802255, "grad_norm": 3.1364095211029053, "learning_rate": 1.8452052807195708e-05, "loss": 0.0254, "step": 2234 }, { "epoch": 0.2423026886383348, "grad_norm": 1.1426291465759277, "learning_rate": 1.845132743362832e-05, "loss": 0.0168, "step": 2235 }, { "epoch": 0.24241110147441458, "grad_norm": 2.2752156257629395, "learning_rate": 1.8450602060060932e-05, "loss": 0.0178, "step": 2236 }, { "epoch": 0.24251951431049437, "grad_norm": 7.5274739265441895, "learning_rate": 1.8449876686493548e-05, "loss": 0.1042, "step": 2237 }, { "epoch": 0.24262792714657416, "grad_norm": 1.8924051523208618, "learning_rate": 1.844915131292616e-05, "loss": 0.0618, "step": 2238 }, { "epoch": 0.24273633998265395, "grad_norm": 3.962908983230591, "learning_rate": 1.8448425939358772e-05, "loss": 0.0568, "step": 2239 }, { "epoch": 0.24284475281873374, "grad_norm": 20.33913803100586, "learning_rate": 1.8447700565791384e-05, "loss": 0.0547, "step": 2240 }, { "epoch": 0.24295316565481354, "grad_norm": 1.3390864133834839, "learning_rate": 1.8446975192223997e-05, "loss": 0.0321, "step": 2241 }, { "epoch": 0.24306157849089333, "grad_norm": 4.0159125328063965, "learning_rate": 1.844624981865661e-05, "loss": 0.0423, "step": 2242 }, { "epoch": 0.24316999132697312, "grad_norm": 2.579294204711914, "learning_rate": 1.844552444508922e-05, "loss": 0.0297, "step": 2243 }, { "epoch": 0.2432784041630529, "grad_norm": 2.3379597663879395, "learning_rate": 1.8444799071521836e-05, "loss": 0.056, "step": 2244 }, { "epoch": 0.2433868169991327, "grad_norm": 3.1433093547821045, "learning_rate": 1.844407369795445e-05, "loss": 0.0383, "step": 2245 }, { "epoch": 0.2434952298352125, "grad_norm": 2.378422975540161, "learning_rate": 1.844334832438706e-05, "loss": 0.0396, "step": 2246 }, { "epoch": 0.24360364267129228, "grad_norm": 3.8050975799560547, "learning_rate": 1.8442622950819673e-05, "loss": 0.0863, "step": 2247 }, { "epoch": 0.24371205550737207, "grad_norm": 6.812331676483154, "learning_rate": 1.8441897577252285e-05, "loss": 0.0857, "step": 2248 }, { "epoch": 0.24382046834345186, "grad_norm": 2.7643563747406006, "learning_rate": 1.8441172203684897e-05, "loss": 0.0524, "step": 2249 }, { "epoch": 0.24392888117953165, "grad_norm": 5.58067512512207, "learning_rate": 1.8440446830117513e-05, "loss": 0.0543, "step": 2250 }, { "epoch": 0.24403729401561144, "grad_norm": 2.939298391342163, "learning_rate": 1.8439721456550125e-05, "loss": 0.0376, "step": 2251 }, { "epoch": 0.24414570685169124, "grad_norm": 2.6815381050109863, "learning_rate": 1.8438996082982737e-05, "loss": 0.0257, "step": 2252 }, { "epoch": 0.24425411968777103, "grad_norm": 2.5759384632110596, "learning_rate": 1.8438270709415352e-05, "loss": 0.0464, "step": 2253 }, { "epoch": 0.24436253252385082, "grad_norm": 2.0780282020568848, "learning_rate": 1.8437545335847965e-05, "loss": 0.023, "step": 2254 }, { "epoch": 0.2444709453599306, "grad_norm": 9.019510269165039, "learning_rate": 1.8436819962280577e-05, "loss": 0.0777, "step": 2255 }, { "epoch": 0.2445793581960104, "grad_norm": 4.862043380737305, "learning_rate": 1.843609458871319e-05, "loss": 0.0618, "step": 2256 }, { "epoch": 0.2446877710320902, "grad_norm": 4.038535118103027, "learning_rate": 1.84353692151458e-05, "loss": 0.0477, "step": 2257 }, { "epoch": 0.24479618386816998, "grad_norm": 4.223560333251953, "learning_rate": 1.8434643841578413e-05, "loss": 0.0417, "step": 2258 }, { "epoch": 0.24490459670424977, "grad_norm": 1.9210426807403564, "learning_rate": 1.843391846801103e-05, "loss": 0.0251, "step": 2259 }, { "epoch": 0.24501300954032956, "grad_norm": 7.3482465744018555, "learning_rate": 1.843319309444364e-05, "loss": 0.0394, "step": 2260 }, { "epoch": 0.24512142237640935, "grad_norm": 4.802293300628662, "learning_rate": 1.8432467720876253e-05, "loss": 0.0509, "step": 2261 }, { "epoch": 0.24522983521248917, "grad_norm": 3.2744247913360596, "learning_rate": 1.8431742347308865e-05, "loss": 0.0614, "step": 2262 }, { "epoch": 0.24533824804856896, "grad_norm": 4.944892883300781, "learning_rate": 1.8431016973741477e-05, "loss": 0.0354, "step": 2263 }, { "epoch": 0.24544666088464875, "grad_norm": 2.9465088844299316, "learning_rate": 1.843029160017409e-05, "loss": 0.0266, "step": 2264 }, { "epoch": 0.24555507372072854, "grad_norm": 5.038067817687988, "learning_rate": 1.8429566226606705e-05, "loss": 0.0538, "step": 2265 }, { "epoch": 0.24566348655680834, "grad_norm": 2.833869457244873, "learning_rate": 1.8428840853039317e-05, "loss": 0.0426, "step": 2266 }, { "epoch": 0.24577189939288813, "grad_norm": 0.8886680603027344, "learning_rate": 1.842811547947193e-05, "loss": 0.014, "step": 2267 }, { "epoch": 0.24588031222896792, "grad_norm": 1.7035900354385376, "learning_rate": 1.842739010590454e-05, "loss": 0.0209, "step": 2268 }, { "epoch": 0.2459887250650477, "grad_norm": 4.781304836273193, "learning_rate": 1.8426664732337157e-05, "loss": 0.0503, "step": 2269 }, { "epoch": 0.2460971379011275, "grad_norm": 4.413846969604492, "learning_rate": 1.842593935876977e-05, "loss": 0.0699, "step": 2270 }, { "epoch": 0.2462055507372073, "grad_norm": 0.9114212989807129, "learning_rate": 1.842521398520238e-05, "loss": 0.0086, "step": 2271 }, { "epoch": 0.24631396357328708, "grad_norm": 3.829106330871582, "learning_rate": 1.8424488611634993e-05, "loss": 0.0556, "step": 2272 }, { "epoch": 0.24642237640936687, "grad_norm": 1.8095567226409912, "learning_rate": 1.8423763238067605e-05, "loss": 0.0399, "step": 2273 }, { "epoch": 0.24653078924544666, "grad_norm": 5.672595977783203, "learning_rate": 1.842303786450022e-05, "loss": 0.0801, "step": 2274 }, { "epoch": 0.24663920208152645, "grad_norm": 2.369130849838257, "learning_rate": 1.8422312490932833e-05, "loss": 0.0633, "step": 2275 }, { "epoch": 0.24674761491760624, "grad_norm": 3.3146238327026367, "learning_rate": 1.8421587117365445e-05, "loss": 0.0591, "step": 2276 }, { "epoch": 0.24685602775368604, "grad_norm": 10.371065139770508, "learning_rate": 1.8420861743798057e-05, "loss": 0.0577, "step": 2277 }, { "epoch": 0.24696444058976583, "grad_norm": 2.3117430210113525, "learning_rate": 1.842013637023067e-05, "loss": 0.0424, "step": 2278 }, { "epoch": 0.24707285342584562, "grad_norm": 4.230005741119385, "learning_rate": 1.8419410996663282e-05, "loss": 0.0175, "step": 2279 }, { "epoch": 0.2471812662619254, "grad_norm": 2.8676185607910156, "learning_rate": 1.8418685623095894e-05, "loss": 0.05, "step": 2280 }, { "epoch": 0.2472896790980052, "grad_norm": 3.0238378047943115, "learning_rate": 1.841796024952851e-05, "loss": 0.0528, "step": 2281 }, { "epoch": 0.247398091934085, "grad_norm": 2.4861834049224854, "learning_rate": 1.841723487596112e-05, "loss": 0.0404, "step": 2282 }, { "epoch": 0.24750650477016478, "grad_norm": 5.164502143859863, "learning_rate": 1.8416509502393734e-05, "loss": 0.0437, "step": 2283 }, { "epoch": 0.24761491760624457, "grad_norm": 2.1726624965667725, "learning_rate": 1.8415784128826346e-05, "loss": 0.0383, "step": 2284 }, { "epoch": 0.24772333044232436, "grad_norm": 1.692067265510559, "learning_rate": 1.8415058755258958e-05, "loss": 0.043, "step": 2285 }, { "epoch": 0.24783174327840415, "grad_norm": 2.6824514865875244, "learning_rate": 1.8414333381691573e-05, "loss": 0.0478, "step": 2286 }, { "epoch": 0.24794015611448394, "grad_norm": 2.7379255294799805, "learning_rate": 1.8413608008124186e-05, "loss": 0.0662, "step": 2287 }, { "epoch": 0.24804856895056374, "grad_norm": 0.9564348459243774, "learning_rate": 1.8412882634556798e-05, "loss": 0.0203, "step": 2288 }, { "epoch": 0.24815698178664353, "grad_norm": 10.468677520751953, "learning_rate": 1.8412157260989413e-05, "loss": 0.0578, "step": 2289 }, { "epoch": 0.24826539462272332, "grad_norm": 2.0435941219329834, "learning_rate": 1.8411431887422025e-05, "loss": 0.0425, "step": 2290 }, { "epoch": 0.2483738074588031, "grad_norm": 5.364309787750244, "learning_rate": 1.8410706513854638e-05, "loss": 0.0829, "step": 2291 }, { "epoch": 0.24848222029488293, "grad_norm": 4.966287136077881, "learning_rate": 1.840998114028725e-05, "loss": 0.0679, "step": 2292 }, { "epoch": 0.24859063313096272, "grad_norm": 1.68585205078125, "learning_rate": 1.8409255766719862e-05, "loss": 0.0384, "step": 2293 }, { "epoch": 0.2486990459670425, "grad_norm": 10.64899730682373, "learning_rate": 1.8408530393152474e-05, "loss": 0.1024, "step": 2294 }, { "epoch": 0.2488074588031223, "grad_norm": 3.3906431198120117, "learning_rate": 1.8407805019585086e-05, "loss": 0.0211, "step": 2295 }, { "epoch": 0.2489158716392021, "grad_norm": 5.83657169342041, "learning_rate": 1.8407079646017702e-05, "loss": 0.0983, "step": 2296 }, { "epoch": 0.24902428447528188, "grad_norm": 3.7445666790008545, "learning_rate": 1.8406354272450314e-05, "loss": 0.0893, "step": 2297 }, { "epoch": 0.24913269731136167, "grad_norm": 3.754603147506714, "learning_rate": 1.8405628898882926e-05, "loss": 0.05, "step": 2298 }, { "epoch": 0.24924111014744146, "grad_norm": 5.255877494812012, "learning_rate": 1.8404903525315538e-05, "loss": 0.0542, "step": 2299 }, { "epoch": 0.24934952298352125, "grad_norm": 4.524463653564453, "learning_rate": 1.840417815174815e-05, "loss": 0.089, "step": 2300 }, { "epoch": 0.24945793581960105, "grad_norm": 2.8242650032043457, "learning_rate": 1.8403452778180762e-05, "loss": 0.0461, "step": 2301 }, { "epoch": 0.24956634865568084, "grad_norm": 2.262608051300049, "learning_rate": 1.8402727404613378e-05, "loss": 0.031, "step": 2302 }, { "epoch": 0.24967476149176063, "grad_norm": 1.6309547424316406, "learning_rate": 1.840200203104599e-05, "loss": 0.0292, "step": 2303 }, { "epoch": 0.24978317432784042, "grad_norm": 3.0528085231781006, "learning_rate": 1.8401276657478602e-05, "loss": 0.0256, "step": 2304 }, { "epoch": 0.2498915871639202, "grad_norm": 3.042881727218628, "learning_rate": 1.8400551283911218e-05, "loss": 0.036, "step": 2305 }, { "epoch": 0.25, "grad_norm": 1.1797157526016235, "learning_rate": 1.839982591034383e-05, "loss": 0.0293, "step": 2306 }, { "epoch": 0.2501084128360798, "grad_norm": 1.3492999076843262, "learning_rate": 1.8399100536776442e-05, "loss": 0.0312, "step": 2307 }, { "epoch": 0.2502168256721596, "grad_norm": 3.8632774353027344, "learning_rate": 1.8398375163209054e-05, "loss": 0.0668, "step": 2308 }, { "epoch": 0.2503252385082394, "grad_norm": 3.542553424835205, "learning_rate": 1.8397649789641666e-05, "loss": 0.046, "step": 2309 }, { "epoch": 0.25043365134431916, "grad_norm": 4.544906139373779, "learning_rate": 1.839692441607428e-05, "loss": 0.0533, "step": 2310 }, { "epoch": 0.250542064180399, "grad_norm": 0.9979420900344849, "learning_rate": 1.8396199042506894e-05, "loss": 0.0291, "step": 2311 }, { "epoch": 0.25065047701647875, "grad_norm": 2.0207762718200684, "learning_rate": 1.8395473668939506e-05, "loss": 0.0507, "step": 2312 }, { "epoch": 0.25075888985255856, "grad_norm": 2.088123321533203, "learning_rate": 1.8394748295372118e-05, "loss": 0.0618, "step": 2313 }, { "epoch": 0.2508673026886383, "grad_norm": 0.7794561982154846, "learning_rate": 1.839402292180473e-05, "loss": 0.0335, "step": 2314 }, { "epoch": 0.25097571552471815, "grad_norm": 3.7435033321380615, "learning_rate": 1.8393297548237343e-05, "loss": 0.0707, "step": 2315 }, { "epoch": 0.2510841283607979, "grad_norm": 1.1078996658325195, "learning_rate": 1.8392572174669955e-05, "loss": 0.0308, "step": 2316 }, { "epoch": 0.2511925411968777, "grad_norm": 4.101029396057129, "learning_rate": 1.839184680110257e-05, "loss": 0.0322, "step": 2317 }, { "epoch": 0.2513009540329575, "grad_norm": 2.431483507156372, "learning_rate": 1.8391121427535182e-05, "loss": 0.0437, "step": 2318 }, { "epoch": 0.2514093668690373, "grad_norm": 3.6937646865844727, "learning_rate": 1.8390396053967795e-05, "loss": 0.0624, "step": 2319 }, { "epoch": 0.2515177797051171, "grad_norm": 3.37628436088562, "learning_rate": 1.838967068040041e-05, "loss": 0.0729, "step": 2320 }, { "epoch": 0.2516261925411969, "grad_norm": 4.695820331573486, "learning_rate": 1.8388945306833022e-05, "loss": 0.043, "step": 2321 }, { "epoch": 0.25173460537727665, "grad_norm": 1.7173906564712524, "learning_rate": 1.8388219933265634e-05, "loss": 0.0487, "step": 2322 }, { "epoch": 0.2518430182133565, "grad_norm": 2.8454768657684326, "learning_rate": 1.8387494559698247e-05, "loss": 0.0343, "step": 2323 }, { "epoch": 0.25195143104943624, "grad_norm": 5.811519145965576, "learning_rate": 1.838676918613086e-05, "loss": 0.0295, "step": 2324 }, { "epoch": 0.25205984388551606, "grad_norm": 2.7988710403442383, "learning_rate": 1.838604381256347e-05, "loss": 0.0407, "step": 2325 }, { "epoch": 0.2521682567215958, "grad_norm": 1.524734377861023, "learning_rate": 1.8385318438996086e-05, "loss": 0.0256, "step": 2326 }, { "epoch": 0.25227666955767564, "grad_norm": 7.2636590003967285, "learning_rate": 1.83845930654287e-05, "loss": 0.0534, "step": 2327 }, { "epoch": 0.2523850823937554, "grad_norm": 3.3826115131378174, "learning_rate": 1.838386769186131e-05, "loss": 0.0418, "step": 2328 }, { "epoch": 0.2524934952298352, "grad_norm": 3.035726547241211, "learning_rate": 1.8383142318293923e-05, "loss": 0.066, "step": 2329 }, { "epoch": 0.252601908065915, "grad_norm": 1.8834582567214966, "learning_rate": 1.8382416944726535e-05, "loss": 0.0122, "step": 2330 }, { "epoch": 0.2527103209019948, "grad_norm": 3.2063543796539307, "learning_rate": 1.8381691571159147e-05, "loss": 0.1133, "step": 2331 }, { "epoch": 0.25281873373807456, "grad_norm": 2.373047351837158, "learning_rate": 1.838096619759176e-05, "loss": 0.0373, "step": 2332 }, { "epoch": 0.2529271465741544, "grad_norm": 1.839297890663147, "learning_rate": 1.8380240824024375e-05, "loss": 0.0354, "step": 2333 }, { "epoch": 0.25303555941023415, "grad_norm": 3.2862372398376465, "learning_rate": 1.8379515450456987e-05, "loss": 0.0696, "step": 2334 }, { "epoch": 0.25314397224631396, "grad_norm": 4.422262191772461, "learning_rate": 1.83787900768896e-05, "loss": 0.0451, "step": 2335 }, { "epoch": 0.2532523850823938, "grad_norm": 3.098156690597534, "learning_rate": 1.837806470332221e-05, "loss": 0.0319, "step": 2336 }, { "epoch": 0.25336079791847355, "grad_norm": 2.5713388919830322, "learning_rate": 1.8377339329754823e-05, "loss": 0.0353, "step": 2337 }, { "epoch": 0.25346921075455336, "grad_norm": 3.0884625911712646, "learning_rate": 1.837661395618744e-05, "loss": 0.0484, "step": 2338 }, { "epoch": 0.2535776235906331, "grad_norm": 2.555636167526245, "learning_rate": 1.837588858262005e-05, "loss": 0.0597, "step": 2339 }, { "epoch": 0.25368603642671295, "grad_norm": 4.498289108276367, "learning_rate": 1.8375163209052663e-05, "loss": 0.0636, "step": 2340 }, { "epoch": 0.2537944492627927, "grad_norm": 4.742649555206299, "learning_rate": 1.837443783548528e-05, "loss": 0.0682, "step": 2341 }, { "epoch": 0.25390286209887253, "grad_norm": 2.3994905948638916, "learning_rate": 1.837371246191789e-05, "loss": 0.0458, "step": 2342 }, { "epoch": 0.2540112749349523, "grad_norm": 1.1088917255401611, "learning_rate": 1.8372987088350503e-05, "loss": 0.0231, "step": 2343 }, { "epoch": 0.2541196877710321, "grad_norm": 3.1589114665985107, "learning_rate": 1.8372261714783115e-05, "loss": 0.0592, "step": 2344 }, { "epoch": 0.2542281006071119, "grad_norm": 2.282548189163208, "learning_rate": 1.8371536341215727e-05, "loss": 0.0614, "step": 2345 }, { "epoch": 0.2543365134431917, "grad_norm": 3.3537304401397705, "learning_rate": 1.837081096764834e-05, "loss": 0.0544, "step": 2346 }, { "epoch": 0.25444492627927146, "grad_norm": 3.070204973220825, "learning_rate": 1.837008559408095e-05, "loss": 0.0365, "step": 2347 }, { "epoch": 0.2545533391153513, "grad_norm": 3.879534959793091, "learning_rate": 1.8369360220513567e-05, "loss": 0.061, "step": 2348 }, { "epoch": 0.25466175195143104, "grad_norm": 1.621515154838562, "learning_rate": 1.836863484694618e-05, "loss": 0.0635, "step": 2349 }, { "epoch": 0.25477016478751086, "grad_norm": 1.4420984983444214, "learning_rate": 1.836790947337879e-05, "loss": 0.0408, "step": 2350 }, { "epoch": 0.2548785776235906, "grad_norm": 1.4691622257232666, "learning_rate": 1.8367184099811403e-05, "loss": 0.0392, "step": 2351 }, { "epoch": 0.25498699045967044, "grad_norm": 1.426093578338623, "learning_rate": 1.8366458726244016e-05, "loss": 0.0387, "step": 2352 }, { "epoch": 0.2550954032957502, "grad_norm": 3.7635433673858643, "learning_rate": 1.8365733352676628e-05, "loss": 0.0378, "step": 2353 }, { "epoch": 0.25520381613183, "grad_norm": 2.432737112045288, "learning_rate": 1.8365007979109243e-05, "loss": 0.052, "step": 2354 }, { "epoch": 0.2553122289679098, "grad_norm": 3.387511730194092, "learning_rate": 1.8364282605541855e-05, "loss": 0.062, "step": 2355 }, { "epoch": 0.2554206418039896, "grad_norm": 2.5584757328033447, "learning_rate": 1.836355723197447e-05, "loss": 0.0598, "step": 2356 }, { "epoch": 0.25552905464006936, "grad_norm": 4.168707370758057, "learning_rate": 1.8362831858407083e-05, "loss": 0.0733, "step": 2357 }, { "epoch": 0.2556374674761492, "grad_norm": 2.008765459060669, "learning_rate": 1.8362106484839695e-05, "loss": 0.0404, "step": 2358 }, { "epoch": 0.25574588031222895, "grad_norm": 5.7227396965026855, "learning_rate": 1.8361381111272307e-05, "loss": 0.0753, "step": 2359 }, { "epoch": 0.25585429314830876, "grad_norm": 3.836402416229248, "learning_rate": 1.836065573770492e-05, "loss": 0.0731, "step": 2360 }, { "epoch": 0.2559627059843885, "grad_norm": 1.5481791496276855, "learning_rate": 1.835993036413753e-05, "loss": 0.0163, "step": 2361 }, { "epoch": 0.25607111882046835, "grad_norm": 2.0579402446746826, "learning_rate": 1.8359204990570144e-05, "loss": 0.0345, "step": 2362 }, { "epoch": 0.2561795316565481, "grad_norm": 2.3073837757110596, "learning_rate": 1.835847961700276e-05, "loss": 0.0691, "step": 2363 }, { "epoch": 0.25628794449262793, "grad_norm": 7.910869121551514, "learning_rate": 1.835775424343537e-05, "loss": 0.0545, "step": 2364 }, { "epoch": 0.2563963573287077, "grad_norm": 2.6892709732055664, "learning_rate": 1.8357028869867984e-05, "loss": 0.0218, "step": 2365 }, { "epoch": 0.2565047701647875, "grad_norm": 1.3717697858810425, "learning_rate": 1.8356303496300596e-05, "loss": 0.0355, "step": 2366 }, { "epoch": 0.25661318300086733, "grad_norm": 1.9145220518112183, "learning_rate": 1.8355578122733208e-05, "loss": 0.0375, "step": 2367 }, { "epoch": 0.2567215958369471, "grad_norm": 3.866741180419922, "learning_rate": 1.835485274916582e-05, "loss": 0.0477, "step": 2368 }, { "epoch": 0.2568300086730269, "grad_norm": 2.934062957763672, "learning_rate": 1.8354127375598436e-05, "loss": 0.0917, "step": 2369 }, { "epoch": 0.2569384215091067, "grad_norm": 1.4151322841644287, "learning_rate": 1.8353402002031048e-05, "loss": 0.018, "step": 2370 }, { "epoch": 0.2570468343451865, "grad_norm": 6.483524322509766, "learning_rate": 1.835267662846366e-05, "loss": 0.032, "step": 2371 }, { "epoch": 0.25715524718126626, "grad_norm": 1.2906163930892944, "learning_rate": 1.8351951254896275e-05, "loss": 0.0401, "step": 2372 }, { "epoch": 0.2572636600173461, "grad_norm": 6.637969970703125, "learning_rate": 1.8351225881328888e-05, "loss": 0.0532, "step": 2373 }, { "epoch": 0.25737207285342584, "grad_norm": 2.9231159687042236, "learning_rate": 1.83505005077615e-05, "loss": 0.0227, "step": 2374 }, { "epoch": 0.25748048568950566, "grad_norm": 1.2646446228027344, "learning_rate": 1.8349775134194112e-05, "loss": 0.031, "step": 2375 }, { "epoch": 0.2575888985255854, "grad_norm": 5.919562816619873, "learning_rate": 1.8349049760626724e-05, "loss": 0.0822, "step": 2376 }, { "epoch": 0.25769731136166524, "grad_norm": 4.007338523864746, "learning_rate": 1.8348324387059336e-05, "loss": 0.0445, "step": 2377 }, { "epoch": 0.257805724197745, "grad_norm": 3.2474918365478516, "learning_rate": 1.834759901349195e-05, "loss": 0.025, "step": 2378 }, { "epoch": 0.2579141370338248, "grad_norm": 1.7985609769821167, "learning_rate": 1.8346873639924564e-05, "loss": 0.0343, "step": 2379 }, { "epoch": 0.2580225498699046, "grad_norm": 3.305677652359009, "learning_rate": 1.8346148266357176e-05, "loss": 0.0287, "step": 2380 }, { "epoch": 0.2581309627059844, "grad_norm": 2.3188562393188477, "learning_rate": 1.8345422892789788e-05, "loss": 0.0107, "step": 2381 }, { "epoch": 0.25823937554206416, "grad_norm": 4.205301761627197, "learning_rate": 1.83446975192224e-05, "loss": 0.055, "step": 2382 }, { "epoch": 0.258347788378144, "grad_norm": 1.8232005834579468, "learning_rate": 1.8343972145655012e-05, "loss": 0.0295, "step": 2383 }, { "epoch": 0.25845620121422375, "grad_norm": 2.0344600677490234, "learning_rate": 1.8343246772087625e-05, "loss": 0.0257, "step": 2384 }, { "epoch": 0.25856461405030357, "grad_norm": 1.8494935035705566, "learning_rate": 1.834252139852024e-05, "loss": 0.0217, "step": 2385 }, { "epoch": 0.25867302688638333, "grad_norm": 3.3549892902374268, "learning_rate": 1.8341796024952852e-05, "loss": 0.0533, "step": 2386 }, { "epoch": 0.25878143972246315, "grad_norm": 2.1669094562530518, "learning_rate": 1.8341070651385464e-05, "loss": 0.04, "step": 2387 }, { "epoch": 0.2588898525585429, "grad_norm": 3.8898379802703857, "learning_rate": 1.8340345277818076e-05, "loss": 0.0342, "step": 2388 }, { "epoch": 0.25899826539462273, "grad_norm": 3.5407073497772217, "learning_rate": 1.833961990425069e-05, "loss": 0.0378, "step": 2389 }, { "epoch": 0.2591066782307025, "grad_norm": 1.555408000946045, "learning_rate": 1.8338894530683304e-05, "loss": 0.0156, "step": 2390 }, { "epoch": 0.2592150910667823, "grad_norm": 2.1355462074279785, "learning_rate": 1.8338169157115916e-05, "loss": 0.0426, "step": 2391 }, { "epoch": 0.2593235039028621, "grad_norm": 0.7171206474304199, "learning_rate": 1.833744378354853e-05, "loss": 0.027, "step": 2392 }, { "epoch": 0.2594319167389419, "grad_norm": 0.9187630414962769, "learning_rate": 1.8336718409981144e-05, "loss": 0.0434, "step": 2393 }, { "epoch": 0.25954032957502166, "grad_norm": 1.4188910722732544, "learning_rate": 1.8335993036413756e-05, "loss": 0.0229, "step": 2394 }, { "epoch": 0.2596487424111015, "grad_norm": 1.5499961376190186, "learning_rate": 1.8335267662846368e-05, "loss": 0.0232, "step": 2395 }, { "epoch": 0.2597571552471813, "grad_norm": 2.5419812202453613, "learning_rate": 1.833454228927898e-05, "loss": 0.0518, "step": 2396 }, { "epoch": 0.25986556808326106, "grad_norm": 0.542569637298584, "learning_rate": 1.8333816915711593e-05, "loss": 0.0043, "step": 2397 }, { "epoch": 0.2599739809193409, "grad_norm": 2.8610775470733643, "learning_rate": 1.8333091542144205e-05, "loss": 0.0201, "step": 2398 }, { "epoch": 0.26008239375542064, "grad_norm": 9.326231002807617, "learning_rate": 1.8332366168576817e-05, "loss": 0.0524, "step": 2399 }, { "epoch": 0.26019080659150046, "grad_norm": 2.9170777797698975, "learning_rate": 1.8331640795009432e-05, "loss": 0.0443, "step": 2400 }, { "epoch": 0.2602992194275802, "grad_norm": 2.5085256099700928, "learning_rate": 1.8330915421442045e-05, "loss": 0.0456, "step": 2401 }, { "epoch": 0.26040763226366004, "grad_norm": 16.359115600585938, "learning_rate": 1.8330190047874657e-05, "loss": 0.1106, "step": 2402 }, { "epoch": 0.2605160450997398, "grad_norm": 6.922853469848633, "learning_rate": 1.832946467430727e-05, "loss": 0.0575, "step": 2403 }, { "epoch": 0.2606244579358196, "grad_norm": 4.906332492828369, "learning_rate": 1.832873930073988e-05, "loss": 0.0566, "step": 2404 }, { "epoch": 0.2607328707718994, "grad_norm": 16.136388778686523, "learning_rate": 1.8328013927172493e-05, "loss": 0.0352, "step": 2405 }, { "epoch": 0.2608412836079792, "grad_norm": 1.33441162109375, "learning_rate": 1.832728855360511e-05, "loss": 0.0307, "step": 2406 }, { "epoch": 0.26094969644405897, "grad_norm": 5.6199846267700195, "learning_rate": 1.832656318003772e-05, "loss": 0.1411, "step": 2407 }, { "epoch": 0.2610581092801388, "grad_norm": 1.8459511995315552, "learning_rate": 1.8325837806470336e-05, "loss": 0.0091, "step": 2408 }, { "epoch": 0.26116652211621855, "grad_norm": 1.2775143384933472, "learning_rate": 1.832511243290295e-05, "loss": 0.0212, "step": 2409 }, { "epoch": 0.26127493495229837, "grad_norm": 2.3041954040527344, "learning_rate": 1.832438705933556e-05, "loss": 0.0365, "step": 2410 }, { "epoch": 0.26138334778837813, "grad_norm": 3.0516059398651123, "learning_rate": 1.8323661685768173e-05, "loss": 0.0247, "step": 2411 }, { "epoch": 0.26149176062445795, "grad_norm": 3.401479721069336, "learning_rate": 1.8322936312200785e-05, "loss": 0.0399, "step": 2412 }, { "epoch": 0.2616001734605377, "grad_norm": 3.513475179672241, "learning_rate": 1.8322210938633397e-05, "loss": 0.0685, "step": 2413 }, { "epoch": 0.26170858629661753, "grad_norm": 2.571826696395874, "learning_rate": 1.832148556506601e-05, "loss": 0.0303, "step": 2414 }, { "epoch": 0.2618169991326973, "grad_norm": 3.5587477684020996, "learning_rate": 1.8320760191498625e-05, "loss": 0.0506, "step": 2415 }, { "epoch": 0.2619254119687771, "grad_norm": 1.0461455583572388, "learning_rate": 1.8320034817931237e-05, "loss": 0.0241, "step": 2416 }, { "epoch": 0.2620338248048569, "grad_norm": 0.3613942563533783, "learning_rate": 1.831930944436385e-05, "loss": 0.0034, "step": 2417 }, { "epoch": 0.2621422376409367, "grad_norm": 2.1905834674835205, "learning_rate": 1.831858407079646e-05, "loss": 0.0122, "step": 2418 }, { "epoch": 0.26225065047701646, "grad_norm": 3.2161448001861572, "learning_rate": 1.8317858697229073e-05, "loss": 0.0441, "step": 2419 }, { "epoch": 0.2623590633130963, "grad_norm": 4.60366153717041, "learning_rate": 1.8317133323661685e-05, "loss": 0.0507, "step": 2420 }, { "epoch": 0.26246747614917604, "grad_norm": 6.655795097351074, "learning_rate": 1.83164079500943e-05, "loss": 0.0562, "step": 2421 }, { "epoch": 0.26257588898525586, "grad_norm": 3.4821672439575195, "learning_rate": 1.8315682576526913e-05, "loss": 0.0345, "step": 2422 }, { "epoch": 0.2626843018213356, "grad_norm": 5.4365339279174805, "learning_rate": 1.8314957202959525e-05, "loss": 0.0425, "step": 2423 }, { "epoch": 0.26279271465741544, "grad_norm": 1.9004970788955688, "learning_rate": 1.831423182939214e-05, "loss": 0.0362, "step": 2424 }, { "epoch": 0.26290112749349526, "grad_norm": 1.7053521871566772, "learning_rate": 1.8313506455824753e-05, "loss": 0.0417, "step": 2425 }, { "epoch": 0.263009540329575, "grad_norm": 2.836522340774536, "learning_rate": 1.8312781082257365e-05, "loss": 0.0304, "step": 2426 }, { "epoch": 0.26311795316565484, "grad_norm": 1.1449517011642456, "learning_rate": 1.8312055708689977e-05, "loss": 0.0359, "step": 2427 }, { "epoch": 0.2632263660017346, "grad_norm": 1.542317509651184, "learning_rate": 1.831133033512259e-05, "loss": 0.0177, "step": 2428 }, { "epoch": 0.2633347788378144, "grad_norm": 4.473903656005859, "learning_rate": 1.83106049615552e-05, "loss": 0.0209, "step": 2429 }, { "epoch": 0.2634431916738942, "grad_norm": 7.4040327072143555, "learning_rate": 1.8309879587987817e-05, "loss": 0.052, "step": 2430 }, { "epoch": 0.263551604509974, "grad_norm": 2.5176947116851807, "learning_rate": 1.830915421442043e-05, "loss": 0.0415, "step": 2431 }, { "epoch": 0.26366001734605377, "grad_norm": 3.6385717391967773, "learning_rate": 1.830842884085304e-05, "loss": 0.0636, "step": 2432 }, { "epoch": 0.2637684301821336, "grad_norm": 1.4961594343185425, "learning_rate": 1.8307703467285653e-05, "loss": 0.0205, "step": 2433 }, { "epoch": 0.26387684301821335, "grad_norm": 1.5650835037231445, "learning_rate": 1.8306978093718266e-05, "loss": 0.0311, "step": 2434 }, { "epoch": 0.26398525585429317, "grad_norm": 8.7254056930542, "learning_rate": 1.8306252720150878e-05, "loss": 0.0788, "step": 2435 }, { "epoch": 0.26409366869037293, "grad_norm": 1.2893818616867065, "learning_rate": 1.830552734658349e-05, "loss": 0.0068, "step": 2436 }, { "epoch": 0.26420208152645275, "grad_norm": 2.202946424484253, "learning_rate": 1.8304801973016105e-05, "loss": 0.0479, "step": 2437 }, { "epoch": 0.2643104943625325, "grad_norm": 3.6398420333862305, "learning_rate": 1.8304076599448718e-05, "loss": 0.0623, "step": 2438 }, { "epoch": 0.26441890719861233, "grad_norm": 6.142880439758301, "learning_rate": 1.830335122588133e-05, "loss": 0.0755, "step": 2439 }, { "epoch": 0.2645273200346921, "grad_norm": 5.688620567321777, "learning_rate": 1.8302625852313942e-05, "loss": 0.0573, "step": 2440 }, { "epoch": 0.2646357328707719, "grad_norm": 3.8163435459136963, "learning_rate": 1.8301900478746557e-05, "loss": 0.0722, "step": 2441 }, { "epoch": 0.2647441457068517, "grad_norm": 2.3056399822235107, "learning_rate": 1.830117510517917e-05, "loss": 0.0552, "step": 2442 }, { "epoch": 0.2648525585429315, "grad_norm": 2.6759610176086426, "learning_rate": 1.830044973161178e-05, "loss": 0.0734, "step": 2443 }, { "epoch": 0.26496097137901126, "grad_norm": 2.0454978942871094, "learning_rate": 1.8299724358044394e-05, "loss": 0.0406, "step": 2444 }, { "epoch": 0.2650693842150911, "grad_norm": 2.4622788429260254, "learning_rate": 1.829899898447701e-05, "loss": 0.0585, "step": 2445 }, { "epoch": 0.26517779705117084, "grad_norm": 1.7303950786590576, "learning_rate": 1.829827361090962e-05, "loss": 0.0651, "step": 2446 }, { "epoch": 0.26528620988725066, "grad_norm": 3.4773805141448975, "learning_rate": 1.8297548237342234e-05, "loss": 0.0375, "step": 2447 }, { "epoch": 0.2653946227233304, "grad_norm": 1.263989806175232, "learning_rate": 1.8296822863774846e-05, "loss": 0.046, "step": 2448 }, { "epoch": 0.26550303555941024, "grad_norm": 4.824807643890381, "learning_rate": 1.8296097490207458e-05, "loss": 0.0697, "step": 2449 }, { "epoch": 0.26561144839549, "grad_norm": 0.8786716461181641, "learning_rate": 1.829537211664007e-05, "loss": 0.0185, "step": 2450 }, { "epoch": 0.2657198612315698, "grad_norm": 2.005157232284546, "learning_rate": 1.8294646743072682e-05, "loss": 0.0582, "step": 2451 }, { "epoch": 0.2658282740676496, "grad_norm": 2.501749277114868, "learning_rate": 1.8293921369505298e-05, "loss": 0.043, "step": 2452 }, { "epoch": 0.2659366869037294, "grad_norm": 1.286709189414978, "learning_rate": 1.829319599593791e-05, "loss": 0.0567, "step": 2453 }, { "epoch": 0.2660450997398092, "grad_norm": 1.0984160900115967, "learning_rate": 1.8292470622370522e-05, "loss": 0.0207, "step": 2454 }, { "epoch": 0.266153512575889, "grad_norm": 5.704442977905273, "learning_rate": 1.8291745248803134e-05, "loss": 0.0381, "step": 2455 }, { "epoch": 0.2662619254119688, "grad_norm": 2.204176425933838, "learning_rate": 1.8291019875235746e-05, "loss": 0.0382, "step": 2456 }, { "epoch": 0.26637033824804857, "grad_norm": 4.887734413146973, "learning_rate": 1.829029450166836e-05, "loss": 0.057, "step": 2457 }, { "epoch": 0.2664787510841284, "grad_norm": 3.528207778930664, "learning_rate": 1.8289569128100974e-05, "loss": 0.0266, "step": 2458 }, { "epoch": 0.26658716392020815, "grad_norm": 1.0690251588821411, "learning_rate": 1.8288843754533586e-05, "loss": 0.0273, "step": 2459 }, { "epoch": 0.26669557675628797, "grad_norm": 2.137162923812866, "learning_rate": 1.82881183809662e-05, "loss": 0.0434, "step": 2460 }, { "epoch": 0.26680398959236773, "grad_norm": 2.634971857070923, "learning_rate": 1.8287393007398814e-05, "loss": 0.0278, "step": 2461 }, { "epoch": 0.26691240242844755, "grad_norm": 1.2359765768051147, "learning_rate": 1.8286667633831426e-05, "loss": 0.0197, "step": 2462 }, { "epoch": 0.2670208152645273, "grad_norm": 1.2272785902023315, "learning_rate": 1.8285942260264038e-05, "loss": 0.0273, "step": 2463 }, { "epoch": 0.26712922810060713, "grad_norm": 2.92299485206604, "learning_rate": 1.828521688669665e-05, "loss": 0.0366, "step": 2464 }, { "epoch": 0.2672376409366869, "grad_norm": 0.7380319833755493, "learning_rate": 1.8284491513129262e-05, "loss": 0.0074, "step": 2465 }, { "epoch": 0.2673460537727667, "grad_norm": 2.596590042114258, "learning_rate": 1.8283766139561874e-05, "loss": 0.0444, "step": 2466 }, { "epoch": 0.2674544666088465, "grad_norm": 0.6617459058761597, "learning_rate": 1.828304076599449e-05, "loss": 0.004, "step": 2467 }, { "epoch": 0.2675628794449263, "grad_norm": 3.0627036094665527, "learning_rate": 1.8282315392427102e-05, "loss": 0.0596, "step": 2468 }, { "epoch": 0.26767129228100606, "grad_norm": 6.697943210601807, "learning_rate": 1.8281590018859714e-05, "loss": 0.1015, "step": 2469 }, { "epoch": 0.2677797051170859, "grad_norm": 1.3368632793426514, "learning_rate": 1.8280864645292326e-05, "loss": 0.0204, "step": 2470 }, { "epoch": 0.26788811795316564, "grad_norm": 1.3692125082015991, "learning_rate": 1.828013927172494e-05, "loss": 0.0146, "step": 2471 }, { "epoch": 0.26799653078924546, "grad_norm": 1.7533409595489502, "learning_rate": 1.827941389815755e-05, "loss": 0.0629, "step": 2472 }, { "epoch": 0.2681049436253252, "grad_norm": 5.038736820220947, "learning_rate": 1.8278688524590166e-05, "loss": 0.0403, "step": 2473 }, { "epoch": 0.26821335646140504, "grad_norm": 6.984476566314697, "learning_rate": 1.827796315102278e-05, "loss": 0.0503, "step": 2474 }, { "epoch": 0.2683217692974848, "grad_norm": 2.111729145050049, "learning_rate": 1.827723777745539e-05, "loss": 0.0478, "step": 2475 }, { "epoch": 0.2684301821335646, "grad_norm": 4.350743293762207, "learning_rate": 1.8276512403888006e-05, "loss": 0.0643, "step": 2476 }, { "epoch": 0.2685385949696444, "grad_norm": 10.254977226257324, "learning_rate": 1.8275787030320618e-05, "loss": 0.1313, "step": 2477 }, { "epoch": 0.2686470078057242, "grad_norm": 1.7169995307922363, "learning_rate": 1.827506165675323e-05, "loss": 0.0125, "step": 2478 }, { "epoch": 0.26875542064180397, "grad_norm": 3.1780664920806885, "learning_rate": 1.8274336283185843e-05, "loss": 0.0289, "step": 2479 }, { "epoch": 0.2688638334778838, "grad_norm": 3.365328550338745, "learning_rate": 1.8273610909618455e-05, "loss": 0.0181, "step": 2480 }, { "epoch": 0.26897224631396355, "grad_norm": 2.8291099071502686, "learning_rate": 1.8272885536051067e-05, "loss": 0.0236, "step": 2481 }, { "epoch": 0.26908065915004337, "grad_norm": 7.172647476196289, "learning_rate": 1.8272160162483682e-05, "loss": 0.0684, "step": 2482 }, { "epoch": 0.26918907198612313, "grad_norm": 1.391855001449585, "learning_rate": 1.8271434788916294e-05, "loss": 0.0125, "step": 2483 }, { "epoch": 0.26929748482220295, "grad_norm": 2.971787929534912, "learning_rate": 1.8270709415348907e-05, "loss": 0.0551, "step": 2484 }, { "epoch": 0.26940589765828277, "grad_norm": 4.990451812744141, "learning_rate": 1.826998404178152e-05, "loss": 0.1025, "step": 2485 }, { "epoch": 0.26951431049436253, "grad_norm": 4.688343048095703, "learning_rate": 1.826925866821413e-05, "loss": 0.0411, "step": 2486 }, { "epoch": 0.26962272333044235, "grad_norm": 4.059597969055176, "learning_rate": 1.8268533294646743e-05, "loss": 0.0557, "step": 2487 }, { "epoch": 0.2697311361665221, "grad_norm": 4.992016792297363, "learning_rate": 1.8267807921079355e-05, "loss": 0.0325, "step": 2488 }, { "epoch": 0.26983954900260193, "grad_norm": 7.315352439880371, "learning_rate": 1.826708254751197e-05, "loss": 0.0655, "step": 2489 }, { "epoch": 0.2699479618386817, "grad_norm": 4.491352558135986, "learning_rate": 1.8266357173944583e-05, "loss": 0.0624, "step": 2490 }, { "epoch": 0.2700563746747615, "grad_norm": 0.9897356629371643, "learning_rate": 1.8265631800377195e-05, "loss": 0.0119, "step": 2491 }, { "epoch": 0.2701647875108413, "grad_norm": 1.2462003231048584, "learning_rate": 1.8264906426809807e-05, "loss": 0.0123, "step": 2492 }, { "epoch": 0.2702732003469211, "grad_norm": 5.291073799133301, "learning_rate": 1.8264181053242423e-05, "loss": 0.1422, "step": 2493 }, { "epoch": 0.27038161318300086, "grad_norm": 6.524372100830078, "learning_rate": 1.8263455679675035e-05, "loss": 0.042, "step": 2494 }, { "epoch": 0.2704900260190807, "grad_norm": 2.582465410232544, "learning_rate": 1.8262730306107647e-05, "loss": 0.0197, "step": 2495 }, { "epoch": 0.27059843885516044, "grad_norm": 5.836318492889404, "learning_rate": 1.826200493254026e-05, "loss": 0.0435, "step": 2496 }, { "epoch": 0.27070685169124026, "grad_norm": 4.261153221130371, "learning_rate": 1.8261279558972875e-05, "loss": 0.0628, "step": 2497 }, { "epoch": 0.27081526452732, "grad_norm": 3.12109375, "learning_rate": 1.8260554185405487e-05, "loss": 0.1192, "step": 2498 }, { "epoch": 0.27092367736339984, "grad_norm": 2.638796329498291, "learning_rate": 1.82598288118381e-05, "loss": 0.0395, "step": 2499 }, { "epoch": 0.2710320901994796, "grad_norm": 4.003046035766602, "learning_rate": 1.825910343827071e-05, "loss": 0.0736, "step": 2500 }, { "epoch": 0.2711405030355594, "grad_norm": 0.9540491104125977, "learning_rate": 1.8258378064703323e-05, "loss": 0.031, "step": 2501 }, { "epoch": 0.2712489158716392, "grad_norm": 4.711818218231201, "learning_rate": 1.8257652691135935e-05, "loss": 0.059, "step": 2502 }, { "epoch": 0.271357328707719, "grad_norm": 2.9236786365509033, "learning_rate": 1.8256927317568548e-05, "loss": 0.1095, "step": 2503 }, { "epoch": 0.27146574154379877, "grad_norm": 2.4855005741119385, "learning_rate": 1.8256201944001163e-05, "loss": 0.0709, "step": 2504 }, { "epoch": 0.2715741543798786, "grad_norm": 1.567025899887085, "learning_rate": 1.8255476570433775e-05, "loss": 0.0362, "step": 2505 }, { "epoch": 0.27168256721595835, "grad_norm": 2.9261231422424316, "learning_rate": 1.8254751196866387e-05, "loss": 0.0356, "step": 2506 }, { "epoch": 0.27179098005203817, "grad_norm": 1.77474045753479, "learning_rate": 1.8254025823299e-05, "loss": 0.0406, "step": 2507 }, { "epoch": 0.27189939288811793, "grad_norm": 1.1989927291870117, "learning_rate": 1.825330044973161e-05, "loss": 0.0308, "step": 2508 }, { "epoch": 0.27200780572419775, "grad_norm": 3.4148130416870117, "learning_rate": 1.8252575076164224e-05, "loss": 0.0429, "step": 2509 }, { "epoch": 0.2721162185602775, "grad_norm": 3.6546711921691895, "learning_rate": 1.825184970259684e-05, "loss": 0.0285, "step": 2510 }, { "epoch": 0.27222463139635733, "grad_norm": 2.567038059234619, "learning_rate": 1.825112432902945e-05, "loss": 0.0608, "step": 2511 }, { "epoch": 0.2723330442324371, "grad_norm": 1.4901784658432007, "learning_rate": 1.8250398955462067e-05, "loss": 0.0254, "step": 2512 }, { "epoch": 0.2724414570685169, "grad_norm": 2.086541175842285, "learning_rate": 1.824967358189468e-05, "loss": 0.0668, "step": 2513 }, { "epoch": 0.27254986990459673, "grad_norm": 5.50830078125, "learning_rate": 1.824894820832729e-05, "loss": 0.0618, "step": 2514 }, { "epoch": 0.2726582827406765, "grad_norm": 2.968564987182617, "learning_rate": 1.8248222834759903e-05, "loss": 0.0794, "step": 2515 }, { "epoch": 0.2727666955767563, "grad_norm": 5.265379905700684, "learning_rate": 1.8247497461192516e-05, "loss": 0.0527, "step": 2516 }, { "epoch": 0.2728751084128361, "grad_norm": 1.896056890487671, "learning_rate": 1.8246772087625128e-05, "loss": 0.0233, "step": 2517 }, { "epoch": 0.2729835212489159, "grad_norm": 3.463764190673828, "learning_rate": 1.824604671405774e-05, "loss": 0.0331, "step": 2518 }, { "epoch": 0.27309193408499566, "grad_norm": 1.4299081563949585, "learning_rate": 1.8245321340490355e-05, "loss": 0.0338, "step": 2519 }, { "epoch": 0.2732003469210755, "grad_norm": 1.7988520860671997, "learning_rate": 1.8244595966922968e-05, "loss": 0.0346, "step": 2520 }, { "epoch": 0.27330875975715524, "grad_norm": 2.4569408893585205, "learning_rate": 1.824387059335558e-05, "loss": 0.0749, "step": 2521 }, { "epoch": 0.27341717259323506, "grad_norm": 2.228292942047119, "learning_rate": 1.8243145219788192e-05, "loss": 0.0552, "step": 2522 }, { "epoch": 0.2735255854293148, "grad_norm": 1.9885984659194946, "learning_rate": 1.8242419846220804e-05, "loss": 0.0371, "step": 2523 }, { "epoch": 0.27363399826539464, "grad_norm": 1.1939033269882202, "learning_rate": 1.8241694472653416e-05, "loss": 0.0126, "step": 2524 }, { "epoch": 0.2737424111014744, "grad_norm": 1.794991135597229, "learning_rate": 1.824096909908603e-05, "loss": 0.04, "step": 2525 }, { "epoch": 0.2738508239375542, "grad_norm": 2.647691011428833, "learning_rate": 1.8240243725518644e-05, "loss": 0.0471, "step": 2526 }, { "epoch": 0.273959236773634, "grad_norm": 3.337278366088867, "learning_rate": 1.8239518351951256e-05, "loss": 0.0449, "step": 2527 }, { "epoch": 0.2740676496097138, "grad_norm": 3.4754748344421387, "learning_rate": 1.823879297838387e-05, "loss": 0.0416, "step": 2528 }, { "epoch": 0.27417606244579357, "grad_norm": 0.5220271348953247, "learning_rate": 1.8238067604816484e-05, "loss": 0.0311, "step": 2529 }, { "epoch": 0.2742844752818734, "grad_norm": 1.9603462219238281, "learning_rate": 1.8237342231249096e-05, "loss": 0.0508, "step": 2530 }, { "epoch": 0.27439288811795315, "grad_norm": 1.1459137201309204, "learning_rate": 1.8236616857681708e-05, "loss": 0.0499, "step": 2531 }, { "epoch": 0.27450130095403297, "grad_norm": 3.7607452869415283, "learning_rate": 1.823589148411432e-05, "loss": 0.0775, "step": 2532 }, { "epoch": 0.27460971379011273, "grad_norm": 6.346685886383057, "learning_rate": 1.8235166110546932e-05, "loss": 0.0342, "step": 2533 }, { "epoch": 0.27471812662619255, "grad_norm": 3.5769340991973877, "learning_rate": 1.8234440736979548e-05, "loss": 0.0582, "step": 2534 }, { "epoch": 0.2748265394622723, "grad_norm": 3.8292078971862793, "learning_rate": 1.823371536341216e-05, "loss": 0.0732, "step": 2535 }, { "epoch": 0.27493495229835213, "grad_norm": 3.2939677238464355, "learning_rate": 1.8232989989844772e-05, "loss": 0.0227, "step": 2536 }, { "epoch": 0.2750433651344319, "grad_norm": 4.280768871307373, "learning_rate": 1.8232264616277384e-05, "loss": 0.025, "step": 2537 }, { "epoch": 0.2751517779705117, "grad_norm": 2.068814754486084, "learning_rate": 1.8231539242709996e-05, "loss": 0.0834, "step": 2538 }, { "epoch": 0.2752601908065915, "grad_norm": 6.2225494384765625, "learning_rate": 1.823081386914261e-05, "loss": 0.0811, "step": 2539 }, { "epoch": 0.2753686036426713, "grad_norm": 3.502819061279297, "learning_rate": 1.823008849557522e-05, "loss": 0.0743, "step": 2540 }, { "epoch": 0.27547701647875106, "grad_norm": 3.2917613983154297, "learning_rate": 1.8229363122007836e-05, "loss": 0.0749, "step": 2541 }, { "epoch": 0.2755854293148309, "grad_norm": 3.2656302452087402, "learning_rate": 1.8228637748440448e-05, "loss": 0.0639, "step": 2542 }, { "epoch": 0.2756938421509107, "grad_norm": 2.596672773361206, "learning_rate": 1.822791237487306e-05, "loss": 0.0628, "step": 2543 }, { "epoch": 0.27580225498699046, "grad_norm": 5.396711826324463, "learning_rate": 1.8227187001305673e-05, "loss": 0.0796, "step": 2544 }, { "epoch": 0.2759106678230703, "grad_norm": 1.8965930938720703, "learning_rate": 1.8226461627738288e-05, "loss": 0.0575, "step": 2545 }, { "epoch": 0.27601908065915004, "grad_norm": 3.6760034561157227, "learning_rate": 1.82257362541709e-05, "loss": 0.0584, "step": 2546 }, { "epoch": 0.27612749349522986, "grad_norm": 2.100379228591919, "learning_rate": 1.8225010880603512e-05, "loss": 0.0783, "step": 2547 }, { "epoch": 0.2762359063313096, "grad_norm": 1.7359803915023804, "learning_rate": 1.8224285507036124e-05, "loss": 0.0507, "step": 2548 }, { "epoch": 0.27634431916738944, "grad_norm": 3.980673313140869, "learning_rate": 1.822356013346874e-05, "loss": 0.063, "step": 2549 }, { "epoch": 0.2764527320034692, "grad_norm": 1.4522349834442139, "learning_rate": 1.8222834759901352e-05, "loss": 0.0523, "step": 2550 }, { "epoch": 0.276561144839549, "grad_norm": 1.368484377861023, "learning_rate": 1.8222109386333964e-05, "loss": 0.0604, "step": 2551 }, { "epoch": 0.2766695576756288, "grad_norm": 1.6213046312332153, "learning_rate": 1.8221384012766576e-05, "loss": 0.0445, "step": 2552 }, { "epoch": 0.2767779705117086, "grad_norm": 0.9601861238479614, "learning_rate": 1.822065863919919e-05, "loss": 0.0269, "step": 2553 }, { "epoch": 0.27688638334778837, "grad_norm": 3.070389747619629, "learning_rate": 1.82199332656318e-05, "loss": 0.0289, "step": 2554 }, { "epoch": 0.2769947961838682, "grad_norm": 2.769318103790283, "learning_rate": 1.8219207892064413e-05, "loss": 0.0322, "step": 2555 }, { "epoch": 0.27710320901994795, "grad_norm": 5.4861531257629395, "learning_rate": 1.821848251849703e-05, "loss": 0.0342, "step": 2556 }, { "epoch": 0.27721162185602777, "grad_norm": 1.9671149253845215, "learning_rate": 1.821775714492964e-05, "loss": 0.0903, "step": 2557 }, { "epoch": 0.27732003469210753, "grad_norm": 2.6022729873657227, "learning_rate": 1.8217031771362253e-05, "loss": 0.0493, "step": 2558 }, { "epoch": 0.27742844752818735, "grad_norm": 0.795097291469574, "learning_rate": 1.8216306397794865e-05, "loss": 0.0195, "step": 2559 }, { "epoch": 0.2775368603642671, "grad_norm": 2.094553232192993, "learning_rate": 1.8215581024227477e-05, "loss": 0.0476, "step": 2560 }, { "epoch": 0.27764527320034693, "grad_norm": 4.747906684875488, "learning_rate": 1.821485565066009e-05, "loss": 0.0512, "step": 2561 }, { "epoch": 0.2777536860364267, "grad_norm": 1.4602046012878418, "learning_rate": 1.8214130277092705e-05, "loss": 0.0217, "step": 2562 }, { "epoch": 0.2778620988725065, "grad_norm": 2.3103604316711426, "learning_rate": 1.8213404903525317e-05, "loss": 0.0427, "step": 2563 }, { "epoch": 0.2779705117085863, "grad_norm": 4.049470901489258, "learning_rate": 1.8212679529957932e-05, "loss": 0.0691, "step": 2564 }, { "epoch": 0.2780789245446661, "grad_norm": 1.1955187320709229, "learning_rate": 1.8211954156390544e-05, "loss": 0.0196, "step": 2565 }, { "epoch": 0.27818733738074586, "grad_norm": 5.912813663482666, "learning_rate": 1.8211228782823157e-05, "loss": 0.0255, "step": 2566 }, { "epoch": 0.2782957502168257, "grad_norm": 1.068504810333252, "learning_rate": 1.821050340925577e-05, "loss": 0.0235, "step": 2567 }, { "epoch": 0.27840416305290544, "grad_norm": 6.832455158233643, "learning_rate": 1.820977803568838e-05, "loss": 0.0345, "step": 2568 }, { "epoch": 0.27851257588898526, "grad_norm": 2.0034894943237305, "learning_rate": 1.8209052662120993e-05, "loss": 0.0217, "step": 2569 }, { "epoch": 0.278620988725065, "grad_norm": 2.7366883754730225, "learning_rate": 1.8208327288553605e-05, "loss": 0.0599, "step": 2570 }, { "epoch": 0.27872940156114484, "grad_norm": 2.4249496459960938, "learning_rate": 1.820760191498622e-05, "loss": 0.0198, "step": 2571 }, { "epoch": 0.2788378143972246, "grad_norm": 4.291657447814941, "learning_rate": 1.8206876541418833e-05, "loss": 0.0862, "step": 2572 }, { "epoch": 0.2789462272333044, "grad_norm": 2.006232738494873, "learning_rate": 1.8206151167851445e-05, "loss": 0.047, "step": 2573 }, { "epoch": 0.27905464006938424, "grad_norm": 7.8950958251953125, "learning_rate": 1.8205425794284057e-05, "loss": 0.0591, "step": 2574 }, { "epoch": 0.279163052905464, "grad_norm": 3.5779902935028076, "learning_rate": 1.820470042071667e-05, "loss": 0.038, "step": 2575 }, { "epoch": 0.2792714657415438, "grad_norm": 3.3573946952819824, "learning_rate": 1.820397504714928e-05, "loss": 0.0631, "step": 2576 }, { "epoch": 0.2793798785776236, "grad_norm": 2.0882468223571777, "learning_rate": 1.8203249673581897e-05, "loss": 0.044, "step": 2577 }, { "epoch": 0.2794882914137034, "grad_norm": 3.5587856769561768, "learning_rate": 1.820252430001451e-05, "loss": 0.0186, "step": 2578 }, { "epoch": 0.27959670424978317, "grad_norm": 4.466032028198242, "learning_rate": 1.820179892644712e-05, "loss": 0.0306, "step": 2579 }, { "epoch": 0.279705117085863, "grad_norm": 5.329821586608887, "learning_rate": 1.8201073552879737e-05, "loss": 0.0569, "step": 2580 }, { "epoch": 0.27981352992194275, "grad_norm": 5.509467601776123, "learning_rate": 1.820034817931235e-05, "loss": 0.0631, "step": 2581 }, { "epoch": 0.27992194275802257, "grad_norm": 2.992870569229126, "learning_rate": 1.819962280574496e-05, "loss": 0.0628, "step": 2582 }, { "epoch": 0.28003035559410233, "grad_norm": 0.5437257289886475, "learning_rate": 1.8198897432177573e-05, "loss": 0.0227, "step": 2583 }, { "epoch": 0.28013876843018215, "grad_norm": 3.1840760707855225, "learning_rate": 1.8198172058610185e-05, "loss": 0.0449, "step": 2584 }, { "epoch": 0.2802471812662619, "grad_norm": 5.993588924407959, "learning_rate": 1.8197446685042797e-05, "loss": 0.1263, "step": 2585 }, { "epoch": 0.28035559410234173, "grad_norm": 1.264582872390747, "learning_rate": 1.8196721311475413e-05, "loss": 0.0196, "step": 2586 }, { "epoch": 0.2804640069384215, "grad_norm": 0.9718048572540283, "learning_rate": 1.8195995937908025e-05, "loss": 0.0296, "step": 2587 }, { "epoch": 0.2805724197745013, "grad_norm": 5.179419040679932, "learning_rate": 1.8195270564340637e-05, "loss": 0.0321, "step": 2588 }, { "epoch": 0.2806808326105811, "grad_norm": 3.286144495010376, "learning_rate": 1.819454519077325e-05, "loss": 0.0535, "step": 2589 }, { "epoch": 0.2807892454466609, "grad_norm": 2.204617738723755, "learning_rate": 1.819381981720586e-05, "loss": 0.0356, "step": 2590 }, { "epoch": 0.28089765828274066, "grad_norm": 4.302628040313721, "learning_rate": 1.8193094443638474e-05, "loss": 0.0618, "step": 2591 }, { "epoch": 0.2810060711188205, "grad_norm": 4.526790142059326, "learning_rate": 1.8192369070071086e-05, "loss": 0.0538, "step": 2592 }, { "epoch": 0.28111448395490024, "grad_norm": 1.872822880744934, "learning_rate": 1.81916436965037e-05, "loss": 0.0492, "step": 2593 }, { "epoch": 0.28122289679098006, "grad_norm": 2.0064706802368164, "learning_rate": 1.8190918322936314e-05, "loss": 0.0246, "step": 2594 }, { "epoch": 0.2813313096270598, "grad_norm": 1.5347541570663452, "learning_rate": 1.8190192949368926e-05, "loss": 0.0372, "step": 2595 }, { "epoch": 0.28143972246313964, "grad_norm": 1.5891941785812378, "learning_rate": 1.8189467575801538e-05, "loss": 0.0203, "step": 2596 }, { "epoch": 0.2815481352992194, "grad_norm": 3.2273757457733154, "learning_rate": 1.8188742202234153e-05, "loss": 0.0344, "step": 2597 }, { "epoch": 0.2816565481352992, "grad_norm": 3.6526052951812744, "learning_rate": 1.8188016828666766e-05, "loss": 0.0228, "step": 2598 }, { "epoch": 0.281764960971379, "grad_norm": 1.353243350982666, "learning_rate": 1.8187291455099378e-05, "loss": 0.0357, "step": 2599 }, { "epoch": 0.2818733738074588, "grad_norm": 1.1323598623275757, "learning_rate": 1.818656608153199e-05, "loss": 0.0358, "step": 2600 }, { "epoch": 0.28198178664353857, "grad_norm": 1.2754943370819092, "learning_rate": 1.8185840707964605e-05, "loss": 0.0085, "step": 2601 }, { "epoch": 0.2820901994796184, "grad_norm": 3.562669515609741, "learning_rate": 1.8185115334397217e-05, "loss": 0.0585, "step": 2602 }, { "epoch": 0.2821986123156982, "grad_norm": 2.80770206451416, "learning_rate": 1.818438996082983e-05, "loss": 0.0372, "step": 2603 }, { "epoch": 0.28230702515177797, "grad_norm": 4.217139720916748, "learning_rate": 1.8183664587262442e-05, "loss": 0.0624, "step": 2604 }, { "epoch": 0.2824154379878578, "grad_norm": 3.3127553462982178, "learning_rate": 1.8182939213695054e-05, "loss": 0.042, "step": 2605 }, { "epoch": 0.28252385082393755, "grad_norm": 1.9067022800445557, "learning_rate": 1.8182213840127666e-05, "loss": 0.034, "step": 2606 }, { "epoch": 0.28263226366001737, "grad_norm": 2.329827308654785, "learning_rate": 1.8181488466560278e-05, "loss": 0.0047, "step": 2607 }, { "epoch": 0.28274067649609713, "grad_norm": 3.7448854446411133, "learning_rate": 1.8180763092992894e-05, "loss": 0.0284, "step": 2608 }, { "epoch": 0.28284908933217695, "grad_norm": 3.4544970989227295, "learning_rate": 1.8180037719425506e-05, "loss": 0.0466, "step": 2609 }, { "epoch": 0.2829575021682567, "grad_norm": 4.609902381896973, "learning_rate": 1.8179312345858118e-05, "loss": 0.0701, "step": 2610 }, { "epoch": 0.28306591500433653, "grad_norm": 0.7607726454734802, "learning_rate": 1.817858697229073e-05, "loss": 0.0062, "step": 2611 }, { "epoch": 0.2831743278404163, "grad_norm": 1.4387778043746948, "learning_rate": 1.8177861598723342e-05, "loss": 0.0296, "step": 2612 }, { "epoch": 0.2832827406764961, "grad_norm": 0.46676936745643616, "learning_rate": 1.8177136225155954e-05, "loss": 0.0149, "step": 2613 }, { "epoch": 0.2833911535125759, "grad_norm": 6.078226566314697, "learning_rate": 1.817641085158857e-05, "loss": 0.0653, "step": 2614 }, { "epoch": 0.2834995663486557, "grad_norm": 3.1621744632720947, "learning_rate": 1.8175685478021182e-05, "loss": 0.0634, "step": 2615 }, { "epoch": 0.28360797918473546, "grad_norm": 4.514850616455078, "learning_rate": 1.8174960104453798e-05, "loss": 0.0784, "step": 2616 }, { "epoch": 0.2837163920208153, "grad_norm": 3.580548048019409, "learning_rate": 1.817423473088641e-05, "loss": 0.0413, "step": 2617 }, { "epoch": 0.28382480485689504, "grad_norm": 7.604640007019043, "learning_rate": 1.8173509357319022e-05, "loss": 0.058, "step": 2618 }, { "epoch": 0.28393321769297486, "grad_norm": 3.047804117202759, "learning_rate": 1.8172783983751634e-05, "loss": 0.0805, "step": 2619 }, { "epoch": 0.2840416305290546, "grad_norm": 3.796596050262451, "learning_rate": 1.8172058610184246e-05, "loss": 0.037, "step": 2620 }, { "epoch": 0.28415004336513444, "grad_norm": 0.9650242328643799, "learning_rate": 1.817133323661686e-05, "loss": 0.019, "step": 2621 }, { "epoch": 0.2842584562012142, "grad_norm": 0.9184617400169373, "learning_rate": 1.817060786304947e-05, "loss": 0.0193, "step": 2622 }, { "epoch": 0.284366869037294, "grad_norm": 3.060595989227295, "learning_rate": 1.8169882489482086e-05, "loss": 0.0694, "step": 2623 }, { "epoch": 0.2844752818733738, "grad_norm": 4.026482105255127, "learning_rate": 1.8169157115914698e-05, "loss": 0.0659, "step": 2624 }, { "epoch": 0.2845836947094536, "grad_norm": 4.631894588470459, "learning_rate": 1.816843174234731e-05, "loss": 0.0481, "step": 2625 }, { "epoch": 0.28469210754553337, "grad_norm": 3.2928881645202637, "learning_rate": 1.8167706368779922e-05, "loss": 0.0527, "step": 2626 }, { "epoch": 0.2848005203816132, "grad_norm": 3.2151379585266113, "learning_rate": 1.8166980995212535e-05, "loss": 0.0606, "step": 2627 }, { "epoch": 0.28490893321769295, "grad_norm": 2.743525505065918, "learning_rate": 1.8166255621645147e-05, "loss": 0.0539, "step": 2628 }, { "epoch": 0.28501734605377277, "grad_norm": 2.8112947940826416, "learning_rate": 1.8165530248077762e-05, "loss": 0.0726, "step": 2629 }, { "epoch": 0.28512575888985253, "grad_norm": 1.523802399635315, "learning_rate": 1.8164804874510374e-05, "loss": 0.0451, "step": 2630 }, { "epoch": 0.28523417172593235, "grad_norm": 2.329040288925171, "learning_rate": 1.8164079500942987e-05, "loss": 0.0325, "step": 2631 }, { "epoch": 0.28534258456201217, "grad_norm": 2.260303020477295, "learning_rate": 1.8163354127375602e-05, "loss": 0.0516, "step": 2632 }, { "epoch": 0.28545099739809193, "grad_norm": 1.88040030002594, "learning_rate": 1.8162628753808214e-05, "loss": 0.0308, "step": 2633 }, { "epoch": 0.28555941023417175, "grad_norm": 3.2522902488708496, "learning_rate": 1.8161903380240826e-05, "loss": 0.0598, "step": 2634 }, { "epoch": 0.2856678230702515, "grad_norm": 2.8029532432556152, "learning_rate": 1.816117800667344e-05, "loss": 0.042, "step": 2635 }, { "epoch": 0.28577623590633133, "grad_norm": 3.5809144973754883, "learning_rate": 1.816045263310605e-05, "loss": 0.05, "step": 2636 }, { "epoch": 0.2858846487424111, "grad_norm": 2.513654947280884, "learning_rate": 1.8159727259538663e-05, "loss": 0.0365, "step": 2637 }, { "epoch": 0.2859930615784909, "grad_norm": 6.053982257843018, "learning_rate": 1.815900188597128e-05, "loss": 0.053, "step": 2638 }, { "epoch": 0.2861014744145707, "grad_norm": 2.163099527359009, "learning_rate": 1.815827651240389e-05, "loss": 0.0529, "step": 2639 }, { "epoch": 0.2862098872506505, "grad_norm": 4.588534355163574, "learning_rate": 1.8157551138836503e-05, "loss": 0.0356, "step": 2640 }, { "epoch": 0.28631830008673026, "grad_norm": 1.9792917966842651, "learning_rate": 1.8156825765269115e-05, "loss": 0.0445, "step": 2641 }, { "epoch": 0.2864267129228101, "grad_norm": 2.1149954795837402, "learning_rate": 1.8156100391701727e-05, "loss": 0.0402, "step": 2642 }, { "epoch": 0.28653512575888984, "grad_norm": 2.0014986991882324, "learning_rate": 1.815537501813434e-05, "loss": 0.0565, "step": 2643 }, { "epoch": 0.28664353859496966, "grad_norm": 1.2110977172851562, "learning_rate": 1.815464964456695e-05, "loss": 0.0301, "step": 2644 }, { "epoch": 0.2867519514310494, "grad_norm": 3.389934778213501, "learning_rate": 1.8153924270999567e-05, "loss": 0.0477, "step": 2645 }, { "epoch": 0.28686036426712924, "grad_norm": 2.4146759510040283, "learning_rate": 1.815319889743218e-05, "loss": 0.0935, "step": 2646 }, { "epoch": 0.286968777103209, "grad_norm": 2.223202705383301, "learning_rate": 1.815247352386479e-05, "loss": 0.0448, "step": 2647 }, { "epoch": 0.2870771899392888, "grad_norm": 1.7847702503204346, "learning_rate": 1.8151748150297403e-05, "loss": 0.0376, "step": 2648 }, { "epoch": 0.2871856027753686, "grad_norm": 2.8249659538269043, "learning_rate": 1.815102277673002e-05, "loss": 0.0362, "step": 2649 }, { "epoch": 0.2872940156114484, "grad_norm": 2.810603618621826, "learning_rate": 1.815029740316263e-05, "loss": 0.0591, "step": 2650 }, { "epoch": 0.28740242844752817, "grad_norm": 1.2566336393356323, "learning_rate": 1.8149572029595243e-05, "loss": 0.0361, "step": 2651 }, { "epoch": 0.287510841283608, "grad_norm": 1.2199214696884155, "learning_rate": 1.8148846656027855e-05, "loss": 0.0341, "step": 2652 }, { "epoch": 0.28761925411968775, "grad_norm": 0.9900665879249573, "learning_rate": 1.814812128246047e-05, "loss": 0.019, "step": 2653 }, { "epoch": 0.28772766695576757, "grad_norm": 1.6664397716522217, "learning_rate": 1.8147395908893083e-05, "loss": 0.0214, "step": 2654 }, { "epoch": 0.28783607979184733, "grad_norm": 9.716283798217773, "learning_rate": 1.8146670535325695e-05, "loss": 0.0868, "step": 2655 }, { "epoch": 0.28794449262792715, "grad_norm": 2.875930070877075, "learning_rate": 1.8145945161758307e-05, "loss": 0.0316, "step": 2656 }, { "epoch": 0.2880529054640069, "grad_norm": 0.5944059491157532, "learning_rate": 1.814521978819092e-05, "loss": 0.029, "step": 2657 }, { "epoch": 0.28816131830008673, "grad_norm": 5.705896854400635, "learning_rate": 1.814449441462353e-05, "loss": 0.0316, "step": 2658 }, { "epoch": 0.2882697311361665, "grad_norm": 1.5680358409881592, "learning_rate": 1.8143769041056144e-05, "loss": 0.0291, "step": 2659 }, { "epoch": 0.2883781439722463, "grad_norm": 3.0324947834014893, "learning_rate": 1.814304366748876e-05, "loss": 0.0709, "step": 2660 }, { "epoch": 0.28848655680832613, "grad_norm": 2.7542355060577393, "learning_rate": 1.814231829392137e-05, "loss": 0.0817, "step": 2661 }, { "epoch": 0.2885949696444059, "grad_norm": 3.498741388320923, "learning_rate": 1.8141592920353983e-05, "loss": 0.0468, "step": 2662 }, { "epoch": 0.2887033824804857, "grad_norm": 3.1206812858581543, "learning_rate": 1.8140867546786595e-05, "loss": 0.0422, "step": 2663 }, { "epoch": 0.2888117953165655, "grad_norm": 4.623449802398682, "learning_rate": 1.8140142173219208e-05, "loss": 0.0518, "step": 2664 }, { "epoch": 0.2889202081526453, "grad_norm": 2.6180830001831055, "learning_rate": 1.813941679965182e-05, "loss": 0.0587, "step": 2665 }, { "epoch": 0.28902862098872506, "grad_norm": 2.6964738368988037, "learning_rate": 1.8138691426084435e-05, "loss": 0.0643, "step": 2666 }, { "epoch": 0.2891370338248049, "grad_norm": 4.861929893493652, "learning_rate": 1.8137966052517047e-05, "loss": 0.0362, "step": 2667 }, { "epoch": 0.28924544666088464, "grad_norm": 3.9920551776885986, "learning_rate": 1.8137240678949663e-05, "loss": 0.0656, "step": 2668 }, { "epoch": 0.28935385949696446, "grad_norm": 5.3861494064331055, "learning_rate": 1.8136515305382275e-05, "loss": 0.0433, "step": 2669 }, { "epoch": 0.2894622723330442, "grad_norm": 1.8577021360397339, "learning_rate": 1.8135789931814887e-05, "loss": 0.0328, "step": 2670 }, { "epoch": 0.28957068516912404, "grad_norm": 0.9382845163345337, "learning_rate": 1.81350645582475e-05, "loss": 0.0273, "step": 2671 }, { "epoch": 0.2896790980052038, "grad_norm": 2.0082285404205322, "learning_rate": 1.813433918468011e-05, "loss": 0.0385, "step": 2672 }, { "epoch": 0.2897875108412836, "grad_norm": 1.4347968101501465, "learning_rate": 1.8133613811112724e-05, "loss": 0.0442, "step": 2673 }, { "epoch": 0.2898959236773634, "grad_norm": 10.05612564086914, "learning_rate": 1.8132888437545336e-05, "loss": 0.0475, "step": 2674 }, { "epoch": 0.2900043365134432, "grad_norm": 10.644655227661133, "learning_rate": 1.813216306397795e-05, "loss": 0.0373, "step": 2675 }, { "epoch": 0.29011274934952297, "grad_norm": 2.6914024353027344, "learning_rate": 1.8131437690410564e-05, "loss": 0.0722, "step": 2676 }, { "epoch": 0.2902211621856028, "grad_norm": 2.2648510932922363, "learning_rate": 1.8130712316843176e-05, "loss": 0.042, "step": 2677 }, { "epoch": 0.29032957502168255, "grad_norm": 7.358752727508545, "learning_rate": 1.8129986943275788e-05, "loss": 0.0623, "step": 2678 }, { "epoch": 0.29043798785776237, "grad_norm": 1.55581796169281, "learning_rate": 1.81292615697084e-05, "loss": 0.0224, "step": 2679 }, { "epoch": 0.29054640069384213, "grad_norm": 3.8306000232696533, "learning_rate": 1.8128536196141012e-05, "loss": 0.0159, "step": 2680 }, { "epoch": 0.29065481352992195, "grad_norm": 1.7715398073196411, "learning_rate": 1.8127810822573628e-05, "loss": 0.0494, "step": 2681 }, { "epoch": 0.2907632263660017, "grad_norm": 5.9304304122924805, "learning_rate": 1.812708544900624e-05, "loss": 0.0712, "step": 2682 }, { "epoch": 0.29087163920208153, "grad_norm": 1.2117218971252441, "learning_rate": 1.8126360075438852e-05, "loss": 0.0269, "step": 2683 }, { "epoch": 0.2909800520381613, "grad_norm": 2.77964186668396, "learning_rate": 1.8125634701871467e-05, "loss": 0.0234, "step": 2684 }, { "epoch": 0.2910884648742411, "grad_norm": 2.3792920112609863, "learning_rate": 1.812490932830408e-05, "loss": 0.0281, "step": 2685 }, { "epoch": 0.2911968777103209, "grad_norm": 1.7460105419158936, "learning_rate": 1.8124183954736692e-05, "loss": 0.0434, "step": 2686 }, { "epoch": 0.2913052905464007, "grad_norm": 1.2455815076828003, "learning_rate": 1.8123458581169304e-05, "loss": 0.0216, "step": 2687 }, { "epoch": 0.29141370338248046, "grad_norm": 2.6608312129974365, "learning_rate": 1.8122733207601916e-05, "loss": 0.0361, "step": 2688 }, { "epoch": 0.2915221162185603, "grad_norm": 2.210660457611084, "learning_rate": 1.8122007834034528e-05, "loss": 0.0581, "step": 2689 }, { "epoch": 0.29163052905464004, "grad_norm": 2.499636173248291, "learning_rate": 1.8121282460467144e-05, "loss": 0.0387, "step": 2690 }, { "epoch": 0.29173894189071986, "grad_norm": 1.654929757118225, "learning_rate": 1.8120557086899756e-05, "loss": 0.029, "step": 2691 }, { "epoch": 0.2918473547267997, "grad_norm": 2.890915870666504, "learning_rate": 1.8119831713332368e-05, "loss": 0.0624, "step": 2692 }, { "epoch": 0.29195576756287944, "grad_norm": 1.093050241470337, "learning_rate": 1.811910633976498e-05, "loss": 0.0267, "step": 2693 }, { "epoch": 0.29206418039895926, "grad_norm": 2.7982404232025146, "learning_rate": 1.8118380966197592e-05, "loss": 0.0307, "step": 2694 }, { "epoch": 0.292172593235039, "grad_norm": 2.4016454219818115, "learning_rate": 1.8117655592630204e-05, "loss": 0.0696, "step": 2695 }, { "epoch": 0.29228100607111884, "grad_norm": 1.132138729095459, "learning_rate": 1.8116930219062817e-05, "loss": 0.0164, "step": 2696 }, { "epoch": 0.2923894189071986, "grad_norm": 2.097791910171509, "learning_rate": 1.8116204845495432e-05, "loss": 0.0425, "step": 2697 }, { "epoch": 0.2924978317432784, "grad_norm": 2.7593371868133545, "learning_rate": 1.8115479471928044e-05, "loss": 0.0568, "step": 2698 }, { "epoch": 0.2926062445793582, "grad_norm": 8.731221199035645, "learning_rate": 1.8114754098360656e-05, "loss": 0.0449, "step": 2699 }, { "epoch": 0.292714657415438, "grad_norm": 3.357619047164917, "learning_rate": 1.8114028724793272e-05, "loss": 0.0347, "step": 2700 }, { "epoch": 0.29282307025151777, "grad_norm": 7.331010341644287, "learning_rate": 1.8113303351225884e-05, "loss": 0.1084, "step": 2701 }, { "epoch": 0.2929314830875976, "grad_norm": 4.018815517425537, "learning_rate": 1.8112577977658496e-05, "loss": 0.0689, "step": 2702 }, { "epoch": 0.29303989592367735, "grad_norm": 2.7923829555511475, "learning_rate": 1.811185260409111e-05, "loss": 0.0318, "step": 2703 }, { "epoch": 0.29314830875975717, "grad_norm": 3.316807746887207, "learning_rate": 1.811112723052372e-05, "loss": 0.0229, "step": 2704 }, { "epoch": 0.29325672159583693, "grad_norm": 1.7445199489593506, "learning_rate": 1.8110401856956336e-05, "loss": 0.0308, "step": 2705 }, { "epoch": 0.29336513443191675, "grad_norm": 2.123377561569214, "learning_rate": 1.8109676483388948e-05, "loss": 0.0339, "step": 2706 }, { "epoch": 0.2934735472679965, "grad_norm": 2.303553581237793, "learning_rate": 1.810895110982156e-05, "loss": 0.0341, "step": 2707 }, { "epoch": 0.29358196010407633, "grad_norm": 4.658823013305664, "learning_rate": 1.8108225736254172e-05, "loss": 0.0957, "step": 2708 }, { "epoch": 0.2936903729401561, "grad_norm": 1.3581557273864746, "learning_rate": 1.8107500362686785e-05, "loss": 0.0362, "step": 2709 }, { "epoch": 0.2937987857762359, "grad_norm": 4.976291656494141, "learning_rate": 1.8106774989119397e-05, "loss": 0.0526, "step": 2710 }, { "epoch": 0.2939071986123157, "grad_norm": 4.8549113273620605, "learning_rate": 1.810604961555201e-05, "loss": 0.041, "step": 2711 }, { "epoch": 0.2940156114483955, "grad_norm": 2.6938979625701904, "learning_rate": 1.8105324241984624e-05, "loss": 0.0553, "step": 2712 }, { "epoch": 0.29412402428447526, "grad_norm": 7.14109468460083, "learning_rate": 1.8104598868417237e-05, "loss": 0.068, "step": 2713 }, { "epoch": 0.2942324371205551, "grad_norm": 1.1501387357711792, "learning_rate": 1.810387349484985e-05, "loss": 0.0052, "step": 2714 }, { "epoch": 0.29434084995663484, "grad_norm": 2.9856932163238525, "learning_rate": 1.810314812128246e-05, "loss": 0.0434, "step": 2715 }, { "epoch": 0.29444926279271466, "grad_norm": 1.2474302053451538, "learning_rate": 1.8102422747715073e-05, "loss": 0.013, "step": 2716 }, { "epoch": 0.2945576756287944, "grad_norm": 0.9099125862121582, "learning_rate": 1.8101697374147685e-05, "loss": 0.0184, "step": 2717 }, { "epoch": 0.29466608846487424, "grad_norm": 0.436085045337677, "learning_rate": 1.81009720005803e-05, "loss": 0.0134, "step": 2718 }, { "epoch": 0.294774501300954, "grad_norm": 2.156698226928711, "learning_rate": 1.8100246627012913e-05, "loss": 0.0479, "step": 2719 }, { "epoch": 0.2948829141370338, "grad_norm": 5.478705883026123, "learning_rate": 1.809952125344553e-05, "loss": 0.0864, "step": 2720 }, { "epoch": 0.29499132697311364, "grad_norm": 3.438929319381714, "learning_rate": 1.809879587987814e-05, "loss": 0.042, "step": 2721 }, { "epoch": 0.2950997398091934, "grad_norm": 4.596712112426758, "learning_rate": 1.8098070506310753e-05, "loss": 0.0956, "step": 2722 }, { "epoch": 0.2952081526452732, "grad_norm": 0.3667796552181244, "learning_rate": 1.8097345132743365e-05, "loss": 0.0044, "step": 2723 }, { "epoch": 0.295316565481353, "grad_norm": 0.8751671314239502, "learning_rate": 1.8096619759175977e-05, "loss": 0.0073, "step": 2724 }, { "epoch": 0.2954249783174328, "grad_norm": 2.700227975845337, "learning_rate": 1.809589438560859e-05, "loss": 0.0712, "step": 2725 }, { "epoch": 0.29553339115351257, "grad_norm": 3.3826189041137695, "learning_rate": 1.80951690120412e-05, "loss": 0.0406, "step": 2726 }, { "epoch": 0.2956418039895924, "grad_norm": 1.9017833471298218, "learning_rate": 1.8094443638473817e-05, "loss": 0.0264, "step": 2727 }, { "epoch": 0.29575021682567215, "grad_norm": 4.030489444732666, "learning_rate": 1.809371826490643e-05, "loss": 0.0456, "step": 2728 }, { "epoch": 0.29585862966175197, "grad_norm": 3.0788846015930176, "learning_rate": 1.809299289133904e-05, "loss": 0.0331, "step": 2729 }, { "epoch": 0.29596704249783173, "grad_norm": 5.958075046539307, "learning_rate": 1.8092267517771653e-05, "loss": 0.0392, "step": 2730 }, { "epoch": 0.29607545533391155, "grad_norm": 2.3875391483306885, "learning_rate": 1.8091542144204265e-05, "loss": 0.0683, "step": 2731 }, { "epoch": 0.2961838681699913, "grad_norm": 4.601038932800293, "learning_rate": 1.8090816770636877e-05, "loss": 0.0296, "step": 2732 }, { "epoch": 0.29629228100607113, "grad_norm": 1.8147629499435425, "learning_rate": 1.8090091397069493e-05, "loss": 0.0241, "step": 2733 }, { "epoch": 0.2964006938421509, "grad_norm": 2.2039191722869873, "learning_rate": 1.8089366023502105e-05, "loss": 0.0487, "step": 2734 }, { "epoch": 0.2965091066782307, "grad_norm": 1.5244836807250977, "learning_rate": 1.8088640649934717e-05, "loss": 0.029, "step": 2735 }, { "epoch": 0.2966175195143105, "grad_norm": 2.747189998626709, "learning_rate": 1.8087915276367333e-05, "loss": 0.0349, "step": 2736 }, { "epoch": 0.2967259323503903, "grad_norm": 1.8825469017028809, "learning_rate": 1.8087189902799945e-05, "loss": 0.0468, "step": 2737 }, { "epoch": 0.29683434518647006, "grad_norm": 5.529403209686279, "learning_rate": 1.8086464529232557e-05, "loss": 0.0779, "step": 2738 }, { "epoch": 0.2969427580225499, "grad_norm": 1.6428452730178833, "learning_rate": 1.808573915566517e-05, "loss": 0.0333, "step": 2739 }, { "epoch": 0.29705117085862964, "grad_norm": 2.1500766277313232, "learning_rate": 1.808501378209778e-05, "loss": 0.0486, "step": 2740 }, { "epoch": 0.29715958369470946, "grad_norm": 5.677361011505127, "learning_rate": 1.8084288408530393e-05, "loss": 0.0875, "step": 2741 }, { "epoch": 0.2972679965307892, "grad_norm": 2.162693738937378, "learning_rate": 1.808356303496301e-05, "loss": 0.0337, "step": 2742 }, { "epoch": 0.29737640936686904, "grad_norm": 1.2262814044952393, "learning_rate": 1.808283766139562e-05, "loss": 0.0121, "step": 2743 }, { "epoch": 0.2974848222029488, "grad_norm": 2.0248844623565674, "learning_rate": 1.8082112287828233e-05, "loss": 0.0494, "step": 2744 }, { "epoch": 0.2975932350390286, "grad_norm": 3.359858989715576, "learning_rate": 1.8081386914260845e-05, "loss": 0.0261, "step": 2745 }, { "epoch": 0.2977016478751084, "grad_norm": 8.94878101348877, "learning_rate": 1.8080661540693458e-05, "loss": 0.0514, "step": 2746 }, { "epoch": 0.2978100607111882, "grad_norm": 1.8758275508880615, "learning_rate": 1.807993616712607e-05, "loss": 0.033, "step": 2747 }, { "epoch": 0.29791847354726797, "grad_norm": 1.9701976776123047, "learning_rate": 1.8079210793558682e-05, "loss": 0.0562, "step": 2748 }, { "epoch": 0.2980268863833478, "grad_norm": 2.185045003890991, "learning_rate": 1.8078485419991297e-05, "loss": 0.0393, "step": 2749 }, { "epoch": 0.2981352992194276, "grad_norm": 2.8945131301879883, "learning_rate": 1.807776004642391e-05, "loss": 0.0613, "step": 2750 }, { "epoch": 0.29824371205550737, "grad_norm": 2.266292095184326, "learning_rate": 1.8077034672856522e-05, "loss": 0.0214, "step": 2751 }, { "epoch": 0.2983521248915872, "grad_norm": 6.891785621643066, "learning_rate": 1.8076309299289137e-05, "loss": 0.0879, "step": 2752 }, { "epoch": 0.29846053772766695, "grad_norm": 3.4330246448516846, "learning_rate": 1.807558392572175e-05, "loss": 0.0363, "step": 2753 }, { "epoch": 0.29856895056374677, "grad_norm": 2.561795473098755, "learning_rate": 1.807485855215436e-05, "loss": 0.0313, "step": 2754 }, { "epoch": 0.29867736339982653, "grad_norm": 1.111541509628296, "learning_rate": 1.8074133178586974e-05, "loss": 0.0309, "step": 2755 }, { "epoch": 0.29878577623590635, "grad_norm": 0.6843072175979614, "learning_rate": 1.8073407805019586e-05, "loss": 0.0158, "step": 2756 }, { "epoch": 0.2988941890719861, "grad_norm": 3.6140518188476562, "learning_rate": 1.80726824314522e-05, "loss": 0.095, "step": 2757 }, { "epoch": 0.29900260190806593, "grad_norm": 4.627827167510986, "learning_rate": 1.8071957057884813e-05, "loss": 0.0365, "step": 2758 }, { "epoch": 0.2991110147441457, "grad_norm": 3.902773857116699, "learning_rate": 1.8071231684317426e-05, "loss": 0.0394, "step": 2759 }, { "epoch": 0.2992194275802255, "grad_norm": 2.658458709716797, "learning_rate": 1.8070506310750038e-05, "loss": 0.0399, "step": 2760 }, { "epoch": 0.2993278404163053, "grad_norm": 2.5355770587921143, "learning_rate": 1.806978093718265e-05, "loss": 0.0502, "step": 2761 }, { "epoch": 0.2994362532523851, "grad_norm": 6.071281433105469, "learning_rate": 1.8069055563615262e-05, "loss": 0.0491, "step": 2762 }, { "epoch": 0.29954466608846486, "grad_norm": 2.428184986114502, "learning_rate": 1.8068330190047874e-05, "loss": 0.0539, "step": 2763 }, { "epoch": 0.2996530789245447, "grad_norm": 0.8455801010131836, "learning_rate": 1.806760481648049e-05, "loss": 0.0219, "step": 2764 }, { "epoch": 0.29976149176062444, "grad_norm": 2.4180123805999756, "learning_rate": 1.8066879442913102e-05, "loss": 0.0176, "step": 2765 }, { "epoch": 0.29986990459670426, "grad_norm": 2.3482048511505127, "learning_rate": 1.8066154069345714e-05, "loss": 0.0335, "step": 2766 }, { "epoch": 0.299978317432784, "grad_norm": 7.143862724304199, "learning_rate": 1.8065428695778326e-05, "loss": 0.0642, "step": 2767 }, { "epoch": 0.30008673026886384, "grad_norm": 0.9544696807861328, "learning_rate": 1.806470332221094e-05, "loss": 0.0106, "step": 2768 }, { "epoch": 0.3001951431049436, "grad_norm": 3.923447847366333, "learning_rate": 1.806397794864355e-05, "loss": 0.0483, "step": 2769 }, { "epoch": 0.3003035559410234, "grad_norm": 8.82554817199707, "learning_rate": 1.8063252575076166e-05, "loss": 0.1247, "step": 2770 }, { "epoch": 0.3004119687771032, "grad_norm": 5.026492595672607, "learning_rate": 1.8062527201508778e-05, "loss": 0.0452, "step": 2771 }, { "epoch": 0.300520381613183, "grad_norm": 1.3101427555084229, "learning_rate": 1.8061801827941394e-05, "loss": 0.0234, "step": 2772 }, { "epoch": 0.30062879444926277, "grad_norm": 3.0686521530151367, "learning_rate": 1.8061076454374006e-05, "loss": 0.0404, "step": 2773 }, { "epoch": 0.3007372072853426, "grad_norm": 1.0717753171920776, "learning_rate": 1.8060351080806618e-05, "loss": 0.0058, "step": 2774 }, { "epoch": 0.30084562012142235, "grad_norm": 2.663888931274414, "learning_rate": 1.805962570723923e-05, "loss": 0.034, "step": 2775 }, { "epoch": 0.30095403295750217, "grad_norm": 5.443084716796875, "learning_rate": 1.8058900333671842e-05, "loss": 0.0959, "step": 2776 }, { "epoch": 0.30106244579358193, "grad_norm": 1.8356767892837524, "learning_rate": 1.8058174960104454e-05, "loss": 0.0379, "step": 2777 }, { "epoch": 0.30117085862966175, "grad_norm": 4.9931321144104, "learning_rate": 1.8057449586537067e-05, "loss": 0.0466, "step": 2778 }, { "epoch": 0.3012792714657415, "grad_norm": 3.1695497035980225, "learning_rate": 1.8056724212969682e-05, "loss": 0.0664, "step": 2779 }, { "epoch": 0.30138768430182133, "grad_norm": 4.383200645446777, "learning_rate": 1.8055998839402294e-05, "loss": 0.0365, "step": 2780 }, { "epoch": 0.30149609713790115, "grad_norm": 5.241037845611572, "learning_rate": 1.8055273465834906e-05, "loss": 0.0891, "step": 2781 }, { "epoch": 0.3016045099739809, "grad_norm": 3.5566582679748535, "learning_rate": 1.805454809226752e-05, "loss": 0.024, "step": 2782 }, { "epoch": 0.30171292281006074, "grad_norm": 1.941341757774353, "learning_rate": 1.805382271870013e-05, "loss": 0.015, "step": 2783 }, { "epoch": 0.3018213356461405, "grad_norm": 2.380605697631836, "learning_rate": 1.8053097345132743e-05, "loss": 0.0465, "step": 2784 }, { "epoch": 0.3019297484822203, "grad_norm": 2.6904425621032715, "learning_rate": 1.8052371971565358e-05, "loss": 0.0764, "step": 2785 }, { "epoch": 0.3020381613183001, "grad_norm": 1.3756234645843506, "learning_rate": 1.805164659799797e-05, "loss": 0.0228, "step": 2786 }, { "epoch": 0.3021465741543799, "grad_norm": 1.8110533952713013, "learning_rate": 1.8050921224430583e-05, "loss": 0.0597, "step": 2787 }, { "epoch": 0.30225498699045966, "grad_norm": 2.368753433227539, "learning_rate": 1.8050195850863198e-05, "loss": 0.0317, "step": 2788 }, { "epoch": 0.3023633998265395, "grad_norm": 2.121304512023926, "learning_rate": 1.804947047729581e-05, "loss": 0.0285, "step": 2789 }, { "epoch": 0.30247181266261924, "grad_norm": 1.772803544998169, "learning_rate": 1.8048745103728422e-05, "loss": 0.0434, "step": 2790 }, { "epoch": 0.30258022549869906, "grad_norm": 8.628003120422363, "learning_rate": 1.8048019730161035e-05, "loss": 0.034, "step": 2791 }, { "epoch": 0.3026886383347788, "grad_norm": 2.631889820098877, "learning_rate": 1.8047294356593647e-05, "loss": 0.0251, "step": 2792 }, { "epoch": 0.30279705117085864, "grad_norm": 1.7089945077896118, "learning_rate": 1.804656898302626e-05, "loss": 0.0207, "step": 2793 }, { "epoch": 0.3029054640069384, "grad_norm": 1.8548797369003296, "learning_rate": 1.8045843609458874e-05, "loss": 0.0353, "step": 2794 }, { "epoch": 0.3030138768430182, "grad_norm": 3.8697738647460938, "learning_rate": 1.8045118235891487e-05, "loss": 0.0328, "step": 2795 }, { "epoch": 0.303122289679098, "grad_norm": 3.0126984119415283, "learning_rate": 1.80443928623241e-05, "loss": 0.0556, "step": 2796 }, { "epoch": 0.3032307025151778, "grad_norm": 2.628809690475464, "learning_rate": 1.804366748875671e-05, "loss": 0.0472, "step": 2797 }, { "epoch": 0.30333911535125757, "grad_norm": 3.534487009048462, "learning_rate": 1.8042942115189323e-05, "loss": 0.0486, "step": 2798 }, { "epoch": 0.3034475281873374, "grad_norm": 3.9681153297424316, "learning_rate": 1.8042216741621935e-05, "loss": 0.03, "step": 2799 }, { "epoch": 0.30355594102341715, "grad_norm": 3.684715747833252, "learning_rate": 1.804149136805455e-05, "loss": 0.0541, "step": 2800 }, { "epoch": 0.30366435385949697, "grad_norm": 2.6113667488098145, "learning_rate": 1.8040765994487163e-05, "loss": 0.0759, "step": 2801 }, { "epoch": 0.30377276669557673, "grad_norm": 4.573042392730713, "learning_rate": 1.8040040620919775e-05, "loss": 0.0251, "step": 2802 }, { "epoch": 0.30388117953165655, "grad_norm": 2.9618113040924072, "learning_rate": 1.8039315247352387e-05, "loss": 0.0559, "step": 2803 }, { "epoch": 0.3039895923677363, "grad_norm": 0.5237324833869934, "learning_rate": 1.8038589873785003e-05, "loss": 0.0148, "step": 2804 }, { "epoch": 0.30409800520381614, "grad_norm": 4.412504196166992, "learning_rate": 1.8037864500217615e-05, "loss": 0.044, "step": 2805 }, { "epoch": 0.3042064180398959, "grad_norm": 2.178485155105591, "learning_rate": 1.8037139126650227e-05, "loss": 0.042, "step": 2806 }, { "epoch": 0.3043148308759757, "grad_norm": 1.2450183629989624, "learning_rate": 1.803641375308284e-05, "loss": 0.0229, "step": 2807 }, { "epoch": 0.3044232437120555, "grad_norm": 1.9448901414871216, "learning_rate": 1.803568837951545e-05, "loss": 0.042, "step": 2808 }, { "epoch": 0.3045316565481353, "grad_norm": 1.837175965309143, "learning_rate": 1.8034963005948067e-05, "loss": 0.0503, "step": 2809 }, { "epoch": 0.3046400693842151, "grad_norm": 3.768650531768799, "learning_rate": 1.803423763238068e-05, "loss": 0.0681, "step": 2810 }, { "epoch": 0.3047484822202949, "grad_norm": 1.8981271982192993, "learning_rate": 1.803351225881329e-05, "loss": 0.0485, "step": 2811 }, { "epoch": 0.3048568950563747, "grad_norm": 2.2807838916778564, "learning_rate": 1.8032786885245903e-05, "loss": 0.045, "step": 2812 }, { "epoch": 0.30496530789245446, "grad_norm": 2.576676368713379, "learning_rate": 1.8032061511678515e-05, "loss": 0.0082, "step": 2813 }, { "epoch": 0.3050737207285343, "grad_norm": 1.3503915071487427, "learning_rate": 1.8031336138111127e-05, "loss": 0.0319, "step": 2814 }, { "epoch": 0.30518213356461404, "grad_norm": 2.3219542503356934, "learning_rate": 1.803061076454374e-05, "loss": 0.0683, "step": 2815 }, { "epoch": 0.30529054640069386, "grad_norm": 3.232304811477661, "learning_rate": 1.8029885390976355e-05, "loss": 0.0358, "step": 2816 }, { "epoch": 0.3053989592367736, "grad_norm": 1.3117178678512573, "learning_rate": 1.8029160017408967e-05, "loss": 0.0214, "step": 2817 }, { "epoch": 0.30550737207285344, "grad_norm": 2.6633083820343018, "learning_rate": 1.802843464384158e-05, "loss": 0.0691, "step": 2818 }, { "epoch": 0.3056157849089332, "grad_norm": 2.0481326580047607, "learning_rate": 1.802770927027419e-05, "loss": 0.0472, "step": 2819 }, { "epoch": 0.305724197745013, "grad_norm": 1.8522998094558716, "learning_rate": 1.8026983896706804e-05, "loss": 0.0339, "step": 2820 }, { "epoch": 0.3058326105810928, "grad_norm": 1.273727536201477, "learning_rate": 1.802625852313942e-05, "loss": 0.0187, "step": 2821 }, { "epoch": 0.3059410234171726, "grad_norm": 1.7072495222091675, "learning_rate": 1.802553314957203e-05, "loss": 0.0482, "step": 2822 }, { "epoch": 0.30604943625325237, "grad_norm": 1.9384346008300781, "learning_rate": 1.8024807776004643e-05, "loss": 0.02, "step": 2823 }, { "epoch": 0.3061578490893322, "grad_norm": 2.2234253883361816, "learning_rate": 1.802408240243726e-05, "loss": 0.0515, "step": 2824 }, { "epoch": 0.30626626192541195, "grad_norm": 4.33506441116333, "learning_rate": 1.802335702886987e-05, "loss": 0.0438, "step": 2825 }, { "epoch": 0.30637467476149177, "grad_norm": 1.753700852394104, "learning_rate": 1.8022631655302483e-05, "loss": 0.0169, "step": 2826 }, { "epoch": 0.30648308759757154, "grad_norm": 2.7688496112823486, "learning_rate": 1.8021906281735095e-05, "loss": 0.0505, "step": 2827 }, { "epoch": 0.30659150043365135, "grad_norm": 2.542426109313965, "learning_rate": 1.8021180908167708e-05, "loss": 0.1047, "step": 2828 }, { "epoch": 0.3066999132697311, "grad_norm": 0.41457995772361755, "learning_rate": 1.802045553460032e-05, "loss": 0.0093, "step": 2829 }, { "epoch": 0.30680832610581094, "grad_norm": 1.8641927242279053, "learning_rate": 1.8019730161032932e-05, "loss": 0.0445, "step": 2830 }, { "epoch": 0.3069167389418907, "grad_norm": 3.62227463722229, "learning_rate": 1.8019004787465547e-05, "loss": 0.0401, "step": 2831 }, { "epoch": 0.3070251517779705, "grad_norm": 1.1042845249176025, "learning_rate": 1.801827941389816e-05, "loss": 0.0105, "step": 2832 }, { "epoch": 0.3071335646140503, "grad_norm": 3.9763646125793457, "learning_rate": 1.801755404033077e-05, "loss": 0.0589, "step": 2833 }, { "epoch": 0.3072419774501301, "grad_norm": 2.057054281234741, "learning_rate": 1.8016828666763384e-05, "loss": 0.0484, "step": 2834 }, { "epoch": 0.30735039028620986, "grad_norm": 8.312311172485352, "learning_rate": 1.8016103293195996e-05, "loss": 0.0771, "step": 2835 }, { "epoch": 0.3074588031222897, "grad_norm": 1.5301769971847534, "learning_rate": 1.8015377919628608e-05, "loss": 0.0677, "step": 2836 }, { "epoch": 0.30756721595836944, "grad_norm": 2.1694767475128174, "learning_rate": 1.8014652546061224e-05, "loss": 0.0575, "step": 2837 }, { "epoch": 0.30767562879444926, "grad_norm": 3.5689690113067627, "learning_rate": 1.8013927172493836e-05, "loss": 0.0598, "step": 2838 }, { "epoch": 0.3077840416305291, "grad_norm": 1.7987473011016846, "learning_rate": 1.8013201798926448e-05, "loss": 0.0413, "step": 2839 }, { "epoch": 0.30789245446660884, "grad_norm": 0.8497005105018616, "learning_rate": 1.8012476425359063e-05, "loss": 0.0149, "step": 2840 }, { "epoch": 0.30800086730268866, "grad_norm": 0.5761507153511047, "learning_rate": 1.8011751051791676e-05, "loss": 0.0152, "step": 2841 }, { "epoch": 0.3081092801387684, "grad_norm": 2.252553939819336, "learning_rate": 1.8011025678224288e-05, "loss": 0.044, "step": 2842 }, { "epoch": 0.30821769297484825, "grad_norm": 3.898939609527588, "learning_rate": 1.80103003046569e-05, "loss": 0.0988, "step": 2843 }, { "epoch": 0.308326105810928, "grad_norm": 1.0056172609329224, "learning_rate": 1.8009574931089512e-05, "loss": 0.0204, "step": 2844 }, { "epoch": 0.3084345186470078, "grad_norm": 1.140045404434204, "learning_rate": 1.8008849557522124e-05, "loss": 0.0275, "step": 2845 }, { "epoch": 0.3085429314830876, "grad_norm": 1.6983896493911743, "learning_rate": 1.800812418395474e-05, "loss": 0.0398, "step": 2846 }, { "epoch": 0.3086513443191674, "grad_norm": 1.4939920902252197, "learning_rate": 1.8007398810387352e-05, "loss": 0.0301, "step": 2847 }, { "epoch": 0.30875975715524717, "grad_norm": 6.795320510864258, "learning_rate": 1.8006673436819964e-05, "loss": 0.045, "step": 2848 }, { "epoch": 0.308868169991327, "grad_norm": 2.8586525917053223, "learning_rate": 1.8005948063252576e-05, "loss": 0.0489, "step": 2849 }, { "epoch": 0.30897658282740675, "grad_norm": 2.955429792404175, "learning_rate": 1.8005222689685188e-05, "loss": 0.0826, "step": 2850 }, { "epoch": 0.3090849956634866, "grad_norm": 2.9933021068573, "learning_rate": 1.80044973161178e-05, "loss": 0.0343, "step": 2851 }, { "epoch": 0.30919340849956634, "grad_norm": 1.2372817993164062, "learning_rate": 1.8003771942550416e-05, "loss": 0.0135, "step": 2852 }, { "epoch": 0.30930182133564615, "grad_norm": 3.98651385307312, "learning_rate": 1.8003046568983028e-05, "loss": 0.0789, "step": 2853 }, { "epoch": 0.3094102341717259, "grad_norm": 2.5399835109710693, "learning_rate": 1.800232119541564e-05, "loss": 0.0299, "step": 2854 }, { "epoch": 0.30951864700780574, "grad_norm": 0.7595511674880981, "learning_rate": 1.8001595821848252e-05, "loss": 0.0301, "step": 2855 }, { "epoch": 0.3096270598438855, "grad_norm": 2.2983615398406982, "learning_rate": 1.8000870448280868e-05, "loss": 0.0432, "step": 2856 }, { "epoch": 0.3097354726799653, "grad_norm": 0.24193207919597626, "learning_rate": 1.800014507471348e-05, "loss": 0.0022, "step": 2857 }, { "epoch": 0.3098438855160451, "grad_norm": 3.0091311931610107, "learning_rate": 1.7999419701146092e-05, "loss": 0.067, "step": 2858 }, { "epoch": 0.3099522983521249, "grad_norm": 2.2417869567871094, "learning_rate": 1.7998694327578704e-05, "loss": 0.0278, "step": 2859 }, { "epoch": 0.31006071118820466, "grad_norm": 5.3329386711120605, "learning_rate": 1.7997968954011316e-05, "loss": 0.025, "step": 2860 }, { "epoch": 0.3101691240242845, "grad_norm": 8.391142845153809, "learning_rate": 1.7997243580443932e-05, "loss": 0.0599, "step": 2861 }, { "epoch": 0.31027753686036424, "grad_norm": 2.3376927375793457, "learning_rate": 1.7996518206876544e-05, "loss": 0.072, "step": 2862 }, { "epoch": 0.31038594969644406, "grad_norm": 0.3779943585395813, "learning_rate": 1.7995792833309156e-05, "loss": 0.0042, "step": 2863 }, { "epoch": 0.3104943625325238, "grad_norm": 4.729840278625488, "learning_rate": 1.799506745974177e-05, "loss": 0.012, "step": 2864 }, { "epoch": 0.31060277536860365, "grad_norm": 1.1629270315170288, "learning_rate": 1.799434208617438e-05, "loss": 0.0223, "step": 2865 }, { "epoch": 0.3107111882046834, "grad_norm": 2.4752206802368164, "learning_rate": 1.7993616712606993e-05, "loss": 0.0228, "step": 2866 }, { "epoch": 0.3108196010407632, "grad_norm": 3.273961067199707, "learning_rate": 1.7992891339039605e-05, "loss": 0.0261, "step": 2867 }, { "epoch": 0.31092801387684305, "grad_norm": 2.1646111011505127, "learning_rate": 1.799216596547222e-05, "loss": 0.0417, "step": 2868 }, { "epoch": 0.3110364267129228, "grad_norm": 2.8788299560546875, "learning_rate": 1.7991440591904833e-05, "loss": 0.0513, "step": 2869 }, { "epoch": 0.3111448395490026, "grad_norm": 2.0139825344085693, "learning_rate": 1.7990715218337445e-05, "loss": 0.0227, "step": 2870 }, { "epoch": 0.3112532523850824, "grad_norm": 4.014468669891357, "learning_rate": 1.7989989844770057e-05, "loss": 0.0463, "step": 2871 }, { "epoch": 0.3113616652211622, "grad_norm": 6.7761311531066895, "learning_rate": 1.798926447120267e-05, "loss": 0.0458, "step": 2872 }, { "epoch": 0.311470078057242, "grad_norm": 0.7224375009536743, "learning_rate": 1.7988539097635285e-05, "loss": 0.0159, "step": 2873 }, { "epoch": 0.3115784908933218, "grad_norm": 4.458789825439453, "learning_rate": 1.7987813724067897e-05, "loss": 0.0779, "step": 2874 }, { "epoch": 0.31168690372940155, "grad_norm": 5.685678482055664, "learning_rate": 1.798708835050051e-05, "loss": 0.0143, "step": 2875 }, { "epoch": 0.3117953165654814, "grad_norm": 3.555609941482544, "learning_rate": 1.7986362976933124e-05, "loss": 0.042, "step": 2876 }, { "epoch": 0.31190372940156114, "grad_norm": 4.840651512145996, "learning_rate": 1.7985637603365736e-05, "loss": 0.0331, "step": 2877 }, { "epoch": 0.31201214223764095, "grad_norm": 4.074406147003174, "learning_rate": 1.798491222979835e-05, "loss": 0.0433, "step": 2878 }, { "epoch": 0.3121205550737207, "grad_norm": 0.8092332482337952, "learning_rate": 1.798418685623096e-05, "loss": 0.0231, "step": 2879 }, { "epoch": 0.31222896790980054, "grad_norm": 8.115175247192383, "learning_rate": 1.7983461482663573e-05, "loss": 0.0206, "step": 2880 }, { "epoch": 0.3123373807458803, "grad_norm": 3.527038335800171, "learning_rate": 1.7982736109096185e-05, "loss": 0.0762, "step": 2881 }, { "epoch": 0.3124457935819601, "grad_norm": 1.1607719659805298, "learning_rate": 1.7982010735528797e-05, "loss": 0.0124, "step": 2882 }, { "epoch": 0.3125542064180399, "grad_norm": 4.729116916656494, "learning_rate": 1.7981285361961413e-05, "loss": 0.1396, "step": 2883 }, { "epoch": 0.3126626192541197, "grad_norm": 2.371586799621582, "learning_rate": 1.7980559988394025e-05, "loss": 0.0395, "step": 2884 }, { "epoch": 0.31277103209019946, "grad_norm": 3.580778121948242, "learning_rate": 1.7979834614826637e-05, "loss": 0.0231, "step": 2885 }, { "epoch": 0.3128794449262793, "grad_norm": 4.97582483291626, "learning_rate": 1.797910924125925e-05, "loss": 0.1018, "step": 2886 }, { "epoch": 0.31298785776235905, "grad_norm": 3.154890298843384, "learning_rate": 1.797838386769186e-05, "loss": 0.0261, "step": 2887 }, { "epoch": 0.31309627059843886, "grad_norm": 0.8550955653190613, "learning_rate": 1.7977658494124473e-05, "loss": 0.0162, "step": 2888 }, { "epoch": 0.3132046834345186, "grad_norm": 2.1251049041748047, "learning_rate": 1.797693312055709e-05, "loss": 0.0232, "step": 2889 }, { "epoch": 0.31331309627059845, "grad_norm": 1.7756139039993286, "learning_rate": 1.79762077469897e-05, "loss": 0.0301, "step": 2890 }, { "epoch": 0.3134215091066782, "grad_norm": 2.78047776222229, "learning_rate": 1.7975482373422313e-05, "loss": 0.0355, "step": 2891 }, { "epoch": 0.313529921942758, "grad_norm": 4.149857521057129, "learning_rate": 1.797475699985493e-05, "loss": 0.0832, "step": 2892 }, { "epoch": 0.3136383347788378, "grad_norm": 3.0218310356140137, "learning_rate": 1.797403162628754e-05, "loss": 0.0699, "step": 2893 }, { "epoch": 0.3137467476149176, "grad_norm": 3.032285213470459, "learning_rate": 1.7973306252720153e-05, "loss": 0.0816, "step": 2894 }, { "epoch": 0.3138551604509974, "grad_norm": 2.5625057220458984, "learning_rate": 1.7972580879152765e-05, "loss": 0.0253, "step": 2895 }, { "epoch": 0.3139635732870772, "grad_norm": 2.0438199043273926, "learning_rate": 1.7971855505585377e-05, "loss": 0.0255, "step": 2896 }, { "epoch": 0.31407198612315695, "grad_norm": 6.373186111450195, "learning_rate": 1.797113013201799e-05, "loss": 0.0229, "step": 2897 }, { "epoch": 0.3141803989592368, "grad_norm": 1.3532977104187012, "learning_rate": 1.7970404758450605e-05, "loss": 0.0078, "step": 2898 }, { "epoch": 0.3142888117953166, "grad_norm": 2.7608275413513184, "learning_rate": 1.7969679384883217e-05, "loss": 0.0562, "step": 2899 }, { "epoch": 0.31439722463139635, "grad_norm": 2.0397021770477295, "learning_rate": 1.796895401131583e-05, "loss": 0.0408, "step": 2900 }, { "epoch": 0.3145056374674762, "grad_norm": 1.7175986766815186, "learning_rate": 1.796822863774844e-05, "loss": 0.0509, "step": 2901 }, { "epoch": 0.31461405030355594, "grad_norm": 3.0812723636627197, "learning_rate": 1.7967503264181054e-05, "loss": 0.0536, "step": 2902 }, { "epoch": 0.31472246313963576, "grad_norm": 2.735365867614746, "learning_rate": 1.7966777890613666e-05, "loss": 0.0335, "step": 2903 }, { "epoch": 0.3148308759757155, "grad_norm": 2.563356876373291, "learning_rate": 1.796605251704628e-05, "loss": 0.082, "step": 2904 }, { "epoch": 0.31493928881179534, "grad_norm": 1.167900800704956, "learning_rate": 1.7965327143478893e-05, "loss": 0.022, "step": 2905 }, { "epoch": 0.3150477016478751, "grad_norm": 3.770143985748291, "learning_rate": 1.7964601769911506e-05, "loss": 0.09, "step": 2906 }, { "epoch": 0.3151561144839549, "grad_norm": 2.616046190261841, "learning_rate": 1.7963876396344118e-05, "loss": 0.0657, "step": 2907 }, { "epoch": 0.3152645273200347, "grad_norm": 1.2405346632003784, "learning_rate": 1.7963151022776733e-05, "loss": 0.032, "step": 2908 }, { "epoch": 0.3153729401561145, "grad_norm": 1.067442774772644, "learning_rate": 1.7962425649209345e-05, "loss": 0.0354, "step": 2909 }, { "epoch": 0.31548135299219426, "grad_norm": 3.5606582164764404, "learning_rate": 1.7961700275641958e-05, "loss": 0.0903, "step": 2910 }, { "epoch": 0.3155897658282741, "grad_norm": 2.2765140533447266, "learning_rate": 1.796097490207457e-05, "loss": 0.0507, "step": 2911 }, { "epoch": 0.31569817866435385, "grad_norm": 2.030407428741455, "learning_rate": 1.7960249528507182e-05, "loss": 0.027, "step": 2912 }, { "epoch": 0.31580659150043366, "grad_norm": 1.160111665725708, "learning_rate": 1.7959524154939797e-05, "loss": 0.041, "step": 2913 }, { "epoch": 0.3159150043365134, "grad_norm": 1.5384725332260132, "learning_rate": 1.795879878137241e-05, "loss": 0.0276, "step": 2914 }, { "epoch": 0.31602341717259325, "grad_norm": 2.3152332305908203, "learning_rate": 1.795807340780502e-05, "loss": 0.0635, "step": 2915 }, { "epoch": 0.316131830008673, "grad_norm": 2.6071505546569824, "learning_rate": 1.7957348034237634e-05, "loss": 0.0553, "step": 2916 }, { "epoch": 0.31624024284475283, "grad_norm": 1.8574007749557495, "learning_rate": 1.7956622660670246e-05, "loss": 0.0512, "step": 2917 }, { "epoch": 0.3163486556808326, "grad_norm": 1.264221429824829, "learning_rate": 1.7955897287102858e-05, "loss": 0.0312, "step": 2918 }, { "epoch": 0.3164570685169124, "grad_norm": 5.534443378448486, "learning_rate": 1.795517191353547e-05, "loss": 0.0337, "step": 2919 }, { "epoch": 0.3165654813529922, "grad_norm": 1.1095532178878784, "learning_rate": 1.7954446539968086e-05, "loss": 0.024, "step": 2920 }, { "epoch": 0.316673894189072, "grad_norm": 5.370068550109863, "learning_rate": 1.7953721166400698e-05, "loss": 0.0351, "step": 2921 }, { "epoch": 0.31678230702515175, "grad_norm": 2.13427996635437, "learning_rate": 1.795299579283331e-05, "loss": 0.064, "step": 2922 }, { "epoch": 0.3168907198612316, "grad_norm": 6.9960455894470215, "learning_rate": 1.7952270419265922e-05, "loss": 0.0472, "step": 2923 }, { "epoch": 0.31699913269731134, "grad_norm": 3.6349527835845947, "learning_rate": 1.7951545045698534e-05, "loss": 0.041, "step": 2924 }, { "epoch": 0.31710754553339116, "grad_norm": 1.6081596612930298, "learning_rate": 1.795081967213115e-05, "loss": 0.0361, "step": 2925 }, { "epoch": 0.3172159583694709, "grad_norm": 2.3843796253204346, "learning_rate": 1.7950094298563762e-05, "loss": 0.0585, "step": 2926 }, { "epoch": 0.31732437120555074, "grad_norm": 0.5572659969329834, "learning_rate": 1.7949368924996374e-05, "loss": 0.0275, "step": 2927 }, { "epoch": 0.31743278404163056, "grad_norm": 3.3537087440490723, "learning_rate": 1.794864355142899e-05, "loss": 0.0657, "step": 2928 }, { "epoch": 0.3175411968777103, "grad_norm": 2.928220748901367, "learning_rate": 1.7947918177861602e-05, "loss": 0.0386, "step": 2929 }, { "epoch": 0.31764960971379014, "grad_norm": 2.0478227138519287, "learning_rate": 1.7947192804294214e-05, "loss": 0.0629, "step": 2930 }, { "epoch": 0.3177580225498699, "grad_norm": 0.8865842819213867, "learning_rate": 1.7946467430726826e-05, "loss": 0.0111, "step": 2931 }, { "epoch": 0.3178664353859497, "grad_norm": 3.311994791030884, "learning_rate": 1.7945742057159438e-05, "loss": 0.0326, "step": 2932 }, { "epoch": 0.3179748482220295, "grad_norm": 0.7979097366333008, "learning_rate": 1.794501668359205e-05, "loss": 0.0099, "step": 2933 }, { "epoch": 0.3180832610581093, "grad_norm": 4.588909149169922, "learning_rate": 1.7944291310024663e-05, "loss": 0.0524, "step": 2934 }, { "epoch": 0.31819167389418906, "grad_norm": 9.06116771697998, "learning_rate": 1.7943565936457278e-05, "loss": 0.077, "step": 2935 }, { "epoch": 0.3183000867302689, "grad_norm": 1.9339215755462646, "learning_rate": 1.794284056288989e-05, "loss": 0.0213, "step": 2936 }, { "epoch": 0.31840849956634865, "grad_norm": 3.983067035675049, "learning_rate": 1.7942115189322502e-05, "loss": 0.0379, "step": 2937 }, { "epoch": 0.31851691240242846, "grad_norm": 4.880102157592773, "learning_rate": 1.7941389815755114e-05, "loss": 0.0233, "step": 2938 }, { "epoch": 0.31862532523850823, "grad_norm": 3.4807937145233154, "learning_rate": 1.7940664442187727e-05, "loss": 0.0347, "step": 2939 }, { "epoch": 0.31873373807458805, "grad_norm": 5.055349349975586, "learning_rate": 1.793993906862034e-05, "loss": 0.0985, "step": 2940 }, { "epoch": 0.3188421509106678, "grad_norm": 8.951018333435059, "learning_rate": 1.7939213695052954e-05, "loss": 0.0777, "step": 2941 }, { "epoch": 0.31895056374674763, "grad_norm": 2.2469513416290283, "learning_rate": 1.7938488321485566e-05, "loss": 0.0182, "step": 2942 }, { "epoch": 0.3190589765828274, "grad_norm": 2.4908056259155273, "learning_rate": 1.793776294791818e-05, "loss": 0.0269, "step": 2943 }, { "epoch": 0.3191673894189072, "grad_norm": 7.954975605010986, "learning_rate": 1.7937037574350794e-05, "loss": 0.0318, "step": 2944 }, { "epoch": 0.319275802254987, "grad_norm": 6.557727336883545, "learning_rate": 1.7936312200783406e-05, "loss": 0.0706, "step": 2945 }, { "epoch": 0.3193842150910668, "grad_norm": 1.3752386569976807, "learning_rate": 1.793558682721602e-05, "loss": 0.0253, "step": 2946 }, { "epoch": 0.31949262792714656, "grad_norm": 4.869333744049072, "learning_rate": 1.793486145364863e-05, "loss": 0.0384, "step": 2947 }, { "epoch": 0.3196010407632264, "grad_norm": 0.7232645153999329, "learning_rate": 1.7934136080081243e-05, "loss": 0.0099, "step": 2948 }, { "epoch": 0.31970945359930614, "grad_norm": 3.0899765491485596, "learning_rate": 1.7933410706513855e-05, "loss": 0.0207, "step": 2949 }, { "epoch": 0.31981786643538596, "grad_norm": 1.854850172996521, "learning_rate": 1.793268533294647e-05, "loss": 0.0329, "step": 2950 }, { "epoch": 0.3199262792714657, "grad_norm": 5.336958408355713, "learning_rate": 1.7931959959379083e-05, "loss": 0.0635, "step": 2951 }, { "epoch": 0.32003469210754554, "grad_norm": 3.3276216983795166, "learning_rate": 1.7931234585811695e-05, "loss": 0.0504, "step": 2952 }, { "epoch": 0.3201431049436253, "grad_norm": 2.685647964477539, "learning_rate": 1.7930509212244307e-05, "loss": 0.05, "step": 2953 }, { "epoch": 0.3202515177797051, "grad_norm": 0.7723896503448486, "learning_rate": 1.792978383867692e-05, "loss": 0.0133, "step": 2954 }, { "epoch": 0.3203599306157849, "grad_norm": 2.211742639541626, "learning_rate": 1.792905846510953e-05, "loss": 0.0414, "step": 2955 }, { "epoch": 0.3204683434518647, "grad_norm": 1.2895170450210571, "learning_rate": 1.7928333091542147e-05, "loss": 0.0094, "step": 2956 }, { "epoch": 0.3205767562879445, "grad_norm": 2.798295259475708, "learning_rate": 1.792760771797476e-05, "loss": 0.038, "step": 2957 }, { "epoch": 0.3206851691240243, "grad_norm": 4.99495792388916, "learning_rate": 1.792688234440737e-05, "loss": 0.0361, "step": 2958 }, { "epoch": 0.3207935819601041, "grad_norm": 1.553117275238037, "learning_rate": 1.7926156970839983e-05, "loss": 0.0203, "step": 2959 }, { "epoch": 0.32090199479618386, "grad_norm": 5.797891616821289, "learning_rate": 1.79254315972726e-05, "loss": 0.0498, "step": 2960 }, { "epoch": 0.3210104076322637, "grad_norm": 4.138906955718994, "learning_rate": 1.792470622370521e-05, "loss": 0.054, "step": 2961 }, { "epoch": 0.32111882046834345, "grad_norm": 1.604718804359436, "learning_rate": 1.7923980850137823e-05, "loss": 0.0273, "step": 2962 }, { "epoch": 0.32122723330442327, "grad_norm": 2.505911111831665, "learning_rate": 1.7923255476570435e-05, "loss": 0.0388, "step": 2963 }, { "epoch": 0.32133564614050303, "grad_norm": 3.9175126552581787, "learning_rate": 1.7922530103003047e-05, "loss": 0.0489, "step": 2964 }, { "epoch": 0.32144405897658285, "grad_norm": 4.219705581665039, "learning_rate": 1.7921804729435663e-05, "loss": 0.0138, "step": 2965 }, { "epoch": 0.3215524718126626, "grad_norm": 2.6312406063079834, "learning_rate": 1.7921079355868275e-05, "loss": 0.0534, "step": 2966 }, { "epoch": 0.32166088464874243, "grad_norm": 6.961330413818359, "learning_rate": 1.7920353982300887e-05, "loss": 0.075, "step": 2967 }, { "epoch": 0.3217692974848222, "grad_norm": 2.745276927947998, "learning_rate": 1.79196286087335e-05, "loss": 0.0291, "step": 2968 }, { "epoch": 0.321877710320902, "grad_norm": 1.84662926197052, "learning_rate": 1.791890323516611e-05, "loss": 0.042, "step": 2969 }, { "epoch": 0.3219861231569818, "grad_norm": 2.4717960357666016, "learning_rate": 1.7918177861598723e-05, "loss": 0.0569, "step": 2970 }, { "epoch": 0.3220945359930616, "grad_norm": 5.231015682220459, "learning_rate": 1.7917452488031336e-05, "loss": 0.0488, "step": 2971 }, { "epoch": 0.32220294882914136, "grad_norm": 3.1245086193084717, "learning_rate": 1.791672711446395e-05, "loss": 0.1073, "step": 2972 }, { "epoch": 0.3223113616652212, "grad_norm": 1.3597874641418457, "learning_rate": 1.7916001740896563e-05, "loss": 0.0351, "step": 2973 }, { "epoch": 0.32241977450130094, "grad_norm": 3.404825448989868, "learning_rate": 1.7915276367329175e-05, "loss": 0.0689, "step": 2974 }, { "epoch": 0.32252818733738076, "grad_norm": 3.771801710128784, "learning_rate": 1.7914550993761788e-05, "loss": 0.0428, "step": 2975 }, { "epoch": 0.3226366001734605, "grad_norm": 1.6620982885360718, "learning_rate": 1.79138256201944e-05, "loss": 0.0245, "step": 2976 }, { "epoch": 0.32274501300954034, "grad_norm": 4.618094444274902, "learning_rate": 1.7913100246627015e-05, "loss": 0.0835, "step": 2977 }, { "epoch": 0.3228534258456201, "grad_norm": 0.8557807207107544, "learning_rate": 1.7912374873059627e-05, "loss": 0.0163, "step": 2978 }, { "epoch": 0.3229618386816999, "grad_norm": 2.706719160079956, "learning_rate": 1.791164949949224e-05, "loss": 0.0362, "step": 2979 }, { "epoch": 0.3230702515177797, "grad_norm": 5.114683628082275, "learning_rate": 1.7910924125924855e-05, "loss": 0.031, "step": 2980 }, { "epoch": 0.3231786643538595, "grad_norm": 0.1666995882987976, "learning_rate": 1.7910198752357467e-05, "loss": 0.0018, "step": 2981 }, { "epoch": 0.32328707718993926, "grad_norm": 1.9585118293762207, "learning_rate": 1.790947337879008e-05, "loss": 0.0405, "step": 2982 }, { "epoch": 0.3233954900260191, "grad_norm": 1.0728180408477783, "learning_rate": 1.790874800522269e-05, "loss": 0.0156, "step": 2983 }, { "epoch": 0.32350390286209885, "grad_norm": 3.6592488288879395, "learning_rate": 1.7908022631655304e-05, "loss": 0.0676, "step": 2984 }, { "epoch": 0.32361231569817867, "grad_norm": 2.3755862712860107, "learning_rate": 1.7907297258087916e-05, "loss": 0.0383, "step": 2985 }, { "epoch": 0.3237207285342585, "grad_norm": 1.5619690418243408, "learning_rate": 1.7906571884520528e-05, "loss": 0.0328, "step": 2986 }, { "epoch": 0.32382914137033825, "grad_norm": 2.3123538494110107, "learning_rate": 1.7905846510953143e-05, "loss": 0.0299, "step": 2987 }, { "epoch": 0.32393755420641807, "grad_norm": 6.742070198059082, "learning_rate": 1.7905121137385756e-05, "loss": 0.0456, "step": 2988 }, { "epoch": 0.32404596704249783, "grad_norm": 2.5743649005889893, "learning_rate": 1.7904395763818368e-05, "loss": 0.041, "step": 2989 }, { "epoch": 0.32415437987857765, "grad_norm": 3.947770595550537, "learning_rate": 1.790367039025098e-05, "loss": 0.0691, "step": 2990 }, { "epoch": 0.3242627927146574, "grad_norm": 1.068336009979248, "learning_rate": 1.7902945016683592e-05, "loss": 0.0284, "step": 2991 }, { "epoch": 0.32437120555073723, "grad_norm": 1.9140032529830933, "learning_rate": 1.7902219643116204e-05, "loss": 0.0236, "step": 2992 }, { "epoch": 0.324479618386817, "grad_norm": 2.133638620376587, "learning_rate": 1.790149426954882e-05, "loss": 0.0471, "step": 2993 }, { "epoch": 0.3245880312228968, "grad_norm": 9.012883186340332, "learning_rate": 1.7900768895981432e-05, "loss": 0.0441, "step": 2994 }, { "epoch": 0.3246964440589766, "grad_norm": 2.234076738357544, "learning_rate": 1.7900043522414047e-05, "loss": 0.0412, "step": 2995 }, { "epoch": 0.3248048568950564, "grad_norm": 2.819051504135132, "learning_rate": 1.789931814884666e-05, "loss": 0.0734, "step": 2996 }, { "epoch": 0.32491326973113616, "grad_norm": 2.2325730323791504, "learning_rate": 1.789859277527927e-05, "loss": 0.0614, "step": 2997 }, { "epoch": 0.325021682567216, "grad_norm": 0.9870586395263672, "learning_rate": 1.7897867401711884e-05, "loss": 0.0201, "step": 2998 }, { "epoch": 0.32513009540329574, "grad_norm": 4.034526348114014, "learning_rate": 1.7897142028144496e-05, "loss": 0.0952, "step": 2999 }, { "epoch": 0.32523850823937556, "grad_norm": 1.4176907539367676, "learning_rate": 1.7896416654577108e-05, "loss": 0.0361, "step": 3000 }, { "epoch": 0.3253469210754553, "grad_norm": 5.285325527191162, "learning_rate": 1.789569128100972e-05, "loss": 0.0255, "step": 3001 }, { "epoch": 0.32545533391153514, "grad_norm": 3.5775227546691895, "learning_rate": 1.7894965907442336e-05, "loss": 0.0528, "step": 3002 }, { "epoch": 0.3255637467476149, "grad_norm": 1.0985182523727417, "learning_rate": 1.7894240533874948e-05, "loss": 0.01, "step": 3003 }, { "epoch": 0.3256721595836947, "grad_norm": 1.9976228475570679, "learning_rate": 1.789351516030756e-05, "loss": 0.0302, "step": 3004 }, { "epoch": 0.3257805724197745, "grad_norm": 0.4558621048927307, "learning_rate": 1.7892789786740172e-05, "loss": 0.0135, "step": 3005 }, { "epoch": 0.3258889852558543, "grad_norm": 1.810655117034912, "learning_rate": 1.7892064413172784e-05, "loss": 0.027, "step": 3006 }, { "epoch": 0.32599739809193407, "grad_norm": 1.0461277961730957, "learning_rate": 1.7891339039605396e-05, "loss": 0.0209, "step": 3007 }, { "epoch": 0.3261058109280139, "grad_norm": 3.241807222366333, "learning_rate": 1.7890613666038012e-05, "loss": 0.0433, "step": 3008 }, { "epoch": 0.32621422376409365, "grad_norm": 1.2920453548431396, "learning_rate": 1.7889888292470624e-05, "loss": 0.0213, "step": 3009 }, { "epoch": 0.32632263660017347, "grad_norm": 3.2627813816070557, "learning_rate": 1.7889162918903236e-05, "loss": 0.0715, "step": 3010 }, { "epoch": 0.32643104943625323, "grad_norm": 1.9883747100830078, "learning_rate": 1.7888437545335852e-05, "loss": 0.0311, "step": 3011 }, { "epoch": 0.32653946227233305, "grad_norm": 1.1475754976272583, "learning_rate": 1.7887712171768464e-05, "loss": 0.0225, "step": 3012 }, { "epoch": 0.3266478751084128, "grad_norm": 4.296020030975342, "learning_rate": 1.7886986798201076e-05, "loss": 0.0379, "step": 3013 }, { "epoch": 0.32675628794449263, "grad_norm": 3.161506414413452, "learning_rate": 1.7886261424633688e-05, "loss": 0.0523, "step": 3014 }, { "epoch": 0.3268647007805724, "grad_norm": 0.7956510186195374, "learning_rate": 1.78855360510663e-05, "loss": 0.0261, "step": 3015 }, { "epoch": 0.3269731136166522, "grad_norm": 6.050285816192627, "learning_rate": 1.7884810677498913e-05, "loss": 0.0568, "step": 3016 }, { "epoch": 0.32708152645273203, "grad_norm": 2.9650917053222656, "learning_rate": 1.7884085303931528e-05, "loss": 0.0745, "step": 3017 }, { "epoch": 0.3271899392888118, "grad_norm": 2.903272867202759, "learning_rate": 1.788335993036414e-05, "loss": 0.0252, "step": 3018 }, { "epoch": 0.3272983521248916, "grad_norm": 1.6064037084579468, "learning_rate": 1.7882634556796752e-05, "loss": 0.0185, "step": 3019 }, { "epoch": 0.3274067649609714, "grad_norm": 2.998322010040283, "learning_rate": 1.7881909183229364e-05, "loss": 0.0888, "step": 3020 }, { "epoch": 0.3275151777970512, "grad_norm": 2.5652670860290527, "learning_rate": 1.7881183809661977e-05, "loss": 0.0912, "step": 3021 }, { "epoch": 0.32762359063313096, "grad_norm": 1.5423938035964966, "learning_rate": 1.788045843609459e-05, "loss": 0.0372, "step": 3022 }, { "epoch": 0.3277320034692108, "grad_norm": 4.742353916168213, "learning_rate": 1.78797330625272e-05, "loss": 0.0818, "step": 3023 }, { "epoch": 0.32784041630529054, "grad_norm": 1.9403313398361206, "learning_rate": 1.7879007688959816e-05, "loss": 0.027, "step": 3024 }, { "epoch": 0.32794882914137036, "grad_norm": 1.7066426277160645, "learning_rate": 1.787828231539243e-05, "loss": 0.0414, "step": 3025 }, { "epoch": 0.3280572419774501, "grad_norm": 2.615689277648926, "learning_rate": 1.787755694182504e-05, "loss": 0.0203, "step": 3026 }, { "epoch": 0.32816565481352994, "grad_norm": 2.494875907897949, "learning_rate": 1.7876831568257653e-05, "loss": 0.035, "step": 3027 }, { "epoch": 0.3282740676496097, "grad_norm": 2.9928338527679443, "learning_rate": 1.7876106194690265e-05, "loss": 0.0147, "step": 3028 }, { "epoch": 0.3283824804856895, "grad_norm": 3.335927724838257, "learning_rate": 1.787538082112288e-05, "loss": 0.0606, "step": 3029 }, { "epoch": 0.3284908933217693, "grad_norm": 1.266322374343872, "learning_rate": 1.7874655447555493e-05, "loss": 0.0288, "step": 3030 }, { "epoch": 0.3285993061578491, "grad_norm": 0.7758939266204834, "learning_rate": 1.7873930073988105e-05, "loss": 0.0065, "step": 3031 }, { "epoch": 0.32870771899392887, "grad_norm": 1.8454564809799194, "learning_rate": 1.787320470042072e-05, "loss": 0.0587, "step": 3032 }, { "epoch": 0.3288161318300087, "grad_norm": 0.7047408819198608, "learning_rate": 1.7872479326853332e-05, "loss": 0.024, "step": 3033 }, { "epoch": 0.32892454466608845, "grad_norm": 1.427189588546753, "learning_rate": 1.7871753953285945e-05, "loss": 0.0379, "step": 3034 }, { "epoch": 0.32903295750216827, "grad_norm": 5.120039463043213, "learning_rate": 1.7871028579718557e-05, "loss": 0.0827, "step": 3035 }, { "epoch": 0.32914137033824803, "grad_norm": 3.267435073852539, "learning_rate": 1.787030320615117e-05, "loss": 0.0517, "step": 3036 }, { "epoch": 0.32924978317432785, "grad_norm": 2.4137303829193115, "learning_rate": 1.786957783258378e-05, "loss": 0.0352, "step": 3037 }, { "epoch": 0.3293581960104076, "grad_norm": 2.0223541259765625, "learning_rate": 1.7868852459016393e-05, "loss": 0.026, "step": 3038 }, { "epoch": 0.32946660884648743, "grad_norm": 0.923394501209259, "learning_rate": 1.786812708544901e-05, "loss": 0.0156, "step": 3039 }, { "epoch": 0.3295750216825672, "grad_norm": 2.9499125480651855, "learning_rate": 1.786740171188162e-05, "loss": 0.0493, "step": 3040 }, { "epoch": 0.329683434518647, "grad_norm": 2.0894198417663574, "learning_rate": 1.7866676338314233e-05, "loss": 0.051, "step": 3041 }, { "epoch": 0.3297918473547268, "grad_norm": 2.401683807373047, "learning_rate": 1.7865950964746845e-05, "loss": 0.0255, "step": 3042 }, { "epoch": 0.3299002601908066, "grad_norm": 4.703960418701172, "learning_rate": 1.7865225591179457e-05, "loss": 0.0459, "step": 3043 }, { "epoch": 0.33000867302688636, "grad_norm": 1.2003473043441772, "learning_rate": 1.786450021761207e-05, "loss": 0.0154, "step": 3044 }, { "epoch": 0.3301170858629662, "grad_norm": 3.2808547019958496, "learning_rate": 1.7863774844044685e-05, "loss": 0.0422, "step": 3045 }, { "epoch": 0.330225498699046, "grad_norm": 2.1124327182769775, "learning_rate": 1.7863049470477297e-05, "loss": 0.0347, "step": 3046 }, { "epoch": 0.33033391153512576, "grad_norm": 3.4784293174743652, "learning_rate": 1.7862324096909913e-05, "loss": 0.0339, "step": 3047 }, { "epoch": 0.3304423243712056, "grad_norm": 5.971360683441162, "learning_rate": 1.7861598723342525e-05, "loss": 0.0241, "step": 3048 }, { "epoch": 0.33055073720728534, "grad_norm": 3.099514961242676, "learning_rate": 1.7860873349775137e-05, "loss": 0.0395, "step": 3049 }, { "epoch": 0.33065915004336516, "grad_norm": 4.383482933044434, "learning_rate": 1.786014797620775e-05, "loss": 0.0315, "step": 3050 }, { "epoch": 0.3307675628794449, "grad_norm": 1.1242482662200928, "learning_rate": 1.785942260264036e-05, "loss": 0.0152, "step": 3051 }, { "epoch": 0.33087597571552474, "grad_norm": 2.0040392875671387, "learning_rate": 1.7858697229072973e-05, "loss": 0.0456, "step": 3052 }, { "epoch": 0.3309843885516045, "grad_norm": 8.608331680297852, "learning_rate": 1.7857971855505586e-05, "loss": 0.1009, "step": 3053 }, { "epoch": 0.3310928013876843, "grad_norm": 3.4332876205444336, "learning_rate": 1.78572464819382e-05, "loss": 0.0455, "step": 3054 }, { "epoch": 0.3312012142237641, "grad_norm": 2.615208864212036, "learning_rate": 1.7856521108370813e-05, "loss": 0.0413, "step": 3055 }, { "epoch": 0.3313096270598439, "grad_norm": 4.909117698669434, "learning_rate": 1.7855795734803425e-05, "loss": 0.079, "step": 3056 }, { "epoch": 0.33141803989592367, "grad_norm": 1.3010791540145874, "learning_rate": 1.7855070361236037e-05, "loss": 0.0095, "step": 3057 }, { "epoch": 0.3315264527320035, "grad_norm": 3.3662073612213135, "learning_rate": 1.785434498766865e-05, "loss": 0.0388, "step": 3058 }, { "epoch": 0.33163486556808325, "grad_norm": 1.7106237411499023, "learning_rate": 1.7853619614101262e-05, "loss": 0.0291, "step": 3059 }, { "epoch": 0.33174327840416307, "grad_norm": 2.2431938648223877, "learning_rate": 1.7852894240533877e-05, "loss": 0.019, "step": 3060 }, { "epoch": 0.33185169124024283, "grad_norm": 2.7484664916992188, "learning_rate": 1.785216886696649e-05, "loss": 0.0367, "step": 3061 }, { "epoch": 0.33196010407632265, "grad_norm": 5.612175941467285, "learning_rate": 1.78514434933991e-05, "loss": 0.05, "step": 3062 }, { "epoch": 0.3320685169124024, "grad_norm": 3.8198163509368896, "learning_rate": 1.7850718119831717e-05, "loss": 0.0793, "step": 3063 }, { "epoch": 0.33217692974848223, "grad_norm": 4.521243572235107, "learning_rate": 1.784999274626433e-05, "loss": 0.0683, "step": 3064 }, { "epoch": 0.332285342584562, "grad_norm": 3.6317355632781982, "learning_rate": 1.784926737269694e-05, "loss": 0.0456, "step": 3065 }, { "epoch": 0.3323937554206418, "grad_norm": 4.69364595413208, "learning_rate": 1.7848541999129554e-05, "loss": 0.0471, "step": 3066 }, { "epoch": 0.3325021682567216, "grad_norm": 0.669378936290741, "learning_rate": 1.7847816625562166e-05, "loss": 0.0156, "step": 3067 }, { "epoch": 0.3326105810928014, "grad_norm": 1.4317842721939087, "learning_rate": 1.7847091251994778e-05, "loss": 0.0256, "step": 3068 }, { "epoch": 0.33271899392888116, "grad_norm": 3.7469282150268555, "learning_rate": 1.7846365878427393e-05, "loss": 0.0349, "step": 3069 }, { "epoch": 0.332827406764961, "grad_norm": 1.4969125986099243, "learning_rate": 1.7845640504860006e-05, "loss": 0.0287, "step": 3070 }, { "epoch": 0.33293581960104074, "grad_norm": 0.5686762928962708, "learning_rate": 1.7844915131292618e-05, "loss": 0.0072, "step": 3071 }, { "epoch": 0.33304423243712056, "grad_norm": 1.2569477558135986, "learning_rate": 1.784418975772523e-05, "loss": 0.0246, "step": 3072 }, { "epoch": 0.3331526452732003, "grad_norm": 1.3558201789855957, "learning_rate": 1.7843464384157842e-05, "loss": 0.0217, "step": 3073 }, { "epoch": 0.33326105810928014, "grad_norm": 2.996443271636963, "learning_rate": 1.7842739010590454e-05, "loss": 0.029, "step": 3074 }, { "epoch": 0.33336947094535996, "grad_norm": 4.118475437164307, "learning_rate": 1.7842013637023066e-05, "loss": 0.0215, "step": 3075 }, { "epoch": 0.3334778837814397, "grad_norm": 1.3178547620773315, "learning_rate": 1.7841288263455682e-05, "loss": 0.0226, "step": 3076 }, { "epoch": 0.33358629661751954, "grad_norm": 2.2369720935821533, "learning_rate": 1.7840562889888294e-05, "loss": 0.0459, "step": 3077 }, { "epoch": 0.3336947094535993, "grad_norm": 1.0505108833312988, "learning_rate": 1.7839837516320906e-05, "loss": 0.0138, "step": 3078 }, { "epoch": 0.3338031222896791, "grad_norm": 7.138694763183594, "learning_rate": 1.7839112142753518e-05, "loss": 0.0237, "step": 3079 }, { "epoch": 0.3339115351257589, "grad_norm": 3.0207626819610596, "learning_rate": 1.783838676918613e-05, "loss": 0.0432, "step": 3080 }, { "epoch": 0.3340199479618387, "grad_norm": 3.103224277496338, "learning_rate": 1.7837661395618746e-05, "loss": 0.0626, "step": 3081 }, { "epoch": 0.33412836079791847, "grad_norm": 1.171137809753418, "learning_rate": 1.7836936022051358e-05, "loss": 0.0138, "step": 3082 }, { "epoch": 0.3342367736339983, "grad_norm": 4.165595531463623, "learning_rate": 1.783621064848397e-05, "loss": 0.0384, "step": 3083 }, { "epoch": 0.33434518647007805, "grad_norm": 5.449735164642334, "learning_rate": 1.7835485274916586e-05, "loss": 0.0286, "step": 3084 }, { "epoch": 0.33445359930615787, "grad_norm": 2.491960287094116, "learning_rate": 1.7834759901349198e-05, "loss": 0.0706, "step": 3085 }, { "epoch": 0.33456201214223763, "grad_norm": 5.791715145111084, "learning_rate": 1.783403452778181e-05, "loss": 0.0584, "step": 3086 }, { "epoch": 0.33467042497831745, "grad_norm": 0.499948114156723, "learning_rate": 1.7833309154214422e-05, "loss": 0.008, "step": 3087 }, { "epoch": 0.3347788378143972, "grad_norm": 3.635878801345825, "learning_rate": 1.7832583780647034e-05, "loss": 0.0395, "step": 3088 }, { "epoch": 0.33488725065047703, "grad_norm": 1.0260330438613892, "learning_rate": 1.7831858407079646e-05, "loss": 0.0096, "step": 3089 }, { "epoch": 0.3349956634865568, "grad_norm": 0.388176292181015, "learning_rate": 1.783113303351226e-05, "loss": 0.0029, "step": 3090 }, { "epoch": 0.3351040763226366, "grad_norm": 9.230364799499512, "learning_rate": 1.7830407659944874e-05, "loss": 0.0976, "step": 3091 }, { "epoch": 0.3352124891587164, "grad_norm": 1.6017290353775024, "learning_rate": 1.7829682286377486e-05, "loss": 0.0452, "step": 3092 }, { "epoch": 0.3353209019947962, "grad_norm": 3.410568952560425, "learning_rate": 1.78289569128101e-05, "loss": 0.0125, "step": 3093 }, { "epoch": 0.33542931483087596, "grad_norm": 0.9402046203613281, "learning_rate": 1.782823153924271e-05, "loss": 0.0223, "step": 3094 }, { "epoch": 0.3355377276669558, "grad_norm": 6.043087005615234, "learning_rate": 1.7827506165675323e-05, "loss": 0.0932, "step": 3095 }, { "epoch": 0.33564614050303554, "grad_norm": 0.9111256003379822, "learning_rate": 1.7826780792107935e-05, "loss": 0.0046, "step": 3096 }, { "epoch": 0.33575455333911536, "grad_norm": 2.1594855785369873, "learning_rate": 1.782605541854055e-05, "loss": 0.0289, "step": 3097 }, { "epoch": 0.3358629661751951, "grad_norm": 2.313297748565674, "learning_rate": 1.7825330044973162e-05, "loss": 0.0524, "step": 3098 }, { "epoch": 0.33597137901127494, "grad_norm": 4.605688095092773, "learning_rate": 1.7824604671405778e-05, "loss": 0.0635, "step": 3099 }, { "epoch": 0.3360797918473547, "grad_norm": 3.1928062438964844, "learning_rate": 1.782387929783839e-05, "loss": 0.0477, "step": 3100 }, { "epoch": 0.3361882046834345, "grad_norm": 7.522801399230957, "learning_rate": 1.7823153924271002e-05, "loss": 0.0544, "step": 3101 }, { "epoch": 0.3362966175195143, "grad_norm": 2.2224912643432617, "learning_rate": 1.7822428550703614e-05, "loss": 0.0186, "step": 3102 }, { "epoch": 0.3364050303555941, "grad_norm": 9.31919002532959, "learning_rate": 1.7821703177136227e-05, "loss": 0.0708, "step": 3103 }, { "epoch": 0.33651344319167387, "grad_norm": 1.0770682096481323, "learning_rate": 1.782097780356884e-05, "loss": 0.0129, "step": 3104 }, { "epoch": 0.3366218560277537, "grad_norm": 2.7902355194091797, "learning_rate": 1.782025243000145e-05, "loss": 0.0528, "step": 3105 }, { "epoch": 0.3367302688638335, "grad_norm": 5.580099582672119, "learning_rate": 1.7819527056434066e-05, "loss": 0.0754, "step": 3106 }, { "epoch": 0.33683868169991327, "grad_norm": 2.219505548477173, "learning_rate": 1.781880168286668e-05, "loss": 0.0423, "step": 3107 }, { "epoch": 0.3369470945359931, "grad_norm": 3.3100533485412598, "learning_rate": 1.781807630929929e-05, "loss": 0.0349, "step": 3108 }, { "epoch": 0.33705550737207285, "grad_norm": 3.0975611209869385, "learning_rate": 1.7817350935731903e-05, "loss": 0.0364, "step": 3109 }, { "epoch": 0.33716392020815267, "grad_norm": 3.8600869178771973, "learning_rate": 1.7816625562164515e-05, "loss": 0.0781, "step": 3110 }, { "epoch": 0.33727233304423243, "grad_norm": 0.6488578915596008, "learning_rate": 1.7815900188597127e-05, "loss": 0.0156, "step": 3111 }, { "epoch": 0.33738074588031225, "grad_norm": 0.7456246614456177, "learning_rate": 1.7815174815029743e-05, "loss": 0.0162, "step": 3112 }, { "epoch": 0.337489158716392, "grad_norm": 1.823073387145996, "learning_rate": 1.7814449441462355e-05, "loss": 0.0259, "step": 3113 }, { "epoch": 0.33759757155247183, "grad_norm": 1.7439723014831543, "learning_rate": 1.7813724067894967e-05, "loss": 0.0217, "step": 3114 }, { "epoch": 0.3377059843885516, "grad_norm": 1.5972843170166016, "learning_rate": 1.7812998694327582e-05, "loss": 0.0198, "step": 3115 }, { "epoch": 0.3378143972246314, "grad_norm": 2.4324734210968018, "learning_rate": 1.7812273320760195e-05, "loss": 0.039, "step": 3116 }, { "epoch": 0.3379228100607112, "grad_norm": 2.292048215866089, "learning_rate": 1.7811547947192807e-05, "loss": 0.0172, "step": 3117 }, { "epoch": 0.338031222896791, "grad_norm": 2.072850465774536, "learning_rate": 1.781082257362542e-05, "loss": 0.0171, "step": 3118 }, { "epoch": 0.33813963573287076, "grad_norm": 4.705867290496826, "learning_rate": 1.781009720005803e-05, "loss": 0.0462, "step": 3119 }, { "epoch": 0.3382480485689506, "grad_norm": 3.360874652862549, "learning_rate": 1.7809371826490643e-05, "loss": 0.0693, "step": 3120 }, { "epoch": 0.33835646140503034, "grad_norm": 1.6587047576904297, "learning_rate": 1.780864645292326e-05, "loss": 0.0247, "step": 3121 }, { "epoch": 0.33846487424111016, "grad_norm": 1.457999587059021, "learning_rate": 1.780792107935587e-05, "loss": 0.0296, "step": 3122 }, { "epoch": 0.3385732870771899, "grad_norm": 3.4057180881500244, "learning_rate": 1.7807195705788483e-05, "loss": 0.0486, "step": 3123 }, { "epoch": 0.33868169991326974, "grad_norm": 6.5116095542907715, "learning_rate": 1.7806470332221095e-05, "loss": 0.0234, "step": 3124 }, { "epoch": 0.3387901127493495, "grad_norm": 2.548415422439575, "learning_rate": 1.7805744958653707e-05, "loss": 0.0283, "step": 3125 }, { "epoch": 0.3388985255854293, "grad_norm": 4.133167266845703, "learning_rate": 1.780501958508632e-05, "loss": 0.0579, "step": 3126 }, { "epoch": 0.3390069384215091, "grad_norm": 1.029006838798523, "learning_rate": 1.780429421151893e-05, "loss": 0.0178, "step": 3127 }, { "epoch": 0.3391153512575889, "grad_norm": 2.515138626098633, "learning_rate": 1.7803568837951547e-05, "loss": 0.0608, "step": 3128 }, { "epoch": 0.33922376409366867, "grad_norm": 3.7579240798950195, "learning_rate": 1.780284346438416e-05, "loss": 0.0192, "step": 3129 }, { "epoch": 0.3393321769297485, "grad_norm": 4.100281238555908, "learning_rate": 1.780211809081677e-05, "loss": 0.0293, "step": 3130 }, { "epoch": 0.33944058976582825, "grad_norm": 4.027987480163574, "learning_rate": 1.7801392717249384e-05, "loss": 0.0404, "step": 3131 }, { "epoch": 0.33954900260190807, "grad_norm": 4.06412410736084, "learning_rate": 1.7800667343682e-05, "loss": 0.0658, "step": 3132 }, { "epoch": 0.33965741543798783, "grad_norm": 3.1434614658355713, "learning_rate": 1.779994197011461e-05, "loss": 0.051, "step": 3133 }, { "epoch": 0.33976582827406765, "grad_norm": 5.215285301208496, "learning_rate": 1.7799216596547223e-05, "loss": 0.0552, "step": 3134 }, { "epoch": 0.33987424111014747, "grad_norm": 2.420541524887085, "learning_rate": 1.7798491222979835e-05, "loss": 0.0641, "step": 3135 }, { "epoch": 0.33998265394622723, "grad_norm": 0.6275432705879211, "learning_rate": 1.779776584941245e-05, "loss": 0.008, "step": 3136 }, { "epoch": 0.34009106678230705, "grad_norm": 5.3303632736206055, "learning_rate": 1.7797040475845063e-05, "loss": 0.0314, "step": 3137 }, { "epoch": 0.3401994796183868, "grad_norm": 1.2859669923782349, "learning_rate": 1.7796315102277675e-05, "loss": 0.0144, "step": 3138 }, { "epoch": 0.34030789245446663, "grad_norm": 1.3937746286392212, "learning_rate": 1.7795589728710287e-05, "loss": 0.0146, "step": 3139 }, { "epoch": 0.3404163052905464, "grad_norm": 4.318231582641602, "learning_rate": 1.77948643551429e-05, "loss": 0.075, "step": 3140 }, { "epoch": 0.3405247181266262, "grad_norm": 2.4758498668670654, "learning_rate": 1.7794138981575512e-05, "loss": 0.0312, "step": 3141 }, { "epoch": 0.340633130962706, "grad_norm": 3.1024043560028076, "learning_rate": 1.7793413608008124e-05, "loss": 0.0372, "step": 3142 }, { "epoch": 0.3407415437987858, "grad_norm": 3.3788294792175293, "learning_rate": 1.779268823444074e-05, "loss": 0.0273, "step": 3143 }, { "epoch": 0.34084995663486556, "grad_norm": 2.6100099086761475, "learning_rate": 1.779196286087335e-05, "loss": 0.0326, "step": 3144 }, { "epoch": 0.3409583694709454, "grad_norm": 6.85964822769165, "learning_rate": 1.7791237487305964e-05, "loss": 0.0503, "step": 3145 }, { "epoch": 0.34106678230702514, "grad_norm": 4.081611633300781, "learning_rate": 1.7790512113738576e-05, "loss": 0.1024, "step": 3146 }, { "epoch": 0.34117519514310496, "grad_norm": 6.815065860748291, "learning_rate": 1.7789786740171188e-05, "loss": 0.0389, "step": 3147 }, { "epoch": 0.3412836079791847, "grad_norm": 3.169703722000122, "learning_rate": 1.77890613666038e-05, "loss": 0.0579, "step": 3148 }, { "epoch": 0.34139202081526454, "grad_norm": 1.2101593017578125, "learning_rate": 1.7788335993036416e-05, "loss": 0.0385, "step": 3149 }, { "epoch": 0.3415004336513443, "grad_norm": 2.5643057823181152, "learning_rate": 1.7787610619469028e-05, "loss": 0.061, "step": 3150 }, { "epoch": 0.3416088464874241, "grad_norm": 0.8732064366340637, "learning_rate": 1.7786885245901643e-05, "loss": 0.0186, "step": 3151 }, { "epoch": 0.3417172593235039, "grad_norm": 0.8909441232681274, "learning_rate": 1.7786159872334255e-05, "loss": 0.0116, "step": 3152 }, { "epoch": 0.3418256721595837, "grad_norm": 4.080467224121094, "learning_rate": 1.7785434498766868e-05, "loss": 0.0678, "step": 3153 }, { "epoch": 0.34193408499566347, "grad_norm": 5.226571559906006, "learning_rate": 1.778470912519948e-05, "loss": 0.0748, "step": 3154 }, { "epoch": 0.3420424978317433, "grad_norm": 2.257305145263672, "learning_rate": 1.7783983751632092e-05, "loss": 0.0291, "step": 3155 }, { "epoch": 0.34215091066782305, "grad_norm": 0.8962790369987488, "learning_rate": 1.7783258378064704e-05, "loss": 0.0221, "step": 3156 }, { "epoch": 0.34225932350390287, "grad_norm": 1.5719021558761597, "learning_rate": 1.7782533004497316e-05, "loss": 0.035, "step": 3157 }, { "epoch": 0.34236773633998263, "grad_norm": 2.788393020629883, "learning_rate": 1.7781807630929932e-05, "loss": 0.0332, "step": 3158 }, { "epoch": 0.34247614917606245, "grad_norm": 1.8871281147003174, "learning_rate": 1.7781082257362544e-05, "loss": 0.0398, "step": 3159 }, { "epoch": 0.3425845620121422, "grad_norm": 0.937585175037384, "learning_rate": 1.7780356883795156e-05, "loss": 0.0388, "step": 3160 }, { "epoch": 0.34269297484822203, "grad_norm": 3.206697702407837, "learning_rate": 1.7779631510227768e-05, "loss": 0.0407, "step": 3161 }, { "epoch": 0.3428013876843018, "grad_norm": 1.4487595558166504, "learning_rate": 1.777890613666038e-05, "loss": 0.0167, "step": 3162 }, { "epoch": 0.3429098005203816, "grad_norm": 3.579288959503174, "learning_rate": 1.7778180763092992e-05, "loss": 0.0467, "step": 3163 }, { "epoch": 0.34301821335646143, "grad_norm": 8.757317543029785, "learning_rate": 1.7777455389525608e-05, "loss": 0.0351, "step": 3164 }, { "epoch": 0.3431266261925412, "grad_norm": 5.374607086181641, "learning_rate": 1.777673001595822e-05, "loss": 0.0332, "step": 3165 }, { "epoch": 0.343235039028621, "grad_norm": 2.565295934677124, "learning_rate": 1.7776004642390832e-05, "loss": 0.0287, "step": 3166 }, { "epoch": 0.3433434518647008, "grad_norm": 0.6589891314506531, "learning_rate": 1.7775279268823448e-05, "loss": 0.0203, "step": 3167 }, { "epoch": 0.3434518647007806, "grad_norm": 4.652326583862305, "learning_rate": 1.777455389525606e-05, "loss": 0.0448, "step": 3168 }, { "epoch": 0.34356027753686036, "grad_norm": 3.672187089920044, "learning_rate": 1.7773828521688672e-05, "loss": 0.0757, "step": 3169 }, { "epoch": 0.3436686903729402, "grad_norm": 2.8604679107666016, "learning_rate": 1.7773103148121284e-05, "loss": 0.0215, "step": 3170 }, { "epoch": 0.34377710320901994, "grad_norm": 3.756725311279297, "learning_rate": 1.7772377774553896e-05, "loss": 0.0473, "step": 3171 }, { "epoch": 0.34388551604509976, "grad_norm": 1.6646900177001953, "learning_rate": 1.777165240098651e-05, "loss": 0.0233, "step": 3172 }, { "epoch": 0.3439939288811795, "grad_norm": 2.122865676879883, "learning_rate": 1.7770927027419124e-05, "loss": 0.0146, "step": 3173 }, { "epoch": 0.34410234171725934, "grad_norm": 1.330749750137329, "learning_rate": 1.7770201653851736e-05, "loss": 0.0189, "step": 3174 }, { "epoch": 0.3442107545533391, "grad_norm": 5.883344650268555, "learning_rate": 1.776947628028435e-05, "loss": 0.0865, "step": 3175 }, { "epoch": 0.3443191673894189, "grad_norm": 3.6616668701171875, "learning_rate": 1.776875090671696e-05, "loss": 0.068, "step": 3176 }, { "epoch": 0.3444275802254987, "grad_norm": 1.7510759830474854, "learning_rate": 1.7768025533149573e-05, "loss": 0.0148, "step": 3177 }, { "epoch": 0.3445359930615785, "grad_norm": 0.4209284484386444, "learning_rate": 1.7767300159582185e-05, "loss": 0.0087, "step": 3178 }, { "epoch": 0.34464440589765827, "grad_norm": 12.438647270202637, "learning_rate": 1.7766574786014797e-05, "loss": 0.0455, "step": 3179 }, { "epoch": 0.3447528187337381, "grad_norm": 3.7290565967559814, "learning_rate": 1.7765849412447412e-05, "loss": 0.0774, "step": 3180 }, { "epoch": 0.34486123156981785, "grad_norm": 3.5154383182525635, "learning_rate": 1.7765124038880025e-05, "loss": 0.0244, "step": 3181 }, { "epoch": 0.34496964440589767, "grad_norm": 9.345235824584961, "learning_rate": 1.7764398665312637e-05, "loss": 0.0274, "step": 3182 }, { "epoch": 0.34507805724197743, "grad_norm": 9.345863342285156, "learning_rate": 1.776367329174525e-05, "loss": 0.032, "step": 3183 }, { "epoch": 0.34518647007805725, "grad_norm": 2.713512897491455, "learning_rate": 1.7762947918177864e-05, "loss": 0.019, "step": 3184 }, { "epoch": 0.345294882914137, "grad_norm": 2.790048360824585, "learning_rate": 1.7762222544610477e-05, "loss": 0.0537, "step": 3185 }, { "epoch": 0.34540329575021683, "grad_norm": 4.974491596221924, "learning_rate": 1.776149717104309e-05, "loss": 0.0558, "step": 3186 }, { "epoch": 0.3455117085862966, "grad_norm": 3.6420910358428955, "learning_rate": 1.77607717974757e-05, "loss": 0.0339, "step": 3187 }, { "epoch": 0.3456201214223764, "grad_norm": 0.4818480610847473, "learning_rate": 1.7760046423908316e-05, "loss": 0.0035, "step": 3188 }, { "epoch": 0.3457285342584562, "grad_norm": 2.603463649749756, "learning_rate": 1.775932105034093e-05, "loss": 0.0295, "step": 3189 }, { "epoch": 0.345836947094536, "grad_norm": 1.5668079853057861, "learning_rate": 1.775859567677354e-05, "loss": 0.0073, "step": 3190 }, { "epoch": 0.34594535993061576, "grad_norm": 3.1724631786346436, "learning_rate": 1.7757870303206153e-05, "loss": 0.0752, "step": 3191 }, { "epoch": 0.3460537727666956, "grad_norm": 2.889686346054077, "learning_rate": 1.7757144929638765e-05, "loss": 0.0371, "step": 3192 }, { "epoch": 0.3461621856027754, "grad_norm": 10.184039115905762, "learning_rate": 1.7756419556071377e-05, "loss": 0.0566, "step": 3193 }, { "epoch": 0.34627059843885516, "grad_norm": 3.84887433052063, "learning_rate": 1.775569418250399e-05, "loss": 0.0574, "step": 3194 }, { "epoch": 0.346379011274935, "grad_norm": 1.9497802257537842, "learning_rate": 1.7754968808936605e-05, "loss": 0.0481, "step": 3195 }, { "epoch": 0.34648742411101474, "grad_norm": 4.785370349884033, "learning_rate": 1.7754243435369217e-05, "loss": 0.0493, "step": 3196 }, { "epoch": 0.34659583694709456, "grad_norm": 1.9907441139221191, "learning_rate": 1.775351806180183e-05, "loss": 0.0311, "step": 3197 }, { "epoch": 0.3467042497831743, "grad_norm": 2.588010311126709, "learning_rate": 1.775279268823444e-05, "loss": 0.043, "step": 3198 }, { "epoch": 0.34681266261925414, "grad_norm": 2.4737491607666016, "learning_rate": 1.7752067314667053e-05, "loss": 0.0666, "step": 3199 }, { "epoch": 0.3469210754553339, "grad_norm": 2.1020054817199707, "learning_rate": 1.7751341941099665e-05, "loss": 0.0483, "step": 3200 }, { "epoch": 0.3470294882914137, "grad_norm": 6.053797245025635, "learning_rate": 1.775061656753228e-05, "loss": 0.0636, "step": 3201 }, { "epoch": 0.3471379011274935, "grad_norm": 5.404990196228027, "learning_rate": 1.7749891193964893e-05, "loss": 0.0597, "step": 3202 }, { "epoch": 0.3472463139635733, "grad_norm": 4.092360019683838, "learning_rate": 1.774916582039751e-05, "loss": 0.0193, "step": 3203 }, { "epoch": 0.34735472679965307, "grad_norm": 1.6679295301437378, "learning_rate": 1.774844044683012e-05, "loss": 0.0312, "step": 3204 }, { "epoch": 0.3474631396357329, "grad_norm": 2.688328266143799, "learning_rate": 1.7747715073262733e-05, "loss": 0.0619, "step": 3205 }, { "epoch": 0.34757155247181265, "grad_norm": 1.1487858295440674, "learning_rate": 1.7746989699695345e-05, "loss": 0.0157, "step": 3206 }, { "epoch": 0.34767996530789247, "grad_norm": 3.8852672576904297, "learning_rate": 1.7746264326127957e-05, "loss": 0.0386, "step": 3207 }, { "epoch": 0.34778837814397223, "grad_norm": 1.5558308362960815, "learning_rate": 1.774553895256057e-05, "loss": 0.0442, "step": 3208 }, { "epoch": 0.34789679098005205, "grad_norm": 2.786409616470337, "learning_rate": 1.774481357899318e-05, "loss": 0.0345, "step": 3209 }, { "epoch": 0.3480052038161318, "grad_norm": 1.986738681793213, "learning_rate": 1.7744088205425797e-05, "loss": 0.0413, "step": 3210 }, { "epoch": 0.34811361665221163, "grad_norm": 1.4156349897384644, "learning_rate": 1.774336283185841e-05, "loss": 0.0299, "step": 3211 }, { "epoch": 0.3482220294882914, "grad_norm": 3.5341274738311768, "learning_rate": 1.774263745829102e-05, "loss": 0.0234, "step": 3212 }, { "epoch": 0.3483304423243712, "grad_norm": 0.9718440175056458, "learning_rate": 1.7741912084723634e-05, "loss": 0.0394, "step": 3213 }, { "epoch": 0.348438855160451, "grad_norm": 3.471628427505493, "learning_rate": 1.7741186711156246e-05, "loss": 0.0635, "step": 3214 }, { "epoch": 0.3485472679965308, "grad_norm": 2.3375957012176514, "learning_rate": 1.7740461337588858e-05, "loss": 0.0773, "step": 3215 }, { "epoch": 0.34865568083261056, "grad_norm": 1.604623794555664, "learning_rate": 1.7739735964021473e-05, "loss": 0.0543, "step": 3216 }, { "epoch": 0.3487640936686904, "grad_norm": 3.5140979290008545, "learning_rate": 1.7739010590454085e-05, "loss": 0.0552, "step": 3217 }, { "epoch": 0.34887250650477014, "grad_norm": 2.3941264152526855, "learning_rate": 1.7738285216886698e-05, "loss": 0.0777, "step": 3218 }, { "epoch": 0.34898091934084996, "grad_norm": 2.8917527198791504, "learning_rate": 1.7737559843319313e-05, "loss": 0.0968, "step": 3219 }, { "epoch": 0.3490893321769297, "grad_norm": 2.3708534240722656, "learning_rate": 1.7736834469751925e-05, "loss": 0.0497, "step": 3220 }, { "epoch": 0.34919774501300954, "grad_norm": 2.4802052974700928, "learning_rate": 1.7736109096184537e-05, "loss": 0.0251, "step": 3221 }, { "epoch": 0.3493061578490893, "grad_norm": 0.6043279767036438, "learning_rate": 1.773538372261715e-05, "loss": 0.0228, "step": 3222 }, { "epoch": 0.3494145706851691, "grad_norm": 3.064588785171509, "learning_rate": 1.7734658349049762e-05, "loss": 0.0381, "step": 3223 }, { "epoch": 0.34952298352124894, "grad_norm": 1.3435554504394531, "learning_rate": 1.7733932975482374e-05, "loss": 0.0373, "step": 3224 }, { "epoch": 0.3496313963573287, "grad_norm": 1.5929911136627197, "learning_rate": 1.773320760191499e-05, "loss": 0.059, "step": 3225 }, { "epoch": 0.3497398091934085, "grad_norm": 3.345524549484253, "learning_rate": 1.77324822283476e-05, "loss": 0.1085, "step": 3226 }, { "epoch": 0.3498482220294883, "grad_norm": 1.9083253145217896, "learning_rate": 1.7731756854780214e-05, "loss": 0.0372, "step": 3227 }, { "epoch": 0.3499566348655681, "grad_norm": 2.0111114978790283, "learning_rate": 1.7731031481212826e-05, "loss": 0.0269, "step": 3228 }, { "epoch": 0.35006504770164787, "grad_norm": 2.9608845710754395, "learning_rate": 1.7730306107645438e-05, "loss": 0.0602, "step": 3229 }, { "epoch": 0.3501734605377277, "grad_norm": 2.670720338821411, "learning_rate": 1.772958073407805e-05, "loss": 0.0633, "step": 3230 }, { "epoch": 0.35028187337380745, "grad_norm": 2.717456340789795, "learning_rate": 1.7728855360510662e-05, "loss": 0.0675, "step": 3231 }, { "epoch": 0.35039028620988727, "grad_norm": 3.234962224960327, "learning_rate": 1.7728129986943278e-05, "loss": 0.0304, "step": 3232 }, { "epoch": 0.35049869904596703, "grad_norm": 1.0458017587661743, "learning_rate": 1.772740461337589e-05, "loss": 0.0246, "step": 3233 }, { "epoch": 0.35060711188204685, "grad_norm": 1.4873024225234985, "learning_rate": 1.7726679239808502e-05, "loss": 0.0256, "step": 3234 }, { "epoch": 0.3507155247181266, "grad_norm": 4.585576057434082, "learning_rate": 1.7725953866241114e-05, "loss": 0.0607, "step": 3235 }, { "epoch": 0.35082393755420643, "grad_norm": 5.197254180908203, "learning_rate": 1.772522849267373e-05, "loss": 0.0974, "step": 3236 }, { "epoch": 0.3509323503902862, "grad_norm": 3.186206817626953, "learning_rate": 1.7724503119106342e-05, "loss": 0.0423, "step": 3237 }, { "epoch": 0.351040763226366, "grad_norm": 0.952198326587677, "learning_rate": 1.7723777745538954e-05, "loss": 0.0327, "step": 3238 }, { "epoch": 0.3511491760624458, "grad_norm": 2.8038110733032227, "learning_rate": 1.7723052371971566e-05, "loss": 0.0615, "step": 3239 }, { "epoch": 0.3512575888985256, "grad_norm": 2.1449766159057617, "learning_rate": 1.7722326998404182e-05, "loss": 0.0663, "step": 3240 }, { "epoch": 0.35136600173460536, "grad_norm": 2.039233446121216, "learning_rate": 1.7721601624836794e-05, "loss": 0.0406, "step": 3241 }, { "epoch": 0.3514744145706852, "grad_norm": 4.14131498336792, "learning_rate": 1.7720876251269406e-05, "loss": 0.0607, "step": 3242 }, { "epoch": 0.35158282740676494, "grad_norm": 1.7590663433074951, "learning_rate": 1.7720150877702018e-05, "loss": 0.0307, "step": 3243 }, { "epoch": 0.35169124024284476, "grad_norm": 1.1587936878204346, "learning_rate": 1.771942550413463e-05, "loss": 0.0462, "step": 3244 }, { "epoch": 0.3517996530789245, "grad_norm": 2.9776289463043213, "learning_rate": 1.7718700130567242e-05, "loss": 0.0286, "step": 3245 }, { "epoch": 0.35190806591500434, "grad_norm": 4.577630519866943, "learning_rate": 1.7717974756999855e-05, "loss": 0.0908, "step": 3246 }, { "epoch": 0.3520164787510841, "grad_norm": 1.6461305618286133, "learning_rate": 1.771724938343247e-05, "loss": 0.0159, "step": 3247 }, { "epoch": 0.3521248915871639, "grad_norm": 0.717430055141449, "learning_rate": 1.7716524009865082e-05, "loss": 0.008, "step": 3248 }, { "epoch": 0.3522333044232437, "grad_norm": 5.616057395935059, "learning_rate": 1.7715798636297694e-05, "loss": 0.092, "step": 3249 }, { "epoch": 0.3523417172593235, "grad_norm": 1.4067966938018799, "learning_rate": 1.7715073262730307e-05, "loss": 0.0351, "step": 3250 }, { "epoch": 0.35245013009540327, "grad_norm": 0.851383626461029, "learning_rate": 1.771434788916292e-05, "loss": 0.0137, "step": 3251 }, { "epoch": 0.3525585429314831, "grad_norm": 0.6894487738609314, "learning_rate": 1.771362251559553e-05, "loss": 0.0186, "step": 3252 }, { "epoch": 0.3526669557675629, "grad_norm": 2.707054853439331, "learning_rate": 1.7712897142028146e-05, "loss": 0.0098, "step": 3253 }, { "epoch": 0.35277536860364267, "grad_norm": 2.342886209487915, "learning_rate": 1.771217176846076e-05, "loss": 0.01, "step": 3254 }, { "epoch": 0.3528837814397225, "grad_norm": 4.150470733642578, "learning_rate": 1.7711446394893374e-05, "loss": 0.0973, "step": 3255 }, { "epoch": 0.35299219427580225, "grad_norm": 2.444720983505249, "learning_rate": 1.7710721021325986e-05, "loss": 0.0222, "step": 3256 }, { "epoch": 0.35310060711188207, "grad_norm": 0.7244285345077515, "learning_rate": 1.7709995647758598e-05, "loss": 0.0119, "step": 3257 }, { "epoch": 0.35320901994796183, "grad_norm": 4.402153015136719, "learning_rate": 1.770927027419121e-05, "loss": 0.0355, "step": 3258 }, { "epoch": 0.35331743278404165, "grad_norm": 0.4532220661640167, "learning_rate": 1.7708544900623823e-05, "loss": 0.0036, "step": 3259 }, { "epoch": 0.3534258456201214, "grad_norm": 2.345306158065796, "learning_rate": 1.7707819527056435e-05, "loss": 0.0362, "step": 3260 }, { "epoch": 0.35353425845620123, "grad_norm": 0.4975920617580414, "learning_rate": 1.7707094153489047e-05, "loss": 0.004, "step": 3261 }, { "epoch": 0.353642671292281, "grad_norm": 3.190009832382202, "learning_rate": 1.7706368779921662e-05, "loss": 0.0646, "step": 3262 }, { "epoch": 0.3537510841283608, "grad_norm": 8.976751327514648, "learning_rate": 1.7705643406354275e-05, "loss": 0.0858, "step": 3263 }, { "epoch": 0.3538594969644406, "grad_norm": 0.5269278883934021, "learning_rate": 1.7704918032786887e-05, "loss": 0.0028, "step": 3264 }, { "epoch": 0.3539679098005204, "grad_norm": 4.739458084106445, "learning_rate": 1.77041926592195e-05, "loss": 0.0564, "step": 3265 }, { "epoch": 0.35407632263660016, "grad_norm": 3.0379810333251953, "learning_rate": 1.770346728565211e-05, "loss": 0.0399, "step": 3266 }, { "epoch": 0.35418473547268, "grad_norm": 7.024932384490967, "learning_rate": 1.7702741912084723e-05, "loss": 0.0828, "step": 3267 }, { "epoch": 0.35429314830875974, "grad_norm": 2.84822416305542, "learning_rate": 1.770201653851734e-05, "loss": 0.0573, "step": 3268 }, { "epoch": 0.35440156114483956, "grad_norm": 2.6892435550689697, "learning_rate": 1.770129116494995e-05, "loss": 0.0591, "step": 3269 }, { "epoch": 0.3545099739809193, "grad_norm": 0.6702439785003662, "learning_rate": 1.7700565791382563e-05, "loss": 0.0171, "step": 3270 }, { "epoch": 0.35461838681699914, "grad_norm": 0.3047809600830078, "learning_rate": 1.769984041781518e-05, "loss": 0.0036, "step": 3271 }, { "epoch": 0.3547267996530789, "grad_norm": 5.431680202484131, "learning_rate": 1.769911504424779e-05, "loss": 0.0273, "step": 3272 }, { "epoch": 0.3548352124891587, "grad_norm": 2.5593533515930176, "learning_rate": 1.7698389670680403e-05, "loss": 0.032, "step": 3273 }, { "epoch": 0.3549436253252385, "grad_norm": 1.8732810020446777, "learning_rate": 1.7697664297113015e-05, "loss": 0.0378, "step": 3274 }, { "epoch": 0.3550520381613183, "grad_norm": 2.2883617877960205, "learning_rate": 1.7696938923545627e-05, "loss": 0.0247, "step": 3275 }, { "epoch": 0.35516045099739807, "grad_norm": 4.123654365539551, "learning_rate": 1.769621354997824e-05, "loss": 0.0465, "step": 3276 }, { "epoch": 0.3552688638334779, "grad_norm": 3.553319215774536, "learning_rate": 1.7695488176410855e-05, "loss": 0.0757, "step": 3277 }, { "epoch": 0.35537727666955765, "grad_norm": 7.7781548500061035, "learning_rate": 1.7694762802843467e-05, "loss": 0.0805, "step": 3278 }, { "epoch": 0.35548568950563747, "grad_norm": 4.610794544219971, "learning_rate": 1.769403742927608e-05, "loss": 0.0663, "step": 3279 }, { "epoch": 0.35559410234171723, "grad_norm": 2.8295481204986572, "learning_rate": 1.769331205570869e-05, "loss": 0.0829, "step": 3280 }, { "epoch": 0.35570251517779705, "grad_norm": 4.948297023773193, "learning_rate": 1.7692586682141303e-05, "loss": 0.0426, "step": 3281 }, { "epoch": 0.35581092801387687, "grad_norm": 3.628624200820923, "learning_rate": 1.7691861308573915e-05, "loss": 0.0328, "step": 3282 }, { "epoch": 0.35591934084995663, "grad_norm": 6.1811699867248535, "learning_rate": 1.7691135935006528e-05, "loss": 0.0297, "step": 3283 }, { "epoch": 0.35602775368603645, "grad_norm": 0.882335901260376, "learning_rate": 1.7690410561439143e-05, "loss": 0.0063, "step": 3284 }, { "epoch": 0.3561361665221162, "grad_norm": 8.311767578125, "learning_rate": 1.7689685187871755e-05, "loss": 0.0227, "step": 3285 }, { "epoch": 0.35624457935819603, "grad_norm": 3.1642909049987793, "learning_rate": 1.7688959814304367e-05, "loss": 0.0347, "step": 3286 }, { "epoch": 0.3563529921942758, "grad_norm": 3.641615867614746, "learning_rate": 1.768823444073698e-05, "loss": 0.0577, "step": 3287 }, { "epoch": 0.3564614050303556, "grad_norm": 3.1506192684173584, "learning_rate": 1.7687509067169595e-05, "loss": 0.0315, "step": 3288 }, { "epoch": 0.3565698178664354, "grad_norm": 2.2510809898376465, "learning_rate": 1.7686783693602207e-05, "loss": 0.0368, "step": 3289 }, { "epoch": 0.3566782307025152, "grad_norm": 1.6198867559432983, "learning_rate": 1.768605832003482e-05, "loss": 0.0415, "step": 3290 }, { "epoch": 0.35678664353859496, "grad_norm": 2.4053444862365723, "learning_rate": 1.768533294646743e-05, "loss": 0.0438, "step": 3291 }, { "epoch": 0.3568950563746748, "grad_norm": 1.903196930885315, "learning_rate": 1.7684607572900047e-05, "loss": 0.0504, "step": 3292 }, { "epoch": 0.35700346921075454, "grad_norm": 2.7287423610687256, "learning_rate": 1.768388219933266e-05, "loss": 0.0483, "step": 3293 }, { "epoch": 0.35711188204683436, "grad_norm": 2.5059397220611572, "learning_rate": 1.768315682576527e-05, "loss": 0.032, "step": 3294 }, { "epoch": 0.3572202948829141, "grad_norm": 2.5512938499450684, "learning_rate": 1.7682431452197883e-05, "loss": 0.0491, "step": 3295 }, { "epoch": 0.35732870771899394, "grad_norm": 2.671139717102051, "learning_rate": 1.7681706078630496e-05, "loss": 0.0486, "step": 3296 }, { "epoch": 0.3574371205550737, "grad_norm": 1.1943166255950928, "learning_rate": 1.7680980705063108e-05, "loss": 0.02, "step": 3297 }, { "epoch": 0.3575455333911535, "grad_norm": 4.542710304260254, "learning_rate": 1.768025533149572e-05, "loss": 0.0926, "step": 3298 }, { "epoch": 0.3576539462272333, "grad_norm": 3.784499406814575, "learning_rate": 1.7679529957928335e-05, "loss": 0.0783, "step": 3299 }, { "epoch": 0.3577623590633131, "grad_norm": 5.7878828048706055, "learning_rate": 1.7678804584360948e-05, "loss": 0.0547, "step": 3300 }, { "epoch": 0.35787077189939287, "grad_norm": 2.303004741668701, "learning_rate": 1.767807921079356e-05, "loss": 0.0458, "step": 3301 }, { "epoch": 0.3579791847354727, "grad_norm": 2.533792495727539, "learning_rate": 1.7677353837226172e-05, "loss": 0.0546, "step": 3302 }, { "epoch": 0.35808759757155245, "grad_norm": 1.8419843912124634, "learning_rate": 1.7676628463658784e-05, "loss": 0.0157, "step": 3303 }, { "epoch": 0.35819601040763227, "grad_norm": 0.7950208187103271, "learning_rate": 1.7675903090091396e-05, "loss": 0.018, "step": 3304 }, { "epoch": 0.35830442324371203, "grad_norm": 2.1880581378936768, "learning_rate": 1.767517771652401e-05, "loss": 0.0255, "step": 3305 }, { "epoch": 0.35841283607979185, "grad_norm": 2.803354501724243, "learning_rate": 1.7674452342956624e-05, "loss": 0.0211, "step": 3306 }, { "epoch": 0.3585212489158716, "grad_norm": 2.201517105102539, "learning_rate": 1.767372696938924e-05, "loss": 0.0352, "step": 3307 }, { "epoch": 0.35862966175195143, "grad_norm": 3.1654436588287354, "learning_rate": 1.767300159582185e-05, "loss": 0.0386, "step": 3308 }, { "epoch": 0.3587380745880312, "grad_norm": 6.467401504516602, "learning_rate": 1.7672276222254464e-05, "loss": 0.0773, "step": 3309 }, { "epoch": 0.358846487424111, "grad_norm": 2.280102252960205, "learning_rate": 1.7671550848687076e-05, "loss": 0.0515, "step": 3310 }, { "epoch": 0.3589549002601908, "grad_norm": 3.0142927169799805, "learning_rate": 1.7670825475119688e-05, "loss": 0.0392, "step": 3311 }, { "epoch": 0.3590633130962706, "grad_norm": 2.7024664878845215, "learning_rate": 1.76701001015523e-05, "loss": 0.0714, "step": 3312 }, { "epoch": 0.3591717259323504, "grad_norm": 1.1937421560287476, "learning_rate": 1.7669374727984912e-05, "loss": 0.0197, "step": 3313 }, { "epoch": 0.3592801387684302, "grad_norm": 3.0768232345581055, "learning_rate": 1.7668649354417528e-05, "loss": 0.0207, "step": 3314 }, { "epoch": 0.35938855160451, "grad_norm": 3.578242063522339, "learning_rate": 1.766792398085014e-05, "loss": 0.0579, "step": 3315 }, { "epoch": 0.35949696444058976, "grad_norm": 3.0481948852539062, "learning_rate": 1.7667198607282752e-05, "loss": 0.077, "step": 3316 }, { "epoch": 0.3596053772766696, "grad_norm": 2.8738510608673096, "learning_rate": 1.7666473233715364e-05, "loss": 0.049, "step": 3317 }, { "epoch": 0.35971379011274934, "grad_norm": 1.9442898035049438, "learning_rate": 1.7665747860147976e-05, "loss": 0.0121, "step": 3318 }, { "epoch": 0.35982220294882916, "grad_norm": 3.5700793266296387, "learning_rate": 1.766502248658059e-05, "loss": 0.0616, "step": 3319 }, { "epoch": 0.3599306157849089, "grad_norm": 2.1438162326812744, "learning_rate": 1.7664297113013204e-05, "loss": 0.0449, "step": 3320 }, { "epoch": 0.36003902862098874, "grad_norm": 1.1622511148452759, "learning_rate": 1.7663571739445816e-05, "loss": 0.0291, "step": 3321 }, { "epoch": 0.3601474414570685, "grad_norm": 0.661365807056427, "learning_rate": 1.7662846365878428e-05, "loss": 0.0118, "step": 3322 }, { "epoch": 0.3602558542931483, "grad_norm": 1.432806372642517, "learning_rate": 1.7662120992311044e-05, "loss": 0.0351, "step": 3323 }, { "epoch": 0.3603642671292281, "grad_norm": 5.7428178787231445, "learning_rate": 1.7661395618743656e-05, "loss": 0.0614, "step": 3324 }, { "epoch": 0.3604726799653079, "grad_norm": 5.162618160247803, "learning_rate": 1.7660670245176268e-05, "loss": 0.0435, "step": 3325 }, { "epoch": 0.36058109280138767, "grad_norm": 2.800185203552246, "learning_rate": 1.765994487160888e-05, "loss": 0.0784, "step": 3326 }, { "epoch": 0.3606895056374675, "grad_norm": 1.5858627557754517, "learning_rate": 1.7659219498041492e-05, "loss": 0.0518, "step": 3327 }, { "epoch": 0.36079791847354725, "grad_norm": 2.859656810760498, "learning_rate": 1.7658494124474105e-05, "loss": 0.0368, "step": 3328 }, { "epoch": 0.36090633130962707, "grad_norm": 1.6132142543792725, "learning_rate": 1.765776875090672e-05, "loss": 0.0345, "step": 3329 }, { "epoch": 0.36101474414570683, "grad_norm": 1.4949593544006348, "learning_rate": 1.7657043377339332e-05, "loss": 0.0173, "step": 3330 }, { "epoch": 0.36112315698178665, "grad_norm": 1.9506267309188843, "learning_rate": 1.7656318003771944e-05, "loss": 0.0599, "step": 3331 }, { "epoch": 0.3612315698178664, "grad_norm": 1.910437822341919, "learning_rate": 1.7655592630204556e-05, "loss": 0.0284, "step": 3332 }, { "epoch": 0.36133998265394623, "grad_norm": 4.14259147644043, "learning_rate": 1.765486725663717e-05, "loss": 0.1037, "step": 3333 }, { "epoch": 0.361448395490026, "grad_norm": 4.60854434967041, "learning_rate": 1.765414188306978e-05, "loss": 0.0674, "step": 3334 }, { "epoch": 0.3615568083261058, "grad_norm": 3.3776915073394775, "learning_rate": 1.7653416509502393e-05, "loss": 0.0558, "step": 3335 }, { "epoch": 0.3616652211621856, "grad_norm": 1.39116370677948, "learning_rate": 1.765269113593501e-05, "loss": 0.0238, "step": 3336 }, { "epoch": 0.3617736339982654, "grad_norm": 1.446150302886963, "learning_rate": 1.765196576236762e-05, "loss": 0.0214, "step": 3337 }, { "epoch": 0.36188204683434516, "grad_norm": 2.7646102905273438, "learning_rate": 1.7651240388800233e-05, "loss": 0.0539, "step": 3338 }, { "epoch": 0.361990459670425, "grad_norm": 1.0258193016052246, "learning_rate": 1.7650515015232845e-05, "loss": 0.0348, "step": 3339 }, { "epoch": 0.36209887250650474, "grad_norm": 2.4567112922668457, "learning_rate": 1.764978964166546e-05, "loss": 0.0635, "step": 3340 }, { "epoch": 0.36220728534258456, "grad_norm": 3.579202175140381, "learning_rate": 1.7649064268098073e-05, "loss": 0.0506, "step": 3341 }, { "epoch": 0.3623156981786644, "grad_norm": 2.1008174419403076, "learning_rate": 1.7648338894530685e-05, "loss": 0.035, "step": 3342 }, { "epoch": 0.36242411101474414, "grad_norm": 2.618886709213257, "learning_rate": 1.7647613520963297e-05, "loss": 0.0497, "step": 3343 }, { "epoch": 0.36253252385082396, "grad_norm": 0.9644128680229187, "learning_rate": 1.7646888147395912e-05, "loss": 0.0163, "step": 3344 }, { "epoch": 0.3626409366869037, "grad_norm": 3.6457526683807373, "learning_rate": 1.7646162773828525e-05, "loss": 0.0551, "step": 3345 }, { "epoch": 0.36274934952298354, "grad_norm": 0.7459282279014587, "learning_rate": 1.7645437400261137e-05, "loss": 0.0128, "step": 3346 }, { "epoch": 0.3628577623590633, "grad_norm": 2.3739571571350098, "learning_rate": 1.764471202669375e-05, "loss": 0.0657, "step": 3347 }, { "epoch": 0.3629661751951431, "grad_norm": 2.889049768447876, "learning_rate": 1.764398665312636e-05, "loss": 0.0436, "step": 3348 }, { "epoch": 0.3630745880312229, "grad_norm": 5.007400989532471, "learning_rate": 1.7643261279558973e-05, "loss": 0.0514, "step": 3349 }, { "epoch": 0.3631830008673027, "grad_norm": 5.492849349975586, "learning_rate": 1.7642535905991585e-05, "loss": 0.0838, "step": 3350 }, { "epoch": 0.36329141370338247, "grad_norm": 3.422548532485962, "learning_rate": 1.76418105324242e-05, "loss": 0.068, "step": 3351 }, { "epoch": 0.3633998265394623, "grad_norm": 2.188232660293579, "learning_rate": 1.7641085158856813e-05, "loss": 0.0447, "step": 3352 }, { "epoch": 0.36350823937554205, "grad_norm": 2.1707777976989746, "learning_rate": 1.7640359785289425e-05, "loss": 0.0376, "step": 3353 }, { "epoch": 0.36361665221162187, "grad_norm": 2.872450113296509, "learning_rate": 1.7639634411722037e-05, "loss": 0.0216, "step": 3354 }, { "epoch": 0.36372506504770163, "grad_norm": 1.8703666925430298, "learning_rate": 1.763890903815465e-05, "loss": 0.0229, "step": 3355 }, { "epoch": 0.36383347788378145, "grad_norm": 2.1233551502227783, "learning_rate": 1.763818366458726e-05, "loss": 0.023, "step": 3356 }, { "epoch": 0.3639418907198612, "grad_norm": 1.7956126928329468, "learning_rate": 1.7637458291019877e-05, "loss": 0.0497, "step": 3357 }, { "epoch": 0.36405030355594103, "grad_norm": 0.5321789979934692, "learning_rate": 1.763673291745249e-05, "loss": 0.0049, "step": 3358 }, { "epoch": 0.3641587163920208, "grad_norm": 2.6235382556915283, "learning_rate": 1.7636007543885105e-05, "loss": 0.0373, "step": 3359 }, { "epoch": 0.3642671292281006, "grad_norm": 4.36507511138916, "learning_rate": 1.7635282170317717e-05, "loss": 0.0263, "step": 3360 }, { "epoch": 0.3643755420641804, "grad_norm": 3.0747878551483154, "learning_rate": 1.763455679675033e-05, "loss": 0.0531, "step": 3361 }, { "epoch": 0.3644839549002602, "grad_norm": 3.484689950942993, "learning_rate": 1.763383142318294e-05, "loss": 0.038, "step": 3362 }, { "epoch": 0.36459236773633996, "grad_norm": 1.4419519901275635, "learning_rate": 1.7633106049615553e-05, "loss": 0.03, "step": 3363 }, { "epoch": 0.3647007805724198, "grad_norm": 7.000133514404297, "learning_rate": 1.7632380676048165e-05, "loss": 0.0449, "step": 3364 }, { "epoch": 0.36480919340849954, "grad_norm": 1.5447324514389038, "learning_rate": 1.7631655302480778e-05, "loss": 0.032, "step": 3365 }, { "epoch": 0.36491760624457936, "grad_norm": 1.361966848373413, "learning_rate": 1.7630929928913393e-05, "loss": 0.0208, "step": 3366 }, { "epoch": 0.3650260190806591, "grad_norm": 2.2185699939727783, "learning_rate": 1.7630204555346005e-05, "loss": 0.0232, "step": 3367 }, { "epoch": 0.36513443191673894, "grad_norm": 3.060534954071045, "learning_rate": 1.7629479181778617e-05, "loss": 0.0413, "step": 3368 }, { "epoch": 0.3652428447528187, "grad_norm": 1.017907738685608, "learning_rate": 1.762875380821123e-05, "loss": 0.0103, "step": 3369 }, { "epoch": 0.3653512575888985, "grad_norm": 4.19007682800293, "learning_rate": 1.762802843464384e-05, "loss": 0.0651, "step": 3370 }, { "epoch": 0.36545967042497834, "grad_norm": 4.1364922523498535, "learning_rate": 1.7627303061076454e-05, "loss": 0.0642, "step": 3371 }, { "epoch": 0.3655680832610581, "grad_norm": 2.3592777252197266, "learning_rate": 1.762657768750907e-05, "loss": 0.02, "step": 3372 }, { "epoch": 0.3656764960971379, "grad_norm": 2.6206963062286377, "learning_rate": 1.762585231394168e-05, "loss": 0.0725, "step": 3373 }, { "epoch": 0.3657849089332177, "grad_norm": 1.1906983852386475, "learning_rate": 1.7625126940374294e-05, "loss": 0.0147, "step": 3374 }, { "epoch": 0.3658933217692975, "grad_norm": 4.941317081451416, "learning_rate": 1.762440156680691e-05, "loss": 0.0637, "step": 3375 }, { "epoch": 0.36600173460537727, "grad_norm": 2.506115198135376, "learning_rate": 1.762367619323952e-05, "loss": 0.0494, "step": 3376 }, { "epoch": 0.3661101474414571, "grad_norm": 2.278244733810425, "learning_rate": 1.7622950819672133e-05, "loss": 0.0149, "step": 3377 }, { "epoch": 0.36621856027753685, "grad_norm": 1.3575491905212402, "learning_rate": 1.7622225446104746e-05, "loss": 0.0116, "step": 3378 }, { "epoch": 0.36632697311361667, "grad_norm": 3.987037181854248, "learning_rate": 1.7621500072537358e-05, "loss": 0.0375, "step": 3379 }, { "epoch": 0.36643538594969643, "grad_norm": 2.6809182167053223, "learning_rate": 1.762077469896997e-05, "loss": 0.0523, "step": 3380 }, { "epoch": 0.36654379878577625, "grad_norm": 5.481452941894531, "learning_rate": 1.7620049325402585e-05, "loss": 0.1273, "step": 3381 }, { "epoch": 0.366652211621856, "grad_norm": 2.6789588928222656, "learning_rate": 1.7619323951835198e-05, "loss": 0.0492, "step": 3382 }, { "epoch": 0.36676062445793584, "grad_norm": 2.985863208770752, "learning_rate": 1.761859857826781e-05, "loss": 0.0325, "step": 3383 }, { "epoch": 0.3668690372940156, "grad_norm": 1.2386305332183838, "learning_rate": 1.7617873204700422e-05, "loss": 0.0146, "step": 3384 }, { "epoch": 0.3669774501300954, "grad_norm": 3.019852638244629, "learning_rate": 1.7617147831133034e-05, "loss": 0.0241, "step": 3385 }, { "epoch": 0.3670858629661752, "grad_norm": 0.5224682688713074, "learning_rate": 1.7616422457565646e-05, "loss": 0.0052, "step": 3386 }, { "epoch": 0.367194275802255, "grad_norm": 1.4883230924606323, "learning_rate": 1.7615697083998258e-05, "loss": 0.0109, "step": 3387 }, { "epoch": 0.36730268863833476, "grad_norm": 1.9835131168365479, "learning_rate": 1.7614971710430874e-05, "loss": 0.0243, "step": 3388 }, { "epoch": 0.3674111014744146, "grad_norm": 7.543875694274902, "learning_rate": 1.7614246336863486e-05, "loss": 0.0457, "step": 3389 }, { "epoch": 0.36751951431049434, "grad_norm": 4.5190300941467285, "learning_rate": 1.7613520963296098e-05, "loss": 0.0371, "step": 3390 }, { "epoch": 0.36762792714657416, "grad_norm": 2.3073880672454834, "learning_rate": 1.761279558972871e-05, "loss": 0.0442, "step": 3391 }, { "epoch": 0.3677363399826539, "grad_norm": 3.573040246963501, "learning_rate": 1.7612070216161326e-05, "loss": 0.0357, "step": 3392 }, { "epoch": 0.36784475281873374, "grad_norm": 2.706681489944458, "learning_rate": 1.7611344842593938e-05, "loss": 0.0368, "step": 3393 }, { "epoch": 0.3679531656548135, "grad_norm": 4.113929748535156, "learning_rate": 1.761061946902655e-05, "loss": 0.0247, "step": 3394 }, { "epoch": 0.3680615784908933, "grad_norm": 2.8712544441223145, "learning_rate": 1.7609894095459162e-05, "loss": 0.0443, "step": 3395 }, { "epoch": 0.3681699913269731, "grad_norm": 8.744576454162598, "learning_rate": 1.7609168721891778e-05, "loss": 0.0513, "step": 3396 }, { "epoch": 0.3682784041630529, "grad_norm": 0.9867718815803528, "learning_rate": 1.760844334832439e-05, "loss": 0.0156, "step": 3397 }, { "epoch": 0.36838681699913267, "grad_norm": 4.3575897216796875, "learning_rate": 1.7607717974757002e-05, "loss": 0.0638, "step": 3398 }, { "epoch": 0.3684952298352125, "grad_norm": 1.698178768157959, "learning_rate": 1.7606992601189614e-05, "loss": 0.0165, "step": 3399 }, { "epoch": 0.3686036426712923, "grad_norm": 4.550554275512695, "learning_rate": 1.7606267227622226e-05, "loss": 0.0497, "step": 3400 }, { "epoch": 0.36871205550737207, "grad_norm": 5.460394859313965, "learning_rate": 1.760554185405484e-05, "loss": 0.0175, "step": 3401 }, { "epoch": 0.3688204683434519, "grad_norm": 2.8686654567718506, "learning_rate": 1.760481648048745e-05, "loss": 0.0242, "step": 3402 }, { "epoch": 0.36892888117953165, "grad_norm": 2.0799262523651123, "learning_rate": 1.7604091106920066e-05, "loss": 0.0215, "step": 3403 }, { "epoch": 0.3690372940156115, "grad_norm": 2.959202527999878, "learning_rate": 1.7603365733352678e-05, "loss": 0.017, "step": 3404 }, { "epoch": 0.36914570685169124, "grad_norm": 0.7163510322570801, "learning_rate": 1.760264035978529e-05, "loss": 0.0124, "step": 3405 }, { "epoch": 0.36925411968777105, "grad_norm": 1.9483075141906738, "learning_rate": 1.7601914986217903e-05, "loss": 0.0191, "step": 3406 }, { "epoch": 0.3693625325238508, "grad_norm": 4.83601713180542, "learning_rate": 1.7601189612650515e-05, "loss": 0.0776, "step": 3407 }, { "epoch": 0.36947094535993064, "grad_norm": 1.6995965242385864, "learning_rate": 1.7600464239083127e-05, "loss": 0.0151, "step": 3408 }, { "epoch": 0.3695793581960104, "grad_norm": 7.437129020690918, "learning_rate": 1.7599738865515742e-05, "loss": 0.0597, "step": 3409 }, { "epoch": 0.3696877710320902, "grad_norm": 3.1194441318511963, "learning_rate": 1.7599013491948354e-05, "loss": 0.0452, "step": 3410 }, { "epoch": 0.36979618386817, "grad_norm": 1.2345818281173706, "learning_rate": 1.759828811838097e-05, "loss": 0.0173, "step": 3411 }, { "epoch": 0.3699045967042498, "grad_norm": 1.5914753675460815, "learning_rate": 1.7597562744813582e-05, "loss": 0.0361, "step": 3412 }, { "epoch": 0.37001300954032956, "grad_norm": 0.6101332306861877, "learning_rate": 1.7596837371246194e-05, "loss": 0.0119, "step": 3413 }, { "epoch": 0.3701214223764094, "grad_norm": 7.855734348297119, "learning_rate": 1.7596111997678806e-05, "loss": 0.0777, "step": 3414 }, { "epoch": 0.37022983521248914, "grad_norm": 3.245562791824341, "learning_rate": 1.759538662411142e-05, "loss": 0.0417, "step": 3415 }, { "epoch": 0.37033824804856896, "grad_norm": 5.694042205810547, "learning_rate": 1.759466125054403e-05, "loss": 0.0325, "step": 3416 }, { "epoch": 0.3704466608846487, "grad_norm": 1.1557406187057495, "learning_rate": 1.7593935876976643e-05, "loss": 0.0456, "step": 3417 }, { "epoch": 0.37055507372072854, "grad_norm": 8.39118766784668, "learning_rate": 1.759321050340926e-05, "loss": 0.0521, "step": 3418 }, { "epoch": 0.3706634865568083, "grad_norm": 7.496553897857666, "learning_rate": 1.759248512984187e-05, "loss": 0.0928, "step": 3419 }, { "epoch": 0.3707718993928881, "grad_norm": 2.3796212673187256, "learning_rate": 1.7591759756274483e-05, "loss": 0.0246, "step": 3420 }, { "epoch": 0.3708803122289679, "grad_norm": 3.212688684463501, "learning_rate": 1.7591034382707095e-05, "loss": 0.0255, "step": 3421 }, { "epoch": 0.3709887250650477, "grad_norm": 3.033801555633545, "learning_rate": 1.7590309009139707e-05, "loss": 0.0332, "step": 3422 }, { "epoch": 0.37109713790112747, "grad_norm": 2.819211959838867, "learning_rate": 1.758958363557232e-05, "loss": 0.0336, "step": 3423 }, { "epoch": 0.3712055507372073, "grad_norm": 3.824970006942749, "learning_rate": 1.7588858262004935e-05, "loss": 0.063, "step": 3424 }, { "epoch": 0.37131396357328705, "grad_norm": 4.31298303604126, "learning_rate": 1.7588132888437547e-05, "loss": 0.0442, "step": 3425 }, { "epoch": 0.3714223764093669, "grad_norm": 14.224492073059082, "learning_rate": 1.758740751487016e-05, "loss": 0.0591, "step": 3426 }, { "epoch": 0.37153078924544664, "grad_norm": 2.2156002521514893, "learning_rate": 1.7586682141302774e-05, "loss": 0.0397, "step": 3427 }, { "epoch": 0.37163920208152645, "grad_norm": 0.4597669243812561, "learning_rate": 1.7585956767735387e-05, "loss": 0.0122, "step": 3428 }, { "epoch": 0.3717476149176062, "grad_norm": 3.4874839782714844, "learning_rate": 1.7585231394168e-05, "loss": 0.094, "step": 3429 }, { "epoch": 0.37185602775368604, "grad_norm": 3.401393175125122, "learning_rate": 1.758450602060061e-05, "loss": 0.0519, "step": 3430 }, { "epoch": 0.37196444058976585, "grad_norm": 3.6853294372558594, "learning_rate": 1.7583780647033223e-05, "loss": 0.0475, "step": 3431 }, { "epoch": 0.3720728534258456, "grad_norm": 5.165797233581543, "learning_rate": 1.7583055273465835e-05, "loss": 0.0887, "step": 3432 }, { "epoch": 0.37218126626192544, "grad_norm": 5.541165351867676, "learning_rate": 1.758232989989845e-05, "loss": 0.0311, "step": 3433 }, { "epoch": 0.3722896790980052, "grad_norm": 1.912506341934204, "learning_rate": 1.7581604526331063e-05, "loss": 0.0497, "step": 3434 }, { "epoch": 0.372398091934085, "grad_norm": 2.8901689052581787, "learning_rate": 1.7580879152763675e-05, "loss": 0.0419, "step": 3435 }, { "epoch": 0.3725065047701648, "grad_norm": 3.4940121173858643, "learning_rate": 1.7580153779196287e-05, "loss": 0.0695, "step": 3436 }, { "epoch": 0.3726149176062446, "grad_norm": 4.830105304718018, "learning_rate": 1.75794284056289e-05, "loss": 0.0172, "step": 3437 }, { "epoch": 0.37272333044232436, "grad_norm": 1.2526832818984985, "learning_rate": 1.757870303206151e-05, "loss": 0.0116, "step": 3438 }, { "epoch": 0.3728317432784042, "grad_norm": 1.6309541463851929, "learning_rate": 1.7577977658494127e-05, "loss": 0.0307, "step": 3439 }, { "epoch": 0.37294015611448394, "grad_norm": 4.485750198364258, "learning_rate": 1.757725228492674e-05, "loss": 0.0581, "step": 3440 }, { "epoch": 0.37304856895056376, "grad_norm": 3.233628034591675, "learning_rate": 1.757652691135935e-05, "loss": 0.0525, "step": 3441 }, { "epoch": 0.3731569817866435, "grad_norm": 4.047743320465088, "learning_rate": 1.7575801537791963e-05, "loss": 0.0557, "step": 3442 }, { "epoch": 0.37326539462272335, "grad_norm": 2.382638692855835, "learning_rate": 1.757507616422458e-05, "loss": 0.0317, "step": 3443 }, { "epoch": 0.3733738074588031, "grad_norm": 2.0921671390533447, "learning_rate": 1.757435079065719e-05, "loss": 0.0262, "step": 3444 }, { "epoch": 0.3734822202948829, "grad_norm": 3.653015613555908, "learning_rate": 1.7573625417089803e-05, "loss": 0.0558, "step": 3445 }, { "epoch": 0.3735906331309627, "grad_norm": 1.4862887859344482, "learning_rate": 1.7572900043522415e-05, "loss": 0.0269, "step": 3446 }, { "epoch": 0.3736990459670425, "grad_norm": 2.432004928588867, "learning_rate": 1.7572174669955028e-05, "loss": 0.0332, "step": 3447 }, { "epoch": 0.3738074588031223, "grad_norm": 2.4541001319885254, "learning_rate": 1.7571449296387643e-05, "loss": 0.0489, "step": 3448 }, { "epoch": 0.3739158716392021, "grad_norm": 3.506176710128784, "learning_rate": 1.7570723922820255e-05, "loss": 0.0973, "step": 3449 }, { "epoch": 0.37402428447528185, "grad_norm": 4.286502361297607, "learning_rate": 1.7569998549252867e-05, "loss": 0.0607, "step": 3450 }, { "epoch": 0.3741326973113617, "grad_norm": 2.884718418121338, "learning_rate": 1.756927317568548e-05, "loss": 0.0374, "step": 3451 }, { "epoch": 0.37424111014744144, "grad_norm": 1.1835278272628784, "learning_rate": 1.756854780211809e-05, "loss": 0.0246, "step": 3452 }, { "epoch": 0.37434952298352125, "grad_norm": 1.9533382654190063, "learning_rate": 1.7567822428550704e-05, "loss": 0.0565, "step": 3453 }, { "epoch": 0.374457935819601, "grad_norm": 3.0781447887420654, "learning_rate": 1.7567097054983316e-05, "loss": 0.0564, "step": 3454 }, { "epoch": 0.37456634865568084, "grad_norm": 1.810557246208191, "learning_rate": 1.756637168141593e-05, "loss": 0.0422, "step": 3455 }, { "epoch": 0.3746747614917606, "grad_norm": 1.3026286363601685, "learning_rate": 1.7565646307848544e-05, "loss": 0.0216, "step": 3456 }, { "epoch": 0.3747831743278404, "grad_norm": 1.0719084739685059, "learning_rate": 1.7564920934281156e-05, "loss": 0.0365, "step": 3457 }, { "epoch": 0.3748915871639202, "grad_norm": 1.7010118961334229, "learning_rate": 1.7564195560713768e-05, "loss": 0.0406, "step": 3458 }, { "epoch": 0.375, "grad_norm": 1.417081356048584, "learning_rate": 1.756347018714638e-05, "loss": 0.0267, "step": 3459 }, { "epoch": 0.3751084128360798, "grad_norm": 1.972625970840454, "learning_rate": 1.7562744813578992e-05, "loss": 0.0229, "step": 3460 }, { "epoch": 0.3752168256721596, "grad_norm": 2.058225393295288, "learning_rate": 1.7562019440011608e-05, "loss": 0.0665, "step": 3461 }, { "epoch": 0.3753252385082394, "grad_norm": 1.757866621017456, "learning_rate": 1.756129406644422e-05, "loss": 0.0158, "step": 3462 }, { "epoch": 0.37543365134431916, "grad_norm": 1.57534921169281, "learning_rate": 1.7560568692876835e-05, "loss": 0.0228, "step": 3463 }, { "epoch": 0.375542064180399, "grad_norm": 1.0670969486236572, "learning_rate": 1.7559843319309448e-05, "loss": 0.0186, "step": 3464 }, { "epoch": 0.37565047701647875, "grad_norm": 2.5407779216766357, "learning_rate": 1.755911794574206e-05, "loss": 0.0523, "step": 3465 }, { "epoch": 0.37575888985255856, "grad_norm": 3.856382131576538, "learning_rate": 1.7558392572174672e-05, "loss": 0.0144, "step": 3466 }, { "epoch": 0.3758673026886383, "grad_norm": 1.9638408422470093, "learning_rate": 1.7557667198607284e-05, "loss": 0.0226, "step": 3467 }, { "epoch": 0.37597571552471815, "grad_norm": 1.0486252307891846, "learning_rate": 1.7556941825039896e-05, "loss": 0.0124, "step": 3468 }, { "epoch": 0.3760841283607979, "grad_norm": 6.136703968048096, "learning_rate": 1.7556216451472508e-05, "loss": 0.0565, "step": 3469 }, { "epoch": 0.3761925411968777, "grad_norm": 3.2032861709594727, "learning_rate": 1.7555491077905124e-05, "loss": 0.0556, "step": 3470 }, { "epoch": 0.3763009540329575, "grad_norm": 3.064696788787842, "learning_rate": 1.7554765704337736e-05, "loss": 0.088, "step": 3471 }, { "epoch": 0.3764093668690373, "grad_norm": 4.815760135650635, "learning_rate": 1.7554040330770348e-05, "loss": 0.0404, "step": 3472 }, { "epoch": 0.3765177797051171, "grad_norm": 0.48090916872024536, "learning_rate": 1.755331495720296e-05, "loss": 0.0035, "step": 3473 }, { "epoch": 0.3766261925411969, "grad_norm": 1.2252329587936401, "learning_rate": 1.7552589583635572e-05, "loss": 0.0231, "step": 3474 }, { "epoch": 0.37673460537727665, "grad_norm": 5.747788906097412, "learning_rate": 1.7551864210068184e-05, "loss": 0.0885, "step": 3475 }, { "epoch": 0.3768430182133565, "grad_norm": 5.000027656555176, "learning_rate": 1.75511388365008e-05, "loss": 0.048, "step": 3476 }, { "epoch": 0.37695143104943624, "grad_norm": 4.1354193687438965, "learning_rate": 1.7550413462933412e-05, "loss": 0.0283, "step": 3477 }, { "epoch": 0.37705984388551606, "grad_norm": 2.107814311981201, "learning_rate": 1.7549688089366024e-05, "loss": 0.0238, "step": 3478 }, { "epoch": 0.3771682567215958, "grad_norm": 3.932170867919922, "learning_rate": 1.754896271579864e-05, "loss": 0.0348, "step": 3479 }, { "epoch": 0.37727666955767564, "grad_norm": 3.452792167663574, "learning_rate": 1.7548237342231252e-05, "loss": 0.0393, "step": 3480 }, { "epoch": 0.3773850823937554, "grad_norm": 2.107940673828125, "learning_rate": 1.7547511968663864e-05, "loss": 0.0222, "step": 3481 }, { "epoch": 0.3774934952298352, "grad_norm": 4.36203670501709, "learning_rate": 1.7546786595096476e-05, "loss": 0.0686, "step": 3482 }, { "epoch": 0.377601908065915, "grad_norm": 1.4979244470596313, "learning_rate": 1.754606122152909e-05, "loss": 0.0298, "step": 3483 }, { "epoch": 0.3777103209019948, "grad_norm": 6.032151222229004, "learning_rate": 1.75453358479617e-05, "loss": 0.0423, "step": 3484 }, { "epoch": 0.37781873373807456, "grad_norm": 6.859500885009766, "learning_rate": 1.7544610474394316e-05, "loss": 0.0658, "step": 3485 }, { "epoch": 0.3779271465741544, "grad_norm": 3.257923126220703, "learning_rate": 1.7543885100826928e-05, "loss": 0.0323, "step": 3486 }, { "epoch": 0.37803555941023415, "grad_norm": 2.7848353385925293, "learning_rate": 1.754315972725954e-05, "loss": 0.0525, "step": 3487 }, { "epoch": 0.37814397224631396, "grad_norm": 4.608953475952148, "learning_rate": 1.7542434353692153e-05, "loss": 0.0325, "step": 3488 }, { "epoch": 0.3782523850823938, "grad_norm": 0.8566356897354126, "learning_rate": 1.7541708980124765e-05, "loss": 0.0234, "step": 3489 }, { "epoch": 0.37836079791847355, "grad_norm": 2.0278372764587402, "learning_rate": 1.7540983606557377e-05, "loss": 0.0222, "step": 3490 }, { "epoch": 0.37846921075455336, "grad_norm": 1.3990824222564697, "learning_rate": 1.7540258232989992e-05, "loss": 0.024, "step": 3491 }, { "epoch": 0.3785776235906331, "grad_norm": 3.56154465675354, "learning_rate": 1.7539532859422604e-05, "loss": 0.0521, "step": 3492 }, { "epoch": 0.37868603642671295, "grad_norm": 1.86797034740448, "learning_rate": 1.7538807485855217e-05, "loss": 0.0505, "step": 3493 }, { "epoch": 0.3787944492627927, "grad_norm": 2.860475540161133, "learning_rate": 1.753808211228783e-05, "loss": 0.0556, "step": 3494 }, { "epoch": 0.37890286209887253, "grad_norm": 7.316224575042725, "learning_rate": 1.7537356738720444e-05, "loss": 0.0411, "step": 3495 }, { "epoch": 0.3790112749349523, "grad_norm": 3.780381441116333, "learning_rate": 1.7536631365153056e-05, "loss": 0.0385, "step": 3496 }, { "epoch": 0.3791196877710321, "grad_norm": 1.0707389116287231, "learning_rate": 1.753590599158567e-05, "loss": 0.0319, "step": 3497 }, { "epoch": 0.3792281006071119, "grad_norm": 1.4719951152801514, "learning_rate": 1.753518061801828e-05, "loss": 0.0263, "step": 3498 }, { "epoch": 0.3793365134431917, "grad_norm": 2.321261167526245, "learning_rate": 1.7534455244450893e-05, "loss": 0.0406, "step": 3499 }, { "epoch": 0.37944492627927146, "grad_norm": 0.6048892140388489, "learning_rate": 1.753372987088351e-05, "loss": 0.0141, "step": 3500 }, { "epoch": 0.3795533391153513, "grad_norm": 1.4578313827514648, "learning_rate": 1.753300449731612e-05, "loss": 0.0064, "step": 3501 }, { "epoch": 0.37966175195143104, "grad_norm": 1.8914051055908203, "learning_rate": 1.7532279123748733e-05, "loss": 0.0258, "step": 3502 }, { "epoch": 0.37977016478751086, "grad_norm": 1.6740155220031738, "learning_rate": 1.7531553750181345e-05, "loss": 0.0388, "step": 3503 }, { "epoch": 0.3798785776235906, "grad_norm": 6.90010929107666, "learning_rate": 1.7530828376613957e-05, "loss": 0.0437, "step": 3504 }, { "epoch": 0.37998699045967044, "grad_norm": 0.896290123462677, "learning_rate": 1.753010300304657e-05, "loss": 0.0082, "step": 3505 }, { "epoch": 0.3800954032957502, "grad_norm": 0.9561831951141357, "learning_rate": 1.752937762947918e-05, "loss": 0.0066, "step": 3506 }, { "epoch": 0.38020381613183, "grad_norm": 1.835693359375, "learning_rate": 1.7528652255911797e-05, "loss": 0.0273, "step": 3507 }, { "epoch": 0.3803122289679098, "grad_norm": 0.890972375869751, "learning_rate": 1.752792688234441e-05, "loss": 0.028, "step": 3508 }, { "epoch": 0.3804206418039896, "grad_norm": 1.9419893026351929, "learning_rate": 1.752720150877702e-05, "loss": 0.0489, "step": 3509 }, { "epoch": 0.38052905464006936, "grad_norm": 1.0913569927215576, "learning_rate": 1.7526476135209633e-05, "loss": 0.0199, "step": 3510 }, { "epoch": 0.3806374674761492, "grad_norm": 4.4586663246154785, "learning_rate": 1.7525750761642245e-05, "loss": 0.1144, "step": 3511 }, { "epoch": 0.38074588031222895, "grad_norm": 3.897123098373413, "learning_rate": 1.752502538807486e-05, "loss": 0.0475, "step": 3512 }, { "epoch": 0.38085429314830876, "grad_norm": 4.798062324523926, "learning_rate": 1.7524300014507473e-05, "loss": 0.0307, "step": 3513 }, { "epoch": 0.3809627059843885, "grad_norm": 3.7207179069519043, "learning_rate": 1.7523574640940085e-05, "loss": 0.0516, "step": 3514 }, { "epoch": 0.38107111882046835, "grad_norm": 3.6979994773864746, "learning_rate": 1.75228492673727e-05, "loss": 0.0283, "step": 3515 }, { "epoch": 0.3811795316565481, "grad_norm": 2.7345471382141113, "learning_rate": 1.7522123893805313e-05, "loss": 0.0561, "step": 3516 }, { "epoch": 0.38128794449262793, "grad_norm": 4.720489978790283, "learning_rate": 1.7521398520237925e-05, "loss": 0.0489, "step": 3517 }, { "epoch": 0.3813963573287077, "grad_norm": 1.9793330430984497, "learning_rate": 1.7520673146670537e-05, "loss": 0.0373, "step": 3518 }, { "epoch": 0.3815047701647875, "grad_norm": 0.8877451419830322, "learning_rate": 1.751994777310315e-05, "loss": 0.0273, "step": 3519 }, { "epoch": 0.38161318300086733, "grad_norm": 2.1995978355407715, "learning_rate": 1.751922239953576e-05, "loss": 0.0257, "step": 3520 }, { "epoch": 0.3817215958369471, "grad_norm": 4.759303569793701, "learning_rate": 1.7518497025968374e-05, "loss": 0.0338, "step": 3521 }, { "epoch": 0.3818300086730269, "grad_norm": 2.245784282684326, "learning_rate": 1.751777165240099e-05, "loss": 0.0277, "step": 3522 }, { "epoch": 0.3819384215091067, "grad_norm": 1.0977060794830322, "learning_rate": 1.75170462788336e-05, "loss": 0.0247, "step": 3523 }, { "epoch": 0.3820468343451865, "grad_norm": 1.9526911973953247, "learning_rate": 1.7516320905266213e-05, "loss": 0.0546, "step": 3524 }, { "epoch": 0.38215524718126626, "grad_norm": 0.737189531326294, "learning_rate": 1.7515595531698826e-05, "loss": 0.0109, "step": 3525 }, { "epoch": 0.3822636600173461, "grad_norm": 2.228228807449341, "learning_rate": 1.7514870158131438e-05, "loss": 0.0398, "step": 3526 }, { "epoch": 0.38237207285342584, "grad_norm": 1.93002450466156, "learning_rate": 1.751414478456405e-05, "loss": 0.0167, "step": 3527 }, { "epoch": 0.38248048568950566, "grad_norm": 2.3830554485321045, "learning_rate": 1.7513419410996665e-05, "loss": 0.02, "step": 3528 }, { "epoch": 0.3825888985255854, "grad_norm": 0.7611172199249268, "learning_rate": 1.7512694037429277e-05, "loss": 0.0059, "step": 3529 }, { "epoch": 0.38269731136166524, "grad_norm": 1.8632367849349976, "learning_rate": 1.751196866386189e-05, "loss": 0.0459, "step": 3530 }, { "epoch": 0.382805724197745, "grad_norm": 0.9141674041748047, "learning_rate": 1.7511243290294505e-05, "loss": 0.0168, "step": 3531 }, { "epoch": 0.3829141370338248, "grad_norm": 2.4046335220336914, "learning_rate": 1.7510517916727117e-05, "loss": 0.0282, "step": 3532 }, { "epoch": 0.3830225498699046, "grad_norm": 3.4087324142456055, "learning_rate": 1.750979254315973e-05, "loss": 0.0211, "step": 3533 }, { "epoch": 0.3831309627059844, "grad_norm": 7.665569305419922, "learning_rate": 1.750906716959234e-05, "loss": 0.0863, "step": 3534 }, { "epoch": 0.38323937554206416, "grad_norm": 6.904253959655762, "learning_rate": 1.7508341796024954e-05, "loss": 0.0812, "step": 3535 }, { "epoch": 0.383347788378144, "grad_norm": 1.4852125644683838, "learning_rate": 1.7507616422457566e-05, "loss": 0.0257, "step": 3536 }, { "epoch": 0.38345620121422375, "grad_norm": 8.452753067016602, "learning_rate": 1.750689104889018e-05, "loss": 0.0939, "step": 3537 }, { "epoch": 0.38356461405030357, "grad_norm": 2.6210567951202393, "learning_rate": 1.7506165675322794e-05, "loss": 0.0332, "step": 3538 }, { "epoch": 0.38367302688638333, "grad_norm": 1.7999299764633179, "learning_rate": 1.7505440301755406e-05, "loss": 0.0533, "step": 3539 }, { "epoch": 0.38378143972246315, "grad_norm": 5.831357479095459, "learning_rate": 1.7504714928188018e-05, "loss": 0.0647, "step": 3540 }, { "epoch": 0.3838898525585429, "grad_norm": 2.8048694133758545, "learning_rate": 1.750398955462063e-05, "loss": 0.0631, "step": 3541 }, { "epoch": 0.38399826539462273, "grad_norm": 0.38367998600006104, "learning_rate": 1.7503264181053242e-05, "loss": 0.0033, "step": 3542 }, { "epoch": 0.3841066782307025, "grad_norm": 2.256136894226074, "learning_rate": 1.7502538807485858e-05, "loss": 0.0188, "step": 3543 }, { "epoch": 0.3842150910667823, "grad_norm": 5.000225067138672, "learning_rate": 1.750181343391847e-05, "loss": 0.0416, "step": 3544 }, { "epoch": 0.3843235039028621, "grad_norm": 2.468393087387085, "learning_rate": 1.7501088060351082e-05, "loss": 0.0369, "step": 3545 }, { "epoch": 0.3844319167389419, "grad_norm": 2.851823091506958, "learning_rate": 1.7500362686783694e-05, "loss": 0.0407, "step": 3546 }, { "epoch": 0.38454032957502166, "grad_norm": 4.533046722412109, "learning_rate": 1.749963731321631e-05, "loss": 0.0561, "step": 3547 }, { "epoch": 0.3846487424111015, "grad_norm": 2.4994473457336426, "learning_rate": 1.7498911939648922e-05, "loss": 0.0328, "step": 3548 }, { "epoch": 0.3847571552471813, "grad_norm": 3.755642890930176, "learning_rate": 1.7498186566081534e-05, "loss": 0.0348, "step": 3549 }, { "epoch": 0.38486556808326106, "grad_norm": 4.055126667022705, "learning_rate": 1.7497461192514146e-05, "loss": 0.1045, "step": 3550 }, { "epoch": 0.3849739809193409, "grad_norm": 2.2977395057678223, "learning_rate": 1.7496735818946758e-05, "loss": 0.0244, "step": 3551 }, { "epoch": 0.38508239375542064, "grad_norm": 1.3864392042160034, "learning_rate": 1.7496010445379374e-05, "loss": 0.0169, "step": 3552 }, { "epoch": 0.38519080659150046, "grad_norm": 4.571689128875732, "learning_rate": 1.7495285071811986e-05, "loss": 0.0575, "step": 3553 }, { "epoch": 0.3852992194275802, "grad_norm": 2.5344810485839844, "learning_rate": 1.7494559698244598e-05, "loss": 0.0474, "step": 3554 }, { "epoch": 0.38540763226366004, "grad_norm": 5.505802631378174, "learning_rate": 1.749383432467721e-05, "loss": 0.0935, "step": 3555 }, { "epoch": 0.3855160450997398, "grad_norm": 4.14180850982666, "learning_rate": 1.7493108951109822e-05, "loss": 0.0931, "step": 3556 }, { "epoch": 0.3856244579358196, "grad_norm": 2.3703744411468506, "learning_rate": 1.7492383577542434e-05, "loss": 0.0487, "step": 3557 }, { "epoch": 0.3857328707718994, "grad_norm": 2.34354305267334, "learning_rate": 1.7491658203975047e-05, "loss": 0.0397, "step": 3558 }, { "epoch": 0.3858412836079792, "grad_norm": 5.90390682220459, "learning_rate": 1.7490932830407662e-05, "loss": 0.0552, "step": 3559 }, { "epoch": 0.38594969644405897, "grad_norm": 1.8608276844024658, "learning_rate": 1.7490207456840274e-05, "loss": 0.0237, "step": 3560 }, { "epoch": 0.3860581092801388, "grad_norm": 1.3537784814834595, "learning_rate": 1.7489482083272886e-05, "loss": 0.0252, "step": 3561 }, { "epoch": 0.38616652211621855, "grad_norm": 1.5353225469589233, "learning_rate": 1.74887567097055e-05, "loss": 0.0296, "step": 3562 }, { "epoch": 0.38627493495229837, "grad_norm": 5.68501091003418, "learning_rate": 1.748803133613811e-05, "loss": 0.0278, "step": 3563 }, { "epoch": 0.38638334778837813, "grad_norm": 1.5693020820617676, "learning_rate": 1.7487305962570726e-05, "loss": 0.0178, "step": 3564 }, { "epoch": 0.38649176062445795, "grad_norm": 1.5825538635253906, "learning_rate": 1.748658058900334e-05, "loss": 0.0582, "step": 3565 }, { "epoch": 0.3866001734605377, "grad_norm": 3.69874906539917, "learning_rate": 1.748585521543595e-05, "loss": 0.1033, "step": 3566 }, { "epoch": 0.38670858629661753, "grad_norm": 2.5193886756896973, "learning_rate": 1.7485129841868566e-05, "loss": 0.0363, "step": 3567 }, { "epoch": 0.3868169991326973, "grad_norm": 2.2629501819610596, "learning_rate": 1.7484404468301178e-05, "loss": 0.0534, "step": 3568 }, { "epoch": 0.3869254119687771, "grad_norm": 3.462144136428833, "learning_rate": 1.748367909473379e-05, "loss": 0.0994, "step": 3569 }, { "epoch": 0.3870338248048569, "grad_norm": 0.8543124198913574, "learning_rate": 1.7482953721166402e-05, "loss": 0.0231, "step": 3570 }, { "epoch": 0.3871422376409367, "grad_norm": 3.535810947418213, "learning_rate": 1.7482228347599015e-05, "loss": 0.033, "step": 3571 }, { "epoch": 0.38725065047701646, "grad_norm": 1.8974180221557617, "learning_rate": 1.7481502974031627e-05, "loss": 0.0413, "step": 3572 }, { "epoch": 0.3873590633130963, "grad_norm": 1.1170408725738525, "learning_rate": 1.748077760046424e-05, "loss": 0.0137, "step": 3573 }, { "epoch": 0.38746747614917604, "grad_norm": 1.8235516548156738, "learning_rate": 1.7480052226896854e-05, "loss": 0.0193, "step": 3574 }, { "epoch": 0.38757588898525586, "grad_norm": 2.774482011795044, "learning_rate": 1.7479326853329467e-05, "loss": 0.0199, "step": 3575 }, { "epoch": 0.3876843018213356, "grad_norm": 5.261894702911377, "learning_rate": 1.747860147976208e-05, "loss": 0.0857, "step": 3576 }, { "epoch": 0.38779271465741544, "grad_norm": 2.7886404991149902, "learning_rate": 1.747787610619469e-05, "loss": 0.0228, "step": 3577 }, { "epoch": 0.38790112749349526, "grad_norm": 1.0947767496109009, "learning_rate": 1.7477150732627303e-05, "loss": 0.0086, "step": 3578 }, { "epoch": 0.388009540329575, "grad_norm": 2.7773101329803467, "learning_rate": 1.7476425359059915e-05, "loss": 0.0533, "step": 3579 }, { "epoch": 0.38811795316565484, "grad_norm": 0.5802661180496216, "learning_rate": 1.747569998549253e-05, "loss": 0.0114, "step": 3580 }, { "epoch": 0.3882263660017346, "grad_norm": 5.150715351104736, "learning_rate": 1.7474974611925143e-05, "loss": 0.0693, "step": 3581 }, { "epoch": 0.3883347788378144, "grad_norm": 2.1094796657562256, "learning_rate": 1.7474249238357755e-05, "loss": 0.0327, "step": 3582 }, { "epoch": 0.3884431916738942, "grad_norm": 5.6963911056518555, "learning_rate": 1.747352386479037e-05, "loss": 0.0524, "step": 3583 }, { "epoch": 0.388551604509974, "grad_norm": 6.712653636932373, "learning_rate": 1.7472798491222983e-05, "loss": 0.044, "step": 3584 }, { "epoch": 0.38866001734605377, "grad_norm": 4.62864875793457, "learning_rate": 1.7472073117655595e-05, "loss": 0.0582, "step": 3585 }, { "epoch": 0.3887684301821336, "grad_norm": 2.0249290466308594, "learning_rate": 1.7471347744088207e-05, "loss": 0.0603, "step": 3586 }, { "epoch": 0.38887684301821335, "grad_norm": 5.753995895385742, "learning_rate": 1.747062237052082e-05, "loss": 0.0593, "step": 3587 }, { "epoch": 0.38898525585429317, "grad_norm": 1.1438781023025513, "learning_rate": 1.746989699695343e-05, "loss": 0.0298, "step": 3588 }, { "epoch": 0.38909366869037293, "grad_norm": 2.4866037368774414, "learning_rate": 1.7469171623386047e-05, "loss": 0.043, "step": 3589 }, { "epoch": 0.38920208152645275, "grad_norm": 3.7594141960144043, "learning_rate": 1.746844624981866e-05, "loss": 0.059, "step": 3590 }, { "epoch": 0.3893104943625325, "grad_norm": 4.418455123901367, "learning_rate": 1.746772087625127e-05, "loss": 0.0701, "step": 3591 }, { "epoch": 0.38941890719861233, "grad_norm": 5.531269550323486, "learning_rate": 1.7466995502683883e-05, "loss": 0.009, "step": 3592 }, { "epoch": 0.3895273200346921, "grad_norm": 3.121734857559204, "learning_rate": 1.7466270129116495e-05, "loss": 0.05, "step": 3593 }, { "epoch": 0.3896357328707719, "grad_norm": 2.274183511734009, "learning_rate": 1.7465544755549107e-05, "loss": 0.0315, "step": 3594 }, { "epoch": 0.3897441457068517, "grad_norm": 2.7304437160491943, "learning_rate": 1.7464819381981723e-05, "loss": 0.0353, "step": 3595 }, { "epoch": 0.3898525585429315, "grad_norm": 3.0942676067352295, "learning_rate": 1.7464094008414335e-05, "loss": 0.0577, "step": 3596 }, { "epoch": 0.38996097137901126, "grad_norm": 3.637108087539673, "learning_rate": 1.7463368634846947e-05, "loss": 0.0496, "step": 3597 }, { "epoch": 0.3900693842150911, "grad_norm": 3.2528178691864014, "learning_rate": 1.746264326127956e-05, "loss": 0.0291, "step": 3598 }, { "epoch": 0.39017779705117084, "grad_norm": 2.9159014225006104, "learning_rate": 1.7461917887712175e-05, "loss": 0.0324, "step": 3599 }, { "epoch": 0.39028620988725066, "grad_norm": 2.3000729084014893, "learning_rate": 1.7461192514144787e-05, "loss": 0.0268, "step": 3600 }, { "epoch": 0.3903946227233304, "grad_norm": 2.3647093772888184, "learning_rate": 1.74604671405774e-05, "loss": 0.0251, "step": 3601 }, { "epoch": 0.39050303555941024, "grad_norm": 5.4481635093688965, "learning_rate": 1.745974176701001e-05, "loss": 0.083, "step": 3602 }, { "epoch": 0.39061144839549, "grad_norm": 2.092068910598755, "learning_rate": 1.7459016393442624e-05, "loss": 0.0602, "step": 3603 }, { "epoch": 0.3907198612315698, "grad_norm": 1.7906464338302612, "learning_rate": 1.745829101987524e-05, "loss": 0.0424, "step": 3604 }, { "epoch": 0.3908282740676496, "grad_norm": 3.503493070602417, "learning_rate": 1.745756564630785e-05, "loss": 0.0504, "step": 3605 }, { "epoch": 0.3909366869037294, "grad_norm": 1.2966547012329102, "learning_rate": 1.7456840272740463e-05, "loss": 0.0275, "step": 3606 }, { "epoch": 0.3910450997398092, "grad_norm": 2.804124116897583, "learning_rate": 1.7456114899173075e-05, "loss": 0.0337, "step": 3607 }, { "epoch": 0.391153512575889, "grad_norm": 2.55934476852417, "learning_rate": 1.7455389525605688e-05, "loss": 0.0527, "step": 3608 }, { "epoch": 0.3912619254119688, "grad_norm": 1.2979553937911987, "learning_rate": 1.74546641520383e-05, "loss": 0.0451, "step": 3609 }, { "epoch": 0.39137033824804857, "grad_norm": 2.533914089202881, "learning_rate": 1.7453938778470912e-05, "loss": 0.0401, "step": 3610 }, { "epoch": 0.3914787510841284, "grad_norm": 2.3220973014831543, "learning_rate": 1.7453213404903527e-05, "loss": 0.0384, "step": 3611 }, { "epoch": 0.39158716392020815, "grad_norm": 4.001220226287842, "learning_rate": 1.745248803133614e-05, "loss": 0.0671, "step": 3612 }, { "epoch": 0.39169557675628797, "grad_norm": 2.8878486156463623, "learning_rate": 1.7451762657768752e-05, "loss": 0.0315, "step": 3613 }, { "epoch": 0.39180398959236773, "grad_norm": 0.8426175713539124, "learning_rate": 1.7451037284201364e-05, "loss": 0.0186, "step": 3614 }, { "epoch": 0.39191240242844755, "grad_norm": 0.5612526535987854, "learning_rate": 1.7450311910633976e-05, "loss": 0.013, "step": 3615 }, { "epoch": 0.3920208152645273, "grad_norm": 0.9218971729278564, "learning_rate": 1.744958653706659e-05, "loss": 0.0073, "step": 3616 }, { "epoch": 0.39212922810060713, "grad_norm": 2.80877685546875, "learning_rate": 1.7448861163499204e-05, "loss": 0.0617, "step": 3617 }, { "epoch": 0.3922376409366869, "grad_norm": 2.8907241821289062, "learning_rate": 1.7448135789931816e-05, "loss": 0.0516, "step": 3618 }, { "epoch": 0.3923460537727667, "grad_norm": 2.882939100265503, "learning_rate": 1.744741041636443e-05, "loss": 0.0217, "step": 3619 }, { "epoch": 0.3924544666088465, "grad_norm": 1.8368360996246338, "learning_rate": 1.7446685042797044e-05, "loss": 0.0509, "step": 3620 }, { "epoch": 0.3925628794449263, "grad_norm": 0.7154641151428223, "learning_rate": 1.7445959669229656e-05, "loss": 0.0231, "step": 3621 }, { "epoch": 0.39267129228100606, "grad_norm": 4.202454090118408, "learning_rate": 1.7445234295662268e-05, "loss": 0.0513, "step": 3622 }, { "epoch": 0.3927797051170859, "grad_norm": 3.4740355014801025, "learning_rate": 1.744450892209488e-05, "loss": 0.0306, "step": 3623 }, { "epoch": 0.39288811795316564, "grad_norm": 3.5955822467803955, "learning_rate": 1.7443783548527492e-05, "loss": 0.0405, "step": 3624 }, { "epoch": 0.39299653078924546, "grad_norm": 4.880216598510742, "learning_rate": 1.7443058174960104e-05, "loss": 0.1184, "step": 3625 }, { "epoch": 0.3931049436253252, "grad_norm": 2.0896878242492676, "learning_rate": 1.744233280139272e-05, "loss": 0.0141, "step": 3626 }, { "epoch": 0.39321335646140504, "grad_norm": 4.342165946960449, "learning_rate": 1.7441607427825332e-05, "loss": 0.0461, "step": 3627 }, { "epoch": 0.3933217692974848, "grad_norm": 3.2600512504577637, "learning_rate": 1.7440882054257944e-05, "loss": 0.0534, "step": 3628 }, { "epoch": 0.3934301821335646, "grad_norm": 3.7391598224639893, "learning_rate": 1.7440156680690556e-05, "loss": 0.0554, "step": 3629 }, { "epoch": 0.3935385949696444, "grad_norm": 1.986478567123413, "learning_rate": 1.743943130712317e-05, "loss": 0.0295, "step": 3630 }, { "epoch": 0.3936470078057242, "grad_norm": 1.9652725458145142, "learning_rate": 1.743870593355578e-05, "loss": 0.0368, "step": 3631 }, { "epoch": 0.39375542064180397, "grad_norm": 0.478317528963089, "learning_rate": 1.7437980559988396e-05, "loss": 0.0145, "step": 3632 }, { "epoch": 0.3938638334778838, "grad_norm": 4.354728698730469, "learning_rate": 1.7437255186421008e-05, "loss": 0.1078, "step": 3633 }, { "epoch": 0.39397224631396355, "grad_norm": 6.549875736236572, "learning_rate": 1.7436529812853624e-05, "loss": 0.0506, "step": 3634 }, { "epoch": 0.39408065915004337, "grad_norm": 6.096012592315674, "learning_rate": 1.7435804439286236e-05, "loss": 0.0395, "step": 3635 }, { "epoch": 0.39418907198612313, "grad_norm": 3.403028726577759, "learning_rate": 1.7435079065718848e-05, "loss": 0.0671, "step": 3636 }, { "epoch": 0.39429748482220295, "grad_norm": 6.98470401763916, "learning_rate": 1.743435369215146e-05, "loss": 0.0322, "step": 3637 }, { "epoch": 0.39440589765828277, "grad_norm": 1.5152299404144287, "learning_rate": 1.7433628318584072e-05, "loss": 0.0121, "step": 3638 }, { "epoch": 0.39451431049436253, "grad_norm": 3.3840532302856445, "learning_rate": 1.7432902945016684e-05, "loss": 0.1316, "step": 3639 }, { "epoch": 0.39462272333044235, "grad_norm": 0.8487409353256226, "learning_rate": 1.7432177571449297e-05, "loss": 0.006, "step": 3640 }, { "epoch": 0.3947311361665221, "grad_norm": 4.367578029632568, "learning_rate": 1.7431452197881912e-05, "loss": 0.0694, "step": 3641 }, { "epoch": 0.39483954900260193, "grad_norm": 3.0667638778686523, "learning_rate": 1.7430726824314524e-05, "loss": 0.0304, "step": 3642 }, { "epoch": 0.3949479618386817, "grad_norm": 2.428441047668457, "learning_rate": 1.7430001450747136e-05, "loss": 0.0415, "step": 3643 }, { "epoch": 0.3950563746747615, "grad_norm": 3.8943967819213867, "learning_rate": 1.742927607717975e-05, "loss": 0.0688, "step": 3644 }, { "epoch": 0.3951647875108413, "grad_norm": 2.802316904067993, "learning_rate": 1.742855070361236e-05, "loss": 0.0534, "step": 3645 }, { "epoch": 0.3952732003469211, "grad_norm": 2.5010735988616943, "learning_rate": 1.7427825330044973e-05, "loss": 0.036, "step": 3646 }, { "epoch": 0.39538161318300086, "grad_norm": 0.7681392431259155, "learning_rate": 1.742709995647759e-05, "loss": 0.0203, "step": 3647 }, { "epoch": 0.3954900260190807, "grad_norm": 2.225067377090454, "learning_rate": 1.74263745829102e-05, "loss": 0.0303, "step": 3648 }, { "epoch": 0.39559843885516044, "grad_norm": 5.29701042175293, "learning_rate": 1.7425649209342813e-05, "loss": 0.0788, "step": 3649 }, { "epoch": 0.39570685169124026, "grad_norm": 1.3891373872756958, "learning_rate": 1.7424923835775425e-05, "loss": 0.0291, "step": 3650 }, { "epoch": 0.39581526452732, "grad_norm": 2.289095878601074, "learning_rate": 1.742419846220804e-05, "loss": 0.0615, "step": 3651 }, { "epoch": 0.39592367736339984, "grad_norm": 2.5950512886047363, "learning_rate": 1.7423473088640652e-05, "loss": 0.0613, "step": 3652 }, { "epoch": 0.3960320901994796, "grad_norm": 1.5810648202896118, "learning_rate": 1.7422747715073265e-05, "loss": 0.0509, "step": 3653 }, { "epoch": 0.3961405030355594, "grad_norm": 1.6969748735427856, "learning_rate": 1.7422022341505877e-05, "loss": 0.0255, "step": 3654 }, { "epoch": 0.3962489158716392, "grad_norm": 1.3704396486282349, "learning_rate": 1.742129696793849e-05, "loss": 0.0232, "step": 3655 }, { "epoch": 0.396357328707719, "grad_norm": 6.236156940460205, "learning_rate": 1.7420571594371104e-05, "loss": 0.0338, "step": 3656 }, { "epoch": 0.39646574154379877, "grad_norm": 1.147537350654602, "learning_rate": 1.7419846220803717e-05, "loss": 0.0425, "step": 3657 }, { "epoch": 0.3965741543798786, "grad_norm": 1.4307888746261597, "learning_rate": 1.741912084723633e-05, "loss": 0.0408, "step": 3658 }, { "epoch": 0.39668256721595835, "grad_norm": 1.485025405883789, "learning_rate": 1.741839547366894e-05, "loss": 0.0189, "step": 3659 }, { "epoch": 0.39679098005203817, "grad_norm": 1.9936959743499756, "learning_rate": 1.7417670100101553e-05, "loss": 0.0343, "step": 3660 }, { "epoch": 0.39689939288811793, "grad_norm": 2.678128957748413, "learning_rate": 1.7416944726534165e-05, "loss": 0.0432, "step": 3661 }, { "epoch": 0.39700780572419775, "grad_norm": 2.1428515911102295, "learning_rate": 1.7416219352966777e-05, "loss": 0.0334, "step": 3662 }, { "epoch": 0.3971162185602775, "grad_norm": 1.4816865921020508, "learning_rate": 1.7415493979399393e-05, "loss": 0.0154, "step": 3663 }, { "epoch": 0.39722463139635733, "grad_norm": 1.536240577697754, "learning_rate": 1.7414768605832005e-05, "loss": 0.0222, "step": 3664 }, { "epoch": 0.3973330442324371, "grad_norm": 1.9189836978912354, "learning_rate": 1.7414043232264617e-05, "loss": 0.0348, "step": 3665 }, { "epoch": 0.3974414570685169, "grad_norm": 5.1824493408203125, "learning_rate": 1.741331785869723e-05, "loss": 0.095, "step": 3666 }, { "epoch": 0.39754986990459673, "grad_norm": 2.051335334777832, "learning_rate": 1.741259248512984e-05, "loss": 0.0296, "step": 3667 }, { "epoch": 0.3976582827406765, "grad_norm": 0.8977745175361633, "learning_rate": 1.7411867111562457e-05, "loss": 0.0221, "step": 3668 }, { "epoch": 0.3977666955767563, "grad_norm": 1.242905855178833, "learning_rate": 1.741114173799507e-05, "loss": 0.0212, "step": 3669 }, { "epoch": 0.3978751084128361, "grad_norm": 2.1811745166778564, "learning_rate": 1.741041636442768e-05, "loss": 0.0621, "step": 3670 }, { "epoch": 0.3979835212489159, "grad_norm": 2.0485754013061523, "learning_rate": 1.7409690990860297e-05, "loss": 0.025, "step": 3671 }, { "epoch": 0.39809193408499566, "grad_norm": 2.8520164489746094, "learning_rate": 1.740896561729291e-05, "loss": 0.0431, "step": 3672 }, { "epoch": 0.3982003469210755, "grad_norm": 2.2912487983703613, "learning_rate": 1.740824024372552e-05, "loss": 0.0232, "step": 3673 }, { "epoch": 0.39830875975715524, "grad_norm": 6.53175163269043, "learning_rate": 1.7407514870158133e-05, "loss": 0.0428, "step": 3674 }, { "epoch": 0.39841717259323506, "grad_norm": 2.211477518081665, "learning_rate": 1.7406789496590745e-05, "loss": 0.056, "step": 3675 }, { "epoch": 0.3985255854293148, "grad_norm": 4.668542385101318, "learning_rate": 1.7406064123023357e-05, "loss": 0.1186, "step": 3676 }, { "epoch": 0.39863399826539464, "grad_norm": 1.9902751445770264, "learning_rate": 1.740533874945597e-05, "loss": 0.0205, "step": 3677 }, { "epoch": 0.3987424111014744, "grad_norm": 2.0458617210388184, "learning_rate": 1.7404613375888585e-05, "loss": 0.0251, "step": 3678 }, { "epoch": 0.3988508239375542, "grad_norm": 0.30434751510620117, "learning_rate": 1.7403888002321197e-05, "loss": 0.001, "step": 3679 }, { "epoch": 0.398959236773634, "grad_norm": 1.4873790740966797, "learning_rate": 1.740316262875381e-05, "loss": 0.0138, "step": 3680 }, { "epoch": 0.3990676496097138, "grad_norm": 6.338332176208496, "learning_rate": 1.740243725518642e-05, "loss": 0.0243, "step": 3681 }, { "epoch": 0.39917606244579357, "grad_norm": 1.580449104309082, "learning_rate": 1.7401711881619034e-05, "loss": 0.0172, "step": 3682 }, { "epoch": 0.3992844752818734, "grad_norm": 2.4940969944000244, "learning_rate": 1.7400986508051646e-05, "loss": 0.0362, "step": 3683 }, { "epoch": 0.39939288811795315, "grad_norm": 1.2331844568252563, "learning_rate": 1.740026113448426e-05, "loss": 0.0452, "step": 3684 }, { "epoch": 0.39950130095403297, "grad_norm": 6.20367956161499, "learning_rate": 1.7399535760916874e-05, "loss": 0.0225, "step": 3685 }, { "epoch": 0.39960971379011273, "grad_norm": 3.3828587532043457, "learning_rate": 1.739881038734949e-05, "loss": 0.0449, "step": 3686 }, { "epoch": 0.39971812662619255, "grad_norm": 6.070315837860107, "learning_rate": 1.73980850137821e-05, "loss": 0.1141, "step": 3687 }, { "epoch": 0.3998265394622723, "grad_norm": 2.296118974685669, "learning_rate": 1.7397359640214713e-05, "loss": 0.0186, "step": 3688 }, { "epoch": 0.39993495229835213, "grad_norm": 2.6371421813964844, "learning_rate": 1.7396634266647325e-05, "loss": 0.0306, "step": 3689 }, { "epoch": 0.4000433651344319, "grad_norm": 2.0762364864349365, "learning_rate": 1.7395908893079938e-05, "loss": 0.0579, "step": 3690 }, { "epoch": 0.4001517779705117, "grad_norm": 2.8182895183563232, "learning_rate": 1.739518351951255e-05, "loss": 0.0255, "step": 3691 }, { "epoch": 0.4002601908065915, "grad_norm": 0.7931435704231262, "learning_rate": 1.7394458145945162e-05, "loss": 0.0127, "step": 3692 }, { "epoch": 0.4003686036426713, "grad_norm": 1.5469199419021606, "learning_rate": 1.7393732772377777e-05, "loss": 0.0226, "step": 3693 }, { "epoch": 0.40047701647875106, "grad_norm": 2.1094391345977783, "learning_rate": 1.739300739881039e-05, "loss": 0.0609, "step": 3694 }, { "epoch": 0.4005854293148309, "grad_norm": 8.193225860595703, "learning_rate": 1.7392282025243002e-05, "loss": 0.0598, "step": 3695 }, { "epoch": 0.4006938421509107, "grad_norm": 3.523939609527588, "learning_rate": 1.7391556651675614e-05, "loss": 0.054, "step": 3696 }, { "epoch": 0.40080225498699046, "grad_norm": 2.3357937335968018, "learning_rate": 1.7390831278108226e-05, "loss": 0.059, "step": 3697 }, { "epoch": 0.4009106678230703, "grad_norm": 2.603294849395752, "learning_rate": 1.7390105904540838e-05, "loss": 0.0246, "step": 3698 }, { "epoch": 0.40101908065915004, "grad_norm": 0.4360423982143402, "learning_rate": 1.7389380530973454e-05, "loss": 0.0015, "step": 3699 }, { "epoch": 0.40112749349522986, "grad_norm": 1.7024794816970825, "learning_rate": 1.7388655157406066e-05, "loss": 0.0157, "step": 3700 }, { "epoch": 0.4012359063313096, "grad_norm": 0.3974096477031708, "learning_rate": 1.7387929783838678e-05, "loss": 0.0086, "step": 3701 }, { "epoch": 0.40134431916738944, "grad_norm": 5.1290602684021, "learning_rate": 1.7387204410271293e-05, "loss": 0.0858, "step": 3702 }, { "epoch": 0.4014527320034692, "grad_norm": 4.128603458404541, "learning_rate": 1.7386479036703906e-05, "loss": 0.0305, "step": 3703 }, { "epoch": 0.401561144839549, "grad_norm": 7.569896697998047, "learning_rate": 1.7385753663136518e-05, "loss": 0.0704, "step": 3704 }, { "epoch": 0.4016695576756288, "grad_norm": 6.132882595062256, "learning_rate": 1.738502828956913e-05, "loss": 0.0521, "step": 3705 }, { "epoch": 0.4017779705117086, "grad_norm": 1.3529136180877686, "learning_rate": 1.7384302916001742e-05, "loss": 0.0382, "step": 3706 }, { "epoch": 0.40188638334778837, "grad_norm": 2.0954067707061768, "learning_rate": 1.7383577542434354e-05, "loss": 0.0262, "step": 3707 }, { "epoch": 0.4019947961838682, "grad_norm": 3.2639031410217285, "learning_rate": 1.738285216886697e-05, "loss": 0.0967, "step": 3708 }, { "epoch": 0.40210320901994795, "grad_norm": 2.273446798324585, "learning_rate": 1.7382126795299582e-05, "loss": 0.0448, "step": 3709 }, { "epoch": 0.40221162185602777, "grad_norm": 0.9743930101394653, "learning_rate": 1.7381401421732194e-05, "loss": 0.0123, "step": 3710 }, { "epoch": 0.40232003469210753, "grad_norm": 2.435955286026001, "learning_rate": 1.7380676048164806e-05, "loss": 0.0645, "step": 3711 }, { "epoch": 0.40242844752818735, "grad_norm": 2.912804126739502, "learning_rate": 1.737995067459742e-05, "loss": 0.0326, "step": 3712 }, { "epoch": 0.4025368603642671, "grad_norm": 3.0204548835754395, "learning_rate": 1.737922530103003e-05, "loss": 0.0788, "step": 3713 }, { "epoch": 0.40264527320034693, "grad_norm": 3.4692325592041016, "learning_rate": 1.7378499927462643e-05, "loss": 0.0615, "step": 3714 }, { "epoch": 0.4027536860364267, "grad_norm": 2.227647066116333, "learning_rate": 1.7377774553895258e-05, "loss": 0.0313, "step": 3715 }, { "epoch": 0.4028620988725065, "grad_norm": 3.05348539352417, "learning_rate": 1.737704918032787e-05, "loss": 0.0441, "step": 3716 }, { "epoch": 0.4029705117085863, "grad_norm": 5.878821849822998, "learning_rate": 1.7376323806760482e-05, "loss": 0.0976, "step": 3717 }, { "epoch": 0.4030789245446661, "grad_norm": 2.693044662475586, "learning_rate": 1.7375598433193095e-05, "loss": 0.0284, "step": 3718 }, { "epoch": 0.40318733738074586, "grad_norm": 3.0900447368621826, "learning_rate": 1.7374873059625707e-05, "loss": 0.0358, "step": 3719 }, { "epoch": 0.4032957502168257, "grad_norm": 2.7444751262664795, "learning_rate": 1.7374147686058322e-05, "loss": 0.0651, "step": 3720 }, { "epoch": 0.40340416305290544, "grad_norm": 2.5330681800842285, "learning_rate": 1.7373422312490934e-05, "loss": 0.0158, "step": 3721 }, { "epoch": 0.40351257588898526, "grad_norm": 2.9286253452301025, "learning_rate": 1.7372696938923547e-05, "loss": 0.0872, "step": 3722 }, { "epoch": 0.403620988725065, "grad_norm": 2.295144557952881, "learning_rate": 1.7371971565356162e-05, "loss": 0.0448, "step": 3723 }, { "epoch": 0.40372940156114484, "grad_norm": 1.5086567401885986, "learning_rate": 1.7371246191788774e-05, "loss": 0.0462, "step": 3724 }, { "epoch": 0.4038378143972246, "grad_norm": 1.7235655784606934, "learning_rate": 1.7370520818221386e-05, "loss": 0.0672, "step": 3725 }, { "epoch": 0.4039462272333044, "grad_norm": 1.7114468812942505, "learning_rate": 1.7369795444654e-05, "loss": 0.0202, "step": 3726 }, { "epoch": 0.40405464006938424, "grad_norm": 2.6277194023132324, "learning_rate": 1.736907007108661e-05, "loss": 0.0301, "step": 3727 }, { "epoch": 0.404163052905464, "grad_norm": 1.6416579484939575, "learning_rate": 1.7368344697519223e-05, "loss": 0.0314, "step": 3728 }, { "epoch": 0.4042714657415438, "grad_norm": 1.9560797214508057, "learning_rate": 1.7367619323951835e-05, "loss": 0.0298, "step": 3729 }, { "epoch": 0.4043798785776236, "grad_norm": 2.155146598815918, "learning_rate": 1.736689395038445e-05, "loss": 0.0485, "step": 3730 }, { "epoch": 0.4044882914137034, "grad_norm": 2.00107741355896, "learning_rate": 1.7366168576817063e-05, "loss": 0.0477, "step": 3731 }, { "epoch": 0.40459670424978317, "grad_norm": 2.8684098720550537, "learning_rate": 1.7365443203249675e-05, "loss": 0.066, "step": 3732 }, { "epoch": 0.404705117085863, "grad_norm": 0.5616599321365356, "learning_rate": 1.7364717829682287e-05, "loss": 0.0083, "step": 3733 }, { "epoch": 0.40481352992194275, "grad_norm": 3.2320303916931152, "learning_rate": 1.73639924561149e-05, "loss": 0.0733, "step": 3734 }, { "epoch": 0.40492194275802257, "grad_norm": 1.2136906385421753, "learning_rate": 1.736326708254751e-05, "loss": 0.0091, "step": 3735 }, { "epoch": 0.40503035559410233, "grad_norm": 1.3673923015594482, "learning_rate": 1.7362541708980127e-05, "loss": 0.0228, "step": 3736 }, { "epoch": 0.40513876843018215, "grad_norm": 3.519225835800171, "learning_rate": 1.736181633541274e-05, "loss": 0.0415, "step": 3737 }, { "epoch": 0.4052471812662619, "grad_norm": 3.3270320892333984, "learning_rate": 1.7361090961845354e-05, "loss": 0.02, "step": 3738 }, { "epoch": 0.40535559410234173, "grad_norm": 2.492558479309082, "learning_rate": 1.7360365588277967e-05, "loss": 0.0379, "step": 3739 }, { "epoch": 0.4054640069384215, "grad_norm": 0.882685124874115, "learning_rate": 1.735964021471058e-05, "loss": 0.0216, "step": 3740 }, { "epoch": 0.4055724197745013, "grad_norm": 3.133577823638916, "learning_rate": 1.735891484114319e-05, "loss": 0.0377, "step": 3741 }, { "epoch": 0.4056808326105811, "grad_norm": 2.27439546585083, "learning_rate": 1.7358189467575803e-05, "loss": 0.0358, "step": 3742 }, { "epoch": 0.4057892454466609, "grad_norm": 2.692136764526367, "learning_rate": 1.7357464094008415e-05, "loss": 0.0444, "step": 3743 }, { "epoch": 0.40589765828274066, "grad_norm": 2.9894118309020996, "learning_rate": 1.7356738720441027e-05, "loss": 0.0524, "step": 3744 }, { "epoch": 0.4060060711188205, "grad_norm": 1.5543410778045654, "learning_rate": 1.7356013346873643e-05, "loss": 0.0335, "step": 3745 }, { "epoch": 0.40611448395490024, "grad_norm": 2.723628282546997, "learning_rate": 1.7355287973306255e-05, "loss": 0.0515, "step": 3746 }, { "epoch": 0.40622289679098006, "grad_norm": 4.520706653594971, "learning_rate": 1.7354562599738867e-05, "loss": 0.0361, "step": 3747 }, { "epoch": 0.4063313096270598, "grad_norm": 2.2783243656158447, "learning_rate": 1.735383722617148e-05, "loss": 0.0354, "step": 3748 }, { "epoch": 0.40643972246313964, "grad_norm": 3.2368571758270264, "learning_rate": 1.735311185260409e-05, "loss": 0.0693, "step": 3749 }, { "epoch": 0.4065481352992194, "grad_norm": 2.203613042831421, "learning_rate": 1.7352386479036703e-05, "loss": 0.0656, "step": 3750 }, { "epoch": 0.4066565481352992, "grad_norm": 3.230863332748413, "learning_rate": 1.735166110546932e-05, "loss": 0.0333, "step": 3751 }, { "epoch": 0.406764960971379, "grad_norm": 4.093863487243652, "learning_rate": 1.735093573190193e-05, "loss": 0.096, "step": 3752 }, { "epoch": 0.4068733738074588, "grad_norm": 4.557138442993164, "learning_rate": 1.7350210358334543e-05, "loss": 0.0491, "step": 3753 }, { "epoch": 0.40698178664353857, "grad_norm": 3.5787835121154785, "learning_rate": 1.734948498476716e-05, "loss": 0.0195, "step": 3754 }, { "epoch": 0.4070901994796184, "grad_norm": 3.0785937309265137, "learning_rate": 1.734875961119977e-05, "loss": 0.0714, "step": 3755 }, { "epoch": 0.4071986123156982, "grad_norm": 1.6948460340499878, "learning_rate": 1.7348034237632383e-05, "loss": 0.0396, "step": 3756 }, { "epoch": 0.40730702515177797, "grad_norm": 1.0696250200271606, "learning_rate": 1.7347308864064995e-05, "loss": 0.007, "step": 3757 }, { "epoch": 0.4074154379878578, "grad_norm": 1.4164924621582031, "learning_rate": 1.7346583490497607e-05, "loss": 0.0099, "step": 3758 }, { "epoch": 0.40752385082393755, "grad_norm": 2.015815258026123, "learning_rate": 1.734585811693022e-05, "loss": 0.0183, "step": 3759 }, { "epoch": 0.40763226366001737, "grad_norm": 1.5047062635421753, "learning_rate": 1.7345132743362835e-05, "loss": 0.0371, "step": 3760 }, { "epoch": 0.40774067649609713, "grad_norm": 7.454853534698486, "learning_rate": 1.7344407369795447e-05, "loss": 0.0518, "step": 3761 }, { "epoch": 0.40784908933217695, "grad_norm": 2.3949291706085205, "learning_rate": 1.734368199622806e-05, "loss": 0.048, "step": 3762 }, { "epoch": 0.4079575021682567, "grad_norm": 8.019828796386719, "learning_rate": 1.734295662266067e-05, "loss": 0.0933, "step": 3763 }, { "epoch": 0.40806591500433653, "grad_norm": 3.5533394813537598, "learning_rate": 1.7342231249093284e-05, "loss": 0.0593, "step": 3764 }, { "epoch": 0.4081743278404163, "grad_norm": 3.112915277481079, "learning_rate": 1.7341505875525896e-05, "loss": 0.0502, "step": 3765 }, { "epoch": 0.4082827406764961, "grad_norm": 1.6011943817138672, "learning_rate": 1.7340780501958508e-05, "loss": 0.0398, "step": 3766 }, { "epoch": 0.4083911535125759, "grad_norm": 0.808015763759613, "learning_rate": 1.7340055128391123e-05, "loss": 0.0156, "step": 3767 }, { "epoch": 0.4084995663486557, "grad_norm": 1.7758655548095703, "learning_rate": 1.7339329754823736e-05, "loss": 0.0369, "step": 3768 }, { "epoch": 0.40860797918473546, "grad_norm": 3.029188871383667, "learning_rate": 1.7338604381256348e-05, "loss": 0.0191, "step": 3769 }, { "epoch": 0.4087163920208153, "grad_norm": 2.790233612060547, "learning_rate": 1.733787900768896e-05, "loss": 0.0465, "step": 3770 }, { "epoch": 0.40882480485689504, "grad_norm": 2.742936372756958, "learning_rate": 1.7337153634121572e-05, "loss": 0.0147, "step": 3771 }, { "epoch": 0.40893321769297486, "grad_norm": 2.1967875957489014, "learning_rate": 1.7336428260554188e-05, "loss": 0.0517, "step": 3772 }, { "epoch": 0.4090416305290546, "grad_norm": 2.076179265975952, "learning_rate": 1.73357028869868e-05, "loss": 0.0397, "step": 3773 }, { "epoch": 0.40915004336513444, "grad_norm": 1.3125604391098022, "learning_rate": 1.7334977513419412e-05, "loss": 0.0202, "step": 3774 }, { "epoch": 0.4092584562012142, "grad_norm": 3.3489673137664795, "learning_rate": 1.7334252139852027e-05, "loss": 0.033, "step": 3775 }, { "epoch": 0.409366869037294, "grad_norm": 1.446570873260498, "learning_rate": 1.733352676628464e-05, "loss": 0.047, "step": 3776 }, { "epoch": 0.4094752818733738, "grad_norm": 3.186082601547241, "learning_rate": 1.733280139271725e-05, "loss": 0.0264, "step": 3777 }, { "epoch": 0.4095836947094536, "grad_norm": 1.3532254695892334, "learning_rate": 1.7332076019149864e-05, "loss": 0.0404, "step": 3778 }, { "epoch": 0.40969210754553337, "grad_norm": 2.5001962184906006, "learning_rate": 1.7331350645582476e-05, "loss": 0.0378, "step": 3779 }, { "epoch": 0.4098005203816132, "grad_norm": 1.1636841297149658, "learning_rate": 1.7330625272015088e-05, "loss": 0.0187, "step": 3780 }, { "epoch": 0.40990893321769295, "grad_norm": 3.615619421005249, "learning_rate": 1.73298998984477e-05, "loss": 0.0567, "step": 3781 }, { "epoch": 0.41001734605377277, "grad_norm": 0.8579983115196228, "learning_rate": 1.7329174524880316e-05, "loss": 0.0069, "step": 3782 }, { "epoch": 0.41012575888985253, "grad_norm": 3.609618902206421, "learning_rate": 1.7328449151312928e-05, "loss": 0.0468, "step": 3783 }, { "epoch": 0.41023417172593235, "grad_norm": 2.410236120223999, "learning_rate": 1.732772377774554e-05, "loss": 0.0569, "step": 3784 }, { "epoch": 0.41034258456201217, "grad_norm": 6.691532611846924, "learning_rate": 1.7326998404178152e-05, "loss": 0.0388, "step": 3785 }, { "epoch": 0.41045099739809193, "grad_norm": 0.7051590085029602, "learning_rate": 1.7326273030610764e-05, "loss": 0.0193, "step": 3786 }, { "epoch": 0.41055941023417175, "grad_norm": 5.996774196624756, "learning_rate": 1.7325547657043377e-05, "loss": 0.0428, "step": 3787 }, { "epoch": 0.4106678230702515, "grad_norm": 8.525338172912598, "learning_rate": 1.7324822283475992e-05, "loss": 0.1626, "step": 3788 }, { "epoch": 0.41077623590633133, "grad_norm": 1.379834771156311, "learning_rate": 1.7324096909908604e-05, "loss": 0.0119, "step": 3789 }, { "epoch": 0.4108846487424111, "grad_norm": 5.288297176361084, "learning_rate": 1.732337153634122e-05, "loss": 0.0428, "step": 3790 }, { "epoch": 0.4109930615784909, "grad_norm": 5.22860860824585, "learning_rate": 1.7322646162773832e-05, "loss": 0.0717, "step": 3791 }, { "epoch": 0.4111014744145707, "grad_norm": 5.431976795196533, "learning_rate": 1.7321920789206444e-05, "loss": 0.0562, "step": 3792 }, { "epoch": 0.4112098872506505, "grad_norm": 2.8364012241363525, "learning_rate": 1.7321195415639056e-05, "loss": 0.0363, "step": 3793 }, { "epoch": 0.41131830008673026, "grad_norm": 6.584090709686279, "learning_rate": 1.7320470042071668e-05, "loss": 0.0341, "step": 3794 }, { "epoch": 0.4114267129228101, "grad_norm": 1.2904428243637085, "learning_rate": 1.731974466850428e-05, "loss": 0.0321, "step": 3795 }, { "epoch": 0.41153512575888984, "grad_norm": 0.5265966057777405, "learning_rate": 1.7319019294936893e-05, "loss": 0.0194, "step": 3796 }, { "epoch": 0.41164353859496966, "grad_norm": 2.900217056274414, "learning_rate": 1.7318293921369508e-05, "loss": 0.036, "step": 3797 }, { "epoch": 0.4117519514310494, "grad_norm": 2.5430996417999268, "learning_rate": 1.731756854780212e-05, "loss": 0.0443, "step": 3798 }, { "epoch": 0.41186036426712924, "grad_norm": 3.760612964630127, "learning_rate": 1.7316843174234732e-05, "loss": 0.0681, "step": 3799 }, { "epoch": 0.411968777103209, "grad_norm": 2.817011594772339, "learning_rate": 1.7316117800667345e-05, "loss": 0.072, "step": 3800 }, { "epoch": 0.4120771899392888, "grad_norm": 2.7976701259613037, "learning_rate": 1.7315392427099957e-05, "loss": 0.058, "step": 3801 }, { "epoch": 0.4121856027753686, "grad_norm": 1.832269310951233, "learning_rate": 1.731466705353257e-05, "loss": 0.0172, "step": 3802 }, { "epoch": 0.4122940156114484, "grad_norm": 0.9687168002128601, "learning_rate": 1.7313941679965184e-05, "loss": 0.0132, "step": 3803 }, { "epoch": 0.41240242844752817, "grad_norm": 7.266855239868164, "learning_rate": 1.7313216306397796e-05, "loss": 0.0191, "step": 3804 }, { "epoch": 0.412510841283608, "grad_norm": 9.610440254211426, "learning_rate": 1.731249093283041e-05, "loss": 0.0266, "step": 3805 }, { "epoch": 0.41261925411968775, "grad_norm": 3.290567398071289, "learning_rate": 1.7311765559263024e-05, "loss": 0.029, "step": 3806 }, { "epoch": 0.41272766695576757, "grad_norm": 0.8528112769126892, "learning_rate": 1.7311040185695636e-05, "loss": 0.0101, "step": 3807 }, { "epoch": 0.41283607979184733, "grad_norm": 5.196075439453125, "learning_rate": 1.731031481212825e-05, "loss": 0.0312, "step": 3808 }, { "epoch": 0.41294449262792715, "grad_norm": 2.931553363800049, "learning_rate": 1.730958943856086e-05, "loss": 0.0625, "step": 3809 }, { "epoch": 0.4130529054640069, "grad_norm": 0.2229873239994049, "learning_rate": 1.7308864064993473e-05, "loss": 0.0023, "step": 3810 }, { "epoch": 0.41316131830008673, "grad_norm": 5.529951095581055, "learning_rate": 1.7308138691426085e-05, "loss": 0.0675, "step": 3811 }, { "epoch": 0.4132697311361665, "grad_norm": 5.5042195320129395, "learning_rate": 1.73074133178587e-05, "loss": 0.0633, "step": 3812 }, { "epoch": 0.4133781439722463, "grad_norm": 3.670104742050171, "learning_rate": 1.7306687944291313e-05, "loss": 0.0405, "step": 3813 }, { "epoch": 0.41348655680832613, "grad_norm": 1.683276891708374, "learning_rate": 1.7305962570723925e-05, "loss": 0.0219, "step": 3814 }, { "epoch": 0.4135949696444059, "grad_norm": 2.942230224609375, "learning_rate": 1.7305237197156537e-05, "loss": 0.0749, "step": 3815 }, { "epoch": 0.4137033824804857, "grad_norm": 4.529637336730957, "learning_rate": 1.730451182358915e-05, "loss": 0.0647, "step": 3816 }, { "epoch": 0.4138117953165655, "grad_norm": 1.90169095993042, "learning_rate": 1.730378645002176e-05, "loss": 0.0232, "step": 3817 }, { "epoch": 0.4139202081526453, "grad_norm": 2.5284249782562256, "learning_rate": 1.7303061076454373e-05, "loss": 0.0425, "step": 3818 }, { "epoch": 0.41402862098872506, "grad_norm": 0.1613054722547531, "learning_rate": 1.730233570288699e-05, "loss": 0.0024, "step": 3819 }, { "epoch": 0.4141370338248049, "grad_norm": 3.830828905105591, "learning_rate": 1.73016103293196e-05, "loss": 0.0399, "step": 3820 }, { "epoch": 0.41424544666088464, "grad_norm": 3.414799928665161, "learning_rate": 1.7300884955752213e-05, "loss": 0.0364, "step": 3821 }, { "epoch": 0.41435385949696446, "grad_norm": 2.7260892391204834, "learning_rate": 1.7300159582184825e-05, "loss": 0.0367, "step": 3822 }, { "epoch": 0.4144622723330442, "grad_norm": 0.7632166743278503, "learning_rate": 1.729943420861744e-05, "loss": 0.0206, "step": 3823 }, { "epoch": 0.41457068516912404, "grad_norm": 1.495160460472107, "learning_rate": 1.7298708835050053e-05, "loss": 0.0083, "step": 3824 }, { "epoch": 0.4146790980052038, "grad_norm": 1.5128567218780518, "learning_rate": 1.7297983461482665e-05, "loss": 0.0193, "step": 3825 }, { "epoch": 0.4147875108412836, "grad_norm": 1.5873644351959229, "learning_rate": 1.7297258087915277e-05, "loss": 0.0469, "step": 3826 }, { "epoch": 0.4148959236773634, "grad_norm": 3.6039540767669678, "learning_rate": 1.7296532714347893e-05, "loss": 0.0523, "step": 3827 }, { "epoch": 0.4150043365134432, "grad_norm": 7.427114486694336, "learning_rate": 1.7295807340780505e-05, "loss": 0.0864, "step": 3828 }, { "epoch": 0.41511274934952297, "grad_norm": 3.084418773651123, "learning_rate": 1.7295081967213117e-05, "loss": 0.0399, "step": 3829 }, { "epoch": 0.4152211621856028, "grad_norm": 2.9850430488586426, "learning_rate": 1.729435659364573e-05, "loss": 0.0592, "step": 3830 }, { "epoch": 0.41532957502168255, "grad_norm": 2.4867732524871826, "learning_rate": 1.729363122007834e-05, "loss": 0.0196, "step": 3831 }, { "epoch": 0.41543798785776237, "grad_norm": 1.2066528797149658, "learning_rate": 1.7292905846510953e-05, "loss": 0.0356, "step": 3832 }, { "epoch": 0.41554640069384213, "grad_norm": 2.3588318824768066, "learning_rate": 1.7292180472943566e-05, "loss": 0.0631, "step": 3833 }, { "epoch": 0.41565481352992195, "grad_norm": 1.1989028453826904, "learning_rate": 1.729145509937618e-05, "loss": 0.0154, "step": 3834 }, { "epoch": 0.4157632263660017, "grad_norm": 4.676646709442139, "learning_rate": 1.7290729725808793e-05, "loss": 0.0622, "step": 3835 }, { "epoch": 0.41587163920208153, "grad_norm": 3.2680249214172363, "learning_rate": 1.7290004352241405e-05, "loss": 0.0247, "step": 3836 }, { "epoch": 0.4159800520381613, "grad_norm": 2.375772476196289, "learning_rate": 1.7289278978674018e-05, "loss": 0.0436, "step": 3837 }, { "epoch": 0.4160884648742411, "grad_norm": 3.1531922817230225, "learning_rate": 1.728855360510663e-05, "loss": 0.044, "step": 3838 }, { "epoch": 0.4161968777103209, "grad_norm": 0.7240849733352661, "learning_rate": 1.7287828231539242e-05, "loss": 0.0097, "step": 3839 }, { "epoch": 0.4163052905464007, "grad_norm": 2.0701115131378174, "learning_rate": 1.7287102857971857e-05, "loss": 0.0332, "step": 3840 }, { "epoch": 0.41641370338248046, "grad_norm": 1.5797351598739624, "learning_rate": 1.728637748440447e-05, "loss": 0.0337, "step": 3841 }, { "epoch": 0.4165221162185603, "grad_norm": 1.763670563697815, "learning_rate": 1.7285652110837085e-05, "loss": 0.0256, "step": 3842 }, { "epoch": 0.41663052905464004, "grad_norm": 3.0733449459075928, "learning_rate": 1.7284926737269697e-05, "loss": 0.0406, "step": 3843 }, { "epoch": 0.41673894189071986, "grad_norm": 4.578733444213867, "learning_rate": 1.728420136370231e-05, "loss": 0.0562, "step": 3844 }, { "epoch": 0.4168473547267997, "grad_norm": 0.7712908983230591, "learning_rate": 1.728347599013492e-05, "loss": 0.0224, "step": 3845 }, { "epoch": 0.41695576756287944, "grad_norm": 3.5290563106536865, "learning_rate": 1.7282750616567534e-05, "loss": 0.041, "step": 3846 }, { "epoch": 0.41706418039895926, "grad_norm": 0.9477789998054504, "learning_rate": 1.7282025243000146e-05, "loss": 0.0102, "step": 3847 }, { "epoch": 0.417172593235039, "grad_norm": 7.8977556228637695, "learning_rate": 1.7281299869432758e-05, "loss": 0.042, "step": 3848 }, { "epoch": 0.41728100607111884, "grad_norm": 2.4570205211639404, "learning_rate": 1.7280574495865373e-05, "loss": 0.0247, "step": 3849 }, { "epoch": 0.4173894189071986, "grad_norm": 3.4197309017181396, "learning_rate": 1.7279849122297986e-05, "loss": 0.0543, "step": 3850 }, { "epoch": 0.4174978317432784, "grad_norm": 1.8840054273605347, "learning_rate": 1.7279123748730598e-05, "loss": 0.0345, "step": 3851 }, { "epoch": 0.4176062445793582, "grad_norm": 5.027442932128906, "learning_rate": 1.727839837516321e-05, "loss": 0.0448, "step": 3852 }, { "epoch": 0.417714657415438, "grad_norm": 1.988006591796875, "learning_rate": 1.7277673001595822e-05, "loss": 0.0264, "step": 3853 }, { "epoch": 0.41782307025151777, "grad_norm": 1.329695463180542, "learning_rate": 1.7276947628028434e-05, "loss": 0.0456, "step": 3854 }, { "epoch": 0.4179314830875976, "grad_norm": 1.9357835054397583, "learning_rate": 1.727622225446105e-05, "loss": 0.0228, "step": 3855 }, { "epoch": 0.41803989592367735, "grad_norm": 3.392714738845825, "learning_rate": 1.7275496880893662e-05, "loss": 0.069, "step": 3856 }, { "epoch": 0.41814830875975717, "grad_norm": 3.441706657409668, "learning_rate": 1.7274771507326274e-05, "loss": 0.0414, "step": 3857 }, { "epoch": 0.41825672159583693, "grad_norm": 4.146743297576904, "learning_rate": 1.727404613375889e-05, "loss": 0.0434, "step": 3858 }, { "epoch": 0.41836513443191675, "grad_norm": 1.748692274093628, "learning_rate": 1.72733207601915e-05, "loss": 0.019, "step": 3859 }, { "epoch": 0.4184735472679965, "grad_norm": 2.289945363998413, "learning_rate": 1.7272595386624114e-05, "loss": 0.0273, "step": 3860 }, { "epoch": 0.41858196010407633, "grad_norm": 3.1472225189208984, "learning_rate": 1.7271870013056726e-05, "loss": 0.0181, "step": 3861 }, { "epoch": 0.4186903729401561, "grad_norm": 2.7055790424346924, "learning_rate": 1.7271144639489338e-05, "loss": 0.0201, "step": 3862 }, { "epoch": 0.4187987857762359, "grad_norm": 1.8431085348129272, "learning_rate": 1.727041926592195e-05, "loss": 0.0093, "step": 3863 }, { "epoch": 0.4189071986123157, "grad_norm": 0.6312961578369141, "learning_rate": 1.7269693892354566e-05, "loss": 0.013, "step": 3864 }, { "epoch": 0.4190156114483955, "grad_norm": 5.559172630310059, "learning_rate": 1.7268968518787178e-05, "loss": 0.0934, "step": 3865 }, { "epoch": 0.41912402428447526, "grad_norm": 3.0793371200561523, "learning_rate": 1.726824314521979e-05, "loss": 0.0113, "step": 3866 }, { "epoch": 0.4192324371205551, "grad_norm": 7.521474361419678, "learning_rate": 1.7267517771652402e-05, "loss": 0.0564, "step": 3867 }, { "epoch": 0.41934084995663484, "grad_norm": 0.5386688709259033, "learning_rate": 1.7266792398085014e-05, "loss": 0.0097, "step": 3868 }, { "epoch": 0.41944926279271466, "grad_norm": 3.739760637283325, "learning_rate": 1.7266067024517626e-05, "loss": 0.0627, "step": 3869 }, { "epoch": 0.4195576756287944, "grad_norm": 4.279928684234619, "learning_rate": 1.726534165095024e-05, "loss": 0.0171, "step": 3870 }, { "epoch": 0.41966608846487424, "grad_norm": 3.1222727298736572, "learning_rate": 1.7264616277382854e-05, "loss": 0.0249, "step": 3871 }, { "epoch": 0.419774501300954, "grad_norm": 2.888275146484375, "learning_rate": 1.7263890903815466e-05, "loss": 0.033, "step": 3872 }, { "epoch": 0.4198829141370338, "grad_norm": 3.568397045135498, "learning_rate": 1.726316553024808e-05, "loss": 0.0962, "step": 3873 }, { "epoch": 0.41999132697311364, "grad_norm": 5.374508380889893, "learning_rate": 1.726244015668069e-05, "loss": 0.0457, "step": 3874 }, { "epoch": 0.4200997398091934, "grad_norm": 2.5744869709014893, "learning_rate": 1.7261714783113306e-05, "loss": 0.0398, "step": 3875 }, { "epoch": 0.4202081526452732, "grad_norm": 3.167720079421997, "learning_rate": 1.7260989409545918e-05, "loss": 0.0607, "step": 3876 }, { "epoch": 0.420316565481353, "grad_norm": 7.305035591125488, "learning_rate": 1.726026403597853e-05, "loss": 0.1258, "step": 3877 }, { "epoch": 0.4204249783174328, "grad_norm": 2.6762595176696777, "learning_rate": 1.7259538662411143e-05, "loss": 0.0249, "step": 3878 }, { "epoch": 0.42053339115351257, "grad_norm": 7.387366771697998, "learning_rate": 1.7258813288843758e-05, "loss": 0.013, "step": 3879 }, { "epoch": 0.4206418039895924, "grad_norm": 3.0897042751312256, "learning_rate": 1.725808791527637e-05, "loss": 0.0219, "step": 3880 }, { "epoch": 0.42075021682567215, "grad_norm": 3.6620934009552, "learning_rate": 1.7257362541708982e-05, "loss": 0.0418, "step": 3881 }, { "epoch": 0.42085862966175197, "grad_norm": 4.516517162322998, "learning_rate": 1.7256637168141594e-05, "loss": 0.1097, "step": 3882 }, { "epoch": 0.42096704249783173, "grad_norm": 5.317453384399414, "learning_rate": 1.7255911794574207e-05, "loss": 0.0948, "step": 3883 }, { "epoch": 0.42107545533391155, "grad_norm": 1.475817322731018, "learning_rate": 1.725518642100682e-05, "loss": 0.0392, "step": 3884 }, { "epoch": 0.4211838681699913, "grad_norm": 1.671354055404663, "learning_rate": 1.725446104743943e-05, "loss": 0.0345, "step": 3885 }, { "epoch": 0.42129228100607113, "grad_norm": 1.1690623760223389, "learning_rate": 1.7253735673872046e-05, "loss": 0.0194, "step": 3886 }, { "epoch": 0.4214006938421509, "grad_norm": 2.869277238845825, "learning_rate": 1.725301030030466e-05, "loss": 0.0159, "step": 3887 }, { "epoch": 0.4215091066782307, "grad_norm": 2.9313504695892334, "learning_rate": 1.725228492673727e-05, "loss": 0.0219, "step": 3888 }, { "epoch": 0.4216175195143105, "grad_norm": 2.0686213970184326, "learning_rate": 1.7251559553169883e-05, "loss": 0.0571, "step": 3889 }, { "epoch": 0.4217259323503903, "grad_norm": 3.043933391571045, "learning_rate": 1.7250834179602495e-05, "loss": 0.0736, "step": 3890 }, { "epoch": 0.42183434518647006, "grad_norm": 1.3630772829055786, "learning_rate": 1.7250108806035107e-05, "loss": 0.0266, "step": 3891 }, { "epoch": 0.4219427580225499, "grad_norm": 2.5190320014953613, "learning_rate": 1.7249383432467723e-05, "loss": 0.0892, "step": 3892 }, { "epoch": 0.42205117085862964, "grad_norm": 1.4802426099777222, "learning_rate": 1.7248658058900335e-05, "loss": 0.0282, "step": 3893 }, { "epoch": 0.42215958369470946, "grad_norm": 4.3567585945129395, "learning_rate": 1.724793268533295e-05, "loss": 0.0445, "step": 3894 }, { "epoch": 0.4222679965307892, "grad_norm": 1.1664702892303467, "learning_rate": 1.7247207311765563e-05, "loss": 0.0347, "step": 3895 }, { "epoch": 0.42237640936686904, "grad_norm": 2.5332841873168945, "learning_rate": 1.7246481938198175e-05, "loss": 0.0539, "step": 3896 }, { "epoch": 0.4224848222029488, "grad_norm": 1.9554708003997803, "learning_rate": 1.7245756564630787e-05, "loss": 0.044, "step": 3897 }, { "epoch": 0.4225932350390286, "grad_norm": 2.27578067779541, "learning_rate": 1.72450311910634e-05, "loss": 0.0452, "step": 3898 }, { "epoch": 0.4227016478751084, "grad_norm": 4.252029895782471, "learning_rate": 1.724430581749601e-05, "loss": 0.0424, "step": 3899 }, { "epoch": 0.4228100607111882, "grad_norm": 3.6179511547088623, "learning_rate": 1.7243580443928623e-05, "loss": 0.0376, "step": 3900 }, { "epoch": 0.42291847354726797, "grad_norm": 1.3115043640136719, "learning_rate": 1.724285507036124e-05, "loss": 0.0388, "step": 3901 }, { "epoch": 0.4230268863833478, "grad_norm": 1.227481484413147, "learning_rate": 1.724212969679385e-05, "loss": 0.0194, "step": 3902 }, { "epoch": 0.4231352992194276, "grad_norm": 1.2322190999984741, "learning_rate": 1.7241404323226463e-05, "loss": 0.0341, "step": 3903 }, { "epoch": 0.42324371205550737, "grad_norm": 4.858956336975098, "learning_rate": 1.7240678949659075e-05, "loss": 0.0723, "step": 3904 }, { "epoch": 0.4233521248915872, "grad_norm": 1.2748368978500366, "learning_rate": 1.7239953576091687e-05, "loss": 0.0171, "step": 3905 }, { "epoch": 0.42346053772766695, "grad_norm": 0.9832856059074402, "learning_rate": 1.72392282025243e-05, "loss": 0.0115, "step": 3906 }, { "epoch": 0.42356895056374677, "grad_norm": 2.3766677379608154, "learning_rate": 1.7238502828956915e-05, "loss": 0.0558, "step": 3907 }, { "epoch": 0.42367736339982653, "grad_norm": 3.898042678833008, "learning_rate": 1.7237777455389527e-05, "loss": 0.0508, "step": 3908 }, { "epoch": 0.42378577623590635, "grad_norm": 1.0154684782028198, "learning_rate": 1.723705208182214e-05, "loss": 0.014, "step": 3909 }, { "epoch": 0.4238941890719861, "grad_norm": 0.8024374842643738, "learning_rate": 1.7236326708254755e-05, "loss": 0.0067, "step": 3910 }, { "epoch": 0.42400260190806593, "grad_norm": 3.3870198726654053, "learning_rate": 1.7235601334687367e-05, "loss": 0.0677, "step": 3911 }, { "epoch": 0.4241110147441457, "grad_norm": 3.7479896545410156, "learning_rate": 1.723487596111998e-05, "loss": 0.0457, "step": 3912 }, { "epoch": 0.4242194275802255, "grad_norm": 1.280005693435669, "learning_rate": 1.723415058755259e-05, "loss": 0.0264, "step": 3913 }, { "epoch": 0.4243278404163053, "grad_norm": 1.8492891788482666, "learning_rate": 1.7233425213985203e-05, "loss": 0.0308, "step": 3914 }, { "epoch": 0.4244362532523851, "grad_norm": 3.3189210891723633, "learning_rate": 1.7232699840417816e-05, "loss": 0.0501, "step": 3915 }, { "epoch": 0.42454466608846486, "grad_norm": 7.162220001220703, "learning_rate": 1.723197446685043e-05, "loss": 0.0742, "step": 3916 }, { "epoch": 0.4246530789245447, "grad_norm": 6.530440807342529, "learning_rate": 1.7231249093283043e-05, "loss": 0.0931, "step": 3917 }, { "epoch": 0.42476149176062444, "grad_norm": 0.44017186760902405, "learning_rate": 1.7230523719715655e-05, "loss": 0.0214, "step": 3918 }, { "epoch": 0.42486990459670426, "grad_norm": 7.039090633392334, "learning_rate": 1.7229798346148268e-05, "loss": 0.0188, "step": 3919 }, { "epoch": 0.424978317432784, "grad_norm": 8.55905818939209, "learning_rate": 1.722907297258088e-05, "loss": 0.0707, "step": 3920 }, { "epoch": 0.42508673026886384, "grad_norm": 3.8016223907470703, "learning_rate": 1.7228347599013492e-05, "loss": 0.0364, "step": 3921 }, { "epoch": 0.4251951431049436, "grad_norm": 5.256477355957031, "learning_rate": 1.7227622225446104e-05, "loss": 0.0625, "step": 3922 }, { "epoch": 0.4253035559410234, "grad_norm": 5.976753234863281, "learning_rate": 1.722689685187872e-05, "loss": 0.0604, "step": 3923 }, { "epoch": 0.4254119687771032, "grad_norm": 1.0638067722320557, "learning_rate": 1.722617147831133e-05, "loss": 0.0382, "step": 3924 }, { "epoch": 0.425520381613183, "grad_norm": 0.33722707629203796, "learning_rate": 1.7225446104743944e-05, "loss": 0.0032, "step": 3925 }, { "epoch": 0.42562879444926277, "grad_norm": 0.5385274291038513, "learning_rate": 1.7224720731176556e-05, "loss": 0.0172, "step": 3926 }, { "epoch": 0.4257372072853426, "grad_norm": 2.6773412227630615, "learning_rate": 1.722399535760917e-05, "loss": 0.0342, "step": 3927 }, { "epoch": 0.42584562012142235, "grad_norm": 1.5947364568710327, "learning_rate": 1.7223269984041784e-05, "loss": 0.0043, "step": 3928 }, { "epoch": 0.42595403295750217, "grad_norm": 1.9264782667160034, "learning_rate": 1.7222544610474396e-05, "loss": 0.0198, "step": 3929 }, { "epoch": 0.42606244579358193, "grad_norm": 3.0829901695251465, "learning_rate": 1.7221819236907008e-05, "loss": 0.0431, "step": 3930 }, { "epoch": 0.42617085862966175, "grad_norm": 0.39959731698036194, "learning_rate": 1.7221093863339623e-05, "loss": 0.011, "step": 3931 }, { "epoch": 0.4262792714657415, "grad_norm": 3.3974366188049316, "learning_rate": 1.7220368489772236e-05, "loss": 0.0574, "step": 3932 }, { "epoch": 0.42638768430182133, "grad_norm": 11.717806816101074, "learning_rate": 1.7219643116204848e-05, "loss": 0.026, "step": 3933 }, { "epoch": 0.42649609713790115, "grad_norm": 0.9951635599136353, "learning_rate": 1.721891774263746e-05, "loss": 0.0084, "step": 3934 }, { "epoch": 0.4266045099739809, "grad_norm": 4.157150745391846, "learning_rate": 1.7218192369070072e-05, "loss": 0.0374, "step": 3935 }, { "epoch": 0.42671292281006074, "grad_norm": 5.999587535858154, "learning_rate": 1.7217466995502684e-05, "loss": 0.0855, "step": 3936 }, { "epoch": 0.4268213356461405, "grad_norm": 2.6011784076690674, "learning_rate": 1.7216741621935296e-05, "loss": 0.0367, "step": 3937 }, { "epoch": 0.4269297484822203, "grad_norm": 6.407433032989502, "learning_rate": 1.7216016248367912e-05, "loss": 0.0978, "step": 3938 }, { "epoch": 0.4270381613183001, "grad_norm": 2.3847131729125977, "learning_rate": 1.7215290874800524e-05, "loss": 0.0298, "step": 3939 }, { "epoch": 0.4271465741543799, "grad_norm": 4.0262603759765625, "learning_rate": 1.7214565501233136e-05, "loss": 0.058, "step": 3940 }, { "epoch": 0.42725498699045966, "grad_norm": 5.2061848640441895, "learning_rate": 1.7213840127665748e-05, "loss": 0.065, "step": 3941 }, { "epoch": 0.4273633998265395, "grad_norm": 2.4605953693389893, "learning_rate": 1.721311475409836e-05, "loss": 0.0488, "step": 3942 }, { "epoch": 0.42747181266261924, "grad_norm": 2.042325735092163, "learning_rate": 1.7212389380530973e-05, "loss": 0.0112, "step": 3943 }, { "epoch": 0.42758022549869906, "grad_norm": 8.298065185546875, "learning_rate": 1.7211664006963588e-05, "loss": 0.0597, "step": 3944 }, { "epoch": 0.4276886383347788, "grad_norm": 0.3247162699699402, "learning_rate": 1.72109386333962e-05, "loss": 0.0068, "step": 3945 }, { "epoch": 0.42779705117085864, "grad_norm": 2.964369297027588, "learning_rate": 1.7210213259828816e-05, "loss": 0.0387, "step": 3946 }, { "epoch": 0.4279054640069384, "grad_norm": 0.9607729315757751, "learning_rate": 1.7209487886261428e-05, "loss": 0.016, "step": 3947 }, { "epoch": 0.4280138768430182, "grad_norm": 5.618923187255859, "learning_rate": 1.720876251269404e-05, "loss": 0.0769, "step": 3948 }, { "epoch": 0.428122289679098, "grad_norm": 1.2394025325775146, "learning_rate": 1.7208037139126652e-05, "loss": 0.0238, "step": 3949 }, { "epoch": 0.4282307025151778, "grad_norm": 1.4266316890716553, "learning_rate": 1.7207311765559264e-05, "loss": 0.0205, "step": 3950 }, { "epoch": 0.42833911535125757, "grad_norm": 3.756162405014038, "learning_rate": 1.7206586391991876e-05, "loss": 0.0421, "step": 3951 }, { "epoch": 0.4284475281873374, "grad_norm": 0.4183262884616852, "learning_rate": 1.720586101842449e-05, "loss": 0.0051, "step": 3952 }, { "epoch": 0.42855594102341715, "grad_norm": 1.7090678215026855, "learning_rate": 1.7205135644857104e-05, "loss": 0.0191, "step": 3953 }, { "epoch": 0.42866435385949697, "grad_norm": 1.046150803565979, "learning_rate": 1.7204410271289716e-05, "loss": 0.0276, "step": 3954 }, { "epoch": 0.42877276669557673, "grad_norm": 1.6393213272094727, "learning_rate": 1.720368489772233e-05, "loss": 0.013, "step": 3955 }, { "epoch": 0.42888117953165655, "grad_norm": 3.201442003250122, "learning_rate": 1.720295952415494e-05, "loss": 0.0624, "step": 3956 }, { "epoch": 0.4289895923677363, "grad_norm": 5.402967929840088, "learning_rate": 1.7202234150587553e-05, "loss": 0.039, "step": 3957 }, { "epoch": 0.42909800520381614, "grad_norm": 2.1213297843933105, "learning_rate": 1.7201508777020165e-05, "loss": 0.0688, "step": 3958 }, { "epoch": 0.4292064180398959, "grad_norm": 4.452892780303955, "learning_rate": 1.720078340345278e-05, "loss": 0.0418, "step": 3959 }, { "epoch": 0.4293148308759757, "grad_norm": 3.6073031425476074, "learning_rate": 1.7200058029885393e-05, "loss": 0.033, "step": 3960 }, { "epoch": 0.4294232437120555, "grad_norm": 2.8427484035491943, "learning_rate": 1.7199332656318005e-05, "loss": 0.072, "step": 3961 }, { "epoch": 0.4295316565481353, "grad_norm": 4.944554805755615, "learning_rate": 1.719860728275062e-05, "loss": 0.0474, "step": 3962 }, { "epoch": 0.4296400693842151, "grad_norm": 1.869627594947815, "learning_rate": 1.7197881909183232e-05, "loss": 0.0459, "step": 3963 }, { "epoch": 0.4297484822202949, "grad_norm": 1.725703239440918, "learning_rate": 1.7197156535615844e-05, "loss": 0.0326, "step": 3964 }, { "epoch": 0.4298568950563747, "grad_norm": 6.194305896759033, "learning_rate": 1.7196431162048457e-05, "loss": 0.0334, "step": 3965 }, { "epoch": 0.42996530789245446, "grad_norm": 0.6566799283027649, "learning_rate": 1.719570578848107e-05, "loss": 0.011, "step": 3966 }, { "epoch": 0.4300737207285343, "grad_norm": 1.616088628768921, "learning_rate": 1.719498041491368e-05, "loss": 0.0703, "step": 3967 }, { "epoch": 0.43018213356461404, "grad_norm": 0.6049470901489258, "learning_rate": 1.7194255041346296e-05, "loss": 0.0054, "step": 3968 }, { "epoch": 0.43029054640069386, "grad_norm": 5.465906620025635, "learning_rate": 1.719352966777891e-05, "loss": 0.0959, "step": 3969 }, { "epoch": 0.4303989592367736, "grad_norm": 0.32841071486473083, "learning_rate": 1.719280429421152e-05, "loss": 0.0026, "step": 3970 }, { "epoch": 0.43050737207285344, "grad_norm": 1.839194893836975, "learning_rate": 1.7192078920644133e-05, "loss": 0.0416, "step": 3971 }, { "epoch": 0.4306157849089332, "grad_norm": 3.726544141769409, "learning_rate": 1.7191353547076745e-05, "loss": 0.0747, "step": 3972 }, { "epoch": 0.430724197745013, "grad_norm": 5.929450035095215, "learning_rate": 1.7190628173509357e-05, "loss": 0.079, "step": 3973 }, { "epoch": 0.4308326105810928, "grad_norm": 3.004753589630127, "learning_rate": 1.718990279994197e-05, "loss": 0.0385, "step": 3974 }, { "epoch": 0.4309410234171726, "grad_norm": 2.4569077491760254, "learning_rate": 1.7189177426374585e-05, "loss": 0.0427, "step": 3975 }, { "epoch": 0.43104943625325237, "grad_norm": 7.488972187042236, "learning_rate": 1.7188452052807197e-05, "loss": 0.0567, "step": 3976 }, { "epoch": 0.4311578490893322, "grad_norm": 3.0175909996032715, "learning_rate": 1.718772667923981e-05, "loss": 0.0737, "step": 3977 }, { "epoch": 0.43126626192541195, "grad_norm": 3.1368355751037598, "learning_rate": 1.718700130567242e-05, "loss": 0.0507, "step": 3978 }, { "epoch": 0.43137467476149177, "grad_norm": 4.842835903167725, "learning_rate": 1.7186275932105037e-05, "loss": 0.0203, "step": 3979 }, { "epoch": 0.43148308759757154, "grad_norm": 1.2319906949996948, "learning_rate": 1.718555055853765e-05, "loss": 0.0578, "step": 3980 }, { "epoch": 0.43159150043365135, "grad_norm": 3.560211181640625, "learning_rate": 1.718482518497026e-05, "loss": 0.0403, "step": 3981 }, { "epoch": 0.4316999132697311, "grad_norm": 1.4914109706878662, "learning_rate": 1.7184099811402873e-05, "loss": 0.0105, "step": 3982 }, { "epoch": 0.43180832610581094, "grad_norm": 1.9072421789169312, "learning_rate": 1.718337443783549e-05, "loss": 0.0298, "step": 3983 }, { "epoch": 0.4319167389418907, "grad_norm": 2.3855295181274414, "learning_rate": 1.71826490642681e-05, "loss": 0.0435, "step": 3984 }, { "epoch": 0.4320251517779705, "grad_norm": 1.555422067642212, "learning_rate": 1.7181923690700713e-05, "loss": 0.063, "step": 3985 }, { "epoch": 0.4321335646140503, "grad_norm": 3.9989688396453857, "learning_rate": 1.7181198317133325e-05, "loss": 0.0335, "step": 3986 }, { "epoch": 0.4322419774501301, "grad_norm": 2.056584358215332, "learning_rate": 1.7180472943565937e-05, "loss": 0.0513, "step": 3987 }, { "epoch": 0.43235039028620986, "grad_norm": 1.9703434705734253, "learning_rate": 1.717974756999855e-05, "loss": 0.0397, "step": 3988 }, { "epoch": 0.4324588031222897, "grad_norm": 1.2406401634216309, "learning_rate": 1.717902219643116e-05, "loss": 0.0294, "step": 3989 }, { "epoch": 0.43256721595836944, "grad_norm": 0.988775372505188, "learning_rate": 1.7178296822863777e-05, "loss": 0.0334, "step": 3990 }, { "epoch": 0.43267562879444926, "grad_norm": 3.4055047035217285, "learning_rate": 1.717757144929639e-05, "loss": 0.0714, "step": 3991 }, { "epoch": 0.4327840416305291, "grad_norm": 2.025782823562622, "learning_rate": 1.7176846075729e-05, "loss": 0.0099, "step": 3992 }, { "epoch": 0.43289245446660884, "grad_norm": 2.358355760574341, "learning_rate": 1.7176120702161614e-05, "loss": 0.0356, "step": 3993 }, { "epoch": 0.43300086730268866, "grad_norm": 2.1371371746063232, "learning_rate": 1.7175395328594226e-05, "loss": 0.0449, "step": 3994 }, { "epoch": 0.4331092801387684, "grad_norm": 2.4509084224700928, "learning_rate": 1.7174669955026838e-05, "loss": 0.0628, "step": 3995 }, { "epoch": 0.43321769297484825, "grad_norm": 5.133479595184326, "learning_rate": 1.7173944581459453e-05, "loss": 0.0268, "step": 3996 }, { "epoch": 0.433326105810928, "grad_norm": 0.8740076422691345, "learning_rate": 1.7173219207892066e-05, "loss": 0.0104, "step": 3997 }, { "epoch": 0.4334345186470078, "grad_norm": 1.7076495885849, "learning_rate": 1.717249383432468e-05, "loss": 0.0227, "step": 3998 }, { "epoch": 0.4335429314830876, "grad_norm": 1.0523003339767456, "learning_rate": 1.7171768460757293e-05, "loss": 0.0405, "step": 3999 }, { "epoch": 0.4336513443191674, "grad_norm": 0.6954062581062317, "learning_rate": 1.7171043087189905e-05, "loss": 0.0254, "step": 4000 }, { "epoch": 0.43375975715524717, "grad_norm": 2.4851999282836914, "learning_rate": 1.7170317713622517e-05, "loss": 0.0441, "step": 4001 }, { "epoch": 0.433868169991327, "grad_norm": 2.0998828411102295, "learning_rate": 1.716959234005513e-05, "loss": 0.041, "step": 4002 }, { "epoch": 0.43397658282740675, "grad_norm": 2.5015008449554443, "learning_rate": 1.7168866966487742e-05, "loss": 0.0573, "step": 4003 }, { "epoch": 0.4340849956634866, "grad_norm": 2.868825912475586, "learning_rate": 1.7168141592920354e-05, "loss": 0.0128, "step": 4004 }, { "epoch": 0.43419340849956634, "grad_norm": 2.2285304069519043, "learning_rate": 1.716741621935297e-05, "loss": 0.0565, "step": 4005 }, { "epoch": 0.43430182133564615, "grad_norm": 2.9005446434020996, "learning_rate": 1.716669084578558e-05, "loss": 0.0359, "step": 4006 }, { "epoch": 0.4344102341717259, "grad_norm": 1.697791337966919, "learning_rate": 1.7165965472218194e-05, "loss": 0.0359, "step": 4007 }, { "epoch": 0.43451864700780574, "grad_norm": 0.6469776034355164, "learning_rate": 1.7165240098650806e-05, "loss": 0.0162, "step": 4008 }, { "epoch": 0.4346270598438855, "grad_norm": 2.4828686714172363, "learning_rate": 1.7164514725083418e-05, "loss": 0.0506, "step": 4009 }, { "epoch": 0.4347354726799653, "grad_norm": 1.351083517074585, "learning_rate": 1.716378935151603e-05, "loss": 0.041, "step": 4010 }, { "epoch": 0.4348438855160451, "grad_norm": 1.6386427879333496, "learning_rate": 1.7163063977948646e-05, "loss": 0.048, "step": 4011 }, { "epoch": 0.4349522983521249, "grad_norm": 5.03855562210083, "learning_rate": 1.7162338604381258e-05, "loss": 0.0866, "step": 4012 }, { "epoch": 0.43506071118820466, "grad_norm": 3.5767366886138916, "learning_rate": 1.716161323081387e-05, "loss": 0.061, "step": 4013 }, { "epoch": 0.4351691240242845, "grad_norm": 2.195721387863159, "learning_rate": 1.7160887857246486e-05, "loss": 0.0376, "step": 4014 }, { "epoch": 0.43527753686036424, "grad_norm": 1.6016957759857178, "learning_rate": 1.7160162483679098e-05, "loss": 0.0071, "step": 4015 }, { "epoch": 0.43538594969644406, "grad_norm": 3.682194471359253, "learning_rate": 1.715943711011171e-05, "loss": 0.0171, "step": 4016 }, { "epoch": 0.4354943625325238, "grad_norm": 3.103516101837158, "learning_rate": 1.7158711736544322e-05, "loss": 0.0236, "step": 4017 }, { "epoch": 0.43560277536860365, "grad_norm": 3.8306338787078857, "learning_rate": 1.7157986362976934e-05, "loss": 0.0463, "step": 4018 }, { "epoch": 0.4357111882046834, "grad_norm": 3.35402512550354, "learning_rate": 1.7157260989409546e-05, "loss": 0.0332, "step": 4019 }, { "epoch": 0.4358196010407632, "grad_norm": 3.951366424560547, "learning_rate": 1.7156535615842162e-05, "loss": 0.0372, "step": 4020 }, { "epoch": 0.43592801387684305, "grad_norm": 2.7045514583587646, "learning_rate": 1.7155810242274774e-05, "loss": 0.0202, "step": 4021 }, { "epoch": 0.4360364267129228, "grad_norm": 4.779821872711182, "learning_rate": 1.7155084868707386e-05, "loss": 0.0557, "step": 4022 }, { "epoch": 0.4361448395490026, "grad_norm": 7.587190628051758, "learning_rate": 1.7154359495139998e-05, "loss": 0.0654, "step": 4023 }, { "epoch": 0.4362532523850824, "grad_norm": 6.700160026550293, "learning_rate": 1.715363412157261e-05, "loss": 0.0528, "step": 4024 }, { "epoch": 0.4363616652211622, "grad_norm": 4.477902412414551, "learning_rate": 1.7152908748005222e-05, "loss": 0.0643, "step": 4025 }, { "epoch": 0.436470078057242, "grad_norm": 2.0910849571228027, "learning_rate": 1.7152183374437838e-05, "loss": 0.0207, "step": 4026 }, { "epoch": 0.4365784908933218, "grad_norm": 4.500982284545898, "learning_rate": 1.715145800087045e-05, "loss": 0.0889, "step": 4027 }, { "epoch": 0.43668690372940155, "grad_norm": 4.49991512298584, "learning_rate": 1.7150732627303062e-05, "loss": 0.0427, "step": 4028 }, { "epoch": 0.4367953165654814, "grad_norm": 3.169245958328247, "learning_rate": 1.7150007253735674e-05, "loss": 0.0692, "step": 4029 }, { "epoch": 0.43690372940156114, "grad_norm": 2.7062978744506836, "learning_rate": 1.7149281880168287e-05, "loss": 0.0496, "step": 4030 }, { "epoch": 0.43701214223764095, "grad_norm": 1.9705978631973267, "learning_rate": 1.7148556506600902e-05, "loss": 0.0316, "step": 4031 }, { "epoch": 0.4371205550737207, "grad_norm": 2.3589985370635986, "learning_rate": 1.7147831133033514e-05, "loss": 0.0419, "step": 4032 }, { "epoch": 0.43722896790980054, "grad_norm": 3.027130603790283, "learning_rate": 1.7147105759466126e-05, "loss": 0.0372, "step": 4033 }, { "epoch": 0.4373373807458803, "grad_norm": 2.1684834957122803, "learning_rate": 1.714638038589874e-05, "loss": 0.0579, "step": 4034 }, { "epoch": 0.4374457935819601, "grad_norm": 1.3729344606399536, "learning_rate": 1.7145655012331354e-05, "loss": 0.0386, "step": 4035 }, { "epoch": 0.4375542064180399, "grad_norm": 2.6560728549957275, "learning_rate": 1.7144929638763966e-05, "loss": 0.0556, "step": 4036 }, { "epoch": 0.4376626192541197, "grad_norm": 3.5318727493286133, "learning_rate": 1.714420426519658e-05, "loss": 0.0153, "step": 4037 }, { "epoch": 0.43777103209019946, "grad_norm": 2.6073648929595947, "learning_rate": 1.714347889162919e-05, "loss": 0.0822, "step": 4038 }, { "epoch": 0.4378794449262793, "grad_norm": 1.9173190593719482, "learning_rate": 1.7142753518061803e-05, "loss": 0.0194, "step": 4039 }, { "epoch": 0.43798785776235905, "grad_norm": 2.949964761734009, "learning_rate": 1.7142028144494415e-05, "loss": 0.0397, "step": 4040 }, { "epoch": 0.43809627059843886, "grad_norm": 2.726497173309326, "learning_rate": 1.7141302770927027e-05, "loss": 0.0413, "step": 4041 }, { "epoch": 0.4382046834345186, "grad_norm": 1.9956257343292236, "learning_rate": 1.7140577397359642e-05, "loss": 0.0444, "step": 4042 }, { "epoch": 0.43831309627059845, "grad_norm": 2.1361427307128906, "learning_rate": 1.7139852023792255e-05, "loss": 0.047, "step": 4043 }, { "epoch": 0.4384215091066782, "grad_norm": 1.8776960372924805, "learning_rate": 1.7139126650224867e-05, "loss": 0.0299, "step": 4044 }, { "epoch": 0.438529921942758, "grad_norm": 1.9214670658111572, "learning_rate": 1.713840127665748e-05, "loss": 0.0284, "step": 4045 }, { "epoch": 0.4386383347788378, "grad_norm": 5.490294456481934, "learning_rate": 1.713767590309009e-05, "loss": 0.0559, "step": 4046 }, { "epoch": 0.4387467476149176, "grad_norm": 1.9311411380767822, "learning_rate": 1.7136950529522703e-05, "loss": 0.0155, "step": 4047 }, { "epoch": 0.4388551604509974, "grad_norm": 3.235219717025757, "learning_rate": 1.713622515595532e-05, "loss": 0.0208, "step": 4048 }, { "epoch": 0.4389635732870772, "grad_norm": 3.1352591514587402, "learning_rate": 1.713549978238793e-05, "loss": 0.0375, "step": 4049 }, { "epoch": 0.43907198612315695, "grad_norm": 2.1568334102630615, "learning_rate": 1.7134774408820546e-05, "loss": 0.0412, "step": 4050 }, { "epoch": 0.4391803989592368, "grad_norm": 0.9505302309989929, "learning_rate": 1.713404903525316e-05, "loss": 0.0234, "step": 4051 }, { "epoch": 0.4392888117953166, "grad_norm": 1.4035077095031738, "learning_rate": 1.713332366168577e-05, "loss": 0.0127, "step": 4052 }, { "epoch": 0.43939722463139635, "grad_norm": 2.305497884750366, "learning_rate": 1.7132598288118383e-05, "loss": 0.0255, "step": 4053 }, { "epoch": 0.4395056374674762, "grad_norm": 1.4112019538879395, "learning_rate": 1.7131872914550995e-05, "loss": 0.0274, "step": 4054 }, { "epoch": 0.43961405030355594, "grad_norm": 3.2528483867645264, "learning_rate": 1.7131147540983607e-05, "loss": 0.0312, "step": 4055 }, { "epoch": 0.43972246313963576, "grad_norm": 3.4793334007263184, "learning_rate": 1.713042216741622e-05, "loss": 0.05, "step": 4056 }, { "epoch": 0.4398308759757155, "grad_norm": 0.5214948058128357, "learning_rate": 1.7129696793848835e-05, "loss": 0.0063, "step": 4057 }, { "epoch": 0.43993928881179534, "grad_norm": 2.2662291526794434, "learning_rate": 1.7128971420281447e-05, "loss": 0.03, "step": 4058 }, { "epoch": 0.4400477016478751, "grad_norm": 2.3205647468566895, "learning_rate": 1.712824604671406e-05, "loss": 0.0566, "step": 4059 }, { "epoch": 0.4401561144839549, "grad_norm": 5.586921691894531, "learning_rate": 1.712752067314667e-05, "loss": 0.0215, "step": 4060 }, { "epoch": 0.4402645273200347, "grad_norm": 2.9690043926239014, "learning_rate": 1.7126795299579283e-05, "loss": 0.0483, "step": 4061 }, { "epoch": 0.4403729401561145, "grad_norm": 1.7101402282714844, "learning_rate": 1.7126069926011896e-05, "loss": 0.023, "step": 4062 }, { "epoch": 0.44048135299219426, "grad_norm": 6.172093391418457, "learning_rate": 1.712534455244451e-05, "loss": 0.0122, "step": 4063 }, { "epoch": 0.4405897658282741, "grad_norm": 4.531774044036865, "learning_rate": 1.7124619178877123e-05, "loss": 0.0552, "step": 4064 }, { "epoch": 0.44069817866435385, "grad_norm": 1.0711787939071655, "learning_rate": 1.7123893805309735e-05, "loss": 0.0345, "step": 4065 }, { "epoch": 0.44080659150043366, "grad_norm": 2.9214026927948, "learning_rate": 1.712316843174235e-05, "loss": 0.0137, "step": 4066 }, { "epoch": 0.4409150043365134, "grad_norm": 1.8300769329071045, "learning_rate": 1.7122443058174963e-05, "loss": 0.077, "step": 4067 }, { "epoch": 0.44102341717259325, "grad_norm": 1.1170402765274048, "learning_rate": 1.7121717684607575e-05, "loss": 0.0366, "step": 4068 }, { "epoch": 0.441131830008673, "grad_norm": 3.8436331748962402, "learning_rate": 1.7120992311040187e-05, "loss": 0.0295, "step": 4069 }, { "epoch": 0.44124024284475283, "grad_norm": 2.792583465576172, "learning_rate": 1.71202669374728e-05, "loss": 0.0238, "step": 4070 }, { "epoch": 0.4413486556808326, "grad_norm": 1.776976466178894, "learning_rate": 1.711954156390541e-05, "loss": 0.0198, "step": 4071 }, { "epoch": 0.4414570685169124, "grad_norm": 4.262516021728516, "learning_rate": 1.7118816190338027e-05, "loss": 0.0332, "step": 4072 }, { "epoch": 0.4415654813529922, "grad_norm": 3.632344961166382, "learning_rate": 1.711809081677064e-05, "loss": 0.0153, "step": 4073 }, { "epoch": 0.441673894189072, "grad_norm": 3.0890491008758545, "learning_rate": 1.711736544320325e-05, "loss": 0.0245, "step": 4074 }, { "epoch": 0.44178230702515175, "grad_norm": 2.8264975547790527, "learning_rate": 1.7116640069635864e-05, "loss": 0.115, "step": 4075 }, { "epoch": 0.4418907198612316, "grad_norm": 2.4254751205444336, "learning_rate": 1.7115914696068476e-05, "loss": 0.05, "step": 4076 }, { "epoch": 0.44199913269731134, "grad_norm": 4.747281551361084, "learning_rate": 1.7115189322501088e-05, "loss": 0.0454, "step": 4077 }, { "epoch": 0.44210754553339116, "grad_norm": 1.8747954368591309, "learning_rate": 1.7114463948933703e-05, "loss": 0.0791, "step": 4078 }, { "epoch": 0.4422159583694709, "grad_norm": 1.6462746858596802, "learning_rate": 1.7113738575366315e-05, "loss": 0.04, "step": 4079 }, { "epoch": 0.44232437120555074, "grad_norm": 1.191345453262329, "learning_rate": 1.7113013201798928e-05, "loss": 0.0061, "step": 4080 }, { "epoch": 0.44243278404163056, "grad_norm": 2.1232645511627197, "learning_rate": 1.711228782823154e-05, "loss": 0.0364, "step": 4081 }, { "epoch": 0.4425411968777103, "grad_norm": 5.583446502685547, "learning_rate": 1.7111562454664152e-05, "loss": 0.0322, "step": 4082 }, { "epoch": 0.44264960971379014, "grad_norm": 2.881477117538452, "learning_rate": 1.7110837081096767e-05, "loss": 0.0338, "step": 4083 }, { "epoch": 0.4427580225498699, "grad_norm": 2.7997066974639893, "learning_rate": 1.711011170752938e-05, "loss": 0.0297, "step": 4084 }, { "epoch": 0.4428664353859497, "grad_norm": 2.215399742126465, "learning_rate": 1.7109386333961992e-05, "loss": 0.0195, "step": 4085 }, { "epoch": 0.4429748482220295, "grad_norm": 0.8736152648925781, "learning_rate": 1.7108660960394604e-05, "loss": 0.0294, "step": 4086 }, { "epoch": 0.4430832610581093, "grad_norm": 2.6539456844329834, "learning_rate": 1.710793558682722e-05, "loss": 0.0409, "step": 4087 }, { "epoch": 0.44319167389418906, "grad_norm": 1.0365619659423828, "learning_rate": 1.710721021325983e-05, "loss": 0.015, "step": 4088 }, { "epoch": 0.4433000867302689, "grad_norm": 1.7912477254867554, "learning_rate": 1.7106484839692444e-05, "loss": 0.0582, "step": 4089 }, { "epoch": 0.44340849956634865, "grad_norm": 3.889514446258545, "learning_rate": 1.7105759466125056e-05, "loss": 0.0389, "step": 4090 }, { "epoch": 0.44351691240242846, "grad_norm": 0.3824092149734497, "learning_rate": 1.7105034092557668e-05, "loss": 0.0092, "step": 4091 }, { "epoch": 0.44362532523850823, "grad_norm": 3.5200836658477783, "learning_rate": 1.710430871899028e-05, "loss": 0.0878, "step": 4092 }, { "epoch": 0.44373373807458805, "grad_norm": 5.608972072601318, "learning_rate": 1.7103583345422892e-05, "loss": 0.0448, "step": 4093 }, { "epoch": 0.4438421509106678, "grad_norm": 2.512906074523926, "learning_rate": 1.7102857971855508e-05, "loss": 0.0526, "step": 4094 }, { "epoch": 0.44395056374674763, "grad_norm": 0.6422743797302246, "learning_rate": 1.710213259828812e-05, "loss": 0.0203, "step": 4095 }, { "epoch": 0.4440589765828274, "grad_norm": 2.054499387741089, "learning_rate": 1.7101407224720732e-05, "loss": 0.0388, "step": 4096 }, { "epoch": 0.4441673894189072, "grad_norm": 0.9149767160415649, "learning_rate": 1.7100681851153344e-05, "loss": 0.0238, "step": 4097 }, { "epoch": 0.444275802254987, "grad_norm": 4.549959659576416, "learning_rate": 1.7099956477585956e-05, "loss": 0.0361, "step": 4098 }, { "epoch": 0.4443842150910668, "grad_norm": 4.692568778991699, "learning_rate": 1.709923110401857e-05, "loss": 0.0501, "step": 4099 }, { "epoch": 0.44449262792714656, "grad_norm": 0.7909969687461853, "learning_rate": 1.7098505730451184e-05, "loss": 0.018, "step": 4100 }, { "epoch": 0.4446010407632264, "grad_norm": 1.4978303909301758, "learning_rate": 1.7097780356883796e-05, "loss": 0.0391, "step": 4101 }, { "epoch": 0.44470945359930614, "grad_norm": 4.388465881347656, "learning_rate": 1.7097054983316412e-05, "loss": 0.0671, "step": 4102 }, { "epoch": 0.44481786643538596, "grad_norm": 1.6174577474594116, "learning_rate": 1.7096329609749024e-05, "loss": 0.0304, "step": 4103 }, { "epoch": 0.4449262792714657, "grad_norm": 2.151261568069458, "learning_rate": 1.7095604236181636e-05, "loss": 0.0439, "step": 4104 }, { "epoch": 0.44503469210754554, "grad_norm": 1.3429919481277466, "learning_rate": 1.7094878862614248e-05, "loss": 0.0456, "step": 4105 }, { "epoch": 0.4451431049436253, "grad_norm": 0.6908811330795288, "learning_rate": 1.709415348904686e-05, "loss": 0.0202, "step": 4106 }, { "epoch": 0.4452515177797051, "grad_norm": 3.9153683185577393, "learning_rate": 1.7093428115479472e-05, "loss": 0.0516, "step": 4107 }, { "epoch": 0.4453599306157849, "grad_norm": 4.150365352630615, "learning_rate": 1.7092702741912085e-05, "loss": 0.0291, "step": 4108 }, { "epoch": 0.4454683434518647, "grad_norm": 2.816427707672119, "learning_rate": 1.70919773683447e-05, "loss": 0.0495, "step": 4109 }, { "epoch": 0.4455767562879445, "grad_norm": 2.227051019668579, "learning_rate": 1.7091251994777312e-05, "loss": 0.0238, "step": 4110 }, { "epoch": 0.4456851691240243, "grad_norm": 1.3126423358917236, "learning_rate": 1.7090526621209924e-05, "loss": 0.0295, "step": 4111 }, { "epoch": 0.4457935819601041, "grad_norm": 0.42188677191734314, "learning_rate": 1.7089801247642537e-05, "loss": 0.0073, "step": 4112 }, { "epoch": 0.44590199479618386, "grad_norm": 2.663299322128296, "learning_rate": 1.708907587407515e-05, "loss": 0.0611, "step": 4113 }, { "epoch": 0.4460104076322637, "grad_norm": 4.201302528381348, "learning_rate": 1.708835050050776e-05, "loss": 0.0347, "step": 4114 }, { "epoch": 0.44611882046834345, "grad_norm": 1.2973098754882812, "learning_rate": 1.7087625126940376e-05, "loss": 0.0259, "step": 4115 }, { "epoch": 0.44622723330442327, "grad_norm": 6.280114650726318, "learning_rate": 1.708689975337299e-05, "loss": 0.0694, "step": 4116 }, { "epoch": 0.44633564614050303, "grad_norm": 2.4953696727752686, "learning_rate": 1.70861743798056e-05, "loss": 0.0425, "step": 4117 }, { "epoch": 0.44644405897658285, "grad_norm": 3.877063274383545, "learning_rate": 1.7085449006238216e-05, "loss": 0.035, "step": 4118 }, { "epoch": 0.4465524718126626, "grad_norm": 7.539348602294922, "learning_rate": 1.708472363267083e-05, "loss": 0.0316, "step": 4119 }, { "epoch": 0.44666088464874243, "grad_norm": 0.8047823905944824, "learning_rate": 1.708399825910344e-05, "loss": 0.0165, "step": 4120 }, { "epoch": 0.4467692974848222, "grad_norm": 3.2665951251983643, "learning_rate": 1.7083272885536053e-05, "loss": 0.0889, "step": 4121 }, { "epoch": 0.446877710320902, "grad_norm": 2.02351713180542, "learning_rate": 1.7082547511968665e-05, "loss": 0.0097, "step": 4122 }, { "epoch": 0.4469861231569818, "grad_norm": 2.8455612659454346, "learning_rate": 1.7081822138401277e-05, "loss": 0.0253, "step": 4123 }, { "epoch": 0.4470945359930616, "grad_norm": 2.656209945678711, "learning_rate": 1.7081096764833892e-05, "loss": 0.0726, "step": 4124 }, { "epoch": 0.44720294882914136, "grad_norm": 4.980937480926514, "learning_rate": 1.7080371391266505e-05, "loss": 0.0449, "step": 4125 }, { "epoch": 0.4473113616652212, "grad_norm": 2.209826946258545, "learning_rate": 1.7079646017699117e-05, "loss": 0.0443, "step": 4126 }, { "epoch": 0.44741977450130094, "grad_norm": 1.8757528066635132, "learning_rate": 1.707892064413173e-05, "loss": 0.0343, "step": 4127 }, { "epoch": 0.44752818733738076, "grad_norm": 2.641447067260742, "learning_rate": 1.707819527056434e-05, "loss": 0.0309, "step": 4128 }, { "epoch": 0.4476366001734605, "grad_norm": 1.1953604221343994, "learning_rate": 1.7077469896996953e-05, "loss": 0.0399, "step": 4129 }, { "epoch": 0.44774501300954034, "grad_norm": 2.3400144577026367, "learning_rate": 1.707674452342957e-05, "loss": 0.0481, "step": 4130 }, { "epoch": 0.4478534258456201, "grad_norm": 1.4176874160766602, "learning_rate": 1.707601914986218e-05, "loss": 0.0532, "step": 4131 }, { "epoch": 0.4479618386816999, "grad_norm": 2.0341615676879883, "learning_rate": 1.7075293776294793e-05, "loss": 0.0377, "step": 4132 }, { "epoch": 0.4480702515177797, "grad_norm": 7.312000751495361, "learning_rate": 1.7074568402727405e-05, "loss": 0.0356, "step": 4133 }, { "epoch": 0.4481786643538595, "grad_norm": 0.8699418902397156, "learning_rate": 1.707384302916002e-05, "loss": 0.02, "step": 4134 }, { "epoch": 0.44828707718993926, "grad_norm": 0.9232418537139893, "learning_rate": 1.7073117655592633e-05, "loss": 0.0144, "step": 4135 }, { "epoch": 0.4483954900260191, "grad_norm": 3.3722712993621826, "learning_rate": 1.7072392282025245e-05, "loss": 0.0559, "step": 4136 }, { "epoch": 0.44850390286209885, "grad_norm": 3.9836013317108154, "learning_rate": 1.7071666908457857e-05, "loss": 0.0691, "step": 4137 }, { "epoch": 0.44861231569817867, "grad_norm": 2.6517107486724854, "learning_rate": 1.707094153489047e-05, "loss": 0.0287, "step": 4138 }, { "epoch": 0.4487207285342585, "grad_norm": 2.424276351928711, "learning_rate": 1.7070216161323085e-05, "loss": 0.0272, "step": 4139 }, { "epoch": 0.44882914137033825, "grad_norm": 1.5060805082321167, "learning_rate": 1.7069490787755697e-05, "loss": 0.0328, "step": 4140 }, { "epoch": 0.44893755420641807, "grad_norm": 1.4769948720932007, "learning_rate": 1.706876541418831e-05, "loss": 0.0528, "step": 4141 }, { "epoch": 0.44904596704249783, "grad_norm": 3.6116793155670166, "learning_rate": 1.706804004062092e-05, "loss": 0.0597, "step": 4142 }, { "epoch": 0.44915437987857765, "grad_norm": 1.9468562602996826, "learning_rate": 1.7067314667053533e-05, "loss": 0.0282, "step": 4143 }, { "epoch": 0.4492627927146574, "grad_norm": 2.410341501235962, "learning_rate": 1.7066589293486145e-05, "loss": 0.0414, "step": 4144 }, { "epoch": 0.44937120555073723, "grad_norm": 2.860220432281494, "learning_rate": 1.7065863919918758e-05, "loss": 0.0699, "step": 4145 }, { "epoch": 0.449479618386817, "grad_norm": 5.622804164886475, "learning_rate": 1.7065138546351373e-05, "loss": 0.0771, "step": 4146 }, { "epoch": 0.4495880312228968, "grad_norm": 5.134733200073242, "learning_rate": 1.7064413172783985e-05, "loss": 0.0557, "step": 4147 }, { "epoch": 0.4496964440589766, "grad_norm": 1.7191590070724487, "learning_rate": 1.7063687799216597e-05, "loss": 0.0575, "step": 4148 }, { "epoch": 0.4498048568950564, "grad_norm": 2.4072792530059814, "learning_rate": 1.706296242564921e-05, "loss": 0.0823, "step": 4149 }, { "epoch": 0.44991326973113616, "grad_norm": 6.51295280456543, "learning_rate": 1.7062237052081822e-05, "loss": 0.023, "step": 4150 }, { "epoch": 0.450021682567216, "grad_norm": 2.192826271057129, "learning_rate": 1.7061511678514434e-05, "loss": 0.0319, "step": 4151 }, { "epoch": 0.45013009540329574, "grad_norm": 1.6214839220046997, "learning_rate": 1.706078630494705e-05, "loss": 0.0263, "step": 4152 }, { "epoch": 0.45023850823937556, "grad_norm": 3.0872483253479004, "learning_rate": 1.706006093137966e-05, "loss": 0.1309, "step": 4153 }, { "epoch": 0.4503469210754553, "grad_norm": 2.5838732719421387, "learning_rate": 1.7059335557812277e-05, "loss": 0.0358, "step": 4154 }, { "epoch": 0.45045533391153514, "grad_norm": 3.2762138843536377, "learning_rate": 1.705861018424489e-05, "loss": 0.0305, "step": 4155 }, { "epoch": 0.4505637467476149, "grad_norm": 1.0841648578643799, "learning_rate": 1.70578848106775e-05, "loss": 0.0227, "step": 4156 }, { "epoch": 0.4506721595836947, "grad_norm": 3.6993837356567383, "learning_rate": 1.7057159437110114e-05, "loss": 0.0891, "step": 4157 }, { "epoch": 0.4507805724197745, "grad_norm": 1.8347760438919067, "learning_rate": 1.7056434063542726e-05, "loss": 0.0301, "step": 4158 }, { "epoch": 0.4508889852558543, "grad_norm": 2.921689033508301, "learning_rate": 1.7055708689975338e-05, "loss": 0.0459, "step": 4159 }, { "epoch": 0.45099739809193407, "grad_norm": 0.9047724008560181, "learning_rate": 1.705498331640795e-05, "loss": 0.023, "step": 4160 }, { "epoch": 0.4511058109280139, "grad_norm": 1.6015448570251465, "learning_rate": 1.7054257942840565e-05, "loss": 0.029, "step": 4161 }, { "epoch": 0.45121422376409365, "grad_norm": 1.7342301607131958, "learning_rate": 1.7053532569273178e-05, "loss": 0.032, "step": 4162 }, { "epoch": 0.45132263660017347, "grad_norm": 4.915634632110596, "learning_rate": 1.705280719570579e-05, "loss": 0.0493, "step": 4163 }, { "epoch": 0.45143104943625323, "grad_norm": 3.04923152923584, "learning_rate": 1.7052081822138402e-05, "loss": 0.0553, "step": 4164 }, { "epoch": 0.45153946227233305, "grad_norm": 3.1209030151367188, "learning_rate": 1.7051356448571014e-05, "loss": 0.0327, "step": 4165 }, { "epoch": 0.4516478751084128, "grad_norm": 3.9464175701141357, "learning_rate": 1.7050631075003626e-05, "loss": 0.0725, "step": 4166 }, { "epoch": 0.45175628794449263, "grad_norm": 1.7728815078735352, "learning_rate": 1.7049905701436242e-05, "loss": 0.0273, "step": 4167 }, { "epoch": 0.4518647007805724, "grad_norm": 1.814340353012085, "learning_rate": 1.7049180327868854e-05, "loss": 0.0326, "step": 4168 }, { "epoch": 0.4519731136166522, "grad_norm": 5.236100673675537, "learning_rate": 1.7048454954301466e-05, "loss": 0.0462, "step": 4169 }, { "epoch": 0.45208152645273203, "grad_norm": 2.3274130821228027, "learning_rate": 1.704772958073408e-05, "loss": 0.0457, "step": 4170 }, { "epoch": 0.4521899392888118, "grad_norm": 3.4954686164855957, "learning_rate": 1.7047004207166694e-05, "loss": 0.0537, "step": 4171 }, { "epoch": 0.4522983521248916, "grad_norm": 1.6593172550201416, "learning_rate": 1.7046278833599306e-05, "loss": 0.0249, "step": 4172 }, { "epoch": 0.4524067649609714, "grad_norm": 2.8891959190368652, "learning_rate": 1.7045553460031918e-05, "loss": 0.0874, "step": 4173 }, { "epoch": 0.4525151777970512, "grad_norm": 3.560152530670166, "learning_rate": 1.704482808646453e-05, "loss": 0.0761, "step": 4174 }, { "epoch": 0.45262359063313096, "grad_norm": 2.050422191619873, "learning_rate": 1.7044102712897142e-05, "loss": 0.036, "step": 4175 }, { "epoch": 0.4527320034692108, "grad_norm": 3.6053266525268555, "learning_rate": 1.7043377339329758e-05, "loss": 0.0546, "step": 4176 }, { "epoch": 0.45284041630529054, "grad_norm": 3.9843978881835938, "learning_rate": 1.704265196576237e-05, "loss": 0.0377, "step": 4177 }, { "epoch": 0.45294882914137036, "grad_norm": 1.1998831033706665, "learning_rate": 1.7041926592194982e-05, "loss": 0.0121, "step": 4178 }, { "epoch": 0.4530572419774501, "grad_norm": 1.9366780519485474, "learning_rate": 1.7041201218627594e-05, "loss": 0.0562, "step": 4179 }, { "epoch": 0.45316565481352994, "grad_norm": 3.5747315883636475, "learning_rate": 1.7040475845060206e-05, "loss": 0.072, "step": 4180 }, { "epoch": 0.4532740676496097, "grad_norm": 2.1172525882720947, "learning_rate": 1.703975047149282e-05, "loss": 0.0473, "step": 4181 }, { "epoch": 0.4533824804856895, "grad_norm": 1.7900042533874512, "learning_rate": 1.7039025097925434e-05, "loss": 0.0407, "step": 4182 }, { "epoch": 0.4534908933217693, "grad_norm": 1.4466557502746582, "learning_rate": 1.7038299724358046e-05, "loss": 0.0233, "step": 4183 }, { "epoch": 0.4535993061578491, "grad_norm": 5.030383110046387, "learning_rate": 1.703757435079066e-05, "loss": 0.0351, "step": 4184 }, { "epoch": 0.45370771899392887, "grad_norm": 1.971287488937378, "learning_rate": 1.703684897722327e-05, "loss": 0.0366, "step": 4185 }, { "epoch": 0.4538161318300087, "grad_norm": 1.6173181533813477, "learning_rate": 1.7036123603655886e-05, "loss": 0.0207, "step": 4186 }, { "epoch": 0.45392454466608845, "grad_norm": 2.093975782394409, "learning_rate": 1.7035398230088498e-05, "loss": 0.023, "step": 4187 }, { "epoch": 0.45403295750216827, "grad_norm": 1.8641973733901978, "learning_rate": 1.703467285652111e-05, "loss": 0.0405, "step": 4188 }, { "epoch": 0.45414137033824803, "grad_norm": 0.9145217537879944, "learning_rate": 1.7033947482953722e-05, "loss": 0.0132, "step": 4189 }, { "epoch": 0.45424978317432785, "grad_norm": 4.236702919006348, "learning_rate": 1.7033222109386335e-05, "loss": 0.0658, "step": 4190 }, { "epoch": 0.4543581960104076, "grad_norm": 2.291621208190918, "learning_rate": 1.703249673581895e-05, "loss": 0.0191, "step": 4191 }, { "epoch": 0.45446660884648743, "grad_norm": 4.493826389312744, "learning_rate": 1.7031771362251562e-05, "loss": 0.0424, "step": 4192 }, { "epoch": 0.4545750216825672, "grad_norm": 0.5808201432228088, "learning_rate": 1.7031045988684174e-05, "loss": 0.0182, "step": 4193 }, { "epoch": 0.454683434518647, "grad_norm": 6.843381404876709, "learning_rate": 1.7030320615116787e-05, "loss": 0.08, "step": 4194 }, { "epoch": 0.4547918473547268, "grad_norm": 2.2693402767181396, "learning_rate": 1.70295952415494e-05, "loss": 0.0536, "step": 4195 }, { "epoch": 0.4549002601908066, "grad_norm": 3.9418489933013916, "learning_rate": 1.702886986798201e-05, "loss": 0.0714, "step": 4196 }, { "epoch": 0.45500867302688636, "grad_norm": 5.141482830047607, "learning_rate": 1.7028144494414623e-05, "loss": 0.0549, "step": 4197 }, { "epoch": 0.4551170858629662, "grad_norm": 0.5586756467819214, "learning_rate": 1.702741912084724e-05, "loss": 0.0091, "step": 4198 }, { "epoch": 0.455225498699046, "grad_norm": 3.6421396732330322, "learning_rate": 1.702669374727985e-05, "loss": 0.0402, "step": 4199 }, { "epoch": 0.45533391153512576, "grad_norm": 0.5039690732955933, "learning_rate": 1.7025968373712463e-05, "loss": 0.0027, "step": 4200 }, { "epoch": 0.4554423243712056, "grad_norm": 5.3744611740112305, "learning_rate": 1.7025243000145075e-05, "loss": 0.0442, "step": 4201 }, { "epoch": 0.45555073720728534, "grad_norm": 4.423074245452881, "learning_rate": 1.7024517626577687e-05, "loss": 0.0376, "step": 4202 }, { "epoch": 0.45565915004336516, "grad_norm": 0.6461367011070251, "learning_rate": 1.7023792253010303e-05, "loss": 0.0095, "step": 4203 }, { "epoch": 0.4557675628794449, "grad_norm": 8.247705459594727, "learning_rate": 1.7023066879442915e-05, "loss": 0.0589, "step": 4204 }, { "epoch": 0.45587597571552474, "grad_norm": 1.7636185884475708, "learning_rate": 1.7022341505875527e-05, "loss": 0.0429, "step": 4205 }, { "epoch": 0.4559843885516045, "grad_norm": 2.080352783203125, "learning_rate": 1.7021616132308142e-05, "loss": 0.0441, "step": 4206 }, { "epoch": 0.4560928013876843, "grad_norm": 1.610559105873108, "learning_rate": 1.7020890758740755e-05, "loss": 0.0364, "step": 4207 }, { "epoch": 0.4562012142237641, "grad_norm": 0.2022242546081543, "learning_rate": 1.7020165385173367e-05, "loss": 0.0017, "step": 4208 }, { "epoch": 0.4563096270598439, "grad_norm": 6.474970817565918, "learning_rate": 1.701944001160598e-05, "loss": 0.0856, "step": 4209 }, { "epoch": 0.45641803989592367, "grad_norm": 2.0625104904174805, "learning_rate": 1.701871463803859e-05, "loss": 0.0093, "step": 4210 }, { "epoch": 0.4565264527320035, "grad_norm": 2.4536876678466797, "learning_rate": 1.7017989264471203e-05, "loss": 0.0461, "step": 4211 }, { "epoch": 0.45663486556808325, "grad_norm": 0.8265190124511719, "learning_rate": 1.7017263890903815e-05, "loss": 0.0202, "step": 4212 }, { "epoch": 0.45674327840416307, "grad_norm": 1.8108584880828857, "learning_rate": 1.701653851733643e-05, "loss": 0.0601, "step": 4213 }, { "epoch": 0.45685169124024283, "grad_norm": 1.846864938735962, "learning_rate": 1.7015813143769043e-05, "loss": 0.0343, "step": 4214 }, { "epoch": 0.45696010407632265, "grad_norm": 3.091521978378296, "learning_rate": 1.7015087770201655e-05, "loss": 0.0361, "step": 4215 }, { "epoch": 0.4570685169124024, "grad_norm": 3.064049243927002, "learning_rate": 1.7014362396634267e-05, "loss": 0.0558, "step": 4216 }, { "epoch": 0.45717692974848223, "grad_norm": 2.240859031677246, "learning_rate": 1.701363702306688e-05, "loss": 0.0309, "step": 4217 }, { "epoch": 0.457285342584562, "grad_norm": 4.456634521484375, "learning_rate": 1.701291164949949e-05, "loss": 0.0663, "step": 4218 }, { "epoch": 0.4573937554206418, "grad_norm": 1.5163387060165405, "learning_rate": 1.7012186275932107e-05, "loss": 0.0077, "step": 4219 }, { "epoch": 0.4575021682567216, "grad_norm": 1.00385582447052, "learning_rate": 1.701146090236472e-05, "loss": 0.0202, "step": 4220 }, { "epoch": 0.4576105810928014, "grad_norm": 2.3306236267089844, "learning_rate": 1.7010735528797335e-05, "loss": 0.05, "step": 4221 }, { "epoch": 0.45771899392888116, "grad_norm": 1.0880568027496338, "learning_rate": 1.7010010155229947e-05, "loss": 0.0259, "step": 4222 }, { "epoch": 0.457827406764961, "grad_norm": 1.8429557085037231, "learning_rate": 1.700928478166256e-05, "loss": 0.0319, "step": 4223 }, { "epoch": 0.45793581960104074, "grad_norm": 2.038769483566284, "learning_rate": 1.700855940809517e-05, "loss": 0.0324, "step": 4224 }, { "epoch": 0.45804423243712056, "grad_norm": 1.9015917778015137, "learning_rate": 1.7007834034527783e-05, "loss": 0.0454, "step": 4225 }, { "epoch": 0.4581526452732003, "grad_norm": 4.576028347015381, "learning_rate": 1.7007108660960395e-05, "loss": 0.0705, "step": 4226 }, { "epoch": 0.45826105810928014, "grad_norm": 1.2785494327545166, "learning_rate": 1.7006383287393008e-05, "loss": 0.0216, "step": 4227 }, { "epoch": 0.45836947094535996, "grad_norm": 2.565514326095581, "learning_rate": 1.7005657913825623e-05, "loss": 0.0337, "step": 4228 }, { "epoch": 0.4584778837814397, "grad_norm": 2.7807862758636475, "learning_rate": 1.7004932540258235e-05, "loss": 0.0299, "step": 4229 }, { "epoch": 0.45858629661751954, "grad_norm": 2.568293333053589, "learning_rate": 1.7004207166690847e-05, "loss": 0.0487, "step": 4230 }, { "epoch": 0.4586947094535993, "grad_norm": 2.7862110137939453, "learning_rate": 1.700348179312346e-05, "loss": 0.0227, "step": 4231 }, { "epoch": 0.4588031222896791, "grad_norm": 1.4013466835021973, "learning_rate": 1.700275641955607e-05, "loss": 0.0347, "step": 4232 }, { "epoch": 0.4589115351257589, "grad_norm": 1.6225496530532837, "learning_rate": 1.7002031045988684e-05, "loss": 0.0154, "step": 4233 }, { "epoch": 0.4590199479618387, "grad_norm": 1.1614794731140137, "learning_rate": 1.70013056724213e-05, "loss": 0.02, "step": 4234 }, { "epoch": 0.45912836079791847, "grad_norm": 3.8633978366851807, "learning_rate": 1.700058029885391e-05, "loss": 0.0562, "step": 4235 }, { "epoch": 0.4592367736339983, "grad_norm": 2.072845697402954, "learning_rate": 1.6999854925286524e-05, "loss": 0.0415, "step": 4236 }, { "epoch": 0.45934518647007805, "grad_norm": 2.8007917404174805, "learning_rate": 1.6999129551719136e-05, "loss": 0.0357, "step": 4237 }, { "epoch": 0.45945359930615787, "grad_norm": 2.0618577003479004, "learning_rate": 1.699840417815175e-05, "loss": 0.043, "step": 4238 }, { "epoch": 0.45956201214223763, "grad_norm": 2.7090160846710205, "learning_rate": 1.6997678804584363e-05, "loss": 0.0621, "step": 4239 }, { "epoch": 0.45967042497831745, "grad_norm": 3.632190227508545, "learning_rate": 1.6996953431016976e-05, "loss": 0.0383, "step": 4240 }, { "epoch": 0.4597788378143972, "grad_norm": 2.413990020751953, "learning_rate": 1.6996228057449588e-05, "loss": 0.0105, "step": 4241 }, { "epoch": 0.45988725065047703, "grad_norm": 1.0150880813598633, "learning_rate": 1.69955026838822e-05, "loss": 0.0064, "step": 4242 }, { "epoch": 0.4599956634865568, "grad_norm": 4.166940212249756, "learning_rate": 1.6994777310314815e-05, "loss": 0.0395, "step": 4243 }, { "epoch": 0.4601040763226366, "grad_norm": 4.825184345245361, "learning_rate": 1.6994051936747428e-05, "loss": 0.0375, "step": 4244 }, { "epoch": 0.4602124891587164, "grad_norm": 5.777692794799805, "learning_rate": 1.699332656318004e-05, "loss": 0.1375, "step": 4245 }, { "epoch": 0.4603209019947962, "grad_norm": 0.9607139229774475, "learning_rate": 1.6992601189612652e-05, "loss": 0.0083, "step": 4246 }, { "epoch": 0.46042931483087596, "grad_norm": 2.4219908714294434, "learning_rate": 1.6991875816045264e-05, "loss": 0.0386, "step": 4247 }, { "epoch": 0.4605377276669558, "grad_norm": 2.9984493255615234, "learning_rate": 1.6991150442477876e-05, "loss": 0.057, "step": 4248 }, { "epoch": 0.46064614050303554, "grad_norm": 0.9550908207893372, "learning_rate": 1.6990425068910488e-05, "loss": 0.0198, "step": 4249 }, { "epoch": 0.46075455333911536, "grad_norm": 4.064596176147461, "learning_rate": 1.6989699695343104e-05, "loss": 0.0331, "step": 4250 }, { "epoch": 0.4608629661751951, "grad_norm": 1.1921106576919556, "learning_rate": 1.6988974321775716e-05, "loss": 0.027, "step": 4251 }, { "epoch": 0.46097137901127494, "grad_norm": 5.952508926391602, "learning_rate": 1.6988248948208328e-05, "loss": 0.018, "step": 4252 }, { "epoch": 0.4610797918473547, "grad_norm": 2.9709854125976562, "learning_rate": 1.698752357464094e-05, "loss": 0.0275, "step": 4253 }, { "epoch": 0.4611882046834345, "grad_norm": 3.7886886596679688, "learning_rate": 1.6986798201073552e-05, "loss": 0.0742, "step": 4254 }, { "epoch": 0.4612966175195143, "grad_norm": 4.221798896789551, "learning_rate": 1.6986072827506168e-05, "loss": 0.069, "step": 4255 }, { "epoch": 0.4614050303555941, "grad_norm": 0.40497440099716187, "learning_rate": 1.698534745393878e-05, "loss": 0.0143, "step": 4256 }, { "epoch": 0.46151344319167387, "grad_norm": 5.936490058898926, "learning_rate": 1.6984622080371392e-05, "loss": 0.0522, "step": 4257 }, { "epoch": 0.4616218560277537, "grad_norm": 0.9298095107078552, "learning_rate": 1.6983896706804008e-05, "loss": 0.0269, "step": 4258 }, { "epoch": 0.4617302688638335, "grad_norm": 4.467740535736084, "learning_rate": 1.698317133323662e-05, "loss": 0.0555, "step": 4259 }, { "epoch": 0.46183868169991327, "grad_norm": 4.130599498748779, "learning_rate": 1.6982445959669232e-05, "loss": 0.065, "step": 4260 }, { "epoch": 0.4619470945359931, "grad_norm": 2.6261579990386963, "learning_rate": 1.6981720586101844e-05, "loss": 0.0254, "step": 4261 }, { "epoch": 0.46205550737207285, "grad_norm": 6.10949182510376, "learning_rate": 1.6980995212534456e-05, "loss": 0.069, "step": 4262 }, { "epoch": 0.46216392020815267, "grad_norm": 2.558134078979492, "learning_rate": 1.698026983896707e-05, "loss": 0.0306, "step": 4263 }, { "epoch": 0.46227233304423243, "grad_norm": 7.658456802368164, "learning_rate": 1.697954446539968e-05, "loss": 0.0354, "step": 4264 }, { "epoch": 0.46238074588031225, "grad_norm": 3.7744252681732178, "learning_rate": 1.6978819091832296e-05, "loss": 0.092, "step": 4265 }, { "epoch": 0.462489158716392, "grad_norm": 4.091119766235352, "learning_rate": 1.6978093718264908e-05, "loss": 0.062, "step": 4266 }, { "epoch": 0.46259757155247183, "grad_norm": 2.9342706203460693, "learning_rate": 1.697736834469752e-05, "loss": 0.0495, "step": 4267 }, { "epoch": 0.4627059843885516, "grad_norm": 3.2454607486724854, "learning_rate": 1.6976642971130133e-05, "loss": 0.0437, "step": 4268 }, { "epoch": 0.4628143972246314, "grad_norm": 4.608704566955566, "learning_rate": 1.6975917597562745e-05, "loss": 0.033, "step": 4269 }, { "epoch": 0.4629228100607112, "grad_norm": 3.4977824687957764, "learning_rate": 1.6975192223995357e-05, "loss": 0.0629, "step": 4270 }, { "epoch": 0.463031222896791, "grad_norm": 9.38178539276123, "learning_rate": 1.6974466850427972e-05, "loss": 0.0457, "step": 4271 }, { "epoch": 0.46313963573287076, "grad_norm": 2.573504686355591, "learning_rate": 1.6973741476860585e-05, "loss": 0.0263, "step": 4272 }, { "epoch": 0.4632480485689506, "grad_norm": 2.4965436458587646, "learning_rate": 1.69730161032932e-05, "loss": 0.0301, "step": 4273 }, { "epoch": 0.46335646140503034, "grad_norm": 1.709629774093628, "learning_rate": 1.6972290729725812e-05, "loss": 0.0222, "step": 4274 }, { "epoch": 0.46346487424111016, "grad_norm": 0.9045200347900391, "learning_rate": 1.6971565356158424e-05, "loss": 0.0129, "step": 4275 }, { "epoch": 0.4635732870771899, "grad_norm": 0.3620319664478302, "learning_rate": 1.6970839982591036e-05, "loss": 0.0045, "step": 4276 }, { "epoch": 0.46368169991326974, "grad_norm": 3.2956581115722656, "learning_rate": 1.697011460902365e-05, "loss": 0.0878, "step": 4277 }, { "epoch": 0.4637901127493495, "grad_norm": 3.240664482116699, "learning_rate": 1.696938923545626e-05, "loss": 0.0274, "step": 4278 }, { "epoch": 0.4638985255854293, "grad_norm": 2.9044151306152344, "learning_rate": 1.6968663861888873e-05, "loss": 0.0526, "step": 4279 }, { "epoch": 0.4640069384215091, "grad_norm": 1.8877217769622803, "learning_rate": 1.696793848832149e-05, "loss": 0.0259, "step": 4280 }, { "epoch": 0.4641153512575889, "grad_norm": 2.2842066287994385, "learning_rate": 1.69672131147541e-05, "loss": 0.0544, "step": 4281 }, { "epoch": 0.46422376409366867, "grad_norm": 1.5218896865844727, "learning_rate": 1.6966487741186713e-05, "loss": 0.046, "step": 4282 }, { "epoch": 0.4643321769297485, "grad_norm": 1.0031781196594238, "learning_rate": 1.6965762367619325e-05, "loss": 0.0191, "step": 4283 }, { "epoch": 0.46444058976582825, "grad_norm": 1.3386012315750122, "learning_rate": 1.6965036994051937e-05, "loss": 0.0397, "step": 4284 }, { "epoch": 0.46454900260190807, "grad_norm": 2.3503711223602295, "learning_rate": 1.696431162048455e-05, "loss": 0.0455, "step": 4285 }, { "epoch": 0.46465741543798783, "grad_norm": 1.431963324546814, "learning_rate": 1.6963586246917165e-05, "loss": 0.041, "step": 4286 }, { "epoch": 0.46476582827406765, "grad_norm": 1.2403810024261475, "learning_rate": 1.6962860873349777e-05, "loss": 0.0399, "step": 4287 }, { "epoch": 0.46487424111014747, "grad_norm": 2.6599557399749756, "learning_rate": 1.696213549978239e-05, "loss": 0.0489, "step": 4288 }, { "epoch": 0.46498265394622723, "grad_norm": 2.8814048767089844, "learning_rate": 1.6961410126215e-05, "loss": 0.0357, "step": 4289 }, { "epoch": 0.46509106678230705, "grad_norm": 2.3886830806732178, "learning_rate": 1.6960684752647617e-05, "loss": 0.0534, "step": 4290 }, { "epoch": 0.4651994796183868, "grad_norm": 0.8630746603012085, "learning_rate": 1.695995937908023e-05, "loss": 0.0133, "step": 4291 }, { "epoch": 0.46530789245446663, "grad_norm": 2.576540231704712, "learning_rate": 1.695923400551284e-05, "loss": 0.0176, "step": 4292 }, { "epoch": 0.4654163052905464, "grad_norm": 6.850473880767822, "learning_rate": 1.6958508631945453e-05, "loss": 0.0939, "step": 4293 }, { "epoch": 0.4655247181266262, "grad_norm": 1.681876540184021, "learning_rate": 1.6957783258378065e-05, "loss": 0.0499, "step": 4294 }, { "epoch": 0.465633130962706, "grad_norm": 2.3997087478637695, "learning_rate": 1.695705788481068e-05, "loss": 0.0481, "step": 4295 }, { "epoch": 0.4657415437987858, "grad_norm": 1.5295761823654175, "learning_rate": 1.6956332511243293e-05, "loss": 0.0307, "step": 4296 }, { "epoch": 0.46584995663486556, "grad_norm": 2.7953896522521973, "learning_rate": 1.6955607137675905e-05, "loss": 0.0339, "step": 4297 }, { "epoch": 0.4659583694709454, "grad_norm": 3.4213061332702637, "learning_rate": 1.6954881764108517e-05, "loss": 0.0417, "step": 4298 }, { "epoch": 0.46606678230702514, "grad_norm": 2.946134567260742, "learning_rate": 1.695415639054113e-05, "loss": 0.0676, "step": 4299 }, { "epoch": 0.46617519514310496, "grad_norm": 1.164265513420105, "learning_rate": 1.695343101697374e-05, "loss": 0.0098, "step": 4300 }, { "epoch": 0.4662836079791847, "grad_norm": 1.6839667558670044, "learning_rate": 1.6952705643406354e-05, "loss": 0.0238, "step": 4301 }, { "epoch": 0.46639202081526454, "grad_norm": 3.1868937015533447, "learning_rate": 1.695198026983897e-05, "loss": 0.075, "step": 4302 }, { "epoch": 0.4665004336513443, "grad_norm": 2.4733526706695557, "learning_rate": 1.695125489627158e-05, "loss": 0.0269, "step": 4303 }, { "epoch": 0.4666088464874241, "grad_norm": 2.303778886795044, "learning_rate": 1.6950529522704193e-05, "loss": 0.0703, "step": 4304 }, { "epoch": 0.4667172593235039, "grad_norm": 4.1667561531066895, "learning_rate": 1.6949804149136806e-05, "loss": 0.0681, "step": 4305 }, { "epoch": 0.4668256721595837, "grad_norm": 2.52824330329895, "learning_rate": 1.6949078775569418e-05, "loss": 0.0447, "step": 4306 }, { "epoch": 0.46693408499566347, "grad_norm": 1.9715672731399536, "learning_rate": 1.6948353402002033e-05, "loss": 0.0526, "step": 4307 }, { "epoch": 0.4670424978317433, "grad_norm": 4.358046054840088, "learning_rate": 1.6947628028434645e-05, "loss": 0.0727, "step": 4308 }, { "epoch": 0.46715091066782305, "grad_norm": 4.164997100830078, "learning_rate": 1.6946902654867258e-05, "loss": 0.0406, "step": 4309 }, { "epoch": 0.46725932350390287, "grad_norm": 5.060085773468018, "learning_rate": 1.6946177281299873e-05, "loss": 0.0183, "step": 4310 }, { "epoch": 0.46736773633998263, "grad_norm": 1.5787506103515625, "learning_rate": 1.6945451907732485e-05, "loss": 0.0164, "step": 4311 }, { "epoch": 0.46747614917606245, "grad_norm": 2.271341562271118, "learning_rate": 1.6944726534165097e-05, "loss": 0.0291, "step": 4312 }, { "epoch": 0.4675845620121422, "grad_norm": 5.063032627105713, "learning_rate": 1.694400116059771e-05, "loss": 0.0951, "step": 4313 }, { "epoch": 0.46769297484822203, "grad_norm": 1.2564510107040405, "learning_rate": 1.694327578703032e-05, "loss": 0.0196, "step": 4314 }, { "epoch": 0.4678013876843018, "grad_norm": 2.245598554611206, "learning_rate": 1.6942550413462934e-05, "loss": 0.0218, "step": 4315 }, { "epoch": 0.4679098005203816, "grad_norm": 1.9345778226852417, "learning_rate": 1.6941825039895546e-05, "loss": 0.0402, "step": 4316 }, { "epoch": 0.46801821335646143, "grad_norm": 1.6845178604125977, "learning_rate": 1.694109966632816e-05, "loss": 0.0367, "step": 4317 }, { "epoch": 0.4681266261925412, "grad_norm": 2.714205026626587, "learning_rate": 1.6940374292760774e-05, "loss": 0.0228, "step": 4318 }, { "epoch": 0.468235039028621, "grad_norm": 3.1953418254852295, "learning_rate": 1.6939648919193386e-05, "loss": 0.0414, "step": 4319 }, { "epoch": 0.4683434518647008, "grad_norm": 1.8729186058044434, "learning_rate": 1.6938923545625998e-05, "loss": 0.0262, "step": 4320 }, { "epoch": 0.4684518647007806, "grad_norm": 2.3229002952575684, "learning_rate": 1.693819817205861e-05, "loss": 0.0448, "step": 4321 }, { "epoch": 0.46856027753686036, "grad_norm": 4.367651462554932, "learning_rate": 1.6937472798491222e-05, "loss": 0.0907, "step": 4322 }, { "epoch": 0.4686686903729402, "grad_norm": 3.230020046234131, "learning_rate": 1.6936747424923838e-05, "loss": 0.0523, "step": 4323 }, { "epoch": 0.46877710320901994, "grad_norm": 1.7637478113174438, "learning_rate": 1.693602205135645e-05, "loss": 0.0172, "step": 4324 }, { "epoch": 0.46888551604509976, "grad_norm": 0.7245608568191528, "learning_rate": 1.6935296677789065e-05, "loss": 0.0232, "step": 4325 }, { "epoch": 0.4689939288811795, "grad_norm": 1.6318364143371582, "learning_rate": 1.6934571304221678e-05, "loss": 0.0246, "step": 4326 }, { "epoch": 0.46910234171725934, "grad_norm": 2.7349507808685303, "learning_rate": 1.693384593065429e-05, "loss": 0.0191, "step": 4327 }, { "epoch": 0.4692107545533391, "grad_norm": 1.6613740921020508, "learning_rate": 1.6933120557086902e-05, "loss": 0.0465, "step": 4328 }, { "epoch": 0.4693191673894189, "grad_norm": 4.625885963439941, "learning_rate": 1.6932395183519514e-05, "loss": 0.0281, "step": 4329 }, { "epoch": 0.4694275802254987, "grad_norm": 2.2580530643463135, "learning_rate": 1.6931669809952126e-05, "loss": 0.0204, "step": 4330 }, { "epoch": 0.4695359930615785, "grad_norm": 4.471071243286133, "learning_rate": 1.6930944436384738e-05, "loss": 0.0606, "step": 4331 }, { "epoch": 0.46964440589765827, "grad_norm": 6.509153366088867, "learning_rate": 1.6930219062817354e-05, "loss": 0.0331, "step": 4332 }, { "epoch": 0.4697528187337381, "grad_norm": 6.7750773429870605, "learning_rate": 1.6929493689249966e-05, "loss": 0.0598, "step": 4333 }, { "epoch": 0.46986123156981785, "grad_norm": 2.3896303176879883, "learning_rate": 1.6928768315682578e-05, "loss": 0.0336, "step": 4334 }, { "epoch": 0.46996964440589767, "grad_norm": 8.406119346618652, "learning_rate": 1.692804294211519e-05, "loss": 0.0574, "step": 4335 }, { "epoch": 0.47007805724197743, "grad_norm": 0.45544007420539856, "learning_rate": 1.6927317568547802e-05, "loss": 0.0299, "step": 4336 }, { "epoch": 0.47018647007805725, "grad_norm": 3.7140555381774902, "learning_rate": 1.6926592194980415e-05, "loss": 0.0891, "step": 4337 }, { "epoch": 0.470294882914137, "grad_norm": 1.8489809036254883, "learning_rate": 1.692586682141303e-05, "loss": 0.0394, "step": 4338 }, { "epoch": 0.47040329575021683, "grad_norm": 6.835998058319092, "learning_rate": 1.6925141447845642e-05, "loss": 0.0953, "step": 4339 }, { "epoch": 0.4705117085862966, "grad_norm": 3.028475046157837, "learning_rate": 1.6924416074278254e-05, "loss": 0.058, "step": 4340 }, { "epoch": 0.4706201214223764, "grad_norm": 1.934751033782959, "learning_rate": 1.6923690700710866e-05, "loss": 0.0321, "step": 4341 }, { "epoch": 0.4707285342584562, "grad_norm": 2.1566967964172363, "learning_rate": 1.6922965327143482e-05, "loss": 0.0306, "step": 4342 }, { "epoch": 0.470836947094536, "grad_norm": 2.5941100120544434, "learning_rate": 1.6922239953576094e-05, "loss": 0.0461, "step": 4343 }, { "epoch": 0.47094535993061576, "grad_norm": 4.265124797821045, "learning_rate": 1.6921514580008706e-05, "loss": 0.0537, "step": 4344 }, { "epoch": 0.4710537727666956, "grad_norm": 0.43962812423706055, "learning_rate": 1.692078920644132e-05, "loss": 0.0056, "step": 4345 }, { "epoch": 0.4711621856027754, "grad_norm": 1.151773452758789, "learning_rate": 1.692006383287393e-05, "loss": 0.0204, "step": 4346 }, { "epoch": 0.47127059843885516, "grad_norm": 3.9302685260772705, "learning_rate": 1.6919338459306546e-05, "loss": 0.0554, "step": 4347 }, { "epoch": 0.471379011274935, "grad_norm": 2.2193522453308105, "learning_rate": 1.6918613085739158e-05, "loss": 0.0888, "step": 4348 }, { "epoch": 0.47148742411101474, "grad_norm": 4.136456489562988, "learning_rate": 1.691788771217177e-05, "loss": 0.0674, "step": 4349 }, { "epoch": 0.47159583694709456, "grad_norm": 1.9756423234939575, "learning_rate": 1.6917162338604383e-05, "loss": 0.047, "step": 4350 }, { "epoch": 0.4717042497831743, "grad_norm": 2.0680623054504395, "learning_rate": 1.6916436965036995e-05, "loss": 0.05, "step": 4351 }, { "epoch": 0.47181266261925414, "grad_norm": 5.67338752746582, "learning_rate": 1.6915711591469607e-05, "loss": 0.0396, "step": 4352 }, { "epoch": 0.4719210754553339, "grad_norm": 2.044703245162964, "learning_rate": 1.691498621790222e-05, "loss": 0.0441, "step": 4353 }, { "epoch": 0.4720294882914137, "grad_norm": 0.6485652327537537, "learning_rate": 1.6914260844334834e-05, "loss": 0.0256, "step": 4354 }, { "epoch": 0.4721379011274935, "grad_norm": 5.613622665405273, "learning_rate": 1.6913535470767447e-05, "loss": 0.0545, "step": 4355 }, { "epoch": 0.4722463139635733, "grad_norm": 1.3519467115402222, "learning_rate": 1.691281009720006e-05, "loss": 0.0236, "step": 4356 }, { "epoch": 0.47235472679965307, "grad_norm": 0.6624916791915894, "learning_rate": 1.691208472363267e-05, "loss": 0.0156, "step": 4357 }, { "epoch": 0.4724631396357329, "grad_norm": 2.6950478553771973, "learning_rate": 1.6911359350065283e-05, "loss": 0.0392, "step": 4358 }, { "epoch": 0.47257155247181265, "grad_norm": 3.7211649417877197, "learning_rate": 1.69106339764979e-05, "loss": 0.0721, "step": 4359 }, { "epoch": 0.47267996530789247, "grad_norm": 5.490581035614014, "learning_rate": 1.690990860293051e-05, "loss": 0.0877, "step": 4360 }, { "epoch": 0.47278837814397223, "grad_norm": 1.6147242784500122, "learning_rate": 1.6909183229363123e-05, "loss": 0.0384, "step": 4361 }, { "epoch": 0.47289679098005205, "grad_norm": 3.03812837600708, "learning_rate": 1.690845785579574e-05, "loss": 0.0718, "step": 4362 }, { "epoch": 0.4730052038161318, "grad_norm": 2.0538766384124756, "learning_rate": 1.690773248222835e-05, "loss": 0.0426, "step": 4363 }, { "epoch": 0.47311361665221163, "grad_norm": 1.356488585472107, "learning_rate": 1.6907007108660963e-05, "loss": 0.0231, "step": 4364 }, { "epoch": 0.4732220294882914, "grad_norm": 2.415693521499634, "learning_rate": 1.6906281735093575e-05, "loss": 0.0449, "step": 4365 }, { "epoch": 0.4733304423243712, "grad_norm": 2.0639684200286865, "learning_rate": 1.6905556361526187e-05, "loss": 0.0255, "step": 4366 }, { "epoch": 0.473438855160451, "grad_norm": 1.5642324686050415, "learning_rate": 1.69048309879588e-05, "loss": 0.0245, "step": 4367 }, { "epoch": 0.4735472679965308, "grad_norm": 3.693575859069824, "learning_rate": 1.690410561439141e-05, "loss": 0.0496, "step": 4368 }, { "epoch": 0.47365568083261056, "grad_norm": 3.6741418838500977, "learning_rate": 1.6903380240824027e-05, "loss": 0.0433, "step": 4369 }, { "epoch": 0.4737640936686904, "grad_norm": 1.4923382997512817, "learning_rate": 1.690265486725664e-05, "loss": 0.0289, "step": 4370 }, { "epoch": 0.47387250650477014, "grad_norm": 0.5577576756477356, "learning_rate": 1.690192949368925e-05, "loss": 0.0136, "step": 4371 }, { "epoch": 0.47398091934084996, "grad_norm": 1.9913235902786255, "learning_rate": 1.6901204120121863e-05, "loss": 0.0365, "step": 4372 }, { "epoch": 0.4740893321769297, "grad_norm": 2.274796724319458, "learning_rate": 1.6900478746554475e-05, "loss": 0.0591, "step": 4373 }, { "epoch": 0.47419774501300954, "grad_norm": 2.371828556060791, "learning_rate": 1.6899753372987088e-05, "loss": 0.0541, "step": 4374 }, { "epoch": 0.4743061578490893, "grad_norm": 5.323863506317139, "learning_rate": 1.6899027999419703e-05, "loss": 0.0575, "step": 4375 }, { "epoch": 0.4744145706851691, "grad_norm": 2.5871007442474365, "learning_rate": 1.6898302625852315e-05, "loss": 0.0361, "step": 4376 }, { "epoch": 0.47452298352124894, "grad_norm": 1.1016637086868286, "learning_rate": 1.689757725228493e-05, "loss": 0.0158, "step": 4377 }, { "epoch": 0.4746313963573287, "grad_norm": 1.330845832824707, "learning_rate": 1.6896851878717543e-05, "loss": 0.0167, "step": 4378 }, { "epoch": 0.4747398091934085, "grad_norm": 1.8727772235870361, "learning_rate": 1.6896126505150155e-05, "loss": 0.0547, "step": 4379 }, { "epoch": 0.4748482220294883, "grad_norm": 3.8832409381866455, "learning_rate": 1.6895401131582767e-05, "loss": 0.0272, "step": 4380 }, { "epoch": 0.4749566348655681, "grad_norm": 2.75656795501709, "learning_rate": 1.689467575801538e-05, "loss": 0.0309, "step": 4381 }, { "epoch": 0.47506504770164787, "grad_norm": 3.108506202697754, "learning_rate": 1.689395038444799e-05, "loss": 0.0423, "step": 4382 }, { "epoch": 0.4751734605377277, "grad_norm": 2.6804304122924805, "learning_rate": 1.6893225010880604e-05, "loss": 0.0682, "step": 4383 }, { "epoch": 0.47528187337380745, "grad_norm": 4.799514293670654, "learning_rate": 1.689249963731322e-05, "loss": 0.0779, "step": 4384 }, { "epoch": 0.47539028620988727, "grad_norm": 4.075442790985107, "learning_rate": 1.689177426374583e-05, "loss": 0.066, "step": 4385 }, { "epoch": 0.47549869904596703, "grad_norm": 2.8338916301727295, "learning_rate": 1.6891048890178443e-05, "loss": 0.0623, "step": 4386 }, { "epoch": 0.47560711188204685, "grad_norm": 2.5581610202789307, "learning_rate": 1.6890323516611056e-05, "loss": 0.0195, "step": 4387 }, { "epoch": 0.4757155247181266, "grad_norm": 2.0111050605773926, "learning_rate": 1.6889598143043668e-05, "loss": 0.0282, "step": 4388 }, { "epoch": 0.47582393755420643, "grad_norm": 4.504203796386719, "learning_rate": 1.688887276947628e-05, "loss": 0.052, "step": 4389 }, { "epoch": 0.4759323503902862, "grad_norm": 6.670331001281738, "learning_rate": 1.6888147395908895e-05, "loss": 0.0649, "step": 4390 }, { "epoch": 0.476040763226366, "grad_norm": 3.432438611984253, "learning_rate": 1.6887422022341508e-05, "loss": 0.0508, "step": 4391 }, { "epoch": 0.4761491760624458, "grad_norm": 2.944662570953369, "learning_rate": 1.688669664877412e-05, "loss": 0.0585, "step": 4392 }, { "epoch": 0.4762575888985256, "grad_norm": 4.347379207611084, "learning_rate": 1.6885971275206732e-05, "loss": 0.0798, "step": 4393 }, { "epoch": 0.47636600173460536, "grad_norm": 1.0138828754425049, "learning_rate": 1.6885245901639347e-05, "loss": 0.0196, "step": 4394 }, { "epoch": 0.4764744145706852, "grad_norm": 1.4417462348937988, "learning_rate": 1.688452052807196e-05, "loss": 0.0112, "step": 4395 }, { "epoch": 0.47658282740676494, "grad_norm": 2.7597475051879883, "learning_rate": 1.688379515450457e-05, "loss": 0.0783, "step": 4396 }, { "epoch": 0.47669124024284476, "grad_norm": 0.8998900055885315, "learning_rate": 1.6883069780937184e-05, "loss": 0.0087, "step": 4397 }, { "epoch": 0.4767996530789245, "grad_norm": 0.5019741058349609, "learning_rate": 1.6882344407369796e-05, "loss": 0.0035, "step": 4398 }, { "epoch": 0.47690806591500434, "grad_norm": 3.4671738147735596, "learning_rate": 1.688161903380241e-05, "loss": 0.0653, "step": 4399 }, { "epoch": 0.4770164787510841, "grad_norm": 1.9735764265060425, "learning_rate": 1.6880893660235024e-05, "loss": 0.0433, "step": 4400 }, { "epoch": 0.4771248915871639, "grad_norm": 6.499273777008057, "learning_rate": 1.6880168286667636e-05, "loss": 0.0392, "step": 4401 }, { "epoch": 0.4772333044232437, "grad_norm": 3.109198808670044, "learning_rate": 1.6879442913100248e-05, "loss": 0.0511, "step": 4402 }, { "epoch": 0.4773417172593235, "grad_norm": 1.6085745096206665, "learning_rate": 1.687871753953286e-05, "loss": 0.045, "step": 4403 }, { "epoch": 0.47745013009540327, "grad_norm": 1.845399022102356, "learning_rate": 1.6877992165965472e-05, "loss": 0.0596, "step": 4404 }, { "epoch": 0.4775585429314831, "grad_norm": 1.4278360605239868, "learning_rate": 1.6877266792398084e-05, "loss": 0.023, "step": 4405 }, { "epoch": 0.4776669557675629, "grad_norm": 2.824406385421753, "learning_rate": 1.68765414188307e-05, "loss": 0.077, "step": 4406 }, { "epoch": 0.47777536860364267, "grad_norm": 0.7555529475212097, "learning_rate": 1.6875816045263312e-05, "loss": 0.018, "step": 4407 }, { "epoch": 0.4778837814397225, "grad_norm": 2.4789276123046875, "learning_rate": 1.6875090671695924e-05, "loss": 0.0303, "step": 4408 }, { "epoch": 0.47799219427580225, "grad_norm": 5.503045558929443, "learning_rate": 1.6874365298128536e-05, "loss": 0.0482, "step": 4409 }, { "epoch": 0.47810060711188207, "grad_norm": 0.6161433458328247, "learning_rate": 1.687363992456115e-05, "loss": 0.0116, "step": 4410 }, { "epoch": 0.47820901994796183, "grad_norm": 3.42220401763916, "learning_rate": 1.6872914550993764e-05, "loss": 0.0836, "step": 4411 }, { "epoch": 0.47831743278404165, "grad_norm": 1.5550992488861084, "learning_rate": 1.6872189177426376e-05, "loss": 0.0313, "step": 4412 }, { "epoch": 0.4784258456201214, "grad_norm": 3.066044569015503, "learning_rate": 1.6871463803858988e-05, "loss": 0.0725, "step": 4413 }, { "epoch": 0.47853425845620123, "grad_norm": 3.486659049987793, "learning_rate": 1.6870738430291604e-05, "loss": 0.0859, "step": 4414 }, { "epoch": 0.478642671292281, "grad_norm": 2.704643726348877, "learning_rate": 1.6870013056724216e-05, "loss": 0.0528, "step": 4415 }, { "epoch": 0.4787510841283608, "grad_norm": 9.409418106079102, "learning_rate": 1.6869287683156828e-05, "loss": 0.0216, "step": 4416 }, { "epoch": 0.4788594969644406, "grad_norm": 1.2913475036621094, "learning_rate": 1.686856230958944e-05, "loss": 0.0297, "step": 4417 }, { "epoch": 0.4789679098005204, "grad_norm": 2.318903923034668, "learning_rate": 1.6867836936022052e-05, "loss": 0.0564, "step": 4418 }, { "epoch": 0.47907632263660016, "grad_norm": 5.266411781311035, "learning_rate": 1.6867111562454664e-05, "loss": 0.0624, "step": 4419 }, { "epoch": 0.47918473547268, "grad_norm": 3.32511830329895, "learning_rate": 1.6866386188887277e-05, "loss": 0.0282, "step": 4420 }, { "epoch": 0.47929314830875974, "grad_norm": 2.031712770462036, "learning_rate": 1.6865660815319892e-05, "loss": 0.0211, "step": 4421 }, { "epoch": 0.47940156114483956, "grad_norm": 1.2341159582138062, "learning_rate": 1.6864935441752504e-05, "loss": 0.0209, "step": 4422 }, { "epoch": 0.4795099739809193, "grad_norm": 2.3311195373535156, "learning_rate": 1.6864210068185116e-05, "loss": 0.042, "step": 4423 }, { "epoch": 0.47961838681699914, "grad_norm": 0.9702398777008057, "learning_rate": 1.686348469461773e-05, "loss": 0.0293, "step": 4424 }, { "epoch": 0.4797267996530789, "grad_norm": 4.11933708190918, "learning_rate": 1.686275932105034e-05, "loss": 0.0678, "step": 4425 }, { "epoch": 0.4798352124891587, "grad_norm": 2.576094388961792, "learning_rate": 1.6862033947482953e-05, "loss": 0.0611, "step": 4426 }, { "epoch": 0.4799436253252385, "grad_norm": 0.8325637578964233, "learning_rate": 1.686130857391557e-05, "loss": 0.0164, "step": 4427 }, { "epoch": 0.4800520381613183, "grad_norm": 5.021759033203125, "learning_rate": 1.686058320034818e-05, "loss": 0.0409, "step": 4428 }, { "epoch": 0.48016045099739807, "grad_norm": 4.54482889175415, "learning_rate": 1.6859857826780796e-05, "loss": 0.0965, "step": 4429 }, { "epoch": 0.4802688638334779, "grad_norm": 1.9884979724884033, "learning_rate": 1.6859132453213408e-05, "loss": 0.0191, "step": 4430 }, { "epoch": 0.48037727666955765, "grad_norm": 2.5680506229400635, "learning_rate": 1.685840707964602e-05, "loss": 0.0298, "step": 4431 }, { "epoch": 0.48048568950563747, "grad_norm": 1.6655057668685913, "learning_rate": 1.6857681706078633e-05, "loss": 0.0314, "step": 4432 }, { "epoch": 0.48059410234171723, "grad_norm": 3.390470504760742, "learning_rate": 1.6856956332511245e-05, "loss": 0.053, "step": 4433 }, { "epoch": 0.48070251517779705, "grad_norm": 2.021397590637207, "learning_rate": 1.6856230958943857e-05, "loss": 0.0461, "step": 4434 }, { "epoch": 0.48081092801387687, "grad_norm": 3.236499547958374, "learning_rate": 1.685550558537647e-05, "loss": 0.0614, "step": 4435 }, { "epoch": 0.48091934084995663, "grad_norm": 5.7600836753845215, "learning_rate": 1.6854780211809084e-05, "loss": 0.1024, "step": 4436 }, { "epoch": 0.48102775368603645, "grad_norm": 2.7753288745880127, "learning_rate": 1.6854054838241697e-05, "loss": 0.0195, "step": 4437 }, { "epoch": 0.4811361665221162, "grad_norm": 2.6198484897613525, "learning_rate": 1.685332946467431e-05, "loss": 0.047, "step": 4438 }, { "epoch": 0.48124457935819603, "grad_norm": 1.7819023132324219, "learning_rate": 1.685260409110692e-05, "loss": 0.0401, "step": 4439 }, { "epoch": 0.4813529921942758, "grad_norm": 2.507128953933716, "learning_rate": 1.6851878717539533e-05, "loss": 0.034, "step": 4440 }, { "epoch": 0.4814614050303556, "grad_norm": 1.1046746969223022, "learning_rate": 1.6851153343972145e-05, "loss": 0.0201, "step": 4441 }, { "epoch": 0.4815698178664354, "grad_norm": 3.2222938537597656, "learning_rate": 1.685042797040476e-05, "loss": 0.0467, "step": 4442 }, { "epoch": 0.4816782307025152, "grad_norm": 2.738499164581299, "learning_rate": 1.6849702596837373e-05, "loss": 0.0325, "step": 4443 }, { "epoch": 0.48178664353859496, "grad_norm": 2.0114221572875977, "learning_rate": 1.6848977223269985e-05, "loss": 0.0294, "step": 4444 }, { "epoch": 0.4818950563746748, "grad_norm": 1.1461902856826782, "learning_rate": 1.68482518497026e-05, "loss": 0.0153, "step": 4445 }, { "epoch": 0.48200346921075454, "grad_norm": 0.5770850777626038, "learning_rate": 1.6847526476135213e-05, "loss": 0.0058, "step": 4446 }, { "epoch": 0.48211188204683436, "grad_norm": 1.3564012050628662, "learning_rate": 1.6846801102567825e-05, "loss": 0.0295, "step": 4447 }, { "epoch": 0.4822202948829141, "grad_norm": 3.52533221244812, "learning_rate": 1.6846075729000437e-05, "loss": 0.078, "step": 4448 }, { "epoch": 0.48232870771899394, "grad_norm": 1.128103256225586, "learning_rate": 1.684535035543305e-05, "loss": 0.0264, "step": 4449 }, { "epoch": 0.4824371205550737, "grad_norm": 3.0966858863830566, "learning_rate": 1.684462498186566e-05, "loss": 0.0257, "step": 4450 }, { "epoch": 0.4825455333911535, "grad_norm": 2.8170671463012695, "learning_rate": 1.6843899608298277e-05, "loss": 0.0614, "step": 4451 }, { "epoch": 0.4826539462272333, "grad_norm": 1.3004765510559082, "learning_rate": 1.684317423473089e-05, "loss": 0.0172, "step": 4452 }, { "epoch": 0.4827623590633131, "grad_norm": 2.755455255508423, "learning_rate": 1.68424488611635e-05, "loss": 0.0408, "step": 4453 }, { "epoch": 0.48287077189939287, "grad_norm": 2.6307241916656494, "learning_rate": 1.6841723487596113e-05, "loss": 0.0399, "step": 4454 }, { "epoch": 0.4829791847354727, "grad_norm": 1.3317819833755493, "learning_rate": 1.6840998114028725e-05, "loss": 0.0106, "step": 4455 }, { "epoch": 0.48308759757155245, "grad_norm": 4.928503036499023, "learning_rate": 1.6840272740461337e-05, "loss": 0.083, "step": 4456 }, { "epoch": 0.48319601040763227, "grad_norm": 1.5501872301101685, "learning_rate": 1.683954736689395e-05, "loss": 0.0189, "step": 4457 }, { "epoch": 0.48330442324371203, "grad_norm": 2.8054864406585693, "learning_rate": 1.6838821993326565e-05, "loss": 0.0391, "step": 4458 }, { "epoch": 0.48341283607979185, "grad_norm": 0.973491370677948, "learning_rate": 1.6838096619759177e-05, "loss": 0.0121, "step": 4459 }, { "epoch": 0.4835212489158716, "grad_norm": 5.296030521392822, "learning_rate": 1.683737124619179e-05, "loss": 0.0664, "step": 4460 }, { "epoch": 0.48362966175195143, "grad_norm": 1.1079118251800537, "learning_rate": 1.68366458726244e-05, "loss": 0.0133, "step": 4461 }, { "epoch": 0.4837380745880312, "grad_norm": 6.69707727432251, "learning_rate": 1.6835920499057014e-05, "loss": 0.0561, "step": 4462 }, { "epoch": 0.483846487424111, "grad_norm": 2.7208399772644043, "learning_rate": 1.683519512548963e-05, "loss": 0.061, "step": 4463 }, { "epoch": 0.4839549002601908, "grad_norm": 4.98520040512085, "learning_rate": 1.683446975192224e-05, "loss": 0.0601, "step": 4464 }, { "epoch": 0.4840633130962706, "grad_norm": 1.279625415802002, "learning_rate": 1.6833744378354854e-05, "loss": 0.0104, "step": 4465 }, { "epoch": 0.4841717259323504, "grad_norm": 1.3545277118682861, "learning_rate": 1.683301900478747e-05, "loss": 0.0372, "step": 4466 }, { "epoch": 0.4842801387684302, "grad_norm": 1.4566622972488403, "learning_rate": 1.683229363122008e-05, "loss": 0.0448, "step": 4467 }, { "epoch": 0.48438855160451, "grad_norm": 4.158965110778809, "learning_rate": 1.6831568257652693e-05, "loss": 0.0706, "step": 4468 }, { "epoch": 0.48449696444058976, "grad_norm": 1.5260132551193237, "learning_rate": 1.6830842884085306e-05, "loss": 0.0326, "step": 4469 }, { "epoch": 0.4846053772766696, "grad_norm": 1.532529592514038, "learning_rate": 1.6830117510517918e-05, "loss": 0.0261, "step": 4470 }, { "epoch": 0.48471379011274934, "grad_norm": 2.003516674041748, "learning_rate": 1.682939213695053e-05, "loss": 0.0215, "step": 4471 }, { "epoch": 0.48482220294882916, "grad_norm": 2.1347949504852295, "learning_rate": 1.6828666763383142e-05, "loss": 0.0603, "step": 4472 }, { "epoch": 0.4849306157849089, "grad_norm": 2.6737709045410156, "learning_rate": 1.6827941389815757e-05, "loss": 0.0389, "step": 4473 }, { "epoch": 0.48503902862098874, "grad_norm": 3.931994676589966, "learning_rate": 1.682721601624837e-05, "loss": 0.0253, "step": 4474 }, { "epoch": 0.4851474414570685, "grad_norm": 3.2984933853149414, "learning_rate": 1.6826490642680982e-05, "loss": 0.0498, "step": 4475 }, { "epoch": 0.4852558542931483, "grad_norm": 1.0607491731643677, "learning_rate": 1.6825765269113594e-05, "loss": 0.0101, "step": 4476 }, { "epoch": 0.4853642671292281, "grad_norm": 3.057821273803711, "learning_rate": 1.6825039895546206e-05, "loss": 0.0492, "step": 4477 }, { "epoch": 0.4854726799653079, "grad_norm": 2.7135798931121826, "learning_rate": 1.6824314521978818e-05, "loss": 0.0361, "step": 4478 }, { "epoch": 0.48558109280138767, "grad_norm": 0.6330222487449646, "learning_rate": 1.6823589148411434e-05, "loss": 0.0256, "step": 4479 }, { "epoch": 0.4856895056374675, "grad_norm": 2.0646190643310547, "learning_rate": 1.6822863774844046e-05, "loss": 0.0541, "step": 4480 }, { "epoch": 0.48579791847354725, "grad_norm": 0.7270298600196838, "learning_rate": 1.682213840127666e-05, "loss": 0.0186, "step": 4481 }, { "epoch": 0.48590633130962707, "grad_norm": 0.8666489720344543, "learning_rate": 1.6821413027709274e-05, "loss": 0.0176, "step": 4482 }, { "epoch": 0.48601474414570683, "grad_norm": 3.9819154739379883, "learning_rate": 1.6820687654141886e-05, "loss": 0.044, "step": 4483 }, { "epoch": 0.48612315698178665, "grad_norm": 3.2703373432159424, "learning_rate": 1.6819962280574498e-05, "loss": 0.028, "step": 4484 }, { "epoch": 0.4862315698178664, "grad_norm": 1.9847304821014404, "learning_rate": 1.681923690700711e-05, "loss": 0.0509, "step": 4485 }, { "epoch": 0.48633998265394623, "grad_norm": 8.857490539550781, "learning_rate": 1.6818511533439722e-05, "loss": 0.0131, "step": 4486 }, { "epoch": 0.486448395490026, "grad_norm": 1.6711311340332031, "learning_rate": 1.6817786159872334e-05, "loss": 0.0196, "step": 4487 }, { "epoch": 0.4865568083261058, "grad_norm": 3.843278646469116, "learning_rate": 1.681706078630495e-05, "loss": 0.0226, "step": 4488 }, { "epoch": 0.4866652211621856, "grad_norm": 3.3572499752044678, "learning_rate": 1.6816335412737562e-05, "loss": 0.0657, "step": 4489 }, { "epoch": 0.4867736339982654, "grad_norm": 4.9474968910217285, "learning_rate": 1.6815610039170174e-05, "loss": 0.0608, "step": 4490 }, { "epoch": 0.48688204683434516, "grad_norm": 3.428194761276245, "learning_rate": 1.6814884665602786e-05, "loss": 0.0342, "step": 4491 }, { "epoch": 0.486990459670425, "grad_norm": 1.2582707405090332, "learning_rate": 1.68141592920354e-05, "loss": 0.0164, "step": 4492 }, { "epoch": 0.48709887250650474, "grad_norm": 2.6507296562194824, "learning_rate": 1.681343391846801e-05, "loss": 0.0388, "step": 4493 }, { "epoch": 0.48720728534258456, "grad_norm": 1.8502781391143799, "learning_rate": 1.6812708544900626e-05, "loss": 0.0436, "step": 4494 }, { "epoch": 0.4873156981786644, "grad_norm": 2.315999746322632, "learning_rate": 1.6811983171333238e-05, "loss": 0.0323, "step": 4495 }, { "epoch": 0.48742411101474414, "grad_norm": 3.043436050415039, "learning_rate": 1.681125779776585e-05, "loss": 0.0522, "step": 4496 }, { "epoch": 0.48753252385082396, "grad_norm": 3.4303793907165527, "learning_rate": 1.6810532424198466e-05, "loss": 0.0828, "step": 4497 }, { "epoch": 0.4876409366869037, "grad_norm": 2.614625930786133, "learning_rate": 1.6809807050631078e-05, "loss": 0.0453, "step": 4498 }, { "epoch": 0.48774934952298354, "grad_norm": 0.9847751259803772, "learning_rate": 1.680908167706369e-05, "loss": 0.0151, "step": 4499 }, { "epoch": 0.4878577623590633, "grad_norm": 3.666177988052368, "learning_rate": 1.6808356303496302e-05, "loss": 0.0401, "step": 4500 }, { "epoch": 0.4879661751951431, "grad_norm": 2.506197214126587, "learning_rate": 1.6807630929928914e-05, "loss": 0.0383, "step": 4501 }, { "epoch": 0.4880745880312229, "grad_norm": 2.7422854900360107, "learning_rate": 1.6806905556361527e-05, "loss": 0.0382, "step": 4502 }, { "epoch": 0.4881830008673027, "grad_norm": 3.5140421390533447, "learning_rate": 1.6806180182794142e-05, "loss": 0.0601, "step": 4503 }, { "epoch": 0.48829141370338247, "grad_norm": 1.7600456476211548, "learning_rate": 1.6805454809226754e-05, "loss": 0.0253, "step": 4504 }, { "epoch": 0.4883998265394623, "grad_norm": 1.9849933385849, "learning_rate": 1.6804729435659366e-05, "loss": 0.0456, "step": 4505 }, { "epoch": 0.48850823937554205, "grad_norm": 4.741661548614502, "learning_rate": 1.680400406209198e-05, "loss": 0.0683, "step": 4506 }, { "epoch": 0.48861665221162187, "grad_norm": 1.7910712957382202, "learning_rate": 1.680327868852459e-05, "loss": 0.0362, "step": 4507 }, { "epoch": 0.48872506504770163, "grad_norm": 0.9404248595237732, "learning_rate": 1.6802553314957203e-05, "loss": 0.0153, "step": 4508 }, { "epoch": 0.48883347788378145, "grad_norm": 1.3954399824142456, "learning_rate": 1.6801827941389815e-05, "loss": 0.0302, "step": 4509 }, { "epoch": 0.4889418907198612, "grad_norm": 0.9363752007484436, "learning_rate": 1.680110256782243e-05, "loss": 0.0234, "step": 4510 }, { "epoch": 0.48905030355594103, "grad_norm": 1.1090446710586548, "learning_rate": 1.6800377194255043e-05, "loss": 0.0128, "step": 4511 }, { "epoch": 0.4891587163920208, "grad_norm": 2.6990649700164795, "learning_rate": 1.6799651820687655e-05, "loss": 0.0473, "step": 4512 }, { "epoch": 0.4892671292281006, "grad_norm": 2.217308521270752, "learning_rate": 1.6798926447120267e-05, "loss": 0.0468, "step": 4513 }, { "epoch": 0.4893755420641804, "grad_norm": 1.6408777236938477, "learning_rate": 1.6798201073552882e-05, "loss": 0.0068, "step": 4514 }, { "epoch": 0.4894839549002602, "grad_norm": 3.6092236042022705, "learning_rate": 1.6797475699985495e-05, "loss": 0.0513, "step": 4515 }, { "epoch": 0.48959236773633996, "grad_norm": 2.5511608123779297, "learning_rate": 1.6796750326418107e-05, "loss": 0.0364, "step": 4516 }, { "epoch": 0.4897007805724198, "grad_norm": 0.6982417106628418, "learning_rate": 1.679602495285072e-05, "loss": 0.0165, "step": 4517 }, { "epoch": 0.48980919340849954, "grad_norm": 3.0524656772613525, "learning_rate": 1.6795299579283334e-05, "loss": 0.0187, "step": 4518 }, { "epoch": 0.48991760624457936, "grad_norm": 2.9368107318878174, "learning_rate": 1.6794574205715947e-05, "loss": 0.0431, "step": 4519 }, { "epoch": 0.4900260190806591, "grad_norm": 2.1945130825042725, "learning_rate": 1.679384883214856e-05, "loss": 0.0198, "step": 4520 }, { "epoch": 0.49013443191673894, "grad_norm": 1.88521409034729, "learning_rate": 1.679312345858117e-05, "loss": 0.0226, "step": 4521 }, { "epoch": 0.4902428447528187, "grad_norm": 6.963931560516357, "learning_rate": 1.6792398085013783e-05, "loss": 0.1312, "step": 4522 }, { "epoch": 0.4903512575888985, "grad_norm": 4.092979907989502, "learning_rate": 1.6791672711446395e-05, "loss": 0.0475, "step": 4523 }, { "epoch": 0.49045967042497834, "grad_norm": 2.9880120754241943, "learning_rate": 1.6790947337879007e-05, "loss": 0.0477, "step": 4524 }, { "epoch": 0.4905680832610581, "grad_norm": 2.2236108779907227, "learning_rate": 1.6790221964311623e-05, "loss": 0.0711, "step": 4525 }, { "epoch": 0.4906764960971379, "grad_norm": 6.332974910736084, "learning_rate": 1.6789496590744235e-05, "loss": 0.0453, "step": 4526 }, { "epoch": 0.4907849089332177, "grad_norm": 1.4137216806411743, "learning_rate": 1.6788771217176847e-05, "loss": 0.0231, "step": 4527 }, { "epoch": 0.4908933217692975, "grad_norm": 4.004007339477539, "learning_rate": 1.678804584360946e-05, "loss": 0.0487, "step": 4528 }, { "epoch": 0.49100173460537727, "grad_norm": 4.762259483337402, "learning_rate": 1.678732047004207e-05, "loss": 0.0729, "step": 4529 }, { "epoch": 0.4911101474414571, "grad_norm": 1.1942411661148071, "learning_rate": 1.6786595096474684e-05, "loss": 0.0077, "step": 4530 }, { "epoch": 0.49121856027753685, "grad_norm": 2.6787333488464355, "learning_rate": 1.67858697229073e-05, "loss": 0.0586, "step": 4531 }, { "epoch": 0.49132697311361667, "grad_norm": 1.4309099912643433, "learning_rate": 1.678514434933991e-05, "loss": 0.0371, "step": 4532 }, { "epoch": 0.49143538594969643, "grad_norm": 1.3209561109542847, "learning_rate": 1.6784418975772527e-05, "loss": 0.0701, "step": 4533 }, { "epoch": 0.49154379878577625, "grad_norm": 3.561617851257324, "learning_rate": 1.678369360220514e-05, "loss": 0.0334, "step": 4534 }, { "epoch": 0.491652211621856, "grad_norm": 1.011401653289795, "learning_rate": 1.678296822863775e-05, "loss": 0.009, "step": 4535 }, { "epoch": 0.49176062445793584, "grad_norm": 2.3631527423858643, "learning_rate": 1.6782242855070363e-05, "loss": 0.0321, "step": 4536 }, { "epoch": 0.4918690372940156, "grad_norm": 1.495410680770874, "learning_rate": 1.6781517481502975e-05, "loss": 0.0339, "step": 4537 }, { "epoch": 0.4919774501300954, "grad_norm": 4.952219486236572, "learning_rate": 1.6780792107935587e-05, "loss": 0.0698, "step": 4538 }, { "epoch": 0.4920858629661752, "grad_norm": 2.583793878555298, "learning_rate": 1.67800667343682e-05, "loss": 0.0226, "step": 4539 }, { "epoch": 0.492194275802255, "grad_norm": 1.5249208211898804, "learning_rate": 1.6779341360800815e-05, "loss": 0.0302, "step": 4540 }, { "epoch": 0.49230268863833476, "grad_norm": 4.506143569946289, "learning_rate": 1.6778615987233427e-05, "loss": 0.0326, "step": 4541 }, { "epoch": 0.4924111014744146, "grad_norm": 3.1555678844451904, "learning_rate": 1.677789061366604e-05, "loss": 0.0338, "step": 4542 }, { "epoch": 0.49251951431049434, "grad_norm": 3.529006242752075, "learning_rate": 1.677716524009865e-05, "loss": 0.0742, "step": 4543 }, { "epoch": 0.49262792714657416, "grad_norm": 4.1559062004089355, "learning_rate": 1.6776439866531264e-05, "loss": 0.0459, "step": 4544 }, { "epoch": 0.4927363399826539, "grad_norm": 3.113442897796631, "learning_rate": 1.6775714492963876e-05, "loss": 0.0473, "step": 4545 }, { "epoch": 0.49284475281873374, "grad_norm": 1.2324492931365967, "learning_rate": 1.677498911939649e-05, "loss": 0.0234, "step": 4546 }, { "epoch": 0.4929531656548135, "grad_norm": 2.626477003097534, "learning_rate": 1.6774263745829104e-05, "loss": 0.0364, "step": 4547 }, { "epoch": 0.4930615784908933, "grad_norm": 5.260581970214844, "learning_rate": 1.6773538372261716e-05, "loss": 0.0916, "step": 4548 }, { "epoch": 0.4931699913269731, "grad_norm": 3.3875768184661865, "learning_rate": 1.677281299869433e-05, "loss": 0.0457, "step": 4549 }, { "epoch": 0.4932784041630529, "grad_norm": 1.2026060819625854, "learning_rate": 1.6772087625126943e-05, "loss": 0.0341, "step": 4550 }, { "epoch": 0.49338681699913267, "grad_norm": 1.8935739994049072, "learning_rate": 1.6771362251559555e-05, "loss": 0.027, "step": 4551 }, { "epoch": 0.4934952298352125, "grad_norm": 3.4250693321228027, "learning_rate": 1.6770636877992168e-05, "loss": 0.0591, "step": 4552 }, { "epoch": 0.4936036426712923, "grad_norm": 1.7008544206619263, "learning_rate": 1.676991150442478e-05, "loss": 0.0615, "step": 4553 }, { "epoch": 0.49371205550737207, "grad_norm": 3.300710678100586, "learning_rate": 1.6769186130857392e-05, "loss": 0.0447, "step": 4554 }, { "epoch": 0.4938204683434519, "grad_norm": 2.1892154216766357, "learning_rate": 1.6768460757290007e-05, "loss": 0.0392, "step": 4555 }, { "epoch": 0.49392888117953165, "grad_norm": 3.4332222938537598, "learning_rate": 1.676773538372262e-05, "loss": 0.0406, "step": 4556 }, { "epoch": 0.4940372940156115, "grad_norm": 1.635794758796692, "learning_rate": 1.6767010010155232e-05, "loss": 0.0296, "step": 4557 }, { "epoch": 0.49414570685169124, "grad_norm": 2.592325210571289, "learning_rate": 1.6766284636587844e-05, "loss": 0.0496, "step": 4558 }, { "epoch": 0.49425411968777105, "grad_norm": 1.134271502494812, "learning_rate": 1.6765559263020456e-05, "loss": 0.0275, "step": 4559 }, { "epoch": 0.4943625325238508, "grad_norm": 1.2329665422439575, "learning_rate": 1.6764833889453068e-05, "loss": 0.0159, "step": 4560 }, { "epoch": 0.49447094535993064, "grad_norm": 2.2918195724487305, "learning_rate": 1.676410851588568e-05, "loss": 0.0197, "step": 4561 }, { "epoch": 0.4945793581960104, "grad_norm": 2.98480486869812, "learning_rate": 1.6763383142318296e-05, "loss": 0.0395, "step": 4562 }, { "epoch": 0.4946877710320902, "grad_norm": 2.3175392150878906, "learning_rate": 1.6762657768750908e-05, "loss": 0.0275, "step": 4563 }, { "epoch": 0.49479618386817, "grad_norm": 2.005474090576172, "learning_rate": 1.676193239518352e-05, "loss": 0.0266, "step": 4564 }, { "epoch": 0.4949045967042498, "grad_norm": 4.247404098510742, "learning_rate": 1.6761207021616132e-05, "loss": 0.0655, "step": 4565 }, { "epoch": 0.49501300954032956, "grad_norm": 4.344419956207275, "learning_rate": 1.6760481648048748e-05, "loss": 0.0353, "step": 4566 }, { "epoch": 0.4951214223764094, "grad_norm": 1.7210333347320557, "learning_rate": 1.675975627448136e-05, "loss": 0.0357, "step": 4567 }, { "epoch": 0.49522983521248914, "grad_norm": 2.103654146194458, "learning_rate": 1.6759030900913972e-05, "loss": 0.0277, "step": 4568 }, { "epoch": 0.49533824804856896, "grad_norm": 3.013103723526001, "learning_rate": 1.6758305527346584e-05, "loss": 0.0307, "step": 4569 }, { "epoch": 0.4954466608846487, "grad_norm": 0.9932206273078918, "learning_rate": 1.67575801537792e-05, "loss": 0.0148, "step": 4570 }, { "epoch": 0.49555507372072854, "grad_norm": 1.9375851154327393, "learning_rate": 1.6756854780211812e-05, "loss": 0.0118, "step": 4571 }, { "epoch": 0.4956634865568083, "grad_norm": 25.80810546875, "learning_rate": 1.6756129406644424e-05, "loss": 0.0499, "step": 4572 }, { "epoch": 0.4957718993928881, "grad_norm": 0.7454195618629456, "learning_rate": 1.6755404033077036e-05, "loss": 0.0128, "step": 4573 }, { "epoch": 0.4958803122289679, "grad_norm": 3.4728875160217285, "learning_rate": 1.675467865950965e-05, "loss": 0.0461, "step": 4574 }, { "epoch": 0.4959887250650477, "grad_norm": 5.218465805053711, "learning_rate": 1.675395328594226e-05, "loss": 0.0403, "step": 4575 }, { "epoch": 0.49609713790112747, "grad_norm": 1.9097219705581665, "learning_rate": 1.6753227912374873e-05, "loss": 0.0132, "step": 4576 }, { "epoch": 0.4962055507372073, "grad_norm": 1.6981875896453857, "learning_rate": 1.6752502538807488e-05, "loss": 0.0187, "step": 4577 }, { "epoch": 0.49631396357328705, "grad_norm": 2.6043903827667236, "learning_rate": 1.67517771652401e-05, "loss": 0.1201, "step": 4578 }, { "epoch": 0.4964223764093669, "grad_norm": 7.15610408782959, "learning_rate": 1.6751051791672712e-05, "loss": 0.0759, "step": 4579 }, { "epoch": 0.49653078924544664, "grad_norm": 2.3526017665863037, "learning_rate": 1.6750326418105325e-05, "loss": 0.022, "step": 4580 }, { "epoch": 0.49663920208152645, "grad_norm": 6.727084636688232, "learning_rate": 1.6749601044537937e-05, "loss": 0.038, "step": 4581 }, { "epoch": 0.4967476149176062, "grad_norm": 2.263925790786743, "learning_rate": 1.674887567097055e-05, "loss": 0.0388, "step": 4582 }, { "epoch": 0.49685602775368604, "grad_norm": 0.6747921705245972, "learning_rate": 1.6748150297403164e-05, "loss": 0.0077, "step": 4583 }, { "epoch": 0.49696444058976585, "grad_norm": 6.067862510681152, "learning_rate": 1.6747424923835777e-05, "loss": 0.015, "step": 4584 }, { "epoch": 0.4970728534258456, "grad_norm": 3.8982980251312256, "learning_rate": 1.6746699550268392e-05, "loss": 0.0404, "step": 4585 }, { "epoch": 0.49718126626192544, "grad_norm": 2.5156784057617188, "learning_rate": 1.6745974176701004e-05, "loss": 0.0484, "step": 4586 }, { "epoch": 0.4972896790980052, "grad_norm": 3.002103328704834, "learning_rate": 1.6745248803133616e-05, "loss": 0.0628, "step": 4587 }, { "epoch": 0.497398091934085, "grad_norm": 3.2216908931732178, "learning_rate": 1.674452342956623e-05, "loss": 0.0549, "step": 4588 }, { "epoch": 0.4975065047701648, "grad_norm": 0.8334619998931885, "learning_rate": 1.674379805599884e-05, "loss": 0.0164, "step": 4589 }, { "epoch": 0.4976149176062446, "grad_norm": 2.2508327960968018, "learning_rate": 1.6743072682431453e-05, "loss": 0.027, "step": 4590 }, { "epoch": 0.49772333044232436, "grad_norm": 1.3298559188842773, "learning_rate": 1.6742347308864065e-05, "loss": 0.016, "step": 4591 }, { "epoch": 0.4978317432784042, "grad_norm": 3.0454161167144775, "learning_rate": 1.674162193529668e-05, "loss": 0.0559, "step": 4592 }, { "epoch": 0.49794015611448394, "grad_norm": 0.9323140382766724, "learning_rate": 1.6740896561729293e-05, "loss": 0.0065, "step": 4593 }, { "epoch": 0.49804856895056376, "grad_norm": 1.3923903703689575, "learning_rate": 1.6740171188161905e-05, "loss": 0.0374, "step": 4594 }, { "epoch": 0.4981569817866435, "grad_norm": 5.118918418884277, "learning_rate": 1.6739445814594517e-05, "loss": 0.0383, "step": 4595 }, { "epoch": 0.49826539462272335, "grad_norm": 3.9827704429626465, "learning_rate": 1.673872044102713e-05, "loss": 0.0772, "step": 4596 }, { "epoch": 0.4983738074588031, "grad_norm": 3.538287401199341, "learning_rate": 1.673799506745974e-05, "loss": 0.0424, "step": 4597 }, { "epoch": 0.4984822202948829, "grad_norm": 1.844081997871399, "learning_rate": 1.6737269693892357e-05, "loss": 0.0425, "step": 4598 }, { "epoch": 0.4985906331309627, "grad_norm": 1.9174487590789795, "learning_rate": 1.673654432032497e-05, "loss": 0.008, "step": 4599 }, { "epoch": 0.4986990459670425, "grad_norm": 1.7515701055526733, "learning_rate": 1.673581894675758e-05, "loss": 0.0225, "step": 4600 }, { "epoch": 0.4988074588031223, "grad_norm": 1.471077561378479, "learning_rate": 1.6735093573190197e-05, "loss": 0.0477, "step": 4601 }, { "epoch": 0.4989158716392021, "grad_norm": 3.2670891284942627, "learning_rate": 1.673436819962281e-05, "loss": 0.0575, "step": 4602 }, { "epoch": 0.49902428447528185, "grad_norm": 6.962626934051514, "learning_rate": 1.673364282605542e-05, "loss": 0.0313, "step": 4603 }, { "epoch": 0.4991326973113617, "grad_norm": 0.867824375629425, "learning_rate": 1.6732917452488033e-05, "loss": 0.0173, "step": 4604 }, { "epoch": 0.49924111014744144, "grad_norm": 1.392978310585022, "learning_rate": 1.6732192078920645e-05, "loss": 0.0088, "step": 4605 }, { "epoch": 0.49934952298352125, "grad_norm": 0.8677394986152649, "learning_rate": 1.6731466705353257e-05, "loss": 0.0108, "step": 4606 }, { "epoch": 0.499457935819601, "grad_norm": 3.1591622829437256, "learning_rate": 1.6730741331785873e-05, "loss": 0.0412, "step": 4607 }, { "epoch": 0.49956634865568084, "grad_norm": 5.43635892868042, "learning_rate": 1.6730015958218485e-05, "loss": 0.0335, "step": 4608 }, { "epoch": 0.4996747614917606, "grad_norm": 1.688206434249878, "learning_rate": 1.6729290584651097e-05, "loss": 0.0247, "step": 4609 }, { "epoch": 0.4997831743278404, "grad_norm": 0.720911979675293, "learning_rate": 1.672856521108371e-05, "loss": 0.028, "step": 4610 }, { "epoch": 0.4998915871639202, "grad_norm": 10.60647964477539, "learning_rate": 1.672783983751632e-05, "loss": 0.027, "step": 4611 }, { "epoch": 0.5, "grad_norm": 4.073796272277832, "learning_rate": 1.6727114463948934e-05, "loss": 0.0345, "step": 4612 }, { "epoch": 0.5001084128360798, "grad_norm": 0.6723598837852478, "learning_rate": 1.6726389090381546e-05, "loss": 0.0154, "step": 4613 }, { "epoch": 0.5002168256721596, "grad_norm": 2.309962511062622, "learning_rate": 1.672566371681416e-05, "loss": 0.0409, "step": 4614 }, { "epoch": 0.5003252385082394, "grad_norm": 2.2098758220672607, "learning_rate": 1.6724938343246773e-05, "loss": 0.0223, "step": 4615 }, { "epoch": 0.5004336513443192, "grad_norm": 3.280367851257324, "learning_rate": 1.6724212969679385e-05, "loss": 0.0172, "step": 4616 }, { "epoch": 0.5005420641803989, "grad_norm": 0.5616666078567505, "learning_rate": 1.6723487596111998e-05, "loss": 0.016, "step": 4617 }, { "epoch": 0.5006504770164788, "grad_norm": 2.4350593090057373, "learning_rate": 1.6722762222544613e-05, "loss": 0.0423, "step": 4618 }, { "epoch": 0.5007588898525586, "grad_norm": 4.410268783569336, "learning_rate": 1.6722036848977225e-05, "loss": 0.0182, "step": 4619 }, { "epoch": 0.5008673026886383, "grad_norm": 1.6487853527069092, "learning_rate": 1.6721311475409837e-05, "loss": 0.0167, "step": 4620 }, { "epoch": 0.5009757155247181, "grad_norm": 3.046135902404785, "learning_rate": 1.672058610184245e-05, "loss": 0.0221, "step": 4621 }, { "epoch": 0.501084128360798, "grad_norm": 2.5321662425994873, "learning_rate": 1.6719860728275065e-05, "loss": 0.023, "step": 4622 }, { "epoch": 0.5011925411968777, "grad_norm": 7.018228530883789, "learning_rate": 1.6719135354707677e-05, "loss": 0.0639, "step": 4623 }, { "epoch": 0.5013009540329575, "grad_norm": 7.022157669067383, "learning_rate": 1.671840998114029e-05, "loss": 0.0815, "step": 4624 }, { "epoch": 0.5014093668690373, "grad_norm": 4.41215705871582, "learning_rate": 1.67176846075729e-05, "loss": 0.0248, "step": 4625 }, { "epoch": 0.5015177797051171, "grad_norm": 2.9087088108062744, "learning_rate": 1.6716959234005514e-05, "loss": 0.0459, "step": 4626 }, { "epoch": 0.5016261925411969, "grad_norm": 5.627305507659912, "learning_rate": 1.6716233860438126e-05, "loss": 0.0313, "step": 4627 }, { "epoch": 0.5017346053772767, "grad_norm": 1.7134910821914673, "learning_rate": 1.6715508486870738e-05, "loss": 0.0197, "step": 4628 }, { "epoch": 0.5018430182133564, "grad_norm": 0.45932257175445557, "learning_rate": 1.6714783113303354e-05, "loss": 0.0043, "step": 4629 }, { "epoch": 0.5019514310494363, "grad_norm": 12.499899864196777, "learning_rate": 1.6714057739735966e-05, "loss": 0.0401, "step": 4630 }, { "epoch": 0.502059843885516, "grad_norm": 0.70970618724823, "learning_rate": 1.6713332366168578e-05, "loss": 0.0123, "step": 4631 }, { "epoch": 0.5021682567215958, "grad_norm": 2.8219242095947266, "learning_rate": 1.671260699260119e-05, "loss": 0.0198, "step": 4632 }, { "epoch": 0.5022766695576756, "grad_norm": 1.999104619026184, "learning_rate": 1.6711881619033802e-05, "loss": 0.0289, "step": 4633 }, { "epoch": 0.5023850823937555, "grad_norm": 3.5128233432769775, "learning_rate": 1.6711156245466414e-05, "loss": 0.0354, "step": 4634 }, { "epoch": 0.5024934952298352, "grad_norm": 4.638782024383545, "learning_rate": 1.671043087189903e-05, "loss": 0.0228, "step": 4635 }, { "epoch": 0.502601908065915, "grad_norm": 2.7950234413146973, "learning_rate": 1.6709705498331642e-05, "loss": 0.0224, "step": 4636 }, { "epoch": 0.5027103209019947, "grad_norm": 1.5124150514602661, "learning_rate": 1.6708980124764257e-05, "loss": 0.0085, "step": 4637 }, { "epoch": 0.5028187337380746, "grad_norm": 2.970531702041626, "learning_rate": 1.670825475119687e-05, "loss": 0.0219, "step": 4638 }, { "epoch": 0.5029271465741544, "grad_norm": 8.397844314575195, "learning_rate": 1.6707529377629482e-05, "loss": 0.0649, "step": 4639 }, { "epoch": 0.5030355594102341, "grad_norm": 1.2227181196212769, "learning_rate": 1.6706804004062094e-05, "loss": 0.0156, "step": 4640 }, { "epoch": 0.5031439722463139, "grad_norm": 0.9516066908836365, "learning_rate": 1.6706078630494706e-05, "loss": 0.0141, "step": 4641 }, { "epoch": 0.5032523850823938, "grad_norm": 1.6972588300704956, "learning_rate": 1.6705353256927318e-05, "loss": 0.0179, "step": 4642 }, { "epoch": 0.5033607979184735, "grad_norm": 2.9484307765960693, "learning_rate": 1.670462788335993e-05, "loss": 0.0122, "step": 4643 }, { "epoch": 0.5034692107545533, "grad_norm": 2.7875354290008545, "learning_rate": 1.6703902509792546e-05, "loss": 0.0621, "step": 4644 }, { "epoch": 0.5035776235906332, "grad_norm": 4.854384422302246, "learning_rate": 1.6703177136225158e-05, "loss": 0.1129, "step": 4645 }, { "epoch": 0.503686036426713, "grad_norm": 0.6415877938270569, "learning_rate": 1.670245176265777e-05, "loss": 0.0187, "step": 4646 }, { "epoch": 0.5037944492627927, "grad_norm": 5.984022617340088, "learning_rate": 1.6701726389090382e-05, "loss": 0.0499, "step": 4647 }, { "epoch": 0.5039028620988725, "grad_norm": 3.235976457595825, "learning_rate": 1.6701001015522994e-05, "loss": 0.0707, "step": 4648 }, { "epoch": 0.5040112749349523, "grad_norm": 4.473579406738281, "learning_rate": 1.6700275641955607e-05, "loss": 0.0266, "step": 4649 }, { "epoch": 0.5041196877710321, "grad_norm": 3.7063522338867188, "learning_rate": 1.6699550268388222e-05, "loss": 0.0632, "step": 4650 }, { "epoch": 0.5042281006071119, "grad_norm": 4.099485397338867, "learning_rate": 1.6698824894820834e-05, "loss": 0.0129, "step": 4651 }, { "epoch": 0.5043365134431916, "grad_norm": 5.246747970581055, "learning_rate": 1.6698099521253446e-05, "loss": 0.0241, "step": 4652 }, { "epoch": 0.5044449262792715, "grad_norm": 5.779518127441406, "learning_rate": 1.6697374147686062e-05, "loss": 0.0764, "step": 4653 }, { "epoch": 0.5045533391153513, "grad_norm": 4.3614349365234375, "learning_rate": 1.6696648774118674e-05, "loss": 0.0809, "step": 4654 }, { "epoch": 0.504661751951431, "grad_norm": 3.4250073432922363, "learning_rate": 1.6695923400551286e-05, "loss": 0.0253, "step": 4655 }, { "epoch": 0.5047701647875108, "grad_norm": 2.691431760787964, "learning_rate": 1.66951980269839e-05, "loss": 0.0335, "step": 4656 }, { "epoch": 0.5048785776235907, "grad_norm": 4.241539478302002, "learning_rate": 1.669447265341651e-05, "loss": 0.0293, "step": 4657 }, { "epoch": 0.5049869904596704, "grad_norm": 2.233398675918579, "learning_rate": 1.6693747279849123e-05, "loss": 0.0581, "step": 4658 }, { "epoch": 0.5050954032957502, "grad_norm": 4.04689359664917, "learning_rate": 1.6693021906281738e-05, "loss": 0.0431, "step": 4659 }, { "epoch": 0.50520381613183, "grad_norm": 3.405289649963379, "learning_rate": 1.669229653271435e-05, "loss": 0.0571, "step": 4660 }, { "epoch": 0.5053122289679098, "grad_norm": 3.614487409591675, "learning_rate": 1.6691571159146962e-05, "loss": 0.0504, "step": 4661 }, { "epoch": 0.5054206418039896, "grad_norm": 1.0096616744995117, "learning_rate": 1.6690845785579575e-05, "loss": 0.0338, "step": 4662 }, { "epoch": 0.5055290546400694, "grad_norm": 2.5340158939361572, "learning_rate": 1.6690120412012187e-05, "loss": 0.0415, "step": 4663 }, { "epoch": 0.5056374674761491, "grad_norm": 3.778398036956787, "learning_rate": 1.66893950384448e-05, "loss": 0.0672, "step": 4664 }, { "epoch": 0.505745880312229, "grad_norm": 2.145612955093384, "learning_rate": 1.6688669664877414e-05, "loss": 0.0572, "step": 4665 }, { "epoch": 0.5058542931483088, "grad_norm": 3.884467124938965, "learning_rate": 1.6687944291310027e-05, "loss": 0.0702, "step": 4666 }, { "epoch": 0.5059627059843885, "grad_norm": 2.617922306060791, "learning_rate": 1.668721891774264e-05, "loss": 0.0707, "step": 4667 }, { "epoch": 0.5060711188204683, "grad_norm": 3.2416791915893555, "learning_rate": 1.668649354417525e-05, "loss": 0.0371, "step": 4668 }, { "epoch": 0.5061795316565482, "grad_norm": 2.20168399810791, "learning_rate": 1.6685768170607863e-05, "loss": 0.0233, "step": 4669 }, { "epoch": 0.5062879444926279, "grad_norm": 1.4148539304733276, "learning_rate": 1.668504279704048e-05, "loss": 0.0321, "step": 4670 }, { "epoch": 0.5063963573287077, "grad_norm": 1.5899767875671387, "learning_rate": 1.668431742347309e-05, "loss": 0.015, "step": 4671 }, { "epoch": 0.5065047701647876, "grad_norm": 1.6136503219604492, "learning_rate": 1.6683592049905703e-05, "loss": 0.0367, "step": 4672 }, { "epoch": 0.5066131830008673, "grad_norm": 3.0047526359558105, "learning_rate": 1.6682866676338315e-05, "loss": 0.0687, "step": 4673 }, { "epoch": 0.5067215958369471, "grad_norm": 1.845011591911316, "learning_rate": 1.668214130277093e-05, "loss": 0.0519, "step": 4674 }, { "epoch": 0.5068300086730269, "grad_norm": 2.558647394180298, "learning_rate": 1.6681415929203543e-05, "loss": 0.0289, "step": 4675 }, { "epoch": 0.5069384215091067, "grad_norm": 1.444347620010376, "learning_rate": 1.6680690555636155e-05, "loss": 0.0283, "step": 4676 }, { "epoch": 0.5070468343451865, "grad_norm": 0.9866913557052612, "learning_rate": 1.6679965182068767e-05, "loss": 0.0409, "step": 4677 }, { "epoch": 0.5071552471812663, "grad_norm": 1.6847624778747559, "learning_rate": 1.667923980850138e-05, "loss": 0.0268, "step": 4678 }, { "epoch": 0.507263660017346, "grad_norm": 1.5474342107772827, "learning_rate": 1.667851443493399e-05, "loss": 0.0187, "step": 4679 }, { "epoch": 0.5073720728534259, "grad_norm": 3.165225028991699, "learning_rate": 1.6677789061366603e-05, "loss": 0.0409, "step": 4680 }, { "epoch": 0.5074804856895057, "grad_norm": 2.419330596923828, "learning_rate": 1.667706368779922e-05, "loss": 0.0405, "step": 4681 }, { "epoch": 0.5075888985255854, "grad_norm": 1.5893912315368652, "learning_rate": 1.667633831423183e-05, "loss": 0.039, "step": 4682 }, { "epoch": 0.5076973113616652, "grad_norm": 2.8203842639923096, "learning_rate": 1.6675612940664443e-05, "loss": 0.05, "step": 4683 }, { "epoch": 0.5078057241977451, "grad_norm": 2.9951653480529785, "learning_rate": 1.6674887567097055e-05, "loss": 0.0718, "step": 4684 }, { "epoch": 0.5079141370338248, "grad_norm": 2.2293975353240967, "learning_rate": 1.6674162193529667e-05, "loss": 0.0568, "step": 4685 }, { "epoch": 0.5080225498699046, "grad_norm": 5.00160551071167, "learning_rate": 1.667343681996228e-05, "loss": 0.0519, "step": 4686 }, { "epoch": 0.5081309627059843, "grad_norm": 3.452150583267212, "learning_rate": 1.6672711446394895e-05, "loss": 0.0577, "step": 4687 }, { "epoch": 0.5082393755420642, "grad_norm": 2.1809494495391846, "learning_rate": 1.6671986072827507e-05, "loss": 0.0204, "step": 4688 }, { "epoch": 0.508347788378144, "grad_norm": 2.1788086891174316, "learning_rate": 1.6671260699260123e-05, "loss": 0.0459, "step": 4689 }, { "epoch": 0.5084562012142237, "grad_norm": 3.536959648132324, "learning_rate": 1.6670535325692735e-05, "loss": 0.0718, "step": 4690 }, { "epoch": 0.5085646140503035, "grad_norm": 1.3483784198760986, "learning_rate": 1.6669809952125347e-05, "loss": 0.0404, "step": 4691 }, { "epoch": 0.5086730268863834, "grad_norm": 2.654895544052124, "learning_rate": 1.666908457855796e-05, "loss": 0.031, "step": 4692 }, { "epoch": 0.5087814397224631, "grad_norm": 3.2487666606903076, "learning_rate": 1.666835920499057e-05, "loss": 0.0334, "step": 4693 }, { "epoch": 0.5088898525585429, "grad_norm": 2.9564268589019775, "learning_rate": 1.6667633831423183e-05, "loss": 0.0348, "step": 4694 }, { "epoch": 0.5089982653946227, "grad_norm": 1.368147373199463, "learning_rate": 1.6666908457855796e-05, "loss": 0.0149, "step": 4695 }, { "epoch": 0.5091066782307025, "grad_norm": 2.069502830505371, "learning_rate": 1.666618308428841e-05, "loss": 0.0568, "step": 4696 }, { "epoch": 0.5092150910667823, "grad_norm": 0.9909101128578186, "learning_rate": 1.6665457710721023e-05, "loss": 0.0123, "step": 4697 }, { "epoch": 0.5093235039028621, "grad_norm": 1.625754714012146, "learning_rate": 1.6664732337153635e-05, "loss": 0.0313, "step": 4698 }, { "epoch": 0.5094319167389418, "grad_norm": 0.6017817854881287, "learning_rate": 1.6664006963586248e-05, "loss": 0.0231, "step": 4699 }, { "epoch": 0.5095403295750217, "grad_norm": 2.3912508487701416, "learning_rate": 1.666328159001886e-05, "loss": 0.0561, "step": 4700 }, { "epoch": 0.5096487424111015, "grad_norm": 0.5490831136703491, "learning_rate": 1.6662556216451472e-05, "loss": 0.0187, "step": 4701 }, { "epoch": 0.5097571552471812, "grad_norm": 0.4307819902896881, "learning_rate": 1.6661830842884087e-05, "loss": 0.0074, "step": 4702 }, { "epoch": 0.5098655680832611, "grad_norm": 3.7201550006866455, "learning_rate": 1.66611054693167e-05, "loss": 0.0933, "step": 4703 }, { "epoch": 0.5099739809193409, "grad_norm": 1.551867961883545, "learning_rate": 1.666038009574931e-05, "loss": 0.022, "step": 4704 }, { "epoch": 0.5100823937554206, "grad_norm": 3.2314035892486572, "learning_rate": 1.6659654722181927e-05, "loss": 0.0379, "step": 4705 }, { "epoch": 0.5101908065915004, "grad_norm": 3.0626111030578613, "learning_rate": 1.665892934861454e-05, "loss": 0.0479, "step": 4706 }, { "epoch": 0.5102992194275803, "grad_norm": 1.9546220302581787, "learning_rate": 1.665820397504715e-05, "loss": 0.0375, "step": 4707 }, { "epoch": 0.51040763226366, "grad_norm": 1.5838518142700195, "learning_rate": 1.6657478601479764e-05, "loss": 0.0378, "step": 4708 }, { "epoch": 0.5105160450997398, "grad_norm": 4.588263034820557, "learning_rate": 1.6656753227912376e-05, "loss": 0.0472, "step": 4709 }, { "epoch": 0.5106244579358196, "grad_norm": 1.243604063987732, "learning_rate": 1.6656027854344988e-05, "loss": 0.0555, "step": 4710 }, { "epoch": 0.5107328707718994, "grad_norm": 0.5268758535385132, "learning_rate": 1.6655302480777603e-05, "loss": 0.0278, "step": 4711 }, { "epoch": 0.5108412836079792, "grad_norm": 2.3707756996154785, "learning_rate": 1.6654577107210216e-05, "loss": 0.0424, "step": 4712 }, { "epoch": 0.510949696444059, "grad_norm": 4.446888446807861, "learning_rate": 1.6653851733642828e-05, "loss": 0.0609, "step": 4713 }, { "epoch": 0.5110581092801387, "grad_norm": 2.3482816219329834, "learning_rate": 1.665312636007544e-05, "loss": 0.0569, "step": 4714 }, { "epoch": 0.5111665221162186, "grad_norm": 5.142110824584961, "learning_rate": 1.6652400986508052e-05, "loss": 0.0611, "step": 4715 }, { "epoch": 0.5112749349522984, "grad_norm": 2.113913059234619, "learning_rate": 1.6651675612940664e-05, "loss": 0.03, "step": 4716 }, { "epoch": 0.5113833477883781, "grad_norm": 6.703291893005371, "learning_rate": 1.665095023937328e-05, "loss": 0.1108, "step": 4717 }, { "epoch": 0.5114917606244579, "grad_norm": 5.146406173706055, "learning_rate": 1.6650224865805892e-05, "loss": 0.0607, "step": 4718 }, { "epoch": 0.5116001734605378, "grad_norm": 2.0933761596679688, "learning_rate": 1.6649499492238504e-05, "loss": 0.0241, "step": 4719 }, { "epoch": 0.5117085862966175, "grad_norm": 2.7565927505493164, "learning_rate": 1.6648774118671116e-05, "loss": 0.0623, "step": 4720 }, { "epoch": 0.5118169991326973, "grad_norm": 1.040407657623291, "learning_rate": 1.6648048745103728e-05, "loss": 0.022, "step": 4721 }, { "epoch": 0.511925411968777, "grad_norm": 3.2635321617126465, "learning_rate": 1.6647323371536344e-05, "loss": 0.0527, "step": 4722 }, { "epoch": 0.5120338248048569, "grad_norm": 2.8258910179138184, "learning_rate": 1.6646597997968956e-05, "loss": 0.0384, "step": 4723 }, { "epoch": 0.5121422376409367, "grad_norm": 4.039077281951904, "learning_rate": 1.6645872624401568e-05, "loss": 0.0366, "step": 4724 }, { "epoch": 0.5122506504770165, "grad_norm": 2.422011137008667, "learning_rate": 1.664514725083418e-05, "loss": 0.0271, "step": 4725 }, { "epoch": 0.5123590633130962, "grad_norm": 4.758168697357178, "learning_rate": 1.6644421877266796e-05, "loss": 0.0509, "step": 4726 }, { "epoch": 0.5124674761491761, "grad_norm": 2.19637131690979, "learning_rate": 1.6643696503699408e-05, "loss": 0.0363, "step": 4727 }, { "epoch": 0.5125758889852559, "grad_norm": 1.2803664207458496, "learning_rate": 1.664297113013202e-05, "loss": 0.032, "step": 4728 }, { "epoch": 0.5126843018213356, "grad_norm": 3.4110770225524902, "learning_rate": 1.6642245756564632e-05, "loss": 0.0311, "step": 4729 }, { "epoch": 0.5127927146574154, "grad_norm": 2.0112977027893066, "learning_rate": 1.6641520382997244e-05, "loss": 0.0395, "step": 4730 }, { "epoch": 0.5129011274934953, "grad_norm": 0.941669762134552, "learning_rate": 1.6640795009429857e-05, "loss": 0.0051, "step": 4731 }, { "epoch": 0.513009540329575, "grad_norm": 3.080129623413086, "learning_rate": 1.664006963586247e-05, "loss": 0.038, "step": 4732 }, { "epoch": 0.5131179531656548, "grad_norm": 3.3003790378570557, "learning_rate": 1.6639344262295084e-05, "loss": 0.0453, "step": 4733 }, { "epoch": 0.5132263660017347, "grad_norm": 2.146815776824951, "learning_rate": 1.6638618888727696e-05, "loss": 0.0075, "step": 4734 }, { "epoch": 0.5133347788378144, "grad_norm": 1.5571080446243286, "learning_rate": 1.663789351516031e-05, "loss": 0.0387, "step": 4735 }, { "epoch": 0.5134431916738942, "grad_norm": 1.4303092956542969, "learning_rate": 1.663716814159292e-05, "loss": 0.0178, "step": 4736 }, { "epoch": 0.513551604509974, "grad_norm": 3.324676275253296, "learning_rate": 1.6636442768025533e-05, "loss": 0.0251, "step": 4737 }, { "epoch": 0.5136600173460538, "grad_norm": 1.698107123374939, "learning_rate": 1.6635717394458145e-05, "loss": 0.0542, "step": 4738 }, { "epoch": 0.5137684301821336, "grad_norm": 2.8751230239868164, "learning_rate": 1.663499202089076e-05, "loss": 0.0229, "step": 4739 }, { "epoch": 0.5138768430182133, "grad_norm": 1.3554778099060059, "learning_rate": 1.6634266647323373e-05, "loss": 0.0096, "step": 4740 }, { "epoch": 0.5139852558542931, "grad_norm": 6.570154666900635, "learning_rate": 1.6633541273755988e-05, "loss": 0.0424, "step": 4741 }, { "epoch": 0.514093668690373, "grad_norm": 2.8197293281555176, "learning_rate": 1.66328159001886e-05, "loss": 0.0392, "step": 4742 }, { "epoch": 0.5142020815264527, "grad_norm": 1.278984785079956, "learning_rate": 1.6632090526621212e-05, "loss": 0.0266, "step": 4743 }, { "epoch": 0.5143104943625325, "grad_norm": 2.630054235458374, "learning_rate": 1.6631365153053825e-05, "loss": 0.0442, "step": 4744 }, { "epoch": 0.5144189071986123, "grad_norm": 1.240576148033142, "learning_rate": 1.6630639779486437e-05, "loss": 0.0178, "step": 4745 }, { "epoch": 0.5145273200346921, "grad_norm": 2.133028507232666, "learning_rate": 1.662991440591905e-05, "loss": 0.0234, "step": 4746 }, { "epoch": 0.5146357328707719, "grad_norm": 3.232893705368042, "learning_rate": 1.662918903235166e-05, "loss": 0.0381, "step": 4747 }, { "epoch": 0.5147441457068517, "grad_norm": 2.0495011806488037, "learning_rate": 1.6628463658784276e-05, "loss": 0.0383, "step": 4748 }, { "epoch": 0.5148525585429314, "grad_norm": 2.614666700363159, "learning_rate": 1.662773828521689e-05, "loss": 0.0506, "step": 4749 }, { "epoch": 0.5149609713790113, "grad_norm": 1.8030248880386353, "learning_rate": 1.66270129116495e-05, "loss": 0.0154, "step": 4750 }, { "epoch": 0.5150693842150911, "grad_norm": 2.3987057209014893, "learning_rate": 1.6626287538082113e-05, "loss": 0.0159, "step": 4751 }, { "epoch": 0.5151777970511708, "grad_norm": 1.5880264043807983, "learning_rate": 1.6625562164514725e-05, "loss": 0.0241, "step": 4752 }, { "epoch": 0.5152862098872506, "grad_norm": 1.3836405277252197, "learning_rate": 1.6624836790947337e-05, "loss": 0.0166, "step": 4753 }, { "epoch": 0.5153946227233305, "grad_norm": 3.720202684402466, "learning_rate": 1.6624111417379953e-05, "loss": 0.0693, "step": 4754 }, { "epoch": 0.5155030355594102, "grad_norm": 2.2324023246765137, "learning_rate": 1.6623386043812565e-05, "loss": 0.0466, "step": 4755 }, { "epoch": 0.51561144839549, "grad_norm": 6.629436016082764, "learning_rate": 1.6622660670245177e-05, "loss": 0.0327, "step": 4756 }, { "epoch": 0.5157198612315698, "grad_norm": 3.3412349224090576, "learning_rate": 1.6621935296677793e-05, "loss": 0.045, "step": 4757 }, { "epoch": 0.5158282740676496, "grad_norm": 1.4772262573242188, "learning_rate": 1.6621209923110405e-05, "loss": 0.0096, "step": 4758 }, { "epoch": 0.5159366869037294, "grad_norm": 3.9154396057128906, "learning_rate": 1.6620484549543017e-05, "loss": 0.0172, "step": 4759 }, { "epoch": 0.5160450997398092, "grad_norm": 0.5951490998268127, "learning_rate": 1.661975917597563e-05, "loss": 0.0161, "step": 4760 }, { "epoch": 0.516153512575889, "grad_norm": 4.587170600891113, "learning_rate": 1.661903380240824e-05, "loss": 0.0321, "step": 4761 }, { "epoch": 0.5162619254119688, "grad_norm": 1.3029872179031372, "learning_rate": 1.6618308428840853e-05, "loss": 0.0141, "step": 4762 }, { "epoch": 0.5163703382480486, "grad_norm": 1.7868202924728394, "learning_rate": 1.661758305527347e-05, "loss": 0.0191, "step": 4763 }, { "epoch": 0.5164787510841283, "grad_norm": 0.7577477693557739, "learning_rate": 1.661685768170608e-05, "loss": 0.0147, "step": 4764 }, { "epoch": 0.5165871639202082, "grad_norm": 3.343825101852417, "learning_rate": 1.6616132308138693e-05, "loss": 0.0417, "step": 4765 }, { "epoch": 0.516695576756288, "grad_norm": 0.7318441867828369, "learning_rate": 1.6615406934571305e-05, "loss": 0.0043, "step": 4766 }, { "epoch": 0.5168039895923677, "grad_norm": 8.763815879821777, "learning_rate": 1.6614681561003917e-05, "loss": 0.0794, "step": 4767 }, { "epoch": 0.5169124024284475, "grad_norm": 6.615494251251221, "learning_rate": 1.661395618743653e-05, "loss": 0.0195, "step": 4768 }, { "epoch": 0.5170208152645274, "grad_norm": 2.2668938636779785, "learning_rate": 1.6613230813869145e-05, "loss": 0.0517, "step": 4769 }, { "epoch": 0.5171292281006071, "grad_norm": 3.889463186264038, "learning_rate": 1.6612505440301757e-05, "loss": 0.0335, "step": 4770 }, { "epoch": 0.5172376409366869, "grad_norm": 0.9488691687583923, "learning_rate": 1.661178006673437e-05, "loss": 0.0171, "step": 4771 }, { "epoch": 0.5173460537727667, "grad_norm": 2.127366781234741, "learning_rate": 1.661105469316698e-05, "loss": 0.0254, "step": 4772 }, { "epoch": 0.5174544666088465, "grad_norm": 1.5138040781021118, "learning_rate": 1.6610329319599594e-05, "loss": 0.0214, "step": 4773 }, { "epoch": 0.5175628794449263, "grad_norm": 3.6421635150909424, "learning_rate": 1.660960394603221e-05, "loss": 0.052, "step": 4774 }, { "epoch": 0.5176712922810061, "grad_norm": 2.600618600845337, "learning_rate": 1.660887857246482e-05, "loss": 0.0192, "step": 4775 }, { "epoch": 0.5177797051170858, "grad_norm": 5.511606216430664, "learning_rate": 1.6608153198897433e-05, "loss": 0.0289, "step": 4776 }, { "epoch": 0.5178881179531657, "grad_norm": 2.478712797164917, "learning_rate": 1.6607427825330046e-05, "loss": 0.0616, "step": 4777 }, { "epoch": 0.5179965307892455, "grad_norm": 3.4207143783569336, "learning_rate": 1.660670245176266e-05, "loss": 0.0862, "step": 4778 }, { "epoch": 0.5181049436253252, "grad_norm": 6.539920806884766, "learning_rate": 1.6605977078195273e-05, "loss": 0.0819, "step": 4779 }, { "epoch": 0.518213356461405, "grad_norm": 5.26842737197876, "learning_rate": 1.6605251704627885e-05, "loss": 0.0766, "step": 4780 }, { "epoch": 0.5183217692974849, "grad_norm": 3.192938804626465, "learning_rate": 1.6604526331060498e-05, "loss": 0.096, "step": 4781 }, { "epoch": 0.5184301821335646, "grad_norm": 1.5982091426849365, "learning_rate": 1.660380095749311e-05, "loss": 0.0206, "step": 4782 }, { "epoch": 0.5185385949696444, "grad_norm": 1.2237695455551147, "learning_rate": 1.6603075583925722e-05, "loss": 0.0087, "step": 4783 }, { "epoch": 0.5186470078057241, "grad_norm": 0.39789193868637085, "learning_rate": 1.6602350210358334e-05, "loss": 0.007, "step": 4784 }, { "epoch": 0.518755420641804, "grad_norm": 3.040741205215454, "learning_rate": 1.660162483679095e-05, "loss": 0.0196, "step": 4785 }, { "epoch": 0.5188638334778838, "grad_norm": 4.976520538330078, "learning_rate": 1.660089946322356e-05, "loss": 0.0569, "step": 4786 }, { "epoch": 0.5189722463139635, "grad_norm": 3.6858766078948975, "learning_rate": 1.6600174089656174e-05, "loss": 0.0468, "step": 4787 }, { "epoch": 0.5190806591500433, "grad_norm": 3.8684022426605225, "learning_rate": 1.6599448716088786e-05, "loss": 0.0269, "step": 4788 }, { "epoch": 0.5191890719861232, "grad_norm": 2.8854048252105713, "learning_rate": 1.6598723342521398e-05, "loss": 0.0593, "step": 4789 }, { "epoch": 0.519297484822203, "grad_norm": 2.2179453372955322, "learning_rate": 1.659799796895401e-05, "loss": 0.0238, "step": 4790 }, { "epoch": 0.5194058976582827, "grad_norm": 5.206043720245361, "learning_rate": 1.6597272595386626e-05, "loss": 0.084, "step": 4791 }, { "epoch": 0.5195143104943626, "grad_norm": 0.38703685998916626, "learning_rate": 1.6596547221819238e-05, "loss": 0.0023, "step": 4792 }, { "epoch": 0.5196227233304423, "grad_norm": 4.158782958984375, "learning_rate": 1.6595821848251853e-05, "loss": 0.0648, "step": 4793 }, { "epoch": 0.5197311361665221, "grad_norm": 6.571321487426758, "learning_rate": 1.6595096474684466e-05, "loss": 0.0654, "step": 4794 }, { "epoch": 0.5198395490026019, "grad_norm": 3.850986957550049, "learning_rate": 1.6594371101117078e-05, "loss": 0.0548, "step": 4795 }, { "epoch": 0.5199479618386817, "grad_norm": 6.811956405639648, "learning_rate": 1.659364572754969e-05, "loss": 0.0433, "step": 4796 }, { "epoch": 0.5200563746747615, "grad_norm": 2.1202142238616943, "learning_rate": 1.6592920353982302e-05, "loss": 0.0302, "step": 4797 }, { "epoch": 0.5201647875108413, "grad_norm": 5.09205436706543, "learning_rate": 1.6592194980414914e-05, "loss": 0.0692, "step": 4798 }, { "epoch": 0.520273200346921, "grad_norm": 2.463858127593994, "learning_rate": 1.6591469606847526e-05, "loss": 0.0414, "step": 4799 }, { "epoch": 0.5203816131830009, "grad_norm": 1.4055241346359253, "learning_rate": 1.6590744233280142e-05, "loss": 0.0174, "step": 4800 }, { "epoch": 0.5204900260190807, "grad_norm": 2.241574287414551, "learning_rate": 1.6590018859712754e-05, "loss": 0.0515, "step": 4801 }, { "epoch": 0.5205984388551604, "grad_norm": 2.878434896469116, "learning_rate": 1.6589293486145366e-05, "loss": 0.0265, "step": 4802 }, { "epoch": 0.5207068516912402, "grad_norm": 2.561225414276123, "learning_rate": 1.6588568112577978e-05, "loss": 0.0255, "step": 4803 }, { "epoch": 0.5208152645273201, "grad_norm": 1.6549104452133179, "learning_rate": 1.658784273901059e-05, "loss": 0.0321, "step": 4804 }, { "epoch": 0.5209236773633998, "grad_norm": 2.4827325344085693, "learning_rate": 1.6587117365443203e-05, "loss": 0.0791, "step": 4805 }, { "epoch": 0.5210320901994796, "grad_norm": 0.8889358043670654, "learning_rate": 1.6586391991875818e-05, "loss": 0.0197, "step": 4806 }, { "epoch": 0.5211405030355594, "grad_norm": 0.8632251024246216, "learning_rate": 1.658566661830843e-05, "loss": 0.028, "step": 4807 }, { "epoch": 0.5212489158716392, "grad_norm": 2.8761301040649414, "learning_rate": 1.6584941244741042e-05, "loss": 0.0307, "step": 4808 }, { "epoch": 0.521357328707719, "grad_norm": 0.28180307149887085, "learning_rate": 1.6584215871173658e-05, "loss": 0.003, "step": 4809 }, { "epoch": 0.5214657415437988, "grad_norm": 2.546570062637329, "learning_rate": 1.658349049760627e-05, "loss": 0.0474, "step": 4810 }, { "epoch": 0.5215741543798785, "grad_norm": 1.2264387607574463, "learning_rate": 1.6582765124038882e-05, "loss": 0.0166, "step": 4811 }, { "epoch": 0.5216825672159584, "grad_norm": 3.640684127807617, "learning_rate": 1.6582039750471494e-05, "loss": 0.0441, "step": 4812 }, { "epoch": 0.5217909800520382, "grad_norm": 2.148681163787842, "learning_rate": 1.6581314376904106e-05, "loss": 0.0264, "step": 4813 }, { "epoch": 0.5218993928881179, "grad_norm": 2.5814874172210693, "learning_rate": 1.658058900333672e-05, "loss": 0.0439, "step": 4814 }, { "epoch": 0.5220078057241977, "grad_norm": 3.4986073970794678, "learning_rate": 1.6579863629769334e-05, "loss": 0.0386, "step": 4815 }, { "epoch": 0.5221162185602776, "grad_norm": 1.6803544759750366, "learning_rate": 1.6579138256201946e-05, "loss": 0.041, "step": 4816 }, { "epoch": 0.5222246313963573, "grad_norm": 2.6592748165130615, "learning_rate": 1.657841288263456e-05, "loss": 0.0656, "step": 4817 }, { "epoch": 0.5223330442324371, "grad_norm": 1.2182376384735107, "learning_rate": 1.657768750906717e-05, "loss": 0.0128, "step": 4818 }, { "epoch": 0.522441457068517, "grad_norm": 2.930098295211792, "learning_rate": 1.6576962135499783e-05, "loss": 0.0577, "step": 4819 }, { "epoch": 0.5225498699045967, "grad_norm": 1.395802617073059, "learning_rate": 1.6576236761932395e-05, "loss": 0.0552, "step": 4820 }, { "epoch": 0.5226582827406765, "grad_norm": 2.1584630012512207, "learning_rate": 1.657551138836501e-05, "loss": 0.0149, "step": 4821 }, { "epoch": 0.5227666955767563, "grad_norm": 1.5434315204620361, "learning_rate": 1.6574786014797623e-05, "loss": 0.0126, "step": 4822 }, { "epoch": 0.5228751084128361, "grad_norm": 2.3089964389801025, "learning_rate": 1.6574060641230235e-05, "loss": 0.0224, "step": 4823 }, { "epoch": 0.5229835212489159, "grad_norm": 1.7199569940567017, "learning_rate": 1.6573335267662847e-05, "loss": 0.02, "step": 4824 }, { "epoch": 0.5230919340849957, "grad_norm": 4.669802665710449, "learning_rate": 1.6572609894095462e-05, "loss": 0.0728, "step": 4825 }, { "epoch": 0.5232003469210754, "grad_norm": 3.8155806064605713, "learning_rate": 1.6571884520528074e-05, "loss": 0.069, "step": 4826 }, { "epoch": 0.5233087597571553, "grad_norm": 0.6854072213172913, "learning_rate": 1.6571159146960687e-05, "loss": 0.0291, "step": 4827 }, { "epoch": 0.5234171725932351, "grad_norm": 1.4976491928100586, "learning_rate": 1.65704337733933e-05, "loss": 0.0188, "step": 4828 }, { "epoch": 0.5235255854293148, "grad_norm": 9.786145210266113, "learning_rate": 1.656970839982591e-05, "loss": 0.0554, "step": 4829 }, { "epoch": 0.5236339982653946, "grad_norm": 1.1341326236724854, "learning_rate": 1.6568983026258526e-05, "loss": 0.0281, "step": 4830 }, { "epoch": 0.5237424111014745, "grad_norm": 0.7541716694831848, "learning_rate": 1.656825765269114e-05, "loss": 0.0177, "step": 4831 }, { "epoch": 0.5238508239375542, "grad_norm": 2.9481186866760254, "learning_rate": 1.656753227912375e-05, "loss": 0.0209, "step": 4832 }, { "epoch": 0.523959236773634, "grad_norm": 0.41247493028640747, "learning_rate": 1.6566806905556363e-05, "loss": 0.0035, "step": 4833 }, { "epoch": 0.5240676496097137, "grad_norm": 2.7307345867156982, "learning_rate": 1.6566081531988975e-05, "loss": 0.0616, "step": 4834 }, { "epoch": 0.5241760624457936, "grad_norm": 5.462207794189453, "learning_rate": 1.6565356158421587e-05, "loss": 0.0295, "step": 4835 }, { "epoch": 0.5242844752818734, "grad_norm": 2.279947519302368, "learning_rate": 1.65646307848542e-05, "loss": 0.042, "step": 4836 }, { "epoch": 0.5243928881179531, "grad_norm": 9.009321212768555, "learning_rate": 1.6563905411286815e-05, "loss": 0.0497, "step": 4837 }, { "epoch": 0.5245013009540329, "grad_norm": 4.325627326965332, "learning_rate": 1.6563180037719427e-05, "loss": 0.0254, "step": 4838 }, { "epoch": 0.5246097137901128, "grad_norm": 4.046699523925781, "learning_rate": 1.656245466415204e-05, "loss": 0.0387, "step": 4839 }, { "epoch": 0.5247181266261925, "grad_norm": 6.007988929748535, "learning_rate": 1.656172929058465e-05, "loss": 0.0681, "step": 4840 }, { "epoch": 0.5248265394622723, "grad_norm": 3.1488282680511475, "learning_rate": 1.6561003917017263e-05, "loss": 0.0652, "step": 4841 }, { "epoch": 0.5249349522983521, "grad_norm": 0.686363697052002, "learning_rate": 1.6560278543449876e-05, "loss": 0.0135, "step": 4842 }, { "epoch": 0.525043365134432, "grad_norm": 12.757492065429688, "learning_rate": 1.655955316988249e-05, "loss": 0.0759, "step": 4843 }, { "epoch": 0.5251517779705117, "grad_norm": 1.5466468334197998, "learning_rate": 1.6558827796315103e-05, "loss": 0.0516, "step": 4844 }, { "epoch": 0.5252601908065915, "grad_norm": 4.417084217071533, "learning_rate": 1.655810242274772e-05, "loss": 0.048, "step": 4845 }, { "epoch": 0.5253686036426712, "grad_norm": 2.004110813140869, "learning_rate": 1.655737704918033e-05, "loss": 0.0164, "step": 4846 }, { "epoch": 0.5254770164787511, "grad_norm": 1.0138249397277832, "learning_rate": 1.6556651675612943e-05, "loss": 0.0329, "step": 4847 }, { "epoch": 0.5255854293148309, "grad_norm": 4.999775409698486, "learning_rate": 1.6555926302045555e-05, "loss": 0.0673, "step": 4848 }, { "epoch": 0.5256938421509106, "grad_norm": 2.656705617904663, "learning_rate": 1.6555200928478167e-05, "loss": 0.0395, "step": 4849 }, { "epoch": 0.5258022549869905, "grad_norm": 4.446071624755859, "learning_rate": 1.655447555491078e-05, "loss": 0.0348, "step": 4850 }, { "epoch": 0.5259106678230703, "grad_norm": 5.093303203582764, "learning_rate": 1.655375018134339e-05, "loss": 0.0459, "step": 4851 }, { "epoch": 0.52601908065915, "grad_norm": 3.855668067932129, "learning_rate": 1.6553024807776007e-05, "loss": 0.0173, "step": 4852 }, { "epoch": 0.5261274934952298, "grad_norm": 8.625114440917969, "learning_rate": 1.655229943420862e-05, "loss": 0.0803, "step": 4853 }, { "epoch": 0.5262359063313097, "grad_norm": 2.6900758743286133, "learning_rate": 1.655157406064123e-05, "loss": 0.0399, "step": 4854 }, { "epoch": 0.5263443191673894, "grad_norm": 2.576582908630371, "learning_rate": 1.6550848687073844e-05, "loss": 0.0439, "step": 4855 }, { "epoch": 0.5264527320034692, "grad_norm": 1.1199408769607544, "learning_rate": 1.6550123313506456e-05, "loss": 0.0172, "step": 4856 }, { "epoch": 0.526561144839549, "grad_norm": 7.2954559326171875, "learning_rate": 1.6549397939939068e-05, "loss": 0.0991, "step": 4857 }, { "epoch": 0.5266695576756288, "grad_norm": 3.57216477394104, "learning_rate": 1.6548672566371683e-05, "loss": 0.0792, "step": 4858 }, { "epoch": 0.5267779705117086, "grad_norm": 6.219527244567871, "learning_rate": 1.6547947192804296e-05, "loss": 0.0461, "step": 4859 }, { "epoch": 0.5268863833477884, "grad_norm": 3.871593475341797, "learning_rate": 1.654722181923691e-05, "loss": 0.034, "step": 4860 }, { "epoch": 0.5269947961838681, "grad_norm": 6.5556721687316895, "learning_rate": 1.6546496445669523e-05, "loss": 0.0752, "step": 4861 }, { "epoch": 0.527103209019948, "grad_norm": 1.184329628944397, "learning_rate": 1.6545771072102135e-05, "loss": 0.035, "step": 4862 }, { "epoch": 0.5272116218560278, "grad_norm": 0.5196889042854309, "learning_rate": 1.6545045698534748e-05, "loss": 0.013, "step": 4863 }, { "epoch": 0.5273200346921075, "grad_norm": 2.5588924884796143, "learning_rate": 1.654432032496736e-05, "loss": 0.0514, "step": 4864 }, { "epoch": 0.5274284475281873, "grad_norm": 2.9014759063720703, "learning_rate": 1.6543594951399972e-05, "loss": 0.0083, "step": 4865 }, { "epoch": 0.5275368603642672, "grad_norm": 0.892989456653595, "learning_rate": 1.6542869577832584e-05, "loss": 0.0124, "step": 4866 }, { "epoch": 0.5276452732003469, "grad_norm": 2.2939999103546143, "learning_rate": 1.65421442042652e-05, "loss": 0.0681, "step": 4867 }, { "epoch": 0.5277536860364267, "grad_norm": 8.537232398986816, "learning_rate": 1.654141883069781e-05, "loss": 0.0457, "step": 4868 }, { "epoch": 0.5278620988725065, "grad_norm": 4.428356170654297, "learning_rate": 1.6540693457130424e-05, "loss": 0.0289, "step": 4869 }, { "epoch": 0.5279705117085863, "grad_norm": 0.8922614455223083, "learning_rate": 1.6539968083563036e-05, "loss": 0.0165, "step": 4870 }, { "epoch": 0.5280789245446661, "grad_norm": 4.118345737457275, "learning_rate": 1.6539242709995648e-05, "loss": 0.0605, "step": 4871 }, { "epoch": 0.5281873373807459, "grad_norm": 5.3826069831848145, "learning_rate": 1.653851733642826e-05, "loss": 0.0489, "step": 4872 }, { "epoch": 0.5282957502168256, "grad_norm": 0.9704894423484802, "learning_rate": 1.6537791962860876e-05, "loss": 0.0185, "step": 4873 }, { "epoch": 0.5284041630529055, "grad_norm": 1.6940116882324219, "learning_rate": 1.6537066589293488e-05, "loss": 0.0346, "step": 4874 }, { "epoch": 0.5285125758889853, "grad_norm": 3.8580732345581055, "learning_rate": 1.65363412157261e-05, "loss": 0.0536, "step": 4875 }, { "epoch": 0.528620988725065, "grad_norm": 4.227573871612549, "learning_rate": 1.6535615842158712e-05, "loss": 0.0697, "step": 4876 }, { "epoch": 0.5287294015611448, "grad_norm": 3.3042407035827637, "learning_rate": 1.6534890468591328e-05, "loss": 0.0248, "step": 4877 }, { "epoch": 0.5288378143972247, "grad_norm": 2.0241687297821045, "learning_rate": 1.653416509502394e-05, "loss": 0.0347, "step": 4878 }, { "epoch": 0.5289462272333044, "grad_norm": 1.2082266807556152, "learning_rate": 1.6533439721456552e-05, "loss": 0.0281, "step": 4879 }, { "epoch": 0.5290546400693842, "grad_norm": 4.893815517425537, "learning_rate": 1.6532714347889164e-05, "loss": 0.0846, "step": 4880 }, { "epoch": 0.5291630529054641, "grad_norm": 9.159770965576172, "learning_rate": 1.6531988974321776e-05, "loss": 0.0486, "step": 4881 }, { "epoch": 0.5292714657415438, "grad_norm": 2.175572633743286, "learning_rate": 1.6531263600754392e-05, "loss": 0.0579, "step": 4882 }, { "epoch": 0.5293798785776236, "grad_norm": 3.392620801925659, "learning_rate": 1.6530538227187004e-05, "loss": 0.0592, "step": 4883 }, { "epoch": 0.5294882914137033, "grad_norm": 0.8980399966239929, "learning_rate": 1.6529812853619616e-05, "loss": 0.0162, "step": 4884 }, { "epoch": 0.5295967042497832, "grad_norm": 10.406543731689453, "learning_rate": 1.6529087480052228e-05, "loss": 0.0927, "step": 4885 }, { "epoch": 0.529705117085863, "grad_norm": 0.6371509432792664, "learning_rate": 1.652836210648484e-05, "loss": 0.0157, "step": 4886 }, { "epoch": 0.5298135299219427, "grad_norm": 0.8844029307365417, "learning_rate": 1.6527636732917453e-05, "loss": 0.0209, "step": 4887 }, { "epoch": 0.5299219427580225, "grad_norm": 0.8563445806503296, "learning_rate": 1.6526911359350065e-05, "loss": 0.009, "step": 4888 }, { "epoch": 0.5300303555941024, "grad_norm": 2.8783016204833984, "learning_rate": 1.652618598578268e-05, "loss": 0.0139, "step": 4889 }, { "epoch": 0.5301387684301822, "grad_norm": 1.3741165399551392, "learning_rate": 1.6525460612215292e-05, "loss": 0.0201, "step": 4890 }, { "epoch": 0.5302471812662619, "grad_norm": 3.586292266845703, "learning_rate": 1.6524735238647904e-05, "loss": 0.0359, "step": 4891 }, { "epoch": 0.5303555941023417, "grad_norm": 0.773844838142395, "learning_rate": 1.6524009865080517e-05, "loss": 0.0217, "step": 4892 }, { "epoch": 0.5304640069384216, "grad_norm": 2.029371500015259, "learning_rate": 1.652328449151313e-05, "loss": 0.0351, "step": 4893 }, { "epoch": 0.5305724197745013, "grad_norm": 1.0808817148208618, "learning_rate": 1.652255911794574e-05, "loss": 0.0124, "step": 4894 }, { "epoch": 0.5306808326105811, "grad_norm": 2.2552578449249268, "learning_rate": 1.6521833744378356e-05, "loss": 0.0338, "step": 4895 }, { "epoch": 0.5307892454466608, "grad_norm": 1.2048685550689697, "learning_rate": 1.652110837081097e-05, "loss": 0.0179, "step": 4896 }, { "epoch": 0.5308976582827407, "grad_norm": 0.21613402664661407, "learning_rate": 1.6520382997243584e-05, "loss": 0.0017, "step": 4897 }, { "epoch": 0.5310060711188205, "grad_norm": 0.7443540692329407, "learning_rate": 1.6519657623676196e-05, "loss": 0.0029, "step": 4898 }, { "epoch": 0.5311144839549002, "grad_norm": 6.86732292175293, "learning_rate": 1.651893225010881e-05, "loss": 0.0235, "step": 4899 }, { "epoch": 0.53122289679098, "grad_norm": 5.361900329589844, "learning_rate": 1.651820687654142e-05, "loss": 0.0755, "step": 4900 }, { "epoch": 0.5313313096270599, "grad_norm": 0.6296424865722656, "learning_rate": 1.6517481502974033e-05, "loss": 0.0017, "step": 4901 }, { "epoch": 0.5314397224631396, "grad_norm": 2.7888309955596924, "learning_rate": 1.6516756129406645e-05, "loss": 0.0247, "step": 4902 }, { "epoch": 0.5315481352992194, "grad_norm": 5.853371620178223, "learning_rate": 1.6516030755839257e-05, "loss": 0.0221, "step": 4903 }, { "epoch": 0.5316565481352992, "grad_norm": 1.4986950159072876, "learning_rate": 1.6515305382271873e-05, "loss": 0.0636, "step": 4904 }, { "epoch": 0.531764960971379, "grad_norm": 5.759965896606445, "learning_rate": 1.6514580008704485e-05, "loss": 0.047, "step": 4905 }, { "epoch": 0.5318733738074588, "grad_norm": 3.5595264434814453, "learning_rate": 1.6513854635137097e-05, "loss": 0.0711, "step": 4906 }, { "epoch": 0.5319817866435386, "grad_norm": 3.7136778831481934, "learning_rate": 1.651312926156971e-05, "loss": 0.017, "step": 4907 }, { "epoch": 0.5320901994796184, "grad_norm": 3.7122414112091064, "learning_rate": 1.651240388800232e-05, "loss": 0.0738, "step": 4908 }, { "epoch": 0.5321986123156982, "grad_norm": 7.437536239624023, "learning_rate": 1.6511678514434933e-05, "loss": 0.1197, "step": 4909 }, { "epoch": 0.532307025151778, "grad_norm": 1.9492765665054321, "learning_rate": 1.651095314086755e-05, "loss": 0.0554, "step": 4910 }, { "epoch": 0.5324154379878577, "grad_norm": 3.550616502761841, "learning_rate": 1.651022776730016e-05, "loss": 0.0558, "step": 4911 }, { "epoch": 0.5325238508239376, "grad_norm": 3.698293447494507, "learning_rate": 1.6509502393732776e-05, "loss": 0.0176, "step": 4912 }, { "epoch": 0.5326322636600174, "grad_norm": 1.118342638015747, "learning_rate": 1.650877702016539e-05, "loss": 0.0169, "step": 4913 }, { "epoch": 0.5327406764960971, "grad_norm": 3.3858449459075928, "learning_rate": 1.6508051646598e-05, "loss": 0.0125, "step": 4914 }, { "epoch": 0.5328490893321769, "grad_norm": 5.5049262046813965, "learning_rate": 1.6507326273030613e-05, "loss": 0.0858, "step": 4915 }, { "epoch": 0.5329575021682568, "grad_norm": 3.238050699234009, "learning_rate": 1.6506600899463225e-05, "loss": 0.0735, "step": 4916 }, { "epoch": 0.5330659150043365, "grad_norm": 2.2523694038391113, "learning_rate": 1.6505875525895837e-05, "loss": 0.0302, "step": 4917 }, { "epoch": 0.5331743278404163, "grad_norm": 4.210996627807617, "learning_rate": 1.650515015232845e-05, "loss": 0.0855, "step": 4918 }, { "epoch": 0.5332827406764961, "grad_norm": 1.316849946975708, "learning_rate": 1.6504424778761065e-05, "loss": 0.0126, "step": 4919 }, { "epoch": 0.5333911535125759, "grad_norm": 0.708992600440979, "learning_rate": 1.6503699405193677e-05, "loss": 0.0239, "step": 4920 }, { "epoch": 0.5334995663486557, "grad_norm": 2.276233196258545, "learning_rate": 1.650297403162629e-05, "loss": 0.0384, "step": 4921 }, { "epoch": 0.5336079791847355, "grad_norm": 17.972082138061523, "learning_rate": 1.65022486580589e-05, "loss": 0.0586, "step": 4922 }, { "epoch": 0.5337163920208152, "grad_norm": 2.426135301589966, "learning_rate": 1.6501523284491513e-05, "loss": 0.0487, "step": 4923 }, { "epoch": 0.5338248048568951, "grad_norm": 4.063471794128418, "learning_rate": 1.6500797910924126e-05, "loss": 0.081, "step": 4924 }, { "epoch": 0.5339332176929749, "grad_norm": 2.1733133792877197, "learning_rate": 1.650007253735674e-05, "loss": 0.023, "step": 4925 }, { "epoch": 0.5340416305290546, "grad_norm": 7.793503761291504, "learning_rate": 1.6499347163789353e-05, "loss": 0.0581, "step": 4926 }, { "epoch": 0.5341500433651344, "grad_norm": 4.763919353485107, "learning_rate": 1.6498621790221965e-05, "loss": 0.0489, "step": 4927 }, { "epoch": 0.5342584562012143, "grad_norm": 2.8081002235412598, "learning_rate": 1.6497896416654577e-05, "loss": 0.0324, "step": 4928 }, { "epoch": 0.534366869037294, "grad_norm": 1.3884551525115967, "learning_rate": 1.6497171043087193e-05, "loss": 0.0395, "step": 4929 }, { "epoch": 0.5344752818733738, "grad_norm": 1.9022552967071533, "learning_rate": 1.6496445669519805e-05, "loss": 0.0662, "step": 4930 }, { "epoch": 0.5345836947094535, "grad_norm": 1.5211238861083984, "learning_rate": 1.6495720295952417e-05, "loss": 0.0109, "step": 4931 }, { "epoch": 0.5346921075455334, "grad_norm": 2.686406373977661, "learning_rate": 1.649499492238503e-05, "loss": 0.0278, "step": 4932 }, { "epoch": 0.5348005203816132, "grad_norm": 2.101771116256714, "learning_rate": 1.649426954881764e-05, "loss": 0.0244, "step": 4933 }, { "epoch": 0.534908933217693, "grad_norm": 2.1245384216308594, "learning_rate": 1.6493544175250257e-05, "loss": 0.0216, "step": 4934 }, { "epoch": 0.5350173460537727, "grad_norm": 3.2011919021606445, "learning_rate": 1.649281880168287e-05, "loss": 0.0797, "step": 4935 }, { "epoch": 0.5351257588898526, "grad_norm": 1.54970121383667, "learning_rate": 1.649209342811548e-05, "loss": 0.0644, "step": 4936 }, { "epoch": 0.5352341717259324, "grad_norm": 0.8893539309501648, "learning_rate": 1.6491368054548094e-05, "loss": 0.0125, "step": 4937 }, { "epoch": 0.5353425845620121, "grad_norm": 3.991703987121582, "learning_rate": 1.6490642680980706e-05, "loss": 0.0418, "step": 4938 }, { "epoch": 0.535450997398092, "grad_norm": 2.699586868286133, "learning_rate": 1.6489917307413318e-05, "loss": 0.0091, "step": 4939 }, { "epoch": 0.5355594102341718, "grad_norm": 4.459163665771484, "learning_rate": 1.648919193384593e-05, "loss": 0.0414, "step": 4940 }, { "epoch": 0.5356678230702515, "grad_norm": 1.2589099407196045, "learning_rate": 1.6488466560278546e-05, "loss": 0.015, "step": 4941 }, { "epoch": 0.5357762359063313, "grad_norm": 2.347491502761841, "learning_rate": 1.6487741186711158e-05, "loss": 0.0464, "step": 4942 }, { "epoch": 0.5358846487424112, "grad_norm": 1.1243704557418823, "learning_rate": 1.648701581314377e-05, "loss": 0.0103, "step": 4943 }, { "epoch": 0.5359930615784909, "grad_norm": 0.9532299041748047, "learning_rate": 1.6486290439576382e-05, "loss": 0.022, "step": 4944 }, { "epoch": 0.5361014744145707, "grad_norm": 1.3415271043777466, "learning_rate": 1.6485565066008994e-05, "loss": 0.0105, "step": 4945 }, { "epoch": 0.5362098872506504, "grad_norm": 1.5306636095046997, "learning_rate": 1.648483969244161e-05, "loss": 0.021, "step": 4946 }, { "epoch": 0.5363183000867303, "grad_norm": 2.278413772583008, "learning_rate": 1.6484114318874222e-05, "loss": 0.0429, "step": 4947 }, { "epoch": 0.5364267129228101, "grad_norm": 2.476881265640259, "learning_rate": 1.6483388945306834e-05, "loss": 0.0284, "step": 4948 }, { "epoch": 0.5365351257588898, "grad_norm": 5.571064472198486, "learning_rate": 1.648266357173945e-05, "loss": 0.043, "step": 4949 }, { "epoch": 0.5366435385949696, "grad_norm": 3.522880792617798, "learning_rate": 1.648193819817206e-05, "loss": 0.0366, "step": 4950 }, { "epoch": 0.5367519514310495, "grad_norm": 3.446202039718628, "learning_rate": 1.6481212824604674e-05, "loss": 0.034, "step": 4951 }, { "epoch": 0.5368603642671292, "grad_norm": 1.367640733718872, "learning_rate": 1.6480487451037286e-05, "loss": 0.025, "step": 4952 }, { "epoch": 0.536968777103209, "grad_norm": 0.8813208937644958, "learning_rate": 1.6479762077469898e-05, "loss": 0.0108, "step": 4953 }, { "epoch": 0.5370771899392888, "grad_norm": 0.36750221252441406, "learning_rate": 1.647903670390251e-05, "loss": 0.0124, "step": 4954 }, { "epoch": 0.5371856027753686, "grad_norm": 3.0766513347625732, "learning_rate": 1.6478311330335122e-05, "loss": 0.0207, "step": 4955 }, { "epoch": 0.5372940156114484, "grad_norm": 4.7988691329956055, "learning_rate": 1.6477585956767738e-05, "loss": 0.0619, "step": 4956 }, { "epoch": 0.5374024284475282, "grad_norm": 1.605661392211914, "learning_rate": 1.647686058320035e-05, "loss": 0.0263, "step": 4957 }, { "epoch": 0.5375108412836079, "grad_norm": 3.599698305130005, "learning_rate": 1.6476135209632962e-05, "loss": 0.0793, "step": 4958 }, { "epoch": 0.5376192541196878, "grad_norm": 1.671757459640503, "learning_rate": 1.6475409836065574e-05, "loss": 0.0353, "step": 4959 }, { "epoch": 0.5377276669557676, "grad_norm": 2.6775243282318115, "learning_rate": 1.6474684462498186e-05, "loss": 0.0225, "step": 4960 }, { "epoch": 0.5378360797918473, "grad_norm": 0.8584907054901123, "learning_rate": 1.64739590889308e-05, "loss": 0.0063, "step": 4961 }, { "epoch": 0.5379444926279271, "grad_norm": 0.3557232916355133, "learning_rate": 1.6473233715363414e-05, "loss": 0.005, "step": 4962 }, { "epoch": 0.538052905464007, "grad_norm": 2.0960171222686768, "learning_rate": 1.6472508341796026e-05, "loss": 0.0096, "step": 4963 }, { "epoch": 0.5381613183000867, "grad_norm": 2.848316192626953, "learning_rate": 1.6471782968228642e-05, "loss": 0.0384, "step": 4964 }, { "epoch": 0.5382697311361665, "grad_norm": 1.8942410945892334, "learning_rate": 1.6471057594661254e-05, "loss": 0.0235, "step": 4965 }, { "epoch": 0.5383781439722463, "grad_norm": 4.193058967590332, "learning_rate": 1.6470332221093866e-05, "loss": 0.0442, "step": 4966 }, { "epoch": 0.5384865568083261, "grad_norm": 2.022921562194824, "learning_rate": 1.6469606847526478e-05, "loss": 0.0249, "step": 4967 }, { "epoch": 0.5385949696444059, "grad_norm": 0.8512078523635864, "learning_rate": 1.646888147395909e-05, "loss": 0.0166, "step": 4968 }, { "epoch": 0.5387033824804857, "grad_norm": 3.8745813369750977, "learning_rate": 1.6468156100391702e-05, "loss": 0.0345, "step": 4969 }, { "epoch": 0.5388117953165655, "grad_norm": 1.6269311904907227, "learning_rate": 1.6467430726824315e-05, "loss": 0.0717, "step": 4970 }, { "epoch": 0.5389202081526453, "grad_norm": 1.1615952253341675, "learning_rate": 1.646670535325693e-05, "loss": 0.0047, "step": 4971 }, { "epoch": 0.5390286209887251, "grad_norm": 2.334136724472046, "learning_rate": 1.6465979979689542e-05, "loss": 0.0097, "step": 4972 }, { "epoch": 0.5391370338248048, "grad_norm": 0.5683445334434509, "learning_rate": 1.6465254606122154e-05, "loss": 0.0067, "step": 4973 }, { "epoch": 0.5392454466608847, "grad_norm": 7.856858730316162, "learning_rate": 1.6464529232554767e-05, "loss": 0.0467, "step": 4974 }, { "epoch": 0.5393538594969645, "grad_norm": 3.888310670852661, "learning_rate": 1.646380385898738e-05, "loss": 0.0331, "step": 4975 }, { "epoch": 0.5394622723330442, "grad_norm": 2.464055299758911, "learning_rate": 1.646307848541999e-05, "loss": 0.0885, "step": 4976 }, { "epoch": 0.539570685169124, "grad_norm": 3.808642625808716, "learning_rate": 1.6462353111852606e-05, "loss": 0.0197, "step": 4977 }, { "epoch": 0.5396790980052039, "grad_norm": 0.9161186218261719, "learning_rate": 1.646162773828522e-05, "loss": 0.0055, "step": 4978 }, { "epoch": 0.5397875108412836, "grad_norm": 3.7072536945343018, "learning_rate": 1.646090236471783e-05, "loss": 0.0491, "step": 4979 }, { "epoch": 0.5398959236773634, "grad_norm": 0.909213125705719, "learning_rate": 1.6460176991150443e-05, "loss": 0.0036, "step": 4980 }, { "epoch": 0.5400043365134432, "grad_norm": 5.192992687225342, "learning_rate": 1.645945161758306e-05, "loss": 0.0574, "step": 4981 }, { "epoch": 0.540112749349523, "grad_norm": 4.310261249542236, "learning_rate": 1.645872624401567e-05, "loss": 0.0746, "step": 4982 }, { "epoch": 0.5402211621856028, "grad_norm": 2.0778801441192627, "learning_rate": 1.6458000870448283e-05, "loss": 0.051, "step": 4983 }, { "epoch": 0.5403295750216826, "grad_norm": 3.529345750808716, "learning_rate": 1.6457275496880895e-05, "loss": 0.0466, "step": 4984 }, { "epoch": 0.5404379878577623, "grad_norm": 2.1594152450561523, "learning_rate": 1.6456550123313507e-05, "loss": 0.0453, "step": 4985 }, { "epoch": 0.5405464006938422, "grad_norm": 1.681780219078064, "learning_rate": 1.6455824749746122e-05, "loss": 0.0141, "step": 4986 }, { "epoch": 0.540654813529922, "grad_norm": 0.43622469902038574, "learning_rate": 1.6455099376178735e-05, "loss": 0.0046, "step": 4987 }, { "epoch": 0.5407632263660017, "grad_norm": 6.0874342918396, "learning_rate": 1.6454374002611347e-05, "loss": 0.0449, "step": 4988 }, { "epoch": 0.5408716392020815, "grad_norm": 1.2590794563293457, "learning_rate": 1.645364862904396e-05, "loss": 0.0128, "step": 4989 }, { "epoch": 0.5409800520381614, "grad_norm": 4.520468711853027, "learning_rate": 1.645292325547657e-05, "loss": 0.0091, "step": 4990 }, { "epoch": 0.5410884648742411, "grad_norm": 4.675743103027344, "learning_rate": 1.6452197881909183e-05, "loss": 0.0209, "step": 4991 }, { "epoch": 0.5411968777103209, "grad_norm": 1.516273856163025, "learning_rate": 1.6451472508341795e-05, "loss": 0.0119, "step": 4992 }, { "epoch": 0.5413052905464006, "grad_norm": 5.678681373596191, "learning_rate": 1.645074713477441e-05, "loss": 0.0458, "step": 4993 }, { "epoch": 0.5414137033824805, "grad_norm": 2.6607799530029297, "learning_rate": 1.6450021761207023e-05, "loss": 0.0475, "step": 4994 }, { "epoch": 0.5415221162185603, "grad_norm": 4.542045593261719, "learning_rate": 1.6449296387639635e-05, "loss": 0.0227, "step": 4995 }, { "epoch": 0.54163052905464, "grad_norm": 3.290955066680908, "learning_rate": 1.6448571014072247e-05, "loss": 0.0287, "step": 4996 }, { "epoch": 0.5417389418907199, "grad_norm": 4.154489040374756, "learning_rate": 1.644784564050486e-05, "loss": 0.0221, "step": 4997 }, { "epoch": 0.5418473547267997, "grad_norm": 1.9538911581039429, "learning_rate": 1.6447120266937475e-05, "loss": 0.0268, "step": 4998 }, { "epoch": 0.5419557675628794, "grad_norm": 3.7169225215911865, "learning_rate": 1.6446394893370087e-05, "loss": 0.0565, "step": 4999 }, { "epoch": 0.5420641803989592, "grad_norm": 1.649086356163025, "learning_rate": 1.64456695198027e-05, "loss": 0.0123, "step": 5000 }, { "epoch": 0.5421725932350391, "grad_norm": 1.474852442741394, "learning_rate": 1.6444944146235315e-05, "loss": 0.0376, "step": 5001 }, { "epoch": 0.5422810060711188, "grad_norm": 4.169290542602539, "learning_rate": 1.6444218772667927e-05, "loss": 0.0399, "step": 5002 }, { "epoch": 0.5423894189071986, "grad_norm": 0.6002543568611145, "learning_rate": 1.644349339910054e-05, "loss": 0.0228, "step": 5003 }, { "epoch": 0.5424978317432784, "grad_norm": 2.115562677383423, "learning_rate": 1.644276802553315e-05, "loss": 0.047, "step": 5004 }, { "epoch": 0.5426062445793582, "grad_norm": 7.714510440826416, "learning_rate": 1.6442042651965763e-05, "loss": 0.0954, "step": 5005 }, { "epoch": 0.542714657415438, "grad_norm": 4.245851516723633, "learning_rate": 1.6441317278398376e-05, "loss": 0.0338, "step": 5006 }, { "epoch": 0.5428230702515178, "grad_norm": 1.6979347467422485, "learning_rate": 1.6440591904830988e-05, "loss": 0.0291, "step": 5007 }, { "epoch": 0.5429314830875975, "grad_norm": 0.9989201426506042, "learning_rate": 1.6439866531263603e-05, "loss": 0.0206, "step": 5008 }, { "epoch": 0.5430398959236774, "grad_norm": 2.1386890411376953, "learning_rate": 1.6439141157696215e-05, "loss": 0.0558, "step": 5009 }, { "epoch": 0.5431483087597572, "grad_norm": 2.3956801891326904, "learning_rate": 1.6438415784128827e-05, "loss": 0.0387, "step": 5010 }, { "epoch": 0.5432567215958369, "grad_norm": 4.722105979919434, "learning_rate": 1.643769041056144e-05, "loss": 0.066, "step": 5011 }, { "epoch": 0.5433651344319167, "grad_norm": 6.838680267333984, "learning_rate": 1.6436965036994052e-05, "loss": 0.0415, "step": 5012 }, { "epoch": 0.5434735472679966, "grad_norm": 0.8987189531326294, "learning_rate": 1.6436239663426664e-05, "loss": 0.0053, "step": 5013 }, { "epoch": 0.5435819601040763, "grad_norm": 1.6930482387542725, "learning_rate": 1.643551428985928e-05, "loss": 0.0103, "step": 5014 }, { "epoch": 0.5436903729401561, "grad_norm": 4.163939952850342, "learning_rate": 1.643478891629189e-05, "loss": 0.0604, "step": 5015 }, { "epoch": 0.5437987857762359, "grad_norm": 5.257479667663574, "learning_rate": 1.6434063542724507e-05, "loss": 0.0845, "step": 5016 }, { "epoch": 0.5439071986123157, "grad_norm": 4.6023173332214355, "learning_rate": 1.643333816915712e-05, "loss": 0.0856, "step": 5017 }, { "epoch": 0.5440156114483955, "grad_norm": 3.7534735202789307, "learning_rate": 1.643261279558973e-05, "loss": 0.0336, "step": 5018 }, { "epoch": 0.5441240242844753, "grad_norm": 2.7363481521606445, "learning_rate": 1.6431887422022344e-05, "loss": 0.0367, "step": 5019 }, { "epoch": 0.544232437120555, "grad_norm": 1.228848934173584, "learning_rate": 1.6431162048454956e-05, "loss": 0.0169, "step": 5020 }, { "epoch": 0.5443408499566349, "grad_norm": 1.9520765542984009, "learning_rate": 1.6430436674887568e-05, "loss": 0.0319, "step": 5021 }, { "epoch": 0.5444492627927147, "grad_norm": 1.4851104021072388, "learning_rate": 1.642971130132018e-05, "loss": 0.0293, "step": 5022 }, { "epoch": 0.5445576756287944, "grad_norm": 1.9691957235336304, "learning_rate": 1.6428985927752795e-05, "loss": 0.0562, "step": 5023 }, { "epoch": 0.5446660884648742, "grad_norm": 1.1585752964019775, "learning_rate": 1.6428260554185408e-05, "loss": 0.0317, "step": 5024 }, { "epoch": 0.5447745013009541, "grad_norm": 3.6996071338653564, "learning_rate": 1.642753518061802e-05, "loss": 0.0814, "step": 5025 }, { "epoch": 0.5448829141370338, "grad_norm": 2.615584373474121, "learning_rate": 1.6426809807050632e-05, "loss": 0.0486, "step": 5026 }, { "epoch": 0.5449913269731136, "grad_norm": 1.9732786417007446, "learning_rate": 1.6426084433483244e-05, "loss": 0.0347, "step": 5027 }, { "epoch": 0.5450997398091935, "grad_norm": 1.8739047050476074, "learning_rate": 1.6425359059915856e-05, "loss": 0.0362, "step": 5028 }, { "epoch": 0.5452081526452732, "grad_norm": 2.378230094909668, "learning_rate": 1.6424633686348472e-05, "loss": 0.0351, "step": 5029 }, { "epoch": 0.545316565481353, "grad_norm": 2.365198850631714, "learning_rate": 1.6423908312781084e-05, "loss": 0.0356, "step": 5030 }, { "epoch": 0.5454249783174328, "grad_norm": 1.3120194673538208, "learning_rate": 1.6423182939213696e-05, "loss": 0.0274, "step": 5031 }, { "epoch": 0.5455333911535126, "grad_norm": 2.4779181480407715, "learning_rate": 1.6422457565646308e-05, "loss": 0.0375, "step": 5032 }, { "epoch": 0.5456418039895924, "grad_norm": 9.00960922241211, "learning_rate": 1.6421732192078924e-05, "loss": 0.1122, "step": 5033 }, { "epoch": 0.5457502168256722, "grad_norm": 1.8537606000900269, "learning_rate": 1.6421006818511536e-05, "loss": 0.049, "step": 5034 }, { "epoch": 0.5458586296617519, "grad_norm": 2.310966968536377, "learning_rate": 1.6420281444944148e-05, "loss": 0.0598, "step": 5035 }, { "epoch": 0.5459670424978318, "grad_norm": 1.5326493978500366, "learning_rate": 1.641955607137676e-05, "loss": 0.0221, "step": 5036 }, { "epoch": 0.5460754553339116, "grad_norm": 2.4593396186828613, "learning_rate": 1.6418830697809372e-05, "loss": 0.0566, "step": 5037 }, { "epoch": 0.5461838681699913, "grad_norm": 3.888810396194458, "learning_rate": 1.6418105324241988e-05, "loss": 0.0674, "step": 5038 }, { "epoch": 0.5462922810060711, "grad_norm": 1.3541456460952759, "learning_rate": 1.64173799506746e-05, "loss": 0.0305, "step": 5039 }, { "epoch": 0.546400693842151, "grad_norm": 4.966612339019775, "learning_rate": 1.6416654577107212e-05, "loss": 0.0634, "step": 5040 }, { "epoch": 0.5465091066782307, "grad_norm": 1.5577768087387085, "learning_rate": 1.6415929203539824e-05, "loss": 0.0245, "step": 5041 }, { "epoch": 0.5466175195143105, "grad_norm": 1.5516546964645386, "learning_rate": 1.6415203829972436e-05, "loss": 0.0273, "step": 5042 }, { "epoch": 0.5467259323503902, "grad_norm": 2.090481758117676, "learning_rate": 1.641447845640505e-05, "loss": 0.0279, "step": 5043 }, { "epoch": 0.5468343451864701, "grad_norm": 1.934254765510559, "learning_rate": 1.641375308283766e-05, "loss": 0.0397, "step": 5044 }, { "epoch": 0.5469427580225499, "grad_norm": 2.647592306137085, "learning_rate": 1.6413027709270276e-05, "loss": 0.0633, "step": 5045 }, { "epoch": 0.5470511708586296, "grad_norm": 3.7030084133148193, "learning_rate": 1.641230233570289e-05, "loss": 0.0276, "step": 5046 }, { "epoch": 0.5471595836947094, "grad_norm": 3.9822676181793213, "learning_rate": 1.64115769621355e-05, "loss": 0.0518, "step": 5047 }, { "epoch": 0.5472679965307893, "grad_norm": 1.9565623998641968, "learning_rate": 1.6410851588568113e-05, "loss": 0.0281, "step": 5048 }, { "epoch": 0.547376409366869, "grad_norm": 3.2398529052734375, "learning_rate": 1.6410126215000725e-05, "loss": 0.0733, "step": 5049 }, { "epoch": 0.5474848222029488, "grad_norm": 1.1828352212905884, "learning_rate": 1.640940084143334e-05, "loss": 0.0256, "step": 5050 }, { "epoch": 0.5475932350390286, "grad_norm": 3.5111591815948486, "learning_rate": 1.6408675467865952e-05, "loss": 0.0423, "step": 5051 }, { "epoch": 0.5477016478751084, "grad_norm": 3.443356513977051, "learning_rate": 1.6407950094298565e-05, "loss": 0.0868, "step": 5052 }, { "epoch": 0.5478100607111882, "grad_norm": 2.364382743835449, "learning_rate": 1.640722472073118e-05, "loss": 0.028, "step": 5053 }, { "epoch": 0.547918473547268, "grad_norm": 1.8880114555358887, "learning_rate": 1.6406499347163792e-05, "loss": 0.0496, "step": 5054 }, { "epoch": 0.5480268863833477, "grad_norm": 1.496038556098938, "learning_rate": 1.6405773973596404e-05, "loss": 0.0431, "step": 5055 }, { "epoch": 0.5481352992194276, "grad_norm": 2.4911179542541504, "learning_rate": 1.6405048600029017e-05, "loss": 0.0428, "step": 5056 }, { "epoch": 0.5482437120555074, "grad_norm": 0.9623023867607117, "learning_rate": 1.640432322646163e-05, "loss": 0.0146, "step": 5057 }, { "epoch": 0.5483521248915871, "grad_norm": 3.3011722564697266, "learning_rate": 1.640359785289424e-05, "loss": 0.0577, "step": 5058 }, { "epoch": 0.548460537727667, "grad_norm": 2.6321561336517334, "learning_rate": 1.6402872479326853e-05, "loss": 0.0236, "step": 5059 }, { "epoch": 0.5485689505637468, "grad_norm": 2.2915337085723877, "learning_rate": 1.640214710575947e-05, "loss": 0.0388, "step": 5060 }, { "epoch": 0.5486773633998265, "grad_norm": 2.0519301891326904, "learning_rate": 1.640142173219208e-05, "loss": 0.0455, "step": 5061 }, { "epoch": 0.5487857762359063, "grad_norm": 1.7248119115829468, "learning_rate": 1.6400696358624693e-05, "loss": 0.0154, "step": 5062 }, { "epoch": 0.5488941890719862, "grad_norm": 1.4660342931747437, "learning_rate": 1.6399970985057305e-05, "loss": 0.0271, "step": 5063 }, { "epoch": 0.5490026019080659, "grad_norm": 2.157135486602783, "learning_rate": 1.6399245611489917e-05, "loss": 0.0506, "step": 5064 }, { "epoch": 0.5491110147441457, "grad_norm": 1.482601523399353, "learning_rate": 1.639852023792253e-05, "loss": 0.0441, "step": 5065 }, { "epoch": 0.5492194275802255, "grad_norm": 0.8864631056785583, "learning_rate": 1.6397794864355145e-05, "loss": 0.0137, "step": 5066 }, { "epoch": 0.5493278404163053, "grad_norm": 2.561509847640991, "learning_rate": 1.6397069490787757e-05, "loss": 0.0252, "step": 5067 }, { "epoch": 0.5494362532523851, "grad_norm": 2.2339768409729004, "learning_rate": 1.6396344117220372e-05, "loss": 0.0491, "step": 5068 }, { "epoch": 0.5495446660884649, "grad_norm": 4.658283710479736, "learning_rate": 1.6395618743652985e-05, "loss": 0.0449, "step": 5069 }, { "epoch": 0.5496530789245446, "grad_norm": 0.5829867124557495, "learning_rate": 1.6394893370085597e-05, "loss": 0.0084, "step": 5070 }, { "epoch": 0.5497614917606245, "grad_norm": 1.5368603467941284, "learning_rate": 1.639416799651821e-05, "loss": 0.0325, "step": 5071 }, { "epoch": 0.5498699045967043, "grad_norm": 0.9212126135826111, "learning_rate": 1.639344262295082e-05, "loss": 0.0187, "step": 5072 }, { "epoch": 0.549978317432784, "grad_norm": 1.476570963859558, "learning_rate": 1.6392717249383433e-05, "loss": 0.0273, "step": 5073 }, { "epoch": 0.5500867302688638, "grad_norm": 2.439697027206421, "learning_rate": 1.6391991875816045e-05, "loss": 0.0782, "step": 5074 }, { "epoch": 0.5501951431049437, "grad_norm": 4.279827117919922, "learning_rate": 1.639126650224866e-05, "loss": 0.0726, "step": 5075 }, { "epoch": 0.5503035559410234, "grad_norm": 4.691141605377197, "learning_rate": 1.6390541128681273e-05, "loss": 0.0614, "step": 5076 }, { "epoch": 0.5504119687771032, "grad_norm": 1.5178337097167969, "learning_rate": 1.6389815755113885e-05, "loss": 0.0326, "step": 5077 }, { "epoch": 0.550520381613183, "grad_norm": 3.064760208129883, "learning_rate": 1.6389090381546497e-05, "loss": 0.0287, "step": 5078 }, { "epoch": 0.5506287944492628, "grad_norm": 1.2932902574539185, "learning_rate": 1.638836500797911e-05, "loss": 0.0173, "step": 5079 }, { "epoch": 0.5507372072853426, "grad_norm": 1.238535761833191, "learning_rate": 1.638763963441172e-05, "loss": 0.0177, "step": 5080 }, { "epoch": 0.5508456201214224, "grad_norm": 2.47029447555542, "learning_rate": 1.6386914260844337e-05, "loss": 0.0335, "step": 5081 }, { "epoch": 0.5509540329575021, "grad_norm": 5.988171577453613, "learning_rate": 1.638618888727695e-05, "loss": 0.055, "step": 5082 }, { "epoch": 0.551062445793582, "grad_norm": 1.995348334312439, "learning_rate": 1.638546351370956e-05, "loss": 0.026, "step": 5083 }, { "epoch": 0.5511708586296618, "grad_norm": 3.88633394241333, "learning_rate": 1.6384738140142174e-05, "loss": 0.0346, "step": 5084 }, { "epoch": 0.5512792714657415, "grad_norm": 2.337913751602173, "learning_rate": 1.638401276657479e-05, "loss": 0.0434, "step": 5085 }, { "epoch": 0.5513876843018214, "grad_norm": 1.8451544046401978, "learning_rate": 1.63832873930074e-05, "loss": 0.0698, "step": 5086 }, { "epoch": 0.5514960971379012, "grad_norm": 3.042032241821289, "learning_rate": 1.6382562019440013e-05, "loss": 0.0262, "step": 5087 }, { "epoch": 0.5516045099739809, "grad_norm": 2.625941753387451, "learning_rate": 1.6381836645872625e-05, "loss": 0.0835, "step": 5088 }, { "epoch": 0.5517129228100607, "grad_norm": 2.7155349254608154, "learning_rate": 1.6381111272305238e-05, "loss": 0.0515, "step": 5089 }, { "epoch": 0.5518213356461406, "grad_norm": 5.825006484985352, "learning_rate": 1.6380385898737853e-05, "loss": 0.0424, "step": 5090 }, { "epoch": 0.5519297484822203, "grad_norm": 5.197133541107178, "learning_rate": 1.6379660525170465e-05, "loss": 0.0866, "step": 5091 }, { "epoch": 0.5520381613183001, "grad_norm": 2.4318346977233887, "learning_rate": 1.6378935151603077e-05, "loss": 0.023, "step": 5092 }, { "epoch": 0.5521465741543798, "grad_norm": 1.8167184591293335, "learning_rate": 1.637820977803569e-05, "loss": 0.0431, "step": 5093 }, { "epoch": 0.5522549869904597, "grad_norm": 3.7687294483184814, "learning_rate": 1.6377484404468302e-05, "loss": 0.0422, "step": 5094 }, { "epoch": 0.5523633998265395, "grad_norm": 5.593832969665527, "learning_rate": 1.6376759030900914e-05, "loss": 0.0852, "step": 5095 }, { "epoch": 0.5524718126626192, "grad_norm": 0.9147458672523499, "learning_rate": 1.6376033657333526e-05, "loss": 0.0222, "step": 5096 }, { "epoch": 0.552580225498699, "grad_norm": 2.1688194274902344, "learning_rate": 1.637530828376614e-05, "loss": 0.0566, "step": 5097 }, { "epoch": 0.5526886383347789, "grad_norm": 1.5940958261489868, "learning_rate": 1.6374582910198754e-05, "loss": 0.0493, "step": 5098 }, { "epoch": 0.5527970511708586, "grad_norm": 1.1145833730697632, "learning_rate": 1.6373857536631366e-05, "loss": 0.0122, "step": 5099 }, { "epoch": 0.5529054640069384, "grad_norm": 3.0540356636047363, "learning_rate": 1.6373132163063978e-05, "loss": 0.0345, "step": 5100 }, { "epoch": 0.5530138768430182, "grad_norm": 1.5795916318893433, "learning_rate": 1.637240678949659e-05, "loss": 0.0255, "step": 5101 }, { "epoch": 0.553122289679098, "grad_norm": 1.0402108430862427, "learning_rate": 1.6371681415929206e-05, "loss": 0.0467, "step": 5102 }, { "epoch": 0.5532307025151778, "grad_norm": 1.799759030342102, "learning_rate": 1.6370956042361818e-05, "loss": 0.0669, "step": 5103 }, { "epoch": 0.5533391153512576, "grad_norm": 2.1882150173187256, "learning_rate": 1.637023066879443e-05, "loss": 0.0415, "step": 5104 }, { "epoch": 0.5534475281873373, "grad_norm": 1.306065559387207, "learning_rate": 1.6369505295227045e-05, "loss": 0.064, "step": 5105 }, { "epoch": 0.5535559410234172, "grad_norm": 2.0577406883239746, "learning_rate": 1.6368779921659658e-05, "loss": 0.0418, "step": 5106 }, { "epoch": 0.553664353859497, "grad_norm": 2.479344367980957, "learning_rate": 1.636805454809227e-05, "loss": 0.0701, "step": 5107 }, { "epoch": 0.5537727666955767, "grad_norm": 1.0812008380889893, "learning_rate": 1.6367329174524882e-05, "loss": 0.0122, "step": 5108 }, { "epoch": 0.5538811795316565, "grad_norm": 1.5880165100097656, "learning_rate": 1.6366603800957494e-05, "loss": 0.0247, "step": 5109 }, { "epoch": 0.5539895923677364, "grad_norm": 4.885863780975342, "learning_rate": 1.6365878427390106e-05, "loss": 0.0458, "step": 5110 }, { "epoch": 0.5540980052038161, "grad_norm": 3.01389479637146, "learning_rate": 1.636515305382272e-05, "loss": 0.0344, "step": 5111 }, { "epoch": 0.5542064180398959, "grad_norm": 3.281635284423828, "learning_rate": 1.6364427680255334e-05, "loss": 0.0178, "step": 5112 }, { "epoch": 0.5543148308759757, "grad_norm": 2.2030270099639893, "learning_rate": 1.6363702306687946e-05, "loss": 0.0677, "step": 5113 }, { "epoch": 0.5544232437120555, "grad_norm": 2.2333106994628906, "learning_rate": 1.6362976933120558e-05, "loss": 0.0428, "step": 5114 }, { "epoch": 0.5545316565481353, "grad_norm": 6.370120525360107, "learning_rate": 1.636225155955317e-05, "loss": 0.0438, "step": 5115 }, { "epoch": 0.5546400693842151, "grad_norm": 9.813862800598145, "learning_rate": 1.6361526185985782e-05, "loss": 0.0435, "step": 5116 }, { "epoch": 0.5547484822202949, "grad_norm": 4.842976093292236, "learning_rate": 1.6360800812418395e-05, "loss": 0.0749, "step": 5117 }, { "epoch": 0.5548568950563747, "grad_norm": 0.922466516494751, "learning_rate": 1.636007543885101e-05, "loss": 0.0256, "step": 5118 }, { "epoch": 0.5549653078924545, "grad_norm": 2.189235210418701, "learning_rate": 1.6359350065283622e-05, "loss": 0.0522, "step": 5119 }, { "epoch": 0.5550737207285342, "grad_norm": 3.2916817665100098, "learning_rate": 1.6358624691716238e-05, "loss": 0.0264, "step": 5120 }, { "epoch": 0.5551821335646141, "grad_norm": 2.747498035430908, "learning_rate": 1.635789931814885e-05, "loss": 0.0384, "step": 5121 }, { "epoch": 0.5552905464006939, "grad_norm": 5.076850414276123, "learning_rate": 1.6357173944581462e-05, "loss": 0.0574, "step": 5122 }, { "epoch": 0.5553989592367736, "grad_norm": 3.152383327484131, "learning_rate": 1.6356448571014074e-05, "loss": 0.0335, "step": 5123 }, { "epoch": 0.5555073720728534, "grad_norm": 4.708315849304199, "learning_rate": 1.6355723197446686e-05, "loss": 0.0486, "step": 5124 }, { "epoch": 0.5556157849089333, "grad_norm": 3.6475539207458496, "learning_rate": 1.63549978238793e-05, "loss": 0.032, "step": 5125 }, { "epoch": 0.555724197745013, "grad_norm": 1.1222507953643799, "learning_rate": 1.635427245031191e-05, "loss": 0.0306, "step": 5126 }, { "epoch": 0.5558326105810928, "grad_norm": 0.8363724946975708, "learning_rate": 1.6353547076744526e-05, "loss": 0.0218, "step": 5127 }, { "epoch": 0.5559410234171726, "grad_norm": 16.779850006103516, "learning_rate": 1.635282170317714e-05, "loss": 0.0478, "step": 5128 }, { "epoch": 0.5560494362532524, "grad_norm": 2.6088106632232666, "learning_rate": 1.635209632960975e-05, "loss": 0.0534, "step": 5129 }, { "epoch": 0.5561578490893322, "grad_norm": 1.0886280536651611, "learning_rate": 1.6351370956042363e-05, "loss": 0.0156, "step": 5130 }, { "epoch": 0.556266261925412, "grad_norm": 1.2531154155731201, "learning_rate": 1.6350645582474975e-05, "loss": 0.0279, "step": 5131 }, { "epoch": 0.5563746747614917, "grad_norm": 1.3497624397277832, "learning_rate": 1.6349920208907587e-05, "loss": 0.0152, "step": 5132 }, { "epoch": 0.5564830875975716, "grad_norm": 2.8636538982391357, "learning_rate": 1.6349194835340202e-05, "loss": 0.0791, "step": 5133 }, { "epoch": 0.5565915004336514, "grad_norm": 1.2900222539901733, "learning_rate": 1.6348469461772815e-05, "loss": 0.0126, "step": 5134 }, { "epoch": 0.5566999132697311, "grad_norm": 1.5883156061172485, "learning_rate": 1.6347744088205427e-05, "loss": 0.0164, "step": 5135 }, { "epoch": 0.5568083261058109, "grad_norm": 2.065227508544922, "learning_rate": 1.6347018714638042e-05, "loss": 0.0214, "step": 5136 }, { "epoch": 0.5569167389418908, "grad_norm": 1.8838157653808594, "learning_rate": 1.6346293341070654e-05, "loss": 0.036, "step": 5137 }, { "epoch": 0.5570251517779705, "grad_norm": 2.6652326583862305, "learning_rate": 1.6345567967503267e-05, "loss": 0.032, "step": 5138 }, { "epoch": 0.5571335646140503, "grad_norm": 3.5843443870544434, "learning_rate": 1.634484259393588e-05, "loss": 0.0231, "step": 5139 }, { "epoch": 0.55724197745013, "grad_norm": 1.7931312322616577, "learning_rate": 1.634411722036849e-05, "loss": 0.0157, "step": 5140 }, { "epoch": 0.5573503902862099, "grad_norm": 3.3847365379333496, "learning_rate": 1.6343391846801103e-05, "loss": 0.0353, "step": 5141 }, { "epoch": 0.5574588031222897, "grad_norm": 1.7128028869628906, "learning_rate": 1.634266647323372e-05, "loss": 0.0337, "step": 5142 }, { "epoch": 0.5575672159583694, "grad_norm": 2.132434844970703, "learning_rate": 1.634194109966633e-05, "loss": 0.0472, "step": 5143 }, { "epoch": 0.5576756287944492, "grad_norm": 3.3361213207244873, "learning_rate": 1.6341215726098943e-05, "loss": 0.023, "step": 5144 }, { "epoch": 0.5577840416305291, "grad_norm": 1.4248710870742798, "learning_rate": 1.6340490352531555e-05, "loss": 0.0217, "step": 5145 }, { "epoch": 0.5578924544666088, "grad_norm": 2.9619414806365967, "learning_rate": 1.6339764978964167e-05, "loss": 0.0346, "step": 5146 }, { "epoch": 0.5580008673026886, "grad_norm": 2.524399757385254, "learning_rate": 1.633903960539678e-05, "loss": 0.0341, "step": 5147 }, { "epoch": 0.5581092801387685, "grad_norm": 1.4899123907089233, "learning_rate": 1.633831423182939e-05, "loss": 0.0236, "step": 5148 }, { "epoch": 0.5582176929748482, "grad_norm": 5.704933166503906, "learning_rate": 1.6337588858262007e-05, "loss": 0.0312, "step": 5149 }, { "epoch": 0.558326105810928, "grad_norm": 2.494537591934204, "learning_rate": 1.633686348469462e-05, "loss": 0.0598, "step": 5150 }, { "epoch": 0.5584345186470078, "grad_norm": 2.5465126037597656, "learning_rate": 1.633613811112723e-05, "loss": 0.0475, "step": 5151 }, { "epoch": 0.5585429314830876, "grad_norm": 2.2555618286132812, "learning_rate": 1.6335412737559843e-05, "loss": 0.0197, "step": 5152 }, { "epoch": 0.5586513443191674, "grad_norm": 4.361448287963867, "learning_rate": 1.6334687363992455e-05, "loss": 0.0681, "step": 5153 }, { "epoch": 0.5587597571552472, "grad_norm": 4.099140644073486, "learning_rate": 1.633396199042507e-05, "loss": 0.0616, "step": 5154 }, { "epoch": 0.5588681699913269, "grad_norm": 2.946967840194702, "learning_rate": 1.6333236616857683e-05, "loss": 0.0686, "step": 5155 }, { "epoch": 0.5589765828274068, "grad_norm": 2.7080612182617188, "learning_rate": 1.6332511243290295e-05, "loss": 0.0191, "step": 5156 }, { "epoch": 0.5590849956634866, "grad_norm": 2.3803796768188477, "learning_rate": 1.633178586972291e-05, "loss": 0.0373, "step": 5157 }, { "epoch": 0.5591934084995663, "grad_norm": 2.1620402336120605, "learning_rate": 1.6331060496155523e-05, "loss": 0.0346, "step": 5158 }, { "epoch": 0.5593018213356461, "grad_norm": 3.7266454696655273, "learning_rate": 1.6330335122588135e-05, "loss": 0.1174, "step": 5159 }, { "epoch": 0.559410234171726, "grad_norm": 3.4500622749328613, "learning_rate": 1.6329609749020747e-05, "loss": 0.0518, "step": 5160 }, { "epoch": 0.5595186470078057, "grad_norm": 2.6627349853515625, "learning_rate": 1.632888437545336e-05, "loss": 0.0173, "step": 5161 }, { "epoch": 0.5596270598438855, "grad_norm": 3.3835270404815674, "learning_rate": 1.632815900188597e-05, "loss": 0.0766, "step": 5162 }, { "epoch": 0.5597354726799653, "grad_norm": 0.8684365153312683, "learning_rate": 1.6327433628318584e-05, "loss": 0.0155, "step": 5163 }, { "epoch": 0.5598438855160451, "grad_norm": 2.510361671447754, "learning_rate": 1.63267082547512e-05, "loss": 0.0637, "step": 5164 }, { "epoch": 0.5599522983521249, "grad_norm": 1.9998728036880493, "learning_rate": 1.632598288118381e-05, "loss": 0.0257, "step": 5165 }, { "epoch": 0.5600607111882047, "grad_norm": 3.2523019313812256, "learning_rate": 1.6325257507616423e-05, "loss": 0.0127, "step": 5166 }, { "epoch": 0.5601691240242844, "grad_norm": 3.516195297241211, "learning_rate": 1.6324532134049036e-05, "loss": 0.0478, "step": 5167 }, { "epoch": 0.5602775368603643, "grad_norm": 2.630148410797119, "learning_rate": 1.6323806760481648e-05, "loss": 0.0734, "step": 5168 }, { "epoch": 0.5603859496964441, "grad_norm": 1.4853366613388062, "learning_rate": 1.632308138691426e-05, "loss": 0.0174, "step": 5169 }, { "epoch": 0.5604943625325238, "grad_norm": 1.7472189664840698, "learning_rate": 1.6322356013346875e-05, "loss": 0.0061, "step": 5170 }, { "epoch": 0.5606027753686036, "grad_norm": 1.7642234563827515, "learning_rate": 1.6321630639779488e-05, "loss": 0.0436, "step": 5171 }, { "epoch": 0.5607111882046835, "grad_norm": 1.4607412815093994, "learning_rate": 1.6320905266212103e-05, "loss": 0.0355, "step": 5172 }, { "epoch": 0.5608196010407632, "grad_norm": 1.0085866451263428, "learning_rate": 1.6320179892644715e-05, "loss": 0.0158, "step": 5173 }, { "epoch": 0.560928013876843, "grad_norm": 1.6380478143692017, "learning_rate": 1.6319454519077327e-05, "loss": 0.0447, "step": 5174 }, { "epoch": 0.5610364267129229, "grad_norm": 2.8992388248443604, "learning_rate": 1.631872914550994e-05, "loss": 0.0621, "step": 5175 }, { "epoch": 0.5611448395490026, "grad_norm": 1.2712373733520508, "learning_rate": 1.631800377194255e-05, "loss": 0.0349, "step": 5176 }, { "epoch": 0.5612532523850824, "grad_norm": 3.5094988346099854, "learning_rate": 1.6317278398375164e-05, "loss": 0.0473, "step": 5177 }, { "epoch": 0.5613616652211622, "grad_norm": 1.1588737964630127, "learning_rate": 1.6316553024807776e-05, "loss": 0.0294, "step": 5178 }, { "epoch": 0.561470078057242, "grad_norm": 2.850642204284668, "learning_rate": 1.631582765124039e-05, "loss": 0.0174, "step": 5179 }, { "epoch": 0.5615784908933218, "grad_norm": 4.136220455169678, "learning_rate": 1.6315102277673004e-05, "loss": 0.0336, "step": 5180 }, { "epoch": 0.5616869037294016, "grad_norm": 3.7432262897491455, "learning_rate": 1.6314376904105616e-05, "loss": 0.0592, "step": 5181 }, { "epoch": 0.5617953165654813, "grad_norm": 3.0783472061157227, "learning_rate": 1.6313651530538228e-05, "loss": 0.0416, "step": 5182 }, { "epoch": 0.5619037294015612, "grad_norm": 2.000615119934082, "learning_rate": 1.631292615697084e-05, "loss": 0.0343, "step": 5183 }, { "epoch": 0.562012142237641, "grad_norm": 1.0947701930999756, "learning_rate": 1.6312200783403452e-05, "loss": 0.0118, "step": 5184 }, { "epoch": 0.5621205550737207, "grad_norm": 1.1278817653656006, "learning_rate": 1.6311475409836068e-05, "loss": 0.0198, "step": 5185 }, { "epoch": 0.5622289679098005, "grad_norm": 3.1313672065734863, "learning_rate": 1.631075003626868e-05, "loss": 0.052, "step": 5186 }, { "epoch": 0.5623373807458804, "grad_norm": 2.0551609992980957, "learning_rate": 1.6310024662701292e-05, "loss": 0.0394, "step": 5187 }, { "epoch": 0.5624457935819601, "grad_norm": 2.774322032928467, "learning_rate": 1.6309299289133908e-05, "loss": 0.0477, "step": 5188 }, { "epoch": 0.5625542064180399, "grad_norm": 3.0275983810424805, "learning_rate": 1.630857391556652e-05, "loss": 0.0757, "step": 5189 }, { "epoch": 0.5626626192541196, "grad_norm": 6.593964099884033, "learning_rate": 1.6307848541999132e-05, "loss": 0.0243, "step": 5190 }, { "epoch": 0.5627710320901995, "grad_norm": 4.4167399406433105, "learning_rate": 1.6307123168431744e-05, "loss": 0.0804, "step": 5191 }, { "epoch": 0.5628794449262793, "grad_norm": 1.3544154167175293, "learning_rate": 1.6306397794864356e-05, "loss": 0.0204, "step": 5192 }, { "epoch": 0.562987857762359, "grad_norm": 1.972520112991333, "learning_rate": 1.6305672421296968e-05, "loss": 0.0245, "step": 5193 }, { "epoch": 0.5630962705984388, "grad_norm": 7.871804714202881, "learning_rate": 1.6304947047729584e-05, "loss": 0.1138, "step": 5194 }, { "epoch": 0.5632046834345187, "grad_norm": 1.9192640781402588, "learning_rate": 1.6304221674162196e-05, "loss": 0.0124, "step": 5195 }, { "epoch": 0.5633130962705984, "grad_norm": 1.6188507080078125, "learning_rate": 1.6303496300594808e-05, "loss": 0.0448, "step": 5196 }, { "epoch": 0.5634215091066782, "grad_norm": 1.386410117149353, "learning_rate": 1.630277092702742e-05, "loss": 0.0097, "step": 5197 }, { "epoch": 0.563529921942758, "grad_norm": 4.078160285949707, "learning_rate": 1.6302045553460032e-05, "loss": 0.0508, "step": 5198 }, { "epoch": 0.5636383347788378, "grad_norm": 1.5509657859802246, "learning_rate": 1.6301320179892645e-05, "loss": 0.0468, "step": 5199 }, { "epoch": 0.5637467476149176, "grad_norm": 7.2553205490112305, "learning_rate": 1.6300594806325257e-05, "loss": 0.1127, "step": 5200 }, { "epoch": 0.5638551604509974, "grad_norm": 4.88713264465332, "learning_rate": 1.6299869432757872e-05, "loss": 0.0301, "step": 5201 }, { "epoch": 0.5639635732870771, "grad_norm": 6.160305500030518, "learning_rate": 1.6299144059190484e-05, "loss": 0.0895, "step": 5202 }, { "epoch": 0.564071986123157, "grad_norm": 2.861266613006592, "learning_rate": 1.6298418685623097e-05, "loss": 0.0303, "step": 5203 }, { "epoch": 0.5641803989592368, "grad_norm": 1.7622042894363403, "learning_rate": 1.629769331205571e-05, "loss": 0.0224, "step": 5204 }, { "epoch": 0.5642888117953165, "grad_norm": 3.561222553253174, "learning_rate": 1.6296967938488324e-05, "loss": 0.079, "step": 5205 }, { "epoch": 0.5643972246313964, "grad_norm": 2.086575984954834, "learning_rate": 1.6296242564920936e-05, "loss": 0.0571, "step": 5206 }, { "epoch": 0.5645056374674762, "grad_norm": 2.6832659244537354, "learning_rate": 1.629551719135355e-05, "loss": 0.022, "step": 5207 }, { "epoch": 0.5646140503035559, "grad_norm": 8.93837833404541, "learning_rate": 1.629479181778616e-05, "loss": 0.0276, "step": 5208 }, { "epoch": 0.5647224631396357, "grad_norm": 0.7841426134109497, "learning_rate": 1.6294066444218776e-05, "loss": 0.0098, "step": 5209 }, { "epoch": 0.5648308759757156, "grad_norm": 5.525722503662109, "learning_rate": 1.6293341070651388e-05, "loss": 0.0676, "step": 5210 }, { "epoch": 0.5649392888117953, "grad_norm": 1.347149133682251, "learning_rate": 1.6292615697084e-05, "loss": 0.0265, "step": 5211 }, { "epoch": 0.5650477016478751, "grad_norm": 4.023346424102783, "learning_rate": 1.6291890323516613e-05, "loss": 0.0419, "step": 5212 }, { "epoch": 0.5651561144839549, "grad_norm": 1.3248422145843506, "learning_rate": 1.6291164949949225e-05, "loss": 0.0148, "step": 5213 }, { "epoch": 0.5652645273200347, "grad_norm": 1.6133791208267212, "learning_rate": 1.6290439576381837e-05, "loss": 0.0194, "step": 5214 }, { "epoch": 0.5653729401561145, "grad_norm": 0.8125328421592712, "learning_rate": 1.628971420281445e-05, "loss": 0.0075, "step": 5215 }, { "epoch": 0.5654813529921943, "grad_norm": 0.8461304306983948, "learning_rate": 1.6288988829247065e-05, "loss": 0.0222, "step": 5216 }, { "epoch": 0.565589765828274, "grad_norm": 4.472258567810059, "learning_rate": 1.6288263455679677e-05, "loss": 0.0277, "step": 5217 }, { "epoch": 0.5656981786643539, "grad_norm": 3.024853229522705, "learning_rate": 1.628753808211229e-05, "loss": 0.0413, "step": 5218 }, { "epoch": 0.5658065915004337, "grad_norm": 4.782208442687988, "learning_rate": 1.62868127085449e-05, "loss": 0.02, "step": 5219 }, { "epoch": 0.5659150043365134, "grad_norm": 3.750657320022583, "learning_rate": 1.6286087334977513e-05, "loss": 0.025, "step": 5220 }, { "epoch": 0.5660234171725932, "grad_norm": 1.719747543334961, "learning_rate": 1.6285361961410125e-05, "loss": 0.0262, "step": 5221 }, { "epoch": 0.5661318300086731, "grad_norm": 1.5786374807357788, "learning_rate": 1.628463658784274e-05, "loss": 0.0201, "step": 5222 }, { "epoch": 0.5662402428447528, "grad_norm": 3.3666677474975586, "learning_rate": 1.6283911214275353e-05, "loss": 0.0259, "step": 5223 }, { "epoch": 0.5663486556808326, "grad_norm": 2.941190719604492, "learning_rate": 1.628318584070797e-05, "loss": 0.0636, "step": 5224 }, { "epoch": 0.5664570685169124, "grad_norm": 8.246116638183594, "learning_rate": 1.628246046714058e-05, "loss": 0.0607, "step": 5225 }, { "epoch": 0.5665654813529922, "grad_norm": 6.2097883224487305, "learning_rate": 1.6281735093573193e-05, "loss": 0.0531, "step": 5226 }, { "epoch": 0.566673894189072, "grad_norm": 3.08660626411438, "learning_rate": 1.6281009720005805e-05, "loss": 0.0196, "step": 5227 }, { "epoch": 0.5667823070251518, "grad_norm": 2.032870054244995, "learning_rate": 1.6280284346438417e-05, "loss": 0.0146, "step": 5228 }, { "epoch": 0.5668907198612315, "grad_norm": 6.369679927825928, "learning_rate": 1.627955897287103e-05, "loss": 0.0881, "step": 5229 }, { "epoch": 0.5669991326973114, "grad_norm": 5.539951324462891, "learning_rate": 1.627883359930364e-05, "loss": 0.0627, "step": 5230 }, { "epoch": 0.5671075455333912, "grad_norm": 2.3747100830078125, "learning_rate": 1.6278108225736257e-05, "loss": 0.0155, "step": 5231 }, { "epoch": 0.5672159583694709, "grad_norm": 3.588204860687256, "learning_rate": 1.627738285216887e-05, "loss": 0.0451, "step": 5232 }, { "epoch": 0.5673243712055508, "grad_norm": 0.2973611354827881, "learning_rate": 1.627665747860148e-05, "loss": 0.0144, "step": 5233 }, { "epoch": 0.5674327840416306, "grad_norm": 1.5738309621810913, "learning_rate": 1.6275932105034093e-05, "loss": 0.0455, "step": 5234 }, { "epoch": 0.5675411968777103, "grad_norm": 0.6969113945960999, "learning_rate": 1.6275206731466705e-05, "loss": 0.0108, "step": 5235 }, { "epoch": 0.5676496097137901, "grad_norm": 3.832191228866577, "learning_rate": 1.6274481357899318e-05, "loss": 0.0518, "step": 5236 }, { "epoch": 0.56775802254987, "grad_norm": 8.962583541870117, "learning_rate": 1.6273755984331933e-05, "loss": 0.0181, "step": 5237 }, { "epoch": 0.5678664353859497, "grad_norm": 5.871152400970459, "learning_rate": 1.6273030610764545e-05, "loss": 0.0463, "step": 5238 }, { "epoch": 0.5679748482220295, "grad_norm": 1.2891918420791626, "learning_rate": 1.6272305237197157e-05, "loss": 0.0312, "step": 5239 }, { "epoch": 0.5680832610581092, "grad_norm": 2.4279236793518066, "learning_rate": 1.6271579863629773e-05, "loss": 0.0861, "step": 5240 }, { "epoch": 0.5681916738941891, "grad_norm": 17.050031661987305, "learning_rate": 1.6270854490062385e-05, "loss": 0.0349, "step": 5241 }, { "epoch": 0.5683000867302689, "grad_norm": 3.371441602706909, "learning_rate": 1.6270129116494997e-05, "loss": 0.0475, "step": 5242 }, { "epoch": 0.5684084995663486, "grad_norm": 4.238305568695068, "learning_rate": 1.626940374292761e-05, "loss": 0.0556, "step": 5243 }, { "epoch": 0.5685169124024284, "grad_norm": 2.020484447479248, "learning_rate": 1.626867836936022e-05, "loss": 0.0399, "step": 5244 }, { "epoch": 0.5686253252385083, "grad_norm": 2.885303497314453, "learning_rate": 1.6267952995792834e-05, "loss": 0.0361, "step": 5245 }, { "epoch": 0.568733738074588, "grad_norm": 8.77847671508789, "learning_rate": 1.626722762222545e-05, "loss": 0.04, "step": 5246 }, { "epoch": 0.5688421509106678, "grad_norm": 0.7872613072395325, "learning_rate": 1.626650224865806e-05, "loss": 0.0188, "step": 5247 }, { "epoch": 0.5689505637467476, "grad_norm": 4.114678859710693, "learning_rate": 1.6265776875090673e-05, "loss": 0.0307, "step": 5248 }, { "epoch": 0.5690589765828274, "grad_norm": 1.50063157081604, "learning_rate": 1.6265051501523286e-05, "loss": 0.0292, "step": 5249 }, { "epoch": 0.5691673894189072, "grad_norm": 1.0851088762283325, "learning_rate": 1.6264326127955898e-05, "loss": 0.0216, "step": 5250 }, { "epoch": 0.569275802254987, "grad_norm": 2.4487082958221436, "learning_rate": 1.626360075438851e-05, "loss": 0.0433, "step": 5251 }, { "epoch": 0.5693842150910667, "grad_norm": 2.5894720554351807, "learning_rate": 1.6262875380821125e-05, "loss": 0.0739, "step": 5252 }, { "epoch": 0.5694926279271466, "grad_norm": 4.17870569229126, "learning_rate": 1.6262150007253738e-05, "loss": 0.0334, "step": 5253 }, { "epoch": 0.5696010407632264, "grad_norm": 1.4308496713638306, "learning_rate": 1.626142463368635e-05, "loss": 0.0489, "step": 5254 }, { "epoch": 0.5697094535993061, "grad_norm": 5.636886119842529, "learning_rate": 1.6260699260118962e-05, "loss": 0.0521, "step": 5255 }, { "epoch": 0.5698178664353859, "grad_norm": 1.1274629831314087, "learning_rate": 1.6259973886551574e-05, "loss": 0.0258, "step": 5256 }, { "epoch": 0.5699262792714658, "grad_norm": 1.6903557777404785, "learning_rate": 1.625924851298419e-05, "loss": 0.0288, "step": 5257 }, { "epoch": 0.5700346921075455, "grad_norm": 2.6939618587493896, "learning_rate": 1.62585231394168e-05, "loss": 0.0283, "step": 5258 }, { "epoch": 0.5701431049436253, "grad_norm": 2.3513541221618652, "learning_rate": 1.6257797765849414e-05, "loss": 0.0522, "step": 5259 }, { "epoch": 0.5702515177797051, "grad_norm": 1.1592096090316772, "learning_rate": 1.6257072392282026e-05, "loss": 0.0133, "step": 5260 }, { "epoch": 0.5703599306157849, "grad_norm": 3.7407641410827637, "learning_rate": 1.625634701871464e-05, "loss": 0.0384, "step": 5261 }, { "epoch": 0.5704683434518647, "grad_norm": 1.9039369821548462, "learning_rate": 1.6255621645147254e-05, "loss": 0.0484, "step": 5262 }, { "epoch": 0.5705767562879445, "grad_norm": 2.466641664505005, "learning_rate": 1.6254896271579866e-05, "loss": 0.0457, "step": 5263 }, { "epoch": 0.5706851691240243, "grad_norm": 2.835786819458008, "learning_rate": 1.6254170898012478e-05, "loss": 0.074, "step": 5264 }, { "epoch": 0.5707935819601041, "grad_norm": 4.359723091125488, "learning_rate": 1.625344552444509e-05, "loss": 0.0495, "step": 5265 }, { "epoch": 0.5709019947961839, "grad_norm": 1.9696494340896606, "learning_rate": 1.6252720150877702e-05, "loss": 0.0371, "step": 5266 }, { "epoch": 0.5710104076322636, "grad_norm": 3.870880603790283, "learning_rate": 1.6251994777310314e-05, "loss": 0.0416, "step": 5267 }, { "epoch": 0.5711188204683435, "grad_norm": 1.241081714630127, "learning_rate": 1.625126940374293e-05, "loss": 0.0276, "step": 5268 }, { "epoch": 0.5712272333044233, "grad_norm": 0.4766845405101776, "learning_rate": 1.6250544030175542e-05, "loss": 0.0053, "step": 5269 }, { "epoch": 0.571335646140503, "grad_norm": 1.5499287843704224, "learning_rate": 1.6249818656608154e-05, "loss": 0.0395, "step": 5270 }, { "epoch": 0.5714440589765828, "grad_norm": 1.1335039138793945, "learning_rate": 1.6249093283040766e-05, "loss": 0.0192, "step": 5271 }, { "epoch": 0.5715524718126627, "grad_norm": 1.3905928134918213, "learning_rate": 1.624836790947338e-05, "loss": 0.0394, "step": 5272 }, { "epoch": 0.5716608846487424, "grad_norm": 2.4143881797790527, "learning_rate": 1.624764253590599e-05, "loss": 0.047, "step": 5273 }, { "epoch": 0.5717692974848222, "grad_norm": 0.831830620765686, "learning_rate": 1.6246917162338606e-05, "loss": 0.0376, "step": 5274 }, { "epoch": 0.571877710320902, "grad_norm": 3.0891244411468506, "learning_rate": 1.6246191788771218e-05, "loss": 0.0413, "step": 5275 }, { "epoch": 0.5719861231569818, "grad_norm": 2.9834165573120117, "learning_rate": 1.6245466415203834e-05, "loss": 0.0416, "step": 5276 }, { "epoch": 0.5720945359930616, "grad_norm": 2.269620180130005, "learning_rate": 1.6244741041636446e-05, "loss": 0.0336, "step": 5277 }, { "epoch": 0.5722029488291414, "grad_norm": 3.4848058223724365, "learning_rate": 1.6244015668069058e-05, "loss": 0.0454, "step": 5278 }, { "epoch": 0.5723113616652211, "grad_norm": 0.6068254113197327, "learning_rate": 1.624329029450167e-05, "loss": 0.0067, "step": 5279 }, { "epoch": 0.572419774501301, "grad_norm": 1.9585928916931152, "learning_rate": 1.6242564920934282e-05, "loss": 0.0465, "step": 5280 }, { "epoch": 0.5725281873373808, "grad_norm": 3.1560661792755127, "learning_rate": 1.6241839547366895e-05, "loss": 0.0246, "step": 5281 }, { "epoch": 0.5726366001734605, "grad_norm": 0.7313416600227356, "learning_rate": 1.6241114173799507e-05, "loss": 0.0198, "step": 5282 }, { "epoch": 0.5727450130095403, "grad_norm": 1.4252623319625854, "learning_rate": 1.6240388800232122e-05, "loss": 0.0097, "step": 5283 }, { "epoch": 0.5728534258456202, "grad_norm": 2.2631523609161377, "learning_rate": 1.6239663426664734e-05, "loss": 0.0424, "step": 5284 }, { "epoch": 0.5729618386816999, "grad_norm": 6.172377109527588, "learning_rate": 1.6238938053097346e-05, "loss": 0.0417, "step": 5285 }, { "epoch": 0.5730702515177797, "grad_norm": 3.4408109188079834, "learning_rate": 1.623821267952996e-05, "loss": 0.0381, "step": 5286 }, { "epoch": 0.5731786643538594, "grad_norm": 3.173572063446045, "learning_rate": 1.623748730596257e-05, "loss": 0.0379, "step": 5287 }, { "epoch": 0.5732870771899393, "grad_norm": 2.9747772216796875, "learning_rate": 1.6236761932395183e-05, "loss": 0.0366, "step": 5288 }, { "epoch": 0.5733954900260191, "grad_norm": 5.336831092834473, "learning_rate": 1.62360365588278e-05, "loss": 0.0273, "step": 5289 }, { "epoch": 0.5735039028620988, "grad_norm": 2.244157075881958, "learning_rate": 1.623531118526041e-05, "loss": 0.0323, "step": 5290 }, { "epoch": 0.5736123156981786, "grad_norm": 2.7446205615997314, "learning_rate": 1.6234585811693023e-05, "loss": 0.0594, "step": 5291 }, { "epoch": 0.5737207285342585, "grad_norm": 4.653918743133545, "learning_rate": 1.6233860438125638e-05, "loss": 0.0688, "step": 5292 }, { "epoch": 0.5738291413703382, "grad_norm": 8.411940574645996, "learning_rate": 1.623313506455825e-05, "loss": 0.0427, "step": 5293 }, { "epoch": 0.573937554206418, "grad_norm": 2.5331740379333496, "learning_rate": 1.6232409690990863e-05, "loss": 0.0356, "step": 5294 }, { "epoch": 0.5740459670424979, "grad_norm": 2.532011032104492, "learning_rate": 1.6231684317423475e-05, "loss": 0.0539, "step": 5295 }, { "epoch": 0.5741543798785776, "grad_norm": 7.0749921798706055, "learning_rate": 1.6230958943856087e-05, "loss": 0.0533, "step": 5296 }, { "epoch": 0.5742627927146574, "grad_norm": 4.422047138214111, "learning_rate": 1.62302335702887e-05, "loss": 0.0381, "step": 5297 }, { "epoch": 0.5743712055507372, "grad_norm": 0.43254709243774414, "learning_rate": 1.6229508196721314e-05, "loss": 0.0049, "step": 5298 }, { "epoch": 0.574479618386817, "grad_norm": 2.047464370727539, "learning_rate": 1.6228782823153927e-05, "loss": 0.0241, "step": 5299 }, { "epoch": 0.5745880312228968, "grad_norm": 2.165015459060669, "learning_rate": 1.622805744958654e-05, "loss": 0.047, "step": 5300 }, { "epoch": 0.5746964440589766, "grad_norm": 7.616374969482422, "learning_rate": 1.622733207601915e-05, "loss": 0.0264, "step": 5301 }, { "epoch": 0.5748048568950563, "grad_norm": 3.9353508949279785, "learning_rate": 1.6226606702451763e-05, "loss": 0.071, "step": 5302 }, { "epoch": 0.5749132697311362, "grad_norm": 2.0389177799224854, "learning_rate": 1.6225881328884375e-05, "loss": 0.0514, "step": 5303 }, { "epoch": 0.575021682567216, "grad_norm": 2.5747926235198975, "learning_rate": 1.622515595531699e-05, "loss": 0.0322, "step": 5304 }, { "epoch": 0.5751300954032957, "grad_norm": 1.9971438646316528, "learning_rate": 1.6224430581749603e-05, "loss": 0.0184, "step": 5305 }, { "epoch": 0.5752385082393755, "grad_norm": 1.232966423034668, "learning_rate": 1.6223705208182215e-05, "loss": 0.0217, "step": 5306 }, { "epoch": 0.5753469210754554, "grad_norm": 0.9555973410606384, "learning_rate": 1.6222979834614827e-05, "loss": 0.0489, "step": 5307 }, { "epoch": 0.5754553339115351, "grad_norm": 3.898606538772583, "learning_rate": 1.622225446104744e-05, "loss": 0.0511, "step": 5308 }, { "epoch": 0.5755637467476149, "grad_norm": 4.207022666931152, "learning_rate": 1.6221529087480055e-05, "loss": 0.0748, "step": 5309 }, { "epoch": 0.5756721595836947, "grad_norm": 3.651495933532715, "learning_rate": 1.6220803713912667e-05, "loss": 0.0355, "step": 5310 }, { "epoch": 0.5757805724197745, "grad_norm": 2.146958351135254, "learning_rate": 1.622007834034528e-05, "loss": 0.0391, "step": 5311 }, { "epoch": 0.5758889852558543, "grad_norm": 1.8144091367721558, "learning_rate": 1.621935296677789e-05, "loss": 0.0665, "step": 5312 }, { "epoch": 0.5759973980919341, "grad_norm": 2.207043170928955, "learning_rate": 1.6218627593210507e-05, "loss": 0.0661, "step": 5313 }, { "epoch": 0.5761058109280138, "grad_norm": 5.301562786102295, "learning_rate": 1.621790221964312e-05, "loss": 0.0955, "step": 5314 }, { "epoch": 0.5762142237640937, "grad_norm": 8.263876914978027, "learning_rate": 1.621717684607573e-05, "loss": 0.0581, "step": 5315 }, { "epoch": 0.5763226366001735, "grad_norm": 5.376153469085693, "learning_rate": 1.6216451472508343e-05, "loss": 0.0456, "step": 5316 }, { "epoch": 0.5764310494362532, "grad_norm": 0.612511932849884, "learning_rate": 1.6215726098940955e-05, "loss": 0.0083, "step": 5317 }, { "epoch": 0.576539462272333, "grad_norm": 1.6612224578857422, "learning_rate": 1.6215000725373568e-05, "loss": 0.0448, "step": 5318 }, { "epoch": 0.5766478751084129, "grad_norm": 1.7105956077575684, "learning_rate": 1.621427535180618e-05, "loss": 0.0221, "step": 5319 }, { "epoch": 0.5767562879444926, "grad_norm": 3.5636003017425537, "learning_rate": 1.6213549978238795e-05, "loss": 0.023, "step": 5320 }, { "epoch": 0.5768647007805724, "grad_norm": 2.061534881591797, "learning_rate": 1.6212824604671407e-05, "loss": 0.0336, "step": 5321 }, { "epoch": 0.5769731136166523, "grad_norm": 2.8595781326293945, "learning_rate": 1.621209923110402e-05, "loss": 0.0622, "step": 5322 }, { "epoch": 0.577081526452732, "grad_norm": 2.3550562858581543, "learning_rate": 1.621137385753663e-05, "loss": 0.0508, "step": 5323 }, { "epoch": 0.5771899392888118, "grad_norm": 3.298471450805664, "learning_rate": 1.6210648483969244e-05, "loss": 0.0557, "step": 5324 }, { "epoch": 0.5772983521248916, "grad_norm": 2.125807285308838, "learning_rate": 1.6209923110401856e-05, "loss": 0.0502, "step": 5325 }, { "epoch": 0.5774067649609714, "grad_norm": 2.642793655395508, "learning_rate": 1.620919773683447e-05, "loss": 0.0536, "step": 5326 }, { "epoch": 0.5775151777970512, "grad_norm": 1.143747091293335, "learning_rate": 1.6208472363267084e-05, "loss": 0.0363, "step": 5327 }, { "epoch": 0.577623590633131, "grad_norm": 2.309100866317749, "learning_rate": 1.62077469896997e-05, "loss": 0.0354, "step": 5328 }, { "epoch": 0.5777320034692107, "grad_norm": 2.7064647674560547, "learning_rate": 1.620702161613231e-05, "loss": 0.0159, "step": 5329 }, { "epoch": 0.5778404163052906, "grad_norm": 3.3355274200439453, "learning_rate": 1.6206296242564923e-05, "loss": 0.0328, "step": 5330 }, { "epoch": 0.5779488291413704, "grad_norm": 1.5646569728851318, "learning_rate": 1.6205570868997536e-05, "loss": 0.0129, "step": 5331 }, { "epoch": 0.5780572419774501, "grad_norm": 3.7858777046203613, "learning_rate": 1.6204845495430148e-05, "loss": 0.046, "step": 5332 }, { "epoch": 0.5781656548135299, "grad_norm": 2.8134171962738037, "learning_rate": 1.620412012186276e-05, "loss": 0.0605, "step": 5333 }, { "epoch": 0.5782740676496098, "grad_norm": 1.826219916343689, "learning_rate": 1.6203394748295372e-05, "loss": 0.0457, "step": 5334 }, { "epoch": 0.5783824804856895, "grad_norm": 2.0335628986358643, "learning_rate": 1.6202669374727988e-05, "loss": 0.0227, "step": 5335 }, { "epoch": 0.5784908933217693, "grad_norm": 3.3866591453552246, "learning_rate": 1.62019440011606e-05, "loss": 0.0457, "step": 5336 }, { "epoch": 0.578599306157849, "grad_norm": 1.015750527381897, "learning_rate": 1.6201218627593212e-05, "loss": 0.0189, "step": 5337 }, { "epoch": 0.5787077189939289, "grad_norm": 0.6093160510063171, "learning_rate": 1.6200493254025824e-05, "loss": 0.013, "step": 5338 }, { "epoch": 0.5788161318300087, "grad_norm": 3.0272274017333984, "learning_rate": 1.6199767880458436e-05, "loss": 0.0352, "step": 5339 }, { "epoch": 0.5789245446660884, "grad_norm": 3.8096964359283447, "learning_rate": 1.6199042506891048e-05, "loss": 0.0753, "step": 5340 }, { "epoch": 0.5790329575021682, "grad_norm": 3.536895751953125, "learning_rate": 1.6198317133323664e-05, "loss": 0.036, "step": 5341 }, { "epoch": 0.5791413703382481, "grad_norm": 0.4554336369037628, "learning_rate": 1.6197591759756276e-05, "loss": 0.0097, "step": 5342 }, { "epoch": 0.5792497831743278, "grad_norm": 3.213266611099243, "learning_rate": 1.6196866386188888e-05, "loss": 0.0387, "step": 5343 }, { "epoch": 0.5793581960104076, "grad_norm": 3.92513370513916, "learning_rate": 1.6196141012621504e-05, "loss": 0.0498, "step": 5344 }, { "epoch": 0.5794666088464874, "grad_norm": 0.5262405276298523, "learning_rate": 1.6195415639054116e-05, "loss": 0.0073, "step": 5345 }, { "epoch": 0.5795750216825672, "grad_norm": 2.9632487297058105, "learning_rate": 1.6194690265486728e-05, "loss": 0.0732, "step": 5346 }, { "epoch": 0.579683434518647, "grad_norm": 1.7832986116409302, "learning_rate": 1.619396489191934e-05, "loss": 0.032, "step": 5347 }, { "epoch": 0.5797918473547268, "grad_norm": 4.980021953582764, "learning_rate": 1.6193239518351952e-05, "loss": 0.0682, "step": 5348 }, { "epoch": 0.5799002601908065, "grad_norm": 1.9875489473342896, "learning_rate": 1.6192514144784564e-05, "loss": 0.0492, "step": 5349 }, { "epoch": 0.5800086730268864, "grad_norm": 2.0231781005859375, "learning_rate": 1.619178877121718e-05, "loss": 0.015, "step": 5350 }, { "epoch": 0.5801170858629662, "grad_norm": 4.458078384399414, "learning_rate": 1.6191063397649792e-05, "loss": 0.0197, "step": 5351 }, { "epoch": 0.5802254986990459, "grad_norm": 0.8644037246704102, "learning_rate": 1.6190338024082404e-05, "loss": 0.0166, "step": 5352 }, { "epoch": 0.5803339115351258, "grad_norm": 1.9743454456329346, "learning_rate": 1.6189612650515016e-05, "loss": 0.0328, "step": 5353 }, { "epoch": 0.5804423243712056, "grad_norm": 1.5969696044921875, "learning_rate": 1.618888727694763e-05, "loss": 0.0289, "step": 5354 }, { "epoch": 0.5805507372072853, "grad_norm": 3.1803319454193115, "learning_rate": 1.618816190338024e-05, "loss": 0.0278, "step": 5355 }, { "epoch": 0.5806591500433651, "grad_norm": 3.908111095428467, "learning_rate": 1.6187436529812856e-05, "loss": 0.0444, "step": 5356 }, { "epoch": 0.580767562879445, "grad_norm": 2.451188802719116, "learning_rate": 1.6186711156245468e-05, "loss": 0.0521, "step": 5357 }, { "epoch": 0.5808759757155247, "grad_norm": 2.6975739002227783, "learning_rate": 1.618598578267808e-05, "loss": 0.025, "step": 5358 }, { "epoch": 0.5809843885516045, "grad_norm": 0.959016740322113, "learning_rate": 1.6185260409110693e-05, "loss": 0.0204, "step": 5359 }, { "epoch": 0.5810928013876843, "grad_norm": 3.6397805213928223, "learning_rate": 1.6184535035543305e-05, "loss": 0.0541, "step": 5360 }, { "epoch": 0.5812012142237641, "grad_norm": 2.214245319366455, "learning_rate": 1.618380966197592e-05, "loss": 0.0374, "step": 5361 }, { "epoch": 0.5813096270598439, "grad_norm": 0.8429578542709351, "learning_rate": 1.6183084288408532e-05, "loss": 0.0126, "step": 5362 }, { "epoch": 0.5814180398959237, "grad_norm": 3.4687721729278564, "learning_rate": 1.6182358914841144e-05, "loss": 0.0581, "step": 5363 }, { "epoch": 0.5815264527320034, "grad_norm": 3.095536470413208, "learning_rate": 1.6181633541273757e-05, "loss": 0.0437, "step": 5364 }, { "epoch": 0.5816348655680833, "grad_norm": 0.2787734568119049, "learning_rate": 1.6180908167706372e-05, "loss": 0.0023, "step": 5365 }, { "epoch": 0.5817432784041631, "grad_norm": 2.788640260696411, "learning_rate": 1.6180182794138984e-05, "loss": 0.0541, "step": 5366 }, { "epoch": 0.5818516912402428, "grad_norm": 3.616495132446289, "learning_rate": 1.6179457420571596e-05, "loss": 0.0515, "step": 5367 }, { "epoch": 0.5819601040763226, "grad_norm": 3.501610517501831, "learning_rate": 1.617873204700421e-05, "loss": 0.0882, "step": 5368 }, { "epoch": 0.5820685169124025, "grad_norm": 1.525067687034607, "learning_rate": 1.617800667343682e-05, "loss": 0.0304, "step": 5369 }, { "epoch": 0.5821769297484822, "grad_norm": 1.391255259513855, "learning_rate": 1.6177281299869433e-05, "loss": 0.0424, "step": 5370 }, { "epoch": 0.582285342584562, "grad_norm": 2.1999175548553467, "learning_rate": 1.6176555926302045e-05, "loss": 0.0498, "step": 5371 }, { "epoch": 0.5823937554206418, "grad_norm": 3.394458770751953, "learning_rate": 1.617583055273466e-05, "loss": 0.0375, "step": 5372 }, { "epoch": 0.5825021682567216, "grad_norm": 2.5009288787841797, "learning_rate": 1.6175105179167273e-05, "loss": 0.0322, "step": 5373 }, { "epoch": 0.5826105810928014, "grad_norm": 1.5276302099227905, "learning_rate": 1.6174379805599885e-05, "loss": 0.0217, "step": 5374 }, { "epoch": 0.5827189939288812, "grad_norm": 2.900317907333374, "learning_rate": 1.6173654432032497e-05, "loss": 0.0366, "step": 5375 }, { "epoch": 0.5828274067649609, "grad_norm": 2.0594265460968018, "learning_rate": 1.617292905846511e-05, "loss": 0.0384, "step": 5376 }, { "epoch": 0.5829358196010408, "grad_norm": 4.08247709274292, "learning_rate": 1.617220368489772e-05, "loss": 0.0406, "step": 5377 }, { "epoch": 0.5830442324371206, "grad_norm": 1.4045206308364868, "learning_rate": 1.6171478311330337e-05, "loss": 0.019, "step": 5378 }, { "epoch": 0.5831526452732003, "grad_norm": 4.507578372955322, "learning_rate": 1.617075293776295e-05, "loss": 0.0416, "step": 5379 }, { "epoch": 0.5832610581092801, "grad_norm": 1.511281967163086, "learning_rate": 1.6170027564195564e-05, "loss": 0.037, "step": 5380 }, { "epoch": 0.58336947094536, "grad_norm": 2.103414535522461, "learning_rate": 1.6169302190628177e-05, "loss": 0.0498, "step": 5381 }, { "epoch": 0.5834778837814397, "grad_norm": 3.0185325145721436, "learning_rate": 1.616857681706079e-05, "loss": 0.0441, "step": 5382 }, { "epoch": 0.5835862966175195, "grad_norm": 2.5058822631835938, "learning_rate": 1.61678514434934e-05, "loss": 0.0318, "step": 5383 }, { "epoch": 0.5836947094535994, "grad_norm": 5.201665878295898, "learning_rate": 1.6167126069926013e-05, "loss": 0.0275, "step": 5384 }, { "epoch": 0.5838031222896791, "grad_norm": 4.888380527496338, "learning_rate": 1.6166400696358625e-05, "loss": 0.0623, "step": 5385 }, { "epoch": 0.5839115351257589, "grad_norm": 3.6907119750976562, "learning_rate": 1.6165675322791237e-05, "loss": 0.058, "step": 5386 }, { "epoch": 0.5840199479618386, "grad_norm": 2.338972806930542, "learning_rate": 1.6164949949223853e-05, "loss": 0.0234, "step": 5387 }, { "epoch": 0.5841283607979185, "grad_norm": 1.7600849866867065, "learning_rate": 1.6164224575656465e-05, "loss": 0.0134, "step": 5388 }, { "epoch": 0.5842367736339983, "grad_norm": 2.058260917663574, "learning_rate": 1.6163499202089077e-05, "loss": 0.0514, "step": 5389 }, { "epoch": 0.584345186470078, "grad_norm": 1.7792413234710693, "learning_rate": 1.616277382852169e-05, "loss": 0.0918, "step": 5390 }, { "epoch": 0.5844535993061578, "grad_norm": 4.73380184173584, "learning_rate": 1.61620484549543e-05, "loss": 0.122, "step": 5391 }, { "epoch": 0.5845620121422377, "grad_norm": 1.2720390558242798, "learning_rate": 1.6161323081386914e-05, "loss": 0.037, "step": 5392 }, { "epoch": 0.5846704249783174, "grad_norm": 2.0727100372314453, "learning_rate": 1.616059770781953e-05, "loss": 0.0524, "step": 5393 }, { "epoch": 0.5847788378143972, "grad_norm": 2.363787889480591, "learning_rate": 1.615987233425214e-05, "loss": 0.0284, "step": 5394 }, { "epoch": 0.584887250650477, "grad_norm": 1.7503042221069336, "learning_rate": 1.6159146960684753e-05, "loss": 0.034, "step": 5395 }, { "epoch": 0.5849956634865568, "grad_norm": 1.7694637775421143, "learning_rate": 1.615842158711737e-05, "loss": 0.0223, "step": 5396 }, { "epoch": 0.5851040763226366, "grad_norm": 1.7693679332733154, "learning_rate": 1.615769621354998e-05, "loss": 0.0489, "step": 5397 }, { "epoch": 0.5852124891587164, "grad_norm": 5.5222344398498535, "learning_rate": 1.6156970839982593e-05, "loss": 0.0348, "step": 5398 }, { "epoch": 0.5853209019947961, "grad_norm": 3.0230793952941895, "learning_rate": 1.6156245466415205e-05, "loss": 0.0575, "step": 5399 }, { "epoch": 0.585429314830876, "grad_norm": 0.8242191672325134, "learning_rate": 1.6155520092847817e-05, "loss": 0.0268, "step": 5400 }, { "epoch": 0.5855377276669558, "grad_norm": 1.1107993125915527, "learning_rate": 1.615479471928043e-05, "loss": 0.0192, "step": 5401 }, { "epoch": 0.5856461405030355, "grad_norm": 1.2558451890945435, "learning_rate": 1.6154069345713045e-05, "loss": 0.0312, "step": 5402 }, { "epoch": 0.5857545533391153, "grad_norm": 1.9819902181625366, "learning_rate": 1.6153343972145657e-05, "loss": 0.0482, "step": 5403 }, { "epoch": 0.5858629661751952, "grad_norm": 2.5539486408233643, "learning_rate": 1.615261859857827e-05, "loss": 0.0557, "step": 5404 }, { "epoch": 0.5859713790112749, "grad_norm": 1.515978455543518, "learning_rate": 1.615189322501088e-05, "loss": 0.0402, "step": 5405 }, { "epoch": 0.5860797918473547, "grad_norm": 1.60758376121521, "learning_rate": 1.6151167851443494e-05, "loss": 0.0351, "step": 5406 }, { "epoch": 0.5861882046834345, "grad_norm": 1.5747932195663452, "learning_rate": 1.6150442477876106e-05, "loss": 0.0497, "step": 5407 }, { "epoch": 0.5862966175195143, "grad_norm": 2.170734167098999, "learning_rate": 1.614971710430872e-05, "loss": 0.0454, "step": 5408 }, { "epoch": 0.5864050303555941, "grad_norm": 7.209428787231445, "learning_rate": 1.6148991730741334e-05, "loss": 0.0363, "step": 5409 }, { "epoch": 0.5865134431916739, "grad_norm": 0.7277272343635559, "learning_rate": 1.6148266357173946e-05, "loss": 0.0228, "step": 5410 }, { "epoch": 0.5866218560277537, "grad_norm": 0.8587915897369385, "learning_rate": 1.6147540983606558e-05, "loss": 0.0154, "step": 5411 }, { "epoch": 0.5867302688638335, "grad_norm": 2.3971188068389893, "learning_rate": 1.614681561003917e-05, "loss": 0.0356, "step": 5412 }, { "epoch": 0.5868386816999133, "grad_norm": 2.0052411556243896, "learning_rate": 1.6146090236471786e-05, "loss": 0.0276, "step": 5413 }, { "epoch": 0.586947094535993, "grad_norm": 3.0747299194335938, "learning_rate": 1.6145364862904398e-05, "loss": 0.0415, "step": 5414 }, { "epoch": 0.5870555073720729, "grad_norm": 3.315342664718628, "learning_rate": 1.614463948933701e-05, "loss": 0.0384, "step": 5415 }, { "epoch": 0.5871639202081527, "grad_norm": 1.594789743423462, "learning_rate": 1.6143914115769622e-05, "loss": 0.0279, "step": 5416 }, { "epoch": 0.5872723330442324, "grad_norm": 0.7832840085029602, "learning_rate": 1.6143188742202237e-05, "loss": 0.0112, "step": 5417 }, { "epoch": 0.5873807458803122, "grad_norm": 4.921454906463623, "learning_rate": 1.614246336863485e-05, "loss": 0.0728, "step": 5418 }, { "epoch": 0.5874891587163921, "grad_norm": 8.221423149108887, "learning_rate": 1.6141737995067462e-05, "loss": 0.0351, "step": 5419 }, { "epoch": 0.5875975715524718, "grad_norm": 5.092695236206055, "learning_rate": 1.6141012621500074e-05, "loss": 0.0319, "step": 5420 }, { "epoch": 0.5877059843885516, "grad_norm": 3.0659234523773193, "learning_rate": 1.6140287247932686e-05, "loss": 0.0302, "step": 5421 }, { "epoch": 0.5878143972246314, "grad_norm": 0.8321399688720703, "learning_rate": 1.6139561874365298e-05, "loss": 0.013, "step": 5422 }, { "epoch": 0.5879228100607112, "grad_norm": 7.353270053863525, "learning_rate": 1.613883650079791e-05, "loss": 0.0887, "step": 5423 }, { "epoch": 0.588031222896791, "grad_norm": 1.752909779548645, "learning_rate": 1.6138111127230526e-05, "loss": 0.0357, "step": 5424 }, { "epoch": 0.5881396357328708, "grad_norm": 2.607586622238159, "learning_rate": 1.6137385753663138e-05, "loss": 0.0289, "step": 5425 }, { "epoch": 0.5882480485689505, "grad_norm": 4.369439125061035, "learning_rate": 1.613666038009575e-05, "loss": 0.0515, "step": 5426 }, { "epoch": 0.5883564614050304, "grad_norm": 2.104149580001831, "learning_rate": 1.6135935006528362e-05, "loss": 0.0199, "step": 5427 }, { "epoch": 0.5884648742411102, "grad_norm": 4.834447383880615, "learning_rate": 1.6135209632960974e-05, "loss": 0.0656, "step": 5428 }, { "epoch": 0.5885732870771899, "grad_norm": 4.4202351570129395, "learning_rate": 1.6134484259393587e-05, "loss": 0.074, "step": 5429 }, { "epoch": 0.5886816999132697, "grad_norm": 1.7335216999053955, "learning_rate": 1.6133758885826202e-05, "loss": 0.0493, "step": 5430 }, { "epoch": 0.5887901127493496, "grad_norm": 3.5467658042907715, "learning_rate": 1.6133033512258814e-05, "loss": 0.0398, "step": 5431 }, { "epoch": 0.5888985255854293, "grad_norm": 2.9135518074035645, "learning_rate": 1.613230813869143e-05, "loss": 0.0418, "step": 5432 }, { "epoch": 0.5890069384215091, "grad_norm": 1.7267022132873535, "learning_rate": 1.6131582765124042e-05, "loss": 0.0411, "step": 5433 }, { "epoch": 0.5891153512575888, "grad_norm": 0.8497400879859924, "learning_rate": 1.6130857391556654e-05, "loss": 0.0222, "step": 5434 }, { "epoch": 0.5892237640936687, "grad_norm": 1.8077391386032104, "learning_rate": 1.6130132017989266e-05, "loss": 0.0234, "step": 5435 }, { "epoch": 0.5893321769297485, "grad_norm": 4.273362636566162, "learning_rate": 1.612940664442188e-05, "loss": 0.0345, "step": 5436 }, { "epoch": 0.5894405897658282, "grad_norm": 5.0852203369140625, "learning_rate": 1.612868127085449e-05, "loss": 0.0515, "step": 5437 }, { "epoch": 0.589549002601908, "grad_norm": 1.4228085279464722, "learning_rate": 1.6127955897287103e-05, "loss": 0.0074, "step": 5438 }, { "epoch": 0.5896574154379879, "grad_norm": 2.9200735092163086, "learning_rate": 1.6127230523719718e-05, "loss": 0.0978, "step": 5439 }, { "epoch": 0.5897658282740676, "grad_norm": 2.595789670944214, "learning_rate": 1.612650515015233e-05, "loss": 0.0513, "step": 5440 }, { "epoch": 0.5898742411101474, "grad_norm": 3.2228288650512695, "learning_rate": 1.6125779776584942e-05, "loss": 0.113, "step": 5441 }, { "epoch": 0.5899826539462273, "grad_norm": 3.9411368370056152, "learning_rate": 1.6125054403017555e-05, "loss": 0.0522, "step": 5442 }, { "epoch": 0.590091066782307, "grad_norm": 2.3517966270446777, "learning_rate": 1.6124329029450167e-05, "loss": 0.0435, "step": 5443 }, { "epoch": 0.5901994796183868, "grad_norm": 2.415329694747925, "learning_rate": 1.612360365588278e-05, "loss": 0.0639, "step": 5444 }, { "epoch": 0.5903078924544666, "grad_norm": 1.6510536670684814, "learning_rate": 1.6122878282315394e-05, "loss": 0.0346, "step": 5445 }, { "epoch": 0.5904163052905465, "grad_norm": 4.368962287902832, "learning_rate": 1.6122152908748007e-05, "loss": 0.071, "step": 5446 }, { "epoch": 0.5905247181266262, "grad_norm": 2.540334463119507, "learning_rate": 1.6121427535180622e-05, "loss": 0.0275, "step": 5447 }, { "epoch": 0.590633130962706, "grad_norm": 6.471439838409424, "learning_rate": 1.6120702161613234e-05, "loss": 0.0611, "step": 5448 }, { "epoch": 0.5907415437987857, "grad_norm": 3.730710506439209, "learning_rate": 1.6119976788045846e-05, "loss": 0.0905, "step": 5449 }, { "epoch": 0.5908499566348656, "grad_norm": 1.9550849199295044, "learning_rate": 1.611925141447846e-05, "loss": 0.0359, "step": 5450 }, { "epoch": 0.5909583694709454, "grad_norm": 2.4698753356933594, "learning_rate": 1.611852604091107e-05, "loss": 0.0453, "step": 5451 }, { "epoch": 0.5910667823070251, "grad_norm": 3.5605905055999756, "learning_rate": 1.6117800667343683e-05, "loss": 0.038, "step": 5452 }, { "epoch": 0.5911751951431049, "grad_norm": 4.645929336547852, "learning_rate": 1.6117075293776295e-05, "loss": 0.0287, "step": 5453 }, { "epoch": 0.5912836079791848, "grad_norm": 1.0472229719161987, "learning_rate": 1.611634992020891e-05, "loss": 0.0367, "step": 5454 }, { "epoch": 0.5913920208152645, "grad_norm": 2.6160247325897217, "learning_rate": 1.6115624546641523e-05, "loss": 0.0586, "step": 5455 }, { "epoch": 0.5915004336513443, "grad_norm": 1.056609034538269, "learning_rate": 1.6114899173074135e-05, "loss": 0.0342, "step": 5456 }, { "epoch": 0.5916088464874241, "grad_norm": 1.3254613876342773, "learning_rate": 1.6114173799506747e-05, "loss": 0.0448, "step": 5457 }, { "epoch": 0.5917172593235039, "grad_norm": 0.6056516766548157, "learning_rate": 1.611344842593936e-05, "loss": 0.0149, "step": 5458 }, { "epoch": 0.5918256721595837, "grad_norm": 1.2115559577941895, "learning_rate": 1.611272305237197e-05, "loss": 0.0429, "step": 5459 }, { "epoch": 0.5919340849956635, "grad_norm": 0.9110471606254578, "learning_rate": 1.6111997678804587e-05, "loss": 0.0257, "step": 5460 }, { "epoch": 0.5920424978317432, "grad_norm": 4.139222621917725, "learning_rate": 1.61112723052372e-05, "loss": 0.0755, "step": 5461 }, { "epoch": 0.5921509106678231, "grad_norm": 0.9901721477508545, "learning_rate": 1.611054693166981e-05, "loss": 0.0258, "step": 5462 }, { "epoch": 0.5922593235039029, "grad_norm": 2.250145196914673, "learning_rate": 1.6109821558102423e-05, "loss": 0.0369, "step": 5463 }, { "epoch": 0.5923677363399826, "grad_norm": 2.1919357776641846, "learning_rate": 1.6109096184535035e-05, "loss": 0.027, "step": 5464 }, { "epoch": 0.5924761491760624, "grad_norm": 1.6624618768692017, "learning_rate": 1.610837081096765e-05, "loss": 0.0315, "step": 5465 }, { "epoch": 0.5925845620121423, "grad_norm": 2.8058948516845703, "learning_rate": 1.6107645437400263e-05, "loss": 0.0356, "step": 5466 }, { "epoch": 0.592692974848222, "grad_norm": 3.128364086151123, "learning_rate": 1.6106920063832875e-05, "loss": 0.0586, "step": 5467 }, { "epoch": 0.5928013876843018, "grad_norm": 2.747513771057129, "learning_rate": 1.6106194690265487e-05, "loss": 0.0456, "step": 5468 }, { "epoch": 0.5929098005203816, "grad_norm": 0.512365460395813, "learning_rate": 1.6105469316698103e-05, "loss": 0.0051, "step": 5469 }, { "epoch": 0.5930182133564614, "grad_norm": 2.5714268684387207, "learning_rate": 1.6104743943130715e-05, "loss": 0.0337, "step": 5470 }, { "epoch": 0.5931266261925412, "grad_norm": 2.0613629817962646, "learning_rate": 1.6104018569563327e-05, "loss": 0.0517, "step": 5471 }, { "epoch": 0.593235039028621, "grad_norm": 2.202636241912842, "learning_rate": 1.610329319599594e-05, "loss": 0.0515, "step": 5472 }, { "epoch": 0.5933434518647008, "grad_norm": 2.943142890930176, "learning_rate": 1.610256782242855e-05, "loss": 0.0342, "step": 5473 }, { "epoch": 0.5934518647007806, "grad_norm": 1.544914960861206, "learning_rate": 1.6101842448861164e-05, "loss": 0.0147, "step": 5474 }, { "epoch": 0.5935602775368604, "grad_norm": 0.8624158501625061, "learning_rate": 1.6101117075293776e-05, "loss": 0.0169, "step": 5475 }, { "epoch": 0.5936686903729401, "grad_norm": 2.3122401237487793, "learning_rate": 1.610039170172639e-05, "loss": 0.0227, "step": 5476 }, { "epoch": 0.59377710320902, "grad_norm": 1.972068428993225, "learning_rate": 1.6099666328159003e-05, "loss": 0.0134, "step": 5477 }, { "epoch": 0.5938855160450998, "grad_norm": 2.371971368789673, "learning_rate": 1.6098940954591616e-05, "loss": 0.0524, "step": 5478 }, { "epoch": 0.5939939288811795, "grad_norm": 6.95225191116333, "learning_rate": 1.6098215581024228e-05, "loss": 0.0651, "step": 5479 }, { "epoch": 0.5941023417172593, "grad_norm": 0.616240382194519, "learning_rate": 1.609749020745684e-05, "loss": 0.0109, "step": 5480 }, { "epoch": 0.5942107545533392, "grad_norm": 2.1773629188537598, "learning_rate": 1.6096764833889452e-05, "loss": 0.0389, "step": 5481 }, { "epoch": 0.5943191673894189, "grad_norm": 2.932011365890503, "learning_rate": 1.6096039460322067e-05, "loss": 0.0344, "step": 5482 }, { "epoch": 0.5944275802254987, "grad_norm": 0.593639612197876, "learning_rate": 1.609531408675468e-05, "loss": 0.0023, "step": 5483 }, { "epoch": 0.5945359930615784, "grad_norm": 1.030544638633728, "learning_rate": 1.6094588713187295e-05, "loss": 0.0243, "step": 5484 }, { "epoch": 0.5946444058976583, "grad_norm": 2.004387378692627, "learning_rate": 1.6093863339619907e-05, "loss": 0.0565, "step": 5485 }, { "epoch": 0.5947528187337381, "grad_norm": 2.820188283920288, "learning_rate": 1.609313796605252e-05, "loss": 0.0357, "step": 5486 }, { "epoch": 0.5948612315698178, "grad_norm": 1.0209336280822754, "learning_rate": 1.609241259248513e-05, "loss": 0.029, "step": 5487 }, { "epoch": 0.5949696444058976, "grad_norm": 4.733990669250488, "learning_rate": 1.6091687218917744e-05, "loss": 0.0338, "step": 5488 }, { "epoch": 0.5950780572419775, "grad_norm": 4.606927871704102, "learning_rate": 1.6090961845350356e-05, "loss": 0.0553, "step": 5489 }, { "epoch": 0.5951864700780573, "grad_norm": 3.945552110671997, "learning_rate": 1.6090236471782968e-05, "loss": 0.0313, "step": 5490 }, { "epoch": 0.595294882914137, "grad_norm": 6.553109645843506, "learning_rate": 1.6089511098215584e-05, "loss": 0.0482, "step": 5491 }, { "epoch": 0.5954032957502168, "grad_norm": 0.40047627687454224, "learning_rate": 1.6088785724648196e-05, "loss": 0.0314, "step": 5492 }, { "epoch": 0.5955117085862967, "grad_norm": 0.9066625237464905, "learning_rate": 1.6088060351080808e-05, "loss": 0.0294, "step": 5493 }, { "epoch": 0.5956201214223764, "grad_norm": 0.4567464590072632, "learning_rate": 1.608733497751342e-05, "loss": 0.0101, "step": 5494 }, { "epoch": 0.5957285342584562, "grad_norm": 2.9073894023895264, "learning_rate": 1.6086609603946032e-05, "loss": 0.0531, "step": 5495 }, { "epoch": 0.5958369470945359, "grad_norm": 0.9525260925292969, "learning_rate": 1.6085884230378644e-05, "loss": 0.0086, "step": 5496 }, { "epoch": 0.5959453599306158, "grad_norm": 1.4149303436279297, "learning_rate": 1.608515885681126e-05, "loss": 0.0221, "step": 5497 }, { "epoch": 0.5960537727666956, "grad_norm": 4.023262023925781, "learning_rate": 1.6084433483243872e-05, "loss": 0.0349, "step": 5498 }, { "epoch": 0.5961621856027753, "grad_norm": 2.3103952407836914, "learning_rate": 1.6083708109676487e-05, "loss": 0.0357, "step": 5499 }, { "epoch": 0.5962705984388552, "grad_norm": 3.0353987216949463, "learning_rate": 1.60829827361091e-05, "loss": 0.0372, "step": 5500 }, { "epoch": 0.596379011274935, "grad_norm": 1.0692740678787231, "learning_rate": 1.6082257362541712e-05, "loss": 0.0159, "step": 5501 }, { "epoch": 0.5964874241110147, "grad_norm": 1.7502177953720093, "learning_rate": 1.6081531988974324e-05, "loss": 0.0584, "step": 5502 }, { "epoch": 0.5965958369470945, "grad_norm": 4.212405204772949, "learning_rate": 1.6080806615406936e-05, "loss": 0.0078, "step": 5503 }, { "epoch": 0.5967042497831744, "grad_norm": 4.063678741455078, "learning_rate": 1.6080081241839548e-05, "loss": 0.0288, "step": 5504 }, { "epoch": 0.5968126626192541, "grad_norm": 7.682085990905762, "learning_rate": 1.607935586827216e-05, "loss": 0.1132, "step": 5505 }, { "epoch": 0.5969210754553339, "grad_norm": 0.20520612597465515, "learning_rate": 1.6078630494704776e-05, "loss": 0.0171, "step": 5506 }, { "epoch": 0.5970294882914137, "grad_norm": 1.3088618516921997, "learning_rate": 1.6077905121137388e-05, "loss": 0.0239, "step": 5507 }, { "epoch": 0.5971379011274935, "grad_norm": 3.815673828125, "learning_rate": 1.607717974757e-05, "loss": 0.041, "step": 5508 }, { "epoch": 0.5972463139635733, "grad_norm": 8.884778022766113, "learning_rate": 1.6076454374002612e-05, "loss": 0.0522, "step": 5509 }, { "epoch": 0.5973547267996531, "grad_norm": 6.311159610748291, "learning_rate": 1.6075729000435224e-05, "loss": 0.0741, "step": 5510 }, { "epoch": 0.5974631396357328, "grad_norm": 1.3165019750595093, "learning_rate": 1.6075003626867837e-05, "loss": 0.0146, "step": 5511 }, { "epoch": 0.5975715524718127, "grad_norm": 2.7032628059387207, "learning_rate": 1.6074278253300452e-05, "loss": 0.0145, "step": 5512 }, { "epoch": 0.5976799653078925, "grad_norm": 0.8783814907073975, "learning_rate": 1.6073552879733064e-05, "loss": 0.0103, "step": 5513 }, { "epoch": 0.5977883781439722, "grad_norm": 3.7264628410339355, "learning_rate": 1.6072827506165676e-05, "loss": 0.0464, "step": 5514 }, { "epoch": 0.597896790980052, "grad_norm": 0.934954047203064, "learning_rate": 1.607210213259829e-05, "loss": 0.0161, "step": 5515 }, { "epoch": 0.5980052038161319, "grad_norm": 0.45617109537124634, "learning_rate": 1.6071376759030904e-05, "loss": 0.0216, "step": 5516 }, { "epoch": 0.5981136166522116, "grad_norm": 8.282215118408203, "learning_rate": 1.6070651385463516e-05, "loss": 0.0512, "step": 5517 }, { "epoch": 0.5982220294882914, "grad_norm": 2.7992935180664062, "learning_rate": 1.606992601189613e-05, "loss": 0.0205, "step": 5518 }, { "epoch": 0.5983304423243712, "grad_norm": 1.8058701753616333, "learning_rate": 1.606920063832874e-05, "loss": 0.0613, "step": 5519 }, { "epoch": 0.598438855160451, "grad_norm": 1.7081698179244995, "learning_rate": 1.6068475264761353e-05, "loss": 0.0252, "step": 5520 }, { "epoch": 0.5985472679965308, "grad_norm": 1.5726584196090698, "learning_rate": 1.6067749891193968e-05, "loss": 0.0241, "step": 5521 }, { "epoch": 0.5986556808326106, "grad_norm": 2.8387773036956787, "learning_rate": 1.606702451762658e-05, "loss": 0.0622, "step": 5522 }, { "epoch": 0.5987640936686903, "grad_norm": 2.866783618927002, "learning_rate": 1.6066299144059192e-05, "loss": 0.0877, "step": 5523 }, { "epoch": 0.5988725065047702, "grad_norm": 6.058095455169678, "learning_rate": 1.6065573770491805e-05, "loss": 0.0619, "step": 5524 }, { "epoch": 0.59898091934085, "grad_norm": 1.779193639755249, "learning_rate": 1.6064848396924417e-05, "loss": 0.0378, "step": 5525 }, { "epoch": 0.5990893321769297, "grad_norm": 1.3041672706604004, "learning_rate": 1.606412302335703e-05, "loss": 0.008, "step": 5526 }, { "epoch": 0.5991977450130095, "grad_norm": 5.25272274017334, "learning_rate": 1.606339764978964e-05, "loss": 0.0528, "step": 5527 }, { "epoch": 0.5993061578490894, "grad_norm": 1.4147015810012817, "learning_rate": 1.6062672276222257e-05, "loss": 0.019, "step": 5528 }, { "epoch": 0.5994145706851691, "grad_norm": 2.3987443447113037, "learning_rate": 1.606194690265487e-05, "loss": 0.0309, "step": 5529 }, { "epoch": 0.5995229835212489, "grad_norm": 4.695638179779053, "learning_rate": 1.606122152908748e-05, "loss": 0.0913, "step": 5530 }, { "epoch": 0.5996313963573288, "grad_norm": 0.5192503333091736, "learning_rate": 1.6060496155520093e-05, "loss": 0.0044, "step": 5531 }, { "epoch": 0.5997398091934085, "grad_norm": 1.5575608015060425, "learning_rate": 1.6059770781952705e-05, "loss": 0.0393, "step": 5532 }, { "epoch": 0.5998482220294883, "grad_norm": 0.7568934559822083, "learning_rate": 1.6059045408385317e-05, "loss": 0.0149, "step": 5533 }, { "epoch": 0.599956634865568, "grad_norm": 2.8095293045043945, "learning_rate": 1.6058320034817933e-05, "loss": 0.0417, "step": 5534 }, { "epoch": 0.6000650477016479, "grad_norm": 2.9142093658447266, "learning_rate": 1.6057594661250545e-05, "loss": 0.0682, "step": 5535 }, { "epoch": 0.6001734605377277, "grad_norm": 1.803117275238037, "learning_rate": 1.605686928768316e-05, "loss": 0.0337, "step": 5536 }, { "epoch": 0.6002818733738075, "grad_norm": 1.5564179420471191, "learning_rate": 1.6056143914115773e-05, "loss": 0.0145, "step": 5537 }, { "epoch": 0.6003902862098872, "grad_norm": 4.689320087432861, "learning_rate": 1.6055418540548385e-05, "loss": 0.061, "step": 5538 }, { "epoch": 0.6004986990459671, "grad_norm": 2.092306613922119, "learning_rate": 1.6054693166980997e-05, "loss": 0.0155, "step": 5539 }, { "epoch": 0.6006071118820469, "grad_norm": 3.6149699687957764, "learning_rate": 1.605396779341361e-05, "loss": 0.0275, "step": 5540 }, { "epoch": 0.6007155247181266, "grad_norm": 2.877488136291504, "learning_rate": 1.605324241984622e-05, "loss": 0.0158, "step": 5541 }, { "epoch": 0.6008239375542064, "grad_norm": 3.4269795417785645, "learning_rate": 1.6052517046278833e-05, "loss": 0.0806, "step": 5542 }, { "epoch": 0.6009323503902863, "grad_norm": 2.9873764514923096, "learning_rate": 1.605179167271145e-05, "loss": 0.0744, "step": 5543 }, { "epoch": 0.601040763226366, "grad_norm": 3.1852762699127197, "learning_rate": 1.605106629914406e-05, "loss": 0.0829, "step": 5544 }, { "epoch": 0.6011491760624458, "grad_norm": 1.9836705923080444, "learning_rate": 1.6050340925576673e-05, "loss": 0.0461, "step": 5545 }, { "epoch": 0.6012575888985255, "grad_norm": 5.759226322174072, "learning_rate": 1.6049615552009285e-05, "loss": 0.0812, "step": 5546 }, { "epoch": 0.6013660017346054, "grad_norm": 1.5301804542541504, "learning_rate": 1.6048890178441897e-05, "loss": 0.0388, "step": 5547 }, { "epoch": 0.6014744145706852, "grad_norm": 1.0413341522216797, "learning_rate": 1.604816480487451e-05, "loss": 0.0098, "step": 5548 }, { "epoch": 0.6015828274067649, "grad_norm": 0.6449833512306213, "learning_rate": 1.6047439431307125e-05, "loss": 0.0143, "step": 5549 }, { "epoch": 0.6016912402428447, "grad_norm": 2.934034585952759, "learning_rate": 1.6046714057739737e-05, "loss": 0.0195, "step": 5550 }, { "epoch": 0.6017996530789246, "grad_norm": 1.2690037488937378, "learning_rate": 1.6045988684172353e-05, "loss": 0.015, "step": 5551 }, { "epoch": 0.6019080659150043, "grad_norm": 2.8185551166534424, "learning_rate": 1.6045263310604965e-05, "loss": 0.0641, "step": 5552 }, { "epoch": 0.6020164787510841, "grad_norm": 0.769207775592804, "learning_rate": 1.6044537937037577e-05, "loss": 0.0261, "step": 5553 }, { "epoch": 0.6021248915871639, "grad_norm": 2.4708893299102783, "learning_rate": 1.604381256347019e-05, "loss": 0.034, "step": 5554 }, { "epoch": 0.6022333044232437, "grad_norm": 5.412479877471924, "learning_rate": 1.60430871899028e-05, "loss": 0.0538, "step": 5555 }, { "epoch": 0.6023417172593235, "grad_norm": 1.741579532623291, "learning_rate": 1.6042361816335414e-05, "loss": 0.0496, "step": 5556 }, { "epoch": 0.6024501300954033, "grad_norm": 3.0771193504333496, "learning_rate": 1.6041636442768026e-05, "loss": 0.0608, "step": 5557 }, { "epoch": 0.602558542931483, "grad_norm": 2.744896411895752, "learning_rate": 1.604091106920064e-05, "loss": 0.0396, "step": 5558 }, { "epoch": 0.6026669557675629, "grad_norm": 2.7260875701904297, "learning_rate": 1.6040185695633253e-05, "loss": 0.05, "step": 5559 }, { "epoch": 0.6027753686036427, "grad_norm": 1.680774450302124, "learning_rate": 1.6039460322065865e-05, "loss": 0.0231, "step": 5560 }, { "epoch": 0.6028837814397224, "grad_norm": 3.1212573051452637, "learning_rate": 1.6038734948498478e-05, "loss": 0.0741, "step": 5561 }, { "epoch": 0.6029921942758023, "grad_norm": 5.117452621459961, "learning_rate": 1.603800957493109e-05, "loss": 0.0638, "step": 5562 }, { "epoch": 0.6031006071118821, "grad_norm": 1.8200849294662476, "learning_rate": 1.6037284201363702e-05, "loss": 0.039, "step": 5563 }, { "epoch": 0.6032090199479618, "grad_norm": 3.5695903301239014, "learning_rate": 1.6036558827796317e-05, "loss": 0.0582, "step": 5564 }, { "epoch": 0.6033174327840416, "grad_norm": 1.5249443054199219, "learning_rate": 1.603583345422893e-05, "loss": 0.0203, "step": 5565 }, { "epoch": 0.6034258456201215, "grad_norm": 0.825475811958313, "learning_rate": 1.6035108080661542e-05, "loss": 0.0196, "step": 5566 }, { "epoch": 0.6035342584562012, "grad_norm": 1.995280146598816, "learning_rate": 1.6034382707094154e-05, "loss": 0.0402, "step": 5567 }, { "epoch": 0.603642671292281, "grad_norm": 4.4190497398376465, "learning_rate": 1.603365733352677e-05, "loss": 0.0632, "step": 5568 }, { "epoch": 0.6037510841283608, "grad_norm": 2.4307315349578857, "learning_rate": 1.603293195995938e-05, "loss": 0.04, "step": 5569 }, { "epoch": 0.6038594969644406, "grad_norm": 1.7408487796783447, "learning_rate": 1.6032206586391994e-05, "loss": 0.014, "step": 5570 }, { "epoch": 0.6039679098005204, "grad_norm": 0.7927572727203369, "learning_rate": 1.6031481212824606e-05, "loss": 0.011, "step": 5571 }, { "epoch": 0.6040763226366002, "grad_norm": 3.904986619949341, "learning_rate": 1.6030755839257218e-05, "loss": 0.041, "step": 5572 }, { "epoch": 0.6041847354726799, "grad_norm": 3.8392348289489746, "learning_rate": 1.6030030465689834e-05, "loss": 0.0444, "step": 5573 }, { "epoch": 0.6042931483087598, "grad_norm": 1.5470178127288818, "learning_rate": 1.6029305092122446e-05, "loss": 0.0388, "step": 5574 }, { "epoch": 0.6044015611448396, "grad_norm": 1.8130712509155273, "learning_rate": 1.6028579718555058e-05, "loss": 0.0448, "step": 5575 }, { "epoch": 0.6045099739809193, "grad_norm": 2.01387095451355, "learning_rate": 1.602785434498767e-05, "loss": 0.0231, "step": 5576 }, { "epoch": 0.6046183868169991, "grad_norm": 1.3776710033416748, "learning_rate": 1.6027128971420282e-05, "loss": 0.0224, "step": 5577 }, { "epoch": 0.604726799653079, "grad_norm": 3.4367213249206543, "learning_rate": 1.6026403597852894e-05, "loss": 0.0371, "step": 5578 }, { "epoch": 0.6048352124891587, "grad_norm": 2.3146257400512695, "learning_rate": 1.6025678224285506e-05, "loss": 0.0622, "step": 5579 }, { "epoch": 0.6049436253252385, "grad_norm": 0.8985017538070679, "learning_rate": 1.6024952850718122e-05, "loss": 0.0175, "step": 5580 }, { "epoch": 0.6050520381613183, "grad_norm": 2.950784921646118, "learning_rate": 1.6024227477150734e-05, "loss": 0.0389, "step": 5581 }, { "epoch": 0.6051604509973981, "grad_norm": 0.9741277098655701, "learning_rate": 1.6023502103583346e-05, "loss": 0.0358, "step": 5582 }, { "epoch": 0.6052688638334779, "grad_norm": 1.608138084411621, "learning_rate": 1.602277673001596e-05, "loss": 0.0303, "step": 5583 }, { "epoch": 0.6053772766695577, "grad_norm": 5.829714298248291, "learning_rate": 1.602205135644857e-05, "loss": 0.0418, "step": 5584 }, { "epoch": 0.6054856895056374, "grad_norm": 3.387040138244629, "learning_rate": 1.6021325982881183e-05, "loss": 0.0627, "step": 5585 }, { "epoch": 0.6055941023417173, "grad_norm": 4.730212688446045, "learning_rate": 1.6020600609313798e-05, "loss": 0.0455, "step": 5586 }, { "epoch": 0.605702515177797, "grad_norm": 5.25546407699585, "learning_rate": 1.601987523574641e-05, "loss": 0.0597, "step": 5587 }, { "epoch": 0.6058109280138768, "grad_norm": 3.806863307952881, "learning_rate": 1.6019149862179026e-05, "loss": 0.0815, "step": 5588 }, { "epoch": 0.6059193408499567, "grad_norm": 2.513624906539917, "learning_rate": 1.6018424488611638e-05, "loss": 0.0746, "step": 5589 }, { "epoch": 0.6060277536860365, "grad_norm": 3.305367946624756, "learning_rate": 1.601769911504425e-05, "loss": 0.0341, "step": 5590 }, { "epoch": 0.6061361665221162, "grad_norm": 4.481930255889893, "learning_rate": 1.6016973741476862e-05, "loss": 0.0315, "step": 5591 }, { "epoch": 0.606244579358196, "grad_norm": 1.6794763803482056, "learning_rate": 1.6016248367909474e-05, "loss": 0.02, "step": 5592 }, { "epoch": 0.6063529921942759, "grad_norm": 4.441989898681641, "learning_rate": 1.6015522994342087e-05, "loss": 0.0247, "step": 5593 }, { "epoch": 0.6064614050303556, "grad_norm": 3.5864853858947754, "learning_rate": 1.60147976207747e-05, "loss": 0.0287, "step": 5594 }, { "epoch": 0.6065698178664354, "grad_norm": 6.25453519821167, "learning_rate": 1.6014072247207314e-05, "loss": 0.0617, "step": 5595 }, { "epoch": 0.6066782307025151, "grad_norm": 1.3640540838241577, "learning_rate": 1.6013346873639926e-05, "loss": 0.027, "step": 5596 }, { "epoch": 0.606786643538595, "grad_norm": 4.882427215576172, "learning_rate": 1.601262150007254e-05, "loss": 0.0437, "step": 5597 }, { "epoch": 0.6068950563746748, "grad_norm": 1.783874273300171, "learning_rate": 1.601189612650515e-05, "loss": 0.03, "step": 5598 }, { "epoch": 0.6070034692107545, "grad_norm": 3.913567066192627, "learning_rate": 1.6011170752937763e-05, "loss": 0.042, "step": 5599 }, { "epoch": 0.6071118820468343, "grad_norm": 8.27513599395752, "learning_rate": 1.6010445379370375e-05, "loss": 0.0794, "step": 5600 }, { "epoch": 0.6072202948829142, "grad_norm": 2.4908816814422607, "learning_rate": 1.600972000580299e-05, "loss": 0.0156, "step": 5601 }, { "epoch": 0.6073287077189939, "grad_norm": 0.7978659868240356, "learning_rate": 1.6008994632235603e-05, "loss": 0.0201, "step": 5602 }, { "epoch": 0.6074371205550737, "grad_norm": 1.3309290409088135, "learning_rate": 1.6008269258668218e-05, "loss": 0.0093, "step": 5603 }, { "epoch": 0.6075455333911535, "grad_norm": 0.572108268737793, "learning_rate": 1.600754388510083e-05, "loss": 0.0079, "step": 5604 }, { "epoch": 0.6076539462272333, "grad_norm": 1.3729177713394165, "learning_rate": 1.6006818511533442e-05, "loss": 0.0219, "step": 5605 }, { "epoch": 0.6077623590633131, "grad_norm": 0.8143964409828186, "learning_rate": 1.6006093137966055e-05, "loss": 0.0094, "step": 5606 }, { "epoch": 0.6078707718993929, "grad_norm": 0.6090654730796814, "learning_rate": 1.6005367764398667e-05, "loss": 0.0106, "step": 5607 }, { "epoch": 0.6079791847354726, "grad_norm": 0.6825733780860901, "learning_rate": 1.600464239083128e-05, "loss": 0.0068, "step": 5608 }, { "epoch": 0.6080875975715525, "grad_norm": 1.5075416564941406, "learning_rate": 1.600391701726389e-05, "loss": 0.051, "step": 5609 }, { "epoch": 0.6081960104076323, "grad_norm": 1.0212644338607788, "learning_rate": 1.6003191643696507e-05, "loss": 0.0232, "step": 5610 }, { "epoch": 0.608304423243712, "grad_norm": 2.2761125564575195, "learning_rate": 1.600246627012912e-05, "loss": 0.0488, "step": 5611 }, { "epoch": 0.6084128360797918, "grad_norm": 5.563344955444336, "learning_rate": 1.600174089656173e-05, "loss": 0.0717, "step": 5612 }, { "epoch": 0.6085212489158717, "grad_norm": 0.4713234007358551, "learning_rate": 1.6001015522994343e-05, "loss": 0.0058, "step": 5613 }, { "epoch": 0.6086296617519514, "grad_norm": 4.494020938873291, "learning_rate": 1.6000290149426955e-05, "loss": 0.0476, "step": 5614 }, { "epoch": 0.6087380745880312, "grad_norm": 1.436490774154663, "learning_rate": 1.5999564775859567e-05, "loss": 0.0272, "step": 5615 }, { "epoch": 0.608846487424111, "grad_norm": 4.8592729568481445, "learning_rate": 1.5998839402292183e-05, "loss": 0.0479, "step": 5616 }, { "epoch": 0.6089549002601908, "grad_norm": 2.6979122161865234, "learning_rate": 1.5998114028724795e-05, "loss": 0.0265, "step": 5617 }, { "epoch": 0.6090633130962706, "grad_norm": 3.453815460205078, "learning_rate": 1.5997388655157407e-05, "loss": 0.0105, "step": 5618 }, { "epoch": 0.6091717259323504, "grad_norm": 2.3140528202056885, "learning_rate": 1.599666328159002e-05, "loss": 0.034, "step": 5619 }, { "epoch": 0.6092801387684302, "grad_norm": 2.78666615486145, "learning_rate": 1.5995937908022635e-05, "loss": 0.0259, "step": 5620 }, { "epoch": 0.60938855160451, "grad_norm": 0.7909703254699707, "learning_rate": 1.5995212534455247e-05, "loss": 0.0071, "step": 5621 }, { "epoch": 0.6094969644405898, "grad_norm": 1.5426374673843384, "learning_rate": 1.599448716088786e-05, "loss": 0.0062, "step": 5622 }, { "epoch": 0.6096053772766695, "grad_norm": 1.1778199672698975, "learning_rate": 1.599376178732047e-05, "loss": 0.0232, "step": 5623 }, { "epoch": 0.6097137901127494, "grad_norm": 3.46173095703125, "learning_rate": 1.5993036413753083e-05, "loss": 0.0614, "step": 5624 }, { "epoch": 0.6098222029488292, "grad_norm": 16.5068302154541, "learning_rate": 1.59923110401857e-05, "loss": 0.0357, "step": 5625 }, { "epoch": 0.6099306157849089, "grad_norm": 7.486351013183594, "learning_rate": 1.599158566661831e-05, "loss": 0.0948, "step": 5626 }, { "epoch": 0.6100390286209887, "grad_norm": 5.178927421569824, "learning_rate": 1.5990860293050923e-05, "loss": 0.0821, "step": 5627 }, { "epoch": 0.6101474414570686, "grad_norm": 0.8838773965835571, "learning_rate": 1.5990134919483535e-05, "loss": 0.0307, "step": 5628 }, { "epoch": 0.6102558542931483, "grad_norm": 2.471798896789551, "learning_rate": 1.5989409545916147e-05, "loss": 0.037, "step": 5629 }, { "epoch": 0.6103642671292281, "grad_norm": 2.991788864135742, "learning_rate": 1.598868417234876e-05, "loss": 0.0409, "step": 5630 }, { "epoch": 0.6104726799653079, "grad_norm": 1.5686076879501343, "learning_rate": 1.5987958798781372e-05, "loss": 0.0207, "step": 5631 }, { "epoch": 0.6105810928013877, "grad_norm": 1.3714693784713745, "learning_rate": 1.5987233425213987e-05, "loss": 0.0174, "step": 5632 }, { "epoch": 0.6106895056374675, "grad_norm": 4.39594030380249, "learning_rate": 1.59865080516466e-05, "loss": 0.0816, "step": 5633 }, { "epoch": 0.6107979184735473, "grad_norm": 3.82497239112854, "learning_rate": 1.598578267807921e-05, "loss": 0.0406, "step": 5634 }, { "epoch": 0.610906331309627, "grad_norm": 2.632361888885498, "learning_rate": 1.5985057304511824e-05, "loss": 0.0369, "step": 5635 }, { "epoch": 0.6110147441457069, "grad_norm": 2.327784776687622, "learning_rate": 1.5984331930944436e-05, "loss": 0.0512, "step": 5636 }, { "epoch": 0.6111231569817867, "grad_norm": 1.6466305255889893, "learning_rate": 1.598360655737705e-05, "loss": 0.0423, "step": 5637 }, { "epoch": 0.6112315698178664, "grad_norm": 3.5859832763671875, "learning_rate": 1.5982881183809663e-05, "loss": 0.0424, "step": 5638 }, { "epoch": 0.6113399826539462, "grad_norm": 1.2305768728256226, "learning_rate": 1.5982155810242276e-05, "loss": 0.0619, "step": 5639 }, { "epoch": 0.611448395490026, "grad_norm": 0.7412703037261963, "learning_rate": 1.598143043667489e-05, "loss": 0.007, "step": 5640 }, { "epoch": 0.6115568083261058, "grad_norm": 1.7932887077331543, "learning_rate": 1.5980705063107503e-05, "loss": 0.0132, "step": 5641 }, { "epoch": 0.6116652211621856, "grad_norm": 2.0444247722625732, "learning_rate": 1.5979979689540115e-05, "loss": 0.0227, "step": 5642 }, { "epoch": 0.6117736339982653, "grad_norm": 3.9727678298950195, "learning_rate": 1.5979254315972728e-05, "loss": 0.0483, "step": 5643 }, { "epoch": 0.6118820468343452, "grad_norm": 4.574518203735352, "learning_rate": 1.597852894240534e-05, "loss": 0.032, "step": 5644 }, { "epoch": 0.611990459670425, "grad_norm": 2.89910626411438, "learning_rate": 1.5977803568837952e-05, "loss": 0.0361, "step": 5645 }, { "epoch": 0.6120988725065047, "grad_norm": 2.669677734375, "learning_rate": 1.5977078195270564e-05, "loss": 0.0475, "step": 5646 }, { "epoch": 0.6122072853425846, "grad_norm": 1.2808960676193237, "learning_rate": 1.597635282170318e-05, "loss": 0.0379, "step": 5647 }, { "epoch": 0.6123156981786644, "grad_norm": 1.378271222114563, "learning_rate": 1.597562744813579e-05, "loss": 0.0231, "step": 5648 }, { "epoch": 0.6124241110147441, "grad_norm": 2.5668697357177734, "learning_rate": 1.5974902074568404e-05, "loss": 0.0489, "step": 5649 }, { "epoch": 0.6125325238508239, "grad_norm": 1.5276209115982056, "learning_rate": 1.5974176701001016e-05, "loss": 0.0387, "step": 5650 }, { "epoch": 0.6126409366869038, "grad_norm": 4.594977378845215, "learning_rate": 1.5973451327433628e-05, "loss": 0.034, "step": 5651 }, { "epoch": 0.6127493495229835, "grad_norm": 2.2355399131774902, "learning_rate": 1.597272595386624e-05, "loss": 0.0449, "step": 5652 }, { "epoch": 0.6128577623590633, "grad_norm": 3.8758199214935303, "learning_rate": 1.5972000580298856e-05, "loss": 0.0519, "step": 5653 }, { "epoch": 0.6129661751951431, "grad_norm": 2.670529365539551, "learning_rate": 1.5971275206731468e-05, "loss": 0.0564, "step": 5654 }, { "epoch": 0.613074588031223, "grad_norm": 4.070440292358398, "learning_rate": 1.5970549833164083e-05, "loss": 0.0395, "step": 5655 }, { "epoch": 0.6131830008673027, "grad_norm": 2.227527618408203, "learning_rate": 1.5969824459596696e-05, "loss": 0.0264, "step": 5656 }, { "epoch": 0.6132914137033825, "grad_norm": 1.3215223550796509, "learning_rate": 1.5969099086029308e-05, "loss": 0.0259, "step": 5657 }, { "epoch": 0.6133998265394622, "grad_norm": 3.472445487976074, "learning_rate": 1.596837371246192e-05, "loss": 0.0411, "step": 5658 }, { "epoch": 0.6135082393755421, "grad_norm": 2.452747106552124, "learning_rate": 1.5967648338894532e-05, "loss": 0.0411, "step": 5659 }, { "epoch": 0.6136166522116219, "grad_norm": 2.1446480751037598, "learning_rate": 1.5966922965327144e-05, "loss": 0.0278, "step": 5660 }, { "epoch": 0.6137250650477016, "grad_norm": 1.5435670614242554, "learning_rate": 1.5966197591759756e-05, "loss": 0.0264, "step": 5661 }, { "epoch": 0.6138334778837814, "grad_norm": 1.2367545366287231, "learning_rate": 1.5965472218192372e-05, "loss": 0.0161, "step": 5662 }, { "epoch": 0.6139418907198613, "grad_norm": 3.9986190795898438, "learning_rate": 1.5964746844624984e-05, "loss": 0.0727, "step": 5663 }, { "epoch": 0.614050303555941, "grad_norm": 1.6163920164108276, "learning_rate": 1.5964021471057596e-05, "loss": 0.0249, "step": 5664 }, { "epoch": 0.6141587163920208, "grad_norm": 0.7734789848327637, "learning_rate": 1.5963296097490208e-05, "loss": 0.0272, "step": 5665 }, { "epoch": 0.6142671292281006, "grad_norm": 1.308065414428711, "learning_rate": 1.596257072392282e-05, "loss": 0.0408, "step": 5666 }, { "epoch": 0.6143755420641804, "grad_norm": 2.9297525882720947, "learning_rate": 1.5961845350355433e-05, "loss": 0.0482, "step": 5667 }, { "epoch": 0.6144839549002602, "grad_norm": 1.0784868001937866, "learning_rate": 1.5961119976788048e-05, "loss": 0.0176, "step": 5668 }, { "epoch": 0.61459236773634, "grad_norm": 1.4731013774871826, "learning_rate": 1.596039460322066e-05, "loss": 0.0295, "step": 5669 }, { "epoch": 0.6147007805724197, "grad_norm": 4.506318092346191, "learning_rate": 1.5959669229653272e-05, "loss": 0.0402, "step": 5670 }, { "epoch": 0.6148091934084996, "grad_norm": 0.6389378905296326, "learning_rate": 1.5958943856085885e-05, "loss": 0.0178, "step": 5671 }, { "epoch": 0.6149176062445794, "grad_norm": 2.69271183013916, "learning_rate": 1.59582184825185e-05, "loss": 0.0341, "step": 5672 }, { "epoch": 0.6150260190806591, "grad_norm": 4.817913055419922, "learning_rate": 1.5957493108951112e-05, "loss": 0.0448, "step": 5673 }, { "epoch": 0.6151344319167389, "grad_norm": 5.705289840698242, "learning_rate": 1.5956767735383724e-05, "loss": 0.1024, "step": 5674 }, { "epoch": 0.6152428447528188, "grad_norm": 4.607203006744385, "learning_rate": 1.5956042361816337e-05, "loss": 0.0693, "step": 5675 }, { "epoch": 0.6153512575888985, "grad_norm": 3.0168685913085938, "learning_rate": 1.595531698824895e-05, "loss": 0.0219, "step": 5676 }, { "epoch": 0.6154596704249783, "grad_norm": 1.6320240497589111, "learning_rate": 1.5954591614681564e-05, "loss": 0.0276, "step": 5677 }, { "epoch": 0.6155680832610582, "grad_norm": 3.0211877822875977, "learning_rate": 1.5953866241114176e-05, "loss": 0.0745, "step": 5678 }, { "epoch": 0.6156764960971379, "grad_norm": 5.530289173126221, "learning_rate": 1.595314086754679e-05, "loss": 0.0579, "step": 5679 }, { "epoch": 0.6157849089332177, "grad_norm": 3.176231622695923, "learning_rate": 1.59524154939794e-05, "loss": 0.0202, "step": 5680 }, { "epoch": 0.6158933217692975, "grad_norm": 0.7449290752410889, "learning_rate": 1.5951690120412013e-05, "loss": 0.0087, "step": 5681 }, { "epoch": 0.6160017346053773, "grad_norm": 1.543864369392395, "learning_rate": 1.5950964746844625e-05, "loss": 0.0252, "step": 5682 }, { "epoch": 0.6161101474414571, "grad_norm": 0.8457866311073303, "learning_rate": 1.5950239373277237e-05, "loss": 0.0156, "step": 5683 }, { "epoch": 0.6162185602775369, "grad_norm": 1.2124531269073486, "learning_rate": 1.5949513999709853e-05, "loss": 0.0201, "step": 5684 }, { "epoch": 0.6163269731136166, "grad_norm": 1.1937522888183594, "learning_rate": 1.5948788626142465e-05, "loss": 0.0163, "step": 5685 }, { "epoch": 0.6164353859496965, "grad_norm": 1.7949323654174805, "learning_rate": 1.5948063252575077e-05, "loss": 0.0302, "step": 5686 }, { "epoch": 0.6165437987857763, "grad_norm": 1.2027610540390015, "learning_rate": 1.594733787900769e-05, "loss": 0.0069, "step": 5687 }, { "epoch": 0.616652211621856, "grad_norm": 5.076691150665283, "learning_rate": 1.59466125054403e-05, "loss": 0.0346, "step": 5688 }, { "epoch": 0.6167606244579358, "grad_norm": 1.0557289123535156, "learning_rate": 1.5945887131872917e-05, "loss": 0.013, "step": 5689 }, { "epoch": 0.6168690372940157, "grad_norm": 2.2726809978485107, "learning_rate": 1.594516175830553e-05, "loss": 0.0294, "step": 5690 }, { "epoch": 0.6169774501300954, "grad_norm": 1.3573542833328247, "learning_rate": 1.594443638473814e-05, "loss": 0.0303, "step": 5691 }, { "epoch": 0.6170858629661752, "grad_norm": 1.414507269859314, "learning_rate": 1.5943711011170756e-05, "loss": 0.0226, "step": 5692 }, { "epoch": 0.6171942758022549, "grad_norm": 1.3111586570739746, "learning_rate": 1.594298563760337e-05, "loss": 0.0122, "step": 5693 }, { "epoch": 0.6173026886383348, "grad_norm": 4.024963855743408, "learning_rate": 1.594226026403598e-05, "loss": 0.0494, "step": 5694 }, { "epoch": 0.6174111014744146, "grad_norm": 1.1545411348342896, "learning_rate": 1.5941534890468593e-05, "loss": 0.0172, "step": 5695 }, { "epoch": 0.6175195143104943, "grad_norm": 2.5739524364471436, "learning_rate": 1.5940809516901205e-05, "loss": 0.0248, "step": 5696 }, { "epoch": 0.6176279271465741, "grad_norm": 1.6735539436340332, "learning_rate": 1.5940084143333817e-05, "loss": 0.0483, "step": 5697 }, { "epoch": 0.617736339982654, "grad_norm": 2.829040050506592, "learning_rate": 1.593935876976643e-05, "loss": 0.0347, "step": 5698 }, { "epoch": 0.6178447528187337, "grad_norm": 2.4362471103668213, "learning_rate": 1.5938633396199045e-05, "loss": 0.08, "step": 5699 }, { "epoch": 0.6179531656548135, "grad_norm": 0.9437357783317566, "learning_rate": 1.5937908022631657e-05, "loss": 0.0077, "step": 5700 }, { "epoch": 0.6180615784908933, "grad_norm": 1.436763048171997, "learning_rate": 1.593718264906427e-05, "loss": 0.0329, "step": 5701 }, { "epoch": 0.6181699913269731, "grad_norm": 4.04925012588501, "learning_rate": 1.593645727549688e-05, "loss": 0.089, "step": 5702 }, { "epoch": 0.6182784041630529, "grad_norm": 4.703114986419678, "learning_rate": 1.5935731901929493e-05, "loss": 0.0553, "step": 5703 }, { "epoch": 0.6183868169991327, "grad_norm": 7.758544445037842, "learning_rate": 1.5935006528362106e-05, "loss": 0.0593, "step": 5704 }, { "epoch": 0.6184952298352124, "grad_norm": 4.922739028930664, "learning_rate": 1.593428115479472e-05, "loss": 0.0631, "step": 5705 }, { "epoch": 0.6186036426712923, "grad_norm": 2.8056318759918213, "learning_rate": 1.5933555781227333e-05, "loss": 0.0225, "step": 5706 }, { "epoch": 0.6187120555073721, "grad_norm": 4.063736438751221, "learning_rate": 1.593283040765995e-05, "loss": 0.0297, "step": 5707 }, { "epoch": 0.6188204683434518, "grad_norm": 2.5831754207611084, "learning_rate": 1.593210503409256e-05, "loss": 0.0952, "step": 5708 }, { "epoch": 0.6189288811795317, "grad_norm": 1.7830240726470947, "learning_rate": 1.5931379660525173e-05, "loss": 0.0279, "step": 5709 }, { "epoch": 0.6190372940156115, "grad_norm": 2.6711184978485107, "learning_rate": 1.5930654286957785e-05, "loss": 0.0433, "step": 5710 }, { "epoch": 0.6191457068516912, "grad_norm": 2.229545831680298, "learning_rate": 1.5929928913390397e-05, "loss": 0.0195, "step": 5711 }, { "epoch": 0.619254119687771, "grad_norm": 4.332931041717529, "learning_rate": 1.592920353982301e-05, "loss": 0.1453, "step": 5712 }, { "epoch": 0.6193625325238509, "grad_norm": 2.8487846851348877, "learning_rate": 1.592847816625562e-05, "loss": 0.0292, "step": 5713 }, { "epoch": 0.6194709453599306, "grad_norm": 1.249898076057434, "learning_rate": 1.5927752792688237e-05, "loss": 0.0403, "step": 5714 }, { "epoch": 0.6195793581960104, "grad_norm": 3.221956491470337, "learning_rate": 1.592702741912085e-05, "loss": 0.0499, "step": 5715 }, { "epoch": 0.6196877710320902, "grad_norm": 4.459357738494873, "learning_rate": 1.592630204555346e-05, "loss": 0.0438, "step": 5716 }, { "epoch": 0.61979618386817, "grad_norm": 0.39000120759010315, "learning_rate": 1.5925576671986074e-05, "loss": 0.0115, "step": 5717 }, { "epoch": 0.6199045967042498, "grad_norm": 2.2410175800323486, "learning_rate": 1.5924851298418686e-05, "loss": 0.0555, "step": 5718 }, { "epoch": 0.6200130095403296, "grad_norm": 4.7823052406311035, "learning_rate": 1.5924125924851298e-05, "loss": 0.0493, "step": 5719 }, { "epoch": 0.6201214223764093, "grad_norm": 1.5632388591766357, "learning_rate": 1.5923400551283913e-05, "loss": 0.0331, "step": 5720 }, { "epoch": 0.6202298352124892, "grad_norm": 1.31956148147583, "learning_rate": 1.5922675177716526e-05, "loss": 0.0187, "step": 5721 }, { "epoch": 0.620338248048569, "grad_norm": 1.026540756225586, "learning_rate": 1.5921949804149138e-05, "loss": 0.0136, "step": 5722 }, { "epoch": 0.6204466608846487, "grad_norm": 2.5321640968322754, "learning_rate": 1.592122443058175e-05, "loss": 0.0288, "step": 5723 }, { "epoch": 0.6205550737207285, "grad_norm": 2.1829590797424316, "learning_rate": 1.5920499057014365e-05, "loss": 0.0229, "step": 5724 }, { "epoch": 0.6206634865568084, "grad_norm": 1.5350303649902344, "learning_rate": 1.5919773683446978e-05, "loss": 0.0357, "step": 5725 }, { "epoch": 0.6207718993928881, "grad_norm": 2.2664406299591064, "learning_rate": 1.591904830987959e-05, "loss": 0.0285, "step": 5726 }, { "epoch": 0.6208803122289679, "grad_norm": 2.807314395904541, "learning_rate": 1.5918322936312202e-05, "loss": 0.0434, "step": 5727 }, { "epoch": 0.6209887250650477, "grad_norm": 2.1221582889556885, "learning_rate": 1.5917597562744814e-05, "loss": 0.0425, "step": 5728 }, { "epoch": 0.6210971379011275, "grad_norm": 2.0636796951293945, "learning_rate": 1.591687218917743e-05, "loss": 0.0351, "step": 5729 }, { "epoch": 0.6212055507372073, "grad_norm": 2.1793508529663086, "learning_rate": 1.591614681561004e-05, "loss": 0.0244, "step": 5730 }, { "epoch": 0.621313963573287, "grad_norm": 3.011720895767212, "learning_rate": 1.5915421442042654e-05, "loss": 0.0454, "step": 5731 }, { "epoch": 0.6214223764093668, "grad_norm": 3.7290093898773193, "learning_rate": 1.5914696068475266e-05, "loss": 0.0806, "step": 5732 }, { "epoch": 0.6215307892454467, "grad_norm": 3.8560304641723633, "learning_rate": 1.5913970694907878e-05, "loss": 0.0717, "step": 5733 }, { "epoch": 0.6216392020815265, "grad_norm": 1.0881266593933105, "learning_rate": 1.591324532134049e-05, "loss": 0.0119, "step": 5734 }, { "epoch": 0.6217476149176062, "grad_norm": 1.5343345403671265, "learning_rate": 1.5912519947773102e-05, "loss": 0.0347, "step": 5735 }, { "epoch": 0.6218560277536861, "grad_norm": 2.380817413330078, "learning_rate": 1.5911794574205718e-05, "loss": 0.0969, "step": 5736 }, { "epoch": 0.6219644405897659, "grad_norm": 1.126901388168335, "learning_rate": 1.591106920063833e-05, "loss": 0.0414, "step": 5737 }, { "epoch": 0.6220728534258456, "grad_norm": 1.1455470323562622, "learning_rate": 1.5910343827070942e-05, "loss": 0.0302, "step": 5738 }, { "epoch": 0.6221812662619254, "grad_norm": 1.6340630054473877, "learning_rate": 1.5909618453503554e-05, "loss": 0.0216, "step": 5739 }, { "epoch": 0.6222896790980053, "grad_norm": 1.2394579648971558, "learning_rate": 1.5908893079936166e-05, "loss": 0.0186, "step": 5740 }, { "epoch": 0.622398091934085, "grad_norm": 2.41145920753479, "learning_rate": 1.5908167706368782e-05, "loss": 0.0233, "step": 5741 }, { "epoch": 0.6225065047701648, "grad_norm": 1.6405519247055054, "learning_rate": 1.5907442332801394e-05, "loss": 0.0311, "step": 5742 }, { "epoch": 0.6226149176062445, "grad_norm": 2.7332608699798584, "learning_rate": 1.5906716959234006e-05, "loss": 0.059, "step": 5743 }, { "epoch": 0.6227233304423244, "grad_norm": 4.222472667694092, "learning_rate": 1.5905991585666622e-05, "loss": 0.0507, "step": 5744 }, { "epoch": 0.6228317432784042, "grad_norm": 2.7874176502227783, "learning_rate": 1.5905266212099234e-05, "loss": 0.0269, "step": 5745 }, { "epoch": 0.622940156114484, "grad_norm": 1.8098492622375488, "learning_rate": 1.5904540838531846e-05, "loss": 0.0259, "step": 5746 }, { "epoch": 0.6230485689505637, "grad_norm": 0.8589420914649963, "learning_rate": 1.5903815464964458e-05, "loss": 0.0235, "step": 5747 }, { "epoch": 0.6231569817866436, "grad_norm": 1.2648661136627197, "learning_rate": 1.590309009139707e-05, "loss": 0.0147, "step": 5748 }, { "epoch": 0.6232653946227233, "grad_norm": 3.5131242275238037, "learning_rate": 1.5902364717829683e-05, "loss": 0.0207, "step": 5749 }, { "epoch": 0.6233738074588031, "grad_norm": 2.1372528076171875, "learning_rate": 1.5901639344262295e-05, "loss": 0.0146, "step": 5750 }, { "epoch": 0.6234822202948829, "grad_norm": 5.211233615875244, "learning_rate": 1.590091397069491e-05, "loss": 0.0498, "step": 5751 }, { "epoch": 0.6235906331309627, "grad_norm": 4.938395023345947, "learning_rate": 1.5900188597127522e-05, "loss": 0.0189, "step": 5752 }, { "epoch": 0.6236990459670425, "grad_norm": 0.8525933027267456, "learning_rate": 1.5899463223560135e-05, "loss": 0.0059, "step": 5753 }, { "epoch": 0.6238074588031223, "grad_norm": 3.8542213439941406, "learning_rate": 1.5898737849992747e-05, "loss": 0.0409, "step": 5754 }, { "epoch": 0.623915871639202, "grad_norm": 1.5384234189987183, "learning_rate": 1.589801247642536e-05, "loss": 0.0162, "step": 5755 }, { "epoch": 0.6240242844752819, "grad_norm": 4.252272605895996, "learning_rate": 1.589728710285797e-05, "loss": 0.0161, "step": 5756 }, { "epoch": 0.6241326973113617, "grad_norm": 1.459502935409546, "learning_rate": 1.5896561729290586e-05, "loss": 0.074, "step": 5757 }, { "epoch": 0.6242411101474414, "grad_norm": 1.0389653444290161, "learning_rate": 1.58958363557232e-05, "loss": 0.0062, "step": 5758 }, { "epoch": 0.6243495229835212, "grad_norm": 1.3039542436599731, "learning_rate": 1.5895110982155814e-05, "loss": 0.0089, "step": 5759 }, { "epoch": 0.6244579358196011, "grad_norm": 0.76076740026474, "learning_rate": 1.5894385608588426e-05, "loss": 0.0058, "step": 5760 }, { "epoch": 0.6245663486556808, "grad_norm": 4.093835830688477, "learning_rate": 1.589366023502104e-05, "loss": 0.0548, "step": 5761 }, { "epoch": 0.6246747614917606, "grad_norm": 2.505847215652466, "learning_rate": 1.589293486145365e-05, "loss": 0.0496, "step": 5762 }, { "epoch": 0.6247831743278404, "grad_norm": 5.336690425872803, "learning_rate": 1.5892209487886263e-05, "loss": 0.0503, "step": 5763 }, { "epoch": 0.6248915871639202, "grad_norm": 4.178478717803955, "learning_rate": 1.5891484114318875e-05, "loss": 0.0713, "step": 5764 }, { "epoch": 0.625, "grad_norm": 2.6181867122650146, "learning_rate": 1.5890758740751487e-05, "loss": 0.0301, "step": 5765 }, { "epoch": 0.6251084128360798, "grad_norm": 3.119985580444336, "learning_rate": 1.5890033367184103e-05, "loss": 0.0284, "step": 5766 }, { "epoch": 0.6252168256721596, "grad_norm": 4.009745121002197, "learning_rate": 1.5889307993616715e-05, "loss": 0.0547, "step": 5767 }, { "epoch": 0.6253252385082394, "grad_norm": 1.750038743019104, "learning_rate": 1.5888582620049327e-05, "loss": 0.0353, "step": 5768 }, { "epoch": 0.6254336513443192, "grad_norm": 3.6283340454101562, "learning_rate": 1.588785724648194e-05, "loss": 0.04, "step": 5769 }, { "epoch": 0.6255420641803989, "grad_norm": 0.3718964457511902, "learning_rate": 1.588713187291455e-05, "loss": 0.0111, "step": 5770 }, { "epoch": 0.6256504770164788, "grad_norm": 7.078014850616455, "learning_rate": 1.5886406499347163e-05, "loss": 0.039, "step": 5771 }, { "epoch": 0.6257588898525586, "grad_norm": 1.6462178230285645, "learning_rate": 1.588568112577978e-05, "loss": 0.05, "step": 5772 }, { "epoch": 0.6258673026886383, "grad_norm": 2.294107437133789, "learning_rate": 1.588495575221239e-05, "loss": 0.0472, "step": 5773 }, { "epoch": 0.6259757155247181, "grad_norm": 1.831059455871582, "learning_rate": 1.5884230378645003e-05, "loss": 0.0248, "step": 5774 }, { "epoch": 0.626084128360798, "grad_norm": 5.327182292938232, "learning_rate": 1.5883505005077615e-05, "loss": 0.029, "step": 5775 }, { "epoch": 0.6261925411968777, "grad_norm": 1.7352570295333862, "learning_rate": 1.588277963151023e-05, "loss": 0.0561, "step": 5776 }, { "epoch": 0.6263009540329575, "grad_norm": 5.8547773361206055, "learning_rate": 1.5882054257942843e-05, "loss": 0.0776, "step": 5777 }, { "epoch": 0.6264093668690373, "grad_norm": 0.665486752986908, "learning_rate": 1.5881328884375455e-05, "loss": 0.022, "step": 5778 }, { "epoch": 0.6265177797051171, "grad_norm": 1.5057414770126343, "learning_rate": 1.5880603510808067e-05, "loss": 0.023, "step": 5779 }, { "epoch": 0.6266261925411969, "grad_norm": 4.744064807891846, "learning_rate": 1.587987813724068e-05, "loss": 0.0233, "step": 5780 }, { "epoch": 0.6267346053772767, "grad_norm": 0.5608115792274475, "learning_rate": 1.5879152763673295e-05, "loss": 0.015, "step": 5781 }, { "epoch": 0.6268430182133564, "grad_norm": 3.4172258377075195, "learning_rate": 1.5878427390105907e-05, "loss": 0.0384, "step": 5782 }, { "epoch": 0.6269514310494363, "grad_norm": 1.465705394744873, "learning_rate": 1.587770201653852e-05, "loss": 0.0319, "step": 5783 }, { "epoch": 0.627059843885516, "grad_norm": 1.1149017810821533, "learning_rate": 1.587697664297113e-05, "loss": 0.0256, "step": 5784 }, { "epoch": 0.6271682567215958, "grad_norm": 5.412172317504883, "learning_rate": 1.5876251269403743e-05, "loss": 0.0176, "step": 5785 }, { "epoch": 0.6272766695576756, "grad_norm": 1.5396549701690674, "learning_rate": 1.5875525895836356e-05, "loss": 0.0189, "step": 5786 }, { "epoch": 0.6273850823937555, "grad_norm": 2.9446804523468018, "learning_rate": 1.5874800522268968e-05, "loss": 0.0785, "step": 5787 }, { "epoch": 0.6274934952298352, "grad_norm": 2.0442028045654297, "learning_rate": 1.5874075148701583e-05, "loss": 0.0524, "step": 5788 }, { "epoch": 0.627601908065915, "grad_norm": 4.776303768157959, "learning_rate": 1.5873349775134195e-05, "loss": 0.0893, "step": 5789 }, { "epoch": 0.6277103209019947, "grad_norm": 6.897060871124268, "learning_rate": 1.5872624401566808e-05, "loss": 0.0739, "step": 5790 }, { "epoch": 0.6278187337380746, "grad_norm": 2.673593044281006, "learning_rate": 1.587189902799942e-05, "loss": 0.054, "step": 5791 }, { "epoch": 0.6279271465741544, "grad_norm": 1.4150705337524414, "learning_rate": 1.5871173654432032e-05, "loss": 0.0157, "step": 5792 }, { "epoch": 0.6280355594102341, "grad_norm": 1.2524023056030273, "learning_rate": 1.5870448280864647e-05, "loss": 0.0135, "step": 5793 }, { "epoch": 0.6281439722463139, "grad_norm": 0.42538022994995117, "learning_rate": 1.586972290729726e-05, "loss": 0.0092, "step": 5794 }, { "epoch": 0.6282523850823938, "grad_norm": 2.2786905765533447, "learning_rate": 1.586899753372987e-05, "loss": 0.042, "step": 5795 }, { "epoch": 0.6283607979184735, "grad_norm": 1.715990424156189, "learning_rate": 1.5868272160162487e-05, "loss": 0.0327, "step": 5796 }, { "epoch": 0.6284692107545533, "grad_norm": 0.8783934712409973, "learning_rate": 1.58675467865951e-05, "loss": 0.0167, "step": 5797 }, { "epoch": 0.6285776235906332, "grad_norm": 1.4479632377624512, "learning_rate": 1.586682141302771e-05, "loss": 0.0132, "step": 5798 }, { "epoch": 0.628686036426713, "grad_norm": 2.2600646018981934, "learning_rate": 1.5866096039460324e-05, "loss": 0.0481, "step": 5799 }, { "epoch": 0.6287944492627927, "grad_norm": 1.4254759550094604, "learning_rate": 1.5865370665892936e-05, "loss": 0.0246, "step": 5800 }, { "epoch": 0.6289028620988725, "grad_norm": 9.543566703796387, "learning_rate": 1.5864645292325548e-05, "loss": 0.1486, "step": 5801 }, { "epoch": 0.6290112749349523, "grad_norm": 4.114129543304443, "learning_rate": 1.586391991875816e-05, "loss": 0.0817, "step": 5802 }, { "epoch": 0.6291196877710321, "grad_norm": 3.0668654441833496, "learning_rate": 1.5863194545190776e-05, "loss": 0.0496, "step": 5803 }, { "epoch": 0.6292281006071119, "grad_norm": 2.3362841606140137, "learning_rate": 1.5862469171623388e-05, "loss": 0.0132, "step": 5804 }, { "epoch": 0.6293365134431916, "grad_norm": 0.8943531513214111, "learning_rate": 1.5861743798056e-05, "loss": 0.0161, "step": 5805 }, { "epoch": 0.6294449262792715, "grad_norm": 1.2603414058685303, "learning_rate": 1.5861018424488612e-05, "loss": 0.0418, "step": 5806 }, { "epoch": 0.6295533391153513, "grad_norm": 2.2915730476379395, "learning_rate": 1.5860293050921224e-05, "loss": 0.031, "step": 5807 }, { "epoch": 0.629661751951431, "grad_norm": 3.4129698276519775, "learning_rate": 1.5859567677353836e-05, "loss": 0.0529, "step": 5808 }, { "epoch": 0.6297701647875108, "grad_norm": 4.868965148925781, "learning_rate": 1.5858842303786452e-05, "loss": 0.0381, "step": 5809 }, { "epoch": 0.6298785776235907, "grad_norm": 3.8729898929595947, "learning_rate": 1.5858116930219064e-05, "loss": 0.0567, "step": 5810 }, { "epoch": 0.6299869904596704, "grad_norm": 2.2245287895202637, "learning_rate": 1.585739155665168e-05, "loss": 0.0317, "step": 5811 }, { "epoch": 0.6300954032957502, "grad_norm": 3.496701955795288, "learning_rate": 1.585666618308429e-05, "loss": 0.082, "step": 5812 }, { "epoch": 0.63020381613183, "grad_norm": 3.2085230350494385, "learning_rate": 1.5855940809516904e-05, "loss": 0.0335, "step": 5813 }, { "epoch": 0.6303122289679098, "grad_norm": 3.7660062313079834, "learning_rate": 1.5855215435949516e-05, "loss": 0.0219, "step": 5814 }, { "epoch": 0.6304206418039896, "grad_norm": 4.029841899871826, "learning_rate": 1.5854490062382128e-05, "loss": 0.0477, "step": 5815 }, { "epoch": 0.6305290546400694, "grad_norm": 1.9276591539382935, "learning_rate": 1.585376468881474e-05, "loss": 0.0568, "step": 5816 }, { "epoch": 0.6306374674761491, "grad_norm": 2.0981791019439697, "learning_rate": 1.5853039315247352e-05, "loss": 0.0597, "step": 5817 }, { "epoch": 0.630745880312229, "grad_norm": 2.7591702938079834, "learning_rate": 1.5852313941679968e-05, "loss": 0.0854, "step": 5818 }, { "epoch": 0.6308542931483088, "grad_norm": 3.0401854515075684, "learning_rate": 1.585158856811258e-05, "loss": 0.027, "step": 5819 }, { "epoch": 0.6309627059843885, "grad_norm": 2.9166676998138428, "learning_rate": 1.5850863194545192e-05, "loss": 0.0241, "step": 5820 }, { "epoch": 0.6310711188204683, "grad_norm": 5.257139205932617, "learning_rate": 1.5850137820977804e-05, "loss": 0.0426, "step": 5821 }, { "epoch": 0.6311795316565482, "grad_norm": 4.678335189819336, "learning_rate": 1.5849412447410416e-05, "loss": 0.0132, "step": 5822 }, { "epoch": 0.6312879444926279, "grad_norm": 2.5787947177886963, "learning_rate": 1.584868707384303e-05, "loss": 0.0713, "step": 5823 }, { "epoch": 0.6313963573287077, "grad_norm": 0.39657530188560486, "learning_rate": 1.5847961700275644e-05, "loss": 0.0032, "step": 5824 }, { "epoch": 0.6315047701647876, "grad_norm": 2.1482670307159424, "learning_rate": 1.5847236326708256e-05, "loss": 0.0234, "step": 5825 }, { "epoch": 0.6316131830008673, "grad_norm": 2.84255051612854, "learning_rate": 1.584651095314087e-05, "loss": 0.0438, "step": 5826 }, { "epoch": 0.6317215958369471, "grad_norm": 1.1968170404434204, "learning_rate": 1.5845785579573484e-05, "loss": 0.0348, "step": 5827 }, { "epoch": 0.6318300086730269, "grad_norm": 0.793670117855072, "learning_rate": 1.5845060206006096e-05, "loss": 0.0183, "step": 5828 }, { "epoch": 0.6319384215091067, "grad_norm": 1.9342310428619385, "learning_rate": 1.5844334832438708e-05, "loss": 0.028, "step": 5829 }, { "epoch": 0.6320468343451865, "grad_norm": 2.728050470352173, "learning_rate": 1.584360945887132e-05, "loss": 0.0634, "step": 5830 }, { "epoch": 0.6321552471812663, "grad_norm": 1.2846770286560059, "learning_rate": 1.5842884085303933e-05, "loss": 0.0253, "step": 5831 }, { "epoch": 0.632263660017346, "grad_norm": 2.3075907230377197, "learning_rate": 1.5842158711736545e-05, "loss": 0.0271, "step": 5832 }, { "epoch": 0.6323720728534259, "grad_norm": 1.90337073802948, "learning_rate": 1.584143333816916e-05, "loss": 0.1015, "step": 5833 }, { "epoch": 0.6324804856895057, "grad_norm": 0.69927579164505, "learning_rate": 1.5840707964601772e-05, "loss": 0.0105, "step": 5834 }, { "epoch": 0.6325888985255854, "grad_norm": 7.455594539642334, "learning_rate": 1.5839982591034384e-05, "loss": 0.0648, "step": 5835 }, { "epoch": 0.6326973113616652, "grad_norm": 2.302048921585083, "learning_rate": 1.5839257217466997e-05, "loss": 0.0406, "step": 5836 }, { "epoch": 0.6328057241977451, "grad_norm": 1.0149997472763062, "learning_rate": 1.583853184389961e-05, "loss": 0.008, "step": 5837 }, { "epoch": 0.6329141370338248, "grad_norm": 0.7989854216575623, "learning_rate": 1.583780647033222e-05, "loss": 0.0141, "step": 5838 }, { "epoch": 0.6330225498699046, "grad_norm": 1.78510582447052, "learning_rate": 1.5837081096764833e-05, "loss": 0.0495, "step": 5839 }, { "epoch": 0.6331309627059843, "grad_norm": 2.3377041816711426, "learning_rate": 1.583635572319745e-05, "loss": 0.0328, "step": 5840 }, { "epoch": 0.6332393755420642, "grad_norm": 3.5756540298461914, "learning_rate": 1.583563034963006e-05, "loss": 0.0342, "step": 5841 }, { "epoch": 0.633347788378144, "grad_norm": 0.9763975143432617, "learning_rate": 1.5834904976062673e-05, "loss": 0.0054, "step": 5842 }, { "epoch": 0.6334562012142237, "grad_norm": 0.4506850242614746, "learning_rate": 1.5834179602495285e-05, "loss": 0.0176, "step": 5843 }, { "epoch": 0.6335646140503035, "grad_norm": 1.5142766237258911, "learning_rate": 1.5833454228927897e-05, "loss": 0.0438, "step": 5844 }, { "epoch": 0.6336730268863834, "grad_norm": 3.266386032104492, "learning_rate": 1.5832728855360513e-05, "loss": 0.0603, "step": 5845 }, { "epoch": 0.6337814397224631, "grad_norm": 2.7608590126037598, "learning_rate": 1.5832003481793125e-05, "loss": 0.0484, "step": 5846 }, { "epoch": 0.6338898525585429, "grad_norm": 1.9177398681640625, "learning_rate": 1.5831278108225737e-05, "loss": 0.0582, "step": 5847 }, { "epoch": 0.6339982653946227, "grad_norm": 1.2325174808502197, "learning_rate": 1.5830552734658353e-05, "loss": 0.0379, "step": 5848 }, { "epoch": 0.6341066782307025, "grad_norm": 1.6710749864578247, "learning_rate": 1.5829827361090965e-05, "loss": 0.0216, "step": 5849 }, { "epoch": 0.6342150910667823, "grad_norm": 3.051342010498047, "learning_rate": 1.5829101987523577e-05, "loss": 0.0581, "step": 5850 }, { "epoch": 0.6343235039028621, "grad_norm": 5.332864761352539, "learning_rate": 1.582837661395619e-05, "loss": 0.0837, "step": 5851 }, { "epoch": 0.6344319167389418, "grad_norm": 1.8193334341049194, "learning_rate": 1.58276512403888e-05, "loss": 0.0329, "step": 5852 }, { "epoch": 0.6345403295750217, "grad_norm": 1.494922399520874, "learning_rate": 1.5826925866821413e-05, "loss": 0.0293, "step": 5853 }, { "epoch": 0.6346487424111015, "grad_norm": 3.356964349746704, "learning_rate": 1.5826200493254025e-05, "loss": 0.0273, "step": 5854 }, { "epoch": 0.6347571552471812, "grad_norm": 0.7716371417045593, "learning_rate": 1.582547511968664e-05, "loss": 0.0164, "step": 5855 }, { "epoch": 0.6348655680832611, "grad_norm": 0.713186502456665, "learning_rate": 1.5824749746119253e-05, "loss": 0.0137, "step": 5856 }, { "epoch": 0.6349739809193409, "grad_norm": 7.409907341003418, "learning_rate": 1.5824024372551865e-05, "loss": 0.0363, "step": 5857 }, { "epoch": 0.6350823937554206, "grad_norm": 7.315070629119873, "learning_rate": 1.5823298998984477e-05, "loss": 0.0897, "step": 5858 }, { "epoch": 0.6351908065915004, "grad_norm": 0.6774212121963501, "learning_rate": 1.582257362541709e-05, "loss": 0.0087, "step": 5859 }, { "epoch": 0.6352992194275803, "grad_norm": 1.4983899593353271, "learning_rate": 1.58218482518497e-05, "loss": 0.0299, "step": 5860 }, { "epoch": 0.63540763226366, "grad_norm": 4.083414554595947, "learning_rate": 1.5821122878282317e-05, "loss": 0.0827, "step": 5861 }, { "epoch": 0.6355160450997398, "grad_norm": 0.7957413196563721, "learning_rate": 1.582039750471493e-05, "loss": 0.02, "step": 5862 }, { "epoch": 0.6356244579358196, "grad_norm": 2.0019173622131348, "learning_rate": 1.5819672131147545e-05, "loss": 0.0294, "step": 5863 }, { "epoch": 0.6357328707718994, "grad_norm": 3.7840077877044678, "learning_rate": 1.5818946757580157e-05, "loss": 0.076, "step": 5864 }, { "epoch": 0.6358412836079792, "grad_norm": 3.4756693840026855, "learning_rate": 1.581822138401277e-05, "loss": 0.0596, "step": 5865 }, { "epoch": 0.635949696444059, "grad_norm": 2.079010486602783, "learning_rate": 1.581749601044538e-05, "loss": 0.0158, "step": 5866 }, { "epoch": 0.6360581092801387, "grad_norm": 3.679152488708496, "learning_rate": 1.5816770636877993e-05, "loss": 0.0321, "step": 5867 }, { "epoch": 0.6361665221162186, "grad_norm": 3.7267887592315674, "learning_rate": 1.5816045263310606e-05, "loss": 0.037, "step": 5868 }, { "epoch": 0.6362749349522984, "grad_norm": 5.828114032745361, "learning_rate": 1.5815319889743218e-05, "loss": 0.04, "step": 5869 }, { "epoch": 0.6363833477883781, "grad_norm": 3.1889541149139404, "learning_rate": 1.5814594516175833e-05, "loss": 0.0272, "step": 5870 }, { "epoch": 0.6364917606244579, "grad_norm": 0.6159899234771729, "learning_rate": 1.5813869142608445e-05, "loss": 0.0084, "step": 5871 }, { "epoch": 0.6366001734605378, "grad_norm": 1.465815782546997, "learning_rate": 1.5813143769041057e-05, "loss": 0.0336, "step": 5872 }, { "epoch": 0.6367085862966175, "grad_norm": 2.025092124938965, "learning_rate": 1.581241839547367e-05, "loss": 0.048, "step": 5873 }, { "epoch": 0.6368169991326973, "grad_norm": 2.3204054832458496, "learning_rate": 1.5811693021906282e-05, "loss": 0.0276, "step": 5874 }, { "epoch": 0.636925411968777, "grad_norm": 1.9372390508651733, "learning_rate": 1.5810967648338894e-05, "loss": 0.0233, "step": 5875 }, { "epoch": 0.6370338248048569, "grad_norm": 1.318209171295166, "learning_rate": 1.581024227477151e-05, "loss": 0.0216, "step": 5876 }, { "epoch": 0.6371422376409367, "grad_norm": 1.1576766967773438, "learning_rate": 1.580951690120412e-05, "loss": 0.0181, "step": 5877 }, { "epoch": 0.6372506504770165, "grad_norm": 2.4127368927001953, "learning_rate": 1.5808791527636734e-05, "loss": 0.0611, "step": 5878 }, { "epoch": 0.6373590633130962, "grad_norm": 0.5800670385360718, "learning_rate": 1.580806615406935e-05, "loss": 0.012, "step": 5879 }, { "epoch": 0.6374674761491761, "grad_norm": 1.1074327230453491, "learning_rate": 1.580734078050196e-05, "loss": 0.0369, "step": 5880 }, { "epoch": 0.6375758889852559, "grad_norm": 3.297293186187744, "learning_rate": 1.5806615406934574e-05, "loss": 0.0201, "step": 5881 }, { "epoch": 0.6376843018213356, "grad_norm": 2.142427921295166, "learning_rate": 1.5805890033367186e-05, "loss": 0.033, "step": 5882 }, { "epoch": 0.6377927146574154, "grad_norm": 1.0469682216644287, "learning_rate": 1.5805164659799798e-05, "loss": 0.0213, "step": 5883 }, { "epoch": 0.6379011274934953, "grad_norm": 1.8661270141601562, "learning_rate": 1.580443928623241e-05, "loss": 0.0168, "step": 5884 }, { "epoch": 0.638009540329575, "grad_norm": 1.258832573890686, "learning_rate": 1.5803713912665026e-05, "loss": 0.0199, "step": 5885 }, { "epoch": 0.6381179531656548, "grad_norm": 4.415274620056152, "learning_rate": 1.5802988539097638e-05, "loss": 0.0415, "step": 5886 }, { "epoch": 0.6382263660017347, "grad_norm": 3.1516056060791016, "learning_rate": 1.580226316553025e-05, "loss": 0.0426, "step": 5887 }, { "epoch": 0.6383347788378144, "grad_norm": 1.557689905166626, "learning_rate": 1.5801537791962862e-05, "loss": 0.0312, "step": 5888 }, { "epoch": 0.6384431916738942, "grad_norm": 2.749910831451416, "learning_rate": 1.5800812418395474e-05, "loss": 0.0259, "step": 5889 }, { "epoch": 0.638551604509974, "grad_norm": 0.7581245303153992, "learning_rate": 1.5800087044828086e-05, "loss": 0.0196, "step": 5890 }, { "epoch": 0.6386600173460538, "grad_norm": 1.2828292846679688, "learning_rate": 1.5799361671260702e-05, "loss": 0.0349, "step": 5891 }, { "epoch": 0.6387684301821336, "grad_norm": 6.499968528747559, "learning_rate": 1.5798636297693314e-05, "loss": 0.0738, "step": 5892 }, { "epoch": 0.6388768430182133, "grad_norm": 5.767934799194336, "learning_rate": 1.5797910924125926e-05, "loss": 0.1221, "step": 5893 }, { "epoch": 0.6389852558542931, "grad_norm": 1.666636347770691, "learning_rate": 1.5797185550558538e-05, "loss": 0.017, "step": 5894 }, { "epoch": 0.639093668690373, "grad_norm": 4.9062275886535645, "learning_rate": 1.579646017699115e-05, "loss": 0.0409, "step": 5895 }, { "epoch": 0.6392020815264527, "grad_norm": 2.5166637897491455, "learning_rate": 1.5795734803423766e-05, "loss": 0.0361, "step": 5896 }, { "epoch": 0.6393104943625325, "grad_norm": 5.665079116821289, "learning_rate": 1.5795009429856378e-05, "loss": 0.0918, "step": 5897 }, { "epoch": 0.6394189071986123, "grad_norm": 7.325791358947754, "learning_rate": 1.579428405628899e-05, "loss": 0.0664, "step": 5898 }, { "epoch": 0.6395273200346921, "grad_norm": 2.1290152072906494, "learning_rate": 1.5793558682721602e-05, "loss": 0.0495, "step": 5899 }, { "epoch": 0.6396357328707719, "grad_norm": 1.7285388708114624, "learning_rate": 1.5792833309154218e-05, "loss": 0.0103, "step": 5900 }, { "epoch": 0.6397441457068517, "grad_norm": 0.6713706254959106, "learning_rate": 1.579210793558683e-05, "loss": 0.0051, "step": 5901 }, { "epoch": 0.6398525585429314, "grad_norm": 1.3438829183578491, "learning_rate": 1.5791382562019442e-05, "loss": 0.0075, "step": 5902 }, { "epoch": 0.6399609713790113, "grad_norm": 0.06025199592113495, "learning_rate": 1.5790657188452054e-05, "loss": 0.0004, "step": 5903 }, { "epoch": 0.6400693842150911, "grad_norm": 2.5132386684417725, "learning_rate": 1.5789931814884666e-05, "loss": 0.0499, "step": 5904 }, { "epoch": 0.6401777970511708, "grad_norm": 3.569532871246338, "learning_rate": 1.578920644131728e-05, "loss": 0.0327, "step": 5905 }, { "epoch": 0.6402862098872506, "grad_norm": 5.551634311676025, "learning_rate": 1.578848106774989e-05, "loss": 0.0427, "step": 5906 }, { "epoch": 0.6403946227233305, "grad_norm": 3.990400552749634, "learning_rate": 1.5787755694182506e-05, "loss": 0.0592, "step": 5907 }, { "epoch": 0.6405030355594102, "grad_norm": 4.462093353271484, "learning_rate": 1.578703032061512e-05, "loss": 0.0718, "step": 5908 }, { "epoch": 0.64061144839549, "grad_norm": 3.066787004470825, "learning_rate": 1.578630494704773e-05, "loss": 0.0381, "step": 5909 }, { "epoch": 0.6407198612315698, "grad_norm": 1.3475372791290283, "learning_rate": 1.5785579573480343e-05, "loss": 0.0339, "step": 5910 }, { "epoch": 0.6408282740676496, "grad_norm": 1.4761534929275513, "learning_rate": 1.5784854199912955e-05, "loss": 0.0279, "step": 5911 }, { "epoch": 0.6409366869037294, "grad_norm": 3.9236323833465576, "learning_rate": 1.5784128826345567e-05, "loss": 0.1132, "step": 5912 }, { "epoch": 0.6410450997398092, "grad_norm": 0.639500081539154, "learning_rate": 1.5783403452778182e-05, "loss": 0.0292, "step": 5913 }, { "epoch": 0.641153512575889, "grad_norm": 0.8044514060020447, "learning_rate": 1.5782678079210795e-05, "loss": 0.0173, "step": 5914 }, { "epoch": 0.6412619254119688, "grad_norm": 3.9762818813323975, "learning_rate": 1.578195270564341e-05, "loss": 0.0538, "step": 5915 }, { "epoch": 0.6413703382480486, "grad_norm": 1.2850911617279053, "learning_rate": 1.5781227332076022e-05, "loss": 0.0135, "step": 5916 }, { "epoch": 0.6414787510841283, "grad_norm": 3.4279463291168213, "learning_rate": 1.5780501958508634e-05, "loss": 0.0911, "step": 5917 }, { "epoch": 0.6415871639202082, "grad_norm": 1.030932903289795, "learning_rate": 1.5779776584941247e-05, "loss": 0.0251, "step": 5918 }, { "epoch": 0.641695576756288, "grad_norm": 1.8722914457321167, "learning_rate": 1.577905121137386e-05, "loss": 0.0235, "step": 5919 }, { "epoch": 0.6418039895923677, "grad_norm": 1.5998436212539673, "learning_rate": 1.577832583780647e-05, "loss": 0.0278, "step": 5920 }, { "epoch": 0.6419124024284475, "grad_norm": 3.511406660079956, "learning_rate": 1.5777600464239083e-05, "loss": 0.0798, "step": 5921 }, { "epoch": 0.6420208152645274, "grad_norm": 0.7035948038101196, "learning_rate": 1.57768750906717e-05, "loss": 0.0074, "step": 5922 }, { "epoch": 0.6421292281006071, "grad_norm": 2.9844725131988525, "learning_rate": 1.577614971710431e-05, "loss": 0.058, "step": 5923 }, { "epoch": 0.6422376409366869, "grad_norm": 4.2923383712768555, "learning_rate": 1.5775424343536923e-05, "loss": 0.0309, "step": 5924 }, { "epoch": 0.6423460537727667, "grad_norm": 4.658617973327637, "learning_rate": 1.5774698969969535e-05, "loss": 0.0293, "step": 5925 }, { "epoch": 0.6424544666088465, "grad_norm": 4.38146448135376, "learning_rate": 1.5773973596402147e-05, "loss": 0.0224, "step": 5926 }, { "epoch": 0.6425628794449263, "grad_norm": 0.8450743556022644, "learning_rate": 1.577324822283476e-05, "loss": 0.014, "step": 5927 }, { "epoch": 0.6426712922810061, "grad_norm": 1.7529592514038086, "learning_rate": 1.5772522849267375e-05, "loss": 0.0239, "step": 5928 }, { "epoch": 0.6427797051170858, "grad_norm": 1.7503379583358765, "learning_rate": 1.5771797475699987e-05, "loss": 0.0285, "step": 5929 }, { "epoch": 0.6428881179531657, "grad_norm": 6.146350383758545, "learning_rate": 1.57710721021326e-05, "loss": 0.0275, "step": 5930 }, { "epoch": 0.6429965307892455, "grad_norm": 1.7193694114685059, "learning_rate": 1.5770346728565215e-05, "loss": 0.0396, "step": 5931 }, { "epoch": 0.6431049436253252, "grad_norm": 4.483680248260498, "learning_rate": 1.5769621354997827e-05, "loss": 0.0889, "step": 5932 }, { "epoch": 0.643213356461405, "grad_norm": 3.2717983722686768, "learning_rate": 1.576889598143044e-05, "loss": 0.1076, "step": 5933 }, { "epoch": 0.6433217692974849, "grad_norm": 2.2444944381713867, "learning_rate": 1.576817060786305e-05, "loss": 0.0326, "step": 5934 }, { "epoch": 0.6434301821335646, "grad_norm": 4.424882888793945, "learning_rate": 1.5767445234295663e-05, "loss": 0.0747, "step": 5935 }, { "epoch": 0.6435385949696444, "grad_norm": 1.7369416952133179, "learning_rate": 1.5766719860728275e-05, "loss": 0.0418, "step": 5936 }, { "epoch": 0.6436470078057241, "grad_norm": 3.0158565044403076, "learning_rate": 1.576599448716089e-05, "loss": 0.0231, "step": 5937 }, { "epoch": 0.643755420641804, "grad_norm": 1.5302475690841675, "learning_rate": 1.5765269113593503e-05, "loss": 0.0106, "step": 5938 }, { "epoch": 0.6438638334778838, "grad_norm": 5.021337509155273, "learning_rate": 1.5764543740026115e-05, "loss": 0.0789, "step": 5939 }, { "epoch": 0.6439722463139635, "grad_norm": 2.2478909492492676, "learning_rate": 1.5763818366458727e-05, "loss": 0.0394, "step": 5940 }, { "epoch": 0.6440806591500433, "grad_norm": 3.773231029510498, "learning_rate": 1.576309299289134e-05, "loss": 0.0376, "step": 5941 }, { "epoch": 0.6441890719861232, "grad_norm": 2.76583194732666, "learning_rate": 1.576236761932395e-05, "loss": 0.0425, "step": 5942 }, { "epoch": 0.644297484822203, "grad_norm": 6.524888038635254, "learning_rate": 1.5761642245756567e-05, "loss": 0.0395, "step": 5943 }, { "epoch": 0.6444058976582827, "grad_norm": 1.3676053285598755, "learning_rate": 1.576091687218918e-05, "loss": 0.036, "step": 5944 }, { "epoch": 0.6445143104943626, "grad_norm": 1.8123160600662231, "learning_rate": 1.576019149862179e-05, "loss": 0.0396, "step": 5945 }, { "epoch": 0.6446227233304423, "grad_norm": 3.698472499847412, "learning_rate": 1.5759466125054404e-05, "loss": 0.0376, "step": 5946 }, { "epoch": 0.6447311361665221, "grad_norm": 2.4111785888671875, "learning_rate": 1.5758740751487016e-05, "loss": 0.0644, "step": 5947 }, { "epoch": 0.6448395490026019, "grad_norm": 1.3703712224960327, "learning_rate": 1.575801537791963e-05, "loss": 0.033, "step": 5948 }, { "epoch": 0.6449479618386817, "grad_norm": 1.6152417659759521, "learning_rate": 1.5757290004352243e-05, "loss": 0.0109, "step": 5949 }, { "epoch": 0.6450563746747615, "grad_norm": 1.3383854627609253, "learning_rate": 1.5756564630784856e-05, "loss": 0.03, "step": 5950 }, { "epoch": 0.6451647875108413, "grad_norm": 1.4642120599746704, "learning_rate": 1.5755839257217468e-05, "loss": 0.0282, "step": 5951 }, { "epoch": 0.645273200346921, "grad_norm": 2.6809372901916504, "learning_rate": 1.5755113883650083e-05, "loss": 0.0943, "step": 5952 }, { "epoch": 0.6453816131830009, "grad_norm": 1.4838814735412598, "learning_rate": 1.5754388510082695e-05, "loss": 0.0324, "step": 5953 }, { "epoch": 0.6454900260190807, "grad_norm": 2.5071980953216553, "learning_rate": 1.5753663136515307e-05, "loss": 0.0603, "step": 5954 }, { "epoch": 0.6455984388551604, "grad_norm": 2.817307233810425, "learning_rate": 1.575293776294792e-05, "loss": 0.0729, "step": 5955 }, { "epoch": 0.6457068516912402, "grad_norm": 3.4921443462371826, "learning_rate": 1.5752212389380532e-05, "loss": 0.0229, "step": 5956 }, { "epoch": 0.6458152645273201, "grad_norm": 5.895246505737305, "learning_rate": 1.5751487015813144e-05, "loss": 0.0504, "step": 5957 }, { "epoch": 0.6459236773633998, "grad_norm": 4.3651838302612305, "learning_rate": 1.5750761642245756e-05, "loss": 0.053, "step": 5958 }, { "epoch": 0.6460320901994796, "grad_norm": 3.638519287109375, "learning_rate": 1.575003626867837e-05, "loss": 0.0527, "step": 5959 }, { "epoch": 0.6461405030355594, "grad_norm": 1.6841723918914795, "learning_rate": 1.5749310895110984e-05, "loss": 0.0319, "step": 5960 }, { "epoch": 0.6462489158716392, "grad_norm": 1.5522353649139404, "learning_rate": 1.5748585521543596e-05, "loss": 0.0177, "step": 5961 }, { "epoch": 0.646357328707719, "grad_norm": 1.3014503717422485, "learning_rate": 1.5747860147976208e-05, "loss": 0.0265, "step": 5962 }, { "epoch": 0.6464657415437988, "grad_norm": 1.3830866813659668, "learning_rate": 1.574713477440882e-05, "loss": 0.0364, "step": 5963 }, { "epoch": 0.6465741543798785, "grad_norm": 3.4549810886383057, "learning_rate": 1.5746409400841432e-05, "loss": 0.0556, "step": 5964 }, { "epoch": 0.6466825672159584, "grad_norm": 2.865199327468872, "learning_rate": 1.5745684027274048e-05, "loss": 0.0387, "step": 5965 }, { "epoch": 0.6467909800520382, "grad_norm": 1.219325304031372, "learning_rate": 1.574495865370666e-05, "loss": 0.0267, "step": 5966 }, { "epoch": 0.6468993928881179, "grad_norm": 3.545252561569214, "learning_rate": 1.5744233280139275e-05, "loss": 0.0335, "step": 5967 }, { "epoch": 0.6470078057241977, "grad_norm": 4.2633056640625, "learning_rate": 1.5743507906571888e-05, "loss": 0.0414, "step": 5968 }, { "epoch": 0.6471162185602776, "grad_norm": 1.2061173915863037, "learning_rate": 1.57427825330045e-05, "loss": 0.0243, "step": 5969 }, { "epoch": 0.6472246313963573, "grad_norm": 5.634023189544678, "learning_rate": 1.5742057159437112e-05, "loss": 0.0236, "step": 5970 }, { "epoch": 0.6473330442324371, "grad_norm": 3.0889527797698975, "learning_rate": 1.5741331785869724e-05, "loss": 0.0232, "step": 5971 }, { "epoch": 0.647441457068517, "grad_norm": 0.9068775773048401, "learning_rate": 1.5740606412302336e-05, "loss": 0.0152, "step": 5972 }, { "epoch": 0.6475498699045967, "grad_norm": 1.8677130937576294, "learning_rate": 1.573988103873495e-05, "loss": 0.0232, "step": 5973 }, { "epoch": 0.6476582827406765, "grad_norm": 3.1371824741363525, "learning_rate": 1.5739155665167564e-05, "loss": 0.0548, "step": 5974 }, { "epoch": 0.6477666955767563, "grad_norm": 0.7156879901885986, "learning_rate": 1.5738430291600176e-05, "loss": 0.007, "step": 5975 }, { "epoch": 0.6478751084128361, "grad_norm": 0.7197954058647156, "learning_rate": 1.5737704918032788e-05, "loss": 0.0203, "step": 5976 }, { "epoch": 0.6479835212489159, "grad_norm": 2.055872917175293, "learning_rate": 1.57369795444654e-05, "loss": 0.026, "step": 5977 }, { "epoch": 0.6480919340849957, "grad_norm": 4.159026145935059, "learning_rate": 1.5736254170898012e-05, "loss": 0.0433, "step": 5978 }, { "epoch": 0.6482003469210754, "grad_norm": 0.3154005706310272, "learning_rate": 1.5735528797330625e-05, "loss": 0.002, "step": 5979 }, { "epoch": 0.6483087597571553, "grad_norm": 1.3792383670806885, "learning_rate": 1.573480342376324e-05, "loss": 0.0145, "step": 5980 }, { "epoch": 0.6484171725932351, "grad_norm": 0.42852163314819336, "learning_rate": 1.5734078050195852e-05, "loss": 0.0105, "step": 5981 }, { "epoch": 0.6485255854293148, "grad_norm": 1.8593448400497437, "learning_rate": 1.5733352676628464e-05, "loss": 0.0476, "step": 5982 }, { "epoch": 0.6486339982653946, "grad_norm": 0.6947349905967712, "learning_rate": 1.573262730306108e-05, "loss": 0.0086, "step": 5983 }, { "epoch": 0.6487424111014745, "grad_norm": 3.443268299102783, "learning_rate": 1.5731901929493692e-05, "loss": 0.0445, "step": 5984 }, { "epoch": 0.6488508239375542, "grad_norm": 2.4639370441436768, "learning_rate": 1.5731176555926304e-05, "loss": 0.016, "step": 5985 }, { "epoch": 0.648959236773634, "grad_norm": 4.649466037750244, "learning_rate": 1.5730451182358916e-05, "loss": 0.0317, "step": 5986 }, { "epoch": 0.6490676496097137, "grad_norm": 4.51639461517334, "learning_rate": 1.572972580879153e-05, "loss": 0.0509, "step": 5987 }, { "epoch": 0.6491760624457936, "grad_norm": 4.988572597503662, "learning_rate": 1.572900043522414e-05, "loss": 0.0408, "step": 5988 }, { "epoch": 0.6492844752818734, "grad_norm": 1.941240668296814, "learning_rate": 1.5728275061656756e-05, "loss": 0.0423, "step": 5989 }, { "epoch": 0.6493928881179531, "grad_norm": 1.8362822532653809, "learning_rate": 1.572754968808937e-05, "loss": 0.0116, "step": 5990 }, { "epoch": 0.6495013009540329, "grad_norm": 2.4355852603912354, "learning_rate": 1.572682431452198e-05, "loss": 0.0382, "step": 5991 }, { "epoch": 0.6496097137901128, "grad_norm": 4.0969157218933105, "learning_rate": 1.5726098940954593e-05, "loss": 0.0388, "step": 5992 }, { "epoch": 0.6497181266261925, "grad_norm": 0.37396103143692017, "learning_rate": 1.5725373567387205e-05, "loss": 0.0061, "step": 5993 }, { "epoch": 0.6498265394622723, "grad_norm": 3.9596126079559326, "learning_rate": 1.5724648193819817e-05, "loss": 0.068, "step": 5994 }, { "epoch": 0.6499349522983521, "grad_norm": 4.833179473876953, "learning_rate": 1.5723922820252432e-05, "loss": 0.0566, "step": 5995 }, { "epoch": 0.650043365134432, "grad_norm": 2.065615177154541, "learning_rate": 1.5723197446685045e-05, "loss": 0.0317, "step": 5996 }, { "epoch": 0.6501517779705117, "grad_norm": 1.7136813402175903, "learning_rate": 1.5722472073117657e-05, "loss": 0.0683, "step": 5997 }, { "epoch": 0.6502601908065915, "grad_norm": 0.4726681411266327, "learning_rate": 1.572174669955027e-05, "loss": 0.0093, "step": 5998 }, { "epoch": 0.6503686036426712, "grad_norm": 1.8381164073944092, "learning_rate": 1.572102132598288e-05, "loss": 0.0371, "step": 5999 }, { "epoch": 0.6504770164787511, "grad_norm": 3.163257122039795, "learning_rate": 1.5720295952415497e-05, "loss": 0.0758, "step": 6000 }, { "epoch": 0.6505854293148309, "grad_norm": 0.6785373687744141, "learning_rate": 1.571957057884811e-05, "loss": 0.0127, "step": 6001 }, { "epoch": 0.6506938421509106, "grad_norm": 2.5879600048065186, "learning_rate": 1.571884520528072e-05, "loss": 0.039, "step": 6002 }, { "epoch": 0.6508022549869905, "grad_norm": 1.9473743438720703, "learning_rate": 1.5718119831713333e-05, "loss": 0.0303, "step": 6003 }, { "epoch": 0.6509106678230703, "grad_norm": 1.0142935514450073, "learning_rate": 1.571739445814595e-05, "loss": 0.0273, "step": 6004 }, { "epoch": 0.65101908065915, "grad_norm": 2.3526992797851562, "learning_rate": 1.571666908457856e-05, "loss": 0.013, "step": 6005 }, { "epoch": 0.6511274934952298, "grad_norm": 1.9215635061264038, "learning_rate": 1.5715943711011173e-05, "loss": 0.0293, "step": 6006 }, { "epoch": 0.6512359063313097, "grad_norm": 2.8821260929107666, "learning_rate": 1.5715218337443785e-05, "loss": 0.0463, "step": 6007 }, { "epoch": 0.6513443191673894, "grad_norm": 2.049889326095581, "learning_rate": 1.5714492963876397e-05, "loss": 0.0189, "step": 6008 }, { "epoch": 0.6514527320034692, "grad_norm": 1.7232017517089844, "learning_rate": 1.571376759030901e-05, "loss": 0.0192, "step": 6009 }, { "epoch": 0.651561144839549, "grad_norm": 0.43538182973861694, "learning_rate": 1.571304221674162e-05, "loss": 0.0065, "step": 6010 }, { "epoch": 0.6516695576756288, "grad_norm": 0.6592133045196533, "learning_rate": 1.5712316843174237e-05, "loss": 0.0069, "step": 6011 }, { "epoch": 0.6517779705117086, "grad_norm": 2.8327410221099854, "learning_rate": 1.571159146960685e-05, "loss": 0.0425, "step": 6012 }, { "epoch": 0.6518863833477884, "grad_norm": 2.7674694061279297, "learning_rate": 1.571086609603946e-05, "loss": 0.0395, "step": 6013 }, { "epoch": 0.6519947961838681, "grad_norm": 2.3710763454437256, "learning_rate": 1.5710140722472073e-05, "loss": 0.0379, "step": 6014 }, { "epoch": 0.652103209019948, "grad_norm": 2.1703827381134033, "learning_rate": 1.5709415348904685e-05, "loss": 0.0366, "step": 6015 }, { "epoch": 0.6522116218560278, "grad_norm": 2.9270620346069336, "learning_rate": 1.5708689975337298e-05, "loss": 0.0865, "step": 6016 }, { "epoch": 0.6523200346921075, "grad_norm": 0.8968841433525085, "learning_rate": 1.5707964601769913e-05, "loss": 0.0126, "step": 6017 }, { "epoch": 0.6524284475281873, "grad_norm": 3.9269864559173584, "learning_rate": 1.5707239228202525e-05, "loss": 0.065, "step": 6018 }, { "epoch": 0.6525368603642672, "grad_norm": 2.6942272186279297, "learning_rate": 1.570651385463514e-05, "loss": 0.0312, "step": 6019 }, { "epoch": 0.6526452732003469, "grad_norm": 1.4657831192016602, "learning_rate": 1.5705788481067753e-05, "loss": 0.0446, "step": 6020 }, { "epoch": 0.6527536860364267, "grad_norm": 1.8410125970840454, "learning_rate": 1.5705063107500365e-05, "loss": 0.0538, "step": 6021 }, { "epoch": 0.6528620988725065, "grad_norm": 2.034855604171753, "learning_rate": 1.5704337733932977e-05, "loss": 0.0573, "step": 6022 }, { "epoch": 0.6529705117085863, "grad_norm": 2.3719046115875244, "learning_rate": 1.570361236036559e-05, "loss": 0.0519, "step": 6023 }, { "epoch": 0.6530789245446661, "grad_norm": 1.8015151023864746, "learning_rate": 1.57028869867982e-05, "loss": 0.0209, "step": 6024 }, { "epoch": 0.6531873373807459, "grad_norm": 3.306574821472168, "learning_rate": 1.5702161613230814e-05, "loss": 0.0354, "step": 6025 }, { "epoch": 0.6532957502168256, "grad_norm": 2.046844959259033, "learning_rate": 1.570143623966343e-05, "loss": 0.037, "step": 6026 }, { "epoch": 0.6534041630529055, "grad_norm": 2.8609044551849365, "learning_rate": 1.570071086609604e-05, "loss": 0.0704, "step": 6027 }, { "epoch": 0.6535125758889853, "grad_norm": 1.6179282665252686, "learning_rate": 1.5699985492528654e-05, "loss": 0.039, "step": 6028 }, { "epoch": 0.653620988725065, "grad_norm": 0.8576993942260742, "learning_rate": 1.5699260118961266e-05, "loss": 0.0162, "step": 6029 }, { "epoch": 0.6537294015611448, "grad_norm": 0.9768182635307312, "learning_rate": 1.5698534745393878e-05, "loss": 0.0234, "step": 6030 }, { "epoch": 0.6538378143972247, "grad_norm": 1.3298115730285645, "learning_rate": 1.569780937182649e-05, "loss": 0.0233, "step": 6031 }, { "epoch": 0.6539462272333044, "grad_norm": 4.078464031219482, "learning_rate": 1.5697083998259105e-05, "loss": 0.0508, "step": 6032 }, { "epoch": 0.6540546400693842, "grad_norm": 2.3410847187042236, "learning_rate": 1.5696358624691718e-05, "loss": 0.0344, "step": 6033 }, { "epoch": 0.6541630529054641, "grad_norm": 1.545935869216919, "learning_rate": 1.569563325112433e-05, "loss": 0.0316, "step": 6034 }, { "epoch": 0.6542714657415438, "grad_norm": 5.163835525512695, "learning_rate": 1.5694907877556945e-05, "loss": 0.0561, "step": 6035 }, { "epoch": 0.6543798785776236, "grad_norm": 1.1565595865249634, "learning_rate": 1.5694182503989557e-05, "loss": 0.0205, "step": 6036 }, { "epoch": 0.6544882914137033, "grad_norm": 6.9456281661987305, "learning_rate": 1.569345713042217e-05, "loss": 0.0454, "step": 6037 }, { "epoch": 0.6545967042497832, "grad_norm": 1.2602839469909668, "learning_rate": 1.5692731756854782e-05, "loss": 0.0425, "step": 6038 }, { "epoch": 0.654705117085863, "grad_norm": 1.340404987335205, "learning_rate": 1.5692006383287394e-05, "loss": 0.0219, "step": 6039 }, { "epoch": 0.6548135299219427, "grad_norm": 0.981153130531311, "learning_rate": 1.5691281009720006e-05, "loss": 0.0151, "step": 6040 }, { "epoch": 0.6549219427580225, "grad_norm": 6.158132076263428, "learning_rate": 1.569055563615262e-05, "loss": 0.0402, "step": 6041 }, { "epoch": 0.6550303555941024, "grad_norm": 2.061732053756714, "learning_rate": 1.5689830262585234e-05, "loss": 0.0444, "step": 6042 }, { "epoch": 0.6551387684301822, "grad_norm": 5.114979267120361, "learning_rate": 1.5689104889017846e-05, "loss": 0.1161, "step": 6043 }, { "epoch": 0.6552471812662619, "grad_norm": 4.614415168762207, "learning_rate": 1.5688379515450458e-05, "loss": 0.0381, "step": 6044 }, { "epoch": 0.6553555941023417, "grad_norm": 1.4585378170013428, "learning_rate": 1.568765414188307e-05, "loss": 0.0112, "step": 6045 }, { "epoch": 0.6554640069384216, "grad_norm": 1.400755524635315, "learning_rate": 1.5686928768315682e-05, "loss": 0.0285, "step": 6046 }, { "epoch": 0.6555724197745013, "grad_norm": 2.321930408477783, "learning_rate": 1.5686203394748298e-05, "loss": 0.0442, "step": 6047 }, { "epoch": 0.6556808326105811, "grad_norm": 3.311962842941284, "learning_rate": 1.568547802118091e-05, "loss": 0.0478, "step": 6048 }, { "epoch": 0.6557892454466608, "grad_norm": 1.1430166959762573, "learning_rate": 1.5684752647613522e-05, "loss": 0.036, "step": 6049 }, { "epoch": 0.6558976582827407, "grad_norm": 0.9158557653427124, "learning_rate": 1.5684027274046134e-05, "loss": 0.0352, "step": 6050 }, { "epoch": 0.6560060711188205, "grad_norm": 1.9901939630508423, "learning_rate": 1.5683301900478746e-05, "loss": 0.0156, "step": 6051 }, { "epoch": 0.6561144839549002, "grad_norm": 1.0629398822784424, "learning_rate": 1.5682576526911362e-05, "loss": 0.0117, "step": 6052 }, { "epoch": 0.65622289679098, "grad_norm": 0.6594254374504089, "learning_rate": 1.5681851153343974e-05, "loss": 0.0063, "step": 6053 }, { "epoch": 0.6563313096270599, "grad_norm": 1.156010627746582, "learning_rate": 1.5681125779776586e-05, "loss": 0.0028, "step": 6054 }, { "epoch": 0.6564397224631396, "grad_norm": 0.8538925051689148, "learning_rate": 1.56804004062092e-05, "loss": 0.0138, "step": 6055 }, { "epoch": 0.6565481352992194, "grad_norm": 3.228797435760498, "learning_rate": 1.5679675032641814e-05, "loss": 0.0321, "step": 6056 }, { "epoch": 0.6566565481352992, "grad_norm": 5.057584762573242, "learning_rate": 1.5678949659074426e-05, "loss": 0.0472, "step": 6057 }, { "epoch": 0.656764960971379, "grad_norm": 2.1523234844207764, "learning_rate": 1.5678224285507038e-05, "loss": 0.0345, "step": 6058 }, { "epoch": 0.6568733738074588, "grad_norm": 5.364363670349121, "learning_rate": 1.567749891193965e-05, "loss": 0.0183, "step": 6059 }, { "epoch": 0.6569817866435386, "grad_norm": 3.5007240772247314, "learning_rate": 1.5676773538372262e-05, "loss": 0.0453, "step": 6060 }, { "epoch": 0.6570901994796184, "grad_norm": 1.0565353631973267, "learning_rate": 1.5676048164804875e-05, "loss": 0.0135, "step": 6061 }, { "epoch": 0.6571986123156982, "grad_norm": 2.7892611026763916, "learning_rate": 1.5675322791237487e-05, "loss": 0.0423, "step": 6062 }, { "epoch": 0.657307025151778, "grad_norm": 2.6959187984466553, "learning_rate": 1.5674597417670102e-05, "loss": 0.0249, "step": 6063 }, { "epoch": 0.6574154379878577, "grad_norm": 1.8285080194473267, "learning_rate": 1.5673872044102714e-05, "loss": 0.008, "step": 6064 }, { "epoch": 0.6575238508239376, "grad_norm": 11.506444931030273, "learning_rate": 1.5673146670535327e-05, "loss": 0.0494, "step": 6065 }, { "epoch": 0.6576322636600174, "grad_norm": 1.0475815534591675, "learning_rate": 1.567242129696794e-05, "loss": 0.0272, "step": 6066 }, { "epoch": 0.6577406764960971, "grad_norm": 3.363004207611084, "learning_rate": 1.567169592340055e-05, "loss": 0.0262, "step": 6067 }, { "epoch": 0.6578490893321769, "grad_norm": 3.298848867416382, "learning_rate": 1.5670970549833163e-05, "loss": 0.0446, "step": 6068 }, { "epoch": 0.6579575021682568, "grad_norm": 2.3260178565979004, "learning_rate": 1.567024517626578e-05, "loss": 0.0743, "step": 6069 }, { "epoch": 0.6580659150043365, "grad_norm": 1.8746765851974487, "learning_rate": 1.566951980269839e-05, "loss": 0.0294, "step": 6070 }, { "epoch": 0.6581743278404163, "grad_norm": 3.593092441558838, "learning_rate": 1.5668794429131006e-05, "loss": 0.0543, "step": 6071 }, { "epoch": 0.6582827406764961, "grad_norm": 6.115507125854492, "learning_rate": 1.566806905556362e-05, "loss": 0.0289, "step": 6072 }, { "epoch": 0.6583911535125759, "grad_norm": 0.5517330169677734, "learning_rate": 1.566734368199623e-05, "loss": 0.0143, "step": 6073 }, { "epoch": 0.6584995663486557, "grad_norm": 3.9617226123809814, "learning_rate": 1.5666618308428843e-05, "loss": 0.017, "step": 6074 }, { "epoch": 0.6586079791847355, "grad_norm": 0.5178637504577637, "learning_rate": 1.5665892934861455e-05, "loss": 0.0109, "step": 6075 }, { "epoch": 0.6587163920208152, "grad_norm": 2.7892520427703857, "learning_rate": 1.5665167561294067e-05, "loss": 0.0593, "step": 6076 }, { "epoch": 0.6588248048568951, "grad_norm": 4.444916725158691, "learning_rate": 1.566444218772668e-05, "loss": 0.0101, "step": 6077 }, { "epoch": 0.6589332176929749, "grad_norm": 7.749855995178223, "learning_rate": 1.5663716814159295e-05, "loss": 0.0363, "step": 6078 }, { "epoch": 0.6590416305290546, "grad_norm": 3.449888229370117, "learning_rate": 1.5662991440591907e-05, "loss": 0.0417, "step": 6079 }, { "epoch": 0.6591500433651344, "grad_norm": 7.817523002624512, "learning_rate": 1.566226606702452e-05, "loss": 0.0418, "step": 6080 }, { "epoch": 0.6592584562012143, "grad_norm": 2.637044906616211, "learning_rate": 1.566154069345713e-05, "loss": 0.0515, "step": 6081 }, { "epoch": 0.659366869037294, "grad_norm": 0.8055978417396545, "learning_rate": 1.5660815319889743e-05, "loss": 0.0158, "step": 6082 }, { "epoch": 0.6594752818733738, "grad_norm": 2.1453447341918945, "learning_rate": 1.5660089946322355e-05, "loss": 0.04, "step": 6083 }, { "epoch": 0.6595836947094535, "grad_norm": 5.4625983238220215, "learning_rate": 1.565936457275497e-05, "loss": 0.0381, "step": 6084 }, { "epoch": 0.6596921075455334, "grad_norm": 1.3946877717971802, "learning_rate": 1.5658639199187583e-05, "loss": 0.0136, "step": 6085 }, { "epoch": 0.6598005203816132, "grad_norm": 4.643239498138428, "learning_rate": 1.5657913825620195e-05, "loss": 0.0178, "step": 6086 }, { "epoch": 0.659908933217693, "grad_norm": 4.312741279602051, "learning_rate": 1.565718845205281e-05, "loss": 0.0148, "step": 6087 }, { "epoch": 0.6600173460537727, "grad_norm": 5.377866744995117, "learning_rate": 1.5656463078485423e-05, "loss": 0.0834, "step": 6088 }, { "epoch": 0.6601257588898526, "grad_norm": 0.8816066384315491, "learning_rate": 1.5655737704918035e-05, "loss": 0.014, "step": 6089 }, { "epoch": 0.6602341717259324, "grad_norm": 3.696803569793701, "learning_rate": 1.5655012331350647e-05, "loss": 0.0332, "step": 6090 }, { "epoch": 0.6603425845620121, "grad_norm": 6.998088359832764, "learning_rate": 1.565428695778326e-05, "loss": 0.066, "step": 6091 }, { "epoch": 0.660450997398092, "grad_norm": 1.0348610877990723, "learning_rate": 1.565356158421587e-05, "loss": 0.0174, "step": 6092 }, { "epoch": 0.6605594102341718, "grad_norm": 0.9343259334564209, "learning_rate": 1.5652836210648487e-05, "loss": 0.0215, "step": 6093 }, { "epoch": 0.6606678230702515, "grad_norm": 3.598690986633301, "learning_rate": 1.56521108370811e-05, "loss": 0.0626, "step": 6094 }, { "epoch": 0.6607762359063313, "grad_norm": 1.2536457777023315, "learning_rate": 1.565138546351371e-05, "loss": 0.0127, "step": 6095 }, { "epoch": 0.6608846487424112, "grad_norm": 2.1339352130889893, "learning_rate": 1.5650660089946323e-05, "loss": 0.0207, "step": 6096 }, { "epoch": 0.6609930615784909, "grad_norm": 2.7857508659362793, "learning_rate": 1.5649934716378935e-05, "loss": 0.1022, "step": 6097 }, { "epoch": 0.6611014744145707, "grad_norm": 1.147702693939209, "learning_rate": 1.5649209342811548e-05, "loss": 0.0349, "step": 6098 }, { "epoch": 0.6612098872506504, "grad_norm": 9.183506965637207, "learning_rate": 1.5648483969244163e-05, "loss": 0.0833, "step": 6099 }, { "epoch": 0.6613183000867303, "grad_norm": 2.436314582824707, "learning_rate": 1.5647758595676775e-05, "loss": 0.0271, "step": 6100 }, { "epoch": 0.6614267129228101, "grad_norm": 1.109975814819336, "learning_rate": 1.5647033222109387e-05, "loss": 0.0156, "step": 6101 }, { "epoch": 0.6615351257588898, "grad_norm": 5.839962482452393, "learning_rate": 1.5646307848542e-05, "loss": 0.1035, "step": 6102 }, { "epoch": 0.6616435385949696, "grad_norm": 0.3158123791217804, "learning_rate": 1.5645582474974612e-05, "loss": 0.007, "step": 6103 }, { "epoch": 0.6617519514310495, "grad_norm": 1.9646739959716797, "learning_rate": 1.5644857101407227e-05, "loss": 0.0254, "step": 6104 }, { "epoch": 0.6618603642671292, "grad_norm": 4.0150604248046875, "learning_rate": 1.564413172783984e-05, "loss": 0.046, "step": 6105 }, { "epoch": 0.661968777103209, "grad_norm": 4.177856922149658, "learning_rate": 1.564340635427245e-05, "loss": 0.087, "step": 6106 }, { "epoch": 0.6620771899392888, "grad_norm": 3.0303707122802734, "learning_rate": 1.5642680980705064e-05, "loss": 0.0221, "step": 6107 }, { "epoch": 0.6621856027753686, "grad_norm": 2.8637592792510986, "learning_rate": 1.564195560713768e-05, "loss": 0.0253, "step": 6108 }, { "epoch": 0.6622940156114484, "grad_norm": 3.4921817779541016, "learning_rate": 1.564123023357029e-05, "loss": 0.0282, "step": 6109 }, { "epoch": 0.6624024284475282, "grad_norm": 8.601495742797852, "learning_rate": 1.5640504860002903e-05, "loss": 0.0618, "step": 6110 }, { "epoch": 0.6625108412836079, "grad_norm": 4.199110507965088, "learning_rate": 1.5639779486435516e-05, "loss": 0.0298, "step": 6111 }, { "epoch": 0.6626192541196878, "grad_norm": 2.5324389934539795, "learning_rate": 1.5639054112868128e-05, "loss": 0.0332, "step": 6112 }, { "epoch": 0.6627276669557676, "grad_norm": 0.06557299196720123, "learning_rate": 1.563832873930074e-05, "loss": 0.0006, "step": 6113 }, { "epoch": 0.6628360797918473, "grad_norm": 3.098504066467285, "learning_rate": 1.5637603365733352e-05, "loss": 0.0521, "step": 6114 }, { "epoch": 0.6629444926279271, "grad_norm": 3.8935225009918213, "learning_rate": 1.5636877992165968e-05, "loss": 0.068, "step": 6115 }, { "epoch": 0.663052905464007, "grad_norm": 3.7430853843688965, "learning_rate": 1.563615261859858e-05, "loss": 0.0619, "step": 6116 }, { "epoch": 0.6631613183000867, "grad_norm": 1.7156527042388916, "learning_rate": 1.5635427245031192e-05, "loss": 0.0403, "step": 6117 }, { "epoch": 0.6632697311361665, "grad_norm": 2.3996176719665527, "learning_rate": 1.5634701871463804e-05, "loss": 0.0237, "step": 6118 }, { "epoch": 0.6633781439722463, "grad_norm": 2.4509530067443848, "learning_rate": 1.5633976497896416e-05, "loss": 0.0356, "step": 6119 }, { "epoch": 0.6634865568083261, "grad_norm": 5.108160495758057, "learning_rate": 1.5633251124329028e-05, "loss": 0.0712, "step": 6120 }, { "epoch": 0.6635949696444059, "grad_norm": 3.9625511169433594, "learning_rate": 1.5632525750761644e-05, "loss": 0.0821, "step": 6121 }, { "epoch": 0.6637033824804857, "grad_norm": 0.5567903518676758, "learning_rate": 1.5631800377194256e-05, "loss": 0.0057, "step": 6122 }, { "epoch": 0.6638117953165655, "grad_norm": 2.063002347946167, "learning_rate": 1.563107500362687e-05, "loss": 0.0406, "step": 6123 }, { "epoch": 0.6639202081526453, "grad_norm": 2.876030206680298, "learning_rate": 1.5630349630059484e-05, "loss": 0.0497, "step": 6124 }, { "epoch": 0.6640286209887251, "grad_norm": 5.8599066734313965, "learning_rate": 1.5629624256492096e-05, "loss": 0.0617, "step": 6125 }, { "epoch": 0.6641370338248048, "grad_norm": 0.9338613748550415, "learning_rate": 1.5628898882924708e-05, "loss": 0.0151, "step": 6126 }, { "epoch": 0.6642454466608847, "grad_norm": 2.008817434310913, "learning_rate": 1.562817350935732e-05, "loss": 0.0422, "step": 6127 }, { "epoch": 0.6643538594969645, "grad_norm": 3.0607151985168457, "learning_rate": 1.5627448135789932e-05, "loss": 0.0333, "step": 6128 }, { "epoch": 0.6644622723330442, "grad_norm": 1.8786921501159668, "learning_rate": 1.5626722762222544e-05, "loss": 0.0531, "step": 6129 }, { "epoch": 0.664570685169124, "grad_norm": 5.95394229888916, "learning_rate": 1.562599738865516e-05, "loss": 0.073, "step": 6130 }, { "epoch": 0.6646790980052039, "grad_norm": 5.58281135559082, "learning_rate": 1.5625272015087772e-05, "loss": 0.064, "step": 6131 }, { "epoch": 0.6647875108412836, "grad_norm": 2.0788300037384033, "learning_rate": 1.5624546641520384e-05, "loss": 0.0187, "step": 6132 }, { "epoch": 0.6648959236773634, "grad_norm": 3.541757583618164, "learning_rate": 1.5623821267952996e-05, "loss": 0.0479, "step": 6133 }, { "epoch": 0.6650043365134432, "grad_norm": 2.3463521003723145, "learning_rate": 1.562309589438561e-05, "loss": 0.0194, "step": 6134 }, { "epoch": 0.665112749349523, "grad_norm": 1.9829082489013672, "learning_rate": 1.562237052081822e-05, "loss": 0.0423, "step": 6135 }, { "epoch": 0.6652211621856028, "grad_norm": 6.709456920623779, "learning_rate": 1.5621645147250836e-05, "loss": 0.0447, "step": 6136 }, { "epoch": 0.6653295750216826, "grad_norm": 1.4655297994613647, "learning_rate": 1.5620919773683448e-05, "loss": 0.0195, "step": 6137 }, { "epoch": 0.6654379878577623, "grad_norm": 2.7704975605010986, "learning_rate": 1.5620194400116064e-05, "loss": 0.0487, "step": 6138 }, { "epoch": 0.6655464006938422, "grad_norm": 0.981963574886322, "learning_rate": 1.5619469026548676e-05, "loss": 0.0124, "step": 6139 }, { "epoch": 0.665654813529922, "grad_norm": 2.1062214374542236, "learning_rate": 1.5618743652981288e-05, "loss": 0.0308, "step": 6140 }, { "epoch": 0.6657632263660017, "grad_norm": 1.297222375869751, "learning_rate": 1.56180182794139e-05, "loss": 0.0164, "step": 6141 }, { "epoch": 0.6658716392020815, "grad_norm": 6.150179862976074, "learning_rate": 1.5617292905846512e-05, "loss": 0.1316, "step": 6142 }, { "epoch": 0.6659800520381614, "grad_norm": 4.016442775726318, "learning_rate": 1.5616567532279125e-05, "loss": 0.049, "step": 6143 }, { "epoch": 0.6660884648742411, "grad_norm": 0.5254237055778503, "learning_rate": 1.5615842158711737e-05, "loss": 0.0168, "step": 6144 }, { "epoch": 0.6661968777103209, "grad_norm": 1.4566900730133057, "learning_rate": 1.5615116785144352e-05, "loss": 0.0217, "step": 6145 }, { "epoch": 0.6663052905464006, "grad_norm": 4.249384880065918, "learning_rate": 1.5614391411576964e-05, "loss": 0.0673, "step": 6146 }, { "epoch": 0.6664137033824805, "grad_norm": 3.172719717025757, "learning_rate": 1.5613666038009577e-05, "loss": 0.0468, "step": 6147 }, { "epoch": 0.6665221162185603, "grad_norm": 2.7495148181915283, "learning_rate": 1.561294066444219e-05, "loss": 0.0551, "step": 6148 }, { "epoch": 0.66663052905464, "grad_norm": 2.361668586730957, "learning_rate": 1.56122152908748e-05, "loss": 0.0199, "step": 6149 }, { "epoch": 0.6667389418907199, "grad_norm": 1.8843101263046265, "learning_rate": 1.5611489917307413e-05, "loss": 0.0339, "step": 6150 }, { "epoch": 0.6668473547267997, "grad_norm": 1.5984110832214355, "learning_rate": 1.561076454374003e-05, "loss": 0.0413, "step": 6151 }, { "epoch": 0.6669557675628794, "grad_norm": 4.828417778015137, "learning_rate": 1.561003917017264e-05, "loss": 0.0174, "step": 6152 }, { "epoch": 0.6670641803989592, "grad_norm": 1.4481583833694458, "learning_rate": 1.5609313796605253e-05, "loss": 0.0384, "step": 6153 }, { "epoch": 0.6671725932350391, "grad_norm": 7.573367595672607, "learning_rate": 1.5608588423037865e-05, "loss": 0.0566, "step": 6154 }, { "epoch": 0.6672810060711188, "grad_norm": 1.9200265407562256, "learning_rate": 1.5607863049470477e-05, "loss": 0.0324, "step": 6155 }, { "epoch": 0.6673894189071986, "grad_norm": 2.572563409805298, "learning_rate": 1.5607137675903093e-05, "loss": 0.0351, "step": 6156 }, { "epoch": 0.6674978317432784, "grad_norm": 0.8120085000991821, "learning_rate": 1.5606412302335705e-05, "loss": 0.0296, "step": 6157 }, { "epoch": 0.6676062445793582, "grad_norm": 2.717172145843506, "learning_rate": 1.5605686928768317e-05, "loss": 0.0523, "step": 6158 }, { "epoch": 0.667714657415438, "grad_norm": 2.3729631900787354, "learning_rate": 1.560496155520093e-05, "loss": 0.0275, "step": 6159 }, { "epoch": 0.6678230702515178, "grad_norm": 0.6350535154342651, "learning_rate": 1.5604236181633545e-05, "loss": 0.008, "step": 6160 }, { "epoch": 0.6679314830875975, "grad_norm": 1.5624208450317383, "learning_rate": 1.5603510808066157e-05, "loss": 0.015, "step": 6161 }, { "epoch": 0.6680398959236774, "grad_norm": 1.2105015516281128, "learning_rate": 1.560278543449877e-05, "loss": 0.036, "step": 6162 }, { "epoch": 0.6681483087597572, "grad_norm": 1.6121692657470703, "learning_rate": 1.560206006093138e-05, "loss": 0.0428, "step": 6163 }, { "epoch": 0.6682567215958369, "grad_norm": 6.789789199829102, "learning_rate": 1.5601334687363993e-05, "loss": 0.0528, "step": 6164 }, { "epoch": 0.6683651344319167, "grad_norm": 2.8087246417999268, "learning_rate": 1.5600609313796605e-05, "loss": 0.0562, "step": 6165 }, { "epoch": 0.6684735472679966, "grad_norm": 1.5919727087020874, "learning_rate": 1.5599883940229217e-05, "loss": 0.0194, "step": 6166 }, { "epoch": 0.6685819601040763, "grad_norm": 2.498699903488159, "learning_rate": 1.5599158566661833e-05, "loss": 0.0164, "step": 6167 }, { "epoch": 0.6686903729401561, "grad_norm": 4.812344551086426, "learning_rate": 1.5598433193094445e-05, "loss": 0.0497, "step": 6168 }, { "epoch": 0.6687987857762359, "grad_norm": 3.0887086391448975, "learning_rate": 1.5597707819527057e-05, "loss": 0.0644, "step": 6169 }, { "epoch": 0.6689071986123157, "grad_norm": 4.956231117248535, "learning_rate": 1.559698244595967e-05, "loss": 0.0881, "step": 6170 }, { "epoch": 0.6690156114483955, "grad_norm": 6.311890602111816, "learning_rate": 1.559625707239228e-05, "loss": 0.0523, "step": 6171 }, { "epoch": 0.6691240242844753, "grad_norm": 2.3880069255828857, "learning_rate": 1.5595531698824894e-05, "loss": 0.0579, "step": 6172 }, { "epoch": 0.669232437120555, "grad_norm": 1.577580451965332, "learning_rate": 1.559480632525751e-05, "loss": 0.0342, "step": 6173 }, { "epoch": 0.6693408499566349, "grad_norm": 4.682582378387451, "learning_rate": 1.559408095169012e-05, "loss": 0.0636, "step": 6174 }, { "epoch": 0.6694492627927147, "grad_norm": 3.2275843620300293, "learning_rate": 1.5593355578122737e-05, "loss": 0.0363, "step": 6175 }, { "epoch": 0.6695576756287944, "grad_norm": 3.1861908435821533, "learning_rate": 1.559263020455535e-05, "loss": 0.0663, "step": 6176 }, { "epoch": 0.6696660884648742, "grad_norm": 0.4741155505180359, "learning_rate": 1.559190483098796e-05, "loss": 0.0225, "step": 6177 }, { "epoch": 0.6697745013009541, "grad_norm": 1.659449577331543, "learning_rate": 1.5591179457420573e-05, "loss": 0.0282, "step": 6178 }, { "epoch": 0.6698829141370338, "grad_norm": 1.5188554525375366, "learning_rate": 1.5590454083853185e-05, "loss": 0.0158, "step": 6179 }, { "epoch": 0.6699913269731136, "grad_norm": 5.535720348358154, "learning_rate": 1.5589728710285798e-05, "loss": 0.0635, "step": 6180 }, { "epoch": 0.6700997398091935, "grad_norm": 1.2643651962280273, "learning_rate": 1.558900333671841e-05, "loss": 0.0283, "step": 6181 }, { "epoch": 0.6702081526452732, "grad_norm": 1.4339498281478882, "learning_rate": 1.5588277963151025e-05, "loss": 0.0118, "step": 6182 }, { "epoch": 0.670316565481353, "grad_norm": 2.0819528102874756, "learning_rate": 1.5587552589583637e-05, "loss": 0.0866, "step": 6183 }, { "epoch": 0.6704249783174328, "grad_norm": 2.226957321166992, "learning_rate": 1.558682721601625e-05, "loss": 0.0507, "step": 6184 }, { "epoch": 0.6705333911535126, "grad_norm": 3.034553289413452, "learning_rate": 1.558610184244886e-05, "loss": 0.0533, "step": 6185 }, { "epoch": 0.6706418039895924, "grad_norm": 1.0615822076797485, "learning_rate": 1.5585376468881474e-05, "loss": 0.0278, "step": 6186 }, { "epoch": 0.6707502168256722, "grad_norm": 1.968295931816101, "learning_rate": 1.5584651095314086e-05, "loss": 0.0489, "step": 6187 }, { "epoch": 0.6708586296617519, "grad_norm": 1.6071754693984985, "learning_rate": 1.55839257217467e-05, "loss": 0.0289, "step": 6188 }, { "epoch": 0.6709670424978318, "grad_norm": 2.1801226139068604, "learning_rate": 1.5583200348179314e-05, "loss": 0.054, "step": 6189 }, { "epoch": 0.6710754553339116, "grad_norm": 3.285607099533081, "learning_rate": 1.558247497461193e-05, "loss": 0.0265, "step": 6190 }, { "epoch": 0.6711838681699913, "grad_norm": 1.5661399364471436, "learning_rate": 1.558174960104454e-05, "loss": 0.0142, "step": 6191 }, { "epoch": 0.6712922810060711, "grad_norm": 0.7364878058433533, "learning_rate": 1.5581024227477153e-05, "loss": 0.0047, "step": 6192 }, { "epoch": 0.671400693842151, "grad_norm": 2.4162144660949707, "learning_rate": 1.5580298853909766e-05, "loss": 0.0202, "step": 6193 }, { "epoch": 0.6715091066782307, "grad_norm": 2.203852891921997, "learning_rate": 1.5579573480342378e-05, "loss": 0.0384, "step": 6194 }, { "epoch": 0.6716175195143105, "grad_norm": 5.217399597167969, "learning_rate": 1.557884810677499e-05, "loss": 0.0379, "step": 6195 }, { "epoch": 0.6717259323503902, "grad_norm": 0.7384206056594849, "learning_rate": 1.5578122733207602e-05, "loss": 0.0082, "step": 6196 }, { "epoch": 0.6718343451864701, "grad_norm": 2.5640175342559814, "learning_rate": 1.5577397359640218e-05, "loss": 0.0723, "step": 6197 }, { "epoch": 0.6719427580225499, "grad_norm": 3.0048325061798096, "learning_rate": 1.557667198607283e-05, "loss": 0.0376, "step": 6198 }, { "epoch": 0.6720511708586296, "grad_norm": 1.8478080034255981, "learning_rate": 1.5575946612505442e-05, "loss": 0.014, "step": 6199 }, { "epoch": 0.6721595836947094, "grad_norm": 3.448723554611206, "learning_rate": 1.5575221238938054e-05, "loss": 0.0472, "step": 6200 }, { "epoch": 0.6722679965307893, "grad_norm": 0.4005613327026367, "learning_rate": 1.5574495865370666e-05, "loss": 0.014, "step": 6201 }, { "epoch": 0.672376409366869, "grad_norm": 4.375622749328613, "learning_rate": 1.5573770491803278e-05, "loss": 0.0392, "step": 6202 }, { "epoch": 0.6724848222029488, "grad_norm": 3.131411075592041, "learning_rate": 1.5573045118235894e-05, "loss": 0.0342, "step": 6203 }, { "epoch": 0.6725932350390286, "grad_norm": 3.184441328048706, "learning_rate": 1.5572319744668506e-05, "loss": 0.0351, "step": 6204 }, { "epoch": 0.6727016478751084, "grad_norm": 3.5382394790649414, "learning_rate": 1.5571594371101118e-05, "loss": 0.0274, "step": 6205 }, { "epoch": 0.6728100607111882, "grad_norm": 2.1842658519744873, "learning_rate": 1.557086899753373e-05, "loss": 0.0691, "step": 6206 }, { "epoch": 0.672918473547268, "grad_norm": 2.481318473815918, "learning_rate": 1.5570143623966346e-05, "loss": 0.027, "step": 6207 }, { "epoch": 0.6730268863833477, "grad_norm": 3.3123738765716553, "learning_rate": 1.5569418250398958e-05, "loss": 0.062, "step": 6208 }, { "epoch": 0.6731352992194276, "grad_norm": 4.788657188415527, "learning_rate": 1.556869287683157e-05, "loss": 0.0756, "step": 6209 }, { "epoch": 0.6732437120555074, "grad_norm": 3.4607086181640625, "learning_rate": 1.5567967503264182e-05, "loss": 0.0759, "step": 6210 }, { "epoch": 0.6733521248915871, "grad_norm": 1.8967446088790894, "learning_rate": 1.5567242129696794e-05, "loss": 0.0369, "step": 6211 }, { "epoch": 0.673460537727667, "grad_norm": 1.7719930410385132, "learning_rate": 1.556651675612941e-05, "loss": 0.0205, "step": 6212 }, { "epoch": 0.6735689505637468, "grad_norm": 2.0953361988067627, "learning_rate": 1.5565791382562022e-05, "loss": 0.0818, "step": 6213 }, { "epoch": 0.6736773633998265, "grad_norm": 0.7148457169532776, "learning_rate": 1.5565066008994634e-05, "loss": 0.0092, "step": 6214 }, { "epoch": 0.6737857762359063, "grad_norm": 2.9496164321899414, "learning_rate": 1.5564340635427246e-05, "loss": 0.0655, "step": 6215 }, { "epoch": 0.6738941890719862, "grad_norm": 2.0144224166870117, "learning_rate": 1.556361526185986e-05, "loss": 0.0328, "step": 6216 }, { "epoch": 0.6740026019080659, "grad_norm": 1.3473950624465942, "learning_rate": 1.556288988829247e-05, "loss": 0.0482, "step": 6217 }, { "epoch": 0.6741110147441457, "grad_norm": 1.3314992189407349, "learning_rate": 1.5562164514725083e-05, "loss": 0.0271, "step": 6218 }, { "epoch": 0.6742194275802255, "grad_norm": 2.1291887760162354, "learning_rate": 1.5561439141157698e-05, "loss": 0.0225, "step": 6219 }, { "epoch": 0.6743278404163053, "grad_norm": 2.5571744441986084, "learning_rate": 1.556071376759031e-05, "loss": 0.0134, "step": 6220 }, { "epoch": 0.6744362532523851, "grad_norm": 1.8258159160614014, "learning_rate": 1.5559988394022923e-05, "loss": 0.0369, "step": 6221 }, { "epoch": 0.6745446660884649, "grad_norm": 0.5102377533912659, "learning_rate": 1.5559263020455535e-05, "loss": 0.0166, "step": 6222 }, { "epoch": 0.6746530789245446, "grad_norm": 2.61210036277771, "learning_rate": 1.5558537646888147e-05, "loss": 0.0298, "step": 6223 }, { "epoch": 0.6747614917606245, "grad_norm": 2.0133817195892334, "learning_rate": 1.555781227332076e-05, "loss": 0.0391, "step": 6224 }, { "epoch": 0.6748699045967043, "grad_norm": 1.9079943895339966, "learning_rate": 1.5557086899753375e-05, "loss": 0.0377, "step": 6225 }, { "epoch": 0.674978317432784, "grad_norm": 0.5940139293670654, "learning_rate": 1.5556361526185987e-05, "loss": 0.006, "step": 6226 }, { "epoch": 0.6750867302688638, "grad_norm": 4.544440746307373, "learning_rate": 1.5555636152618602e-05, "loss": 0.0457, "step": 6227 }, { "epoch": 0.6751951431049437, "grad_norm": 1.1493589878082275, "learning_rate": 1.5554910779051214e-05, "loss": 0.0333, "step": 6228 }, { "epoch": 0.6753035559410234, "grad_norm": 2.997349739074707, "learning_rate": 1.5554185405483826e-05, "loss": 0.0549, "step": 6229 }, { "epoch": 0.6754119687771032, "grad_norm": 6.1722283363342285, "learning_rate": 1.555346003191644e-05, "loss": 0.0448, "step": 6230 }, { "epoch": 0.675520381613183, "grad_norm": 2.054063558578491, "learning_rate": 1.555273465834905e-05, "loss": 0.0516, "step": 6231 }, { "epoch": 0.6756287944492628, "grad_norm": 5.531498432159424, "learning_rate": 1.5552009284781663e-05, "loss": 0.0395, "step": 6232 }, { "epoch": 0.6757372072853426, "grad_norm": 4.467756748199463, "learning_rate": 1.5551283911214275e-05, "loss": 0.0454, "step": 6233 }, { "epoch": 0.6758456201214224, "grad_norm": 2.1222872734069824, "learning_rate": 1.555055853764689e-05, "loss": 0.0397, "step": 6234 }, { "epoch": 0.6759540329575021, "grad_norm": 1.7751879692077637, "learning_rate": 1.5549833164079503e-05, "loss": 0.0203, "step": 6235 }, { "epoch": 0.676062445793582, "grad_norm": 6.588987350463867, "learning_rate": 1.5549107790512115e-05, "loss": 0.0389, "step": 6236 }, { "epoch": 0.6761708586296618, "grad_norm": 3.190140962600708, "learning_rate": 1.5548382416944727e-05, "loss": 0.0421, "step": 6237 }, { "epoch": 0.6762792714657415, "grad_norm": 3.534895896911621, "learning_rate": 1.554765704337734e-05, "loss": 0.0464, "step": 6238 }, { "epoch": 0.6763876843018214, "grad_norm": 4.418449878692627, "learning_rate": 1.554693166980995e-05, "loss": 0.027, "step": 6239 }, { "epoch": 0.6764960971379012, "grad_norm": 2.5480120182037354, "learning_rate": 1.5546206296242567e-05, "loss": 0.0212, "step": 6240 }, { "epoch": 0.6766045099739809, "grad_norm": 5.027189254760742, "learning_rate": 1.554548092267518e-05, "loss": 0.0671, "step": 6241 }, { "epoch": 0.6767129228100607, "grad_norm": 7.505736351013184, "learning_rate": 1.5544755549107794e-05, "loss": 0.0333, "step": 6242 }, { "epoch": 0.6768213356461406, "grad_norm": 4.553447723388672, "learning_rate": 1.5544030175540407e-05, "loss": 0.0339, "step": 6243 }, { "epoch": 0.6769297484822203, "grad_norm": 8.865666389465332, "learning_rate": 1.554330480197302e-05, "loss": 0.0941, "step": 6244 }, { "epoch": 0.6770381613183001, "grad_norm": 1.185862421989441, "learning_rate": 1.554257942840563e-05, "loss": 0.026, "step": 6245 }, { "epoch": 0.6771465741543798, "grad_norm": 2.8875532150268555, "learning_rate": 1.5541854054838243e-05, "loss": 0.0767, "step": 6246 }, { "epoch": 0.6772549869904597, "grad_norm": 1.2288799285888672, "learning_rate": 1.5541128681270855e-05, "loss": 0.0168, "step": 6247 }, { "epoch": 0.6773633998265395, "grad_norm": 1.0228793621063232, "learning_rate": 1.5540403307703467e-05, "loss": 0.0069, "step": 6248 }, { "epoch": 0.6774718126626192, "grad_norm": 3.761200189590454, "learning_rate": 1.5539677934136083e-05, "loss": 0.0839, "step": 6249 }, { "epoch": 0.677580225498699, "grad_norm": 4.318882942199707, "learning_rate": 1.5538952560568695e-05, "loss": 0.0452, "step": 6250 }, { "epoch": 0.6776886383347789, "grad_norm": 3.533691883087158, "learning_rate": 1.5538227187001307e-05, "loss": 0.0784, "step": 6251 }, { "epoch": 0.6777970511708586, "grad_norm": 1.2997759580612183, "learning_rate": 1.553750181343392e-05, "loss": 0.0467, "step": 6252 }, { "epoch": 0.6779054640069384, "grad_norm": 5.0535712242126465, "learning_rate": 1.553677643986653e-05, "loss": 0.0766, "step": 6253 }, { "epoch": 0.6780138768430182, "grad_norm": 3.0593857765197754, "learning_rate": 1.5536051066299144e-05, "loss": 0.045, "step": 6254 }, { "epoch": 0.678122289679098, "grad_norm": 2.7986350059509277, "learning_rate": 1.553532569273176e-05, "loss": 0.0346, "step": 6255 }, { "epoch": 0.6782307025151778, "grad_norm": 3.6345245838165283, "learning_rate": 1.553460031916437e-05, "loss": 0.0425, "step": 6256 }, { "epoch": 0.6783391153512576, "grad_norm": 1.838793396949768, "learning_rate": 1.5533874945596983e-05, "loss": 0.0481, "step": 6257 }, { "epoch": 0.6784475281873373, "grad_norm": 0.6680998206138611, "learning_rate": 1.5533149572029596e-05, "loss": 0.0107, "step": 6258 }, { "epoch": 0.6785559410234172, "grad_norm": 4.118919849395752, "learning_rate": 1.553242419846221e-05, "loss": 0.0765, "step": 6259 }, { "epoch": 0.678664353859497, "grad_norm": 2.3136301040649414, "learning_rate": 1.5531698824894823e-05, "loss": 0.0714, "step": 6260 }, { "epoch": 0.6787727666955767, "grad_norm": 2.3305504322052, "learning_rate": 1.5530973451327435e-05, "loss": 0.0241, "step": 6261 }, { "epoch": 0.6788811795316565, "grad_norm": 2.8829715251922607, "learning_rate": 1.5530248077760048e-05, "loss": 0.0502, "step": 6262 }, { "epoch": 0.6789895923677364, "grad_norm": 0.9023724794387817, "learning_rate": 1.552952270419266e-05, "loss": 0.0116, "step": 6263 }, { "epoch": 0.6790980052038161, "grad_norm": 2.0835671424865723, "learning_rate": 1.5528797330625275e-05, "loss": 0.0398, "step": 6264 }, { "epoch": 0.6792064180398959, "grad_norm": 3.074652671813965, "learning_rate": 1.5528071957057887e-05, "loss": 0.0866, "step": 6265 }, { "epoch": 0.6793148308759757, "grad_norm": 2.1839263439178467, "learning_rate": 1.55273465834905e-05, "loss": 0.0326, "step": 6266 }, { "epoch": 0.6794232437120555, "grad_norm": 1.9112528562545776, "learning_rate": 1.552662120992311e-05, "loss": 0.0434, "step": 6267 }, { "epoch": 0.6795316565481353, "grad_norm": 2.7461659908294678, "learning_rate": 1.5525895836355724e-05, "loss": 0.0411, "step": 6268 }, { "epoch": 0.6796400693842151, "grad_norm": 1.4646518230438232, "learning_rate": 1.5525170462788336e-05, "loss": 0.0241, "step": 6269 }, { "epoch": 0.6797484822202949, "grad_norm": 1.2423628568649292, "learning_rate": 1.5524445089220948e-05, "loss": 0.0218, "step": 6270 }, { "epoch": 0.6798568950563747, "grad_norm": 2.006364583969116, "learning_rate": 1.5523719715653564e-05, "loss": 0.0333, "step": 6271 }, { "epoch": 0.6799653078924545, "grad_norm": 2.9473915100097656, "learning_rate": 1.5522994342086176e-05, "loss": 0.0749, "step": 6272 }, { "epoch": 0.6800737207285342, "grad_norm": 2.7493128776550293, "learning_rate": 1.5522268968518788e-05, "loss": 0.0678, "step": 6273 }, { "epoch": 0.6801821335646141, "grad_norm": 2.640374183654785, "learning_rate": 1.55215435949514e-05, "loss": 0.0657, "step": 6274 }, { "epoch": 0.6802905464006939, "grad_norm": 1.9766991138458252, "learning_rate": 1.5520818221384012e-05, "loss": 0.0587, "step": 6275 }, { "epoch": 0.6803989592367736, "grad_norm": 0.5994882583618164, "learning_rate": 1.5520092847816624e-05, "loss": 0.0094, "step": 6276 }, { "epoch": 0.6805073720728534, "grad_norm": 1.1844371557235718, "learning_rate": 1.551936747424924e-05, "loss": 0.0208, "step": 6277 }, { "epoch": 0.6806157849089333, "grad_norm": 2.1183114051818848, "learning_rate": 1.5518642100681852e-05, "loss": 0.0282, "step": 6278 }, { "epoch": 0.680724197745013, "grad_norm": 1.8371423482894897, "learning_rate": 1.5517916727114468e-05, "loss": 0.0235, "step": 6279 }, { "epoch": 0.6808326105810928, "grad_norm": 2.2182295322418213, "learning_rate": 1.551719135354708e-05, "loss": 0.0149, "step": 6280 }, { "epoch": 0.6809410234171726, "grad_norm": 1.1045044660568237, "learning_rate": 1.5516465979979692e-05, "loss": 0.0299, "step": 6281 }, { "epoch": 0.6810494362532524, "grad_norm": 0.4136558771133423, "learning_rate": 1.5515740606412304e-05, "loss": 0.0094, "step": 6282 }, { "epoch": 0.6811578490893322, "grad_norm": 2.0722506046295166, "learning_rate": 1.5515015232844916e-05, "loss": 0.0421, "step": 6283 }, { "epoch": 0.681266261925412, "grad_norm": 1.0790083408355713, "learning_rate": 1.5514289859277528e-05, "loss": 0.0231, "step": 6284 }, { "epoch": 0.6813746747614917, "grad_norm": 3.2103612422943115, "learning_rate": 1.551356448571014e-05, "loss": 0.0614, "step": 6285 }, { "epoch": 0.6814830875975716, "grad_norm": 2.5480706691741943, "learning_rate": 1.5512839112142756e-05, "loss": 0.0776, "step": 6286 }, { "epoch": 0.6815915004336514, "grad_norm": 1.8565428256988525, "learning_rate": 1.5512113738575368e-05, "loss": 0.0397, "step": 6287 }, { "epoch": 0.6816999132697311, "grad_norm": 2.0163414478302, "learning_rate": 1.551138836500798e-05, "loss": 0.0537, "step": 6288 }, { "epoch": 0.6818083261058109, "grad_norm": 3.2265522480010986, "learning_rate": 1.5510662991440592e-05, "loss": 0.0371, "step": 6289 }, { "epoch": 0.6819167389418908, "grad_norm": 2.066340208053589, "learning_rate": 1.5509937617873204e-05, "loss": 0.0477, "step": 6290 }, { "epoch": 0.6820251517779705, "grad_norm": 3.152362108230591, "learning_rate": 1.5509212244305817e-05, "loss": 0.0181, "step": 6291 }, { "epoch": 0.6821335646140503, "grad_norm": 1.11388099193573, "learning_rate": 1.5508486870738432e-05, "loss": 0.0225, "step": 6292 }, { "epoch": 0.68224197745013, "grad_norm": 1.7567970752716064, "learning_rate": 1.5507761497171044e-05, "loss": 0.0146, "step": 6293 }, { "epoch": 0.6823503902862099, "grad_norm": 0.7189561724662781, "learning_rate": 1.550703612360366e-05, "loss": 0.0136, "step": 6294 }, { "epoch": 0.6824588031222897, "grad_norm": 1.9954618215560913, "learning_rate": 1.5506310750036272e-05, "loss": 0.0216, "step": 6295 }, { "epoch": 0.6825672159583694, "grad_norm": 1.4274389743804932, "learning_rate": 1.5505585376468884e-05, "loss": 0.0186, "step": 6296 }, { "epoch": 0.6826756287944492, "grad_norm": 3.5117244720458984, "learning_rate": 1.5504860002901496e-05, "loss": 0.0515, "step": 6297 }, { "epoch": 0.6827840416305291, "grad_norm": 4.318310737609863, "learning_rate": 1.550413462933411e-05, "loss": 0.0647, "step": 6298 }, { "epoch": 0.6828924544666088, "grad_norm": 2.630152463912964, "learning_rate": 1.550340925576672e-05, "loss": 0.0149, "step": 6299 }, { "epoch": 0.6830008673026886, "grad_norm": 1.5546846389770508, "learning_rate": 1.5502683882199333e-05, "loss": 0.0317, "step": 6300 }, { "epoch": 0.6831092801387685, "grad_norm": 1.4597309827804565, "learning_rate": 1.5501958508631948e-05, "loss": 0.01, "step": 6301 }, { "epoch": 0.6832176929748482, "grad_norm": 1.972609519958496, "learning_rate": 1.550123313506456e-05, "loss": 0.0139, "step": 6302 }, { "epoch": 0.683326105810928, "grad_norm": 3.972626209259033, "learning_rate": 1.5500507761497173e-05, "loss": 0.0659, "step": 6303 }, { "epoch": 0.6834345186470078, "grad_norm": 5.26975154876709, "learning_rate": 1.5499782387929785e-05, "loss": 0.0626, "step": 6304 }, { "epoch": 0.6835429314830876, "grad_norm": 2.328242301940918, "learning_rate": 1.5499057014362397e-05, "loss": 0.047, "step": 6305 }, { "epoch": 0.6836513443191674, "grad_norm": 1.5364903211593628, "learning_rate": 1.549833164079501e-05, "loss": 0.0268, "step": 6306 }, { "epoch": 0.6837597571552472, "grad_norm": 1.0845463275909424, "learning_rate": 1.5497606267227624e-05, "loss": 0.025, "step": 6307 }, { "epoch": 0.6838681699913269, "grad_norm": 9.041102409362793, "learning_rate": 1.5496880893660237e-05, "loss": 0.0401, "step": 6308 }, { "epoch": 0.6839765828274068, "grad_norm": 3.8950366973876953, "learning_rate": 1.549615552009285e-05, "loss": 0.0335, "step": 6309 }, { "epoch": 0.6840849956634866, "grad_norm": 6.213386535644531, "learning_rate": 1.549543014652546e-05, "loss": 0.0382, "step": 6310 }, { "epoch": 0.6841934084995663, "grad_norm": 1.6822937726974487, "learning_rate": 1.5494704772958076e-05, "loss": 0.0111, "step": 6311 }, { "epoch": 0.6843018213356461, "grad_norm": 1.0711230039596558, "learning_rate": 1.549397939939069e-05, "loss": 0.0231, "step": 6312 }, { "epoch": 0.684410234171726, "grad_norm": 8.353927612304688, "learning_rate": 1.54932540258233e-05, "loss": 0.04, "step": 6313 }, { "epoch": 0.6845186470078057, "grad_norm": 4.141488075256348, "learning_rate": 1.5492528652255913e-05, "loss": 0.0354, "step": 6314 }, { "epoch": 0.6846270598438855, "grad_norm": 3.562082529067993, "learning_rate": 1.5491803278688525e-05, "loss": 0.0348, "step": 6315 }, { "epoch": 0.6847354726799653, "grad_norm": 2.0800411701202393, "learning_rate": 1.549107790512114e-05, "loss": 0.0391, "step": 6316 }, { "epoch": 0.6848438855160451, "grad_norm": 1.3316453695297241, "learning_rate": 1.5490352531553753e-05, "loss": 0.015, "step": 6317 }, { "epoch": 0.6849522983521249, "grad_norm": 2.3120827674865723, "learning_rate": 1.5489627157986365e-05, "loss": 0.0385, "step": 6318 }, { "epoch": 0.6850607111882047, "grad_norm": 5.466778755187988, "learning_rate": 1.5488901784418977e-05, "loss": 0.0232, "step": 6319 }, { "epoch": 0.6851691240242844, "grad_norm": 7.556248188018799, "learning_rate": 1.548817641085159e-05, "loss": 0.0484, "step": 6320 }, { "epoch": 0.6852775368603643, "grad_norm": 2.0293776988983154, "learning_rate": 1.54874510372842e-05, "loss": 0.0299, "step": 6321 }, { "epoch": 0.6853859496964441, "grad_norm": 4.78512716293335, "learning_rate": 1.5486725663716813e-05, "loss": 0.1145, "step": 6322 }, { "epoch": 0.6854943625325238, "grad_norm": 3.074976921081543, "learning_rate": 1.548600029014943e-05, "loss": 0.0414, "step": 6323 }, { "epoch": 0.6856027753686036, "grad_norm": 1.0992894172668457, "learning_rate": 1.548527491658204e-05, "loss": 0.0211, "step": 6324 }, { "epoch": 0.6857111882046835, "grad_norm": 3.2810328006744385, "learning_rate": 1.5484549543014653e-05, "loss": 0.0517, "step": 6325 }, { "epoch": 0.6858196010407632, "grad_norm": 1.9072024822235107, "learning_rate": 1.5483824169447265e-05, "loss": 0.0178, "step": 6326 }, { "epoch": 0.685928013876843, "grad_norm": 2.5140485763549805, "learning_rate": 1.5483098795879878e-05, "loss": 0.0368, "step": 6327 }, { "epoch": 0.6860364267129229, "grad_norm": 0.41370806097984314, "learning_rate": 1.5482373422312493e-05, "loss": 0.0109, "step": 6328 }, { "epoch": 0.6861448395490026, "grad_norm": 1.9236350059509277, "learning_rate": 1.5481648048745105e-05, "loss": 0.0329, "step": 6329 }, { "epoch": 0.6862532523850824, "grad_norm": 1.780253529548645, "learning_rate": 1.5480922675177717e-05, "loss": 0.0261, "step": 6330 }, { "epoch": 0.6863616652211622, "grad_norm": 1.5697609186172485, "learning_rate": 1.5480197301610333e-05, "loss": 0.0196, "step": 6331 }, { "epoch": 0.686470078057242, "grad_norm": 2.2492427825927734, "learning_rate": 1.5479471928042945e-05, "loss": 0.0326, "step": 6332 }, { "epoch": 0.6865784908933218, "grad_norm": 2.811805486679077, "learning_rate": 1.5478746554475557e-05, "loss": 0.0317, "step": 6333 }, { "epoch": 0.6866869037294016, "grad_norm": 2.4531331062316895, "learning_rate": 1.547802118090817e-05, "loss": 0.0244, "step": 6334 }, { "epoch": 0.6867953165654813, "grad_norm": 1.6033064126968384, "learning_rate": 1.547729580734078e-05, "loss": 0.04, "step": 6335 }, { "epoch": 0.6869037294015612, "grad_norm": 3.496811628341675, "learning_rate": 1.5476570433773394e-05, "loss": 0.0315, "step": 6336 }, { "epoch": 0.687012142237641, "grad_norm": 3.6649351119995117, "learning_rate": 1.5475845060206006e-05, "loss": 0.0157, "step": 6337 }, { "epoch": 0.6871205550737207, "grad_norm": 1.9454090595245361, "learning_rate": 1.547511968663862e-05, "loss": 0.0304, "step": 6338 }, { "epoch": 0.6872289679098005, "grad_norm": 2.0504703521728516, "learning_rate": 1.5474394313071233e-05, "loss": 0.0281, "step": 6339 }, { "epoch": 0.6873373807458804, "grad_norm": 7.8204193115234375, "learning_rate": 1.5473668939503846e-05, "loss": 0.039, "step": 6340 }, { "epoch": 0.6874457935819601, "grad_norm": 0.21253541111946106, "learning_rate": 1.5472943565936458e-05, "loss": 0.0019, "step": 6341 }, { "epoch": 0.6875542064180399, "grad_norm": 2.697347640991211, "learning_rate": 1.547221819236907e-05, "loss": 0.0404, "step": 6342 }, { "epoch": 0.6876626192541196, "grad_norm": 0.8933144211769104, "learning_rate": 1.5471492818801682e-05, "loss": 0.0152, "step": 6343 }, { "epoch": 0.6877710320901995, "grad_norm": 4.912659645080566, "learning_rate": 1.5470767445234297e-05, "loss": 0.0409, "step": 6344 }, { "epoch": 0.6878794449262793, "grad_norm": 2.1885156631469727, "learning_rate": 1.547004207166691e-05, "loss": 0.0343, "step": 6345 }, { "epoch": 0.687987857762359, "grad_norm": 5.874892234802246, "learning_rate": 1.5469316698099525e-05, "loss": 0.0514, "step": 6346 }, { "epoch": 0.6880962705984388, "grad_norm": 1.5038245916366577, "learning_rate": 1.5468591324532137e-05, "loss": 0.04, "step": 6347 }, { "epoch": 0.6882046834345187, "grad_norm": 5.8769850730896, "learning_rate": 1.546786595096475e-05, "loss": 0.0497, "step": 6348 }, { "epoch": 0.6883130962705984, "grad_norm": 1.635311245918274, "learning_rate": 1.546714057739736e-05, "loss": 0.0133, "step": 6349 }, { "epoch": 0.6884215091066782, "grad_norm": 3.318399667739868, "learning_rate": 1.5466415203829974e-05, "loss": 0.0683, "step": 6350 }, { "epoch": 0.688529921942758, "grad_norm": 4.112983226776123, "learning_rate": 1.5465689830262586e-05, "loss": 0.0444, "step": 6351 }, { "epoch": 0.6886383347788378, "grad_norm": 3.402355909347534, "learning_rate": 1.5464964456695198e-05, "loss": 0.0592, "step": 6352 }, { "epoch": 0.6887467476149176, "grad_norm": 0.4132062494754791, "learning_rate": 1.5464239083127814e-05, "loss": 0.0143, "step": 6353 }, { "epoch": 0.6888551604509974, "grad_norm": 8.564096450805664, "learning_rate": 1.5463513709560426e-05, "loss": 0.0638, "step": 6354 }, { "epoch": 0.6889635732870771, "grad_norm": 0.8954260945320129, "learning_rate": 1.5462788335993038e-05, "loss": 0.0167, "step": 6355 }, { "epoch": 0.689071986123157, "grad_norm": 0.7464190125465393, "learning_rate": 1.546206296242565e-05, "loss": 0.0062, "step": 6356 }, { "epoch": 0.6891803989592368, "grad_norm": 9.868799209594727, "learning_rate": 1.5461337588858262e-05, "loss": 0.0448, "step": 6357 }, { "epoch": 0.6892888117953165, "grad_norm": 1.818411946296692, "learning_rate": 1.5460612215290874e-05, "loss": 0.0483, "step": 6358 }, { "epoch": 0.6893972246313964, "grad_norm": 4.3481245040893555, "learning_rate": 1.545988684172349e-05, "loss": 0.0441, "step": 6359 }, { "epoch": 0.6895056374674762, "grad_norm": 2.5397207736968994, "learning_rate": 1.5459161468156102e-05, "loss": 0.0275, "step": 6360 }, { "epoch": 0.6896140503035559, "grad_norm": 4.800454616546631, "learning_rate": 1.5458436094588714e-05, "loss": 0.0512, "step": 6361 }, { "epoch": 0.6897224631396357, "grad_norm": 1.4384690523147583, "learning_rate": 1.5457710721021326e-05, "loss": 0.0283, "step": 6362 }, { "epoch": 0.6898308759757156, "grad_norm": 3.350599765777588, "learning_rate": 1.5456985347453942e-05, "loss": 0.0224, "step": 6363 }, { "epoch": 0.6899392888117953, "grad_norm": 2.7035255432128906, "learning_rate": 1.5456259973886554e-05, "loss": 0.0305, "step": 6364 }, { "epoch": 0.6900477016478751, "grad_norm": 1.2351707220077515, "learning_rate": 1.5455534600319166e-05, "loss": 0.0272, "step": 6365 }, { "epoch": 0.6901561144839549, "grad_norm": 3.498579740524292, "learning_rate": 1.5454809226751778e-05, "loss": 0.0543, "step": 6366 }, { "epoch": 0.6902645273200347, "grad_norm": 0.5462286472320557, "learning_rate": 1.545408385318439e-05, "loss": 0.0051, "step": 6367 }, { "epoch": 0.6903729401561145, "grad_norm": 4.222938060760498, "learning_rate": 1.5453358479617006e-05, "loss": 0.0427, "step": 6368 }, { "epoch": 0.6904813529921943, "grad_norm": 6.496833324432373, "learning_rate": 1.5452633106049618e-05, "loss": 0.0498, "step": 6369 }, { "epoch": 0.690589765828274, "grad_norm": 1.96712327003479, "learning_rate": 1.545190773248223e-05, "loss": 0.0291, "step": 6370 }, { "epoch": 0.6906981786643539, "grad_norm": 7.166554927825928, "learning_rate": 1.5451182358914842e-05, "loss": 0.0458, "step": 6371 }, { "epoch": 0.6908065915004337, "grad_norm": 3.4271950721740723, "learning_rate": 1.5450456985347454e-05, "loss": 0.0732, "step": 6372 }, { "epoch": 0.6909150043365134, "grad_norm": 2.302427291870117, "learning_rate": 1.5449731611780067e-05, "loss": 0.0344, "step": 6373 }, { "epoch": 0.6910234171725932, "grad_norm": 4.970489978790283, "learning_rate": 1.544900623821268e-05, "loss": 0.0302, "step": 6374 }, { "epoch": 0.6911318300086731, "grad_norm": 3.976546287536621, "learning_rate": 1.5448280864645294e-05, "loss": 0.0426, "step": 6375 }, { "epoch": 0.6912402428447528, "grad_norm": 0.7080246210098267, "learning_rate": 1.5447555491077906e-05, "loss": 0.0111, "step": 6376 }, { "epoch": 0.6913486556808326, "grad_norm": 6.0543622970581055, "learning_rate": 1.544683011751052e-05, "loss": 0.0565, "step": 6377 }, { "epoch": 0.6914570685169124, "grad_norm": 3.2535343170166016, "learning_rate": 1.544610474394313e-05, "loss": 0.0458, "step": 6378 }, { "epoch": 0.6915654813529922, "grad_norm": 3.4925637245178223, "learning_rate": 1.5445379370375743e-05, "loss": 0.0355, "step": 6379 }, { "epoch": 0.691673894189072, "grad_norm": 3.5065996646881104, "learning_rate": 1.544465399680836e-05, "loss": 0.0603, "step": 6380 }, { "epoch": 0.6917823070251518, "grad_norm": 2.1063430309295654, "learning_rate": 1.544392862324097e-05, "loss": 0.0108, "step": 6381 }, { "epoch": 0.6918907198612315, "grad_norm": 1.6253585815429688, "learning_rate": 1.5443203249673583e-05, "loss": 0.0351, "step": 6382 }, { "epoch": 0.6919991326973114, "grad_norm": 2.212463617324829, "learning_rate": 1.5442477876106198e-05, "loss": 0.0339, "step": 6383 }, { "epoch": 0.6921075455333912, "grad_norm": 3.454627275466919, "learning_rate": 1.544175250253881e-05, "loss": 0.07, "step": 6384 }, { "epoch": 0.6922159583694709, "grad_norm": 6.6940789222717285, "learning_rate": 1.5441027128971422e-05, "loss": 0.033, "step": 6385 }, { "epoch": 0.6923243712055508, "grad_norm": 1.3050345182418823, "learning_rate": 1.5440301755404035e-05, "loss": 0.0071, "step": 6386 }, { "epoch": 0.6924327840416306, "grad_norm": 0.6972947120666504, "learning_rate": 1.5439576381836647e-05, "loss": 0.0284, "step": 6387 }, { "epoch": 0.6925411968777103, "grad_norm": 1.9210879802703857, "learning_rate": 1.543885100826926e-05, "loss": 0.0244, "step": 6388 }, { "epoch": 0.6926496097137901, "grad_norm": 2.4976627826690674, "learning_rate": 1.543812563470187e-05, "loss": 0.0245, "step": 6389 }, { "epoch": 0.69275802254987, "grad_norm": 4.648207664489746, "learning_rate": 1.5437400261134487e-05, "loss": 0.03, "step": 6390 }, { "epoch": 0.6928664353859497, "grad_norm": 0.8146288990974426, "learning_rate": 1.54366748875671e-05, "loss": 0.0294, "step": 6391 }, { "epoch": 0.6929748482220295, "grad_norm": 4.824600696563721, "learning_rate": 1.543594951399971e-05, "loss": 0.079, "step": 6392 }, { "epoch": 0.6930832610581092, "grad_norm": 2.7086825370788574, "learning_rate": 1.5435224140432323e-05, "loss": 0.0465, "step": 6393 }, { "epoch": 0.6931916738941891, "grad_norm": 1.969772458076477, "learning_rate": 1.5434498766864935e-05, "loss": 0.0287, "step": 6394 }, { "epoch": 0.6933000867302689, "grad_norm": 0.43922653794288635, "learning_rate": 1.5433773393297547e-05, "loss": 0.0145, "step": 6395 }, { "epoch": 0.6934084995663486, "grad_norm": 1.9105329513549805, "learning_rate": 1.5433048019730163e-05, "loss": 0.0377, "step": 6396 }, { "epoch": 0.6935169124024284, "grad_norm": 1.9558682441711426, "learning_rate": 1.5432322646162775e-05, "loss": 0.0369, "step": 6397 }, { "epoch": 0.6936253252385083, "grad_norm": 0.6979974508285522, "learning_rate": 1.543159727259539e-05, "loss": 0.0163, "step": 6398 }, { "epoch": 0.693733738074588, "grad_norm": 0.7980437874794006, "learning_rate": 1.5430871899028003e-05, "loss": 0.0027, "step": 6399 }, { "epoch": 0.6938421509106678, "grad_norm": 2.1363563537597656, "learning_rate": 1.5430146525460615e-05, "loss": 0.0474, "step": 6400 }, { "epoch": 0.6939505637467476, "grad_norm": 10.623748779296875, "learning_rate": 1.5429421151893227e-05, "loss": 0.0259, "step": 6401 }, { "epoch": 0.6940589765828274, "grad_norm": 7.901752471923828, "learning_rate": 1.542869577832584e-05, "loss": 0.0935, "step": 6402 }, { "epoch": 0.6941673894189072, "grad_norm": 5.410195827484131, "learning_rate": 1.542797040475845e-05, "loss": 0.121, "step": 6403 }, { "epoch": 0.694275802254987, "grad_norm": 3.444336414337158, "learning_rate": 1.5427245031191063e-05, "loss": 0.077, "step": 6404 }, { "epoch": 0.6943842150910667, "grad_norm": 4.335376262664795, "learning_rate": 1.542651965762368e-05, "loss": 0.0735, "step": 6405 }, { "epoch": 0.6944926279271466, "grad_norm": 6.189162731170654, "learning_rate": 1.542579428405629e-05, "loss": 0.0405, "step": 6406 }, { "epoch": 0.6946010407632264, "grad_norm": 3.1364634037017822, "learning_rate": 1.5425068910488903e-05, "loss": 0.0543, "step": 6407 }, { "epoch": 0.6947094535993061, "grad_norm": 1.688596487045288, "learning_rate": 1.5424343536921515e-05, "loss": 0.0281, "step": 6408 }, { "epoch": 0.6948178664353859, "grad_norm": 4.668508052825928, "learning_rate": 1.5423618163354127e-05, "loss": 0.0334, "step": 6409 }, { "epoch": 0.6949262792714658, "grad_norm": 3.2016985416412354, "learning_rate": 1.542289278978674e-05, "loss": 0.0384, "step": 6410 }, { "epoch": 0.6950346921075455, "grad_norm": 1.4710347652435303, "learning_rate": 1.5422167416219355e-05, "loss": 0.0315, "step": 6411 }, { "epoch": 0.6951431049436253, "grad_norm": 1.334486722946167, "learning_rate": 1.5421442042651967e-05, "loss": 0.0408, "step": 6412 }, { "epoch": 0.6952515177797051, "grad_norm": 4.0365495681762695, "learning_rate": 1.542071666908458e-05, "loss": 0.0845, "step": 6413 }, { "epoch": 0.6953599306157849, "grad_norm": 2.8798153400421143, "learning_rate": 1.541999129551719e-05, "loss": 0.0762, "step": 6414 }, { "epoch": 0.6954683434518647, "grad_norm": 5.807030200958252, "learning_rate": 1.5419265921949807e-05, "loss": 0.0541, "step": 6415 }, { "epoch": 0.6955767562879445, "grad_norm": 2.2321183681488037, "learning_rate": 1.541854054838242e-05, "loss": 0.0266, "step": 6416 }, { "epoch": 0.6956851691240243, "grad_norm": 1.4345214366912842, "learning_rate": 1.541781517481503e-05, "loss": 0.0248, "step": 6417 }, { "epoch": 0.6957935819601041, "grad_norm": 3.5550801753997803, "learning_rate": 1.5417089801247644e-05, "loss": 0.0755, "step": 6418 }, { "epoch": 0.6959019947961839, "grad_norm": 2.9337875843048096, "learning_rate": 1.5416364427680256e-05, "loss": 0.0361, "step": 6419 }, { "epoch": 0.6960104076322636, "grad_norm": 2.7294018268585205, "learning_rate": 1.541563905411287e-05, "loss": 0.0501, "step": 6420 }, { "epoch": 0.6961188204683435, "grad_norm": 1.2554007768630981, "learning_rate": 1.5414913680545483e-05, "loss": 0.0312, "step": 6421 }, { "epoch": 0.6962272333044233, "grad_norm": 2.4125990867614746, "learning_rate": 1.5414188306978096e-05, "loss": 0.0129, "step": 6422 }, { "epoch": 0.696335646140503, "grad_norm": 1.915705680847168, "learning_rate": 1.5413462933410708e-05, "loss": 0.054, "step": 6423 }, { "epoch": 0.6964440589765828, "grad_norm": 1.4006152153015137, "learning_rate": 1.541273755984332e-05, "loss": 0.0262, "step": 6424 }, { "epoch": 0.6965524718126627, "grad_norm": 1.0310523509979248, "learning_rate": 1.5412012186275932e-05, "loss": 0.0294, "step": 6425 }, { "epoch": 0.6966608846487424, "grad_norm": 2.7455637454986572, "learning_rate": 1.5411286812708544e-05, "loss": 0.0453, "step": 6426 }, { "epoch": 0.6967692974848222, "grad_norm": 2.369006872177124, "learning_rate": 1.541056143914116e-05, "loss": 0.0215, "step": 6427 }, { "epoch": 0.696877710320902, "grad_norm": 1.621395468711853, "learning_rate": 1.5409836065573772e-05, "loss": 0.0182, "step": 6428 }, { "epoch": 0.6969861231569818, "grad_norm": 1.35612154006958, "learning_rate": 1.5409110692006384e-05, "loss": 0.0263, "step": 6429 }, { "epoch": 0.6970945359930616, "grad_norm": 1.1370481252670288, "learning_rate": 1.5408385318438996e-05, "loss": 0.0201, "step": 6430 }, { "epoch": 0.6972029488291414, "grad_norm": 2.3025341033935547, "learning_rate": 1.5407659944871608e-05, "loss": 0.0172, "step": 6431 }, { "epoch": 0.6973113616652211, "grad_norm": 2.6073555946350098, "learning_rate": 1.5406934571304224e-05, "loss": 0.0292, "step": 6432 }, { "epoch": 0.697419774501301, "grad_norm": 2.5799460411071777, "learning_rate": 1.5406209197736836e-05, "loss": 0.0342, "step": 6433 }, { "epoch": 0.6975281873373808, "grad_norm": 5.7299323081970215, "learning_rate": 1.5405483824169448e-05, "loss": 0.0491, "step": 6434 }, { "epoch": 0.6976366001734605, "grad_norm": 1.5978339910507202, "learning_rate": 1.5404758450602064e-05, "loss": 0.0331, "step": 6435 }, { "epoch": 0.6977450130095403, "grad_norm": 4.534636974334717, "learning_rate": 1.5404033077034676e-05, "loss": 0.0668, "step": 6436 }, { "epoch": 0.6978534258456202, "grad_norm": 4.278945446014404, "learning_rate": 1.5403307703467288e-05, "loss": 0.0889, "step": 6437 }, { "epoch": 0.6979618386816999, "grad_norm": 3.844377279281616, "learning_rate": 1.54025823298999e-05, "loss": 0.0632, "step": 6438 }, { "epoch": 0.6980702515177797, "grad_norm": 1.7674322128295898, "learning_rate": 1.5401856956332512e-05, "loss": 0.031, "step": 6439 }, { "epoch": 0.6981786643538594, "grad_norm": 1.3948017358779907, "learning_rate": 1.5401131582765124e-05, "loss": 0.0448, "step": 6440 }, { "epoch": 0.6982870771899393, "grad_norm": 1.8801676034927368, "learning_rate": 1.5400406209197736e-05, "loss": 0.0298, "step": 6441 }, { "epoch": 0.6983954900260191, "grad_norm": 1.3208328485488892, "learning_rate": 1.5399680835630352e-05, "loss": 0.0259, "step": 6442 }, { "epoch": 0.6985039028620988, "grad_norm": 2.357801914215088, "learning_rate": 1.5398955462062964e-05, "loss": 0.0875, "step": 6443 }, { "epoch": 0.6986123156981786, "grad_norm": 1.0450180768966675, "learning_rate": 1.5398230088495576e-05, "loss": 0.0314, "step": 6444 }, { "epoch": 0.6987207285342585, "grad_norm": 1.3989592790603638, "learning_rate": 1.539750471492819e-05, "loss": 0.0404, "step": 6445 }, { "epoch": 0.6988291413703382, "grad_norm": 1.66252863407135, "learning_rate": 1.53967793413608e-05, "loss": 0.0238, "step": 6446 }, { "epoch": 0.698937554206418, "grad_norm": 6.976204872131348, "learning_rate": 1.5396053967793413e-05, "loss": 0.1184, "step": 6447 }, { "epoch": 0.6990459670424979, "grad_norm": 0.7069690823554993, "learning_rate": 1.5395328594226028e-05, "loss": 0.0214, "step": 6448 }, { "epoch": 0.6991543798785776, "grad_norm": 2.9261274337768555, "learning_rate": 1.539460322065864e-05, "loss": 0.0593, "step": 6449 }, { "epoch": 0.6992627927146574, "grad_norm": 1.275661826133728, "learning_rate": 1.5393877847091256e-05, "loss": 0.0327, "step": 6450 }, { "epoch": 0.6993712055507372, "grad_norm": 1.5178173780441284, "learning_rate": 1.5393152473523868e-05, "loss": 0.0249, "step": 6451 }, { "epoch": 0.699479618386817, "grad_norm": 1.0287507772445679, "learning_rate": 1.539242709995648e-05, "loss": 0.0161, "step": 6452 }, { "epoch": 0.6995880312228968, "grad_norm": 4.648040771484375, "learning_rate": 1.5391701726389092e-05, "loss": 0.0915, "step": 6453 }, { "epoch": 0.6996964440589766, "grad_norm": 2.6432459354400635, "learning_rate": 1.5390976352821704e-05, "loss": 0.041, "step": 6454 }, { "epoch": 0.6998048568950563, "grad_norm": 3.183009624481201, "learning_rate": 1.5390250979254317e-05, "loss": 0.0714, "step": 6455 }, { "epoch": 0.6999132697311362, "grad_norm": 2.01147198677063, "learning_rate": 1.538952560568693e-05, "loss": 0.0274, "step": 6456 }, { "epoch": 0.700021682567216, "grad_norm": 2.145744800567627, "learning_rate": 1.5388800232119544e-05, "loss": 0.0387, "step": 6457 }, { "epoch": 0.7001300954032957, "grad_norm": 1.8843454122543335, "learning_rate": 1.5388074858552156e-05, "loss": 0.0651, "step": 6458 }, { "epoch": 0.7002385082393755, "grad_norm": 0.7881611585617065, "learning_rate": 1.538734948498477e-05, "loss": 0.0119, "step": 6459 }, { "epoch": 0.7003469210754554, "grad_norm": 2.506704092025757, "learning_rate": 1.538662411141738e-05, "loss": 0.0273, "step": 6460 }, { "epoch": 0.7004553339115351, "grad_norm": 1.9347975254058838, "learning_rate": 1.5385898737849993e-05, "loss": 0.0296, "step": 6461 }, { "epoch": 0.7005637467476149, "grad_norm": 2.159005641937256, "learning_rate": 1.5385173364282605e-05, "loss": 0.055, "step": 6462 }, { "epoch": 0.7006721595836947, "grad_norm": 3.394172430038452, "learning_rate": 1.538444799071522e-05, "loss": 0.0598, "step": 6463 }, { "epoch": 0.7007805724197745, "grad_norm": 1.72151517868042, "learning_rate": 1.5383722617147833e-05, "loss": 0.046, "step": 6464 }, { "epoch": 0.7008889852558543, "grad_norm": 2.1674084663391113, "learning_rate": 1.5382997243580445e-05, "loss": 0.0472, "step": 6465 }, { "epoch": 0.7009973980919341, "grad_norm": 0.5115740299224854, "learning_rate": 1.5382271870013057e-05, "loss": 0.0061, "step": 6466 }, { "epoch": 0.7011058109280138, "grad_norm": 0.8190004229545593, "learning_rate": 1.5381546496445672e-05, "loss": 0.0217, "step": 6467 }, { "epoch": 0.7012142237640937, "grad_norm": 1.4194380044937134, "learning_rate": 1.5380821122878285e-05, "loss": 0.0462, "step": 6468 }, { "epoch": 0.7013226366001735, "grad_norm": 3.7424206733703613, "learning_rate": 1.5380095749310897e-05, "loss": 0.0604, "step": 6469 }, { "epoch": 0.7014310494362532, "grad_norm": 1.7853285074234009, "learning_rate": 1.537937037574351e-05, "loss": 0.0254, "step": 6470 }, { "epoch": 0.701539462272333, "grad_norm": 5.820122718811035, "learning_rate": 1.537864500217612e-05, "loss": 0.0344, "step": 6471 }, { "epoch": 0.7016478751084129, "grad_norm": 2.771068572998047, "learning_rate": 1.5377919628608737e-05, "loss": 0.0354, "step": 6472 }, { "epoch": 0.7017562879444926, "grad_norm": 7.004798412322998, "learning_rate": 1.537719425504135e-05, "loss": 0.1007, "step": 6473 }, { "epoch": 0.7018647007805724, "grad_norm": 1.5588666200637817, "learning_rate": 1.537646888147396e-05, "loss": 0.0299, "step": 6474 }, { "epoch": 0.7019731136166523, "grad_norm": 0.9436423778533936, "learning_rate": 1.5375743507906573e-05, "loss": 0.0231, "step": 6475 }, { "epoch": 0.702081526452732, "grad_norm": 0.8207408785820007, "learning_rate": 1.5375018134339185e-05, "loss": 0.007, "step": 6476 }, { "epoch": 0.7021899392888118, "grad_norm": 0.9585299491882324, "learning_rate": 1.5374292760771797e-05, "loss": 0.026, "step": 6477 }, { "epoch": 0.7022983521248916, "grad_norm": 2.940669298171997, "learning_rate": 1.537356738720441e-05, "loss": 0.0564, "step": 6478 }, { "epoch": 0.7024067649609714, "grad_norm": 4.32949686050415, "learning_rate": 1.5372842013637025e-05, "loss": 0.0303, "step": 6479 }, { "epoch": 0.7025151777970512, "grad_norm": 2.1658387184143066, "learning_rate": 1.5372116640069637e-05, "loss": 0.0582, "step": 6480 }, { "epoch": 0.702623590633131, "grad_norm": 1.9390313625335693, "learning_rate": 1.537139126650225e-05, "loss": 0.0204, "step": 6481 }, { "epoch": 0.7027320034692107, "grad_norm": 4.2701334953308105, "learning_rate": 1.537066589293486e-05, "loss": 0.0481, "step": 6482 }, { "epoch": 0.7028404163052906, "grad_norm": 0.6670863628387451, "learning_rate": 1.5369940519367474e-05, "loss": 0.0114, "step": 6483 }, { "epoch": 0.7029488291413704, "grad_norm": 2.4228270053863525, "learning_rate": 1.536921514580009e-05, "loss": 0.0368, "step": 6484 }, { "epoch": 0.7030572419774501, "grad_norm": 1.523970127105713, "learning_rate": 1.53684897722327e-05, "loss": 0.0114, "step": 6485 }, { "epoch": 0.7031656548135299, "grad_norm": 1.8900837898254395, "learning_rate": 1.5367764398665313e-05, "loss": 0.044, "step": 6486 }, { "epoch": 0.7032740676496098, "grad_norm": 2.4750242233276367, "learning_rate": 1.536703902509793e-05, "loss": 0.037, "step": 6487 }, { "epoch": 0.7033824804856895, "grad_norm": 1.5696065425872803, "learning_rate": 1.536631365153054e-05, "loss": 0.025, "step": 6488 }, { "epoch": 0.7034908933217693, "grad_norm": 7.1247382164001465, "learning_rate": 1.5365588277963153e-05, "loss": 0.0223, "step": 6489 }, { "epoch": 0.703599306157849, "grad_norm": 1.5798828601837158, "learning_rate": 1.5364862904395765e-05, "loss": 0.0203, "step": 6490 }, { "epoch": 0.7037077189939289, "grad_norm": 1.5471882820129395, "learning_rate": 1.5364137530828377e-05, "loss": 0.0389, "step": 6491 }, { "epoch": 0.7038161318300087, "grad_norm": 0.8161094784736633, "learning_rate": 1.536341215726099e-05, "loss": 0.0119, "step": 6492 }, { "epoch": 0.7039245446660884, "grad_norm": 2.082202434539795, "learning_rate": 1.5362686783693602e-05, "loss": 0.0219, "step": 6493 }, { "epoch": 0.7040329575021682, "grad_norm": 6.641467571258545, "learning_rate": 1.5361961410126217e-05, "loss": 0.0494, "step": 6494 }, { "epoch": 0.7041413703382481, "grad_norm": 1.1208758354187012, "learning_rate": 1.536123603655883e-05, "loss": 0.0254, "step": 6495 }, { "epoch": 0.7042497831743278, "grad_norm": 0.3515578508377075, "learning_rate": 1.536051066299144e-05, "loss": 0.0026, "step": 6496 }, { "epoch": 0.7043581960104076, "grad_norm": 2.8969616889953613, "learning_rate": 1.5359785289424054e-05, "loss": 0.0282, "step": 6497 }, { "epoch": 0.7044666088464874, "grad_norm": 0.6529015302658081, "learning_rate": 1.5359059915856666e-05, "loss": 0.0156, "step": 6498 }, { "epoch": 0.7045750216825672, "grad_norm": 1.4913356304168701, "learning_rate": 1.5358334542289278e-05, "loss": 0.0269, "step": 6499 }, { "epoch": 0.704683434518647, "grad_norm": 1.0970885753631592, "learning_rate": 1.5357609168721894e-05, "loss": 0.0358, "step": 6500 }, { "epoch": 0.7047918473547268, "grad_norm": 3.267869710922241, "learning_rate": 1.5356883795154506e-05, "loss": 0.0298, "step": 6501 }, { "epoch": 0.7049002601908065, "grad_norm": 2.3354744911193848, "learning_rate": 1.535615842158712e-05, "loss": 0.0726, "step": 6502 }, { "epoch": 0.7050086730268864, "grad_norm": 0.3514156639575958, "learning_rate": 1.5355433048019733e-05, "loss": 0.0036, "step": 6503 }, { "epoch": 0.7051170858629662, "grad_norm": 3.5856497287750244, "learning_rate": 1.5354707674452345e-05, "loss": 0.0623, "step": 6504 }, { "epoch": 0.7052254986990459, "grad_norm": 2.8140265941619873, "learning_rate": 1.5353982300884958e-05, "loss": 0.0571, "step": 6505 }, { "epoch": 0.7053339115351258, "grad_norm": 1.670891284942627, "learning_rate": 1.535325692731757e-05, "loss": 0.0677, "step": 6506 }, { "epoch": 0.7054423243712056, "grad_norm": 3.723348617553711, "learning_rate": 1.5352531553750182e-05, "loss": 0.0486, "step": 6507 }, { "epoch": 0.7055507372072853, "grad_norm": 6.534306526184082, "learning_rate": 1.5351806180182794e-05, "loss": 0.0758, "step": 6508 }, { "epoch": 0.7056591500433651, "grad_norm": 1.6185556650161743, "learning_rate": 1.535108080661541e-05, "loss": 0.0133, "step": 6509 }, { "epoch": 0.705767562879445, "grad_norm": 5.4790120124816895, "learning_rate": 1.5350355433048022e-05, "loss": 0.0637, "step": 6510 }, { "epoch": 0.7058759757155247, "grad_norm": 1.0995310544967651, "learning_rate": 1.5349630059480634e-05, "loss": 0.0157, "step": 6511 }, { "epoch": 0.7059843885516045, "grad_norm": 2.2177226543426514, "learning_rate": 1.5348904685913246e-05, "loss": 0.0281, "step": 6512 }, { "epoch": 0.7060928013876843, "grad_norm": 4.603241920471191, "learning_rate": 1.5348179312345858e-05, "loss": 0.0813, "step": 6513 }, { "epoch": 0.7062012142237641, "grad_norm": 5.857466697692871, "learning_rate": 1.534745393877847e-05, "loss": 0.0875, "step": 6514 }, { "epoch": 0.7063096270598439, "grad_norm": 0.9873718023300171, "learning_rate": 1.5346728565211086e-05, "loss": 0.0285, "step": 6515 }, { "epoch": 0.7064180398959237, "grad_norm": 5.3144073486328125, "learning_rate": 1.5346003191643698e-05, "loss": 0.0482, "step": 6516 }, { "epoch": 0.7065264527320034, "grad_norm": 4.566690921783447, "learning_rate": 1.534527781807631e-05, "loss": 0.0848, "step": 6517 }, { "epoch": 0.7066348655680833, "grad_norm": 1.1458688974380493, "learning_rate": 1.5344552444508926e-05, "loss": 0.011, "step": 6518 }, { "epoch": 0.7067432784041631, "grad_norm": 2.7065622806549072, "learning_rate": 1.5343827070941538e-05, "loss": 0.0581, "step": 6519 }, { "epoch": 0.7068516912402428, "grad_norm": 5.24940299987793, "learning_rate": 1.534310169737415e-05, "loss": 0.0676, "step": 6520 }, { "epoch": 0.7069601040763226, "grad_norm": 1.8807958364486694, "learning_rate": 1.5342376323806762e-05, "loss": 0.0598, "step": 6521 }, { "epoch": 0.7070685169124025, "grad_norm": 2.693732976913452, "learning_rate": 1.5341650950239374e-05, "loss": 0.0498, "step": 6522 }, { "epoch": 0.7071769297484822, "grad_norm": 2.671246290206909, "learning_rate": 1.5340925576671986e-05, "loss": 0.0226, "step": 6523 }, { "epoch": 0.707285342584562, "grad_norm": 3.7498488426208496, "learning_rate": 1.5340200203104602e-05, "loss": 0.0427, "step": 6524 }, { "epoch": 0.7073937554206418, "grad_norm": 2.2917582988739014, "learning_rate": 1.5339474829537214e-05, "loss": 0.0514, "step": 6525 }, { "epoch": 0.7075021682567216, "grad_norm": 1.2115800380706787, "learning_rate": 1.5338749455969826e-05, "loss": 0.0184, "step": 6526 }, { "epoch": 0.7076105810928014, "grad_norm": 0.7308741211891174, "learning_rate": 1.533802408240244e-05, "loss": 0.0336, "step": 6527 }, { "epoch": 0.7077189939288812, "grad_norm": 2.528709650039673, "learning_rate": 1.533729870883505e-05, "loss": 0.0521, "step": 6528 }, { "epoch": 0.7078274067649609, "grad_norm": 1.1127679347991943, "learning_rate": 1.5336573335267663e-05, "loss": 0.0131, "step": 6529 }, { "epoch": 0.7079358196010408, "grad_norm": 1.376329779624939, "learning_rate": 1.5335847961700278e-05, "loss": 0.0254, "step": 6530 }, { "epoch": 0.7080442324371206, "grad_norm": 3.1843841075897217, "learning_rate": 1.533512258813289e-05, "loss": 0.0197, "step": 6531 }, { "epoch": 0.7081526452732003, "grad_norm": 2.2998299598693848, "learning_rate": 1.5334397214565502e-05, "loss": 0.0186, "step": 6532 }, { "epoch": 0.7082610581092801, "grad_norm": 1.71563720703125, "learning_rate": 1.5333671840998115e-05, "loss": 0.0372, "step": 6533 }, { "epoch": 0.70836947094536, "grad_norm": 2.2548885345458984, "learning_rate": 1.5332946467430727e-05, "loss": 0.0839, "step": 6534 }, { "epoch": 0.7084778837814397, "grad_norm": 2.028604507446289, "learning_rate": 1.533222109386334e-05, "loss": 0.0417, "step": 6535 }, { "epoch": 0.7085862966175195, "grad_norm": 1.1069533824920654, "learning_rate": 1.5331495720295954e-05, "loss": 0.0174, "step": 6536 }, { "epoch": 0.7086947094535994, "grad_norm": 4.092308521270752, "learning_rate": 1.5330770346728567e-05, "loss": 0.0693, "step": 6537 }, { "epoch": 0.7088031222896791, "grad_norm": 2.8349714279174805, "learning_rate": 1.533004497316118e-05, "loss": 0.0569, "step": 6538 }, { "epoch": 0.7089115351257589, "grad_norm": 1.422399878501892, "learning_rate": 1.5329319599593794e-05, "loss": 0.0198, "step": 6539 }, { "epoch": 0.7090199479618386, "grad_norm": 1.0999252796173096, "learning_rate": 1.5328594226026406e-05, "loss": 0.0137, "step": 6540 }, { "epoch": 0.7091283607979185, "grad_norm": 1.417082667350769, "learning_rate": 1.532786885245902e-05, "loss": 0.0319, "step": 6541 }, { "epoch": 0.7092367736339983, "grad_norm": 5.852000713348389, "learning_rate": 1.532714347889163e-05, "loss": 0.024, "step": 6542 }, { "epoch": 0.709345186470078, "grad_norm": 1.739246129989624, "learning_rate": 1.5326418105324243e-05, "loss": 0.0258, "step": 6543 }, { "epoch": 0.7094535993061578, "grad_norm": 4.394303321838379, "learning_rate": 1.5325692731756855e-05, "loss": 0.0276, "step": 6544 }, { "epoch": 0.7095620121422377, "grad_norm": 3.6306312084198, "learning_rate": 1.5324967358189467e-05, "loss": 0.0448, "step": 6545 }, { "epoch": 0.7096704249783174, "grad_norm": 0.5439679026603699, "learning_rate": 1.5324241984622083e-05, "loss": 0.0093, "step": 6546 }, { "epoch": 0.7097788378143972, "grad_norm": 1.0613421201705933, "learning_rate": 1.5323516611054695e-05, "loss": 0.0178, "step": 6547 }, { "epoch": 0.709887250650477, "grad_norm": 1.7910246849060059, "learning_rate": 1.5322791237487307e-05, "loss": 0.0241, "step": 6548 }, { "epoch": 0.7099956634865568, "grad_norm": 0.8796201348304749, "learning_rate": 1.532206586391992e-05, "loss": 0.0184, "step": 6549 }, { "epoch": 0.7101040763226366, "grad_norm": 9.965840339660645, "learning_rate": 1.532134049035253e-05, "loss": 0.0857, "step": 6550 }, { "epoch": 0.7102124891587164, "grad_norm": 4.29813814163208, "learning_rate": 1.5320615116785143e-05, "loss": 0.0498, "step": 6551 }, { "epoch": 0.7103209019947961, "grad_norm": 1.1273037195205688, "learning_rate": 1.531988974321776e-05, "loss": 0.0172, "step": 6552 }, { "epoch": 0.710429314830876, "grad_norm": 3.5021631717681885, "learning_rate": 1.531916436965037e-05, "loss": 0.0483, "step": 6553 }, { "epoch": 0.7105377276669558, "grad_norm": 2.7613415718078613, "learning_rate": 1.5318438996082987e-05, "loss": 0.0361, "step": 6554 }, { "epoch": 0.7106461405030355, "grad_norm": 1.3818682432174683, "learning_rate": 1.53177136225156e-05, "loss": 0.0209, "step": 6555 }, { "epoch": 0.7107545533391153, "grad_norm": 4.460722923278809, "learning_rate": 1.531698824894821e-05, "loss": 0.025, "step": 6556 }, { "epoch": 0.7108629661751952, "grad_norm": 4.727636337280273, "learning_rate": 1.5316262875380823e-05, "loss": 0.0333, "step": 6557 }, { "epoch": 0.7109713790112749, "grad_norm": 1.6987501382827759, "learning_rate": 1.5315537501813435e-05, "loss": 0.0252, "step": 6558 }, { "epoch": 0.7110797918473547, "grad_norm": 5.259948253631592, "learning_rate": 1.5314812128246047e-05, "loss": 0.0502, "step": 6559 }, { "epoch": 0.7111882046834345, "grad_norm": 4.415209770202637, "learning_rate": 1.531408675467866e-05, "loss": 0.1166, "step": 6560 }, { "epoch": 0.7112966175195143, "grad_norm": 3.492880344390869, "learning_rate": 1.5313361381111275e-05, "loss": 0.0375, "step": 6561 }, { "epoch": 0.7114050303555941, "grad_norm": 2.192490816116333, "learning_rate": 1.5312636007543887e-05, "loss": 0.0187, "step": 6562 }, { "epoch": 0.7115134431916739, "grad_norm": 1.7467138767242432, "learning_rate": 1.53119106339765e-05, "loss": 0.0208, "step": 6563 }, { "epoch": 0.7116218560277537, "grad_norm": 2.7407000064849854, "learning_rate": 1.531118526040911e-05, "loss": 0.0798, "step": 6564 }, { "epoch": 0.7117302688638335, "grad_norm": 1.23361337184906, "learning_rate": 1.5310459886841723e-05, "loss": 0.0054, "step": 6565 }, { "epoch": 0.7118386816999133, "grad_norm": 1.189759373664856, "learning_rate": 1.5309734513274336e-05, "loss": 0.0398, "step": 6566 }, { "epoch": 0.711947094535993, "grad_norm": 4.952749729156494, "learning_rate": 1.530900913970695e-05, "loss": 0.04, "step": 6567 }, { "epoch": 0.7120555073720729, "grad_norm": 0.5539998412132263, "learning_rate": 1.5308283766139563e-05, "loss": 0.0197, "step": 6568 }, { "epoch": 0.7121639202081527, "grad_norm": 0.6113380789756775, "learning_rate": 1.5307558392572175e-05, "loss": 0.0082, "step": 6569 }, { "epoch": 0.7122723330442324, "grad_norm": 3.788027763366699, "learning_rate": 1.530683301900479e-05, "loss": 0.0775, "step": 6570 }, { "epoch": 0.7123807458803122, "grad_norm": 3.582916498184204, "learning_rate": 1.5306107645437403e-05, "loss": 0.0494, "step": 6571 }, { "epoch": 0.7124891587163921, "grad_norm": 2.5287983417510986, "learning_rate": 1.5305382271870015e-05, "loss": 0.0655, "step": 6572 }, { "epoch": 0.7125975715524718, "grad_norm": 5.600586891174316, "learning_rate": 1.5304656898302627e-05, "loss": 0.0855, "step": 6573 }, { "epoch": 0.7127059843885516, "grad_norm": 1.234372854232788, "learning_rate": 1.530393152473524e-05, "loss": 0.0414, "step": 6574 }, { "epoch": 0.7128143972246314, "grad_norm": 3.7036361694335938, "learning_rate": 1.5303206151167852e-05, "loss": 0.0247, "step": 6575 }, { "epoch": 0.7129228100607112, "grad_norm": 2.0548412799835205, "learning_rate": 1.5302480777600467e-05, "loss": 0.0104, "step": 6576 }, { "epoch": 0.713031222896791, "grad_norm": 1.7008062601089478, "learning_rate": 1.530175540403308e-05, "loss": 0.0375, "step": 6577 }, { "epoch": 0.7131396357328708, "grad_norm": 3.166505813598633, "learning_rate": 1.530103003046569e-05, "loss": 0.0313, "step": 6578 }, { "epoch": 0.7132480485689505, "grad_norm": 1.3687540292739868, "learning_rate": 1.5300304656898304e-05, "loss": 0.0235, "step": 6579 }, { "epoch": 0.7133564614050304, "grad_norm": 0.5985596179962158, "learning_rate": 1.5299579283330916e-05, "loss": 0.0211, "step": 6580 }, { "epoch": 0.7134648742411102, "grad_norm": 2.3587591648101807, "learning_rate": 1.5298853909763528e-05, "loss": 0.0356, "step": 6581 }, { "epoch": 0.7135732870771899, "grad_norm": 1.1008851528167725, "learning_rate": 1.5298128536196143e-05, "loss": 0.0074, "step": 6582 }, { "epoch": 0.7136816999132697, "grad_norm": 2.1202399730682373, "learning_rate": 1.5297403162628756e-05, "loss": 0.007, "step": 6583 }, { "epoch": 0.7137901127493496, "grad_norm": 2.666534900665283, "learning_rate": 1.5296677789061368e-05, "loss": 0.0712, "step": 6584 }, { "epoch": 0.7138985255854293, "grad_norm": 2.780245304107666, "learning_rate": 1.529595241549398e-05, "loss": 0.036, "step": 6585 }, { "epoch": 0.7140069384215091, "grad_norm": 5.632691383361816, "learning_rate": 1.5295227041926592e-05, "loss": 0.0956, "step": 6586 }, { "epoch": 0.7141153512575888, "grad_norm": 4.722858428955078, "learning_rate": 1.5294501668359204e-05, "loss": 0.0247, "step": 6587 }, { "epoch": 0.7142237640936687, "grad_norm": 2.119436025619507, "learning_rate": 1.529377629479182e-05, "loss": 0.0099, "step": 6588 }, { "epoch": 0.7143321769297485, "grad_norm": 1.0697243213653564, "learning_rate": 1.5293050921224432e-05, "loss": 0.0379, "step": 6589 }, { "epoch": 0.7144405897658282, "grad_norm": 0.35640949010849, "learning_rate": 1.5292325547657044e-05, "loss": 0.004, "step": 6590 }, { "epoch": 0.714549002601908, "grad_norm": 3.191392421722412, "learning_rate": 1.529160017408966e-05, "loss": 0.0459, "step": 6591 }, { "epoch": 0.7146574154379879, "grad_norm": 1.2639495134353638, "learning_rate": 1.529087480052227e-05, "loss": 0.034, "step": 6592 }, { "epoch": 0.7147658282740676, "grad_norm": 0.6702236533164978, "learning_rate": 1.5290149426954884e-05, "loss": 0.0089, "step": 6593 }, { "epoch": 0.7148742411101474, "grad_norm": 5.834944725036621, "learning_rate": 1.5289424053387496e-05, "loss": 0.0672, "step": 6594 }, { "epoch": 0.7149826539462273, "grad_norm": 1.51447594165802, "learning_rate": 1.5288698679820108e-05, "loss": 0.0403, "step": 6595 }, { "epoch": 0.715091066782307, "grad_norm": 1.8710063695907593, "learning_rate": 1.528797330625272e-05, "loss": 0.0245, "step": 6596 }, { "epoch": 0.7151994796183868, "grad_norm": 1.6418697834014893, "learning_rate": 1.5287247932685332e-05, "loss": 0.0169, "step": 6597 }, { "epoch": 0.7153078924544666, "grad_norm": 3.692732095718384, "learning_rate": 1.5286522559117948e-05, "loss": 0.0221, "step": 6598 }, { "epoch": 0.7154163052905465, "grad_norm": 2.742082118988037, "learning_rate": 1.528579718555056e-05, "loss": 0.0313, "step": 6599 }, { "epoch": 0.7155247181266262, "grad_norm": 5.708698272705078, "learning_rate": 1.5285071811983172e-05, "loss": 0.0168, "step": 6600 }, { "epoch": 0.715633130962706, "grad_norm": 3.8536877632141113, "learning_rate": 1.5284346438415784e-05, "loss": 0.0772, "step": 6601 }, { "epoch": 0.7157415437987857, "grad_norm": 3.5791213512420654, "learning_rate": 1.5283621064848397e-05, "loss": 0.046, "step": 6602 }, { "epoch": 0.7158499566348656, "grad_norm": 4.03754997253418, "learning_rate": 1.528289569128101e-05, "loss": 0.0506, "step": 6603 }, { "epoch": 0.7159583694709454, "grad_norm": 2.0892627239227295, "learning_rate": 1.5282170317713624e-05, "loss": 0.011, "step": 6604 }, { "epoch": 0.7160667823070251, "grad_norm": 2.939044237136841, "learning_rate": 1.5281444944146236e-05, "loss": 0.0248, "step": 6605 }, { "epoch": 0.7161751951431049, "grad_norm": 0.5315196514129639, "learning_rate": 1.5280719570578852e-05, "loss": 0.0027, "step": 6606 }, { "epoch": 0.7162836079791848, "grad_norm": 6.143918991088867, "learning_rate": 1.5279994197011464e-05, "loss": 0.1111, "step": 6607 }, { "epoch": 0.7163920208152645, "grad_norm": 0.775556743144989, "learning_rate": 1.5279268823444076e-05, "loss": 0.0179, "step": 6608 }, { "epoch": 0.7165004336513443, "grad_norm": 6.8943562507629395, "learning_rate": 1.5278543449876688e-05, "loss": 0.0563, "step": 6609 }, { "epoch": 0.7166088464874241, "grad_norm": 2.87412166595459, "learning_rate": 1.52778180763093e-05, "loss": 0.0457, "step": 6610 }, { "epoch": 0.7167172593235039, "grad_norm": 10.688435554504395, "learning_rate": 1.5277092702741913e-05, "loss": 0.0402, "step": 6611 }, { "epoch": 0.7168256721595837, "grad_norm": 2.9358718395233154, "learning_rate": 1.5276367329174525e-05, "loss": 0.0218, "step": 6612 }, { "epoch": 0.7169340849956635, "grad_norm": 0.4974524974822998, "learning_rate": 1.527564195560714e-05, "loss": 0.0059, "step": 6613 }, { "epoch": 0.7170424978317432, "grad_norm": 3.5267317295074463, "learning_rate": 1.5274916582039752e-05, "loss": 0.0453, "step": 6614 }, { "epoch": 0.7171509106678231, "grad_norm": 2.3117306232452393, "learning_rate": 1.5274191208472365e-05, "loss": 0.0812, "step": 6615 }, { "epoch": 0.7172593235039029, "grad_norm": 1.6414228677749634, "learning_rate": 1.5273465834904977e-05, "loss": 0.0107, "step": 6616 }, { "epoch": 0.7173677363399826, "grad_norm": 4.014375686645508, "learning_rate": 1.527274046133759e-05, "loss": 0.0279, "step": 6617 }, { "epoch": 0.7174761491760624, "grad_norm": 4.931699752807617, "learning_rate": 1.52720150877702e-05, "loss": 0.0417, "step": 6618 }, { "epoch": 0.7175845620121423, "grad_norm": 0.1062442883849144, "learning_rate": 1.5271289714202817e-05, "loss": 0.0012, "step": 6619 }, { "epoch": 0.717692974848222, "grad_norm": 1.87273371219635, "learning_rate": 1.527056434063543e-05, "loss": 0.0805, "step": 6620 }, { "epoch": 0.7178013876843018, "grad_norm": 0.33420512080192566, "learning_rate": 1.526983896706804e-05, "loss": 0.0026, "step": 6621 }, { "epoch": 0.7179098005203816, "grad_norm": 0.9792523980140686, "learning_rate": 1.5269113593500656e-05, "loss": 0.0149, "step": 6622 }, { "epoch": 0.7180182133564614, "grad_norm": 0.9325674176216125, "learning_rate": 1.526838821993327e-05, "loss": 0.0054, "step": 6623 }, { "epoch": 0.7181266261925412, "grad_norm": 2.9477014541625977, "learning_rate": 1.526766284636588e-05, "loss": 0.0235, "step": 6624 }, { "epoch": 0.718235039028621, "grad_norm": 2.802672863006592, "learning_rate": 1.5266937472798493e-05, "loss": 0.0407, "step": 6625 }, { "epoch": 0.7183434518647008, "grad_norm": 2.0470566749572754, "learning_rate": 1.5266212099231105e-05, "loss": 0.0202, "step": 6626 }, { "epoch": 0.7184518647007806, "grad_norm": 5.238465309143066, "learning_rate": 1.5265486725663717e-05, "loss": 0.0647, "step": 6627 }, { "epoch": 0.7185602775368604, "grad_norm": 1.2049286365509033, "learning_rate": 1.5264761352096333e-05, "loss": 0.0182, "step": 6628 }, { "epoch": 0.7186686903729401, "grad_norm": 3.9633891582489014, "learning_rate": 1.5264035978528945e-05, "loss": 0.0824, "step": 6629 }, { "epoch": 0.71877710320902, "grad_norm": 0.8661350607872009, "learning_rate": 1.5263310604961557e-05, "loss": 0.0264, "step": 6630 }, { "epoch": 0.7188855160450998, "grad_norm": 0.95668625831604, "learning_rate": 1.526258523139417e-05, "loss": 0.0093, "step": 6631 }, { "epoch": 0.7189939288811795, "grad_norm": 0.40953654050827026, "learning_rate": 1.526185985782678e-05, "loss": 0.0075, "step": 6632 }, { "epoch": 0.7191023417172593, "grad_norm": 3.557279109954834, "learning_rate": 1.5261134484259393e-05, "loss": 0.0206, "step": 6633 }, { "epoch": 0.7192107545533392, "grad_norm": 9.330904006958008, "learning_rate": 1.526040911069201e-05, "loss": 0.0573, "step": 6634 }, { "epoch": 0.7193191673894189, "grad_norm": 8.466632843017578, "learning_rate": 1.525968373712462e-05, "loss": 0.0284, "step": 6635 }, { "epoch": 0.7194275802254987, "grad_norm": 2.0038177967071533, "learning_rate": 1.5258958363557233e-05, "loss": 0.0172, "step": 6636 }, { "epoch": 0.7195359930615784, "grad_norm": 0.8684471845626831, "learning_rate": 1.5258232989989845e-05, "loss": 0.0138, "step": 6637 }, { "epoch": 0.7196444058976583, "grad_norm": 2.5750277042388916, "learning_rate": 1.5257507616422457e-05, "loss": 0.0375, "step": 6638 }, { "epoch": 0.7197528187337381, "grad_norm": 11.579888343811035, "learning_rate": 1.5256782242855073e-05, "loss": 0.0467, "step": 6639 }, { "epoch": 0.7198612315698178, "grad_norm": 6.618727207183838, "learning_rate": 1.5256056869287685e-05, "loss": 0.0644, "step": 6640 }, { "epoch": 0.7199696444058976, "grad_norm": 4.956480026245117, "learning_rate": 1.5255331495720297e-05, "loss": 0.0794, "step": 6641 }, { "epoch": 0.7200780572419775, "grad_norm": 5.960089683532715, "learning_rate": 1.5254606122152911e-05, "loss": 0.0851, "step": 6642 }, { "epoch": 0.7201864700780573, "grad_norm": 2.725215435028076, "learning_rate": 1.5253880748585523e-05, "loss": 0.058, "step": 6643 }, { "epoch": 0.720294882914137, "grad_norm": 3.517139196395874, "learning_rate": 1.5253155375018135e-05, "loss": 0.0656, "step": 6644 }, { "epoch": 0.7204032957502168, "grad_norm": 4.016201496124268, "learning_rate": 1.525243000145075e-05, "loss": 0.0533, "step": 6645 }, { "epoch": 0.7205117085862967, "grad_norm": 1.912377953529358, "learning_rate": 1.5251704627883361e-05, "loss": 0.0229, "step": 6646 }, { "epoch": 0.7206201214223764, "grad_norm": 0.8599995374679565, "learning_rate": 1.5250979254315973e-05, "loss": 0.014, "step": 6647 }, { "epoch": 0.7207285342584562, "grad_norm": 6.520524978637695, "learning_rate": 1.5250253880748587e-05, "loss": 0.0658, "step": 6648 }, { "epoch": 0.7208369470945359, "grad_norm": 1.6879431009292603, "learning_rate": 1.52495285071812e-05, "loss": 0.0481, "step": 6649 }, { "epoch": 0.7209453599306158, "grad_norm": 4.349456787109375, "learning_rate": 1.5248803133613812e-05, "loss": 0.0507, "step": 6650 }, { "epoch": 0.7210537727666956, "grad_norm": 3.7911787033081055, "learning_rate": 1.5248077760046425e-05, "loss": 0.0442, "step": 6651 }, { "epoch": 0.7211621856027753, "grad_norm": 3.4214179515838623, "learning_rate": 1.5247352386479038e-05, "loss": 0.0549, "step": 6652 }, { "epoch": 0.7212705984388552, "grad_norm": 2.925508499145508, "learning_rate": 1.524662701291165e-05, "loss": 0.0552, "step": 6653 }, { "epoch": 0.721379011274935, "grad_norm": 1.2326672077178955, "learning_rate": 1.5245901639344264e-05, "loss": 0.0243, "step": 6654 }, { "epoch": 0.7214874241110147, "grad_norm": 2.2055437564849854, "learning_rate": 1.5245176265776876e-05, "loss": 0.019, "step": 6655 }, { "epoch": 0.7215958369470945, "grad_norm": 3.859755277633667, "learning_rate": 1.5244450892209488e-05, "loss": 0.0796, "step": 6656 }, { "epoch": 0.7217042497831744, "grad_norm": 1.8456016778945923, "learning_rate": 1.5243725518642103e-05, "loss": 0.0363, "step": 6657 }, { "epoch": 0.7218126626192541, "grad_norm": 1.314358115196228, "learning_rate": 1.5243000145074716e-05, "loss": 0.0189, "step": 6658 }, { "epoch": 0.7219210754553339, "grad_norm": 2.3606326580047607, "learning_rate": 1.5242274771507328e-05, "loss": 0.0452, "step": 6659 }, { "epoch": 0.7220294882914137, "grad_norm": 1.7670958042144775, "learning_rate": 1.5241549397939941e-05, "loss": 0.0508, "step": 6660 }, { "epoch": 0.7221379011274935, "grad_norm": 2.569084644317627, "learning_rate": 1.5240824024372554e-05, "loss": 0.0414, "step": 6661 }, { "epoch": 0.7222463139635733, "grad_norm": 2.2536654472351074, "learning_rate": 1.5240098650805166e-05, "loss": 0.0558, "step": 6662 }, { "epoch": 0.7223547267996531, "grad_norm": 0.5541701912879944, "learning_rate": 1.523937327723778e-05, "loss": 0.0094, "step": 6663 }, { "epoch": 0.7224631396357328, "grad_norm": 2.096566915512085, "learning_rate": 1.5238647903670392e-05, "loss": 0.0533, "step": 6664 }, { "epoch": 0.7225715524718127, "grad_norm": 1.636957049369812, "learning_rate": 1.5237922530103004e-05, "loss": 0.0336, "step": 6665 }, { "epoch": 0.7226799653078925, "grad_norm": 2.5205495357513428, "learning_rate": 1.5237197156535618e-05, "loss": 0.0239, "step": 6666 }, { "epoch": 0.7227883781439722, "grad_norm": 4.971327781677246, "learning_rate": 1.523647178296823e-05, "loss": 0.0628, "step": 6667 }, { "epoch": 0.722896790980052, "grad_norm": 0.927675187587738, "learning_rate": 1.5235746409400842e-05, "loss": 0.0132, "step": 6668 }, { "epoch": 0.7230052038161319, "grad_norm": 1.92433762550354, "learning_rate": 1.5235021035833454e-05, "loss": 0.0232, "step": 6669 }, { "epoch": 0.7231136166522116, "grad_norm": 0.5019639730453491, "learning_rate": 1.5234295662266068e-05, "loss": 0.0092, "step": 6670 }, { "epoch": 0.7232220294882914, "grad_norm": 3.488759756088257, "learning_rate": 1.523357028869868e-05, "loss": 0.1109, "step": 6671 }, { "epoch": 0.7233304423243712, "grad_norm": 1.8802345991134644, "learning_rate": 1.5232844915131292e-05, "loss": 0.0298, "step": 6672 }, { "epoch": 0.723438855160451, "grad_norm": 2.277339458465576, "learning_rate": 1.5232119541563906e-05, "loss": 0.0332, "step": 6673 }, { "epoch": 0.7235472679965308, "grad_norm": 0.4580039978027344, "learning_rate": 1.523139416799652e-05, "loss": 0.0127, "step": 6674 }, { "epoch": 0.7236556808326106, "grad_norm": 0.5913442969322205, "learning_rate": 1.5230668794429134e-05, "loss": 0.0052, "step": 6675 }, { "epoch": 0.7237640936686903, "grad_norm": 1.4803388118743896, "learning_rate": 1.5229943420861746e-05, "loss": 0.0188, "step": 6676 }, { "epoch": 0.7238725065047702, "grad_norm": 2.803704261779785, "learning_rate": 1.5229218047294358e-05, "loss": 0.0495, "step": 6677 }, { "epoch": 0.72398091934085, "grad_norm": 1.4103316068649292, "learning_rate": 1.5228492673726972e-05, "loss": 0.0222, "step": 6678 }, { "epoch": 0.7240893321769297, "grad_norm": 1.499503493309021, "learning_rate": 1.5227767300159584e-05, "loss": 0.0592, "step": 6679 }, { "epoch": 0.7241977450130095, "grad_norm": 3.3911619186401367, "learning_rate": 1.5227041926592196e-05, "loss": 0.0413, "step": 6680 }, { "epoch": 0.7243061578490894, "grad_norm": 2.8109230995178223, "learning_rate": 1.5226316553024808e-05, "loss": 0.0478, "step": 6681 }, { "epoch": 0.7244145706851691, "grad_norm": 3.060147285461426, "learning_rate": 1.5225591179457422e-05, "loss": 0.0392, "step": 6682 }, { "epoch": 0.7245229835212489, "grad_norm": 10.247428894042969, "learning_rate": 1.5224865805890034e-05, "loss": 0.0658, "step": 6683 }, { "epoch": 0.7246313963573288, "grad_norm": 3.4150002002716064, "learning_rate": 1.5224140432322646e-05, "loss": 0.0512, "step": 6684 }, { "epoch": 0.7247398091934085, "grad_norm": 1.0769367218017578, "learning_rate": 1.522341505875526e-05, "loss": 0.0131, "step": 6685 }, { "epoch": 0.7248482220294883, "grad_norm": 2.998500108718872, "learning_rate": 1.5222689685187872e-05, "loss": 0.0428, "step": 6686 }, { "epoch": 0.724956634865568, "grad_norm": 2.043710708618164, "learning_rate": 1.5221964311620485e-05, "loss": 0.0428, "step": 6687 }, { "epoch": 0.7250650477016479, "grad_norm": 4.095914363861084, "learning_rate": 1.5221238938053098e-05, "loss": 0.0606, "step": 6688 }, { "epoch": 0.7251734605377277, "grad_norm": 3.143038749694824, "learning_rate": 1.522051356448571e-05, "loss": 0.0491, "step": 6689 }, { "epoch": 0.7252818733738075, "grad_norm": 14.982831954956055, "learning_rate": 1.5219788190918323e-05, "loss": 0.0199, "step": 6690 }, { "epoch": 0.7253902862098872, "grad_norm": 6.033532619476318, "learning_rate": 1.5219062817350938e-05, "loss": 0.0585, "step": 6691 }, { "epoch": 0.7254986990459671, "grad_norm": 1.7372468709945679, "learning_rate": 1.521833744378355e-05, "loss": 0.0397, "step": 6692 }, { "epoch": 0.7256071118820469, "grad_norm": 2.8381869792938232, "learning_rate": 1.5217612070216163e-05, "loss": 0.0228, "step": 6693 }, { "epoch": 0.7257155247181266, "grad_norm": 2.161670684814453, "learning_rate": 1.5216886696648776e-05, "loss": 0.0424, "step": 6694 }, { "epoch": 0.7258239375542064, "grad_norm": 0.6472815275192261, "learning_rate": 1.5216161323081389e-05, "loss": 0.0073, "step": 6695 }, { "epoch": 0.7259323503902863, "grad_norm": 1.207442283630371, "learning_rate": 1.5215435949514e-05, "loss": 0.0144, "step": 6696 }, { "epoch": 0.726040763226366, "grad_norm": 3.2431278228759766, "learning_rate": 1.5214710575946615e-05, "loss": 0.049, "step": 6697 }, { "epoch": 0.7261491760624458, "grad_norm": 2.5647950172424316, "learning_rate": 1.5213985202379227e-05, "loss": 0.0542, "step": 6698 }, { "epoch": 0.7262575888985255, "grad_norm": 10.296209335327148, "learning_rate": 1.5213259828811839e-05, "loss": 0.0583, "step": 6699 }, { "epoch": 0.7263660017346054, "grad_norm": 5.6246256828308105, "learning_rate": 1.5212534455244453e-05, "loss": 0.0263, "step": 6700 }, { "epoch": 0.7264744145706852, "grad_norm": 1.8339940309524536, "learning_rate": 1.5211809081677065e-05, "loss": 0.0207, "step": 6701 }, { "epoch": 0.7265828274067649, "grad_norm": 2.2208573818206787, "learning_rate": 1.5211083708109677e-05, "loss": 0.0237, "step": 6702 }, { "epoch": 0.7266912402428447, "grad_norm": 4.170096397399902, "learning_rate": 1.521035833454229e-05, "loss": 0.0752, "step": 6703 }, { "epoch": 0.7267996530789246, "grad_norm": 6.463716506958008, "learning_rate": 1.5209632960974903e-05, "loss": 0.0592, "step": 6704 }, { "epoch": 0.7269080659150043, "grad_norm": 3.226668357849121, "learning_rate": 1.5208907587407515e-05, "loss": 0.0248, "step": 6705 }, { "epoch": 0.7270164787510841, "grad_norm": 2.221557855606079, "learning_rate": 1.5208182213840129e-05, "loss": 0.0189, "step": 6706 }, { "epoch": 0.7271248915871639, "grad_norm": 1.2005062103271484, "learning_rate": 1.5207456840272741e-05, "loss": 0.0274, "step": 6707 }, { "epoch": 0.7272333044232437, "grad_norm": 2.4389476776123047, "learning_rate": 1.5206731466705355e-05, "loss": 0.0532, "step": 6708 }, { "epoch": 0.7273417172593235, "grad_norm": 4.120566368103027, "learning_rate": 1.5206006093137969e-05, "loss": 0.0935, "step": 6709 }, { "epoch": 0.7274501300954033, "grad_norm": 4.159060478210449, "learning_rate": 1.520528071957058e-05, "loss": 0.036, "step": 6710 }, { "epoch": 0.727558542931483, "grad_norm": 6.468141555786133, "learning_rate": 1.5204555346003193e-05, "loss": 0.0514, "step": 6711 }, { "epoch": 0.7276669557675629, "grad_norm": 2.875736713409424, "learning_rate": 1.5203829972435807e-05, "loss": 0.0273, "step": 6712 }, { "epoch": 0.7277753686036427, "grad_norm": 2.519009590148926, "learning_rate": 1.5203104598868419e-05, "loss": 0.0188, "step": 6713 }, { "epoch": 0.7278837814397224, "grad_norm": 0.8686773777008057, "learning_rate": 1.5202379225301031e-05, "loss": 0.0247, "step": 6714 }, { "epoch": 0.7279921942758023, "grad_norm": 5.342012882232666, "learning_rate": 1.5201653851733645e-05, "loss": 0.0818, "step": 6715 }, { "epoch": 0.7281006071118821, "grad_norm": 1.3177714347839355, "learning_rate": 1.5200928478166257e-05, "loss": 0.0405, "step": 6716 }, { "epoch": 0.7282090199479618, "grad_norm": 2.325274705886841, "learning_rate": 1.520020310459887e-05, "loss": 0.0688, "step": 6717 }, { "epoch": 0.7283174327840416, "grad_norm": 2.2220139503479004, "learning_rate": 1.5199477731031483e-05, "loss": 0.0511, "step": 6718 }, { "epoch": 0.7284258456201215, "grad_norm": 3.696354627609253, "learning_rate": 1.5198752357464095e-05, "loss": 0.0425, "step": 6719 }, { "epoch": 0.7285342584562012, "grad_norm": 2.123810052871704, "learning_rate": 1.5198026983896707e-05, "loss": 0.0409, "step": 6720 }, { "epoch": 0.728642671292281, "grad_norm": 0.9324665069580078, "learning_rate": 1.519730161032932e-05, "loss": 0.0356, "step": 6721 }, { "epoch": 0.7287510841283608, "grad_norm": 2.481003761291504, "learning_rate": 1.5196576236761933e-05, "loss": 0.0541, "step": 6722 }, { "epoch": 0.7288594969644406, "grad_norm": 2.411632537841797, "learning_rate": 1.5195850863194545e-05, "loss": 0.0321, "step": 6723 }, { "epoch": 0.7289679098005204, "grad_norm": 2.0542142391204834, "learning_rate": 1.5195125489627158e-05, "loss": 0.0603, "step": 6724 }, { "epoch": 0.7290763226366002, "grad_norm": 3.6768126487731934, "learning_rate": 1.5194400116059771e-05, "loss": 0.0262, "step": 6725 }, { "epoch": 0.7291847354726799, "grad_norm": 1.4117320775985718, "learning_rate": 1.5193674742492385e-05, "loss": 0.0256, "step": 6726 }, { "epoch": 0.7292931483087598, "grad_norm": 2.000110626220703, "learning_rate": 1.5192949368924999e-05, "loss": 0.0286, "step": 6727 }, { "epoch": 0.7294015611448396, "grad_norm": 2.754660129547119, "learning_rate": 1.5192223995357611e-05, "loss": 0.0769, "step": 6728 }, { "epoch": 0.7295099739809193, "grad_norm": 0.7898893356323242, "learning_rate": 1.5191498621790223e-05, "loss": 0.0087, "step": 6729 }, { "epoch": 0.7296183868169991, "grad_norm": 1.4923157691955566, "learning_rate": 1.5190773248222837e-05, "loss": 0.028, "step": 6730 }, { "epoch": 0.729726799653079, "grad_norm": 0.811703085899353, "learning_rate": 1.519004787465545e-05, "loss": 0.0205, "step": 6731 }, { "epoch": 0.7298352124891587, "grad_norm": 3.9671311378479004, "learning_rate": 1.5189322501088062e-05, "loss": 0.0733, "step": 6732 }, { "epoch": 0.7299436253252385, "grad_norm": 1.0394729375839233, "learning_rate": 1.5188597127520674e-05, "loss": 0.03, "step": 6733 }, { "epoch": 0.7300520381613183, "grad_norm": 1.136306881904602, "learning_rate": 1.5187871753953288e-05, "loss": 0.0116, "step": 6734 }, { "epoch": 0.7301604509973981, "grad_norm": 2.2200591564178467, "learning_rate": 1.51871463803859e-05, "loss": 0.0355, "step": 6735 }, { "epoch": 0.7302688638334779, "grad_norm": 4.923301696777344, "learning_rate": 1.5186421006818512e-05, "loss": 0.0522, "step": 6736 }, { "epoch": 0.7303772766695577, "grad_norm": 4.5644612312316895, "learning_rate": 1.5185695633251126e-05, "loss": 0.0705, "step": 6737 }, { "epoch": 0.7304856895056374, "grad_norm": 2.339421272277832, "learning_rate": 1.5184970259683738e-05, "loss": 0.0304, "step": 6738 }, { "epoch": 0.7305941023417173, "grad_norm": 1.2288999557495117, "learning_rate": 1.518424488611635e-05, "loss": 0.0166, "step": 6739 }, { "epoch": 0.730702515177797, "grad_norm": 2.5053415298461914, "learning_rate": 1.5183519512548964e-05, "loss": 0.029, "step": 6740 }, { "epoch": 0.7308109280138768, "grad_norm": 3.2363951206207275, "learning_rate": 1.5182794138981576e-05, "loss": 0.0386, "step": 6741 }, { "epoch": 0.7309193408499567, "grad_norm": 2.0879013538360596, "learning_rate": 1.5182068765414188e-05, "loss": 0.0219, "step": 6742 }, { "epoch": 0.7310277536860365, "grad_norm": 3.709576368331909, "learning_rate": 1.5181343391846804e-05, "loss": 0.0455, "step": 6743 }, { "epoch": 0.7311361665221162, "grad_norm": 3.4322893619537354, "learning_rate": 1.5180618018279416e-05, "loss": 0.0357, "step": 6744 }, { "epoch": 0.731244579358196, "grad_norm": 0.9875735640525818, "learning_rate": 1.5179892644712028e-05, "loss": 0.0074, "step": 6745 }, { "epoch": 0.7313529921942759, "grad_norm": 1.909488558769226, "learning_rate": 1.5179167271144642e-05, "loss": 0.0625, "step": 6746 }, { "epoch": 0.7314614050303556, "grad_norm": 1.7796779870986938, "learning_rate": 1.5178441897577254e-05, "loss": 0.0162, "step": 6747 }, { "epoch": 0.7315698178664354, "grad_norm": 1.7191342115402222, "learning_rate": 1.5177716524009866e-05, "loss": 0.0218, "step": 6748 }, { "epoch": 0.7316782307025151, "grad_norm": 1.4997601509094238, "learning_rate": 1.517699115044248e-05, "loss": 0.0282, "step": 6749 }, { "epoch": 0.731786643538595, "grad_norm": 1.4483983516693115, "learning_rate": 1.5176265776875092e-05, "loss": 0.0363, "step": 6750 }, { "epoch": 0.7318950563746748, "grad_norm": 0.813562273979187, "learning_rate": 1.5175540403307704e-05, "loss": 0.012, "step": 6751 }, { "epoch": 0.7320034692107545, "grad_norm": 0.9380598068237305, "learning_rate": 1.5174815029740318e-05, "loss": 0.0498, "step": 6752 }, { "epoch": 0.7321118820468343, "grad_norm": 0.25576847791671753, "learning_rate": 1.517408965617293e-05, "loss": 0.0067, "step": 6753 }, { "epoch": 0.7322202948829142, "grad_norm": 2.6987504959106445, "learning_rate": 1.5173364282605542e-05, "loss": 0.0408, "step": 6754 }, { "epoch": 0.7323287077189939, "grad_norm": 4.199209690093994, "learning_rate": 1.5172638909038156e-05, "loss": 0.0296, "step": 6755 }, { "epoch": 0.7324371205550737, "grad_norm": 6.113672256469727, "learning_rate": 1.5171913535470768e-05, "loss": 0.0393, "step": 6756 }, { "epoch": 0.7325455333911535, "grad_norm": 5.230867385864258, "learning_rate": 1.517118816190338e-05, "loss": 0.033, "step": 6757 }, { "epoch": 0.7326539462272333, "grad_norm": 2.640284538269043, "learning_rate": 1.5170462788335994e-05, "loss": 0.0203, "step": 6758 }, { "epoch": 0.7327623590633131, "grad_norm": 1.2606749534606934, "learning_rate": 1.5169737414768606e-05, "loss": 0.0222, "step": 6759 }, { "epoch": 0.7328707718993929, "grad_norm": 1.62392258644104, "learning_rate": 1.516901204120122e-05, "loss": 0.022, "step": 6760 }, { "epoch": 0.7329791847354726, "grad_norm": 2.8217532634735107, "learning_rate": 1.5168286667633834e-05, "loss": 0.0501, "step": 6761 }, { "epoch": 0.7330875975715525, "grad_norm": 4.205286979675293, "learning_rate": 1.5167561294066446e-05, "loss": 0.0567, "step": 6762 }, { "epoch": 0.7331960104076323, "grad_norm": 3.7178194522857666, "learning_rate": 1.5166835920499058e-05, "loss": 0.0517, "step": 6763 }, { "epoch": 0.733304423243712, "grad_norm": 1.8514158725738525, "learning_rate": 1.5166110546931672e-05, "loss": 0.0162, "step": 6764 }, { "epoch": 0.7334128360797918, "grad_norm": 2.3674092292785645, "learning_rate": 1.5165385173364284e-05, "loss": 0.0323, "step": 6765 }, { "epoch": 0.7335212489158717, "grad_norm": 4.049951076507568, "learning_rate": 1.5164659799796896e-05, "loss": 0.0411, "step": 6766 }, { "epoch": 0.7336296617519514, "grad_norm": 1.9132367372512817, "learning_rate": 1.516393442622951e-05, "loss": 0.0348, "step": 6767 }, { "epoch": 0.7337380745880312, "grad_norm": 2.4023218154907227, "learning_rate": 1.5163209052662122e-05, "loss": 0.0347, "step": 6768 }, { "epoch": 0.733846487424111, "grad_norm": 0.5678877234458923, "learning_rate": 1.5162483679094735e-05, "loss": 0.0062, "step": 6769 }, { "epoch": 0.7339549002601908, "grad_norm": 1.41378915309906, "learning_rate": 1.5161758305527348e-05, "loss": 0.032, "step": 6770 }, { "epoch": 0.7340633130962706, "grad_norm": 0.34753280878067017, "learning_rate": 1.516103293195996e-05, "loss": 0.0008, "step": 6771 }, { "epoch": 0.7341717259323504, "grad_norm": 3.987016201019287, "learning_rate": 1.5160307558392573e-05, "loss": 0.0342, "step": 6772 }, { "epoch": 0.7342801387684302, "grad_norm": 4.4430832862854, "learning_rate": 1.5159582184825185e-05, "loss": 0.1022, "step": 6773 }, { "epoch": 0.73438855160451, "grad_norm": 4.306236267089844, "learning_rate": 1.5158856811257799e-05, "loss": 0.1096, "step": 6774 }, { "epoch": 0.7344969644405898, "grad_norm": 2.153353691101074, "learning_rate": 1.515813143769041e-05, "loss": 0.0472, "step": 6775 }, { "epoch": 0.7346053772766695, "grad_norm": 3.9264228343963623, "learning_rate": 1.5157406064123023e-05, "loss": 0.042, "step": 6776 }, { "epoch": 0.7347137901127494, "grad_norm": 3.8805336952209473, "learning_rate": 1.5156680690555637e-05, "loss": 0.0288, "step": 6777 }, { "epoch": 0.7348222029488292, "grad_norm": 2.445352077484131, "learning_rate": 1.515595531698825e-05, "loss": 0.0254, "step": 6778 }, { "epoch": 0.7349306157849089, "grad_norm": 6.342884540557861, "learning_rate": 1.5155229943420864e-05, "loss": 0.0667, "step": 6779 }, { "epoch": 0.7350390286209887, "grad_norm": 2.1035866737365723, "learning_rate": 1.5154504569853477e-05, "loss": 0.0561, "step": 6780 }, { "epoch": 0.7351474414570686, "grad_norm": 2.4568827152252197, "learning_rate": 1.5153779196286089e-05, "loss": 0.02, "step": 6781 }, { "epoch": 0.7352558542931483, "grad_norm": 5.171073913574219, "learning_rate": 1.5153053822718703e-05, "loss": 0.0216, "step": 6782 }, { "epoch": 0.7353642671292281, "grad_norm": 0.5418777465820312, "learning_rate": 1.5152328449151315e-05, "loss": 0.0196, "step": 6783 }, { "epoch": 0.7354726799653079, "grad_norm": 1.5378644466400146, "learning_rate": 1.5151603075583927e-05, "loss": 0.0286, "step": 6784 }, { "epoch": 0.7355810928013877, "grad_norm": 6.032355785369873, "learning_rate": 1.5150877702016539e-05, "loss": 0.0456, "step": 6785 }, { "epoch": 0.7356895056374675, "grad_norm": 1.9474146366119385, "learning_rate": 1.5150152328449153e-05, "loss": 0.0297, "step": 6786 }, { "epoch": 0.7357979184735473, "grad_norm": 0.965621829032898, "learning_rate": 1.5149426954881765e-05, "loss": 0.0137, "step": 6787 }, { "epoch": 0.735906331309627, "grad_norm": 4.930842876434326, "learning_rate": 1.5148701581314377e-05, "loss": 0.0466, "step": 6788 }, { "epoch": 0.7360147441457069, "grad_norm": 2.7943923473358154, "learning_rate": 1.5147976207746991e-05, "loss": 0.048, "step": 6789 }, { "epoch": 0.7361231569817867, "grad_norm": 4.142658233642578, "learning_rate": 1.5147250834179603e-05, "loss": 0.1049, "step": 6790 }, { "epoch": 0.7362315698178664, "grad_norm": 1.8563405275344849, "learning_rate": 1.5146525460612215e-05, "loss": 0.0175, "step": 6791 }, { "epoch": 0.7363399826539462, "grad_norm": 2.5081281661987305, "learning_rate": 1.5145800087044829e-05, "loss": 0.058, "step": 6792 }, { "epoch": 0.736448395490026, "grad_norm": 2.367609739303589, "learning_rate": 1.5145074713477441e-05, "loss": 0.0157, "step": 6793 }, { "epoch": 0.7365568083261058, "grad_norm": 1.7852911949157715, "learning_rate": 1.5144349339910053e-05, "loss": 0.0071, "step": 6794 }, { "epoch": 0.7366652211621856, "grad_norm": 1.0007344484329224, "learning_rate": 1.5143623966342669e-05, "loss": 0.0183, "step": 6795 }, { "epoch": 0.7367736339982653, "grad_norm": 1.6616638898849487, "learning_rate": 1.5142898592775281e-05, "loss": 0.0441, "step": 6796 }, { "epoch": 0.7368820468343452, "grad_norm": 1.9908615350723267, "learning_rate": 1.5142173219207895e-05, "loss": 0.0389, "step": 6797 }, { "epoch": 0.736990459670425, "grad_norm": 3.260906457901001, "learning_rate": 1.5141447845640507e-05, "loss": 0.0163, "step": 6798 }, { "epoch": 0.7370988725065047, "grad_norm": 2.4800572395324707, "learning_rate": 1.514072247207312e-05, "loss": 0.0446, "step": 6799 }, { "epoch": 0.7372072853425846, "grad_norm": 0.41057515144348145, "learning_rate": 1.5139997098505731e-05, "loss": 0.019, "step": 6800 }, { "epoch": 0.7373156981786644, "grad_norm": 1.4390565156936646, "learning_rate": 1.5139271724938345e-05, "loss": 0.0221, "step": 6801 }, { "epoch": 0.7374241110147441, "grad_norm": 1.8194222450256348, "learning_rate": 1.5138546351370957e-05, "loss": 0.0424, "step": 6802 }, { "epoch": 0.7375325238508239, "grad_norm": 2.7169017791748047, "learning_rate": 1.513782097780357e-05, "loss": 0.0487, "step": 6803 }, { "epoch": 0.7376409366869038, "grad_norm": 1.8343732357025146, "learning_rate": 1.5137095604236183e-05, "loss": 0.0245, "step": 6804 }, { "epoch": 0.7377493495229835, "grad_norm": 2.9080605506896973, "learning_rate": 1.5136370230668795e-05, "loss": 0.1059, "step": 6805 }, { "epoch": 0.7378577623590633, "grad_norm": 5.437178134918213, "learning_rate": 1.5135644857101408e-05, "loss": 0.0803, "step": 6806 }, { "epoch": 0.7379661751951431, "grad_norm": 3.2527711391448975, "learning_rate": 1.5134919483534021e-05, "loss": 0.0408, "step": 6807 }, { "epoch": 0.738074588031223, "grad_norm": 4.09775447845459, "learning_rate": 1.5134194109966634e-05, "loss": 0.0241, "step": 6808 }, { "epoch": 0.7381830008673027, "grad_norm": 1.8983101844787598, "learning_rate": 1.5133468736399246e-05, "loss": 0.0464, "step": 6809 }, { "epoch": 0.7382914137033825, "grad_norm": 3.964928150177002, "learning_rate": 1.513274336283186e-05, "loss": 0.0478, "step": 6810 }, { "epoch": 0.7383998265394622, "grad_norm": 3.4102470874786377, "learning_rate": 1.5132017989264472e-05, "loss": 0.0223, "step": 6811 }, { "epoch": 0.7385082393755421, "grad_norm": 1.1307927370071411, "learning_rate": 1.5131292615697086e-05, "loss": 0.0197, "step": 6812 }, { "epoch": 0.7386166522116219, "grad_norm": 3.1307148933410645, "learning_rate": 1.51305672421297e-05, "loss": 0.0734, "step": 6813 }, { "epoch": 0.7387250650477016, "grad_norm": 3.4683780670166016, "learning_rate": 1.5129841868562312e-05, "loss": 0.0599, "step": 6814 }, { "epoch": 0.7388334778837814, "grad_norm": 2.3690879344940186, "learning_rate": 1.5129116494994924e-05, "loss": 0.0359, "step": 6815 }, { "epoch": 0.7389418907198613, "grad_norm": 0.6636857986450195, "learning_rate": 1.5128391121427537e-05, "loss": 0.0065, "step": 6816 }, { "epoch": 0.739050303555941, "grad_norm": 1.5777679681777954, "learning_rate": 1.512766574786015e-05, "loss": 0.0385, "step": 6817 }, { "epoch": 0.7391587163920208, "grad_norm": 1.0537865161895752, "learning_rate": 1.5126940374292762e-05, "loss": 0.0211, "step": 6818 }, { "epoch": 0.7392671292281006, "grad_norm": 1.2097980976104736, "learning_rate": 1.5126215000725376e-05, "loss": 0.0225, "step": 6819 }, { "epoch": 0.7393755420641804, "grad_norm": 1.494883418083191, "learning_rate": 1.5125489627157988e-05, "loss": 0.0493, "step": 6820 }, { "epoch": 0.7394839549002602, "grad_norm": 1.7305501699447632, "learning_rate": 1.51247642535906e-05, "loss": 0.0474, "step": 6821 }, { "epoch": 0.73959236773634, "grad_norm": 3.0998666286468506, "learning_rate": 1.5124038880023214e-05, "loss": 0.0853, "step": 6822 }, { "epoch": 0.7397007805724197, "grad_norm": 4.857917308807373, "learning_rate": 1.5123313506455826e-05, "loss": 0.0278, "step": 6823 }, { "epoch": 0.7398091934084996, "grad_norm": 6.560347080230713, "learning_rate": 1.5122588132888438e-05, "loss": 0.0845, "step": 6824 }, { "epoch": 0.7399176062445794, "grad_norm": 3.0117111206054688, "learning_rate": 1.512186275932105e-05, "loss": 0.0728, "step": 6825 }, { "epoch": 0.7400260190806591, "grad_norm": 1.7139439582824707, "learning_rate": 1.5121137385753664e-05, "loss": 0.0339, "step": 6826 }, { "epoch": 0.7401344319167389, "grad_norm": 1.2744060754776, "learning_rate": 1.5120412012186276e-05, "loss": 0.0401, "step": 6827 }, { "epoch": 0.7402428447528188, "grad_norm": 4.367711067199707, "learning_rate": 1.5119686638618888e-05, "loss": 0.0488, "step": 6828 }, { "epoch": 0.7403512575888985, "grad_norm": 3.8653571605682373, "learning_rate": 1.5118961265051504e-05, "loss": 0.0349, "step": 6829 }, { "epoch": 0.7404596704249783, "grad_norm": 3.1161460876464844, "learning_rate": 1.5118235891484116e-05, "loss": 0.0611, "step": 6830 }, { "epoch": 0.7405680832610582, "grad_norm": 3.0622942447662354, "learning_rate": 1.511751051791673e-05, "loss": 0.0473, "step": 6831 }, { "epoch": 0.7406764960971379, "grad_norm": 1.2693489789962769, "learning_rate": 1.5116785144349342e-05, "loss": 0.0183, "step": 6832 }, { "epoch": 0.7407849089332177, "grad_norm": 4.070211410522461, "learning_rate": 1.5116059770781954e-05, "loss": 0.0506, "step": 6833 }, { "epoch": 0.7408933217692975, "grad_norm": 1.4264475107192993, "learning_rate": 1.5115334397214568e-05, "loss": 0.0327, "step": 6834 }, { "epoch": 0.7410017346053773, "grad_norm": 1.6209489107131958, "learning_rate": 1.511460902364718e-05, "loss": 0.0186, "step": 6835 }, { "epoch": 0.7411101474414571, "grad_norm": 1.0544875860214233, "learning_rate": 1.5113883650079792e-05, "loss": 0.0148, "step": 6836 }, { "epoch": 0.7412185602775369, "grad_norm": 1.1838696002960205, "learning_rate": 1.5113158276512404e-05, "loss": 0.0292, "step": 6837 }, { "epoch": 0.7413269731136166, "grad_norm": 0.8610125184059143, "learning_rate": 1.5112432902945018e-05, "loss": 0.0268, "step": 6838 }, { "epoch": 0.7414353859496965, "grad_norm": 3.1953470706939697, "learning_rate": 1.511170752937763e-05, "loss": 0.0387, "step": 6839 }, { "epoch": 0.7415437987857763, "grad_norm": 2.914215326309204, "learning_rate": 1.5110982155810242e-05, "loss": 0.0189, "step": 6840 }, { "epoch": 0.741652211621856, "grad_norm": 3.4745280742645264, "learning_rate": 1.5110256782242856e-05, "loss": 0.0225, "step": 6841 }, { "epoch": 0.7417606244579358, "grad_norm": 1.330583095550537, "learning_rate": 1.5109531408675468e-05, "loss": 0.0228, "step": 6842 }, { "epoch": 0.7418690372940157, "grad_norm": 3.3831565380096436, "learning_rate": 1.510880603510808e-05, "loss": 0.0475, "step": 6843 }, { "epoch": 0.7419774501300954, "grad_norm": 2.6843535900115967, "learning_rate": 1.5108080661540694e-05, "loss": 0.0329, "step": 6844 }, { "epoch": 0.7420858629661752, "grad_norm": 1.4933037757873535, "learning_rate": 1.5107355287973307e-05, "loss": 0.011, "step": 6845 }, { "epoch": 0.7421942758022549, "grad_norm": 1.4486273527145386, "learning_rate": 1.5106629914405919e-05, "loss": 0.0329, "step": 6846 }, { "epoch": 0.7423026886383348, "grad_norm": 4.199230194091797, "learning_rate": 1.5105904540838534e-05, "loss": 0.0717, "step": 6847 }, { "epoch": 0.7424111014744146, "grad_norm": 1.3114700317382812, "learning_rate": 1.5105179167271146e-05, "loss": 0.04, "step": 6848 }, { "epoch": 0.7425195143104943, "grad_norm": 2.2384555339813232, "learning_rate": 1.510445379370376e-05, "loss": 0.0365, "step": 6849 }, { "epoch": 0.7426279271465741, "grad_norm": 3.8833372592926025, "learning_rate": 1.5103728420136372e-05, "loss": 0.0463, "step": 6850 }, { "epoch": 0.742736339982654, "grad_norm": 2.626457929611206, "learning_rate": 1.5103003046568985e-05, "loss": 0.029, "step": 6851 }, { "epoch": 0.7428447528187337, "grad_norm": 1.1798560619354248, "learning_rate": 1.5102277673001597e-05, "loss": 0.0175, "step": 6852 }, { "epoch": 0.7429531656548135, "grad_norm": 3.312044143676758, "learning_rate": 1.510155229943421e-05, "loss": 0.0437, "step": 6853 }, { "epoch": 0.7430615784908933, "grad_norm": 3.4717636108398438, "learning_rate": 1.5100826925866823e-05, "loss": 0.0526, "step": 6854 }, { "epoch": 0.7431699913269731, "grad_norm": 5.005003452301025, "learning_rate": 1.5100101552299435e-05, "loss": 0.0541, "step": 6855 }, { "epoch": 0.7432784041630529, "grad_norm": 2.5315680503845215, "learning_rate": 1.5099376178732049e-05, "loss": 0.0612, "step": 6856 }, { "epoch": 0.7433868169991327, "grad_norm": 2.016953706741333, "learning_rate": 1.509865080516466e-05, "loss": 0.0462, "step": 6857 }, { "epoch": 0.7434952298352124, "grad_norm": 0.9303341507911682, "learning_rate": 1.5097925431597273e-05, "loss": 0.0177, "step": 6858 }, { "epoch": 0.7436036426712923, "grad_norm": 2.246246337890625, "learning_rate": 1.5097200058029887e-05, "loss": 0.0538, "step": 6859 }, { "epoch": 0.7437120555073721, "grad_norm": 0.7103111743927002, "learning_rate": 1.5096474684462499e-05, "loss": 0.0199, "step": 6860 }, { "epoch": 0.7438204683434518, "grad_norm": 2.580106019973755, "learning_rate": 1.5095749310895111e-05, "loss": 0.0292, "step": 6861 }, { "epoch": 0.7439288811795317, "grad_norm": 1.0738270282745361, "learning_rate": 1.5095023937327725e-05, "loss": 0.0372, "step": 6862 }, { "epoch": 0.7440372940156115, "grad_norm": 3.729860305786133, "learning_rate": 1.5094298563760337e-05, "loss": 0.0389, "step": 6863 }, { "epoch": 0.7441457068516912, "grad_norm": 5.309930801391602, "learning_rate": 1.5093573190192951e-05, "loss": 0.0443, "step": 6864 }, { "epoch": 0.744254119687771, "grad_norm": 1.399530291557312, "learning_rate": 1.5092847816625565e-05, "loss": 0.0214, "step": 6865 }, { "epoch": 0.7443625325238509, "grad_norm": 2.7506117820739746, "learning_rate": 1.5092122443058177e-05, "loss": 0.0709, "step": 6866 }, { "epoch": 0.7444709453599306, "grad_norm": 1.121443748474121, "learning_rate": 1.5091397069490789e-05, "loss": 0.0675, "step": 6867 }, { "epoch": 0.7445793581960104, "grad_norm": 3.8042922019958496, "learning_rate": 1.5090671695923403e-05, "loss": 0.0687, "step": 6868 }, { "epoch": 0.7446877710320902, "grad_norm": 1.1608835458755493, "learning_rate": 1.5089946322356015e-05, "loss": 0.0205, "step": 6869 }, { "epoch": 0.74479618386817, "grad_norm": 2.9199931621551514, "learning_rate": 1.5089220948788627e-05, "loss": 0.0559, "step": 6870 }, { "epoch": 0.7449045967042498, "grad_norm": 9.66712474822998, "learning_rate": 1.5088495575221241e-05, "loss": 0.0479, "step": 6871 }, { "epoch": 0.7450130095403296, "grad_norm": 1.550669550895691, "learning_rate": 1.5087770201653853e-05, "loss": 0.0206, "step": 6872 }, { "epoch": 0.7451214223764093, "grad_norm": 1.8958179950714111, "learning_rate": 1.5087044828086465e-05, "loss": 0.0434, "step": 6873 }, { "epoch": 0.7452298352124892, "grad_norm": 1.8445355892181396, "learning_rate": 1.5086319454519079e-05, "loss": 0.0361, "step": 6874 }, { "epoch": 0.745338248048569, "grad_norm": 1.9020938873291016, "learning_rate": 1.5085594080951691e-05, "loss": 0.0239, "step": 6875 }, { "epoch": 0.7454466608846487, "grad_norm": 2.776273727416992, "learning_rate": 1.5084868707384303e-05, "loss": 0.0402, "step": 6876 }, { "epoch": 0.7455550737207285, "grad_norm": 2.173734188079834, "learning_rate": 1.5084143333816916e-05, "loss": 0.051, "step": 6877 }, { "epoch": 0.7456634865568084, "grad_norm": 1.9076687097549438, "learning_rate": 1.508341796024953e-05, "loss": 0.0347, "step": 6878 }, { "epoch": 0.7457718993928881, "grad_norm": 1.8490055799484253, "learning_rate": 1.5082692586682141e-05, "loss": 0.0321, "step": 6879 }, { "epoch": 0.7458803122289679, "grad_norm": 0.9533063769340515, "learning_rate": 1.5081967213114754e-05, "loss": 0.0211, "step": 6880 }, { "epoch": 0.7459887250650477, "grad_norm": 4.44788122177124, "learning_rate": 1.508124183954737e-05, "loss": 0.0481, "step": 6881 }, { "epoch": 0.7460971379011275, "grad_norm": 1.19774329662323, "learning_rate": 1.5080516465979981e-05, "loss": 0.021, "step": 6882 }, { "epoch": 0.7462055507372073, "grad_norm": 2.368448495864868, "learning_rate": 1.5079791092412595e-05, "loss": 0.0543, "step": 6883 }, { "epoch": 0.746313963573287, "grad_norm": 2.7651376724243164, "learning_rate": 1.5079065718845207e-05, "loss": 0.0259, "step": 6884 }, { "epoch": 0.7464223764093668, "grad_norm": 0.9799730777740479, "learning_rate": 1.507834034527782e-05, "loss": 0.0102, "step": 6885 }, { "epoch": 0.7465307892454467, "grad_norm": 0.37003442645072937, "learning_rate": 1.5077614971710433e-05, "loss": 0.0175, "step": 6886 }, { "epoch": 0.7466392020815265, "grad_norm": 4.027294635772705, "learning_rate": 1.5076889598143045e-05, "loss": 0.0224, "step": 6887 }, { "epoch": 0.7467476149176062, "grad_norm": 1.6694607734680176, "learning_rate": 1.5076164224575658e-05, "loss": 0.0247, "step": 6888 }, { "epoch": 0.7468560277536861, "grad_norm": 1.864428162574768, "learning_rate": 1.507543885100827e-05, "loss": 0.0146, "step": 6889 }, { "epoch": 0.7469644405897659, "grad_norm": 3.0501928329467773, "learning_rate": 1.5074713477440884e-05, "loss": 0.0409, "step": 6890 }, { "epoch": 0.7470728534258456, "grad_norm": 3.3586819171905518, "learning_rate": 1.5073988103873496e-05, "loss": 0.0491, "step": 6891 }, { "epoch": 0.7471812662619254, "grad_norm": 5.604455471038818, "learning_rate": 1.5073262730306108e-05, "loss": 0.1169, "step": 6892 }, { "epoch": 0.7472896790980053, "grad_norm": 3.737891912460327, "learning_rate": 1.5072537356738722e-05, "loss": 0.0403, "step": 6893 }, { "epoch": 0.747398091934085, "grad_norm": 2.2815663814544678, "learning_rate": 1.5071811983171334e-05, "loss": 0.0265, "step": 6894 }, { "epoch": 0.7475065047701648, "grad_norm": 1.6029634475708008, "learning_rate": 1.5071086609603946e-05, "loss": 0.0152, "step": 6895 }, { "epoch": 0.7476149176062445, "grad_norm": 3.4724061489105225, "learning_rate": 1.507036123603656e-05, "loss": 0.0478, "step": 6896 }, { "epoch": 0.7477233304423244, "grad_norm": 2.9591729640960693, "learning_rate": 1.5069635862469172e-05, "loss": 0.0505, "step": 6897 }, { "epoch": 0.7478317432784042, "grad_norm": 2.2557971477508545, "learning_rate": 1.5068910488901787e-05, "loss": 0.0558, "step": 6898 }, { "epoch": 0.747940156114484, "grad_norm": 10.204141616821289, "learning_rate": 1.50681851153344e-05, "loss": 0.0697, "step": 6899 }, { "epoch": 0.7480485689505637, "grad_norm": 0.511874258518219, "learning_rate": 1.5067459741767012e-05, "loss": 0.0146, "step": 6900 }, { "epoch": 0.7481569817866436, "grad_norm": 1.9244418144226074, "learning_rate": 1.5066734368199626e-05, "loss": 0.0244, "step": 6901 }, { "epoch": 0.7482653946227233, "grad_norm": 1.903429627418518, "learning_rate": 1.5066008994632238e-05, "loss": 0.0453, "step": 6902 }, { "epoch": 0.7483738074588031, "grad_norm": 1.8106776475906372, "learning_rate": 1.506528362106485e-05, "loss": 0.0155, "step": 6903 }, { "epoch": 0.7484822202948829, "grad_norm": 1.5841082334518433, "learning_rate": 1.5064558247497462e-05, "loss": 0.0476, "step": 6904 }, { "epoch": 0.7485906331309627, "grad_norm": 4.110623359680176, "learning_rate": 1.5063832873930076e-05, "loss": 0.0783, "step": 6905 }, { "epoch": 0.7486990459670425, "grad_norm": 3.466686487197876, "learning_rate": 1.5063107500362688e-05, "loss": 0.0624, "step": 6906 }, { "epoch": 0.7488074588031223, "grad_norm": 2.0518383979797363, "learning_rate": 1.50623821267953e-05, "loss": 0.0158, "step": 6907 }, { "epoch": 0.748915871639202, "grad_norm": 2.456156015396118, "learning_rate": 1.5061656753227914e-05, "loss": 0.0337, "step": 6908 }, { "epoch": 0.7490242844752819, "grad_norm": 4.206280708312988, "learning_rate": 1.5060931379660526e-05, "loss": 0.0861, "step": 6909 }, { "epoch": 0.7491326973113617, "grad_norm": 2.907832622528076, "learning_rate": 1.5060206006093138e-05, "loss": 0.0609, "step": 6910 }, { "epoch": 0.7492411101474414, "grad_norm": 2.3392748832702637, "learning_rate": 1.5059480632525752e-05, "loss": 0.043, "step": 6911 }, { "epoch": 0.7493495229835212, "grad_norm": 2.856138229370117, "learning_rate": 1.5058755258958364e-05, "loss": 0.0507, "step": 6912 }, { "epoch": 0.7494579358196011, "grad_norm": 2.3569869995117188, "learning_rate": 1.5058029885390976e-05, "loss": 0.0197, "step": 6913 }, { "epoch": 0.7495663486556808, "grad_norm": 1.4791970252990723, "learning_rate": 1.505730451182359e-05, "loss": 0.0273, "step": 6914 }, { "epoch": 0.7496747614917606, "grad_norm": 1.6511445045471191, "learning_rate": 1.5056579138256202e-05, "loss": 0.0355, "step": 6915 }, { "epoch": 0.7497831743278404, "grad_norm": 2.200425624847412, "learning_rate": 1.5055853764688816e-05, "loss": 0.0635, "step": 6916 }, { "epoch": 0.7498915871639202, "grad_norm": 4.71242618560791, "learning_rate": 1.505512839112143e-05, "loss": 0.0599, "step": 6917 }, { "epoch": 0.75, "grad_norm": 2.379730463027954, "learning_rate": 1.5054403017554042e-05, "loss": 0.0341, "step": 6918 }, { "epoch": 0.7501084128360798, "grad_norm": 1.8262932300567627, "learning_rate": 1.5053677643986654e-05, "loss": 0.0296, "step": 6919 }, { "epoch": 0.7502168256721596, "grad_norm": 1.1276897192001343, "learning_rate": 1.5052952270419268e-05, "loss": 0.0188, "step": 6920 }, { "epoch": 0.7503252385082394, "grad_norm": 0.995851993560791, "learning_rate": 1.505222689685188e-05, "loss": 0.0129, "step": 6921 }, { "epoch": 0.7504336513443192, "grad_norm": 1.8854690790176392, "learning_rate": 1.5051501523284492e-05, "loss": 0.0485, "step": 6922 }, { "epoch": 0.7505420641803989, "grad_norm": 5.958863735198975, "learning_rate": 1.5050776149717106e-05, "loss": 0.0739, "step": 6923 }, { "epoch": 0.7506504770164788, "grad_norm": 2.5531728267669678, "learning_rate": 1.5050050776149718e-05, "loss": 0.0401, "step": 6924 }, { "epoch": 0.7507588898525586, "grad_norm": 5.035049915313721, "learning_rate": 1.504932540258233e-05, "loss": 0.0459, "step": 6925 }, { "epoch": 0.7508673026886383, "grad_norm": 2.584609270095825, "learning_rate": 1.5048600029014944e-05, "loss": 0.0378, "step": 6926 }, { "epoch": 0.7509757155247181, "grad_norm": 1.2373337745666504, "learning_rate": 1.5047874655447557e-05, "loss": 0.032, "step": 6927 }, { "epoch": 0.751084128360798, "grad_norm": 2.0847244262695312, "learning_rate": 1.5047149281880169e-05, "loss": 0.0563, "step": 6928 }, { "epoch": 0.7511925411968777, "grad_norm": 1.0323563814163208, "learning_rate": 1.5046423908312781e-05, "loss": 0.0394, "step": 6929 }, { "epoch": 0.7513009540329575, "grad_norm": 0.16138717532157898, "learning_rate": 1.5045698534745395e-05, "loss": 0.0025, "step": 6930 }, { "epoch": 0.7514093668690373, "grad_norm": 1.3753653764724731, "learning_rate": 1.5044973161178007e-05, "loss": 0.0183, "step": 6931 }, { "epoch": 0.7515177797051171, "grad_norm": 2.0621280670166016, "learning_rate": 1.5044247787610619e-05, "loss": 0.018, "step": 6932 }, { "epoch": 0.7516261925411969, "grad_norm": 1.2717573642730713, "learning_rate": 1.5043522414043235e-05, "loss": 0.0677, "step": 6933 }, { "epoch": 0.7517346053772767, "grad_norm": 1.5586564540863037, "learning_rate": 1.5042797040475847e-05, "loss": 0.0153, "step": 6934 }, { "epoch": 0.7518430182133564, "grad_norm": 0.6102261543273926, "learning_rate": 1.504207166690846e-05, "loss": 0.0244, "step": 6935 }, { "epoch": 0.7519514310494363, "grad_norm": 1.3226947784423828, "learning_rate": 1.5041346293341073e-05, "loss": 0.0269, "step": 6936 }, { "epoch": 0.752059843885516, "grad_norm": 1.5322598218917847, "learning_rate": 1.5040620919773685e-05, "loss": 0.0317, "step": 6937 }, { "epoch": 0.7521682567215958, "grad_norm": 1.4414606094360352, "learning_rate": 1.5039895546206299e-05, "loss": 0.0626, "step": 6938 }, { "epoch": 0.7522766695576756, "grad_norm": 4.234394550323486, "learning_rate": 1.503917017263891e-05, "loss": 0.0443, "step": 6939 }, { "epoch": 0.7523850823937555, "grad_norm": 1.0007401704788208, "learning_rate": 1.5038444799071523e-05, "loss": 0.0234, "step": 6940 }, { "epoch": 0.7524934952298352, "grad_norm": 2.107722520828247, "learning_rate": 1.5037719425504135e-05, "loss": 0.013, "step": 6941 }, { "epoch": 0.752601908065915, "grad_norm": 4.488519668579102, "learning_rate": 1.5036994051936749e-05, "loss": 0.085, "step": 6942 }, { "epoch": 0.7527103209019947, "grad_norm": 3.5917181968688965, "learning_rate": 1.5036268678369361e-05, "loss": 0.0488, "step": 6943 }, { "epoch": 0.7528187337380746, "grad_norm": 2.4693727493286133, "learning_rate": 1.5035543304801973e-05, "loss": 0.0614, "step": 6944 }, { "epoch": 0.7529271465741544, "grad_norm": 1.9418115615844727, "learning_rate": 1.5034817931234587e-05, "loss": 0.0179, "step": 6945 }, { "epoch": 0.7530355594102341, "grad_norm": 3.084685802459717, "learning_rate": 1.5034092557667199e-05, "loss": 0.016, "step": 6946 }, { "epoch": 0.7531439722463139, "grad_norm": 5.314968585968018, "learning_rate": 1.5033367184099811e-05, "loss": 0.0755, "step": 6947 }, { "epoch": 0.7532523850823938, "grad_norm": 6.10956335067749, "learning_rate": 1.5032641810532425e-05, "loss": 0.0664, "step": 6948 }, { "epoch": 0.7533607979184735, "grad_norm": 3.849925994873047, "learning_rate": 1.5031916436965037e-05, "loss": 0.0419, "step": 6949 }, { "epoch": 0.7534692107545533, "grad_norm": 2.327413320541382, "learning_rate": 1.5031191063397653e-05, "loss": 0.0971, "step": 6950 }, { "epoch": 0.7535776235906332, "grad_norm": 3.9025142192840576, "learning_rate": 1.5030465689830265e-05, "loss": 0.0809, "step": 6951 }, { "epoch": 0.753686036426713, "grad_norm": 3.316828727722168, "learning_rate": 1.5029740316262877e-05, "loss": 0.0446, "step": 6952 }, { "epoch": 0.7537944492627927, "grad_norm": 3.140136241912842, "learning_rate": 1.5029014942695491e-05, "loss": 0.0308, "step": 6953 }, { "epoch": 0.7539028620988725, "grad_norm": 5.209475994110107, "learning_rate": 1.5028289569128103e-05, "loss": 0.0843, "step": 6954 }, { "epoch": 0.7540112749349523, "grad_norm": 1.445209264755249, "learning_rate": 1.5027564195560715e-05, "loss": 0.0241, "step": 6955 }, { "epoch": 0.7541196877710321, "grad_norm": 1.0291026830673218, "learning_rate": 1.5026838821993327e-05, "loss": 0.0259, "step": 6956 }, { "epoch": 0.7542281006071119, "grad_norm": 3.9011647701263428, "learning_rate": 1.5026113448425941e-05, "loss": 0.0381, "step": 6957 }, { "epoch": 0.7543365134431916, "grad_norm": 2.232104539871216, "learning_rate": 1.5025388074858553e-05, "loss": 0.0267, "step": 6958 }, { "epoch": 0.7544449262792715, "grad_norm": 3.800670623779297, "learning_rate": 1.5024662701291165e-05, "loss": 0.0218, "step": 6959 }, { "epoch": 0.7545533391153513, "grad_norm": 0.9305667281150818, "learning_rate": 1.502393732772378e-05, "loss": 0.0288, "step": 6960 }, { "epoch": 0.754661751951431, "grad_norm": 8.165789604187012, "learning_rate": 1.5023211954156391e-05, "loss": 0.053, "step": 6961 }, { "epoch": 0.7547701647875108, "grad_norm": 4.356983184814453, "learning_rate": 1.5022486580589004e-05, "loss": 0.0334, "step": 6962 }, { "epoch": 0.7548785776235907, "grad_norm": 31.193403244018555, "learning_rate": 1.5021761207021617e-05, "loss": 0.0879, "step": 6963 }, { "epoch": 0.7549869904596704, "grad_norm": 3.519162178039551, "learning_rate": 1.502103583345423e-05, "loss": 0.0828, "step": 6964 }, { "epoch": 0.7550954032957502, "grad_norm": 5.076742172241211, "learning_rate": 1.5020310459886842e-05, "loss": 0.0414, "step": 6965 }, { "epoch": 0.75520381613183, "grad_norm": 1.9586607217788696, "learning_rate": 1.5019585086319456e-05, "loss": 0.048, "step": 6966 }, { "epoch": 0.7553122289679098, "grad_norm": 1.4018607139587402, "learning_rate": 1.5018859712752068e-05, "loss": 0.0285, "step": 6967 }, { "epoch": 0.7554206418039896, "grad_norm": 3.3838367462158203, "learning_rate": 1.5018134339184682e-05, "loss": 0.0339, "step": 6968 }, { "epoch": 0.7555290546400694, "grad_norm": 3.1114935874938965, "learning_rate": 1.5017408965617295e-05, "loss": 0.0668, "step": 6969 }, { "epoch": 0.7556374674761491, "grad_norm": 0.6839630007743835, "learning_rate": 1.5016683592049908e-05, "loss": 0.0205, "step": 6970 }, { "epoch": 0.755745880312229, "grad_norm": 4.8583550453186035, "learning_rate": 1.501595821848252e-05, "loss": 0.0271, "step": 6971 }, { "epoch": 0.7558542931483088, "grad_norm": 2.0074286460876465, "learning_rate": 1.5015232844915134e-05, "loss": 0.0337, "step": 6972 }, { "epoch": 0.7559627059843885, "grad_norm": 2.745988607406616, "learning_rate": 1.5014507471347746e-05, "loss": 0.0312, "step": 6973 }, { "epoch": 0.7560711188204683, "grad_norm": 3.849738121032715, "learning_rate": 1.5013782097780358e-05, "loss": 0.0575, "step": 6974 }, { "epoch": 0.7561795316565482, "grad_norm": 1.3582730293273926, "learning_rate": 1.5013056724212972e-05, "loss": 0.0239, "step": 6975 }, { "epoch": 0.7562879444926279, "grad_norm": 1.995600938796997, "learning_rate": 1.5012331350645584e-05, "loss": 0.0373, "step": 6976 }, { "epoch": 0.7563963573287077, "grad_norm": 4.729765892028809, "learning_rate": 1.5011605977078196e-05, "loss": 0.0296, "step": 6977 }, { "epoch": 0.7565047701647876, "grad_norm": 1.1392521858215332, "learning_rate": 1.501088060351081e-05, "loss": 0.0135, "step": 6978 }, { "epoch": 0.7566131830008673, "grad_norm": 2.064143657684326, "learning_rate": 1.5010155229943422e-05, "loss": 0.0602, "step": 6979 }, { "epoch": 0.7567215958369471, "grad_norm": 0.8120319247245789, "learning_rate": 1.5009429856376034e-05, "loss": 0.006, "step": 6980 }, { "epoch": 0.7568300086730269, "grad_norm": 2.3205173015594482, "learning_rate": 1.5008704482808646e-05, "loss": 0.0385, "step": 6981 }, { "epoch": 0.7569384215091067, "grad_norm": 2.3402323722839355, "learning_rate": 1.500797910924126e-05, "loss": 0.0601, "step": 6982 }, { "epoch": 0.7570468343451865, "grad_norm": 0.6853066682815552, "learning_rate": 1.5007253735673872e-05, "loss": 0.0181, "step": 6983 }, { "epoch": 0.7571552471812663, "grad_norm": 1.878201961517334, "learning_rate": 1.5006528362106484e-05, "loss": 0.039, "step": 6984 }, { "epoch": 0.757263660017346, "grad_norm": 2.228705883026123, "learning_rate": 1.50058029885391e-05, "loss": 0.0205, "step": 6985 }, { "epoch": 0.7573720728534259, "grad_norm": 2.259089231491089, "learning_rate": 1.5005077614971712e-05, "loss": 0.0524, "step": 6986 }, { "epoch": 0.7574804856895057, "grad_norm": 3.2789182662963867, "learning_rate": 1.5004352241404326e-05, "loss": 0.0252, "step": 6987 }, { "epoch": 0.7575888985255854, "grad_norm": 5.886814594268799, "learning_rate": 1.5003626867836938e-05, "loss": 0.0154, "step": 6988 }, { "epoch": 0.7576973113616652, "grad_norm": 2.3258755207061768, "learning_rate": 1.500290149426955e-05, "loss": 0.0536, "step": 6989 }, { "epoch": 0.7578057241977451, "grad_norm": 2.0488719940185547, "learning_rate": 1.5002176120702164e-05, "loss": 0.026, "step": 6990 }, { "epoch": 0.7579141370338248, "grad_norm": 5.20925235748291, "learning_rate": 1.5001450747134776e-05, "loss": 0.078, "step": 6991 }, { "epoch": 0.7580225498699046, "grad_norm": 1.9431917667388916, "learning_rate": 1.5000725373567388e-05, "loss": 0.0313, "step": 6992 }, { "epoch": 0.7581309627059843, "grad_norm": 3.101797103881836, "learning_rate": 1.5000000000000002e-05, "loss": 0.0175, "step": 6993 }, { "epoch": 0.7582393755420642, "grad_norm": 4.276869773864746, "learning_rate": 1.4999274626432614e-05, "loss": 0.0254, "step": 6994 }, { "epoch": 0.758347788378144, "grad_norm": 0.9609035849571228, "learning_rate": 1.4998549252865226e-05, "loss": 0.0202, "step": 6995 }, { "epoch": 0.7584562012142237, "grad_norm": 1.405021071434021, "learning_rate": 1.4997823879297839e-05, "loss": 0.0392, "step": 6996 }, { "epoch": 0.7585646140503035, "grad_norm": 1.734400987625122, "learning_rate": 1.4997098505730452e-05, "loss": 0.039, "step": 6997 }, { "epoch": 0.7586730268863834, "grad_norm": 1.4414633512496948, "learning_rate": 1.4996373132163064e-05, "loss": 0.0286, "step": 6998 }, { "epoch": 0.7587814397224631, "grad_norm": 2.7827045917510986, "learning_rate": 1.4995647758595677e-05, "loss": 0.0233, "step": 6999 }, { "epoch": 0.7588898525585429, "grad_norm": 2.4246649742126465, "learning_rate": 1.499492238502829e-05, "loss": 0.0681, "step": 7000 }, { "epoch": 0.7589982653946227, "grad_norm": 2.7934207916259766, "learning_rate": 1.4994197011460903e-05, "loss": 0.0752, "step": 7001 }, { "epoch": 0.7591066782307025, "grad_norm": 5.463825225830078, "learning_rate": 1.4993471637893518e-05, "loss": 0.0557, "step": 7002 }, { "epoch": 0.7592150910667823, "grad_norm": 2.010129690170288, "learning_rate": 1.499274626432613e-05, "loss": 0.0428, "step": 7003 }, { "epoch": 0.7593235039028621, "grad_norm": 3.0817384719848633, "learning_rate": 1.4992020890758742e-05, "loss": 0.0504, "step": 7004 }, { "epoch": 0.7594319167389418, "grad_norm": 1.5094221830368042, "learning_rate": 1.4991295517191356e-05, "loss": 0.0308, "step": 7005 }, { "epoch": 0.7595403295750217, "grad_norm": 1.0339715480804443, "learning_rate": 1.4990570143623968e-05, "loss": 0.0164, "step": 7006 }, { "epoch": 0.7596487424111015, "grad_norm": 10.368961334228516, "learning_rate": 1.498984477005658e-05, "loss": 0.0409, "step": 7007 }, { "epoch": 0.7597571552471812, "grad_norm": 1.4087146520614624, "learning_rate": 1.4989119396489193e-05, "loss": 0.0416, "step": 7008 }, { "epoch": 0.7598655680832611, "grad_norm": 3.4901106357574463, "learning_rate": 1.4988394022921807e-05, "loss": 0.0467, "step": 7009 }, { "epoch": 0.7599739809193409, "grad_norm": 2.3249757289886475, "learning_rate": 1.4987668649354419e-05, "loss": 0.0525, "step": 7010 }, { "epoch": 0.7600823937554206, "grad_norm": 4.864637851715088, "learning_rate": 1.498694327578703e-05, "loss": 0.0962, "step": 7011 }, { "epoch": 0.7601908065915004, "grad_norm": 1.2886468172073364, "learning_rate": 1.4986217902219645e-05, "loss": 0.0315, "step": 7012 }, { "epoch": 0.7602992194275803, "grad_norm": 5.983847141265869, "learning_rate": 1.4985492528652257e-05, "loss": 0.1118, "step": 7013 }, { "epoch": 0.76040763226366, "grad_norm": 4.542652606964111, "learning_rate": 1.4984767155084869e-05, "loss": 0.0372, "step": 7014 }, { "epoch": 0.7605160450997398, "grad_norm": 1.4600722789764404, "learning_rate": 1.4984041781517483e-05, "loss": 0.0248, "step": 7015 }, { "epoch": 0.7606244579358196, "grad_norm": 0.49207761883735657, "learning_rate": 1.4983316407950095e-05, "loss": 0.0091, "step": 7016 }, { "epoch": 0.7607328707718994, "grad_norm": 3.2150681018829346, "learning_rate": 1.4982591034382707e-05, "loss": 0.0632, "step": 7017 }, { "epoch": 0.7608412836079792, "grad_norm": 1.4717175960540771, "learning_rate": 1.4981865660815321e-05, "loss": 0.0276, "step": 7018 }, { "epoch": 0.760949696444059, "grad_norm": 1.0882749557495117, "learning_rate": 1.4981140287247935e-05, "loss": 0.0382, "step": 7019 }, { "epoch": 0.7610581092801387, "grad_norm": 2.8413164615631104, "learning_rate": 1.4980414913680547e-05, "loss": 0.032, "step": 7020 }, { "epoch": 0.7611665221162186, "grad_norm": 3.2795803546905518, "learning_rate": 1.497968954011316e-05, "loss": 0.0269, "step": 7021 }, { "epoch": 0.7612749349522984, "grad_norm": 1.699267864227295, "learning_rate": 1.4978964166545773e-05, "loss": 0.0329, "step": 7022 }, { "epoch": 0.7613833477883781, "grad_norm": 5.0462327003479, "learning_rate": 1.4978238792978385e-05, "loss": 0.0569, "step": 7023 }, { "epoch": 0.7614917606244579, "grad_norm": 5.859446048736572, "learning_rate": 1.4977513419410999e-05, "loss": 0.1097, "step": 7024 }, { "epoch": 0.7616001734605378, "grad_norm": 2.8738558292388916, "learning_rate": 1.4976788045843611e-05, "loss": 0.0401, "step": 7025 }, { "epoch": 0.7617085862966175, "grad_norm": 5.983516216278076, "learning_rate": 1.4976062672276223e-05, "loss": 0.0951, "step": 7026 }, { "epoch": 0.7618169991326973, "grad_norm": 1.1183831691741943, "learning_rate": 1.4975337298708837e-05, "loss": 0.0274, "step": 7027 }, { "epoch": 0.761925411968777, "grad_norm": 2.190769910812378, "learning_rate": 1.4974611925141449e-05, "loss": 0.0901, "step": 7028 }, { "epoch": 0.7620338248048569, "grad_norm": 2.1655542850494385, "learning_rate": 1.4973886551574061e-05, "loss": 0.0067, "step": 7029 }, { "epoch": 0.7621422376409367, "grad_norm": 0.6205065250396729, "learning_rate": 1.4973161178006675e-05, "loss": 0.0092, "step": 7030 }, { "epoch": 0.7622506504770165, "grad_norm": 0.7034154534339905, "learning_rate": 1.4972435804439287e-05, "loss": 0.008, "step": 7031 }, { "epoch": 0.7623590633130962, "grad_norm": 1.364152193069458, "learning_rate": 1.49717104308719e-05, "loss": 0.0311, "step": 7032 }, { "epoch": 0.7624674761491761, "grad_norm": 5.902780532836914, "learning_rate": 1.4970985057304512e-05, "loss": 0.019, "step": 7033 }, { "epoch": 0.7625758889852559, "grad_norm": 1.2314143180847168, "learning_rate": 1.4970259683737125e-05, "loss": 0.026, "step": 7034 }, { "epoch": 0.7626843018213356, "grad_norm": 2.822411298751831, "learning_rate": 1.4969534310169738e-05, "loss": 0.0544, "step": 7035 }, { "epoch": 0.7627927146574154, "grad_norm": 2.8768229484558105, "learning_rate": 1.496880893660235e-05, "loss": 0.0464, "step": 7036 }, { "epoch": 0.7629011274934953, "grad_norm": 2.5990993976593018, "learning_rate": 1.4968083563034965e-05, "loss": 0.0268, "step": 7037 }, { "epoch": 0.763009540329575, "grad_norm": 0.7444362640380859, "learning_rate": 1.4967358189467577e-05, "loss": 0.0211, "step": 7038 }, { "epoch": 0.7631179531656548, "grad_norm": 3.675602912902832, "learning_rate": 1.4966632815900191e-05, "loss": 0.0832, "step": 7039 }, { "epoch": 0.7632263660017347, "grad_norm": 3.119431495666504, "learning_rate": 1.4965907442332803e-05, "loss": 0.098, "step": 7040 }, { "epoch": 0.7633347788378144, "grad_norm": 1.7927945852279663, "learning_rate": 1.4965182068765415e-05, "loss": 0.019, "step": 7041 }, { "epoch": 0.7634431916738942, "grad_norm": 1.6284072399139404, "learning_rate": 1.496445669519803e-05, "loss": 0.0197, "step": 7042 }, { "epoch": 0.763551604509974, "grad_norm": 4.8613104820251465, "learning_rate": 1.4963731321630641e-05, "loss": 0.0554, "step": 7043 }, { "epoch": 0.7636600173460538, "grad_norm": 3.5591139793395996, "learning_rate": 1.4963005948063254e-05, "loss": 0.0459, "step": 7044 }, { "epoch": 0.7637684301821336, "grad_norm": 2.762779474258423, "learning_rate": 1.4962280574495867e-05, "loss": 0.0357, "step": 7045 }, { "epoch": 0.7638768430182133, "grad_norm": 2.970524549484253, "learning_rate": 1.496155520092848e-05, "loss": 0.0337, "step": 7046 }, { "epoch": 0.7639852558542931, "grad_norm": 1.2064146995544434, "learning_rate": 1.4960829827361092e-05, "loss": 0.0139, "step": 7047 }, { "epoch": 0.764093668690373, "grad_norm": 4.543856143951416, "learning_rate": 1.4960104453793704e-05, "loss": 0.0236, "step": 7048 }, { "epoch": 0.7642020815264527, "grad_norm": 2.4490370750427246, "learning_rate": 1.4959379080226318e-05, "loss": 0.06, "step": 7049 }, { "epoch": 0.7643104943625325, "grad_norm": 0.7389101982116699, "learning_rate": 1.495865370665893e-05, "loss": 0.0178, "step": 7050 }, { "epoch": 0.7644189071986123, "grad_norm": 3.110086441040039, "learning_rate": 1.4957928333091542e-05, "loss": 0.0226, "step": 7051 }, { "epoch": 0.7645273200346921, "grad_norm": 2.2671427726745605, "learning_rate": 1.4957202959524156e-05, "loss": 0.0196, "step": 7052 }, { "epoch": 0.7646357328707719, "grad_norm": 1.9954346418380737, "learning_rate": 1.4956477585956768e-05, "loss": 0.0218, "step": 7053 }, { "epoch": 0.7647441457068517, "grad_norm": 1.3248507976531982, "learning_rate": 1.4955752212389383e-05, "loss": 0.011, "step": 7054 }, { "epoch": 0.7648525585429314, "grad_norm": 2.1308488845825195, "learning_rate": 1.4955026838821996e-05, "loss": 0.0551, "step": 7055 }, { "epoch": 0.7649609713790113, "grad_norm": 7.294936656951904, "learning_rate": 1.4954301465254608e-05, "loss": 0.1299, "step": 7056 }, { "epoch": 0.7650693842150911, "grad_norm": 3.1340503692626953, "learning_rate": 1.4953576091687222e-05, "loss": 0.0189, "step": 7057 }, { "epoch": 0.7651777970511708, "grad_norm": 0.462787926197052, "learning_rate": 1.4952850718119834e-05, "loss": 0.0067, "step": 7058 }, { "epoch": 0.7652862098872506, "grad_norm": 3.1694304943084717, "learning_rate": 1.4952125344552446e-05, "loss": 0.0191, "step": 7059 }, { "epoch": 0.7653946227233305, "grad_norm": 1.927847981452942, "learning_rate": 1.4951399970985058e-05, "loss": 0.0298, "step": 7060 }, { "epoch": 0.7655030355594102, "grad_norm": 0.608572244644165, "learning_rate": 1.4950674597417672e-05, "loss": 0.0127, "step": 7061 }, { "epoch": 0.76561144839549, "grad_norm": 5.482871055603027, "learning_rate": 1.4949949223850284e-05, "loss": 0.0386, "step": 7062 }, { "epoch": 0.7657198612315698, "grad_norm": 1.8556092977523804, "learning_rate": 1.4949223850282896e-05, "loss": 0.0183, "step": 7063 }, { "epoch": 0.7658282740676496, "grad_norm": 2.8905179500579834, "learning_rate": 1.494849847671551e-05, "loss": 0.0405, "step": 7064 }, { "epoch": 0.7659366869037294, "grad_norm": 4.186559200286865, "learning_rate": 1.4947773103148122e-05, "loss": 0.0599, "step": 7065 }, { "epoch": 0.7660450997398092, "grad_norm": 1.1022685766220093, "learning_rate": 1.4947047729580734e-05, "loss": 0.0207, "step": 7066 }, { "epoch": 0.766153512575889, "grad_norm": 0.8414877653121948, "learning_rate": 1.4946322356013348e-05, "loss": 0.0344, "step": 7067 }, { "epoch": 0.7662619254119688, "grad_norm": 3.377225399017334, "learning_rate": 1.494559698244596e-05, "loss": 0.0612, "step": 7068 }, { "epoch": 0.7663703382480486, "grad_norm": 3.633744955062866, "learning_rate": 1.4944871608878572e-05, "loss": 0.0624, "step": 7069 }, { "epoch": 0.7664787510841283, "grad_norm": 1.0973869562149048, "learning_rate": 1.4944146235311186e-05, "loss": 0.0409, "step": 7070 }, { "epoch": 0.7665871639202082, "grad_norm": 4.947967529296875, "learning_rate": 1.49434208617438e-05, "loss": 0.0202, "step": 7071 }, { "epoch": 0.766695576756288, "grad_norm": 2.704719305038452, "learning_rate": 1.4942695488176412e-05, "loss": 0.0875, "step": 7072 }, { "epoch": 0.7668039895923677, "grad_norm": 1.029467225074768, "learning_rate": 1.4941970114609026e-05, "loss": 0.0165, "step": 7073 }, { "epoch": 0.7669124024284475, "grad_norm": 3.4509472846984863, "learning_rate": 1.4941244741041638e-05, "loss": 0.0654, "step": 7074 }, { "epoch": 0.7670208152645274, "grad_norm": 1.5691351890563965, "learning_rate": 1.494051936747425e-05, "loss": 0.0256, "step": 7075 }, { "epoch": 0.7671292281006071, "grad_norm": 0.7772897481918335, "learning_rate": 1.4939793993906864e-05, "loss": 0.0255, "step": 7076 }, { "epoch": 0.7672376409366869, "grad_norm": 2.1541872024536133, "learning_rate": 1.4939068620339476e-05, "loss": 0.0575, "step": 7077 }, { "epoch": 0.7673460537727667, "grad_norm": 3.5120224952697754, "learning_rate": 1.4938343246772088e-05, "loss": 0.03, "step": 7078 }, { "epoch": 0.7674544666088465, "grad_norm": 2.056581497192383, "learning_rate": 1.4937617873204702e-05, "loss": 0.0226, "step": 7079 }, { "epoch": 0.7675628794449263, "grad_norm": 1.9716583490371704, "learning_rate": 1.4936892499637314e-05, "loss": 0.0346, "step": 7080 }, { "epoch": 0.7676712922810061, "grad_norm": 1.7872120141983032, "learning_rate": 1.4936167126069927e-05, "loss": 0.0161, "step": 7081 }, { "epoch": 0.7677797051170858, "grad_norm": 1.776415228843689, "learning_rate": 1.493544175250254e-05, "loss": 0.0304, "step": 7082 }, { "epoch": 0.7678881179531657, "grad_norm": 3.0082221031188965, "learning_rate": 1.4934716378935153e-05, "loss": 0.0858, "step": 7083 }, { "epoch": 0.7679965307892455, "grad_norm": 0.5656991004943848, "learning_rate": 1.4933991005367765e-05, "loss": 0.0082, "step": 7084 }, { "epoch": 0.7681049436253252, "grad_norm": 3.8892910480499268, "learning_rate": 1.4933265631800377e-05, "loss": 0.0945, "step": 7085 }, { "epoch": 0.768213356461405, "grad_norm": 0.7475916147232056, "learning_rate": 1.493254025823299e-05, "loss": 0.0199, "step": 7086 }, { "epoch": 0.7683217692974849, "grad_norm": 6.24436092376709, "learning_rate": 1.4931814884665603e-05, "loss": 0.0292, "step": 7087 }, { "epoch": 0.7684301821335646, "grad_norm": 0.7963841557502747, "learning_rate": 1.4931089511098215e-05, "loss": 0.0177, "step": 7088 }, { "epoch": 0.7685385949696444, "grad_norm": 3.4233837127685547, "learning_rate": 1.493036413753083e-05, "loss": 0.0245, "step": 7089 }, { "epoch": 0.7686470078057241, "grad_norm": 3.5476431846618652, "learning_rate": 1.4929638763963443e-05, "loss": 0.0334, "step": 7090 }, { "epoch": 0.768755420641804, "grad_norm": 0.7859734892845154, "learning_rate": 1.4928913390396057e-05, "loss": 0.0136, "step": 7091 }, { "epoch": 0.7688638334778838, "grad_norm": 1.8078235387802124, "learning_rate": 1.4928188016828669e-05, "loss": 0.0338, "step": 7092 }, { "epoch": 0.7689722463139635, "grad_norm": 2.5001747608184814, "learning_rate": 1.492746264326128e-05, "loss": 0.0349, "step": 7093 }, { "epoch": 0.7690806591500433, "grad_norm": 0.44683101773262024, "learning_rate": 1.4926737269693895e-05, "loss": 0.0137, "step": 7094 }, { "epoch": 0.7691890719861232, "grad_norm": 4.099231243133545, "learning_rate": 1.4926011896126507e-05, "loss": 0.0679, "step": 7095 }, { "epoch": 0.769297484822203, "grad_norm": 0.3989224135875702, "learning_rate": 1.4925286522559119e-05, "loss": 0.0201, "step": 7096 }, { "epoch": 0.7694058976582827, "grad_norm": 0.3283509314060211, "learning_rate": 1.4924561148991733e-05, "loss": 0.0038, "step": 7097 }, { "epoch": 0.7695143104943626, "grad_norm": 5.0875372886657715, "learning_rate": 1.4923835775424345e-05, "loss": 0.0966, "step": 7098 }, { "epoch": 0.7696227233304423, "grad_norm": 4.335901260375977, "learning_rate": 1.4923110401856957e-05, "loss": 0.038, "step": 7099 }, { "epoch": 0.7697311361665221, "grad_norm": 8.929465293884277, "learning_rate": 1.492238502828957e-05, "loss": 0.1272, "step": 7100 }, { "epoch": 0.7698395490026019, "grad_norm": 1.1884013414382935, "learning_rate": 1.4921659654722183e-05, "loss": 0.0117, "step": 7101 }, { "epoch": 0.7699479618386817, "grad_norm": 3.6984143257141113, "learning_rate": 1.4920934281154795e-05, "loss": 0.0386, "step": 7102 }, { "epoch": 0.7700563746747615, "grad_norm": 4.081668853759766, "learning_rate": 1.4920208907587407e-05, "loss": 0.0414, "step": 7103 }, { "epoch": 0.7701647875108413, "grad_norm": 2.731553792953491, "learning_rate": 1.4919483534020021e-05, "loss": 0.0447, "step": 7104 }, { "epoch": 0.770273200346921, "grad_norm": 2.2875940799713135, "learning_rate": 1.4918758160452633e-05, "loss": 0.019, "step": 7105 }, { "epoch": 0.7703816131830009, "grad_norm": 2.3964056968688965, "learning_rate": 1.4918032786885249e-05, "loss": 0.0246, "step": 7106 }, { "epoch": 0.7704900260190807, "grad_norm": 2.2197344303131104, "learning_rate": 1.4917307413317861e-05, "loss": 0.0203, "step": 7107 }, { "epoch": 0.7705984388551604, "grad_norm": 9.40040397644043, "learning_rate": 1.4916582039750473e-05, "loss": 0.085, "step": 7108 }, { "epoch": 0.7707068516912402, "grad_norm": 0.3591688573360443, "learning_rate": 1.4915856666183087e-05, "loss": 0.0082, "step": 7109 }, { "epoch": 0.7708152645273201, "grad_norm": 2.537837505340576, "learning_rate": 1.4915131292615699e-05, "loss": 0.0403, "step": 7110 }, { "epoch": 0.7709236773633998, "grad_norm": 3.5296339988708496, "learning_rate": 1.4914405919048311e-05, "loss": 0.07, "step": 7111 }, { "epoch": 0.7710320901994796, "grad_norm": 2.394172430038452, "learning_rate": 1.4913680545480923e-05, "loss": 0.0677, "step": 7112 }, { "epoch": 0.7711405030355594, "grad_norm": 3.489595651626587, "learning_rate": 1.4912955171913537e-05, "loss": 0.0847, "step": 7113 }, { "epoch": 0.7712489158716392, "grad_norm": 2.0790860652923584, "learning_rate": 1.491222979834615e-05, "loss": 0.031, "step": 7114 }, { "epoch": 0.771357328707719, "grad_norm": 4.380008220672607, "learning_rate": 1.4911504424778761e-05, "loss": 0.0471, "step": 7115 }, { "epoch": 0.7714657415437988, "grad_norm": 1.3697559833526611, "learning_rate": 1.4910779051211375e-05, "loss": 0.0237, "step": 7116 }, { "epoch": 0.7715741543798785, "grad_norm": 3.1444010734558105, "learning_rate": 1.4910053677643987e-05, "loss": 0.0362, "step": 7117 }, { "epoch": 0.7716825672159584, "grad_norm": 0.5573845505714417, "learning_rate": 1.49093283040766e-05, "loss": 0.0114, "step": 7118 }, { "epoch": 0.7717909800520382, "grad_norm": 1.1554348468780518, "learning_rate": 1.4908602930509213e-05, "loss": 0.0123, "step": 7119 }, { "epoch": 0.7718993928881179, "grad_norm": 2.810413122177124, "learning_rate": 1.4907877556941826e-05, "loss": 0.026, "step": 7120 }, { "epoch": 0.7720078057241977, "grad_norm": 2.4604923725128174, "learning_rate": 1.4907152183374438e-05, "loss": 0.0675, "step": 7121 }, { "epoch": 0.7721162185602776, "grad_norm": 4.366931438446045, "learning_rate": 1.4906426809807052e-05, "loss": 0.0619, "step": 7122 }, { "epoch": 0.7722246313963573, "grad_norm": 2.482496976852417, "learning_rate": 1.4905701436239665e-05, "loss": 0.0296, "step": 7123 }, { "epoch": 0.7723330442324371, "grad_norm": 1.403544306755066, "learning_rate": 1.4904976062672278e-05, "loss": 0.0281, "step": 7124 }, { "epoch": 0.772441457068517, "grad_norm": 1.3277299404144287, "learning_rate": 1.4904250689104891e-05, "loss": 0.0509, "step": 7125 }, { "epoch": 0.7725498699045967, "grad_norm": 0.8708651065826416, "learning_rate": 1.4903525315537504e-05, "loss": 0.007, "step": 7126 }, { "epoch": 0.7726582827406765, "grad_norm": 1.9712580442428589, "learning_rate": 1.4902799941970116e-05, "loss": 0.0311, "step": 7127 }, { "epoch": 0.7727666955767563, "grad_norm": 4.9510884284973145, "learning_rate": 1.490207456840273e-05, "loss": 0.0902, "step": 7128 }, { "epoch": 0.7728751084128361, "grad_norm": 0.5923239588737488, "learning_rate": 1.4901349194835342e-05, "loss": 0.0253, "step": 7129 }, { "epoch": 0.7729835212489159, "grad_norm": 0.5762630701065063, "learning_rate": 1.4900623821267954e-05, "loss": 0.0153, "step": 7130 }, { "epoch": 0.7730919340849957, "grad_norm": 1.6260169744491577, "learning_rate": 1.4899898447700568e-05, "loss": 0.0208, "step": 7131 }, { "epoch": 0.7732003469210754, "grad_norm": 1.1613479852676392, "learning_rate": 1.489917307413318e-05, "loss": 0.021, "step": 7132 }, { "epoch": 0.7733087597571553, "grad_norm": 1.7045859098434448, "learning_rate": 1.4898447700565792e-05, "loss": 0.0187, "step": 7133 }, { "epoch": 0.7734171725932351, "grad_norm": 2.9240760803222656, "learning_rate": 1.4897722326998406e-05, "loss": 0.0512, "step": 7134 }, { "epoch": 0.7735255854293148, "grad_norm": 2.1579790115356445, "learning_rate": 1.4896996953431018e-05, "loss": 0.0249, "step": 7135 }, { "epoch": 0.7736339982653946, "grad_norm": 4.817594528198242, "learning_rate": 1.489627157986363e-05, "loss": 0.0367, "step": 7136 }, { "epoch": 0.7737424111014745, "grad_norm": 2.090439796447754, "learning_rate": 1.4895546206296242e-05, "loss": 0.0385, "step": 7137 }, { "epoch": 0.7738508239375542, "grad_norm": 6.700855255126953, "learning_rate": 1.4894820832728856e-05, "loss": 0.0552, "step": 7138 }, { "epoch": 0.773959236773634, "grad_norm": 4.548102378845215, "learning_rate": 1.4894095459161468e-05, "loss": 0.0455, "step": 7139 }, { "epoch": 0.7740676496097137, "grad_norm": 2.4372963905334473, "learning_rate": 1.4893370085594084e-05, "loss": 0.0439, "step": 7140 }, { "epoch": 0.7741760624457936, "grad_norm": 4.071413993835449, "learning_rate": 1.4892644712026696e-05, "loss": 0.0855, "step": 7141 }, { "epoch": 0.7742844752818734, "grad_norm": 1.0907107591629028, "learning_rate": 1.4891919338459308e-05, "loss": 0.0169, "step": 7142 }, { "epoch": 0.7743928881179531, "grad_norm": 1.055619239807129, "learning_rate": 1.4891193964891922e-05, "loss": 0.0097, "step": 7143 }, { "epoch": 0.7745013009540329, "grad_norm": 1.4049427509307861, "learning_rate": 1.4890468591324534e-05, "loss": 0.0255, "step": 7144 }, { "epoch": 0.7746097137901128, "grad_norm": 3.024822473526001, "learning_rate": 1.4889743217757146e-05, "loss": 0.0274, "step": 7145 }, { "epoch": 0.7747181266261925, "grad_norm": 3.7135848999023438, "learning_rate": 1.488901784418976e-05, "loss": 0.1178, "step": 7146 }, { "epoch": 0.7748265394622723, "grad_norm": 0.8257648944854736, "learning_rate": 1.4888292470622372e-05, "loss": 0.0092, "step": 7147 }, { "epoch": 0.7749349522983521, "grad_norm": 3.7109434604644775, "learning_rate": 1.4887567097054984e-05, "loss": 0.0445, "step": 7148 }, { "epoch": 0.775043365134432, "grad_norm": 3.4346961975097656, "learning_rate": 1.4886841723487598e-05, "loss": 0.0598, "step": 7149 }, { "epoch": 0.7751517779705117, "grad_norm": 1.7815176248550415, "learning_rate": 1.488611634992021e-05, "loss": 0.0279, "step": 7150 }, { "epoch": 0.7752601908065915, "grad_norm": 2.4110500812530518, "learning_rate": 1.4885390976352822e-05, "loss": 0.0653, "step": 7151 }, { "epoch": 0.7753686036426712, "grad_norm": 1.9651907682418823, "learning_rate": 1.4884665602785435e-05, "loss": 0.0453, "step": 7152 }, { "epoch": 0.7754770164787511, "grad_norm": 2.7594640254974365, "learning_rate": 1.4883940229218048e-05, "loss": 0.0562, "step": 7153 }, { "epoch": 0.7755854293148309, "grad_norm": 5.158276081085205, "learning_rate": 1.488321485565066e-05, "loss": 0.0076, "step": 7154 }, { "epoch": 0.7756938421509106, "grad_norm": 1.5126044750213623, "learning_rate": 1.4882489482083273e-05, "loss": 0.0379, "step": 7155 }, { "epoch": 0.7758022549869905, "grad_norm": 1.6114568710327148, "learning_rate": 1.4881764108515886e-05, "loss": 0.0856, "step": 7156 }, { "epoch": 0.7759106678230703, "grad_norm": 2.4548940658569336, "learning_rate": 1.4881038734948499e-05, "loss": 0.0462, "step": 7157 }, { "epoch": 0.77601908065915, "grad_norm": 2.0807414054870605, "learning_rate": 1.4880313361381114e-05, "loss": 0.0741, "step": 7158 }, { "epoch": 0.7761274934952298, "grad_norm": 2.2523512840270996, "learning_rate": 1.4879587987813726e-05, "loss": 0.0469, "step": 7159 }, { "epoch": 0.7762359063313097, "grad_norm": 1.1330864429473877, "learning_rate": 1.4878862614246338e-05, "loss": 0.015, "step": 7160 }, { "epoch": 0.7763443191673894, "grad_norm": 1.0497878789901733, "learning_rate": 1.4878137240678952e-05, "loss": 0.0213, "step": 7161 }, { "epoch": 0.7764527320034692, "grad_norm": 0.9804829359054565, "learning_rate": 1.4877411867111564e-05, "loss": 0.012, "step": 7162 }, { "epoch": 0.776561144839549, "grad_norm": 3.511430501937866, "learning_rate": 1.4876686493544177e-05, "loss": 0.0461, "step": 7163 }, { "epoch": 0.7766695576756288, "grad_norm": 1.4172418117523193, "learning_rate": 1.4875961119976789e-05, "loss": 0.0486, "step": 7164 }, { "epoch": 0.7767779705117086, "grad_norm": 3.899444818496704, "learning_rate": 1.4875235746409403e-05, "loss": 0.0633, "step": 7165 }, { "epoch": 0.7768863833477884, "grad_norm": 2.4944136142730713, "learning_rate": 1.4874510372842015e-05, "loss": 0.0407, "step": 7166 }, { "epoch": 0.7769947961838681, "grad_norm": 3.010653257369995, "learning_rate": 1.4873784999274627e-05, "loss": 0.0305, "step": 7167 }, { "epoch": 0.777103209019948, "grad_norm": 2.0635619163513184, "learning_rate": 1.487305962570724e-05, "loss": 0.0389, "step": 7168 }, { "epoch": 0.7772116218560278, "grad_norm": 1.040834665298462, "learning_rate": 1.4872334252139853e-05, "loss": 0.0413, "step": 7169 }, { "epoch": 0.7773200346921075, "grad_norm": 1.3756272792816162, "learning_rate": 1.4871608878572465e-05, "loss": 0.0327, "step": 7170 }, { "epoch": 0.7774284475281873, "grad_norm": 1.236547827720642, "learning_rate": 1.4870883505005079e-05, "loss": 0.0249, "step": 7171 }, { "epoch": 0.7775368603642672, "grad_norm": 1.4807426929473877, "learning_rate": 1.4870158131437691e-05, "loss": 0.0216, "step": 7172 }, { "epoch": 0.7776452732003469, "grad_norm": 1.730705976486206, "learning_rate": 1.4869432757870303e-05, "loss": 0.0087, "step": 7173 }, { "epoch": 0.7777536860364267, "grad_norm": 2.4068970680236816, "learning_rate": 1.4868707384302917e-05, "loss": 0.0322, "step": 7174 }, { "epoch": 0.7778620988725065, "grad_norm": 0.8525661826133728, "learning_rate": 1.486798201073553e-05, "loss": 0.0213, "step": 7175 }, { "epoch": 0.7779705117085863, "grad_norm": 0.7970444560050964, "learning_rate": 1.4867256637168143e-05, "loss": 0.0178, "step": 7176 }, { "epoch": 0.7780789245446661, "grad_norm": 1.626834511756897, "learning_rate": 1.4866531263600757e-05, "loss": 0.03, "step": 7177 }, { "epoch": 0.7781873373807459, "grad_norm": 1.3392653465270996, "learning_rate": 1.4865805890033369e-05, "loss": 0.0116, "step": 7178 }, { "epoch": 0.7782957502168256, "grad_norm": 1.8642032146453857, "learning_rate": 1.4865080516465981e-05, "loss": 0.0327, "step": 7179 }, { "epoch": 0.7784041630529055, "grad_norm": 3.0650811195373535, "learning_rate": 1.4864355142898595e-05, "loss": 0.0301, "step": 7180 }, { "epoch": 0.7785125758889853, "grad_norm": 6.481222152709961, "learning_rate": 1.4863629769331207e-05, "loss": 0.0605, "step": 7181 }, { "epoch": 0.778620988725065, "grad_norm": 1.4560561180114746, "learning_rate": 1.4862904395763819e-05, "loss": 0.0478, "step": 7182 }, { "epoch": 0.7787294015611448, "grad_norm": 1.8332586288452148, "learning_rate": 1.4862179022196433e-05, "loss": 0.0353, "step": 7183 }, { "epoch": 0.7788378143972247, "grad_norm": 3.842479705810547, "learning_rate": 1.4861453648629045e-05, "loss": 0.0595, "step": 7184 }, { "epoch": 0.7789462272333044, "grad_norm": 1.2874248027801514, "learning_rate": 1.4860728275061657e-05, "loss": 0.0306, "step": 7185 }, { "epoch": 0.7790546400693842, "grad_norm": 0.6459070444107056, "learning_rate": 1.4860002901494271e-05, "loss": 0.011, "step": 7186 }, { "epoch": 0.7791630529054641, "grad_norm": 1.7847273349761963, "learning_rate": 1.4859277527926883e-05, "loss": 0.037, "step": 7187 }, { "epoch": 0.7792714657415438, "grad_norm": 4.126032829284668, "learning_rate": 1.4858552154359495e-05, "loss": 0.0803, "step": 7188 }, { "epoch": 0.7793798785776236, "grad_norm": 7.602262496948242, "learning_rate": 1.4857826780792108e-05, "loss": 0.0187, "step": 7189 }, { "epoch": 0.7794882914137033, "grad_norm": 5.0502824783325195, "learning_rate": 1.4857101407224721e-05, "loss": 0.0413, "step": 7190 }, { "epoch": 0.7795967042497832, "grad_norm": 3.2149832248687744, "learning_rate": 1.4856376033657334e-05, "loss": 0.0541, "step": 7191 }, { "epoch": 0.779705117085863, "grad_norm": 4.321658611297607, "learning_rate": 1.4855650660089949e-05, "loss": 0.0871, "step": 7192 }, { "epoch": 0.7798135299219427, "grad_norm": 0.6103509068489075, "learning_rate": 1.4854925286522561e-05, "loss": 0.0109, "step": 7193 }, { "epoch": 0.7799219427580225, "grad_norm": 2.3636209964752197, "learning_rate": 1.4854199912955173e-05, "loss": 0.0859, "step": 7194 }, { "epoch": 0.7800303555941024, "grad_norm": 1.2329226732254028, "learning_rate": 1.4853474539387787e-05, "loss": 0.019, "step": 7195 }, { "epoch": 0.7801387684301822, "grad_norm": 2.334106683731079, "learning_rate": 1.48527491658204e-05, "loss": 0.0328, "step": 7196 }, { "epoch": 0.7802471812662619, "grad_norm": 1.3041492700576782, "learning_rate": 1.4852023792253011e-05, "loss": 0.0308, "step": 7197 }, { "epoch": 0.7803555941023417, "grad_norm": 4.077573776245117, "learning_rate": 1.4851298418685625e-05, "loss": 0.0652, "step": 7198 }, { "epoch": 0.7804640069384216, "grad_norm": 1.502652883529663, "learning_rate": 1.4850573045118237e-05, "loss": 0.0264, "step": 7199 }, { "epoch": 0.7805724197745013, "grad_norm": 1.50059974193573, "learning_rate": 1.484984767155085e-05, "loss": 0.0227, "step": 7200 }, { "epoch": 0.7806808326105811, "grad_norm": 3.5923404693603516, "learning_rate": 1.4849122297983463e-05, "loss": 0.0455, "step": 7201 }, { "epoch": 0.7807892454466608, "grad_norm": 3.045184850692749, "learning_rate": 1.4848396924416076e-05, "loss": 0.0403, "step": 7202 }, { "epoch": 0.7808976582827407, "grad_norm": 6.694420337677002, "learning_rate": 1.4847671550848688e-05, "loss": 0.0365, "step": 7203 }, { "epoch": 0.7810060711188205, "grad_norm": 1.920149564743042, "learning_rate": 1.48469461772813e-05, "loss": 0.0455, "step": 7204 }, { "epoch": 0.7811144839549002, "grad_norm": 0.5821015238761902, "learning_rate": 1.4846220803713914e-05, "loss": 0.0176, "step": 7205 }, { "epoch": 0.78122289679098, "grad_norm": 3.4075329303741455, "learning_rate": 1.4845495430146526e-05, "loss": 0.0423, "step": 7206 }, { "epoch": 0.7813313096270599, "grad_norm": 1.4490352869033813, "learning_rate": 1.4844770056579138e-05, "loss": 0.031, "step": 7207 }, { "epoch": 0.7814397224631396, "grad_norm": 1.6824225187301636, "learning_rate": 1.4844044683011752e-05, "loss": 0.0397, "step": 7208 }, { "epoch": 0.7815481352992194, "grad_norm": 2.18235182762146, "learning_rate": 1.4843319309444366e-05, "loss": 0.027, "step": 7209 }, { "epoch": 0.7816565481352992, "grad_norm": 1.9830330610275269, "learning_rate": 1.484259393587698e-05, "loss": 0.0397, "step": 7210 }, { "epoch": 0.781764960971379, "grad_norm": 2.4862918853759766, "learning_rate": 1.4841868562309592e-05, "loss": 0.0253, "step": 7211 }, { "epoch": 0.7818733738074588, "grad_norm": 1.8300896883010864, "learning_rate": 1.4841143188742204e-05, "loss": 0.0191, "step": 7212 }, { "epoch": 0.7819817866435386, "grad_norm": 0.6457175612449646, "learning_rate": 1.4840417815174818e-05, "loss": 0.0159, "step": 7213 }, { "epoch": 0.7820901994796184, "grad_norm": 1.7358357906341553, "learning_rate": 1.483969244160743e-05, "loss": 0.0318, "step": 7214 }, { "epoch": 0.7821986123156982, "grad_norm": 3.1996686458587646, "learning_rate": 1.4838967068040042e-05, "loss": 0.033, "step": 7215 }, { "epoch": 0.782307025151778, "grad_norm": 2.8833162784576416, "learning_rate": 1.4838241694472654e-05, "loss": 0.0465, "step": 7216 }, { "epoch": 0.7824154379878577, "grad_norm": 0.9913215637207031, "learning_rate": 1.4837516320905268e-05, "loss": 0.0178, "step": 7217 }, { "epoch": 0.7825238508239376, "grad_norm": 4.060728073120117, "learning_rate": 1.483679094733788e-05, "loss": 0.0552, "step": 7218 }, { "epoch": 0.7826322636600174, "grad_norm": 6.53836727142334, "learning_rate": 1.4836065573770492e-05, "loss": 0.0314, "step": 7219 }, { "epoch": 0.7827406764960971, "grad_norm": 0.9754031896591187, "learning_rate": 1.4835340200203106e-05, "loss": 0.0087, "step": 7220 }, { "epoch": 0.7828490893321769, "grad_norm": 3.6099753379821777, "learning_rate": 1.4834614826635718e-05, "loss": 0.0632, "step": 7221 }, { "epoch": 0.7829575021682568, "grad_norm": 1.091518759727478, "learning_rate": 1.483388945306833e-05, "loss": 0.0142, "step": 7222 }, { "epoch": 0.7830659150043365, "grad_norm": 1.111801266670227, "learning_rate": 1.4833164079500944e-05, "loss": 0.0192, "step": 7223 }, { "epoch": 0.7831743278404163, "grad_norm": 6.190621376037598, "learning_rate": 1.4832438705933556e-05, "loss": 0.0512, "step": 7224 }, { "epoch": 0.7832827406764961, "grad_norm": 4.446232795715332, "learning_rate": 1.4831713332366168e-05, "loss": 0.0358, "step": 7225 }, { "epoch": 0.7833911535125759, "grad_norm": 1.1163474321365356, "learning_rate": 1.4830987958798782e-05, "loss": 0.0097, "step": 7226 }, { "epoch": 0.7834995663486557, "grad_norm": 4.511983871459961, "learning_rate": 1.4830262585231396e-05, "loss": 0.0263, "step": 7227 }, { "epoch": 0.7836079791847355, "grad_norm": 1.2370868921279907, "learning_rate": 1.4829537211664008e-05, "loss": 0.0255, "step": 7228 }, { "epoch": 0.7837163920208152, "grad_norm": 2.326427459716797, "learning_rate": 1.4828811838096622e-05, "loss": 0.0435, "step": 7229 }, { "epoch": 0.7838248048568951, "grad_norm": 4.000243663787842, "learning_rate": 1.4828086464529234e-05, "loss": 0.0451, "step": 7230 }, { "epoch": 0.7839332176929749, "grad_norm": 2.5583655834198, "learning_rate": 1.4827361090961846e-05, "loss": 0.0424, "step": 7231 }, { "epoch": 0.7840416305290546, "grad_norm": 3.265003204345703, "learning_rate": 1.482663571739446e-05, "loss": 0.0624, "step": 7232 }, { "epoch": 0.7841500433651344, "grad_norm": 1.5800001621246338, "learning_rate": 1.4825910343827072e-05, "loss": 0.0304, "step": 7233 }, { "epoch": 0.7842584562012143, "grad_norm": 1.488728642463684, "learning_rate": 1.4825184970259684e-05, "loss": 0.0405, "step": 7234 }, { "epoch": 0.784366869037294, "grad_norm": 0.13242144882678986, "learning_rate": 1.4824459596692298e-05, "loss": 0.0013, "step": 7235 }, { "epoch": 0.7844752818733738, "grad_norm": 2.3500826358795166, "learning_rate": 1.482373422312491e-05, "loss": 0.0873, "step": 7236 }, { "epoch": 0.7845836947094535, "grad_norm": 2.7544496059417725, "learning_rate": 1.4823008849557523e-05, "loss": 0.0059, "step": 7237 }, { "epoch": 0.7846921075455334, "grad_norm": 3.124185800552368, "learning_rate": 1.4822283475990136e-05, "loss": 0.0388, "step": 7238 }, { "epoch": 0.7848005203816132, "grad_norm": 1.2174584865570068, "learning_rate": 1.4821558102422749e-05, "loss": 0.0241, "step": 7239 }, { "epoch": 0.784908933217693, "grad_norm": 3.958803415298462, "learning_rate": 1.482083272885536e-05, "loss": 0.0564, "step": 7240 }, { "epoch": 0.7850173460537727, "grad_norm": 1.47871732711792, "learning_rate": 1.4820107355287975e-05, "loss": 0.0253, "step": 7241 }, { "epoch": 0.7851257588898526, "grad_norm": 1.5943900346755981, "learning_rate": 1.4819381981720587e-05, "loss": 0.0659, "step": 7242 }, { "epoch": 0.7852341717259324, "grad_norm": 3.473331928253174, "learning_rate": 1.4818656608153199e-05, "loss": 0.0424, "step": 7243 }, { "epoch": 0.7853425845620121, "grad_norm": 1.7572031021118164, "learning_rate": 1.4817931234585814e-05, "loss": 0.0314, "step": 7244 }, { "epoch": 0.785450997398092, "grad_norm": 2.220207691192627, "learning_rate": 1.4817205861018427e-05, "loss": 0.0497, "step": 7245 }, { "epoch": 0.7855594102341718, "grad_norm": 0.7081645727157593, "learning_rate": 1.4816480487451039e-05, "loss": 0.02, "step": 7246 }, { "epoch": 0.7856678230702515, "grad_norm": 3.8483834266662598, "learning_rate": 1.4815755113883653e-05, "loss": 0.038, "step": 7247 }, { "epoch": 0.7857762359063313, "grad_norm": 1.9871065616607666, "learning_rate": 1.4815029740316265e-05, "loss": 0.0286, "step": 7248 }, { "epoch": 0.7858846487424112, "grad_norm": 2.1800127029418945, "learning_rate": 1.4814304366748877e-05, "loss": 0.0214, "step": 7249 }, { "epoch": 0.7859930615784909, "grad_norm": 1.0385466814041138, "learning_rate": 1.481357899318149e-05, "loss": 0.0087, "step": 7250 }, { "epoch": 0.7861014744145707, "grad_norm": 6.359253883361816, "learning_rate": 1.4812853619614103e-05, "loss": 0.045, "step": 7251 }, { "epoch": 0.7862098872506504, "grad_norm": 1.2802832126617432, "learning_rate": 1.4812128246046715e-05, "loss": 0.0158, "step": 7252 }, { "epoch": 0.7863183000867303, "grad_norm": 4.91099739074707, "learning_rate": 1.4811402872479329e-05, "loss": 0.0402, "step": 7253 }, { "epoch": 0.7864267129228101, "grad_norm": 1.40901517868042, "learning_rate": 1.4810677498911941e-05, "loss": 0.0138, "step": 7254 }, { "epoch": 0.7865351257588898, "grad_norm": 1.794016718864441, "learning_rate": 1.4809952125344553e-05, "loss": 0.0236, "step": 7255 }, { "epoch": 0.7866435385949696, "grad_norm": 2.503628730773926, "learning_rate": 1.4809226751777165e-05, "loss": 0.0335, "step": 7256 }, { "epoch": 0.7867519514310495, "grad_norm": 2.1348156929016113, "learning_rate": 1.4808501378209779e-05, "loss": 0.0302, "step": 7257 }, { "epoch": 0.7868603642671292, "grad_norm": 3.1254220008850098, "learning_rate": 1.4807776004642391e-05, "loss": 0.0439, "step": 7258 }, { "epoch": 0.786968777103209, "grad_norm": 3.340156316757202, "learning_rate": 1.4807050631075003e-05, "loss": 0.0749, "step": 7259 }, { "epoch": 0.7870771899392888, "grad_norm": 0.6504594683647156, "learning_rate": 1.4806325257507617e-05, "loss": 0.011, "step": 7260 }, { "epoch": 0.7871856027753686, "grad_norm": 4.145257949829102, "learning_rate": 1.4805599883940231e-05, "loss": 0.1046, "step": 7261 }, { "epoch": 0.7872940156114484, "grad_norm": 3.7596004009246826, "learning_rate": 1.4804874510372845e-05, "loss": 0.0543, "step": 7262 }, { "epoch": 0.7874024284475282, "grad_norm": 0.6743737459182739, "learning_rate": 1.4804149136805457e-05, "loss": 0.0126, "step": 7263 }, { "epoch": 0.7875108412836079, "grad_norm": 1.7589654922485352, "learning_rate": 1.4803423763238069e-05, "loss": 0.0239, "step": 7264 }, { "epoch": 0.7876192541196878, "grad_norm": 1.361156702041626, "learning_rate": 1.4802698389670683e-05, "loss": 0.0178, "step": 7265 }, { "epoch": 0.7877276669557676, "grad_norm": 3.0581929683685303, "learning_rate": 1.4801973016103295e-05, "loss": 0.0092, "step": 7266 }, { "epoch": 0.7878360797918473, "grad_norm": 3.131403684616089, "learning_rate": 1.4801247642535907e-05, "loss": 0.038, "step": 7267 }, { "epoch": 0.7879444926279271, "grad_norm": 4.540439128875732, "learning_rate": 1.480052226896852e-05, "loss": 0.0265, "step": 7268 }, { "epoch": 0.788052905464007, "grad_norm": 5.103579044342041, "learning_rate": 1.4799796895401133e-05, "loss": 0.1103, "step": 7269 }, { "epoch": 0.7881613183000867, "grad_norm": 4.777495861053467, "learning_rate": 1.4799071521833745e-05, "loss": 0.0659, "step": 7270 }, { "epoch": 0.7882697311361665, "grad_norm": 2.048508405685425, "learning_rate": 1.4798346148266358e-05, "loss": 0.0349, "step": 7271 }, { "epoch": 0.7883781439722463, "grad_norm": 2.4277150630950928, "learning_rate": 1.4797620774698971e-05, "loss": 0.0226, "step": 7272 }, { "epoch": 0.7884865568083261, "grad_norm": 3.6939713954925537, "learning_rate": 1.4796895401131583e-05, "loss": 0.02, "step": 7273 }, { "epoch": 0.7885949696444059, "grad_norm": 5.1267218589782715, "learning_rate": 1.4796170027564196e-05, "loss": 0.0528, "step": 7274 }, { "epoch": 0.7887033824804857, "grad_norm": 0.4020121097564697, "learning_rate": 1.479544465399681e-05, "loss": 0.0036, "step": 7275 }, { "epoch": 0.7888117953165655, "grad_norm": 2.5916800498962402, "learning_rate": 1.4794719280429422e-05, "loss": 0.0575, "step": 7276 }, { "epoch": 0.7889202081526453, "grad_norm": 4.5080790519714355, "learning_rate": 1.4793993906862034e-05, "loss": 0.1071, "step": 7277 }, { "epoch": 0.7890286209887251, "grad_norm": 1.0931849479675293, "learning_rate": 1.4793268533294648e-05, "loss": 0.0088, "step": 7278 }, { "epoch": 0.7891370338248048, "grad_norm": 0.22464992105960846, "learning_rate": 1.4792543159727261e-05, "loss": 0.0018, "step": 7279 }, { "epoch": 0.7892454466608847, "grad_norm": 1.7326712608337402, "learning_rate": 1.4791817786159874e-05, "loss": 0.0366, "step": 7280 }, { "epoch": 0.7893538594969645, "grad_norm": 2.0595273971557617, "learning_rate": 1.4791092412592487e-05, "loss": 0.0305, "step": 7281 }, { "epoch": 0.7894622723330442, "grad_norm": 0.6810624003410339, "learning_rate": 1.47903670390251e-05, "loss": 0.017, "step": 7282 }, { "epoch": 0.789570685169124, "grad_norm": 2.435361862182617, "learning_rate": 1.4789641665457712e-05, "loss": 0.0301, "step": 7283 }, { "epoch": 0.7896790980052039, "grad_norm": 4.44297981262207, "learning_rate": 1.4788916291890326e-05, "loss": 0.1038, "step": 7284 }, { "epoch": 0.7897875108412836, "grad_norm": 1.351958990097046, "learning_rate": 1.4788190918322938e-05, "loss": 0.0684, "step": 7285 }, { "epoch": 0.7898959236773634, "grad_norm": 0.4723299443721771, "learning_rate": 1.478746554475555e-05, "loss": 0.0183, "step": 7286 }, { "epoch": 0.7900043365134432, "grad_norm": 1.662345290184021, "learning_rate": 1.4786740171188164e-05, "loss": 0.0233, "step": 7287 }, { "epoch": 0.790112749349523, "grad_norm": 3.9861767292022705, "learning_rate": 1.4786014797620776e-05, "loss": 0.0353, "step": 7288 }, { "epoch": 0.7902211621856028, "grad_norm": 1.8141698837280273, "learning_rate": 1.4785289424053388e-05, "loss": 0.0401, "step": 7289 }, { "epoch": 0.7903295750216826, "grad_norm": 2.4570159912109375, "learning_rate": 1.4784564050486002e-05, "loss": 0.0482, "step": 7290 }, { "epoch": 0.7904379878577623, "grad_norm": 2.170297145843506, "learning_rate": 1.4783838676918614e-05, "loss": 0.0411, "step": 7291 }, { "epoch": 0.7905464006938422, "grad_norm": 5.775479316711426, "learning_rate": 1.4783113303351226e-05, "loss": 0.0628, "step": 7292 }, { "epoch": 0.790654813529922, "grad_norm": 2.0676660537719727, "learning_rate": 1.478238792978384e-05, "loss": 0.0376, "step": 7293 }, { "epoch": 0.7907632263660017, "grad_norm": 1.960037112236023, "learning_rate": 1.4781662556216452e-05, "loss": 0.0216, "step": 7294 }, { "epoch": 0.7908716392020815, "grad_norm": 3.7719810009002686, "learning_rate": 1.4780937182649064e-05, "loss": 0.0393, "step": 7295 }, { "epoch": 0.7909800520381614, "grad_norm": 4.375003337860107, "learning_rate": 1.478021180908168e-05, "loss": 0.0876, "step": 7296 }, { "epoch": 0.7910884648742411, "grad_norm": 2.1825807094573975, "learning_rate": 1.4779486435514292e-05, "loss": 0.0183, "step": 7297 }, { "epoch": 0.7911968777103209, "grad_norm": 1.2885634899139404, "learning_rate": 1.4778761061946904e-05, "loss": 0.0458, "step": 7298 }, { "epoch": 0.7913052905464006, "grad_norm": 1.8226561546325684, "learning_rate": 1.4778035688379518e-05, "loss": 0.0324, "step": 7299 }, { "epoch": 0.7914137033824805, "grad_norm": 2.1587486267089844, "learning_rate": 1.477731031481213e-05, "loss": 0.0294, "step": 7300 }, { "epoch": 0.7915221162185603, "grad_norm": 3.853130578994751, "learning_rate": 1.4776584941244742e-05, "loss": 0.0364, "step": 7301 }, { "epoch": 0.79163052905464, "grad_norm": 2.5205960273742676, "learning_rate": 1.4775859567677356e-05, "loss": 0.0328, "step": 7302 }, { "epoch": 0.7917389418907199, "grad_norm": 6.14669132232666, "learning_rate": 1.4775134194109968e-05, "loss": 0.0415, "step": 7303 }, { "epoch": 0.7918473547267997, "grad_norm": 0.3667096495628357, "learning_rate": 1.477440882054258e-05, "loss": 0.0039, "step": 7304 }, { "epoch": 0.7919557675628794, "grad_norm": 2.6567983627319336, "learning_rate": 1.4773683446975194e-05, "loss": 0.0816, "step": 7305 }, { "epoch": 0.7920641803989592, "grad_norm": 2.0711047649383545, "learning_rate": 1.4772958073407806e-05, "loss": 0.0151, "step": 7306 }, { "epoch": 0.7921725932350391, "grad_norm": 1.7621479034423828, "learning_rate": 1.4772232699840418e-05, "loss": 0.0445, "step": 7307 }, { "epoch": 0.7922810060711188, "grad_norm": 1.7140867710113525, "learning_rate": 1.477150732627303e-05, "loss": 0.0667, "step": 7308 }, { "epoch": 0.7923894189071986, "grad_norm": 1.3281440734863281, "learning_rate": 1.4770781952705644e-05, "loss": 0.0119, "step": 7309 }, { "epoch": 0.7924978317432784, "grad_norm": 1.2419344186782837, "learning_rate": 1.4770056579138257e-05, "loss": 0.0352, "step": 7310 }, { "epoch": 0.7926062445793582, "grad_norm": 2.6033730506896973, "learning_rate": 1.4769331205570869e-05, "loss": 0.0987, "step": 7311 }, { "epoch": 0.792714657415438, "grad_norm": 1.1451048851013184, "learning_rate": 1.4768605832003482e-05, "loss": 0.0333, "step": 7312 }, { "epoch": 0.7928230702515178, "grad_norm": 1.160057783126831, "learning_rate": 1.4767880458436096e-05, "loss": 0.0299, "step": 7313 }, { "epoch": 0.7929314830875975, "grad_norm": 2.8342790603637695, "learning_rate": 1.476715508486871e-05, "loss": 0.0285, "step": 7314 }, { "epoch": 0.7930398959236774, "grad_norm": 2.18890118598938, "learning_rate": 1.4766429711301322e-05, "loss": 0.0455, "step": 7315 }, { "epoch": 0.7931483087597572, "grad_norm": 1.839590072631836, "learning_rate": 1.4765704337733934e-05, "loss": 0.0222, "step": 7316 }, { "epoch": 0.7932567215958369, "grad_norm": 0.3184478282928467, "learning_rate": 1.4764978964166548e-05, "loss": 0.0045, "step": 7317 }, { "epoch": 0.7933651344319167, "grad_norm": 1.4989954233169556, "learning_rate": 1.476425359059916e-05, "loss": 0.0406, "step": 7318 }, { "epoch": 0.7934735472679966, "grad_norm": 4.894773960113525, "learning_rate": 1.4763528217031773e-05, "loss": 0.0267, "step": 7319 }, { "epoch": 0.7935819601040763, "grad_norm": 1.0554486513137817, "learning_rate": 1.4762802843464385e-05, "loss": 0.0216, "step": 7320 }, { "epoch": 0.7936903729401561, "grad_norm": 1.1170752048492432, "learning_rate": 1.4762077469896999e-05, "loss": 0.0396, "step": 7321 }, { "epoch": 0.7937987857762359, "grad_norm": 2.104360818862915, "learning_rate": 1.476135209632961e-05, "loss": 0.0241, "step": 7322 }, { "epoch": 0.7939071986123157, "grad_norm": 0.3584396243095398, "learning_rate": 1.4760626722762223e-05, "loss": 0.0036, "step": 7323 }, { "epoch": 0.7940156114483955, "grad_norm": 2.6033425331115723, "learning_rate": 1.4759901349194837e-05, "loss": 0.041, "step": 7324 }, { "epoch": 0.7941240242844753, "grad_norm": 4.326061248779297, "learning_rate": 1.4759175975627449e-05, "loss": 0.0745, "step": 7325 }, { "epoch": 0.794232437120555, "grad_norm": 1.9267204999923706, "learning_rate": 1.4758450602060061e-05, "loss": 0.0494, "step": 7326 }, { "epoch": 0.7943408499566349, "grad_norm": 1.8768693208694458, "learning_rate": 1.4757725228492675e-05, "loss": 0.0464, "step": 7327 }, { "epoch": 0.7944492627927147, "grad_norm": 1.9062132835388184, "learning_rate": 1.4756999854925287e-05, "loss": 0.0223, "step": 7328 }, { "epoch": 0.7945576756287944, "grad_norm": 2.1545515060424805, "learning_rate": 1.4756274481357899e-05, "loss": 0.0547, "step": 7329 }, { "epoch": 0.7946660884648742, "grad_norm": 5.078322410583496, "learning_rate": 1.4755549107790515e-05, "loss": 0.0228, "step": 7330 }, { "epoch": 0.7947745013009541, "grad_norm": 5.040393352508545, "learning_rate": 1.4754823734223127e-05, "loss": 0.0275, "step": 7331 }, { "epoch": 0.7948829141370338, "grad_norm": 15.207469940185547, "learning_rate": 1.4754098360655739e-05, "loss": 0.0437, "step": 7332 }, { "epoch": 0.7949913269731136, "grad_norm": 3.090289354324341, "learning_rate": 1.4753372987088353e-05, "loss": 0.0269, "step": 7333 }, { "epoch": 0.7950997398091935, "grad_norm": 1.7588971853256226, "learning_rate": 1.4752647613520965e-05, "loss": 0.0244, "step": 7334 }, { "epoch": 0.7952081526452732, "grad_norm": 4.264991283416748, "learning_rate": 1.4751922239953577e-05, "loss": 0.1379, "step": 7335 }, { "epoch": 0.795316565481353, "grad_norm": 2.8773117065429688, "learning_rate": 1.4751196866386191e-05, "loss": 0.0481, "step": 7336 }, { "epoch": 0.7954249783174328, "grad_norm": 2.39567232131958, "learning_rate": 1.4750471492818803e-05, "loss": 0.0308, "step": 7337 }, { "epoch": 0.7955333911535126, "grad_norm": 1.7640223503112793, "learning_rate": 1.4749746119251415e-05, "loss": 0.0261, "step": 7338 }, { "epoch": 0.7956418039895924, "grad_norm": 2.1090385913848877, "learning_rate": 1.4749020745684029e-05, "loss": 0.0329, "step": 7339 }, { "epoch": 0.7957502168256722, "grad_norm": 1.9654251337051392, "learning_rate": 1.4748295372116641e-05, "loss": 0.0232, "step": 7340 }, { "epoch": 0.7958586296617519, "grad_norm": 0.8682138919830322, "learning_rate": 1.4747569998549253e-05, "loss": 0.0166, "step": 7341 }, { "epoch": 0.7959670424978318, "grad_norm": 3.0957934856414795, "learning_rate": 1.4746844624981867e-05, "loss": 0.0107, "step": 7342 }, { "epoch": 0.7960754553339116, "grad_norm": 1.5700396299362183, "learning_rate": 1.474611925141448e-05, "loss": 0.039, "step": 7343 }, { "epoch": 0.7961838681699913, "grad_norm": 2.7093863487243652, "learning_rate": 1.4745393877847091e-05, "loss": 0.0392, "step": 7344 }, { "epoch": 0.7962922810060711, "grad_norm": 0.8072781562805176, "learning_rate": 1.4744668504279705e-05, "loss": 0.0174, "step": 7345 }, { "epoch": 0.796400693842151, "grad_norm": 1.3870484828948975, "learning_rate": 1.4743943130712317e-05, "loss": 0.0179, "step": 7346 }, { "epoch": 0.7965091066782307, "grad_norm": 6.919134140014648, "learning_rate": 1.474321775714493e-05, "loss": 0.0447, "step": 7347 }, { "epoch": 0.7966175195143105, "grad_norm": 6.637925148010254, "learning_rate": 1.4742492383577545e-05, "loss": 0.0818, "step": 7348 }, { "epoch": 0.7967259323503902, "grad_norm": 0.1800239533185959, "learning_rate": 1.4741767010010157e-05, "loss": 0.0017, "step": 7349 }, { "epoch": 0.7968343451864701, "grad_norm": 6.287868022918701, "learning_rate": 1.474104163644277e-05, "loss": 0.0425, "step": 7350 }, { "epoch": 0.7969427580225499, "grad_norm": 2.099630832672119, "learning_rate": 1.4740316262875383e-05, "loss": 0.0438, "step": 7351 }, { "epoch": 0.7970511708586296, "grad_norm": 4.429477691650391, "learning_rate": 1.4739590889307995e-05, "loss": 0.0528, "step": 7352 }, { "epoch": 0.7971595836947094, "grad_norm": 0.8433004021644592, "learning_rate": 1.4738865515740607e-05, "loss": 0.0139, "step": 7353 }, { "epoch": 0.7972679965307893, "grad_norm": 1.681042194366455, "learning_rate": 1.4738140142173221e-05, "loss": 0.0381, "step": 7354 }, { "epoch": 0.797376409366869, "grad_norm": 3.639204740524292, "learning_rate": 1.4737414768605833e-05, "loss": 0.0653, "step": 7355 }, { "epoch": 0.7974848222029488, "grad_norm": 1.1780247688293457, "learning_rate": 1.4736689395038446e-05, "loss": 0.0259, "step": 7356 }, { "epoch": 0.7975932350390286, "grad_norm": 5.252540111541748, "learning_rate": 1.473596402147106e-05, "loss": 0.0937, "step": 7357 }, { "epoch": 0.7977016478751084, "grad_norm": 3.832502841949463, "learning_rate": 1.4735238647903672e-05, "loss": 0.0511, "step": 7358 }, { "epoch": 0.7978100607111882, "grad_norm": 5.249119281768799, "learning_rate": 1.4734513274336284e-05, "loss": 0.0588, "step": 7359 }, { "epoch": 0.797918473547268, "grad_norm": 4.970529079437256, "learning_rate": 1.4733787900768896e-05, "loss": 0.061, "step": 7360 }, { "epoch": 0.7980268863833477, "grad_norm": 2.429973840713501, "learning_rate": 1.473306252720151e-05, "loss": 0.0379, "step": 7361 }, { "epoch": 0.7981352992194276, "grad_norm": 3.897284507751465, "learning_rate": 1.4732337153634122e-05, "loss": 0.0466, "step": 7362 }, { "epoch": 0.7982437120555074, "grad_norm": 1.905495285987854, "learning_rate": 1.4731611780066734e-05, "loss": 0.0293, "step": 7363 }, { "epoch": 0.7983521248915871, "grad_norm": 0.8275286555290222, "learning_rate": 1.4730886406499348e-05, "loss": 0.0078, "step": 7364 }, { "epoch": 0.798460537727667, "grad_norm": 7.959173679351807, "learning_rate": 1.4730161032931962e-05, "loss": 0.0588, "step": 7365 }, { "epoch": 0.7985689505637468, "grad_norm": 3.271023988723755, "learning_rate": 1.4729435659364576e-05, "loss": 0.0373, "step": 7366 }, { "epoch": 0.7986773633998265, "grad_norm": 3.321749687194824, "learning_rate": 1.4728710285797188e-05, "loss": 0.0405, "step": 7367 }, { "epoch": 0.7987857762359063, "grad_norm": 4.687636852264404, "learning_rate": 1.47279849122298e-05, "loss": 0.0222, "step": 7368 }, { "epoch": 0.7988941890719862, "grad_norm": 1.8940064907073975, "learning_rate": 1.4727259538662414e-05, "loss": 0.049, "step": 7369 }, { "epoch": 0.7990026019080659, "grad_norm": 2.2162039279937744, "learning_rate": 1.4726534165095026e-05, "loss": 0.038, "step": 7370 }, { "epoch": 0.7991110147441457, "grad_norm": 1.619350790977478, "learning_rate": 1.4725808791527638e-05, "loss": 0.0332, "step": 7371 }, { "epoch": 0.7992194275802255, "grad_norm": 1.7683041095733643, "learning_rate": 1.472508341796025e-05, "loss": 0.0279, "step": 7372 }, { "epoch": 0.7993278404163053, "grad_norm": 2.282498836517334, "learning_rate": 1.4724358044392864e-05, "loss": 0.0317, "step": 7373 }, { "epoch": 0.7994362532523851, "grad_norm": 3.4293270111083984, "learning_rate": 1.4723632670825476e-05, "loss": 0.0721, "step": 7374 }, { "epoch": 0.7995446660884649, "grad_norm": 1.967128038406372, "learning_rate": 1.4722907297258088e-05, "loss": 0.0509, "step": 7375 }, { "epoch": 0.7996530789245446, "grad_norm": 2.0264368057250977, "learning_rate": 1.4722181923690702e-05, "loss": 0.0347, "step": 7376 }, { "epoch": 0.7997614917606245, "grad_norm": 2.1490232944488525, "learning_rate": 1.4721456550123314e-05, "loss": 0.0324, "step": 7377 }, { "epoch": 0.7998699045967043, "grad_norm": 3.878885507583618, "learning_rate": 1.4720731176555926e-05, "loss": 0.0336, "step": 7378 }, { "epoch": 0.799978317432784, "grad_norm": 3.6943325996398926, "learning_rate": 1.472000580298854e-05, "loss": 0.0306, "step": 7379 }, { "epoch": 0.8000867302688638, "grad_norm": 2.6406049728393555, "learning_rate": 1.4719280429421152e-05, "loss": 0.0523, "step": 7380 }, { "epoch": 0.8001951431049437, "grad_norm": 3.602928638458252, "learning_rate": 1.4718555055853764e-05, "loss": 0.0581, "step": 7381 }, { "epoch": 0.8003035559410234, "grad_norm": 0.2792077660560608, "learning_rate": 1.471782968228638e-05, "loss": 0.0043, "step": 7382 }, { "epoch": 0.8004119687771032, "grad_norm": 3.3032140731811523, "learning_rate": 1.4717104308718992e-05, "loss": 0.0296, "step": 7383 }, { "epoch": 0.800520381613183, "grad_norm": 0.8306502103805542, "learning_rate": 1.4716378935151604e-05, "loss": 0.0102, "step": 7384 }, { "epoch": 0.8006287944492628, "grad_norm": 2.5124263763427734, "learning_rate": 1.4715653561584218e-05, "loss": 0.0594, "step": 7385 }, { "epoch": 0.8007372072853426, "grad_norm": 2.9732766151428223, "learning_rate": 1.471492818801683e-05, "loss": 0.0243, "step": 7386 }, { "epoch": 0.8008456201214224, "grad_norm": 2.269350051879883, "learning_rate": 1.4714202814449442e-05, "loss": 0.0125, "step": 7387 }, { "epoch": 0.8009540329575021, "grad_norm": 2.76354718208313, "learning_rate": 1.4713477440882056e-05, "loss": 0.0298, "step": 7388 }, { "epoch": 0.801062445793582, "grad_norm": 2.611882209777832, "learning_rate": 1.4712752067314668e-05, "loss": 0.0491, "step": 7389 }, { "epoch": 0.8011708586296618, "grad_norm": 2.0278708934783936, "learning_rate": 1.471202669374728e-05, "loss": 0.0332, "step": 7390 }, { "epoch": 0.8012792714657415, "grad_norm": 2.66349720954895, "learning_rate": 1.4711301320179894e-05, "loss": 0.0225, "step": 7391 }, { "epoch": 0.8013876843018214, "grad_norm": 1.8120660781860352, "learning_rate": 1.4710575946612506e-05, "loss": 0.0267, "step": 7392 }, { "epoch": 0.8014960971379012, "grad_norm": 11.864376068115234, "learning_rate": 1.4709850573045119e-05, "loss": 0.0602, "step": 7393 }, { "epoch": 0.8016045099739809, "grad_norm": 1.5633525848388672, "learning_rate": 1.4709125199477732e-05, "loss": 0.0356, "step": 7394 }, { "epoch": 0.8017129228100607, "grad_norm": 3.245147228240967, "learning_rate": 1.4708399825910345e-05, "loss": 0.0108, "step": 7395 }, { "epoch": 0.8018213356461406, "grad_norm": 1.4299777746200562, "learning_rate": 1.4707674452342957e-05, "loss": 0.0182, "step": 7396 }, { "epoch": 0.8019297484822203, "grad_norm": 3.8055734634399414, "learning_rate": 1.470694907877557e-05, "loss": 0.03, "step": 7397 }, { "epoch": 0.8020381613183001, "grad_norm": 3.2276012897491455, "learning_rate": 1.4706223705208183e-05, "loss": 0.012, "step": 7398 }, { "epoch": 0.8021465741543798, "grad_norm": 2.4191675186157227, "learning_rate": 1.4705498331640797e-05, "loss": 0.0425, "step": 7399 }, { "epoch": 0.8022549869904597, "grad_norm": 1.4638891220092773, "learning_rate": 1.470477295807341e-05, "loss": 0.0272, "step": 7400 }, { "epoch": 0.8023633998265395, "grad_norm": 3.527648687362671, "learning_rate": 1.4704047584506023e-05, "loss": 0.068, "step": 7401 }, { "epoch": 0.8024718126626192, "grad_norm": 1.255759358406067, "learning_rate": 1.4703322210938635e-05, "loss": 0.0303, "step": 7402 }, { "epoch": 0.802580225498699, "grad_norm": 2.479172468185425, "learning_rate": 1.4702596837371249e-05, "loss": 0.0348, "step": 7403 }, { "epoch": 0.8026886383347789, "grad_norm": 10.22083568572998, "learning_rate": 1.470187146380386e-05, "loss": 0.0542, "step": 7404 }, { "epoch": 0.8027970511708586, "grad_norm": 2.572855234146118, "learning_rate": 1.4701146090236473e-05, "loss": 0.0204, "step": 7405 }, { "epoch": 0.8029054640069384, "grad_norm": 3.4424386024475098, "learning_rate": 1.4700420716669087e-05, "loss": 0.0177, "step": 7406 }, { "epoch": 0.8030138768430182, "grad_norm": 3.353353500366211, "learning_rate": 1.4699695343101699e-05, "loss": 0.0523, "step": 7407 }, { "epoch": 0.803122289679098, "grad_norm": 1.2120585441589355, "learning_rate": 1.4698969969534311e-05, "loss": 0.0323, "step": 7408 }, { "epoch": 0.8032307025151778, "grad_norm": 2.4332070350646973, "learning_rate": 1.4698244595966925e-05, "loss": 0.0313, "step": 7409 }, { "epoch": 0.8033391153512576, "grad_norm": 5.601504802703857, "learning_rate": 1.4697519222399537e-05, "loss": 0.061, "step": 7410 }, { "epoch": 0.8034475281873373, "grad_norm": 1.8834869861602783, "learning_rate": 1.4696793848832149e-05, "loss": 0.0295, "step": 7411 }, { "epoch": 0.8035559410234172, "grad_norm": 0.6017666459083557, "learning_rate": 1.4696068475264761e-05, "loss": 0.0206, "step": 7412 }, { "epoch": 0.803664353859497, "grad_norm": 3.5189716815948486, "learning_rate": 1.4695343101697375e-05, "loss": 0.0373, "step": 7413 }, { "epoch": 0.8037727666955767, "grad_norm": 3.1705269813537598, "learning_rate": 1.4694617728129987e-05, "loss": 0.0285, "step": 7414 }, { "epoch": 0.8038811795316565, "grad_norm": 7.08400821685791, "learning_rate": 1.46938923545626e-05, "loss": 0.0676, "step": 7415 }, { "epoch": 0.8039895923677364, "grad_norm": 12.930665969848633, "learning_rate": 1.4693166980995213e-05, "loss": 0.0396, "step": 7416 }, { "epoch": 0.8040980052038161, "grad_norm": 0.9345747828483582, "learning_rate": 1.4692441607427827e-05, "loss": 0.0113, "step": 7417 }, { "epoch": 0.8042064180398959, "grad_norm": 3.185244560241699, "learning_rate": 1.469171623386044e-05, "loss": 0.0683, "step": 7418 }, { "epoch": 0.8043148308759757, "grad_norm": 4.768292427062988, "learning_rate": 1.4690990860293053e-05, "loss": 0.0278, "step": 7419 }, { "epoch": 0.8044232437120555, "grad_norm": 0.7953506708145142, "learning_rate": 1.4690265486725665e-05, "loss": 0.0187, "step": 7420 }, { "epoch": 0.8045316565481353, "grad_norm": 2.3461861610412598, "learning_rate": 1.4689540113158279e-05, "loss": 0.0488, "step": 7421 }, { "epoch": 0.8046400693842151, "grad_norm": 2.1840100288391113, "learning_rate": 1.4688814739590891e-05, "loss": 0.0127, "step": 7422 }, { "epoch": 0.8047484822202949, "grad_norm": 2.3670313358306885, "learning_rate": 1.4688089366023503e-05, "loss": 0.0395, "step": 7423 }, { "epoch": 0.8048568950563747, "grad_norm": 0.9271336793899536, "learning_rate": 1.4687363992456115e-05, "loss": 0.0185, "step": 7424 }, { "epoch": 0.8049653078924545, "grad_norm": 1.2907969951629639, "learning_rate": 1.468663861888873e-05, "loss": 0.0242, "step": 7425 }, { "epoch": 0.8050737207285342, "grad_norm": 2.297893524169922, "learning_rate": 1.4685913245321341e-05, "loss": 0.0398, "step": 7426 }, { "epoch": 0.8051821335646141, "grad_norm": 3.8397765159606934, "learning_rate": 1.4685187871753954e-05, "loss": 0.0737, "step": 7427 }, { "epoch": 0.8052905464006939, "grad_norm": 7.428755760192871, "learning_rate": 1.4684462498186567e-05, "loss": 0.017, "step": 7428 }, { "epoch": 0.8053989592367736, "grad_norm": 0.29806652665138245, "learning_rate": 1.468373712461918e-05, "loss": 0.0031, "step": 7429 }, { "epoch": 0.8055073720728534, "grad_norm": 3.264422655105591, "learning_rate": 1.4683011751051792e-05, "loss": 0.0295, "step": 7430 }, { "epoch": 0.8056157849089333, "grad_norm": 1.3917715549468994, "learning_rate": 1.4682286377484405e-05, "loss": 0.0466, "step": 7431 }, { "epoch": 0.805724197745013, "grad_norm": 1.4791491031646729, "learning_rate": 1.4681561003917018e-05, "loss": 0.026, "step": 7432 }, { "epoch": 0.8058326105810928, "grad_norm": 3.351896286010742, "learning_rate": 1.468083563034963e-05, "loss": 0.0568, "step": 7433 }, { "epoch": 0.8059410234171726, "grad_norm": 2.01967716217041, "learning_rate": 1.4680110256782245e-05, "loss": 0.0087, "step": 7434 }, { "epoch": 0.8060494362532524, "grad_norm": 0.8134462237358093, "learning_rate": 1.4679384883214857e-05, "loss": 0.0165, "step": 7435 }, { "epoch": 0.8061578490893322, "grad_norm": 7.7808709144592285, "learning_rate": 1.4678659509647471e-05, "loss": 0.0898, "step": 7436 }, { "epoch": 0.806266261925412, "grad_norm": 2.7345995903015137, "learning_rate": 1.4677934136080083e-05, "loss": 0.0937, "step": 7437 }, { "epoch": 0.8063746747614917, "grad_norm": 3.4368996620178223, "learning_rate": 1.4677208762512696e-05, "loss": 0.0305, "step": 7438 }, { "epoch": 0.8064830875975716, "grad_norm": 2.2323906421661377, "learning_rate": 1.4676483388945308e-05, "loss": 0.0228, "step": 7439 }, { "epoch": 0.8065915004336514, "grad_norm": 2.7470169067382812, "learning_rate": 1.4675758015377922e-05, "loss": 0.0453, "step": 7440 }, { "epoch": 0.8066999132697311, "grad_norm": 1.5691744089126587, "learning_rate": 1.4675032641810534e-05, "loss": 0.0182, "step": 7441 }, { "epoch": 0.8068083261058109, "grad_norm": 7.604193687438965, "learning_rate": 1.4674307268243146e-05, "loss": 0.09, "step": 7442 }, { "epoch": 0.8069167389418908, "grad_norm": 2.922368049621582, "learning_rate": 1.467358189467576e-05, "loss": 0.0618, "step": 7443 }, { "epoch": 0.8070251517779705, "grad_norm": 2.9364113807678223, "learning_rate": 1.4672856521108372e-05, "loss": 0.0874, "step": 7444 }, { "epoch": 0.8071335646140503, "grad_norm": 0.8219984769821167, "learning_rate": 1.4672131147540984e-05, "loss": 0.0156, "step": 7445 }, { "epoch": 0.80724197745013, "grad_norm": 3.613880157470703, "learning_rate": 1.4671405773973598e-05, "loss": 0.0348, "step": 7446 }, { "epoch": 0.8073503902862099, "grad_norm": 0.8764395713806152, "learning_rate": 1.467068040040621e-05, "loss": 0.0225, "step": 7447 }, { "epoch": 0.8074588031222897, "grad_norm": 0.48517322540283203, "learning_rate": 1.4669955026838822e-05, "loss": 0.0046, "step": 7448 }, { "epoch": 0.8075672159583694, "grad_norm": 2.6518499851226807, "learning_rate": 1.4669229653271436e-05, "loss": 0.055, "step": 7449 }, { "epoch": 0.8076756287944492, "grad_norm": 3.055706739425659, "learning_rate": 1.4668504279704048e-05, "loss": 0.04, "step": 7450 }, { "epoch": 0.8077840416305291, "grad_norm": 1.6582415103912354, "learning_rate": 1.4667778906136662e-05, "loss": 0.0444, "step": 7451 }, { "epoch": 0.8078924544666088, "grad_norm": 2.777707815170288, "learning_rate": 1.4667053532569276e-05, "loss": 0.0245, "step": 7452 }, { "epoch": 0.8080008673026886, "grad_norm": 2.584343194961548, "learning_rate": 1.4666328159001888e-05, "loss": 0.0447, "step": 7453 }, { "epoch": 0.8081092801387685, "grad_norm": 7.342785835266113, "learning_rate": 1.46656027854345e-05, "loss": 0.0611, "step": 7454 }, { "epoch": 0.8082176929748482, "grad_norm": 2.427593946456909, "learning_rate": 1.4664877411867114e-05, "loss": 0.0126, "step": 7455 }, { "epoch": 0.808326105810928, "grad_norm": 1.7487285137176514, "learning_rate": 1.4664152038299726e-05, "loss": 0.0201, "step": 7456 }, { "epoch": 0.8084345186470078, "grad_norm": 2.0955240726470947, "learning_rate": 1.4663426664732338e-05, "loss": 0.0256, "step": 7457 }, { "epoch": 0.8085429314830876, "grad_norm": 1.9847807884216309, "learning_rate": 1.4662701291164952e-05, "loss": 0.037, "step": 7458 }, { "epoch": 0.8086513443191674, "grad_norm": 3.6514501571655273, "learning_rate": 1.4661975917597564e-05, "loss": 0.0566, "step": 7459 }, { "epoch": 0.8087597571552472, "grad_norm": 1.485471248626709, "learning_rate": 1.4661250544030176e-05, "loss": 0.0324, "step": 7460 }, { "epoch": 0.8088681699913269, "grad_norm": 1.4724831581115723, "learning_rate": 1.466052517046279e-05, "loss": 0.0189, "step": 7461 }, { "epoch": 0.8089765828274068, "grad_norm": 1.5253909826278687, "learning_rate": 1.4659799796895402e-05, "loss": 0.0285, "step": 7462 }, { "epoch": 0.8090849956634866, "grad_norm": 2.019132614135742, "learning_rate": 1.4659074423328014e-05, "loss": 0.0214, "step": 7463 }, { "epoch": 0.8091934084995663, "grad_norm": 2.6450934410095215, "learning_rate": 1.4658349049760627e-05, "loss": 0.0388, "step": 7464 }, { "epoch": 0.8093018213356461, "grad_norm": 2.8454861640930176, "learning_rate": 1.465762367619324e-05, "loss": 0.0254, "step": 7465 }, { "epoch": 0.809410234171726, "grad_norm": 3.7269322872161865, "learning_rate": 1.4656898302625853e-05, "loss": 0.0338, "step": 7466 }, { "epoch": 0.8095186470078057, "grad_norm": 4.0216264724731445, "learning_rate": 1.4656172929058465e-05, "loss": 0.0373, "step": 7467 }, { "epoch": 0.8096270598438855, "grad_norm": 2.731898069381714, "learning_rate": 1.4655447555491079e-05, "loss": 0.0268, "step": 7468 }, { "epoch": 0.8097354726799653, "grad_norm": 2.9385900497436523, "learning_rate": 1.4654722181923692e-05, "loss": 0.0301, "step": 7469 }, { "epoch": 0.8098438855160451, "grad_norm": 0.8618904948234558, "learning_rate": 1.4653996808356306e-05, "loss": 0.0126, "step": 7470 }, { "epoch": 0.8099522983521249, "grad_norm": 2.3427276611328125, "learning_rate": 1.4653271434788918e-05, "loss": 0.0399, "step": 7471 }, { "epoch": 0.8100607111882047, "grad_norm": 9.708152770996094, "learning_rate": 1.465254606122153e-05, "loss": 0.0412, "step": 7472 }, { "epoch": 0.8101691240242844, "grad_norm": 2.625464916229248, "learning_rate": 1.4651820687654144e-05, "loss": 0.0489, "step": 7473 }, { "epoch": 0.8102775368603643, "grad_norm": 2.906892776489258, "learning_rate": 1.4651095314086756e-05, "loss": 0.0525, "step": 7474 }, { "epoch": 0.8103859496964441, "grad_norm": 2.0294859409332275, "learning_rate": 1.4650369940519369e-05, "loss": 0.0422, "step": 7475 }, { "epoch": 0.8104943625325238, "grad_norm": 5.014775276184082, "learning_rate": 1.464964456695198e-05, "loss": 0.091, "step": 7476 }, { "epoch": 0.8106027753686036, "grad_norm": 1.19895601272583, "learning_rate": 1.4648919193384595e-05, "loss": 0.0186, "step": 7477 }, { "epoch": 0.8107111882046835, "grad_norm": 2.693784236907959, "learning_rate": 1.4648193819817207e-05, "loss": 0.0548, "step": 7478 }, { "epoch": 0.8108196010407632, "grad_norm": 4.661581516265869, "learning_rate": 1.4647468446249819e-05, "loss": 0.0781, "step": 7479 }, { "epoch": 0.810928013876843, "grad_norm": 3.661285877227783, "learning_rate": 1.4646743072682433e-05, "loss": 0.0503, "step": 7480 }, { "epoch": 0.8110364267129229, "grad_norm": 1.3561416864395142, "learning_rate": 1.4646017699115045e-05, "loss": 0.0271, "step": 7481 }, { "epoch": 0.8111448395490026, "grad_norm": 2.124124765396118, "learning_rate": 1.4645292325547657e-05, "loss": 0.0488, "step": 7482 }, { "epoch": 0.8112532523850824, "grad_norm": 2.91467547416687, "learning_rate": 1.464456695198027e-05, "loss": 0.0463, "step": 7483 }, { "epoch": 0.8113616652211622, "grad_norm": 4.1637139320373535, "learning_rate": 1.4643841578412883e-05, "loss": 0.0229, "step": 7484 }, { "epoch": 0.811470078057242, "grad_norm": 4.6619439125061035, "learning_rate": 1.4643116204845495e-05, "loss": 0.0212, "step": 7485 }, { "epoch": 0.8115784908933218, "grad_norm": 4.101515769958496, "learning_rate": 1.464239083127811e-05, "loss": 0.0599, "step": 7486 }, { "epoch": 0.8116869037294016, "grad_norm": 1.07630455493927, "learning_rate": 1.4641665457710723e-05, "loss": 0.0191, "step": 7487 }, { "epoch": 0.8117953165654813, "grad_norm": 2.3645431995391846, "learning_rate": 1.4640940084143337e-05, "loss": 0.054, "step": 7488 }, { "epoch": 0.8119037294015612, "grad_norm": 0.5776128172874451, "learning_rate": 1.4640214710575949e-05, "loss": 0.0056, "step": 7489 }, { "epoch": 0.812012142237641, "grad_norm": 1.0631694793701172, "learning_rate": 1.4639489337008561e-05, "loss": 0.0371, "step": 7490 }, { "epoch": 0.8121205550737207, "grad_norm": 2.5479846000671387, "learning_rate": 1.4638763963441173e-05, "loss": 0.0334, "step": 7491 }, { "epoch": 0.8122289679098005, "grad_norm": 4.40256929397583, "learning_rate": 1.4638038589873787e-05, "loss": 0.072, "step": 7492 }, { "epoch": 0.8123373807458804, "grad_norm": 1.1031484603881836, "learning_rate": 1.4637313216306399e-05, "loss": 0.0212, "step": 7493 }, { "epoch": 0.8124457935819601, "grad_norm": 2.2209396362304688, "learning_rate": 1.4636587842739011e-05, "loss": 0.032, "step": 7494 }, { "epoch": 0.8125542064180399, "grad_norm": 1.438402771949768, "learning_rate": 1.4635862469171625e-05, "loss": 0.0116, "step": 7495 }, { "epoch": 0.8126626192541196, "grad_norm": 1.1008656024932861, "learning_rate": 1.4635137095604237e-05, "loss": 0.0648, "step": 7496 }, { "epoch": 0.8127710320901995, "grad_norm": 2.5051755905151367, "learning_rate": 1.463441172203685e-05, "loss": 0.0248, "step": 7497 }, { "epoch": 0.8128794449262793, "grad_norm": 2.1394460201263428, "learning_rate": 1.4633686348469463e-05, "loss": 0.0606, "step": 7498 }, { "epoch": 0.812987857762359, "grad_norm": 3.6333835124969482, "learning_rate": 1.4632960974902075e-05, "loss": 0.0355, "step": 7499 }, { "epoch": 0.8130962705984388, "grad_norm": 3.029951572418213, "learning_rate": 1.4632235601334687e-05, "loss": 0.0821, "step": 7500 }, { "epoch": 0.8132046834345187, "grad_norm": 1.923235297203064, "learning_rate": 1.4631510227767301e-05, "loss": 0.0415, "step": 7501 }, { "epoch": 0.8133130962705984, "grad_norm": 2.850855827331543, "learning_rate": 1.4630784854199913e-05, "loss": 0.0473, "step": 7502 }, { "epoch": 0.8134215091066782, "grad_norm": 1.2554510831832886, "learning_rate": 1.4630059480632527e-05, "loss": 0.0367, "step": 7503 }, { "epoch": 0.813529921942758, "grad_norm": 9.474595069885254, "learning_rate": 1.4629334107065141e-05, "loss": 0.0381, "step": 7504 }, { "epoch": 0.8136383347788378, "grad_norm": 3.45708966255188, "learning_rate": 1.4628608733497753e-05, "loss": 0.0329, "step": 7505 }, { "epoch": 0.8137467476149176, "grad_norm": 1.6276206970214844, "learning_rate": 1.4627883359930365e-05, "loss": 0.0209, "step": 7506 }, { "epoch": 0.8138551604509974, "grad_norm": 1.211529016494751, "learning_rate": 1.462715798636298e-05, "loss": 0.0103, "step": 7507 }, { "epoch": 0.8139635732870771, "grad_norm": 2.4703335762023926, "learning_rate": 1.4626432612795591e-05, "loss": 0.0547, "step": 7508 }, { "epoch": 0.814071986123157, "grad_norm": 2.8776817321777344, "learning_rate": 1.4625707239228203e-05, "loss": 0.0793, "step": 7509 }, { "epoch": 0.8141803989592368, "grad_norm": 4.636091232299805, "learning_rate": 1.4624981865660817e-05, "loss": 0.0661, "step": 7510 }, { "epoch": 0.8142888117953165, "grad_norm": 5.2196736335754395, "learning_rate": 1.462425649209343e-05, "loss": 0.039, "step": 7511 }, { "epoch": 0.8143972246313964, "grad_norm": 4.631572246551514, "learning_rate": 1.4623531118526042e-05, "loss": 0.0206, "step": 7512 }, { "epoch": 0.8145056374674762, "grad_norm": 2.437126874923706, "learning_rate": 1.4622805744958655e-05, "loss": 0.0208, "step": 7513 }, { "epoch": 0.8146140503035559, "grad_norm": 0.9391622543334961, "learning_rate": 1.4622080371391268e-05, "loss": 0.0247, "step": 7514 }, { "epoch": 0.8147224631396357, "grad_norm": 9.681453704833984, "learning_rate": 1.462135499782388e-05, "loss": 0.0387, "step": 7515 }, { "epoch": 0.8148308759757156, "grad_norm": 1.3284776210784912, "learning_rate": 1.4620629624256492e-05, "loss": 0.0319, "step": 7516 }, { "epoch": 0.8149392888117953, "grad_norm": 1.129873514175415, "learning_rate": 1.4619904250689106e-05, "loss": 0.031, "step": 7517 }, { "epoch": 0.8150477016478751, "grad_norm": 1.7665523290634155, "learning_rate": 1.4619178877121718e-05, "loss": 0.0605, "step": 7518 }, { "epoch": 0.8151561144839549, "grad_norm": 2.969928503036499, "learning_rate": 1.461845350355433e-05, "loss": 0.0153, "step": 7519 }, { "epoch": 0.8152645273200347, "grad_norm": 1.7305662631988525, "learning_rate": 1.4617728129986946e-05, "loss": 0.0136, "step": 7520 }, { "epoch": 0.8153729401561145, "grad_norm": 2.863581657409668, "learning_rate": 1.4617002756419558e-05, "loss": 0.0218, "step": 7521 }, { "epoch": 0.8154813529921943, "grad_norm": 0.9962496757507324, "learning_rate": 1.4616277382852172e-05, "loss": 0.0212, "step": 7522 }, { "epoch": 0.815589765828274, "grad_norm": 22.63390350341797, "learning_rate": 1.4615552009284784e-05, "loss": 0.0665, "step": 7523 }, { "epoch": 0.8156981786643539, "grad_norm": 2.634053945541382, "learning_rate": 1.4614826635717396e-05, "loss": 0.0484, "step": 7524 }, { "epoch": 0.8158065915004337, "grad_norm": 4.402463436126709, "learning_rate": 1.461410126215001e-05, "loss": 0.065, "step": 7525 }, { "epoch": 0.8159150043365134, "grad_norm": 0.5702108144760132, "learning_rate": 1.4613375888582622e-05, "loss": 0.0058, "step": 7526 }, { "epoch": 0.8160234171725932, "grad_norm": 1.3619738817214966, "learning_rate": 1.4612650515015234e-05, "loss": 0.0117, "step": 7527 }, { "epoch": 0.8161318300086731, "grad_norm": 5.881645679473877, "learning_rate": 1.4611925141447846e-05, "loss": 0.0615, "step": 7528 }, { "epoch": 0.8162402428447528, "grad_norm": 2.977437734603882, "learning_rate": 1.461119976788046e-05, "loss": 0.0419, "step": 7529 }, { "epoch": 0.8163486556808326, "grad_norm": 2.366708993911743, "learning_rate": 1.4610474394313072e-05, "loss": 0.0349, "step": 7530 }, { "epoch": 0.8164570685169124, "grad_norm": 3.25227952003479, "learning_rate": 1.4609749020745684e-05, "loss": 0.0457, "step": 7531 }, { "epoch": 0.8165654813529922, "grad_norm": 2.8433573246002197, "learning_rate": 1.4609023647178298e-05, "loss": 0.0329, "step": 7532 }, { "epoch": 0.816673894189072, "grad_norm": 0.6061180830001831, "learning_rate": 1.460829827361091e-05, "loss": 0.0069, "step": 7533 }, { "epoch": 0.8167823070251518, "grad_norm": 3.727801561355591, "learning_rate": 1.4607572900043522e-05, "loss": 0.0797, "step": 7534 }, { "epoch": 0.8168907198612315, "grad_norm": 0.7123563885688782, "learning_rate": 1.4606847526476136e-05, "loss": 0.0202, "step": 7535 }, { "epoch": 0.8169991326973114, "grad_norm": 5.94475793838501, "learning_rate": 1.4606122152908748e-05, "loss": 0.1142, "step": 7536 }, { "epoch": 0.8171075455333912, "grad_norm": 3.1790812015533447, "learning_rate": 1.460539677934136e-05, "loss": 0.0548, "step": 7537 }, { "epoch": 0.8172159583694709, "grad_norm": 2.0606775283813477, "learning_rate": 1.4604671405773976e-05, "loss": 0.0281, "step": 7538 }, { "epoch": 0.8173243712055508, "grad_norm": 1.720301866531372, "learning_rate": 1.4603946032206588e-05, "loss": 0.0292, "step": 7539 }, { "epoch": 0.8174327840416306, "grad_norm": 2.2515783309936523, "learning_rate": 1.4603220658639202e-05, "loss": 0.0128, "step": 7540 }, { "epoch": 0.8175411968777103, "grad_norm": 1.5120643377304077, "learning_rate": 1.4602495285071814e-05, "loss": 0.0276, "step": 7541 }, { "epoch": 0.8176496097137901, "grad_norm": 7.333964824676514, "learning_rate": 1.4601769911504426e-05, "loss": 0.0546, "step": 7542 }, { "epoch": 0.81775802254987, "grad_norm": 2.509059190750122, "learning_rate": 1.4601044537937038e-05, "loss": 0.0395, "step": 7543 }, { "epoch": 0.8178664353859497, "grad_norm": 2.8642194271087646, "learning_rate": 1.4600319164369652e-05, "loss": 0.0246, "step": 7544 }, { "epoch": 0.8179748482220295, "grad_norm": 2.0711264610290527, "learning_rate": 1.4599593790802264e-05, "loss": 0.0612, "step": 7545 }, { "epoch": 0.8180832610581092, "grad_norm": 0.908037006855011, "learning_rate": 1.4598868417234877e-05, "loss": 0.0188, "step": 7546 }, { "epoch": 0.8181916738941891, "grad_norm": 1.5599621534347534, "learning_rate": 1.459814304366749e-05, "loss": 0.0234, "step": 7547 }, { "epoch": 0.8183000867302689, "grad_norm": 1.5958664417266846, "learning_rate": 1.4597417670100102e-05, "loss": 0.0517, "step": 7548 }, { "epoch": 0.8184084995663486, "grad_norm": 2.920579671859741, "learning_rate": 1.4596692296532715e-05, "loss": 0.0296, "step": 7549 }, { "epoch": 0.8185169124024284, "grad_norm": 2.298768997192383, "learning_rate": 1.4595966922965328e-05, "loss": 0.021, "step": 7550 }, { "epoch": 0.8186253252385083, "grad_norm": 3.21612286567688, "learning_rate": 1.459524154939794e-05, "loss": 0.0323, "step": 7551 }, { "epoch": 0.818733738074588, "grad_norm": 1.6515885591506958, "learning_rate": 1.4594516175830553e-05, "loss": 0.0349, "step": 7552 }, { "epoch": 0.8188421509106678, "grad_norm": 2.8358516693115234, "learning_rate": 1.4593790802263167e-05, "loss": 0.0514, "step": 7553 }, { "epoch": 0.8189505637467476, "grad_norm": 5.158514499664307, "learning_rate": 1.4593065428695779e-05, "loss": 0.0298, "step": 7554 }, { "epoch": 0.8190589765828274, "grad_norm": 4.093610763549805, "learning_rate": 1.4592340055128393e-05, "loss": 0.0435, "step": 7555 }, { "epoch": 0.8191673894189072, "grad_norm": 0.8052939176559448, "learning_rate": 1.4591614681561006e-05, "loss": 0.0061, "step": 7556 }, { "epoch": 0.819275802254987, "grad_norm": 1.678462266921997, "learning_rate": 1.4590889307993619e-05, "loss": 0.0382, "step": 7557 }, { "epoch": 0.8193842150910667, "grad_norm": 6.6689887046813965, "learning_rate": 1.459016393442623e-05, "loss": 0.0685, "step": 7558 }, { "epoch": 0.8194926279271466, "grad_norm": 2.6543056964874268, "learning_rate": 1.4589438560858845e-05, "loss": 0.042, "step": 7559 }, { "epoch": 0.8196010407632264, "grad_norm": 1.8368160724639893, "learning_rate": 1.4588713187291457e-05, "loss": 0.051, "step": 7560 }, { "epoch": 0.8197094535993061, "grad_norm": 6.2339043617248535, "learning_rate": 1.4587987813724069e-05, "loss": 0.0546, "step": 7561 }, { "epoch": 0.8198178664353859, "grad_norm": 1.9476113319396973, "learning_rate": 1.4587262440156683e-05, "loss": 0.052, "step": 7562 }, { "epoch": 0.8199262792714658, "grad_norm": 1.083762288093567, "learning_rate": 1.4586537066589295e-05, "loss": 0.009, "step": 7563 }, { "epoch": 0.8200346921075455, "grad_norm": 2.76418399810791, "learning_rate": 1.4585811693021907e-05, "loss": 0.0349, "step": 7564 }, { "epoch": 0.8201431049436253, "grad_norm": 0.6401739716529846, "learning_rate": 1.458508631945452e-05, "loss": 0.0079, "step": 7565 }, { "epoch": 0.8202515177797051, "grad_norm": 0.5077617764472961, "learning_rate": 1.4584360945887133e-05, "loss": 0.0168, "step": 7566 }, { "epoch": 0.8203599306157849, "grad_norm": 3.2044246196746826, "learning_rate": 1.4583635572319745e-05, "loss": 0.0564, "step": 7567 }, { "epoch": 0.8204683434518647, "grad_norm": 4.541947364807129, "learning_rate": 1.4582910198752357e-05, "loss": 0.0182, "step": 7568 }, { "epoch": 0.8205767562879445, "grad_norm": 1.267189860343933, "learning_rate": 1.4582184825184971e-05, "loss": 0.0294, "step": 7569 }, { "epoch": 0.8206851691240243, "grad_norm": 3.7325150966644287, "learning_rate": 1.4581459451617583e-05, "loss": 0.0609, "step": 7570 }, { "epoch": 0.8207935819601041, "grad_norm": 1.7807230949401855, "learning_rate": 1.4580734078050195e-05, "loss": 0.0314, "step": 7571 }, { "epoch": 0.8209019947961839, "grad_norm": 5.606831073760986, "learning_rate": 1.4580008704482811e-05, "loss": 0.0331, "step": 7572 }, { "epoch": 0.8210104076322636, "grad_norm": 1.29137122631073, "learning_rate": 1.4579283330915423e-05, "loss": 0.0234, "step": 7573 }, { "epoch": 0.8211188204683435, "grad_norm": 0.6355518102645874, "learning_rate": 1.4578557957348037e-05, "loss": 0.0178, "step": 7574 }, { "epoch": 0.8212272333044233, "grad_norm": 1.4677907228469849, "learning_rate": 1.4577832583780649e-05, "loss": 0.0288, "step": 7575 }, { "epoch": 0.821335646140503, "grad_norm": 1.1966886520385742, "learning_rate": 1.4577107210213261e-05, "loss": 0.024, "step": 7576 }, { "epoch": 0.8214440589765828, "grad_norm": 2.9755802154541016, "learning_rate": 1.4576381836645875e-05, "loss": 0.0417, "step": 7577 }, { "epoch": 0.8215524718126627, "grad_norm": 1.8162094354629517, "learning_rate": 1.4575656463078487e-05, "loss": 0.0239, "step": 7578 }, { "epoch": 0.8216608846487424, "grad_norm": 4.302316665649414, "learning_rate": 1.45749310895111e-05, "loss": 0.0279, "step": 7579 }, { "epoch": 0.8217692974848222, "grad_norm": 2.938328504562378, "learning_rate": 1.4574205715943711e-05, "loss": 0.0213, "step": 7580 }, { "epoch": 0.821877710320902, "grad_norm": 1.1613227128982544, "learning_rate": 1.4573480342376325e-05, "loss": 0.0212, "step": 7581 }, { "epoch": 0.8219861231569818, "grad_norm": 2.1028599739074707, "learning_rate": 1.4572754968808937e-05, "loss": 0.0317, "step": 7582 }, { "epoch": 0.8220945359930616, "grad_norm": 2.913407325744629, "learning_rate": 1.457202959524155e-05, "loss": 0.0214, "step": 7583 }, { "epoch": 0.8222029488291414, "grad_norm": 1.2352089881896973, "learning_rate": 1.4571304221674163e-05, "loss": 0.0246, "step": 7584 }, { "epoch": 0.8223113616652211, "grad_norm": 3.0610761642456055, "learning_rate": 1.4570578848106776e-05, "loss": 0.0361, "step": 7585 }, { "epoch": 0.822419774501301, "grad_norm": 3.3661062717437744, "learning_rate": 1.4569853474539388e-05, "loss": 0.0312, "step": 7586 }, { "epoch": 0.8225281873373808, "grad_norm": 0.33693867921829224, "learning_rate": 1.4569128100972001e-05, "loss": 0.0077, "step": 7587 }, { "epoch": 0.8226366001734605, "grad_norm": 3.6577131748199463, "learning_rate": 1.4568402727404614e-05, "loss": 0.0691, "step": 7588 }, { "epoch": 0.8227450130095403, "grad_norm": 1.4598734378814697, "learning_rate": 1.4567677353837226e-05, "loss": 0.0175, "step": 7589 }, { "epoch": 0.8228534258456202, "grad_norm": 3.264315366744995, "learning_rate": 1.4566951980269841e-05, "loss": 0.0342, "step": 7590 }, { "epoch": 0.8229618386816999, "grad_norm": 0.4149165749549866, "learning_rate": 1.4566226606702453e-05, "loss": 0.0029, "step": 7591 }, { "epoch": 0.8230702515177797, "grad_norm": 6.373319149017334, "learning_rate": 1.4565501233135067e-05, "loss": 0.0237, "step": 7592 }, { "epoch": 0.8231786643538594, "grad_norm": 3.6513359546661377, "learning_rate": 1.456477585956768e-05, "loss": 0.0805, "step": 7593 }, { "epoch": 0.8232870771899393, "grad_norm": 0.42534568905830383, "learning_rate": 1.4564050486000292e-05, "loss": 0.0034, "step": 7594 }, { "epoch": 0.8233954900260191, "grad_norm": 4.9658732414245605, "learning_rate": 1.4563325112432904e-05, "loss": 0.0735, "step": 7595 }, { "epoch": 0.8235039028620988, "grad_norm": 6.059704780578613, "learning_rate": 1.4562599738865518e-05, "loss": 0.0827, "step": 7596 }, { "epoch": 0.8236123156981786, "grad_norm": 0.7223392724990845, "learning_rate": 1.456187436529813e-05, "loss": 0.0181, "step": 7597 }, { "epoch": 0.8237207285342585, "grad_norm": 1.4969878196716309, "learning_rate": 1.4561148991730742e-05, "loss": 0.046, "step": 7598 }, { "epoch": 0.8238291413703382, "grad_norm": 4.472140789031982, "learning_rate": 1.4560423618163356e-05, "loss": 0.0381, "step": 7599 }, { "epoch": 0.823937554206418, "grad_norm": 3.616239070892334, "learning_rate": 1.4559698244595968e-05, "loss": 0.037, "step": 7600 }, { "epoch": 0.8240459670424979, "grad_norm": 1.785622000694275, "learning_rate": 1.455897287102858e-05, "loss": 0.0145, "step": 7601 }, { "epoch": 0.8241543798785776, "grad_norm": 1.1516680717468262, "learning_rate": 1.4558247497461194e-05, "loss": 0.0338, "step": 7602 }, { "epoch": 0.8242627927146574, "grad_norm": 2.38808012008667, "learning_rate": 1.4557522123893806e-05, "loss": 0.0377, "step": 7603 }, { "epoch": 0.8243712055507372, "grad_norm": 1.8069671392440796, "learning_rate": 1.4556796750326418e-05, "loss": 0.064, "step": 7604 }, { "epoch": 0.824479618386817, "grad_norm": 5.286204814910889, "learning_rate": 1.4556071376759032e-05, "loss": 0.0191, "step": 7605 }, { "epoch": 0.8245880312228968, "grad_norm": 4.224722862243652, "learning_rate": 1.4555346003191644e-05, "loss": 0.0799, "step": 7606 }, { "epoch": 0.8246964440589766, "grad_norm": 4.888130187988281, "learning_rate": 1.4554620629624258e-05, "loss": 0.0385, "step": 7607 }, { "epoch": 0.8248048568950563, "grad_norm": 3.5419695377349854, "learning_rate": 1.4553895256056872e-05, "loss": 0.0911, "step": 7608 }, { "epoch": 0.8249132697311362, "grad_norm": 1.567664623260498, "learning_rate": 1.4553169882489484e-05, "loss": 0.0168, "step": 7609 }, { "epoch": 0.825021682567216, "grad_norm": 6.114418029785156, "learning_rate": 1.4552444508922096e-05, "loss": 0.0898, "step": 7610 }, { "epoch": 0.8251300954032957, "grad_norm": 3.8896286487579346, "learning_rate": 1.455171913535471e-05, "loss": 0.0567, "step": 7611 }, { "epoch": 0.8252385082393755, "grad_norm": 1.4627233743667603, "learning_rate": 1.4550993761787322e-05, "loss": 0.0433, "step": 7612 }, { "epoch": 0.8253469210754554, "grad_norm": 3.4290835857391357, "learning_rate": 1.4550268388219934e-05, "loss": 0.0483, "step": 7613 }, { "epoch": 0.8254553339115351, "grad_norm": 0.6228319406509399, "learning_rate": 1.4549543014652548e-05, "loss": 0.0132, "step": 7614 }, { "epoch": 0.8255637467476149, "grad_norm": 1.7691465616226196, "learning_rate": 1.454881764108516e-05, "loss": 0.0165, "step": 7615 }, { "epoch": 0.8256721595836947, "grad_norm": 0.9309958815574646, "learning_rate": 1.4548092267517772e-05, "loss": 0.0268, "step": 7616 }, { "epoch": 0.8257805724197745, "grad_norm": 2.851771593093872, "learning_rate": 1.4547366893950386e-05, "loss": 0.052, "step": 7617 }, { "epoch": 0.8258889852558543, "grad_norm": 4.329989910125732, "learning_rate": 1.4546641520382998e-05, "loss": 0.0269, "step": 7618 }, { "epoch": 0.8259973980919341, "grad_norm": 3.066990613937378, "learning_rate": 1.454591614681561e-05, "loss": 0.0435, "step": 7619 }, { "epoch": 0.8261058109280138, "grad_norm": 1.5751270055770874, "learning_rate": 1.4545190773248223e-05, "loss": 0.0737, "step": 7620 }, { "epoch": 0.8262142237640937, "grad_norm": 0.43571674823760986, "learning_rate": 1.4544465399680836e-05, "loss": 0.0217, "step": 7621 }, { "epoch": 0.8263226366001735, "grad_norm": 4.916717529296875, "learning_rate": 1.4543740026113449e-05, "loss": 0.0467, "step": 7622 }, { "epoch": 0.8264310494362532, "grad_norm": 3.6659066677093506, "learning_rate": 1.454301465254606e-05, "loss": 0.0443, "step": 7623 }, { "epoch": 0.826539462272333, "grad_norm": 1.9976303577423096, "learning_rate": 1.4542289278978676e-05, "loss": 0.0191, "step": 7624 }, { "epoch": 0.8266478751084129, "grad_norm": 2.8584983348846436, "learning_rate": 1.4541563905411288e-05, "loss": 0.0753, "step": 7625 }, { "epoch": 0.8267562879444926, "grad_norm": 0.9754735827445984, "learning_rate": 1.4540838531843902e-05, "loss": 0.0101, "step": 7626 }, { "epoch": 0.8268647007805724, "grad_norm": 1.3172580003738403, "learning_rate": 1.4540113158276514e-05, "loss": 0.0141, "step": 7627 }, { "epoch": 0.8269731136166523, "grad_norm": 3.2221927642822266, "learning_rate": 1.4539387784709126e-05, "loss": 0.0727, "step": 7628 }, { "epoch": 0.827081526452732, "grad_norm": 2.162903308868408, "learning_rate": 1.453866241114174e-05, "loss": 0.0484, "step": 7629 }, { "epoch": 0.8271899392888118, "grad_norm": 2.1483428478240967, "learning_rate": 1.4537937037574352e-05, "loss": 0.0305, "step": 7630 }, { "epoch": 0.8272983521248916, "grad_norm": 1.4259846210479736, "learning_rate": 1.4537211664006965e-05, "loss": 0.0223, "step": 7631 }, { "epoch": 0.8274067649609714, "grad_norm": 2.5833041667938232, "learning_rate": 1.4536486290439578e-05, "loss": 0.029, "step": 7632 }, { "epoch": 0.8275151777970512, "grad_norm": 0.7691939473152161, "learning_rate": 1.453576091687219e-05, "loss": 0.0094, "step": 7633 }, { "epoch": 0.827623590633131, "grad_norm": 1.8208458423614502, "learning_rate": 1.4535035543304803e-05, "loss": 0.0121, "step": 7634 }, { "epoch": 0.8277320034692107, "grad_norm": 3.0810532569885254, "learning_rate": 1.4534310169737415e-05, "loss": 0.0353, "step": 7635 }, { "epoch": 0.8278404163052906, "grad_norm": 0.2934591472148895, "learning_rate": 1.4533584796170029e-05, "loss": 0.0041, "step": 7636 }, { "epoch": 0.8279488291413704, "grad_norm": 1.3549602031707764, "learning_rate": 1.4532859422602641e-05, "loss": 0.0094, "step": 7637 }, { "epoch": 0.8280572419774501, "grad_norm": 2.1604855060577393, "learning_rate": 1.4532134049035253e-05, "loss": 0.0366, "step": 7638 }, { "epoch": 0.8281656548135299, "grad_norm": 2.069974899291992, "learning_rate": 1.4531408675467867e-05, "loss": 0.0344, "step": 7639 }, { "epoch": 0.8282740676496098, "grad_norm": 3.235337495803833, "learning_rate": 1.4530683301900479e-05, "loss": 0.0261, "step": 7640 }, { "epoch": 0.8283824804856895, "grad_norm": 2.725832462310791, "learning_rate": 1.4529957928333095e-05, "loss": 0.0444, "step": 7641 }, { "epoch": 0.8284908933217693, "grad_norm": 0.938290536403656, "learning_rate": 1.4529232554765707e-05, "loss": 0.0211, "step": 7642 }, { "epoch": 0.828599306157849, "grad_norm": 2.6649088859558105, "learning_rate": 1.4528507181198319e-05, "loss": 0.0427, "step": 7643 }, { "epoch": 0.8287077189939289, "grad_norm": 0.6667219996452332, "learning_rate": 1.4527781807630933e-05, "loss": 0.0212, "step": 7644 }, { "epoch": 0.8288161318300087, "grad_norm": 4.345589637756348, "learning_rate": 1.4527056434063545e-05, "loss": 0.0537, "step": 7645 }, { "epoch": 0.8289245446660884, "grad_norm": 1.266364336013794, "learning_rate": 1.4526331060496157e-05, "loss": 0.0074, "step": 7646 }, { "epoch": 0.8290329575021682, "grad_norm": 0.7066763043403625, "learning_rate": 1.4525605686928769e-05, "loss": 0.0135, "step": 7647 }, { "epoch": 0.8291413703382481, "grad_norm": 0.5233468413352966, "learning_rate": 1.4524880313361383e-05, "loss": 0.0021, "step": 7648 }, { "epoch": 0.8292497831743278, "grad_norm": 6.11190938949585, "learning_rate": 1.4524154939793995e-05, "loss": 0.0886, "step": 7649 }, { "epoch": 0.8293581960104076, "grad_norm": 2.573697090148926, "learning_rate": 1.4523429566226607e-05, "loss": 0.0602, "step": 7650 }, { "epoch": 0.8294666088464874, "grad_norm": 3.7580504417419434, "learning_rate": 1.4522704192659221e-05, "loss": 0.0485, "step": 7651 }, { "epoch": 0.8295750216825672, "grad_norm": 2.9872117042541504, "learning_rate": 1.4521978819091833e-05, "loss": 0.0199, "step": 7652 }, { "epoch": 0.829683434518647, "grad_norm": 1.8226041793823242, "learning_rate": 1.4521253445524445e-05, "loss": 0.0244, "step": 7653 }, { "epoch": 0.8297918473547268, "grad_norm": 2.4908366203308105, "learning_rate": 1.4520528071957059e-05, "loss": 0.0237, "step": 7654 }, { "epoch": 0.8299002601908065, "grad_norm": 0.9695826172828674, "learning_rate": 1.4519802698389671e-05, "loss": 0.0189, "step": 7655 }, { "epoch": 0.8300086730268864, "grad_norm": 2.7151966094970703, "learning_rate": 1.4519077324822283e-05, "loss": 0.0315, "step": 7656 }, { "epoch": 0.8301170858629662, "grad_norm": 1.6215153932571411, "learning_rate": 1.4518351951254897e-05, "loss": 0.0242, "step": 7657 }, { "epoch": 0.8302254986990459, "grad_norm": 1.3158690929412842, "learning_rate": 1.451762657768751e-05, "loss": 0.017, "step": 7658 }, { "epoch": 0.8303339115351258, "grad_norm": 4.899900913238525, "learning_rate": 1.4516901204120123e-05, "loss": 0.0483, "step": 7659 }, { "epoch": 0.8304423243712056, "grad_norm": 2.449690580368042, "learning_rate": 1.4516175830552737e-05, "loss": 0.0376, "step": 7660 }, { "epoch": 0.8305507372072853, "grad_norm": 1.796509861946106, "learning_rate": 1.451545045698535e-05, "loss": 0.014, "step": 7661 }, { "epoch": 0.8306591500433651, "grad_norm": 4.087615966796875, "learning_rate": 1.4514725083417961e-05, "loss": 0.0864, "step": 7662 }, { "epoch": 0.830767562879445, "grad_norm": 1.282423973083496, "learning_rate": 1.4513999709850575e-05, "loss": 0.0183, "step": 7663 }, { "epoch": 0.8308759757155247, "grad_norm": 1.9652756452560425, "learning_rate": 1.4513274336283187e-05, "loss": 0.0405, "step": 7664 }, { "epoch": 0.8309843885516045, "grad_norm": 1.6701141595840454, "learning_rate": 1.45125489627158e-05, "loss": 0.026, "step": 7665 }, { "epoch": 0.8310928013876843, "grad_norm": 0.4532533884048462, "learning_rate": 1.4511823589148413e-05, "loss": 0.0113, "step": 7666 }, { "epoch": 0.8312012142237641, "grad_norm": 0.8756096363067627, "learning_rate": 1.4511098215581025e-05, "loss": 0.0077, "step": 7667 }, { "epoch": 0.8313096270598439, "grad_norm": 8.051863670349121, "learning_rate": 1.4510372842013638e-05, "loss": 0.1418, "step": 7668 }, { "epoch": 0.8314180398959237, "grad_norm": 2.510385751724243, "learning_rate": 1.4509647468446251e-05, "loss": 0.0424, "step": 7669 }, { "epoch": 0.8315264527320034, "grad_norm": 4.202641487121582, "learning_rate": 1.4508922094878864e-05, "loss": 0.0489, "step": 7670 }, { "epoch": 0.8316348655680833, "grad_norm": 3.0357439517974854, "learning_rate": 1.4508196721311476e-05, "loss": 0.0289, "step": 7671 }, { "epoch": 0.8317432784041631, "grad_norm": 2.650383949279785, "learning_rate": 1.4507471347744088e-05, "loss": 0.0203, "step": 7672 }, { "epoch": 0.8318516912402428, "grad_norm": 2.5947718620300293, "learning_rate": 1.4506745974176702e-05, "loss": 0.0516, "step": 7673 }, { "epoch": 0.8319601040763226, "grad_norm": 6.227081775665283, "learning_rate": 1.4506020600609314e-05, "loss": 0.106, "step": 7674 }, { "epoch": 0.8320685169124025, "grad_norm": 3.3611936569213867, "learning_rate": 1.4505295227041926e-05, "loss": 0.0389, "step": 7675 }, { "epoch": 0.8321769297484822, "grad_norm": 11.460087776184082, "learning_rate": 1.4504569853474542e-05, "loss": 0.0284, "step": 7676 }, { "epoch": 0.832285342584562, "grad_norm": 4.608936309814453, "learning_rate": 1.4503844479907154e-05, "loss": 0.0367, "step": 7677 }, { "epoch": 0.8323937554206418, "grad_norm": 3.8873043060302734, "learning_rate": 1.4503119106339768e-05, "loss": 0.0327, "step": 7678 }, { "epoch": 0.8325021682567216, "grad_norm": 2.3878939151763916, "learning_rate": 1.450239373277238e-05, "loss": 0.0313, "step": 7679 }, { "epoch": 0.8326105810928014, "grad_norm": 0.7442169785499573, "learning_rate": 1.4501668359204992e-05, "loss": 0.0063, "step": 7680 }, { "epoch": 0.8327189939288812, "grad_norm": 1.4730690717697144, "learning_rate": 1.4500942985637606e-05, "loss": 0.0131, "step": 7681 }, { "epoch": 0.8328274067649609, "grad_norm": 2.9507577419281006, "learning_rate": 1.4500217612070218e-05, "loss": 0.0483, "step": 7682 }, { "epoch": 0.8329358196010408, "grad_norm": 4.087472915649414, "learning_rate": 1.449949223850283e-05, "loss": 0.0799, "step": 7683 }, { "epoch": 0.8330442324371206, "grad_norm": 1.1227678060531616, "learning_rate": 1.4498766864935444e-05, "loss": 0.0107, "step": 7684 }, { "epoch": 0.8331526452732003, "grad_norm": 2.4923219680786133, "learning_rate": 1.4498041491368056e-05, "loss": 0.0414, "step": 7685 }, { "epoch": 0.8332610581092801, "grad_norm": 1.0094516277313232, "learning_rate": 1.4497316117800668e-05, "loss": 0.0219, "step": 7686 }, { "epoch": 0.83336947094536, "grad_norm": 2.389533281326294, "learning_rate": 1.449659074423328e-05, "loss": 0.0584, "step": 7687 }, { "epoch": 0.8334778837814397, "grad_norm": 7.610772132873535, "learning_rate": 1.4495865370665894e-05, "loss": 0.0636, "step": 7688 }, { "epoch": 0.8335862966175195, "grad_norm": 5.171246528625488, "learning_rate": 1.4495139997098506e-05, "loss": 0.0971, "step": 7689 }, { "epoch": 0.8336947094535994, "grad_norm": 3.051957845687866, "learning_rate": 1.4494414623531118e-05, "loss": 0.0287, "step": 7690 }, { "epoch": 0.8338031222896791, "grad_norm": 3.6726653575897217, "learning_rate": 1.4493689249963732e-05, "loss": 0.0132, "step": 7691 }, { "epoch": 0.8339115351257589, "grad_norm": 5.309534072875977, "learning_rate": 1.4492963876396344e-05, "loss": 0.0268, "step": 7692 }, { "epoch": 0.8340199479618386, "grad_norm": 3.6724236011505127, "learning_rate": 1.449223850282896e-05, "loss": 0.0597, "step": 7693 }, { "epoch": 0.8341283607979185, "grad_norm": 0.8327425718307495, "learning_rate": 1.4491513129261572e-05, "loss": 0.0143, "step": 7694 }, { "epoch": 0.8342367736339983, "grad_norm": 2.1067049503326416, "learning_rate": 1.4490787755694184e-05, "loss": 0.0201, "step": 7695 }, { "epoch": 0.834345186470078, "grad_norm": 1.4612563848495483, "learning_rate": 1.4490062382126798e-05, "loss": 0.0414, "step": 7696 }, { "epoch": 0.8344535993061578, "grad_norm": 2.767756938934326, "learning_rate": 1.448933700855941e-05, "loss": 0.0393, "step": 7697 }, { "epoch": 0.8345620121422377, "grad_norm": 5.554281711578369, "learning_rate": 1.4488611634992022e-05, "loss": 0.0655, "step": 7698 }, { "epoch": 0.8346704249783174, "grad_norm": 4.072072982788086, "learning_rate": 1.4487886261424634e-05, "loss": 0.0327, "step": 7699 }, { "epoch": 0.8347788378143972, "grad_norm": 1.6816565990447998, "learning_rate": 1.4487160887857248e-05, "loss": 0.0279, "step": 7700 }, { "epoch": 0.834887250650477, "grad_norm": 3.0193698406219482, "learning_rate": 1.448643551428986e-05, "loss": 0.032, "step": 7701 }, { "epoch": 0.8349956634865568, "grad_norm": 0.6146589517593384, "learning_rate": 1.4485710140722473e-05, "loss": 0.0087, "step": 7702 }, { "epoch": 0.8351040763226366, "grad_norm": 3.4668447971343994, "learning_rate": 1.4484984767155086e-05, "loss": 0.0403, "step": 7703 }, { "epoch": 0.8352124891587164, "grad_norm": 1.3368556499481201, "learning_rate": 1.4484259393587699e-05, "loss": 0.0213, "step": 7704 }, { "epoch": 0.8353209019947961, "grad_norm": 4.23341703414917, "learning_rate": 1.448353402002031e-05, "loss": 0.0344, "step": 7705 }, { "epoch": 0.835429314830876, "grad_norm": 4.86683464050293, "learning_rate": 1.4482808646452924e-05, "loss": 0.054, "step": 7706 }, { "epoch": 0.8355377276669558, "grad_norm": 2.1347081661224365, "learning_rate": 1.4482083272885537e-05, "loss": 0.0352, "step": 7707 }, { "epoch": 0.8356461405030355, "grad_norm": 2.5002570152282715, "learning_rate": 1.4481357899318149e-05, "loss": 0.0504, "step": 7708 }, { "epoch": 0.8357545533391153, "grad_norm": 2.098775625228882, "learning_rate": 1.4480632525750763e-05, "loss": 0.0286, "step": 7709 }, { "epoch": 0.8358629661751952, "grad_norm": 2.662959098815918, "learning_rate": 1.4479907152183376e-05, "loss": 0.0372, "step": 7710 }, { "epoch": 0.8359713790112749, "grad_norm": 5.272482872009277, "learning_rate": 1.4479181778615989e-05, "loss": 0.0549, "step": 7711 }, { "epoch": 0.8360797918473547, "grad_norm": 2.1402361392974854, "learning_rate": 1.4478456405048602e-05, "loss": 0.0201, "step": 7712 }, { "epoch": 0.8361882046834345, "grad_norm": 3.3613171577453613, "learning_rate": 1.4477731031481215e-05, "loss": 0.0271, "step": 7713 }, { "epoch": 0.8362966175195143, "grad_norm": 2.150831699371338, "learning_rate": 1.4477005657913827e-05, "loss": 0.045, "step": 7714 }, { "epoch": 0.8364050303555941, "grad_norm": 2.846318006515503, "learning_rate": 1.447628028434644e-05, "loss": 0.0378, "step": 7715 }, { "epoch": 0.8365134431916739, "grad_norm": 5.660454273223877, "learning_rate": 1.4475554910779053e-05, "loss": 0.0392, "step": 7716 }, { "epoch": 0.8366218560277537, "grad_norm": 0.7087425589561462, "learning_rate": 1.4474829537211665e-05, "loss": 0.0157, "step": 7717 }, { "epoch": 0.8367302688638335, "grad_norm": 1.195578694343567, "learning_rate": 1.4474104163644279e-05, "loss": 0.0251, "step": 7718 }, { "epoch": 0.8368386816999133, "grad_norm": 1.6152852773666382, "learning_rate": 1.447337879007689e-05, "loss": 0.0164, "step": 7719 }, { "epoch": 0.836947094535993, "grad_norm": 3.934096336364746, "learning_rate": 1.4472653416509503e-05, "loss": 0.0507, "step": 7720 }, { "epoch": 0.8370555073720729, "grad_norm": 13.031949996948242, "learning_rate": 1.4471928042942117e-05, "loss": 0.098, "step": 7721 }, { "epoch": 0.8371639202081527, "grad_norm": 3.7423861026763916, "learning_rate": 1.4471202669374729e-05, "loss": 0.0337, "step": 7722 }, { "epoch": 0.8372723330442324, "grad_norm": 0.5759287476539612, "learning_rate": 1.4470477295807341e-05, "loss": 0.0072, "step": 7723 }, { "epoch": 0.8373807458803122, "grad_norm": 3.10174298286438, "learning_rate": 1.4469751922239953e-05, "loss": 0.0469, "step": 7724 }, { "epoch": 0.8374891587163921, "grad_norm": 1.2307277917861938, "learning_rate": 1.4469026548672567e-05, "loss": 0.0255, "step": 7725 }, { "epoch": 0.8375975715524718, "grad_norm": 2.721863031387329, "learning_rate": 1.446830117510518e-05, "loss": 0.0374, "step": 7726 }, { "epoch": 0.8377059843885516, "grad_norm": 3.5014452934265137, "learning_rate": 1.4467575801537791e-05, "loss": 0.0176, "step": 7727 }, { "epoch": 0.8378143972246314, "grad_norm": 3.1737849712371826, "learning_rate": 1.4466850427970407e-05, "loss": 0.0591, "step": 7728 }, { "epoch": 0.8379228100607112, "grad_norm": 2.019073247909546, "learning_rate": 1.4466125054403019e-05, "loss": 0.0377, "step": 7729 }, { "epoch": 0.838031222896791, "grad_norm": 1.6378589868545532, "learning_rate": 1.4465399680835633e-05, "loss": 0.0361, "step": 7730 }, { "epoch": 0.8381396357328708, "grad_norm": 1.8822990655899048, "learning_rate": 1.4464674307268245e-05, "loss": 0.0369, "step": 7731 }, { "epoch": 0.8382480485689505, "grad_norm": 1.4829810857772827, "learning_rate": 1.4463948933700857e-05, "loss": 0.0178, "step": 7732 }, { "epoch": 0.8383564614050304, "grad_norm": 2.888023614883423, "learning_rate": 1.4463223560133471e-05, "loss": 0.0207, "step": 7733 }, { "epoch": 0.8384648742411102, "grad_norm": 3.73431134223938, "learning_rate": 1.4462498186566083e-05, "loss": 0.0606, "step": 7734 }, { "epoch": 0.8385732870771899, "grad_norm": 2.4504361152648926, "learning_rate": 1.4461772812998695e-05, "loss": 0.0258, "step": 7735 }, { "epoch": 0.8386816999132697, "grad_norm": 4.595053672790527, "learning_rate": 1.4461047439431309e-05, "loss": 0.0372, "step": 7736 }, { "epoch": 0.8387901127493496, "grad_norm": 2.863325595855713, "learning_rate": 1.4460322065863921e-05, "loss": 0.0232, "step": 7737 }, { "epoch": 0.8388985255854293, "grad_norm": 3.6690409183502197, "learning_rate": 1.4459596692296533e-05, "loss": 0.0983, "step": 7738 }, { "epoch": 0.8390069384215091, "grad_norm": 4.790167808532715, "learning_rate": 1.4458871318729146e-05, "loss": 0.0211, "step": 7739 }, { "epoch": 0.8391153512575888, "grad_norm": 3.115015983581543, "learning_rate": 1.445814594516176e-05, "loss": 0.0362, "step": 7740 }, { "epoch": 0.8392237640936687, "grad_norm": 0.7343436479568481, "learning_rate": 1.4457420571594372e-05, "loss": 0.0164, "step": 7741 }, { "epoch": 0.8393321769297485, "grad_norm": 2.2825098037719727, "learning_rate": 1.4456695198026984e-05, "loss": 0.0255, "step": 7742 }, { "epoch": 0.8394405897658282, "grad_norm": 3.369027614593506, "learning_rate": 1.4455969824459598e-05, "loss": 0.0305, "step": 7743 }, { "epoch": 0.839549002601908, "grad_norm": 4.099157333374023, "learning_rate": 1.445524445089221e-05, "loss": 0.0735, "step": 7744 }, { "epoch": 0.8396574154379879, "grad_norm": 2.3348636627197266, "learning_rate": 1.4454519077324825e-05, "loss": 0.0489, "step": 7745 }, { "epoch": 0.8397658282740676, "grad_norm": 0.39408791065216064, "learning_rate": 1.4453793703757437e-05, "loss": 0.0181, "step": 7746 }, { "epoch": 0.8398742411101474, "grad_norm": 5.355527400970459, "learning_rate": 1.445306833019005e-05, "loss": 0.0546, "step": 7747 }, { "epoch": 0.8399826539462273, "grad_norm": 1.8250144720077515, "learning_rate": 1.4452342956622663e-05, "loss": 0.05, "step": 7748 }, { "epoch": 0.840091066782307, "grad_norm": 4.890260696411133, "learning_rate": 1.4451617583055275e-05, "loss": 0.0849, "step": 7749 }, { "epoch": 0.8401994796183868, "grad_norm": 1.468627691268921, "learning_rate": 1.4450892209487888e-05, "loss": 0.0395, "step": 7750 }, { "epoch": 0.8403078924544666, "grad_norm": 3.909935235977173, "learning_rate": 1.44501668359205e-05, "loss": 0.0768, "step": 7751 }, { "epoch": 0.8404163052905465, "grad_norm": 3.2218918800354004, "learning_rate": 1.4449441462353114e-05, "loss": 0.025, "step": 7752 }, { "epoch": 0.8405247181266262, "grad_norm": 1.4728916883468628, "learning_rate": 1.4448716088785726e-05, "loss": 0.0126, "step": 7753 }, { "epoch": 0.840633130962706, "grad_norm": 1.7101905345916748, "learning_rate": 1.4447990715218338e-05, "loss": 0.0539, "step": 7754 }, { "epoch": 0.8407415437987857, "grad_norm": 0.8134397268295288, "learning_rate": 1.4447265341650952e-05, "loss": 0.03, "step": 7755 }, { "epoch": 0.8408499566348656, "grad_norm": 1.2472096681594849, "learning_rate": 1.4446539968083564e-05, "loss": 0.0282, "step": 7756 }, { "epoch": 0.8409583694709454, "grad_norm": 1.6625324487686157, "learning_rate": 1.4445814594516176e-05, "loss": 0.0279, "step": 7757 }, { "epoch": 0.8410667823070251, "grad_norm": 5.132262229919434, "learning_rate": 1.444508922094879e-05, "loss": 0.043, "step": 7758 }, { "epoch": 0.8411751951431049, "grad_norm": 0.7404937148094177, "learning_rate": 1.4444363847381402e-05, "loss": 0.0162, "step": 7759 }, { "epoch": 0.8412836079791848, "grad_norm": 0.9357375502586365, "learning_rate": 1.4443638473814014e-05, "loss": 0.0173, "step": 7760 }, { "epoch": 0.8413920208152645, "grad_norm": 1.3719525337219238, "learning_rate": 1.4442913100246628e-05, "loss": 0.0097, "step": 7761 }, { "epoch": 0.8415004336513443, "grad_norm": 1.436344027519226, "learning_rate": 1.4442187726679242e-05, "loss": 0.015, "step": 7762 }, { "epoch": 0.8416088464874241, "grad_norm": 7.3505024909973145, "learning_rate": 1.4441462353111854e-05, "loss": 0.0383, "step": 7763 }, { "epoch": 0.8417172593235039, "grad_norm": 1.9809178113937378, "learning_rate": 1.4440736979544468e-05, "loss": 0.0217, "step": 7764 }, { "epoch": 0.8418256721595837, "grad_norm": 2.2942402362823486, "learning_rate": 1.444001160597708e-05, "loss": 0.0205, "step": 7765 }, { "epoch": 0.8419340849956635, "grad_norm": 5.854365825653076, "learning_rate": 1.4439286232409692e-05, "loss": 0.0802, "step": 7766 }, { "epoch": 0.8420424978317432, "grad_norm": 2.8826394081115723, "learning_rate": 1.4438560858842306e-05, "loss": 0.0465, "step": 7767 }, { "epoch": 0.8421509106678231, "grad_norm": 1.8397434949874878, "learning_rate": 1.4437835485274918e-05, "loss": 0.0184, "step": 7768 }, { "epoch": 0.8422593235039029, "grad_norm": 3.38736629486084, "learning_rate": 1.443711011170753e-05, "loss": 0.0152, "step": 7769 }, { "epoch": 0.8423677363399826, "grad_norm": 2.5979137420654297, "learning_rate": 1.4436384738140144e-05, "loss": 0.0622, "step": 7770 }, { "epoch": 0.8424761491760624, "grad_norm": 5.181448936462402, "learning_rate": 1.4435659364572756e-05, "loss": 0.0144, "step": 7771 }, { "epoch": 0.8425845620121423, "grad_norm": 7.228602409362793, "learning_rate": 1.4434933991005368e-05, "loss": 0.0381, "step": 7772 }, { "epoch": 0.842692974848222, "grad_norm": 6.479517936706543, "learning_rate": 1.4434208617437982e-05, "loss": 0.0355, "step": 7773 }, { "epoch": 0.8428013876843018, "grad_norm": 1.7372756004333496, "learning_rate": 1.4433483243870594e-05, "loss": 0.0567, "step": 7774 }, { "epoch": 0.8429098005203816, "grad_norm": 2.8091611862182617, "learning_rate": 1.4432757870303206e-05, "loss": 0.0569, "step": 7775 }, { "epoch": 0.8430182133564614, "grad_norm": 2.344705104827881, "learning_rate": 1.4432032496735819e-05, "loss": 0.0377, "step": 7776 }, { "epoch": 0.8431266261925412, "grad_norm": 3.132903814315796, "learning_rate": 1.4431307123168432e-05, "loss": 0.0079, "step": 7777 }, { "epoch": 0.843235039028621, "grad_norm": 7.892641067504883, "learning_rate": 1.4430581749601045e-05, "loss": 0.0628, "step": 7778 }, { "epoch": 0.8433434518647008, "grad_norm": 5.332473278045654, "learning_rate": 1.4429856376033657e-05, "loss": 0.0417, "step": 7779 }, { "epoch": 0.8434518647007806, "grad_norm": 1.384737253189087, "learning_rate": 1.4429131002466272e-05, "loss": 0.0264, "step": 7780 }, { "epoch": 0.8435602775368604, "grad_norm": 4.727534770965576, "learning_rate": 1.4428405628898884e-05, "loss": 0.0209, "step": 7781 }, { "epoch": 0.8436686903729401, "grad_norm": 2.7280895709991455, "learning_rate": 1.4427680255331498e-05, "loss": 0.0588, "step": 7782 }, { "epoch": 0.84377710320902, "grad_norm": 1.3331443071365356, "learning_rate": 1.442695488176411e-05, "loss": 0.0423, "step": 7783 }, { "epoch": 0.8438855160450998, "grad_norm": 0.6335434913635254, "learning_rate": 1.4426229508196722e-05, "loss": 0.0055, "step": 7784 }, { "epoch": 0.8439939288811795, "grad_norm": 1.220085859298706, "learning_rate": 1.4425504134629336e-05, "loss": 0.0255, "step": 7785 }, { "epoch": 0.8441023417172593, "grad_norm": 3.1604106426239014, "learning_rate": 1.4424778761061948e-05, "loss": 0.0302, "step": 7786 }, { "epoch": 0.8442107545533392, "grad_norm": 4.123033046722412, "learning_rate": 1.442405338749456e-05, "loss": 0.0652, "step": 7787 }, { "epoch": 0.8443191673894189, "grad_norm": 2.438953161239624, "learning_rate": 1.4423328013927174e-05, "loss": 0.0454, "step": 7788 }, { "epoch": 0.8444275802254987, "grad_norm": 1.6696487665176392, "learning_rate": 1.4422602640359787e-05, "loss": 0.0265, "step": 7789 }, { "epoch": 0.8445359930615784, "grad_norm": 1.6914844512939453, "learning_rate": 1.4421877266792399e-05, "loss": 0.0234, "step": 7790 }, { "epoch": 0.8446444058976583, "grad_norm": 2.4465205669403076, "learning_rate": 1.4421151893225011e-05, "loss": 0.0322, "step": 7791 }, { "epoch": 0.8447528187337381, "grad_norm": 1.3017476797103882, "learning_rate": 1.4420426519657625e-05, "loss": 0.0284, "step": 7792 }, { "epoch": 0.8448612315698178, "grad_norm": 3.8226113319396973, "learning_rate": 1.4419701146090237e-05, "loss": 0.1019, "step": 7793 }, { "epoch": 0.8449696444058976, "grad_norm": 2.1627771854400635, "learning_rate": 1.4418975772522849e-05, "loss": 0.0266, "step": 7794 }, { "epoch": 0.8450780572419775, "grad_norm": 1.928151249885559, "learning_rate": 1.4418250398955463e-05, "loss": 0.0172, "step": 7795 }, { "epoch": 0.8451864700780573, "grad_norm": 2.791259527206421, "learning_rate": 1.4417525025388075e-05, "loss": 0.0543, "step": 7796 }, { "epoch": 0.845294882914137, "grad_norm": 1.2184923887252808, "learning_rate": 1.441679965182069e-05, "loss": 0.022, "step": 7797 }, { "epoch": 0.8454032957502168, "grad_norm": 2.678586006164551, "learning_rate": 1.4416074278253303e-05, "loss": 0.031, "step": 7798 }, { "epoch": 0.8455117085862967, "grad_norm": 1.055680751800537, "learning_rate": 1.4415348904685915e-05, "loss": 0.0186, "step": 7799 }, { "epoch": 0.8456201214223764, "grad_norm": 0.3166171610355377, "learning_rate": 1.4414623531118529e-05, "loss": 0.0039, "step": 7800 }, { "epoch": 0.8457285342584562, "grad_norm": 2.0464141368865967, "learning_rate": 1.441389815755114e-05, "loss": 0.0467, "step": 7801 }, { "epoch": 0.8458369470945359, "grad_norm": 1.003382921218872, "learning_rate": 1.4413172783983753e-05, "loss": 0.0172, "step": 7802 }, { "epoch": 0.8459453599306158, "grad_norm": 5.714642524719238, "learning_rate": 1.4412447410416365e-05, "loss": 0.0963, "step": 7803 }, { "epoch": 0.8460537727666956, "grad_norm": 3.4555046558380127, "learning_rate": 1.4411722036848979e-05, "loss": 0.0381, "step": 7804 }, { "epoch": 0.8461621856027753, "grad_norm": 3.4450488090515137, "learning_rate": 1.4410996663281591e-05, "loss": 0.0605, "step": 7805 }, { "epoch": 0.8462705984388552, "grad_norm": 3.14210844039917, "learning_rate": 1.4410271289714203e-05, "loss": 0.0154, "step": 7806 }, { "epoch": 0.846379011274935, "grad_norm": 2.073488235473633, "learning_rate": 1.4409545916146817e-05, "loss": 0.0452, "step": 7807 }, { "epoch": 0.8464874241110147, "grad_norm": 2.0360610485076904, "learning_rate": 1.440882054257943e-05, "loss": 0.05, "step": 7808 }, { "epoch": 0.8465958369470945, "grad_norm": 0.3418780267238617, "learning_rate": 1.4408095169012041e-05, "loss": 0.0075, "step": 7809 }, { "epoch": 0.8467042497831744, "grad_norm": 1.8115208148956299, "learning_rate": 1.4407369795444655e-05, "loss": 0.0267, "step": 7810 }, { "epoch": 0.8468126626192541, "grad_norm": 5.98789644241333, "learning_rate": 1.4406644421877267e-05, "loss": 0.038, "step": 7811 }, { "epoch": 0.8469210754553339, "grad_norm": 1.5409396886825562, "learning_rate": 1.440591904830988e-05, "loss": 0.0464, "step": 7812 }, { "epoch": 0.8470294882914137, "grad_norm": 1.4791744947433472, "learning_rate": 1.4405193674742493e-05, "loss": 0.0218, "step": 7813 }, { "epoch": 0.8471379011274935, "grad_norm": 2.0195534229278564, "learning_rate": 1.4404468301175107e-05, "loss": 0.0304, "step": 7814 }, { "epoch": 0.8472463139635733, "grad_norm": 0.9792048931121826, "learning_rate": 1.440374292760772e-05, "loss": 0.0124, "step": 7815 }, { "epoch": 0.8473547267996531, "grad_norm": 1.662519097328186, "learning_rate": 1.4403017554040333e-05, "loss": 0.0259, "step": 7816 }, { "epoch": 0.8474631396357328, "grad_norm": 0.3378449082374573, "learning_rate": 1.4402292180472945e-05, "loss": 0.0126, "step": 7817 }, { "epoch": 0.8475715524718127, "grad_norm": 1.0974857807159424, "learning_rate": 1.4401566806905557e-05, "loss": 0.0249, "step": 7818 }, { "epoch": 0.8476799653078925, "grad_norm": 0.29895374178886414, "learning_rate": 1.4400841433338171e-05, "loss": 0.0144, "step": 7819 }, { "epoch": 0.8477883781439722, "grad_norm": 2.387584924697876, "learning_rate": 1.4400116059770783e-05, "loss": 0.0132, "step": 7820 }, { "epoch": 0.847896790980052, "grad_norm": 2.3453571796417236, "learning_rate": 1.4399390686203396e-05, "loss": 0.0233, "step": 7821 }, { "epoch": 0.8480052038161319, "grad_norm": 3.527333974838257, "learning_rate": 1.439866531263601e-05, "loss": 0.0799, "step": 7822 }, { "epoch": 0.8481136166522116, "grad_norm": 5.62752103805542, "learning_rate": 1.4397939939068621e-05, "loss": 0.0432, "step": 7823 }, { "epoch": 0.8482220294882914, "grad_norm": 1.561280608177185, "learning_rate": 1.4397214565501234e-05, "loss": 0.0128, "step": 7824 }, { "epoch": 0.8483304423243712, "grad_norm": 3.998690605163574, "learning_rate": 1.4396489191933847e-05, "loss": 0.0484, "step": 7825 }, { "epoch": 0.848438855160451, "grad_norm": 5.727916717529297, "learning_rate": 1.439576381836646e-05, "loss": 0.0206, "step": 7826 }, { "epoch": 0.8485472679965308, "grad_norm": 4.605086803436279, "learning_rate": 1.4395038444799072e-05, "loss": 0.0508, "step": 7827 }, { "epoch": 0.8486556808326106, "grad_norm": 3.1527535915374756, "learning_rate": 1.4394313071231686e-05, "loss": 0.0523, "step": 7828 }, { "epoch": 0.8487640936686903, "grad_norm": 2.5705888271331787, "learning_rate": 1.4393587697664298e-05, "loss": 0.0248, "step": 7829 }, { "epoch": 0.8488725065047702, "grad_norm": 2.0541038513183594, "learning_rate": 1.439286232409691e-05, "loss": 0.0329, "step": 7830 }, { "epoch": 0.84898091934085, "grad_norm": 2.5437700748443604, "learning_rate": 1.4392136950529525e-05, "loss": 0.0435, "step": 7831 }, { "epoch": 0.8490893321769297, "grad_norm": 2.1883063316345215, "learning_rate": 1.4391411576962138e-05, "loss": 0.0373, "step": 7832 }, { "epoch": 0.8491977450130095, "grad_norm": 1.1127691268920898, "learning_rate": 1.439068620339475e-05, "loss": 0.0219, "step": 7833 }, { "epoch": 0.8493061578490894, "grad_norm": 3.6785290241241455, "learning_rate": 1.4389960829827364e-05, "loss": 0.0421, "step": 7834 }, { "epoch": 0.8494145706851691, "grad_norm": 4.855587005615234, "learning_rate": 1.4389235456259976e-05, "loss": 0.0481, "step": 7835 }, { "epoch": 0.8495229835212489, "grad_norm": 1.3135778903961182, "learning_rate": 1.4388510082692588e-05, "loss": 0.0334, "step": 7836 }, { "epoch": 0.8496313963573288, "grad_norm": 2.1466526985168457, "learning_rate": 1.4387784709125202e-05, "loss": 0.0486, "step": 7837 }, { "epoch": 0.8497398091934085, "grad_norm": 0.6848601698875427, "learning_rate": 1.4387059335557814e-05, "loss": 0.0146, "step": 7838 }, { "epoch": 0.8498482220294883, "grad_norm": 0.6397877931594849, "learning_rate": 1.4386333961990426e-05, "loss": 0.0165, "step": 7839 }, { "epoch": 0.849956634865568, "grad_norm": 2.58969783782959, "learning_rate": 1.438560858842304e-05, "loss": 0.0385, "step": 7840 }, { "epoch": 0.8500650477016479, "grad_norm": 2.4322493076324463, "learning_rate": 1.4384883214855652e-05, "loss": 0.0323, "step": 7841 }, { "epoch": 0.8501734605377277, "grad_norm": 2.7524056434631348, "learning_rate": 1.4384157841288264e-05, "loss": 0.0448, "step": 7842 }, { "epoch": 0.8502818733738075, "grad_norm": 0.9600659012794495, "learning_rate": 1.4383432467720876e-05, "loss": 0.0181, "step": 7843 }, { "epoch": 0.8503902862098872, "grad_norm": 2.913025140762329, "learning_rate": 1.438270709415349e-05, "loss": 0.0726, "step": 7844 }, { "epoch": 0.8504986990459671, "grad_norm": 4.299576282501221, "learning_rate": 1.4381981720586102e-05, "loss": 0.0526, "step": 7845 }, { "epoch": 0.8506071118820469, "grad_norm": 1.744206190109253, "learning_rate": 1.4381256347018714e-05, "loss": 0.0136, "step": 7846 }, { "epoch": 0.8507155247181266, "grad_norm": 2.217249631881714, "learning_rate": 1.4380530973451328e-05, "loss": 0.0399, "step": 7847 }, { "epoch": 0.8508239375542064, "grad_norm": 5.920917510986328, "learning_rate": 1.437980559988394e-05, "loss": 0.04, "step": 7848 }, { "epoch": 0.8509323503902863, "grad_norm": 4.023726463317871, "learning_rate": 1.4379080226316556e-05, "loss": 0.072, "step": 7849 }, { "epoch": 0.851040763226366, "grad_norm": 2.7967512607574463, "learning_rate": 1.4378354852749168e-05, "loss": 0.0675, "step": 7850 }, { "epoch": 0.8511491760624458, "grad_norm": 1.1532061100006104, "learning_rate": 1.437762947918178e-05, "loss": 0.0209, "step": 7851 }, { "epoch": 0.8512575888985255, "grad_norm": 2.778616428375244, "learning_rate": 1.4376904105614394e-05, "loss": 0.0277, "step": 7852 }, { "epoch": 0.8513660017346054, "grad_norm": 1.45601487159729, "learning_rate": 1.4376178732047006e-05, "loss": 0.0222, "step": 7853 }, { "epoch": 0.8514744145706852, "grad_norm": 1.5088468790054321, "learning_rate": 1.4375453358479618e-05, "loss": 0.0223, "step": 7854 }, { "epoch": 0.8515828274067649, "grad_norm": 2.4209582805633545, "learning_rate": 1.437472798491223e-05, "loss": 0.0394, "step": 7855 }, { "epoch": 0.8516912402428447, "grad_norm": 3.0444424152374268, "learning_rate": 1.4374002611344844e-05, "loss": 0.0463, "step": 7856 }, { "epoch": 0.8517996530789246, "grad_norm": 1.075174331665039, "learning_rate": 1.4373277237777456e-05, "loss": 0.0271, "step": 7857 }, { "epoch": 0.8519080659150043, "grad_norm": 3.724353790283203, "learning_rate": 1.4372551864210069e-05, "loss": 0.0751, "step": 7858 }, { "epoch": 0.8520164787510841, "grad_norm": 0.8310491442680359, "learning_rate": 1.4371826490642682e-05, "loss": 0.0129, "step": 7859 }, { "epoch": 0.8521248915871639, "grad_norm": 4.171200275421143, "learning_rate": 1.4371101117075295e-05, "loss": 0.0742, "step": 7860 }, { "epoch": 0.8522333044232437, "grad_norm": 1.320491909980774, "learning_rate": 1.4370375743507907e-05, "loss": 0.0227, "step": 7861 }, { "epoch": 0.8523417172593235, "grad_norm": 3.626473903656006, "learning_rate": 1.436965036994052e-05, "loss": 0.0482, "step": 7862 }, { "epoch": 0.8524501300954033, "grad_norm": 1.5084596872329712, "learning_rate": 1.4368924996373133e-05, "loss": 0.0333, "step": 7863 }, { "epoch": 0.852558542931483, "grad_norm": 1.3997552394866943, "learning_rate": 1.4368199622805745e-05, "loss": 0.0289, "step": 7864 }, { "epoch": 0.8526669557675629, "grad_norm": 1.2833043336868286, "learning_rate": 1.4367474249238359e-05, "loss": 0.0179, "step": 7865 }, { "epoch": 0.8527753686036427, "grad_norm": 1.0089895725250244, "learning_rate": 1.4366748875670972e-05, "loss": 0.0383, "step": 7866 }, { "epoch": 0.8528837814397224, "grad_norm": 1.2866237163543701, "learning_rate": 1.4366023502103585e-05, "loss": 0.0124, "step": 7867 }, { "epoch": 0.8529921942758023, "grad_norm": 0.7608217597007751, "learning_rate": 1.4365298128536198e-05, "loss": 0.0123, "step": 7868 }, { "epoch": 0.8531006071118821, "grad_norm": 2.8762590885162354, "learning_rate": 1.436457275496881e-05, "loss": 0.0333, "step": 7869 }, { "epoch": 0.8532090199479618, "grad_norm": 0.8361130952835083, "learning_rate": 1.4363847381401423e-05, "loss": 0.0364, "step": 7870 }, { "epoch": 0.8533174327840416, "grad_norm": 1.8713176250457764, "learning_rate": 1.4363122007834037e-05, "loss": 0.0207, "step": 7871 }, { "epoch": 0.8534258456201215, "grad_norm": 2.0673043727874756, "learning_rate": 1.4362396634266649e-05, "loss": 0.0207, "step": 7872 }, { "epoch": 0.8535342584562012, "grad_norm": 7.250340938568115, "learning_rate": 1.4361671260699261e-05, "loss": 0.0304, "step": 7873 }, { "epoch": 0.853642671292281, "grad_norm": 1.3336946964263916, "learning_rate": 1.4360945887131875e-05, "loss": 0.0168, "step": 7874 }, { "epoch": 0.8537510841283608, "grad_norm": 1.9332295656204224, "learning_rate": 1.4360220513564487e-05, "loss": 0.0688, "step": 7875 }, { "epoch": 0.8538594969644406, "grad_norm": 5.4598069190979, "learning_rate": 1.4359495139997099e-05, "loss": 0.0673, "step": 7876 }, { "epoch": 0.8539679098005204, "grad_norm": 3.4660959243774414, "learning_rate": 1.4358769766429713e-05, "loss": 0.0337, "step": 7877 }, { "epoch": 0.8540763226366002, "grad_norm": 0.781355082988739, "learning_rate": 1.4358044392862325e-05, "loss": 0.015, "step": 7878 }, { "epoch": 0.8541847354726799, "grad_norm": 2.3394618034362793, "learning_rate": 1.4357319019294937e-05, "loss": 0.0462, "step": 7879 }, { "epoch": 0.8542931483087598, "grad_norm": 0.5726883411407471, "learning_rate": 1.4356593645727551e-05, "loss": 0.0244, "step": 7880 }, { "epoch": 0.8544015611448396, "grad_norm": 4.005661964416504, "learning_rate": 1.4355868272160163e-05, "loss": 0.0439, "step": 7881 }, { "epoch": 0.8545099739809193, "grad_norm": 4.569991111755371, "learning_rate": 1.4355142898592775e-05, "loss": 0.0488, "step": 7882 }, { "epoch": 0.8546183868169991, "grad_norm": 4.999753475189209, "learning_rate": 1.435441752502539e-05, "loss": 0.0342, "step": 7883 }, { "epoch": 0.854726799653079, "grad_norm": 2.3314127922058105, "learning_rate": 1.4353692151458003e-05, "loss": 0.0404, "step": 7884 }, { "epoch": 0.8548352124891587, "grad_norm": 3.6091246604919434, "learning_rate": 1.4352966777890615e-05, "loss": 0.0681, "step": 7885 }, { "epoch": 0.8549436253252385, "grad_norm": 2.791903495788574, "learning_rate": 1.4352241404323229e-05, "loss": 0.038, "step": 7886 }, { "epoch": 0.8550520381613183, "grad_norm": 2.248356342315674, "learning_rate": 1.4351516030755841e-05, "loss": 0.0137, "step": 7887 }, { "epoch": 0.8551604509973981, "grad_norm": 0.7588630318641663, "learning_rate": 1.4350790657188453e-05, "loss": 0.0129, "step": 7888 }, { "epoch": 0.8552688638334779, "grad_norm": 2.0299978256225586, "learning_rate": 1.4350065283621067e-05, "loss": 0.0083, "step": 7889 }, { "epoch": 0.8553772766695577, "grad_norm": 2.329390287399292, "learning_rate": 1.4349339910053679e-05, "loss": 0.0346, "step": 7890 }, { "epoch": 0.8554856895056374, "grad_norm": 2.2844152450561523, "learning_rate": 1.4348614536486291e-05, "loss": 0.0207, "step": 7891 }, { "epoch": 0.8555941023417173, "grad_norm": 2.593461036682129, "learning_rate": 1.4347889162918905e-05, "loss": 0.0405, "step": 7892 }, { "epoch": 0.855702515177797, "grad_norm": 4.699047565460205, "learning_rate": 1.4347163789351517e-05, "loss": 0.0553, "step": 7893 }, { "epoch": 0.8558109280138768, "grad_norm": 0.6554625034332275, "learning_rate": 1.434643841578413e-05, "loss": 0.0119, "step": 7894 }, { "epoch": 0.8559193408499567, "grad_norm": 2.596870183944702, "learning_rate": 1.4345713042216742e-05, "loss": 0.0459, "step": 7895 }, { "epoch": 0.8560277536860365, "grad_norm": 1.9355887174606323, "learning_rate": 1.4344987668649355e-05, "loss": 0.0227, "step": 7896 }, { "epoch": 0.8561361665221162, "grad_norm": 1.056273102760315, "learning_rate": 1.4344262295081968e-05, "loss": 0.0109, "step": 7897 }, { "epoch": 0.856244579358196, "grad_norm": 2.1277992725372314, "learning_rate": 1.434353692151458e-05, "loss": 0.0525, "step": 7898 }, { "epoch": 0.8563529921942759, "grad_norm": 2.7900540828704834, "learning_rate": 1.4342811547947194e-05, "loss": 0.053, "step": 7899 }, { "epoch": 0.8564614050303556, "grad_norm": 3.4091780185699463, "learning_rate": 1.4342086174379807e-05, "loss": 0.0417, "step": 7900 }, { "epoch": 0.8565698178664354, "grad_norm": 0.3131111264228821, "learning_rate": 1.4341360800812421e-05, "loss": 0.0033, "step": 7901 }, { "epoch": 0.8566782307025151, "grad_norm": 2.8424906730651855, "learning_rate": 1.4340635427245033e-05, "loss": 0.0214, "step": 7902 }, { "epoch": 0.856786643538595, "grad_norm": 2.0403482913970947, "learning_rate": 1.4339910053677645e-05, "loss": 0.0336, "step": 7903 }, { "epoch": 0.8568950563746748, "grad_norm": 1.4164621829986572, "learning_rate": 1.433918468011026e-05, "loss": 0.0555, "step": 7904 }, { "epoch": 0.8570034692107545, "grad_norm": 6.3264970779418945, "learning_rate": 1.4338459306542871e-05, "loss": 0.0274, "step": 7905 }, { "epoch": 0.8571118820468343, "grad_norm": 3.870671510696411, "learning_rate": 1.4337733932975484e-05, "loss": 0.0782, "step": 7906 }, { "epoch": 0.8572202948829142, "grad_norm": 4.4419050216674805, "learning_rate": 1.4337008559408096e-05, "loss": 0.0233, "step": 7907 }, { "epoch": 0.8573287077189939, "grad_norm": 0.6380200982093811, "learning_rate": 1.433628318584071e-05, "loss": 0.0146, "step": 7908 }, { "epoch": 0.8574371205550737, "grad_norm": 3.4068524837493896, "learning_rate": 1.4335557812273322e-05, "loss": 0.0646, "step": 7909 }, { "epoch": 0.8575455333911535, "grad_norm": 3.170560359954834, "learning_rate": 1.4334832438705934e-05, "loss": 0.0502, "step": 7910 }, { "epoch": 0.8576539462272333, "grad_norm": 1.7749098539352417, "learning_rate": 1.4334107065138548e-05, "loss": 0.0378, "step": 7911 }, { "epoch": 0.8577623590633131, "grad_norm": 2.5446395874023438, "learning_rate": 1.433338169157116e-05, "loss": 0.0742, "step": 7912 }, { "epoch": 0.8578707718993929, "grad_norm": 8.127424240112305, "learning_rate": 1.4332656318003772e-05, "loss": 0.0534, "step": 7913 }, { "epoch": 0.8579791847354726, "grad_norm": 3.2956156730651855, "learning_rate": 1.4331930944436386e-05, "loss": 0.0407, "step": 7914 }, { "epoch": 0.8580875975715525, "grad_norm": 1.8967548608779907, "learning_rate": 1.4331205570868998e-05, "loss": 0.0225, "step": 7915 }, { "epoch": 0.8581960104076323, "grad_norm": 3.797395706176758, "learning_rate": 1.433048019730161e-05, "loss": 0.0622, "step": 7916 }, { "epoch": 0.858304423243712, "grad_norm": 1.3175208568572998, "learning_rate": 1.4329754823734224e-05, "loss": 0.0087, "step": 7917 }, { "epoch": 0.8584128360797918, "grad_norm": 0.8619345426559448, "learning_rate": 1.4329029450166838e-05, "loss": 0.0155, "step": 7918 }, { "epoch": 0.8585212489158717, "grad_norm": 1.742645025253296, "learning_rate": 1.432830407659945e-05, "loss": 0.0494, "step": 7919 }, { "epoch": 0.8586296617519514, "grad_norm": 2.149244785308838, "learning_rate": 1.4327578703032064e-05, "loss": 0.0353, "step": 7920 }, { "epoch": 0.8587380745880312, "grad_norm": 3.6396257877349854, "learning_rate": 1.4326853329464676e-05, "loss": 0.0596, "step": 7921 }, { "epoch": 0.858846487424111, "grad_norm": 0.8856493830680847, "learning_rate": 1.4326127955897288e-05, "loss": 0.0178, "step": 7922 }, { "epoch": 0.8589549002601908, "grad_norm": 2.003355026245117, "learning_rate": 1.4325402582329902e-05, "loss": 0.0301, "step": 7923 }, { "epoch": 0.8590633130962706, "grad_norm": 1.590441346168518, "learning_rate": 1.4324677208762514e-05, "loss": 0.0372, "step": 7924 }, { "epoch": 0.8591717259323504, "grad_norm": 0.8110508322715759, "learning_rate": 1.4323951835195126e-05, "loss": 0.0147, "step": 7925 }, { "epoch": 0.8592801387684302, "grad_norm": 1.2276794910430908, "learning_rate": 1.432322646162774e-05, "loss": 0.0142, "step": 7926 }, { "epoch": 0.85938855160451, "grad_norm": 2.3790407180786133, "learning_rate": 1.4322501088060352e-05, "loss": 0.0599, "step": 7927 }, { "epoch": 0.8594969644405898, "grad_norm": 2.9520885944366455, "learning_rate": 1.4321775714492964e-05, "loss": 0.0527, "step": 7928 }, { "epoch": 0.8596053772766695, "grad_norm": 0.8095361590385437, "learning_rate": 1.4321050340925578e-05, "loss": 0.01, "step": 7929 }, { "epoch": 0.8597137901127494, "grad_norm": 1.5968719720840454, "learning_rate": 1.432032496735819e-05, "loss": 0.021, "step": 7930 }, { "epoch": 0.8598222029488292, "grad_norm": 2.249164342880249, "learning_rate": 1.4319599593790802e-05, "loss": 0.0408, "step": 7931 }, { "epoch": 0.8599306157849089, "grad_norm": 1.9196239709854126, "learning_rate": 1.4318874220223416e-05, "loss": 0.0137, "step": 7932 }, { "epoch": 0.8600390286209887, "grad_norm": 1.8918339014053345, "learning_rate": 1.4318148846656028e-05, "loss": 0.0176, "step": 7933 }, { "epoch": 0.8601474414570686, "grad_norm": 1.4203602075576782, "learning_rate": 1.431742347308864e-05, "loss": 0.0101, "step": 7934 }, { "epoch": 0.8602558542931483, "grad_norm": 2.1310458183288574, "learning_rate": 1.4316698099521256e-05, "loss": 0.019, "step": 7935 }, { "epoch": 0.8603642671292281, "grad_norm": 2.6311416625976562, "learning_rate": 1.4315972725953868e-05, "loss": 0.0266, "step": 7936 }, { "epoch": 0.8604726799653079, "grad_norm": 2.761718511581421, "learning_rate": 1.431524735238648e-05, "loss": 0.0821, "step": 7937 }, { "epoch": 0.8605810928013877, "grad_norm": 3.4546701908111572, "learning_rate": 1.4314521978819094e-05, "loss": 0.0392, "step": 7938 }, { "epoch": 0.8606895056374675, "grad_norm": 1.5925132036209106, "learning_rate": 1.4313796605251706e-05, "loss": 0.0148, "step": 7939 }, { "epoch": 0.8607979184735473, "grad_norm": 5.723888874053955, "learning_rate": 1.4313071231684319e-05, "loss": 0.0538, "step": 7940 }, { "epoch": 0.860906331309627, "grad_norm": 2.2359471321105957, "learning_rate": 1.4312345858116932e-05, "loss": 0.0292, "step": 7941 }, { "epoch": 0.8610147441457069, "grad_norm": 0.5396251082420349, "learning_rate": 1.4311620484549544e-05, "loss": 0.0117, "step": 7942 }, { "epoch": 0.8611231569817867, "grad_norm": 1.1633650064468384, "learning_rate": 1.4310895110982157e-05, "loss": 0.0104, "step": 7943 }, { "epoch": 0.8612315698178664, "grad_norm": 5.31704044342041, "learning_rate": 1.431016973741477e-05, "loss": 0.1268, "step": 7944 }, { "epoch": 0.8613399826539462, "grad_norm": 1.7268213033676147, "learning_rate": 1.4309444363847383e-05, "loss": 0.0217, "step": 7945 }, { "epoch": 0.861448395490026, "grad_norm": 0.33341288566589355, "learning_rate": 1.4308718990279995e-05, "loss": 0.006, "step": 7946 }, { "epoch": 0.8615568083261058, "grad_norm": 5.00016450881958, "learning_rate": 1.4307993616712607e-05, "loss": 0.0375, "step": 7947 }, { "epoch": 0.8616652211621856, "grad_norm": 1.582535743713379, "learning_rate": 1.430726824314522e-05, "loss": 0.0142, "step": 7948 }, { "epoch": 0.8617736339982653, "grad_norm": 5.927709579467773, "learning_rate": 1.4306542869577833e-05, "loss": 0.0519, "step": 7949 }, { "epoch": 0.8618820468343452, "grad_norm": 2.4111955165863037, "learning_rate": 1.4305817496010445e-05, "loss": 0.063, "step": 7950 }, { "epoch": 0.861990459670425, "grad_norm": 4.711329936981201, "learning_rate": 1.4305092122443059e-05, "loss": 0.1441, "step": 7951 }, { "epoch": 0.8620988725065047, "grad_norm": 0.7541309595108032, "learning_rate": 1.4304366748875673e-05, "loss": 0.0099, "step": 7952 }, { "epoch": 0.8622072853425846, "grad_norm": 4.458071231842041, "learning_rate": 1.4303641375308287e-05, "loss": 0.0359, "step": 7953 }, { "epoch": 0.8623156981786644, "grad_norm": 6.86018705368042, "learning_rate": 1.4302916001740899e-05, "loss": 0.0292, "step": 7954 }, { "epoch": 0.8624241110147441, "grad_norm": 1.305107593536377, "learning_rate": 1.430219062817351e-05, "loss": 0.0219, "step": 7955 }, { "epoch": 0.8625325238508239, "grad_norm": 3.8442769050598145, "learning_rate": 1.4301465254606125e-05, "loss": 0.0483, "step": 7956 }, { "epoch": 0.8626409366869038, "grad_norm": 2.0378997325897217, "learning_rate": 1.4300739881038737e-05, "loss": 0.0057, "step": 7957 }, { "epoch": 0.8627493495229835, "grad_norm": 5.188742637634277, "learning_rate": 1.4300014507471349e-05, "loss": 0.0277, "step": 7958 }, { "epoch": 0.8628577623590633, "grad_norm": 2.3237311840057373, "learning_rate": 1.4299289133903961e-05, "loss": 0.0439, "step": 7959 }, { "epoch": 0.8629661751951431, "grad_norm": 3.8490991592407227, "learning_rate": 1.4298563760336575e-05, "loss": 0.0456, "step": 7960 }, { "epoch": 0.863074588031223, "grad_norm": 5.1846418380737305, "learning_rate": 1.4297838386769187e-05, "loss": 0.1502, "step": 7961 }, { "epoch": 0.8631830008673027, "grad_norm": 2.8211700916290283, "learning_rate": 1.42971130132018e-05, "loss": 0.0774, "step": 7962 }, { "epoch": 0.8632914137033825, "grad_norm": 1.704890251159668, "learning_rate": 1.4296387639634413e-05, "loss": 0.0185, "step": 7963 }, { "epoch": 0.8633998265394622, "grad_norm": 3.775412082672119, "learning_rate": 1.4295662266067025e-05, "loss": 0.0828, "step": 7964 }, { "epoch": 0.8635082393755421, "grad_norm": 5.122465133666992, "learning_rate": 1.4294936892499637e-05, "loss": 0.0638, "step": 7965 }, { "epoch": 0.8636166522116219, "grad_norm": 2.1847729682922363, "learning_rate": 1.4294211518932251e-05, "loss": 0.0229, "step": 7966 }, { "epoch": 0.8637250650477016, "grad_norm": 3.661789655685425, "learning_rate": 1.4293486145364863e-05, "loss": 0.063, "step": 7967 }, { "epoch": 0.8638334778837814, "grad_norm": 6.030697345733643, "learning_rate": 1.4292760771797475e-05, "loss": 0.0511, "step": 7968 }, { "epoch": 0.8639418907198613, "grad_norm": 1.5640172958374023, "learning_rate": 1.429203539823009e-05, "loss": 0.045, "step": 7969 }, { "epoch": 0.864050303555941, "grad_norm": 0.5372598767280579, "learning_rate": 1.4291310024662703e-05, "loss": 0.0125, "step": 7970 }, { "epoch": 0.8641587163920208, "grad_norm": 2.175018310546875, "learning_rate": 1.4290584651095315e-05, "loss": 0.0391, "step": 7971 }, { "epoch": 0.8642671292281006, "grad_norm": 0.7085787653923035, "learning_rate": 1.4289859277527929e-05, "loss": 0.0142, "step": 7972 }, { "epoch": 0.8643755420641804, "grad_norm": 1.5143722295761108, "learning_rate": 1.4289133903960541e-05, "loss": 0.0081, "step": 7973 }, { "epoch": 0.8644839549002602, "grad_norm": 1.3740907907485962, "learning_rate": 1.4288408530393153e-05, "loss": 0.0332, "step": 7974 }, { "epoch": 0.86459236773634, "grad_norm": 1.7942980527877808, "learning_rate": 1.4287683156825767e-05, "loss": 0.0417, "step": 7975 }, { "epoch": 0.8647007805724197, "grad_norm": 1.336540699005127, "learning_rate": 1.428695778325838e-05, "loss": 0.0426, "step": 7976 }, { "epoch": 0.8648091934084996, "grad_norm": 3.2899975776672363, "learning_rate": 1.4286232409690992e-05, "loss": 0.0343, "step": 7977 }, { "epoch": 0.8649176062445794, "grad_norm": 4.324069023132324, "learning_rate": 1.4285507036123605e-05, "loss": 0.0465, "step": 7978 }, { "epoch": 0.8650260190806591, "grad_norm": 0.6216616630554199, "learning_rate": 1.4284781662556218e-05, "loss": 0.0096, "step": 7979 }, { "epoch": 0.8651344319167389, "grad_norm": 0.7912998795509338, "learning_rate": 1.428405628898883e-05, "loss": 0.0384, "step": 7980 }, { "epoch": 0.8652428447528188, "grad_norm": 1.5871801376342773, "learning_rate": 1.4283330915421443e-05, "loss": 0.0105, "step": 7981 }, { "epoch": 0.8653512575888985, "grad_norm": 0.3759511113166809, "learning_rate": 1.4282605541854056e-05, "loss": 0.0132, "step": 7982 }, { "epoch": 0.8654596704249783, "grad_norm": 4.008275508880615, "learning_rate": 1.4281880168286668e-05, "loss": 0.0648, "step": 7983 }, { "epoch": 0.8655680832610582, "grad_norm": 1.0879040956497192, "learning_rate": 1.4281154794719282e-05, "loss": 0.032, "step": 7984 }, { "epoch": 0.8656764960971379, "grad_norm": 2.012477397918701, "learning_rate": 1.4280429421151894e-05, "loss": 0.0478, "step": 7985 }, { "epoch": 0.8657849089332177, "grad_norm": 2.203341007232666, "learning_rate": 1.4279704047584506e-05, "loss": 0.0268, "step": 7986 }, { "epoch": 0.8658933217692975, "grad_norm": 2.3670706748962402, "learning_rate": 1.4278978674017121e-05, "loss": 0.0554, "step": 7987 }, { "epoch": 0.8660017346053773, "grad_norm": 1.2758408784866333, "learning_rate": 1.4278253300449734e-05, "loss": 0.0305, "step": 7988 }, { "epoch": 0.8661101474414571, "grad_norm": 2.0058248043060303, "learning_rate": 1.4277527926882346e-05, "loss": 0.0234, "step": 7989 }, { "epoch": 0.8662185602775369, "grad_norm": 4.832606315612793, "learning_rate": 1.427680255331496e-05, "loss": 0.0318, "step": 7990 }, { "epoch": 0.8663269731136166, "grad_norm": 3.8600268363952637, "learning_rate": 1.4276077179747572e-05, "loss": 0.0823, "step": 7991 }, { "epoch": 0.8664353859496965, "grad_norm": 3.7016544342041016, "learning_rate": 1.4275351806180184e-05, "loss": 0.0659, "step": 7992 }, { "epoch": 0.8665437987857763, "grad_norm": 1.989766001701355, "learning_rate": 1.4274626432612798e-05, "loss": 0.0306, "step": 7993 }, { "epoch": 0.866652211621856, "grad_norm": 1.368395209312439, "learning_rate": 1.427390105904541e-05, "loss": 0.0334, "step": 7994 }, { "epoch": 0.8667606244579358, "grad_norm": 0.4460444152355194, "learning_rate": 1.4273175685478022e-05, "loss": 0.0058, "step": 7995 }, { "epoch": 0.8668690372940157, "grad_norm": 1.7804995775222778, "learning_rate": 1.4272450311910636e-05, "loss": 0.0237, "step": 7996 }, { "epoch": 0.8669774501300954, "grad_norm": 3.0486013889312744, "learning_rate": 1.4271724938343248e-05, "loss": 0.0461, "step": 7997 }, { "epoch": 0.8670858629661752, "grad_norm": 2.51411771774292, "learning_rate": 1.427099956477586e-05, "loss": 0.0525, "step": 7998 }, { "epoch": 0.8671942758022549, "grad_norm": 3.9269397258758545, "learning_rate": 1.4270274191208472e-05, "loss": 0.0471, "step": 7999 }, { "epoch": 0.8673026886383348, "grad_norm": 5.060004234313965, "learning_rate": 1.4269548817641086e-05, "loss": 0.0559, "step": 8000 }, { "epoch": 0.8674111014744146, "grad_norm": 2.3048160076141357, "learning_rate": 1.4268823444073698e-05, "loss": 0.0237, "step": 8001 }, { "epoch": 0.8675195143104943, "grad_norm": 2.971775531768799, "learning_rate": 1.426809807050631e-05, "loss": 0.0677, "step": 8002 }, { "epoch": 0.8676279271465741, "grad_norm": 1.8784558773040771, "learning_rate": 1.4267372696938924e-05, "loss": 0.0219, "step": 8003 }, { "epoch": 0.867736339982654, "grad_norm": 3.4581215381622314, "learning_rate": 1.4266647323371538e-05, "loss": 0.0653, "step": 8004 }, { "epoch": 0.8678447528187337, "grad_norm": 1.6493992805480957, "learning_rate": 1.4265921949804152e-05, "loss": 0.0115, "step": 8005 }, { "epoch": 0.8679531656548135, "grad_norm": 2.956303834915161, "learning_rate": 1.4265196576236764e-05, "loss": 0.049, "step": 8006 }, { "epoch": 0.8680615784908933, "grad_norm": 1.1635135412216187, "learning_rate": 1.4264471202669376e-05, "loss": 0.0091, "step": 8007 }, { "epoch": 0.8681699913269731, "grad_norm": 3.437730312347412, "learning_rate": 1.426374582910199e-05, "loss": 0.0558, "step": 8008 }, { "epoch": 0.8682784041630529, "grad_norm": 1.6971757411956787, "learning_rate": 1.4263020455534602e-05, "loss": 0.032, "step": 8009 }, { "epoch": 0.8683868169991327, "grad_norm": 0.7581056356430054, "learning_rate": 1.4262295081967214e-05, "loss": 0.0076, "step": 8010 }, { "epoch": 0.8684952298352124, "grad_norm": 3.059907913208008, "learning_rate": 1.4261569708399826e-05, "loss": 0.0915, "step": 8011 }, { "epoch": 0.8686036426712923, "grad_norm": 2.6435928344726562, "learning_rate": 1.426084433483244e-05, "loss": 0.0603, "step": 8012 }, { "epoch": 0.8687120555073721, "grad_norm": 4.6530537605285645, "learning_rate": 1.4260118961265052e-05, "loss": 0.0493, "step": 8013 }, { "epoch": 0.8688204683434518, "grad_norm": 1.9499857425689697, "learning_rate": 1.4259393587697665e-05, "loss": 0.0195, "step": 8014 }, { "epoch": 0.8689288811795317, "grad_norm": 0.7740224599838257, "learning_rate": 1.4258668214130278e-05, "loss": 0.0099, "step": 8015 }, { "epoch": 0.8690372940156115, "grad_norm": 2.329242706298828, "learning_rate": 1.425794284056289e-05, "loss": 0.0088, "step": 8016 }, { "epoch": 0.8691457068516912, "grad_norm": 1.5963560342788696, "learning_rate": 1.4257217466995503e-05, "loss": 0.0482, "step": 8017 }, { "epoch": 0.869254119687771, "grad_norm": 1.455012321472168, "learning_rate": 1.4256492093428117e-05, "loss": 0.0282, "step": 8018 }, { "epoch": 0.8693625325238509, "grad_norm": 2.239081621170044, "learning_rate": 1.4255766719860729e-05, "loss": 0.0322, "step": 8019 }, { "epoch": 0.8694709453599306, "grad_norm": 1.812936544418335, "learning_rate": 1.425504134629334e-05, "loss": 0.0473, "step": 8020 }, { "epoch": 0.8695793581960104, "grad_norm": 0.8190416097640991, "learning_rate": 1.4254315972725956e-05, "loss": 0.0149, "step": 8021 }, { "epoch": 0.8696877710320902, "grad_norm": 1.627938985824585, "learning_rate": 1.4253590599158568e-05, "loss": 0.0485, "step": 8022 }, { "epoch": 0.86979618386817, "grad_norm": 0.7005493640899658, "learning_rate": 1.4252865225591182e-05, "loss": 0.0191, "step": 8023 }, { "epoch": 0.8699045967042498, "grad_norm": 2.237460136413574, "learning_rate": 1.4252139852023794e-05, "loss": 0.0473, "step": 8024 }, { "epoch": 0.8700130095403296, "grad_norm": 2.2432243824005127, "learning_rate": 1.4251414478456407e-05, "loss": 0.0203, "step": 8025 }, { "epoch": 0.8701214223764093, "grad_norm": 7.81601095199585, "learning_rate": 1.4250689104889019e-05, "loss": 0.0758, "step": 8026 }, { "epoch": 0.8702298352124892, "grad_norm": 1.4165481328964233, "learning_rate": 1.4249963731321633e-05, "loss": 0.0227, "step": 8027 }, { "epoch": 0.870338248048569, "grad_norm": 1.2642923593521118, "learning_rate": 1.4249238357754245e-05, "loss": 0.0084, "step": 8028 }, { "epoch": 0.8704466608846487, "grad_norm": 3.584392786026001, "learning_rate": 1.4248512984186857e-05, "loss": 0.0372, "step": 8029 }, { "epoch": 0.8705550737207285, "grad_norm": 2.3891139030456543, "learning_rate": 1.424778761061947e-05, "loss": 0.0314, "step": 8030 }, { "epoch": 0.8706634865568084, "grad_norm": 4.206315994262695, "learning_rate": 1.4247062237052083e-05, "loss": 0.0383, "step": 8031 }, { "epoch": 0.8707718993928881, "grad_norm": 0.49154260754585266, "learning_rate": 1.4246336863484695e-05, "loss": 0.0287, "step": 8032 }, { "epoch": 0.8708803122289679, "grad_norm": 0.5824975967407227, "learning_rate": 1.4245611489917309e-05, "loss": 0.0274, "step": 8033 }, { "epoch": 0.8709887250650477, "grad_norm": 1.3049019575119019, "learning_rate": 1.4244886116349921e-05, "loss": 0.0239, "step": 8034 }, { "epoch": 0.8710971379011275, "grad_norm": 0.7505710124969482, "learning_rate": 1.4244160742782533e-05, "loss": 0.0273, "step": 8035 }, { "epoch": 0.8712055507372073, "grad_norm": 6.490268707275391, "learning_rate": 1.4243435369215147e-05, "loss": 0.1034, "step": 8036 }, { "epoch": 0.871313963573287, "grad_norm": 1.2819592952728271, "learning_rate": 1.4242709995647759e-05, "loss": 0.0086, "step": 8037 }, { "epoch": 0.8714223764093668, "grad_norm": 5.21129846572876, "learning_rate": 1.4241984622080371e-05, "loss": 0.0347, "step": 8038 }, { "epoch": 0.8715307892454467, "grad_norm": 1.2269725799560547, "learning_rate": 1.4241259248512987e-05, "loss": 0.0194, "step": 8039 }, { "epoch": 0.8716392020815265, "grad_norm": 0.8150596618652344, "learning_rate": 1.4240533874945599e-05, "loss": 0.019, "step": 8040 }, { "epoch": 0.8717476149176062, "grad_norm": 0.3896048665046692, "learning_rate": 1.4239808501378211e-05, "loss": 0.0025, "step": 8041 }, { "epoch": 0.8718560277536861, "grad_norm": 1.2361735105514526, "learning_rate": 1.4239083127810825e-05, "loss": 0.0181, "step": 8042 }, { "epoch": 0.8719644405897659, "grad_norm": 1.2699921131134033, "learning_rate": 1.4238357754243437e-05, "loss": 0.0335, "step": 8043 }, { "epoch": 0.8720728534258456, "grad_norm": 0.7751911878585815, "learning_rate": 1.423763238067605e-05, "loss": 0.0054, "step": 8044 }, { "epoch": 0.8721812662619254, "grad_norm": 2.622572660446167, "learning_rate": 1.4236907007108663e-05, "loss": 0.0362, "step": 8045 }, { "epoch": 0.8722896790980053, "grad_norm": 3.5080344676971436, "learning_rate": 1.4236181633541275e-05, "loss": 0.0426, "step": 8046 }, { "epoch": 0.872398091934085, "grad_norm": 2.401693820953369, "learning_rate": 1.4235456259973887e-05, "loss": 0.0482, "step": 8047 }, { "epoch": 0.8725065047701648, "grad_norm": 1.830289602279663, "learning_rate": 1.4234730886406501e-05, "loss": 0.0254, "step": 8048 }, { "epoch": 0.8726149176062445, "grad_norm": 2.715035915374756, "learning_rate": 1.4234005512839113e-05, "loss": 0.08, "step": 8049 }, { "epoch": 0.8727233304423244, "grad_norm": 4.826967716217041, "learning_rate": 1.4233280139271725e-05, "loss": 0.0788, "step": 8050 }, { "epoch": 0.8728317432784042, "grad_norm": 0.19969235360622406, "learning_rate": 1.4232554765704338e-05, "loss": 0.0028, "step": 8051 }, { "epoch": 0.872940156114484, "grad_norm": 2.4279541969299316, "learning_rate": 1.4231829392136951e-05, "loss": 0.0441, "step": 8052 }, { "epoch": 0.8730485689505637, "grad_norm": 0.9442045092582703, "learning_rate": 1.4231104018569564e-05, "loss": 0.0294, "step": 8053 }, { "epoch": 0.8731569817866436, "grad_norm": 3.5271406173706055, "learning_rate": 1.4230378645002176e-05, "loss": 0.0448, "step": 8054 }, { "epoch": 0.8732653946227233, "grad_norm": 0.2305496335029602, "learning_rate": 1.422965327143479e-05, "loss": 0.0014, "step": 8055 }, { "epoch": 0.8733738074588031, "grad_norm": 1.8833463191986084, "learning_rate": 1.4228927897867403e-05, "loss": 0.018, "step": 8056 }, { "epoch": 0.8734822202948829, "grad_norm": 1.3195995092391968, "learning_rate": 1.4228202524300017e-05, "loss": 0.023, "step": 8057 }, { "epoch": 0.8735906331309627, "grad_norm": 1.8359928131103516, "learning_rate": 1.422747715073263e-05, "loss": 0.0281, "step": 8058 }, { "epoch": 0.8736990459670425, "grad_norm": 2.491288661956787, "learning_rate": 1.4226751777165241e-05, "loss": 0.0495, "step": 8059 }, { "epoch": 0.8738074588031223, "grad_norm": 7.034435272216797, "learning_rate": 1.4226026403597855e-05, "loss": 0.0643, "step": 8060 }, { "epoch": 0.873915871639202, "grad_norm": 1.069267988204956, "learning_rate": 1.4225301030030467e-05, "loss": 0.0437, "step": 8061 }, { "epoch": 0.8740242844752819, "grad_norm": 2.9873745441436768, "learning_rate": 1.422457565646308e-05, "loss": 0.0341, "step": 8062 }, { "epoch": 0.8741326973113617, "grad_norm": 2.2653675079345703, "learning_rate": 1.4223850282895692e-05, "loss": 0.0367, "step": 8063 }, { "epoch": 0.8742411101474414, "grad_norm": 0.5695919990539551, "learning_rate": 1.4223124909328306e-05, "loss": 0.0121, "step": 8064 }, { "epoch": 0.8743495229835212, "grad_norm": 1.7588367462158203, "learning_rate": 1.4222399535760918e-05, "loss": 0.018, "step": 8065 }, { "epoch": 0.8744579358196011, "grad_norm": 1.9642565250396729, "learning_rate": 1.422167416219353e-05, "loss": 0.0507, "step": 8066 }, { "epoch": 0.8745663486556808, "grad_norm": 3.0161561965942383, "learning_rate": 1.4220948788626144e-05, "loss": 0.0204, "step": 8067 }, { "epoch": 0.8746747614917606, "grad_norm": 2.143346071243286, "learning_rate": 1.4220223415058756e-05, "loss": 0.0126, "step": 8068 }, { "epoch": 0.8747831743278404, "grad_norm": 2.5749459266662598, "learning_rate": 1.4219498041491368e-05, "loss": 0.0582, "step": 8069 }, { "epoch": 0.8748915871639202, "grad_norm": 1.7931797504425049, "learning_rate": 1.4218772667923982e-05, "loss": 0.0276, "step": 8070 }, { "epoch": 0.875, "grad_norm": 3.640681028366089, "learning_rate": 1.4218047294356594e-05, "loss": 0.0508, "step": 8071 }, { "epoch": 0.8751084128360798, "grad_norm": 3.221275568008423, "learning_rate": 1.4217321920789206e-05, "loss": 0.0266, "step": 8072 }, { "epoch": 0.8752168256721596, "grad_norm": 3.9380481243133545, "learning_rate": 1.4216596547221822e-05, "loss": 0.0573, "step": 8073 }, { "epoch": 0.8753252385082394, "grad_norm": 0.6425238847732544, "learning_rate": 1.4215871173654434e-05, "loss": 0.0119, "step": 8074 }, { "epoch": 0.8754336513443192, "grad_norm": 1.1823647022247314, "learning_rate": 1.4215145800087048e-05, "loss": 0.0257, "step": 8075 }, { "epoch": 0.8755420641803989, "grad_norm": 3.980466365814209, "learning_rate": 1.421442042651966e-05, "loss": 0.0756, "step": 8076 }, { "epoch": 0.8756504770164788, "grad_norm": 4.662801742553711, "learning_rate": 1.4213695052952272e-05, "loss": 0.0347, "step": 8077 }, { "epoch": 0.8757588898525586, "grad_norm": 3.085780143737793, "learning_rate": 1.4212969679384884e-05, "loss": 0.0223, "step": 8078 }, { "epoch": 0.8758673026886383, "grad_norm": 2.0305774211883545, "learning_rate": 1.4212244305817498e-05, "loss": 0.0461, "step": 8079 }, { "epoch": 0.8759757155247181, "grad_norm": 0.9853245615959167, "learning_rate": 1.421151893225011e-05, "loss": 0.0329, "step": 8080 }, { "epoch": 0.876084128360798, "grad_norm": 3.1230080127716064, "learning_rate": 1.4210793558682722e-05, "loss": 0.0301, "step": 8081 }, { "epoch": 0.8761925411968777, "grad_norm": 2.145052909851074, "learning_rate": 1.4210068185115336e-05, "loss": 0.0708, "step": 8082 }, { "epoch": 0.8763009540329575, "grad_norm": 2.7465460300445557, "learning_rate": 1.4209342811547948e-05, "loss": 0.0742, "step": 8083 }, { "epoch": 0.8764093668690373, "grad_norm": 1.9304271936416626, "learning_rate": 1.420861743798056e-05, "loss": 0.038, "step": 8084 }, { "epoch": 0.8765177797051171, "grad_norm": 2.5519673824310303, "learning_rate": 1.4207892064413174e-05, "loss": 0.0635, "step": 8085 }, { "epoch": 0.8766261925411969, "grad_norm": 3.217938184738159, "learning_rate": 1.4207166690845786e-05, "loss": 0.0217, "step": 8086 }, { "epoch": 0.8767346053772767, "grad_norm": 3.43426251411438, "learning_rate": 1.4206441317278398e-05, "loss": 0.0283, "step": 8087 }, { "epoch": 0.8768430182133564, "grad_norm": 1.9282466173171997, "learning_rate": 1.4205715943711012e-05, "loss": 0.0492, "step": 8088 }, { "epoch": 0.8769514310494363, "grad_norm": 2.177001953125, "learning_rate": 1.4204990570143624e-05, "loss": 0.0312, "step": 8089 }, { "epoch": 0.877059843885516, "grad_norm": 1.1182564496994019, "learning_rate": 1.4204265196576237e-05, "loss": 0.0114, "step": 8090 }, { "epoch": 0.8771682567215958, "grad_norm": 1.8708008527755737, "learning_rate": 1.4203539823008852e-05, "loss": 0.0148, "step": 8091 }, { "epoch": 0.8772766695576756, "grad_norm": 2.399564266204834, "learning_rate": 1.4202814449441464e-05, "loss": 0.0442, "step": 8092 }, { "epoch": 0.8773850823937555, "grad_norm": 1.14308500289917, "learning_rate": 1.4202089075874076e-05, "loss": 0.0585, "step": 8093 }, { "epoch": 0.8774934952298352, "grad_norm": 1.1186962127685547, "learning_rate": 1.420136370230669e-05, "loss": 0.0172, "step": 8094 }, { "epoch": 0.877601908065915, "grad_norm": 2.08892822265625, "learning_rate": 1.4200638328739302e-05, "loss": 0.0461, "step": 8095 }, { "epoch": 0.8777103209019947, "grad_norm": 1.9151304960250854, "learning_rate": 1.4199912955171915e-05, "loss": 0.0377, "step": 8096 }, { "epoch": 0.8778187337380746, "grad_norm": 3.6979622840881348, "learning_rate": 1.4199187581604528e-05, "loss": 0.0497, "step": 8097 }, { "epoch": 0.8779271465741544, "grad_norm": 1.6974685192108154, "learning_rate": 1.419846220803714e-05, "loss": 0.0127, "step": 8098 }, { "epoch": 0.8780355594102341, "grad_norm": 0.4228705167770386, "learning_rate": 1.4197736834469753e-05, "loss": 0.0124, "step": 8099 }, { "epoch": 0.8781439722463139, "grad_norm": 1.9110758304595947, "learning_rate": 1.4197011460902366e-05, "loss": 0.0429, "step": 8100 }, { "epoch": 0.8782523850823938, "grad_norm": 1.3823661804199219, "learning_rate": 1.4196286087334979e-05, "loss": 0.0172, "step": 8101 }, { "epoch": 0.8783607979184735, "grad_norm": 6.960862636566162, "learning_rate": 1.419556071376759e-05, "loss": 0.0527, "step": 8102 }, { "epoch": 0.8784692107545533, "grad_norm": 1.120564341545105, "learning_rate": 1.4194835340200203e-05, "loss": 0.0542, "step": 8103 }, { "epoch": 0.8785776235906332, "grad_norm": 0.5404911637306213, "learning_rate": 1.4194109966632817e-05, "loss": 0.0148, "step": 8104 }, { "epoch": 0.878686036426713, "grad_norm": 2.8074212074279785, "learning_rate": 1.4193384593065429e-05, "loss": 0.0426, "step": 8105 }, { "epoch": 0.8787944492627927, "grad_norm": 5.115703582763672, "learning_rate": 1.4192659219498041e-05, "loss": 0.0283, "step": 8106 }, { "epoch": 0.8789028620988725, "grad_norm": 0.5363705158233643, "learning_rate": 1.4191933845930655e-05, "loss": 0.0149, "step": 8107 }, { "epoch": 0.8790112749349523, "grad_norm": 0.9008330702781677, "learning_rate": 1.4191208472363269e-05, "loss": 0.0195, "step": 8108 }, { "epoch": 0.8791196877710321, "grad_norm": 2.8013553619384766, "learning_rate": 1.4190483098795883e-05, "loss": 0.0754, "step": 8109 }, { "epoch": 0.8792281006071119, "grad_norm": 1.5205248594284058, "learning_rate": 1.4189757725228495e-05, "loss": 0.0421, "step": 8110 }, { "epoch": 0.8793365134431916, "grad_norm": 2.3401899337768555, "learning_rate": 1.4189032351661107e-05, "loss": 0.0533, "step": 8111 }, { "epoch": 0.8794449262792715, "grad_norm": 1.0174224376678467, "learning_rate": 1.418830697809372e-05, "loss": 0.0139, "step": 8112 }, { "epoch": 0.8795533391153513, "grad_norm": 1.4390273094177246, "learning_rate": 1.4187581604526333e-05, "loss": 0.0123, "step": 8113 }, { "epoch": 0.879661751951431, "grad_norm": 2.9250354766845703, "learning_rate": 1.4186856230958945e-05, "loss": 0.0399, "step": 8114 }, { "epoch": 0.8797701647875108, "grad_norm": 1.209985613822937, "learning_rate": 1.4186130857391557e-05, "loss": 0.0357, "step": 8115 }, { "epoch": 0.8798785776235907, "grad_norm": 0.6963673233985901, "learning_rate": 1.4185405483824171e-05, "loss": 0.0364, "step": 8116 }, { "epoch": 0.8799869904596704, "grad_norm": 1.6751407384872437, "learning_rate": 1.4184680110256783e-05, "loss": 0.0361, "step": 8117 }, { "epoch": 0.8800954032957502, "grad_norm": 4.654283046722412, "learning_rate": 1.4183954736689395e-05, "loss": 0.0733, "step": 8118 }, { "epoch": 0.88020381613183, "grad_norm": 4.600047588348389, "learning_rate": 1.4183229363122009e-05, "loss": 0.0549, "step": 8119 }, { "epoch": 0.8803122289679098, "grad_norm": 2.09578800201416, "learning_rate": 1.4182503989554621e-05, "loss": 0.0463, "step": 8120 }, { "epoch": 0.8804206418039896, "grad_norm": 1.8246562480926514, "learning_rate": 1.4181778615987233e-05, "loss": 0.0251, "step": 8121 }, { "epoch": 0.8805290546400694, "grad_norm": 2.769501209259033, "learning_rate": 1.4181053242419847e-05, "loss": 0.035, "step": 8122 }, { "epoch": 0.8806374674761491, "grad_norm": 3.1662275791168213, "learning_rate": 1.418032786885246e-05, "loss": 0.1045, "step": 8123 }, { "epoch": 0.880745880312229, "grad_norm": 0.9679413437843323, "learning_rate": 1.4179602495285071e-05, "loss": 0.0339, "step": 8124 }, { "epoch": 0.8808542931483088, "grad_norm": 0.624873697757721, "learning_rate": 1.4178877121717687e-05, "loss": 0.0219, "step": 8125 }, { "epoch": 0.8809627059843885, "grad_norm": 4.642431735992432, "learning_rate": 1.4178151748150299e-05, "loss": 0.0551, "step": 8126 }, { "epoch": 0.8810711188204683, "grad_norm": 3.9159843921661377, "learning_rate": 1.4177426374582913e-05, "loss": 0.0453, "step": 8127 }, { "epoch": 0.8811795316565482, "grad_norm": 2.972102165222168, "learning_rate": 1.4176701001015525e-05, "loss": 0.0248, "step": 8128 }, { "epoch": 0.8812879444926279, "grad_norm": 3.1238279342651367, "learning_rate": 1.4175975627448137e-05, "loss": 0.0479, "step": 8129 }, { "epoch": 0.8813963573287077, "grad_norm": 1.5992869138717651, "learning_rate": 1.417525025388075e-05, "loss": 0.0262, "step": 8130 }, { "epoch": 0.8815047701647876, "grad_norm": 3.1938655376434326, "learning_rate": 1.4174524880313363e-05, "loss": 0.0405, "step": 8131 }, { "epoch": 0.8816131830008673, "grad_norm": 0.738466203212738, "learning_rate": 1.4173799506745975e-05, "loss": 0.0193, "step": 8132 }, { "epoch": 0.8817215958369471, "grad_norm": 2.7689194679260254, "learning_rate": 1.4173074133178588e-05, "loss": 0.0531, "step": 8133 }, { "epoch": 0.8818300086730269, "grad_norm": 8.690560340881348, "learning_rate": 1.4172348759611201e-05, "loss": 0.0571, "step": 8134 }, { "epoch": 0.8819384215091067, "grad_norm": 3.577118396759033, "learning_rate": 1.4171623386043814e-05, "loss": 0.0157, "step": 8135 }, { "epoch": 0.8820468343451865, "grad_norm": 2.776097297668457, "learning_rate": 1.4170898012476426e-05, "loss": 0.1083, "step": 8136 }, { "epoch": 0.8821552471812663, "grad_norm": 3.3091328144073486, "learning_rate": 1.417017263890904e-05, "loss": 0.0371, "step": 8137 }, { "epoch": 0.882263660017346, "grad_norm": 0.44500723481178284, "learning_rate": 1.4169447265341652e-05, "loss": 0.0261, "step": 8138 }, { "epoch": 0.8823720728534259, "grad_norm": 4.630529880523682, "learning_rate": 1.4168721891774264e-05, "loss": 0.0568, "step": 8139 }, { "epoch": 0.8824804856895057, "grad_norm": 1.1380705833435059, "learning_rate": 1.4167996518206878e-05, "loss": 0.0286, "step": 8140 }, { "epoch": 0.8825888985255854, "grad_norm": 3.6591176986694336, "learning_rate": 1.416727114463949e-05, "loss": 0.0568, "step": 8141 }, { "epoch": 0.8826973113616652, "grad_norm": 1.7739790678024292, "learning_rate": 1.4166545771072104e-05, "loss": 0.0217, "step": 8142 }, { "epoch": 0.8828057241977451, "grad_norm": 2.0650060176849365, "learning_rate": 1.4165820397504717e-05, "loss": 0.0336, "step": 8143 }, { "epoch": 0.8829141370338248, "grad_norm": 2.415738582611084, "learning_rate": 1.416509502393733e-05, "loss": 0.0108, "step": 8144 }, { "epoch": 0.8830225498699046, "grad_norm": 0.44311270117759705, "learning_rate": 1.4164369650369942e-05, "loss": 0.0143, "step": 8145 }, { "epoch": 0.8831309627059843, "grad_norm": 1.1026086807250977, "learning_rate": 1.4163644276802556e-05, "loss": 0.0144, "step": 8146 }, { "epoch": 0.8832393755420642, "grad_norm": 10.319235801696777, "learning_rate": 1.4162918903235168e-05, "loss": 0.0649, "step": 8147 }, { "epoch": 0.883347788378144, "grad_norm": 1.43741774559021, "learning_rate": 1.416219352966778e-05, "loss": 0.0342, "step": 8148 }, { "epoch": 0.8834562012142237, "grad_norm": 1.1975719928741455, "learning_rate": 1.4161468156100394e-05, "loss": 0.0328, "step": 8149 }, { "epoch": 0.8835646140503035, "grad_norm": 4.215717315673828, "learning_rate": 1.4160742782533006e-05, "loss": 0.0574, "step": 8150 }, { "epoch": 0.8836730268863834, "grad_norm": 2.4665040969848633, "learning_rate": 1.4160017408965618e-05, "loss": 0.0642, "step": 8151 }, { "epoch": 0.8837814397224631, "grad_norm": 1.3733872175216675, "learning_rate": 1.4159292035398232e-05, "loss": 0.0215, "step": 8152 }, { "epoch": 0.8838898525585429, "grad_norm": 1.665300726890564, "learning_rate": 1.4158566661830844e-05, "loss": 0.0249, "step": 8153 }, { "epoch": 0.8839982653946227, "grad_norm": 2.092365026473999, "learning_rate": 1.4157841288263456e-05, "loss": 0.0258, "step": 8154 }, { "epoch": 0.8841066782307025, "grad_norm": 3.0524988174438477, "learning_rate": 1.4157115914696068e-05, "loss": 0.0974, "step": 8155 }, { "epoch": 0.8842150910667823, "grad_norm": 0.49288931488990784, "learning_rate": 1.4156390541128682e-05, "loss": 0.0127, "step": 8156 }, { "epoch": 0.8843235039028621, "grad_norm": 2.7626194953918457, "learning_rate": 1.4155665167561294e-05, "loss": 0.0258, "step": 8157 }, { "epoch": 0.8844319167389418, "grad_norm": 1.7703137397766113, "learning_rate": 1.4154939793993906e-05, "loss": 0.0559, "step": 8158 }, { "epoch": 0.8845403295750217, "grad_norm": 2.757835626602173, "learning_rate": 1.415421442042652e-05, "loss": 0.0611, "step": 8159 }, { "epoch": 0.8846487424111015, "grad_norm": 2.8438222408294678, "learning_rate": 1.4153489046859134e-05, "loss": 0.0625, "step": 8160 }, { "epoch": 0.8847571552471812, "grad_norm": 3.0950920581817627, "learning_rate": 1.4152763673291748e-05, "loss": 0.0245, "step": 8161 }, { "epoch": 0.8848655680832611, "grad_norm": 2.697099208831787, "learning_rate": 1.415203829972436e-05, "loss": 0.0229, "step": 8162 }, { "epoch": 0.8849739809193409, "grad_norm": 0.9097659587860107, "learning_rate": 1.4151312926156972e-05, "loss": 0.0201, "step": 8163 }, { "epoch": 0.8850823937554206, "grad_norm": 1.0622237920761108, "learning_rate": 1.4150587552589586e-05, "loss": 0.0139, "step": 8164 }, { "epoch": 0.8851908065915004, "grad_norm": 5.4586310386657715, "learning_rate": 1.4149862179022198e-05, "loss": 0.0244, "step": 8165 }, { "epoch": 0.8852992194275803, "grad_norm": 1.462203860282898, "learning_rate": 1.414913680545481e-05, "loss": 0.0345, "step": 8166 }, { "epoch": 0.88540763226366, "grad_norm": 2.7485620975494385, "learning_rate": 1.4148411431887422e-05, "loss": 0.04, "step": 8167 }, { "epoch": 0.8855160450997398, "grad_norm": 5.254604816436768, "learning_rate": 1.4147686058320036e-05, "loss": 0.0394, "step": 8168 }, { "epoch": 0.8856244579358196, "grad_norm": 1.6458263397216797, "learning_rate": 1.4146960684752648e-05, "loss": 0.029, "step": 8169 }, { "epoch": 0.8857328707718994, "grad_norm": 2.643801689147949, "learning_rate": 1.414623531118526e-05, "loss": 0.0502, "step": 8170 }, { "epoch": 0.8858412836079792, "grad_norm": 2.248610258102417, "learning_rate": 1.4145509937617874e-05, "loss": 0.0219, "step": 8171 }, { "epoch": 0.885949696444059, "grad_norm": 3.5102107524871826, "learning_rate": 1.4144784564050487e-05, "loss": 0.0478, "step": 8172 }, { "epoch": 0.8860581092801387, "grad_norm": 1.7504301071166992, "learning_rate": 1.4144059190483099e-05, "loss": 0.0297, "step": 8173 }, { "epoch": 0.8861665221162186, "grad_norm": 1.7859188318252563, "learning_rate": 1.4143333816915713e-05, "loss": 0.0533, "step": 8174 }, { "epoch": 0.8862749349522984, "grad_norm": 3.5384042263031006, "learning_rate": 1.4142608443348325e-05, "loss": 0.0339, "step": 8175 }, { "epoch": 0.8863833477883781, "grad_norm": 3.471513509750366, "learning_rate": 1.4141883069780937e-05, "loss": 0.0807, "step": 8176 }, { "epoch": 0.8864917606244579, "grad_norm": 1.6591006517410278, "learning_rate": 1.4141157696213552e-05, "loss": 0.0417, "step": 8177 }, { "epoch": 0.8866001734605378, "grad_norm": 2.5835275650024414, "learning_rate": 1.4140432322646164e-05, "loss": 0.0259, "step": 8178 }, { "epoch": 0.8867085862966175, "grad_norm": 1.053515076637268, "learning_rate": 1.4139706949078778e-05, "loss": 0.0221, "step": 8179 }, { "epoch": 0.8868169991326973, "grad_norm": 3.9805381298065186, "learning_rate": 1.413898157551139e-05, "loss": 0.0375, "step": 8180 }, { "epoch": 0.886925411968777, "grad_norm": 1.4047681093215942, "learning_rate": 1.4138256201944003e-05, "loss": 0.0162, "step": 8181 }, { "epoch": 0.8870338248048569, "grad_norm": 2.283818483352661, "learning_rate": 1.4137530828376615e-05, "loss": 0.0332, "step": 8182 }, { "epoch": 0.8871422376409367, "grad_norm": 3.2067923545837402, "learning_rate": 1.4136805454809229e-05, "loss": 0.0536, "step": 8183 }, { "epoch": 0.8872506504770165, "grad_norm": 2.9882493019104004, "learning_rate": 1.413608008124184e-05, "loss": 0.016, "step": 8184 }, { "epoch": 0.8873590633130962, "grad_norm": 0.8794456124305725, "learning_rate": 1.4135354707674453e-05, "loss": 0.0214, "step": 8185 }, { "epoch": 0.8874674761491761, "grad_norm": 1.4080965518951416, "learning_rate": 1.4134629334107067e-05, "loss": 0.0203, "step": 8186 }, { "epoch": 0.8875758889852559, "grad_norm": 2.298433542251587, "learning_rate": 1.4133903960539679e-05, "loss": 0.0387, "step": 8187 }, { "epoch": 0.8876843018213356, "grad_norm": 0.6520380973815918, "learning_rate": 1.4133178586972291e-05, "loss": 0.0209, "step": 8188 }, { "epoch": 0.8877927146574154, "grad_norm": 3.1448781490325928, "learning_rate": 1.4132453213404905e-05, "loss": 0.0707, "step": 8189 }, { "epoch": 0.8879011274934953, "grad_norm": 1.190123438835144, "learning_rate": 1.4131727839837517e-05, "loss": 0.0278, "step": 8190 }, { "epoch": 0.888009540329575, "grad_norm": 2.6515040397644043, "learning_rate": 1.4131002466270129e-05, "loss": 0.0519, "step": 8191 }, { "epoch": 0.8881179531656548, "grad_norm": 1.9985023736953735, "learning_rate": 1.4130277092702743e-05, "loss": 0.0583, "step": 8192 }, { "epoch": 0.8882263660017347, "grad_norm": 4.987871170043945, "learning_rate": 1.4129551719135355e-05, "loss": 0.0683, "step": 8193 }, { "epoch": 0.8883347788378144, "grad_norm": 2.680427312850952, "learning_rate": 1.4128826345567969e-05, "loss": 0.0279, "step": 8194 }, { "epoch": 0.8884431916738942, "grad_norm": 2.7657017707824707, "learning_rate": 1.4128100972000583e-05, "loss": 0.0317, "step": 8195 }, { "epoch": 0.888551604509974, "grad_norm": 4.572019577026367, "learning_rate": 1.4127375598433195e-05, "loss": 0.0682, "step": 8196 }, { "epoch": 0.8886600173460538, "grad_norm": 1.0716297626495361, "learning_rate": 1.4126650224865807e-05, "loss": 0.0255, "step": 8197 }, { "epoch": 0.8887684301821336, "grad_norm": 1.2891583442687988, "learning_rate": 1.4125924851298421e-05, "loss": 0.0364, "step": 8198 }, { "epoch": 0.8888768430182133, "grad_norm": 0.8252458572387695, "learning_rate": 1.4125199477731033e-05, "loss": 0.0176, "step": 8199 }, { "epoch": 0.8889852558542931, "grad_norm": 3.4785282611846924, "learning_rate": 1.4124474104163645e-05, "loss": 0.0731, "step": 8200 }, { "epoch": 0.889093668690373, "grad_norm": 3.142012596130371, "learning_rate": 1.4123748730596259e-05, "loss": 0.0193, "step": 8201 }, { "epoch": 0.8892020815264527, "grad_norm": 2.93003511428833, "learning_rate": 1.4123023357028871e-05, "loss": 0.0532, "step": 8202 }, { "epoch": 0.8893104943625325, "grad_norm": 0.8463981747627258, "learning_rate": 1.4122297983461483e-05, "loss": 0.0113, "step": 8203 }, { "epoch": 0.8894189071986123, "grad_norm": 2.513619899749756, "learning_rate": 1.4121572609894097e-05, "loss": 0.0271, "step": 8204 }, { "epoch": 0.8895273200346921, "grad_norm": 2.8790390491485596, "learning_rate": 1.412084723632671e-05, "loss": 0.0552, "step": 8205 }, { "epoch": 0.8896357328707719, "grad_norm": 4.913691997528076, "learning_rate": 1.4120121862759321e-05, "loss": 0.0484, "step": 8206 }, { "epoch": 0.8897441457068517, "grad_norm": 3.3248043060302734, "learning_rate": 1.4119396489191934e-05, "loss": 0.0325, "step": 8207 }, { "epoch": 0.8898525585429314, "grad_norm": 2.2952072620391846, "learning_rate": 1.4118671115624547e-05, "loss": 0.0241, "step": 8208 }, { "epoch": 0.8899609713790113, "grad_norm": 11.659014701843262, "learning_rate": 1.411794574205716e-05, "loss": 0.1446, "step": 8209 }, { "epoch": 0.8900693842150911, "grad_norm": 1.503167748451233, "learning_rate": 1.4117220368489772e-05, "loss": 0.0407, "step": 8210 }, { "epoch": 0.8901777970511708, "grad_norm": 2.2726433277130127, "learning_rate": 1.4116494994922387e-05, "loss": 0.0456, "step": 8211 }, { "epoch": 0.8902862098872506, "grad_norm": 4.776848793029785, "learning_rate": 1.4115769621355e-05, "loss": 0.0262, "step": 8212 }, { "epoch": 0.8903946227233305, "grad_norm": 1.4755703210830688, "learning_rate": 1.4115044247787613e-05, "loss": 0.0294, "step": 8213 }, { "epoch": 0.8905030355594102, "grad_norm": 2.819031238555908, "learning_rate": 1.4114318874220225e-05, "loss": 0.0624, "step": 8214 }, { "epoch": 0.89061144839549, "grad_norm": 0.7291708588600159, "learning_rate": 1.4113593500652838e-05, "loss": 0.031, "step": 8215 }, { "epoch": 0.8907198612315698, "grad_norm": 3.5829432010650635, "learning_rate": 1.4112868127085451e-05, "loss": 0.1274, "step": 8216 }, { "epoch": 0.8908282740676496, "grad_norm": 1.9352926015853882, "learning_rate": 1.4112142753518063e-05, "loss": 0.0285, "step": 8217 }, { "epoch": 0.8909366869037294, "grad_norm": 1.9766550064086914, "learning_rate": 1.4111417379950676e-05, "loss": 0.0281, "step": 8218 }, { "epoch": 0.8910450997398092, "grad_norm": 1.5799757242202759, "learning_rate": 1.4110692006383288e-05, "loss": 0.033, "step": 8219 }, { "epoch": 0.891153512575889, "grad_norm": 2.4868505001068115, "learning_rate": 1.4109966632815902e-05, "loss": 0.0191, "step": 8220 }, { "epoch": 0.8912619254119688, "grad_norm": 1.6776032447814941, "learning_rate": 1.4109241259248514e-05, "loss": 0.0561, "step": 8221 }, { "epoch": 0.8913703382480486, "grad_norm": 2.8472018241882324, "learning_rate": 1.4108515885681126e-05, "loss": 0.0508, "step": 8222 }, { "epoch": 0.8914787510841283, "grad_norm": 0.4993368685245514, "learning_rate": 1.410779051211374e-05, "loss": 0.0121, "step": 8223 }, { "epoch": 0.8915871639202082, "grad_norm": 0.9787811636924744, "learning_rate": 1.4107065138546352e-05, "loss": 0.0178, "step": 8224 }, { "epoch": 0.891695576756288, "grad_norm": 1.4040212631225586, "learning_rate": 1.4106339764978964e-05, "loss": 0.0307, "step": 8225 }, { "epoch": 0.8918039895923677, "grad_norm": 2.1403815746307373, "learning_rate": 1.4105614391411578e-05, "loss": 0.0366, "step": 8226 }, { "epoch": 0.8919124024284475, "grad_norm": 3.7483015060424805, "learning_rate": 1.410488901784419e-05, "loss": 0.0818, "step": 8227 }, { "epoch": 0.8920208152645274, "grad_norm": 0.14754325151443481, "learning_rate": 1.4104163644276802e-05, "loss": 0.0026, "step": 8228 }, { "epoch": 0.8921292281006071, "grad_norm": 4.187467098236084, "learning_rate": 1.4103438270709418e-05, "loss": 0.0546, "step": 8229 }, { "epoch": 0.8922376409366869, "grad_norm": 2.1822669506073, "learning_rate": 1.410271289714203e-05, "loss": 0.0309, "step": 8230 }, { "epoch": 0.8923460537727667, "grad_norm": 1.9461897611618042, "learning_rate": 1.4101987523574644e-05, "loss": 0.0462, "step": 8231 }, { "epoch": 0.8924544666088465, "grad_norm": 1.7879347801208496, "learning_rate": 1.4101262150007256e-05, "loss": 0.0425, "step": 8232 }, { "epoch": 0.8925628794449263, "grad_norm": 3.2766246795654297, "learning_rate": 1.4100536776439868e-05, "loss": 0.0566, "step": 8233 }, { "epoch": 0.8926712922810061, "grad_norm": 1.7306431531906128, "learning_rate": 1.409981140287248e-05, "loss": 0.0751, "step": 8234 }, { "epoch": 0.8927797051170858, "grad_norm": 2.858586311340332, "learning_rate": 1.4099086029305094e-05, "loss": 0.052, "step": 8235 }, { "epoch": 0.8928881179531657, "grad_norm": 2.391082286834717, "learning_rate": 1.4098360655737706e-05, "loss": 0.027, "step": 8236 }, { "epoch": 0.8929965307892455, "grad_norm": 1.4942259788513184, "learning_rate": 1.4097635282170318e-05, "loss": 0.0402, "step": 8237 }, { "epoch": 0.8931049436253252, "grad_norm": 2.0679478645324707, "learning_rate": 1.4096909908602932e-05, "loss": 0.0397, "step": 8238 }, { "epoch": 0.893213356461405, "grad_norm": 2.854905843734741, "learning_rate": 1.4096184535035544e-05, "loss": 0.0319, "step": 8239 }, { "epoch": 0.8933217692974849, "grad_norm": 1.0247361660003662, "learning_rate": 1.4095459161468156e-05, "loss": 0.018, "step": 8240 }, { "epoch": 0.8934301821335646, "grad_norm": 0.7335030436515808, "learning_rate": 1.409473378790077e-05, "loss": 0.0251, "step": 8241 }, { "epoch": 0.8935385949696444, "grad_norm": 2.263826608657837, "learning_rate": 1.4094008414333382e-05, "loss": 0.0463, "step": 8242 }, { "epoch": 0.8936470078057241, "grad_norm": 1.901896357536316, "learning_rate": 1.4093283040765994e-05, "loss": 0.0533, "step": 8243 }, { "epoch": 0.893755420641804, "grad_norm": 1.05595064163208, "learning_rate": 1.4092557667198608e-05, "loss": 0.0132, "step": 8244 }, { "epoch": 0.8938638334778838, "grad_norm": 1.8256242275238037, "learning_rate": 1.409183229363122e-05, "loss": 0.021, "step": 8245 }, { "epoch": 0.8939722463139635, "grad_norm": 3.4419305324554443, "learning_rate": 1.4091106920063834e-05, "loss": 0.0666, "step": 8246 }, { "epoch": 0.8940806591500433, "grad_norm": 0.6310603618621826, "learning_rate": 1.4090381546496448e-05, "loss": 0.0142, "step": 8247 }, { "epoch": 0.8941890719861232, "grad_norm": 1.4760961532592773, "learning_rate": 1.408965617292906e-05, "loss": 0.0141, "step": 8248 }, { "epoch": 0.894297484822203, "grad_norm": 3.164740562438965, "learning_rate": 1.4088930799361672e-05, "loss": 0.0609, "step": 8249 }, { "epoch": 0.8944058976582827, "grad_norm": 4.081722736358643, "learning_rate": 1.4088205425794286e-05, "loss": 0.0446, "step": 8250 }, { "epoch": 0.8945143104943626, "grad_norm": 5.539366722106934, "learning_rate": 1.4087480052226898e-05, "loss": 0.0996, "step": 8251 }, { "epoch": 0.8946227233304423, "grad_norm": 2.071413278579712, "learning_rate": 1.408675467865951e-05, "loss": 0.0248, "step": 8252 }, { "epoch": 0.8947311361665221, "grad_norm": 1.490682601928711, "learning_rate": 1.4086029305092124e-05, "loss": 0.0395, "step": 8253 }, { "epoch": 0.8948395490026019, "grad_norm": 1.0211830139160156, "learning_rate": 1.4085303931524737e-05, "loss": 0.0205, "step": 8254 }, { "epoch": 0.8949479618386817, "grad_norm": 2.0914831161499023, "learning_rate": 1.4084578557957349e-05, "loss": 0.0315, "step": 8255 }, { "epoch": 0.8950563746747615, "grad_norm": 3.08935284614563, "learning_rate": 1.4083853184389962e-05, "loss": 0.0137, "step": 8256 }, { "epoch": 0.8951647875108413, "grad_norm": 1.6613733768463135, "learning_rate": 1.4083127810822575e-05, "loss": 0.02, "step": 8257 }, { "epoch": 0.895273200346921, "grad_norm": 2.199171304702759, "learning_rate": 1.4082402437255187e-05, "loss": 0.0298, "step": 8258 }, { "epoch": 0.8953816131830009, "grad_norm": 1.404603123664856, "learning_rate": 1.4081677063687799e-05, "loss": 0.0314, "step": 8259 }, { "epoch": 0.8954900260190807, "grad_norm": 1.1128959655761719, "learning_rate": 1.4080951690120413e-05, "loss": 0.0093, "step": 8260 }, { "epoch": 0.8955984388551604, "grad_norm": 1.4266358613967896, "learning_rate": 1.4080226316553025e-05, "loss": 0.0119, "step": 8261 }, { "epoch": 0.8957068516912402, "grad_norm": 1.7530649900436401, "learning_rate": 1.4079500942985637e-05, "loss": 0.0297, "step": 8262 }, { "epoch": 0.8958152645273201, "grad_norm": 4.236114025115967, "learning_rate": 1.4078775569418253e-05, "loss": 0.0356, "step": 8263 }, { "epoch": 0.8959236773633998, "grad_norm": 2.4083847999572754, "learning_rate": 1.4078050195850865e-05, "loss": 0.0606, "step": 8264 }, { "epoch": 0.8960320901994796, "grad_norm": 3.898621082305908, "learning_rate": 1.4077324822283479e-05, "loss": 0.0366, "step": 8265 }, { "epoch": 0.8961405030355594, "grad_norm": 2.8021318912506104, "learning_rate": 1.407659944871609e-05, "loss": 0.0243, "step": 8266 }, { "epoch": 0.8962489158716392, "grad_norm": 3.6293580532073975, "learning_rate": 1.4075874075148703e-05, "loss": 0.0428, "step": 8267 }, { "epoch": 0.896357328707719, "grad_norm": 1.3265424966812134, "learning_rate": 1.4075148701581317e-05, "loss": 0.0118, "step": 8268 }, { "epoch": 0.8964657415437988, "grad_norm": 1.5061935186386108, "learning_rate": 1.4074423328013929e-05, "loss": 0.0077, "step": 8269 }, { "epoch": 0.8965741543798785, "grad_norm": 5.188292503356934, "learning_rate": 1.4073697954446541e-05, "loss": 0.0341, "step": 8270 }, { "epoch": 0.8966825672159584, "grad_norm": 1.7237130403518677, "learning_rate": 1.4072972580879155e-05, "loss": 0.0207, "step": 8271 }, { "epoch": 0.8967909800520382, "grad_norm": 3.708582878112793, "learning_rate": 1.4072247207311767e-05, "loss": 0.046, "step": 8272 }, { "epoch": 0.8968993928881179, "grad_norm": 3.8510520458221436, "learning_rate": 1.4071521833744379e-05, "loss": 0.11, "step": 8273 }, { "epoch": 0.8970078057241977, "grad_norm": 0.5057426691055298, "learning_rate": 1.4070796460176991e-05, "loss": 0.0178, "step": 8274 }, { "epoch": 0.8971162185602776, "grad_norm": 1.5914503335952759, "learning_rate": 1.4070071086609605e-05, "loss": 0.0273, "step": 8275 }, { "epoch": 0.8972246313963573, "grad_norm": 1.8696620464324951, "learning_rate": 1.4069345713042217e-05, "loss": 0.0409, "step": 8276 }, { "epoch": 0.8973330442324371, "grad_norm": 0.6461164355278015, "learning_rate": 1.406862033947483e-05, "loss": 0.0302, "step": 8277 }, { "epoch": 0.897441457068517, "grad_norm": 1.153374195098877, "learning_rate": 1.4067894965907443e-05, "loss": 0.0202, "step": 8278 }, { "epoch": 0.8975498699045967, "grad_norm": 1.7034845352172852, "learning_rate": 1.4067169592340055e-05, "loss": 0.055, "step": 8279 }, { "epoch": 0.8976582827406765, "grad_norm": 1.7474838495254517, "learning_rate": 1.4066444218772667e-05, "loss": 0.0428, "step": 8280 }, { "epoch": 0.8977666955767563, "grad_norm": 4.293319225311279, "learning_rate": 1.4065718845205283e-05, "loss": 0.0713, "step": 8281 }, { "epoch": 0.8978751084128361, "grad_norm": 1.7515738010406494, "learning_rate": 1.4064993471637895e-05, "loss": 0.0291, "step": 8282 }, { "epoch": 0.8979835212489159, "grad_norm": 4.646170139312744, "learning_rate": 1.4064268098070509e-05, "loss": 0.0445, "step": 8283 }, { "epoch": 0.8980919340849957, "grad_norm": 1.4154895544052124, "learning_rate": 1.4063542724503121e-05, "loss": 0.0203, "step": 8284 }, { "epoch": 0.8982003469210754, "grad_norm": 1.7459315061569214, "learning_rate": 1.4062817350935733e-05, "loss": 0.0278, "step": 8285 }, { "epoch": 0.8983087597571553, "grad_norm": 1.8133418560028076, "learning_rate": 1.4062091977368345e-05, "loss": 0.0168, "step": 8286 }, { "epoch": 0.8984171725932351, "grad_norm": 3.0659902095794678, "learning_rate": 1.406136660380096e-05, "loss": 0.0419, "step": 8287 }, { "epoch": 0.8985255854293148, "grad_norm": 1.8204962015151978, "learning_rate": 1.4060641230233571e-05, "loss": 0.0221, "step": 8288 }, { "epoch": 0.8986339982653946, "grad_norm": 2.3422276973724365, "learning_rate": 1.4059915856666184e-05, "loss": 0.0303, "step": 8289 }, { "epoch": 0.8987424111014745, "grad_norm": 4.650077819824219, "learning_rate": 1.4059190483098797e-05, "loss": 0.0286, "step": 8290 }, { "epoch": 0.8988508239375542, "grad_norm": 3.146296262741089, "learning_rate": 1.405846510953141e-05, "loss": 0.0436, "step": 8291 }, { "epoch": 0.898959236773634, "grad_norm": 2.7499873638153076, "learning_rate": 1.4057739735964022e-05, "loss": 0.0171, "step": 8292 }, { "epoch": 0.8990676496097137, "grad_norm": 0.4291859567165375, "learning_rate": 1.4057014362396636e-05, "loss": 0.0105, "step": 8293 }, { "epoch": 0.8991760624457936, "grad_norm": 1.2345715761184692, "learning_rate": 1.4056288988829248e-05, "loss": 0.0243, "step": 8294 }, { "epoch": 0.8992844752818734, "grad_norm": 5.369596481323242, "learning_rate": 1.405556361526186e-05, "loss": 0.0393, "step": 8295 }, { "epoch": 0.8993928881179531, "grad_norm": 1.4823453426361084, "learning_rate": 1.4054838241694474e-05, "loss": 0.0534, "step": 8296 }, { "epoch": 0.8995013009540329, "grad_norm": 2.4487626552581787, "learning_rate": 1.4054112868127086e-05, "loss": 0.0367, "step": 8297 }, { "epoch": 0.8996097137901128, "grad_norm": 0.5960486531257629, "learning_rate": 1.40533874945597e-05, "loss": 0.0063, "step": 8298 }, { "epoch": 0.8997181266261925, "grad_norm": 0.9043692350387573, "learning_rate": 1.4052662120992313e-05, "loss": 0.0148, "step": 8299 }, { "epoch": 0.8998265394622723, "grad_norm": 2.49131178855896, "learning_rate": 1.4051936747424926e-05, "loss": 0.0518, "step": 8300 }, { "epoch": 0.8999349522983521, "grad_norm": 1.8373241424560547, "learning_rate": 1.4051211373857538e-05, "loss": 0.0449, "step": 8301 }, { "epoch": 0.900043365134432, "grad_norm": 2.2400155067443848, "learning_rate": 1.4050486000290152e-05, "loss": 0.0217, "step": 8302 }, { "epoch": 0.9001517779705117, "grad_norm": 5.90880012512207, "learning_rate": 1.4049760626722764e-05, "loss": 0.0326, "step": 8303 }, { "epoch": 0.9002601908065915, "grad_norm": 3.8039650917053223, "learning_rate": 1.4049035253155376e-05, "loss": 0.0537, "step": 8304 }, { "epoch": 0.9003686036426712, "grad_norm": 2.8772530555725098, "learning_rate": 1.404830987958799e-05, "loss": 0.0354, "step": 8305 }, { "epoch": 0.9004770164787511, "grad_norm": 1.7136329412460327, "learning_rate": 1.4047584506020602e-05, "loss": 0.0097, "step": 8306 }, { "epoch": 0.9005854293148309, "grad_norm": 0.9703714847564697, "learning_rate": 1.4046859132453214e-05, "loss": 0.0308, "step": 8307 }, { "epoch": 0.9006938421509106, "grad_norm": 2.049757957458496, "learning_rate": 1.4046133758885828e-05, "loss": 0.0407, "step": 8308 }, { "epoch": 0.9008022549869905, "grad_norm": 0.36981508135795593, "learning_rate": 1.404540838531844e-05, "loss": 0.0039, "step": 8309 }, { "epoch": 0.9009106678230703, "grad_norm": 3.149810314178467, "learning_rate": 1.4044683011751052e-05, "loss": 0.0466, "step": 8310 }, { "epoch": 0.90101908065915, "grad_norm": 1.7830129861831665, "learning_rate": 1.4043957638183664e-05, "loss": 0.0216, "step": 8311 }, { "epoch": 0.9011274934952298, "grad_norm": 2.262889862060547, "learning_rate": 1.4043232264616278e-05, "loss": 0.0416, "step": 8312 }, { "epoch": 0.9012359063313097, "grad_norm": 0.3389241397380829, "learning_rate": 1.404250689104889e-05, "loss": 0.005, "step": 8313 }, { "epoch": 0.9013443191673894, "grad_norm": 0.9514628648757935, "learning_rate": 1.4041781517481502e-05, "loss": 0.0055, "step": 8314 }, { "epoch": 0.9014527320034692, "grad_norm": 1.2249242067337036, "learning_rate": 1.4041056143914118e-05, "loss": 0.0136, "step": 8315 }, { "epoch": 0.901561144839549, "grad_norm": 0.4531153440475464, "learning_rate": 1.404033077034673e-05, "loss": 0.0054, "step": 8316 }, { "epoch": 0.9016695576756288, "grad_norm": 0.8813421726226807, "learning_rate": 1.4039605396779344e-05, "loss": 0.0261, "step": 8317 }, { "epoch": 0.9017779705117086, "grad_norm": 1.2532154321670532, "learning_rate": 1.4038880023211956e-05, "loss": 0.031, "step": 8318 }, { "epoch": 0.9018863833477884, "grad_norm": 0.514251172542572, "learning_rate": 1.4038154649644568e-05, "loss": 0.0077, "step": 8319 }, { "epoch": 0.9019947961838681, "grad_norm": 4.643055438995361, "learning_rate": 1.4037429276077182e-05, "loss": 0.0348, "step": 8320 }, { "epoch": 0.902103209019948, "grad_norm": 1.066137671470642, "learning_rate": 1.4036703902509794e-05, "loss": 0.0171, "step": 8321 }, { "epoch": 0.9022116218560278, "grad_norm": 1.713767170906067, "learning_rate": 1.4035978528942406e-05, "loss": 0.0303, "step": 8322 }, { "epoch": 0.9023200346921075, "grad_norm": 2.1490304470062256, "learning_rate": 1.403525315537502e-05, "loss": 0.0836, "step": 8323 }, { "epoch": 0.9024284475281873, "grad_norm": 6.372574806213379, "learning_rate": 1.4034527781807632e-05, "loss": 0.0367, "step": 8324 }, { "epoch": 0.9025368603642672, "grad_norm": 1.71448814868927, "learning_rate": 1.4033802408240244e-05, "loss": 0.0195, "step": 8325 }, { "epoch": 0.9026452732003469, "grad_norm": 1.0940574407577515, "learning_rate": 1.4033077034672857e-05, "loss": 0.0136, "step": 8326 }, { "epoch": 0.9027536860364267, "grad_norm": 3.237929344177246, "learning_rate": 1.403235166110547e-05, "loss": 0.0284, "step": 8327 }, { "epoch": 0.9028620988725065, "grad_norm": 3.4153683185577393, "learning_rate": 1.4031626287538083e-05, "loss": 0.069, "step": 8328 }, { "epoch": 0.9029705117085863, "grad_norm": 9.361586570739746, "learning_rate": 1.4030900913970695e-05, "loss": 0.0202, "step": 8329 }, { "epoch": 0.9030789245446661, "grad_norm": 4.152696132659912, "learning_rate": 1.4030175540403309e-05, "loss": 0.0981, "step": 8330 }, { "epoch": 0.9031873373807459, "grad_norm": 6.9471516609191895, "learning_rate": 1.402945016683592e-05, "loss": 0.0693, "step": 8331 }, { "epoch": 0.9032957502168256, "grad_norm": 0.9601320028305054, "learning_rate": 1.4028724793268536e-05, "loss": 0.0231, "step": 8332 }, { "epoch": 0.9034041630529055, "grad_norm": 1.5680257081985474, "learning_rate": 1.4027999419701148e-05, "loss": 0.0238, "step": 8333 }, { "epoch": 0.9035125758889853, "grad_norm": 1.024437427520752, "learning_rate": 1.402727404613376e-05, "loss": 0.0067, "step": 8334 }, { "epoch": 0.903620988725065, "grad_norm": 2.6371750831604004, "learning_rate": 1.4026548672566374e-05, "loss": 0.0262, "step": 8335 }, { "epoch": 0.9037294015611448, "grad_norm": 2.9808642864227295, "learning_rate": 1.4025823298998986e-05, "loss": 0.0923, "step": 8336 }, { "epoch": 0.9038378143972247, "grad_norm": 2.1549975872039795, "learning_rate": 1.4025097925431599e-05, "loss": 0.0142, "step": 8337 }, { "epoch": 0.9039462272333044, "grad_norm": 3.632037401199341, "learning_rate": 1.402437255186421e-05, "loss": 0.0459, "step": 8338 }, { "epoch": 0.9040546400693842, "grad_norm": 3.083470344543457, "learning_rate": 1.4023647178296825e-05, "loss": 0.0185, "step": 8339 }, { "epoch": 0.9041630529054641, "grad_norm": 4.07413911819458, "learning_rate": 1.4022921804729437e-05, "loss": 0.0841, "step": 8340 }, { "epoch": 0.9042714657415438, "grad_norm": 6.894989967346191, "learning_rate": 1.4022196431162049e-05, "loss": 0.0571, "step": 8341 }, { "epoch": 0.9043798785776236, "grad_norm": 4.463385581970215, "learning_rate": 1.4021471057594663e-05, "loss": 0.0668, "step": 8342 }, { "epoch": 0.9044882914137033, "grad_norm": 1.0840801000595093, "learning_rate": 1.4020745684027275e-05, "loss": 0.0088, "step": 8343 }, { "epoch": 0.9045967042497832, "grad_norm": 1.1630797386169434, "learning_rate": 1.4020020310459887e-05, "loss": 0.0327, "step": 8344 }, { "epoch": 0.904705117085863, "grad_norm": 1.426598072052002, "learning_rate": 1.4019294936892501e-05, "loss": 0.0216, "step": 8345 }, { "epoch": 0.9048135299219427, "grad_norm": 0.8303424715995789, "learning_rate": 1.4018569563325113e-05, "loss": 0.0465, "step": 8346 }, { "epoch": 0.9049219427580225, "grad_norm": 3.4279165267944336, "learning_rate": 1.4017844189757725e-05, "loss": 0.0439, "step": 8347 }, { "epoch": 0.9050303555941024, "grad_norm": 5.925168514251709, "learning_rate": 1.4017118816190339e-05, "loss": 0.0496, "step": 8348 }, { "epoch": 0.9051387684301822, "grad_norm": 1.0972877740859985, "learning_rate": 1.4016393442622951e-05, "loss": 0.0297, "step": 8349 }, { "epoch": 0.9052471812662619, "grad_norm": 3.6317553520202637, "learning_rate": 1.4015668069055565e-05, "loss": 0.0405, "step": 8350 }, { "epoch": 0.9053555941023417, "grad_norm": 1.4430513381958008, "learning_rate": 1.4014942695488179e-05, "loss": 0.0418, "step": 8351 }, { "epoch": 0.9054640069384216, "grad_norm": 1.5992079973220825, "learning_rate": 1.4014217321920791e-05, "loss": 0.03, "step": 8352 }, { "epoch": 0.9055724197745013, "grad_norm": 2.1984801292419434, "learning_rate": 1.4013491948353403e-05, "loss": 0.0556, "step": 8353 }, { "epoch": 0.9056808326105811, "grad_norm": 2.003969430923462, "learning_rate": 1.4012766574786017e-05, "loss": 0.0357, "step": 8354 }, { "epoch": 0.9057892454466608, "grad_norm": 1.470772624015808, "learning_rate": 1.4012041201218629e-05, "loss": 0.0545, "step": 8355 }, { "epoch": 0.9058976582827407, "grad_norm": 2.226196527481079, "learning_rate": 1.4011315827651241e-05, "loss": 0.0274, "step": 8356 }, { "epoch": 0.9060060711188205, "grad_norm": 1.7826358079910278, "learning_rate": 1.4010590454083855e-05, "loss": 0.0553, "step": 8357 }, { "epoch": 0.9061144839549002, "grad_norm": 0.8954988121986389, "learning_rate": 1.4009865080516467e-05, "loss": 0.0303, "step": 8358 }, { "epoch": 0.90622289679098, "grad_norm": 2.5133562088012695, "learning_rate": 1.400913970694908e-05, "loss": 0.0358, "step": 8359 }, { "epoch": 0.9063313096270599, "grad_norm": 2.0499863624572754, "learning_rate": 1.4008414333381693e-05, "loss": 0.0669, "step": 8360 }, { "epoch": 0.9064397224631396, "grad_norm": 0.6068353056907654, "learning_rate": 1.4007688959814305e-05, "loss": 0.0114, "step": 8361 }, { "epoch": 0.9065481352992194, "grad_norm": 1.740957498550415, "learning_rate": 1.4006963586246917e-05, "loss": 0.0218, "step": 8362 }, { "epoch": 0.9066565481352992, "grad_norm": 0.6878123879432678, "learning_rate": 1.400623821267953e-05, "loss": 0.0154, "step": 8363 }, { "epoch": 0.906764960971379, "grad_norm": 2.285616636276245, "learning_rate": 1.4005512839112143e-05, "loss": 0.0516, "step": 8364 }, { "epoch": 0.9068733738074588, "grad_norm": 1.2377870082855225, "learning_rate": 1.4004787465544756e-05, "loss": 0.0221, "step": 8365 }, { "epoch": 0.9069817866435386, "grad_norm": 2.519101142883301, "learning_rate": 1.4004062091977368e-05, "loss": 0.0514, "step": 8366 }, { "epoch": 0.9070901994796184, "grad_norm": 2.632387399673462, "learning_rate": 1.4003336718409983e-05, "loss": 0.0393, "step": 8367 }, { "epoch": 0.9071986123156982, "grad_norm": 1.4918887615203857, "learning_rate": 1.4002611344842595e-05, "loss": 0.0341, "step": 8368 }, { "epoch": 0.907307025151778, "grad_norm": 2.25828218460083, "learning_rate": 1.400188597127521e-05, "loss": 0.0532, "step": 8369 }, { "epoch": 0.9074154379878577, "grad_norm": 1.6215249300003052, "learning_rate": 1.4001160597707821e-05, "loss": 0.0448, "step": 8370 }, { "epoch": 0.9075238508239376, "grad_norm": 0.9436188340187073, "learning_rate": 1.4000435224140434e-05, "loss": 0.0256, "step": 8371 }, { "epoch": 0.9076322636600174, "grad_norm": 2.9293389320373535, "learning_rate": 1.3999709850573047e-05, "loss": 0.0445, "step": 8372 }, { "epoch": 0.9077406764960971, "grad_norm": 5.21458101272583, "learning_rate": 1.399898447700566e-05, "loss": 0.0711, "step": 8373 }, { "epoch": 0.9078490893321769, "grad_norm": 0.6759172081947327, "learning_rate": 1.3998259103438272e-05, "loss": 0.0228, "step": 8374 }, { "epoch": 0.9079575021682568, "grad_norm": 1.4202055931091309, "learning_rate": 1.3997533729870885e-05, "loss": 0.0269, "step": 8375 }, { "epoch": 0.9080659150043365, "grad_norm": 8.509138107299805, "learning_rate": 1.3996808356303498e-05, "loss": 0.0509, "step": 8376 }, { "epoch": 0.9081743278404163, "grad_norm": 0.860762357711792, "learning_rate": 1.399608298273611e-05, "loss": 0.0103, "step": 8377 }, { "epoch": 0.9082827406764961, "grad_norm": 4.7233567237854, "learning_rate": 1.3995357609168722e-05, "loss": 0.134, "step": 8378 }, { "epoch": 0.9083911535125759, "grad_norm": 4.218101501464844, "learning_rate": 1.3994632235601336e-05, "loss": 0.0515, "step": 8379 }, { "epoch": 0.9084995663486557, "grad_norm": 2.732952833175659, "learning_rate": 1.3993906862033948e-05, "loss": 0.0626, "step": 8380 }, { "epoch": 0.9086079791847355, "grad_norm": 1.2337231636047363, "learning_rate": 1.399318148846656e-05, "loss": 0.0243, "step": 8381 }, { "epoch": 0.9087163920208152, "grad_norm": 1.1484664678573608, "learning_rate": 1.3992456114899174e-05, "loss": 0.0297, "step": 8382 }, { "epoch": 0.9088248048568951, "grad_norm": 2.911869525909424, "learning_rate": 1.3991730741331786e-05, "loss": 0.0348, "step": 8383 }, { "epoch": 0.9089332176929749, "grad_norm": 2.3359508514404297, "learning_rate": 1.3991005367764402e-05, "loss": 0.0549, "step": 8384 }, { "epoch": 0.9090416305290546, "grad_norm": 2.5795090198516846, "learning_rate": 1.3990279994197014e-05, "loss": 0.0302, "step": 8385 }, { "epoch": 0.9091500433651344, "grad_norm": 2.81355619430542, "learning_rate": 1.3989554620629626e-05, "loss": 0.0326, "step": 8386 }, { "epoch": 0.9092584562012143, "grad_norm": 0.5831828713417053, "learning_rate": 1.398882924706224e-05, "loss": 0.0182, "step": 8387 }, { "epoch": 0.909366869037294, "grad_norm": 5.355520248413086, "learning_rate": 1.3988103873494852e-05, "loss": 0.0452, "step": 8388 }, { "epoch": 0.9094752818733738, "grad_norm": 1.3925482034683228, "learning_rate": 1.3987378499927464e-05, "loss": 0.0102, "step": 8389 }, { "epoch": 0.9095836947094535, "grad_norm": 1.753212332725525, "learning_rate": 1.3986653126360076e-05, "loss": 0.0255, "step": 8390 }, { "epoch": 0.9096921075455334, "grad_norm": 0.5685348510742188, "learning_rate": 1.398592775279269e-05, "loss": 0.0143, "step": 8391 }, { "epoch": 0.9098005203816132, "grad_norm": 0.7699397206306458, "learning_rate": 1.3985202379225302e-05, "loss": 0.0172, "step": 8392 }, { "epoch": 0.909908933217693, "grad_norm": 2.9899203777313232, "learning_rate": 1.3984477005657914e-05, "loss": 0.0805, "step": 8393 }, { "epoch": 0.9100173460537727, "grad_norm": 2.271130084991455, "learning_rate": 1.3983751632090528e-05, "loss": 0.0269, "step": 8394 }, { "epoch": 0.9101257588898526, "grad_norm": 3.0910139083862305, "learning_rate": 1.398302625852314e-05, "loss": 0.0522, "step": 8395 }, { "epoch": 0.9102341717259324, "grad_norm": 0.9130125641822815, "learning_rate": 1.3982300884955752e-05, "loss": 0.0043, "step": 8396 }, { "epoch": 0.9103425845620121, "grad_norm": 1.3777705430984497, "learning_rate": 1.3981575511388366e-05, "loss": 0.0297, "step": 8397 }, { "epoch": 0.910450997398092, "grad_norm": 1.1748160123825073, "learning_rate": 1.3980850137820978e-05, "loss": 0.0366, "step": 8398 }, { "epoch": 0.9105594102341718, "grad_norm": 7.832603454589844, "learning_rate": 1.398012476425359e-05, "loss": 0.0601, "step": 8399 }, { "epoch": 0.9106678230702515, "grad_norm": 2.257004499435425, "learning_rate": 1.3979399390686204e-05, "loss": 0.0636, "step": 8400 }, { "epoch": 0.9107762359063313, "grad_norm": 1.5414812564849854, "learning_rate": 1.3978674017118818e-05, "loss": 0.0348, "step": 8401 }, { "epoch": 0.9108846487424112, "grad_norm": 4.768926620483398, "learning_rate": 1.397794864355143e-05, "loss": 0.0293, "step": 8402 }, { "epoch": 0.9109930615784909, "grad_norm": 1.159966230392456, "learning_rate": 1.3977223269984044e-05, "loss": 0.0436, "step": 8403 }, { "epoch": 0.9111014744145707, "grad_norm": 0.8542189002037048, "learning_rate": 1.3976497896416656e-05, "loss": 0.0121, "step": 8404 }, { "epoch": 0.9112098872506504, "grad_norm": 3.1725118160247803, "learning_rate": 1.3975772522849268e-05, "loss": 0.0229, "step": 8405 }, { "epoch": 0.9113183000867303, "grad_norm": 3.287496328353882, "learning_rate": 1.3975047149281882e-05, "loss": 0.0745, "step": 8406 }, { "epoch": 0.9114267129228101, "grad_norm": 3.106501817703247, "learning_rate": 1.3974321775714494e-05, "loss": 0.0415, "step": 8407 }, { "epoch": 0.9115351257588898, "grad_norm": 1.8977187871932983, "learning_rate": 1.3973596402147107e-05, "loss": 0.0203, "step": 8408 }, { "epoch": 0.9116435385949696, "grad_norm": 1.6523510217666626, "learning_rate": 1.397287102857972e-05, "loss": 0.0117, "step": 8409 }, { "epoch": 0.9117519514310495, "grad_norm": 3.6483781337738037, "learning_rate": 1.3972145655012333e-05, "loss": 0.1274, "step": 8410 }, { "epoch": 0.9118603642671292, "grad_norm": 1.0802134275436401, "learning_rate": 1.3971420281444945e-05, "loss": 0.021, "step": 8411 }, { "epoch": 0.911968777103209, "grad_norm": 1.7584819793701172, "learning_rate": 1.3970694907877559e-05, "loss": 0.0318, "step": 8412 }, { "epoch": 0.9120771899392888, "grad_norm": 4.398706912994385, "learning_rate": 1.396996953431017e-05, "loss": 0.1127, "step": 8413 }, { "epoch": 0.9121856027753686, "grad_norm": 1.0590250492095947, "learning_rate": 1.3969244160742783e-05, "loss": 0.0281, "step": 8414 }, { "epoch": 0.9122940156114484, "grad_norm": 4.339182376861572, "learning_rate": 1.3968518787175395e-05, "loss": 0.0145, "step": 8415 }, { "epoch": 0.9124024284475282, "grad_norm": 2.7237088680267334, "learning_rate": 1.3967793413608009e-05, "loss": 0.0291, "step": 8416 }, { "epoch": 0.9125108412836079, "grad_norm": 2.874310255050659, "learning_rate": 1.3967068040040621e-05, "loss": 0.04, "step": 8417 }, { "epoch": 0.9126192541196878, "grad_norm": 3.674494743347168, "learning_rate": 1.3966342666473233e-05, "loss": 0.0409, "step": 8418 }, { "epoch": 0.9127276669557676, "grad_norm": 3.908358573913574, "learning_rate": 1.3965617292905849e-05, "loss": 0.0245, "step": 8419 }, { "epoch": 0.9128360797918473, "grad_norm": 0.9556687474250793, "learning_rate": 1.396489191933846e-05, "loss": 0.0085, "step": 8420 }, { "epoch": 0.9129444926279271, "grad_norm": 1.7271095514297485, "learning_rate": 1.3964166545771075e-05, "loss": 0.0129, "step": 8421 }, { "epoch": 0.913052905464007, "grad_norm": 0.9873594045639038, "learning_rate": 1.3963441172203687e-05, "loss": 0.0197, "step": 8422 }, { "epoch": 0.9131613183000867, "grad_norm": 2.1581833362579346, "learning_rate": 1.3962715798636299e-05, "loss": 0.0234, "step": 8423 }, { "epoch": 0.9132697311361665, "grad_norm": 1.0871610641479492, "learning_rate": 1.3961990425068913e-05, "loss": 0.0153, "step": 8424 }, { "epoch": 0.9133781439722463, "grad_norm": 2.369835138320923, "learning_rate": 1.3961265051501525e-05, "loss": 0.0381, "step": 8425 }, { "epoch": 0.9134865568083261, "grad_norm": 1.7753194570541382, "learning_rate": 1.3960539677934137e-05, "loss": 0.0324, "step": 8426 }, { "epoch": 0.9135949696444059, "grad_norm": 3.296874523162842, "learning_rate": 1.395981430436675e-05, "loss": 0.0218, "step": 8427 }, { "epoch": 0.9137033824804857, "grad_norm": 0.9732568860054016, "learning_rate": 1.3959088930799363e-05, "loss": 0.028, "step": 8428 }, { "epoch": 0.9138117953165655, "grad_norm": 2.0289008617401123, "learning_rate": 1.3958363557231975e-05, "loss": 0.0145, "step": 8429 }, { "epoch": 0.9139202081526453, "grad_norm": 2.0039775371551514, "learning_rate": 1.3957638183664587e-05, "loss": 0.0335, "step": 8430 }, { "epoch": 0.9140286209887251, "grad_norm": 2.578813314437866, "learning_rate": 1.3956912810097201e-05, "loss": 0.0335, "step": 8431 }, { "epoch": 0.9141370338248048, "grad_norm": 2.632233142852783, "learning_rate": 1.3956187436529813e-05, "loss": 0.0349, "step": 8432 }, { "epoch": 0.9142454466608847, "grad_norm": 0.5160049200057983, "learning_rate": 1.3955462062962425e-05, "loss": 0.0075, "step": 8433 }, { "epoch": 0.9143538594969645, "grad_norm": 5.505561351776123, "learning_rate": 1.395473668939504e-05, "loss": 0.036, "step": 8434 }, { "epoch": 0.9144622723330442, "grad_norm": 2.02065372467041, "learning_rate": 1.3954011315827651e-05, "loss": 0.0192, "step": 8435 }, { "epoch": 0.914570685169124, "grad_norm": 2.5754992961883545, "learning_rate": 1.3953285942260267e-05, "loss": 0.0572, "step": 8436 }, { "epoch": 0.9146790980052039, "grad_norm": 1.466325283050537, "learning_rate": 1.3952560568692879e-05, "loss": 0.0348, "step": 8437 }, { "epoch": 0.9147875108412836, "grad_norm": 4.295619487762451, "learning_rate": 1.3951835195125491e-05, "loss": 0.0132, "step": 8438 }, { "epoch": 0.9148959236773634, "grad_norm": 3.842074394226074, "learning_rate": 1.3951109821558105e-05, "loss": 0.0309, "step": 8439 }, { "epoch": 0.9150043365134432, "grad_norm": 1.976240634918213, "learning_rate": 1.3950384447990717e-05, "loss": 0.0261, "step": 8440 }, { "epoch": 0.915112749349523, "grad_norm": 5.27584981918335, "learning_rate": 1.394965907442333e-05, "loss": 0.063, "step": 8441 }, { "epoch": 0.9152211621856028, "grad_norm": 8.240487098693848, "learning_rate": 1.3948933700855941e-05, "loss": 0.0595, "step": 8442 }, { "epoch": 0.9153295750216826, "grad_norm": 1.6143853664398193, "learning_rate": 1.3948208327288555e-05, "loss": 0.0374, "step": 8443 }, { "epoch": 0.9154379878577623, "grad_norm": 2.347226858139038, "learning_rate": 1.3947482953721167e-05, "loss": 0.0308, "step": 8444 }, { "epoch": 0.9155464006938422, "grad_norm": 1.250130534172058, "learning_rate": 1.394675758015378e-05, "loss": 0.0254, "step": 8445 }, { "epoch": 0.915654813529922, "grad_norm": 2.0128374099731445, "learning_rate": 1.3946032206586393e-05, "loss": 0.0321, "step": 8446 }, { "epoch": 0.9157632263660017, "grad_norm": 1.4317429065704346, "learning_rate": 1.3945306833019006e-05, "loss": 0.0439, "step": 8447 }, { "epoch": 0.9158716392020815, "grad_norm": 0.4695618748664856, "learning_rate": 1.3944581459451618e-05, "loss": 0.0069, "step": 8448 }, { "epoch": 0.9159800520381614, "grad_norm": 1.1432857513427734, "learning_rate": 1.3943856085884232e-05, "loss": 0.0359, "step": 8449 }, { "epoch": 0.9160884648742411, "grad_norm": 1.996779203414917, "learning_rate": 1.3943130712316844e-05, "loss": 0.0245, "step": 8450 }, { "epoch": 0.9161968777103209, "grad_norm": 3.646860122680664, "learning_rate": 1.3942405338749456e-05, "loss": 0.0378, "step": 8451 }, { "epoch": 0.9163052905464006, "grad_norm": 0.620581865310669, "learning_rate": 1.394167996518207e-05, "loss": 0.005, "step": 8452 }, { "epoch": 0.9164137033824805, "grad_norm": 2.405488967895508, "learning_rate": 1.3940954591614683e-05, "loss": 0.021, "step": 8453 }, { "epoch": 0.9165221162185603, "grad_norm": 2.901467800140381, "learning_rate": 1.3940229218047296e-05, "loss": 0.0357, "step": 8454 }, { "epoch": 0.91663052905464, "grad_norm": 4.247167587280273, "learning_rate": 1.393950384447991e-05, "loss": 0.0453, "step": 8455 }, { "epoch": 0.9167389418907199, "grad_norm": 3.292783498764038, "learning_rate": 1.3938778470912522e-05, "loss": 0.0365, "step": 8456 }, { "epoch": 0.9168473547267997, "grad_norm": 2.518894910812378, "learning_rate": 1.3938053097345134e-05, "loss": 0.0153, "step": 8457 }, { "epoch": 0.9169557675628794, "grad_norm": 0.5730555057525635, "learning_rate": 1.3937327723777748e-05, "loss": 0.0145, "step": 8458 }, { "epoch": 0.9170641803989592, "grad_norm": 5.310911655426025, "learning_rate": 1.393660235021036e-05, "loss": 0.0701, "step": 8459 }, { "epoch": 0.9171725932350391, "grad_norm": 1.5671309232711792, "learning_rate": 1.3935876976642972e-05, "loss": 0.028, "step": 8460 }, { "epoch": 0.9172810060711188, "grad_norm": 0.9611075520515442, "learning_rate": 1.3935151603075586e-05, "loss": 0.0218, "step": 8461 }, { "epoch": 0.9173894189071986, "grad_norm": 1.8810392618179321, "learning_rate": 1.3934426229508198e-05, "loss": 0.0646, "step": 8462 }, { "epoch": 0.9174978317432784, "grad_norm": 3.142502784729004, "learning_rate": 1.393370085594081e-05, "loss": 0.0367, "step": 8463 }, { "epoch": 0.9176062445793582, "grad_norm": 0.15418626368045807, "learning_rate": 1.3932975482373424e-05, "loss": 0.0018, "step": 8464 }, { "epoch": 0.917714657415438, "grad_norm": 1.7578558921813965, "learning_rate": 1.3932250108806036e-05, "loss": 0.014, "step": 8465 }, { "epoch": 0.9178230702515178, "grad_norm": 0.556182324886322, "learning_rate": 1.3931524735238648e-05, "loss": 0.0186, "step": 8466 }, { "epoch": 0.9179314830875975, "grad_norm": 5.502771377563477, "learning_rate": 1.3930799361671262e-05, "loss": 0.0232, "step": 8467 }, { "epoch": 0.9180398959236774, "grad_norm": 2.2661497592926025, "learning_rate": 1.3930073988103874e-05, "loss": 0.0567, "step": 8468 }, { "epoch": 0.9181483087597572, "grad_norm": 4.416314601898193, "learning_rate": 1.3929348614536486e-05, "loss": 0.0953, "step": 8469 }, { "epoch": 0.9182567215958369, "grad_norm": 6.330834865570068, "learning_rate": 1.3928623240969098e-05, "loss": 0.0728, "step": 8470 }, { "epoch": 0.9183651344319167, "grad_norm": 1.979494571685791, "learning_rate": 1.3927897867401714e-05, "loss": 0.0292, "step": 8471 }, { "epoch": 0.9184735472679966, "grad_norm": 2.9342830181121826, "learning_rate": 1.3927172493834326e-05, "loss": 0.0758, "step": 8472 }, { "epoch": 0.9185819601040763, "grad_norm": 1.6065961122512817, "learning_rate": 1.392644712026694e-05, "loss": 0.0335, "step": 8473 }, { "epoch": 0.9186903729401561, "grad_norm": 1.2183243036270142, "learning_rate": 1.3925721746699552e-05, "loss": 0.0281, "step": 8474 }, { "epoch": 0.9187987857762359, "grad_norm": 3.535433053970337, "learning_rate": 1.3924996373132164e-05, "loss": 0.0782, "step": 8475 }, { "epoch": 0.9189071986123157, "grad_norm": 2.2581787109375, "learning_rate": 1.3924270999564778e-05, "loss": 0.0428, "step": 8476 }, { "epoch": 0.9190156114483955, "grad_norm": 4.963884353637695, "learning_rate": 1.392354562599739e-05, "loss": 0.0803, "step": 8477 }, { "epoch": 0.9191240242844753, "grad_norm": 3.1233506202697754, "learning_rate": 1.3922820252430002e-05, "loss": 0.0253, "step": 8478 }, { "epoch": 0.919232437120555, "grad_norm": 2.5477535724639893, "learning_rate": 1.3922094878862616e-05, "loss": 0.041, "step": 8479 }, { "epoch": 0.9193408499566349, "grad_norm": 0.30731967091560364, "learning_rate": 1.3921369505295228e-05, "loss": 0.0047, "step": 8480 }, { "epoch": 0.9194492627927147, "grad_norm": 1.6848931312561035, "learning_rate": 1.392064413172784e-05, "loss": 0.0164, "step": 8481 }, { "epoch": 0.9195576756287944, "grad_norm": 1.618739128112793, "learning_rate": 1.3919918758160453e-05, "loss": 0.0305, "step": 8482 }, { "epoch": 0.9196660884648742, "grad_norm": 1.1762275695800781, "learning_rate": 1.3919193384593066e-05, "loss": 0.0228, "step": 8483 }, { "epoch": 0.9197745013009541, "grad_norm": 3.623002290725708, "learning_rate": 1.3918468011025679e-05, "loss": 0.0395, "step": 8484 }, { "epoch": 0.9198829141370338, "grad_norm": 2.176908016204834, "learning_rate": 1.391774263745829e-05, "loss": 0.0284, "step": 8485 }, { "epoch": 0.9199913269731136, "grad_norm": 1.0026015043258667, "learning_rate": 1.3917017263890905e-05, "loss": 0.0167, "step": 8486 }, { "epoch": 0.9200997398091935, "grad_norm": 5.329020023345947, "learning_rate": 1.3916291890323517e-05, "loss": 0.0367, "step": 8487 }, { "epoch": 0.9202081526452732, "grad_norm": 1.4116437435150146, "learning_rate": 1.3915566516756132e-05, "loss": 0.0325, "step": 8488 }, { "epoch": 0.920316565481353, "grad_norm": 0.6833784580230713, "learning_rate": 1.3914841143188744e-05, "loss": 0.0149, "step": 8489 }, { "epoch": 0.9204249783174328, "grad_norm": 1.6900891065597534, "learning_rate": 1.3914115769621357e-05, "loss": 0.0192, "step": 8490 }, { "epoch": 0.9205333911535126, "grad_norm": 0.6202974915504456, "learning_rate": 1.391339039605397e-05, "loss": 0.0138, "step": 8491 }, { "epoch": 0.9206418039895924, "grad_norm": 5.277013301849365, "learning_rate": 1.3912665022486582e-05, "loss": 0.0767, "step": 8492 }, { "epoch": 0.9207502168256722, "grad_norm": 0.6606448292732239, "learning_rate": 1.3911939648919195e-05, "loss": 0.0156, "step": 8493 }, { "epoch": 0.9208586296617519, "grad_norm": 0.706046998500824, "learning_rate": 1.3911214275351807e-05, "loss": 0.0331, "step": 8494 }, { "epoch": 0.9209670424978318, "grad_norm": 2.4067702293395996, "learning_rate": 1.391048890178442e-05, "loss": 0.0567, "step": 8495 }, { "epoch": 0.9210754553339116, "grad_norm": 0.7845496535301208, "learning_rate": 1.3909763528217033e-05, "loss": 0.0209, "step": 8496 }, { "epoch": 0.9211838681699913, "grad_norm": 3.9900336265563965, "learning_rate": 1.3909038154649645e-05, "loss": 0.0522, "step": 8497 }, { "epoch": 0.9212922810060711, "grad_norm": 1.071343183517456, "learning_rate": 1.3908312781082259e-05, "loss": 0.02, "step": 8498 }, { "epoch": 0.921400693842151, "grad_norm": 1.4673922061920166, "learning_rate": 1.3907587407514871e-05, "loss": 0.0474, "step": 8499 }, { "epoch": 0.9215091066782307, "grad_norm": 1.6775696277618408, "learning_rate": 1.3906862033947483e-05, "loss": 0.0193, "step": 8500 }, { "epoch": 0.9216175195143105, "grad_norm": 2.0157241821289062, "learning_rate": 1.3906136660380097e-05, "loss": 0.0452, "step": 8501 }, { "epoch": 0.9217259323503902, "grad_norm": 3.695042133331299, "learning_rate": 1.3905411286812709e-05, "loss": 0.0586, "step": 8502 }, { "epoch": 0.9218343451864701, "grad_norm": 6.2164177894592285, "learning_rate": 1.3904685913245321e-05, "loss": 0.0315, "step": 8503 }, { "epoch": 0.9219427580225499, "grad_norm": 2.746223211288452, "learning_rate": 1.3903960539677935e-05, "loss": 0.0226, "step": 8504 }, { "epoch": 0.9220511708586296, "grad_norm": 2.270765781402588, "learning_rate": 1.3903235166110549e-05, "loss": 0.039, "step": 8505 }, { "epoch": 0.9221595836947094, "grad_norm": 3.2498037815093994, "learning_rate": 1.3902509792543161e-05, "loss": 0.0693, "step": 8506 }, { "epoch": 0.9222679965307893, "grad_norm": 2.4164493083953857, "learning_rate": 1.3901784418975775e-05, "loss": 0.0261, "step": 8507 }, { "epoch": 0.922376409366869, "grad_norm": 1.6743508577346802, "learning_rate": 1.3901059045408387e-05, "loss": 0.0437, "step": 8508 }, { "epoch": 0.9224848222029488, "grad_norm": 3.3922715187072754, "learning_rate": 1.3900333671840999e-05, "loss": 0.0346, "step": 8509 }, { "epoch": 0.9225932350390286, "grad_norm": 2.5430827140808105, "learning_rate": 1.3899608298273613e-05, "loss": 0.0374, "step": 8510 }, { "epoch": 0.9227016478751084, "grad_norm": 1.5933634042739868, "learning_rate": 1.3898882924706225e-05, "loss": 0.0273, "step": 8511 }, { "epoch": 0.9228100607111882, "grad_norm": 3.5318803787231445, "learning_rate": 1.3898157551138837e-05, "loss": 0.0428, "step": 8512 }, { "epoch": 0.922918473547268, "grad_norm": 2.15714955329895, "learning_rate": 1.3897432177571451e-05, "loss": 0.0504, "step": 8513 }, { "epoch": 0.9230268863833477, "grad_norm": 2.274625539779663, "learning_rate": 1.3896706804004063e-05, "loss": 0.0232, "step": 8514 }, { "epoch": 0.9231352992194276, "grad_norm": 2.523036003112793, "learning_rate": 1.3895981430436675e-05, "loss": 0.0196, "step": 8515 }, { "epoch": 0.9232437120555074, "grad_norm": 3.207883358001709, "learning_rate": 1.389525605686929e-05, "loss": 0.0454, "step": 8516 }, { "epoch": 0.9233521248915871, "grad_norm": 1.50212562084198, "learning_rate": 1.3894530683301901e-05, "loss": 0.0363, "step": 8517 }, { "epoch": 0.923460537727667, "grad_norm": 2.8428945541381836, "learning_rate": 1.3893805309734513e-05, "loss": 0.0711, "step": 8518 }, { "epoch": 0.9235689505637468, "grad_norm": 1.309156060218811, "learning_rate": 1.3893079936167127e-05, "loss": 0.0116, "step": 8519 }, { "epoch": 0.9236773633998265, "grad_norm": 2.557647705078125, "learning_rate": 1.389235456259974e-05, "loss": 0.0522, "step": 8520 }, { "epoch": 0.9237857762359063, "grad_norm": 1.7750277519226074, "learning_rate": 1.3891629189032352e-05, "loss": 0.0086, "step": 8521 }, { "epoch": 0.9238941890719862, "grad_norm": 1.7410287857055664, "learning_rate": 1.3890903815464967e-05, "loss": 0.0236, "step": 8522 }, { "epoch": 0.9240026019080659, "grad_norm": 0.6170380115509033, "learning_rate": 1.389017844189758e-05, "loss": 0.0111, "step": 8523 }, { "epoch": 0.9241110147441457, "grad_norm": 1.8050258159637451, "learning_rate": 1.3889453068330191e-05, "loss": 0.0186, "step": 8524 }, { "epoch": 0.9242194275802255, "grad_norm": 1.877638816833496, "learning_rate": 1.3888727694762805e-05, "loss": 0.0459, "step": 8525 }, { "epoch": 0.9243278404163053, "grad_norm": 0.806684672832489, "learning_rate": 1.3888002321195417e-05, "loss": 0.0172, "step": 8526 }, { "epoch": 0.9244362532523851, "grad_norm": 3.92775297164917, "learning_rate": 1.388727694762803e-05, "loss": 0.0504, "step": 8527 }, { "epoch": 0.9245446660884649, "grad_norm": 1.5992121696472168, "learning_rate": 1.3886551574060643e-05, "loss": 0.0262, "step": 8528 }, { "epoch": 0.9246530789245446, "grad_norm": 2.0836985111236572, "learning_rate": 1.3885826200493256e-05, "loss": 0.0522, "step": 8529 }, { "epoch": 0.9247614917606245, "grad_norm": 2.2595431804656982, "learning_rate": 1.3885100826925868e-05, "loss": 0.0399, "step": 8530 }, { "epoch": 0.9248699045967043, "grad_norm": 5.292272567749023, "learning_rate": 1.3884375453358481e-05, "loss": 0.0327, "step": 8531 }, { "epoch": 0.924978317432784, "grad_norm": 3.268470287322998, "learning_rate": 1.3883650079791094e-05, "loss": 0.0488, "step": 8532 }, { "epoch": 0.9250867302688638, "grad_norm": 4.603903293609619, "learning_rate": 1.3882924706223706e-05, "loss": 0.1132, "step": 8533 }, { "epoch": 0.9251951431049437, "grad_norm": 2.748138904571533, "learning_rate": 1.3882199332656318e-05, "loss": 0.0274, "step": 8534 }, { "epoch": 0.9253035559410234, "grad_norm": 4.402390003204346, "learning_rate": 1.3881473959088932e-05, "loss": 0.1193, "step": 8535 }, { "epoch": 0.9254119687771032, "grad_norm": 2.5961456298828125, "learning_rate": 1.3880748585521544e-05, "loss": 0.0262, "step": 8536 }, { "epoch": 0.925520381613183, "grad_norm": 1.5720094442367554, "learning_rate": 1.3880023211954156e-05, "loss": 0.0169, "step": 8537 }, { "epoch": 0.9256287944492628, "grad_norm": 3.2837724685668945, "learning_rate": 1.387929783838677e-05, "loss": 0.0644, "step": 8538 }, { "epoch": 0.9257372072853426, "grad_norm": 0.8303350210189819, "learning_rate": 1.3878572464819382e-05, "loss": 0.015, "step": 8539 }, { "epoch": 0.9258456201214224, "grad_norm": 1.4758163690567017, "learning_rate": 1.3877847091251998e-05, "loss": 0.0335, "step": 8540 }, { "epoch": 0.9259540329575021, "grad_norm": 5.2917351722717285, "learning_rate": 1.387712171768461e-05, "loss": 0.0401, "step": 8541 }, { "epoch": 0.926062445793582, "grad_norm": 1.740087628364563, "learning_rate": 1.3876396344117222e-05, "loss": 0.0307, "step": 8542 }, { "epoch": 0.9261708586296618, "grad_norm": 3.266124963760376, "learning_rate": 1.3875670970549836e-05, "loss": 0.057, "step": 8543 }, { "epoch": 0.9262792714657415, "grad_norm": 3.613456964492798, "learning_rate": 1.3874945596982448e-05, "loss": 0.0117, "step": 8544 }, { "epoch": 0.9263876843018214, "grad_norm": 2.960214853286743, "learning_rate": 1.387422022341506e-05, "loss": 0.0295, "step": 8545 }, { "epoch": 0.9264960971379012, "grad_norm": 1.4547712802886963, "learning_rate": 1.3873494849847672e-05, "loss": 0.0142, "step": 8546 }, { "epoch": 0.9266045099739809, "grad_norm": 0.7630785703659058, "learning_rate": 1.3872769476280286e-05, "loss": 0.0296, "step": 8547 }, { "epoch": 0.9267129228100607, "grad_norm": 1.7257426977157593, "learning_rate": 1.3872044102712898e-05, "loss": 0.0355, "step": 8548 }, { "epoch": 0.9268213356461406, "grad_norm": 0.5296071767807007, "learning_rate": 1.387131872914551e-05, "loss": 0.0233, "step": 8549 }, { "epoch": 0.9269297484822203, "grad_norm": 2.6967928409576416, "learning_rate": 1.3870593355578124e-05, "loss": 0.034, "step": 8550 }, { "epoch": 0.9270381613183001, "grad_norm": 2.100094795227051, "learning_rate": 1.3869867982010736e-05, "loss": 0.012, "step": 8551 }, { "epoch": 0.9271465741543798, "grad_norm": 4.163510322570801, "learning_rate": 1.3869142608443348e-05, "loss": 0.0364, "step": 8552 }, { "epoch": 0.9272549869904597, "grad_norm": 2.741142988204956, "learning_rate": 1.3868417234875962e-05, "loss": 0.0619, "step": 8553 }, { "epoch": 0.9273633998265395, "grad_norm": 2.2313895225524902, "learning_rate": 1.3867691861308574e-05, "loss": 0.0308, "step": 8554 }, { "epoch": 0.9274718126626192, "grad_norm": 0.6482556462287903, "learning_rate": 1.3866966487741186e-05, "loss": 0.026, "step": 8555 }, { "epoch": 0.927580225498699, "grad_norm": 2.4296441078186035, "learning_rate": 1.38662411141738e-05, "loss": 0.0401, "step": 8556 }, { "epoch": 0.9276886383347789, "grad_norm": 3.09734845161438, "learning_rate": 1.3865515740606414e-05, "loss": 0.0686, "step": 8557 }, { "epoch": 0.9277970511708586, "grad_norm": 1.5953162908554077, "learning_rate": 1.3864790367039026e-05, "loss": 0.0333, "step": 8558 }, { "epoch": 0.9279054640069384, "grad_norm": 8.56212329864502, "learning_rate": 1.386406499347164e-05, "loss": 0.0454, "step": 8559 }, { "epoch": 0.9280138768430182, "grad_norm": 8.747836112976074, "learning_rate": 1.3863339619904252e-05, "loss": 0.0474, "step": 8560 }, { "epoch": 0.928122289679098, "grad_norm": 1.589890718460083, "learning_rate": 1.3862614246336864e-05, "loss": 0.0238, "step": 8561 }, { "epoch": 0.9282307025151778, "grad_norm": 0.9139382243156433, "learning_rate": 1.3861888872769478e-05, "loss": 0.0161, "step": 8562 }, { "epoch": 0.9283391153512576, "grad_norm": 1.8919464349746704, "learning_rate": 1.386116349920209e-05, "loss": 0.0164, "step": 8563 }, { "epoch": 0.9284475281873373, "grad_norm": 0.6242110133171082, "learning_rate": 1.3860438125634703e-05, "loss": 0.0039, "step": 8564 }, { "epoch": 0.9285559410234172, "grad_norm": 2.6744837760925293, "learning_rate": 1.3859712752067316e-05, "loss": 0.0113, "step": 8565 }, { "epoch": 0.928664353859497, "grad_norm": 2.870377540588379, "learning_rate": 1.3858987378499929e-05, "loss": 0.0823, "step": 8566 }, { "epoch": 0.9287727666955767, "grad_norm": 3.599372148513794, "learning_rate": 1.385826200493254e-05, "loss": 0.0489, "step": 8567 }, { "epoch": 0.9288811795316565, "grad_norm": 1.7099000215530396, "learning_rate": 1.3857536631365155e-05, "loss": 0.03, "step": 8568 }, { "epoch": 0.9289895923677364, "grad_norm": 2.450592041015625, "learning_rate": 1.3856811257797767e-05, "loss": 0.0466, "step": 8569 }, { "epoch": 0.9290980052038161, "grad_norm": 4.133527755737305, "learning_rate": 1.3856085884230379e-05, "loss": 0.0544, "step": 8570 }, { "epoch": 0.9292064180398959, "grad_norm": 3.215651035308838, "learning_rate": 1.3855360510662993e-05, "loss": 0.048, "step": 8571 }, { "epoch": 0.9293148308759757, "grad_norm": 2.42881441116333, "learning_rate": 1.3854635137095605e-05, "loss": 0.0155, "step": 8572 }, { "epoch": 0.9294232437120555, "grad_norm": 1.9095710515975952, "learning_rate": 1.3853909763528217e-05, "loss": 0.0314, "step": 8573 }, { "epoch": 0.9295316565481353, "grad_norm": 2.7649505138397217, "learning_rate": 1.3853184389960832e-05, "loss": 0.0945, "step": 8574 }, { "epoch": 0.9296400693842151, "grad_norm": 1.3549995422363281, "learning_rate": 1.3852459016393445e-05, "loss": 0.0447, "step": 8575 }, { "epoch": 0.9297484822202949, "grad_norm": 1.196387529373169, "learning_rate": 1.3851733642826057e-05, "loss": 0.0243, "step": 8576 }, { "epoch": 0.9298568950563747, "grad_norm": 1.392562985420227, "learning_rate": 1.385100826925867e-05, "loss": 0.0377, "step": 8577 }, { "epoch": 0.9299653078924545, "grad_norm": 1.5887205600738525, "learning_rate": 1.3850282895691283e-05, "loss": 0.0195, "step": 8578 }, { "epoch": 0.9300737207285342, "grad_norm": 1.1549125909805298, "learning_rate": 1.3849557522123895e-05, "loss": 0.0078, "step": 8579 }, { "epoch": 0.9301821335646141, "grad_norm": 3.1087048053741455, "learning_rate": 1.3848832148556509e-05, "loss": 0.0319, "step": 8580 }, { "epoch": 0.9302905464006939, "grad_norm": 3.2342166900634766, "learning_rate": 1.3848106774989121e-05, "loss": 0.0473, "step": 8581 }, { "epoch": 0.9303989592367736, "grad_norm": 2.990643262863159, "learning_rate": 1.3847381401421733e-05, "loss": 0.0389, "step": 8582 }, { "epoch": 0.9305073720728534, "grad_norm": 1.7612022161483765, "learning_rate": 1.3846656027854347e-05, "loss": 0.0268, "step": 8583 }, { "epoch": 0.9306157849089333, "grad_norm": 3.206082582473755, "learning_rate": 1.3845930654286959e-05, "loss": 0.0802, "step": 8584 }, { "epoch": 0.930724197745013, "grad_norm": 1.241625428199768, "learning_rate": 1.3845205280719571e-05, "loss": 0.0118, "step": 8585 }, { "epoch": 0.9308326105810928, "grad_norm": 0.9702046513557434, "learning_rate": 1.3844479907152183e-05, "loss": 0.0084, "step": 8586 }, { "epoch": 0.9309410234171726, "grad_norm": 0.9198194742202759, "learning_rate": 1.3843754533584797e-05, "loss": 0.025, "step": 8587 }, { "epoch": 0.9310494362532524, "grad_norm": 3.742159366607666, "learning_rate": 1.384302916001741e-05, "loss": 0.0441, "step": 8588 }, { "epoch": 0.9311578490893322, "grad_norm": 1.176159143447876, "learning_rate": 1.3842303786450021e-05, "loss": 0.0208, "step": 8589 }, { "epoch": 0.931266261925412, "grad_norm": 2.4440412521362305, "learning_rate": 1.3841578412882635e-05, "loss": 0.0246, "step": 8590 }, { "epoch": 0.9313746747614917, "grad_norm": 3.5356016159057617, "learning_rate": 1.3840853039315247e-05, "loss": 0.018, "step": 8591 }, { "epoch": 0.9314830875975716, "grad_norm": 1.2604548931121826, "learning_rate": 1.3840127665747863e-05, "loss": 0.0263, "step": 8592 }, { "epoch": 0.9315915004336514, "grad_norm": 2.446388006210327, "learning_rate": 1.3839402292180475e-05, "loss": 0.0427, "step": 8593 }, { "epoch": 0.9316999132697311, "grad_norm": 2.337949514389038, "learning_rate": 1.3838676918613087e-05, "loss": 0.0295, "step": 8594 }, { "epoch": 0.9318083261058109, "grad_norm": 2.7371184825897217, "learning_rate": 1.3837951545045701e-05, "loss": 0.0287, "step": 8595 }, { "epoch": 0.9319167389418908, "grad_norm": 1.204284906387329, "learning_rate": 1.3837226171478313e-05, "loss": 0.0094, "step": 8596 }, { "epoch": 0.9320251517779705, "grad_norm": 2.807631015777588, "learning_rate": 1.3836500797910925e-05, "loss": 0.0531, "step": 8597 }, { "epoch": 0.9321335646140503, "grad_norm": 2.1321239471435547, "learning_rate": 1.3835775424343537e-05, "loss": 0.05, "step": 8598 }, { "epoch": 0.93224197745013, "grad_norm": 2.511660575866699, "learning_rate": 1.3835050050776151e-05, "loss": 0.0163, "step": 8599 }, { "epoch": 0.9323503902862099, "grad_norm": 1.1687923669815063, "learning_rate": 1.3834324677208763e-05, "loss": 0.0169, "step": 8600 }, { "epoch": 0.9324588031222897, "grad_norm": 3.853854179382324, "learning_rate": 1.3833599303641376e-05, "loss": 0.0459, "step": 8601 }, { "epoch": 0.9325672159583694, "grad_norm": 1.920576572418213, "learning_rate": 1.383287393007399e-05, "loss": 0.035, "step": 8602 }, { "epoch": 0.9326756287944492, "grad_norm": 0.3668023347854614, "learning_rate": 1.3832148556506602e-05, "loss": 0.0056, "step": 8603 }, { "epoch": 0.9327840416305291, "grad_norm": 2.8632943630218506, "learning_rate": 1.3831423182939214e-05, "loss": 0.01, "step": 8604 }, { "epoch": 0.9328924544666088, "grad_norm": 2.8440661430358887, "learning_rate": 1.3830697809371828e-05, "loss": 0.0213, "step": 8605 }, { "epoch": 0.9330008673026886, "grad_norm": 3.548954486846924, "learning_rate": 1.382997243580444e-05, "loss": 0.0272, "step": 8606 }, { "epoch": 0.9331092801387685, "grad_norm": 3.566431999206543, "learning_rate": 1.3829247062237052e-05, "loss": 0.0578, "step": 8607 }, { "epoch": 0.9332176929748482, "grad_norm": 1.6082308292388916, "learning_rate": 1.3828521688669666e-05, "loss": 0.0549, "step": 8608 }, { "epoch": 0.933326105810928, "grad_norm": 2.3978071212768555, "learning_rate": 1.382779631510228e-05, "loss": 0.0181, "step": 8609 }, { "epoch": 0.9334345186470078, "grad_norm": 1.5661945343017578, "learning_rate": 1.3827070941534892e-05, "loss": 0.0455, "step": 8610 }, { "epoch": 0.9335429314830876, "grad_norm": 2.3745524883270264, "learning_rate": 1.3826345567967505e-05, "loss": 0.031, "step": 8611 }, { "epoch": 0.9336513443191674, "grad_norm": 3.7123072147369385, "learning_rate": 1.3825620194400118e-05, "loss": 0.0743, "step": 8612 }, { "epoch": 0.9337597571552472, "grad_norm": 2.15726900100708, "learning_rate": 1.382489482083273e-05, "loss": 0.0245, "step": 8613 }, { "epoch": 0.9338681699913269, "grad_norm": 2.0327939987182617, "learning_rate": 1.3824169447265344e-05, "loss": 0.0276, "step": 8614 }, { "epoch": 0.9339765828274068, "grad_norm": 2.298553466796875, "learning_rate": 1.3823444073697956e-05, "loss": 0.0222, "step": 8615 }, { "epoch": 0.9340849956634866, "grad_norm": 1.2931773662567139, "learning_rate": 1.3822718700130568e-05, "loss": 0.0185, "step": 8616 }, { "epoch": 0.9341934084995663, "grad_norm": 1.8030277490615845, "learning_rate": 1.3821993326563182e-05, "loss": 0.0085, "step": 8617 }, { "epoch": 0.9343018213356461, "grad_norm": 5.209865093231201, "learning_rate": 1.3821267952995794e-05, "loss": 0.0399, "step": 8618 }, { "epoch": 0.934410234171726, "grad_norm": 2.9728875160217285, "learning_rate": 1.3820542579428406e-05, "loss": 0.0319, "step": 8619 }, { "epoch": 0.9345186470078057, "grad_norm": 1.2983288764953613, "learning_rate": 1.381981720586102e-05, "loss": 0.0537, "step": 8620 }, { "epoch": 0.9346270598438855, "grad_norm": 6.20657205581665, "learning_rate": 1.3819091832293632e-05, "loss": 0.077, "step": 8621 }, { "epoch": 0.9347354726799653, "grad_norm": 2.5850107669830322, "learning_rate": 1.3818366458726244e-05, "loss": 0.0308, "step": 8622 }, { "epoch": 0.9348438855160451, "grad_norm": 3.037051200866699, "learning_rate": 1.3817641085158858e-05, "loss": 0.0371, "step": 8623 }, { "epoch": 0.9349522983521249, "grad_norm": 0.4638214707374573, "learning_rate": 1.381691571159147e-05, "loss": 0.0166, "step": 8624 }, { "epoch": 0.9350607111882047, "grad_norm": 1.4530658721923828, "learning_rate": 1.3816190338024082e-05, "loss": 0.02, "step": 8625 }, { "epoch": 0.9351691240242844, "grad_norm": 3.065890312194824, "learning_rate": 1.3815464964456698e-05, "loss": 0.0136, "step": 8626 }, { "epoch": 0.9352775368603643, "grad_norm": 1.6046892404556274, "learning_rate": 1.381473959088931e-05, "loss": 0.0139, "step": 8627 }, { "epoch": 0.9353859496964441, "grad_norm": 0.5714161992073059, "learning_rate": 1.3814014217321922e-05, "loss": 0.0194, "step": 8628 }, { "epoch": 0.9354943625325238, "grad_norm": 2.4359793663024902, "learning_rate": 1.3813288843754536e-05, "loss": 0.0195, "step": 8629 }, { "epoch": 0.9356027753686036, "grad_norm": 1.3660578727722168, "learning_rate": 1.3812563470187148e-05, "loss": 0.0307, "step": 8630 }, { "epoch": 0.9357111882046835, "grad_norm": 2.5724852085113525, "learning_rate": 1.381183809661976e-05, "loss": 0.0147, "step": 8631 }, { "epoch": 0.9358196010407632, "grad_norm": 4.731133937835693, "learning_rate": 1.3811112723052374e-05, "loss": 0.0367, "step": 8632 }, { "epoch": 0.935928013876843, "grad_norm": 4.05128812789917, "learning_rate": 1.3810387349484986e-05, "loss": 0.0248, "step": 8633 }, { "epoch": 0.9360364267129229, "grad_norm": 5.086112976074219, "learning_rate": 1.3809661975917598e-05, "loss": 0.0277, "step": 8634 }, { "epoch": 0.9361448395490026, "grad_norm": 1.9334702491760254, "learning_rate": 1.3808936602350212e-05, "loss": 0.0289, "step": 8635 }, { "epoch": 0.9362532523850824, "grad_norm": 3.3886940479278564, "learning_rate": 1.3808211228782824e-05, "loss": 0.0423, "step": 8636 }, { "epoch": 0.9363616652211622, "grad_norm": 2.433323621749878, "learning_rate": 1.3807485855215436e-05, "loss": 0.0094, "step": 8637 }, { "epoch": 0.936470078057242, "grad_norm": 1.1931025981903076, "learning_rate": 1.3806760481648049e-05, "loss": 0.062, "step": 8638 }, { "epoch": 0.9365784908933218, "grad_norm": 3.10538649559021, "learning_rate": 1.3806035108080662e-05, "loss": 0.048, "step": 8639 }, { "epoch": 0.9366869037294016, "grad_norm": 4.181024551391602, "learning_rate": 1.3805309734513275e-05, "loss": 0.0591, "step": 8640 }, { "epoch": 0.9367953165654813, "grad_norm": 0.7976246476173401, "learning_rate": 1.3804584360945887e-05, "loss": 0.0448, "step": 8641 }, { "epoch": 0.9369037294015612, "grad_norm": 3.036518096923828, "learning_rate": 1.38038589873785e-05, "loss": 0.0394, "step": 8642 }, { "epoch": 0.937012142237641, "grad_norm": 2.3065123558044434, "learning_rate": 1.3803133613811114e-05, "loss": 0.0344, "step": 8643 }, { "epoch": 0.9371205550737207, "grad_norm": 1.2583224773406982, "learning_rate": 1.3802408240243728e-05, "loss": 0.0155, "step": 8644 }, { "epoch": 0.9372289679098005, "grad_norm": 3.333574056625366, "learning_rate": 1.380168286667634e-05, "loss": 0.0454, "step": 8645 }, { "epoch": 0.9373373807458804, "grad_norm": 2.9487388134002686, "learning_rate": 1.3800957493108953e-05, "loss": 0.0363, "step": 8646 }, { "epoch": 0.9374457935819601, "grad_norm": 0.4853103756904602, "learning_rate": 1.3800232119541566e-05, "loss": 0.0268, "step": 8647 }, { "epoch": 0.9375542064180399, "grad_norm": 2.717064619064331, "learning_rate": 1.3799506745974179e-05, "loss": 0.0222, "step": 8648 }, { "epoch": 0.9376626192541196, "grad_norm": 3.1742563247680664, "learning_rate": 1.379878137240679e-05, "loss": 0.0164, "step": 8649 }, { "epoch": 0.9377710320901995, "grad_norm": 1.8603121042251587, "learning_rate": 1.3798055998839403e-05, "loss": 0.0108, "step": 8650 }, { "epoch": 0.9378794449262793, "grad_norm": 3.208526611328125, "learning_rate": 1.3797330625272017e-05, "loss": 0.0515, "step": 8651 }, { "epoch": 0.937987857762359, "grad_norm": 3.0374093055725098, "learning_rate": 1.3796605251704629e-05, "loss": 0.028, "step": 8652 }, { "epoch": 0.9380962705984388, "grad_norm": 4.205315589904785, "learning_rate": 1.3795879878137241e-05, "loss": 0.1284, "step": 8653 }, { "epoch": 0.9382046834345187, "grad_norm": 3.559767723083496, "learning_rate": 1.3795154504569855e-05, "loss": 0.0393, "step": 8654 }, { "epoch": 0.9383130962705984, "grad_norm": 1.1579105854034424, "learning_rate": 1.3794429131002467e-05, "loss": 0.0208, "step": 8655 }, { "epoch": 0.9384215091066782, "grad_norm": 2.566239356994629, "learning_rate": 1.3793703757435079e-05, "loss": 0.0064, "step": 8656 }, { "epoch": 0.938529921942758, "grad_norm": 1.6096422672271729, "learning_rate": 1.3792978383867693e-05, "loss": 0.0443, "step": 8657 }, { "epoch": 0.9386383347788378, "grad_norm": 1.0719581842422485, "learning_rate": 1.3792253010300305e-05, "loss": 0.0311, "step": 8658 }, { "epoch": 0.9387467476149176, "grad_norm": 0.5227487087249756, "learning_rate": 1.3791527636732917e-05, "loss": 0.0024, "step": 8659 }, { "epoch": 0.9388551604509974, "grad_norm": 2.52437162399292, "learning_rate": 1.3790802263165531e-05, "loss": 0.0312, "step": 8660 }, { "epoch": 0.9389635732870771, "grad_norm": 2.1950082778930664, "learning_rate": 1.3790076889598145e-05, "loss": 0.0313, "step": 8661 }, { "epoch": 0.939071986123157, "grad_norm": 1.439882516860962, "learning_rate": 1.3789351516030759e-05, "loss": 0.0174, "step": 8662 }, { "epoch": 0.9391803989592368, "grad_norm": 1.0356101989746094, "learning_rate": 1.378862614246337e-05, "loss": 0.0402, "step": 8663 }, { "epoch": 0.9392888117953165, "grad_norm": 3.5007166862487793, "learning_rate": 1.3787900768895983e-05, "loss": 0.0642, "step": 8664 }, { "epoch": 0.9393972246313964, "grad_norm": 5.72214412689209, "learning_rate": 1.3787175395328595e-05, "loss": 0.0687, "step": 8665 }, { "epoch": 0.9395056374674762, "grad_norm": 0.5931270122528076, "learning_rate": 1.3786450021761209e-05, "loss": 0.0047, "step": 8666 }, { "epoch": 0.9396140503035559, "grad_norm": 2.3664627075195312, "learning_rate": 1.3785724648193821e-05, "loss": 0.0322, "step": 8667 }, { "epoch": 0.9397224631396357, "grad_norm": 2.1873881816864014, "learning_rate": 1.3784999274626433e-05, "loss": 0.0278, "step": 8668 }, { "epoch": 0.9398308759757156, "grad_norm": 2.037238836288452, "learning_rate": 1.3784273901059047e-05, "loss": 0.0265, "step": 8669 }, { "epoch": 0.9399392888117953, "grad_norm": 1.107082724571228, "learning_rate": 1.378354852749166e-05, "loss": 0.0106, "step": 8670 }, { "epoch": 0.9400477016478751, "grad_norm": 3.120319128036499, "learning_rate": 1.3782823153924271e-05, "loss": 0.047, "step": 8671 }, { "epoch": 0.9401561144839549, "grad_norm": 2.11509108543396, "learning_rate": 1.3782097780356885e-05, "loss": 0.0379, "step": 8672 }, { "epoch": 0.9402645273200347, "grad_norm": 1.1125235557556152, "learning_rate": 1.3781372406789497e-05, "loss": 0.0062, "step": 8673 }, { "epoch": 0.9403729401561145, "grad_norm": 3.920182704925537, "learning_rate": 1.378064703322211e-05, "loss": 0.0883, "step": 8674 }, { "epoch": 0.9404813529921943, "grad_norm": 0.4655478000640869, "learning_rate": 1.3779921659654723e-05, "loss": 0.0092, "step": 8675 }, { "epoch": 0.940589765828274, "grad_norm": 2.6178464889526367, "learning_rate": 1.3779196286087335e-05, "loss": 0.0568, "step": 8676 }, { "epoch": 0.9406981786643539, "grad_norm": 9.765904426574707, "learning_rate": 1.3778470912519948e-05, "loss": 0.0398, "step": 8677 }, { "epoch": 0.9408065915004337, "grad_norm": 1.0716333389282227, "learning_rate": 1.3777745538952563e-05, "loss": 0.0085, "step": 8678 }, { "epoch": 0.9409150043365134, "grad_norm": 6.16426420211792, "learning_rate": 1.3777020165385175e-05, "loss": 0.042, "step": 8679 }, { "epoch": 0.9410234171725932, "grad_norm": 2.359231472015381, "learning_rate": 1.3776294791817787e-05, "loss": 0.04, "step": 8680 }, { "epoch": 0.9411318300086731, "grad_norm": 2.9035043716430664, "learning_rate": 1.3775569418250401e-05, "loss": 0.0395, "step": 8681 }, { "epoch": 0.9412402428447528, "grad_norm": 1.6066200733184814, "learning_rate": 1.3774844044683013e-05, "loss": 0.0384, "step": 8682 }, { "epoch": 0.9413486556808326, "grad_norm": 0.8219196200370789, "learning_rate": 1.3774118671115626e-05, "loss": 0.0153, "step": 8683 }, { "epoch": 0.9414570685169124, "grad_norm": 2.5092079639434814, "learning_rate": 1.377339329754824e-05, "loss": 0.0756, "step": 8684 }, { "epoch": 0.9415654813529922, "grad_norm": 0.6712855100631714, "learning_rate": 1.3772667923980852e-05, "loss": 0.0171, "step": 8685 }, { "epoch": 0.941673894189072, "grad_norm": 2.0735270977020264, "learning_rate": 1.3771942550413464e-05, "loss": 0.0468, "step": 8686 }, { "epoch": 0.9417823070251518, "grad_norm": 1.1591894626617432, "learning_rate": 1.3771217176846078e-05, "loss": 0.0191, "step": 8687 }, { "epoch": 0.9418907198612315, "grad_norm": 2.310805082321167, "learning_rate": 1.377049180327869e-05, "loss": 0.0205, "step": 8688 }, { "epoch": 0.9419991326973114, "grad_norm": 1.4172390699386597, "learning_rate": 1.3769766429711302e-05, "loss": 0.0338, "step": 8689 }, { "epoch": 0.9421075455333912, "grad_norm": 3.6367173194885254, "learning_rate": 1.3769041056143914e-05, "loss": 0.071, "step": 8690 }, { "epoch": 0.9422159583694709, "grad_norm": 2.672851800918579, "learning_rate": 1.3768315682576528e-05, "loss": 0.0225, "step": 8691 }, { "epoch": 0.9423243712055508, "grad_norm": 1.7088122367858887, "learning_rate": 1.376759030900914e-05, "loss": 0.0609, "step": 8692 }, { "epoch": 0.9424327840416306, "grad_norm": 1.7109447717666626, "learning_rate": 1.3766864935441752e-05, "loss": 0.0207, "step": 8693 }, { "epoch": 0.9425411968777103, "grad_norm": 4.5093512535095215, "learning_rate": 1.3766139561874366e-05, "loss": 0.0833, "step": 8694 }, { "epoch": 0.9426496097137901, "grad_norm": 0.8024038076400757, "learning_rate": 1.376541418830698e-05, "loss": 0.029, "step": 8695 }, { "epoch": 0.94275802254987, "grad_norm": 3.0627002716064453, "learning_rate": 1.3764688814739594e-05, "loss": 0.0615, "step": 8696 }, { "epoch": 0.9428664353859497, "grad_norm": 1.6997394561767578, "learning_rate": 1.3763963441172206e-05, "loss": 0.0285, "step": 8697 }, { "epoch": 0.9429748482220295, "grad_norm": 1.0257409811019897, "learning_rate": 1.3763238067604818e-05, "loss": 0.0273, "step": 8698 }, { "epoch": 0.9430832610581092, "grad_norm": 0.8746684789657593, "learning_rate": 1.3762512694037432e-05, "loss": 0.0202, "step": 8699 }, { "epoch": 0.9431916738941891, "grad_norm": 1.748139500617981, "learning_rate": 1.3761787320470044e-05, "loss": 0.0177, "step": 8700 }, { "epoch": 0.9433000867302689, "grad_norm": 4.069674968719482, "learning_rate": 1.3761061946902656e-05, "loss": 0.0371, "step": 8701 }, { "epoch": 0.9434084995663486, "grad_norm": 1.130999207496643, "learning_rate": 1.3760336573335268e-05, "loss": 0.0189, "step": 8702 }, { "epoch": 0.9435169124024284, "grad_norm": 4.601190090179443, "learning_rate": 1.3759611199767882e-05, "loss": 0.0184, "step": 8703 }, { "epoch": 0.9436253252385083, "grad_norm": 1.990169644355774, "learning_rate": 1.3758885826200494e-05, "loss": 0.0243, "step": 8704 }, { "epoch": 0.943733738074588, "grad_norm": 0.7260334491729736, "learning_rate": 1.3758160452633106e-05, "loss": 0.0159, "step": 8705 }, { "epoch": 0.9438421509106678, "grad_norm": 0.5143265724182129, "learning_rate": 1.375743507906572e-05, "loss": 0.0172, "step": 8706 }, { "epoch": 0.9439505637467476, "grad_norm": 2.973217248916626, "learning_rate": 1.3756709705498332e-05, "loss": 0.058, "step": 8707 }, { "epoch": 0.9440589765828274, "grad_norm": 2.945237636566162, "learning_rate": 1.3755984331930944e-05, "loss": 0.0593, "step": 8708 }, { "epoch": 0.9441673894189072, "grad_norm": 0.183694988489151, "learning_rate": 1.3755258958363558e-05, "loss": 0.0132, "step": 8709 }, { "epoch": 0.944275802254987, "grad_norm": 1.1365300416946411, "learning_rate": 1.375453358479617e-05, "loss": 0.0204, "step": 8710 }, { "epoch": 0.9443842150910667, "grad_norm": 4.964504241943359, "learning_rate": 1.3753808211228783e-05, "loss": 0.052, "step": 8711 }, { "epoch": 0.9444926279271466, "grad_norm": 3.321671724319458, "learning_rate": 1.3753082837661398e-05, "loss": 0.0295, "step": 8712 }, { "epoch": 0.9446010407632264, "grad_norm": 0.5104036927223206, "learning_rate": 1.375235746409401e-05, "loss": 0.0074, "step": 8713 }, { "epoch": 0.9447094535993061, "grad_norm": 2.095870018005371, "learning_rate": 1.3751632090526624e-05, "loss": 0.0415, "step": 8714 }, { "epoch": 0.9448178664353859, "grad_norm": 1.8825113773345947, "learning_rate": 1.3750906716959236e-05, "loss": 0.0135, "step": 8715 }, { "epoch": 0.9449262792714658, "grad_norm": 4.117137908935547, "learning_rate": 1.3750181343391848e-05, "loss": 0.0233, "step": 8716 }, { "epoch": 0.9450346921075455, "grad_norm": 2.825833320617676, "learning_rate": 1.374945596982446e-05, "loss": 0.0314, "step": 8717 }, { "epoch": 0.9451431049436253, "grad_norm": 10.094178199768066, "learning_rate": 1.3748730596257074e-05, "loss": 0.0788, "step": 8718 }, { "epoch": 0.9452515177797051, "grad_norm": 3.6362645626068115, "learning_rate": 1.3748005222689686e-05, "loss": 0.023, "step": 8719 }, { "epoch": 0.9453599306157849, "grad_norm": 4.337198734283447, "learning_rate": 1.3747279849122299e-05, "loss": 0.0299, "step": 8720 }, { "epoch": 0.9454683434518647, "grad_norm": 4.290066719055176, "learning_rate": 1.3746554475554912e-05, "loss": 0.0215, "step": 8721 }, { "epoch": 0.9455767562879445, "grad_norm": 0.8741763830184937, "learning_rate": 1.3745829101987525e-05, "loss": 0.0172, "step": 8722 }, { "epoch": 0.9456851691240243, "grad_norm": 3.047531843185425, "learning_rate": 1.3745103728420137e-05, "loss": 0.0435, "step": 8723 }, { "epoch": 0.9457935819601041, "grad_norm": 1.9945675134658813, "learning_rate": 1.374437835485275e-05, "loss": 0.0272, "step": 8724 }, { "epoch": 0.9459019947961839, "grad_norm": 1.681244969367981, "learning_rate": 1.3743652981285363e-05, "loss": 0.0172, "step": 8725 }, { "epoch": 0.9460104076322636, "grad_norm": 1.264197587966919, "learning_rate": 1.3742927607717975e-05, "loss": 0.02, "step": 8726 }, { "epoch": 0.9461188204683435, "grad_norm": 0.8104239106178284, "learning_rate": 1.3742202234150589e-05, "loss": 0.0244, "step": 8727 }, { "epoch": 0.9462272333044233, "grad_norm": 0.4884665608406067, "learning_rate": 1.37414768605832e-05, "loss": 0.0072, "step": 8728 }, { "epoch": 0.946335646140503, "grad_norm": 4.441947937011719, "learning_rate": 1.3740751487015813e-05, "loss": 0.0561, "step": 8729 }, { "epoch": 0.9464440589765828, "grad_norm": 3.6516642570495605, "learning_rate": 1.3740026113448428e-05, "loss": 0.073, "step": 8730 }, { "epoch": 0.9465524718126627, "grad_norm": 1.2199214696884155, "learning_rate": 1.373930073988104e-05, "loss": 0.022, "step": 8731 }, { "epoch": 0.9466608846487424, "grad_norm": 1.9491091966629028, "learning_rate": 1.3738575366313653e-05, "loss": 0.0477, "step": 8732 }, { "epoch": 0.9467692974848222, "grad_norm": 3.5063397884368896, "learning_rate": 1.3737849992746267e-05, "loss": 0.0537, "step": 8733 }, { "epoch": 0.946877710320902, "grad_norm": 1.2264657020568848, "learning_rate": 1.3737124619178879e-05, "loss": 0.0402, "step": 8734 }, { "epoch": 0.9469861231569818, "grad_norm": 4.304168224334717, "learning_rate": 1.3736399245611491e-05, "loss": 0.0526, "step": 8735 }, { "epoch": 0.9470945359930616, "grad_norm": 1.9382134675979614, "learning_rate": 1.3735673872044105e-05, "loss": 0.0635, "step": 8736 }, { "epoch": 0.9472029488291414, "grad_norm": 2.251389980316162, "learning_rate": 1.3734948498476717e-05, "loss": 0.038, "step": 8737 }, { "epoch": 0.9473113616652211, "grad_norm": 1.6755189895629883, "learning_rate": 1.3734223124909329e-05, "loss": 0.0199, "step": 8738 }, { "epoch": 0.947419774501301, "grad_norm": 4.564231872558594, "learning_rate": 1.3733497751341943e-05, "loss": 0.0979, "step": 8739 }, { "epoch": 0.9475281873373808, "grad_norm": 5.46632719039917, "learning_rate": 1.3732772377774555e-05, "loss": 0.0801, "step": 8740 }, { "epoch": 0.9476366001734605, "grad_norm": 1.2171063423156738, "learning_rate": 1.3732047004207167e-05, "loss": 0.0178, "step": 8741 }, { "epoch": 0.9477450130095403, "grad_norm": 1.8411039113998413, "learning_rate": 1.373132163063978e-05, "loss": 0.0266, "step": 8742 }, { "epoch": 0.9478534258456202, "grad_norm": 3.8140311241149902, "learning_rate": 1.3730596257072393e-05, "loss": 0.036, "step": 8743 }, { "epoch": 0.9479618386816999, "grad_norm": 2.2947494983673096, "learning_rate": 1.3729870883505005e-05, "loss": 0.0521, "step": 8744 }, { "epoch": 0.9480702515177797, "grad_norm": 3.28240966796875, "learning_rate": 1.3729145509937617e-05, "loss": 0.0648, "step": 8745 }, { "epoch": 0.9481786643538594, "grad_norm": 2.4790663719177246, "learning_rate": 1.3728420136370231e-05, "loss": 0.0376, "step": 8746 }, { "epoch": 0.9482870771899393, "grad_norm": 2.6620280742645264, "learning_rate": 1.3727694762802845e-05, "loss": 0.0257, "step": 8747 }, { "epoch": 0.9483954900260191, "grad_norm": 7.854915142059326, "learning_rate": 1.3726969389235459e-05, "loss": 0.0619, "step": 8748 }, { "epoch": 0.9485039028620988, "grad_norm": 1.8568439483642578, "learning_rate": 1.3726244015668071e-05, "loss": 0.0568, "step": 8749 }, { "epoch": 0.9486123156981786, "grad_norm": 2.039388656616211, "learning_rate": 1.3725518642100683e-05, "loss": 0.0439, "step": 8750 }, { "epoch": 0.9487207285342585, "grad_norm": 3.5202527046203613, "learning_rate": 1.3724793268533297e-05, "loss": 0.0534, "step": 8751 }, { "epoch": 0.9488291413703382, "grad_norm": 2.2609691619873047, "learning_rate": 1.372406789496591e-05, "loss": 0.0305, "step": 8752 }, { "epoch": 0.948937554206418, "grad_norm": 0.8456473350524902, "learning_rate": 1.3723342521398521e-05, "loss": 0.0275, "step": 8753 }, { "epoch": 0.9490459670424979, "grad_norm": 0.6529988050460815, "learning_rate": 1.3722617147831133e-05, "loss": 0.0153, "step": 8754 }, { "epoch": 0.9491543798785776, "grad_norm": 2.097604990005493, "learning_rate": 1.3721891774263747e-05, "loss": 0.039, "step": 8755 }, { "epoch": 0.9492627927146574, "grad_norm": 1.1402517557144165, "learning_rate": 1.372116640069636e-05, "loss": 0.0368, "step": 8756 }, { "epoch": 0.9493712055507372, "grad_norm": 2.3556151390075684, "learning_rate": 1.3720441027128972e-05, "loss": 0.0455, "step": 8757 }, { "epoch": 0.949479618386817, "grad_norm": 1.5823620557785034, "learning_rate": 1.3719715653561585e-05, "loss": 0.0406, "step": 8758 }, { "epoch": 0.9495880312228968, "grad_norm": 0.955348551273346, "learning_rate": 1.3718990279994198e-05, "loss": 0.0161, "step": 8759 }, { "epoch": 0.9496964440589766, "grad_norm": 2.105759859085083, "learning_rate": 1.371826490642681e-05, "loss": 0.0219, "step": 8760 }, { "epoch": 0.9498048568950563, "grad_norm": 4.6040825843811035, "learning_rate": 1.3717539532859424e-05, "loss": 0.0278, "step": 8761 }, { "epoch": 0.9499132697311362, "grad_norm": 2.085900068283081, "learning_rate": 1.3716814159292036e-05, "loss": 0.0349, "step": 8762 }, { "epoch": 0.950021682567216, "grad_norm": 1.206300139427185, "learning_rate": 1.3716088785724648e-05, "loss": 0.0255, "step": 8763 }, { "epoch": 0.9501300954032957, "grad_norm": 1.1789714097976685, "learning_rate": 1.3715363412157263e-05, "loss": 0.0245, "step": 8764 }, { "epoch": 0.9502385082393755, "grad_norm": 2.453028678894043, "learning_rate": 1.3714638038589876e-05, "loss": 0.0441, "step": 8765 }, { "epoch": 0.9503469210754554, "grad_norm": 6.859647274017334, "learning_rate": 1.371391266502249e-05, "loss": 0.0404, "step": 8766 }, { "epoch": 0.9504553339115351, "grad_norm": 0.8797872066497803, "learning_rate": 1.3713187291455101e-05, "loss": 0.0135, "step": 8767 }, { "epoch": 0.9505637467476149, "grad_norm": 1.6072499752044678, "learning_rate": 1.3712461917887714e-05, "loss": 0.0519, "step": 8768 }, { "epoch": 0.9506721595836947, "grad_norm": 1.6069551706314087, "learning_rate": 1.3711736544320326e-05, "loss": 0.0366, "step": 8769 }, { "epoch": 0.9507805724197745, "grad_norm": 1.01071298122406, "learning_rate": 1.371101117075294e-05, "loss": 0.0312, "step": 8770 }, { "epoch": 0.9508889852558543, "grad_norm": 3.4233224391937256, "learning_rate": 1.3710285797185552e-05, "loss": 0.0166, "step": 8771 }, { "epoch": 0.9509973980919341, "grad_norm": 3.4483115673065186, "learning_rate": 1.3709560423618164e-05, "loss": 0.0256, "step": 8772 }, { "epoch": 0.9511058109280138, "grad_norm": 3.744981288909912, "learning_rate": 1.3708835050050778e-05, "loss": 0.0612, "step": 8773 }, { "epoch": 0.9512142237640937, "grad_norm": 0.9706977605819702, "learning_rate": 1.370810967648339e-05, "loss": 0.0194, "step": 8774 }, { "epoch": 0.9513226366001735, "grad_norm": 3.4024510383605957, "learning_rate": 1.3707384302916002e-05, "loss": 0.0506, "step": 8775 }, { "epoch": 0.9514310494362532, "grad_norm": 0.4584977924823761, "learning_rate": 1.3706658929348616e-05, "loss": 0.0195, "step": 8776 }, { "epoch": 0.951539462272333, "grad_norm": 0.8538906574249268, "learning_rate": 1.3705933555781228e-05, "loss": 0.014, "step": 8777 }, { "epoch": 0.9516478751084129, "grad_norm": 5.726755142211914, "learning_rate": 1.370520818221384e-05, "loss": 0.036, "step": 8778 }, { "epoch": 0.9517562879444926, "grad_norm": 2.045536518096924, "learning_rate": 1.3704482808646454e-05, "loss": 0.0482, "step": 8779 }, { "epoch": 0.9518647007805724, "grad_norm": 1.1831265687942505, "learning_rate": 1.3703757435079066e-05, "loss": 0.0046, "step": 8780 }, { "epoch": 0.9519731136166523, "grad_norm": 1.1120874881744385, "learning_rate": 1.3703032061511678e-05, "loss": 0.0294, "step": 8781 }, { "epoch": 0.952081526452732, "grad_norm": 1.7151798009872437, "learning_rate": 1.3702306687944294e-05, "loss": 0.0385, "step": 8782 }, { "epoch": 0.9521899392888118, "grad_norm": 2.820746898651123, "learning_rate": 1.3701581314376906e-05, "loss": 0.0549, "step": 8783 }, { "epoch": 0.9522983521248916, "grad_norm": 0.43040722608566284, "learning_rate": 1.3700855940809518e-05, "loss": 0.0086, "step": 8784 }, { "epoch": 0.9524067649609714, "grad_norm": 6.909896373748779, "learning_rate": 1.3700130567242132e-05, "loss": 0.058, "step": 8785 }, { "epoch": 0.9525151777970512, "grad_norm": 2.143156051635742, "learning_rate": 1.3699405193674744e-05, "loss": 0.0698, "step": 8786 }, { "epoch": 0.952623590633131, "grad_norm": 1.9314463138580322, "learning_rate": 1.3698679820107356e-05, "loss": 0.0374, "step": 8787 }, { "epoch": 0.9527320034692107, "grad_norm": 0.3942400813102722, "learning_rate": 1.369795444653997e-05, "loss": 0.0036, "step": 8788 }, { "epoch": 0.9528404163052906, "grad_norm": 3.709426164627075, "learning_rate": 1.3697229072972582e-05, "loss": 0.0643, "step": 8789 }, { "epoch": 0.9529488291413704, "grad_norm": 1.5048754215240479, "learning_rate": 1.3696503699405194e-05, "loss": 0.0281, "step": 8790 }, { "epoch": 0.9530572419774501, "grad_norm": 2.4995555877685547, "learning_rate": 1.3695778325837808e-05, "loss": 0.039, "step": 8791 }, { "epoch": 0.9531656548135299, "grad_norm": 2.2991764545440674, "learning_rate": 1.369505295227042e-05, "loss": 0.0266, "step": 8792 }, { "epoch": 0.9532740676496098, "grad_norm": 1.1400502920150757, "learning_rate": 1.3694327578703032e-05, "loss": 0.0263, "step": 8793 }, { "epoch": 0.9533824804856895, "grad_norm": 1.1217411756515503, "learning_rate": 1.3693602205135645e-05, "loss": 0.0545, "step": 8794 }, { "epoch": 0.9534908933217693, "grad_norm": 1.9086617231369019, "learning_rate": 1.3692876831568258e-05, "loss": 0.0262, "step": 8795 }, { "epoch": 0.953599306157849, "grad_norm": 2.176400661468506, "learning_rate": 1.369215145800087e-05, "loss": 0.0176, "step": 8796 }, { "epoch": 0.9537077189939289, "grad_norm": 2.545994997024536, "learning_rate": 1.3691426084433483e-05, "loss": 0.0628, "step": 8797 }, { "epoch": 0.9538161318300087, "grad_norm": 2.658336639404297, "learning_rate": 1.3690700710866097e-05, "loss": 0.056, "step": 8798 }, { "epoch": 0.9539245446660884, "grad_norm": 3.541633129119873, "learning_rate": 1.368997533729871e-05, "loss": 0.0228, "step": 8799 }, { "epoch": 0.9540329575021682, "grad_norm": 1.6553654670715332, "learning_rate": 1.3689249963731324e-05, "loss": 0.0196, "step": 8800 }, { "epoch": 0.9541413703382481, "grad_norm": 2.475940465927124, "learning_rate": 1.3688524590163936e-05, "loss": 0.0565, "step": 8801 }, { "epoch": 0.9542497831743278, "grad_norm": 3.4796574115753174, "learning_rate": 1.3687799216596549e-05, "loss": 0.0336, "step": 8802 }, { "epoch": 0.9543581960104076, "grad_norm": 2.117178440093994, "learning_rate": 1.3687073843029162e-05, "loss": 0.0283, "step": 8803 }, { "epoch": 0.9544666088464874, "grad_norm": 0.7021660208702087, "learning_rate": 1.3686348469461775e-05, "loss": 0.0101, "step": 8804 }, { "epoch": 0.9545750216825672, "grad_norm": 1.7048717737197876, "learning_rate": 1.3685623095894387e-05, "loss": 0.0187, "step": 8805 }, { "epoch": 0.954683434518647, "grad_norm": 3.548799753189087, "learning_rate": 1.3684897722326999e-05, "loss": 0.0436, "step": 8806 }, { "epoch": 0.9547918473547268, "grad_norm": 1.5603982210159302, "learning_rate": 1.3684172348759613e-05, "loss": 0.0479, "step": 8807 }, { "epoch": 0.9549002601908065, "grad_norm": 2.3843209743499756, "learning_rate": 1.3683446975192225e-05, "loss": 0.0316, "step": 8808 }, { "epoch": 0.9550086730268864, "grad_norm": 1.0474086999893188, "learning_rate": 1.3682721601624837e-05, "loss": 0.0149, "step": 8809 }, { "epoch": 0.9551170858629662, "grad_norm": 2.131039619445801, "learning_rate": 1.368199622805745e-05, "loss": 0.0143, "step": 8810 }, { "epoch": 0.9552254986990459, "grad_norm": 1.5740282535552979, "learning_rate": 1.3681270854490063e-05, "loss": 0.0343, "step": 8811 }, { "epoch": 0.9553339115351258, "grad_norm": 3.2507529258728027, "learning_rate": 1.3680545480922675e-05, "loss": 0.0405, "step": 8812 }, { "epoch": 0.9554423243712056, "grad_norm": 0.7673196792602539, "learning_rate": 1.3679820107355289e-05, "loss": 0.0183, "step": 8813 }, { "epoch": 0.9555507372072853, "grad_norm": 0.796749472618103, "learning_rate": 1.3679094733787901e-05, "loss": 0.0149, "step": 8814 }, { "epoch": 0.9556591500433651, "grad_norm": 2.319223403930664, "learning_rate": 1.3678369360220513e-05, "loss": 0.0445, "step": 8815 }, { "epoch": 0.955767562879445, "grad_norm": 3.097954034805298, "learning_rate": 1.3677643986653129e-05, "loss": 0.0784, "step": 8816 }, { "epoch": 0.9558759757155247, "grad_norm": 1.5678638219833374, "learning_rate": 1.3676918613085741e-05, "loss": 0.0249, "step": 8817 }, { "epoch": 0.9559843885516045, "grad_norm": 0.7720779776573181, "learning_rate": 1.3676193239518355e-05, "loss": 0.0067, "step": 8818 }, { "epoch": 0.9560928013876843, "grad_norm": 6.009801864624023, "learning_rate": 1.3675467865950967e-05, "loss": 0.0636, "step": 8819 }, { "epoch": 0.9562012142237641, "grad_norm": 1.3318058252334595, "learning_rate": 1.3674742492383579e-05, "loss": 0.0101, "step": 8820 }, { "epoch": 0.9563096270598439, "grad_norm": 0.4443390369415283, "learning_rate": 1.3674017118816191e-05, "loss": 0.0021, "step": 8821 }, { "epoch": 0.9564180398959237, "grad_norm": 2.316166877746582, "learning_rate": 1.3673291745248805e-05, "loss": 0.0606, "step": 8822 }, { "epoch": 0.9565264527320034, "grad_norm": 2.324476957321167, "learning_rate": 1.3672566371681417e-05, "loss": 0.0343, "step": 8823 }, { "epoch": 0.9566348655680833, "grad_norm": 3.3300206661224365, "learning_rate": 1.367184099811403e-05, "loss": 0.0584, "step": 8824 }, { "epoch": 0.9567432784041631, "grad_norm": 5.443485260009766, "learning_rate": 1.3671115624546643e-05, "loss": 0.024, "step": 8825 }, { "epoch": 0.9568516912402428, "grad_norm": 0.46453744173049927, "learning_rate": 1.3670390250979255e-05, "loss": 0.0069, "step": 8826 }, { "epoch": 0.9569601040763226, "grad_norm": 3.125852346420288, "learning_rate": 1.3669664877411867e-05, "loss": 0.0782, "step": 8827 }, { "epoch": 0.9570685169124025, "grad_norm": 5.236395359039307, "learning_rate": 1.3668939503844481e-05, "loss": 0.0172, "step": 8828 }, { "epoch": 0.9571769297484822, "grad_norm": 0.4056030511856079, "learning_rate": 1.3668214130277093e-05, "loss": 0.0109, "step": 8829 }, { "epoch": 0.957285342584562, "grad_norm": 0.9813648462295532, "learning_rate": 1.3667488756709705e-05, "loss": 0.0311, "step": 8830 }, { "epoch": 0.9573937554206418, "grad_norm": 7.5233259201049805, "learning_rate": 1.366676338314232e-05, "loss": 0.1443, "step": 8831 }, { "epoch": 0.9575021682567216, "grad_norm": 1.9243334531784058, "learning_rate": 1.3666038009574931e-05, "loss": 0.0349, "step": 8832 }, { "epoch": 0.9576105810928014, "grad_norm": 5.421128749847412, "learning_rate": 1.3665312636007545e-05, "loss": 0.0392, "step": 8833 }, { "epoch": 0.9577189939288812, "grad_norm": 4.5902557373046875, "learning_rate": 1.3664587262440159e-05, "loss": 0.0449, "step": 8834 }, { "epoch": 0.9578274067649609, "grad_norm": 1.679832935333252, "learning_rate": 1.3663861888872771e-05, "loss": 0.0212, "step": 8835 }, { "epoch": 0.9579358196010408, "grad_norm": 3.0691444873809814, "learning_rate": 1.3663136515305383e-05, "loss": 0.0725, "step": 8836 }, { "epoch": 0.9580442324371206, "grad_norm": 0.6532723903656006, "learning_rate": 1.3662411141737997e-05, "loss": 0.0114, "step": 8837 }, { "epoch": 0.9581526452732003, "grad_norm": 1.9546966552734375, "learning_rate": 1.366168576817061e-05, "loss": 0.0225, "step": 8838 }, { "epoch": 0.9582610581092801, "grad_norm": 2.506354331970215, "learning_rate": 1.3660960394603222e-05, "loss": 0.0392, "step": 8839 }, { "epoch": 0.95836947094536, "grad_norm": 1.9289311170578003, "learning_rate": 1.3660235021035835e-05, "loss": 0.0182, "step": 8840 }, { "epoch": 0.9584778837814397, "grad_norm": 0.4771251678466797, "learning_rate": 1.3659509647468448e-05, "loss": 0.0128, "step": 8841 }, { "epoch": 0.9585862966175195, "grad_norm": 3.934729814529419, "learning_rate": 1.365878427390106e-05, "loss": 0.0771, "step": 8842 }, { "epoch": 0.9586947094535994, "grad_norm": 5.915515899658203, "learning_rate": 1.3658058900333674e-05, "loss": 0.0336, "step": 8843 }, { "epoch": 0.9588031222896791, "grad_norm": 1.849253535270691, "learning_rate": 1.3657333526766286e-05, "loss": 0.0671, "step": 8844 }, { "epoch": 0.9589115351257589, "grad_norm": 3.3380091190338135, "learning_rate": 1.3656608153198898e-05, "loss": 0.031, "step": 8845 }, { "epoch": 0.9590199479618386, "grad_norm": 1.9227687120437622, "learning_rate": 1.365588277963151e-05, "loss": 0.019, "step": 8846 }, { "epoch": 0.9591283607979185, "grad_norm": 4.1858439445495605, "learning_rate": 1.3655157406064124e-05, "loss": 0.0447, "step": 8847 }, { "epoch": 0.9592367736339983, "grad_norm": 2.1918437480926514, "learning_rate": 1.3654432032496736e-05, "loss": 0.0482, "step": 8848 }, { "epoch": 0.959345186470078, "grad_norm": 4.9768385887146, "learning_rate": 1.3653706658929348e-05, "loss": 0.0849, "step": 8849 }, { "epoch": 0.9594535993061578, "grad_norm": 2.8469204902648926, "learning_rate": 1.3652981285361962e-05, "loss": 0.0596, "step": 8850 }, { "epoch": 0.9595620121422377, "grad_norm": 3.170682907104492, "learning_rate": 1.3652255911794576e-05, "loss": 0.0512, "step": 8851 }, { "epoch": 0.9596704249783174, "grad_norm": 0.7169297933578491, "learning_rate": 1.365153053822719e-05, "loss": 0.0266, "step": 8852 }, { "epoch": 0.9597788378143972, "grad_norm": 4.018579483032227, "learning_rate": 1.3650805164659802e-05, "loss": 0.042, "step": 8853 }, { "epoch": 0.959887250650477, "grad_norm": 1.1496548652648926, "learning_rate": 1.3650079791092414e-05, "loss": 0.0243, "step": 8854 }, { "epoch": 0.9599956634865568, "grad_norm": 2.692201614379883, "learning_rate": 1.3649354417525028e-05, "loss": 0.0185, "step": 8855 }, { "epoch": 0.9601040763226366, "grad_norm": 1.264520287513733, "learning_rate": 1.364862904395764e-05, "loss": 0.0219, "step": 8856 }, { "epoch": 0.9602124891587164, "grad_norm": 1.7952934503555298, "learning_rate": 1.3647903670390252e-05, "loss": 0.0494, "step": 8857 }, { "epoch": 0.9603209019947961, "grad_norm": 4.552087783813477, "learning_rate": 1.3647178296822866e-05, "loss": 0.0455, "step": 8858 }, { "epoch": 0.960429314830876, "grad_norm": 3.8385672569274902, "learning_rate": 1.3646452923255478e-05, "loss": 0.0423, "step": 8859 }, { "epoch": 0.9605377276669558, "grad_norm": 1.4164202213287354, "learning_rate": 1.364572754968809e-05, "loss": 0.0189, "step": 8860 }, { "epoch": 0.9606461405030355, "grad_norm": 3.012514352798462, "learning_rate": 1.3645002176120702e-05, "loss": 0.0191, "step": 8861 }, { "epoch": 0.9607545533391153, "grad_norm": 1.9885077476501465, "learning_rate": 1.3644276802553316e-05, "loss": 0.0247, "step": 8862 }, { "epoch": 0.9608629661751952, "grad_norm": 2.0201046466827393, "learning_rate": 1.3643551428985928e-05, "loss": 0.0232, "step": 8863 }, { "epoch": 0.9609713790112749, "grad_norm": 2.8973491191864014, "learning_rate": 1.364282605541854e-05, "loss": 0.0338, "step": 8864 }, { "epoch": 0.9610797918473547, "grad_norm": 3.9712023735046387, "learning_rate": 1.3642100681851154e-05, "loss": 0.0458, "step": 8865 }, { "epoch": 0.9611882046834345, "grad_norm": 1.3805431127548218, "learning_rate": 1.3641375308283766e-05, "loss": 0.0352, "step": 8866 }, { "epoch": 0.9612966175195143, "grad_norm": 0.8995699882507324, "learning_rate": 1.3640649934716379e-05, "loss": 0.0282, "step": 8867 }, { "epoch": 0.9614050303555941, "grad_norm": 4.317062854766846, "learning_rate": 1.3639924561148994e-05, "loss": 0.0486, "step": 8868 }, { "epoch": 0.9615134431916739, "grad_norm": 2.60315203666687, "learning_rate": 1.3639199187581606e-05, "loss": 0.0253, "step": 8869 }, { "epoch": 0.9616218560277537, "grad_norm": 3.8597378730773926, "learning_rate": 1.363847381401422e-05, "loss": 0.0144, "step": 8870 }, { "epoch": 0.9617302688638335, "grad_norm": 2.5610554218292236, "learning_rate": 1.3637748440446832e-05, "loss": 0.0384, "step": 8871 }, { "epoch": 0.9618386816999133, "grad_norm": 2.809450149536133, "learning_rate": 1.3637023066879444e-05, "loss": 0.0326, "step": 8872 }, { "epoch": 0.961947094535993, "grad_norm": 1.582634687423706, "learning_rate": 1.3636297693312056e-05, "loss": 0.0191, "step": 8873 }, { "epoch": 0.9620555073720729, "grad_norm": 8.217493057250977, "learning_rate": 1.363557231974467e-05, "loss": 0.0513, "step": 8874 }, { "epoch": 0.9621639202081527, "grad_norm": 1.0418193340301514, "learning_rate": 1.3634846946177282e-05, "loss": 0.0221, "step": 8875 }, { "epoch": 0.9622723330442324, "grad_norm": 2.781264543533325, "learning_rate": 1.3634121572609895e-05, "loss": 0.0283, "step": 8876 }, { "epoch": 0.9623807458803122, "grad_norm": 0.9480435848236084, "learning_rate": 1.3633396199042508e-05, "loss": 0.0176, "step": 8877 }, { "epoch": 0.9624891587163921, "grad_norm": 2.068311929702759, "learning_rate": 1.363267082547512e-05, "loss": 0.0237, "step": 8878 }, { "epoch": 0.9625975715524718, "grad_norm": 3.0408527851104736, "learning_rate": 1.3631945451907733e-05, "loss": 0.036, "step": 8879 }, { "epoch": 0.9627059843885516, "grad_norm": 2.5869226455688477, "learning_rate": 1.3631220078340347e-05, "loss": 0.0786, "step": 8880 }, { "epoch": 0.9628143972246314, "grad_norm": 2.837805986404419, "learning_rate": 1.3630494704772959e-05, "loss": 0.0572, "step": 8881 }, { "epoch": 0.9629228100607112, "grad_norm": 2.429102897644043, "learning_rate": 1.362976933120557e-05, "loss": 0.0481, "step": 8882 }, { "epoch": 0.963031222896791, "grad_norm": 0.50271075963974, "learning_rate": 1.3629043957638185e-05, "loss": 0.0212, "step": 8883 }, { "epoch": 0.9631396357328708, "grad_norm": 1.0291404724121094, "learning_rate": 1.3628318584070797e-05, "loss": 0.0243, "step": 8884 }, { "epoch": 0.9632480485689505, "grad_norm": 7.060874938964844, "learning_rate": 1.362759321050341e-05, "loss": 0.036, "step": 8885 }, { "epoch": 0.9633564614050304, "grad_norm": 4.0897746086120605, "learning_rate": 1.3626867836936024e-05, "loss": 0.037, "step": 8886 }, { "epoch": 0.9634648742411102, "grad_norm": 3.0325896739959717, "learning_rate": 1.3626142463368637e-05, "loss": 0.0307, "step": 8887 }, { "epoch": 0.9635732870771899, "grad_norm": 3.2308077812194824, "learning_rate": 1.3625417089801249e-05, "loss": 0.0479, "step": 8888 }, { "epoch": 0.9636816999132697, "grad_norm": 1.6033622026443481, "learning_rate": 1.3624691716233863e-05, "loss": 0.0585, "step": 8889 }, { "epoch": 0.9637901127493496, "grad_norm": 0.5644727945327759, "learning_rate": 1.3623966342666475e-05, "loss": 0.0126, "step": 8890 }, { "epoch": 0.9638985255854293, "grad_norm": 0.7946133613586426, "learning_rate": 1.3623240969099087e-05, "loss": 0.0339, "step": 8891 }, { "epoch": 0.9640069384215091, "grad_norm": 2.5621135234832764, "learning_rate": 1.36225155955317e-05, "loss": 0.0537, "step": 8892 }, { "epoch": 0.9641153512575888, "grad_norm": 0.8609674572944641, "learning_rate": 1.3621790221964313e-05, "loss": 0.011, "step": 8893 }, { "epoch": 0.9642237640936687, "grad_norm": 2.15120792388916, "learning_rate": 1.3621064848396925e-05, "loss": 0.0499, "step": 8894 }, { "epoch": 0.9643321769297485, "grad_norm": 2.534268617630005, "learning_rate": 1.3620339474829539e-05, "loss": 0.053, "step": 8895 }, { "epoch": 0.9644405897658282, "grad_norm": 4.114903450012207, "learning_rate": 1.3619614101262151e-05, "loss": 0.0384, "step": 8896 }, { "epoch": 0.964549002601908, "grad_norm": 2.577876329421997, "learning_rate": 1.3618888727694763e-05, "loss": 0.014, "step": 8897 }, { "epoch": 0.9646574154379879, "grad_norm": 1.324697732925415, "learning_rate": 1.3618163354127375e-05, "loss": 0.027, "step": 8898 }, { "epoch": 0.9647658282740676, "grad_norm": 1.4715442657470703, "learning_rate": 1.3617437980559989e-05, "loss": 0.0476, "step": 8899 }, { "epoch": 0.9648742411101474, "grad_norm": 1.5662708282470703, "learning_rate": 1.3616712606992601e-05, "loss": 0.0253, "step": 8900 }, { "epoch": 0.9649826539462273, "grad_norm": 1.0804553031921387, "learning_rate": 1.3615987233425213e-05, "loss": 0.0048, "step": 8901 }, { "epoch": 0.965091066782307, "grad_norm": 1.157240867614746, "learning_rate": 1.3615261859857829e-05, "loss": 0.0223, "step": 8902 }, { "epoch": 0.9651994796183868, "grad_norm": 1.146071434020996, "learning_rate": 1.3614536486290441e-05, "loss": 0.0137, "step": 8903 }, { "epoch": 0.9653078924544666, "grad_norm": 1.2459676265716553, "learning_rate": 1.3613811112723055e-05, "loss": 0.0294, "step": 8904 }, { "epoch": 0.9654163052905465, "grad_norm": 1.9106724262237549, "learning_rate": 1.3613085739155667e-05, "loss": 0.056, "step": 8905 }, { "epoch": 0.9655247181266262, "grad_norm": 0.8968461155891418, "learning_rate": 1.361236036558828e-05, "loss": 0.0349, "step": 8906 }, { "epoch": 0.965633130962706, "grad_norm": 3.178011655807495, "learning_rate": 1.3611634992020893e-05, "loss": 0.0931, "step": 8907 }, { "epoch": 0.9657415437987857, "grad_norm": 2.632952928543091, "learning_rate": 1.3610909618453505e-05, "loss": 0.0529, "step": 8908 }, { "epoch": 0.9658499566348656, "grad_norm": 3.180612087249756, "learning_rate": 1.3610184244886117e-05, "loss": 0.0817, "step": 8909 }, { "epoch": 0.9659583694709454, "grad_norm": 1.5131746530532837, "learning_rate": 1.3609458871318731e-05, "loss": 0.0218, "step": 8910 }, { "epoch": 0.9660667823070251, "grad_norm": 1.3970588445663452, "learning_rate": 1.3608733497751343e-05, "loss": 0.0141, "step": 8911 }, { "epoch": 0.9661751951431049, "grad_norm": 1.451733946800232, "learning_rate": 1.3608008124183955e-05, "loss": 0.0293, "step": 8912 }, { "epoch": 0.9662836079791848, "grad_norm": 2.696319341659546, "learning_rate": 1.3607282750616568e-05, "loss": 0.0339, "step": 8913 }, { "epoch": 0.9663920208152645, "grad_norm": 2.170839548110962, "learning_rate": 1.3606557377049181e-05, "loss": 0.0387, "step": 8914 }, { "epoch": 0.9665004336513443, "grad_norm": 7.4752678871154785, "learning_rate": 1.3605832003481794e-05, "loss": 0.0662, "step": 8915 }, { "epoch": 0.9666088464874241, "grad_norm": 2.484614610671997, "learning_rate": 1.3605106629914406e-05, "loss": 0.0526, "step": 8916 }, { "epoch": 0.9667172593235039, "grad_norm": 0.7402966022491455, "learning_rate": 1.360438125634702e-05, "loss": 0.0088, "step": 8917 }, { "epoch": 0.9668256721595837, "grad_norm": 1.7108423709869385, "learning_rate": 1.3603655882779632e-05, "loss": 0.0247, "step": 8918 }, { "epoch": 0.9669340849956635, "grad_norm": 2.7388906478881836, "learning_rate": 1.3602930509212244e-05, "loss": 0.1114, "step": 8919 }, { "epoch": 0.9670424978317432, "grad_norm": 2.3648922443389893, "learning_rate": 1.360220513564486e-05, "loss": 0.041, "step": 8920 }, { "epoch": 0.9671509106678231, "grad_norm": 2.996425151824951, "learning_rate": 1.3601479762077472e-05, "loss": 0.0245, "step": 8921 }, { "epoch": 0.9672593235039029, "grad_norm": 1.473972201347351, "learning_rate": 1.3600754388510085e-05, "loss": 0.0225, "step": 8922 }, { "epoch": 0.9673677363399826, "grad_norm": 4.366210460662842, "learning_rate": 1.3600029014942698e-05, "loss": 0.047, "step": 8923 }, { "epoch": 0.9674761491760624, "grad_norm": 1.7403526306152344, "learning_rate": 1.359930364137531e-05, "loss": 0.0484, "step": 8924 }, { "epoch": 0.9675845620121423, "grad_norm": 2.5838024616241455, "learning_rate": 1.3598578267807922e-05, "loss": 0.0188, "step": 8925 }, { "epoch": 0.967692974848222, "grad_norm": 1.1756227016448975, "learning_rate": 1.3597852894240536e-05, "loss": 0.0219, "step": 8926 }, { "epoch": 0.9678013876843018, "grad_norm": 5.103893756866455, "learning_rate": 1.3597127520673148e-05, "loss": 0.0829, "step": 8927 }, { "epoch": 0.9679098005203816, "grad_norm": 2.9233973026275635, "learning_rate": 1.359640214710576e-05, "loss": 0.07, "step": 8928 }, { "epoch": 0.9680182133564614, "grad_norm": 2.2532036304473877, "learning_rate": 1.3595676773538374e-05, "loss": 0.0103, "step": 8929 }, { "epoch": 0.9681266261925412, "grad_norm": 1.9442493915557861, "learning_rate": 1.3594951399970986e-05, "loss": 0.0254, "step": 8930 }, { "epoch": 0.968235039028621, "grad_norm": 0.9825525283813477, "learning_rate": 1.3594226026403598e-05, "loss": 0.0105, "step": 8931 }, { "epoch": 0.9683434518647008, "grad_norm": 1.7190382480621338, "learning_rate": 1.3593500652836212e-05, "loss": 0.0252, "step": 8932 }, { "epoch": 0.9684518647007806, "grad_norm": 1.6534284353256226, "learning_rate": 1.3592775279268824e-05, "loss": 0.0665, "step": 8933 }, { "epoch": 0.9685602775368604, "grad_norm": 6.85966682434082, "learning_rate": 1.3592049905701436e-05, "loss": 0.0633, "step": 8934 }, { "epoch": 0.9686686903729401, "grad_norm": 2.8826205730438232, "learning_rate": 1.359132453213405e-05, "loss": 0.0389, "step": 8935 }, { "epoch": 0.96877710320902, "grad_norm": 1.0714792013168335, "learning_rate": 1.3590599158566662e-05, "loss": 0.0103, "step": 8936 }, { "epoch": 0.9688855160450998, "grad_norm": 1.665893793106079, "learning_rate": 1.3589873784999276e-05, "loss": 0.0386, "step": 8937 }, { "epoch": 0.9689939288811795, "grad_norm": 1.763328194618225, "learning_rate": 1.358914841143189e-05, "loss": 0.026, "step": 8938 }, { "epoch": 0.9691023417172593, "grad_norm": 3.1143252849578857, "learning_rate": 1.3588423037864502e-05, "loss": 0.0943, "step": 8939 }, { "epoch": 0.9692107545533392, "grad_norm": 0.7998878359794617, "learning_rate": 1.3587697664297114e-05, "loss": 0.0138, "step": 8940 }, { "epoch": 0.9693191673894189, "grad_norm": 3.677680730819702, "learning_rate": 1.3586972290729728e-05, "loss": 0.0441, "step": 8941 }, { "epoch": 0.9694275802254987, "grad_norm": 2.6754043102264404, "learning_rate": 1.358624691716234e-05, "loss": 0.0567, "step": 8942 }, { "epoch": 0.9695359930615784, "grad_norm": 1.401915192604065, "learning_rate": 1.3585521543594952e-05, "loss": 0.0157, "step": 8943 }, { "epoch": 0.9696444058976583, "grad_norm": 1.9158905744552612, "learning_rate": 1.3584796170027566e-05, "loss": 0.0194, "step": 8944 }, { "epoch": 0.9697528187337381, "grad_norm": 1.224670648574829, "learning_rate": 1.3584070796460178e-05, "loss": 0.0327, "step": 8945 }, { "epoch": 0.9698612315698178, "grad_norm": 1.0722734928131104, "learning_rate": 1.358334542289279e-05, "loss": 0.0208, "step": 8946 }, { "epoch": 0.9699696444058976, "grad_norm": 2.373718500137329, "learning_rate": 1.3582620049325404e-05, "loss": 0.0738, "step": 8947 }, { "epoch": 0.9700780572419775, "grad_norm": 2.8846237659454346, "learning_rate": 1.3581894675758016e-05, "loss": 0.0248, "step": 8948 }, { "epoch": 0.9701864700780573, "grad_norm": 2.487226963043213, "learning_rate": 1.3581169302190628e-05, "loss": 0.047, "step": 8949 }, { "epoch": 0.970294882914137, "grad_norm": 0.9232132434844971, "learning_rate": 1.358044392862324e-05, "loss": 0.0129, "step": 8950 }, { "epoch": 0.9704032957502168, "grad_norm": 9.819997787475586, "learning_rate": 1.3579718555055854e-05, "loss": 0.057, "step": 8951 }, { "epoch": 0.9705117085862967, "grad_norm": 2.7934982776641846, "learning_rate": 1.3578993181488467e-05, "loss": 0.0278, "step": 8952 }, { "epoch": 0.9706201214223764, "grad_norm": 1.790417194366455, "learning_rate": 1.3578267807921079e-05, "loss": 0.0423, "step": 8953 }, { "epoch": 0.9707285342584562, "grad_norm": 7.38576602935791, "learning_rate": 1.3577542434353694e-05, "loss": 0.0534, "step": 8954 }, { "epoch": 0.9708369470945359, "grad_norm": 0.8436934947967529, "learning_rate": 1.3576817060786306e-05, "loss": 0.0445, "step": 8955 }, { "epoch": 0.9709453599306158, "grad_norm": 3.020076036453247, "learning_rate": 1.357609168721892e-05, "loss": 0.0264, "step": 8956 }, { "epoch": 0.9710537727666956, "grad_norm": 1.0565299987792969, "learning_rate": 1.3575366313651532e-05, "loss": 0.0183, "step": 8957 }, { "epoch": 0.9711621856027753, "grad_norm": 0.9615487456321716, "learning_rate": 1.3574640940084145e-05, "loss": 0.0179, "step": 8958 }, { "epoch": 0.9712705984388552, "grad_norm": 2.2535688877105713, "learning_rate": 1.3573915566516758e-05, "loss": 0.0094, "step": 8959 }, { "epoch": 0.971379011274935, "grad_norm": 0.9424329996109009, "learning_rate": 1.357319019294937e-05, "loss": 0.011, "step": 8960 }, { "epoch": 0.9714874241110147, "grad_norm": 0.3177720606327057, "learning_rate": 1.3572464819381983e-05, "loss": 0.0065, "step": 8961 }, { "epoch": 0.9715958369470945, "grad_norm": 1.8002351522445679, "learning_rate": 1.3571739445814597e-05, "loss": 0.0094, "step": 8962 }, { "epoch": 0.9717042497831744, "grad_norm": 4.655573844909668, "learning_rate": 1.3571014072247209e-05, "loss": 0.0588, "step": 8963 }, { "epoch": 0.9718126626192541, "grad_norm": 3.0259580612182617, "learning_rate": 1.357028869867982e-05, "loss": 0.0605, "step": 8964 }, { "epoch": 0.9719210754553339, "grad_norm": 2.1011152267456055, "learning_rate": 1.3569563325112433e-05, "loss": 0.041, "step": 8965 }, { "epoch": 0.9720294882914137, "grad_norm": 1.8592205047607422, "learning_rate": 1.3568837951545047e-05, "loss": 0.0363, "step": 8966 }, { "epoch": 0.9721379011274935, "grad_norm": 3.4215800762176514, "learning_rate": 1.3568112577977659e-05, "loss": 0.0518, "step": 8967 }, { "epoch": 0.9722463139635733, "grad_norm": 3.416551113128662, "learning_rate": 1.3567387204410271e-05, "loss": 0.0517, "step": 8968 }, { "epoch": 0.9723547267996531, "grad_norm": 2.373206377029419, "learning_rate": 1.3566661830842885e-05, "loss": 0.0751, "step": 8969 }, { "epoch": 0.9724631396357328, "grad_norm": 5.436093330383301, "learning_rate": 1.3565936457275497e-05, "loss": 0.1166, "step": 8970 }, { "epoch": 0.9725715524718127, "grad_norm": 1.0993695259094238, "learning_rate": 1.356521108370811e-05, "loss": 0.019, "step": 8971 }, { "epoch": 0.9726799653078925, "grad_norm": 4.194721698760986, "learning_rate": 1.3564485710140725e-05, "loss": 0.0795, "step": 8972 }, { "epoch": 0.9727883781439722, "grad_norm": 3.2381057739257812, "learning_rate": 1.3563760336573337e-05, "loss": 0.0189, "step": 8973 }, { "epoch": 0.972896790980052, "grad_norm": 6.103817939758301, "learning_rate": 1.356303496300595e-05, "loss": 0.0389, "step": 8974 }, { "epoch": 0.9730052038161319, "grad_norm": 0.5281330943107605, "learning_rate": 1.3562309589438563e-05, "loss": 0.0228, "step": 8975 }, { "epoch": 0.9731136166522116, "grad_norm": 1.4316678047180176, "learning_rate": 1.3561584215871175e-05, "loss": 0.024, "step": 8976 }, { "epoch": 0.9732220294882914, "grad_norm": 4.764274597167969, "learning_rate": 1.3560858842303787e-05, "loss": 0.022, "step": 8977 }, { "epoch": 0.9733304423243712, "grad_norm": 2.726827383041382, "learning_rate": 1.3560133468736401e-05, "loss": 0.0438, "step": 8978 }, { "epoch": 0.973438855160451, "grad_norm": 3.138866662979126, "learning_rate": 1.3559408095169013e-05, "loss": 0.045, "step": 8979 }, { "epoch": 0.9735472679965308, "grad_norm": 5.253053188323975, "learning_rate": 1.3558682721601625e-05, "loss": 0.0303, "step": 8980 }, { "epoch": 0.9736556808326106, "grad_norm": 1.804242730140686, "learning_rate": 1.3557957348034239e-05, "loss": 0.0439, "step": 8981 }, { "epoch": 0.9737640936686903, "grad_norm": 3.846229314804077, "learning_rate": 1.3557231974466851e-05, "loss": 0.0385, "step": 8982 }, { "epoch": 0.9738725065047702, "grad_norm": 0.9603638052940369, "learning_rate": 1.3556506600899463e-05, "loss": 0.0115, "step": 8983 }, { "epoch": 0.97398091934085, "grad_norm": 3.1341769695281982, "learning_rate": 1.3555781227332077e-05, "loss": 0.0421, "step": 8984 }, { "epoch": 0.9740893321769297, "grad_norm": 5.299373626708984, "learning_rate": 1.355505585376469e-05, "loss": 0.0586, "step": 8985 }, { "epoch": 0.9741977450130095, "grad_norm": 7.736644744873047, "learning_rate": 1.3554330480197302e-05, "loss": 0.0966, "step": 8986 }, { "epoch": 0.9743061578490894, "grad_norm": 2.272494316101074, "learning_rate": 1.3553605106629915e-05, "loss": 0.025, "step": 8987 }, { "epoch": 0.9744145706851691, "grad_norm": 3.4968268871307373, "learning_rate": 1.3552879733062527e-05, "loss": 0.0372, "step": 8988 }, { "epoch": 0.9745229835212489, "grad_norm": 3.624920129776001, "learning_rate": 1.3552154359495141e-05, "loss": 0.0506, "step": 8989 }, { "epoch": 0.9746313963573288, "grad_norm": 0.9079505205154419, "learning_rate": 1.3551428985927755e-05, "loss": 0.0129, "step": 8990 }, { "epoch": 0.9747398091934085, "grad_norm": 2.3569514751434326, "learning_rate": 1.3550703612360367e-05, "loss": 0.0301, "step": 8991 }, { "epoch": 0.9748482220294883, "grad_norm": 1.6555936336517334, "learning_rate": 1.354997823879298e-05, "loss": 0.0265, "step": 8992 }, { "epoch": 0.974956634865568, "grad_norm": 1.9554953575134277, "learning_rate": 1.3549252865225593e-05, "loss": 0.0254, "step": 8993 }, { "epoch": 0.9750650477016479, "grad_norm": 3.3813397884368896, "learning_rate": 1.3548527491658205e-05, "loss": 0.0687, "step": 8994 }, { "epoch": 0.9751734605377277, "grad_norm": 1.6538716554641724, "learning_rate": 1.3547802118090818e-05, "loss": 0.033, "step": 8995 }, { "epoch": 0.9752818733738075, "grad_norm": 6.581789493560791, "learning_rate": 1.3547076744523431e-05, "loss": 0.1099, "step": 8996 }, { "epoch": 0.9753902862098872, "grad_norm": 2.077277898788452, "learning_rate": 1.3546351370956044e-05, "loss": 0.0189, "step": 8997 }, { "epoch": 0.9754986990459671, "grad_norm": 1.5673305988311768, "learning_rate": 1.3545625997388656e-05, "loss": 0.0384, "step": 8998 }, { "epoch": 0.9756071118820469, "grad_norm": 2.855759382247925, "learning_rate": 1.354490062382127e-05, "loss": 0.0207, "step": 8999 }, { "epoch": 0.9757155247181266, "grad_norm": 2.5234947204589844, "learning_rate": 1.3544175250253882e-05, "loss": 0.0608, "step": 9000 }, { "epoch": 0.9758239375542064, "grad_norm": 1.6069328784942627, "learning_rate": 1.3543449876686494e-05, "loss": 0.0329, "step": 9001 }, { "epoch": 0.9759323503902863, "grad_norm": 1.7884843349456787, "learning_rate": 1.3542724503119106e-05, "loss": 0.032, "step": 9002 }, { "epoch": 0.976040763226366, "grad_norm": 3.9030888080596924, "learning_rate": 1.354199912955172e-05, "loss": 0.0283, "step": 9003 }, { "epoch": 0.9761491760624458, "grad_norm": 1.5872572660446167, "learning_rate": 1.3541273755984332e-05, "loss": 0.0436, "step": 9004 }, { "epoch": 0.9762575888985255, "grad_norm": 1.6080561876296997, "learning_rate": 1.3540548382416944e-05, "loss": 0.0176, "step": 9005 }, { "epoch": 0.9763660017346054, "grad_norm": 1.4843549728393555, "learning_rate": 1.353982300884956e-05, "loss": 0.0254, "step": 9006 }, { "epoch": 0.9764744145706852, "grad_norm": 3.329052686691284, "learning_rate": 1.3539097635282172e-05, "loss": 0.0361, "step": 9007 }, { "epoch": 0.9765828274067649, "grad_norm": 1.9769713878631592, "learning_rate": 1.3538372261714786e-05, "loss": 0.0332, "step": 9008 }, { "epoch": 0.9766912402428447, "grad_norm": 1.038824200630188, "learning_rate": 1.3537646888147398e-05, "loss": 0.0234, "step": 9009 }, { "epoch": 0.9767996530789246, "grad_norm": 2.0035383701324463, "learning_rate": 1.353692151458001e-05, "loss": 0.0319, "step": 9010 }, { "epoch": 0.9769080659150043, "grad_norm": 3.55777645111084, "learning_rate": 1.3536196141012624e-05, "loss": 0.0425, "step": 9011 }, { "epoch": 0.9770164787510841, "grad_norm": 0.4315906763076782, "learning_rate": 1.3535470767445236e-05, "loss": 0.013, "step": 9012 }, { "epoch": 0.9771248915871639, "grad_norm": 1.5095345973968506, "learning_rate": 1.3534745393877848e-05, "loss": 0.043, "step": 9013 }, { "epoch": 0.9772333044232437, "grad_norm": 1.5223695039749146, "learning_rate": 1.3534020020310462e-05, "loss": 0.0225, "step": 9014 }, { "epoch": 0.9773417172593235, "grad_norm": 4.215207099914551, "learning_rate": 1.3533294646743074e-05, "loss": 0.067, "step": 9015 }, { "epoch": 0.9774501300954033, "grad_norm": 0.9534411430358887, "learning_rate": 1.3532569273175686e-05, "loss": 0.0146, "step": 9016 }, { "epoch": 0.977558542931483, "grad_norm": 1.3415906429290771, "learning_rate": 1.3531843899608298e-05, "loss": 0.0479, "step": 9017 }, { "epoch": 0.9776669557675629, "grad_norm": 2.3439130783081055, "learning_rate": 1.3531118526040912e-05, "loss": 0.037, "step": 9018 }, { "epoch": 0.9777753686036427, "grad_norm": 0.8306970000267029, "learning_rate": 1.3530393152473524e-05, "loss": 0.019, "step": 9019 }, { "epoch": 0.9778837814397224, "grad_norm": 3.161949872970581, "learning_rate": 1.3529667778906136e-05, "loss": 0.0636, "step": 9020 }, { "epoch": 0.9779921942758023, "grad_norm": 1.6712816953659058, "learning_rate": 1.352894240533875e-05, "loss": 0.0129, "step": 9021 }, { "epoch": 0.9781006071118821, "grad_norm": 0.4837920069694519, "learning_rate": 1.3528217031771362e-05, "loss": 0.012, "step": 9022 }, { "epoch": 0.9782090199479618, "grad_norm": 2.3839471340179443, "learning_rate": 1.3527491658203978e-05, "loss": 0.0102, "step": 9023 }, { "epoch": 0.9783174327840416, "grad_norm": 1.6353651285171509, "learning_rate": 1.352676628463659e-05, "loss": 0.0377, "step": 9024 }, { "epoch": 0.9784258456201215, "grad_norm": 1.9513072967529297, "learning_rate": 1.3526040911069202e-05, "loss": 0.0447, "step": 9025 }, { "epoch": 0.9785342584562012, "grad_norm": 2.734929084777832, "learning_rate": 1.3525315537501816e-05, "loss": 0.0381, "step": 9026 }, { "epoch": 0.978642671292281, "grad_norm": 1.2294795513153076, "learning_rate": 1.3524590163934428e-05, "loss": 0.0242, "step": 9027 }, { "epoch": 0.9787510841283608, "grad_norm": 2.0679116249084473, "learning_rate": 1.352386479036704e-05, "loss": 0.0234, "step": 9028 }, { "epoch": 0.9788594969644406, "grad_norm": 1.7563740015029907, "learning_rate": 1.3523139416799652e-05, "loss": 0.0286, "step": 9029 }, { "epoch": 0.9789679098005204, "grad_norm": 1.4525697231292725, "learning_rate": 1.3522414043232266e-05, "loss": 0.0492, "step": 9030 }, { "epoch": 0.9790763226366002, "grad_norm": 3.910109758377075, "learning_rate": 1.3521688669664878e-05, "loss": 0.0774, "step": 9031 }, { "epoch": 0.9791847354726799, "grad_norm": 1.3274552822113037, "learning_rate": 1.352096329609749e-05, "loss": 0.035, "step": 9032 }, { "epoch": 0.9792931483087598, "grad_norm": 0.9394749402999878, "learning_rate": 1.3520237922530104e-05, "loss": 0.0126, "step": 9033 }, { "epoch": 0.9794015611448396, "grad_norm": 2.6257736682891846, "learning_rate": 1.3519512548962717e-05, "loss": 0.0405, "step": 9034 }, { "epoch": 0.9795099739809193, "grad_norm": 2.8472585678100586, "learning_rate": 1.3518787175395329e-05, "loss": 0.0244, "step": 9035 }, { "epoch": 0.9796183868169991, "grad_norm": 3.039682149887085, "learning_rate": 1.3518061801827943e-05, "loss": 0.0318, "step": 9036 }, { "epoch": 0.979726799653079, "grad_norm": 4.007814407348633, "learning_rate": 1.3517336428260555e-05, "loss": 0.0317, "step": 9037 }, { "epoch": 0.9798352124891587, "grad_norm": 1.6310936212539673, "learning_rate": 1.3516611054693167e-05, "loss": 0.0064, "step": 9038 }, { "epoch": 0.9799436253252385, "grad_norm": 2.8715763092041016, "learning_rate": 1.351588568112578e-05, "loss": 0.0085, "step": 9039 }, { "epoch": 0.9800520381613183, "grad_norm": 4.189711570739746, "learning_rate": 1.3515160307558393e-05, "loss": 0.0445, "step": 9040 }, { "epoch": 0.9801604509973981, "grad_norm": 1.1780604124069214, "learning_rate": 1.3514434933991007e-05, "loss": 0.0401, "step": 9041 }, { "epoch": 0.9802688638334779, "grad_norm": 2.2955288887023926, "learning_rate": 1.351370956042362e-05, "loss": 0.0401, "step": 9042 }, { "epoch": 0.9803772766695577, "grad_norm": 0.8680179715156555, "learning_rate": 1.3512984186856233e-05, "loss": 0.0115, "step": 9043 }, { "epoch": 0.9804856895056374, "grad_norm": 2.7411439418792725, "learning_rate": 1.3512258813288845e-05, "loss": 0.0503, "step": 9044 }, { "epoch": 0.9805941023417173, "grad_norm": 4.654412746429443, "learning_rate": 1.3511533439721459e-05, "loss": 0.0512, "step": 9045 }, { "epoch": 0.980702515177797, "grad_norm": 8.36481761932373, "learning_rate": 1.351080806615407e-05, "loss": 0.0522, "step": 9046 }, { "epoch": 0.9808109280138768, "grad_norm": 1.8173484802246094, "learning_rate": 1.3510082692586683e-05, "loss": 0.044, "step": 9047 }, { "epoch": 0.9809193408499567, "grad_norm": 1.0763413906097412, "learning_rate": 1.3509357319019297e-05, "loss": 0.0165, "step": 9048 }, { "epoch": 0.9810277536860365, "grad_norm": 1.988569736480713, "learning_rate": 1.3508631945451909e-05, "loss": 0.0577, "step": 9049 }, { "epoch": 0.9811361665221162, "grad_norm": 3.370368480682373, "learning_rate": 1.3507906571884521e-05, "loss": 0.0363, "step": 9050 }, { "epoch": 0.981244579358196, "grad_norm": 6.230110168457031, "learning_rate": 1.3507181198317135e-05, "loss": 0.0447, "step": 9051 }, { "epoch": 0.9813529921942759, "grad_norm": 1.3870253562927246, "learning_rate": 1.3506455824749747e-05, "loss": 0.0495, "step": 9052 }, { "epoch": 0.9814614050303556, "grad_norm": 3.056591749191284, "learning_rate": 1.350573045118236e-05, "loss": 0.0382, "step": 9053 }, { "epoch": 0.9815698178664354, "grad_norm": 4.059561729431152, "learning_rate": 1.3505005077614971e-05, "loss": 0.0267, "step": 9054 }, { "epoch": 0.9816782307025151, "grad_norm": 0.864116370677948, "learning_rate": 1.3504279704047585e-05, "loss": 0.0149, "step": 9055 }, { "epoch": 0.981786643538595, "grad_norm": 2.082624912261963, "learning_rate": 1.3503554330480197e-05, "loss": 0.0185, "step": 9056 }, { "epoch": 0.9818950563746748, "grad_norm": 4.839231491088867, "learning_rate": 1.350282895691281e-05, "loss": 0.0228, "step": 9057 }, { "epoch": 0.9820034692107545, "grad_norm": 1.030593991279602, "learning_rate": 1.3502103583345425e-05, "loss": 0.0063, "step": 9058 }, { "epoch": 0.9821118820468343, "grad_norm": 1.1961805820465088, "learning_rate": 1.3501378209778037e-05, "loss": 0.0208, "step": 9059 }, { "epoch": 0.9822202948829142, "grad_norm": 7.502965927124023, "learning_rate": 1.3500652836210651e-05, "loss": 0.0357, "step": 9060 }, { "epoch": 0.9823287077189939, "grad_norm": 2.316706895828247, "learning_rate": 1.3499927462643263e-05, "loss": 0.0347, "step": 9061 }, { "epoch": 0.9824371205550737, "grad_norm": 2.342743158340454, "learning_rate": 1.3499202089075875e-05, "loss": 0.071, "step": 9062 }, { "epoch": 0.9825455333911535, "grad_norm": 2.132955312728882, "learning_rate": 1.3498476715508489e-05, "loss": 0.0261, "step": 9063 }, { "epoch": 0.9826539462272333, "grad_norm": 1.6095638275146484, "learning_rate": 1.3497751341941101e-05, "loss": 0.0308, "step": 9064 }, { "epoch": 0.9827623590633131, "grad_norm": 2.916311740875244, "learning_rate": 1.3497025968373713e-05, "loss": 0.0354, "step": 9065 }, { "epoch": 0.9828707718993929, "grad_norm": 2.902928590774536, "learning_rate": 1.3496300594806327e-05, "loss": 0.0549, "step": 9066 }, { "epoch": 0.9829791847354726, "grad_norm": 1.1533737182617188, "learning_rate": 1.349557522123894e-05, "loss": 0.0256, "step": 9067 }, { "epoch": 0.9830875975715525, "grad_norm": 1.9885634183883667, "learning_rate": 1.3494849847671551e-05, "loss": 0.039, "step": 9068 }, { "epoch": 0.9831960104076323, "grad_norm": 3.5873608589172363, "learning_rate": 1.3494124474104164e-05, "loss": 0.0629, "step": 9069 }, { "epoch": 0.983304423243712, "grad_norm": 1.929785966873169, "learning_rate": 1.3493399100536777e-05, "loss": 0.0269, "step": 9070 }, { "epoch": 0.9834128360797918, "grad_norm": 1.1243793964385986, "learning_rate": 1.349267372696939e-05, "loss": 0.0198, "step": 9071 }, { "epoch": 0.9835212489158717, "grad_norm": 1.4847935438156128, "learning_rate": 1.3491948353402002e-05, "loss": 0.0093, "step": 9072 }, { "epoch": 0.9836296617519514, "grad_norm": 1.111545443534851, "learning_rate": 1.3491222979834616e-05, "loss": 0.012, "step": 9073 }, { "epoch": 0.9837380745880312, "grad_norm": 2.4780962467193604, "learning_rate": 1.3490497606267228e-05, "loss": 0.027, "step": 9074 }, { "epoch": 0.983846487424111, "grad_norm": 1.6022710800170898, "learning_rate": 1.3489772232699843e-05, "loss": 0.0135, "step": 9075 }, { "epoch": 0.9839549002601908, "grad_norm": 2.9497337341308594, "learning_rate": 1.3489046859132455e-05, "loss": 0.0377, "step": 9076 }, { "epoch": 0.9840633130962706, "grad_norm": 7.225473880767822, "learning_rate": 1.3488321485565068e-05, "loss": 0.0709, "step": 9077 }, { "epoch": 0.9841717259323504, "grad_norm": 6.701991558074951, "learning_rate": 1.3487596111997681e-05, "loss": 0.0259, "step": 9078 }, { "epoch": 0.9842801387684302, "grad_norm": 0.9721494913101196, "learning_rate": 1.3486870738430294e-05, "loss": 0.0117, "step": 9079 }, { "epoch": 0.98438855160451, "grad_norm": 1.840410590171814, "learning_rate": 1.3486145364862906e-05, "loss": 0.0297, "step": 9080 }, { "epoch": 0.9844969644405898, "grad_norm": 2.4029812812805176, "learning_rate": 1.3485419991295518e-05, "loss": 0.0381, "step": 9081 }, { "epoch": 0.9846053772766695, "grad_norm": 2.8369438648223877, "learning_rate": 1.3484694617728132e-05, "loss": 0.0317, "step": 9082 }, { "epoch": 0.9847137901127494, "grad_norm": 0.9013267159461975, "learning_rate": 1.3483969244160744e-05, "loss": 0.0199, "step": 9083 }, { "epoch": 0.9848222029488292, "grad_norm": 5.807422161102295, "learning_rate": 1.3483243870593356e-05, "loss": 0.0376, "step": 9084 }, { "epoch": 0.9849306157849089, "grad_norm": 3.753819704055786, "learning_rate": 1.348251849702597e-05, "loss": 0.0391, "step": 9085 }, { "epoch": 0.9850390286209887, "grad_norm": 1.695373296737671, "learning_rate": 1.3481793123458582e-05, "loss": 0.0166, "step": 9086 }, { "epoch": 0.9851474414570686, "grad_norm": 1.1345385313034058, "learning_rate": 1.3481067749891194e-05, "loss": 0.0095, "step": 9087 }, { "epoch": 0.9852558542931483, "grad_norm": 4.891382217407227, "learning_rate": 1.3480342376323808e-05, "loss": 0.0344, "step": 9088 }, { "epoch": 0.9853642671292281, "grad_norm": 0.7256417870521545, "learning_rate": 1.347961700275642e-05, "loss": 0.0089, "step": 9089 }, { "epoch": 0.9854726799653079, "grad_norm": 5.496345043182373, "learning_rate": 1.3478891629189032e-05, "loss": 0.0137, "step": 9090 }, { "epoch": 0.9855810928013877, "grad_norm": 16.732542037963867, "learning_rate": 1.3478166255621646e-05, "loss": 0.0652, "step": 9091 }, { "epoch": 0.9856895056374675, "grad_norm": 3.4295530319213867, "learning_rate": 1.347744088205426e-05, "loss": 0.0425, "step": 9092 }, { "epoch": 0.9857979184735473, "grad_norm": 2.9981367588043213, "learning_rate": 1.3476715508486872e-05, "loss": 0.0407, "step": 9093 }, { "epoch": 0.985906331309627, "grad_norm": 1.7187827825546265, "learning_rate": 1.3475990134919486e-05, "loss": 0.0644, "step": 9094 }, { "epoch": 0.9860147441457069, "grad_norm": 1.023971438407898, "learning_rate": 1.3475264761352098e-05, "loss": 0.0123, "step": 9095 }, { "epoch": 0.9861231569817867, "grad_norm": 4.458873748779297, "learning_rate": 1.347453938778471e-05, "loss": 0.0254, "step": 9096 }, { "epoch": 0.9862315698178664, "grad_norm": 1.47965407371521, "learning_rate": 1.3473814014217324e-05, "loss": 0.041, "step": 9097 }, { "epoch": 0.9863399826539462, "grad_norm": 5.753917694091797, "learning_rate": 1.3473088640649936e-05, "loss": 0.042, "step": 9098 }, { "epoch": 0.986448395490026, "grad_norm": 7.599332809448242, "learning_rate": 1.3472363267082548e-05, "loss": 0.0746, "step": 9099 }, { "epoch": 0.9865568083261058, "grad_norm": 2.200463056564331, "learning_rate": 1.3471637893515162e-05, "loss": 0.0137, "step": 9100 }, { "epoch": 0.9866652211621856, "grad_norm": 4.619117259979248, "learning_rate": 1.3470912519947774e-05, "loss": 0.0653, "step": 9101 }, { "epoch": 0.9867736339982653, "grad_norm": 0.5758630037307739, "learning_rate": 1.3470187146380386e-05, "loss": 0.0156, "step": 9102 }, { "epoch": 0.9868820468343452, "grad_norm": 1.675722360610962, "learning_rate": 1.3469461772813e-05, "loss": 0.0525, "step": 9103 }, { "epoch": 0.986990459670425, "grad_norm": 2.2366435527801514, "learning_rate": 1.3468736399245612e-05, "loss": 0.0231, "step": 9104 }, { "epoch": 0.9870988725065047, "grad_norm": 1.2535223960876465, "learning_rate": 1.3468011025678224e-05, "loss": 0.0136, "step": 9105 }, { "epoch": 0.9872072853425846, "grad_norm": 2.5311992168426514, "learning_rate": 1.3467285652110838e-05, "loss": 0.0078, "step": 9106 }, { "epoch": 0.9873156981786644, "grad_norm": 3.0210349559783936, "learning_rate": 1.346656027854345e-05, "loss": 0.0277, "step": 9107 }, { "epoch": 0.9874241110147441, "grad_norm": 1.3974342346191406, "learning_rate": 1.3465834904976063e-05, "loss": 0.0148, "step": 9108 }, { "epoch": 0.9875325238508239, "grad_norm": 4.216785907745361, "learning_rate": 1.3465109531408675e-05, "loss": 0.0378, "step": 9109 }, { "epoch": 0.9876409366869038, "grad_norm": 4.941182613372803, "learning_rate": 1.346438415784129e-05, "loss": 0.0442, "step": 9110 }, { "epoch": 0.9877493495229835, "grad_norm": 3.434451103210449, "learning_rate": 1.3463658784273902e-05, "loss": 0.0339, "step": 9111 }, { "epoch": 0.9878577623590633, "grad_norm": 4.232774257659912, "learning_rate": 1.3462933410706516e-05, "loss": 0.0271, "step": 9112 }, { "epoch": 0.9879661751951431, "grad_norm": 3.1057322025299072, "learning_rate": 1.3462208037139128e-05, "loss": 0.0677, "step": 9113 }, { "epoch": 0.988074588031223, "grad_norm": 2.0881388187408447, "learning_rate": 1.346148266357174e-05, "loss": 0.0208, "step": 9114 }, { "epoch": 0.9881830008673027, "grad_norm": 1.3067185878753662, "learning_rate": 1.3460757290004354e-05, "loss": 0.0192, "step": 9115 }, { "epoch": 0.9882914137033825, "grad_norm": 0.6778854727745056, "learning_rate": 1.3460031916436967e-05, "loss": 0.005, "step": 9116 }, { "epoch": 0.9883998265394622, "grad_norm": 2.8964998722076416, "learning_rate": 1.3459306542869579e-05, "loss": 0.0543, "step": 9117 }, { "epoch": 0.9885082393755421, "grad_norm": 2.17151141166687, "learning_rate": 1.3458581169302193e-05, "loss": 0.0388, "step": 9118 }, { "epoch": 0.9886166522116219, "grad_norm": 1.9875364303588867, "learning_rate": 1.3457855795734805e-05, "loss": 0.0273, "step": 9119 }, { "epoch": 0.9887250650477016, "grad_norm": 1.3160264492034912, "learning_rate": 1.3457130422167417e-05, "loss": 0.0224, "step": 9120 }, { "epoch": 0.9888334778837814, "grad_norm": 1.3131295442581177, "learning_rate": 1.3456405048600029e-05, "loss": 0.032, "step": 9121 }, { "epoch": 0.9889418907198613, "grad_norm": 2.0313401222229004, "learning_rate": 1.3455679675032643e-05, "loss": 0.0048, "step": 9122 }, { "epoch": 0.989050303555941, "grad_norm": 1.4484885931015015, "learning_rate": 1.3454954301465255e-05, "loss": 0.0358, "step": 9123 }, { "epoch": 0.9891587163920208, "grad_norm": 3.7902133464813232, "learning_rate": 1.3454228927897867e-05, "loss": 0.0965, "step": 9124 }, { "epoch": 0.9892671292281006, "grad_norm": 4.637089729309082, "learning_rate": 1.3453503554330481e-05, "loss": 0.0526, "step": 9125 }, { "epoch": 0.9893755420641804, "grad_norm": 3.7071399688720703, "learning_rate": 1.3452778180763093e-05, "loss": 0.0504, "step": 9126 }, { "epoch": 0.9894839549002602, "grad_norm": 0.835309624671936, "learning_rate": 1.3452052807195709e-05, "loss": 0.0217, "step": 9127 }, { "epoch": 0.98959236773634, "grad_norm": 0.8945609927177429, "learning_rate": 1.345132743362832e-05, "loss": 0.0293, "step": 9128 }, { "epoch": 0.9897007805724197, "grad_norm": 2.7681615352630615, "learning_rate": 1.3450602060060933e-05, "loss": 0.0294, "step": 9129 }, { "epoch": 0.9898091934084996, "grad_norm": 3.596224308013916, "learning_rate": 1.3449876686493547e-05, "loss": 0.0384, "step": 9130 }, { "epoch": 0.9899176062445794, "grad_norm": 2.2011241912841797, "learning_rate": 1.3449151312926159e-05, "loss": 0.0416, "step": 9131 }, { "epoch": 0.9900260190806591, "grad_norm": 1.3045223951339722, "learning_rate": 1.3448425939358771e-05, "loss": 0.0085, "step": 9132 }, { "epoch": 0.9901344319167389, "grad_norm": 1.1098355054855347, "learning_rate": 1.3447700565791383e-05, "loss": 0.0205, "step": 9133 }, { "epoch": 0.9902428447528188, "grad_norm": 3.4689536094665527, "learning_rate": 1.3446975192223997e-05, "loss": 0.0575, "step": 9134 }, { "epoch": 0.9903512575888985, "grad_norm": 3.1142258644104004, "learning_rate": 1.3446249818656609e-05, "loss": 0.0411, "step": 9135 }, { "epoch": 0.9904596704249783, "grad_norm": 5.413689136505127, "learning_rate": 1.3445524445089221e-05, "loss": 0.0705, "step": 9136 }, { "epoch": 0.9905680832610582, "grad_norm": 1.683387041091919, "learning_rate": 1.3444799071521835e-05, "loss": 0.0558, "step": 9137 }, { "epoch": 0.9906764960971379, "grad_norm": 1.5667705535888672, "learning_rate": 1.3444073697954447e-05, "loss": 0.0526, "step": 9138 }, { "epoch": 0.9907849089332177, "grad_norm": 2.1184797286987305, "learning_rate": 1.344334832438706e-05, "loss": 0.0372, "step": 9139 }, { "epoch": 0.9908933217692975, "grad_norm": 0.5783866047859192, "learning_rate": 1.3442622950819673e-05, "loss": 0.0184, "step": 9140 }, { "epoch": 0.9910017346053773, "grad_norm": 0.5692380666732788, "learning_rate": 1.3441897577252285e-05, "loss": 0.0122, "step": 9141 }, { "epoch": 0.9911101474414571, "grad_norm": 1.0031529664993286, "learning_rate": 1.3441172203684898e-05, "loss": 0.0174, "step": 9142 }, { "epoch": 0.9912185602775369, "grad_norm": 1.183524250984192, "learning_rate": 1.3440446830117511e-05, "loss": 0.0422, "step": 9143 }, { "epoch": 0.9913269731136166, "grad_norm": 0.47120875120162964, "learning_rate": 1.3439721456550125e-05, "loss": 0.0102, "step": 9144 }, { "epoch": 0.9914353859496965, "grad_norm": 4.173432350158691, "learning_rate": 1.3438996082982737e-05, "loss": 0.0306, "step": 9145 }, { "epoch": 0.9915437987857763, "grad_norm": 0.8588484525680542, "learning_rate": 1.3438270709415351e-05, "loss": 0.0241, "step": 9146 }, { "epoch": 0.991652211621856, "grad_norm": 1.690294861793518, "learning_rate": 1.3437545335847963e-05, "loss": 0.0385, "step": 9147 }, { "epoch": 0.9917606244579358, "grad_norm": 3.71650767326355, "learning_rate": 1.3436819962280575e-05, "loss": 0.0467, "step": 9148 }, { "epoch": 0.9918690372940157, "grad_norm": 3.417840003967285, "learning_rate": 1.343609458871319e-05, "loss": 0.0461, "step": 9149 }, { "epoch": 0.9919774501300954, "grad_norm": 1.4056041240692139, "learning_rate": 1.3435369215145801e-05, "loss": 0.0404, "step": 9150 }, { "epoch": 0.9920858629661752, "grad_norm": 0.8508855700492859, "learning_rate": 1.3434643841578414e-05, "loss": 0.0185, "step": 9151 }, { "epoch": 0.9921942758022549, "grad_norm": 3.0527796745300293, "learning_rate": 1.3433918468011027e-05, "loss": 0.0132, "step": 9152 }, { "epoch": 0.9923026886383348, "grad_norm": 0.4600517153739929, "learning_rate": 1.343319309444364e-05, "loss": 0.0125, "step": 9153 }, { "epoch": 0.9924111014744146, "grad_norm": 1.808220386505127, "learning_rate": 1.3432467720876252e-05, "loss": 0.0272, "step": 9154 }, { "epoch": 0.9925195143104943, "grad_norm": 3.4880521297454834, "learning_rate": 1.3431742347308866e-05, "loss": 0.0846, "step": 9155 }, { "epoch": 0.9926279271465741, "grad_norm": 1.7297526597976685, "learning_rate": 1.3431016973741478e-05, "loss": 0.0365, "step": 9156 }, { "epoch": 0.992736339982654, "grad_norm": 2.8245291709899902, "learning_rate": 1.343029160017409e-05, "loss": 0.0338, "step": 9157 }, { "epoch": 0.9928447528187337, "grad_norm": 0.8214278221130371, "learning_rate": 1.3429566226606704e-05, "loss": 0.011, "step": 9158 }, { "epoch": 0.9929531656548135, "grad_norm": 2.5034260749816895, "learning_rate": 1.3428840853039316e-05, "loss": 0.0117, "step": 9159 }, { "epoch": 0.9930615784908933, "grad_norm": 4.644093990325928, "learning_rate": 1.3428115479471928e-05, "loss": 0.0334, "step": 9160 }, { "epoch": 0.9931699913269731, "grad_norm": 0.9197945594787598, "learning_rate": 1.342739010590454e-05, "loss": 0.0173, "step": 9161 }, { "epoch": 0.9932784041630529, "grad_norm": 1.1287389993667603, "learning_rate": 1.3426664732337156e-05, "loss": 0.0274, "step": 9162 }, { "epoch": 0.9933868169991327, "grad_norm": 4.690813064575195, "learning_rate": 1.3425939358769768e-05, "loss": 0.0536, "step": 9163 }, { "epoch": 0.9934952298352124, "grad_norm": 2.869736909866333, "learning_rate": 1.3425213985202382e-05, "loss": 0.0344, "step": 9164 }, { "epoch": 0.9936036426712923, "grad_norm": 1.7362375259399414, "learning_rate": 1.3424488611634994e-05, "loss": 0.032, "step": 9165 }, { "epoch": 0.9937120555073721, "grad_norm": 5.084672927856445, "learning_rate": 1.3423763238067606e-05, "loss": 0.0518, "step": 9166 }, { "epoch": 0.9938204683434518, "grad_norm": 0.7563862204551697, "learning_rate": 1.342303786450022e-05, "loss": 0.0303, "step": 9167 }, { "epoch": 0.9939288811795317, "grad_norm": 0.8490155935287476, "learning_rate": 1.3422312490932832e-05, "loss": 0.0086, "step": 9168 }, { "epoch": 0.9940372940156115, "grad_norm": 0.9300787448883057, "learning_rate": 1.3421587117365444e-05, "loss": 0.0152, "step": 9169 }, { "epoch": 0.9941457068516912, "grad_norm": 2.719688892364502, "learning_rate": 1.3420861743798058e-05, "loss": 0.0475, "step": 9170 }, { "epoch": 0.994254119687771, "grad_norm": 2.9905190467834473, "learning_rate": 1.342013637023067e-05, "loss": 0.0697, "step": 9171 }, { "epoch": 0.9943625325238509, "grad_norm": 4.623388290405273, "learning_rate": 1.3419410996663282e-05, "loss": 0.0437, "step": 9172 }, { "epoch": 0.9944709453599306, "grad_norm": 11.160117149353027, "learning_rate": 1.3418685623095894e-05, "loss": 0.0281, "step": 9173 }, { "epoch": 0.9945793581960104, "grad_norm": 3.0767204761505127, "learning_rate": 1.3417960249528508e-05, "loss": 0.0287, "step": 9174 }, { "epoch": 0.9946877710320902, "grad_norm": 7.113481044769287, "learning_rate": 1.341723487596112e-05, "loss": 0.0427, "step": 9175 }, { "epoch": 0.99479618386817, "grad_norm": 6.142299652099609, "learning_rate": 1.3416509502393732e-05, "loss": 0.1, "step": 9176 }, { "epoch": 0.9949045967042498, "grad_norm": 1.8431521654129028, "learning_rate": 1.3415784128826346e-05, "loss": 0.0123, "step": 9177 }, { "epoch": 0.9950130095403296, "grad_norm": 3.413325309753418, "learning_rate": 1.3415058755258958e-05, "loss": 0.0755, "step": 9178 }, { "epoch": 0.9951214223764093, "grad_norm": 1.0625168085098267, "learning_rate": 1.3414333381691574e-05, "loss": 0.0296, "step": 9179 }, { "epoch": 0.9952298352124892, "grad_norm": 0.5343732237815857, "learning_rate": 1.3413608008124186e-05, "loss": 0.017, "step": 9180 }, { "epoch": 0.995338248048569, "grad_norm": 4.553608417510986, "learning_rate": 1.3412882634556798e-05, "loss": 0.0494, "step": 9181 }, { "epoch": 0.9954466608846487, "grad_norm": 2.0292842388153076, "learning_rate": 1.3412157260989412e-05, "loss": 0.0413, "step": 9182 }, { "epoch": 0.9955550737207285, "grad_norm": 1.6032705307006836, "learning_rate": 1.3411431887422024e-05, "loss": 0.0372, "step": 9183 }, { "epoch": 0.9956634865568084, "grad_norm": 0.5154082775115967, "learning_rate": 1.3410706513854636e-05, "loss": 0.0157, "step": 9184 }, { "epoch": 0.9957718993928881, "grad_norm": 1.7086303234100342, "learning_rate": 1.3409981140287248e-05, "loss": 0.0248, "step": 9185 }, { "epoch": 0.9958803122289679, "grad_norm": 3.062920570373535, "learning_rate": 1.3409255766719862e-05, "loss": 0.0809, "step": 9186 }, { "epoch": 0.9959887250650477, "grad_norm": 0.8390081524848938, "learning_rate": 1.3408530393152474e-05, "loss": 0.025, "step": 9187 }, { "epoch": 0.9960971379011275, "grad_norm": 4.840274333953857, "learning_rate": 1.3407805019585087e-05, "loss": 0.0385, "step": 9188 }, { "epoch": 0.9962055507372073, "grad_norm": 3.6274259090423584, "learning_rate": 1.34070796460177e-05, "loss": 0.0365, "step": 9189 }, { "epoch": 0.996313963573287, "grad_norm": 5.384641170501709, "learning_rate": 1.3406354272450313e-05, "loss": 0.0705, "step": 9190 }, { "epoch": 0.9964223764093668, "grad_norm": 1.6884198188781738, "learning_rate": 1.3405628898882925e-05, "loss": 0.054, "step": 9191 }, { "epoch": 0.9965307892454467, "grad_norm": 0.3593965470790863, "learning_rate": 1.3404903525315539e-05, "loss": 0.0049, "step": 9192 }, { "epoch": 0.9966392020815265, "grad_norm": 2.6792802810668945, "learning_rate": 1.340417815174815e-05, "loss": 0.0488, "step": 9193 }, { "epoch": 0.9967476149176062, "grad_norm": 4.204979419708252, "learning_rate": 1.3403452778180763e-05, "loss": 0.0468, "step": 9194 }, { "epoch": 0.9968560277536861, "grad_norm": 1.9924951791763306, "learning_rate": 1.3402727404613377e-05, "loss": 0.0224, "step": 9195 }, { "epoch": 0.9969644405897659, "grad_norm": 0.44417059421539307, "learning_rate": 1.340200203104599e-05, "loss": 0.0075, "step": 9196 }, { "epoch": 0.9970728534258456, "grad_norm": 4.681076526641846, "learning_rate": 1.3401276657478603e-05, "loss": 0.0454, "step": 9197 }, { "epoch": 0.9971812662619254, "grad_norm": 1.7073866128921509, "learning_rate": 1.3400551283911217e-05, "loss": 0.0241, "step": 9198 }, { "epoch": 0.9972896790980053, "grad_norm": 1.9633103609085083, "learning_rate": 1.3399825910343829e-05, "loss": 0.0147, "step": 9199 }, { "epoch": 0.997398091934085, "grad_norm": 1.696662425994873, "learning_rate": 1.339910053677644e-05, "loss": 0.042, "step": 9200 }, { "epoch": 0.9975065047701648, "grad_norm": 2.1947336196899414, "learning_rate": 1.3398375163209055e-05, "loss": 0.0419, "step": 9201 }, { "epoch": 0.9976149176062445, "grad_norm": 1.8563106060028076, "learning_rate": 1.3397649789641667e-05, "loss": 0.0099, "step": 9202 }, { "epoch": 0.9977233304423244, "grad_norm": 7.10136079788208, "learning_rate": 1.3396924416074279e-05, "loss": 0.0648, "step": 9203 }, { "epoch": 0.9978317432784042, "grad_norm": 1.4113268852233887, "learning_rate": 1.3396199042506893e-05, "loss": 0.0341, "step": 9204 }, { "epoch": 0.997940156114484, "grad_norm": 2.9828433990478516, "learning_rate": 1.3395473668939505e-05, "loss": 0.0157, "step": 9205 }, { "epoch": 0.9980485689505637, "grad_norm": 5.939035892486572, "learning_rate": 1.3394748295372117e-05, "loss": 0.0712, "step": 9206 }, { "epoch": 0.9981569817866436, "grad_norm": 2.486107110977173, "learning_rate": 1.3394022921804731e-05, "loss": 0.0399, "step": 9207 }, { "epoch": 0.9982653946227233, "grad_norm": 1.433416485786438, "learning_rate": 1.3393297548237343e-05, "loss": 0.0428, "step": 9208 }, { "epoch": 0.9983738074588031, "grad_norm": 1.2704049348831177, "learning_rate": 1.3392572174669955e-05, "loss": 0.0111, "step": 9209 }, { "epoch": 0.9984822202948829, "grad_norm": 1.102683186531067, "learning_rate": 1.3391846801102569e-05, "loss": 0.0227, "step": 9210 }, { "epoch": 0.9985906331309627, "grad_norm": 2.261756181716919, "learning_rate": 1.3391121427535181e-05, "loss": 0.0422, "step": 9211 }, { "epoch": 0.9986990459670425, "grad_norm": 6.022621154785156, "learning_rate": 1.3390396053967793e-05, "loss": 0.0467, "step": 9212 }, { "epoch": 0.9988074588031223, "grad_norm": 3.7990760803222656, "learning_rate": 1.3389670680400409e-05, "loss": 0.0367, "step": 9213 }, { "epoch": 0.998915871639202, "grad_norm": 2.646127462387085, "learning_rate": 1.3388945306833021e-05, "loss": 0.0302, "step": 9214 }, { "epoch": 0.9990242844752819, "grad_norm": 5.295278549194336, "learning_rate": 1.3388219933265633e-05, "loss": 0.0501, "step": 9215 }, { "epoch": 0.9991326973113617, "grad_norm": 2.768740177154541, "learning_rate": 1.3387494559698247e-05, "loss": 0.0308, "step": 9216 }, { "epoch": 0.9992411101474414, "grad_norm": 7.205316543579102, "learning_rate": 1.3386769186130859e-05, "loss": 0.0697, "step": 9217 }, { "epoch": 0.9993495229835212, "grad_norm": 3.4344544410705566, "learning_rate": 1.3386043812563471e-05, "loss": 0.0154, "step": 9218 }, { "epoch": 0.9994579358196011, "grad_norm": 3.02193021774292, "learning_rate": 1.3385318438996085e-05, "loss": 0.0382, "step": 9219 }, { "epoch": 0.9995663486556808, "grad_norm": 1.4274216890335083, "learning_rate": 1.3384593065428697e-05, "loss": 0.0219, "step": 9220 }, { "epoch": 0.9996747614917606, "grad_norm": 1.1479383707046509, "learning_rate": 1.338386769186131e-05, "loss": 0.0222, "step": 9221 }, { "epoch": 0.9997831743278404, "grad_norm": 1.6105575561523438, "learning_rate": 1.3383142318293923e-05, "loss": 0.0264, "step": 9222 }, { "epoch": 0.9998915871639202, "grad_norm": 3.9579708576202393, "learning_rate": 1.3382416944726535e-05, "loss": 0.0698, "step": 9223 }, { "epoch": 1.0, "grad_norm": 2.5868847370147705, "learning_rate": 1.3381691571159147e-05, "loss": 0.0272, "step": 9224 }, { "epoch": 1.0001084128360798, "grad_norm": 0.6271238923072815, "learning_rate": 1.338096619759176e-05, "loss": 0.0116, "step": 9225 }, { "epoch": 1.0002168256721595, "grad_norm": 0.7333528995513916, "learning_rate": 1.3380240824024373e-05, "loss": 0.0039, "step": 9226 }, { "epoch": 1.0003252385082393, "grad_norm": 0.5451459288597107, "learning_rate": 1.3379515450456986e-05, "loss": 0.0146, "step": 9227 }, { "epoch": 1.0004336513443193, "grad_norm": 1.2713881731033325, "learning_rate": 1.3378790076889598e-05, "loss": 0.0182, "step": 9228 }, { "epoch": 1.000542064180399, "grad_norm": 1.202480673789978, "learning_rate": 1.3378064703322212e-05, "loss": 0.0218, "step": 9229 }, { "epoch": 1.0006504770164788, "grad_norm": 3.6712515354156494, "learning_rate": 1.3377339329754824e-05, "loss": 0.0511, "step": 9230 }, { "epoch": 1.0007588898525586, "grad_norm": 0.5653786063194275, "learning_rate": 1.337661395618744e-05, "loss": 0.012, "step": 9231 }, { "epoch": 1.0008673026886383, "grad_norm": 4.00275993347168, "learning_rate": 1.3375888582620051e-05, "loss": 0.0668, "step": 9232 }, { "epoch": 1.000975715524718, "grad_norm": 1.4154622554779053, "learning_rate": 1.3375163209052664e-05, "loss": 0.0513, "step": 9233 }, { "epoch": 1.0010841283607979, "grad_norm": 0.4258038401603699, "learning_rate": 1.3374437835485277e-05, "loss": 0.0102, "step": 9234 }, { "epoch": 1.0011925411968776, "grad_norm": 0.37554487586021423, "learning_rate": 1.337371246191789e-05, "loss": 0.0069, "step": 9235 }, { "epoch": 1.0013009540329576, "grad_norm": 3.1124155521392822, "learning_rate": 1.3372987088350502e-05, "loss": 0.0323, "step": 9236 }, { "epoch": 1.0014093668690374, "grad_norm": 0.1123809963464737, "learning_rate": 1.3372261714783114e-05, "loss": 0.0015, "step": 9237 }, { "epoch": 1.0015177797051171, "grad_norm": 0.7458950281143188, "learning_rate": 1.3371536341215728e-05, "loss": 0.0063, "step": 9238 }, { "epoch": 1.001626192541197, "grad_norm": 1.5473226308822632, "learning_rate": 1.337081096764834e-05, "loss": 0.0198, "step": 9239 }, { "epoch": 1.0017346053772767, "grad_norm": 4.284530162811279, "learning_rate": 1.3370085594080952e-05, "loss": 0.0797, "step": 9240 }, { "epoch": 1.0018430182133564, "grad_norm": 1.271921992301941, "learning_rate": 1.3369360220513566e-05, "loss": 0.022, "step": 9241 }, { "epoch": 1.0019514310494362, "grad_norm": 0.6552806496620178, "learning_rate": 1.3368634846946178e-05, "loss": 0.0252, "step": 9242 }, { "epoch": 1.002059843885516, "grad_norm": 0.5106589794158936, "learning_rate": 1.336790947337879e-05, "loss": 0.0052, "step": 9243 }, { "epoch": 1.002168256721596, "grad_norm": 4.683409690856934, "learning_rate": 1.3367184099811404e-05, "loss": 0.0662, "step": 9244 }, { "epoch": 1.0022766695576757, "grad_norm": 2.889756917953491, "learning_rate": 1.3366458726244016e-05, "loss": 0.0542, "step": 9245 }, { "epoch": 1.0023850823937555, "grad_norm": 3.571415662765503, "learning_rate": 1.3365733352676628e-05, "loss": 0.0127, "step": 9246 }, { "epoch": 1.0024934952298352, "grad_norm": 0.2689647078514099, "learning_rate": 1.3365007979109242e-05, "loss": 0.0031, "step": 9247 }, { "epoch": 1.002601908065915, "grad_norm": 7.720685005187988, "learning_rate": 1.3364282605541856e-05, "loss": 0.0291, "step": 9248 }, { "epoch": 1.0027103209019947, "grad_norm": 1.4107921123504639, "learning_rate": 1.336355723197447e-05, "loss": 0.0231, "step": 9249 }, { "epoch": 1.0028187337380745, "grad_norm": 1.845859169960022, "learning_rate": 1.3362831858407082e-05, "loss": 0.0115, "step": 9250 }, { "epoch": 1.0029271465741543, "grad_norm": 4.774386882781982, "learning_rate": 1.3362106484839694e-05, "loss": 0.0118, "step": 9251 }, { "epoch": 1.0030355594102343, "grad_norm": 1.820234775543213, "learning_rate": 1.3361381111272306e-05, "loss": 0.0478, "step": 9252 }, { "epoch": 1.003143972246314, "grad_norm": 1.7794691324234009, "learning_rate": 1.336065573770492e-05, "loss": 0.0157, "step": 9253 }, { "epoch": 1.0032523850823938, "grad_norm": 0.30080363154411316, "learning_rate": 1.3359930364137532e-05, "loss": 0.0033, "step": 9254 }, { "epoch": 1.0033607979184735, "grad_norm": 4.627859592437744, "learning_rate": 1.3359204990570144e-05, "loss": 0.0149, "step": 9255 }, { "epoch": 1.0034692107545533, "grad_norm": 0.6629464626312256, "learning_rate": 1.3358479617002758e-05, "loss": 0.0055, "step": 9256 }, { "epoch": 1.003577623590633, "grad_norm": 3.5511317253112793, "learning_rate": 1.335775424343537e-05, "loss": 0.0215, "step": 9257 }, { "epoch": 1.0036860364267128, "grad_norm": 1.5987787246704102, "learning_rate": 1.3357028869867982e-05, "loss": 0.0606, "step": 9258 }, { "epoch": 1.0037944492627928, "grad_norm": 8.922318458557129, "learning_rate": 1.3356303496300596e-05, "loss": 0.0313, "step": 9259 }, { "epoch": 1.0039028620988726, "grad_norm": 2.2624104022979736, "learning_rate": 1.3355578122733208e-05, "loss": 0.0395, "step": 9260 }, { "epoch": 1.0040112749349523, "grad_norm": 1.7118175029754639, "learning_rate": 1.335485274916582e-05, "loss": 0.0129, "step": 9261 }, { "epoch": 1.004119687771032, "grad_norm": 1.7331072092056274, "learning_rate": 1.3354127375598434e-05, "loss": 0.0626, "step": 9262 }, { "epoch": 1.0042281006071119, "grad_norm": 0.5035593509674072, "learning_rate": 1.3353402002031046e-05, "loss": 0.0105, "step": 9263 }, { "epoch": 1.0043365134431916, "grad_norm": 0.3548838794231415, "learning_rate": 1.3352676628463659e-05, "loss": 0.0025, "step": 9264 }, { "epoch": 1.0044449262792714, "grad_norm": 2.20461368560791, "learning_rate": 1.3351951254896274e-05, "loss": 0.0184, "step": 9265 }, { "epoch": 1.0045533391153512, "grad_norm": 0.773459792137146, "learning_rate": 1.3351225881328886e-05, "loss": 0.0139, "step": 9266 }, { "epoch": 1.0046617519514311, "grad_norm": 3.098508358001709, "learning_rate": 1.3350500507761498e-05, "loss": 0.0306, "step": 9267 }, { "epoch": 1.004770164787511, "grad_norm": 1.506425380706787, "learning_rate": 1.3349775134194112e-05, "loss": 0.0244, "step": 9268 }, { "epoch": 1.0048785776235907, "grad_norm": 0.9106695652008057, "learning_rate": 1.3349049760626724e-05, "loss": 0.0175, "step": 9269 }, { "epoch": 1.0049869904596704, "grad_norm": 1.075599193572998, "learning_rate": 1.3348324387059337e-05, "loss": 0.0122, "step": 9270 }, { "epoch": 1.0050954032957502, "grad_norm": 2.277285099029541, "learning_rate": 1.334759901349195e-05, "loss": 0.0084, "step": 9271 }, { "epoch": 1.00520381613183, "grad_norm": 1.271340250968933, "learning_rate": 1.3346873639924563e-05, "loss": 0.0041, "step": 9272 }, { "epoch": 1.0053122289679097, "grad_norm": 0.6936512589454651, "learning_rate": 1.3346148266357175e-05, "loss": 0.004, "step": 9273 }, { "epoch": 1.0054206418039895, "grad_norm": 0.46445536613464355, "learning_rate": 1.3345422892789789e-05, "loss": 0.0134, "step": 9274 }, { "epoch": 1.0055290546400695, "grad_norm": 3.6951699256896973, "learning_rate": 1.33446975192224e-05, "loss": 0.0208, "step": 9275 }, { "epoch": 1.0056374674761492, "grad_norm": 1.1162103414535522, "learning_rate": 1.3343972145655013e-05, "loss": 0.0194, "step": 9276 }, { "epoch": 1.005745880312229, "grad_norm": 3.00752592086792, "learning_rate": 1.3343246772087625e-05, "loss": 0.0553, "step": 9277 }, { "epoch": 1.0058542931483088, "grad_norm": 2.062613010406494, "learning_rate": 1.3342521398520239e-05, "loss": 0.0148, "step": 9278 }, { "epoch": 1.0059627059843885, "grad_norm": 0.43541219830513, "learning_rate": 1.3341796024952851e-05, "loss": 0.0015, "step": 9279 }, { "epoch": 1.0060711188204683, "grad_norm": 0.46501100063323975, "learning_rate": 1.3341070651385463e-05, "loss": 0.0244, "step": 9280 }, { "epoch": 1.006179531656548, "grad_norm": 2.9758379459381104, "learning_rate": 1.3340345277818077e-05, "loss": 0.04, "step": 9281 }, { "epoch": 1.0062879444926278, "grad_norm": 0.344830721616745, "learning_rate": 1.3339619904250689e-05, "loss": 0.0012, "step": 9282 }, { "epoch": 1.0063963573287078, "grad_norm": 3.894989252090454, "learning_rate": 1.3338894530683305e-05, "loss": 0.0253, "step": 9283 }, { "epoch": 1.0065047701647876, "grad_norm": 6.8443427085876465, "learning_rate": 1.3338169157115917e-05, "loss": 0.0346, "step": 9284 }, { "epoch": 1.0066131830008673, "grad_norm": 2.4437549114227295, "learning_rate": 1.3337443783548529e-05, "loss": 0.0108, "step": 9285 }, { "epoch": 1.006721595836947, "grad_norm": 0.2887539863586426, "learning_rate": 1.3336718409981143e-05, "loss": 0.0039, "step": 9286 }, { "epoch": 1.0068300086730269, "grad_norm": 2.673990488052368, "learning_rate": 1.3335993036413755e-05, "loss": 0.0195, "step": 9287 }, { "epoch": 1.0069384215091066, "grad_norm": 0.50516277551651, "learning_rate": 1.3335267662846367e-05, "loss": 0.0109, "step": 9288 }, { "epoch": 1.0070468343451864, "grad_norm": 4.720067977905273, "learning_rate": 1.333454228927898e-05, "loss": 0.0512, "step": 9289 }, { "epoch": 1.0071552471812664, "grad_norm": 2.210390090942383, "learning_rate": 1.3333816915711593e-05, "loss": 0.0201, "step": 9290 }, { "epoch": 1.0072636600173461, "grad_norm": 0.4205301105976105, "learning_rate": 1.3333091542144205e-05, "loss": 0.0015, "step": 9291 }, { "epoch": 1.007372072853426, "grad_norm": 2.4765613079071045, "learning_rate": 1.3332366168576817e-05, "loss": 0.017, "step": 9292 }, { "epoch": 1.0074804856895057, "grad_norm": 1.7385146617889404, "learning_rate": 1.3331640795009431e-05, "loss": 0.0369, "step": 9293 }, { "epoch": 1.0075888985255854, "grad_norm": 2.3839001655578613, "learning_rate": 1.3330915421442043e-05, "loss": 0.0371, "step": 9294 }, { "epoch": 1.0076973113616652, "grad_norm": 1.6880987882614136, "learning_rate": 1.3330190047874655e-05, "loss": 0.0094, "step": 9295 }, { "epoch": 1.007805724197745, "grad_norm": 2.436349630355835, "learning_rate": 1.332946467430727e-05, "loss": 0.0396, "step": 9296 }, { "epoch": 1.0079141370338247, "grad_norm": 0.40183186531066895, "learning_rate": 1.3328739300739881e-05, "loss": 0.0066, "step": 9297 }, { "epoch": 1.0080225498699047, "grad_norm": 3.2066256999969482, "learning_rate": 1.3328013927172494e-05, "loss": 0.0186, "step": 9298 }, { "epoch": 1.0081309627059845, "grad_norm": 1.1779595613479614, "learning_rate": 1.3327288553605107e-05, "loss": 0.0064, "step": 9299 }, { "epoch": 1.0082393755420642, "grad_norm": 0.10294840484857559, "learning_rate": 1.3326563180037721e-05, "loss": 0.0009, "step": 9300 }, { "epoch": 1.008347788378144, "grad_norm": 1.456398844718933, "learning_rate": 1.3325837806470335e-05, "loss": 0.0305, "step": 9301 }, { "epoch": 1.0084562012142237, "grad_norm": 0.3638860583305359, "learning_rate": 1.3325112432902947e-05, "loss": 0.0068, "step": 9302 }, { "epoch": 1.0085646140503035, "grad_norm": 0.5874691605567932, "learning_rate": 1.332438705933556e-05, "loss": 0.0027, "step": 9303 }, { "epoch": 1.0086730268863833, "grad_norm": 3.341787815093994, "learning_rate": 1.3323661685768171e-05, "loss": 0.0257, "step": 9304 }, { "epoch": 1.008781439722463, "grad_norm": 1.0833193063735962, "learning_rate": 1.3322936312200785e-05, "loss": 0.0035, "step": 9305 }, { "epoch": 1.008889852558543, "grad_norm": 2.810171127319336, "learning_rate": 1.3322210938633397e-05, "loss": 0.0183, "step": 9306 }, { "epoch": 1.0089982653946228, "grad_norm": 0.5815998911857605, "learning_rate": 1.332148556506601e-05, "loss": 0.0142, "step": 9307 }, { "epoch": 1.0091066782307025, "grad_norm": 0.9517871737480164, "learning_rate": 1.3320760191498623e-05, "loss": 0.0143, "step": 9308 }, { "epoch": 1.0092150910667823, "grad_norm": 2.3781967163085938, "learning_rate": 1.3320034817931236e-05, "loss": 0.0286, "step": 9309 }, { "epoch": 1.009323503902862, "grad_norm": 2.45475172996521, "learning_rate": 1.3319309444363848e-05, "loss": 0.012, "step": 9310 }, { "epoch": 1.0094319167389418, "grad_norm": 1.4584271907806396, "learning_rate": 1.3318584070796462e-05, "loss": 0.0109, "step": 9311 }, { "epoch": 1.0095403295750216, "grad_norm": 2.612616539001465, "learning_rate": 1.3317858697229074e-05, "loss": 0.03, "step": 9312 }, { "epoch": 1.0096487424111014, "grad_norm": 9.938859939575195, "learning_rate": 1.3317133323661686e-05, "loss": 0.0494, "step": 9313 }, { "epoch": 1.0097571552471813, "grad_norm": 1.0546663999557495, "learning_rate": 1.33164079500943e-05, "loss": 0.0097, "step": 9314 }, { "epoch": 1.0098655680832611, "grad_norm": 5.0326337814331055, "learning_rate": 1.3315682576526912e-05, "loss": 0.123, "step": 9315 }, { "epoch": 1.0099739809193409, "grad_norm": 3.2564520835876465, "learning_rate": 1.3314957202959524e-05, "loss": 0.0198, "step": 9316 }, { "epoch": 1.0100823937554206, "grad_norm": 3.6688218116760254, "learning_rate": 1.331423182939214e-05, "loss": 0.0384, "step": 9317 }, { "epoch": 1.0101908065915004, "grad_norm": 3.165292978286743, "learning_rate": 1.3313506455824752e-05, "loss": 0.0248, "step": 9318 }, { "epoch": 1.0102992194275802, "grad_norm": 1.1611989736557007, "learning_rate": 1.3312781082257364e-05, "loss": 0.0244, "step": 9319 }, { "epoch": 1.01040763226366, "grad_norm": 0.9679598212242126, "learning_rate": 1.3312055708689978e-05, "loss": 0.0079, "step": 9320 }, { "epoch": 1.01051604509974, "grad_norm": 3.6602251529693604, "learning_rate": 1.331133033512259e-05, "loss": 0.0367, "step": 9321 }, { "epoch": 1.0106244579358197, "grad_norm": 3.0755727291107178, "learning_rate": 1.3310604961555202e-05, "loss": 0.0208, "step": 9322 }, { "epoch": 1.0107328707718994, "grad_norm": 0.61830735206604, "learning_rate": 1.3309879587987816e-05, "loss": 0.0061, "step": 9323 }, { "epoch": 1.0108412836079792, "grad_norm": 1.3473620414733887, "learning_rate": 1.3309154214420428e-05, "loss": 0.0252, "step": 9324 }, { "epoch": 1.010949696444059, "grad_norm": 4.705120086669922, "learning_rate": 1.330842884085304e-05, "loss": 0.0242, "step": 9325 }, { "epoch": 1.0110581092801387, "grad_norm": 0.6761757731437683, "learning_rate": 1.3307703467285654e-05, "loss": 0.0058, "step": 9326 }, { "epoch": 1.0111665221162185, "grad_norm": 2.1128690242767334, "learning_rate": 1.3306978093718266e-05, "loss": 0.0097, "step": 9327 }, { "epoch": 1.0112749349522983, "grad_norm": 2.394164800643921, "learning_rate": 1.3306252720150878e-05, "loss": 0.0359, "step": 9328 }, { "epoch": 1.0113833477883782, "grad_norm": 3.8097002506256104, "learning_rate": 1.330552734658349e-05, "loss": 0.0468, "step": 9329 }, { "epoch": 1.011491760624458, "grad_norm": 1.0256199836730957, "learning_rate": 1.3304801973016104e-05, "loss": 0.0145, "step": 9330 }, { "epoch": 1.0116001734605378, "grad_norm": 0.5499988198280334, "learning_rate": 1.3304076599448716e-05, "loss": 0.0035, "step": 9331 }, { "epoch": 1.0117085862966175, "grad_norm": 2.080979347229004, "learning_rate": 1.3303351225881328e-05, "loss": 0.0184, "step": 9332 }, { "epoch": 1.0118169991326973, "grad_norm": 3.0619637966156006, "learning_rate": 1.3302625852313942e-05, "loss": 0.0317, "step": 9333 }, { "epoch": 1.011925411968777, "grad_norm": 1.358562707901001, "learning_rate": 1.3301900478746556e-05, "loss": 0.0115, "step": 9334 }, { "epoch": 1.0120338248048568, "grad_norm": 1.0252662897109985, "learning_rate": 1.330117510517917e-05, "loss": 0.0355, "step": 9335 }, { "epoch": 1.0121422376409366, "grad_norm": 0.7356091737747192, "learning_rate": 1.3300449731611782e-05, "loss": 0.0133, "step": 9336 }, { "epoch": 1.0122506504770166, "grad_norm": 0.40801802277565, "learning_rate": 1.3299724358044394e-05, "loss": 0.0115, "step": 9337 }, { "epoch": 1.0123590633130963, "grad_norm": 0.2617858052253723, "learning_rate": 1.3298998984477008e-05, "loss": 0.0016, "step": 9338 }, { "epoch": 1.012467476149176, "grad_norm": 2.408686399459839, "learning_rate": 1.329827361090962e-05, "loss": 0.0556, "step": 9339 }, { "epoch": 1.0125758889852559, "grad_norm": 2.0488853454589844, "learning_rate": 1.3297548237342232e-05, "loss": 0.0236, "step": 9340 }, { "epoch": 1.0126843018213356, "grad_norm": 1.2910748720169067, "learning_rate": 1.3296822863774844e-05, "loss": 0.0391, "step": 9341 }, { "epoch": 1.0127927146574154, "grad_norm": 0.35455018281936646, "learning_rate": 1.3296097490207458e-05, "loss": 0.0021, "step": 9342 }, { "epoch": 1.0129011274934951, "grad_norm": 2.373521566390991, "learning_rate": 1.329537211664007e-05, "loss": 0.0298, "step": 9343 }, { "epoch": 1.0130095403295751, "grad_norm": 5.504181861877441, "learning_rate": 1.3294646743072683e-05, "loss": 0.0131, "step": 9344 }, { "epoch": 1.013117953165655, "grad_norm": 4.283162593841553, "learning_rate": 1.3293921369505296e-05, "loss": 0.0186, "step": 9345 }, { "epoch": 1.0132263660017347, "grad_norm": 1.2418800592422485, "learning_rate": 1.3293195995937909e-05, "loss": 0.0166, "step": 9346 }, { "epoch": 1.0133347788378144, "grad_norm": 1.8958699703216553, "learning_rate": 1.329247062237052e-05, "loss": 0.0225, "step": 9347 }, { "epoch": 1.0134431916738942, "grad_norm": 4.264674663543701, "learning_rate": 1.3291745248803135e-05, "loss": 0.0381, "step": 9348 }, { "epoch": 1.013551604509974, "grad_norm": 5.147700786590576, "learning_rate": 1.3291019875235747e-05, "loss": 0.0386, "step": 9349 }, { "epoch": 1.0136600173460537, "grad_norm": 3.422966718673706, "learning_rate": 1.3290294501668359e-05, "loss": 0.037, "step": 9350 }, { "epoch": 1.0137684301821335, "grad_norm": 10.081551551818848, "learning_rate": 1.3289569128100973e-05, "loss": 0.0127, "step": 9351 }, { "epoch": 1.0138768430182135, "grad_norm": 2.154238224029541, "learning_rate": 1.3288843754533587e-05, "loss": 0.0783, "step": 9352 }, { "epoch": 1.0139852558542932, "grad_norm": 2.177427291870117, "learning_rate": 1.32881183809662e-05, "loss": 0.0138, "step": 9353 }, { "epoch": 1.014093668690373, "grad_norm": 0.9087066650390625, "learning_rate": 1.3287393007398813e-05, "loss": 0.0203, "step": 9354 }, { "epoch": 1.0142020815264527, "grad_norm": 6.247570991516113, "learning_rate": 1.3286667633831425e-05, "loss": 0.0492, "step": 9355 }, { "epoch": 1.0143104943625325, "grad_norm": 6.750767230987549, "learning_rate": 1.3285942260264037e-05, "loss": 0.0294, "step": 9356 }, { "epoch": 1.0144189071986123, "grad_norm": 3.366105794906616, "learning_rate": 1.328521688669665e-05, "loss": 0.0059, "step": 9357 }, { "epoch": 1.014527320034692, "grad_norm": 3.341885805130005, "learning_rate": 1.3284491513129263e-05, "loss": 0.0098, "step": 9358 }, { "epoch": 1.0146357328707718, "grad_norm": 6.384799957275391, "learning_rate": 1.3283766139561875e-05, "loss": 0.0793, "step": 9359 }, { "epoch": 1.0147441457068518, "grad_norm": 0.6344824433326721, "learning_rate": 1.3283040765994489e-05, "loss": 0.0126, "step": 9360 }, { "epoch": 1.0148525585429315, "grad_norm": 0.6828880906105042, "learning_rate": 1.3282315392427101e-05, "loss": 0.018, "step": 9361 }, { "epoch": 1.0149609713790113, "grad_norm": 0.37745535373687744, "learning_rate": 1.3281590018859713e-05, "loss": 0.0105, "step": 9362 }, { "epoch": 1.015069384215091, "grad_norm": 0.10570042580366135, "learning_rate": 1.3280864645292327e-05, "loss": 0.0008, "step": 9363 }, { "epoch": 1.0151777970511708, "grad_norm": 1.5955431461334229, "learning_rate": 1.3280139271724939e-05, "loss": 0.0207, "step": 9364 }, { "epoch": 1.0152862098872506, "grad_norm": 0.025287926197052002, "learning_rate": 1.3279413898157551e-05, "loss": 0.0002, "step": 9365 }, { "epoch": 1.0153946227233304, "grad_norm": 14.215418815612793, "learning_rate": 1.3278688524590165e-05, "loss": 0.0337, "step": 9366 }, { "epoch": 1.0155030355594101, "grad_norm": 2.3650362491607666, "learning_rate": 1.3277963151022777e-05, "loss": 0.0019, "step": 9367 }, { "epoch": 1.0156114483954901, "grad_norm": 2.524819850921631, "learning_rate": 1.327723777745539e-05, "loss": 0.0407, "step": 9368 }, { "epoch": 1.0157198612315699, "grad_norm": 2.036526918411255, "learning_rate": 1.3276512403888005e-05, "loss": 0.0442, "step": 9369 }, { "epoch": 1.0158282740676496, "grad_norm": 2.8777008056640625, "learning_rate": 1.3275787030320617e-05, "loss": 0.0176, "step": 9370 }, { "epoch": 1.0159366869037294, "grad_norm": 0.15895597636699677, "learning_rate": 1.3275061656753229e-05, "loss": 0.0009, "step": 9371 }, { "epoch": 1.0160450997398092, "grad_norm": 3.4775569438934326, "learning_rate": 1.3274336283185843e-05, "loss": 0.0339, "step": 9372 }, { "epoch": 1.016153512575889, "grad_norm": 0.3334898054599762, "learning_rate": 1.3273610909618455e-05, "loss": 0.0066, "step": 9373 }, { "epoch": 1.0162619254119687, "grad_norm": 0.9487817883491516, "learning_rate": 1.3272885536051067e-05, "loss": 0.0255, "step": 9374 }, { "epoch": 1.0163703382480487, "grad_norm": 2.5064096450805664, "learning_rate": 1.3272160162483681e-05, "loss": 0.0622, "step": 9375 }, { "epoch": 1.0164787510841284, "grad_norm": 2.0381479263305664, "learning_rate": 1.3271434788916293e-05, "loss": 0.0264, "step": 9376 }, { "epoch": 1.0165871639202082, "grad_norm": 0.11259004473686218, "learning_rate": 1.3270709415348905e-05, "loss": 0.0006, "step": 9377 }, { "epoch": 1.016695576756288, "grad_norm": 8.205733299255371, "learning_rate": 1.326998404178152e-05, "loss": 0.0442, "step": 9378 }, { "epoch": 1.0168039895923677, "grad_norm": 4.266232490539551, "learning_rate": 1.3269258668214131e-05, "loss": 0.0446, "step": 9379 }, { "epoch": 1.0169124024284475, "grad_norm": 0.43645229935646057, "learning_rate": 1.3268533294646743e-05, "loss": 0.0136, "step": 9380 }, { "epoch": 1.0170208152645273, "grad_norm": 0.2960815131664276, "learning_rate": 1.3267807921079356e-05, "loss": 0.0009, "step": 9381 }, { "epoch": 1.017129228100607, "grad_norm": 2.354936122894287, "learning_rate": 1.326708254751197e-05, "loss": 0.0231, "step": 9382 }, { "epoch": 1.017237640936687, "grad_norm": 2.4253430366516113, "learning_rate": 1.3266357173944582e-05, "loss": 0.0217, "step": 9383 }, { "epoch": 1.0173460537727668, "grad_norm": 1.4187004566192627, "learning_rate": 1.3265631800377194e-05, "loss": 0.0219, "step": 9384 }, { "epoch": 1.0174544666088465, "grad_norm": 0.8743941187858582, "learning_rate": 1.3264906426809808e-05, "loss": 0.0265, "step": 9385 }, { "epoch": 1.0175628794449263, "grad_norm": 0.2460465133190155, "learning_rate": 1.3264181053242421e-05, "loss": 0.0051, "step": 9386 }, { "epoch": 1.017671292281006, "grad_norm": 5.031276702880859, "learning_rate": 1.3263455679675035e-05, "loss": 0.0292, "step": 9387 }, { "epoch": 1.0177797051170858, "grad_norm": 1.8320412635803223, "learning_rate": 1.3262730306107647e-05, "loss": 0.029, "step": 9388 }, { "epoch": 1.0178881179531656, "grad_norm": 0.16361360251903534, "learning_rate": 1.326200493254026e-05, "loss": 0.0019, "step": 9389 }, { "epoch": 1.0179965307892453, "grad_norm": 0.5066527128219604, "learning_rate": 1.3261279558972873e-05, "loss": 0.0129, "step": 9390 }, { "epoch": 1.0181049436253253, "grad_norm": 0.40400460362434387, "learning_rate": 1.3260554185405486e-05, "loss": 0.0014, "step": 9391 }, { "epoch": 1.018213356461405, "grad_norm": 11.00551700592041, "learning_rate": 1.3259828811838098e-05, "loss": 0.0742, "step": 9392 }, { "epoch": 1.0183217692974849, "grad_norm": 0.5432352423667908, "learning_rate": 1.325910343827071e-05, "loss": 0.0038, "step": 9393 }, { "epoch": 1.0184301821335646, "grad_norm": 0.334687739610672, "learning_rate": 1.3258378064703324e-05, "loss": 0.0091, "step": 9394 }, { "epoch": 1.0185385949696444, "grad_norm": 0.7933147549629211, "learning_rate": 1.3257652691135936e-05, "loss": 0.0075, "step": 9395 }, { "epoch": 1.0186470078057241, "grad_norm": 12.35422420501709, "learning_rate": 1.3256927317568548e-05, "loss": 0.0267, "step": 9396 }, { "epoch": 1.018755420641804, "grad_norm": 1.9854596853256226, "learning_rate": 1.3256201944001162e-05, "loss": 0.0109, "step": 9397 }, { "epoch": 1.0188638334778837, "grad_norm": 0.3926622271537781, "learning_rate": 1.3255476570433774e-05, "loss": 0.0088, "step": 9398 }, { "epoch": 1.0189722463139637, "grad_norm": 2.0265743732452393, "learning_rate": 1.3254751196866386e-05, "loss": 0.0196, "step": 9399 }, { "epoch": 1.0190806591500434, "grad_norm": 2.280947208404541, "learning_rate": 1.3254025823299e-05, "loss": 0.0263, "step": 9400 }, { "epoch": 1.0191890719861232, "grad_norm": 1.7174869775772095, "learning_rate": 1.3253300449731612e-05, "loss": 0.0169, "step": 9401 }, { "epoch": 1.019297484822203, "grad_norm": 4.377513408660889, "learning_rate": 1.3252575076164224e-05, "loss": 0.0374, "step": 9402 }, { "epoch": 1.0194058976582827, "grad_norm": 4.53632116317749, "learning_rate": 1.325184970259684e-05, "loss": 0.0731, "step": 9403 }, { "epoch": 1.0195143104943625, "grad_norm": 0.307345449924469, "learning_rate": 1.3251124329029452e-05, "loss": 0.0181, "step": 9404 }, { "epoch": 1.0196227233304422, "grad_norm": 1.6500953435897827, "learning_rate": 1.3250398955462066e-05, "loss": 0.0547, "step": 9405 }, { "epoch": 1.0197311361665222, "grad_norm": 3.4982597827911377, "learning_rate": 1.3249673581894678e-05, "loss": 0.0379, "step": 9406 }, { "epoch": 1.019839549002602, "grad_norm": 2.1775639057159424, "learning_rate": 1.324894820832729e-05, "loss": 0.0354, "step": 9407 }, { "epoch": 1.0199479618386817, "grad_norm": 4.544557571411133, "learning_rate": 1.3248222834759902e-05, "loss": 0.0298, "step": 9408 }, { "epoch": 1.0200563746747615, "grad_norm": 1.6316020488739014, "learning_rate": 1.3247497461192516e-05, "loss": 0.0197, "step": 9409 }, { "epoch": 1.0201647875108413, "grad_norm": 1.653757929801941, "learning_rate": 1.3246772087625128e-05, "loss": 0.0147, "step": 9410 }, { "epoch": 1.020273200346921, "grad_norm": 4.661936283111572, "learning_rate": 1.324604671405774e-05, "loss": 0.0388, "step": 9411 }, { "epoch": 1.0203816131830008, "grad_norm": 3.018101215362549, "learning_rate": 1.3245321340490354e-05, "loss": 0.0424, "step": 9412 }, { "epoch": 1.0204900260190806, "grad_norm": 0.6550666093826294, "learning_rate": 1.3244595966922966e-05, "loss": 0.0117, "step": 9413 }, { "epoch": 1.0205984388551606, "grad_norm": 0.3288068175315857, "learning_rate": 1.3243870593355578e-05, "loss": 0.0025, "step": 9414 }, { "epoch": 1.0207068516912403, "grad_norm": 0.3029256761074066, "learning_rate": 1.3243145219788192e-05, "loss": 0.0021, "step": 9415 }, { "epoch": 1.02081526452732, "grad_norm": 1.0488696098327637, "learning_rate": 1.3242419846220804e-05, "loss": 0.0018, "step": 9416 }, { "epoch": 1.0209236773633998, "grad_norm": 1.9018160104751587, "learning_rate": 1.3241694472653417e-05, "loss": 0.0248, "step": 9417 }, { "epoch": 1.0210320901994796, "grad_norm": 1.321307897567749, "learning_rate": 1.324096909908603e-05, "loss": 0.0074, "step": 9418 }, { "epoch": 1.0211405030355594, "grad_norm": 3.186765193939209, "learning_rate": 1.3240243725518643e-05, "loss": 0.0062, "step": 9419 }, { "epoch": 1.0212489158716391, "grad_norm": 2.049888849258423, "learning_rate": 1.3239518351951255e-05, "loss": 0.0246, "step": 9420 }, { "epoch": 1.021357328707719, "grad_norm": 1.6911410093307495, "learning_rate": 1.323879297838387e-05, "loss": 0.0329, "step": 9421 }, { "epoch": 1.0214657415437989, "grad_norm": 0.9529488682746887, "learning_rate": 1.3238067604816482e-05, "loss": 0.025, "step": 9422 }, { "epoch": 1.0215741543798786, "grad_norm": 0.0897672176361084, "learning_rate": 1.3237342231249094e-05, "loss": 0.0006, "step": 9423 }, { "epoch": 1.0216825672159584, "grad_norm": 4.144759178161621, "learning_rate": 1.3236616857681708e-05, "loss": 0.0169, "step": 9424 }, { "epoch": 1.0217909800520382, "grad_norm": 1.6256872415542603, "learning_rate": 1.323589148411432e-05, "loss": 0.0059, "step": 9425 }, { "epoch": 1.021899392888118, "grad_norm": 0.8651663064956665, "learning_rate": 1.3235166110546933e-05, "loss": 0.0031, "step": 9426 }, { "epoch": 1.0220078057241977, "grad_norm": 5.14158821105957, "learning_rate": 1.3234440736979546e-05, "loss": 0.0714, "step": 9427 }, { "epoch": 1.0221162185602775, "grad_norm": 1.7570545673370361, "learning_rate": 1.3233715363412159e-05, "loss": 0.0156, "step": 9428 }, { "epoch": 1.0222246313963572, "grad_norm": 1.497058391571045, "learning_rate": 1.323298998984477e-05, "loss": 0.0196, "step": 9429 }, { "epoch": 1.0223330442324372, "grad_norm": 2.950854778289795, "learning_rate": 1.3232264616277385e-05, "loss": 0.0223, "step": 9430 }, { "epoch": 1.022441457068517, "grad_norm": 1.805001974105835, "learning_rate": 1.3231539242709997e-05, "loss": 0.0193, "step": 9431 }, { "epoch": 1.0225498699045967, "grad_norm": 0.7657067179679871, "learning_rate": 1.3230813869142609e-05, "loss": 0.026, "step": 9432 }, { "epoch": 1.0226582827406765, "grad_norm": 2.412046432495117, "learning_rate": 1.3230088495575221e-05, "loss": 0.0171, "step": 9433 }, { "epoch": 1.0227666955767563, "grad_norm": 0.9519621133804321, "learning_rate": 1.3229363122007835e-05, "loss": 0.0109, "step": 9434 }, { "epoch": 1.022875108412836, "grad_norm": 0.24722886085510254, "learning_rate": 1.3228637748440447e-05, "loss": 0.0019, "step": 9435 }, { "epoch": 1.0229835212489158, "grad_norm": 3.162975788116455, "learning_rate": 1.3227912374873059e-05, "loss": 0.0292, "step": 9436 }, { "epoch": 1.0230919340849958, "grad_norm": 1.7104769945144653, "learning_rate": 1.3227187001305673e-05, "loss": 0.037, "step": 9437 }, { "epoch": 1.0232003469210755, "grad_norm": 0.7699774503707886, "learning_rate": 1.3226461627738287e-05, "loss": 0.0028, "step": 9438 }, { "epoch": 1.0233087597571553, "grad_norm": 3.500216245651245, "learning_rate": 1.32257362541709e-05, "loss": 0.0445, "step": 9439 }, { "epoch": 1.023417172593235, "grad_norm": 2.1926441192626953, "learning_rate": 1.3225010880603513e-05, "loss": 0.0369, "step": 9440 }, { "epoch": 1.0235255854293148, "grad_norm": 8.710352897644043, "learning_rate": 1.3224285507036125e-05, "loss": 0.0181, "step": 9441 }, { "epoch": 1.0236339982653946, "grad_norm": 1.1724269390106201, "learning_rate": 1.3223560133468739e-05, "loss": 0.0127, "step": 9442 }, { "epoch": 1.0237424111014743, "grad_norm": 2.699519157409668, "learning_rate": 1.3222834759901351e-05, "loss": 0.0553, "step": 9443 }, { "epoch": 1.023850823937554, "grad_norm": 1.2144542932510376, "learning_rate": 1.3222109386333963e-05, "loss": 0.0316, "step": 9444 }, { "epoch": 1.023959236773634, "grad_norm": 1.4398514032363892, "learning_rate": 1.3221384012766575e-05, "loss": 0.0146, "step": 9445 }, { "epoch": 1.0240676496097139, "grad_norm": 2.7719767093658447, "learning_rate": 1.3220658639199189e-05, "loss": 0.025, "step": 9446 }, { "epoch": 1.0241760624457936, "grad_norm": 0.7797651886940002, "learning_rate": 1.3219933265631801e-05, "loss": 0.0034, "step": 9447 }, { "epoch": 1.0242844752818734, "grad_norm": 3.242371082305908, "learning_rate": 1.3219207892064413e-05, "loss": 0.0176, "step": 9448 }, { "epoch": 1.0243928881179531, "grad_norm": 1.2921786308288574, "learning_rate": 1.3218482518497027e-05, "loss": 0.0107, "step": 9449 }, { "epoch": 1.024501300954033, "grad_norm": 0.4053630232810974, "learning_rate": 1.321775714492964e-05, "loss": 0.0107, "step": 9450 }, { "epoch": 1.0246097137901127, "grad_norm": 0.8505823016166687, "learning_rate": 1.3217031771362251e-05, "loss": 0.0179, "step": 9451 }, { "epoch": 1.0247181266261924, "grad_norm": 5.031578063964844, "learning_rate": 1.3216306397794865e-05, "loss": 0.0296, "step": 9452 }, { "epoch": 1.0248265394622724, "grad_norm": 1.77387535572052, "learning_rate": 1.3215581024227477e-05, "loss": 0.0086, "step": 9453 }, { "epoch": 1.0249349522983522, "grad_norm": 0.7277616858482361, "learning_rate": 1.321485565066009e-05, "loss": 0.0176, "step": 9454 }, { "epoch": 1.025043365134432, "grad_norm": 1.6382100582122803, "learning_rate": 1.3214130277092705e-05, "loss": 0.0818, "step": 9455 }, { "epoch": 1.0251517779705117, "grad_norm": 1.5169456005096436, "learning_rate": 1.3213404903525317e-05, "loss": 0.015, "step": 9456 }, { "epoch": 1.0252601908065915, "grad_norm": 3.186310291290283, "learning_rate": 1.3212679529957931e-05, "loss": 0.0818, "step": 9457 }, { "epoch": 1.0253686036426712, "grad_norm": 3.070282459259033, "learning_rate": 1.3211954156390543e-05, "loss": 0.0331, "step": 9458 }, { "epoch": 1.025477016478751, "grad_norm": 0.7400879859924316, "learning_rate": 1.3211228782823155e-05, "loss": 0.0301, "step": 9459 }, { "epoch": 1.0255854293148308, "grad_norm": 1.3668426275253296, "learning_rate": 1.3210503409255767e-05, "loss": 0.0073, "step": 9460 }, { "epoch": 1.0256938421509108, "grad_norm": 0.41702520847320557, "learning_rate": 1.3209778035688381e-05, "loss": 0.0025, "step": 9461 }, { "epoch": 1.0258022549869905, "grad_norm": 0.9205644726753235, "learning_rate": 1.3209052662120993e-05, "loss": 0.0125, "step": 9462 }, { "epoch": 1.0259106678230703, "grad_norm": 0.299262136220932, "learning_rate": 1.3208327288553606e-05, "loss": 0.0008, "step": 9463 }, { "epoch": 1.02601908065915, "grad_norm": 0.4770534932613373, "learning_rate": 1.320760191498622e-05, "loss": 0.0198, "step": 9464 }, { "epoch": 1.0261274934952298, "grad_norm": 3.877173900604248, "learning_rate": 1.3206876541418832e-05, "loss": 0.0706, "step": 9465 }, { "epoch": 1.0262359063313096, "grad_norm": 2.2828867435455322, "learning_rate": 1.3206151167851444e-05, "loss": 0.0199, "step": 9466 }, { "epoch": 1.0263443191673893, "grad_norm": 5.307802677154541, "learning_rate": 1.3205425794284058e-05, "loss": 0.0194, "step": 9467 }, { "epoch": 1.0264527320034693, "grad_norm": 2.3201422691345215, "learning_rate": 1.320470042071667e-05, "loss": 0.0497, "step": 9468 }, { "epoch": 1.026561144839549, "grad_norm": 5.978251934051514, "learning_rate": 1.3203975047149282e-05, "loss": 0.0635, "step": 9469 }, { "epoch": 1.0266695576756288, "grad_norm": 1.6096981763839722, "learning_rate": 1.3203249673581896e-05, "loss": 0.0019, "step": 9470 }, { "epoch": 1.0267779705117086, "grad_norm": 0.12463278323411942, "learning_rate": 1.3202524300014508e-05, "loss": 0.0011, "step": 9471 }, { "epoch": 1.0268863833477884, "grad_norm": 0.4437178373336792, "learning_rate": 1.320179892644712e-05, "loss": 0.0186, "step": 9472 }, { "epoch": 1.0269947961838681, "grad_norm": 2.515230178833008, "learning_rate": 1.3201073552879736e-05, "loss": 0.0345, "step": 9473 }, { "epoch": 1.027103209019948, "grad_norm": 2.1209919452667236, "learning_rate": 1.3200348179312348e-05, "loss": 0.0197, "step": 9474 }, { "epoch": 1.0272116218560277, "grad_norm": 0.08235795050859451, "learning_rate": 1.319962280574496e-05, "loss": 0.0016, "step": 9475 }, { "epoch": 1.0273200346921076, "grad_norm": 1.8828084468841553, "learning_rate": 1.3198897432177574e-05, "loss": 0.0348, "step": 9476 }, { "epoch": 1.0274284475281874, "grad_norm": 0.7242196202278137, "learning_rate": 1.3198172058610186e-05, "loss": 0.0168, "step": 9477 }, { "epoch": 1.0275368603642672, "grad_norm": 1.4820002317428589, "learning_rate": 1.3197446685042798e-05, "loss": 0.0051, "step": 9478 }, { "epoch": 1.027645273200347, "grad_norm": 2.386916160583496, "learning_rate": 1.3196721311475412e-05, "loss": 0.0286, "step": 9479 }, { "epoch": 1.0277536860364267, "grad_norm": 2.3359627723693848, "learning_rate": 1.3195995937908024e-05, "loss": 0.0816, "step": 9480 }, { "epoch": 1.0278620988725065, "grad_norm": 0.06800765544176102, "learning_rate": 1.3195270564340636e-05, "loss": 0.0004, "step": 9481 }, { "epoch": 1.0279705117085862, "grad_norm": 9.421839714050293, "learning_rate": 1.319454519077325e-05, "loss": 0.0304, "step": 9482 }, { "epoch": 1.028078924544666, "grad_norm": 0.13627177476882935, "learning_rate": 1.3193819817205862e-05, "loss": 0.0009, "step": 9483 }, { "epoch": 1.028187337380746, "grad_norm": 1.9159680604934692, "learning_rate": 1.3193094443638474e-05, "loss": 0.0109, "step": 9484 }, { "epoch": 1.0282957502168257, "grad_norm": 11.545697212219238, "learning_rate": 1.3192369070071086e-05, "loss": 0.0414, "step": 9485 }, { "epoch": 1.0284041630529055, "grad_norm": 1.0715094804763794, "learning_rate": 1.31916436965037e-05, "loss": 0.0052, "step": 9486 }, { "epoch": 1.0285125758889853, "grad_norm": 4.732147216796875, "learning_rate": 1.3190918322936312e-05, "loss": 0.0281, "step": 9487 }, { "epoch": 1.028620988725065, "grad_norm": 1.86595618724823, "learning_rate": 1.3190192949368924e-05, "loss": 0.0169, "step": 9488 }, { "epoch": 1.0287294015611448, "grad_norm": 0.7271178364753723, "learning_rate": 1.3189467575801538e-05, "loss": 0.0136, "step": 9489 }, { "epoch": 1.0288378143972245, "grad_norm": 6.329071998596191, "learning_rate": 1.3188742202234152e-05, "loss": 0.0444, "step": 9490 }, { "epoch": 1.0289462272333045, "grad_norm": 2.3626937866210938, "learning_rate": 1.3188016828666766e-05, "loss": 0.0314, "step": 9491 }, { "epoch": 1.0290546400693843, "grad_norm": 4.978950023651123, "learning_rate": 1.3187291455099378e-05, "loss": 0.0547, "step": 9492 }, { "epoch": 1.029163052905464, "grad_norm": 4.366977214813232, "learning_rate": 1.318656608153199e-05, "loss": 0.0278, "step": 9493 }, { "epoch": 1.0292714657415438, "grad_norm": 1.7117637395858765, "learning_rate": 1.3185840707964604e-05, "loss": 0.0126, "step": 9494 }, { "epoch": 1.0293798785776236, "grad_norm": 0.3978598415851593, "learning_rate": 1.3185115334397216e-05, "loss": 0.0072, "step": 9495 }, { "epoch": 1.0294882914137033, "grad_norm": 2.8162896633148193, "learning_rate": 1.3184389960829828e-05, "loss": 0.0357, "step": 9496 }, { "epoch": 1.0295967042497831, "grad_norm": 0.8311901688575745, "learning_rate": 1.3183664587262442e-05, "loss": 0.0089, "step": 9497 }, { "epoch": 1.0297051170858629, "grad_norm": 0.3603347837924957, "learning_rate": 1.3182939213695054e-05, "loss": 0.006, "step": 9498 }, { "epoch": 1.0298135299219429, "grad_norm": 7.081571102142334, "learning_rate": 1.3182213840127666e-05, "loss": 0.0226, "step": 9499 }, { "epoch": 1.0299219427580226, "grad_norm": 2.1861369609832764, "learning_rate": 1.3181488466560279e-05, "loss": 0.0458, "step": 9500 }, { "epoch": 1.0300303555941024, "grad_norm": 1.668532133102417, "learning_rate": 1.3180763092992892e-05, "loss": 0.0243, "step": 9501 }, { "epoch": 1.0301387684301822, "grad_norm": 0.6975701451301575, "learning_rate": 1.3180037719425505e-05, "loss": 0.0108, "step": 9502 }, { "epoch": 1.030247181266262, "grad_norm": 2.3260293006896973, "learning_rate": 1.3179312345858117e-05, "loss": 0.01, "step": 9503 }, { "epoch": 1.0303555941023417, "grad_norm": 2.7223849296569824, "learning_rate": 1.317858697229073e-05, "loss": 0.0192, "step": 9504 }, { "epoch": 1.0304640069384214, "grad_norm": 0.33833152055740356, "learning_rate": 1.3177861598723343e-05, "loss": 0.0116, "step": 9505 }, { "epoch": 1.0305724197745012, "grad_norm": 2.1227166652679443, "learning_rate": 1.3177136225155955e-05, "loss": 0.0336, "step": 9506 }, { "epoch": 1.0306808326105812, "grad_norm": 0.1610575020313263, "learning_rate": 1.317641085158857e-05, "loss": 0.0012, "step": 9507 }, { "epoch": 1.030789245446661, "grad_norm": 1.9397251605987549, "learning_rate": 1.3175685478021183e-05, "loss": 0.0162, "step": 9508 }, { "epoch": 1.0308976582827407, "grad_norm": 7.498363971710205, "learning_rate": 1.3174960104453796e-05, "loss": 0.0713, "step": 9509 }, { "epoch": 1.0310060711188205, "grad_norm": 1.294293761253357, "learning_rate": 1.3174234730886409e-05, "loss": 0.0115, "step": 9510 }, { "epoch": 1.0311144839549002, "grad_norm": 1.5989114046096802, "learning_rate": 1.317350935731902e-05, "loss": 0.0021, "step": 9511 }, { "epoch": 1.03122289679098, "grad_norm": 3.3361527919769287, "learning_rate": 1.3172783983751633e-05, "loss": 0.0721, "step": 9512 }, { "epoch": 1.0313313096270598, "grad_norm": 2.100346803665161, "learning_rate": 1.3172058610184247e-05, "loss": 0.038, "step": 9513 }, { "epoch": 1.0314397224631395, "grad_norm": 2.182663679122925, "learning_rate": 1.3171333236616859e-05, "loss": 0.0574, "step": 9514 }, { "epoch": 1.0315481352992195, "grad_norm": 0.6589436531066895, "learning_rate": 1.3170607863049471e-05, "loss": 0.0225, "step": 9515 }, { "epoch": 1.0316565481352993, "grad_norm": 0.24953316152095795, "learning_rate": 1.3169882489482085e-05, "loss": 0.0071, "step": 9516 }, { "epoch": 1.031764960971379, "grad_norm": 1.0300370454788208, "learning_rate": 1.3169157115914697e-05, "loss": 0.0323, "step": 9517 }, { "epoch": 1.0318733738074588, "grad_norm": 19.86367416381836, "learning_rate": 1.3168431742347309e-05, "loss": 0.0509, "step": 9518 }, { "epoch": 1.0319817866435386, "grad_norm": 0.22433403134346008, "learning_rate": 1.3167706368779923e-05, "loss": 0.0016, "step": 9519 }, { "epoch": 1.0320901994796183, "grad_norm": 0.4948521554470062, "learning_rate": 1.3166980995212535e-05, "loss": 0.0014, "step": 9520 }, { "epoch": 1.032198612315698, "grad_norm": 0.30227383971214294, "learning_rate": 1.3166255621645147e-05, "loss": 0.0015, "step": 9521 }, { "epoch": 1.0323070251517779, "grad_norm": 0.5932102799415588, "learning_rate": 1.3165530248077761e-05, "loss": 0.0073, "step": 9522 }, { "epoch": 1.0324154379878578, "grad_norm": 5.519566059112549, "learning_rate": 1.3164804874510373e-05, "loss": 0.0273, "step": 9523 }, { "epoch": 1.0325238508239376, "grad_norm": 2.6992204189300537, "learning_rate": 1.3164079500942987e-05, "loss": 0.0328, "step": 9524 }, { "epoch": 1.0326322636600174, "grad_norm": 0.1123848482966423, "learning_rate": 1.3163354127375601e-05, "loss": 0.0007, "step": 9525 }, { "epoch": 1.0327406764960971, "grad_norm": 2.5054750442504883, "learning_rate": 1.3162628753808213e-05, "loss": 0.0217, "step": 9526 }, { "epoch": 1.032849089332177, "grad_norm": 3.3442201614379883, "learning_rate": 1.3161903380240825e-05, "loss": 0.0166, "step": 9527 }, { "epoch": 1.0329575021682567, "grad_norm": 0.9420766234397888, "learning_rate": 1.3161178006673439e-05, "loss": 0.0162, "step": 9528 }, { "epoch": 1.0330659150043364, "grad_norm": 6.114983558654785, "learning_rate": 1.3160452633106051e-05, "loss": 0.018, "step": 9529 }, { "epoch": 1.0331743278404164, "grad_norm": 3.0170202255249023, "learning_rate": 1.3159727259538663e-05, "loss": 0.0219, "step": 9530 }, { "epoch": 1.0332827406764962, "grad_norm": 1.943117380142212, "learning_rate": 1.3159001885971277e-05, "loss": 0.0823, "step": 9531 }, { "epoch": 1.033391153512576, "grad_norm": 2.838935613632202, "learning_rate": 1.315827651240389e-05, "loss": 0.0274, "step": 9532 }, { "epoch": 1.0334995663486557, "grad_norm": 0.36606743931770325, "learning_rate": 1.3157551138836501e-05, "loss": 0.0019, "step": 9533 }, { "epoch": 1.0336079791847355, "grad_norm": 4.886092662811279, "learning_rate": 1.3156825765269115e-05, "loss": 0.0844, "step": 9534 }, { "epoch": 1.0337163920208152, "grad_norm": 0.46374228596687317, "learning_rate": 1.3156100391701727e-05, "loss": 0.0174, "step": 9535 }, { "epoch": 1.033824804856895, "grad_norm": 3.574962854385376, "learning_rate": 1.315537501813434e-05, "loss": 0.0323, "step": 9536 }, { "epoch": 1.0339332176929747, "grad_norm": 0.4274231195449829, "learning_rate": 1.3154649644566952e-05, "loss": 0.0106, "step": 9537 }, { "epoch": 1.0340416305290547, "grad_norm": 1.9693917036056519, "learning_rate": 1.3153924270999565e-05, "loss": 0.0078, "step": 9538 }, { "epoch": 1.0341500433651345, "grad_norm": 2.0817651748657227, "learning_rate": 1.3153198897432178e-05, "loss": 0.0222, "step": 9539 }, { "epoch": 1.0342584562012143, "grad_norm": 2.724282741546631, "learning_rate": 1.315247352386479e-05, "loss": 0.012, "step": 9540 }, { "epoch": 1.034366869037294, "grad_norm": 1.8214479684829712, "learning_rate": 1.3151748150297404e-05, "loss": 0.009, "step": 9541 }, { "epoch": 1.0344752818733738, "grad_norm": 1.0092777013778687, "learning_rate": 1.3151022776730017e-05, "loss": 0.0253, "step": 9542 }, { "epoch": 1.0345836947094535, "grad_norm": 3.5180187225341797, "learning_rate": 1.3150297403162631e-05, "loss": 0.0416, "step": 9543 }, { "epoch": 1.0346921075455333, "grad_norm": 3.6618542671203613, "learning_rate": 1.3149572029595243e-05, "loss": 0.0321, "step": 9544 }, { "epoch": 1.034800520381613, "grad_norm": 7.5645341873168945, "learning_rate": 1.3148846656027856e-05, "loss": 0.0699, "step": 9545 }, { "epoch": 1.034908933217693, "grad_norm": 0.4065364897251129, "learning_rate": 1.314812128246047e-05, "loss": 0.0047, "step": 9546 }, { "epoch": 1.0350173460537728, "grad_norm": 7.133181095123291, "learning_rate": 1.3147395908893082e-05, "loss": 0.068, "step": 9547 }, { "epoch": 1.0351257588898526, "grad_norm": 0.20589418709278107, "learning_rate": 1.3146670535325694e-05, "loss": 0.0051, "step": 9548 }, { "epoch": 1.0352341717259324, "grad_norm": 0.4857299029827118, "learning_rate": 1.3145945161758308e-05, "loss": 0.002, "step": 9549 }, { "epoch": 1.0353425845620121, "grad_norm": 2.190272331237793, "learning_rate": 1.314521978819092e-05, "loss": 0.0443, "step": 9550 }, { "epoch": 1.0354509973980919, "grad_norm": 3.12076735496521, "learning_rate": 1.3144494414623532e-05, "loss": 0.0185, "step": 9551 }, { "epoch": 1.0355594102341716, "grad_norm": 1.8431357145309448, "learning_rate": 1.3143769041056144e-05, "loss": 0.0208, "step": 9552 }, { "epoch": 1.0356678230702516, "grad_norm": 2.189878225326538, "learning_rate": 1.3143043667488758e-05, "loss": 0.0144, "step": 9553 }, { "epoch": 1.0357762359063314, "grad_norm": 0.3623190224170685, "learning_rate": 1.314231829392137e-05, "loss": 0.0022, "step": 9554 }, { "epoch": 1.0358846487424112, "grad_norm": 15.660726547241211, "learning_rate": 1.3141592920353982e-05, "loss": 0.0489, "step": 9555 }, { "epoch": 1.035993061578491, "grad_norm": 5.085871696472168, "learning_rate": 1.3140867546786596e-05, "loss": 0.0677, "step": 9556 }, { "epoch": 1.0361014744145707, "grad_norm": 1.7711859941482544, "learning_rate": 1.3140142173219208e-05, "loss": 0.0083, "step": 9557 }, { "epoch": 1.0362098872506504, "grad_norm": 0.2151520699262619, "learning_rate": 1.313941679965182e-05, "loss": 0.0015, "step": 9558 }, { "epoch": 1.0363183000867302, "grad_norm": 0.22094480693340302, "learning_rate": 1.3138691426084436e-05, "loss": 0.002, "step": 9559 }, { "epoch": 1.03642671292281, "grad_norm": 1.1073728799819946, "learning_rate": 1.3137966052517048e-05, "loss": 0.0182, "step": 9560 }, { "epoch": 1.03653512575889, "grad_norm": 6.063858509063721, "learning_rate": 1.3137240678949662e-05, "loss": 0.0599, "step": 9561 }, { "epoch": 1.0366435385949697, "grad_norm": 1.1468383073806763, "learning_rate": 1.3136515305382274e-05, "loss": 0.0147, "step": 9562 }, { "epoch": 1.0367519514310495, "grad_norm": 5.150400638580322, "learning_rate": 1.3135789931814886e-05, "loss": 0.0206, "step": 9563 }, { "epoch": 1.0368603642671292, "grad_norm": 1.1021158695220947, "learning_rate": 1.3135064558247498e-05, "loss": 0.0023, "step": 9564 }, { "epoch": 1.036968777103209, "grad_norm": 1.826781153678894, "learning_rate": 1.3134339184680112e-05, "loss": 0.0247, "step": 9565 }, { "epoch": 1.0370771899392888, "grad_norm": 6.071022033691406, "learning_rate": 1.3133613811112724e-05, "loss": 0.0112, "step": 9566 }, { "epoch": 1.0371856027753685, "grad_norm": 0.08254723250865936, "learning_rate": 1.3132888437545336e-05, "loss": 0.001, "step": 9567 }, { "epoch": 1.0372940156114483, "grad_norm": 0.3071206212043762, "learning_rate": 1.313216306397795e-05, "loss": 0.0022, "step": 9568 }, { "epoch": 1.0374024284475283, "grad_norm": 1.2838704586029053, "learning_rate": 1.3131437690410562e-05, "loss": 0.0046, "step": 9569 }, { "epoch": 1.037510841283608, "grad_norm": 2.4066197872161865, "learning_rate": 1.3130712316843174e-05, "loss": 0.0282, "step": 9570 }, { "epoch": 1.0376192541196878, "grad_norm": 0.15372343361377716, "learning_rate": 1.3129986943275788e-05, "loss": 0.0017, "step": 9571 }, { "epoch": 1.0377276669557676, "grad_norm": 0.5132650136947632, "learning_rate": 1.31292615697084e-05, "loss": 0.0177, "step": 9572 }, { "epoch": 1.0378360797918473, "grad_norm": 3.8894999027252197, "learning_rate": 1.3128536196141013e-05, "loss": 0.0865, "step": 9573 }, { "epoch": 1.037944492627927, "grad_norm": 1.6789565086364746, "learning_rate": 1.3127810822573626e-05, "loss": 0.0235, "step": 9574 }, { "epoch": 1.0380529054640069, "grad_norm": 0.9441251158714294, "learning_rate": 1.3127085449006239e-05, "loss": 0.0078, "step": 9575 }, { "epoch": 1.0381613183000868, "grad_norm": 4.9645915031433105, "learning_rate": 1.3126360075438852e-05, "loss": 0.0684, "step": 9576 }, { "epoch": 1.0382697311361666, "grad_norm": 4.088911056518555, "learning_rate": 1.3125634701871466e-05, "loss": 0.0466, "step": 9577 }, { "epoch": 1.0383781439722464, "grad_norm": 7.322424411773682, "learning_rate": 1.3124909328304078e-05, "loss": 0.0192, "step": 9578 }, { "epoch": 1.0384865568083261, "grad_norm": 3.961365222930908, "learning_rate": 1.312418395473669e-05, "loss": 0.0124, "step": 9579 }, { "epoch": 1.038594969644406, "grad_norm": 1.7133359909057617, "learning_rate": 1.3123458581169304e-05, "loss": 0.0058, "step": 9580 }, { "epoch": 1.0387033824804857, "grad_norm": 4.381832122802734, "learning_rate": 1.3122733207601916e-05, "loss": 0.0184, "step": 9581 }, { "epoch": 1.0388117953165654, "grad_norm": 5.418706893920898, "learning_rate": 1.3122007834034529e-05, "loss": 0.0394, "step": 9582 }, { "epoch": 1.0389202081526452, "grad_norm": 1.8612366914749146, "learning_rate": 1.3121282460467142e-05, "loss": 0.0274, "step": 9583 }, { "epoch": 1.0390286209887252, "grad_norm": 0.5312594771385193, "learning_rate": 1.3120557086899755e-05, "loss": 0.0102, "step": 9584 }, { "epoch": 1.039137033824805, "grad_norm": 2.159263849258423, "learning_rate": 1.3119831713332367e-05, "loss": 0.0205, "step": 9585 }, { "epoch": 1.0392454466608847, "grad_norm": 0.7147873640060425, "learning_rate": 1.311910633976498e-05, "loss": 0.0051, "step": 9586 }, { "epoch": 1.0393538594969645, "grad_norm": 2.596038579940796, "learning_rate": 1.3118380966197593e-05, "loss": 0.0178, "step": 9587 }, { "epoch": 1.0394622723330442, "grad_norm": 3.545832633972168, "learning_rate": 1.3117655592630205e-05, "loss": 0.0271, "step": 9588 }, { "epoch": 1.039570685169124, "grad_norm": 2.092759609222412, "learning_rate": 1.3116930219062817e-05, "loss": 0.0219, "step": 9589 }, { "epoch": 1.0396790980052038, "grad_norm": 4.332213401794434, "learning_rate": 1.311620484549543e-05, "loss": 0.0249, "step": 9590 }, { "epoch": 1.0397875108412835, "grad_norm": 0.11580104380846024, "learning_rate": 1.3115479471928043e-05, "loss": 0.0008, "step": 9591 }, { "epoch": 1.0398959236773635, "grad_norm": 3.5673165321350098, "learning_rate": 1.3114754098360655e-05, "loss": 0.0448, "step": 9592 }, { "epoch": 1.0400043365134433, "grad_norm": 0.6529038548469543, "learning_rate": 1.311402872479327e-05, "loss": 0.0255, "step": 9593 }, { "epoch": 1.040112749349523, "grad_norm": 5.341263294219971, "learning_rate": 1.3113303351225883e-05, "loss": 0.0215, "step": 9594 }, { "epoch": 1.0402211621856028, "grad_norm": 0.7256187796592712, "learning_rate": 1.3112577977658497e-05, "loss": 0.0109, "step": 9595 }, { "epoch": 1.0403295750216826, "grad_norm": 0.5663480758666992, "learning_rate": 1.3111852604091109e-05, "loss": 0.0094, "step": 9596 }, { "epoch": 1.0404379878577623, "grad_norm": 4.047170639038086, "learning_rate": 1.3111127230523721e-05, "loss": 0.0265, "step": 9597 }, { "epoch": 1.040546400693842, "grad_norm": 3.0901107788085938, "learning_rate": 1.3110401856956335e-05, "loss": 0.0241, "step": 9598 }, { "epoch": 1.0406548135299218, "grad_norm": 1.1178470849990845, "learning_rate": 1.3109676483388947e-05, "loss": 0.0138, "step": 9599 }, { "epoch": 1.0407632263660018, "grad_norm": 0.3359074890613556, "learning_rate": 1.3108951109821559e-05, "loss": 0.0022, "step": 9600 }, { "epoch": 1.0408716392020816, "grad_norm": 1.7394651174545288, "learning_rate": 1.3108225736254173e-05, "loss": 0.0173, "step": 9601 }, { "epoch": 1.0409800520381614, "grad_norm": 4.09516716003418, "learning_rate": 1.3107500362686785e-05, "loss": 0.1008, "step": 9602 }, { "epoch": 1.0410884648742411, "grad_norm": 0.7847360372543335, "learning_rate": 1.3106774989119397e-05, "loss": 0.0032, "step": 9603 }, { "epoch": 1.0411968777103209, "grad_norm": 0.33822959661483765, "learning_rate": 1.310604961555201e-05, "loss": 0.0084, "step": 9604 }, { "epoch": 1.0413052905464006, "grad_norm": 0.456688791513443, "learning_rate": 1.3105324241984623e-05, "loss": 0.0032, "step": 9605 }, { "epoch": 1.0414137033824804, "grad_norm": 2.966648578643799, "learning_rate": 1.3104598868417235e-05, "loss": 0.0519, "step": 9606 }, { "epoch": 1.0415221162185602, "grad_norm": 3.5058066844940186, "learning_rate": 1.3103873494849847e-05, "loss": 0.0289, "step": 9607 }, { "epoch": 1.0416305290546402, "grad_norm": 1.9366189241409302, "learning_rate": 1.3103148121282461e-05, "loss": 0.0292, "step": 9608 }, { "epoch": 1.04173894189072, "grad_norm": 2.6218678951263428, "learning_rate": 1.3102422747715073e-05, "loss": 0.0596, "step": 9609 }, { "epoch": 1.0418473547267997, "grad_norm": 1.87125825881958, "learning_rate": 1.3101697374147686e-05, "loss": 0.024, "step": 9610 }, { "epoch": 1.0419557675628794, "grad_norm": 1.7413030862808228, "learning_rate": 1.3100972000580301e-05, "loss": 0.0302, "step": 9611 }, { "epoch": 1.0420641803989592, "grad_norm": 1.3163701295852661, "learning_rate": 1.3100246627012913e-05, "loss": 0.0213, "step": 9612 }, { "epoch": 1.042172593235039, "grad_norm": 5.11528205871582, "learning_rate": 1.3099521253445527e-05, "loss": 0.0532, "step": 9613 }, { "epoch": 1.0422810060711187, "grad_norm": 8.589293479919434, "learning_rate": 1.309879587987814e-05, "loss": 0.0487, "step": 9614 }, { "epoch": 1.0423894189071987, "grad_norm": 0.3034510016441345, "learning_rate": 1.3098070506310751e-05, "loss": 0.0048, "step": 9615 }, { "epoch": 1.0424978317432785, "grad_norm": 0.7533201575279236, "learning_rate": 1.3097345132743363e-05, "loss": 0.0124, "step": 9616 }, { "epoch": 1.0426062445793582, "grad_norm": 3.771925926208496, "learning_rate": 1.3096619759175977e-05, "loss": 0.0116, "step": 9617 }, { "epoch": 1.042714657415438, "grad_norm": 2.736859083175659, "learning_rate": 1.309589438560859e-05, "loss": 0.0398, "step": 9618 }, { "epoch": 1.0428230702515178, "grad_norm": 1.2046737670898438, "learning_rate": 1.3095169012041202e-05, "loss": 0.0361, "step": 9619 }, { "epoch": 1.0429314830875975, "grad_norm": 2.2397284507751465, "learning_rate": 1.3094443638473815e-05, "loss": 0.0233, "step": 9620 }, { "epoch": 1.0430398959236773, "grad_norm": 1.006395697593689, "learning_rate": 1.3093718264906428e-05, "loss": 0.02, "step": 9621 }, { "epoch": 1.043148308759757, "grad_norm": 8.480888366699219, "learning_rate": 1.309299289133904e-05, "loss": 0.0772, "step": 9622 }, { "epoch": 1.043256721595837, "grad_norm": 0.9167270660400391, "learning_rate": 1.3092267517771654e-05, "loss": 0.0193, "step": 9623 }, { "epoch": 1.0433651344319168, "grad_norm": 4.213002681732178, "learning_rate": 1.3091542144204266e-05, "loss": 0.0633, "step": 9624 }, { "epoch": 1.0434735472679966, "grad_norm": 1.2907108068466187, "learning_rate": 1.3090816770636878e-05, "loss": 0.0206, "step": 9625 }, { "epoch": 1.0435819601040763, "grad_norm": 1.0483615398406982, "learning_rate": 1.3090091397069492e-05, "loss": 0.025, "step": 9626 }, { "epoch": 1.043690372940156, "grad_norm": 0.7403331995010376, "learning_rate": 1.3089366023502104e-05, "loss": 0.0244, "step": 9627 }, { "epoch": 1.0437987857762359, "grad_norm": 2.603494644165039, "learning_rate": 1.3088640649934718e-05, "loss": 0.0384, "step": 9628 }, { "epoch": 1.0439071986123156, "grad_norm": 0.39964887499809265, "learning_rate": 1.3087915276367332e-05, "loss": 0.0086, "step": 9629 }, { "epoch": 1.0440156114483954, "grad_norm": 1.6789218187332153, "learning_rate": 1.3087189902799944e-05, "loss": 0.0492, "step": 9630 }, { "epoch": 1.0441240242844754, "grad_norm": 0.791985809803009, "learning_rate": 1.3086464529232556e-05, "loss": 0.0173, "step": 9631 }, { "epoch": 1.0442324371205551, "grad_norm": 1.5353477001190186, "learning_rate": 1.308573915566517e-05, "loss": 0.0175, "step": 9632 }, { "epoch": 1.044340849956635, "grad_norm": 0.35027000308036804, "learning_rate": 1.3085013782097782e-05, "loss": 0.0075, "step": 9633 }, { "epoch": 1.0444492627927147, "grad_norm": 2.091930627822876, "learning_rate": 1.3084288408530394e-05, "loss": 0.0476, "step": 9634 }, { "epoch": 1.0445576756287944, "grad_norm": 5.130664348602295, "learning_rate": 1.3083563034963008e-05, "loss": 0.06, "step": 9635 }, { "epoch": 1.0446660884648742, "grad_norm": 6.207822322845459, "learning_rate": 1.308283766139562e-05, "loss": 0.0674, "step": 9636 }, { "epoch": 1.044774501300954, "grad_norm": 5.548070907592773, "learning_rate": 1.3082112287828232e-05, "loss": 0.0405, "step": 9637 }, { "epoch": 1.044882914137034, "grad_norm": 0.6872706413269043, "learning_rate": 1.3081386914260846e-05, "loss": 0.0151, "step": 9638 }, { "epoch": 1.0449913269731137, "grad_norm": 1.9625935554504395, "learning_rate": 1.3080661540693458e-05, "loss": 0.0296, "step": 9639 }, { "epoch": 1.0450997398091935, "grad_norm": 2.5404484272003174, "learning_rate": 1.307993616712607e-05, "loss": 0.0214, "step": 9640 }, { "epoch": 1.0452081526452732, "grad_norm": 0.6463337540626526, "learning_rate": 1.3079210793558682e-05, "loss": 0.0028, "step": 9641 }, { "epoch": 1.045316565481353, "grad_norm": 1.077194333076477, "learning_rate": 1.3078485419991296e-05, "loss": 0.0218, "step": 9642 }, { "epoch": 1.0454249783174328, "grad_norm": 2.377124786376953, "learning_rate": 1.3077760046423908e-05, "loss": 0.0361, "step": 9643 }, { "epoch": 1.0455333911535125, "grad_norm": 0.782995343208313, "learning_rate": 1.307703467285652e-05, "loss": 0.0118, "step": 9644 }, { "epoch": 1.0456418039895923, "grad_norm": 4.28220272064209, "learning_rate": 1.3076309299289136e-05, "loss": 0.0227, "step": 9645 }, { "epoch": 1.0457502168256723, "grad_norm": 2.1624579429626465, "learning_rate": 1.3075583925721748e-05, "loss": 0.039, "step": 9646 }, { "epoch": 1.045858629661752, "grad_norm": 1.6227492094039917, "learning_rate": 1.3074858552154362e-05, "loss": 0.0058, "step": 9647 }, { "epoch": 1.0459670424978318, "grad_norm": 1.0255464315414429, "learning_rate": 1.3074133178586974e-05, "loss": 0.0278, "step": 9648 }, { "epoch": 1.0460754553339116, "grad_norm": 2.8231236934661865, "learning_rate": 1.3073407805019586e-05, "loss": 0.0064, "step": 9649 }, { "epoch": 1.0461838681699913, "grad_norm": 13.069660186767578, "learning_rate": 1.30726824314522e-05, "loss": 0.0748, "step": 9650 }, { "epoch": 1.046292281006071, "grad_norm": 6.63764762878418, "learning_rate": 1.3071957057884812e-05, "loss": 0.0562, "step": 9651 }, { "epoch": 1.0464006938421508, "grad_norm": 0.4214865267276764, "learning_rate": 1.3071231684317424e-05, "loss": 0.0042, "step": 9652 }, { "epoch": 1.0465091066782306, "grad_norm": 1.390793800354004, "learning_rate": 1.3070506310750038e-05, "loss": 0.0153, "step": 9653 }, { "epoch": 1.0466175195143106, "grad_norm": 7.380241394042969, "learning_rate": 1.306978093718265e-05, "loss": 0.0253, "step": 9654 }, { "epoch": 1.0467259323503904, "grad_norm": 2.3108878135681152, "learning_rate": 1.3069055563615263e-05, "loss": 0.0406, "step": 9655 }, { "epoch": 1.0468343451864701, "grad_norm": 2.085853338241577, "learning_rate": 1.3068330190047875e-05, "loss": 0.0433, "step": 9656 }, { "epoch": 1.0469427580225499, "grad_norm": 3.056539297103882, "learning_rate": 1.3067604816480488e-05, "loss": 0.0417, "step": 9657 }, { "epoch": 1.0470511708586296, "grad_norm": 2.3756041526794434, "learning_rate": 1.30668794429131e-05, "loss": 0.0305, "step": 9658 }, { "epoch": 1.0471595836947094, "grad_norm": 0.8882442712783813, "learning_rate": 1.3066154069345713e-05, "loss": 0.0091, "step": 9659 }, { "epoch": 1.0472679965307892, "grad_norm": 0.09951068460941315, "learning_rate": 1.3065428695778327e-05, "loss": 0.0008, "step": 9660 }, { "epoch": 1.047376409366869, "grad_norm": 0.9020751118659973, "learning_rate": 1.3064703322210939e-05, "loss": 0.0225, "step": 9661 }, { "epoch": 1.047484822202949, "grad_norm": 3.2771897315979004, "learning_rate": 1.3063977948643551e-05, "loss": 0.0522, "step": 9662 }, { "epoch": 1.0475932350390287, "grad_norm": 0.06536763906478882, "learning_rate": 1.3063252575076166e-05, "loss": 0.0008, "step": 9663 }, { "epoch": 1.0477016478751084, "grad_norm": 0.5218431353569031, "learning_rate": 1.3062527201508779e-05, "loss": 0.0102, "step": 9664 }, { "epoch": 1.0478100607111882, "grad_norm": 3.4443037509918213, "learning_rate": 1.3061801827941392e-05, "loss": 0.0506, "step": 9665 }, { "epoch": 1.047918473547268, "grad_norm": 3.4271838665008545, "learning_rate": 1.3061076454374005e-05, "loss": 0.012, "step": 9666 }, { "epoch": 1.0480268863833477, "grad_norm": 1.4979788064956665, "learning_rate": 1.3060351080806617e-05, "loss": 0.0102, "step": 9667 }, { "epoch": 1.0481352992194275, "grad_norm": 0.718792200088501, "learning_rate": 1.3059625707239229e-05, "loss": 0.0102, "step": 9668 }, { "epoch": 1.0482437120555073, "grad_norm": 2.08526349067688, "learning_rate": 1.3058900333671843e-05, "loss": 0.0082, "step": 9669 }, { "epoch": 1.0483521248915872, "grad_norm": 1.2374794483184814, "learning_rate": 1.3058174960104455e-05, "loss": 0.0109, "step": 9670 }, { "epoch": 1.048460537727667, "grad_norm": 0.5054830312728882, "learning_rate": 1.3057449586537067e-05, "loss": 0.0254, "step": 9671 }, { "epoch": 1.0485689505637468, "grad_norm": 5.897187232971191, "learning_rate": 1.305672421296968e-05, "loss": 0.0294, "step": 9672 }, { "epoch": 1.0486773633998265, "grad_norm": 0.5735303163528442, "learning_rate": 1.3055998839402293e-05, "loss": 0.0046, "step": 9673 }, { "epoch": 1.0487857762359063, "grad_norm": 0.6810314655303955, "learning_rate": 1.3055273465834905e-05, "loss": 0.0036, "step": 9674 }, { "epoch": 1.048894189071986, "grad_norm": 7.318222522735596, "learning_rate": 1.3054548092267519e-05, "loss": 0.053, "step": 9675 }, { "epoch": 1.0490026019080658, "grad_norm": 5.272810459136963, "learning_rate": 1.3053822718700131e-05, "loss": 0.0424, "step": 9676 }, { "epoch": 1.0491110147441458, "grad_norm": 0.41762909293174744, "learning_rate": 1.3053097345132743e-05, "loss": 0.0106, "step": 9677 }, { "epoch": 1.0492194275802256, "grad_norm": 0.29189106822013855, "learning_rate": 1.3052371971565357e-05, "loss": 0.0065, "step": 9678 }, { "epoch": 1.0493278404163053, "grad_norm": 3.2063674926757812, "learning_rate": 1.305164659799797e-05, "loss": 0.0172, "step": 9679 }, { "epoch": 1.049436253252385, "grad_norm": 3.1609556674957275, "learning_rate": 1.3050921224430583e-05, "loss": 0.0239, "step": 9680 }, { "epoch": 1.0495446660884649, "grad_norm": 17.42000961303711, "learning_rate": 1.3050195850863197e-05, "loss": 0.0293, "step": 9681 }, { "epoch": 1.0496530789245446, "grad_norm": 2.681676149368286, "learning_rate": 1.3049470477295809e-05, "loss": 0.0291, "step": 9682 }, { "epoch": 1.0497614917606244, "grad_norm": 1.7649190425872803, "learning_rate": 1.3048745103728421e-05, "loss": 0.0024, "step": 9683 }, { "epoch": 1.0498699045967042, "grad_norm": 6.404138088226318, "learning_rate": 1.3048019730161035e-05, "loss": 0.0691, "step": 9684 }, { "epoch": 1.0499783174327841, "grad_norm": 0.8139180541038513, "learning_rate": 1.3047294356593647e-05, "loss": 0.0068, "step": 9685 }, { "epoch": 1.050086730268864, "grad_norm": 0.6035231947898865, "learning_rate": 1.304656898302626e-05, "loss": 0.0223, "step": 9686 }, { "epoch": 1.0501951431049437, "grad_norm": 4.04175329208374, "learning_rate": 1.3045843609458873e-05, "loss": 0.0454, "step": 9687 }, { "epoch": 1.0503035559410234, "grad_norm": 1.52590012550354, "learning_rate": 1.3045118235891485e-05, "loss": 0.0073, "step": 9688 }, { "epoch": 1.0504119687771032, "grad_norm": 5.639049530029297, "learning_rate": 1.3044392862324097e-05, "loss": 0.024, "step": 9689 }, { "epoch": 1.050520381613183, "grad_norm": 2.6591591835021973, "learning_rate": 1.3043667488756711e-05, "loss": 0.0263, "step": 9690 }, { "epoch": 1.0506287944492627, "grad_norm": 0.238646999001503, "learning_rate": 1.3042942115189323e-05, "loss": 0.0011, "step": 9691 }, { "epoch": 1.0507372072853425, "grad_norm": 3.5002284049987793, "learning_rate": 1.3042216741621936e-05, "loss": 0.033, "step": 9692 }, { "epoch": 1.0508456201214225, "grad_norm": 0.775521993637085, "learning_rate": 1.304149136805455e-05, "loss": 0.009, "step": 9693 }, { "epoch": 1.0509540329575022, "grad_norm": 0.20200732350349426, "learning_rate": 1.3040765994487162e-05, "loss": 0.0013, "step": 9694 }, { "epoch": 1.051062445793582, "grad_norm": 0.6138750314712524, "learning_rate": 1.3040040620919774e-05, "loss": 0.0123, "step": 9695 }, { "epoch": 1.0511708586296618, "grad_norm": 2.9956748485565186, "learning_rate": 1.3039315247352386e-05, "loss": 0.0218, "step": 9696 }, { "epoch": 1.0512792714657415, "grad_norm": 3.0982062816619873, "learning_rate": 1.3038589873785001e-05, "loss": 0.0186, "step": 9697 }, { "epoch": 1.0513876843018213, "grad_norm": 4.19765043258667, "learning_rate": 1.3037864500217613e-05, "loss": 0.0567, "step": 9698 }, { "epoch": 1.051496097137901, "grad_norm": 1.5428694486618042, "learning_rate": 1.3037139126650227e-05, "loss": 0.0088, "step": 9699 }, { "epoch": 1.051604509973981, "grad_norm": 0.7998387813568115, "learning_rate": 1.303641375308284e-05, "loss": 0.0253, "step": 9700 }, { "epoch": 1.0517129228100608, "grad_norm": 0.6606804728507996, "learning_rate": 1.3035688379515452e-05, "loss": 0.0022, "step": 9701 }, { "epoch": 1.0518213356461406, "grad_norm": 0.962634801864624, "learning_rate": 1.3034963005948065e-05, "loss": 0.0115, "step": 9702 }, { "epoch": 1.0519297484822203, "grad_norm": 2.214792013168335, "learning_rate": 1.3034237632380678e-05, "loss": 0.0169, "step": 9703 }, { "epoch": 1.0520381613183, "grad_norm": 4.574423789978027, "learning_rate": 1.303351225881329e-05, "loss": 0.0384, "step": 9704 }, { "epoch": 1.0521465741543798, "grad_norm": 0.242960587143898, "learning_rate": 1.3032786885245904e-05, "loss": 0.001, "step": 9705 }, { "epoch": 1.0522549869904596, "grad_norm": 2.4867055416107178, "learning_rate": 1.3032061511678516e-05, "loss": 0.0447, "step": 9706 }, { "epoch": 1.0523633998265394, "grad_norm": 1.9587417840957642, "learning_rate": 1.3031336138111128e-05, "loss": 0.0226, "step": 9707 }, { "epoch": 1.0524718126626194, "grad_norm": 0.024133428931236267, "learning_rate": 1.303061076454374e-05, "loss": 0.0002, "step": 9708 }, { "epoch": 1.0525802254986991, "grad_norm": 0.35381606221199036, "learning_rate": 1.3029885390976354e-05, "loss": 0.0116, "step": 9709 }, { "epoch": 1.0526886383347789, "grad_norm": 5.342977523803711, "learning_rate": 1.3029160017408966e-05, "loss": 0.083, "step": 9710 }, { "epoch": 1.0527970511708586, "grad_norm": 3.452035427093506, "learning_rate": 1.3028434643841578e-05, "loss": 0.0352, "step": 9711 }, { "epoch": 1.0529054640069384, "grad_norm": 3.548915147781372, "learning_rate": 1.3027709270274192e-05, "loss": 0.074, "step": 9712 }, { "epoch": 1.0530138768430182, "grad_norm": 6.96862268447876, "learning_rate": 1.3026983896706804e-05, "loss": 0.015, "step": 9713 }, { "epoch": 1.053122289679098, "grad_norm": 0.18433010578155518, "learning_rate": 1.302625852313942e-05, "loss": 0.0004, "step": 9714 }, { "epoch": 1.0532307025151777, "grad_norm": 3.6810836791992188, "learning_rate": 1.3025533149572032e-05, "loss": 0.0279, "step": 9715 }, { "epoch": 1.0533391153512577, "grad_norm": 10.13861083984375, "learning_rate": 1.3024807776004644e-05, "loss": 0.0254, "step": 9716 }, { "epoch": 1.0534475281873374, "grad_norm": 1.8279597759246826, "learning_rate": 1.3024082402437258e-05, "loss": 0.0122, "step": 9717 }, { "epoch": 1.0535559410234172, "grad_norm": 1.6306157112121582, "learning_rate": 1.302335702886987e-05, "loss": 0.0209, "step": 9718 }, { "epoch": 1.053664353859497, "grad_norm": 1.4688185453414917, "learning_rate": 1.3022631655302482e-05, "loss": 0.0346, "step": 9719 }, { "epoch": 1.0537727666955767, "grad_norm": 6.603394031524658, "learning_rate": 1.3021906281735094e-05, "loss": 0.0564, "step": 9720 }, { "epoch": 1.0538811795316565, "grad_norm": 0.14857834577560425, "learning_rate": 1.3021180908167708e-05, "loss": 0.0013, "step": 9721 }, { "epoch": 1.0539895923677363, "grad_norm": 1.4376970529556274, "learning_rate": 1.302045553460032e-05, "loss": 0.0315, "step": 9722 }, { "epoch": 1.0540980052038162, "grad_norm": 1.6940279006958008, "learning_rate": 1.3019730161032932e-05, "loss": 0.0261, "step": 9723 }, { "epoch": 1.054206418039896, "grad_norm": 1.796778917312622, "learning_rate": 1.3019004787465546e-05, "loss": 0.0171, "step": 9724 }, { "epoch": 1.0543148308759758, "grad_norm": 4.468644618988037, "learning_rate": 1.3018279413898158e-05, "loss": 0.027, "step": 9725 }, { "epoch": 1.0544232437120555, "grad_norm": 1.574357509613037, "learning_rate": 1.301755404033077e-05, "loss": 0.019, "step": 9726 }, { "epoch": 1.0545316565481353, "grad_norm": 3.922041177749634, "learning_rate": 1.3016828666763384e-05, "loss": 0.0319, "step": 9727 }, { "epoch": 1.054640069384215, "grad_norm": 0.8122105598449707, "learning_rate": 1.3016103293195996e-05, "loss": 0.0028, "step": 9728 }, { "epoch": 1.0547484822202948, "grad_norm": 8.10763168334961, "learning_rate": 1.3015377919628609e-05, "loss": 0.0569, "step": 9729 }, { "epoch": 1.0548568950563746, "grad_norm": 0.7490761876106262, "learning_rate": 1.3014652546061222e-05, "loss": 0.0116, "step": 9730 }, { "epoch": 1.0549653078924546, "grad_norm": 2.03739595413208, "learning_rate": 1.3013927172493835e-05, "loss": 0.0177, "step": 9731 }, { "epoch": 1.0550737207285343, "grad_norm": 0.4052888751029968, "learning_rate": 1.3013201798926448e-05, "loss": 0.0133, "step": 9732 }, { "epoch": 1.055182133564614, "grad_norm": 1.2614654302597046, "learning_rate": 1.3012476425359062e-05, "loss": 0.0182, "step": 9733 }, { "epoch": 1.0552905464006939, "grad_norm": 2.395493745803833, "learning_rate": 1.3011751051791674e-05, "loss": 0.022, "step": 9734 }, { "epoch": 1.0553989592367736, "grad_norm": 3.6563007831573486, "learning_rate": 1.3011025678224286e-05, "loss": 0.0206, "step": 9735 }, { "epoch": 1.0555073720728534, "grad_norm": 2.4304099082946777, "learning_rate": 1.30103003046569e-05, "loss": 0.01, "step": 9736 }, { "epoch": 1.0556157849089332, "grad_norm": 0.35879117250442505, "learning_rate": 1.3009574931089512e-05, "loss": 0.0105, "step": 9737 }, { "epoch": 1.055724197745013, "grad_norm": 4.653168201446533, "learning_rate": 1.3008849557522125e-05, "loss": 0.0275, "step": 9738 }, { "epoch": 1.055832610581093, "grad_norm": 1.232223391532898, "learning_rate": 1.3008124183954738e-05, "loss": 0.0178, "step": 9739 }, { "epoch": 1.0559410234171727, "grad_norm": 0.2218286693096161, "learning_rate": 1.300739881038735e-05, "loss": 0.0044, "step": 9740 }, { "epoch": 1.0560494362532524, "grad_norm": 2.023500442504883, "learning_rate": 1.3006673436819963e-05, "loss": 0.0496, "step": 9741 }, { "epoch": 1.0561578490893322, "grad_norm": 3.479125738143921, "learning_rate": 1.3005948063252577e-05, "loss": 0.0255, "step": 9742 }, { "epoch": 1.056266261925412, "grad_norm": 5.330687046051025, "learning_rate": 1.3005222689685189e-05, "loss": 0.0454, "step": 9743 }, { "epoch": 1.0563746747614917, "grad_norm": 4.565203666687012, "learning_rate": 1.3004497316117801e-05, "loss": 0.0235, "step": 9744 }, { "epoch": 1.0564830875975715, "grad_norm": 1.8946781158447266, "learning_rate": 1.3003771942550415e-05, "loss": 0.0272, "step": 9745 }, { "epoch": 1.0565915004336512, "grad_norm": 5.047449111938477, "learning_rate": 1.3003046568983027e-05, "loss": 0.0282, "step": 9746 }, { "epoch": 1.0566999132697312, "grad_norm": 0.8607661128044128, "learning_rate": 1.3002321195415639e-05, "loss": 0.0268, "step": 9747 }, { "epoch": 1.056808326105811, "grad_norm": 3.923684597015381, "learning_rate": 1.3001595821848251e-05, "loss": 0.041, "step": 9748 }, { "epoch": 1.0569167389418908, "grad_norm": 0.4166490137577057, "learning_rate": 1.3000870448280867e-05, "loss": 0.0022, "step": 9749 }, { "epoch": 1.0570251517779705, "grad_norm": 1.8405981063842773, "learning_rate": 1.3000145074713479e-05, "loss": 0.018, "step": 9750 }, { "epoch": 1.0571335646140503, "grad_norm": 0.4882224500179291, "learning_rate": 1.2999419701146093e-05, "loss": 0.0157, "step": 9751 }, { "epoch": 1.05724197745013, "grad_norm": 5.40364933013916, "learning_rate": 1.2998694327578705e-05, "loss": 0.0168, "step": 9752 }, { "epoch": 1.0573503902862098, "grad_norm": 2.723231792449951, "learning_rate": 1.2997968954011317e-05, "loss": 0.013, "step": 9753 }, { "epoch": 1.0574588031222896, "grad_norm": 3.828049898147583, "learning_rate": 1.299724358044393e-05, "loss": 0.0185, "step": 9754 }, { "epoch": 1.0575672159583696, "grad_norm": 7.155186176300049, "learning_rate": 1.2996518206876543e-05, "loss": 0.02, "step": 9755 }, { "epoch": 1.0576756287944493, "grad_norm": 6.342182159423828, "learning_rate": 1.2995792833309155e-05, "loss": 0.0228, "step": 9756 }, { "epoch": 1.057784041630529, "grad_norm": 1.7421985864639282, "learning_rate": 1.2995067459741769e-05, "loss": 0.023, "step": 9757 }, { "epoch": 1.0578924544666088, "grad_norm": 1.160927176475525, "learning_rate": 1.2994342086174381e-05, "loss": 0.0121, "step": 9758 }, { "epoch": 1.0580008673026886, "grad_norm": 1.4694057703018188, "learning_rate": 1.2993616712606993e-05, "loss": 0.0411, "step": 9759 }, { "epoch": 1.0581092801387684, "grad_norm": 2.226461172103882, "learning_rate": 1.2992891339039605e-05, "loss": 0.0166, "step": 9760 }, { "epoch": 1.0582176929748481, "grad_norm": 0.8196007013320923, "learning_rate": 1.299216596547222e-05, "loss": 0.0048, "step": 9761 }, { "epoch": 1.0583261058109281, "grad_norm": 3.351454973220825, "learning_rate": 1.2991440591904831e-05, "loss": 0.0524, "step": 9762 }, { "epoch": 1.0584345186470079, "grad_norm": 2.5085740089416504, "learning_rate": 1.2990715218337443e-05, "loss": 0.0107, "step": 9763 }, { "epoch": 1.0585429314830876, "grad_norm": 0.3157958984375, "learning_rate": 1.2989989844770057e-05, "loss": 0.0019, "step": 9764 }, { "epoch": 1.0586513443191674, "grad_norm": 2.059983253479004, "learning_rate": 1.298926447120267e-05, "loss": 0.0404, "step": 9765 }, { "epoch": 1.0587597571552472, "grad_norm": 3.3989667892456055, "learning_rate": 1.2988539097635285e-05, "loss": 0.022, "step": 9766 }, { "epoch": 1.058868169991327, "grad_norm": 0.24224641919136047, "learning_rate": 1.2987813724067897e-05, "loss": 0.0016, "step": 9767 }, { "epoch": 1.0589765828274067, "grad_norm": 0.5452288389205933, "learning_rate": 1.298708835050051e-05, "loss": 0.0056, "step": 9768 }, { "epoch": 1.0590849956634865, "grad_norm": 1.918105125427246, "learning_rate": 1.2986362976933123e-05, "loss": 0.0468, "step": 9769 }, { "epoch": 1.0591934084995664, "grad_norm": 1.020900011062622, "learning_rate": 1.2985637603365735e-05, "loss": 0.0074, "step": 9770 }, { "epoch": 1.0593018213356462, "grad_norm": 7.351837635040283, "learning_rate": 1.2984912229798347e-05, "loss": 0.0335, "step": 9771 }, { "epoch": 1.059410234171726, "grad_norm": 0.3882102370262146, "learning_rate": 1.298418685623096e-05, "loss": 0.0094, "step": 9772 }, { "epoch": 1.0595186470078057, "grad_norm": 13.619375228881836, "learning_rate": 1.2983461482663573e-05, "loss": 0.0474, "step": 9773 }, { "epoch": 1.0596270598438855, "grad_norm": 2.5054850578308105, "learning_rate": 1.2982736109096185e-05, "loss": 0.012, "step": 9774 }, { "epoch": 1.0597354726799653, "grad_norm": 5.142969608306885, "learning_rate": 1.2982010735528798e-05, "loss": 0.0356, "step": 9775 }, { "epoch": 1.059843885516045, "grad_norm": 8.802706718444824, "learning_rate": 1.2981285361961411e-05, "loss": 0.052, "step": 9776 }, { "epoch": 1.0599522983521248, "grad_norm": 2.4116334915161133, "learning_rate": 1.2980559988394024e-05, "loss": 0.0506, "step": 9777 }, { "epoch": 1.0600607111882048, "grad_norm": 3.396712064743042, "learning_rate": 1.2979834614826636e-05, "loss": 0.0268, "step": 9778 }, { "epoch": 1.0601691240242845, "grad_norm": 0.23273764550685883, "learning_rate": 1.297910924125925e-05, "loss": 0.0043, "step": 9779 }, { "epoch": 1.0602775368603643, "grad_norm": 3.5450470447540283, "learning_rate": 1.2978383867691862e-05, "loss": 0.015, "step": 9780 }, { "epoch": 1.060385949696444, "grad_norm": 0.4446010887622833, "learning_rate": 1.2977658494124474e-05, "loss": 0.0121, "step": 9781 }, { "epoch": 1.0604943625325238, "grad_norm": 4.620891571044922, "learning_rate": 1.2976933120557088e-05, "loss": 0.0237, "step": 9782 }, { "epoch": 1.0606027753686036, "grad_norm": 1.730422854423523, "learning_rate": 1.29762077469897e-05, "loss": 0.0248, "step": 9783 }, { "epoch": 1.0607111882046834, "grad_norm": 0.4458130896091461, "learning_rate": 1.2975482373422314e-05, "loss": 0.0111, "step": 9784 }, { "epoch": 1.0608196010407633, "grad_norm": 0.412193238735199, "learning_rate": 1.2974756999854928e-05, "loss": 0.0231, "step": 9785 }, { "epoch": 1.060928013876843, "grad_norm": 1.1842131614685059, "learning_rate": 1.297403162628754e-05, "loss": 0.0171, "step": 9786 }, { "epoch": 1.0610364267129229, "grad_norm": 2.0396578311920166, "learning_rate": 1.2973306252720152e-05, "loss": 0.0062, "step": 9787 }, { "epoch": 1.0611448395490026, "grad_norm": 1.2340376377105713, "learning_rate": 1.2972580879152766e-05, "loss": 0.0069, "step": 9788 }, { "epoch": 1.0612532523850824, "grad_norm": 0.28754106163978577, "learning_rate": 1.2971855505585378e-05, "loss": 0.0047, "step": 9789 }, { "epoch": 1.0613616652211622, "grad_norm": 1.699949026107788, "learning_rate": 1.297113013201799e-05, "loss": 0.0177, "step": 9790 }, { "epoch": 1.061470078057242, "grad_norm": 1.4993401765823364, "learning_rate": 1.2970404758450604e-05, "loss": 0.009, "step": 9791 }, { "epoch": 1.0615784908933217, "grad_norm": 3.1304571628570557, "learning_rate": 1.2969679384883216e-05, "loss": 0.0265, "step": 9792 }, { "epoch": 1.0616869037294017, "grad_norm": 0.8349982500076294, "learning_rate": 1.2968954011315828e-05, "loss": 0.0059, "step": 9793 }, { "epoch": 1.0617953165654814, "grad_norm": 6.838306903839111, "learning_rate": 1.2968228637748442e-05, "loss": 0.0845, "step": 9794 }, { "epoch": 1.0619037294015612, "grad_norm": 0.6538352370262146, "learning_rate": 1.2967503264181054e-05, "loss": 0.0227, "step": 9795 }, { "epoch": 1.062012142237641, "grad_norm": 1.4347741603851318, "learning_rate": 1.2966777890613666e-05, "loss": 0.0422, "step": 9796 }, { "epoch": 1.0621205550737207, "grad_norm": 0.3710731863975525, "learning_rate": 1.296605251704628e-05, "loss": 0.0021, "step": 9797 }, { "epoch": 1.0622289679098005, "grad_norm": 1.1108065843582153, "learning_rate": 1.2965327143478892e-05, "loss": 0.0055, "step": 9798 }, { "epoch": 1.0623373807458802, "grad_norm": 1.3924403190612793, "learning_rate": 1.2964601769911504e-05, "loss": 0.0209, "step": 9799 }, { "epoch": 1.06244579358196, "grad_norm": 0.48631349205970764, "learning_rate": 1.2963876396344116e-05, "loss": 0.0136, "step": 9800 }, { "epoch": 1.06255420641804, "grad_norm": 1.791516900062561, "learning_rate": 1.2963151022776732e-05, "loss": 0.0166, "step": 9801 }, { "epoch": 1.0626626192541198, "grad_norm": 1.8029626607894897, "learning_rate": 1.2962425649209344e-05, "loss": 0.0165, "step": 9802 }, { "epoch": 1.0627710320901995, "grad_norm": 1.106937050819397, "learning_rate": 1.2961700275641958e-05, "loss": 0.0144, "step": 9803 }, { "epoch": 1.0628794449262793, "grad_norm": 0.3362376391887665, "learning_rate": 1.296097490207457e-05, "loss": 0.0113, "step": 9804 }, { "epoch": 1.062987857762359, "grad_norm": 1.929789662361145, "learning_rate": 1.2960249528507182e-05, "loss": 0.0313, "step": 9805 }, { "epoch": 1.0630962705984388, "grad_norm": 1.2884215116500854, "learning_rate": 1.2959524154939796e-05, "loss": 0.0227, "step": 9806 }, { "epoch": 1.0632046834345186, "grad_norm": 1.4326391220092773, "learning_rate": 1.2958798781372408e-05, "loss": 0.0929, "step": 9807 }, { "epoch": 1.0633130962705986, "grad_norm": 2.992779493331909, "learning_rate": 1.295807340780502e-05, "loss": 0.0185, "step": 9808 }, { "epoch": 1.0634215091066783, "grad_norm": 0.12773382663726807, "learning_rate": 1.2957348034237634e-05, "loss": 0.0009, "step": 9809 }, { "epoch": 1.063529921942758, "grad_norm": 4.610313892364502, "learning_rate": 1.2956622660670246e-05, "loss": 0.0077, "step": 9810 }, { "epoch": 1.0636383347788378, "grad_norm": 2.084372043609619, "learning_rate": 1.2955897287102859e-05, "loss": 0.0319, "step": 9811 }, { "epoch": 1.0637467476149176, "grad_norm": 2.299898624420166, "learning_rate": 1.295517191353547e-05, "loss": 0.0245, "step": 9812 }, { "epoch": 1.0638551604509974, "grad_norm": 3.780410051345825, "learning_rate": 1.2954446539968084e-05, "loss": 0.061, "step": 9813 }, { "epoch": 1.0639635732870771, "grad_norm": 0.5381193161010742, "learning_rate": 1.2953721166400697e-05, "loss": 0.0293, "step": 9814 }, { "epoch": 1.064071986123157, "grad_norm": 0.343709260225296, "learning_rate": 1.2952995792833309e-05, "loss": 0.0076, "step": 9815 }, { "epoch": 1.0641803989592367, "grad_norm": 1.6821426153182983, "learning_rate": 1.2952270419265923e-05, "loss": 0.0415, "step": 9816 }, { "epoch": 1.0642888117953166, "grad_norm": 0.5648136138916016, "learning_rate": 1.2951545045698535e-05, "loss": 0.0141, "step": 9817 }, { "epoch": 1.0643972246313964, "grad_norm": 0.17528392374515533, "learning_rate": 1.295081967213115e-05, "loss": 0.001, "step": 9818 }, { "epoch": 1.0645056374674762, "grad_norm": 1.5792356729507446, "learning_rate": 1.2950094298563762e-05, "loss": 0.0274, "step": 9819 }, { "epoch": 1.064614050303556, "grad_norm": 3.449887275695801, "learning_rate": 1.2949368924996375e-05, "loss": 0.0378, "step": 9820 }, { "epoch": 1.0647224631396357, "grad_norm": 8.257076263427734, "learning_rate": 1.2948643551428988e-05, "loss": 0.0146, "step": 9821 }, { "epoch": 1.0648308759757155, "grad_norm": 4.958632469177246, "learning_rate": 1.29479181778616e-05, "loss": 0.0471, "step": 9822 }, { "epoch": 1.0649392888117952, "grad_norm": 2.9282822608947754, "learning_rate": 1.2947192804294213e-05, "loss": 0.0173, "step": 9823 }, { "epoch": 1.0650477016478752, "grad_norm": 4.268789768218994, "learning_rate": 1.2946467430726825e-05, "loss": 0.0444, "step": 9824 }, { "epoch": 1.065156114483955, "grad_norm": 1.3057026863098145, "learning_rate": 1.2945742057159439e-05, "loss": 0.0161, "step": 9825 }, { "epoch": 1.0652645273200347, "grad_norm": 1.235019564628601, "learning_rate": 1.294501668359205e-05, "loss": 0.0176, "step": 9826 }, { "epoch": 1.0653729401561145, "grad_norm": 2.4051504135131836, "learning_rate": 1.2944291310024663e-05, "loss": 0.0169, "step": 9827 }, { "epoch": 1.0654813529921943, "grad_norm": 1.307982087135315, "learning_rate": 1.2943565936457277e-05, "loss": 0.0037, "step": 9828 }, { "epoch": 1.065589765828274, "grad_norm": 0.4607962965965271, "learning_rate": 1.2942840562889889e-05, "loss": 0.016, "step": 9829 }, { "epoch": 1.0656981786643538, "grad_norm": 0.8354153037071228, "learning_rate": 1.2942115189322501e-05, "loss": 0.0012, "step": 9830 }, { "epoch": 1.0658065915004336, "grad_norm": 2.291322708129883, "learning_rate": 1.2941389815755115e-05, "loss": 0.0218, "step": 9831 }, { "epoch": 1.0659150043365135, "grad_norm": 0.4181390404701233, "learning_rate": 1.2940664442187727e-05, "loss": 0.0049, "step": 9832 }, { "epoch": 1.0660234171725933, "grad_norm": 0.14602603018283844, "learning_rate": 1.293993906862034e-05, "loss": 0.001, "step": 9833 }, { "epoch": 1.066131830008673, "grad_norm": 0.06804364174604416, "learning_rate": 1.2939213695052953e-05, "loss": 0.0008, "step": 9834 }, { "epoch": 1.0662402428447528, "grad_norm": 2.3896772861480713, "learning_rate": 1.2938488321485567e-05, "loss": 0.0595, "step": 9835 }, { "epoch": 1.0663486556808326, "grad_norm": 13.193070411682129, "learning_rate": 1.2937762947918179e-05, "loss": 0.0559, "step": 9836 }, { "epoch": 1.0664570685169124, "grad_norm": 0.4356280565261841, "learning_rate": 1.2937037574350793e-05, "loss": 0.0076, "step": 9837 }, { "epoch": 1.0665654813529921, "grad_norm": 0.48206743597984314, "learning_rate": 1.2936312200783405e-05, "loss": 0.0186, "step": 9838 }, { "epoch": 1.0666738941890719, "grad_norm": 2.8451781272888184, "learning_rate": 1.2935586827216017e-05, "loss": 0.0265, "step": 9839 }, { "epoch": 1.0667823070251519, "grad_norm": 0.6541950106620789, "learning_rate": 1.2934861453648631e-05, "loss": 0.0072, "step": 9840 }, { "epoch": 1.0668907198612316, "grad_norm": 2.579160451889038, "learning_rate": 1.2934136080081243e-05, "loss": 0.0338, "step": 9841 }, { "epoch": 1.0669991326973114, "grad_norm": 4.396290302276611, "learning_rate": 1.2933410706513855e-05, "loss": 0.0198, "step": 9842 }, { "epoch": 1.0671075455333912, "grad_norm": 0.05630582943558693, "learning_rate": 1.2932685332946469e-05, "loss": 0.0005, "step": 9843 }, { "epoch": 1.067215958369471, "grad_norm": 5.791938304901123, "learning_rate": 1.2931959959379081e-05, "loss": 0.0554, "step": 9844 }, { "epoch": 1.0673243712055507, "grad_norm": 7.495187759399414, "learning_rate": 1.2931234585811693e-05, "loss": 0.0446, "step": 9845 }, { "epoch": 1.0674327840416304, "grad_norm": 3.8185157775878906, "learning_rate": 1.2930509212244307e-05, "loss": 0.0032, "step": 9846 }, { "epoch": 1.0675411968777104, "grad_norm": 0.8245841264724731, "learning_rate": 1.292978383867692e-05, "loss": 0.019, "step": 9847 }, { "epoch": 1.0676496097137902, "grad_norm": 5.220889568328857, "learning_rate": 1.2929058465109532e-05, "loss": 0.0226, "step": 9848 }, { "epoch": 1.06775802254987, "grad_norm": 7.495218753814697, "learning_rate": 1.2928333091542145e-05, "loss": 0.0355, "step": 9849 }, { "epoch": 1.0678664353859497, "grad_norm": 3.688051223754883, "learning_rate": 1.2927607717974758e-05, "loss": 0.0273, "step": 9850 }, { "epoch": 1.0679748482220295, "grad_norm": 4.789533615112305, "learning_rate": 1.292688234440737e-05, "loss": 0.0372, "step": 9851 }, { "epoch": 1.0680832610581092, "grad_norm": 3.417404890060425, "learning_rate": 1.2926156970839982e-05, "loss": 0.0342, "step": 9852 }, { "epoch": 1.068191673894189, "grad_norm": 1.362295150756836, "learning_rate": 1.2925431597272597e-05, "loss": 0.0121, "step": 9853 }, { "epoch": 1.0683000867302688, "grad_norm": 15.600964546203613, "learning_rate": 1.292470622370521e-05, "loss": 0.0209, "step": 9854 }, { "epoch": 1.0684084995663488, "grad_norm": 2.215575933456421, "learning_rate": 1.2923980850137823e-05, "loss": 0.0169, "step": 9855 }, { "epoch": 1.0685169124024285, "grad_norm": 1.8399637937545776, "learning_rate": 1.2923255476570435e-05, "loss": 0.0204, "step": 9856 }, { "epoch": 1.0686253252385083, "grad_norm": 0.4809853434562683, "learning_rate": 1.2922530103003048e-05, "loss": 0.0127, "step": 9857 }, { "epoch": 1.068733738074588, "grad_norm": 3.4508121013641357, "learning_rate": 1.2921804729435661e-05, "loss": 0.0491, "step": 9858 }, { "epoch": 1.0688421509106678, "grad_norm": 1.2078665494918823, "learning_rate": 1.2921079355868274e-05, "loss": 0.0084, "step": 9859 }, { "epoch": 1.0689505637467476, "grad_norm": 1.3256171941757202, "learning_rate": 1.2920353982300886e-05, "loss": 0.0022, "step": 9860 }, { "epoch": 1.0690589765828273, "grad_norm": 0.7049620747566223, "learning_rate": 1.29196286087335e-05, "loss": 0.015, "step": 9861 }, { "epoch": 1.069167389418907, "grad_norm": 0.6882070302963257, "learning_rate": 1.2918903235166112e-05, "loss": 0.0086, "step": 9862 }, { "epoch": 1.069275802254987, "grad_norm": 4.88173770904541, "learning_rate": 1.2918177861598724e-05, "loss": 0.0291, "step": 9863 }, { "epoch": 1.0693842150910668, "grad_norm": 0.6052717566490173, "learning_rate": 1.2917452488031336e-05, "loss": 0.0129, "step": 9864 }, { "epoch": 1.0694926279271466, "grad_norm": 0.6502029299736023, "learning_rate": 1.291672711446395e-05, "loss": 0.009, "step": 9865 }, { "epoch": 1.0696010407632264, "grad_norm": 0.3433762490749359, "learning_rate": 1.2916001740896562e-05, "loss": 0.0055, "step": 9866 }, { "epoch": 1.0697094535993061, "grad_norm": 2.247105836868286, "learning_rate": 1.2915276367329174e-05, "loss": 0.0183, "step": 9867 }, { "epoch": 1.069817866435386, "grad_norm": 4.304882526397705, "learning_rate": 1.2914550993761788e-05, "loss": 0.0441, "step": 9868 }, { "epoch": 1.0699262792714657, "grad_norm": 3.794795274734497, "learning_rate": 1.29138256201944e-05, "loss": 0.009, "step": 9869 }, { "epoch": 1.0700346921075456, "grad_norm": 0.30128082633018494, "learning_rate": 1.2913100246627016e-05, "loss": 0.0012, "step": 9870 }, { "epoch": 1.0701431049436254, "grad_norm": 6.489210605621338, "learning_rate": 1.2912374873059628e-05, "loss": 0.0515, "step": 9871 }, { "epoch": 1.0702515177797052, "grad_norm": 2.5669941902160645, "learning_rate": 1.291164949949224e-05, "loss": 0.0664, "step": 9872 }, { "epoch": 1.070359930615785, "grad_norm": 3.307081937789917, "learning_rate": 1.2910924125924854e-05, "loss": 0.0147, "step": 9873 }, { "epoch": 1.0704683434518647, "grad_norm": 4.73938512802124, "learning_rate": 1.2910198752357466e-05, "loss": 0.0209, "step": 9874 }, { "epoch": 1.0705767562879445, "grad_norm": 1.6454994678497314, "learning_rate": 1.2909473378790078e-05, "loss": 0.0292, "step": 9875 }, { "epoch": 1.0706851691240242, "grad_norm": 2.015333652496338, "learning_rate": 1.290874800522269e-05, "loss": 0.0251, "step": 9876 }, { "epoch": 1.070793581960104, "grad_norm": 4.119876861572266, "learning_rate": 1.2908022631655304e-05, "loss": 0.0144, "step": 9877 }, { "epoch": 1.0709019947961838, "grad_norm": 3.5581183433532715, "learning_rate": 1.2907297258087916e-05, "loss": 0.0151, "step": 9878 }, { "epoch": 1.0710104076322637, "grad_norm": 3.039170742034912, "learning_rate": 1.2906571884520528e-05, "loss": 0.0301, "step": 9879 }, { "epoch": 1.0711188204683435, "grad_norm": 1.6171725988388062, "learning_rate": 1.2905846510953142e-05, "loss": 0.0356, "step": 9880 }, { "epoch": 1.0712272333044233, "grad_norm": 3.6301891803741455, "learning_rate": 1.2905121137385754e-05, "loss": 0.0423, "step": 9881 }, { "epoch": 1.071335646140503, "grad_norm": 3.5146875381469727, "learning_rate": 1.2904395763818366e-05, "loss": 0.0291, "step": 9882 }, { "epoch": 1.0714440589765828, "grad_norm": 0.07337800413370132, "learning_rate": 1.290367039025098e-05, "loss": 0.0005, "step": 9883 }, { "epoch": 1.0715524718126626, "grad_norm": 1.5671337842941284, "learning_rate": 1.2902945016683592e-05, "loss": 0.0105, "step": 9884 }, { "epoch": 1.0716608846487423, "grad_norm": 2.2038793563842773, "learning_rate": 1.2902219643116205e-05, "loss": 0.0416, "step": 9885 }, { "epoch": 1.0717692974848223, "grad_norm": 2.409990072250366, "learning_rate": 1.2901494269548818e-05, "loss": 0.0447, "step": 9886 }, { "epoch": 1.071877710320902, "grad_norm": 7.031850337982178, "learning_rate": 1.2900768895981432e-05, "loss": 0.0158, "step": 9887 }, { "epoch": 1.0719861231569818, "grad_norm": 2.519615650177002, "learning_rate": 1.2900043522414046e-05, "loss": 0.0231, "step": 9888 }, { "epoch": 1.0720945359930616, "grad_norm": 1.804700493812561, "learning_rate": 1.2899318148846658e-05, "loss": 0.0556, "step": 9889 }, { "epoch": 1.0722029488291414, "grad_norm": 1.02056086063385, "learning_rate": 1.289859277527927e-05, "loss": 0.0163, "step": 9890 }, { "epoch": 1.0723113616652211, "grad_norm": 2.433009147644043, "learning_rate": 1.2897867401711883e-05, "loss": 0.0261, "step": 9891 }, { "epoch": 1.0724197745013009, "grad_norm": 0.4210943281650543, "learning_rate": 1.2897142028144496e-05, "loss": 0.0081, "step": 9892 }, { "epoch": 1.0725281873373806, "grad_norm": 0.3769594430923462, "learning_rate": 1.2896416654577108e-05, "loss": 0.0067, "step": 9893 }, { "epoch": 1.0726366001734606, "grad_norm": 1.590501070022583, "learning_rate": 1.289569128100972e-05, "loss": 0.0131, "step": 9894 }, { "epoch": 1.0727450130095404, "grad_norm": 5.475620269775391, "learning_rate": 1.2894965907442334e-05, "loss": 0.0464, "step": 9895 }, { "epoch": 1.0728534258456202, "grad_norm": 0.38677412271499634, "learning_rate": 1.2894240533874947e-05, "loss": 0.003, "step": 9896 }, { "epoch": 1.0729618386817, "grad_norm": 0.3736484944820404, "learning_rate": 1.2893515160307559e-05, "loss": 0.018, "step": 9897 }, { "epoch": 1.0730702515177797, "grad_norm": 0.9262869954109192, "learning_rate": 1.2892789786740173e-05, "loss": 0.0123, "step": 9898 }, { "epoch": 1.0731786643538594, "grad_norm": 2.5420446395874023, "learning_rate": 1.2892064413172785e-05, "loss": 0.0135, "step": 9899 }, { "epoch": 1.0732870771899392, "grad_norm": 0.6090350151062012, "learning_rate": 1.2891339039605397e-05, "loss": 0.0021, "step": 9900 }, { "epoch": 1.073395490026019, "grad_norm": 0.3730151653289795, "learning_rate": 1.289061366603801e-05, "loss": 0.0123, "step": 9901 }, { "epoch": 1.073503902862099, "grad_norm": 7.102900505065918, "learning_rate": 1.2889888292470623e-05, "loss": 0.0493, "step": 9902 }, { "epoch": 1.0736123156981787, "grad_norm": 2.141451120376587, "learning_rate": 1.2889162918903235e-05, "loss": 0.0233, "step": 9903 }, { "epoch": 1.0737207285342585, "grad_norm": 0.03809472545981407, "learning_rate": 1.288843754533585e-05, "loss": 0.0004, "step": 9904 }, { "epoch": 1.0738291413703382, "grad_norm": 4.618420600891113, "learning_rate": 1.2887712171768463e-05, "loss": 0.0204, "step": 9905 }, { "epoch": 1.073937554206418, "grad_norm": 1.0832035541534424, "learning_rate": 1.2886986798201075e-05, "loss": 0.0216, "step": 9906 }, { "epoch": 1.0740459670424978, "grad_norm": 5.284449100494385, "learning_rate": 1.2886261424633689e-05, "loss": 0.0226, "step": 9907 }, { "epoch": 1.0741543798785775, "grad_norm": 0.38356253504753113, "learning_rate": 1.28855360510663e-05, "loss": 0.0114, "step": 9908 }, { "epoch": 1.0742627927146575, "grad_norm": 1.6647498607635498, "learning_rate": 1.2884810677498913e-05, "loss": 0.0258, "step": 9909 }, { "epoch": 1.0743712055507373, "grad_norm": 2.346545457839966, "learning_rate": 1.2884085303931527e-05, "loss": 0.0165, "step": 9910 }, { "epoch": 1.074479618386817, "grad_norm": 0.7547392249107361, "learning_rate": 1.2883359930364139e-05, "loss": 0.014, "step": 9911 }, { "epoch": 1.0745880312228968, "grad_norm": 3.967381715774536, "learning_rate": 1.2882634556796751e-05, "loss": 0.0417, "step": 9912 }, { "epoch": 1.0746964440589766, "grad_norm": 5.041220188140869, "learning_rate": 1.2881909183229365e-05, "loss": 0.0435, "step": 9913 }, { "epoch": 1.0748048568950563, "grad_norm": 13.326567649841309, "learning_rate": 1.2881183809661977e-05, "loss": 0.0726, "step": 9914 }, { "epoch": 1.074913269731136, "grad_norm": 1.1501550674438477, "learning_rate": 1.288045843609459e-05, "loss": 0.0034, "step": 9915 }, { "epoch": 1.0750216825672159, "grad_norm": 0.7812813520431519, "learning_rate": 1.2879733062527201e-05, "loss": 0.0019, "step": 9916 }, { "epoch": 1.0751300954032958, "grad_norm": 3.9892561435699463, "learning_rate": 1.2879007688959815e-05, "loss": 0.0686, "step": 9917 }, { "epoch": 1.0752385082393756, "grad_norm": 0.7510021328926086, "learning_rate": 1.2878282315392427e-05, "loss": 0.0064, "step": 9918 }, { "epoch": 1.0753469210754554, "grad_norm": 0.9699954390525818, "learning_rate": 1.287755694182504e-05, "loss": 0.0189, "step": 9919 }, { "epoch": 1.0754553339115351, "grad_norm": 0.8107630014419556, "learning_rate": 1.2876831568257653e-05, "loss": 0.0025, "step": 9920 }, { "epoch": 1.075563746747615, "grad_norm": 0.805063784122467, "learning_rate": 1.2876106194690265e-05, "loss": 0.0145, "step": 9921 }, { "epoch": 1.0756721595836947, "grad_norm": 1.088112711906433, "learning_rate": 1.2875380821122881e-05, "loss": 0.0033, "step": 9922 }, { "epoch": 1.0757805724197744, "grad_norm": 2.5427231788635254, "learning_rate": 1.2874655447555493e-05, "loss": 0.006, "step": 9923 }, { "epoch": 1.0758889852558542, "grad_norm": 1.5970996618270874, "learning_rate": 1.2873930073988105e-05, "loss": 0.0131, "step": 9924 }, { "epoch": 1.0759973980919342, "grad_norm": 4.6690263748168945, "learning_rate": 1.2873204700420719e-05, "loss": 0.0293, "step": 9925 }, { "epoch": 1.076105810928014, "grad_norm": 2.742994546890259, "learning_rate": 1.2872479326853331e-05, "loss": 0.0037, "step": 9926 }, { "epoch": 1.0762142237640937, "grad_norm": 4.793099880218506, "learning_rate": 1.2871753953285943e-05, "loss": 0.0271, "step": 9927 }, { "epoch": 1.0763226366001735, "grad_norm": 1.2671722173690796, "learning_rate": 1.2871028579718556e-05, "loss": 0.0115, "step": 9928 }, { "epoch": 1.0764310494362532, "grad_norm": 0.06345885992050171, "learning_rate": 1.287030320615117e-05, "loss": 0.0004, "step": 9929 }, { "epoch": 1.076539462272333, "grad_norm": 0.13367438316345215, "learning_rate": 1.2869577832583782e-05, "loss": 0.0005, "step": 9930 }, { "epoch": 1.0766478751084128, "grad_norm": 0.29186734557151794, "learning_rate": 1.2868852459016394e-05, "loss": 0.0061, "step": 9931 }, { "epoch": 1.0767562879444927, "grad_norm": 6.678895950317383, "learning_rate": 1.2868127085449007e-05, "loss": 0.0166, "step": 9932 }, { "epoch": 1.0768647007805725, "grad_norm": 2.6819260120391846, "learning_rate": 1.286740171188162e-05, "loss": 0.0512, "step": 9933 }, { "epoch": 1.0769731136166523, "grad_norm": 0.2565784752368927, "learning_rate": 1.2866676338314232e-05, "loss": 0.0008, "step": 9934 }, { "epoch": 1.077081526452732, "grad_norm": 1.5420989990234375, "learning_rate": 1.2865950964746846e-05, "loss": 0.0588, "step": 9935 }, { "epoch": 1.0771899392888118, "grad_norm": 0.23012828826904297, "learning_rate": 1.2865225591179458e-05, "loss": 0.0008, "step": 9936 }, { "epoch": 1.0772983521248916, "grad_norm": 0.7168587446212769, "learning_rate": 1.286450021761207e-05, "loss": 0.0107, "step": 9937 }, { "epoch": 1.0774067649609713, "grad_norm": 2.6725661754608154, "learning_rate": 1.2863774844044684e-05, "loss": 0.0255, "step": 9938 }, { "epoch": 1.077515177797051, "grad_norm": 3.6691300868988037, "learning_rate": 1.2863049470477298e-05, "loss": 0.0489, "step": 9939 }, { "epoch": 1.0776235906331308, "grad_norm": 2.778684377670288, "learning_rate": 1.2862324096909911e-05, "loss": 0.0185, "step": 9940 }, { "epoch": 1.0777320034692108, "grad_norm": 11.096457481384277, "learning_rate": 1.2861598723342524e-05, "loss": 0.0164, "step": 9941 }, { "epoch": 1.0778404163052906, "grad_norm": 1.610951542854309, "learning_rate": 1.2860873349775136e-05, "loss": 0.0034, "step": 9942 }, { "epoch": 1.0779488291413704, "grad_norm": 2.9346511363983154, "learning_rate": 1.2860147976207748e-05, "loss": 0.0474, "step": 9943 }, { "epoch": 1.0780572419774501, "grad_norm": 8.811359405517578, "learning_rate": 1.2859422602640362e-05, "loss": 0.1254, "step": 9944 }, { "epoch": 1.0781656548135299, "grad_norm": 2.821640729904175, "learning_rate": 1.2858697229072974e-05, "loss": 0.0037, "step": 9945 }, { "epoch": 1.0782740676496096, "grad_norm": 1.3396227359771729, "learning_rate": 1.2857971855505586e-05, "loss": 0.0081, "step": 9946 }, { "epoch": 1.0783824804856894, "grad_norm": 0.003767340676859021, "learning_rate": 1.28572464819382e-05, "loss": 0.0, "step": 9947 }, { "epoch": 1.0784908933217694, "grad_norm": 3.405557632446289, "learning_rate": 1.2856521108370812e-05, "loss": 0.053, "step": 9948 }, { "epoch": 1.0785993061578492, "grad_norm": 0.21793563663959503, "learning_rate": 1.2855795734803424e-05, "loss": 0.0016, "step": 9949 }, { "epoch": 1.078707718993929, "grad_norm": 0.5515692830085754, "learning_rate": 1.2855070361236038e-05, "loss": 0.0022, "step": 9950 }, { "epoch": 1.0788161318300087, "grad_norm": 6.690428256988525, "learning_rate": 1.285434498766865e-05, "loss": 0.0487, "step": 9951 }, { "epoch": 1.0789245446660884, "grad_norm": 4.824116230010986, "learning_rate": 1.2853619614101262e-05, "loss": 0.0587, "step": 9952 }, { "epoch": 1.0790329575021682, "grad_norm": 0.3828524947166443, "learning_rate": 1.2852894240533876e-05, "loss": 0.0054, "step": 9953 }, { "epoch": 1.079141370338248, "grad_norm": 0.020818499848246574, "learning_rate": 1.2852168866966488e-05, "loss": 0.0002, "step": 9954 }, { "epoch": 1.0792497831743277, "grad_norm": 0.3644108474254608, "learning_rate": 1.28514434933991e-05, "loss": 0.0101, "step": 9955 }, { "epoch": 1.0793581960104077, "grad_norm": 2.4605042934417725, "learning_rate": 1.2850718119831716e-05, "loss": 0.0181, "step": 9956 }, { "epoch": 1.0794666088464875, "grad_norm": 3.409778356552124, "learning_rate": 1.2849992746264328e-05, "loss": 0.0599, "step": 9957 }, { "epoch": 1.0795750216825672, "grad_norm": 0.32555699348449707, "learning_rate": 1.284926737269694e-05, "loss": 0.0063, "step": 9958 }, { "epoch": 1.079683434518647, "grad_norm": 2.112271547317505, "learning_rate": 1.2848541999129554e-05, "loss": 0.0511, "step": 9959 }, { "epoch": 1.0797918473547268, "grad_norm": 0.4080270528793335, "learning_rate": 1.2847816625562166e-05, "loss": 0.0086, "step": 9960 }, { "epoch": 1.0799002601908065, "grad_norm": 0.46728670597076416, "learning_rate": 1.2847091251994778e-05, "loss": 0.0311, "step": 9961 }, { "epoch": 1.0800086730268863, "grad_norm": 3.352463960647583, "learning_rate": 1.2846365878427392e-05, "loss": 0.0142, "step": 9962 }, { "epoch": 1.080117085862966, "grad_norm": 14.263521194458008, "learning_rate": 1.2845640504860004e-05, "loss": 0.0271, "step": 9963 }, { "epoch": 1.080225498699046, "grad_norm": 0.42305630445480347, "learning_rate": 1.2844915131292616e-05, "loss": 0.0096, "step": 9964 }, { "epoch": 1.0803339115351258, "grad_norm": 2.3150501251220703, "learning_rate": 1.284418975772523e-05, "loss": 0.0289, "step": 9965 }, { "epoch": 1.0804423243712056, "grad_norm": 5.088493824005127, "learning_rate": 1.2843464384157842e-05, "loss": 0.0285, "step": 9966 }, { "epoch": 1.0805507372072853, "grad_norm": 2.0311219692230225, "learning_rate": 1.2842739010590455e-05, "loss": 0.0213, "step": 9967 }, { "epoch": 1.080659150043365, "grad_norm": 1.5263954401016235, "learning_rate": 1.2842013637023067e-05, "loss": 0.017, "step": 9968 }, { "epoch": 1.0807675628794449, "grad_norm": 0.10107234120368958, "learning_rate": 1.284128826345568e-05, "loss": 0.0008, "step": 9969 }, { "epoch": 1.0808759757155246, "grad_norm": 1.3376201391220093, "learning_rate": 1.2840562889888293e-05, "loss": 0.0393, "step": 9970 }, { "epoch": 1.0809843885516046, "grad_norm": 1.7887628078460693, "learning_rate": 1.2839837516320905e-05, "loss": 0.0123, "step": 9971 }, { "epoch": 1.0810928013876844, "grad_norm": 0.85093754529953, "learning_rate": 1.2839112142753519e-05, "loss": 0.011, "step": 9972 }, { "epoch": 1.0812012142237641, "grad_norm": 3.6075363159179688, "learning_rate": 1.283838676918613e-05, "loss": 0.02, "step": 9973 }, { "epoch": 1.081309627059844, "grad_norm": 1.2174779176712036, "learning_rate": 1.2837661395618746e-05, "loss": 0.003, "step": 9974 }, { "epoch": 1.0814180398959237, "grad_norm": 1.466941475868225, "learning_rate": 1.2836936022051358e-05, "loss": 0.0563, "step": 9975 }, { "epoch": 1.0815264527320034, "grad_norm": 3.8468141555786133, "learning_rate": 1.283621064848397e-05, "loss": 0.0328, "step": 9976 }, { "epoch": 1.0816348655680832, "grad_norm": 4.607918739318848, "learning_rate": 1.2835485274916584e-05, "loss": 0.0418, "step": 9977 }, { "epoch": 1.081743278404163, "grad_norm": 0.5598578453063965, "learning_rate": 1.2834759901349197e-05, "loss": 0.0026, "step": 9978 }, { "epoch": 1.081851691240243, "grad_norm": 0.7246946096420288, "learning_rate": 1.2834034527781809e-05, "loss": 0.0126, "step": 9979 }, { "epoch": 1.0819601040763227, "grad_norm": 0.9817759990692139, "learning_rate": 1.2833309154214421e-05, "loss": 0.005, "step": 9980 }, { "epoch": 1.0820685169124025, "grad_norm": 0.47456932067871094, "learning_rate": 1.2832583780647035e-05, "loss": 0.0115, "step": 9981 }, { "epoch": 1.0821769297484822, "grad_norm": 10.728769302368164, "learning_rate": 1.2831858407079647e-05, "loss": 0.0617, "step": 9982 }, { "epoch": 1.082285342584562, "grad_norm": 2.620926856994629, "learning_rate": 1.2831133033512259e-05, "loss": 0.0049, "step": 9983 }, { "epoch": 1.0823937554206418, "grad_norm": 1.3776856660842896, "learning_rate": 1.2830407659944873e-05, "loss": 0.0081, "step": 9984 }, { "epoch": 1.0825021682567215, "grad_norm": 3.891362428665161, "learning_rate": 1.2829682286377485e-05, "loss": 0.0247, "step": 9985 }, { "epoch": 1.0826105810928013, "grad_norm": 0.06496604532003403, "learning_rate": 1.2828956912810097e-05, "loss": 0.0006, "step": 9986 }, { "epoch": 1.0827189939288813, "grad_norm": 0.7607097625732422, "learning_rate": 1.2828231539242711e-05, "loss": 0.0212, "step": 9987 }, { "epoch": 1.082827406764961, "grad_norm": 0.5064356923103333, "learning_rate": 1.2827506165675323e-05, "loss": 0.013, "step": 9988 }, { "epoch": 1.0829358196010408, "grad_norm": 4.916568279266357, "learning_rate": 1.2826780792107935e-05, "loss": 0.0376, "step": 9989 }, { "epoch": 1.0830442324371206, "grad_norm": 2.1306135654449463, "learning_rate": 1.2826055418540549e-05, "loss": 0.0446, "step": 9990 }, { "epoch": 1.0831526452732003, "grad_norm": 1.946634292602539, "learning_rate": 1.2825330044973163e-05, "loss": 0.0239, "step": 9991 }, { "epoch": 1.08326105810928, "grad_norm": 1.109544038772583, "learning_rate": 1.2824604671405777e-05, "loss": 0.0166, "step": 9992 }, { "epoch": 1.0833694709453598, "grad_norm": 1.5627821683883667, "learning_rate": 1.2823879297838389e-05, "loss": 0.0258, "step": 9993 }, { "epoch": 1.0834778837814398, "grad_norm": 1.188489317893982, "learning_rate": 1.2823153924271001e-05, "loss": 0.016, "step": 9994 }, { "epoch": 1.0835862966175196, "grad_norm": 4.540463447570801, "learning_rate": 1.2822428550703613e-05, "loss": 0.0359, "step": 9995 }, { "epoch": 1.0836947094535994, "grad_norm": 0.19484087824821472, "learning_rate": 1.2821703177136227e-05, "loss": 0.0011, "step": 9996 }, { "epoch": 1.0838031222896791, "grad_norm": 3.0051496028900146, "learning_rate": 1.282097780356884e-05, "loss": 0.0729, "step": 9997 }, { "epoch": 1.0839115351257589, "grad_norm": 2.5111513137817383, "learning_rate": 1.2820252430001451e-05, "loss": 0.0367, "step": 9998 }, { "epoch": 1.0840199479618386, "grad_norm": 0.09057524055242538, "learning_rate": 1.2819527056434065e-05, "loss": 0.0009, "step": 9999 }, { "epoch": 1.0841283607979184, "grad_norm": 0.16989871859550476, "learning_rate": 1.2818801682866677e-05, "loss": 0.0017, "step": 10000 }, { "epoch": 1.0842367736339982, "grad_norm": 0.8128799796104431, "learning_rate": 1.281807630929929e-05, "loss": 0.0054, "step": 10001 }, { "epoch": 1.0843451864700782, "grad_norm": 4.123354911804199, "learning_rate": 1.2817350935731903e-05, "loss": 0.0347, "step": 10002 }, { "epoch": 1.084453599306158, "grad_norm": 1.9153056144714355, "learning_rate": 1.2816625562164515e-05, "loss": 0.0363, "step": 10003 }, { "epoch": 1.0845620121422377, "grad_norm": 4.00861120223999, "learning_rate": 1.2815900188597128e-05, "loss": 0.0246, "step": 10004 }, { "epoch": 1.0846704249783174, "grad_norm": 1.5132513046264648, "learning_rate": 1.2815174815029741e-05, "loss": 0.0188, "step": 10005 }, { "epoch": 1.0847788378143972, "grad_norm": 0.5164042711257935, "learning_rate": 1.2814449441462354e-05, "loss": 0.0182, "step": 10006 }, { "epoch": 1.084887250650477, "grad_norm": 1.6113098859786987, "learning_rate": 1.2813724067894966e-05, "loss": 0.0172, "step": 10007 }, { "epoch": 1.0849956634865567, "grad_norm": 0.7064293622970581, "learning_rate": 1.2812998694327581e-05, "loss": 0.0211, "step": 10008 }, { "epoch": 1.0851040763226365, "grad_norm": 0.5161158442497253, "learning_rate": 1.2812273320760193e-05, "loss": 0.0084, "step": 10009 }, { "epoch": 1.0852124891587165, "grad_norm": 0.25608211755752563, "learning_rate": 1.2811547947192805e-05, "loss": 0.0026, "step": 10010 }, { "epoch": 1.0853209019947962, "grad_norm": 4.112806797027588, "learning_rate": 1.281082257362542e-05, "loss": 0.0428, "step": 10011 }, { "epoch": 1.085429314830876, "grad_norm": 9.430105209350586, "learning_rate": 1.2810097200058031e-05, "loss": 0.0212, "step": 10012 }, { "epoch": 1.0855377276669558, "grad_norm": 1.2020952701568604, "learning_rate": 1.2809371826490644e-05, "loss": 0.0514, "step": 10013 }, { "epoch": 1.0856461405030355, "grad_norm": 0.30774715542793274, "learning_rate": 1.2808646452923257e-05, "loss": 0.0055, "step": 10014 }, { "epoch": 1.0857545533391153, "grad_norm": 9.481781959533691, "learning_rate": 1.280792107935587e-05, "loss": 0.0289, "step": 10015 }, { "epoch": 1.085862966175195, "grad_norm": 0.11767084896564484, "learning_rate": 1.2807195705788482e-05, "loss": 0.0009, "step": 10016 }, { "epoch": 1.085971379011275, "grad_norm": 1.7448865175247192, "learning_rate": 1.2806470332221096e-05, "loss": 0.0053, "step": 10017 }, { "epoch": 1.0860797918473548, "grad_norm": 0.024813514202833176, "learning_rate": 1.2805744958653708e-05, "loss": 0.0003, "step": 10018 }, { "epoch": 1.0861882046834346, "grad_norm": 0.12390144914388657, "learning_rate": 1.280501958508632e-05, "loss": 0.0014, "step": 10019 }, { "epoch": 1.0862966175195143, "grad_norm": 13.934171676635742, "learning_rate": 1.2804294211518932e-05, "loss": 0.0537, "step": 10020 }, { "epoch": 1.086405030355594, "grad_norm": 5.520460605621338, "learning_rate": 1.2803568837951546e-05, "loss": 0.0378, "step": 10021 }, { "epoch": 1.0865134431916739, "grad_norm": 1.7307405471801758, "learning_rate": 1.2802843464384158e-05, "loss": 0.0305, "step": 10022 }, { "epoch": 1.0866218560277536, "grad_norm": 0.4072401523590088, "learning_rate": 1.280211809081677e-05, "loss": 0.0222, "step": 10023 }, { "epoch": 1.0867302688638334, "grad_norm": 0.7271900177001953, "learning_rate": 1.2801392717249384e-05, "loss": 0.0285, "step": 10024 }, { "epoch": 1.0868386816999132, "grad_norm": 1.827413558959961, "learning_rate": 1.2800667343681998e-05, "loss": 0.0311, "step": 10025 }, { "epoch": 1.0869470945359931, "grad_norm": 0.06961102038621902, "learning_rate": 1.2799941970114612e-05, "loss": 0.0006, "step": 10026 }, { "epoch": 1.087055507372073, "grad_norm": 2.0992674827575684, "learning_rate": 1.2799216596547224e-05, "loss": 0.0083, "step": 10027 }, { "epoch": 1.0871639202081527, "grad_norm": 5.190422534942627, "learning_rate": 1.2798491222979836e-05, "loss": 0.0157, "step": 10028 }, { "epoch": 1.0872723330442324, "grad_norm": 0.7017732262611389, "learning_rate": 1.279776584941245e-05, "loss": 0.0094, "step": 10029 }, { "epoch": 1.0873807458803122, "grad_norm": 0.18517525494098663, "learning_rate": 1.2797040475845062e-05, "loss": 0.0019, "step": 10030 }, { "epoch": 1.087489158716392, "grad_norm": 1.2258511781692505, "learning_rate": 1.2796315102277674e-05, "loss": 0.014, "step": 10031 }, { "epoch": 1.0875975715524717, "grad_norm": 3.5789341926574707, "learning_rate": 1.2795589728710286e-05, "loss": 0.0207, "step": 10032 }, { "epoch": 1.0877059843885517, "grad_norm": 2.344564437866211, "learning_rate": 1.27948643551429e-05, "loss": 0.0211, "step": 10033 }, { "epoch": 1.0878143972246315, "grad_norm": 0.1865938901901245, "learning_rate": 1.2794138981575512e-05, "loss": 0.0015, "step": 10034 }, { "epoch": 1.0879228100607112, "grad_norm": 0.41668078303337097, "learning_rate": 1.2793413608008124e-05, "loss": 0.0118, "step": 10035 }, { "epoch": 1.088031222896791, "grad_norm": 3.1892600059509277, "learning_rate": 1.2792688234440738e-05, "loss": 0.0154, "step": 10036 }, { "epoch": 1.0881396357328708, "grad_norm": 2.2070960998535156, "learning_rate": 1.279196286087335e-05, "loss": 0.0403, "step": 10037 }, { "epoch": 1.0882480485689505, "grad_norm": 1.5523484945297241, "learning_rate": 1.2791237487305962e-05, "loss": 0.0243, "step": 10038 }, { "epoch": 1.0883564614050303, "grad_norm": 4.078056335449219, "learning_rate": 1.2790512113738576e-05, "loss": 0.0153, "step": 10039 }, { "epoch": 1.08846487424111, "grad_norm": 0.41876697540283203, "learning_rate": 1.2789786740171188e-05, "loss": 0.0114, "step": 10040 }, { "epoch": 1.08857328707719, "grad_norm": 0.2722753584384918, "learning_rate": 1.27890613666038e-05, "loss": 0.002, "step": 10041 }, { "epoch": 1.0886816999132698, "grad_norm": 3.237166166305542, "learning_rate": 1.2788335993036414e-05, "loss": 0.0197, "step": 10042 }, { "epoch": 1.0887901127493496, "grad_norm": 2.4029452800750732, "learning_rate": 1.2787610619469028e-05, "loss": 0.0103, "step": 10043 }, { "epoch": 1.0888985255854293, "grad_norm": 1.9439324140548706, "learning_rate": 1.2786885245901642e-05, "loss": 0.0124, "step": 10044 }, { "epoch": 1.089006938421509, "grad_norm": 0.08084328472614288, "learning_rate": 1.2786159872334254e-05, "loss": 0.0006, "step": 10045 }, { "epoch": 1.0891153512575888, "grad_norm": 1.7422386407852173, "learning_rate": 1.2785434498766866e-05, "loss": 0.0014, "step": 10046 }, { "epoch": 1.0892237640936686, "grad_norm": 4.506454944610596, "learning_rate": 1.2784709125199479e-05, "loss": 0.0244, "step": 10047 }, { "epoch": 1.0893321769297484, "grad_norm": 1.335930347442627, "learning_rate": 1.2783983751632092e-05, "loss": 0.0222, "step": 10048 }, { "epoch": 1.0894405897658284, "grad_norm": 3.496673107147217, "learning_rate": 1.2783258378064704e-05, "loss": 0.0217, "step": 10049 }, { "epoch": 1.0895490026019081, "grad_norm": 6.353552341461182, "learning_rate": 1.2782533004497317e-05, "loss": 0.0825, "step": 10050 }, { "epoch": 1.0896574154379879, "grad_norm": 3.4903688430786133, "learning_rate": 1.278180763092993e-05, "loss": 0.0197, "step": 10051 }, { "epoch": 1.0897658282740676, "grad_norm": 0.5278635621070862, "learning_rate": 1.2781082257362543e-05, "loss": 0.0152, "step": 10052 }, { "epoch": 1.0898742411101474, "grad_norm": 5.170781135559082, "learning_rate": 1.2780356883795155e-05, "loss": 0.0397, "step": 10053 }, { "epoch": 1.0899826539462272, "grad_norm": 1.9493188858032227, "learning_rate": 1.2779631510227769e-05, "loss": 0.0157, "step": 10054 }, { "epoch": 1.090091066782307, "grad_norm": 2.537480592727661, "learning_rate": 1.277890613666038e-05, "loss": 0.0261, "step": 10055 }, { "epoch": 1.090199479618387, "grad_norm": 0.2427615076303482, "learning_rate": 1.2778180763092993e-05, "loss": 0.0123, "step": 10056 }, { "epoch": 1.0903078924544667, "grad_norm": 0.489601731300354, "learning_rate": 1.2777455389525607e-05, "loss": 0.0016, "step": 10057 }, { "epoch": 1.0904163052905465, "grad_norm": 5.797685146331787, "learning_rate": 1.2776730015958219e-05, "loss": 0.0091, "step": 10058 }, { "epoch": 1.0905247181266262, "grad_norm": 10.855294227600098, "learning_rate": 1.2776004642390831e-05, "loss": 0.0126, "step": 10059 }, { "epoch": 1.090633130962706, "grad_norm": 5.205140113830566, "learning_rate": 1.2775279268823447e-05, "loss": 0.0563, "step": 10060 }, { "epoch": 1.0907415437987857, "grad_norm": 0.5313560962677002, "learning_rate": 1.2774553895256059e-05, "loss": 0.0039, "step": 10061 }, { "epoch": 1.0908499566348655, "grad_norm": 5.067217826843262, "learning_rate": 1.277382852168867e-05, "loss": 0.0636, "step": 10062 }, { "epoch": 1.0909583694709453, "grad_norm": 0.5211114287376404, "learning_rate": 1.2773103148121285e-05, "loss": 0.0018, "step": 10063 }, { "epoch": 1.0910667823070253, "grad_norm": 2.340449333190918, "learning_rate": 1.2772377774553897e-05, "loss": 0.0162, "step": 10064 }, { "epoch": 1.091175195143105, "grad_norm": 2.1042251586914062, "learning_rate": 1.2771652400986509e-05, "loss": 0.0125, "step": 10065 }, { "epoch": 1.0912836079791848, "grad_norm": 0.2893965542316437, "learning_rate": 1.2770927027419123e-05, "loss": 0.0076, "step": 10066 }, { "epoch": 1.0913920208152645, "grad_norm": 2.2556769847869873, "learning_rate": 1.2770201653851735e-05, "loss": 0.0193, "step": 10067 }, { "epoch": 1.0915004336513443, "grad_norm": 5.345326900482178, "learning_rate": 1.2769476280284347e-05, "loss": 0.015, "step": 10068 }, { "epoch": 1.091608846487424, "grad_norm": 2.705315351486206, "learning_rate": 1.2768750906716961e-05, "loss": 0.0644, "step": 10069 }, { "epoch": 1.0917172593235038, "grad_norm": 0.09112769365310669, "learning_rate": 1.2768025533149573e-05, "loss": 0.0007, "step": 10070 }, { "epoch": 1.0918256721595836, "grad_norm": 0.44356444478034973, "learning_rate": 1.2767300159582185e-05, "loss": 0.0091, "step": 10071 }, { "epoch": 1.0919340849956636, "grad_norm": 10.242758750915527, "learning_rate": 1.2766574786014797e-05, "loss": 0.016, "step": 10072 }, { "epoch": 1.0920424978317433, "grad_norm": 2.175384283065796, "learning_rate": 1.2765849412447411e-05, "loss": 0.0519, "step": 10073 }, { "epoch": 1.092150910667823, "grad_norm": 2.2454373836517334, "learning_rate": 1.2765124038880023e-05, "loss": 0.0043, "step": 10074 }, { "epoch": 1.0922593235039029, "grad_norm": 3.345734119415283, "learning_rate": 1.2764398665312635e-05, "loss": 0.036, "step": 10075 }, { "epoch": 1.0923677363399826, "grad_norm": 0.27437496185302734, "learning_rate": 1.276367329174525e-05, "loss": 0.0011, "step": 10076 }, { "epoch": 1.0924761491760624, "grad_norm": 9.092126846313477, "learning_rate": 1.2762947918177863e-05, "loss": 0.027, "step": 10077 }, { "epoch": 1.0925845620121422, "grad_norm": 3.501424551010132, "learning_rate": 1.2762222544610477e-05, "loss": 0.029, "step": 10078 }, { "epoch": 1.0926929748482221, "grad_norm": 0.9344944953918457, "learning_rate": 1.2761497171043089e-05, "loss": 0.0022, "step": 10079 }, { "epoch": 1.092801387684302, "grad_norm": 2.7168612480163574, "learning_rate": 1.2760771797475701e-05, "loss": 0.0191, "step": 10080 }, { "epoch": 1.0929098005203817, "grad_norm": 10.893120765686035, "learning_rate": 1.2760046423908315e-05, "loss": 0.0357, "step": 10081 }, { "epoch": 1.0930182133564614, "grad_norm": 1.4572404623031616, "learning_rate": 1.2759321050340927e-05, "loss": 0.0453, "step": 10082 }, { "epoch": 1.0931266261925412, "grad_norm": 0.03446921706199646, "learning_rate": 1.275859567677354e-05, "loss": 0.0004, "step": 10083 }, { "epoch": 1.093235039028621, "grad_norm": 2.967978000640869, "learning_rate": 1.2757870303206153e-05, "loss": 0.0604, "step": 10084 }, { "epoch": 1.0933434518647007, "grad_norm": 2.5561020374298096, "learning_rate": 1.2757144929638765e-05, "loss": 0.0083, "step": 10085 }, { "epoch": 1.0934518647007805, "grad_norm": 4.355749607086182, "learning_rate": 1.2756419556071378e-05, "loss": 0.0906, "step": 10086 }, { "epoch": 1.0935602775368602, "grad_norm": 4.227186679840088, "learning_rate": 1.275569418250399e-05, "loss": 0.1189, "step": 10087 }, { "epoch": 1.0936686903729402, "grad_norm": 0.5398635268211365, "learning_rate": 1.2754968808936603e-05, "loss": 0.0035, "step": 10088 }, { "epoch": 1.09377710320902, "grad_norm": 1.754058599472046, "learning_rate": 1.2754243435369216e-05, "loss": 0.0241, "step": 10089 }, { "epoch": 1.0938855160450998, "grad_norm": 1.6601440906524658, "learning_rate": 1.2753518061801828e-05, "loss": 0.0134, "step": 10090 }, { "epoch": 1.0939939288811795, "grad_norm": 0.1009831577539444, "learning_rate": 1.2752792688234442e-05, "loss": 0.0009, "step": 10091 }, { "epoch": 1.0941023417172593, "grad_norm": 4.973974704742432, "learning_rate": 1.2752067314667054e-05, "loss": 0.048, "step": 10092 }, { "epoch": 1.094210754553339, "grad_norm": 0.6890907287597656, "learning_rate": 1.2751341941099666e-05, "loss": 0.0194, "step": 10093 }, { "epoch": 1.0943191673894188, "grad_norm": 0.5941234230995178, "learning_rate": 1.2750616567532281e-05, "loss": 0.0151, "step": 10094 }, { "epoch": 1.0944275802254988, "grad_norm": 0.9497804641723633, "learning_rate": 1.2749891193964894e-05, "loss": 0.0079, "step": 10095 }, { "epoch": 1.0945359930615786, "grad_norm": 0.403821736574173, "learning_rate": 1.2749165820397507e-05, "loss": 0.014, "step": 10096 }, { "epoch": 1.0946444058976583, "grad_norm": 1.108542799949646, "learning_rate": 1.274844044683012e-05, "loss": 0.0163, "step": 10097 }, { "epoch": 1.094752818733738, "grad_norm": 4.220431804656982, "learning_rate": 1.2747715073262732e-05, "loss": 0.005, "step": 10098 }, { "epoch": 1.0948612315698178, "grad_norm": 0.4703908860683441, "learning_rate": 1.2746989699695344e-05, "loss": 0.0123, "step": 10099 }, { "epoch": 1.0949696444058976, "grad_norm": 1.5979630947113037, "learning_rate": 1.2746264326127958e-05, "loss": 0.0648, "step": 10100 }, { "epoch": 1.0950780572419774, "grad_norm": 0.5171712636947632, "learning_rate": 1.274553895256057e-05, "loss": 0.0053, "step": 10101 }, { "epoch": 1.0951864700780571, "grad_norm": 1.4563153982162476, "learning_rate": 1.2744813578993182e-05, "loss": 0.0182, "step": 10102 }, { "epoch": 1.0952948829141371, "grad_norm": 3.7808945178985596, "learning_rate": 1.2744088205425796e-05, "loss": 0.0306, "step": 10103 }, { "epoch": 1.0954032957502169, "grad_norm": 6.284337043762207, "learning_rate": 1.2743362831858408e-05, "loss": 0.0312, "step": 10104 }, { "epoch": 1.0955117085862967, "grad_norm": 5.169218063354492, "learning_rate": 1.274263745829102e-05, "loss": 0.0214, "step": 10105 }, { "epoch": 1.0956201214223764, "grad_norm": 2.5361435413360596, "learning_rate": 1.2741912084723634e-05, "loss": 0.0232, "step": 10106 }, { "epoch": 1.0957285342584562, "grad_norm": 0.3279423117637634, "learning_rate": 1.2741186711156246e-05, "loss": 0.0085, "step": 10107 }, { "epoch": 1.095836947094536, "grad_norm": 1.913529872894287, "learning_rate": 1.2740461337588858e-05, "loss": 0.027, "step": 10108 }, { "epoch": 1.0959453599306157, "grad_norm": 0.9801921248435974, "learning_rate": 1.2739735964021472e-05, "loss": 0.0123, "step": 10109 }, { "epoch": 1.0960537727666955, "grad_norm": 2.9701855182647705, "learning_rate": 1.2739010590454084e-05, "loss": 0.0425, "step": 10110 }, { "epoch": 1.0961621856027755, "grad_norm": 2.0097496509552, "learning_rate": 1.2738285216886696e-05, "loss": 0.0177, "step": 10111 }, { "epoch": 1.0962705984388552, "grad_norm": 3.97574520111084, "learning_rate": 1.2737559843319312e-05, "loss": 0.04, "step": 10112 }, { "epoch": 1.096379011274935, "grad_norm": 2.8176488876342773, "learning_rate": 1.2736834469751924e-05, "loss": 0.0569, "step": 10113 }, { "epoch": 1.0964874241110147, "grad_norm": 3.8333756923675537, "learning_rate": 1.2736109096184536e-05, "loss": 0.0149, "step": 10114 }, { "epoch": 1.0965958369470945, "grad_norm": 1.1487557888031006, "learning_rate": 1.273538372261715e-05, "loss": 0.0171, "step": 10115 }, { "epoch": 1.0967042497831743, "grad_norm": 1.26828932762146, "learning_rate": 1.2734658349049762e-05, "loss": 0.0227, "step": 10116 }, { "epoch": 1.096812662619254, "grad_norm": 2.315419912338257, "learning_rate": 1.2733932975482374e-05, "loss": 0.0556, "step": 10117 }, { "epoch": 1.096921075455334, "grad_norm": 0.6657808423042297, "learning_rate": 1.2733207601914988e-05, "loss": 0.005, "step": 10118 }, { "epoch": 1.0970294882914138, "grad_norm": 2.0036938190460205, "learning_rate": 1.27324822283476e-05, "loss": 0.0198, "step": 10119 }, { "epoch": 1.0971379011274935, "grad_norm": 4.168649196624756, "learning_rate": 1.2731756854780212e-05, "loss": 0.0159, "step": 10120 }, { "epoch": 1.0972463139635733, "grad_norm": 0.6002011299133301, "learning_rate": 1.2731031481212826e-05, "loss": 0.0133, "step": 10121 }, { "epoch": 1.097354726799653, "grad_norm": 0.11510282009840012, "learning_rate": 1.2730306107645438e-05, "loss": 0.0115, "step": 10122 }, { "epoch": 1.0974631396357328, "grad_norm": 8.116156578063965, "learning_rate": 1.272958073407805e-05, "loss": 0.0311, "step": 10123 }, { "epoch": 1.0975715524718126, "grad_norm": 0.3728950023651123, "learning_rate": 1.2728855360510663e-05, "loss": 0.003, "step": 10124 }, { "epoch": 1.0976799653078924, "grad_norm": 0.26490187644958496, "learning_rate": 1.2728129986943277e-05, "loss": 0.0069, "step": 10125 }, { "epoch": 1.0977883781439723, "grad_norm": 4.164005279541016, "learning_rate": 1.2727404613375889e-05, "loss": 0.0532, "step": 10126 }, { "epoch": 1.097896790980052, "grad_norm": 1.9544600248336792, "learning_rate": 1.27266792398085e-05, "loss": 0.0144, "step": 10127 }, { "epoch": 1.0980052038161319, "grad_norm": 0.5698323249816895, "learning_rate": 1.2725953866241115e-05, "loss": 0.0052, "step": 10128 }, { "epoch": 1.0981136166522116, "grad_norm": 2.876386880874634, "learning_rate": 1.2725228492673728e-05, "loss": 0.0259, "step": 10129 }, { "epoch": 1.0982220294882914, "grad_norm": 1.2544795274734497, "learning_rate": 1.2724503119106342e-05, "loss": 0.0322, "step": 10130 }, { "epoch": 1.0983304423243712, "grad_norm": 3.121474027633667, "learning_rate": 1.2723777745538954e-05, "loss": 0.0082, "step": 10131 }, { "epoch": 1.098438855160451, "grad_norm": 1.006396770477295, "learning_rate": 1.2723052371971567e-05, "loss": 0.0109, "step": 10132 }, { "epoch": 1.0985472679965307, "grad_norm": 2.314570426940918, "learning_rate": 1.272232699840418e-05, "loss": 0.0739, "step": 10133 }, { "epoch": 1.0986556808326107, "grad_norm": 0.4390665888786316, "learning_rate": 1.2721601624836793e-05, "loss": 0.017, "step": 10134 }, { "epoch": 1.0987640936686904, "grad_norm": 6.04393196105957, "learning_rate": 1.2720876251269405e-05, "loss": 0.0504, "step": 10135 }, { "epoch": 1.0988725065047702, "grad_norm": 0.09035855531692505, "learning_rate": 1.2720150877702019e-05, "loss": 0.0112, "step": 10136 }, { "epoch": 1.09898091934085, "grad_norm": 1.489833950996399, "learning_rate": 1.271942550413463e-05, "loss": 0.0223, "step": 10137 }, { "epoch": 1.0990893321769297, "grad_norm": 8.231744766235352, "learning_rate": 1.2718700130567243e-05, "loss": 0.0254, "step": 10138 }, { "epoch": 1.0991977450130095, "grad_norm": 1.427884817123413, "learning_rate": 1.2717974756999855e-05, "loss": 0.0096, "step": 10139 }, { "epoch": 1.0993061578490892, "grad_norm": 6.216740608215332, "learning_rate": 1.2717249383432469e-05, "loss": 0.0269, "step": 10140 }, { "epoch": 1.0994145706851692, "grad_norm": 0.4675293564796448, "learning_rate": 1.2716524009865081e-05, "loss": 0.0081, "step": 10141 }, { "epoch": 1.099522983521249, "grad_norm": 0.9842858910560608, "learning_rate": 1.2715798636297693e-05, "loss": 0.0145, "step": 10142 }, { "epoch": 1.0996313963573288, "grad_norm": 4.835000514984131, "learning_rate": 1.2715073262730307e-05, "loss": 0.0327, "step": 10143 }, { "epoch": 1.0997398091934085, "grad_norm": 2.7032968997955322, "learning_rate": 1.2714347889162919e-05, "loss": 0.0144, "step": 10144 }, { "epoch": 1.0998482220294883, "grad_norm": 3.4244751930236816, "learning_rate": 1.2713622515595531e-05, "loss": 0.0187, "step": 10145 }, { "epoch": 1.099956634865568, "grad_norm": 0.19045013189315796, "learning_rate": 1.2712897142028147e-05, "loss": 0.0013, "step": 10146 }, { "epoch": 1.1000650477016478, "grad_norm": 4.8453803062438965, "learning_rate": 1.2712171768460759e-05, "loss": 0.0266, "step": 10147 }, { "epoch": 1.1001734605377276, "grad_norm": 1.5003070831298828, "learning_rate": 1.2711446394893373e-05, "loss": 0.0316, "step": 10148 }, { "epoch": 1.1002818733738076, "grad_norm": 3.9218854904174805, "learning_rate": 1.2710721021325985e-05, "loss": 0.0219, "step": 10149 }, { "epoch": 1.1003902862098873, "grad_norm": 1.7048447132110596, "learning_rate": 1.2709995647758597e-05, "loss": 0.0452, "step": 10150 }, { "epoch": 1.100498699045967, "grad_norm": 6.135798454284668, "learning_rate": 1.270927027419121e-05, "loss": 0.0203, "step": 10151 }, { "epoch": 1.1006071118820469, "grad_norm": 2.9992897510528564, "learning_rate": 1.2708544900623823e-05, "loss": 0.0336, "step": 10152 }, { "epoch": 1.1007155247181266, "grad_norm": 0.5123120546340942, "learning_rate": 1.2707819527056435e-05, "loss": 0.0298, "step": 10153 }, { "epoch": 1.1008239375542064, "grad_norm": 0.697451651096344, "learning_rate": 1.2707094153489047e-05, "loss": 0.0183, "step": 10154 }, { "epoch": 1.1009323503902861, "grad_norm": 4.655820369720459, "learning_rate": 1.2706368779921661e-05, "loss": 0.0528, "step": 10155 }, { "epoch": 1.101040763226366, "grad_norm": 4.766585826873779, "learning_rate": 1.2705643406354273e-05, "loss": 0.0172, "step": 10156 }, { "epoch": 1.101149176062446, "grad_norm": 1.6683365106582642, "learning_rate": 1.2704918032786885e-05, "loss": 0.0354, "step": 10157 }, { "epoch": 1.1012575888985257, "grad_norm": 1.9959454536437988, "learning_rate": 1.27041926592195e-05, "loss": 0.009, "step": 10158 }, { "epoch": 1.1013660017346054, "grad_norm": 0.4169074594974518, "learning_rate": 1.2703467285652111e-05, "loss": 0.0108, "step": 10159 }, { "epoch": 1.1014744145706852, "grad_norm": 0.25146931409835815, "learning_rate": 1.2702741912084724e-05, "loss": 0.0026, "step": 10160 }, { "epoch": 1.101582827406765, "grad_norm": 0.5746005177497864, "learning_rate": 1.2702016538517337e-05, "loss": 0.0135, "step": 10161 }, { "epoch": 1.1016912402428447, "grad_norm": 0.2639920115470886, "learning_rate": 1.270129116494995e-05, "loss": 0.0022, "step": 10162 }, { "epoch": 1.1017996530789245, "grad_norm": 0.5420212745666504, "learning_rate": 1.2700565791382562e-05, "loss": 0.0233, "step": 10163 }, { "epoch": 1.1019080659150045, "grad_norm": 1.6850152015686035, "learning_rate": 1.2699840417815177e-05, "loss": 0.0851, "step": 10164 }, { "epoch": 1.1020164787510842, "grad_norm": 2.7732608318328857, "learning_rate": 1.269911504424779e-05, "loss": 0.0281, "step": 10165 }, { "epoch": 1.102124891587164, "grad_norm": 2.7924611568450928, "learning_rate": 1.2698389670680402e-05, "loss": 0.0201, "step": 10166 }, { "epoch": 1.1022333044232437, "grad_norm": 0.5327793955802917, "learning_rate": 1.2697664297113015e-05, "loss": 0.0034, "step": 10167 }, { "epoch": 1.1023417172593235, "grad_norm": 1.1668907403945923, "learning_rate": 1.2696938923545627e-05, "loss": 0.0242, "step": 10168 }, { "epoch": 1.1024501300954033, "grad_norm": 2.223766803741455, "learning_rate": 1.269621354997824e-05, "loss": 0.0234, "step": 10169 }, { "epoch": 1.102558542931483, "grad_norm": 0.1840181052684784, "learning_rate": 1.2695488176410853e-05, "loss": 0.0014, "step": 10170 }, { "epoch": 1.1026669557675628, "grad_norm": 0.7746832370758057, "learning_rate": 1.2694762802843466e-05, "loss": 0.0078, "step": 10171 }, { "epoch": 1.1027753686036426, "grad_norm": 0.6088758707046509, "learning_rate": 1.2694037429276078e-05, "loss": 0.0015, "step": 10172 }, { "epoch": 1.1028837814397225, "grad_norm": 1.9287258386611938, "learning_rate": 1.2693312055708692e-05, "loss": 0.0209, "step": 10173 }, { "epoch": 1.1029921942758023, "grad_norm": 0.07010968774557114, "learning_rate": 1.2692586682141304e-05, "loss": 0.0007, "step": 10174 }, { "epoch": 1.103100607111882, "grad_norm": 1.1016343832015991, "learning_rate": 1.2691861308573916e-05, "loss": 0.0429, "step": 10175 }, { "epoch": 1.1032090199479618, "grad_norm": 1.8925700187683105, "learning_rate": 1.2691135935006528e-05, "loss": 0.0394, "step": 10176 }, { "epoch": 1.1033174327840416, "grad_norm": 0.31392359733581543, "learning_rate": 1.2690410561439142e-05, "loss": 0.0063, "step": 10177 }, { "epoch": 1.1034258456201214, "grad_norm": 1.1379740238189697, "learning_rate": 1.2689685187871754e-05, "loss": 0.0063, "step": 10178 }, { "epoch": 1.1035342584562011, "grad_norm": 0.4595695734024048, "learning_rate": 1.2688959814304366e-05, "loss": 0.0153, "step": 10179 }, { "epoch": 1.103642671292281, "grad_norm": 3.6214470863342285, "learning_rate": 1.268823444073698e-05, "loss": 0.0228, "step": 10180 }, { "epoch": 1.1037510841283609, "grad_norm": 2.6078689098358154, "learning_rate": 1.2687509067169594e-05, "loss": 0.0162, "step": 10181 }, { "epoch": 1.1038594969644406, "grad_norm": 5.440478324890137, "learning_rate": 1.2686783693602208e-05, "loss": 0.0049, "step": 10182 }, { "epoch": 1.1039679098005204, "grad_norm": 0.328935444355011, "learning_rate": 1.268605832003482e-05, "loss": 0.0086, "step": 10183 }, { "epoch": 1.1040763226366002, "grad_norm": 14.489599227905273, "learning_rate": 1.2685332946467432e-05, "loss": 0.0401, "step": 10184 }, { "epoch": 1.10418473547268, "grad_norm": 1.1205227375030518, "learning_rate": 1.2684607572900046e-05, "loss": 0.0028, "step": 10185 }, { "epoch": 1.1042931483087597, "grad_norm": 8.338656425476074, "learning_rate": 1.2683882199332658e-05, "loss": 0.0257, "step": 10186 }, { "epoch": 1.1044015611448394, "grad_norm": 0.04802761599421501, "learning_rate": 1.268315682576527e-05, "loss": 0.0005, "step": 10187 }, { "epoch": 1.1045099739809194, "grad_norm": 1.5735841989517212, "learning_rate": 1.2682431452197884e-05, "loss": 0.0034, "step": 10188 }, { "epoch": 1.1046183868169992, "grad_norm": 4.371750831604004, "learning_rate": 1.2681706078630496e-05, "loss": 0.0433, "step": 10189 }, { "epoch": 1.104726799653079, "grad_norm": 2.233551263809204, "learning_rate": 1.2680980705063108e-05, "loss": 0.0223, "step": 10190 }, { "epoch": 1.1048352124891587, "grad_norm": 2.9351367950439453, "learning_rate": 1.268025533149572e-05, "loss": 0.0292, "step": 10191 }, { "epoch": 1.1049436253252385, "grad_norm": 1.3448765277862549, "learning_rate": 1.2679529957928334e-05, "loss": 0.0278, "step": 10192 }, { "epoch": 1.1050520381613183, "grad_norm": 1.833247423171997, "learning_rate": 1.2678804584360946e-05, "loss": 0.0217, "step": 10193 }, { "epoch": 1.105160450997398, "grad_norm": 1.101340413093567, "learning_rate": 1.2678079210793558e-05, "loss": 0.0071, "step": 10194 }, { "epoch": 1.1052688638334778, "grad_norm": 2.1659739017486572, "learning_rate": 1.2677353837226172e-05, "loss": 0.0632, "step": 10195 }, { "epoch": 1.1053772766695578, "grad_norm": 0.4085281193256378, "learning_rate": 1.2676628463658784e-05, "loss": 0.0098, "step": 10196 }, { "epoch": 1.1054856895056375, "grad_norm": 1.8403617143630981, "learning_rate": 1.2675903090091397e-05, "loss": 0.0352, "step": 10197 }, { "epoch": 1.1055941023417173, "grad_norm": 1.6690516471862793, "learning_rate": 1.2675177716524012e-05, "loss": 0.0456, "step": 10198 }, { "epoch": 1.105702515177797, "grad_norm": 1.3309595584869385, "learning_rate": 1.2674452342956624e-05, "loss": 0.0123, "step": 10199 }, { "epoch": 1.1058109280138768, "grad_norm": 0.4463779926300049, "learning_rate": 1.2673726969389238e-05, "loss": 0.0064, "step": 10200 }, { "epoch": 1.1059193408499566, "grad_norm": 2.5923430919647217, "learning_rate": 1.267300159582185e-05, "loss": 0.0095, "step": 10201 }, { "epoch": 1.1060277536860363, "grad_norm": 1.7677421569824219, "learning_rate": 1.2672276222254462e-05, "loss": 0.0245, "step": 10202 }, { "epoch": 1.1061361665221163, "grad_norm": 2.761086940765381, "learning_rate": 1.2671550848687075e-05, "loss": 0.0206, "step": 10203 }, { "epoch": 1.106244579358196, "grad_norm": 0.452732652425766, "learning_rate": 1.2670825475119688e-05, "loss": 0.009, "step": 10204 }, { "epoch": 1.1063529921942759, "grad_norm": 0.155154287815094, "learning_rate": 1.26701001015523e-05, "loss": 0.002, "step": 10205 }, { "epoch": 1.1064614050303556, "grad_norm": 2.162461042404175, "learning_rate": 1.2669374727984913e-05, "loss": 0.015, "step": 10206 }, { "epoch": 1.1065698178664354, "grad_norm": 0.7195900082588196, "learning_rate": 1.2668649354417526e-05, "loss": 0.0214, "step": 10207 }, { "epoch": 1.1066782307025151, "grad_norm": 0.5041993856430054, "learning_rate": 1.2667923980850139e-05, "loss": 0.0213, "step": 10208 }, { "epoch": 1.106786643538595, "grad_norm": 4.846608638763428, "learning_rate": 1.266719860728275e-05, "loss": 0.0445, "step": 10209 }, { "epoch": 1.1068950563746747, "grad_norm": 1.5086936950683594, "learning_rate": 1.2666473233715365e-05, "loss": 0.0183, "step": 10210 }, { "epoch": 1.1070034692107547, "grad_norm": 5.427215099334717, "learning_rate": 1.2665747860147977e-05, "loss": 0.0332, "step": 10211 }, { "epoch": 1.1071118820468344, "grad_norm": 1.6817816495895386, "learning_rate": 1.2665022486580589e-05, "loss": 0.0053, "step": 10212 }, { "epoch": 1.1072202948829142, "grad_norm": 0.472013384103775, "learning_rate": 1.2664297113013203e-05, "loss": 0.0091, "step": 10213 }, { "epoch": 1.107328707718994, "grad_norm": 0.7171529531478882, "learning_rate": 1.2663571739445815e-05, "loss": 0.0132, "step": 10214 }, { "epoch": 1.1074371205550737, "grad_norm": 0.14293272793293, "learning_rate": 1.2662846365878429e-05, "loss": 0.0009, "step": 10215 }, { "epoch": 1.1075455333911535, "grad_norm": 0.6655650734901428, "learning_rate": 1.2662120992311043e-05, "loss": 0.0125, "step": 10216 }, { "epoch": 1.1076539462272332, "grad_norm": 11.666810989379883, "learning_rate": 1.2661395618743655e-05, "loss": 0.0498, "step": 10217 }, { "epoch": 1.107762359063313, "grad_norm": 0.36364510655403137, "learning_rate": 1.2660670245176267e-05, "loss": 0.0068, "step": 10218 }, { "epoch": 1.107870771899393, "grad_norm": 3.209585189819336, "learning_rate": 1.265994487160888e-05, "loss": 0.0158, "step": 10219 }, { "epoch": 1.1079791847354727, "grad_norm": 0.47487780451774597, "learning_rate": 1.2659219498041493e-05, "loss": 0.009, "step": 10220 }, { "epoch": 1.1080875975715525, "grad_norm": 0.27984172105789185, "learning_rate": 1.2658494124474105e-05, "loss": 0.003, "step": 10221 }, { "epoch": 1.1081960104076323, "grad_norm": 1.7580064535140991, "learning_rate": 1.2657768750906719e-05, "loss": 0.0105, "step": 10222 }, { "epoch": 1.108304423243712, "grad_norm": 1.6100971698760986, "learning_rate": 1.2657043377339331e-05, "loss": 0.015, "step": 10223 }, { "epoch": 1.1084128360797918, "grad_norm": 8.649822235107422, "learning_rate": 1.2656318003771943e-05, "loss": 0.0388, "step": 10224 }, { "epoch": 1.1085212489158716, "grad_norm": 4.248824119567871, "learning_rate": 1.2655592630204557e-05, "loss": 0.0407, "step": 10225 }, { "epoch": 1.1086296617519515, "grad_norm": 0.5286899209022522, "learning_rate": 1.2654867256637169e-05, "loss": 0.0095, "step": 10226 }, { "epoch": 1.1087380745880313, "grad_norm": 8.351020812988281, "learning_rate": 1.2654141883069781e-05, "loss": 0.0237, "step": 10227 }, { "epoch": 1.108846487424111, "grad_norm": 8.160548210144043, "learning_rate": 1.2653416509502393e-05, "loss": 0.0517, "step": 10228 }, { "epoch": 1.1089549002601908, "grad_norm": 11.297676086425781, "learning_rate": 1.2652691135935007e-05, "loss": 0.0394, "step": 10229 }, { "epoch": 1.1090633130962706, "grad_norm": 1.9734835624694824, "learning_rate": 1.265196576236762e-05, "loss": 0.0097, "step": 10230 }, { "epoch": 1.1091717259323504, "grad_norm": 7.789968013763428, "learning_rate": 1.2651240388800231e-05, "loss": 0.0434, "step": 10231 }, { "epoch": 1.1092801387684301, "grad_norm": 2.2494089603424072, "learning_rate": 1.2650515015232845e-05, "loss": 0.0064, "step": 10232 }, { "epoch": 1.1093885516045099, "grad_norm": 1.0414018630981445, "learning_rate": 1.264978964166546e-05, "loss": 0.0117, "step": 10233 }, { "epoch": 1.1094969644405897, "grad_norm": 0.4354168176651001, "learning_rate": 1.2649064268098073e-05, "loss": 0.0074, "step": 10234 }, { "epoch": 1.1096053772766696, "grad_norm": 4.162149429321289, "learning_rate": 1.2648338894530685e-05, "loss": 0.065, "step": 10235 }, { "epoch": 1.1097137901127494, "grad_norm": 0.9983661770820618, "learning_rate": 1.2647613520963297e-05, "loss": 0.0157, "step": 10236 }, { "epoch": 1.1098222029488292, "grad_norm": 3.5310096740722656, "learning_rate": 1.2646888147395911e-05, "loss": 0.0424, "step": 10237 }, { "epoch": 1.109930615784909, "grad_norm": 4.144296169281006, "learning_rate": 1.2646162773828523e-05, "loss": 0.0589, "step": 10238 }, { "epoch": 1.1100390286209887, "grad_norm": 3.767232894897461, "learning_rate": 1.2645437400261135e-05, "loss": 0.0389, "step": 10239 }, { "epoch": 1.1101474414570685, "grad_norm": 3.8790783882141113, "learning_rate": 1.264471202669375e-05, "loss": 0.0281, "step": 10240 }, { "epoch": 1.1102558542931482, "grad_norm": 0.3320175111293793, "learning_rate": 1.2643986653126361e-05, "loss": 0.0068, "step": 10241 }, { "epoch": 1.1103642671292282, "grad_norm": 2.1666688919067383, "learning_rate": 1.2643261279558974e-05, "loss": 0.0076, "step": 10242 }, { "epoch": 1.110472679965308, "grad_norm": 0.7255966067314148, "learning_rate": 1.2642535905991586e-05, "loss": 0.024, "step": 10243 }, { "epoch": 1.1105810928013877, "grad_norm": 9.92307186126709, "learning_rate": 1.26418105324242e-05, "loss": 0.0362, "step": 10244 }, { "epoch": 1.1106895056374675, "grad_norm": 0.4310791492462158, "learning_rate": 1.2641085158856812e-05, "loss": 0.0076, "step": 10245 }, { "epoch": 1.1107979184735473, "grad_norm": 2.3412816524505615, "learning_rate": 1.2640359785289424e-05, "loss": 0.0718, "step": 10246 }, { "epoch": 1.110906331309627, "grad_norm": 8.866535186767578, "learning_rate": 1.2639634411722038e-05, "loss": 0.0319, "step": 10247 }, { "epoch": 1.1110147441457068, "grad_norm": 3.7920191287994385, "learning_rate": 1.263890903815465e-05, "loss": 0.0693, "step": 10248 }, { "epoch": 1.1111231569817865, "grad_norm": 0.6355178356170654, "learning_rate": 1.2638183664587262e-05, "loss": 0.0183, "step": 10249 }, { "epoch": 1.1112315698178665, "grad_norm": 0.9483978152275085, "learning_rate": 1.2637458291019877e-05, "loss": 0.0094, "step": 10250 }, { "epoch": 1.1113399826539463, "grad_norm": 1.8324017524719238, "learning_rate": 1.263673291745249e-05, "loss": 0.042, "step": 10251 }, { "epoch": 1.111448395490026, "grad_norm": 0.587326169013977, "learning_rate": 1.2636007543885103e-05, "loss": 0.017, "step": 10252 }, { "epoch": 1.1115568083261058, "grad_norm": 0.20300033688545227, "learning_rate": 1.2635282170317716e-05, "loss": 0.0009, "step": 10253 }, { "epoch": 1.1116652211621856, "grad_norm": 3.8878886699676514, "learning_rate": 1.2634556796750328e-05, "loss": 0.0223, "step": 10254 }, { "epoch": 1.1117736339982653, "grad_norm": 0.2824340760707855, "learning_rate": 1.263383142318294e-05, "loss": 0.0015, "step": 10255 }, { "epoch": 1.111882046834345, "grad_norm": 3.7081902027130127, "learning_rate": 1.2633106049615554e-05, "loss": 0.0482, "step": 10256 }, { "epoch": 1.1119904596704249, "grad_norm": 0.6778024435043335, "learning_rate": 1.2632380676048166e-05, "loss": 0.0219, "step": 10257 }, { "epoch": 1.1120988725065049, "grad_norm": 4.028826713562012, "learning_rate": 1.2631655302480778e-05, "loss": 0.0526, "step": 10258 }, { "epoch": 1.1122072853425846, "grad_norm": 2.1450371742248535, "learning_rate": 1.2630929928913392e-05, "loss": 0.0622, "step": 10259 }, { "epoch": 1.1123156981786644, "grad_norm": 11.523038864135742, "learning_rate": 1.2630204555346004e-05, "loss": 0.0214, "step": 10260 }, { "epoch": 1.1124241110147441, "grad_norm": 1.2785080671310425, "learning_rate": 1.2629479181778616e-05, "loss": 0.0036, "step": 10261 }, { "epoch": 1.112532523850824, "grad_norm": 2.847346782684326, "learning_rate": 1.262875380821123e-05, "loss": 0.0245, "step": 10262 }, { "epoch": 1.1126409366869037, "grad_norm": 0.12993504106998444, "learning_rate": 1.2628028434643842e-05, "loss": 0.001, "step": 10263 }, { "epoch": 1.1127493495229834, "grad_norm": 0.26191261410713196, "learning_rate": 1.2627303061076454e-05, "loss": 0.0022, "step": 10264 }, { "epoch": 1.1128577623590634, "grad_norm": 0.28195610642433167, "learning_rate": 1.2626577687509068e-05, "loss": 0.006, "step": 10265 }, { "epoch": 1.1129661751951432, "grad_norm": 0.6139281392097473, "learning_rate": 1.262585231394168e-05, "loss": 0.0011, "step": 10266 }, { "epoch": 1.113074588031223, "grad_norm": 3.3094851970672607, "learning_rate": 1.2625126940374294e-05, "loss": 0.0194, "step": 10267 }, { "epoch": 1.1131830008673027, "grad_norm": 2.4506819248199463, "learning_rate": 1.2624401566806908e-05, "loss": 0.0222, "step": 10268 }, { "epoch": 1.1132914137033825, "grad_norm": 1.9161567687988281, "learning_rate": 1.262367619323952e-05, "loss": 0.0258, "step": 10269 }, { "epoch": 1.1133998265394622, "grad_norm": 5.9394707679748535, "learning_rate": 1.2622950819672132e-05, "loss": 0.0444, "step": 10270 }, { "epoch": 1.113508239375542, "grad_norm": 4.963822841644287, "learning_rate": 1.2622225446104746e-05, "loss": 0.0862, "step": 10271 }, { "epoch": 1.1136166522116218, "grad_norm": 2.9067251682281494, "learning_rate": 1.2621500072537358e-05, "loss": 0.0368, "step": 10272 }, { "epoch": 1.1137250650477017, "grad_norm": 1.9569422006607056, "learning_rate": 1.262077469896997e-05, "loss": 0.0144, "step": 10273 }, { "epoch": 1.1138334778837815, "grad_norm": 1.3907582759857178, "learning_rate": 1.2620049325402584e-05, "loss": 0.0061, "step": 10274 }, { "epoch": 1.1139418907198613, "grad_norm": 6.590664386749268, "learning_rate": 1.2619323951835196e-05, "loss": 0.0186, "step": 10275 }, { "epoch": 1.114050303555941, "grad_norm": 2.9117801189422607, "learning_rate": 1.2618598578267808e-05, "loss": 0.0086, "step": 10276 }, { "epoch": 1.1141587163920208, "grad_norm": 0.8895557522773743, "learning_rate": 1.2617873204700422e-05, "loss": 0.0279, "step": 10277 }, { "epoch": 1.1142671292281006, "grad_norm": 0.453035444021225, "learning_rate": 1.2617147831133034e-05, "loss": 0.0123, "step": 10278 }, { "epoch": 1.1143755420641803, "grad_norm": 1.780776858329773, "learning_rate": 1.2616422457565647e-05, "loss": 0.0281, "step": 10279 }, { "epoch": 1.11448395490026, "grad_norm": 1.2408047914505005, "learning_rate": 1.2615697083998259e-05, "loss": 0.0148, "step": 10280 }, { "epoch": 1.11459236773634, "grad_norm": 0.3579881191253662, "learning_rate": 1.2614971710430873e-05, "loss": 0.0108, "step": 10281 }, { "epoch": 1.1147007805724198, "grad_norm": 4.250640869140625, "learning_rate": 1.2614246336863485e-05, "loss": 0.0563, "step": 10282 }, { "epoch": 1.1148091934084996, "grad_norm": 0.18385832011699677, "learning_rate": 1.2613520963296097e-05, "loss": 0.0018, "step": 10283 }, { "epoch": 1.1149176062445794, "grad_norm": 1.1253374814987183, "learning_rate": 1.261279558972871e-05, "loss": 0.0016, "step": 10284 }, { "epoch": 1.1150260190806591, "grad_norm": 4.548178195953369, "learning_rate": 1.2612070216161324e-05, "loss": 0.0126, "step": 10285 }, { "epoch": 1.1151344319167389, "grad_norm": 2.281508684158325, "learning_rate": 1.2611344842593938e-05, "loss": 0.044, "step": 10286 }, { "epoch": 1.1152428447528187, "grad_norm": 0.13692188262939453, "learning_rate": 1.261061946902655e-05, "loss": 0.001, "step": 10287 }, { "epoch": 1.1153512575888986, "grad_norm": 1.185463547706604, "learning_rate": 1.2609894095459163e-05, "loss": 0.0183, "step": 10288 }, { "epoch": 1.1154596704249784, "grad_norm": 0.32568275928497314, "learning_rate": 1.2609168721891776e-05, "loss": 0.0077, "step": 10289 }, { "epoch": 1.1155680832610582, "grad_norm": 0.1491917222738266, "learning_rate": 1.2608443348324389e-05, "loss": 0.0004, "step": 10290 }, { "epoch": 1.115676496097138, "grad_norm": 0.9688519835472107, "learning_rate": 1.2607717974757e-05, "loss": 0.015, "step": 10291 }, { "epoch": 1.1157849089332177, "grad_norm": 1.1741039752960205, "learning_rate": 1.2606992601189615e-05, "loss": 0.0206, "step": 10292 }, { "epoch": 1.1158933217692975, "grad_norm": 1.0148769617080688, "learning_rate": 1.2606267227622227e-05, "loss": 0.0038, "step": 10293 }, { "epoch": 1.1160017346053772, "grad_norm": 0.3164820373058319, "learning_rate": 1.2605541854054839e-05, "loss": 0.0021, "step": 10294 }, { "epoch": 1.116110147441457, "grad_norm": 0.13153716921806335, "learning_rate": 1.2604816480487451e-05, "loss": 0.001, "step": 10295 }, { "epoch": 1.116218560277537, "grad_norm": 3.4680473804473877, "learning_rate": 1.2604091106920065e-05, "loss": 0.0528, "step": 10296 }, { "epoch": 1.1163269731136167, "grad_norm": 1.917960524559021, "learning_rate": 1.2603365733352677e-05, "loss": 0.0282, "step": 10297 }, { "epoch": 1.1164353859496965, "grad_norm": 2.1210343837738037, "learning_rate": 1.2602640359785289e-05, "loss": 0.0053, "step": 10298 }, { "epoch": 1.1165437987857763, "grad_norm": 4.141436576843262, "learning_rate": 1.2601914986217903e-05, "loss": 0.0272, "step": 10299 }, { "epoch": 1.116652211621856, "grad_norm": 4.34027624130249, "learning_rate": 1.2601189612650515e-05, "loss": 0.0231, "step": 10300 }, { "epoch": 1.1167606244579358, "grad_norm": 8.654833793640137, "learning_rate": 1.2600464239083127e-05, "loss": 0.0173, "step": 10301 }, { "epoch": 1.1168690372940155, "grad_norm": 2.1254537105560303, "learning_rate": 1.2599738865515743e-05, "loss": 0.045, "step": 10302 }, { "epoch": 1.1169774501300953, "grad_norm": 2.684269666671753, "learning_rate": 1.2599013491948355e-05, "loss": 0.0803, "step": 10303 }, { "epoch": 1.1170858629661753, "grad_norm": 2.748960256576538, "learning_rate": 1.2598288118380969e-05, "loss": 0.0666, "step": 10304 }, { "epoch": 1.117194275802255, "grad_norm": 2.2635936737060547, "learning_rate": 1.2597562744813581e-05, "loss": 0.0281, "step": 10305 }, { "epoch": 1.1173026886383348, "grad_norm": 1.739843726158142, "learning_rate": 1.2596837371246193e-05, "loss": 0.0256, "step": 10306 }, { "epoch": 1.1174111014744146, "grad_norm": 8.623891830444336, "learning_rate": 1.2596111997678805e-05, "loss": 0.0562, "step": 10307 }, { "epoch": 1.1175195143104943, "grad_norm": 1.0063914060592651, "learning_rate": 1.2595386624111419e-05, "loss": 0.0018, "step": 10308 }, { "epoch": 1.117627927146574, "grad_norm": 1.372256875038147, "learning_rate": 1.2594661250544031e-05, "loss": 0.0213, "step": 10309 }, { "epoch": 1.1177363399826539, "grad_norm": 0.4630522131919861, "learning_rate": 1.2593935876976643e-05, "loss": 0.0188, "step": 10310 }, { "epoch": 1.1178447528187339, "grad_norm": 2.5910415649414062, "learning_rate": 1.2593210503409257e-05, "loss": 0.0282, "step": 10311 }, { "epoch": 1.1179531656548136, "grad_norm": 0.1852601021528244, "learning_rate": 1.259248512984187e-05, "loss": 0.0014, "step": 10312 }, { "epoch": 1.1180615784908934, "grad_norm": 1.76344895362854, "learning_rate": 1.2591759756274481e-05, "loss": 0.0171, "step": 10313 }, { "epoch": 1.1181699913269731, "grad_norm": 6.519124507904053, "learning_rate": 1.2591034382707095e-05, "loss": 0.0149, "step": 10314 }, { "epoch": 1.118278404163053, "grad_norm": 1.6868172883987427, "learning_rate": 1.2590309009139707e-05, "loss": 0.0396, "step": 10315 }, { "epoch": 1.1183868169991327, "grad_norm": 1.0184909105300903, "learning_rate": 1.258958363557232e-05, "loss": 0.0205, "step": 10316 }, { "epoch": 1.1184952298352124, "grad_norm": 1.0907717943191528, "learning_rate": 1.2588858262004933e-05, "loss": 0.0186, "step": 10317 }, { "epoch": 1.1186036426712922, "grad_norm": 1.3289682865142822, "learning_rate": 1.2588132888437546e-05, "loss": 0.0464, "step": 10318 }, { "epoch": 1.118712055507372, "grad_norm": 0.20965750515460968, "learning_rate": 1.258740751487016e-05, "loss": 0.0017, "step": 10319 }, { "epoch": 1.118820468343452, "grad_norm": 2.1585049629211426, "learning_rate": 1.2586682141302773e-05, "loss": 0.0336, "step": 10320 }, { "epoch": 1.1189288811795317, "grad_norm": 0.3621286451816559, "learning_rate": 1.2585956767735385e-05, "loss": 0.0089, "step": 10321 }, { "epoch": 1.1190372940156115, "grad_norm": 0.1580570489168167, "learning_rate": 1.2585231394167998e-05, "loss": 0.0115, "step": 10322 }, { "epoch": 1.1191457068516912, "grad_norm": 0.46779727935791016, "learning_rate": 1.2584506020600611e-05, "loss": 0.0064, "step": 10323 }, { "epoch": 1.119254119687771, "grad_norm": 1.4514744281768799, "learning_rate": 1.2583780647033223e-05, "loss": 0.022, "step": 10324 }, { "epoch": 1.1193625325238508, "grad_norm": 0.3812352418899536, "learning_rate": 1.2583055273465836e-05, "loss": 0.0183, "step": 10325 }, { "epoch": 1.1194709453599305, "grad_norm": 0.29805320501327515, "learning_rate": 1.258232989989845e-05, "loss": 0.0096, "step": 10326 }, { "epoch": 1.1195793581960105, "grad_norm": 3.002350330352783, "learning_rate": 1.2581604526331062e-05, "loss": 0.0203, "step": 10327 }, { "epoch": 1.1196877710320903, "grad_norm": 1.0144257545471191, "learning_rate": 1.2580879152763674e-05, "loss": 0.0166, "step": 10328 }, { "epoch": 1.11979618386817, "grad_norm": 1.3438340425491333, "learning_rate": 1.2580153779196288e-05, "loss": 0.0055, "step": 10329 }, { "epoch": 1.1199045967042498, "grad_norm": 0.3667878210544586, "learning_rate": 1.25794284056289e-05, "loss": 0.0095, "step": 10330 }, { "epoch": 1.1200130095403296, "grad_norm": 0.5365878939628601, "learning_rate": 1.2578703032061512e-05, "loss": 0.0123, "step": 10331 }, { "epoch": 1.1201214223764093, "grad_norm": 0.09420301020145416, "learning_rate": 1.2577977658494126e-05, "loss": 0.0006, "step": 10332 }, { "epoch": 1.120229835212489, "grad_norm": 0.3832864463329315, "learning_rate": 1.2577252284926738e-05, "loss": 0.0125, "step": 10333 }, { "epoch": 1.1203382480485689, "grad_norm": 1.283463716506958, "learning_rate": 1.257652691135935e-05, "loss": 0.0139, "step": 10334 }, { "epoch": 1.1204466608846488, "grad_norm": 2.1743826866149902, "learning_rate": 1.2575801537791962e-05, "loss": 0.0384, "step": 10335 }, { "epoch": 1.1205550737207286, "grad_norm": 2.3534507751464844, "learning_rate": 1.2575076164224578e-05, "loss": 0.0093, "step": 10336 }, { "epoch": 1.1206634865568084, "grad_norm": 5.496932029724121, "learning_rate": 1.257435079065719e-05, "loss": 0.0173, "step": 10337 }, { "epoch": 1.1207718993928881, "grad_norm": 0.2626018524169922, "learning_rate": 1.2573625417089804e-05, "loss": 0.0028, "step": 10338 }, { "epoch": 1.120880312228968, "grad_norm": 1.0500097274780273, "learning_rate": 1.2572900043522416e-05, "loss": 0.0029, "step": 10339 }, { "epoch": 1.1209887250650477, "grad_norm": 0.3182556927204132, "learning_rate": 1.2572174669955028e-05, "loss": 0.0008, "step": 10340 }, { "epoch": 1.1210971379011274, "grad_norm": 0.33632728457450867, "learning_rate": 1.2571449296387642e-05, "loss": 0.0116, "step": 10341 }, { "epoch": 1.1212055507372072, "grad_norm": 1.976782202720642, "learning_rate": 1.2570723922820254e-05, "loss": 0.0197, "step": 10342 }, { "epoch": 1.1213139635732872, "grad_norm": 2.0191099643707275, "learning_rate": 1.2569998549252866e-05, "loss": 0.0415, "step": 10343 }, { "epoch": 1.121422376409367, "grad_norm": 1.3600658178329468, "learning_rate": 1.256927317568548e-05, "loss": 0.0854, "step": 10344 }, { "epoch": 1.1215307892454467, "grad_norm": 11.001748085021973, "learning_rate": 1.2568547802118092e-05, "loss": 0.0709, "step": 10345 }, { "epoch": 1.1216392020815265, "grad_norm": 0.7274342775344849, "learning_rate": 1.2567822428550704e-05, "loss": 0.0246, "step": 10346 }, { "epoch": 1.1217476149176062, "grad_norm": 0.8175974488258362, "learning_rate": 1.2567097054983316e-05, "loss": 0.0159, "step": 10347 }, { "epoch": 1.121856027753686, "grad_norm": 0.39581218361854553, "learning_rate": 1.256637168141593e-05, "loss": 0.0035, "step": 10348 }, { "epoch": 1.1219644405897657, "grad_norm": 2.494074583053589, "learning_rate": 1.2565646307848542e-05, "loss": 0.0244, "step": 10349 }, { "epoch": 1.1220728534258457, "grad_norm": 6.0434393882751465, "learning_rate": 1.2564920934281154e-05, "loss": 0.0229, "step": 10350 }, { "epoch": 1.1221812662619255, "grad_norm": 0.2827724814414978, "learning_rate": 1.2564195560713768e-05, "loss": 0.0065, "step": 10351 }, { "epoch": 1.1222896790980053, "grad_norm": 2.0097837448120117, "learning_rate": 1.256347018714638e-05, "loss": 0.0528, "step": 10352 }, { "epoch": 1.122398091934085, "grad_norm": 2.559936285018921, "learning_rate": 1.2562744813578993e-05, "loss": 0.0251, "step": 10353 }, { "epoch": 1.1225065047701648, "grad_norm": 0.43180927634239197, "learning_rate": 1.2562019440011608e-05, "loss": 0.0242, "step": 10354 }, { "epoch": 1.1226149176062445, "grad_norm": 1.2035435438156128, "learning_rate": 1.256129406644422e-05, "loss": 0.0061, "step": 10355 }, { "epoch": 1.1227233304423243, "grad_norm": 4.8855509757995605, "learning_rate": 1.2560568692876834e-05, "loss": 0.0317, "step": 10356 }, { "epoch": 1.122831743278404, "grad_norm": 1.371067762374878, "learning_rate": 1.2559843319309446e-05, "loss": 0.0184, "step": 10357 }, { "epoch": 1.122940156114484, "grad_norm": 2.1799371242523193, "learning_rate": 1.2559117945742058e-05, "loss": 0.0218, "step": 10358 }, { "epoch": 1.1230485689505638, "grad_norm": 0.52211594581604, "learning_rate": 1.255839257217467e-05, "loss": 0.0175, "step": 10359 }, { "epoch": 1.1231569817866436, "grad_norm": 0.5379668474197388, "learning_rate": 1.2557667198607284e-05, "loss": 0.0042, "step": 10360 }, { "epoch": 1.1232653946227233, "grad_norm": 0.3426607549190521, "learning_rate": 1.2556941825039897e-05, "loss": 0.0019, "step": 10361 }, { "epoch": 1.123373807458803, "grad_norm": 2.2276947498321533, "learning_rate": 1.2556216451472509e-05, "loss": 0.0315, "step": 10362 }, { "epoch": 1.1234822202948829, "grad_norm": 0.039823174476623535, "learning_rate": 1.2555491077905123e-05, "loss": 0.0004, "step": 10363 }, { "epoch": 1.1235906331309626, "grad_norm": 3.785975933074951, "learning_rate": 1.2554765704337735e-05, "loss": 0.0825, "step": 10364 }, { "epoch": 1.1236990459670424, "grad_norm": 0.5519446730613708, "learning_rate": 1.2554040330770347e-05, "loss": 0.0087, "step": 10365 }, { "epoch": 1.1238074588031224, "grad_norm": 2.682020425796509, "learning_rate": 1.255331495720296e-05, "loss": 0.0124, "step": 10366 }, { "epoch": 1.1239158716392021, "grad_norm": 2.6835498809814453, "learning_rate": 1.2552589583635573e-05, "loss": 0.0114, "step": 10367 }, { "epoch": 1.124024284475282, "grad_norm": 3.2358975410461426, "learning_rate": 1.2551864210068185e-05, "loss": 0.0203, "step": 10368 }, { "epoch": 1.1241326973113617, "grad_norm": 0.9066609144210815, "learning_rate": 1.2551138836500799e-05, "loss": 0.0109, "step": 10369 }, { "epoch": 1.1242411101474414, "grad_norm": 3.09002423286438, "learning_rate": 1.2550413462933411e-05, "loss": 0.0414, "step": 10370 }, { "epoch": 1.1243495229835212, "grad_norm": 0.27879539132118225, "learning_rate": 1.2549688089366025e-05, "loss": 0.0071, "step": 10371 }, { "epoch": 1.124457935819601, "grad_norm": 5.472727298736572, "learning_rate": 1.2548962715798639e-05, "loss": 0.0672, "step": 10372 }, { "epoch": 1.124566348655681, "grad_norm": 2.825174331665039, "learning_rate": 1.254823734223125e-05, "loss": 0.0083, "step": 10373 }, { "epoch": 1.1246747614917607, "grad_norm": 4.480884075164795, "learning_rate": 1.2547511968663863e-05, "loss": 0.0227, "step": 10374 }, { "epoch": 1.1247831743278405, "grad_norm": 0.27752575278282166, "learning_rate": 1.2546786595096477e-05, "loss": 0.0069, "step": 10375 }, { "epoch": 1.1248915871639202, "grad_norm": 0.34554269909858704, "learning_rate": 1.2546061221529089e-05, "loss": 0.005, "step": 10376 }, { "epoch": 1.125, "grad_norm": 22.307762145996094, "learning_rate": 1.2545335847961701e-05, "loss": 0.0307, "step": 10377 }, { "epoch": 1.1251084128360798, "grad_norm": 0.9442766308784485, "learning_rate": 1.2544610474394315e-05, "loss": 0.0243, "step": 10378 }, { "epoch": 1.1252168256721595, "grad_norm": 5.057775020599365, "learning_rate": 1.2543885100826927e-05, "loss": 0.0251, "step": 10379 }, { "epoch": 1.1253252385082393, "grad_norm": 2.6488711833953857, "learning_rate": 1.2543159727259539e-05, "loss": 0.0086, "step": 10380 }, { "epoch": 1.125433651344319, "grad_norm": 4.881179332733154, "learning_rate": 1.2542434353692153e-05, "loss": 0.0209, "step": 10381 }, { "epoch": 1.125542064180399, "grad_norm": 3.6033873558044434, "learning_rate": 1.2541708980124765e-05, "loss": 0.0408, "step": 10382 }, { "epoch": 1.1256504770164788, "grad_norm": 4.117387771606445, "learning_rate": 1.2540983606557377e-05, "loss": 0.0339, "step": 10383 }, { "epoch": 1.1257588898525586, "grad_norm": 0.4082711338996887, "learning_rate": 1.2540258232989991e-05, "loss": 0.014, "step": 10384 }, { "epoch": 1.1258673026886383, "grad_norm": 0.13055747747421265, "learning_rate": 1.2539532859422603e-05, "loss": 0.0012, "step": 10385 }, { "epoch": 1.125975715524718, "grad_norm": 0.44970813393592834, "learning_rate": 1.2538807485855215e-05, "loss": 0.0064, "step": 10386 }, { "epoch": 1.1260841283607979, "grad_norm": 7.409921169281006, "learning_rate": 1.2538082112287827e-05, "loss": 0.0338, "step": 10387 }, { "epoch": 1.1261925411968776, "grad_norm": 1.5426515340805054, "learning_rate": 1.2537356738720443e-05, "loss": 0.0131, "step": 10388 }, { "epoch": 1.1263009540329576, "grad_norm": 9.90496826171875, "learning_rate": 1.2536631365153055e-05, "loss": 0.0833, "step": 10389 }, { "epoch": 1.1264093668690374, "grad_norm": 1.858407974243164, "learning_rate": 1.2535905991585669e-05, "loss": 0.041, "step": 10390 }, { "epoch": 1.1265177797051171, "grad_norm": 1.341581106185913, "learning_rate": 1.2535180618018281e-05, "loss": 0.0184, "step": 10391 }, { "epoch": 1.126626192541197, "grad_norm": 2.886662006378174, "learning_rate": 1.2534455244450893e-05, "loss": 0.026, "step": 10392 }, { "epoch": 1.1267346053772767, "grad_norm": 2.0695624351501465, "learning_rate": 1.2533729870883507e-05, "loss": 0.0143, "step": 10393 }, { "epoch": 1.1268430182133564, "grad_norm": 0.8602486252784729, "learning_rate": 1.253300449731612e-05, "loss": 0.0203, "step": 10394 }, { "epoch": 1.1269514310494362, "grad_norm": 0.566030740737915, "learning_rate": 1.2532279123748731e-05, "loss": 0.0039, "step": 10395 }, { "epoch": 1.1270598438855162, "grad_norm": 0.3026905655860901, "learning_rate": 1.2531553750181345e-05, "loss": 0.0164, "step": 10396 }, { "epoch": 1.127168256721596, "grad_norm": 3.7795426845550537, "learning_rate": 1.2530828376613957e-05, "loss": 0.0416, "step": 10397 }, { "epoch": 1.1272766695576757, "grad_norm": 0.4221295118331909, "learning_rate": 1.253010300304657e-05, "loss": 0.0132, "step": 10398 }, { "epoch": 1.1273850823937555, "grad_norm": 2.7044386863708496, "learning_rate": 1.2529377629479182e-05, "loss": 0.0794, "step": 10399 }, { "epoch": 1.1274934952298352, "grad_norm": 0.559027910232544, "learning_rate": 1.2528652255911796e-05, "loss": 0.0203, "step": 10400 }, { "epoch": 1.127601908065915, "grad_norm": 6.027595520019531, "learning_rate": 1.2527926882344408e-05, "loss": 0.0393, "step": 10401 }, { "epoch": 1.1277103209019947, "grad_norm": 1.1855428218841553, "learning_rate": 1.252720150877702e-05, "loss": 0.0225, "step": 10402 }, { "epoch": 1.1278187337380745, "grad_norm": 0.1312667727470398, "learning_rate": 1.2526476135209634e-05, "loss": 0.0012, "step": 10403 }, { "epoch": 1.1279271465741543, "grad_norm": 2.029597043991089, "learning_rate": 1.2525750761642246e-05, "loss": 0.0027, "step": 10404 }, { "epoch": 1.1280355594102343, "grad_norm": 0.34016481041908264, "learning_rate": 1.2525025388074861e-05, "loss": 0.0023, "step": 10405 }, { "epoch": 1.128143972246314, "grad_norm": 1.0747004747390747, "learning_rate": 1.2524300014507473e-05, "loss": 0.0596, "step": 10406 }, { "epoch": 1.1282523850823938, "grad_norm": 2.8191277980804443, "learning_rate": 1.2523574640940086e-05, "loss": 0.0187, "step": 10407 }, { "epoch": 1.1283607979184735, "grad_norm": 1.5668574571609497, "learning_rate": 1.25228492673727e-05, "loss": 0.0158, "step": 10408 }, { "epoch": 1.1284692107545533, "grad_norm": 0.974341094493866, "learning_rate": 1.2522123893805312e-05, "loss": 0.0115, "step": 10409 }, { "epoch": 1.128577623590633, "grad_norm": 5.659423828125, "learning_rate": 1.2521398520237924e-05, "loss": 0.005, "step": 10410 }, { "epoch": 1.1286860364267128, "grad_norm": 0.8851567506790161, "learning_rate": 1.2520673146670536e-05, "loss": 0.0144, "step": 10411 }, { "epoch": 1.1287944492627928, "grad_norm": 0.5225546360015869, "learning_rate": 1.251994777310315e-05, "loss": 0.0153, "step": 10412 }, { "epoch": 1.1289028620988726, "grad_norm": 0.9858400225639343, "learning_rate": 1.2519222399535762e-05, "loss": 0.0165, "step": 10413 }, { "epoch": 1.1290112749349523, "grad_norm": 4.119184494018555, "learning_rate": 1.2518497025968374e-05, "loss": 0.0225, "step": 10414 }, { "epoch": 1.129119687771032, "grad_norm": 2.1354966163635254, "learning_rate": 1.2517771652400988e-05, "loss": 0.0326, "step": 10415 }, { "epoch": 1.1292281006071119, "grad_norm": 3.6060147285461426, "learning_rate": 1.25170462788336e-05, "loss": 0.0074, "step": 10416 }, { "epoch": 1.1293365134431916, "grad_norm": 2.7923243045806885, "learning_rate": 1.2516320905266212e-05, "loss": 0.023, "step": 10417 }, { "epoch": 1.1294449262792714, "grad_norm": 3.458789825439453, "learning_rate": 1.2515595531698826e-05, "loss": 0.0054, "step": 10418 }, { "epoch": 1.1295533391153512, "grad_norm": 0.32770782709121704, "learning_rate": 1.2514870158131438e-05, "loss": 0.0089, "step": 10419 }, { "epoch": 1.129661751951431, "grad_norm": 2.5011022090911865, "learning_rate": 1.251414478456405e-05, "loss": 0.0201, "step": 10420 }, { "epoch": 1.129770164787511, "grad_norm": 4.29337215423584, "learning_rate": 1.2513419410996664e-05, "loss": 0.0442, "step": 10421 }, { "epoch": 1.1298785776235907, "grad_norm": 0.8786622285842896, "learning_rate": 1.2512694037429276e-05, "loss": 0.0192, "step": 10422 }, { "epoch": 1.1299869904596704, "grad_norm": 0.08503840118646622, "learning_rate": 1.251196866386189e-05, "loss": 0.0009, "step": 10423 }, { "epoch": 1.1300954032957502, "grad_norm": 2.1133158206939697, "learning_rate": 1.2511243290294504e-05, "loss": 0.0176, "step": 10424 }, { "epoch": 1.13020381613183, "grad_norm": 0.9814158082008362, "learning_rate": 1.2510517916727116e-05, "loss": 0.0064, "step": 10425 }, { "epoch": 1.1303122289679097, "grad_norm": 1.8237043619155884, "learning_rate": 1.2509792543159728e-05, "loss": 0.0173, "step": 10426 }, { "epoch": 1.1304206418039895, "grad_norm": 2.358771800994873, "learning_rate": 1.2509067169592342e-05, "loss": 0.0153, "step": 10427 }, { "epoch": 1.1305290546400695, "grad_norm": 1.9242796897888184, "learning_rate": 1.2508341796024954e-05, "loss": 0.0555, "step": 10428 }, { "epoch": 1.1306374674761492, "grad_norm": 3.558171272277832, "learning_rate": 1.2507616422457566e-05, "loss": 0.0252, "step": 10429 }, { "epoch": 1.130745880312229, "grad_norm": 3.582554817199707, "learning_rate": 1.250689104889018e-05, "loss": 0.0199, "step": 10430 }, { "epoch": 1.1308542931483088, "grad_norm": 0.6442551612854004, "learning_rate": 1.2506165675322792e-05, "loss": 0.0022, "step": 10431 }, { "epoch": 1.1309627059843885, "grad_norm": 1.4818068742752075, "learning_rate": 1.2505440301755404e-05, "loss": 0.0043, "step": 10432 }, { "epoch": 1.1310711188204683, "grad_norm": 0.6579238772392273, "learning_rate": 1.2504714928188018e-05, "loss": 0.0099, "step": 10433 }, { "epoch": 1.131179531656548, "grad_norm": 0.09682419151067734, "learning_rate": 1.250398955462063e-05, "loss": 0.0009, "step": 10434 }, { "epoch": 1.131287944492628, "grad_norm": 1.0421032905578613, "learning_rate": 1.2503264181053243e-05, "loss": 0.0056, "step": 10435 }, { "epoch": 1.1313963573287078, "grad_norm": 0.23384420573711395, "learning_rate": 1.2502538807485856e-05, "loss": 0.0064, "step": 10436 }, { "epoch": 1.1315047701647876, "grad_norm": 5.908698558807373, "learning_rate": 1.2501813433918469e-05, "loss": 0.0405, "step": 10437 }, { "epoch": 1.1316131830008673, "grad_norm": 2.843492031097412, "learning_rate": 1.250108806035108e-05, "loss": 0.0074, "step": 10438 }, { "epoch": 1.131721595836947, "grad_norm": 3.1270668506622314, "learning_rate": 1.2500362686783693e-05, "loss": 0.0387, "step": 10439 }, { "epoch": 1.1318300086730269, "grad_norm": 1.278210163116455, "learning_rate": 1.2499637313216308e-05, "loss": 0.0049, "step": 10440 }, { "epoch": 1.1319384215091066, "grad_norm": 0.3713797330856323, "learning_rate": 1.249891193964892e-05, "loss": 0.0074, "step": 10441 }, { "epoch": 1.1320468343451864, "grad_norm": 6.816887855529785, "learning_rate": 1.2498186566081534e-05, "loss": 0.0401, "step": 10442 }, { "epoch": 1.1321552471812661, "grad_norm": 0.9880887269973755, "learning_rate": 1.2497461192514146e-05, "loss": 0.0137, "step": 10443 }, { "epoch": 1.1322636600173461, "grad_norm": 3.455789566040039, "learning_rate": 1.2496735818946759e-05, "loss": 0.0277, "step": 10444 }, { "epoch": 1.132372072853426, "grad_norm": 0.649738609790802, "learning_rate": 1.2496010445379372e-05, "loss": 0.0172, "step": 10445 }, { "epoch": 1.1324804856895057, "grad_norm": 1.1698142290115356, "learning_rate": 1.2495285071811985e-05, "loss": 0.011, "step": 10446 }, { "epoch": 1.1325888985255854, "grad_norm": 0.48083439469337463, "learning_rate": 1.2494559698244597e-05, "loss": 0.004, "step": 10447 }, { "epoch": 1.1326973113616652, "grad_norm": 6.485860347747803, "learning_rate": 1.249383432467721e-05, "loss": 0.0175, "step": 10448 }, { "epoch": 1.132805724197745, "grad_norm": 1.3634297847747803, "learning_rate": 1.2493108951109823e-05, "loss": 0.0125, "step": 10449 }, { "epoch": 1.1329141370338247, "grad_norm": 3.455838918685913, "learning_rate": 1.2492383577542435e-05, "loss": 0.0099, "step": 10450 }, { "epoch": 1.1330225498699047, "grad_norm": 0.22714976966381073, "learning_rate": 1.2491658203975047e-05, "loss": 0.0042, "step": 10451 }, { "epoch": 1.1331309627059845, "grad_norm": 4.443262577056885, "learning_rate": 1.2490932830407661e-05, "loss": 0.0087, "step": 10452 }, { "epoch": 1.1332393755420642, "grad_norm": 0.4791901409626007, "learning_rate": 1.2490207456840273e-05, "loss": 0.0121, "step": 10453 }, { "epoch": 1.133347788378144, "grad_norm": 3.532335042953491, "learning_rate": 1.2489482083272885e-05, "loss": 0.0052, "step": 10454 }, { "epoch": 1.1334562012142237, "grad_norm": 0.7816373109817505, "learning_rate": 1.2488756709705499e-05, "loss": 0.0037, "step": 10455 }, { "epoch": 1.1335646140503035, "grad_norm": 0.5621979832649231, "learning_rate": 1.2488031336138111e-05, "loss": 0.0259, "step": 10456 }, { "epoch": 1.1336730268863833, "grad_norm": 0.31737443804740906, "learning_rate": 1.2487305962570727e-05, "loss": 0.0054, "step": 10457 }, { "epoch": 1.1337814397224633, "grad_norm": 11.031896591186523, "learning_rate": 1.2486580589003339e-05, "loss": 0.0163, "step": 10458 }, { "epoch": 1.133889852558543, "grad_norm": 1.383758783340454, "learning_rate": 1.2485855215435951e-05, "loss": 0.0177, "step": 10459 }, { "epoch": 1.1339982653946228, "grad_norm": 2.2999658584594727, "learning_rate": 1.2485129841868565e-05, "loss": 0.0652, "step": 10460 }, { "epoch": 1.1341066782307025, "grad_norm": 3.3439416885375977, "learning_rate": 1.2484404468301177e-05, "loss": 0.0222, "step": 10461 }, { "epoch": 1.1342150910667823, "grad_norm": 4.570712089538574, "learning_rate": 1.2483679094733789e-05, "loss": 0.0396, "step": 10462 }, { "epoch": 1.134323503902862, "grad_norm": 1.8791340589523315, "learning_rate": 1.2482953721166401e-05, "loss": 0.0604, "step": 10463 }, { "epoch": 1.1344319167389418, "grad_norm": 15.312854766845703, "learning_rate": 1.2482228347599015e-05, "loss": 0.0269, "step": 10464 }, { "epoch": 1.1345403295750216, "grad_norm": 4.875741004943848, "learning_rate": 1.2481502974031627e-05, "loss": 0.0073, "step": 10465 }, { "epoch": 1.1346487424111014, "grad_norm": 9.70179557800293, "learning_rate": 1.248077760046424e-05, "loss": 0.0306, "step": 10466 }, { "epoch": 1.1347571552471813, "grad_norm": 3.443135976791382, "learning_rate": 1.2480052226896853e-05, "loss": 0.0564, "step": 10467 }, { "epoch": 1.1348655680832611, "grad_norm": 0.5242893099784851, "learning_rate": 1.2479326853329465e-05, "loss": 0.017, "step": 10468 }, { "epoch": 1.1349739809193409, "grad_norm": 1.7528836727142334, "learning_rate": 1.2478601479762077e-05, "loss": 0.02, "step": 10469 }, { "epoch": 1.1350823937554206, "grad_norm": 2.697319269180298, "learning_rate": 1.2477876106194691e-05, "loss": 0.0287, "step": 10470 }, { "epoch": 1.1351908065915004, "grad_norm": 9.712557792663574, "learning_rate": 1.2477150732627303e-05, "loss": 0.0202, "step": 10471 }, { "epoch": 1.1352992194275802, "grad_norm": 3.5770363807678223, "learning_rate": 1.2476425359059916e-05, "loss": 0.0311, "step": 10472 }, { "epoch": 1.13540763226366, "grad_norm": 1.6564860343933105, "learning_rate": 1.247569998549253e-05, "loss": 0.0261, "step": 10473 }, { "epoch": 1.13551604509974, "grad_norm": 3.311258316040039, "learning_rate": 1.2474974611925142e-05, "loss": 0.0789, "step": 10474 }, { "epoch": 1.1356244579358197, "grad_norm": 1.0611567497253418, "learning_rate": 1.2474249238357755e-05, "loss": 0.0051, "step": 10475 }, { "epoch": 1.1357328707718994, "grad_norm": 0.3061218857765198, "learning_rate": 1.247352386479037e-05, "loss": 0.0066, "step": 10476 }, { "epoch": 1.1358412836079792, "grad_norm": 4.32620906829834, "learning_rate": 1.2472798491222981e-05, "loss": 0.0102, "step": 10477 }, { "epoch": 1.135949696444059, "grad_norm": 2.508978843688965, "learning_rate": 1.2472073117655594e-05, "loss": 0.006, "step": 10478 }, { "epoch": 1.1360581092801387, "grad_norm": 2.2972447872161865, "learning_rate": 1.2471347744088207e-05, "loss": 0.0239, "step": 10479 }, { "epoch": 1.1361665221162185, "grad_norm": 1.700519323348999, "learning_rate": 1.247062237052082e-05, "loss": 0.0334, "step": 10480 }, { "epoch": 1.1362749349522985, "grad_norm": 1.0487827062606812, "learning_rate": 1.2469896996953432e-05, "loss": 0.0109, "step": 10481 }, { "epoch": 1.1363833477883782, "grad_norm": 0.31280282139778137, "learning_rate": 1.2469171623386045e-05, "loss": 0.0094, "step": 10482 }, { "epoch": 1.136491760624458, "grad_norm": 0.24808749556541443, "learning_rate": 1.2468446249818658e-05, "loss": 0.0023, "step": 10483 }, { "epoch": 1.1366001734605378, "grad_norm": 0.13998787105083466, "learning_rate": 1.246772087625127e-05, "loss": 0.0011, "step": 10484 }, { "epoch": 1.1367085862966175, "grad_norm": 0.8410853743553162, "learning_rate": 1.2466995502683884e-05, "loss": 0.0041, "step": 10485 }, { "epoch": 1.1368169991326973, "grad_norm": 2.8688347339630127, "learning_rate": 1.2466270129116496e-05, "loss": 0.0918, "step": 10486 }, { "epoch": 1.136925411968777, "grad_norm": 2.119229316711426, "learning_rate": 1.2465544755549108e-05, "loss": 0.0476, "step": 10487 }, { "epoch": 1.1370338248048568, "grad_norm": 7.358556270599365, "learning_rate": 1.2464819381981722e-05, "loss": 0.0288, "step": 10488 }, { "epoch": 1.1371422376409366, "grad_norm": 2.0117781162261963, "learning_rate": 1.2464094008414334e-05, "loss": 0.0063, "step": 10489 }, { "epoch": 1.1372506504770166, "grad_norm": 1.804250955581665, "learning_rate": 1.2463368634846946e-05, "loss": 0.0085, "step": 10490 }, { "epoch": 1.1373590633130963, "grad_norm": 0.2434483915567398, "learning_rate": 1.2462643261279558e-05, "loss": 0.0013, "step": 10491 }, { "epoch": 1.137467476149176, "grad_norm": 0.0441192165017128, "learning_rate": 1.2461917887712174e-05, "loss": 0.0005, "step": 10492 }, { "epoch": 1.1375758889852559, "grad_norm": 1.902610421180725, "learning_rate": 1.2461192514144786e-05, "loss": 0.0539, "step": 10493 }, { "epoch": 1.1376843018213356, "grad_norm": 3.4612913131713867, "learning_rate": 1.24604671405774e-05, "loss": 0.027, "step": 10494 }, { "epoch": 1.1377927146574154, "grad_norm": 0.7095637321472168, "learning_rate": 1.2459741767010012e-05, "loss": 0.0069, "step": 10495 }, { "epoch": 1.1379011274934951, "grad_norm": 5.69302225112915, "learning_rate": 1.2459016393442624e-05, "loss": 0.0425, "step": 10496 }, { "epoch": 1.1380095403295751, "grad_norm": 2.0917088985443115, "learning_rate": 1.2458291019875238e-05, "loss": 0.0291, "step": 10497 }, { "epoch": 1.138117953165655, "grad_norm": 12.184344291687012, "learning_rate": 1.245756564630785e-05, "loss": 0.0695, "step": 10498 }, { "epoch": 1.1382263660017347, "grad_norm": 0.8182512521743774, "learning_rate": 1.2456840272740462e-05, "loss": 0.0231, "step": 10499 }, { "epoch": 1.1383347788378144, "grad_norm": 0.3856706917285919, "learning_rate": 1.2456114899173076e-05, "loss": 0.0074, "step": 10500 }, { "epoch": 1.1384431916738942, "grad_norm": 2.3777916431427, "learning_rate": 1.2455389525605688e-05, "loss": 0.0192, "step": 10501 }, { "epoch": 1.138551604509974, "grad_norm": 5.525095462799072, "learning_rate": 1.24546641520383e-05, "loss": 0.0656, "step": 10502 }, { "epoch": 1.1386600173460537, "grad_norm": 0.35826316475868225, "learning_rate": 1.2453938778470912e-05, "loss": 0.019, "step": 10503 }, { "epoch": 1.1387684301821335, "grad_norm": 5.472493648529053, "learning_rate": 1.2453213404903526e-05, "loss": 0.0295, "step": 10504 }, { "epoch": 1.1388768430182132, "grad_norm": 3.6913866996765137, "learning_rate": 1.2452488031336138e-05, "loss": 0.0112, "step": 10505 }, { "epoch": 1.1389852558542932, "grad_norm": 0.030102605000138283, "learning_rate": 1.245176265776875e-05, "loss": 0.0004, "step": 10506 }, { "epoch": 1.139093668690373, "grad_norm": 2.4417734146118164, "learning_rate": 1.2451037284201364e-05, "loss": 0.0327, "step": 10507 }, { "epoch": 1.1392020815264527, "grad_norm": 4.316529750823975, "learning_rate": 1.2450311910633976e-05, "loss": 0.0454, "step": 10508 }, { "epoch": 1.1393104943625325, "grad_norm": 12.256725311279297, "learning_rate": 1.2449586537066592e-05, "loss": 0.024, "step": 10509 }, { "epoch": 1.1394189071986123, "grad_norm": 0.9267774820327759, "learning_rate": 1.2448861163499204e-05, "loss": 0.0052, "step": 10510 }, { "epoch": 1.139527320034692, "grad_norm": 0.4705113470554352, "learning_rate": 1.2448135789931816e-05, "loss": 0.0033, "step": 10511 }, { "epoch": 1.1396357328707718, "grad_norm": 0.11209579557180405, "learning_rate": 1.244741041636443e-05, "loss": 0.0009, "step": 10512 }, { "epoch": 1.1397441457068518, "grad_norm": 2.442225933074951, "learning_rate": 1.2446685042797042e-05, "loss": 0.0498, "step": 10513 }, { "epoch": 1.1398525585429315, "grad_norm": 0.8953227400779724, "learning_rate": 1.2445959669229654e-05, "loss": 0.0054, "step": 10514 }, { "epoch": 1.1399609713790113, "grad_norm": 0.0503297857940197, "learning_rate": 1.2445234295662267e-05, "loss": 0.0004, "step": 10515 }, { "epoch": 1.140069384215091, "grad_norm": 4.0525078773498535, "learning_rate": 1.244450892209488e-05, "loss": 0.0624, "step": 10516 }, { "epoch": 1.1401777970511708, "grad_norm": 0.7132616639137268, "learning_rate": 1.2443783548527493e-05, "loss": 0.0101, "step": 10517 }, { "epoch": 1.1402862098872506, "grad_norm": 0.09665247797966003, "learning_rate": 1.2443058174960105e-05, "loss": 0.001, "step": 10518 }, { "epoch": 1.1403946227233304, "grad_norm": 0.836480975151062, "learning_rate": 1.2442332801392719e-05, "loss": 0.0128, "step": 10519 }, { "epoch": 1.1405030355594103, "grad_norm": 7.137210845947266, "learning_rate": 1.244160742782533e-05, "loss": 0.0271, "step": 10520 }, { "epoch": 1.1406114483954901, "grad_norm": 0.3239494860172272, "learning_rate": 1.2440882054257943e-05, "loss": 0.0093, "step": 10521 }, { "epoch": 1.1407198612315699, "grad_norm": 6.072994709014893, "learning_rate": 1.2440156680690557e-05, "loss": 0.0421, "step": 10522 }, { "epoch": 1.1408282740676496, "grad_norm": 0.30172818899154663, "learning_rate": 1.2439431307123169e-05, "loss": 0.0308, "step": 10523 }, { "epoch": 1.1409366869037294, "grad_norm": 1.3908344507217407, "learning_rate": 1.2438705933555781e-05, "loss": 0.022, "step": 10524 }, { "epoch": 1.1410450997398092, "grad_norm": 4.762768268585205, "learning_rate": 1.2437980559988395e-05, "loss": 0.0249, "step": 10525 }, { "epoch": 1.141153512575889, "grad_norm": 2.2647993564605713, "learning_rate": 1.2437255186421009e-05, "loss": 0.04, "step": 10526 }, { "epoch": 1.1412619254119687, "grad_norm": 7.495922088623047, "learning_rate": 1.2436529812853622e-05, "loss": 0.0386, "step": 10527 }, { "epoch": 1.1413703382480485, "grad_norm": 0.554620623588562, "learning_rate": 1.2435804439286235e-05, "loss": 0.0031, "step": 10528 }, { "epoch": 1.1414787510841284, "grad_norm": 9.132238388061523, "learning_rate": 1.2435079065718847e-05, "loss": 0.0319, "step": 10529 }, { "epoch": 1.1415871639202082, "grad_norm": 1.310931921005249, "learning_rate": 1.2434353692151459e-05, "loss": 0.0164, "step": 10530 }, { "epoch": 1.141695576756288, "grad_norm": 4.035055160522461, "learning_rate": 1.2433628318584073e-05, "loss": 0.0219, "step": 10531 }, { "epoch": 1.1418039895923677, "grad_norm": 1.125093936920166, "learning_rate": 1.2432902945016685e-05, "loss": 0.0024, "step": 10532 }, { "epoch": 1.1419124024284475, "grad_norm": 1.5405887365341187, "learning_rate": 1.2432177571449297e-05, "loss": 0.0353, "step": 10533 }, { "epoch": 1.1420208152645273, "grad_norm": 1.9627419710159302, "learning_rate": 1.243145219788191e-05, "loss": 0.0113, "step": 10534 }, { "epoch": 1.142129228100607, "grad_norm": 2.144775629043579, "learning_rate": 1.2430726824314523e-05, "loss": 0.0148, "step": 10535 }, { "epoch": 1.142237640936687, "grad_norm": 5.786046504974365, "learning_rate": 1.2430001450747135e-05, "loss": 0.0318, "step": 10536 }, { "epoch": 1.1423460537727668, "grad_norm": 4.402706146240234, "learning_rate": 1.2429276077179749e-05, "loss": 0.0183, "step": 10537 }, { "epoch": 1.1424544666088465, "grad_norm": 1.719185709953308, "learning_rate": 1.2428550703612361e-05, "loss": 0.0114, "step": 10538 }, { "epoch": 1.1425628794449263, "grad_norm": 0.5430685877799988, "learning_rate": 1.2427825330044973e-05, "loss": 0.0032, "step": 10539 }, { "epoch": 1.142671292281006, "grad_norm": 3.2279176712036133, "learning_rate": 1.2427099956477587e-05, "loss": 0.0613, "step": 10540 }, { "epoch": 1.1427797051170858, "grad_norm": 0.1293559968471527, "learning_rate": 1.24263745829102e-05, "loss": 0.0007, "step": 10541 }, { "epoch": 1.1428881179531656, "grad_norm": 6.092804431915283, "learning_rate": 1.2425649209342811e-05, "loss": 0.0274, "step": 10542 }, { "epoch": 1.1429965307892456, "grad_norm": 2.771338939666748, "learning_rate": 1.2424923835775424e-05, "loss": 0.0195, "step": 10543 }, { "epoch": 1.1431049436253253, "grad_norm": 0.5665154457092285, "learning_rate": 1.2424198462208039e-05, "loss": 0.0222, "step": 10544 }, { "epoch": 1.143213356461405, "grad_norm": 2.4341204166412354, "learning_rate": 1.2423473088640651e-05, "loss": 0.0387, "step": 10545 }, { "epoch": 1.1433217692974849, "grad_norm": 3.84700608253479, "learning_rate": 1.2422747715073265e-05, "loss": 0.0168, "step": 10546 }, { "epoch": 1.1434301821335646, "grad_norm": 2.036345958709717, "learning_rate": 1.2422022341505877e-05, "loss": 0.0423, "step": 10547 }, { "epoch": 1.1435385949696444, "grad_norm": 1.3810131549835205, "learning_rate": 1.242129696793849e-05, "loss": 0.03, "step": 10548 }, { "epoch": 1.1436470078057241, "grad_norm": 5.513139247894287, "learning_rate": 1.2420571594371103e-05, "loss": 0.0171, "step": 10549 }, { "epoch": 1.143755420641804, "grad_norm": 2.492495059967041, "learning_rate": 1.2419846220803715e-05, "loss": 0.0478, "step": 10550 }, { "epoch": 1.1438638334778837, "grad_norm": 4.704056262969971, "learning_rate": 1.2419120847236327e-05, "loss": 0.01, "step": 10551 }, { "epoch": 1.1439722463139637, "grad_norm": 4.186091899871826, "learning_rate": 1.2418395473668941e-05, "loss": 0.0374, "step": 10552 }, { "epoch": 1.1440806591500434, "grad_norm": 0.6637603044509888, "learning_rate": 1.2417670100101553e-05, "loss": 0.0119, "step": 10553 }, { "epoch": 1.1441890719861232, "grad_norm": 0.4321880340576172, "learning_rate": 1.2416944726534166e-05, "loss": 0.0132, "step": 10554 }, { "epoch": 1.144297484822203, "grad_norm": 0.073900006711483, "learning_rate": 1.2416219352966778e-05, "loss": 0.0006, "step": 10555 }, { "epoch": 1.1444058976582827, "grad_norm": 2.53753399848938, "learning_rate": 1.2415493979399392e-05, "loss": 0.0317, "step": 10556 }, { "epoch": 1.1445143104943625, "grad_norm": 0.21227507293224335, "learning_rate": 1.2414768605832004e-05, "loss": 0.0021, "step": 10557 }, { "epoch": 1.1446227233304422, "grad_norm": 3.423175096511841, "learning_rate": 1.2414043232264616e-05, "loss": 0.0214, "step": 10558 }, { "epoch": 1.1447311361665222, "grad_norm": 2.2220547199249268, "learning_rate": 1.241331785869723e-05, "loss": 0.0048, "step": 10559 }, { "epoch": 1.144839549002602, "grad_norm": 1.3257725238800049, "learning_rate": 1.2412592485129842e-05, "loss": 0.0173, "step": 10560 }, { "epoch": 1.1449479618386817, "grad_norm": 2.7678401470184326, "learning_rate": 1.2411867111562457e-05, "loss": 0.0337, "step": 10561 }, { "epoch": 1.1450563746747615, "grad_norm": 0.9068416357040405, "learning_rate": 1.241114173799507e-05, "loss": 0.0102, "step": 10562 }, { "epoch": 1.1451647875108413, "grad_norm": 3.9161479473114014, "learning_rate": 1.2410416364427682e-05, "loss": 0.0417, "step": 10563 }, { "epoch": 1.145273200346921, "grad_norm": 1.132385015487671, "learning_rate": 1.2409690990860295e-05, "loss": 0.0139, "step": 10564 }, { "epoch": 1.1453816131830008, "grad_norm": 2.904233455657959, "learning_rate": 1.2408965617292908e-05, "loss": 0.0493, "step": 10565 }, { "epoch": 1.1454900260190806, "grad_norm": 1.0372207164764404, "learning_rate": 1.240824024372552e-05, "loss": 0.0182, "step": 10566 }, { "epoch": 1.1455984388551603, "grad_norm": 3.8064029216766357, "learning_rate": 1.2407514870158132e-05, "loss": 0.0225, "step": 10567 }, { "epoch": 1.1457068516912403, "grad_norm": 2.012054920196533, "learning_rate": 1.2406789496590746e-05, "loss": 0.0202, "step": 10568 }, { "epoch": 1.14581526452732, "grad_norm": 3.980055093765259, "learning_rate": 1.2406064123023358e-05, "loss": 0.0482, "step": 10569 }, { "epoch": 1.1459236773633998, "grad_norm": 2.2009716033935547, "learning_rate": 1.240533874945597e-05, "loss": 0.0097, "step": 10570 }, { "epoch": 1.1460320901994796, "grad_norm": 5.997243881225586, "learning_rate": 1.2404613375888584e-05, "loss": 0.0093, "step": 10571 }, { "epoch": 1.1461405030355594, "grad_norm": 3.6236727237701416, "learning_rate": 1.2403888002321196e-05, "loss": 0.051, "step": 10572 }, { "epoch": 1.1462489158716391, "grad_norm": 4.050356388092041, "learning_rate": 1.2403162628753808e-05, "loss": 0.0231, "step": 10573 }, { "epoch": 1.146357328707719, "grad_norm": 0.37431421875953674, "learning_rate": 1.2402437255186422e-05, "loss": 0.032, "step": 10574 }, { "epoch": 1.1464657415437989, "grad_norm": 0.27064749598503113, "learning_rate": 1.2401711881619034e-05, "loss": 0.0057, "step": 10575 }, { "epoch": 1.1465741543798786, "grad_norm": 6.045248508453369, "learning_rate": 1.2400986508051646e-05, "loss": 0.0239, "step": 10576 }, { "epoch": 1.1466825672159584, "grad_norm": 2.260554790496826, "learning_rate": 1.240026113448426e-05, "loss": 0.0191, "step": 10577 }, { "epoch": 1.1467909800520382, "grad_norm": 1.7210527658462524, "learning_rate": 1.2399535760916874e-05, "loss": 0.0389, "step": 10578 }, { "epoch": 1.146899392888118, "grad_norm": 1.3975141048431396, "learning_rate": 1.2398810387349488e-05, "loss": 0.0052, "step": 10579 }, { "epoch": 1.1470078057241977, "grad_norm": 0.11616061627864838, "learning_rate": 1.23980850137821e-05, "loss": 0.0013, "step": 10580 }, { "epoch": 1.1471162185602775, "grad_norm": 1.5027486085891724, "learning_rate": 1.2397359640214712e-05, "loss": 0.0227, "step": 10581 }, { "epoch": 1.1472246313963574, "grad_norm": 0.5395842790603638, "learning_rate": 1.2396634266647324e-05, "loss": 0.0146, "step": 10582 }, { "epoch": 1.1473330442324372, "grad_norm": 2.091209888458252, "learning_rate": 1.2395908893079938e-05, "loss": 0.0358, "step": 10583 }, { "epoch": 1.147441457068517, "grad_norm": 2.0232393741607666, "learning_rate": 1.239518351951255e-05, "loss": 0.0299, "step": 10584 }, { "epoch": 1.1475498699045967, "grad_norm": 4.064876556396484, "learning_rate": 1.2394458145945162e-05, "loss": 0.0272, "step": 10585 }, { "epoch": 1.1476582827406765, "grad_norm": 1.0884668827056885, "learning_rate": 1.2393732772377776e-05, "loss": 0.0277, "step": 10586 }, { "epoch": 1.1477666955767563, "grad_norm": 3.900261640548706, "learning_rate": 1.2393007398810388e-05, "loss": 0.027, "step": 10587 }, { "epoch": 1.147875108412836, "grad_norm": 0.7864914536476135, "learning_rate": 1.2392282025243e-05, "loss": 0.0069, "step": 10588 }, { "epoch": 1.1479835212489158, "grad_norm": 4.3245720863342285, "learning_rate": 1.2391556651675614e-05, "loss": 0.0412, "step": 10589 }, { "epoch": 1.1480919340849955, "grad_norm": 3.9431753158569336, "learning_rate": 1.2390831278108226e-05, "loss": 0.0313, "step": 10590 }, { "epoch": 1.1482003469210755, "grad_norm": 1.6068191528320312, "learning_rate": 1.2390105904540839e-05, "loss": 0.0049, "step": 10591 }, { "epoch": 1.1483087597571553, "grad_norm": 3.126695156097412, "learning_rate": 1.2389380530973452e-05, "loss": 0.012, "step": 10592 }, { "epoch": 1.148417172593235, "grad_norm": 2.413947582244873, "learning_rate": 1.2388655157406065e-05, "loss": 0.0041, "step": 10593 }, { "epoch": 1.1485255854293148, "grad_norm": 0.6590264439582825, "learning_rate": 1.2387929783838677e-05, "loss": 0.0048, "step": 10594 }, { "epoch": 1.1486339982653946, "grad_norm": 2.208580255508423, "learning_rate": 1.2387204410271292e-05, "loss": 0.0322, "step": 10595 }, { "epoch": 1.1487424111014743, "grad_norm": 1.3710708618164062, "learning_rate": 1.2386479036703904e-05, "loss": 0.0095, "step": 10596 }, { "epoch": 1.148850823937554, "grad_norm": 0.3100741505622864, "learning_rate": 1.2385753663136517e-05, "loss": 0.0022, "step": 10597 }, { "epoch": 1.148959236773634, "grad_norm": 0.8239582777023315, "learning_rate": 1.238502828956913e-05, "loss": 0.015, "step": 10598 }, { "epoch": 1.1490676496097139, "grad_norm": 3.4787957668304443, "learning_rate": 1.2384302916001743e-05, "loss": 0.0108, "step": 10599 }, { "epoch": 1.1491760624457936, "grad_norm": 5.255945205688477, "learning_rate": 1.2383577542434355e-05, "loss": 0.0519, "step": 10600 }, { "epoch": 1.1492844752818734, "grad_norm": 10.328628540039062, "learning_rate": 1.2382852168866968e-05, "loss": 0.0252, "step": 10601 }, { "epoch": 1.1493928881179531, "grad_norm": 1.6029882431030273, "learning_rate": 1.238212679529958e-05, "loss": 0.0071, "step": 10602 }, { "epoch": 1.149501300954033, "grad_norm": 0.8784891366958618, "learning_rate": 1.2381401421732193e-05, "loss": 0.027, "step": 10603 }, { "epoch": 1.1496097137901127, "grad_norm": 2.307964324951172, "learning_rate": 1.2380676048164807e-05, "loss": 0.042, "step": 10604 }, { "epoch": 1.1497181266261927, "grad_norm": 0.29261788725852966, "learning_rate": 1.2379950674597419e-05, "loss": 0.0085, "step": 10605 }, { "epoch": 1.1498265394622724, "grad_norm": 1.948144555091858, "learning_rate": 1.2379225301030031e-05, "loss": 0.0257, "step": 10606 }, { "epoch": 1.1499349522983522, "grad_norm": 1.5814342498779297, "learning_rate": 1.2378499927462643e-05, "loss": 0.0256, "step": 10607 }, { "epoch": 1.150043365134432, "grad_norm": 1.5902718305587769, "learning_rate": 1.2377774553895257e-05, "loss": 0.0206, "step": 10608 }, { "epoch": 1.1501517779705117, "grad_norm": 1.8364219665527344, "learning_rate": 1.2377049180327869e-05, "loss": 0.0233, "step": 10609 }, { "epoch": 1.1502601908065915, "grad_norm": 1.615221381187439, "learning_rate": 1.2376323806760481e-05, "loss": 0.0556, "step": 10610 }, { "epoch": 1.1503686036426712, "grad_norm": 2.7054593563079834, "learning_rate": 1.2375598433193095e-05, "loss": 0.0521, "step": 10611 }, { "epoch": 1.150477016478751, "grad_norm": 3.0695886611938477, "learning_rate": 1.2374873059625707e-05, "loss": 0.0463, "step": 10612 }, { "epoch": 1.1505854293148308, "grad_norm": 0.8912034630775452, "learning_rate": 1.2374147686058323e-05, "loss": 0.0185, "step": 10613 }, { "epoch": 1.1506938421509108, "grad_norm": 0.7159791588783264, "learning_rate": 1.2373422312490935e-05, "loss": 0.0137, "step": 10614 }, { "epoch": 1.1508022549869905, "grad_norm": 0.10610870271921158, "learning_rate": 1.2372696938923547e-05, "loss": 0.0013, "step": 10615 }, { "epoch": 1.1509106678230703, "grad_norm": 0.17581404745578766, "learning_rate": 1.237197156535616e-05, "loss": 0.0019, "step": 10616 }, { "epoch": 1.15101908065915, "grad_norm": 2.6576199531555176, "learning_rate": 1.2371246191788773e-05, "loss": 0.036, "step": 10617 }, { "epoch": 1.1511274934952298, "grad_norm": 0.368318110704422, "learning_rate": 1.2370520818221385e-05, "loss": 0.0047, "step": 10618 }, { "epoch": 1.1512359063313096, "grad_norm": 0.837104320526123, "learning_rate": 1.2369795444653997e-05, "loss": 0.0036, "step": 10619 }, { "epoch": 1.1513443191673893, "grad_norm": 6.31784725189209, "learning_rate": 1.2369070071086611e-05, "loss": 0.0088, "step": 10620 }, { "epoch": 1.1514527320034693, "grad_norm": 2.14203143119812, "learning_rate": 1.2368344697519223e-05, "loss": 0.012, "step": 10621 }, { "epoch": 1.151561144839549, "grad_norm": 0.8307805061340332, "learning_rate": 1.2367619323951835e-05, "loss": 0.0073, "step": 10622 }, { "epoch": 1.1516695576756288, "grad_norm": 2.4943366050720215, "learning_rate": 1.236689395038445e-05, "loss": 0.0198, "step": 10623 }, { "epoch": 1.1517779705117086, "grad_norm": 4.788029670715332, "learning_rate": 1.2366168576817061e-05, "loss": 0.0402, "step": 10624 }, { "epoch": 1.1518863833477884, "grad_norm": 2.892270088195801, "learning_rate": 1.2365443203249673e-05, "loss": 0.0201, "step": 10625 }, { "epoch": 1.1519947961838681, "grad_norm": 1.7087719440460205, "learning_rate": 1.2364717829682287e-05, "loss": 0.0317, "step": 10626 }, { "epoch": 1.152103209019948, "grad_norm": 0.8983746767044067, "learning_rate": 1.23639924561149e-05, "loss": 0.0278, "step": 10627 }, { "epoch": 1.1522116218560277, "grad_norm": 2.2818386554718018, "learning_rate": 1.2363267082547512e-05, "loss": 0.0461, "step": 10628 }, { "epoch": 1.1523200346921076, "grad_norm": 2.4734625816345215, "learning_rate": 1.2362541708980125e-05, "loss": 0.0419, "step": 10629 }, { "epoch": 1.1524284475281874, "grad_norm": 6.887143135070801, "learning_rate": 1.236181633541274e-05, "loss": 0.0295, "step": 10630 }, { "epoch": 1.1525368603642672, "grad_norm": 0.17915505170822144, "learning_rate": 1.2361090961845353e-05, "loss": 0.0005, "step": 10631 }, { "epoch": 1.152645273200347, "grad_norm": 1.0276601314544678, "learning_rate": 1.2360365588277965e-05, "loss": 0.013, "step": 10632 }, { "epoch": 1.1527536860364267, "grad_norm": 4.330392837524414, "learning_rate": 1.2359640214710577e-05, "loss": 0.0445, "step": 10633 }, { "epoch": 1.1528620988725065, "grad_norm": 2.884093761444092, "learning_rate": 1.235891484114319e-05, "loss": 0.0428, "step": 10634 }, { "epoch": 1.1529705117085862, "grad_norm": 2.429352283477783, "learning_rate": 1.2358189467575803e-05, "loss": 0.0203, "step": 10635 }, { "epoch": 1.153078924544666, "grad_norm": 3.4883413314819336, "learning_rate": 1.2357464094008416e-05, "loss": 0.012, "step": 10636 }, { "epoch": 1.153187337380746, "grad_norm": 10.640593528747559, "learning_rate": 1.2356738720441028e-05, "loss": 0.0106, "step": 10637 }, { "epoch": 1.1532957502168257, "grad_norm": 0.2305629849433899, "learning_rate": 1.2356013346873642e-05, "loss": 0.0128, "step": 10638 }, { "epoch": 1.1534041630529055, "grad_norm": 0.3299785256385803, "learning_rate": 1.2355287973306254e-05, "loss": 0.0125, "step": 10639 }, { "epoch": 1.1535125758889853, "grad_norm": 2.5797200202941895, "learning_rate": 1.2354562599738866e-05, "loss": 0.0204, "step": 10640 }, { "epoch": 1.153620988725065, "grad_norm": 0.4010950028896332, "learning_rate": 1.235383722617148e-05, "loss": 0.001, "step": 10641 }, { "epoch": 1.1537294015611448, "grad_norm": 2.074427366256714, "learning_rate": 1.2353111852604092e-05, "loss": 0.0249, "step": 10642 }, { "epoch": 1.1538378143972245, "grad_norm": 4.170165061950684, "learning_rate": 1.2352386479036704e-05, "loss": 0.0193, "step": 10643 }, { "epoch": 1.1539462272333045, "grad_norm": 0.6952863931655884, "learning_rate": 1.2351661105469318e-05, "loss": 0.002, "step": 10644 }, { "epoch": 1.1540546400693843, "grad_norm": 0.19398754835128784, "learning_rate": 1.235093573190193e-05, "loss": 0.0137, "step": 10645 }, { "epoch": 1.154163052905464, "grad_norm": 1.394538402557373, "learning_rate": 1.2350210358334542e-05, "loss": 0.0084, "step": 10646 }, { "epoch": 1.1542714657415438, "grad_norm": 5.674915790557861, "learning_rate": 1.2349484984767158e-05, "loss": 0.0197, "step": 10647 }, { "epoch": 1.1543798785776236, "grad_norm": 0.8705136775970459, "learning_rate": 1.234875961119977e-05, "loss": 0.0207, "step": 10648 }, { "epoch": 1.1544882914137033, "grad_norm": 1.4198615550994873, "learning_rate": 1.2348034237632382e-05, "loss": 0.0087, "step": 10649 }, { "epoch": 1.1545967042497831, "grad_norm": 0.40272149443626404, "learning_rate": 1.2347308864064996e-05, "loss": 0.0041, "step": 10650 }, { "epoch": 1.1547051170858629, "grad_norm": 3.842033863067627, "learning_rate": 1.2346583490497608e-05, "loss": 0.0233, "step": 10651 }, { "epoch": 1.1548135299219426, "grad_norm": 3.440856695175171, "learning_rate": 1.234585811693022e-05, "loss": 0.019, "step": 10652 }, { "epoch": 1.1549219427580226, "grad_norm": 0.7359222173690796, "learning_rate": 1.2345132743362834e-05, "loss": 0.0029, "step": 10653 }, { "epoch": 1.1550303555941024, "grad_norm": 1.8425463438034058, "learning_rate": 1.2344407369795446e-05, "loss": 0.0197, "step": 10654 }, { "epoch": 1.1551387684301822, "grad_norm": 1.6157609224319458, "learning_rate": 1.2343681996228058e-05, "loss": 0.025, "step": 10655 }, { "epoch": 1.155247181266262, "grad_norm": 3.588937759399414, "learning_rate": 1.2342956622660672e-05, "loss": 0.0087, "step": 10656 }, { "epoch": 1.1553555941023417, "grad_norm": 0.6645163297653198, "learning_rate": 1.2342231249093284e-05, "loss": 0.0028, "step": 10657 }, { "epoch": 1.1554640069384214, "grad_norm": 4.360535621643066, "learning_rate": 1.2341505875525896e-05, "loss": 0.0602, "step": 10658 }, { "epoch": 1.1555724197745012, "grad_norm": 1.2514333724975586, "learning_rate": 1.2340780501958508e-05, "loss": 0.0057, "step": 10659 }, { "epoch": 1.1556808326105812, "grad_norm": 0.688158392906189, "learning_rate": 1.2340055128391122e-05, "loss": 0.0171, "step": 10660 }, { "epoch": 1.155789245446661, "grad_norm": 0.047112416476011276, "learning_rate": 1.2339329754823734e-05, "loss": 0.0004, "step": 10661 }, { "epoch": 1.1558976582827407, "grad_norm": 20.388147354125977, "learning_rate": 1.2338604381256346e-05, "loss": 0.0405, "step": 10662 }, { "epoch": 1.1560060711188205, "grad_norm": 1.1331238746643066, "learning_rate": 1.233787900768896e-05, "loss": 0.0063, "step": 10663 }, { "epoch": 1.1561144839549002, "grad_norm": 13.044633865356445, "learning_rate": 1.2337153634121572e-05, "loss": 0.0418, "step": 10664 }, { "epoch": 1.15622289679098, "grad_norm": 2.081923723220825, "learning_rate": 1.2336428260554188e-05, "loss": 0.0205, "step": 10665 }, { "epoch": 1.1563313096270598, "grad_norm": 0.9973017573356628, "learning_rate": 1.23357028869868e-05, "loss": 0.0054, "step": 10666 }, { "epoch": 1.1564397224631398, "grad_norm": 0.3167414367198944, "learning_rate": 1.2334977513419412e-05, "loss": 0.0064, "step": 10667 }, { "epoch": 1.1565481352992195, "grad_norm": 2.2004430294036865, "learning_rate": 1.2334252139852026e-05, "loss": 0.0148, "step": 10668 }, { "epoch": 1.1566565481352993, "grad_norm": 0.9598429203033447, "learning_rate": 1.2333526766284638e-05, "loss": 0.0191, "step": 10669 }, { "epoch": 1.156764960971379, "grad_norm": 2.186718463897705, "learning_rate": 1.233280139271725e-05, "loss": 0.0292, "step": 10670 }, { "epoch": 1.1568733738074588, "grad_norm": 6.191115856170654, "learning_rate": 1.2332076019149863e-05, "loss": 0.0111, "step": 10671 }, { "epoch": 1.1569817866435386, "grad_norm": 9.976150512695312, "learning_rate": 1.2331350645582476e-05, "loss": 0.0143, "step": 10672 }, { "epoch": 1.1570901994796183, "grad_norm": 0.28788843750953674, "learning_rate": 1.2330625272015089e-05, "loss": 0.0057, "step": 10673 }, { "epoch": 1.157198612315698, "grad_norm": 0.4286693036556244, "learning_rate": 1.23298998984477e-05, "loss": 0.006, "step": 10674 }, { "epoch": 1.1573070251517779, "grad_norm": 2.633958578109741, "learning_rate": 1.2329174524880315e-05, "loss": 0.0499, "step": 10675 }, { "epoch": 1.1574154379878578, "grad_norm": 6.465961933135986, "learning_rate": 1.2328449151312927e-05, "loss": 0.0341, "step": 10676 }, { "epoch": 1.1575238508239376, "grad_norm": 2.546757698059082, "learning_rate": 1.2327723777745539e-05, "loss": 0.0246, "step": 10677 }, { "epoch": 1.1576322636600174, "grad_norm": 2.0684401988983154, "learning_rate": 1.2326998404178153e-05, "loss": 0.0114, "step": 10678 }, { "epoch": 1.1577406764960971, "grad_norm": 0.23065002262592316, "learning_rate": 1.2326273030610765e-05, "loss": 0.0049, "step": 10679 }, { "epoch": 1.157849089332177, "grad_norm": 0.48555707931518555, "learning_rate": 1.2325547657043377e-05, "loss": 0.0146, "step": 10680 }, { "epoch": 1.1579575021682567, "grad_norm": 1.9317561388015747, "learning_rate": 1.232482228347599e-05, "loss": 0.0316, "step": 10681 }, { "epoch": 1.1580659150043364, "grad_norm": 2.147916078567505, "learning_rate": 1.2324096909908605e-05, "loss": 0.0126, "step": 10682 }, { "epoch": 1.1581743278404164, "grad_norm": 0.5644909739494324, "learning_rate": 1.2323371536341218e-05, "loss": 0.0253, "step": 10683 }, { "epoch": 1.1582827406764962, "grad_norm": 3.165898561477661, "learning_rate": 1.232264616277383e-05, "loss": 0.0061, "step": 10684 }, { "epoch": 1.158391153512576, "grad_norm": 1.2766298055648804, "learning_rate": 1.2321920789206443e-05, "loss": 0.0246, "step": 10685 }, { "epoch": 1.1584995663486557, "grad_norm": 1.8730931282043457, "learning_rate": 1.2321195415639055e-05, "loss": 0.0424, "step": 10686 }, { "epoch": 1.1586079791847355, "grad_norm": 2.9624714851379395, "learning_rate": 1.2320470042071669e-05, "loss": 0.0337, "step": 10687 }, { "epoch": 1.1587163920208152, "grad_norm": 0.3781150281429291, "learning_rate": 1.2319744668504281e-05, "loss": 0.0108, "step": 10688 }, { "epoch": 1.158824804856895, "grad_norm": 0.44809970259666443, "learning_rate": 1.2319019294936893e-05, "loss": 0.001, "step": 10689 }, { "epoch": 1.158933217692975, "grad_norm": 1.8432655334472656, "learning_rate": 1.2318293921369507e-05, "loss": 0.0243, "step": 10690 }, { "epoch": 1.1590416305290547, "grad_norm": 7.064820766448975, "learning_rate": 1.2317568547802119e-05, "loss": 0.0559, "step": 10691 }, { "epoch": 1.1591500433651345, "grad_norm": 0.09252114593982697, "learning_rate": 1.2316843174234731e-05, "loss": 0.0008, "step": 10692 }, { "epoch": 1.1592584562012143, "grad_norm": 0.5783925652503967, "learning_rate": 1.2316117800667345e-05, "loss": 0.0154, "step": 10693 }, { "epoch": 1.159366869037294, "grad_norm": 1.3722431659698486, "learning_rate": 1.2315392427099957e-05, "loss": 0.0065, "step": 10694 }, { "epoch": 1.1594752818733738, "grad_norm": 0.2929612100124359, "learning_rate": 1.231466705353257e-05, "loss": 0.0063, "step": 10695 }, { "epoch": 1.1595836947094535, "grad_norm": 2.6541545391082764, "learning_rate": 1.2313941679965183e-05, "loss": 0.0589, "step": 10696 }, { "epoch": 1.1596921075455333, "grad_norm": 7.570868015289307, "learning_rate": 1.2313216306397795e-05, "loss": 0.0254, "step": 10697 }, { "epoch": 1.159800520381613, "grad_norm": 3.655616521835327, "learning_rate": 1.2312490932830407e-05, "loss": 0.0333, "step": 10698 }, { "epoch": 1.159908933217693, "grad_norm": 1.6305217742919922, "learning_rate": 1.2311765559263023e-05, "loss": 0.0049, "step": 10699 }, { "epoch": 1.1600173460537728, "grad_norm": 1.9358030557632446, "learning_rate": 1.2311040185695635e-05, "loss": 0.0528, "step": 10700 }, { "epoch": 1.1601257588898526, "grad_norm": 0.40350255370140076, "learning_rate": 1.2310314812128247e-05, "loss": 0.0066, "step": 10701 }, { "epoch": 1.1602341717259324, "grad_norm": 1.5948437452316284, "learning_rate": 1.2309589438560861e-05, "loss": 0.0284, "step": 10702 }, { "epoch": 1.1603425845620121, "grad_norm": 5.299556732177734, "learning_rate": 1.2308864064993473e-05, "loss": 0.0601, "step": 10703 }, { "epoch": 1.1604509973980919, "grad_norm": 4.288453102111816, "learning_rate": 1.2308138691426085e-05, "loss": 0.0713, "step": 10704 }, { "epoch": 1.1605594102341716, "grad_norm": 3.5839333534240723, "learning_rate": 1.23074133178587e-05, "loss": 0.014, "step": 10705 }, { "epoch": 1.1606678230702516, "grad_norm": 3.5984678268432617, "learning_rate": 1.2306687944291311e-05, "loss": 0.0186, "step": 10706 }, { "epoch": 1.1607762359063314, "grad_norm": 4.647922992706299, "learning_rate": 1.2305962570723923e-05, "loss": 0.0361, "step": 10707 }, { "epoch": 1.1608846487424112, "grad_norm": 7.199489593505859, "learning_rate": 1.2305237197156537e-05, "loss": 0.0357, "step": 10708 }, { "epoch": 1.160993061578491, "grad_norm": 0.3567824363708496, "learning_rate": 1.230451182358915e-05, "loss": 0.0068, "step": 10709 }, { "epoch": 1.1611014744145707, "grad_norm": 5.342402458190918, "learning_rate": 1.2303786450021762e-05, "loss": 0.0388, "step": 10710 }, { "epoch": 1.1612098872506504, "grad_norm": 0.274841845035553, "learning_rate": 1.2303061076454374e-05, "loss": 0.0057, "step": 10711 }, { "epoch": 1.1613183000867302, "grad_norm": 0.28033915162086487, "learning_rate": 1.2302335702886988e-05, "loss": 0.0088, "step": 10712 }, { "epoch": 1.16142671292281, "grad_norm": 3.7319843769073486, "learning_rate": 1.23016103293196e-05, "loss": 0.0317, "step": 10713 }, { "epoch": 1.1615351257588897, "grad_norm": 2.1092841625213623, "learning_rate": 1.2300884955752212e-05, "loss": 0.024, "step": 10714 }, { "epoch": 1.1616435385949697, "grad_norm": 1.0580967664718628, "learning_rate": 1.2300159582184826e-05, "loss": 0.0165, "step": 10715 }, { "epoch": 1.1617519514310495, "grad_norm": 0.4272522032260895, "learning_rate": 1.229943420861744e-05, "loss": 0.0034, "step": 10716 }, { "epoch": 1.1618603642671292, "grad_norm": 1.1739733219146729, "learning_rate": 1.2298708835050053e-05, "loss": 0.0151, "step": 10717 }, { "epoch": 1.161968777103209, "grad_norm": 0.5253989100456238, "learning_rate": 1.2297983461482665e-05, "loss": 0.0067, "step": 10718 }, { "epoch": 1.1620771899392888, "grad_norm": 0.31284019351005554, "learning_rate": 1.2297258087915278e-05, "loss": 0.0011, "step": 10719 }, { "epoch": 1.1621856027753685, "grad_norm": 1.6440943479537964, "learning_rate": 1.2296532714347891e-05, "loss": 0.0223, "step": 10720 }, { "epoch": 1.1622940156114483, "grad_norm": 0.48683083057403564, "learning_rate": 1.2295807340780504e-05, "loss": 0.0238, "step": 10721 }, { "epoch": 1.1624024284475283, "grad_norm": 0.5581669211387634, "learning_rate": 1.2295081967213116e-05, "loss": 0.0072, "step": 10722 }, { "epoch": 1.162510841283608, "grad_norm": 0.7419750690460205, "learning_rate": 1.229435659364573e-05, "loss": 0.004, "step": 10723 }, { "epoch": 1.1626192541196878, "grad_norm": 0.7183884382247925, "learning_rate": 1.2293631220078342e-05, "loss": 0.0244, "step": 10724 }, { "epoch": 1.1627276669557676, "grad_norm": 0.34062549471855164, "learning_rate": 1.2292905846510954e-05, "loss": 0.0105, "step": 10725 }, { "epoch": 1.1628360797918473, "grad_norm": 0.5063143372535706, "learning_rate": 1.2292180472943566e-05, "loss": 0.0098, "step": 10726 }, { "epoch": 1.162944492627927, "grad_norm": 2.92075514793396, "learning_rate": 1.229145509937618e-05, "loss": 0.0236, "step": 10727 }, { "epoch": 1.1630529054640069, "grad_norm": 1.6625925302505493, "learning_rate": 1.2290729725808792e-05, "loss": 0.0335, "step": 10728 }, { "epoch": 1.1631613183000868, "grad_norm": 1.0608305931091309, "learning_rate": 1.2290004352241404e-05, "loss": 0.0092, "step": 10729 }, { "epoch": 1.1632697311361666, "grad_norm": 0.08664074540138245, "learning_rate": 1.2289278978674018e-05, "loss": 0.0008, "step": 10730 }, { "epoch": 1.1633781439722464, "grad_norm": 0.8388910889625549, "learning_rate": 1.228855360510663e-05, "loss": 0.0194, "step": 10731 }, { "epoch": 1.1634865568083261, "grad_norm": 5.102963447570801, "learning_rate": 1.2287828231539242e-05, "loss": 0.0709, "step": 10732 }, { "epoch": 1.163594969644406, "grad_norm": 1.5285005569458008, "learning_rate": 1.2287102857971856e-05, "loss": 0.005, "step": 10733 }, { "epoch": 1.1637033824804857, "grad_norm": 1.8251911401748657, "learning_rate": 1.228637748440447e-05, "loss": 0.028, "step": 10734 }, { "epoch": 1.1638117953165654, "grad_norm": 1.445844054222107, "learning_rate": 1.2285652110837084e-05, "loss": 0.0204, "step": 10735 }, { "epoch": 1.1639202081526452, "grad_norm": 10.364396095275879, "learning_rate": 1.2284926737269696e-05, "loss": 0.0431, "step": 10736 }, { "epoch": 1.164028620988725, "grad_norm": 2.463773012161255, "learning_rate": 1.2284201363702308e-05, "loss": 0.055, "step": 10737 }, { "epoch": 1.164137033824805, "grad_norm": 4.149074077606201, "learning_rate": 1.228347599013492e-05, "loss": 0.0294, "step": 10738 }, { "epoch": 1.1642454466608847, "grad_norm": 1.5363640785217285, "learning_rate": 1.2282750616567534e-05, "loss": 0.0061, "step": 10739 }, { "epoch": 1.1643538594969645, "grad_norm": 0.5879969000816345, "learning_rate": 1.2282025243000146e-05, "loss": 0.0017, "step": 10740 }, { "epoch": 1.1644622723330442, "grad_norm": 4.227755069732666, "learning_rate": 1.2281299869432758e-05, "loss": 0.042, "step": 10741 }, { "epoch": 1.164570685169124, "grad_norm": 1.9128406047821045, "learning_rate": 1.2280574495865372e-05, "loss": 0.0863, "step": 10742 }, { "epoch": 1.1646790980052038, "grad_norm": 0.9637722373008728, "learning_rate": 1.2279849122297984e-05, "loss": 0.019, "step": 10743 }, { "epoch": 1.1647875108412835, "grad_norm": 5.172754764556885, "learning_rate": 1.2279123748730596e-05, "loss": 0.0385, "step": 10744 }, { "epoch": 1.1648959236773635, "grad_norm": 6.48955774307251, "learning_rate": 1.227839837516321e-05, "loss": 0.0246, "step": 10745 }, { "epoch": 1.1650043365134433, "grad_norm": 2.054006576538086, "learning_rate": 1.2277673001595822e-05, "loss": 0.0276, "step": 10746 }, { "epoch": 1.165112749349523, "grad_norm": 3.762632369995117, "learning_rate": 1.2276947628028435e-05, "loss": 0.0216, "step": 10747 }, { "epoch": 1.1652211621856028, "grad_norm": 0.5499668121337891, "learning_rate": 1.2276222254461048e-05, "loss": 0.0066, "step": 10748 }, { "epoch": 1.1653295750216826, "grad_norm": 2.5189266204833984, "learning_rate": 1.227549688089366e-05, "loss": 0.0366, "step": 10749 }, { "epoch": 1.1654379878577623, "grad_norm": 0.5112559795379639, "learning_rate": 1.2274771507326273e-05, "loss": 0.0188, "step": 10750 }, { "epoch": 1.165546400693842, "grad_norm": 1.8535680770874023, "learning_rate": 1.2274046133758888e-05, "loss": 0.0123, "step": 10751 }, { "epoch": 1.165654813529922, "grad_norm": 0.13848331570625305, "learning_rate": 1.22733207601915e-05, "loss": 0.0018, "step": 10752 }, { "epoch": 1.1657632263660018, "grad_norm": 0.9878071546554565, "learning_rate": 1.2272595386624113e-05, "loss": 0.024, "step": 10753 }, { "epoch": 1.1658716392020816, "grad_norm": 1.2500433921813965, "learning_rate": 1.2271870013056726e-05, "loss": 0.0026, "step": 10754 }, { "epoch": 1.1659800520381614, "grad_norm": 1.256819725036621, "learning_rate": 1.2271144639489339e-05, "loss": 0.0182, "step": 10755 }, { "epoch": 1.1660884648742411, "grad_norm": 0.3696213662624359, "learning_rate": 1.227041926592195e-05, "loss": 0.0086, "step": 10756 }, { "epoch": 1.1661968777103209, "grad_norm": 0.45930153131484985, "learning_rate": 1.2269693892354564e-05, "loss": 0.0147, "step": 10757 }, { "epoch": 1.1663052905464006, "grad_norm": 0.4500036835670471, "learning_rate": 1.2268968518787177e-05, "loss": 0.0112, "step": 10758 }, { "epoch": 1.1664137033824804, "grad_norm": 2.22509765625, "learning_rate": 1.2268243145219789e-05, "loss": 0.0326, "step": 10759 }, { "epoch": 1.1665221162185602, "grad_norm": 3.4354331493377686, "learning_rate": 1.2267517771652403e-05, "loss": 0.0538, "step": 10760 }, { "epoch": 1.1666305290546402, "grad_norm": 2.406233549118042, "learning_rate": 1.2266792398085015e-05, "loss": 0.0163, "step": 10761 }, { "epoch": 1.16673894189072, "grad_norm": 1.1506333351135254, "learning_rate": 1.2266067024517627e-05, "loss": 0.0329, "step": 10762 }, { "epoch": 1.1668473547267997, "grad_norm": 6.1524882316589355, "learning_rate": 1.2265341650950239e-05, "loss": 0.0152, "step": 10763 }, { "epoch": 1.1669557675628794, "grad_norm": 0.023769954219460487, "learning_rate": 1.2264616277382853e-05, "loss": 0.0002, "step": 10764 }, { "epoch": 1.1670641803989592, "grad_norm": 3.227067470550537, "learning_rate": 1.2263890903815465e-05, "loss": 0.0047, "step": 10765 }, { "epoch": 1.167172593235039, "grad_norm": 3.663919687271118, "learning_rate": 1.2263165530248077e-05, "loss": 0.0162, "step": 10766 }, { "epoch": 1.1672810060711187, "grad_norm": 0.09558005630970001, "learning_rate": 1.2262440156680691e-05, "loss": 0.0006, "step": 10767 }, { "epoch": 1.1673894189071987, "grad_norm": 0.5487382411956787, "learning_rate": 1.2261714783113305e-05, "loss": 0.0046, "step": 10768 }, { "epoch": 1.1674978317432785, "grad_norm": 0.44957759976387024, "learning_rate": 1.2260989409545919e-05, "loss": 0.0093, "step": 10769 }, { "epoch": 1.1676062445793582, "grad_norm": 0.6690652966499329, "learning_rate": 1.226026403597853e-05, "loss": 0.0038, "step": 10770 }, { "epoch": 1.167714657415438, "grad_norm": 0.9184154868125916, "learning_rate": 1.2259538662411143e-05, "loss": 0.0261, "step": 10771 }, { "epoch": 1.1678230702515178, "grad_norm": 0.3735913634300232, "learning_rate": 1.2258813288843757e-05, "loss": 0.0099, "step": 10772 }, { "epoch": 1.1679314830875975, "grad_norm": 4.760108947753906, "learning_rate": 1.2258087915276369e-05, "loss": 0.0382, "step": 10773 }, { "epoch": 1.1680398959236773, "grad_norm": 0.3902168571949005, "learning_rate": 1.2257362541708981e-05, "loss": 0.0104, "step": 10774 }, { "epoch": 1.168148308759757, "grad_norm": 0.3447076082229614, "learning_rate": 1.2256637168141595e-05, "loss": 0.02, "step": 10775 }, { "epoch": 1.1682567215958368, "grad_norm": 0.15172691643238068, "learning_rate": 1.2255911794574207e-05, "loss": 0.0012, "step": 10776 }, { "epoch": 1.1683651344319168, "grad_norm": 0.3877066373825073, "learning_rate": 1.225518642100682e-05, "loss": 0.0129, "step": 10777 }, { "epoch": 1.1684735472679966, "grad_norm": 2.9283692836761475, "learning_rate": 1.2254461047439431e-05, "loss": 0.0614, "step": 10778 }, { "epoch": 1.1685819601040763, "grad_norm": 0.6499618887901306, "learning_rate": 1.2253735673872045e-05, "loss": 0.0042, "step": 10779 }, { "epoch": 1.168690372940156, "grad_norm": 0.347257524728775, "learning_rate": 1.2253010300304657e-05, "loss": 0.0125, "step": 10780 }, { "epoch": 1.1687987857762359, "grad_norm": 1.8300191164016724, "learning_rate": 1.225228492673727e-05, "loss": 0.0022, "step": 10781 }, { "epoch": 1.1689071986123156, "grad_norm": 0.9814660549163818, "learning_rate": 1.2251559553169883e-05, "loss": 0.0223, "step": 10782 }, { "epoch": 1.1690156114483954, "grad_norm": 0.13529610633850098, "learning_rate": 1.2250834179602495e-05, "loss": 0.0014, "step": 10783 }, { "epoch": 1.1691240242844754, "grad_norm": 2.583087205886841, "learning_rate": 1.2250108806035108e-05, "loss": 0.0167, "step": 10784 }, { "epoch": 1.1692324371205551, "grad_norm": 2.6931607723236084, "learning_rate": 1.2249383432467721e-05, "loss": 0.0145, "step": 10785 }, { "epoch": 1.169340849956635, "grad_norm": 6.010319232940674, "learning_rate": 1.2248658058900335e-05, "loss": 0.0296, "step": 10786 }, { "epoch": 1.1694492627927147, "grad_norm": 1.671138882637024, "learning_rate": 1.2247932685332949e-05, "loss": 0.0052, "step": 10787 }, { "epoch": 1.1695576756287944, "grad_norm": 2.601155996322632, "learning_rate": 1.2247207311765561e-05, "loss": 0.0866, "step": 10788 }, { "epoch": 1.1696660884648742, "grad_norm": 0.3992711901664734, "learning_rate": 1.2246481938198173e-05, "loss": 0.0098, "step": 10789 }, { "epoch": 1.169774501300954, "grad_norm": 2.4439890384674072, "learning_rate": 1.2245756564630786e-05, "loss": 0.0567, "step": 10790 }, { "epoch": 1.169882914137034, "grad_norm": 1.7362638711929321, "learning_rate": 1.22450311910634e-05, "loss": 0.017, "step": 10791 }, { "epoch": 1.1699913269731137, "grad_norm": 0.3661291301250458, "learning_rate": 1.2244305817496012e-05, "loss": 0.0125, "step": 10792 }, { "epoch": 1.1700997398091935, "grad_norm": 0.29788675904273987, "learning_rate": 1.2243580443928624e-05, "loss": 0.0051, "step": 10793 }, { "epoch": 1.1702081526452732, "grad_norm": 1.5366531610488892, "learning_rate": 1.2242855070361238e-05, "loss": 0.0327, "step": 10794 }, { "epoch": 1.170316565481353, "grad_norm": 1.0145256519317627, "learning_rate": 1.224212969679385e-05, "loss": 0.0206, "step": 10795 }, { "epoch": 1.1704249783174328, "grad_norm": 6.933171272277832, "learning_rate": 1.2241404323226462e-05, "loss": 0.0545, "step": 10796 }, { "epoch": 1.1705333911535125, "grad_norm": 1.6544857025146484, "learning_rate": 1.2240678949659076e-05, "loss": 0.0228, "step": 10797 }, { "epoch": 1.1706418039895923, "grad_norm": 0.456603467464447, "learning_rate": 1.2239953576091688e-05, "loss": 0.002, "step": 10798 }, { "epoch": 1.170750216825672, "grad_norm": 2.445953607559204, "learning_rate": 1.22392282025243e-05, "loss": 0.0215, "step": 10799 }, { "epoch": 1.170858629661752, "grad_norm": 0.5197497606277466, "learning_rate": 1.2238502828956914e-05, "loss": 0.0093, "step": 10800 }, { "epoch": 1.1709670424978318, "grad_norm": 3.0305233001708984, "learning_rate": 1.2237777455389526e-05, "loss": 0.0116, "step": 10801 }, { "epoch": 1.1710754553339116, "grad_norm": 1.888885736465454, "learning_rate": 1.2237052081822138e-05, "loss": 0.0269, "step": 10802 }, { "epoch": 1.1711838681699913, "grad_norm": 0.4241541922092438, "learning_rate": 1.2236326708254754e-05, "loss": 0.0081, "step": 10803 }, { "epoch": 1.171292281006071, "grad_norm": 4.387640476226807, "learning_rate": 1.2235601334687366e-05, "loss": 0.0307, "step": 10804 }, { "epoch": 1.1714006938421508, "grad_norm": 2.7363150119781494, "learning_rate": 1.2234875961119978e-05, "loss": 0.0137, "step": 10805 }, { "epoch": 1.1715091066782306, "grad_norm": 3.1535515785217285, "learning_rate": 1.2234150587552592e-05, "loss": 0.0138, "step": 10806 }, { "epoch": 1.1716175195143106, "grad_norm": 1.5016988515853882, "learning_rate": 1.2233425213985204e-05, "loss": 0.0083, "step": 10807 }, { "epoch": 1.1717259323503904, "grad_norm": 0.8064522743225098, "learning_rate": 1.2232699840417816e-05, "loss": 0.0129, "step": 10808 }, { "epoch": 1.1718343451864701, "grad_norm": 0.8774905800819397, "learning_rate": 1.223197446685043e-05, "loss": 0.0053, "step": 10809 }, { "epoch": 1.1719427580225499, "grad_norm": 1.731755018234253, "learning_rate": 1.2231249093283042e-05, "loss": 0.0405, "step": 10810 }, { "epoch": 1.1720511708586296, "grad_norm": 0.421753853559494, "learning_rate": 1.2230523719715654e-05, "loss": 0.0218, "step": 10811 }, { "epoch": 1.1721595836947094, "grad_norm": 6.58620023727417, "learning_rate": 1.2229798346148268e-05, "loss": 0.0226, "step": 10812 }, { "epoch": 1.1722679965307892, "grad_norm": 2.407170534133911, "learning_rate": 1.222907297258088e-05, "loss": 0.0109, "step": 10813 }, { "epoch": 1.1723764093668692, "grad_norm": 0.12321178615093231, "learning_rate": 1.2228347599013492e-05, "loss": 0.0008, "step": 10814 }, { "epoch": 1.172484822202949, "grad_norm": 0.21238552033901215, "learning_rate": 1.2227622225446104e-05, "loss": 0.0043, "step": 10815 }, { "epoch": 1.1725932350390287, "grad_norm": 4.181194305419922, "learning_rate": 1.2226896851878718e-05, "loss": 0.0036, "step": 10816 }, { "epoch": 1.1727016478751084, "grad_norm": 1.3224767446517944, "learning_rate": 1.222617147831133e-05, "loss": 0.0156, "step": 10817 }, { "epoch": 1.1728100607111882, "grad_norm": 0.47441431879997253, "learning_rate": 1.2225446104743943e-05, "loss": 0.0027, "step": 10818 }, { "epoch": 1.172918473547268, "grad_norm": 6.33229923248291, "learning_rate": 1.2224720731176556e-05, "loss": 0.0267, "step": 10819 }, { "epoch": 1.1730268863833477, "grad_norm": 4.228228569030762, "learning_rate": 1.222399535760917e-05, "loss": 0.0576, "step": 10820 }, { "epoch": 1.1731352992194275, "grad_norm": 4.905219078063965, "learning_rate": 1.2223269984041784e-05, "loss": 0.0447, "step": 10821 }, { "epoch": 1.1732437120555073, "grad_norm": 0.8685855865478516, "learning_rate": 1.2222544610474396e-05, "loss": 0.0258, "step": 10822 }, { "epoch": 1.1733521248915872, "grad_norm": 9.382949829101562, "learning_rate": 1.2221819236907008e-05, "loss": 0.0529, "step": 10823 }, { "epoch": 1.173460537727667, "grad_norm": 0.6617307662963867, "learning_rate": 1.2221093863339622e-05, "loss": 0.0227, "step": 10824 }, { "epoch": 1.1735689505637468, "grad_norm": 4.3032026290893555, "learning_rate": 1.2220368489772234e-05, "loss": 0.0459, "step": 10825 }, { "epoch": 1.1736773633998265, "grad_norm": 0.36922091245651245, "learning_rate": 1.2219643116204846e-05, "loss": 0.0033, "step": 10826 }, { "epoch": 1.1737857762359063, "grad_norm": 6.937429428100586, "learning_rate": 1.221891774263746e-05, "loss": 0.0334, "step": 10827 }, { "epoch": 1.173894189071986, "grad_norm": 1.489320993423462, "learning_rate": 1.2218192369070072e-05, "loss": 0.022, "step": 10828 }, { "epoch": 1.1740026019080658, "grad_norm": 0.060309998691082, "learning_rate": 1.2217466995502685e-05, "loss": 0.0005, "step": 10829 }, { "epoch": 1.1741110147441458, "grad_norm": 0.4333968162536621, "learning_rate": 1.2216741621935297e-05, "loss": 0.0082, "step": 10830 }, { "epoch": 1.1742194275802256, "grad_norm": 4.689543724060059, "learning_rate": 1.221601624836791e-05, "loss": 0.0301, "step": 10831 }, { "epoch": 1.1743278404163053, "grad_norm": 1.1370458602905273, "learning_rate": 1.2215290874800523e-05, "loss": 0.0126, "step": 10832 }, { "epoch": 1.174436253252385, "grad_norm": 3.0572831630706787, "learning_rate": 1.2214565501233135e-05, "loss": 0.0275, "step": 10833 }, { "epoch": 1.1745446660884649, "grad_norm": 0.41408178210258484, "learning_rate": 1.2213840127665749e-05, "loss": 0.0096, "step": 10834 }, { "epoch": 1.1746530789245446, "grad_norm": 6.276247978210449, "learning_rate": 1.221311475409836e-05, "loss": 0.0519, "step": 10835 }, { "epoch": 1.1747614917606244, "grad_norm": 1.3364089727401733, "learning_rate": 1.2212389380530973e-05, "loss": 0.0164, "step": 10836 }, { "epoch": 1.1748699045967044, "grad_norm": 0.7173839807510376, "learning_rate": 1.2211664006963588e-05, "loss": 0.0225, "step": 10837 }, { "epoch": 1.1749783174327841, "grad_norm": 0.4164884090423584, "learning_rate": 1.22109386333962e-05, "loss": 0.0136, "step": 10838 }, { "epoch": 1.175086730268864, "grad_norm": 4.629250526428223, "learning_rate": 1.2210213259828814e-05, "loss": 0.0602, "step": 10839 }, { "epoch": 1.1751951431049437, "grad_norm": 4.210160255432129, "learning_rate": 1.2209487886261427e-05, "loss": 0.049, "step": 10840 }, { "epoch": 1.1753035559410234, "grad_norm": 0.5899997353553772, "learning_rate": 1.2208762512694039e-05, "loss": 0.0119, "step": 10841 }, { "epoch": 1.1754119687771032, "grad_norm": 1.0582891702651978, "learning_rate": 1.2208037139126651e-05, "loss": 0.0054, "step": 10842 }, { "epoch": 1.175520381613183, "grad_norm": 3.1518914699554443, "learning_rate": 1.2207311765559265e-05, "loss": 0.0082, "step": 10843 }, { "epoch": 1.1756287944492627, "grad_norm": 1.6735725402832031, "learning_rate": 1.2206586391991877e-05, "loss": 0.013, "step": 10844 }, { "epoch": 1.1757372072853425, "grad_norm": 0.8467465043067932, "learning_rate": 1.2205861018424489e-05, "loss": 0.0151, "step": 10845 }, { "epoch": 1.1758456201214225, "grad_norm": 11.019859313964844, "learning_rate": 1.2205135644857103e-05, "loss": 0.0739, "step": 10846 }, { "epoch": 1.1759540329575022, "grad_norm": 2.2854013442993164, "learning_rate": 1.2204410271289715e-05, "loss": 0.0198, "step": 10847 }, { "epoch": 1.176062445793582, "grad_norm": 1.1156728267669678, "learning_rate": 1.2203684897722327e-05, "loss": 0.031, "step": 10848 }, { "epoch": 1.1761708586296618, "grad_norm": 0.31838321685791016, "learning_rate": 1.2202959524154941e-05, "loss": 0.0105, "step": 10849 }, { "epoch": 1.1762792714657415, "grad_norm": 1.6453354358673096, "learning_rate": 1.2202234150587553e-05, "loss": 0.0246, "step": 10850 }, { "epoch": 1.1763876843018213, "grad_norm": 0.7016268968582153, "learning_rate": 1.2201508777020165e-05, "loss": 0.002, "step": 10851 }, { "epoch": 1.176496097137901, "grad_norm": 3.7942581176757812, "learning_rate": 1.2200783403452779e-05, "loss": 0.026, "step": 10852 }, { "epoch": 1.176604509973981, "grad_norm": 1.8771686553955078, "learning_rate": 1.2200058029885391e-05, "loss": 0.0048, "step": 10853 }, { "epoch": 1.1767129228100608, "grad_norm": 0.0887540802359581, "learning_rate": 1.2199332656318003e-05, "loss": 0.0006, "step": 10854 }, { "epoch": 1.1768213356461406, "grad_norm": 0.2689228653907776, "learning_rate": 1.2198607282750619e-05, "loss": 0.0024, "step": 10855 }, { "epoch": 1.1769297484822203, "grad_norm": 10.96830940246582, "learning_rate": 1.2197881909183231e-05, "loss": 0.0197, "step": 10856 }, { "epoch": 1.1770381613183, "grad_norm": 6.956709384918213, "learning_rate": 1.2197156535615843e-05, "loss": 0.0052, "step": 10857 }, { "epoch": 1.1771465741543798, "grad_norm": 0.8958228230476379, "learning_rate": 1.2196431162048457e-05, "loss": 0.034, "step": 10858 }, { "epoch": 1.1772549869904596, "grad_norm": 7.00978946685791, "learning_rate": 1.219570578848107e-05, "loss": 0.0154, "step": 10859 }, { "epoch": 1.1773633998265394, "grad_norm": 3.8792901039123535, "learning_rate": 1.2194980414913681e-05, "loss": 0.0217, "step": 10860 }, { "epoch": 1.1774718126626191, "grad_norm": 0.08751966804265976, "learning_rate": 1.2194255041346295e-05, "loss": 0.0006, "step": 10861 }, { "epoch": 1.1775802254986991, "grad_norm": 3.1915361881256104, "learning_rate": 1.2193529667778907e-05, "loss": 0.0092, "step": 10862 }, { "epoch": 1.1776886383347789, "grad_norm": 7.163654804229736, "learning_rate": 1.219280429421152e-05, "loss": 0.036, "step": 10863 }, { "epoch": 1.1777970511708586, "grad_norm": 8.05302619934082, "learning_rate": 1.2192078920644133e-05, "loss": 0.1024, "step": 10864 }, { "epoch": 1.1779054640069384, "grad_norm": 4.526006698608398, "learning_rate": 1.2191353547076745e-05, "loss": 0.0562, "step": 10865 }, { "epoch": 1.1780138768430182, "grad_norm": 3.059131145477295, "learning_rate": 1.2190628173509358e-05, "loss": 0.0398, "step": 10866 }, { "epoch": 1.178122289679098, "grad_norm": 2.9359989166259766, "learning_rate": 1.218990279994197e-05, "loss": 0.0234, "step": 10867 }, { "epoch": 1.1782307025151777, "grad_norm": 4.475697040557861, "learning_rate": 1.2189177426374584e-05, "loss": 0.0068, "step": 10868 }, { "epoch": 1.1783391153512577, "grad_norm": 3.2144975662231445, "learning_rate": 1.2188452052807196e-05, "loss": 0.0502, "step": 10869 }, { "epoch": 1.1784475281873374, "grad_norm": 0.6659774780273438, "learning_rate": 1.2187726679239808e-05, "loss": 0.0132, "step": 10870 }, { "epoch": 1.1785559410234172, "grad_norm": 1.982386589050293, "learning_rate": 1.2187001305672422e-05, "loss": 0.0307, "step": 10871 }, { "epoch": 1.178664353859497, "grad_norm": 0.3341493606567383, "learning_rate": 1.2186275932105036e-05, "loss": 0.0109, "step": 10872 }, { "epoch": 1.1787727666955767, "grad_norm": 0.6546970009803772, "learning_rate": 1.218555055853765e-05, "loss": 0.0035, "step": 10873 }, { "epoch": 1.1788811795316565, "grad_norm": 0.30023130774497986, "learning_rate": 1.2184825184970262e-05, "loss": 0.0032, "step": 10874 }, { "epoch": 1.1789895923677363, "grad_norm": 10.72288703918457, "learning_rate": 1.2184099811402874e-05, "loss": 0.0202, "step": 10875 }, { "epoch": 1.1790980052038162, "grad_norm": 10.014193534851074, "learning_rate": 1.2183374437835487e-05, "loss": 0.0196, "step": 10876 }, { "epoch": 1.179206418039896, "grad_norm": 0.19598403573036194, "learning_rate": 1.21826490642681e-05, "loss": 0.0012, "step": 10877 }, { "epoch": 1.1793148308759758, "grad_norm": 0.3062592148780823, "learning_rate": 1.2181923690700712e-05, "loss": 0.0095, "step": 10878 }, { "epoch": 1.1794232437120555, "grad_norm": 8.213872909545898, "learning_rate": 1.2181198317133326e-05, "loss": 0.038, "step": 10879 }, { "epoch": 1.1795316565481353, "grad_norm": 4.132203578948975, "learning_rate": 1.2180472943565938e-05, "loss": 0.0255, "step": 10880 }, { "epoch": 1.179640069384215, "grad_norm": 0.6139547824859619, "learning_rate": 1.217974756999855e-05, "loss": 0.0039, "step": 10881 }, { "epoch": 1.1797484822202948, "grad_norm": 8.8012113571167, "learning_rate": 1.2179022196431162e-05, "loss": 0.038, "step": 10882 }, { "epoch": 1.1798568950563746, "grad_norm": 0.7462108731269836, "learning_rate": 1.2178296822863776e-05, "loss": 0.0178, "step": 10883 }, { "epoch": 1.1799653078924544, "grad_norm": 7.287258625030518, "learning_rate": 1.2177571449296388e-05, "loss": 0.0189, "step": 10884 }, { "epoch": 1.1800737207285343, "grad_norm": 3.697911500930786, "learning_rate": 1.2176846075729e-05, "loss": 0.0411, "step": 10885 }, { "epoch": 1.180182133564614, "grad_norm": 0.37713754177093506, "learning_rate": 1.2176120702161614e-05, "loss": 0.0005, "step": 10886 }, { "epoch": 1.1802905464006939, "grad_norm": 0.12951703369617462, "learning_rate": 1.2175395328594226e-05, "loss": 0.0008, "step": 10887 }, { "epoch": 1.1803989592367736, "grad_norm": 0.31037598848342896, "learning_rate": 1.2174669955026838e-05, "loss": 0.0102, "step": 10888 }, { "epoch": 1.1805073720728534, "grad_norm": 2.234471559524536, "learning_rate": 1.2173944581459454e-05, "loss": 0.0337, "step": 10889 }, { "epoch": 1.1806157849089332, "grad_norm": 2.380847215652466, "learning_rate": 1.2173219207892066e-05, "loss": 0.0579, "step": 10890 }, { "epoch": 1.180724197745013, "grad_norm": 1.7546379566192627, "learning_rate": 1.217249383432468e-05, "loss": 0.0275, "step": 10891 }, { "epoch": 1.180832610581093, "grad_norm": 1.8456965684890747, "learning_rate": 1.2171768460757292e-05, "loss": 0.0028, "step": 10892 }, { "epoch": 1.1809410234171727, "grad_norm": 0.20102845132350922, "learning_rate": 1.2171043087189904e-05, "loss": 0.0015, "step": 10893 }, { "epoch": 1.1810494362532524, "grad_norm": 4.4223551750183105, "learning_rate": 1.2170317713622516e-05, "loss": 0.0218, "step": 10894 }, { "epoch": 1.1811578490893322, "grad_norm": 1.4018263816833496, "learning_rate": 1.216959234005513e-05, "loss": 0.0126, "step": 10895 }, { "epoch": 1.181266261925412, "grad_norm": 3.4012975692749023, "learning_rate": 1.2168866966487742e-05, "loss": 0.0114, "step": 10896 }, { "epoch": 1.1813746747614917, "grad_norm": 2.221717596054077, "learning_rate": 1.2168141592920354e-05, "loss": 0.0294, "step": 10897 }, { "epoch": 1.1814830875975715, "grad_norm": 0.31131094694137573, "learning_rate": 1.2167416219352968e-05, "loss": 0.0029, "step": 10898 }, { "epoch": 1.1815915004336515, "grad_norm": 2.683051347732544, "learning_rate": 1.216669084578558e-05, "loss": 0.0037, "step": 10899 }, { "epoch": 1.1816999132697312, "grad_norm": 3.269559860229492, "learning_rate": 1.2165965472218192e-05, "loss": 0.0073, "step": 10900 }, { "epoch": 1.181808326105811, "grad_norm": 0.2343035340309143, "learning_rate": 1.2165240098650806e-05, "loss": 0.0013, "step": 10901 }, { "epoch": 1.1819167389418908, "grad_norm": 0.46079105138778687, "learning_rate": 1.2164514725083418e-05, "loss": 0.0114, "step": 10902 }, { "epoch": 1.1820251517779705, "grad_norm": 5.094196319580078, "learning_rate": 1.216378935151603e-05, "loss": 0.0065, "step": 10903 }, { "epoch": 1.1821335646140503, "grad_norm": 4.8120832443237305, "learning_rate": 1.2163063977948644e-05, "loss": 0.0576, "step": 10904 }, { "epoch": 1.18224197745013, "grad_norm": 5.009016990661621, "learning_rate": 1.2162338604381257e-05, "loss": 0.0728, "step": 10905 }, { "epoch": 1.1823503902862098, "grad_norm": 6.152103900909424, "learning_rate": 1.216161323081387e-05, "loss": 0.1219, "step": 10906 }, { "epoch": 1.1824588031222896, "grad_norm": 3.2040023803710938, "learning_rate": 1.2160887857246484e-05, "loss": 0.0477, "step": 10907 }, { "epoch": 1.1825672159583696, "grad_norm": 0.3609168827533722, "learning_rate": 1.2160162483679096e-05, "loss": 0.0218, "step": 10908 }, { "epoch": 1.1826756287944493, "grad_norm": 0.7117453813552856, "learning_rate": 1.2159437110111709e-05, "loss": 0.0217, "step": 10909 }, { "epoch": 1.182784041630529, "grad_norm": 0.06043717637658119, "learning_rate": 1.2158711736544322e-05, "loss": 0.0004, "step": 10910 }, { "epoch": 1.1828924544666088, "grad_norm": 1.9262036085128784, "learning_rate": 1.2157986362976935e-05, "loss": 0.0176, "step": 10911 }, { "epoch": 1.1830008673026886, "grad_norm": 0.35704556107521057, "learning_rate": 1.2157260989409547e-05, "loss": 0.0215, "step": 10912 }, { "epoch": 1.1831092801387684, "grad_norm": 1.792039155960083, "learning_rate": 1.215653561584216e-05, "loss": 0.0457, "step": 10913 }, { "epoch": 1.1832176929748481, "grad_norm": 0.6916185617446899, "learning_rate": 1.2155810242274773e-05, "loss": 0.0197, "step": 10914 }, { "epoch": 1.1833261058109281, "grad_norm": 1.084333062171936, "learning_rate": 1.2155084868707385e-05, "loss": 0.0025, "step": 10915 }, { "epoch": 1.1834345186470079, "grad_norm": 3.895570993423462, "learning_rate": 1.2154359495139999e-05, "loss": 0.0375, "step": 10916 }, { "epoch": 1.1835429314830876, "grad_norm": 0.3156333565711975, "learning_rate": 1.215363412157261e-05, "loss": 0.0091, "step": 10917 }, { "epoch": 1.1836513443191674, "grad_norm": 2.850224018096924, "learning_rate": 1.2152908748005223e-05, "loss": 0.017, "step": 10918 }, { "epoch": 1.1837597571552472, "grad_norm": 0.7900972366333008, "learning_rate": 1.2152183374437837e-05, "loss": 0.0164, "step": 10919 }, { "epoch": 1.183868169991327, "grad_norm": 7.066811561584473, "learning_rate": 1.2151458000870449e-05, "loss": 0.0447, "step": 10920 }, { "epoch": 1.1839765828274067, "grad_norm": 2.5851099491119385, "learning_rate": 1.2150732627303061e-05, "loss": 0.0246, "step": 10921 }, { "epoch": 1.1840849956634865, "grad_norm": 0.8743633031845093, "learning_rate": 1.2150007253735673e-05, "loss": 0.0285, "step": 10922 }, { "epoch": 1.1841934084995662, "grad_norm": 1.669737696647644, "learning_rate": 1.2149281880168287e-05, "loss": 0.0355, "step": 10923 }, { "epoch": 1.1843018213356462, "grad_norm": 2.8853046894073486, "learning_rate": 1.2148556506600901e-05, "loss": 0.0554, "step": 10924 }, { "epoch": 1.184410234171726, "grad_norm": 1.0858983993530273, "learning_rate": 1.2147831133033515e-05, "loss": 0.0034, "step": 10925 }, { "epoch": 1.1845186470078057, "grad_norm": 14.820440292358398, "learning_rate": 1.2147105759466127e-05, "loss": 0.1259, "step": 10926 }, { "epoch": 1.1846270598438855, "grad_norm": 0.2620268762111664, "learning_rate": 1.2146380385898739e-05, "loss": 0.0153, "step": 10927 }, { "epoch": 1.1847354726799653, "grad_norm": 0.7475166320800781, "learning_rate": 1.2145655012331353e-05, "loss": 0.0043, "step": 10928 }, { "epoch": 1.184843885516045, "grad_norm": 3.8306727409362793, "learning_rate": 1.2144929638763965e-05, "loss": 0.0344, "step": 10929 }, { "epoch": 1.1849522983521248, "grad_norm": 5.467525959014893, "learning_rate": 1.2144204265196577e-05, "loss": 0.0598, "step": 10930 }, { "epoch": 1.1850607111882048, "grad_norm": 5.826955795288086, "learning_rate": 1.2143478891629191e-05, "loss": 0.039, "step": 10931 }, { "epoch": 1.1851691240242845, "grad_norm": 6.418527603149414, "learning_rate": 1.2142753518061803e-05, "loss": 0.0599, "step": 10932 }, { "epoch": 1.1852775368603643, "grad_norm": 2.260674238204956, "learning_rate": 1.2142028144494415e-05, "loss": 0.0281, "step": 10933 }, { "epoch": 1.185385949696444, "grad_norm": 2.884397268295288, "learning_rate": 1.2141302770927027e-05, "loss": 0.0518, "step": 10934 }, { "epoch": 1.1854943625325238, "grad_norm": 0.4027702808380127, "learning_rate": 1.2140577397359641e-05, "loss": 0.0218, "step": 10935 }, { "epoch": 1.1856027753686036, "grad_norm": 3.7839598655700684, "learning_rate": 1.2139852023792253e-05, "loss": 0.0536, "step": 10936 }, { "epoch": 1.1857111882046834, "grad_norm": 2.687976360321045, "learning_rate": 1.2139126650224866e-05, "loss": 0.0233, "step": 10937 }, { "epoch": 1.1858196010407633, "grad_norm": 0.8814360499382019, "learning_rate": 1.213840127665748e-05, "loss": 0.0165, "step": 10938 }, { "epoch": 1.185928013876843, "grad_norm": 2.107940912246704, "learning_rate": 1.2137675903090091e-05, "loss": 0.054, "step": 10939 }, { "epoch": 1.1860364267129229, "grad_norm": 1.1723990440368652, "learning_rate": 1.2136950529522704e-05, "loss": 0.0211, "step": 10940 }, { "epoch": 1.1861448395490026, "grad_norm": 0.32059043645858765, "learning_rate": 1.213622515595532e-05, "loss": 0.0084, "step": 10941 }, { "epoch": 1.1862532523850824, "grad_norm": 0.879500687122345, "learning_rate": 1.2135499782387931e-05, "loss": 0.0136, "step": 10942 }, { "epoch": 1.1863616652211622, "grad_norm": 2.659217357635498, "learning_rate": 1.2134774408820545e-05, "loss": 0.0227, "step": 10943 }, { "epoch": 1.186470078057242, "grad_norm": 2.546699047088623, "learning_rate": 1.2134049035253157e-05, "loss": 0.0348, "step": 10944 }, { "epoch": 1.1865784908933217, "grad_norm": 2.8122262954711914, "learning_rate": 1.213332366168577e-05, "loss": 0.0632, "step": 10945 }, { "epoch": 1.1866869037294014, "grad_norm": 3.2642264366149902, "learning_rate": 1.2132598288118382e-05, "loss": 0.0206, "step": 10946 }, { "epoch": 1.1867953165654814, "grad_norm": 0.4182312786579132, "learning_rate": 1.2131872914550995e-05, "loss": 0.0039, "step": 10947 }, { "epoch": 1.1869037294015612, "grad_norm": 0.17622017860412598, "learning_rate": 1.2131147540983608e-05, "loss": 0.0116, "step": 10948 }, { "epoch": 1.187012142237641, "grad_norm": 0.49784812331199646, "learning_rate": 1.213042216741622e-05, "loss": 0.0246, "step": 10949 }, { "epoch": 1.1871205550737207, "grad_norm": 1.5747320652008057, "learning_rate": 1.2129696793848834e-05, "loss": 0.0276, "step": 10950 }, { "epoch": 1.1872289679098005, "grad_norm": 1.8145320415496826, "learning_rate": 1.2128971420281446e-05, "loss": 0.0247, "step": 10951 }, { "epoch": 1.1873373807458802, "grad_norm": 1.1935433149337769, "learning_rate": 1.2128246046714058e-05, "loss": 0.0274, "step": 10952 }, { "epoch": 1.18744579358196, "grad_norm": 0.9010871648788452, "learning_rate": 1.2127520673146672e-05, "loss": 0.0131, "step": 10953 }, { "epoch": 1.18755420641804, "grad_norm": 3.1622843742370605, "learning_rate": 1.2126795299579284e-05, "loss": 0.0515, "step": 10954 }, { "epoch": 1.1876626192541198, "grad_norm": 0.29802218079566956, "learning_rate": 1.2126069926011896e-05, "loss": 0.0066, "step": 10955 }, { "epoch": 1.1877710320901995, "grad_norm": 0.9158298969268799, "learning_rate": 1.212534455244451e-05, "loss": 0.0096, "step": 10956 }, { "epoch": 1.1878794449262793, "grad_norm": 5.20123291015625, "learning_rate": 1.2124619178877122e-05, "loss": 0.0422, "step": 10957 }, { "epoch": 1.187987857762359, "grad_norm": 3.7382006645202637, "learning_rate": 1.2123893805309736e-05, "loss": 0.0167, "step": 10958 }, { "epoch": 1.1880962705984388, "grad_norm": 3.0011677742004395, "learning_rate": 1.212316843174235e-05, "loss": 0.0297, "step": 10959 }, { "epoch": 1.1882046834345186, "grad_norm": 0.059563469141721725, "learning_rate": 1.2122443058174962e-05, "loss": 0.0007, "step": 10960 }, { "epoch": 1.1883130962705986, "grad_norm": 3.0457656383514404, "learning_rate": 1.2121717684607574e-05, "loss": 0.0352, "step": 10961 }, { "epoch": 1.1884215091066783, "grad_norm": 0.887307345867157, "learning_rate": 1.2120992311040188e-05, "loss": 0.007, "step": 10962 }, { "epoch": 1.188529921942758, "grad_norm": 2.325676441192627, "learning_rate": 1.21202669374728e-05, "loss": 0.0101, "step": 10963 }, { "epoch": 1.1886383347788378, "grad_norm": 6.690056800842285, "learning_rate": 1.2119541563905412e-05, "loss": 0.0191, "step": 10964 }, { "epoch": 1.1887467476149176, "grad_norm": 1.8523470163345337, "learning_rate": 1.2118816190338026e-05, "loss": 0.0161, "step": 10965 }, { "epoch": 1.1888551604509974, "grad_norm": 0.20353230834007263, "learning_rate": 1.2118090816770638e-05, "loss": 0.002, "step": 10966 }, { "epoch": 1.1889635732870771, "grad_norm": 1.2181720733642578, "learning_rate": 1.211736544320325e-05, "loss": 0.0063, "step": 10967 }, { "epoch": 1.189071986123157, "grad_norm": 2.6880710124969482, "learning_rate": 1.2116640069635864e-05, "loss": 0.0127, "step": 10968 }, { "epoch": 1.1891803989592367, "grad_norm": 0.8958750367164612, "learning_rate": 1.2115914696068476e-05, "loss": 0.0171, "step": 10969 }, { "epoch": 1.1892888117953166, "grad_norm": 0.14274784922599792, "learning_rate": 1.2115189322501088e-05, "loss": 0.0015, "step": 10970 }, { "epoch": 1.1893972246313964, "grad_norm": 2.454131603240967, "learning_rate": 1.2114463948933702e-05, "loss": 0.0382, "step": 10971 }, { "epoch": 1.1895056374674762, "grad_norm": 5.155940055847168, "learning_rate": 1.2113738575366314e-05, "loss": 0.0501, "step": 10972 }, { "epoch": 1.189614050303556, "grad_norm": 0.08362032473087311, "learning_rate": 1.2113013201798926e-05, "loss": 0.0009, "step": 10973 }, { "epoch": 1.1897224631396357, "grad_norm": 12.289737701416016, "learning_rate": 1.2112287828231539e-05, "loss": 0.0402, "step": 10974 }, { "epoch": 1.1898308759757155, "grad_norm": 0.8616237640380859, "learning_rate": 1.2111562454664152e-05, "loss": 0.0102, "step": 10975 }, { "epoch": 1.1899392888117952, "grad_norm": 0.48096466064453125, "learning_rate": 1.2110837081096766e-05, "loss": 0.0029, "step": 10976 }, { "epoch": 1.1900477016478752, "grad_norm": 0.4555591940879822, "learning_rate": 1.211011170752938e-05, "loss": 0.0039, "step": 10977 }, { "epoch": 1.190156114483955, "grad_norm": 0.33990171551704407, "learning_rate": 1.2109386333961992e-05, "loss": 0.0101, "step": 10978 }, { "epoch": 1.1902645273200347, "grad_norm": 1.3741071224212646, "learning_rate": 1.2108660960394604e-05, "loss": 0.0278, "step": 10979 }, { "epoch": 1.1903729401561145, "grad_norm": 1.7862639427185059, "learning_rate": 1.2107935586827218e-05, "loss": 0.0146, "step": 10980 }, { "epoch": 1.1904813529921943, "grad_norm": 2.0177664756774902, "learning_rate": 1.210721021325983e-05, "loss": 0.0311, "step": 10981 }, { "epoch": 1.190589765828274, "grad_norm": 0.368604451417923, "learning_rate": 1.2106484839692442e-05, "loss": 0.0019, "step": 10982 }, { "epoch": 1.1906981786643538, "grad_norm": 0.4544275999069214, "learning_rate": 1.2105759466125056e-05, "loss": 0.0124, "step": 10983 }, { "epoch": 1.1908065915004338, "grad_norm": 2.0449514389038086, "learning_rate": 1.2105034092557668e-05, "loss": 0.0189, "step": 10984 }, { "epoch": 1.1909150043365135, "grad_norm": 5.120380878448486, "learning_rate": 1.210430871899028e-05, "loss": 0.0501, "step": 10985 }, { "epoch": 1.1910234171725933, "grad_norm": 4.105486869812012, "learning_rate": 1.2103583345422893e-05, "loss": 0.0126, "step": 10986 }, { "epoch": 1.191131830008673, "grad_norm": 0.40995872020721436, "learning_rate": 1.2102857971855507e-05, "loss": 0.0077, "step": 10987 }, { "epoch": 1.1912402428447528, "grad_norm": 3.384737491607666, "learning_rate": 1.2102132598288119e-05, "loss": 0.0206, "step": 10988 }, { "epoch": 1.1913486556808326, "grad_norm": 0.7105897068977356, "learning_rate": 1.210140722472073e-05, "loss": 0.019, "step": 10989 }, { "epoch": 1.1914570685169124, "grad_norm": 4.274063587188721, "learning_rate": 1.2100681851153345e-05, "loss": 0.0276, "step": 10990 }, { "epoch": 1.1915654813529921, "grad_norm": 0.8367263674736023, "learning_rate": 1.2099956477585957e-05, "loss": 0.0172, "step": 10991 }, { "epoch": 1.1916738941890719, "grad_norm": 2.309887170791626, "learning_rate": 1.2099231104018569e-05, "loss": 0.0182, "step": 10992 }, { "epoch": 1.1917823070251519, "grad_norm": 5.156077861785889, "learning_rate": 1.2098505730451184e-05, "loss": 0.0336, "step": 10993 }, { "epoch": 1.1918907198612316, "grad_norm": 0.4786612391471863, "learning_rate": 1.2097780356883797e-05, "loss": 0.0144, "step": 10994 }, { "epoch": 1.1919991326973114, "grad_norm": 4.224212169647217, "learning_rate": 1.209705498331641e-05, "loss": 0.059, "step": 10995 }, { "epoch": 1.1921075455333912, "grad_norm": 2.32861328125, "learning_rate": 1.2096329609749023e-05, "loss": 0.0517, "step": 10996 }, { "epoch": 1.192215958369471, "grad_norm": 4.917468070983887, "learning_rate": 1.2095604236181635e-05, "loss": 0.0227, "step": 10997 }, { "epoch": 1.1923243712055507, "grad_norm": 3.087841510772705, "learning_rate": 1.2094878862614247e-05, "loss": 0.0251, "step": 10998 }, { "epoch": 1.1924327840416304, "grad_norm": 1.2302680015563965, "learning_rate": 1.209415348904686e-05, "loss": 0.0079, "step": 10999 }, { "epoch": 1.1925411968777104, "grad_norm": 2.1245408058166504, "learning_rate": 1.2093428115479473e-05, "loss": 0.0115, "step": 11000 }, { "epoch": 1.1926496097137902, "grad_norm": 24.662195205688477, "learning_rate": 1.2092702741912085e-05, "loss": 0.0121, "step": 11001 }, { "epoch": 1.19275802254987, "grad_norm": 6.448363780975342, "learning_rate": 1.2091977368344699e-05, "loss": 0.0358, "step": 11002 }, { "epoch": 1.1928664353859497, "grad_norm": 1.925129771232605, "learning_rate": 1.2091251994777311e-05, "loss": 0.0516, "step": 11003 }, { "epoch": 1.1929748482220295, "grad_norm": 4.067893028259277, "learning_rate": 1.2090526621209923e-05, "loss": 0.0068, "step": 11004 }, { "epoch": 1.1930832610581092, "grad_norm": 1.386080026626587, "learning_rate": 1.2089801247642537e-05, "loss": 0.0307, "step": 11005 }, { "epoch": 1.193191673894189, "grad_norm": 2.6034488677978516, "learning_rate": 1.2089075874075149e-05, "loss": 0.0182, "step": 11006 }, { "epoch": 1.1933000867302688, "grad_norm": 3.3347408771514893, "learning_rate": 1.2088350500507761e-05, "loss": 0.0583, "step": 11007 }, { "epoch": 1.1934084995663485, "grad_norm": 2.179253339767456, "learning_rate": 1.2087625126940375e-05, "loss": 0.0416, "step": 11008 }, { "epoch": 1.1935169124024285, "grad_norm": 3.2642171382904053, "learning_rate": 1.2086899753372987e-05, "loss": 0.0204, "step": 11009 }, { "epoch": 1.1936253252385083, "grad_norm": 4.703514099121094, "learning_rate": 1.2086174379805601e-05, "loss": 0.0603, "step": 11010 }, { "epoch": 1.193733738074588, "grad_norm": 0.9097206592559814, "learning_rate": 1.2085449006238215e-05, "loss": 0.0196, "step": 11011 }, { "epoch": 1.1938421509106678, "grad_norm": 5.892024993896484, "learning_rate": 1.2084723632670827e-05, "loss": 0.0769, "step": 11012 }, { "epoch": 1.1939505637467476, "grad_norm": 2.8726966381073, "learning_rate": 1.208399825910344e-05, "loss": 0.0214, "step": 11013 }, { "epoch": 1.1940589765828273, "grad_norm": 1.6018805503845215, "learning_rate": 1.2083272885536053e-05, "loss": 0.0203, "step": 11014 }, { "epoch": 1.194167389418907, "grad_norm": 2.1942832469940186, "learning_rate": 1.2082547511968665e-05, "loss": 0.0191, "step": 11015 }, { "epoch": 1.194275802254987, "grad_norm": 0.659352719783783, "learning_rate": 1.2081822138401277e-05, "loss": 0.0203, "step": 11016 }, { "epoch": 1.1943842150910668, "grad_norm": 0.25806593894958496, "learning_rate": 1.2081096764833891e-05, "loss": 0.0056, "step": 11017 }, { "epoch": 1.1944926279271466, "grad_norm": 2.389418840408325, "learning_rate": 1.2080371391266503e-05, "loss": 0.0367, "step": 11018 }, { "epoch": 1.1946010407632264, "grad_norm": 0.055011067539453506, "learning_rate": 1.2079646017699115e-05, "loss": 0.0006, "step": 11019 }, { "epoch": 1.1947094535993061, "grad_norm": 4.802459716796875, "learning_rate": 1.207892064413173e-05, "loss": 0.0092, "step": 11020 }, { "epoch": 1.194817866435386, "grad_norm": 1.408866047859192, "learning_rate": 1.2078195270564341e-05, "loss": 0.0175, "step": 11021 }, { "epoch": 1.1949262792714657, "grad_norm": 2.361346483230591, "learning_rate": 1.2077469896996954e-05, "loss": 0.0227, "step": 11022 }, { "epoch": 1.1950346921075456, "grad_norm": 5.451864719390869, "learning_rate": 1.2076744523429567e-05, "loss": 0.0703, "step": 11023 }, { "epoch": 1.1951431049436254, "grad_norm": 3.064772367477417, "learning_rate": 1.207601914986218e-05, "loss": 0.0282, "step": 11024 }, { "epoch": 1.1952515177797052, "grad_norm": 6.420792579650879, "learning_rate": 1.2075293776294792e-05, "loss": 0.0227, "step": 11025 }, { "epoch": 1.195359930615785, "grad_norm": 0.5299970507621765, "learning_rate": 1.2074568402727404e-05, "loss": 0.0158, "step": 11026 }, { "epoch": 1.1954683434518647, "grad_norm": 2.2303617000579834, "learning_rate": 1.207384302916002e-05, "loss": 0.0218, "step": 11027 }, { "epoch": 1.1955767562879445, "grad_norm": 0.42438745498657227, "learning_rate": 1.2073117655592632e-05, "loss": 0.0107, "step": 11028 }, { "epoch": 1.1956851691240242, "grad_norm": 5.094810485839844, "learning_rate": 1.2072392282025245e-05, "loss": 0.0735, "step": 11029 }, { "epoch": 1.195793581960104, "grad_norm": 0.3138744533061981, "learning_rate": 1.2071666908457858e-05, "loss": 0.0066, "step": 11030 }, { "epoch": 1.1959019947961838, "grad_norm": 0.43891438841819763, "learning_rate": 1.207094153489047e-05, "loss": 0.0031, "step": 11031 }, { "epoch": 1.1960104076322637, "grad_norm": 3.9607303142547607, "learning_rate": 1.2070216161323083e-05, "loss": 0.0112, "step": 11032 }, { "epoch": 1.1961188204683435, "grad_norm": 2.4908668994903564, "learning_rate": 1.2069490787755696e-05, "loss": 0.0559, "step": 11033 }, { "epoch": 1.1962272333044233, "grad_norm": 1.0769332647323608, "learning_rate": 1.2068765414188308e-05, "loss": 0.0338, "step": 11034 }, { "epoch": 1.196335646140503, "grad_norm": 0.5685699582099915, "learning_rate": 1.2068040040620922e-05, "loss": 0.0131, "step": 11035 }, { "epoch": 1.1964440589765828, "grad_norm": 3.6591103076934814, "learning_rate": 1.2067314667053534e-05, "loss": 0.0291, "step": 11036 }, { "epoch": 1.1965524718126626, "grad_norm": 0.5752103328704834, "learning_rate": 1.2066589293486146e-05, "loss": 0.0053, "step": 11037 }, { "epoch": 1.1966608846487423, "grad_norm": 1.9198836088180542, "learning_rate": 1.2065863919918758e-05, "loss": 0.041, "step": 11038 }, { "epoch": 1.1967692974848223, "grad_norm": 0.8623551726341248, "learning_rate": 1.2065138546351372e-05, "loss": 0.0131, "step": 11039 }, { "epoch": 1.196877710320902, "grad_norm": 5.035588264465332, "learning_rate": 1.2064413172783984e-05, "loss": 0.025, "step": 11040 }, { "epoch": 1.1969861231569818, "grad_norm": 2.072519302368164, "learning_rate": 1.2063687799216596e-05, "loss": 0.0278, "step": 11041 }, { "epoch": 1.1970945359930616, "grad_norm": 3.993783712387085, "learning_rate": 1.206296242564921e-05, "loss": 0.0288, "step": 11042 }, { "epoch": 1.1972029488291414, "grad_norm": 0.1339857280254364, "learning_rate": 1.2062237052081822e-05, "loss": 0.0014, "step": 11043 }, { "epoch": 1.1973113616652211, "grad_norm": 2.5578503608703613, "learning_rate": 1.2061511678514434e-05, "loss": 0.0328, "step": 11044 }, { "epoch": 1.1974197745013009, "grad_norm": 0.6887350678443909, "learning_rate": 1.206078630494705e-05, "loss": 0.0206, "step": 11045 }, { "epoch": 1.1975281873373809, "grad_norm": 1.3040188550949097, "learning_rate": 1.2060060931379662e-05, "loss": 0.0097, "step": 11046 }, { "epoch": 1.1976366001734606, "grad_norm": 2.50569224357605, "learning_rate": 1.2059335557812276e-05, "loss": 0.0389, "step": 11047 }, { "epoch": 1.1977450130095404, "grad_norm": 2.4984655380249023, "learning_rate": 1.2058610184244888e-05, "loss": 0.0348, "step": 11048 }, { "epoch": 1.1978534258456202, "grad_norm": 0.7843446135520935, "learning_rate": 1.20578848106775e-05, "loss": 0.0101, "step": 11049 }, { "epoch": 1.1979618386817, "grad_norm": 2.0557878017425537, "learning_rate": 1.2057159437110112e-05, "loss": 0.032, "step": 11050 }, { "epoch": 1.1980702515177797, "grad_norm": 13.157063484191895, "learning_rate": 1.2056434063542726e-05, "loss": 0.0511, "step": 11051 }, { "epoch": 1.1981786643538594, "grad_norm": 4.200435638427734, "learning_rate": 1.2055708689975338e-05, "loss": 0.0296, "step": 11052 }, { "epoch": 1.1982870771899392, "grad_norm": 0.15218237042427063, "learning_rate": 1.205498331640795e-05, "loss": 0.0017, "step": 11053 }, { "epoch": 1.198395490026019, "grad_norm": 4.608287334442139, "learning_rate": 1.2054257942840564e-05, "loss": 0.0225, "step": 11054 }, { "epoch": 1.198503902862099, "grad_norm": 0.7370682954788208, "learning_rate": 1.2053532569273176e-05, "loss": 0.0075, "step": 11055 }, { "epoch": 1.1986123156981787, "grad_norm": 0.6042704582214355, "learning_rate": 1.2052807195705788e-05, "loss": 0.0124, "step": 11056 }, { "epoch": 1.1987207285342585, "grad_norm": 3.500532388687134, "learning_rate": 1.2052081822138402e-05, "loss": 0.0766, "step": 11057 }, { "epoch": 1.1988291413703382, "grad_norm": 0.7267184853553772, "learning_rate": 1.2051356448571014e-05, "loss": 0.003, "step": 11058 }, { "epoch": 1.198937554206418, "grad_norm": 0.5785245895385742, "learning_rate": 1.2050631075003627e-05, "loss": 0.0211, "step": 11059 }, { "epoch": 1.1990459670424978, "grad_norm": 1.1978647708892822, "learning_rate": 1.204990570143624e-05, "loss": 0.03, "step": 11060 }, { "epoch": 1.1991543798785775, "grad_norm": 0.9000385999679565, "learning_rate": 1.2049180327868853e-05, "loss": 0.0035, "step": 11061 }, { "epoch": 1.1992627927146575, "grad_norm": 2.784205198287964, "learning_rate": 1.2048454954301466e-05, "loss": 0.0312, "step": 11062 }, { "epoch": 1.1993712055507373, "grad_norm": 1.293662428855896, "learning_rate": 1.204772958073408e-05, "loss": 0.0127, "step": 11063 }, { "epoch": 1.199479618386817, "grad_norm": 8.205911636352539, "learning_rate": 1.2047004207166692e-05, "loss": 0.087, "step": 11064 }, { "epoch": 1.1995880312228968, "grad_norm": 2.4011430740356445, "learning_rate": 1.2046278833599305e-05, "loss": 0.0703, "step": 11065 }, { "epoch": 1.1996964440589766, "grad_norm": 1.6516951322555542, "learning_rate": 1.2045553460031918e-05, "loss": 0.0463, "step": 11066 }, { "epoch": 1.1998048568950563, "grad_norm": 1.2065579891204834, "learning_rate": 1.204482808646453e-05, "loss": 0.0077, "step": 11067 }, { "epoch": 1.199913269731136, "grad_norm": 1.9163798093795776, "learning_rate": 1.2044102712897143e-05, "loss": 0.0292, "step": 11068 }, { "epoch": 1.2000216825672159, "grad_norm": 0.7030028700828552, "learning_rate": 1.2043377339329757e-05, "loss": 0.0055, "step": 11069 }, { "epoch": 1.2001300954032956, "grad_norm": 1.4296965599060059, "learning_rate": 1.2042651965762369e-05, "loss": 0.0062, "step": 11070 }, { "epoch": 1.2002385082393756, "grad_norm": 2.6811556816101074, "learning_rate": 1.204192659219498e-05, "loss": 0.0367, "step": 11071 }, { "epoch": 1.2003469210754554, "grad_norm": 0.7470754981040955, "learning_rate": 1.2041201218627595e-05, "loss": 0.0078, "step": 11072 }, { "epoch": 1.2004553339115351, "grad_norm": 2.928356885910034, "learning_rate": 1.2040475845060207e-05, "loss": 0.0292, "step": 11073 }, { "epoch": 1.200563746747615, "grad_norm": 1.0612714290618896, "learning_rate": 1.2039750471492819e-05, "loss": 0.0132, "step": 11074 }, { "epoch": 1.2006721595836947, "grad_norm": 0.32056620717048645, "learning_rate": 1.2039025097925433e-05, "loss": 0.0096, "step": 11075 }, { "epoch": 1.2007805724197744, "grad_norm": 1.3226492404937744, "learning_rate": 1.2038299724358045e-05, "loss": 0.0212, "step": 11076 }, { "epoch": 1.2008889852558542, "grad_norm": 0.9794641137123108, "learning_rate": 1.2037574350790657e-05, "loss": 0.0213, "step": 11077 }, { "epoch": 1.2009973980919342, "grad_norm": 0.4109565317630768, "learning_rate": 1.203684897722327e-05, "loss": 0.0119, "step": 11078 }, { "epoch": 1.201105810928014, "grad_norm": 2.371436834335327, "learning_rate": 1.2036123603655885e-05, "loss": 0.0109, "step": 11079 }, { "epoch": 1.2012142237640937, "grad_norm": 2.994112253189087, "learning_rate": 1.2035398230088497e-05, "loss": 0.0428, "step": 11080 }, { "epoch": 1.2013226366001735, "grad_norm": 2.4719321727752686, "learning_rate": 1.203467285652111e-05, "loss": 0.0094, "step": 11081 }, { "epoch": 1.2014310494362532, "grad_norm": 0.18991950154304504, "learning_rate": 1.2033947482953723e-05, "loss": 0.002, "step": 11082 }, { "epoch": 1.201539462272333, "grad_norm": 4.1416401863098145, "learning_rate": 1.2033222109386335e-05, "loss": 0.0269, "step": 11083 }, { "epoch": 1.2016478751084128, "grad_norm": 0.7135887742042542, "learning_rate": 1.2032496735818949e-05, "loss": 0.0027, "step": 11084 }, { "epoch": 1.2017562879444927, "grad_norm": 1.8020752668380737, "learning_rate": 1.2031771362251561e-05, "loss": 0.0105, "step": 11085 }, { "epoch": 1.2018647007805725, "grad_norm": 0.6090044975280762, "learning_rate": 1.2031045988684173e-05, "loss": 0.0018, "step": 11086 }, { "epoch": 1.2019731136166523, "grad_norm": 0.8841310143470764, "learning_rate": 1.2030320615116787e-05, "loss": 0.0064, "step": 11087 }, { "epoch": 1.202081526452732, "grad_norm": 2.3528318405151367, "learning_rate": 1.2029595241549399e-05, "loss": 0.015, "step": 11088 }, { "epoch": 1.2021899392888118, "grad_norm": 1.2569005489349365, "learning_rate": 1.2028869867982011e-05, "loss": 0.0127, "step": 11089 }, { "epoch": 1.2022983521248916, "grad_norm": 2.019397497177124, "learning_rate": 1.2028144494414623e-05, "loss": 0.0214, "step": 11090 }, { "epoch": 1.2024067649609713, "grad_norm": 2.1946232318878174, "learning_rate": 1.2027419120847237e-05, "loss": 0.022, "step": 11091 }, { "epoch": 1.202515177797051, "grad_norm": 3.3470282554626465, "learning_rate": 1.202669374727985e-05, "loss": 0.0296, "step": 11092 }, { "epoch": 1.2026235906331308, "grad_norm": 0.12353949248790741, "learning_rate": 1.2025968373712462e-05, "loss": 0.001, "step": 11093 }, { "epoch": 1.2027320034692108, "grad_norm": 3.7096588611602783, "learning_rate": 1.2025243000145075e-05, "loss": 0.0532, "step": 11094 }, { "epoch": 1.2028404163052906, "grad_norm": 0.5632321238517761, "learning_rate": 1.2024517626577687e-05, "loss": 0.0087, "step": 11095 }, { "epoch": 1.2029488291413704, "grad_norm": 1.736954689025879, "learning_rate": 1.2023792253010303e-05, "loss": 0.0435, "step": 11096 }, { "epoch": 1.2030572419774501, "grad_norm": 2.3007709980010986, "learning_rate": 1.2023066879442915e-05, "loss": 0.0131, "step": 11097 }, { "epoch": 1.2031656548135299, "grad_norm": 1.231887936592102, "learning_rate": 1.2022341505875527e-05, "loss": 0.0187, "step": 11098 }, { "epoch": 1.2032740676496096, "grad_norm": 3.8786652088165283, "learning_rate": 1.2021616132308141e-05, "loss": 0.0062, "step": 11099 }, { "epoch": 1.2033824804856894, "grad_norm": 0.9000112414360046, "learning_rate": 1.2020890758740753e-05, "loss": 0.0247, "step": 11100 }, { "epoch": 1.2034908933217694, "grad_norm": 0.4238200783729553, "learning_rate": 1.2020165385173365e-05, "loss": 0.0119, "step": 11101 }, { "epoch": 1.2035993061578492, "grad_norm": 2.634173631668091, "learning_rate": 1.2019440011605978e-05, "loss": 0.0123, "step": 11102 }, { "epoch": 1.203707718993929, "grad_norm": 1.3709951639175415, "learning_rate": 1.2018714638038591e-05, "loss": 0.011, "step": 11103 }, { "epoch": 1.2038161318300087, "grad_norm": 0.2867696285247803, "learning_rate": 1.2017989264471204e-05, "loss": 0.0052, "step": 11104 }, { "epoch": 1.2039245446660884, "grad_norm": 5.042664527893066, "learning_rate": 1.2017263890903816e-05, "loss": 0.0802, "step": 11105 }, { "epoch": 1.2040329575021682, "grad_norm": 0.06679622083902359, "learning_rate": 1.201653851733643e-05, "loss": 0.0003, "step": 11106 }, { "epoch": 1.204141370338248, "grad_norm": 0.27367928624153137, "learning_rate": 1.2015813143769042e-05, "loss": 0.0014, "step": 11107 }, { "epoch": 1.204249783174328, "grad_norm": 2.252990245819092, "learning_rate": 1.2015087770201654e-05, "loss": 0.0097, "step": 11108 }, { "epoch": 1.2043581960104077, "grad_norm": 7.381346702575684, "learning_rate": 1.2014362396634268e-05, "loss": 0.0327, "step": 11109 }, { "epoch": 1.2044666088464875, "grad_norm": 0.4334496557712555, "learning_rate": 1.201363702306688e-05, "loss": 0.0017, "step": 11110 }, { "epoch": 1.2045750216825672, "grad_norm": 1.4541051387786865, "learning_rate": 1.2012911649499492e-05, "loss": 0.0402, "step": 11111 }, { "epoch": 1.204683434518647, "grad_norm": 3.1589713096618652, "learning_rate": 1.2012186275932106e-05, "loss": 0.0351, "step": 11112 }, { "epoch": 1.2047918473547268, "grad_norm": 8.28279972076416, "learning_rate": 1.2011460902364718e-05, "loss": 0.0966, "step": 11113 }, { "epoch": 1.2049002601908065, "grad_norm": 6.155050277709961, "learning_rate": 1.2010735528797333e-05, "loss": 0.0376, "step": 11114 }, { "epoch": 1.2050086730268863, "grad_norm": 1.2160016298294067, "learning_rate": 1.2010010155229946e-05, "loss": 0.0242, "step": 11115 }, { "epoch": 1.205117085862966, "grad_norm": 0.41815030574798584, "learning_rate": 1.2009284781662558e-05, "loss": 0.023, "step": 11116 }, { "epoch": 1.205225498699046, "grad_norm": 4.199515342712402, "learning_rate": 1.200855940809517e-05, "loss": 0.0262, "step": 11117 }, { "epoch": 1.2053339115351258, "grad_norm": 0.4128738045692444, "learning_rate": 1.2007834034527784e-05, "loss": 0.0118, "step": 11118 }, { "epoch": 1.2054423243712056, "grad_norm": 0.3732626438140869, "learning_rate": 1.2007108660960396e-05, "loss": 0.0124, "step": 11119 }, { "epoch": 1.2055507372072853, "grad_norm": 0.36236023902893066, "learning_rate": 1.2006383287393008e-05, "loss": 0.0103, "step": 11120 }, { "epoch": 1.205659150043365, "grad_norm": 0.47890612483024597, "learning_rate": 1.2005657913825622e-05, "loss": 0.0089, "step": 11121 }, { "epoch": 1.2057675628794449, "grad_norm": 0.3836487829685211, "learning_rate": 1.2004932540258234e-05, "loss": 0.017, "step": 11122 }, { "epoch": 1.2058759757155246, "grad_norm": 3.8176584243774414, "learning_rate": 1.2004207166690846e-05, "loss": 0.036, "step": 11123 }, { "epoch": 1.2059843885516046, "grad_norm": 1.4744937419891357, "learning_rate": 1.200348179312346e-05, "loss": 0.0375, "step": 11124 }, { "epoch": 1.2060928013876844, "grad_norm": 4.035163402557373, "learning_rate": 1.2002756419556072e-05, "loss": 0.0399, "step": 11125 }, { "epoch": 1.2062012142237641, "grad_norm": 1.4305086135864258, "learning_rate": 1.2002031045988684e-05, "loss": 0.0046, "step": 11126 }, { "epoch": 1.206309627059844, "grad_norm": 1.3187066316604614, "learning_rate": 1.2001305672421298e-05, "loss": 0.0194, "step": 11127 }, { "epoch": 1.2064180398959237, "grad_norm": 5.906400680541992, "learning_rate": 1.200058029885391e-05, "loss": 0.0277, "step": 11128 }, { "epoch": 1.2065264527320034, "grad_norm": 1.969512701034546, "learning_rate": 1.1999854925286522e-05, "loss": 0.0185, "step": 11129 }, { "epoch": 1.2066348655680832, "grad_norm": 0.4686046838760376, "learning_rate": 1.1999129551719135e-05, "loss": 0.0122, "step": 11130 }, { "epoch": 1.2067432784041632, "grad_norm": 3.2201521396636963, "learning_rate": 1.199840417815175e-05, "loss": 0.0251, "step": 11131 }, { "epoch": 1.206851691240243, "grad_norm": 2.4350461959838867, "learning_rate": 1.1997678804584362e-05, "loss": 0.0549, "step": 11132 }, { "epoch": 1.2069601040763227, "grad_norm": 2.6204869747161865, "learning_rate": 1.1996953431016976e-05, "loss": 0.024, "step": 11133 }, { "epoch": 1.2070685169124025, "grad_norm": 0.6625131368637085, "learning_rate": 1.1996228057449588e-05, "loss": 0.0106, "step": 11134 }, { "epoch": 1.2071769297484822, "grad_norm": 0.8440901637077332, "learning_rate": 1.19955026838822e-05, "loss": 0.0105, "step": 11135 }, { "epoch": 1.207285342584562, "grad_norm": 2.118129253387451, "learning_rate": 1.1994777310314814e-05, "loss": 0.0132, "step": 11136 }, { "epoch": 1.2073937554206418, "grad_norm": 1.0675232410430908, "learning_rate": 1.1994051936747426e-05, "loss": 0.0064, "step": 11137 }, { "epoch": 1.2075021682567215, "grad_norm": 6.575364589691162, "learning_rate": 1.1993326563180038e-05, "loss": 0.0314, "step": 11138 }, { "epoch": 1.2076105810928013, "grad_norm": 1.3280353546142578, "learning_rate": 1.1992601189612652e-05, "loss": 0.0319, "step": 11139 }, { "epoch": 1.2077189939288813, "grad_norm": 2.69642972946167, "learning_rate": 1.1991875816045264e-05, "loss": 0.0564, "step": 11140 }, { "epoch": 1.207827406764961, "grad_norm": 0.684725284576416, "learning_rate": 1.1991150442477877e-05, "loss": 0.0349, "step": 11141 }, { "epoch": 1.2079358196010408, "grad_norm": 3.442523241043091, "learning_rate": 1.1990425068910489e-05, "loss": 0.0302, "step": 11142 }, { "epoch": 1.2080442324371206, "grad_norm": 0.36989283561706543, "learning_rate": 1.1989699695343103e-05, "loss": 0.0096, "step": 11143 }, { "epoch": 1.2081526452732003, "grad_norm": 7.55681037902832, "learning_rate": 1.1988974321775715e-05, "loss": 0.034, "step": 11144 }, { "epoch": 1.20826105810928, "grad_norm": 3.560567855834961, "learning_rate": 1.1988248948208327e-05, "loss": 0.0531, "step": 11145 }, { "epoch": 1.2083694709453598, "grad_norm": 3.4074883460998535, "learning_rate": 1.198752357464094e-05, "loss": 0.0217, "step": 11146 }, { "epoch": 1.2084778837814398, "grad_norm": 0.18900258839130402, "learning_rate": 1.1986798201073553e-05, "loss": 0.0017, "step": 11147 }, { "epoch": 1.2085862966175196, "grad_norm": 4.764148712158203, "learning_rate": 1.1986072827506168e-05, "loss": 0.0306, "step": 11148 }, { "epoch": 1.2086947094535994, "grad_norm": 4.929210186004639, "learning_rate": 1.198534745393878e-05, "loss": 0.0496, "step": 11149 }, { "epoch": 1.2088031222896791, "grad_norm": 1.8535611629486084, "learning_rate": 1.1984622080371393e-05, "loss": 0.0054, "step": 11150 }, { "epoch": 1.2089115351257589, "grad_norm": 1.0730507373809814, "learning_rate": 1.1983896706804006e-05, "loss": 0.0257, "step": 11151 }, { "epoch": 1.2090199479618386, "grad_norm": 2.8409624099731445, "learning_rate": 1.1983171333236619e-05, "loss": 0.0557, "step": 11152 }, { "epoch": 1.2091283607979184, "grad_norm": 0.4091412425041199, "learning_rate": 1.198244595966923e-05, "loss": 0.013, "step": 11153 }, { "epoch": 1.2092367736339982, "grad_norm": 0.2576581835746765, "learning_rate": 1.1981720586101843e-05, "loss": 0.0011, "step": 11154 }, { "epoch": 1.209345186470078, "grad_norm": 10.303729057312012, "learning_rate": 1.1980995212534457e-05, "loss": 0.0117, "step": 11155 }, { "epoch": 1.209453599306158, "grad_norm": 0.7999076247215271, "learning_rate": 1.1980269838967069e-05, "loss": 0.0124, "step": 11156 }, { "epoch": 1.2095620121422377, "grad_norm": 0.4499988853931427, "learning_rate": 1.1979544465399681e-05, "loss": 0.0156, "step": 11157 }, { "epoch": 1.2096704249783174, "grad_norm": 3.5753543376922607, "learning_rate": 1.1978819091832295e-05, "loss": 0.0659, "step": 11158 }, { "epoch": 1.2097788378143972, "grad_norm": 9.602116584777832, "learning_rate": 1.1978093718264907e-05, "loss": 0.0163, "step": 11159 }, { "epoch": 1.209887250650477, "grad_norm": 2.233006715774536, "learning_rate": 1.197736834469752e-05, "loss": 0.0375, "step": 11160 }, { "epoch": 1.2099956634865567, "grad_norm": 1.624811053276062, "learning_rate": 1.1976642971130133e-05, "loss": 0.0276, "step": 11161 }, { "epoch": 1.2101040763226365, "grad_norm": 2.3312394618988037, "learning_rate": 1.1975917597562745e-05, "loss": 0.0176, "step": 11162 }, { "epoch": 1.2102124891587165, "grad_norm": 1.625910997390747, "learning_rate": 1.1975192223995357e-05, "loss": 0.0742, "step": 11163 }, { "epoch": 1.2103209019947962, "grad_norm": 1.2703801393508911, "learning_rate": 1.1974466850427971e-05, "loss": 0.0226, "step": 11164 }, { "epoch": 1.210429314830876, "grad_norm": 1.0877554416656494, "learning_rate": 1.1973741476860583e-05, "loss": 0.018, "step": 11165 }, { "epoch": 1.2105377276669558, "grad_norm": 3.0193049907684326, "learning_rate": 1.1973016103293199e-05, "loss": 0.0157, "step": 11166 }, { "epoch": 1.2106461405030355, "grad_norm": 1.0524048805236816, "learning_rate": 1.1972290729725811e-05, "loss": 0.0144, "step": 11167 }, { "epoch": 1.2107545533391153, "grad_norm": 0.2787300944328308, "learning_rate": 1.1971565356158423e-05, "loss": 0.0082, "step": 11168 }, { "epoch": 1.210862966175195, "grad_norm": 23.71962547302246, "learning_rate": 1.1970839982591035e-05, "loss": 0.0381, "step": 11169 }, { "epoch": 1.210971379011275, "grad_norm": 0.4031292200088501, "learning_rate": 1.1970114609023649e-05, "loss": 0.0208, "step": 11170 }, { "epoch": 1.2110797918473548, "grad_norm": 1.0238126516342163, "learning_rate": 1.1969389235456261e-05, "loss": 0.0057, "step": 11171 }, { "epoch": 1.2111882046834346, "grad_norm": 0.23330263793468475, "learning_rate": 1.1968663861888873e-05, "loss": 0.0022, "step": 11172 }, { "epoch": 1.2112966175195143, "grad_norm": 2.341010808944702, "learning_rate": 1.1967938488321487e-05, "loss": 0.0084, "step": 11173 }, { "epoch": 1.211405030355594, "grad_norm": 1.3557229042053223, "learning_rate": 1.19672131147541e-05, "loss": 0.0112, "step": 11174 }, { "epoch": 1.2115134431916739, "grad_norm": 0.2912724018096924, "learning_rate": 1.1966487741186711e-05, "loss": 0.0209, "step": 11175 }, { "epoch": 1.2116218560277536, "grad_norm": 2.5985801219940186, "learning_rate": 1.1965762367619325e-05, "loss": 0.0643, "step": 11176 }, { "epoch": 1.2117302688638334, "grad_norm": 6.127293586730957, "learning_rate": 1.1965036994051937e-05, "loss": 0.0638, "step": 11177 }, { "epoch": 1.2118386816999132, "grad_norm": 0.8777729272842407, "learning_rate": 1.196431162048455e-05, "loss": 0.0105, "step": 11178 }, { "epoch": 1.2119470945359931, "grad_norm": 1.3455837965011597, "learning_rate": 1.1963586246917163e-05, "loss": 0.0298, "step": 11179 }, { "epoch": 1.212055507372073, "grad_norm": 1.4492448568344116, "learning_rate": 1.1962860873349776e-05, "loss": 0.0079, "step": 11180 }, { "epoch": 1.2121639202081527, "grad_norm": 6.140752792358398, "learning_rate": 1.1962135499782388e-05, "loss": 0.0613, "step": 11181 }, { "epoch": 1.2122723330442324, "grad_norm": 0.06803646683692932, "learning_rate": 1.1961410126215e-05, "loss": 0.0009, "step": 11182 }, { "epoch": 1.2123807458803122, "grad_norm": 0.2857019007205963, "learning_rate": 1.1960684752647615e-05, "loss": 0.0022, "step": 11183 }, { "epoch": 1.212489158716392, "grad_norm": 1.2473772764205933, "learning_rate": 1.1959959379080228e-05, "loss": 0.0171, "step": 11184 }, { "epoch": 1.2125975715524717, "grad_norm": 3.696889638900757, "learning_rate": 1.1959234005512841e-05, "loss": 0.0147, "step": 11185 }, { "epoch": 1.2127059843885517, "grad_norm": 1.2526370286941528, "learning_rate": 1.1958508631945454e-05, "loss": 0.0128, "step": 11186 }, { "epoch": 1.2128143972246315, "grad_norm": 0.04921228811144829, "learning_rate": 1.1957783258378066e-05, "loss": 0.0005, "step": 11187 }, { "epoch": 1.2129228100607112, "grad_norm": 1.9999510049819946, "learning_rate": 1.195705788481068e-05, "loss": 0.0121, "step": 11188 }, { "epoch": 1.213031222896791, "grad_norm": 1.6045854091644287, "learning_rate": 1.1956332511243292e-05, "loss": 0.037, "step": 11189 }, { "epoch": 1.2131396357328708, "grad_norm": 3.3285107612609863, "learning_rate": 1.1955607137675904e-05, "loss": 0.0616, "step": 11190 }, { "epoch": 1.2132480485689505, "grad_norm": 1.6048904657363892, "learning_rate": 1.1954881764108518e-05, "loss": 0.0189, "step": 11191 }, { "epoch": 1.2133564614050303, "grad_norm": 1.1249525547027588, "learning_rate": 1.195415639054113e-05, "loss": 0.0283, "step": 11192 }, { "epoch": 1.2134648742411103, "grad_norm": 1.753074049949646, "learning_rate": 1.1953431016973742e-05, "loss": 0.051, "step": 11193 }, { "epoch": 1.21357328707719, "grad_norm": 0.8204514980316162, "learning_rate": 1.1952705643406354e-05, "loss": 0.0218, "step": 11194 }, { "epoch": 1.2136816999132698, "grad_norm": 0.8900035619735718, "learning_rate": 1.1951980269838968e-05, "loss": 0.0026, "step": 11195 }, { "epoch": 1.2137901127493496, "grad_norm": 2.34635329246521, "learning_rate": 1.195125489627158e-05, "loss": 0.0712, "step": 11196 }, { "epoch": 1.2138985255854293, "grad_norm": 7.900373935699463, "learning_rate": 1.1950529522704192e-05, "loss": 0.0535, "step": 11197 }, { "epoch": 1.214006938421509, "grad_norm": 4.890043258666992, "learning_rate": 1.1949804149136806e-05, "loss": 0.0239, "step": 11198 }, { "epoch": 1.2141153512575888, "grad_norm": 1.2411268949508667, "learning_rate": 1.1949078775569418e-05, "loss": 0.0158, "step": 11199 }, { "epoch": 1.2142237640936686, "grad_norm": 0.19343821704387665, "learning_rate": 1.1948353402002034e-05, "loss": 0.0025, "step": 11200 }, { "epoch": 1.2143321769297484, "grad_norm": 1.970428228378296, "learning_rate": 1.1947628028434646e-05, "loss": 0.0484, "step": 11201 }, { "epoch": 1.2144405897658284, "grad_norm": 0.3716067373752594, "learning_rate": 1.1946902654867258e-05, "loss": 0.0107, "step": 11202 }, { "epoch": 1.2145490026019081, "grad_norm": 3.642321825027466, "learning_rate": 1.1946177281299872e-05, "loss": 0.0396, "step": 11203 }, { "epoch": 1.2146574154379879, "grad_norm": 5.236907005310059, "learning_rate": 1.1945451907732484e-05, "loss": 0.083, "step": 11204 }, { "epoch": 1.2147658282740676, "grad_norm": 15.237751960754395, "learning_rate": 1.1944726534165096e-05, "loss": 0.1227, "step": 11205 }, { "epoch": 1.2148742411101474, "grad_norm": 0.1861383616924286, "learning_rate": 1.1944001160597708e-05, "loss": 0.0016, "step": 11206 }, { "epoch": 1.2149826539462272, "grad_norm": 1.3972667455673218, "learning_rate": 1.1943275787030322e-05, "loss": 0.0226, "step": 11207 }, { "epoch": 1.215091066782307, "grad_norm": 1.646689772605896, "learning_rate": 1.1942550413462934e-05, "loss": 0.0493, "step": 11208 }, { "epoch": 1.215199479618387, "grad_norm": 0.4350988268852234, "learning_rate": 1.1941825039895546e-05, "loss": 0.0085, "step": 11209 }, { "epoch": 1.2153078924544667, "grad_norm": 0.47953730821609497, "learning_rate": 1.194109966632816e-05, "loss": 0.0034, "step": 11210 }, { "epoch": 1.2154163052905465, "grad_norm": 3.404912233352661, "learning_rate": 1.1940374292760772e-05, "loss": 0.0612, "step": 11211 }, { "epoch": 1.2155247181266262, "grad_norm": 2.834697723388672, "learning_rate": 1.1939648919193385e-05, "loss": 0.0415, "step": 11212 }, { "epoch": 1.215633130962706, "grad_norm": 0.4053260385990143, "learning_rate": 1.1938923545625998e-05, "loss": 0.0182, "step": 11213 }, { "epoch": 1.2157415437987857, "grad_norm": 4.357780456542969, "learning_rate": 1.193819817205861e-05, "loss": 0.0333, "step": 11214 }, { "epoch": 1.2158499566348655, "grad_norm": 0.485012024641037, "learning_rate": 1.1937472798491223e-05, "loss": 0.0163, "step": 11215 }, { "epoch": 1.2159583694709453, "grad_norm": 3.2342326641082764, "learning_rate": 1.1936747424923836e-05, "loss": 0.0363, "step": 11216 }, { "epoch": 1.216066782307025, "grad_norm": 2.2092809677124023, "learning_rate": 1.193602205135645e-05, "loss": 0.0353, "step": 11217 }, { "epoch": 1.216175195143105, "grad_norm": 0.44532880187034607, "learning_rate": 1.1935296677789064e-05, "loss": 0.0142, "step": 11218 }, { "epoch": 1.2162836079791848, "grad_norm": 1.653409719467163, "learning_rate": 1.1934571304221676e-05, "loss": 0.0141, "step": 11219 }, { "epoch": 1.2163920208152645, "grad_norm": 4.555911064147949, "learning_rate": 1.1933845930654288e-05, "loss": 0.0674, "step": 11220 }, { "epoch": 1.2165004336513443, "grad_norm": 4.196483612060547, "learning_rate": 1.19331205570869e-05, "loss": 0.0252, "step": 11221 }, { "epoch": 1.216608846487424, "grad_norm": 1.4975906610488892, "learning_rate": 1.1932395183519514e-05, "loss": 0.0097, "step": 11222 }, { "epoch": 1.2167172593235038, "grad_norm": 1.066821575164795, "learning_rate": 1.1931669809952127e-05, "loss": 0.0258, "step": 11223 }, { "epoch": 1.2168256721595836, "grad_norm": 0.8152394890785217, "learning_rate": 1.1930944436384739e-05, "loss": 0.0045, "step": 11224 }, { "epoch": 1.2169340849956636, "grad_norm": 2.0703487396240234, "learning_rate": 1.1930219062817353e-05, "loss": 0.0282, "step": 11225 }, { "epoch": 1.2170424978317433, "grad_norm": 1.0586308240890503, "learning_rate": 1.1929493689249965e-05, "loss": 0.0127, "step": 11226 }, { "epoch": 1.217150910667823, "grad_norm": 1.4588708877563477, "learning_rate": 1.1928768315682577e-05, "loss": 0.0173, "step": 11227 }, { "epoch": 1.2172593235039029, "grad_norm": 1.8180655241012573, "learning_rate": 1.192804294211519e-05, "loss": 0.0266, "step": 11228 }, { "epoch": 1.2173677363399826, "grad_norm": 2.0849649906158447, "learning_rate": 1.1927317568547803e-05, "loss": 0.0229, "step": 11229 }, { "epoch": 1.2174761491760624, "grad_norm": 0.911196768283844, "learning_rate": 1.1926592194980415e-05, "loss": 0.0186, "step": 11230 }, { "epoch": 1.2175845620121422, "grad_norm": 3.380221366882324, "learning_rate": 1.1925866821413029e-05, "loss": 0.0151, "step": 11231 }, { "epoch": 1.2176929748482221, "grad_norm": 1.5912365913391113, "learning_rate": 1.1925141447845641e-05, "loss": 0.0456, "step": 11232 }, { "epoch": 1.217801387684302, "grad_norm": 2.7283027172088623, "learning_rate": 1.1924416074278253e-05, "loss": 0.0108, "step": 11233 }, { "epoch": 1.2179098005203817, "grad_norm": 0.12448639422655106, "learning_rate": 1.1923690700710865e-05, "loss": 0.0016, "step": 11234 }, { "epoch": 1.2180182133564614, "grad_norm": 0.4735221862792969, "learning_rate": 1.192296532714348e-05, "loss": 0.0254, "step": 11235 }, { "epoch": 1.2181266261925412, "grad_norm": 1.6708617210388184, "learning_rate": 1.1922239953576093e-05, "loss": 0.0404, "step": 11236 }, { "epoch": 1.218235039028621, "grad_norm": 0.5332473516464233, "learning_rate": 1.1921514580008707e-05, "loss": 0.0132, "step": 11237 }, { "epoch": 1.2183434518647007, "grad_norm": 4.328149795532227, "learning_rate": 1.1920789206441319e-05, "loss": 0.0524, "step": 11238 }, { "epoch": 1.2184518647007805, "grad_norm": 3.329049825668335, "learning_rate": 1.1920063832873931e-05, "loss": 0.0146, "step": 11239 }, { "epoch": 1.2185602775368602, "grad_norm": 0.09414795786142349, "learning_rate": 1.1919338459306545e-05, "loss": 0.001, "step": 11240 }, { "epoch": 1.2186686903729402, "grad_norm": 0.5300240516662598, "learning_rate": 1.1918613085739157e-05, "loss": 0.0156, "step": 11241 }, { "epoch": 1.21877710320902, "grad_norm": 2.3989768028259277, "learning_rate": 1.1917887712171769e-05, "loss": 0.0284, "step": 11242 }, { "epoch": 1.2188855160450998, "grad_norm": 0.3540501296520233, "learning_rate": 1.1917162338604383e-05, "loss": 0.0213, "step": 11243 }, { "epoch": 1.2189939288811795, "grad_norm": 0.6270464658737183, "learning_rate": 1.1916436965036995e-05, "loss": 0.0103, "step": 11244 }, { "epoch": 1.2191023417172593, "grad_norm": 5.067623138427734, "learning_rate": 1.1915711591469607e-05, "loss": 0.0552, "step": 11245 }, { "epoch": 1.219210754553339, "grad_norm": 0.7582765817642212, "learning_rate": 1.191498621790222e-05, "loss": 0.0134, "step": 11246 }, { "epoch": 1.2193191673894188, "grad_norm": 1.653190016746521, "learning_rate": 1.1914260844334833e-05, "loss": 0.0335, "step": 11247 }, { "epoch": 1.2194275802254988, "grad_norm": 0.5208827257156372, "learning_rate": 1.1913535470767445e-05, "loss": 0.0187, "step": 11248 }, { "epoch": 1.2195359930615786, "grad_norm": 1.0381388664245605, "learning_rate": 1.1912810097200058e-05, "loss": 0.0131, "step": 11249 }, { "epoch": 1.2196444058976583, "grad_norm": 0.21705326437950134, "learning_rate": 1.1912084723632671e-05, "loss": 0.0016, "step": 11250 }, { "epoch": 1.219752818733738, "grad_norm": 1.8144021034240723, "learning_rate": 1.1911359350065284e-05, "loss": 0.0303, "step": 11251 }, { "epoch": 1.2198612315698178, "grad_norm": 8.040262222290039, "learning_rate": 1.1910633976497899e-05, "loss": 0.0629, "step": 11252 }, { "epoch": 1.2199696444058976, "grad_norm": 1.3389482498168945, "learning_rate": 1.1909908602930511e-05, "loss": 0.0162, "step": 11253 }, { "epoch": 1.2200780572419774, "grad_norm": 2.3020825386047363, "learning_rate": 1.1909183229363123e-05, "loss": 0.0404, "step": 11254 }, { "epoch": 1.2201864700780574, "grad_norm": 5.733405113220215, "learning_rate": 1.1908457855795737e-05, "loss": 0.0488, "step": 11255 }, { "epoch": 1.2202948829141371, "grad_norm": 0.35222044587135315, "learning_rate": 1.190773248222835e-05, "loss": 0.0117, "step": 11256 }, { "epoch": 1.2204032957502169, "grad_norm": 0.6301643252372742, "learning_rate": 1.1907007108660961e-05, "loss": 0.0107, "step": 11257 }, { "epoch": 1.2205117085862967, "grad_norm": 1.6576387882232666, "learning_rate": 1.1906281735093574e-05, "loss": 0.0237, "step": 11258 }, { "epoch": 1.2206201214223764, "grad_norm": 1.4039156436920166, "learning_rate": 1.1905556361526187e-05, "loss": 0.0288, "step": 11259 }, { "epoch": 1.2207285342584562, "grad_norm": 1.0775914192199707, "learning_rate": 1.19048309879588e-05, "loss": 0.0198, "step": 11260 }, { "epoch": 1.220836947094536, "grad_norm": 1.8618450164794922, "learning_rate": 1.1904105614391412e-05, "loss": 0.0449, "step": 11261 }, { "epoch": 1.2209453599306157, "grad_norm": 6.926957130432129, "learning_rate": 1.1903380240824026e-05, "loss": 0.0441, "step": 11262 }, { "epoch": 1.2210537727666955, "grad_norm": 4.823750972747803, "learning_rate": 1.1902654867256638e-05, "loss": 0.0575, "step": 11263 }, { "epoch": 1.2211621856027755, "grad_norm": 1.067522644996643, "learning_rate": 1.190192949368925e-05, "loss": 0.0217, "step": 11264 }, { "epoch": 1.2212705984388552, "grad_norm": 1.242092490196228, "learning_rate": 1.1901204120121864e-05, "loss": 0.0407, "step": 11265 }, { "epoch": 1.221379011274935, "grad_norm": 2.3829357624053955, "learning_rate": 1.1900478746554476e-05, "loss": 0.0182, "step": 11266 }, { "epoch": 1.2214874241110147, "grad_norm": 1.7339857816696167, "learning_rate": 1.1899753372987088e-05, "loss": 0.0546, "step": 11267 }, { "epoch": 1.2215958369470945, "grad_norm": 1.0382106304168701, "learning_rate": 1.1899027999419702e-05, "loss": 0.024, "step": 11268 }, { "epoch": 1.2217042497831743, "grad_norm": 0.15367057919502258, "learning_rate": 1.1898302625852316e-05, "loss": 0.0008, "step": 11269 }, { "epoch": 1.221812662619254, "grad_norm": 2.6632492542266846, "learning_rate": 1.189757725228493e-05, "loss": 0.017, "step": 11270 }, { "epoch": 1.221921075455334, "grad_norm": 0.44683194160461426, "learning_rate": 1.1896851878717542e-05, "loss": 0.0126, "step": 11271 }, { "epoch": 1.2220294882914138, "grad_norm": 6.235104560852051, "learning_rate": 1.1896126505150154e-05, "loss": 0.0504, "step": 11272 }, { "epoch": 1.2221379011274935, "grad_norm": 0.6292608380317688, "learning_rate": 1.1895401131582766e-05, "loss": 0.0197, "step": 11273 }, { "epoch": 1.2222463139635733, "grad_norm": 1.5728025436401367, "learning_rate": 1.189467575801538e-05, "loss": 0.0363, "step": 11274 }, { "epoch": 1.222354726799653, "grad_norm": 0.689207136631012, "learning_rate": 1.1893950384447992e-05, "loss": 0.0088, "step": 11275 }, { "epoch": 1.2224631396357328, "grad_norm": 1.5859953165054321, "learning_rate": 1.1893225010880604e-05, "loss": 0.0145, "step": 11276 }, { "epoch": 1.2225715524718126, "grad_norm": 1.00786292552948, "learning_rate": 1.1892499637313218e-05, "loss": 0.0109, "step": 11277 }, { "epoch": 1.2226799653078924, "grad_norm": 12.415237426757812, "learning_rate": 1.189177426374583e-05, "loss": 0.0795, "step": 11278 }, { "epoch": 1.2227883781439723, "grad_norm": 0.19605514407157898, "learning_rate": 1.1891048890178442e-05, "loss": 0.0019, "step": 11279 }, { "epoch": 1.222896790980052, "grad_norm": 2.244802236557007, "learning_rate": 1.1890323516611056e-05, "loss": 0.0381, "step": 11280 }, { "epoch": 1.2230052038161319, "grad_norm": 1.8871040344238281, "learning_rate": 1.1889598143043668e-05, "loss": 0.0373, "step": 11281 }, { "epoch": 1.2231136166522116, "grad_norm": 2.5810413360595703, "learning_rate": 1.188887276947628e-05, "loss": 0.0218, "step": 11282 }, { "epoch": 1.2232220294882914, "grad_norm": 0.4815288186073303, "learning_rate": 1.1888147395908894e-05, "loss": 0.0082, "step": 11283 }, { "epoch": 1.2233304423243712, "grad_norm": 1.090466022491455, "learning_rate": 1.1887422022341506e-05, "loss": 0.0321, "step": 11284 }, { "epoch": 1.223438855160451, "grad_norm": 3.5528171062469482, "learning_rate": 1.1886696648774118e-05, "loss": 0.0544, "step": 11285 }, { "epoch": 1.2235472679965307, "grad_norm": 0.6166350841522217, "learning_rate": 1.188597127520673e-05, "loss": 0.0126, "step": 11286 }, { "epoch": 1.2236556808326107, "grad_norm": 3.8198373317718506, "learning_rate": 1.1885245901639346e-05, "loss": 0.0103, "step": 11287 }, { "epoch": 1.2237640936686904, "grad_norm": 4.457541465759277, "learning_rate": 1.1884520528071958e-05, "loss": 0.029, "step": 11288 }, { "epoch": 1.2238725065047702, "grad_norm": 3.569916009902954, "learning_rate": 1.1883795154504572e-05, "loss": 0.0218, "step": 11289 }, { "epoch": 1.22398091934085, "grad_norm": 0.40314382314682007, "learning_rate": 1.1883069780937184e-05, "loss": 0.0018, "step": 11290 }, { "epoch": 1.2240893321769297, "grad_norm": 2.32440447807312, "learning_rate": 1.1882344407369796e-05, "loss": 0.0386, "step": 11291 }, { "epoch": 1.2241977450130095, "grad_norm": 8.636808395385742, "learning_rate": 1.188161903380241e-05, "loss": 0.0548, "step": 11292 }, { "epoch": 1.2243061578490892, "grad_norm": 3.466951847076416, "learning_rate": 1.1880893660235022e-05, "loss": 0.0287, "step": 11293 }, { "epoch": 1.2244145706851692, "grad_norm": 2.401601552963257, "learning_rate": 1.1880168286667634e-05, "loss": 0.0207, "step": 11294 }, { "epoch": 1.224522983521249, "grad_norm": 2.3556880950927734, "learning_rate": 1.1879442913100248e-05, "loss": 0.0193, "step": 11295 }, { "epoch": 1.2246313963573288, "grad_norm": 1.3567092418670654, "learning_rate": 1.187871753953286e-05, "loss": 0.0281, "step": 11296 }, { "epoch": 1.2247398091934085, "grad_norm": 1.8871372938156128, "learning_rate": 1.1877992165965473e-05, "loss": 0.0664, "step": 11297 }, { "epoch": 1.2248482220294883, "grad_norm": 2.2051899433135986, "learning_rate": 1.1877266792398085e-05, "loss": 0.0285, "step": 11298 }, { "epoch": 1.224956634865568, "grad_norm": 2.394637107849121, "learning_rate": 1.1876541418830699e-05, "loss": 0.003, "step": 11299 }, { "epoch": 1.2250650477016478, "grad_norm": 0.9998143315315247, "learning_rate": 1.187581604526331e-05, "loss": 0.0207, "step": 11300 }, { "epoch": 1.2251734605377276, "grad_norm": 2.0751771926879883, "learning_rate": 1.1875090671695923e-05, "loss": 0.0287, "step": 11301 }, { "epoch": 1.2252818733738073, "grad_norm": 0.3742099106311798, "learning_rate": 1.1874365298128537e-05, "loss": 0.0021, "step": 11302 }, { "epoch": 1.2253902862098873, "grad_norm": 5.322146415710449, "learning_rate": 1.1873639924561149e-05, "loss": 0.035, "step": 11303 }, { "epoch": 1.225498699045967, "grad_norm": 1.7625257968902588, "learning_rate": 1.1872914550993764e-05, "loss": 0.0191, "step": 11304 }, { "epoch": 1.2256071118820469, "grad_norm": 0.8345404863357544, "learning_rate": 1.1872189177426377e-05, "loss": 0.0232, "step": 11305 }, { "epoch": 1.2257155247181266, "grad_norm": 0.8075301647186279, "learning_rate": 1.1871463803858989e-05, "loss": 0.0146, "step": 11306 }, { "epoch": 1.2258239375542064, "grad_norm": 0.6989781260490417, "learning_rate": 1.1870738430291603e-05, "loss": 0.0188, "step": 11307 }, { "epoch": 1.2259323503902861, "grad_norm": 1.3083453178405762, "learning_rate": 1.1870013056724215e-05, "loss": 0.0036, "step": 11308 }, { "epoch": 1.226040763226366, "grad_norm": 1.1059393882751465, "learning_rate": 1.1869287683156827e-05, "loss": 0.0199, "step": 11309 }, { "epoch": 1.226149176062446, "grad_norm": 1.2647268772125244, "learning_rate": 1.1868562309589439e-05, "loss": 0.0181, "step": 11310 }, { "epoch": 1.2262575888985257, "grad_norm": 1.9484864473342896, "learning_rate": 1.1867836936022053e-05, "loss": 0.0156, "step": 11311 }, { "epoch": 1.2263660017346054, "grad_norm": 1.671934723854065, "learning_rate": 1.1867111562454665e-05, "loss": 0.0063, "step": 11312 }, { "epoch": 1.2264744145706852, "grad_norm": 0.4894309639930725, "learning_rate": 1.1866386188887277e-05, "loss": 0.0269, "step": 11313 }, { "epoch": 1.226582827406765, "grad_norm": 6.593571662902832, "learning_rate": 1.1865660815319891e-05, "loss": 0.0329, "step": 11314 }, { "epoch": 1.2266912402428447, "grad_norm": 0.3679896295070648, "learning_rate": 1.1864935441752503e-05, "loss": 0.0046, "step": 11315 }, { "epoch": 1.2267996530789245, "grad_norm": 1.506295919418335, "learning_rate": 1.1864210068185115e-05, "loss": 0.0259, "step": 11316 }, { "epoch": 1.2269080659150045, "grad_norm": 0.17271815240383148, "learning_rate": 1.1863484694617729e-05, "loss": 0.0019, "step": 11317 }, { "epoch": 1.2270164787510842, "grad_norm": 1.7829029560089111, "learning_rate": 1.1862759321050341e-05, "loss": 0.0265, "step": 11318 }, { "epoch": 1.227124891587164, "grad_norm": 6.470808506011963, "learning_rate": 1.1862033947482953e-05, "loss": 0.033, "step": 11319 }, { "epoch": 1.2272333044232437, "grad_norm": 2.7702298164367676, "learning_rate": 1.1861308573915567e-05, "loss": 0.0245, "step": 11320 }, { "epoch": 1.2273417172593235, "grad_norm": 0.6708892583847046, "learning_rate": 1.1860583200348181e-05, "loss": 0.0079, "step": 11321 }, { "epoch": 1.2274501300954033, "grad_norm": 6.724693775177002, "learning_rate": 1.1859857826780795e-05, "loss": 0.0627, "step": 11322 }, { "epoch": 1.227558542931483, "grad_norm": 2.5210988521575928, "learning_rate": 1.1859132453213407e-05, "loss": 0.0095, "step": 11323 }, { "epoch": 1.2276669557675628, "grad_norm": 1.2408114671707153, "learning_rate": 1.1858407079646019e-05, "loss": 0.0199, "step": 11324 }, { "epoch": 1.2277753686036426, "grad_norm": 3.4725725650787354, "learning_rate": 1.1857681706078631e-05, "loss": 0.0429, "step": 11325 }, { "epoch": 1.2278837814397225, "grad_norm": 0.24672992527484894, "learning_rate": 1.1856956332511245e-05, "loss": 0.005, "step": 11326 }, { "epoch": 1.2279921942758023, "grad_norm": 3.9267256259918213, "learning_rate": 1.1856230958943857e-05, "loss": 0.0569, "step": 11327 }, { "epoch": 1.228100607111882, "grad_norm": 4.25592041015625, "learning_rate": 1.185550558537647e-05, "loss": 0.0108, "step": 11328 }, { "epoch": 1.2282090199479618, "grad_norm": 0.3864511549472809, "learning_rate": 1.1854780211809083e-05, "loss": 0.0105, "step": 11329 }, { "epoch": 1.2283174327840416, "grad_norm": 3.4238150119781494, "learning_rate": 1.1854054838241695e-05, "loss": 0.0293, "step": 11330 }, { "epoch": 1.2284258456201214, "grad_norm": 4.742404937744141, "learning_rate": 1.1853329464674307e-05, "loss": 0.0241, "step": 11331 }, { "epoch": 1.2285342584562011, "grad_norm": 9.479101181030273, "learning_rate": 1.1852604091106921e-05, "loss": 0.0208, "step": 11332 }, { "epoch": 1.228642671292281, "grad_norm": 11.938019752502441, "learning_rate": 1.1851878717539533e-05, "loss": 0.023, "step": 11333 }, { "epoch": 1.2287510841283609, "grad_norm": 1.139640212059021, "learning_rate": 1.1851153343972146e-05, "loss": 0.0321, "step": 11334 }, { "epoch": 1.2288594969644406, "grad_norm": 3.6642680168151855, "learning_rate": 1.185042797040476e-05, "loss": 0.1254, "step": 11335 }, { "epoch": 1.2289679098005204, "grad_norm": 0.32945942878723145, "learning_rate": 1.1849702596837372e-05, "loss": 0.0039, "step": 11336 }, { "epoch": 1.2290763226366002, "grad_norm": 1.9385607242584229, "learning_rate": 1.1848977223269984e-05, "loss": 0.0163, "step": 11337 }, { "epoch": 1.22918473547268, "grad_norm": 1.5916694402694702, "learning_rate": 1.18482518497026e-05, "loss": 0.0034, "step": 11338 }, { "epoch": 1.2292931483087597, "grad_norm": 3.1025335788726807, "learning_rate": 1.1847526476135211e-05, "loss": 0.0261, "step": 11339 }, { "epoch": 1.2294015611448397, "grad_norm": 1.170405387878418, "learning_rate": 1.1846801102567824e-05, "loss": 0.0097, "step": 11340 }, { "epoch": 1.2295099739809194, "grad_norm": 1.4926027059555054, "learning_rate": 1.1846075729000437e-05, "loss": 0.0398, "step": 11341 }, { "epoch": 1.2296183868169992, "grad_norm": 2.9352831840515137, "learning_rate": 1.184535035543305e-05, "loss": 0.0846, "step": 11342 }, { "epoch": 1.229726799653079, "grad_norm": 1.4817471504211426, "learning_rate": 1.1844624981865662e-05, "loss": 0.0192, "step": 11343 }, { "epoch": 1.2298352124891587, "grad_norm": 0.5399120450019836, "learning_rate": 1.1843899608298276e-05, "loss": 0.011, "step": 11344 }, { "epoch": 1.2299436253252385, "grad_norm": 7.821362495422363, "learning_rate": 1.1843174234730888e-05, "loss": 0.0821, "step": 11345 }, { "epoch": 1.2300520381613183, "grad_norm": 5.943788528442383, "learning_rate": 1.18424488611635e-05, "loss": 0.0369, "step": 11346 }, { "epoch": 1.230160450997398, "grad_norm": 0.5461013317108154, "learning_rate": 1.1841723487596114e-05, "loss": 0.0286, "step": 11347 }, { "epoch": 1.2302688638334778, "grad_norm": 3.4253346920013428, "learning_rate": 1.1840998114028726e-05, "loss": 0.0127, "step": 11348 }, { "epoch": 1.2303772766695578, "grad_norm": 1.7588986158370972, "learning_rate": 1.1840272740461338e-05, "loss": 0.0043, "step": 11349 }, { "epoch": 1.2304856895056375, "grad_norm": 0.8769294023513794, "learning_rate": 1.183954736689395e-05, "loss": 0.0149, "step": 11350 }, { "epoch": 1.2305941023417173, "grad_norm": 1.944635033607483, "learning_rate": 1.1838821993326564e-05, "loss": 0.018, "step": 11351 }, { "epoch": 1.230702515177797, "grad_norm": 6.351424217224121, "learning_rate": 1.1838096619759176e-05, "loss": 0.0337, "step": 11352 }, { "epoch": 1.2308109280138768, "grad_norm": 1.2144604921340942, "learning_rate": 1.1837371246191788e-05, "loss": 0.0133, "step": 11353 }, { "epoch": 1.2309193408499566, "grad_norm": 1.3683472871780396, "learning_rate": 1.1836645872624402e-05, "loss": 0.0101, "step": 11354 }, { "epoch": 1.2310277536860363, "grad_norm": 2.803273916244507, "learning_rate": 1.1835920499057014e-05, "loss": 0.0335, "step": 11355 }, { "epoch": 1.2311361665221163, "grad_norm": 0.3845394551753998, "learning_rate": 1.183519512548963e-05, "loss": 0.0104, "step": 11356 }, { "epoch": 1.231244579358196, "grad_norm": 0.4333466589450836, "learning_rate": 1.1834469751922242e-05, "loss": 0.0148, "step": 11357 }, { "epoch": 1.2313529921942759, "grad_norm": 1.616421103477478, "learning_rate": 1.1833744378354854e-05, "loss": 0.0246, "step": 11358 }, { "epoch": 1.2314614050303556, "grad_norm": 3.5077359676361084, "learning_rate": 1.1833019004787468e-05, "loss": 0.021, "step": 11359 }, { "epoch": 1.2315698178664354, "grad_norm": 5.7360992431640625, "learning_rate": 1.183229363122008e-05, "loss": 0.0359, "step": 11360 }, { "epoch": 1.2316782307025151, "grad_norm": 0.4038110375404358, "learning_rate": 1.1831568257652692e-05, "loss": 0.0092, "step": 11361 }, { "epoch": 1.231786643538595, "grad_norm": 1.53436279296875, "learning_rate": 1.1830842884085306e-05, "loss": 0.0062, "step": 11362 }, { "epoch": 1.2318950563746747, "grad_norm": 4.332874298095703, "learning_rate": 1.1830117510517918e-05, "loss": 0.0156, "step": 11363 }, { "epoch": 1.2320034692107544, "grad_norm": 2.965618133544922, "learning_rate": 1.182939213695053e-05, "loss": 0.0315, "step": 11364 }, { "epoch": 1.2321118820468344, "grad_norm": 5.991425037384033, "learning_rate": 1.1828666763383142e-05, "loss": 0.0168, "step": 11365 }, { "epoch": 1.2322202948829142, "grad_norm": 0.4391186535358429, "learning_rate": 1.1827941389815756e-05, "loss": 0.0069, "step": 11366 }, { "epoch": 1.232328707718994, "grad_norm": 0.8649458885192871, "learning_rate": 1.1827216016248368e-05, "loss": 0.0132, "step": 11367 }, { "epoch": 1.2324371205550737, "grad_norm": 2.808849334716797, "learning_rate": 1.182649064268098e-05, "loss": 0.0267, "step": 11368 }, { "epoch": 1.2325455333911535, "grad_norm": 16.84714126586914, "learning_rate": 1.1825765269113594e-05, "loss": 0.0125, "step": 11369 }, { "epoch": 1.2326539462272332, "grad_norm": 3.988628387451172, "learning_rate": 1.1825039895546206e-05, "loss": 0.0508, "step": 11370 }, { "epoch": 1.232762359063313, "grad_norm": 8.205649375915527, "learning_rate": 1.1824314521978819e-05, "loss": 0.0376, "step": 11371 }, { "epoch": 1.232870771899393, "grad_norm": 0.3735044300556183, "learning_rate": 1.1823589148411432e-05, "loss": 0.0131, "step": 11372 }, { "epoch": 1.2329791847354727, "grad_norm": 0.023564567789435387, "learning_rate": 1.1822863774844046e-05, "loss": 0.0002, "step": 11373 }, { "epoch": 1.2330875975715525, "grad_norm": 2.3855791091918945, "learning_rate": 1.182213840127666e-05, "loss": 0.0684, "step": 11374 }, { "epoch": 1.2331960104076323, "grad_norm": 0.17237915098667145, "learning_rate": 1.1821413027709272e-05, "loss": 0.0023, "step": 11375 }, { "epoch": 1.233304423243712, "grad_norm": 1.8112049102783203, "learning_rate": 1.1820687654141884e-05, "loss": 0.0399, "step": 11376 }, { "epoch": 1.2334128360797918, "grad_norm": 3.194561719894409, "learning_rate": 1.1819962280574497e-05, "loss": 0.0144, "step": 11377 }, { "epoch": 1.2335212489158716, "grad_norm": 0.9171454310417175, "learning_rate": 1.181923690700711e-05, "loss": 0.0336, "step": 11378 }, { "epoch": 1.2336296617519515, "grad_norm": 0.8646206259727478, "learning_rate": 1.1818511533439723e-05, "loss": 0.0145, "step": 11379 }, { "epoch": 1.2337380745880313, "grad_norm": 3.3305609226226807, "learning_rate": 1.1817786159872335e-05, "loss": 0.0052, "step": 11380 }, { "epoch": 1.233846487424111, "grad_norm": 2.497968912124634, "learning_rate": 1.1817060786304949e-05, "loss": 0.0403, "step": 11381 }, { "epoch": 1.2339549002601908, "grad_norm": 1.0573699474334717, "learning_rate": 1.181633541273756e-05, "loss": 0.0227, "step": 11382 }, { "epoch": 1.2340633130962706, "grad_norm": 0.8838213086128235, "learning_rate": 1.1815610039170173e-05, "loss": 0.02, "step": 11383 }, { "epoch": 1.2341717259323504, "grad_norm": 7.592560291290283, "learning_rate": 1.1814884665602787e-05, "loss": 0.0237, "step": 11384 }, { "epoch": 1.2342801387684301, "grad_norm": 9.685275077819824, "learning_rate": 1.1814159292035399e-05, "loss": 0.0186, "step": 11385 }, { "epoch": 1.2343885516045099, "grad_norm": 1.3430458307266235, "learning_rate": 1.1813433918468011e-05, "loss": 0.0098, "step": 11386 }, { "epoch": 1.2344969644405897, "grad_norm": 3.050231695175171, "learning_rate": 1.1812708544900625e-05, "loss": 0.0736, "step": 11387 }, { "epoch": 1.2346053772766696, "grad_norm": 1.1230393648147583, "learning_rate": 1.1811983171333237e-05, "loss": 0.0261, "step": 11388 }, { "epoch": 1.2347137901127494, "grad_norm": 0.33614128828048706, "learning_rate": 1.1811257797765849e-05, "loss": 0.004, "step": 11389 }, { "epoch": 1.2348222029488292, "grad_norm": 4.920724391937256, "learning_rate": 1.1810532424198465e-05, "loss": 0.0258, "step": 11390 }, { "epoch": 1.234930615784909, "grad_norm": 0.9063054323196411, "learning_rate": 1.1809807050631077e-05, "loss": 0.0294, "step": 11391 }, { "epoch": 1.2350390286209887, "grad_norm": 0.9878413677215576, "learning_rate": 1.1809081677063689e-05, "loss": 0.0058, "step": 11392 }, { "epoch": 1.2351474414570685, "grad_norm": 4.258705139160156, "learning_rate": 1.1808356303496303e-05, "loss": 0.0189, "step": 11393 }, { "epoch": 1.2352558542931482, "grad_norm": 1.4232804775238037, "learning_rate": 1.1807630929928915e-05, "loss": 0.0275, "step": 11394 }, { "epoch": 1.2353642671292282, "grad_norm": 0.7106199264526367, "learning_rate": 1.1806905556361527e-05, "loss": 0.0216, "step": 11395 }, { "epoch": 1.235472679965308, "grad_norm": 1.4620558023452759, "learning_rate": 1.1806180182794141e-05, "loss": 0.035, "step": 11396 }, { "epoch": 1.2355810928013877, "grad_norm": 1.0789471864700317, "learning_rate": 1.1805454809226753e-05, "loss": 0.0084, "step": 11397 }, { "epoch": 1.2356895056374675, "grad_norm": 8.856524467468262, "learning_rate": 1.1804729435659365e-05, "loss": 0.0201, "step": 11398 }, { "epoch": 1.2357979184735473, "grad_norm": 1.3766748905181885, "learning_rate": 1.1804004062091979e-05, "loss": 0.0347, "step": 11399 }, { "epoch": 1.235906331309627, "grad_norm": 5.8675456047058105, "learning_rate": 1.1803278688524591e-05, "loss": 0.0365, "step": 11400 }, { "epoch": 1.2360147441457068, "grad_norm": 7.350146770477295, "learning_rate": 1.1802553314957203e-05, "loss": 0.0463, "step": 11401 }, { "epoch": 1.2361231569817868, "grad_norm": 5.323853492736816, "learning_rate": 1.1801827941389815e-05, "loss": 0.0351, "step": 11402 }, { "epoch": 1.2362315698178665, "grad_norm": 1.9761384725570679, "learning_rate": 1.180110256782243e-05, "loss": 0.0515, "step": 11403 }, { "epoch": 1.2363399826539463, "grad_norm": 0.6863661408424377, "learning_rate": 1.1800377194255041e-05, "loss": 0.0169, "step": 11404 }, { "epoch": 1.236448395490026, "grad_norm": 3.3857202529907227, "learning_rate": 1.1799651820687654e-05, "loss": 0.026, "step": 11405 }, { "epoch": 1.2365568083261058, "grad_norm": 0.45939403772354126, "learning_rate": 1.1798926447120267e-05, "loss": 0.0093, "step": 11406 }, { "epoch": 1.2366652211621856, "grad_norm": 0.3264886438846588, "learning_rate": 1.1798201073552881e-05, "loss": 0.0164, "step": 11407 }, { "epoch": 1.2367736339982653, "grad_norm": 10.246191024780273, "learning_rate": 1.1797475699985495e-05, "loss": 0.049, "step": 11408 }, { "epoch": 1.236882046834345, "grad_norm": 5.762904644012451, "learning_rate": 1.1796750326418107e-05, "loss": 0.0176, "step": 11409 }, { "epoch": 1.2369904596704249, "grad_norm": 0.5164905190467834, "learning_rate": 1.179602495285072e-05, "loss": 0.0148, "step": 11410 }, { "epoch": 1.2370988725065049, "grad_norm": 0.18547172844409943, "learning_rate": 1.1795299579283333e-05, "loss": 0.0011, "step": 11411 }, { "epoch": 1.2372072853425846, "grad_norm": 0.36258983612060547, "learning_rate": 1.1794574205715945e-05, "loss": 0.0045, "step": 11412 }, { "epoch": 1.2373156981786644, "grad_norm": 0.5007626414299011, "learning_rate": 1.1793848832148557e-05, "loss": 0.0155, "step": 11413 }, { "epoch": 1.2374241110147441, "grad_norm": 1.3058085441589355, "learning_rate": 1.1793123458581171e-05, "loss": 0.0047, "step": 11414 }, { "epoch": 1.237532523850824, "grad_norm": 3.216865301132202, "learning_rate": 1.1792398085013783e-05, "loss": 0.0524, "step": 11415 }, { "epoch": 1.2376409366869037, "grad_norm": 2.2311692237854004, "learning_rate": 1.1791672711446396e-05, "loss": 0.0248, "step": 11416 }, { "epoch": 1.2377493495229834, "grad_norm": 0.7247829437255859, "learning_rate": 1.1790947337879008e-05, "loss": 0.0188, "step": 11417 }, { "epoch": 1.2378577623590634, "grad_norm": 0.4457912743091583, "learning_rate": 1.1790221964311622e-05, "loss": 0.0167, "step": 11418 }, { "epoch": 1.2379661751951432, "grad_norm": 0.30380570888519287, "learning_rate": 1.1789496590744234e-05, "loss": 0.001, "step": 11419 }, { "epoch": 1.238074588031223, "grad_norm": 1.980797290802002, "learning_rate": 1.1788771217176846e-05, "loss": 0.0279, "step": 11420 }, { "epoch": 1.2381830008673027, "grad_norm": 14.59085464477539, "learning_rate": 1.178804584360946e-05, "loss": 0.0434, "step": 11421 }, { "epoch": 1.2382914137033825, "grad_norm": 10.606270790100098, "learning_rate": 1.1787320470042072e-05, "loss": 0.0672, "step": 11422 }, { "epoch": 1.2383998265394622, "grad_norm": 0.6840663552284241, "learning_rate": 1.1786595096474684e-05, "loss": 0.0206, "step": 11423 }, { "epoch": 1.238508239375542, "grad_norm": 7.547423362731934, "learning_rate": 1.1785869722907298e-05, "loss": 0.0323, "step": 11424 }, { "epoch": 1.2386166522116218, "grad_norm": 2.6944329738616943, "learning_rate": 1.1785144349339912e-05, "loss": 0.0391, "step": 11425 }, { "epoch": 1.2387250650477015, "grad_norm": 0.21831412613391876, "learning_rate": 1.1784418975772525e-05, "loss": 0.0016, "step": 11426 }, { "epoch": 1.2388334778837815, "grad_norm": 1.1233524084091187, "learning_rate": 1.1783693602205138e-05, "loss": 0.0303, "step": 11427 }, { "epoch": 1.2389418907198613, "grad_norm": 1.5416842699050903, "learning_rate": 1.178296822863775e-05, "loss": 0.0269, "step": 11428 }, { "epoch": 1.239050303555941, "grad_norm": 0.4833320677280426, "learning_rate": 1.1782242855070362e-05, "loss": 0.0025, "step": 11429 }, { "epoch": 1.2391587163920208, "grad_norm": 4.680027008056641, "learning_rate": 1.1781517481502976e-05, "loss": 0.062, "step": 11430 }, { "epoch": 1.2392671292281006, "grad_norm": 13.496797561645508, "learning_rate": 1.1780792107935588e-05, "loss": 0.0101, "step": 11431 }, { "epoch": 1.2393755420641803, "grad_norm": 3.3171844482421875, "learning_rate": 1.17800667343682e-05, "loss": 0.0313, "step": 11432 }, { "epoch": 1.23948395490026, "grad_norm": 3.7106821537017822, "learning_rate": 1.1779341360800814e-05, "loss": 0.057, "step": 11433 }, { "epoch": 1.23959236773634, "grad_norm": 0.8450610041618347, "learning_rate": 1.1778615987233426e-05, "loss": 0.0124, "step": 11434 }, { "epoch": 1.2397007805724198, "grad_norm": 4.725439071655273, "learning_rate": 1.1777890613666038e-05, "loss": 0.0179, "step": 11435 }, { "epoch": 1.2398091934084996, "grad_norm": 1.1906591653823853, "learning_rate": 1.1777165240098652e-05, "loss": 0.0194, "step": 11436 }, { "epoch": 1.2399176062445794, "grad_norm": 0.8140407800674438, "learning_rate": 1.1776439866531264e-05, "loss": 0.0062, "step": 11437 }, { "epoch": 1.2400260190806591, "grad_norm": 1.592211127281189, "learning_rate": 1.1775714492963876e-05, "loss": 0.0619, "step": 11438 }, { "epoch": 1.2401344319167389, "grad_norm": 2.143948554992676, "learning_rate": 1.177498911939649e-05, "loss": 0.0703, "step": 11439 }, { "epoch": 1.2402428447528187, "grad_norm": 0.6430805325508118, "learning_rate": 1.1774263745829102e-05, "loss": 0.0172, "step": 11440 }, { "epoch": 1.2403512575888986, "grad_norm": 0.3278186023235321, "learning_rate": 1.1773538372261714e-05, "loss": 0.012, "step": 11441 }, { "epoch": 1.2404596704249784, "grad_norm": 1.465704083442688, "learning_rate": 1.177281299869433e-05, "loss": 0.0043, "step": 11442 }, { "epoch": 1.2405680832610582, "grad_norm": 1.8673489093780518, "learning_rate": 1.1772087625126942e-05, "loss": 0.0362, "step": 11443 }, { "epoch": 1.240676496097138, "grad_norm": 0.5852158069610596, "learning_rate": 1.1771362251559554e-05, "loss": 0.0051, "step": 11444 }, { "epoch": 1.2407849089332177, "grad_norm": 1.443178653717041, "learning_rate": 1.1770636877992168e-05, "loss": 0.0254, "step": 11445 }, { "epoch": 1.2408933217692975, "grad_norm": 1.2894827127456665, "learning_rate": 1.176991150442478e-05, "loss": 0.0072, "step": 11446 }, { "epoch": 1.2410017346053772, "grad_norm": 1.6336349248886108, "learning_rate": 1.1769186130857392e-05, "loss": 0.0457, "step": 11447 }, { "epoch": 1.241110147441457, "grad_norm": 0.29030701518058777, "learning_rate": 1.1768460757290006e-05, "loss": 0.0168, "step": 11448 }, { "epoch": 1.2412185602775367, "grad_norm": 0.3087719976902008, "learning_rate": 1.1767735383722618e-05, "loss": 0.01, "step": 11449 }, { "epoch": 1.2413269731136167, "grad_norm": 0.1803489774465561, "learning_rate": 1.176701001015523e-05, "loss": 0.002, "step": 11450 }, { "epoch": 1.2414353859496965, "grad_norm": 0.9101794362068176, "learning_rate": 1.1766284636587844e-05, "loss": 0.0058, "step": 11451 }, { "epoch": 1.2415437987857763, "grad_norm": 0.13379673659801483, "learning_rate": 1.1765559263020456e-05, "loss": 0.0006, "step": 11452 }, { "epoch": 1.241652211621856, "grad_norm": 0.7633090615272522, "learning_rate": 1.1764833889453069e-05, "loss": 0.0175, "step": 11453 }, { "epoch": 1.2417606244579358, "grad_norm": 0.3801858127117157, "learning_rate": 1.176410851588568e-05, "loss": 0.0043, "step": 11454 }, { "epoch": 1.2418690372940155, "grad_norm": 0.5539543032646179, "learning_rate": 1.1763383142318295e-05, "loss": 0.0294, "step": 11455 }, { "epoch": 1.2419774501300953, "grad_norm": 6.920376300811768, "learning_rate": 1.1762657768750907e-05, "loss": 0.0203, "step": 11456 }, { "epoch": 1.2420858629661753, "grad_norm": 1.8348472118377686, "learning_rate": 1.1761932395183519e-05, "loss": 0.0157, "step": 11457 }, { "epoch": 1.242194275802255, "grad_norm": 1.1627821922302246, "learning_rate": 1.1761207021616133e-05, "loss": 0.0081, "step": 11458 }, { "epoch": 1.2423026886383348, "grad_norm": 0.16504773497581482, "learning_rate": 1.1760481648048747e-05, "loss": 0.0017, "step": 11459 }, { "epoch": 1.2424111014744146, "grad_norm": 2.730712890625, "learning_rate": 1.175975627448136e-05, "loss": 0.0263, "step": 11460 }, { "epoch": 1.2425195143104943, "grad_norm": 4.921864032745361, "learning_rate": 1.1759030900913973e-05, "loss": 0.0624, "step": 11461 }, { "epoch": 1.242627927146574, "grad_norm": 2.7874693870544434, "learning_rate": 1.1758305527346585e-05, "loss": 0.048, "step": 11462 }, { "epoch": 1.2427363399826539, "grad_norm": 5.268733501434326, "learning_rate": 1.1757580153779199e-05, "loss": 0.015, "step": 11463 }, { "epoch": 1.2428447528187339, "grad_norm": 3.0133771896362305, "learning_rate": 1.175685478021181e-05, "loss": 0.0166, "step": 11464 }, { "epoch": 1.2429531656548136, "grad_norm": 26.329206466674805, "learning_rate": 1.1756129406644423e-05, "loss": 0.0728, "step": 11465 }, { "epoch": 1.2430615784908934, "grad_norm": 2.1792900562286377, "learning_rate": 1.1755404033077037e-05, "loss": 0.0307, "step": 11466 }, { "epoch": 1.2431699913269731, "grad_norm": 1.3231372833251953, "learning_rate": 1.1754678659509649e-05, "loss": 0.0068, "step": 11467 }, { "epoch": 1.243278404163053, "grad_norm": 0.2657129466533661, "learning_rate": 1.1753953285942261e-05, "loss": 0.0019, "step": 11468 }, { "epoch": 1.2433868169991327, "grad_norm": 11.242035865783691, "learning_rate": 1.1753227912374873e-05, "loss": 0.0271, "step": 11469 }, { "epoch": 1.2434952298352124, "grad_norm": 4.598475933074951, "learning_rate": 1.1752502538807487e-05, "loss": 0.0272, "step": 11470 }, { "epoch": 1.2436036426712922, "grad_norm": 0.6368927955627441, "learning_rate": 1.1751777165240099e-05, "loss": 0.0129, "step": 11471 }, { "epoch": 1.243712055507372, "grad_norm": 0.328725129365921, "learning_rate": 1.1751051791672711e-05, "loss": 0.0023, "step": 11472 }, { "epoch": 1.243820468343452, "grad_norm": 7.401363849639893, "learning_rate": 1.1750326418105325e-05, "loss": 0.0383, "step": 11473 }, { "epoch": 1.2439288811795317, "grad_norm": 4.0471320152282715, "learning_rate": 1.1749601044537937e-05, "loss": 0.0436, "step": 11474 }, { "epoch": 1.2440372940156115, "grad_norm": 0.5805746912956238, "learning_rate": 1.174887567097055e-05, "loss": 0.0037, "step": 11475 }, { "epoch": 1.2441457068516912, "grad_norm": 1.8191195726394653, "learning_rate": 1.1748150297403163e-05, "loss": 0.0313, "step": 11476 }, { "epoch": 1.244254119687771, "grad_norm": 1.783729910850525, "learning_rate": 1.1747424923835777e-05, "loss": 0.0143, "step": 11477 }, { "epoch": 1.2443625325238508, "grad_norm": 0.5169678330421448, "learning_rate": 1.174669955026839e-05, "loss": 0.0017, "step": 11478 }, { "epoch": 1.2444709453599305, "grad_norm": 0.16059793531894684, "learning_rate": 1.1745974176701003e-05, "loss": 0.0011, "step": 11479 }, { "epoch": 1.2445793581960105, "grad_norm": 0.0972558856010437, "learning_rate": 1.1745248803133615e-05, "loss": 0.0008, "step": 11480 }, { "epoch": 1.2446877710320903, "grad_norm": 0.32514384388923645, "learning_rate": 1.1744523429566227e-05, "loss": 0.0111, "step": 11481 }, { "epoch": 1.24479618386817, "grad_norm": 3.8657684326171875, "learning_rate": 1.1743798055998841e-05, "loss": 0.0338, "step": 11482 }, { "epoch": 1.2449045967042498, "grad_norm": 2.4264349937438965, "learning_rate": 1.1743072682431453e-05, "loss": 0.0283, "step": 11483 }, { "epoch": 1.2450130095403296, "grad_norm": 1.9958698749542236, "learning_rate": 1.1742347308864065e-05, "loss": 0.0358, "step": 11484 }, { "epoch": 1.2451214223764093, "grad_norm": 1.0359182357788086, "learning_rate": 1.174162193529668e-05, "loss": 0.0147, "step": 11485 }, { "epoch": 1.245229835212489, "grad_norm": 0.55156010389328, "learning_rate": 1.1740896561729291e-05, "loss": 0.0053, "step": 11486 }, { "epoch": 1.245338248048569, "grad_norm": 3.074960470199585, "learning_rate": 1.1740171188161904e-05, "loss": 0.0098, "step": 11487 }, { "epoch": 1.2454466608846488, "grad_norm": 10.827045440673828, "learning_rate": 1.1739445814594517e-05, "loss": 0.0601, "step": 11488 }, { "epoch": 1.2455550737207286, "grad_norm": 3.5624589920043945, "learning_rate": 1.173872044102713e-05, "loss": 0.0751, "step": 11489 }, { "epoch": 1.2456634865568084, "grad_norm": 0.5126752853393555, "learning_rate": 1.1737995067459742e-05, "loss": 0.0146, "step": 11490 }, { "epoch": 1.2457718993928881, "grad_norm": 1.342349886894226, "learning_rate": 1.1737269693892355e-05, "loss": 0.0254, "step": 11491 }, { "epoch": 1.245880312228968, "grad_norm": 10.308816909790039, "learning_rate": 1.1736544320324968e-05, "loss": 0.0808, "step": 11492 }, { "epoch": 1.2459887250650477, "grad_norm": 0.5564707517623901, "learning_rate": 1.173581894675758e-05, "loss": 0.0155, "step": 11493 }, { "epoch": 1.2460971379011274, "grad_norm": 2.8121159076690674, "learning_rate": 1.1735093573190195e-05, "loss": 0.0896, "step": 11494 }, { "epoch": 1.2462055507372072, "grad_norm": 7.461474418640137, "learning_rate": 1.1734368199622807e-05, "loss": 0.0203, "step": 11495 }, { "epoch": 1.2463139635732872, "grad_norm": 1.124923586845398, "learning_rate": 1.173364282605542e-05, "loss": 0.0148, "step": 11496 }, { "epoch": 1.246422376409367, "grad_norm": 0.30767586827278137, "learning_rate": 1.1732917452488033e-05, "loss": 0.0101, "step": 11497 }, { "epoch": 1.2465307892454467, "grad_norm": 2.3968982696533203, "learning_rate": 1.1732192078920646e-05, "loss": 0.0643, "step": 11498 }, { "epoch": 1.2466392020815265, "grad_norm": 1.962020754814148, "learning_rate": 1.1731466705353258e-05, "loss": 0.0459, "step": 11499 }, { "epoch": 1.2467476149176062, "grad_norm": 3.725825548171997, "learning_rate": 1.1730741331785872e-05, "loss": 0.0267, "step": 11500 }, { "epoch": 1.246856027753686, "grad_norm": 0.5680530071258545, "learning_rate": 1.1730015958218484e-05, "loss": 0.0024, "step": 11501 }, { "epoch": 1.2469644405897657, "grad_norm": 0.40130937099456787, "learning_rate": 1.1729290584651096e-05, "loss": 0.0114, "step": 11502 }, { "epoch": 1.2470728534258457, "grad_norm": 0.1798042505979538, "learning_rate": 1.172856521108371e-05, "loss": 0.014, "step": 11503 }, { "epoch": 1.2471812662619255, "grad_norm": 4.703001022338867, "learning_rate": 1.1727839837516322e-05, "loss": 0.0457, "step": 11504 }, { "epoch": 1.2472896790980053, "grad_norm": 2.9338057041168213, "learning_rate": 1.1727114463948934e-05, "loss": 0.0317, "step": 11505 }, { "epoch": 1.247398091934085, "grad_norm": 1.4048246145248413, "learning_rate": 1.1726389090381546e-05, "loss": 0.0159, "step": 11506 }, { "epoch": 1.2475065047701648, "grad_norm": 2.070378541946411, "learning_rate": 1.172566371681416e-05, "loss": 0.0316, "step": 11507 }, { "epoch": 1.2476149176062445, "grad_norm": 0.235159769654274, "learning_rate": 1.1724938343246772e-05, "loss": 0.0008, "step": 11508 }, { "epoch": 1.2477233304423243, "grad_norm": 1.1777092218399048, "learning_rate": 1.1724212969679384e-05, "loss": 0.0106, "step": 11509 }, { "epoch": 1.247831743278404, "grad_norm": 0.5376105904579163, "learning_rate": 1.1723487596111998e-05, "loss": 0.0186, "step": 11510 }, { "epoch": 1.2479401561144838, "grad_norm": 6.660817623138428, "learning_rate": 1.1722762222544612e-05, "loss": 0.032, "step": 11511 }, { "epoch": 1.2480485689505638, "grad_norm": 0.29259636998176575, "learning_rate": 1.1722036848977226e-05, "loss": 0.006, "step": 11512 }, { "epoch": 1.2481569817866436, "grad_norm": 0.8582717180252075, "learning_rate": 1.1721311475409838e-05, "loss": 0.0199, "step": 11513 }, { "epoch": 1.2482653946227233, "grad_norm": 0.505871057510376, "learning_rate": 1.172058610184245e-05, "loss": 0.006, "step": 11514 }, { "epoch": 1.248373807458803, "grad_norm": 7.369187831878662, "learning_rate": 1.1719860728275064e-05, "loss": 0.014, "step": 11515 }, { "epoch": 1.2484822202948829, "grad_norm": 1.3551548719406128, "learning_rate": 1.1719135354707676e-05, "loss": 0.054, "step": 11516 }, { "epoch": 1.2485906331309626, "grad_norm": 4.022509574890137, "learning_rate": 1.1718409981140288e-05, "loss": 0.0688, "step": 11517 }, { "epoch": 1.2486990459670424, "grad_norm": 0.12736797332763672, "learning_rate": 1.1717684607572902e-05, "loss": 0.0007, "step": 11518 }, { "epoch": 1.2488074588031224, "grad_norm": 1.24708092212677, "learning_rate": 1.1716959234005514e-05, "loss": 0.0218, "step": 11519 }, { "epoch": 1.2489158716392021, "grad_norm": 1.1695022583007812, "learning_rate": 1.1716233860438126e-05, "loss": 0.0177, "step": 11520 }, { "epoch": 1.249024284475282, "grad_norm": 0.6452866196632385, "learning_rate": 1.1715508486870738e-05, "loss": 0.006, "step": 11521 }, { "epoch": 1.2491326973113617, "grad_norm": 1.887833595275879, "learning_rate": 1.1714783113303352e-05, "loss": 0.0255, "step": 11522 }, { "epoch": 1.2492411101474414, "grad_norm": 0.5753704905509949, "learning_rate": 1.1714057739735964e-05, "loss": 0.0171, "step": 11523 }, { "epoch": 1.2493495229835212, "grad_norm": 1.9158047437667847, "learning_rate": 1.1713332366168577e-05, "loss": 0.0231, "step": 11524 }, { "epoch": 1.249457935819601, "grad_norm": 2.4048449993133545, "learning_rate": 1.171260699260119e-05, "loss": 0.0554, "step": 11525 }, { "epoch": 1.249566348655681, "grad_norm": 0.3323412537574768, "learning_rate": 1.1711881619033803e-05, "loss": 0.0054, "step": 11526 }, { "epoch": 1.2496747614917607, "grad_norm": 1.406032919883728, "learning_rate": 1.1711156245466415e-05, "loss": 0.0273, "step": 11527 }, { "epoch": 1.2497831743278405, "grad_norm": 3.9656388759613037, "learning_rate": 1.171043087189903e-05, "loss": 0.0282, "step": 11528 }, { "epoch": 1.2498915871639202, "grad_norm": 3.2849650382995605, "learning_rate": 1.1709705498331642e-05, "loss": 0.04, "step": 11529 }, { "epoch": 1.25, "grad_norm": 4.670889854431152, "learning_rate": 1.1708980124764256e-05, "loss": 0.0249, "step": 11530 }, { "epoch": 1.2501084128360798, "grad_norm": 0.9362373352050781, "learning_rate": 1.1708254751196868e-05, "loss": 0.0354, "step": 11531 }, { "epoch": 1.2502168256721595, "grad_norm": 0.4248203635215759, "learning_rate": 1.170752937762948e-05, "loss": 0.0116, "step": 11532 }, { "epoch": 1.2503252385082395, "grad_norm": 2.643928050994873, "learning_rate": 1.1706804004062093e-05, "loss": 0.0442, "step": 11533 }, { "epoch": 1.250433651344319, "grad_norm": 2.4228672981262207, "learning_rate": 1.1706078630494706e-05, "loss": 0.0286, "step": 11534 }, { "epoch": 1.250542064180399, "grad_norm": 14.951740264892578, "learning_rate": 1.1705353256927319e-05, "loss": 0.0536, "step": 11535 }, { "epoch": 1.2506504770164788, "grad_norm": 1.6172258853912354, "learning_rate": 1.170462788335993e-05, "loss": 0.0292, "step": 11536 }, { "epoch": 1.2507588898525586, "grad_norm": 5.400867938995361, "learning_rate": 1.1703902509792545e-05, "loss": 0.084, "step": 11537 }, { "epoch": 1.2508673026886383, "grad_norm": 6.022961139678955, "learning_rate": 1.1703177136225157e-05, "loss": 0.0152, "step": 11538 }, { "epoch": 1.250975715524718, "grad_norm": 1.191752552986145, "learning_rate": 1.1702451762657769e-05, "loss": 0.0199, "step": 11539 }, { "epoch": 1.2510841283607979, "grad_norm": 0.8515707850456238, "learning_rate": 1.1701726389090383e-05, "loss": 0.0125, "step": 11540 }, { "epoch": 1.2511925411968776, "grad_norm": 4.280216693878174, "learning_rate": 1.1701001015522995e-05, "loss": 0.0477, "step": 11541 }, { "epoch": 1.2513009540329576, "grad_norm": 0.6773744225502014, "learning_rate": 1.1700275641955607e-05, "loss": 0.009, "step": 11542 }, { "epoch": 1.2514093668690374, "grad_norm": 4.4032769203186035, "learning_rate": 1.169955026838822e-05, "loss": 0.0289, "step": 11543 }, { "epoch": 1.2515177797051171, "grad_norm": 3.8060760498046875, "learning_rate": 1.1698824894820833e-05, "loss": 0.0467, "step": 11544 }, { "epoch": 1.251626192541197, "grad_norm": 0.6084786057472229, "learning_rate": 1.1698099521253445e-05, "loss": 0.015, "step": 11545 }, { "epoch": 1.2517346053772767, "grad_norm": 0.1007440835237503, "learning_rate": 1.169737414768606e-05, "loss": 0.0008, "step": 11546 }, { "epoch": 1.2518430182133564, "grad_norm": 3.991689443588257, "learning_rate": 1.1696648774118673e-05, "loss": 0.0718, "step": 11547 }, { "epoch": 1.2519514310494362, "grad_norm": 0.3829799294471741, "learning_rate": 1.1695923400551285e-05, "loss": 0.0333, "step": 11548 }, { "epoch": 1.2520598438855162, "grad_norm": 0.4124218225479126, "learning_rate": 1.1695198026983899e-05, "loss": 0.0048, "step": 11549 }, { "epoch": 1.2521682567215957, "grad_norm": 1.5316007137298584, "learning_rate": 1.1694472653416511e-05, "loss": 0.0134, "step": 11550 }, { "epoch": 1.2522766695576757, "grad_norm": 5.3334455490112305, "learning_rate": 1.1693747279849123e-05, "loss": 0.0162, "step": 11551 }, { "epoch": 1.2523850823937555, "grad_norm": 1.3592625856399536, "learning_rate": 1.1693021906281737e-05, "loss": 0.0169, "step": 11552 }, { "epoch": 1.2524934952298352, "grad_norm": 1.2724473476409912, "learning_rate": 1.1692296532714349e-05, "loss": 0.0381, "step": 11553 }, { "epoch": 1.252601908065915, "grad_norm": 10.665823936462402, "learning_rate": 1.1691571159146961e-05, "loss": 0.0652, "step": 11554 }, { "epoch": 1.2527103209019947, "grad_norm": 6.732754230499268, "learning_rate": 1.1690845785579575e-05, "loss": 0.0298, "step": 11555 }, { "epoch": 1.2528187337380745, "grad_norm": 0.5602063536643982, "learning_rate": 1.1690120412012187e-05, "loss": 0.0041, "step": 11556 }, { "epoch": 1.2529271465741543, "grad_norm": 0.43460702896118164, "learning_rate": 1.16893950384448e-05, "loss": 0.0026, "step": 11557 }, { "epoch": 1.2530355594102343, "grad_norm": 7.899319648742676, "learning_rate": 1.1688669664877413e-05, "loss": 0.038, "step": 11558 }, { "epoch": 1.253143972246314, "grad_norm": 0.04896154627203941, "learning_rate": 1.1687944291310025e-05, "loss": 0.0005, "step": 11559 }, { "epoch": 1.2532523850823938, "grad_norm": 0.8987048268318176, "learning_rate": 1.1687218917742637e-05, "loss": 0.0053, "step": 11560 }, { "epoch": 1.2533607979184735, "grad_norm": 0.12475289404392242, "learning_rate": 1.168649354417525e-05, "loss": 0.0013, "step": 11561 }, { "epoch": 1.2534692107545533, "grad_norm": 1.261684536933899, "learning_rate": 1.1685768170607863e-05, "loss": 0.0146, "step": 11562 }, { "epoch": 1.253577623590633, "grad_norm": 3.7115015983581543, "learning_rate": 1.1685042797040477e-05, "loss": 0.0274, "step": 11563 }, { "epoch": 1.2536860364267128, "grad_norm": 3.3855206966400146, "learning_rate": 1.1684317423473091e-05, "loss": 0.049, "step": 11564 }, { "epoch": 1.2537944492627928, "grad_norm": 1.7619708776474, "learning_rate": 1.1683592049905703e-05, "loss": 0.0228, "step": 11565 }, { "epoch": 1.2539028620988726, "grad_norm": 2.63362193107605, "learning_rate": 1.1682866676338315e-05, "loss": 0.0536, "step": 11566 }, { "epoch": 1.2540112749349523, "grad_norm": 6.345800876617432, "learning_rate": 1.168214130277093e-05, "loss": 0.0449, "step": 11567 }, { "epoch": 1.254119687771032, "grad_norm": 0.7514756917953491, "learning_rate": 1.1681415929203541e-05, "loss": 0.0132, "step": 11568 }, { "epoch": 1.2542281006071119, "grad_norm": 1.8494629859924316, "learning_rate": 1.1680690555636153e-05, "loss": 0.0198, "step": 11569 }, { "epoch": 1.2543365134431916, "grad_norm": 0.7426497936248779, "learning_rate": 1.1679965182068767e-05, "loss": 0.0145, "step": 11570 }, { "epoch": 1.2544449262792714, "grad_norm": 2.320350170135498, "learning_rate": 1.167923980850138e-05, "loss": 0.0658, "step": 11571 }, { "epoch": 1.2545533391153514, "grad_norm": 0.4210459589958191, "learning_rate": 1.1678514434933992e-05, "loss": 0.0085, "step": 11572 }, { "epoch": 1.254661751951431, "grad_norm": 0.9136366844177246, "learning_rate": 1.1677789061366604e-05, "loss": 0.0137, "step": 11573 }, { "epoch": 1.254770164787511, "grad_norm": 0.8272304534912109, "learning_rate": 1.1677063687799218e-05, "loss": 0.0073, "step": 11574 }, { "epoch": 1.2548785776235907, "grad_norm": 3.7178826332092285, "learning_rate": 1.167633831423183e-05, "loss": 0.031, "step": 11575 }, { "epoch": 1.2549869904596704, "grad_norm": 4.669285774230957, "learning_rate": 1.1675612940664442e-05, "loss": 0.0361, "step": 11576 }, { "epoch": 1.2550954032957502, "grad_norm": 3.4335505962371826, "learning_rate": 1.1674887567097056e-05, "loss": 0.0303, "step": 11577 }, { "epoch": 1.25520381613183, "grad_norm": 0.05897742137312889, "learning_rate": 1.1674162193529668e-05, "loss": 0.0007, "step": 11578 }, { "epoch": 1.2553122289679097, "grad_norm": 1.8534170389175415, "learning_rate": 1.167343681996228e-05, "loss": 0.0295, "step": 11579 }, { "epoch": 1.2554206418039895, "grad_norm": 0.6155359148979187, "learning_rate": 1.1672711446394896e-05, "loss": 0.0141, "step": 11580 }, { "epoch": 1.2555290546400695, "grad_norm": 7.262537479400635, "learning_rate": 1.1671986072827508e-05, "loss": 0.0135, "step": 11581 }, { "epoch": 1.2556374674761492, "grad_norm": 1.092545986175537, "learning_rate": 1.1671260699260122e-05, "loss": 0.0123, "step": 11582 }, { "epoch": 1.255745880312229, "grad_norm": 5.276129245758057, "learning_rate": 1.1670535325692734e-05, "loss": 0.0283, "step": 11583 }, { "epoch": 1.2558542931483088, "grad_norm": 1.4463614225387573, "learning_rate": 1.1669809952125346e-05, "loss": 0.0432, "step": 11584 }, { "epoch": 1.2559627059843885, "grad_norm": 1.1603800058364868, "learning_rate": 1.1669084578557958e-05, "loss": 0.0338, "step": 11585 }, { "epoch": 1.2560711188204683, "grad_norm": 2.8425207138061523, "learning_rate": 1.1668359204990572e-05, "loss": 0.0466, "step": 11586 }, { "epoch": 1.256179531656548, "grad_norm": 1.7357113361358643, "learning_rate": 1.1667633831423184e-05, "loss": 0.0092, "step": 11587 }, { "epoch": 1.256287944492628, "grad_norm": 3.725757360458374, "learning_rate": 1.1666908457855796e-05, "loss": 0.0242, "step": 11588 }, { "epoch": 1.2563963573287076, "grad_norm": 0.9589143395423889, "learning_rate": 1.166618308428841e-05, "loss": 0.0149, "step": 11589 }, { "epoch": 1.2565047701647876, "grad_norm": 1.215027093887329, "learning_rate": 1.1665457710721022e-05, "loss": 0.0184, "step": 11590 }, { "epoch": 1.2566131830008673, "grad_norm": 3.9114935398101807, "learning_rate": 1.1664732337153634e-05, "loss": 0.0212, "step": 11591 }, { "epoch": 1.256721595836947, "grad_norm": 0.524928629398346, "learning_rate": 1.1664006963586248e-05, "loss": 0.0201, "step": 11592 }, { "epoch": 1.2568300086730269, "grad_norm": 1.7863647937774658, "learning_rate": 1.166328159001886e-05, "loss": 0.0227, "step": 11593 }, { "epoch": 1.2569384215091066, "grad_norm": 0.5139811038970947, "learning_rate": 1.1662556216451472e-05, "loss": 0.008, "step": 11594 }, { "epoch": 1.2570468343451866, "grad_norm": 2.364304542541504, "learning_rate": 1.1661830842884086e-05, "loss": 0.0353, "step": 11595 }, { "epoch": 1.2571552471812661, "grad_norm": 3.0755958557128906, "learning_rate": 1.1661105469316698e-05, "loss": 0.0544, "step": 11596 }, { "epoch": 1.2572636600173461, "grad_norm": 2.447596549987793, "learning_rate": 1.1660380095749312e-05, "loss": 0.0227, "step": 11597 }, { "epoch": 1.257372072853426, "grad_norm": 0.9381093382835388, "learning_rate": 1.1659654722181926e-05, "loss": 0.0235, "step": 11598 }, { "epoch": 1.2574804856895057, "grad_norm": 2.9549646377563477, "learning_rate": 1.1658929348614538e-05, "loss": 0.012, "step": 11599 }, { "epoch": 1.2575888985255854, "grad_norm": 1.3651354312896729, "learning_rate": 1.165820397504715e-05, "loss": 0.0308, "step": 11600 }, { "epoch": 1.2576973113616652, "grad_norm": 1.7185343503952026, "learning_rate": 1.1657478601479764e-05, "loss": 0.0344, "step": 11601 }, { "epoch": 1.257805724197745, "grad_norm": 1.567839503288269, "learning_rate": 1.1656753227912376e-05, "loss": 0.0091, "step": 11602 }, { "epoch": 1.2579141370338247, "grad_norm": 2.175079345703125, "learning_rate": 1.1656027854344988e-05, "loss": 0.0442, "step": 11603 }, { "epoch": 1.2580225498699047, "grad_norm": 0.2552185356616974, "learning_rate": 1.1655302480777602e-05, "loss": 0.0039, "step": 11604 }, { "epoch": 1.2581309627059845, "grad_norm": 2.1017391681671143, "learning_rate": 1.1654577107210214e-05, "loss": 0.0162, "step": 11605 }, { "epoch": 1.2582393755420642, "grad_norm": 0.6479749083518982, "learning_rate": 1.1653851733642826e-05, "loss": 0.0102, "step": 11606 }, { "epoch": 1.258347788378144, "grad_norm": 1.466694951057434, "learning_rate": 1.165312636007544e-05, "loss": 0.0234, "step": 11607 }, { "epoch": 1.2584562012142237, "grad_norm": 4.656468391418457, "learning_rate": 1.1652400986508052e-05, "loss": 0.0725, "step": 11608 }, { "epoch": 1.2585646140503035, "grad_norm": 1.79865562915802, "learning_rate": 1.1651675612940665e-05, "loss": 0.0237, "step": 11609 }, { "epoch": 1.2586730268863833, "grad_norm": 8.275565147399902, "learning_rate": 1.1650950239373278e-05, "loss": 0.0902, "step": 11610 }, { "epoch": 1.2587814397224633, "grad_norm": 9.119538307189941, "learning_rate": 1.165022486580589e-05, "loss": 0.1003, "step": 11611 }, { "epoch": 1.2588898525585428, "grad_norm": 5.775468826293945, "learning_rate": 1.1649499492238503e-05, "loss": 0.0084, "step": 11612 }, { "epoch": 1.2589982653946228, "grad_norm": 1.352510690689087, "learning_rate": 1.1648774118671115e-05, "loss": 0.0145, "step": 11613 }, { "epoch": 1.2591066782307025, "grad_norm": 1.3584043979644775, "learning_rate": 1.1648048745103729e-05, "loss": 0.0061, "step": 11614 }, { "epoch": 1.2592150910667823, "grad_norm": 1.8770959377288818, "learning_rate": 1.1647323371536343e-05, "loss": 0.0172, "step": 11615 }, { "epoch": 1.259323503902862, "grad_norm": 7.311827659606934, "learning_rate": 1.1646597997968956e-05, "loss": 0.0749, "step": 11616 }, { "epoch": 1.2594319167389418, "grad_norm": 1.9361746311187744, "learning_rate": 1.1645872624401569e-05, "loss": 0.0269, "step": 11617 }, { "epoch": 1.2595403295750216, "grad_norm": 1.1269601583480835, "learning_rate": 1.164514725083418e-05, "loss": 0.0109, "step": 11618 }, { "epoch": 1.2596487424111014, "grad_norm": 9.37637996673584, "learning_rate": 1.1644421877266795e-05, "loss": 0.045, "step": 11619 }, { "epoch": 1.2597571552471813, "grad_norm": 0.5067290663719177, "learning_rate": 1.1643696503699407e-05, "loss": 0.0023, "step": 11620 }, { "epoch": 1.2598655680832611, "grad_norm": 1.5236955881118774, "learning_rate": 1.1642971130132019e-05, "loss": 0.0214, "step": 11621 }, { "epoch": 1.2599739809193409, "grad_norm": 0.8241915106773376, "learning_rate": 1.1642245756564633e-05, "loss": 0.0087, "step": 11622 }, { "epoch": 1.2600823937554206, "grad_norm": 5.403717041015625, "learning_rate": 1.1641520382997245e-05, "loss": 0.0269, "step": 11623 }, { "epoch": 1.2601908065915004, "grad_norm": 1.8234187364578247, "learning_rate": 1.1640795009429857e-05, "loss": 0.0487, "step": 11624 }, { "epoch": 1.2602992194275802, "grad_norm": 0.5279690027236938, "learning_rate": 1.1640069635862469e-05, "loss": 0.0241, "step": 11625 }, { "epoch": 1.26040763226366, "grad_norm": 2.078507423400879, "learning_rate": 1.1639344262295083e-05, "loss": 0.0074, "step": 11626 }, { "epoch": 1.26051604509974, "grad_norm": 7.401886463165283, "learning_rate": 1.1638618888727695e-05, "loss": 0.0563, "step": 11627 }, { "epoch": 1.2606244579358197, "grad_norm": 6.085056781768799, "learning_rate": 1.1637893515160307e-05, "loss": 0.0573, "step": 11628 }, { "epoch": 1.2607328707718994, "grad_norm": 2.0394904613494873, "learning_rate": 1.1637168141592921e-05, "loss": 0.0704, "step": 11629 }, { "epoch": 1.2608412836079792, "grad_norm": 0.4579128324985504, "learning_rate": 1.1636442768025533e-05, "loss": 0.0191, "step": 11630 }, { "epoch": 1.260949696444059, "grad_norm": 4.45080041885376, "learning_rate": 1.1635717394458145e-05, "loss": 0.0138, "step": 11631 }, { "epoch": 1.2610581092801387, "grad_norm": 0.22360728681087494, "learning_rate": 1.1634992020890761e-05, "loss": 0.0017, "step": 11632 }, { "epoch": 1.2611665221162185, "grad_norm": 0.39081838726997375, "learning_rate": 1.1634266647323373e-05, "loss": 0.0134, "step": 11633 }, { "epoch": 1.2612749349522985, "grad_norm": 3.1962034702301025, "learning_rate": 1.1633541273755987e-05, "loss": 0.0321, "step": 11634 }, { "epoch": 1.261383347788378, "grad_norm": 6.417614459991455, "learning_rate": 1.1632815900188599e-05, "loss": 0.0181, "step": 11635 }, { "epoch": 1.261491760624458, "grad_norm": 1.5337165594100952, "learning_rate": 1.1632090526621211e-05, "loss": 0.0117, "step": 11636 }, { "epoch": 1.2616001734605378, "grad_norm": 1.5517436265945435, "learning_rate": 1.1631365153053823e-05, "loss": 0.0615, "step": 11637 }, { "epoch": 1.2617085862966175, "grad_norm": 0.36657318472862244, "learning_rate": 1.1630639779486437e-05, "loss": 0.009, "step": 11638 }, { "epoch": 1.2618169991326973, "grad_norm": 1.281580924987793, "learning_rate": 1.162991440591905e-05, "loss": 0.0155, "step": 11639 }, { "epoch": 1.261925411968777, "grad_norm": 0.41593170166015625, "learning_rate": 1.1629189032351661e-05, "loss": 0.0173, "step": 11640 }, { "epoch": 1.2620338248048568, "grad_norm": 1.0609934329986572, "learning_rate": 1.1628463658784275e-05, "loss": 0.015, "step": 11641 }, { "epoch": 1.2621422376409366, "grad_norm": 5.95967960357666, "learning_rate": 1.1627738285216887e-05, "loss": 0.0213, "step": 11642 }, { "epoch": 1.2622506504770166, "grad_norm": 0.9813323020935059, "learning_rate": 1.16270129116495e-05, "loss": 0.0055, "step": 11643 }, { "epoch": 1.2623590633130963, "grad_norm": 0.25827956199645996, "learning_rate": 1.1626287538082113e-05, "loss": 0.0013, "step": 11644 }, { "epoch": 1.262467476149176, "grad_norm": 0.8774774670600891, "learning_rate": 1.1625562164514726e-05, "loss": 0.0108, "step": 11645 }, { "epoch": 1.2625758889852559, "grad_norm": 1.2560619115829468, "learning_rate": 1.1624836790947338e-05, "loss": 0.0253, "step": 11646 }, { "epoch": 1.2626843018213356, "grad_norm": 2.7519893646240234, "learning_rate": 1.1624111417379951e-05, "loss": 0.041, "step": 11647 }, { "epoch": 1.2627927146574154, "grad_norm": 1.4581882953643799, "learning_rate": 1.1623386043812564e-05, "loss": 0.0133, "step": 11648 }, { "epoch": 1.2629011274934951, "grad_norm": 0.21947194635868073, "learning_rate": 1.1622660670245177e-05, "loss": 0.0017, "step": 11649 }, { "epoch": 1.2630095403295751, "grad_norm": 4.049546718597412, "learning_rate": 1.1621935296677791e-05, "loss": 0.0242, "step": 11650 }, { "epoch": 1.263117953165655, "grad_norm": 1.1488574743270874, "learning_rate": 1.1621209923110403e-05, "loss": 0.0171, "step": 11651 }, { "epoch": 1.2632263660017347, "grad_norm": 5.749540328979492, "learning_rate": 1.1620484549543016e-05, "loss": 0.04, "step": 11652 }, { "epoch": 1.2633347788378144, "grad_norm": 4.038445472717285, "learning_rate": 1.161975917597563e-05, "loss": 0.0287, "step": 11653 }, { "epoch": 1.2634431916738942, "grad_norm": 0.411436527967453, "learning_rate": 1.1619033802408242e-05, "loss": 0.0041, "step": 11654 }, { "epoch": 1.263551604509974, "grad_norm": 5.568243503570557, "learning_rate": 1.1618308428840854e-05, "loss": 0.012, "step": 11655 }, { "epoch": 1.2636600173460537, "grad_norm": 0.7856410145759583, "learning_rate": 1.1617583055273468e-05, "loss": 0.014, "step": 11656 }, { "epoch": 1.2637684301821337, "grad_norm": 5.99152946472168, "learning_rate": 1.161685768170608e-05, "loss": 0.0476, "step": 11657 }, { "epoch": 1.2638768430182132, "grad_norm": 3.9797329902648926, "learning_rate": 1.1616132308138692e-05, "loss": 0.0258, "step": 11658 }, { "epoch": 1.2639852558542932, "grad_norm": 0.48559123277664185, "learning_rate": 1.1615406934571306e-05, "loss": 0.0086, "step": 11659 }, { "epoch": 1.264093668690373, "grad_norm": 1.9720189571380615, "learning_rate": 1.1614681561003918e-05, "loss": 0.0348, "step": 11660 }, { "epoch": 1.2642020815264527, "grad_norm": 0.6364951729774475, "learning_rate": 1.161395618743653e-05, "loss": 0.0037, "step": 11661 }, { "epoch": 1.2643104943625325, "grad_norm": 1.8759725093841553, "learning_rate": 1.1613230813869144e-05, "loss": 0.0238, "step": 11662 }, { "epoch": 1.2644189071986123, "grad_norm": 1.371801495552063, "learning_rate": 1.1612505440301756e-05, "loss": 0.0444, "step": 11663 }, { "epoch": 1.264527320034692, "grad_norm": 2.5053439140319824, "learning_rate": 1.1611780066734368e-05, "loss": 0.0345, "step": 11664 }, { "epoch": 1.2646357328707718, "grad_norm": 2.4469547271728516, "learning_rate": 1.161105469316698e-05, "loss": 0.0073, "step": 11665 }, { "epoch": 1.2647441457068518, "grad_norm": 0.6190223693847656, "learning_rate": 1.1610329319599594e-05, "loss": 0.0104, "step": 11666 }, { "epoch": 1.2648525585429315, "grad_norm": 1.079531192779541, "learning_rate": 1.1609603946032208e-05, "loss": 0.0133, "step": 11667 }, { "epoch": 1.2649609713790113, "grad_norm": 4.2157063484191895, "learning_rate": 1.1608878572464822e-05, "loss": 0.0129, "step": 11668 }, { "epoch": 1.265069384215091, "grad_norm": 2.8257932662963867, "learning_rate": 1.1608153198897434e-05, "loss": 0.0145, "step": 11669 }, { "epoch": 1.2651777970511708, "grad_norm": 6.663112163543701, "learning_rate": 1.1607427825330046e-05, "loss": 0.0202, "step": 11670 }, { "epoch": 1.2652862098872506, "grad_norm": 5.639617443084717, "learning_rate": 1.160670245176266e-05, "loss": 0.0109, "step": 11671 }, { "epoch": 1.2653946227233304, "grad_norm": 0.5475360155105591, "learning_rate": 1.1605977078195272e-05, "loss": 0.004, "step": 11672 }, { "epoch": 1.2655030355594103, "grad_norm": 0.5634406208992004, "learning_rate": 1.1605251704627884e-05, "loss": 0.0104, "step": 11673 }, { "epoch": 1.26561144839549, "grad_norm": 0.290113240480423, "learning_rate": 1.1604526331060498e-05, "loss": 0.0126, "step": 11674 }, { "epoch": 1.2657198612315699, "grad_norm": 0.24899519979953766, "learning_rate": 1.160380095749311e-05, "loss": 0.0025, "step": 11675 }, { "epoch": 1.2658282740676496, "grad_norm": 7.837523460388184, "learning_rate": 1.1603075583925722e-05, "loss": 0.0293, "step": 11676 }, { "epoch": 1.2659366869037294, "grad_norm": 1.4848605394363403, "learning_rate": 1.1602350210358334e-05, "loss": 0.0265, "step": 11677 }, { "epoch": 1.2660450997398092, "grad_norm": 5.111189842224121, "learning_rate": 1.1601624836790948e-05, "loss": 0.0462, "step": 11678 }, { "epoch": 1.266153512575889, "grad_norm": 2.043771743774414, "learning_rate": 1.160089946322356e-05, "loss": 0.0085, "step": 11679 }, { "epoch": 1.266261925411969, "grad_norm": 0.4934147000312805, "learning_rate": 1.1600174089656173e-05, "loss": 0.0168, "step": 11680 }, { "epoch": 1.2663703382480485, "grad_norm": 0.8218798637390137, "learning_rate": 1.1599448716088786e-05, "loss": 0.0272, "step": 11681 }, { "epoch": 1.2664787510841284, "grad_norm": 0.23516806960105896, "learning_rate": 1.1598723342521399e-05, "loss": 0.0052, "step": 11682 }, { "epoch": 1.2665871639202082, "grad_norm": 1.7980579137802124, "learning_rate": 1.159799796895401e-05, "loss": 0.0324, "step": 11683 }, { "epoch": 1.266695576756288, "grad_norm": 4.340123653411865, "learning_rate": 1.1597272595386626e-05, "loss": 0.0635, "step": 11684 }, { "epoch": 1.2668039895923677, "grad_norm": 0.5485577583312988, "learning_rate": 1.1596547221819238e-05, "loss": 0.0164, "step": 11685 }, { "epoch": 1.2669124024284475, "grad_norm": 4.282956600189209, "learning_rate": 1.1595821848251852e-05, "loss": 0.0183, "step": 11686 }, { "epoch": 1.2670208152645273, "grad_norm": 0.9645175933837891, "learning_rate": 1.1595096474684464e-05, "loss": 0.0037, "step": 11687 }, { "epoch": 1.267129228100607, "grad_norm": 0.5727105736732483, "learning_rate": 1.1594371101117076e-05, "loss": 0.0131, "step": 11688 }, { "epoch": 1.267237640936687, "grad_norm": 1.4587088823318481, "learning_rate": 1.1593645727549689e-05, "loss": 0.0241, "step": 11689 }, { "epoch": 1.2673460537727668, "grad_norm": 5.0032830238342285, "learning_rate": 1.1592920353982302e-05, "loss": 0.0099, "step": 11690 }, { "epoch": 1.2674544666088465, "grad_norm": 1.4845900535583496, "learning_rate": 1.1592194980414915e-05, "loss": 0.036, "step": 11691 }, { "epoch": 1.2675628794449263, "grad_norm": 2.394366979598999, "learning_rate": 1.1591469606847527e-05, "loss": 0.008, "step": 11692 }, { "epoch": 1.267671292281006, "grad_norm": 3.57580828666687, "learning_rate": 1.159074423328014e-05, "loss": 0.0461, "step": 11693 }, { "epoch": 1.2677797051170858, "grad_norm": 2.16882586479187, "learning_rate": 1.1590018859712753e-05, "loss": 0.0424, "step": 11694 }, { "epoch": 1.2678881179531656, "grad_norm": 0.3523558974266052, "learning_rate": 1.1589293486145365e-05, "loss": 0.0019, "step": 11695 }, { "epoch": 1.2679965307892456, "grad_norm": 5.8326029777526855, "learning_rate": 1.1588568112577979e-05, "loss": 0.0403, "step": 11696 }, { "epoch": 1.268104943625325, "grad_norm": 7.450520992279053, "learning_rate": 1.158784273901059e-05, "loss": 0.0894, "step": 11697 }, { "epoch": 1.268213356461405, "grad_norm": 5.63214111328125, "learning_rate": 1.1587117365443203e-05, "loss": 0.0559, "step": 11698 }, { "epoch": 1.2683217692974849, "grad_norm": 1.781421422958374, "learning_rate": 1.1586391991875817e-05, "loss": 0.0356, "step": 11699 }, { "epoch": 1.2684301821335646, "grad_norm": 1.3507951498031616, "learning_rate": 1.1585666618308429e-05, "loss": 0.0022, "step": 11700 }, { "epoch": 1.2685385949696444, "grad_norm": 1.1475061178207397, "learning_rate": 1.1584941244741043e-05, "loss": 0.0136, "step": 11701 }, { "epoch": 1.2686470078057241, "grad_norm": 2.338012456893921, "learning_rate": 1.1584215871173657e-05, "loss": 0.0202, "step": 11702 }, { "epoch": 1.268755420641804, "grad_norm": 2.407444477081299, "learning_rate": 1.1583490497606269e-05, "loss": 0.0499, "step": 11703 }, { "epoch": 1.2688638334778837, "grad_norm": 3.5966012477874756, "learning_rate": 1.1582765124038881e-05, "loss": 0.0279, "step": 11704 }, { "epoch": 1.2689722463139637, "grad_norm": 3.318027973175049, "learning_rate": 1.1582039750471495e-05, "loss": 0.035, "step": 11705 }, { "epoch": 1.2690806591500434, "grad_norm": 9.276013374328613, "learning_rate": 1.1581314376904107e-05, "loss": 0.0209, "step": 11706 }, { "epoch": 1.2691890719861232, "grad_norm": 0.6649794578552246, "learning_rate": 1.1580589003336719e-05, "loss": 0.0136, "step": 11707 }, { "epoch": 1.269297484822203, "grad_norm": 1.8097642660140991, "learning_rate": 1.1579863629769333e-05, "loss": 0.0182, "step": 11708 }, { "epoch": 1.2694058976582827, "grad_norm": 1.4471596479415894, "learning_rate": 1.1579138256201945e-05, "loss": 0.0126, "step": 11709 }, { "epoch": 1.2695143104943625, "grad_norm": 2.9517180919647217, "learning_rate": 1.1578412882634557e-05, "loss": 0.0264, "step": 11710 }, { "epoch": 1.2696227233304422, "grad_norm": 2.0816967487335205, "learning_rate": 1.1577687509067171e-05, "loss": 0.0144, "step": 11711 }, { "epoch": 1.2697311361665222, "grad_norm": 2.364035129547119, "learning_rate": 1.1576962135499783e-05, "loss": 0.0478, "step": 11712 }, { "epoch": 1.269839549002602, "grad_norm": 1.3126291036605835, "learning_rate": 1.1576236761932395e-05, "loss": 0.0317, "step": 11713 }, { "epoch": 1.2699479618386817, "grad_norm": 0.30004438757896423, "learning_rate": 1.1575511388365009e-05, "loss": 0.0098, "step": 11714 }, { "epoch": 1.2700563746747615, "grad_norm": 0.6924656629562378, "learning_rate": 1.1574786014797621e-05, "loss": 0.0202, "step": 11715 }, { "epoch": 1.2701647875108413, "grad_norm": 2.494326114654541, "learning_rate": 1.1574060641230233e-05, "loss": 0.0153, "step": 11716 }, { "epoch": 1.270273200346921, "grad_norm": 3.282778739929199, "learning_rate": 1.1573335267662846e-05, "loss": 0.044, "step": 11717 }, { "epoch": 1.2703816131830008, "grad_norm": 0.8661418557167053, "learning_rate": 1.1572609894095461e-05, "loss": 0.0103, "step": 11718 }, { "epoch": 1.2704900260190808, "grad_norm": 3.055284261703491, "learning_rate": 1.1571884520528073e-05, "loss": 0.0259, "step": 11719 }, { "epoch": 1.2705984388551603, "grad_norm": 0.43691328167915344, "learning_rate": 1.1571159146960687e-05, "loss": 0.0095, "step": 11720 }, { "epoch": 1.2707068516912403, "grad_norm": 0.9574017524719238, "learning_rate": 1.15704337733933e-05, "loss": 0.0234, "step": 11721 }, { "epoch": 1.27081526452732, "grad_norm": 1.2015740871429443, "learning_rate": 1.1569708399825911e-05, "loss": 0.007, "step": 11722 }, { "epoch": 1.2709236773633998, "grad_norm": 2.332998037338257, "learning_rate": 1.1568983026258525e-05, "loss": 0.0152, "step": 11723 }, { "epoch": 1.2710320901994796, "grad_norm": 1.8877367973327637, "learning_rate": 1.1568257652691137e-05, "loss": 0.0362, "step": 11724 }, { "epoch": 1.2711405030355594, "grad_norm": 7.1265177726745605, "learning_rate": 1.156753227912375e-05, "loss": 0.0325, "step": 11725 }, { "epoch": 1.2712489158716391, "grad_norm": 5.394772529602051, "learning_rate": 1.1566806905556363e-05, "loss": 0.007, "step": 11726 }, { "epoch": 1.271357328707719, "grad_norm": 1.5565176010131836, "learning_rate": 1.1566081531988975e-05, "loss": 0.0428, "step": 11727 }, { "epoch": 1.2714657415437989, "grad_norm": 1.3213165998458862, "learning_rate": 1.1565356158421588e-05, "loss": 0.0052, "step": 11728 }, { "epoch": 1.2715741543798786, "grad_norm": 1.1445313692092896, "learning_rate": 1.15646307848542e-05, "loss": 0.0793, "step": 11729 }, { "epoch": 1.2716825672159584, "grad_norm": 3.6415750980377197, "learning_rate": 1.1563905411286814e-05, "loss": 0.0264, "step": 11730 }, { "epoch": 1.2717909800520382, "grad_norm": 1.3925343751907349, "learning_rate": 1.1563180037719426e-05, "loss": 0.0061, "step": 11731 }, { "epoch": 1.271899392888118, "grad_norm": 1.4261023998260498, "learning_rate": 1.1562454664152038e-05, "loss": 0.021, "step": 11732 }, { "epoch": 1.2720078057241977, "grad_norm": 1.109451413154602, "learning_rate": 1.1561729290584652e-05, "loss": 0.0269, "step": 11733 }, { "epoch": 1.2721162185602775, "grad_norm": 0.49439501762390137, "learning_rate": 1.1561003917017264e-05, "loss": 0.0017, "step": 11734 }, { "epoch": 1.2722246313963574, "grad_norm": 0.17422227561473846, "learning_rate": 1.1560278543449876e-05, "loss": 0.0012, "step": 11735 }, { "epoch": 1.272333044232437, "grad_norm": 4.580779552459717, "learning_rate": 1.1559553169882492e-05, "loss": 0.0597, "step": 11736 }, { "epoch": 1.272441457068517, "grad_norm": 3.1922266483306885, "learning_rate": 1.1558827796315104e-05, "loss": 0.027, "step": 11737 }, { "epoch": 1.2725498699045967, "grad_norm": 2.8589916229248047, "learning_rate": 1.1558102422747718e-05, "loss": 0.0176, "step": 11738 }, { "epoch": 1.2726582827406765, "grad_norm": 3.7778172492980957, "learning_rate": 1.155737704918033e-05, "loss": 0.0104, "step": 11739 }, { "epoch": 1.2727666955767563, "grad_norm": 3.6660702228546143, "learning_rate": 1.1556651675612942e-05, "loss": 0.0173, "step": 11740 }, { "epoch": 1.272875108412836, "grad_norm": 0.7275188565254211, "learning_rate": 1.1555926302045554e-05, "loss": 0.0194, "step": 11741 }, { "epoch": 1.272983521248916, "grad_norm": 2.847399950027466, "learning_rate": 1.1555200928478168e-05, "loss": 0.0177, "step": 11742 }, { "epoch": 1.2730919340849955, "grad_norm": 8.161880493164062, "learning_rate": 1.155447555491078e-05, "loss": 0.0172, "step": 11743 }, { "epoch": 1.2732003469210755, "grad_norm": 2.9556562900543213, "learning_rate": 1.1553750181343392e-05, "loss": 0.0176, "step": 11744 }, { "epoch": 1.2733087597571553, "grad_norm": 1.4619834423065186, "learning_rate": 1.1553024807776006e-05, "loss": 0.0278, "step": 11745 }, { "epoch": 1.273417172593235, "grad_norm": 0.3691202402114868, "learning_rate": 1.1552299434208618e-05, "loss": 0.0124, "step": 11746 }, { "epoch": 1.2735255854293148, "grad_norm": 2.4339935779571533, "learning_rate": 1.155157406064123e-05, "loss": 0.0266, "step": 11747 }, { "epoch": 1.2736339982653946, "grad_norm": 0.40821021795272827, "learning_rate": 1.1550848687073844e-05, "loss": 0.0159, "step": 11748 }, { "epoch": 1.2737424111014743, "grad_norm": 0.3539700508117676, "learning_rate": 1.1550123313506456e-05, "loss": 0.0099, "step": 11749 }, { "epoch": 1.273850823937554, "grad_norm": 1.7614589929580688, "learning_rate": 1.1549397939939068e-05, "loss": 0.0086, "step": 11750 }, { "epoch": 1.273959236773634, "grad_norm": 1.2376031875610352, "learning_rate": 1.1548672566371682e-05, "loss": 0.0424, "step": 11751 }, { "epoch": 1.2740676496097139, "grad_norm": 0.3769613206386566, "learning_rate": 1.1547947192804294e-05, "loss": 0.0222, "step": 11752 }, { "epoch": 1.2741760624457936, "grad_norm": 0.40041840076446533, "learning_rate": 1.154722181923691e-05, "loss": 0.003, "step": 11753 }, { "epoch": 1.2742844752818734, "grad_norm": 1.1892063617706299, "learning_rate": 1.1546496445669522e-05, "loss": 0.0094, "step": 11754 }, { "epoch": 1.2743928881179531, "grad_norm": 2.677137851715088, "learning_rate": 1.1545771072102134e-05, "loss": 0.0409, "step": 11755 }, { "epoch": 1.274501300954033, "grad_norm": 0.7710829973220825, "learning_rate": 1.1545045698534746e-05, "loss": 0.0084, "step": 11756 }, { "epoch": 1.2746097137901127, "grad_norm": 3.1982812881469727, "learning_rate": 1.154432032496736e-05, "loss": 0.0245, "step": 11757 }, { "epoch": 1.2747181266261927, "grad_norm": 2.175248622894287, "learning_rate": 1.1543594951399972e-05, "loss": 0.0307, "step": 11758 }, { "epoch": 1.2748265394622722, "grad_norm": 6.105396270751953, "learning_rate": 1.1542869577832584e-05, "loss": 0.0122, "step": 11759 }, { "epoch": 1.2749349522983522, "grad_norm": 2.078312635421753, "learning_rate": 1.1542144204265198e-05, "loss": 0.0525, "step": 11760 }, { "epoch": 1.275043365134432, "grad_norm": 1.5707597732543945, "learning_rate": 1.154141883069781e-05, "loss": 0.0451, "step": 11761 }, { "epoch": 1.2751517779705117, "grad_norm": 1.2972042560577393, "learning_rate": 1.1540693457130423e-05, "loss": 0.027, "step": 11762 }, { "epoch": 1.2752601908065915, "grad_norm": 3.238976240158081, "learning_rate": 1.1539968083563036e-05, "loss": 0.0572, "step": 11763 }, { "epoch": 1.2753686036426712, "grad_norm": 0.2264994978904724, "learning_rate": 1.1539242709995648e-05, "loss": 0.0013, "step": 11764 }, { "epoch": 1.275477016478751, "grad_norm": 1.49667489528656, "learning_rate": 1.153851733642826e-05, "loss": 0.0123, "step": 11765 }, { "epoch": 1.2755854293148308, "grad_norm": 0.7027671933174133, "learning_rate": 1.1537791962860874e-05, "loss": 0.0148, "step": 11766 }, { "epoch": 1.2756938421509108, "grad_norm": 3.892265558242798, "learning_rate": 1.1537066589293487e-05, "loss": 0.0268, "step": 11767 }, { "epoch": 1.2758022549869905, "grad_norm": 2.581389904022217, "learning_rate": 1.1536341215726099e-05, "loss": 0.0334, "step": 11768 }, { "epoch": 1.2759106678230703, "grad_norm": 2.889894485473633, "learning_rate": 1.1535615842158711e-05, "loss": 0.0149, "step": 11769 }, { "epoch": 1.27601908065915, "grad_norm": 1.7559860944747925, "learning_rate": 1.1534890468591326e-05, "loss": 0.0142, "step": 11770 }, { "epoch": 1.2761274934952298, "grad_norm": 2.4838619232177734, "learning_rate": 1.1534165095023939e-05, "loss": 0.0265, "step": 11771 }, { "epoch": 1.2762359063313096, "grad_norm": 6.088444709777832, "learning_rate": 1.1533439721456552e-05, "loss": 0.0121, "step": 11772 }, { "epoch": 1.2763443191673893, "grad_norm": 0.6731381416320801, "learning_rate": 1.1532714347889165e-05, "loss": 0.0057, "step": 11773 }, { "epoch": 1.2764527320034693, "grad_norm": 3.8708086013793945, "learning_rate": 1.1531988974321777e-05, "loss": 0.0346, "step": 11774 }, { "epoch": 1.276561144839549, "grad_norm": 4.718855857849121, "learning_rate": 1.153126360075439e-05, "loss": 0.0089, "step": 11775 }, { "epoch": 1.2766695576756288, "grad_norm": 6.410114765167236, "learning_rate": 1.1530538227187003e-05, "loss": 0.039, "step": 11776 }, { "epoch": 1.2767779705117086, "grad_norm": 0.9184079170227051, "learning_rate": 1.1529812853619615e-05, "loss": 0.0127, "step": 11777 }, { "epoch": 1.2768863833477884, "grad_norm": 0.653152585029602, "learning_rate": 1.1529087480052229e-05, "loss": 0.0134, "step": 11778 }, { "epoch": 1.2769947961838681, "grad_norm": 0.14091399312019348, "learning_rate": 1.152836210648484e-05, "loss": 0.0013, "step": 11779 }, { "epoch": 1.277103209019948, "grad_norm": 2.2092461585998535, "learning_rate": 1.1527636732917453e-05, "loss": 0.0494, "step": 11780 }, { "epoch": 1.2772116218560279, "grad_norm": 0.8447210192680359, "learning_rate": 1.1526911359350065e-05, "loss": 0.009, "step": 11781 }, { "epoch": 1.2773200346921074, "grad_norm": 2.3486857414245605, "learning_rate": 1.1526185985782679e-05, "loss": 0.0178, "step": 11782 }, { "epoch": 1.2774284475281874, "grad_norm": 1.321887493133545, "learning_rate": 1.1525460612215291e-05, "loss": 0.0111, "step": 11783 }, { "epoch": 1.2775368603642672, "grad_norm": 2.022571563720703, "learning_rate": 1.1524735238647903e-05, "loss": 0.0046, "step": 11784 }, { "epoch": 1.277645273200347, "grad_norm": 2.301635980606079, "learning_rate": 1.1524009865080517e-05, "loss": 0.0226, "step": 11785 }, { "epoch": 1.2777536860364267, "grad_norm": 3.104599952697754, "learning_rate": 1.152328449151313e-05, "loss": 0.0672, "step": 11786 }, { "epoch": 1.2778620988725065, "grad_norm": 2.990755558013916, "learning_rate": 1.1522559117945741e-05, "loss": 0.0125, "step": 11787 }, { "epoch": 1.2779705117085862, "grad_norm": 1.240925669670105, "learning_rate": 1.1521833744378357e-05, "loss": 0.0046, "step": 11788 }, { "epoch": 1.278078924544666, "grad_norm": 3.6523184776306152, "learning_rate": 1.1521108370810969e-05, "loss": 0.005, "step": 11789 }, { "epoch": 1.278187337380746, "grad_norm": 1.6642109155654907, "learning_rate": 1.1520382997243583e-05, "loss": 0.0128, "step": 11790 }, { "epoch": 1.2782957502168257, "grad_norm": 0.6839498281478882, "learning_rate": 1.1519657623676195e-05, "loss": 0.0219, "step": 11791 }, { "epoch": 1.2784041630529055, "grad_norm": 3.61942720413208, "learning_rate": 1.1518932250108807e-05, "loss": 0.0713, "step": 11792 }, { "epoch": 1.2785125758889853, "grad_norm": 0.5478349328041077, "learning_rate": 1.151820687654142e-05, "loss": 0.0172, "step": 11793 }, { "epoch": 1.278620988725065, "grad_norm": 1.2968333959579468, "learning_rate": 1.1517481502974033e-05, "loss": 0.0715, "step": 11794 }, { "epoch": 1.2787294015611448, "grad_norm": 3.67387318611145, "learning_rate": 1.1516756129406645e-05, "loss": 0.0132, "step": 11795 }, { "epoch": 1.2788378143972245, "grad_norm": 5.664700984954834, "learning_rate": 1.1516030755839257e-05, "loss": 0.021, "step": 11796 }, { "epoch": 1.2789462272333045, "grad_norm": 0.5157437920570374, "learning_rate": 1.1515305382271871e-05, "loss": 0.0028, "step": 11797 }, { "epoch": 1.2790546400693843, "grad_norm": 6.495853424072266, "learning_rate": 1.1514580008704483e-05, "loss": 0.033, "step": 11798 }, { "epoch": 1.279163052905464, "grad_norm": 2.5209131240844727, "learning_rate": 1.1513854635137096e-05, "loss": 0.0678, "step": 11799 }, { "epoch": 1.2792714657415438, "grad_norm": 4.631521224975586, "learning_rate": 1.151312926156971e-05, "loss": 0.0069, "step": 11800 }, { "epoch": 1.2793798785776236, "grad_norm": 2.5258896350860596, "learning_rate": 1.1512403888002322e-05, "loss": 0.0507, "step": 11801 }, { "epoch": 1.2794882914137033, "grad_norm": 1.9148154258728027, "learning_rate": 1.1511678514434934e-05, "loss": 0.037, "step": 11802 }, { "epoch": 1.2795967042497831, "grad_norm": 1.1531000137329102, "learning_rate": 1.1510953140867547e-05, "loss": 0.0109, "step": 11803 }, { "epoch": 1.279705117085863, "grad_norm": 2.171989679336548, "learning_rate": 1.151022776730016e-05, "loss": 0.0567, "step": 11804 }, { "epoch": 1.2798135299219426, "grad_norm": 1.3691009283065796, "learning_rate": 1.1509502393732775e-05, "loss": 0.0178, "step": 11805 }, { "epoch": 1.2799219427580226, "grad_norm": 2.2394533157348633, "learning_rate": 1.1508777020165387e-05, "loss": 0.0197, "step": 11806 }, { "epoch": 1.2800303555941024, "grad_norm": 0.22642071545124054, "learning_rate": 1.1508051646598e-05, "loss": 0.0017, "step": 11807 }, { "epoch": 1.2801387684301822, "grad_norm": 0.31284990906715393, "learning_rate": 1.1507326273030612e-05, "loss": 0.0081, "step": 11808 }, { "epoch": 1.280247181266262, "grad_norm": 1.0944654941558838, "learning_rate": 1.1506600899463225e-05, "loss": 0.0343, "step": 11809 }, { "epoch": 1.2803555941023417, "grad_norm": 2.2081785202026367, "learning_rate": 1.1505875525895838e-05, "loss": 0.0135, "step": 11810 }, { "epoch": 1.2804640069384214, "grad_norm": 7.773329257965088, "learning_rate": 1.150515015232845e-05, "loss": 0.0612, "step": 11811 }, { "epoch": 1.2805724197745012, "grad_norm": 0.3085293471813202, "learning_rate": 1.1504424778761064e-05, "loss": 0.0074, "step": 11812 }, { "epoch": 1.2806808326105812, "grad_norm": 0.8457627892494202, "learning_rate": 1.1503699405193676e-05, "loss": 0.0034, "step": 11813 }, { "epoch": 1.280789245446661, "grad_norm": 1.6470081806182861, "learning_rate": 1.1502974031626288e-05, "loss": 0.0168, "step": 11814 }, { "epoch": 1.2808976582827407, "grad_norm": 4.867800235748291, "learning_rate": 1.1502248658058902e-05, "loss": 0.0547, "step": 11815 }, { "epoch": 1.2810060711188205, "grad_norm": 2.4255106449127197, "learning_rate": 1.1501523284491514e-05, "loss": 0.0434, "step": 11816 }, { "epoch": 1.2811144839549002, "grad_norm": 0.5906890034675598, "learning_rate": 1.1500797910924126e-05, "loss": 0.0201, "step": 11817 }, { "epoch": 1.28122289679098, "grad_norm": 1.162699818611145, "learning_rate": 1.150007253735674e-05, "loss": 0.0234, "step": 11818 }, { "epoch": 1.2813313096270598, "grad_norm": 0.46818381547927856, "learning_rate": 1.1499347163789352e-05, "loss": 0.0044, "step": 11819 }, { "epoch": 1.2814397224631398, "grad_norm": 5.106358051300049, "learning_rate": 1.1498621790221964e-05, "loss": 0.0247, "step": 11820 }, { "epoch": 1.2815481352992193, "grad_norm": 0.10503140836954117, "learning_rate": 1.1497896416654576e-05, "loss": 0.0014, "step": 11821 }, { "epoch": 1.2816565481352993, "grad_norm": 0.3341064155101776, "learning_rate": 1.1497171043087192e-05, "loss": 0.0014, "step": 11822 }, { "epoch": 1.281764960971379, "grad_norm": 3.3514199256896973, "learning_rate": 1.1496445669519804e-05, "loss": 0.0196, "step": 11823 }, { "epoch": 1.2818733738074588, "grad_norm": 0.8220013380050659, "learning_rate": 1.1495720295952418e-05, "loss": 0.0048, "step": 11824 }, { "epoch": 1.2819817866435386, "grad_norm": 5.431563854217529, "learning_rate": 1.149499492238503e-05, "loss": 0.0787, "step": 11825 }, { "epoch": 1.2820901994796183, "grad_norm": 0.416015088558197, "learning_rate": 1.1494269548817642e-05, "loss": 0.0196, "step": 11826 }, { "epoch": 1.2821986123156983, "grad_norm": 0.5054962635040283, "learning_rate": 1.1493544175250256e-05, "loss": 0.003, "step": 11827 }, { "epoch": 1.2823070251517779, "grad_norm": 0.48391565680503845, "learning_rate": 1.1492818801682868e-05, "loss": 0.0035, "step": 11828 }, { "epoch": 1.2824154379878578, "grad_norm": 3.222538471221924, "learning_rate": 1.149209342811548e-05, "loss": 0.0181, "step": 11829 }, { "epoch": 1.2825238508239376, "grad_norm": 1.3290935754776, "learning_rate": 1.1491368054548094e-05, "loss": 0.0101, "step": 11830 }, { "epoch": 1.2826322636600174, "grad_norm": 3.046868324279785, "learning_rate": 1.1490642680980706e-05, "loss": 0.0283, "step": 11831 }, { "epoch": 1.2827406764960971, "grad_norm": 1.403693675994873, "learning_rate": 1.1489917307413318e-05, "loss": 0.0139, "step": 11832 }, { "epoch": 1.282849089332177, "grad_norm": 2.4544854164123535, "learning_rate": 1.148919193384593e-05, "loss": 0.0409, "step": 11833 }, { "epoch": 1.2829575021682567, "grad_norm": 2.8032150268554688, "learning_rate": 1.1488466560278544e-05, "loss": 0.0187, "step": 11834 }, { "epoch": 1.2830659150043364, "grad_norm": 3.307701349258423, "learning_rate": 1.1487741186711156e-05, "loss": 0.0394, "step": 11835 }, { "epoch": 1.2831743278404164, "grad_norm": 1.451332449913025, "learning_rate": 1.1487015813143769e-05, "loss": 0.0492, "step": 11836 }, { "epoch": 1.2832827406764962, "grad_norm": 3.0637896060943604, "learning_rate": 1.1486290439576382e-05, "loss": 0.0034, "step": 11837 }, { "epoch": 1.283391153512576, "grad_norm": 0.7254160642623901, "learning_rate": 1.1485565066008995e-05, "loss": 0.0062, "step": 11838 }, { "epoch": 1.2834995663486557, "grad_norm": 0.5056540369987488, "learning_rate": 1.148483969244161e-05, "loss": 0.0011, "step": 11839 }, { "epoch": 1.2836079791847355, "grad_norm": 1.8799614906311035, "learning_rate": 1.1484114318874222e-05, "loss": 0.0344, "step": 11840 }, { "epoch": 1.2837163920208152, "grad_norm": 1.9308682680130005, "learning_rate": 1.1483388945306834e-05, "loss": 0.028, "step": 11841 }, { "epoch": 1.283824804856895, "grad_norm": 0.4741562604904175, "learning_rate": 1.1482663571739448e-05, "loss": 0.0157, "step": 11842 }, { "epoch": 1.283933217692975, "grad_norm": 0.8082515001296997, "learning_rate": 1.148193819817206e-05, "loss": 0.0369, "step": 11843 }, { "epoch": 1.2840416305290545, "grad_norm": 1.3435801267623901, "learning_rate": 1.1481212824604672e-05, "loss": 0.0238, "step": 11844 }, { "epoch": 1.2841500433651345, "grad_norm": 1.8820717334747314, "learning_rate": 1.1480487451037285e-05, "loss": 0.0212, "step": 11845 }, { "epoch": 1.2842584562012143, "grad_norm": 2.323009729385376, "learning_rate": 1.1479762077469898e-05, "loss": 0.0735, "step": 11846 }, { "epoch": 1.284366869037294, "grad_norm": 3.3035695552825928, "learning_rate": 1.147903670390251e-05, "loss": 0.0296, "step": 11847 }, { "epoch": 1.2844752818733738, "grad_norm": 0.8190964460372925, "learning_rate": 1.1478311330335123e-05, "loss": 0.0116, "step": 11848 }, { "epoch": 1.2845836947094535, "grad_norm": 0.6546860933303833, "learning_rate": 1.1477585956767737e-05, "loss": 0.0044, "step": 11849 }, { "epoch": 1.2846921075455333, "grad_norm": 2.487347364425659, "learning_rate": 1.1476860583200349e-05, "loss": 0.017, "step": 11850 }, { "epoch": 1.284800520381613, "grad_norm": 1.1773099899291992, "learning_rate": 1.1476135209632961e-05, "loss": 0.0261, "step": 11851 }, { "epoch": 1.284908933217693, "grad_norm": 0.1142742857336998, "learning_rate": 1.1475409836065575e-05, "loss": 0.0011, "step": 11852 }, { "epoch": 1.2850173460537728, "grad_norm": 4.933173656463623, "learning_rate": 1.1474684462498187e-05, "loss": 0.0214, "step": 11853 }, { "epoch": 1.2851257588898526, "grad_norm": 3.317312002182007, "learning_rate": 1.1473959088930799e-05, "loss": 0.0264, "step": 11854 }, { "epoch": 1.2852341717259324, "grad_norm": 1.1147953271865845, "learning_rate": 1.1473233715363413e-05, "loss": 0.0046, "step": 11855 }, { "epoch": 1.2853425845620121, "grad_norm": 2.959604024887085, "learning_rate": 1.1472508341796025e-05, "loss": 0.0206, "step": 11856 }, { "epoch": 1.2854509973980919, "grad_norm": 2.737826347351074, "learning_rate": 1.147178296822864e-05, "loss": 0.0276, "step": 11857 }, { "epoch": 1.2855594102341716, "grad_norm": 0.7121634483337402, "learning_rate": 1.1471057594661253e-05, "loss": 0.0036, "step": 11858 }, { "epoch": 1.2856678230702516, "grad_norm": 0.05654435604810715, "learning_rate": 1.1470332221093865e-05, "loss": 0.0008, "step": 11859 }, { "epoch": 1.2857762359063314, "grad_norm": 0.44024422764778137, "learning_rate": 1.1469606847526477e-05, "loss": 0.0137, "step": 11860 }, { "epoch": 1.2858846487424112, "grad_norm": 10.446000099182129, "learning_rate": 1.146888147395909e-05, "loss": 0.0409, "step": 11861 }, { "epoch": 1.285993061578491, "grad_norm": 0.7456680536270142, "learning_rate": 1.1468156100391703e-05, "loss": 0.0055, "step": 11862 }, { "epoch": 1.2861014744145707, "grad_norm": 3.6644413471221924, "learning_rate": 1.1467430726824315e-05, "loss": 0.0198, "step": 11863 }, { "epoch": 1.2862098872506504, "grad_norm": 0.7511047720909119, "learning_rate": 1.1466705353256929e-05, "loss": 0.0175, "step": 11864 }, { "epoch": 1.2863183000867302, "grad_norm": 2.993720531463623, "learning_rate": 1.1465979979689541e-05, "loss": 0.0105, "step": 11865 }, { "epoch": 1.2864267129228102, "grad_norm": 7.515557289123535, "learning_rate": 1.1465254606122153e-05, "loss": 0.0171, "step": 11866 }, { "epoch": 1.2865351257588897, "grad_norm": 0.539653480052948, "learning_rate": 1.1464529232554767e-05, "loss": 0.0123, "step": 11867 }, { "epoch": 1.2866435385949697, "grad_norm": 1.5169848203659058, "learning_rate": 1.146380385898738e-05, "loss": 0.0232, "step": 11868 }, { "epoch": 1.2867519514310495, "grad_norm": 0.5265146493911743, "learning_rate": 1.1463078485419991e-05, "loss": 0.0043, "step": 11869 }, { "epoch": 1.2868603642671292, "grad_norm": 7.094411849975586, "learning_rate": 1.1462353111852605e-05, "loss": 0.0463, "step": 11870 }, { "epoch": 1.286968777103209, "grad_norm": 2.595686435699463, "learning_rate": 1.1461627738285217e-05, "loss": 0.0293, "step": 11871 }, { "epoch": 1.2870771899392888, "grad_norm": 1.7268491983413696, "learning_rate": 1.146090236471783e-05, "loss": 0.0273, "step": 11872 }, { "epoch": 1.2871856027753685, "grad_norm": 3.90789794921875, "learning_rate": 1.1460176991150442e-05, "loss": 0.0268, "step": 11873 }, { "epoch": 1.2872940156114483, "grad_norm": 0.838852047920227, "learning_rate": 1.1459451617583057e-05, "loss": 0.009, "step": 11874 }, { "epoch": 1.2874024284475283, "grad_norm": 0.409391313791275, "learning_rate": 1.145872624401567e-05, "loss": 0.0123, "step": 11875 }, { "epoch": 1.287510841283608, "grad_norm": 0.6337389945983887, "learning_rate": 1.1458000870448283e-05, "loss": 0.0239, "step": 11876 }, { "epoch": 1.2876192541196878, "grad_norm": 0.08829737454652786, "learning_rate": 1.1457275496880895e-05, "loss": 0.0006, "step": 11877 }, { "epoch": 1.2877276669557676, "grad_norm": 8.819745063781738, "learning_rate": 1.1456550123313507e-05, "loss": 0.0295, "step": 11878 }, { "epoch": 1.2878360797918473, "grad_norm": 0.25015461444854736, "learning_rate": 1.1455824749746121e-05, "loss": 0.0016, "step": 11879 }, { "epoch": 1.287944492627927, "grad_norm": 1.143637776374817, "learning_rate": 1.1455099376178733e-05, "loss": 0.006, "step": 11880 }, { "epoch": 1.2880529054640069, "grad_norm": 3.948780059814453, "learning_rate": 1.1454374002611346e-05, "loss": 0.0113, "step": 11881 }, { "epoch": 1.2881613183000868, "grad_norm": 5.3135175704956055, "learning_rate": 1.145364862904396e-05, "loss": 0.0133, "step": 11882 }, { "epoch": 1.2882697311361664, "grad_norm": 2.809990406036377, "learning_rate": 1.1452923255476571e-05, "loss": 0.0525, "step": 11883 }, { "epoch": 1.2883781439722464, "grad_norm": 0.43392395973205566, "learning_rate": 1.1452197881909184e-05, "loss": 0.0262, "step": 11884 }, { "epoch": 1.2884865568083261, "grad_norm": 5.377943992614746, "learning_rate": 1.1451472508341796e-05, "loss": 0.0322, "step": 11885 }, { "epoch": 1.288594969644406, "grad_norm": 11.167054176330566, "learning_rate": 1.145074713477441e-05, "loss": 0.0228, "step": 11886 }, { "epoch": 1.2887033824804857, "grad_norm": 4.909409523010254, "learning_rate": 1.1450021761207022e-05, "loss": 0.09, "step": 11887 }, { "epoch": 1.2888117953165654, "grad_norm": 0.9559733271598816, "learning_rate": 1.1449296387639634e-05, "loss": 0.0195, "step": 11888 }, { "epoch": 1.2889202081526454, "grad_norm": 2.2730486392974854, "learning_rate": 1.1448571014072248e-05, "loss": 0.0715, "step": 11889 }, { "epoch": 1.289028620988725, "grad_norm": 3.9811267852783203, "learning_rate": 1.144784564050486e-05, "loss": 0.016, "step": 11890 }, { "epoch": 1.289137033824805, "grad_norm": 11.131088256835938, "learning_rate": 1.1447120266937475e-05, "loss": 0.0462, "step": 11891 }, { "epoch": 1.2892454466608847, "grad_norm": 4.043120861053467, "learning_rate": 1.1446394893370088e-05, "loss": 0.0649, "step": 11892 }, { "epoch": 1.2893538594969645, "grad_norm": 1.8818881511688232, "learning_rate": 1.14456695198027e-05, "loss": 0.018, "step": 11893 }, { "epoch": 1.2894622723330442, "grad_norm": 1.230560064315796, "learning_rate": 1.1444944146235314e-05, "loss": 0.0175, "step": 11894 }, { "epoch": 1.289570685169124, "grad_norm": 3.1202163696289062, "learning_rate": 1.1444218772667926e-05, "loss": 0.0565, "step": 11895 }, { "epoch": 1.2896790980052038, "grad_norm": 8.453857421875, "learning_rate": 1.1443493399100538e-05, "loss": 0.0125, "step": 11896 }, { "epoch": 1.2897875108412835, "grad_norm": 0.35637953877449036, "learning_rate": 1.144276802553315e-05, "loss": 0.0102, "step": 11897 }, { "epoch": 1.2898959236773635, "grad_norm": 3.6194281578063965, "learning_rate": 1.1442042651965764e-05, "loss": 0.0277, "step": 11898 }, { "epoch": 1.2900043365134433, "grad_norm": 6.364652633666992, "learning_rate": 1.1441317278398376e-05, "loss": 0.0413, "step": 11899 }, { "epoch": 1.290112749349523, "grad_norm": 4.148703575134277, "learning_rate": 1.1440591904830988e-05, "loss": 0.0709, "step": 11900 }, { "epoch": 1.2902211621856028, "grad_norm": 0.290427029132843, "learning_rate": 1.1439866531263602e-05, "loss": 0.0064, "step": 11901 }, { "epoch": 1.2903295750216826, "grad_norm": 1.5255012512207031, "learning_rate": 1.1439141157696214e-05, "loss": 0.0061, "step": 11902 }, { "epoch": 1.2904379878577623, "grad_norm": 0.7341137528419495, "learning_rate": 1.1438415784128826e-05, "loss": 0.003, "step": 11903 }, { "epoch": 1.290546400693842, "grad_norm": 8.179959297180176, "learning_rate": 1.143769041056144e-05, "loss": 0.0254, "step": 11904 }, { "epoch": 1.290654813529922, "grad_norm": 8.646160125732422, "learning_rate": 1.1436965036994052e-05, "loss": 0.0772, "step": 11905 }, { "epoch": 1.2907632263660016, "grad_norm": 3.7257511615753174, "learning_rate": 1.1436239663426664e-05, "loss": 0.0549, "step": 11906 }, { "epoch": 1.2908716392020816, "grad_norm": 2.6217939853668213, "learning_rate": 1.1435514289859278e-05, "loss": 0.0225, "step": 11907 }, { "epoch": 1.2909800520381614, "grad_norm": 0.5934320688247681, "learning_rate": 1.1434788916291892e-05, "loss": 0.0085, "step": 11908 }, { "epoch": 1.2910884648742411, "grad_norm": 1.339996337890625, "learning_rate": 1.1434063542724506e-05, "loss": 0.0358, "step": 11909 }, { "epoch": 1.2911968777103209, "grad_norm": 3.99906063079834, "learning_rate": 1.1433338169157118e-05, "loss": 0.0106, "step": 11910 }, { "epoch": 1.2913052905464006, "grad_norm": 2.720050811767578, "learning_rate": 1.143261279558973e-05, "loss": 0.0813, "step": 11911 }, { "epoch": 1.2914137033824804, "grad_norm": 0.13726241886615753, "learning_rate": 1.1431887422022342e-05, "loss": 0.0009, "step": 11912 }, { "epoch": 1.2915221162185602, "grad_norm": 0.8780321478843689, "learning_rate": 1.1431162048454956e-05, "loss": 0.0217, "step": 11913 }, { "epoch": 1.2916305290546402, "grad_norm": 3.3145899772644043, "learning_rate": 1.1430436674887568e-05, "loss": 0.0243, "step": 11914 }, { "epoch": 1.29173894189072, "grad_norm": 0.4390909969806671, "learning_rate": 1.142971130132018e-05, "loss": 0.009, "step": 11915 }, { "epoch": 1.2918473547267997, "grad_norm": 0.9241986870765686, "learning_rate": 1.1428985927752794e-05, "loss": 0.015, "step": 11916 }, { "epoch": 1.2919557675628794, "grad_norm": 1.2990293502807617, "learning_rate": 1.1428260554185406e-05, "loss": 0.021, "step": 11917 }, { "epoch": 1.2920641803989592, "grad_norm": 5.016598701477051, "learning_rate": 1.1427535180618019e-05, "loss": 0.0622, "step": 11918 }, { "epoch": 1.292172593235039, "grad_norm": 2.876492738723755, "learning_rate": 1.1426809807050632e-05, "loss": 0.0121, "step": 11919 }, { "epoch": 1.2922810060711187, "grad_norm": 0.37981685996055603, "learning_rate": 1.1426084433483245e-05, "loss": 0.0023, "step": 11920 }, { "epoch": 1.2923894189071987, "grad_norm": 9.677388191223145, "learning_rate": 1.1425359059915857e-05, "loss": 0.0236, "step": 11921 }, { "epoch": 1.2924978317432785, "grad_norm": 0.718985378742218, "learning_rate": 1.142463368634847e-05, "loss": 0.0024, "step": 11922 }, { "epoch": 1.2926062445793582, "grad_norm": 3.8087379932403564, "learning_rate": 1.1423908312781083e-05, "loss": 0.0334, "step": 11923 }, { "epoch": 1.292714657415438, "grad_norm": 2.1228606700897217, "learning_rate": 1.1423182939213695e-05, "loss": 0.0179, "step": 11924 }, { "epoch": 1.2928230702515178, "grad_norm": 1.580731987953186, "learning_rate": 1.1422457565646307e-05, "loss": 0.0121, "step": 11925 }, { "epoch": 1.2929314830875975, "grad_norm": 1.382831335067749, "learning_rate": 1.1421732192078922e-05, "loss": 0.011, "step": 11926 }, { "epoch": 1.2930398959236773, "grad_norm": 3.160372734069824, "learning_rate": 1.1421006818511535e-05, "loss": 0.0524, "step": 11927 }, { "epoch": 1.2931483087597573, "grad_norm": 1.1441149711608887, "learning_rate": 1.1420281444944148e-05, "loss": 0.0208, "step": 11928 }, { "epoch": 1.2932567215958368, "grad_norm": 2.222565174102783, "learning_rate": 1.141955607137676e-05, "loss": 0.0607, "step": 11929 }, { "epoch": 1.2933651344319168, "grad_norm": 6.670602798461914, "learning_rate": 1.1418830697809373e-05, "loss": 0.0283, "step": 11930 }, { "epoch": 1.2934735472679966, "grad_norm": 7.717975616455078, "learning_rate": 1.1418105324241987e-05, "loss": 0.0307, "step": 11931 }, { "epoch": 1.2935819601040763, "grad_norm": 8.399221420288086, "learning_rate": 1.1417379950674599e-05, "loss": 0.0292, "step": 11932 }, { "epoch": 1.293690372940156, "grad_norm": 1.1768406629562378, "learning_rate": 1.141665457710721e-05, "loss": 0.0358, "step": 11933 }, { "epoch": 1.2937987857762359, "grad_norm": 5.107679843902588, "learning_rate": 1.1415929203539825e-05, "loss": 0.0319, "step": 11934 }, { "epoch": 1.2939071986123156, "grad_norm": 1.6808550357818604, "learning_rate": 1.1415203829972437e-05, "loss": 0.0285, "step": 11935 }, { "epoch": 1.2940156114483954, "grad_norm": 0.39204278588294983, "learning_rate": 1.1414478456405049e-05, "loss": 0.0132, "step": 11936 }, { "epoch": 1.2941240242844754, "grad_norm": 5.320147514343262, "learning_rate": 1.1413753082837661e-05, "loss": 0.0659, "step": 11937 }, { "epoch": 1.2942324371205551, "grad_norm": 6.955322265625, "learning_rate": 1.1413027709270275e-05, "loss": 0.0454, "step": 11938 }, { "epoch": 1.294340849956635, "grad_norm": 5.714385032653809, "learning_rate": 1.1412302335702887e-05, "loss": 0.0638, "step": 11939 }, { "epoch": 1.2944492627927147, "grad_norm": 0.349884033203125, "learning_rate": 1.14115769621355e-05, "loss": 0.0035, "step": 11940 }, { "epoch": 1.2945576756287944, "grad_norm": 5.694552421569824, "learning_rate": 1.1410851588568113e-05, "loss": 0.048, "step": 11941 }, { "epoch": 1.2946660884648742, "grad_norm": 0.48735564947128296, "learning_rate": 1.1410126215000725e-05, "loss": 0.0114, "step": 11942 }, { "epoch": 1.294774501300954, "grad_norm": 1.8150473833084106, "learning_rate": 1.140940084143334e-05, "loss": 0.0146, "step": 11943 }, { "epoch": 1.294882914137034, "grad_norm": 2.853611469268799, "learning_rate": 1.1408675467865953e-05, "loss": 0.0139, "step": 11944 }, { "epoch": 1.2949913269731137, "grad_norm": 3.1592588424682617, "learning_rate": 1.1407950094298565e-05, "loss": 0.0379, "step": 11945 }, { "epoch": 1.2950997398091935, "grad_norm": 1.2610211372375488, "learning_rate": 1.1407224720731179e-05, "loss": 0.0134, "step": 11946 }, { "epoch": 1.2952081526452732, "grad_norm": 6.483944416046143, "learning_rate": 1.1406499347163791e-05, "loss": 0.0743, "step": 11947 }, { "epoch": 1.295316565481353, "grad_norm": 0.8167458176612854, "learning_rate": 1.1405773973596403e-05, "loss": 0.0061, "step": 11948 }, { "epoch": 1.2954249783174328, "grad_norm": 1.9286071062088013, "learning_rate": 1.1405048600029017e-05, "loss": 0.0227, "step": 11949 }, { "epoch": 1.2955333911535125, "grad_norm": 2.600191831588745, "learning_rate": 1.1404323226461629e-05, "loss": 0.0574, "step": 11950 }, { "epoch": 1.2956418039895925, "grad_norm": 2.0644195079803467, "learning_rate": 1.1403597852894241e-05, "loss": 0.0111, "step": 11951 }, { "epoch": 1.295750216825672, "grad_norm": 4.5502777099609375, "learning_rate": 1.1402872479326853e-05, "loss": 0.028, "step": 11952 }, { "epoch": 1.295858629661752, "grad_norm": 0.7355906367301941, "learning_rate": 1.1402147105759467e-05, "loss": 0.0049, "step": 11953 }, { "epoch": 1.2959670424978318, "grad_norm": 3.9565882682800293, "learning_rate": 1.140142173219208e-05, "loss": 0.0404, "step": 11954 }, { "epoch": 1.2960754553339116, "grad_norm": 0.14663565158843994, "learning_rate": 1.1400696358624692e-05, "loss": 0.0015, "step": 11955 }, { "epoch": 1.2961838681699913, "grad_norm": 6.2519001960754395, "learning_rate": 1.1399970985057305e-05, "loss": 0.0354, "step": 11956 }, { "epoch": 1.296292281006071, "grad_norm": 0.6606003642082214, "learning_rate": 1.1399245611489918e-05, "loss": 0.0207, "step": 11957 }, { "epoch": 1.2964006938421508, "grad_norm": 1.7365790605545044, "learning_rate": 1.139852023792253e-05, "loss": 0.0257, "step": 11958 }, { "epoch": 1.2965091066782306, "grad_norm": 1.9667041301727295, "learning_rate": 1.1397794864355144e-05, "loss": 0.0113, "step": 11959 }, { "epoch": 1.2966175195143106, "grad_norm": 0.421368271112442, "learning_rate": 1.1397069490787757e-05, "loss": 0.0127, "step": 11960 }, { "epoch": 1.2967259323503904, "grad_norm": 5.217456817626953, "learning_rate": 1.1396344117220371e-05, "loss": 0.0329, "step": 11961 }, { "epoch": 1.2968343451864701, "grad_norm": 2.864820718765259, "learning_rate": 1.1395618743652983e-05, "loss": 0.0388, "step": 11962 }, { "epoch": 1.2969427580225499, "grad_norm": 2.045578956604004, "learning_rate": 1.1394893370085595e-05, "loss": 0.0355, "step": 11963 }, { "epoch": 1.2970511708586296, "grad_norm": 3.480156183242798, "learning_rate": 1.1394167996518208e-05, "loss": 0.0498, "step": 11964 }, { "epoch": 1.2971595836947094, "grad_norm": 1.3424603939056396, "learning_rate": 1.1393442622950821e-05, "loss": 0.0195, "step": 11965 }, { "epoch": 1.2972679965307892, "grad_norm": 1.2890592813491821, "learning_rate": 1.1392717249383434e-05, "loss": 0.0285, "step": 11966 }, { "epoch": 1.2973764093668692, "grad_norm": 4.164579391479492, "learning_rate": 1.1391991875816046e-05, "loss": 0.0504, "step": 11967 }, { "epoch": 1.2974848222029487, "grad_norm": 1.5619988441467285, "learning_rate": 1.139126650224866e-05, "loss": 0.0119, "step": 11968 }, { "epoch": 1.2975932350390287, "grad_norm": 1.9611949920654297, "learning_rate": 1.1390541128681272e-05, "loss": 0.0124, "step": 11969 }, { "epoch": 1.2977016478751084, "grad_norm": 2.863109827041626, "learning_rate": 1.1389815755113884e-05, "loss": 0.0218, "step": 11970 }, { "epoch": 1.2978100607111882, "grad_norm": 6.350564479827881, "learning_rate": 1.1389090381546498e-05, "loss": 0.0475, "step": 11971 }, { "epoch": 1.297918473547268, "grad_norm": 1.6601922512054443, "learning_rate": 1.138836500797911e-05, "loss": 0.0276, "step": 11972 }, { "epoch": 1.2980268863833477, "grad_norm": 3.255676507949829, "learning_rate": 1.1387639634411722e-05, "loss": 0.0407, "step": 11973 }, { "epoch": 1.2981352992194277, "grad_norm": 4.702002048492432, "learning_rate": 1.1386914260844336e-05, "loss": 0.0294, "step": 11974 }, { "epoch": 1.2982437120555073, "grad_norm": 2.4619600772857666, "learning_rate": 1.1386188887276948e-05, "loss": 0.0319, "step": 11975 }, { "epoch": 1.2983521248915872, "grad_norm": 6.239012718200684, "learning_rate": 1.138546351370956e-05, "loss": 0.0488, "step": 11976 }, { "epoch": 1.298460537727667, "grad_norm": 2.329146146774292, "learning_rate": 1.1384738140142172e-05, "loss": 0.028, "step": 11977 }, { "epoch": 1.2985689505637468, "grad_norm": 4.4594407081604, "learning_rate": 1.1384012766574788e-05, "loss": 0.018, "step": 11978 }, { "epoch": 1.2986773633998265, "grad_norm": 0.4687112867832184, "learning_rate": 1.13832873930074e-05, "loss": 0.0041, "step": 11979 }, { "epoch": 1.2987857762359063, "grad_norm": 0.22966592013835907, "learning_rate": 1.1382562019440014e-05, "loss": 0.0024, "step": 11980 }, { "epoch": 1.298894189071986, "grad_norm": 1.1221386194229126, "learning_rate": 1.1381836645872626e-05, "loss": 0.0125, "step": 11981 }, { "epoch": 1.2990026019080658, "grad_norm": 1.4894838333129883, "learning_rate": 1.1381111272305238e-05, "loss": 0.0256, "step": 11982 }, { "epoch": 1.2991110147441458, "grad_norm": 1.1853667497634888, "learning_rate": 1.1380385898737852e-05, "loss": 0.0089, "step": 11983 }, { "epoch": 1.2992194275802256, "grad_norm": 8.961874961853027, "learning_rate": 1.1379660525170464e-05, "loss": 0.0287, "step": 11984 }, { "epoch": 1.2993278404163053, "grad_norm": 2.556180000305176, "learning_rate": 1.1378935151603076e-05, "loss": 0.0277, "step": 11985 }, { "epoch": 1.299436253252385, "grad_norm": 2.9724133014678955, "learning_rate": 1.137820977803569e-05, "loss": 0.0274, "step": 11986 }, { "epoch": 1.2995446660884649, "grad_norm": 2.1988027095794678, "learning_rate": 1.1377484404468302e-05, "loss": 0.0148, "step": 11987 }, { "epoch": 1.2996530789245446, "grad_norm": 2.668973207473755, "learning_rate": 1.1376759030900914e-05, "loss": 0.0381, "step": 11988 }, { "epoch": 1.2997614917606244, "grad_norm": 1.0934367179870605, "learning_rate": 1.1376033657333526e-05, "loss": 0.0152, "step": 11989 }, { "epoch": 1.2998699045967044, "grad_norm": 1.4701708555221558, "learning_rate": 1.137530828376614e-05, "loss": 0.0515, "step": 11990 }, { "epoch": 1.299978317432784, "grad_norm": 0.6817007660865784, "learning_rate": 1.1374582910198752e-05, "loss": 0.0083, "step": 11991 }, { "epoch": 1.300086730268864, "grad_norm": 2.708559513092041, "learning_rate": 1.1373857536631365e-05, "loss": 0.0164, "step": 11992 }, { "epoch": 1.3001951431049437, "grad_norm": 0.5385591387748718, "learning_rate": 1.1373132163063978e-05, "loss": 0.0045, "step": 11993 }, { "epoch": 1.3003035559410234, "grad_norm": 1.6411206722259521, "learning_rate": 1.137240678949659e-05, "loss": 0.0623, "step": 11994 }, { "epoch": 1.3004119687771032, "grad_norm": 4.525465965270996, "learning_rate": 1.1371681415929206e-05, "loss": 0.041, "step": 11995 }, { "epoch": 1.300520381613183, "grad_norm": 3.8541476726531982, "learning_rate": 1.1370956042361818e-05, "loss": 0.0233, "step": 11996 }, { "epoch": 1.3006287944492627, "grad_norm": 0.12119737267494202, "learning_rate": 1.137023066879443e-05, "loss": 0.001, "step": 11997 }, { "epoch": 1.3007372072853425, "grad_norm": 2.304417610168457, "learning_rate": 1.1369505295227044e-05, "loss": 0.0144, "step": 11998 }, { "epoch": 1.3008456201214225, "grad_norm": 1.0622278451919556, "learning_rate": 1.1368779921659656e-05, "loss": 0.0096, "step": 11999 }, { "epoch": 1.3009540329575022, "grad_norm": 0.8062612414360046, "learning_rate": 1.1368054548092268e-05, "loss": 0.0089, "step": 12000 }, { "epoch": 1.301062445793582, "grad_norm": 2.352654218673706, "learning_rate": 1.1367329174524882e-05, "loss": 0.0138, "step": 12001 }, { "epoch": 1.3011708586296618, "grad_norm": 5.568779945373535, "learning_rate": 1.1366603800957494e-05, "loss": 0.0423, "step": 12002 }, { "epoch": 1.3012792714657415, "grad_norm": 2.560126781463623, "learning_rate": 1.1365878427390107e-05, "loss": 0.0293, "step": 12003 }, { "epoch": 1.3013876843018213, "grad_norm": 0.49132946133613586, "learning_rate": 1.1365153053822719e-05, "loss": 0.0046, "step": 12004 }, { "epoch": 1.301496097137901, "grad_norm": 1.4667900800704956, "learning_rate": 1.1364427680255333e-05, "loss": 0.0197, "step": 12005 }, { "epoch": 1.301604509973981, "grad_norm": 1.6059297323226929, "learning_rate": 1.1363702306687945e-05, "loss": 0.0255, "step": 12006 }, { "epoch": 1.3017129228100608, "grad_norm": 2.4145727157592773, "learning_rate": 1.1362976933120557e-05, "loss": 0.0482, "step": 12007 }, { "epoch": 1.3018213356461406, "grad_norm": 5.0270771980285645, "learning_rate": 1.136225155955317e-05, "loss": 0.0274, "step": 12008 }, { "epoch": 1.3019297484822203, "grad_norm": 4.887167453765869, "learning_rate": 1.1361526185985783e-05, "loss": 0.0496, "step": 12009 }, { "epoch": 1.3020381613183, "grad_norm": 1.2170222997665405, "learning_rate": 1.1360800812418395e-05, "loss": 0.0252, "step": 12010 }, { "epoch": 1.3021465741543798, "grad_norm": 2.0237393379211426, "learning_rate": 1.1360075438851009e-05, "loss": 0.016, "step": 12011 }, { "epoch": 1.3022549869904596, "grad_norm": 2.7726986408233643, "learning_rate": 1.1359350065283623e-05, "loss": 0.0515, "step": 12012 }, { "epoch": 1.3023633998265396, "grad_norm": 2.357834815979004, "learning_rate": 1.1358624691716237e-05, "loss": 0.0207, "step": 12013 }, { "epoch": 1.3024718126626191, "grad_norm": 1.3096758127212524, "learning_rate": 1.1357899318148849e-05, "loss": 0.0131, "step": 12014 }, { "epoch": 1.3025802254986991, "grad_norm": 1.3651412725448608, "learning_rate": 1.135717394458146e-05, "loss": 0.0141, "step": 12015 }, { "epoch": 1.3026886383347789, "grad_norm": 4.551332473754883, "learning_rate": 1.1356448571014073e-05, "loss": 0.0224, "step": 12016 }, { "epoch": 1.3027970511708586, "grad_norm": 2.1540398597717285, "learning_rate": 1.1355723197446687e-05, "loss": 0.0785, "step": 12017 }, { "epoch": 1.3029054640069384, "grad_norm": 1.3532488346099854, "learning_rate": 1.1354997823879299e-05, "loss": 0.0362, "step": 12018 }, { "epoch": 1.3030138768430182, "grad_norm": 11.03811264038086, "learning_rate": 1.1354272450311911e-05, "loss": 0.0426, "step": 12019 }, { "epoch": 1.303122289679098, "grad_norm": 2.481322765350342, "learning_rate": 1.1353547076744525e-05, "loss": 0.0141, "step": 12020 }, { "epoch": 1.3032307025151777, "grad_norm": 0.9341323375701904, "learning_rate": 1.1352821703177137e-05, "loss": 0.0175, "step": 12021 }, { "epoch": 1.3033391153512577, "grad_norm": 0.47415047883987427, "learning_rate": 1.135209632960975e-05, "loss": 0.004, "step": 12022 }, { "epoch": 1.3034475281873374, "grad_norm": 2.552260398864746, "learning_rate": 1.1351370956042363e-05, "loss": 0.0503, "step": 12023 }, { "epoch": 1.3035559410234172, "grad_norm": 0.4503195881843567, "learning_rate": 1.1350645582474975e-05, "loss": 0.0129, "step": 12024 }, { "epoch": 1.303664353859497, "grad_norm": 3.078672170639038, "learning_rate": 1.1349920208907587e-05, "loss": 0.1104, "step": 12025 }, { "epoch": 1.3037727666955767, "grad_norm": 0.40792566537857056, "learning_rate": 1.1349194835340201e-05, "loss": 0.0033, "step": 12026 }, { "epoch": 1.3038811795316565, "grad_norm": 0.9431314468383789, "learning_rate": 1.1348469461772813e-05, "loss": 0.0127, "step": 12027 }, { "epoch": 1.3039895923677363, "grad_norm": 0.8928254246711731, "learning_rate": 1.1347744088205425e-05, "loss": 0.007, "step": 12028 }, { "epoch": 1.3040980052038162, "grad_norm": 0.6648698449134827, "learning_rate": 1.1347018714638041e-05, "loss": 0.038, "step": 12029 }, { "epoch": 1.3042064180398958, "grad_norm": 2.531553268432617, "learning_rate": 1.1346293341070653e-05, "loss": 0.0738, "step": 12030 }, { "epoch": 1.3043148308759758, "grad_norm": 2.5343527793884277, "learning_rate": 1.1345567967503265e-05, "loss": 0.0117, "step": 12031 }, { "epoch": 1.3044232437120555, "grad_norm": 1.5855624675750732, "learning_rate": 1.1344842593935879e-05, "loss": 0.0295, "step": 12032 }, { "epoch": 1.3045316565481353, "grad_norm": 1.9574483633041382, "learning_rate": 1.1344117220368491e-05, "loss": 0.0234, "step": 12033 }, { "epoch": 1.304640069384215, "grad_norm": 2.9595143795013428, "learning_rate": 1.1343391846801103e-05, "loss": 0.0616, "step": 12034 }, { "epoch": 1.3047484822202948, "grad_norm": 3.489567756652832, "learning_rate": 1.1342666473233717e-05, "loss": 0.0275, "step": 12035 }, { "epoch": 1.3048568950563748, "grad_norm": 3.183800458908081, "learning_rate": 1.134194109966633e-05, "loss": 0.0345, "step": 12036 }, { "epoch": 1.3049653078924544, "grad_norm": 0.1599392294883728, "learning_rate": 1.1341215726098942e-05, "loss": 0.0021, "step": 12037 }, { "epoch": 1.3050737207285343, "grad_norm": 1.5053597688674927, "learning_rate": 1.1340490352531555e-05, "loss": 0.0231, "step": 12038 }, { "epoch": 1.305182133564614, "grad_norm": 3.379140615463257, "learning_rate": 1.1339764978964167e-05, "loss": 0.0242, "step": 12039 }, { "epoch": 1.3052905464006939, "grad_norm": 2.962252378463745, "learning_rate": 1.133903960539678e-05, "loss": 0.0183, "step": 12040 }, { "epoch": 1.3053989592367736, "grad_norm": 0.9513176679611206, "learning_rate": 1.1338314231829392e-05, "loss": 0.0075, "step": 12041 }, { "epoch": 1.3055073720728534, "grad_norm": 4.061813831329346, "learning_rate": 1.1337588858262006e-05, "loss": 0.0198, "step": 12042 }, { "epoch": 1.3056157849089332, "grad_norm": 1.195298433303833, "learning_rate": 1.1336863484694618e-05, "loss": 0.0171, "step": 12043 }, { "epoch": 1.305724197745013, "grad_norm": 2.7161097526550293, "learning_rate": 1.133613811112723e-05, "loss": 0.008, "step": 12044 }, { "epoch": 1.305832610581093, "grad_norm": 2.1499922275543213, "learning_rate": 1.1335412737559844e-05, "loss": 0.047, "step": 12045 }, { "epoch": 1.3059410234171727, "grad_norm": 2.1680686473846436, "learning_rate": 1.1334687363992456e-05, "loss": 0.0266, "step": 12046 }, { "epoch": 1.3060494362532524, "grad_norm": 1.6634753942489624, "learning_rate": 1.1333961990425071e-05, "loss": 0.0247, "step": 12047 }, { "epoch": 1.3061578490893322, "grad_norm": 0.028943980112671852, "learning_rate": 1.1333236616857684e-05, "loss": 0.0005, "step": 12048 }, { "epoch": 1.306266261925412, "grad_norm": 4.324159145355225, "learning_rate": 1.1332511243290296e-05, "loss": 0.0444, "step": 12049 }, { "epoch": 1.3063746747614917, "grad_norm": 1.7345566749572754, "learning_rate": 1.133178586972291e-05, "loss": 0.0273, "step": 12050 }, { "epoch": 1.3064830875975715, "grad_norm": 0.6100361943244934, "learning_rate": 1.1331060496155522e-05, "loss": 0.0253, "step": 12051 }, { "epoch": 1.3065915004336515, "grad_norm": 1.5103492736816406, "learning_rate": 1.1330335122588134e-05, "loss": 0.0329, "step": 12052 }, { "epoch": 1.306699913269731, "grad_norm": 0.5115548968315125, "learning_rate": 1.1329609749020748e-05, "loss": 0.0048, "step": 12053 }, { "epoch": 1.306808326105811, "grad_norm": 0.5352603197097778, "learning_rate": 1.132888437545336e-05, "loss": 0.0031, "step": 12054 }, { "epoch": 1.3069167389418908, "grad_norm": 0.11811567097902298, "learning_rate": 1.1328159001885972e-05, "loss": 0.0013, "step": 12055 }, { "epoch": 1.3070251517779705, "grad_norm": 0.33449673652648926, "learning_rate": 1.1327433628318584e-05, "loss": 0.0027, "step": 12056 }, { "epoch": 1.3071335646140503, "grad_norm": 2.9470505714416504, "learning_rate": 1.1326708254751198e-05, "loss": 0.0355, "step": 12057 }, { "epoch": 1.30724197745013, "grad_norm": 0.8178216218948364, "learning_rate": 1.132598288118381e-05, "loss": 0.0089, "step": 12058 }, { "epoch": 1.3073503902862098, "grad_norm": 2.1728315353393555, "learning_rate": 1.1325257507616422e-05, "loss": 0.0167, "step": 12059 }, { "epoch": 1.3074588031222896, "grad_norm": 2.8575832843780518, "learning_rate": 1.1324532134049036e-05, "loss": 0.0177, "step": 12060 }, { "epoch": 1.3075672159583696, "grad_norm": 4.9112043380737305, "learning_rate": 1.1323806760481648e-05, "loss": 0.0505, "step": 12061 }, { "epoch": 1.3076756287944493, "grad_norm": 3.048051595687866, "learning_rate": 1.132308138691426e-05, "loss": 0.0223, "step": 12062 }, { "epoch": 1.307784041630529, "grad_norm": 0.8217023611068726, "learning_rate": 1.1322356013346874e-05, "loss": 0.0046, "step": 12063 }, { "epoch": 1.3078924544666088, "grad_norm": 0.8960667252540588, "learning_rate": 1.1321630639779488e-05, "loss": 0.0099, "step": 12064 }, { "epoch": 1.3080008673026886, "grad_norm": 3.2532007694244385, "learning_rate": 1.1320905266212102e-05, "loss": 0.0187, "step": 12065 }, { "epoch": 1.3081092801387684, "grad_norm": 2.218454360961914, "learning_rate": 1.1320179892644714e-05, "loss": 0.012, "step": 12066 }, { "epoch": 1.3082176929748481, "grad_norm": 1.329828143119812, "learning_rate": 1.1319454519077326e-05, "loss": 0.0478, "step": 12067 }, { "epoch": 1.3083261058109281, "grad_norm": 5.887802600860596, "learning_rate": 1.1318729145509938e-05, "loss": 0.0324, "step": 12068 }, { "epoch": 1.3084345186470079, "grad_norm": 0.6855884790420532, "learning_rate": 1.1318003771942552e-05, "loss": 0.0126, "step": 12069 }, { "epoch": 1.3085429314830876, "grad_norm": 0.5090818405151367, "learning_rate": 1.1317278398375164e-05, "loss": 0.0124, "step": 12070 }, { "epoch": 1.3086513443191674, "grad_norm": 0.49622195959091187, "learning_rate": 1.1316553024807776e-05, "loss": 0.0067, "step": 12071 }, { "epoch": 1.3087597571552472, "grad_norm": 2.219012975692749, "learning_rate": 1.131582765124039e-05, "loss": 0.0417, "step": 12072 }, { "epoch": 1.308868169991327, "grad_norm": 0.7056358456611633, "learning_rate": 1.1315102277673002e-05, "loss": 0.0077, "step": 12073 }, { "epoch": 1.3089765828274067, "grad_norm": 1.900364637374878, "learning_rate": 1.1314376904105615e-05, "loss": 0.0213, "step": 12074 }, { "epoch": 1.3090849956634867, "grad_norm": 3.2385952472686768, "learning_rate": 1.1313651530538228e-05, "loss": 0.0194, "step": 12075 }, { "epoch": 1.3091934084995662, "grad_norm": 7.4166340827941895, "learning_rate": 1.131292615697084e-05, "loss": 0.0433, "step": 12076 }, { "epoch": 1.3093018213356462, "grad_norm": 0.18589074909687042, "learning_rate": 1.1312200783403453e-05, "loss": 0.0011, "step": 12077 }, { "epoch": 1.309410234171726, "grad_norm": 3.194828510284424, "learning_rate": 1.1311475409836066e-05, "loss": 0.0282, "step": 12078 }, { "epoch": 1.3095186470078057, "grad_norm": 0.23145481944084167, "learning_rate": 1.1310750036268679e-05, "loss": 0.0008, "step": 12079 }, { "epoch": 1.3096270598438855, "grad_norm": 4.177999019622803, "learning_rate": 1.131002466270129e-05, "loss": 0.0545, "step": 12080 }, { "epoch": 1.3097354726799653, "grad_norm": 2.5997564792633057, "learning_rate": 1.1309299289133906e-05, "loss": 0.0335, "step": 12081 }, { "epoch": 1.309843885516045, "grad_norm": 4.830793380737305, "learning_rate": 1.1308573915566518e-05, "loss": 0.0919, "step": 12082 }, { "epoch": 1.3099522983521248, "grad_norm": 0.2740618884563446, "learning_rate": 1.130784854199913e-05, "loss": 0.0015, "step": 12083 }, { "epoch": 1.3100607111882048, "grad_norm": 16.0936336517334, "learning_rate": 1.1307123168431744e-05, "loss": 0.0482, "step": 12084 }, { "epoch": 1.3101691240242845, "grad_norm": 3.286172389984131, "learning_rate": 1.1306397794864357e-05, "loss": 0.0439, "step": 12085 }, { "epoch": 1.3102775368603643, "grad_norm": 0.4426724910736084, "learning_rate": 1.1305672421296969e-05, "loss": 0.0216, "step": 12086 }, { "epoch": 1.310385949696444, "grad_norm": 8.222915649414062, "learning_rate": 1.1304947047729583e-05, "loss": 0.0596, "step": 12087 }, { "epoch": 1.3104943625325238, "grad_norm": 4.422583103179932, "learning_rate": 1.1304221674162195e-05, "loss": 0.0556, "step": 12088 }, { "epoch": 1.3106027753686036, "grad_norm": 2.562670946121216, "learning_rate": 1.1303496300594807e-05, "loss": 0.01, "step": 12089 }, { "epoch": 1.3107111882046834, "grad_norm": 2.6501660346984863, "learning_rate": 1.130277092702742e-05, "loss": 0.0278, "step": 12090 }, { "epoch": 1.3108196010407633, "grad_norm": 2.546755313873291, "learning_rate": 1.1302045553460033e-05, "loss": 0.0102, "step": 12091 }, { "epoch": 1.310928013876843, "grad_norm": 4.520971775054932, "learning_rate": 1.1301320179892645e-05, "loss": 0.0307, "step": 12092 }, { "epoch": 1.3110364267129229, "grad_norm": 0.6321756839752197, "learning_rate": 1.1300594806325257e-05, "loss": 0.0044, "step": 12093 }, { "epoch": 1.3111448395490026, "grad_norm": 3.878641128540039, "learning_rate": 1.1299869432757871e-05, "loss": 0.01, "step": 12094 }, { "epoch": 1.3112532523850824, "grad_norm": 0.15785622596740723, "learning_rate": 1.1299144059190483e-05, "loss": 0.0129, "step": 12095 }, { "epoch": 1.3113616652211622, "grad_norm": 2.51889705657959, "learning_rate": 1.1298418685623095e-05, "loss": 0.0367, "step": 12096 }, { "epoch": 1.311470078057242, "grad_norm": 4.490137100219727, "learning_rate": 1.1297693312055709e-05, "loss": 0.0071, "step": 12097 }, { "epoch": 1.311578490893322, "grad_norm": 3.286207914352417, "learning_rate": 1.1296967938488323e-05, "loss": 0.02, "step": 12098 }, { "epoch": 1.3116869037294014, "grad_norm": 2.7716712951660156, "learning_rate": 1.1296242564920937e-05, "loss": 0.027, "step": 12099 }, { "epoch": 1.3117953165654814, "grad_norm": 3.429687976837158, "learning_rate": 1.1295517191353549e-05, "loss": 0.0845, "step": 12100 }, { "epoch": 1.3119037294015612, "grad_norm": 0.3253709077835083, "learning_rate": 1.1294791817786161e-05, "loss": 0.024, "step": 12101 }, { "epoch": 1.312012142237641, "grad_norm": 2.223783016204834, "learning_rate": 1.1294066444218775e-05, "loss": 0.0257, "step": 12102 }, { "epoch": 1.3121205550737207, "grad_norm": 1.7906314134597778, "learning_rate": 1.1293341070651387e-05, "loss": 0.007, "step": 12103 }, { "epoch": 1.3122289679098005, "grad_norm": 1.8214640617370605, "learning_rate": 1.1292615697084e-05, "loss": 0.029, "step": 12104 }, { "epoch": 1.3123373807458802, "grad_norm": 3.192185163497925, "learning_rate": 1.1291890323516613e-05, "loss": 0.022, "step": 12105 }, { "epoch": 1.31244579358196, "grad_norm": 0.8999619483947754, "learning_rate": 1.1291164949949225e-05, "loss": 0.0175, "step": 12106 }, { "epoch": 1.31255420641804, "grad_norm": 0.21491074562072754, "learning_rate": 1.1290439576381837e-05, "loss": 0.0014, "step": 12107 }, { "epoch": 1.3126626192541198, "grad_norm": 1.3444874286651611, "learning_rate": 1.128971420281445e-05, "loss": 0.0251, "step": 12108 }, { "epoch": 1.3127710320901995, "grad_norm": 5.251614570617676, "learning_rate": 1.1288988829247063e-05, "loss": 0.0252, "step": 12109 }, { "epoch": 1.3128794449262793, "grad_norm": 0.4090370535850525, "learning_rate": 1.1288263455679675e-05, "loss": 0.004, "step": 12110 }, { "epoch": 1.312987857762359, "grad_norm": 3.023258924484253, "learning_rate": 1.1287538082112288e-05, "loss": 0.0214, "step": 12111 }, { "epoch": 1.3130962705984388, "grad_norm": 2.8677167892456055, "learning_rate": 1.1286812708544901e-05, "loss": 0.0481, "step": 12112 }, { "epoch": 1.3132046834345186, "grad_norm": 8.475367546081543, "learning_rate": 1.1286087334977514e-05, "loss": 0.0201, "step": 12113 }, { "epoch": 1.3133130962705986, "grad_norm": 1.0601214170455933, "learning_rate": 1.1285361961410126e-05, "loss": 0.0039, "step": 12114 }, { "epoch": 1.313421509106678, "grad_norm": 0.6024901866912842, "learning_rate": 1.128463658784274e-05, "loss": 0.0088, "step": 12115 }, { "epoch": 1.313529921942758, "grad_norm": 0.34689822793006897, "learning_rate": 1.1283911214275353e-05, "loss": 0.005, "step": 12116 }, { "epoch": 1.3136383347788378, "grad_norm": 1.9840508699417114, "learning_rate": 1.1283185840707967e-05, "loss": 0.0137, "step": 12117 }, { "epoch": 1.3137467476149176, "grad_norm": 0.190218985080719, "learning_rate": 1.128246046714058e-05, "loss": 0.002, "step": 12118 }, { "epoch": 1.3138551604509974, "grad_norm": 1.3841469287872314, "learning_rate": 1.1281735093573191e-05, "loss": 0.0133, "step": 12119 }, { "epoch": 1.3139635732870771, "grad_norm": 0.08999159187078476, "learning_rate": 1.1281009720005804e-05, "loss": 0.0006, "step": 12120 }, { "epoch": 1.314071986123157, "grad_norm": 0.9560346603393555, "learning_rate": 1.1280284346438417e-05, "loss": 0.0181, "step": 12121 }, { "epoch": 1.3141803989592367, "grad_norm": 2.9529528617858887, "learning_rate": 1.127955897287103e-05, "loss": 0.0361, "step": 12122 }, { "epoch": 1.3142888117953166, "grad_norm": 3.9457356929779053, "learning_rate": 1.1278833599303642e-05, "loss": 0.0854, "step": 12123 }, { "epoch": 1.3143972246313964, "grad_norm": 0.35037761926651, "learning_rate": 1.1278108225736256e-05, "loss": 0.0106, "step": 12124 }, { "epoch": 1.3145056374674762, "grad_norm": 3.6330041885375977, "learning_rate": 1.1277382852168868e-05, "loss": 0.0395, "step": 12125 }, { "epoch": 1.314614050303556, "grad_norm": 0.5387045741081238, "learning_rate": 1.127665747860148e-05, "loss": 0.0042, "step": 12126 }, { "epoch": 1.3147224631396357, "grad_norm": 0.4374832510948181, "learning_rate": 1.1275932105034094e-05, "loss": 0.011, "step": 12127 }, { "epoch": 1.3148308759757155, "grad_norm": 0.40846318006515503, "learning_rate": 1.1275206731466706e-05, "loss": 0.0073, "step": 12128 }, { "epoch": 1.3149392888117952, "grad_norm": 0.3310736119747162, "learning_rate": 1.1274481357899318e-05, "loss": 0.0077, "step": 12129 }, { "epoch": 1.3150477016478752, "grad_norm": 1.451669454574585, "learning_rate": 1.1273755984331932e-05, "loss": 0.0173, "step": 12130 }, { "epoch": 1.315156114483955, "grad_norm": 1.666538953781128, "learning_rate": 1.1273030610764544e-05, "loss": 0.034, "step": 12131 }, { "epoch": 1.3152645273200347, "grad_norm": 0.3485575020313263, "learning_rate": 1.1272305237197156e-05, "loss": 0.0015, "step": 12132 }, { "epoch": 1.3153729401561145, "grad_norm": 8.188716888427734, "learning_rate": 1.1271579863629772e-05, "loss": 0.0267, "step": 12133 }, { "epoch": 1.3154813529921943, "grad_norm": 4.478043079376221, "learning_rate": 1.1270854490062384e-05, "loss": 0.0436, "step": 12134 }, { "epoch": 1.315589765828274, "grad_norm": 0.31822240352630615, "learning_rate": 1.1270129116494996e-05, "loss": 0.0096, "step": 12135 }, { "epoch": 1.3156981786643538, "grad_norm": 7.153264045715332, "learning_rate": 1.126940374292761e-05, "loss": 0.0123, "step": 12136 }, { "epoch": 1.3158065915004338, "grad_norm": 0.6615201234817505, "learning_rate": 1.1268678369360222e-05, "loss": 0.0084, "step": 12137 }, { "epoch": 1.3159150043365133, "grad_norm": 0.044770944863557816, "learning_rate": 1.1267952995792834e-05, "loss": 0.0005, "step": 12138 }, { "epoch": 1.3160234171725933, "grad_norm": 1.613013744354248, "learning_rate": 1.1267227622225448e-05, "loss": 0.0381, "step": 12139 }, { "epoch": 1.316131830008673, "grad_norm": 3.2049176692962646, "learning_rate": 1.126650224865806e-05, "loss": 0.0358, "step": 12140 }, { "epoch": 1.3162402428447528, "grad_norm": 3.1305360794067383, "learning_rate": 1.1265776875090672e-05, "loss": 0.0533, "step": 12141 }, { "epoch": 1.3163486556808326, "grad_norm": 9.188389778137207, "learning_rate": 1.1265051501523286e-05, "loss": 0.0776, "step": 12142 }, { "epoch": 1.3164570685169124, "grad_norm": 2.8401477336883545, "learning_rate": 1.1264326127955898e-05, "loss": 0.0439, "step": 12143 }, { "epoch": 1.3165654813529921, "grad_norm": 1.4755356311798096, "learning_rate": 1.126360075438851e-05, "loss": 0.0167, "step": 12144 }, { "epoch": 1.3166738941890719, "grad_norm": 3.190364122390747, "learning_rate": 1.1262875380821124e-05, "loss": 0.0233, "step": 12145 }, { "epoch": 1.3167823070251519, "grad_norm": 2.967484712600708, "learning_rate": 1.1262150007253736e-05, "loss": 0.0443, "step": 12146 }, { "epoch": 1.3168907198612316, "grad_norm": 1.9461430311203003, "learning_rate": 1.1261424633686348e-05, "loss": 0.0206, "step": 12147 }, { "epoch": 1.3169991326973114, "grad_norm": 2.7809102535247803, "learning_rate": 1.126069926011896e-05, "loss": 0.0059, "step": 12148 }, { "epoch": 1.3171075455333912, "grad_norm": 0.45175808668136597, "learning_rate": 1.1259973886551574e-05, "loss": 0.0148, "step": 12149 }, { "epoch": 1.317215958369471, "grad_norm": 1.8147071599960327, "learning_rate": 1.1259248512984188e-05, "loss": 0.0203, "step": 12150 }, { "epoch": 1.3173243712055507, "grad_norm": 0.49388280510902405, "learning_rate": 1.1258523139416802e-05, "loss": 0.0061, "step": 12151 }, { "epoch": 1.3174327840416304, "grad_norm": 2.8892605304718018, "learning_rate": 1.1257797765849414e-05, "loss": 0.035, "step": 12152 }, { "epoch": 1.3175411968777104, "grad_norm": 0.38775166869163513, "learning_rate": 1.1257072392282026e-05, "loss": 0.0225, "step": 12153 }, { "epoch": 1.3176496097137902, "grad_norm": 4.653488636016846, "learning_rate": 1.125634701871464e-05, "loss": 0.0144, "step": 12154 }, { "epoch": 1.31775802254987, "grad_norm": 0.6934651732444763, "learning_rate": 1.1255621645147252e-05, "loss": 0.0033, "step": 12155 }, { "epoch": 1.3178664353859497, "grad_norm": 0.765185534954071, "learning_rate": 1.1254896271579865e-05, "loss": 0.0075, "step": 12156 }, { "epoch": 1.3179748482220295, "grad_norm": 0.35013407468795776, "learning_rate": 1.1254170898012478e-05, "loss": 0.001, "step": 12157 }, { "epoch": 1.3180832610581092, "grad_norm": 3.160342216491699, "learning_rate": 1.125344552444509e-05, "loss": 0.0682, "step": 12158 }, { "epoch": 1.318191673894189, "grad_norm": 1.313878059387207, "learning_rate": 1.1252720150877703e-05, "loss": 0.0077, "step": 12159 }, { "epoch": 1.318300086730269, "grad_norm": 3.147975206375122, "learning_rate": 1.1251994777310315e-05, "loss": 0.0194, "step": 12160 }, { "epoch": 1.3184084995663485, "grad_norm": 1.458349585533142, "learning_rate": 1.1251269403742929e-05, "loss": 0.0275, "step": 12161 }, { "epoch": 1.3185169124024285, "grad_norm": 4.098687648773193, "learning_rate": 1.125054403017554e-05, "loss": 0.0171, "step": 12162 }, { "epoch": 1.3186253252385083, "grad_norm": 0.7048590183258057, "learning_rate": 1.1249818656608153e-05, "loss": 0.0069, "step": 12163 }, { "epoch": 1.318733738074588, "grad_norm": 5.225002765655518, "learning_rate": 1.1249093283040767e-05, "loss": 0.0148, "step": 12164 }, { "epoch": 1.3188421509106678, "grad_norm": 0.43215978145599365, "learning_rate": 1.1248367909473379e-05, "loss": 0.0129, "step": 12165 }, { "epoch": 1.3189505637467476, "grad_norm": 7.261262893676758, "learning_rate": 1.1247642535905991e-05, "loss": 0.0356, "step": 12166 }, { "epoch": 1.3190589765828273, "grad_norm": 0.4690222442150116, "learning_rate": 1.1246917162338605e-05, "loss": 0.0067, "step": 12167 }, { "epoch": 1.319167389418907, "grad_norm": 1.006040096282959, "learning_rate": 1.1246191788771219e-05, "loss": 0.0125, "step": 12168 }, { "epoch": 1.319275802254987, "grad_norm": 2.6748769283294678, "learning_rate": 1.1245466415203833e-05, "loss": 0.047, "step": 12169 }, { "epoch": 1.3193842150910668, "grad_norm": 0.5400314331054688, "learning_rate": 1.1244741041636445e-05, "loss": 0.0151, "step": 12170 }, { "epoch": 1.3194926279271466, "grad_norm": 0.7638609409332275, "learning_rate": 1.1244015668069057e-05, "loss": 0.0191, "step": 12171 }, { "epoch": 1.3196010407632264, "grad_norm": 2.2841949462890625, "learning_rate": 1.1243290294501669e-05, "loss": 0.0992, "step": 12172 }, { "epoch": 1.3197094535993061, "grad_norm": 1.3171026706695557, "learning_rate": 1.1242564920934283e-05, "loss": 0.0171, "step": 12173 }, { "epoch": 1.319817866435386, "grad_norm": 0.16352629661560059, "learning_rate": 1.1241839547366895e-05, "loss": 0.0011, "step": 12174 }, { "epoch": 1.3199262792714657, "grad_norm": 1.0796438455581665, "learning_rate": 1.1241114173799507e-05, "loss": 0.0064, "step": 12175 }, { "epoch": 1.3200346921075456, "grad_norm": 1.9879230260849, "learning_rate": 1.1240388800232121e-05, "loss": 0.0432, "step": 12176 }, { "epoch": 1.3201431049436252, "grad_norm": 0.5463106632232666, "learning_rate": 1.1239663426664733e-05, "loss": 0.0089, "step": 12177 }, { "epoch": 1.3202515177797052, "grad_norm": 5.36577844619751, "learning_rate": 1.1238938053097345e-05, "loss": 0.0107, "step": 12178 }, { "epoch": 1.320359930615785, "grad_norm": 6.5918755531311035, "learning_rate": 1.1238212679529959e-05, "loss": 0.0159, "step": 12179 }, { "epoch": 1.3204683434518647, "grad_norm": 3.6269631385803223, "learning_rate": 1.1237487305962571e-05, "loss": 0.049, "step": 12180 }, { "epoch": 1.3205767562879445, "grad_norm": 1.3588038682937622, "learning_rate": 1.1236761932395183e-05, "loss": 0.0175, "step": 12181 }, { "epoch": 1.3206851691240242, "grad_norm": 2.0040297508239746, "learning_rate": 1.1236036558827797e-05, "loss": 0.051, "step": 12182 }, { "epoch": 1.3207935819601042, "grad_norm": 1.7110648155212402, "learning_rate": 1.123531118526041e-05, "loss": 0.0121, "step": 12183 }, { "epoch": 1.3209019947961838, "grad_norm": 2.1353518962860107, "learning_rate": 1.1234585811693021e-05, "loss": 0.0214, "step": 12184 }, { "epoch": 1.3210104076322637, "grad_norm": 1.4683451652526855, "learning_rate": 1.1233860438125637e-05, "loss": 0.0318, "step": 12185 }, { "epoch": 1.3211188204683435, "grad_norm": 1.6417537927627563, "learning_rate": 1.1233135064558249e-05, "loss": 0.0206, "step": 12186 }, { "epoch": 1.3212272333044233, "grad_norm": 0.8159938454627991, "learning_rate": 1.1232409690990861e-05, "loss": 0.0141, "step": 12187 }, { "epoch": 1.321335646140503, "grad_norm": 0.9334996342658997, "learning_rate": 1.1231684317423475e-05, "loss": 0.014, "step": 12188 }, { "epoch": 1.3214440589765828, "grad_norm": 3.8961384296417236, "learning_rate": 1.1230958943856087e-05, "loss": 0.0346, "step": 12189 }, { "epoch": 1.3215524718126626, "grad_norm": 1.5436573028564453, "learning_rate": 1.12302335702887e-05, "loss": 0.0147, "step": 12190 }, { "epoch": 1.3216608846487423, "grad_norm": 0.8930156826972961, "learning_rate": 1.1229508196721313e-05, "loss": 0.013, "step": 12191 }, { "epoch": 1.3217692974848223, "grad_norm": 2.020688056945801, "learning_rate": 1.1228782823153925e-05, "loss": 0.0189, "step": 12192 }, { "epoch": 1.321877710320902, "grad_norm": 0.7886160016059875, "learning_rate": 1.1228057449586538e-05, "loss": 0.0209, "step": 12193 }, { "epoch": 1.3219861231569818, "grad_norm": 1.6681703329086304, "learning_rate": 1.1227332076019151e-05, "loss": 0.0148, "step": 12194 }, { "epoch": 1.3220945359930616, "grad_norm": 2.4134347438812256, "learning_rate": 1.1226606702451764e-05, "loss": 0.0171, "step": 12195 }, { "epoch": 1.3222029488291414, "grad_norm": 1.022861123085022, "learning_rate": 1.1225881328884376e-05, "loss": 0.0047, "step": 12196 }, { "epoch": 1.3223113616652211, "grad_norm": 0.9555264711380005, "learning_rate": 1.122515595531699e-05, "loss": 0.0183, "step": 12197 }, { "epoch": 1.3224197745013009, "grad_norm": 2.5602664947509766, "learning_rate": 1.1224430581749602e-05, "loss": 0.0196, "step": 12198 }, { "epoch": 1.3225281873373809, "grad_norm": 1.207651972770691, "learning_rate": 1.1223705208182214e-05, "loss": 0.0247, "step": 12199 }, { "epoch": 1.3226366001734604, "grad_norm": 0.1354813575744629, "learning_rate": 1.1222979834614826e-05, "loss": 0.0013, "step": 12200 }, { "epoch": 1.3227450130095404, "grad_norm": 7.346349716186523, "learning_rate": 1.122225446104744e-05, "loss": 0.0685, "step": 12201 }, { "epoch": 1.3228534258456202, "grad_norm": 0.45830631256103516, "learning_rate": 1.1221529087480054e-05, "loss": 0.0142, "step": 12202 }, { "epoch": 1.3229618386817, "grad_norm": 0.9419376850128174, "learning_rate": 1.1220803713912667e-05, "loss": 0.0174, "step": 12203 }, { "epoch": 1.3230702515177797, "grad_norm": 0.27297624945640564, "learning_rate": 1.122007834034528e-05, "loss": 0.0061, "step": 12204 }, { "epoch": 1.3231786643538594, "grad_norm": 0.883253276348114, "learning_rate": 1.1219352966777892e-05, "loss": 0.0048, "step": 12205 }, { "epoch": 1.3232870771899392, "grad_norm": 2.4413909912109375, "learning_rate": 1.1218627593210506e-05, "loss": 0.0163, "step": 12206 }, { "epoch": 1.323395490026019, "grad_norm": 0.8968837261199951, "learning_rate": 1.1217902219643118e-05, "loss": 0.0123, "step": 12207 }, { "epoch": 1.323503902862099, "grad_norm": 2.4376649856567383, "learning_rate": 1.121717684607573e-05, "loss": 0.0157, "step": 12208 }, { "epoch": 1.3236123156981787, "grad_norm": 1.9665765762329102, "learning_rate": 1.1216451472508344e-05, "loss": 0.0642, "step": 12209 }, { "epoch": 1.3237207285342585, "grad_norm": 0.37796393036842346, "learning_rate": 1.1215726098940956e-05, "loss": 0.0141, "step": 12210 }, { "epoch": 1.3238291413703382, "grad_norm": 2.977715015411377, "learning_rate": 1.1215000725373568e-05, "loss": 0.0577, "step": 12211 }, { "epoch": 1.323937554206418, "grad_norm": 0.7148692011833191, "learning_rate": 1.121427535180618e-05, "loss": 0.0099, "step": 12212 }, { "epoch": 1.3240459670424978, "grad_norm": 2.8680973052978516, "learning_rate": 1.1213549978238794e-05, "loss": 0.0508, "step": 12213 }, { "epoch": 1.3241543798785775, "grad_norm": 6.458461284637451, "learning_rate": 1.1212824604671406e-05, "loss": 0.011, "step": 12214 }, { "epoch": 1.3242627927146575, "grad_norm": 4.951766014099121, "learning_rate": 1.1212099231104018e-05, "loss": 0.022, "step": 12215 }, { "epoch": 1.3243712055507373, "grad_norm": 1.333493709564209, "learning_rate": 1.1211373857536632e-05, "loss": 0.0222, "step": 12216 }, { "epoch": 1.324479618386817, "grad_norm": 7.520349979400635, "learning_rate": 1.1210648483969244e-05, "loss": 0.048, "step": 12217 }, { "epoch": 1.3245880312228968, "grad_norm": 0.26280316710472107, "learning_rate": 1.1209923110401856e-05, "loss": 0.0026, "step": 12218 }, { "epoch": 1.3246964440589766, "grad_norm": 0.3711051642894745, "learning_rate": 1.1209197736834472e-05, "loss": 0.022, "step": 12219 }, { "epoch": 1.3248048568950563, "grad_norm": 0.32506367564201355, "learning_rate": 1.1208472363267084e-05, "loss": 0.0097, "step": 12220 }, { "epoch": 1.324913269731136, "grad_norm": 0.5153440833091736, "learning_rate": 1.1207746989699698e-05, "loss": 0.0229, "step": 12221 }, { "epoch": 1.325021682567216, "grad_norm": 1.6222885847091675, "learning_rate": 1.120702161613231e-05, "loss": 0.0057, "step": 12222 }, { "epoch": 1.3251300954032956, "grad_norm": 0.5429849028587341, "learning_rate": 1.1206296242564922e-05, "loss": 0.0178, "step": 12223 }, { "epoch": 1.3252385082393756, "grad_norm": 1.800706386566162, "learning_rate": 1.1205570868997534e-05, "loss": 0.0416, "step": 12224 }, { "epoch": 1.3253469210754554, "grad_norm": 1.5160645246505737, "learning_rate": 1.1204845495430148e-05, "loss": 0.0029, "step": 12225 }, { "epoch": 1.3254553339115351, "grad_norm": 1.315646767616272, "learning_rate": 1.120412012186276e-05, "loss": 0.005, "step": 12226 }, { "epoch": 1.325563746747615, "grad_norm": 0.28222280740737915, "learning_rate": 1.1203394748295372e-05, "loss": 0.0066, "step": 12227 }, { "epoch": 1.3256721595836947, "grad_norm": 0.4884192645549774, "learning_rate": 1.1202669374727986e-05, "loss": 0.0154, "step": 12228 }, { "epoch": 1.3257805724197744, "grad_norm": 0.6575177311897278, "learning_rate": 1.1201944001160598e-05, "loss": 0.0093, "step": 12229 }, { "epoch": 1.3258889852558542, "grad_norm": 0.39055269956588745, "learning_rate": 1.120121862759321e-05, "loss": 0.0138, "step": 12230 }, { "epoch": 1.3259973980919342, "grad_norm": 2.506833076477051, "learning_rate": 1.1200493254025824e-05, "loss": 0.0537, "step": 12231 }, { "epoch": 1.326105810928014, "grad_norm": 12.280877113342285, "learning_rate": 1.1199767880458437e-05, "loss": 0.0387, "step": 12232 }, { "epoch": 1.3262142237640937, "grad_norm": 0.03698943555355072, "learning_rate": 1.1199042506891049e-05, "loss": 0.0004, "step": 12233 }, { "epoch": 1.3263226366001735, "grad_norm": 1.5822240114212036, "learning_rate": 1.1198317133323663e-05, "loss": 0.0281, "step": 12234 }, { "epoch": 1.3264310494362532, "grad_norm": 0.3075900375843048, "learning_rate": 1.1197591759756275e-05, "loss": 0.0086, "step": 12235 }, { "epoch": 1.326539462272333, "grad_norm": 15.388504981994629, "learning_rate": 1.1196866386188887e-05, "loss": 0.0613, "step": 12236 }, { "epoch": 1.3266478751084128, "grad_norm": 1.0844494104385376, "learning_rate": 1.1196141012621502e-05, "loss": 0.0038, "step": 12237 }, { "epoch": 1.3267562879444927, "grad_norm": 6.344170093536377, "learning_rate": 1.1195415639054114e-05, "loss": 0.0503, "step": 12238 }, { "epoch": 1.3268647007805723, "grad_norm": 0.03822372108697891, "learning_rate": 1.1194690265486727e-05, "loss": 0.0004, "step": 12239 }, { "epoch": 1.3269731136166523, "grad_norm": 2.5980305671691895, "learning_rate": 1.119396489191934e-05, "loss": 0.0257, "step": 12240 }, { "epoch": 1.327081526452732, "grad_norm": 0.40684273838996887, "learning_rate": 1.1193239518351953e-05, "loss": 0.0151, "step": 12241 }, { "epoch": 1.3271899392888118, "grad_norm": 10.528546333312988, "learning_rate": 1.1192514144784565e-05, "loss": 0.0491, "step": 12242 }, { "epoch": 1.3272983521248916, "grad_norm": 3.188948392868042, "learning_rate": 1.1191788771217179e-05, "loss": 0.0178, "step": 12243 }, { "epoch": 1.3274067649609713, "grad_norm": 0.03835615888237953, "learning_rate": 1.119106339764979e-05, "loss": 0.0003, "step": 12244 }, { "epoch": 1.3275151777970513, "grad_norm": 4.923505783081055, "learning_rate": 1.1190338024082403e-05, "loss": 0.02, "step": 12245 }, { "epoch": 1.3276235906331308, "grad_norm": 1.8958964347839355, "learning_rate": 1.1189612650515017e-05, "loss": 0.0092, "step": 12246 }, { "epoch": 1.3277320034692108, "grad_norm": 1.6200158596038818, "learning_rate": 1.1188887276947629e-05, "loss": 0.0302, "step": 12247 }, { "epoch": 1.3278404163052906, "grad_norm": 2.9929540157318115, "learning_rate": 1.1188161903380241e-05, "loss": 0.0736, "step": 12248 }, { "epoch": 1.3279488291413704, "grad_norm": 0.4099046289920807, "learning_rate": 1.1187436529812855e-05, "loss": 0.0075, "step": 12249 }, { "epoch": 1.3280572419774501, "grad_norm": 4.193930625915527, "learning_rate": 1.1186711156245467e-05, "loss": 0.0387, "step": 12250 }, { "epoch": 1.3281656548135299, "grad_norm": 1.9069560766220093, "learning_rate": 1.1185985782678079e-05, "loss": 0.0781, "step": 12251 }, { "epoch": 1.3282740676496096, "grad_norm": 3.0277981758117676, "learning_rate": 1.1185260409110691e-05, "loss": 0.0473, "step": 12252 }, { "epoch": 1.3283824804856894, "grad_norm": 5.791854381561279, "learning_rate": 1.1184535035543305e-05, "loss": 0.0395, "step": 12253 }, { "epoch": 1.3284908933217694, "grad_norm": 17.502700805664062, "learning_rate": 1.1183809661975919e-05, "loss": 0.0324, "step": 12254 }, { "epoch": 1.3285993061578492, "grad_norm": 3.90354323387146, "learning_rate": 1.1183084288408533e-05, "loss": 0.0367, "step": 12255 }, { "epoch": 1.328707718993929, "grad_norm": 1.9281907081604004, "learning_rate": 1.1182358914841145e-05, "loss": 0.0235, "step": 12256 }, { "epoch": 1.3288161318300087, "grad_norm": 0.3751288950443268, "learning_rate": 1.1181633541273757e-05, "loss": 0.0133, "step": 12257 }, { "epoch": 1.3289245446660884, "grad_norm": 0.42573651671409607, "learning_rate": 1.1180908167706371e-05, "loss": 0.0226, "step": 12258 }, { "epoch": 1.3290329575021682, "grad_norm": 0.09388042241334915, "learning_rate": 1.1180182794138983e-05, "loss": 0.0007, "step": 12259 }, { "epoch": 1.329141370338248, "grad_norm": 0.4440506100654602, "learning_rate": 1.1179457420571595e-05, "loss": 0.0162, "step": 12260 }, { "epoch": 1.329249783174328, "grad_norm": 0.06777733564376831, "learning_rate": 1.1178732047004209e-05, "loss": 0.0008, "step": 12261 }, { "epoch": 1.3293581960104075, "grad_norm": 4.003735065460205, "learning_rate": 1.1178006673436821e-05, "loss": 0.0235, "step": 12262 }, { "epoch": 1.3294666088464875, "grad_norm": 6.027027130126953, "learning_rate": 1.1177281299869433e-05, "loss": 0.0513, "step": 12263 }, { "epoch": 1.3295750216825672, "grad_norm": 1.172899603843689, "learning_rate": 1.1176555926302045e-05, "loss": 0.0078, "step": 12264 }, { "epoch": 1.329683434518647, "grad_norm": 0.9919085502624512, "learning_rate": 1.117583055273466e-05, "loss": 0.0363, "step": 12265 }, { "epoch": 1.3297918473547268, "grad_norm": 0.22407397627830505, "learning_rate": 1.1175105179167271e-05, "loss": 0.0052, "step": 12266 }, { "epoch": 1.3299002601908065, "grad_norm": 0.36368563771247864, "learning_rate": 1.1174379805599884e-05, "loss": 0.0057, "step": 12267 }, { "epoch": 1.3300086730268863, "grad_norm": 0.33866098523139954, "learning_rate": 1.1173654432032497e-05, "loss": 0.0082, "step": 12268 }, { "epoch": 1.330117085862966, "grad_norm": 0.14508645236492157, "learning_rate": 1.117292905846511e-05, "loss": 0.0008, "step": 12269 }, { "epoch": 1.330225498699046, "grad_norm": 2.0390093326568604, "learning_rate": 1.1172203684897722e-05, "loss": 0.0316, "step": 12270 }, { "epoch": 1.3303339115351258, "grad_norm": 5.28831672668457, "learning_rate": 1.1171478311330337e-05, "loss": 0.018, "step": 12271 }, { "epoch": 1.3304423243712056, "grad_norm": 0.1503859907388687, "learning_rate": 1.117075293776295e-05, "loss": 0.0011, "step": 12272 }, { "epoch": 1.3305507372072853, "grad_norm": 3.998990058898926, "learning_rate": 1.1170027564195563e-05, "loss": 0.061, "step": 12273 }, { "epoch": 1.330659150043365, "grad_norm": 0.05346939340233803, "learning_rate": 1.1169302190628175e-05, "loss": 0.0004, "step": 12274 }, { "epoch": 1.3307675628794449, "grad_norm": 4.386535167694092, "learning_rate": 1.1168576817060787e-05, "loss": 0.102, "step": 12275 }, { "epoch": 1.3308759757155246, "grad_norm": 0.40697601437568665, "learning_rate": 1.11678514434934e-05, "loss": 0.0124, "step": 12276 }, { "epoch": 1.3309843885516046, "grad_norm": 4.285304069519043, "learning_rate": 1.1167126069926013e-05, "loss": 0.026, "step": 12277 }, { "epoch": 1.3310928013876844, "grad_norm": 1.0583091974258423, "learning_rate": 1.1166400696358626e-05, "loss": 0.0033, "step": 12278 }, { "epoch": 1.3312012142237641, "grad_norm": 0.5387941002845764, "learning_rate": 1.1165675322791238e-05, "loss": 0.0031, "step": 12279 }, { "epoch": 1.331309627059844, "grad_norm": 4.107535362243652, "learning_rate": 1.1164949949223852e-05, "loss": 0.0331, "step": 12280 }, { "epoch": 1.3314180398959237, "grad_norm": 2.281899929046631, "learning_rate": 1.1164224575656464e-05, "loss": 0.0192, "step": 12281 }, { "epoch": 1.3315264527320034, "grad_norm": 4.025112628936768, "learning_rate": 1.1163499202089076e-05, "loss": 0.0355, "step": 12282 }, { "epoch": 1.3316348655680832, "grad_norm": 2.3640177249908447, "learning_rate": 1.116277382852169e-05, "loss": 0.0272, "step": 12283 }, { "epoch": 1.3317432784041632, "grad_norm": 5.829911231994629, "learning_rate": 1.1162048454954302e-05, "loss": 0.0133, "step": 12284 }, { "epoch": 1.3318516912402427, "grad_norm": 0.41428273916244507, "learning_rate": 1.1161323081386914e-05, "loss": 0.0125, "step": 12285 }, { "epoch": 1.3319601040763227, "grad_norm": 2.756307363510132, "learning_rate": 1.1160597707819528e-05, "loss": 0.0119, "step": 12286 }, { "epoch": 1.3320685169124025, "grad_norm": 0.9858490824699402, "learning_rate": 1.115987233425214e-05, "loss": 0.0092, "step": 12287 }, { "epoch": 1.3321769297484822, "grad_norm": 5.582947731018066, "learning_rate": 1.1159146960684754e-05, "loss": 0.0352, "step": 12288 }, { "epoch": 1.332285342584562, "grad_norm": 0.5046470761299133, "learning_rate": 1.1158421587117368e-05, "loss": 0.0257, "step": 12289 }, { "epoch": 1.3323937554206418, "grad_norm": 3.119633436203003, "learning_rate": 1.115769621354998e-05, "loss": 0.0357, "step": 12290 }, { "epoch": 1.3325021682567215, "grad_norm": 0.8430535197257996, "learning_rate": 1.1156970839982592e-05, "loss": 0.01, "step": 12291 }, { "epoch": 1.3326105810928013, "grad_norm": 2.157080888748169, "learning_rate": 1.1156245466415206e-05, "loss": 0.0049, "step": 12292 }, { "epoch": 1.3327189939288813, "grad_norm": 5.000466346740723, "learning_rate": 1.1155520092847818e-05, "loss": 0.0321, "step": 12293 }, { "epoch": 1.332827406764961, "grad_norm": 1.0483285188674927, "learning_rate": 1.115479471928043e-05, "loss": 0.0121, "step": 12294 }, { "epoch": 1.3329358196010408, "grad_norm": 5.684298515319824, "learning_rate": 1.1154069345713044e-05, "loss": 0.0355, "step": 12295 }, { "epoch": 1.3330442324371206, "grad_norm": 1.352524757385254, "learning_rate": 1.1153343972145656e-05, "loss": 0.0099, "step": 12296 }, { "epoch": 1.3331526452732003, "grad_norm": 0.21022039651870728, "learning_rate": 1.1152618598578268e-05, "loss": 0.0019, "step": 12297 }, { "epoch": 1.33326105810928, "grad_norm": 0.7752771377563477, "learning_rate": 1.1151893225010882e-05, "loss": 0.0029, "step": 12298 }, { "epoch": 1.3333694709453598, "grad_norm": 2.4360642433166504, "learning_rate": 1.1151167851443494e-05, "loss": 0.0146, "step": 12299 }, { "epoch": 1.3334778837814398, "grad_norm": 3.3120272159576416, "learning_rate": 1.1150442477876106e-05, "loss": 0.0345, "step": 12300 }, { "epoch": 1.3335862966175196, "grad_norm": 2.4877800941467285, "learning_rate": 1.114971710430872e-05, "loss": 0.018, "step": 12301 }, { "epoch": 1.3336947094535994, "grad_norm": 4.2943549156188965, "learning_rate": 1.1148991730741332e-05, "loss": 0.0283, "step": 12302 }, { "epoch": 1.3338031222896791, "grad_norm": 4.748287200927734, "learning_rate": 1.1148266357173944e-05, "loss": 0.0305, "step": 12303 }, { "epoch": 1.3339115351257589, "grad_norm": 1.35645592212677, "learning_rate": 1.1147540983606557e-05, "loss": 0.022, "step": 12304 }, { "epoch": 1.3340199479618386, "grad_norm": 1.0931986570358276, "learning_rate": 1.114681561003917e-05, "loss": 0.0087, "step": 12305 }, { "epoch": 1.3341283607979184, "grad_norm": 0.38770851492881775, "learning_rate": 1.1146090236471784e-05, "loss": 0.0164, "step": 12306 }, { "epoch": 1.3342367736339984, "grad_norm": 0.24696242809295654, "learning_rate": 1.1145364862904398e-05, "loss": 0.0014, "step": 12307 }, { "epoch": 1.334345186470078, "grad_norm": 9.410857200622559, "learning_rate": 1.114463948933701e-05, "loss": 0.0308, "step": 12308 }, { "epoch": 1.334453599306158, "grad_norm": 0.4892594814300537, "learning_rate": 1.1143914115769622e-05, "loss": 0.0129, "step": 12309 }, { "epoch": 1.3345620121422377, "grad_norm": 5.626303195953369, "learning_rate": 1.1143188742202236e-05, "loss": 0.0961, "step": 12310 }, { "epoch": 1.3346704249783174, "grad_norm": 10.634540557861328, "learning_rate": 1.1142463368634848e-05, "loss": 0.0346, "step": 12311 }, { "epoch": 1.3347788378143972, "grad_norm": 2.327418088912964, "learning_rate": 1.114173799506746e-05, "loss": 0.0847, "step": 12312 }, { "epoch": 1.334887250650477, "grad_norm": 1.5467510223388672, "learning_rate": 1.1141012621500074e-05, "loss": 0.0318, "step": 12313 }, { "epoch": 1.3349956634865567, "grad_norm": 2.9061145782470703, "learning_rate": 1.1140287247932686e-05, "loss": 0.0131, "step": 12314 }, { "epoch": 1.3351040763226365, "grad_norm": 1.407687783241272, "learning_rate": 1.1139561874365299e-05, "loss": 0.0235, "step": 12315 }, { "epoch": 1.3352124891587165, "grad_norm": 2.784978151321411, "learning_rate": 1.113883650079791e-05, "loss": 0.0346, "step": 12316 }, { "epoch": 1.3353209019947962, "grad_norm": 1.2823127508163452, "learning_rate": 1.1138111127230525e-05, "loss": 0.021, "step": 12317 }, { "epoch": 1.335429314830876, "grad_norm": 2.701319932937622, "learning_rate": 1.1137385753663137e-05, "loss": 0.0174, "step": 12318 }, { "epoch": 1.3355377276669558, "grad_norm": 7.796424388885498, "learning_rate": 1.1136660380095749e-05, "loss": 0.0195, "step": 12319 }, { "epoch": 1.3356461405030355, "grad_norm": 0.3284701406955719, "learning_rate": 1.1135935006528363e-05, "loss": 0.0067, "step": 12320 }, { "epoch": 1.3357545533391153, "grad_norm": 1.2067129611968994, "learning_rate": 1.1135209632960975e-05, "loss": 0.0251, "step": 12321 }, { "epoch": 1.335862966175195, "grad_norm": 0.7397680878639221, "learning_rate": 1.1134484259393587e-05, "loss": 0.0047, "step": 12322 }, { "epoch": 1.335971379011275, "grad_norm": 10.361900329589844, "learning_rate": 1.1133758885826203e-05, "loss": 0.0125, "step": 12323 }, { "epoch": 1.3360797918473546, "grad_norm": 2.3722715377807617, "learning_rate": 1.1133033512258815e-05, "loss": 0.0373, "step": 12324 }, { "epoch": 1.3361882046834346, "grad_norm": 2.2170495986938477, "learning_rate": 1.1132308138691429e-05, "loss": 0.0172, "step": 12325 }, { "epoch": 1.3362966175195143, "grad_norm": 0.2952810227870941, "learning_rate": 1.113158276512404e-05, "loss": 0.0133, "step": 12326 }, { "epoch": 1.336405030355594, "grad_norm": 0.38429298996925354, "learning_rate": 1.1130857391556653e-05, "loss": 0.0205, "step": 12327 }, { "epoch": 1.3365134431916739, "grad_norm": 0.7580405473709106, "learning_rate": 1.1130132017989265e-05, "loss": 0.0058, "step": 12328 }, { "epoch": 1.3366218560277536, "grad_norm": 3.735250234603882, "learning_rate": 1.1129406644421879e-05, "loss": 0.0323, "step": 12329 }, { "epoch": 1.3367302688638336, "grad_norm": 2.375180244445801, "learning_rate": 1.1128681270854491e-05, "loss": 0.0246, "step": 12330 }, { "epoch": 1.3368386816999132, "grad_norm": 0.39486658573150635, "learning_rate": 1.1127955897287103e-05, "loss": 0.0124, "step": 12331 }, { "epoch": 1.3369470945359931, "grad_norm": 2.301032781600952, "learning_rate": 1.1127230523719717e-05, "loss": 0.0486, "step": 12332 }, { "epoch": 1.337055507372073, "grad_norm": 0.7256770133972168, "learning_rate": 1.1126505150152329e-05, "loss": 0.009, "step": 12333 }, { "epoch": 1.3371639202081527, "grad_norm": 4.169882774353027, "learning_rate": 1.1125779776584941e-05, "loss": 0.0566, "step": 12334 }, { "epoch": 1.3372723330442324, "grad_norm": 1.0964032411575317, "learning_rate": 1.1125054403017555e-05, "loss": 0.014, "step": 12335 }, { "epoch": 1.3373807458803122, "grad_norm": 2.0979955196380615, "learning_rate": 1.1124329029450167e-05, "loss": 0.0197, "step": 12336 }, { "epoch": 1.337489158716392, "grad_norm": 0.05607980117201805, "learning_rate": 1.112360365588278e-05, "loss": 0.0006, "step": 12337 }, { "epoch": 1.3375975715524717, "grad_norm": 2.8300492763519287, "learning_rate": 1.1122878282315393e-05, "loss": 0.0317, "step": 12338 }, { "epoch": 1.3377059843885517, "grad_norm": 2.148481845855713, "learning_rate": 1.1122152908748005e-05, "loss": 0.0367, "step": 12339 }, { "epoch": 1.3378143972246315, "grad_norm": 2.245722770690918, "learning_rate": 1.1121427535180621e-05, "loss": 0.0475, "step": 12340 }, { "epoch": 1.3379228100607112, "grad_norm": 1.468888282775879, "learning_rate": 1.1120702161613233e-05, "loss": 0.0317, "step": 12341 }, { "epoch": 1.338031222896791, "grad_norm": 0.6719309687614441, "learning_rate": 1.1119976788045845e-05, "loss": 0.0123, "step": 12342 }, { "epoch": 1.3381396357328708, "grad_norm": 2.3863070011138916, "learning_rate": 1.1119251414478457e-05, "loss": 0.0295, "step": 12343 }, { "epoch": 1.3382480485689505, "grad_norm": 1.1948176622390747, "learning_rate": 1.1118526040911071e-05, "loss": 0.0081, "step": 12344 }, { "epoch": 1.3383564614050303, "grad_norm": 4.259799957275391, "learning_rate": 1.1117800667343683e-05, "loss": 0.0276, "step": 12345 }, { "epoch": 1.3384648742411103, "grad_norm": 1.24160897731781, "learning_rate": 1.1117075293776295e-05, "loss": 0.0208, "step": 12346 }, { "epoch": 1.3385732870771898, "grad_norm": 5.127392292022705, "learning_rate": 1.111634992020891e-05, "loss": 0.009, "step": 12347 }, { "epoch": 1.3386816999132698, "grad_norm": 5.244153022766113, "learning_rate": 1.1115624546641521e-05, "loss": 0.0335, "step": 12348 }, { "epoch": 1.3387901127493496, "grad_norm": 2.079350471496582, "learning_rate": 1.1114899173074134e-05, "loss": 0.0085, "step": 12349 }, { "epoch": 1.3388985255854293, "grad_norm": 2.040583372116089, "learning_rate": 1.1114173799506747e-05, "loss": 0.0255, "step": 12350 }, { "epoch": 1.339006938421509, "grad_norm": 0.34585461020469666, "learning_rate": 1.111344842593936e-05, "loss": 0.0073, "step": 12351 }, { "epoch": 1.3391153512575888, "grad_norm": 2.9654784202575684, "learning_rate": 1.1112723052371972e-05, "loss": 0.0195, "step": 12352 }, { "epoch": 1.3392237640936686, "grad_norm": 3.038510322570801, "learning_rate": 1.1111997678804586e-05, "loss": 0.0163, "step": 12353 }, { "epoch": 1.3393321769297484, "grad_norm": 2.8491618633270264, "learning_rate": 1.1111272305237198e-05, "loss": 0.0213, "step": 12354 }, { "epoch": 1.3394405897658284, "grad_norm": 1.4494272470474243, "learning_rate": 1.111054693166981e-05, "loss": 0.0345, "step": 12355 }, { "epoch": 1.3395490026019081, "grad_norm": 1.1679470539093018, "learning_rate": 1.1109821558102422e-05, "loss": 0.0187, "step": 12356 }, { "epoch": 1.3396574154379879, "grad_norm": 2.4805705547332764, "learning_rate": 1.1109096184535036e-05, "loss": 0.016, "step": 12357 }, { "epoch": 1.3397658282740676, "grad_norm": 4.278524398803711, "learning_rate": 1.110837081096765e-05, "loss": 0.0428, "step": 12358 }, { "epoch": 1.3398742411101474, "grad_norm": 0.5805615782737732, "learning_rate": 1.1107645437400263e-05, "loss": 0.0015, "step": 12359 }, { "epoch": 1.3399826539462272, "grad_norm": 6.399397373199463, "learning_rate": 1.1106920063832876e-05, "loss": 0.0422, "step": 12360 }, { "epoch": 1.340091066782307, "grad_norm": 0.6887749433517456, "learning_rate": 1.1106194690265488e-05, "loss": 0.0191, "step": 12361 }, { "epoch": 1.340199479618387, "grad_norm": 0.8130899667739868, "learning_rate": 1.1105469316698102e-05, "loss": 0.0128, "step": 12362 }, { "epoch": 1.3403078924544667, "grad_norm": 2.2171976566314697, "learning_rate": 1.1104743943130714e-05, "loss": 0.0458, "step": 12363 }, { "epoch": 1.3404163052905465, "grad_norm": 4.7283244132995605, "learning_rate": 1.1104018569563326e-05, "loss": 0.0287, "step": 12364 }, { "epoch": 1.3405247181266262, "grad_norm": 2.2842910289764404, "learning_rate": 1.110329319599594e-05, "loss": 0.0066, "step": 12365 }, { "epoch": 1.340633130962706, "grad_norm": 2.069981098175049, "learning_rate": 1.1102567822428552e-05, "loss": 0.0057, "step": 12366 }, { "epoch": 1.3407415437987857, "grad_norm": 0.8900951147079468, "learning_rate": 1.1101842448861164e-05, "loss": 0.003, "step": 12367 }, { "epoch": 1.3408499566348655, "grad_norm": 1.5228129625320435, "learning_rate": 1.1101117075293776e-05, "loss": 0.0102, "step": 12368 }, { "epoch": 1.3409583694709455, "grad_norm": 0.5567495822906494, "learning_rate": 1.110039170172639e-05, "loss": 0.022, "step": 12369 }, { "epoch": 1.341066782307025, "grad_norm": 3.744415044784546, "learning_rate": 1.1099666328159002e-05, "loss": 0.0213, "step": 12370 }, { "epoch": 1.341175195143105, "grad_norm": 2.7958760261535645, "learning_rate": 1.1098940954591614e-05, "loss": 0.0326, "step": 12371 }, { "epoch": 1.3412836079791848, "grad_norm": 7.448126316070557, "learning_rate": 1.1098215581024228e-05, "loss": 0.0563, "step": 12372 }, { "epoch": 1.3413920208152645, "grad_norm": 1.0551584959030151, "learning_rate": 1.109749020745684e-05, "loss": 0.0122, "step": 12373 }, { "epoch": 1.3415004336513443, "grad_norm": 0.793107271194458, "learning_rate": 1.1096764833889452e-05, "loss": 0.0137, "step": 12374 }, { "epoch": 1.341608846487424, "grad_norm": 6.976545333862305, "learning_rate": 1.1096039460322068e-05, "loss": 0.0495, "step": 12375 }, { "epoch": 1.3417172593235038, "grad_norm": 1.0143795013427734, "learning_rate": 1.109531408675468e-05, "loss": 0.0109, "step": 12376 }, { "epoch": 1.3418256721595836, "grad_norm": 2.9813334941864014, "learning_rate": 1.1094588713187294e-05, "loss": 0.0146, "step": 12377 }, { "epoch": 1.3419340849956636, "grad_norm": 0.48127859830856323, "learning_rate": 1.1093863339619906e-05, "loss": 0.0096, "step": 12378 }, { "epoch": 1.3420424978317433, "grad_norm": 1.7654504776000977, "learning_rate": 1.1093137966052518e-05, "loss": 0.0278, "step": 12379 }, { "epoch": 1.342150910667823, "grad_norm": 2.8043124675750732, "learning_rate": 1.109241259248513e-05, "loss": 0.0327, "step": 12380 }, { "epoch": 1.3422593235039029, "grad_norm": 3.546814203262329, "learning_rate": 1.1091687218917744e-05, "loss": 0.0429, "step": 12381 }, { "epoch": 1.3423677363399826, "grad_norm": 1.3364510536193848, "learning_rate": 1.1090961845350356e-05, "loss": 0.0201, "step": 12382 }, { "epoch": 1.3424761491760624, "grad_norm": 2.6671290397644043, "learning_rate": 1.1090236471782968e-05, "loss": 0.0056, "step": 12383 }, { "epoch": 1.3425845620121422, "grad_norm": 1.071080207824707, "learning_rate": 1.1089511098215582e-05, "loss": 0.0122, "step": 12384 }, { "epoch": 1.3426929748482221, "grad_norm": 5.1929802894592285, "learning_rate": 1.1088785724648194e-05, "loss": 0.0389, "step": 12385 }, { "epoch": 1.3428013876843017, "grad_norm": 12.817906379699707, "learning_rate": 1.1088060351080807e-05, "loss": 0.0209, "step": 12386 }, { "epoch": 1.3429098005203817, "grad_norm": 0.7770147919654846, "learning_rate": 1.108733497751342e-05, "loss": 0.0074, "step": 12387 }, { "epoch": 1.3430182133564614, "grad_norm": 0.05849982425570488, "learning_rate": 1.1086609603946033e-05, "loss": 0.0005, "step": 12388 }, { "epoch": 1.3431266261925412, "grad_norm": 5.583581447601318, "learning_rate": 1.1085884230378645e-05, "loss": 0.0981, "step": 12389 }, { "epoch": 1.343235039028621, "grad_norm": 5.5937604904174805, "learning_rate": 1.1085158856811259e-05, "loss": 0.0243, "step": 12390 }, { "epoch": 1.3433434518647007, "grad_norm": 0.35968974232673645, "learning_rate": 1.108443348324387e-05, "loss": 0.0099, "step": 12391 }, { "epoch": 1.3434518647007807, "grad_norm": 4.466398239135742, "learning_rate": 1.1083708109676486e-05, "loss": 0.0355, "step": 12392 }, { "epoch": 1.3435602775368602, "grad_norm": 0.41432875394821167, "learning_rate": 1.1082982736109098e-05, "loss": 0.0123, "step": 12393 }, { "epoch": 1.3436686903729402, "grad_norm": 0.02769717201590538, "learning_rate": 1.108225736254171e-05, "loss": 0.0003, "step": 12394 }, { "epoch": 1.34377710320902, "grad_norm": 0.6423577666282654, "learning_rate": 1.1081531988974323e-05, "loss": 0.018, "step": 12395 }, { "epoch": 1.3438855160450998, "grad_norm": 0.7838678956031799, "learning_rate": 1.1080806615406936e-05, "loss": 0.0105, "step": 12396 }, { "epoch": 1.3439939288811795, "grad_norm": 4.740433692932129, "learning_rate": 1.1080081241839549e-05, "loss": 0.0458, "step": 12397 }, { "epoch": 1.3441023417172593, "grad_norm": 3.8255088329315186, "learning_rate": 1.107935586827216e-05, "loss": 0.0924, "step": 12398 }, { "epoch": 1.344210754553339, "grad_norm": 2.0995707511901855, "learning_rate": 1.1078630494704775e-05, "loss": 0.0292, "step": 12399 }, { "epoch": 1.3443191673894188, "grad_norm": 1.2183257341384888, "learning_rate": 1.1077905121137387e-05, "loss": 0.0061, "step": 12400 }, { "epoch": 1.3444275802254988, "grad_norm": 1.3718912601470947, "learning_rate": 1.1077179747569999e-05, "loss": 0.0631, "step": 12401 }, { "epoch": 1.3445359930615786, "grad_norm": 2.6643714904785156, "learning_rate": 1.1076454374002613e-05, "loss": 0.0773, "step": 12402 }, { "epoch": 1.3446444058976583, "grad_norm": 4.712092399597168, "learning_rate": 1.1075729000435225e-05, "loss": 0.0664, "step": 12403 }, { "epoch": 1.344752818733738, "grad_norm": 5.123619079589844, "learning_rate": 1.1075003626867837e-05, "loss": 0.031, "step": 12404 }, { "epoch": 1.3448612315698178, "grad_norm": 0.40736910700798035, "learning_rate": 1.107427825330045e-05, "loss": 0.0097, "step": 12405 }, { "epoch": 1.3449696444058976, "grad_norm": 3.1368026733398438, "learning_rate": 1.1073552879733063e-05, "loss": 0.0188, "step": 12406 }, { "epoch": 1.3450780572419774, "grad_norm": 1.5478107929229736, "learning_rate": 1.1072827506165675e-05, "loss": 0.0076, "step": 12407 }, { "epoch": 1.3451864700780574, "grad_norm": 2.8627569675445557, "learning_rate": 1.1072102132598287e-05, "loss": 0.0303, "step": 12408 }, { "epoch": 1.345294882914137, "grad_norm": 4.198646545410156, "learning_rate": 1.1071376759030903e-05, "loss": 0.0681, "step": 12409 }, { "epoch": 1.3454032957502169, "grad_norm": 0.42207440733909607, "learning_rate": 1.1070651385463515e-05, "loss": 0.0086, "step": 12410 }, { "epoch": 1.3455117085862967, "grad_norm": 1.8075282573699951, "learning_rate": 1.1069926011896129e-05, "loss": 0.0227, "step": 12411 }, { "epoch": 1.3456201214223764, "grad_norm": 0.9016944169998169, "learning_rate": 1.1069200638328741e-05, "loss": 0.0127, "step": 12412 }, { "epoch": 1.3457285342584562, "grad_norm": 0.7099564075469971, "learning_rate": 1.1068475264761353e-05, "loss": 0.0056, "step": 12413 }, { "epoch": 1.345836947094536, "grad_norm": 0.2206236869096756, "learning_rate": 1.1067749891193967e-05, "loss": 0.0012, "step": 12414 }, { "epoch": 1.3459453599306157, "grad_norm": 0.6638776659965515, "learning_rate": 1.1067024517626579e-05, "loss": 0.0126, "step": 12415 }, { "epoch": 1.3460537727666955, "grad_norm": 2.4679250717163086, "learning_rate": 1.1066299144059191e-05, "loss": 0.0096, "step": 12416 }, { "epoch": 1.3461621856027755, "grad_norm": 7.415048599243164, "learning_rate": 1.1065573770491805e-05, "loss": 0.0895, "step": 12417 }, { "epoch": 1.3462705984388552, "grad_norm": 0.40961721539497375, "learning_rate": 1.1064848396924417e-05, "loss": 0.0083, "step": 12418 }, { "epoch": 1.346379011274935, "grad_norm": 2.8525538444519043, "learning_rate": 1.106412302335703e-05, "loss": 0.0125, "step": 12419 }, { "epoch": 1.3464874241110147, "grad_norm": 6.236695289611816, "learning_rate": 1.1063397649789641e-05, "loss": 0.039, "step": 12420 }, { "epoch": 1.3465958369470945, "grad_norm": 0.5613144636154175, "learning_rate": 1.1062672276222255e-05, "loss": 0.0058, "step": 12421 }, { "epoch": 1.3467042497831743, "grad_norm": 1.5501114130020142, "learning_rate": 1.1061946902654867e-05, "loss": 0.0089, "step": 12422 }, { "epoch": 1.346812662619254, "grad_norm": 8.1350679397583, "learning_rate": 1.106122152908748e-05, "loss": 0.0416, "step": 12423 }, { "epoch": 1.346921075455334, "grad_norm": 6.061282157897949, "learning_rate": 1.1060496155520093e-05, "loss": 0.0536, "step": 12424 }, { "epoch": 1.3470294882914138, "grad_norm": 7.883142471313477, "learning_rate": 1.1059770781952706e-05, "loss": 0.043, "step": 12425 }, { "epoch": 1.3471379011274935, "grad_norm": 0.2528552711009979, "learning_rate": 1.1059045408385318e-05, "loss": 0.0022, "step": 12426 }, { "epoch": 1.3472463139635733, "grad_norm": 0.36286404728889465, "learning_rate": 1.1058320034817933e-05, "loss": 0.0085, "step": 12427 }, { "epoch": 1.347354726799653, "grad_norm": 0.9040703773498535, "learning_rate": 1.1057594661250545e-05, "loss": 0.0261, "step": 12428 }, { "epoch": 1.3474631396357328, "grad_norm": 1.934842824935913, "learning_rate": 1.105686928768316e-05, "loss": 0.0385, "step": 12429 }, { "epoch": 1.3475715524718126, "grad_norm": 1.9717130661010742, "learning_rate": 1.1056143914115771e-05, "loss": 0.0286, "step": 12430 }, { "epoch": 1.3476799653078926, "grad_norm": 0.848335862159729, "learning_rate": 1.1055418540548384e-05, "loss": 0.0163, "step": 12431 }, { "epoch": 1.3477883781439721, "grad_norm": 3.8485794067382812, "learning_rate": 1.1054693166980996e-05, "loss": 0.0206, "step": 12432 }, { "epoch": 1.347896790980052, "grad_norm": 5.251617908477783, "learning_rate": 1.105396779341361e-05, "loss": 0.0355, "step": 12433 }, { "epoch": 1.3480052038161319, "grad_norm": 1.1926014423370361, "learning_rate": 1.1053242419846222e-05, "loss": 0.0179, "step": 12434 }, { "epoch": 1.3481136166522116, "grad_norm": 1.2049165964126587, "learning_rate": 1.1052517046278834e-05, "loss": 0.033, "step": 12435 }, { "epoch": 1.3482220294882914, "grad_norm": 6.286747455596924, "learning_rate": 1.1051791672711448e-05, "loss": 0.0802, "step": 12436 }, { "epoch": 1.3483304423243712, "grad_norm": 0.3535017669200897, "learning_rate": 1.105106629914406e-05, "loss": 0.0072, "step": 12437 }, { "epoch": 1.348438855160451, "grad_norm": 0.9073785543441772, "learning_rate": 1.1050340925576672e-05, "loss": 0.0122, "step": 12438 }, { "epoch": 1.3485472679965307, "grad_norm": 1.3539167642593384, "learning_rate": 1.1049615552009286e-05, "loss": 0.027, "step": 12439 }, { "epoch": 1.3486556808326107, "grad_norm": 2.1949026584625244, "learning_rate": 1.1048890178441898e-05, "loss": 0.0128, "step": 12440 }, { "epoch": 1.3487640936686904, "grad_norm": 1.974877953529358, "learning_rate": 1.104816480487451e-05, "loss": 0.0219, "step": 12441 }, { "epoch": 1.3488725065047702, "grad_norm": 2.796315908432007, "learning_rate": 1.1047439431307124e-05, "loss": 0.0426, "step": 12442 }, { "epoch": 1.34898091934085, "grad_norm": 2.163970708847046, "learning_rate": 1.1046714057739736e-05, "loss": 0.05, "step": 12443 }, { "epoch": 1.3490893321769297, "grad_norm": 4.990420341491699, "learning_rate": 1.1045988684172352e-05, "loss": 0.0323, "step": 12444 }, { "epoch": 1.3491977450130095, "grad_norm": 3.722914218902588, "learning_rate": 1.1045263310604964e-05, "loss": 0.0578, "step": 12445 }, { "epoch": 1.3493061578490892, "grad_norm": 0.41108080744743347, "learning_rate": 1.1044537937037576e-05, "loss": 0.0076, "step": 12446 }, { "epoch": 1.3494145706851692, "grad_norm": 2.224228858947754, "learning_rate": 1.1043812563470188e-05, "loss": 0.0381, "step": 12447 }, { "epoch": 1.349522983521249, "grad_norm": 1.8439202308654785, "learning_rate": 1.1043087189902802e-05, "loss": 0.0167, "step": 12448 }, { "epoch": 1.3496313963573288, "grad_norm": 0.3253989517688751, "learning_rate": 1.1042361816335414e-05, "loss": 0.0167, "step": 12449 }, { "epoch": 1.3497398091934085, "grad_norm": 0.6761869192123413, "learning_rate": 1.1041636442768026e-05, "loss": 0.0036, "step": 12450 }, { "epoch": 1.3498482220294883, "grad_norm": 3.3944859504699707, "learning_rate": 1.104091106920064e-05, "loss": 0.0284, "step": 12451 }, { "epoch": 1.349956634865568, "grad_norm": 0.46924781799316406, "learning_rate": 1.1040185695633252e-05, "loss": 0.0045, "step": 12452 }, { "epoch": 1.3500650477016478, "grad_norm": 5.3033127784729, "learning_rate": 1.1039460322065864e-05, "loss": 0.044, "step": 12453 }, { "epoch": 1.3501734605377278, "grad_norm": 6.583380699157715, "learning_rate": 1.1038734948498478e-05, "loss": 0.0582, "step": 12454 }, { "epoch": 1.3502818733738073, "grad_norm": 1.89506995677948, "learning_rate": 1.103800957493109e-05, "loss": 0.0117, "step": 12455 }, { "epoch": 1.3503902862098873, "grad_norm": 2.9344482421875, "learning_rate": 1.1037284201363702e-05, "loss": 0.0146, "step": 12456 }, { "epoch": 1.350498699045967, "grad_norm": 6.564068794250488, "learning_rate": 1.1036558827796316e-05, "loss": 0.0114, "step": 12457 }, { "epoch": 1.3506071118820469, "grad_norm": 2.7078680992126465, "learning_rate": 1.1035833454228928e-05, "loss": 0.041, "step": 12458 }, { "epoch": 1.3507155247181266, "grad_norm": 6.629741668701172, "learning_rate": 1.103510808066154e-05, "loss": 0.0408, "step": 12459 }, { "epoch": 1.3508239375542064, "grad_norm": 1.728696584701538, "learning_rate": 1.1034382707094153e-05, "loss": 0.0111, "step": 12460 }, { "epoch": 1.3509323503902861, "grad_norm": 0.34801799058914185, "learning_rate": 1.1033657333526768e-05, "loss": 0.0068, "step": 12461 }, { "epoch": 1.351040763226366, "grad_norm": 4.319174766540527, "learning_rate": 1.103293195995938e-05, "loss": 0.0726, "step": 12462 }, { "epoch": 1.351149176062446, "grad_norm": 0.9215168952941895, "learning_rate": 1.1032206586391994e-05, "loss": 0.0255, "step": 12463 }, { "epoch": 1.3512575888985257, "grad_norm": 0.4547075033187866, "learning_rate": 1.1031481212824606e-05, "loss": 0.0109, "step": 12464 }, { "epoch": 1.3513660017346054, "grad_norm": 10.061135292053223, "learning_rate": 1.1030755839257218e-05, "loss": 0.0198, "step": 12465 }, { "epoch": 1.3514744145706852, "grad_norm": 0.5235917568206787, "learning_rate": 1.1030030465689832e-05, "loss": 0.0126, "step": 12466 }, { "epoch": 1.351582827406765, "grad_norm": 0.06482855975627899, "learning_rate": 1.1029305092122444e-05, "loss": 0.0007, "step": 12467 }, { "epoch": 1.3516912402428447, "grad_norm": 0.7259129881858826, "learning_rate": 1.1028579718555057e-05, "loss": 0.0028, "step": 12468 }, { "epoch": 1.3517996530789245, "grad_norm": 0.8942561149597168, "learning_rate": 1.102785434498767e-05, "loss": 0.0156, "step": 12469 }, { "epoch": 1.3519080659150045, "grad_norm": 4.219942092895508, "learning_rate": 1.1027128971420283e-05, "loss": 0.0083, "step": 12470 }, { "epoch": 1.352016478751084, "grad_norm": 1.2492886781692505, "learning_rate": 1.1026403597852895e-05, "loss": 0.0153, "step": 12471 }, { "epoch": 1.352124891587164, "grad_norm": 2.0793092250823975, "learning_rate": 1.1025678224285507e-05, "loss": 0.0216, "step": 12472 }, { "epoch": 1.3522333044232437, "grad_norm": 1.7154573202133179, "learning_rate": 1.102495285071812e-05, "loss": 0.0428, "step": 12473 }, { "epoch": 1.3523417172593235, "grad_norm": 1.8504096269607544, "learning_rate": 1.1024227477150733e-05, "loss": 0.0409, "step": 12474 }, { "epoch": 1.3524501300954033, "grad_norm": 1.0230803489685059, "learning_rate": 1.1023502103583345e-05, "loss": 0.0293, "step": 12475 }, { "epoch": 1.352558542931483, "grad_norm": 1.4804795980453491, "learning_rate": 1.1022776730015959e-05, "loss": 0.0335, "step": 12476 }, { "epoch": 1.352666955767563, "grad_norm": 1.7326421737670898, "learning_rate": 1.1022051356448571e-05, "loss": 0.0262, "step": 12477 }, { "epoch": 1.3527753686036426, "grad_norm": 0.037193529307842255, "learning_rate": 1.1021325982881183e-05, "loss": 0.0003, "step": 12478 }, { "epoch": 1.3528837814397225, "grad_norm": 3.6545627117156982, "learning_rate": 1.1020600609313799e-05, "loss": 0.0238, "step": 12479 }, { "epoch": 1.3529921942758023, "grad_norm": 1.7205955982208252, "learning_rate": 1.101987523574641e-05, "loss": 0.0138, "step": 12480 }, { "epoch": 1.353100607111882, "grad_norm": 3.573281764984131, "learning_rate": 1.1019149862179025e-05, "loss": 0.0795, "step": 12481 }, { "epoch": 1.3532090199479618, "grad_norm": 6.96619987487793, "learning_rate": 1.1018424488611637e-05, "loss": 0.0466, "step": 12482 }, { "epoch": 1.3533174327840416, "grad_norm": 0.32917520403862, "learning_rate": 1.1017699115044249e-05, "loss": 0.0091, "step": 12483 }, { "epoch": 1.3534258456201214, "grad_norm": 4.439688205718994, "learning_rate": 1.1016973741476861e-05, "loss": 0.007, "step": 12484 }, { "epoch": 1.3535342584562011, "grad_norm": 5.766302108764648, "learning_rate": 1.1016248367909475e-05, "loss": 0.0281, "step": 12485 }, { "epoch": 1.353642671292281, "grad_norm": 3.31373929977417, "learning_rate": 1.1015522994342087e-05, "loss": 0.0257, "step": 12486 }, { "epoch": 1.3537510841283609, "grad_norm": 3.6227104663848877, "learning_rate": 1.1014797620774699e-05, "loss": 0.0208, "step": 12487 }, { "epoch": 1.3538594969644406, "grad_norm": 2.4178898334503174, "learning_rate": 1.1014072247207313e-05, "loss": 0.0478, "step": 12488 }, { "epoch": 1.3539679098005204, "grad_norm": 3.0778634548187256, "learning_rate": 1.1013346873639925e-05, "loss": 0.057, "step": 12489 }, { "epoch": 1.3540763226366002, "grad_norm": 1.2375566959381104, "learning_rate": 1.1012621500072537e-05, "loss": 0.0149, "step": 12490 }, { "epoch": 1.35418473547268, "grad_norm": 1.1536641120910645, "learning_rate": 1.1011896126505151e-05, "loss": 0.0298, "step": 12491 }, { "epoch": 1.3542931483087597, "grad_norm": 1.5757173299789429, "learning_rate": 1.1011170752937763e-05, "loss": 0.0128, "step": 12492 }, { "epoch": 1.3544015611448397, "grad_norm": 1.5986852645874023, "learning_rate": 1.1010445379370375e-05, "loss": 0.0715, "step": 12493 }, { "epoch": 1.3545099739809192, "grad_norm": 0.6243651509284973, "learning_rate": 1.100972000580299e-05, "loss": 0.0081, "step": 12494 }, { "epoch": 1.3546183868169992, "grad_norm": 1.6524266004562378, "learning_rate": 1.1008994632235601e-05, "loss": 0.0026, "step": 12495 }, { "epoch": 1.354726799653079, "grad_norm": 5.683779716491699, "learning_rate": 1.1008269258668217e-05, "loss": 0.0328, "step": 12496 }, { "epoch": 1.3548352124891587, "grad_norm": 1.1610031127929688, "learning_rate": 1.1007543885100829e-05, "loss": 0.0169, "step": 12497 }, { "epoch": 1.3549436253252385, "grad_norm": 5.252737045288086, "learning_rate": 1.1006818511533441e-05, "loss": 0.0573, "step": 12498 }, { "epoch": 1.3550520381613183, "grad_norm": 0.3145187795162201, "learning_rate": 1.1006093137966053e-05, "loss": 0.0026, "step": 12499 }, { "epoch": 1.355160450997398, "grad_norm": 2.1971864700317383, "learning_rate": 1.1005367764398667e-05, "loss": 0.0334, "step": 12500 }, { "epoch": 1.3552688638334778, "grad_norm": 2.20589542388916, "learning_rate": 1.100464239083128e-05, "loss": 0.0267, "step": 12501 }, { "epoch": 1.3553772766695578, "grad_norm": 0.22547408938407898, "learning_rate": 1.1003917017263891e-05, "loss": 0.0021, "step": 12502 }, { "epoch": 1.3554856895056375, "grad_norm": 0.9695308208465576, "learning_rate": 1.1003191643696505e-05, "loss": 0.0131, "step": 12503 }, { "epoch": 1.3555941023417173, "grad_norm": 5.244540691375732, "learning_rate": 1.1002466270129117e-05, "loss": 0.0281, "step": 12504 }, { "epoch": 1.355702515177797, "grad_norm": 0.6634459495544434, "learning_rate": 1.100174089656173e-05, "loss": 0.0111, "step": 12505 }, { "epoch": 1.3558109280138768, "grad_norm": 0.6900047063827515, "learning_rate": 1.1001015522994343e-05, "loss": 0.0069, "step": 12506 }, { "epoch": 1.3559193408499566, "grad_norm": 0.960128664970398, "learning_rate": 1.1000290149426956e-05, "loss": 0.0207, "step": 12507 }, { "epoch": 1.3560277536860363, "grad_norm": 3.1237456798553467, "learning_rate": 1.0999564775859568e-05, "loss": 0.0188, "step": 12508 }, { "epoch": 1.3561361665221163, "grad_norm": 1.196207880973816, "learning_rate": 1.0998839402292182e-05, "loss": 0.0103, "step": 12509 }, { "epoch": 1.356244579358196, "grad_norm": 0.400212824344635, "learning_rate": 1.0998114028724794e-05, "loss": 0.0035, "step": 12510 }, { "epoch": 1.3563529921942759, "grad_norm": 7.199605464935303, "learning_rate": 1.0997388655157406e-05, "loss": 0.0341, "step": 12511 }, { "epoch": 1.3564614050303556, "grad_norm": 2.0632166862487793, "learning_rate": 1.0996663281590018e-05, "loss": 0.0143, "step": 12512 }, { "epoch": 1.3565698178664354, "grad_norm": 2.8845996856689453, "learning_rate": 1.0995937908022633e-05, "loss": 0.0418, "step": 12513 }, { "epoch": 1.3566782307025151, "grad_norm": 1.9822007417678833, "learning_rate": 1.0995212534455246e-05, "loss": 0.0387, "step": 12514 }, { "epoch": 1.356786643538595, "grad_norm": 0.6998735666275024, "learning_rate": 1.099448716088786e-05, "loss": 0.0059, "step": 12515 }, { "epoch": 1.356895056374675, "grad_norm": 1.4780880212783813, "learning_rate": 1.0993761787320472e-05, "loss": 0.0114, "step": 12516 }, { "epoch": 1.3570034692107544, "grad_norm": 2.903836488723755, "learning_rate": 1.0993036413753084e-05, "loss": 0.0563, "step": 12517 }, { "epoch": 1.3571118820468344, "grad_norm": 2.5587756633758545, "learning_rate": 1.0992311040185698e-05, "loss": 0.0343, "step": 12518 }, { "epoch": 1.3572202948829142, "grad_norm": 4.389158248901367, "learning_rate": 1.099158566661831e-05, "loss": 0.0153, "step": 12519 }, { "epoch": 1.357328707718994, "grad_norm": 0.08797124773263931, "learning_rate": 1.0990860293050922e-05, "loss": 0.0008, "step": 12520 }, { "epoch": 1.3574371205550737, "grad_norm": 1.739242434501648, "learning_rate": 1.0990134919483536e-05, "loss": 0.0083, "step": 12521 }, { "epoch": 1.3575455333911535, "grad_norm": 2.2375893592834473, "learning_rate": 1.0989409545916148e-05, "loss": 0.0163, "step": 12522 }, { "epoch": 1.3576539462272332, "grad_norm": 1.1894900798797607, "learning_rate": 1.098868417234876e-05, "loss": 0.0072, "step": 12523 }, { "epoch": 1.357762359063313, "grad_norm": 0.7792953848838806, "learning_rate": 1.0987958798781372e-05, "loss": 0.0189, "step": 12524 }, { "epoch": 1.357870771899393, "grad_norm": 1.3763991594314575, "learning_rate": 1.0987233425213986e-05, "loss": 0.0044, "step": 12525 }, { "epoch": 1.3579791847354727, "grad_norm": 4.586109161376953, "learning_rate": 1.0986508051646598e-05, "loss": 0.0325, "step": 12526 }, { "epoch": 1.3580875975715525, "grad_norm": 0.10724648833274841, "learning_rate": 1.098578267807921e-05, "loss": 0.0011, "step": 12527 }, { "epoch": 1.3581960104076323, "grad_norm": 0.38203349709510803, "learning_rate": 1.0985057304511824e-05, "loss": 0.0025, "step": 12528 }, { "epoch": 1.358304423243712, "grad_norm": 0.47840237617492676, "learning_rate": 1.0984331930944436e-05, "loss": 0.0134, "step": 12529 }, { "epoch": 1.3584128360797918, "grad_norm": 1.340746283531189, "learning_rate": 1.0983606557377052e-05, "loss": 0.009, "step": 12530 }, { "epoch": 1.3585212489158716, "grad_norm": 5.693720817565918, "learning_rate": 1.0982881183809664e-05, "loss": 0.021, "step": 12531 }, { "epoch": 1.3586296617519515, "grad_norm": 2.8531885147094727, "learning_rate": 1.0982155810242276e-05, "loss": 0.0362, "step": 12532 }, { "epoch": 1.358738074588031, "grad_norm": 0.4498688280582428, "learning_rate": 1.098143043667489e-05, "loss": 0.0166, "step": 12533 }, { "epoch": 1.358846487424111, "grad_norm": 0.23520812392234802, "learning_rate": 1.0980705063107502e-05, "loss": 0.0018, "step": 12534 }, { "epoch": 1.3589549002601908, "grad_norm": 1.5517874956130981, "learning_rate": 1.0979979689540114e-05, "loss": 0.0135, "step": 12535 }, { "epoch": 1.3590633130962706, "grad_norm": 0.6341421008110046, "learning_rate": 1.0979254315972726e-05, "loss": 0.0127, "step": 12536 }, { "epoch": 1.3591717259323504, "grad_norm": 0.2789842188358307, "learning_rate": 1.097852894240534e-05, "loss": 0.0022, "step": 12537 }, { "epoch": 1.3592801387684301, "grad_norm": 2.358745813369751, "learning_rate": 1.0977803568837952e-05, "loss": 0.0154, "step": 12538 }, { "epoch": 1.35938855160451, "grad_norm": 1.7442682981491089, "learning_rate": 1.0977078195270564e-05, "loss": 0.0106, "step": 12539 }, { "epoch": 1.3594969644405897, "grad_norm": 2.009369373321533, "learning_rate": 1.0976352821703178e-05, "loss": 0.0136, "step": 12540 }, { "epoch": 1.3596053772766696, "grad_norm": 0.8091204166412354, "learning_rate": 1.097562744813579e-05, "loss": 0.0048, "step": 12541 }, { "epoch": 1.3597137901127494, "grad_norm": 4.170993804931641, "learning_rate": 1.0974902074568403e-05, "loss": 0.0278, "step": 12542 }, { "epoch": 1.3598222029488292, "grad_norm": 2.4520084857940674, "learning_rate": 1.0974176701001016e-05, "loss": 0.021, "step": 12543 }, { "epoch": 1.359930615784909, "grad_norm": 2.1315534114837646, "learning_rate": 1.0973451327433629e-05, "loss": 0.0478, "step": 12544 }, { "epoch": 1.3600390286209887, "grad_norm": 0.3223886489868164, "learning_rate": 1.097272595386624e-05, "loss": 0.0107, "step": 12545 }, { "epoch": 1.3601474414570685, "grad_norm": 1.6632390022277832, "learning_rate": 1.0972000580298855e-05, "loss": 0.0135, "step": 12546 }, { "epoch": 1.3602558542931482, "grad_norm": 0.5789124369621277, "learning_rate": 1.0971275206731467e-05, "loss": 0.0175, "step": 12547 }, { "epoch": 1.3603642671292282, "grad_norm": 0.04865679889917374, "learning_rate": 1.0970549833164082e-05, "loss": 0.0003, "step": 12548 }, { "epoch": 1.360472679965308, "grad_norm": 0.5975942611694336, "learning_rate": 1.0969824459596694e-05, "loss": 0.031, "step": 12549 }, { "epoch": 1.3605810928013877, "grad_norm": 1.6558889150619507, "learning_rate": 1.0969099086029306e-05, "loss": 0.0062, "step": 12550 }, { "epoch": 1.3606895056374675, "grad_norm": 0.04449913278222084, "learning_rate": 1.0968373712461919e-05, "loss": 0.0003, "step": 12551 }, { "epoch": 1.3607979184735473, "grad_norm": 4.914567470550537, "learning_rate": 1.0967648338894532e-05, "loss": 0.0247, "step": 12552 }, { "epoch": 1.360906331309627, "grad_norm": 3.3344671726226807, "learning_rate": 1.0966922965327145e-05, "loss": 0.0513, "step": 12553 }, { "epoch": 1.3610147441457068, "grad_norm": 0.375039279460907, "learning_rate": 1.0966197591759757e-05, "loss": 0.0059, "step": 12554 }, { "epoch": 1.3611231569817868, "grad_norm": 4.112647533416748, "learning_rate": 1.096547221819237e-05, "loss": 0.0311, "step": 12555 }, { "epoch": 1.3612315698178663, "grad_norm": 2.070923328399658, "learning_rate": 1.0964746844624983e-05, "loss": 0.0429, "step": 12556 }, { "epoch": 1.3613399826539463, "grad_norm": 0.5975987315177917, "learning_rate": 1.0964021471057595e-05, "loss": 0.0151, "step": 12557 }, { "epoch": 1.361448395490026, "grad_norm": 0.23313669860363007, "learning_rate": 1.0963296097490209e-05, "loss": 0.0006, "step": 12558 }, { "epoch": 1.3615568083261058, "grad_norm": 3.3284881114959717, "learning_rate": 1.0962570723922821e-05, "loss": 0.027, "step": 12559 }, { "epoch": 1.3616652211621856, "grad_norm": 0.6450234651565552, "learning_rate": 1.0961845350355433e-05, "loss": 0.0212, "step": 12560 }, { "epoch": 1.3617736339982653, "grad_norm": 2.2859082221984863, "learning_rate": 1.0961119976788047e-05, "loss": 0.0162, "step": 12561 }, { "epoch": 1.361882046834345, "grad_norm": 0.3276672959327698, "learning_rate": 1.0960394603220659e-05, "loss": 0.0019, "step": 12562 }, { "epoch": 1.3619904596704249, "grad_norm": 0.4843534529209137, "learning_rate": 1.0959669229653271e-05, "loss": 0.0282, "step": 12563 }, { "epoch": 1.3620988725065049, "grad_norm": 2.273861885070801, "learning_rate": 1.0958943856085883e-05, "loss": 0.0918, "step": 12564 }, { "epoch": 1.3622072853425846, "grad_norm": 0.804473876953125, "learning_rate": 1.0958218482518499e-05, "loss": 0.0144, "step": 12565 }, { "epoch": 1.3623156981786644, "grad_norm": 2.5903234481811523, "learning_rate": 1.0957493108951111e-05, "loss": 0.0386, "step": 12566 }, { "epoch": 1.3624241110147441, "grad_norm": 2.0874319076538086, "learning_rate": 1.0956767735383725e-05, "loss": 0.0334, "step": 12567 }, { "epoch": 1.362532523850824, "grad_norm": 4.811905860900879, "learning_rate": 1.0956042361816337e-05, "loss": 0.039, "step": 12568 }, { "epoch": 1.3626409366869037, "grad_norm": 4.033801078796387, "learning_rate": 1.0955316988248949e-05, "loss": 0.0365, "step": 12569 }, { "epoch": 1.3627493495229834, "grad_norm": 3.082409143447876, "learning_rate": 1.0954591614681563e-05, "loss": 0.0244, "step": 12570 }, { "epoch": 1.3628577623590634, "grad_norm": 3.2256903648376465, "learning_rate": 1.0953866241114175e-05, "loss": 0.0397, "step": 12571 }, { "epoch": 1.3629661751951432, "grad_norm": 0.08409981429576874, "learning_rate": 1.0953140867546787e-05, "loss": 0.0004, "step": 12572 }, { "epoch": 1.363074588031223, "grad_norm": 3.445040702819824, "learning_rate": 1.0952415493979401e-05, "loss": 0.0482, "step": 12573 }, { "epoch": 1.3631830008673027, "grad_norm": 3.8880510330200195, "learning_rate": 1.0951690120412013e-05, "loss": 0.0633, "step": 12574 }, { "epoch": 1.3632914137033825, "grad_norm": 1.527127742767334, "learning_rate": 1.0950964746844625e-05, "loss": 0.0183, "step": 12575 }, { "epoch": 1.3633998265394622, "grad_norm": 1.5581552982330322, "learning_rate": 1.0950239373277237e-05, "loss": 0.0151, "step": 12576 }, { "epoch": 1.363508239375542, "grad_norm": 0.6655043363571167, "learning_rate": 1.0949513999709851e-05, "loss": 0.0068, "step": 12577 }, { "epoch": 1.363616652211622, "grad_norm": 1.6065030097961426, "learning_rate": 1.0948788626142463e-05, "loss": 0.0338, "step": 12578 }, { "epoch": 1.3637250650477015, "grad_norm": 3.7087199687957764, "learning_rate": 1.0948063252575076e-05, "loss": 0.0123, "step": 12579 }, { "epoch": 1.3638334778837815, "grad_norm": 1.6959632635116577, "learning_rate": 1.094733787900769e-05, "loss": 0.0239, "step": 12580 }, { "epoch": 1.3639418907198613, "grad_norm": 3.4767022132873535, "learning_rate": 1.0946612505440302e-05, "loss": 0.0698, "step": 12581 }, { "epoch": 1.364050303555941, "grad_norm": 2.262803792953491, "learning_rate": 1.0945887131872917e-05, "loss": 0.0062, "step": 12582 }, { "epoch": 1.3641587163920208, "grad_norm": 1.5069847106933594, "learning_rate": 1.094516175830553e-05, "loss": 0.0204, "step": 12583 }, { "epoch": 1.3642671292281006, "grad_norm": 6.02863073348999, "learning_rate": 1.0944436384738141e-05, "loss": 0.0432, "step": 12584 }, { "epoch": 1.3643755420641803, "grad_norm": 3.530547857284546, "learning_rate": 1.0943711011170755e-05, "loss": 0.0237, "step": 12585 }, { "epoch": 1.36448395490026, "grad_norm": 2.9498255252838135, "learning_rate": 1.0942985637603367e-05, "loss": 0.0352, "step": 12586 }, { "epoch": 1.36459236773634, "grad_norm": 0.954502522945404, "learning_rate": 1.094226026403598e-05, "loss": 0.028, "step": 12587 }, { "epoch": 1.3647007805724198, "grad_norm": 0.7205872535705566, "learning_rate": 1.0941534890468593e-05, "loss": 0.0268, "step": 12588 }, { "epoch": 1.3648091934084996, "grad_norm": 2.3826711177825928, "learning_rate": 1.0940809516901206e-05, "loss": 0.0545, "step": 12589 }, { "epoch": 1.3649176062445794, "grad_norm": 7.947179317474365, "learning_rate": 1.0940084143333818e-05, "loss": 0.0471, "step": 12590 }, { "epoch": 1.3650260190806591, "grad_norm": 0.5611388087272644, "learning_rate": 1.093935876976643e-05, "loss": 0.0046, "step": 12591 }, { "epoch": 1.3651344319167389, "grad_norm": 0.5669108033180237, "learning_rate": 1.0938633396199044e-05, "loss": 0.0076, "step": 12592 }, { "epoch": 1.3652428447528187, "grad_norm": 6.439491271972656, "learning_rate": 1.0937908022631656e-05, "loss": 0.0498, "step": 12593 }, { "epoch": 1.3653512575888986, "grad_norm": 2.581895589828491, "learning_rate": 1.0937182649064268e-05, "loss": 0.0132, "step": 12594 }, { "epoch": 1.3654596704249784, "grad_norm": 1.6365028619766235, "learning_rate": 1.0936457275496882e-05, "loss": 0.0355, "step": 12595 }, { "epoch": 1.3655680832610582, "grad_norm": 0.28772053122520447, "learning_rate": 1.0935731901929494e-05, "loss": 0.0174, "step": 12596 }, { "epoch": 1.365676496097138, "grad_norm": 2.018413543701172, "learning_rate": 1.0935006528362106e-05, "loss": 0.0095, "step": 12597 }, { "epoch": 1.3657849089332177, "grad_norm": 0.5621086359024048, "learning_rate": 1.093428115479472e-05, "loss": 0.0156, "step": 12598 }, { "epoch": 1.3658933217692975, "grad_norm": 2.5747671127319336, "learning_rate": 1.0933555781227334e-05, "loss": 0.0185, "step": 12599 }, { "epoch": 1.3660017346053772, "grad_norm": 3.054666519165039, "learning_rate": 1.0932830407659948e-05, "loss": 0.0325, "step": 12600 }, { "epoch": 1.3661101474414572, "grad_norm": 4.4281005859375, "learning_rate": 1.093210503409256e-05, "loss": 0.0403, "step": 12601 }, { "epoch": 1.3662185602775367, "grad_norm": 2.310636043548584, "learning_rate": 1.0931379660525172e-05, "loss": 0.0237, "step": 12602 }, { "epoch": 1.3663269731136167, "grad_norm": 0.878125786781311, "learning_rate": 1.0930654286957784e-05, "loss": 0.0161, "step": 12603 }, { "epoch": 1.3664353859496965, "grad_norm": 0.8566964864730835, "learning_rate": 1.0929928913390398e-05, "loss": 0.0094, "step": 12604 }, { "epoch": 1.3665437987857763, "grad_norm": 0.4879777431488037, "learning_rate": 1.092920353982301e-05, "loss": 0.0024, "step": 12605 }, { "epoch": 1.366652211621856, "grad_norm": 5.738369464874268, "learning_rate": 1.0928478166255622e-05, "loss": 0.0203, "step": 12606 }, { "epoch": 1.3667606244579358, "grad_norm": 10.283358573913574, "learning_rate": 1.0927752792688236e-05, "loss": 0.0247, "step": 12607 }, { "epoch": 1.3668690372940155, "grad_norm": 1.4032539129257202, "learning_rate": 1.0927027419120848e-05, "loss": 0.0213, "step": 12608 }, { "epoch": 1.3669774501300953, "grad_norm": 3.924254894256592, "learning_rate": 1.092630204555346e-05, "loss": 0.007, "step": 12609 }, { "epoch": 1.3670858629661753, "grad_norm": 2.4000301361083984, "learning_rate": 1.0925576671986074e-05, "loss": 0.065, "step": 12610 }, { "epoch": 1.367194275802255, "grad_norm": 3.3634276390075684, "learning_rate": 1.0924851298418686e-05, "loss": 0.0256, "step": 12611 }, { "epoch": 1.3673026886383348, "grad_norm": 0.7449190020561218, "learning_rate": 1.0924125924851298e-05, "loss": 0.0075, "step": 12612 }, { "epoch": 1.3674111014744146, "grad_norm": 1.0063140392303467, "learning_rate": 1.0923400551283912e-05, "loss": 0.0032, "step": 12613 }, { "epoch": 1.3675195143104943, "grad_norm": 2.0047106742858887, "learning_rate": 1.0922675177716524e-05, "loss": 0.0046, "step": 12614 }, { "epoch": 1.367627927146574, "grad_norm": 0.896315336227417, "learning_rate": 1.0921949804149136e-05, "loss": 0.0153, "step": 12615 }, { "epoch": 1.3677363399826539, "grad_norm": 0.11084898561239243, "learning_rate": 1.0921224430581749e-05, "loss": 0.0009, "step": 12616 }, { "epoch": 1.3678447528187339, "grad_norm": 1.4431798458099365, "learning_rate": 1.0920499057014364e-05, "loss": 0.0273, "step": 12617 }, { "epoch": 1.3679531656548134, "grad_norm": 0.13251030445098877, "learning_rate": 1.0919773683446976e-05, "loss": 0.0011, "step": 12618 }, { "epoch": 1.3680615784908934, "grad_norm": 5.30898380279541, "learning_rate": 1.091904830987959e-05, "loss": 0.0405, "step": 12619 }, { "epoch": 1.3681699913269731, "grad_norm": 0.2020263671875, "learning_rate": 1.0918322936312202e-05, "loss": 0.014, "step": 12620 }, { "epoch": 1.368278404163053, "grad_norm": 8.108973503112793, "learning_rate": 1.0917597562744814e-05, "loss": 0.0434, "step": 12621 }, { "epoch": 1.3683868169991327, "grad_norm": 1.6626625061035156, "learning_rate": 1.0916872189177428e-05, "loss": 0.0321, "step": 12622 }, { "epoch": 1.3684952298352124, "grad_norm": 0.6866579651832581, "learning_rate": 1.091614681561004e-05, "loss": 0.0078, "step": 12623 }, { "epoch": 1.3686036426712924, "grad_norm": 3.5356967449188232, "learning_rate": 1.0915421442042653e-05, "loss": 0.034, "step": 12624 }, { "epoch": 1.368712055507372, "grad_norm": 0.8574448823928833, "learning_rate": 1.0914696068475266e-05, "loss": 0.0076, "step": 12625 }, { "epoch": 1.368820468343452, "grad_norm": 4.014377117156982, "learning_rate": 1.0913970694907879e-05, "loss": 0.0379, "step": 12626 }, { "epoch": 1.3689288811795317, "grad_norm": 0.7876272201538086, "learning_rate": 1.091324532134049e-05, "loss": 0.0095, "step": 12627 }, { "epoch": 1.3690372940156115, "grad_norm": 0.4422430098056793, "learning_rate": 1.0912519947773103e-05, "loss": 0.0075, "step": 12628 }, { "epoch": 1.3691457068516912, "grad_norm": 0.6382347941398621, "learning_rate": 1.0911794574205717e-05, "loss": 0.0202, "step": 12629 }, { "epoch": 1.369254119687771, "grad_norm": 1.1293126344680786, "learning_rate": 1.0911069200638329e-05, "loss": 0.0081, "step": 12630 }, { "epoch": 1.3693625325238508, "grad_norm": 4.460083484649658, "learning_rate": 1.0910343827070941e-05, "loss": 0.0267, "step": 12631 }, { "epoch": 1.3694709453599305, "grad_norm": 0.16878139972686768, "learning_rate": 1.0909618453503555e-05, "loss": 0.0007, "step": 12632 }, { "epoch": 1.3695793581960105, "grad_norm": 0.8399560451507568, "learning_rate": 1.0908893079936167e-05, "loss": 0.0222, "step": 12633 }, { "epoch": 1.3696877710320903, "grad_norm": 1.7371610403060913, "learning_rate": 1.0908167706368782e-05, "loss": 0.0312, "step": 12634 }, { "epoch": 1.36979618386817, "grad_norm": 2.7725555896759033, "learning_rate": 1.0907442332801395e-05, "loss": 0.061, "step": 12635 }, { "epoch": 1.3699045967042498, "grad_norm": 0.443777859210968, "learning_rate": 1.0906716959234007e-05, "loss": 0.0215, "step": 12636 }, { "epoch": 1.3700130095403296, "grad_norm": 0.30106690526008606, "learning_rate": 1.090599158566662e-05, "loss": 0.0091, "step": 12637 }, { "epoch": 1.3701214223764093, "grad_norm": 5.277652263641357, "learning_rate": 1.0905266212099233e-05, "loss": 0.0507, "step": 12638 }, { "epoch": 1.370229835212489, "grad_norm": 3.915916681289673, "learning_rate": 1.0904540838531845e-05, "loss": 0.0521, "step": 12639 }, { "epoch": 1.370338248048569, "grad_norm": 2.0238308906555176, "learning_rate": 1.0903815464964459e-05, "loss": 0.03, "step": 12640 }, { "epoch": 1.3704466608846486, "grad_norm": 3.0500032901763916, "learning_rate": 1.090309009139707e-05, "loss": 0.0368, "step": 12641 }, { "epoch": 1.3705550737207286, "grad_norm": 5.842222690582275, "learning_rate": 1.0902364717829683e-05, "loss": 0.0232, "step": 12642 }, { "epoch": 1.3706634865568084, "grad_norm": 4.9470133781433105, "learning_rate": 1.0901639344262295e-05, "loss": 0.0588, "step": 12643 }, { "epoch": 1.3707718993928881, "grad_norm": 1.8673983812332153, "learning_rate": 1.0900913970694909e-05, "loss": 0.0065, "step": 12644 }, { "epoch": 1.370880312228968, "grad_norm": 1.1876764297485352, "learning_rate": 1.0900188597127521e-05, "loss": 0.0247, "step": 12645 }, { "epoch": 1.3709887250650477, "grad_norm": 4.299071311950684, "learning_rate": 1.0899463223560133e-05, "loss": 0.0243, "step": 12646 }, { "epoch": 1.3710971379011274, "grad_norm": 1.0262553691864014, "learning_rate": 1.0898737849992747e-05, "loss": 0.01, "step": 12647 }, { "epoch": 1.3712055507372072, "grad_norm": 0.6284225583076477, "learning_rate": 1.089801247642536e-05, "loss": 0.0079, "step": 12648 }, { "epoch": 1.3713139635732872, "grad_norm": 0.2214677929878235, "learning_rate": 1.0897287102857971e-05, "loss": 0.0033, "step": 12649 }, { "epoch": 1.371422376409367, "grad_norm": 0.2919841408729553, "learning_rate": 1.0896561729290585e-05, "loss": 0.003, "step": 12650 }, { "epoch": 1.3715307892454467, "grad_norm": 2.118349552154541, "learning_rate": 1.0895836355723199e-05, "loss": 0.0113, "step": 12651 }, { "epoch": 1.3716392020815265, "grad_norm": 4.523299694061279, "learning_rate": 1.0895110982155813e-05, "loss": 0.0297, "step": 12652 }, { "epoch": 1.3717476149176062, "grad_norm": 0.4847501218318939, "learning_rate": 1.0894385608588425e-05, "loss": 0.0149, "step": 12653 }, { "epoch": 1.371856027753686, "grad_norm": 1.7706191539764404, "learning_rate": 1.0893660235021037e-05, "loss": 0.0337, "step": 12654 }, { "epoch": 1.3719644405897657, "grad_norm": 6.989010810852051, "learning_rate": 1.089293486145365e-05, "loss": 0.0298, "step": 12655 }, { "epoch": 1.3720728534258457, "grad_norm": 2.3151695728302, "learning_rate": 1.0892209487886263e-05, "loss": 0.0455, "step": 12656 }, { "epoch": 1.3721812662619255, "grad_norm": 0.8025867938995361, "learning_rate": 1.0891484114318875e-05, "loss": 0.0227, "step": 12657 }, { "epoch": 1.3722896790980053, "grad_norm": 3.1162490844726562, "learning_rate": 1.0890758740751487e-05, "loss": 0.0567, "step": 12658 }, { "epoch": 1.372398091934085, "grad_norm": 2.394912004470825, "learning_rate": 1.0890033367184101e-05, "loss": 0.0355, "step": 12659 }, { "epoch": 1.3725065047701648, "grad_norm": 0.5751938223838806, "learning_rate": 1.0889307993616713e-05, "loss": 0.0263, "step": 12660 }, { "epoch": 1.3726149176062445, "grad_norm": 0.8468310236930847, "learning_rate": 1.0888582620049326e-05, "loss": 0.0168, "step": 12661 }, { "epoch": 1.3727233304423243, "grad_norm": 6.646358489990234, "learning_rate": 1.088785724648194e-05, "loss": 0.0434, "step": 12662 }, { "epoch": 1.3728317432784043, "grad_norm": 8.68865966796875, "learning_rate": 1.0887131872914552e-05, "loss": 0.0368, "step": 12663 }, { "epoch": 1.3729401561144838, "grad_norm": 2.3097341060638428, "learning_rate": 1.0886406499347164e-05, "loss": 0.0538, "step": 12664 }, { "epoch": 1.3730485689505638, "grad_norm": 1.2288520336151123, "learning_rate": 1.0885681125779778e-05, "loss": 0.0063, "step": 12665 }, { "epoch": 1.3731569817866436, "grad_norm": 4.571734428405762, "learning_rate": 1.088495575221239e-05, "loss": 0.0137, "step": 12666 }, { "epoch": 1.3732653946227233, "grad_norm": 0.9514390826225281, "learning_rate": 1.0884230378645002e-05, "loss": 0.0126, "step": 12667 }, { "epoch": 1.373373807458803, "grad_norm": 2.9897406101226807, "learning_rate": 1.0883505005077614e-05, "loss": 0.012, "step": 12668 }, { "epoch": 1.3734822202948829, "grad_norm": 2.0455682277679443, "learning_rate": 1.088277963151023e-05, "loss": 0.0473, "step": 12669 }, { "epoch": 1.3735906331309626, "grad_norm": 2.0000572204589844, "learning_rate": 1.0882054257942842e-05, "loss": 0.0166, "step": 12670 }, { "epoch": 1.3736990459670424, "grad_norm": 0.026887882500886917, "learning_rate": 1.0881328884375455e-05, "loss": 0.0002, "step": 12671 }, { "epoch": 1.3738074588031224, "grad_norm": 2.082444429397583, "learning_rate": 1.0880603510808068e-05, "loss": 0.0422, "step": 12672 }, { "epoch": 1.3739158716392021, "grad_norm": 1.7146155834197998, "learning_rate": 1.087987813724068e-05, "loss": 0.0142, "step": 12673 }, { "epoch": 1.374024284475282, "grad_norm": 0.4078371822834015, "learning_rate": 1.0879152763673294e-05, "loss": 0.0008, "step": 12674 }, { "epoch": 1.3741326973113617, "grad_norm": 3.5558812618255615, "learning_rate": 1.0878427390105906e-05, "loss": 0.0158, "step": 12675 }, { "epoch": 1.3742411101474414, "grad_norm": 0.9008092284202576, "learning_rate": 1.0877702016538518e-05, "loss": 0.0236, "step": 12676 }, { "epoch": 1.3743495229835212, "grad_norm": 2.097227096557617, "learning_rate": 1.0876976642971132e-05, "loss": 0.015, "step": 12677 }, { "epoch": 1.374457935819601, "grad_norm": 3.5837225914001465, "learning_rate": 1.0876251269403744e-05, "loss": 0.0227, "step": 12678 }, { "epoch": 1.374566348655681, "grad_norm": 0.08112526684999466, "learning_rate": 1.0875525895836356e-05, "loss": 0.0007, "step": 12679 }, { "epoch": 1.3746747614917605, "grad_norm": 1.4249849319458008, "learning_rate": 1.0874800522268968e-05, "loss": 0.0151, "step": 12680 }, { "epoch": 1.3747831743278405, "grad_norm": 6.939162254333496, "learning_rate": 1.0874075148701582e-05, "loss": 0.0281, "step": 12681 }, { "epoch": 1.3748915871639202, "grad_norm": 2.2147083282470703, "learning_rate": 1.0873349775134194e-05, "loss": 0.0229, "step": 12682 }, { "epoch": 1.375, "grad_norm": 1.9473345279693604, "learning_rate": 1.0872624401566806e-05, "loss": 0.036, "step": 12683 }, { "epoch": 1.3751084128360798, "grad_norm": 0.574220597743988, "learning_rate": 1.087189902799942e-05, "loss": 0.0093, "step": 12684 }, { "epoch": 1.3752168256721595, "grad_norm": 3.6630358695983887, "learning_rate": 1.0871173654432032e-05, "loss": 0.0372, "step": 12685 }, { "epoch": 1.3753252385082395, "grad_norm": 1.453598976135254, "learning_rate": 1.0870448280864648e-05, "loss": 0.0136, "step": 12686 }, { "epoch": 1.375433651344319, "grad_norm": 1.893458604812622, "learning_rate": 1.086972290729726e-05, "loss": 0.0171, "step": 12687 }, { "epoch": 1.375542064180399, "grad_norm": 4.341336250305176, "learning_rate": 1.0868997533729872e-05, "loss": 0.0588, "step": 12688 }, { "epoch": 1.3756504770164788, "grad_norm": 17.292762756347656, "learning_rate": 1.0868272160162486e-05, "loss": 0.0244, "step": 12689 }, { "epoch": 1.3757588898525586, "grad_norm": 6.975484848022461, "learning_rate": 1.0867546786595098e-05, "loss": 0.0275, "step": 12690 }, { "epoch": 1.3758673026886383, "grad_norm": 10.186700820922852, "learning_rate": 1.086682141302771e-05, "loss": 0.044, "step": 12691 }, { "epoch": 1.375975715524718, "grad_norm": 3.186189651489258, "learning_rate": 1.0866096039460324e-05, "loss": 0.0325, "step": 12692 }, { "epoch": 1.3760841283607979, "grad_norm": 0.4046144485473633, "learning_rate": 1.0865370665892936e-05, "loss": 0.0129, "step": 12693 }, { "epoch": 1.3761925411968776, "grad_norm": 3.7195823192596436, "learning_rate": 1.0864645292325548e-05, "loss": 0.0556, "step": 12694 }, { "epoch": 1.3763009540329576, "grad_norm": 1.7657136917114258, "learning_rate": 1.086391991875816e-05, "loss": 0.0211, "step": 12695 }, { "epoch": 1.3764093668690374, "grad_norm": 1.9606596231460571, "learning_rate": 1.0863194545190774e-05, "loss": 0.0343, "step": 12696 }, { "epoch": 1.3765177797051171, "grad_norm": 1.4132051467895508, "learning_rate": 1.0862469171623386e-05, "loss": 0.0342, "step": 12697 }, { "epoch": 1.376626192541197, "grad_norm": 0.2301432490348816, "learning_rate": 1.0861743798055999e-05, "loss": 0.001, "step": 12698 }, { "epoch": 1.3767346053772767, "grad_norm": 1.834007978439331, "learning_rate": 1.0861018424488612e-05, "loss": 0.0179, "step": 12699 }, { "epoch": 1.3768430182133564, "grad_norm": 0.3513193130493164, "learning_rate": 1.0860293050921225e-05, "loss": 0.0098, "step": 12700 }, { "epoch": 1.3769514310494362, "grad_norm": 0.20893238484859467, "learning_rate": 1.0859567677353837e-05, "loss": 0.0256, "step": 12701 }, { "epoch": 1.3770598438855162, "grad_norm": 0.3924466669559479, "learning_rate": 1.085884230378645e-05, "loss": 0.0016, "step": 12702 }, { "epoch": 1.3771682567215957, "grad_norm": 2.67893385887146, "learning_rate": 1.0858116930219064e-05, "loss": 0.0357, "step": 12703 }, { "epoch": 1.3772766695576757, "grad_norm": 7.401793956756592, "learning_rate": 1.0857391556651678e-05, "loss": 0.041, "step": 12704 }, { "epoch": 1.3773850823937555, "grad_norm": 0.2606763541698456, "learning_rate": 1.085666618308429e-05, "loss": 0.0207, "step": 12705 }, { "epoch": 1.3774934952298352, "grad_norm": 1.6749629974365234, "learning_rate": 1.0855940809516903e-05, "loss": 0.0134, "step": 12706 }, { "epoch": 1.377601908065915, "grad_norm": 5.2347588539123535, "learning_rate": 1.0855215435949515e-05, "loss": 0.0672, "step": 12707 }, { "epoch": 1.3777103209019947, "grad_norm": 0.391765296459198, "learning_rate": 1.0854490062382128e-05, "loss": 0.0029, "step": 12708 }, { "epoch": 1.3778187337380745, "grad_norm": 3.7545273303985596, "learning_rate": 1.085376468881474e-05, "loss": 0.0312, "step": 12709 }, { "epoch": 1.3779271465741543, "grad_norm": 2.218092203140259, "learning_rate": 1.0853039315247353e-05, "loss": 0.0187, "step": 12710 }, { "epoch": 1.3780355594102343, "grad_norm": 6.1622209548950195, "learning_rate": 1.0852313941679967e-05, "loss": 0.0757, "step": 12711 }, { "epoch": 1.378143972246314, "grad_norm": 1.8663418292999268, "learning_rate": 1.0851588568112579e-05, "loss": 0.0579, "step": 12712 }, { "epoch": 1.3782523850823938, "grad_norm": 0.8319501876831055, "learning_rate": 1.0850863194545191e-05, "loss": 0.0248, "step": 12713 }, { "epoch": 1.3783607979184735, "grad_norm": 0.5388902425765991, "learning_rate": 1.0850137820977805e-05, "loss": 0.0111, "step": 12714 }, { "epoch": 1.3784692107545533, "grad_norm": 0.05831567570567131, "learning_rate": 1.0849412447410417e-05, "loss": 0.0004, "step": 12715 }, { "epoch": 1.378577623590633, "grad_norm": 8.04116439819336, "learning_rate": 1.0848687073843029e-05, "loss": 0.0371, "step": 12716 }, { "epoch": 1.3786860364267128, "grad_norm": 1.1167354583740234, "learning_rate": 1.0847961700275643e-05, "loss": 0.0321, "step": 12717 }, { "epoch": 1.3787944492627928, "grad_norm": 4.038622856140137, "learning_rate": 1.0847236326708255e-05, "loss": 0.0472, "step": 12718 }, { "epoch": 1.3789028620988726, "grad_norm": 6.670390605926514, "learning_rate": 1.0846510953140867e-05, "loss": 0.0386, "step": 12719 }, { "epoch": 1.3790112749349523, "grad_norm": 1.392380952835083, "learning_rate": 1.0845785579573483e-05, "loss": 0.0219, "step": 12720 }, { "epoch": 1.379119687771032, "grad_norm": 2.9067516326904297, "learning_rate": 1.0845060206006095e-05, "loss": 0.0316, "step": 12721 }, { "epoch": 1.3792281006071119, "grad_norm": 0.5809608697891235, "learning_rate": 1.0844334832438707e-05, "loss": 0.0078, "step": 12722 }, { "epoch": 1.3793365134431916, "grad_norm": 5.0484514236450195, "learning_rate": 1.084360945887132e-05, "loss": 0.0341, "step": 12723 }, { "epoch": 1.3794449262792714, "grad_norm": 0.14266912639141083, "learning_rate": 1.0842884085303933e-05, "loss": 0.0007, "step": 12724 }, { "epoch": 1.3795533391153514, "grad_norm": 0.9926408529281616, "learning_rate": 1.0842158711736545e-05, "loss": 0.0077, "step": 12725 }, { "epoch": 1.379661751951431, "grad_norm": 2.0986671447753906, "learning_rate": 1.0841433338169159e-05, "loss": 0.0108, "step": 12726 }, { "epoch": 1.379770164787511, "grad_norm": 2.0372583866119385, "learning_rate": 1.0840707964601771e-05, "loss": 0.0219, "step": 12727 }, { "epoch": 1.3798785776235907, "grad_norm": 1.1499251127243042, "learning_rate": 1.0839982591034383e-05, "loss": 0.0218, "step": 12728 }, { "epoch": 1.3799869904596704, "grad_norm": 3.700928211212158, "learning_rate": 1.0839257217466997e-05, "loss": 0.0453, "step": 12729 }, { "epoch": 1.3800954032957502, "grad_norm": 4.368139266967773, "learning_rate": 1.083853184389961e-05, "loss": 0.0393, "step": 12730 }, { "epoch": 1.38020381613183, "grad_norm": 1.2371402978897095, "learning_rate": 1.0837806470332221e-05, "loss": 0.0223, "step": 12731 }, { "epoch": 1.3803122289679097, "grad_norm": 2.0161919593811035, "learning_rate": 1.0837081096764833e-05, "loss": 0.013, "step": 12732 }, { "epoch": 1.3804206418039895, "grad_norm": 0.2783471345901489, "learning_rate": 1.0836355723197447e-05, "loss": 0.0028, "step": 12733 }, { "epoch": 1.3805290546400695, "grad_norm": 1.5416648387908936, "learning_rate": 1.083563034963006e-05, "loss": 0.0302, "step": 12734 }, { "epoch": 1.3806374674761492, "grad_norm": 0.5779917240142822, "learning_rate": 1.0834904976062672e-05, "loss": 0.0135, "step": 12735 }, { "epoch": 1.380745880312229, "grad_norm": 8.057777404785156, "learning_rate": 1.0834179602495285e-05, "loss": 0.0097, "step": 12736 }, { "epoch": 1.3808542931483088, "grad_norm": 2.5039124488830566, "learning_rate": 1.0833454228927898e-05, "loss": 0.0066, "step": 12737 }, { "epoch": 1.3809627059843885, "grad_norm": 2.0883517265319824, "learning_rate": 1.0832728855360513e-05, "loss": 0.0611, "step": 12738 }, { "epoch": 1.3810711188204683, "grad_norm": 0.975925087928772, "learning_rate": 1.0832003481793125e-05, "loss": 0.0124, "step": 12739 }, { "epoch": 1.381179531656548, "grad_norm": 3.2893929481506348, "learning_rate": 1.0831278108225737e-05, "loss": 0.0491, "step": 12740 }, { "epoch": 1.381287944492628, "grad_norm": 2.402061700820923, "learning_rate": 1.0830552734658351e-05, "loss": 0.0323, "step": 12741 }, { "epoch": 1.3813963573287076, "grad_norm": 0.7932988405227661, "learning_rate": 1.0829827361090963e-05, "loss": 0.0155, "step": 12742 }, { "epoch": 1.3815047701647876, "grad_norm": 0.340804785490036, "learning_rate": 1.0829101987523576e-05, "loss": 0.0072, "step": 12743 }, { "epoch": 1.3816131830008673, "grad_norm": 1.4857598543167114, "learning_rate": 1.082837661395619e-05, "loss": 0.0291, "step": 12744 }, { "epoch": 1.381721595836947, "grad_norm": 1.3910218477249146, "learning_rate": 1.0827651240388802e-05, "loss": 0.0257, "step": 12745 }, { "epoch": 1.3818300086730269, "grad_norm": 8.992488861083984, "learning_rate": 1.0826925866821414e-05, "loss": 0.0136, "step": 12746 }, { "epoch": 1.3819384215091066, "grad_norm": 0.22828896343708038, "learning_rate": 1.0826200493254026e-05, "loss": 0.0017, "step": 12747 }, { "epoch": 1.3820468343451866, "grad_norm": 2.4871137142181396, "learning_rate": 1.082547511968664e-05, "loss": 0.0337, "step": 12748 }, { "epoch": 1.3821552471812661, "grad_norm": 1.010132074356079, "learning_rate": 1.0824749746119252e-05, "loss": 0.0153, "step": 12749 }, { "epoch": 1.3822636600173461, "grad_norm": 4.598804473876953, "learning_rate": 1.0824024372551864e-05, "loss": 0.0273, "step": 12750 }, { "epoch": 1.382372072853426, "grad_norm": 0.4025903046131134, "learning_rate": 1.0823298998984478e-05, "loss": 0.0126, "step": 12751 }, { "epoch": 1.3824804856895057, "grad_norm": 0.4009324014186859, "learning_rate": 1.082257362541709e-05, "loss": 0.0036, "step": 12752 }, { "epoch": 1.3825888985255854, "grad_norm": 4.9647369384765625, "learning_rate": 1.0821848251849702e-05, "loss": 0.0292, "step": 12753 }, { "epoch": 1.3826973113616652, "grad_norm": 1.3205173015594482, "learning_rate": 1.0821122878282316e-05, "loss": 0.0194, "step": 12754 }, { "epoch": 1.382805724197745, "grad_norm": 1.057423710823059, "learning_rate": 1.082039750471493e-05, "loss": 0.0252, "step": 12755 }, { "epoch": 1.3829141370338247, "grad_norm": 7.172843933105469, "learning_rate": 1.0819672131147544e-05, "loss": 0.0777, "step": 12756 }, { "epoch": 1.3830225498699047, "grad_norm": 6.237669467926025, "learning_rate": 1.0818946757580156e-05, "loss": 0.0574, "step": 12757 }, { "epoch": 1.3831309627059845, "grad_norm": 0.6706156730651855, "learning_rate": 1.0818221384012768e-05, "loss": 0.0183, "step": 12758 }, { "epoch": 1.3832393755420642, "grad_norm": 3.4390761852264404, "learning_rate": 1.081749601044538e-05, "loss": 0.0831, "step": 12759 }, { "epoch": 1.383347788378144, "grad_norm": 3.666515827178955, "learning_rate": 1.0816770636877994e-05, "loss": 0.0121, "step": 12760 }, { "epoch": 1.3834562012142237, "grad_norm": 0.7538495659828186, "learning_rate": 1.0816045263310606e-05, "loss": 0.0114, "step": 12761 }, { "epoch": 1.3835646140503035, "grad_norm": 2.663355827331543, "learning_rate": 1.0815319889743218e-05, "loss": 0.054, "step": 12762 }, { "epoch": 1.3836730268863833, "grad_norm": 0.3232938051223755, "learning_rate": 1.0814594516175832e-05, "loss": 0.0028, "step": 12763 }, { "epoch": 1.3837814397224633, "grad_norm": 5.248039722442627, "learning_rate": 1.0813869142608444e-05, "loss": 0.0177, "step": 12764 }, { "epoch": 1.3838898525585428, "grad_norm": 1.0627518892288208, "learning_rate": 1.0813143769041056e-05, "loss": 0.0153, "step": 12765 }, { "epoch": 1.3839982653946228, "grad_norm": 3.800133466720581, "learning_rate": 1.081241839547367e-05, "loss": 0.0525, "step": 12766 }, { "epoch": 1.3841066782307025, "grad_norm": 1.116392970085144, "learning_rate": 1.0811693021906282e-05, "loss": 0.0089, "step": 12767 }, { "epoch": 1.3842150910667823, "grad_norm": 1.3594053983688354, "learning_rate": 1.0810967648338894e-05, "loss": 0.0165, "step": 12768 }, { "epoch": 1.384323503902862, "grad_norm": 1.6399577856063843, "learning_rate": 1.0810242274771508e-05, "loss": 0.0078, "step": 12769 }, { "epoch": 1.3844319167389418, "grad_norm": 7.299533367156982, "learning_rate": 1.080951690120412e-05, "loss": 0.0493, "step": 12770 }, { "epoch": 1.3845403295750216, "grad_norm": 2.9593405723571777, "learning_rate": 1.0808791527636732e-05, "loss": 0.0464, "step": 12771 }, { "epoch": 1.3846487424111014, "grad_norm": 1.8149951696395874, "learning_rate": 1.0808066154069348e-05, "loss": 0.0195, "step": 12772 }, { "epoch": 1.3847571552471813, "grad_norm": 9.424322128295898, "learning_rate": 1.080734078050196e-05, "loss": 0.0407, "step": 12773 }, { "epoch": 1.3848655680832611, "grad_norm": 1.1673214435577393, "learning_rate": 1.0806615406934572e-05, "loss": 0.0128, "step": 12774 }, { "epoch": 1.3849739809193409, "grad_norm": 2.0708725452423096, "learning_rate": 1.0805890033367186e-05, "loss": 0.0253, "step": 12775 }, { "epoch": 1.3850823937554206, "grad_norm": 0.7574790716171265, "learning_rate": 1.0805164659799798e-05, "loss": 0.0218, "step": 12776 }, { "epoch": 1.3851908065915004, "grad_norm": 1.3009800910949707, "learning_rate": 1.080443928623241e-05, "loss": 0.0238, "step": 12777 }, { "epoch": 1.3852992194275802, "grad_norm": 3.3177437782287598, "learning_rate": 1.0803713912665024e-05, "loss": 0.0135, "step": 12778 }, { "epoch": 1.38540763226366, "grad_norm": 2.2821478843688965, "learning_rate": 1.0802988539097636e-05, "loss": 0.0304, "step": 12779 }, { "epoch": 1.38551604509974, "grad_norm": 0.4437834918498993, "learning_rate": 1.0802263165530249e-05, "loss": 0.0137, "step": 12780 }, { "epoch": 1.3856244579358197, "grad_norm": 1.3345627784729004, "learning_rate": 1.0801537791962862e-05, "loss": 0.0289, "step": 12781 }, { "epoch": 1.3857328707718994, "grad_norm": 0.4897924065589905, "learning_rate": 1.0800812418395475e-05, "loss": 0.0189, "step": 12782 }, { "epoch": 1.3858412836079792, "grad_norm": 2.384711742401123, "learning_rate": 1.0800087044828087e-05, "loss": 0.0181, "step": 12783 }, { "epoch": 1.385949696444059, "grad_norm": 2.284052848815918, "learning_rate": 1.07993616712607e-05, "loss": 0.0166, "step": 12784 }, { "epoch": 1.3860581092801387, "grad_norm": 1.3309494256973267, "learning_rate": 1.0798636297693313e-05, "loss": 0.0349, "step": 12785 }, { "epoch": 1.3861665221162185, "grad_norm": 1.0424175262451172, "learning_rate": 1.0797910924125925e-05, "loss": 0.015, "step": 12786 }, { "epoch": 1.3862749349522985, "grad_norm": 0.4914921522140503, "learning_rate": 1.0797185550558537e-05, "loss": 0.0189, "step": 12787 }, { "epoch": 1.386383347788378, "grad_norm": 0.19931307435035706, "learning_rate": 1.079646017699115e-05, "loss": 0.0013, "step": 12788 }, { "epoch": 1.386491760624458, "grad_norm": 2.2917964458465576, "learning_rate": 1.0795734803423765e-05, "loss": 0.0274, "step": 12789 }, { "epoch": 1.3866001734605378, "grad_norm": 4.636282920837402, "learning_rate": 1.0795009429856378e-05, "loss": 0.0227, "step": 12790 }, { "epoch": 1.3867085862966175, "grad_norm": 1.6217681169509888, "learning_rate": 1.079428405628899e-05, "loss": 0.0276, "step": 12791 }, { "epoch": 1.3868169991326973, "grad_norm": 3.15616774559021, "learning_rate": 1.0793558682721603e-05, "loss": 0.0263, "step": 12792 }, { "epoch": 1.386925411968777, "grad_norm": 5.292498588562012, "learning_rate": 1.0792833309154217e-05, "loss": 0.0494, "step": 12793 }, { "epoch": 1.3870338248048568, "grad_norm": 0.8297560811042786, "learning_rate": 1.0792107935586829e-05, "loss": 0.0187, "step": 12794 }, { "epoch": 1.3871422376409366, "grad_norm": 2.466599941253662, "learning_rate": 1.0791382562019441e-05, "loss": 0.0145, "step": 12795 }, { "epoch": 1.3872506504770166, "grad_norm": 1.0625791549682617, "learning_rate": 1.0790657188452055e-05, "loss": 0.0051, "step": 12796 }, { "epoch": 1.3873590633130963, "grad_norm": 13.842034339904785, "learning_rate": 1.0789931814884667e-05, "loss": 0.0484, "step": 12797 }, { "epoch": 1.387467476149176, "grad_norm": 0.7325395345687866, "learning_rate": 1.0789206441317279e-05, "loss": 0.0077, "step": 12798 }, { "epoch": 1.3875758889852559, "grad_norm": 3.789306640625, "learning_rate": 1.0788481067749891e-05, "loss": 0.0089, "step": 12799 }, { "epoch": 1.3876843018213356, "grad_norm": 1.950097680091858, "learning_rate": 1.0787755694182505e-05, "loss": 0.0254, "step": 12800 }, { "epoch": 1.3877927146574154, "grad_norm": 1.6230918169021606, "learning_rate": 1.0787030320615117e-05, "loss": 0.0115, "step": 12801 }, { "epoch": 1.3879011274934951, "grad_norm": 1.4657667875289917, "learning_rate": 1.078630494704773e-05, "loss": 0.0167, "step": 12802 }, { "epoch": 1.3880095403295751, "grad_norm": 0.965739369392395, "learning_rate": 1.0785579573480343e-05, "loss": 0.0159, "step": 12803 }, { "epoch": 1.388117953165655, "grad_norm": 3.241032838821411, "learning_rate": 1.0784854199912955e-05, "loss": 0.0705, "step": 12804 }, { "epoch": 1.3882263660017347, "grad_norm": 1.7682647705078125, "learning_rate": 1.0784128826345567e-05, "loss": 0.0336, "step": 12805 }, { "epoch": 1.3883347788378144, "grad_norm": 1.9411243200302124, "learning_rate": 1.0783403452778181e-05, "loss": 0.0096, "step": 12806 }, { "epoch": 1.3884431916738942, "grad_norm": 0.5577396750450134, "learning_rate": 1.0782678079210795e-05, "loss": 0.0167, "step": 12807 }, { "epoch": 1.388551604509974, "grad_norm": 2.0644919872283936, "learning_rate": 1.0781952705643409e-05, "loss": 0.03, "step": 12808 }, { "epoch": 1.3886600173460537, "grad_norm": 1.5761792659759521, "learning_rate": 1.0781227332076021e-05, "loss": 0.0168, "step": 12809 }, { "epoch": 1.3887684301821337, "grad_norm": 4.808876037597656, "learning_rate": 1.0780501958508633e-05, "loss": 0.0359, "step": 12810 }, { "epoch": 1.3888768430182132, "grad_norm": 0.45639434456825256, "learning_rate": 1.0779776584941245e-05, "loss": 0.0162, "step": 12811 }, { "epoch": 1.3889852558542932, "grad_norm": 1.8680191040039062, "learning_rate": 1.077905121137386e-05, "loss": 0.0133, "step": 12812 }, { "epoch": 1.389093668690373, "grad_norm": 0.6626560688018799, "learning_rate": 1.0778325837806471e-05, "loss": 0.0072, "step": 12813 }, { "epoch": 1.3892020815264527, "grad_norm": 19.640993118286133, "learning_rate": 1.0777600464239083e-05, "loss": 0.0461, "step": 12814 }, { "epoch": 1.3893104943625325, "grad_norm": 0.19570821523666382, "learning_rate": 1.0776875090671697e-05, "loss": 0.001, "step": 12815 }, { "epoch": 1.3894189071986123, "grad_norm": 0.9477192759513855, "learning_rate": 1.077614971710431e-05, "loss": 0.0219, "step": 12816 }, { "epoch": 1.389527320034692, "grad_norm": 0.3273494243621826, "learning_rate": 1.0775424343536922e-05, "loss": 0.0057, "step": 12817 }, { "epoch": 1.3896357328707718, "grad_norm": 3.8312387466430664, "learning_rate": 1.0774698969969535e-05, "loss": 0.0608, "step": 12818 }, { "epoch": 1.3897441457068518, "grad_norm": 2.0525124073028564, "learning_rate": 1.0773973596402148e-05, "loss": 0.0178, "step": 12819 }, { "epoch": 1.3898525585429315, "grad_norm": 2.466050148010254, "learning_rate": 1.077324822283476e-05, "loss": 0.0336, "step": 12820 }, { "epoch": 1.3899609713790113, "grad_norm": 1.884217381477356, "learning_rate": 1.0772522849267374e-05, "loss": 0.0203, "step": 12821 }, { "epoch": 1.390069384215091, "grad_norm": 0.47043919563293457, "learning_rate": 1.0771797475699986e-05, "loss": 0.0136, "step": 12822 }, { "epoch": 1.3901777970511708, "grad_norm": 0.2338608205318451, "learning_rate": 1.0771072102132598e-05, "loss": 0.0019, "step": 12823 }, { "epoch": 1.3902862098872506, "grad_norm": 0.32610100507736206, "learning_rate": 1.0770346728565213e-05, "loss": 0.0122, "step": 12824 }, { "epoch": 1.3903946227233304, "grad_norm": 5.169716835021973, "learning_rate": 1.0769621354997826e-05, "loss": 0.0212, "step": 12825 }, { "epoch": 1.3905030355594103, "grad_norm": 4.519085884094238, "learning_rate": 1.0768895981430438e-05, "loss": 0.0173, "step": 12826 }, { "epoch": 1.39061144839549, "grad_norm": 0.2230484038591385, "learning_rate": 1.0768170607863051e-05, "loss": 0.0015, "step": 12827 }, { "epoch": 1.3907198612315699, "grad_norm": 0.6043306589126587, "learning_rate": 1.0767445234295664e-05, "loss": 0.0151, "step": 12828 }, { "epoch": 1.3908282740676496, "grad_norm": 0.4413702189922333, "learning_rate": 1.0766719860728276e-05, "loss": 0.0263, "step": 12829 }, { "epoch": 1.3909366869037294, "grad_norm": 1.8850606679916382, "learning_rate": 1.076599448716089e-05, "loss": 0.0031, "step": 12830 }, { "epoch": 1.3910450997398092, "grad_norm": 0.6564436554908752, "learning_rate": 1.0765269113593502e-05, "loss": 0.0307, "step": 12831 }, { "epoch": 1.391153512575889, "grad_norm": 1.989389419555664, "learning_rate": 1.0764543740026114e-05, "loss": 0.0129, "step": 12832 }, { "epoch": 1.391261925411969, "grad_norm": 7.6991448402404785, "learning_rate": 1.0763818366458728e-05, "loss": 0.0515, "step": 12833 }, { "epoch": 1.3913703382480485, "grad_norm": 0.2763608396053314, "learning_rate": 1.076309299289134e-05, "loss": 0.0018, "step": 12834 }, { "epoch": 1.3914787510841284, "grad_norm": 0.1640772521495819, "learning_rate": 1.0762367619323952e-05, "loss": 0.0008, "step": 12835 }, { "epoch": 1.3915871639202082, "grad_norm": 2.210470676422119, "learning_rate": 1.0761642245756566e-05, "loss": 0.0123, "step": 12836 }, { "epoch": 1.391695576756288, "grad_norm": 1.3857814073562622, "learning_rate": 1.0760916872189178e-05, "loss": 0.0275, "step": 12837 }, { "epoch": 1.3918039895923677, "grad_norm": 0.4940783679485321, "learning_rate": 1.076019149862179e-05, "loss": 0.0129, "step": 12838 }, { "epoch": 1.3919124024284475, "grad_norm": 1.4820599555969238, "learning_rate": 1.0759466125054402e-05, "loss": 0.0167, "step": 12839 }, { "epoch": 1.3920208152645273, "grad_norm": 0.33159247040748596, "learning_rate": 1.0758740751487016e-05, "loss": 0.0178, "step": 12840 }, { "epoch": 1.392129228100607, "grad_norm": 1.723348617553711, "learning_rate": 1.075801537791963e-05, "loss": 0.015, "step": 12841 }, { "epoch": 1.392237640936687, "grad_norm": 2.453076124191284, "learning_rate": 1.0757290004352244e-05, "loss": 0.0395, "step": 12842 }, { "epoch": 1.3923460537727668, "grad_norm": 5.283194065093994, "learning_rate": 1.0756564630784856e-05, "loss": 0.0252, "step": 12843 }, { "epoch": 1.3924544666088465, "grad_norm": 3.4539685249328613, "learning_rate": 1.0755839257217468e-05, "loss": 0.0322, "step": 12844 }, { "epoch": 1.3925628794449263, "grad_norm": 3.6450355052948, "learning_rate": 1.0755113883650082e-05, "loss": 0.0365, "step": 12845 }, { "epoch": 1.392671292281006, "grad_norm": 1.3987326622009277, "learning_rate": 1.0754388510082694e-05, "loss": 0.055, "step": 12846 }, { "epoch": 1.3927797051170858, "grad_norm": 0.2853524386882782, "learning_rate": 1.0753663136515306e-05, "loss": 0.0059, "step": 12847 }, { "epoch": 1.3928881179531656, "grad_norm": 1.5907020568847656, "learning_rate": 1.075293776294792e-05, "loss": 0.0268, "step": 12848 }, { "epoch": 1.3929965307892456, "grad_norm": 4.173847198486328, "learning_rate": 1.0752212389380532e-05, "loss": 0.1139, "step": 12849 }, { "epoch": 1.393104943625325, "grad_norm": 0.5531826019287109, "learning_rate": 1.0751487015813144e-05, "loss": 0.0102, "step": 12850 }, { "epoch": 1.393213356461405, "grad_norm": 1.3682835102081299, "learning_rate": 1.0750761642245756e-05, "loss": 0.0106, "step": 12851 }, { "epoch": 1.3933217692974849, "grad_norm": 2.6320297718048096, "learning_rate": 1.075003626867837e-05, "loss": 0.0086, "step": 12852 }, { "epoch": 1.3934301821335646, "grad_norm": 0.2312026470899582, "learning_rate": 1.0749310895110982e-05, "loss": 0.0014, "step": 12853 }, { "epoch": 1.3935385949696444, "grad_norm": 1.1083950996398926, "learning_rate": 1.0748585521543595e-05, "loss": 0.016, "step": 12854 }, { "epoch": 1.3936470078057241, "grad_norm": 0.5121146440505981, "learning_rate": 1.0747860147976208e-05, "loss": 0.0074, "step": 12855 }, { "epoch": 1.393755420641804, "grad_norm": 2.2578024864196777, "learning_rate": 1.074713477440882e-05, "loss": 0.0283, "step": 12856 }, { "epoch": 1.3938638334778837, "grad_norm": 1.4121530055999756, "learning_rate": 1.0746409400841433e-05, "loss": 0.0037, "step": 12857 }, { "epoch": 1.3939722463139637, "grad_norm": 3.049748659133911, "learning_rate": 1.0745684027274047e-05, "loss": 0.0632, "step": 12858 }, { "epoch": 1.3940806591500434, "grad_norm": 0.08194592595100403, "learning_rate": 1.074495865370666e-05, "loss": 0.0005, "step": 12859 }, { "epoch": 1.3941890719861232, "grad_norm": 2.008117437362671, "learning_rate": 1.0744233280139274e-05, "loss": 0.0207, "step": 12860 }, { "epoch": 1.394297484822203, "grad_norm": 0.5345215797424316, "learning_rate": 1.0743507906571886e-05, "loss": 0.0064, "step": 12861 }, { "epoch": 1.3944058976582827, "grad_norm": 2.2262825965881348, "learning_rate": 1.0742782533004499e-05, "loss": 0.0114, "step": 12862 }, { "epoch": 1.3945143104943625, "grad_norm": 2.4769363403320312, "learning_rate": 1.074205715943711e-05, "loss": 0.0217, "step": 12863 }, { "epoch": 1.3946227233304422, "grad_norm": 1.428581953048706, "learning_rate": 1.0741331785869725e-05, "loss": 0.0211, "step": 12864 }, { "epoch": 1.3947311361665222, "grad_norm": 0.4729292094707489, "learning_rate": 1.0740606412302337e-05, "loss": 0.0016, "step": 12865 }, { "epoch": 1.394839549002602, "grad_norm": 1.2629120349884033, "learning_rate": 1.0739881038734949e-05, "loss": 0.037, "step": 12866 }, { "epoch": 1.3949479618386817, "grad_norm": 5.289929389953613, "learning_rate": 1.0739155665167563e-05, "loss": 0.0964, "step": 12867 }, { "epoch": 1.3950563746747615, "grad_norm": 1.9186160564422607, "learning_rate": 1.0738430291600175e-05, "loss": 0.0201, "step": 12868 }, { "epoch": 1.3951647875108413, "grad_norm": 0.9259864687919617, "learning_rate": 1.0737704918032787e-05, "loss": 0.0196, "step": 12869 }, { "epoch": 1.395273200346921, "grad_norm": 1.710052728652954, "learning_rate": 1.07369795444654e-05, "loss": 0.0377, "step": 12870 }, { "epoch": 1.3953816131830008, "grad_norm": 1.3009209632873535, "learning_rate": 1.0736254170898013e-05, "loss": 0.0226, "step": 12871 }, { "epoch": 1.3954900260190808, "grad_norm": 0.4135991036891937, "learning_rate": 1.0735528797330625e-05, "loss": 0.0118, "step": 12872 }, { "epoch": 1.3955984388551603, "grad_norm": 5.861530780792236, "learning_rate": 1.0734803423763239e-05, "loss": 0.0421, "step": 12873 }, { "epoch": 1.3957068516912403, "grad_norm": 2.0721447467803955, "learning_rate": 1.0734078050195851e-05, "loss": 0.0103, "step": 12874 }, { "epoch": 1.39581526452732, "grad_norm": 2.4156229496002197, "learning_rate": 1.0733352676628463e-05, "loss": 0.0235, "step": 12875 }, { "epoch": 1.3959236773633998, "grad_norm": 4.177497863769531, "learning_rate": 1.0732627303061079e-05, "loss": 0.022, "step": 12876 }, { "epoch": 1.3960320901994796, "grad_norm": 1.8008650541305542, "learning_rate": 1.073190192949369e-05, "loss": 0.0383, "step": 12877 }, { "epoch": 1.3961405030355594, "grad_norm": 4.382083892822266, "learning_rate": 1.0731176555926303e-05, "loss": 0.0361, "step": 12878 }, { "epoch": 1.3962489158716391, "grad_norm": 2.2753708362579346, "learning_rate": 1.0730451182358917e-05, "loss": 0.0117, "step": 12879 }, { "epoch": 1.396357328707719, "grad_norm": 0.43636617064476013, "learning_rate": 1.0729725808791529e-05, "loss": 0.003, "step": 12880 }, { "epoch": 1.3964657415437989, "grad_norm": 1.1496727466583252, "learning_rate": 1.0729000435224141e-05, "loss": 0.0205, "step": 12881 }, { "epoch": 1.3965741543798786, "grad_norm": 1.585526943206787, "learning_rate": 1.0728275061656755e-05, "loss": 0.0094, "step": 12882 }, { "epoch": 1.3966825672159584, "grad_norm": 0.09037578105926514, "learning_rate": 1.0727549688089367e-05, "loss": 0.0008, "step": 12883 }, { "epoch": 1.3967909800520382, "grad_norm": 7.38157320022583, "learning_rate": 1.072682431452198e-05, "loss": 0.025, "step": 12884 }, { "epoch": 1.396899392888118, "grad_norm": 3.5338528156280518, "learning_rate": 1.0726098940954593e-05, "loss": 0.0579, "step": 12885 }, { "epoch": 1.3970078057241977, "grad_norm": 1.0254392623901367, "learning_rate": 1.0725373567387205e-05, "loss": 0.0109, "step": 12886 }, { "epoch": 1.3971162185602775, "grad_norm": 0.9911329746246338, "learning_rate": 1.0724648193819817e-05, "loss": 0.0152, "step": 12887 }, { "epoch": 1.3972246313963574, "grad_norm": 3.24623441696167, "learning_rate": 1.0723922820252431e-05, "loss": 0.0589, "step": 12888 }, { "epoch": 1.397333044232437, "grad_norm": 0.539262592792511, "learning_rate": 1.0723197446685043e-05, "loss": 0.0131, "step": 12889 }, { "epoch": 1.397441457068517, "grad_norm": 0.9694265723228455, "learning_rate": 1.0722472073117655e-05, "loss": 0.0244, "step": 12890 }, { "epoch": 1.3975498699045967, "grad_norm": 0.28383082151412964, "learning_rate": 1.0721746699550268e-05, "loss": 0.0182, "step": 12891 }, { "epoch": 1.3976582827406765, "grad_norm": 1.4275745153427124, "learning_rate": 1.0721021325982881e-05, "loss": 0.0047, "step": 12892 }, { "epoch": 1.3977666955767563, "grad_norm": 0.4707970917224884, "learning_rate": 1.0720295952415495e-05, "loss": 0.0128, "step": 12893 }, { "epoch": 1.397875108412836, "grad_norm": 0.273250937461853, "learning_rate": 1.0719570578848109e-05, "loss": 0.0151, "step": 12894 }, { "epoch": 1.397983521248916, "grad_norm": 0.524208128452301, "learning_rate": 1.0718845205280721e-05, "loss": 0.0225, "step": 12895 }, { "epoch": 1.3980919340849955, "grad_norm": 18.207677841186523, "learning_rate": 1.0718119831713333e-05, "loss": 0.067, "step": 12896 }, { "epoch": 1.3982003469210755, "grad_norm": 6.291947364807129, "learning_rate": 1.0717394458145947e-05, "loss": 0.0257, "step": 12897 }, { "epoch": 1.3983087597571553, "grad_norm": 2.8209264278411865, "learning_rate": 1.071666908457856e-05, "loss": 0.0463, "step": 12898 }, { "epoch": 1.398417172593235, "grad_norm": 2.3063318729400635, "learning_rate": 1.0715943711011172e-05, "loss": 0.0491, "step": 12899 }, { "epoch": 1.3985255854293148, "grad_norm": 0.6129186153411865, "learning_rate": 1.0715218337443785e-05, "loss": 0.0197, "step": 12900 }, { "epoch": 1.3986339982653946, "grad_norm": 2.6321914196014404, "learning_rate": 1.0714492963876398e-05, "loss": 0.0237, "step": 12901 }, { "epoch": 1.3987424111014743, "grad_norm": 0.48182550072669983, "learning_rate": 1.071376759030901e-05, "loss": 0.0027, "step": 12902 }, { "epoch": 1.398850823937554, "grad_norm": 0.5891686081886292, "learning_rate": 1.0713042216741622e-05, "loss": 0.0156, "step": 12903 }, { "epoch": 1.398959236773634, "grad_norm": 3.563793897628784, "learning_rate": 1.0712316843174236e-05, "loss": 0.0311, "step": 12904 }, { "epoch": 1.3990676496097139, "grad_norm": 5.8787522315979, "learning_rate": 1.0711591469606848e-05, "loss": 0.0216, "step": 12905 }, { "epoch": 1.3991760624457936, "grad_norm": 4.687880039215088, "learning_rate": 1.071086609603946e-05, "loss": 0.0219, "step": 12906 }, { "epoch": 1.3992844752818734, "grad_norm": 2.8271822929382324, "learning_rate": 1.0710140722472074e-05, "loss": 0.0586, "step": 12907 }, { "epoch": 1.3993928881179531, "grad_norm": 4.0111870765686035, "learning_rate": 1.0709415348904686e-05, "loss": 0.0531, "step": 12908 }, { "epoch": 1.399501300954033, "grad_norm": 2.2525157928466797, "learning_rate": 1.0708689975337298e-05, "loss": 0.0497, "step": 12909 }, { "epoch": 1.3996097137901127, "grad_norm": 1.2050721645355225, "learning_rate": 1.0707964601769914e-05, "loss": 0.0291, "step": 12910 }, { "epoch": 1.3997181266261927, "grad_norm": 2.0631263256073, "learning_rate": 1.0707239228202526e-05, "loss": 0.0302, "step": 12911 }, { "epoch": 1.3998265394622722, "grad_norm": 4.651644229888916, "learning_rate": 1.070651385463514e-05, "loss": 0.0269, "step": 12912 }, { "epoch": 1.3999349522983522, "grad_norm": 3.3645215034484863, "learning_rate": 1.0705788481067752e-05, "loss": 0.0332, "step": 12913 }, { "epoch": 1.400043365134432, "grad_norm": 2.1411242485046387, "learning_rate": 1.0705063107500364e-05, "loss": 0.045, "step": 12914 }, { "epoch": 1.4001517779705117, "grad_norm": 3.1316094398498535, "learning_rate": 1.0704337733932976e-05, "loss": 0.0321, "step": 12915 }, { "epoch": 1.4002601908065915, "grad_norm": 2.534743547439575, "learning_rate": 1.070361236036559e-05, "loss": 0.1086, "step": 12916 }, { "epoch": 1.4003686036426712, "grad_norm": 1.031879186630249, "learning_rate": 1.0702886986798202e-05, "loss": 0.0047, "step": 12917 }, { "epoch": 1.400477016478751, "grad_norm": 1.6677794456481934, "learning_rate": 1.0702161613230814e-05, "loss": 0.0428, "step": 12918 }, { "epoch": 1.4005854293148308, "grad_norm": 2.6821649074554443, "learning_rate": 1.0701436239663428e-05, "loss": 0.0305, "step": 12919 }, { "epoch": 1.4006938421509108, "grad_norm": 4.035801410675049, "learning_rate": 1.070071086609604e-05, "loss": 0.0165, "step": 12920 }, { "epoch": 1.4008022549869905, "grad_norm": 2.1925199031829834, "learning_rate": 1.0699985492528652e-05, "loss": 0.0486, "step": 12921 }, { "epoch": 1.4009106678230703, "grad_norm": 1.1338403224945068, "learning_rate": 1.0699260118961266e-05, "loss": 0.0083, "step": 12922 }, { "epoch": 1.40101908065915, "grad_norm": 1.2157503366470337, "learning_rate": 1.0698534745393878e-05, "loss": 0.0208, "step": 12923 }, { "epoch": 1.4011274934952298, "grad_norm": 0.554736852645874, "learning_rate": 1.069780937182649e-05, "loss": 0.0227, "step": 12924 }, { "epoch": 1.4012359063313096, "grad_norm": 1.5042009353637695, "learning_rate": 1.0697083998259104e-05, "loss": 0.0263, "step": 12925 }, { "epoch": 1.4013443191673893, "grad_norm": 1.2419270277023315, "learning_rate": 1.0696358624691716e-05, "loss": 0.0149, "step": 12926 }, { "epoch": 1.4014527320034693, "grad_norm": 2.1089282035827637, "learning_rate": 1.0695633251124329e-05, "loss": 0.0413, "step": 12927 }, { "epoch": 1.401561144839549, "grad_norm": 3.062971830368042, "learning_rate": 1.0694907877556944e-05, "loss": 0.0285, "step": 12928 }, { "epoch": 1.4016695576756288, "grad_norm": 1.6454510688781738, "learning_rate": 1.0694182503989556e-05, "loss": 0.028, "step": 12929 }, { "epoch": 1.4017779705117086, "grad_norm": 2.4890024662017822, "learning_rate": 1.0693457130422168e-05, "loss": 0.0653, "step": 12930 }, { "epoch": 1.4018863833477884, "grad_norm": 1.6388136148452759, "learning_rate": 1.0692731756854782e-05, "loss": 0.0344, "step": 12931 }, { "epoch": 1.4019947961838681, "grad_norm": 1.2270703315734863, "learning_rate": 1.0692006383287394e-05, "loss": 0.0081, "step": 12932 }, { "epoch": 1.402103209019948, "grad_norm": 2.246534585952759, "learning_rate": 1.0691281009720006e-05, "loss": 0.0209, "step": 12933 }, { "epoch": 1.4022116218560279, "grad_norm": 2.443997621536255, "learning_rate": 1.069055563615262e-05, "loss": 0.0155, "step": 12934 }, { "epoch": 1.4023200346921074, "grad_norm": 1.3039300441741943, "learning_rate": 1.0689830262585232e-05, "loss": 0.0166, "step": 12935 }, { "epoch": 1.4024284475281874, "grad_norm": 1.4970009326934814, "learning_rate": 1.0689104889017845e-05, "loss": 0.0211, "step": 12936 }, { "epoch": 1.4025368603642672, "grad_norm": 0.7532984018325806, "learning_rate": 1.0688379515450458e-05, "loss": 0.0054, "step": 12937 }, { "epoch": 1.402645273200347, "grad_norm": 2.4677913188934326, "learning_rate": 1.068765414188307e-05, "loss": 0.0325, "step": 12938 }, { "epoch": 1.4027536860364267, "grad_norm": 2.8186886310577393, "learning_rate": 1.0686928768315683e-05, "loss": 0.0368, "step": 12939 }, { "epoch": 1.4028620988725065, "grad_norm": 0.5833777785301208, "learning_rate": 1.0686203394748297e-05, "loss": 0.0034, "step": 12940 }, { "epoch": 1.4029705117085862, "grad_norm": 8.75539493560791, "learning_rate": 1.0685478021180909e-05, "loss": 0.0487, "step": 12941 }, { "epoch": 1.403078924544666, "grad_norm": 0.9600929617881775, "learning_rate": 1.068475264761352e-05, "loss": 0.0219, "step": 12942 }, { "epoch": 1.403187337380746, "grad_norm": 8.191596984863281, "learning_rate": 1.0684027274046133e-05, "loss": 0.0347, "step": 12943 }, { "epoch": 1.4032957502168257, "grad_norm": 0.3848296105861664, "learning_rate": 1.0683301900478747e-05, "loss": 0.0077, "step": 12944 }, { "epoch": 1.4034041630529055, "grad_norm": 1.7000138759613037, "learning_rate": 1.068257652691136e-05, "loss": 0.0077, "step": 12945 }, { "epoch": 1.4035125758889853, "grad_norm": 0.757773756980896, "learning_rate": 1.0681851153343974e-05, "loss": 0.0072, "step": 12946 }, { "epoch": 1.403620988725065, "grad_norm": 2.9894766807556152, "learning_rate": 1.0681125779776587e-05, "loss": 0.0433, "step": 12947 }, { "epoch": 1.4037294015611448, "grad_norm": 1.6170098781585693, "learning_rate": 1.0680400406209199e-05, "loss": 0.0337, "step": 12948 }, { "epoch": 1.4038378143972245, "grad_norm": 2.3576884269714355, "learning_rate": 1.0679675032641813e-05, "loss": 0.0423, "step": 12949 }, { "epoch": 1.4039462272333045, "grad_norm": 1.7235534191131592, "learning_rate": 1.0678949659074425e-05, "loss": 0.0111, "step": 12950 }, { "epoch": 1.4040546400693843, "grad_norm": 0.32140877842903137, "learning_rate": 1.0678224285507037e-05, "loss": 0.0126, "step": 12951 }, { "epoch": 1.404163052905464, "grad_norm": 5.856876373291016, "learning_rate": 1.067749891193965e-05, "loss": 0.0392, "step": 12952 }, { "epoch": 1.4042714657415438, "grad_norm": 3.2140629291534424, "learning_rate": 1.0676773538372263e-05, "loss": 0.0275, "step": 12953 }, { "epoch": 1.4043798785776236, "grad_norm": 0.4875439405441284, "learning_rate": 1.0676048164804875e-05, "loss": 0.003, "step": 12954 }, { "epoch": 1.4044882914137033, "grad_norm": 1.3939204216003418, "learning_rate": 1.0675322791237487e-05, "loss": 0.0111, "step": 12955 }, { "epoch": 1.4045967042497831, "grad_norm": 1.0426602363586426, "learning_rate": 1.0674597417670101e-05, "loss": 0.0237, "step": 12956 }, { "epoch": 1.404705117085863, "grad_norm": 0.8880680799484253, "learning_rate": 1.0673872044102713e-05, "loss": 0.0078, "step": 12957 }, { "epoch": 1.4048135299219426, "grad_norm": 2.915682077407837, "learning_rate": 1.0673146670535325e-05, "loss": 0.0824, "step": 12958 }, { "epoch": 1.4049219427580226, "grad_norm": 2.0285823345184326, "learning_rate": 1.0672421296967939e-05, "loss": 0.0225, "step": 12959 }, { "epoch": 1.4050303555941024, "grad_norm": 0.7003351449966431, "learning_rate": 1.0671695923400551e-05, "loss": 0.0264, "step": 12960 }, { "epoch": 1.4051387684301822, "grad_norm": 3.352952003479004, "learning_rate": 1.0670970549833163e-05, "loss": 0.0254, "step": 12961 }, { "epoch": 1.405247181266262, "grad_norm": 1.7501246929168701, "learning_rate": 1.0670245176265779e-05, "loss": 0.0089, "step": 12962 }, { "epoch": 1.4053555941023417, "grad_norm": 0.8218085169792175, "learning_rate": 1.0669519802698391e-05, "loss": 0.0214, "step": 12963 }, { "epoch": 1.4054640069384214, "grad_norm": 4.346510410308838, "learning_rate": 1.0668794429131005e-05, "loss": 0.0168, "step": 12964 }, { "epoch": 1.4055724197745012, "grad_norm": 0.30578121542930603, "learning_rate": 1.0668069055563617e-05, "loss": 0.0103, "step": 12965 }, { "epoch": 1.4056808326105812, "grad_norm": 0.7385688424110413, "learning_rate": 1.066734368199623e-05, "loss": 0.0067, "step": 12966 }, { "epoch": 1.405789245446661, "grad_norm": 0.702896773815155, "learning_rate": 1.0666618308428841e-05, "loss": 0.0111, "step": 12967 }, { "epoch": 1.4058976582827407, "grad_norm": 1.5927926301956177, "learning_rate": 1.0665892934861455e-05, "loss": 0.0122, "step": 12968 }, { "epoch": 1.4060060711188205, "grad_norm": 0.16326312720775604, "learning_rate": 1.0665167561294067e-05, "loss": 0.0009, "step": 12969 }, { "epoch": 1.4061144839549002, "grad_norm": 5.489437103271484, "learning_rate": 1.066444218772668e-05, "loss": 0.0116, "step": 12970 }, { "epoch": 1.40622289679098, "grad_norm": 2.9994661808013916, "learning_rate": 1.0663716814159293e-05, "loss": 0.0213, "step": 12971 }, { "epoch": 1.4063313096270598, "grad_norm": 3.0944666862487793, "learning_rate": 1.0662991440591905e-05, "loss": 0.0151, "step": 12972 }, { "epoch": 1.4064397224631398, "grad_norm": 2.4189670085906982, "learning_rate": 1.0662266067024518e-05, "loss": 0.0165, "step": 12973 }, { "epoch": 1.4065481352992193, "grad_norm": 4.362511157989502, "learning_rate": 1.0661540693457131e-05, "loss": 0.0146, "step": 12974 }, { "epoch": 1.4066565481352993, "grad_norm": 2.7459559440612793, "learning_rate": 1.0660815319889744e-05, "loss": 0.0888, "step": 12975 }, { "epoch": 1.406764960971379, "grad_norm": 6.050456523895264, "learning_rate": 1.0660089946322356e-05, "loss": 0.0115, "step": 12976 }, { "epoch": 1.4068733738074588, "grad_norm": 7.249133110046387, "learning_rate": 1.065936457275497e-05, "loss": 0.0323, "step": 12977 }, { "epoch": 1.4069817866435386, "grad_norm": 0.34915614128112793, "learning_rate": 1.0658639199187582e-05, "loss": 0.0178, "step": 12978 }, { "epoch": 1.4070901994796183, "grad_norm": 1.8684420585632324, "learning_rate": 1.0657913825620194e-05, "loss": 0.064, "step": 12979 }, { "epoch": 1.4071986123156983, "grad_norm": 4.0497002601623535, "learning_rate": 1.065718845205281e-05, "loss": 0.0203, "step": 12980 }, { "epoch": 1.4073070251517779, "grad_norm": 1.187911033630371, "learning_rate": 1.0656463078485422e-05, "loss": 0.0476, "step": 12981 }, { "epoch": 1.4074154379878578, "grad_norm": 0.620119035243988, "learning_rate": 1.0655737704918034e-05, "loss": 0.0192, "step": 12982 }, { "epoch": 1.4075238508239376, "grad_norm": 3.9842722415924072, "learning_rate": 1.0655012331350647e-05, "loss": 0.0176, "step": 12983 }, { "epoch": 1.4076322636600174, "grad_norm": 2.1114776134490967, "learning_rate": 1.065428695778326e-05, "loss": 0.0211, "step": 12984 }, { "epoch": 1.4077406764960971, "grad_norm": 0.7680261135101318, "learning_rate": 1.0653561584215872e-05, "loss": 0.0039, "step": 12985 }, { "epoch": 1.407849089332177, "grad_norm": 0.06851585954427719, "learning_rate": 1.0652836210648486e-05, "loss": 0.0005, "step": 12986 }, { "epoch": 1.4079575021682567, "grad_norm": 0.2662808299064636, "learning_rate": 1.0652110837081098e-05, "loss": 0.0063, "step": 12987 }, { "epoch": 1.4080659150043364, "grad_norm": 0.8945752382278442, "learning_rate": 1.065138546351371e-05, "loss": 0.0083, "step": 12988 }, { "epoch": 1.4081743278404164, "grad_norm": 0.585071861743927, "learning_rate": 1.0650660089946324e-05, "loss": 0.0088, "step": 12989 }, { "epoch": 1.4082827406764962, "grad_norm": 0.8518219590187073, "learning_rate": 1.0649934716378936e-05, "loss": 0.0066, "step": 12990 }, { "epoch": 1.408391153512576, "grad_norm": 1.1144516468048096, "learning_rate": 1.0649209342811548e-05, "loss": 0.0162, "step": 12991 }, { "epoch": 1.4084995663486557, "grad_norm": 1.2357112169265747, "learning_rate": 1.0648483969244162e-05, "loss": 0.0371, "step": 12992 }, { "epoch": 1.4086079791847355, "grad_norm": 1.9319536685943604, "learning_rate": 1.0647758595676774e-05, "loss": 0.0239, "step": 12993 }, { "epoch": 1.4087163920208152, "grad_norm": 1.5769575834274292, "learning_rate": 1.0647033222109386e-05, "loss": 0.0102, "step": 12994 }, { "epoch": 1.408824804856895, "grad_norm": 2.371061325073242, "learning_rate": 1.0646307848541998e-05, "loss": 0.0255, "step": 12995 }, { "epoch": 1.408933217692975, "grad_norm": 1.3553495407104492, "learning_rate": 1.0645582474974612e-05, "loss": 0.0037, "step": 12996 }, { "epoch": 1.4090416305290545, "grad_norm": 2.13053560256958, "learning_rate": 1.0644857101407226e-05, "loss": 0.0278, "step": 12997 }, { "epoch": 1.4091500433651345, "grad_norm": 0.7298462390899658, "learning_rate": 1.064413172783984e-05, "loss": 0.0098, "step": 12998 }, { "epoch": 1.4092584562012143, "grad_norm": 4.325020790100098, "learning_rate": 1.0643406354272452e-05, "loss": 0.0601, "step": 12999 }, { "epoch": 1.409366869037294, "grad_norm": 0.8304789662361145, "learning_rate": 1.0642680980705064e-05, "loss": 0.0124, "step": 13000 }, { "epoch": 1.4094752818733738, "grad_norm": 3.1144821643829346, "learning_rate": 1.0641955607137678e-05, "loss": 0.0777, "step": 13001 }, { "epoch": 1.4095836947094535, "grad_norm": 2.4491515159606934, "learning_rate": 1.064123023357029e-05, "loss": 0.0249, "step": 13002 }, { "epoch": 1.4096921075455333, "grad_norm": 8.597721099853516, "learning_rate": 1.0640504860002902e-05, "loss": 0.0212, "step": 13003 }, { "epoch": 1.409800520381613, "grad_norm": 4.690474987030029, "learning_rate": 1.0639779486435516e-05, "loss": 0.022, "step": 13004 }, { "epoch": 1.409908933217693, "grad_norm": 0.21865063905715942, "learning_rate": 1.0639054112868128e-05, "loss": 0.0022, "step": 13005 }, { "epoch": 1.4100173460537728, "grad_norm": 3.066761016845703, "learning_rate": 1.063832873930074e-05, "loss": 0.0437, "step": 13006 }, { "epoch": 1.4101257588898526, "grad_norm": 0.9529425501823425, "learning_rate": 1.0637603365733352e-05, "loss": 0.0159, "step": 13007 }, { "epoch": 1.4102341717259324, "grad_norm": 2.1438255310058594, "learning_rate": 1.0636877992165966e-05, "loss": 0.0297, "step": 13008 }, { "epoch": 1.4103425845620121, "grad_norm": 0.7045460939407349, "learning_rate": 1.0636152618598578e-05, "loss": 0.0124, "step": 13009 }, { "epoch": 1.4104509973980919, "grad_norm": 4.687535762786865, "learning_rate": 1.063542724503119e-05, "loss": 0.0161, "step": 13010 }, { "epoch": 1.4105594102341716, "grad_norm": 10.989405632019043, "learning_rate": 1.0634701871463804e-05, "loss": 0.0275, "step": 13011 }, { "epoch": 1.4106678230702516, "grad_norm": 0.7162541151046753, "learning_rate": 1.0633976497896417e-05, "loss": 0.0079, "step": 13012 }, { "epoch": 1.4107762359063314, "grad_norm": 1.9676493406295776, "learning_rate": 1.0633251124329029e-05, "loss": 0.0328, "step": 13013 }, { "epoch": 1.4108846487424112, "grad_norm": 2.1752097606658936, "learning_rate": 1.0632525750761644e-05, "loss": 0.025, "step": 13014 }, { "epoch": 1.410993061578491, "grad_norm": 1.2910445928573608, "learning_rate": 1.0631800377194256e-05, "loss": 0.0328, "step": 13015 }, { "epoch": 1.4111014744145707, "grad_norm": 1.305593490600586, "learning_rate": 1.063107500362687e-05, "loss": 0.0417, "step": 13016 }, { "epoch": 1.4112098872506504, "grad_norm": 0.6531866192817688, "learning_rate": 1.0630349630059482e-05, "loss": 0.0118, "step": 13017 }, { "epoch": 1.4113183000867302, "grad_norm": 2.6562225818634033, "learning_rate": 1.0629624256492095e-05, "loss": 0.016, "step": 13018 }, { "epoch": 1.4114267129228102, "grad_norm": 4.230174541473389, "learning_rate": 1.0628898882924707e-05, "loss": 0.0223, "step": 13019 }, { "epoch": 1.4115351257588897, "grad_norm": 0.3362794518470764, "learning_rate": 1.062817350935732e-05, "loss": 0.0212, "step": 13020 }, { "epoch": 1.4116435385949697, "grad_norm": 3.0452260971069336, "learning_rate": 1.0627448135789933e-05, "loss": 0.0437, "step": 13021 }, { "epoch": 1.4117519514310495, "grad_norm": 10.722339630126953, "learning_rate": 1.0626722762222545e-05, "loss": 0.0349, "step": 13022 }, { "epoch": 1.4118603642671292, "grad_norm": 0.756566047668457, "learning_rate": 1.0625997388655159e-05, "loss": 0.0056, "step": 13023 }, { "epoch": 1.411968777103209, "grad_norm": 0.7726079225540161, "learning_rate": 1.062527201508777e-05, "loss": 0.0131, "step": 13024 }, { "epoch": 1.4120771899392888, "grad_norm": 2.590593099594116, "learning_rate": 1.0624546641520383e-05, "loss": 0.0094, "step": 13025 }, { "epoch": 1.4121856027753685, "grad_norm": 5.15300989151001, "learning_rate": 1.0623821267952997e-05, "loss": 0.0364, "step": 13026 }, { "epoch": 1.4122940156114483, "grad_norm": 4.986574649810791, "learning_rate": 1.0623095894385609e-05, "loss": 0.0318, "step": 13027 }, { "epoch": 1.4124024284475283, "grad_norm": 1.4791247844696045, "learning_rate": 1.0622370520818221e-05, "loss": 0.0197, "step": 13028 }, { "epoch": 1.412510841283608, "grad_norm": 4.473242282867432, "learning_rate": 1.0621645147250835e-05, "loss": 0.0771, "step": 13029 }, { "epoch": 1.4126192541196878, "grad_norm": 3.0871267318725586, "learning_rate": 1.0620919773683447e-05, "loss": 0.0393, "step": 13030 }, { "epoch": 1.4127276669557676, "grad_norm": 3.4235785007476807, "learning_rate": 1.0620194400116063e-05, "loss": 0.0338, "step": 13031 }, { "epoch": 1.4128360797918473, "grad_norm": 0.7461102604866028, "learning_rate": 1.0619469026548675e-05, "loss": 0.0264, "step": 13032 }, { "epoch": 1.412944492627927, "grad_norm": 0.4053988754749298, "learning_rate": 1.0618743652981287e-05, "loss": 0.0089, "step": 13033 }, { "epoch": 1.4130529054640069, "grad_norm": 3.5481276512145996, "learning_rate": 1.0618018279413899e-05, "loss": 0.0824, "step": 13034 }, { "epoch": 1.4131613183000868, "grad_norm": 0.3512735664844513, "learning_rate": 1.0617292905846513e-05, "loss": 0.0287, "step": 13035 }, { "epoch": 1.4132697311361664, "grad_norm": 7.858354568481445, "learning_rate": 1.0616567532279125e-05, "loss": 0.0331, "step": 13036 }, { "epoch": 1.4133781439722464, "grad_norm": 0.3327564299106598, "learning_rate": 1.0615842158711737e-05, "loss": 0.0034, "step": 13037 }, { "epoch": 1.4134865568083261, "grad_norm": 1.5343213081359863, "learning_rate": 1.0615116785144351e-05, "loss": 0.0269, "step": 13038 }, { "epoch": 1.413594969644406, "grad_norm": 0.9410562515258789, "learning_rate": 1.0614391411576963e-05, "loss": 0.0165, "step": 13039 }, { "epoch": 1.4137033824804857, "grad_norm": 0.6975739598274231, "learning_rate": 1.0613666038009575e-05, "loss": 0.0078, "step": 13040 }, { "epoch": 1.4138117953165654, "grad_norm": 6.620089530944824, "learning_rate": 1.0612940664442189e-05, "loss": 0.052, "step": 13041 }, { "epoch": 1.4139202081526454, "grad_norm": 2.2774994373321533, "learning_rate": 1.0612215290874801e-05, "loss": 0.017, "step": 13042 }, { "epoch": 1.414028620988725, "grad_norm": 0.28843235969543457, "learning_rate": 1.0611489917307413e-05, "loss": 0.0174, "step": 13043 }, { "epoch": 1.414137033824805, "grad_norm": 0.34438300132751465, "learning_rate": 1.0610764543740027e-05, "loss": 0.0076, "step": 13044 }, { "epoch": 1.4142454466608847, "grad_norm": 1.882943868637085, "learning_rate": 1.061003917017264e-05, "loss": 0.0476, "step": 13045 }, { "epoch": 1.4143538594969645, "grad_norm": 1.8774806261062622, "learning_rate": 1.0609313796605251e-05, "loss": 0.0194, "step": 13046 }, { "epoch": 1.4144622723330442, "grad_norm": 3.9189558029174805, "learning_rate": 1.0608588423037864e-05, "loss": 0.0183, "step": 13047 }, { "epoch": 1.414570685169124, "grad_norm": 0.6046788096427917, "learning_rate": 1.0607863049470477e-05, "loss": 0.0077, "step": 13048 }, { "epoch": 1.4146790980052038, "grad_norm": 1.8748877048492432, "learning_rate": 1.0607137675903091e-05, "loss": 0.0163, "step": 13049 }, { "epoch": 1.4147875108412835, "grad_norm": 1.9882582426071167, "learning_rate": 1.0606412302335705e-05, "loss": 0.0161, "step": 13050 }, { "epoch": 1.4148959236773635, "grad_norm": 3.111201286315918, "learning_rate": 1.0605686928768317e-05, "loss": 0.0115, "step": 13051 }, { "epoch": 1.4150043365134433, "grad_norm": 3.164936065673828, "learning_rate": 1.060496155520093e-05, "loss": 0.0241, "step": 13052 }, { "epoch": 1.415112749349523, "grad_norm": 0.38377970457077026, "learning_rate": 1.0604236181633543e-05, "loss": 0.0144, "step": 13053 }, { "epoch": 1.4152211621856028, "grad_norm": 2.5148699283599854, "learning_rate": 1.0603510808066155e-05, "loss": 0.034, "step": 13054 }, { "epoch": 1.4153295750216826, "grad_norm": 3.7031776905059814, "learning_rate": 1.0602785434498768e-05, "loss": 0.0231, "step": 13055 }, { "epoch": 1.4154379878577623, "grad_norm": 1.1886786222457886, "learning_rate": 1.0602060060931381e-05, "loss": 0.0043, "step": 13056 }, { "epoch": 1.415546400693842, "grad_norm": 4.107936859130859, "learning_rate": 1.0601334687363994e-05, "loss": 0.0453, "step": 13057 }, { "epoch": 1.415654813529922, "grad_norm": 2.252455711364746, "learning_rate": 1.0600609313796606e-05, "loss": 0.0172, "step": 13058 }, { "epoch": 1.4157632263660016, "grad_norm": 5.0571393966674805, "learning_rate": 1.0599883940229218e-05, "loss": 0.0164, "step": 13059 }, { "epoch": 1.4158716392020816, "grad_norm": 0.6944025754928589, "learning_rate": 1.0599158566661832e-05, "loss": 0.0281, "step": 13060 }, { "epoch": 1.4159800520381614, "grad_norm": 1.2215396165847778, "learning_rate": 1.0598433193094444e-05, "loss": 0.0178, "step": 13061 }, { "epoch": 1.4160884648742411, "grad_norm": 0.34631532430648804, "learning_rate": 1.0597707819527056e-05, "loss": 0.0038, "step": 13062 }, { "epoch": 1.4161968777103209, "grad_norm": 1.457776427268982, "learning_rate": 1.059698244595967e-05, "loss": 0.01, "step": 13063 }, { "epoch": 1.4163052905464006, "grad_norm": 0.8548278212547302, "learning_rate": 1.0596257072392282e-05, "loss": 0.0052, "step": 13064 }, { "epoch": 1.4164137033824804, "grad_norm": 2.2803330421447754, "learning_rate": 1.0595531698824894e-05, "loss": 0.0488, "step": 13065 }, { "epoch": 1.4165221162185602, "grad_norm": 1.0514941215515137, "learning_rate": 1.059480632525751e-05, "loss": 0.0167, "step": 13066 }, { "epoch": 1.4166305290546402, "grad_norm": 15.29360580444336, "learning_rate": 1.0594080951690122e-05, "loss": 0.054, "step": 13067 }, { "epoch": 1.41673894189072, "grad_norm": 1.3348362445831299, "learning_rate": 1.0593355578122736e-05, "loss": 0.0073, "step": 13068 }, { "epoch": 1.4168473547267997, "grad_norm": 2.243145227432251, "learning_rate": 1.0592630204555348e-05, "loss": 0.0542, "step": 13069 }, { "epoch": 1.4169557675628794, "grad_norm": 1.1747187376022339, "learning_rate": 1.059190483098796e-05, "loss": 0.0251, "step": 13070 }, { "epoch": 1.4170641803989592, "grad_norm": 1.9049921035766602, "learning_rate": 1.0591179457420572e-05, "loss": 0.0226, "step": 13071 }, { "epoch": 1.417172593235039, "grad_norm": 0.9267534017562866, "learning_rate": 1.0590454083853186e-05, "loss": 0.0095, "step": 13072 }, { "epoch": 1.4172810060711187, "grad_norm": 0.4624527394771576, "learning_rate": 1.0589728710285798e-05, "loss": 0.0131, "step": 13073 }, { "epoch": 1.4173894189071987, "grad_norm": 0.29963019490242004, "learning_rate": 1.058900333671841e-05, "loss": 0.0022, "step": 13074 }, { "epoch": 1.4174978317432785, "grad_norm": 2.0938937664031982, "learning_rate": 1.0588277963151024e-05, "loss": 0.0275, "step": 13075 }, { "epoch": 1.4176062445793582, "grad_norm": 9.244601249694824, "learning_rate": 1.0587552589583636e-05, "loss": 0.0442, "step": 13076 }, { "epoch": 1.417714657415438, "grad_norm": 1.0732289552688599, "learning_rate": 1.0586827216016248e-05, "loss": 0.0078, "step": 13077 }, { "epoch": 1.4178230702515178, "grad_norm": 2.4611568450927734, "learning_rate": 1.0586101842448862e-05, "loss": 0.0542, "step": 13078 }, { "epoch": 1.4179314830875975, "grad_norm": 3.008270025253296, "learning_rate": 1.0585376468881474e-05, "loss": 0.0396, "step": 13079 }, { "epoch": 1.4180398959236773, "grad_norm": 5.9760966300964355, "learning_rate": 1.0584651095314086e-05, "loss": 0.0636, "step": 13080 }, { "epoch": 1.4181483087597573, "grad_norm": 12.008304595947266, "learning_rate": 1.05839257217467e-05, "loss": 0.0436, "step": 13081 }, { "epoch": 1.4182567215958368, "grad_norm": 1.6759063005447388, "learning_rate": 1.0583200348179312e-05, "loss": 0.0245, "step": 13082 }, { "epoch": 1.4183651344319168, "grad_norm": 3.0213913917541504, "learning_rate": 1.0582474974611928e-05, "loss": 0.0381, "step": 13083 }, { "epoch": 1.4184735472679966, "grad_norm": 4.113759994506836, "learning_rate": 1.058174960104454e-05, "loss": 0.0148, "step": 13084 }, { "epoch": 1.4185819601040763, "grad_norm": 4.822463035583496, "learning_rate": 1.0581024227477152e-05, "loss": 0.0334, "step": 13085 }, { "epoch": 1.418690372940156, "grad_norm": 5.798211574554443, "learning_rate": 1.0580298853909764e-05, "loss": 0.0206, "step": 13086 }, { "epoch": 1.4187987857762359, "grad_norm": 1.1060186624526978, "learning_rate": 1.0579573480342378e-05, "loss": 0.0121, "step": 13087 }, { "epoch": 1.4189071986123156, "grad_norm": 9.68399715423584, "learning_rate": 1.057884810677499e-05, "loss": 0.013, "step": 13088 }, { "epoch": 1.4190156114483954, "grad_norm": 5.965898036956787, "learning_rate": 1.0578122733207602e-05, "loss": 0.0183, "step": 13089 }, { "epoch": 1.4191240242844754, "grad_norm": 6.444133281707764, "learning_rate": 1.0577397359640216e-05, "loss": 0.0128, "step": 13090 }, { "epoch": 1.4192324371205551, "grad_norm": 2.2751686573028564, "learning_rate": 1.0576671986072828e-05, "loss": 0.0027, "step": 13091 }, { "epoch": 1.419340849956635, "grad_norm": 0.6058176159858704, "learning_rate": 1.057594661250544e-05, "loss": 0.004, "step": 13092 }, { "epoch": 1.4194492627927147, "grad_norm": 2.9658257961273193, "learning_rate": 1.0575221238938054e-05, "loss": 0.0445, "step": 13093 }, { "epoch": 1.4195576756287944, "grad_norm": 4.728202819824219, "learning_rate": 1.0574495865370667e-05, "loss": 0.0241, "step": 13094 }, { "epoch": 1.4196660884648742, "grad_norm": 0.47600021958351135, "learning_rate": 1.0573770491803279e-05, "loss": 0.0011, "step": 13095 }, { "epoch": 1.419774501300954, "grad_norm": 1.5337101221084595, "learning_rate": 1.0573045118235893e-05, "loss": 0.0307, "step": 13096 }, { "epoch": 1.419882914137034, "grad_norm": 0.4975338280200958, "learning_rate": 1.0572319744668505e-05, "loss": 0.0031, "step": 13097 }, { "epoch": 1.4199913269731137, "grad_norm": 2.1128921508789062, "learning_rate": 1.0571594371101117e-05, "loss": 0.0368, "step": 13098 }, { "epoch": 1.4200997398091935, "grad_norm": 3.6721620559692383, "learning_rate": 1.0570868997533729e-05, "loss": 0.015, "step": 13099 }, { "epoch": 1.4202081526452732, "grad_norm": 1.358906626701355, "learning_rate": 1.0570143623966345e-05, "loss": 0.0236, "step": 13100 }, { "epoch": 1.420316565481353, "grad_norm": 0.29617199301719666, "learning_rate": 1.0569418250398957e-05, "loss": 0.0088, "step": 13101 }, { "epoch": 1.4204249783174328, "grad_norm": 0.8888416290283203, "learning_rate": 1.056869287683157e-05, "loss": 0.0035, "step": 13102 }, { "epoch": 1.4205333911535125, "grad_norm": 0.19361303746700287, "learning_rate": 1.0567967503264183e-05, "loss": 0.004, "step": 13103 }, { "epoch": 1.4206418039895925, "grad_norm": 1.7980364561080933, "learning_rate": 1.0567242129696795e-05, "loss": 0.0154, "step": 13104 }, { "epoch": 1.420750216825672, "grad_norm": 2.5586133003234863, "learning_rate": 1.0566516756129409e-05, "loss": 0.0244, "step": 13105 }, { "epoch": 1.420858629661752, "grad_norm": 0.969007134437561, "learning_rate": 1.056579138256202e-05, "loss": 0.0063, "step": 13106 }, { "epoch": 1.4209670424978318, "grad_norm": 0.6442322134971619, "learning_rate": 1.0565066008994633e-05, "loss": 0.0171, "step": 13107 }, { "epoch": 1.4210754553339116, "grad_norm": 1.0856581926345825, "learning_rate": 1.0564340635427247e-05, "loss": 0.0169, "step": 13108 }, { "epoch": 1.4211838681699913, "grad_norm": 0.33792194724082947, "learning_rate": 1.0563615261859859e-05, "loss": 0.0115, "step": 13109 }, { "epoch": 1.421292281006071, "grad_norm": 2.0811872482299805, "learning_rate": 1.0562889888292471e-05, "loss": 0.028, "step": 13110 }, { "epoch": 1.4214006938421508, "grad_norm": 0.6307767629623413, "learning_rate": 1.0562164514725083e-05, "loss": 0.0313, "step": 13111 }, { "epoch": 1.4215091066782306, "grad_norm": 1.917904019355774, "learning_rate": 1.0561439141157697e-05, "loss": 0.0295, "step": 13112 }, { "epoch": 1.4216175195143106, "grad_norm": 2.2386703491210938, "learning_rate": 1.0560713767590309e-05, "loss": 0.0161, "step": 13113 }, { "epoch": 1.4217259323503904, "grad_norm": 4.966413974761963, "learning_rate": 1.0559988394022921e-05, "loss": 0.0282, "step": 13114 }, { "epoch": 1.4218343451864701, "grad_norm": 0.2888721525669098, "learning_rate": 1.0559263020455535e-05, "loss": 0.0011, "step": 13115 }, { "epoch": 1.4219427580225499, "grad_norm": 3.451777935028076, "learning_rate": 1.0558537646888147e-05, "loss": 0.0269, "step": 13116 }, { "epoch": 1.4220511708586296, "grad_norm": 3.5909364223480225, "learning_rate": 1.055781227332076e-05, "loss": 0.0067, "step": 13117 }, { "epoch": 1.4221595836947094, "grad_norm": 0.4186944365501404, "learning_rate": 1.0557086899753375e-05, "loss": 0.0024, "step": 13118 }, { "epoch": 1.4222679965307892, "grad_norm": 7.084299087524414, "learning_rate": 1.0556361526185987e-05, "loss": 0.0123, "step": 13119 }, { "epoch": 1.4223764093668692, "grad_norm": 0.7532768249511719, "learning_rate": 1.0555636152618601e-05, "loss": 0.0195, "step": 13120 }, { "epoch": 1.4224848222029487, "grad_norm": 0.3436429798603058, "learning_rate": 1.0554910779051213e-05, "loss": 0.0106, "step": 13121 }, { "epoch": 1.4225932350390287, "grad_norm": 1.7939462661743164, "learning_rate": 1.0554185405483825e-05, "loss": 0.0174, "step": 13122 }, { "epoch": 1.4227016478751084, "grad_norm": 0.8960286378860474, "learning_rate": 1.0553460031916437e-05, "loss": 0.0116, "step": 13123 }, { "epoch": 1.4228100607111882, "grad_norm": 5.3328094482421875, "learning_rate": 1.0552734658349051e-05, "loss": 0.0353, "step": 13124 }, { "epoch": 1.422918473547268, "grad_norm": 1.4513218402862549, "learning_rate": 1.0552009284781663e-05, "loss": 0.0196, "step": 13125 }, { "epoch": 1.4230268863833477, "grad_norm": 3.2952680587768555, "learning_rate": 1.0551283911214275e-05, "loss": 0.0697, "step": 13126 }, { "epoch": 1.4231352992194277, "grad_norm": 1.8830602169036865, "learning_rate": 1.055055853764689e-05, "loss": 0.003, "step": 13127 }, { "epoch": 1.4232437120555073, "grad_norm": 3.8666694164276123, "learning_rate": 1.0549833164079501e-05, "loss": 0.0129, "step": 13128 }, { "epoch": 1.4233521248915872, "grad_norm": 3.204402446746826, "learning_rate": 1.0549107790512114e-05, "loss": 0.0481, "step": 13129 }, { "epoch": 1.423460537727667, "grad_norm": 1.947135090827942, "learning_rate": 1.0548382416944727e-05, "loss": 0.0148, "step": 13130 }, { "epoch": 1.4235689505637468, "grad_norm": 0.7461889982223511, "learning_rate": 1.054765704337734e-05, "loss": 0.0206, "step": 13131 }, { "epoch": 1.4236773633998265, "grad_norm": 1.3682256937026978, "learning_rate": 1.0546931669809952e-05, "loss": 0.0065, "step": 13132 }, { "epoch": 1.4237857762359063, "grad_norm": 0.30629774928092957, "learning_rate": 1.0546206296242566e-05, "loss": 0.0062, "step": 13133 }, { "epoch": 1.423894189071986, "grad_norm": 0.4722682535648346, "learning_rate": 1.0545480922675178e-05, "loss": 0.0176, "step": 13134 }, { "epoch": 1.4240026019080658, "grad_norm": 11.829771995544434, "learning_rate": 1.0544755549107793e-05, "loss": 0.0494, "step": 13135 }, { "epoch": 1.4241110147441458, "grad_norm": 8.315764427185059, "learning_rate": 1.0544030175540405e-05, "loss": 0.0648, "step": 13136 }, { "epoch": 1.4242194275802256, "grad_norm": 1.7185781002044678, "learning_rate": 1.0543304801973018e-05, "loss": 0.0158, "step": 13137 }, { "epoch": 1.4243278404163053, "grad_norm": 0.09586384147405624, "learning_rate": 1.054257942840563e-05, "loss": 0.0008, "step": 13138 }, { "epoch": 1.424436253252385, "grad_norm": 0.29149070382118225, "learning_rate": 1.0541854054838244e-05, "loss": 0.0068, "step": 13139 }, { "epoch": 1.4245446660884649, "grad_norm": 1.1081757545471191, "learning_rate": 1.0541128681270856e-05, "loss": 0.0072, "step": 13140 }, { "epoch": 1.4246530789245446, "grad_norm": 3.215179443359375, "learning_rate": 1.0540403307703468e-05, "loss": 0.0092, "step": 13141 }, { "epoch": 1.4247614917606244, "grad_norm": 0.33433881402015686, "learning_rate": 1.0539677934136082e-05, "loss": 0.0185, "step": 13142 }, { "epoch": 1.4248699045967044, "grad_norm": 1.792860507965088, "learning_rate": 1.0538952560568694e-05, "loss": 0.02, "step": 13143 }, { "epoch": 1.424978317432784, "grad_norm": 7.856632232666016, "learning_rate": 1.0538227187001306e-05, "loss": 0.0171, "step": 13144 }, { "epoch": 1.425086730268864, "grad_norm": 3.6217377185821533, "learning_rate": 1.053750181343392e-05, "loss": 0.0047, "step": 13145 }, { "epoch": 1.4251951431049437, "grad_norm": 0.3146536350250244, "learning_rate": 1.0536776439866532e-05, "loss": 0.0015, "step": 13146 }, { "epoch": 1.4253035559410234, "grad_norm": 2.9275753498077393, "learning_rate": 1.0536051066299144e-05, "loss": 0.0618, "step": 13147 }, { "epoch": 1.4254119687771032, "grad_norm": 1.347872018814087, "learning_rate": 1.0535325692731758e-05, "loss": 0.0302, "step": 13148 }, { "epoch": 1.425520381613183, "grad_norm": 2.9084510803222656, "learning_rate": 1.053460031916437e-05, "loss": 0.0479, "step": 13149 }, { "epoch": 1.4256287944492627, "grad_norm": 0.12409090250730515, "learning_rate": 1.0533874945596982e-05, "loss": 0.0008, "step": 13150 }, { "epoch": 1.4257372072853425, "grad_norm": 0.472788006067276, "learning_rate": 1.0533149572029594e-05, "loss": 0.0096, "step": 13151 }, { "epoch": 1.4258456201214225, "grad_norm": 6.885107517242432, "learning_rate": 1.053242419846221e-05, "loss": 0.0632, "step": 13152 }, { "epoch": 1.4259540329575022, "grad_norm": 4.207721710205078, "learning_rate": 1.0531698824894822e-05, "loss": 0.0594, "step": 13153 }, { "epoch": 1.426062445793582, "grad_norm": 1.915959358215332, "learning_rate": 1.0530973451327436e-05, "loss": 0.0465, "step": 13154 }, { "epoch": 1.4261708586296618, "grad_norm": 2.2525155544281006, "learning_rate": 1.0530248077760048e-05, "loss": 0.0235, "step": 13155 }, { "epoch": 1.4262792714657415, "grad_norm": 2.017435073852539, "learning_rate": 1.052952270419266e-05, "loss": 0.0012, "step": 13156 }, { "epoch": 1.4263876843018213, "grad_norm": 0.6080421805381775, "learning_rate": 1.0528797330625274e-05, "loss": 0.0207, "step": 13157 }, { "epoch": 1.426496097137901, "grad_norm": 0.5539584159851074, "learning_rate": 1.0528071957057886e-05, "loss": 0.0265, "step": 13158 }, { "epoch": 1.426604509973981, "grad_norm": 8.964938163757324, "learning_rate": 1.0527346583490498e-05, "loss": 0.0352, "step": 13159 }, { "epoch": 1.4267129228100608, "grad_norm": 7.521317958831787, "learning_rate": 1.0526621209923112e-05, "loss": 0.0787, "step": 13160 }, { "epoch": 1.4268213356461406, "grad_norm": 0.7675606608390808, "learning_rate": 1.0525895836355724e-05, "loss": 0.0108, "step": 13161 }, { "epoch": 1.4269297484822203, "grad_norm": 1.2444732189178467, "learning_rate": 1.0525170462788336e-05, "loss": 0.007, "step": 13162 }, { "epoch": 1.4270381613183, "grad_norm": 4.705690383911133, "learning_rate": 1.0524445089220949e-05, "loss": 0.0181, "step": 13163 }, { "epoch": 1.4271465741543798, "grad_norm": 0.9380009174346924, "learning_rate": 1.0523719715653562e-05, "loss": 0.0093, "step": 13164 }, { "epoch": 1.4272549869904596, "grad_norm": 1.527179479598999, "learning_rate": 1.0522994342086174e-05, "loss": 0.0144, "step": 13165 }, { "epoch": 1.4273633998265396, "grad_norm": 0.4154603183269501, "learning_rate": 1.0522268968518787e-05, "loss": 0.0137, "step": 13166 }, { "epoch": 1.4274718126626191, "grad_norm": 0.704947829246521, "learning_rate": 1.05215435949514e-05, "loss": 0.0133, "step": 13167 }, { "epoch": 1.4275802254986991, "grad_norm": 0.3329516351222992, "learning_rate": 1.0520818221384013e-05, "loss": 0.008, "step": 13168 }, { "epoch": 1.4276886383347789, "grad_norm": 4.5988240242004395, "learning_rate": 1.0520092847816625e-05, "loss": 0.0938, "step": 13169 }, { "epoch": 1.4277970511708586, "grad_norm": 20.442340850830078, "learning_rate": 1.051936747424924e-05, "loss": 0.0358, "step": 13170 }, { "epoch": 1.4279054640069384, "grad_norm": 2.6842987537384033, "learning_rate": 1.0518642100681852e-05, "loss": 0.0568, "step": 13171 }, { "epoch": 1.4280138768430182, "grad_norm": 1.7541372776031494, "learning_rate": 1.0517916727114466e-05, "loss": 0.0518, "step": 13172 }, { "epoch": 1.428122289679098, "grad_norm": 0.8432877659797668, "learning_rate": 1.0517191353547078e-05, "loss": 0.0182, "step": 13173 }, { "epoch": 1.4282307025151777, "grad_norm": 1.3307321071624756, "learning_rate": 1.051646597997969e-05, "loss": 0.0407, "step": 13174 }, { "epoch": 1.4283391153512577, "grad_norm": 3.976769208908081, "learning_rate": 1.0515740606412304e-05, "loss": 0.0584, "step": 13175 }, { "epoch": 1.4284475281873374, "grad_norm": 2.2791523933410645, "learning_rate": 1.0515015232844917e-05, "loss": 0.0215, "step": 13176 }, { "epoch": 1.4285559410234172, "grad_norm": 1.3546032905578613, "learning_rate": 1.0514289859277529e-05, "loss": 0.0168, "step": 13177 }, { "epoch": 1.428664353859497, "grad_norm": 0.908111035823822, "learning_rate": 1.051356448571014e-05, "loss": 0.0062, "step": 13178 }, { "epoch": 1.4287727666955767, "grad_norm": 1.9126626253128052, "learning_rate": 1.0512839112142755e-05, "loss": 0.0112, "step": 13179 }, { "epoch": 1.4288811795316565, "grad_norm": 5.188838958740234, "learning_rate": 1.0512113738575367e-05, "loss": 0.036, "step": 13180 }, { "epoch": 1.4289895923677363, "grad_norm": 0.42475369572639465, "learning_rate": 1.0511388365007979e-05, "loss": 0.0023, "step": 13181 }, { "epoch": 1.4290980052038162, "grad_norm": 1.5094341039657593, "learning_rate": 1.0510662991440593e-05, "loss": 0.0311, "step": 13182 }, { "epoch": 1.4292064180398958, "grad_norm": 0.27053022384643555, "learning_rate": 1.0509937617873205e-05, "loss": 0.0093, "step": 13183 }, { "epoch": 1.4293148308759758, "grad_norm": 1.6512343883514404, "learning_rate": 1.0509212244305817e-05, "loss": 0.0262, "step": 13184 }, { "epoch": 1.4294232437120555, "grad_norm": 3.6209638118743896, "learning_rate": 1.0508486870738431e-05, "loss": 0.0386, "step": 13185 }, { "epoch": 1.4295316565481353, "grad_norm": 1.3358947038650513, "learning_rate": 1.0507761497171043e-05, "loss": 0.0035, "step": 13186 }, { "epoch": 1.429640069384215, "grad_norm": 1.829614281654358, "learning_rate": 1.0507036123603659e-05, "loss": 0.0368, "step": 13187 }, { "epoch": 1.4297484822202948, "grad_norm": 9.105447769165039, "learning_rate": 1.050631075003627e-05, "loss": 0.0413, "step": 13188 }, { "epoch": 1.4298568950563748, "grad_norm": 0.5832311511039734, "learning_rate": 1.0505585376468883e-05, "loss": 0.0191, "step": 13189 }, { "epoch": 1.4299653078924544, "grad_norm": 1.3382453918457031, "learning_rate": 1.0504860002901495e-05, "loss": 0.0168, "step": 13190 }, { "epoch": 1.4300737207285343, "grad_norm": 0.994892418384552, "learning_rate": 1.0504134629334109e-05, "loss": 0.0235, "step": 13191 }, { "epoch": 1.430182133564614, "grad_norm": 5.73836088180542, "learning_rate": 1.0503409255766721e-05, "loss": 0.0084, "step": 13192 }, { "epoch": 1.4302905464006939, "grad_norm": 1.9391021728515625, "learning_rate": 1.0502683882199333e-05, "loss": 0.0459, "step": 13193 }, { "epoch": 1.4303989592367736, "grad_norm": 0.5454114675521851, "learning_rate": 1.0501958508631947e-05, "loss": 0.0173, "step": 13194 }, { "epoch": 1.4305073720728534, "grad_norm": 0.5020231604576111, "learning_rate": 1.0501233135064559e-05, "loss": 0.0141, "step": 13195 }, { "epoch": 1.4306157849089332, "grad_norm": 3.3253204822540283, "learning_rate": 1.0500507761497171e-05, "loss": 0.0116, "step": 13196 }, { "epoch": 1.430724197745013, "grad_norm": 0.3671189248561859, "learning_rate": 1.0499782387929785e-05, "loss": 0.0121, "step": 13197 }, { "epoch": 1.430832610581093, "grad_norm": 2.303237199783325, "learning_rate": 1.0499057014362397e-05, "loss": 0.0127, "step": 13198 }, { "epoch": 1.4309410234171727, "grad_norm": 0.8595816493034363, "learning_rate": 1.049833164079501e-05, "loss": 0.0048, "step": 13199 }, { "epoch": 1.4310494362532524, "grad_norm": 1.2253137826919556, "learning_rate": 1.0497606267227623e-05, "loss": 0.0151, "step": 13200 }, { "epoch": 1.4311578490893322, "grad_norm": 3.750494956970215, "learning_rate": 1.0496880893660235e-05, "loss": 0.0323, "step": 13201 }, { "epoch": 1.431266261925412, "grad_norm": 1.1211402416229248, "learning_rate": 1.0496155520092848e-05, "loss": 0.0067, "step": 13202 }, { "epoch": 1.4313746747614917, "grad_norm": 0.5623771548271179, "learning_rate": 1.049543014652546e-05, "loss": 0.0106, "step": 13203 }, { "epoch": 1.4314830875975715, "grad_norm": 0.8713539242744446, "learning_rate": 1.0494704772958075e-05, "loss": 0.017, "step": 13204 }, { "epoch": 1.4315915004336515, "grad_norm": 1.9861955642700195, "learning_rate": 1.0493979399390687e-05, "loss": 0.0384, "step": 13205 }, { "epoch": 1.431699913269731, "grad_norm": 0.6274131536483765, "learning_rate": 1.0493254025823301e-05, "loss": 0.0028, "step": 13206 }, { "epoch": 1.431808326105811, "grad_norm": 0.9873430728912354, "learning_rate": 1.0492528652255913e-05, "loss": 0.0081, "step": 13207 }, { "epoch": 1.4319167389418908, "grad_norm": 0.28313586115837097, "learning_rate": 1.0491803278688525e-05, "loss": 0.0031, "step": 13208 }, { "epoch": 1.4320251517779705, "grad_norm": 1.6208391189575195, "learning_rate": 1.049107790512114e-05, "loss": 0.0165, "step": 13209 }, { "epoch": 1.4321335646140503, "grad_norm": 2.8884382247924805, "learning_rate": 1.0490352531553751e-05, "loss": 0.07, "step": 13210 }, { "epoch": 1.43224197745013, "grad_norm": 4.01675271987915, "learning_rate": 1.0489627157986364e-05, "loss": 0.0247, "step": 13211 }, { "epoch": 1.4323503902862098, "grad_norm": 3.551809310913086, "learning_rate": 1.0488901784418977e-05, "loss": 0.0066, "step": 13212 }, { "epoch": 1.4324588031222896, "grad_norm": 9.221222877502441, "learning_rate": 1.048817641085159e-05, "loss": 0.038, "step": 13213 }, { "epoch": 1.4325672159583696, "grad_norm": 0.2817084491252899, "learning_rate": 1.0487451037284202e-05, "loss": 0.0009, "step": 13214 }, { "epoch": 1.4326756287944493, "grad_norm": 1.5750998258590698, "learning_rate": 1.0486725663716814e-05, "loss": 0.0136, "step": 13215 }, { "epoch": 1.432784041630529, "grad_norm": 2.4198498725891113, "learning_rate": 1.0486000290149428e-05, "loss": 0.0491, "step": 13216 }, { "epoch": 1.4328924544666088, "grad_norm": 0.16930913925170898, "learning_rate": 1.048527491658204e-05, "loss": 0.0026, "step": 13217 }, { "epoch": 1.4330008673026886, "grad_norm": 0.34279343485832214, "learning_rate": 1.0484549543014652e-05, "loss": 0.0072, "step": 13218 }, { "epoch": 1.4331092801387684, "grad_norm": 2.638033151626587, "learning_rate": 1.0483824169447266e-05, "loss": 0.0496, "step": 13219 }, { "epoch": 1.4332176929748481, "grad_norm": 4.879482269287109, "learning_rate": 1.0483098795879878e-05, "loss": 0.0227, "step": 13220 }, { "epoch": 1.4333261058109281, "grad_norm": 0.7352001667022705, "learning_rate": 1.0482373422312493e-05, "loss": 0.0147, "step": 13221 }, { "epoch": 1.4334345186470079, "grad_norm": 1.9314124584197998, "learning_rate": 1.0481648048745106e-05, "loss": 0.0196, "step": 13222 }, { "epoch": 1.4335429314830876, "grad_norm": 4.722945213317871, "learning_rate": 1.0480922675177718e-05, "loss": 0.0215, "step": 13223 }, { "epoch": 1.4336513443191674, "grad_norm": 5.935818195343018, "learning_rate": 1.0480197301610332e-05, "loss": 0.0304, "step": 13224 }, { "epoch": 1.4337597571552472, "grad_norm": 4.094651222229004, "learning_rate": 1.0479471928042944e-05, "loss": 0.0463, "step": 13225 }, { "epoch": 1.433868169991327, "grad_norm": 2.520848512649536, "learning_rate": 1.0478746554475556e-05, "loss": 0.0591, "step": 13226 }, { "epoch": 1.4339765828274067, "grad_norm": 1.8884189128875732, "learning_rate": 1.047802118090817e-05, "loss": 0.025, "step": 13227 }, { "epoch": 1.4340849956634867, "grad_norm": 9.153964042663574, "learning_rate": 1.0477295807340782e-05, "loss": 0.0576, "step": 13228 }, { "epoch": 1.4341934084995662, "grad_norm": 5.863832950592041, "learning_rate": 1.0476570433773394e-05, "loss": 0.0164, "step": 13229 }, { "epoch": 1.4343018213356462, "grad_norm": 0.47727906703948975, "learning_rate": 1.0475845060206006e-05, "loss": 0.016, "step": 13230 }, { "epoch": 1.434410234171726, "grad_norm": 5.558053016662598, "learning_rate": 1.047511968663862e-05, "loss": 0.0289, "step": 13231 }, { "epoch": 1.4345186470078057, "grad_norm": 2.127023458480835, "learning_rate": 1.0474394313071232e-05, "loss": 0.0363, "step": 13232 }, { "epoch": 1.4346270598438855, "grad_norm": 0.5849302411079407, "learning_rate": 1.0473668939503844e-05, "loss": 0.0074, "step": 13233 }, { "epoch": 1.4347354726799653, "grad_norm": 0.7858381271362305, "learning_rate": 1.0472943565936458e-05, "loss": 0.0044, "step": 13234 }, { "epoch": 1.434843885516045, "grad_norm": 0.057780373841524124, "learning_rate": 1.047221819236907e-05, "loss": 0.0005, "step": 13235 }, { "epoch": 1.4349522983521248, "grad_norm": 5.355767250061035, "learning_rate": 1.0471492818801682e-05, "loss": 0.027, "step": 13236 }, { "epoch": 1.4350607111882048, "grad_norm": 1.4080954790115356, "learning_rate": 1.0470767445234296e-05, "loss": 0.0269, "step": 13237 }, { "epoch": 1.4351691240242845, "grad_norm": 2.0530059337615967, "learning_rate": 1.0470042071666908e-05, "loss": 0.03, "step": 13238 }, { "epoch": 1.4352775368603643, "grad_norm": 2.158280849456787, "learning_rate": 1.0469316698099524e-05, "loss": 0.0282, "step": 13239 }, { "epoch": 1.435385949696444, "grad_norm": 0.5708619952201843, "learning_rate": 1.0468591324532136e-05, "loss": 0.0028, "step": 13240 }, { "epoch": 1.4354943625325238, "grad_norm": 0.27947041392326355, "learning_rate": 1.0467865950964748e-05, "loss": 0.0068, "step": 13241 }, { "epoch": 1.4356027753686036, "grad_norm": 5.310966491699219, "learning_rate": 1.046714057739736e-05, "loss": 0.0515, "step": 13242 }, { "epoch": 1.4357111882046834, "grad_norm": 0.7997069954872131, "learning_rate": 1.0466415203829974e-05, "loss": 0.0107, "step": 13243 }, { "epoch": 1.4358196010407633, "grad_norm": 2.1600911617279053, "learning_rate": 1.0465689830262586e-05, "loss": 0.0155, "step": 13244 }, { "epoch": 1.435928013876843, "grad_norm": 0.2719329595565796, "learning_rate": 1.0464964456695198e-05, "loss": 0.0065, "step": 13245 }, { "epoch": 1.4360364267129229, "grad_norm": 0.26238179206848145, "learning_rate": 1.0464239083127812e-05, "loss": 0.0192, "step": 13246 }, { "epoch": 1.4361448395490026, "grad_norm": 4.352330684661865, "learning_rate": 1.0463513709560424e-05, "loss": 0.0447, "step": 13247 }, { "epoch": 1.4362532523850824, "grad_norm": 5.453233242034912, "learning_rate": 1.0462788335993037e-05, "loss": 0.0248, "step": 13248 }, { "epoch": 1.4363616652211622, "grad_norm": 0.3217136859893799, "learning_rate": 1.046206296242565e-05, "loss": 0.0068, "step": 13249 }, { "epoch": 1.436470078057242, "grad_norm": 1.936288833618164, "learning_rate": 1.0461337588858263e-05, "loss": 0.0153, "step": 13250 }, { "epoch": 1.436578490893322, "grad_norm": 4.953063011169434, "learning_rate": 1.0460612215290875e-05, "loss": 0.0154, "step": 13251 }, { "epoch": 1.4366869037294014, "grad_norm": 5.947673320770264, "learning_rate": 1.0459886841723489e-05, "loss": 0.0459, "step": 13252 }, { "epoch": 1.4367953165654814, "grad_norm": 0.43656644225120544, "learning_rate": 1.04591614681561e-05, "loss": 0.0091, "step": 13253 }, { "epoch": 1.4369037294015612, "grad_norm": 1.04696786403656, "learning_rate": 1.0458436094588713e-05, "loss": 0.0022, "step": 13254 }, { "epoch": 1.437012142237641, "grad_norm": 0.3246101438999176, "learning_rate": 1.0457710721021325e-05, "loss": 0.0083, "step": 13255 }, { "epoch": 1.4371205550737207, "grad_norm": 0.4637770354747772, "learning_rate": 1.045698534745394e-05, "loss": 0.0009, "step": 13256 }, { "epoch": 1.4372289679098005, "grad_norm": 2.0471935272216797, "learning_rate": 1.0456259973886553e-05, "loss": 0.0094, "step": 13257 }, { "epoch": 1.4373373807458802, "grad_norm": 6.814828395843506, "learning_rate": 1.0455534600319166e-05, "loss": 0.0339, "step": 13258 }, { "epoch": 1.43744579358196, "grad_norm": 0.7389218807220459, "learning_rate": 1.0454809226751779e-05, "loss": 0.0027, "step": 13259 }, { "epoch": 1.43755420641804, "grad_norm": 0.6135815382003784, "learning_rate": 1.045408385318439e-05, "loss": 0.0095, "step": 13260 }, { "epoch": 1.4376626192541198, "grad_norm": 6.746767520904541, "learning_rate": 1.0453358479617005e-05, "loss": 0.0227, "step": 13261 }, { "epoch": 1.4377710320901995, "grad_norm": 1.0739580392837524, "learning_rate": 1.0452633106049617e-05, "loss": 0.0117, "step": 13262 }, { "epoch": 1.4378794449262793, "grad_norm": 6.122714996337891, "learning_rate": 1.0451907732482229e-05, "loss": 0.0596, "step": 13263 }, { "epoch": 1.437987857762359, "grad_norm": 4.4617109298706055, "learning_rate": 1.0451182358914843e-05, "loss": 0.0584, "step": 13264 }, { "epoch": 1.4380962705984388, "grad_norm": 0.16506630182266235, "learning_rate": 1.0450456985347455e-05, "loss": 0.0012, "step": 13265 }, { "epoch": 1.4382046834345186, "grad_norm": 3.2216434478759766, "learning_rate": 1.0449731611780067e-05, "loss": 0.0058, "step": 13266 }, { "epoch": 1.4383130962705986, "grad_norm": 0.9481397271156311, "learning_rate": 1.044900623821268e-05, "loss": 0.0168, "step": 13267 }, { "epoch": 1.438421509106678, "grad_norm": 2.203906297683716, "learning_rate": 1.0448280864645293e-05, "loss": 0.0152, "step": 13268 }, { "epoch": 1.438529921942758, "grad_norm": 3.4585423469543457, "learning_rate": 1.0447555491077905e-05, "loss": 0.0171, "step": 13269 }, { "epoch": 1.4386383347788378, "grad_norm": 0.11709737032651901, "learning_rate": 1.0446830117510517e-05, "loss": 0.0006, "step": 13270 }, { "epoch": 1.4387467476149176, "grad_norm": 1.6349576711654663, "learning_rate": 1.0446104743943131e-05, "loss": 0.0274, "step": 13271 }, { "epoch": 1.4388551604509974, "grad_norm": 3.6234750747680664, "learning_rate": 1.0445379370375743e-05, "loss": 0.054, "step": 13272 }, { "epoch": 1.4389635732870771, "grad_norm": 3.225034713745117, "learning_rate": 1.0444653996808359e-05, "loss": 0.0261, "step": 13273 }, { "epoch": 1.439071986123157, "grad_norm": 1.0137463808059692, "learning_rate": 1.0443928623240971e-05, "loss": 0.008, "step": 13274 }, { "epoch": 1.4391803989592367, "grad_norm": 0.22351385653018951, "learning_rate": 1.0443203249673583e-05, "loss": 0.0162, "step": 13275 }, { "epoch": 1.4392888117953166, "grad_norm": 1.0934169292449951, "learning_rate": 1.0442477876106197e-05, "loss": 0.0041, "step": 13276 }, { "epoch": 1.4393972246313964, "grad_norm": 5.970538139343262, "learning_rate": 1.0441752502538809e-05, "loss": 0.0267, "step": 13277 }, { "epoch": 1.4395056374674762, "grad_norm": 2.840947389602661, "learning_rate": 1.0441027128971421e-05, "loss": 0.0451, "step": 13278 }, { "epoch": 1.439614050303556, "grad_norm": 12.875006675720215, "learning_rate": 1.0440301755404035e-05, "loss": 0.0504, "step": 13279 }, { "epoch": 1.4397224631396357, "grad_norm": 1.307105541229248, "learning_rate": 1.0439576381836647e-05, "loss": 0.0213, "step": 13280 }, { "epoch": 1.4398308759757155, "grad_norm": 1.4346809387207031, "learning_rate": 1.043885100826926e-05, "loss": 0.0233, "step": 13281 }, { "epoch": 1.4399392888117952, "grad_norm": 0.5282348394393921, "learning_rate": 1.0438125634701871e-05, "loss": 0.0079, "step": 13282 }, { "epoch": 1.4400477016478752, "grad_norm": 3.764153480529785, "learning_rate": 1.0437400261134485e-05, "loss": 0.0177, "step": 13283 }, { "epoch": 1.440156114483955, "grad_norm": 0.4992634057998657, "learning_rate": 1.0436674887567097e-05, "loss": 0.0167, "step": 13284 }, { "epoch": 1.4402645273200347, "grad_norm": 12.803938865661621, "learning_rate": 1.043594951399971e-05, "loss": 0.0368, "step": 13285 }, { "epoch": 1.4403729401561145, "grad_norm": 0.7922650575637817, "learning_rate": 1.0435224140432323e-05, "loss": 0.0177, "step": 13286 }, { "epoch": 1.4404813529921943, "grad_norm": 0.5808444619178772, "learning_rate": 1.0434498766864936e-05, "loss": 0.0158, "step": 13287 }, { "epoch": 1.440589765828274, "grad_norm": 2.2907299995422363, "learning_rate": 1.0433773393297548e-05, "loss": 0.0134, "step": 13288 }, { "epoch": 1.4406981786643538, "grad_norm": 6.118897914886475, "learning_rate": 1.0433048019730162e-05, "loss": 0.0684, "step": 13289 }, { "epoch": 1.4408065915004338, "grad_norm": 9.389077186584473, "learning_rate": 1.0432322646162775e-05, "loss": 0.0551, "step": 13290 }, { "epoch": 1.4409150043365133, "grad_norm": 1.2471837997436523, "learning_rate": 1.043159727259539e-05, "loss": 0.0054, "step": 13291 }, { "epoch": 1.4410234171725933, "grad_norm": 2.382840633392334, "learning_rate": 1.0430871899028001e-05, "loss": 0.015, "step": 13292 }, { "epoch": 1.441131830008673, "grad_norm": 0.2913439869880676, "learning_rate": 1.0430146525460614e-05, "loss": 0.0107, "step": 13293 }, { "epoch": 1.4412402428447528, "grad_norm": 0.47189632058143616, "learning_rate": 1.0429421151893226e-05, "loss": 0.0132, "step": 13294 }, { "epoch": 1.4413486556808326, "grad_norm": 2.933000326156616, "learning_rate": 1.042869577832584e-05, "loss": 0.0111, "step": 13295 }, { "epoch": 1.4414570685169124, "grad_norm": 0.7844198942184448, "learning_rate": 1.0427970404758452e-05, "loss": 0.0041, "step": 13296 }, { "epoch": 1.4415654813529921, "grad_norm": 4.5027689933776855, "learning_rate": 1.0427245031191064e-05, "loss": 0.0187, "step": 13297 }, { "epoch": 1.4416738941890719, "grad_norm": 4.955831527709961, "learning_rate": 1.0426519657623678e-05, "loss": 0.0327, "step": 13298 }, { "epoch": 1.4417823070251519, "grad_norm": 10.774580955505371, "learning_rate": 1.042579428405629e-05, "loss": 0.0559, "step": 13299 }, { "epoch": 1.4418907198612316, "grad_norm": 0.26451951265335083, "learning_rate": 1.0425068910488902e-05, "loss": 0.0055, "step": 13300 }, { "epoch": 1.4419991326973114, "grad_norm": 2.3621435165405273, "learning_rate": 1.0424343536921516e-05, "loss": 0.0753, "step": 13301 }, { "epoch": 1.4421075455333912, "grad_norm": 3.631242036819458, "learning_rate": 1.0423618163354128e-05, "loss": 0.0314, "step": 13302 }, { "epoch": 1.442215958369471, "grad_norm": 2.069611072540283, "learning_rate": 1.042289278978674e-05, "loss": 0.0723, "step": 13303 }, { "epoch": 1.4423243712055507, "grad_norm": 3.4550395011901855, "learning_rate": 1.0422167416219354e-05, "loss": 0.0313, "step": 13304 }, { "epoch": 1.4424327840416304, "grad_norm": 2.6771366596221924, "learning_rate": 1.0421442042651966e-05, "loss": 0.0438, "step": 13305 }, { "epoch": 1.4425411968777104, "grad_norm": 4.332179069519043, "learning_rate": 1.0420716669084578e-05, "loss": 0.0488, "step": 13306 }, { "epoch": 1.4426496097137902, "grad_norm": 0.25692665576934814, "learning_rate": 1.041999129551719e-05, "loss": 0.0019, "step": 13307 }, { "epoch": 1.44275802254987, "grad_norm": 0.6626154184341431, "learning_rate": 1.0419265921949806e-05, "loss": 0.0029, "step": 13308 }, { "epoch": 1.4428664353859497, "grad_norm": 3.6981234550476074, "learning_rate": 1.0418540548382418e-05, "loss": 0.0264, "step": 13309 }, { "epoch": 1.4429748482220295, "grad_norm": 1.3964451551437378, "learning_rate": 1.0417815174815032e-05, "loss": 0.0063, "step": 13310 }, { "epoch": 1.4430832610581092, "grad_norm": 3.751260280609131, "learning_rate": 1.0417089801247644e-05, "loss": 0.0344, "step": 13311 }, { "epoch": 1.443191673894189, "grad_norm": 0.29064998030662537, "learning_rate": 1.0416364427680256e-05, "loss": 0.0171, "step": 13312 }, { "epoch": 1.443300086730269, "grad_norm": 2.0129926204681396, "learning_rate": 1.041563905411287e-05, "loss": 0.051, "step": 13313 }, { "epoch": 1.4434084995663485, "grad_norm": 4.910007476806641, "learning_rate": 1.0414913680545482e-05, "loss": 0.0183, "step": 13314 }, { "epoch": 1.4435169124024285, "grad_norm": 0.42424386739730835, "learning_rate": 1.0414188306978094e-05, "loss": 0.012, "step": 13315 }, { "epoch": 1.4436253252385083, "grad_norm": 1.1224629878997803, "learning_rate": 1.0413462933410708e-05, "loss": 0.0162, "step": 13316 }, { "epoch": 1.443733738074588, "grad_norm": 1.4493916034698486, "learning_rate": 1.041273755984332e-05, "loss": 0.0154, "step": 13317 }, { "epoch": 1.4438421509106678, "grad_norm": 4.684647560119629, "learning_rate": 1.0412012186275932e-05, "loss": 0.0081, "step": 13318 }, { "epoch": 1.4439505637467476, "grad_norm": 0.3335331678390503, "learning_rate": 1.0411286812708545e-05, "loss": 0.0132, "step": 13319 }, { "epoch": 1.4440589765828273, "grad_norm": 1.3118855953216553, "learning_rate": 1.0410561439141158e-05, "loss": 0.0176, "step": 13320 }, { "epoch": 1.444167389418907, "grad_norm": 5.564804553985596, "learning_rate": 1.040983606557377e-05, "loss": 0.0042, "step": 13321 }, { "epoch": 1.444275802254987, "grad_norm": 1.038251519203186, "learning_rate": 1.0409110692006383e-05, "loss": 0.0178, "step": 13322 }, { "epoch": 1.4443842150910668, "grad_norm": 2.1993448734283447, "learning_rate": 1.0408385318438996e-05, "loss": 0.0306, "step": 13323 }, { "epoch": 1.4444926279271466, "grad_norm": 1.050670862197876, "learning_rate": 1.0407659944871609e-05, "loss": 0.0042, "step": 13324 }, { "epoch": 1.4446010407632264, "grad_norm": 1.2652276754379272, "learning_rate": 1.0406934571304224e-05, "loss": 0.0067, "step": 13325 }, { "epoch": 1.4447094535993061, "grad_norm": 1.3046791553497314, "learning_rate": 1.0406209197736836e-05, "loss": 0.0252, "step": 13326 }, { "epoch": 1.444817866435386, "grad_norm": 1.8104983568191528, "learning_rate": 1.0405483824169448e-05, "loss": 0.0355, "step": 13327 }, { "epoch": 1.4449262792714657, "grad_norm": 2.847114086151123, "learning_rate": 1.0404758450602062e-05, "loss": 0.051, "step": 13328 }, { "epoch": 1.4450346921075456, "grad_norm": 0.05786336958408356, "learning_rate": 1.0404033077034674e-05, "loss": 0.0005, "step": 13329 }, { "epoch": 1.4451431049436252, "grad_norm": 0.45941224694252014, "learning_rate": 1.0403307703467287e-05, "loss": 0.0062, "step": 13330 }, { "epoch": 1.4452515177797052, "grad_norm": 2.675936222076416, "learning_rate": 1.04025823298999e-05, "loss": 0.0397, "step": 13331 }, { "epoch": 1.445359930615785, "grad_norm": 0.1312810331583023, "learning_rate": 1.0401856956332513e-05, "loss": 0.0015, "step": 13332 }, { "epoch": 1.4454683434518647, "grad_norm": 1.3976316452026367, "learning_rate": 1.0401131582765125e-05, "loss": 0.0336, "step": 13333 }, { "epoch": 1.4455767562879445, "grad_norm": 1.069734811782837, "learning_rate": 1.0400406209197737e-05, "loss": 0.0318, "step": 13334 }, { "epoch": 1.4456851691240242, "grad_norm": 2.23032546043396, "learning_rate": 1.039968083563035e-05, "loss": 0.004, "step": 13335 }, { "epoch": 1.4457935819601042, "grad_norm": 0.40689605474472046, "learning_rate": 1.0398955462062963e-05, "loss": 0.0011, "step": 13336 }, { "epoch": 1.4459019947961838, "grad_norm": 0.31315019726753235, "learning_rate": 1.0398230088495575e-05, "loss": 0.0029, "step": 13337 }, { "epoch": 1.4460104076322637, "grad_norm": 0.2574698328971863, "learning_rate": 1.0397504714928189e-05, "loss": 0.0006, "step": 13338 }, { "epoch": 1.4461188204683435, "grad_norm": 1.7667627334594727, "learning_rate": 1.0396779341360801e-05, "loss": 0.0296, "step": 13339 }, { "epoch": 1.4462272333044233, "grad_norm": 3.1127912998199463, "learning_rate": 1.0396053967793413e-05, "loss": 0.034, "step": 13340 }, { "epoch": 1.446335646140503, "grad_norm": 2.2532904148101807, "learning_rate": 1.0395328594226027e-05, "loss": 0.0479, "step": 13341 }, { "epoch": 1.4464440589765828, "grad_norm": 20.58791732788086, "learning_rate": 1.039460322065864e-05, "loss": 0.0324, "step": 13342 }, { "epoch": 1.4465524718126626, "grad_norm": 2.2091307640075684, "learning_rate": 1.0393877847091255e-05, "loss": 0.0156, "step": 13343 }, { "epoch": 1.4466608846487423, "grad_norm": 0.5019559264183044, "learning_rate": 1.0393152473523867e-05, "loss": 0.0027, "step": 13344 }, { "epoch": 1.4467692974848223, "grad_norm": 5.033097743988037, "learning_rate": 1.0392427099956479e-05, "loss": 0.0381, "step": 13345 }, { "epoch": 1.446877710320902, "grad_norm": 0.5840543508529663, "learning_rate": 1.0391701726389091e-05, "loss": 0.0058, "step": 13346 }, { "epoch": 1.4469861231569818, "grad_norm": 1.0784207582473755, "learning_rate": 1.0390976352821705e-05, "loss": 0.0043, "step": 13347 }, { "epoch": 1.4470945359930616, "grad_norm": 3.265512228012085, "learning_rate": 1.0390250979254317e-05, "loss": 0.0147, "step": 13348 }, { "epoch": 1.4472029488291414, "grad_norm": 0.47436943650245667, "learning_rate": 1.0389525605686929e-05, "loss": 0.0078, "step": 13349 }, { "epoch": 1.4473113616652211, "grad_norm": 0.32156646251678467, "learning_rate": 1.0388800232119543e-05, "loss": 0.0024, "step": 13350 }, { "epoch": 1.4474197745013009, "grad_norm": 1.8937655687332153, "learning_rate": 1.0388074858552155e-05, "loss": 0.015, "step": 13351 }, { "epoch": 1.4475281873373809, "grad_norm": 3.955409049987793, "learning_rate": 1.0387349484984767e-05, "loss": 0.0142, "step": 13352 }, { "epoch": 1.4476366001734604, "grad_norm": 2.6875927448272705, "learning_rate": 1.0386624111417381e-05, "loss": 0.0227, "step": 13353 }, { "epoch": 1.4477450130095404, "grad_norm": 2.624441385269165, "learning_rate": 1.0385898737849993e-05, "loss": 0.012, "step": 13354 }, { "epoch": 1.4478534258456202, "grad_norm": 0.5585238933563232, "learning_rate": 1.0385173364282605e-05, "loss": 0.0058, "step": 13355 }, { "epoch": 1.4479618386817, "grad_norm": 2.8684754371643066, "learning_rate": 1.038444799071522e-05, "loss": 0.0215, "step": 13356 }, { "epoch": 1.4480702515177797, "grad_norm": 2.4655158519744873, "learning_rate": 1.0383722617147831e-05, "loss": 0.012, "step": 13357 }, { "epoch": 1.4481786643538594, "grad_norm": 2.795443534851074, "learning_rate": 1.0382997243580444e-05, "loss": 0.0186, "step": 13358 }, { "epoch": 1.4482870771899392, "grad_norm": 0.11217087507247925, "learning_rate": 1.0382271870013056e-05, "loss": 0.001, "step": 13359 }, { "epoch": 1.448395490026019, "grad_norm": 0.8987830281257629, "learning_rate": 1.0381546496445671e-05, "loss": 0.0246, "step": 13360 }, { "epoch": 1.448503902862099, "grad_norm": 0.8764113783836365, "learning_rate": 1.0380821122878283e-05, "loss": 0.0185, "step": 13361 }, { "epoch": 1.4486123156981787, "grad_norm": 15.546088218688965, "learning_rate": 1.0380095749310897e-05, "loss": 0.0437, "step": 13362 }, { "epoch": 1.4487207285342585, "grad_norm": 0.5881134867668152, "learning_rate": 1.037937037574351e-05, "loss": 0.0042, "step": 13363 }, { "epoch": 1.4488291413703382, "grad_norm": 0.642757773399353, "learning_rate": 1.0378645002176121e-05, "loss": 0.0014, "step": 13364 }, { "epoch": 1.448937554206418, "grad_norm": 2.672647714614868, "learning_rate": 1.0377919628608735e-05, "loss": 0.0434, "step": 13365 }, { "epoch": 1.4490459670424978, "grad_norm": 0.7072030901908875, "learning_rate": 1.0377194255041347e-05, "loss": 0.0019, "step": 13366 }, { "epoch": 1.4491543798785775, "grad_norm": 5.890017032623291, "learning_rate": 1.037646888147396e-05, "loss": 0.0578, "step": 13367 }, { "epoch": 1.4492627927146575, "grad_norm": 2.2896530628204346, "learning_rate": 1.0375743507906573e-05, "loss": 0.0268, "step": 13368 }, { "epoch": 1.4493712055507373, "grad_norm": 3.5421090126037598, "learning_rate": 1.0375018134339186e-05, "loss": 0.0091, "step": 13369 }, { "epoch": 1.449479618386817, "grad_norm": 0.5070486664772034, "learning_rate": 1.0374292760771798e-05, "loss": 0.0016, "step": 13370 }, { "epoch": 1.4495880312228968, "grad_norm": 0.7891908288002014, "learning_rate": 1.037356738720441e-05, "loss": 0.004, "step": 13371 }, { "epoch": 1.4496964440589766, "grad_norm": 10.785022735595703, "learning_rate": 1.0372842013637024e-05, "loss": 0.0234, "step": 13372 }, { "epoch": 1.4498048568950563, "grad_norm": 2.3802237510681152, "learning_rate": 1.0372116640069636e-05, "loss": 0.0252, "step": 13373 }, { "epoch": 1.449913269731136, "grad_norm": 2.8609538078308105, "learning_rate": 1.0371391266502248e-05, "loss": 0.0199, "step": 13374 }, { "epoch": 1.450021682567216, "grad_norm": 2.1897361278533936, "learning_rate": 1.0370665892934862e-05, "loss": 0.0297, "step": 13375 }, { "epoch": 1.4501300954032956, "grad_norm": 0.5170254111289978, "learning_rate": 1.0369940519367474e-05, "loss": 0.0024, "step": 13376 }, { "epoch": 1.4502385082393756, "grad_norm": 0.6781315803527832, "learning_rate": 1.036921514580009e-05, "loss": 0.0136, "step": 13377 }, { "epoch": 1.4503469210754554, "grad_norm": 6.70890474319458, "learning_rate": 1.0368489772232702e-05, "loss": 0.011, "step": 13378 }, { "epoch": 1.4504553339115351, "grad_norm": 0.6434826850891113, "learning_rate": 1.0367764398665314e-05, "loss": 0.0127, "step": 13379 }, { "epoch": 1.450563746747615, "grad_norm": 2.5327627658843994, "learning_rate": 1.0367039025097928e-05, "loss": 0.0455, "step": 13380 }, { "epoch": 1.4506721595836947, "grad_norm": 2.4176738262176514, "learning_rate": 1.036631365153054e-05, "loss": 0.0164, "step": 13381 }, { "epoch": 1.4507805724197744, "grad_norm": 0.1812494397163391, "learning_rate": 1.0365588277963152e-05, "loss": 0.0042, "step": 13382 }, { "epoch": 1.4508889852558542, "grad_norm": 2.389768362045288, "learning_rate": 1.0364862904395766e-05, "loss": 0.0123, "step": 13383 }, { "epoch": 1.4509973980919342, "grad_norm": 0.0398842953145504, "learning_rate": 1.0364137530828378e-05, "loss": 0.0001, "step": 13384 }, { "epoch": 1.451105810928014, "grad_norm": 2.5042436122894287, "learning_rate": 1.036341215726099e-05, "loss": 0.0226, "step": 13385 }, { "epoch": 1.4512142237640937, "grad_norm": 7.268534183502197, "learning_rate": 1.0362686783693602e-05, "loss": 0.0137, "step": 13386 }, { "epoch": 1.4513226366001735, "grad_norm": 0.3908085823059082, "learning_rate": 1.0361961410126216e-05, "loss": 0.007, "step": 13387 }, { "epoch": 1.4514310494362532, "grad_norm": 0.5766713619232178, "learning_rate": 1.0361236036558828e-05, "loss": 0.0023, "step": 13388 }, { "epoch": 1.451539462272333, "grad_norm": 2.574855327606201, "learning_rate": 1.036051066299144e-05, "loss": 0.049, "step": 13389 }, { "epoch": 1.4516478751084128, "grad_norm": 3.5908732414245605, "learning_rate": 1.0359785289424054e-05, "loss": 0.0339, "step": 13390 }, { "epoch": 1.4517562879444927, "grad_norm": 1.3427973985671997, "learning_rate": 1.0359059915856666e-05, "loss": 0.034, "step": 13391 }, { "epoch": 1.4518647007805723, "grad_norm": 1.8248565196990967, "learning_rate": 1.0358334542289278e-05, "loss": 0.0208, "step": 13392 }, { "epoch": 1.4519731136166523, "grad_norm": 0.575787365436554, "learning_rate": 1.0357609168721892e-05, "loss": 0.0204, "step": 13393 }, { "epoch": 1.452081526452732, "grad_norm": 3.2733569145202637, "learning_rate": 1.0356883795154506e-05, "loss": 0.0138, "step": 13394 }, { "epoch": 1.4521899392888118, "grad_norm": 2.1493027210235596, "learning_rate": 1.035615842158712e-05, "loss": 0.0095, "step": 13395 }, { "epoch": 1.4522983521248916, "grad_norm": 2.5756795406341553, "learning_rate": 1.0355433048019732e-05, "loss": 0.0293, "step": 13396 }, { "epoch": 1.4524067649609713, "grad_norm": 1.8452988862991333, "learning_rate": 1.0354707674452344e-05, "loss": 0.0123, "step": 13397 }, { "epoch": 1.4525151777970513, "grad_norm": 0.6882520318031311, "learning_rate": 1.0353982300884956e-05, "loss": 0.0109, "step": 13398 }, { "epoch": 1.4526235906331308, "grad_norm": 5.236673831939697, "learning_rate": 1.035325692731757e-05, "loss": 0.0147, "step": 13399 }, { "epoch": 1.4527320034692108, "grad_norm": 0.8311687707901001, "learning_rate": 1.0352531553750182e-05, "loss": 0.0046, "step": 13400 }, { "epoch": 1.4528404163052906, "grad_norm": 12.687910079956055, "learning_rate": 1.0351806180182794e-05, "loss": 0.0492, "step": 13401 }, { "epoch": 1.4529488291413704, "grad_norm": 0.5615220665931702, "learning_rate": 1.0351080806615408e-05, "loss": 0.0158, "step": 13402 }, { "epoch": 1.4530572419774501, "grad_norm": 0.9733361601829529, "learning_rate": 1.035035543304802e-05, "loss": 0.0131, "step": 13403 }, { "epoch": 1.4531656548135299, "grad_norm": 1.1590861082077026, "learning_rate": 1.0349630059480633e-05, "loss": 0.0224, "step": 13404 }, { "epoch": 1.4532740676496096, "grad_norm": 6.0628814697265625, "learning_rate": 1.0348904685913246e-05, "loss": 0.0192, "step": 13405 }, { "epoch": 1.4533824804856894, "grad_norm": 0.37508177757263184, "learning_rate": 1.0348179312345859e-05, "loss": 0.0011, "step": 13406 }, { "epoch": 1.4534908933217694, "grad_norm": 0.65721595287323, "learning_rate": 1.034745393877847e-05, "loss": 0.0155, "step": 13407 }, { "epoch": 1.4535993061578492, "grad_norm": 2.3597822189331055, "learning_rate": 1.0346728565211085e-05, "loss": 0.0224, "step": 13408 }, { "epoch": 1.453707718993929, "grad_norm": 5.221830368041992, "learning_rate": 1.0346003191643697e-05, "loss": 0.0391, "step": 13409 }, { "epoch": 1.4538161318300087, "grad_norm": 4.801540374755859, "learning_rate": 1.0345277818076309e-05, "loss": 0.0376, "step": 13410 }, { "epoch": 1.4539245446660884, "grad_norm": 0.2014438509941101, "learning_rate": 1.0344552444508924e-05, "loss": 0.001, "step": 13411 }, { "epoch": 1.4540329575021682, "grad_norm": 1.2817957401275635, "learning_rate": 1.0343827070941537e-05, "loss": 0.0039, "step": 13412 }, { "epoch": 1.454141370338248, "grad_norm": 1.303157925605774, "learning_rate": 1.0343101697374149e-05, "loss": 0.0054, "step": 13413 }, { "epoch": 1.454249783174328, "grad_norm": 0.11611157655715942, "learning_rate": 1.0342376323806763e-05, "loss": 0.0006, "step": 13414 }, { "epoch": 1.4543581960104075, "grad_norm": 0.3092537224292755, "learning_rate": 1.0341650950239375e-05, "loss": 0.0096, "step": 13415 }, { "epoch": 1.4544666088464875, "grad_norm": 0.7654658555984497, "learning_rate": 1.0340925576671987e-05, "loss": 0.0039, "step": 13416 }, { "epoch": 1.4545750216825672, "grad_norm": 6.13742733001709, "learning_rate": 1.03402002031046e-05, "loss": 0.0259, "step": 13417 }, { "epoch": 1.454683434518647, "grad_norm": 4.0880208015441895, "learning_rate": 1.0339474829537213e-05, "loss": 0.0327, "step": 13418 }, { "epoch": 1.4547918473547268, "grad_norm": 0.637058436870575, "learning_rate": 1.0338749455969825e-05, "loss": 0.0127, "step": 13419 }, { "epoch": 1.4549002601908065, "grad_norm": 1.8986577987670898, "learning_rate": 1.0338024082402439e-05, "loss": 0.0591, "step": 13420 }, { "epoch": 1.4550086730268863, "grad_norm": 2.562727928161621, "learning_rate": 1.0337298708835051e-05, "loss": 0.01, "step": 13421 }, { "epoch": 1.455117085862966, "grad_norm": 3.8914072513580322, "learning_rate": 1.0336573335267663e-05, "loss": 0.0109, "step": 13422 }, { "epoch": 1.455225498699046, "grad_norm": 0.7385918498039246, "learning_rate": 1.0335847961700277e-05, "loss": 0.0311, "step": 13423 }, { "epoch": 1.4553339115351258, "grad_norm": 0.2526257038116455, "learning_rate": 1.0335122588132889e-05, "loss": 0.003, "step": 13424 }, { "epoch": 1.4554423243712056, "grad_norm": 0.6817057728767395, "learning_rate": 1.0334397214565501e-05, "loss": 0.0191, "step": 13425 }, { "epoch": 1.4555507372072853, "grad_norm": 0.0728871077299118, "learning_rate": 1.0333671840998113e-05, "loss": 0.0007, "step": 13426 }, { "epoch": 1.455659150043365, "grad_norm": 6.332141399383545, "learning_rate": 1.0332946467430727e-05, "loss": 0.038, "step": 13427 }, { "epoch": 1.4557675628794449, "grad_norm": 4.362967014312744, "learning_rate": 1.033222109386334e-05, "loss": 0.0486, "step": 13428 }, { "epoch": 1.4558759757155246, "grad_norm": 1.1278555393218994, "learning_rate": 1.0331495720295955e-05, "loss": 0.0082, "step": 13429 }, { "epoch": 1.4559843885516046, "grad_norm": 0.19973039627075195, "learning_rate": 1.0330770346728567e-05, "loss": 0.0012, "step": 13430 }, { "epoch": 1.4560928013876844, "grad_norm": 0.15712587535381317, "learning_rate": 1.0330044973161179e-05, "loss": 0.0013, "step": 13431 }, { "epoch": 1.4562012142237641, "grad_norm": 0.0523340068757534, "learning_rate": 1.0329319599593793e-05, "loss": 0.0004, "step": 13432 }, { "epoch": 1.456309627059844, "grad_norm": 2.6876678466796875, "learning_rate": 1.0328594226026405e-05, "loss": 0.0466, "step": 13433 }, { "epoch": 1.4564180398959237, "grad_norm": 1.1958248615264893, "learning_rate": 1.0327868852459017e-05, "loss": 0.0151, "step": 13434 }, { "epoch": 1.4565264527320034, "grad_norm": 9.732090950012207, "learning_rate": 1.0327143478891631e-05, "loss": 0.0065, "step": 13435 }, { "epoch": 1.4566348655680832, "grad_norm": 1.054110050201416, "learning_rate": 1.0326418105324243e-05, "loss": 0.0033, "step": 13436 }, { "epoch": 1.4567432784041632, "grad_norm": 8.051785469055176, "learning_rate": 1.0325692731756855e-05, "loss": 0.0517, "step": 13437 }, { "epoch": 1.4568516912402427, "grad_norm": 2.844435691833496, "learning_rate": 1.0324967358189468e-05, "loss": 0.0083, "step": 13438 }, { "epoch": 1.4569601040763227, "grad_norm": 10.95680046081543, "learning_rate": 1.0324241984622081e-05, "loss": 0.0238, "step": 13439 }, { "epoch": 1.4570685169124025, "grad_norm": 0.7133119106292725, "learning_rate": 1.0323516611054693e-05, "loss": 0.0057, "step": 13440 }, { "epoch": 1.4571769297484822, "grad_norm": 1.0167953968048096, "learning_rate": 1.0322791237487306e-05, "loss": 0.0049, "step": 13441 }, { "epoch": 1.457285342584562, "grad_norm": 3.8608295917510986, "learning_rate": 1.032206586391992e-05, "loss": 0.0238, "step": 13442 }, { "epoch": 1.4573937554206418, "grad_norm": 1.9779317378997803, "learning_rate": 1.0321340490352532e-05, "loss": 0.0178, "step": 13443 }, { "epoch": 1.4575021682567215, "grad_norm": 1.5681042671203613, "learning_rate": 1.0320615116785144e-05, "loss": 0.0267, "step": 13444 }, { "epoch": 1.4576105810928013, "grad_norm": 4.524564266204834, "learning_rate": 1.0319889743217758e-05, "loss": 0.0153, "step": 13445 }, { "epoch": 1.4577189939288813, "grad_norm": 1.8362287282943726, "learning_rate": 1.0319164369650371e-05, "loss": 0.0104, "step": 13446 }, { "epoch": 1.457827406764961, "grad_norm": 0.23286668956279755, "learning_rate": 1.0318438996082985e-05, "loss": 0.005, "step": 13447 }, { "epoch": 1.4579358196010408, "grad_norm": 0.28588956594467163, "learning_rate": 1.0317713622515597e-05, "loss": 0.0063, "step": 13448 }, { "epoch": 1.4580442324371206, "grad_norm": 2.4274909496307373, "learning_rate": 1.031698824894821e-05, "loss": 0.0124, "step": 13449 }, { "epoch": 1.4581526452732003, "grad_norm": 2.7718679904937744, "learning_rate": 1.0316262875380822e-05, "loss": 0.0131, "step": 13450 }, { "epoch": 1.45826105810928, "grad_norm": 12.229447364807129, "learning_rate": 1.0315537501813436e-05, "loss": 0.0409, "step": 13451 }, { "epoch": 1.4583694709453598, "grad_norm": 2.240480899810791, "learning_rate": 1.0314812128246048e-05, "loss": 0.0625, "step": 13452 }, { "epoch": 1.4584778837814398, "grad_norm": 1.8185817003250122, "learning_rate": 1.031408675467866e-05, "loss": 0.016, "step": 13453 }, { "epoch": 1.4585862966175196, "grad_norm": 0.7182130217552185, "learning_rate": 1.0313361381111274e-05, "loss": 0.0087, "step": 13454 }, { "epoch": 1.4586947094535994, "grad_norm": 0.26858070492744446, "learning_rate": 1.0312636007543886e-05, "loss": 0.0059, "step": 13455 }, { "epoch": 1.4588031222896791, "grad_norm": 11.097847938537598, "learning_rate": 1.0311910633976498e-05, "loss": 0.0349, "step": 13456 }, { "epoch": 1.4589115351257589, "grad_norm": 3.556274652481079, "learning_rate": 1.0311185260409112e-05, "loss": 0.0356, "step": 13457 }, { "epoch": 1.4590199479618386, "grad_norm": 1.3266407251358032, "learning_rate": 1.0310459886841724e-05, "loss": 0.0219, "step": 13458 }, { "epoch": 1.4591283607979184, "grad_norm": 8.740081787109375, "learning_rate": 1.0309734513274336e-05, "loss": 0.0418, "step": 13459 }, { "epoch": 1.4592367736339984, "grad_norm": 1.6009819507598877, "learning_rate": 1.030900913970695e-05, "loss": 0.0454, "step": 13460 }, { "epoch": 1.459345186470078, "grad_norm": 0.6199125647544861, "learning_rate": 1.0308283766139562e-05, "loss": 0.0055, "step": 13461 }, { "epoch": 1.459453599306158, "grad_norm": 3.0225188732147217, "learning_rate": 1.0307558392572174e-05, "loss": 0.0709, "step": 13462 }, { "epoch": 1.4595620121422377, "grad_norm": 2.5308523178100586, "learning_rate": 1.030683301900479e-05, "loss": 0.0676, "step": 13463 }, { "epoch": 1.4596704249783174, "grad_norm": 0.8421883583068848, "learning_rate": 1.0306107645437402e-05, "loss": 0.016, "step": 13464 }, { "epoch": 1.4597788378143972, "grad_norm": 0.9476563930511475, "learning_rate": 1.0305382271870014e-05, "loss": 0.0354, "step": 13465 }, { "epoch": 1.459887250650477, "grad_norm": 2.153853416442871, "learning_rate": 1.0304656898302628e-05, "loss": 0.0518, "step": 13466 }, { "epoch": 1.4599956634865567, "grad_norm": 4.193877220153809, "learning_rate": 1.030393152473524e-05, "loss": 0.0277, "step": 13467 }, { "epoch": 1.4601040763226365, "grad_norm": 1.0240395069122314, "learning_rate": 1.0303206151167852e-05, "loss": 0.0162, "step": 13468 }, { "epoch": 1.4602124891587165, "grad_norm": 12.08859634399414, "learning_rate": 1.0302480777600466e-05, "loss": 0.0305, "step": 13469 }, { "epoch": 1.4603209019947962, "grad_norm": 2.21737003326416, "learning_rate": 1.0301755404033078e-05, "loss": 0.043, "step": 13470 }, { "epoch": 1.460429314830876, "grad_norm": 0.8588833212852478, "learning_rate": 1.030103003046569e-05, "loss": 0.0206, "step": 13471 }, { "epoch": 1.4605377276669558, "grad_norm": 0.21813614666461945, "learning_rate": 1.0300304656898304e-05, "loss": 0.0047, "step": 13472 }, { "epoch": 1.4606461405030355, "grad_norm": 0.13877762854099274, "learning_rate": 1.0299579283330916e-05, "loss": 0.0012, "step": 13473 }, { "epoch": 1.4607545533391153, "grad_norm": 1.8419848680496216, "learning_rate": 1.0298853909763528e-05, "loss": 0.0154, "step": 13474 }, { "epoch": 1.460862966175195, "grad_norm": 1.6490373611450195, "learning_rate": 1.0298128536196142e-05, "loss": 0.0306, "step": 13475 }, { "epoch": 1.460971379011275, "grad_norm": 2.2106435298919678, "learning_rate": 1.0297403162628754e-05, "loss": 0.0103, "step": 13476 }, { "epoch": 1.4610797918473546, "grad_norm": 0.9483835101127625, "learning_rate": 1.0296677789061367e-05, "loss": 0.0226, "step": 13477 }, { "epoch": 1.4611882046834346, "grad_norm": 1.4240400791168213, "learning_rate": 1.0295952415493979e-05, "loss": 0.0187, "step": 13478 }, { "epoch": 1.4612966175195143, "grad_norm": 1.4506704807281494, "learning_rate": 1.0295227041926592e-05, "loss": 0.0114, "step": 13479 }, { "epoch": 1.461405030355594, "grad_norm": 0.8002123236656189, "learning_rate": 1.0294501668359205e-05, "loss": 0.0167, "step": 13480 }, { "epoch": 1.4615134431916739, "grad_norm": 3.124519109725952, "learning_rate": 1.029377629479182e-05, "loss": 0.0231, "step": 13481 }, { "epoch": 1.4616218560277536, "grad_norm": 0.34235596656799316, "learning_rate": 1.0293050921224432e-05, "loss": 0.0087, "step": 13482 }, { "epoch": 1.4617302688638336, "grad_norm": 2.539184808731079, "learning_rate": 1.0292325547657044e-05, "loss": 0.0314, "step": 13483 }, { "epoch": 1.4618386816999132, "grad_norm": 1.6830257177352905, "learning_rate": 1.0291600174089658e-05, "loss": 0.0166, "step": 13484 }, { "epoch": 1.4619470945359931, "grad_norm": 4.281980514526367, "learning_rate": 1.029087480052227e-05, "loss": 0.0229, "step": 13485 }, { "epoch": 1.462055507372073, "grad_norm": 0.6630450487136841, "learning_rate": 1.0290149426954883e-05, "loss": 0.0153, "step": 13486 }, { "epoch": 1.4621639202081527, "grad_norm": 2.3498027324676514, "learning_rate": 1.0289424053387496e-05, "loss": 0.0348, "step": 13487 }, { "epoch": 1.4622723330442324, "grad_norm": 0.9702057838439941, "learning_rate": 1.0288698679820109e-05, "loss": 0.0054, "step": 13488 }, { "epoch": 1.4623807458803122, "grad_norm": 5.5265960693359375, "learning_rate": 1.028797330625272e-05, "loss": 0.0384, "step": 13489 }, { "epoch": 1.462489158716392, "grad_norm": 0.49374639987945557, "learning_rate": 1.0287247932685333e-05, "loss": 0.0142, "step": 13490 }, { "epoch": 1.4625975715524717, "grad_norm": 0.2968825101852417, "learning_rate": 1.0286522559117947e-05, "loss": 0.0013, "step": 13491 }, { "epoch": 1.4627059843885517, "grad_norm": 4.659385681152344, "learning_rate": 1.0285797185550559e-05, "loss": 0.0411, "step": 13492 }, { "epoch": 1.4628143972246315, "grad_norm": 0.3662410080432892, "learning_rate": 1.0285071811983171e-05, "loss": 0.0082, "step": 13493 }, { "epoch": 1.4629228100607112, "grad_norm": 0.5217183232307434, "learning_rate": 1.0284346438415785e-05, "loss": 0.0196, "step": 13494 }, { "epoch": 1.463031222896791, "grad_norm": 1.8700058460235596, "learning_rate": 1.0283621064848397e-05, "loss": 0.0476, "step": 13495 }, { "epoch": 1.4631396357328708, "grad_norm": 0.5536752343177795, "learning_rate": 1.0282895691281009e-05, "loss": 0.0026, "step": 13496 }, { "epoch": 1.4632480485689505, "grad_norm": 1.5694550275802612, "learning_rate": 1.0282170317713623e-05, "loss": 0.0204, "step": 13497 }, { "epoch": 1.4633564614050303, "grad_norm": 0.12949341535568237, "learning_rate": 1.0281444944146237e-05, "loss": 0.0009, "step": 13498 }, { "epoch": 1.4634648742411103, "grad_norm": 7.843405246734619, "learning_rate": 1.028071957057885e-05, "loss": 0.022, "step": 13499 }, { "epoch": 1.4635732870771898, "grad_norm": 3.2351934909820557, "learning_rate": 1.0279994197011463e-05, "loss": 0.0342, "step": 13500 }, { "epoch": 1.4636816999132698, "grad_norm": 2.433819055557251, "learning_rate": 1.0279268823444075e-05, "loss": 0.0326, "step": 13501 }, { "epoch": 1.4637901127493496, "grad_norm": 0.5039979815483093, "learning_rate": 1.0278543449876687e-05, "loss": 0.0017, "step": 13502 }, { "epoch": 1.4638985255854293, "grad_norm": 7.048642158508301, "learning_rate": 1.0277818076309301e-05, "loss": 0.0227, "step": 13503 }, { "epoch": 1.464006938421509, "grad_norm": 0.6512885093688965, "learning_rate": 1.0277092702741913e-05, "loss": 0.0184, "step": 13504 }, { "epoch": 1.4641153512575888, "grad_norm": 2.6870622634887695, "learning_rate": 1.0276367329174525e-05, "loss": 0.049, "step": 13505 }, { "epoch": 1.4642237640936686, "grad_norm": 6.926236629486084, "learning_rate": 1.0275641955607139e-05, "loss": 0.0602, "step": 13506 }, { "epoch": 1.4643321769297484, "grad_norm": 0.46192020177841187, "learning_rate": 1.0274916582039751e-05, "loss": 0.0062, "step": 13507 }, { "epoch": 1.4644405897658284, "grad_norm": 0.9429023265838623, "learning_rate": 1.0274191208472363e-05, "loss": 0.0182, "step": 13508 }, { "epoch": 1.4645490026019081, "grad_norm": 1.7908767461776733, "learning_rate": 1.0273465834904977e-05, "loss": 0.0296, "step": 13509 }, { "epoch": 1.4646574154379879, "grad_norm": 3.1496667861938477, "learning_rate": 1.027274046133759e-05, "loss": 0.0308, "step": 13510 }, { "epoch": 1.4647658282740676, "grad_norm": 2.855433464050293, "learning_rate": 1.0272015087770201e-05, "loss": 0.0489, "step": 13511 }, { "epoch": 1.4648742411101474, "grad_norm": 3.6502180099487305, "learning_rate": 1.0271289714202815e-05, "loss": 0.0056, "step": 13512 }, { "epoch": 1.4649826539462272, "grad_norm": 2.037581205368042, "learning_rate": 1.0270564340635427e-05, "loss": 0.0129, "step": 13513 }, { "epoch": 1.465091066782307, "grad_norm": 4.164394855499268, "learning_rate": 1.026983896706804e-05, "loss": 0.0757, "step": 13514 }, { "epoch": 1.465199479618387, "grad_norm": 1.488874077796936, "learning_rate": 1.0269113593500655e-05, "loss": 0.0121, "step": 13515 }, { "epoch": 1.4653078924544667, "grad_norm": 0.14614108204841614, "learning_rate": 1.0268388219933267e-05, "loss": 0.0013, "step": 13516 }, { "epoch": 1.4654163052905465, "grad_norm": 0.1246655285358429, "learning_rate": 1.026766284636588e-05, "loss": 0.0012, "step": 13517 }, { "epoch": 1.4655247181266262, "grad_norm": 1.311924695968628, "learning_rate": 1.0266937472798493e-05, "loss": 0.0176, "step": 13518 }, { "epoch": 1.465633130962706, "grad_norm": 1.2324022054672241, "learning_rate": 1.0266212099231105e-05, "loss": 0.0066, "step": 13519 }, { "epoch": 1.4657415437987857, "grad_norm": 3.717362642288208, "learning_rate": 1.0265486725663717e-05, "loss": 0.0399, "step": 13520 }, { "epoch": 1.4658499566348655, "grad_norm": 1.7473145723342896, "learning_rate": 1.0264761352096331e-05, "loss": 0.0059, "step": 13521 }, { "epoch": 1.4659583694709455, "grad_norm": 6.71177339553833, "learning_rate": 1.0264035978528943e-05, "loss": 0.0346, "step": 13522 }, { "epoch": 1.466066782307025, "grad_norm": 1.0817989110946655, "learning_rate": 1.0263310604961556e-05, "loss": 0.0134, "step": 13523 }, { "epoch": 1.466175195143105, "grad_norm": 9.139788627624512, "learning_rate": 1.026258523139417e-05, "loss": 0.0656, "step": 13524 }, { "epoch": 1.4662836079791848, "grad_norm": 0.8911800384521484, "learning_rate": 1.0261859857826782e-05, "loss": 0.0139, "step": 13525 }, { "epoch": 1.4663920208152645, "grad_norm": 0.38443127274513245, "learning_rate": 1.0261134484259394e-05, "loss": 0.0028, "step": 13526 }, { "epoch": 1.4665004336513443, "grad_norm": 1.960610270500183, "learning_rate": 1.0260409110692008e-05, "loss": 0.0139, "step": 13527 }, { "epoch": 1.466608846487424, "grad_norm": 2.8385767936706543, "learning_rate": 1.025968373712462e-05, "loss": 0.0308, "step": 13528 }, { "epoch": 1.4667172593235038, "grad_norm": 3.9050707817077637, "learning_rate": 1.0258958363557232e-05, "loss": 0.0592, "step": 13529 }, { "epoch": 1.4668256721595836, "grad_norm": 1.6853519678115845, "learning_rate": 1.0258232989989844e-05, "loss": 0.0094, "step": 13530 }, { "epoch": 1.4669340849956636, "grad_norm": 0.7381689548492432, "learning_rate": 1.0257507616422458e-05, "loss": 0.0183, "step": 13531 }, { "epoch": 1.4670424978317433, "grad_norm": 1.4241876602172852, "learning_rate": 1.0256782242855072e-05, "loss": 0.0138, "step": 13532 }, { "epoch": 1.467150910667823, "grad_norm": 1.3819855451583862, "learning_rate": 1.0256056869287686e-05, "loss": 0.0192, "step": 13533 }, { "epoch": 1.4672593235039029, "grad_norm": 0.24318727850914001, "learning_rate": 1.0255331495720298e-05, "loss": 0.0014, "step": 13534 }, { "epoch": 1.4673677363399826, "grad_norm": 5.6090593338012695, "learning_rate": 1.025460612215291e-05, "loss": 0.0358, "step": 13535 }, { "epoch": 1.4674761491760624, "grad_norm": 4.398367404937744, "learning_rate": 1.0253880748585524e-05, "loss": 0.0696, "step": 13536 }, { "epoch": 1.4675845620121422, "grad_norm": 1.021519660949707, "learning_rate": 1.0253155375018136e-05, "loss": 0.0129, "step": 13537 }, { "epoch": 1.4676929748482221, "grad_norm": 1.4811267852783203, "learning_rate": 1.0252430001450748e-05, "loss": 0.0462, "step": 13538 }, { "epoch": 1.4678013876843017, "grad_norm": 3.1237881183624268, "learning_rate": 1.0251704627883362e-05, "loss": 0.0175, "step": 13539 }, { "epoch": 1.4679098005203817, "grad_norm": 1.3637831211090088, "learning_rate": 1.0250979254315974e-05, "loss": 0.01, "step": 13540 }, { "epoch": 1.4680182133564614, "grad_norm": 1.3179235458374023, "learning_rate": 1.0250253880748586e-05, "loss": 0.0127, "step": 13541 }, { "epoch": 1.4681266261925412, "grad_norm": 1.0233492851257324, "learning_rate": 1.0249528507181198e-05, "loss": 0.0022, "step": 13542 }, { "epoch": 1.468235039028621, "grad_norm": 1.240057349205017, "learning_rate": 1.0248803133613812e-05, "loss": 0.0047, "step": 13543 }, { "epoch": 1.4683434518647007, "grad_norm": 4.5914483070373535, "learning_rate": 1.0248077760046424e-05, "loss": 0.0234, "step": 13544 }, { "epoch": 1.4684518647007807, "grad_norm": 1.5192800760269165, "learning_rate": 1.0247352386479036e-05, "loss": 0.0169, "step": 13545 }, { "epoch": 1.4685602775368602, "grad_norm": 0.1953750103712082, "learning_rate": 1.024662701291165e-05, "loss": 0.0022, "step": 13546 }, { "epoch": 1.4686686903729402, "grad_norm": 0.4764021635055542, "learning_rate": 1.0245901639344262e-05, "loss": 0.0021, "step": 13547 }, { "epoch": 1.46877710320902, "grad_norm": 0.21487565338611603, "learning_rate": 1.0245176265776874e-05, "loss": 0.0011, "step": 13548 }, { "epoch": 1.4688855160450998, "grad_norm": 2.5748157501220703, "learning_rate": 1.0244450892209488e-05, "loss": 0.0193, "step": 13549 }, { "epoch": 1.4689939288811795, "grad_norm": 5.129122734069824, "learning_rate": 1.0243725518642102e-05, "loss": 0.0076, "step": 13550 }, { "epoch": 1.4691023417172593, "grad_norm": 0.3654600977897644, "learning_rate": 1.0243000145074716e-05, "loss": 0.0018, "step": 13551 }, { "epoch": 1.469210754553339, "grad_norm": 0.9535527229309082, "learning_rate": 1.0242274771507328e-05, "loss": 0.0057, "step": 13552 }, { "epoch": 1.4693191673894188, "grad_norm": 0.5462039709091187, "learning_rate": 1.024154939793994e-05, "loss": 0.0137, "step": 13553 }, { "epoch": 1.4694275802254988, "grad_norm": 0.40584051609039307, "learning_rate": 1.0240824024372552e-05, "loss": 0.0119, "step": 13554 }, { "epoch": 1.4695359930615786, "grad_norm": 4.436368942260742, "learning_rate": 1.0240098650805166e-05, "loss": 0.0389, "step": 13555 }, { "epoch": 1.4696444058976583, "grad_norm": 9.620180130004883, "learning_rate": 1.0239373277237778e-05, "loss": 0.0195, "step": 13556 }, { "epoch": 1.469752818733738, "grad_norm": 0.31125709414482117, "learning_rate": 1.023864790367039e-05, "loss": 0.0072, "step": 13557 }, { "epoch": 1.4698612315698178, "grad_norm": 0.8081915974617004, "learning_rate": 1.0237922530103004e-05, "loss": 0.011, "step": 13558 }, { "epoch": 1.4699696444058976, "grad_norm": 0.522743284702301, "learning_rate": 1.0237197156535616e-05, "loss": 0.0176, "step": 13559 }, { "epoch": 1.4700780572419774, "grad_norm": 8.10914134979248, "learning_rate": 1.0236471782968229e-05, "loss": 0.0127, "step": 13560 }, { "epoch": 1.4701864700780574, "grad_norm": 1.860416293144226, "learning_rate": 1.0235746409400842e-05, "loss": 0.0206, "step": 13561 }, { "epoch": 1.470294882914137, "grad_norm": 0.7459973692893982, "learning_rate": 1.0235021035833455e-05, "loss": 0.0041, "step": 13562 }, { "epoch": 1.4704032957502169, "grad_norm": 1.4671368598937988, "learning_rate": 1.0234295662266067e-05, "loss": 0.0028, "step": 13563 }, { "epoch": 1.4705117085862967, "grad_norm": 7.23916482925415, "learning_rate": 1.023357028869868e-05, "loss": 0.0744, "step": 13564 }, { "epoch": 1.4706201214223764, "grad_norm": 1.067887783050537, "learning_rate": 1.0232844915131293e-05, "loss": 0.0136, "step": 13565 }, { "epoch": 1.4707285342584562, "grad_norm": 8.555538177490234, "learning_rate": 1.0232119541563905e-05, "loss": 0.0265, "step": 13566 }, { "epoch": 1.470836947094536, "grad_norm": 1.2981617450714111, "learning_rate": 1.023139416799652e-05, "loss": 0.0063, "step": 13567 }, { "epoch": 1.4709453599306157, "grad_norm": 10.607885360717773, "learning_rate": 1.0230668794429133e-05, "loss": 0.0211, "step": 13568 }, { "epoch": 1.4710537727666955, "grad_norm": 2.834594964981079, "learning_rate": 1.0229943420861745e-05, "loss": 0.0021, "step": 13569 }, { "epoch": 1.4711621856027755, "grad_norm": 0.6594696044921875, "learning_rate": 1.0229218047294359e-05, "loss": 0.0036, "step": 13570 }, { "epoch": 1.4712705984388552, "grad_norm": 0.29548630118370056, "learning_rate": 1.022849267372697e-05, "loss": 0.0086, "step": 13571 }, { "epoch": 1.471379011274935, "grad_norm": 9.153505325317383, "learning_rate": 1.0227767300159583e-05, "loss": 0.0104, "step": 13572 }, { "epoch": 1.4714874241110147, "grad_norm": 0.44600290060043335, "learning_rate": 1.0227041926592197e-05, "loss": 0.0094, "step": 13573 }, { "epoch": 1.4715958369470945, "grad_norm": 0.294338583946228, "learning_rate": 1.0226316553024809e-05, "loss": 0.0091, "step": 13574 }, { "epoch": 1.4717042497831743, "grad_norm": 3.1327381134033203, "learning_rate": 1.0225591179457421e-05, "loss": 0.0109, "step": 13575 }, { "epoch": 1.471812662619254, "grad_norm": 0.04912383109331131, "learning_rate": 1.0224865805890035e-05, "loss": 0.0003, "step": 13576 }, { "epoch": 1.471921075455334, "grad_norm": 2.9589388370513916, "learning_rate": 1.0224140432322647e-05, "loss": 0.0666, "step": 13577 }, { "epoch": 1.4720294882914138, "grad_norm": 2.4354193210601807, "learning_rate": 1.0223415058755259e-05, "loss": 0.0225, "step": 13578 }, { "epoch": 1.4721379011274935, "grad_norm": 11.762218475341797, "learning_rate": 1.0222689685187873e-05, "loss": 0.0154, "step": 13579 }, { "epoch": 1.4722463139635733, "grad_norm": 8.784704208374023, "learning_rate": 1.0221964311620485e-05, "loss": 0.08, "step": 13580 }, { "epoch": 1.472354726799653, "grad_norm": 2.8590922355651855, "learning_rate": 1.0221238938053097e-05, "loss": 0.0503, "step": 13581 }, { "epoch": 1.4724631396357328, "grad_norm": 1.9234074354171753, "learning_rate": 1.022051356448571e-05, "loss": 0.0325, "step": 13582 }, { "epoch": 1.4725715524718126, "grad_norm": 8.124296188354492, "learning_rate": 1.0219788190918323e-05, "loss": 0.0274, "step": 13583 }, { "epoch": 1.4726799653078926, "grad_norm": 1.4754765033721924, "learning_rate": 1.0219062817350937e-05, "loss": 0.0428, "step": 13584 }, { "epoch": 1.4727883781439721, "grad_norm": 4.318846225738525, "learning_rate": 1.021833744378355e-05, "loss": 0.0202, "step": 13585 }, { "epoch": 1.472896790980052, "grad_norm": 6.695424556732178, "learning_rate": 1.0217612070216163e-05, "loss": 0.0249, "step": 13586 }, { "epoch": 1.4730052038161319, "grad_norm": 0.470731645822525, "learning_rate": 1.0216886696648775e-05, "loss": 0.0133, "step": 13587 }, { "epoch": 1.4731136166522116, "grad_norm": 6.841675281524658, "learning_rate": 1.0216161323081389e-05, "loss": 0.0517, "step": 13588 }, { "epoch": 1.4732220294882914, "grad_norm": 6.427421569824219, "learning_rate": 1.0215435949514001e-05, "loss": 0.1112, "step": 13589 }, { "epoch": 1.4733304423243712, "grad_norm": 4.048673152923584, "learning_rate": 1.0214710575946613e-05, "loss": 0.0233, "step": 13590 }, { "epoch": 1.473438855160451, "grad_norm": 0.09215247631072998, "learning_rate": 1.0213985202379227e-05, "loss": 0.0005, "step": 13591 }, { "epoch": 1.4735472679965307, "grad_norm": 4.63618803024292, "learning_rate": 1.021325982881184e-05, "loss": 0.0274, "step": 13592 }, { "epoch": 1.4736556808326107, "grad_norm": 4.064499378204346, "learning_rate": 1.0212534455244451e-05, "loss": 0.0144, "step": 13593 }, { "epoch": 1.4737640936686904, "grad_norm": 7.03343391418457, "learning_rate": 1.0211809081677064e-05, "loss": 0.0289, "step": 13594 }, { "epoch": 1.4738725065047702, "grad_norm": 0.7054116725921631, "learning_rate": 1.0211083708109677e-05, "loss": 0.0026, "step": 13595 }, { "epoch": 1.47398091934085, "grad_norm": 1.3229643106460571, "learning_rate": 1.021035833454229e-05, "loss": 0.025, "step": 13596 }, { "epoch": 1.4740893321769297, "grad_norm": 0.40751567482948303, "learning_rate": 1.0209632960974902e-05, "loss": 0.0155, "step": 13597 }, { "epoch": 1.4741977450130095, "grad_norm": 0.8480643630027771, "learning_rate": 1.0208907587407515e-05, "loss": 0.0029, "step": 13598 }, { "epoch": 1.4743061578490892, "grad_norm": 2.1717705726623535, "learning_rate": 1.0208182213840128e-05, "loss": 0.0619, "step": 13599 }, { "epoch": 1.4744145706851692, "grad_norm": 5.957804203033447, "learning_rate": 1.020745684027274e-05, "loss": 0.0125, "step": 13600 }, { "epoch": 1.474522983521249, "grad_norm": 3.7017264366149902, "learning_rate": 1.0206731466705355e-05, "loss": 0.0263, "step": 13601 }, { "epoch": 1.4746313963573288, "grad_norm": 0.28409716486930847, "learning_rate": 1.0206006093137967e-05, "loss": 0.0102, "step": 13602 }, { "epoch": 1.4747398091934085, "grad_norm": 1.3075554370880127, "learning_rate": 1.0205280719570581e-05, "loss": 0.0232, "step": 13603 }, { "epoch": 1.4748482220294883, "grad_norm": 3.5019490718841553, "learning_rate": 1.0204555346003193e-05, "loss": 0.0845, "step": 13604 }, { "epoch": 1.474956634865568, "grad_norm": 2.9096786975860596, "learning_rate": 1.0203829972435806e-05, "loss": 0.0339, "step": 13605 }, { "epoch": 1.4750650477016478, "grad_norm": 1.9000768661499023, "learning_rate": 1.0203104598868418e-05, "loss": 0.0509, "step": 13606 }, { "epoch": 1.4751734605377278, "grad_norm": 3.579679012298584, "learning_rate": 1.0202379225301032e-05, "loss": 0.0661, "step": 13607 }, { "epoch": 1.4752818733738073, "grad_norm": 8.206664085388184, "learning_rate": 1.0201653851733644e-05, "loss": 0.0389, "step": 13608 }, { "epoch": 1.4753902862098873, "grad_norm": 0.4935655891895294, "learning_rate": 1.0200928478166256e-05, "loss": 0.0072, "step": 13609 }, { "epoch": 1.475498699045967, "grad_norm": 5.505314826965332, "learning_rate": 1.020020310459887e-05, "loss": 0.0833, "step": 13610 }, { "epoch": 1.4756071118820469, "grad_norm": 4.032709121704102, "learning_rate": 1.0199477731031482e-05, "loss": 0.0478, "step": 13611 }, { "epoch": 1.4757155247181266, "grad_norm": 2.6003530025482178, "learning_rate": 1.0198752357464094e-05, "loss": 0.0188, "step": 13612 }, { "epoch": 1.4758239375542064, "grad_norm": 0.46686801314353943, "learning_rate": 1.0198026983896708e-05, "loss": 0.0086, "step": 13613 }, { "epoch": 1.4759323503902861, "grad_norm": 1.761765480041504, "learning_rate": 1.019730161032932e-05, "loss": 0.0191, "step": 13614 }, { "epoch": 1.476040763226366, "grad_norm": 2.7600605487823486, "learning_rate": 1.0196576236761932e-05, "loss": 0.0177, "step": 13615 }, { "epoch": 1.476149176062446, "grad_norm": 0.2542237639427185, "learning_rate": 1.0195850863194546e-05, "loss": 0.0017, "step": 13616 }, { "epoch": 1.4762575888985257, "grad_norm": 3.8380777835845947, "learning_rate": 1.0195125489627158e-05, "loss": 0.0333, "step": 13617 }, { "epoch": 1.4763660017346054, "grad_norm": 0.23556479811668396, "learning_rate": 1.019440011605977e-05, "loss": 0.0019, "step": 13618 }, { "epoch": 1.4764744145706852, "grad_norm": 3.0945911407470703, "learning_rate": 1.0193674742492386e-05, "loss": 0.0308, "step": 13619 }, { "epoch": 1.476582827406765, "grad_norm": 0.7110846638679504, "learning_rate": 1.0192949368924998e-05, "loss": 0.0049, "step": 13620 }, { "epoch": 1.4766912402428447, "grad_norm": 0.09327814728021622, "learning_rate": 1.019222399535761e-05, "loss": 0.0007, "step": 13621 }, { "epoch": 1.4767996530789245, "grad_norm": 0.39297887682914734, "learning_rate": 1.0191498621790224e-05, "loss": 0.0097, "step": 13622 }, { "epoch": 1.4769080659150045, "grad_norm": 0.9196930527687073, "learning_rate": 1.0190773248222836e-05, "loss": 0.0081, "step": 13623 }, { "epoch": 1.477016478751084, "grad_norm": 1.4899924993515015, "learning_rate": 1.0190047874655448e-05, "loss": 0.048, "step": 13624 }, { "epoch": 1.477124891587164, "grad_norm": 2.009028196334839, "learning_rate": 1.0189322501088062e-05, "loss": 0.0236, "step": 13625 }, { "epoch": 1.4772333044232437, "grad_norm": 2.9177865982055664, "learning_rate": 1.0188597127520674e-05, "loss": 0.04, "step": 13626 }, { "epoch": 1.4773417172593235, "grad_norm": 2.007664203643799, "learning_rate": 1.0187871753953286e-05, "loss": 0.0087, "step": 13627 }, { "epoch": 1.4774501300954033, "grad_norm": 0.5523643493652344, "learning_rate": 1.01871463803859e-05, "loss": 0.0237, "step": 13628 }, { "epoch": 1.477558542931483, "grad_norm": 2.9070065021514893, "learning_rate": 1.0186421006818512e-05, "loss": 0.0404, "step": 13629 }, { "epoch": 1.477666955767563, "grad_norm": 1.638540506362915, "learning_rate": 1.0185695633251124e-05, "loss": 0.0308, "step": 13630 }, { "epoch": 1.4777753686036426, "grad_norm": 1.3316619396209717, "learning_rate": 1.0184970259683738e-05, "loss": 0.0137, "step": 13631 }, { "epoch": 1.4778837814397225, "grad_norm": 2.1452553272247314, "learning_rate": 1.018424488611635e-05, "loss": 0.0263, "step": 13632 }, { "epoch": 1.4779921942758023, "grad_norm": 1.1143889427185059, "learning_rate": 1.0183519512548963e-05, "loss": 0.0111, "step": 13633 }, { "epoch": 1.478100607111882, "grad_norm": 0.6474800705909729, "learning_rate": 1.0182794138981575e-05, "loss": 0.0044, "step": 13634 }, { "epoch": 1.4782090199479618, "grad_norm": 0.9184459447860718, "learning_rate": 1.0182068765414189e-05, "loss": 0.018, "step": 13635 }, { "epoch": 1.4783174327840416, "grad_norm": 3.1463305950164795, "learning_rate": 1.0181343391846802e-05, "loss": 0.0136, "step": 13636 }, { "epoch": 1.4784258456201214, "grad_norm": 1.9359099864959717, "learning_rate": 1.0180618018279416e-05, "loss": 0.0432, "step": 13637 }, { "epoch": 1.4785342584562011, "grad_norm": 3.1788506507873535, "learning_rate": 1.0179892644712028e-05, "loss": 0.0437, "step": 13638 }, { "epoch": 1.478642671292281, "grad_norm": 1.7080880403518677, "learning_rate": 1.017916727114464e-05, "loss": 0.0242, "step": 13639 }, { "epoch": 1.4787510841283609, "grad_norm": 2.4962799549102783, "learning_rate": 1.0178441897577254e-05, "loss": 0.028, "step": 13640 }, { "epoch": 1.4788594969644406, "grad_norm": 1.1202187538146973, "learning_rate": 1.0177716524009866e-05, "loss": 0.0373, "step": 13641 }, { "epoch": 1.4789679098005204, "grad_norm": 0.3213821351528168, "learning_rate": 1.0176991150442479e-05, "loss": 0.0027, "step": 13642 }, { "epoch": 1.4790763226366002, "grad_norm": 1.6726617813110352, "learning_rate": 1.0176265776875092e-05, "loss": 0.0091, "step": 13643 }, { "epoch": 1.47918473547268, "grad_norm": 1.1620681285858154, "learning_rate": 1.0175540403307705e-05, "loss": 0.0175, "step": 13644 }, { "epoch": 1.4792931483087597, "grad_norm": 0.3610716164112091, "learning_rate": 1.0174815029740317e-05, "loss": 0.0035, "step": 13645 }, { "epoch": 1.4794015611448397, "grad_norm": 2.6275930404663086, "learning_rate": 1.0174089656172929e-05, "loss": 0.0288, "step": 13646 }, { "epoch": 1.4795099739809192, "grad_norm": 2.788815498352051, "learning_rate": 1.0173364282605543e-05, "loss": 0.0372, "step": 13647 }, { "epoch": 1.4796183868169992, "grad_norm": 4.58290433883667, "learning_rate": 1.0172638909038155e-05, "loss": 0.0214, "step": 13648 }, { "epoch": 1.479726799653079, "grad_norm": 0.34691786766052246, "learning_rate": 1.0171913535470767e-05, "loss": 0.0019, "step": 13649 }, { "epoch": 1.4798352124891587, "grad_norm": 1.3700364828109741, "learning_rate": 1.017118816190338e-05, "loss": 0.0319, "step": 13650 }, { "epoch": 1.4799436253252385, "grad_norm": 2.8018321990966797, "learning_rate": 1.0170462788335993e-05, "loss": 0.0077, "step": 13651 }, { "epoch": 1.4800520381613183, "grad_norm": 3.9696664810180664, "learning_rate": 1.0169737414768605e-05, "loss": 0.0544, "step": 13652 }, { "epoch": 1.480160450997398, "grad_norm": 5.725498676300049, "learning_rate": 1.016901204120122e-05, "loss": 0.0908, "step": 13653 }, { "epoch": 1.4802688638334778, "grad_norm": 4.664687633514404, "learning_rate": 1.0168286667633833e-05, "loss": 0.0332, "step": 13654 }, { "epoch": 1.4803772766695578, "grad_norm": 3.331533670425415, "learning_rate": 1.0167561294066447e-05, "loss": 0.0084, "step": 13655 }, { "epoch": 1.4804856895056375, "grad_norm": 3.317589044570923, "learning_rate": 1.0166835920499059e-05, "loss": 0.064, "step": 13656 }, { "epoch": 1.4805941023417173, "grad_norm": 4.266674041748047, "learning_rate": 1.0166110546931671e-05, "loss": 0.0221, "step": 13657 }, { "epoch": 1.480702515177797, "grad_norm": 0.33110374212265015, "learning_rate": 1.0165385173364283e-05, "loss": 0.0195, "step": 13658 }, { "epoch": 1.4808109280138768, "grad_norm": 4.365935802459717, "learning_rate": 1.0164659799796897e-05, "loss": 0.0211, "step": 13659 }, { "epoch": 1.4809193408499566, "grad_norm": 4.558294296264648, "learning_rate": 1.0163934426229509e-05, "loss": 0.0334, "step": 13660 }, { "epoch": 1.4810277536860363, "grad_norm": 5.9684648513793945, "learning_rate": 1.0163209052662121e-05, "loss": 0.0258, "step": 13661 }, { "epoch": 1.4811361665221163, "grad_norm": 1.7510194778442383, "learning_rate": 1.0162483679094735e-05, "loss": 0.0304, "step": 13662 }, { "epoch": 1.481244579358196, "grad_norm": 0.8152337670326233, "learning_rate": 1.0161758305527347e-05, "loss": 0.0175, "step": 13663 }, { "epoch": 1.4813529921942759, "grad_norm": 2.6054975986480713, "learning_rate": 1.016103293195996e-05, "loss": 0.0277, "step": 13664 }, { "epoch": 1.4814614050303556, "grad_norm": 2.314627170562744, "learning_rate": 1.0160307558392573e-05, "loss": 0.0258, "step": 13665 }, { "epoch": 1.4815698178664354, "grad_norm": 1.779099702835083, "learning_rate": 1.0159582184825185e-05, "loss": 0.0064, "step": 13666 }, { "epoch": 1.4816782307025151, "grad_norm": 2.0666189193725586, "learning_rate": 1.0158856811257797e-05, "loss": 0.0158, "step": 13667 }, { "epoch": 1.481786643538595, "grad_norm": 0.4085031747817993, "learning_rate": 1.0158131437690411e-05, "loss": 0.0027, "step": 13668 }, { "epoch": 1.481895056374675, "grad_norm": 3.5692710876464844, "learning_rate": 1.0157406064123023e-05, "loss": 0.0424, "step": 13669 }, { "epoch": 1.4820034692107544, "grad_norm": 2.0755088329315186, "learning_rate": 1.0156680690555636e-05, "loss": 0.0098, "step": 13670 }, { "epoch": 1.4821118820468344, "grad_norm": 0.2538906931877136, "learning_rate": 1.0155955316988251e-05, "loss": 0.0024, "step": 13671 }, { "epoch": 1.4822202948829142, "grad_norm": 3.4516825675964355, "learning_rate": 1.0155229943420863e-05, "loss": 0.0208, "step": 13672 }, { "epoch": 1.482328707718994, "grad_norm": 2.197465658187866, "learning_rate": 1.0154504569853475e-05, "loss": 0.0075, "step": 13673 }, { "epoch": 1.4824371205550737, "grad_norm": 0.08554312586784363, "learning_rate": 1.015377919628609e-05, "loss": 0.0011, "step": 13674 }, { "epoch": 1.4825455333911535, "grad_norm": 1.5331066846847534, "learning_rate": 1.0153053822718701e-05, "loss": 0.0084, "step": 13675 }, { "epoch": 1.4826539462272332, "grad_norm": 0.3560749292373657, "learning_rate": 1.0152328449151313e-05, "loss": 0.0085, "step": 13676 }, { "epoch": 1.482762359063313, "grad_norm": 2.796405792236328, "learning_rate": 1.0151603075583927e-05, "loss": 0.0299, "step": 13677 }, { "epoch": 1.482870771899393, "grad_norm": 0.3857100307941437, "learning_rate": 1.015087770201654e-05, "loss": 0.0016, "step": 13678 }, { "epoch": 1.4829791847354727, "grad_norm": 0.3137039542198181, "learning_rate": 1.0150152328449152e-05, "loss": 0.0054, "step": 13679 }, { "epoch": 1.4830875975715525, "grad_norm": 0.3558512330055237, "learning_rate": 1.0149426954881765e-05, "loss": 0.017, "step": 13680 }, { "epoch": 1.4831960104076323, "grad_norm": 1.3776317834854126, "learning_rate": 1.0148701581314378e-05, "loss": 0.0048, "step": 13681 }, { "epoch": 1.483304423243712, "grad_norm": 2.5051872730255127, "learning_rate": 1.014797620774699e-05, "loss": 0.0423, "step": 13682 }, { "epoch": 1.4834128360797918, "grad_norm": 6.81102991104126, "learning_rate": 1.0147250834179604e-05, "loss": 0.0216, "step": 13683 }, { "epoch": 1.4835212489158716, "grad_norm": 3.6102025508880615, "learning_rate": 1.0146525460612216e-05, "loss": 0.0212, "step": 13684 }, { "epoch": 1.4836296617519515, "grad_norm": 0.6232573986053467, "learning_rate": 1.0145800087044828e-05, "loss": 0.0111, "step": 13685 }, { "epoch": 1.483738074588031, "grad_norm": 1.1593399047851562, "learning_rate": 1.014507471347744e-05, "loss": 0.0345, "step": 13686 }, { "epoch": 1.483846487424111, "grad_norm": 0.38407444953918457, "learning_rate": 1.0144349339910054e-05, "loss": 0.0082, "step": 13687 }, { "epoch": 1.4839549002601908, "grad_norm": 2.8701913356781006, "learning_rate": 1.0143623966342668e-05, "loss": 0.0394, "step": 13688 }, { "epoch": 1.4840633130962706, "grad_norm": 0.5116901397705078, "learning_rate": 1.0142898592775282e-05, "loss": 0.0222, "step": 13689 }, { "epoch": 1.4841717259323504, "grad_norm": 1.506543517112732, "learning_rate": 1.0142173219207894e-05, "loss": 0.0117, "step": 13690 }, { "epoch": 1.4842801387684301, "grad_norm": 0.9128370881080627, "learning_rate": 1.0141447845640506e-05, "loss": 0.0288, "step": 13691 }, { "epoch": 1.48438855160451, "grad_norm": 0.3663076162338257, "learning_rate": 1.014072247207312e-05, "loss": 0.0285, "step": 13692 }, { "epoch": 1.4844969644405897, "grad_norm": 4.665919303894043, "learning_rate": 1.0139997098505732e-05, "loss": 0.021, "step": 13693 }, { "epoch": 1.4846053772766696, "grad_norm": 0.4476304352283478, "learning_rate": 1.0139271724938344e-05, "loss": 0.0029, "step": 13694 }, { "epoch": 1.4847137901127494, "grad_norm": 3.302410840988159, "learning_rate": 1.0138546351370958e-05, "loss": 0.0241, "step": 13695 }, { "epoch": 1.4848222029488292, "grad_norm": 0.2523941397666931, "learning_rate": 1.013782097780357e-05, "loss": 0.0007, "step": 13696 }, { "epoch": 1.484930615784909, "grad_norm": 0.6845107078552246, "learning_rate": 1.0137095604236182e-05, "loss": 0.01, "step": 13697 }, { "epoch": 1.4850390286209887, "grad_norm": 4.396280765533447, "learning_rate": 1.0136370230668794e-05, "loss": 0.0504, "step": 13698 }, { "epoch": 1.4851474414570685, "grad_norm": 1.943698763847351, "learning_rate": 1.0135644857101408e-05, "loss": 0.0228, "step": 13699 }, { "epoch": 1.4852558542931482, "grad_norm": 4.06968355178833, "learning_rate": 1.013491948353402e-05, "loss": 0.0204, "step": 13700 }, { "epoch": 1.4853642671292282, "grad_norm": 0.3755668103694916, "learning_rate": 1.0134194109966632e-05, "loss": 0.0102, "step": 13701 }, { "epoch": 1.485472679965308, "grad_norm": 0.10331109166145325, "learning_rate": 1.0133468736399246e-05, "loss": 0.001, "step": 13702 }, { "epoch": 1.4855810928013877, "grad_norm": 2.773892641067505, "learning_rate": 1.0132743362831858e-05, "loss": 0.0258, "step": 13703 }, { "epoch": 1.4856895056374675, "grad_norm": 1.6637996435165405, "learning_rate": 1.013201798926447e-05, "loss": 0.0538, "step": 13704 }, { "epoch": 1.4857979184735473, "grad_norm": 0.5027179718017578, "learning_rate": 1.0131292615697086e-05, "loss": 0.0072, "step": 13705 }, { "epoch": 1.485906331309627, "grad_norm": 2.6604700088500977, "learning_rate": 1.0130567242129698e-05, "loss": 0.0475, "step": 13706 }, { "epoch": 1.4860147441457068, "grad_norm": 1.3206946849822998, "learning_rate": 1.0129841868562312e-05, "loss": 0.033, "step": 13707 }, { "epoch": 1.4861231569817868, "grad_norm": 0.30248111486434937, "learning_rate": 1.0129116494994924e-05, "loss": 0.0028, "step": 13708 }, { "epoch": 1.4862315698178663, "grad_norm": 0.6688490509986877, "learning_rate": 1.0128391121427536e-05, "loss": 0.014, "step": 13709 }, { "epoch": 1.4863399826539463, "grad_norm": 5.133955955505371, "learning_rate": 1.0127665747860148e-05, "loss": 0.0929, "step": 13710 }, { "epoch": 1.486448395490026, "grad_norm": 2.265848398208618, "learning_rate": 1.0126940374292762e-05, "loss": 0.0426, "step": 13711 }, { "epoch": 1.4865568083261058, "grad_norm": 0.20051433145999908, "learning_rate": 1.0126215000725374e-05, "loss": 0.0015, "step": 13712 }, { "epoch": 1.4866652211621856, "grad_norm": 11.461959838867188, "learning_rate": 1.0125489627157987e-05, "loss": 0.0337, "step": 13713 }, { "epoch": 1.4867736339982653, "grad_norm": 3.597683906555176, "learning_rate": 1.01247642535906e-05, "loss": 0.0388, "step": 13714 }, { "epoch": 1.486882046834345, "grad_norm": 1.5373787879943848, "learning_rate": 1.0124038880023212e-05, "loss": 0.0088, "step": 13715 }, { "epoch": 1.4869904596704249, "grad_norm": 6.380814075469971, "learning_rate": 1.0123313506455825e-05, "loss": 0.0602, "step": 13716 }, { "epoch": 1.4870988725065049, "grad_norm": 3.469592332839966, "learning_rate": 1.0122588132888438e-05, "loss": 0.0903, "step": 13717 }, { "epoch": 1.4872072853425846, "grad_norm": 0.022548912093043327, "learning_rate": 1.012186275932105e-05, "loss": 0.0002, "step": 13718 }, { "epoch": 1.4873156981786644, "grad_norm": 0.05738642439246178, "learning_rate": 1.0121137385753663e-05, "loss": 0.0005, "step": 13719 }, { "epoch": 1.4874241110147441, "grad_norm": 3.2083823680877686, "learning_rate": 1.0120412012186277e-05, "loss": 0.012, "step": 13720 }, { "epoch": 1.487532523850824, "grad_norm": 0.3475770056247711, "learning_rate": 1.0119686638618889e-05, "loss": 0.0126, "step": 13721 }, { "epoch": 1.4876409366869037, "grad_norm": 0.09094204008579254, "learning_rate": 1.0118961265051504e-05, "loss": 0.0009, "step": 13722 }, { "epoch": 1.4877493495229834, "grad_norm": 0.3513452410697937, "learning_rate": 1.0118235891484116e-05, "loss": 0.0065, "step": 13723 }, { "epoch": 1.4878577623590634, "grad_norm": 0.739683985710144, "learning_rate": 1.0117510517916729e-05, "loss": 0.0145, "step": 13724 }, { "epoch": 1.4879661751951432, "grad_norm": 7.533625602722168, "learning_rate": 1.011678514434934e-05, "loss": 0.0649, "step": 13725 }, { "epoch": 1.488074588031223, "grad_norm": 0.5930241942405701, "learning_rate": 1.0116059770781955e-05, "loss": 0.0229, "step": 13726 }, { "epoch": 1.4881830008673027, "grad_norm": 0.34871938824653625, "learning_rate": 1.0115334397214567e-05, "loss": 0.0118, "step": 13727 }, { "epoch": 1.4882914137033825, "grad_norm": 2.6628260612487793, "learning_rate": 1.0114609023647179e-05, "loss": 0.064, "step": 13728 }, { "epoch": 1.4883998265394622, "grad_norm": 0.12465977668762207, "learning_rate": 1.0113883650079793e-05, "loss": 0.0014, "step": 13729 }, { "epoch": 1.488508239375542, "grad_norm": 3.4411263465881348, "learning_rate": 1.0113158276512405e-05, "loss": 0.0208, "step": 13730 }, { "epoch": 1.488616652211622, "grad_norm": 0.561435341835022, "learning_rate": 1.0112432902945017e-05, "loss": 0.0022, "step": 13731 }, { "epoch": 1.4887250650477015, "grad_norm": 4.60944128036499, "learning_rate": 1.011170752937763e-05, "loss": 0.0429, "step": 13732 }, { "epoch": 1.4888334778837815, "grad_norm": 2.1801199913024902, "learning_rate": 1.0110982155810243e-05, "loss": 0.0352, "step": 13733 }, { "epoch": 1.4889418907198613, "grad_norm": 4.557387828826904, "learning_rate": 1.0110256782242855e-05, "loss": 0.0548, "step": 13734 }, { "epoch": 1.489050303555941, "grad_norm": 1.1355677843093872, "learning_rate": 1.0109531408675469e-05, "loss": 0.004, "step": 13735 }, { "epoch": 1.4891587163920208, "grad_norm": 0.25335806608200073, "learning_rate": 1.0108806035108081e-05, "loss": 0.0019, "step": 13736 }, { "epoch": 1.4892671292281006, "grad_norm": 2.315394878387451, "learning_rate": 1.0108080661540693e-05, "loss": 0.0424, "step": 13737 }, { "epoch": 1.4893755420641803, "grad_norm": 0.4445308446884155, "learning_rate": 1.0107355287973305e-05, "loss": 0.0015, "step": 13738 }, { "epoch": 1.48948395490026, "grad_norm": 4.508067607879639, "learning_rate": 1.010662991440592e-05, "loss": 0.0599, "step": 13739 }, { "epoch": 1.48959236773634, "grad_norm": 3.652918577194214, "learning_rate": 1.0105904540838533e-05, "loss": 0.0322, "step": 13740 }, { "epoch": 1.4897007805724198, "grad_norm": 0.5760942697525024, "learning_rate": 1.0105179167271147e-05, "loss": 0.0047, "step": 13741 }, { "epoch": 1.4898091934084996, "grad_norm": 1.3905127048492432, "learning_rate": 1.0104453793703759e-05, "loss": 0.0116, "step": 13742 }, { "epoch": 1.4899176062445794, "grad_norm": 5.587434768676758, "learning_rate": 1.0103728420136371e-05, "loss": 0.044, "step": 13743 }, { "epoch": 1.4900260190806591, "grad_norm": 5.4740519523620605, "learning_rate": 1.0103003046568985e-05, "loss": 0.015, "step": 13744 }, { "epoch": 1.4901344319167389, "grad_norm": 11.69371509552002, "learning_rate": 1.0102277673001597e-05, "loss": 0.0413, "step": 13745 }, { "epoch": 1.4902428447528187, "grad_norm": 3.285024642944336, "learning_rate": 1.010155229943421e-05, "loss": 0.0302, "step": 13746 }, { "epoch": 1.4903512575888986, "grad_norm": 4.89230489730835, "learning_rate": 1.0100826925866823e-05, "loss": 0.0055, "step": 13747 }, { "epoch": 1.4904596704249784, "grad_norm": 4.5872111320495605, "learning_rate": 1.0100101552299435e-05, "loss": 0.0245, "step": 13748 }, { "epoch": 1.4905680832610582, "grad_norm": 3.296687602996826, "learning_rate": 1.0099376178732047e-05, "loss": 0.0325, "step": 13749 }, { "epoch": 1.490676496097138, "grad_norm": 0.12039536237716675, "learning_rate": 1.009865080516466e-05, "loss": 0.0017, "step": 13750 }, { "epoch": 1.4907849089332177, "grad_norm": 3.1914420127868652, "learning_rate": 1.0097925431597273e-05, "loss": 0.0479, "step": 13751 }, { "epoch": 1.4908933217692975, "grad_norm": 1.2439687252044678, "learning_rate": 1.0097200058029886e-05, "loss": 0.0209, "step": 13752 }, { "epoch": 1.4910017346053772, "grad_norm": 0.3730373680591583, "learning_rate": 1.0096474684462498e-05, "loss": 0.0131, "step": 13753 }, { "epoch": 1.4911101474414572, "grad_norm": 0.7060497999191284, "learning_rate": 1.0095749310895111e-05, "loss": 0.0104, "step": 13754 }, { "epoch": 1.4912185602775367, "grad_norm": 0.7244200110435486, "learning_rate": 1.0095023937327724e-05, "loss": 0.0215, "step": 13755 }, { "epoch": 1.4913269731136167, "grad_norm": 5.706031322479248, "learning_rate": 1.0094298563760336e-05, "loss": 0.0669, "step": 13756 }, { "epoch": 1.4914353859496965, "grad_norm": 3.8127281665802, "learning_rate": 1.0093573190192951e-05, "loss": 0.0166, "step": 13757 }, { "epoch": 1.4915437987857763, "grad_norm": 3.021254062652588, "learning_rate": 1.0092847816625563e-05, "loss": 0.0196, "step": 13758 }, { "epoch": 1.491652211621856, "grad_norm": 2.5833446979522705, "learning_rate": 1.0092122443058177e-05, "loss": 0.0551, "step": 13759 }, { "epoch": 1.4917606244579358, "grad_norm": 5.285458564758301, "learning_rate": 1.009139706949079e-05, "loss": 0.0771, "step": 13760 }, { "epoch": 1.4918690372940155, "grad_norm": 0.4167274534702301, "learning_rate": 1.0090671695923402e-05, "loss": 0.0115, "step": 13761 }, { "epoch": 1.4919774501300953, "grad_norm": 2.1074106693267822, "learning_rate": 1.0089946322356014e-05, "loss": 0.0556, "step": 13762 }, { "epoch": 1.4920858629661753, "grad_norm": 0.6360434889793396, "learning_rate": 1.0089220948788628e-05, "loss": 0.0042, "step": 13763 }, { "epoch": 1.492194275802255, "grad_norm": 2.656080484390259, "learning_rate": 1.008849557522124e-05, "loss": 0.0137, "step": 13764 }, { "epoch": 1.4923026886383348, "grad_norm": 0.2790141999721527, "learning_rate": 1.0087770201653852e-05, "loss": 0.0077, "step": 13765 }, { "epoch": 1.4924111014744146, "grad_norm": 7.5367112159729, "learning_rate": 1.0087044828086466e-05, "loss": 0.0688, "step": 13766 }, { "epoch": 1.4925195143104943, "grad_norm": 2.5828819274902344, "learning_rate": 1.0086319454519078e-05, "loss": 0.0399, "step": 13767 }, { "epoch": 1.492627927146574, "grad_norm": 4.366225242614746, "learning_rate": 1.008559408095169e-05, "loss": 0.0314, "step": 13768 }, { "epoch": 1.4927363399826539, "grad_norm": 0.28570762276649475, "learning_rate": 1.0084868707384304e-05, "loss": 0.002, "step": 13769 }, { "epoch": 1.4928447528187339, "grad_norm": 2.0764665603637695, "learning_rate": 1.0084143333816916e-05, "loss": 0.0118, "step": 13770 }, { "epoch": 1.4929531656548134, "grad_norm": 3.0420870780944824, "learning_rate": 1.0083417960249528e-05, "loss": 0.02, "step": 13771 }, { "epoch": 1.4930615784908934, "grad_norm": 0.33681559562683105, "learning_rate": 1.0082692586682142e-05, "loss": 0.0125, "step": 13772 }, { "epoch": 1.4931699913269731, "grad_norm": 2.231656312942505, "learning_rate": 1.0081967213114754e-05, "loss": 0.0251, "step": 13773 }, { "epoch": 1.493278404163053, "grad_norm": 8.084162712097168, "learning_rate": 1.008124183954737e-05, "loss": 0.0153, "step": 13774 }, { "epoch": 1.4933868169991327, "grad_norm": 2.684765338897705, "learning_rate": 1.0080516465979982e-05, "loss": 0.0164, "step": 13775 }, { "epoch": 1.4934952298352124, "grad_norm": 0.9551674723625183, "learning_rate": 1.0079791092412594e-05, "loss": 0.0098, "step": 13776 }, { "epoch": 1.4936036426712924, "grad_norm": 0.46389690041542053, "learning_rate": 1.0079065718845206e-05, "loss": 0.0128, "step": 13777 }, { "epoch": 1.493712055507372, "grad_norm": 0.513100266456604, "learning_rate": 1.007834034527782e-05, "loss": 0.0139, "step": 13778 }, { "epoch": 1.493820468343452, "grad_norm": 2.0975656509399414, "learning_rate": 1.0077614971710432e-05, "loss": 0.0383, "step": 13779 }, { "epoch": 1.4939288811795317, "grad_norm": 2.4184250831604004, "learning_rate": 1.0076889598143044e-05, "loss": 0.0168, "step": 13780 }, { "epoch": 1.4940372940156115, "grad_norm": 5.980062961578369, "learning_rate": 1.0076164224575658e-05, "loss": 0.0596, "step": 13781 }, { "epoch": 1.4941457068516912, "grad_norm": 0.9933544993400574, "learning_rate": 1.007543885100827e-05, "loss": 0.0144, "step": 13782 }, { "epoch": 1.494254119687771, "grad_norm": 1.320034146308899, "learning_rate": 1.0074713477440882e-05, "loss": 0.0164, "step": 13783 }, { "epoch": 1.4943625325238508, "grad_norm": 0.596872091293335, "learning_rate": 1.0073988103873496e-05, "loss": 0.0088, "step": 13784 }, { "epoch": 1.4944709453599305, "grad_norm": 3.560882568359375, "learning_rate": 1.0073262730306108e-05, "loss": 0.0409, "step": 13785 }, { "epoch": 1.4945793581960105, "grad_norm": 0.15246039628982544, "learning_rate": 1.007253735673872e-05, "loss": 0.0005, "step": 13786 }, { "epoch": 1.4946877710320903, "grad_norm": 1.1221696138381958, "learning_rate": 1.0071811983171334e-05, "loss": 0.0349, "step": 13787 }, { "epoch": 1.49479618386817, "grad_norm": 7.4962849617004395, "learning_rate": 1.0071086609603946e-05, "loss": 0.0156, "step": 13788 }, { "epoch": 1.4949045967042498, "grad_norm": 1.4570245742797852, "learning_rate": 1.0070361236036559e-05, "loss": 0.0082, "step": 13789 }, { "epoch": 1.4950130095403296, "grad_norm": 2.2640316486358643, "learning_rate": 1.006963586246917e-05, "loss": 0.0206, "step": 13790 }, { "epoch": 1.4951214223764093, "grad_norm": 1.617462396621704, "learning_rate": 1.0068910488901786e-05, "loss": 0.0199, "step": 13791 }, { "epoch": 1.495229835212489, "grad_norm": 3.985607385635376, "learning_rate": 1.0068185115334398e-05, "loss": 0.0168, "step": 13792 }, { "epoch": 1.495338248048569, "grad_norm": 0.4499663710594177, "learning_rate": 1.0067459741767012e-05, "loss": 0.0196, "step": 13793 }, { "epoch": 1.4954466608846486, "grad_norm": 2.3834633827209473, "learning_rate": 1.0066734368199624e-05, "loss": 0.0202, "step": 13794 }, { "epoch": 1.4955550737207286, "grad_norm": 0.481518417596817, "learning_rate": 1.0066008994632236e-05, "loss": 0.0031, "step": 13795 }, { "epoch": 1.4956634865568084, "grad_norm": 1.5376328229904175, "learning_rate": 1.006528362106485e-05, "loss": 0.0118, "step": 13796 }, { "epoch": 1.4957718993928881, "grad_norm": 0.20638127624988556, "learning_rate": 1.0064558247497462e-05, "loss": 0.0018, "step": 13797 }, { "epoch": 1.495880312228968, "grad_norm": 0.6546648740768433, "learning_rate": 1.0063832873930075e-05, "loss": 0.0142, "step": 13798 }, { "epoch": 1.4959887250650477, "grad_norm": 3.90934157371521, "learning_rate": 1.0063107500362688e-05, "loss": 0.0588, "step": 13799 }, { "epoch": 1.4960971379011274, "grad_norm": 0.9032189846038818, "learning_rate": 1.00623821267953e-05, "loss": 0.0054, "step": 13800 }, { "epoch": 1.4962055507372072, "grad_norm": 2.1658172607421875, "learning_rate": 1.0061656753227913e-05, "loss": 0.0172, "step": 13801 }, { "epoch": 1.4963139635732872, "grad_norm": 0.9458690881729126, "learning_rate": 1.0060931379660525e-05, "loss": 0.0078, "step": 13802 }, { "epoch": 1.496422376409367, "grad_norm": 1.8556146621704102, "learning_rate": 1.0060206006093139e-05, "loss": 0.0564, "step": 13803 }, { "epoch": 1.4965307892454467, "grad_norm": 4.524037837982178, "learning_rate": 1.0059480632525751e-05, "loss": 0.0323, "step": 13804 }, { "epoch": 1.4966392020815265, "grad_norm": 2.129345178604126, "learning_rate": 1.0058755258958363e-05, "loss": 0.0318, "step": 13805 }, { "epoch": 1.4967476149176062, "grad_norm": 2.9694764614105225, "learning_rate": 1.0058029885390977e-05, "loss": 0.0324, "step": 13806 }, { "epoch": 1.496856027753686, "grad_norm": 5.1907639503479, "learning_rate": 1.0057304511823589e-05, "loss": 0.022, "step": 13807 }, { "epoch": 1.4969644405897657, "grad_norm": 3.8247666358947754, "learning_rate": 1.0056579138256201e-05, "loss": 0.0474, "step": 13808 }, { "epoch": 1.4970728534258457, "grad_norm": 3.296967029571533, "learning_rate": 1.0055853764688817e-05, "loss": 0.0289, "step": 13809 }, { "epoch": 1.4971812662619255, "grad_norm": 3.9531729221343994, "learning_rate": 1.0055128391121429e-05, "loss": 0.007, "step": 13810 }, { "epoch": 1.4972896790980053, "grad_norm": 0.3323369324207306, "learning_rate": 1.0054403017554043e-05, "loss": 0.0104, "step": 13811 }, { "epoch": 1.497398091934085, "grad_norm": 0.4989723265171051, "learning_rate": 1.0053677643986655e-05, "loss": 0.0237, "step": 13812 }, { "epoch": 1.4975065047701648, "grad_norm": 0.45681893825531006, "learning_rate": 1.0052952270419267e-05, "loss": 0.0079, "step": 13813 }, { "epoch": 1.4976149176062445, "grad_norm": 1.7704211473464966, "learning_rate": 1.005222689685188e-05, "loss": 0.0268, "step": 13814 }, { "epoch": 1.4977233304423243, "grad_norm": 0.46965402364730835, "learning_rate": 1.0051501523284493e-05, "loss": 0.0217, "step": 13815 }, { "epoch": 1.4978317432784043, "grad_norm": 0.8625825047492981, "learning_rate": 1.0050776149717105e-05, "loss": 0.0243, "step": 13816 }, { "epoch": 1.4979401561144838, "grad_norm": 0.11100202798843384, "learning_rate": 1.0050050776149717e-05, "loss": 0.0009, "step": 13817 }, { "epoch": 1.4980485689505638, "grad_norm": 2.1705713272094727, "learning_rate": 1.0049325402582331e-05, "loss": 0.016, "step": 13818 }, { "epoch": 1.4981569817866436, "grad_norm": 0.35413238406181335, "learning_rate": 1.0048600029014943e-05, "loss": 0.0076, "step": 13819 }, { "epoch": 1.4982653946227233, "grad_norm": 2.015434980392456, "learning_rate": 1.0047874655447555e-05, "loss": 0.0195, "step": 13820 }, { "epoch": 1.498373807458803, "grad_norm": 0.33766117691993713, "learning_rate": 1.0047149281880169e-05, "loss": 0.003, "step": 13821 }, { "epoch": 1.4984822202948829, "grad_norm": 6.938407897949219, "learning_rate": 1.0046423908312781e-05, "loss": 0.0296, "step": 13822 }, { "epoch": 1.4985906331309626, "grad_norm": 1.248772144317627, "learning_rate": 1.0045698534745393e-05, "loss": 0.0123, "step": 13823 }, { "epoch": 1.4986990459670424, "grad_norm": 0.6474801898002625, "learning_rate": 1.0044973161178007e-05, "loss": 0.0024, "step": 13824 }, { "epoch": 1.4988074588031224, "grad_norm": 1.651252269744873, "learning_rate": 1.004424778761062e-05, "loss": 0.0405, "step": 13825 }, { "epoch": 1.4989158716392021, "grad_norm": 3.7906436920166016, "learning_rate": 1.0043522414043235e-05, "loss": 0.0407, "step": 13826 }, { "epoch": 1.499024284475282, "grad_norm": 1.8675390481948853, "learning_rate": 1.0042797040475847e-05, "loss": 0.0044, "step": 13827 }, { "epoch": 1.4991326973113617, "grad_norm": 1.7233072519302368, "learning_rate": 1.004207166690846e-05, "loss": 0.0217, "step": 13828 }, { "epoch": 1.4992411101474414, "grad_norm": 1.9276301860809326, "learning_rate": 1.0041346293341071e-05, "loss": 0.0391, "step": 13829 }, { "epoch": 1.4993495229835212, "grad_norm": 2.037313461303711, "learning_rate": 1.0040620919773685e-05, "loss": 0.0308, "step": 13830 }, { "epoch": 1.499457935819601, "grad_norm": 3.6397974491119385, "learning_rate": 1.0039895546206297e-05, "loss": 0.021, "step": 13831 }, { "epoch": 1.499566348655681, "grad_norm": 1.6868091821670532, "learning_rate": 1.003917017263891e-05, "loss": 0.0264, "step": 13832 }, { "epoch": 1.4996747614917605, "grad_norm": 0.3223305940628052, "learning_rate": 1.0038444799071523e-05, "loss": 0.0067, "step": 13833 }, { "epoch": 1.4997831743278405, "grad_norm": 0.1517861783504486, "learning_rate": 1.0037719425504135e-05, "loss": 0.0018, "step": 13834 }, { "epoch": 1.4998915871639202, "grad_norm": 1.0676000118255615, "learning_rate": 1.0036994051936748e-05, "loss": 0.0393, "step": 13835 }, { "epoch": 1.5, "grad_norm": 0.7051855325698853, "learning_rate": 1.0036268678369361e-05, "loss": 0.0155, "step": 13836 }, { "epoch": 1.5001084128360798, "grad_norm": 3.436760663986206, "learning_rate": 1.0035543304801974e-05, "loss": 0.0411, "step": 13837 }, { "epoch": 1.5002168256721595, "grad_norm": 7.41394567489624, "learning_rate": 1.0034817931234586e-05, "loss": 0.0159, "step": 13838 }, { "epoch": 1.5003252385082395, "grad_norm": 0.28366127610206604, "learning_rate": 1.00340925576672e-05, "loss": 0.0016, "step": 13839 }, { "epoch": 1.500433651344319, "grad_norm": 1.1678847074508667, "learning_rate": 1.0033367184099812e-05, "loss": 0.0326, "step": 13840 }, { "epoch": 1.500542064180399, "grad_norm": 1.8186894655227661, "learning_rate": 1.0032641810532424e-05, "loss": 0.0388, "step": 13841 }, { "epoch": 1.5006504770164788, "grad_norm": 10.312520980834961, "learning_rate": 1.0031916436965036e-05, "loss": 0.0256, "step": 13842 }, { "epoch": 1.5007588898525586, "grad_norm": 0.7432416677474976, "learning_rate": 1.0031191063397652e-05, "loss": 0.0092, "step": 13843 }, { "epoch": 1.5008673026886383, "grad_norm": 0.49750804901123047, "learning_rate": 1.0030465689830264e-05, "loss": 0.0075, "step": 13844 }, { "epoch": 1.500975715524718, "grad_norm": 1.4281541109085083, "learning_rate": 1.0029740316262878e-05, "loss": 0.0188, "step": 13845 }, { "epoch": 1.501084128360798, "grad_norm": 8.503618240356445, "learning_rate": 1.002901494269549e-05, "loss": 0.0524, "step": 13846 }, { "epoch": 1.5011925411968776, "grad_norm": 0.7712326049804688, "learning_rate": 1.0028289569128102e-05, "loss": 0.0127, "step": 13847 }, { "epoch": 1.5013009540329576, "grad_norm": 3.0937154293060303, "learning_rate": 1.0027564195560716e-05, "loss": 0.0177, "step": 13848 }, { "epoch": 1.5014093668690371, "grad_norm": 0.5750963687896729, "learning_rate": 1.0026838821993328e-05, "loss": 0.0132, "step": 13849 }, { "epoch": 1.5015177797051171, "grad_norm": 2.5984113216400146, "learning_rate": 1.002611344842594e-05, "loss": 0.0166, "step": 13850 }, { "epoch": 1.501626192541197, "grad_norm": 0.5946915745735168, "learning_rate": 1.0025388074858554e-05, "loss": 0.009, "step": 13851 }, { "epoch": 1.5017346053772767, "grad_norm": 0.5171538591384888, "learning_rate": 1.0024662701291166e-05, "loss": 0.0238, "step": 13852 }, { "epoch": 1.5018430182133564, "grad_norm": 4.097198486328125, "learning_rate": 1.0023937327723778e-05, "loss": 0.0383, "step": 13853 }, { "epoch": 1.5019514310494362, "grad_norm": 5.426459312438965, "learning_rate": 1.002321195415639e-05, "loss": 0.0072, "step": 13854 }, { "epoch": 1.5020598438855162, "grad_norm": 2.726030111312866, "learning_rate": 1.0022486580589004e-05, "loss": 0.0179, "step": 13855 }, { "epoch": 1.5021682567215957, "grad_norm": 1.5224454402923584, "learning_rate": 1.0021761207021616e-05, "loss": 0.0164, "step": 13856 }, { "epoch": 1.5022766695576757, "grad_norm": 2.1432647705078125, "learning_rate": 1.0021035833454228e-05, "loss": 0.0123, "step": 13857 }, { "epoch": 1.5023850823937555, "grad_norm": 3.197638511657715, "learning_rate": 1.0020310459886842e-05, "loss": 0.0064, "step": 13858 }, { "epoch": 1.5024934952298352, "grad_norm": 4.054254055023193, "learning_rate": 1.0019585086319454e-05, "loss": 0.0516, "step": 13859 }, { "epoch": 1.502601908065915, "grad_norm": 5.036928176879883, "learning_rate": 1.0018859712752066e-05, "loss": 0.0512, "step": 13860 }, { "epoch": 1.5027103209019947, "grad_norm": 0.27880996465682983, "learning_rate": 1.0018134339184682e-05, "loss": 0.0016, "step": 13861 }, { "epoch": 1.5028187337380747, "grad_norm": 2.5431106090545654, "learning_rate": 1.0017408965617294e-05, "loss": 0.013, "step": 13862 }, { "epoch": 1.5029271465741543, "grad_norm": 12.323360443115234, "learning_rate": 1.0016683592049908e-05, "loss": 0.0551, "step": 13863 }, { "epoch": 1.5030355594102343, "grad_norm": 0.5260310173034668, "learning_rate": 1.001595821848252e-05, "loss": 0.0086, "step": 13864 }, { "epoch": 1.5031439722463138, "grad_norm": 2.221123456954956, "learning_rate": 1.0015232844915132e-05, "loss": 0.011, "step": 13865 }, { "epoch": 1.5032523850823938, "grad_norm": 3.8488733768463135, "learning_rate": 1.0014507471347746e-05, "loss": 0.0148, "step": 13866 }, { "epoch": 1.5033607979184735, "grad_norm": 2.780586004257202, "learning_rate": 1.0013782097780358e-05, "loss": 0.0088, "step": 13867 }, { "epoch": 1.5034692107545533, "grad_norm": 10.296000480651855, "learning_rate": 1.001305672421297e-05, "loss": 0.0155, "step": 13868 }, { "epoch": 1.5035776235906333, "grad_norm": 1.79684579372406, "learning_rate": 1.0012331350645583e-05, "loss": 0.0324, "step": 13869 }, { "epoch": 1.5036860364267128, "grad_norm": 4.052270889282227, "learning_rate": 1.0011605977078196e-05, "loss": 0.0284, "step": 13870 }, { "epoch": 1.5037944492627928, "grad_norm": 2.9536170959472656, "learning_rate": 1.0010880603510809e-05, "loss": 0.0137, "step": 13871 }, { "epoch": 1.5039028620988724, "grad_norm": 0.7538491487503052, "learning_rate": 1.001015522994342e-05, "loss": 0.019, "step": 13872 }, { "epoch": 1.5040112749349523, "grad_norm": 1.7023768424987793, "learning_rate": 1.0009429856376034e-05, "loss": 0.0054, "step": 13873 }, { "epoch": 1.504119687771032, "grad_norm": 3.8175320625305176, "learning_rate": 1.0008704482808647e-05, "loss": 0.0209, "step": 13874 }, { "epoch": 1.5042281006071119, "grad_norm": 15.786690711975098, "learning_rate": 1.0007979109241259e-05, "loss": 0.1469, "step": 13875 }, { "epoch": 1.5043365134431916, "grad_norm": 5.4783034324646, "learning_rate": 1.0007253735673873e-05, "loss": 0.0523, "step": 13876 }, { "epoch": 1.5044449262792714, "grad_norm": 1.4969773292541504, "learning_rate": 1.0006528362106485e-05, "loss": 0.0751, "step": 13877 }, { "epoch": 1.5045533391153514, "grad_norm": 0.04127814993262291, "learning_rate": 1.00058029885391e-05, "loss": 0.0003, "step": 13878 }, { "epoch": 1.504661751951431, "grad_norm": 2.7052884101867676, "learning_rate": 1.0005077614971712e-05, "loss": 0.0374, "step": 13879 }, { "epoch": 1.504770164787511, "grad_norm": 3.9304587841033936, "learning_rate": 1.0004352241404325e-05, "loss": 0.0847, "step": 13880 }, { "epoch": 1.5048785776235907, "grad_norm": 1.5521050691604614, "learning_rate": 1.0003626867836937e-05, "loss": 0.016, "step": 13881 }, { "epoch": 1.5049869904596704, "grad_norm": 0.1452474594116211, "learning_rate": 1.000290149426955e-05, "loss": 0.0011, "step": 13882 }, { "epoch": 1.5050954032957502, "grad_norm": 1.045074224472046, "learning_rate": 1.0002176120702163e-05, "loss": 0.0083, "step": 13883 }, { "epoch": 1.50520381613183, "grad_norm": 1.1200170516967773, "learning_rate": 1.0001450747134775e-05, "loss": 0.006, "step": 13884 }, { "epoch": 1.50531222896791, "grad_norm": 0.26765939593315125, "learning_rate": 1.0000725373567389e-05, "loss": 0.006, "step": 13885 }, { "epoch": 1.5054206418039895, "grad_norm": 0.3657315671443939, "learning_rate": 1e-05, "loss": 0.0025, "step": 13886 }, { "epoch": 1.5055290546400695, "grad_norm": 3.7674288749694824, "learning_rate": 9.999274626432613e-06, "loss": 0.1279, "step": 13887 }, { "epoch": 1.505637467476149, "grad_norm": 1.7140631675720215, "learning_rate": 9.998549252865227e-06, "loss": 0.011, "step": 13888 }, { "epoch": 1.505745880312229, "grad_norm": 4.54549503326416, "learning_rate": 9.997823879297839e-06, "loss": 0.075, "step": 13889 }, { "epoch": 1.5058542931483088, "grad_norm": 8.35997200012207, "learning_rate": 9.997098505730451e-06, "loss": 0.0284, "step": 13890 }, { "epoch": 1.5059627059843885, "grad_norm": 0.9899147748947144, "learning_rate": 9.996373132163065e-06, "loss": 0.0355, "step": 13891 }, { "epoch": 1.5060711188204683, "grad_norm": 4.953623294830322, "learning_rate": 9.995647758595679e-06, "loss": 0.0371, "step": 13892 }, { "epoch": 1.506179531656548, "grad_norm": 0.29348450899124146, "learning_rate": 9.994922385028291e-06, "loss": 0.0061, "step": 13893 }, { "epoch": 1.506287944492628, "grad_norm": 0.8729523420333862, "learning_rate": 9.994197011460903e-06, "loss": 0.0134, "step": 13894 }, { "epoch": 1.5063963573287076, "grad_norm": 1.077600121498108, "learning_rate": 9.993471637893515e-06, "loss": 0.0086, "step": 13895 }, { "epoch": 1.5065047701647876, "grad_norm": 3.4711337089538574, "learning_rate": 9.992746264326129e-06, "loss": 0.0052, "step": 13896 }, { "epoch": 1.5066131830008673, "grad_norm": 2.014873743057251, "learning_rate": 9.992020890758741e-06, "loss": 0.0199, "step": 13897 }, { "epoch": 1.506721595836947, "grad_norm": 7.096180438995361, "learning_rate": 9.991295517191353e-06, "loss": 0.0208, "step": 13898 }, { "epoch": 1.5068300086730269, "grad_norm": 2.6225037574768066, "learning_rate": 9.990570143623967e-06, "loss": 0.0348, "step": 13899 }, { "epoch": 1.5069384215091066, "grad_norm": 3.5435023307800293, "learning_rate": 9.989844770056581e-06, "loss": 0.0377, "step": 13900 }, { "epoch": 1.5070468343451866, "grad_norm": 3.7984604835510254, "learning_rate": 9.989119396489193e-06, "loss": 0.0382, "step": 13901 }, { "epoch": 1.5071552471812661, "grad_norm": 0.24275729060173035, "learning_rate": 9.988394022921805e-06, "loss": 0.0051, "step": 13902 }, { "epoch": 1.5072636600173461, "grad_norm": 1.9495078325271606, "learning_rate": 9.987668649354419e-06, "loss": 0.0544, "step": 13903 }, { "epoch": 1.507372072853426, "grad_norm": 9.405433654785156, "learning_rate": 9.986943275787031e-06, "loss": 0.0383, "step": 13904 }, { "epoch": 1.5074804856895057, "grad_norm": 0.744618833065033, "learning_rate": 9.986217902219643e-06, "loss": 0.0176, "step": 13905 }, { "epoch": 1.5075888985255854, "grad_norm": 6.936422824859619, "learning_rate": 9.985492528652256e-06, "loss": 0.0405, "step": 13906 }, { "epoch": 1.5076973113616652, "grad_norm": 5.494248867034912, "learning_rate": 9.98476715508487e-06, "loss": 0.0614, "step": 13907 }, { "epoch": 1.5078057241977452, "grad_norm": 1.980735182762146, "learning_rate": 9.984041781517483e-06, "loss": 0.0125, "step": 13908 }, { "epoch": 1.5079141370338247, "grad_norm": 0.4292881488800049, "learning_rate": 9.983316407950095e-06, "loss": 0.0261, "step": 13909 }, { "epoch": 1.5080225498699047, "grad_norm": 0.4167192876338959, "learning_rate": 9.982591034382708e-06, "loss": 0.0087, "step": 13910 }, { "epoch": 1.5081309627059842, "grad_norm": 1.9355133771896362, "learning_rate": 9.981865660815321e-06, "loss": 0.038, "step": 13911 }, { "epoch": 1.5082393755420642, "grad_norm": 1.0410475730895996, "learning_rate": 9.981140287247933e-06, "loss": 0.0335, "step": 13912 }, { "epoch": 1.508347788378144, "grad_norm": 0.47653812170028687, "learning_rate": 9.980414913680546e-06, "loss": 0.0024, "step": 13913 }, { "epoch": 1.5084562012142237, "grad_norm": 0.41624146699905396, "learning_rate": 9.97968954011316e-06, "loss": 0.0194, "step": 13914 }, { "epoch": 1.5085646140503035, "grad_norm": 3.2473080158233643, "learning_rate": 9.978964166545772e-06, "loss": 0.0376, "step": 13915 }, { "epoch": 1.5086730268863833, "grad_norm": 0.5940911173820496, "learning_rate": 9.978238792978384e-06, "loss": 0.0155, "step": 13916 }, { "epoch": 1.5087814397224633, "grad_norm": 1.844508171081543, "learning_rate": 9.977513419410998e-06, "loss": 0.037, "step": 13917 }, { "epoch": 1.5088898525585428, "grad_norm": 0.8053896427154541, "learning_rate": 9.976788045843611e-06, "loss": 0.0203, "step": 13918 }, { "epoch": 1.5089982653946228, "grad_norm": 2.478363513946533, "learning_rate": 9.976062672276224e-06, "loss": 0.0181, "step": 13919 }, { "epoch": 1.5091066782307025, "grad_norm": 2.3573648929595947, "learning_rate": 9.975337298708836e-06, "loss": 0.0082, "step": 13920 }, { "epoch": 1.5092150910667823, "grad_norm": 4.474584102630615, "learning_rate": 9.974611925141448e-06, "loss": 0.0612, "step": 13921 }, { "epoch": 1.509323503902862, "grad_norm": 1.327294945716858, "learning_rate": 9.973886551574062e-06, "loss": 0.013, "step": 13922 }, { "epoch": 1.5094319167389418, "grad_norm": 1.9396398067474365, "learning_rate": 9.973161178006674e-06, "loss": 0.0636, "step": 13923 }, { "epoch": 1.5095403295750218, "grad_norm": 1.1545356512069702, "learning_rate": 9.972435804439286e-06, "loss": 0.0074, "step": 13924 }, { "epoch": 1.5096487424111014, "grad_norm": 1.1374541521072388, "learning_rate": 9.9717104308719e-06, "loss": 0.0328, "step": 13925 }, { "epoch": 1.5097571552471813, "grad_norm": 4.227858543395996, "learning_rate": 9.970985057304514e-06, "loss": 0.0453, "step": 13926 }, { "epoch": 1.5098655680832611, "grad_norm": 0.03343602642416954, "learning_rate": 9.970259683737126e-06, "loss": 0.0005, "step": 13927 }, { "epoch": 1.5099739809193409, "grad_norm": 5.6143717765808105, "learning_rate": 9.969534310169738e-06, "loss": 0.0526, "step": 13928 }, { "epoch": 1.5100823937554206, "grad_norm": 2.4222450256347656, "learning_rate": 9.968808936602352e-06, "loss": 0.0246, "step": 13929 }, { "epoch": 1.5101908065915004, "grad_norm": 1.6233917474746704, "learning_rate": 9.968083563034964e-06, "loss": 0.0247, "step": 13930 }, { "epoch": 1.5102992194275804, "grad_norm": 1.159044861793518, "learning_rate": 9.967358189467576e-06, "loss": 0.0227, "step": 13931 }, { "epoch": 1.51040763226366, "grad_norm": 7.646029949188232, "learning_rate": 9.966632815900188e-06, "loss": 0.0823, "step": 13932 }, { "epoch": 1.51051604509974, "grad_norm": 2.9438042640686035, "learning_rate": 9.965907442332802e-06, "loss": 0.0776, "step": 13933 }, { "epoch": 1.5106244579358195, "grad_norm": 0.3573271930217743, "learning_rate": 9.965182068765416e-06, "loss": 0.0127, "step": 13934 }, { "epoch": 1.5107328707718994, "grad_norm": 0.2870725691318512, "learning_rate": 9.964456695198028e-06, "loss": 0.0022, "step": 13935 }, { "epoch": 1.5108412836079792, "grad_norm": 1.7633572816848755, "learning_rate": 9.96373132163064e-06, "loss": 0.022, "step": 13936 }, { "epoch": 1.510949696444059, "grad_norm": 0.5842339992523193, "learning_rate": 9.963005948063254e-06, "loss": 0.0044, "step": 13937 }, { "epoch": 1.5110581092801387, "grad_norm": 1.531343698501587, "learning_rate": 9.962280574495866e-06, "loss": 0.0438, "step": 13938 }, { "epoch": 1.5111665221162185, "grad_norm": 5.038231372833252, "learning_rate": 9.961555200928478e-06, "loss": 0.0426, "step": 13939 }, { "epoch": 1.5112749349522985, "grad_norm": 2.510457992553711, "learning_rate": 9.960829827361092e-06, "loss": 0.0353, "step": 13940 }, { "epoch": 1.511383347788378, "grad_norm": 1.8005698919296265, "learning_rate": 9.960104453793704e-06, "loss": 0.0489, "step": 13941 }, { "epoch": 1.511491760624458, "grad_norm": 0.5783145427703857, "learning_rate": 9.959379080226316e-06, "loss": 0.0127, "step": 13942 }, { "epoch": 1.5116001734605378, "grad_norm": 2.2473244667053223, "learning_rate": 9.95865370665893e-06, "loss": 0.0136, "step": 13943 }, { "epoch": 1.5117085862966175, "grad_norm": 0.7504259943962097, "learning_rate": 9.957928333091544e-06, "loss": 0.0155, "step": 13944 }, { "epoch": 1.5118169991326973, "grad_norm": 13.636663436889648, "learning_rate": 9.957202959524156e-06, "loss": 0.014, "step": 13945 }, { "epoch": 1.511925411968777, "grad_norm": 1.6801081895828247, "learning_rate": 9.956477585956768e-06, "loss": 0.0285, "step": 13946 }, { "epoch": 1.512033824804857, "grad_norm": 0.642959475517273, "learning_rate": 9.95575221238938e-06, "loss": 0.0058, "step": 13947 }, { "epoch": 1.5121422376409366, "grad_norm": 0.7251641750335693, "learning_rate": 9.955026838821994e-06, "loss": 0.018, "step": 13948 }, { "epoch": 1.5122506504770166, "grad_norm": 0.9100024104118347, "learning_rate": 9.954301465254607e-06, "loss": 0.0184, "step": 13949 }, { "epoch": 1.512359063313096, "grad_norm": 1.0276719331741333, "learning_rate": 9.953576091687219e-06, "loss": 0.0143, "step": 13950 }, { "epoch": 1.512467476149176, "grad_norm": 0.18802979588508606, "learning_rate": 9.952850718119832e-06, "loss": 0.0021, "step": 13951 }, { "epoch": 1.5125758889852559, "grad_norm": 2.027817964553833, "learning_rate": 9.952125344552446e-06, "loss": 0.0166, "step": 13952 }, { "epoch": 1.5126843018213356, "grad_norm": 1.194875955581665, "learning_rate": 9.951399970985058e-06, "loss": 0.0079, "step": 13953 }, { "epoch": 1.5127927146574154, "grad_norm": 0.166417196393013, "learning_rate": 9.95067459741767e-06, "loss": 0.0013, "step": 13954 }, { "epoch": 1.5129011274934951, "grad_norm": 0.4546383023262024, "learning_rate": 9.949949223850284e-06, "loss": 0.0053, "step": 13955 }, { "epoch": 1.5130095403295751, "grad_norm": 0.2434258908033371, "learning_rate": 9.949223850282897e-06, "loss": 0.002, "step": 13956 }, { "epoch": 1.5131179531656547, "grad_norm": 2.3790531158447266, "learning_rate": 9.948498476715509e-06, "loss": 0.0379, "step": 13957 }, { "epoch": 1.5132263660017347, "grad_norm": 0.44283556938171387, "learning_rate": 9.947773103148121e-06, "loss": 0.0105, "step": 13958 }, { "epoch": 1.5133347788378144, "grad_norm": 2.7368037700653076, "learning_rate": 9.947047729580735e-06, "loss": 0.0366, "step": 13959 }, { "epoch": 1.5134431916738942, "grad_norm": 1.5016497373580933, "learning_rate": 9.946322356013349e-06, "loss": 0.0196, "step": 13960 }, { "epoch": 1.513551604509974, "grad_norm": 0.14673607051372528, "learning_rate": 9.94559698244596e-06, "loss": 0.0021, "step": 13961 }, { "epoch": 1.5136600173460537, "grad_norm": 0.37360963225364685, "learning_rate": 9.944871608878573e-06, "loss": 0.0062, "step": 13962 }, { "epoch": 1.5137684301821337, "grad_norm": 0.7379931211471558, "learning_rate": 9.944146235311187e-06, "loss": 0.006, "step": 13963 }, { "epoch": 1.5138768430182132, "grad_norm": 2.024502992630005, "learning_rate": 9.943420861743799e-06, "loss": 0.0611, "step": 13964 }, { "epoch": 1.5139852558542932, "grad_norm": 5.024960041046143, "learning_rate": 9.942695488176411e-06, "loss": 0.0402, "step": 13965 }, { "epoch": 1.514093668690373, "grad_norm": 1.7353565692901611, "learning_rate": 9.941970114609025e-06, "loss": 0.0379, "step": 13966 }, { "epoch": 1.5142020815264527, "grad_norm": 0.5356326699256897, "learning_rate": 9.941244741041637e-06, "loss": 0.0156, "step": 13967 }, { "epoch": 1.5143104943625325, "grad_norm": 0.04134971648454666, "learning_rate": 9.940519367474249e-06, "loss": 0.0006, "step": 13968 }, { "epoch": 1.5144189071986123, "grad_norm": 3.6958792209625244, "learning_rate": 9.939793993906863e-06, "loss": 0.0052, "step": 13969 }, { "epoch": 1.5145273200346923, "grad_norm": 0.3974068760871887, "learning_rate": 9.939068620339477e-06, "loss": 0.0086, "step": 13970 }, { "epoch": 1.5146357328707718, "grad_norm": 0.1053774431347847, "learning_rate": 9.938343246772089e-06, "loss": 0.0009, "step": 13971 }, { "epoch": 1.5147441457068518, "grad_norm": 2.29227352142334, "learning_rate": 9.937617873204701e-06, "loss": 0.0758, "step": 13972 }, { "epoch": 1.5148525585429313, "grad_norm": 0.8829266428947449, "learning_rate": 9.936892499637313e-06, "loss": 0.0018, "step": 13973 }, { "epoch": 1.5149609713790113, "grad_norm": 1.870162844657898, "learning_rate": 9.936167126069927e-06, "loss": 0.0247, "step": 13974 }, { "epoch": 1.515069384215091, "grad_norm": 2.2086970806121826, "learning_rate": 9.93544175250254e-06, "loss": 0.0586, "step": 13975 }, { "epoch": 1.5151777970511708, "grad_norm": 1.0562766790390015, "learning_rate": 9.934716378935151e-06, "loss": 0.0117, "step": 13976 }, { "epoch": 1.5152862098872506, "grad_norm": 1.042328119277954, "learning_rate": 9.933991005367765e-06, "loss": 0.012, "step": 13977 }, { "epoch": 1.5153946227233304, "grad_norm": 1.1213607788085938, "learning_rate": 9.933265631800379e-06, "loss": 0.0043, "step": 13978 }, { "epoch": 1.5155030355594103, "grad_norm": 0.9558759331703186, "learning_rate": 9.932540258232991e-06, "loss": 0.0025, "step": 13979 }, { "epoch": 1.51561144839549, "grad_norm": 0.9690811634063721, "learning_rate": 9.931814884665603e-06, "loss": 0.0127, "step": 13980 }, { "epoch": 1.5157198612315699, "grad_norm": 0.9250116944313049, "learning_rate": 9.931089511098217e-06, "loss": 0.0423, "step": 13981 }, { "epoch": 1.5158282740676496, "grad_norm": 0.33265620470046997, "learning_rate": 9.93036413753083e-06, "loss": 0.0048, "step": 13982 }, { "epoch": 1.5159366869037294, "grad_norm": 1.2632486820220947, "learning_rate": 9.929638763963441e-06, "loss": 0.0123, "step": 13983 }, { "epoch": 1.5160450997398092, "grad_norm": 0.9771695137023926, "learning_rate": 9.928913390396054e-06, "loss": 0.024, "step": 13984 }, { "epoch": 1.516153512575889, "grad_norm": 1.3054014444351196, "learning_rate": 9.928188016828667e-06, "loss": 0.0049, "step": 13985 }, { "epoch": 1.516261925411969, "grad_norm": 2.4902749061584473, "learning_rate": 9.927462643261281e-06, "loss": 0.0013, "step": 13986 }, { "epoch": 1.5163703382480485, "grad_norm": 5.590590000152588, "learning_rate": 9.926737269693893e-06, "loss": 0.0419, "step": 13987 }, { "epoch": 1.5164787510841284, "grad_norm": 0.9676462411880493, "learning_rate": 9.926011896126506e-06, "loss": 0.0366, "step": 13988 }, { "epoch": 1.5165871639202082, "grad_norm": 1.9799569845199585, "learning_rate": 9.92528652255912e-06, "loss": 0.0331, "step": 13989 }, { "epoch": 1.516695576756288, "grad_norm": 1.3724899291992188, "learning_rate": 9.924561148991731e-06, "loss": 0.0896, "step": 13990 }, { "epoch": 1.5168039895923677, "grad_norm": 0.5532281398773193, "learning_rate": 9.923835775424344e-06, "loss": 0.0044, "step": 13991 }, { "epoch": 1.5169124024284475, "grad_norm": 2.200374126434326, "learning_rate": 9.923110401856957e-06, "loss": 0.0216, "step": 13992 }, { "epoch": 1.5170208152645275, "grad_norm": 1.4373701810836792, "learning_rate": 9.92238502828957e-06, "loss": 0.0318, "step": 13993 }, { "epoch": 1.517129228100607, "grad_norm": 4.287402629852295, "learning_rate": 9.921659654722183e-06, "loss": 0.0651, "step": 13994 }, { "epoch": 1.517237640936687, "grad_norm": 0.3089795708656311, "learning_rate": 9.920934281154796e-06, "loss": 0.0062, "step": 13995 }, { "epoch": 1.5173460537727665, "grad_norm": 1.035556674003601, "learning_rate": 9.92020890758741e-06, "loss": 0.0198, "step": 13996 }, { "epoch": 1.5174544666088465, "grad_norm": 8.091975212097168, "learning_rate": 9.919483534020022e-06, "loss": 0.0211, "step": 13997 }, { "epoch": 1.5175628794449263, "grad_norm": 2.2994158267974854, "learning_rate": 9.918758160452634e-06, "loss": 0.0267, "step": 13998 }, { "epoch": 1.517671292281006, "grad_norm": 13.13562297821045, "learning_rate": 9.918032786885246e-06, "loss": 0.0298, "step": 13999 }, { "epoch": 1.5177797051170858, "grad_norm": 10.121044158935547, "learning_rate": 9.91730741331786e-06, "loss": 0.0871, "step": 14000 }, { "epoch": 1.5178881179531656, "grad_norm": 1.963808298110962, "learning_rate": 9.916582039750472e-06, "loss": 0.0371, "step": 14001 }, { "epoch": 1.5179965307892456, "grad_norm": 4.412830352783203, "learning_rate": 9.915856666183084e-06, "loss": 0.0244, "step": 14002 }, { "epoch": 1.518104943625325, "grad_norm": 1.3159362077713013, "learning_rate": 9.915131292615698e-06, "loss": 0.0041, "step": 14003 }, { "epoch": 1.518213356461405, "grad_norm": 2.341092586517334, "learning_rate": 9.914405919048312e-06, "loss": 0.0447, "step": 14004 }, { "epoch": 1.5183217692974849, "grad_norm": 0.8169490098953247, "learning_rate": 9.913680545480924e-06, "loss": 0.0046, "step": 14005 }, { "epoch": 1.5184301821335646, "grad_norm": 1.4196280241012573, "learning_rate": 9.912955171913536e-06, "loss": 0.0115, "step": 14006 }, { "epoch": 1.5185385949696444, "grad_norm": 0.3176099359989166, "learning_rate": 9.91222979834615e-06, "loss": 0.0136, "step": 14007 }, { "epoch": 1.5186470078057241, "grad_norm": 0.28935450315475464, "learning_rate": 9.911504424778762e-06, "loss": 0.0195, "step": 14008 }, { "epoch": 1.5187554206418041, "grad_norm": 3.226079225540161, "learning_rate": 9.910779051211374e-06, "loss": 0.0214, "step": 14009 }, { "epoch": 1.5188638334778837, "grad_norm": 0.3500001132488251, "learning_rate": 9.910053677643988e-06, "loss": 0.0033, "step": 14010 }, { "epoch": 1.5189722463139637, "grad_norm": 4.4571309089660645, "learning_rate": 9.9093283040766e-06, "loss": 0.0292, "step": 14011 }, { "epoch": 1.5190806591500432, "grad_norm": 0.5843936204910278, "learning_rate": 9.908602930509214e-06, "loss": 0.0165, "step": 14012 }, { "epoch": 1.5191890719861232, "grad_norm": 4.525378704071045, "learning_rate": 9.907877556941826e-06, "loss": 0.0084, "step": 14013 }, { "epoch": 1.519297484822203, "grad_norm": 0.972446858882904, "learning_rate": 9.907152183374438e-06, "loss": 0.0216, "step": 14014 }, { "epoch": 1.5194058976582827, "grad_norm": 1.1304885149002075, "learning_rate": 9.906426809807052e-06, "loss": 0.0082, "step": 14015 }, { "epoch": 1.5195143104943627, "grad_norm": 4.922208786010742, "learning_rate": 9.905701436239664e-06, "loss": 0.0185, "step": 14016 }, { "epoch": 1.5196227233304422, "grad_norm": 8.915943145751953, "learning_rate": 9.904976062672276e-06, "loss": 0.0757, "step": 14017 }, { "epoch": 1.5197311361665222, "grad_norm": 0.4400412440299988, "learning_rate": 9.90425068910489e-06, "loss": 0.0164, "step": 14018 }, { "epoch": 1.5198395490026018, "grad_norm": 1.2253830432891846, "learning_rate": 9.903525315537502e-06, "loss": 0.0067, "step": 14019 }, { "epoch": 1.5199479618386817, "grad_norm": 1.186362862586975, "learning_rate": 9.902799941970116e-06, "loss": 0.026, "step": 14020 }, { "epoch": 1.5200563746747615, "grad_norm": 5.774421215057373, "learning_rate": 9.902074568402728e-06, "loss": 0.0174, "step": 14021 }, { "epoch": 1.5201647875108413, "grad_norm": 8.015257835388184, "learning_rate": 9.901349194835342e-06, "loss": 0.0556, "step": 14022 }, { "epoch": 1.520273200346921, "grad_norm": 1.229756474494934, "learning_rate": 9.900623821267954e-06, "loss": 0.0428, "step": 14023 }, { "epoch": 1.5203816131830008, "grad_norm": 0.9344032406806946, "learning_rate": 9.899898447700566e-06, "loss": 0.0339, "step": 14024 }, { "epoch": 1.5204900260190808, "grad_norm": 8.651483535766602, "learning_rate": 9.899173074133179e-06, "loss": 0.0226, "step": 14025 }, { "epoch": 1.5205984388551603, "grad_norm": 0.1443266123533249, "learning_rate": 9.898447700565792e-06, "loss": 0.0011, "step": 14026 }, { "epoch": 1.5207068516912403, "grad_norm": 0.33373257517814636, "learning_rate": 9.897722326998405e-06, "loss": 0.0137, "step": 14027 }, { "epoch": 1.52081526452732, "grad_norm": 3.5696470737457275, "learning_rate": 9.896996953431017e-06, "loss": 0.0491, "step": 14028 }, { "epoch": 1.5209236773633998, "grad_norm": 0.5794336199760437, "learning_rate": 9.89627157986363e-06, "loss": 0.009, "step": 14029 }, { "epoch": 1.5210320901994796, "grad_norm": 0.3354976773262024, "learning_rate": 9.895546206296244e-06, "loss": 0.0051, "step": 14030 }, { "epoch": 1.5211405030355594, "grad_norm": 1.4250465631484985, "learning_rate": 9.894820832728856e-06, "loss": 0.0216, "step": 14031 }, { "epoch": 1.5212489158716394, "grad_norm": 18.74901008605957, "learning_rate": 9.894095459161469e-06, "loss": 0.0295, "step": 14032 }, { "epoch": 1.521357328707719, "grad_norm": 7.9372124671936035, "learning_rate": 9.893370085594082e-06, "loss": 0.0158, "step": 14033 }, { "epoch": 1.5214657415437989, "grad_norm": 0.414783239364624, "learning_rate": 9.892644712026695e-06, "loss": 0.0022, "step": 14034 }, { "epoch": 1.5215741543798784, "grad_norm": 2.303431987762451, "learning_rate": 9.891919338459307e-06, "loss": 0.0287, "step": 14035 }, { "epoch": 1.5216825672159584, "grad_norm": 1.032713770866394, "learning_rate": 9.89119396489192e-06, "loss": 0.0242, "step": 14036 }, { "epoch": 1.5217909800520382, "grad_norm": 0.2530244290828705, "learning_rate": 9.890468591324533e-06, "loss": 0.0074, "step": 14037 }, { "epoch": 1.521899392888118, "grad_norm": 3.6997716426849365, "learning_rate": 9.889743217757147e-06, "loss": 0.0511, "step": 14038 }, { "epoch": 1.5220078057241977, "grad_norm": 0.3549843430519104, "learning_rate": 9.889017844189759e-06, "loss": 0.0147, "step": 14039 }, { "epoch": 1.5221162185602775, "grad_norm": 0.19486045837402344, "learning_rate": 9.888292470622371e-06, "loss": 0.0183, "step": 14040 }, { "epoch": 1.5222246313963574, "grad_norm": 0.9780137538909912, "learning_rate": 9.887567097054985e-06, "loss": 0.0077, "step": 14041 }, { "epoch": 1.522333044232437, "grad_norm": 2.183335065841675, "learning_rate": 9.886841723487597e-06, "loss": 0.0465, "step": 14042 }, { "epoch": 1.522441457068517, "grad_norm": 0.9195458292961121, "learning_rate": 9.886116349920209e-06, "loss": 0.0135, "step": 14043 }, { "epoch": 1.5225498699045967, "grad_norm": 3.302295684814453, "learning_rate": 9.885390976352823e-06, "loss": 0.0355, "step": 14044 }, { "epoch": 1.5226582827406765, "grad_norm": 2.004056453704834, "learning_rate": 9.884665602785435e-06, "loss": 0.0638, "step": 14045 }, { "epoch": 1.5227666955767563, "grad_norm": 0.5255159139633179, "learning_rate": 9.883940229218049e-06, "loss": 0.0025, "step": 14046 }, { "epoch": 1.522875108412836, "grad_norm": 0.035793352872133255, "learning_rate": 9.883214855650661e-06, "loss": 0.0001, "step": 14047 }, { "epoch": 1.522983521248916, "grad_norm": 1.9095183610916138, "learning_rate": 9.882489482083275e-06, "loss": 0.0197, "step": 14048 }, { "epoch": 1.5230919340849955, "grad_norm": 2.8926820755004883, "learning_rate": 9.881764108515887e-06, "loss": 0.0474, "step": 14049 }, { "epoch": 1.5232003469210755, "grad_norm": 0.3516842722892761, "learning_rate": 9.881038734948499e-06, "loss": 0.0119, "step": 14050 }, { "epoch": 1.5233087597571553, "grad_norm": 0.05690811201930046, "learning_rate": 9.880313361381111e-06, "loss": 0.0007, "step": 14051 }, { "epoch": 1.523417172593235, "grad_norm": 1.4659616947174072, "learning_rate": 9.879587987813725e-06, "loss": 0.03, "step": 14052 }, { "epoch": 1.5235255854293148, "grad_norm": 1.7716917991638184, "learning_rate": 9.878862614246337e-06, "loss": 0.007, "step": 14053 }, { "epoch": 1.5236339982653946, "grad_norm": 1.4729738235473633, "learning_rate": 9.87813724067895e-06, "loss": 0.0066, "step": 14054 }, { "epoch": 1.5237424111014746, "grad_norm": 0.4664026200771332, "learning_rate": 9.877411867111563e-06, "loss": 0.0021, "step": 14055 }, { "epoch": 1.523850823937554, "grad_norm": 0.9915744066238403, "learning_rate": 9.876686493544177e-06, "loss": 0.0044, "step": 14056 }, { "epoch": 1.523959236773634, "grad_norm": 1.0783694982528687, "learning_rate": 9.875961119976789e-06, "loss": 0.0104, "step": 14057 }, { "epoch": 1.5240676496097136, "grad_norm": 7.284003257751465, "learning_rate": 9.875235746409401e-06, "loss": 0.0218, "step": 14058 }, { "epoch": 1.5241760624457936, "grad_norm": 0.3577990233898163, "learning_rate": 9.874510372842015e-06, "loss": 0.0063, "step": 14059 }, { "epoch": 1.5242844752818734, "grad_norm": 1.6168972253799438, "learning_rate": 9.873784999274627e-06, "loss": 0.0256, "step": 14060 }, { "epoch": 1.5243928881179531, "grad_norm": 1.061929702758789, "learning_rate": 9.87305962570724e-06, "loss": 0.0134, "step": 14061 }, { "epoch": 1.524501300954033, "grad_norm": 0.11972072720527649, "learning_rate": 9.872334252139853e-06, "loss": 0.001, "step": 14062 }, { "epoch": 1.5246097137901127, "grad_norm": 5.874689102172852, "learning_rate": 9.871608878572465e-06, "loss": 0.0416, "step": 14063 }, { "epoch": 1.5247181266261927, "grad_norm": 1.5427422523498535, "learning_rate": 9.87088350500508e-06, "loss": 0.0357, "step": 14064 }, { "epoch": 1.5248265394622722, "grad_norm": 4.385651588439941, "learning_rate": 9.870158131437691e-06, "loss": 0.0241, "step": 14065 }, { "epoch": 1.5249349522983522, "grad_norm": 2.946599245071411, "learning_rate": 9.869432757870304e-06, "loss": 0.044, "step": 14066 }, { "epoch": 1.525043365134432, "grad_norm": 2.268113136291504, "learning_rate": 9.868707384302917e-06, "loss": 0.0163, "step": 14067 }, { "epoch": 1.5251517779705117, "grad_norm": 0.7841516137123108, "learning_rate": 9.86798201073553e-06, "loss": 0.0112, "step": 14068 }, { "epoch": 1.5252601908065915, "grad_norm": 5.42081880569458, "learning_rate": 9.867256637168142e-06, "loss": 0.0221, "step": 14069 }, { "epoch": 1.5253686036426712, "grad_norm": 5.113323211669922, "learning_rate": 9.866531263600755e-06, "loss": 0.0432, "step": 14070 }, { "epoch": 1.5254770164787512, "grad_norm": 2.220133066177368, "learning_rate": 9.865805890033368e-06, "loss": 0.0233, "step": 14071 }, { "epoch": 1.5255854293148308, "grad_norm": 3.07249116897583, "learning_rate": 9.865080516465981e-06, "loss": 0.0394, "step": 14072 }, { "epoch": 1.5256938421509108, "grad_norm": 5.271185874938965, "learning_rate": 9.864355142898594e-06, "loss": 0.0199, "step": 14073 }, { "epoch": 1.5258022549869905, "grad_norm": 0.3088279962539673, "learning_rate": 9.863629769331207e-06, "loss": 0.0179, "step": 14074 }, { "epoch": 1.5259106678230703, "grad_norm": 8.281686782836914, "learning_rate": 9.86290439576382e-06, "loss": 0.0179, "step": 14075 }, { "epoch": 1.52601908065915, "grad_norm": 2.355510711669922, "learning_rate": 9.862179022196432e-06, "loss": 0.0426, "step": 14076 }, { "epoch": 1.5261274934952298, "grad_norm": 6.092412948608398, "learning_rate": 9.861453648629044e-06, "loss": 0.0334, "step": 14077 }, { "epoch": 1.5262359063313098, "grad_norm": 0.33503973484039307, "learning_rate": 9.860728275061658e-06, "loss": 0.0126, "step": 14078 }, { "epoch": 1.5263443191673893, "grad_norm": 2.1221163272857666, "learning_rate": 9.86000290149427e-06, "loss": 0.0199, "step": 14079 }, { "epoch": 1.5264527320034693, "grad_norm": 3.5650475025177, "learning_rate": 9.859277527926882e-06, "loss": 0.0201, "step": 14080 }, { "epoch": 1.5265611448395489, "grad_norm": 1.19663667678833, "learning_rate": 9.858552154359496e-06, "loss": 0.0085, "step": 14081 }, { "epoch": 1.5266695576756288, "grad_norm": 1.4783203601837158, "learning_rate": 9.85782678079211e-06, "loss": 0.0178, "step": 14082 }, { "epoch": 1.5267779705117086, "grad_norm": 10.46694278717041, "learning_rate": 9.857101407224722e-06, "loss": 0.015, "step": 14083 }, { "epoch": 1.5268863833477884, "grad_norm": 0.6804354190826416, "learning_rate": 9.856376033657334e-06, "loss": 0.0152, "step": 14084 }, { "epoch": 1.5269947961838681, "grad_norm": 0.3595980405807495, "learning_rate": 9.855650660089948e-06, "loss": 0.0138, "step": 14085 }, { "epoch": 1.527103209019948, "grad_norm": 0.2727457284927368, "learning_rate": 9.85492528652256e-06, "loss": 0.0029, "step": 14086 }, { "epoch": 1.5272116218560279, "grad_norm": 7.605820178985596, "learning_rate": 9.854199912955172e-06, "loss": 0.0556, "step": 14087 }, { "epoch": 1.5273200346921074, "grad_norm": 0.7393295168876648, "learning_rate": 9.853474539387786e-06, "loss": 0.0104, "step": 14088 }, { "epoch": 1.5274284475281874, "grad_norm": 3.7057695388793945, "learning_rate": 9.852749165820398e-06, "loss": 0.0392, "step": 14089 }, { "epoch": 1.5275368603642672, "grad_norm": 3.745863437652588, "learning_rate": 9.852023792253012e-06, "loss": 0.0183, "step": 14090 }, { "epoch": 1.527645273200347, "grad_norm": 4.899933815002441, "learning_rate": 9.851298418685624e-06, "loss": 0.0451, "step": 14091 }, { "epoch": 1.5277536860364267, "grad_norm": 2.836862087249756, "learning_rate": 9.850573045118236e-06, "loss": 0.0031, "step": 14092 }, { "epoch": 1.5278620988725065, "grad_norm": 0.41574332118034363, "learning_rate": 9.84984767155085e-06, "loss": 0.0191, "step": 14093 }, { "epoch": 1.5279705117085864, "grad_norm": 1.0971691608428955, "learning_rate": 9.849122297983462e-06, "loss": 0.0068, "step": 14094 }, { "epoch": 1.528078924544666, "grad_norm": 3.064652681350708, "learning_rate": 9.848396924416074e-06, "loss": 0.0325, "step": 14095 }, { "epoch": 1.528187337380746, "grad_norm": 4.878243923187256, "learning_rate": 9.847671550848688e-06, "loss": 0.0339, "step": 14096 }, { "epoch": 1.5282957502168255, "grad_norm": 5.765390872955322, "learning_rate": 9.8469461772813e-06, "loss": 0.0386, "step": 14097 }, { "epoch": 1.5284041630529055, "grad_norm": 3.057081460952759, "learning_rate": 9.846220803713914e-06, "loss": 0.0587, "step": 14098 }, { "epoch": 1.5285125758889853, "grad_norm": 1.3034204244613647, "learning_rate": 9.845495430146526e-06, "loss": 0.0262, "step": 14099 }, { "epoch": 1.528620988725065, "grad_norm": 7.038536071777344, "learning_rate": 9.84477005657914e-06, "loss": 0.0198, "step": 14100 }, { "epoch": 1.5287294015611448, "grad_norm": 1.7179368734359741, "learning_rate": 9.844044683011752e-06, "loss": 0.0281, "step": 14101 }, { "epoch": 1.5288378143972245, "grad_norm": 4.654073238372803, "learning_rate": 9.843319309444364e-06, "loss": 0.006, "step": 14102 }, { "epoch": 1.5289462272333045, "grad_norm": 3.6925575733184814, "learning_rate": 9.842593935876977e-06, "loss": 0.0219, "step": 14103 }, { "epoch": 1.529054640069384, "grad_norm": 9.5132474899292, "learning_rate": 9.84186856230959e-06, "loss": 0.0555, "step": 14104 }, { "epoch": 1.529163052905464, "grad_norm": 3.235464572906494, "learning_rate": 9.841143188742203e-06, "loss": 0.0286, "step": 14105 }, { "epoch": 1.5292714657415438, "grad_norm": 6.742948532104492, "learning_rate": 9.840417815174815e-06, "loss": 0.0389, "step": 14106 }, { "epoch": 1.5293798785776236, "grad_norm": 3.799631357192993, "learning_rate": 9.839692441607429e-06, "loss": 0.0296, "step": 14107 }, { "epoch": 1.5294882914137033, "grad_norm": 0.8680015206336975, "learning_rate": 9.838967068040042e-06, "loss": 0.0091, "step": 14108 }, { "epoch": 1.5295967042497831, "grad_norm": 2.5961477756500244, "learning_rate": 9.838241694472654e-06, "loss": 0.0319, "step": 14109 }, { "epoch": 1.529705117085863, "grad_norm": 2.9924488067626953, "learning_rate": 9.837516320905267e-06, "loss": 0.0557, "step": 14110 }, { "epoch": 1.5298135299219426, "grad_norm": 0.5738507509231567, "learning_rate": 9.83679094733788e-06, "loss": 0.0141, "step": 14111 }, { "epoch": 1.5299219427580226, "grad_norm": 0.4073778986930847, "learning_rate": 9.836065573770493e-06, "loss": 0.0115, "step": 14112 }, { "epoch": 1.5300303555941024, "grad_norm": 3.2063024044036865, "learning_rate": 9.835340200203105e-06, "loss": 0.0336, "step": 14113 }, { "epoch": 1.5301387684301822, "grad_norm": 2.3342208862304688, "learning_rate": 9.834614826635719e-06, "loss": 0.0609, "step": 14114 }, { "epoch": 1.530247181266262, "grad_norm": 12.654125213623047, "learning_rate": 9.83388945306833e-06, "loss": 0.022, "step": 14115 }, { "epoch": 1.5303555941023417, "grad_norm": 0.31898704171180725, "learning_rate": 9.833164079500945e-06, "loss": 0.0124, "step": 14116 }, { "epoch": 1.5304640069384217, "grad_norm": 1.1726064682006836, "learning_rate": 9.832438705933557e-06, "loss": 0.0386, "step": 14117 }, { "epoch": 1.5305724197745012, "grad_norm": 1.8104488849639893, "learning_rate": 9.831713332366169e-06, "loss": 0.0149, "step": 14118 }, { "epoch": 1.5306808326105812, "grad_norm": 11.07103157043457, "learning_rate": 9.830987958798783e-06, "loss": 0.0136, "step": 14119 }, { "epoch": 1.5307892454466607, "grad_norm": 7.182590484619141, "learning_rate": 9.830262585231395e-06, "loss": 0.0443, "step": 14120 }, { "epoch": 1.5308976582827407, "grad_norm": 2.291147470474243, "learning_rate": 9.829537211664007e-06, "loss": 0.0407, "step": 14121 }, { "epoch": 1.5310060711188205, "grad_norm": 6.500705718994141, "learning_rate": 9.82881183809662e-06, "loss": 0.0332, "step": 14122 }, { "epoch": 1.5311144839549002, "grad_norm": 0.9915448427200317, "learning_rate": 9.828086464529233e-06, "loss": 0.0122, "step": 14123 }, { "epoch": 1.53122289679098, "grad_norm": 1.5615832805633545, "learning_rate": 9.827361090961847e-06, "loss": 0.0561, "step": 14124 }, { "epoch": 1.5313313096270598, "grad_norm": 2.6590323448181152, "learning_rate": 9.826635717394459e-06, "loss": 0.0444, "step": 14125 }, { "epoch": 1.5314397224631398, "grad_norm": 11.874258995056152, "learning_rate": 9.825910343827073e-06, "loss": 0.072, "step": 14126 }, { "epoch": 1.5315481352992193, "grad_norm": 0.9090726971626282, "learning_rate": 9.825184970259685e-06, "loss": 0.0167, "step": 14127 }, { "epoch": 1.5316565481352993, "grad_norm": 0.3905547559261322, "learning_rate": 9.824459596692297e-06, "loss": 0.0123, "step": 14128 }, { "epoch": 1.531764960971379, "grad_norm": 2.620966672897339, "learning_rate": 9.82373422312491e-06, "loss": 0.0391, "step": 14129 }, { "epoch": 1.5318733738074588, "grad_norm": 0.6413389444351196, "learning_rate": 9.823008849557523e-06, "loss": 0.0113, "step": 14130 }, { "epoch": 1.5319817866435386, "grad_norm": 2.489929676055908, "learning_rate": 9.822283475990135e-06, "loss": 0.0134, "step": 14131 }, { "epoch": 1.5320901994796183, "grad_norm": 0.2530302405357361, "learning_rate": 9.821558102422747e-06, "loss": 0.0048, "step": 14132 }, { "epoch": 1.5321986123156983, "grad_norm": 1.8314844369888306, "learning_rate": 9.820832728855361e-06, "loss": 0.0117, "step": 14133 }, { "epoch": 1.5323070251517779, "grad_norm": 2.0737855434417725, "learning_rate": 9.820107355287975e-06, "loss": 0.0271, "step": 14134 }, { "epoch": 1.5324154379878578, "grad_norm": 1.5710663795471191, "learning_rate": 9.819381981720587e-06, "loss": 0.0212, "step": 14135 }, { "epoch": 1.5325238508239376, "grad_norm": 5.278850555419922, "learning_rate": 9.8186566081532e-06, "loss": 0.0186, "step": 14136 }, { "epoch": 1.5326322636600174, "grad_norm": 2.775195360183716, "learning_rate": 9.817931234585813e-06, "loss": 0.0241, "step": 14137 }, { "epoch": 1.5327406764960971, "grad_norm": 2.283586025238037, "learning_rate": 9.817205861018425e-06, "loss": 0.024, "step": 14138 }, { "epoch": 1.532849089332177, "grad_norm": 2.2665908336639404, "learning_rate": 9.816480487451037e-06, "loss": 0.0425, "step": 14139 }, { "epoch": 1.5329575021682569, "grad_norm": 0.5811182856559753, "learning_rate": 9.815755113883651e-06, "loss": 0.0104, "step": 14140 }, { "epoch": 1.5330659150043364, "grad_norm": 4.496435165405273, "learning_rate": 9.815029740316263e-06, "loss": 0.0318, "step": 14141 }, { "epoch": 1.5331743278404164, "grad_norm": 3.1094164848327637, "learning_rate": 9.814304366748877e-06, "loss": 0.0504, "step": 14142 }, { "epoch": 1.533282740676496, "grad_norm": 0.0919840857386589, "learning_rate": 9.81357899318149e-06, "loss": 0.0011, "step": 14143 }, { "epoch": 1.533391153512576, "grad_norm": 1.6852803230285645, "learning_rate": 9.812853619614102e-06, "loss": 0.0564, "step": 14144 }, { "epoch": 1.5334995663486557, "grad_norm": 0.5061220526695251, "learning_rate": 9.812128246046715e-06, "loss": 0.0112, "step": 14145 }, { "epoch": 1.5336079791847355, "grad_norm": 2.406061887741089, "learning_rate": 9.811402872479328e-06, "loss": 0.0148, "step": 14146 }, { "epoch": 1.5337163920208152, "grad_norm": 1.064393162727356, "learning_rate": 9.81067749891194e-06, "loss": 0.0228, "step": 14147 }, { "epoch": 1.533824804856895, "grad_norm": 0.9168946146965027, "learning_rate": 9.809952125344553e-06, "loss": 0.0158, "step": 14148 }, { "epoch": 1.533933217692975, "grad_norm": 0.22945860028266907, "learning_rate": 9.809226751777166e-06, "loss": 0.0023, "step": 14149 }, { "epoch": 1.5340416305290545, "grad_norm": 0.20469613373279572, "learning_rate": 9.80850137820978e-06, "loss": 0.0013, "step": 14150 }, { "epoch": 1.5341500433651345, "grad_norm": 3.0988473892211914, "learning_rate": 9.807776004642392e-06, "loss": 0.033, "step": 14151 }, { "epoch": 1.5342584562012143, "grad_norm": 1.94771146774292, "learning_rate": 9.807050631075005e-06, "loss": 0.0484, "step": 14152 }, { "epoch": 1.534366869037294, "grad_norm": 1.2774301767349243, "learning_rate": 9.806325257507618e-06, "loss": 0.0324, "step": 14153 }, { "epoch": 1.5344752818733738, "grad_norm": 1.660470724105835, "learning_rate": 9.80559988394023e-06, "loss": 0.0352, "step": 14154 }, { "epoch": 1.5345836947094535, "grad_norm": 1.9577908515930176, "learning_rate": 9.804874510372842e-06, "loss": 0.0167, "step": 14155 }, { "epoch": 1.5346921075455335, "grad_norm": 1.0718907117843628, "learning_rate": 9.804149136805456e-06, "loss": 0.0207, "step": 14156 }, { "epoch": 1.534800520381613, "grad_norm": 4.317446231842041, "learning_rate": 9.803423763238068e-06, "loss": 0.0076, "step": 14157 }, { "epoch": 1.534908933217693, "grad_norm": 1.887144684791565, "learning_rate": 9.80269838967068e-06, "loss": 0.0154, "step": 14158 }, { "epoch": 1.5350173460537726, "grad_norm": 5.387886047363281, "learning_rate": 9.801973016103294e-06, "loss": 0.0338, "step": 14159 }, { "epoch": 1.5351257588898526, "grad_norm": 3.9545106887817383, "learning_rate": 9.801247642535908e-06, "loss": 0.035, "step": 14160 }, { "epoch": 1.5352341717259324, "grad_norm": 0.4820278584957123, "learning_rate": 9.80052226896852e-06, "loss": 0.0076, "step": 14161 }, { "epoch": 1.5353425845620121, "grad_norm": 1.4881162643432617, "learning_rate": 9.799796895401132e-06, "loss": 0.0164, "step": 14162 }, { "epoch": 1.535450997398092, "grad_norm": 1.9527273178100586, "learning_rate": 9.799071521833746e-06, "loss": 0.0131, "step": 14163 }, { "epoch": 1.5355594102341716, "grad_norm": 1.3417109251022339, "learning_rate": 9.798346148266358e-06, "loss": 0.034, "step": 14164 }, { "epoch": 1.5356678230702516, "grad_norm": 0.4087957739830017, "learning_rate": 9.79762077469897e-06, "loss": 0.0183, "step": 14165 }, { "epoch": 1.5357762359063312, "grad_norm": 5.656579494476318, "learning_rate": 9.796895401131584e-06, "loss": 0.0265, "step": 14166 }, { "epoch": 1.5358846487424112, "grad_norm": 0.3845059871673584, "learning_rate": 9.796170027564196e-06, "loss": 0.0088, "step": 14167 }, { "epoch": 1.535993061578491, "grad_norm": 0.5110628008842468, "learning_rate": 9.79544465399681e-06, "loss": 0.0133, "step": 14168 }, { "epoch": 1.5361014744145707, "grad_norm": 0.7832975387573242, "learning_rate": 9.794719280429422e-06, "loss": 0.0213, "step": 14169 }, { "epoch": 1.5362098872506504, "grad_norm": 0.3969884514808655, "learning_rate": 9.793993906862034e-06, "loss": 0.0104, "step": 14170 }, { "epoch": 1.5363183000867302, "grad_norm": 2.2762680053710938, "learning_rate": 9.793268533294648e-06, "loss": 0.0741, "step": 14171 }, { "epoch": 1.5364267129228102, "grad_norm": 1.7226719856262207, "learning_rate": 9.79254315972726e-06, "loss": 0.0092, "step": 14172 }, { "epoch": 1.5365351257588897, "grad_norm": 0.5983474850654602, "learning_rate": 9.791817786159872e-06, "loss": 0.0254, "step": 14173 }, { "epoch": 1.5366435385949697, "grad_norm": 0.15636974573135376, "learning_rate": 9.791092412592486e-06, "loss": 0.002, "step": 14174 }, { "epoch": 1.5367519514310495, "grad_norm": 3.5100886821746826, "learning_rate": 9.790367039025098e-06, "loss": 0.0139, "step": 14175 }, { "epoch": 1.5368603642671292, "grad_norm": 1.3131223917007446, "learning_rate": 9.789641665457712e-06, "loss": 0.0364, "step": 14176 }, { "epoch": 1.536968777103209, "grad_norm": 5.046243667602539, "learning_rate": 9.788916291890324e-06, "loss": 0.0324, "step": 14177 }, { "epoch": 1.5370771899392888, "grad_norm": 3.348933696746826, "learning_rate": 9.788190918322938e-06, "loss": 0.0456, "step": 14178 }, { "epoch": 1.5371856027753688, "grad_norm": 1.9305286407470703, "learning_rate": 9.78746554475555e-06, "loss": 0.0407, "step": 14179 }, { "epoch": 1.5372940156114483, "grad_norm": 0.5359037518501282, "learning_rate": 9.786740171188162e-06, "loss": 0.0036, "step": 14180 }, { "epoch": 1.5374024284475283, "grad_norm": 1.4307565689086914, "learning_rate": 9.786014797620775e-06, "loss": 0.0454, "step": 14181 }, { "epoch": 1.5375108412836078, "grad_norm": 1.681573510169983, "learning_rate": 9.785289424053388e-06, "loss": 0.0128, "step": 14182 }, { "epoch": 1.5376192541196878, "grad_norm": 0.8539377450942993, "learning_rate": 9.784564050486e-06, "loss": 0.0173, "step": 14183 }, { "epoch": 1.5377276669557676, "grad_norm": 1.814848780632019, "learning_rate": 9.783838676918613e-06, "loss": 0.0362, "step": 14184 }, { "epoch": 1.5378360797918473, "grad_norm": 3.216571569442749, "learning_rate": 9.783113303351227e-06, "loss": 0.0452, "step": 14185 }, { "epoch": 1.537944492627927, "grad_norm": 1.5659964084625244, "learning_rate": 9.78238792978384e-06, "loss": 0.018, "step": 14186 }, { "epoch": 1.5380529054640069, "grad_norm": 0.34081560373306274, "learning_rate": 9.781662556216452e-06, "loss": 0.0076, "step": 14187 }, { "epoch": 1.5381613183000868, "grad_norm": 3.6170084476470947, "learning_rate": 9.780937182649065e-06, "loss": 0.0418, "step": 14188 }, { "epoch": 1.5382697311361664, "grad_norm": 0.8063557147979736, "learning_rate": 9.780211809081678e-06, "loss": 0.0153, "step": 14189 }, { "epoch": 1.5383781439722464, "grad_norm": 0.7018044590950012, "learning_rate": 9.77948643551429e-06, "loss": 0.0253, "step": 14190 }, { "epoch": 1.5384865568083261, "grad_norm": 0.7329677939414978, "learning_rate": 9.778761061946903e-06, "loss": 0.0226, "step": 14191 }, { "epoch": 1.538594969644406, "grad_norm": 0.683428168296814, "learning_rate": 9.778035688379517e-06, "loss": 0.0078, "step": 14192 }, { "epoch": 1.5387033824804857, "grad_norm": 0.4713193476200104, "learning_rate": 9.777310314812129e-06, "loss": 0.0097, "step": 14193 }, { "epoch": 1.5388117953165654, "grad_norm": 0.40563878417015076, "learning_rate": 9.776584941244743e-06, "loss": 0.0029, "step": 14194 }, { "epoch": 1.5389202081526454, "grad_norm": 3.840785503387451, "learning_rate": 9.775859567677355e-06, "loss": 0.0545, "step": 14195 }, { "epoch": 1.539028620988725, "grad_norm": 1.7307487726211548, "learning_rate": 9.775134194109967e-06, "loss": 0.0315, "step": 14196 }, { "epoch": 1.539137033824805, "grad_norm": 1.6894508600234985, "learning_rate": 9.77440882054258e-06, "loss": 0.0196, "step": 14197 }, { "epoch": 1.5392454466608847, "grad_norm": 10.10802173614502, "learning_rate": 9.773683446975193e-06, "loss": 0.0338, "step": 14198 }, { "epoch": 1.5393538594969645, "grad_norm": 1.6282193660736084, "learning_rate": 9.772958073407805e-06, "loss": 0.0295, "step": 14199 }, { "epoch": 1.5394622723330442, "grad_norm": 2.5938868522644043, "learning_rate": 9.772232699840419e-06, "loss": 0.0308, "step": 14200 }, { "epoch": 1.539570685169124, "grad_norm": 0.5703275203704834, "learning_rate": 9.771507326273031e-06, "loss": 0.0087, "step": 14201 }, { "epoch": 1.539679098005204, "grad_norm": 0.28570839762687683, "learning_rate": 9.770781952705645e-06, "loss": 0.0056, "step": 14202 }, { "epoch": 1.5397875108412835, "grad_norm": 7.754755020141602, "learning_rate": 9.770056579138257e-06, "loss": 0.0909, "step": 14203 }, { "epoch": 1.5398959236773635, "grad_norm": 0.8202523589134216, "learning_rate": 9.76933120557087e-06, "loss": 0.0074, "step": 14204 }, { "epoch": 1.540004336513443, "grad_norm": 6.040383338928223, "learning_rate": 9.768605832003483e-06, "loss": 0.0211, "step": 14205 }, { "epoch": 1.540112749349523, "grad_norm": 2.0398380756378174, "learning_rate": 9.767880458436095e-06, "loss": 0.0637, "step": 14206 }, { "epoch": 1.5402211621856028, "grad_norm": 2.0276756286621094, "learning_rate": 9.767155084868707e-06, "loss": 0.0242, "step": 14207 }, { "epoch": 1.5403295750216826, "grad_norm": 2.30865478515625, "learning_rate": 9.766429711301321e-06, "loss": 0.0613, "step": 14208 }, { "epoch": 1.5404379878577623, "grad_norm": 5.5703582763671875, "learning_rate": 9.765704337733933e-06, "loss": 0.0377, "step": 14209 }, { "epoch": 1.540546400693842, "grad_norm": 0.09865517914295197, "learning_rate": 9.764978964166547e-06, "loss": 0.0014, "step": 14210 }, { "epoch": 1.540654813529922, "grad_norm": 0.6455672979354858, "learning_rate": 9.76425359059916e-06, "loss": 0.0049, "step": 14211 }, { "epoch": 1.5407632263660016, "grad_norm": 3.350445508956909, "learning_rate": 9.763528217031773e-06, "loss": 0.0208, "step": 14212 }, { "epoch": 1.5408716392020816, "grad_norm": 3.2730677127838135, "learning_rate": 9.762802843464385e-06, "loss": 0.0666, "step": 14213 }, { "epoch": 1.5409800520381614, "grad_norm": 0.313088595867157, "learning_rate": 9.762077469896997e-06, "loss": 0.0084, "step": 14214 }, { "epoch": 1.5410884648742411, "grad_norm": 3.1464035511016846, "learning_rate": 9.761352096329611e-06, "loss": 0.0225, "step": 14215 }, { "epoch": 1.5411968777103209, "grad_norm": 1.2715319395065308, "learning_rate": 9.760626722762223e-06, "loss": 0.0258, "step": 14216 }, { "epoch": 1.5413052905464006, "grad_norm": 1.0077072381973267, "learning_rate": 9.759901349194835e-06, "loss": 0.0371, "step": 14217 }, { "epoch": 1.5414137033824806, "grad_norm": 0.538901686668396, "learning_rate": 9.75917597562745e-06, "loss": 0.0047, "step": 14218 }, { "epoch": 1.5415221162185602, "grad_norm": 0.17507821321487427, "learning_rate": 9.758450602060061e-06, "loss": 0.0021, "step": 14219 }, { "epoch": 1.5416305290546402, "grad_norm": 3.040862560272217, "learning_rate": 9.757725228492675e-06, "loss": 0.0221, "step": 14220 }, { "epoch": 1.54173894189072, "grad_norm": 1.1855416297912598, "learning_rate": 9.756999854925287e-06, "loss": 0.0333, "step": 14221 }, { "epoch": 1.5418473547267997, "grad_norm": 7.301022529602051, "learning_rate": 9.7562744813579e-06, "loss": 0.0713, "step": 14222 }, { "epoch": 1.5419557675628794, "grad_norm": 3.768073558807373, "learning_rate": 9.755549107790513e-06, "loss": 0.0166, "step": 14223 }, { "epoch": 1.5420641803989592, "grad_norm": 1.2309479713439941, "learning_rate": 9.754823734223126e-06, "loss": 0.0233, "step": 14224 }, { "epoch": 1.5421725932350392, "grad_norm": 9.727949142456055, "learning_rate": 9.754098360655738e-06, "loss": 0.0289, "step": 14225 }, { "epoch": 1.5422810060711187, "grad_norm": 0.23818820714950562, "learning_rate": 9.753372987088351e-06, "loss": 0.0057, "step": 14226 }, { "epoch": 1.5423894189071987, "grad_norm": 6.333088397979736, "learning_rate": 9.752647613520964e-06, "loss": 0.0224, "step": 14227 }, { "epoch": 1.5424978317432783, "grad_norm": 3.9880950450897217, "learning_rate": 9.751922239953577e-06, "loss": 0.0386, "step": 14228 }, { "epoch": 1.5426062445793582, "grad_norm": 3.540902614593506, "learning_rate": 9.75119686638619e-06, "loss": 0.0308, "step": 14229 }, { "epoch": 1.542714657415438, "grad_norm": 2.225921154022217, "learning_rate": 9.750471492818803e-06, "loss": 0.029, "step": 14230 }, { "epoch": 1.5428230702515178, "grad_norm": 1.5677889585494995, "learning_rate": 9.749746119251416e-06, "loss": 0.0168, "step": 14231 }, { "epoch": 1.5429314830875975, "grad_norm": 1.761297345161438, "learning_rate": 9.749020745684028e-06, "loss": 0.0208, "step": 14232 }, { "epoch": 1.5430398959236773, "grad_norm": 0.25906676054000854, "learning_rate": 9.74829537211664e-06, "loss": 0.0035, "step": 14233 }, { "epoch": 1.5431483087597573, "grad_norm": 1.5235917568206787, "learning_rate": 9.747569998549254e-06, "loss": 0.0379, "step": 14234 }, { "epoch": 1.5432567215958368, "grad_norm": 3.985288619995117, "learning_rate": 9.746844624981866e-06, "loss": 0.0848, "step": 14235 }, { "epoch": 1.5433651344319168, "grad_norm": 0.22136467695236206, "learning_rate": 9.74611925141448e-06, "loss": 0.0016, "step": 14236 }, { "epoch": 1.5434735472679966, "grad_norm": 0.5903112888336182, "learning_rate": 9.745393877847092e-06, "loss": 0.0239, "step": 14237 }, { "epoch": 1.5435819601040763, "grad_norm": 0.8035132884979248, "learning_rate": 9.744668504279706e-06, "loss": 0.0149, "step": 14238 }, { "epoch": 1.543690372940156, "grad_norm": 0.8369925618171692, "learning_rate": 9.743943130712318e-06, "loss": 0.0119, "step": 14239 }, { "epoch": 1.5437987857762359, "grad_norm": 12.302245140075684, "learning_rate": 9.74321775714493e-06, "loss": 0.0181, "step": 14240 }, { "epoch": 1.5439071986123158, "grad_norm": 6.088554382324219, "learning_rate": 9.742492383577544e-06, "loss": 0.0407, "step": 14241 }, { "epoch": 1.5440156114483954, "grad_norm": 0.27411317825317383, "learning_rate": 9.741767010010156e-06, "loss": 0.014, "step": 14242 }, { "epoch": 1.5441240242844754, "grad_norm": 1.8577672243118286, "learning_rate": 9.741041636442768e-06, "loss": 0.027, "step": 14243 }, { "epoch": 1.544232437120555, "grad_norm": 1.4278125762939453, "learning_rate": 9.740316262875382e-06, "loss": 0.0061, "step": 14244 }, { "epoch": 1.544340849956635, "grad_norm": 1.5784997940063477, "learning_rate": 9.739590889307994e-06, "loss": 0.023, "step": 14245 }, { "epoch": 1.5444492627927147, "grad_norm": 1.0061922073364258, "learning_rate": 9.738865515740608e-06, "loss": 0.0033, "step": 14246 }, { "epoch": 1.5445576756287944, "grad_norm": 1.6510159969329834, "learning_rate": 9.73814014217322e-06, "loss": 0.0217, "step": 14247 }, { "epoch": 1.5446660884648742, "grad_norm": 1.188815951347351, "learning_rate": 9.737414768605832e-06, "loss": 0.0274, "step": 14248 }, { "epoch": 1.544774501300954, "grad_norm": 0.6939932703971863, "learning_rate": 9.736689395038446e-06, "loss": 0.0055, "step": 14249 }, { "epoch": 1.544882914137034, "grad_norm": 0.206160768866539, "learning_rate": 9.735964021471058e-06, "loss": 0.0046, "step": 14250 }, { "epoch": 1.5449913269731135, "grad_norm": 7.90289831161499, "learning_rate": 9.73523864790367e-06, "loss": 0.0456, "step": 14251 }, { "epoch": 1.5450997398091935, "grad_norm": 3.565187454223633, "learning_rate": 9.734513274336284e-06, "loss": 0.0043, "step": 14252 }, { "epoch": 1.5452081526452732, "grad_norm": 2.035597324371338, "learning_rate": 9.733787900768896e-06, "loss": 0.0247, "step": 14253 }, { "epoch": 1.545316565481353, "grad_norm": 0.4691403806209564, "learning_rate": 9.73306252720151e-06, "loss": 0.0037, "step": 14254 }, { "epoch": 1.5454249783174328, "grad_norm": 3.05180287361145, "learning_rate": 9.732337153634122e-06, "loss": 0.0223, "step": 14255 }, { "epoch": 1.5455333911535125, "grad_norm": 8.630744934082031, "learning_rate": 9.731611780066736e-06, "loss": 0.0317, "step": 14256 }, { "epoch": 1.5456418039895925, "grad_norm": 0.4514458477497101, "learning_rate": 9.730886406499348e-06, "loss": 0.0157, "step": 14257 }, { "epoch": 1.545750216825672, "grad_norm": 5.610213279724121, "learning_rate": 9.73016103293196e-06, "loss": 0.0298, "step": 14258 }, { "epoch": 1.545858629661752, "grad_norm": 3.099494218826294, "learning_rate": 9.729435659364573e-06, "loss": 0.0377, "step": 14259 }, { "epoch": 1.5459670424978318, "grad_norm": 1.0864976644515991, "learning_rate": 9.728710285797186e-06, "loss": 0.0048, "step": 14260 }, { "epoch": 1.5460754553339116, "grad_norm": 0.864417552947998, "learning_rate": 9.727984912229799e-06, "loss": 0.0123, "step": 14261 }, { "epoch": 1.5461838681699913, "grad_norm": 0.933826744556427, "learning_rate": 9.727259538662412e-06, "loss": 0.006, "step": 14262 }, { "epoch": 1.546292281006071, "grad_norm": 16.29800796508789, "learning_rate": 9.726534165095025e-06, "loss": 0.0244, "step": 14263 }, { "epoch": 1.546400693842151, "grad_norm": 0.21541021764278412, "learning_rate": 9.725808791527638e-06, "loss": 0.0017, "step": 14264 }, { "epoch": 1.5465091066782306, "grad_norm": 0.24870146811008453, "learning_rate": 9.72508341796025e-06, "loss": 0.0031, "step": 14265 }, { "epoch": 1.5466175195143106, "grad_norm": 5.878173828125, "learning_rate": 9.724358044392863e-06, "loss": 0.0111, "step": 14266 }, { "epoch": 1.5467259323503901, "grad_norm": 0.2257332056760788, "learning_rate": 9.723632670825476e-06, "loss": 0.0023, "step": 14267 }, { "epoch": 1.5468343451864701, "grad_norm": 0.38451868295669556, "learning_rate": 9.722907297258089e-06, "loss": 0.0097, "step": 14268 }, { "epoch": 1.5469427580225499, "grad_norm": 0.4832967221736908, "learning_rate": 9.7221819236907e-06, "loss": 0.0116, "step": 14269 }, { "epoch": 1.5470511708586296, "grad_norm": 0.4503129720687866, "learning_rate": 9.721456550123315e-06, "loss": 0.0051, "step": 14270 }, { "epoch": 1.5471595836947094, "grad_norm": 2.1460459232330322, "learning_rate": 9.720731176555927e-06, "loss": 0.0124, "step": 14271 }, { "epoch": 1.5472679965307892, "grad_norm": 3.771519422531128, "learning_rate": 9.72000580298854e-06, "loss": 0.0337, "step": 14272 }, { "epoch": 1.5473764093668692, "grad_norm": 2.524446725845337, "learning_rate": 9.719280429421153e-06, "loss": 0.0108, "step": 14273 }, { "epoch": 1.5474848222029487, "grad_norm": 0.4135926067829132, "learning_rate": 9.718555055853765e-06, "loss": 0.0032, "step": 14274 }, { "epoch": 1.5475932350390287, "grad_norm": 1.872139811515808, "learning_rate": 9.717829682286379e-06, "loss": 0.0359, "step": 14275 }, { "epoch": 1.5477016478751084, "grad_norm": 0.7308757305145264, "learning_rate": 9.717104308718991e-06, "loss": 0.0027, "step": 14276 }, { "epoch": 1.5478100607111882, "grad_norm": 1.489060878753662, "learning_rate": 9.716378935151603e-06, "loss": 0.0083, "step": 14277 }, { "epoch": 1.547918473547268, "grad_norm": 0.031891416758298874, "learning_rate": 9.715653561584217e-06, "loss": 0.0003, "step": 14278 }, { "epoch": 1.5480268863833477, "grad_norm": 2.3057682514190674, "learning_rate": 9.714928188016829e-06, "loss": 0.0424, "step": 14279 }, { "epoch": 1.5481352992194277, "grad_norm": 3.43430233001709, "learning_rate": 9.714202814449443e-06, "loss": 0.046, "step": 14280 }, { "epoch": 1.5482437120555073, "grad_norm": 0.6119400858879089, "learning_rate": 9.713477440882055e-06, "loss": 0.0263, "step": 14281 }, { "epoch": 1.5483521248915872, "grad_norm": 1.8252965211868286, "learning_rate": 9.712752067314669e-06, "loss": 0.0173, "step": 14282 }, { "epoch": 1.548460537727667, "grad_norm": 1.1919240951538086, "learning_rate": 9.712026693747281e-06, "loss": 0.0421, "step": 14283 }, { "epoch": 1.5485689505637468, "grad_norm": 0.7420903444290161, "learning_rate": 9.711301320179893e-06, "loss": 0.0024, "step": 14284 }, { "epoch": 1.5486773633998265, "grad_norm": 2.3880302906036377, "learning_rate": 9.710575946612505e-06, "loss": 0.0268, "step": 14285 }, { "epoch": 1.5487857762359063, "grad_norm": 4.141277313232422, "learning_rate": 9.709850573045119e-06, "loss": 0.0538, "step": 14286 }, { "epoch": 1.5488941890719863, "grad_norm": 2.234837532043457, "learning_rate": 9.709125199477731e-06, "loss": 0.0143, "step": 14287 }, { "epoch": 1.5490026019080658, "grad_norm": 2.9684622287750244, "learning_rate": 9.708399825910345e-06, "loss": 0.035, "step": 14288 }, { "epoch": 1.5491110147441458, "grad_norm": 6.323471546173096, "learning_rate": 9.707674452342957e-06, "loss": 0.0071, "step": 14289 }, { "epoch": 1.5492194275802254, "grad_norm": 5.853045463562012, "learning_rate": 9.706949078775571e-06, "loss": 0.0432, "step": 14290 }, { "epoch": 1.5493278404163053, "grad_norm": 2.5343146324157715, "learning_rate": 9.706223705208183e-06, "loss": 0.0729, "step": 14291 }, { "epoch": 1.549436253252385, "grad_norm": 2.0574333667755127, "learning_rate": 9.705498331640795e-06, "loss": 0.0743, "step": 14292 }, { "epoch": 1.5495446660884649, "grad_norm": 3.0196924209594727, "learning_rate": 9.704772958073409e-06, "loss": 0.0266, "step": 14293 }, { "epoch": 1.5496530789245446, "grad_norm": 1.5590107440948486, "learning_rate": 9.704047584506021e-06, "loss": 0.0051, "step": 14294 }, { "epoch": 1.5497614917606244, "grad_norm": 5.6481523513793945, "learning_rate": 9.703322210938633e-06, "loss": 0.0317, "step": 14295 }, { "epoch": 1.5498699045967044, "grad_norm": 0.23361477255821228, "learning_rate": 9.702596837371247e-06, "loss": 0.0009, "step": 14296 }, { "epoch": 1.549978317432784, "grad_norm": 1.150643229484558, "learning_rate": 9.70187146380386e-06, "loss": 0.0123, "step": 14297 }, { "epoch": 1.550086730268864, "grad_norm": 0.07433068752288818, "learning_rate": 9.701146090236473e-06, "loss": 0.0007, "step": 14298 }, { "epoch": 1.5501951431049437, "grad_norm": 4.353437423706055, "learning_rate": 9.700420716669085e-06, "loss": 0.0088, "step": 14299 }, { "epoch": 1.5503035559410234, "grad_norm": 0.6696186661720276, "learning_rate": 9.699695343101698e-06, "loss": 0.0161, "step": 14300 }, { "epoch": 1.5504119687771032, "grad_norm": 1.533119797706604, "learning_rate": 9.698969969534311e-06, "loss": 0.0096, "step": 14301 }, { "epoch": 1.550520381613183, "grad_norm": 0.541222333908081, "learning_rate": 9.698244595966924e-06, "loss": 0.0162, "step": 14302 }, { "epoch": 1.550628794449263, "grad_norm": 3.2466251850128174, "learning_rate": 9.697519222399536e-06, "loss": 0.0152, "step": 14303 }, { "epoch": 1.5507372072853425, "grad_norm": 0.046361714601516724, "learning_rate": 9.69679384883215e-06, "loss": 0.0004, "step": 14304 }, { "epoch": 1.5508456201214225, "grad_norm": 0.2714187502861023, "learning_rate": 9.696068475264763e-06, "loss": 0.0159, "step": 14305 }, { "epoch": 1.550954032957502, "grad_norm": 3.4604899883270264, "learning_rate": 9.695343101697375e-06, "loss": 0.0633, "step": 14306 }, { "epoch": 1.551062445793582, "grad_norm": 0.23542198538780212, "learning_rate": 9.694617728129988e-06, "loss": 0.0048, "step": 14307 }, { "epoch": 1.5511708586296618, "grad_norm": 3.6953725814819336, "learning_rate": 9.693892354562601e-06, "loss": 0.035, "step": 14308 }, { "epoch": 1.5512792714657415, "grad_norm": 1.974953055381775, "learning_rate": 9.693166980995214e-06, "loss": 0.0133, "step": 14309 }, { "epoch": 1.5513876843018215, "grad_norm": 0.5281949639320374, "learning_rate": 9.692441607427826e-06, "loss": 0.0128, "step": 14310 }, { "epoch": 1.551496097137901, "grad_norm": 0.7389776110649109, "learning_rate": 9.691716233860438e-06, "loss": 0.0114, "step": 14311 }, { "epoch": 1.551604509973981, "grad_norm": 1.4978034496307373, "learning_rate": 9.690990860293052e-06, "loss": 0.0242, "step": 14312 }, { "epoch": 1.5517129228100606, "grad_norm": 0.904170572757721, "learning_rate": 9.690265486725664e-06, "loss": 0.0172, "step": 14313 }, { "epoch": 1.5518213356461406, "grad_norm": 0.610355794429779, "learning_rate": 9.689540113158278e-06, "loss": 0.0257, "step": 14314 }, { "epoch": 1.5519297484822203, "grad_norm": 0.9704105257987976, "learning_rate": 9.68881473959089e-06, "loss": 0.0058, "step": 14315 }, { "epoch": 1.5520381613183, "grad_norm": 1.1740732192993164, "learning_rate": 9.688089366023504e-06, "loss": 0.0032, "step": 14316 }, { "epoch": 1.5521465741543798, "grad_norm": 0.6114950776100159, "learning_rate": 9.687363992456116e-06, "loss": 0.0047, "step": 14317 }, { "epoch": 1.5522549869904596, "grad_norm": 3.5958330631256104, "learning_rate": 9.686638618888728e-06, "loss": 0.0509, "step": 14318 }, { "epoch": 1.5523633998265396, "grad_norm": 2.3642823696136475, "learning_rate": 9.685913245321342e-06, "loss": 0.0083, "step": 14319 }, { "epoch": 1.5524718126626191, "grad_norm": 0.40794843435287476, "learning_rate": 9.685187871753954e-06, "loss": 0.0106, "step": 14320 }, { "epoch": 1.5525802254986991, "grad_norm": 3.4767239093780518, "learning_rate": 9.684462498186566e-06, "loss": 0.0232, "step": 14321 }, { "epoch": 1.5526886383347789, "grad_norm": 1.9811172485351562, "learning_rate": 9.68373712461918e-06, "loss": 0.0047, "step": 14322 }, { "epoch": 1.5527970511708586, "grad_norm": 5.38528299331665, "learning_rate": 9.683011751051792e-06, "loss": 0.0444, "step": 14323 }, { "epoch": 1.5529054640069384, "grad_norm": 3.650357961654663, "learning_rate": 9.682286377484406e-06, "loss": 0.0325, "step": 14324 }, { "epoch": 1.5530138768430182, "grad_norm": 0.15946388244628906, "learning_rate": 9.681561003917018e-06, "loss": 0.0128, "step": 14325 }, { "epoch": 1.5531222896790982, "grad_norm": 8.95676326751709, "learning_rate": 9.68083563034963e-06, "loss": 0.0401, "step": 14326 }, { "epoch": 1.5532307025151777, "grad_norm": 2.226480722427368, "learning_rate": 9.680110256782244e-06, "loss": 0.012, "step": 14327 }, { "epoch": 1.5533391153512577, "grad_norm": 1.905281901359558, "learning_rate": 9.679384883214856e-06, "loss": 0.0217, "step": 14328 }, { "epoch": 1.5534475281873372, "grad_norm": 0.49072858691215515, "learning_rate": 9.678659509647468e-06, "loss": 0.0147, "step": 14329 }, { "epoch": 1.5535559410234172, "grad_norm": 0.9969756603240967, "learning_rate": 9.677934136080082e-06, "loss": 0.0047, "step": 14330 }, { "epoch": 1.553664353859497, "grad_norm": 2.9911415576934814, "learning_rate": 9.677208762512696e-06, "loss": 0.0304, "step": 14331 }, { "epoch": 1.5537727666955767, "grad_norm": 0.06538892537355423, "learning_rate": 9.676483388945308e-06, "loss": 0.0005, "step": 14332 }, { "epoch": 1.5538811795316565, "grad_norm": 12.223066329956055, "learning_rate": 9.67575801537792e-06, "loss": 0.0162, "step": 14333 }, { "epoch": 1.5539895923677363, "grad_norm": 10.870772361755371, "learning_rate": 9.675032641810534e-06, "loss": 0.0397, "step": 14334 }, { "epoch": 1.5540980052038162, "grad_norm": 2.8037242889404297, "learning_rate": 9.674307268243146e-06, "loss": 0.0412, "step": 14335 }, { "epoch": 1.5542064180398958, "grad_norm": 4.609831809997559, "learning_rate": 9.673581894675758e-06, "loss": 0.0514, "step": 14336 }, { "epoch": 1.5543148308759758, "grad_norm": 3.675938129425049, "learning_rate": 9.67285652110837e-06, "loss": 0.0218, "step": 14337 }, { "epoch": 1.5544232437120555, "grad_norm": 0.3455454111099243, "learning_rate": 9.672131147540984e-06, "loss": 0.0069, "step": 14338 }, { "epoch": 1.5545316565481353, "grad_norm": 0.4486558735370636, "learning_rate": 9.671405773973597e-06, "loss": 0.004, "step": 14339 }, { "epoch": 1.554640069384215, "grad_norm": 2.525952100753784, "learning_rate": 9.67068040040621e-06, "loss": 0.053, "step": 14340 }, { "epoch": 1.5547484822202948, "grad_norm": 0.5383776426315308, "learning_rate": 9.669955026838823e-06, "loss": 0.008, "step": 14341 }, { "epoch": 1.5548568950563748, "grad_norm": 2.1520462036132812, "learning_rate": 9.669229653271436e-06, "loss": 0.0341, "step": 14342 }, { "epoch": 1.5549653078924544, "grad_norm": 0.052467286586761475, "learning_rate": 9.668504279704049e-06, "loss": 0.0003, "step": 14343 }, { "epoch": 1.5550737207285343, "grad_norm": 1.6642946004867554, "learning_rate": 9.66777890613666e-06, "loss": 0.0268, "step": 14344 }, { "epoch": 1.555182133564614, "grad_norm": 2.6325864791870117, "learning_rate": 9.667053532569274e-06, "loss": 0.0036, "step": 14345 }, { "epoch": 1.5552905464006939, "grad_norm": 3.5127804279327393, "learning_rate": 9.666328159001887e-06, "loss": 0.0085, "step": 14346 }, { "epoch": 1.5553989592367736, "grad_norm": 1.7888787984848022, "learning_rate": 9.665602785434499e-06, "loss": 0.0105, "step": 14347 }, { "epoch": 1.5555073720728534, "grad_norm": 3.1201491355895996, "learning_rate": 9.664877411867113e-06, "loss": 0.0812, "step": 14348 }, { "epoch": 1.5556157849089334, "grad_norm": 0.5273770093917847, "learning_rate": 9.664152038299725e-06, "loss": 0.0051, "step": 14349 }, { "epoch": 1.555724197745013, "grad_norm": 1.1897083520889282, "learning_rate": 9.663426664732339e-06, "loss": 0.0207, "step": 14350 }, { "epoch": 1.555832610581093, "grad_norm": 6.093393325805664, "learning_rate": 9.66270129116495e-06, "loss": 0.0715, "step": 14351 }, { "epoch": 1.5559410234171724, "grad_norm": 1.895269751548767, "learning_rate": 9.661975917597563e-06, "loss": 0.0306, "step": 14352 }, { "epoch": 1.5560494362532524, "grad_norm": 6.086984634399414, "learning_rate": 9.661250544030177e-06, "loss": 0.0436, "step": 14353 }, { "epoch": 1.5561578490893322, "grad_norm": 6.178284645080566, "learning_rate": 9.660525170462789e-06, "loss": 0.0241, "step": 14354 }, { "epoch": 1.556266261925412, "grad_norm": 3.0202174186706543, "learning_rate": 9.659799796895401e-06, "loss": 0.0119, "step": 14355 }, { "epoch": 1.5563746747614917, "grad_norm": 1.9669135808944702, "learning_rate": 9.659074423328015e-06, "loss": 0.0336, "step": 14356 }, { "epoch": 1.5564830875975715, "grad_norm": 0.5082223415374756, "learning_rate": 9.658349049760629e-06, "loss": 0.0106, "step": 14357 }, { "epoch": 1.5565915004336515, "grad_norm": 0.5002356171607971, "learning_rate": 9.65762367619324e-06, "loss": 0.0179, "step": 14358 }, { "epoch": 1.556699913269731, "grad_norm": 0.3752776086330414, "learning_rate": 9.656898302625853e-06, "loss": 0.0109, "step": 14359 }, { "epoch": 1.556808326105811, "grad_norm": 0.03822239488363266, "learning_rate": 9.656172929058467e-06, "loss": 0.0005, "step": 14360 }, { "epoch": 1.5569167389418908, "grad_norm": 0.40669119358062744, "learning_rate": 9.655447555491079e-06, "loss": 0.0036, "step": 14361 }, { "epoch": 1.5570251517779705, "grad_norm": 0.4503086507320404, "learning_rate": 9.654722181923691e-06, "loss": 0.0037, "step": 14362 }, { "epoch": 1.5571335646140503, "grad_norm": 0.046380408108234406, "learning_rate": 9.653996808356303e-06, "loss": 0.0005, "step": 14363 }, { "epoch": 1.55724197745013, "grad_norm": 0.9373039603233337, "learning_rate": 9.653271434788917e-06, "loss": 0.0376, "step": 14364 }, { "epoch": 1.55735039028621, "grad_norm": 11.239120483398438, "learning_rate": 9.65254606122153e-06, "loss": 0.0187, "step": 14365 }, { "epoch": 1.5574588031222896, "grad_norm": 9.531968116760254, "learning_rate": 9.651820687654143e-06, "loss": 0.0629, "step": 14366 }, { "epoch": 1.5575672159583696, "grad_norm": 0.3495888411998749, "learning_rate": 9.651095314086755e-06, "loss": 0.002, "step": 14367 }, { "epoch": 1.557675628794449, "grad_norm": 3.728285074234009, "learning_rate": 9.650369940519369e-06, "loss": 0.041, "step": 14368 }, { "epoch": 1.557784041630529, "grad_norm": 4.281483173370361, "learning_rate": 9.649644566951981e-06, "loss": 0.0113, "step": 14369 }, { "epoch": 1.5578924544666088, "grad_norm": 1.8054176568984985, "learning_rate": 9.648919193384593e-06, "loss": 0.0327, "step": 14370 }, { "epoch": 1.5580008673026886, "grad_norm": 1.486520767211914, "learning_rate": 9.648193819817207e-06, "loss": 0.0066, "step": 14371 }, { "epoch": 1.5581092801387686, "grad_norm": 1.5271008014678955, "learning_rate": 9.64746844624982e-06, "loss": 0.0137, "step": 14372 }, { "epoch": 1.5582176929748481, "grad_norm": 1.3454235792160034, "learning_rate": 9.646743072682431e-06, "loss": 0.0061, "step": 14373 }, { "epoch": 1.5583261058109281, "grad_norm": 1.3295780420303345, "learning_rate": 9.646017699115045e-06, "loss": 0.0077, "step": 14374 }, { "epoch": 1.5584345186470077, "grad_norm": 3.1204400062561035, "learning_rate": 9.645292325547657e-06, "loss": 0.0538, "step": 14375 }, { "epoch": 1.5585429314830876, "grad_norm": 7.565263748168945, "learning_rate": 9.644566951980271e-06, "loss": 0.0342, "step": 14376 }, { "epoch": 1.5586513443191674, "grad_norm": 2.573546886444092, "learning_rate": 9.643841578412883e-06, "loss": 0.0186, "step": 14377 }, { "epoch": 1.5587597571552472, "grad_norm": 2.3434412479400635, "learning_rate": 9.643116204845496e-06, "loss": 0.0426, "step": 14378 }, { "epoch": 1.558868169991327, "grad_norm": 0.38375118374824524, "learning_rate": 9.64239083127811e-06, "loss": 0.0011, "step": 14379 }, { "epoch": 1.5589765828274067, "grad_norm": 0.2054760903120041, "learning_rate": 9.641665457710722e-06, "loss": 0.0037, "step": 14380 }, { "epoch": 1.5590849956634867, "grad_norm": 3.7460343837738037, "learning_rate": 9.640940084143334e-06, "loss": 0.032, "step": 14381 }, { "epoch": 1.5591934084995662, "grad_norm": 2.937819719314575, "learning_rate": 9.640214710575948e-06, "loss": 0.0086, "step": 14382 }, { "epoch": 1.5593018213356462, "grad_norm": 3.0120484828948975, "learning_rate": 9.639489337008561e-06, "loss": 0.0305, "step": 14383 }, { "epoch": 1.559410234171726, "grad_norm": 0.4193115532398224, "learning_rate": 9.638763963441173e-06, "loss": 0.0112, "step": 14384 }, { "epoch": 1.5595186470078057, "grad_norm": 2.5313467979431152, "learning_rate": 9.638038589873786e-06, "loss": 0.0174, "step": 14385 }, { "epoch": 1.5596270598438855, "grad_norm": 1.4166337251663208, "learning_rate": 9.6373132163064e-06, "loss": 0.0124, "step": 14386 }, { "epoch": 1.5597354726799653, "grad_norm": 2.9997663497924805, "learning_rate": 9.636587842739012e-06, "loss": 0.0098, "step": 14387 }, { "epoch": 1.5598438855160452, "grad_norm": 4.984210014343262, "learning_rate": 9.635862469171624e-06, "loss": 0.0159, "step": 14388 }, { "epoch": 1.5599522983521248, "grad_norm": 9.140443801879883, "learning_rate": 9.635137095604236e-06, "loss": 0.0108, "step": 14389 }, { "epoch": 1.5600607111882048, "grad_norm": 0.4314298927783966, "learning_rate": 9.63441172203685e-06, "loss": 0.0149, "step": 14390 }, { "epoch": 1.5601691240242843, "grad_norm": 0.9957219362258911, "learning_rate": 9.633686348469462e-06, "loss": 0.0019, "step": 14391 }, { "epoch": 1.5602775368603643, "grad_norm": 0.6558157801628113, "learning_rate": 9.632960974902076e-06, "loss": 0.0199, "step": 14392 }, { "epoch": 1.560385949696444, "grad_norm": 1.668637990951538, "learning_rate": 9.632235601334688e-06, "loss": 0.0331, "step": 14393 }, { "epoch": 1.5604943625325238, "grad_norm": 1.3094990253448486, "learning_rate": 9.631510227767302e-06, "loss": 0.0069, "step": 14394 }, { "epoch": 1.5606027753686036, "grad_norm": 1.6844195127487183, "learning_rate": 9.630784854199914e-06, "loss": 0.0301, "step": 14395 }, { "epoch": 1.5607111882046834, "grad_norm": 3.1712186336517334, "learning_rate": 9.630059480632526e-06, "loss": 0.0228, "step": 14396 }, { "epoch": 1.5608196010407633, "grad_norm": 1.0150134563446045, "learning_rate": 9.62933410706514e-06, "loss": 0.0128, "step": 14397 }, { "epoch": 1.5609280138768429, "grad_norm": 6.997344017028809, "learning_rate": 9.628608733497752e-06, "loss": 0.0268, "step": 14398 }, { "epoch": 1.5610364267129229, "grad_norm": 3.4626715183258057, "learning_rate": 9.627883359930364e-06, "loss": 0.0738, "step": 14399 }, { "epoch": 1.5611448395490026, "grad_norm": 6.126176357269287, "learning_rate": 9.627157986362978e-06, "loss": 0.0296, "step": 14400 }, { "epoch": 1.5612532523850824, "grad_norm": 1.8677048683166504, "learning_rate": 9.626432612795592e-06, "loss": 0.0274, "step": 14401 }, { "epoch": 1.5613616652211622, "grad_norm": 0.45111018419265747, "learning_rate": 9.625707239228204e-06, "loss": 0.0091, "step": 14402 }, { "epoch": 1.561470078057242, "grad_norm": 1.8629016876220703, "learning_rate": 9.624981865660816e-06, "loss": 0.0097, "step": 14403 }, { "epoch": 1.561578490893322, "grad_norm": 0.6833471059799194, "learning_rate": 9.624256492093428e-06, "loss": 0.021, "step": 14404 }, { "epoch": 1.5616869037294014, "grad_norm": 6.531404972076416, "learning_rate": 9.623531118526042e-06, "loss": 0.0254, "step": 14405 }, { "epoch": 1.5617953165654814, "grad_norm": 1.6254407167434692, "learning_rate": 9.622805744958654e-06, "loss": 0.0397, "step": 14406 }, { "epoch": 1.5619037294015612, "grad_norm": 2.51265287399292, "learning_rate": 9.622080371391266e-06, "loss": 0.0277, "step": 14407 }, { "epoch": 1.562012142237641, "grad_norm": 2.3379554748535156, "learning_rate": 9.62135499782388e-06, "loss": 0.0157, "step": 14408 }, { "epoch": 1.5621205550737207, "grad_norm": 0.32115086913108826, "learning_rate": 9.620629624256494e-06, "loss": 0.0083, "step": 14409 }, { "epoch": 1.5622289679098005, "grad_norm": 2.4604852199554443, "learning_rate": 9.619904250689106e-06, "loss": 0.0284, "step": 14410 }, { "epoch": 1.5623373807458805, "grad_norm": 1.8563295602798462, "learning_rate": 9.619178877121718e-06, "loss": 0.0313, "step": 14411 }, { "epoch": 1.56244579358196, "grad_norm": 0.865429699420929, "learning_rate": 9.618453503554332e-06, "loss": 0.0259, "step": 14412 }, { "epoch": 1.56255420641804, "grad_norm": 0.4073095917701721, "learning_rate": 9.617728129986944e-06, "loss": 0.0132, "step": 14413 }, { "epoch": 1.5626626192541195, "grad_norm": 8.673422813415527, "learning_rate": 9.617002756419556e-06, "loss": 0.0251, "step": 14414 }, { "epoch": 1.5627710320901995, "grad_norm": 0.16268165409564972, "learning_rate": 9.616277382852169e-06, "loss": 0.0015, "step": 14415 }, { "epoch": 1.5628794449262793, "grad_norm": 1.7898951768875122, "learning_rate": 9.615552009284782e-06, "loss": 0.0273, "step": 14416 }, { "epoch": 1.562987857762359, "grad_norm": 3.8337535858154297, "learning_rate": 9.614826635717395e-06, "loss": 0.0507, "step": 14417 }, { "epoch": 1.5630962705984388, "grad_norm": 0.5284180641174316, "learning_rate": 9.614101262150008e-06, "loss": 0.005, "step": 14418 }, { "epoch": 1.5632046834345186, "grad_norm": 3.4288759231567383, "learning_rate": 9.61337588858262e-06, "loss": 0.0372, "step": 14419 }, { "epoch": 1.5633130962705986, "grad_norm": 7.262107849121094, "learning_rate": 9.612650515015234e-06, "loss": 0.0577, "step": 14420 }, { "epoch": 1.563421509106678, "grad_norm": 0.9713724255561829, "learning_rate": 9.611925141447847e-06, "loss": 0.0215, "step": 14421 }, { "epoch": 1.563529921942758, "grad_norm": 9.141190528869629, "learning_rate": 9.611199767880459e-06, "loss": 0.0294, "step": 14422 }, { "epoch": 1.5636383347788378, "grad_norm": 2.326024293899536, "learning_rate": 9.610474394313072e-06, "loss": 0.0205, "step": 14423 }, { "epoch": 1.5637467476149176, "grad_norm": 1.948662281036377, "learning_rate": 9.609749020745685e-06, "loss": 0.0087, "step": 14424 }, { "epoch": 1.5638551604509974, "grad_norm": 5.558991432189941, "learning_rate": 9.609023647178297e-06, "loss": 0.0453, "step": 14425 }, { "epoch": 1.5639635732870771, "grad_norm": 6.012519359588623, "learning_rate": 9.60829827361091e-06, "loss": 0.0245, "step": 14426 }, { "epoch": 1.5640719861231571, "grad_norm": 1.4496495723724365, "learning_rate": 9.607572900043524e-06, "loss": 0.024, "step": 14427 }, { "epoch": 1.5641803989592367, "grad_norm": 1.9359664916992188, "learning_rate": 9.606847526476137e-06, "loss": 0.0148, "step": 14428 }, { "epoch": 1.5642888117953166, "grad_norm": 0.3559400737285614, "learning_rate": 9.606122152908749e-06, "loss": 0.0026, "step": 14429 }, { "epoch": 1.5643972246313964, "grad_norm": 3.880748987197876, "learning_rate": 9.605396779341361e-06, "loss": 0.0556, "step": 14430 }, { "epoch": 1.5645056374674762, "grad_norm": 9.360146522521973, "learning_rate": 9.604671405773975e-06, "loss": 0.105, "step": 14431 }, { "epoch": 1.564614050303556, "grad_norm": 2.3903348445892334, "learning_rate": 9.603946032206587e-06, "loss": 0.0232, "step": 14432 }, { "epoch": 1.5647224631396357, "grad_norm": 2.512202739715576, "learning_rate": 9.603220658639199e-06, "loss": 0.0125, "step": 14433 }, { "epoch": 1.5648308759757157, "grad_norm": 2.4260780811309814, "learning_rate": 9.602495285071813e-06, "loss": 0.0124, "step": 14434 }, { "epoch": 1.5649392888117952, "grad_norm": 4.294937610626221, "learning_rate": 9.601769911504427e-06, "loss": 0.0436, "step": 14435 }, { "epoch": 1.5650477016478752, "grad_norm": 1.8146424293518066, "learning_rate": 9.601044537937039e-06, "loss": 0.044, "step": 14436 }, { "epoch": 1.5651561144839548, "grad_norm": 0.19387856125831604, "learning_rate": 9.600319164369651e-06, "loss": 0.0013, "step": 14437 }, { "epoch": 1.5652645273200347, "grad_norm": 7.391541957855225, "learning_rate": 9.599593790802265e-06, "loss": 0.045, "step": 14438 }, { "epoch": 1.5653729401561145, "grad_norm": 0.46647122502326965, "learning_rate": 9.598868417234877e-06, "loss": 0.0041, "step": 14439 }, { "epoch": 1.5654813529921943, "grad_norm": 3.4060921669006348, "learning_rate": 9.598143043667489e-06, "loss": 0.0215, "step": 14440 }, { "epoch": 1.565589765828274, "grad_norm": 2.3473737239837646, "learning_rate": 9.597417670100101e-06, "loss": 0.0345, "step": 14441 }, { "epoch": 1.5656981786643538, "grad_norm": 0.8441020250320435, "learning_rate": 9.596692296532715e-06, "loss": 0.0043, "step": 14442 }, { "epoch": 1.5658065915004338, "grad_norm": 1.5361312627792358, "learning_rate": 9.595966922965327e-06, "loss": 0.0266, "step": 14443 }, { "epoch": 1.5659150043365133, "grad_norm": 0.12536346912384033, "learning_rate": 9.595241549397941e-06, "loss": 0.0011, "step": 14444 }, { "epoch": 1.5660234171725933, "grad_norm": 1.0528333187103271, "learning_rate": 9.594516175830553e-06, "loss": 0.0199, "step": 14445 }, { "epoch": 1.566131830008673, "grad_norm": 5.605066776275635, "learning_rate": 9.593790802263167e-06, "loss": 0.0489, "step": 14446 }, { "epoch": 1.5662402428447528, "grad_norm": 4.580202102661133, "learning_rate": 9.59306542869578e-06, "loss": 0.0255, "step": 14447 }, { "epoch": 1.5663486556808326, "grad_norm": 0.34805548191070557, "learning_rate": 9.592340055128391e-06, "loss": 0.0139, "step": 14448 }, { "epoch": 1.5664570685169124, "grad_norm": 0.6992433667182922, "learning_rate": 9.591614681561005e-06, "loss": 0.0018, "step": 14449 }, { "epoch": 1.5665654813529923, "grad_norm": 2.0498270988464355, "learning_rate": 9.590889307993617e-06, "loss": 0.0229, "step": 14450 }, { "epoch": 1.5666738941890719, "grad_norm": 1.390475869178772, "learning_rate": 9.59016393442623e-06, "loss": 0.0094, "step": 14451 }, { "epoch": 1.5667823070251519, "grad_norm": 1.3695735931396484, "learning_rate": 9.589438560858843e-06, "loss": 0.0211, "step": 14452 }, { "epoch": 1.5668907198612314, "grad_norm": 0.5494533181190491, "learning_rate": 9.588713187291457e-06, "loss": 0.003, "step": 14453 }, { "epoch": 1.5669991326973114, "grad_norm": 0.9719621539115906, "learning_rate": 9.58798781372407e-06, "loss": 0.0111, "step": 14454 }, { "epoch": 1.5671075455333912, "grad_norm": 2.507697105407715, "learning_rate": 9.587262440156681e-06, "loss": 0.0165, "step": 14455 }, { "epoch": 1.567215958369471, "grad_norm": 1.1603904962539673, "learning_rate": 9.586537066589294e-06, "loss": 0.0544, "step": 14456 }, { "epoch": 1.567324371205551, "grad_norm": 1.8870559930801392, "learning_rate": 9.585811693021907e-06, "loss": 0.0174, "step": 14457 }, { "epoch": 1.5674327840416304, "grad_norm": 7.571075916290283, "learning_rate": 9.58508631945452e-06, "loss": 0.0063, "step": 14458 }, { "epoch": 1.5675411968777104, "grad_norm": 2.811107635498047, "learning_rate": 9.584360945887132e-06, "loss": 0.019, "step": 14459 }, { "epoch": 1.56764960971379, "grad_norm": 0.7941247224807739, "learning_rate": 9.583635572319746e-06, "loss": 0.0097, "step": 14460 }, { "epoch": 1.56775802254987, "grad_norm": 0.9017319083213806, "learning_rate": 9.58291019875236e-06, "loss": 0.0052, "step": 14461 }, { "epoch": 1.5678664353859497, "grad_norm": 3.417567014694214, "learning_rate": 9.582184825184971e-06, "loss": 0.0535, "step": 14462 }, { "epoch": 1.5679748482220295, "grad_norm": 3.607456684112549, "learning_rate": 9.581459451617584e-06, "loss": 0.0217, "step": 14463 }, { "epoch": 1.5680832610581092, "grad_norm": 5.732057094573975, "learning_rate": 9.580734078050197e-06, "loss": 0.0648, "step": 14464 }, { "epoch": 1.568191673894189, "grad_norm": 4.65419864654541, "learning_rate": 9.58000870448281e-06, "loss": 0.0261, "step": 14465 }, { "epoch": 1.568300086730269, "grad_norm": 0.9011474847793579, "learning_rate": 9.579283330915422e-06, "loss": 0.0029, "step": 14466 }, { "epoch": 1.5684084995663485, "grad_norm": 0.22856630384922028, "learning_rate": 9.578557957348034e-06, "loss": 0.004, "step": 14467 }, { "epoch": 1.5685169124024285, "grad_norm": 2.393688201904297, "learning_rate": 9.577832583780648e-06, "loss": 0.0708, "step": 14468 }, { "epoch": 1.5686253252385083, "grad_norm": 1.5591691732406616, "learning_rate": 9.57710721021326e-06, "loss": 0.027, "step": 14469 }, { "epoch": 1.568733738074588, "grad_norm": 4.731564998626709, "learning_rate": 9.576381836645874e-06, "loss": 0.0415, "step": 14470 }, { "epoch": 1.5688421509106678, "grad_norm": 1.2748302221298218, "learning_rate": 9.575656463078486e-06, "loss": 0.0196, "step": 14471 }, { "epoch": 1.5689505637467476, "grad_norm": 2.45306134223938, "learning_rate": 9.5749310895111e-06, "loss": 0.0621, "step": 14472 }, { "epoch": 1.5690589765828276, "grad_norm": 2.8850326538085938, "learning_rate": 9.574205715943712e-06, "loss": 0.0553, "step": 14473 }, { "epoch": 1.569167389418907, "grad_norm": 1.4818485975265503, "learning_rate": 9.573480342376324e-06, "loss": 0.03, "step": 14474 }, { "epoch": 1.569275802254987, "grad_norm": 3.913912296295166, "learning_rate": 9.572754968808938e-06, "loss": 0.0616, "step": 14475 }, { "epoch": 1.5693842150910666, "grad_norm": 0.44865795969963074, "learning_rate": 9.57202959524155e-06, "loss": 0.0143, "step": 14476 }, { "epoch": 1.5694926279271466, "grad_norm": 0.9207632541656494, "learning_rate": 9.571304221674162e-06, "loss": 0.0054, "step": 14477 }, { "epoch": 1.5696010407632264, "grad_norm": 1.2503573894500732, "learning_rate": 9.570578848106776e-06, "loss": 0.0276, "step": 14478 }, { "epoch": 1.5697094535993061, "grad_norm": 3.8999950885772705, "learning_rate": 9.56985347453939e-06, "loss": 0.033, "step": 14479 }, { "epoch": 1.569817866435386, "grad_norm": 0.35703331232070923, "learning_rate": 9.569128100972002e-06, "loss": 0.0033, "step": 14480 }, { "epoch": 1.5699262792714657, "grad_norm": 2.7710297107696533, "learning_rate": 9.568402727404614e-06, "loss": 0.036, "step": 14481 }, { "epoch": 1.5700346921075456, "grad_norm": 0.5688058733940125, "learning_rate": 9.567677353837226e-06, "loss": 0.0168, "step": 14482 }, { "epoch": 1.5701431049436252, "grad_norm": 1.036841630935669, "learning_rate": 9.56695198026984e-06, "loss": 0.0403, "step": 14483 }, { "epoch": 1.5702515177797052, "grad_norm": 1.326138973236084, "learning_rate": 9.566226606702452e-06, "loss": 0.0167, "step": 14484 }, { "epoch": 1.570359930615785, "grad_norm": 1.2180185317993164, "learning_rate": 9.565501233135064e-06, "loss": 0.0031, "step": 14485 }, { "epoch": 1.5704683434518647, "grad_norm": 3.1521315574645996, "learning_rate": 9.564775859567678e-06, "loss": 0.0941, "step": 14486 }, { "epoch": 1.5705767562879445, "grad_norm": 0.6919737458229065, "learning_rate": 9.564050486000292e-06, "loss": 0.0095, "step": 14487 }, { "epoch": 1.5706851691240242, "grad_norm": 1.3647950887680054, "learning_rate": 9.563325112432904e-06, "loss": 0.0495, "step": 14488 }, { "epoch": 1.5707935819601042, "grad_norm": 0.4032716751098633, "learning_rate": 9.562599738865516e-06, "loss": 0.0227, "step": 14489 }, { "epoch": 1.5709019947961838, "grad_norm": 0.08634667098522186, "learning_rate": 9.56187436529813e-06, "loss": 0.0011, "step": 14490 }, { "epoch": 1.5710104076322637, "grad_norm": 2.5289359092712402, "learning_rate": 9.561148991730742e-06, "loss": 0.0177, "step": 14491 }, { "epoch": 1.5711188204683435, "grad_norm": 3.5915727615356445, "learning_rate": 9.560423618163354e-06, "loss": 0.0113, "step": 14492 }, { "epoch": 1.5712272333044233, "grad_norm": 5.259051322937012, "learning_rate": 9.559698244595967e-06, "loss": 0.0382, "step": 14493 }, { "epoch": 1.571335646140503, "grad_norm": 1.726240634918213, "learning_rate": 9.55897287102858e-06, "loss": 0.0237, "step": 14494 }, { "epoch": 1.5714440589765828, "grad_norm": 7.499872207641602, "learning_rate": 9.558247497461194e-06, "loss": 0.0761, "step": 14495 }, { "epoch": 1.5715524718126628, "grad_norm": 0.3200010359287262, "learning_rate": 9.557522123893806e-06, "loss": 0.0071, "step": 14496 }, { "epoch": 1.5716608846487423, "grad_norm": 1.6739119291305542, "learning_rate": 9.556796750326419e-06, "loss": 0.0234, "step": 14497 }, { "epoch": 1.5717692974848223, "grad_norm": 0.17716354131698608, "learning_rate": 9.556071376759032e-06, "loss": 0.0018, "step": 14498 }, { "epoch": 1.5718777103209018, "grad_norm": 2.4532177448272705, "learning_rate": 9.555346003191645e-06, "loss": 0.0404, "step": 14499 }, { "epoch": 1.5719861231569818, "grad_norm": 4.116982460021973, "learning_rate": 9.554620629624257e-06, "loss": 0.0407, "step": 14500 }, { "epoch": 1.5720945359930616, "grad_norm": 1.2393110990524292, "learning_rate": 9.55389525605687e-06, "loss": 0.0301, "step": 14501 }, { "epoch": 1.5722029488291414, "grad_norm": 3.013037919998169, "learning_rate": 9.553169882489483e-06, "loss": 0.0186, "step": 14502 }, { "epoch": 1.5723113616652211, "grad_norm": 0.42680954933166504, "learning_rate": 9.552444508922095e-06, "loss": 0.0272, "step": 14503 }, { "epoch": 1.5724197745013009, "grad_norm": 0.8253358602523804, "learning_rate": 9.551719135354709e-06, "loss": 0.0134, "step": 14504 }, { "epoch": 1.5725281873373809, "grad_norm": 2.7805817127227783, "learning_rate": 9.550993761787322e-06, "loss": 0.0118, "step": 14505 }, { "epoch": 1.5726366001734604, "grad_norm": 2.439757823944092, "learning_rate": 9.550268388219935e-06, "loss": 0.0159, "step": 14506 }, { "epoch": 1.5727450130095404, "grad_norm": 0.08344579488039017, "learning_rate": 9.549543014652547e-06, "loss": 0.0008, "step": 14507 }, { "epoch": 1.5728534258456202, "grad_norm": 4.049241065979004, "learning_rate": 9.548817641085159e-06, "loss": 0.0331, "step": 14508 }, { "epoch": 1.5729618386817, "grad_norm": 3.235241413116455, "learning_rate": 9.548092267517773e-06, "loss": 0.0185, "step": 14509 }, { "epoch": 1.5730702515177797, "grad_norm": 4.236340045928955, "learning_rate": 9.547366893950385e-06, "loss": 0.0755, "step": 14510 }, { "epoch": 1.5731786643538594, "grad_norm": 3.9747085571289062, "learning_rate": 9.546641520382997e-06, "loss": 0.0393, "step": 14511 }, { "epoch": 1.5732870771899394, "grad_norm": 0.04081229865550995, "learning_rate": 9.545916146815611e-06, "loss": 0.0004, "step": 14512 }, { "epoch": 1.573395490026019, "grad_norm": 0.6658350229263306, "learning_rate": 9.545190773248225e-06, "loss": 0.0084, "step": 14513 }, { "epoch": 1.573503902862099, "grad_norm": 6.339938163757324, "learning_rate": 9.544465399680837e-06, "loss": 0.0232, "step": 14514 }, { "epoch": 1.5736123156981785, "grad_norm": 4.399850368499756, "learning_rate": 9.543740026113449e-06, "loss": 0.0319, "step": 14515 }, { "epoch": 1.5737207285342585, "grad_norm": 3.821119785308838, "learning_rate": 9.543014652546063e-06, "loss": 0.0248, "step": 14516 }, { "epoch": 1.5738291413703382, "grad_norm": 0.9864227175712585, "learning_rate": 9.542289278978675e-06, "loss": 0.027, "step": 14517 }, { "epoch": 1.573937554206418, "grad_norm": 1.943791151046753, "learning_rate": 9.541563905411287e-06, "loss": 0.02, "step": 14518 }, { "epoch": 1.574045967042498, "grad_norm": 1.5147291421890259, "learning_rate": 9.5408385318439e-06, "loss": 0.005, "step": 14519 }, { "epoch": 1.5741543798785775, "grad_norm": 1.6907765865325928, "learning_rate": 9.540113158276513e-06, "loss": 0.0033, "step": 14520 }, { "epoch": 1.5742627927146575, "grad_norm": 1.9803903102874756, "learning_rate": 9.539387784709127e-06, "loss": 0.0098, "step": 14521 }, { "epoch": 1.574371205550737, "grad_norm": 0.38131222128868103, "learning_rate": 9.538662411141739e-06, "loss": 0.0077, "step": 14522 }, { "epoch": 1.574479618386817, "grad_norm": 2.820483446121216, "learning_rate": 9.537937037574351e-06, "loss": 0.0033, "step": 14523 }, { "epoch": 1.5745880312228968, "grad_norm": 3.294663667678833, "learning_rate": 9.537211664006965e-06, "loss": 0.0315, "step": 14524 }, { "epoch": 1.5746964440589766, "grad_norm": 7.58444881439209, "learning_rate": 9.536486290439577e-06, "loss": 0.0344, "step": 14525 }, { "epoch": 1.5748048568950563, "grad_norm": 0.7655863761901855, "learning_rate": 9.53576091687219e-06, "loss": 0.0166, "step": 14526 }, { "epoch": 1.574913269731136, "grad_norm": 0.8803068399429321, "learning_rate": 9.535035543304803e-06, "loss": 0.0172, "step": 14527 }, { "epoch": 1.575021682567216, "grad_norm": 1.3548486232757568, "learning_rate": 9.534310169737415e-06, "loss": 0.0167, "step": 14528 }, { "epoch": 1.5751300954032956, "grad_norm": 0.8981338739395142, "learning_rate": 9.533584796170027e-06, "loss": 0.0245, "step": 14529 }, { "epoch": 1.5752385082393756, "grad_norm": 0.8565184473991394, "learning_rate": 9.532859422602641e-06, "loss": 0.0059, "step": 14530 }, { "epoch": 1.5753469210754554, "grad_norm": 1.5396161079406738, "learning_rate": 9.532134049035255e-06, "loss": 0.0656, "step": 14531 }, { "epoch": 1.5754553339115351, "grad_norm": 0.5127642750740051, "learning_rate": 9.531408675467867e-06, "loss": 0.0014, "step": 14532 }, { "epoch": 1.575563746747615, "grad_norm": 0.49219006299972534, "learning_rate": 9.53068330190048e-06, "loss": 0.009, "step": 14533 }, { "epoch": 1.5756721595836947, "grad_norm": 2.4565865993499756, "learning_rate": 9.529957928333092e-06, "loss": 0.0161, "step": 14534 }, { "epoch": 1.5757805724197746, "grad_norm": 0.44288164377212524, "learning_rate": 9.529232554765705e-06, "loss": 0.0129, "step": 14535 }, { "epoch": 1.5758889852558542, "grad_norm": 2.4127678871154785, "learning_rate": 9.528507181198318e-06, "loss": 0.0435, "step": 14536 }, { "epoch": 1.5759973980919342, "grad_norm": 1.7393927574157715, "learning_rate": 9.52778180763093e-06, "loss": 0.0228, "step": 14537 }, { "epoch": 1.5761058109280137, "grad_norm": 0.13761332631111145, "learning_rate": 9.527056434063544e-06, "loss": 0.0012, "step": 14538 }, { "epoch": 1.5762142237640937, "grad_norm": 0.8810299634933472, "learning_rate": 9.526331060496157e-06, "loss": 0.0122, "step": 14539 }, { "epoch": 1.5763226366001735, "grad_norm": 0.6688612699508667, "learning_rate": 9.52560568692877e-06, "loss": 0.0159, "step": 14540 }, { "epoch": 1.5764310494362532, "grad_norm": 1.1403052806854248, "learning_rate": 9.524880313361382e-06, "loss": 0.024, "step": 14541 }, { "epoch": 1.576539462272333, "grad_norm": 3.1514973640441895, "learning_rate": 9.524154939793995e-06, "loss": 0.0316, "step": 14542 }, { "epoch": 1.5766478751084128, "grad_norm": 5.78656005859375, "learning_rate": 9.523429566226608e-06, "loss": 0.0156, "step": 14543 }, { "epoch": 1.5767562879444927, "grad_norm": 5.953794479370117, "learning_rate": 9.52270419265922e-06, "loss": 0.0338, "step": 14544 }, { "epoch": 1.5768647007805723, "grad_norm": 2.797720432281494, "learning_rate": 9.521978819091832e-06, "loss": 0.0211, "step": 14545 }, { "epoch": 1.5769731136166523, "grad_norm": 1.8436506986618042, "learning_rate": 9.521253445524446e-06, "loss": 0.0207, "step": 14546 }, { "epoch": 1.577081526452732, "grad_norm": 0.7425704598426819, "learning_rate": 9.52052807195706e-06, "loss": 0.0042, "step": 14547 }, { "epoch": 1.5771899392888118, "grad_norm": 2.3111653327941895, "learning_rate": 9.519802698389672e-06, "loss": 0.0532, "step": 14548 }, { "epoch": 1.5772983521248916, "grad_norm": 1.1334377527236938, "learning_rate": 9.519077324822284e-06, "loss": 0.0112, "step": 14549 }, { "epoch": 1.5774067649609713, "grad_norm": 0.13657809793949127, "learning_rate": 9.518351951254898e-06, "loss": 0.001, "step": 14550 }, { "epoch": 1.5775151777970513, "grad_norm": 2.588122606277466, "learning_rate": 9.51762657768751e-06, "loss": 0.0498, "step": 14551 }, { "epoch": 1.5776235906331308, "grad_norm": 1.1076468229293823, "learning_rate": 9.516901204120122e-06, "loss": 0.0296, "step": 14552 }, { "epoch": 1.5777320034692108, "grad_norm": 1.6539288759231567, "learning_rate": 9.516175830552736e-06, "loss": 0.0468, "step": 14553 }, { "epoch": 1.5778404163052906, "grad_norm": 7.410676002502441, "learning_rate": 9.515450456985348e-06, "loss": 0.0192, "step": 14554 }, { "epoch": 1.5779488291413704, "grad_norm": 1.831323266029358, "learning_rate": 9.51472508341796e-06, "loss": 0.0306, "step": 14555 }, { "epoch": 1.5780572419774501, "grad_norm": 8.190498352050781, "learning_rate": 9.513999709850574e-06, "loss": 0.045, "step": 14556 }, { "epoch": 1.5781656548135299, "grad_norm": 0.4145088791847229, "learning_rate": 9.513274336283188e-06, "loss": 0.0021, "step": 14557 }, { "epoch": 1.5782740676496099, "grad_norm": 0.7190539240837097, "learning_rate": 9.5125489627158e-06, "loss": 0.007, "step": 14558 }, { "epoch": 1.5783824804856894, "grad_norm": 8.521746635437012, "learning_rate": 9.511823589148412e-06, "loss": 0.0092, "step": 14559 }, { "epoch": 1.5784908933217694, "grad_norm": 0.4900094270706177, "learning_rate": 9.511098215581024e-06, "loss": 0.0093, "step": 14560 }, { "epoch": 1.578599306157849, "grad_norm": 0.49343541264533997, "learning_rate": 9.510372842013638e-06, "loss": 0.007, "step": 14561 }, { "epoch": 1.578707718993929, "grad_norm": 2.756122589111328, "learning_rate": 9.50964746844625e-06, "loss": 0.0228, "step": 14562 }, { "epoch": 1.5788161318300087, "grad_norm": 0.44472408294677734, "learning_rate": 9.508922094878862e-06, "loss": 0.0103, "step": 14563 }, { "epoch": 1.5789245446660884, "grad_norm": 0.6349124908447266, "learning_rate": 9.508196721311476e-06, "loss": 0.0079, "step": 14564 }, { "epoch": 1.5790329575021682, "grad_norm": 4.049480438232422, "learning_rate": 9.50747134774409e-06, "loss": 0.0584, "step": 14565 }, { "epoch": 1.579141370338248, "grad_norm": 0.4593397378921509, "learning_rate": 9.506745974176702e-06, "loss": 0.0111, "step": 14566 }, { "epoch": 1.579249783174328, "grad_norm": 1.1658051013946533, "learning_rate": 9.506020600609314e-06, "loss": 0.006, "step": 14567 }, { "epoch": 1.5793581960104075, "grad_norm": 0.5690330266952515, "learning_rate": 9.505295227041928e-06, "loss": 0.0072, "step": 14568 }, { "epoch": 1.5794666088464875, "grad_norm": 1.4250177145004272, "learning_rate": 9.50456985347454e-06, "loss": 0.0121, "step": 14569 }, { "epoch": 1.5795750216825672, "grad_norm": 0.22954696416854858, "learning_rate": 9.503844479907152e-06, "loss": 0.0048, "step": 14570 }, { "epoch": 1.579683434518647, "grad_norm": 0.9464309215545654, "learning_rate": 9.503119106339765e-06, "loss": 0.0218, "step": 14571 }, { "epoch": 1.5797918473547268, "grad_norm": 0.5286070108413696, "learning_rate": 9.502393732772378e-06, "loss": 0.0174, "step": 14572 }, { "epoch": 1.5799002601908065, "grad_norm": 4.2034149169921875, "learning_rate": 9.501668359204992e-06, "loss": 0.0561, "step": 14573 }, { "epoch": 1.5800086730268865, "grad_norm": 0.41480645537376404, "learning_rate": 9.500942985637604e-06, "loss": 0.0026, "step": 14574 }, { "epoch": 1.580117085862966, "grad_norm": 0.2727297246456146, "learning_rate": 9.500217612070217e-06, "loss": 0.0019, "step": 14575 }, { "epoch": 1.580225498699046, "grad_norm": 8.236145973205566, "learning_rate": 9.49949223850283e-06, "loss": 0.0193, "step": 14576 }, { "epoch": 1.5803339115351258, "grad_norm": 4.6479315757751465, "learning_rate": 9.498766864935443e-06, "loss": 0.023, "step": 14577 }, { "epoch": 1.5804423243712056, "grad_norm": 1.924680233001709, "learning_rate": 9.498041491368055e-06, "loss": 0.0136, "step": 14578 }, { "epoch": 1.5805507372072853, "grad_norm": 9.918758392333984, "learning_rate": 9.497316117800669e-06, "loss": 0.0207, "step": 14579 }, { "epoch": 1.580659150043365, "grad_norm": 0.37921732664108276, "learning_rate": 9.49659074423328e-06, "loss": 0.002, "step": 14580 }, { "epoch": 1.580767562879445, "grad_norm": 5.377986907958984, "learning_rate": 9.495865370665893e-06, "loss": 0.0501, "step": 14581 }, { "epoch": 1.5808759757155246, "grad_norm": 1.400795817375183, "learning_rate": 9.495139997098507e-06, "loss": 0.0055, "step": 14582 }, { "epoch": 1.5809843885516046, "grad_norm": 0.15175829827785492, "learning_rate": 9.49441462353112e-06, "loss": 0.0014, "step": 14583 }, { "epoch": 1.5810928013876842, "grad_norm": 0.31559091806411743, "learning_rate": 9.493689249963733e-06, "loss": 0.0018, "step": 14584 }, { "epoch": 1.5812012142237641, "grad_norm": 1.0987131595611572, "learning_rate": 9.492963876396345e-06, "loss": 0.0159, "step": 14585 }, { "epoch": 1.581309627059844, "grad_norm": 0.6174569725990295, "learning_rate": 9.492238502828957e-06, "loss": 0.007, "step": 14586 }, { "epoch": 1.5814180398959237, "grad_norm": 1.9524565935134888, "learning_rate": 9.49151312926157e-06, "loss": 0.0365, "step": 14587 }, { "epoch": 1.5815264527320034, "grad_norm": 0.8155305981636047, "learning_rate": 9.490787755694183e-06, "loss": 0.017, "step": 14588 }, { "epoch": 1.5816348655680832, "grad_norm": 0.6494739055633545, "learning_rate": 9.490062382126795e-06, "loss": 0.0213, "step": 14589 }, { "epoch": 1.5817432784041632, "grad_norm": 0.33598265051841736, "learning_rate": 9.489337008559409e-06, "loss": 0.0078, "step": 14590 }, { "epoch": 1.5818516912402427, "grad_norm": 0.0784834697842598, "learning_rate": 9.488611634992023e-06, "loss": 0.0006, "step": 14591 }, { "epoch": 1.5819601040763227, "grad_norm": 1.2912390232086182, "learning_rate": 9.487886261424635e-06, "loss": 0.004, "step": 14592 }, { "epoch": 1.5820685169124025, "grad_norm": 4.778634071350098, "learning_rate": 9.487160887857247e-06, "loss": 0.03, "step": 14593 }, { "epoch": 1.5821769297484822, "grad_norm": 1.7381937503814697, "learning_rate": 9.48643551428986e-06, "loss": 0.0124, "step": 14594 }, { "epoch": 1.582285342584562, "grad_norm": 0.4573255479335785, "learning_rate": 9.485710140722473e-06, "loss": 0.0127, "step": 14595 }, { "epoch": 1.5823937554206418, "grad_norm": 3.567411422729492, "learning_rate": 9.484984767155085e-06, "loss": 0.0507, "step": 14596 }, { "epoch": 1.5825021682567217, "grad_norm": 2.6444268226623535, "learning_rate": 9.484259393587697e-06, "loss": 0.0092, "step": 14597 }, { "epoch": 1.5826105810928013, "grad_norm": 3.102569818496704, "learning_rate": 9.483534020020311e-06, "loss": 0.0452, "step": 14598 }, { "epoch": 1.5827189939288813, "grad_norm": 0.7871925234794617, "learning_rate": 9.482808646452925e-06, "loss": 0.0051, "step": 14599 }, { "epoch": 1.5828274067649608, "grad_norm": 0.5929619669914246, "learning_rate": 9.482083272885537e-06, "loss": 0.0125, "step": 14600 }, { "epoch": 1.5829358196010408, "grad_norm": 4.314620494842529, "learning_rate": 9.48135789931815e-06, "loss": 0.0267, "step": 14601 }, { "epoch": 1.5830442324371206, "grad_norm": 3.844881057739258, "learning_rate": 9.480632525750763e-06, "loss": 0.0361, "step": 14602 }, { "epoch": 1.5831526452732003, "grad_norm": 2.1229617595672607, "learning_rate": 9.479907152183375e-06, "loss": 0.027, "step": 14603 }, { "epoch": 1.58326105810928, "grad_norm": 0.03147972747683525, "learning_rate": 9.479181778615987e-06, "loss": 0.0003, "step": 14604 }, { "epoch": 1.5833694709453598, "grad_norm": 0.7003017663955688, "learning_rate": 9.478456405048601e-06, "loss": 0.0092, "step": 14605 }, { "epoch": 1.5834778837814398, "grad_norm": 2.9617488384246826, "learning_rate": 9.477731031481213e-06, "loss": 0.0278, "step": 14606 }, { "epoch": 1.5835862966175194, "grad_norm": 1.386372447013855, "learning_rate": 9.477005657913825e-06, "loss": 0.017, "step": 14607 }, { "epoch": 1.5836947094535994, "grad_norm": 0.604851245880127, "learning_rate": 9.47628028434644e-06, "loss": 0.0094, "step": 14608 }, { "epoch": 1.5838031222896791, "grad_norm": 0.3562578558921814, "learning_rate": 9.475554910779053e-06, "loss": 0.0132, "step": 14609 }, { "epoch": 1.5839115351257589, "grad_norm": 0.3864821195602417, "learning_rate": 9.474829537211665e-06, "loss": 0.0085, "step": 14610 }, { "epoch": 1.5840199479618386, "grad_norm": 3.1316275596618652, "learning_rate": 9.474104163644277e-06, "loss": 0.0464, "step": 14611 }, { "epoch": 1.5841283607979184, "grad_norm": 0.1274021714925766, "learning_rate": 9.47337879007689e-06, "loss": 0.0003, "step": 14612 }, { "epoch": 1.5842367736339984, "grad_norm": 0.9487230181694031, "learning_rate": 9.472653416509503e-06, "loss": 0.0056, "step": 14613 }, { "epoch": 1.584345186470078, "grad_norm": 0.81511390209198, "learning_rate": 9.471928042942116e-06, "loss": 0.0098, "step": 14614 }, { "epoch": 1.584453599306158, "grad_norm": 0.23495806753635406, "learning_rate": 9.471202669374728e-06, "loss": 0.0035, "step": 14615 }, { "epoch": 1.5845620121422377, "grad_norm": 0.2712582051753998, "learning_rate": 9.470477295807342e-06, "loss": 0.0025, "step": 14616 }, { "epoch": 1.5846704249783174, "grad_norm": 1.6175436973571777, "learning_rate": 9.469751922239955e-06, "loss": 0.004, "step": 14617 }, { "epoch": 1.5847788378143972, "grad_norm": 2.4163379669189453, "learning_rate": 9.469026548672568e-06, "loss": 0.015, "step": 14618 }, { "epoch": 1.584887250650477, "grad_norm": 0.42019978165626526, "learning_rate": 9.46830117510518e-06, "loss": 0.0013, "step": 14619 }, { "epoch": 1.584995663486557, "grad_norm": 5.793758869171143, "learning_rate": 9.467575801537793e-06, "loss": 0.0232, "step": 14620 }, { "epoch": 1.5851040763226365, "grad_norm": 0.9039399027824402, "learning_rate": 9.466850427970406e-06, "loss": 0.0159, "step": 14621 }, { "epoch": 1.5852124891587165, "grad_norm": 1.2673976421356201, "learning_rate": 9.466125054403018e-06, "loss": 0.0034, "step": 14622 }, { "epoch": 1.585320901994796, "grad_norm": 8.454240798950195, "learning_rate": 9.465399680835632e-06, "loss": 0.0306, "step": 14623 }, { "epoch": 1.585429314830876, "grad_norm": 0.5225427150726318, "learning_rate": 9.464674307268244e-06, "loss": 0.0079, "step": 14624 }, { "epoch": 1.5855377276669558, "grad_norm": 0.5392827391624451, "learning_rate": 9.463948933700858e-06, "loss": 0.0088, "step": 14625 }, { "epoch": 1.5856461405030355, "grad_norm": 0.619300901889801, "learning_rate": 9.46322356013347e-06, "loss": 0.0244, "step": 14626 }, { "epoch": 1.5857545533391153, "grad_norm": 1.9816216230392456, "learning_rate": 9.462498186566082e-06, "loss": 0.0171, "step": 14627 }, { "epoch": 1.585862966175195, "grad_norm": 1.5422192811965942, "learning_rate": 9.461772812998696e-06, "loss": 0.0246, "step": 14628 }, { "epoch": 1.585971379011275, "grad_norm": 1.512855887413025, "learning_rate": 9.461047439431308e-06, "loss": 0.0115, "step": 14629 }, { "epoch": 1.5860797918473546, "grad_norm": 2.5192830562591553, "learning_rate": 9.46032206586392e-06, "loss": 0.0248, "step": 14630 }, { "epoch": 1.5861882046834346, "grad_norm": 1.1893787384033203, "learning_rate": 9.459596692296534e-06, "loss": 0.0088, "step": 14631 }, { "epoch": 1.5862966175195143, "grad_norm": 1.701856255531311, "learning_rate": 9.458871318729146e-06, "loss": 0.0017, "step": 14632 }, { "epoch": 1.586405030355594, "grad_norm": 6.5653557777404785, "learning_rate": 9.458145945161758e-06, "loss": 0.0247, "step": 14633 }, { "epoch": 1.5865134431916739, "grad_norm": 4.102325439453125, "learning_rate": 9.457420571594372e-06, "loss": 0.0286, "step": 14634 }, { "epoch": 1.5866218560277536, "grad_norm": 0.6257241368293762, "learning_rate": 9.456695198026986e-06, "loss": 0.026, "step": 14635 }, { "epoch": 1.5867302688638336, "grad_norm": 2.0816023349761963, "learning_rate": 9.455969824459598e-06, "loss": 0.0179, "step": 14636 }, { "epoch": 1.5868386816999132, "grad_norm": 0.8496229648590088, "learning_rate": 9.45524445089221e-06, "loss": 0.0165, "step": 14637 }, { "epoch": 1.5869470945359931, "grad_norm": 1.1070655584335327, "learning_rate": 9.454519077324822e-06, "loss": 0.025, "step": 14638 }, { "epoch": 1.587055507372073, "grad_norm": 1.0748063325881958, "learning_rate": 9.453793703757436e-06, "loss": 0.009, "step": 14639 }, { "epoch": 1.5871639202081527, "grad_norm": 8.370574951171875, "learning_rate": 9.453068330190048e-06, "loss": 0.0524, "step": 14640 }, { "epoch": 1.5872723330442324, "grad_norm": 2.2629141807556152, "learning_rate": 9.45234295662266e-06, "loss": 0.0061, "step": 14641 }, { "epoch": 1.5873807458803122, "grad_norm": 0.30486345291137695, "learning_rate": 9.451617583055274e-06, "loss": 0.0052, "step": 14642 }, { "epoch": 1.5874891587163922, "grad_norm": 5.344525337219238, "learning_rate": 9.450892209487888e-06, "loss": 0.0224, "step": 14643 }, { "epoch": 1.5875975715524717, "grad_norm": 0.1459997296333313, "learning_rate": 9.4501668359205e-06, "loss": 0.0012, "step": 14644 }, { "epoch": 1.5877059843885517, "grad_norm": 6.333653926849365, "learning_rate": 9.449441462353112e-06, "loss": 0.0155, "step": 14645 }, { "epoch": 1.5878143972246312, "grad_norm": 0.43257853388786316, "learning_rate": 9.448716088785726e-06, "loss": 0.0075, "step": 14646 }, { "epoch": 1.5879228100607112, "grad_norm": 0.28230223059654236, "learning_rate": 9.447990715218338e-06, "loss": 0.0059, "step": 14647 }, { "epoch": 1.588031222896791, "grad_norm": 1.217248558998108, "learning_rate": 9.44726534165095e-06, "loss": 0.0092, "step": 14648 }, { "epoch": 1.5881396357328708, "grad_norm": 0.6110410094261169, "learning_rate": 9.446539968083564e-06, "loss": 0.0179, "step": 14649 }, { "epoch": 1.5882480485689505, "grad_norm": 2.321521520614624, "learning_rate": 9.445814594516176e-06, "loss": 0.0049, "step": 14650 }, { "epoch": 1.5883564614050303, "grad_norm": 0.09963274002075195, "learning_rate": 9.44508922094879e-06, "loss": 0.0007, "step": 14651 }, { "epoch": 1.5884648742411103, "grad_norm": 3.1964099407196045, "learning_rate": 9.444363847381402e-06, "loss": 0.0178, "step": 14652 }, { "epoch": 1.5885732870771898, "grad_norm": 9.447402000427246, "learning_rate": 9.443638473814015e-06, "loss": 0.0389, "step": 14653 }, { "epoch": 1.5886816999132698, "grad_norm": 0.28393542766571045, "learning_rate": 9.442913100246628e-06, "loss": 0.0059, "step": 14654 }, { "epoch": 1.5887901127493496, "grad_norm": 0.47949540615081787, "learning_rate": 9.44218772667924e-06, "loss": 0.0075, "step": 14655 }, { "epoch": 1.5888985255854293, "grad_norm": 1.2971712350845337, "learning_rate": 9.441462353111853e-06, "loss": 0.0239, "step": 14656 }, { "epoch": 1.589006938421509, "grad_norm": 7.173391819000244, "learning_rate": 9.440736979544467e-06, "loss": 0.0313, "step": 14657 }, { "epoch": 1.5891153512575888, "grad_norm": 0.47732335329055786, "learning_rate": 9.440011605977079e-06, "loss": 0.0016, "step": 14658 }, { "epoch": 1.5892237640936688, "grad_norm": 4.7624287605285645, "learning_rate": 9.43928623240969e-06, "loss": 0.0675, "step": 14659 }, { "epoch": 1.5893321769297484, "grad_norm": 2.0608937740325928, "learning_rate": 9.438560858842305e-06, "loss": 0.0125, "step": 14660 }, { "epoch": 1.5894405897658284, "grad_norm": 7.0090107917785645, "learning_rate": 9.437835485274918e-06, "loss": 0.0128, "step": 14661 }, { "epoch": 1.589549002601908, "grad_norm": 2.9198169708251953, "learning_rate": 9.43711011170753e-06, "loss": 0.0091, "step": 14662 }, { "epoch": 1.5896574154379879, "grad_norm": 0.6489585638046265, "learning_rate": 9.436384738140143e-06, "loss": 0.0193, "step": 14663 }, { "epoch": 1.5897658282740676, "grad_norm": 0.06057659164071083, "learning_rate": 9.435659364572755e-06, "loss": 0.0004, "step": 14664 }, { "epoch": 1.5898742411101474, "grad_norm": 0.06687064468860626, "learning_rate": 9.434933991005369e-06, "loss": 0.0006, "step": 14665 }, { "epoch": 1.5899826539462274, "grad_norm": 0.33099010586738586, "learning_rate": 9.434208617437981e-06, "loss": 0.0047, "step": 14666 }, { "epoch": 1.590091066782307, "grad_norm": 0.26981624960899353, "learning_rate": 9.433483243870593e-06, "loss": 0.0017, "step": 14667 }, { "epoch": 1.590199479618387, "grad_norm": 5.538473129272461, "learning_rate": 9.432757870303207e-06, "loss": 0.0366, "step": 14668 }, { "epoch": 1.5903078924544665, "grad_norm": 0.3727226257324219, "learning_rate": 9.43203249673582e-06, "loss": 0.0082, "step": 14669 }, { "epoch": 1.5904163052905465, "grad_norm": 0.8384011387825012, "learning_rate": 9.431307123168433e-06, "loss": 0.0037, "step": 14670 }, { "epoch": 1.5905247181266262, "grad_norm": 9.904664039611816, "learning_rate": 9.430581749601045e-06, "loss": 0.0403, "step": 14671 }, { "epoch": 1.590633130962706, "grad_norm": 11.626673698425293, "learning_rate": 9.429856376033659e-06, "loss": 0.0124, "step": 14672 }, { "epoch": 1.5907415437987857, "grad_norm": 4.730937480926514, "learning_rate": 9.429131002466271e-06, "loss": 0.0298, "step": 14673 }, { "epoch": 1.5908499566348655, "grad_norm": 0.5562030673027039, "learning_rate": 9.428405628898883e-06, "loss": 0.0136, "step": 14674 }, { "epoch": 1.5909583694709455, "grad_norm": 8.71495532989502, "learning_rate": 9.427680255331497e-06, "loss": 0.0086, "step": 14675 }, { "epoch": 1.591066782307025, "grad_norm": 0.6380831599235535, "learning_rate": 9.426954881764109e-06, "loss": 0.012, "step": 14676 }, { "epoch": 1.591175195143105, "grad_norm": 4.135817050933838, "learning_rate": 9.426229508196723e-06, "loss": 0.0158, "step": 14677 }, { "epoch": 1.5912836079791848, "grad_norm": 2.8032119274139404, "learning_rate": 9.425504134629335e-06, "loss": 0.0592, "step": 14678 }, { "epoch": 1.5913920208152645, "grad_norm": 0.27547842264175415, "learning_rate": 9.424778761061947e-06, "loss": 0.0051, "step": 14679 }, { "epoch": 1.5915004336513443, "grad_norm": 0.42757079005241394, "learning_rate": 9.424053387494561e-06, "loss": 0.017, "step": 14680 }, { "epoch": 1.591608846487424, "grad_norm": 3.357668399810791, "learning_rate": 9.423328013927173e-06, "loss": 0.0356, "step": 14681 }, { "epoch": 1.591717259323504, "grad_norm": 3.2805521488189697, "learning_rate": 9.422602640359785e-06, "loss": 0.0242, "step": 14682 }, { "epoch": 1.5918256721595836, "grad_norm": 4.489615440368652, "learning_rate": 9.4218772667924e-06, "loss": 0.0443, "step": 14683 }, { "epoch": 1.5919340849956636, "grad_norm": 3.3258187770843506, "learning_rate": 9.421151893225011e-06, "loss": 0.0474, "step": 14684 }, { "epoch": 1.5920424978317431, "grad_norm": 0.611504316329956, "learning_rate": 9.420426519657623e-06, "loss": 0.0017, "step": 14685 }, { "epoch": 1.592150910667823, "grad_norm": 6.909637451171875, "learning_rate": 9.419701146090237e-06, "loss": 0.0603, "step": 14686 }, { "epoch": 1.5922593235039029, "grad_norm": 3.186534881591797, "learning_rate": 9.418975772522851e-06, "loss": 0.0874, "step": 14687 }, { "epoch": 1.5923677363399826, "grad_norm": 0.43415477871894836, "learning_rate": 9.418250398955463e-06, "loss": 0.0122, "step": 14688 }, { "epoch": 1.5924761491760624, "grad_norm": 0.8105874061584473, "learning_rate": 9.417525025388075e-06, "loss": 0.0194, "step": 14689 }, { "epoch": 1.5925845620121422, "grad_norm": 1.8515045642852783, "learning_rate": 9.416799651820688e-06, "loss": 0.0342, "step": 14690 }, { "epoch": 1.5926929748482221, "grad_norm": 4.0725603103637695, "learning_rate": 9.416074278253301e-06, "loss": 0.0347, "step": 14691 }, { "epoch": 1.5928013876843017, "grad_norm": 7.355429172515869, "learning_rate": 9.415348904685914e-06, "loss": 0.0608, "step": 14692 }, { "epoch": 1.5929098005203817, "grad_norm": 1.0179959535598755, "learning_rate": 9.414623531118526e-06, "loss": 0.0067, "step": 14693 }, { "epoch": 1.5930182133564614, "grad_norm": 1.6108942031860352, "learning_rate": 9.41389815755114e-06, "loss": 0.0144, "step": 14694 }, { "epoch": 1.5931266261925412, "grad_norm": 1.3112882375717163, "learning_rate": 9.413172783983753e-06, "loss": 0.0276, "step": 14695 }, { "epoch": 1.593235039028621, "grad_norm": 2.153346538543701, "learning_rate": 9.412447410416366e-06, "loss": 0.0322, "step": 14696 }, { "epoch": 1.5933434518647007, "grad_norm": 3.530473232269287, "learning_rate": 9.411722036848978e-06, "loss": 0.0564, "step": 14697 }, { "epoch": 1.5934518647007807, "grad_norm": 2.1979165077209473, "learning_rate": 9.410996663281591e-06, "loss": 0.0439, "step": 14698 }, { "epoch": 1.5935602775368602, "grad_norm": 0.31279927492141724, "learning_rate": 9.410271289714204e-06, "loss": 0.0052, "step": 14699 }, { "epoch": 1.5936686903729402, "grad_norm": 0.6024447083473206, "learning_rate": 9.409545916146816e-06, "loss": 0.0024, "step": 14700 }, { "epoch": 1.59377710320902, "grad_norm": 3.6740562915802, "learning_rate": 9.40882054257943e-06, "loss": 0.0179, "step": 14701 }, { "epoch": 1.5938855160450998, "grad_norm": 0.32148686051368713, "learning_rate": 9.408095169012042e-06, "loss": 0.0115, "step": 14702 }, { "epoch": 1.5939939288811795, "grad_norm": 0.3501225411891937, "learning_rate": 9.407369795444656e-06, "loss": 0.0079, "step": 14703 }, { "epoch": 1.5941023417172593, "grad_norm": 0.49249231815338135, "learning_rate": 9.406644421877268e-06, "loss": 0.0024, "step": 14704 }, { "epoch": 1.5942107545533393, "grad_norm": 0.3662688434123993, "learning_rate": 9.40591904830988e-06, "loss": 0.0067, "step": 14705 }, { "epoch": 1.5943191673894188, "grad_norm": 2.9802141189575195, "learning_rate": 9.405193674742494e-06, "loss": 0.048, "step": 14706 }, { "epoch": 1.5944275802254988, "grad_norm": 10.014412879943848, "learning_rate": 9.404468301175106e-06, "loss": 0.0414, "step": 14707 }, { "epoch": 1.5945359930615783, "grad_norm": 1.6323156356811523, "learning_rate": 9.403742927607718e-06, "loss": 0.0039, "step": 14708 }, { "epoch": 1.5946444058976583, "grad_norm": 10.546412467956543, "learning_rate": 9.403017554040332e-06, "loss": 0.0389, "step": 14709 }, { "epoch": 1.594752818733738, "grad_norm": 2.163172960281372, "learning_rate": 9.402292180472944e-06, "loss": 0.0666, "step": 14710 }, { "epoch": 1.5948612315698178, "grad_norm": 0.30152925848960876, "learning_rate": 9.401566806905558e-06, "loss": 0.0064, "step": 14711 }, { "epoch": 1.5949696444058976, "grad_norm": 0.41005316376686096, "learning_rate": 9.40084143333817e-06, "loss": 0.011, "step": 14712 }, { "epoch": 1.5950780572419774, "grad_norm": 0.10407562553882599, "learning_rate": 9.400116059770784e-06, "loss": 0.0005, "step": 14713 }, { "epoch": 1.5951864700780574, "grad_norm": 4.270118713378906, "learning_rate": 9.399390686203396e-06, "loss": 0.0234, "step": 14714 }, { "epoch": 1.595294882914137, "grad_norm": 3.0469002723693848, "learning_rate": 9.398665312636008e-06, "loss": 0.0078, "step": 14715 }, { "epoch": 1.5954032957502169, "grad_norm": 0.2703753709793091, "learning_rate": 9.39793993906862e-06, "loss": 0.0065, "step": 14716 }, { "epoch": 1.5955117085862967, "grad_norm": 3.4771504402160645, "learning_rate": 9.397214565501234e-06, "loss": 0.0046, "step": 14717 }, { "epoch": 1.5956201214223764, "grad_norm": 3.4887263774871826, "learning_rate": 9.396489191933846e-06, "loss": 0.0382, "step": 14718 }, { "epoch": 1.5957285342584562, "grad_norm": 4.539914131164551, "learning_rate": 9.395763818366458e-06, "loss": 0.0283, "step": 14719 }, { "epoch": 1.595836947094536, "grad_norm": 2.568533420562744, "learning_rate": 9.395038444799072e-06, "loss": 0.0517, "step": 14720 }, { "epoch": 1.595945359930616, "grad_norm": 1.774488925933838, "learning_rate": 9.394313071231686e-06, "loss": 0.0269, "step": 14721 }, { "epoch": 1.5960537727666955, "grad_norm": 3.788522481918335, "learning_rate": 9.393587697664298e-06, "loss": 0.0313, "step": 14722 }, { "epoch": 1.5961621856027755, "grad_norm": 0.1219652071595192, "learning_rate": 9.39286232409691e-06, "loss": 0.0011, "step": 14723 }, { "epoch": 1.5962705984388552, "grad_norm": 0.9085721373558044, "learning_rate": 9.392136950529524e-06, "loss": 0.0135, "step": 14724 }, { "epoch": 1.596379011274935, "grad_norm": 11.090314865112305, "learning_rate": 9.391411576962136e-06, "loss": 0.0127, "step": 14725 }, { "epoch": 1.5964874241110147, "grad_norm": 0.9464656114578247, "learning_rate": 9.390686203394748e-06, "loss": 0.0145, "step": 14726 }, { "epoch": 1.5965958369470945, "grad_norm": 2.102123260498047, "learning_rate": 9.389960829827362e-06, "loss": 0.0545, "step": 14727 }, { "epoch": 1.5967042497831745, "grad_norm": 1.1540800333023071, "learning_rate": 9.389235456259974e-06, "loss": 0.0412, "step": 14728 }, { "epoch": 1.596812662619254, "grad_norm": 1.4745832681655884, "learning_rate": 9.388510082692588e-06, "loss": 0.046, "step": 14729 }, { "epoch": 1.596921075455334, "grad_norm": 2.6572437286376953, "learning_rate": 9.3877847091252e-06, "loss": 0.0581, "step": 14730 }, { "epoch": 1.5970294882914136, "grad_norm": 0.4824415147304535, "learning_rate": 9.387059335557813e-06, "loss": 0.0064, "step": 14731 }, { "epoch": 1.5971379011274935, "grad_norm": 2.28255033493042, "learning_rate": 9.386333961990426e-06, "loss": 0.0397, "step": 14732 }, { "epoch": 1.5972463139635733, "grad_norm": 0.8061087131500244, "learning_rate": 9.385608588423039e-06, "loss": 0.0218, "step": 14733 }, { "epoch": 1.597354726799653, "grad_norm": 1.8222482204437256, "learning_rate": 9.38488321485565e-06, "loss": 0.0174, "step": 14734 }, { "epoch": 1.5974631396357328, "grad_norm": 11.350746154785156, "learning_rate": 9.384157841288265e-06, "loss": 0.0399, "step": 14735 }, { "epoch": 1.5975715524718126, "grad_norm": 0.10371964424848557, "learning_rate": 9.383432467720877e-06, "loss": 0.0009, "step": 14736 }, { "epoch": 1.5976799653078926, "grad_norm": 1.4677103757858276, "learning_rate": 9.38270709415349e-06, "loss": 0.0017, "step": 14737 }, { "epoch": 1.5977883781439721, "grad_norm": 1.9614983797073364, "learning_rate": 9.381981720586103e-06, "loss": 0.0091, "step": 14738 }, { "epoch": 1.597896790980052, "grad_norm": 1.6396235227584839, "learning_rate": 9.381256347018716e-06, "loss": 0.0262, "step": 14739 }, { "epoch": 1.5980052038161319, "grad_norm": 0.13614167273044586, "learning_rate": 9.380530973451329e-06, "loss": 0.0014, "step": 14740 }, { "epoch": 1.5981136166522116, "grad_norm": 8.835854530334473, "learning_rate": 9.37980559988394e-06, "loss": 0.025, "step": 14741 }, { "epoch": 1.5982220294882914, "grad_norm": 1.4156543016433716, "learning_rate": 9.379080226316553e-06, "loss": 0.0206, "step": 14742 }, { "epoch": 1.5983304423243712, "grad_norm": 3.189561367034912, "learning_rate": 9.378354852749167e-06, "loss": 0.055, "step": 14743 }, { "epoch": 1.5984388551604511, "grad_norm": 2.992276430130005, "learning_rate": 9.377629479181779e-06, "loss": 0.038, "step": 14744 }, { "epoch": 1.5985472679965307, "grad_norm": 0.43117934465408325, "learning_rate": 9.376904105614391e-06, "loss": 0.0101, "step": 14745 }, { "epoch": 1.5986556808326107, "grad_norm": 2.9954137802124023, "learning_rate": 9.376178732047005e-06, "loss": 0.008, "step": 14746 }, { "epoch": 1.5987640936686902, "grad_norm": 0.9124048948287964, "learning_rate": 9.375453358479619e-06, "loss": 0.0054, "step": 14747 }, { "epoch": 1.5988725065047702, "grad_norm": 1.9383907318115234, "learning_rate": 9.374727984912231e-06, "loss": 0.0422, "step": 14748 }, { "epoch": 1.59898091934085, "grad_norm": 1.8869236707687378, "learning_rate": 9.374002611344843e-06, "loss": 0.0114, "step": 14749 }, { "epoch": 1.5990893321769297, "grad_norm": 0.4446215033531189, "learning_rate": 9.373277237777457e-06, "loss": 0.0058, "step": 14750 }, { "epoch": 1.5991977450130095, "grad_norm": 4.646246910095215, "learning_rate": 9.372551864210069e-06, "loss": 0.0398, "step": 14751 }, { "epoch": 1.5993061578490892, "grad_norm": 2.373492479324341, "learning_rate": 9.371826490642681e-06, "loss": 0.0082, "step": 14752 }, { "epoch": 1.5994145706851692, "grad_norm": 2.2935516834259033, "learning_rate": 9.371101117075295e-06, "loss": 0.0463, "step": 14753 }, { "epoch": 1.5995229835212488, "grad_norm": 2.821133852005005, "learning_rate": 9.370375743507907e-06, "loss": 0.053, "step": 14754 }, { "epoch": 1.5996313963573288, "grad_norm": 0.5107030272483826, "learning_rate": 9.369650369940521e-06, "loss": 0.0149, "step": 14755 }, { "epoch": 1.5997398091934085, "grad_norm": 2.1303529739379883, "learning_rate": 9.368924996373133e-06, "loss": 0.0279, "step": 14756 }, { "epoch": 1.5998482220294883, "grad_norm": 1.553466796875, "learning_rate": 9.368199622805745e-06, "loss": 0.0374, "step": 14757 }, { "epoch": 1.599956634865568, "grad_norm": 3.266404390335083, "learning_rate": 9.367474249238359e-06, "loss": 0.0128, "step": 14758 }, { "epoch": 1.6000650477016478, "grad_norm": 4.623193264007568, "learning_rate": 9.366748875670971e-06, "loss": 0.0251, "step": 14759 }, { "epoch": 1.6001734605377278, "grad_norm": 4.750661373138428, "learning_rate": 9.366023502103583e-06, "loss": 0.0259, "step": 14760 }, { "epoch": 1.6002818733738073, "grad_norm": 0.990058958530426, "learning_rate": 9.365298128536197e-06, "loss": 0.0064, "step": 14761 }, { "epoch": 1.6003902862098873, "grad_norm": 1.0448342561721802, "learning_rate": 9.36457275496881e-06, "loss": 0.0119, "step": 14762 }, { "epoch": 1.600498699045967, "grad_norm": 0.5901542901992798, "learning_rate": 9.363847381401423e-06, "loss": 0.0238, "step": 14763 }, { "epoch": 1.6006071118820469, "grad_norm": 3.3765268325805664, "learning_rate": 9.363122007834035e-06, "loss": 0.0234, "step": 14764 }, { "epoch": 1.6007155247181266, "grad_norm": 2.3607840538024902, "learning_rate": 9.362396634266649e-06, "loss": 0.0352, "step": 14765 }, { "epoch": 1.6008239375542064, "grad_norm": 1.566666603088379, "learning_rate": 9.361671260699261e-06, "loss": 0.0133, "step": 14766 }, { "epoch": 1.6009323503902864, "grad_norm": 2.282609224319458, "learning_rate": 9.360945887131873e-06, "loss": 0.0568, "step": 14767 }, { "epoch": 1.601040763226366, "grad_norm": 1.660489797592163, "learning_rate": 9.360220513564486e-06, "loss": 0.0202, "step": 14768 }, { "epoch": 1.601149176062446, "grad_norm": 0.49263790249824524, "learning_rate": 9.3594951399971e-06, "loss": 0.0149, "step": 14769 }, { "epoch": 1.6012575888985254, "grad_norm": 1.6760557889938354, "learning_rate": 9.358769766429712e-06, "loss": 0.0454, "step": 14770 }, { "epoch": 1.6013660017346054, "grad_norm": 1.5582228899002075, "learning_rate": 9.358044392862324e-06, "loss": 0.0207, "step": 14771 }, { "epoch": 1.6014744145706852, "grad_norm": 2.2533082962036133, "learning_rate": 9.357319019294938e-06, "loss": 0.0314, "step": 14772 }, { "epoch": 1.601582827406765, "grad_norm": 4.097076892852783, "learning_rate": 9.356593645727551e-06, "loss": 0.0426, "step": 14773 }, { "epoch": 1.6016912402428447, "grad_norm": 1.0519201755523682, "learning_rate": 9.355868272160164e-06, "loss": 0.0229, "step": 14774 }, { "epoch": 1.6017996530789245, "grad_norm": 0.7660413980484009, "learning_rate": 9.355142898592776e-06, "loss": 0.015, "step": 14775 }, { "epoch": 1.6019080659150045, "grad_norm": 1.4567582607269287, "learning_rate": 9.35441752502539e-06, "loss": 0.0077, "step": 14776 }, { "epoch": 1.602016478751084, "grad_norm": 2.8436758518218994, "learning_rate": 9.353692151458002e-06, "loss": 0.011, "step": 14777 }, { "epoch": 1.602124891587164, "grad_norm": 1.8726153373718262, "learning_rate": 9.352966777890614e-06, "loss": 0.0517, "step": 14778 }, { "epoch": 1.6022333044232437, "grad_norm": 1.1098909378051758, "learning_rate": 9.352241404323228e-06, "loss": 0.0079, "step": 14779 }, { "epoch": 1.6023417172593235, "grad_norm": 1.3202334642410278, "learning_rate": 9.35151603075584e-06, "loss": 0.0193, "step": 14780 }, { "epoch": 1.6024501300954033, "grad_norm": 2.9340596199035645, "learning_rate": 9.350790657188454e-06, "loss": 0.0316, "step": 14781 }, { "epoch": 1.602558542931483, "grad_norm": 0.6105090975761414, "learning_rate": 9.350065283621066e-06, "loss": 0.003, "step": 14782 }, { "epoch": 1.602666955767563, "grad_norm": 2.0828864574432373, "learning_rate": 9.349339910053678e-06, "loss": 0.0623, "step": 14783 }, { "epoch": 1.6027753686036426, "grad_norm": 4.100461483001709, "learning_rate": 9.348614536486292e-06, "loss": 0.032, "step": 14784 }, { "epoch": 1.6028837814397225, "grad_norm": 1.0424282550811768, "learning_rate": 9.347889162918904e-06, "loss": 0.0246, "step": 14785 }, { "epoch": 1.6029921942758023, "grad_norm": 0.14315950870513916, "learning_rate": 9.347163789351516e-06, "loss": 0.0007, "step": 14786 }, { "epoch": 1.603100607111882, "grad_norm": 0.41850998997688293, "learning_rate": 9.34643841578413e-06, "loss": 0.0034, "step": 14787 }, { "epoch": 1.6032090199479618, "grad_norm": 1.5260027647018433, "learning_rate": 9.345713042216742e-06, "loss": 0.0295, "step": 14788 }, { "epoch": 1.6033174327840416, "grad_norm": 1.9145371913909912, "learning_rate": 9.344987668649356e-06, "loss": 0.021, "step": 14789 }, { "epoch": 1.6034258456201216, "grad_norm": 0.5407087802886963, "learning_rate": 9.344262295081968e-06, "loss": 0.0029, "step": 14790 }, { "epoch": 1.6035342584562011, "grad_norm": 2.7836666107177734, "learning_rate": 9.343536921514582e-06, "loss": 0.0076, "step": 14791 }, { "epoch": 1.603642671292281, "grad_norm": 1.5510265827178955, "learning_rate": 9.342811547947194e-06, "loss": 0.0117, "step": 14792 }, { "epoch": 1.6037510841283606, "grad_norm": 8.153518676757812, "learning_rate": 9.342086174379806e-06, "loss": 0.0148, "step": 14793 }, { "epoch": 1.6038594969644406, "grad_norm": 1.2978507280349731, "learning_rate": 9.341360800812418e-06, "loss": 0.0318, "step": 14794 }, { "epoch": 1.6039679098005204, "grad_norm": 2.8621695041656494, "learning_rate": 9.340635427245032e-06, "loss": 0.0103, "step": 14795 }, { "epoch": 1.6040763226366002, "grad_norm": 0.4379744827747345, "learning_rate": 9.339910053677644e-06, "loss": 0.0233, "step": 14796 }, { "epoch": 1.60418473547268, "grad_norm": 3.135005235671997, "learning_rate": 9.339184680110256e-06, "loss": 0.0702, "step": 14797 }, { "epoch": 1.6042931483087597, "grad_norm": 1.5875698328018188, "learning_rate": 9.33845930654287e-06, "loss": 0.0309, "step": 14798 }, { "epoch": 1.6044015611448397, "grad_norm": 2.714545726776123, "learning_rate": 9.337733932975484e-06, "loss": 0.0271, "step": 14799 }, { "epoch": 1.6045099739809192, "grad_norm": 1.8252079486846924, "learning_rate": 9.337008559408096e-06, "loss": 0.0099, "step": 14800 }, { "epoch": 1.6046183868169992, "grad_norm": 2.7596588134765625, "learning_rate": 9.336283185840708e-06, "loss": 0.0112, "step": 14801 }, { "epoch": 1.604726799653079, "grad_norm": 1.3999842405319214, "learning_rate": 9.335557812273322e-06, "loss": 0.052, "step": 14802 }, { "epoch": 1.6048352124891587, "grad_norm": 1.0834006071090698, "learning_rate": 9.334832438705934e-06, "loss": 0.016, "step": 14803 }, { "epoch": 1.6049436253252385, "grad_norm": 2.995370864868164, "learning_rate": 9.334107065138546e-06, "loss": 0.0609, "step": 14804 }, { "epoch": 1.6050520381613183, "grad_norm": 3.594076156616211, "learning_rate": 9.33338169157116e-06, "loss": 0.0383, "step": 14805 }, { "epoch": 1.6051604509973982, "grad_norm": 5.191169738769531, "learning_rate": 9.332656318003772e-06, "loss": 0.0175, "step": 14806 }, { "epoch": 1.6052688638334778, "grad_norm": 0.26999473571777344, "learning_rate": 9.331930944436386e-06, "loss": 0.0155, "step": 14807 }, { "epoch": 1.6053772766695578, "grad_norm": 2.070812225341797, "learning_rate": 9.331205570868998e-06, "loss": 0.0251, "step": 14808 }, { "epoch": 1.6054856895056373, "grad_norm": 3.346665143966675, "learning_rate": 9.33048019730161e-06, "loss": 0.0171, "step": 14809 }, { "epoch": 1.6055941023417173, "grad_norm": 1.2561566829681396, "learning_rate": 9.329754823734224e-06, "loss": 0.0048, "step": 14810 }, { "epoch": 1.605702515177797, "grad_norm": 1.169304609298706, "learning_rate": 9.329029450166837e-06, "loss": 0.0159, "step": 14811 }, { "epoch": 1.6058109280138768, "grad_norm": 1.5653395652770996, "learning_rate": 9.328304076599449e-06, "loss": 0.0278, "step": 14812 }, { "epoch": 1.6059193408499568, "grad_norm": 2.962590456008911, "learning_rate": 9.327578703032063e-06, "loss": 0.0708, "step": 14813 }, { "epoch": 1.6060277536860363, "grad_norm": 1.449797511100769, "learning_rate": 9.326853329464675e-06, "loss": 0.0057, "step": 14814 }, { "epoch": 1.6061361665221163, "grad_norm": 2.1556742191314697, "learning_rate": 9.326127955897289e-06, "loss": 0.0388, "step": 14815 }, { "epoch": 1.6062445793581959, "grad_norm": 3.4229812622070312, "learning_rate": 9.3254025823299e-06, "loss": 0.0195, "step": 14816 }, { "epoch": 1.6063529921942759, "grad_norm": 1.5386533737182617, "learning_rate": 9.324677208762514e-06, "loss": 0.0208, "step": 14817 }, { "epoch": 1.6064614050303556, "grad_norm": 0.050673697143793106, "learning_rate": 9.323951835195127e-06, "loss": 0.0006, "step": 14818 }, { "epoch": 1.6065698178664354, "grad_norm": 6.427593231201172, "learning_rate": 9.323226461627739e-06, "loss": 0.0264, "step": 14819 }, { "epoch": 1.6066782307025151, "grad_norm": 0.21876685321331024, "learning_rate": 9.322501088060351e-06, "loss": 0.0016, "step": 14820 }, { "epoch": 1.606786643538595, "grad_norm": 1.6135811805725098, "learning_rate": 9.321775714492965e-06, "loss": 0.0345, "step": 14821 }, { "epoch": 1.606895056374675, "grad_norm": 1.3172587156295776, "learning_rate": 9.321050340925577e-06, "loss": 0.0046, "step": 14822 }, { "epoch": 1.6070034692107544, "grad_norm": 2.5119550228118896, "learning_rate": 9.320324967358189e-06, "loss": 0.02, "step": 14823 }, { "epoch": 1.6071118820468344, "grad_norm": 2.562187910079956, "learning_rate": 9.319599593790803e-06, "loss": 0.0233, "step": 14824 }, { "epoch": 1.6072202948829142, "grad_norm": 1.3905848264694214, "learning_rate": 9.318874220223417e-06, "loss": 0.0179, "step": 14825 }, { "epoch": 1.607328707718994, "grad_norm": 2.868614673614502, "learning_rate": 9.318148846656029e-06, "loss": 0.0455, "step": 14826 }, { "epoch": 1.6074371205550737, "grad_norm": 12.394783020019531, "learning_rate": 9.317423473088641e-06, "loss": 0.0544, "step": 14827 }, { "epoch": 1.6075455333911535, "grad_norm": 0.3445468544960022, "learning_rate": 9.316698099521255e-06, "loss": 0.0082, "step": 14828 }, { "epoch": 1.6076539462272335, "grad_norm": 16.617647171020508, "learning_rate": 9.315972725953867e-06, "loss": 0.0185, "step": 14829 }, { "epoch": 1.607762359063313, "grad_norm": 2.760601282119751, "learning_rate": 9.315247352386479e-06, "loss": 0.0939, "step": 14830 }, { "epoch": 1.607870771899393, "grad_norm": 8.15281867980957, "learning_rate": 9.314521978819093e-06, "loss": 0.026, "step": 14831 }, { "epoch": 1.6079791847354725, "grad_norm": 6.845024585723877, "learning_rate": 9.313796605251705e-06, "loss": 0.0532, "step": 14832 }, { "epoch": 1.6080875975715525, "grad_norm": 0.07110576331615448, "learning_rate": 9.313071231684319e-06, "loss": 0.0009, "step": 14833 }, { "epoch": 1.6081960104076323, "grad_norm": 1.5715851783752441, "learning_rate": 9.312345858116931e-06, "loss": 0.026, "step": 14834 }, { "epoch": 1.608304423243712, "grad_norm": 3.044804334640503, "learning_rate": 9.311620484549543e-06, "loss": 0.0612, "step": 14835 }, { "epoch": 1.6084128360797918, "grad_norm": 2.2724862098693848, "learning_rate": 9.310895110982157e-06, "loss": 0.0124, "step": 14836 }, { "epoch": 1.6085212489158716, "grad_norm": 0.847544252872467, "learning_rate": 9.31016973741477e-06, "loss": 0.0072, "step": 14837 }, { "epoch": 1.6086296617519515, "grad_norm": 0.5364713668823242, "learning_rate": 9.309444363847381e-06, "loss": 0.0078, "step": 14838 }, { "epoch": 1.608738074588031, "grad_norm": 0.3732278645038605, "learning_rate": 9.308718990279995e-06, "loss": 0.0012, "step": 14839 }, { "epoch": 1.608846487424111, "grad_norm": 1.6881780624389648, "learning_rate": 9.307993616712607e-06, "loss": 0.0076, "step": 14840 }, { "epoch": 1.6089549002601908, "grad_norm": 2.8134231567382812, "learning_rate": 9.307268243145221e-06, "loss": 0.0277, "step": 14841 }, { "epoch": 1.6090633130962706, "grad_norm": 3.6048176288604736, "learning_rate": 9.306542869577833e-06, "loss": 0.0278, "step": 14842 }, { "epoch": 1.6091717259323504, "grad_norm": 3.619532585144043, "learning_rate": 9.305817496010447e-06, "loss": 0.0609, "step": 14843 }, { "epoch": 1.6092801387684301, "grad_norm": 1.146141767501831, "learning_rate": 9.30509212244306e-06, "loss": 0.0209, "step": 14844 }, { "epoch": 1.60938855160451, "grad_norm": 5.418017864227295, "learning_rate": 9.304366748875671e-06, "loss": 0.0397, "step": 14845 }, { "epoch": 1.6094969644405897, "grad_norm": 0.8772813677787781, "learning_rate": 9.303641375308284e-06, "loss": 0.0165, "step": 14846 }, { "epoch": 1.6096053772766696, "grad_norm": 2.469841480255127, "learning_rate": 9.302916001740897e-06, "loss": 0.0589, "step": 14847 }, { "epoch": 1.6097137901127494, "grad_norm": 0.29374802112579346, "learning_rate": 9.30219062817351e-06, "loss": 0.0109, "step": 14848 }, { "epoch": 1.6098222029488292, "grad_norm": 0.8805221319198608, "learning_rate": 9.301465254606122e-06, "loss": 0.0071, "step": 14849 }, { "epoch": 1.609930615784909, "grad_norm": 0.5712875723838806, "learning_rate": 9.300739881038736e-06, "loss": 0.023, "step": 14850 }, { "epoch": 1.6100390286209887, "grad_norm": 2.318298578262329, "learning_rate": 9.30001450747135e-06, "loss": 0.0408, "step": 14851 }, { "epoch": 1.6101474414570687, "grad_norm": 1.7255523204803467, "learning_rate": 9.299289133903962e-06, "loss": 0.0282, "step": 14852 }, { "epoch": 1.6102558542931482, "grad_norm": 1.0427402257919312, "learning_rate": 9.298563760336574e-06, "loss": 0.0017, "step": 14853 }, { "epoch": 1.6103642671292282, "grad_norm": 0.33571088314056396, "learning_rate": 9.297838386769188e-06, "loss": 0.0121, "step": 14854 }, { "epoch": 1.6104726799653077, "grad_norm": 1.2831485271453857, "learning_rate": 9.2971130132018e-06, "loss": 0.0063, "step": 14855 }, { "epoch": 1.6105810928013877, "grad_norm": 5.539121627807617, "learning_rate": 9.296387639634412e-06, "loss": 0.0407, "step": 14856 }, { "epoch": 1.6106895056374675, "grad_norm": 1.7647923231124878, "learning_rate": 9.295662266067026e-06, "loss": 0.031, "step": 14857 }, { "epoch": 1.6107979184735473, "grad_norm": 2.007667303085327, "learning_rate": 9.294936892499638e-06, "loss": 0.035, "step": 14858 }, { "epoch": 1.610906331309627, "grad_norm": 0.5799288153648376, "learning_rate": 9.294211518932252e-06, "loss": 0.0165, "step": 14859 }, { "epoch": 1.6110147441457068, "grad_norm": 0.4025472104549408, "learning_rate": 9.293486145364864e-06, "loss": 0.0038, "step": 14860 }, { "epoch": 1.6111231569817868, "grad_norm": 0.8401881456375122, "learning_rate": 9.292760771797476e-06, "loss": 0.033, "step": 14861 }, { "epoch": 1.6112315698178663, "grad_norm": 0.348540335893631, "learning_rate": 9.29203539823009e-06, "loss": 0.0069, "step": 14862 }, { "epoch": 1.6113399826539463, "grad_norm": 0.7473383545875549, "learning_rate": 9.291310024662702e-06, "loss": 0.0137, "step": 14863 }, { "epoch": 1.611448395490026, "grad_norm": 0.5259153842926025, "learning_rate": 9.290584651095314e-06, "loss": 0.008, "step": 14864 }, { "epoch": 1.6115568083261058, "grad_norm": 0.06685755401849747, "learning_rate": 9.289859277527928e-06, "loss": 0.0009, "step": 14865 }, { "epoch": 1.6116652211621856, "grad_norm": 0.8379815816879272, "learning_rate": 9.28913390396054e-06, "loss": 0.0059, "step": 14866 }, { "epoch": 1.6117736339982653, "grad_norm": 1.881235122680664, "learning_rate": 9.288408530393154e-06, "loss": 0.0216, "step": 14867 }, { "epoch": 1.6118820468343453, "grad_norm": 0.6838498115539551, "learning_rate": 9.287683156825766e-06, "loss": 0.0125, "step": 14868 }, { "epoch": 1.6119904596704249, "grad_norm": 0.5450702905654907, "learning_rate": 9.28695778325838e-06, "loss": 0.0143, "step": 14869 }, { "epoch": 1.6120988725065049, "grad_norm": 3.497087240219116, "learning_rate": 9.286232409690992e-06, "loss": 0.0686, "step": 14870 }, { "epoch": 1.6122072853425846, "grad_norm": 3.411726236343384, "learning_rate": 9.285507036123604e-06, "loss": 0.0257, "step": 14871 }, { "epoch": 1.6123156981786644, "grad_norm": 0.18737471103668213, "learning_rate": 9.284781662556216e-06, "loss": 0.0005, "step": 14872 }, { "epoch": 1.6124241110147441, "grad_norm": 16.001855850219727, "learning_rate": 9.28405628898883e-06, "loss": 0.0164, "step": 14873 }, { "epoch": 1.612532523850824, "grad_norm": 12.938973426818848, "learning_rate": 9.283330915421442e-06, "loss": 0.0244, "step": 14874 }, { "epoch": 1.612640936686904, "grad_norm": 0.5213188529014587, "learning_rate": 9.282605541854054e-06, "loss": 0.0213, "step": 14875 }, { "epoch": 1.6127493495229834, "grad_norm": 4.662551403045654, "learning_rate": 9.281880168286668e-06, "loss": 0.0411, "step": 14876 }, { "epoch": 1.6128577623590634, "grad_norm": 1.627806544303894, "learning_rate": 9.281154794719282e-06, "loss": 0.0148, "step": 14877 }, { "epoch": 1.612966175195143, "grad_norm": 0.06288397312164307, "learning_rate": 9.280429421151894e-06, "loss": 0.0003, "step": 14878 }, { "epoch": 1.613074588031223, "grad_norm": 0.6944822072982788, "learning_rate": 9.279704047584506e-06, "loss": 0.0286, "step": 14879 }, { "epoch": 1.6131830008673027, "grad_norm": 0.6213815212249756, "learning_rate": 9.27897867401712e-06, "loss": 0.0072, "step": 14880 }, { "epoch": 1.6132914137033825, "grad_norm": 1.6053225994110107, "learning_rate": 9.278253300449732e-06, "loss": 0.0231, "step": 14881 }, { "epoch": 1.6133998265394622, "grad_norm": 4.222893238067627, "learning_rate": 9.277527926882344e-06, "loss": 0.0283, "step": 14882 }, { "epoch": 1.613508239375542, "grad_norm": 3.6934688091278076, "learning_rate": 9.276802553314958e-06, "loss": 0.0449, "step": 14883 }, { "epoch": 1.613616652211622, "grad_norm": 4.117877960205078, "learning_rate": 9.27607717974757e-06, "loss": 0.0173, "step": 14884 }, { "epoch": 1.6137250650477015, "grad_norm": 5.938858509063721, "learning_rate": 9.275351806180184e-06, "loss": 0.0888, "step": 14885 }, { "epoch": 1.6138334778837815, "grad_norm": 1.1409358978271484, "learning_rate": 9.274626432612796e-06, "loss": 0.0153, "step": 14886 }, { "epoch": 1.6139418907198613, "grad_norm": 3.6140296459198, "learning_rate": 9.273901059045409e-06, "loss": 0.04, "step": 14887 }, { "epoch": 1.614050303555941, "grad_norm": 0.4123804271221161, "learning_rate": 9.273175685478022e-06, "loss": 0.0101, "step": 14888 }, { "epoch": 1.6141587163920208, "grad_norm": 1.3501633405685425, "learning_rate": 9.272450311910635e-06, "loss": 0.0275, "step": 14889 }, { "epoch": 1.6142671292281006, "grad_norm": 1.9534307718276978, "learning_rate": 9.271724938343247e-06, "loss": 0.0301, "step": 14890 }, { "epoch": 1.6143755420641805, "grad_norm": 1.193055272102356, "learning_rate": 9.27099956477586e-06, "loss": 0.0359, "step": 14891 }, { "epoch": 1.61448395490026, "grad_norm": 0.4612313508987427, "learning_rate": 9.270274191208473e-06, "loss": 0.0073, "step": 14892 }, { "epoch": 1.61459236773634, "grad_norm": 0.37083181738853455, "learning_rate": 9.269548817641087e-06, "loss": 0.0108, "step": 14893 }, { "epoch": 1.6147007805724196, "grad_norm": 1.6807440519332886, "learning_rate": 9.268823444073699e-06, "loss": 0.0357, "step": 14894 }, { "epoch": 1.6148091934084996, "grad_norm": 3.706573009490967, "learning_rate": 9.268098070506312e-06, "loss": 0.0407, "step": 14895 }, { "epoch": 1.6149176062445794, "grad_norm": 2.076324224472046, "learning_rate": 9.267372696938925e-06, "loss": 0.0461, "step": 14896 }, { "epoch": 1.6150260190806591, "grad_norm": 1.1770384311676025, "learning_rate": 9.266647323371537e-06, "loss": 0.0061, "step": 14897 }, { "epoch": 1.6151344319167389, "grad_norm": 1.3462129831314087, "learning_rate": 9.265921949804149e-06, "loss": 0.0074, "step": 14898 }, { "epoch": 1.6152428447528187, "grad_norm": 0.6937117576599121, "learning_rate": 9.265196576236763e-06, "loss": 0.0055, "step": 14899 }, { "epoch": 1.6153512575888986, "grad_norm": 1.6253166198730469, "learning_rate": 9.264471202669375e-06, "loss": 0.0165, "step": 14900 }, { "epoch": 1.6154596704249782, "grad_norm": 1.4507194757461548, "learning_rate": 9.263745829101989e-06, "loss": 0.0482, "step": 14901 }, { "epoch": 1.6155680832610582, "grad_norm": 0.04472178965806961, "learning_rate": 9.263020455534601e-06, "loss": 0.0004, "step": 14902 }, { "epoch": 1.615676496097138, "grad_norm": 1.642988681793213, "learning_rate": 9.262295081967215e-06, "loss": 0.0125, "step": 14903 }, { "epoch": 1.6157849089332177, "grad_norm": 2.698352575302124, "learning_rate": 9.261569708399827e-06, "loss": 0.0373, "step": 14904 }, { "epoch": 1.6158933217692975, "grad_norm": 1.9554401636123657, "learning_rate": 9.260844334832439e-06, "loss": 0.0257, "step": 14905 }, { "epoch": 1.6160017346053772, "grad_norm": 2.613283395767212, "learning_rate": 9.260118961265053e-06, "loss": 0.0445, "step": 14906 }, { "epoch": 1.6161101474414572, "grad_norm": 2.7602996826171875, "learning_rate": 9.259393587697665e-06, "loss": 0.017, "step": 14907 }, { "epoch": 1.6162185602775367, "grad_norm": 0.6933571696281433, "learning_rate": 9.258668214130277e-06, "loss": 0.027, "step": 14908 }, { "epoch": 1.6163269731136167, "grad_norm": 1.339288353919983, "learning_rate": 9.257942840562891e-06, "loss": 0.0393, "step": 14909 }, { "epoch": 1.6164353859496965, "grad_norm": 2.887007474899292, "learning_rate": 9.257217466995503e-06, "loss": 0.0751, "step": 14910 }, { "epoch": 1.6165437987857763, "grad_norm": 1.8858469724655151, "learning_rate": 9.256492093428117e-06, "loss": 0.0329, "step": 14911 }, { "epoch": 1.616652211621856, "grad_norm": 3.089254856109619, "learning_rate": 9.255766719860729e-06, "loss": 0.0519, "step": 14912 }, { "epoch": 1.6167606244579358, "grad_norm": 0.11042002588510513, "learning_rate": 9.255041346293341e-06, "loss": 0.0008, "step": 14913 }, { "epoch": 1.6168690372940158, "grad_norm": 1.5834957361221313, "learning_rate": 9.254315972725955e-06, "loss": 0.0239, "step": 14914 }, { "epoch": 1.6169774501300953, "grad_norm": 0.5501869320869446, "learning_rate": 9.253590599158567e-06, "loss": 0.0039, "step": 14915 }, { "epoch": 1.6170858629661753, "grad_norm": 2.7872872352600098, "learning_rate": 9.25286522559118e-06, "loss": 0.0322, "step": 14916 }, { "epoch": 1.6171942758022548, "grad_norm": 0.16943615674972534, "learning_rate": 9.252139852023793e-06, "loss": 0.002, "step": 14917 }, { "epoch": 1.6173026886383348, "grad_norm": 0.8001420497894287, "learning_rate": 9.251414478456405e-06, "loss": 0.0134, "step": 14918 }, { "epoch": 1.6174111014744146, "grad_norm": 2.0864436626434326, "learning_rate": 9.25068910488902e-06, "loss": 0.0339, "step": 14919 }, { "epoch": 1.6175195143104943, "grad_norm": 0.9199328422546387, "learning_rate": 9.249963731321631e-06, "loss": 0.0077, "step": 14920 }, { "epoch": 1.617627927146574, "grad_norm": 2.779510974884033, "learning_rate": 9.249238357754245e-06, "loss": 0.0326, "step": 14921 }, { "epoch": 1.6177363399826539, "grad_norm": 0.5048143863677979, "learning_rate": 9.248512984186857e-06, "loss": 0.0173, "step": 14922 }, { "epoch": 1.6178447528187339, "grad_norm": 0.7320012450218201, "learning_rate": 9.24778761061947e-06, "loss": 0.0226, "step": 14923 }, { "epoch": 1.6179531656548134, "grad_norm": 4.844354629516602, "learning_rate": 9.247062237052082e-06, "loss": 0.0317, "step": 14924 }, { "epoch": 1.6180615784908934, "grad_norm": 0.2418760061264038, "learning_rate": 9.246336863484695e-06, "loss": 0.0113, "step": 14925 }, { "epoch": 1.6181699913269731, "grad_norm": 5.556978702545166, "learning_rate": 9.245611489917308e-06, "loss": 0.0215, "step": 14926 }, { "epoch": 1.618278404163053, "grad_norm": 2.0046792030334473, "learning_rate": 9.244886116349921e-06, "loss": 0.0373, "step": 14927 }, { "epoch": 1.6183868169991327, "grad_norm": 2.291682004928589, "learning_rate": 9.244160742782534e-06, "loss": 0.0271, "step": 14928 }, { "epoch": 1.6184952298352124, "grad_norm": 0.12187682837247849, "learning_rate": 9.243435369215147e-06, "loss": 0.0011, "step": 14929 }, { "epoch": 1.6186036426712924, "grad_norm": 0.15971669554710388, "learning_rate": 9.24270999564776e-06, "loss": 0.0013, "step": 14930 }, { "epoch": 1.618712055507372, "grad_norm": 1.1784533262252808, "learning_rate": 9.241984622080372e-06, "loss": 0.0381, "step": 14931 }, { "epoch": 1.618820468343452, "grad_norm": 0.33108100295066833, "learning_rate": 9.241259248512986e-06, "loss": 0.0105, "step": 14932 }, { "epoch": 1.6189288811795317, "grad_norm": 6.130397796630859, "learning_rate": 9.240533874945598e-06, "loss": 0.1067, "step": 14933 }, { "epoch": 1.6190372940156115, "grad_norm": 5.682862758636475, "learning_rate": 9.23980850137821e-06, "loss": 0.0218, "step": 14934 }, { "epoch": 1.6191457068516912, "grad_norm": 1.2691255807876587, "learning_rate": 9.239083127810824e-06, "loss": 0.0277, "step": 14935 }, { "epoch": 1.619254119687771, "grad_norm": 0.023370450362563133, "learning_rate": 9.238357754243436e-06, "loss": 0.0003, "step": 14936 }, { "epoch": 1.619362532523851, "grad_norm": 0.742117166519165, "learning_rate": 9.23763238067605e-06, "loss": 0.0303, "step": 14937 }, { "epoch": 1.6194709453599305, "grad_norm": 1.117810606956482, "learning_rate": 9.236907007108662e-06, "loss": 0.0182, "step": 14938 }, { "epoch": 1.6195793581960105, "grad_norm": 1.586653709411621, "learning_rate": 9.236181633541274e-06, "loss": 0.0114, "step": 14939 }, { "epoch": 1.61968777103209, "grad_norm": 0.7870058417320251, "learning_rate": 9.235456259973888e-06, "loss": 0.0179, "step": 14940 }, { "epoch": 1.61979618386817, "grad_norm": 3.948976516723633, "learning_rate": 9.2347308864065e-06, "loss": 0.0293, "step": 14941 }, { "epoch": 1.6199045967042498, "grad_norm": 3.6653900146484375, "learning_rate": 9.234005512839112e-06, "loss": 0.0229, "step": 14942 }, { "epoch": 1.6200130095403296, "grad_norm": 0.5650979280471802, "learning_rate": 9.233280139271726e-06, "loss": 0.0186, "step": 14943 }, { "epoch": 1.6201214223764093, "grad_norm": 0.4403104782104492, "learning_rate": 9.232554765704338e-06, "loss": 0.0151, "step": 14944 }, { "epoch": 1.620229835212489, "grad_norm": 0.428617000579834, "learning_rate": 9.231829392136952e-06, "loss": 0.0114, "step": 14945 }, { "epoch": 1.620338248048569, "grad_norm": 3.9609429836273193, "learning_rate": 9.231104018569564e-06, "loss": 0.014, "step": 14946 }, { "epoch": 1.6204466608846486, "grad_norm": 2.0504133701324463, "learning_rate": 9.230378645002178e-06, "loss": 0.0276, "step": 14947 }, { "epoch": 1.6205550737207286, "grad_norm": 1.357041597366333, "learning_rate": 9.22965327143479e-06, "loss": 0.0204, "step": 14948 }, { "epoch": 1.6206634865568084, "grad_norm": 0.9090738892555237, "learning_rate": 9.228927897867402e-06, "loss": 0.0209, "step": 14949 }, { "epoch": 1.6207718993928881, "grad_norm": 2.0285558700561523, "learning_rate": 9.228202524300014e-06, "loss": 0.0495, "step": 14950 }, { "epoch": 1.620880312228968, "grad_norm": 0.3682752549648285, "learning_rate": 9.227477150732628e-06, "loss": 0.0036, "step": 14951 }, { "epoch": 1.6209887250650477, "grad_norm": 1.0959607362747192, "learning_rate": 9.22675177716524e-06, "loss": 0.018, "step": 14952 }, { "epoch": 1.6210971379011276, "grad_norm": 1.4601951837539673, "learning_rate": 9.226026403597854e-06, "loss": 0.0384, "step": 14953 }, { "epoch": 1.6212055507372072, "grad_norm": 4.019326686859131, "learning_rate": 9.225301030030466e-06, "loss": 0.0194, "step": 14954 }, { "epoch": 1.6213139635732872, "grad_norm": 6.885995864868164, "learning_rate": 9.22457565646308e-06, "loss": 0.0864, "step": 14955 }, { "epoch": 1.6214223764093667, "grad_norm": 2.122363328933716, "learning_rate": 9.223850282895692e-06, "loss": 0.0419, "step": 14956 }, { "epoch": 1.6215307892454467, "grad_norm": 0.6419770121574402, "learning_rate": 9.223124909328304e-06, "loss": 0.01, "step": 14957 }, { "epoch": 1.6216392020815265, "grad_norm": 4.299265384674072, "learning_rate": 9.222399535760918e-06, "loss": 0.017, "step": 14958 }, { "epoch": 1.6217476149176062, "grad_norm": 1.5921943187713623, "learning_rate": 9.22167416219353e-06, "loss": 0.0131, "step": 14959 }, { "epoch": 1.6218560277536862, "grad_norm": 2.5811076164245605, "learning_rate": 9.220948788626142e-06, "loss": 0.0191, "step": 14960 }, { "epoch": 1.6219644405897657, "grad_norm": 3.6489200592041016, "learning_rate": 9.220223415058756e-06, "loss": 0.0274, "step": 14961 }, { "epoch": 1.6220728534258457, "grad_norm": 0.7719200253486633, "learning_rate": 9.219498041491368e-06, "loss": 0.0061, "step": 14962 }, { "epoch": 1.6221812662619253, "grad_norm": 5.331376075744629, "learning_rate": 9.218772667923982e-06, "loss": 0.027, "step": 14963 }, { "epoch": 1.6222896790980053, "grad_norm": 4.011160850524902, "learning_rate": 9.218047294356594e-06, "loss": 0.0612, "step": 14964 }, { "epoch": 1.622398091934085, "grad_norm": 2.505711793899536, "learning_rate": 9.217321920789207e-06, "loss": 0.0505, "step": 14965 }, { "epoch": 1.6225065047701648, "grad_norm": 4.2467827796936035, "learning_rate": 9.21659654722182e-06, "loss": 0.0409, "step": 14966 }, { "epoch": 1.6226149176062445, "grad_norm": 3.760450601577759, "learning_rate": 9.215871173654433e-06, "loss": 0.0264, "step": 14967 }, { "epoch": 1.6227233304423243, "grad_norm": 0.32902854681015015, "learning_rate": 9.215145800087045e-06, "loss": 0.0196, "step": 14968 }, { "epoch": 1.6228317432784043, "grad_norm": 0.7332794666290283, "learning_rate": 9.214420426519659e-06, "loss": 0.0171, "step": 14969 }, { "epoch": 1.6229401561144838, "grad_norm": 1.0387052297592163, "learning_rate": 9.21369505295227e-06, "loss": 0.0247, "step": 14970 }, { "epoch": 1.6230485689505638, "grad_norm": 0.4594515860080719, "learning_rate": 9.212969679384885e-06, "loss": 0.0034, "step": 14971 }, { "epoch": 1.6231569817866436, "grad_norm": 10.927248001098633, "learning_rate": 9.212244305817497e-06, "loss": 0.0666, "step": 14972 }, { "epoch": 1.6232653946227233, "grad_norm": 9.386734008789062, "learning_rate": 9.21151893225011e-06, "loss": 0.0288, "step": 14973 }, { "epoch": 1.623373807458803, "grad_norm": 0.24386265873908997, "learning_rate": 9.210793558682723e-06, "loss": 0.005, "step": 14974 }, { "epoch": 1.6234822202948829, "grad_norm": 2.980271100997925, "learning_rate": 9.210068185115335e-06, "loss": 0.0309, "step": 14975 }, { "epoch": 1.6235906331309629, "grad_norm": 0.8896777033805847, "learning_rate": 9.209342811547947e-06, "loss": 0.02, "step": 14976 }, { "epoch": 1.6236990459670424, "grad_norm": 2.214050531387329, "learning_rate": 9.20861743798056e-06, "loss": 0.0068, "step": 14977 }, { "epoch": 1.6238074588031224, "grad_norm": 0.9646321535110474, "learning_rate": 9.207892064413173e-06, "loss": 0.0185, "step": 14978 }, { "epoch": 1.623915871639202, "grad_norm": 2.6782569885253906, "learning_rate": 9.207166690845787e-06, "loss": 0.0272, "step": 14979 }, { "epoch": 1.624024284475282, "grad_norm": 2.602649211883545, "learning_rate": 9.206441317278399e-06, "loss": 0.0185, "step": 14980 }, { "epoch": 1.6241326973113617, "grad_norm": 0.267912358045578, "learning_rate": 9.205715943711013e-06, "loss": 0.0062, "step": 14981 }, { "epoch": 1.6242411101474414, "grad_norm": 2.855048894882202, "learning_rate": 9.204990570143625e-06, "loss": 0.0191, "step": 14982 }, { "epoch": 1.6243495229835212, "grad_norm": 0.10651564598083496, "learning_rate": 9.204265196576237e-06, "loss": 0.0006, "step": 14983 }, { "epoch": 1.624457935819601, "grad_norm": 1.8108904361724854, "learning_rate": 9.203539823008851e-06, "loss": 0.0152, "step": 14984 }, { "epoch": 1.624566348655681, "grad_norm": 2.9103174209594727, "learning_rate": 9.202814449441463e-06, "loss": 0.0201, "step": 14985 }, { "epoch": 1.6246747614917605, "grad_norm": 1.9704322814941406, "learning_rate": 9.202089075874075e-06, "loss": 0.0164, "step": 14986 }, { "epoch": 1.6247831743278405, "grad_norm": 2.010896921157837, "learning_rate": 9.201363702306689e-06, "loss": 0.0245, "step": 14987 }, { "epoch": 1.6248915871639202, "grad_norm": 1.5807812213897705, "learning_rate": 9.200638328739301e-06, "loss": 0.0323, "step": 14988 }, { "epoch": 1.625, "grad_norm": 0.43429434299468994, "learning_rate": 9.199912955171915e-06, "loss": 0.0037, "step": 14989 }, { "epoch": 1.6251084128360798, "grad_norm": 1.190622091293335, "learning_rate": 9.199187581604527e-06, "loss": 0.0259, "step": 14990 }, { "epoch": 1.6252168256721595, "grad_norm": 1.4771478176116943, "learning_rate": 9.19846220803714e-06, "loss": 0.0214, "step": 14991 }, { "epoch": 1.6253252385082395, "grad_norm": 0.9082456231117249, "learning_rate": 9.197736834469753e-06, "loss": 0.0163, "step": 14992 }, { "epoch": 1.625433651344319, "grad_norm": 1.1985348463058472, "learning_rate": 9.197011460902365e-06, "loss": 0.0041, "step": 14993 }, { "epoch": 1.625542064180399, "grad_norm": 0.4474407434463501, "learning_rate": 9.196286087334977e-06, "loss": 0.0084, "step": 14994 }, { "epoch": 1.6256504770164788, "grad_norm": 0.4679880738258362, "learning_rate": 9.195560713767591e-06, "loss": 0.0107, "step": 14995 }, { "epoch": 1.6257588898525586, "grad_norm": 0.43310660123825073, "learning_rate": 9.194835340200205e-06, "loss": 0.009, "step": 14996 }, { "epoch": 1.6258673026886383, "grad_norm": 2.579946517944336, "learning_rate": 9.194109966632817e-06, "loss": 0.0222, "step": 14997 }, { "epoch": 1.625975715524718, "grad_norm": 1.635884404182434, "learning_rate": 9.19338459306543e-06, "loss": 0.0288, "step": 14998 }, { "epoch": 1.626084128360798, "grad_norm": 3.1900722980499268, "learning_rate": 9.192659219498043e-06, "loss": 0.0653, "step": 14999 }, { "epoch": 1.6261925411968776, "grad_norm": 0.30576464533805847, "learning_rate": 9.191933845930655e-06, "loss": 0.0081, "step": 15000 }, { "epoch": 1.6263009540329576, "grad_norm": 0.4358081519603729, "learning_rate": 9.191208472363267e-06, "loss": 0.0212, "step": 15001 }, { "epoch": 1.6264093668690371, "grad_norm": 0.6356130242347717, "learning_rate": 9.19048309879588e-06, "loss": 0.0111, "step": 15002 }, { "epoch": 1.6265177797051171, "grad_norm": 0.4664842188358307, "learning_rate": 9.189757725228493e-06, "loss": 0.0029, "step": 15003 }, { "epoch": 1.626626192541197, "grad_norm": 2.5348074436187744, "learning_rate": 9.189032351661106e-06, "loss": 0.0121, "step": 15004 }, { "epoch": 1.6267346053772767, "grad_norm": 0.044157177209854126, "learning_rate": 9.18830697809372e-06, "loss": 0.0003, "step": 15005 }, { "epoch": 1.6268430182133564, "grad_norm": 0.013910694047808647, "learning_rate": 9.187581604526332e-06, "loss": 0.0001, "step": 15006 }, { "epoch": 1.6269514310494362, "grad_norm": 3.9563987255096436, "learning_rate": 9.186856230958945e-06, "loss": 0.0129, "step": 15007 }, { "epoch": 1.6270598438855162, "grad_norm": 1.061083197593689, "learning_rate": 9.186130857391558e-06, "loss": 0.009, "step": 15008 }, { "epoch": 1.6271682567215957, "grad_norm": 2.3152108192443848, "learning_rate": 9.18540548382417e-06, "loss": 0.0265, "step": 15009 }, { "epoch": 1.6272766695576757, "grad_norm": 2.946207046508789, "learning_rate": 9.184680110256784e-06, "loss": 0.0324, "step": 15010 }, { "epoch": 1.6273850823937555, "grad_norm": 7.005158424377441, "learning_rate": 9.183954736689396e-06, "loss": 0.0485, "step": 15011 }, { "epoch": 1.6274934952298352, "grad_norm": 0.635008692741394, "learning_rate": 9.183229363122008e-06, "loss": 0.0038, "step": 15012 }, { "epoch": 1.627601908065915, "grad_norm": 2.6235766410827637, "learning_rate": 9.182503989554622e-06, "loss": 0.0551, "step": 15013 }, { "epoch": 1.6277103209019947, "grad_norm": 3.3889927864074707, "learning_rate": 9.181778615987235e-06, "loss": 0.0341, "step": 15014 }, { "epoch": 1.6278187337380747, "grad_norm": 1.82490074634552, "learning_rate": 9.181053242419848e-06, "loss": 0.0189, "step": 15015 }, { "epoch": 1.6279271465741543, "grad_norm": 0.8394601941108704, "learning_rate": 9.18032786885246e-06, "loss": 0.0222, "step": 15016 }, { "epoch": 1.6280355594102343, "grad_norm": 3.9997408390045166, "learning_rate": 9.179602495285072e-06, "loss": 0.0402, "step": 15017 }, { "epoch": 1.6281439722463138, "grad_norm": 0.34633928537368774, "learning_rate": 9.178877121717686e-06, "loss": 0.0084, "step": 15018 }, { "epoch": 1.6282523850823938, "grad_norm": 0.07596265524625778, "learning_rate": 9.178151748150298e-06, "loss": 0.0004, "step": 15019 }, { "epoch": 1.6283607979184735, "grad_norm": 0.099813312292099, "learning_rate": 9.17742637458291e-06, "loss": 0.0008, "step": 15020 }, { "epoch": 1.6284692107545533, "grad_norm": 4.421648025512695, "learning_rate": 9.176701001015524e-06, "loss": 0.017, "step": 15021 }, { "epoch": 1.6285776235906333, "grad_norm": 0.31549713015556335, "learning_rate": 9.175975627448138e-06, "loss": 0.0054, "step": 15022 }, { "epoch": 1.6286860364267128, "grad_norm": 3.753142833709717, "learning_rate": 9.17525025388075e-06, "loss": 0.0184, "step": 15023 }, { "epoch": 1.6287944492627928, "grad_norm": 3.3688008785247803, "learning_rate": 9.174524880313362e-06, "loss": 0.015, "step": 15024 }, { "epoch": 1.6289028620988724, "grad_norm": 0.31713277101516724, "learning_rate": 9.173799506745976e-06, "loss": 0.004, "step": 15025 }, { "epoch": 1.6290112749349523, "grad_norm": 4.608153343200684, "learning_rate": 9.173074133178588e-06, "loss": 0.0477, "step": 15026 }, { "epoch": 1.629119687771032, "grad_norm": 0.15944886207580566, "learning_rate": 9.1723487596112e-06, "loss": 0.0007, "step": 15027 }, { "epoch": 1.6292281006071119, "grad_norm": 4.449586868286133, "learning_rate": 9.171623386043812e-06, "loss": 0.0784, "step": 15028 }, { "epoch": 1.6293365134431916, "grad_norm": 8.79099178314209, "learning_rate": 9.170898012476426e-06, "loss": 0.0429, "step": 15029 }, { "epoch": 1.6294449262792714, "grad_norm": 13.619941711425781, "learning_rate": 9.170172638909038e-06, "loss": 0.0253, "step": 15030 }, { "epoch": 1.6295533391153514, "grad_norm": 7.670693874359131, "learning_rate": 9.169447265341652e-06, "loss": 0.0858, "step": 15031 }, { "epoch": 1.629661751951431, "grad_norm": 2.601672887802124, "learning_rate": 9.168721891774264e-06, "loss": 0.022, "step": 15032 }, { "epoch": 1.629770164787511, "grad_norm": 1.4502352476119995, "learning_rate": 9.167996518206878e-06, "loss": 0.0064, "step": 15033 }, { "epoch": 1.6298785776235907, "grad_norm": 4.517495632171631, "learning_rate": 9.16727114463949e-06, "loss": 0.0582, "step": 15034 }, { "epoch": 1.6299869904596704, "grad_norm": 2.0792627334594727, "learning_rate": 9.166545771072102e-06, "loss": 0.0068, "step": 15035 }, { "epoch": 1.6300954032957502, "grad_norm": 1.6585570573806763, "learning_rate": 9.165820397504716e-06, "loss": 0.0177, "step": 15036 }, { "epoch": 1.63020381613183, "grad_norm": 0.1091306284070015, "learning_rate": 9.165095023937328e-06, "loss": 0.001, "step": 15037 }, { "epoch": 1.63031222896791, "grad_norm": 0.47267836332321167, "learning_rate": 9.16436965036994e-06, "loss": 0.0113, "step": 15038 }, { "epoch": 1.6304206418039895, "grad_norm": 3.3905177116394043, "learning_rate": 9.163644276802554e-06, "loss": 0.009, "step": 15039 }, { "epoch": 1.6305290546400695, "grad_norm": 0.2947938144207001, "learning_rate": 9.162918903235168e-06, "loss": 0.0168, "step": 15040 }, { "epoch": 1.630637467476149, "grad_norm": 0.5739681720733643, "learning_rate": 9.16219352966778e-06, "loss": 0.0292, "step": 15041 }, { "epoch": 1.630745880312229, "grad_norm": 2.224637269973755, "learning_rate": 9.161468156100392e-06, "loss": 0.0186, "step": 15042 }, { "epoch": 1.6308542931483088, "grad_norm": 0.4107353389263153, "learning_rate": 9.160742782533005e-06, "loss": 0.0102, "step": 15043 }, { "epoch": 1.6309627059843885, "grad_norm": 0.45084312558174133, "learning_rate": 9.160017408965618e-06, "loss": 0.0133, "step": 15044 }, { "epoch": 1.6310711188204683, "grad_norm": 7.5139994621276855, "learning_rate": 9.15929203539823e-06, "loss": 0.0514, "step": 15045 }, { "epoch": 1.631179531656548, "grad_norm": 0.2992960214614868, "learning_rate": 9.158566661830843e-06, "loss": 0.0017, "step": 15046 }, { "epoch": 1.631287944492628, "grad_norm": 1.0022008419036865, "learning_rate": 9.157841288263457e-06, "loss": 0.0026, "step": 15047 }, { "epoch": 1.6313963573287076, "grad_norm": 2.022706985473633, "learning_rate": 9.15711591469607e-06, "loss": 0.0207, "step": 15048 }, { "epoch": 1.6315047701647876, "grad_norm": 2.968740940093994, "learning_rate": 9.156390541128683e-06, "loss": 0.0116, "step": 15049 }, { "epoch": 1.6316131830008673, "grad_norm": 2.6398122310638428, "learning_rate": 9.155665167561295e-06, "loss": 0.013, "step": 15050 }, { "epoch": 1.631721595836947, "grad_norm": 0.9360402226448059, "learning_rate": 9.154939793993909e-06, "loss": 0.0046, "step": 15051 }, { "epoch": 1.6318300086730269, "grad_norm": 0.9000341296195984, "learning_rate": 9.15421442042652e-06, "loss": 0.0054, "step": 15052 }, { "epoch": 1.6319384215091066, "grad_norm": 1.779038429260254, "learning_rate": 9.153489046859133e-06, "loss": 0.0689, "step": 15053 }, { "epoch": 1.6320468343451866, "grad_norm": 1.1011064052581787, "learning_rate": 9.152763673291745e-06, "loss": 0.0364, "step": 15054 }, { "epoch": 1.6321552471812661, "grad_norm": 1.1543909311294556, "learning_rate": 9.152038299724359e-06, "loss": 0.0077, "step": 15055 }, { "epoch": 1.6322636600173461, "grad_norm": 0.5734848380088806, "learning_rate": 9.151312926156971e-06, "loss": 0.006, "step": 15056 }, { "epoch": 1.632372072853426, "grad_norm": 0.2593822479248047, "learning_rate": 9.150587552589585e-06, "loss": 0.0044, "step": 15057 }, { "epoch": 1.6324804856895057, "grad_norm": 0.7315007448196411, "learning_rate": 9.149862179022197e-06, "loss": 0.0074, "step": 15058 }, { "epoch": 1.6325888985255854, "grad_norm": 3.0633506774902344, "learning_rate": 9.14913680545481e-06, "loss": 0.0803, "step": 15059 }, { "epoch": 1.6326973113616652, "grad_norm": 3.4991798400878906, "learning_rate": 9.148411431887423e-06, "loss": 0.0265, "step": 15060 }, { "epoch": 1.6328057241977452, "grad_norm": 0.49356168508529663, "learning_rate": 9.147686058320035e-06, "loss": 0.0146, "step": 15061 }, { "epoch": 1.6329141370338247, "grad_norm": 5.285974502563477, "learning_rate": 9.146960684752649e-06, "loss": 0.0275, "step": 15062 }, { "epoch": 1.6330225498699047, "grad_norm": 2.8872740268707275, "learning_rate": 9.146235311185261e-06, "loss": 0.0281, "step": 15063 }, { "epoch": 1.6331309627059842, "grad_norm": 2.699631929397583, "learning_rate": 9.145509937617873e-06, "loss": 0.0287, "step": 15064 }, { "epoch": 1.6332393755420642, "grad_norm": 0.1059093028306961, "learning_rate": 9.144784564050487e-06, "loss": 0.0009, "step": 15065 }, { "epoch": 1.633347788378144, "grad_norm": 7.084626197814941, "learning_rate": 9.1440591904831e-06, "loss": 0.0282, "step": 15066 }, { "epoch": 1.6334562012142237, "grad_norm": 1.9738085269927979, "learning_rate": 9.143333816915713e-06, "loss": 0.0545, "step": 15067 }, { "epoch": 1.6335646140503035, "grad_norm": 0.43955206871032715, "learning_rate": 9.142608443348325e-06, "loss": 0.0033, "step": 15068 }, { "epoch": 1.6336730268863833, "grad_norm": 2.7171902656555176, "learning_rate": 9.141883069780937e-06, "loss": 0.0294, "step": 15069 }, { "epoch": 1.6337814397224633, "grad_norm": 2.291829824447632, "learning_rate": 9.141157696213551e-06, "loss": 0.0355, "step": 15070 }, { "epoch": 1.6338898525585428, "grad_norm": 5.461297035217285, "learning_rate": 9.140432322646163e-06, "loss": 0.0734, "step": 15071 }, { "epoch": 1.6339982653946228, "grad_norm": 0.6328120231628418, "learning_rate": 9.139706949078775e-06, "loss": 0.0129, "step": 15072 }, { "epoch": 1.6341066782307025, "grad_norm": 0.08205047249794006, "learning_rate": 9.13898157551139e-06, "loss": 0.0011, "step": 15073 }, { "epoch": 1.6342150910667823, "grad_norm": 3.761012554168701, "learning_rate": 9.138256201944003e-06, "loss": 0.0359, "step": 15074 }, { "epoch": 1.634323503902862, "grad_norm": 0.5348849892616272, "learning_rate": 9.137530828376615e-06, "loss": 0.0115, "step": 15075 }, { "epoch": 1.6344319167389418, "grad_norm": 3.311734437942505, "learning_rate": 9.136805454809227e-06, "loss": 0.0552, "step": 15076 }, { "epoch": 1.6345403295750218, "grad_norm": 0.5403863787651062, "learning_rate": 9.136080081241841e-06, "loss": 0.0146, "step": 15077 }, { "epoch": 1.6346487424111014, "grad_norm": 1.639883279800415, "learning_rate": 9.135354707674453e-06, "loss": 0.0133, "step": 15078 }, { "epoch": 1.6347571552471813, "grad_norm": 2.6548054218292236, "learning_rate": 9.134629334107065e-06, "loss": 0.0294, "step": 15079 }, { "epoch": 1.6348655680832611, "grad_norm": 1.3712936639785767, "learning_rate": 9.133903960539678e-06, "loss": 0.0286, "step": 15080 }, { "epoch": 1.6349739809193409, "grad_norm": 1.689637541770935, "learning_rate": 9.133178586972291e-06, "loss": 0.0465, "step": 15081 }, { "epoch": 1.6350823937554206, "grad_norm": 1.5088915824890137, "learning_rate": 9.132453213404904e-06, "loss": 0.0183, "step": 15082 }, { "epoch": 1.6351908065915004, "grad_norm": 2.0311460494995117, "learning_rate": 9.131727839837517e-06, "loss": 0.0208, "step": 15083 }, { "epoch": 1.6352992194275804, "grad_norm": 0.8072732090950012, "learning_rate": 9.13100246627013e-06, "loss": 0.03, "step": 15084 }, { "epoch": 1.63540763226366, "grad_norm": 1.92581045627594, "learning_rate": 9.130277092702743e-06, "loss": 0.0127, "step": 15085 }, { "epoch": 1.63551604509974, "grad_norm": 1.2346136569976807, "learning_rate": 9.129551719135356e-06, "loss": 0.011, "step": 15086 }, { "epoch": 1.6356244579358195, "grad_norm": 2.0150578022003174, "learning_rate": 9.128826345567968e-06, "loss": 0.0266, "step": 15087 }, { "epoch": 1.6357328707718994, "grad_norm": 5.136524677276611, "learning_rate": 9.128100972000582e-06, "loss": 0.0533, "step": 15088 }, { "epoch": 1.6358412836079792, "grad_norm": 5.5715813636779785, "learning_rate": 9.127375598433194e-06, "loss": 0.0321, "step": 15089 }, { "epoch": 1.635949696444059, "grad_norm": 1.7135881185531616, "learning_rate": 9.126650224865806e-06, "loss": 0.0198, "step": 15090 }, { "epoch": 1.6360581092801387, "grad_norm": 8.758293151855469, "learning_rate": 9.12592485129842e-06, "loss": 0.0511, "step": 15091 }, { "epoch": 1.6361665221162185, "grad_norm": 0.4509046673774719, "learning_rate": 9.125199477731033e-06, "loss": 0.0096, "step": 15092 }, { "epoch": 1.6362749349522985, "grad_norm": 0.4873078167438507, "learning_rate": 9.124474104163646e-06, "loss": 0.0036, "step": 15093 }, { "epoch": 1.636383347788378, "grad_norm": 0.31217527389526367, "learning_rate": 9.123748730596258e-06, "loss": 0.0034, "step": 15094 }, { "epoch": 1.636491760624458, "grad_norm": 2.4727494716644287, "learning_rate": 9.12302335702887e-06, "loss": 0.0218, "step": 15095 }, { "epoch": 1.6366001734605378, "grad_norm": 3.443746328353882, "learning_rate": 9.122297983461484e-06, "loss": 0.023, "step": 15096 }, { "epoch": 1.6367085862966175, "grad_norm": 0.4256133735179901, "learning_rate": 9.121572609894096e-06, "loss": 0.003, "step": 15097 }, { "epoch": 1.6368169991326973, "grad_norm": 1.033676028251648, "learning_rate": 9.120847236326708e-06, "loss": 0.0369, "step": 15098 }, { "epoch": 1.636925411968777, "grad_norm": 0.44479963183403015, "learning_rate": 9.120121862759322e-06, "loss": 0.0067, "step": 15099 }, { "epoch": 1.637033824804857, "grad_norm": 0.30320829153060913, "learning_rate": 9.119396489191936e-06, "loss": 0.0115, "step": 15100 }, { "epoch": 1.6371422376409366, "grad_norm": 0.7857376933097839, "learning_rate": 9.118671115624548e-06, "loss": 0.0057, "step": 15101 }, { "epoch": 1.6372506504770166, "grad_norm": 4.778924465179443, "learning_rate": 9.11794574205716e-06, "loss": 0.063, "step": 15102 }, { "epoch": 1.637359063313096, "grad_norm": 0.7949631810188293, "learning_rate": 9.117220368489774e-06, "loss": 0.0208, "step": 15103 }, { "epoch": 1.637467476149176, "grad_norm": 0.07962155342102051, "learning_rate": 9.116494994922386e-06, "loss": 0.0007, "step": 15104 }, { "epoch": 1.6375758889852559, "grad_norm": 1.887002944946289, "learning_rate": 9.115769621354998e-06, "loss": 0.0159, "step": 15105 }, { "epoch": 1.6376843018213356, "grad_norm": 7.480710029602051, "learning_rate": 9.11504424778761e-06, "loss": 0.0503, "step": 15106 }, { "epoch": 1.6377927146574154, "grad_norm": 0.5916144251823425, "learning_rate": 9.114318874220224e-06, "loss": 0.0081, "step": 15107 }, { "epoch": 1.6379011274934951, "grad_norm": 7.705348014831543, "learning_rate": 9.113593500652836e-06, "loss": 0.0269, "step": 15108 }, { "epoch": 1.6380095403295751, "grad_norm": 5.7895073890686035, "learning_rate": 9.11286812708545e-06, "loss": 0.0251, "step": 15109 }, { "epoch": 1.6381179531656547, "grad_norm": 0.23303361237049103, "learning_rate": 9.112142753518062e-06, "loss": 0.0005, "step": 15110 }, { "epoch": 1.6382263660017347, "grad_norm": 2.7282943725585938, "learning_rate": 9.111417379950676e-06, "loss": 0.0648, "step": 15111 }, { "epoch": 1.6383347788378144, "grad_norm": 1.3531041145324707, "learning_rate": 9.110692006383288e-06, "loss": 0.0096, "step": 15112 }, { "epoch": 1.6384431916738942, "grad_norm": 0.8096768260002136, "learning_rate": 9.1099666328159e-06, "loss": 0.0151, "step": 15113 }, { "epoch": 1.638551604509974, "grad_norm": 0.845488965511322, "learning_rate": 9.109241259248514e-06, "loss": 0.0047, "step": 15114 }, { "epoch": 1.6386600173460537, "grad_norm": 1.1111836433410645, "learning_rate": 9.108515885681126e-06, "loss": 0.0193, "step": 15115 }, { "epoch": 1.6387684301821337, "grad_norm": 4.535961151123047, "learning_rate": 9.107790512113738e-06, "loss": 0.043, "step": 15116 }, { "epoch": 1.6388768430182132, "grad_norm": 0.9152356386184692, "learning_rate": 9.107065138546352e-06, "loss": 0.0067, "step": 15117 }, { "epoch": 1.6389852558542932, "grad_norm": 10.438447952270508, "learning_rate": 9.106339764978966e-06, "loss": 0.0547, "step": 15118 }, { "epoch": 1.639093668690373, "grad_norm": 0.5856316089630127, "learning_rate": 9.105614391411578e-06, "loss": 0.0275, "step": 15119 }, { "epoch": 1.6392020815264527, "grad_norm": 5.328556537628174, "learning_rate": 9.10488901784419e-06, "loss": 0.0226, "step": 15120 }, { "epoch": 1.6393104943625325, "grad_norm": 2.465139389038086, "learning_rate": 9.104163644276803e-06, "loss": 0.0551, "step": 15121 }, { "epoch": 1.6394189071986123, "grad_norm": 1.6079120635986328, "learning_rate": 9.103438270709416e-06, "loss": 0.0272, "step": 15122 }, { "epoch": 1.6395273200346923, "grad_norm": 0.6726207137107849, "learning_rate": 9.102712897142029e-06, "loss": 0.0111, "step": 15123 }, { "epoch": 1.6396357328707718, "grad_norm": 2.822312116622925, "learning_rate": 9.10198752357464e-06, "loss": 0.0185, "step": 15124 }, { "epoch": 1.6397441457068518, "grad_norm": 8.110525131225586, "learning_rate": 9.101262150007255e-06, "loss": 0.0137, "step": 15125 }, { "epoch": 1.6398525585429313, "grad_norm": 4.905274391174316, "learning_rate": 9.100536776439868e-06, "loss": 0.0198, "step": 15126 }, { "epoch": 1.6399609713790113, "grad_norm": 3.675321578979492, "learning_rate": 9.09981140287248e-06, "loss": 0.0735, "step": 15127 }, { "epoch": 1.640069384215091, "grad_norm": 3.3702921867370605, "learning_rate": 9.099086029305093e-06, "loss": 0.0344, "step": 15128 }, { "epoch": 1.6401777970511708, "grad_norm": 2.557238817214966, "learning_rate": 9.098360655737707e-06, "loss": 0.0203, "step": 15129 }, { "epoch": 1.6402862098872506, "grad_norm": 0.4569324254989624, "learning_rate": 9.097635282170319e-06, "loss": 0.013, "step": 15130 }, { "epoch": 1.6403946227233304, "grad_norm": 1.6976395845413208, "learning_rate": 9.09690990860293e-06, "loss": 0.0041, "step": 15131 }, { "epoch": 1.6405030355594103, "grad_norm": 1.967970371246338, "learning_rate": 9.096184535035543e-06, "loss": 0.0642, "step": 15132 }, { "epoch": 1.64061144839549, "grad_norm": 5.196765899658203, "learning_rate": 9.095459161468157e-06, "loss": 0.0262, "step": 15133 }, { "epoch": 1.6407198612315699, "grad_norm": 0.1502516120672226, "learning_rate": 9.094733787900769e-06, "loss": 0.001, "step": 15134 }, { "epoch": 1.6408282740676496, "grad_norm": 2.1520636081695557, "learning_rate": 9.094008414333383e-06, "loss": 0.0173, "step": 15135 }, { "epoch": 1.6409366869037294, "grad_norm": 4.057706832885742, "learning_rate": 9.093283040765995e-06, "loss": 0.0273, "step": 15136 }, { "epoch": 1.6410450997398092, "grad_norm": 0.304562509059906, "learning_rate": 9.092557667198609e-06, "loss": 0.0062, "step": 15137 }, { "epoch": 1.641153512575889, "grad_norm": 1.4759938716888428, "learning_rate": 9.091832293631221e-06, "loss": 0.0275, "step": 15138 }, { "epoch": 1.641261925411969, "grad_norm": 0.42087873816490173, "learning_rate": 9.091106920063833e-06, "loss": 0.0096, "step": 15139 }, { "epoch": 1.6413703382480485, "grad_norm": 0.6669497489929199, "learning_rate": 9.090381546496447e-06, "loss": 0.0038, "step": 15140 }, { "epoch": 1.6414787510841284, "grad_norm": 0.5608685612678528, "learning_rate": 9.089656172929059e-06, "loss": 0.0031, "step": 15141 }, { "epoch": 1.6415871639202082, "grad_norm": 1.9985949993133545, "learning_rate": 9.088930799361671e-06, "loss": 0.028, "step": 15142 }, { "epoch": 1.641695576756288, "grad_norm": 1.0188071727752686, "learning_rate": 9.088205425794285e-06, "loss": 0.0069, "step": 15143 }, { "epoch": 1.6418039895923677, "grad_norm": 3.4982826709747314, "learning_rate": 9.087480052226899e-06, "loss": 0.0599, "step": 15144 }, { "epoch": 1.6419124024284475, "grad_norm": 4.128875732421875, "learning_rate": 9.086754678659511e-06, "loss": 0.0237, "step": 15145 }, { "epoch": 1.6420208152645275, "grad_norm": 0.4746415317058563, "learning_rate": 9.086029305092123e-06, "loss": 0.0064, "step": 15146 }, { "epoch": 1.642129228100607, "grad_norm": 2.761202812194824, "learning_rate": 9.085303931524735e-06, "loss": 0.0238, "step": 15147 }, { "epoch": 1.642237640936687, "grad_norm": 0.43875789642333984, "learning_rate": 9.084578557957349e-06, "loss": 0.0083, "step": 15148 }, { "epoch": 1.6423460537727665, "grad_norm": 6.34018087387085, "learning_rate": 9.083853184389961e-06, "loss": 0.0564, "step": 15149 }, { "epoch": 1.6424544666088465, "grad_norm": 0.5342543125152588, "learning_rate": 9.083127810822573e-06, "loss": 0.0158, "step": 15150 }, { "epoch": 1.6425628794449263, "grad_norm": 0.4003608226776123, "learning_rate": 9.082402437255187e-06, "loss": 0.0143, "step": 15151 }, { "epoch": 1.642671292281006, "grad_norm": 4.8562164306640625, "learning_rate": 9.081677063687801e-06, "loss": 0.0162, "step": 15152 }, { "epoch": 1.6427797051170858, "grad_norm": 2.4369773864746094, "learning_rate": 9.080951690120413e-06, "loss": 0.0436, "step": 15153 }, { "epoch": 1.6428881179531656, "grad_norm": 0.3995938301086426, "learning_rate": 9.080226316553025e-06, "loss": 0.0213, "step": 15154 }, { "epoch": 1.6429965307892456, "grad_norm": 0.5571122169494629, "learning_rate": 9.07950094298564e-06, "loss": 0.0044, "step": 15155 }, { "epoch": 1.643104943625325, "grad_norm": 3.3440613746643066, "learning_rate": 9.078775569418251e-06, "loss": 0.0344, "step": 15156 }, { "epoch": 1.643213356461405, "grad_norm": 0.2402973175048828, "learning_rate": 9.078050195850863e-06, "loss": 0.005, "step": 15157 }, { "epoch": 1.6433217692974849, "grad_norm": 2.164006471633911, "learning_rate": 9.077324822283476e-06, "loss": 0.0757, "step": 15158 }, { "epoch": 1.6434301821335646, "grad_norm": 0.7419070601463318, "learning_rate": 9.07659944871609e-06, "loss": 0.0067, "step": 15159 }, { "epoch": 1.6435385949696444, "grad_norm": 3.6964995861053467, "learning_rate": 9.075874075148702e-06, "loss": 0.0115, "step": 15160 }, { "epoch": 1.6436470078057241, "grad_norm": 1.614608645439148, "learning_rate": 9.075148701581315e-06, "loss": 0.0115, "step": 15161 }, { "epoch": 1.6437554206418041, "grad_norm": 14.44118595123291, "learning_rate": 9.074423328013928e-06, "loss": 0.0241, "step": 15162 }, { "epoch": 1.6438638334778837, "grad_norm": 3.547405958175659, "learning_rate": 9.073697954446541e-06, "loss": 0.0133, "step": 15163 }, { "epoch": 1.6439722463139637, "grad_norm": 0.19325628876686096, "learning_rate": 9.072972580879154e-06, "loss": 0.0018, "step": 15164 }, { "epoch": 1.6440806591500432, "grad_norm": 3.7973098754882812, "learning_rate": 9.072247207311766e-06, "loss": 0.0359, "step": 15165 }, { "epoch": 1.6441890719861232, "grad_norm": 4.237241268157959, "learning_rate": 9.07152183374438e-06, "loss": 0.021, "step": 15166 }, { "epoch": 1.644297484822203, "grad_norm": 5.362905025482178, "learning_rate": 9.070796460176992e-06, "loss": 0.0516, "step": 15167 }, { "epoch": 1.6444058976582827, "grad_norm": 5.883849620819092, "learning_rate": 9.070071086609604e-06, "loss": 0.0473, "step": 15168 }, { "epoch": 1.6445143104943627, "grad_norm": 3.9835665225982666, "learning_rate": 9.069345713042218e-06, "loss": 0.0368, "step": 15169 }, { "epoch": 1.6446227233304422, "grad_norm": 1.9913963079452515, "learning_rate": 9.068620339474831e-06, "loss": 0.0111, "step": 15170 }, { "epoch": 1.6447311361665222, "grad_norm": 3.4903085231781006, "learning_rate": 9.067894965907444e-06, "loss": 0.0398, "step": 15171 }, { "epoch": 1.6448395490026018, "grad_norm": 4.878653049468994, "learning_rate": 9.067169592340056e-06, "loss": 0.0284, "step": 15172 }, { "epoch": 1.6449479618386817, "grad_norm": 5.299010753631592, "learning_rate": 9.066444218772668e-06, "loss": 0.0589, "step": 15173 }, { "epoch": 1.6450563746747615, "grad_norm": 2.595729351043701, "learning_rate": 9.065718845205282e-06, "loss": 0.0212, "step": 15174 }, { "epoch": 1.6451647875108413, "grad_norm": 7.235758304595947, "learning_rate": 9.064993471637894e-06, "loss": 0.1071, "step": 15175 }, { "epoch": 1.645273200346921, "grad_norm": 1.0751421451568604, "learning_rate": 9.064268098070506e-06, "loss": 0.0041, "step": 15176 }, { "epoch": 1.6453816131830008, "grad_norm": 2.2155075073242188, "learning_rate": 9.06354272450312e-06, "loss": 0.0067, "step": 15177 }, { "epoch": 1.6454900260190808, "grad_norm": 1.6243664026260376, "learning_rate": 9.062817350935734e-06, "loss": 0.0132, "step": 15178 }, { "epoch": 1.6455984388551603, "grad_norm": 2.3775339126586914, "learning_rate": 9.062091977368346e-06, "loss": 0.0124, "step": 15179 }, { "epoch": 1.6457068516912403, "grad_norm": 0.17755165696144104, "learning_rate": 9.061366603800958e-06, "loss": 0.0023, "step": 15180 }, { "epoch": 1.64581526452732, "grad_norm": 1.3830711841583252, "learning_rate": 9.060641230233572e-06, "loss": 0.033, "step": 15181 }, { "epoch": 1.6459236773633998, "grad_norm": 3.5517308712005615, "learning_rate": 9.059915856666184e-06, "loss": 0.0074, "step": 15182 }, { "epoch": 1.6460320901994796, "grad_norm": 3.9147555828094482, "learning_rate": 9.059190483098796e-06, "loss": 0.0132, "step": 15183 }, { "epoch": 1.6461405030355594, "grad_norm": 2.074711561203003, "learning_rate": 9.058465109531408e-06, "loss": 0.0254, "step": 15184 }, { "epoch": 1.6462489158716394, "grad_norm": 6.536477565765381, "learning_rate": 9.057739735964022e-06, "loss": 0.0586, "step": 15185 }, { "epoch": 1.646357328707719, "grad_norm": 0.624811053276062, "learning_rate": 9.057014362396636e-06, "loss": 0.0047, "step": 15186 }, { "epoch": 1.6464657415437989, "grad_norm": 3.0970959663391113, "learning_rate": 9.056288988829248e-06, "loss": 0.0221, "step": 15187 }, { "epoch": 1.6465741543798784, "grad_norm": 2.155362606048584, "learning_rate": 9.05556361526186e-06, "loss": 0.0227, "step": 15188 }, { "epoch": 1.6466825672159584, "grad_norm": 3.4328646659851074, "learning_rate": 9.054838241694474e-06, "loss": 0.0453, "step": 15189 }, { "epoch": 1.6467909800520382, "grad_norm": 0.5634642839431763, "learning_rate": 9.054112868127086e-06, "loss": 0.0214, "step": 15190 }, { "epoch": 1.646899392888118, "grad_norm": 0.44597843289375305, "learning_rate": 9.053387494559698e-06, "loss": 0.0136, "step": 15191 }, { "epoch": 1.6470078057241977, "grad_norm": 0.40385517477989197, "learning_rate": 9.052662120992312e-06, "loss": 0.003, "step": 15192 }, { "epoch": 1.6471162185602775, "grad_norm": 5.570011138916016, "learning_rate": 9.051936747424924e-06, "loss": 0.0265, "step": 15193 }, { "epoch": 1.6472246313963574, "grad_norm": 0.11122237890958786, "learning_rate": 9.051211373857536e-06, "loss": 0.0012, "step": 15194 }, { "epoch": 1.647333044232437, "grad_norm": 1.014851450920105, "learning_rate": 9.05048600029015e-06, "loss": 0.0117, "step": 15195 }, { "epoch": 1.647441457068517, "grad_norm": 2.1714112758636475, "learning_rate": 9.049760626722764e-06, "loss": 0.0321, "step": 15196 }, { "epoch": 1.6475498699045967, "grad_norm": 5.697597503662109, "learning_rate": 9.049035253155376e-06, "loss": 0.0243, "step": 15197 }, { "epoch": 1.6476582827406765, "grad_norm": 1.6977931261062622, "learning_rate": 9.048309879587988e-06, "loss": 0.0297, "step": 15198 }, { "epoch": 1.6477666955767563, "grad_norm": 1.5418094396591187, "learning_rate": 9.0475845060206e-06, "loss": 0.0286, "step": 15199 }, { "epoch": 1.647875108412836, "grad_norm": 0.37908297777175903, "learning_rate": 9.046859132453214e-06, "loss": 0.0025, "step": 15200 }, { "epoch": 1.647983521248916, "grad_norm": 1.3095498085021973, "learning_rate": 9.046133758885827e-06, "loss": 0.02, "step": 15201 }, { "epoch": 1.6480919340849955, "grad_norm": 1.9293938875198364, "learning_rate": 9.045408385318439e-06, "loss": 0.0397, "step": 15202 }, { "epoch": 1.6482003469210755, "grad_norm": 3.0668914318084717, "learning_rate": 9.044683011751053e-06, "loss": 0.034, "step": 15203 }, { "epoch": 1.6483087597571553, "grad_norm": 2.2822329998016357, "learning_rate": 9.043957638183666e-06, "loss": 0.0318, "step": 15204 }, { "epoch": 1.648417172593235, "grad_norm": 0.43194523453712463, "learning_rate": 9.043232264616279e-06, "loss": 0.005, "step": 15205 }, { "epoch": 1.6485255854293148, "grad_norm": 4.249887943267822, "learning_rate": 9.04250689104889e-06, "loss": 0.0231, "step": 15206 }, { "epoch": 1.6486339982653946, "grad_norm": 3.5933420658111572, "learning_rate": 9.041781517481505e-06, "loss": 0.0211, "step": 15207 }, { "epoch": 1.6487424111014746, "grad_norm": 0.2607746720314026, "learning_rate": 9.041056143914117e-06, "loss": 0.0059, "step": 15208 }, { "epoch": 1.648850823937554, "grad_norm": 2.564521074295044, "learning_rate": 9.040330770346729e-06, "loss": 0.0147, "step": 15209 }, { "epoch": 1.648959236773634, "grad_norm": 1.193798303604126, "learning_rate": 9.039605396779341e-06, "loss": 0.0192, "step": 15210 }, { "epoch": 1.6490676496097136, "grad_norm": 0.7613924741744995, "learning_rate": 9.038880023211955e-06, "loss": 0.0068, "step": 15211 }, { "epoch": 1.6491760624457936, "grad_norm": 5.701979637145996, "learning_rate": 9.038154649644569e-06, "loss": 0.0316, "step": 15212 }, { "epoch": 1.6492844752818734, "grad_norm": 1.2366249561309814, "learning_rate": 9.03742927607718e-06, "loss": 0.0339, "step": 15213 }, { "epoch": 1.6493928881179531, "grad_norm": 2.755603551864624, "learning_rate": 9.036703902509793e-06, "loss": 0.0334, "step": 15214 }, { "epoch": 1.649501300954033, "grad_norm": 2.995062828063965, "learning_rate": 9.035978528942407e-06, "loss": 0.0198, "step": 15215 }, { "epoch": 1.6496097137901127, "grad_norm": 0.9950438737869263, "learning_rate": 9.035253155375019e-06, "loss": 0.0025, "step": 15216 }, { "epoch": 1.6497181266261927, "grad_norm": 1.1662960052490234, "learning_rate": 9.034527781807631e-06, "loss": 0.0187, "step": 15217 }, { "epoch": 1.6498265394622722, "grad_norm": 1.6433879137039185, "learning_rate": 9.033802408240245e-06, "loss": 0.0187, "step": 15218 }, { "epoch": 1.6499349522983522, "grad_norm": 0.4177266061306, "learning_rate": 9.033077034672857e-06, "loss": 0.0075, "step": 15219 }, { "epoch": 1.650043365134432, "grad_norm": 4.263974189758301, "learning_rate": 9.03235166110547e-06, "loss": 0.0258, "step": 15220 }, { "epoch": 1.6501517779705117, "grad_norm": 0.6523081660270691, "learning_rate": 9.031626287538083e-06, "loss": 0.0133, "step": 15221 }, { "epoch": 1.6502601908065915, "grad_norm": 3.3512203693389893, "learning_rate": 9.030900913970697e-06, "loss": 0.0465, "step": 15222 }, { "epoch": 1.6503686036426712, "grad_norm": 0.6989319920539856, "learning_rate": 9.030175540403309e-06, "loss": 0.0087, "step": 15223 }, { "epoch": 1.6504770164787512, "grad_norm": 0.6540137529373169, "learning_rate": 9.029450166835921e-06, "loss": 0.0148, "step": 15224 }, { "epoch": 1.6505854293148308, "grad_norm": 0.7764877676963806, "learning_rate": 9.028724793268533e-06, "loss": 0.0121, "step": 15225 }, { "epoch": 1.6506938421509108, "grad_norm": 2.602947950363159, "learning_rate": 9.027999419701147e-06, "loss": 0.0295, "step": 15226 }, { "epoch": 1.6508022549869905, "grad_norm": 2.6663460731506348, "learning_rate": 9.02727404613376e-06, "loss": 0.0345, "step": 15227 }, { "epoch": 1.6509106678230703, "grad_norm": 1.6453355550765991, "learning_rate": 9.026548672566371e-06, "loss": 0.022, "step": 15228 }, { "epoch": 1.65101908065915, "grad_norm": 1.4197635650634766, "learning_rate": 9.025823298998985e-06, "loss": 0.0024, "step": 15229 }, { "epoch": 1.6511274934952298, "grad_norm": 1.4717135429382324, "learning_rate": 9.025097925431599e-06, "loss": 0.0172, "step": 15230 }, { "epoch": 1.6512359063313098, "grad_norm": 0.40369030833244324, "learning_rate": 9.024372551864211e-06, "loss": 0.0131, "step": 15231 }, { "epoch": 1.6513443191673893, "grad_norm": 2.309429883956909, "learning_rate": 9.023647178296823e-06, "loss": 0.0176, "step": 15232 }, { "epoch": 1.6514527320034693, "grad_norm": 5.6566877365112305, "learning_rate": 9.022921804729437e-06, "loss": 0.0841, "step": 15233 }, { "epoch": 1.6515611448395489, "grad_norm": 0.29305389523506165, "learning_rate": 9.02219643116205e-06, "loss": 0.0013, "step": 15234 }, { "epoch": 1.6516695576756288, "grad_norm": 3.563155174255371, "learning_rate": 9.021471057594661e-06, "loss": 0.019, "step": 15235 }, { "epoch": 1.6517779705117086, "grad_norm": 5.6866230964660645, "learning_rate": 9.020745684027275e-06, "loss": 0.0177, "step": 15236 }, { "epoch": 1.6518863833477884, "grad_norm": 6.353486061096191, "learning_rate": 9.020020310459887e-06, "loss": 0.0601, "step": 15237 }, { "epoch": 1.6519947961838681, "grad_norm": 0.3283397853374481, "learning_rate": 9.019294936892501e-06, "loss": 0.0053, "step": 15238 }, { "epoch": 1.652103209019948, "grad_norm": 6.285064697265625, "learning_rate": 9.018569563325113e-06, "loss": 0.0239, "step": 15239 }, { "epoch": 1.6522116218560279, "grad_norm": 1.1064270734786987, "learning_rate": 9.017844189757726e-06, "loss": 0.0039, "step": 15240 }, { "epoch": 1.6523200346921074, "grad_norm": 0.8604133725166321, "learning_rate": 9.01711881619034e-06, "loss": 0.0047, "step": 15241 }, { "epoch": 1.6524284475281874, "grad_norm": 2.4813146591186523, "learning_rate": 9.016393442622952e-06, "loss": 0.0226, "step": 15242 }, { "epoch": 1.6525368603642672, "grad_norm": 1.068954586982727, "learning_rate": 9.015668069055564e-06, "loss": 0.0271, "step": 15243 }, { "epoch": 1.652645273200347, "grad_norm": 5.187743186950684, "learning_rate": 9.014942695488178e-06, "loss": 0.0808, "step": 15244 }, { "epoch": 1.6527536860364267, "grad_norm": 2.380321741104126, "learning_rate": 9.01421732192079e-06, "loss": 0.0136, "step": 15245 }, { "epoch": 1.6528620988725065, "grad_norm": 1.547420859336853, "learning_rate": 9.013491948353402e-06, "loss": 0.0462, "step": 15246 }, { "epoch": 1.6529705117085864, "grad_norm": 3.764968156814575, "learning_rate": 9.012766574786016e-06, "loss": 0.029, "step": 15247 }, { "epoch": 1.653078924544666, "grad_norm": 5.603547096252441, "learning_rate": 9.01204120121863e-06, "loss": 0.0194, "step": 15248 }, { "epoch": 1.653187337380746, "grad_norm": 11.308661460876465, "learning_rate": 9.011315827651242e-06, "loss": 0.013, "step": 15249 }, { "epoch": 1.6532957502168255, "grad_norm": 0.8118690848350525, "learning_rate": 9.010590454083854e-06, "loss": 0.0067, "step": 15250 }, { "epoch": 1.6534041630529055, "grad_norm": 2.197563409805298, "learning_rate": 9.009865080516466e-06, "loss": 0.0326, "step": 15251 }, { "epoch": 1.6535125758889853, "grad_norm": 4.905447959899902, "learning_rate": 9.00913970694908e-06, "loss": 0.0214, "step": 15252 }, { "epoch": 1.653620988725065, "grad_norm": 6.7401652336120605, "learning_rate": 9.008414333381692e-06, "loss": 0.0277, "step": 15253 }, { "epoch": 1.6537294015611448, "grad_norm": 2.701380729675293, "learning_rate": 9.007688959814304e-06, "loss": 0.0199, "step": 15254 }, { "epoch": 1.6538378143972245, "grad_norm": 0.31807032227516174, "learning_rate": 9.006963586246918e-06, "loss": 0.0016, "step": 15255 }, { "epoch": 1.6539462272333045, "grad_norm": 0.08640467375516891, "learning_rate": 9.006238212679532e-06, "loss": 0.0008, "step": 15256 }, { "epoch": 1.654054640069384, "grad_norm": 3.721628427505493, "learning_rate": 9.005512839112144e-06, "loss": 0.0471, "step": 15257 }, { "epoch": 1.654163052905464, "grad_norm": 3.402817964553833, "learning_rate": 9.004787465544756e-06, "loss": 0.0326, "step": 15258 }, { "epoch": 1.6542714657415438, "grad_norm": 1.1718705892562866, "learning_rate": 9.00406209197737e-06, "loss": 0.0084, "step": 15259 }, { "epoch": 1.6543798785776236, "grad_norm": 4.782650947570801, "learning_rate": 9.003336718409982e-06, "loss": 0.0485, "step": 15260 }, { "epoch": 1.6544882914137033, "grad_norm": 1.1300381422042847, "learning_rate": 9.002611344842594e-06, "loss": 0.0316, "step": 15261 }, { "epoch": 1.6545967042497831, "grad_norm": 0.6010677218437195, "learning_rate": 9.001885971275208e-06, "loss": 0.0039, "step": 15262 }, { "epoch": 1.654705117085863, "grad_norm": 4.116256237030029, "learning_rate": 9.00116059770782e-06, "loss": 0.0335, "step": 15263 }, { "epoch": 1.6548135299219426, "grad_norm": 5.0148396492004395, "learning_rate": 9.000435224140434e-06, "loss": 0.0145, "step": 15264 }, { "epoch": 1.6549219427580226, "grad_norm": 0.7231183648109436, "learning_rate": 8.999709850573046e-06, "loss": 0.0163, "step": 15265 }, { "epoch": 1.6550303555941024, "grad_norm": 1.6425065994262695, "learning_rate": 8.998984477005658e-06, "loss": 0.008, "step": 15266 }, { "epoch": 1.6551387684301822, "grad_norm": 1.8447152376174927, "learning_rate": 8.998259103438272e-06, "loss": 0.0249, "step": 15267 }, { "epoch": 1.655247181266262, "grad_norm": 6.051000118255615, "learning_rate": 8.997533729870884e-06, "loss": 0.0201, "step": 15268 }, { "epoch": 1.6553555941023417, "grad_norm": 13.652944564819336, "learning_rate": 8.996808356303496e-06, "loss": 0.0477, "step": 15269 }, { "epoch": 1.6554640069384217, "grad_norm": 0.42058685421943665, "learning_rate": 8.99608298273611e-06, "loss": 0.0091, "step": 15270 }, { "epoch": 1.6555724197745012, "grad_norm": 0.4650594890117645, "learning_rate": 8.995357609168722e-06, "loss": 0.0092, "step": 15271 }, { "epoch": 1.6556808326105812, "grad_norm": 17.595563888549805, "learning_rate": 8.994632235601334e-06, "loss": 0.0407, "step": 15272 }, { "epoch": 1.6557892454466607, "grad_norm": 1.8427406549453735, "learning_rate": 8.993906862033948e-06, "loss": 0.0199, "step": 15273 }, { "epoch": 1.6558976582827407, "grad_norm": 1.645296335220337, "learning_rate": 8.993181488466562e-06, "loss": 0.0141, "step": 15274 }, { "epoch": 1.6560060711188205, "grad_norm": 3.9159886837005615, "learning_rate": 8.992456114899174e-06, "loss": 0.0184, "step": 15275 }, { "epoch": 1.6561144839549002, "grad_norm": 0.487057089805603, "learning_rate": 8.991730741331786e-06, "loss": 0.0017, "step": 15276 }, { "epoch": 1.65622289679098, "grad_norm": 1.602333426475525, "learning_rate": 8.991005367764399e-06, "loss": 0.0296, "step": 15277 }, { "epoch": 1.6563313096270598, "grad_norm": 0.6247276663780212, "learning_rate": 8.990279994197012e-06, "loss": 0.0261, "step": 15278 }, { "epoch": 1.6564397224631398, "grad_norm": 0.5143165588378906, "learning_rate": 8.989554620629625e-06, "loss": 0.0122, "step": 15279 }, { "epoch": 1.6565481352992193, "grad_norm": 1.5785270929336548, "learning_rate": 8.988829247062237e-06, "loss": 0.0375, "step": 15280 }, { "epoch": 1.6566565481352993, "grad_norm": 12.021574020385742, "learning_rate": 8.98810387349485e-06, "loss": 0.0245, "step": 15281 }, { "epoch": 1.656764960971379, "grad_norm": 5.538603782653809, "learning_rate": 8.987378499927464e-06, "loss": 0.036, "step": 15282 }, { "epoch": 1.6568733738074588, "grad_norm": 0.386054664850235, "learning_rate": 8.986653126360077e-06, "loss": 0.0038, "step": 15283 }, { "epoch": 1.6569817866435386, "grad_norm": 0.862480103969574, "learning_rate": 8.985927752792689e-06, "loss": 0.0208, "step": 15284 }, { "epoch": 1.6570901994796183, "grad_norm": 4.179012298583984, "learning_rate": 8.985202379225303e-06, "loss": 0.0458, "step": 15285 }, { "epoch": 1.6571986123156983, "grad_norm": 1.146348476409912, "learning_rate": 8.984477005657915e-06, "loss": 0.0127, "step": 15286 }, { "epoch": 1.6573070251517779, "grad_norm": 0.2966979146003723, "learning_rate": 8.983751632090527e-06, "loss": 0.0023, "step": 15287 }, { "epoch": 1.6574154379878578, "grad_norm": 6.012899875640869, "learning_rate": 8.98302625852314e-06, "loss": 0.0608, "step": 15288 }, { "epoch": 1.6575238508239376, "grad_norm": 0.5600704550743103, "learning_rate": 8.982300884955753e-06, "loss": 0.0178, "step": 15289 }, { "epoch": 1.6576322636600174, "grad_norm": 0.2144586592912674, "learning_rate": 8.981575511388367e-06, "loss": 0.0007, "step": 15290 }, { "epoch": 1.6577406764960971, "grad_norm": 0.3484652042388916, "learning_rate": 8.980850137820979e-06, "loss": 0.008, "step": 15291 }, { "epoch": 1.657849089332177, "grad_norm": 0.2801911532878876, "learning_rate": 8.980124764253591e-06, "loss": 0.0019, "step": 15292 }, { "epoch": 1.6579575021682569, "grad_norm": 0.3200439214706421, "learning_rate": 8.979399390686205e-06, "loss": 0.0083, "step": 15293 }, { "epoch": 1.6580659150043364, "grad_norm": 0.1675795018672943, "learning_rate": 8.978674017118817e-06, "loss": 0.0012, "step": 15294 }, { "epoch": 1.6581743278404164, "grad_norm": 3.8517708778381348, "learning_rate": 8.977948643551429e-06, "loss": 0.0173, "step": 15295 }, { "epoch": 1.658282740676496, "grad_norm": 11.681336402893066, "learning_rate": 8.977223269984043e-06, "loss": 0.0171, "step": 15296 }, { "epoch": 1.658391153512576, "grad_norm": 5.151548385620117, "learning_rate": 8.976497896416655e-06, "loss": 0.0318, "step": 15297 }, { "epoch": 1.6584995663486557, "grad_norm": 0.7075296640396118, "learning_rate": 8.975772522849267e-06, "loss": 0.0044, "step": 15298 }, { "epoch": 1.6586079791847355, "grad_norm": 0.4613589346408844, "learning_rate": 8.975047149281881e-06, "loss": 0.012, "step": 15299 }, { "epoch": 1.6587163920208152, "grad_norm": 3.1105055809020996, "learning_rate": 8.974321775714495e-06, "loss": 0.0773, "step": 15300 }, { "epoch": 1.658824804856895, "grad_norm": 2.796860933303833, "learning_rate": 8.973596402147107e-06, "loss": 0.0362, "step": 15301 }, { "epoch": 1.658933217692975, "grad_norm": 0.4189758896827698, "learning_rate": 8.972871028579719e-06, "loss": 0.0114, "step": 15302 }, { "epoch": 1.6590416305290545, "grad_norm": 1.1924593448638916, "learning_rate": 8.972145655012331e-06, "loss": 0.0042, "step": 15303 }, { "epoch": 1.6591500433651345, "grad_norm": 1.0698264837265015, "learning_rate": 8.971420281444945e-06, "loss": 0.0042, "step": 15304 }, { "epoch": 1.6592584562012143, "grad_norm": 0.44352155923843384, "learning_rate": 8.970694907877557e-06, "loss": 0.014, "step": 15305 }, { "epoch": 1.659366869037294, "grad_norm": 8.890639305114746, "learning_rate": 8.96996953431017e-06, "loss": 0.0537, "step": 15306 }, { "epoch": 1.6594752818733738, "grad_norm": 0.610059916973114, "learning_rate": 8.969244160742783e-06, "loss": 0.0015, "step": 15307 }, { "epoch": 1.6595836947094535, "grad_norm": 0.6707718968391418, "learning_rate": 8.968518787175397e-06, "loss": 0.0166, "step": 15308 }, { "epoch": 1.6596921075455335, "grad_norm": 0.027420639991760254, "learning_rate": 8.96779341360801e-06, "loss": 0.0002, "step": 15309 }, { "epoch": 1.659800520381613, "grad_norm": 0.6525850296020508, "learning_rate": 8.967068040040621e-06, "loss": 0.0149, "step": 15310 }, { "epoch": 1.659908933217693, "grad_norm": 0.8987362384796143, "learning_rate": 8.966342666473235e-06, "loss": 0.0367, "step": 15311 }, { "epoch": 1.6600173460537726, "grad_norm": 1.2603747844696045, "learning_rate": 8.965617292905847e-06, "loss": 0.0173, "step": 15312 }, { "epoch": 1.6601257588898526, "grad_norm": 3.852091073989868, "learning_rate": 8.96489191933846e-06, "loss": 0.0259, "step": 15313 }, { "epoch": 1.6602341717259324, "grad_norm": 0.2496701180934906, "learning_rate": 8.964166545771073e-06, "loss": 0.0017, "step": 15314 }, { "epoch": 1.6603425845620121, "grad_norm": 2.300368309020996, "learning_rate": 8.963441172203685e-06, "loss": 0.0192, "step": 15315 }, { "epoch": 1.660450997398092, "grad_norm": 4.62888240814209, "learning_rate": 8.9627157986363e-06, "loss": 0.0599, "step": 15316 }, { "epoch": 1.6605594102341716, "grad_norm": 3.9790236949920654, "learning_rate": 8.961990425068911e-06, "loss": 0.0637, "step": 15317 }, { "epoch": 1.6606678230702516, "grad_norm": 2.8800222873687744, "learning_rate": 8.961265051501524e-06, "loss": 0.0302, "step": 15318 }, { "epoch": 1.6607762359063312, "grad_norm": 4.075161457061768, "learning_rate": 8.960539677934137e-06, "loss": 0.0163, "step": 15319 }, { "epoch": 1.6608846487424112, "grad_norm": 1.1014915704727173, "learning_rate": 8.95981430436675e-06, "loss": 0.0279, "step": 15320 }, { "epoch": 1.660993061578491, "grad_norm": 3.2447848320007324, "learning_rate": 8.959088930799362e-06, "loss": 0.0085, "step": 15321 }, { "epoch": 1.6611014744145707, "grad_norm": 2.9973349571228027, "learning_rate": 8.958363557231976e-06, "loss": 0.0249, "step": 15322 }, { "epoch": 1.6612098872506504, "grad_norm": 12.18868637084961, "learning_rate": 8.957638183664588e-06, "loss": 0.0726, "step": 15323 }, { "epoch": 1.6613183000867302, "grad_norm": 7.9551920890808105, "learning_rate": 8.9569128100972e-06, "loss": 0.0769, "step": 15324 }, { "epoch": 1.6614267129228102, "grad_norm": 4.454019069671631, "learning_rate": 8.956187436529814e-06, "loss": 0.0539, "step": 15325 }, { "epoch": 1.6615351257588897, "grad_norm": 0.3487131595611572, "learning_rate": 8.955462062962428e-06, "loss": 0.023, "step": 15326 }, { "epoch": 1.6616435385949697, "grad_norm": 7.995986461639404, "learning_rate": 8.95473668939504e-06, "loss": 0.0283, "step": 15327 }, { "epoch": 1.6617519514310495, "grad_norm": 5.162187099456787, "learning_rate": 8.954011315827652e-06, "loss": 0.1004, "step": 15328 }, { "epoch": 1.6618603642671292, "grad_norm": 2.5195775032043457, "learning_rate": 8.953285942260264e-06, "loss": 0.0475, "step": 15329 }, { "epoch": 1.661968777103209, "grad_norm": 0.9965161681175232, "learning_rate": 8.952560568692878e-06, "loss": 0.0061, "step": 15330 }, { "epoch": 1.6620771899392888, "grad_norm": 1.554923415184021, "learning_rate": 8.95183519512549e-06, "loss": 0.0071, "step": 15331 }, { "epoch": 1.6621856027753688, "grad_norm": 3.4743218421936035, "learning_rate": 8.951109821558102e-06, "loss": 0.0459, "step": 15332 }, { "epoch": 1.6622940156114483, "grad_norm": 0.1810048520565033, "learning_rate": 8.950384447990716e-06, "loss": 0.0109, "step": 15333 }, { "epoch": 1.6624024284475283, "grad_norm": 0.45892420411109924, "learning_rate": 8.94965907442333e-06, "loss": 0.0107, "step": 15334 }, { "epoch": 1.6625108412836078, "grad_norm": 1.5950088500976562, "learning_rate": 8.948933700855942e-06, "loss": 0.0123, "step": 15335 }, { "epoch": 1.6626192541196878, "grad_norm": 2.8802266120910645, "learning_rate": 8.948208327288554e-06, "loss": 0.0177, "step": 15336 }, { "epoch": 1.6627276669557676, "grad_norm": 4.2861328125, "learning_rate": 8.947482953721168e-06, "loss": 0.049, "step": 15337 }, { "epoch": 1.6628360797918473, "grad_norm": 0.5148818492889404, "learning_rate": 8.94675758015378e-06, "loss": 0.01, "step": 15338 }, { "epoch": 1.662944492627927, "grad_norm": 1.156944990158081, "learning_rate": 8.946032206586392e-06, "loss": 0.0136, "step": 15339 }, { "epoch": 1.6630529054640069, "grad_norm": 2.0413284301757812, "learning_rate": 8.945306833019006e-06, "loss": 0.0479, "step": 15340 }, { "epoch": 1.6631613183000868, "grad_norm": 2.501765489578247, "learning_rate": 8.944581459451618e-06, "loss": 0.0467, "step": 15341 }, { "epoch": 1.6632697311361664, "grad_norm": 4.129952430725098, "learning_rate": 8.943856085884232e-06, "loss": 0.0083, "step": 15342 }, { "epoch": 1.6633781439722464, "grad_norm": 2.4151511192321777, "learning_rate": 8.943130712316844e-06, "loss": 0.0161, "step": 15343 }, { "epoch": 1.6634865568083261, "grad_norm": 0.012417992576956749, "learning_rate": 8.942405338749456e-06, "loss": 0.0001, "step": 15344 }, { "epoch": 1.663594969644406, "grad_norm": 2.183434009552002, "learning_rate": 8.94167996518207e-06, "loss": 0.0264, "step": 15345 }, { "epoch": 1.6637033824804857, "grad_norm": 2.382978916168213, "learning_rate": 8.940954591614682e-06, "loss": 0.0192, "step": 15346 }, { "epoch": 1.6638117953165654, "grad_norm": 0.30705341696739197, "learning_rate": 8.940229218047294e-06, "loss": 0.0015, "step": 15347 }, { "epoch": 1.6639202081526454, "grad_norm": 0.6386613845825195, "learning_rate": 8.939503844479908e-06, "loss": 0.0024, "step": 15348 }, { "epoch": 1.664028620988725, "grad_norm": 0.03944487124681473, "learning_rate": 8.93877847091252e-06, "loss": 0.0003, "step": 15349 }, { "epoch": 1.664137033824805, "grad_norm": 1.158856749534607, "learning_rate": 8.938053097345133e-06, "loss": 0.0033, "step": 15350 }, { "epoch": 1.6642454466608847, "grad_norm": 0.3399236500263214, "learning_rate": 8.937327723777746e-06, "loss": 0.0078, "step": 15351 }, { "epoch": 1.6643538594969645, "grad_norm": 0.5711873769760132, "learning_rate": 8.93660235021036e-06, "loss": 0.0175, "step": 15352 }, { "epoch": 1.6644622723330442, "grad_norm": 0.7408980131149292, "learning_rate": 8.935876976642972e-06, "loss": 0.0164, "step": 15353 }, { "epoch": 1.664570685169124, "grad_norm": 3.667860269546509, "learning_rate": 8.935151603075584e-06, "loss": 0.0049, "step": 15354 }, { "epoch": 1.664679098005204, "grad_norm": 1.459122657775879, "learning_rate": 8.934426229508197e-06, "loss": 0.0173, "step": 15355 }, { "epoch": 1.6647875108412835, "grad_norm": 2.0005524158477783, "learning_rate": 8.93370085594081e-06, "loss": 0.0272, "step": 15356 }, { "epoch": 1.6648959236773635, "grad_norm": 2.704219102859497, "learning_rate": 8.932975482373423e-06, "loss": 0.0059, "step": 15357 }, { "epoch": 1.665004336513443, "grad_norm": 0.22198282182216644, "learning_rate": 8.932250108806035e-06, "loss": 0.0022, "step": 15358 }, { "epoch": 1.665112749349523, "grad_norm": 1.9475640058517456, "learning_rate": 8.931524735238649e-06, "loss": 0.0125, "step": 15359 }, { "epoch": 1.6652211621856028, "grad_norm": 0.2892754375934601, "learning_rate": 8.930799361671262e-06, "loss": 0.0019, "step": 15360 }, { "epoch": 1.6653295750216826, "grad_norm": 0.6898569464683533, "learning_rate": 8.930073988103875e-06, "loss": 0.0109, "step": 15361 }, { "epoch": 1.6654379878577623, "grad_norm": 1.4949002265930176, "learning_rate": 8.929348614536487e-06, "loss": 0.0035, "step": 15362 }, { "epoch": 1.665546400693842, "grad_norm": 3.210584878921509, "learning_rate": 8.9286232409691e-06, "loss": 0.0491, "step": 15363 }, { "epoch": 1.665654813529922, "grad_norm": 3.2877681255340576, "learning_rate": 8.927897867401713e-06, "loss": 0.037, "step": 15364 }, { "epoch": 1.6657632263660016, "grad_norm": 1.6723976135253906, "learning_rate": 8.927172493834325e-06, "loss": 0.0227, "step": 15365 }, { "epoch": 1.6658716392020816, "grad_norm": 1.8183788061141968, "learning_rate": 8.926447120266939e-06, "loss": 0.0343, "step": 15366 }, { "epoch": 1.6659800520381614, "grad_norm": 2.0046815872192383, "learning_rate": 8.92572174669955e-06, "loss": 0.0107, "step": 15367 }, { "epoch": 1.6660884648742411, "grad_norm": 3.495856523513794, "learning_rate": 8.924996373132165e-06, "loss": 0.0311, "step": 15368 }, { "epoch": 1.6661968777103209, "grad_norm": 0.13449715077877045, "learning_rate": 8.924270999564777e-06, "loss": 0.0015, "step": 15369 }, { "epoch": 1.6663052905464006, "grad_norm": 0.6616106033325195, "learning_rate": 8.923545625997389e-06, "loss": 0.0214, "step": 15370 }, { "epoch": 1.6664137033824806, "grad_norm": 0.38499557971954346, "learning_rate": 8.922820252430003e-06, "loss": 0.0102, "step": 15371 }, { "epoch": 1.6665221162185602, "grad_norm": 0.4415135383605957, "learning_rate": 8.922094878862615e-06, "loss": 0.0109, "step": 15372 }, { "epoch": 1.6666305290546402, "grad_norm": 1.176956057548523, "learning_rate": 8.921369505295227e-06, "loss": 0.0066, "step": 15373 }, { "epoch": 1.66673894189072, "grad_norm": 2.6597325801849365, "learning_rate": 8.920644131727841e-06, "loss": 0.0182, "step": 15374 }, { "epoch": 1.6668473547267997, "grad_norm": 1.2847782373428345, "learning_rate": 8.919918758160453e-06, "loss": 0.0109, "step": 15375 }, { "epoch": 1.6669557675628794, "grad_norm": 8.531048774719238, "learning_rate": 8.919193384593065e-06, "loss": 0.0547, "step": 15376 }, { "epoch": 1.6670641803989592, "grad_norm": 0.9740511178970337, "learning_rate": 8.918468011025679e-06, "loss": 0.0102, "step": 15377 }, { "epoch": 1.6671725932350392, "grad_norm": 0.642761766910553, "learning_rate": 8.917742637458293e-06, "loss": 0.0154, "step": 15378 }, { "epoch": 1.6672810060711187, "grad_norm": 0.5459675192832947, "learning_rate": 8.917017263890905e-06, "loss": 0.0181, "step": 15379 }, { "epoch": 1.6673894189071987, "grad_norm": 0.05231787636876106, "learning_rate": 8.916291890323517e-06, "loss": 0.0003, "step": 15380 }, { "epoch": 1.6674978317432783, "grad_norm": 6.451568603515625, "learning_rate": 8.91556651675613e-06, "loss": 0.0707, "step": 15381 }, { "epoch": 1.6676062445793582, "grad_norm": 0.47113320231437683, "learning_rate": 8.914841143188743e-06, "loss": 0.0092, "step": 15382 }, { "epoch": 1.667714657415438, "grad_norm": 2.842968225479126, "learning_rate": 8.914115769621355e-06, "loss": 0.0155, "step": 15383 }, { "epoch": 1.6678230702515178, "grad_norm": 0.7546283602714539, "learning_rate": 8.913390396053967e-06, "loss": 0.0116, "step": 15384 }, { "epoch": 1.6679314830875975, "grad_norm": 0.39232805371284485, "learning_rate": 8.912665022486581e-06, "loss": 0.0016, "step": 15385 }, { "epoch": 1.6680398959236773, "grad_norm": 4.5329365730285645, "learning_rate": 8.911939648919195e-06, "loss": 0.0385, "step": 15386 }, { "epoch": 1.6681483087597573, "grad_norm": 0.04972190037369728, "learning_rate": 8.911214275351807e-06, "loss": 0.0003, "step": 15387 }, { "epoch": 1.6682567215958368, "grad_norm": 4.426724433898926, "learning_rate": 8.91048890178442e-06, "loss": 0.0529, "step": 15388 }, { "epoch": 1.6683651344319168, "grad_norm": 1.9920072555541992, "learning_rate": 8.909763528217033e-06, "loss": 0.0187, "step": 15389 }, { "epoch": 1.6684735472679966, "grad_norm": 2.5979604721069336, "learning_rate": 8.909038154649645e-06, "loss": 0.0523, "step": 15390 }, { "epoch": 1.6685819601040763, "grad_norm": 0.40176165103912354, "learning_rate": 8.908312781082257e-06, "loss": 0.0278, "step": 15391 }, { "epoch": 1.668690372940156, "grad_norm": 8.901411056518555, "learning_rate": 8.907587407514871e-06, "loss": 0.0284, "step": 15392 }, { "epoch": 1.6687987857762359, "grad_norm": 0.4122732877731323, "learning_rate": 8.906862033947483e-06, "loss": 0.0153, "step": 15393 }, { "epoch": 1.6689071986123158, "grad_norm": 2.520632743835449, "learning_rate": 8.906136660380097e-06, "loss": 0.0032, "step": 15394 }, { "epoch": 1.6690156114483954, "grad_norm": 1.4173754453659058, "learning_rate": 8.90541128681271e-06, "loss": 0.006, "step": 15395 }, { "epoch": 1.6691240242844754, "grad_norm": 5.832826614379883, "learning_rate": 8.904685913245322e-06, "loss": 0.1206, "step": 15396 }, { "epoch": 1.669232437120555, "grad_norm": 1.9173858165740967, "learning_rate": 8.903960539677935e-06, "loss": 0.0516, "step": 15397 }, { "epoch": 1.669340849956635, "grad_norm": 1.2384545803070068, "learning_rate": 8.903235166110548e-06, "loss": 0.0168, "step": 15398 }, { "epoch": 1.6694492627927147, "grad_norm": 0.13738732039928436, "learning_rate": 8.90250979254316e-06, "loss": 0.0006, "step": 15399 }, { "epoch": 1.6695576756287944, "grad_norm": 2.1785788536071777, "learning_rate": 8.901784418975774e-06, "loss": 0.0249, "step": 15400 }, { "epoch": 1.6696660884648742, "grad_norm": 2.372044563293457, "learning_rate": 8.901059045408386e-06, "loss": 0.0468, "step": 15401 }, { "epoch": 1.669774501300954, "grad_norm": 1.9754221439361572, "learning_rate": 8.900333671841e-06, "loss": 0.0203, "step": 15402 }, { "epoch": 1.669882914137034, "grad_norm": 0.6148597598075867, "learning_rate": 8.899608298273612e-06, "loss": 0.0191, "step": 15403 }, { "epoch": 1.6699913269731135, "grad_norm": 0.5111438035964966, "learning_rate": 8.898882924706226e-06, "loss": 0.0108, "step": 15404 }, { "epoch": 1.6700997398091935, "grad_norm": 0.4468175768852234, "learning_rate": 8.898157551138838e-06, "loss": 0.0013, "step": 15405 }, { "epoch": 1.6702081526452732, "grad_norm": 1.6146175861358643, "learning_rate": 8.89743217757145e-06, "loss": 0.0258, "step": 15406 }, { "epoch": 1.670316565481353, "grad_norm": 2.373854637145996, "learning_rate": 8.896706804004062e-06, "loss": 0.0205, "step": 15407 }, { "epoch": 1.6704249783174328, "grad_norm": 6.083525657653809, "learning_rate": 8.895981430436676e-06, "loss": 0.0496, "step": 15408 }, { "epoch": 1.6705333911535125, "grad_norm": 7.661657810211182, "learning_rate": 8.895256056869288e-06, "loss": 0.029, "step": 15409 }, { "epoch": 1.6706418039895925, "grad_norm": 2.3917787075042725, "learning_rate": 8.8945306833019e-06, "loss": 0.0478, "step": 15410 }, { "epoch": 1.670750216825672, "grad_norm": 0.3990125060081482, "learning_rate": 8.893805309734514e-06, "loss": 0.0149, "step": 15411 }, { "epoch": 1.670858629661752, "grad_norm": 0.7036777138710022, "learning_rate": 8.893079936167128e-06, "loss": 0.0311, "step": 15412 }, { "epoch": 1.6709670424978318, "grad_norm": 0.368774950504303, "learning_rate": 8.89235456259974e-06, "loss": 0.0132, "step": 15413 }, { "epoch": 1.6710754553339116, "grad_norm": 0.4221065640449524, "learning_rate": 8.891629189032352e-06, "loss": 0.0227, "step": 15414 }, { "epoch": 1.6711838681699913, "grad_norm": 1.9234267473220825, "learning_rate": 8.890903815464966e-06, "loss": 0.0068, "step": 15415 }, { "epoch": 1.671292281006071, "grad_norm": 1.1821413040161133, "learning_rate": 8.890178441897578e-06, "loss": 0.0084, "step": 15416 }, { "epoch": 1.671400693842151, "grad_norm": 0.7376672029495239, "learning_rate": 8.88945306833019e-06, "loss": 0.023, "step": 15417 }, { "epoch": 1.6715091066782306, "grad_norm": 3.1993582248687744, "learning_rate": 8.888727694762804e-06, "loss": 0.0191, "step": 15418 }, { "epoch": 1.6716175195143106, "grad_norm": 3.391974687576294, "learning_rate": 8.888002321195416e-06, "loss": 0.0201, "step": 15419 }, { "epoch": 1.6717259323503901, "grad_norm": 1.6826082468032837, "learning_rate": 8.88727694762803e-06, "loss": 0.0277, "step": 15420 }, { "epoch": 1.6718343451864701, "grad_norm": 2.053011894226074, "learning_rate": 8.886551574060642e-06, "loss": 0.0388, "step": 15421 }, { "epoch": 1.6719427580225499, "grad_norm": 5.537627220153809, "learning_rate": 8.885826200493254e-06, "loss": 0.034, "step": 15422 }, { "epoch": 1.6720511708586296, "grad_norm": 2.3267900943756104, "learning_rate": 8.885100826925868e-06, "loss": 0.0631, "step": 15423 }, { "epoch": 1.6721595836947094, "grad_norm": 2.699954032897949, "learning_rate": 8.88437545335848e-06, "loss": 0.0127, "step": 15424 }, { "epoch": 1.6722679965307892, "grad_norm": 2.5441842079162598, "learning_rate": 8.883650079791092e-06, "loss": 0.019, "step": 15425 }, { "epoch": 1.6723764093668692, "grad_norm": 1.402482271194458, "learning_rate": 8.882924706223706e-06, "loss": 0.0415, "step": 15426 }, { "epoch": 1.6724848222029487, "grad_norm": 0.4367145597934723, "learning_rate": 8.882199332656318e-06, "loss": 0.0124, "step": 15427 }, { "epoch": 1.6725932350390287, "grad_norm": 1.1005420684814453, "learning_rate": 8.881473959088932e-06, "loss": 0.0267, "step": 15428 }, { "epoch": 1.6727016478751084, "grad_norm": 5.058130741119385, "learning_rate": 8.880748585521544e-06, "loss": 0.0475, "step": 15429 }, { "epoch": 1.6728100607111882, "grad_norm": 5.428210735321045, "learning_rate": 8.880023211954158e-06, "loss": 0.0356, "step": 15430 }, { "epoch": 1.672918473547268, "grad_norm": 3.1267337799072266, "learning_rate": 8.87929783838677e-06, "loss": 0.0059, "step": 15431 }, { "epoch": 1.6730268863833477, "grad_norm": 1.6633797883987427, "learning_rate": 8.878572464819382e-06, "loss": 0.0121, "step": 15432 }, { "epoch": 1.6731352992194277, "grad_norm": 1.8155196905136108, "learning_rate": 8.877847091251995e-06, "loss": 0.0358, "step": 15433 }, { "epoch": 1.6732437120555073, "grad_norm": 2.360666275024414, "learning_rate": 8.877121717684608e-06, "loss": 0.0282, "step": 15434 }, { "epoch": 1.6733521248915872, "grad_norm": 2.832165479660034, "learning_rate": 8.87639634411722e-06, "loss": 0.0111, "step": 15435 }, { "epoch": 1.673460537727667, "grad_norm": 3.522404193878174, "learning_rate": 8.875670970549833e-06, "loss": 0.0142, "step": 15436 }, { "epoch": 1.6735689505637468, "grad_norm": 2.2835378646850586, "learning_rate": 8.874945596982447e-06, "loss": 0.03, "step": 15437 }, { "epoch": 1.6736773633998265, "grad_norm": 1.4911556243896484, "learning_rate": 8.87422022341506e-06, "loss": 0.0063, "step": 15438 }, { "epoch": 1.6737857762359063, "grad_norm": 0.41800764203071594, "learning_rate": 8.873494849847673e-06, "loss": 0.0019, "step": 15439 }, { "epoch": 1.6738941890719863, "grad_norm": 2.648254632949829, "learning_rate": 8.872769476280285e-06, "loss": 0.0362, "step": 15440 }, { "epoch": 1.6740026019080658, "grad_norm": 4.524896144866943, "learning_rate": 8.872044102712899e-06, "loss": 0.0327, "step": 15441 }, { "epoch": 1.6741110147441458, "grad_norm": 0.3075612187385559, "learning_rate": 8.87131872914551e-06, "loss": 0.0022, "step": 15442 }, { "epoch": 1.6742194275802254, "grad_norm": 1.0905908346176147, "learning_rate": 8.870593355578123e-06, "loss": 0.042, "step": 15443 }, { "epoch": 1.6743278404163053, "grad_norm": 0.10415586084127426, "learning_rate": 8.869867982010737e-06, "loss": 0.0014, "step": 15444 }, { "epoch": 1.674436253252385, "grad_norm": 0.3080449104309082, "learning_rate": 8.869142608443349e-06, "loss": 0.0113, "step": 15445 }, { "epoch": 1.6745446660884649, "grad_norm": 1.0862183570861816, "learning_rate": 8.868417234875963e-06, "loss": 0.0256, "step": 15446 }, { "epoch": 1.6746530789245446, "grad_norm": 1.8726569414138794, "learning_rate": 8.867691861308575e-06, "loss": 0.0316, "step": 15447 }, { "epoch": 1.6747614917606244, "grad_norm": 0.7821620106697083, "learning_rate": 8.866966487741187e-06, "loss": 0.0074, "step": 15448 }, { "epoch": 1.6748699045967044, "grad_norm": 2.1908555030822754, "learning_rate": 8.8662411141738e-06, "loss": 0.0285, "step": 15449 }, { "epoch": 1.674978317432784, "grad_norm": 0.6234317421913147, "learning_rate": 8.865515740606413e-06, "loss": 0.0275, "step": 15450 }, { "epoch": 1.675086730268864, "grad_norm": 4.634093761444092, "learning_rate": 8.864790367039025e-06, "loss": 0.0338, "step": 15451 }, { "epoch": 1.6751951431049437, "grad_norm": 5.3099212646484375, "learning_rate": 8.864064993471639e-06, "loss": 0.0531, "step": 15452 }, { "epoch": 1.6753035559410234, "grad_norm": 4.5259623527526855, "learning_rate": 8.863339619904251e-06, "loss": 0.0332, "step": 15453 }, { "epoch": 1.6754119687771032, "grad_norm": 2.6458096504211426, "learning_rate": 8.862614246336865e-06, "loss": 0.0583, "step": 15454 }, { "epoch": 1.675520381613183, "grad_norm": 5.815067291259766, "learning_rate": 8.861888872769477e-06, "loss": 0.0212, "step": 15455 }, { "epoch": 1.675628794449263, "grad_norm": 0.11736686527729034, "learning_rate": 8.861163499202091e-06, "loss": 0.0014, "step": 15456 }, { "epoch": 1.6757372072853425, "grad_norm": 0.17477352917194366, "learning_rate": 8.860438125634703e-06, "loss": 0.002, "step": 15457 }, { "epoch": 1.6758456201214225, "grad_norm": 0.1823408454656601, "learning_rate": 8.859712752067315e-06, "loss": 0.0012, "step": 15458 }, { "epoch": 1.675954032957502, "grad_norm": 2.4737608432769775, "learning_rate": 8.858987378499927e-06, "loss": 0.0138, "step": 15459 }, { "epoch": 1.676062445793582, "grad_norm": 0.3162197172641754, "learning_rate": 8.858262004932541e-06, "loss": 0.0063, "step": 15460 }, { "epoch": 1.6761708586296618, "grad_norm": 0.28633546829223633, "learning_rate": 8.857536631365153e-06, "loss": 0.006, "step": 15461 }, { "epoch": 1.6762792714657415, "grad_norm": 0.58266282081604, "learning_rate": 8.856811257797765e-06, "loss": 0.0215, "step": 15462 }, { "epoch": 1.6763876843018215, "grad_norm": 0.28829601407051086, "learning_rate": 8.85608588423038e-06, "loss": 0.0062, "step": 15463 }, { "epoch": 1.676496097137901, "grad_norm": 5.004398345947266, "learning_rate": 8.855360510662993e-06, "loss": 0.0176, "step": 15464 }, { "epoch": 1.676604509973981, "grad_norm": 0.2995142936706543, "learning_rate": 8.854635137095605e-06, "loss": 0.0064, "step": 15465 }, { "epoch": 1.6767129228100606, "grad_norm": 0.4138076901435852, "learning_rate": 8.853909763528217e-06, "loss": 0.0129, "step": 15466 }, { "epoch": 1.6768213356461406, "grad_norm": 1.5873974561691284, "learning_rate": 8.853184389960831e-06, "loss": 0.0159, "step": 15467 }, { "epoch": 1.6769297484822203, "grad_norm": 1.0636690855026245, "learning_rate": 8.852459016393443e-06, "loss": 0.0435, "step": 15468 }, { "epoch": 1.6770381613183, "grad_norm": 1.9234431982040405, "learning_rate": 8.851733642826055e-06, "loss": 0.0143, "step": 15469 }, { "epoch": 1.6771465741543798, "grad_norm": 0.71645587682724, "learning_rate": 8.85100826925867e-06, "loss": 0.019, "step": 15470 }, { "epoch": 1.6772549869904596, "grad_norm": 1.3235446214675903, "learning_rate": 8.850282895691281e-06, "loss": 0.0394, "step": 15471 }, { "epoch": 1.6773633998265396, "grad_norm": 4.6610589027404785, "learning_rate": 8.849557522123895e-06, "loss": 0.0275, "step": 15472 }, { "epoch": 1.6774718126626191, "grad_norm": 2.282548189163208, "learning_rate": 8.848832148556507e-06, "loss": 0.0237, "step": 15473 }, { "epoch": 1.6775802254986991, "grad_norm": 0.7250590324401855, "learning_rate": 8.84810677498912e-06, "loss": 0.0195, "step": 15474 }, { "epoch": 1.6776886383347789, "grad_norm": 1.6438024044036865, "learning_rate": 8.847381401421733e-06, "loss": 0.014, "step": 15475 }, { "epoch": 1.6777970511708586, "grad_norm": 3.523895740509033, "learning_rate": 8.846656027854346e-06, "loss": 0.0197, "step": 15476 }, { "epoch": 1.6779054640069384, "grad_norm": 0.9232774376869202, "learning_rate": 8.845930654286958e-06, "loss": 0.0038, "step": 15477 }, { "epoch": 1.6780138768430182, "grad_norm": 1.7708165645599365, "learning_rate": 8.845205280719572e-06, "loss": 0.0156, "step": 15478 }, { "epoch": 1.6781222896790982, "grad_norm": 3.189880609512329, "learning_rate": 8.844479907152184e-06, "loss": 0.062, "step": 15479 }, { "epoch": 1.6782307025151777, "grad_norm": 4.989212989807129, "learning_rate": 8.843754533584798e-06, "loss": 0.02, "step": 15480 }, { "epoch": 1.6783391153512577, "grad_norm": 1.8204680681228638, "learning_rate": 8.84302916001741e-06, "loss": 0.0381, "step": 15481 }, { "epoch": 1.6784475281873372, "grad_norm": 1.3293540477752686, "learning_rate": 8.842303786450024e-06, "loss": 0.004, "step": 15482 }, { "epoch": 1.6785559410234172, "grad_norm": 0.3941785991191864, "learning_rate": 8.841578412882636e-06, "loss": 0.0218, "step": 15483 }, { "epoch": 1.678664353859497, "grad_norm": 5.440786361694336, "learning_rate": 8.840853039315248e-06, "loss": 0.0467, "step": 15484 }, { "epoch": 1.6787727666955767, "grad_norm": 0.3047320544719696, "learning_rate": 8.84012766574786e-06, "loss": 0.0063, "step": 15485 }, { "epoch": 1.6788811795316565, "grad_norm": 1.325002670288086, "learning_rate": 8.839402292180474e-06, "loss": 0.0192, "step": 15486 }, { "epoch": 1.6789895923677363, "grad_norm": 3.2195239067077637, "learning_rate": 8.838676918613086e-06, "loss": 0.0575, "step": 15487 }, { "epoch": 1.6790980052038162, "grad_norm": 2.43921160697937, "learning_rate": 8.837951545045698e-06, "loss": 0.0388, "step": 15488 }, { "epoch": 1.6792064180398958, "grad_norm": 1.9784729480743408, "learning_rate": 8.837226171478312e-06, "loss": 0.0254, "step": 15489 }, { "epoch": 1.6793148308759758, "grad_norm": 0.36299988627433777, "learning_rate": 8.836500797910926e-06, "loss": 0.0202, "step": 15490 }, { "epoch": 1.6794232437120555, "grad_norm": 2.9380199909210205, "learning_rate": 8.835775424343538e-06, "loss": 0.0773, "step": 15491 }, { "epoch": 1.6795316565481353, "grad_norm": 2.2828779220581055, "learning_rate": 8.83505005077615e-06, "loss": 0.0718, "step": 15492 }, { "epoch": 1.679640069384215, "grad_norm": 2.2958385944366455, "learning_rate": 8.834324677208764e-06, "loss": 0.0185, "step": 15493 }, { "epoch": 1.6797484822202948, "grad_norm": 4.526198863983154, "learning_rate": 8.833599303641376e-06, "loss": 0.0452, "step": 15494 }, { "epoch": 1.6798568950563748, "grad_norm": 3.583008050918579, "learning_rate": 8.832873930073988e-06, "loss": 0.0467, "step": 15495 }, { "epoch": 1.6799653078924544, "grad_norm": 9.338337898254395, "learning_rate": 8.832148556506602e-06, "loss": 0.009, "step": 15496 }, { "epoch": 1.6800737207285343, "grad_norm": 2.041952610015869, "learning_rate": 8.831423182939214e-06, "loss": 0.0266, "step": 15497 }, { "epoch": 1.680182133564614, "grad_norm": 6.130582332611084, "learning_rate": 8.830697809371828e-06, "loss": 0.0161, "step": 15498 }, { "epoch": 1.6802905464006939, "grad_norm": 2.432117462158203, "learning_rate": 8.82997243580444e-06, "loss": 0.0357, "step": 15499 }, { "epoch": 1.6803989592367736, "grad_norm": 6.109475135803223, "learning_rate": 8.829247062237052e-06, "loss": 0.0347, "step": 15500 }, { "epoch": 1.6805073720728534, "grad_norm": 1.5020747184753418, "learning_rate": 8.828521688669666e-06, "loss": 0.0616, "step": 15501 }, { "epoch": 1.6806157849089334, "grad_norm": 0.2004055380821228, "learning_rate": 8.827796315102278e-06, "loss": 0.0018, "step": 15502 }, { "epoch": 1.680724197745013, "grad_norm": 0.7590656876564026, "learning_rate": 8.82707094153489e-06, "loss": 0.0095, "step": 15503 }, { "epoch": 1.680832610581093, "grad_norm": 2.321474313735962, "learning_rate": 8.826345567967504e-06, "loss": 0.0259, "step": 15504 }, { "epoch": 1.6809410234171724, "grad_norm": 3.4980390071868896, "learning_rate": 8.825620194400116e-06, "loss": 0.0304, "step": 15505 }, { "epoch": 1.6810494362532524, "grad_norm": 8.844280242919922, "learning_rate": 8.82489482083273e-06, "loss": 0.0276, "step": 15506 }, { "epoch": 1.6811578490893322, "grad_norm": 2.4060614109039307, "learning_rate": 8.824169447265342e-06, "loss": 0.0418, "step": 15507 }, { "epoch": 1.681266261925412, "grad_norm": 0.7403106689453125, "learning_rate": 8.823444073697956e-06, "loss": 0.0115, "step": 15508 }, { "epoch": 1.6813746747614917, "grad_norm": 1.2412151098251343, "learning_rate": 8.822718700130568e-06, "loss": 0.0182, "step": 15509 }, { "epoch": 1.6814830875975715, "grad_norm": 0.31593555212020874, "learning_rate": 8.82199332656318e-06, "loss": 0.0021, "step": 15510 }, { "epoch": 1.6815915004336515, "grad_norm": 0.6617424488067627, "learning_rate": 8.821267952995793e-06, "loss": 0.0151, "step": 15511 }, { "epoch": 1.681699913269731, "grad_norm": 5.545982360839844, "learning_rate": 8.820542579428406e-06, "loss": 0.1029, "step": 15512 }, { "epoch": 1.681808326105811, "grad_norm": 1.9020377397537231, "learning_rate": 8.819817205861019e-06, "loss": 0.0214, "step": 15513 }, { "epoch": 1.6819167389418908, "grad_norm": 0.3639196753501892, "learning_rate": 8.81909183229363e-06, "loss": 0.0087, "step": 15514 }, { "epoch": 1.6820251517779705, "grad_norm": 0.33613696694374084, "learning_rate": 8.818366458726245e-06, "loss": 0.0127, "step": 15515 }, { "epoch": 1.6821335646140503, "grad_norm": 0.7518830895423889, "learning_rate": 8.817641085158858e-06, "loss": 0.0311, "step": 15516 }, { "epoch": 1.68224197745013, "grad_norm": 6.029175281524658, "learning_rate": 8.81691571159147e-06, "loss": 0.031, "step": 15517 }, { "epoch": 1.68235039028621, "grad_norm": 7.2503533363342285, "learning_rate": 8.816190338024083e-06, "loss": 0.0162, "step": 15518 }, { "epoch": 1.6824588031222896, "grad_norm": 0.8606199622154236, "learning_rate": 8.815464964456697e-06, "loss": 0.0161, "step": 15519 }, { "epoch": 1.6825672159583696, "grad_norm": 6.382972240447998, "learning_rate": 8.814739590889309e-06, "loss": 0.0323, "step": 15520 }, { "epoch": 1.682675628794449, "grad_norm": 2.8645362854003906, "learning_rate": 8.81401421732192e-06, "loss": 0.0578, "step": 15521 }, { "epoch": 1.682784041630529, "grad_norm": 2.3859338760375977, "learning_rate": 8.813288843754535e-06, "loss": 0.0326, "step": 15522 }, { "epoch": 1.6828924544666088, "grad_norm": 1.8662595748901367, "learning_rate": 8.812563470187147e-06, "loss": 0.0517, "step": 15523 }, { "epoch": 1.6830008673026886, "grad_norm": 1.6403993368148804, "learning_rate": 8.81183809661976e-06, "loss": 0.0327, "step": 15524 }, { "epoch": 1.6831092801387686, "grad_norm": 5.645172595977783, "learning_rate": 8.811112723052373e-06, "loss": 0.0323, "step": 15525 }, { "epoch": 1.6832176929748481, "grad_norm": 3.208956718444824, "learning_rate": 8.810387349484985e-06, "loss": 0.0202, "step": 15526 }, { "epoch": 1.6833261058109281, "grad_norm": 0.09411982446908951, "learning_rate": 8.809661975917599e-06, "loss": 0.001, "step": 15527 }, { "epoch": 1.6834345186470077, "grad_norm": 3.58697247505188, "learning_rate": 8.808936602350211e-06, "loss": 0.0175, "step": 15528 }, { "epoch": 1.6835429314830876, "grad_norm": 1.8395675420761108, "learning_rate": 8.808211228782823e-06, "loss": 0.0187, "step": 15529 }, { "epoch": 1.6836513443191674, "grad_norm": 4.344844341278076, "learning_rate": 8.807485855215437e-06, "loss": 0.0151, "step": 15530 }, { "epoch": 1.6837597571552472, "grad_norm": 7.35872220993042, "learning_rate": 8.806760481648049e-06, "loss": 0.0251, "step": 15531 }, { "epoch": 1.683868169991327, "grad_norm": 0.3281245231628418, "learning_rate": 8.806035108080663e-06, "loss": 0.0109, "step": 15532 }, { "epoch": 1.6839765828274067, "grad_norm": 3.4782843589782715, "learning_rate": 8.805309734513275e-06, "loss": 0.0137, "step": 15533 }, { "epoch": 1.6840849956634867, "grad_norm": 0.23530855774879456, "learning_rate": 8.804584360945889e-06, "loss": 0.0023, "step": 15534 }, { "epoch": 1.6841934084995662, "grad_norm": 3.3395822048187256, "learning_rate": 8.803858987378501e-06, "loss": 0.0237, "step": 15535 }, { "epoch": 1.6843018213356462, "grad_norm": 1.1155898571014404, "learning_rate": 8.803133613811113e-06, "loss": 0.0328, "step": 15536 }, { "epoch": 1.684410234171726, "grad_norm": 2.1072628498077393, "learning_rate": 8.802408240243725e-06, "loss": 0.0158, "step": 15537 }, { "epoch": 1.6845186470078057, "grad_norm": 4.472982883453369, "learning_rate": 8.801682866676339e-06, "loss": 0.022, "step": 15538 }, { "epoch": 1.6846270598438855, "grad_norm": 3.0472471714019775, "learning_rate": 8.800957493108951e-06, "loss": 0.0146, "step": 15539 }, { "epoch": 1.6847354726799653, "grad_norm": 1.194154143333435, "learning_rate": 8.800232119541563e-06, "loss": 0.0026, "step": 15540 }, { "epoch": 1.6848438855160452, "grad_norm": 2.7148287296295166, "learning_rate": 8.799506745974177e-06, "loss": 0.0406, "step": 15541 }, { "epoch": 1.6849522983521248, "grad_norm": 12.387070655822754, "learning_rate": 8.798781372406791e-06, "loss": 0.0476, "step": 15542 }, { "epoch": 1.6850607111882048, "grad_norm": 4.647090911865234, "learning_rate": 8.798055998839403e-06, "loss": 0.0185, "step": 15543 }, { "epoch": 1.6851691240242843, "grad_norm": 0.805517852306366, "learning_rate": 8.797330625272015e-06, "loss": 0.0098, "step": 15544 }, { "epoch": 1.6852775368603643, "grad_norm": 3.386786699295044, "learning_rate": 8.79660525170463e-06, "loss": 0.0306, "step": 15545 }, { "epoch": 1.685385949696444, "grad_norm": 1.3288089036941528, "learning_rate": 8.795879878137241e-06, "loss": 0.0171, "step": 15546 }, { "epoch": 1.6854943625325238, "grad_norm": 1.5745530128479004, "learning_rate": 8.795154504569853e-06, "loss": 0.0244, "step": 15547 }, { "epoch": 1.6856027753686036, "grad_norm": 0.32173290848731995, "learning_rate": 8.794429131002467e-06, "loss": 0.0058, "step": 15548 }, { "epoch": 1.6857111882046834, "grad_norm": 4.087828636169434, "learning_rate": 8.79370375743508e-06, "loss": 0.0119, "step": 15549 }, { "epoch": 1.6858196010407633, "grad_norm": 1.5000237226486206, "learning_rate": 8.792978383867693e-06, "loss": 0.006, "step": 15550 }, { "epoch": 1.6859280138768429, "grad_norm": 2.7634947299957275, "learning_rate": 8.792253010300305e-06, "loss": 0.0307, "step": 15551 }, { "epoch": 1.6860364267129229, "grad_norm": 0.7150548696517944, "learning_rate": 8.791527636732918e-06, "loss": 0.0033, "step": 15552 }, { "epoch": 1.6861448395490026, "grad_norm": 1.2573217153549194, "learning_rate": 8.790802263165531e-06, "loss": 0.021, "step": 15553 }, { "epoch": 1.6862532523850824, "grad_norm": 0.9113923907279968, "learning_rate": 8.790076889598144e-06, "loss": 0.0125, "step": 15554 }, { "epoch": 1.6863616652211622, "grad_norm": 1.3390345573425293, "learning_rate": 8.789351516030756e-06, "loss": 0.04, "step": 15555 }, { "epoch": 1.686470078057242, "grad_norm": 4.8538289070129395, "learning_rate": 8.78862614246337e-06, "loss": 0.0153, "step": 15556 }, { "epoch": 1.686578490893322, "grad_norm": 1.2444573640823364, "learning_rate": 8.787900768895982e-06, "loss": 0.0262, "step": 15557 }, { "epoch": 1.6866869037294014, "grad_norm": 0.39825716614723206, "learning_rate": 8.787175395328596e-06, "loss": 0.0178, "step": 15558 }, { "epoch": 1.6867953165654814, "grad_norm": 0.93556147813797, "learning_rate": 8.786450021761208e-06, "loss": 0.0208, "step": 15559 }, { "epoch": 1.6869037294015612, "grad_norm": 1.9062963724136353, "learning_rate": 8.785724648193822e-06, "loss": 0.0252, "step": 15560 }, { "epoch": 1.687012142237641, "grad_norm": 0.8156787157058716, "learning_rate": 8.784999274626434e-06, "loss": 0.0053, "step": 15561 }, { "epoch": 1.6871205550737207, "grad_norm": 1.377826452255249, "learning_rate": 8.784273901059046e-06, "loss": 0.0263, "step": 15562 }, { "epoch": 1.6872289679098005, "grad_norm": 3.9016823768615723, "learning_rate": 8.783548527491658e-06, "loss": 0.0116, "step": 15563 }, { "epoch": 1.6873373807458805, "grad_norm": 0.03459421917796135, "learning_rate": 8.782823153924272e-06, "loss": 0.0003, "step": 15564 }, { "epoch": 1.68744579358196, "grad_norm": 4.081727027893066, "learning_rate": 8.782097780356884e-06, "loss": 0.0311, "step": 15565 }, { "epoch": 1.68755420641804, "grad_norm": 0.23092862963676453, "learning_rate": 8.781372406789496e-06, "loss": 0.0004, "step": 15566 }, { "epoch": 1.6876626192541195, "grad_norm": 3.564652681350708, "learning_rate": 8.78064703322211e-06, "loss": 0.0183, "step": 15567 }, { "epoch": 1.6877710320901995, "grad_norm": 4.680166721343994, "learning_rate": 8.779921659654724e-06, "loss": 0.0271, "step": 15568 }, { "epoch": 1.6878794449262793, "grad_norm": 0.7901824116706848, "learning_rate": 8.779196286087336e-06, "loss": 0.0073, "step": 15569 }, { "epoch": 1.687987857762359, "grad_norm": 1.9371798038482666, "learning_rate": 8.778470912519948e-06, "loss": 0.0282, "step": 15570 }, { "epoch": 1.6880962705984388, "grad_norm": 1.4129612445831299, "learning_rate": 8.777745538952562e-06, "loss": 0.01, "step": 15571 }, { "epoch": 1.6882046834345186, "grad_norm": 2.585416793823242, "learning_rate": 8.777020165385174e-06, "loss": 0.0287, "step": 15572 }, { "epoch": 1.6883130962705986, "grad_norm": 9.009903907775879, "learning_rate": 8.776294791817786e-06, "loss": 0.0389, "step": 15573 }, { "epoch": 1.688421509106678, "grad_norm": 0.5087922811508179, "learning_rate": 8.7755694182504e-06, "loss": 0.0023, "step": 15574 }, { "epoch": 1.688529921942758, "grad_norm": 1.3715660572052002, "learning_rate": 8.774844044683012e-06, "loss": 0.032, "step": 15575 }, { "epoch": 1.6886383347788378, "grad_norm": 2.6316263675689697, "learning_rate": 8.774118671115626e-06, "loss": 0.0794, "step": 15576 }, { "epoch": 1.6887467476149176, "grad_norm": 2.040562868118286, "learning_rate": 8.773393297548238e-06, "loss": 0.0069, "step": 15577 }, { "epoch": 1.6888551604509974, "grad_norm": 2.641364097595215, "learning_rate": 8.77266792398085e-06, "loss": 0.0172, "step": 15578 }, { "epoch": 1.6889635732870771, "grad_norm": 5.480940341949463, "learning_rate": 8.771942550413464e-06, "loss": 0.0664, "step": 15579 }, { "epoch": 1.6890719861231571, "grad_norm": 0.3155122697353363, "learning_rate": 8.771217176846076e-06, "loss": 0.0078, "step": 15580 }, { "epoch": 1.6891803989592367, "grad_norm": 4.057413101196289, "learning_rate": 8.770491803278688e-06, "loss": 0.0114, "step": 15581 }, { "epoch": 1.6892888117953166, "grad_norm": 3.343637704849243, "learning_rate": 8.769766429711302e-06, "loss": 0.0336, "step": 15582 }, { "epoch": 1.6893972246313964, "grad_norm": 3.2086596488952637, "learning_rate": 8.769041056143914e-06, "loss": 0.0274, "step": 15583 }, { "epoch": 1.6895056374674762, "grad_norm": 1.0170810222625732, "learning_rate": 8.768315682576528e-06, "loss": 0.0049, "step": 15584 }, { "epoch": 1.689614050303556, "grad_norm": 0.394169420003891, "learning_rate": 8.76759030900914e-06, "loss": 0.0229, "step": 15585 }, { "epoch": 1.6897224631396357, "grad_norm": 0.513087272644043, "learning_rate": 8.766864935441754e-06, "loss": 0.0178, "step": 15586 }, { "epoch": 1.6898308759757157, "grad_norm": 0.5290887951850891, "learning_rate": 8.766139561874366e-06, "loss": 0.0145, "step": 15587 }, { "epoch": 1.6899392888117952, "grad_norm": 2.0807957649230957, "learning_rate": 8.765414188306978e-06, "loss": 0.0038, "step": 15588 }, { "epoch": 1.6900477016478752, "grad_norm": 1.8581063747406006, "learning_rate": 8.76468881473959e-06, "loss": 0.0156, "step": 15589 }, { "epoch": 1.6901561144839548, "grad_norm": 3.1298134326934814, "learning_rate": 8.763963441172204e-06, "loss": 0.0171, "step": 15590 }, { "epoch": 1.6902645273200347, "grad_norm": 0.21718262135982513, "learning_rate": 8.763238067604817e-06, "loss": 0.0012, "step": 15591 }, { "epoch": 1.6903729401561145, "grad_norm": 0.8581893444061279, "learning_rate": 8.76251269403743e-06, "loss": 0.0079, "step": 15592 }, { "epoch": 1.6904813529921943, "grad_norm": 1.5348697900772095, "learning_rate": 8.761787320470043e-06, "loss": 0.0037, "step": 15593 }, { "epoch": 1.690589765828274, "grad_norm": 1.0930334329605103, "learning_rate": 8.761061946902656e-06, "loss": 0.0041, "step": 15594 }, { "epoch": 1.6906981786643538, "grad_norm": 1.6425446271896362, "learning_rate": 8.760336573335269e-06, "loss": 0.0293, "step": 15595 }, { "epoch": 1.6908065915004338, "grad_norm": 0.5392507910728455, "learning_rate": 8.75961119976788e-06, "loss": 0.0049, "step": 15596 }, { "epoch": 1.6909150043365133, "grad_norm": 1.515643835067749, "learning_rate": 8.758885826200495e-06, "loss": 0.0028, "step": 15597 }, { "epoch": 1.6910234171725933, "grad_norm": 3.350118637084961, "learning_rate": 8.758160452633107e-06, "loss": 0.0322, "step": 15598 }, { "epoch": 1.691131830008673, "grad_norm": 2.5506935119628906, "learning_rate": 8.757435079065719e-06, "loss": 0.0386, "step": 15599 }, { "epoch": 1.6912402428447528, "grad_norm": 1.7319926023483276, "learning_rate": 8.756709705498333e-06, "loss": 0.0185, "step": 15600 }, { "epoch": 1.6913486556808326, "grad_norm": 0.7895534634590149, "learning_rate": 8.755984331930945e-06, "loss": 0.0138, "step": 15601 }, { "epoch": 1.6914570685169124, "grad_norm": 0.057364754378795624, "learning_rate": 8.755258958363559e-06, "loss": 0.0004, "step": 15602 }, { "epoch": 1.6915654813529923, "grad_norm": 2.6846768856048584, "learning_rate": 8.75453358479617e-06, "loss": 0.0323, "step": 15603 }, { "epoch": 1.6916738941890719, "grad_norm": 3.876166343688965, "learning_rate": 8.753808211228783e-06, "loss": 0.0209, "step": 15604 }, { "epoch": 1.6917823070251519, "grad_norm": 0.5221475958824158, "learning_rate": 8.753082837661397e-06, "loss": 0.0079, "step": 15605 }, { "epoch": 1.6918907198612314, "grad_norm": 0.7041590213775635, "learning_rate": 8.752357464094009e-06, "loss": 0.0017, "step": 15606 }, { "epoch": 1.6919991326973114, "grad_norm": 0.6003170013427734, "learning_rate": 8.751632090526621e-06, "loss": 0.0176, "step": 15607 }, { "epoch": 1.6921075455333912, "grad_norm": 0.23478323221206665, "learning_rate": 8.750906716959235e-06, "loss": 0.0012, "step": 15608 }, { "epoch": 1.692215958369471, "grad_norm": 5.836124420166016, "learning_rate": 8.750181343391847e-06, "loss": 0.025, "step": 15609 }, { "epoch": 1.692324371205551, "grad_norm": 6.339295864105225, "learning_rate": 8.749455969824461e-06, "loss": 0.0133, "step": 15610 }, { "epoch": 1.6924327840416304, "grad_norm": 2.993443727493286, "learning_rate": 8.748730596257073e-06, "loss": 0.0776, "step": 15611 }, { "epoch": 1.6925411968777104, "grad_norm": 4.471772193908691, "learning_rate": 8.748005222689687e-06, "loss": 0.0162, "step": 15612 }, { "epoch": 1.69264960971379, "grad_norm": 2.0369482040405273, "learning_rate": 8.747279849122299e-06, "loss": 0.038, "step": 15613 }, { "epoch": 1.69275802254987, "grad_norm": 4.898540496826172, "learning_rate": 8.746554475554911e-06, "loss": 0.0324, "step": 15614 }, { "epoch": 1.6928664353859497, "grad_norm": 0.810990035533905, "learning_rate": 8.745829101987523e-06, "loss": 0.0115, "step": 15615 }, { "epoch": 1.6929748482220295, "grad_norm": 4.853838920593262, "learning_rate": 8.745103728420137e-06, "loss": 0.0465, "step": 15616 }, { "epoch": 1.6930832610581092, "grad_norm": 1.5833746194839478, "learning_rate": 8.74437835485275e-06, "loss": 0.0072, "step": 15617 }, { "epoch": 1.693191673894189, "grad_norm": 3.898742914199829, "learning_rate": 8.743652981285363e-06, "loss": 0.0326, "step": 15618 }, { "epoch": 1.693300086730269, "grad_norm": 0.5143082737922668, "learning_rate": 8.742927607717975e-06, "loss": 0.012, "step": 15619 }, { "epoch": 1.6934084995663485, "grad_norm": 9.973628044128418, "learning_rate": 8.742202234150589e-06, "loss": 0.0423, "step": 15620 }, { "epoch": 1.6935169124024285, "grad_norm": 4.1122145652771, "learning_rate": 8.741476860583201e-06, "loss": 0.0089, "step": 15621 }, { "epoch": 1.6936253252385083, "grad_norm": 3.0288803577423096, "learning_rate": 8.740751487015813e-06, "loss": 0.0075, "step": 15622 }, { "epoch": 1.693733738074588, "grad_norm": 2.4803571701049805, "learning_rate": 8.740026113448427e-06, "loss": 0.0445, "step": 15623 }, { "epoch": 1.6938421509106678, "grad_norm": 0.40056654810905457, "learning_rate": 8.73930073988104e-06, "loss": 0.0132, "step": 15624 }, { "epoch": 1.6939505637467476, "grad_norm": 6.182928085327148, "learning_rate": 8.738575366313652e-06, "loss": 0.0379, "step": 15625 }, { "epoch": 1.6940589765828276, "grad_norm": 3.716174840927124, "learning_rate": 8.737849992746265e-06, "loss": 0.0099, "step": 15626 }, { "epoch": 1.694167389418907, "grad_norm": 0.7554640173912048, "learning_rate": 8.737124619178877e-06, "loss": 0.0242, "step": 15627 }, { "epoch": 1.694275802254987, "grad_norm": 1.3098772764205933, "learning_rate": 8.736399245611491e-06, "loss": 0.0366, "step": 15628 }, { "epoch": 1.6943842150910666, "grad_norm": 0.33180975914001465, "learning_rate": 8.735673872044103e-06, "loss": 0.0009, "step": 15629 }, { "epoch": 1.6944926279271466, "grad_norm": 0.4433070719242096, "learning_rate": 8.734948498476716e-06, "loss": 0.0158, "step": 15630 }, { "epoch": 1.6946010407632264, "grad_norm": 0.43432921171188354, "learning_rate": 8.73422312490933e-06, "loss": 0.0096, "step": 15631 }, { "epoch": 1.6947094535993061, "grad_norm": 0.04835217818617821, "learning_rate": 8.733497751341942e-06, "loss": 0.0002, "step": 15632 }, { "epoch": 1.694817866435386, "grad_norm": 1.4976415634155273, "learning_rate": 8.732772377774554e-06, "loss": 0.0303, "step": 15633 }, { "epoch": 1.6949262792714657, "grad_norm": 2.838146924972534, "learning_rate": 8.732047004207168e-06, "loss": 0.0164, "step": 15634 }, { "epoch": 1.6950346921075456, "grad_norm": 2.0861239433288574, "learning_rate": 8.73132163063978e-06, "loss": 0.0354, "step": 15635 }, { "epoch": 1.6951431049436252, "grad_norm": 1.468245267868042, "learning_rate": 8.730596257072394e-06, "loss": 0.0081, "step": 15636 }, { "epoch": 1.6952515177797052, "grad_norm": 1.5784032344818115, "learning_rate": 8.729870883505006e-06, "loss": 0.027, "step": 15637 }, { "epoch": 1.695359930615785, "grad_norm": 1.9144260883331299, "learning_rate": 8.72914550993762e-06, "loss": 0.0412, "step": 15638 }, { "epoch": 1.6954683434518647, "grad_norm": 1.612510085105896, "learning_rate": 8.728420136370232e-06, "loss": 0.0171, "step": 15639 }, { "epoch": 1.6955767562879445, "grad_norm": 0.2683279812335968, "learning_rate": 8.727694762802844e-06, "loss": 0.0046, "step": 15640 }, { "epoch": 1.6956851691240242, "grad_norm": 9.421296119689941, "learning_rate": 8.726969389235456e-06, "loss": 0.0157, "step": 15641 }, { "epoch": 1.6957935819601042, "grad_norm": 2.537778377532959, "learning_rate": 8.72624401566807e-06, "loss": 0.0279, "step": 15642 }, { "epoch": 1.6959019947961838, "grad_norm": 0.2218884378671646, "learning_rate": 8.725518642100682e-06, "loss": 0.0016, "step": 15643 }, { "epoch": 1.6960104076322637, "grad_norm": 0.4433078467845917, "learning_rate": 8.724793268533296e-06, "loss": 0.0088, "step": 15644 }, { "epoch": 1.6961188204683435, "grad_norm": 1.5471704006195068, "learning_rate": 8.724067894965908e-06, "loss": 0.0475, "step": 15645 }, { "epoch": 1.6962272333044233, "grad_norm": 4.815232276916504, "learning_rate": 8.723342521398522e-06, "loss": 0.0461, "step": 15646 }, { "epoch": 1.696335646140503, "grad_norm": 1.3497387170791626, "learning_rate": 8.722617147831134e-06, "loss": 0.0039, "step": 15647 }, { "epoch": 1.6964440589765828, "grad_norm": 6.819809436798096, "learning_rate": 8.721891774263746e-06, "loss": 0.0359, "step": 15648 }, { "epoch": 1.6965524718126628, "grad_norm": 1.2673882246017456, "learning_rate": 8.72116640069636e-06, "loss": 0.0063, "step": 15649 }, { "epoch": 1.6966608846487423, "grad_norm": 0.8626117706298828, "learning_rate": 8.720441027128972e-06, "loss": 0.026, "step": 15650 }, { "epoch": 1.6967692974848223, "grad_norm": 1.4031496047973633, "learning_rate": 8.719715653561584e-06, "loss": 0.0078, "step": 15651 }, { "epoch": 1.6968777103209018, "grad_norm": 6.769156455993652, "learning_rate": 8.718990279994198e-06, "loss": 0.0254, "step": 15652 }, { "epoch": 1.6969861231569818, "grad_norm": 2.259446859359741, "learning_rate": 8.718264906426812e-06, "loss": 0.0418, "step": 15653 }, { "epoch": 1.6970945359930616, "grad_norm": 0.3528255224227905, "learning_rate": 8.717539532859424e-06, "loss": 0.0144, "step": 15654 }, { "epoch": 1.6972029488291414, "grad_norm": 3.8847806453704834, "learning_rate": 8.716814159292036e-06, "loss": 0.0115, "step": 15655 }, { "epoch": 1.6973113616652211, "grad_norm": 0.13518323004245758, "learning_rate": 8.716088785724648e-06, "loss": 0.001, "step": 15656 }, { "epoch": 1.6974197745013009, "grad_norm": 0.4232494533061981, "learning_rate": 8.715363412157262e-06, "loss": 0.0207, "step": 15657 }, { "epoch": 1.6975281873373809, "grad_norm": 11.655574798583984, "learning_rate": 8.714638038589874e-06, "loss": 0.0376, "step": 15658 }, { "epoch": 1.6976366001734604, "grad_norm": 1.1745578050613403, "learning_rate": 8.713912665022486e-06, "loss": 0.0038, "step": 15659 }, { "epoch": 1.6977450130095404, "grad_norm": 0.6204564571380615, "learning_rate": 8.7131872914551e-06, "loss": 0.0174, "step": 15660 }, { "epoch": 1.6978534258456202, "grad_norm": 2.455766439437866, "learning_rate": 8.712461917887712e-06, "loss": 0.0045, "step": 15661 }, { "epoch": 1.6979618386817, "grad_norm": 1.5443602800369263, "learning_rate": 8.711736544320326e-06, "loss": 0.0317, "step": 15662 }, { "epoch": 1.6980702515177797, "grad_norm": 0.36924588680267334, "learning_rate": 8.711011170752938e-06, "loss": 0.0024, "step": 15663 }, { "epoch": 1.6981786643538594, "grad_norm": 2.1377103328704834, "learning_rate": 8.710285797185552e-06, "loss": 0.0127, "step": 15664 }, { "epoch": 1.6982870771899394, "grad_norm": 8.769771575927734, "learning_rate": 8.709560423618164e-06, "loss": 0.0539, "step": 15665 }, { "epoch": 1.698395490026019, "grad_norm": 0.9235367774963379, "learning_rate": 8.708835050050776e-06, "loss": 0.013, "step": 15666 }, { "epoch": 1.698503902862099, "grad_norm": 2.725797414779663, "learning_rate": 8.708109676483389e-06, "loss": 0.0217, "step": 15667 }, { "epoch": 1.6986123156981785, "grad_norm": 0.6076398491859436, "learning_rate": 8.707384302916002e-06, "loss": 0.0175, "step": 15668 }, { "epoch": 1.6987207285342585, "grad_norm": 5.7844343185424805, "learning_rate": 8.706658929348615e-06, "loss": 0.0511, "step": 15669 }, { "epoch": 1.6988291413703382, "grad_norm": 5.873962879180908, "learning_rate": 8.705933555781228e-06, "loss": 0.0108, "step": 15670 }, { "epoch": 1.698937554206418, "grad_norm": 2.4593987464904785, "learning_rate": 8.70520818221384e-06, "loss": 0.0912, "step": 15671 }, { "epoch": 1.699045967042498, "grad_norm": 1.0133002996444702, "learning_rate": 8.704482808646454e-06, "loss": 0.0311, "step": 15672 }, { "epoch": 1.6991543798785775, "grad_norm": 1.0693359375, "learning_rate": 8.703757435079067e-06, "loss": 0.0133, "step": 15673 }, { "epoch": 1.6992627927146575, "grad_norm": 1.9676703214645386, "learning_rate": 8.703032061511679e-06, "loss": 0.0251, "step": 15674 }, { "epoch": 1.699371205550737, "grad_norm": 4.433228492736816, "learning_rate": 8.702306687944293e-06, "loss": 0.0595, "step": 15675 }, { "epoch": 1.699479618386817, "grad_norm": 0.05680453032255173, "learning_rate": 8.701581314376905e-06, "loss": 0.0004, "step": 15676 }, { "epoch": 1.6995880312228968, "grad_norm": 0.44111892580986023, "learning_rate": 8.700855940809517e-06, "loss": 0.0143, "step": 15677 }, { "epoch": 1.6996964440589766, "grad_norm": 1.8828316926956177, "learning_rate": 8.70013056724213e-06, "loss": 0.0178, "step": 15678 }, { "epoch": 1.6998048568950563, "grad_norm": 2.0619003772735596, "learning_rate": 8.699405193674745e-06, "loss": 0.036, "step": 15679 }, { "epoch": 1.699913269731136, "grad_norm": 0.3771032392978668, "learning_rate": 8.698679820107357e-06, "loss": 0.0136, "step": 15680 }, { "epoch": 1.700021682567216, "grad_norm": 0.39183661341667175, "learning_rate": 8.697954446539969e-06, "loss": 0.0145, "step": 15681 }, { "epoch": 1.7001300954032956, "grad_norm": 2.7312278747558594, "learning_rate": 8.697229072972581e-06, "loss": 0.0763, "step": 15682 }, { "epoch": 1.7002385082393756, "grad_norm": 3.4070701599121094, "learning_rate": 8.696503699405195e-06, "loss": 0.0139, "step": 15683 }, { "epoch": 1.7003469210754554, "grad_norm": 0.1666911095380783, "learning_rate": 8.695778325837807e-06, "loss": 0.0011, "step": 15684 }, { "epoch": 1.7004553339115351, "grad_norm": 1.7685410976409912, "learning_rate": 8.695052952270419e-06, "loss": 0.0058, "step": 15685 }, { "epoch": 1.700563746747615, "grad_norm": 9.73998737335205, "learning_rate": 8.694327578703033e-06, "loss": 0.0486, "step": 15686 }, { "epoch": 1.7006721595836947, "grad_norm": 0.8319499492645264, "learning_rate": 8.693602205135647e-06, "loss": 0.0117, "step": 15687 }, { "epoch": 1.7007805724197746, "grad_norm": 5.568907260894775, "learning_rate": 8.692876831568259e-06, "loss": 0.0219, "step": 15688 }, { "epoch": 1.7008889852558542, "grad_norm": 1.3891371488571167, "learning_rate": 8.692151458000871e-06, "loss": 0.0323, "step": 15689 }, { "epoch": 1.7009973980919342, "grad_norm": 4.295403003692627, "learning_rate": 8.691426084433485e-06, "loss": 0.0111, "step": 15690 }, { "epoch": 1.7011058109280137, "grad_norm": 3.0901358127593994, "learning_rate": 8.690700710866097e-06, "loss": 0.0602, "step": 15691 }, { "epoch": 1.7012142237640937, "grad_norm": 0.19553694128990173, "learning_rate": 8.68997533729871e-06, "loss": 0.0015, "step": 15692 }, { "epoch": 1.7013226366001735, "grad_norm": 5.940379619598389, "learning_rate": 8.689249963731321e-06, "loss": 0.0651, "step": 15693 }, { "epoch": 1.7014310494362532, "grad_norm": 0.42441096901893616, "learning_rate": 8.688524590163935e-06, "loss": 0.0209, "step": 15694 }, { "epoch": 1.701539462272333, "grad_norm": 15.812790870666504, "learning_rate": 8.687799216596547e-06, "loss": 0.0408, "step": 15695 }, { "epoch": 1.7016478751084128, "grad_norm": 5.177221775054932, "learning_rate": 8.687073843029161e-06, "loss": 0.0659, "step": 15696 }, { "epoch": 1.7017562879444927, "grad_norm": 2.771554946899414, "learning_rate": 8.686348469461773e-06, "loss": 0.0292, "step": 15697 }, { "epoch": 1.7018647007805723, "grad_norm": 4.9676008224487305, "learning_rate": 8.685623095894387e-06, "loss": 0.0211, "step": 15698 }, { "epoch": 1.7019731136166523, "grad_norm": 0.4422203600406647, "learning_rate": 8.684897722327e-06, "loss": 0.0164, "step": 15699 }, { "epoch": 1.702081526452732, "grad_norm": 0.7758595943450928, "learning_rate": 8.684172348759611e-06, "loss": 0.0063, "step": 15700 }, { "epoch": 1.7021899392888118, "grad_norm": 2.8114898204803467, "learning_rate": 8.683446975192225e-06, "loss": 0.0448, "step": 15701 }, { "epoch": 1.7022983521248916, "grad_norm": 1.954424262046814, "learning_rate": 8.682721601624837e-06, "loss": 0.021, "step": 15702 }, { "epoch": 1.7024067649609713, "grad_norm": 8.814002990722656, "learning_rate": 8.68199622805745e-06, "loss": 0.0699, "step": 15703 }, { "epoch": 1.7025151777970513, "grad_norm": 2.260345220565796, "learning_rate": 8.681270854490063e-06, "loss": 0.0544, "step": 15704 }, { "epoch": 1.7026235906331308, "grad_norm": 1.1967614889144897, "learning_rate": 8.680545480922677e-06, "loss": 0.0093, "step": 15705 }, { "epoch": 1.7027320034692108, "grad_norm": 2.84369158744812, "learning_rate": 8.67982010735529e-06, "loss": 0.0433, "step": 15706 }, { "epoch": 1.7028404163052906, "grad_norm": 0.9026446342468262, "learning_rate": 8.679094733787901e-06, "loss": 0.0229, "step": 15707 }, { "epoch": 1.7029488291413704, "grad_norm": 0.43238624930381775, "learning_rate": 8.678369360220514e-06, "loss": 0.0082, "step": 15708 }, { "epoch": 1.7030572419774501, "grad_norm": 0.3442783057689667, "learning_rate": 8.677643986653127e-06, "loss": 0.0038, "step": 15709 }, { "epoch": 1.7031656548135299, "grad_norm": 1.8496429920196533, "learning_rate": 8.67691861308574e-06, "loss": 0.0258, "step": 15710 }, { "epoch": 1.7032740676496099, "grad_norm": 0.2871474027633667, "learning_rate": 8.676193239518352e-06, "loss": 0.0071, "step": 15711 }, { "epoch": 1.7033824804856894, "grad_norm": 0.4893612861633301, "learning_rate": 8.675467865950966e-06, "loss": 0.0133, "step": 15712 }, { "epoch": 1.7034908933217694, "grad_norm": 0.5163834095001221, "learning_rate": 8.67474249238358e-06, "loss": 0.0026, "step": 15713 }, { "epoch": 1.703599306157849, "grad_norm": 0.12257236242294312, "learning_rate": 8.674017118816192e-06, "loss": 0.0014, "step": 15714 }, { "epoch": 1.703707718993929, "grad_norm": 1.4919719696044922, "learning_rate": 8.673291745248804e-06, "loss": 0.0191, "step": 15715 }, { "epoch": 1.7038161318300087, "grad_norm": 1.1432462930679321, "learning_rate": 8.672566371681418e-06, "loss": 0.0107, "step": 15716 }, { "epoch": 1.7039245446660884, "grad_norm": 2.6554250717163086, "learning_rate": 8.67184099811403e-06, "loss": 0.0215, "step": 15717 }, { "epoch": 1.7040329575021682, "grad_norm": 1.9243881702423096, "learning_rate": 8.671115624546642e-06, "loss": 0.0459, "step": 15718 }, { "epoch": 1.704141370338248, "grad_norm": 4.609610557556152, "learning_rate": 8.670390250979254e-06, "loss": 0.0283, "step": 15719 }, { "epoch": 1.704249783174328, "grad_norm": 0.39007267355918884, "learning_rate": 8.669664877411868e-06, "loss": 0.0055, "step": 15720 }, { "epoch": 1.7043581960104075, "grad_norm": 1.648861050605774, "learning_rate": 8.66893950384448e-06, "loss": 0.0296, "step": 15721 }, { "epoch": 1.7044666088464875, "grad_norm": 2.9429304599761963, "learning_rate": 8.668214130277094e-06, "loss": 0.0282, "step": 15722 }, { "epoch": 1.7045750216825672, "grad_norm": 0.8442559838294983, "learning_rate": 8.667488756709706e-06, "loss": 0.0028, "step": 15723 }, { "epoch": 1.704683434518647, "grad_norm": 0.15564468502998352, "learning_rate": 8.66676338314232e-06, "loss": 0.0017, "step": 15724 }, { "epoch": 1.7047918473547268, "grad_norm": 1.4766067266464233, "learning_rate": 8.666038009574932e-06, "loss": 0.0132, "step": 15725 }, { "epoch": 1.7049002601908065, "grad_norm": 0.37318599224090576, "learning_rate": 8.665312636007544e-06, "loss": 0.0083, "step": 15726 }, { "epoch": 1.7050086730268865, "grad_norm": 0.5105865597724915, "learning_rate": 8.664587262440158e-06, "loss": 0.003, "step": 15727 }, { "epoch": 1.705117085862966, "grad_norm": 9.633598327636719, "learning_rate": 8.66386188887277e-06, "loss": 0.0513, "step": 15728 }, { "epoch": 1.705225498699046, "grad_norm": 3.2848961353302, "learning_rate": 8.663136515305382e-06, "loss": 0.0275, "step": 15729 }, { "epoch": 1.7053339115351258, "grad_norm": 1.0106269121170044, "learning_rate": 8.662411141737996e-06, "loss": 0.0409, "step": 15730 }, { "epoch": 1.7054423243712056, "grad_norm": 4.9278435707092285, "learning_rate": 8.66168576817061e-06, "loss": 0.029, "step": 15731 }, { "epoch": 1.7055507372072853, "grad_norm": 0.8050466179847717, "learning_rate": 8.660960394603222e-06, "loss": 0.0098, "step": 15732 }, { "epoch": 1.705659150043365, "grad_norm": 1.3333560228347778, "learning_rate": 8.660235021035834e-06, "loss": 0.0011, "step": 15733 }, { "epoch": 1.705767562879445, "grad_norm": 2.281282424926758, "learning_rate": 8.659509647468446e-06, "loss": 0.0105, "step": 15734 }, { "epoch": 1.7058759757155246, "grad_norm": 0.3875800371170044, "learning_rate": 8.65878427390106e-06, "loss": 0.0031, "step": 15735 }, { "epoch": 1.7059843885516046, "grad_norm": 2.354724407196045, "learning_rate": 8.658058900333672e-06, "loss": 0.0117, "step": 15736 }, { "epoch": 1.7060928013876842, "grad_norm": 1.1956356763839722, "learning_rate": 8.657333526766284e-06, "loss": 0.0054, "step": 15737 }, { "epoch": 1.7062012142237641, "grad_norm": 1.3496644496917725, "learning_rate": 8.656608153198898e-06, "loss": 0.0232, "step": 15738 }, { "epoch": 1.706309627059844, "grad_norm": 0.5182804465293884, "learning_rate": 8.655882779631512e-06, "loss": 0.0116, "step": 15739 }, { "epoch": 1.7064180398959237, "grad_norm": 0.937991738319397, "learning_rate": 8.655157406064124e-06, "loss": 0.0066, "step": 15740 }, { "epoch": 1.7065264527320034, "grad_norm": 3.6070218086242676, "learning_rate": 8.654432032496736e-06, "loss": 0.0302, "step": 15741 }, { "epoch": 1.7066348655680832, "grad_norm": 0.5501512289047241, "learning_rate": 8.65370665892935e-06, "loss": 0.0053, "step": 15742 }, { "epoch": 1.7067432784041632, "grad_norm": 0.7964781522750854, "learning_rate": 8.652981285361962e-06, "loss": 0.0124, "step": 15743 }, { "epoch": 1.7068516912402427, "grad_norm": 1.4230467081069946, "learning_rate": 8.652255911794574e-06, "loss": 0.0586, "step": 15744 }, { "epoch": 1.7069601040763227, "grad_norm": 7.1316070556640625, "learning_rate": 8.651530538227187e-06, "loss": 0.0315, "step": 15745 }, { "epoch": 1.7070685169124025, "grad_norm": 0.3566938042640686, "learning_rate": 8.6508051646598e-06, "loss": 0.0092, "step": 15746 }, { "epoch": 1.7071769297484822, "grad_norm": 0.6295198202133179, "learning_rate": 8.650079791092413e-06, "loss": 0.0223, "step": 15747 }, { "epoch": 1.707285342584562, "grad_norm": 0.5409960746765137, "learning_rate": 8.649354417525026e-06, "loss": 0.01, "step": 15748 }, { "epoch": 1.7073937554206418, "grad_norm": 8.090262413024902, "learning_rate": 8.648629043957639e-06, "loss": 0.0643, "step": 15749 }, { "epoch": 1.7075021682567217, "grad_norm": 1.8871567249298096, "learning_rate": 8.647903670390252e-06, "loss": 0.029, "step": 15750 }, { "epoch": 1.7076105810928013, "grad_norm": 0.18021085858345032, "learning_rate": 8.647178296822865e-06, "loss": 0.0009, "step": 15751 }, { "epoch": 1.7077189939288813, "grad_norm": 1.229019045829773, "learning_rate": 8.646452923255477e-06, "loss": 0.0274, "step": 15752 }, { "epoch": 1.7078274067649608, "grad_norm": 0.9528196454048157, "learning_rate": 8.64572754968809e-06, "loss": 0.0077, "step": 15753 }, { "epoch": 1.7079358196010408, "grad_norm": 2.973667860031128, "learning_rate": 8.645002176120703e-06, "loss": 0.0851, "step": 15754 }, { "epoch": 1.7080442324371206, "grad_norm": 1.9699134826660156, "learning_rate": 8.644276802553315e-06, "loss": 0.0527, "step": 15755 }, { "epoch": 1.7081526452732003, "grad_norm": 2.871464490890503, "learning_rate": 8.643551428985929e-06, "loss": 0.0369, "step": 15756 }, { "epoch": 1.70826105810928, "grad_norm": 0.26223745942115784, "learning_rate": 8.642826055418543e-06, "loss": 0.0056, "step": 15757 }, { "epoch": 1.7083694709453598, "grad_norm": 0.46152234077453613, "learning_rate": 8.642100681851155e-06, "loss": 0.0077, "step": 15758 }, { "epoch": 1.7084778837814398, "grad_norm": 2.662764549255371, "learning_rate": 8.641375308283767e-06, "loss": 0.0378, "step": 15759 }, { "epoch": 1.7085862966175194, "grad_norm": 0.9972325563430786, "learning_rate": 8.640649934716379e-06, "loss": 0.0027, "step": 15760 }, { "epoch": 1.7086947094535994, "grad_norm": 0.2942749261856079, "learning_rate": 8.639924561148993e-06, "loss": 0.0106, "step": 15761 }, { "epoch": 1.7088031222896791, "grad_norm": 1.07906973361969, "learning_rate": 8.639199187581605e-06, "loss": 0.0401, "step": 15762 }, { "epoch": 1.7089115351257589, "grad_norm": 0.7926595211029053, "learning_rate": 8.638473814014217e-06, "loss": 0.0265, "step": 15763 }, { "epoch": 1.7090199479618386, "grad_norm": 2.610973834991455, "learning_rate": 8.637748440446831e-06, "loss": 0.0088, "step": 15764 }, { "epoch": 1.7091283607979184, "grad_norm": 0.3944208025932312, "learning_rate": 8.637023066879445e-06, "loss": 0.0139, "step": 15765 }, { "epoch": 1.7092367736339984, "grad_norm": 1.946732521057129, "learning_rate": 8.636297693312057e-06, "loss": 0.038, "step": 15766 }, { "epoch": 1.709345186470078, "grad_norm": 5.464410305023193, "learning_rate": 8.635572319744669e-06, "loss": 0.0096, "step": 15767 }, { "epoch": 1.709453599306158, "grad_norm": 0.5471584796905518, "learning_rate": 8.634846946177283e-06, "loss": 0.0014, "step": 15768 }, { "epoch": 1.7095620121422377, "grad_norm": 0.5374216437339783, "learning_rate": 8.634121572609895e-06, "loss": 0.0025, "step": 15769 }, { "epoch": 1.7096704249783174, "grad_norm": 1.2173792123794556, "learning_rate": 8.633396199042507e-06, "loss": 0.0058, "step": 15770 }, { "epoch": 1.7097788378143972, "grad_norm": 2.065753221511841, "learning_rate": 8.63267082547512e-06, "loss": 0.0312, "step": 15771 }, { "epoch": 1.709887250650477, "grad_norm": 1.2780309915542603, "learning_rate": 8.631945451907733e-06, "loss": 0.0157, "step": 15772 }, { "epoch": 1.709995663486557, "grad_norm": 2.279691457748413, "learning_rate": 8.631220078340345e-06, "loss": 0.0105, "step": 15773 }, { "epoch": 1.7101040763226365, "grad_norm": 2.1048736572265625, "learning_rate": 8.630494704772959e-06, "loss": 0.0347, "step": 15774 }, { "epoch": 1.7102124891587165, "grad_norm": 0.1490352898836136, "learning_rate": 8.629769331205571e-06, "loss": 0.0015, "step": 15775 }, { "epoch": 1.710320901994796, "grad_norm": 4.971882343292236, "learning_rate": 8.629043957638185e-06, "loss": 0.0528, "step": 15776 }, { "epoch": 1.710429314830876, "grad_norm": 1.223503828048706, "learning_rate": 8.628318584070797e-06, "loss": 0.026, "step": 15777 }, { "epoch": 1.7105377276669558, "grad_norm": 5.4751105308532715, "learning_rate": 8.62759321050341e-06, "loss": 0.0363, "step": 15778 }, { "epoch": 1.7106461405030355, "grad_norm": 1.296781063079834, "learning_rate": 8.626867836936023e-06, "loss": 0.0112, "step": 15779 }, { "epoch": 1.7107545533391153, "grad_norm": 3.707904577255249, "learning_rate": 8.626142463368635e-06, "loss": 0.0081, "step": 15780 }, { "epoch": 1.710862966175195, "grad_norm": 0.6016910672187805, "learning_rate": 8.625417089801248e-06, "loss": 0.0042, "step": 15781 }, { "epoch": 1.710971379011275, "grad_norm": 2.2386558055877686, "learning_rate": 8.624691716233861e-06, "loss": 0.0045, "step": 15782 }, { "epoch": 1.7110797918473546, "grad_norm": 0.5645791888237, "learning_rate": 8.623966342666475e-06, "loss": 0.003, "step": 15783 }, { "epoch": 1.7111882046834346, "grad_norm": 1.9622148275375366, "learning_rate": 8.623240969099087e-06, "loss": 0.0115, "step": 15784 }, { "epoch": 1.7112966175195143, "grad_norm": 2.1725106239318848, "learning_rate": 8.6225155955317e-06, "loss": 0.0271, "step": 15785 }, { "epoch": 1.711405030355594, "grad_norm": 1.2837038040161133, "learning_rate": 8.621790221964312e-06, "loss": 0.0051, "step": 15786 }, { "epoch": 1.7115134431916739, "grad_norm": 9.32487964630127, "learning_rate": 8.621064848396925e-06, "loss": 0.1155, "step": 15787 }, { "epoch": 1.7116218560277536, "grad_norm": 0.2662983536720276, "learning_rate": 8.620339474829538e-06, "loss": 0.0082, "step": 15788 }, { "epoch": 1.7117302688638336, "grad_norm": 4.871973037719727, "learning_rate": 8.61961410126215e-06, "loss": 0.0162, "step": 15789 }, { "epoch": 1.7118386816999132, "grad_norm": 2.3985016345977783, "learning_rate": 8.618888727694764e-06, "loss": 0.0096, "step": 15790 }, { "epoch": 1.7119470945359931, "grad_norm": 2.916564464569092, "learning_rate": 8.618163354127377e-06, "loss": 0.0219, "step": 15791 }, { "epoch": 1.712055507372073, "grad_norm": 2.070352792739868, "learning_rate": 8.61743798055999e-06, "loss": 0.0172, "step": 15792 }, { "epoch": 1.7121639202081527, "grad_norm": 0.2930643558502197, "learning_rate": 8.616712606992602e-06, "loss": 0.0072, "step": 15793 }, { "epoch": 1.7122723330442324, "grad_norm": 2.991781234741211, "learning_rate": 8.615987233425216e-06, "loss": 0.0118, "step": 15794 }, { "epoch": 1.7123807458803122, "grad_norm": 3.0521044731140137, "learning_rate": 8.615261859857828e-06, "loss": 0.0068, "step": 15795 }, { "epoch": 1.7124891587163922, "grad_norm": 1.197708010673523, "learning_rate": 8.61453648629044e-06, "loss": 0.0279, "step": 15796 }, { "epoch": 1.7125975715524717, "grad_norm": 1.4840680360794067, "learning_rate": 8.613811112723052e-06, "loss": 0.0515, "step": 15797 }, { "epoch": 1.7127059843885517, "grad_norm": 2.273135185241699, "learning_rate": 8.613085739155666e-06, "loss": 0.0352, "step": 15798 }, { "epoch": 1.7128143972246312, "grad_norm": 0.548369824886322, "learning_rate": 8.612360365588278e-06, "loss": 0.018, "step": 15799 }, { "epoch": 1.7129228100607112, "grad_norm": 1.8365763425827026, "learning_rate": 8.611634992020892e-06, "loss": 0.0498, "step": 15800 }, { "epoch": 1.713031222896791, "grad_norm": 0.13767895102500916, "learning_rate": 8.610909618453504e-06, "loss": 0.0015, "step": 15801 }, { "epoch": 1.7131396357328708, "grad_norm": 4.725630760192871, "learning_rate": 8.610184244886118e-06, "loss": 0.0345, "step": 15802 }, { "epoch": 1.7132480485689505, "grad_norm": 4.624423503875732, "learning_rate": 8.60945887131873e-06, "loss": 0.0751, "step": 15803 }, { "epoch": 1.7133564614050303, "grad_norm": 3.2140798568725586, "learning_rate": 8.608733497751342e-06, "loss": 0.0178, "step": 15804 }, { "epoch": 1.7134648742411103, "grad_norm": 0.29423847794532776, "learning_rate": 8.608008124183956e-06, "loss": 0.0062, "step": 15805 }, { "epoch": 1.7135732870771898, "grad_norm": 0.7636950016021729, "learning_rate": 8.607282750616568e-06, "loss": 0.003, "step": 15806 }, { "epoch": 1.7136816999132698, "grad_norm": 0.08736766129732132, "learning_rate": 8.60655737704918e-06, "loss": 0.0006, "step": 15807 }, { "epoch": 1.7137901127493496, "grad_norm": 5.915948390960693, "learning_rate": 8.605832003481794e-06, "loss": 0.0579, "step": 15808 }, { "epoch": 1.7138985255854293, "grad_norm": 2.4216086864471436, "learning_rate": 8.605106629914408e-06, "loss": 0.0017, "step": 15809 }, { "epoch": 1.714006938421509, "grad_norm": 1.903836727142334, "learning_rate": 8.60438125634702e-06, "loss": 0.0184, "step": 15810 }, { "epoch": 1.7141153512575888, "grad_norm": 2.1133368015289307, "learning_rate": 8.603655882779632e-06, "loss": 0.0463, "step": 15811 }, { "epoch": 1.7142237640936688, "grad_norm": 8.765978813171387, "learning_rate": 8.602930509212244e-06, "loss": 0.0546, "step": 15812 }, { "epoch": 1.7143321769297484, "grad_norm": 1.1524972915649414, "learning_rate": 8.602205135644858e-06, "loss": 0.009, "step": 15813 }, { "epoch": 1.7144405897658284, "grad_norm": 1.560789704322815, "learning_rate": 8.60147976207747e-06, "loss": 0.0127, "step": 15814 }, { "epoch": 1.714549002601908, "grad_norm": 4.438934326171875, "learning_rate": 8.600754388510082e-06, "loss": 0.0305, "step": 15815 }, { "epoch": 1.7146574154379879, "grad_norm": 1.8576427698135376, "learning_rate": 8.600029014942696e-06, "loss": 0.0751, "step": 15816 }, { "epoch": 1.7147658282740676, "grad_norm": 1.446632742881775, "learning_rate": 8.59930364137531e-06, "loss": 0.0174, "step": 15817 }, { "epoch": 1.7148742411101474, "grad_norm": 1.06002676486969, "learning_rate": 8.598578267807922e-06, "loss": 0.0151, "step": 15818 }, { "epoch": 1.7149826539462274, "grad_norm": 1.9934008121490479, "learning_rate": 8.597852894240534e-06, "loss": 0.0173, "step": 15819 }, { "epoch": 1.715091066782307, "grad_norm": 2.1718311309814453, "learning_rate": 8.597127520673148e-06, "loss": 0.014, "step": 15820 }, { "epoch": 1.715199479618387, "grad_norm": 1.6033198833465576, "learning_rate": 8.59640214710576e-06, "loss": 0.0143, "step": 15821 }, { "epoch": 1.7153078924544665, "grad_norm": 1.0100146532058716, "learning_rate": 8.595676773538373e-06, "loss": 0.0227, "step": 15822 }, { "epoch": 1.7154163052905465, "grad_norm": 2.9703609943389893, "learning_rate": 8.594951399970985e-06, "loss": 0.0273, "step": 15823 }, { "epoch": 1.7155247181266262, "grad_norm": 5.920437812805176, "learning_rate": 8.594226026403598e-06, "loss": 0.0647, "step": 15824 }, { "epoch": 1.715633130962706, "grad_norm": 0.7186580300331116, "learning_rate": 8.59350065283621e-06, "loss": 0.0202, "step": 15825 }, { "epoch": 1.7157415437987857, "grad_norm": 0.23113323748111725, "learning_rate": 8.592775279268824e-06, "loss": 0.0017, "step": 15826 }, { "epoch": 1.7158499566348655, "grad_norm": 0.01269131526350975, "learning_rate": 8.592049905701437e-06, "loss": 0.0002, "step": 15827 }, { "epoch": 1.7159583694709455, "grad_norm": 1.5746169090270996, "learning_rate": 8.59132453213405e-06, "loss": 0.0087, "step": 15828 }, { "epoch": 1.716066782307025, "grad_norm": 1.985710620880127, "learning_rate": 8.590599158566663e-06, "loss": 0.03, "step": 15829 }, { "epoch": 1.716175195143105, "grad_norm": 1.5961464643478394, "learning_rate": 8.589873784999275e-06, "loss": 0.0114, "step": 15830 }, { "epoch": 1.7162836079791848, "grad_norm": 0.24271902441978455, "learning_rate": 8.589148411431889e-06, "loss": 0.0023, "step": 15831 }, { "epoch": 1.7163920208152645, "grad_norm": 0.3091757297515869, "learning_rate": 8.5884230378645e-06, "loss": 0.0076, "step": 15832 }, { "epoch": 1.7165004336513443, "grad_norm": 2.2713265419006348, "learning_rate": 8.587697664297113e-06, "loss": 0.0271, "step": 15833 }, { "epoch": 1.716608846487424, "grad_norm": 1.7396551370620728, "learning_rate": 8.586972290729727e-06, "loss": 0.0416, "step": 15834 }, { "epoch": 1.716717259323504, "grad_norm": 2.416806221008301, "learning_rate": 8.58624691716234e-06, "loss": 0.0111, "step": 15835 }, { "epoch": 1.7168256721595836, "grad_norm": 2.3760299682617188, "learning_rate": 8.585521543594953e-06, "loss": 0.0111, "step": 15836 }, { "epoch": 1.7169340849956636, "grad_norm": 0.8236096501350403, "learning_rate": 8.584796170027565e-06, "loss": 0.0073, "step": 15837 }, { "epoch": 1.7170424978317431, "grad_norm": 1.7369006872177124, "learning_rate": 8.584070796460177e-06, "loss": 0.0122, "step": 15838 }, { "epoch": 1.717150910667823, "grad_norm": 3.3070943355560303, "learning_rate": 8.58334542289279e-06, "loss": 0.0179, "step": 15839 }, { "epoch": 1.7172593235039029, "grad_norm": 0.41948968172073364, "learning_rate": 8.582620049325403e-06, "loss": 0.02, "step": 15840 }, { "epoch": 1.7173677363399826, "grad_norm": 1.6286927461624146, "learning_rate": 8.581894675758015e-06, "loss": 0.0263, "step": 15841 }, { "epoch": 1.7174761491760624, "grad_norm": 1.296305537223816, "learning_rate": 8.581169302190629e-06, "loss": 0.0251, "step": 15842 }, { "epoch": 1.7175845620121422, "grad_norm": 3.1453981399536133, "learning_rate": 8.580443928623243e-06, "loss": 0.0298, "step": 15843 }, { "epoch": 1.7176929748482221, "grad_norm": 0.3882908225059509, "learning_rate": 8.579718555055855e-06, "loss": 0.0027, "step": 15844 }, { "epoch": 1.7178013876843017, "grad_norm": 0.3506493866443634, "learning_rate": 8.578993181488467e-06, "loss": 0.0073, "step": 15845 }, { "epoch": 1.7179098005203817, "grad_norm": 0.7687737345695496, "learning_rate": 8.578267807921081e-06, "loss": 0.0231, "step": 15846 }, { "epoch": 1.7180182133564614, "grad_norm": 0.5997355580329895, "learning_rate": 8.577542434353693e-06, "loss": 0.0144, "step": 15847 }, { "epoch": 1.7181266261925412, "grad_norm": 0.5686501264572144, "learning_rate": 8.576817060786305e-06, "loss": 0.0108, "step": 15848 }, { "epoch": 1.718235039028621, "grad_norm": 10.433791160583496, "learning_rate": 8.576091687218919e-06, "loss": 0.0257, "step": 15849 }, { "epoch": 1.7183434518647007, "grad_norm": 1.566926121711731, "learning_rate": 8.575366313651531e-06, "loss": 0.0118, "step": 15850 }, { "epoch": 1.7184518647007807, "grad_norm": 1.2701719999313354, "learning_rate": 8.574640940084143e-06, "loss": 0.0155, "step": 15851 }, { "epoch": 1.7185602775368602, "grad_norm": 1.546052098274231, "learning_rate": 8.573915566516757e-06, "loss": 0.0186, "step": 15852 }, { "epoch": 1.7186686903729402, "grad_norm": 1.8304450511932373, "learning_rate": 8.57319019294937e-06, "loss": 0.0173, "step": 15853 }, { "epoch": 1.71877710320902, "grad_norm": 3.8055014610290527, "learning_rate": 8.572464819381983e-06, "loss": 0.0721, "step": 15854 }, { "epoch": 1.7188855160450998, "grad_norm": 0.4987938404083252, "learning_rate": 8.571739445814595e-06, "loss": 0.0023, "step": 15855 }, { "epoch": 1.7189939288811795, "grad_norm": 3.7703328132629395, "learning_rate": 8.571014072247207e-06, "loss": 0.0211, "step": 15856 }, { "epoch": 1.7191023417172593, "grad_norm": 0.6919175982475281, "learning_rate": 8.570288698679821e-06, "loss": 0.0087, "step": 15857 }, { "epoch": 1.7192107545533393, "grad_norm": 3.1497912406921387, "learning_rate": 8.569563325112433e-06, "loss": 0.0682, "step": 15858 }, { "epoch": 1.7193191673894188, "grad_norm": 0.4570425748825073, "learning_rate": 8.568837951545046e-06, "loss": 0.0041, "step": 15859 }, { "epoch": 1.7194275802254988, "grad_norm": 3.3345794677734375, "learning_rate": 8.56811257797766e-06, "loss": 0.0226, "step": 15860 }, { "epoch": 1.7195359930615783, "grad_norm": 1.574837565422058, "learning_rate": 8.567387204410273e-06, "loss": 0.0195, "step": 15861 }, { "epoch": 1.7196444058976583, "grad_norm": 4.331833362579346, "learning_rate": 8.566661830842885e-06, "loss": 0.0285, "step": 15862 }, { "epoch": 1.719752818733738, "grad_norm": 0.31070154905319214, "learning_rate": 8.565936457275497e-06, "loss": 0.0092, "step": 15863 }, { "epoch": 1.7198612315698178, "grad_norm": 0.2769026458263397, "learning_rate": 8.56521108370811e-06, "loss": 0.014, "step": 15864 }, { "epoch": 1.7199696444058976, "grad_norm": 4.381880283355713, "learning_rate": 8.564485710140723e-06, "loss": 0.0352, "step": 15865 }, { "epoch": 1.7200780572419774, "grad_norm": 4.740592956542969, "learning_rate": 8.563760336573336e-06, "loss": 0.0492, "step": 15866 }, { "epoch": 1.7201864700780574, "grad_norm": 3.937652826309204, "learning_rate": 8.563034963005948e-06, "loss": 0.0078, "step": 15867 }, { "epoch": 1.720294882914137, "grad_norm": 0.6272957921028137, "learning_rate": 8.562309589438562e-06, "loss": 0.0089, "step": 15868 }, { "epoch": 1.7204032957502169, "grad_norm": 4.499197959899902, "learning_rate": 8.561584215871175e-06, "loss": 0.0159, "step": 15869 }, { "epoch": 1.7205117085862967, "grad_norm": 2.248424768447876, "learning_rate": 8.560858842303788e-06, "loss": 0.0247, "step": 15870 }, { "epoch": 1.7206201214223764, "grad_norm": 5.86781644821167, "learning_rate": 8.5601334687364e-06, "loss": 0.0166, "step": 15871 }, { "epoch": 1.7207285342584562, "grad_norm": 0.401395320892334, "learning_rate": 8.559408095169014e-06, "loss": 0.0197, "step": 15872 }, { "epoch": 1.720836947094536, "grad_norm": 7.581278324127197, "learning_rate": 8.558682721601626e-06, "loss": 0.0954, "step": 15873 }, { "epoch": 1.720945359930616, "grad_norm": 3.4720592498779297, "learning_rate": 8.557957348034238e-06, "loss": 0.0313, "step": 15874 }, { "epoch": 1.7210537727666955, "grad_norm": 1.4666255712509155, "learning_rate": 8.557231974466852e-06, "loss": 0.0081, "step": 15875 }, { "epoch": 1.7211621856027755, "grad_norm": 1.8909010887145996, "learning_rate": 8.556506600899464e-06, "loss": 0.0371, "step": 15876 }, { "epoch": 1.7212705984388552, "grad_norm": 0.08775193244218826, "learning_rate": 8.555781227332076e-06, "loss": 0.0004, "step": 15877 }, { "epoch": 1.721379011274935, "grad_norm": 4.520221710205078, "learning_rate": 8.55505585376469e-06, "loss": 0.081, "step": 15878 }, { "epoch": 1.7214874241110147, "grad_norm": 13.173072814941406, "learning_rate": 8.554330480197302e-06, "loss": 0.019, "step": 15879 }, { "epoch": 1.7215958369470945, "grad_norm": 5.223920822143555, "learning_rate": 8.553605106629916e-06, "loss": 0.0422, "step": 15880 }, { "epoch": 1.7217042497831745, "grad_norm": 4.651695251464844, "learning_rate": 8.552879733062528e-06, "loss": 0.0336, "step": 15881 }, { "epoch": 1.721812662619254, "grad_norm": 2.1136698722839355, "learning_rate": 8.55215435949514e-06, "loss": 0.0496, "step": 15882 }, { "epoch": 1.721921075455334, "grad_norm": 6.569188117980957, "learning_rate": 8.551428985927754e-06, "loss": 0.0234, "step": 15883 }, { "epoch": 1.7220294882914136, "grad_norm": 0.49200886487960815, "learning_rate": 8.550703612360366e-06, "loss": 0.0099, "step": 15884 }, { "epoch": 1.7221379011274935, "grad_norm": 0.3576403260231018, "learning_rate": 8.549978238792978e-06, "loss": 0.0136, "step": 15885 }, { "epoch": 1.7222463139635733, "grad_norm": 3.6136133670806885, "learning_rate": 8.549252865225592e-06, "loss": 0.0137, "step": 15886 }, { "epoch": 1.722354726799653, "grad_norm": 0.348673552274704, "learning_rate": 8.548527491658206e-06, "loss": 0.007, "step": 15887 }, { "epoch": 1.7224631396357328, "grad_norm": 2.5535967350006104, "learning_rate": 8.547802118090818e-06, "loss": 0.0182, "step": 15888 }, { "epoch": 1.7225715524718126, "grad_norm": 2.0968711376190186, "learning_rate": 8.54707674452343e-06, "loss": 0.0217, "step": 15889 }, { "epoch": 1.7226799653078926, "grad_norm": 4.841239929199219, "learning_rate": 8.546351370956042e-06, "loss": 0.0251, "step": 15890 }, { "epoch": 1.7227883781439721, "grad_norm": 3.2341878414154053, "learning_rate": 8.545625997388656e-06, "loss": 0.0435, "step": 15891 }, { "epoch": 1.722896790980052, "grad_norm": 0.2964121997356415, "learning_rate": 8.544900623821268e-06, "loss": 0.0014, "step": 15892 }, { "epoch": 1.7230052038161319, "grad_norm": 0.9349746108055115, "learning_rate": 8.54417525025388e-06, "loss": 0.0037, "step": 15893 }, { "epoch": 1.7231136166522116, "grad_norm": 1.5410182476043701, "learning_rate": 8.543449876686494e-06, "loss": 0.0498, "step": 15894 }, { "epoch": 1.7232220294882914, "grad_norm": 0.6019025444984436, "learning_rate": 8.542724503119108e-06, "loss": 0.0102, "step": 15895 }, { "epoch": 1.7233304423243712, "grad_norm": 0.5539785623550415, "learning_rate": 8.54199912955172e-06, "loss": 0.0126, "step": 15896 }, { "epoch": 1.7234388551604511, "grad_norm": 0.8113543391227722, "learning_rate": 8.541273755984332e-06, "loss": 0.0325, "step": 15897 }, { "epoch": 1.7235472679965307, "grad_norm": 0.5449151992797852, "learning_rate": 8.540548382416946e-06, "loss": 0.0144, "step": 15898 }, { "epoch": 1.7236556808326107, "grad_norm": 1.4806104898452759, "learning_rate": 8.539823008849558e-06, "loss": 0.0209, "step": 15899 }, { "epoch": 1.7237640936686902, "grad_norm": 5.793206214904785, "learning_rate": 8.53909763528217e-06, "loss": 0.0246, "step": 15900 }, { "epoch": 1.7238725065047702, "grad_norm": 1.4737157821655273, "learning_rate": 8.538372261714784e-06, "loss": 0.0181, "step": 15901 }, { "epoch": 1.72398091934085, "grad_norm": 0.7212010622024536, "learning_rate": 8.537646888147396e-06, "loss": 0.0037, "step": 15902 }, { "epoch": 1.7240893321769297, "grad_norm": 2.3603968620300293, "learning_rate": 8.53692151458001e-06, "loss": 0.0299, "step": 15903 }, { "epoch": 1.7241977450130095, "grad_norm": 2.242375373840332, "learning_rate": 8.536196141012622e-06, "loss": 0.0124, "step": 15904 }, { "epoch": 1.7243061578490892, "grad_norm": 5.389328956604004, "learning_rate": 8.535470767445235e-06, "loss": 0.0249, "step": 15905 }, { "epoch": 1.7244145706851692, "grad_norm": 4.937601089477539, "learning_rate": 8.534745393877848e-06, "loss": 0.0412, "step": 15906 }, { "epoch": 1.7245229835212488, "grad_norm": 2.5270583629608154, "learning_rate": 8.53402002031046e-06, "loss": 0.0184, "step": 15907 }, { "epoch": 1.7246313963573288, "grad_norm": 1.1036142110824585, "learning_rate": 8.533294646743073e-06, "loss": 0.0067, "step": 15908 }, { "epoch": 1.7247398091934085, "grad_norm": 1.0290783643722534, "learning_rate": 8.532569273175687e-06, "loss": 0.0195, "step": 15909 }, { "epoch": 1.7248482220294883, "grad_norm": 0.8185569643974304, "learning_rate": 8.531843899608299e-06, "loss": 0.0075, "step": 15910 }, { "epoch": 1.724956634865568, "grad_norm": 2.333233594894409, "learning_rate": 8.531118526040911e-06, "loss": 0.0624, "step": 15911 }, { "epoch": 1.7250650477016478, "grad_norm": 1.0352320671081543, "learning_rate": 8.530393152473525e-06, "loss": 0.0251, "step": 15912 }, { "epoch": 1.7251734605377278, "grad_norm": 2.45870304107666, "learning_rate": 8.529667778906139e-06, "loss": 0.0487, "step": 15913 }, { "epoch": 1.7252818733738073, "grad_norm": 0.3010920286178589, "learning_rate": 8.52894240533875e-06, "loss": 0.0063, "step": 15914 }, { "epoch": 1.7253902862098873, "grad_norm": 0.11289025843143463, "learning_rate": 8.528217031771363e-06, "loss": 0.0011, "step": 15915 }, { "epoch": 1.725498699045967, "grad_norm": 0.22922970354557037, "learning_rate": 8.527491658203975e-06, "loss": 0.0132, "step": 15916 }, { "epoch": 1.7256071118820469, "grad_norm": 0.9086204171180725, "learning_rate": 8.526766284636589e-06, "loss": 0.007, "step": 15917 }, { "epoch": 1.7257155247181266, "grad_norm": 0.3007561266422272, "learning_rate": 8.526040911069201e-06, "loss": 0.0052, "step": 15918 }, { "epoch": 1.7258239375542064, "grad_norm": 4.63594388961792, "learning_rate": 8.525315537501813e-06, "loss": 0.0259, "step": 15919 }, { "epoch": 1.7259323503902864, "grad_norm": 3.464603900909424, "learning_rate": 8.524590163934427e-06, "loss": 0.022, "step": 15920 }, { "epoch": 1.726040763226366, "grad_norm": 0.2970826327800751, "learning_rate": 8.52386479036704e-06, "loss": 0.0065, "step": 15921 }, { "epoch": 1.726149176062446, "grad_norm": 6.045563220977783, "learning_rate": 8.523139416799653e-06, "loss": 0.0544, "step": 15922 }, { "epoch": 1.7262575888985254, "grad_norm": 5.997888088226318, "learning_rate": 8.522414043232265e-06, "loss": 0.0108, "step": 15923 }, { "epoch": 1.7263660017346054, "grad_norm": 0.126688152551651, "learning_rate": 8.521688669664879e-06, "loss": 0.0017, "step": 15924 }, { "epoch": 1.7264744145706852, "grad_norm": 4.148280620574951, "learning_rate": 8.520963296097491e-06, "loss": 0.0439, "step": 15925 }, { "epoch": 1.726582827406765, "grad_norm": 2.3374719619750977, "learning_rate": 8.520237922530103e-06, "loss": 0.0154, "step": 15926 }, { "epoch": 1.7266912402428447, "grad_norm": 2.2095448970794678, "learning_rate": 8.519512548962717e-06, "loss": 0.0224, "step": 15927 }, { "epoch": 1.7267996530789245, "grad_norm": 0.39971923828125, "learning_rate": 8.51878717539533e-06, "loss": 0.0037, "step": 15928 }, { "epoch": 1.7269080659150045, "grad_norm": 2.3828494548797607, "learning_rate": 8.518061801827943e-06, "loss": 0.0276, "step": 15929 }, { "epoch": 1.727016478751084, "grad_norm": 7.294161319732666, "learning_rate": 8.517336428260555e-06, "loss": 0.0241, "step": 15930 }, { "epoch": 1.727124891587164, "grad_norm": 2.861323118209839, "learning_rate": 8.516611054693167e-06, "loss": 0.013, "step": 15931 }, { "epoch": 1.7272333044232437, "grad_norm": 1.188820242881775, "learning_rate": 8.515885681125781e-06, "loss": 0.0218, "step": 15932 }, { "epoch": 1.7273417172593235, "grad_norm": 5.281932353973389, "learning_rate": 8.515160307558393e-06, "loss": 0.0189, "step": 15933 }, { "epoch": 1.7274501300954033, "grad_norm": 0.5523979067802429, "learning_rate": 8.514434933991005e-06, "loss": 0.0119, "step": 15934 }, { "epoch": 1.727558542931483, "grad_norm": 0.7081233263015747, "learning_rate": 8.51370956042362e-06, "loss": 0.0134, "step": 15935 }, { "epoch": 1.727666955767563, "grad_norm": 6.081823825836182, "learning_rate": 8.512984186856231e-06, "loss": 0.0542, "step": 15936 }, { "epoch": 1.7277753686036426, "grad_norm": 6.9949212074279785, "learning_rate": 8.512258813288844e-06, "loss": 0.0147, "step": 15937 }, { "epoch": 1.7278837814397225, "grad_norm": 0.4576459527015686, "learning_rate": 8.511533439721457e-06, "loss": 0.0119, "step": 15938 }, { "epoch": 1.7279921942758023, "grad_norm": 2.8114731311798096, "learning_rate": 8.510808066154071e-06, "loss": 0.05, "step": 15939 }, { "epoch": 1.728100607111882, "grad_norm": 2.482726573944092, "learning_rate": 8.510082692586683e-06, "loss": 0.0328, "step": 15940 }, { "epoch": 1.7282090199479618, "grad_norm": 0.07597978413105011, "learning_rate": 8.509357319019295e-06, "loss": 0.0006, "step": 15941 }, { "epoch": 1.7283174327840416, "grad_norm": 0.910781979560852, "learning_rate": 8.508631945451908e-06, "loss": 0.019, "step": 15942 }, { "epoch": 1.7284258456201216, "grad_norm": 3.4928171634674072, "learning_rate": 8.507906571884521e-06, "loss": 0.0219, "step": 15943 }, { "epoch": 1.7285342584562011, "grad_norm": 4.239562034606934, "learning_rate": 8.507181198317134e-06, "loss": 0.0265, "step": 15944 }, { "epoch": 1.728642671292281, "grad_norm": 11.423895835876465, "learning_rate": 8.506455824749746e-06, "loss": 0.0503, "step": 15945 }, { "epoch": 1.7287510841283606, "grad_norm": 0.513465166091919, "learning_rate": 8.50573045118236e-06, "loss": 0.0182, "step": 15946 }, { "epoch": 1.7288594969644406, "grad_norm": 3.7451906204223633, "learning_rate": 8.505005077614973e-06, "loss": 0.0225, "step": 15947 }, { "epoch": 1.7289679098005204, "grad_norm": 1.1851528882980347, "learning_rate": 8.504279704047586e-06, "loss": 0.0199, "step": 15948 }, { "epoch": 1.7290763226366002, "grad_norm": 3.614959955215454, "learning_rate": 8.503554330480198e-06, "loss": 0.0083, "step": 15949 }, { "epoch": 1.72918473547268, "grad_norm": 4.626496315002441, "learning_rate": 8.502828956912812e-06, "loss": 0.0766, "step": 15950 }, { "epoch": 1.7292931483087597, "grad_norm": 1.4622429609298706, "learning_rate": 8.502103583345424e-06, "loss": 0.0046, "step": 15951 }, { "epoch": 1.7294015611448397, "grad_norm": 2.2768638134002686, "learning_rate": 8.501378209778036e-06, "loss": 0.0053, "step": 15952 }, { "epoch": 1.7295099739809192, "grad_norm": 8.024125099182129, "learning_rate": 8.50065283621065e-06, "loss": 0.0673, "step": 15953 }, { "epoch": 1.7296183868169992, "grad_norm": 1.3215017318725586, "learning_rate": 8.499927462643262e-06, "loss": 0.0392, "step": 15954 }, { "epoch": 1.729726799653079, "grad_norm": 0.28759118914604187, "learning_rate": 8.499202089075876e-06, "loss": 0.0081, "step": 15955 }, { "epoch": 1.7298352124891587, "grad_norm": 2.036097526550293, "learning_rate": 8.498476715508488e-06, "loss": 0.0329, "step": 15956 }, { "epoch": 1.7299436253252385, "grad_norm": 5.158695697784424, "learning_rate": 8.4977513419411e-06, "loss": 0.0597, "step": 15957 }, { "epoch": 1.7300520381613183, "grad_norm": 0.9920631647109985, "learning_rate": 8.497025968373714e-06, "loss": 0.0213, "step": 15958 }, { "epoch": 1.7301604509973982, "grad_norm": 0.30461427569389343, "learning_rate": 8.496300594806326e-06, "loss": 0.0051, "step": 15959 }, { "epoch": 1.7302688638334778, "grad_norm": 13.737156867980957, "learning_rate": 8.495575221238938e-06, "loss": 0.0302, "step": 15960 }, { "epoch": 1.7303772766695578, "grad_norm": 1.7075449228286743, "learning_rate": 8.494849847671552e-06, "loss": 0.011, "step": 15961 }, { "epoch": 1.7304856895056373, "grad_norm": 4.149733543395996, "learning_rate": 8.494124474104164e-06, "loss": 0.0903, "step": 15962 }, { "epoch": 1.7305941023417173, "grad_norm": 3.9920146465301514, "learning_rate": 8.493399100536776e-06, "loss": 0.0504, "step": 15963 }, { "epoch": 1.730702515177797, "grad_norm": 8.704174041748047, "learning_rate": 8.49267372696939e-06, "loss": 0.0617, "step": 15964 }, { "epoch": 1.7308109280138768, "grad_norm": 2.238420248031616, "learning_rate": 8.491948353402004e-06, "loss": 0.008, "step": 15965 }, { "epoch": 1.7309193408499568, "grad_norm": 2.857436180114746, "learning_rate": 8.491222979834616e-06, "loss": 0.0292, "step": 15966 }, { "epoch": 1.7310277536860363, "grad_norm": 8.575277328491211, "learning_rate": 8.490497606267228e-06, "loss": 0.0629, "step": 15967 }, { "epoch": 1.7311361665221163, "grad_norm": 5.862160682678223, "learning_rate": 8.48977223269984e-06, "loss": 0.0123, "step": 15968 }, { "epoch": 1.7312445793581959, "grad_norm": 2.347874402999878, "learning_rate": 8.489046859132454e-06, "loss": 0.0211, "step": 15969 }, { "epoch": 1.7313529921942759, "grad_norm": 10.162626266479492, "learning_rate": 8.488321485565066e-06, "loss": 0.031, "step": 15970 }, { "epoch": 1.7314614050303556, "grad_norm": 1.9041897058486938, "learning_rate": 8.487596111997678e-06, "loss": 0.0149, "step": 15971 }, { "epoch": 1.7315698178664354, "grad_norm": 4.61678409576416, "learning_rate": 8.486870738430292e-06, "loss": 0.0139, "step": 15972 }, { "epoch": 1.7316782307025151, "grad_norm": 1.9089280366897583, "learning_rate": 8.486145364862906e-06, "loss": 0.0369, "step": 15973 }, { "epoch": 1.731786643538595, "grad_norm": 0.3010008931159973, "learning_rate": 8.485419991295518e-06, "loss": 0.0096, "step": 15974 }, { "epoch": 1.731895056374675, "grad_norm": 0.8678640127182007, "learning_rate": 8.48469461772813e-06, "loss": 0.0056, "step": 15975 }, { "epoch": 1.7320034692107544, "grad_norm": 1.4383043050765991, "learning_rate": 8.483969244160744e-06, "loss": 0.0145, "step": 15976 }, { "epoch": 1.7321118820468344, "grad_norm": 1.5642579793930054, "learning_rate": 8.483243870593356e-06, "loss": 0.0273, "step": 15977 }, { "epoch": 1.7322202948829142, "grad_norm": 1.5085357427597046, "learning_rate": 8.482518497025969e-06, "loss": 0.006, "step": 15978 }, { "epoch": 1.732328707718994, "grad_norm": 1.9288216829299927, "learning_rate": 8.481793123458582e-06, "loss": 0.0431, "step": 15979 }, { "epoch": 1.7324371205550737, "grad_norm": 2.6573843955993652, "learning_rate": 8.481067749891194e-06, "loss": 0.0145, "step": 15980 }, { "epoch": 1.7325455333911535, "grad_norm": 6.31217622756958, "learning_rate": 8.480342376323808e-06, "loss": 0.0092, "step": 15981 }, { "epoch": 1.7326539462272335, "grad_norm": 1.5910232067108154, "learning_rate": 8.47961700275642e-06, "loss": 0.0229, "step": 15982 }, { "epoch": 1.732762359063313, "grad_norm": 3.1935606002807617, "learning_rate": 8.478891629189033e-06, "loss": 0.0256, "step": 15983 }, { "epoch": 1.732870771899393, "grad_norm": 7.19254207611084, "learning_rate": 8.478166255621646e-06, "loss": 0.0887, "step": 15984 }, { "epoch": 1.7329791847354725, "grad_norm": 0.23295995593070984, "learning_rate": 8.477440882054259e-06, "loss": 0.002, "step": 15985 }, { "epoch": 1.7330875975715525, "grad_norm": 1.7612519264221191, "learning_rate": 8.47671550848687e-06, "loss": 0.0185, "step": 15986 }, { "epoch": 1.7331960104076323, "grad_norm": 1.8176790475845337, "learning_rate": 8.475990134919485e-06, "loss": 0.014, "step": 15987 }, { "epoch": 1.733304423243712, "grad_norm": 3.969895124435425, "learning_rate": 8.475264761352097e-06, "loss": 0.0253, "step": 15988 }, { "epoch": 1.7334128360797918, "grad_norm": 4.731590747833252, "learning_rate": 8.474539387784709e-06, "loss": 0.0763, "step": 15989 }, { "epoch": 1.7335212489158716, "grad_norm": 1.9801441431045532, "learning_rate": 8.473814014217323e-06, "loss": 0.0306, "step": 15990 }, { "epoch": 1.7336296617519515, "grad_norm": 0.7150025963783264, "learning_rate": 8.473088640649937e-06, "loss": 0.025, "step": 15991 }, { "epoch": 1.733738074588031, "grad_norm": 1.4537707567214966, "learning_rate": 8.472363267082549e-06, "loss": 0.0192, "step": 15992 }, { "epoch": 1.733846487424111, "grad_norm": 3.6471238136291504, "learning_rate": 8.47163789351516e-06, "loss": 0.0075, "step": 15993 }, { "epoch": 1.7339549002601908, "grad_norm": 1.8270219564437866, "learning_rate": 8.470912519947773e-06, "loss": 0.0222, "step": 15994 }, { "epoch": 1.7340633130962706, "grad_norm": 2.1054155826568604, "learning_rate": 8.470187146380387e-06, "loss": 0.0465, "step": 15995 }, { "epoch": 1.7341717259323504, "grad_norm": 1.2957066297531128, "learning_rate": 8.469461772812999e-06, "loss": 0.0185, "step": 15996 }, { "epoch": 1.7342801387684301, "grad_norm": 3.29443621635437, "learning_rate": 8.468736399245611e-06, "loss": 0.0544, "step": 15997 }, { "epoch": 1.73438855160451, "grad_norm": 2.228027582168579, "learning_rate": 8.468011025678225e-06, "loss": 0.0206, "step": 15998 }, { "epoch": 1.7344969644405897, "grad_norm": 5.95639705657959, "learning_rate": 8.467285652110839e-06, "loss": 0.0294, "step": 15999 }, { "epoch": 1.7346053772766696, "grad_norm": 0.9071144461631775, "learning_rate": 8.466560278543451e-06, "loss": 0.012, "step": 16000 }, { "epoch": 1.7347137901127494, "grad_norm": 3.8503735065460205, "learning_rate": 8.465834904976063e-06, "loss": 0.046, "step": 16001 }, { "epoch": 1.7348222029488292, "grad_norm": 3.9070398807525635, "learning_rate": 8.465109531408677e-06, "loss": 0.0391, "step": 16002 }, { "epoch": 1.734930615784909, "grad_norm": 2.0010268688201904, "learning_rate": 8.464384157841289e-06, "loss": 0.0359, "step": 16003 }, { "epoch": 1.7350390286209887, "grad_norm": 2.815840244293213, "learning_rate": 8.463658784273901e-06, "loss": 0.0229, "step": 16004 }, { "epoch": 1.7351474414570687, "grad_norm": 3.0598692893981934, "learning_rate": 8.462933410706515e-06, "loss": 0.0685, "step": 16005 }, { "epoch": 1.7352558542931482, "grad_norm": 2.808056354522705, "learning_rate": 8.462208037139127e-06, "loss": 0.0062, "step": 16006 }, { "epoch": 1.7353642671292282, "grad_norm": 0.41070520877838135, "learning_rate": 8.461482663571741e-06, "loss": 0.0027, "step": 16007 }, { "epoch": 1.7354726799653077, "grad_norm": 0.918938159942627, "learning_rate": 8.460757290004353e-06, "loss": 0.0116, "step": 16008 }, { "epoch": 1.7355810928013877, "grad_norm": 3.1401376724243164, "learning_rate": 8.460031916436965e-06, "loss": 0.0276, "step": 16009 }, { "epoch": 1.7356895056374675, "grad_norm": 0.4646398723125458, "learning_rate": 8.459306542869579e-06, "loss": 0.0205, "step": 16010 }, { "epoch": 1.7357979184735473, "grad_norm": 1.9390079975128174, "learning_rate": 8.458581169302191e-06, "loss": 0.0073, "step": 16011 }, { "epoch": 1.735906331309627, "grad_norm": 1.1072580814361572, "learning_rate": 8.457855795734803e-06, "loss": 0.0096, "step": 16012 }, { "epoch": 1.7360147441457068, "grad_norm": 0.43041178584098816, "learning_rate": 8.457130422167417e-06, "loss": 0.0114, "step": 16013 }, { "epoch": 1.7361231569817868, "grad_norm": 1.9017428159713745, "learning_rate": 8.45640504860003e-06, "loss": 0.0172, "step": 16014 }, { "epoch": 1.7362315698178663, "grad_norm": 2.8894309997558594, "learning_rate": 8.455679675032642e-06, "loss": 0.0139, "step": 16015 }, { "epoch": 1.7363399826539463, "grad_norm": 1.4782490730285645, "learning_rate": 8.454954301465255e-06, "loss": 0.0079, "step": 16016 }, { "epoch": 1.736448395490026, "grad_norm": 0.7029299139976501, "learning_rate": 8.45422892789787e-06, "loss": 0.0254, "step": 16017 }, { "epoch": 1.7365568083261058, "grad_norm": 2.9147162437438965, "learning_rate": 8.453503554330481e-06, "loss": 0.0541, "step": 16018 }, { "epoch": 1.7366652211621856, "grad_norm": 1.2291808128356934, "learning_rate": 8.452778180763093e-06, "loss": 0.0148, "step": 16019 }, { "epoch": 1.7367736339982653, "grad_norm": 6.526785373687744, "learning_rate": 8.452052807195706e-06, "loss": 0.0201, "step": 16020 }, { "epoch": 1.7368820468343453, "grad_norm": 0.27513250708580017, "learning_rate": 8.45132743362832e-06, "loss": 0.003, "step": 16021 }, { "epoch": 1.7369904596704249, "grad_norm": 4.480197429656982, "learning_rate": 8.450602060060932e-06, "loss": 0.015, "step": 16022 }, { "epoch": 1.7370988725065049, "grad_norm": 3.0992298126220703, "learning_rate": 8.449876686493544e-06, "loss": 0.0815, "step": 16023 }, { "epoch": 1.7372072853425846, "grad_norm": 2.102965831756592, "learning_rate": 8.449151312926158e-06, "loss": 0.0098, "step": 16024 }, { "epoch": 1.7373156981786644, "grad_norm": 1.1993565559387207, "learning_rate": 8.448425939358771e-06, "loss": 0.0347, "step": 16025 }, { "epoch": 1.7374241110147441, "grad_norm": 0.7974923253059387, "learning_rate": 8.447700565791384e-06, "loss": 0.0077, "step": 16026 }, { "epoch": 1.737532523850824, "grad_norm": 0.6418524980545044, "learning_rate": 8.446975192223996e-06, "loss": 0.0068, "step": 16027 }, { "epoch": 1.737640936686904, "grad_norm": 1.115614414215088, "learning_rate": 8.44624981865661e-06, "loss": 0.0128, "step": 16028 }, { "epoch": 1.7377493495229834, "grad_norm": 1.79929780960083, "learning_rate": 8.445524445089222e-06, "loss": 0.034, "step": 16029 }, { "epoch": 1.7378577623590634, "grad_norm": 0.8259658813476562, "learning_rate": 8.444799071521834e-06, "loss": 0.0142, "step": 16030 }, { "epoch": 1.737966175195143, "grad_norm": 1.3635257482528687, "learning_rate": 8.444073697954448e-06, "loss": 0.0048, "step": 16031 }, { "epoch": 1.738074588031223, "grad_norm": 2.963789224624634, "learning_rate": 8.44334832438706e-06, "loss": 0.0199, "step": 16032 }, { "epoch": 1.7381830008673027, "grad_norm": 2.8849010467529297, "learning_rate": 8.442622950819674e-06, "loss": 0.0387, "step": 16033 }, { "epoch": 1.7382914137033825, "grad_norm": 2.1868104934692383, "learning_rate": 8.441897577252286e-06, "loss": 0.0251, "step": 16034 }, { "epoch": 1.7383998265394622, "grad_norm": 0.7047473192214966, "learning_rate": 8.441172203684898e-06, "loss": 0.0066, "step": 16035 }, { "epoch": 1.738508239375542, "grad_norm": 3.126549005508423, "learning_rate": 8.440446830117512e-06, "loss": 0.005, "step": 16036 }, { "epoch": 1.738616652211622, "grad_norm": 2.020291328430176, "learning_rate": 8.439721456550124e-06, "loss": 0.0155, "step": 16037 }, { "epoch": 1.7387250650477015, "grad_norm": 9.030537605285645, "learning_rate": 8.438996082982736e-06, "loss": 0.0199, "step": 16038 }, { "epoch": 1.7388334778837815, "grad_norm": 3.630167007446289, "learning_rate": 8.43827070941535e-06, "loss": 0.0146, "step": 16039 }, { "epoch": 1.7389418907198613, "grad_norm": 0.15429389476776123, "learning_rate": 8.437545335847962e-06, "loss": 0.001, "step": 16040 }, { "epoch": 1.739050303555941, "grad_norm": 0.4233706295490265, "learning_rate": 8.436819962280574e-06, "loss": 0.0135, "step": 16041 }, { "epoch": 1.7391587163920208, "grad_norm": 0.8322369456291199, "learning_rate": 8.436094588713188e-06, "loss": 0.0255, "step": 16042 }, { "epoch": 1.7392671292281006, "grad_norm": 3.5020036697387695, "learning_rate": 8.435369215145802e-06, "loss": 0.047, "step": 16043 }, { "epoch": 1.7393755420641805, "grad_norm": 2.687062978744507, "learning_rate": 8.434643841578414e-06, "loss": 0.0286, "step": 16044 }, { "epoch": 1.73948395490026, "grad_norm": 0.44138607382774353, "learning_rate": 8.433918468011026e-06, "loss": 0.0206, "step": 16045 }, { "epoch": 1.73959236773634, "grad_norm": 1.6492465734481812, "learning_rate": 8.433193094443638e-06, "loss": 0.0119, "step": 16046 }, { "epoch": 1.7397007805724196, "grad_norm": 1.8483577966690063, "learning_rate": 8.432467720876252e-06, "loss": 0.0233, "step": 16047 }, { "epoch": 1.7398091934084996, "grad_norm": 0.4797469675540924, "learning_rate": 8.431742347308864e-06, "loss": 0.0034, "step": 16048 }, { "epoch": 1.7399176062445794, "grad_norm": 2.397010087966919, "learning_rate": 8.431016973741476e-06, "loss": 0.0324, "step": 16049 }, { "epoch": 1.7400260190806591, "grad_norm": 0.28405314683914185, "learning_rate": 8.43029160017409e-06, "loss": 0.0021, "step": 16050 }, { "epoch": 1.7401344319167389, "grad_norm": 1.8611317873001099, "learning_rate": 8.429566226606704e-06, "loss": 0.0075, "step": 16051 }, { "epoch": 1.7402428447528187, "grad_norm": 0.4313695728778839, "learning_rate": 8.428840853039316e-06, "loss": 0.0132, "step": 16052 }, { "epoch": 1.7403512575888986, "grad_norm": 3.1433260440826416, "learning_rate": 8.428115479471928e-06, "loss": 0.079, "step": 16053 }, { "epoch": 1.7404596704249782, "grad_norm": 0.2703007459640503, "learning_rate": 8.427390105904542e-06, "loss": 0.0061, "step": 16054 }, { "epoch": 1.7405680832610582, "grad_norm": 7.086662292480469, "learning_rate": 8.426664732337154e-06, "loss": 0.0321, "step": 16055 }, { "epoch": 1.740676496097138, "grad_norm": 0.18690016865730286, "learning_rate": 8.425939358769767e-06, "loss": 0.0016, "step": 16056 }, { "epoch": 1.7407849089332177, "grad_norm": 3.6826488971710205, "learning_rate": 8.42521398520238e-06, "loss": 0.119, "step": 16057 }, { "epoch": 1.7408933217692975, "grad_norm": 2.2696807384490967, "learning_rate": 8.424488611634993e-06, "loss": 0.0305, "step": 16058 }, { "epoch": 1.7410017346053772, "grad_norm": 3.2656779289245605, "learning_rate": 8.423763238067606e-06, "loss": 0.0717, "step": 16059 }, { "epoch": 1.7411101474414572, "grad_norm": 1.108519196510315, "learning_rate": 8.423037864500218e-06, "loss": 0.0211, "step": 16060 }, { "epoch": 1.7412185602775367, "grad_norm": 5.033763408660889, "learning_rate": 8.42231249093283e-06, "loss": 0.0538, "step": 16061 }, { "epoch": 1.7413269731136167, "grad_norm": 10.917778015136719, "learning_rate": 8.421587117365444e-06, "loss": 0.0113, "step": 16062 }, { "epoch": 1.7414353859496965, "grad_norm": 2.0232110023498535, "learning_rate": 8.420861743798057e-06, "loss": 0.0243, "step": 16063 }, { "epoch": 1.7415437987857763, "grad_norm": 1.89762282371521, "learning_rate": 8.420136370230669e-06, "loss": 0.0262, "step": 16064 }, { "epoch": 1.741652211621856, "grad_norm": 1.4794867038726807, "learning_rate": 8.419410996663283e-06, "loss": 0.0051, "step": 16065 }, { "epoch": 1.7417606244579358, "grad_norm": 0.9173302054405212, "learning_rate": 8.418685623095895e-06, "loss": 0.0125, "step": 16066 }, { "epoch": 1.7418690372940158, "grad_norm": 1.46211576461792, "learning_rate": 8.417960249528507e-06, "loss": 0.0294, "step": 16067 }, { "epoch": 1.7419774501300953, "grad_norm": 8.868647575378418, "learning_rate": 8.41723487596112e-06, "loss": 0.0661, "step": 16068 }, { "epoch": 1.7420858629661753, "grad_norm": 6.158055305480957, "learning_rate": 8.416509502393735e-06, "loss": 0.0231, "step": 16069 }, { "epoch": 1.7421942758022548, "grad_norm": 2.6097702980041504, "learning_rate": 8.415784128826347e-06, "loss": 0.026, "step": 16070 }, { "epoch": 1.7423026886383348, "grad_norm": 3.097576856613159, "learning_rate": 8.415058755258959e-06, "loss": 0.0122, "step": 16071 }, { "epoch": 1.7424111014744146, "grad_norm": 10.173165321350098, "learning_rate": 8.414333381691571e-06, "loss": 0.0243, "step": 16072 }, { "epoch": 1.7425195143104943, "grad_norm": 0.49629199504852295, "learning_rate": 8.413608008124185e-06, "loss": 0.0057, "step": 16073 }, { "epoch": 1.742627927146574, "grad_norm": 0.26756057143211365, "learning_rate": 8.412882634556797e-06, "loss": 0.0056, "step": 16074 }, { "epoch": 1.7427363399826539, "grad_norm": 0.042358603328466415, "learning_rate": 8.412157260989409e-06, "loss": 0.0005, "step": 16075 }, { "epoch": 1.7428447528187339, "grad_norm": 1.0806455612182617, "learning_rate": 8.411431887422023e-06, "loss": 0.0129, "step": 16076 }, { "epoch": 1.7429531656548134, "grad_norm": 0.15822575986385345, "learning_rate": 8.410706513854637e-06, "loss": 0.0021, "step": 16077 }, { "epoch": 1.7430615784908934, "grad_norm": 3.0265371799468994, "learning_rate": 8.409981140287249e-06, "loss": 0.0188, "step": 16078 }, { "epoch": 1.7431699913269731, "grad_norm": 0.46663323044776917, "learning_rate": 8.409255766719861e-06, "loss": 0.0192, "step": 16079 }, { "epoch": 1.743278404163053, "grad_norm": 1.869884967803955, "learning_rate": 8.408530393152475e-06, "loss": 0.0017, "step": 16080 }, { "epoch": 1.7433868169991327, "grad_norm": 1.6784170866012573, "learning_rate": 8.407805019585087e-06, "loss": 0.023, "step": 16081 }, { "epoch": 1.7434952298352124, "grad_norm": 6.120704650878906, "learning_rate": 8.4070796460177e-06, "loss": 0.0299, "step": 16082 }, { "epoch": 1.7436036426712924, "grad_norm": 0.3423967659473419, "learning_rate": 8.406354272450313e-06, "loss": 0.0122, "step": 16083 }, { "epoch": 1.743712055507372, "grad_norm": 2.426560163497925, "learning_rate": 8.405628898882925e-06, "loss": 0.0349, "step": 16084 }, { "epoch": 1.743820468343452, "grad_norm": 1.1344244480133057, "learning_rate": 8.404903525315539e-06, "loss": 0.0108, "step": 16085 }, { "epoch": 1.7439288811795317, "grad_norm": 0.27991291880607605, "learning_rate": 8.404178151748151e-06, "loss": 0.0088, "step": 16086 }, { "epoch": 1.7440372940156115, "grad_norm": 1.0707234144210815, "learning_rate": 8.403452778180763e-06, "loss": 0.0102, "step": 16087 }, { "epoch": 1.7441457068516912, "grad_norm": 0.9458991289138794, "learning_rate": 8.402727404613377e-06, "loss": 0.0371, "step": 16088 }, { "epoch": 1.744254119687771, "grad_norm": 1.1459779739379883, "learning_rate": 8.40200203104599e-06, "loss": 0.0167, "step": 16089 }, { "epoch": 1.744362532523851, "grad_norm": 2.0795884132385254, "learning_rate": 8.401276657478601e-06, "loss": 0.0288, "step": 16090 }, { "epoch": 1.7444709453599305, "grad_norm": 1.2277189493179321, "learning_rate": 8.400551283911215e-06, "loss": 0.0295, "step": 16091 }, { "epoch": 1.7445793581960105, "grad_norm": 3.417664051055908, "learning_rate": 8.399825910343827e-06, "loss": 0.0157, "step": 16092 }, { "epoch": 1.74468777103209, "grad_norm": 1.3487495183944702, "learning_rate": 8.399100536776441e-06, "loss": 0.0131, "step": 16093 }, { "epoch": 1.74479618386817, "grad_norm": 0.7293980717658997, "learning_rate": 8.398375163209053e-06, "loss": 0.0329, "step": 16094 }, { "epoch": 1.7449045967042498, "grad_norm": 3.308755397796631, "learning_rate": 8.397649789641667e-06, "loss": 0.035, "step": 16095 }, { "epoch": 1.7450130095403296, "grad_norm": 1.5440144538879395, "learning_rate": 8.39692441607428e-06, "loss": 0.0328, "step": 16096 }, { "epoch": 1.7451214223764093, "grad_norm": 2.4230690002441406, "learning_rate": 8.396199042506892e-06, "loss": 0.0236, "step": 16097 }, { "epoch": 1.745229835212489, "grad_norm": 4.362038612365723, "learning_rate": 8.395473668939504e-06, "loss": 0.0364, "step": 16098 }, { "epoch": 1.745338248048569, "grad_norm": 9.616585731506348, "learning_rate": 8.394748295372117e-06, "loss": 0.0578, "step": 16099 }, { "epoch": 1.7454466608846486, "grad_norm": 1.0546059608459473, "learning_rate": 8.39402292180473e-06, "loss": 0.0061, "step": 16100 }, { "epoch": 1.7455550737207286, "grad_norm": 0.9121033549308777, "learning_rate": 8.393297548237342e-06, "loss": 0.003, "step": 16101 }, { "epoch": 1.7456634865568084, "grad_norm": 4.373749256134033, "learning_rate": 8.392572174669956e-06, "loss": 0.0667, "step": 16102 }, { "epoch": 1.7457718993928881, "grad_norm": 1.4074299335479736, "learning_rate": 8.39184680110257e-06, "loss": 0.0475, "step": 16103 }, { "epoch": 1.745880312228968, "grad_norm": 0.9342390298843384, "learning_rate": 8.391121427535182e-06, "loss": 0.023, "step": 16104 }, { "epoch": 1.7459887250650477, "grad_norm": 0.5238152742385864, "learning_rate": 8.390396053967794e-06, "loss": 0.0162, "step": 16105 }, { "epoch": 1.7460971379011276, "grad_norm": 0.6111493110656738, "learning_rate": 8.389670680400408e-06, "loss": 0.0018, "step": 16106 }, { "epoch": 1.7462055507372072, "grad_norm": 1.4094905853271484, "learning_rate": 8.38894530683302e-06, "loss": 0.0096, "step": 16107 }, { "epoch": 1.7463139635732872, "grad_norm": 5.333425521850586, "learning_rate": 8.388219933265632e-06, "loss": 0.0166, "step": 16108 }, { "epoch": 1.7464223764093667, "grad_norm": 2.317856550216675, "learning_rate": 8.387494559698246e-06, "loss": 0.0105, "step": 16109 }, { "epoch": 1.7465307892454467, "grad_norm": 2.050131320953369, "learning_rate": 8.386769186130858e-06, "loss": 0.0264, "step": 16110 }, { "epoch": 1.7466392020815265, "grad_norm": 0.4501124322414398, "learning_rate": 8.386043812563472e-06, "loss": 0.0108, "step": 16111 }, { "epoch": 1.7467476149176062, "grad_norm": 2.0181431770324707, "learning_rate": 8.385318438996084e-06, "loss": 0.0303, "step": 16112 }, { "epoch": 1.7468560277536862, "grad_norm": 1.5389593839645386, "learning_rate": 8.384593065428696e-06, "loss": 0.0428, "step": 16113 }, { "epoch": 1.7469644405897657, "grad_norm": 0.45716458559036255, "learning_rate": 8.38386769186131e-06, "loss": 0.0131, "step": 16114 }, { "epoch": 1.7470728534258457, "grad_norm": 1.8168696165084839, "learning_rate": 8.383142318293922e-06, "loss": 0.0175, "step": 16115 }, { "epoch": 1.7471812662619253, "grad_norm": 0.3272008001804352, "learning_rate": 8.382416944726534e-06, "loss": 0.0112, "step": 16116 }, { "epoch": 1.7472896790980053, "grad_norm": 1.379759430885315, "learning_rate": 8.381691571159148e-06, "loss": 0.0122, "step": 16117 }, { "epoch": 1.747398091934085, "grad_norm": 0.7556582093238831, "learning_rate": 8.38096619759176e-06, "loss": 0.0224, "step": 16118 }, { "epoch": 1.7475065047701648, "grad_norm": 0.8184324502944946, "learning_rate": 8.380240824024374e-06, "loss": 0.0196, "step": 16119 }, { "epoch": 1.7476149176062445, "grad_norm": 5.778987407684326, "learning_rate": 8.379515450456986e-06, "loss": 0.007, "step": 16120 }, { "epoch": 1.7477233304423243, "grad_norm": 0.46596017479896545, "learning_rate": 8.3787900768896e-06, "loss": 0.0084, "step": 16121 }, { "epoch": 1.7478317432784043, "grad_norm": 0.7685713768005371, "learning_rate": 8.378064703322212e-06, "loss": 0.0122, "step": 16122 }, { "epoch": 1.7479401561144838, "grad_norm": 0.1025814637541771, "learning_rate": 8.377339329754824e-06, "loss": 0.001, "step": 16123 }, { "epoch": 1.7480485689505638, "grad_norm": 0.5399781465530396, "learning_rate": 8.376613956187436e-06, "loss": 0.018, "step": 16124 }, { "epoch": 1.7481569817866436, "grad_norm": 2.929180860519409, "learning_rate": 8.37588858262005e-06, "loss": 0.0401, "step": 16125 }, { "epoch": 1.7482653946227233, "grad_norm": 1.7359005212783813, "learning_rate": 8.375163209052662e-06, "loss": 0.0433, "step": 16126 }, { "epoch": 1.748373807458803, "grad_norm": 0.22379744052886963, "learning_rate": 8.374437835485274e-06, "loss": 0.0017, "step": 16127 }, { "epoch": 1.7484822202948829, "grad_norm": 1.3913062810897827, "learning_rate": 8.373712461917888e-06, "loss": 0.0163, "step": 16128 }, { "epoch": 1.7485906331309629, "grad_norm": 2.053203821182251, "learning_rate": 8.372987088350502e-06, "loss": 0.0482, "step": 16129 }, { "epoch": 1.7486990459670424, "grad_norm": 0.3763275742530823, "learning_rate": 8.372261714783114e-06, "loss": 0.0065, "step": 16130 }, { "epoch": 1.7488074588031224, "grad_norm": 1.6566190719604492, "learning_rate": 8.371536341215726e-06, "loss": 0.0037, "step": 16131 }, { "epoch": 1.748915871639202, "grad_norm": 2.588257074356079, "learning_rate": 8.37081096764834e-06, "loss": 0.0482, "step": 16132 }, { "epoch": 1.749024284475282, "grad_norm": 2.020634412765503, "learning_rate": 8.370085594080952e-06, "loss": 0.0205, "step": 16133 }, { "epoch": 1.7491326973113617, "grad_norm": 0.9138292670249939, "learning_rate": 8.369360220513565e-06, "loss": 0.0033, "step": 16134 }, { "epoch": 1.7492411101474414, "grad_norm": 2.3832356929779053, "learning_rate": 8.368634846946178e-06, "loss": 0.0174, "step": 16135 }, { "epoch": 1.7493495229835212, "grad_norm": 5.0068359375, "learning_rate": 8.36790947337879e-06, "loss": 0.006, "step": 16136 }, { "epoch": 1.749457935819601, "grad_norm": 1.8812718391418457, "learning_rate": 8.367184099811404e-06, "loss": 0.0668, "step": 16137 }, { "epoch": 1.749566348655681, "grad_norm": 1.1102526187896729, "learning_rate": 8.366458726244016e-06, "loss": 0.0136, "step": 16138 }, { "epoch": 1.7496747614917605, "grad_norm": 0.5145768523216248, "learning_rate": 8.365733352676629e-06, "loss": 0.0075, "step": 16139 }, { "epoch": 1.7497831743278405, "grad_norm": 0.17459021508693695, "learning_rate": 8.365007979109242e-06, "loss": 0.0016, "step": 16140 }, { "epoch": 1.7498915871639202, "grad_norm": 3.1902236938476562, "learning_rate": 8.364282605541855e-06, "loss": 0.0259, "step": 16141 }, { "epoch": 1.75, "grad_norm": 1.912839651107788, "learning_rate": 8.363557231974467e-06, "loss": 0.0248, "step": 16142 }, { "epoch": 1.7501084128360798, "grad_norm": 5.479602336883545, "learning_rate": 8.36283185840708e-06, "loss": 0.0316, "step": 16143 }, { "epoch": 1.7502168256721595, "grad_norm": 2.9909403324127197, "learning_rate": 8.362106484839693e-06, "loss": 0.0394, "step": 16144 }, { "epoch": 1.7503252385082395, "grad_norm": 3.870386838912964, "learning_rate": 8.361381111272307e-06, "loss": 0.0622, "step": 16145 }, { "epoch": 1.750433651344319, "grad_norm": 2.474350929260254, "learning_rate": 8.360655737704919e-06, "loss": 0.0068, "step": 16146 }, { "epoch": 1.750542064180399, "grad_norm": 4.315378189086914, "learning_rate": 8.359930364137533e-06, "loss": 0.0305, "step": 16147 }, { "epoch": 1.7506504770164788, "grad_norm": 0.4241470396518707, "learning_rate": 8.359204990570145e-06, "loss": 0.0037, "step": 16148 }, { "epoch": 1.7507588898525586, "grad_norm": 5.215141296386719, "learning_rate": 8.358479617002757e-06, "loss": 0.0175, "step": 16149 }, { "epoch": 1.7508673026886383, "grad_norm": 6.456576824188232, "learning_rate": 8.357754243435369e-06, "loss": 0.0171, "step": 16150 }, { "epoch": 1.750975715524718, "grad_norm": 0.43607577681541443, "learning_rate": 8.357028869867983e-06, "loss": 0.0136, "step": 16151 }, { "epoch": 1.751084128360798, "grad_norm": 1.0026347637176514, "learning_rate": 8.356303496300595e-06, "loss": 0.0566, "step": 16152 }, { "epoch": 1.7511925411968776, "grad_norm": 0.4534107744693756, "learning_rate": 8.355578122733207e-06, "loss": 0.0249, "step": 16153 }, { "epoch": 1.7513009540329576, "grad_norm": 7.185702800750732, "learning_rate": 8.354852749165821e-06, "loss": 0.0127, "step": 16154 }, { "epoch": 1.7514093668690371, "grad_norm": 0.40740644931793213, "learning_rate": 8.354127375598435e-06, "loss": 0.0194, "step": 16155 }, { "epoch": 1.7515177797051171, "grad_norm": 0.41948461532592773, "learning_rate": 8.353402002031047e-06, "loss": 0.0155, "step": 16156 }, { "epoch": 1.751626192541197, "grad_norm": 1.4913816452026367, "learning_rate": 8.352676628463659e-06, "loss": 0.0348, "step": 16157 }, { "epoch": 1.7517346053772767, "grad_norm": 0.014361578971147537, "learning_rate": 8.351951254896273e-06, "loss": 0.0002, "step": 16158 }, { "epoch": 1.7518430182133564, "grad_norm": 5.032920837402344, "learning_rate": 8.351225881328885e-06, "loss": 0.038, "step": 16159 }, { "epoch": 1.7519514310494362, "grad_norm": 0.392117440700531, "learning_rate": 8.350500507761497e-06, "loss": 0.0091, "step": 16160 }, { "epoch": 1.7520598438855162, "grad_norm": 3.3967196941375732, "learning_rate": 8.349775134194111e-06, "loss": 0.0541, "step": 16161 }, { "epoch": 1.7521682567215957, "grad_norm": 0.26935768127441406, "learning_rate": 8.349049760626723e-06, "loss": 0.0055, "step": 16162 }, { "epoch": 1.7522766695576757, "grad_norm": 0.5310776829719543, "learning_rate": 8.348324387059337e-06, "loss": 0.0135, "step": 16163 }, { "epoch": 1.7523850823937555, "grad_norm": 3.4140288829803467, "learning_rate": 8.34759901349195e-06, "loss": 0.0201, "step": 16164 }, { "epoch": 1.7524934952298352, "grad_norm": 0.10637146979570389, "learning_rate": 8.346873639924561e-06, "loss": 0.001, "step": 16165 }, { "epoch": 1.752601908065915, "grad_norm": 0.4198998510837555, "learning_rate": 8.346148266357175e-06, "loss": 0.0069, "step": 16166 }, { "epoch": 1.7527103209019947, "grad_norm": 5.131176948547363, "learning_rate": 8.345422892789787e-06, "loss": 0.0602, "step": 16167 }, { "epoch": 1.7528187337380747, "grad_norm": 1.2159602642059326, "learning_rate": 8.3446975192224e-06, "loss": 0.0156, "step": 16168 }, { "epoch": 1.7529271465741543, "grad_norm": 5.253089904785156, "learning_rate": 8.343972145655013e-06, "loss": 0.0394, "step": 16169 }, { "epoch": 1.7530355594102343, "grad_norm": 6.135360240936279, "learning_rate": 8.343246772087625e-06, "loss": 0.0703, "step": 16170 }, { "epoch": 1.7531439722463138, "grad_norm": 0.5186441540718079, "learning_rate": 8.34252139852024e-06, "loss": 0.0021, "step": 16171 }, { "epoch": 1.7532523850823938, "grad_norm": 0.1611536741256714, "learning_rate": 8.341796024952851e-06, "loss": 0.0014, "step": 16172 }, { "epoch": 1.7533607979184735, "grad_norm": 2.1719110012054443, "learning_rate": 8.341070651385465e-06, "loss": 0.0167, "step": 16173 }, { "epoch": 1.7534692107545533, "grad_norm": 0.748630702495575, "learning_rate": 8.340345277818077e-06, "loss": 0.0021, "step": 16174 }, { "epoch": 1.7535776235906333, "grad_norm": 0.6601061820983887, "learning_rate": 8.33961990425069e-06, "loss": 0.0125, "step": 16175 }, { "epoch": 1.7536860364267128, "grad_norm": 0.7170270681381226, "learning_rate": 8.338894530683302e-06, "loss": 0.0289, "step": 16176 }, { "epoch": 1.7537944492627928, "grad_norm": 1.632405161857605, "learning_rate": 8.338169157115915e-06, "loss": 0.0265, "step": 16177 }, { "epoch": 1.7539028620988724, "grad_norm": 0.24758239090442657, "learning_rate": 8.337443783548528e-06, "loss": 0.0014, "step": 16178 }, { "epoch": 1.7540112749349523, "grad_norm": 0.30927786231040955, "learning_rate": 8.33671840998114e-06, "loss": 0.0056, "step": 16179 }, { "epoch": 1.754119687771032, "grad_norm": 1.4047244787216187, "learning_rate": 8.335993036413754e-06, "loss": 0.0255, "step": 16180 }, { "epoch": 1.7542281006071119, "grad_norm": 0.658437192440033, "learning_rate": 8.335267662846367e-06, "loss": 0.0037, "step": 16181 }, { "epoch": 1.7543365134431916, "grad_norm": 1.358826756477356, "learning_rate": 8.33454228927898e-06, "loss": 0.0075, "step": 16182 }, { "epoch": 1.7544449262792714, "grad_norm": 1.5834364891052246, "learning_rate": 8.333816915711592e-06, "loss": 0.0043, "step": 16183 }, { "epoch": 1.7545533391153514, "grad_norm": 0.7203183174133301, "learning_rate": 8.333091542144206e-06, "loss": 0.0189, "step": 16184 }, { "epoch": 1.754661751951431, "grad_norm": 0.18040050566196442, "learning_rate": 8.332366168576818e-06, "loss": 0.0022, "step": 16185 }, { "epoch": 1.754770164787511, "grad_norm": 4.584592819213867, "learning_rate": 8.33164079500943e-06, "loss": 0.0135, "step": 16186 }, { "epoch": 1.7548785776235907, "grad_norm": 0.1388993263244629, "learning_rate": 8.330915421442044e-06, "loss": 0.0015, "step": 16187 }, { "epoch": 1.7549869904596704, "grad_norm": 0.3407028317451477, "learning_rate": 8.330190047874656e-06, "loss": 0.0062, "step": 16188 }, { "epoch": 1.7550954032957502, "grad_norm": 1.3415781259536743, "learning_rate": 8.32946467430727e-06, "loss": 0.0242, "step": 16189 }, { "epoch": 1.75520381613183, "grad_norm": 5.5664963722229, "learning_rate": 8.328739300739882e-06, "loss": 0.0155, "step": 16190 }, { "epoch": 1.75531222896791, "grad_norm": 3.577676296234131, "learning_rate": 8.328013927172494e-06, "loss": 0.0247, "step": 16191 }, { "epoch": 1.7554206418039895, "grad_norm": 3.20348858833313, "learning_rate": 8.327288553605108e-06, "loss": 0.0272, "step": 16192 }, { "epoch": 1.7555290546400695, "grad_norm": 2.271178722381592, "learning_rate": 8.32656318003772e-06, "loss": 0.0298, "step": 16193 }, { "epoch": 1.755637467476149, "grad_norm": 0.34406983852386475, "learning_rate": 8.325837806470332e-06, "loss": 0.0063, "step": 16194 }, { "epoch": 1.755745880312229, "grad_norm": 4.927204608917236, "learning_rate": 8.325112432902946e-06, "loss": 0.0278, "step": 16195 }, { "epoch": 1.7558542931483088, "grad_norm": 0.9666191935539246, "learning_rate": 8.324387059335558e-06, "loss": 0.0129, "step": 16196 }, { "epoch": 1.7559627059843885, "grad_norm": 0.7533310055732727, "learning_rate": 8.323661685768172e-06, "loss": 0.0256, "step": 16197 }, { "epoch": 1.7560711188204683, "grad_norm": 3.249732255935669, "learning_rate": 8.322936312200784e-06, "loss": 0.0273, "step": 16198 }, { "epoch": 1.756179531656548, "grad_norm": 2.9650909900665283, "learning_rate": 8.322210938633398e-06, "loss": 0.0471, "step": 16199 }, { "epoch": 1.756287944492628, "grad_norm": 2.6898114681243896, "learning_rate": 8.32148556506601e-06, "loss": 0.0357, "step": 16200 }, { "epoch": 1.7563963573287076, "grad_norm": 1.2827448844909668, "learning_rate": 8.320760191498622e-06, "loss": 0.0149, "step": 16201 }, { "epoch": 1.7565047701647876, "grad_norm": 0.27481961250305176, "learning_rate": 8.320034817931234e-06, "loss": 0.0064, "step": 16202 }, { "epoch": 1.7566131830008673, "grad_norm": 1.1205196380615234, "learning_rate": 8.319309444363848e-06, "loss": 0.0265, "step": 16203 }, { "epoch": 1.756721595836947, "grad_norm": 1.95314621925354, "learning_rate": 8.31858407079646e-06, "loss": 0.0214, "step": 16204 }, { "epoch": 1.7568300086730269, "grad_norm": 2.8069372177124023, "learning_rate": 8.317858697229072e-06, "loss": 0.0589, "step": 16205 }, { "epoch": 1.7569384215091066, "grad_norm": 1.3836885690689087, "learning_rate": 8.317133323661686e-06, "loss": 0.0179, "step": 16206 }, { "epoch": 1.7570468343451866, "grad_norm": 5.684413433074951, "learning_rate": 8.3164079500943e-06, "loss": 0.0562, "step": 16207 }, { "epoch": 1.7571552471812661, "grad_norm": 3.8828237056732178, "learning_rate": 8.315682576526912e-06, "loss": 0.058, "step": 16208 }, { "epoch": 1.7572636600173461, "grad_norm": 1.4504832029342651, "learning_rate": 8.314957202959524e-06, "loss": 0.0046, "step": 16209 }, { "epoch": 1.757372072853426, "grad_norm": 2.211888074874878, "learning_rate": 8.314231829392138e-06, "loss": 0.0044, "step": 16210 }, { "epoch": 1.7574804856895057, "grad_norm": 0.5565624833106995, "learning_rate": 8.31350645582475e-06, "loss": 0.0085, "step": 16211 }, { "epoch": 1.7575888985255854, "grad_norm": 0.7660475373268127, "learning_rate": 8.312781082257363e-06, "loss": 0.004, "step": 16212 }, { "epoch": 1.7576973113616652, "grad_norm": 2.272672176361084, "learning_rate": 8.312055708689976e-06, "loss": 0.0092, "step": 16213 }, { "epoch": 1.7578057241977452, "grad_norm": 0.12301980704069138, "learning_rate": 8.311330335122589e-06, "loss": 0.001, "step": 16214 }, { "epoch": 1.7579141370338247, "grad_norm": 0.386548787355423, "learning_rate": 8.310604961555202e-06, "loss": 0.0093, "step": 16215 }, { "epoch": 1.7580225498699047, "grad_norm": 4.530596733093262, "learning_rate": 8.309879587987814e-06, "loss": 0.0791, "step": 16216 }, { "epoch": 1.7581309627059842, "grad_norm": 4.7955217361450195, "learning_rate": 8.309154214420427e-06, "loss": 0.0603, "step": 16217 }, { "epoch": 1.7582393755420642, "grad_norm": 5.029377460479736, "learning_rate": 8.30842884085304e-06, "loss": 0.0182, "step": 16218 }, { "epoch": 1.758347788378144, "grad_norm": 5.248482704162598, "learning_rate": 8.307703467285653e-06, "loss": 0.055, "step": 16219 }, { "epoch": 1.7584562012142237, "grad_norm": 7.5606231689453125, "learning_rate": 8.306978093718265e-06, "loss": 0.0364, "step": 16220 }, { "epoch": 1.7585646140503035, "grad_norm": 1.5969746112823486, "learning_rate": 8.306252720150879e-06, "loss": 0.0244, "step": 16221 }, { "epoch": 1.7586730268863833, "grad_norm": 0.49193644523620605, "learning_rate": 8.30552734658349e-06, "loss": 0.0176, "step": 16222 }, { "epoch": 1.7587814397224633, "grad_norm": 4.769722938537598, "learning_rate": 8.304801973016105e-06, "loss": 0.0825, "step": 16223 }, { "epoch": 1.7588898525585428, "grad_norm": 3.000967264175415, "learning_rate": 8.304076599448717e-06, "loss": 0.0321, "step": 16224 }, { "epoch": 1.7589982653946228, "grad_norm": 0.28987544775009155, "learning_rate": 8.30335122588133e-06, "loss": 0.0026, "step": 16225 }, { "epoch": 1.7591066782307025, "grad_norm": 3.139740467071533, "learning_rate": 8.302625852313943e-06, "loss": 0.0462, "step": 16226 }, { "epoch": 1.7592150910667823, "grad_norm": 0.635214626789093, "learning_rate": 8.301900478746555e-06, "loss": 0.0217, "step": 16227 }, { "epoch": 1.759323503902862, "grad_norm": 0.39955276250839233, "learning_rate": 8.301175105179167e-06, "loss": 0.003, "step": 16228 }, { "epoch": 1.7594319167389418, "grad_norm": 1.9003390073776245, "learning_rate": 8.30044973161178e-06, "loss": 0.033, "step": 16229 }, { "epoch": 1.7595403295750218, "grad_norm": 1.7077996730804443, "learning_rate": 8.299724358044393e-06, "loss": 0.0068, "step": 16230 }, { "epoch": 1.7596487424111014, "grad_norm": 27.279022216796875, "learning_rate": 8.298998984477005e-06, "loss": 0.0277, "step": 16231 }, { "epoch": 1.7597571552471813, "grad_norm": 0.8764214515686035, "learning_rate": 8.298273610909619e-06, "loss": 0.0137, "step": 16232 }, { "epoch": 1.7598655680832611, "grad_norm": 0.890884280204773, "learning_rate": 8.297548237342233e-06, "loss": 0.0167, "step": 16233 }, { "epoch": 1.7599739809193409, "grad_norm": 2.8155930042266846, "learning_rate": 8.296822863774845e-06, "loss": 0.0569, "step": 16234 }, { "epoch": 1.7600823937554206, "grad_norm": 3.2189438343048096, "learning_rate": 8.296097490207457e-06, "loss": 0.0648, "step": 16235 }, { "epoch": 1.7601908065915004, "grad_norm": 0.7943809032440186, "learning_rate": 8.295372116640071e-06, "loss": 0.0135, "step": 16236 }, { "epoch": 1.7602992194275804, "grad_norm": 1.7815673351287842, "learning_rate": 8.294646743072683e-06, "loss": 0.0336, "step": 16237 }, { "epoch": 1.76040763226366, "grad_norm": 2.6773483753204346, "learning_rate": 8.293921369505295e-06, "loss": 0.0157, "step": 16238 }, { "epoch": 1.76051604509974, "grad_norm": 1.306476354598999, "learning_rate": 8.293195995937909e-06, "loss": 0.0266, "step": 16239 }, { "epoch": 1.7606244579358195, "grad_norm": 7.30010986328125, "learning_rate": 8.292470622370521e-06, "loss": 0.0348, "step": 16240 }, { "epoch": 1.7607328707718994, "grad_norm": 0.4432743787765503, "learning_rate": 8.291745248803135e-06, "loss": 0.0164, "step": 16241 }, { "epoch": 1.7608412836079792, "grad_norm": 0.23746086657047272, "learning_rate": 8.291019875235747e-06, "loss": 0.0053, "step": 16242 }, { "epoch": 1.760949696444059, "grad_norm": 0.9588786363601685, "learning_rate": 8.29029450166836e-06, "loss": 0.0093, "step": 16243 }, { "epoch": 1.7610581092801387, "grad_norm": 0.37871772050857544, "learning_rate": 8.289569128100973e-06, "loss": 0.0032, "step": 16244 }, { "epoch": 1.7611665221162185, "grad_norm": 0.15976403653621674, "learning_rate": 8.288843754533585e-06, "loss": 0.0013, "step": 16245 }, { "epoch": 1.7612749349522985, "grad_norm": 0.7383745908737183, "learning_rate": 8.288118380966197e-06, "loss": 0.0016, "step": 16246 }, { "epoch": 1.761383347788378, "grad_norm": 4.383888244628906, "learning_rate": 8.287393007398811e-06, "loss": 0.0651, "step": 16247 }, { "epoch": 1.761491760624458, "grad_norm": 0.5773365497589111, "learning_rate": 8.286667633831423e-06, "loss": 0.0191, "step": 16248 }, { "epoch": 1.7616001734605378, "grad_norm": 2.599294662475586, "learning_rate": 8.285942260264037e-06, "loss": 0.0185, "step": 16249 }, { "epoch": 1.7617085862966175, "grad_norm": 6.41776704788208, "learning_rate": 8.28521688669665e-06, "loss": 0.0275, "step": 16250 }, { "epoch": 1.7618169991326973, "grad_norm": 1.3596431016921997, "learning_rate": 8.284491513129263e-06, "loss": 0.0145, "step": 16251 }, { "epoch": 1.761925411968777, "grad_norm": 2.2071878910064697, "learning_rate": 8.283766139561875e-06, "loss": 0.0371, "step": 16252 }, { "epoch": 1.762033824804857, "grad_norm": 0.5750311017036438, "learning_rate": 8.283040765994488e-06, "loss": 0.0125, "step": 16253 }, { "epoch": 1.7621422376409366, "grad_norm": 6.363739013671875, "learning_rate": 8.2823153924271e-06, "loss": 0.0172, "step": 16254 }, { "epoch": 1.7622506504770166, "grad_norm": 2.303394317626953, "learning_rate": 8.281590018859713e-06, "loss": 0.0312, "step": 16255 }, { "epoch": 1.762359063313096, "grad_norm": 3.18160080909729, "learning_rate": 8.280864645292326e-06, "loss": 0.0647, "step": 16256 }, { "epoch": 1.762467476149176, "grad_norm": 0.1630815863609314, "learning_rate": 8.280139271724938e-06, "loss": 0.0018, "step": 16257 }, { "epoch": 1.7625758889852559, "grad_norm": 0.045034538954496384, "learning_rate": 8.279413898157552e-06, "loss": 0.0004, "step": 16258 }, { "epoch": 1.7626843018213356, "grad_norm": 0.20106597244739532, "learning_rate": 8.278688524590165e-06, "loss": 0.0018, "step": 16259 }, { "epoch": 1.7627927146574154, "grad_norm": 0.8356142640113831, "learning_rate": 8.277963151022778e-06, "loss": 0.024, "step": 16260 }, { "epoch": 1.7629011274934951, "grad_norm": 2.6242990493774414, "learning_rate": 8.27723777745539e-06, "loss": 0.0634, "step": 16261 }, { "epoch": 1.7630095403295751, "grad_norm": 2.195889949798584, "learning_rate": 8.276512403888004e-06, "loss": 0.0227, "step": 16262 }, { "epoch": 1.7631179531656547, "grad_norm": 2.5434844493865967, "learning_rate": 8.275787030320616e-06, "loss": 0.0833, "step": 16263 }, { "epoch": 1.7632263660017347, "grad_norm": 0.4325208067893982, "learning_rate": 8.275061656753228e-06, "loss": 0.0157, "step": 16264 }, { "epoch": 1.7633347788378144, "grad_norm": 0.46301454305648804, "learning_rate": 8.274336283185842e-06, "loss": 0.0103, "step": 16265 }, { "epoch": 1.7634431916738942, "grad_norm": 0.8132782578468323, "learning_rate": 8.273610909618456e-06, "loss": 0.0056, "step": 16266 }, { "epoch": 1.763551604509974, "grad_norm": 2.49587082862854, "learning_rate": 8.272885536051068e-06, "loss": 0.0213, "step": 16267 }, { "epoch": 1.7636600173460537, "grad_norm": 0.7358472943305969, "learning_rate": 8.27216016248368e-06, "loss": 0.016, "step": 16268 }, { "epoch": 1.7637684301821337, "grad_norm": 0.35992130637168884, "learning_rate": 8.271434788916292e-06, "loss": 0.0078, "step": 16269 }, { "epoch": 1.7638768430182132, "grad_norm": 2.624350070953369, "learning_rate": 8.270709415348906e-06, "loss": 0.0233, "step": 16270 }, { "epoch": 1.7639852558542932, "grad_norm": 0.34301403164863586, "learning_rate": 8.269984041781518e-06, "loss": 0.0105, "step": 16271 }, { "epoch": 1.764093668690373, "grad_norm": 3.998561382293701, "learning_rate": 8.26925866821413e-06, "loss": 0.0721, "step": 16272 }, { "epoch": 1.7642020815264527, "grad_norm": 1.2580652236938477, "learning_rate": 8.268533294646744e-06, "loss": 0.0142, "step": 16273 }, { "epoch": 1.7643104943625325, "grad_norm": 1.9482192993164062, "learning_rate": 8.267807921079356e-06, "loss": 0.0317, "step": 16274 }, { "epoch": 1.7644189071986123, "grad_norm": 3.7797372341156006, "learning_rate": 8.26708254751197e-06, "loss": 0.0328, "step": 16275 }, { "epoch": 1.7645273200346923, "grad_norm": 7.4522318840026855, "learning_rate": 8.266357173944582e-06, "loss": 0.0316, "step": 16276 }, { "epoch": 1.7646357328707718, "grad_norm": 0.8590108156204224, "learning_rate": 8.265631800377196e-06, "loss": 0.0083, "step": 16277 }, { "epoch": 1.7647441457068518, "grad_norm": 0.5717258453369141, "learning_rate": 8.264906426809808e-06, "loss": 0.0141, "step": 16278 }, { "epoch": 1.7648525585429313, "grad_norm": 1.523486614227295, "learning_rate": 8.26418105324242e-06, "loss": 0.0398, "step": 16279 }, { "epoch": 1.7649609713790113, "grad_norm": 3.0332083702087402, "learning_rate": 8.263455679675032e-06, "loss": 0.0292, "step": 16280 }, { "epoch": 1.765069384215091, "grad_norm": 0.1172342523932457, "learning_rate": 8.262730306107646e-06, "loss": 0.0014, "step": 16281 }, { "epoch": 1.7651777970511708, "grad_norm": 4.11051082611084, "learning_rate": 8.262004932540258e-06, "loss": 0.023, "step": 16282 }, { "epoch": 1.7652862098872506, "grad_norm": 0.585260808467865, "learning_rate": 8.26127955897287e-06, "loss": 0.0145, "step": 16283 }, { "epoch": 1.7653946227233304, "grad_norm": 0.5123074054718018, "learning_rate": 8.260554185405484e-06, "loss": 0.0102, "step": 16284 }, { "epoch": 1.7655030355594103, "grad_norm": 7.3430633544921875, "learning_rate": 8.259828811838098e-06, "loss": 0.0174, "step": 16285 }, { "epoch": 1.76561144839549, "grad_norm": 0.13532654941082, "learning_rate": 8.25910343827071e-06, "loss": 0.0012, "step": 16286 }, { "epoch": 1.7657198612315699, "grad_norm": 4.052873611450195, "learning_rate": 8.258378064703322e-06, "loss": 0.0434, "step": 16287 }, { "epoch": 1.7658282740676496, "grad_norm": 1.354218602180481, "learning_rate": 8.257652691135936e-06, "loss": 0.0131, "step": 16288 }, { "epoch": 1.7659366869037294, "grad_norm": 0.7424793839454651, "learning_rate": 8.256927317568548e-06, "loss": 0.0122, "step": 16289 }, { "epoch": 1.7660450997398092, "grad_norm": 2.9443936347961426, "learning_rate": 8.25620194400116e-06, "loss": 0.033, "step": 16290 }, { "epoch": 1.766153512575889, "grad_norm": 4.914475440979004, "learning_rate": 8.255476570433774e-06, "loss": 0.0375, "step": 16291 }, { "epoch": 1.766261925411969, "grad_norm": 0.5710111260414124, "learning_rate": 8.254751196866388e-06, "loss": 0.0105, "step": 16292 }, { "epoch": 1.7663703382480485, "grad_norm": 4.262892723083496, "learning_rate": 8.254025823299e-06, "loss": 0.0524, "step": 16293 }, { "epoch": 1.7664787510841284, "grad_norm": 0.6534938216209412, "learning_rate": 8.253300449731613e-06, "loss": 0.0472, "step": 16294 }, { "epoch": 1.7665871639202082, "grad_norm": 2.3295257091522217, "learning_rate": 8.252575076164225e-06, "loss": 0.0391, "step": 16295 }, { "epoch": 1.766695576756288, "grad_norm": 3.1549925804138184, "learning_rate": 8.251849702596838e-06, "loss": 0.0339, "step": 16296 }, { "epoch": 1.7668039895923677, "grad_norm": 2.861804962158203, "learning_rate": 8.25112432902945e-06, "loss": 0.037, "step": 16297 }, { "epoch": 1.7669124024284475, "grad_norm": 2.0767948627471924, "learning_rate": 8.250398955462063e-06, "loss": 0.0244, "step": 16298 }, { "epoch": 1.7670208152645275, "grad_norm": 0.6375948190689087, "learning_rate": 8.249673581894677e-06, "loss": 0.0139, "step": 16299 }, { "epoch": 1.767129228100607, "grad_norm": 5.119906902313232, "learning_rate": 8.248948208327289e-06, "loss": 0.0384, "step": 16300 }, { "epoch": 1.767237640936687, "grad_norm": 6.837708473205566, "learning_rate": 8.248222834759903e-06, "loss": 0.0346, "step": 16301 }, { "epoch": 1.7673460537727665, "grad_norm": 1.634505033493042, "learning_rate": 8.247497461192515e-06, "loss": 0.0316, "step": 16302 }, { "epoch": 1.7674544666088465, "grad_norm": 5.423619270324707, "learning_rate": 8.246772087625129e-06, "loss": 0.029, "step": 16303 }, { "epoch": 1.7675628794449263, "grad_norm": 5.121882915496826, "learning_rate": 8.24604671405774e-06, "loss": 0.042, "step": 16304 }, { "epoch": 1.767671292281006, "grad_norm": 0.957722544670105, "learning_rate": 8.245321340490353e-06, "loss": 0.0238, "step": 16305 }, { "epoch": 1.7677797051170858, "grad_norm": 1.1368571519851685, "learning_rate": 8.244595966922965e-06, "loss": 0.039, "step": 16306 }, { "epoch": 1.7678881179531656, "grad_norm": 3.267482280731201, "learning_rate": 8.243870593355579e-06, "loss": 0.0146, "step": 16307 }, { "epoch": 1.7679965307892456, "grad_norm": 1.757615089416504, "learning_rate": 8.243145219788191e-06, "loss": 0.0226, "step": 16308 }, { "epoch": 1.768104943625325, "grad_norm": 1.256695032119751, "learning_rate": 8.242419846220805e-06, "loss": 0.0084, "step": 16309 }, { "epoch": 1.768213356461405, "grad_norm": 2.4125373363494873, "learning_rate": 8.241694472653417e-06, "loss": 0.0363, "step": 16310 }, { "epoch": 1.7683217692974849, "grad_norm": 1.050856113433838, "learning_rate": 8.24096909908603e-06, "loss": 0.0254, "step": 16311 }, { "epoch": 1.7684301821335646, "grad_norm": 0.8093042373657227, "learning_rate": 8.240243725518643e-06, "loss": 0.0061, "step": 16312 }, { "epoch": 1.7685385949696444, "grad_norm": 0.5447174310684204, "learning_rate": 8.239518351951255e-06, "loss": 0.0048, "step": 16313 }, { "epoch": 1.7686470078057241, "grad_norm": 8.42240047454834, "learning_rate": 8.238792978383869e-06, "loss": 0.0332, "step": 16314 }, { "epoch": 1.7687554206418041, "grad_norm": 0.36344897747039795, "learning_rate": 8.238067604816481e-06, "loss": 0.0043, "step": 16315 }, { "epoch": 1.7688638334778837, "grad_norm": 0.4928210377693176, "learning_rate": 8.237342231249093e-06, "loss": 0.0048, "step": 16316 }, { "epoch": 1.7689722463139637, "grad_norm": 0.4208053946495056, "learning_rate": 8.236616857681707e-06, "loss": 0.0017, "step": 16317 }, { "epoch": 1.7690806591500432, "grad_norm": 2.56307315826416, "learning_rate": 8.235891484114321e-06, "loss": 0.0067, "step": 16318 }, { "epoch": 1.7691890719861232, "grad_norm": 1.7096856832504272, "learning_rate": 8.235166110546933e-06, "loss": 0.0177, "step": 16319 }, { "epoch": 1.769297484822203, "grad_norm": 1.0723124742507935, "learning_rate": 8.234440736979545e-06, "loss": 0.005, "step": 16320 }, { "epoch": 1.7694058976582827, "grad_norm": 1.4777640104293823, "learning_rate": 8.233715363412157e-06, "loss": 0.0063, "step": 16321 }, { "epoch": 1.7695143104943627, "grad_norm": 1.5652912855148315, "learning_rate": 8.232989989844771e-06, "loss": 0.039, "step": 16322 }, { "epoch": 1.7696227233304422, "grad_norm": 2.6882553100585938, "learning_rate": 8.232264616277383e-06, "loss": 0.0192, "step": 16323 }, { "epoch": 1.7697311361665222, "grad_norm": 0.5531238913536072, "learning_rate": 8.231539242709995e-06, "loss": 0.0031, "step": 16324 }, { "epoch": 1.7698395490026018, "grad_norm": 1.5559159517288208, "learning_rate": 8.23081386914261e-06, "loss": 0.033, "step": 16325 }, { "epoch": 1.7699479618386817, "grad_norm": 2.398380756378174, "learning_rate": 8.230088495575221e-06, "loss": 0.0491, "step": 16326 }, { "epoch": 1.7700563746747615, "grad_norm": 0.7129753232002258, "learning_rate": 8.229363122007835e-06, "loss": 0.0204, "step": 16327 }, { "epoch": 1.7701647875108413, "grad_norm": 3.4664604663848877, "learning_rate": 8.228637748440447e-06, "loss": 0.0148, "step": 16328 }, { "epoch": 1.770273200346921, "grad_norm": 3.1556928157806396, "learning_rate": 8.227912374873061e-06, "loss": 0.039, "step": 16329 }, { "epoch": 1.7703816131830008, "grad_norm": 2.621640682220459, "learning_rate": 8.227187001305673e-06, "loss": 0.0128, "step": 16330 }, { "epoch": 1.7704900260190808, "grad_norm": 6.926424503326416, "learning_rate": 8.226461627738286e-06, "loss": 0.0286, "step": 16331 }, { "epoch": 1.7705984388551603, "grad_norm": 2.4305436611175537, "learning_rate": 8.225736254170898e-06, "loss": 0.0206, "step": 16332 }, { "epoch": 1.7707068516912403, "grad_norm": 3.632007360458374, "learning_rate": 8.225010880603512e-06, "loss": 0.0345, "step": 16333 }, { "epoch": 1.77081526452732, "grad_norm": 4.041248321533203, "learning_rate": 8.224285507036124e-06, "loss": 0.0441, "step": 16334 }, { "epoch": 1.7709236773633998, "grad_norm": 3.698286294937134, "learning_rate": 8.223560133468737e-06, "loss": 0.0371, "step": 16335 }, { "epoch": 1.7710320901994796, "grad_norm": 0.2737027108669281, "learning_rate": 8.22283475990135e-06, "loss": 0.0203, "step": 16336 }, { "epoch": 1.7711405030355594, "grad_norm": 2.645050048828125, "learning_rate": 8.222109386333963e-06, "loss": 0.015, "step": 16337 }, { "epoch": 1.7712489158716394, "grad_norm": 0.28035768866539, "learning_rate": 8.221384012766576e-06, "loss": 0.0046, "step": 16338 }, { "epoch": 1.771357328707719, "grad_norm": 1.9624735116958618, "learning_rate": 8.220658639199188e-06, "loss": 0.0301, "step": 16339 }, { "epoch": 1.7714657415437989, "grad_norm": 0.2510398030281067, "learning_rate": 8.219933265631802e-06, "loss": 0.0045, "step": 16340 }, { "epoch": 1.7715741543798784, "grad_norm": 4.2752299308776855, "learning_rate": 8.219207892064414e-06, "loss": 0.0274, "step": 16341 }, { "epoch": 1.7716825672159584, "grad_norm": 1.4524928331375122, "learning_rate": 8.218482518497026e-06, "loss": 0.0439, "step": 16342 }, { "epoch": 1.7717909800520382, "grad_norm": 1.6553783416748047, "learning_rate": 8.21775714492964e-06, "loss": 0.0093, "step": 16343 }, { "epoch": 1.771899392888118, "grad_norm": 1.5396138429641724, "learning_rate": 8.217031771362254e-06, "loss": 0.0155, "step": 16344 }, { "epoch": 1.7720078057241977, "grad_norm": 3.9450435638427734, "learning_rate": 8.216306397794866e-06, "loss": 0.0293, "step": 16345 }, { "epoch": 1.7721162185602775, "grad_norm": 2.875330686569214, "learning_rate": 8.215581024227478e-06, "loss": 0.042, "step": 16346 }, { "epoch": 1.7722246313963574, "grad_norm": 1.9244494438171387, "learning_rate": 8.21485565066009e-06, "loss": 0.0278, "step": 16347 }, { "epoch": 1.772333044232437, "grad_norm": 0.5168846845626831, "learning_rate": 8.214130277092704e-06, "loss": 0.0051, "step": 16348 }, { "epoch": 1.772441457068517, "grad_norm": 4.035820484161377, "learning_rate": 8.213404903525316e-06, "loss": 0.0233, "step": 16349 }, { "epoch": 1.7725498699045967, "grad_norm": 6.420061111450195, "learning_rate": 8.212679529957928e-06, "loss": 0.0303, "step": 16350 }, { "epoch": 1.7726582827406765, "grad_norm": 1.2476811408996582, "learning_rate": 8.211954156390542e-06, "loss": 0.0135, "step": 16351 }, { "epoch": 1.7727666955767563, "grad_norm": 1.1408021450042725, "learning_rate": 8.211228782823154e-06, "loss": 0.0136, "step": 16352 }, { "epoch": 1.772875108412836, "grad_norm": 0.6073698997497559, "learning_rate": 8.210503409255768e-06, "loss": 0.0051, "step": 16353 }, { "epoch": 1.772983521248916, "grad_norm": 0.07200471311807632, "learning_rate": 8.20977803568838e-06, "loss": 0.0006, "step": 16354 }, { "epoch": 1.7730919340849955, "grad_norm": 0.928810179233551, "learning_rate": 8.209052662120994e-06, "loss": 0.005, "step": 16355 }, { "epoch": 1.7732003469210755, "grad_norm": 4.963531017303467, "learning_rate": 8.208327288553606e-06, "loss": 0.0325, "step": 16356 }, { "epoch": 1.7733087597571553, "grad_norm": 0.5570686459541321, "learning_rate": 8.207601914986218e-06, "loss": 0.0187, "step": 16357 }, { "epoch": 1.773417172593235, "grad_norm": 4.986204147338867, "learning_rate": 8.20687654141883e-06, "loss": 0.0208, "step": 16358 }, { "epoch": 1.7735255854293148, "grad_norm": 3.1883740425109863, "learning_rate": 8.206151167851444e-06, "loss": 0.0346, "step": 16359 }, { "epoch": 1.7736339982653946, "grad_norm": 6.230443000793457, "learning_rate": 8.205425794284056e-06, "loss": 0.0083, "step": 16360 }, { "epoch": 1.7737424111014746, "grad_norm": 0.5005254745483398, "learning_rate": 8.20470042071667e-06, "loss": 0.0242, "step": 16361 }, { "epoch": 1.773850823937554, "grad_norm": 0.38520246744155884, "learning_rate": 8.203975047149282e-06, "loss": 0.0092, "step": 16362 }, { "epoch": 1.773959236773634, "grad_norm": 1.1301857233047485, "learning_rate": 8.203249673581896e-06, "loss": 0.0257, "step": 16363 }, { "epoch": 1.7740676496097136, "grad_norm": 0.9157657623291016, "learning_rate": 8.202524300014508e-06, "loss": 0.0036, "step": 16364 }, { "epoch": 1.7741760624457936, "grad_norm": 1.3145328760147095, "learning_rate": 8.20179892644712e-06, "loss": 0.0124, "step": 16365 }, { "epoch": 1.7742844752818734, "grad_norm": 0.995822548866272, "learning_rate": 8.201073552879734e-06, "loss": 0.0035, "step": 16366 }, { "epoch": 1.7743928881179531, "grad_norm": 1.0013959407806396, "learning_rate": 8.200348179312346e-06, "loss": 0.0148, "step": 16367 }, { "epoch": 1.774501300954033, "grad_norm": 0.30319744348526, "learning_rate": 8.199622805744959e-06, "loss": 0.0109, "step": 16368 }, { "epoch": 1.7746097137901127, "grad_norm": 0.6131674647331238, "learning_rate": 8.198897432177572e-06, "loss": 0.0021, "step": 16369 }, { "epoch": 1.7747181266261927, "grad_norm": 0.2116938829421997, "learning_rate": 8.198172058610186e-06, "loss": 0.0042, "step": 16370 }, { "epoch": 1.7748265394622722, "grad_norm": 0.3243589997291565, "learning_rate": 8.197446685042798e-06, "loss": 0.0022, "step": 16371 }, { "epoch": 1.7749349522983522, "grad_norm": 5.233452796936035, "learning_rate": 8.19672131147541e-06, "loss": 0.026, "step": 16372 }, { "epoch": 1.775043365134432, "grad_norm": 0.48457419872283936, "learning_rate": 8.195995937908023e-06, "loss": 0.0128, "step": 16373 }, { "epoch": 1.7751517779705117, "grad_norm": 0.8846068978309631, "learning_rate": 8.195270564340636e-06, "loss": 0.0078, "step": 16374 }, { "epoch": 1.7752601908065915, "grad_norm": 0.3346388041973114, "learning_rate": 8.194545190773249e-06, "loss": 0.0108, "step": 16375 }, { "epoch": 1.7753686036426712, "grad_norm": 2.7522566318511963, "learning_rate": 8.19381981720586e-06, "loss": 0.0473, "step": 16376 }, { "epoch": 1.7754770164787512, "grad_norm": 0.6746419072151184, "learning_rate": 8.193094443638475e-06, "loss": 0.0173, "step": 16377 }, { "epoch": 1.7755854293148308, "grad_norm": 0.771036684513092, "learning_rate": 8.192369070071087e-06, "loss": 0.0016, "step": 16378 }, { "epoch": 1.7756938421509108, "grad_norm": 7.406538486480713, "learning_rate": 8.1916436965037e-06, "loss": 0.0101, "step": 16379 }, { "epoch": 1.7758022549869905, "grad_norm": 0.29024237394332886, "learning_rate": 8.190918322936313e-06, "loss": 0.0015, "step": 16380 }, { "epoch": 1.7759106678230703, "grad_norm": 7.08880090713501, "learning_rate": 8.190192949368927e-06, "loss": 0.0249, "step": 16381 }, { "epoch": 1.77601908065915, "grad_norm": 1.3654487133026123, "learning_rate": 8.189467575801539e-06, "loss": 0.0134, "step": 16382 }, { "epoch": 1.7761274934952298, "grad_norm": 0.930522620677948, "learning_rate": 8.188742202234151e-06, "loss": 0.0023, "step": 16383 }, { "epoch": 1.7762359063313098, "grad_norm": 7.100279331207275, "learning_rate": 8.188016828666763e-06, "loss": 0.0194, "step": 16384 }, { "epoch": 1.7763443191673893, "grad_norm": 1.5640065670013428, "learning_rate": 8.187291455099377e-06, "loss": 0.0065, "step": 16385 }, { "epoch": 1.7764527320034693, "grad_norm": 5.38766622543335, "learning_rate": 8.186566081531989e-06, "loss": 0.0251, "step": 16386 }, { "epoch": 1.7765611448395489, "grad_norm": 1.775382399559021, "learning_rate": 8.185840707964603e-06, "loss": 0.0173, "step": 16387 }, { "epoch": 1.7766695576756288, "grad_norm": 2.85263729095459, "learning_rate": 8.185115334397215e-06, "loss": 0.0187, "step": 16388 }, { "epoch": 1.7767779705117086, "grad_norm": 3.425032377243042, "learning_rate": 8.184389960829829e-06, "loss": 0.0225, "step": 16389 }, { "epoch": 1.7768863833477884, "grad_norm": 2.625356912612915, "learning_rate": 8.183664587262441e-06, "loss": 0.0482, "step": 16390 }, { "epoch": 1.7769947961838681, "grad_norm": 0.6289578080177307, "learning_rate": 8.182939213695053e-06, "loss": 0.0041, "step": 16391 }, { "epoch": 1.777103209019948, "grad_norm": 1.1573853492736816, "learning_rate": 8.182213840127667e-06, "loss": 0.014, "step": 16392 }, { "epoch": 1.7772116218560279, "grad_norm": 0.4048413336277008, "learning_rate": 8.181488466560279e-06, "loss": 0.0092, "step": 16393 }, { "epoch": 1.7773200346921074, "grad_norm": 3.0892536640167236, "learning_rate": 8.180763092992891e-06, "loss": 0.045, "step": 16394 }, { "epoch": 1.7774284475281874, "grad_norm": 0.34900200366973877, "learning_rate": 8.180037719425505e-06, "loss": 0.0163, "step": 16395 }, { "epoch": 1.7775368603642672, "grad_norm": 3.0373408794403076, "learning_rate": 8.179312345858119e-06, "loss": 0.0172, "step": 16396 }, { "epoch": 1.777645273200347, "grad_norm": 0.2110893577337265, "learning_rate": 8.178586972290731e-06, "loss": 0.001, "step": 16397 }, { "epoch": 1.7777536860364267, "grad_norm": 2.0593533515930176, "learning_rate": 8.177861598723343e-06, "loss": 0.0358, "step": 16398 }, { "epoch": 1.7778620988725065, "grad_norm": 1.8171193599700928, "learning_rate": 8.177136225155955e-06, "loss": 0.0067, "step": 16399 }, { "epoch": 1.7779705117085864, "grad_norm": 1.2399733066558838, "learning_rate": 8.17641085158857e-06, "loss": 0.0043, "step": 16400 }, { "epoch": 1.778078924544666, "grad_norm": 0.6112927198410034, "learning_rate": 8.175685478021181e-06, "loss": 0.0032, "step": 16401 }, { "epoch": 1.778187337380746, "grad_norm": 0.290296345949173, "learning_rate": 8.174960104453793e-06, "loss": 0.0192, "step": 16402 }, { "epoch": 1.7782957502168255, "grad_norm": 0.5547267198562622, "learning_rate": 8.174234730886407e-06, "loss": 0.0073, "step": 16403 }, { "epoch": 1.7784041630529055, "grad_norm": 0.3389599621295929, "learning_rate": 8.173509357319021e-06, "loss": 0.0055, "step": 16404 }, { "epoch": 1.7785125758889853, "grad_norm": 1.0736057758331299, "learning_rate": 8.172783983751633e-06, "loss": 0.0159, "step": 16405 }, { "epoch": 1.778620988725065, "grad_norm": 3.219564914703369, "learning_rate": 8.172058610184245e-06, "loss": 0.0395, "step": 16406 }, { "epoch": 1.7787294015611448, "grad_norm": 0.38638433814048767, "learning_rate": 8.17133323661686e-06, "loss": 0.0118, "step": 16407 }, { "epoch": 1.7788378143972245, "grad_norm": 1.3351346254348755, "learning_rate": 8.170607863049471e-06, "loss": 0.0159, "step": 16408 }, { "epoch": 1.7789462272333045, "grad_norm": 1.8265899419784546, "learning_rate": 8.169882489482084e-06, "loss": 0.0195, "step": 16409 }, { "epoch": 1.779054640069384, "grad_norm": 0.5098327994346619, "learning_rate": 8.169157115914696e-06, "loss": 0.0143, "step": 16410 }, { "epoch": 1.779163052905464, "grad_norm": 0.9562959671020508, "learning_rate": 8.16843174234731e-06, "loss": 0.0029, "step": 16411 }, { "epoch": 1.7792714657415438, "grad_norm": 2.021974563598633, "learning_rate": 8.167706368779922e-06, "loss": 0.0251, "step": 16412 }, { "epoch": 1.7793798785776236, "grad_norm": 1.1471999883651733, "learning_rate": 8.166980995212535e-06, "loss": 0.0357, "step": 16413 }, { "epoch": 1.7794882914137033, "grad_norm": 5.329068183898926, "learning_rate": 8.166255621645148e-06, "loss": 0.0347, "step": 16414 }, { "epoch": 1.7795967042497831, "grad_norm": 3.689772367477417, "learning_rate": 8.165530248077761e-06, "loss": 0.0666, "step": 16415 }, { "epoch": 1.779705117085863, "grad_norm": 1.5060149431228638, "learning_rate": 8.164804874510374e-06, "loss": 0.0196, "step": 16416 }, { "epoch": 1.7798135299219426, "grad_norm": 1.7062727212905884, "learning_rate": 8.164079500942986e-06, "loss": 0.0369, "step": 16417 }, { "epoch": 1.7799219427580226, "grad_norm": 1.3991895914077759, "learning_rate": 8.1633541273756e-06, "loss": 0.0341, "step": 16418 }, { "epoch": 1.7800303555941024, "grad_norm": 0.3221709132194519, "learning_rate": 8.162628753808212e-06, "loss": 0.0043, "step": 16419 }, { "epoch": 1.7801387684301822, "grad_norm": 3.145838975906372, "learning_rate": 8.161903380240824e-06, "loss": 0.0463, "step": 16420 }, { "epoch": 1.780247181266262, "grad_norm": 1.1220004558563232, "learning_rate": 8.161178006673438e-06, "loss": 0.0243, "step": 16421 }, { "epoch": 1.7803555941023417, "grad_norm": 0.0368419773876667, "learning_rate": 8.160452633106052e-06, "loss": 0.0002, "step": 16422 }, { "epoch": 1.7804640069384217, "grad_norm": 2.612186908721924, "learning_rate": 8.159727259538664e-06, "loss": 0.0125, "step": 16423 }, { "epoch": 1.7805724197745012, "grad_norm": 6.098625183105469, "learning_rate": 8.159001885971276e-06, "loss": 0.0324, "step": 16424 }, { "epoch": 1.7806808326105812, "grad_norm": 2.262968063354492, "learning_rate": 8.158276512403888e-06, "loss": 0.0188, "step": 16425 }, { "epoch": 1.7807892454466607, "grad_norm": 1.9435486793518066, "learning_rate": 8.157551138836502e-06, "loss": 0.0299, "step": 16426 }, { "epoch": 1.7808976582827407, "grad_norm": 2.4491536617279053, "learning_rate": 8.156825765269114e-06, "loss": 0.0034, "step": 16427 }, { "epoch": 1.7810060711188205, "grad_norm": 8.63616943359375, "learning_rate": 8.156100391701726e-06, "loss": 0.0504, "step": 16428 }, { "epoch": 1.7811144839549002, "grad_norm": 5.312628746032715, "learning_rate": 8.15537501813434e-06, "loss": 0.0337, "step": 16429 }, { "epoch": 1.78122289679098, "grad_norm": 12.550161361694336, "learning_rate": 8.154649644566954e-06, "loss": 0.0492, "step": 16430 }, { "epoch": 1.7813313096270598, "grad_norm": 2.263486385345459, "learning_rate": 8.153924270999566e-06, "loss": 0.009, "step": 16431 }, { "epoch": 1.7814397224631398, "grad_norm": 5.730537414550781, "learning_rate": 8.153198897432178e-06, "loss": 0.0294, "step": 16432 }, { "epoch": 1.7815481352992193, "grad_norm": 10.9924898147583, "learning_rate": 8.152473523864792e-06, "loss": 0.0144, "step": 16433 }, { "epoch": 1.7816565481352993, "grad_norm": 0.856234073638916, "learning_rate": 8.151748150297404e-06, "loss": 0.0028, "step": 16434 }, { "epoch": 1.781764960971379, "grad_norm": 0.7109564542770386, "learning_rate": 8.151022776730016e-06, "loss": 0.0091, "step": 16435 }, { "epoch": 1.7818733738074588, "grad_norm": 3.451127290725708, "learning_rate": 8.150297403162628e-06, "loss": 0.0254, "step": 16436 }, { "epoch": 1.7819817866435386, "grad_norm": 0.6685019731521606, "learning_rate": 8.149572029595242e-06, "loss": 0.0208, "step": 16437 }, { "epoch": 1.7820901994796183, "grad_norm": 14.94678020477295, "learning_rate": 8.148846656027854e-06, "loss": 0.0206, "step": 16438 }, { "epoch": 1.7821986123156983, "grad_norm": 2.189912796020508, "learning_rate": 8.148121282460468e-06, "loss": 0.0174, "step": 16439 }, { "epoch": 1.7823070251517779, "grad_norm": 0.28969717025756836, "learning_rate": 8.14739590889308e-06, "loss": 0.0059, "step": 16440 }, { "epoch": 1.7824154379878578, "grad_norm": 1.2546228170394897, "learning_rate": 8.146670535325694e-06, "loss": 0.0174, "step": 16441 }, { "epoch": 1.7825238508239376, "grad_norm": 3.0468125343322754, "learning_rate": 8.145945161758306e-06, "loss": 0.0317, "step": 16442 }, { "epoch": 1.7826322636600174, "grad_norm": 1.1334552764892578, "learning_rate": 8.145219788190918e-06, "loss": 0.0023, "step": 16443 }, { "epoch": 1.7827406764960971, "grad_norm": 7.4444780349731445, "learning_rate": 8.144494414623532e-06, "loss": 0.0299, "step": 16444 }, { "epoch": 1.782849089332177, "grad_norm": 0.7583389282226562, "learning_rate": 8.143769041056144e-06, "loss": 0.0025, "step": 16445 }, { "epoch": 1.7829575021682569, "grad_norm": 0.2464001178741455, "learning_rate": 8.143043667488757e-06, "loss": 0.0014, "step": 16446 }, { "epoch": 1.7830659150043364, "grad_norm": 0.6311344504356384, "learning_rate": 8.14231829392137e-06, "loss": 0.0049, "step": 16447 }, { "epoch": 1.7831743278404164, "grad_norm": 1.4513027667999268, "learning_rate": 8.141592920353984e-06, "loss": 0.0298, "step": 16448 }, { "epoch": 1.783282740676496, "grad_norm": 0.34543460607528687, "learning_rate": 8.140867546786596e-06, "loss": 0.0076, "step": 16449 }, { "epoch": 1.783391153512576, "grad_norm": 0.8228883743286133, "learning_rate": 8.140142173219209e-06, "loss": 0.0273, "step": 16450 }, { "epoch": 1.7834995663486557, "grad_norm": 2.8930323123931885, "learning_rate": 8.13941679965182e-06, "loss": 0.0401, "step": 16451 }, { "epoch": 1.7836079791847355, "grad_norm": 3.0288052558898926, "learning_rate": 8.138691426084434e-06, "loss": 0.0294, "step": 16452 }, { "epoch": 1.7837163920208152, "grad_norm": 1.3369516134262085, "learning_rate": 8.137966052517047e-06, "loss": 0.0381, "step": 16453 }, { "epoch": 1.783824804856895, "grad_norm": 4.197065353393555, "learning_rate": 8.137240678949659e-06, "loss": 0.0777, "step": 16454 }, { "epoch": 1.783933217692975, "grad_norm": 0.07895119488239288, "learning_rate": 8.136515305382273e-06, "loss": 0.0004, "step": 16455 }, { "epoch": 1.7840416305290545, "grad_norm": 1.7405610084533691, "learning_rate": 8.135789931814886e-06, "loss": 0.0207, "step": 16456 }, { "epoch": 1.7841500433651345, "grad_norm": 3.141008138656616, "learning_rate": 8.135064558247499e-06, "loss": 0.044, "step": 16457 }, { "epoch": 1.7842584562012143, "grad_norm": 0.39322683215141296, "learning_rate": 8.13433918468011e-06, "loss": 0.0224, "step": 16458 }, { "epoch": 1.784366869037294, "grad_norm": 0.2246096283197403, "learning_rate": 8.133613811112725e-06, "loss": 0.0042, "step": 16459 }, { "epoch": 1.7844752818733738, "grad_norm": 1.4622104167938232, "learning_rate": 8.132888437545337e-06, "loss": 0.0163, "step": 16460 }, { "epoch": 1.7845836947094535, "grad_norm": 0.6403701305389404, "learning_rate": 8.132163063977949e-06, "loss": 0.0137, "step": 16461 }, { "epoch": 1.7846921075455335, "grad_norm": 0.5527397394180298, "learning_rate": 8.131437690410563e-06, "loss": 0.0166, "step": 16462 }, { "epoch": 1.784800520381613, "grad_norm": 1.4105764627456665, "learning_rate": 8.130712316843175e-06, "loss": 0.0257, "step": 16463 }, { "epoch": 1.784908933217693, "grad_norm": 0.385247141122818, "learning_rate": 8.129986943275787e-06, "loss": 0.0017, "step": 16464 }, { "epoch": 1.7850173460537726, "grad_norm": 4.339653491973877, "learning_rate": 8.1292615697084e-06, "loss": 0.0393, "step": 16465 }, { "epoch": 1.7851257588898526, "grad_norm": 6.1200947761535645, "learning_rate": 8.128536196141013e-06, "loss": 0.0842, "step": 16466 }, { "epoch": 1.7852341717259324, "grad_norm": 5.3043904304504395, "learning_rate": 8.127810822573627e-06, "loss": 0.0191, "step": 16467 }, { "epoch": 1.7853425845620121, "grad_norm": 2.12804913520813, "learning_rate": 8.127085449006239e-06, "loss": 0.0401, "step": 16468 }, { "epoch": 1.785450997398092, "grad_norm": 11.307563781738281, "learning_rate": 8.126360075438851e-06, "loss": 0.0409, "step": 16469 }, { "epoch": 1.7855594102341716, "grad_norm": 0.9892942905426025, "learning_rate": 8.125634701871465e-06, "loss": 0.0331, "step": 16470 }, { "epoch": 1.7856678230702516, "grad_norm": 2.963538408279419, "learning_rate": 8.124909328304077e-06, "loss": 0.012, "step": 16471 }, { "epoch": 1.7857762359063312, "grad_norm": 2.2463533878326416, "learning_rate": 8.12418395473669e-06, "loss": 0.0605, "step": 16472 }, { "epoch": 1.7858846487424112, "grad_norm": 2.0295026302337646, "learning_rate": 8.123458581169303e-06, "loss": 0.0143, "step": 16473 }, { "epoch": 1.785993061578491, "grad_norm": 1.4231066703796387, "learning_rate": 8.122733207601917e-06, "loss": 0.0227, "step": 16474 }, { "epoch": 1.7861014744145707, "grad_norm": 0.9038627743721008, "learning_rate": 8.122007834034529e-06, "loss": 0.0203, "step": 16475 }, { "epoch": 1.7862098872506504, "grad_norm": 0.6503868103027344, "learning_rate": 8.121282460467141e-06, "loss": 0.0153, "step": 16476 }, { "epoch": 1.7863183000867302, "grad_norm": 0.3398912847042084, "learning_rate": 8.120557086899753e-06, "loss": 0.0123, "step": 16477 }, { "epoch": 1.7864267129228102, "grad_norm": 0.7790567278862, "learning_rate": 8.119831713332367e-06, "loss": 0.0057, "step": 16478 }, { "epoch": 1.7865351257588897, "grad_norm": 0.4623921811580658, "learning_rate": 8.11910633976498e-06, "loss": 0.0039, "step": 16479 }, { "epoch": 1.7866435385949697, "grad_norm": 0.8098152875900269, "learning_rate": 8.118380966197591e-06, "loss": 0.0121, "step": 16480 }, { "epoch": 1.7867519514310495, "grad_norm": 4.270035743713379, "learning_rate": 8.117655592630205e-06, "loss": 0.0199, "step": 16481 }, { "epoch": 1.7868603642671292, "grad_norm": 1.5734673738479614, "learning_rate": 8.116930219062819e-06, "loss": 0.0272, "step": 16482 }, { "epoch": 1.786968777103209, "grad_norm": 0.2409801036119461, "learning_rate": 8.116204845495431e-06, "loss": 0.0019, "step": 16483 }, { "epoch": 1.7870771899392888, "grad_norm": 3.197852611541748, "learning_rate": 8.115479471928043e-06, "loss": 0.0158, "step": 16484 }, { "epoch": 1.7871856027753688, "grad_norm": 0.7575893402099609, "learning_rate": 8.114754098360657e-06, "loss": 0.0048, "step": 16485 }, { "epoch": 1.7872940156114483, "grad_norm": 0.5853466391563416, "learning_rate": 8.11402872479327e-06, "loss": 0.0239, "step": 16486 }, { "epoch": 1.7874024284475283, "grad_norm": 0.3327122926712036, "learning_rate": 8.113303351225882e-06, "loss": 0.0018, "step": 16487 }, { "epoch": 1.7875108412836078, "grad_norm": 1.6032038927078247, "learning_rate": 8.112577977658495e-06, "loss": 0.0164, "step": 16488 }, { "epoch": 1.7876192541196878, "grad_norm": 0.5520350337028503, "learning_rate": 8.111852604091108e-06, "loss": 0.0237, "step": 16489 }, { "epoch": 1.7877276669557676, "grad_norm": 0.07045824080705643, "learning_rate": 8.11112723052372e-06, "loss": 0.0006, "step": 16490 }, { "epoch": 1.7878360797918473, "grad_norm": 5.430344581604004, "learning_rate": 8.110401856956333e-06, "loss": 0.0448, "step": 16491 }, { "epoch": 1.787944492627927, "grad_norm": 2.7131378650665283, "learning_rate": 8.109676483388946e-06, "loss": 0.0633, "step": 16492 }, { "epoch": 1.7880529054640069, "grad_norm": 7.083757400512695, "learning_rate": 8.10895110982156e-06, "loss": 0.0431, "step": 16493 }, { "epoch": 1.7881613183000868, "grad_norm": 2.176177740097046, "learning_rate": 8.108225736254172e-06, "loss": 0.0138, "step": 16494 }, { "epoch": 1.7882697311361664, "grad_norm": 0.3327833414077759, "learning_rate": 8.107500362686784e-06, "loss": 0.0076, "step": 16495 }, { "epoch": 1.7883781439722464, "grad_norm": 6.303095817565918, "learning_rate": 8.106774989119398e-06, "loss": 0.0298, "step": 16496 }, { "epoch": 1.7884865568083261, "grad_norm": 0.5585796236991882, "learning_rate": 8.10604961555201e-06, "loss": 0.0092, "step": 16497 }, { "epoch": 1.788594969644406, "grad_norm": 0.5225658416748047, "learning_rate": 8.105324241984622e-06, "loss": 0.0145, "step": 16498 }, { "epoch": 1.7887033824804857, "grad_norm": 4.563345909118652, "learning_rate": 8.104598868417236e-06, "loss": 0.0201, "step": 16499 }, { "epoch": 1.7888117953165654, "grad_norm": 0.7588600516319275, "learning_rate": 8.10387349484985e-06, "loss": 0.0032, "step": 16500 }, { "epoch": 1.7889202081526454, "grad_norm": 2.5062835216522217, "learning_rate": 8.103148121282462e-06, "loss": 0.0251, "step": 16501 }, { "epoch": 1.789028620988725, "grad_norm": 0.3354646861553192, "learning_rate": 8.102422747715074e-06, "loss": 0.0195, "step": 16502 }, { "epoch": 1.789137033824805, "grad_norm": 4.808438301086426, "learning_rate": 8.101697374147686e-06, "loss": 0.0143, "step": 16503 }, { "epoch": 1.7892454466608847, "grad_norm": 0.2013993263244629, "learning_rate": 8.1009720005803e-06, "loss": 0.0019, "step": 16504 }, { "epoch": 1.7893538594969645, "grad_norm": 1.497483253479004, "learning_rate": 8.100246627012912e-06, "loss": 0.0165, "step": 16505 }, { "epoch": 1.7894622723330442, "grad_norm": 0.5588499903678894, "learning_rate": 8.099521253445524e-06, "loss": 0.0033, "step": 16506 }, { "epoch": 1.789570685169124, "grad_norm": 0.32970353960990906, "learning_rate": 8.098795879878138e-06, "loss": 0.0185, "step": 16507 }, { "epoch": 1.789679098005204, "grad_norm": 1.4741415977478027, "learning_rate": 8.098070506310752e-06, "loss": 0.0324, "step": 16508 }, { "epoch": 1.7897875108412835, "grad_norm": 1.856790542602539, "learning_rate": 8.097345132743364e-06, "loss": 0.0244, "step": 16509 }, { "epoch": 1.7898959236773635, "grad_norm": 0.5661352276802063, "learning_rate": 8.096619759175976e-06, "loss": 0.0099, "step": 16510 }, { "epoch": 1.790004336513443, "grad_norm": 0.5713467597961426, "learning_rate": 8.09589438560859e-06, "loss": 0.0083, "step": 16511 }, { "epoch": 1.790112749349523, "grad_norm": 0.49037447571754456, "learning_rate": 8.095169012041202e-06, "loss": 0.023, "step": 16512 }, { "epoch": 1.7902211621856028, "grad_norm": 1.5372974872589111, "learning_rate": 8.094443638473814e-06, "loss": 0.0054, "step": 16513 }, { "epoch": 1.7903295750216826, "grad_norm": 0.3345696032047272, "learning_rate": 8.093718264906428e-06, "loss": 0.0078, "step": 16514 }, { "epoch": 1.7904379878577623, "grad_norm": 0.3292703926563263, "learning_rate": 8.09299289133904e-06, "loss": 0.0115, "step": 16515 }, { "epoch": 1.790546400693842, "grad_norm": 4.687192916870117, "learning_rate": 8.092267517771652e-06, "loss": 0.0269, "step": 16516 }, { "epoch": 1.790654813529922, "grad_norm": 0.5835212469100952, "learning_rate": 8.091542144204266e-06, "loss": 0.0034, "step": 16517 }, { "epoch": 1.7907632263660016, "grad_norm": 2.6472654342651367, "learning_rate": 8.090816770636878e-06, "loss": 0.0239, "step": 16518 }, { "epoch": 1.7908716392020816, "grad_norm": 3.9108636379241943, "learning_rate": 8.090091397069492e-06, "loss": 0.0101, "step": 16519 }, { "epoch": 1.7909800520381614, "grad_norm": 2.0654289722442627, "learning_rate": 8.089366023502104e-06, "loss": 0.0407, "step": 16520 }, { "epoch": 1.7910884648742411, "grad_norm": 2.056192636489868, "learning_rate": 8.088640649934716e-06, "loss": 0.0028, "step": 16521 }, { "epoch": 1.7911968777103209, "grad_norm": 4.770427703857422, "learning_rate": 8.08791527636733e-06, "loss": 0.021, "step": 16522 }, { "epoch": 1.7913052905464006, "grad_norm": 0.8410387635231018, "learning_rate": 8.087189902799942e-06, "loss": 0.0046, "step": 16523 }, { "epoch": 1.7914137033824806, "grad_norm": 1.1155283451080322, "learning_rate": 8.086464529232555e-06, "loss": 0.0037, "step": 16524 }, { "epoch": 1.7915221162185602, "grad_norm": 0.3499533236026764, "learning_rate": 8.085739155665168e-06, "loss": 0.0067, "step": 16525 }, { "epoch": 1.7916305290546402, "grad_norm": 1.921849250793457, "learning_rate": 8.085013782097782e-06, "loss": 0.0269, "step": 16526 }, { "epoch": 1.79173894189072, "grad_norm": 0.09609290957450867, "learning_rate": 8.084288408530394e-06, "loss": 0.0004, "step": 16527 }, { "epoch": 1.7918473547267997, "grad_norm": 0.30411624908447266, "learning_rate": 8.083563034963007e-06, "loss": 0.0099, "step": 16528 }, { "epoch": 1.7919557675628794, "grad_norm": 8.76191520690918, "learning_rate": 8.082837661395619e-06, "loss": 0.0407, "step": 16529 }, { "epoch": 1.7920641803989592, "grad_norm": 4.000940322875977, "learning_rate": 8.082112287828233e-06, "loss": 0.0115, "step": 16530 }, { "epoch": 1.7921725932350392, "grad_norm": 8.463537216186523, "learning_rate": 8.081386914260845e-06, "loss": 0.0353, "step": 16531 }, { "epoch": 1.7922810060711187, "grad_norm": 0.2973676323890686, "learning_rate": 8.080661540693457e-06, "loss": 0.0123, "step": 16532 }, { "epoch": 1.7923894189071987, "grad_norm": 3.2293829917907715, "learning_rate": 8.07993616712607e-06, "loss": 0.047, "step": 16533 }, { "epoch": 1.7924978317432783, "grad_norm": 3.4786648750305176, "learning_rate": 8.079210793558684e-06, "loss": 0.0284, "step": 16534 }, { "epoch": 1.7926062445793582, "grad_norm": 0.7369047999382019, "learning_rate": 8.078485419991297e-06, "loss": 0.0138, "step": 16535 }, { "epoch": 1.792714657415438, "grad_norm": 2.097276449203491, "learning_rate": 8.077760046423909e-06, "loss": 0.0358, "step": 16536 }, { "epoch": 1.7928230702515178, "grad_norm": 1.0495558977127075, "learning_rate": 8.077034672856523e-06, "loss": 0.0111, "step": 16537 }, { "epoch": 1.7929314830875975, "grad_norm": 5.192250728607178, "learning_rate": 8.076309299289135e-06, "loss": 0.0566, "step": 16538 }, { "epoch": 1.7930398959236773, "grad_norm": 0.11465122550725937, "learning_rate": 8.075583925721747e-06, "loss": 0.0009, "step": 16539 }, { "epoch": 1.7931483087597573, "grad_norm": 0.09035471081733704, "learning_rate": 8.07485855215436e-06, "loss": 0.0006, "step": 16540 }, { "epoch": 1.7932567215958368, "grad_norm": 0.5607683062553406, "learning_rate": 8.074133178586973e-06, "loss": 0.0185, "step": 16541 }, { "epoch": 1.7933651344319168, "grad_norm": 8.939566612243652, "learning_rate": 8.073407805019585e-06, "loss": 0.0175, "step": 16542 }, { "epoch": 1.7934735472679966, "grad_norm": 5.0130720138549805, "learning_rate": 8.072682431452199e-06, "loss": 0.0306, "step": 16543 }, { "epoch": 1.7935819601040763, "grad_norm": 4.377624034881592, "learning_rate": 8.071957057884811e-06, "loss": 0.013, "step": 16544 }, { "epoch": 1.793690372940156, "grad_norm": 3.0930426120758057, "learning_rate": 8.071231684317425e-06, "loss": 0.0106, "step": 16545 }, { "epoch": 1.7937987857762359, "grad_norm": 0.6954178214073181, "learning_rate": 8.070506310750037e-06, "loss": 0.0098, "step": 16546 }, { "epoch": 1.7939071986123158, "grad_norm": 2.057575225830078, "learning_rate": 8.069780937182649e-06, "loss": 0.0398, "step": 16547 }, { "epoch": 1.7940156114483954, "grad_norm": 1.5595141649246216, "learning_rate": 8.069055563615263e-06, "loss": 0.0104, "step": 16548 }, { "epoch": 1.7941240242844754, "grad_norm": 5.361660480499268, "learning_rate": 8.068330190047875e-06, "loss": 0.061, "step": 16549 }, { "epoch": 1.794232437120555, "grad_norm": 0.4103001356124878, "learning_rate": 8.067604816480487e-06, "loss": 0.0022, "step": 16550 }, { "epoch": 1.794340849956635, "grad_norm": 0.9311861395835876, "learning_rate": 8.066879442913101e-06, "loss": 0.0164, "step": 16551 }, { "epoch": 1.7944492627927147, "grad_norm": 0.6683372259140015, "learning_rate": 8.066154069345715e-06, "loss": 0.0195, "step": 16552 }, { "epoch": 1.7945576756287944, "grad_norm": 1.9273722171783447, "learning_rate": 8.065428695778327e-06, "loss": 0.0037, "step": 16553 }, { "epoch": 1.7946660884648742, "grad_norm": 1.9351818561553955, "learning_rate": 8.06470332221094e-06, "loss": 0.0071, "step": 16554 }, { "epoch": 1.794774501300954, "grad_norm": 0.47911524772644043, "learning_rate": 8.063977948643551e-06, "loss": 0.0024, "step": 16555 }, { "epoch": 1.794882914137034, "grad_norm": 4.108949661254883, "learning_rate": 8.063252575076165e-06, "loss": 0.0372, "step": 16556 }, { "epoch": 1.7949913269731135, "grad_norm": 2.5591344833374023, "learning_rate": 8.062527201508777e-06, "loss": 0.0253, "step": 16557 }, { "epoch": 1.7950997398091935, "grad_norm": 0.13444966077804565, "learning_rate": 8.06180182794139e-06, "loss": 0.0013, "step": 16558 }, { "epoch": 1.7952081526452732, "grad_norm": 0.11867327988147736, "learning_rate": 8.061076454374003e-06, "loss": 0.001, "step": 16559 }, { "epoch": 1.795316565481353, "grad_norm": 0.795049250125885, "learning_rate": 8.060351080806617e-06, "loss": 0.0161, "step": 16560 }, { "epoch": 1.7954249783174328, "grad_norm": 7.726803779602051, "learning_rate": 8.05962570723923e-06, "loss": 0.0337, "step": 16561 }, { "epoch": 1.7955333911535125, "grad_norm": 4.413461685180664, "learning_rate": 8.058900333671841e-06, "loss": 0.0438, "step": 16562 }, { "epoch": 1.7956418039895925, "grad_norm": 0.29982057213783264, "learning_rate": 8.058174960104455e-06, "loss": 0.0088, "step": 16563 }, { "epoch": 1.795750216825672, "grad_norm": 3.0533206462860107, "learning_rate": 8.057449586537067e-06, "loss": 0.0462, "step": 16564 }, { "epoch": 1.795858629661752, "grad_norm": 6.804557800292969, "learning_rate": 8.05672421296968e-06, "loss": 0.0537, "step": 16565 }, { "epoch": 1.7959670424978318, "grad_norm": 0.26821038126945496, "learning_rate": 8.055998839402293e-06, "loss": 0.0008, "step": 16566 }, { "epoch": 1.7960754553339116, "grad_norm": 2.1653854846954346, "learning_rate": 8.055273465834906e-06, "loss": 0.0302, "step": 16567 }, { "epoch": 1.7961838681699913, "grad_norm": 16.42745018005371, "learning_rate": 8.054548092267518e-06, "loss": 0.0261, "step": 16568 }, { "epoch": 1.796292281006071, "grad_norm": 9.900450706481934, "learning_rate": 8.053822718700132e-06, "loss": 0.079, "step": 16569 }, { "epoch": 1.796400693842151, "grad_norm": 0.3498712480068207, "learning_rate": 8.053097345132744e-06, "loss": 0.0074, "step": 16570 }, { "epoch": 1.7965091066782306, "grad_norm": 12.272270202636719, "learning_rate": 8.052371971565357e-06, "loss": 0.0477, "step": 16571 }, { "epoch": 1.7966175195143106, "grad_norm": 3.4408457279205322, "learning_rate": 8.05164659799797e-06, "loss": 0.0182, "step": 16572 }, { "epoch": 1.7967259323503901, "grad_norm": 0.23218977451324463, "learning_rate": 8.050921224430582e-06, "loss": 0.0003, "step": 16573 }, { "epoch": 1.7968343451864701, "grad_norm": 1.9061392545700073, "learning_rate": 8.050195850863196e-06, "loss": 0.0085, "step": 16574 }, { "epoch": 1.7969427580225499, "grad_norm": 2.7446184158325195, "learning_rate": 8.049470477295808e-06, "loss": 0.0465, "step": 16575 }, { "epoch": 1.7970511708586296, "grad_norm": 1.5563650131225586, "learning_rate": 8.04874510372842e-06, "loss": 0.0329, "step": 16576 }, { "epoch": 1.7971595836947094, "grad_norm": 1.0476022958755493, "learning_rate": 8.048019730161034e-06, "loss": 0.0164, "step": 16577 }, { "epoch": 1.7972679965307892, "grad_norm": 0.44256314635276794, "learning_rate": 8.047294356593648e-06, "loss": 0.0221, "step": 16578 }, { "epoch": 1.7973764093668692, "grad_norm": 4.128744602203369, "learning_rate": 8.04656898302626e-06, "loss": 0.0071, "step": 16579 }, { "epoch": 1.7974848222029487, "grad_norm": 0.39885836839675903, "learning_rate": 8.045843609458872e-06, "loss": 0.0128, "step": 16580 }, { "epoch": 1.7975932350390287, "grad_norm": 2.6904757022857666, "learning_rate": 8.045118235891484e-06, "loss": 0.0729, "step": 16581 }, { "epoch": 1.7977016478751084, "grad_norm": 0.44276222586631775, "learning_rate": 8.044392862324098e-06, "loss": 0.0149, "step": 16582 }, { "epoch": 1.7978100607111882, "grad_norm": 2.2425267696380615, "learning_rate": 8.04366748875671e-06, "loss": 0.0084, "step": 16583 }, { "epoch": 1.797918473547268, "grad_norm": 0.5341245532035828, "learning_rate": 8.042942115189322e-06, "loss": 0.0194, "step": 16584 }, { "epoch": 1.7980268863833477, "grad_norm": 0.45105311274528503, "learning_rate": 8.042216741621936e-06, "loss": 0.0061, "step": 16585 }, { "epoch": 1.7981352992194277, "grad_norm": 0.5040265917778015, "learning_rate": 8.04149136805455e-06, "loss": 0.0023, "step": 16586 }, { "epoch": 1.7982437120555073, "grad_norm": 5.011329174041748, "learning_rate": 8.040765994487162e-06, "loss": 0.0271, "step": 16587 }, { "epoch": 1.7983521248915872, "grad_norm": 3.122875690460205, "learning_rate": 8.040040620919774e-06, "loss": 0.0631, "step": 16588 }, { "epoch": 1.798460537727667, "grad_norm": 0.15082065761089325, "learning_rate": 8.039315247352388e-06, "loss": 0.0016, "step": 16589 }, { "epoch": 1.7985689505637468, "grad_norm": 2.743412733078003, "learning_rate": 8.038589873785e-06, "loss": 0.0179, "step": 16590 }, { "epoch": 1.7986773633998265, "grad_norm": 0.3825589716434479, "learning_rate": 8.037864500217612e-06, "loss": 0.0024, "step": 16591 }, { "epoch": 1.7987857762359063, "grad_norm": 0.06665976345539093, "learning_rate": 8.037139126650226e-06, "loss": 0.0005, "step": 16592 }, { "epoch": 1.7988941890719863, "grad_norm": 0.1307278424501419, "learning_rate": 8.036413753082838e-06, "loss": 0.0114, "step": 16593 }, { "epoch": 1.7990026019080658, "grad_norm": 7.889293670654297, "learning_rate": 8.035688379515452e-06, "loss": 0.0622, "step": 16594 }, { "epoch": 1.7991110147441458, "grad_norm": 0.5979624390602112, "learning_rate": 8.034963005948064e-06, "loss": 0.0151, "step": 16595 }, { "epoch": 1.7992194275802254, "grad_norm": 2.3318796157836914, "learning_rate": 8.034237632380676e-06, "loss": 0.0106, "step": 16596 }, { "epoch": 1.7993278404163053, "grad_norm": 4.349052429199219, "learning_rate": 8.03351225881329e-06, "loss": 0.0073, "step": 16597 }, { "epoch": 1.799436253252385, "grad_norm": 1.1195350885391235, "learning_rate": 8.032786885245902e-06, "loss": 0.063, "step": 16598 }, { "epoch": 1.7995446660884649, "grad_norm": 2.470534324645996, "learning_rate": 8.032061511678514e-06, "loss": 0.0226, "step": 16599 }, { "epoch": 1.7996530789245446, "grad_norm": 3.749366044998169, "learning_rate": 8.031336138111128e-06, "loss": 0.0235, "step": 16600 }, { "epoch": 1.7997614917606244, "grad_norm": 2.6454365253448486, "learning_rate": 8.03061076454374e-06, "loss": 0.034, "step": 16601 }, { "epoch": 1.7998699045967044, "grad_norm": 3.066671133041382, "learning_rate": 8.029885390976353e-06, "loss": 0.0659, "step": 16602 }, { "epoch": 1.799978317432784, "grad_norm": 0.5782780051231384, "learning_rate": 8.029160017408966e-06, "loss": 0.0191, "step": 16603 }, { "epoch": 1.800086730268864, "grad_norm": 2.3948211669921875, "learning_rate": 8.02843464384158e-06, "loss": 0.0364, "step": 16604 }, { "epoch": 1.8001951431049437, "grad_norm": 0.0941288098692894, "learning_rate": 8.027709270274192e-06, "loss": 0.0011, "step": 16605 }, { "epoch": 1.8003035559410234, "grad_norm": 3.3434407711029053, "learning_rate": 8.026983896706805e-06, "loss": 0.0374, "step": 16606 }, { "epoch": 1.8004119687771032, "grad_norm": 1.3338696956634521, "learning_rate": 8.026258523139417e-06, "loss": 0.0255, "step": 16607 }, { "epoch": 1.800520381613183, "grad_norm": 3.1426446437835693, "learning_rate": 8.02553314957203e-06, "loss": 0.059, "step": 16608 }, { "epoch": 1.800628794449263, "grad_norm": 3.5454328060150146, "learning_rate": 8.024807776004643e-06, "loss": 0.018, "step": 16609 }, { "epoch": 1.8007372072853425, "grad_norm": 0.39580029249191284, "learning_rate": 8.024082402437255e-06, "loss": 0.0038, "step": 16610 }, { "epoch": 1.8008456201214225, "grad_norm": 0.3570394814014435, "learning_rate": 8.023357028869869e-06, "loss": 0.0118, "step": 16611 }, { "epoch": 1.800954032957502, "grad_norm": 11.23450756072998, "learning_rate": 8.022631655302482e-06, "loss": 0.0513, "step": 16612 }, { "epoch": 1.801062445793582, "grad_norm": 0.7823687791824341, "learning_rate": 8.021906281735095e-06, "loss": 0.0139, "step": 16613 }, { "epoch": 1.8011708586296618, "grad_norm": 8.217854499816895, "learning_rate": 8.021180908167707e-06, "loss": 0.0368, "step": 16614 }, { "epoch": 1.8012792714657415, "grad_norm": 0.3189030885696411, "learning_rate": 8.02045553460032e-06, "loss": 0.0028, "step": 16615 }, { "epoch": 1.8013876843018215, "grad_norm": 1.4832836389541626, "learning_rate": 8.019730161032933e-06, "loss": 0.0351, "step": 16616 }, { "epoch": 1.801496097137901, "grad_norm": 0.03842972591519356, "learning_rate": 8.019004787465545e-06, "loss": 0.0003, "step": 16617 }, { "epoch": 1.801604509973981, "grad_norm": 5.052258491516113, "learning_rate": 8.018279413898159e-06, "loss": 0.0595, "step": 16618 }, { "epoch": 1.8017129228100606, "grad_norm": 1.3142729997634888, "learning_rate": 8.017554040330771e-06, "loss": 0.0293, "step": 16619 }, { "epoch": 1.8018213356461406, "grad_norm": 0.30306968092918396, "learning_rate": 8.016828666763385e-06, "loss": 0.0064, "step": 16620 }, { "epoch": 1.8019297484822203, "grad_norm": 1.3465094566345215, "learning_rate": 8.016103293195997e-06, "loss": 0.0162, "step": 16621 }, { "epoch": 1.8020381613183, "grad_norm": 0.6445974707603455, "learning_rate": 8.015377919628609e-06, "loss": 0.0088, "step": 16622 }, { "epoch": 1.8021465741543798, "grad_norm": 0.6750607490539551, "learning_rate": 8.014652546061223e-06, "loss": 0.0154, "step": 16623 }, { "epoch": 1.8022549869904596, "grad_norm": 1.827782392501831, "learning_rate": 8.013927172493835e-06, "loss": 0.0404, "step": 16624 }, { "epoch": 1.8023633998265396, "grad_norm": 1.0718300342559814, "learning_rate": 8.013201798926447e-06, "loss": 0.0156, "step": 16625 }, { "epoch": 1.8024718126626191, "grad_norm": 0.13960124552249908, "learning_rate": 8.012476425359061e-06, "loss": 0.0012, "step": 16626 }, { "epoch": 1.8025802254986991, "grad_norm": 1.6179757118225098, "learning_rate": 8.011751051791673e-06, "loss": 0.0109, "step": 16627 }, { "epoch": 1.8026886383347789, "grad_norm": 0.9174948930740356, "learning_rate": 8.011025678224285e-06, "loss": 0.016, "step": 16628 }, { "epoch": 1.8027970511708586, "grad_norm": 2.1129531860351562, "learning_rate": 8.010300304656899e-06, "loss": 0.0132, "step": 16629 }, { "epoch": 1.8029054640069384, "grad_norm": 4.694336891174316, "learning_rate": 8.009574931089513e-06, "loss": 0.0379, "step": 16630 }, { "epoch": 1.8030138768430182, "grad_norm": 3.8354578018188477, "learning_rate": 8.008849557522125e-06, "loss": 0.0074, "step": 16631 }, { "epoch": 1.8031222896790982, "grad_norm": 4.272554397583008, "learning_rate": 8.008124183954737e-06, "loss": 0.0715, "step": 16632 }, { "epoch": 1.8032307025151777, "grad_norm": 0.15875376760959625, "learning_rate": 8.00739881038735e-06, "loss": 0.0012, "step": 16633 }, { "epoch": 1.8033391153512577, "grad_norm": 0.7176689505577087, "learning_rate": 8.006673436819963e-06, "loss": 0.0032, "step": 16634 }, { "epoch": 1.8034475281873372, "grad_norm": 1.0271849632263184, "learning_rate": 8.005948063252575e-06, "loss": 0.0062, "step": 16635 }, { "epoch": 1.8035559410234172, "grad_norm": 1.465708613395691, "learning_rate": 8.005222689685187e-06, "loss": 0.0157, "step": 16636 }, { "epoch": 1.803664353859497, "grad_norm": 8.990524291992188, "learning_rate": 8.004497316117801e-06, "loss": 0.0383, "step": 16637 }, { "epoch": 1.8037727666955767, "grad_norm": 0.9614115357398987, "learning_rate": 8.003771942550415e-06, "loss": 0.0172, "step": 16638 }, { "epoch": 1.8038811795316565, "grad_norm": 1.2187799215316772, "learning_rate": 8.003046568983027e-06, "loss": 0.0269, "step": 16639 }, { "epoch": 1.8039895923677363, "grad_norm": 1.7839678525924683, "learning_rate": 8.00232119541564e-06, "loss": 0.0092, "step": 16640 }, { "epoch": 1.8040980052038162, "grad_norm": 0.7330020666122437, "learning_rate": 8.001595821848253e-06, "loss": 0.0099, "step": 16641 }, { "epoch": 1.8042064180398958, "grad_norm": 2.782214403152466, "learning_rate": 8.000870448280865e-06, "loss": 0.0273, "step": 16642 }, { "epoch": 1.8043148308759758, "grad_norm": 1.4249727725982666, "learning_rate": 8.000145074713478e-06, "loss": 0.005, "step": 16643 }, { "epoch": 1.8044232437120555, "grad_norm": 1.31476628780365, "learning_rate": 7.999419701146091e-06, "loss": 0.0223, "step": 16644 }, { "epoch": 1.8045316565481353, "grad_norm": 5.089469909667969, "learning_rate": 7.998694327578704e-06, "loss": 0.0199, "step": 16645 }, { "epoch": 1.804640069384215, "grad_norm": 0.5605581402778625, "learning_rate": 7.997968954011317e-06, "loss": 0.0126, "step": 16646 }, { "epoch": 1.8047484822202948, "grad_norm": 4.868312835693359, "learning_rate": 7.99724358044393e-06, "loss": 0.0335, "step": 16647 }, { "epoch": 1.8048568950563748, "grad_norm": 0.40771520137786865, "learning_rate": 7.996518206876542e-06, "loss": 0.0088, "step": 16648 }, { "epoch": 1.8049653078924544, "grad_norm": 1.8017469644546509, "learning_rate": 7.995792833309155e-06, "loss": 0.0362, "step": 16649 }, { "epoch": 1.8050737207285343, "grad_norm": 1.747198224067688, "learning_rate": 7.995067459741768e-06, "loss": 0.0236, "step": 16650 }, { "epoch": 1.805182133564614, "grad_norm": 4.075907230377197, "learning_rate": 7.99434208617438e-06, "loss": 0.0225, "step": 16651 }, { "epoch": 1.8052905464006939, "grad_norm": 1.240859031677246, "learning_rate": 7.993616712606994e-06, "loss": 0.0151, "step": 16652 }, { "epoch": 1.8053989592367736, "grad_norm": 2.30554461479187, "learning_rate": 7.992891339039606e-06, "loss": 0.0493, "step": 16653 }, { "epoch": 1.8055073720728534, "grad_norm": 4.5917487144470215, "learning_rate": 7.992165965472218e-06, "loss": 0.0478, "step": 16654 }, { "epoch": 1.8056157849089334, "grad_norm": 0.43776166439056396, "learning_rate": 7.991440591904832e-06, "loss": 0.0056, "step": 16655 }, { "epoch": 1.805724197745013, "grad_norm": 0.4706901013851166, "learning_rate": 7.990715218337446e-06, "loss": 0.0012, "step": 16656 }, { "epoch": 1.805832610581093, "grad_norm": 0.3726716935634613, "learning_rate": 7.989989844770058e-06, "loss": 0.0095, "step": 16657 }, { "epoch": 1.8059410234171724, "grad_norm": 1.074822187423706, "learning_rate": 7.98926447120267e-06, "loss": 0.0151, "step": 16658 }, { "epoch": 1.8060494362532524, "grad_norm": 3.6804165840148926, "learning_rate": 7.988539097635282e-06, "loss": 0.0114, "step": 16659 }, { "epoch": 1.8061578490893322, "grad_norm": 1.5066031217575073, "learning_rate": 7.987813724067896e-06, "loss": 0.0561, "step": 16660 }, { "epoch": 1.806266261925412, "grad_norm": 0.3270654082298279, "learning_rate": 7.987088350500508e-06, "loss": 0.0023, "step": 16661 }, { "epoch": 1.8063746747614917, "grad_norm": 2.558474540710449, "learning_rate": 7.98636297693312e-06, "loss": 0.0431, "step": 16662 }, { "epoch": 1.8064830875975715, "grad_norm": 4.999547481536865, "learning_rate": 7.985637603365734e-06, "loss": 0.0179, "step": 16663 }, { "epoch": 1.8065915004336515, "grad_norm": 1.8099523782730103, "learning_rate": 7.984912229798348e-06, "loss": 0.0217, "step": 16664 }, { "epoch": 1.806699913269731, "grad_norm": 3.1686174869537354, "learning_rate": 7.98418685623096e-06, "loss": 0.04, "step": 16665 }, { "epoch": 1.806808326105811, "grad_norm": 0.6305993795394897, "learning_rate": 7.983461482663572e-06, "loss": 0.0078, "step": 16666 }, { "epoch": 1.8069167389418908, "grad_norm": 1.6376714706420898, "learning_rate": 7.982736109096186e-06, "loss": 0.046, "step": 16667 }, { "epoch": 1.8070251517779705, "grad_norm": 0.13579730689525604, "learning_rate": 7.982010735528798e-06, "loss": 0.0008, "step": 16668 }, { "epoch": 1.8071335646140503, "grad_norm": 0.6900201439857483, "learning_rate": 7.98128536196141e-06, "loss": 0.0137, "step": 16669 }, { "epoch": 1.80724197745013, "grad_norm": 0.19594350457191467, "learning_rate": 7.980559988394024e-06, "loss": 0.0012, "step": 16670 }, { "epoch": 1.80735039028621, "grad_norm": 1.0689713954925537, "learning_rate": 7.979834614826636e-06, "loss": 0.0077, "step": 16671 }, { "epoch": 1.8074588031222896, "grad_norm": 1.2158682346343994, "learning_rate": 7.97910924125925e-06, "loss": 0.0022, "step": 16672 }, { "epoch": 1.8075672159583696, "grad_norm": 2.3454642295837402, "learning_rate": 7.978383867691862e-06, "loss": 0.036, "step": 16673 }, { "epoch": 1.807675628794449, "grad_norm": 1.0080066919326782, "learning_rate": 7.977658494124474e-06, "loss": 0.0077, "step": 16674 }, { "epoch": 1.807784041630529, "grad_norm": 1.6223913431167603, "learning_rate": 7.976933120557088e-06, "loss": 0.0132, "step": 16675 }, { "epoch": 1.8078924544666088, "grad_norm": 2.31968355178833, "learning_rate": 7.9762077469897e-06, "loss": 0.0057, "step": 16676 }, { "epoch": 1.8080008673026886, "grad_norm": 1.5268069505691528, "learning_rate": 7.975482373422312e-06, "loss": 0.0121, "step": 16677 }, { "epoch": 1.8081092801387686, "grad_norm": 5.04645299911499, "learning_rate": 7.974756999854926e-06, "loss": 0.0472, "step": 16678 }, { "epoch": 1.8082176929748481, "grad_norm": 1.2264182567596436, "learning_rate": 7.974031626287538e-06, "loss": 0.0554, "step": 16679 }, { "epoch": 1.8083261058109281, "grad_norm": 2.933915853500366, "learning_rate": 7.97330625272015e-06, "loss": 0.0617, "step": 16680 }, { "epoch": 1.8084345186470077, "grad_norm": 4.1835618019104, "learning_rate": 7.972580879152764e-06, "loss": 0.0061, "step": 16681 }, { "epoch": 1.8085429314830876, "grad_norm": 3.3683948516845703, "learning_rate": 7.971855505585378e-06, "loss": 0.0382, "step": 16682 }, { "epoch": 1.8086513443191674, "grad_norm": 2.8041484355926514, "learning_rate": 7.97113013201799e-06, "loss": 0.0135, "step": 16683 }, { "epoch": 1.8087597571552472, "grad_norm": 0.6071380376815796, "learning_rate": 7.970404758450603e-06, "loss": 0.0144, "step": 16684 }, { "epoch": 1.808868169991327, "grad_norm": 2.716142177581787, "learning_rate": 7.969679384883215e-06, "loss": 0.0366, "step": 16685 }, { "epoch": 1.8089765828274067, "grad_norm": 3.9846479892730713, "learning_rate": 7.968954011315829e-06, "loss": 0.0256, "step": 16686 }, { "epoch": 1.8090849956634867, "grad_norm": 2.776310682296753, "learning_rate": 7.96822863774844e-06, "loss": 0.0279, "step": 16687 }, { "epoch": 1.8091934084995662, "grad_norm": 0.8878976106643677, "learning_rate": 7.967503264181053e-06, "loss": 0.031, "step": 16688 }, { "epoch": 1.8093018213356462, "grad_norm": 0.5660370588302612, "learning_rate": 7.966777890613667e-06, "loss": 0.0271, "step": 16689 }, { "epoch": 1.809410234171726, "grad_norm": 0.5584537386894226, "learning_rate": 7.96605251704628e-06, "loss": 0.0056, "step": 16690 }, { "epoch": 1.8095186470078057, "grad_norm": 4.665583610534668, "learning_rate": 7.965327143478893e-06, "loss": 0.016, "step": 16691 }, { "epoch": 1.8096270598438855, "grad_norm": 0.36289849877357483, "learning_rate": 7.964601769911505e-06, "loss": 0.0077, "step": 16692 }, { "epoch": 1.8097354726799653, "grad_norm": 1.7590065002441406, "learning_rate": 7.963876396344119e-06, "loss": 0.0352, "step": 16693 }, { "epoch": 1.8098438855160452, "grad_norm": 1.8277826309204102, "learning_rate": 7.96315102277673e-06, "loss": 0.0094, "step": 16694 }, { "epoch": 1.8099522983521248, "grad_norm": 0.9037254452705383, "learning_rate": 7.962425649209343e-06, "loss": 0.0034, "step": 16695 }, { "epoch": 1.8100607111882048, "grad_norm": 0.9411906599998474, "learning_rate": 7.961700275641957e-06, "loss": 0.0221, "step": 16696 }, { "epoch": 1.8101691240242843, "grad_norm": 2.3052194118499756, "learning_rate": 7.960974902074569e-06, "loss": 0.0253, "step": 16697 }, { "epoch": 1.8102775368603643, "grad_norm": 2.9677810668945312, "learning_rate": 7.960249528507183e-06, "loss": 0.0364, "step": 16698 }, { "epoch": 1.810385949696444, "grad_norm": 0.10363168269395828, "learning_rate": 7.959524154939795e-06, "loss": 0.0012, "step": 16699 }, { "epoch": 1.8104943625325238, "grad_norm": 3.585024833679199, "learning_rate": 7.958798781372407e-06, "loss": 0.0274, "step": 16700 }, { "epoch": 1.8106027753686036, "grad_norm": 1.8330405950546265, "learning_rate": 7.95807340780502e-06, "loss": 0.0138, "step": 16701 }, { "epoch": 1.8107111882046834, "grad_norm": 0.17990249395370483, "learning_rate": 7.957348034237633e-06, "loss": 0.0011, "step": 16702 }, { "epoch": 1.8108196010407633, "grad_norm": 0.6508957147598267, "learning_rate": 7.956622660670245e-06, "loss": 0.0228, "step": 16703 }, { "epoch": 1.8109280138768429, "grad_norm": 0.3309211730957031, "learning_rate": 7.955897287102859e-06, "loss": 0.0114, "step": 16704 }, { "epoch": 1.8110364267129229, "grad_norm": 0.5005684494972229, "learning_rate": 7.955171913535471e-06, "loss": 0.0149, "step": 16705 }, { "epoch": 1.8111448395490026, "grad_norm": 2.8861730098724365, "learning_rate": 7.954446539968083e-06, "loss": 0.0179, "step": 16706 }, { "epoch": 1.8112532523850824, "grad_norm": 1.8619199991226196, "learning_rate": 7.953721166400697e-06, "loss": 0.0256, "step": 16707 }, { "epoch": 1.8113616652211622, "grad_norm": 0.4012764096260071, "learning_rate": 7.952995792833311e-06, "loss": 0.0129, "step": 16708 }, { "epoch": 1.811470078057242, "grad_norm": 3.383150339126587, "learning_rate": 7.952270419265923e-06, "loss": 0.0278, "step": 16709 }, { "epoch": 1.811578490893322, "grad_norm": 6.604656219482422, "learning_rate": 7.951545045698535e-06, "loss": 0.0643, "step": 16710 }, { "epoch": 1.8116869037294014, "grad_norm": 9.83855152130127, "learning_rate": 7.950819672131147e-06, "loss": 0.0159, "step": 16711 }, { "epoch": 1.8117953165654814, "grad_norm": 3.594874382019043, "learning_rate": 7.950094298563761e-06, "loss": 0.0712, "step": 16712 }, { "epoch": 1.8119037294015612, "grad_norm": 1.3857005834579468, "learning_rate": 7.949368924996373e-06, "loss": 0.0095, "step": 16713 }, { "epoch": 1.812012142237641, "grad_norm": 1.6918540000915527, "learning_rate": 7.948643551428985e-06, "loss": 0.0616, "step": 16714 }, { "epoch": 1.8121205550737207, "grad_norm": 3.240785598754883, "learning_rate": 7.9479181778616e-06, "loss": 0.0112, "step": 16715 }, { "epoch": 1.8122289679098005, "grad_norm": 0.4314250349998474, "learning_rate": 7.947192804294213e-06, "loss": 0.0166, "step": 16716 }, { "epoch": 1.8123373807458805, "grad_norm": 3.69534969329834, "learning_rate": 7.946467430726825e-06, "loss": 0.0639, "step": 16717 }, { "epoch": 1.81244579358196, "grad_norm": 6.954265117645264, "learning_rate": 7.945742057159437e-06, "loss": 0.0485, "step": 16718 }, { "epoch": 1.81255420641804, "grad_norm": 2.8993656635284424, "learning_rate": 7.945016683592051e-06, "loss": 0.0497, "step": 16719 }, { "epoch": 1.8126626192541195, "grad_norm": 0.2910110652446747, "learning_rate": 7.944291310024663e-06, "loss": 0.0076, "step": 16720 }, { "epoch": 1.8127710320901995, "grad_norm": 0.2621178925037384, "learning_rate": 7.943565936457276e-06, "loss": 0.0058, "step": 16721 }, { "epoch": 1.8128794449262793, "grad_norm": 0.37332990765571594, "learning_rate": 7.94284056288989e-06, "loss": 0.0225, "step": 16722 }, { "epoch": 1.812987857762359, "grad_norm": 1.4207172393798828, "learning_rate": 7.942115189322502e-06, "loss": 0.0077, "step": 16723 }, { "epoch": 1.8130962705984388, "grad_norm": 0.35578852891921997, "learning_rate": 7.941389815755115e-06, "loss": 0.0124, "step": 16724 }, { "epoch": 1.8132046834345186, "grad_norm": 1.7145054340362549, "learning_rate": 7.940664442187728e-06, "loss": 0.0066, "step": 16725 }, { "epoch": 1.8133130962705986, "grad_norm": 0.34663689136505127, "learning_rate": 7.93993906862034e-06, "loss": 0.0017, "step": 16726 }, { "epoch": 1.813421509106678, "grad_norm": 3.6297786235809326, "learning_rate": 7.939213695052953e-06, "loss": 0.0333, "step": 16727 }, { "epoch": 1.813529921942758, "grad_norm": 2.9052531719207764, "learning_rate": 7.938488321485566e-06, "loss": 0.0482, "step": 16728 }, { "epoch": 1.8136383347788378, "grad_norm": 2.0948801040649414, "learning_rate": 7.937762947918178e-06, "loss": 0.0071, "step": 16729 }, { "epoch": 1.8137467476149176, "grad_norm": 0.5358269214630127, "learning_rate": 7.937037574350792e-06, "loss": 0.0049, "step": 16730 }, { "epoch": 1.8138551604509974, "grad_norm": 1.3138834238052368, "learning_rate": 7.936312200783404e-06, "loss": 0.0028, "step": 16731 }, { "epoch": 1.8139635732870771, "grad_norm": 2.293509006500244, "learning_rate": 7.935586827216016e-06, "loss": 0.0039, "step": 16732 }, { "epoch": 1.8140719861231571, "grad_norm": 1.7586313486099243, "learning_rate": 7.93486145364863e-06, "loss": 0.0177, "step": 16733 }, { "epoch": 1.8141803989592367, "grad_norm": 0.5393751859664917, "learning_rate": 7.934136080081244e-06, "loss": 0.0224, "step": 16734 }, { "epoch": 1.8142888117953166, "grad_norm": 0.4429246187210083, "learning_rate": 7.933410706513856e-06, "loss": 0.0094, "step": 16735 }, { "epoch": 1.8143972246313964, "grad_norm": 5.499034404754639, "learning_rate": 7.932685332946468e-06, "loss": 0.0513, "step": 16736 }, { "epoch": 1.8145056374674762, "grad_norm": 3.390540838241577, "learning_rate": 7.93195995937908e-06, "loss": 0.0198, "step": 16737 }, { "epoch": 1.814614050303556, "grad_norm": 6.967928409576416, "learning_rate": 7.931234585811694e-06, "loss": 0.0224, "step": 16738 }, { "epoch": 1.8147224631396357, "grad_norm": 3.699517250061035, "learning_rate": 7.930509212244306e-06, "loss": 0.0294, "step": 16739 }, { "epoch": 1.8148308759757157, "grad_norm": 1.8267362117767334, "learning_rate": 7.929783838676918e-06, "loss": 0.0356, "step": 16740 }, { "epoch": 1.8149392888117952, "grad_norm": 7.875589847564697, "learning_rate": 7.929058465109532e-06, "loss": 0.0655, "step": 16741 }, { "epoch": 1.8150477016478752, "grad_norm": 1.5123642683029175, "learning_rate": 7.928333091542146e-06, "loss": 0.0289, "step": 16742 }, { "epoch": 1.8151561144839548, "grad_norm": 0.538303554058075, "learning_rate": 7.927607717974758e-06, "loss": 0.0035, "step": 16743 }, { "epoch": 1.8152645273200347, "grad_norm": 5.265900135040283, "learning_rate": 7.92688234440737e-06, "loss": 0.0444, "step": 16744 }, { "epoch": 1.8153729401561145, "grad_norm": 6.85777473449707, "learning_rate": 7.926156970839984e-06, "loss": 0.0135, "step": 16745 }, { "epoch": 1.8154813529921943, "grad_norm": 4.439815521240234, "learning_rate": 7.925431597272596e-06, "loss": 0.0467, "step": 16746 }, { "epoch": 1.815589765828274, "grad_norm": 0.8867642283439636, "learning_rate": 7.924706223705208e-06, "loss": 0.0023, "step": 16747 }, { "epoch": 1.8156981786643538, "grad_norm": 7.628986358642578, "learning_rate": 7.923980850137822e-06, "loss": 0.0301, "step": 16748 }, { "epoch": 1.8158065915004338, "grad_norm": 4.253026008605957, "learning_rate": 7.923255476570434e-06, "loss": 0.0205, "step": 16749 }, { "epoch": 1.8159150043365133, "grad_norm": 0.14077533781528473, "learning_rate": 7.922530103003048e-06, "loss": 0.0017, "step": 16750 }, { "epoch": 1.8160234171725933, "grad_norm": 2.5826504230499268, "learning_rate": 7.92180472943566e-06, "loss": 0.0357, "step": 16751 }, { "epoch": 1.816131830008673, "grad_norm": 0.11378385126590729, "learning_rate": 7.921079355868272e-06, "loss": 0.0009, "step": 16752 }, { "epoch": 1.8162402428447528, "grad_norm": 2.726426601409912, "learning_rate": 7.920353982300886e-06, "loss": 0.0194, "step": 16753 }, { "epoch": 1.8163486556808326, "grad_norm": 0.7242252826690674, "learning_rate": 7.919628608733498e-06, "loss": 0.0022, "step": 16754 }, { "epoch": 1.8164570685169124, "grad_norm": 1.4258744716644287, "learning_rate": 7.91890323516611e-06, "loss": 0.0262, "step": 16755 }, { "epoch": 1.8165654813529923, "grad_norm": 0.7245040535926819, "learning_rate": 7.918177861598724e-06, "loss": 0.0147, "step": 16756 }, { "epoch": 1.8166738941890719, "grad_norm": 12.164769172668457, "learning_rate": 7.917452488031336e-06, "loss": 0.0499, "step": 16757 }, { "epoch": 1.8167823070251519, "grad_norm": 1.8236571550369263, "learning_rate": 7.916727114463949e-06, "loss": 0.0388, "step": 16758 }, { "epoch": 1.8168907198612314, "grad_norm": 1.0794765949249268, "learning_rate": 7.916001740896562e-06, "loss": 0.0074, "step": 16759 }, { "epoch": 1.8169991326973114, "grad_norm": 0.31221264600753784, "learning_rate": 7.915276367329176e-06, "loss": 0.007, "step": 16760 }, { "epoch": 1.8171075455333912, "grad_norm": 0.8706511855125427, "learning_rate": 7.914550993761788e-06, "loss": 0.0161, "step": 16761 }, { "epoch": 1.817215958369471, "grad_norm": 0.5483095049858093, "learning_rate": 7.9138256201944e-06, "loss": 0.0272, "step": 16762 }, { "epoch": 1.817324371205551, "grad_norm": 0.3554437756538391, "learning_rate": 7.913100246627013e-06, "loss": 0.0024, "step": 16763 }, { "epoch": 1.8174327840416304, "grad_norm": 0.7349015474319458, "learning_rate": 7.912374873059627e-06, "loss": 0.0169, "step": 16764 }, { "epoch": 1.8175411968777104, "grad_norm": 1.3153852224349976, "learning_rate": 7.911649499492239e-06, "loss": 0.025, "step": 16765 }, { "epoch": 1.81764960971379, "grad_norm": 2.684025287628174, "learning_rate": 7.91092412592485e-06, "loss": 0.0131, "step": 16766 }, { "epoch": 1.81775802254987, "grad_norm": 0.4402956962585449, "learning_rate": 7.910198752357465e-06, "loss": 0.0133, "step": 16767 }, { "epoch": 1.8178664353859497, "grad_norm": 0.6452608108520508, "learning_rate": 7.909473378790078e-06, "loss": 0.0043, "step": 16768 }, { "epoch": 1.8179748482220295, "grad_norm": 2.6695289611816406, "learning_rate": 7.90874800522269e-06, "loss": 0.0276, "step": 16769 }, { "epoch": 1.8180832610581092, "grad_norm": 3.4263200759887695, "learning_rate": 7.908022631655303e-06, "loss": 0.0115, "step": 16770 }, { "epoch": 1.818191673894189, "grad_norm": 0.3758145868778229, "learning_rate": 7.907297258087917e-06, "loss": 0.0088, "step": 16771 }, { "epoch": 1.818300086730269, "grad_norm": 1.0836198329925537, "learning_rate": 7.906571884520529e-06, "loss": 0.0543, "step": 16772 }, { "epoch": 1.8184084995663485, "grad_norm": 1.154797077178955, "learning_rate": 7.905846510953141e-06, "loss": 0.013, "step": 16773 }, { "epoch": 1.8185169124024285, "grad_norm": 0.04709064960479736, "learning_rate": 7.905121137385755e-06, "loss": 0.0005, "step": 16774 }, { "epoch": 1.8186253252385083, "grad_norm": 4.971432685852051, "learning_rate": 7.904395763818367e-06, "loss": 0.0191, "step": 16775 }, { "epoch": 1.818733738074588, "grad_norm": 4.297083854675293, "learning_rate": 7.90367039025098e-06, "loss": 0.0135, "step": 16776 }, { "epoch": 1.8188421509106678, "grad_norm": 0.41495683789253235, "learning_rate": 7.902945016683593e-06, "loss": 0.0091, "step": 16777 }, { "epoch": 1.8189505637467476, "grad_norm": 3.236931085586548, "learning_rate": 7.902219643116205e-06, "loss": 0.0385, "step": 16778 }, { "epoch": 1.8190589765828276, "grad_norm": 0.4433155655860901, "learning_rate": 7.901494269548819e-06, "loss": 0.0155, "step": 16779 }, { "epoch": 1.819167389418907, "grad_norm": 0.5662212371826172, "learning_rate": 7.900768895981431e-06, "loss": 0.0169, "step": 16780 }, { "epoch": 1.819275802254987, "grad_norm": 7.314748287200928, "learning_rate": 7.900043522414043e-06, "loss": 0.0216, "step": 16781 }, { "epoch": 1.8193842150910666, "grad_norm": 1.443576455116272, "learning_rate": 7.899318148846657e-06, "loss": 0.0336, "step": 16782 }, { "epoch": 1.8194926279271466, "grad_norm": 5.789901256561279, "learning_rate": 7.898592775279269e-06, "loss": 0.0278, "step": 16783 }, { "epoch": 1.8196010407632264, "grad_norm": 7.238768100738525, "learning_rate": 7.897867401711883e-06, "loss": 0.0217, "step": 16784 }, { "epoch": 1.8197094535993061, "grad_norm": 1.1914341449737549, "learning_rate": 7.897142028144495e-06, "loss": 0.008, "step": 16785 }, { "epoch": 1.819817866435386, "grad_norm": 0.38505300879478455, "learning_rate": 7.896416654577109e-06, "loss": 0.0018, "step": 16786 }, { "epoch": 1.8199262792714657, "grad_norm": 13.677854537963867, "learning_rate": 7.895691281009721e-06, "loss": 0.0599, "step": 16787 }, { "epoch": 1.8200346921075456, "grad_norm": 3.0970284938812256, "learning_rate": 7.894965907442333e-06, "loss": 0.046, "step": 16788 }, { "epoch": 1.8201431049436252, "grad_norm": 2.099757671356201, "learning_rate": 7.894240533874945e-06, "loss": 0.0149, "step": 16789 }, { "epoch": 1.8202515177797052, "grad_norm": 0.14975588023662567, "learning_rate": 7.89351516030756e-06, "loss": 0.0016, "step": 16790 }, { "epoch": 1.820359930615785, "grad_norm": 3.277264356613159, "learning_rate": 7.892789786740171e-06, "loss": 0.0422, "step": 16791 }, { "epoch": 1.8204683434518647, "grad_norm": 6.2005510330200195, "learning_rate": 7.892064413172783e-06, "loss": 0.035, "step": 16792 }, { "epoch": 1.8205767562879445, "grad_norm": 4.294259548187256, "learning_rate": 7.891339039605397e-06, "loss": 0.0188, "step": 16793 }, { "epoch": 1.8206851691240242, "grad_norm": 4.560187816619873, "learning_rate": 7.890613666038011e-06, "loss": 0.0338, "step": 16794 }, { "epoch": 1.8207935819601042, "grad_norm": 2.2278895378112793, "learning_rate": 7.889888292470623e-06, "loss": 0.0246, "step": 16795 }, { "epoch": 1.8209019947961838, "grad_norm": 3.4093518257141113, "learning_rate": 7.889162918903235e-06, "loss": 0.0075, "step": 16796 }, { "epoch": 1.8210104076322637, "grad_norm": 2.4091873168945312, "learning_rate": 7.88843754533585e-06, "loss": 0.0183, "step": 16797 }, { "epoch": 1.8211188204683435, "grad_norm": 0.4109621047973633, "learning_rate": 7.887712171768461e-06, "loss": 0.0167, "step": 16798 }, { "epoch": 1.8212272333044233, "grad_norm": 3.600771188735962, "learning_rate": 7.886986798201074e-06, "loss": 0.0286, "step": 16799 }, { "epoch": 1.821335646140503, "grad_norm": 0.36364802718162537, "learning_rate": 7.886261424633687e-06, "loss": 0.003, "step": 16800 }, { "epoch": 1.8214440589765828, "grad_norm": 1.0835685729980469, "learning_rate": 7.8855360510663e-06, "loss": 0.0631, "step": 16801 }, { "epoch": 1.8215524718126628, "grad_norm": 0.449999064207077, "learning_rate": 7.884810677498913e-06, "loss": 0.0168, "step": 16802 }, { "epoch": 1.8216608846487423, "grad_norm": 3.50886607170105, "learning_rate": 7.884085303931526e-06, "loss": 0.0458, "step": 16803 }, { "epoch": 1.8217692974848223, "grad_norm": 3.3045787811279297, "learning_rate": 7.883359930364138e-06, "loss": 0.008, "step": 16804 }, { "epoch": 1.8218777103209018, "grad_norm": 0.5939558744430542, "learning_rate": 7.882634556796752e-06, "loss": 0.0116, "step": 16805 }, { "epoch": 1.8219861231569818, "grad_norm": 3.6118319034576416, "learning_rate": 7.881909183229364e-06, "loss": 0.0336, "step": 16806 }, { "epoch": 1.8220945359930616, "grad_norm": 4.925454139709473, "learning_rate": 7.881183809661976e-06, "loss": 0.0095, "step": 16807 }, { "epoch": 1.8222029488291414, "grad_norm": 2.014768600463867, "learning_rate": 7.88045843609459e-06, "loss": 0.051, "step": 16808 }, { "epoch": 1.8223113616652211, "grad_norm": 0.3774493336677551, "learning_rate": 7.879733062527202e-06, "loss": 0.0055, "step": 16809 }, { "epoch": 1.8224197745013009, "grad_norm": 1.895029902458191, "learning_rate": 7.879007688959816e-06, "loss": 0.0408, "step": 16810 }, { "epoch": 1.8225281873373809, "grad_norm": 3.2825913429260254, "learning_rate": 7.878282315392428e-06, "loss": 0.0211, "step": 16811 }, { "epoch": 1.8226366001734604, "grad_norm": 0.28079336881637573, "learning_rate": 7.877556941825042e-06, "loss": 0.0062, "step": 16812 }, { "epoch": 1.8227450130095404, "grad_norm": 0.4381869435310364, "learning_rate": 7.876831568257654e-06, "loss": 0.0064, "step": 16813 }, { "epoch": 1.8228534258456202, "grad_norm": 2.336719274520874, "learning_rate": 7.876106194690266e-06, "loss": 0.0434, "step": 16814 }, { "epoch": 1.8229618386817, "grad_norm": 2.470935583114624, "learning_rate": 7.875380821122878e-06, "loss": 0.0172, "step": 16815 }, { "epoch": 1.8230702515177797, "grad_norm": 0.7695043087005615, "learning_rate": 7.874655447555492e-06, "loss": 0.0175, "step": 16816 }, { "epoch": 1.8231786643538594, "grad_norm": 4.859504699707031, "learning_rate": 7.873930073988104e-06, "loss": 0.0369, "step": 16817 }, { "epoch": 1.8232870771899394, "grad_norm": 0.10126293450593948, "learning_rate": 7.873204700420716e-06, "loss": 0.001, "step": 16818 }, { "epoch": 1.823395490026019, "grad_norm": 2.6989400386810303, "learning_rate": 7.87247932685333e-06, "loss": 0.0437, "step": 16819 }, { "epoch": 1.823503902862099, "grad_norm": 0.6799686551094055, "learning_rate": 7.871753953285944e-06, "loss": 0.0025, "step": 16820 }, { "epoch": 1.8236123156981785, "grad_norm": 0.10532412678003311, "learning_rate": 7.871028579718556e-06, "loss": 0.0009, "step": 16821 }, { "epoch": 1.8237207285342585, "grad_norm": 4.60087251663208, "learning_rate": 7.870303206151168e-06, "loss": 0.0476, "step": 16822 }, { "epoch": 1.8238291413703382, "grad_norm": 0.4411631226539612, "learning_rate": 7.869577832583782e-06, "loss": 0.0098, "step": 16823 }, { "epoch": 1.823937554206418, "grad_norm": 0.6347993612289429, "learning_rate": 7.868852459016394e-06, "loss": 0.0036, "step": 16824 }, { "epoch": 1.824045967042498, "grad_norm": 3.336348533630371, "learning_rate": 7.868127085449006e-06, "loss": 0.0465, "step": 16825 }, { "epoch": 1.8241543798785775, "grad_norm": 1.066330909729004, "learning_rate": 7.86740171188162e-06, "loss": 0.0395, "step": 16826 }, { "epoch": 1.8242627927146575, "grad_norm": 2.9295566082000732, "learning_rate": 7.866676338314232e-06, "loss": 0.0373, "step": 16827 }, { "epoch": 1.824371205550737, "grad_norm": 5.870022773742676, "learning_rate": 7.865950964746846e-06, "loss": 0.0341, "step": 16828 }, { "epoch": 1.824479618386817, "grad_norm": 0.3243667781352997, "learning_rate": 7.865225591179458e-06, "loss": 0.002, "step": 16829 }, { "epoch": 1.8245880312228968, "grad_norm": 3.1209611892700195, "learning_rate": 7.86450021761207e-06, "loss": 0.0058, "step": 16830 }, { "epoch": 1.8246964440589766, "grad_norm": 0.44997236132621765, "learning_rate": 7.863774844044684e-06, "loss": 0.0162, "step": 16831 }, { "epoch": 1.8248048568950563, "grad_norm": 4.023565292358398, "learning_rate": 7.863049470477296e-06, "loss": 0.0147, "step": 16832 }, { "epoch": 1.824913269731136, "grad_norm": 0.44098541140556335, "learning_rate": 7.862324096909908e-06, "loss": 0.0021, "step": 16833 }, { "epoch": 1.825021682567216, "grad_norm": 5.380909442901611, "learning_rate": 7.861598723342522e-06, "loss": 0.0236, "step": 16834 }, { "epoch": 1.8251300954032956, "grad_norm": 1.2709153890609741, "learning_rate": 7.860873349775134e-06, "loss": 0.0046, "step": 16835 }, { "epoch": 1.8252385082393756, "grad_norm": 2.4944632053375244, "learning_rate": 7.860147976207748e-06, "loss": 0.0228, "step": 16836 }, { "epoch": 1.8253469210754554, "grad_norm": 3.7601842880249023, "learning_rate": 7.85942260264036e-06, "loss": 0.0158, "step": 16837 }, { "epoch": 1.8254553339115351, "grad_norm": 3.7145822048187256, "learning_rate": 7.858697229072974e-06, "loss": 0.0223, "step": 16838 }, { "epoch": 1.825563746747615, "grad_norm": 0.1313195675611496, "learning_rate": 7.857971855505586e-06, "loss": 0.0015, "step": 16839 }, { "epoch": 1.8256721595836947, "grad_norm": 0.554865837097168, "learning_rate": 7.857246481938199e-06, "loss": 0.0078, "step": 16840 }, { "epoch": 1.8257805724197746, "grad_norm": 2.4996683597564697, "learning_rate": 7.85652110837081e-06, "loss": 0.0116, "step": 16841 }, { "epoch": 1.8258889852558542, "grad_norm": 1.3159326314926147, "learning_rate": 7.855795734803425e-06, "loss": 0.0181, "step": 16842 }, { "epoch": 1.8259973980919342, "grad_norm": 0.14429399371147156, "learning_rate": 7.855070361236037e-06, "loss": 0.0016, "step": 16843 }, { "epoch": 1.8261058109280137, "grad_norm": 0.6371038556098938, "learning_rate": 7.854344987668649e-06, "loss": 0.0115, "step": 16844 }, { "epoch": 1.8262142237640937, "grad_norm": 4.265449047088623, "learning_rate": 7.853619614101263e-06, "loss": 0.0476, "step": 16845 }, { "epoch": 1.8263226366001735, "grad_norm": 0.9923971891403198, "learning_rate": 7.852894240533876e-06, "loss": 0.0051, "step": 16846 }, { "epoch": 1.8264310494362532, "grad_norm": 0.30329737067222595, "learning_rate": 7.852168866966489e-06, "loss": 0.0107, "step": 16847 }, { "epoch": 1.826539462272333, "grad_norm": 5.058859348297119, "learning_rate": 7.8514434933991e-06, "loss": 0.0176, "step": 16848 }, { "epoch": 1.8266478751084128, "grad_norm": 2.6866822242736816, "learning_rate": 7.850718119831715e-06, "loss": 0.025, "step": 16849 }, { "epoch": 1.8267562879444927, "grad_norm": 3.9834742546081543, "learning_rate": 7.849992746264327e-06, "loss": 0.0531, "step": 16850 }, { "epoch": 1.8268647007805723, "grad_norm": 0.2824135720729828, "learning_rate": 7.849267372696939e-06, "loss": 0.0048, "step": 16851 }, { "epoch": 1.8269731136166523, "grad_norm": 2.3472940921783447, "learning_rate": 7.848541999129553e-06, "loss": 0.0282, "step": 16852 }, { "epoch": 1.827081526452732, "grad_norm": 4.210781097412109, "learning_rate": 7.847816625562165e-06, "loss": 0.015, "step": 16853 }, { "epoch": 1.8271899392888118, "grad_norm": 4.555943965911865, "learning_rate": 7.847091251994779e-06, "loss": 0.0608, "step": 16854 }, { "epoch": 1.8272983521248916, "grad_norm": 5.286530017852783, "learning_rate": 7.846365878427391e-06, "loss": 0.0218, "step": 16855 }, { "epoch": 1.8274067649609713, "grad_norm": 2.7776286602020264, "learning_rate": 7.845640504860003e-06, "loss": 0.0181, "step": 16856 }, { "epoch": 1.8275151777970513, "grad_norm": 2.6066172122955322, "learning_rate": 7.844915131292617e-06, "loss": 0.0075, "step": 16857 }, { "epoch": 1.8276235906331308, "grad_norm": 5.560667037963867, "learning_rate": 7.844189757725229e-06, "loss": 0.0188, "step": 16858 }, { "epoch": 1.8277320034692108, "grad_norm": 5.703892707824707, "learning_rate": 7.843464384157841e-06, "loss": 0.0331, "step": 16859 }, { "epoch": 1.8278404163052906, "grad_norm": 4.90786075592041, "learning_rate": 7.842739010590455e-06, "loss": 0.0399, "step": 16860 }, { "epoch": 1.8279488291413704, "grad_norm": 2.3557217121124268, "learning_rate": 7.842013637023067e-06, "loss": 0.0268, "step": 16861 }, { "epoch": 1.8280572419774501, "grad_norm": 3.6423768997192383, "learning_rate": 7.841288263455681e-06, "loss": 0.0115, "step": 16862 }, { "epoch": 1.8281656548135299, "grad_norm": 1.7967299222946167, "learning_rate": 7.840562889888293e-06, "loss": 0.0081, "step": 16863 }, { "epoch": 1.8282740676496099, "grad_norm": 5.341029167175293, "learning_rate": 7.839837516320907e-06, "loss": 0.0125, "step": 16864 }, { "epoch": 1.8283824804856894, "grad_norm": 7.766292572021484, "learning_rate": 7.839112142753519e-06, "loss": 0.0481, "step": 16865 }, { "epoch": 1.8284908933217694, "grad_norm": 1.4116878509521484, "learning_rate": 7.838386769186131e-06, "loss": 0.0088, "step": 16866 }, { "epoch": 1.828599306157849, "grad_norm": 4.149320602416992, "learning_rate": 7.837661395618743e-06, "loss": 0.0316, "step": 16867 }, { "epoch": 1.828707718993929, "grad_norm": 0.48480087518692017, "learning_rate": 7.836936022051357e-06, "loss": 0.0074, "step": 16868 }, { "epoch": 1.8288161318300087, "grad_norm": 0.828852653503418, "learning_rate": 7.83621064848397e-06, "loss": 0.0107, "step": 16869 }, { "epoch": 1.8289245446660884, "grad_norm": 0.4173383116722107, "learning_rate": 7.835485274916581e-06, "loss": 0.0131, "step": 16870 }, { "epoch": 1.8290329575021682, "grad_norm": 4.690745830535889, "learning_rate": 7.834759901349195e-06, "loss": 0.0169, "step": 16871 }, { "epoch": 1.829141370338248, "grad_norm": 0.3292469382286072, "learning_rate": 7.83403452778181e-06, "loss": 0.0069, "step": 16872 }, { "epoch": 1.829249783174328, "grad_norm": 0.6475711464881897, "learning_rate": 7.833309154214421e-06, "loss": 0.0229, "step": 16873 }, { "epoch": 1.8293581960104075, "grad_norm": 0.9238314628601074, "learning_rate": 7.832583780647033e-06, "loss": 0.0028, "step": 16874 }, { "epoch": 1.8294666088464875, "grad_norm": 0.29633477330207825, "learning_rate": 7.831858407079647e-06, "loss": 0.0057, "step": 16875 }, { "epoch": 1.8295750216825672, "grad_norm": 0.6146421432495117, "learning_rate": 7.83113303351226e-06, "loss": 0.0044, "step": 16876 }, { "epoch": 1.829683434518647, "grad_norm": 0.03981032967567444, "learning_rate": 7.830407659944872e-06, "loss": 0.0004, "step": 16877 }, { "epoch": 1.8297918473547268, "grad_norm": 2.4258859157562256, "learning_rate": 7.829682286377485e-06, "loss": 0.0157, "step": 16878 }, { "epoch": 1.8299002601908065, "grad_norm": 9.377248764038086, "learning_rate": 7.828956912810098e-06, "loss": 0.045, "step": 16879 }, { "epoch": 1.8300086730268865, "grad_norm": 3.6966819763183594, "learning_rate": 7.828231539242711e-06, "loss": 0.088, "step": 16880 }, { "epoch": 1.830117085862966, "grad_norm": 4.502748489379883, "learning_rate": 7.827506165675324e-06, "loss": 0.0137, "step": 16881 }, { "epoch": 1.830225498699046, "grad_norm": 2.106448173522949, "learning_rate": 7.826780792107936e-06, "loss": 0.0231, "step": 16882 }, { "epoch": 1.8303339115351258, "grad_norm": 0.6708443760871887, "learning_rate": 7.82605541854055e-06, "loss": 0.0148, "step": 16883 }, { "epoch": 1.8304423243712056, "grad_norm": 0.03392064571380615, "learning_rate": 7.825330044973162e-06, "loss": 0.0003, "step": 16884 }, { "epoch": 1.8305507372072853, "grad_norm": 3.2154340744018555, "learning_rate": 7.824604671405774e-06, "loss": 0.0174, "step": 16885 }, { "epoch": 1.830659150043365, "grad_norm": 4.8164544105529785, "learning_rate": 7.823879297838388e-06, "loss": 0.0349, "step": 16886 }, { "epoch": 1.830767562879445, "grad_norm": 2.915160655975342, "learning_rate": 7.823153924271e-06, "loss": 0.0161, "step": 16887 }, { "epoch": 1.8308759757155246, "grad_norm": 0.9387792944908142, "learning_rate": 7.822428550703614e-06, "loss": 0.0169, "step": 16888 }, { "epoch": 1.8309843885516046, "grad_norm": 0.0710882842540741, "learning_rate": 7.821703177136226e-06, "loss": 0.0005, "step": 16889 }, { "epoch": 1.8310928013876842, "grad_norm": 0.37962475419044495, "learning_rate": 7.82097780356884e-06, "loss": 0.0055, "step": 16890 }, { "epoch": 1.8312012142237641, "grad_norm": 0.4428533613681793, "learning_rate": 7.820252430001452e-06, "loss": 0.0241, "step": 16891 }, { "epoch": 1.831309627059844, "grad_norm": 0.893385648727417, "learning_rate": 7.819527056434064e-06, "loss": 0.0094, "step": 16892 }, { "epoch": 1.8314180398959237, "grad_norm": 0.4243585169315338, "learning_rate": 7.818801682866676e-06, "loss": 0.0172, "step": 16893 }, { "epoch": 1.8315264527320034, "grad_norm": 3.8028922080993652, "learning_rate": 7.81807630929929e-06, "loss": 0.044, "step": 16894 }, { "epoch": 1.8316348655680832, "grad_norm": 4.3328142166137695, "learning_rate": 7.817350935731902e-06, "loss": 0.0452, "step": 16895 }, { "epoch": 1.8317432784041632, "grad_norm": 4.092642784118652, "learning_rate": 7.816625562164514e-06, "loss": 0.0232, "step": 16896 }, { "epoch": 1.8318516912402427, "grad_norm": 1.2772908210754395, "learning_rate": 7.815900188597128e-06, "loss": 0.0193, "step": 16897 }, { "epoch": 1.8319601040763227, "grad_norm": 3.7007369995117188, "learning_rate": 7.815174815029742e-06, "loss": 0.0476, "step": 16898 }, { "epoch": 1.8320685169124025, "grad_norm": 3.6574504375457764, "learning_rate": 7.814449441462354e-06, "loss": 0.0479, "step": 16899 }, { "epoch": 1.8321769297484822, "grad_norm": 0.7038118839263916, "learning_rate": 7.813724067894966e-06, "loss": 0.0024, "step": 16900 }, { "epoch": 1.832285342584562, "grad_norm": 0.7773370742797852, "learning_rate": 7.81299869432758e-06, "loss": 0.0217, "step": 16901 }, { "epoch": 1.8323937554206418, "grad_norm": 1.4946969747543335, "learning_rate": 7.812273320760192e-06, "loss": 0.0173, "step": 16902 }, { "epoch": 1.8325021682567217, "grad_norm": 2.4221043586730957, "learning_rate": 7.811547947192804e-06, "loss": 0.0435, "step": 16903 }, { "epoch": 1.8326105810928013, "grad_norm": 5.4116530418396, "learning_rate": 7.810822573625418e-06, "loss": 0.0197, "step": 16904 }, { "epoch": 1.8327189939288813, "grad_norm": 2.6639890670776367, "learning_rate": 7.810097200058032e-06, "loss": 0.0204, "step": 16905 }, { "epoch": 1.8328274067649608, "grad_norm": 3.0183160305023193, "learning_rate": 7.809371826490644e-06, "loss": 0.0287, "step": 16906 }, { "epoch": 1.8329358196010408, "grad_norm": 0.5500087141990662, "learning_rate": 7.808646452923256e-06, "loss": 0.0135, "step": 16907 }, { "epoch": 1.8330442324371206, "grad_norm": 1.1927611827850342, "learning_rate": 7.807921079355868e-06, "loss": 0.0282, "step": 16908 }, { "epoch": 1.8331526452732003, "grad_norm": 1.1789661645889282, "learning_rate": 7.807195705788482e-06, "loss": 0.0043, "step": 16909 }, { "epoch": 1.83326105810928, "grad_norm": 4.598733425140381, "learning_rate": 7.806470332221094e-06, "loss": 0.0307, "step": 16910 }, { "epoch": 1.8333694709453598, "grad_norm": 1.1619740724563599, "learning_rate": 7.805744958653706e-06, "loss": 0.0051, "step": 16911 }, { "epoch": 1.8334778837814398, "grad_norm": 0.33515146374702454, "learning_rate": 7.80501958508632e-06, "loss": 0.0087, "step": 16912 }, { "epoch": 1.8335862966175194, "grad_norm": 2.939531087875366, "learning_rate": 7.804294211518932e-06, "loss": 0.0467, "step": 16913 }, { "epoch": 1.8336947094535994, "grad_norm": 2.902137041091919, "learning_rate": 7.803568837951546e-06, "loss": 0.0289, "step": 16914 }, { "epoch": 1.8338031222896791, "grad_norm": 2.983959436416626, "learning_rate": 7.802843464384158e-06, "loss": 0.0185, "step": 16915 }, { "epoch": 1.8339115351257589, "grad_norm": 1.8754807710647583, "learning_rate": 7.802118090816772e-06, "loss": 0.0021, "step": 16916 }, { "epoch": 1.8340199479618386, "grad_norm": 6.864595413208008, "learning_rate": 7.801392717249384e-06, "loss": 0.0347, "step": 16917 }, { "epoch": 1.8341283607979184, "grad_norm": 1.3650938272476196, "learning_rate": 7.800667343681997e-06, "loss": 0.029, "step": 16918 }, { "epoch": 1.8342367736339984, "grad_norm": 1.049461007118225, "learning_rate": 7.799941970114609e-06, "loss": 0.0053, "step": 16919 }, { "epoch": 1.834345186470078, "grad_norm": 0.46591511368751526, "learning_rate": 7.799216596547223e-06, "loss": 0.024, "step": 16920 }, { "epoch": 1.834453599306158, "grad_norm": 0.6937958598136902, "learning_rate": 7.798491222979835e-06, "loss": 0.0144, "step": 16921 }, { "epoch": 1.8345620121422377, "grad_norm": 3.275550365447998, "learning_rate": 7.797765849412447e-06, "loss": 0.015, "step": 16922 }, { "epoch": 1.8346704249783174, "grad_norm": 1.6776494979858398, "learning_rate": 7.79704047584506e-06, "loss": 0.0652, "step": 16923 }, { "epoch": 1.8347788378143972, "grad_norm": 4.035539150238037, "learning_rate": 7.796315102277674e-06, "loss": 0.0255, "step": 16924 }, { "epoch": 1.834887250650477, "grad_norm": 0.5502243041992188, "learning_rate": 7.795589728710287e-06, "loss": 0.0014, "step": 16925 }, { "epoch": 1.834995663486557, "grad_norm": 17.803165435791016, "learning_rate": 7.794864355142899e-06, "loss": 0.0155, "step": 16926 }, { "epoch": 1.8351040763226365, "grad_norm": 1.4103403091430664, "learning_rate": 7.794138981575513e-06, "loss": 0.0074, "step": 16927 }, { "epoch": 1.8352124891587165, "grad_norm": 6.386130332946777, "learning_rate": 7.793413608008125e-06, "loss": 0.0587, "step": 16928 }, { "epoch": 1.835320901994796, "grad_norm": 0.34065258502960205, "learning_rate": 7.792688234440737e-06, "loss": 0.0223, "step": 16929 }, { "epoch": 1.835429314830876, "grad_norm": 0.5083040595054626, "learning_rate": 7.79196286087335e-06, "loss": 0.0028, "step": 16930 }, { "epoch": 1.8355377276669558, "grad_norm": 0.49201714992523193, "learning_rate": 7.791237487305965e-06, "loss": 0.0085, "step": 16931 }, { "epoch": 1.8356461405030355, "grad_norm": 0.48561835289001465, "learning_rate": 7.790512113738577e-06, "loss": 0.0028, "step": 16932 }, { "epoch": 1.8357545533391153, "grad_norm": 2.8282413482666016, "learning_rate": 7.789786740171189e-06, "loss": 0.0096, "step": 16933 }, { "epoch": 1.835862966175195, "grad_norm": 0.3509075939655304, "learning_rate": 7.789061366603801e-06, "loss": 0.0024, "step": 16934 }, { "epoch": 1.835971379011275, "grad_norm": 2.731231689453125, "learning_rate": 7.788335993036415e-06, "loss": 0.0335, "step": 16935 }, { "epoch": 1.8360797918473546, "grad_norm": 0.41663092374801636, "learning_rate": 7.787610619469027e-06, "loss": 0.0142, "step": 16936 }, { "epoch": 1.8361882046834346, "grad_norm": 0.2715320289134979, "learning_rate": 7.786885245901639e-06, "loss": 0.0064, "step": 16937 }, { "epoch": 1.8362966175195143, "grad_norm": 0.23692694306373596, "learning_rate": 7.786159872334253e-06, "loss": 0.0054, "step": 16938 }, { "epoch": 1.836405030355594, "grad_norm": 6.001845359802246, "learning_rate": 7.785434498766865e-06, "loss": 0.0299, "step": 16939 }, { "epoch": 1.8365134431916739, "grad_norm": 0.26689261198043823, "learning_rate": 7.784709125199479e-06, "loss": 0.0067, "step": 16940 }, { "epoch": 1.8366218560277536, "grad_norm": 4.360345363616943, "learning_rate": 7.783983751632091e-06, "loss": 0.0343, "step": 16941 }, { "epoch": 1.8367302688638336, "grad_norm": 3.6035470962524414, "learning_rate": 7.783258378064705e-06, "loss": 0.016, "step": 16942 }, { "epoch": 1.8368386816999132, "grad_norm": 1.9023767709732056, "learning_rate": 7.782533004497317e-06, "loss": 0.0219, "step": 16943 }, { "epoch": 1.8369470945359931, "grad_norm": 2.1731948852539062, "learning_rate": 7.78180763092993e-06, "loss": 0.0305, "step": 16944 }, { "epoch": 1.837055507372073, "grad_norm": 1.615236759185791, "learning_rate": 7.781082257362541e-06, "loss": 0.0481, "step": 16945 }, { "epoch": 1.8371639202081527, "grad_norm": 0.5814985632896423, "learning_rate": 7.780356883795155e-06, "loss": 0.0248, "step": 16946 }, { "epoch": 1.8372723330442324, "grad_norm": 1.9637633562088013, "learning_rate": 7.779631510227767e-06, "loss": 0.0205, "step": 16947 }, { "epoch": 1.8373807458803122, "grad_norm": 0.46151190996170044, "learning_rate": 7.77890613666038e-06, "loss": 0.0085, "step": 16948 }, { "epoch": 1.8374891587163922, "grad_norm": 7.210976600646973, "learning_rate": 7.778180763092993e-06, "loss": 0.0519, "step": 16949 }, { "epoch": 1.8375975715524717, "grad_norm": 0.3199470639228821, "learning_rate": 7.777455389525607e-06, "loss": 0.0015, "step": 16950 }, { "epoch": 1.8377059843885517, "grad_norm": 1.9438704252243042, "learning_rate": 7.77673001595822e-06, "loss": 0.0442, "step": 16951 }, { "epoch": 1.8378143972246312, "grad_norm": 3.3125946521759033, "learning_rate": 7.776004642390831e-06, "loss": 0.0399, "step": 16952 }, { "epoch": 1.8379228100607112, "grad_norm": 2.446523666381836, "learning_rate": 7.775279268823445e-06, "loss": 0.0299, "step": 16953 }, { "epoch": 1.838031222896791, "grad_norm": 3.187615156173706, "learning_rate": 7.774553895256057e-06, "loss": 0.0313, "step": 16954 }, { "epoch": 1.8381396357328708, "grad_norm": 6.337727069854736, "learning_rate": 7.77382852168867e-06, "loss": 0.0137, "step": 16955 }, { "epoch": 1.8382480485689505, "grad_norm": 1.3728069067001343, "learning_rate": 7.773103148121283e-06, "loss": 0.0266, "step": 16956 }, { "epoch": 1.8383564614050303, "grad_norm": 11.77234172821045, "learning_rate": 7.772377774553897e-06, "loss": 0.0349, "step": 16957 }, { "epoch": 1.8384648742411103, "grad_norm": 0.7224940061569214, "learning_rate": 7.77165240098651e-06, "loss": 0.019, "step": 16958 }, { "epoch": 1.8385732870771898, "grad_norm": 2.2503843307495117, "learning_rate": 7.770927027419122e-06, "loss": 0.0042, "step": 16959 }, { "epoch": 1.8386816999132698, "grad_norm": 1.2513386011123657, "learning_rate": 7.770201653851734e-06, "loss": 0.019, "step": 16960 }, { "epoch": 1.8387901127493496, "grad_norm": 4.797248363494873, "learning_rate": 7.769476280284348e-06, "loss": 0.0162, "step": 16961 }, { "epoch": 1.8388985255854293, "grad_norm": 0.07409842312335968, "learning_rate": 7.76875090671696e-06, "loss": 0.0005, "step": 16962 }, { "epoch": 1.839006938421509, "grad_norm": 0.3321739137172699, "learning_rate": 7.768025533149572e-06, "loss": 0.0075, "step": 16963 }, { "epoch": 1.8391153512575888, "grad_norm": 0.6015564799308777, "learning_rate": 7.767300159582186e-06, "loss": 0.0113, "step": 16964 }, { "epoch": 1.8392237640936688, "grad_norm": 1.8812062740325928, "learning_rate": 7.766574786014798e-06, "loss": 0.0438, "step": 16965 }, { "epoch": 1.8393321769297484, "grad_norm": 0.8116024136543274, "learning_rate": 7.765849412447412e-06, "loss": 0.0166, "step": 16966 }, { "epoch": 1.8394405897658284, "grad_norm": 0.8963645696640015, "learning_rate": 7.765124038880024e-06, "loss": 0.0067, "step": 16967 }, { "epoch": 1.839549002601908, "grad_norm": 1.3010201454162598, "learning_rate": 7.764398665312638e-06, "loss": 0.0087, "step": 16968 }, { "epoch": 1.8396574154379879, "grad_norm": 0.06928595900535583, "learning_rate": 7.76367329174525e-06, "loss": 0.0006, "step": 16969 }, { "epoch": 1.8397658282740676, "grad_norm": 0.7749621868133545, "learning_rate": 7.762947918177862e-06, "loss": 0.0049, "step": 16970 }, { "epoch": 1.8398742411101474, "grad_norm": 2.7408714294433594, "learning_rate": 7.762222544610474e-06, "loss": 0.0475, "step": 16971 }, { "epoch": 1.8399826539462274, "grad_norm": 0.6670923233032227, "learning_rate": 7.761497171043088e-06, "loss": 0.0137, "step": 16972 }, { "epoch": 1.840091066782307, "grad_norm": 1.786331057548523, "learning_rate": 7.7607717974757e-06, "loss": 0.0039, "step": 16973 }, { "epoch": 1.840199479618387, "grad_norm": 6.266730308532715, "learning_rate": 7.760046423908312e-06, "loss": 0.0125, "step": 16974 }, { "epoch": 1.8403078924544665, "grad_norm": 11.687625885009766, "learning_rate": 7.759321050340926e-06, "loss": 0.0432, "step": 16975 }, { "epoch": 1.8404163052905465, "grad_norm": 0.3892693519592285, "learning_rate": 7.75859567677354e-06, "loss": 0.0023, "step": 16976 }, { "epoch": 1.8405247181266262, "grad_norm": 1.9479162693023682, "learning_rate": 7.757870303206152e-06, "loss": 0.0379, "step": 16977 }, { "epoch": 1.840633130962706, "grad_norm": 0.08652554452419281, "learning_rate": 7.757144929638764e-06, "loss": 0.0009, "step": 16978 }, { "epoch": 1.8407415437987857, "grad_norm": 1.441481113433838, "learning_rate": 7.756419556071378e-06, "loss": 0.0105, "step": 16979 }, { "epoch": 1.8408499566348655, "grad_norm": 0.15575823187828064, "learning_rate": 7.75569418250399e-06, "loss": 0.001, "step": 16980 }, { "epoch": 1.8409583694709455, "grad_norm": 2.292327880859375, "learning_rate": 7.754968808936602e-06, "loss": 0.0185, "step": 16981 }, { "epoch": 1.841066782307025, "grad_norm": 1.3437145948410034, "learning_rate": 7.754243435369216e-06, "loss": 0.0018, "step": 16982 }, { "epoch": 1.841175195143105, "grad_norm": 0.263229638338089, "learning_rate": 7.75351806180183e-06, "loss": 0.0051, "step": 16983 }, { "epoch": 1.8412836079791848, "grad_norm": 3.068326711654663, "learning_rate": 7.752792688234442e-06, "loss": 0.0183, "step": 16984 }, { "epoch": 1.8413920208152645, "grad_norm": 0.34969034790992737, "learning_rate": 7.752067314667054e-06, "loss": 0.0125, "step": 16985 }, { "epoch": 1.8415004336513443, "grad_norm": 0.6465548276901245, "learning_rate": 7.751341941099666e-06, "loss": 0.0012, "step": 16986 }, { "epoch": 1.841608846487424, "grad_norm": 3.2901811599731445, "learning_rate": 7.75061656753228e-06, "loss": 0.1068, "step": 16987 }, { "epoch": 1.841717259323504, "grad_norm": 12.583723068237305, "learning_rate": 7.749891193964892e-06, "loss": 0.073, "step": 16988 }, { "epoch": 1.8418256721595836, "grad_norm": 6.071331024169922, "learning_rate": 7.749165820397504e-06, "loss": 0.0661, "step": 16989 }, { "epoch": 1.8419340849956636, "grad_norm": 0.4700002074241638, "learning_rate": 7.748440446830118e-06, "loss": 0.0092, "step": 16990 }, { "epoch": 1.8420424978317431, "grad_norm": 3.474302053451538, "learning_rate": 7.74771507326273e-06, "loss": 0.0275, "step": 16991 }, { "epoch": 1.842150910667823, "grad_norm": 1.6300595998764038, "learning_rate": 7.746989699695344e-06, "loss": 0.0052, "step": 16992 }, { "epoch": 1.8422593235039029, "grad_norm": 1.9060158729553223, "learning_rate": 7.746264326127956e-06, "loss": 0.0237, "step": 16993 }, { "epoch": 1.8423677363399826, "grad_norm": 0.2920624017715454, "learning_rate": 7.74553895256057e-06, "loss": 0.0062, "step": 16994 }, { "epoch": 1.8424761491760624, "grad_norm": 0.8991216421127319, "learning_rate": 7.744813578993182e-06, "loss": 0.0161, "step": 16995 }, { "epoch": 1.8425845620121422, "grad_norm": 2.749938726425171, "learning_rate": 7.744088205425795e-06, "loss": 0.0615, "step": 16996 }, { "epoch": 1.8426929748482221, "grad_norm": 4.6077494621276855, "learning_rate": 7.743362831858407e-06, "loss": 0.0223, "step": 16997 }, { "epoch": 1.8428013876843017, "grad_norm": 3.564406394958496, "learning_rate": 7.74263745829102e-06, "loss": 0.0263, "step": 16998 }, { "epoch": 1.8429098005203817, "grad_norm": 4.3040852546691895, "learning_rate": 7.741912084723633e-06, "loss": 0.0358, "step": 16999 }, { "epoch": 1.8430182133564614, "grad_norm": 0.12409649044275284, "learning_rate": 7.741186711156247e-06, "loss": 0.0009, "step": 17000 }, { "epoch": 1.8431266261925412, "grad_norm": 1.1193534135818481, "learning_rate": 7.740461337588859e-06, "loss": 0.0188, "step": 17001 }, { "epoch": 1.843235039028621, "grad_norm": 1.7217998504638672, "learning_rate": 7.739735964021473e-06, "loss": 0.0306, "step": 17002 }, { "epoch": 1.8433434518647007, "grad_norm": 0.46332019567489624, "learning_rate": 7.739010590454085e-06, "loss": 0.0232, "step": 17003 }, { "epoch": 1.8434518647007807, "grad_norm": 0.46034491062164307, "learning_rate": 7.738285216886697e-06, "loss": 0.0056, "step": 17004 }, { "epoch": 1.8435602775368602, "grad_norm": 0.4205484092235565, "learning_rate": 7.73755984331931e-06, "loss": 0.0086, "step": 17005 }, { "epoch": 1.8436686903729402, "grad_norm": 5.457734107971191, "learning_rate": 7.736834469751923e-06, "loss": 0.0368, "step": 17006 }, { "epoch": 1.84377710320902, "grad_norm": 3.39772891998291, "learning_rate": 7.736109096184535e-06, "loss": 0.0309, "step": 17007 }, { "epoch": 1.8438855160450998, "grad_norm": 0.6588850021362305, "learning_rate": 7.735383722617149e-06, "loss": 0.0081, "step": 17008 }, { "epoch": 1.8439939288811795, "grad_norm": 4.637040138244629, "learning_rate": 7.734658349049763e-06, "loss": 0.0153, "step": 17009 }, { "epoch": 1.8441023417172593, "grad_norm": 3.1079959869384766, "learning_rate": 7.733932975482375e-06, "loss": 0.0036, "step": 17010 }, { "epoch": 1.8442107545533393, "grad_norm": 1.9282913208007812, "learning_rate": 7.733207601914987e-06, "loss": 0.023, "step": 17011 }, { "epoch": 1.8443191673894188, "grad_norm": 2.545966148376465, "learning_rate": 7.732482228347599e-06, "loss": 0.0303, "step": 17012 }, { "epoch": 1.8444275802254988, "grad_norm": 9.821939468383789, "learning_rate": 7.731756854780213e-06, "loss": 0.0275, "step": 17013 }, { "epoch": 1.8445359930615783, "grad_norm": 0.7362653017044067, "learning_rate": 7.731031481212825e-06, "loss": 0.0215, "step": 17014 }, { "epoch": 1.8446444058976583, "grad_norm": 0.692424476146698, "learning_rate": 7.730306107645437e-06, "loss": 0.0159, "step": 17015 }, { "epoch": 1.844752818733738, "grad_norm": 3.717207431793213, "learning_rate": 7.729580734078051e-06, "loss": 0.0444, "step": 17016 }, { "epoch": 1.8448612315698178, "grad_norm": 1.9356437921524048, "learning_rate": 7.728855360510663e-06, "loss": 0.0061, "step": 17017 }, { "epoch": 1.8449696444058976, "grad_norm": 0.06887682527303696, "learning_rate": 7.728129986943277e-06, "loss": 0.0007, "step": 17018 }, { "epoch": 1.8450780572419774, "grad_norm": 0.28599366545677185, "learning_rate": 7.727404613375889e-06, "loss": 0.0067, "step": 17019 }, { "epoch": 1.8451864700780574, "grad_norm": 2.2386152744293213, "learning_rate": 7.726679239808503e-06, "loss": 0.0193, "step": 17020 }, { "epoch": 1.845294882914137, "grad_norm": 1.8618935346603394, "learning_rate": 7.725953866241115e-06, "loss": 0.0102, "step": 17021 }, { "epoch": 1.8454032957502169, "grad_norm": 0.6614826321601868, "learning_rate": 7.725228492673727e-06, "loss": 0.0092, "step": 17022 }, { "epoch": 1.8455117085862967, "grad_norm": 6.778225898742676, "learning_rate": 7.72450311910634e-06, "loss": 0.0179, "step": 17023 }, { "epoch": 1.8456201214223764, "grad_norm": 4.7637553215026855, "learning_rate": 7.723777745538953e-06, "loss": 0.0217, "step": 17024 }, { "epoch": 1.8457285342584562, "grad_norm": 2.107952117919922, "learning_rate": 7.723052371971565e-06, "loss": 0.0372, "step": 17025 }, { "epoch": 1.845836947094536, "grad_norm": 0.7081575393676758, "learning_rate": 7.72232699840418e-06, "loss": 0.0143, "step": 17026 }, { "epoch": 1.845945359930616, "grad_norm": 0.7614714503288269, "learning_rate": 7.721601624836791e-06, "loss": 0.007, "step": 17027 }, { "epoch": 1.8460537727666955, "grad_norm": 1.9174087047576904, "learning_rate": 7.720876251269405e-06, "loss": 0.0156, "step": 17028 }, { "epoch": 1.8461621856027755, "grad_norm": 1.0619640350341797, "learning_rate": 7.720150877702017e-06, "loss": 0.0147, "step": 17029 }, { "epoch": 1.8462705984388552, "grad_norm": 5.758350849151611, "learning_rate": 7.71942550413463e-06, "loss": 0.0361, "step": 17030 }, { "epoch": 1.846379011274935, "grad_norm": 0.6623355746269226, "learning_rate": 7.718700130567243e-06, "loss": 0.0044, "step": 17031 }, { "epoch": 1.8464874241110147, "grad_norm": 7.263010501861572, "learning_rate": 7.717974756999855e-06, "loss": 0.0277, "step": 17032 }, { "epoch": 1.8465958369470945, "grad_norm": 1.7437143325805664, "learning_rate": 7.717249383432468e-06, "loss": 0.0139, "step": 17033 }, { "epoch": 1.8467042497831745, "grad_norm": 0.37925416231155396, "learning_rate": 7.716524009865081e-06, "loss": 0.0103, "step": 17034 }, { "epoch": 1.846812662619254, "grad_norm": 0.539841890335083, "learning_rate": 7.715798636297695e-06, "loss": 0.0094, "step": 17035 }, { "epoch": 1.846921075455334, "grad_norm": 1.9569569826126099, "learning_rate": 7.715073262730307e-06, "loss": 0.03, "step": 17036 }, { "epoch": 1.8470294882914136, "grad_norm": 8.349591255187988, "learning_rate": 7.71434788916292e-06, "loss": 0.0183, "step": 17037 }, { "epoch": 1.8471379011274935, "grad_norm": 5.924037933349609, "learning_rate": 7.713622515595532e-06, "loss": 0.0252, "step": 17038 }, { "epoch": 1.8472463139635733, "grad_norm": 2.218311309814453, "learning_rate": 7.712897142028146e-06, "loss": 0.0316, "step": 17039 }, { "epoch": 1.847354726799653, "grad_norm": 3.290003776550293, "learning_rate": 7.712171768460758e-06, "loss": 0.037, "step": 17040 }, { "epoch": 1.8474631396357328, "grad_norm": 0.4425044357776642, "learning_rate": 7.71144639489337e-06, "loss": 0.0085, "step": 17041 }, { "epoch": 1.8475715524718126, "grad_norm": 0.2870739996433258, "learning_rate": 7.710721021325984e-06, "loss": 0.0019, "step": 17042 }, { "epoch": 1.8476799653078926, "grad_norm": 2.535141944885254, "learning_rate": 7.709995647758596e-06, "loss": 0.0153, "step": 17043 }, { "epoch": 1.8477883781439721, "grad_norm": 0.40948304533958435, "learning_rate": 7.70927027419121e-06, "loss": 0.002, "step": 17044 }, { "epoch": 1.847896790980052, "grad_norm": 0.12124698609113693, "learning_rate": 7.708544900623822e-06, "loss": 0.0006, "step": 17045 }, { "epoch": 1.8480052038161319, "grad_norm": 8.133979797363281, "learning_rate": 7.707819527056436e-06, "loss": 0.0604, "step": 17046 }, { "epoch": 1.8481136166522116, "grad_norm": 1.1347346305847168, "learning_rate": 7.707094153489048e-06, "loss": 0.0277, "step": 17047 }, { "epoch": 1.8482220294882914, "grad_norm": 4.742656707763672, "learning_rate": 7.70636877992166e-06, "loss": 0.0323, "step": 17048 }, { "epoch": 1.8483304423243712, "grad_norm": 2.0495285987854004, "learning_rate": 7.705643406354272e-06, "loss": 0.0145, "step": 17049 }, { "epoch": 1.8484388551604511, "grad_norm": 3.0475943088531494, "learning_rate": 7.704918032786886e-06, "loss": 0.0036, "step": 17050 }, { "epoch": 1.8485472679965307, "grad_norm": 1.6539193391799927, "learning_rate": 7.704192659219498e-06, "loss": 0.0167, "step": 17051 }, { "epoch": 1.8486556808326107, "grad_norm": 1.8859107494354248, "learning_rate": 7.703467285652112e-06, "loss": 0.0224, "step": 17052 }, { "epoch": 1.8487640936686902, "grad_norm": 0.2757059931755066, "learning_rate": 7.702741912084724e-06, "loss": 0.0054, "step": 17053 }, { "epoch": 1.8488725065047702, "grad_norm": 0.5209680199623108, "learning_rate": 7.702016538517338e-06, "loss": 0.0197, "step": 17054 }, { "epoch": 1.84898091934085, "grad_norm": 5.251359462738037, "learning_rate": 7.70129116494995e-06, "loss": 0.0474, "step": 17055 }, { "epoch": 1.8490893321769297, "grad_norm": 5.478834629058838, "learning_rate": 7.700565791382562e-06, "loss": 0.026, "step": 17056 }, { "epoch": 1.8491977450130095, "grad_norm": 0.10092934221029282, "learning_rate": 7.699840417815176e-06, "loss": 0.0007, "step": 17057 }, { "epoch": 1.8493061578490892, "grad_norm": 2.327704668045044, "learning_rate": 7.699115044247788e-06, "loss": 0.0182, "step": 17058 }, { "epoch": 1.8494145706851692, "grad_norm": 1.0484405755996704, "learning_rate": 7.6983896706804e-06, "loss": 0.0162, "step": 17059 }, { "epoch": 1.8495229835212488, "grad_norm": 0.29137492179870605, "learning_rate": 7.697664297113014e-06, "loss": 0.0062, "step": 17060 }, { "epoch": 1.8496313963573288, "grad_norm": 1.7976652383804321, "learning_rate": 7.696938923545628e-06, "loss": 0.0195, "step": 17061 }, { "epoch": 1.8497398091934085, "grad_norm": 1.9312820434570312, "learning_rate": 7.69621354997824e-06, "loss": 0.017, "step": 17062 }, { "epoch": 1.8498482220294883, "grad_norm": 0.5168769955635071, "learning_rate": 7.695488176410852e-06, "loss": 0.016, "step": 17063 }, { "epoch": 1.849956634865568, "grad_norm": 2.470547914505005, "learning_rate": 7.694762802843464e-06, "loss": 0.038, "step": 17064 }, { "epoch": 1.8500650477016478, "grad_norm": 0.3691755533218384, "learning_rate": 7.694037429276078e-06, "loss": 0.0056, "step": 17065 }, { "epoch": 1.8501734605377278, "grad_norm": 2.9474782943725586, "learning_rate": 7.69331205570869e-06, "loss": 0.0537, "step": 17066 }, { "epoch": 1.8502818733738073, "grad_norm": 0.5804237127304077, "learning_rate": 7.692586682141302e-06, "loss": 0.0127, "step": 17067 }, { "epoch": 1.8503902862098873, "grad_norm": 0.23734109103679657, "learning_rate": 7.691861308573916e-06, "loss": 0.0011, "step": 17068 }, { "epoch": 1.850498699045967, "grad_norm": 0.892844557762146, "learning_rate": 7.691135935006528e-06, "loss": 0.0049, "step": 17069 }, { "epoch": 1.8506071118820469, "grad_norm": 7.260867595672607, "learning_rate": 7.690410561439142e-06, "loss": 0.0584, "step": 17070 }, { "epoch": 1.8507155247181266, "grad_norm": 1.7272095680236816, "learning_rate": 7.689685187871754e-06, "loss": 0.0295, "step": 17071 }, { "epoch": 1.8508239375542064, "grad_norm": 1.0438193082809448, "learning_rate": 7.688959814304368e-06, "loss": 0.02, "step": 17072 }, { "epoch": 1.8509323503902864, "grad_norm": 1.701474666595459, "learning_rate": 7.68823444073698e-06, "loss": 0.02, "step": 17073 }, { "epoch": 1.851040763226366, "grad_norm": 1.8608170747756958, "learning_rate": 7.687509067169593e-06, "loss": 0.0071, "step": 17074 }, { "epoch": 1.851149176062446, "grad_norm": 0.955700695514679, "learning_rate": 7.686783693602205e-06, "loss": 0.0154, "step": 17075 }, { "epoch": 1.8512575888985254, "grad_norm": 8.186729431152344, "learning_rate": 7.686058320034819e-06, "loss": 0.0364, "step": 17076 }, { "epoch": 1.8513660017346054, "grad_norm": 0.3729745149612427, "learning_rate": 7.68533294646743e-06, "loss": 0.0063, "step": 17077 }, { "epoch": 1.8514744145706852, "grad_norm": 0.5280081629753113, "learning_rate": 7.684607572900045e-06, "loss": 0.018, "step": 17078 }, { "epoch": 1.851582827406765, "grad_norm": 0.2756633460521698, "learning_rate": 7.683882199332657e-06, "loss": 0.0063, "step": 17079 }, { "epoch": 1.8516912402428447, "grad_norm": 1.1022716760635376, "learning_rate": 7.68315682576527e-06, "loss": 0.024, "step": 17080 }, { "epoch": 1.8517996530789245, "grad_norm": 3.295250177383423, "learning_rate": 7.682431452197883e-06, "loss": 0.0704, "step": 17081 }, { "epoch": 1.8519080659150045, "grad_norm": 2.3618528842926025, "learning_rate": 7.681706078630495e-06, "loss": 0.0265, "step": 17082 }, { "epoch": 1.852016478751084, "grad_norm": 0.5525144934654236, "learning_rate": 7.680980705063109e-06, "loss": 0.0025, "step": 17083 }, { "epoch": 1.852124891587164, "grad_norm": 3.459808588027954, "learning_rate": 7.68025533149572e-06, "loss": 0.0222, "step": 17084 }, { "epoch": 1.8522333044232437, "grad_norm": 2.616814613342285, "learning_rate": 7.679529957928333e-06, "loss": 0.0235, "step": 17085 }, { "epoch": 1.8523417172593235, "grad_norm": 0.704574465751648, "learning_rate": 7.678804584360947e-06, "loss": 0.0008, "step": 17086 }, { "epoch": 1.8524501300954033, "grad_norm": 2.2688958644866943, "learning_rate": 7.67807921079356e-06, "loss": 0.034, "step": 17087 }, { "epoch": 1.852558542931483, "grad_norm": 1.3368769884109497, "learning_rate": 7.677353837226173e-06, "loss": 0.0124, "step": 17088 }, { "epoch": 1.852666955767563, "grad_norm": 1.9137377738952637, "learning_rate": 7.676628463658785e-06, "loss": 0.0152, "step": 17089 }, { "epoch": 1.8527753686036426, "grad_norm": 2.741668462753296, "learning_rate": 7.675903090091397e-06, "loss": 0.0264, "step": 17090 }, { "epoch": 1.8528837814397225, "grad_norm": 2.071582794189453, "learning_rate": 7.675177716524011e-06, "loss": 0.0166, "step": 17091 }, { "epoch": 1.8529921942758023, "grad_norm": 4.460122108459473, "learning_rate": 7.674452342956623e-06, "loss": 0.0511, "step": 17092 }, { "epoch": 1.853100607111882, "grad_norm": 0.478912889957428, "learning_rate": 7.673726969389235e-06, "loss": 0.0079, "step": 17093 }, { "epoch": 1.8532090199479618, "grad_norm": 2.3380565643310547, "learning_rate": 7.673001595821849e-06, "loss": 0.0226, "step": 17094 }, { "epoch": 1.8533174327840416, "grad_norm": 0.7146459221839905, "learning_rate": 7.672276222254463e-06, "loss": 0.0036, "step": 17095 }, { "epoch": 1.8534258456201216, "grad_norm": 1.6581147909164429, "learning_rate": 7.671550848687075e-06, "loss": 0.0069, "step": 17096 }, { "epoch": 1.8535342584562011, "grad_norm": 4.401130676269531, "learning_rate": 7.670825475119687e-06, "loss": 0.0607, "step": 17097 }, { "epoch": 1.853642671292281, "grad_norm": 0.28841638565063477, "learning_rate": 7.670100101552301e-06, "loss": 0.0105, "step": 17098 }, { "epoch": 1.8537510841283606, "grad_norm": 3.285973072052002, "learning_rate": 7.669374727984913e-06, "loss": 0.0736, "step": 17099 }, { "epoch": 1.8538594969644406, "grad_norm": 1.4041558504104614, "learning_rate": 7.668649354417525e-06, "loss": 0.0066, "step": 17100 }, { "epoch": 1.8539679098005204, "grad_norm": 4.690195560455322, "learning_rate": 7.667923980850139e-06, "loss": 0.0433, "step": 17101 }, { "epoch": 1.8540763226366002, "grad_norm": 5.49570369720459, "learning_rate": 7.667198607282751e-06, "loss": 0.0646, "step": 17102 }, { "epoch": 1.85418473547268, "grad_norm": 0.054872266948223114, "learning_rate": 7.666473233715363e-06, "loss": 0.0006, "step": 17103 }, { "epoch": 1.8542931483087597, "grad_norm": 0.3116517663002014, "learning_rate": 7.665747860147977e-06, "loss": 0.0064, "step": 17104 }, { "epoch": 1.8544015611448397, "grad_norm": 4.001355171203613, "learning_rate": 7.66502248658059e-06, "loss": 0.0158, "step": 17105 }, { "epoch": 1.8545099739809192, "grad_norm": 1.0281636714935303, "learning_rate": 7.664297113013203e-06, "loss": 0.0132, "step": 17106 }, { "epoch": 1.8546183868169992, "grad_norm": 4.694221496582031, "learning_rate": 7.663571739445815e-06, "loss": 0.0236, "step": 17107 }, { "epoch": 1.854726799653079, "grad_norm": 0.09797199815511703, "learning_rate": 7.662846365878427e-06, "loss": 0.0007, "step": 17108 }, { "epoch": 1.8548352124891587, "grad_norm": 0.6669718623161316, "learning_rate": 7.662120992311041e-06, "loss": 0.0182, "step": 17109 }, { "epoch": 1.8549436253252385, "grad_norm": 1.2118127346038818, "learning_rate": 7.661395618743653e-06, "loss": 0.0107, "step": 17110 }, { "epoch": 1.8550520381613183, "grad_norm": 5.124387264251709, "learning_rate": 7.660670245176266e-06, "loss": 0.025, "step": 17111 }, { "epoch": 1.8551604509973982, "grad_norm": 6.495660305023193, "learning_rate": 7.65994487160888e-06, "loss": 0.0261, "step": 17112 }, { "epoch": 1.8552688638334778, "grad_norm": 0.37461403012275696, "learning_rate": 7.659219498041493e-06, "loss": 0.0027, "step": 17113 }, { "epoch": 1.8553772766695578, "grad_norm": 2.262303352355957, "learning_rate": 7.658494124474105e-06, "loss": 0.0085, "step": 17114 }, { "epoch": 1.8554856895056373, "grad_norm": 5.341168403625488, "learning_rate": 7.657768750906718e-06, "loss": 0.0168, "step": 17115 }, { "epoch": 1.8555941023417173, "grad_norm": 5.391664505004883, "learning_rate": 7.65704337733933e-06, "loss": 0.0405, "step": 17116 }, { "epoch": 1.855702515177797, "grad_norm": 1.2687641382217407, "learning_rate": 7.656318003771944e-06, "loss": 0.024, "step": 17117 }, { "epoch": 1.8558109280138768, "grad_norm": 0.5073751211166382, "learning_rate": 7.655592630204556e-06, "loss": 0.0164, "step": 17118 }, { "epoch": 1.8559193408499568, "grad_norm": 0.48307672142982483, "learning_rate": 7.654867256637168e-06, "loss": 0.0028, "step": 17119 }, { "epoch": 1.8560277536860363, "grad_norm": 6.337152481079102, "learning_rate": 7.654141883069782e-06, "loss": 0.0402, "step": 17120 }, { "epoch": 1.8561361665221163, "grad_norm": 3.9663610458374023, "learning_rate": 7.653416509502395e-06, "loss": 0.0337, "step": 17121 }, { "epoch": 1.8562445793581959, "grad_norm": 6.861896991729736, "learning_rate": 7.652691135935008e-06, "loss": 0.0057, "step": 17122 }, { "epoch": 1.8563529921942759, "grad_norm": 0.34244105219841003, "learning_rate": 7.65196576236762e-06, "loss": 0.004, "step": 17123 }, { "epoch": 1.8564614050303556, "grad_norm": 1.7768951654434204, "learning_rate": 7.651240388800234e-06, "loss": 0.0032, "step": 17124 }, { "epoch": 1.8565698178664354, "grad_norm": 2.6207730770111084, "learning_rate": 7.650515015232846e-06, "loss": 0.0302, "step": 17125 }, { "epoch": 1.8566782307025151, "grad_norm": 1.629948616027832, "learning_rate": 7.649789641665458e-06, "loss": 0.0521, "step": 17126 }, { "epoch": 1.856786643538595, "grad_norm": 1.6529004573822021, "learning_rate": 7.649064268098072e-06, "loss": 0.0301, "step": 17127 }, { "epoch": 1.856895056374675, "grad_norm": 0.45044606924057007, "learning_rate": 7.648338894530684e-06, "loss": 0.0256, "step": 17128 }, { "epoch": 1.8570034692107544, "grad_norm": 2.5096235275268555, "learning_rate": 7.647613520963296e-06, "loss": 0.0159, "step": 17129 }, { "epoch": 1.8571118820468344, "grad_norm": 0.3027070164680481, "learning_rate": 7.64688814739591e-06, "loss": 0.0131, "step": 17130 }, { "epoch": 1.8572202948829142, "grad_norm": 0.7495800852775574, "learning_rate": 7.646162773828522e-06, "loss": 0.0174, "step": 17131 }, { "epoch": 1.857328707718994, "grad_norm": 7.1156110763549805, "learning_rate": 7.645437400261136e-06, "loss": 0.0334, "step": 17132 }, { "epoch": 1.8574371205550737, "grad_norm": 2.3646292686462402, "learning_rate": 7.644712026693748e-06, "loss": 0.0454, "step": 17133 }, { "epoch": 1.8575455333911535, "grad_norm": 0.9955110549926758, "learning_rate": 7.64398665312636e-06, "loss": 0.0099, "step": 17134 }, { "epoch": 1.8576539462272335, "grad_norm": 4.640987873077393, "learning_rate": 7.643261279558974e-06, "loss": 0.0152, "step": 17135 }, { "epoch": 1.857762359063313, "grad_norm": 7.466670513153076, "learning_rate": 7.642535905991586e-06, "loss": 0.0136, "step": 17136 }, { "epoch": 1.857870771899393, "grad_norm": 1.276297688484192, "learning_rate": 7.641810532424198e-06, "loss": 0.0039, "step": 17137 }, { "epoch": 1.8579791847354725, "grad_norm": 5.853169918060303, "learning_rate": 7.641085158856812e-06, "loss": 0.0177, "step": 17138 }, { "epoch": 1.8580875975715525, "grad_norm": 4.460516452789307, "learning_rate": 7.640359785289426e-06, "loss": 0.0525, "step": 17139 }, { "epoch": 1.8581960104076323, "grad_norm": 0.7945716381072998, "learning_rate": 7.639634411722038e-06, "loss": 0.0053, "step": 17140 }, { "epoch": 1.858304423243712, "grad_norm": 5.838634967803955, "learning_rate": 7.63890903815465e-06, "loss": 0.0141, "step": 17141 }, { "epoch": 1.8584128360797918, "grad_norm": 0.31849929690361023, "learning_rate": 7.638183664587262e-06, "loss": 0.0014, "step": 17142 }, { "epoch": 1.8585212489158716, "grad_norm": 3.0513014793395996, "learning_rate": 7.637458291019876e-06, "loss": 0.0193, "step": 17143 }, { "epoch": 1.8586296617519515, "grad_norm": 1.2597781419754028, "learning_rate": 7.636732917452488e-06, "loss": 0.0247, "step": 17144 }, { "epoch": 1.858738074588031, "grad_norm": 0.46217864751815796, "learning_rate": 7.6360075438851e-06, "loss": 0.0025, "step": 17145 }, { "epoch": 1.858846487424111, "grad_norm": 2.230465888977051, "learning_rate": 7.635282170317714e-06, "loss": 0.0314, "step": 17146 }, { "epoch": 1.8589549002601908, "grad_norm": 1.0678480863571167, "learning_rate": 7.634556796750328e-06, "loss": 0.0018, "step": 17147 }, { "epoch": 1.8590633130962706, "grad_norm": 2.3527066707611084, "learning_rate": 7.63383142318294e-06, "loss": 0.015, "step": 17148 }, { "epoch": 1.8591717259323504, "grad_norm": 1.1582239866256714, "learning_rate": 7.633106049615552e-06, "loss": 0.0064, "step": 17149 }, { "epoch": 1.8592801387684301, "grad_norm": 2.222419261932373, "learning_rate": 7.632380676048166e-06, "loss": 0.012, "step": 17150 }, { "epoch": 1.85938855160451, "grad_norm": 0.5436112880706787, "learning_rate": 7.631655302480778e-06, "loss": 0.0129, "step": 17151 }, { "epoch": 1.8594969644405897, "grad_norm": 6.460822582244873, "learning_rate": 7.63092992891339e-06, "loss": 0.0361, "step": 17152 }, { "epoch": 1.8596053772766696, "grad_norm": 2.6202120780944824, "learning_rate": 7.630204555346004e-06, "loss": 0.0448, "step": 17153 }, { "epoch": 1.8597137901127494, "grad_norm": 3.584228754043579, "learning_rate": 7.629479181778617e-06, "loss": 0.0195, "step": 17154 }, { "epoch": 1.8598222029488292, "grad_norm": 0.29914340376853943, "learning_rate": 7.628753808211229e-06, "loss": 0.0017, "step": 17155 }, { "epoch": 1.859930615784909, "grad_norm": 0.18954700231552124, "learning_rate": 7.6280284346438425e-06, "loss": 0.0014, "step": 17156 }, { "epoch": 1.8600390286209887, "grad_norm": 4.430742263793945, "learning_rate": 7.6273030610764555e-06, "loss": 0.0032, "step": 17157 }, { "epoch": 1.8601474414570687, "grad_norm": 1.0066885948181152, "learning_rate": 7.626577687509068e-06, "loss": 0.016, "step": 17158 }, { "epoch": 1.8602558542931482, "grad_norm": 2.9126181602478027, "learning_rate": 7.625852313941681e-06, "loss": 0.0072, "step": 17159 }, { "epoch": 1.8603642671292282, "grad_norm": 0.9151864647865295, "learning_rate": 7.625126940374294e-06, "loss": 0.0105, "step": 17160 }, { "epoch": 1.8604726799653077, "grad_norm": 4.06859827041626, "learning_rate": 7.624401566806906e-06, "loss": 0.0322, "step": 17161 }, { "epoch": 1.8605810928013877, "grad_norm": 4.449569225311279, "learning_rate": 7.623676193239519e-06, "loss": 0.0272, "step": 17162 }, { "epoch": 1.8606895056374675, "grad_norm": 7.980531215667725, "learning_rate": 7.622950819672132e-06, "loss": 0.0261, "step": 17163 }, { "epoch": 1.8607979184735473, "grad_norm": 0.05334484577178955, "learning_rate": 7.622225446104744e-06, "loss": 0.0002, "step": 17164 }, { "epoch": 1.860906331309627, "grad_norm": 0.7808350920677185, "learning_rate": 7.621500072537358e-06, "loss": 0.0038, "step": 17165 }, { "epoch": 1.8610147441457068, "grad_norm": 0.40730658173561096, "learning_rate": 7.620774698969971e-06, "loss": 0.028, "step": 17166 }, { "epoch": 1.8611231569817868, "grad_norm": 0.05111568793654442, "learning_rate": 7.620049325402583e-06, "loss": 0.0006, "step": 17167 }, { "epoch": 1.8612315698178663, "grad_norm": 3.7800910472869873, "learning_rate": 7.619323951835196e-06, "loss": 0.0306, "step": 17168 }, { "epoch": 1.8613399826539463, "grad_norm": 2.9829227924346924, "learning_rate": 7.618598578267809e-06, "loss": 0.0696, "step": 17169 }, { "epoch": 1.861448395490026, "grad_norm": 0.6806905269622803, "learning_rate": 7.617873204700421e-06, "loss": 0.0071, "step": 17170 }, { "epoch": 1.8615568083261058, "grad_norm": 1.578525185585022, "learning_rate": 7.617147831133034e-06, "loss": 0.0163, "step": 17171 }, { "epoch": 1.8616652211621856, "grad_norm": 0.1791577786207199, "learning_rate": 7.616422457565646e-06, "loss": 0.0018, "step": 17172 }, { "epoch": 1.8617736339982653, "grad_norm": 2.839191198348999, "learning_rate": 7.61569708399826e-06, "loss": 0.0214, "step": 17173 }, { "epoch": 1.8618820468343453, "grad_norm": 4.5514044761657715, "learning_rate": 7.614971710430873e-06, "loss": 0.0186, "step": 17174 }, { "epoch": 1.8619904596704249, "grad_norm": 0.8718957304954529, "learning_rate": 7.614246336863486e-06, "loss": 0.0484, "step": 17175 }, { "epoch": 1.8620988725065049, "grad_norm": 1.9276078939437866, "learning_rate": 7.613520963296098e-06, "loss": 0.0255, "step": 17176 }, { "epoch": 1.8622072853425846, "grad_norm": 6.628270626068115, "learning_rate": 7.612795589728711e-06, "loss": 0.0252, "step": 17177 }, { "epoch": 1.8623156981786644, "grad_norm": 0.3203030526638031, "learning_rate": 7.612070216161323e-06, "loss": 0.0063, "step": 17178 }, { "epoch": 1.8624241110147441, "grad_norm": 1.999302864074707, "learning_rate": 7.611344842593936e-06, "loss": 0.014, "step": 17179 }, { "epoch": 1.862532523850824, "grad_norm": 1.5238087177276611, "learning_rate": 7.610619469026549e-06, "loss": 0.0028, "step": 17180 }, { "epoch": 1.862640936686904, "grad_norm": 7.480490684509277, "learning_rate": 7.609894095459161e-06, "loss": 0.0403, "step": 17181 }, { "epoch": 1.8627493495229834, "grad_norm": 0.7502231001853943, "learning_rate": 7.609168721891775e-06, "loss": 0.0212, "step": 17182 }, { "epoch": 1.8628577623590634, "grad_norm": 0.3434661626815796, "learning_rate": 7.608443348324388e-06, "loss": 0.0069, "step": 17183 }, { "epoch": 1.862966175195143, "grad_norm": 0.122219517827034, "learning_rate": 7.607717974757e-06, "loss": 0.0008, "step": 17184 }, { "epoch": 1.863074588031223, "grad_norm": 0.4289276599884033, "learning_rate": 7.606992601189613e-06, "loss": 0.0028, "step": 17185 }, { "epoch": 1.8631830008673027, "grad_norm": 0.6793042421340942, "learning_rate": 7.606267227622226e-06, "loss": 0.019, "step": 17186 }, { "epoch": 1.8632914137033825, "grad_norm": 1.5869245529174805, "learning_rate": 7.6055418540548385e-06, "loss": 0.0175, "step": 17187 }, { "epoch": 1.8633998265394622, "grad_norm": 2.249077558517456, "learning_rate": 7.6048164804874514e-06, "loss": 0.0232, "step": 17188 }, { "epoch": 1.863508239375542, "grad_norm": 0.8713834881782532, "learning_rate": 7.6040911069200644e-06, "loss": 0.0158, "step": 17189 }, { "epoch": 1.863616652211622, "grad_norm": 2.606482982635498, "learning_rate": 7.6033657333526774e-06, "loss": 0.0303, "step": 17190 }, { "epoch": 1.8637250650477015, "grad_norm": 4.792291641235352, "learning_rate": 7.60264035978529e-06, "loss": 0.0177, "step": 17191 }, { "epoch": 1.8638334778837815, "grad_norm": 0.912223219871521, "learning_rate": 7.601914986217903e-06, "loss": 0.0105, "step": 17192 }, { "epoch": 1.8639418907198613, "grad_norm": 0.3979629874229431, "learning_rate": 7.6011896126505156e-06, "loss": 0.0116, "step": 17193 }, { "epoch": 1.864050303555941, "grad_norm": 0.24275179207324982, "learning_rate": 7.6004642390831285e-06, "loss": 0.0022, "step": 17194 }, { "epoch": 1.8641587163920208, "grad_norm": 2.737917184829712, "learning_rate": 7.5997388655157415e-06, "loss": 0.0903, "step": 17195 }, { "epoch": 1.8642671292281006, "grad_norm": 0.6706730723381042, "learning_rate": 7.599013491948354e-06, "loss": 0.0208, "step": 17196 }, { "epoch": 1.8643755420641805, "grad_norm": 6.343048095703125, "learning_rate": 7.598288118380967e-06, "loss": 0.005, "step": 17197 }, { "epoch": 1.86448395490026, "grad_norm": 1.790963053703308, "learning_rate": 7.597562744813579e-06, "loss": 0.0292, "step": 17198 }, { "epoch": 1.86459236773634, "grad_norm": 0.8440129160881042, "learning_rate": 7.596837371246193e-06, "loss": 0.0082, "step": 17199 }, { "epoch": 1.8647007805724196, "grad_norm": 4.526053428649902, "learning_rate": 7.596111997678806e-06, "loss": 0.0419, "step": 17200 }, { "epoch": 1.8648091934084996, "grad_norm": 5.072968482971191, "learning_rate": 7.595386624111419e-06, "loss": 0.0454, "step": 17201 }, { "epoch": 1.8649176062445794, "grad_norm": 5.150339126586914, "learning_rate": 7.594661250544031e-06, "loss": 0.0266, "step": 17202 }, { "epoch": 1.8650260190806591, "grad_norm": 4.133177757263184, "learning_rate": 7.593935876976644e-06, "loss": 0.0708, "step": 17203 }, { "epoch": 1.8651344319167389, "grad_norm": 7.3041462898254395, "learning_rate": 7.593210503409256e-06, "loss": 0.0373, "step": 17204 }, { "epoch": 1.8652428447528187, "grad_norm": 2.1489622592926025, "learning_rate": 7.592485129841869e-06, "loss": 0.0251, "step": 17205 }, { "epoch": 1.8653512575888986, "grad_norm": 0.3319332003593445, "learning_rate": 7.591759756274482e-06, "loss": 0.0132, "step": 17206 }, { "epoch": 1.8654596704249782, "grad_norm": 4.642762184143066, "learning_rate": 7.591034382707094e-06, "loss": 0.0393, "step": 17207 }, { "epoch": 1.8655680832610582, "grad_norm": 6.284135341644287, "learning_rate": 7.590309009139708e-06, "loss": 0.055, "step": 17208 }, { "epoch": 1.865676496097138, "grad_norm": 1.6864103078842163, "learning_rate": 7.589583635572321e-06, "loss": 0.0398, "step": 17209 }, { "epoch": 1.8657849089332177, "grad_norm": 1.4912117719650269, "learning_rate": 7.588858262004933e-06, "loss": 0.0439, "step": 17210 }, { "epoch": 1.8658933217692975, "grad_norm": 0.7997227907180786, "learning_rate": 7.588132888437546e-06, "loss": 0.0143, "step": 17211 }, { "epoch": 1.8660017346053772, "grad_norm": 0.11216980963945389, "learning_rate": 7.587407514870159e-06, "loss": 0.0009, "step": 17212 }, { "epoch": 1.8661101474414572, "grad_norm": 2.5928900241851807, "learning_rate": 7.586682141302771e-06, "loss": 0.0298, "step": 17213 }, { "epoch": 1.8662185602775367, "grad_norm": 1.250890851020813, "learning_rate": 7.585956767735384e-06, "loss": 0.0706, "step": 17214 }, { "epoch": 1.8663269731136167, "grad_norm": 0.5352864265441895, "learning_rate": 7.585231394167997e-06, "loss": 0.0139, "step": 17215 }, { "epoch": 1.8664353859496965, "grad_norm": 2.6906585693359375, "learning_rate": 7.58450602060061e-06, "loss": 0.0249, "step": 17216 }, { "epoch": 1.8665437987857763, "grad_norm": 0.9670906066894531, "learning_rate": 7.583780647033223e-06, "loss": 0.0173, "step": 17217 }, { "epoch": 1.866652211621856, "grad_norm": 1.2746752500534058, "learning_rate": 7.583055273465836e-06, "loss": 0.013, "step": 17218 }, { "epoch": 1.8667606244579358, "grad_norm": 1.2130320072174072, "learning_rate": 7.582329899898448e-06, "loss": 0.0168, "step": 17219 }, { "epoch": 1.8668690372940158, "grad_norm": 0.23742294311523438, "learning_rate": 7.581604526331061e-06, "loss": 0.0052, "step": 17220 }, { "epoch": 1.8669774501300953, "grad_norm": 0.5222061276435852, "learning_rate": 7.580879152763674e-06, "loss": 0.0171, "step": 17221 }, { "epoch": 1.8670858629661753, "grad_norm": 0.28326815366744995, "learning_rate": 7.580153779196286e-06, "loss": 0.0029, "step": 17222 }, { "epoch": 1.8671942758022548, "grad_norm": 0.9899402856826782, "learning_rate": 7.579428405628899e-06, "loss": 0.0397, "step": 17223 }, { "epoch": 1.8673026886383348, "grad_norm": 1.422886610031128, "learning_rate": 7.5787030320615115e-06, "loss": 0.045, "step": 17224 }, { "epoch": 1.8674111014744146, "grad_norm": 0.03857401758432388, "learning_rate": 7.577977658494125e-06, "loss": 0.0006, "step": 17225 }, { "epoch": 1.8675195143104943, "grad_norm": 0.09015129506587982, "learning_rate": 7.577252284926738e-06, "loss": 0.0012, "step": 17226 }, { "epoch": 1.867627927146574, "grad_norm": 4.554002285003662, "learning_rate": 7.576526911359351e-06, "loss": 0.03, "step": 17227 }, { "epoch": 1.8677363399826539, "grad_norm": 2.273486375808716, "learning_rate": 7.5758015377919634e-06, "loss": 0.0273, "step": 17228 }, { "epoch": 1.8678447528187339, "grad_norm": 0.7971234321594238, "learning_rate": 7.5750761642245764e-06, "loss": 0.0043, "step": 17229 }, { "epoch": 1.8679531656548134, "grad_norm": 7.263484477996826, "learning_rate": 7.574350790657189e-06, "loss": 0.0154, "step": 17230 }, { "epoch": 1.8680615784908934, "grad_norm": 1.4138250350952148, "learning_rate": 7.5736254170898016e-06, "loss": 0.0164, "step": 17231 }, { "epoch": 1.8681699913269731, "grad_norm": 2.9637510776519775, "learning_rate": 7.5729000435224146e-06, "loss": 0.031, "step": 17232 }, { "epoch": 1.868278404163053, "grad_norm": 0.5376128554344177, "learning_rate": 7.572174669955027e-06, "loss": 0.0118, "step": 17233 }, { "epoch": 1.8683868169991327, "grad_norm": 6.544656276702881, "learning_rate": 7.5714492963876405e-06, "loss": 0.0122, "step": 17234 }, { "epoch": 1.8684952298352124, "grad_norm": 1.6002681255340576, "learning_rate": 7.5707239228202535e-06, "loss": 0.0293, "step": 17235 }, { "epoch": 1.8686036426712924, "grad_norm": 0.0842636376619339, "learning_rate": 7.569998549252866e-06, "loss": 0.0009, "step": 17236 }, { "epoch": 1.868712055507372, "grad_norm": 1.5297960042953491, "learning_rate": 7.569273175685479e-06, "loss": 0.0244, "step": 17237 }, { "epoch": 1.868820468343452, "grad_norm": 1.0701805353164673, "learning_rate": 7.568547802118092e-06, "loss": 0.0174, "step": 17238 }, { "epoch": 1.8689288811795317, "grad_norm": 4.259766101837158, "learning_rate": 7.567822428550704e-06, "loss": 0.0685, "step": 17239 }, { "epoch": 1.8690372940156115, "grad_norm": 2.3959951400756836, "learning_rate": 7.567097054983317e-06, "loss": 0.0138, "step": 17240 }, { "epoch": 1.8691457068516912, "grad_norm": 0.13937947154045105, "learning_rate": 7.56637168141593e-06, "loss": 0.0014, "step": 17241 }, { "epoch": 1.869254119687771, "grad_norm": 0.4308558702468872, "learning_rate": 7.565646307848543e-06, "loss": 0.0134, "step": 17242 }, { "epoch": 1.869362532523851, "grad_norm": 1.5547226667404175, "learning_rate": 7.564920934281156e-06, "loss": 0.0373, "step": 17243 }, { "epoch": 1.8694709453599305, "grad_norm": 2.4393537044525146, "learning_rate": 7.564195560713769e-06, "loss": 0.0529, "step": 17244 }, { "epoch": 1.8695793581960105, "grad_norm": 1.3289728164672852, "learning_rate": 7.563470187146381e-06, "loss": 0.0309, "step": 17245 }, { "epoch": 1.86968777103209, "grad_norm": 3.6008124351501465, "learning_rate": 7.562744813578994e-06, "loss": 0.0196, "step": 17246 }, { "epoch": 1.86979618386817, "grad_norm": 0.539354681968689, "learning_rate": 7.562019440011607e-06, "loss": 0.0256, "step": 17247 }, { "epoch": 1.8699045967042498, "grad_norm": 0.7945030331611633, "learning_rate": 7.561294066444219e-06, "loss": 0.0222, "step": 17248 }, { "epoch": 1.8700130095403296, "grad_norm": 1.3967676162719727, "learning_rate": 7.560568692876832e-06, "loss": 0.0697, "step": 17249 }, { "epoch": 1.8701214223764093, "grad_norm": 1.5669245719909668, "learning_rate": 7.559843319309444e-06, "loss": 0.056, "step": 17250 }, { "epoch": 1.870229835212489, "grad_norm": 0.48290035128593445, "learning_rate": 7.559117945742058e-06, "loss": 0.0199, "step": 17251 }, { "epoch": 1.870338248048569, "grad_norm": 4.659204483032227, "learning_rate": 7.558392572174671e-06, "loss": 0.0418, "step": 17252 }, { "epoch": 1.8704466608846486, "grad_norm": 1.5992341041564941, "learning_rate": 7.557667198607284e-06, "loss": 0.0325, "step": 17253 }, { "epoch": 1.8705550737207286, "grad_norm": 1.004790186882019, "learning_rate": 7.556941825039896e-06, "loss": 0.0488, "step": 17254 }, { "epoch": 1.8706634865568084, "grad_norm": 2.9664535522460938, "learning_rate": 7.556216451472509e-06, "loss": 0.0278, "step": 17255 }, { "epoch": 1.8707718993928881, "grad_norm": 1.433631420135498, "learning_rate": 7.555491077905121e-06, "loss": 0.0347, "step": 17256 }, { "epoch": 1.870880312228968, "grad_norm": 0.34426286816596985, "learning_rate": 7.554765704337734e-06, "loss": 0.0033, "step": 17257 }, { "epoch": 1.8709887250650477, "grad_norm": 0.6591077446937561, "learning_rate": 7.554040330770347e-06, "loss": 0.0067, "step": 17258 }, { "epoch": 1.8710971379011276, "grad_norm": 2.031803846359253, "learning_rate": 7.553314957202959e-06, "loss": 0.0621, "step": 17259 }, { "epoch": 1.8712055507372072, "grad_norm": 1.662615418434143, "learning_rate": 7.552589583635573e-06, "loss": 0.0086, "step": 17260 }, { "epoch": 1.8713139635732872, "grad_norm": 3.995530843734741, "learning_rate": 7.551864210068186e-06, "loss": 0.0226, "step": 17261 }, { "epoch": 1.8714223764093667, "grad_norm": 2.7271900177001953, "learning_rate": 7.551138836500798e-06, "loss": 0.0363, "step": 17262 }, { "epoch": 1.8715307892454467, "grad_norm": 1.31010103225708, "learning_rate": 7.550413462933411e-06, "loss": 0.0063, "step": 17263 }, { "epoch": 1.8716392020815265, "grad_norm": 1.9382728338241577, "learning_rate": 7.549688089366024e-06, "loss": 0.0253, "step": 17264 }, { "epoch": 1.8717476149176062, "grad_norm": 3.8744258880615234, "learning_rate": 7.5489627157986365e-06, "loss": 0.0856, "step": 17265 }, { "epoch": 1.8718560277536862, "grad_norm": 0.34896671772003174, "learning_rate": 7.5482373422312495e-06, "loss": 0.0071, "step": 17266 }, { "epoch": 1.8719644405897657, "grad_norm": 3.0910274982452393, "learning_rate": 7.5475119686638624e-06, "loss": 0.0172, "step": 17267 }, { "epoch": 1.8720728534258457, "grad_norm": 0.4531962275505066, "learning_rate": 7.5467865950964754e-06, "loss": 0.0151, "step": 17268 }, { "epoch": 1.8721812662619253, "grad_norm": 0.9015445709228516, "learning_rate": 7.5460612215290884e-06, "loss": 0.0208, "step": 17269 }, { "epoch": 1.8722896790980053, "grad_norm": 0.2714443802833557, "learning_rate": 7.545335847961701e-06, "loss": 0.0071, "step": 17270 }, { "epoch": 1.872398091934085, "grad_norm": 2.026132583618164, "learning_rate": 7.5446104743943136e-06, "loss": 0.0138, "step": 17271 }, { "epoch": 1.8725065047701648, "grad_norm": 0.1580987572669983, "learning_rate": 7.5438851008269266e-06, "loss": 0.0021, "step": 17272 }, { "epoch": 1.8726149176062445, "grad_norm": 2.8887953758239746, "learning_rate": 7.5431597272595395e-06, "loss": 0.0262, "step": 17273 }, { "epoch": 1.8727233304423243, "grad_norm": 1.3161418437957764, "learning_rate": 7.542434353692152e-06, "loss": 0.0078, "step": 17274 }, { "epoch": 1.8728317432784043, "grad_norm": 3.7470617294311523, "learning_rate": 7.541708980124765e-06, "loss": 0.0183, "step": 17275 }, { "epoch": 1.8729401561144838, "grad_norm": 2.375523805618286, "learning_rate": 7.540983606557377e-06, "loss": 0.0125, "step": 17276 }, { "epoch": 1.8730485689505638, "grad_norm": 3.9297449588775635, "learning_rate": 7.540258232989991e-06, "loss": 0.0185, "step": 17277 }, { "epoch": 1.8731569817866436, "grad_norm": 0.4172329008579254, "learning_rate": 7.539532859422604e-06, "loss": 0.0048, "step": 17278 }, { "epoch": 1.8732653946227233, "grad_norm": 3.6752500534057617, "learning_rate": 7.538807485855217e-06, "loss": 0.0148, "step": 17279 }, { "epoch": 1.873373807458803, "grad_norm": 0.08046703785657883, "learning_rate": 7.538082112287829e-06, "loss": 0.0011, "step": 17280 }, { "epoch": 1.8734822202948829, "grad_norm": 1.7238059043884277, "learning_rate": 7.537356738720442e-06, "loss": 0.0505, "step": 17281 }, { "epoch": 1.8735906331309629, "grad_norm": 2.8925182819366455, "learning_rate": 7.536631365153054e-06, "loss": 0.0333, "step": 17282 }, { "epoch": 1.8736990459670424, "grad_norm": 0.11281169950962067, "learning_rate": 7.535905991585667e-06, "loss": 0.0008, "step": 17283 }, { "epoch": 1.8738074588031224, "grad_norm": 2.819385051727295, "learning_rate": 7.53518061801828e-06, "loss": 0.0223, "step": 17284 }, { "epoch": 1.873915871639202, "grad_norm": 0.10198016464710236, "learning_rate": 7.534455244450894e-06, "loss": 0.001, "step": 17285 }, { "epoch": 1.874024284475282, "grad_norm": 0.3578528165817261, "learning_rate": 7.533729870883506e-06, "loss": 0.0129, "step": 17286 }, { "epoch": 1.8741326973113617, "grad_norm": 2.4982895851135254, "learning_rate": 7.533004497316119e-06, "loss": 0.0217, "step": 17287 }, { "epoch": 1.8742411101474414, "grad_norm": 4.252399444580078, "learning_rate": 7.532279123748731e-06, "loss": 0.0554, "step": 17288 }, { "epoch": 1.8743495229835212, "grad_norm": 0.3247673809528351, "learning_rate": 7.531553750181344e-06, "loss": 0.0061, "step": 17289 }, { "epoch": 1.874457935819601, "grad_norm": 0.38027501106262207, "learning_rate": 7.530828376613957e-06, "loss": 0.0172, "step": 17290 }, { "epoch": 1.874566348655681, "grad_norm": 0.6149821281433105, "learning_rate": 7.530103003046569e-06, "loss": 0.0039, "step": 17291 }, { "epoch": 1.8746747614917605, "grad_norm": 1.629131555557251, "learning_rate": 7.529377629479182e-06, "loss": 0.0261, "step": 17292 }, { "epoch": 1.8747831743278405, "grad_norm": 2.2303543090820312, "learning_rate": 7.528652255911795e-06, "loss": 0.0515, "step": 17293 }, { "epoch": 1.8748915871639202, "grad_norm": 0.5735266804695129, "learning_rate": 7.527926882344408e-06, "loss": 0.008, "step": 17294 }, { "epoch": 1.875, "grad_norm": 0.5161371827125549, "learning_rate": 7.527201508777021e-06, "loss": 0.0055, "step": 17295 }, { "epoch": 1.8751084128360798, "grad_norm": 1.876737356185913, "learning_rate": 7.526476135209634e-06, "loss": 0.0231, "step": 17296 }, { "epoch": 1.8752168256721595, "grad_norm": 3.671755790710449, "learning_rate": 7.525750761642246e-06, "loss": 0.0344, "step": 17297 }, { "epoch": 1.8753252385082395, "grad_norm": 0.4047134816646576, "learning_rate": 7.525025388074859e-06, "loss": 0.0112, "step": 17298 }, { "epoch": 1.875433651344319, "grad_norm": 3.4503839015960693, "learning_rate": 7.524300014507472e-06, "loss": 0.0676, "step": 17299 }, { "epoch": 1.875542064180399, "grad_norm": 0.3485827147960663, "learning_rate": 7.523574640940084e-06, "loss": 0.0063, "step": 17300 }, { "epoch": 1.8756504770164788, "grad_norm": 1.7350128889083862, "learning_rate": 7.522849267372697e-06, "loss": 0.0088, "step": 17301 }, { "epoch": 1.8757588898525586, "grad_norm": 1.747628927230835, "learning_rate": 7.5221238938053095e-06, "loss": 0.0313, "step": 17302 }, { "epoch": 1.8758673026886383, "grad_norm": 1.217864751815796, "learning_rate": 7.521398520237923e-06, "loss": 0.0274, "step": 17303 }, { "epoch": 1.875975715524718, "grad_norm": 2.727678060531616, "learning_rate": 7.520673146670536e-06, "loss": 0.0455, "step": 17304 }, { "epoch": 1.876084128360798, "grad_norm": 3.0619516372680664, "learning_rate": 7.519947773103149e-06, "loss": 0.0224, "step": 17305 }, { "epoch": 1.8761925411968776, "grad_norm": 4.55052375793457, "learning_rate": 7.5192223995357614e-06, "loss": 0.0068, "step": 17306 }, { "epoch": 1.8763009540329576, "grad_norm": 0.41813865303993225, "learning_rate": 7.5184970259683744e-06, "loss": 0.004, "step": 17307 }, { "epoch": 1.8764093668690371, "grad_norm": 1.08945631980896, "learning_rate": 7.517771652400987e-06, "loss": 0.0225, "step": 17308 }, { "epoch": 1.8765177797051171, "grad_norm": 2.2120542526245117, "learning_rate": 7.5170462788335996e-06, "loss": 0.0295, "step": 17309 }, { "epoch": 1.876626192541197, "grad_norm": 0.43973761796951294, "learning_rate": 7.5163209052662126e-06, "loss": 0.0047, "step": 17310 }, { "epoch": 1.8767346053772767, "grad_norm": 2.0062685012817383, "learning_rate": 7.515595531698826e-06, "loss": 0.0404, "step": 17311 }, { "epoch": 1.8768430182133564, "grad_norm": 3.6748642921447754, "learning_rate": 7.5148701581314385e-06, "loss": 0.042, "step": 17312 }, { "epoch": 1.8769514310494362, "grad_norm": 7.162006378173828, "learning_rate": 7.5141447845640515e-06, "loss": 0.0375, "step": 17313 }, { "epoch": 1.8770598438855162, "grad_norm": 2.8534739017486572, "learning_rate": 7.513419410996664e-06, "loss": 0.0333, "step": 17314 }, { "epoch": 1.8771682567215957, "grad_norm": 0.4150286614894867, "learning_rate": 7.512694037429277e-06, "loss": 0.0245, "step": 17315 }, { "epoch": 1.8772766695576757, "grad_norm": 0.8494009971618652, "learning_rate": 7.51196866386189e-06, "loss": 0.0175, "step": 17316 }, { "epoch": 1.8773850823937555, "grad_norm": 4.4918928146362305, "learning_rate": 7.511243290294502e-06, "loss": 0.0324, "step": 17317 }, { "epoch": 1.8774934952298352, "grad_norm": 4.731874465942383, "learning_rate": 7.510517916727115e-06, "loss": 0.0432, "step": 17318 }, { "epoch": 1.877601908065915, "grad_norm": 0.8500758409500122, "learning_rate": 7.509792543159728e-06, "loss": 0.0377, "step": 17319 }, { "epoch": 1.8777103209019947, "grad_norm": 0.035090502351522446, "learning_rate": 7.509067169592341e-06, "loss": 0.0004, "step": 17320 }, { "epoch": 1.8778187337380747, "grad_norm": 5.473740577697754, "learning_rate": 7.508341796024954e-06, "loss": 0.0466, "step": 17321 }, { "epoch": 1.8779271465741543, "grad_norm": 2.0268123149871826, "learning_rate": 7.507616422457567e-06, "loss": 0.0465, "step": 17322 }, { "epoch": 1.8780355594102343, "grad_norm": 1.7630033493041992, "learning_rate": 7.506891048890179e-06, "loss": 0.0273, "step": 17323 }, { "epoch": 1.8781439722463138, "grad_norm": 0.31374409794807434, "learning_rate": 7.506165675322792e-06, "loss": 0.0138, "step": 17324 }, { "epoch": 1.8782523850823938, "grad_norm": 2.523392915725708, "learning_rate": 7.505440301755405e-06, "loss": 0.0133, "step": 17325 }, { "epoch": 1.8783607979184735, "grad_norm": 7.88416862487793, "learning_rate": 7.504714928188017e-06, "loss": 0.0321, "step": 17326 }, { "epoch": 1.8784692107545533, "grad_norm": 0.582746148109436, "learning_rate": 7.50398955462063e-06, "loss": 0.0044, "step": 17327 }, { "epoch": 1.8785776235906333, "grad_norm": 1.5935485363006592, "learning_rate": 7.503264181053242e-06, "loss": 0.0142, "step": 17328 }, { "epoch": 1.8786860364267128, "grad_norm": 0.7620798945426941, "learning_rate": 7.502538807485856e-06, "loss": 0.0208, "step": 17329 }, { "epoch": 1.8787944492627928, "grad_norm": 0.5493718385696411, "learning_rate": 7.501813433918469e-06, "loss": 0.0044, "step": 17330 }, { "epoch": 1.8789028620988724, "grad_norm": 0.05435894429683685, "learning_rate": 7.501088060351082e-06, "loss": 0.0007, "step": 17331 }, { "epoch": 1.8790112749349523, "grad_norm": 0.8389462232589722, "learning_rate": 7.500362686783694e-06, "loss": 0.0022, "step": 17332 }, { "epoch": 1.879119687771032, "grad_norm": 2.3500044345855713, "learning_rate": 7.499637313216307e-06, "loss": 0.0227, "step": 17333 }, { "epoch": 1.8792281006071119, "grad_norm": 7.486987113952637, "learning_rate": 7.498911939648919e-06, "loss": 0.0205, "step": 17334 }, { "epoch": 1.8793365134431916, "grad_norm": 0.11651443690061569, "learning_rate": 7.498186566081532e-06, "loss": 0.0012, "step": 17335 }, { "epoch": 1.8794449262792714, "grad_norm": 2.135338306427002, "learning_rate": 7.497461192514145e-06, "loss": 0.0058, "step": 17336 }, { "epoch": 1.8795533391153514, "grad_norm": 1.6092313528060913, "learning_rate": 7.496735818946759e-06, "loss": 0.065, "step": 17337 }, { "epoch": 1.879661751951431, "grad_norm": 3.7716751098632812, "learning_rate": 7.496010445379371e-06, "loss": 0.0076, "step": 17338 }, { "epoch": 1.879770164787511, "grad_norm": 3.5104925632476807, "learning_rate": 7.495285071811984e-06, "loss": 0.0728, "step": 17339 }, { "epoch": 1.8798785776235907, "grad_norm": 0.059001434594392776, "learning_rate": 7.494559698244596e-06, "loss": 0.0006, "step": 17340 }, { "epoch": 1.8799869904596704, "grad_norm": 0.8686181306838989, "learning_rate": 7.493834324677209e-06, "loss": 0.0252, "step": 17341 }, { "epoch": 1.8800954032957502, "grad_norm": 2.096755266189575, "learning_rate": 7.493108951109822e-06, "loss": 0.032, "step": 17342 }, { "epoch": 1.88020381613183, "grad_norm": 0.47348493337631226, "learning_rate": 7.4923835775424345e-06, "loss": 0.0215, "step": 17343 }, { "epoch": 1.88031222896791, "grad_norm": 0.9643246531486511, "learning_rate": 7.4916582039750475e-06, "loss": 0.0476, "step": 17344 }, { "epoch": 1.8804206418039895, "grad_norm": 0.5442063212394714, "learning_rate": 7.4909328304076605e-06, "loss": 0.0203, "step": 17345 }, { "epoch": 1.8805290546400695, "grad_norm": 0.4569055438041687, "learning_rate": 7.4902074568402734e-06, "loss": 0.0045, "step": 17346 }, { "epoch": 1.880637467476149, "grad_norm": 0.0690959095954895, "learning_rate": 7.4894820832728864e-06, "loss": 0.0006, "step": 17347 }, { "epoch": 1.880745880312229, "grad_norm": 5.236357688903809, "learning_rate": 7.4887567097054994e-06, "loss": 0.0493, "step": 17348 }, { "epoch": 1.8808542931483088, "grad_norm": 5.371456623077393, "learning_rate": 7.4880313361381116e-06, "loss": 0.0401, "step": 17349 }, { "epoch": 1.8809627059843885, "grad_norm": 2.475891590118408, "learning_rate": 7.4873059625707246e-06, "loss": 0.029, "step": 17350 }, { "epoch": 1.8810711188204683, "grad_norm": 2.8991668224334717, "learning_rate": 7.4865805890033375e-06, "loss": 0.059, "step": 17351 }, { "epoch": 1.881179531656548, "grad_norm": 0.8993618488311768, "learning_rate": 7.48585521543595e-06, "loss": 0.0108, "step": 17352 }, { "epoch": 1.881287944492628, "grad_norm": 0.4249767065048218, "learning_rate": 7.485129841868563e-06, "loss": 0.0023, "step": 17353 }, { "epoch": 1.8813963573287076, "grad_norm": 0.6386200785636902, "learning_rate": 7.484404468301175e-06, "loss": 0.0042, "step": 17354 }, { "epoch": 1.8815047701647876, "grad_norm": 2.123076915740967, "learning_rate": 7.483679094733789e-06, "loss": 0.0233, "step": 17355 }, { "epoch": 1.8816131830008673, "grad_norm": 2.1116156578063965, "learning_rate": 7.482953721166402e-06, "loss": 0.0457, "step": 17356 }, { "epoch": 1.881721595836947, "grad_norm": 0.06831014156341553, "learning_rate": 7.482228347599015e-06, "loss": 0.0009, "step": 17357 }, { "epoch": 1.8818300086730269, "grad_norm": 1.5106109380722046, "learning_rate": 7.481502974031627e-06, "loss": 0.0036, "step": 17358 }, { "epoch": 1.8819384215091066, "grad_norm": 0.42425915598869324, "learning_rate": 7.48077760046424e-06, "loss": 0.0143, "step": 17359 }, { "epoch": 1.8820468343451866, "grad_norm": 0.611132800579071, "learning_rate": 7.480052226896852e-06, "loss": 0.0257, "step": 17360 }, { "epoch": 1.8821552471812661, "grad_norm": 1.8244010210037231, "learning_rate": 7.479326853329465e-06, "loss": 0.0159, "step": 17361 }, { "epoch": 1.8822636600173461, "grad_norm": 3.111778736114502, "learning_rate": 7.478601479762078e-06, "loss": 0.031, "step": 17362 }, { "epoch": 1.882372072853426, "grad_norm": 1.3581284284591675, "learning_rate": 7.477876106194692e-06, "loss": 0.0097, "step": 17363 }, { "epoch": 1.8824804856895057, "grad_norm": 2.0580081939697266, "learning_rate": 7.477150732627304e-06, "loss": 0.0575, "step": 17364 }, { "epoch": 1.8825888985255854, "grad_norm": 1.6137394905090332, "learning_rate": 7.476425359059917e-06, "loss": 0.0096, "step": 17365 }, { "epoch": 1.8826973113616652, "grad_norm": 3.13843035697937, "learning_rate": 7.475699985492529e-06, "loss": 0.0255, "step": 17366 }, { "epoch": 1.8828057241977452, "grad_norm": 0.3520061671733856, "learning_rate": 7.474974611925142e-06, "loss": 0.0038, "step": 17367 }, { "epoch": 1.8829141370338247, "grad_norm": 3.4419429302215576, "learning_rate": 7.474249238357755e-06, "loss": 0.0299, "step": 17368 }, { "epoch": 1.8830225498699047, "grad_norm": 8.662712097167969, "learning_rate": 7.473523864790367e-06, "loss": 0.0197, "step": 17369 }, { "epoch": 1.8831309627059842, "grad_norm": 3.085822582244873, "learning_rate": 7.47279849122298e-06, "loss": 0.013, "step": 17370 }, { "epoch": 1.8832393755420642, "grad_norm": 0.07241560518741608, "learning_rate": 7.472073117655593e-06, "loss": 0.0009, "step": 17371 }, { "epoch": 1.883347788378144, "grad_norm": 0.2307482659816742, "learning_rate": 7.471347744088206e-06, "loss": 0.0022, "step": 17372 }, { "epoch": 1.8834562012142237, "grad_norm": 1.3502520322799683, "learning_rate": 7.470622370520819e-06, "loss": 0.0227, "step": 17373 }, { "epoch": 1.8835646140503035, "grad_norm": 0.05260458588600159, "learning_rate": 7.469896996953432e-06, "loss": 0.0006, "step": 17374 }, { "epoch": 1.8836730268863833, "grad_norm": 1.758176326751709, "learning_rate": 7.469171623386044e-06, "loss": 0.0048, "step": 17375 }, { "epoch": 1.8837814397224633, "grad_norm": 2.032276153564453, "learning_rate": 7.468446249818657e-06, "loss": 0.0275, "step": 17376 }, { "epoch": 1.8838898525585428, "grad_norm": 3.7198500633239746, "learning_rate": 7.46772087625127e-06, "loss": 0.0238, "step": 17377 }, { "epoch": 1.8839982653946228, "grad_norm": 0.20802535116672516, "learning_rate": 7.466995502683882e-06, "loss": 0.0012, "step": 17378 }, { "epoch": 1.8841066782307025, "grad_norm": 1.995954155921936, "learning_rate": 7.466270129116495e-06, "loss": 0.0276, "step": 17379 }, { "epoch": 1.8842150910667823, "grad_norm": 3.4616854190826416, "learning_rate": 7.4655447555491075e-06, "loss": 0.021, "step": 17380 }, { "epoch": 1.884323503902862, "grad_norm": 0.16535137593746185, "learning_rate": 7.464819381981721e-06, "loss": 0.0021, "step": 17381 }, { "epoch": 1.8844319167389418, "grad_norm": 2.2226295471191406, "learning_rate": 7.464094008414334e-06, "loss": 0.0146, "step": 17382 }, { "epoch": 1.8845403295750218, "grad_norm": 0.6134275197982788, "learning_rate": 7.463368634846947e-06, "loss": 0.0153, "step": 17383 }, { "epoch": 1.8846487424111014, "grad_norm": 1.2856042385101318, "learning_rate": 7.4626432612795595e-06, "loss": 0.0033, "step": 17384 }, { "epoch": 1.8847571552471813, "grad_norm": 0.858081579208374, "learning_rate": 7.4619178877121724e-06, "loss": 0.0099, "step": 17385 }, { "epoch": 1.8848655680832611, "grad_norm": 2.0801541805267334, "learning_rate": 7.461192514144785e-06, "loss": 0.0204, "step": 17386 }, { "epoch": 1.8849739809193409, "grad_norm": 0.40290868282318115, "learning_rate": 7.460467140577398e-06, "loss": 0.01, "step": 17387 }, { "epoch": 1.8850823937554206, "grad_norm": 2.567955732345581, "learning_rate": 7.4597417670100106e-06, "loss": 0.0346, "step": 17388 }, { "epoch": 1.8851908065915004, "grad_norm": 1.561455249786377, "learning_rate": 7.459016393442624e-06, "loss": 0.0167, "step": 17389 }, { "epoch": 1.8852992194275804, "grad_norm": 2.814411163330078, "learning_rate": 7.4582910198752366e-06, "loss": 0.0179, "step": 17390 }, { "epoch": 1.88540763226366, "grad_norm": 2.7825803756713867, "learning_rate": 7.4575656463078495e-06, "loss": 0.0063, "step": 17391 }, { "epoch": 1.88551604509974, "grad_norm": 1.122225046157837, "learning_rate": 7.456840272740462e-06, "loss": 0.0273, "step": 17392 }, { "epoch": 1.8856244579358195, "grad_norm": 0.2703920602798462, "learning_rate": 7.456114899173075e-06, "loss": 0.0016, "step": 17393 }, { "epoch": 1.8857328707718994, "grad_norm": 1.787561058998108, "learning_rate": 7.455389525605688e-06, "loss": 0.0622, "step": 17394 }, { "epoch": 1.8858412836079792, "grad_norm": 0.6500192284584045, "learning_rate": 7.4546641520383e-06, "loss": 0.0116, "step": 17395 }, { "epoch": 1.885949696444059, "grad_norm": 3.962284803390503, "learning_rate": 7.453938778470913e-06, "loss": 0.0401, "step": 17396 }, { "epoch": 1.8860581092801387, "grad_norm": 0.5108076930046082, "learning_rate": 7.453213404903526e-06, "loss": 0.015, "step": 17397 }, { "epoch": 1.8861665221162185, "grad_norm": 1.305304765701294, "learning_rate": 7.452488031336139e-06, "loss": 0.0324, "step": 17398 }, { "epoch": 1.8862749349522985, "grad_norm": 5.219715118408203, "learning_rate": 7.451762657768752e-06, "loss": 0.0192, "step": 17399 }, { "epoch": 1.886383347788378, "grad_norm": 12.725650787353516, "learning_rate": 7.451037284201365e-06, "loss": 0.0521, "step": 17400 }, { "epoch": 1.886491760624458, "grad_norm": 1.3012820482254028, "learning_rate": 7.450311910633977e-06, "loss": 0.0144, "step": 17401 }, { "epoch": 1.8866001734605378, "grad_norm": 0.18251045048236847, "learning_rate": 7.44958653706659e-06, "loss": 0.0013, "step": 17402 }, { "epoch": 1.8867085862966175, "grad_norm": 3.082393169403076, "learning_rate": 7.448861163499203e-06, "loss": 0.0358, "step": 17403 }, { "epoch": 1.8868169991326973, "grad_norm": 0.5580188632011414, "learning_rate": 7.448135789931815e-06, "loss": 0.0189, "step": 17404 }, { "epoch": 1.886925411968777, "grad_norm": 0.4100677967071533, "learning_rate": 7.447410416364428e-06, "loss": 0.0197, "step": 17405 }, { "epoch": 1.887033824804857, "grad_norm": 2.6508138179779053, "learning_rate": 7.446685042797042e-06, "loss": 0.0021, "step": 17406 }, { "epoch": 1.8871422376409366, "grad_norm": 0.38032975792884827, "learning_rate": 7.445959669229654e-06, "loss": 0.003, "step": 17407 }, { "epoch": 1.8872506504770166, "grad_norm": 1.992750644683838, "learning_rate": 7.445234295662267e-06, "loss": 0.0198, "step": 17408 }, { "epoch": 1.887359063313096, "grad_norm": 1.0953388214111328, "learning_rate": 7.44450892209488e-06, "loss": 0.0166, "step": 17409 }, { "epoch": 1.887467476149176, "grad_norm": 2.011916399002075, "learning_rate": 7.443783548527492e-06, "loss": 0.0159, "step": 17410 }, { "epoch": 1.8875758889852559, "grad_norm": 0.4633195996284485, "learning_rate": 7.443058174960105e-06, "loss": 0.0188, "step": 17411 }, { "epoch": 1.8876843018213356, "grad_norm": 2.8644766807556152, "learning_rate": 7.442332801392717e-06, "loss": 0.0694, "step": 17412 }, { "epoch": 1.8877927146574154, "grad_norm": 3.347245454788208, "learning_rate": 7.44160742782533e-06, "loss": 0.0639, "step": 17413 }, { "epoch": 1.8879011274934951, "grad_norm": 0.02266843616962433, "learning_rate": 7.440882054257943e-06, "loss": 0.0002, "step": 17414 }, { "epoch": 1.8880095403295751, "grad_norm": 1.528092622756958, "learning_rate": 7.440156680690557e-06, "loss": 0.0089, "step": 17415 }, { "epoch": 1.8881179531656547, "grad_norm": 8.124428749084473, "learning_rate": 7.439431307123169e-06, "loss": 0.0207, "step": 17416 }, { "epoch": 1.8882263660017347, "grad_norm": 1.7941854000091553, "learning_rate": 7.438705933555782e-06, "loss": 0.0364, "step": 17417 }, { "epoch": 1.8883347788378144, "grad_norm": 0.963239848613739, "learning_rate": 7.437980559988394e-06, "loss": 0.0207, "step": 17418 }, { "epoch": 1.8884431916738942, "grad_norm": 0.010005644522607327, "learning_rate": 7.437255186421007e-06, "loss": 0.0001, "step": 17419 }, { "epoch": 1.888551604509974, "grad_norm": 3.630415678024292, "learning_rate": 7.43652981285362e-06, "loss": 0.0486, "step": 17420 }, { "epoch": 1.8886600173460537, "grad_norm": 1.3778467178344727, "learning_rate": 7.4358044392862325e-06, "loss": 0.0232, "step": 17421 }, { "epoch": 1.8887684301821337, "grad_norm": 0.7361356616020203, "learning_rate": 7.4350790657188455e-06, "loss": 0.0191, "step": 17422 }, { "epoch": 1.8888768430182132, "grad_norm": 1.315354824066162, "learning_rate": 7.4343536921514585e-06, "loss": 0.0208, "step": 17423 }, { "epoch": 1.8889852558542932, "grad_norm": 0.39216479659080505, "learning_rate": 7.4336283185840714e-06, "loss": 0.0131, "step": 17424 }, { "epoch": 1.889093668690373, "grad_norm": 7.234725475311279, "learning_rate": 7.4329029450166844e-06, "loss": 0.0383, "step": 17425 }, { "epoch": 1.8892020815264527, "grad_norm": 1.605044960975647, "learning_rate": 7.4321775714492974e-06, "loss": 0.0114, "step": 17426 }, { "epoch": 1.8893104943625325, "grad_norm": 0.1506781429052353, "learning_rate": 7.4314521978819096e-06, "loss": 0.0007, "step": 17427 }, { "epoch": 1.8894189071986123, "grad_norm": 2.5349230766296387, "learning_rate": 7.4307268243145226e-06, "loss": 0.0457, "step": 17428 }, { "epoch": 1.8895273200346923, "grad_norm": 0.692786693572998, "learning_rate": 7.4300014507471356e-06, "loss": 0.0228, "step": 17429 }, { "epoch": 1.8896357328707718, "grad_norm": 0.3423049747943878, "learning_rate": 7.429276077179748e-06, "loss": 0.0031, "step": 17430 }, { "epoch": 1.8897441457068518, "grad_norm": 2.689328908920288, "learning_rate": 7.428550703612361e-06, "loss": 0.0395, "step": 17431 }, { "epoch": 1.8898525585429313, "grad_norm": 5.122376441955566, "learning_rate": 7.4278253300449745e-06, "loss": 0.0455, "step": 17432 }, { "epoch": 1.8899609713790113, "grad_norm": 1.2903697490692139, "learning_rate": 7.427099956477587e-06, "loss": 0.0016, "step": 17433 }, { "epoch": 1.890069384215091, "grad_norm": 1.4923691749572754, "learning_rate": 7.4263745829102e-06, "loss": 0.0211, "step": 17434 }, { "epoch": 1.8901777970511708, "grad_norm": 7.27031135559082, "learning_rate": 7.425649209342813e-06, "loss": 0.0705, "step": 17435 }, { "epoch": 1.8902862098872506, "grad_norm": 1.3029615879058838, "learning_rate": 7.424923835775425e-06, "loss": 0.0148, "step": 17436 }, { "epoch": 1.8903946227233304, "grad_norm": 1.2019013166427612, "learning_rate": 7.424198462208038e-06, "loss": 0.0454, "step": 17437 }, { "epoch": 1.8905030355594103, "grad_norm": 5.521629333496094, "learning_rate": 7.42347308864065e-06, "loss": 0.0273, "step": 17438 }, { "epoch": 1.89061144839549, "grad_norm": 2.1829476356506348, "learning_rate": 7.422747715073263e-06, "loss": 0.0261, "step": 17439 }, { "epoch": 1.8907198612315699, "grad_norm": 0.3563210964202881, "learning_rate": 7.422022341505876e-06, "loss": 0.0085, "step": 17440 }, { "epoch": 1.8908282740676496, "grad_norm": 9.424786567687988, "learning_rate": 7.42129696793849e-06, "loss": 0.0214, "step": 17441 }, { "epoch": 1.8909366869037294, "grad_norm": 1.7428816556930542, "learning_rate": 7.420571594371102e-06, "loss": 0.0132, "step": 17442 }, { "epoch": 1.8910450997398092, "grad_norm": 2.530000925064087, "learning_rate": 7.419846220803715e-06, "loss": 0.0338, "step": 17443 }, { "epoch": 1.891153512575889, "grad_norm": 6.674912452697754, "learning_rate": 7.419120847236327e-06, "loss": 0.032, "step": 17444 }, { "epoch": 1.891261925411969, "grad_norm": 0.054487500339746475, "learning_rate": 7.41839547366894e-06, "loss": 0.0005, "step": 17445 }, { "epoch": 1.8913703382480485, "grad_norm": 0.05107206478714943, "learning_rate": 7.417670100101553e-06, "loss": 0.0005, "step": 17446 }, { "epoch": 1.8914787510841284, "grad_norm": 0.36251574754714966, "learning_rate": 7.416944726534165e-06, "loss": 0.0035, "step": 17447 }, { "epoch": 1.8915871639202082, "grad_norm": 2.921656608581543, "learning_rate": 7.416219352966778e-06, "loss": 0.0372, "step": 17448 }, { "epoch": 1.891695576756288, "grad_norm": 6.0704545974731445, "learning_rate": 7.415493979399391e-06, "loss": 0.0649, "step": 17449 }, { "epoch": 1.8918039895923677, "grad_norm": 0.47330906987190247, "learning_rate": 7.414768605832004e-06, "loss": 0.0031, "step": 17450 }, { "epoch": 1.8919124024284475, "grad_norm": 0.6417796015739441, "learning_rate": 7.414043232264617e-06, "loss": 0.0288, "step": 17451 }, { "epoch": 1.8920208152645275, "grad_norm": 2.413393974304199, "learning_rate": 7.41331785869723e-06, "loss": 0.0145, "step": 17452 }, { "epoch": 1.892129228100607, "grad_norm": 1.5012842416763306, "learning_rate": 7.412592485129842e-06, "loss": 0.0391, "step": 17453 }, { "epoch": 1.892237640936687, "grad_norm": 0.3695718050003052, "learning_rate": 7.411867111562455e-06, "loss": 0.0172, "step": 17454 }, { "epoch": 1.8923460537727665, "grad_norm": 0.05482667684555054, "learning_rate": 7.411141737995068e-06, "loss": 0.0005, "step": 17455 }, { "epoch": 1.8924544666088465, "grad_norm": 1.848348617553711, "learning_rate": 7.41041636442768e-06, "loss": 0.026, "step": 17456 }, { "epoch": 1.8925628794449263, "grad_norm": 0.5336444973945618, "learning_rate": 7.409690990860293e-06, "loss": 0.0268, "step": 17457 }, { "epoch": 1.892671292281006, "grad_norm": 1.4122540950775146, "learning_rate": 7.408965617292907e-06, "loss": 0.0566, "step": 17458 }, { "epoch": 1.8927797051170858, "grad_norm": 0.6913909316062927, "learning_rate": 7.408240243725519e-06, "loss": 0.0131, "step": 17459 }, { "epoch": 1.8928881179531656, "grad_norm": 3.4613521099090576, "learning_rate": 7.407514870158132e-06, "loss": 0.0234, "step": 17460 }, { "epoch": 1.8929965307892456, "grad_norm": 2.632033586502075, "learning_rate": 7.406789496590745e-06, "loss": 0.0373, "step": 17461 }, { "epoch": 1.893104943625325, "grad_norm": 19.15367889404297, "learning_rate": 7.4060641230233575e-06, "loss": 0.0695, "step": 17462 }, { "epoch": 1.893213356461405, "grad_norm": 3.286661386489868, "learning_rate": 7.4053387494559705e-06, "loss": 0.0455, "step": 17463 }, { "epoch": 1.8933217692974849, "grad_norm": 3.4388105869293213, "learning_rate": 7.404613375888583e-06, "loss": 0.038, "step": 17464 }, { "epoch": 1.8934301821335646, "grad_norm": 3.974011182785034, "learning_rate": 7.403888002321196e-06, "loss": 0.0236, "step": 17465 }, { "epoch": 1.8935385949696444, "grad_norm": 1.8440085649490356, "learning_rate": 7.403162628753809e-06, "loss": 0.0224, "step": 17466 }, { "epoch": 1.8936470078057241, "grad_norm": 1.763463020324707, "learning_rate": 7.402437255186422e-06, "loss": 0.0289, "step": 17467 }, { "epoch": 1.8937554206418041, "grad_norm": 4.858684539794922, "learning_rate": 7.4017118816190346e-06, "loss": 0.0084, "step": 17468 }, { "epoch": 1.8938638334778837, "grad_norm": 0.9195191860198975, "learning_rate": 7.4009865080516475e-06, "loss": 0.0178, "step": 17469 }, { "epoch": 1.8939722463139637, "grad_norm": 0.4672034978866577, "learning_rate": 7.40026113448426e-06, "loss": 0.0191, "step": 17470 }, { "epoch": 1.8940806591500432, "grad_norm": 9.881436347961426, "learning_rate": 7.399535760916873e-06, "loss": 0.0351, "step": 17471 }, { "epoch": 1.8941890719861232, "grad_norm": 0.5589181780815125, "learning_rate": 7.398810387349486e-06, "loss": 0.0044, "step": 17472 }, { "epoch": 1.894297484822203, "grad_norm": 0.4679892063140869, "learning_rate": 7.398085013782098e-06, "loss": 0.0047, "step": 17473 }, { "epoch": 1.8944058976582827, "grad_norm": 1.3160889148712158, "learning_rate": 7.397359640214711e-06, "loss": 0.0072, "step": 17474 }, { "epoch": 1.8945143104943627, "grad_norm": 0.05601911619305611, "learning_rate": 7.396634266647324e-06, "loss": 0.0008, "step": 17475 }, { "epoch": 1.8946227233304422, "grad_norm": 0.22089441120624542, "learning_rate": 7.395908893079937e-06, "loss": 0.0028, "step": 17476 }, { "epoch": 1.8947311361665222, "grad_norm": 7.202025890350342, "learning_rate": 7.39518351951255e-06, "loss": 0.0157, "step": 17477 }, { "epoch": 1.8948395490026018, "grad_norm": 2.5836000442504883, "learning_rate": 7.394458145945163e-06, "loss": 0.0152, "step": 17478 }, { "epoch": 1.8949479618386817, "grad_norm": 1.3322162628173828, "learning_rate": 7.393732772377775e-06, "loss": 0.0044, "step": 17479 }, { "epoch": 1.8950563746747615, "grad_norm": 0.628828763961792, "learning_rate": 7.393007398810388e-06, "loss": 0.0078, "step": 17480 }, { "epoch": 1.8951647875108413, "grad_norm": 0.8096179962158203, "learning_rate": 7.392282025243001e-06, "loss": 0.0145, "step": 17481 }, { "epoch": 1.895273200346921, "grad_norm": 2.95802640914917, "learning_rate": 7.391556651675613e-06, "loss": 0.0172, "step": 17482 }, { "epoch": 1.8953816131830008, "grad_norm": 1.3368834257125854, "learning_rate": 7.390831278108226e-06, "loss": 0.0345, "step": 17483 }, { "epoch": 1.8954900260190808, "grad_norm": 0.4029300808906555, "learning_rate": 7.39010590454084e-06, "loss": 0.0046, "step": 17484 }, { "epoch": 1.8955984388551603, "grad_norm": 1.128838062286377, "learning_rate": 7.389380530973452e-06, "loss": 0.0216, "step": 17485 }, { "epoch": 1.8957068516912403, "grad_norm": 1.2124698162078857, "learning_rate": 7.388655157406065e-06, "loss": 0.0295, "step": 17486 }, { "epoch": 1.89581526452732, "grad_norm": 3.1067545413970947, "learning_rate": 7.387929783838678e-06, "loss": 0.0222, "step": 17487 }, { "epoch": 1.8959236773633998, "grad_norm": 1.6881922483444214, "learning_rate": 7.38720441027129e-06, "loss": 0.0308, "step": 17488 }, { "epoch": 1.8960320901994796, "grad_norm": 6.139901638031006, "learning_rate": 7.386479036703903e-06, "loss": 0.0134, "step": 17489 }, { "epoch": 1.8961405030355594, "grad_norm": 2.533815860748291, "learning_rate": 7.385753663136515e-06, "loss": 0.031, "step": 17490 }, { "epoch": 1.8962489158716394, "grad_norm": 0.7340039014816284, "learning_rate": 7.385028289569128e-06, "loss": 0.0115, "step": 17491 }, { "epoch": 1.896357328707719, "grad_norm": 5.805150032043457, "learning_rate": 7.384302916001741e-06, "loss": 0.0141, "step": 17492 }, { "epoch": 1.8964657415437989, "grad_norm": 5.078738689422607, "learning_rate": 7.383577542434355e-06, "loss": 0.0307, "step": 17493 }, { "epoch": 1.8965741543798784, "grad_norm": 1.7308847904205322, "learning_rate": 7.382852168866967e-06, "loss": 0.0028, "step": 17494 }, { "epoch": 1.8966825672159584, "grad_norm": 0.4204413890838623, "learning_rate": 7.38212679529958e-06, "loss": 0.0125, "step": 17495 }, { "epoch": 1.8967909800520382, "grad_norm": 0.43083563446998596, "learning_rate": 7.381401421732192e-06, "loss": 0.0028, "step": 17496 }, { "epoch": 1.896899392888118, "grad_norm": 0.7287760972976685, "learning_rate": 7.380676048164805e-06, "loss": 0.0247, "step": 17497 }, { "epoch": 1.8970078057241977, "grad_norm": 0.6452992558479309, "learning_rate": 7.379950674597418e-06, "loss": 0.0146, "step": 17498 }, { "epoch": 1.8971162185602775, "grad_norm": 0.36068111658096313, "learning_rate": 7.3792253010300305e-06, "loss": 0.0013, "step": 17499 }, { "epoch": 1.8972246313963574, "grad_norm": 3.0960493087768555, "learning_rate": 7.3784999274626435e-06, "loss": 0.0345, "step": 17500 }, { "epoch": 1.897333044232437, "grad_norm": 0.32533887028694153, "learning_rate": 7.377774553895257e-06, "loss": 0.0071, "step": 17501 }, { "epoch": 1.897441457068517, "grad_norm": 0.49728861451148987, "learning_rate": 7.3770491803278695e-06, "loss": 0.0131, "step": 17502 }, { "epoch": 1.8975498699045967, "grad_norm": 8.137223243713379, "learning_rate": 7.3763238067604824e-06, "loss": 0.0112, "step": 17503 }, { "epoch": 1.8976582827406765, "grad_norm": 0.6085871458053589, "learning_rate": 7.3755984331930954e-06, "loss": 0.0148, "step": 17504 }, { "epoch": 1.8977666955767563, "grad_norm": 1.4488383531570435, "learning_rate": 7.374873059625708e-06, "loss": 0.0032, "step": 17505 }, { "epoch": 1.897875108412836, "grad_norm": 4.923340797424316, "learning_rate": 7.3741476860583206e-06, "loss": 0.0182, "step": 17506 }, { "epoch": 1.897983521248916, "grad_norm": 0.6792129278182983, "learning_rate": 7.3734223124909336e-06, "loss": 0.0134, "step": 17507 }, { "epoch": 1.8980919340849955, "grad_norm": 5.534414291381836, "learning_rate": 7.372696938923546e-06, "loss": 0.0248, "step": 17508 }, { "epoch": 1.8982003469210755, "grad_norm": 4.466676712036133, "learning_rate": 7.371971565356159e-06, "loss": 0.0253, "step": 17509 }, { "epoch": 1.8983087597571553, "grad_norm": 1.808672308921814, "learning_rate": 7.3712461917887725e-06, "loss": 0.0358, "step": 17510 }, { "epoch": 1.898417172593235, "grad_norm": 0.7340937256813049, "learning_rate": 7.370520818221385e-06, "loss": 0.0048, "step": 17511 }, { "epoch": 1.8985255854293148, "grad_norm": 0.30803072452545166, "learning_rate": 7.369795444653998e-06, "loss": 0.0119, "step": 17512 }, { "epoch": 1.8986339982653946, "grad_norm": 1.5283172130584717, "learning_rate": 7.369070071086611e-06, "loss": 0.0166, "step": 17513 }, { "epoch": 1.8987424111014746, "grad_norm": 0.6721521615982056, "learning_rate": 7.368344697519223e-06, "loss": 0.0018, "step": 17514 }, { "epoch": 1.898850823937554, "grad_norm": 0.3638254702091217, "learning_rate": 7.367619323951836e-06, "loss": 0.0026, "step": 17515 }, { "epoch": 1.898959236773634, "grad_norm": 4.3137922286987305, "learning_rate": 7.366893950384448e-06, "loss": 0.0433, "step": 17516 }, { "epoch": 1.8990676496097136, "grad_norm": 3.1262338161468506, "learning_rate": 7.366168576817061e-06, "loss": 0.0531, "step": 17517 }, { "epoch": 1.8991760624457936, "grad_norm": 3.1034984588623047, "learning_rate": 7.365443203249674e-06, "loss": 0.0267, "step": 17518 }, { "epoch": 1.8992844752818734, "grad_norm": 0.5135335922241211, "learning_rate": 7.364717829682288e-06, "loss": 0.0101, "step": 17519 }, { "epoch": 1.8993928881179531, "grad_norm": 0.6575534343719482, "learning_rate": 7.3639924561149e-06, "loss": 0.0071, "step": 17520 }, { "epoch": 1.899501300954033, "grad_norm": 6.2041754722595215, "learning_rate": 7.363267082547513e-06, "loss": 0.0087, "step": 17521 }, { "epoch": 1.8996097137901127, "grad_norm": 0.3553836941719055, "learning_rate": 7.362541708980125e-06, "loss": 0.0107, "step": 17522 }, { "epoch": 1.8997181266261927, "grad_norm": 1.8285810947418213, "learning_rate": 7.361816335412738e-06, "loss": 0.036, "step": 17523 }, { "epoch": 1.8998265394622722, "grad_norm": 4.139588832855225, "learning_rate": 7.361090961845351e-06, "loss": 0.0989, "step": 17524 }, { "epoch": 1.8999349522983522, "grad_norm": 0.014352784492075443, "learning_rate": 7.360365588277963e-06, "loss": 0.0002, "step": 17525 }, { "epoch": 1.900043365134432, "grad_norm": 0.10344044864177704, "learning_rate": 7.359640214710576e-06, "loss": 0.0012, "step": 17526 }, { "epoch": 1.9001517779705117, "grad_norm": 0.812056303024292, "learning_rate": 7.35891484114319e-06, "loss": 0.0122, "step": 17527 }, { "epoch": 1.9002601908065915, "grad_norm": 0.47883540391921997, "learning_rate": 7.358189467575802e-06, "loss": 0.0026, "step": 17528 }, { "epoch": 1.9003686036426712, "grad_norm": 4.8004536628723145, "learning_rate": 7.357464094008415e-06, "loss": 0.0614, "step": 17529 }, { "epoch": 1.9004770164787512, "grad_norm": 3.71032977104187, "learning_rate": 7.356738720441028e-06, "loss": 0.0389, "step": 17530 }, { "epoch": 1.9005854293148308, "grad_norm": 2.006957769393921, "learning_rate": 7.35601334687364e-06, "loss": 0.0232, "step": 17531 }, { "epoch": 1.9006938421509108, "grad_norm": 4.7975873947143555, "learning_rate": 7.355287973306253e-06, "loss": 0.0379, "step": 17532 }, { "epoch": 1.9008022549869905, "grad_norm": 0.2571498453617096, "learning_rate": 7.354562599738866e-06, "loss": 0.0047, "step": 17533 }, { "epoch": 1.9009106678230703, "grad_norm": 6.105103015899658, "learning_rate": 7.353837226171478e-06, "loss": 0.035, "step": 17534 }, { "epoch": 1.90101908065915, "grad_norm": 6.7156805992126465, "learning_rate": 7.353111852604091e-06, "loss": 0.0533, "step": 17535 }, { "epoch": 1.9011274934952298, "grad_norm": 4.557792663574219, "learning_rate": 7.352386479036705e-06, "loss": 0.0403, "step": 17536 }, { "epoch": 1.9012359063313098, "grad_norm": 3.0766541957855225, "learning_rate": 7.351661105469317e-06, "loss": 0.0181, "step": 17537 }, { "epoch": 1.9013443191673893, "grad_norm": 1.8921992778778076, "learning_rate": 7.35093573190193e-06, "loss": 0.0206, "step": 17538 }, { "epoch": 1.9014527320034693, "grad_norm": 5.723480224609375, "learning_rate": 7.350210358334543e-06, "loss": 0.013, "step": 17539 }, { "epoch": 1.9015611448395489, "grad_norm": 0.28490644693374634, "learning_rate": 7.3494849847671555e-06, "loss": 0.0011, "step": 17540 }, { "epoch": 1.9016695576756288, "grad_norm": 1.437004566192627, "learning_rate": 7.3487596111997685e-06, "loss": 0.0115, "step": 17541 }, { "epoch": 1.9017779705117086, "grad_norm": 1.1045048236846924, "learning_rate": 7.348034237632381e-06, "loss": 0.0269, "step": 17542 }, { "epoch": 1.9018863833477884, "grad_norm": 5.179756164550781, "learning_rate": 7.347308864064994e-06, "loss": 0.0451, "step": 17543 }, { "epoch": 1.9019947961838681, "grad_norm": 2.3017451763153076, "learning_rate": 7.346583490497607e-06, "loss": 0.0102, "step": 17544 }, { "epoch": 1.902103209019948, "grad_norm": 6.556440830230713, "learning_rate": 7.34585811693022e-06, "loss": 0.059, "step": 17545 }, { "epoch": 1.9022116218560279, "grad_norm": 0.4337715208530426, "learning_rate": 7.3451327433628326e-06, "loss": 0.0088, "step": 17546 }, { "epoch": 1.9023200346921074, "grad_norm": 2.257054090499878, "learning_rate": 7.3444073697954456e-06, "loss": 0.0215, "step": 17547 }, { "epoch": 1.9024284475281874, "grad_norm": 0.5946452617645264, "learning_rate": 7.343681996228058e-06, "loss": 0.0032, "step": 17548 }, { "epoch": 1.9025368603642672, "grad_norm": 2.098827362060547, "learning_rate": 7.342956622660671e-06, "loss": 0.0131, "step": 17549 }, { "epoch": 1.902645273200347, "grad_norm": 2.354613780975342, "learning_rate": 7.342231249093284e-06, "loss": 0.0187, "step": 17550 }, { "epoch": 1.9027536860364267, "grad_norm": 1.212676763534546, "learning_rate": 7.341505875525896e-06, "loss": 0.0186, "step": 17551 }, { "epoch": 1.9028620988725065, "grad_norm": 2.223600149154663, "learning_rate": 7.340780501958509e-06, "loss": 0.0516, "step": 17552 }, { "epoch": 1.9029705117085864, "grad_norm": 1.417933464050293, "learning_rate": 7.340055128391123e-06, "loss": 0.0109, "step": 17553 }, { "epoch": 1.903078924544666, "grad_norm": 0.08208717405796051, "learning_rate": 7.339329754823736e-06, "loss": 0.0005, "step": 17554 }, { "epoch": 1.903187337380746, "grad_norm": 4.312765121459961, "learning_rate": 7.338604381256348e-06, "loss": 0.0391, "step": 17555 }, { "epoch": 1.9032957502168255, "grad_norm": 3.156146287918091, "learning_rate": 7.337879007688961e-06, "loss": 0.0226, "step": 17556 }, { "epoch": 1.9034041630529055, "grad_norm": 1.9243499040603638, "learning_rate": 7.337153634121573e-06, "loss": 0.029, "step": 17557 }, { "epoch": 1.9035125758889853, "grad_norm": 2.5255935192108154, "learning_rate": 7.336428260554186e-06, "loss": 0.0244, "step": 17558 }, { "epoch": 1.903620988725065, "grad_norm": 1.8610050678253174, "learning_rate": 7.335702886986799e-06, "loss": 0.0137, "step": 17559 }, { "epoch": 1.9037294015611448, "grad_norm": 9.91826343536377, "learning_rate": 7.334977513419411e-06, "loss": 0.0672, "step": 17560 }, { "epoch": 1.9038378143972245, "grad_norm": 1.136445164680481, "learning_rate": 7.334252139852024e-06, "loss": 0.0251, "step": 17561 }, { "epoch": 1.9039462272333045, "grad_norm": 0.5483034253120422, "learning_rate": 7.333526766284638e-06, "loss": 0.0187, "step": 17562 }, { "epoch": 1.904054640069384, "grad_norm": 0.7545689344406128, "learning_rate": 7.33280139271725e-06, "loss": 0.0315, "step": 17563 }, { "epoch": 1.904163052905464, "grad_norm": 5.6189985275268555, "learning_rate": 7.332076019149863e-06, "loss": 0.0542, "step": 17564 }, { "epoch": 1.9042714657415438, "grad_norm": 0.39623206853866577, "learning_rate": 7.331350645582476e-06, "loss": 0.0131, "step": 17565 }, { "epoch": 1.9043798785776236, "grad_norm": 2.3777213096618652, "learning_rate": 7.330625272015088e-06, "loss": 0.0241, "step": 17566 }, { "epoch": 1.9044882914137033, "grad_norm": 5.101234436035156, "learning_rate": 7.329899898447701e-06, "loss": 0.0378, "step": 17567 }, { "epoch": 1.9045967042497831, "grad_norm": 9.289639472961426, "learning_rate": 7.329174524880313e-06, "loss": 0.0628, "step": 17568 }, { "epoch": 1.904705117085863, "grad_norm": 0.9687122702598572, "learning_rate": 7.328449151312926e-06, "loss": 0.0134, "step": 17569 }, { "epoch": 1.9048135299219426, "grad_norm": 1.5927674770355225, "learning_rate": 7.327723777745539e-06, "loss": 0.0234, "step": 17570 }, { "epoch": 1.9049219427580226, "grad_norm": 0.61771160364151, "learning_rate": 7.326998404178153e-06, "loss": 0.006, "step": 17571 }, { "epoch": 1.9050303555941024, "grad_norm": 3.9315378665924072, "learning_rate": 7.326273030610765e-06, "loss": 0.0233, "step": 17572 }, { "epoch": 1.9051387684301822, "grad_norm": 1.7769535779953003, "learning_rate": 7.325547657043378e-06, "loss": 0.0107, "step": 17573 }, { "epoch": 1.905247181266262, "grad_norm": 2.123159646987915, "learning_rate": 7.32482228347599e-06, "loss": 0.0408, "step": 17574 }, { "epoch": 1.9053555941023417, "grad_norm": 0.4178181290626526, "learning_rate": 7.324096909908603e-06, "loss": 0.0132, "step": 17575 }, { "epoch": 1.9054640069384217, "grad_norm": 0.6397672891616821, "learning_rate": 7.323371536341216e-06, "loss": 0.0148, "step": 17576 }, { "epoch": 1.9055724197745012, "grad_norm": 0.8164721131324768, "learning_rate": 7.3226461627738285e-06, "loss": 0.0114, "step": 17577 }, { "epoch": 1.9056808326105812, "grad_norm": 2.6047089099884033, "learning_rate": 7.3219207892064415e-06, "loss": 0.0245, "step": 17578 }, { "epoch": 1.9057892454466607, "grad_norm": 3.667264223098755, "learning_rate": 7.321195415639055e-06, "loss": 0.0187, "step": 17579 }, { "epoch": 1.9058976582827407, "grad_norm": 7.47035551071167, "learning_rate": 7.320470042071668e-06, "loss": 0.1209, "step": 17580 }, { "epoch": 1.9060060711188205, "grad_norm": 1.7879737615585327, "learning_rate": 7.3197446685042805e-06, "loss": 0.0291, "step": 17581 }, { "epoch": 1.9061144839549002, "grad_norm": 2.4127891063690186, "learning_rate": 7.3190192949368934e-06, "loss": 0.0289, "step": 17582 }, { "epoch": 1.90622289679098, "grad_norm": 3.362945556640625, "learning_rate": 7.318293921369506e-06, "loss": 0.0178, "step": 17583 }, { "epoch": 1.9063313096270598, "grad_norm": 0.8666655421257019, "learning_rate": 7.317568547802119e-06, "loss": 0.0122, "step": 17584 }, { "epoch": 1.9064397224631398, "grad_norm": 1.3697401285171509, "learning_rate": 7.3168431742347316e-06, "loss": 0.0529, "step": 17585 }, { "epoch": 1.9065481352992193, "grad_norm": 5.576757907867432, "learning_rate": 7.316117800667344e-06, "loss": 0.072, "step": 17586 }, { "epoch": 1.9066565481352993, "grad_norm": 1.3890169858932495, "learning_rate": 7.315392427099957e-06, "loss": 0.0267, "step": 17587 }, { "epoch": 1.906764960971379, "grad_norm": 2.526010751724243, "learning_rate": 7.3146670535325705e-06, "loss": 0.0132, "step": 17588 }, { "epoch": 1.9068733738074588, "grad_norm": 6.144186496734619, "learning_rate": 7.313941679965183e-06, "loss": 0.0502, "step": 17589 }, { "epoch": 1.9069817866435386, "grad_norm": 2.113435745239258, "learning_rate": 7.313216306397796e-06, "loss": 0.0289, "step": 17590 }, { "epoch": 1.9070901994796183, "grad_norm": 3.1988933086395264, "learning_rate": 7.312490932830409e-06, "loss": 0.0082, "step": 17591 }, { "epoch": 1.9071986123156983, "grad_norm": 4.467253684997559, "learning_rate": 7.311765559263021e-06, "loss": 0.0143, "step": 17592 }, { "epoch": 1.9073070251517779, "grad_norm": 1.530000925064087, "learning_rate": 7.311040185695634e-06, "loss": 0.005, "step": 17593 }, { "epoch": 1.9074154379878578, "grad_norm": 3.0559887886047363, "learning_rate": 7.310314812128246e-06, "loss": 0.0148, "step": 17594 }, { "epoch": 1.9075238508239376, "grad_norm": 1.0005502700805664, "learning_rate": 7.309589438560859e-06, "loss": 0.0465, "step": 17595 }, { "epoch": 1.9076322636600174, "grad_norm": 4.1767096519470215, "learning_rate": 7.308864064993473e-06, "loss": 0.0349, "step": 17596 }, { "epoch": 1.9077406764960971, "grad_norm": 8.780299186706543, "learning_rate": 7.308138691426086e-06, "loss": 0.0135, "step": 17597 }, { "epoch": 1.907849089332177, "grad_norm": 0.3813754618167877, "learning_rate": 7.307413317858698e-06, "loss": 0.0074, "step": 17598 }, { "epoch": 1.9079575021682569, "grad_norm": 3.0697996616363525, "learning_rate": 7.306687944291311e-06, "loss": 0.0275, "step": 17599 }, { "epoch": 1.9080659150043364, "grad_norm": 0.4783734381198883, "learning_rate": 7.305962570723923e-06, "loss": 0.0092, "step": 17600 }, { "epoch": 1.9081743278404164, "grad_norm": 0.4801097810268402, "learning_rate": 7.305237197156536e-06, "loss": 0.0121, "step": 17601 }, { "epoch": 1.908282740676496, "grad_norm": 1.0651277303695679, "learning_rate": 7.304511823589149e-06, "loss": 0.0361, "step": 17602 }, { "epoch": 1.908391153512576, "grad_norm": 1.2605319023132324, "learning_rate": 7.303786450021761e-06, "loss": 0.0158, "step": 17603 }, { "epoch": 1.9084995663486557, "grad_norm": 0.39005473256111145, "learning_rate": 7.303061076454374e-06, "loss": 0.0154, "step": 17604 }, { "epoch": 1.9086079791847355, "grad_norm": 2.70137357711792, "learning_rate": 7.302335702886988e-06, "loss": 0.0097, "step": 17605 }, { "epoch": 1.9087163920208152, "grad_norm": 2.114687442779541, "learning_rate": 7.301610329319601e-06, "loss": 0.0375, "step": 17606 }, { "epoch": 1.908824804856895, "grad_norm": 0.14840556681156158, "learning_rate": 7.300884955752213e-06, "loss": 0.0019, "step": 17607 }, { "epoch": 1.908933217692975, "grad_norm": 1.726503849029541, "learning_rate": 7.300159582184826e-06, "loss": 0.0194, "step": 17608 }, { "epoch": 1.9090416305290545, "grad_norm": 1.4626139402389526, "learning_rate": 7.299434208617438e-06, "loss": 0.0319, "step": 17609 }, { "epoch": 1.9091500433651345, "grad_norm": 2.7431063652038574, "learning_rate": 7.298708835050051e-06, "loss": 0.0248, "step": 17610 }, { "epoch": 1.9092584562012143, "grad_norm": 0.900685727596283, "learning_rate": 7.297983461482664e-06, "loss": 0.0176, "step": 17611 }, { "epoch": 1.909366869037294, "grad_norm": 1.4252177476882935, "learning_rate": 7.297258087915276e-06, "loss": 0.0205, "step": 17612 }, { "epoch": 1.9094752818733738, "grad_norm": 1.868881344795227, "learning_rate": 7.296532714347889e-06, "loss": 0.0198, "step": 17613 }, { "epoch": 1.9095836947094535, "grad_norm": 0.6639248132705688, "learning_rate": 7.295807340780503e-06, "loss": 0.0033, "step": 17614 }, { "epoch": 1.9096921075455335, "grad_norm": 0.6260640621185303, "learning_rate": 7.295081967213115e-06, "loss": 0.0036, "step": 17615 }, { "epoch": 1.909800520381613, "grad_norm": 0.7520203590393066, "learning_rate": 7.294356593645728e-06, "loss": 0.0091, "step": 17616 }, { "epoch": 1.909908933217693, "grad_norm": 0.8161993026733398, "learning_rate": 7.293631220078341e-06, "loss": 0.008, "step": 17617 }, { "epoch": 1.9100173460537726, "grad_norm": 1.5829185247421265, "learning_rate": 7.2929058465109535e-06, "loss": 0.0193, "step": 17618 }, { "epoch": 1.9101257588898526, "grad_norm": 1.9313218593597412, "learning_rate": 7.2921804729435665e-06, "loss": 0.0231, "step": 17619 }, { "epoch": 1.9102341717259324, "grad_norm": 2.7648775577545166, "learning_rate": 7.291455099376179e-06, "loss": 0.0059, "step": 17620 }, { "epoch": 1.9103425845620121, "grad_norm": 2.017270803451538, "learning_rate": 7.290729725808792e-06, "loss": 0.0234, "step": 17621 }, { "epoch": 1.910450997398092, "grad_norm": 2.430813789367676, "learning_rate": 7.2900043522414054e-06, "loss": 0.0495, "step": 17622 }, { "epoch": 1.9105594102341716, "grad_norm": 0.7469125390052795, "learning_rate": 7.2892789786740184e-06, "loss": 0.0079, "step": 17623 }, { "epoch": 1.9106678230702516, "grad_norm": 4.9788031578063965, "learning_rate": 7.2885536051066306e-06, "loss": 0.0403, "step": 17624 }, { "epoch": 1.9107762359063312, "grad_norm": 3.4128286838531494, "learning_rate": 7.2878282315392436e-06, "loss": 0.0355, "step": 17625 }, { "epoch": 1.9108846487424112, "grad_norm": 1.1801429986953735, "learning_rate": 7.287102857971856e-06, "loss": 0.0076, "step": 17626 }, { "epoch": 1.910993061578491, "grad_norm": 3.2949254512786865, "learning_rate": 7.286377484404469e-06, "loss": 0.012, "step": 17627 }, { "epoch": 1.9111014744145707, "grad_norm": 1.3538216352462769, "learning_rate": 7.285652110837082e-06, "loss": 0.0215, "step": 17628 }, { "epoch": 1.9112098872506504, "grad_norm": 1.5073940753936768, "learning_rate": 7.284926737269694e-06, "loss": 0.0193, "step": 17629 }, { "epoch": 1.9113183000867302, "grad_norm": 1.0914673805236816, "learning_rate": 7.284201363702307e-06, "loss": 0.0074, "step": 17630 }, { "epoch": 1.9114267129228102, "grad_norm": 0.36804911494255066, "learning_rate": 7.283475990134921e-06, "loss": 0.0108, "step": 17631 }, { "epoch": 1.9115351257588897, "grad_norm": 2.294168710708618, "learning_rate": 7.282750616567534e-06, "loss": 0.0281, "step": 17632 }, { "epoch": 1.9116435385949697, "grad_norm": 1.5400285720825195, "learning_rate": 7.282025243000146e-06, "loss": 0.0051, "step": 17633 }, { "epoch": 1.9117519514310495, "grad_norm": 3.4578428268432617, "learning_rate": 7.281299869432759e-06, "loss": 0.0161, "step": 17634 }, { "epoch": 1.9118603642671292, "grad_norm": 1.3252211809158325, "learning_rate": 7.280574495865371e-06, "loss": 0.0115, "step": 17635 }, { "epoch": 1.911968777103209, "grad_norm": 1.6053202152252197, "learning_rate": 7.279849122297984e-06, "loss": 0.0204, "step": 17636 }, { "epoch": 1.9120771899392888, "grad_norm": 4.937232494354248, "learning_rate": 7.279123748730597e-06, "loss": 0.0181, "step": 17637 }, { "epoch": 1.9121856027753688, "grad_norm": 1.315826177597046, "learning_rate": 7.278398375163209e-06, "loss": 0.0147, "step": 17638 }, { "epoch": 1.9122940156114483, "grad_norm": 0.47291702032089233, "learning_rate": 7.277673001595822e-06, "loss": 0.0079, "step": 17639 }, { "epoch": 1.9124024284475283, "grad_norm": 0.3529238998889923, "learning_rate": 7.276947628028436e-06, "loss": 0.012, "step": 17640 }, { "epoch": 1.9125108412836078, "grad_norm": 6.811412334442139, "learning_rate": 7.276222254461048e-06, "loss": 0.0104, "step": 17641 }, { "epoch": 1.9126192541196878, "grad_norm": 0.7446550726890564, "learning_rate": 7.275496880893661e-06, "loss": 0.0061, "step": 17642 }, { "epoch": 1.9127276669557676, "grad_norm": 3.0581159591674805, "learning_rate": 7.274771507326274e-06, "loss": 0.0105, "step": 17643 }, { "epoch": 1.9128360797918473, "grad_norm": 0.6836767792701721, "learning_rate": 7.274046133758886e-06, "loss": 0.0341, "step": 17644 }, { "epoch": 1.912944492627927, "grad_norm": 2.2548413276672363, "learning_rate": 7.273320760191499e-06, "loss": 0.025, "step": 17645 }, { "epoch": 1.9130529054640069, "grad_norm": 4.829050064086914, "learning_rate": 7.272595386624111e-06, "loss": 0.0685, "step": 17646 }, { "epoch": 1.9131613183000868, "grad_norm": 0.6347755193710327, "learning_rate": 7.271870013056724e-06, "loss": 0.0304, "step": 17647 }, { "epoch": 1.9132697311361664, "grad_norm": 3.15216064453125, "learning_rate": 7.271144639489338e-06, "loss": 0.0321, "step": 17648 }, { "epoch": 1.9133781439722464, "grad_norm": 0.5060856342315674, "learning_rate": 7.270419265921951e-06, "loss": 0.0141, "step": 17649 }, { "epoch": 1.9134865568083261, "grad_norm": 0.11324483901262283, "learning_rate": 7.269693892354563e-06, "loss": 0.0008, "step": 17650 }, { "epoch": 1.913594969644406, "grad_norm": 0.38298094272613525, "learning_rate": 7.268968518787176e-06, "loss": 0.0066, "step": 17651 }, { "epoch": 1.9137033824804857, "grad_norm": 0.9322221279144287, "learning_rate": 7.268243145219789e-06, "loss": 0.0118, "step": 17652 }, { "epoch": 1.9138117953165654, "grad_norm": 0.4279971122741699, "learning_rate": 7.267517771652401e-06, "loss": 0.0064, "step": 17653 }, { "epoch": 1.9139202081526454, "grad_norm": 1.4513570070266724, "learning_rate": 7.266792398085014e-06, "loss": 0.0455, "step": 17654 }, { "epoch": 1.914028620988725, "grad_norm": 0.27013275027275085, "learning_rate": 7.2660670245176265e-06, "loss": 0.006, "step": 17655 }, { "epoch": 1.914137033824805, "grad_norm": 1.1948133707046509, "learning_rate": 7.2653416509502395e-06, "loss": 0.0064, "step": 17656 }, { "epoch": 1.9142454466608847, "grad_norm": 4.518542766571045, "learning_rate": 7.264616277382853e-06, "loss": 0.0096, "step": 17657 }, { "epoch": 1.9143538594969645, "grad_norm": 3.2030141353607178, "learning_rate": 7.263890903815466e-06, "loss": 0.0301, "step": 17658 }, { "epoch": 1.9144622723330442, "grad_norm": 5.430473327636719, "learning_rate": 7.2631655302480785e-06, "loss": 0.0283, "step": 17659 }, { "epoch": 1.914570685169124, "grad_norm": 2.40765380859375, "learning_rate": 7.2624401566806914e-06, "loss": 0.032, "step": 17660 }, { "epoch": 1.914679098005204, "grad_norm": 0.6171637177467346, "learning_rate": 7.261714783113304e-06, "loss": 0.004, "step": 17661 }, { "epoch": 1.9147875108412835, "grad_norm": 0.61688232421875, "learning_rate": 7.260989409545917e-06, "loss": 0.0147, "step": 17662 }, { "epoch": 1.9148959236773635, "grad_norm": 0.048139531165361404, "learning_rate": 7.2602640359785296e-06, "loss": 0.0002, "step": 17663 }, { "epoch": 1.915004336513443, "grad_norm": 0.9906501770019531, "learning_rate": 7.259538662411142e-06, "loss": 0.0035, "step": 17664 }, { "epoch": 1.915112749349523, "grad_norm": 1.4556541442871094, "learning_rate": 7.258813288843755e-06, "loss": 0.0177, "step": 17665 }, { "epoch": 1.9152211621856028, "grad_norm": 4.180774211883545, "learning_rate": 7.2580879152763685e-06, "loss": 0.0093, "step": 17666 }, { "epoch": 1.9153295750216826, "grad_norm": 2.3450820446014404, "learning_rate": 7.257362541708981e-06, "loss": 0.0102, "step": 17667 }, { "epoch": 1.9154379878577623, "grad_norm": 2.0039708614349365, "learning_rate": 7.256637168141594e-06, "loss": 0.0103, "step": 17668 }, { "epoch": 1.915546400693842, "grad_norm": 5.242007255554199, "learning_rate": 7.255911794574207e-06, "loss": 0.0313, "step": 17669 }, { "epoch": 1.915654813529922, "grad_norm": 10.099082946777344, "learning_rate": 7.255186421006819e-06, "loss": 0.0129, "step": 17670 }, { "epoch": 1.9157632263660016, "grad_norm": 0.13296058773994446, "learning_rate": 7.254461047439432e-06, "loss": 0.0126, "step": 17671 }, { "epoch": 1.9158716392020816, "grad_norm": 0.476785272359848, "learning_rate": 7.253735673872044e-06, "loss": 0.0074, "step": 17672 }, { "epoch": 1.9159800520381614, "grad_norm": 0.6947687268257141, "learning_rate": 7.253010300304657e-06, "loss": 0.0055, "step": 17673 }, { "epoch": 1.9160884648742411, "grad_norm": 3.2703194618225098, "learning_rate": 7.252284926737271e-06, "loss": 0.029, "step": 17674 }, { "epoch": 1.9161968777103209, "grad_norm": 0.037798892706632614, "learning_rate": 7.251559553169884e-06, "loss": 0.0003, "step": 17675 }, { "epoch": 1.9163052905464006, "grad_norm": 0.28680557012557983, "learning_rate": 7.250834179602496e-06, "loss": 0.0059, "step": 17676 }, { "epoch": 1.9164137033824806, "grad_norm": 4.339040756225586, "learning_rate": 7.250108806035109e-06, "loss": 0.0373, "step": 17677 }, { "epoch": 1.9165221162185602, "grad_norm": 1.0716619491577148, "learning_rate": 7.249383432467722e-06, "loss": 0.0296, "step": 17678 }, { "epoch": 1.9166305290546402, "grad_norm": 0.5340219140052795, "learning_rate": 7.248658058900334e-06, "loss": 0.0042, "step": 17679 }, { "epoch": 1.91673894189072, "grad_norm": 1.3952481746673584, "learning_rate": 7.247932685332947e-06, "loss": 0.0229, "step": 17680 }, { "epoch": 1.9168473547267997, "grad_norm": 0.3972775340080261, "learning_rate": 7.247207311765559e-06, "loss": 0.0023, "step": 17681 }, { "epoch": 1.9169557675628794, "grad_norm": 5.875853061676025, "learning_rate": 7.246481938198172e-06, "loss": 0.0159, "step": 17682 }, { "epoch": 1.9170641803989592, "grad_norm": 1.1558839082717896, "learning_rate": 7.245756564630786e-06, "loss": 0.003, "step": 17683 }, { "epoch": 1.9171725932350392, "grad_norm": 1.771074652671814, "learning_rate": 7.245031191063399e-06, "loss": 0.0221, "step": 17684 }, { "epoch": 1.9172810060711187, "grad_norm": 1.2068403959274292, "learning_rate": 7.244305817496011e-06, "loss": 0.0104, "step": 17685 }, { "epoch": 1.9173894189071987, "grad_norm": 1.2658302783966064, "learning_rate": 7.243580443928624e-06, "loss": 0.0096, "step": 17686 }, { "epoch": 1.9174978317432783, "grad_norm": 0.3302075266838074, "learning_rate": 7.242855070361236e-06, "loss": 0.0051, "step": 17687 }, { "epoch": 1.9176062445793582, "grad_norm": 0.1855383962392807, "learning_rate": 7.242129696793849e-06, "loss": 0.0007, "step": 17688 }, { "epoch": 1.917714657415438, "grad_norm": 0.44993099570274353, "learning_rate": 7.241404323226462e-06, "loss": 0.0045, "step": 17689 }, { "epoch": 1.9178230702515178, "grad_norm": 4.817838668823242, "learning_rate": 7.240678949659074e-06, "loss": 0.0347, "step": 17690 }, { "epoch": 1.9179314830875975, "grad_norm": 2.4080984592437744, "learning_rate": 7.239953576091688e-06, "loss": 0.1065, "step": 17691 }, { "epoch": 1.9180398959236773, "grad_norm": 2.755943775177002, "learning_rate": 7.239228202524301e-06, "loss": 0.0228, "step": 17692 }, { "epoch": 1.9181483087597573, "grad_norm": 3.7796428203582764, "learning_rate": 7.238502828956913e-06, "loss": 0.0244, "step": 17693 }, { "epoch": 1.9182567215958368, "grad_norm": 2.639265775680542, "learning_rate": 7.237777455389526e-06, "loss": 0.0472, "step": 17694 }, { "epoch": 1.9183651344319168, "grad_norm": 6.379082202911377, "learning_rate": 7.237052081822139e-06, "loss": 0.0195, "step": 17695 }, { "epoch": 1.9184735472679966, "grad_norm": 0.49830156564712524, "learning_rate": 7.2363267082547515e-06, "loss": 0.0153, "step": 17696 }, { "epoch": 1.9185819601040763, "grad_norm": 6.752969741821289, "learning_rate": 7.2356013346873645e-06, "loss": 0.0581, "step": 17697 }, { "epoch": 1.918690372940156, "grad_norm": 1.7268654108047485, "learning_rate": 7.234875961119977e-06, "loss": 0.006, "step": 17698 }, { "epoch": 1.9187987857762359, "grad_norm": 7.329611778259277, "learning_rate": 7.23415058755259e-06, "loss": 0.0967, "step": 17699 }, { "epoch": 1.9189071986123158, "grad_norm": 4.109506607055664, "learning_rate": 7.2334252139852034e-06, "loss": 0.0187, "step": 17700 }, { "epoch": 1.9190156114483954, "grad_norm": 3.8676488399505615, "learning_rate": 7.2326998404178164e-06, "loss": 0.051, "step": 17701 }, { "epoch": 1.9191240242844754, "grad_norm": 0.9703428149223328, "learning_rate": 7.231974466850429e-06, "loss": 0.0124, "step": 17702 }, { "epoch": 1.919232437120555, "grad_norm": 5.190497398376465, "learning_rate": 7.2312490932830416e-06, "loss": 0.1124, "step": 17703 }, { "epoch": 1.919340849956635, "grad_norm": 6.410442352294922, "learning_rate": 7.2305237197156546e-06, "loss": 0.0393, "step": 17704 }, { "epoch": 1.9194492627927147, "grad_norm": 15.017233848571777, "learning_rate": 7.229798346148267e-06, "loss": 0.0361, "step": 17705 }, { "epoch": 1.9195576756287944, "grad_norm": 0.4439379870891571, "learning_rate": 7.22907297258088e-06, "loss": 0.0012, "step": 17706 }, { "epoch": 1.9196660884648742, "grad_norm": 8.921953201293945, "learning_rate": 7.228347599013492e-06, "loss": 0.0076, "step": 17707 }, { "epoch": 1.919774501300954, "grad_norm": 2.6290783882141113, "learning_rate": 7.227622225446105e-06, "loss": 0.032, "step": 17708 }, { "epoch": 1.919882914137034, "grad_norm": 2.4372708797454834, "learning_rate": 7.226896851878719e-06, "loss": 0.0571, "step": 17709 }, { "epoch": 1.9199913269731135, "grad_norm": 0.01116550900042057, "learning_rate": 7.226171478311332e-06, "loss": 0.0001, "step": 17710 }, { "epoch": 1.9200997398091935, "grad_norm": 2.5726850032806396, "learning_rate": 7.225446104743944e-06, "loss": 0.0402, "step": 17711 }, { "epoch": 1.9202081526452732, "grad_norm": 7.029848098754883, "learning_rate": 7.224720731176557e-06, "loss": 0.0212, "step": 17712 }, { "epoch": 1.920316565481353, "grad_norm": 0.5596956610679626, "learning_rate": 7.223995357609169e-06, "loss": 0.0142, "step": 17713 }, { "epoch": 1.9204249783174328, "grad_norm": 3.939084768295288, "learning_rate": 7.223269984041782e-06, "loss": 0.0388, "step": 17714 }, { "epoch": 1.9205333911535125, "grad_norm": 0.2854551374912262, "learning_rate": 7.222544610474395e-06, "loss": 0.0058, "step": 17715 }, { "epoch": 1.9206418039895925, "grad_norm": 0.6122543811798096, "learning_rate": 7.221819236907007e-06, "loss": 0.0022, "step": 17716 }, { "epoch": 1.920750216825672, "grad_norm": 10.768678665161133, "learning_rate": 7.221093863339621e-06, "loss": 0.035, "step": 17717 }, { "epoch": 1.920858629661752, "grad_norm": 1.9615998268127441, "learning_rate": 7.220368489772234e-06, "loss": 0.0346, "step": 17718 }, { "epoch": 1.9209670424978318, "grad_norm": 6.6164445877075195, "learning_rate": 7.219643116204846e-06, "loss": 0.0266, "step": 17719 }, { "epoch": 1.9210754553339116, "grad_norm": 0.4368365406990051, "learning_rate": 7.218917742637459e-06, "loss": 0.0074, "step": 17720 }, { "epoch": 1.9211838681699913, "grad_norm": 4.098673343658447, "learning_rate": 7.218192369070072e-06, "loss": 0.0378, "step": 17721 }, { "epoch": 1.921292281006071, "grad_norm": 0.9400911927223206, "learning_rate": 7.217466995502684e-06, "loss": 0.0019, "step": 17722 }, { "epoch": 1.921400693842151, "grad_norm": 2.914036750793457, "learning_rate": 7.216741621935297e-06, "loss": 0.0493, "step": 17723 }, { "epoch": 1.9215091066782306, "grad_norm": 1.9792189598083496, "learning_rate": 7.216016248367909e-06, "loss": 0.03, "step": 17724 }, { "epoch": 1.9216175195143106, "grad_norm": 0.7370694875717163, "learning_rate": 7.215290874800522e-06, "loss": 0.0083, "step": 17725 }, { "epoch": 1.9217259323503901, "grad_norm": 0.0957689881324768, "learning_rate": 7.214565501233136e-06, "loss": 0.0008, "step": 17726 }, { "epoch": 1.9218343451864701, "grad_norm": 0.4526130259037018, "learning_rate": 7.213840127665749e-06, "loss": 0.0174, "step": 17727 }, { "epoch": 1.9219427580225499, "grad_norm": 1.0091460943222046, "learning_rate": 7.213114754098361e-06, "loss": 0.0065, "step": 17728 }, { "epoch": 1.9220511708586296, "grad_norm": 0.7210672497749329, "learning_rate": 7.212389380530974e-06, "loss": 0.0059, "step": 17729 }, { "epoch": 1.9221595836947094, "grad_norm": 8.412817001342773, "learning_rate": 7.211664006963587e-06, "loss": 0.0222, "step": 17730 }, { "epoch": 1.9222679965307892, "grad_norm": 1.743098497390747, "learning_rate": 7.210938633396199e-06, "loss": 0.0362, "step": 17731 }, { "epoch": 1.9223764093668692, "grad_norm": 5.162940979003906, "learning_rate": 7.210213259828812e-06, "loss": 0.0055, "step": 17732 }, { "epoch": 1.9224848222029487, "grad_norm": 2.18595290184021, "learning_rate": 7.2094878862614245e-06, "loss": 0.0076, "step": 17733 }, { "epoch": 1.9225932350390287, "grad_norm": 0.30830585956573486, "learning_rate": 7.2087625126940375e-06, "loss": 0.0067, "step": 17734 }, { "epoch": 1.9227016478751084, "grad_norm": 0.5190162658691406, "learning_rate": 7.208037139126651e-06, "loss": 0.0093, "step": 17735 }, { "epoch": 1.9228100607111882, "grad_norm": 5.774763584136963, "learning_rate": 7.207311765559264e-06, "loss": 0.049, "step": 17736 }, { "epoch": 1.922918473547268, "grad_norm": 2.461683988571167, "learning_rate": 7.2065863919918765e-06, "loss": 0.0206, "step": 17737 }, { "epoch": 1.9230268863833477, "grad_norm": 0.1997661292552948, "learning_rate": 7.2058610184244895e-06, "loss": 0.0019, "step": 17738 }, { "epoch": 1.9231352992194277, "grad_norm": 2.1718530654907227, "learning_rate": 7.205135644857102e-06, "loss": 0.0059, "step": 17739 }, { "epoch": 1.9232437120555073, "grad_norm": 5.998303413391113, "learning_rate": 7.204410271289715e-06, "loss": 0.0372, "step": 17740 }, { "epoch": 1.9233521248915872, "grad_norm": 4.238367080688477, "learning_rate": 7.203684897722328e-06, "loss": 0.0241, "step": 17741 }, { "epoch": 1.923460537727667, "grad_norm": 0.6654507517814636, "learning_rate": 7.20295952415494e-06, "loss": 0.0256, "step": 17742 }, { "epoch": 1.9235689505637468, "grad_norm": 0.4120662808418274, "learning_rate": 7.2022341505875536e-06, "loss": 0.0102, "step": 17743 }, { "epoch": 1.9236773633998265, "grad_norm": 0.317881315946579, "learning_rate": 7.2015087770201666e-06, "loss": 0.0029, "step": 17744 }, { "epoch": 1.9237857762359063, "grad_norm": 1.676861047744751, "learning_rate": 7.200783403452779e-06, "loss": 0.0092, "step": 17745 }, { "epoch": 1.9238941890719863, "grad_norm": 2.2697277069091797, "learning_rate": 7.200058029885392e-06, "loss": 0.0122, "step": 17746 }, { "epoch": 1.9240026019080658, "grad_norm": 0.38446691632270813, "learning_rate": 7.199332656318005e-06, "loss": 0.014, "step": 17747 }, { "epoch": 1.9241110147441458, "grad_norm": 3.7955517768859863, "learning_rate": 7.198607282750617e-06, "loss": 0.0161, "step": 17748 }, { "epoch": 1.9242194275802254, "grad_norm": 0.3165431618690491, "learning_rate": 7.19788190918323e-06, "loss": 0.0065, "step": 17749 }, { "epoch": 1.9243278404163053, "grad_norm": 2.3020389080047607, "learning_rate": 7.197156535615843e-06, "loss": 0.0218, "step": 17750 }, { "epoch": 1.924436253252385, "grad_norm": 7.658707618713379, "learning_rate": 7.196431162048455e-06, "loss": 0.0361, "step": 17751 }, { "epoch": 1.9245446660884649, "grad_norm": 4.937124729156494, "learning_rate": 7.195705788481069e-06, "loss": 0.0497, "step": 17752 }, { "epoch": 1.9246530789245446, "grad_norm": 1.908339023590088, "learning_rate": 7.194980414913682e-06, "loss": 0.0085, "step": 17753 }, { "epoch": 1.9247614917606244, "grad_norm": 2.6199018955230713, "learning_rate": 7.194255041346294e-06, "loss": 0.0277, "step": 17754 }, { "epoch": 1.9248699045967044, "grad_norm": 0.021149734035134315, "learning_rate": 7.193529667778907e-06, "loss": 0.0002, "step": 17755 }, { "epoch": 1.924978317432784, "grad_norm": 0.18960970640182495, "learning_rate": 7.19280429421152e-06, "loss": 0.0006, "step": 17756 }, { "epoch": 1.925086730268864, "grad_norm": 1.5498617887496948, "learning_rate": 7.192078920644132e-06, "loss": 0.0351, "step": 17757 }, { "epoch": 1.9251951431049437, "grad_norm": 3.8467063903808594, "learning_rate": 7.191353547076745e-06, "loss": 0.0031, "step": 17758 }, { "epoch": 1.9253035559410234, "grad_norm": 5.892819881439209, "learning_rate": 7.190628173509357e-06, "loss": 0.0445, "step": 17759 }, { "epoch": 1.9254119687771032, "grad_norm": 0.2910963296890259, "learning_rate": 7.18990279994197e-06, "loss": 0.0103, "step": 17760 }, { "epoch": 1.925520381613183, "grad_norm": 0.3356376588344574, "learning_rate": 7.189177426374584e-06, "loss": 0.0048, "step": 17761 }, { "epoch": 1.925628794449263, "grad_norm": 6.825380325317383, "learning_rate": 7.188452052807197e-06, "loss": 0.0507, "step": 17762 }, { "epoch": 1.9257372072853425, "grad_norm": 2.2009971141815186, "learning_rate": 7.187726679239809e-06, "loss": 0.0807, "step": 17763 }, { "epoch": 1.9258456201214225, "grad_norm": 1.6089955568313599, "learning_rate": 7.187001305672422e-06, "loss": 0.0343, "step": 17764 }, { "epoch": 1.925954032957502, "grad_norm": 0.30954495072364807, "learning_rate": 7.186275932105034e-06, "loss": 0.013, "step": 17765 }, { "epoch": 1.926062445793582, "grad_norm": 2.4425904750823975, "learning_rate": 7.185550558537647e-06, "loss": 0.0125, "step": 17766 }, { "epoch": 1.9261708586296618, "grad_norm": 0.645159125328064, "learning_rate": 7.18482518497026e-06, "loss": 0.0243, "step": 17767 }, { "epoch": 1.9262792714657415, "grad_norm": 0.6929438710212708, "learning_rate": 7.184099811402872e-06, "loss": 0.012, "step": 17768 }, { "epoch": 1.9263876843018215, "grad_norm": 0.12192791700363159, "learning_rate": 7.183374437835486e-06, "loss": 0.0005, "step": 17769 }, { "epoch": 1.926496097137901, "grad_norm": 2.1290760040283203, "learning_rate": 7.182649064268099e-06, "loss": 0.0274, "step": 17770 }, { "epoch": 1.926604509973981, "grad_norm": 0.4070360064506531, "learning_rate": 7.181923690700711e-06, "loss": 0.0154, "step": 17771 }, { "epoch": 1.9267129228100606, "grad_norm": 2.312922954559326, "learning_rate": 7.181198317133324e-06, "loss": 0.0308, "step": 17772 }, { "epoch": 1.9268213356461406, "grad_norm": 3.8411459922790527, "learning_rate": 7.180472943565937e-06, "loss": 0.029, "step": 17773 }, { "epoch": 1.9269297484822203, "grad_norm": 5.245419025421143, "learning_rate": 7.1797475699985495e-06, "loss": 0.0468, "step": 17774 }, { "epoch": 1.9270381613183, "grad_norm": 0.3144073188304901, "learning_rate": 7.1790221964311625e-06, "loss": 0.0085, "step": 17775 }, { "epoch": 1.9271465741543798, "grad_norm": 7.487030029296875, "learning_rate": 7.1782968228637755e-06, "loss": 0.0278, "step": 17776 }, { "epoch": 1.9272549869904596, "grad_norm": 0.19193938374519348, "learning_rate": 7.177571449296388e-06, "loss": 0.0023, "step": 17777 }, { "epoch": 1.9273633998265396, "grad_norm": 0.03425540775060654, "learning_rate": 7.1768460757290014e-06, "loss": 0.0003, "step": 17778 }, { "epoch": 1.9274718126626191, "grad_norm": 1.797532081604004, "learning_rate": 7.1761207021616144e-06, "loss": 0.0296, "step": 17779 }, { "epoch": 1.9275802254986991, "grad_norm": 2.0269250869750977, "learning_rate": 7.175395328594227e-06, "loss": 0.0239, "step": 17780 }, { "epoch": 1.9276886383347789, "grad_norm": 1.9273343086242676, "learning_rate": 7.1746699550268396e-06, "loss": 0.0214, "step": 17781 }, { "epoch": 1.9277970511708586, "grad_norm": 6.727912425994873, "learning_rate": 7.1739445814594526e-06, "loss": 0.0286, "step": 17782 }, { "epoch": 1.9279054640069384, "grad_norm": 2.583083391189575, "learning_rate": 7.173219207892065e-06, "loss": 0.0079, "step": 17783 }, { "epoch": 1.9280138768430182, "grad_norm": 1.3504053354263306, "learning_rate": 7.172493834324678e-06, "loss": 0.0113, "step": 17784 }, { "epoch": 1.9281222896790982, "grad_norm": 2.5667567253112793, "learning_rate": 7.17176846075729e-06, "loss": 0.0176, "step": 17785 }, { "epoch": 1.9282307025151777, "grad_norm": 5.931547164916992, "learning_rate": 7.171043087189904e-06, "loss": 0.0319, "step": 17786 }, { "epoch": 1.9283391153512577, "grad_norm": 2.244981050491333, "learning_rate": 7.170317713622517e-06, "loss": 0.0226, "step": 17787 }, { "epoch": 1.9284475281873372, "grad_norm": 4.064821243286133, "learning_rate": 7.16959234005513e-06, "loss": 0.0374, "step": 17788 }, { "epoch": 1.9285559410234172, "grad_norm": 0.35283172130584717, "learning_rate": 7.168866966487742e-06, "loss": 0.0019, "step": 17789 }, { "epoch": 1.928664353859497, "grad_norm": 1.7763296365737915, "learning_rate": 7.168141592920355e-06, "loss": 0.0165, "step": 17790 }, { "epoch": 1.9287727666955767, "grad_norm": 1.8800699710845947, "learning_rate": 7.167416219352967e-06, "loss": 0.0134, "step": 17791 }, { "epoch": 1.9288811795316565, "grad_norm": 4.410950660705566, "learning_rate": 7.16669084578558e-06, "loss": 0.0371, "step": 17792 }, { "epoch": 1.9289895923677363, "grad_norm": 1.4604747295379639, "learning_rate": 7.165965472218193e-06, "loss": 0.0352, "step": 17793 }, { "epoch": 1.9290980052038162, "grad_norm": 1.4182764291763306, "learning_rate": 7.165240098650805e-06, "loss": 0.0385, "step": 17794 }, { "epoch": 1.9292064180398958, "grad_norm": 3.77427339553833, "learning_rate": 7.164514725083419e-06, "loss": 0.0137, "step": 17795 }, { "epoch": 1.9293148308759758, "grad_norm": 5.051896095275879, "learning_rate": 7.163789351516032e-06, "loss": 0.0112, "step": 17796 }, { "epoch": 1.9294232437120555, "grad_norm": 1.6898859739303589, "learning_rate": 7.163063977948644e-06, "loss": 0.0143, "step": 17797 }, { "epoch": 1.9295316565481353, "grad_norm": 0.09339646250009537, "learning_rate": 7.162338604381257e-06, "loss": 0.0006, "step": 17798 }, { "epoch": 1.929640069384215, "grad_norm": 5.322567939758301, "learning_rate": 7.16161323081387e-06, "loss": 0.0679, "step": 17799 }, { "epoch": 1.9297484822202948, "grad_norm": 0.6677443981170654, "learning_rate": 7.160887857246482e-06, "loss": 0.004, "step": 17800 }, { "epoch": 1.9298568950563748, "grad_norm": 1.4935038089752197, "learning_rate": 7.160162483679095e-06, "loss": 0.0019, "step": 17801 }, { "epoch": 1.9299653078924544, "grad_norm": 1.26284658908844, "learning_rate": 7.159437110111708e-06, "loss": 0.0241, "step": 17802 }, { "epoch": 1.9300737207285343, "grad_norm": 0.10249380022287369, "learning_rate": 7.15871173654432e-06, "loss": 0.0013, "step": 17803 }, { "epoch": 1.930182133564614, "grad_norm": 0.33731314539909363, "learning_rate": 7.157986362976934e-06, "loss": 0.0031, "step": 17804 }, { "epoch": 1.9302905464006939, "grad_norm": 3.974757432937622, "learning_rate": 7.157260989409547e-06, "loss": 0.1424, "step": 17805 }, { "epoch": 1.9303989592367736, "grad_norm": 0.6590971350669861, "learning_rate": 7.156535615842159e-06, "loss": 0.0051, "step": 17806 }, { "epoch": 1.9305073720728534, "grad_norm": 2.2999379634857178, "learning_rate": 7.155810242274772e-06, "loss": 0.0098, "step": 17807 }, { "epoch": 1.9306157849089334, "grad_norm": 0.13116097450256348, "learning_rate": 7.155084868707385e-06, "loss": 0.0013, "step": 17808 }, { "epoch": 1.930724197745013, "grad_norm": 0.18690979480743408, "learning_rate": 7.154359495139997e-06, "loss": 0.0014, "step": 17809 }, { "epoch": 1.930832610581093, "grad_norm": 1.6275864839553833, "learning_rate": 7.15363412157261e-06, "loss": 0.0219, "step": 17810 }, { "epoch": 1.9309410234171724, "grad_norm": 0.34578126668930054, "learning_rate": 7.1529087480052225e-06, "loss": 0.0015, "step": 17811 }, { "epoch": 1.9310494362532524, "grad_norm": 0.34175652265548706, "learning_rate": 7.152183374437836e-06, "loss": 0.0123, "step": 17812 }, { "epoch": 1.9311578490893322, "grad_norm": 2.209228992462158, "learning_rate": 7.151458000870449e-06, "loss": 0.0643, "step": 17813 }, { "epoch": 1.931266261925412, "grad_norm": 7.824172019958496, "learning_rate": 7.150732627303062e-06, "loss": 0.0176, "step": 17814 }, { "epoch": 1.9313746747614917, "grad_norm": 0.4461808204650879, "learning_rate": 7.1500072537356745e-06, "loss": 0.0029, "step": 17815 }, { "epoch": 1.9314830875975715, "grad_norm": 2.2744357585906982, "learning_rate": 7.1492818801682875e-06, "loss": 0.0297, "step": 17816 }, { "epoch": 1.9315915004336515, "grad_norm": 0.5905119776725769, "learning_rate": 7.1485565066009e-06, "loss": 0.0088, "step": 17817 }, { "epoch": 1.931699913269731, "grad_norm": 11.970545768737793, "learning_rate": 7.147831133033513e-06, "loss": 0.0736, "step": 17818 }, { "epoch": 1.931808326105811, "grad_norm": 1.8663941621780396, "learning_rate": 7.147105759466126e-06, "loss": 0.0041, "step": 17819 }, { "epoch": 1.9319167389418908, "grad_norm": 1.017825961112976, "learning_rate": 7.146380385898738e-06, "loss": 0.0057, "step": 17820 }, { "epoch": 1.9320251517779705, "grad_norm": 1.811147928237915, "learning_rate": 7.1456550123313516e-06, "loss": 0.0393, "step": 17821 }, { "epoch": 1.9321335646140503, "grad_norm": 3.2477238178253174, "learning_rate": 7.1449296387639646e-06, "loss": 0.087, "step": 17822 }, { "epoch": 1.93224197745013, "grad_norm": 8.402175903320312, "learning_rate": 7.144204265196577e-06, "loss": 0.0232, "step": 17823 }, { "epoch": 1.93235039028621, "grad_norm": 0.07534389942884445, "learning_rate": 7.14347889162919e-06, "loss": 0.0008, "step": 17824 }, { "epoch": 1.9324588031222896, "grad_norm": 3.1357078552246094, "learning_rate": 7.142753518061803e-06, "loss": 0.0686, "step": 17825 }, { "epoch": 1.9325672159583696, "grad_norm": 0.604830265045166, "learning_rate": 7.142028144494415e-06, "loss": 0.0221, "step": 17826 }, { "epoch": 1.932675628794449, "grad_norm": 7.789571285247803, "learning_rate": 7.141302770927028e-06, "loss": 0.0673, "step": 17827 }, { "epoch": 1.932784041630529, "grad_norm": 4.921297550201416, "learning_rate": 7.140577397359641e-06, "loss": 0.0256, "step": 17828 }, { "epoch": 1.9328924544666088, "grad_norm": 3.5274627208709717, "learning_rate": 7.139852023792253e-06, "loss": 0.0291, "step": 17829 }, { "epoch": 1.9330008673026886, "grad_norm": 3.563023090362549, "learning_rate": 7.139126650224867e-06, "loss": 0.016, "step": 17830 }, { "epoch": 1.9331092801387686, "grad_norm": 1.6705775260925293, "learning_rate": 7.13840127665748e-06, "loss": 0.0344, "step": 17831 }, { "epoch": 1.9332176929748481, "grad_norm": 1.9043751955032349, "learning_rate": 7.137675903090092e-06, "loss": 0.0298, "step": 17832 }, { "epoch": 1.9333261058109281, "grad_norm": 0.13145047426223755, "learning_rate": 7.136950529522705e-06, "loss": 0.0016, "step": 17833 }, { "epoch": 1.9334345186470077, "grad_norm": 3.3195314407348633, "learning_rate": 7.136225155955318e-06, "loss": 0.0452, "step": 17834 }, { "epoch": 1.9335429314830876, "grad_norm": 2.1351284980773926, "learning_rate": 7.13549978238793e-06, "loss": 0.0309, "step": 17835 }, { "epoch": 1.9336513443191674, "grad_norm": 1.2030398845672607, "learning_rate": 7.134774408820543e-06, "loss": 0.0273, "step": 17836 }, { "epoch": 1.9337597571552472, "grad_norm": 4.3369526863098145, "learning_rate": 7.134049035253155e-06, "loss": 0.0237, "step": 17837 }, { "epoch": 1.933868169991327, "grad_norm": 0.1340438723564148, "learning_rate": 7.133323661685769e-06, "loss": 0.0011, "step": 17838 }, { "epoch": 1.9339765828274067, "grad_norm": 0.10285194218158722, "learning_rate": 7.132598288118382e-06, "loss": 0.0009, "step": 17839 }, { "epoch": 1.9340849956634867, "grad_norm": 1.6178550720214844, "learning_rate": 7.131872914550995e-06, "loss": 0.0229, "step": 17840 }, { "epoch": 1.9341934084995662, "grad_norm": 4.054009914398193, "learning_rate": 7.131147540983607e-06, "loss": 0.0228, "step": 17841 }, { "epoch": 1.9343018213356462, "grad_norm": 2.8053765296936035, "learning_rate": 7.13042216741622e-06, "loss": 0.0264, "step": 17842 }, { "epoch": 1.934410234171726, "grad_norm": 1.8523370027542114, "learning_rate": 7.129696793848832e-06, "loss": 0.0308, "step": 17843 }, { "epoch": 1.9345186470078057, "grad_norm": 2.4798781871795654, "learning_rate": 7.128971420281445e-06, "loss": 0.0561, "step": 17844 }, { "epoch": 1.9346270598438855, "grad_norm": 1.2485129833221436, "learning_rate": 7.128246046714058e-06, "loss": 0.0163, "step": 17845 }, { "epoch": 1.9347354726799653, "grad_norm": 1.8641122579574585, "learning_rate": 7.12752067314667e-06, "loss": 0.0257, "step": 17846 }, { "epoch": 1.9348438855160452, "grad_norm": 0.27225080132484436, "learning_rate": 7.126795299579284e-06, "loss": 0.0062, "step": 17847 }, { "epoch": 1.9349522983521248, "grad_norm": 0.2148987203836441, "learning_rate": 7.126069926011897e-06, "loss": 0.0037, "step": 17848 }, { "epoch": 1.9350607111882048, "grad_norm": 1.2154881954193115, "learning_rate": 7.125344552444509e-06, "loss": 0.0314, "step": 17849 }, { "epoch": 1.9351691240242843, "grad_norm": 1.9773900508880615, "learning_rate": 7.124619178877122e-06, "loss": 0.0461, "step": 17850 }, { "epoch": 1.9352775368603643, "grad_norm": 4.212021350860596, "learning_rate": 7.123893805309735e-06, "loss": 0.0267, "step": 17851 }, { "epoch": 1.935385949696444, "grad_norm": 0.06033720448613167, "learning_rate": 7.1231684317423475e-06, "loss": 0.0112, "step": 17852 }, { "epoch": 1.9354943625325238, "grad_norm": 1.0818469524383545, "learning_rate": 7.1224430581749605e-06, "loss": 0.0266, "step": 17853 }, { "epoch": 1.9356027753686036, "grad_norm": 6.086261749267578, "learning_rate": 7.1217176846075735e-06, "loss": 0.0919, "step": 17854 }, { "epoch": 1.9357111882046834, "grad_norm": 0.426606684923172, "learning_rate": 7.120992311040186e-06, "loss": 0.0093, "step": 17855 }, { "epoch": 1.9358196010407633, "grad_norm": 4.44518518447876, "learning_rate": 7.1202669374727995e-06, "loss": 0.0459, "step": 17856 }, { "epoch": 1.9359280138768429, "grad_norm": 0.9089160561561584, "learning_rate": 7.1195415639054124e-06, "loss": 0.0356, "step": 17857 }, { "epoch": 1.9360364267129229, "grad_norm": 2.347402334213257, "learning_rate": 7.118816190338025e-06, "loss": 0.0203, "step": 17858 }, { "epoch": 1.9361448395490026, "grad_norm": 0.5821120142936707, "learning_rate": 7.118090816770638e-06, "loss": 0.0154, "step": 17859 }, { "epoch": 1.9362532523850824, "grad_norm": 0.5112180709838867, "learning_rate": 7.1173654432032506e-06, "loss": 0.0149, "step": 17860 }, { "epoch": 1.9363616652211622, "grad_norm": 10.539016723632812, "learning_rate": 7.116640069635863e-06, "loss": 0.0335, "step": 17861 }, { "epoch": 1.936470078057242, "grad_norm": 0.8519940972328186, "learning_rate": 7.115914696068476e-06, "loss": 0.0064, "step": 17862 }, { "epoch": 1.936578490893322, "grad_norm": 1.4539121389389038, "learning_rate": 7.115189322501088e-06, "loss": 0.0198, "step": 17863 }, { "epoch": 1.9366869037294014, "grad_norm": 1.377925992012024, "learning_rate": 7.114463948933702e-06, "loss": 0.0076, "step": 17864 }, { "epoch": 1.9367953165654814, "grad_norm": 1.267730951309204, "learning_rate": 7.113738575366315e-06, "loss": 0.0374, "step": 17865 }, { "epoch": 1.9369037294015612, "grad_norm": 0.4602888524532318, "learning_rate": 7.113013201798928e-06, "loss": 0.0247, "step": 17866 }, { "epoch": 1.937012142237641, "grad_norm": 1.22465980052948, "learning_rate": 7.11228782823154e-06, "loss": 0.0208, "step": 17867 }, { "epoch": 1.9371205550737207, "grad_norm": 0.6139552593231201, "learning_rate": 7.111562454664153e-06, "loss": 0.0183, "step": 17868 }, { "epoch": 1.9372289679098005, "grad_norm": 2.1653892993927, "learning_rate": 7.110837081096765e-06, "loss": 0.0329, "step": 17869 }, { "epoch": 1.9373373807458805, "grad_norm": 0.9990324378013611, "learning_rate": 7.110111707529378e-06, "loss": 0.0273, "step": 17870 }, { "epoch": 1.93744579358196, "grad_norm": 1.2338368892669678, "learning_rate": 7.109386333961991e-06, "loss": 0.0126, "step": 17871 }, { "epoch": 1.93755420641804, "grad_norm": 0.8331577181816101, "learning_rate": 7.108660960394603e-06, "loss": 0.0105, "step": 17872 }, { "epoch": 1.9376626192541195, "grad_norm": 1.2783024311065674, "learning_rate": 7.107935586827217e-06, "loss": 0.0134, "step": 17873 }, { "epoch": 1.9377710320901995, "grad_norm": 1.539586067199707, "learning_rate": 7.10721021325983e-06, "loss": 0.0383, "step": 17874 }, { "epoch": 1.9378794449262793, "grad_norm": 0.7774860262870789, "learning_rate": 7.106484839692442e-06, "loss": 0.0206, "step": 17875 }, { "epoch": 1.937987857762359, "grad_norm": 2.889517068862915, "learning_rate": 7.105759466125055e-06, "loss": 0.018, "step": 17876 }, { "epoch": 1.9380962705984388, "grad_norm": 0.7139133214950562, "learning_rate": 7.105034092557668e-06, "loss": 0.0041, "step": 17877 }, { "epoch": 1.9382046834345186, "grad_norm": 1.3767812252044678, "learning_rate": 7.10430871899028e-06, "loss": 0.0438, "step": 17878 }, { "epoch": 1.9383130962705986, "grad_norm": 0.4870764911174774, "learning_rate": 7.103583345422893e-06, "loss": 0.0178, "step": 17879 }, { "epoch": 1.938421509106678, "grad_norm": 2.707761287689209, "learning_rate": 7.102857971855506e-06, "loss": 0.0369, "step": 17880 }, { "epoch": 1.938529921942758, "grad_norm": 1.6030296087265015, "learning_rate": 7.102132598288118e-06, "loss": 0.0166, "step": 17881 }, { "epoch": 1.9386383347788378, "grad_norm": 3.2071523666381836, "learning_rate": 7.101407224720732e-06, "loss": 0.046, "step": 17882 }, { "epoch": 1.9387467476149176, "grad_norm": 0.3290970027446747, "learning_rate": 7.100681851153345e-06, "loss": 0.0016, "step": 17883 }, { "epoch": 1.9388551604509974, "grad_norm": 0.4052515923976898, "learning_rate": 7.099956477585957e-06, "loss": 0.0106, "step": 17884 }, { "epoch": 1.9389635732870771, "grad_norm": 1.1396433115005493, "learning_rate": 7.09923110401857e-06, "loss": 0.0063, "step": 17885 }, { "epoch": 1.9390719861231571, "grad_norm": 0.5861196517944336, "learning_rate": 7.098505730451183e-06, "loss": 0.025, "step": 17886 }, { "epoch": 1.9391803989592367, "grad_norm": 0.4077323079109192, "learning_rate": 7.097780356883795e-06, "loss": 0.0182, "step": 17887 }, { "epoch": 1.9392888117953166, "grad_norm": 0.7796560525894165, "learning_rate": 7.097054983316408e-06, "loss": 0.0018, "step": 17888 }, { "epoch": 1.9393972246313964, "grad_norm": 7.783126354217529, "learning_rate": 7.0963296097490205e-06, "loss": 0.0831, "step": 17889 }, { "epoch": 1.9395056374674762, "grad_norm": 4.413112640380859, "learning_rate": 7.095604236181634e-06, "loss": 0.0498, "step": 17890 }, { "epoch": 1.939614050303556, "grad_norm": 6.768350601196289, "learning_rate": 7.094878862614247e-06, "loss": 0.029, "step": 17891 }, { "epoch": 1.9397224631396357, "grad_norm": 1.567381501197815, "learning_rate": 7.09415348904686e-06, "loss": 0.0665, "step": 17892 }, { "epoch": 1.9398308759757157, "grad_norm": 4.806014060974121, "learning_rate": 7.0934281154794725e-06, "loss": 0.0688, "step": 17893 }, { "epoch": 1.9399392888117952, "grad_norm": 1.3733723163604736, "learning_rate": 7.0927027419120855e-06, "loss": 0.0202, "step": 17894 }, { "epoch": 1.9400477016478752, "grad_norm": 0.4227800667285919, "learning_rate": 7.091977368344698e-06, "loss": 0.0148, "step": 17895 }, { "epoch": 1.9401561144839548, "grad_norm": 6.8520402908325195, "learning_rate": 7.091251994777311e-06, "loss": 0.0645, "step": 17896 }, { "epoch": 1.9402645273200347, "grad_norm": 0.4509485960006714, "learning_rate": 7.090526621209924e-06, "loss": 0.0051, "step": 17897 }, { "epoch": 1.9403729401561145, "grad_norm": 1.864398717880249, "learning_rate": 7.089801247642536e-06, "loss": 0.0182, "step": 17898 }, { "epoch": 1.9404813529921943, "grad_norm": 1.6053931713104248, "learning_rate": 7.0890758740751496e-06, "loss": 0.0158, "step": 17899 }, { "epoch": 1.940589765828274, "grad_norm": 2.7106399536132812, "learning_rate": 7.0883505005077626e-06, "loss": 0.0236, "step": 17900 }, { "epoch": 1.9406981786643538, "grad_norm": 0.5847746729850769, "learning_rate": 7.087625126940375e-06, "loss": 0.0151, "step": 17901 }, { "epoch": 1.9408065915004338, "grad_norm": 5.919960975646973, "learning_rate": 7.086899753372988e-06, "loss": 0.0703, "step": 17902 }, { "epoch": 1.9409150043365133, "grad_norm": 0.25604048371315, "learning_rate": 7.086174379805601e-06, "loss": 0.0052, "step": 17903 }, { "epoch": 1.9410234171725933, "grad_norm": 8.592806816101074, "learning_rate": 7.085449006238213e-06, "loss": 0.0649, "step": 17904 }, { "epoch": 1.941131830008673, "grad_norm": 3.3240222930908203, "learning_rate": 7.084723632670826e-06, "loss": 0.0548, "step": 17905 }, { "epoch": 1.9412402428447528, "grad_norm": 1.0999730825424194, "learning_rate": 7.083998259103439e-06, "loss": 0.0132, "step": 17906 }, { "epoch": 1.9413486556808326, "grad_norm": 1.8762462139129639, "learning_rate": 7.083272885536052e-06, "loss": 0.0221, "step": 17907 }, { "epoch": 1.9414570685169124, "grad_norm": 0.6001587510108948, "learning_rate": 7.082547511968665e-06, "loss": 0.0143, "step": 17908 }, { "epoch": 1.9415654813529923, "grad_norm": 2.106137990951538, "learning_rate": 7.081822138401278e-06, "loss": 0.0283, "step": 17909 }, { "epoch": 1.9416738941890719, "grad_norm": 0.10646636039018631, "learning_rate": 7.08109676483389e-06, "loss": 0.0017, "step": 17910 }, { "epoch": 1.9417823070251519, "grad_norm": 0.5445530414581299, "learning_rate": 7.080371391266503e-06, "loss": 0.0113, "step": 17911 }, { "epoch": 1.9418907198612314, "grad_norm": 6.69057035446167, "learning_rate": 7.079646017699116e-06, "loss": 0.0326, "step": 17912 }, { "epoch": 1.9419991326973114, "grad_norm": 7.6312255859375, "learning_rate": 7.078920644131728e-06, "loss": 0.0496, "step": 17913 }, { "epoch": 1.9421075455333912, "grad_norm": 0.6794564127922058, "learning_rate": 7.078195270564341e-06, "loss": 0.0141, "step": 17914 }, { "epoch": 1.942215958369471, "grad_norm": 0.5141822099685669, "learning_rate": 7.077469896996953e-06, "loss": 0.0028, "step": 17915 }, { "epoch": 1.942324371205551, "grad_norm": 1.9105221033096313, "learning_rate": 7.076744523429567e-06, "loss": 0.0658, "step": 17916 }, { "epoch": 1.9424327840416304, "grad_norm": 10.197920799255371, "learning_rate": 7.07601914986218e-06, "loss": 0.0529, "step": 17917 }, { "epoch": 1.9425411968777104, "grad_norm": 1.4540891647338867, "learning_rate": 7.075293776294793e-06, "loss": 0.0188, "step": 17918 }, { "epoch": 1.94264960971379, "grad_norm": 1.0540779829025269, "learning_rate": 7.074568402727405e-06, "loss": 0.0069, "step": 17919 }, { "epoch": 1.94275802254987, "grad_norm": 0.9500809907913208, "learning_rate": 7.073843029160018e-06, "loss": 0.0207, "step": 17920 }, { "epoch": 1.9428664353859497, "grad_norm": 0.14155197143554688, "learning_rate": 7.07311765559263e-06, "loss": 0.0011, "step": 17921 }, { "epoch": 1.9429748482220295, "grad_norm": 0.3303461968898773, "learning_rate": 7.072392282025243e-06, "loss": 0.0037, "step": 17922 }, { "epoch": 1.9430832610581092, "grad_norm": 10.115174293518066, "learning_rate": 7.071666908457856e-06, "loss": 0.0266, "step": 17923 }, { "epoch": 1.943191673894189, "grad_norm": 0.08433511853218079, "learning_rate": 7.070941534890468e-06, "loss": 0.0006, "step": 17924 }, { "epoch": 1.943300086730269, "grad_norm": 2.1046156883239746, "learning_rate": 7.070216161323082e-06, "loss": 0.0269, "step": 17925 }, { "epoch": 1.9434084995663485, "grad_norm": 0.3806993067264557, "learning_rate": 7.069490787755695e-06, "loss": 0.016, "step": 17926 }, { "epoch": 1.9435169124024285, "grad_norm": 2.4203500747680664, "learning_rate": 7.068765414188307e-06, "loss": 0.0244, "step": 17927 }, { "epoch": 1.9436253252385083, "grad_norm": 2.251296281814575, "learning_rate": 7.06804004062092e-06, "loss": 0.027, "step": 17928 }, { "epoch": 1.943733738074588, "grad_norm": 1.5364627838134766, "learning_rate": 7.067314667053533e-06, "loss": 0.0123, "step": 17929 }, { "epoch": 1.9438421509106678, "grad_norm": 0.7516927719116211, "learning_rate": 7.0665892934861455e-06, "loss": 0.003, "step": 17930 }, { "epoch": 1.9439505637467476, "grad_norm": 4.111992835998535, "learning_rate": 7.0658639199187585e-06, "loss": 0.0061, "step": 17931 }, { "epoch": 1.9440589765828276, "grad_norm": 0.5345110893249512, "learning_rate": 7.0651385463513715e-06, "loss": 0.0166, "step": 17932 }, { "epoch": 1.944167389418907, "grad_norm": 1.9567410945892334, "learning_rate": 7.0644131727839845e-06, "loss": 0.0192, "step": 17933 }, { "epoch": 1.944275802254987, "grad_norm": 1.9799160957336426, "learning_rate": 7.0636877992165975e-06, "loss": 0.0355, "step": 17934 }, { "epoch": 1.9443842150910666, "grad_norm": 8.780467987060547, "learning_rate": 7.0629624256492105e-06, "loss": 0.0134, "step": 17935 }, { "epoch": 1.9444926279271466, "grad_norm": 0.5075684785842896, "learning_rate": 7.062237052081823e-06, "loss": 0.0034, "step": 17936 }, { "epoch": 1.9446010407632264, "grad_norm": 0.33057913184165955, "learning_rate": 7.061511678514436e-06, "loss": 0.0067, "step": 17937 }, { "epoch": 1.9447094535993061, "grad_norm": 0.4644230902194977, "learning_rate": 7.060786304947049e-06, "loss": 0.0073, "step": 17938 }, { "epoch": 1.944817866435386, "grad_norm": 1.522096037864685, "learning_rate": 7.060060931379661e-06, "loss": 0.0113, "step": 17939 }, { "epoch": 1.9449262792714657, "grad_norm": 0.4164460599422455, "learning_rate": 7.059335557812274e-06, "loss": 0.0155, "step": 17940 }, { "epoch": 1.9450346921075456, "grad_norm": 0.6294049620628357, "learning_rate": 7.058610184244886e-06, "loss": 0.005, "step": 17941 }, { "epoch": 1.9451431049436252, "grad_norm": 2.6346771717071533, "learning_rate": 7.0578848106775e-06, "loss": 0.0189, "step": 17942 }, { "epoch": 1.9452515177797052, "grad_norm": 2.571641445159912, "learning_rate": 7.057159437110113e-06, "loss": 0.031, "step": 17943 }, { "epoch": 1.945359930615785, "grad_norm": 1.6391189098358154, "learning_rate": 7.056434063542726e-06, "loss": 0.0126, "step": 17944 }, { "epoch": 1.9454683434518647, "grad_norm": 2.5809133052825928, "learning_rate": 7.055708689975338e-06, "loss": 0.022, "step": 17945 }, { "epoch": 1.9455767562879445, "grad_norm": 0.681209146976471, "learning_rate": 7.054983316407951e-06, "loss": 0.0035, "step": 17946 }, { "epoch": 1.9456851691240242, "grad_norm": 4.0851969718933105, "learning_rate": 7.054257942840563e-06, "loss": 0.0161, "step": 17947 }, { "epoch": 1.9457935819601042, "grad_norm": 3.8748559951782227, "learning_rate": 7.053532569273176e-06, "loss": 0.0506, "step": 17948 }, { "epoch": 1.9459019947961838, "grad_norm": 1.7124994993209839, "learning_rate": 7.052807195705789e-06, "loss": 0.0145, "step": 17949 }, { "epoch": 1.9460104076322637, "grad_norm": 1.0112624168395996, "learning_rate": 7.052081822138401e-06, "loss": 0.0018, "step": 17950 }, { "epoch": 1.9461188204683435, "grad_norm": 0.4764690697193146, "learning_rate": 7.051356448571015e-06, "loss": 0.002, "step": 17951 }, { "epoch": 1.9462272333044233, "grad_norm": 0.7443278431892395, "learning_rate": 7.050631075003628e-06, "loss": 0.0106, "step": 17952 }, { "epoch": 1.946335646140503, "grad_norm": 2.0057995319366455, "learning_rate": 7.04990570143624e-06, "loss": 0.0218, "step": 17953 }, { "epoch": 1.9464440589765828, "grad_norm": 0.3140258193016052, "learning_rate": 7.049180327868853e-06, "loss": 0.0117, "step": 17954 }, { "epoch": 1.9465524718126628, "grad_norm": 0.3432687819004059, "learning_rate": 7.048454954301466e-06, "loss": 0.0059, "step": 17955 }, { "epoch": 1.9466608846487423, "grad_norm": 0.3410005271434784, "learning_rate": 7.047729580734078e-06, "loss": 0.0063, "step": 17956 }, { "epoch": 1.9467692974848223, "grad_norm": 0.8208169937133789, "learning_rate": 7.047004207166691e-06, "loss": 0.002, "step": 17957 }, { "epoch": 1.9468777103209018, "grad_norm": 0.42632925510406494, "learning_rate": 7.046278833599304e-06, "loss": 0.0135, "step": 17958 }, { "epoch": 1.9469861231569818, "grad_norm": 2.3502886295318604, "learning_rate": 7.045553460031917e-06, "loss": 0.0304, "step": 17959 }, { "epoch": 1.9470945359930616, "grad_norm": 16.028886795043945, "learning_rate": 7.04482808646453e-06, "loss": 0.0291, "step": 17960 }, { "epoch": 1.9472029488291414, "grad_norm": 2.0101757049560547, "learning_rate": 7.044102712897143e-06, "loss": 0.0611, "step": 17961 }, { "epoch": 1.9473113616652211, "grad_norm": 3.061547040939331, "learning_rate": 7.043377339329755e-06, "loss": 0.0541, "step": 17962 }, { "epoch": 1.9474197745013009, "grad_norm": 0.6436639428138733, "learning_rate": 7.042651965762368e-06, "loss": 0.0142, "step": 17963 }, { "epoch": 1.9475281873373809, "grad_norm": 1.3682308197021484, "learning_rate": 7.041926592194981e-06, "loss": 0.0022, "step": 17964 }, { "epoch": 1.9476366001734604, "grad_norm": 2.4951493740081787, "learning_rate": 7.041201218627593e-06, "loss": 0.0358, "step": 17965 }, { "epoch": 1.9477450130095404, "grad_norm": 8.724632263183594, "learning_rate": 7.040475845060206e-06, "loss": 0.0563, "step": 17966 }, { "epoch": 1.9478534258456202, "grad_norm": 0.33534595370292664, "learning_rate": 7.0397504714928185e-06, "loss": 0.0115, "step": 17967 }, { "epoch": 1.9479618386817, "grad_norm": 2.937208414077759, "learning_rate": 7.039025097925432e-06, "loss": 0.0552, "step": 17968 }, { "epoch": 1.9480702515177797, "grad_norm": 0.46649351716041565, "learning_rate": 7.038299724358045e-06, "loss": 0.0025, "step": 17969 }, { "epoch": 1.9481786643538594, "grad_norm": 0.0797097310423851, "learning_rate": 7.037574350790658e-06, "loss": 0.0005, "step": 17970 }, { "epoch": 1.9482870771899394, "grad_norm": 3.1803128719329834, "learning_rate": 7.0368489772232705e-06, "loss": 0.0332, "step": 17971 }, { "epoch": 1.948395490026019, "grad_norm": 1.0413906574249268, "learning_rate": 7.0361236036558835e-06, "loss": 0.0095, "step": 17972 }, { "epoch": 1.948503902862099, "grad_norm": 2.8042612075805664, "learning_rate": 7.035398230088496e-06, "loss": 0.0488, "step": 17973 }, { "epoch": 1.9486123156981785, "grad_norm": 2.9372317790985107, "learning_rate": 7.034672856521109e-06, "loss": 0.0173, "step": 17974 }, { "epoch": 1.9487207285342585, "grad_norm": 0.3586968183517456, "learning_rate": 7.033947482953722e-06, "loss": 0.0076, "step": 17975 }, { "epoch": 1.9488291413703382, "grad_norm": 4.507453441619873, "learning_rate": 7.033222109386334e-06, "loss": 0.0209, "step": 17976 }, { "epoch": 1.948937554206418, "grad_norm": 2.1356022357940674, "learning_rate": 7.032496735818948e-06, "loss": 0.0442, "step": 17977 }, { "epoch": 1.949045967042498, "grad_norm": 7.573163032531738, "learning_rate": 7.0317713622515606e-06, "loss": 0.0259, "step": 17978 }, { "epoch": 1.9491543798785775, "grad_norm": 2.4779977798461914, "learning_rate": 7.031045988684173e-06, "loss": 0.0267, "step": 17979 }, { "epoch": 1.9492627927146575, "grad_norm": 0.421432763338089, "learning_rate": 7.030320615116786e-06, "loss": 0.0171, "step": 17980 }, { "epoch": 1.949371205550737, "grad_norm": 0.7578642964363098, "learning_rate": 7.029595241549399e-06, "loss": 0.0043, "step": 17981 }, { "epoch": 1.949479618386817, "grad_norm": 14.488973617553711, "learning_rate": 7.028869867982011e-06, "loss": 0.0398, "step": 17982 }, { "epoch": 1.9495880312228968, "grad_norm": 1.8714479207992554, "learning_rate": 7.028144494414624e-06, "loss": 0.0243, "step": 17983 }, { "epoch": 1.9496964440589766, "grad_norm": 2.7462689876556396, "learning_rate": 7.027419120847237e-06, "loss": 0.0064, "step": 17984 }, { "epoch": 1.9498048568950563, "grad_norm": 8.422550201416016, "learning_rate": 7.02669374727985e-06, "loss": 0.0296, "step": 17985 }, { "epoch": 1.949913269731136, "grad_norm": 0.36446377635002136, "learning_rate": 7.025968373712463e-06, "loss": 0.0072, "step": 17986 }, { "epoch": 1.950021682567216, "grad_norm": 0.06657546758651733, "learning_rate": 7.025243000145076e-06, "loss": 0.0002, "step": 17987 }, { "epoch": 1.9501300954032956, "grad_norm": 5.419763565063477, "learning_rate": 7.024517626577688e-06, "loss": 0.0163, "step": 17988 }, { "epoch": 1.9502385082393756, "grad_norm": 27.862810134887695, "learning_rate": 7.023792253010301e-06, "loss": 0.0418, "step": 17989 }, { "epoch": 1.9503469210754554, "grad_norm": 1.1672452688217163, "learning_rate": 7.023066879442914e-06, "loss": 0.018, "step": 17990 }, { "epoch": 1.9504553339115351, "grad_norm": 0.6128671169281006, "learning_rate": 7.022341505875526e-06, "loss": 0.0179, "step": 17991 }, { "epoch": 1.950563746747615, "grad_norm": 5.794752597808838, "learning_rate": 7.021616132308139e-06, "loss": 0.0234, "step": 17992 }, { "epoch": 1.9506721595836947, "grad_norm": 2.537470579147339, "learning_rate": 7.020890758740751e-06, "loss": 0.0388, "step": 17993 }, { "epoch": 1.9507805724197746, "grad_norm": 1.0493804216384888, "learning_rate": 7.020165385173365e-06, "loss": 0.0212, "step": 17994 }, { "epoch": 1.9508889852558542, "grad_norm": 1.6701246500015259, "learning_rate": 7.019440011605978e-06, "loss": 0.0178, "step": 17995 }, { "epoch": 1.9509973980919342, "grad_norm": 5.011599540710449, "learning_rate": 7.018714638038591e-06, "loss": 0.0087, "step": 17996 }, { "epoch": 1.9511058109280137, "grad_norm": 3.317749500274658, "learning_rate": 7.017989264471203e-06, "loss": 0.0474, "step": 17997 }, { "epoch": 1.9512142237640937, "grad_norm": 0.8722836971282959, "learning_rate": 7.017263890903816e-06, "loss": 0.0252, "step": 17998 }, { "epoch": 1.9513226366001735, "grad_norm": 0.3365720212459564, "learning_rate": 7.016538517336428e-06, "loss": 0.0079, "step": 17999 }, { "epoch": 1.9514310494362532, "grad_norm": 4.098448276519775, "learning_rate": 7.015813143769041e-06, "loss": 0.0338, "step": 18000 }, { "epoch": 1.951539462272333, "grad_norm": 0.42040959000587463, "learning_rate": 7.015087770201654e-06, "loss": 0.0074, "step": 18001 }, { "epoch": 1.9516478751084128, "grad_norm": 5.5074968338012695, "learning_rate": 7.014362396634268e-06, "loss": 0.0369, "step": 18002 }, { "epoch": 1.9517562879444927, "grad_norm": 1.109816551208496, "learning_rate": 7.01363702306688e-06, "loss": 0.0033, "step": 18003 }, { "epoch": 1.9518647007805723, "grad_norm": 0.3371262550354004, "learning_rate": 7.012911649499493e-06, "loss": 0.0078, "step": 18004 }, { "epoch": 1.9519731136166523, "grad_norm": 2.655045986175537, "learning_rate": 7.012186275932105e-06, "loss": 0.0172, "step": 18005 }, { "epoch": 1.952081526452732, "grad_norm": 1.5900228023529053, "learning_rate": 7.011460902364718e-06, "loss": 0.0236, "step": 18006 }, { "epoch": 1.9521899392888118, "grad_norm": 2.564786672592163, "learning_rate": 7.010735528797331e-06, "loss": 0.0491, "step": 18007 }, { "epoch": 1.9522983521248916, "grad_norm": 2.2516257762908936, "learning_rate": 7.0100101552299435e-06, "loss": 0.0144, "step": 18008 }, { "epoch": 1.9524067649609713, "grad_norm": 1.626622200012207, "learning_rate": 7.0092847816625565e-06, "loss": 0.032, "step": 18009 }, { "epoch": 1.9525151777970513, "grad_norm": 1.9479432106018066, "learning_rate": 7.0085594080951695e-06, "loss": 0.0115, "step": 18010 }, { "epoch": 1.9526235906331308, "grad_norm": 1.4928902387619019, "learning_rate": 7.0078340345277825e-06, "loss": 0.0303, "step": 18011 }, { "epoch": 1.9527320034692108, "grad_norm": 0.4280032217502594, "learning_rate": 7.0071086609603955e-06, "loss": 0.0031, "step": 18012 }, { "epoch": 1.9528404163052906, "grad_norm": 1.5415548086166382, "learning_rate": 7.0063832873930085e-06, "loss": 0.0218, "step": 18013 }, { "epoch": 1.9529488291413704, "grad_norm": 1.856990098953247, "learning_rate": 7.005657913825621e-06, "loss": 0.019, "step": 18014 }, { "epoch": 1.9530572419774501, "grad_norm": 0.5252339243888855, "learning_rate": 7.004932540258234e-06, "loss": 0.0115, "step": 18015 }, { "epoch": 1.9531656548135299, "grad_norm": 30.772520065307617, "learning_rate": 7.004207166690847e-06, "loss": 0.0479, "step": 18016 }, { "epoch": 1.9532740676496099, "grad_norm": 2.93784499168396, "learning_rate": 7.003481793123459e-06, "loss": 0.0506, "step": 18017 }, { "epoch": 1.9533824804856894, "grad_norm": 2.6187777519226074, "learning_rate": 7.002756419556072e-06, "loss": 0.0213, "step": 18018 }, { "epoch": 1.9534908933217694, "grad_norm": 0.8966375589370728, "learning_rate": 7.002031045988684e-06, "loss": 0.0063, "step": 18019 }, { "epoch": 1.953599306157849, "grad_norm": 8.16641616821289, "learning_rate": 7.001305672421298e-06, "loss": 0.0247, "step": 18020 }, { "epoch": 1.953707718993929, "grad_norm": 9.04319953918457, "learning_rate": 7.000580298853911e-06, "loss": 0.014, "step": 18021 }, { "epoch": 1.9538161318300087, "grad_norm": 1.4833004474639893, "learning_rate": 6.999854925286524e-06, "loss": 0.0103, "step": 18022 }, { "epoch": 1.9539245446660884, "grad_norm": 6.506290435791016, "learning_rate": 6.999129551719136e-06, "loss": 0.0808, "step": 18023 }, { "epoch": 1.9540329575021682, "grad_norm": 0.34509363770484924, "learning_rate": 6.998404178151749e-06, "loss": 0.0072, "step": 18024 }, { "epoch": 1.954141370338248, "grad_norm": 5.377808094024658, "learning_rate": 6.997678804584361e-06, "loss": 0.0432, "step": 18025 }, { "epoch": 1.954249783174328, "grad_norm": 7.4040846824646, "learning_rate": 6.996953431016974e-06, "loss": 0.0496, "step": 18026 }, { "epoch": 1.9543581960104075, "grad_norm": 0.8745601773262024, "learning_rate": 6.996228057449587e-06, "loss": 0.0144, "step": 18027 }, { "epoch": 1.9544666088464875, "grad_norm": 5.071586608886719, "learning_rate": 6.995502683882201e-06, "loss": 0.0371, "step": 18028 }, { "epoch": 1.9545750216825672, "grad_norm": 1.3173741102218628, "learning_rate": 6.994777310314813e-06, "loss": 0.0018, "step": 18029 }, { "epoch": 1.954683434518647, "grad_norm": 8.863094329833984, "learning_rate": 6.994051936747426e-06, "loss": 0.0406, "step": 18030 }, { "epoch": 1.9547918473547268, "grad_norm": 1.696510672569275, "learning_rate": 6.993326563180038e-06, "loss": 0.0088, "step": 18031 }, { "epoch": 1.9549002601908065, "grad_norm": 1.8339518308639526, "learning_rate": 6.992601189612651e-06, "loss": 0.0105, "step": 18032 }, { "epoch": 1.9550086730268865, "grad_norm": 1.1681170463562012, "learning_rate": 6.991875816045264e-06, "loss": 0.0028, "step": 18033 }, { "epoch": 1.955117085862966, "grad_norm": 4.36221981048584, "learning_rate": 6.991150442477876e-06, "loss": 0.0035, "step": 18034 }, { "epoch": 1.955225498699046, "grad_norm": 6.325088977813721, "learning_rate": 6.990425068910489e-06, "loss": 0.0358, "step": 18035 }, { "epoch": 1.9553339115351258, "grad_norm": 0.5588573813438416, "learning_rate": 6.989699695343102e-06, "loss": 0.0135, "step": 18036 }, { "epoch": 1.9554423243712056, "grad_norm": 2.4358372688293457, "learning_rate": 6.988974321775715e-06, "loss": 0.041, "step": 18037 }, { "epoch": 1.9555507372072853, "grad_norm": 1.8778645992279053, "learning_rate": 6.988248948208328e-06, "loss": 0.0202, "step": 18038 }, { "epoch": 1.955659150043365, "grad_norm": 5.265295505523682, "learning_rate": 6.987523574640941e-06, "loss": 0.0601, "step": 18039 }, { "epoch": 1.955767562879445, "grad_norm": 2.652597665786743, "learning_rate": 6.986798201073553e-06, "loss": 0.0023, "step": 18040 }, { "epoch": 1.9558759757155246, "grad_norm": 1.2332792282104492, "learning_rate": 6.986072827506166e-06, "loss": 0.0425, "step": 18041 }, { "epoch": 1.9559843885516046, "grad_norm": 0.3535912036895752, "learning_rate": 6.985347453938779e-06, "loss": 0.008, "step": 18042 }, { "epoch": 1.9560928013876842, "grad_norm": 0.4257231056690216, "learning_rate": 6.984622080371391e-06, "loss": 0.0125, "step": 18043 }, { "epoch": 1.9562012142237641, "grad_norm": 0.6423915028572083, "learning_rate": 6.983896706804004e-06, "loss": 0.0041, "step": 18044 }, { "epoch": 1.956309627059844, "grad_norm": 2.55403995513916, "learning_rate": 6.9831713332366165e-06, "loss": 0.0293, "step": 18045 }, { "epoch": 1.9564180398959237, "grad_norm": 0.7723326086997986, "learning_rate": 6.98244595966923e-06, "loss": 0.0177, "step": 18046 }, { "epoch": 1.9565264527320034, "grad_norm": 0.8186415433883667, "learning_rate": 6.981720586101843e-06, "loss": 0.0095, "step": 18047 }, { "epoch": 1.9566348655680832, "grad_norm": 2.9037163257598877, "learning_rate": 6.980995212534456e-06, "loss": 0.0304, "step": 18048 }, { "epoch": 1.9567432784041632, "grad_norm": 5.480710506439209, "learning_rate": 6.9802698389670685e-06, "loss": 0.0162, "step": 18049 }, { "epoch": 1.9568516912402427, "grad_norm": 0.4524533450603485, "learning_rate": 6.9795444653996815e-06, "loss": 0.0161, "step": 18050 }, { "epoch": 1.9569601040763227, "grad_norm": 5.666116237640381, "learning_rate": 6.978819091832294e-06, "loss": 0.0469, "step": 18051 }, { "epoch": 1.9570685169124025, "grad_norm": 0.49612510204315186, "learning_rate": 6.978093718264907e-06, "loss": 0.0135, "step": 18052 }, { "epoch": 1.9571769297484822, "grad_norm": 5.059928894042969, "learning_rate": 6.97736834469752e-06, "loss": 0.0355, "step": 18053 }, { "epoch": 1.957285342584562, "grad_norm": 0.4355916678905487, "learning_rate": 6.9766429711301334e-06, "loss": 0.0076, "step": 18054 }, { "epoch": 1.9573937554206418, "grad_norm": 0.7033308744430542, "learning_rate": 6.975917597562746e-06, "loss": 0.0248, "step": 18055 }, { "epoch": 1.9575021682567217, "grad_norm": 7.41295862197876, "learning_rate": 6.975192223995359e-06, "loss": 0.0473, "step": 18056 }, { "epoch": 1.9576105810928013, "grad_norm": 0.39178046584129333, "learning_rate": 6.974466850427971e-06, "loss": 0.011, "step": 18057 }, { "epoch": 1.9577189939288813, "grad_norm": 3.9135420322418213, "learning_rate": 6.973741476860584e-06, "loss": 0.0161, "step": 18058 }, { "epoch": 1.9578274067649608, "grad_norm": 1.098740816116333, "learning_rate": 6.973016103293197e-06, "loss": 0.0214, "step": 18059 }, { "epoch": 1.9579358196010408, "grad_norm": 1.4772109985351562, "learning_rate": 6.972290729725809e-06, "loss": 0.0331, "step": 18060 }, { "epoch": 1.9580442324371206, "grad_norm": 0.3198680579662323, "learning_rate": 6.971565356158422e-06, "loss": 0.0104, "step": 18061 }, { "epoch": 1.9581526452732003, "grad_norm": 5.464626312255859, "learning_rate": 6.970839982591035e-06, "loss": 0.0396, "step": 18062 }, { "epoch": 1.95826105810928, "grad_norm": 0.36457473039627075, "learning_rate": 6.970114609023648e-06, "loss": 0.0116, "step": 18063 }, { "epoch": 1.9583694709453598, "grad_norm": 2.2051641941070557, "learning_rate": 6.969389235456261e-06, "loss": 0.027, "step": 18064 }, { "epoch": 1.9584778837814398, "grad_norm": 2.639961004257202, "learning_rate": 6.968663861888874e-06, "loss": 0.0117, "step": 18065 }, { "epoch": 1.9585862966175194, "grad_norm": 0.419223815202713, "learning_rate": 6.967938488321486e-06, "loss": 0.0107, "step": 18066 }, { "epoch": 1.9586947094535994, "grad_norm": 7.99282693862915, "learning_rate": 6.967213114754099e-06, "loss": 0.0052, "step": 18067 }, { "epoch": 1.9588031222896791, "grad_norm": 0.8299877047538757, "learning_rate": 6.966487741186712e-06, "loss": 0.019, "step": 18068 }, { "epoch": 1.9589115351257589, "grad_norm": 0.9253062605857849, "learning_rate": 6.965762367619324e-06, "loss": 0.0162, "step": 18069 }, { "epoch": 1.9590199479618386, "grad_norm": 0.219559907913208, "learning_rate": 6.965036994051937e-06, "loss": 0.0031, "step": 18070 }, { "epoch": 1.9591283607979184, "grad_norm": 0.7009627819061279, "learning_rate": 6.964311620484549e-06, "loss": 0.0156, "step": 18071 }, { "epoch": 1.9592367736339984, "grad_norm": 3.9192681312561035, "learning_rate": 6.963586246917163e-06, "loss": 0.0531, "step": 18072 }, { "epoch": 1.959345186470078, "grad_norm": 0.4931579828262329, "learning_rate": 6.962860873349776e-06, "loss": 0.0055, "step": 18073 }, { "epoch": 1.959453599306158, "grad_norm": 2.195791721343994, "learning_rate": 6.962135499782389e-06, "loss": 0.0204, "step": 18074 }, { "epoch": 1.9595620121422377, "grad_norm": 12.054974555969238, "learning_rate": 6.961410126215001e-06, "loss": 0.0307, "step": 18075 }, { "epoch": 1.9596704249783174, "grad_norm": 0.5499779582023621, "learning_rate": 6.960684752647614e-06, "loss": 0.0078, "step": 18076 }, { "epoch": 1.9597788378143972, "grad_norm": 1.1913855075836182, "learning_rate": 6.959959379080226e-06, "loss": 0.012, "step": 18077 }, { "epoch": 1.959887250650477, "grad_norm": 1.39149010181427, "learning_rate": 6.959234005512839e-06, "loss": 0.0188, "step": 18078 }, { "epoch": 1.959995663486557, "grad_norm": 1.0852737426757812, "learning_rate": 6.958508631945452e-06, "loss": 0.0358, "step": 18079 }, { "epoch": 1.9601040763226365, "grad_norm": 2.8083364963531494, "learning_rate": 6.957783258378066e-06, "loss": 0.0593, "step": 18080 }, { "epoch": 1.9602124891587165, "grad_norm": 0.08576326072216034, "learning_rate": 6.957057884810678e-06, "loss": 0.0009, "step": 18081 }, { "epoch": 1.960320901994796, "grad_norm": 0.3214322328567505, "learning_rate": 6.956332511243291e-06, "loss": 0.0071, "step": 18082 }, { "epoch": 1.960429314830876, "grad_norm": 0.6109359860420227, "learning_rate": 6.955607137675903e-06, "loss": 0.0034, "step": 18083 }, { "epoch": 1.9605377276669558, "grad_norm": 5.1210150718688965, "learning_rate": 6.954881764108516e-06, "loss": 0.0106, "step": 18084 }, { "epoch": 1.9606461405030355, "grad_norm": 2.961550712585449, "learning_rate": 6.954156390541129e-06, "loss": 0.0264, "step": 18085 }, { "epoch": 1.9607545533391153, "grad_norm": 0.3082554042339325, "learning_rate": 6.9534310169737415e-06, "loss": 0.0061, "step": 18086 }, { "epoch": 1.960862966175195, "grad_norm": 2.485197067260742, "learning_rate": 6.9527056434063545e-06, "loss": 0.0554, "step": 18087 }, { "epoch": 1.960971379011275, "grad_norm": 5.365561008453369, "learning_rate": 6.9519802698389675e-06, "loss": 0.0178, "step": 18088 }, { "epoch": 1.9610797918473546, "grad_norm": 5.969494342803955, "learning_rate": 6.9512548962715805e-06, "loss": 0.0715, "step": 18089 }, { "epoch": 1.9611882046834346, "grad_norm": 1.9206246137619019, "learning_rate": 6.9505295227041935e-06, "loss": 0.0325, "step": 18090 }, { "epoch": 1.9612966175195143, "grad_norm": 2.9812867641448975, "learning_rate": 6.9498041491368065e-06, "loss": 0.0506, "step": 18091 }, { "epoch": 1.961405030355594, "grad_norm": 0.9657650589942932, "learning_rate": 6.949078775569419e-06, "loss": 0.0049, "step": 18092 }, { "epoch": 1.9615134431916739, "grad_norm": 0.4915368854999542, "learning_rate": 6.948353402002032e-06, "loss": 0.0131, "step": 18093 }, { "epoch": 1.9616218560277536, "grad_norm": 0.5185753703117371, "learning_rate": 6.947628028434645e-06, "loss": 0.0313, "step": 18094 }, { "epoch": 1.9617302688638336, "grad_norm": 1.2641950845718384, "learning_rate": 6.946902654867257e-06, "loss": 0.0024, "step": 18095 }, { "epoch": 1.9618386816999132, "grad_norm": 5.219112873077393, "learning_rate": 6.94617728129987e-06, "loss": 0.0226, "step": 18096 }, { "epoch": 1.9619470945359931, "grad_norm": 0.5410592555999756, "learning_rate": 6.9454519077324836e-06, "loss": 0.0102, "step": 18097 }, { "epoch": 1.962055507372073, "grad_norm": 0.42284467816352844, "learning_rate": 6.944726534165096e-06, "loss": 0.0041, "step": 18098 }, { "epoch": 1.9621639202081527, "grad_norm": 0.8554395437240601, "learning_rate": 6.944001160597709e-06, "loss": 0.0048, "step": 18099 }, { "epoch": 1.9622723330442324, "grad_norm": 4.214117050170898, "learning_rate": 6.943275787030322e-06, "loss": 0.0208, "step": 18100 }, { "epoch": 1.9623807458803122, "grad_norm": 0.7054004669189453, "learning_rate": 6.942550413462934e-06, "loss": 0.0023, "step": 18101 }, { "epoch": 1.9624891587163922, "grad_norm": 2.206554651260376, "learning_rate": 6.941825039895547e-06, "loss": 0.0129, "step": 18102 }, { "epoch": 1.9625975715524717, "grad_norm": 1.2863075733184814, "learning_rate": 6.941099666328159e-06, "loss": 0.0138, "step": 18103 }, { "epoch": 1.9627059843885517, "grad_norm": 0.6052761077880859, "learning_rate": 6.940374292760772e-06, "loss": 0.0086, "step": 18104 }, { "epoch": 1.9628143972246312, "grad_norm": 4.065529823303223, "learning_rate": 6.939648919193385e-06, "loss": 0.0293, "step": 18105 }, { "epoch": 1.9629228100607112, "grad_norm": 6.638633728027344, "learning_rate": 6.938923545625999e-06, "loss": 0.0315, "step": 18106 }, { "epoch": 1.963031222896791, "grad_norm": 2.0459463596343994, "learning_rate": 6.938198172058611e-06, "loss": 0.0225, "step": 18107 }, { "epoch": 1.9631396357328708, "grad_norm": 9.370384216308594, "learning_rate": 6.937472798491224e-06, "loss": 0.0235, "step": 18108 }, { "epoch": 1.9632480485689505, "grad_norm": 3.1950325965881348, "learning_rate": 6.936747424923836e-06, "loss": 0.0437, "step": 18109 }, { "epoch": 1.9633564614050303, "grad_norm": 2.575000047683716, "learning_rate": 6.936022051356449e-06, "loss": 0.02, "step": 18110 }, { "epoch": 1.9634648742411103, "grad_norm": 1.2241226434707642, "learning_rate": 6.935296677789062e-06, "loss": 0.0285, "step": 18111 }, { "epoch": 1.9635732870771898, "grad_norm": 0.7820103764533997, "learning_rate": 6.934571304221674e-06, "loss": 0.004, "step": 18112 }, { "epoch": 1.9636816999132698, "grad_norm": 3.1956892013549805, "learning_rate": 6.933845930654287e-06, "loss": 0.0676, "step": 18113 }, { "epoch": 1.9637901127493496, "grad_norm": 5.548873424530029, "learning_rate": 6.9331205570869e-06, "loss": 0.0353, "step": 18114 }, { "epoch": 1.9638985255854293, "grad_norm": 0.1717398315668106, "learning_rate": 6.932395183519513e-06, "loss": 0.0009, "step": 18115 }, { "epoch": 1.964006938421509, "grad_norm": 0.052410345524549484, "learning_rate": 6.931669809952126e-06, "loss": 0.0003, "step": 18116 }, { "epoch": 1.9641153512575888, "grad_norm": 0.4059673845767975, "learning_rate": 6.930944436384739e-06, "loss": 0.0103, "step": 18117 }, { "epoch": 1.9642237640936688, "grad_norm": 1.2996454238891602, "learning_rate": 6.930219062817351e-06, "loss": 0.0078, "step": 18118 }, { "epoch": 1.9643321769297484, "grad_norm": 0.033111680299043655, "learning_rate": 6.929493689249964e-06, "loss": 0.0002, "step": 18119 }, { "epoch": 1.9644405897658284, "grad_norm": 4.698306083679199, "learning_rate": 6.928768315682577e-06, "loss": 0.0283, "step": 18120 }, { "epoch": 1.964549002601908, "grad_norm": 4.339698314666748, "learning_rate": 6.928042942115189e-06, "loss": 0.0411, "step": 18121 }, { "epoch": 1.9646574154379879, "grad_norm": 4.807851314544678, "learning_rate": 6.927317568547802e-06, "loss": 0.065, "step": 18122 }, { "epoch": 1.9647658282740676, "grad_norm": 2.6080970764160156, "learning_rate": 6.926592194980416e-06, "loss": 0.0192, "step": 18123 }, { "epoch": 1.9648742411101474, "grad_norm": 0.03788890689611435, "learning_rate": 6.925866821413028e-06, "loss": 0.0003, "step": 18124 }, { "epoch": 1.9649826539462274, "grad_norm": 1.1758205890655518, "learning_rate": 6.925141447845641e-06, "loss": 0.0329, "step": 18125 }, { "epoch": 1.965091066782307, "grad_norm": 2.256470203399658, "learning_rate": 6.924416074278254e-06, "loss": 0.0285, "step": 18126 }, { "epoch": 1.965199479618387, "grad_norm": 0.5771596431732178, "learning_rate": 6.9236907007108665e-06, "loss": 0.0126, "step": 18127 }, { "epoch": 1.9653078924544665, "grad_norm": 0.7194493412971497, "learning_rate": 6.9229653271434795e-06, "loss": 0.0019, "step": 18128 }, { "epoch": 1.9654163052905465, "grad_norm": 3.158522844314575, "learning_rate": 6.922239953576092e-06, "loss": 0.0199, "step": 18129 }, { "epoch": 1.9655247181266262, "grad_norm": 2.178842306137085, "learning_rate": 6.921514580008705e-06, "loss": 0.022, "step": 18130 }, { "epoch": 1.965633130962706, "grad_norm": 5.523011207580566, "learning_rate": 6.920789206441318e-06, "loss": 0.0404, "step": 18131 }, { "epoch": 1.9657415437987857, "grad_norm": 0.3612278997898102, "learning_rate": 6.9200638328739314e-06, "loss": 0.0138, "step": 18132 }, { "epoch": 1.9658499566348655, "grad_norm": 0.039458923041820526, "learning_rate": 6.919338459306544e-06, "loss": 0.0004, "step": 18133 }, { "epoch": 1.9659583694709455, "grad_norm": 3.7197718620300293, "learning_rate": 6.918613085739157e-06, "loss": 0.0419, "step": 18134 }, { "epoch": 1.966066782307025, "grad_norm": 2.8159947395324707, "learning_rate": 6.917887712171769e-06, "loss": 0.0209, "step": 18135 }, { "epoch": 1.966175195143105, "grad_norm": 4.312593460083008, "learning_rate": 6.917162338604382e-06, "loss": 0.0367, "step": 18136 }, { "epoch": 1.9662836079791848, "grad_norm": 0.10100182145833969, "learning_rate": 6.916436965036995e-06, "loss": 0.0011, "step": 18137 }, { "epoch": 1.9663920208152645, "grad_norm": 1.4087512493133545, "learning_rate": 6.915711591469607e-06, "loss": 0.0183, "step": 18138 }, { "epoch": 1.9665004336513443, "grad_norm": 8.435134887695312, "learning_rate": 6.91498621790222e-06, "loss": 0.0138, "step": 18139 }, { "epoch": 1.966608846487424, "grad_norm": 2.1310982704162598, "learning_rate": 6.914260844334833e-06, "loss": 0.0056, "step": 18140 }, { "epoch": 1.966717259323504, "grad_norm": 1.9966455698013306, "learning_rate": 6.913535470767446e-06, "loss": 0.0213, "step": 18141 }, { "epoch": 1.9668256721595836, "grad_norm": 0.4829460084438324, "learning_rate": 6.912810097200059e-06, "loss": 0.0085, "step": 18142 }, { "epoch": 1.9669340849956636, "grad_norm": 0.2704712748527527, "learning_rate": 6.912084723632672e-06, "loss": 0.006, "step": 18143 }, { "epoch": 1.9670424978317431, "grad_norm": 7.354250431060791, "learning_rate": 6.911359350065284e-06, "loss": 0.0762, "step": 18144 }, { "epoch": 1.967150910667823, "grad_norm": 9.321880340576172, "learning_rate": 6.910633976497897e-06, "loss": 0.0348, "step": 18145 }, { "epoch": 1.9672593235039029, "grad_norm": 4.665500640869141, "learning_rate": 6.90990860293051e-06, "loss": 0.0553, "step": 18146 }, { "epoch": 1.9673677363399826, "grad_norm": 0.7639163136482239, "learning_rate": 6.909183229363122e-06, "loss": 0.014, "step": 18147 }, { "epoch": 1.9674761491760624, "grad_norm": 1.946462869644165, "learning_rate": 6.908457855795735e-06, "loss": 0.0301, "step": 18148 }, { "epoch": 1.9675845620121422, "grad_norm": 2.103982448577881, "learning_rate": 6.907732482228349e-06, "loss": 0.0181, "step": 18149 }, { "epoch": 1.9676929748482221, "grad_norm": 5.3951005935668945, "learning_rate": 6.907007108660961e-06, "loss": 0.0509, "step": 18150 }, { "epoch": 1.9678013876843017, "grad_norm": 1.0111346244812012, "learning_rate": 6.906281735093574e-06, "loss": 0.0189, "step": 18151 }, { "epoch": 1.9679098005203817, "grad_norm": 3.062429428100586, "learning_rate": 6.905556361526187e-06, "loss": 0.0437, "step": 18152 }, { "epoch": 1.9680182133564614, "grad_norm": 0.037662114948034286, "learning_rate": 6.904830987958799e-06, "loss": 0.0004, "step": 18153 }, { "epoch": 1.9681266261925412, "grad_norm": 4.465389251708984, "learning_rate": 6.904105614391412e-06, "loss": 0.0438, "step": 18154 }, { "epoch": 1.968235039028621, "grad_norm": 3.890794515609741, "learning_rate": 6.903380240824024e-06, "loss": 0.0889, "step": 18155 }, { "epoch": 1.9683434518647007, "grad_norm": 3.316368579864502, "learning_rate": 6.902654867256637e-06, "loss": 0.0277, "step": 18156 }, { "epoch": 1.9684518647007807, "grad_norm": 3.760145425796509, "learning_rate": 6.90192949368925e-06, "loss": 0.0409, "step": 18157 }, { "epoch": 1.9685602775368602, "grad_norm": 0.5422763228416443, "learning_rate": 6.901204120121864e-06, "loss": 0.0045, "step": 18158 }, { "epoch": 1.9686686903729402, "grad_norm": 0.4772178828716278, "learning_rate": 6.900478746554476e-06, "loss": 0.0291, "step": 18159 }, { "epoch": 1.96877710320902, "grad_norm": 8.294501304626465, "learning_rate": 6.899753372987089e-06, "loss": 0.0206, "step": 18160 }, { "epoch": 1.9688855160450998, "grad_norm": 0.4927404522895813, "learning_rate": 6.899027999419701e-06, "loss": 0.0026, "step": 18161 }, { "epoch": 1.9689939288811795, "grad_norm": 0.14716966450214386, "learning_rate": 6.898302625852314e-06, "loss": 0.0013, "step": 18162 }, { "epoch": 1.9691023417172593, "grad_norm": 3.5346157550811768, "learning_rate": 6.897577252284927e-06, "loss": 0.0302, "step": 18163 }, { "epoch": 1.9692107545533393, "grad_norm": 0.030909258872270584, "learning_rate": 6.8968518787175395e-06, "loss": 0.0002, "step": 18164 }, { "epoch": 1.9693191673894188, "grad_norm": 0.9671289324760437, "learning_rate": 6.8961265051501525e-06, "loss": 0.0357, "step": 18165 }, { "epoch": 1.9694275802254988, "grad_norm": 1.8476877212524414, "learning_rate": 6.8954011315827655e-06, "loss": 0.0222, "step": 18166 }, { "epoch": 1.9695359930615783, "grad_norm": 1.3804943561553955, "learning_rate": 6.894675758015379e-06, "loss": 0.0214, "step": 18167 }, { "epoch": 1.9696444058976583, "grad_norm": 0.2635296583175659, "learning_rate": 6.8939503844479915e-06, "loss": 0.0017, "step": 18168 }, { "epoch": 1.969752818733738, "grad_norm": 0.3211367428302765, "learning_rate": 6.8932250108806045e-06, "loss": 0.0051, "step": 18169 }, { "epoch": 1.9698612315698178, "grad_norm": 3.9838109016418457, "learning_rate": 6.892499637313217e-06, "loss": 0.0476, "step": 18170 }, { "epoch": 1.9699696444058976, "grad_norm": 2.0853817462921143, "learning_rate": 6.89177426374583e-06, "loss": 0.0154, "step": 18171 }, { "epoch": 1.9700780572419774, "grad_norm": 0.7273909449577332, "learning_rate": 6.891048890178443e-06, "loss": 0.0181, "step": 18172 }, { "epoch": 1.9701864700780574, "grad_norm": 0.5277756452560425, "learning_rate": 6.890323516611055e-06, "loss": 0.0056, "step": 18173 }, { "epoch": 1.970294882914137, "grad_norm": 4.314625263214111, "learning_rate": 6.889598143043668e-06, "loss": 0.0274, "step": 18174 }, { "epoch": 1.9704032957502169, "grad_norm": 14.82829761505127, "learning_rate": 6.8888727694762816e-06, "loss": 0.0322, "step": 18175 }, { "epoch": 1.9705117085862967, "grad_norm": 0.47211650013923645, "learning_rate": 6.888147395908894e-06, "loss": 0.0023, "step": 18176 }, { "epoch": 1.9706201214223764, "grad_norm": 3.6558361053466797, "learning_rate": 6.887422022341507e-06, "loss": 0.0441, "step": 18177 }, { "epoch": 1.9707285342584562, "grad_norm": 1.401824712753296, "learning_rate": 6.88669664877412e-06, "loss": 0.0041, "step": 18178 }, { "epoch": 1.970836947094536, "grad_norm": 0.4359285533428192, "learning_rate": 6.885971275206732e-06, "loss": 0.0015, "step": 18179 }, { "epoch": 1.970945359930616, "grad_norm": 2.573650360107422, "learning_rate": 6.885245901639345e-06, "loss": 0.0388, "step": 18180 }, { "epoch": 1.9710537727666955, "grad_norm": 0.4596036672592163, "learning_rate": 6.884520528071957e-06, "loss": 0.0012, "step": 18181 }, { "epoch": 1.9711621856027755, "grad_norm": 6.343008518218994, "learning_rate": 6.88379515450457e-06, "loss": 0.0399, "step": 18182 }, { "epoch": 1.9712705984388552, "grad_norm": 1.3988304138183594, "learning_rate": 6.883069780937183e-06, "loss": 0.0184, "step": 18183 }, { "epoch": 1.971379011274935, "grad_norm": 2.0636422634124756, "learning_rate": 6.882344407369797e-06, "loss": 0.0284, "step": 18184 }, { "epoch": 1.9714874241110147, "grad_norm": 2.9587841033935547, "learning_rate": 6.881619033802409e-06, "loss": 0.022, "step": 18185 }, { "epoch": 1.9715958369470945, "grad_norm": 0.17257075011730194, "learning_rate": 6.880893660235022e-06, "loss": 0.0021, "step": 18186 }, { "epoch": 1.9717042497831745, "grad_norm": 0.6108195185661316, "learning_rate": 6.880168286667634e-06, "loss": 0.0273, "step": 18187 }, { "epoch": 1.971812662619254, "grad_norm": 0.591010570526123, "learning_rate": 6.879442913100247e-06, "loss": 0.0135, "step": 18188 }, { "epoch": 1.971921075455334, "grad_norm": 2.605588912963867, "learning_rate": 6.87871753953286e-06, "loss": 0.0381, "step": 18189 }, { "epoch": 1.9720294882914136, "grad_norm": 1.4519555568695068, "learning_rate": 6.877992165965472e-06, "loss": 0.0316, "step": 18190 }, { "epoch": 1.9721379011274935, "grad_norm": 0.6753460168838501, "learning_rate": 6.877266792398085e-06, "loss": 0.0015, "step": 18191 }, { "epoch": 1.9722463139635733, "grad_norm": 0.5030945539474487, "learning_rate": 6.876541418830699e-06, "loss": 0.0253, "step": 18192 }, { "epoch": 1.972354726799653, "grad_norm": 5.241085529327393, "learning_rate": 6.875816045263312e-06, "loss": 0.0111, "step": 18193 }, { "epoch": 1.9724631396357328, "grad_norm": 0.2918631434440613, "learning_rate": 6.875090671695924e-06, "loss": 0.01, "step": 18194 }, { "epoch": 1.9725715524718126, "grad_norm": 2.0925090312957764, "learning_rate": 6.874365298128537e-06, "loss": 0.0065, "step": 18195 }, { "epoch": 1.9726799653078926, "grad_norm": 1.7424265146255493, "learning_rate": 6.873639924561149e-06, "loss": 0.0219, "step": 18196 }, { "epoch": 1.9727883781439721, "grad_norm": 2.0914249420166016, "learning_rate": 6.872914550993762e-06, "loss": 0.0063, "step": 18197 }, { "epoch": 1.972896790980052, "grad_norm": 7.179256916046143, "learning_rate": 6.872189177426375e-06, "loss": 0.0258, "step": 18198 }, { "epoch": 1.9730052038161319, "grad_norm": 0.48458728194236755, "learning_rate": 6.871463803858987e-06, "loss": 0.0016, "step": 18199 }, { "epoch": 1.9731136166522116, "grad_norm": 0.5879507064819336, "learning_rate": 6.8707384302916e-06, "loss": 0.0047, "step": 18200 }, { "epoch": 1.9732220294882914, "grad_norm": 3.3218202590942383, "learning_rate": 6.870013056724214e-06, "loss": 0.0307, "step": 18201 }, { "epoch": 1.9733304423243712, "grad_norm": 0.5328900218009949, "learning_rate": 6.869287683156826e-06, "loss": 0.0138, "step": 18202 }, { "epoch": 1.9734388551604511, "grad_norm": 0.4425312876701355, "learning_rate": 6.868562309589439e-06, "loss": 0.0168, "step": 18203 }, { "epoch": 1.9735472679965307, "grad_norm": 1.4690091609954834, "learning_rate": 6.867836936022052e-06, "loss": 0.0071, "step": 18204 }, { "epoch": 1.9736556808326107, "grad_norm": 0.958533763885498, "learning_rate": 6.8671115624546645e-06, "loss": 0.0097, "step": 18205 }, { "epoch": 1.9737640936686902, "grad_norm": 1.576126217842102, "learning_rate": 6.8663861888872775e-06, "loss": 0.0067, "step": 18206 }, { "epoch": 1.9738725065047702, "grad_norm": 12.576306343078613, "learning_rate": 6.86566081531989e-06, "loss": 0.0535, "step": 18207 }, { "epoch": 1.97398091934085, "grad_norm": 0.31277620792388916, "learning_rate": 6.864935441752503e-06, "loss": 0.0067, "step": 18208 }, { "epoch": 1.9740893321769297, "grad_norm": 0.17903932929039001, "learning_rate": 6.864210068185116e-06, "loss": 0.0014, "step": 18209 }, { "epoch": 1.9741977450130095, "grad_norm": 4.500082015991211, "learning_rate": 6.8634846946177295e-06, "loss": 0.028, "step": 18210 }, { "epoch": 1.9743061578490892, "grad_norm": 0.4560146927833557, "learning_rate": 6.862759321050342e-06, "loss": 0.0027, "step": 18211 }, { "epoch": 1.9744145706851692, "grad_norm": 3.79620099067688, "learning_rate": 6.862033947482955e-06, "loss": 0.0591, "step": 18212 }, { "epoch": 1.9745229835212488, "grad_norm": 2.3936927318573, "learning_rate": 6.861308573915567e-06, "loss": 0.0849, "step": 18213 }, { "epoch": 1.9746313963573288, "grad_norm": 0.5260933637619019, "learning_rate": 6.86058320034818e-06, "loss": 0.0161, "step": 18214 }, { "epoch": 1.9747398091934085, "grad_norm": 3.777622699737549, "learning_rate": 6.859857826780793e-06, "loss": 0.0304, "step": 18215 }, { "epoch": 1.9748482220294883, "grad_norm": 11.297466278076172, "learning_rate": 6.859132453213405e-06, "loss": 0.0647, "step": 18216 }, { "epoch": 1.974956634865568, "grad_norm": 9.055525779724121, "learning_rate": 6.858407079646018e-06, "loss": 0.0428, "step": 18217 }, { "epoch": 1.9750650477016478, "grad_norm": 28.27457618713379, "learning_rate": 6.857681706078632e-06, "loss": 0.0224, "step": 18218 }, { "epoch": 1.9751734605377278, "grad_norm": 1.9908137321472168, "learning_rate": 6.856956332511245e-06, "loss": 0.0317, "step": 18219 }, { "epoch": 1.9752818733738073, "grad_norm": 0.28338706493377686, "learning_rate": 6.856230958943857e-06, "loss": 0.0145, "step": 18220 }, { "epoch": 1.9753902862098873, "grad_norm": 1.4776480197906494, "learning_rate": 6.85550558537647e-06, "loss": 0.0116, "step": 18221 }, { "epoch": 1.975498699045967, "grad_norm": 0.2225300669670105, "learning_rate": 6.854780211809082e-06, "loss": 0.0036, "step": 18222 }, { "epoch": 1.9756071118820469, "grad_norm": 2.55692195892334, "learning_rate": 6.854054838241695e-06, "loss": 0.0225, "step": 18223 }, { "epoch": 1.9757155247181266, "grad_norm": 3.3331949710845947, "learning_rate": 6.853329464674308e-06, "loss": 0.0417, "step": 18224 }, { "epoch": 1.9758239375542064, "grad_norm": 2.8794798851013184, "learning_rate": 6.85260409110692e-06, "loss": 0.0723, "step": 18225 }, { "epoch": 1.9759323503902864, "grad_norm": 2.0164847373962402, "learning_rate": 6.851878717539533e-06, "loss": 0.0121, "step": 18226 }, { "epoch": 1.976040763226366, "grad_norm": 0.3763352930545807, "learning_rate": 6.851153343972147e-06, "loss": 0.0103, "step": 18227 }, { "epoch": 1.976149176062446, "grad_norm": 1.8154240846633911, "learning_rate": 6.850427970404759e-06, "loss": 0.0118, "step": 18228 }, { "epoch": 1.9762575888985254, "grad_norm": 0.41967180371284485, "learning_rate": 6.849702596837372e-06, "loss": 0.0016, "step": 18229 }, { "epoch": 1.9763660017346054, "grad_norm": 0.31283095479011536, "learning_rate": 6.848977223269985e-06, "loss": 0.0018, "step": 18230 }, { "epoch": 1.9764744145706852, "grad_norm": 1.564738392829895, "learning_rate": 6.848251849702597e-06, "loss": 0.1023, "step": 18231 }, { "epoch": 1.976582827406765, "grad_norm": 3.7270078659057617, "learning_rate": 6.84752647613521e-06, "loss": 0.0482, "step": 18232 }, { "epoch": 1.9766912402428447, "grad_norm": 1.919651985168457, "learning_rate": 6.846801102567822e-06, "loss": 0.037, "step": 18233 }, { "epoch": 1.9767996530789245, "grad_norm": 3.1957452297210693, "learning_rate": 6.846075729000435e-06, "loss": 0.0724, "step": 18234 }, { "epoch": 1.9769080659150045, "grad_norm": 0.5043485164642334, "learning_rate": 6.845350355433048e-06, "loss": 0.0023, "step": 18235 }, { "epoch": 1.977016478751084, "grad_norm": 1.5680558681488037, "learning_rate": 6.844624981865662e-06, "loss": 0.0405, "step": 18236 }, { "epoch": 1.977124891587164, "grad_norm": 2.860352039337158, "learning_rate": 6.843899608298274e-06, "loss": 0.0612, "step": 18237 }, { "epoch": 1.9772333044232437, "grad_norm": 6.680774211883545, "learning_rate": 6.843174234730887e-06, "loss": 0.028, "step": 18238 }, { "epoch": 1.9773417172593235, "grad_norm": 0.08945457637310028, "learning_rate": 6.842448861163499e-06, "loss": 0.0009, "step": 18239 }, { "epoch": 1.9774501300954033, "grad_norm": 1.7406671047210693, "learning_rate": 6.841723487596112e-06, "loss": 0.0279, "step": 18240 }, { "epoch": 1.977558542931483, "grad_norm": 0.3005421757698059, "learning_rate": 6.840998114028725e-06, "loss": 0.0065, "step": 18241 }, { "epoch": 1.977666955767563, "grad_norm": 4.00197172164917, "learning_rate": 6.8402727404613375e-06, "loss": 0.0253, "step": 18242 }, { "epoch": 1.9777753686036426, "grad_norm": 0.15052345395088196, "learning_rate": 6.8395473668939505e-06, "loss": 0.0016, "step": 18243 }, { "epoch": 1.9778837814397225, "grad_norm": 0.8054745197296143, "learning_rate": 6.838821993326564e-06, "loss": 0.0116, "step": 18244 }, { "epoch": 1.9779921942758023, "grad_norm": 0.3310889005661011, "learning_rate": 6.838096619759177e-06, "loss": 0.0011, "step": 18245 }, { "epoch": 1.978100607111882, "grad_norm": 0.3393137753009796, "learning_rate": 6.8373712461917895e-06, "loss": 0.0062, "step": 18246 }, { "epoch": 1.9782090199479618, "grad_norm": 0.5844441056251526, "learning_rate": 6.8366458726244025e-06, "loss": 0.0222, "step": 18247 }, { "epoch": 1.9783174327840416, "grad_norm": 3.6537158489227295, "learning_rate": 6.835920499057015e-06, "loss": 0.0465, "step": 18248 }, { "epoch": 1.9784258456201216, "grad_norm": 5.929421901702881, "learning_rate": 6.835195125489628e-06, "loss": 0.0178, "step": 18249 }, { "epoch": 1.9785342584562011, "grad_norm": 2.083376884460449, "learning_rate": 6.834469751922241e-06, "loss": 0.0423, "step": 18250 }, { "epoch": 1.978642671292281, "grad_norm": 1.2980937957763672, "learning_rate": 6.833744378354853e-06, "loss": 0.0254, "step": 18251 }, { "epoch": 1.9787510841283606, "grad_norm": 0.5343608856201172, "learning_rate": 6.833019004787466e-06, "loss": 0.02, "step": 18252 }, { "epoch": 1.9788594969644406, "grad_norm": 1.3483290672302246, "learning_rate": 6.8322936312200796e-06, "loss": 0.0068, "step": 18253 }, { "epoch": 1.9789679098005204, "grad_norm": 4.378442764282227, "learning_rate": 6.831568257652692e-06, "loss": 0.062, "step": 18254 }, { "epoch": 1.9790763226366002, "grad_norm": 1.2530882358551025, "learning_rate": 6.830842884085305e-06, "loss": 0.0099, "step": 18255 }, { "epoch": 1.97918473547268, "grad_norm": 2.3668880462646484, "learning_rate": 6.830117510517918e-06, "loss": 0.0222, "step": 18256 }, { "epoch": 1.9792931483087597, "grad_norm": 4.45868444442749, "learning_rate": 6.82939213695053e-06, "loss": 0.0164, "step": 18257 }, { "epoch": 1.9794015611448397, "grad_norm": 1.821117639541626, "learning_rate": 6.828666763383143e-06, "loss": 0.0184, "step": 18258 }, { "epoch": 1.9795099739809192, "grad_norm": 0.29929378628730774, "learning_rate": 6.827941389815755e-06, "loss": 0.0092, "step": 18259 }, { "epoch": 1.9796183868169992, "grad_norm": 1.6043545007705688, "learning_rate": 6.827216016248368e-06, "loss": 0.0226, "step": 18260 }, { "epoch": 1.979726799653079, "grad_norm": 2.8911759853363037, "learning_rate": 6.826490642680981e-06, "loss": 0.0354, "step": 18261 }, { "epoch": 1.9798352124891587, "grad_norm": 0.6768787503242493, "learning_rate": 6.825765269113595e-06, "loss": 0.0097, "step": 18262 }, { "epoch": 1.9799436253252385, "grad_norm": 0.2508951723575592, "learning_rate": 6.825039895546207e-06, "loss": 0.0019, "step": 18263 }, { "epoch": 1.9800520381613183, "grad_norm": 0.18814241886138916, "learning_rate": 6.82431452197882e-06, "loss": 0.0027, "step": 18264 }, { "epoch": 1.9801604509973982, "grad_norm": 0.03328770399093628, "learning_rate": 6.823589148411433e-06, "loss": 0.0005, "step": 18265 }, { "epoch": 1.9802688638334778, "grad_norm": 0.21696503460407257, "learning_rate": 6.822863774844045e-06, "loss": 0.0016, "step": 18266 }, { "epoch": 1.9803772766695578, "grad_norm": 1.5152696371078491, "learning_rate": 6.822138401276658e-06, "loss": 0.0232, "step": 18267 }, { "epoch": 1.9804856895056373, "grad_norm": 2.3020219802856445, "learning_rate": 6.82141302770927e-06, "loss": 0.031, "step": 18268 }, { "epoch": 1.9805941023417173, "grad_norm": 1.514855146408081, "learning_rate": 6.820687654141883e-06, "loss": 0.0466, "step": 18269 }, { "epoch": 1.980702515177797, "grad_norm": 0.2455117404460907, "learning_rate": 6.819962280574497e-06, "loss": 0.0038, "step": 18270 }, { "epoch": 1.9808109280138768, "grad_norm": 1.9192421436309814, "learning_rate": 6.81923690700711e-06, "loss": 0.0506, "step": 18271 }, { "epoch": 1.9809193408499568, "grad_norm": 0.8340036273002625, "learning_rate": 6.818511533439722e-06, "loss": 0.0019, "step": 18272 }, { "epoch": 1.9810277536860363, "grad_norm": 0.2559850811958313, "learning_rate": 6.817786159872335e-06, "loss": 0.0021, "step": 18273 }, { "epoch": 1.9811361665221163, "grad_norm": 6.887842178344727, "learning_rate": 6.817060786304947e-06, "loss": 0.0624, "step": 18274 }, { "epoch": 1.9812445793581959, "grad_norm": 6.168846607208252, "learning_rate": 6.81633541273756e-06, "loss": 0.0381, "step": 18275 }, { "epoch": 1.9813529921942759, "grad_norm": 2.9561307430267334, "learning_rate": 6.815610039170173e-06, "loss": 0.064, "step": 18276 }, { "epoch": 1.9814614050303556, "grad_norm": 0.632433295249939, "learning_rate": 6.814884665602785e-06, "loss": 0.0054, "step": 18277 }, { "epoch": 1.9815698178664354, "grad_norm": 0.9991042017936707, "learning_rate": 6.814159292035398e-06, "loss": 0.0156, "step": 18278 }, { "epoch": 1.9816782307025151, "grad_norm": 0.09949919581413269, "learning_rate": 6.813433918468012e-06, "loss": 0.0008, "step": 18279 }, { "epoch": 1.981786643538595, "grad_norm": 2.7968602180480957, "learning_rate": 6.812708544900624e-06, "loss": 0.0436, "step": 18280 }, { "epoch": 1.981895056374675, "grad_norm": 0.12838158011436462, "learning_rate": 6.811983171333237e-06, "loss": 0.0009, "step": 18281 }, { "epoch": 1.9820034692107544, "grad_norm": 0.686017632484436, "learning_rate": 6.81125779776585e-06, "loss": 0.0048, "step": 18282 }, { "epoch": 1.9821118820468344, "grad_norm": 1.696161150932312, "learning_rate": 6.8105324241984625e-06, "loss": 0.0182, "step": 18283 }, { "epoch": 1.9822202948829142, "grad_norm": 1.2948836088180542, "learning_rate": 6.8098070506310755e-06, "loss": 0.0038, "step": 18284 }, { "epoch": 1.982328707718994, "grad_norm": 1.926350712776184, "learning_rate": 6.809081677063688e-06, "loss": 0.031, "step": 18285 }, { "epoch": 1.9824371205550737, "grad_norm": 4.455101013183594, "learning_rate": 6.808356303496301e-06, "loss": 0.0773, "step": 18286 }, { "epoch": 1.9825455333911535, "grad_norm": 4.94576358795166, "learning_rate": 6.8076309299289145e-06, "loss": 0.0227, "step": 18287 }, { "epoch": 1.9826539462272335, "grad_norm": 2.9443912506103516, "learning_rate": 6.8069055563615275e-06, "loss": 0.0211, "step": 18288 }, { "epoch": 1.982762359063313, "grad_norm": 4.998694896697998, "learning_rate": 6.80618018279414e-06, "loss": 0.0458, "step": 18289 }, { "epoch": 1.982870771899393, "grad_norm": 1.3059337139129639, "learning_rate": 6.805454809226753e-06, "loss": 0.0274, "step": 18290 }, { "epoch": 1.9829791847354725, "grad_norm": 0.5313696265220642, "learning_rate": 6.804729435659366e-06, "loss": 0.008, "step": 18291 }, { "epoch": 1.9830875975715525, "grad_norm": 0.3799651861190796, "learning_rate": 6.804004062091978e-06, "loss": 0.0073, "step": 18292 }, { "epoch": 1.9831960104076323, "grad_norm": 5.350579261779785, "learning_rate": 6.803278688524591e-06, "loss": 0.0543, "step": 18293 }, { "epoch": 1.983304423243712, "grad_norm": 0.6992322206497192, "learning_rate": 6.802553314957203e-06, "loss": 0.0051, "step": 18294 }, { "epoch": 1.9834128360797918, "grad_norm": 1.2226132154464722, "learning_rate": 6.801827941389816e-06, "loss": 0.0137, "step": 18295 }, { "epoch": 1.9835212489158716, "grad_norm": 0.7115071415901184, "learning_rate": 6.80110256782243e-06, "loss": 0.0216, "step": 18296 }, { "epoch": 1.9836296617519515, "grad_norm": 6.2139387130737305, "learning_rate": 6.800377194255043e-06, "loss": 0.0437, "step": 18297 }, { "epoch": 1.983738074588031, "grad_norm": 1.1029651165008545, "learning_rate": 6.799651820687655e-06, "loss": 0.0105, "step": 18298 }, { "epoch": 1.983846487424111, "grad_norm": 1.5809797048568726, "learning_rate": 6.798926447120268e-06, "loss": 0.0323, "step": 18299 }, { "epoch": 1.9839549002601908, "grad_norm": 0.3572099506855011, "learning_rate": 6.79820107355288e-06, "loss": 0.0017, "step": 18300 }, { "epoch": 1.9840633130962706, "grad_norm": 0.1119864359498024, "learning_rate": 6.797475699985493e-06, "loss": 0.001, "step": 18301 }, { "epoch": 1.9841717259323504, "grad_norm": 1.479862093925476, "learning_rate": 6.796750326418106e-06, "loss": 0.0347, "step": 18302 }, { "epoch": 1.9842801387684301, "grad_norm": 0.6380778551101685, "learning_rate": 6.796024952850718e-06, "loss": 0.0161, "step": 18303 }, { "epoch": 1.98438855160451, "grad_norm": 3.8638546466827393, "learning_rate": 6.795299579283331e-06, "loss": 0.0135, "step": 18304 }, { "epoch": 1.9844969644405897, "grad_norm": 5.235713958740234, "learning_rate": 6.794574205715945e-06, "loss": 0.0341, "step": 18305 }, { "epoch": 1.9846053772766696, "grad_norm": 2.534285545349121, "learning_rate": 6.793848832148557e-06, "loss": 0.0493, "step": 18306 }, { "epoch": 1.9847137901127494, "grad_norm": 2.886582851409912, "learning_rate": 6.79312345858117e-06, "loss": 0.0063, "step": 18307 }, { "epoch": 1.9848222029488292, "grad_norm": 1.463727593421936, "learning_rate": 6.792398085013783e-06, "loss": 0.038, "step": 18308 }, { "epoch": 1.984930615784909, "grad_norm": 0.8346675038337708, "learning_rate": 6.791672711446395e-06, "loss": 0.0149, "step": 18309 }, { "epoch": 1.9850390286209887, "grad_norm": 0.45533764362335205, "learning_rate": 6.790947337879008e-06, "loss": 0.0125, "step": 18310 }, { "epoch": 1.9851474414570687, "grad_norm": 0.5425202250480652, "learning_rate": 6.79022196431162e-06, "loss": 0.0201, "step": 18311 }, { "epoch": 1.9852558542931482, "grad_norm": 0.34004291892051697, "learning_rate": 6.789496590744233e-06, "loss": 0.0068, "step": 18312 }, { "epoch": 1.9853642671292282, "grad_norm": 5.69736385345459, "learning_rate": 6.788771217176847e-06, "loss": 0.0147, "step": 18313 }, { "epoch": 1.9854726799653077, "grad_norm": 1.5815750360488892, "learning_rate": 6.78804584360946e-06, "loss": 0.0146, "step": 18314 }, { "epoch": 1.9855810928013877, "grad_norm": 0.053373899310827255, "learning_rate": 6.787320470042072e-06, "loss": 0.0004, "step": 18315 }, { "epoch": 1.9856895056374675, "grad_norm": 1.7597203254699707, "learning_rate": 6.786595096474685e-06, "loss": 0.0114, "step": 18316 }, { "epoch": 1.9857979184735473, "grad_norm": 1.4668442010879517, "learning_rate": 6.785869722907298e-06, "loss": 0.0154, "step": 18317 }, { "epoch": 1.985906331309627, "grad_norm": 4.058829307556152, "learning_rate": 6.78514434933991e-06, "loss": 0.0562, "step": 18318 }, { "epoch": 1.9860147441457068, "grad_norm": 1.0382859706878662, "learning_rate": 6.784418975772523e-06, "loss": 0.0166, "step": 18319 }, { "epoch": 1.9861231569817868, "grad_norm": 4.995181083679199, "learning_rate": 6.7836936022051355e-06, "loss": 0.0462, "step": 18320 }, { "epoch": 1.9862315698178663, "grad_norm": 0.39191240072250366, "learning_rate": 6.7829682286377485e-06, "loss": 0.0027, "step": 18321 }, { "epoch": 1.9863399826539463, "grad_norm": 0.2654420733451843, "learning_rate": 6.782242855070362e-06, "loss": 0.0028, "step": 18322 }, { "epoch": 1.986448395490026, "grad_norm": 5.739865303039551, "learning_rate": 6.781517481502975e-06, "loss": 0.0209, "step": 18323 }, { "epoch": 1.9865568083261058, "grad_norm": 0.4258524179458618, "learning_rate": 6.7807921079355875e-06, "loss": 0.0129, "step": 18324 }, { "epoch": 1.9866652211621856, "grad_norm": 0.9852182269096375, "learning_rate": 6.7800667343682005e-06, "loss": 0.0046, "step": 18325 }, { "epoch": 1.9867736339982653, "grad_norm": 0.47801294922828674, "learning_rate": 6.779341360800813e-06, "loss": 0.0087, "step": 18326 }, { "epoch": 1.9868820468343453, "grad_norm": 0.06274553388357162, "learning_rate": 6.778615987233426e-06, "loss": 0.0008, "step": 18327 }, { "epoch": 1.9869904596704249, "grad_norm": 2.8332386016845703, "learning_rate": 6.777890613666039e-06, "loss": 0.0122, "step": 18328 }, { "epoch": 1.9870988725065049, "grad_norm": 3.1851632595062256, "learning_rate": 6.777165240098651e-06, "loss": 0.045, "step": 18329 }, { "epoch": 1.9872072853425846, "grad_norm": 2.235309600830078, "learning_rate": 6.776439866531264e-06, "loss": 0.0539, "step": 18330 }, { "epoch": 1.9873156981786644, "grad_norm": 2.4376296997070312, "learning_rate": 6.775714492963878e-06, "loss": 0.0407, "step": 18331 }, { "epoch": 1.9874241110147441, "grad_norm": 0.30828657746315, "learning_rate": 6.77498911939649e-06, "loss": 0.0016, "step": 18332 }, { "epoch": 1.987532523850824, "grad_norm": 1.9709466695785522, "learning_rate": 6.774263745829103e-06, "loss": 0.015, "step": 18333 }, { "epoch": 1.987640936686904, "grad_norm": 0.6378077864646912, "learning_rate": 6.773538372261716e-06, "loss": 0.0079, "step": 18334 }, { "epoch": 1.9877493495229834, "grad_norm": 7.445310592651367, "learning_rate": 6.772812998694328e-06, "loss": 0.0373, "step": 18335 }, { "epoch": 1.9878577623590634, "grad_norm": 2.5627527236938477, "learning_rate": 6.772087625126941e-06, "loss": 0.0373, "step": 18336 }, { "epoch": 1.987966175195143, "grad_norm": 1.2373480796813965, "learning_rate": 6.771362251559553e-06, "loss": 0.0307, "step": 18337 }, { "epoch": 1.988074588031223, "grad_norm": 1.947519302368164, "learning_rate": 6.770636877992166e-06, "loss": 0.0589, "step": 18338 }, { "epoch": 1.9881830008673027, "grad_norm": 0.09992365539073944, "learning_rate": 6.76991150442478e-06, "loss": 0.0008, "step": 18339 }, { "epoch": 1.9882914137033825, "grad_norm": 0.2928069233894348, "learning_rate": 6.769186130857393e-06, "loss": 0.0024, "step": 18340 }, { "epoch": 1.9883998265394622, "grad_norm": 3.26544451713562, "learning_rate": 6.768460757290005e-06, "loss": 0.0282, "step": 18341 }, { "epoch": 1.988508239375542, "grad_norm": 3.425689458847046, "learning_rate": 6.767735383722618e-06, "loss": 0.0272, "step": 18342 }, { "epoch": 1.988616652211622, "grad_norm": 0.8618786931037903, "learning_rate": 6.767010010155231e-06, "loss": 0.0069, "step": 18343 }, { "epoch": 1.9887250650477015, "grad_norm": 1.5879011154174805, "learning_rate": 6.766284636587843e-06, "loss": 0.0233, "step": 18344 }, { "epoch": 1.9888334778837815, "grad_norm": 3.5438270568847656, "learning_rate": 6.765559263020456e-06, "loss": 0.0217, "step": 18345 }, { "epoch": 1.9889418907198613, "grad_norm": 3.221853733062744, "learning_rate": 6.764833889453068e-06, "loss": 0.0237, "step": 18346 }, { "epoch": 1.989050303555941, "grad_norm": 2.6026744842529297, "learning_rate": 6.764108515885681e-06, "loss": 0.034, "step": 18347 }, { "epoch": 1.9891587163920208, "grad_norm": 0.5831708908081055, "learning_rate": 6.763383142318295e-06, "loss": 0.0109, "step": 18348 }, { "epoch": 1.9892671292281006, "grad_norm": 6.7384796142578125, "learning_rate": 6.762657768750908e-06, "loss": 0.0461, "step": 18349 }, { "epoch": 1.9893755420641805, "grad_norm": 5.917725563049316, "learning_rate": 6.76193239518352e-06, "loss": 0.0276, "step": 18350 }, { "epoch": 1.98948395490026, "grad_norm": 4.447926044464111, "learning_rate": 6.761207021616133e-06, "loss": 0.04, "step": 18351 }, { "epoch": 1.98959236773634, "grad_norm": 2.2031049728393555, "learning_rate": 6.760481648048745e-06, "loss": 0.0385, "step": 18352 }, { "epoch": 1.9897007805724196, "grad_norm": 1.4858123064041138, "learning_rate": 6.759756274481358e-06, "loss": 0.0075, "step": 18353 }, { "epoch": 1.9898091934084996, "grad_norm": 5.347671031951904, "learning_rate": 6.759030900913971e-06, "loss": 0.041, "step": 18354 }, { "epoch": 1.9899176062445794, "grad_norm": 7.798589706420898, "learning_rate": 6.758305527346583e-06, "loss": 0.0264, "step": 18355 }, { "epoch": 1.9900260190806591, "grad_norm": 1.2623564004898071, "learning_rate": 6.757580153779196e-06, "loss": 0.0235, "step": 18356 }, { "epoch": 1.9901344319167389, "grad_norm": 3.579619884490967, "learning_rate": 6.75685478021181e-06, "loss": 0.0465, "step": 18357 }, { "epoch": 1.9902428447528187, "grad_norm": 0.28381073474884033, "learning_rate": 6.756129406644422e-06, "loss": 0.0063, "step": 18358 }, { "epoch": 1.9903512575888986, "grad_norm": 0.11061770468950272, "learning_rate": 6.755404033077035e-06, "loss": 0.0017, "step": 18359 }, { "epoch": 1.9904596704249782, "grad_norm": 0.16948024928569794, "learning_rate": 6.754678659509648e-06, "loss": 0.0018, "step": 18360 }, { "epoch": 1.9905680832610582, "grad_norm": 4.5268449783325195, "learning_rate": 6.7539532859422605e-06, "loss": 0.0286, "step": 18361 }, { "epoch": 1.990676496097138, "grad_norm": 2.5838911533355713, "learning_rate": 6.7532279123748735e-06, "loss": 0.02, "step": 18362 }, { "epoch": 1.9907849089332177, "grad_norm": 0.9062396287918091, "learning_rate": 6.752502538807486e-06, "loss": 0.0174, "step": 18363 }, { "epoch": 1.9908933217692975, "grad_norm": 0.009358149021863937, "learning_rate": 6.751777165240099e-06, "loss": 0.0001, "step": 18364 }, { "epoch": 1.9910017346053772, "grad_norm": 3.5826447010040283, "learning_rate": 6.7510517916727125e-06, "loss": 0.0464, "step": 18365 }, { "epoch": 1.9911101474414572, "grad_norm": 2.1766672134399414, "learning_rate": 6.7503264181053255e-06, "loss": 0.0251, "step": 18366 }, { "epoch": 1.9912185602775367, "grad_norm": 1.2075529098510742, "learning_rate": 6.749601044537938e-06, "loss": 0.0245, "step": 18367 }, { "epoch": 1.9913269731136167, "grad_norm": 3.281276226043701, "learning_rate": 6.748875670970551e-06, "loss": 0.0139, "step": 18368 }, { "epoch": 1.9914353859496965, "grad_norm": 2.395712375640869, "learning_rate": 6.748150297403164e-06, "loss": 0.0133, "step": 18369 }, { "epoch": 1.9915437987857763, "grad_norm": 2.3933253288269043, "learning_rate": 6.747424923835776e-06, "loss": 0.0145, "step": 18370 }, { "epoch": 1.991652211621856, "grad_norm": 4.26458215713501, "learning_rate": 6.746699550268389e-06, "loss": 0.0402, "step": 18371 }, { "epoch": 1.9917606244579358, "grad_norm": 0.2075081467628479, "learning_rate": 6.745974176701001e-06, "loss": 0.0009, "step": 18372 }, { "epoch": 1.9918690372940158, "grad_norm": 9.196683883666992, "learning_rate": 6.745248803133614e-06, "loss": 0.0378, "step": 18373 }, { "epoch": 1.9919774501300953, "grad_norm": 6.0869340896606445, "learning_rate": 6.744523429566228e-06, "loss": 0.0465, "step": 18374 }, { "epoch": 1.9920858629661753, "grad_norm": 2.9236435890197754, "learning_rate": 6.743798055998841e-06, "loss": 0.0102, "step": 18375 }, { "epoch": 1.9921942758022548, "grad_norm": 1.3113417625427246, "learning_rate": 6.743072682431453e-06, "loss": 0.0108, "step": 18376 }, { "epoch": 1.9923026886383348, "grad_norm": 0.27779123187065125, "learning_rate": 6.742347308864066e-06, "loss": 0.0062, "step": 18377 }, { "epoch": 1.9924111014744146, "grad_norm": 0.7981836199760437, "learning_rate": 6.741621935296678e-06, "loss": 0.008, "step": 18378 }, { "epoch": 1.9925195143104943, "grad_norm": 1.5868617296218872, "learning_rate": 6.740896561729291e-06, "loss": 0.0082, "step": 18379 }, { "epoch": 1.992627927146574, "grad_norm": 1.3588258028030396, "learning_rate": 6.740171188161904e-06, "loss": 0.0219, "step": 18380 }, { "epoch": 1.9927363399826539, "grad_norm": 0.3239632248878479, "learning_rate": 6.739445814594516e-06, "loss": 0.0075, "step": 18381 }, { "epoch": 1.9928447528187339, "grad_norm": 0.5977206826210022, "learning_rate": 6.73872044102713e-06, "loss": 0.005, "step": 18382 }, { "epoch": 1.9929531656548134, "grad_norm": 1.0865057706832886, "learning_rate": 6.737995067459743e-06, "loss": 0.0136, "step": 18383 }, { "epoch": 1.9930615784908934, "grad_norm": 3.3432977199554443, "learning_rate": 6.737269693892355e-06, "loss": 0.0197, "step": 18384 }, { "epoch": 1.9931699913269731, "grad_norm": 6.438643455505371, "learning_rate": 6.736544320324968e-06, "loss": 0.0467, "step": 18385 }, { "epoch": 1.993278404163053, "grad_norm": 0.42483171820640564, "learning_rate": 6.735818946757581e-06, "loss": 0.009, "step": 18386 }, { "epoch": 1.9933868169991327, "grad_norm": 0.5919297933578491, "learning_rate": 6.735093573190193e-06, "loss": 0.0328, "step": 18387 }, { "epoch": 1.9934952298352124, "grad_norm": 0.3981974720954895, "learning_rate": 6.734368199622806e-06, "loss": 0.0113, "step": 18388 }, { "epoch": 1.9936036426712924, "grad_norm": 0.5355836153030396, "learning_rate": 6.733642826055419e-06, "loss": 0.0379, "step": 18389 }, { "epoch": 1.993712055507372, "grad_norm": 5.976578235626221, "learning_rate": 6.732917452488031e-06, "loss": 0.0443, "step": 18390 }, { "epoch": 1.993820468343452, "grad_norm": 2.5998899936676025, "learning_rate": 6.732192078920645e-06, "loss": 0.0363, "step": 18391 }, { "epoch": 1.9939288811795317, "grad_norm": 2.1368408203125, "learning_rate": 6.731466705353258e-06, "loss": 0.0188, "step": 18392 }, { "epoch": 1.9940372940156115, "grad_norm": 0.5338265895843506, "learning_rate": 6.73074133178587e-06, "loss": 0.0113, "step": 18393 }, { "epoch": 1.9941457068516912, "grad_norm": 0.446918249130249, "learning_rate": 6.730015958218483e-06, "loss": 0.0068, "step": 18394 }, { "epoch": 1.994254119687771, "grad_norm": 0.48953482508659363, "learning_rate": 6.729290584651096e-06, "loss": 0.0101, "step": 18395 }, { "epoch": 1.994362532523851, "grad_norm": 2.3971054553985596, "learning_rate": 6.728565211083708e-06, "loss": 0.005, "step": 18396 }, { "epoch": 1.9944709453599305, "grad_norm": 0.768899142742157, "learning_rate": 6.727839837516321e-06, "loss": 0.0156, "step": 18397 }, { "epoch": 1.9945793581960105, "grad_norm": 1.8720852136611938, "learning_rate": 6.7271144639489335e-06, "loss": 0.0233, "step": 18398 }, { "epoch": 1.99468777103209, "grad_norm": 0.5814356803894043, "learning_rate": 6.7263890903815465e-06, "loss": 0.0228, "step": 18399 }, { "epoch": 1.99479618386817, "grad_norm": 1.7886149883270264, "learning_rate": 6.72566371681416e-06, "loss": 0.0142, "step": 18400 }, { "epoch": 1.9949045967042498, "grad_norm": 0.9695405960083008, "learning_rate": 6.724938343246773e-06, "loss": 0.0102, "step": 18401 }, { "epoch": 1.9950130095403296, "grad_norm": 7.698095798492432, "learning_rate": 6.7242129696793855e-06, "loss": 0.0678, "step": 18402 }, { "epoch": 1.9951214223764093, "grad_norm": 2.8463990688323975, "learning_rate": 6.7234875961119985e-06, "loss": 0.0298, "step": 18403 }, { "epoch": 1.995229835212489, "grad_norm": 2.714362859725952, "learning_rate": 6.722762222544611e-06, "loss": 0.0528, "step": 18404 }, { "epoch": 1.995338248048569, "grad_norm": 1.9631662368774414, "learning_rate": 6.722036848977224e-06, "loss": 0.0351, "step": 18405 }, { "epoch": 1.9954466608846486, "grad_norm": 0.306039959192276, "learning_rate": 6.721311475409837e-06, "loss": 0.0049, "step": 18406 }, { "epoch": 1.9955550737207286, "grad_norm": 1.1114825010299683, "learning_rate": 6.720586101842449e-06, "loss": 0.0021, "step": 18407 }, { "epoch": 1.9956634865568084, "grad_norm": 1.4530255794525146, "learning_rate": 6.719860728275063e-06, "loss": 0.0222, "step": 18408 }, { "epoch": 1.9957718993928881, "grad_norm": 1.0679956674575806, "learning_rate": 6.719135354707676e-06, "loss": 0.0241, "step": 18409 }, { "epoch": 1.995880312228968, "grad_norm": 2.311995267868042, "learning_rate": 6.718409981140288e-06, "loss": 0.0196, "step": 18410 }, { "epoch": 1.9959887250650477, "grad_norm": 1.3284395933151245, "learning_rate": 6.717684607572901e-06, "loss": 0.0332, "step": 18411 }, { "epoch": 1.9960971379011276, "grad_norm": 2.2652640342712402, "learning_rate": 6.716959234005514e-06, "loss": 0.045, "step": 18412 }, { "epoch": 1.9962055507372072, "grad_norm": 0.6717034578323364, "learning_rate": 6.716233860438126e-06, "loss": 0.0082, "step": 18413 }, { "epoch": 1.9963139635732872, "grad_norm": 1.5099339485168457, "learning_rate": 6.715508486870739e-06, "loss": 0.0157, "step": 18414 }, { "epoch": 1.9964223764093667, "grad_norm": 0.47351783514022827, "learning_rate": 6.714783113303352e-06, "loss": 0.0269, "step": 18415 }, { "epoch": 1.9965307892454467, "grad_norm": 0.48212486505508423, "learning_rate": 6.714057739735964e-06, "loss": 0.0212, "step": 18416 }, { "epoch": 1.9966392020815265, "grad_norm": 10.830495834350586, "learning_rate": 6.713332366168578e-06, "loss": 0.0315, "step": 18417 }, { "epoch": 1.9967476149176062, "grad_norm": 0.5151516199111938, "learning_rate": 6.712606992601191e-06, "loss": 0.016, "step": 18418 }, { "epoch": 1.9968560277536862, "grad_norm": 3.9585301876068115, "learning_rate": 6.711881619033803e-06, "loss": 0.0242, "step": 18419 }, { "epoch": 1.9969644405897657, "grad_norm": 2.109262704849243, "learning_rate": 6.711156245466416e-06, "loss": 0.0166, "step": 18420 }, { "epoch": 1.9970728534258457, "grad_norm": 0.3349097967147827, "learning_rate": 6.710430871899029e-06, "loss": 0.0061, "step": 18421 }, { "epoch": 1.9971812662619253, "grad_norm": 1.1299095153808594, "learning_rate": 6.709705498331641e-06, "loss": 0.0254, "step": 18422 }, { "epoch": 1.9972896790980053, "grad_norm": 1.234511137008667, "learning_rate": 6.708980124764254e-06, "loss": 0.0026, "step": 18423 }, { "epoch": 1.997398091934085, "grad_norm": 0.30855730175971985, "learning_rate": 6.708254751196866e-06, "loss": 0.0064, "step": 18424 }, { "epoch": 1.9975065047701648, "grad_norm": 1.7493960857391357, "learning_rate": 6.707529377629479e-06, "loss": 0.0114, "step": 18425 }, { "epoch": 1.9976149176062445, "grad_norm": 1.8088979721069336, "learning_rate": 6.706804004062093e-06, "loss": 0.0226, "step": 18426 }, { "epoch": 1.9977233304423243, "grad_norm": 2.6340854167938232, "learning_rate": 6.706078630494706e-06, "loss": 0.012, "step": 18427 }, { "epoch": 1.9978317432784043, "grad_norm": 1.9005624055862427, "learning_rate": 6.705353256927318e-06, "loss": 0.0225, "step": 18428 }, { "epoch": 1.9979401561144838, "grad_norm": 2.021052360534668, "learning_rate": 6.704627883359931e-06, "loss": 0.0073, "step": 18429 }, { "epoch": 1.9980485689505638, "grad_norm": 4.974847793579102, "learning_rate": 6.703902509792543e-06, "loss": 0.0147, "step": 18430 }, { "epoch": 1.9981569817866436, "grad_norm": 2.228914499282837, "learning_rate": 6.703177136225156e-06, "loss": 0.0323, "step": 18431 }, { "epoch": 1.9982653946227233, "grad_norm": 3.8879287242889404, "learning_rate": 6.702451762657769e-06, "loss": 0.0284, "step": 18432 }, { "epoch": 1.998373807458803, "grad_norm": 1.328416109085083, "learning_rate": 6.7017263890903814e-06, "loss": 0.0062, "step": 18433 }, { "epoch": 1.9984822202948829, "grad_norm": 0.5383108258247375, "learning_rate": 6.701001015522995e-06, "loss": 0.0028, "step": 18434 }, { "epoch": 1.9985906331309629, "grad_norm": 17.09422492980957, "learning_rate": 6.700275641955608e-06, "loss": 0.0487, "step": 18435 }, { "epoch": 1.9986990459670424, "grad_norm": 2.6367623805999756, "learning_rate": 6.69955026838822e-06, "loss": 0.01, "step": 18436 }, { "epoch": 1.9988074588031224, "grad_norm": 0.34053775668144226, "learning_rate": 6.698824894820833e-06, "loss": 0.0112, "step": 18437 }, { "epoch": 1.998915871639202, "grad_norm": 0.918017566204071, "learning_rate": 6.698099521253446e-06, "loss": 0.0213, "step": 18438 }, { "epoch": 1.999024284475282, "grad_norm": 3.6240763664245605, "learning_rate": 6.6973741476860585e-06, "loss": 0.0311, "step": 18439 }, { "epoch": 1.9991326973113617, "grad_norm": 0.502051591873169, "learning_rate": 6.6966487741186715e-06, "loss": 0.0031, "step": 18440 }, { "epoch": 1.9992411101474414, "grad_norm": 3.499380111694336, "learning_rate": 6.6959234005512845e-06, "loss": 0.0201, "step": 18441 }, { "epoch": 1.9993495229835212, "grad_norm": 0.6872121095657349, "learning_rate": 6.695198026983897e-06, "loss": 0.0084, "step": 18442 }, { "epoch": 1.999457935819601, "grad_norm": 0.3425630033016205, "learning_rate": 6.6944726534165105e-06, "loss": 0.0092, "step": 18443 }, { "epoch": 1.999566348655681, "grad_norm": 0.5476936101913452, "learning_rate": 6.6937472798491235e-06, "loss": 0.0094, "step": 18444 }, { "epoch": 1.9996747614917605, "grad_norm": 2.515338182449341, "learning_rate": 6.693021906281736e-06, "loss": 0.0368, "step": 18445 }, { "epoch": 1.9997831743278405, "grad_norm": 3.9283668994903564, "learning_rate": 6.692296532714349e-06, "loss": 0.0333, "step": 18446 }, { "epoch": 1.9998915871639202, "grad_norm": 1.3927348852157593, "learning_rate": 6.691571159146962e-06, "loss": 0.0331, "step": 18447 }, { "epoch": 2.0, "grad_norm": 0.38355711102485657, "learning_rate": 6.690845785579574e-06, "loss": 0.0015, "step": 18448 }, { "epoch": 2.00010841283608, "grad_norm": 0.6122352480888367, "learning_rate": 6.690120412012187e-06, "loss": 0.0021, "step": 18449 }, { "epoch": 2.0002168256721595, "grad_norm": 0.08043181151151657, "learning_rate": 6.689395038444799e-06, "loss": 0.0003, "step": 18450 }, { "epoch": 2.0003252385082395, "grad_norm": 1.152809977531433, "learning_rate": 6.688669664877412e-06, "loss": 0.0093, "step": 18451 }, { "epoch": 2.000433651344319, "grad_norm": 0.3359609544277191, "learning_rate": 6.687944291310026e-06, "loss": 0.0108, "step": 18452 }, { "epoch": 2.000542064180399, "grad_norm": 0.3894190788269043, "learning_rate": 6.687218917742639e-06, "loss": 0.0125, "step": 18453 }, { "epoch": 2.0006504770164786, "grad_norm": 0.6365669369697571, "learning_rate": 6.686493544175251e-06, "loss": 0.0077, "step": 18454 }, { "epoch": 2.0007588898525586, "grad_norm": 1.9806995391845703, "learning_rate": 6.685768170607864e-06, "loss": 0.0191, "step": 18455 }, { "epoch": 2.0008673026886385, "grad_norm": 2.8344430923461914, "learning_rate": 6.685042797040476e-06, "loss": 0.0696, "step": 18456 }, { "epoch": 2.000975715524718, "grad_norm": 1.3655712604522705, "learning_rate": 6.684317423473089e-06, "loss": 0.0084, "step": 18457 }, { "epoch": 2.001084128360798, "grad_norm": 0.03368523716926575, "learning_rate": 6.683592049905702e-06, "loss": 0.0003, "step": 18458 }, { "epoch": 2.0011925411968776, "grad_norm": 0.34055382013320923, "learning_rate": 6.682866676338314e-06, "loss": 0.0016, "step": 18459 }, { "epoch": 2.0013009540329576, "grad_norm": 1.0576115846633911, "learning_rate": 6.682141302770928e-06, "loss": 0.0075, "step": 18460 }, { "epoch": 2.001409366869037, "grad_norm": 0.12392938882112503, "learning_rate": 6.681415929203541e-06, "loss": 0.0013, "step": 18461 }, { "epoch": 2.001517779705117, "grad_norm": 0.8716841340065002, "learning_rate": 6.680690555636153e-06, "loss": 0.0209, "step": 18462 }, { "epoch": 2.0016261925411967, "grad_norm": 0.8171087503433228, "learning_rate": 6.679965182068766e-06, "loss": 0.0043, "step": 18463 }, { "epoch": 2.0017346053772767, "grad_norm": 0.6210882067680359, "learning_rate": 6.679239808501379e-06, "loss": 0.0275, "step": 18464 }, { "epoch": 2.0018430182133566, "grad_norm": 0.5316013693809509, "learning_rate": 6.678514434933991e-06, "loss": 0.0052, "step": 18465 }, { "epoch": 2.001951431049436, "grad_norm": 0.6438783407211304, "learning_rate": 6.677789061366604e-06, "loss": 0.0097, "step": 18466 }, { "epoch": 2.002059843885516, "grad_norm": 1.772873878479004, "learning_rate": 6.677063687799217e-06, "loss": 0.0355, "step": 18467 }, { "epoch": 2.0021682567215957, "grad_norm": 0.1172189936041832, "learning_rate": 6.676338314231829e-06, "loss": 0.0011, "step": 18468 }, { "epoch": 2.0022766695576757, "grad_norm": 0.5094093680381775, "learning_rate": 6.675612940664443e-06, "loss": 0.0045, "step": 18469 }, { "epoch": 2.0023850823937552, "grad_norm": 0.7487057447433472, "learning_rate": 6.674887567097056e-06, "loss": 0.0124, "step": 18470 }, { "epoch": 2.002493495229835, "grad_norm": 1.210382342338562, "learning_rate": 6.674162193529668e-06, "loss": 0.0068, "step": 18471 }, { "epoch": 2.002601908065915, "grad_norm": 0.9964191317558289, "learning_rate": 6.673436819962281e-06, "loss": 0.0067, "step": 18472 }, { "epoch": 2.0027103209019947, "grad_norm": 7.424534320831299, "learning_rate": 6.672711446394894e-06, "loss": 0.0181, "step": 18473 }, { "epoch": 2.0028187337380747, "grad_norm": 0.2603756785392761, "learning_rate": 6.671986072827506e-06, "loss": 0.0018, "step": 18474 }, { "epoch": 2.0029271465741543, "grad_norm": 0.34123313426971436, "learning_rate": 6.671260699260119e-06, "loss": 0.0066, "step": 18475 }, { "epoch": 2.0030355594102343, "grad_norm": 0.07498179376125336, "learning_rate": 6.6705353256927315e-06, "loss": 0.0006, "step": 18476 }, { "epoch": 2.003143972246314, "grad_norm": 0.03985412046313286, "learning_rate": 6.6698099521253445e-06, "loss": 0.0004, "step": 18477 }, { "epoch": 2.003252385082394, "grad_norm": 0.6153527498245239, "learning_rate": 6.669084578557958e-06, "loss": 0.0052, "step": 18478 }, { "epoch": 2.0033607979184738, "grad_norm": 0.3941037654876709, "learning_rate": 6.668359204990571e-06, "loss": 0.0257, "step": 18479 }, { "epoch": 2.0034692107545533, "grad_norm": 1.3159477710723877, "learning_rate": 6.6676338314231835e-06, "loss": 0.0081, "step": 18480 }, { "epoch": 2.0035776235906333, "grad_norm": 0.15804220736026764, "learning_rate": 6.6669084578557965e-06, "loss": 0.0009, "step": 18481 }, { "epoch": 2.003686036426713, "grad_norm": 0.4169195890426636, "learning_rate": 6.666183084288409e-06, "loss": 0.0324, "step": 18482 }, { "epoch": 2.003794449262793, "grad_norm": 0.03697223216295242, "learning_rate": 6.665457710721022e-06, "loss": 0.0003, "step": 18483 }, { "epoch": 2.0039028620988724, "grad_norm": 0.9169265627861023, "learning_rate": 6.664732337153635e-06, "loss": 0.0037, "step": 18484 }, { "epoch": 2.0040112749349523, "grad_norm": 0.5446261167526245, "learning_rate": 6.664006963586247e-06, "loss": 0.0117, "step": 18485 }, { "epoch": 2.004119687771032, "grad_norm": 0.16674113273620605, "learning_rate": 6.663281590018861e-06, "loss": 0.0027, "step": 18486 }, { "epoch": 2.004228100607112, "grad_norm": 0.7230163216590881, "learning_rate": 6.662556216451474e-06, "loss": 0.0015, "step": 18487 }, { "epoch": 2.004336513443192, "grad_norm": 2.0665066242218018, "learning_rate": 6.661830842884086e-06, "loss": 0.0292, "step": 18488 }, { "epoch": 2.0044449262792714, "grad_norm": 0.0618109293282032, "learning_rate": 6.661105469316699e-06, "loss": 0.0001, "step": 18489 }, { "epoch": 2.0045533391153514, "grad_norm": 1.8718469142913818, "learning_rate": 6.660380095749312e-06, "loss": 0.0106, "step": 18490 }, { "epoch": 2.004661751951431, "grad_norm": 0.12113521248102188, "learning_rate": 6.659654722181924e-06, "loss": 0.0106, "step": 18491 }, { "epoch": 2.004770164787511, "grad_norm": 0.9324817657470703, "learning_rate": 6.658929348614537e-06, "loss": 0.0031, "step": 18492 }, { "epoch": 2.0048785776235905, "grad_norm": 0.835646390914917, "learning_rate": 6.65820397504715e-06, "loss": 0.0023, "step": 18493 }, { "epoch": 2.0049869904596704, "grad_norm": 2.338793992996216, "learning_rate": 6.657478601479762e-06, "loss": 0.0324, "step": 18494 }, { "epoch": 2.0050954032957504, "grad_norm": 0.9173580408096313, "learning_rate": 6.656753227912376e-06, "loss": 0.0163, "step": 18495 }, { "epoch": 2.00520381613183, "grad_norm": 0.7839045524597168, "learning_rate": 6.656027854344989e-06, "loss": 0.0221, "step": 18496 }, { "epoch": 2.00531222896791, "grad_norm": 0.3044144809246063, "learning_rate": 6.655302480777601e-06, "loss": 0.0062, "step": 18497 }, { "epoch": 2.0054206418039895, "grad_norm": 0.0594874769449234, "learning_rate": 6.654577107210214e-06, "loss": 0.0004, "step": 18498 }, { "epoch": 2.0055290546400695, "grad_norm": 0.2668127119541168, "learning_rate": 6.653851733642827e-06, "loss": 0.0024, "step": 18499 }, { "epoch": 2.005637467476149, "grad_norm": 0.33086666464805603, "learning_rate": 6.653126360075439e-06, "loss": 0.002, "step": 18500 }, { "epoch": 2.005745880312229, "grad_norm": 0.760938286781311, "learning_rate": 6.652400986508052e-06, "loss": 0.0198, "step": 18501 }, { "epoch": 2.0058542931483085, "grad_norm": 0.7309538722038269, "learning_rate": 6.651675612940664e-06, "loss": 0.0161, "step": 18502 }, { "epoch": 2.0059627059843885, "grad_norm": 3.9360878467559814, "learning_rate": 6.650950239373278e-06, "loss": 0.033, "step": 18503 }, { "epoch": 2.0060711188204685, "grad_norm": 0.3573399782180786, "learning_rate": 6.650224865805891e-06, "loss": 0.0019, "step": 18504 }, { "epoch": 2.006179531656548, "grad_norm": 0.4784281253814697, "learning_rate": 6.649499492238504e-06, "loss": 0.022, "step": 18505 }, { "epoch": 2.006287944492628, "grad_norm": 1.190382719039917, "learning_rate": 6.648774118671116e-06, "loss": 0.0049, "step": 18506 }, { "epoch": 2.0063963573287076, "grad_norm": 0.2888100743293762, "learning_rate": 6.648048745103729e-06, "loss": 0.0117, "step": 18507 }, { "epoch": 2.0065047701647876, "grad_norm": 0.24254755675792694, "learning_rate": 6.647323371536341e-06, "loss": 0.005, "step": 18508 }, { "epoch": 2.006613183000867, "grad_norm": 0.4772666394710541, "learning_rate": 6.646597997968954e-06, "loss": 0.0094, "step": 18509 }, { "epoch": 2.006721595836947, "grad_norm": 3.057997703552246, "learning_rate": 6.645872624401567e-06, "loss": 0.0123, "step": 18510 }, { "epoch": 2.006830008673027, "grad_norm": 0.36774942278862, "learning_rate": 6.6451472508341794e-06, "loss": 0.0136, "step": 18511 }, { "epoch": 2.0069384215091066, "grad_norm": 0.11301496624946594, "learning_rate": 6.644421877266793e-06, "loss": 0.0006, "step": 18512 }, { "epoch": 2.0070468343451866, "grad_norm": 3.8677234649658203, "learning_rate": 6.643696503699406e-06, "loss": 0.0232, "step": 18513 }, { "epoch": 2.007155247181266, "grad_norm": 0.5124951004981995, "learning_rate": 6.642971130132018e-06, "loss": 0.0197, "step": 18514 }, { "epoch": 2.007263660017346, "grad_norm": 11.468936920166016, "learning_rate": 6.642245756564631e-06, "loss": 0.0152, "step": 18515 }, { "epoch": 2.0073720728534257, "grad_norm": 0.37753310799598694, "learning_rate": 6.641520382997244e-06, "loss": 0.0072, "step": 18516 }, { "epoch": 2.0074804856895057, "grad_norm": 3.9830570220947266, "learning_rate": 6.6407950094298565e-06, "loss": 0.0062, "step": 18517 }, { "epoch": 2.0075888985255856, "grad_norm": 0.15723201632499695, "learning_rate": 6.6400696358624695e-06, "loss": 0.0005, "step": 18518 }, { "epoch": 2.007697311361665, "grad_norm": 1.0076814889907837, "learning_rate": 6.6393442622950825e-06, "loss": 0.0022, "step": 18519 }, { "epoch": 2.007805724197745, "grad_norm": 1.3463733196258545, "learning_rate": 6.638618888727695e-06, "loss": 0.0098, "step": 18520 }, { "epoch": 2.0079141370338247, "grad_norm": 0.4694475531578064, "learning_rate": 6.6378935151603085e-06, "loss": 0.0006, "step": 18521 }, { "epoch": 2.0080225498699047, "grad_norm": 0.06867952644824982, "learning_rate": 6.6371681415929215e-06, "loss": 0.0004, "step": 18522 }, { "epoch": 2.0081309627059842, "grad_norm": 0.6709952354431152, "learning_rate": 6.636442768025534e-06, "loss": 0.0287, "step": 18523 }, { "epoch": 2.008239375542064, "grad_norm": 0.2833941578865051, "learning_rate": 6.635717394458147e-06, "loss": 0.0058, "step": 18524 }, { "epoch": 2.0083477883781438, "grad_norm": 1.325862169265747, "learning_rate": 6.63499202089076e-06, "loss": 0.0042, "step": 18525 }, { "epoch": 2.0084562012142237, "grad_norm": 2.4346890449523926, "learning_rate": 6.634266647323372e-06, "loss": 0.0092, "step": 18526 }, { "epoch": 2.0085646140503037, "grad_norm": 12.011093139648438, "learning_rate": 6.633541273755985e-06, "loss": 0.0178, "step": 18527 }, { "epoch": 2.0086730268863833, "grad_norm": 1.1277674436569214, "learning_rate": 6.632815900188597e-06, "loss": 0.0042, "step": 18528 }, { "epoch": 2.0087814397224633, "grad_norm": 0.6124005913734436, "learning_rate": 6.632090526621211e-06, "loss": 0.0246, "step": 18529 }, { "epoch": 2.008889852558543, "grad_norm": 0.3917793035507202, "learning_rate": 6.631365153053824e-06, "loss": 0.0226, "step": 18530 }, { "epoch": 2.008998265394623, "grad_norm": 0.642614483833313, "learning_rate": 6.630639779486437e-06, "loss": 0.0321, "step": 18531 }, { "epoch": 2.0091066782307023, "grad_norm": 0.12006977200508118, "learning_rate": 6.629914405919049e-06, "loss": 0.0118, "step": 18532 }, { "epoch": 2.0092150910667823, "grad_norm": 0.6296345591545105, "learning_rate": 6.629189032351662e-06, "loss": 0.0167, "step": 18533 }, { "epoch": 2.0093235039028623, "grad_norm": 0.3593909740447998, "learning_rate": 6.628463658784274e-06, "loss": 0.0133, "step": 18534 }, { "epoch": 2.009431916738942, "grad_norm": 0.030381111428141594, "learning_rate": 6.627738285216887e-06, "loss": 0.0003, "step": 18535 }, { "epoch": 2.009540329575022, "grad_norm": 1.0712419748306274, "learning_rate": 6.6270129116495e-06, "loss": 0.0055, "step": 18536 }, { "epoch": 2.0096487424111014, "grad_norm": 1.5881203413009644, "learning_rate": 6.626287538082112e-06, "loss": 0.0075, "step": 18537 }, { "epoch": 2.0097571552471813, "grad_norm": 0.29665443301200867, "learning_rate": 6.625562164514726e-06, "loss": 0.0178, "step": 18538 }, { "epoch": 2.009865568083261, "grad_norm": 5.035015106201172, "learning_rate": 6.624836790947339e-06, "loss": 0.0087, "step": 18539 }, { "epoch": 2.009973980919341, "grad_norm": 15.685863494873047, "learning_rate": 6.624111417379951e-06, "loss": 0.0414, "step": 18540 }, { "epoch": 2.010082393755421, "grad_norm": 0.3452596068382263, "learning_rate": 6.623386043812564e-06, "loss": 0.0012, "step": 18541 }, { "epoch": 2.0101908065915004, "grad_norm": 0.5433038473129272, "learning_rate": 6.622660670245177e-06, "loss": 0.0104, "step": 18542 }, { "epoch": 2.0102992194275804, "grad_norm": 0.6875519752502441, "learning_rate": 6.621935296677789e-06, "loss": 0.0216, "step": 18543 }, { "epoch": 2.01040763226366, "grad_norm": 0.3586939573287964, "learning_rate": 6.621209923110402e-06, "loss": 0.0106, "step": 18544 }, { "epoch": 2.01051604509974, "grad_norm": 0.5652293562889099, "learning_rate": 6.620484549543015e-06, "loss": 0.0031, "step": 18545 }, { "epoch": 2.0106244579358195, "grad_norm": 3.7297141551971436, "learning_rate": 6.619759175975627e-06, "loss": 0.0421, "step": 18546 }, { "epoch": 2.0107328707718994, "grad_norm": 8.085708618164062, "learning_rate": 6.619033802408241e-06, "loss": 0.0396, "step": 18547 }, { "epoch": 2.010841283607979, "grad_norm": 2.7728774547576904, "learning_rate": 6.618308428840854e-06, "loss": 0.0146, "step": 18548 }, { "epoch": 2.010949696444059, "grad_norm": 0.6905974745750427, "learning_rate": 6.617583055273466e-06, "loss": 0.0092, "step": 18549 }, { "epoch": 2.011058109280139, "grad_norm": 0.6503647565841675, "learning_rate": 6.616857681706079e-06, "loss": 0.0066, "step": 18550 }, { "epoch": 2.0111665221162185, "grad_norm": 0.5189678072929382, "learning_rate": 6.616132308138692e-06, "loss": 0.0145, "step": 18551 }, { "epoch": 2.0112749349522985, "grad_norm": 0.24448370933532715, "learning_rate": 6.615406934571304e-06, "loss": 0.0048, "step": 18552 }, { "epoch": 2.011383347788378, "grad_norm": 2.7132081985473633, "learning_rate": 6.614681561003917e-06, "loss": 0.0193, "step": 18553 }, { "epoch": 2.011491760624458, "grad_norm": 0.23075422644615173, "learning_rate": 6.6139561874365296e-06, "loss": 0.0182, "step": 18554 }, { "epoch": 2.0116001734605375, "grad_norm": 1.0692048072814941, "learning_rate": 6.613230813869143e-06, "loss": 0.0488, "step": 18555 }, { "epoch": 2.0117085862966175, "grad_norm": 5.021009922027588, "learning_rate": 6.612505440301756e-06, "loss": 0.0399, "step": 18556 }, { "epoch": 2.0118169991326975, "grad_norm": 0.3207659125328064, "learning_rate": 6.611780066734369e-06, "loss": 0.0085, "step": 18557 }, { "epoch": 2.011925411968777, "grad_norm": 0.0669420063495636, "learning_rate": 6.6110546931669815e-06, "loss": 0.0003, "step": 18558 }, { "epoch": 2.012033824804857, "grad_norm": 0.01672394759953022, "learning_rate": 6.6103293195995945e-06, "loss": 0.0001, "step": 18559 }, { "epoch": 2.0121422376409366, "grad_norm": 1.9610157012939453, "learning_rate": 6.609603946032207e-06, "loss": 0.0128, "step": 18560 }, { "epoch": 2.0122506504770166, "grad_norm": 0.125815287232399, "learning_rate": 6.60887857246482e-06, "loss": 0.0009, "step": 18561 }, { "epoch": 2.012359063313096, "grad_norm": 0.2985195517539978, "learning_rate": 6.608153198897433e-06, "loss": 0.0102, "step": 18562 }, { "epoch": 2.012467476149176, "grad_norm": 0.36852508783340454, "learning_rate": 6.607427825330045e-06, "loss": 0.0077, "step": 18563 }, { "epoch": 2.0125758889852556, "grad_norm": 0.3019046485424042, "learning_rate": 6.606702451762659e-06, "loss": 0.0009, "step": 18564 }, { "epoch": 2.0126843018213356, "grad_norm": 0.12581825256347656, "learning_rate": 6.605977078195272e-06, "loss": 0.0005, "step": 18565 }, { "epoch": 2.0127927146574156, "grad_norm": 0.18360771238803864, "learning_rate": 6.605251704627884e-06, "loss": 0.0027, "step": 18566 }, { "epoch": 2.012901127493495, "grad_norm": 0.2964757978916168, "learning_rate": 6.604526331060497e-06, "loss": 0.0107, "step": 18567 }, { "epoch": 2.013009540329575, "grad_norm": 1.889190435409546, "learning_rate": 6.60380095749311e-06, "loss": 0.0069, "step": 18568 }, { "epoch": 2.0131179531656547, "grad_norm": 2.705996036529541, "learning_rate": 6.603075583925722e-06, "loss": 0.018, "step": 18569 }, { "epoch": 2.0132263660017347, "grad_norm": 1.8630928993225098, "learning_rate": 6.602350210358335e-06, "loss": 0.0277, "step": 18570 }, { "epoch": 2.013334778837814, "grad_norm": 0.7061126232147217, "learning_rate": 6.601624836790948e-06, "loss": 0.0132, "step": 18571 }, { "epoch": 2.013443191673894, "grad_norm": 2.887774705886841, "learning_rate": 6.60089946322356e-06, "loss": 0.0029, "step": 18572 }, { "epoch": 2.013551604509974, "grad_norm": 0.3170143961906433, "learning_rate": 6.600174089656174e-06, "loss": 0.0275, "step": 18573 }, { "epoch": 2.0136600173460537, "grad_norm": 0.377619206905365, "learning_rate": 6.599448716088787e-06, "loss": 0.0114, "step": 18574 }, { "epoch": 2.0137684301821337, "grad_norm": 0.3564954102039337, "learning_rate": 6.598723342521399e-06, "loss": 0.0087, "step": 18575 }, { "epoch": 2.0138768430182132, "grad_norm": 0.46441879868507385, "learning_rate": 6.597997968954012e-06, "loss": 0.0118, "step": 18576 }, { "epoch": 2.013985255854293, "grad_norm": 1.9447879791259766, "learning_rate": 6.597272595386625e-06, "loss": 0.0198, "step": 18577 }, { "epoch": 2.0140936686903728, "grad_norm": 1.749289631843567, "learning_rate": 6.596547221819237e-06, "loss": 0.0167, "step": 18578 }, { "epoch": 2.0142020815264527, "grad_norm": 0.22346684336662292, "learning_rate": 6.59582184825185e-06, "loss": 0.0012, "step": 18579 }, { "epoch": 2.0143104943625327, "grad_norm": 2.4586386680603027, "learning_rate": 6.595096474684462e-06, "loss": 0.029, "step": 18580 }, { "epoch": 2.0144189071986123, "grad_norm": 10.184507369995117, "learning_rate": 6.594371101117076e-06, "loss": 0.0061, "step": 18581 }, { "epoch": 2.0145273200346923, "grad_norm": 0.2079792469739914, "learning_rate": 6.593645727549689e-06, "loss": 0.0007, "step": 18582 }, { "epoch": 2.014635732870772, "grad_norm": 0.567031741142273, "learning_rate": 6.592920353982302e-06, "loss": 0.0016, "step": 18583 }, { "epoch": 2.014744145706852, "grad_norm": 0.009860320948064327, "learning_rate": 6.592194980414914e-06, "loss": 0.0001, "step": 18584 }, { "epoch": 2.0148525585429313, "grad_norm": 0.6514960527420044, "learning_rate": 6.591469606847527e-06, "loss": 0.0198, "step": 18585 }, { "epoch": 2.0149609713790113, "grad_norm": 0.03889114409685135, "learning_rate": 6.590744233280139e-06, "loss": 0.0002, "step": 18586 }, { "epoch": 2.015069384215091, "grad_norm": 0.8484267592430115, "learning_rate": 6.590018859712752e-06, "loss": 0.0351, "step": 18587 }, { "epoch": 2.015177797051171, "grad_norm": 0.13331648707389832, "learning_rate": 6.589293486145365e-06, "loss": 0.0009, "step": 18588 }, { "epoch": 2.015286209887251, "grad_norm": 0.11473390460014343, "learning_rate": 6.5885681125779774e-06, "loss": 0.0003, "step": 18589 }, { "epoch": 2.0153946227233304, "grad_norm": 0.36655187606811523, "learning_rate": 6.587842739010591e-06, "loss": 0.0014, "step": 18590 }, { "epoch": 2.0155030355594103, "grad_norm": 0.7560021877288818, "learning_rate": 6.587117365443204e-06, "loss": 0.0154, "step": 18591 }, { "epoch": 2.01561144839549, "grad_norm": 0.411600798368454, "learning_rate": 6.586391991875816e-06, "loss": 0.0076, "step": 18592 }, { "epoch": 2.01571986123157, "grad_norm": 2.585284948348999, "learning_rate": 6.585666618308429e-06, "loss": 0.0471, "step": 18593 }, { "epoch": 2.0158282740676494, "grad_norm": 4.005432605743408, "learning_rate": 6.584941244741042e-06, "loss": 0.0234, "step": 18594 }, { "epoch": 2.0159366869037294, "grad_norm": 0.02394542656838894, "learning_rate": 6.5842158711736545e-06, "loss": 0.0001, "step": 18595 }, { "epoch": 2.0160450997398094, "grad_norm": 0.28594112396240234, "learning_rate": 6.5834904976062675e-06, "loss": 0.0043, "step": 18596 }, { "epoch": 2.016153512575889, "grad_norm": 0.9133322238922119, "learning_rate": 6.5827651240388805e-06, "loss": 0.0039, "step": 18597 }, { "epoch": 2.016261925411969, "grad_norm": 0.21128566563129425, "learning_rate": 6.5820397504714935e-06, "loss": 0.0037, "step": 18598 }, { "epoch": 2.0163703382480485, "grad_norm": 0.25488296151161194, "learning_rate": 6.5813143769041065e-06, "loss": 0.0051, "step": 18599 }, { "epoch": 2.0164787510841284, "grad_norm": 0.04976912587881088, "learning_rate": 6.5805890033367195e-06, "loss": 0.0004, "step": 18600 }, { "epoch": 2.016587163920208, "grad_norm": 0.2261964976787567, "learning_rate": 6.579863629769332e-06, "loss": 0.0168, "step": 18601 }, { "epoch": 2.016695576756288, "grad_norm": 1.7937791347503662, "learning_rate": 6.579138256201945e-06, "loss": 0.0552, "step": 18602 }, { "epoch": 2.016803989592368, "grad_norm": 0.4470209777355194, "learning_rate": 6.578412882634558e-06, "loss": 0.0015, "step": 18603 }, { "epoch": 2.0169124024284475, "grad_norm": 0.3901258707046509, "learning_rate": 6.57768750906717e-06, "loss": 0.0125, "step": 18604 }, { "epoch": 2.0170208152645275, "grad_norm": 1.5546913146972656, "learning_rate": 6.576962135499783e-06, "loss": 0.0099, "step": 18605 }, { "epoch": 2.017129228100607, "grad_norm": 0.07912880182266235, "learning_rate": 6.576236761932395e-06, "loss": 0.0004, "step": 18606 }, { "epoch": 2.017237640936687, "grad_norm": 0.45476794242858887, "learning_rate": 6.575511388365009e-06, "loss": 0.018, "step": 18607 }, { "epoch": 2.0173460537727665, "grad_norm": 0.4089871048927307, "learning_rate": 6.574786014797622e-06, "loss": 0.0074, "step": 18608 }, { "epoch": 2.0174544666088465, "grad_norm": 0.20194482803344727, "learning_rate": 6.574060641230235e-06, "loss": 0.0009, "step": 18609 }, { "epoch": 2.017562879444926, "grad_norm": 0.5070132613182068, "learning_rate": 6.573335267662847e-06, "loss": 0.0203, "step": 18610 }, { "epoch": 2.017671292281006, "grad_norm": 0.3471248149871826, "learning_rate": 6.57260989409546e-06, "loss": 0.0121, "step": 18611 }, { "epoch": 2.017779705117086, "grad_norm": 0.0963912233710289, "learning_rate": 6.571884520528072e-06, "loss": 0.0008, "step": 18612 }, { "epoch": 2.0178881179531656, "grad_norm": 0.13876955211162567, "learning_rate": 6.571159146960685e-06, "loss": 0.0006, "step": 18613 }, { "epoch": 2.0179965307892456, "grad_norm": 0.3697851002216339, "learning_rate": 6.570433773393298e-06, "loss": 0.0017, "step": 18614 }, { "epoch": 2.018104943625325, "grad_norm": 0.3626031279563904, "learning_rate": 6.56970839982591e-06, "loss": 0.0117, "step": 18615 }, { "epoch": 2.018213356461405, "grad_norm": 0.6563114523887634, "learning_rate": 6.568983026258524e-06, "loss": 0.0085, "step": 18616 }, { "epoch": 2.0183217692974846, "grad_norm": 0.6028862595558167, "learning_rate": 6.568257652691137e-06, "loss": 0.0188, "step": 18617 }, { "epoch": 2.0184301821335646, "grad_norm": 0.8895212411880493, "learning_rate": 6.567532279123749e-06, "loss": 0.0151, "step": 18618 }, { "epoch": 2.0185385949696446, "grad_norm": 0.6115695238113403, "learning_rate": 6.566806905556362e-06, "loss": 0.0183, "step": 18619 }, { "epoch": 2.018647007805724, "grad_norm": 0.37019842863082886, "learning_rate": 6.566081531988975e-06, "loss": 0.0109, "step": 18620 }, { "epoch": 2.018755420641804, "grad_norm": 0.03939791023731232, "learning_rate": 6.565356158421587e-06, "loss": 0.0001, "step": 18621 }, { "epoch": 2.0188638334778837, "grad_norm": 0.46722856163978577, "learning_rate": 6.5646307848542e-06, "loss": 0.0102, "step": 18622 }, { "epoch": 2.0189722463139637, "grad_norm": 0.09052428603172302, "learning_rate": 6.563905411286813e-06, "loss": 0.0006, "step": 18623 }, { "epoch": 2.019080659150043, "grad_norm": 0.05276096612215042, "learning_rate": 6.563180037719426e-06, "loss": 0.0002, "step": 18624 }, { "epoch": 2.019189071986123, "grad_norm": 0.036491766571998596, "learning_rate": 6.562454664152039e-06, "loss": 0.0003, "step": 18625 }, { "epoch": 2.0192974848222027, "grad_norm": 1.2013239860534668, "learning_rate": 6.561729290584652e-06, "loss": 0.012, "step": 18626 }, { "epoch": 2.0194058976582827, "grad_norm": 0.018495948985219002, "learning_rate": 6.561003917017264e-06, "loss": 0.0002, "step": 18627 }, { "epoch": 2.0195143104943627, "grad_norm": 1.1023584604263306, "learning_rate": 6.560278543449877e-06, "loss": 0.002, "step": 18628 }, { "epoch": 2.0196227233304422, "grad_norm": 0.009389178827404976, "learning_rate": 6.55955316988249e-06, "loss": 0.0001, "step": 18629 }, { "epoch": 2.0197311361665222, "grad_norm": 4.353659152984619, "learning_rate": 6.5588277963151024e-06, "loss": 0.0301, "step": 18630 }, { "epoch": 2.0198395490026018, "grad_norm": 0.3622344732284546, "learning_rate": 6.558102422747715e-06, "loss": 0.0102, "step": 18631 }, { "epoch": 2.0199479618386817, "grad_norm": 0.6074655652046204, "learning_rate": 6.5573770491803276e-06, "loss": 0.0042, "step": 18632 }, { "epoch": 2.0200563746747613, "grad_norm": 0.34505540132522583, "learning_rate": 6.556651675612941e-06, "loss": 0.0111, "step": 18633 }, { "epoch": 2.0201647875108413, "grad_norm": 0.6578876972198486, "learning_rate": 6.555926302045554e-06, "loss": 0.0098, "step": 18634 }, { "epoch": 2.0202732003469213, "grad_norm": 0.03282856196165085, "learning_rate": 6.555200928478167e-06, "loss": 0.0002, "step": 18635 }, { "epoch": 2.020381613183001, "grad_norm": 0.4282343089580536, "learning_rate": 6.5544755549107795e-06, "loss": 0.0091, "step": 18636 }, { "epoch": 2.020490026019081, "grad_norm": 1.0648587942123413, "learning_rate": 6.5537501813433925e-06, "loss": 0.0025, "step": 18637 }, { "epoch": 2.0205984388551603, "grad_norm": 0.5046409964561462, "learning_rate": 6.553024807776005e-06, "loss": 0.0116, "step": 18638 }, { "epoch": 2.0207068516912403, "grad_norm": 5.25960111618042, "learning_rate": 6.552299434208618e-06, "loss": 0.0074, "step": 18639 }, { "epoch": 2.02081526452732, "grad_norm": 3.901149034500122, "learning_rate": 6.551574060641231e-06, "loss": 0.0206, "step": 18640 }, { "epoch": 2.0209236773634, "grad_norm": 0.46164995431900024, "learning_rate": 6.550848687073843e-06, "loss": 0.0105, "step": 18641 }, { "epoch": 2.02103209019948, "grad_norm": 0.07219438254833221, "learning_rate": 6.550123313506457e-06, "loss": 0.0004, "step": 18642 }, { "epoch": 2.0211405030355594, "grad_norm": 3.9534764289855957, "learning_rate": 6.54939793993907e-06, "loss": 0.0326, "step": 18643 }, { "epoch": 2.0212489158716394, "grad_norm": 0.05435287952423096, "learning_rate": 6.548672566371682e-06, "loss": 0.0003, "step": 18644 }, { "epoch": 2.021357328707719, "grad_norm": 1.5211948156356812, "learning_rate": 6.547947192804295e-06, "loss": 0.0136, "step": 18645 }, { "epoch": 2.021465741543799, "grad_norm": 0.4840509593486786, "learning_rate": 6.547221819236908e-06, "loss": 0.0196, "step": 18646 }, { "epoch": 2.0215741543798784, "grad_norm": 14.620840072631836, "learning_rate": 6.54649644566952e-06, "loss": 0.0107, "step": 18647 }, { "epoch": 2.0216825672159584, "grad_norm": 0.039362650364637375, "learning_rate": 6.545771072102133e-06, "loss": 0.0001, "step": 18648 }, { "epoch": 2.021790980052038, "grad_norm": 6.006960868835449, "learning_rate": 6.545045698534746e-06, "loss": 0.0404, "step": 18649 }, { "epoch": 2.021899392888118, "grad_norm": 0.07146108150482178, "learning_rate": 6.544320324967359e-06, "loss": 0.0005, "step": 18650 }, { "epoch": 2.022007805724198, "grad_norm": 3.2960736751556396, "learning_rate": 6.543594951399972e-06, "loss": 0.071, "step": 18651 }, { "epoch": 2.0221162185602775, "grad_norm": 0.4114208221435547, "learning_rate": 6.542869577832585e-06, "loss": 0.0071, "step": 18652 }, { "epoch": 2.0222246313963574, "grad_norm": 1.4328800439834595, "learning_rate": 6.542144204265197e-06, "loss": 0.0086, "step": 18653 }, { "epoch": 2.022333044232437, "grad_norm": 0.29882127046585083, "learning_rate": 6.54141883069781e-06, "loss": 0.0141, "step": 18654 }, { "epoch": 2.022441457068517, "grad_norm": 0.08837521076202393, "learning_rate": 6.540693457130423e-06, "loss": 0.0002, "step": 18655 }, { "epoch": 2.0225498699045965, "grad_norm": 5.470886707305908, "learning_rate": 6.539968083563035e-06, "loss": 0.0117, "step": 18656 }, { "epoch": 2.0226582827406765, "grad_norm": 4.204073429107666, "learning_rate": 6.539242709995648e-06, "loss": 0.0291, "step": 18657 }, { "epoch": 2.0227666955767565, "grad_norm": 0.3024824857711792, "learning_rate": 6.53851733642826e-06, "loss": 0.0007, "step": 18658 }, { "epoch": 2.022875108412836, "grad_norm": 0.007131460588425398, "learning_rate": 6.537791962860874e-06, "loss": 0.0, "step": 18659 }, { "epoch": 2.022983521248916, "grad_norm": 0.013961010612547398, "learning_rate": 6.537066589293487e-06, "loss": 0.0001, "step": 18660 }, { "epoch": 2.0230919340849955, "grad_norm": 0.6919848322868347, "learning_rate": 6.5363412157261e-06, "loss": 0.0101, "step": 18661 }, { "epoch": 2.0232003469210755, "grad_norm": 0.756521999835968, "learning_rate": 6.535615842158712e-06, "loss": 0.0007, "step": 18662 }, { "epoch": 2.023308759757155, "grad_norm": 2.132453441619873, "learning_rate": 6.534890468591325e-06, "loss": 0.0393, "step": 18663 }, { "epoch": 2.023417172593235, "grad_norm": 0.2810741364955902, "learning_rate": 6.534165095023937e-06, "loss": 0.0056, "step": 18664 }, { "epoch": 2.023525585429315, "grad_norm": 0.3060592710971832, "learning_rate": 6.53343972145655e-06, "loss": 0.0079, "step": 18665 }, { "epoch": 2.0236339982653946, "grad_norm": 0.18550686538219452, "learning_rate": 6.532714347889163e-06, "loss": 0.0009, "step": 18666 }, { "epoch": 2.0237424111014746, "grad_norm": 0.4362712502479553, "learning_rate": 6.5319889743217754e-06, "loss": 0.0147, "step": 18667 }, { "epoch": 2.023850823937554, "grad_norm": 5.977529525756836, "learning_rate": 6.531263600754389e-06, "loss": 0.023, "step": 18668 }, { "epoch": 2.023959236773634, "grad_norm": 1.163537859916687, "learning_rate": 6.530538227187002e-06, "loss": 0.0026, "step": 18669 }, { "epoch": 2.0240676496097136, "grad_norm": 2.3670730590820312, "learning_rate": 6.529812853619614e-06, "loss": 0.0415, "step": 18670 }, { "epoch": 2.0241760624457936, "grad_norm": 0.43581023812294006, "learning_rate": 6.529087480052227e-06, "loss": 0.015, "step": 18671 }, { "epoch": 2.024284475281873, "grad_norm": 0.608040988445282, "learning_rate": 6.52836210648484e-06, "loss": 0.025, "step": 18672 }, { "epoch": 2.024392888117953, "grad_norm": 0.009635201655328274, "learning_rate": 6.5276367329174525e-06, "loss": 0.0001, "step": 18673 }, { "epoch": 2.024501300954033, "grad_norm": 7.061888694763184, "learning_rate": 6.5269113593500655e-06, "loss": 0.0461, "step": 18674 }, { "epoch": 2.0246097137901127, "grad_norm": 0.3488618731498718, "learning_rate": 6.5261859857826785e-06, "loss": 0.0007, "step": 18675 }, { "epoch": 2.0247181266261927, "grad_norm": 0.5013133883476257, "learning_rate": 6.5254606122152915e-06, "loss": 0.027, "step": 18676 }, { "epoch": 2.024826539462272, "grad_norm": 2.111633062362671, "learning_rate": 6.5247352386479045e-06, "loss": 0.021, "step": 18677 }, { "epoch": 2.024934952298352, "grad_norm": 0.09363090246915817, "learning_rate": 6.5240098650805175e-06, "loss": 0.0004, "step": 18678 }, { "epoch": 2.0250433651344317, "grad_norm": 0.40319445729255676, "learning_rate": 6.52328449151313e-06, "loss": 0.0117, "step": 18679 }, { "epoch": 2.0251517779705117, "grad_norm": 0.36691007018089294, "learning_rate": 6.522559117945743e-06, "loss": 0.0009, "step": 18680 }, { "epoch": 2.0252601908065917, "grad_norm": 0.016700949519872665, "learning_rate": 6.521833744378356e-06, "loss": 0.0001, "step": 18681 }, { "epoch": 2.0253686036426712, "grad_norm": 0.046589452773332596, "learning_rate": 6.521108370810968e-06, "loss": 0.0003, "step": 18682 }, { "epoch": 2.0254770164787512, "grad_norm": 4.058444976806641, "learning_rate": 6.520382997243581e-06, "loss": 0.0426, "step": 18683 }, { "epoch": 2.0255854293148308, "grad_norm": 0.017501279711723328, "learning_rate": 6.519657623676193e-06, "loss": 0.0002, "step": 18684 }, { "epoch": 2.0256938421509108, "grad_norm": 0.31172701716423035, "learning_rate": 6.518932250108807e-06, "loss": 0.0013, "step": 18685 }, { "epoch": 2.0258022549869903, "grad_norm": 0.7046546936035156, "learning_rate": 6.51820687654142e-06, "loss": 0.0156, "step": 18686 }, { "epoch": 2.0259106678230703, "grad_norm": 0.8700048327445984, "learning_rate": 6.517481502974033e-06, "loss": 0.0029, "step": 18687 }, { "epoch": 2.0260190806591503, "grad_norm": 5.143415927886963, "learning_rate": 6.516756129406645e-06, "loss": 0.0231, "step": 18688 }, { "epoch": 2.02612749349523, "grad_norm": 0.2524588406085968, "learning_rate": 6.516030755839258e-06, "loss": 0.0156, "step": 18689 }, { "epoch": 2.02623590633131, "grad_norm": 0.38625290989875793, "learning_rate": 6.51530538227187e-06, "loss": 0.0245, "step": 18690 }, { "epoch": 2.0263443191673893, "grad_norm": 0.019950976595282555, "learning_rate": 6.514580008704483e-06, "loss": 0.0001, "step": 18691 }, { "epoch": 2.0264527320034693, "grad_norm": 0.9237937927246094, "learning_rate": 6.513854635137096e-06, "loss": 0.0069, "step": 18692 }, { "epoch": 2.026561144839549, "grad_norm": 1.6683275699615479, "learning_rate": 6.51312926156971e-06, "loss": 0.0102, "step": 18693 }, { "epoch": 2.026669557675629, "grad_norm": 0.22046025097370148, "learning_rate": 6.512403888002322e-06, "loss": 0.0007, "step": 18694 }, { "epoch": 2.0267779705117084, "grad_norm": 0.022654511034488678, "learning_rate": 6.511678514434935e-06, "loss": 0.0001, "step": 18695 }, { "epoch": 2.0268863833477884, "grad_norm": 0.253730833530426, "learning_rate": 6.510953140867547e-06, "loss": 0.0046, "step": 18696 }, { "epoch": 2.0269947961838684, "grad_norm": 5.734254837036133, "learning_rate": 6.51022776730016e-06, "loss": 0.0122, "step": 18697 }, { "epoch": 2.027103209019948, "grad_norm": 2.8471641540527344, "learning_rate": 6.509502393732773e-06, "loss": 0.0407, "step": 18698 }, { "epoch": 2.027211621856028, "grad_norm": 5.804933547973633, "learning_rate": 6.508777020165385e-06, "loss": 0.0264, "step": 18699 }, { "epoch": 2.0273200346921074, "grad_norm": 0.0592949204146862, "learning_rate": 6.508051646597998e-06, "loss": 0.0002, "step": 18700 }, { "epoch": 2.0274284475281874, "grad_norm": 0.34019914269447327, "learning_rate": 6.507326273030611e-06, "loss": 0.0126, "step": 18701 }, { "epoch": 2.027536860364267, "grad_norm": 0.35240989923477173, "learning_rate": 6.506600899463224e-06, "loss": 0.0041, "step": 18702 }, { "epoch": 2.027645273200347, "grad_norm": 0.08024990558624268, "learning_rate": 6.505875525895837e-06, "loss": 0.0004, "step": 18703 }, { "epoch": 2.027753686036427, "grad_norm": 5.813528060913086, "learning_rate": 6.50515015232845e-06, "loss": 0.0111, "step": 18704 }, { "epoch": 2.0278620988725065, "grad_norm": 6.128537654876709, "learning_rate": 6.504424778761062e-06, "loss": 0.0047, "step": 18705 }, { "epoch": 2.0279705117085864, "grad_norm": 0.3520165979862213, "learning_rate": 6.503699405193675e-06, "loss": 0.0117, "step": 18706 }, { "epoch": 2.028078924544666, "grad_norm": 0.48158854246139526, "learning_rate": 6.502974031626288e-06, "loss": 0.0097, "step": 18707 }, { "epoch": 2.028187337380746, "grad_norm": 0.12658604979515076, "learning_rate": 6.5022486580589004e-06, "loss": 0.0005, "step": 18708 }, { "epoch": 2.0282957502168255, "grad_norm": 0.5741860270500183, "learning_rate": 6.501523284491513e-06, "loss": 0.0136, "step": 18709 }, { "epoch": 2.0284041630529055, "grad_norm": 0.05736108124256134, "learning_rate": 6.5007979109241256e-06, "loss": 0.0002, "step": 18710 }, { "epoch": 2.028512575888985, "grad_norm": 0.3321278691291809, "learning_rate": 6.500072537356739e-06, "loss": 0.0119, "step": 18711 }, { "epoch": 2.028620988725065, "grad_norm": 0.13034480810165405, "learning_rate": 6.499347163789352e-06, "loss": 0.0004, "step": 18712 }, { "epoch": 2.028729401561145, "grad_norm": 0.0059066955000162125, "learning_rate": 6.498621790221965e-06, "loss": 0.0, "step": 18713 }, { "epoch": 2.0288378143972245, "grad_norm": 0.4801068603992462, "learning_rate": 6.4978964166545775e-06, "loss": 0.0083, "step": 18714 }, { "epoch": 2.0289462272333045, "grad_norm": 0.16101379692554474, "learning_rate": 6.4971710430871905e-06, "loss": 0.0007, "step": 18715 }, { "epoch": 2.029054640069384, "grad_norm": 1.2224489450454712, "learning_rate": 6.496445669519803e-06, "loss": 0.0059, "step": 18716 }, { "epoch": 2.029163052905464, "grad_norm": 2.9820475578308105, "learning_rate": 6.495720295952416e-06, "loss": 0.0156, "step": 18717 }, { "epoch": 2.0292714657415436, "grad_norm": 0.7024425268173218, "learning_rate": 6.494994922385029e-06, "loss": 0.0163, "step": 18718 }, { "epoch": 2.0293798785776236, "grad_norm": 9.468441009521484, "learning_rate": 6.4942695488176425e-06, "loss": 0.0448, "step": 18719 }, { "epoch": 2.0294882914137036, "grad_norm": 0.2632649540901184, "learning_rate": 6.493544175250255e-06, "loss": 0.0074, "step": 18720 }, { "epoch": 2.029596704249783, "grad_norm": 0.5856404304504395, "learning_rate": 6.492818801682868e-06, "loss": 0.0185, "step": 18721 }, { "epoch": 2.029705117085863, "grad_norm": 1.9030669927597046, "learning_rate": 6.49209342811548e-06, "loss": 0.0439, "step": 18722 }, { "epoch": 2.0298135299219426, "grad_norm": 0.23533104360103607, "learning_rate": 6.491368054548093e-06, "loss": 0.0008, "step": 18723 }, { "epoch": 2.0299219427580226, "grad_norm": 1.2515543699264526, "learning_rate": 6.490642680980706e-06, "loss": 0.0088, "step": 18724 }, { "epoch": 2.030030355594102, "grad_norm": 0.3650374412536621, "learning_rate": 6.489917307413318e-06, "loss": 0.005, "step": 18725 }, { "epoch": 2.030138768430182, "grad_norm": 4.3013224601745605, "learning_rate": 6.489191933845931e-06, "loss": 0.0337, "step": 18726 }, { "epoch": 2.030247181266262, "grad_norm": 3.68328857421875, "learning_rate": 6.488466560278544e-06, "loss": 0.0066, "step": 18727 }, { "epoch": 2.0303555941023417, "grad_norm": 0.010080848820507526, "learning_rate": 6.487741186711157e-06, "loss": 0.0001, "step": 18728 }, { "epoch": 2.0304640069384217, "grad_norm": 0.4218764007091522, "learning_rate": 6.48701581314377e-06, "loss": 0.0112, "step": 18729 }, { "epoch": 2.030572419774501, "grad_norm": 1.1328628063201904, "learning_rate": 6.486290439576383e-06, "loss": 0.0023, "step": 18730 }, { "epoch": 2.030680832610581, "grad_norm": 0.3268081545829773, "learning_rate": 6.485565066008995e-06, "loss": 0.0099, "step": 18731 }, { "epoch": 2.0307892454466607, "grad_norm": 0.9624691605567932, "learning_rate": 6.484839692441608e-06, "loss": 0.0285, "step": 18732 }, { "epoch": 2.0308976582827407, "grad_norm": 5.843120098114014, "learning_rate": 6.484114318874221e-06, "loss": 0.0263, "step": 18733 }, { "epoch": 2.0310060711188203, "grad_norm": 0.051577307283878326, "learning_rate": 6.483388945306833e-06, "loss": 0.0002, "step": 18734 }, { "epoch": 2.0311144839549002, "grad_norm": 0.07191735506057739, "learning_rate": 6.482663571739446e-06, "loss": 0.0003, "step": 18735 }, { "epoch": 2.0312228967909802, "grad_norm": 0.6374619007110596, "learning_rate": 6.481938198172058e-06, "loss": 0.0115, "step": 18736 }, { "epoch": 2.0313313096270598, "grad_norm": 0.016469160094857216, "learning_rate": 6.481212824604672e-06, "loss": 0.0001, "step": 18737 }, { "epoch": 2.0314397224631398, "grad_norm": 1.5885149240493774, "learning_rate": 6.480487451037285e-06, "loss": 0.0634, "step": 18738 }, { "epoch": 2.0315481352992193, "grad_norm": 0.006067750975489616, "learning_rate": 6.479762077469898e-06, "loss": 0.0, "step": 18739 }, { "epoch": 2.0316565481352993, "grad_norm": 0.06234822794795036, "learning_rate": 6.47903670390251e-06, "loss": 0.0002, "step": 18740 }, { "epoch": 2.031764960971379, "grad_norm": 0.11998767405748367, "learning_rate": 6.478311330335123e-06, "loss": 0.0007, "step": 18741 }, { "epoch": 2.031873373807459, "grad_norm": 0.225261852145195, "learning_rate": 6.477585956767735e-06, "loss": 0.0006, "step": 18742 }, { "epoch": 2.031981786643539, "grad_norm": 7.271164417266846, "learning_rate": 6.476860583200348e-06, "loss": 0.0189, "step": 18743 }, { "epoch": 2.0320901994796183, "grad_norm": 0.28449010848999023, "learning_rate": 6.476135209632961e-06, "loss": 0.009, "step": 18744 }, { "epoch": 2.0321986123156983, "grad_norm": 2.4620022773742676, "learning_rate": 6.475409836065575e-06, "loss": 0.0507, "step": 18745 }, { "epoch": 2.032307025151778, "grad_norm": 0.22911567986011505, "learning_rate": 6.474684462498187e-06, "loss": 0.0042, "step": 18746 }, { "epoch": 2.032415437987858, "grad_norm": 0.19352105259895325, "learning_rate": 6.4739590889308e-06, "loss": 0.0034, "step": 18747 }, { "epoch": 2.0325238508239374, "grad_norm": 0.01853061094880104, "learning_rate": 6.4732337153634124e-06, "loss": 0.0001, "step": 18748 }, { "epoch": 2.0326322636600174, "grad_norm": 3.2459933757781982, "learning_rate": 6.472508341796025e-06, "loss": 0.0095, "step": 18749 }, { "epoch": 2.0327406764960974, "grad_norm": 0.20445241034030914, "learning_rate": 6.471782968228638e-06, "loss": 0.0032, "step": 18750 }, { "epoch": 2.032849089332177, "grad_norm": 0.00826869998127222, "learning_rate": 6.4710575946612505e-06, "loss": 0.0001, "step": 18751 }, { "epoch": 2.032957502168257, "grad_norm": 0.11814314872026443, "learning_rate": 6.4703322210938635e-06, "loss": 0.0004, "step": 18752 }, { "epoch": 2.0330659150043364, "grad_norm": 0.4193361699581146, "learning_rate": 6.4696068475264765e-06, "loss": 0.0094, "step": 18753 }, { "epoch": 2.0331743278404164, "grad_norm": 3.372309684753418, "learning_rate": 6.4688814739590895e-06, "loss": 0.0047, "step": 18754 }, { "epoch": 2.033282740676496, "grad_norm": 0.781477153301239, "learning_rate": 6.4681561003917025e-06, "loss": 0.0149, "step": 18755 }, { "epoch": 2.033391153512576, "grad_norm": 1.1671102046966553, "learning_rate": 6.4674307268243155e-06, "loss": 0.0085, "step": 18756 }, { "epoch": 2.0334995663486555, "grad_norm": 0.24233701825141907, "learning_rate": 6.466705353256928e-06, "loss": 0.0002, "step": 18757 }, { "epoch": 2.0336079791847355, "grad_norm": 6.712564945220947, "learning_rate": 6.465979979689541e-06, "loss": 0.0089, "step": 18758 }, { "epoch": 2.0337163920208154, "grad_norm": 0.35353678464889526, "learning_rate": 6.465254606122154e-06, "loss": 0.0121, "step": 18759 }, { "epoch": 2.033824804856895, "grad_norm": 0.03482615575194359, "learning_rate": 6.464529232554766e-06, "loss": 0.0002, "step": 18760 }, { "epoch": 2.033933217692975, "grad_norm": 0.03714291751384735, "learning_rate": 6.463803858987379e-06, "loss": 0.0002, "step": 18761 }, { "epoch": 2.0340416305290545, "grad_norm": 2.9146604537963867, "learning_rate": 6.463078485419991e-06, "loss": 0.0013, "step": 18762 }, { "epoch": 2.0341500433651345, "grad_norm": 2.497908592224121, "learning_rate": 6.462353111852605e-06, "loss": 0.0245, "step": 18763 }, { "epoch": 2.034258456201214, "grad_norm": 0.8060997128486633, "learning_rate": 6.461627738285218e-06, "loss": 0.004, "step": 18764 }, { "epoch": 2.034366869037294, "grad_norm": 0.040347080677747726, "learning_rate": 6.460902364717831e-06, "loss": 0.0002, "step": 18765 }, { "epoch": 2.034475281873374, "grad_norm": 0.3836052417755127, "learning_rate": 6.460176991150443e-06, "loss": 0.012, "step": 18766 }, { "epoch": 2.0345836947094535, "grad_norm": 23.466960906982422, "learning_rate": 6.459451617583056e-06, "loss": 0.0519, "step": 18767 }, { "epoch": 2.0346921075455335, "grad_norm": 0.8619051575660706, "learning_rate": 6.458726244015668e-06, "loss": 0.0264, "step": 18768 }, { "epoch": 2.034800520381613, "grad_norm": 0.5772110819816589, "learning_rate": 6.458000870448281e-06, "loss": 0.023, "step": 18769 }, { "epoch": 2.034908933217693, "grad_norm": 0.28235405683517456, "learning_rate": 6.457275496880894e-06, "loss": 0.0032, "step": 18770 }, { "epoch": 2.0350173460537726, "grad_norm": 2.7404439449310303, "learning_rate": 6.456550123313508e-06, "loss": 0.0206, "step": 18771 }, { "epoch": 2.0351257588898526, "grad_norm": 0.753961443901062, "learning_rate": 6.45582474974612e-06, "loss": 0.0214, "step": 18772 }, { "epoch": 2.035234171725932, "grad_norm": 0.5330139994621277, "learning_rate": 6.455099376178733e-06, "loss": 0.0152, "step": 18773 }, { "epoch": 2.035342584562012, "grad_norm": 0.3355467915534973, "learning_rate": 6.454374002611345e-06, "loss": 0.0095, "step": 18774 }, { "epoch": 2.035450997398092, "grad_norm": 0.7251346707344055, "learning_rate": 6.453648629043958e-06, "loss": 0.0025, "step": 18775 }, { "epoch": 2.0355594102341716, "grad_norm": 0.20452278852462769, "learning_rate": 6.452923255476571e-06, "loss": 0.0036, "step": 18776 }, { "epoch": 2.0356678230702516, "grad_norm": 0.9393226504325867, "learning_rate": 6.452197881909183e-06, "loss": 0.0014, "step": 18777 }, { "epoch": 2.035776235906331, "grad_norm": 0.37541359663009644, "learning_rate": 6.451472508341796e-06, "loss": 0.0067, "step": 18778 }, { "epoch": 2.035884648742411, "grad_norm": 0.6919270753860474, "learning_rate": 6.450747134774409e-06, "loss": 0.0179, "step": 18779 }, { "epoch": 2.0359930615784907, "grad_norm": 0.38531941175460815, "learning_rate": 6.450021761207023e-06, "loss": 0.0126, "step": 18780 }, { "epoch": 2.0361014744145707, "grad_norm": 0.00584869971498847, "learning_rate": 6.449296387639635e-06, "loss": 0.0, "step": 18781 }, { "epoch": 2.0362098872506507, "grad_norm": 2.8416998386383057, "learning_rate": 6.448571014072248e-06, "loss": 0.0061, "step": 18782 }, { "epoch": 2.03631830008673, "grad_norm": 0.14929050207138062, "learning_rate": 6.44784564050486e-06, "loss": 0.0004, "step": 18783 }, { "epoch": 2.03642671292281, "grad_norm": 0.0026372892316430807, "learning_rate": 6.447120266937473e-06, "loss": 0.0, "step": 18784 }, { "epoch": 2.0365351257588897, "grad_norm": 0.0017243953188881278, "learning_rate": 6.446394893370086e-06, "loss": 0.0, "step": 18785 }, { "epoch": 2.0366435385949697, "grad_norm": 1.1646151542663574, "learning_rate": 6.4456695198026984e-06, "loss": 0.0021, "step": 18786 }, { "epoch": 2.0367519514310493, "grad_norm": 2.831441640853882, "learning_rate": 6.4449441462353114e-06, "loss": 0.0316, "step": 18787 }, { "epoch": 2.0368603642671292, "grad_norm": 0.03325735032558441, "learning_rate": 6.444218772667925e-06, "loss": 0.0002, "step": 18788 }, { "epoch": 2.0369687771032092, "grad_norm": 0.33921298384666443, "learning_rate": 6.443493399100537e-06, "loss": 0.0082, "step": 18789 }, { "epoch": 2.0370771899392888, "grad_norm": 0.34329354763031006, "learning_rate": 6.44276802553315e-06, "loss": 0.012, "step": 18790 }, { "epoch": 2.0371856027753688, "grad_norm": 0.3287389576435089, "learning_rate": 6.442042651965763e-06, "loss": 0.0114, "step": 18791 }, { "epoch": 2.0372940156114483, "grad_norm": 0.006300782784819603, "learning_rate": 6.4413172783983755e-06, "loss": 0.0, "step": 18792 }, { "epoch": 2.0374024284475283, "grad_norm": 0.17278172075748444, "learning_rate": 6.4405919048309885e-06, "loss": 0.0018, "step": 18793 }, { "epoch": 2.037510841283608, "grad_norm": 0.01712542586028576, "learning_rate": 6.439866531263601e-06, "loss": 0.0001, "step": 18794 }, { "epoch": 2.037619254119688, "grad_norm": 0.34827157855033875, "learning_rate": 6.439141157696214e-06, "loss": 0.0072, "step": 18795 }, { "epoch": 2.0377276669557673, "grad_norm": 0.015017525292932987, "learning_rate": 6.438415784128827e-06, "loss": 0.0001, "step": 18796 }, { "epoch": 2.0378360797918473, "grad_norm": 1.6337699890136719, "learning_rate": 6.4376904105614405e-06, "loss": 0.0234, "step": 18797 }, { "epoch": 2.0379444926279273, "grad_norm": 0.5164227485656738, "learning_rate": 6.436965036994053e-06, "loss": 0.0024, "step": 18798 }, { "epoch": 2.038052905464007, "grad_norm": 0.40759336948394775, "learning_rate": 6.436239663426666e-06, "loss": 0.0078, "step": 18799 }, { "epoch": 2.038161318300087, "grad_norm": 0.017443666234612465, "learning_rate": 6.435514289859278e-06, "loss": 0.0001, "step": 18800 }, { "epoch": 2.0382697311361664, "grad_norm": 0.014130456373095512, "learning_rate": 6.434788916291891e-06, "loss": 0.0001, "step": 18801 }, { "epoch": 2.0383781439722464, "grad_norm": 1.8463209867477417, "learning_rate": 6.434063542724504e-06, "loss": 0.0067, "step": 18802 }, { "epoch": 2.038486556808326, "grad_norm": 0.01560972910374403, "learning_rate": 6.433338169157116e-06, "loss": 0.0001, "step": 18803 }, { "epoch": 2.038594969644406, "grad_norm": 0.1645561307668686, "learning_rate": 6.432612795589729e-06, "loss": 0.0003, "step": 18804 }, { "epoch": 2.038703382480486, "grad_norm": 0.7424262762069702, "learning_rate": 6.431887422022342e-06, "loss": 0.0169, "step": 18805 }, { "epoch": 2.0388117953165654, "grad_norm": 0.4767621159553528, "learning_rate": 6.431162048454956e-06, "loss": 0.0012, "step": 18806 }, { "epoch": 2.0389202081526454, "grad_norm": 0.4358499050140381, "learning_rate": 6.430436674887568e-06, "loss": 0.0049, "step": 18807 }, { "epoch": 2.039028620988725, "grad_norm": 0.2952995300292969, "learning_rate": 6.429711301320181e-06, "loss": 0.0011, "step": 18808 }, { "epoch": 2.039137033824805, "grad_norm": 0.39081665873527527, "learning_rate": 6.428985927752793e-06, "loss": 0.0171, "step": 18809 }, { "epoch": 2.0392454466608845, "grad_norm": 0.019172141328454018, "learning_rate": 6.428260554185406e-06, "loss": 0.0001, "step": 18810 }, { "epoch": 2.0393538594969645, "grad_norm": 16.543977737426758, "learning_rate": 6.427535180618019e-06, "loss": 0.0694, "step": 18811 }, { "epoch": 2.0394622723330444, "grad_norm": 23.244861602783203, "learning_rate": 6.426809807050631e-06, "loss": 0.0314, "step": 18812 }, { "epoch": 2.039570685169124, "grad_norm": 0.23893150687217712, "learning_rate": 6.426084433483244e-06, "loss": 0.0045, "step": 18813 }, { "epoch": 2.039679098005204, "grad_norm": 3.31715989112854, "learning_rate": 6.425359059915858e-06, "loss": 0.0283, "step": 18814 }, { "epoch": 2.0397875108412835, "grad_norm": 0.09731339663267136, "learning_rate": 6.42463368634847e-06, "loss": 0.0002, "step": 18815 }, { "epoch": 2.0398959236773635, "grad_norm": 0.33828726410865784, "learning_rate": 6.423908312781083e-06, "loss": 0.0047, "step": 18816 }, { "epoch": 2.040004336513443, "grad_norm": 0.2568559944629669, "learning_rate": 6.423182939213696e-06, "loss": 0.0009, "step": 18817 }, { "epoch": 2.040112749349523, "grad_norm": 0.2914339601993561, "learning_rate": 6.422457565646308e-06, "loss": 0.0044, "step": 18818 }, { "epoch": 2.0402211621856026, "grad_norm": 0.3887019157409668, "learning_rate": 6.421732192078921e-06, "loss": 0.0128, "step": 18819 }, { "epoch": 2.0403295750216826, "grad_norm": 4.88659143447876, "learning_rate": 6.421006818511533e-06, "loss": 0.0405, "step": 18820 }, { "epoch": 2.0404379878577625, "grad_norm": 0.759515106678009, "learning_rate": 6.420281444944146e-06, "loss": 0.0018, "step": 18821 }, { "epoch": 2.040546400693842, "grad_norm": 0.03157854080200195, "learning_rate": 6.419556071376759e-06, "loss": 0.0001, "step": 18822 }, { "epoch": 2.040654813529922, "grad_norm": 0.24201153218746185, "learning_rate": 6.418830697809373e-06, "loss": 0.0045, "step": 18823 }, { "epoch": 2.0407632263660016, "grad_norm": 0.36116307973861694, "learning_rate": 6.418105324241985e-06, "loss": 0.0078, "step": 18824 }, { "epoch": 2.0408716392020816, "grad_norm": 0.5501447916030884, "learning_rate": 6.417379950674598e-06, "loss": 0.0111, "step": 18825 }, { "epoch": 2.040980052038161, "grad_norm": 0.004648265894502401, "learning_rate": 6.4166545771072104e-06, "loss": 0.0, "step": 18826 }, { "epoch": 2.041088464874241, "grad_norm": 0.21743957698345184, "learning_rate": 6.415929203539823e-06, "loss": 0.0042, "step": 18827 }, { "epoch": 2.041196877710321, "grad_norm": 0.4195869266986847, "learning_rate": 6.415203829972436e-06, "loss": 0.0125, "step": 18828 }, { "epoch": 2.0413052905464006, "grad_norm": 0.5180922746658325, "learning_rate": 6.4144784564050486e-06, "loss": 0.0176, "step": 18829 }, { "epoch": 2.0414137033824806, "grad_norm": 5.445912837982178, "learning_rate": 6.4137530828376615e-06, "loss": 0.0613, "step": 18830 }, { "epoch": 2.04152211621856, "grad_norm": 0.0037580153439193964, "learning_rate": 6.4130277092702745e-06, "loss": 0.0, "step": 18831 }, { "epoch": 2.04163052905464, "grad_norm": 2.2610738277435303, "learning_rate": 6.412302335702888e-06, "loss": 0.0274, "step": 18832 }, { "epoch": 2.0417389418907197, "grad_norm": 1.7422771453857422, "learning_rate": 6.4115769621355005e-06, "loss": 0.0064, "step": 18833 }, { "epoch": 2.0418473547267997, "grad_norm": 1.8083585500717163, "learning_rate": 6.4108515885681135e-06, "loss": 0.0606, "step": 18834 }, { "epoch": 2.0419557675628797, "grad_norm": 0.3321373164653778, "learning_rate": 6.410126215000726e-06, "loss": 0.0006, "step": 18835 }, { "epoch": 2.042064180398959, "grad_norm": 1.0091429948806763, "learning_rate": 6.409400841433339e-06, "loss": 0.0164, "step": 18836 }, { "epoch": 2.042172593235039, "grad_norm": 0.01963597536087036, "learning_rate": 6.408675467865952e-06, "loss": 0.0001, "step": 18837 }, { "epoch": 2.0422810060711187, "grad_norm": 0.35103175044059753, "learning_rate": 6.407950094298564e-06, "loss": 0.0125, "step": 18838 }, { "epoch": 2.0423894189071987, "grad_norm": 0.2850116193294525, "learning_rate": 6.407224720731177e-06, "loss": 0.0009, "step": 18839 }, { "epoch": 2.0424978317432783, "grad_norm": 2.284412384033203, "learning_rate": 6.406499347163791e-06, "loss": 0.0234, "step": 18840 }, { "epoch": 2.0426062445793582, "grad_norm": 0.3061205744743347, "learning_rate": 6.405773973596403e-06, "loss": 0.0005, "step": 18841 }, { "epoch": 2.042714657415438, "grad_norm": 0.07033565640449524, "learning_rate": 6.405048600029016e-06, "loss": 0.0004, "step": 18842 }, { "epoch": 2.0428230702515178, "grad_norm": 4.415495872497559, "learning_rate": 6.404323226461629e-06, "loss": 0.0412, "step": 18843 }, { "epoch": 2.0429314830875978, "grad_norm": 5.099344253540039, "learning_rate": 6.403597852894241e-06, "loss": 0.025, "step": 18844 }, { "epoch": 2.0430398959236773, "grad_norm": 0.3692920506000519, "learning_rate": 6.402872479326854e-06, "loss": 0.0077, "step": 18845 }, { "epoch": 2.0431483087597573, "grad_norm": 0.7841330170631409, "learning_rate": 6.402147105759466e-06, "loss": 0.015, "step": 18846 }, { "epoch": 2.043256721595837, "grad_norm": 14.574803352355957, "learning_rate": 6.401421732192079e-06, "loss": 0.0199, "step": 18847 }, { "epoch": 2.043365134431917, "grad_norm": 3.378838539123535, "learning_rate": 6.400696358624692e-06, "loss": 0.0219, "step": 18848 }, { "epoch": 2.0434735472679963, "grad_norm": 0.09721849858760834, "learning_rate": 6.399970985057306e-06, "loss": 0.0005, "step": 18849 }, { "epoch": 2.0435819601040763, "grad_norm": 0.3493732213973999, "learning_rate": 6.399245611489918e-06, "loss": 0.0134, "step": 18850 }, { "epoch": 2.0436903729401563, "grad_norm": 0.009564264677464962, "learning_rate": 6.398520237922531e-06, "loss": 0.0001, "step": 18851 }, { "epoch": 2.043798785776236, "grad_norm": 0.5032451152801514, "learning_rate": 6.397794864355143e-06, "loss": 0.0113, "step": 18852 }, { "epoch": 2.043907198612316, "grad_norm": 0.0011936299270018935, "learning_rate": 6.397069490787756e-06, "loss": 0.0, "step": 18853 }, { "epoch": 2.0440156114483954, "grad_norm": 10.816368103027344, "learning_rate": 6.396344117220369e-06, "loss": 0.015, "step": 18854 }, { "epoch": 2.0441240242844754, "grad_norm": 0.5932748317718506, "learning_rate": 6.395618743652981e-06, "loss": 0.0203, "step": 18855 }, { "epoch": 2.044232437120555, "grad_norm": 1.1219303607940674, "learning_rate": 6.394893370085594e-06, "loss": 0.0123, "step": 18856 }, { "epoch": 2.044340849956635, "grad_norm": 0.7893503904342651, "learning_rate": 6.394167996518207e-06, "loss": 0.0114, "step": 18857 }, { "epoch": 2.0444492627927144, "grad_norm": 22.26360321044922, "learning_rate": 6.393442622950821e-06, "loss": 0.0249, "step": 18858 }, { "epoch": 2.0445576756287944, "grad_norm": 0.368467777967453, "learning_rate": 6.392717249383433e-06, "loss": 0.0059, "step": 18859 }, { "epoch": 2.0446660884648744, "grad_norm": 0.45918986201286316, "learning_rate": 6.391991875816046e-06, "loss": 0.0089, "step": 18860 }, { "epoch": 2.044774501300954, "grad_norm": 0.49747732281684875, "learning_rate": 6.391266502248658e-06, "loss": 0.0178, "step": 18861 }, { "epoch": 2.044882914137034, "grad_norm": 0.6292567253112793, "learning_rate": 6.390541128681271e-06, "loss": 0.0197, "step": 18862 }, { "epoch": 2.0449913269731135, "grad_norm": 0.7546218633651733, "learning_rate": 6.389815755113884e-06, "loss": 0.0048, "step": 18863 }, { "epoch": 2.0450997398091935, "grad_norm": 0.046370889991521835, "learning_rate": 6.3890903815464964e-06, "loss": 0.0002, "step": 18864 }, { "epoch": 2.045208152645273, "grad_norm": 0.18502749502658844, "learning_rate": 6.3883650079791094e-06, "loss": 0.0003, "step": 18865 }, { "epoch": 2.045316565481353, "grad_norm": 0.6505342125892639, "learning_rate": 6.387639634411723e-06, "loss": 0.0016, "step": 18866 }, { "epoch": 2.045424978317433, "grad_norm": 0.3025074899196625, "learning_rate": 6.386914260844335e-06, "loss": 0.0065, "step": 18867 }, { "epoch": 2.0455333911535125, "grad_norm": 0.2421104609966278, "learning_rate": 6.386188887276948e-06, "loss": 0.0046, "step": 18868 }, { "epoch": 2.0456418039895925, "grad_norm": 0.23237740993499756, "learning_rate": 6.385463513709561e-06, "loss": 0.0005, "step": 18869 }, { "epoch": 2.045750216825672, "grad_norm": 0.6537210941314697, "learning_rate": 6.3847381401421735e-06, "loss": 0.0121, "step": 18870 }, { "epoch": 2.045858629661752, "grad_norm": 0.6817895770072937, "learning_rate": 6.3840127665747865e-06, "loss": 0.0372, "step": 18871 }, { "epoch": 2.0459670424978316, "grad_norm": 4.801248550415039, "learning_rate": 6.383287393007399e-06, "loss": 0.0128, "step": 18872 }, { "epoch": 2.0460754553339116, "grad_norm": 3.4255480766296387, "learning_rate": 6.382562019440012e-06, "loss": 0.0455, "step": 18873 }, { "epoch": 2.0461838681699915, "grad_norm": 2.190028429031372, "learning_rate": 6.381836645872625e-06, "loss": 0.0231, "step": 18874 }, { "epoch": 2.046292281006071, "grad_norm": 2.899111270904541, "learning_rate": 6.3811112723052385e-06, "loss": 0.0565, "step": 18875 }, { "epoch": 2.046400693842151, "grad_norm": 0.046426232904195786, "learning_rate": 6.380385898737851e-06, "loss": 0.0001, "step": 18876 }, { "epoch": 2.0465091066782306, "grad_norm": 3.2149715423583984, "learning_rate": 6.379660525170464e-06, "loss": 0.0223, "step": 18877 }, { "epoch": 2.0466175195143106, "grad_norm": 0.32242345809936523, "learning_rate": 6.378935151603077e-06, "loss": 0.0114, "step": 18878 }, { "epoch": 2.04672593235039, "grad_norm": 3.5843448638916016, "learning_rate": 6.378209778035689e-06, "loss": 0.0343, "step": 18879 }, { "epoch": 2.04683434518647, "grad_norm": 4.006797790527344, "learning_rate": 6.377484404468302e-06, "loss": 0.0225, "step": 18880 }, { "epoch": 2.0469427580225497, "grad_norm": 0.4139193296432495, "learning_rate": 6.376759030900914e-06, "loss": 0.0102, "step": 18881 }, { "epoch": 2.0470511708586296, "grad_norm": 9.572025299072266, "learning_rate": 6.376033657333527e-06, "loss": 0.0141, "step": 18882 }, { "epoch": 2.0471595836947096, "grad_norm": 0.24754247069358826, "learning_rate": 6.375308283766141e-06, "loss": 0.0047, "step": 18883 }, { "epoch": 2.047267996530789, "grad_norm": 2.7739975452423096, "learning_rate": 6.374582910198754e-06, "loss": 0.0123, "step": 18884 }, { "epoch": 2.047376409366869, "grad_norm": 0.03757978603243828, "learning_rate": 6.373857536631366e-06, "loss": 0.0002, "step": 18885 }, { "epoch": 2.0474848222029487, "grad_norm": 1.314828634262085, "learning_rate": 6.373132163063979e-06, "loss": 0.0017, "step": 18886 }, { "epoch": 2.0475932350390287, "grad_norm": 0.5167147517204285, "learning_rate": 6.372406789496591e-06, "loss": 0.0165, "step": 18887 }, { "epoch": 2.047701647875108, "grad_norm": 0.36277565360069275, "learning_rate": 6.371681415929204e-06, "loss": 0.0115, "step": 18888 }, { "epoch": 2.047810060711188, "grad_norm": 0.5112881660461426, "learning_rate": 6.370956042361817e-06, "loss": 0.0253, "step": 18889 }, { "epoch": 2.047918473547268, "grad_norm": 0.2991999089717865, "learning_rate": 6.370230668794429e-06, "loss": 0.0107, "step": 18890 }, { "epoch": 2.0480268863833477, "grad_norm": 2.7856242656707764, "learning_rate": 6.369505295227042e-06, "loss": 0.0523, "step": 18891 }, { "epoch": 2.0481352992194277, "grad_norm": 2.3354010581970215, "learning_rate": 6.368779921659656e-06, "loss": 0.0078, "step": 18892 }, { "epoch": 2.0482437120555073, "grad_norm": 3.6636080741882324, "learning_rate": 6.368054548092268e-06, "loss": 0.0378, "step": 18893 }, { "epoch": 2.0483521248915872, "grad_norm": 0.22714263200759888, "learning_rate": 6.367329174524881e-06, "loss": 0.004, "step": 18894 }, { "epoch": 2.048460537727667, "grad_norm": 0.35929447412490845, "learning_rate": 6.366603800957494e-06, "loss": 0.0107, "step": 18895 }, { "epoch": 2.0485689505637468, "grad_norm": 4.133220672607422, "learning_rate": 6.365878427390106e-06, "loss": 0.0647, "step": 18896 }, { "epoch": 2.0486773633998268, "grad_norm": 0.23243623971939087, "learning_rate": 6.365153053822719e-06, "loss": 0.0042, "step": 18897 }, { "epoch": 2.0487857762359063, "grad_norm": 0.01373094879090786, "learning_rate": 6.364427680255331e-06, "loss": 0.0001, "step": 18898 }, { "epoch": 2.0488941890719863, "grad_norm": 0.36393582820892334, "learning_rate": 6.363702306687944e-06, "loss": 0.0077, "step": 18899 }, { "epoch": 2.049002601908066, "grad_norm": 0.11758144199848175, "learning_rate": 6.362976933120557e-06, "loss": 0.0008, "step": 18900 }, { "epoch": 2.049111014744146, "grad_norm": 0.4720686972141266, "learning_rate": 6.362251559553171e-06, "loss": 0.0071, "step": 18901 }, { "epoch": 2.0492194275802254, "grad_norm": 8.222095489501953, "learning_rate": 6.361526185985783e-06, "loss": 0.0398, "step": 18902 }, { "epoch": 2.0493278404163053, "grad_norm": 1.6429991722106934, "learning_rate": 6.360800812418396e-06, "loss": 0.0102, "step": 18903 }, { "epoch": 2.049436253252385, "grad_norm": 0.18352392315864563, "learning_rate": 6.360075438851009e-06, "loss": 0.0033, "step": 18904 }, { "epoch": 2.049544666088465, "grad_norm": 0.4635339677333832, "learning_rate": 6.3593500652836214e-06, "loss": 0.0145, "step": 18905 }, { "epoch": 2.049653078924545, "grad_norm": 5.029561519622803, "learning_rate": 6.358624691716234e-06, "loss": 0.0281, "step": 18906 }, { "epoch": 2.0497614917606244, "grad_norm": 2.942765951156616, "learning_rate": 6.3578993181488466e-06, "loss": 0.0233, "step": 18907 }, { "epoch": 2.0498699045967044, "grad_norm": 0.2624431252479553, "learning_rate": 6.3571739445814596e-06, "loss": 0.0046, "step": 18908 }, { "epoch": 2.049978317432784, "grad_norm": 0.5500547289848328, "learning_rate": 6.356448571014073e-06, "loss": 0.0132, "step": 18909 }, { "epoch": 2.050086730268864, "grad_norm": 0.4867381453514099, "learning_rate": 6.355723197446686e-06, "loss": 0.0092, "step": 18910 }, { "epoch": 2.0501951431049434, "grad_norm": 0.36691370606422424, "learning_rate": 6.3549978238792985e-06, "loss": 0.0111, "step": 18911 }, { "epoch": 2.0503035559410234, "grad_norm": 0.21910162270069122, "learning_rate": 6.3542724503119115e-06, "loss": 0.0045, "step": 18912 }, { "epoch": 2.0504119687771034, "grad_norm": 0.004185180179774761, "learning_rate": 6.353547076744524e-06, "loss": 0.0, "step": 18913 }, { "epoch": 2.050520381613183, "grad_norm": 0.031038450077176094, "learning_rate": 6.352821703177137e-06, "loss": 0.0002, "step": 18914 }, { "epoch": 2.050628794449263, "grad_norm": 0.1723726987838745, "learning_rate": 6.35209632960975e-06, "loss": 0.0006, "step": 18915 }, { "epoch": 2.0507372072853425, "grad_norm": 6.862821578979492, "learning_rate": 6.351370956042362e-06, "loss": 0.0371, "step": 18916 }, { "epoch": 2.0508456201214225, "grad_norm": 0.3216134011745453, "learning_rate": 6.350645582474975e-06, "loss": 0.0053, "step": 18917 }, { "epoch": 2.050954032957502, "grad_norm": 3.8148021697998047, "learning_rate": 6.349920208907589e-06, "loss": 0.0309, "step": 18918 }, { "epoch": 2.051062445793582, "grad_norm": 1.0263981819152832, "learning_rate": 6.349194835340201e-06, "loss": 0.0016, "step": 18919 }, { "epoch": 2.0511708586296615, "grad_norm": 0.3917987048625946, "learning_rate": 6.348469461772814e-06, "loss": 0.0107, "step": 18920 }, { "epoch": 2.0512792714657415, "grad_norm": 1.1393139362335205, "learning_rate": 6.347744088205427e-06, "loss": 0.0096, "step": 18921 }, { "epoch": 2.0513876843018215, "grad_norm": 8.80575942993164, "learning_rate": 6.347018714638039e-06, "loss": 0.0369, "step": 18922 }, { "epoch": 2.051496097137901, "grad_norm": 0.4561397135257721, "learning_rate": 6.346293341070652e-06, "loss": 0.0161, "step": 18923 }, { "epoch": 2.051604509973981, "grad_norm": 0.7472440004348755, "learning_rate": 6.345567967503264e-06, "loss": 0.0009, "step": 18924 }, { "epoch": 2.0517129228100606, "grad_norm": 0.011784739792346954, "learning_rate": 6.344842593935877e-06, "loss": 0.0001, "step": 18925 }, { "epoch": 2.0518213356461406, "grad_norm": 0.22767433524131775, "learning_rate": 6.34411722036849e-06, "loss": 0.0046, "step": 18926 }, { "epoch": 2.05192974848222, "grad_norm": 0.6534292101860046, "learning_rate": 6.343391846801104e-06, "loss": 0.009, "step": 18927 }, { "epoch": 2.0520381613183, "grad_norm": 0.7008994817733765, "learning_rate": 6.342666473233716e-06, "loss": 0.0022, "step": 18928 }, { "epoch": 2.05214657415438, "grad_norm": 0.780398428440094, "learning_rate": 6.341941099666329e-06, "loss": 0.0192, "step": 18929 }, { "epoch": 2.0522549869904596, "grad_norm": 1.099307656288147, "learning_rate": 6.341215726098942e-06, "loss": 0.0543, "step": 18930 }, { "epoch": 2.0523633998265396, "grad_norm": 0.03286723792552948, "learning_rate": 6.340490352531554e-06, "loss": 0.0002, "step": 18931 }, { "epoch": 2.052471812662619, "grad_norm": 0.32478809356689453, "learning_rate": 6.339764978964167e-06, "loss": 0.0115, "step": 18932 }, { "epoch": 2.052580225498699, "grad_norm": 1.1655460596084595, "learning_rate": 6.339039605396779e-06, "loss": 0.013, "step": 18933 }, { "epoch": 2.0526886383347787, "grad_norm": 5.764469623565674, "learning_rate": 6.338314231829392e-06, "loss": 0.02, "step": 18934 }, { "epoch": 2.0527970511708586, "grad_norm": 0.511121928691864, "learning_rate": 6.337588858262006e-06, "loss": 0.0175, "step": 18935 }, { "epoch": 2.0529054640069386, "grad_norm": 0.3102698028087616, "learning_rate": 6.336863484694619e-06, "loss": 0.0121, "step": 18936 }, { "epoch": 2.053013876843018, "grad_norm": 0.30955347418785095, "learning_rate": 6.336138111127231e-06, "loss": 0.0117, "step": 18937 }, { "epoch": 2.053122289679098, "grad_norm": 0.46166476607322693, "learning_rate": 6.335412737559844e-06, "loss": 0.015, "step": 18938 }, { "epoch": 2.0532307025151777, "grad_norm": 0.2432766854763031, "learning_rate": 6.334687363992456e-06, "loss": 0.005, "step": 18939 }, { "epoch": 2.0533391153512577, "grad_norm": 1.3041919469833374, "learning_rate": 6.333961990425069e-06, "loss": 0.0108, "step": 18940 }, { "epoch": 2.0534475281873372, "grad_norm": 1.1819584369659424, "learning_rate": 6.333236616857682e-06, "loss": 0.0127, "step": 18941 }, { "epoch": 2.053555941023417, "grad_norm": 0.25749143958091736, "learning_rate": 6.3325112432902945e-06, "loss": 0.0052, "step": 18942 }, { "epoch": 2.0536643538594967, "grad_norm": 0.36339810490608215, "learning_rate": 6.3317858697229074e-06, "loss": 0.0093, "step": 18943 }, { "epoch": 2.0537727666955767, "grad_norm": 0.2769409716129303, "learning_rate": 6.331060496155521e-06, "loss": 0.0061, "step": 18944 }, { "epoch": 2.0538811795316567, "grad_norm": 0.007717546075582504, "learning_rate": 6.330335122588133e-06, "loss": 0.0, "step": 18945 }, { "epoch": 2.0539895923677363, "grad_norm": 0.25026842951774597, "learning_rate": 6.329609749020746e-06, "loss": 0.0051, "step": 18946 }, { "epoch": 2.0540980052038162, "grad_norm": 0.4685315489768982, "learning_rate": 6.328884375453359e-06, "loss": 0.0007, "step": 18947 }, { "epoch": 2.054206418039896, "grad_norm": 0.24939477443695068, "learning_rate": 6.3281590018859715e-06, "loss": 0.0006, "step": 18948 }, { "epoch": 2.0543148308759758, "grad_norm": 0.9400825500488281, "learning_rate": 6.3274336283185845e-06, "loss": 0.0321, "step": 18949 }, { "epoch": 2.0544232437120553, "grad_norm": 0.4889639616012573, "learning_rate": 6.326708254751197e-06, "loss": 0.0291, "step": 18950 }, { "epoch": 2.0545316565481353, "grad_norm": 4.359314918518066, "learning_rate": 6.32598288118381e-06, "loss": 0.0083, "step": 18951 }, { "epoch": 2.0546400693842153, "grad_norm": 0.17979396879673004, "learning_rate": 6.325257507616423e-06, "loss": 0.0006, "step": 18952 }, { "epoch": 2.054748482220295, "grad_norm": 0.36330312490463257, "learning_rate": 6.3245321340490365e-06, "loss": 0.0137, "step": 18953 }, { "epoch": 2.054856895056375, "grad_norm": 0.3456159830093384, "learning_rate": 6.323806760481649e-06, "loss": 0.0213, "step": 18954 }, { "epoch": 2.0549653078924544, "grad_norm": 0.24824687838554382, "learning_rate": 6.323081386914262e-06, "loss": 0.0036, "step": 18955 }, { "epoch": 2.0550737207285343, "grad_norm": 1.0979382991790771, "learning_rate": 6.322356013346875e-06, "loss": 0.0121, "step": 18956 }, { "epoch": 2.055182133564614, "grad_norm": 0.12014967948198318, "learning_rate": 6.321630639779487e-06, "loss": 0.0002, "step": 18957 }, { "epoch": 2.055290546400694, "grad_norm": 0.5805631279945374, "learning_rate": 6.3209052662121e-06, "loss": 0.0187, "step": 18958 }, { "epoch": 2.055398959236774, "grad_norm": 3.9080893993377686, "learning_rate": 6.320179892644712e-06, "loss": 0.0172, "step": 18959 }, { "epoch": 2.0555073720728534, "grad_norm": 8.0487699508667, "learning_rate": 6.319454519077325e-06, "loss": 0.0134, "step": 18960 }, { "epoch": 2.0556157849089334, "grad_norm": 0.019764719530940056, "learning_rate": 6.318729145509939e-06, "loss": 0.0001, "step": 18961 }, { "epoch": 2.055724197745013, "grad_norm": 0.9878969192504883, "learning_rate": 6.318003771942552e-06, "loss": 0.0031, "step": 18962 }, { "epoch": 2.055832610581093, "grad_norm": 0.389151006937027, "learning_rate": 6.317278398375164e-06, "loss": 0.0008, "step": 18963 }, { "epoch": 2.0559410234171724, "grad_norm": 0.2584276497364044, "learning_rate": 6.316553024807777e-06, "loss": 0.0053, "step": 18964 }, { "epoch": 2.0560494362532524, "grad_norm": 3.9074809551239014, "learning_rate": 6.315827651240389e-06, "loss": 0.0335, "step": 18965 }, { "epoch": 2.056157849089332, "grad_norm": 18.298137664794922, "learning_rate": 6.315102277673002e-06, "loss": 0.0104, "step": 18966 }, { "epoch": 2.056266261925412, "grad_norm": 0.21097701787948608, "learning_rate": 6.314376904105615e-06, "loss": 0.0031, "step": 18967 }, { "epoch": 2.056374674761492, "grad_norm": 0.6580141186714172, "learning_rate": 6.313651530538227e-06, "loss": 0.0128, "step": 18968 }, { "epoch": 2.0564830875975715, "grad_norm": 0.4162302613258362, "learning_rate": 6.31292615697084e-06, "loss": 0.0219, "step": 18969 }, { "epoch": 2.0565915004336515, "grad_norm": 0.011589797213673592, "learning_rate": 6.312200783403454e-06, "loss": 0.0001, "step": 18970 }, { "epoch": 2.056699913269731, "grad_norm": 0.511162281036377, "learning_rate": 6.311475409836066e-06, "loss": 0.0251, "step": 18971 }, { "epoch": 2.056808326105811, "grad_norm": 1.1602745056152344, "learning_rate": 6.310750036268679e-06, "loss": 0.0066, "step": 18972 }, { "epoch": 2.0569167389418905, "grad_norm": 0.4933418035507202, "learning_rate": 6.310024662701292e-06, "loss": 0.0215, "step": 18973 }, { "epoch": 2.0570251517779705, "grad_norm": 6.138609886169434, "learning_rate": 6.309299289133904e-06, "loss": 0.0222, "step": 18974 }, { "epoch": 2.0571335646140505, "grad_norm": 21.177383422851562, "learning_rate": 6.308573915566517e-06, "loss": 0.022, "step": 18975 }, { "epoch": 2.05724197745013, "grad_norm": 0.15521280467510223, "learning_rate": 6.307848541999129e-06, "loss": 0.0005, "step": 18976 }, { "epoch": 2.05735039028621, "grad_norm": 0.024552740156650543, "learning_rate": 6.307123168431742e-06, "loss": 0.0003, "step": 18977 }, { "epoch": 2.0574588031222896, "grad_norm": 0.036592066287994385, "learning_rate": 6.306397794864355e-06, "loss": 0.0002, "step": 18978 }, { "epoch": 2.0575672159583696, "grad_norm": 0.367051362991333, "learning_rate": 6.305672421296969e-06, "loss": 0.0116, "step": 18979 }, { "epoch": 2.057675628794449, "grad_norm": 0.37381768226623535, "learning_rate": 6.304947047729581e-06, "loss": 0.012, "step": 18980 }, { "epoch": 2.057784041630529, "grad_norm": 4.96714448928833, "learning_rate": 6.304221674162194e-06, "loss": 0.0418, "step": 18981 }, { "epoch": 2.057892454466609, "grad_norm": 0.6069086790084839, "learning_rate": 6.303496300594807e-06, "loss": 0.0178, "step": 18982 }, { "epoch": 2.0580008673026886, "grad_norm": 0.28479477763175964, "learning_rate": 6.3027709270274194e-06, "loss": 0.0061, "step": 18983 }, { "epoch": 2.0581092801387686, "grad_norm": 3.047346591949463, "learning_rate": 6.3020455534600324e-06, "loss": 0.008, "step": 18984 }, { "epoch": 2.058217692974848, "grad_norm": 2.9039366245269775, "learning_rate": 6.3013201798926446e-06, "loss": 0.007, "step": 18985 }, { "epoch": 2.058326105810928, "grad_norm": 0.5105989575386047, "learning_rate": 6.3005948063252576e-06, "loss": 0.0015, "step": 18986 }, { "epoch": 2.0584345186470077, "grad_norm": 0.27006715536117554, "learning_rate": 6.299869432757871e-06, "loss": 0.0087, "step": 18987 }, { "epoch": 2.0585429314830876, "grad_norm": 4.1372199058532715, "learning_rate": 6.299144059190484e-06, "loss": 0.004, "step": 18988 }, { "epoch": 2.058651344319167, "grad_norm": 0.28540822863578796, "learning_rate": 6.2984186856230965e-06, "loss": 0.0046, "step": 18989 }, { "epoch": 2.058759757155247, "grad_norm": 0.36246639490127563, "learning_rate": 6.2976933120557095e-06, "loss": 0.0105, "step": 18990 }, { "epoch": 2.058868169991327, "grad_norm": 3.3264927864074707, "learning_rate": 6.296967938488322e-06, "loss": 0.1001, "step": 18991 }, { "epoch": 2.0589765828274067, "grad_norm": 0.02419106848537922, "learning_rate": 6.296242564920935e-06, "loss": 0.0002, "step": 18992 }, { "epoch": 2.0590849956634867, "grad_norm": 0.2544577121734619, "learning_rate": 6.295517191353548e-06, "loss": 0.0169, "step": 18993 }, { "epoch": 2.0591934084995662, "grad_norm": 4.980714797973633, "learning_rate": 6.29479181778616e-06, "loss": 0.0036, "step": 18994 }, { "epoch": 2.059301821335646, "grad_norm": 0.32281044125556946, "learning_rate": 6.294066444218773e-06, "loss": 0.0059, "step": 18995 }, { "epoch": 2.0594102341717258, "grad_norm": 0.481900155544281, "learning_rate": 6.293341070651387e-06, "loss": 0.0216, "step": 18996 }, { "epoch": 2.0595186470078057, "grad_norm": 0.18493130803108215, "learning_rate": 6.292615697083999e-06, "loss": 0.0006, "step": 18997 }, { "epoch": 2.0596270598438857, "grad_norm": 1.3813886642456055, "learning_rate": 6.291890323516612e-06, "loss": 0.0063, "step": 18998 }, { "epoch": 2.0597354726799653, "grad_norm": 0.1980239748954773, "learning_rate": 6.291164949949225e-06, "loss": 0.0008, "step": 18999 }, { "epoch": 2.0598438855160452, "grad_norm": 0.15899798274040222, "learning_rate": 6.290439576381837e-06, "loss": 0.0009, "step": 19000 }, { "epoch": 2.059952298352125, "grad_norm": 0.5491641759872437, "learning_rate": 6.28971420281445e-06, "loss": 0.0153, "step": 19001 }, { "epoch": 2.0600607111882048, "grad_norm": 2.0140998363494873, "learning_rate": 6.288988829247063e-06, "loss": 0.0132, "step": 19002 }, { "epoch": 2.0601691240242843, "grad_norm": 0.32665929198265076, "learning_rate": 6.288263455679675e-06, "loss": 0.0122, "step": 19003 }, { "epoch": 2.0602775368603643, "grad_norm": 0.5393140912055969, "learning_rate": 6.287538082112289e-06, "loss": 0.0008, "step": 19004 }, { "epoch": 2.060385949696444, "grad_norm": 0.2801160216331482, "learning_rate": 6.286812708544902e-06, "loss": 0.006, "step": 19005 }, { "epoch": 2.060494362532524, "grad_norm": 0.002005422255024314, "learning_rate": 6.286087334977514e-06, "loss": 0.0, "step": 19006 }, { "epoch": 2.060602775368604, "grad_norm": 0.7116613984107971, "learning_rate": 6.285361961410127e-06, "loss": 0.0096, "step": 19007 }, { "epoch": 2.0607111882046834, "grad_norm": 8.090059280395508, "learning_rate": 6.28463658784274e-06, "loss": 0.0301, "step": 19008 }, { "epoch": 2.0608196010407633, "grad_norm": 0.07970666140317917, "learning_rate": 6.283911214275352e-06, "loss": 0.0003, "step": 19009 }, { "epoch": 2.060928013876843, "grad_norm": 0.025494223460555077, "learning_rate": 6.283185840707965e-06, "loss": 0.0001, "step": 19010 }, { "epoch": 2.061036426712923, "grad_norm": 0.5280333757400513, "learning_rate": 6.282460467140577e-06, "loss": 0.0136, "step": 19011 }, { "epoch": 2.0611448395490024, "grad_norm": 0.4013572633266449, "learning_rate": 6.28173509357319e-06, "loss": 0.024, "step": 19012 }, { "epoch": 2.0612532523850824, "grad_norm": 3.1086266040802, "learning_rate": 6.281009720005804e-06, "loss": 0.0411, "step": 19013 }, { "epoch": 2.0613616652211624, "grad_norm": 18.068775177001953, "learning_rate": 6.280284346438417e-06, "loss": 0.0281, "step": 19014 }, { "epoch": 2.061470078057242, "grad_norm": 7.815779209136963, "learning_rate": 6.279558972871029e-06, "loss": 0.0226, "step": 19015 }, { "epoch": 2.061578490893322, "grad_norm": 0.49839478731155396, "learning_rate": 6.278833599303642e-06, "loss": 0.0059, "step": 19016 }, { "epoch": 2.0616869037294014, "grad_norm": 0.018746359273791313, "learning_rate": 6.278108225736254e-06, "loss": 0.0002, "step": 19017 }, { "epoch": 2.0617953165654814, "grad_norm": 0.018677031621336937, "learning_rate": 6.277382852168867e-06, "loss": 0.0001, "step": 19018 }, { "epoch": 2.061903729401561, "grad_norm": 0.541005551815033, "learning_rate": 6.27665747860148e-06, "loss": 0.0168, "step": 19019 }, { "epoch": 2.062012142237641, "grad_norm": 0.6040480136871338, "learning_rate": 6.2759321050340925e-06, "loss": 0.0015, "step": 19020 }, { "epoch": 2.062120555073721, "grad_norm": 0.19546005129814148, "learning_rate": 6.2752067314667054e-06, "loss": 0.0008, "step": 19021 }, { "epoch": 2.0622289679098005, "grad_norm": 0.6275780200958252, "learning_rate": 6.274481357899319e-06, "loss": 0.0349, "step": 19022 }, { "epoch": 2.0623373807458805, "grad_norm": 2.5666322708129883, "learning_rate": 6.2737559843319314e-06, "loss": 0.0083, "step": 19023 }, { "epoch": 2.06244579358196, "grad_norm": 0.044265713542699814, "learning_rate": 6.273030610764544e-06, "loss": 0.0001, "step": 19024 }, { "epoch": 2.06255420641804, "grad_norm": 0.2886579930782318, "learning_rate": 6.272305237197157e-06, "loss": 0.0008, "step": 19025 }, { "epoch": 2.0626626192541195, "grad_norm": 1.812898874282837, "learning_rate": 6.2715798636297696e-06, "loss": 0.0421, "step": 19026 }, { "epoch": 2.0627710320901995, "grad_norm": 0.3116930425167084, "learning_rate": 6.2708544900623825e-06, "loss": 0.0058, "step": 19027 }, { "epoch": 2.062879444926279, "grad_norm": 0.3471522927284241, "learning_rate": 6.2701291164949955e-06, "loss": 0.0015, "step": 19028 }, { "epoch": 2.062987857762359, "grad_norm": 4.317888259887695, "learning_rate": 6.269403742927608e-06, "loss": 0.0115, "step": 19029 }, { "epoch": 2.063096270598439, "grad_norm": 0.48978468775749207, "learning_rate": 6.2686783693602215e-06, "loss": 0.019, "step": 19030 }, { "epoch": 2.0632046834345186, "grad_norm": 0.039277952164411545, "learning_rate": 6.2679529957928345e-06, "loss": 0.0002, "step": 19031 }, { "epoch": 2.0633130962705986, "grad_norm": 0.0894092321395874, "learning_rate": 6.267227622225447e-06, "loss": 0.0004, "step": 19032 }, { "epoch": 2.063421509106678, "grad_norm": 0.4940645694732666, "learning_rate": 6.26650224865806e-06, "loss": 0.001, "step": 19033 }, { "epoch": 2.063529921942758, "grad_norm": 0.09878428280353546, "learning_rate": 6.265776875090673e-06, "loss": 0.0004, "step": 19034 }, { "epoch": 2.0636383347788376, "grad_norm": 1.422479510307312, "learning_rate": 6.265051501523285e-06, "loss": 0.0048, "step": 19035 }, { "epoch": 2.0637467476149176, "grad_norm": 0.687405526638031, "learning_rate": 6.264326127955898e-06, "loss": 0.0253, "step": 19036 }, { "epoch": 2.0638551604509976, "grad_norm": 0.4678427577018738, "learning_rate": 6.26360075438851e-06, "loss": 0.0059, "step": 19037 }, { "epoch": 2.063963573287077, "grad_norm": 0.017550397664308548, "learning_rate": 6.262875380821123e-06, "loss": 0.0001, "step": 19038 }, { "epoch": 2.064071986123157, "grad_norm": 0.661899209022522, "learning_rate": 6.262150007253737e-06, "loss": 0.0017, "step": 19039 }, { "epoch": 2.0641803989592367, "grad_norm": 0.28809309005737305, "learning_rate": 6.26142463368635e-06, "loss": 0.0088, "step": 19040 }, { "epoch": 2.0642888117953166, "grad_norm": 0.020577948540449142, "learning_rate": 6.260699260118962e-06, "loss": 0.0001, "step": 19041 }, { "epoch": 2.064397224631396, "grad_norm": 0.31304165720939636, "learning_rate": 6.259973886551575e-06, "loss": 0.0118, "step": 19042 }, { "epoch": 2.064505637467476, "grad_norm": 2.0565247535705566, "learning_rate": 6.259248512984187e-06, "loss": 0.0508, "step": 19043 }, { "epoch": 2.0646140503035557, "grad_norm": 0.6551663279533386, "learning_rate": 6.2585231394168e-06, "loss": 0.0181, "step": 19044 }, { "epoch": 2.0647224631396357, "grad_norm": 0.2940312623977661, "learning_rate": 6.257797765849413e-06, "loss": 0.0054, "step": 19045 }, { "epoch": 2.0648308759757157, "grad_norm": 3.1890721321105957, "learning_rate": 6.257072392282025e-06, "loss": 0.0204, "step": 19046 }, { "epoch": 2.0649392888117952, "grad_norm": 0.29620394110679626, "learning_rate": 6.256347018714638e-06, "loss": 0.0048, "step": 19047 }, { "epoch": 2.065047701647875, "grad_norm": 0.0169181227684021, "learning_rate": 6.255621645147252e-06, "loss": 0.0001, "step": 19048 }, { "epoch": 2.0651561144839548, "grad_norm": 0.43260207772254944, "learning_rate": 6.254896271579864e-06, "loss": 0.0261, "step": 19049 }, { "epoch": 2.0652645273200347, "grad_norm": 0.030313463881611824, "learning_rate": 6.254170898012477e-06, "loss": 0.0002, "step": 19050 }, { "epoch": 2.0653729401561143, "grad_norm": 0.24506786465644836, "learning_rate": 6.25344552444509e-06, "loss": 0.0053, "step": 19051 }, { "epoch": 2.0654813529921943, "grad_norm": 0.016149280592799187, "learning_rate": 6.252720150877702e-06, "loss": 0.0001, "step": 19052 }, { "epoch": 2.0655897658282742, "grad_norm": 1.4901994466781616, "learning_rate": 6.251994777310315e-06, "loss": 0.0009, "step": 19053 }, { "epoch": 2.065698178664354, "grad_norm": 7.51464319229126, "learning_rate": 6.251269403742928e-06, "loss": 0.0846, "step": 19054 }, { "epoch": 2.0658065915004338, "grad_norm": 0.12032455205917358, "learning_rate": 6.25054403017554e-06, "loss": 0.0005, "step": 19055 }, { "epoch": 2.0659150043365133, "grad_norm": 1.5777480602264404, "learning_rate": 6.249818656608154e-06, "loss": 0.0009, "step": 19056 }, { "epoch": 2.0660234171725933, "grad_norm": 0.48878154158592224, "learning_rate": 6.249093283040767e-06, "loss": 0.0172, "step": 19057 }, { "epoch": 2.066131830008673, "grad_norm": 0.7001110315322876, "learning_rate": 6.248367909473379e-06, "loss": 0.0016, "step": 19058 }, { "epoch": 2.066240242844753, "grad_norm": 7.8183722496032715, "learning_rate": 6.247642535905992e-06, "loss": 0.0469, "step": 19059 }, { "epoch": 2.066348655680833, "grad_norm": 4.391632080078125, "learning_rate": 6.246917162338605e-06, "loss": 0.0344, "step": 19060 }, { "epoch": 2.0664570685169124, "grad_norm": 3.7232720851898193, "learning_rate": 6.2461917887712174e-06, "loss": 0.0308, "step": 19061 }, { "epoch": 2.0665654813529923, "grad_norm": 0.6167246103286743, "learning_rate": 6.2454664152038304e-06, "loss": 0.0097, "step": 19062 }, { "epoch": 2.066673894189072, "grad_norm": 0.32286080718040466, "learning_rate": 6.2447410416364426e-06, "loss": 0.0072, "step": 19063 }, { "epoch": 2.066782307025152, "grad_norm": 0.02132934331893921, "learning_rate": 6.2440156680690556e-06, "loss": 0.0001, "step": 19064 }, { "epoch": 2.0668907198612314, "grad_norm": 0.0312550812959671, "learning_rate": 6.243290294501669e-06, "loss": 0.0002, "step": 19065 }, { "epoch": 2.0669991326973114, "grad_norm": 0.47841399908065796, "learning_rate": 6.242564920934282e-06, "loss": 0.0096, "step": 19066 }, { "epoch": 2.067107545533391, "grad_norm": 5.103862762451172, "learning_rate": 6.2418395473668945e-06, "loss": 0.0274, "step": 19067 }, { "epoch": 2.067215958369471, "grad_norm": 0.228728249669075, "learning_rate": 6.2411141737995075e-06, "loss": 0.0041, "step": 19068 }, { "epoch": 2.067324371205551, "grad_norm": 2.5327236652374268, "learning_rate": 6.24038880023212e-06, "loss": 0.0089, "step": 19069 }, { "epoch": 2.0674327840416304, "grad_norm": 1.0432133674621582, "learning_rate": 6.239663426664733e-06, "loss": 0.0019, "step": 19070 }, { "epoch": 2.0675411968777104, "grad_norm": 0.129823237657547, "learning_rate": 6.238938053097346e-06, "loss": 0.0004, "step": 19071 }, { "epoch": 2.06764960971379, "grad_norm": 3.101489305496216, "learning_rate": 6.238212679529958e-06, "loss": 0.0188, "step": 19072 }, { "epoch": 2.06775802254987, "grad_norm": 0.8272630572319031, "learning_rate": 6.237487305962571e-06, "loss": 0.0013, "step": 19073 }, { "epoch": 2.0678664353859495, "grad_norm": 4.009174823760986, "learning_rate": 6.236761932395185e-06, "loss": 0.0028, "step": 19074 }, { "epoch": 2.0679748482220295, "grad_norm": 0.288480669260025, "learning_rate": 6.236036558827797e-06, "loss": 0.0233, "step": 19075 }, { "epoch": 2.0680832610581095, "grad_norm": 0.27445122599601746, "learning_rate": 6.23531118526041e-06, "loss": 0.0046, "step": 19076 }, { "epoch": 2.068191673894189, "grad_norm": 2.3529844284057617, "learning_rate": 6.234585811693023e-06, "loss": 0.0029, "step": 19077 }, { "epoch": 2.068300086730269, "grad_norm": 0.42838776111602783, "learning_rate": 6.233860438125635e-06, "loss": 0.0011, "step": 19078 }, { "epoch": 2.0684084995663485, "grad_norm": 0.3382244110107422, "learning_rate": 6.233135064558248e-06, "loss": 0.0115, "step": 19079 }, { "epoch": 2.0685169124024285, "grad_norm": 2.8440046310424805, "learning_rate": 6.232409690990861e-06, "loss": 0.0144, "step": 19080 }, { "epoch": 2.068625325238508, "grad_norm": 0.211544468998909, "learning_rate": 6.231684317423473e-06, "loss": 0.0039, "step": 19081 }, { "epoch": 2.068733738074588, "grad_norm": 12.135592460632324, "learning_rate": 6.230958943856087e-06, "loss": 0.0425, "step": 19082 }, { "epoch": 2.068842150910668, "grad_norm": 0.9262174367904663, "learning_rate": 6.2302335702887e-06, "loss": 0.002, "step": 19083 }, { "epoch": 2.0689505637467476, "grad_norm": 0.5335162281990051, "learning_rate": 6.229508196721312e-06, "loss": 0.0346, "step": 19084 }, { "epoch": 2.0690589765828276, "grad_norm": 0.33326447010040283, "learning_rate": 6.228782823153925e-06, "loss": 0.0059, "step": 19085 }, { "epoch": 2.069167389418907, "grad_norm": 0.37032490968704224, "learning_rate": 6.228057449586538e-06, "loss": 0.013, "step": 19086 }, { "epoch": 2.069275802254987, "grad_norm": 0.37628886103630066, "learning_rate": 6.22733207601915e-06, "loss": 0.012, "step": 19087 }, { "epoch": 2.0693842150910666, "grad_norm": 1.9638639688491821, "learning_rate": 6.226606702451763e-06, "loss": 0.0012, "step": 19088 }, { "epoch": 2.0694926279271466, "grad_norm": 0.3058273196220398, "learning_rate": 6.225881328884375e-06, "loss": 0.0065, "step": 19089 }, { "epoch": 2.069601040763226, "grad_norm": 0.2323349118232727, "learning_rate": 6.225155955316988e-06, "loss": 0.0044, "step": 19090 }, { "epoch": 2.069709453599306, "grad_norm": 0.27847346663475037, "learning_rate": 6.224430581749602e-06, "loss": 0.0154, "step": 19091 }, { "epoch": 2.069817866435386, "grad_norm": 0.013226219452917576, "learning_rate": 6.223705208182215e-06, "loss": 0.0001, "step": 19092 }, { "epoch": 2.0699262792714657, "grad_norm": 0.2085294872522354, "learning_rate": 6.222979834614827e-06, "loss": 0.0041, "step": 19093 }, { "epoch": 2.0700346921075456, "grad_norm": 0.6529048085212708, "learning_rate": 6.22225446104744e-06, "loss": 0.0283, "step": 19094 }, { "epoch": 2.070143104943625, "grad_norm": 7.552515983581543, "learning_rate": 6.221529087480052e-06, "loss": 0.0696, "step": 19095 }, { "epoch": 2.070251517779705, "grad_norm": 0.3265049457550049, "learning_rate": 6.220803713912665e-06, "loss": 0.0058, "step": 19096 }, { "epoch": 2.0703599306157847, "grad_norm": 0.036007508635520935, "learning_rate": 6.220078340345278e-06, "loss": 0.0003, "step": 19097 }, { "epoch": 2.0704683434518647, "grad_norm": 0.36286377906799316, "learning_rate": 6.2193529667778905e-06, "loss": 0.0121, "step": 19098 }, { "epoch": 2.0705767562879447, "grad_norm": 0.1875964105129242, "learning_rate": 6.218627593210504e-06, "loss": 0.0005, "step": 19099 }, { "epoch": 2.0706851691240242, "grad_norm": 2.5618088245391846, "learning_rate": 6.217902219643117e-06, "loss": 0.0449, "step": 19100 }, { "epoch": 2.070793581960104, "grad_norm": 0.007438457105308771, "learning_rate": 6.2171768460757294e-06, "loss": 0.0, "step": 19101 }, { "epoch": 2.0709019947961838, "grad_norm": 1.3482416868209839, "learning_rate": 6.2164514725083424e-06, "loss": 0.0031, "step": 19102 }, { "epoch": 2.0710104076322637, "grad_norm": 0.6791034936904907, "learning_rate": 6.215726098940955e-06, "loss": 0.0094, "step": 19103 }, { "epoch": 2.0711188204683433, "grad_norm": 0.03535345941781998, "learning_rate": 6.2150007253735676e-06, "loss": 0.0003, "step": 19104 }, { "epoch": 2.0712272333044233, "grad_norm": 0.004633828066289425, "learning_rate": 6.2142753518061805e-06, "loss": 0.0, "step": 19105 }, { "epoch": 2.0713356461405033, "grad_norm": 3.139669179916382, "learning_rate": 6.2135499782387935e-06, "loss": 0.0279, "step": 19106 }, { "epoch": 2.071444058976583, "grad_norm": 6.822243690490723, "learning_rate": 6.212824604671406e-06, "loss": 0.0596, "step": 19107 }, { "epoch": 2.0715524718126628, "grad_norm": 0.65694659948349, "learning_rate": 6.2120992311040195e-06, "loss": 0.0077, "step": 19108 }, { "epoch": 2.0716608846487423, "grad_norm": 0.37483730912208557, "learning_rate": 6.2113738575366325e-06, "loss": 0.0009, "step": 19109 }, { "epoch": 2.0717692974848223, "grad_norm": 0.5459954142570496, "learning_rate": 6.210648483969245e-06, "loss": 0.0163, "step": 19110 }, { "epoch": 2.071877710320902, "grad_norm": 0.7816259860992432, "learning_rate": 6.209923110401858e-06, "loss": 0.0014, "step": 19111 }, { "epoch": 2.071986123156982, "grad_norm": 0.07765655219554901, "learning_rate": 6.209197736834471e-06, "loss": 0.0003, "step": 19112 }, { "epoch": 2.0720945359930614, "grad_norm": 4.680686950683594, "learning_rate": 6.208472363267083e-06, "loss": 0.0331, "step": 19113 }, { "epoch": 2.0722029488291414, "grad_norm": 12.751998901367188, "learning_rate": 6.207746989699696e-06, "loss": 0.064, "step": 19114 }, { "epoch": 2.0723113616652213, "grad_norm": 0.393551230430603, "learning_rate": 6.207021616132308e-06, "loss": 0.0107, "step": 19115 }, { "epoch": 2.072419774501301, "grad_norm": 0.3398526906967163, "learning_rate": 6.206296242564921e-06, "loss": 0.0062, "step": 19116 }, { "epoch": 2.072528187337381, "grad_norm": 0.2221784144639969, "learning_rate": 6.205570868997535e-06, "loss": 0.0004, "step": 19117 }, { "epoch": 2.0726366001734604, "grad_norm": 2.8571057319641113, "learning_rate": 6.204845495430148e-06, "loss": 0.0317, "step": 19118 }, { "epoch": 2.0727450130095404, "grad_norm": 1.2567847967147827, "learning_rate": 6.20412012186276e-06, "loss": 0.0206, "step": 19119 }, { "epoch": 2.07285342584562, "grad_norm": 16.542848587036133, "learning_rate": 6.203394748295373e-06, "loss": 0.0199, "step": 19120 }, { "epoch": 2.0729618386817, "grad_norm": 0.3724603056907654, "learning_rate": 6.202669374727985e-06, "loss": 0.0093, "step": 19121 }, { "epoch": 2.07307025151778, "grad_norm": 0.48426589369773865, "learning_rate": 6.201944001160598e-06, "loss": 0.0095, "step": 19122 }, { "epoch": 2.0731786643538594, "grad_norm": 0.001110748271457851, "learning_rate": 6.201218627593211e-06, "loss": 0.0, "step": 19123 }, { "epoch": 2.0732870771899394, "grad_norm": 0.8264920115470886, "learning_rate": 6.200493254025823e-06, "loss": 0.0111, "step": 19124 }, { "epoch": 2.073395490026019, "grad_norm": 0.24359917640686035, "learning_rate": 6.199767880458437e-06, "loss": 0.0042, "step": 19125 }, { "epoch": 2.073503902862099, "grad_norm": 0.3945618271827698, "learning_rate": 6.19904250689105e-06, "loss": 0.001, "step": 19126 }, { "epoch": 2.0736123156981785, "grad_norm": 5.012815952301025, "learning_rate": 6.198317133323662e-06, "loss": 0.0362, "step": 19127 }, { "epoch": 2.0737207285342585, "grad_norm": 0.732185959815979, "learning_rate": 6.197591759756275e-06, "loss": 0.0017, "step": 19128 }, { "epoch": 2.0738291413703385, "grad_norm": 2.682687997817993, "learning_rate": 6.196866386188888e-06, "loss": 0.0008, "step": 19129 }, { "epoch": 2.073937554206418, "grad_norm": 1.1548963785171509, "learning_rate": 6.1961410126215e-06, "loss": 0.001, "step": 19130 }, { "epoch": 2.074045967042498, "grad_norm": 1.2079347372055054, "learning_rate": 6.195415639054113e-06, "loss": 0.0064, "step": 19131 }, { "epoch": 2.0741543798785775, "grad_norm": 0.24776984751224518, "learning_rate": 6.194690265486726e-06, "loss": 0.0044, "step": 19132 }, { "epoch": 2.0742627927146575, "grad_norm": 2.391711950302124, "learning_rate": 6.193964891919338e-06, "loss": 0.0482, "step": 19133 }, { "epoch": 2.074371205550737, "grad_norm": 15.956344604492188, "learning_rate": 6.193239518351952e-06, "loss": 0.033, "step": 19134 }, { "epoch": 2.074479618386817, "grad_norm": 1.0587877035140991, "learning_rate": 6.192514144784565e-06, "loss": 0.0111, "step": 19135 }, { "epoch": 2.0745880312228966, "grad_norm": 0.8485131859779358, "learning_rate": 6.191788771217177e-06, "loss": 0.0127, "step": 19136 }, { "epoch": 2.0746964440589766, "grad_norm": 0.3683811128139496, "learning_rate": 6.19106339764979e-06, "loss": 0.0061, "step": 19137 }, { "epoch": 2.0748048568950566, "grad_norm": 0.3639148771762848, "learning_rate": 6.190338024082403e-06, "loss": 0.0065, "step": 19138 }, { "epoch": 2.074913269731136, "grad_norm": 2.010507106781006, "learning_rate": 6.1896126505150154e-06, "loss": 0.0042, "step": 19139 }, { "epoch": 2.075021682567216, "grad_norm": 1.0137890577316284, "learning_rate": 6.1888872769476284e-06, "loss": 0.0027, "step": 19140 }, { "epoch": 2.0751300954032956, "grad_norm": 0.45814940333366394, "learning_rate": 6.188161903380241e-06, "loss": 0.0052, "step": 19141 }, { "epoch": 2.0752385082393756, "grad_norm": 8.075505256652832, "learning_rate": 6.1874365298128536e-06, "loss": 0.018, "step": 19142 }, { "epoch": 2.075346921075455, "grad_norm": 0.011431621387600899, "learning_rate": 6.186711156245467e-06, "loss": 0.0001, "step": 19143 }, { "epoch": 2.075455333911535, "grad_norm": 0.06382421404123306, "learning_rate": 6.18598578267808e-06, "loss": 0.0002, "step": 19144 }, { "epoch": 2.075563746747615, "grad_norm": 4.276103973388672, "learning_rate": 6.1852604091106925e-06, "loss": 0.023, "step": 19145 }, { "epoch": 2.0756721595836947, "grad_norm": 1.9607881307601929, "learning_rate": 6.1845350355433055e-06, "loss": 0.0095, "step": 19146 }, { "epoch": 2.0757805724197746, "grad_norm": 2.055997133255005, "learning_rate": 6.183809661975918e-06, "loss": 0.0226, "step": 19147 }, { "epoch": 2.075888985255854, "grad_norm": 0.1769610047340393, "learning_rate": 6.183084288408531e-06, "loss": 0.0031, "step": 19148 }, { "epoch": 2.075997398091934, "grad_norm": 0.3763732314109802, "learning_rate": 6.182358914841144e-06, "loss": 0.0167, "step": 19149 }, { "epoch": 2.0761058109280137, "grad_norm": 0.41454842686653137, "learning_rate": 6.181633541273756e-06, "loss": 0.0123, "step": 19150 }, { "epoch": 2.0762142237640937, "grad_norm": 0.5592687726020813, "learning_rate": 6.18090816770637e-06, "loss": 0.0277, "step": 19151 }, { "epoch": 2.0763226366001737, "grad_norm": 6.264283180236816, "learning_rate": 6.180182794138983e-06, "loss": 0.015, "step": 19152 }, { "epoch": 2.0764310494362532, "grad_norm": 0.014726627618074417, "learning_rate": 6.179457420571595e-06, "loss": 0.0001, "step": 19153 }, { "epoch": 2.076539462272333, "grad_norm": 4.258795261383057, "learning_rate": 6.178732047004208e-06, "loss": 0.0151, "step": 19154 }, { "epoch": 2.0766478751084128, "grad_norm": 0.0031467003282159567, "learning_rate": 6.178006673436821e-06, "loss": 0.0, "step": 19155 }, { "epoch": 2.0767562879444927, "grad_norm": 0.35093918442726135, "learning_rate": 6.177281299869433e-06, "loss": 0.0127, "step": 19156 }, { "epoch": 2.0768647007805723, "grad_norm": 0.11165641248226166, "learning_rate": 6.176555926302046e-06, "loss": 0.0005, "step": 19157 }, { "epoch": 2.0769731136166523, "grad_norm": 0.032568346709012985, "learning_rate": 6.175830552734659e-06, "loss": 0.0002, "step": 19158 }, { "epoch": 2.077081526452732, "grad_norm": 3.580113649368286, "learning_rate": 6.175105179167271e-06, "loss": 0.0655, "step": 19159 }, { "epoch": 2.077189939288812, "grad_norm": 3.4743080139160156, "learning_rate": 6.174379805599885e-06, "loss": 0.0145, "step": 19160 }, { "epoch": 2.077298352124892, "grad_norm": 0.013822236098349094, "learning_rate": 6.173654432032498e-06, "loss": 0.0001, "step": 19161 }, { "epoch": 2.0774067649609713, "grad_norm": 0.024948105216026306, "learning_rate": 6.17292905846511e-06, "loss": 0.0001, "step": 19162 }, { "epoch": 2.0775151777970513, "grad_norm": 0.49550899863243103, "learning_rate": 6.172203684897723e-06, "loss": 0.0097, "step": 19163 }, { "epoch": 2.077623590633131, "grad_norm": 1.0112324953079224, "learning_rate": 6.171478311330336e-06, "loss": 0.0397, "step": 19164 }, { "epoch": 2.077732003469211, "grad_norm": 0.02651507779955864, "learning_rate": 6.170752937762948e-06, "loss": 0.0002, "step": 19165 }, { "epoch": 2.0778404163052904, "grad_norm": 1.8281581401824951, "learning_rate": 6.170027564195561e-06, "loss": 0.0526, "step": 19166 }, { "epoch": 2.0779488291413704, "grad_norm": 0.5192272663116455, "learning_rate": 6.169302190628173e-06, "loss": 0.0103, "step": 19167 }, { "epoch": 2.0780572419774503, "grad_norm": 0.4304587244987488, "learning_rate": 6.168576817060786e-06, "loss": 0.01, "step": 19168 }, { "epoch": 2.07816565481353, "grad_norm": 0.5193438529968262, "learning_rate": 6.1678514434934e-06, "loss": 0.0189, "step": 19169 }, { "epoch": 2.07827406764961, "grad_norm": 6.5848069190979, "learning_rate": 6.167126069926013e-06, "loss": 0.0521, "step": 19170 }, { "epoch": 2.0783824804856894, "grad_norm": 3.6372897624969482, "learning_rate": 6.166400696358625e-06, "loss": 0.047, "step": 19171 }, { "epoch": 2.0784908933217694, "grad_norm": 0.5977292656898499, "learning_rate": 6.165675322791238e-06, "loss": 0.0045, "step": 19172 }, { "epoch": 2.078599306157849, "grad_norm": 13.299240112304688, "learning_rate": 6.16494994922385e-06, "loss": 0.0185, "step": 19173 }, { "epoch": 2.078707718993929, "grad_norm": 0.3879539966583252, "learning_rate": 6.164224575656463e-06, "loss": 0.0096, "step": 19174 }, { "epoch": 2.0788161318300085, "grad_norm": 0.29280996322631836, "learning_rate": 6.163499202089076e-06, "loss": 0.0039, "step": 19175 }, { "epoch": 2.0789245446660884, "grad_norm": 0.6305917501449585, "learning_rate": 6.1627738285216885e-06, "loss": 0.0176, "step": 19176 }, { "epoch": 2.0790329575021684, "grad_norm": 6.5343499183654785, "learning_rate": 6.162048454954302e-06, "loss": 0.0493, "step": 19177 }, { "epoch": 2.079141370338248, "grad_norm": 0.03463393449783325, "learning_rate": 6.161323081386915e-06, "loss": 0.0003, "step": 19178 }, { "epoch": 2.079249783174328, "grad_norm": 7.808938980102539, "learning_rate": 6.1605977078195274e-06, "loss": 0.027, "step": 19179 }, { "epoch": 2.0793581960104075, "grad_norm": 0.8174359202384949, "learning_rate": 6.1598723342521404e-06, "loss": 0.0009, "step": 19180 }, { "epoch": 2.0794666088464875, "grad_norm": 0.47954273223876953, "learning_rate": 6.159146960684753e-06, "loss": 0.0125, "step": 19181 }, { "epoch": 2.079575021682567, "grad_norm": 0.371977835893631, "learning_rate": 6.1584215871173656e-06, "loss": 0.0102, "step": 19182 }, { "epoch": 2.079683434518647, "grad_norm": 0.03620713949203491, "learning_rate": 6.1576962135499786e-06, "loss": 0.0002, "step": 19183 }, { "epoch": 2.079791847354727, "grad_norm": 1.1651349067687988, "learning_rate": 6.1569708399825915e-06, "loss": 0.0026, "step": 19184 }, { "epoch": 2.0799002601908065, "grad_norm": 0.7831078171730042, "learning_rate": 6.156245466415204e-06, "loss": 0.0204, "step": 19185 }, { "epoch": 2.0800086730268865, "grad_norm": 0.3090008497238159, "learning_rate": 6.1555200928478175e-06, "loss": 0.0066, "step": 19186 }, { "epoch": 2.080117085862966, "grad_norm": 0.18319135904312134, "learning_rate": 6.1547947192804305e-06, "loss": 0.0035, "step": 19187 }, { "epoch": 2.080225498699046, "grad_norm": 5.055363655090332, "learning_rate": 6.154069345713043e-06, "loss": 0.0913, "step": 19188 }, { "epoch": 2.0803339115351256, "grad_norm": 0.021503882482647896, "learning_rate": 6.153343972145656e-06, "loss": 0.0001, "step": 19189 }, { "epoch": 2.0804423243712056, "grad_norm": 0.36643683910369873, "learning_rate": 6.152618598578269e-06, "loss": 0.0095, "step": 19190 }, { "epoch": 2.080550737207285, "grad_norm": 0.42269349098205566, "learning_rate": 6.151893225010881e-06, "loss": 0.0084, "step": 19191 }, { "epoch": 2.080659150043365, "grad_norm": 11.21512222290039, "learning_rate": 6.151167851443494e-06, "loss": 0.0237, "step": 19192 }, { "epoch": 2.080767562879445, "grad_norm": 11.761100769042969, "learning_rate": 6.150442477876106e-06, "loss": 0.0406, "step": 19193 }, { "epoch": 2.0808759757155246, "grad_norm": 1.5755531787872314, "learning_rate": 6.14971710430872e-06, "loss": 0.015, "step": 19194 }, { "epoch": 2.0809843885516046, "grad_norm": 9.80647087097168, "learning_rate": 6.148991730741333e-06, "loss": 0.0296, "step": 19195 }, { "epoch": 2.081092801387684, "grad_norm": 20.85367202758789, "learning_rate": 6.148266357173946e-06, "loss": 0.0082, "step": 19196 }, { "epoch": 2.081201214223764, "grad_norm": 0.4880266487598419, "learning_rate": 6.147540983606558e-06, "loss": 0.0195, "step": 19197 }, { "epoch": 2.0813096270598437, "grad_norm": 7.897656440734863, "learning_rate": 6.146815610039171e-06, "loss": 0.0305, "step": 19198 }, { "epoch": 2.0814180398959237, "grad_norm": 8.561052322387695, "learning_rate": 6.146090236471783e-06, "loss": 0.0356, "step": 19199 }, { "epoch": 2.0815264527320037, "grad_norm": 0.3223089575767517, "learning_rate": 6.145364862904396e-06, "loss": 0.0104, "step": 19200 }, { "epoch": 2.081634865568083, "grad_norm": 0.09769245982170105, "learning_rate": 6.144639489337009e-06, "loss": 0.0003, "step": 19201 }, { "epoch": 2.081743278404163, "grad_norm": 0.26705509424209595, "learning_rate": 6.143914115769621e-06, "loss": 0.0042, "step": 19202 }, { "epoch": 2.0818516912402427, "grad_norm": 0.10107006132602692, "learning_rate": 6.143188742202235e-06, "loss": 0.0006, "step": 19203 }, { "epoch": 2.0819601040763227, "grad_norm": 0.48146188259124756, "learning_rate": 6.142463368634848e-06, "loss": 0.0134, "step": 19204 }, { "epoch": 2.0820685169124022, "grad_norm": 0.3147105574607849, "learning_rate": 6.14173799506746e-06, "loss": 0.0083, "step": 19205 }, { "epoch": 2.0821769297484822, "grad_norm": 1.606352686882019, "learning_rate": 6.141012621500073e-06, "loss": 0.0121, "step": 19206 }, { "epoch": 2.082285342584562, "grad_norm": 0.24905706942081451, "learning_rate": 6.140287247932686e-06, "loss": 0.0009, "step": 19207 }, { "epoch": 2.0823937554206418, "grad_norm": 2.066272020339966, "learning_rate": 6.139561874365298e-06, "loss": 0.0135, "step": 19208 }, { "epoch": 2.0825021682567217, "grad_norm": 0.07696746289730072, "learning_rate": 6.138836500797911e-06, "loss": 0.0004, "step": 19209 }, { "epoch": 2.0826105810928013, "grad_norm": 0.7327773571014404, "learning_rate": 6.138111127230524e-06, "loss": 0.0022, "step": 19210 }, { "epoch": 2.0827189939288813, "grad_norm": 0.5477282404899597, "learning_rate": 6.137385753663136e-06, "loss": 0.0123, "step": 19211 }, { "epoch": 2.082827406764961, "grad_norm": 0.2318548858165741, "learning_rate": 6.13666038009575e-06, "loss": 0.0047, "step": 19212 }, { "epoch": 2.082935819601041, "grad_norm": 0.18036307394504547, "learning_rate": 6.135935006528363e-06, "loss": 0.0009, "step": 19213 }, { "epoch": 2.0830442324371203, "grad_norm": 3.136274814605713, "learning_rate": 6.135209632960975e-06, "loss": 0.0267, "step": 19214 }, { "epoch": 2.0831526452732003, "grad_norm": 0.027129068970680237, "learning_rate": 6.134484259393588e-06, "loss": 0.0002, "step": 19215 }, { "epoch": 2.0832610581092803, "grad_norm": 1.5360521078109741, "learning_rate": 6.133758885826201e-06, "loss": 0.0039, "step": 19216 }, { "epoch": 2.08336947094536, "grad_norm": 0.01700753904879093, "learning_rate": 6.1330335122588135e-06, "loss": 0.0001, "step": 19217 }, { "epoch": 2.08347788378144, "grad_norm": 1.1032973527908325, "learning_rate": 6.1323081386914264e-06, "loss": 0.0427, "step": 19218 }, { "epoch": 2.0835862966175194, "grad_norm": 0.2914431095123291, "learning_rate": 6.131582765124039e-06, "loss": 0.0066, "step": 19219 }, { "epoch": 2.0836947094535994, "grad_norm": 0.40223485231399536, "learning_rate": 6.1308573915566524e-06, "loss": 0.0021, "step": 19220 }, { "epoch": 2.083803122289679, "grad_norm": 0.961129903793335, "learning_rate": 6.130132017989265e-06, "loss": 0.0021, "step": 19221 }, { "epoch": 2.083911535125759, "grad_norm": 2.4540374279022217, "learning_rate": 6.129406644421878e-06, "loss": 0.0692, "step": 19222 }, { "epoch": 2.084019947961839, "grad_norm": 0.45759522914886475, "learning_rate": 6.1286812708544905e-06, "loss": 0.0066, "step": 19223 }, { "epoch": 2.0841283607979184, "grad_norm": 9.537064552307129, "learning_rate": 6.1279558972871035e-06, "loss": 0.0268, "step": 19224 }, { "epoch": 2.0842367736339984, "grad_norm": 0.43061700463294983, "learning_rate": 6.127230523719716e-06, "loss": 0.0221, "step": 19225 }, { "epoch": 2.084345186470078, "grad_norm": 0.4729282259941101, "learning_rate": 6.126505150152329e-06, "loss": 0.0052, "step": 19226 }, { "epoch": 2.084453599306158, "grad_norm": 3.63737416267395, "learning_rate": 6.125779776584942e-06, "loss": 0.0134, "step": 19227 }, { "epoch": 2.0845620121422375, "grad_norm": 1.1324113607406616, "learning_rate": 6.125054403017554e-06, "loss": 0.0146, "step": 19228 }, { "epoch": 2.0846704249783174, "grad_norm": 0.264049232006073, "learning_rate": 6.124329029450168e-06, "loss": 0.0049, "step": 19229 }, { "epoch": 2.0847788378143974, "grad_norm": 0.3098289966583252, "learning_rate": 6.123603655882781e-06, "loss": 0.01, "step": 19230 }, { "epoch": 2.084887250650477, "grad_norm": 0.02471640333533287, "learning_rate": 6.122878282315393e-06, "loss": 0.0001, "step": 19231 }, { "epoch": 2.084995663486557, "grad_norm": 0.6064726114273071, "learning_rate": 6.122152908748006e-06, "loss": 0.0201, "step": 19232 }, { "epoch": 2.0851040763226365, "grad_norm": 0.4054226279258728, "learning_rate": 6.121427535180619e-06, "loss": 0.0076, "step": 19233 }, { "epoch": 2.0852124891587165, "grad_norm": 0.31519177556037903, "learning_rate": 6.120702161613231e-06, "loss": 0.0106, "step": 19234 }, { "epoch": 2.085320901994796, "grad_norm": 1.554705023765564, "learning_rate": 6.119976788045844e-06, "loss": 0.0153, "step": 19235 }, { "epoch": 2.085429314830876, "grad_norm": 0.6816611289978027, "learning_rate": 6.119251414478457e-06, "loss": 0.0006, "step": 19236 }, { "epoch": 2.0855377276669556, "grad_norm": 0.4561242461204529, "learning_rate": 6.118526040911069e-06, "loss": 0.0152, "step": 19237 }, { "epoch": 2.0856461405030355, "grad_norm": 0.22545869648456573, "learning_rate": 6.117800667343683e-06, "loss": 0.0045, "step": 19238 }, { "epoch": 2.0857545533391155, "grad_norm": 0.3351520597934723, "learning_rate": 6.117075293776296e-06, "loss": 0.0159, "step": 19239 }, { "epoch": 2.085862966175195, "grad_norm": 0.6767316460609436, "learning_rate": 6.116349920208908e-06, "loss": 0.0023, "step": 19240 }, { "epoch": 2.085971379011275, "grad_norm": 0.0012881498550996184, "learning_rate": 6.115624546641521e-06, "loss": 0.0, "step": 19241 }, { "epoch": 2.0860797918473546, "grad_norm": 1.3513164520263672, "learning_rate": 6.114899173074134e-06, "loss": 0.0098, "step": 19242 }, { "epoch": 2.0861882046834346, "grad_norm": 0.08793219178915024, "learning_rate": 6.114173799506746e-06, "loss": 0.0003, "step": 19243 }, { "epoch": 2.086296617519514, "grad_norm": 3.5035364627838135, "learning_rate": 6.113448425939359e-06, "loss": 0.024, "step": 19244 }, { "epoch": 2.086405030355594, "grad_norm": 0.3696168065071106, "learning_rate": 6.112723052371971e-06, "loss": 0.0134, "step": 19245 }, { "epoch": 2.086513443191674, "grad_norm": 0.32414892315864563, "learning_rate": 6.111997678804585e-06, "loss": 0.0053, "step": 19246 }, { "epoch": 2.0866218560277536, "grad_norm": 1.2274187803268433, "learning_rate": 6.111272305237198e-06, "loss": 0.0698, "step": 19247 }, { "epoch": 2.0867302688638336, "grad_norm": 0.6016245484352112, "learning_rate": 6.110546931669811e-06, "loss": 0.0015, "step": 19248 }, { "epoch": 2.086838681699913, "grad_norm": 0.29110997915267944, "learning_rate": 6.109821558102423e-06, "loss": 0.0064, "step": 19249 }, { "epoch": 2.086947094535993, "grad_norm": 2.9881575107574463, "learning_rate": 6.109096184535036e-06, "loss": 0.0379, "step": 19250 }, { "epoch": 2.0870555073720727, "grad_norm": 1.6351979970932007, "learning_rate": 6.108370810967648e-06, "loss": 0.0444, "step": 19251 }, { "epoch": 2.0871639202081527, "grad_norm": 0.1154986023902893, "learning_rate": 6.107645437400261e-06, "loss": 0.0006, "step": 19252 }, { "epoch": 2.0872723330442327, "grad_norm": 0.029081864282488823, "learning_rate": 6.106920063832874e-06, "loss": 0.0003, "step": 19253 }, { "epoch": 2.087380745880312, "grad_norm": 0.04876978322863579, "learning_rate": 6.1061946902654865e-06, "loss": 0.0001, "step": 19254 }, { "epoch": 2.087489158716392, "grad_norm": 0.022308802232146263, "learning_rate": 6.1054693166981e-06, "loss": 0.0001, "step": 19255 }, { "epoch": 2.0875975715524717, "grad_norm": 3.1020612716674805, "learning_rate": 6.104743943130713e-06, "loss": 0.012, "step": 19256 }, { "epoch": 2.0877059843885517, "grad_norm": 0.05913533642888069, "learning_rate": 6.1040185695633254e-06, "loss": 0.0004, "step": 19257 }, { "epoch": 2.0878143972246312, "grad_norm": 0.7036241292953491, "learning_rate": 6.1032931959959384e-06, "loss": 0.0157, "step": 19258 }, { "epoch": 2.0879228100607112, "grad_norm": 7.372473239898682, "learning_rate": 6.1025678224285514e-06, "loss": 0.0452, "step": 19259 }, { "epoch": 2.0880312228967908, "grad_norm": 1.1090984344482422, "learning_rate": 6.1018424488611636e-06, "loss": 0.0024, "step": 19260 }, { "epoch": 2.0881396357328708, "grad_norm": 0.023200705647468567, "learning_rate": 6.1011170752937766e-06, "loss": 0.0001, "step": 19261 }, { "epoch": 2.0882480485689507, "grad_norm": 7.770406723022461, "learning_rate": 6.1003917017263896e-06, "loss": 0.0063, "step": 19262 }, { "epoch": 2.0883564614050303, "grad_norm": 14.137158393859863, "learning_rate": 6.099666328159002e-06, "loss": 0.0207, "step": 19263 }, { "epoch": 2.0884648742411103, "grad_norm": 0.1019495502114296, "learning_rate": 6.0989409545916155e-06, "loss": 0.0005, "step": 19264 }, { "epoch": 2.08857328707719, "grad_norm": 0.27414974570274353, "learning_rate": 6.0982155810242285e-06, "loss": 0.0063, "step": 19265 }, { "epoch": 2.08868169991327, "grad_norm": 0.7911152243614197, "learning_rate": 6.097490207456841e-06, "loss": 0.0031, "step": 19266 }, { "epoch": 2.0887901127493493, "grad_norm": 2.0538406372070312, "learning_rate": 6.096764833889454e-06, "loss": 0.0338, "step": 19267 }, { "epoch": 2.0888985255854293, "grad_norm": 2.311184883117676, "learning_rate": 6.096039460322067e-06, "loss": 0.0322, "step": 19268 }, { "epoch": 2.0890069384215093, "grad_norm": 3.632319211959839, "learning_rate": 6.095314086754679e-06, "loss": 0.0233, "step": 19269 }, { "epoch": 2.089115351257589, "grad_norm": 3.080040693283081, "learning_rate": 6.094588713187292e-06, "loss": 0.0154, "step": 19270 }, { "epoch": 2.089223764093669, "grad_norm": 0.0161887239664793, "learning_rate": 6.093863339619904e-06, "loss": 0.0001, "step": 19271 }, { "epoch": 2.0893321769297484, "grad_norm": 0.2461652308702469, "learning_rate": 6.093137966052518e-06, "loss": 0.0054, "step": 19272 }, { "epoch": 2.0894405897658284, "grad_norm": 0.3854878842830658, "learning_rate": 6.092412592485131e-06, "loss": 0.0312, "step": 19273 }, { "epoch": 2.089549002601908, "grad_norm": 0.35008326172828674, "learning_rate": 6.091687218917744e-06, "loss": 0.0021, "step": 19274 }, { "epoch": 2.089657415437988, "grad_norm": 2.0284249782562256, "learning_rate": 6.090961845350356e-06, "loss": 0.0095, "step": 19275 }, { "epoch": 2.089765828274068, "grad_norm": 0.021117007359862328, "learning_rate": 6.090236471782969e-06, "loss": 0.0003, "step": 19276 }, { "epoch": 2.0898742411101474, "grad_norm": 1.7480469942092896, "learning_rate": 6.089511098215581e-06, "loss": 0.0076, "step": 19277 }, { "epoch": 2.0899826539462274, "grad_norm": 0.0313163623213768, "learning_rate": 6.088785724648194e-06, "loss": 0.0003, "step": 19278 }, { "epoch": 2.090091066782307, "grad_norm": 0.5124592185020447, "learning_rate": 6.088060351080807e-06, "loss": 0.02, "step": 19279 }, { "epoch": 2.090199479618387, "grad_norm": 3.039372205734253, "learning_rate": 6.087334977513419e-06, "loss": 0.0096, "step": 19280 }, { "epoch": 2.0903078924544665, "grad_norm": 0.4786014258861542, "learning_rate": 6.086609603946033e-06, "loss": 0.0114, "step": 19281 }, { "epoch": 2.0904163052905465, "grad_norm": 4.019374370574951, "learning_rate": 6.085884230378646e-06, "loss": 0.0127, "step": 19282 }, { "epoch": 2.090524718126626, "grad_norm": 6.865487575531006, "learning_rate": 6.085158856811258e-06, "loss": 0.0351, "step": 19283 }, { "epoch": 2.090633130962706, "grad_norm": 0.29752883315086365, "learning_rate": 6.084433483243871e-06, "loss": 0.0114, "step": 19284 }, { "epoch": 2.090741543798786, "grad_norm": 1.088956594467163, "learning_rate": 6.083708109676484e-06, "loss": 0.0007, "step": 19285 }, { "epoch": 2.0908499566348655, "grad_norm": 0.4754190742969513, "learning_rate": 6.082982736109096e-06, "loss": 0.0103, "step": 19286 }, { "epoch": 2.0909583694709455, "grad_norm": 0.5359010100364685, "learning_rate": 6.082257362541709e-06, "loss": 0.017, "step": 19287 }, { "epoch": 2.091066782307025, "grad_norm": 0.29860129952430725, "learning_rate": 6.081531988974322e-06, "loss": 0.0118, "step": 19288 }, { "epoch": 2.091175195143105, "grad_norm": 0.255855917930603, "learning_rate": 6.080806615406935e-06, "loss": 0.0165, "step": 19289 }, { "epoch": 2.0912836079791846, "grad_norm": 0.0962829515337944, "learning_rate": 6.080081241839548e-06, "loss": 0.0102, "step": 19290 }, { "epoch": 2.0913920208152645, "grad_norm": 0.0063149528577923775, "learning_rate": 6.079355868272161e-06, "loss": 0.0, "step": 19291 }, { "epoch": 2.0915004336513445, "grad_norm": 0.17801283299922943, "learning_rate": 6.078630494704773e-06, "loss": 0.0002, "step": 19292 }, { "epoch": 2.091608846487424, "grad_norm": 0.03575189784169197, "learning_rate": 6.077905121137386e-06, "loss": 0.0003, "step": 19293 }, { "epoch": 2.091717259323504, "grad_norm": 0.461883008480072, "learning_rate": 6.077179747569999e-06, "loss": 0.007, "step": 19294 }, { "epoch": 2.0918256721595836, "grad_norm": 0.2330690175294876, "learning_rate": 6.0764543740026115e-06, "loss": 0.0009, "step": 19295 }, { "epoch": 2.0919340849956636, "grad_norm": 0.3273262083530426, "learning_rate": 6.0757290004352245e-06, "loss": 0.0119, "step": 19296 }, { "epoch": 2.092042497831743, "grad_norm": 0.49603939056396484, "learning_rate": 6.075003626867837e-06, "loss": 0.0131, "step": 19297 }, { "epoch": 2.092150910667823, "grad_norm": 3.604207992553711, "learning_rate": 6.0742782533004504e-06, "loss": 0.0639, "step": 19298 }, { "epoch": 2.092259323503903, "grad_norm": 0.21194729208946228, "learning_rate": 6.073552879733063e-06, "loss": 0.0014, "step": 19299 }, { "epoch": 2.0923677363399826, "grad_norm": 0.5493357181549072, "learning_rate": 6.072827506165676e-06, "loss": 0.0245, "step": 19300 }, { "epoch": 2.0924761491760626, "grad_norm": 0.1261715143918991, "learning_rate": 6.0721021325982886e-06, "loss": 0.0007, "step": 19301 }, { "epoch": 2.092584562012142, "grad_norm": 0.2846578061580658, "learning_rate": 6.0713767590309015e-06, "loss": 0.0079, "step": 19302 }, { "epoch": 2.092692974848222, "grad_norm": 1.0745586156845093, "learning_rate": 6.070651385463514e-06, "loss": 0.0247, "step": 19303 }, { "epoch": 2.0928013876843017, "grad_norm": 4.29286527633667, "learning_rate": 6.069926011896127e-06, "loss": 0.0414, "step": 19304 }, { "epoch": 2.0929098005203817, "grad_norm": 4.652688980102539, "learning_rate": 6.06920063832874e-06, "loss": 0.0112, "step": 19305 }, { "epoch": 2.093018213356461, "grad_norm": 19.27511215209961, "learning_rate": 6.068475264761352e-06, "loss": 0.0236, "step": 19306 }, { "epoch": 2.093126626192541, "grad_norm": 0.3231011927127838, "learning_rate": 6.067749891193966e-06, "loss": 0.0332, "step": 19307 }, { "epoch": 2.093235039028621, "grad_norm": 2.3728771209716797, "learning_rate": 6.067024517626579e-06, "loss": 0.0273, "step": 19308 }, { "epoch": 2.0933434518647007, "grad_norm": 0.3920026123523712, "learning_rate": 6.066299144059191e-06, "loss": 0.0072, "step": 19309 }, { "epoch": 2.0934518647007807, "grad_norm": 0.8725764155387878, "learning_rate": 6.065573770491804e-06, "loss": 0.001, "step": 19310 }, { "epoch": 2.0935602775368602, "grad_norm": 0.37594524025917053, "learning_rate": 6.064848396924417e-06, "loss": 0.012, "step": 19311 }, { "epoch": 2.0936686903729402, "grad_norm": 0.5820456147193909, "learning_rate": 6.064123023357029e-06, "loss": 0.0279, "step": 19312 }, { "epoch": 2.0937771032090198, "grad_norm": 0.3878690004348755, "learning_rate": 6.063397649789642e-06, "loss": 0.0016, "step": 19313 }, { "epoch": 2.0938855160450998, "grad_norm": 4.1398091316223145, "learning_rate": 6.062672276222255e-06, "loss": 0.0617, "step": 19314 }, { "epoch": 2.0939939288811797, "grad_norm": 3.188662528991699, "learning_rate": 6.061946902654868e-06, "loss": 0.0092, "step": 19315 }, { "epoch": 2.0941023417172593, "grad_norm": 5.080784797668457, "learning_rate": 6.061221529087481e-06, "loss": 0.0349, "step": 19316 }, { "epoch": 2.0942107545533393, "grad_norm": 0.0457816906273365, "learning_rate": 6.060496155520094e-06, "loss": 0.0003, "step": 19317 }, { "epoch": 2.094319167389419, "grad_norm": 0.28856563568115234, "learning_rate": 6.059770781952706e-06, "loss": 0.0058, "step": 19318 }, { "epoch": 2.094427580225499, "grad_norm": 2.324286460876465, "learning_rate": 6.059045408385319e-06, "loss": 0.0256, "step": 19319 }, { "epoch": 2.0945359930615783, "grad_norm": 1.5461101531982422, "learning_rate": 6.058320034817932e-06, "loss": 0.035, "step": 19320 }, { "epoch": 2.0946444058976583, "grad_norm": 0.4462130665779114, "learning_rate": 6.057594661250544e-06, "loss": 0.0103, "step": 19321 }, { "epoch": 2.094752818733738, "grad_norm": 0.24499982595443726, "learning_rate": 6.056869287683157e-06, "loss": 0.0054, "step": 19322 }, { "epoch": 2.094861231569818, "grad_norm": 0.34155911207199097, "learning_rate": 6.056143914115769e-06, "loss": 0.0118, "step": 19323 }, { "epoch": 2.094969644405898, "grad_norm": 0.5559288859367371, "learning_rate": 6.055418540548383e-06, "loss": 0.0012, "step": 19324 }, { "epoch": 2.0950780572419774, "grad_norm": 2.139242172241211, "learning_rate": 6.054693166980996e-06, "loss": 0.0372, "step": 19325 }, { "epoch": 2.0951864700780574, "grad_norm": 4.225952625274658, "learning_rate": 6.053967793413609e-06, "loss": 0.0368, "step": 19326 }, { "epoch": 2.095294882914137, "grad_norm": 1.100217580795288, "learning_rate": 6.053242419846221e-06, "loss": 0.0026, "step": 19327 }, { "epoch": 2.095403295750217, "grad_norm": 0.3694135546684265, "learning_rate": 6.052517046278834e-06, "loss": 0.0128, "step": 19328 }, { "epoch": 2.0955117085862964, "grad_norm": 0.032025281339883804, "learning_rate": 6.051791672711446e-06, "loss": 0.0004, "step": 19329 }, { "epoch": 2.0956201214223764, "grad_norm": 0.19920766353607178, "learning_rate": 6.051066299144059e-06, "loss": 0.001, "step": 19330 }, { "epoch": 2.0957285342584564, "grad_norm": 1.3394631147384644, "learning_rate": 6.050340925576672e-06, "loss": 0.043, "step": 19331 }, { "epoch": 2.095836947094536, "grad_norm": 2.2979867458343506, "learning_rate": 6.0496155520092845e-06, "loss": 0.006, "step": 19332 }, { "epoch": 2.095945359930616, "grad_norm": 0.3179014325141907, "learning_rate": 6.048890178441898e-06, "loss": 0.0112, "step": 19333 }, { "epoch": 2.0960537727666955, "grad_norm": 0.3193737268447876, "learning_rate": 6.048164804874511e-06, "loss": 0.0055, "step": 19334 }, { "epoch": 2.0961621856027755, "grad_norm": 0.08558054268360138, "learning_rate": 6.0474394313071235e-06, "loss": 0.0006, "step": 19335 }, { "epoch": 2.096270598438855, "grad_norm": 1.8283506631851196, "learning_rate": 6.0467140577397364e-06, "loss": 0.0194, "step": 19336 }, { "epoch": 2.096379011274935, "grad_norm": 0.22281233966350555, "learning_rate": 6.0459886841723494e-06, "loss": 0.0046, "step": 19337 }, { "epoch": 2.0964874241110145, "grad_norm": 0.629075288772583, "learning_rate": 6.045263310604962e-06, "loss": 0.0213, "step": 19338 }, { "epoch": 2.0965958369470945, "grad_norm": 0.050010308623313904, "learning_rate": 6.0445379370375746e-06, "loss": 0.0003, "step": 19339 }, { "epoch": 2.0967042497831745, "grad_norm": 2.517076253890991, "learning_rate": 6.0438125634701876e-06, "loss": 0.0188, "step": 19340 }, { "epoch": 2.096812662619254, "grad_norm": 0.06596328318119049, "learning_rate": 6.0430871899028005e-06, "loss": 0.0003, "step": 19341 }, { "epoch": 2.096921075455334, "grad_norm": 0.07223314791917801, "learning_rate": 6.0423618163354135e-06, "loss": 0.0004, "step": 19342 }, { "epoch": 2.0970294882914136, "grad_norm": 0.3213094472885132, "learning_rate": 6.0416364427680265e-06, "loss": 0.0075, "step": 19343 }, { "epoch": 2.0971379011274935, "grad_norm": 0.2415613979101181, "learning_rate": 6.040911069200639e-06, "loss": 0.0052, "step": 19344 }, { "epoch": 2.097246313963573, "grad_norm": 2.5762276649475098, "learning_rate": 6.040185695633252e-06, "loss": 0.039, "step": 19345 }, { "epoch": 2.097354726799653, "grad_norm": 1.8004623651504517, "learning_rate": 6.039460322065865e-06, "loss": 0.0088, "step": 19346 }, { "epoch": 2.097463139635733, "grad_norm": 0.2611943185329437, "learning_rate": 6.038734948498477e-06, "loss": 0.0007, "step": 19347 }, { "epoch": 2.0975715524718126, "grad_norm": 0.06330502778291702, "learning_rate": 6.03800957493109e-06, "loss": 0.0006, "step": 19348 }, { "epoch": 2.0976799653078926, "grad_norm": 5.538384437561035, "learning_rate": 6.037284201363702e-06, "loss": 0.0272, "step": 19349 }, { "epoch": 2.097788378143972, "grad_norm": 0.4485527276992798, "learning_rate": 6.036558827796316e-06, "loss": 0.0266, "step": 19350 }, { "epoch": 2.097896790980052, "grad_norm": 0.28724005818367004, "learning_rate": 6.035833454228929e-06, "loss": 0.0064, "step": 19351 }, { "epoch": 2.0980052038161316, "grad_norm": 0.39738425612449646, "learning_rate": 6.035108080661542e-06, "loss": 0.0006, "step": 19352 }, { "epoch": 2.0981136166522116, "grad_norm": 0.3463897109031677, "learning_rate": 6.034382707094154e-06, "loss": 0.0069, "step": 19353 }, { "epoch": 2.0982220294882916, "grad_norm": 0.015138044022023678, "learning_rate": 6.033657333526767e-06, "loss": 0.0, "step": 19354 }, { "epoch": 2.098330442324371, "grad_norm": 0.3882938325405121, "learning_rate": 6.032931959959379e-06, "loss": 0.0067, "step": 19355 }, { "epoch": 2.098438855160451, "grad_norm": 0.31840017437934875, "learning_rate": 6.032206586391992e-06, "loss": 0.0101, "step": 19356 }, { "epoch": 2.0985472679965307, "grad_norm": 0.35187214612960815, "learning_rate": 6.031481212824605e-06, "loss": 0.0131, "step": 19357 }, { "epoch": 2.0986556808326107, "grad_norm": 0.42120659351348877, "learning_rate": 6.030755839257217e-06, "loss": 0.0148, "step": 19358 }, { "epoch": 2.09876409366869, "grad_norm": 1.6885069608688354, "learning_rate": 6.030030465689831e-06, "loss": 0.0369, "step": 19359 }, { "epoch": 2.09887250650477, "grad_norm": 0.20062653720378876, "learning_rate": 6.029305092122444e-06, "loss": 0.0007, "step": 19360 }, { "epoch": 2.0989809193408497, "grad_norm": 0.06739925593137741, "learning_rate": 6.028579718555056e-06, "loss": 0.0005, "step": 19361 }, { "epoch": 2.0990893321769297, "grad_norm": 0.3427749276161194, "learning_rate": 6.027854344987669e-06, "loss": 0.0217, "step": 19362 }, { "epoch": 2.0991977450130097, "grad_norm": 0.6842244267463684, "learning_rate": 6.027128971420282e-06, "loss": 0.0248, "step": 19363 }, { "epoch": 2.0993061578490892, "grad_norm": 5.9584269523620605, "learning_rate": 6.026403597852894e-06, "loss": 0.0421, "step": 19364 }, { "epoch": 2.0994145706851692, "grad_norm": 2.744105100631714, "learning_rate": 6.025678224285507e-06, "loss": 0.0209, "step": 19365 }, { "epoch": 2.0995229835212488, "grad_norm": 1.7212505340576172, "learning_rate": 6.02495285071812e-06, "loss": 0.0301, "step": 19366 }, { "epoch": 2.0996313963573288, "grad_norm": 4.436058521270752, "learning_rate": 6.024227477150733e-06, "loss": 0.0151, "step": 19367 }, { "epoch": 2.0997398091934083, "grad_norm": 0.004114754963666201, "learning_rate": 6.023502103583346e-06, "loss": 0.0, "step": 19368 }, { "epoch": 2.0998482220294883, "grad_norm": 0.30973580479621887, "learning_rate": 6.022776730015959e-06, "loss": 0.0069, "step": 19369 }, { "epoch": 2.0999566348655683, "grad_norm": 0.023892726749181747, "learning_rate": 6.022051356448571e-06, "loss": 0.0001, "step": 19370 }, { "epoch": 2.100065047701648, "grad_norm": 0.27881118655204773, "learning_rate": 6.021325982881184e-06, "loss": 0.0052, "step": 19371 }, { "epoch": 2.100173460537728, "grad_norm": 0.28764981031417847, "learning_rate": 6.020600609313797e-06, "loss": 0.0086, "step": 19372 }, { "epoch": 2.1002818733738073, "grad_norm": 0.6640120148658752, "learning_rate": 6.0198752357464095e-06, "loss": 0.0198, "step": 19373 }, { "epoch": 2.1003902862098873, "grad_norm": 3.197436571121216, "learning_rate": 6.0191498621790225e-06, "loss": 0.0383, "step": 19374 }, { "epoch": 2.100498699045967, "grad_norm": 4.134912967681885, "learning_rate": 6.018424488611635e-06, "loss": 0.0195, "step": 19375 }, { "epoch": 2.100607111882047, "grad_norm": 0.9004355072975159, "learning_rate": 6.0176991150442484e-06, "loss": 0.0129, "step": 19376 }, { "epoch": 2.100715524718127, "grad_norm": 1.437552571296692, "learning_rate": 6.0169737414768614e-06, "loss": 0.0086, "step": 19377 }, { "epoch": 2.1008239375542064, "grad_norm": 0.24907219409942627, "learning_rate": 6.016248367909474e-06, "loss": 0.0175, "step": 19378 }, { "epoch": 2.1009323503902864, "grad_norm": 0.04491749405860901, "learning_rate": 6.0155229943420866e-06, "loss": 0.0005, "step": 19379 }, { "epoch": 2.101040763226366, "grad_norm": 14.096508979797363, "learning_rate": 6.0147976207746996e-06, "loss": 0.0329, "step": 19380 }, { "epoch": 2.101149176062446, "grad_norm": 7.261989593505859, "learning_rate": 6.014072247207312e-06, "loss": 0.029, "step": 19381 }, { "epoch": 2.1012575888985254, "grad_norm": 0.2870766222476959, "learning_rate": 6.013346873639925e-06, "loss": 0.0007, "step": 19382 }, { "epoch": 2.1013660017346054, "grad_norm": 0.08229617029428482, "learning_rate": 6.012621500072538e-06, "loss": 0.0005, "step": 19383 }, { "epoch": 2.101474414570685, "grad_norm": 0.2459239959716797, "learning_rate": 6.0118961265051515e-06, "loss": 0.0011, "step": 19384 }, { "epoch": 2.101582827406765, "grad_norm": 0.456592857837677, "learning_rate": 6.011170752937764e-06, "loss": 0.0147, "step": 19385 }, { "epoch": 2.101691240242845, "grad_norm": 0.6221083998680115, "learning_rate": 6.010445379370377e-06, "loss": 0.0166, "step": 19386 }, { "epoch": 2.1017996530789245, "grad_norm": 0.4532686769962311, "learning_rate": 6.009720005802989e-06, "loss": 0.0212, "step": 19387 }, { "epoch": 2.1019080659150045, "grad_norm": 2.8571970462799072, "learning_rate": 6.008994632235602e-06, "loss": 0.0041, "step": 19388 }, { "epoch": 2.102016478751084, "grad_norm": 1.348012089729309, "learning_rate": 6.008269258668215e-06, "loss": 0.0155, "step": 19389 }, { "epoch": 2.102124891587164, "grad_norm": 1.462746024131775, "learning_rate": 6.007543885100827e-06, "loss": 0.0352, "step": 19390 }, { "epoch": 2.1022333044232435, "grad_norm": 0.07674440741539001, "learning_rate": 6.00681851153344e-06, "loss": 0.0007, "step": 19391 }, { "epoch": 2.1023417172593235, "grad_norm": 10.759973526000977, "learning_rate": 6.006093137966053e-06, "loss": 0.0145, "step": 19392 }, { "epoch": 2.1024501300954035, "grad_norm": 0.3066852390766144, "learning_rate": 6.005367764398667e-06, "loss": 0.0121, "step": 19393 }, { "epoch": 2.102558542931483, "grad_norm": 1.2752525806427002, "learning_rate": 6.004642390831279e-06, "loss": 0.0083, "step": 19394 }, { "epoch": 2.102666955767563, "grad_norm": 0.32005438208580017, "learning_rate": 6.003917017263892e-06, "loss": 0.0013, "step": 19395 }, { "epoch": 2.1027753686036426, "grad_norm": 5.579611301422119, "learning_rate": 6.003191643696504e-06, "loss": 0.0282, "step": 19396 }, { "epoch": 2.1028837814397225, "grad_norm": 0.02405666932463646, "learning_rate": 6.002466270129117e-06, "loss": 0.0002, "step": 19397 }, { "epoch": 2.102992194275802, "grad_norm": 5.942249298095703, "learning_rate": 6.00174089656173e-06, "loss": 0.0392, "step": 19398 }, { "epoch": 2.103100607111882, "grad_norm": 4.608067512512207, "learning_rate": 6.001015522994342e-06, "loss": 0.0149, "step": 19399 }, { "epoch": 2.103209019947962, "grad_norm": 0.4255863130092621, "learning_rate": 6.000290149426955e-06, "loss": 0.0227, "step": 19400 }, { "epoch": 2.1033174327840416, "grad_norm": 0.42798981070518494, "learning_rate": 5.999564775859567e-06, "loss": 0.0008, "step": 19401 }, { "epoch": 2.1034258456201216, "grad_norm": 7.486379623413086, "learning_rate": 5.998839402292181e-06, "loss": 0.036, "step": 19402 }, { "epoch": 2.103534258456201, "grad_norm": 0.32217898964881897, "learning_rate": 5.998114028724794e-06, "loss": 0.01, "step": 19403 }, { "epoch": 2.103642671292281, "grad_norm": 1.1324424743652344, "learning_rate": 5.997388655157407e-06, "loss": 0.0042, "step": 19404 }, { "epoch": 2.1037510841283606, "grad_norm": 0.15309986472129822, "learning_rate": 5.996663281590019e-06, "loss": 0.0006, "step": 19405 }, { "epoch": 2.1038594969644406, "grad_norm": 0.13124676048755646, "learning_rate": 5.995937908022632e-06, "loss": 0.0005, "step": 19406 }, { "epoch": 2.10396790980052, "grad_norm": 8.426505088806152, "learning_rate": 5.995212534455244e-06, "loss": 0.058, "step": 19407 }, { "epoch": 2.1040763226366, "grad_norm": 0.22435764968395233, "learning_rate": 5.994487160887857e-06, "loss": 0.0019, "step": 19408 }, { "epoch": 2.10418473547268, "grad_norm": 0.2565912902355194, "learning_rate": 5.99376178732047e-06, "loss": 0.0141, "step": 19409 }, { "epoch": 2.1042931483087597, "grad_norm": 17.479814529418945, "learning_rate": 5.993036413753084e-06, "loss": 0.0456, "step": 19410 }, { "epoch": 2.1044015611448397, "grad_norm": 1.075408697128296, "learning_rate": 5.992311040185696e-06, "loss": 0.0161, "step": 19411 }, { "epoch": 2.104509973980919, "grad_norm": 3.4236323833465576, "learning_rate": 5.991585666618309e-06, "loss": 0.019, "step": 19412 }, { "epoch": 2.104618386816999, "grad_norm": 0.046000100672245026, "learning_rate": 5.9908602930509215e-06, "loss": 0.0002, "step": 19413 }, { "epoch": 2.1047267996530787, "grad_norm": 9.887569427490234, "learning_rate": 5.9901349194835345e-06, "loss": 0.0134, "step": 19414 }, { "epoch": 2.1048352124891587, "grad_norm": 0.026563361287117004, "learning_rate": 5.9894095459161474e-06, "loss": 0.0002, "step": 19415 }, { "epoch": 2.1049436253252387, "grad_norm": 0.3226328194141388, "learning_rate": 5.98868417234876e-06, "loss": 0.0077, "step": 19416 }, { "epoch": 2.1050520381613183, "grad_norm": 3.748216152191162, "learning_rate": 5.9879587987813726e-06, "loss": 0.0276, "step": 19417 }, { "epoch": 2.1051604509973982, "grad_norm": 0.32627740502357483, "learning_rate": 5.9872334252139856e-06, "loss": 0.0125, "step": 19418 }, { "epoch": 2.1052688638334778, "grad_norm": 0.120844766497612, "learning_rate": 5.986508051646599e-06, "loss": 0.0005, "step": 19419 }, { "epoch": 2.1053772766695578, "grad_norm": 0.9456669688224792, "learning_rate": 5.9857826780792115e-06, "loss": 0.0115, "step": 19420 }, { "epoch": 2.1054856895056373, "grad_norm": 0.21048663556575775, "learning_rate": 5.9850573045118245e-06, "loss": 0.0045, "step": 19421 }, { "epoch": 2.1055941023417173, "grad_norm": 1.2714091539382935, "learning_rate": 5.984331930944437e-06, "loss": 0.0081, "step": 19422 }, { "epoch": 2.1057025151777973, "grad_norm": 1.4467201232910156, "learning_rate": 5.98360655737705e-06, "loss": 0.0249, "step": 19423 }, { "epoch": 2.105810928013877, "grad_norm": 0.8890173435211182, "learning_rate": 5.982881183809663e-06, "loss": 0.0183, "step": 19424 }, { "epoch": 2.105919340849957, "grad_norm": 0.7552171945571899, "learning_rate": 5.982155810242275e-06, "loss": 0.027, "step": 19425 }, { "epoch": 2.1060277536860363, "grad_norm": 0.027398347854614258, "learning_rate": 5.981430436674888e-06, "loss": 0.0003, "step": 19426 }, { "epoch": 2.1061361665221163, "grad_norm": 0.010623826645314693, "learning_rate": 5.9807050631075e-06, "loss": 0.0001, "step": 19427 }, { "epoch": 2.106244579358196, "grad_norm": 0.1833367645740509, "learning_rate": 5.979979689540114e-06, "loss": 0.0007, "step": 19428 }, { "epoch": 2.106352992194276, "grad_norm": 0.2319502979516983, "learning_rate": 5.979254315972727e-06, "loss": 0.0046, "step": 19429 }, { "epoch": 2.1064614050303554, "grad_norm": 0.11678516119718552, "learning_rate": 5.97852894240534e-06, "loss": 0.0004, "step": 19430 }, { "epoch": 2.1065698178664354, "grad_norm": 0.9951105117797852, "learning_rate": 5.977803568837952e-06, "loss": 0.0028, "step": 19431 }, { "epoch": 2.1066782307025154, "grad_norm": 4.026498794555664, "learning_rate": 5.977078195270565e-06, "loss": 0.025, "step": 19432 }, { "epoch": 2.106786643538595, "grad_norm": 0.55599445104599, "learning_rate": 5.976352821703177e-06, "loss": 0.0019, "step": 19433 }, { "epoch": 2.106895056374675, "grad_norm": 2.0856950283050537, "learning_rate": 5.97562744813579e-06, "loss": 0.0517, "step": 19434 }, { "epoch": 2.1070034692107544, "grad_norm": 0.3046685457229614, "learning_rate": 5.974902074568403e-06, "loss": 0.01, "step": 19435 }, { "epoch": 2.1071118820468344, "grad_norm": 6.856200218200684, "learning_rate": 5.974176701001017e-06, "loss": 0.0254, "step": 19436 }, { "epoch": 2.107220294882914, "grad_norm": 3.1136324405670166, "learning_rate": 5.973451327433629e-06, "loss": 0.0303, "step": 19437 }, { "epoch": 2.107328707718994, "grad_norm": 0.2964106798171997, "learning_rate": 5.972725953866242e-06, "loss": 0.0117, "step": 19438 }, { "epoch": 2.107437120555074, "grad_norm": 0.014453694224357605, "learning_rate": 5.972000580298854e-06, "loss": 0.0001, "step": 19439 }, { "epoch": 2.1075455333911535, "grad_norm": 0.1324848085641861, "learning_rate": 5.971275206731467e-06, "loss": 0.0004, "step": 19440 }, { "epoch": 2.1076539462272335, "grad_norm": 5.915916442871094, "learning_rate": 5.97054983316408e-06, "loss": 0.0318, "step": 19441 }, { "epoch": 2.107762359063313, "grad_norm": 0.3385451138019562, "learning_rate": 5.969824459596692e-06, "loss": 0.0195, "step": 19442 }, { "epoch": 2.107870771899393, "grad_norm": 0.8501534461975098, "learning_rate": 5.969099086029305e-06, "loss": 0.0204, "step": 19443 }, { "epoch": 2.1079791847354725, "grad_norm": 0.6028839945793152, "learning_rate": 5.968373712461918e-06, "loss": 0.0017, "step": 19444 }, { "epoch": 2.1080875975715525, "grad_norm": 1.4782127141952515, "learning_rate": 5.967648338894532e-06, "loss": 0.0025, "step": 19445 }, { "epoch": 2.1081960104076325, "grad_norm": 12.661494255065918, "learning_rate": 5.966922965327144e-06, "loss": 0.0197, "step": 19446 }, { "epoch": 2.108304423243712, "grad_norm": 0.02666224166750908, "learning_rate": 5.966197591759757e-06, "loss": 0.0002, "step": 19447 }, { "epoch": 2.108412836079792, "grad_norm": 0.004180779680609703, "learning_rate": 5.965472218192369e-06, "loss": 0.0, "step": 19448 }, { "epoch": 2.1085212489158716, "grad_norm": 1.3565384149551392, "learning_rate": 5.964746844624982e-06, "loss": 0.0054, "step": 19449 }, { "epoch": 2.1086296617519515, "grad_norm": 4.055633544921875, "learning_rate": 5.964021471057595e-06, "loss": 0.0359, "step": 19450 }, { "epoch": 2.108738074588031, "grad_norm": 8.15384578704834, "learning_rate": 5.9632960974902075e-06, "loss": 0.0275, "step": 19451 }, { "epoch": 2.108846487424111, "grad_norm": 2.2293288707733154, "learning_rate": 5.9625707239228205e-06, "loss": 0.0348, "step": 19452 }, { "epoch": 2.1089549002601906, "grad_norm": 15.1298828125, "learning_rate": 5.961845350355433e-06, "loss": 0.0078, "step": 19453 }, { "epoch": 2.1090633130962706, "grad_norm": 1.665777325630188, "learning_rate": 5.9611199767880464e-06, "loss": 0.0033, "step": 19454 }, { "epoch": 2.1091717259323506, "grad_norm": 0.3067239224910736, "learning_rate": 5.9603946032206594e-06, "loss": 0.0008, "step": 19455 }, { "epoch": 2.10928013876843, "grad_norm": 0.13322222232818604, "learning_rate": 5.9596692296532724e-06, "loss": 0.0004, "step": 19456 }, { "epoch": 2.10938855160451, "grad_norm": 3.209838390350342, "learning_rate": 5.9589438560858846e-06, "loss": 0.0319, "step": 19457 }, { "epoch": 2.1094969644405897, "grad_norm": 0.3555794954299927, "learning_rate": 5.9582184825184976e-06, "loss": 0.0124, "step": 19458 }, { "epoch": 2.1096053772766696, "grad_norm": 0.32679614424705505, "learning_rate": 5.95749310895111e-06, "loss": 0.0287, "step": 19459 }, { "epoch": 2.109713790112749, "grad_norm": 0.08759576082229614, "learning_rate": 5.956767735383723e-06, "loss": 0.0006, "step": 19460 }, { "epoch": 2.109822202948829, "grad_norm": 0.27159592509269714, "learning_rate": 5.956042361816336e-06, "loss": 0.0063, "step": 19461 }, { "epoch": 2.109930615784909, "grad_norm": 0.03580167144536972, "learning_rate": 5.9553169882489495e-06, "loss": 0.0002, "step": 19462 }, { "epoch": 2.1100390286209887, "grad_norm": 3.946169137954712, "learning_rate": 5.954591614681562e-06, "loss": 0.0231, "step": 19463 }, { "epoch": 2.1101474414570687, "grad_norm": 0.7460979223251343, "learning_rate": 5.953866241114175e-06, "loss": 0.0065, "step": 19464 }, { "epoch": 2.110255854293148, "grad_norm": 3.2427568435668945, "learning_rate": 5.953140867546787e-06, "loss": 0.0187, "step": 19465 }, { "epoch": 2.110364267129228, "grad_norm": 0.5162734985351562, "learning_rate": 5.9524154939794e-06, "loss": 0.0147, "step": 19466 }, { "epoch": 2.1104726799653077, "grad_norm": 3.9389185905456543, "learning_rate": 5.951690120412013e-06, "loss": 0.041, "step": 19467 }, { "epoch": 2.1105810928013877, "grad_norm": 3.449711799621582, "learning_rate": 5.950964746844625e-06, "loss": 0.0242, "step": 19468 }, { "epoch": 2.1106895056374673, "grad_norm": 0.5131258368492126, "learning_rate": 5.950239373277238e-06, "loss": 0.0134, "step": 19469 }, { "epoch": 2.1107979184735473, "grad_norm": 0.051329582929611206, "learning_rate": 5.949513999709851e-06, "loss": 0.0002, "step": 19470 }, { "epoch": 2.1109063313096272, "grad_norm": 0.2868936359882355, "learning_rate": 5.948788626142465e-06, "loss": 0.0055, "step": 19471 }, { "epoch": 2.111014744145707, "grad_norm": 0.04558253660798073, "learning_rate": 5.948063252575077e-06, "loss": 0.0003, "step": 19472 }, { "epoch": 2.1111231569817868, "grad_norm": 0.02952902764081955, "learning_rate": 5.94733787900769e-06, "loss": 0.0002, "step": 19473 }, { "epoch": 2.1112315698178663, "grad_norm": 0.2120039165019989, "learning_rate": 5.946612505440302e-06, "loss": 0.0003, "step": 19474 }, { "epoch": 2.1113399826539463, "grad_norm": 0.018556030467152596, "learning_rate": 5.945887131872915e-06, "loss": 0.0001, "step": 19475 }, { "epoch": 2.111448395490026, "grad_norm": 1.555545687675476, "learning_rate": 5.945161758305528e-06, "loss": 0.0221, "step": 19476 }, { "epoch": 2.111556808326106, "grad_norm": 0.36923912167549133, "learning_rate": 5.94443638473814e-06, "loss": 0.0064, "step": 19477 }, { "epoch": 2.111665221162186, "grad_norm": 0.1124117523431778, "learning_rate": 5.943711011170753e-06, "loss": 0.0003, "step": 19478 }, { "epoch": 2.1117736339982653, "grad_norm": 18.06307029724121, "learning_rate": 5.942985637603365e-06, "loss": 0.0225, "step": 19479 }, { "epoch": 2.1118820468343453, "grad_norm": 7.998801231384277, "learning_rate": 5.942260264035979e-06, "loss": 0.005, "step": 19480 }, { "epoch": 2.111990459670425, "grad_norm": 2.6758852005004883, "learning_rate": 5.941534890468592e-06, "loss": 0.013, "step": 19481 }, { "epoch": 2.112098872506505, "grad_norm": 0.2945537269115448, "learning_rate": 5.940809516901205e-06, "loss": 0.0045, "step": 19482 }, { "epoch": 2.1122072853425844, "grad_norm": 0.3812185823917389, "learning_rate": 5.940084143333817e-06, "loss": 0.0158, "step": 19483 }, { "epoch": 2.1123156981786644, "grad_norm": 0.754750669002533, "learning_rate": 5.93935876976643e-06, "loss": 0.0011, "step": 19484 }, { "epoch": 2.112424111014744, "grad_norm": 0.042343419045209885, "learning_rate": 5.938633396199042e-06, "loss": 0.0001, "step": 19485 }, { "epoch": 2.112532523850824, "grad_norm": 0.04169193282723427, "learning_rate": 5.937908022631655e-06, "loss": 0.0002, "step": 19486 }, { "epoch": 2.112640936686904, "grad_norm": 0.4673953950405121, "learning_rate": 5.937182649064268e-06, "loss": 0.017, "step": 19487 }, { "epoch": 2.1127493495229834, "grad_norm": 4.092884063720703, "learning_rate": 5.936457275496882e-06, "loss": 0.0135, "step": 19488 }, { "epoch": 2.1128577623590634, "grad_norm": 5.342578887939453, "learning_rate": 5.935731901929494e-06, "loss": 0.0142, "step": 19489 }, { "epoch": 2.112966175195143, "grad_norm": 2.565929651260376, "learning_rate": 5.935006528362107e-06, "loss": 0.0046, "step": 19490 }, { "epoch": 2.113074588031223, "grad_norm": 0.009012448601424694, "learning_rate": 5.9342811547947195e-06, "loss": 0.0, "step": 19491 }, { "epoch": 2.1131830008673025, "grad_norm": 0.014944328926503658, "learning_rate": 5.9335557812273325e-06, "loss": 0.0002, "step": 19492 }, { "epoch": 2.1132914137033825, "grad_norm": 0.393378883600235, "learning_rate": 5.9328304076599454e-06, "loss": 0.0057, "step": 19493 }, { "epoch": 2.1133998265394625, "grad_norm": 0.340198278427124, "learning_rate": 5.932105034092558e-06, "loss": 0.0086, "step": 19494 }, { "epoch": 2.113508239375542, "grad_norm": 0.27006441354751587, "learning_rate": 5.931379660525171e-06, "loss": 0.0004, "step": 19495 }, { "epoch": 2.113616652211622, "grad_norm": 2.1828508377075195, "learning_rate": 5.9306542869577836e-06, "loss": 0.0135, "step": 19496 }, { "epoch": 2.1137250650477015, "grad_norm": 4.357468605041504, "learning_rate": 5.929928913390397e-06, "loss": 0.0154, "step": 19497 }, { "epoch": 2.1138334778837815, "grad_norm": 0.0329500213265419, "learning_rate": 5.9292035398230096e-06, "loss": 0.0002, "step": 19498 }, { "epoch": 2.113941890719861, "grad_norm": 2.0665454864501953, "learning_rate": 5.9284781662556225e-06, "loss": 0.0261, "step": 19499 }, { "epoch": 2.114050303555941, "grad_norm": 0.25953832268714905, "learning_rate": 5.927752792688235e-06, "loss": 0.0055, "step": 19500 }, { "epoch": 2.114158716392021, "grad_norm": 0.3710080683231354, "learning_rate": 5.927027419120848e-06, "loss": 0.0066, "step": 19501 }, { "epoch": 2.1142671292281006, "grad_norm": 2.7085938453674316, "learning_rate": 5.926302045553461e-06, "loss": 0.0222, "step": 19502 }, { "epoch": 2.1143755420641805, "grad_norm": 0.3586137294769287, "learning_rate": 5.925576671986073e-06, "loss": 0.0063, "step": 19503 }, { "epoch": 2.11448395490026, "grad_norm": 0.0020304934587329626, "learning_rate": 5.924851298418686e-06, "loss": 0.0, "step": 19504 }, { "epoch": 2.11459236773634, "grad_norm": 0.5773712396621704, "learning_rate": 5.9241259248513e-06, "loss": 0.012, "step": 19505 }, { "epoch": 2.1147007805724196, "grad_norm": 1.212108850479126, "learning_rate": 5.923400551283912e-06, "loss": 0.0077, "step": 19506 }, { "epoch": 2.1148091934084996, "grad_norm": 1.7814610004425049, "learning_rate": 5.922675177716525e-06, "loss": 0.0266, "step": 19507 }, { "epoch": 2.114917606244579, "grad_norm": 0.06286115944385529, "learning_rate": 5.921949804149138e-06, "loss": 0.0002, "step": 19508 }, { "epoch": 2.115026019080659, "grad_norm": 0.34275102615356445, "learning_rate": 5.92122443058175e-06, "loss": 0.014, "step": 19509 }, { "epoch": 2.115134431916739, "grad_norm": 0.02640637569129467, "learning_rate": 5.920499057014363e-06, "loss": 0.0002, "step": 19510 }, { "epoch": 2.1152428447528187, "grad_norm": 0.01108788512647152, "learning_rate": 5.919773683446975e-06, "loss": 0.0001, "step": 19511 }, { "epoch": 2.1153512575888986, "grad_norm": 1.0692843198776245, "learning_rate": 5.919048309879588e-06, "loss": 0.0136, "step": 19512 }, { "epoch": 2.115459670424978, "grad_norm": 0.06808355450630188, "learning_rate": 5.918322936312201e-06, "loss": 0.0002, "step": 19513 }, { "epoch": 2.115568083261058, "grad_norm": 2.981851100921631, "learning_rate": 5.917597562744815e-06, "loss": 0.0492, "step": 19514 }, { "epoch": 2.1156764960971377, "grad_norm": 0.5155304670333862, "learning_rate": 5.916872189177427e-06, "loss": 0.0168, "step": 19515 }, { "epoch": 2.1157849089332177, "grad_norm": 0.7692033648490906, "learning_rate": 5.91614681561004e-06, "loss": 0.0033, "step": 19516 }, { "epoch": 2.1158933217692977, "grad_norm": 9.652780532836914, "learning_rate": 5.915421442042653e-06, "loss": 0.0166, "step": 19517 }, { "epoch": 2.116001734605377, "grad_norm": 0.29684358835220337, "learning_rate": 5.914696068475265e-06, "loss": 0.0014, "step": 19518 }, { "epoch": 2.116110147441457, "grad_norm": 0.2749001979827881, "learning_rate": 5.913970694907878e-06, "loss": 0.0052, "step": 19519 }, { "epoch": 2.1162185602775367, "grad_norm": 5.001869201660156, "learning_rate": 5.91324532134049e-06, "loss": 0.0121, "step": 19520 }, { "epoch": 2.1163269731136167, "grad_norm": 0.02288762480020523, "learning_rate": 5.912519947773103e-06, "loss": 0.0002, "step": 19521 }, { "epoch": 2.1164353859496963, "grad_norm": 0.3818246126174927, "learning_rate": 5.911794574205716e-06, "loss": 0.007, "step": 19522 }, { "epoch": 2.1165437987857763, "grad_norm": 5.873222351074219, "learning_rate": 5.91106920063833e-06, "loss": 0.0134, "step": 19523 }, { "epoch": 2.1166522116218562, "grad_norm": 0.18204182386398315, "learning_rate": 5.910343827070942e-06, "loss": 0.0004, "step": 19524 }, { "epoch": 2.116760624457936, "grad_norm": 0.07996872812509537, "learning_rate": 5.909618453503555e-06, "loss": 0.0006, "step": 19525 }, { "epoch": 2.1168690372940158, "grad_norm": 0.2792631685733795, "learning_rate": 5.908893079936167e-06, "loss": 0.0054, "step": 19526 }, { "epoch": 2.1169774501300953, "grad_norm": 3.226893186569214, "learning_rate": 5.90816770636878e-06, "loss": 0.0124, "step": 19527 }, { "epoch": 2.1170858629661753, "grad_norm": 0.23261645436286926, "learning_rate": 5.907442332801393e-06, "loss": 0.0046, "step": 19528 }, { "epoch": 2.117194275802255, "grad_norm": 1.9288811683654785, "learning_rate": 5.9067169592340055e-06, "loss": 0.0285, "step": 19529 }, { "epoch": 2.117302688638335, "grad_norm": 1.9939671754837036, "learning_rate": 5.9059915856666185e-06, "loss": 0.002, "step": 19530 }, { "epoch": 2.1174111014744144, "grad_norm": 0.007436971180140972, "learning_rate": 5.905266212099232e-06, "loss": 0.0001, "step": 19531 }, { "epoch": 2.1175195143104943, "grad_norm": 2.2082529067993164, "learning_rate": 5.9045408385318445e-06, "loss": 0.003, "step": 19532 }, { "epoch": 2.1176279271465743, "grad_norm": 11.570651054382324, "learning_rate": 5.9038154649644574e-06, "loss": 0.0053, "step": 19533 }, { "epoch": 2.117736339982654, "grad_norm": 1.97227942943573, "learning_rate": 5.9030900913970704e-06, "loss": 0.0487, "step": 19534 }, { "epoch": 2.117844752818734, "grad_norm": 0.36941689252853394, "learning_rate": 5.9023647178296826e-06, "loss": 0.0219, "step": 19535 }, { "epoch": 2.1179531656548134, "grad_norm": 3.0781116485595703, "learning_rate": 5.9016393442622956e-06, "loss": 0.0225, "step": 19536 }, { "epoch": 2.1180615784908934, "grad_norm": 0.015861811116337776, "learning_rate": 5.900913970694908e-06, "loss": 0.0002, "step": 19537 }, { "epoch": 2.118169991326973, "grad_norm": 0.3333609402179718, "learning_rate": 5.900188597127521e-06, "loss": 0.0073, "step": 19538 }, { "epoch": 2.118278404163053, "grad_norm": 7.752725124359131, "learning_rate": 5.899463223560134e-06, "loss": 0.0908, "step": 19539 }, { "epoch": 2.118386816999133, "grad_norm": 0.48305004835128784, "learning_rate": 5.8987378499927475e-06, "loss": 0.0129, "step": 19540 }, { "epoch": 2.1184952298352124, "grad_norm": 0.5974724292755127, "learning_rate": 5.89801247642536e-06, "loss": 0.014, "step": 19541 }, { "epoch": 2.1186036426712924, "grad_norm": 0.017632126808166504, "learning_rate": 5.897287102857973e-06, "loss": 0.0001, "step": 19542 }, { "epoch": 2.118712055507372, "grad_norm": 2.508159875869751, "learning_rate": 5.896561729290586e-06, "loss": 0.0165, "step": 19543 }, { "epoch": 2.118820468343452, "grad_norm": 0.17024199664592743, "learning_rate": 5.895836355723198e-06, "loss": 0.0004, "step": 19544 }, { "epoch": 2.1189288811795315, "grad_norm": 0.014852040447294712, "learning_rate": 5.895110982155811e-06, "loss": 0.0001, "step": 19545 }, { "epoch": 2.1190372940156115, "grad_norm": 2.706817626953125, "learning_rate": 5.894385608588423e-06, "loss": 0.0897, "step": 19546 }, { "epoch": 2.1191457068516915, "grad_norm": 0.21875478327274323, "learning_rate": 5.893660235021036e-06, "loss": 0.0007, "step": 19547 }, { "epoch": 2.119254119687771, "grad_norm": 0.021182632073760033, "learning_rate": 5.892934861453649e-06, "loss": 0.0002, "step": 19548 }, { "epoch": 2.119362532523851, "grad_norm": 0.029148614034056664, "learning_rate": 5.892209487886263e-06, "loss": 0.0002, "step": 19549 }, { "epoch": 2.1194709453599305, "grad_norm": 6.39862585067749, "learning_rate": 5.891484114318875e-06, "loss": 0.0377, "step": 19550 }, { "epoch": 2.1195793581960105, "grad_norm": 0.015020083636045456, "learning_rate": 5.890758740751488e-06, "loss": 0.0001, "step": 19551 }, { "epoch": 2.11968777103209, "grad_norm": 1.2732782363891602, "learning_rate": 5.8900333671841e-06, "loss": 0.0024, "step": 19552 }, { "epoch": 2.11979618386817, "grad_norm": 0.043259769678115845, "learning_rate": 5.889307993616713e-06, "loss": 0.0002, "step": 19553 }, { "epoch": 2.1199045967042496, "grad_norm": 0.5727802515029907, "learning_rate": 5.888582620049326e-06, "loss": 0.0254, "step": 19554 }, { "epoch": 2.1200130095403296, "grad_norm": 0.46775415539741516, "learning_rate": 5.887857246481938e-06, "loss": 0.0092, "step": 19555 }, { "epoch": 2.1201214223764095, "grad_norm": 0.28925761580467224, "learning_rate": 5.887131872914551e-06, "loss": 0.0104, "step": 19556 }, { "epoch": 2.120229835212489, "grad_norm": 2.939293146133423, "learning_rate": 5.886406499347165e-06, "loss": 0.0442, "step": 19557 }, { "epoch": 2.120338248048569, "grad_norm": 0.0029572939965873957, "learning_rate": 5.885681125779777e-06, "loss": 0.0, "step": 19558 }, { "epoch": 2.1204466608846486, "grad_norm": 0.3477439880371094, "learning_rate": 5.88495575221239e-06, "loss": 0.0009, "step": 19559 }, { "epoch": 2.1205550737207286, "grad_norm": 0.9599785804748535, "learning_rate": 5.884230378645003e-06, "loss": 0.0021, "step": 19560 }, { "epoch": 2.120663486556808, "grad_norm": 1.01102614402771, "learning_rate": 5.883505005077615e-06, "loss": 0.0013, "step": 19561 }, { "epoch": 2.120771899392888, "grad_norm": 2.832165002822876, "learning_rate": 5.882779631510228e-06, "loss": 0.033, "step": 19562 }, { "epoch": 2.120880312228968, "grad_norm": 15.869142532348633, "learning_rate": 5.88205425794284e-06, "loss": 0.0291, "step": 19563 }, { "epoch": 2.1209887250650477, "grad_norm": 0.2602067291736603, "learning_rate": 5.881328884375453e-06, "loss": 0.0013, "step": 19564 }, { "epoch": 2.1210971379011276, "grad_norm": 0.26004308462142944, "learning_rate": 5.880603510808066e-06, "loss": 0.005, "step": 19565 }, { "epoch": 2.121205550737207, "grad_norm": 6.119291305541992, "learning_rate": 5.87987813724068e-06, "loss": 0.0353, "step": 19566 }, { "epoch": 2.121313963573287, "grad_norm": 0.01833210326731205, "learning_rate": 5.879152763673292e-06, "loss": 0.0001, "step": 19567 }, { "epoch": 2.1214223764093667, "grad_norm": 0.06328710168600082, "learning_rate": 5.878427390105905e-06, "loss": 0.0003, "step": 19568 }, { "epoch": 2.1215307892454467, "grad_norm": 0.4603751003742218, "learning_rate": 5.877702016538518e-06, "loss": 0.0169, "step": 19569 }, { "epoch": 2.1216392020815267, "grad_norm": 0.008355841971933842, "learning_rate": 5.8769766429711305e-06, "loss": 0.0001, "step": 19570 }, { "epoch": 2.121747614917606, "grad_norm": 0.5565474033355713, "learning_rate": 5.8762512694037435e-06, "loss": 0.0125, "step": 19571 }, { "epoch": 2.121856027753686, "grad_norm": 0.12749192118644714, "learning_rate": 5.875525895836356e-06, "loss": 0.0008, "step": 19572 }, { "epoch": 2.1219644405897657, "grad_norm": 0.6427545547485352, "learning_rate": 5.874800522268969e-06, "loss": 0.0267, "step": 19573 }, { "epoch": 2.1220728534258457, "grad_norm": 5.664020538330078, "learning_rate": 5.874075148701582e-06, "loss": 0.0369, "step": 19574 }, { "epoch": 2.1221812662619253, "grad_norm": 0.4528435170650482, "learning_rate": 5.873349775134195e-06, "loss": 0.0171, "step": 19575 }, { "epoch": 2.1222896790980053, "grad_norm": 17.80399513244629, "learning_rate": 5.8726244015668076e-06, "loss": 0.0339, "step": 19576 }, { "epoch": 2.122398091934085, "grad_norm": 0.3423251807689667, "learning_rate": 5.8718990279994205e-06, "loss": 0.0056, "step": 19577 }, { "epoch": 2.122506504770165, "grad_norm": 0.2568267285823822, "learning_rate": 5.871173654432033e-06, "loss": 0.0009, "step": 19578 }, { "epoch": 2.1226149176062448, "grad_norm": 0.22879528999328613, "learning_rate": 5.870448280864646e-06, "loss": 0.0048, "step": 19579 }, { "epoch": 2.1227233304423243, "grad_norm": 0.3017609715461731, "learning_rate": 5.869722907297259e-06, "loss": 0.0109, "step": 19580 }, { "epoch": 2.1228317432784043, "grad_norm": 3.3646957874298096, "learning_rate": 5.868997533729871e-06, "loss": 0.0212, "step": 19581 }, { "epoch": 2.122940156114484, "grad_norm": 10.700647354125977, "learning_rate": 5.868272160162484e-06, "loss": 0.0536, "step": 19582 }, { "epoch": 2.123048568950564, "grad_norm": 0.3304556906223297, "learning_rate": 5.867546786595098e-06, "loss": 0.0119, "step": 19583 }, { "epoch": 2.1231569817866434, "grad_norm": 15.18225383758545, "learning_rate": 5.86682141302771e-06, "loss": 0.1045, "step": 19584 }, { "epoch": 2.1232653946227233, "grad_norm": 0.5516979098320007, "learning_rate": 5.866096039460323e-06, "loss": 0.0154, "step": 19585 }, { "epoch": 2.1233738074588033, "grad_norm": 0.41807788610458374, "learning_rate": 5.865370665892936e-06, "loss": 0.0072, "step": 19586 }, { "epoch": 2.123482220294883, "grad_norm": 1.3652617931365967, "learning_rate": 5.864645292325548e-06, "loss": 0.0361, "step": 19587 }, { "epoch": 2.123590633130963, "grad_norm": 0.12178952246904373, "learning_rate": 5.863919918758161e-06, "loss": 0.0004, "step": 19588 }, { "epoch": 2.1236990459670424, "grad_norm": 0.20271486043930054, "learning_rate": 5.863194545190773e-06, "loss": 0.0005, "step": 19589 }, { "epoch": 2.1238074588031224, "grad_norm": 2.6049466133117676, "learning_rate": 5.862469171623386e-06, "loss": 0.0226, "step": 19590 }, { "epoch": 2.123915871639202, "grad_norm": 0.14265388250350952, "learning_rate": 5.861743798055999e-06, "loss": 0.0023, "step": 19591 }, { "epoch": 2.124024284475282, "grad_norm": 4.55855131149292, "learning_rate": 5.861018424488613e-06, "loss": 0.0162, "step": 19592 }, { "epoch": 2.1241326973113615, "grad_norm": 4.050947189331055, "learning_rate": 5.860293050921225e-06, "loss": 0.0248, "step": 19593 }, { "epoch": 2.1242411101474414, "grad_norm": 1.761530876159668, "learning_rate": 5.859567677353838e-06, "loss": 0.0074, "step": 19594 }, { "epoch": 2.1243495229835214, "grad_norm": 0.024686258286237717, "learning_rate": 5.858842303786451e-06, "loss": 0.0001, "step": 19595 }, { "epoch": 2.124457935819601, "grad_norm": 0.4229811131954193, "learning_rate": 5.858116930219063e-06, "loss": 0.0179, "step": 19596 }, { "epoch": 2.124566348655681, "grad_norm": 0.9885755181312561, "learning_rate": 5.857391556651676e-06, "loss": 0.0032, "step": 19597 }, { "epoch": 2.1246747614917605, "grad_norm": 5.227764129638672, "learning_rate": 5.856666183084288e-06, "loss": 0.0199, "step": 19598 }, { "epoch": 2.1247831743278405, "grad_norm": 0.3426004946231842, "learning_rate": 5.855940809516901e-06, "loss": 0.0141, "step": 19599 }, { "epoch": 2.12489158716392, "grad_norm": 0.2682301700115204, "learning_rate": 5.855215435949515e-06, "loss": 0.0012, "step": 19600 }, { "epoch": 2.125, "grad_norm": 0.005328039638698101, "learning_rate": 5.854490062382128e-06, "loss": 0.0001, "step": 19601 }, { "epoch": 2.12510841283608, "grad_norm": 0.7838541269302368, "learning_rate": 5.85376468881474e-06, "loss": 0.0119, "step": 19602 }, { "epoch": 2.1252168256721595, "grad_norm": 0.35934358835220337, "learning_rate": 5.853039315247353e-06, "loss": 0.0109, "step": 19603 }, { "epoch": 2.1253252385082395, "grad_norm": 0.02816581353545189, "learning_rate": 5.852313941679965e-06, "loss": 0.0001, "step": 19604 }, { "epoch": 2.125433651344319, "grad_norm": 0.935644268989563, "learning_rate": 5.851588568112578e-06, "loss": 0.002, "step": 19605 }, { "epoch": 2.125542064180399, "grad_norm": 3.6772572994232178, "learning_rate": 5.850863194545191e-06, "loss": 0.0396, "step": 19606 }, { "epoch": 2.1256504770164786, "grad_norm": 0.08932032436132431, "learning_rate": 5.8501378209778035e-06, "loss": 0.0007, "step": 19607 }, { "epoch": 2.1257588898525586, "grad_norm": 0.2486647218465805, "learning_rate": 5.8494124474104165e-06, "loss": 0.0051, "step": 19608 }, { "epoch": 2.125867302688638, "grad_norm": 1.913732647895813, "learning_rate": 5.84868707384303e-06, "loss": 0.0052, "step": 19609 }, { "epoch": 2.125975715524718, "grad_norm": 1.3447265625, "learning_rate": 5.8479617002756425e-06, "loss": 0.0137, "step": 19610 }, { "epoch": 2.126084128360798, "grad_norm": 0.3943750560283661, "learning_rate": 5.8472363267082554e-06, "loss": 0.0121, "step": 19611 }, { "epoch": 2.1261925411968776, "grad_norm": 0.13459645211696625, "learning_rate": 5.8465109531408684e-06, "loss": 0.0011, "step": 19612 }, { "epoch": 2.1263009540329576, "grad_norm": 0.25171419978141785, "learning_rate": 5.845785579573481e-06, "loss": 0.0031, "step": 19613 }, { "epoch": 2.126409366869037, "grad_norm": 1.5892415046691895, "learning_rate": 5.8450602060060936e-06, "loss": 0.0032, "step": 19614 }, { "epoch": 2.126517779705117, "grad_norm": 0.3983924388885498, "learning_rate": 5.8443348324387066e-06, "loss": 0.0118, "step": 19615 }, { "epoch": 2.126626192541197, "grad_norm": 5.272879600524902, "learning_rate": 5.843609458871319e-06, "loss": 0.0074, "step": 19616 }, { "epoch": 2.1267346053772767, "grad_norm": 3.0367517471313477, "learning_rate": 5.842884085303932e-06, "loss": 0.0312, "step": 19617 }, { "epoch": 2.1268430182133566, "grad_norm": 0.31422460079193115, "learning_rate": 5.8421587117365455e-06, "loss": 0.0066, "step": 19618 }, { "epoch": 2.126951431049436, "grad_norm": 0.29136383533477783, "learning_rate": 5.841433338169158e-06, "loss": 0.0061, "step": 19619 }, { "epoch": 2.127059843885516, "grad_norm": 0.006526959128677845, "learning_rate": 5.840707964601771e-06, "loss": 0.0, "step": 19620 }, { "epoch": 2.1271682567215957, "grad_norm": 0.4220638871192932, "learning_rate": 5.839982591034384e-06, "loss": 0.0072, "step": 19621 }, { "epoch": 2.1272766695576757, "grad_norm": 0.3252708315849304, "learning_rate": 5.839257217466996e-06, "loss": 0.007, "step": 19622 }, { "epoch": 2.1273850823937552, "grad_norm": 0.35350075364112854, "learning_rate": 5.838531843899609e-06, "loss": 0.0113, "step": 19623 }, { "epoch": 2.127493495229835, "grad_norm": 0.7677720189094543, "learning_rate": 5.837806470332221e-06, "loss": 0.0133, "step": 19624 }, { "epoch": 2.127601908065915, "grad_norm": 0.22185762226581573, "learning_rate": 5.837081096764834e-06, "loss": 0.0004, "step": 19625 }, { "epoch": 2.1277103209019947, "grad_norm": 0.44735822081565857, "learning_rate": 5.836355723197448e-06, "loss": 0.0022, "step": 19626 }, { "epoch": 2.1278187337380747, "grad_norm": 0.25088775157928467, "learning_rate": 5.835630349630061e-06, "loss": 0.0051, "step": 19627 }, { "epoch": 2.1279271465741543, "grad_norm": 1.2304928302764893, "learning_rate": 5.834904976062673e-06, "loss": 0.0059, "step": 19628 }, { "epoch": 2.1280355594102343, "grad_norm": 0.14565156400203705, "learning_rate": 5.834179602495286e-06, "loss": 0.0007, "step": 19629 }, { "epoch": 2.128143972246314, "grad_norm": 2.8803374767303467, "learning_rate": 5.833454228927898e-06, "loss": 0.0533, "step": 19630 }, { "epoch": 2.128252385082394, "grad_norm": 0.40688425302505493, "learning_rate": 5.832728855360511e-06, "loss": 0.0073, "step": 19631 }, { "epoch": 2.1283607979184733, "grad_norm": 0.2642396092414856, "learning_rate": 5.832003481793124e-06, "loss": 0.0051, "step": 19632 }, { "epoch": 2.1284692107545533, "grad_norm": 3.1340725421905518, "learning_rate": 5.831278108225736e-06, "loss": 0.0344, "step": 19633 }, { "epoch": 2.1285776235906333, "grad_norm": 7.121674537658691, "learning_rate": 5.830552734658349e-06, "loss": 0.0451, "step": 19634 }, { "epoch": 2.128686036426713, "grad_norm": 0.39337316155433655, "learning_rate": 5.829827361090963e-06, "loss": 0.0078, "step": 19635 }, { "epoch": 2.128794449262793, "grad_norm": 0.3235127925872803, "learning_rate": 5.829101987523575e-06, "loss": 0.0115, "step": 19636 }, { "epoch": 2.1289028620988724, "grad_norm": 0.04046735167503357, "learning_rate": 5.828376613956188e-06, "loss": 0.0001, "step": 19637 }, { "epoch": 2.1290112749349523, "grad_norm": 0.30066150426864624, "learning_rate": 5.827651240388801e-06, "loss": 0.0161, "step": 19638 }, { "epoch": 2.129119687771032, "grad_norm": 2.7820887565612793, "learning_rate": 5.826925866821413e-06, "loss": 0.0265, "step": 19639 }, { "epoch": 2.129228100607112, "grad_norm": 0.3541220426559448, "learning_rate": 5.826200493254026e-06, "loss": 0.0092, "step": 19640 }, { "epoch": 2.129336513443192, "grad_norm": 12.640274047851562, "learning_rate": 5.825475119686639e-06, "loss": 0.0415, "step": 19641 }, { "epoch": 2.1294449262792714, "grad_norm": 1.5580384731292725, "learning_rate": 5.824749746119251e-06, "loss": 0.0132, "step": 19642 }, { "epoch": 2.1295533391153514, "grad_norm": 0.5279702544212341, "learning_rate": 5.824024372551864e-06, "loss": 0.0052, "step": 19643 }, { "epoch": 2.129661751951431, "grad_norm": 8.865723609924316, "learning_rate": 5.823298998984478e-06, "loss": 0.0397, "step": 19644 }, { "epoch": 2.129770164787511, "grad_norm": 0.7932853698730469, "learning_rate": 5.82257362541709e-06, "loss": 0.0021, "step": 19645 }, { "epoch": 2.1298785776235905, "grad_norm": 0.3487800061702728, "learning_rate": 5.821848251849703e-06, "loss": 0.0005, "step": 19646 }, { "epoch": 2.1299869904596704, "grad_norm": 0.717025101184845, "learning_rate": 5.821122878282316e-06, "loss": 0.0159, "step": 19647 }, { "epoch": 2.1300954032957504, "grad_norm": 0.3087014853954315, "learning_rate": 5.8203975047149285e-06, "loss": 0.0114, "step": 19648 }, { "epoch": 2.13020381613183, "grad_norm": 3.3957200050354004, "learning_rate": 5.8196721311475415e-06, "loss": 0.012, "step": 19649 }, { "epoch": 2.13031222896791, "grad_norm": 0.3494822382926941, "learning_rate": 5.818946757580154e-06, "loss": 0.012, "step": 19650 }, { "epoch": 2.1304206418039895, "grad_norm": 0.005334843415766954, "learning_rate": 5.818221384012767e-06, "loss": 0.0001, "step": 19651 }, { "epoch": 2.1305290546400695, "grad_norm": 0.26508620381355286, "learning_rate": 5.8174960104453804e-06, "loss": 0.0052, "step": 19652 }, { "epoch": 2.130637467476149, "grad_norm": 3.489133596420288, "learning_rate": 5.816770636877993e-06, "loss": 0.0492, "step": 19653 }, { "epoch": 2.130745880312229, "grad_norm": 2.8598477840423584, "learning_rate": 5.8160452633106056e-06, "loss": 0.0153, "step": 19654 }, { "epoch": 2.1308542931483085, "grad_norm": 0.029512029141187668, "learning_rate": 5.8153198897432186e-06, "loss": 0.0001, "step": 19655 }, { "epoch": 2.1309627059843885, "grad_norm": 0.03527960926294327, "learning_rate": 5.814594516175831e-06, "loss": 0.0002, "step": 19656 }, { "epoch": 2.1310711188204685, "grad_norm": 1.1616922616958618, "learning_rate": 5.813869142608444e-06, "loss": 0.005, "step": 19657 }, { "epoch": 2.131179531656548, "grad_norm": 0.17813141644001007, "learning_rate": 5.813143769041057e-06, "loss": 0.0016, "step": 19658 }, { "epoch": 2.131287944492628, "grad_norm": 0.2352234274148941, "learning_rate": 5.812418395473669e-06, "loss": 0.0015, "step": 19659 }, { "epoch": 2.1313963573287076, "grad_norm": 0.3629254996776581, "learning_rate": 5.811693021906282e-06, "loss": 0.0106, "step": 19660 }, { "epoch": 2.1315047701647876, "grad_norm": 1.8566118478775024, "learning_rate": 5.810967648338896e-06, "loss": 0.029, "step": 19661 }, { "epoch": 2.131613183000867, "grad_norm": 1.4616494178771973, "learning_rate": 5.810242274771508e-06, "loss": 0.0217, "step": 19662 }, { "epoch": 2.131721595836947, "grad_norm": 0.01437662448734045, "learning_rate": 5.809516901204121e-06, "loss": 0.0001, "step": 19663 }, { "epoch": 2.131830008673027, "grad_norm": 0.5117490291595459, "learning_rate": 5.808791527636734e-06, "loss": 0.0161, "step": 19664 }, { "epoch": 2.1319384215091066, "grad_norm": 0.11149817705154419, "learning_rate": 5.808066154069346e-06, "loss": 0.0008, "step": 19665 }, { "epoch": 2.1320468343451866, "grad_norm": 0.029471950605511665, "learning_rate": 5.807340780501959e-06, "loss": 0.0002, "step": 19666 }, { "epoch": 2.132155247181266, "grad_norm": 0.2365017980337143, "learning_rate": 5.806615406934572e-06, "loss": 0.0006, "step": 19667 }, { "epoch": 2.132263660017346, "grad_norm": 13.10945987701416, "learning_rate": 5.805890033367184e-06, "loss": 0.0186, "step": 19668 }, { "epoch": 2.1323720728534257, "grad_norm": 2.539182186126709, "learning_rate": 5.805164659799797e-06, "loss": 0.0013, "step": 19669 }, { "epoch": 2.1324804856895057, "grad_norm": 1.6914315223693848, "learning_rate": 5.804439286232411e-06, "loss": 0.0424, "step": 19670 }, { "epoch": 2.1325888985255856, "grad_norm": 0.378801167011261, "learning_rate": 5.803713912665023e-06, "loss": 0.0105, "step": 19671 }, { "epoch": 2.132697311361665, "grad_norm": 0.3106669485569, "learning_rate": 5.802988539097636e-06, "loss": 0.0061, "step": 19672 }, { "epoch": 2.132805724197745, "grad_norm": 0.5132223963737488, "learning_rate": 5.802263165530249e-06, "loss": 0.0056, "step": 19673 }, { "epoch": 2.1329141370338247, "grad_norm": 0.2881225049495697, "learning_rate": 5.801537791962861e-06, "loss": 0.0051, "step": 19674 }, { "epoch": 2.1330225498699047, "grad_norm": 0.4609180688858032, "learning_rate": 5.800812418395474e-06, "loss": 0.0007, "step": 19675 }, { "epoch": 2.1331309627059842, "grad_norm": 10.005468368530273, "learning_rate": 5.800087044828086e-06, "loss": 0.0118, "step": 19676 }, { "epoch": 2.133239375542064, "grad_norm": 0.06620212644338608, "learning_rate": 5.799361671260699e-06, "loss": 0.0002, "step": 19677 }, { "epoch": 2.1333477883781438, "grad_norm": 0.3307555615901947, "learning_rate": 5.798636297693313e-06, "loss": 0.001, "step": 19678 }, { "epoch": 2.1334562012142237, "grad_norm": 2.5690462589263916, "learning_rate": 5.797910924125926e-06, "loss": 0.0226, "step": 19679 }, { "epoch": 2.1335646140503037, "grad_norm": 0.4507180154323578, "learning_rate": 5.797185550558538e-06, "loss": 0.0116, "step": 19680 }, { "epoch": 2.1336730268863833, "grad_norm": 0.33956092596054077, "learning_rate": 5.796460176991151e-06, "loss": 0.0015, "step": 19681 }, { "epoch": 2.1337814397224633, "grad_norm": 6.2139997482299805, "learning_rate": 5.795734803423763e-06, "loss": 0.0117, "step": 19682 }, { "epoch": 2.133889852558543, "grad_norm": 0.1669800579547882, "learning_rate": 5.795009429856376e-06, "loss": 0.0126, "step": 19683 }, { "epoch": 2.133998265394623, "grad_norm": 0.15016207098960876, "learning_rate": 5.794284056288989e-06, "loss": 0.0006, "step": 19684 }, { "epoch": 2.1341066782307023, "grad_norm": 4.091774940490723, "learning_rate": 5.7935586827216015e-06, "loss": 0.0297, "step": 19685 }, { "epoch": 2.1342150910667823, "grad_norm": 4.065063953399658, "learning_rate": 5.7928333091542145e-06, "loss": 0.0579, "step": 19686 }, { "epoch": 2.1343235039028623, "grad_norm": 1.2364637851715088, "learning_rate": 5.792107935586828e-06, "loss": 0.0448, "step": 19687 }, { "epoch": 2.134431916738942, "grad_norm": 0.4477856457233429, "learning_rate": 5.7913825620194405e-06, "loss": 0.0078, "step": 19688 }, { "epoch": 2.134540329575022, "grad_norm": 0.3058364987373352, "learning_rate": 5.7906571884520535e-06, "loss": 0.0059, "step": 19689 }, { "epoch": 2.1346487424111014, "grad_norm": 0.39745500683784485, "learning_rate": 5.7899318148846664e-06, "loss": 0.0134, "step": 19690 }, { "epoch": 2.1347571552471813, "grad_norm": 0.28091415762901306, "learning_rate": 5.789206441317279e-06, "loss": 0.0008, "step": 19691 }, { "epoch": 2.134865568083261, "grad_norm": 2.4647274017333984, "learning_rate": 5.788481067749892e-06, "loss": 0.0262, "step": 19692 }, { "epoch": 2.134973980919341, "grad_norm": 0.13251186907291412, "learning_rate": 5.7877556941825046e-06, "loss": 0.0004, "step": 19693 }, { "epoch": 2.135082393755421, "grad_norm": 0.5783589482307434, "learning_rate": 5.787030320615117e-06, "loss": 0.01, "step": 19694 }, { "epoch": 2.1351908065915004, "grad_norm": 10.772628784179688, "learning_rate": 5.7863049470477305e-06, "loss": 0.0265, "step": 19695 }, { "epoch": 2.1352992194275804, "grad_norm": 10.367121696472168, "learning_rate": 5.7855795734803435e-06, "loss": 0.0192, "step": 19696 }, { "epoch": 2.13540763226366, "grad_norm": 0.7442491054534912, "learning_rate": 5.784854199912956e-06, "loss": 0.017, "step": 19697 }, { "epoch": 2.13551604509974, "grad_norm": 0.5374119281768799, "learning_rate": 5.784128826345569e-06, "loss": 0.0022, "step": 19698 }, { "epoch": 2.1356244579358195, "grad_norm": 0.09846776723861694, "learning_rate": 5.783403452778182e-06, "loss": 0.0007, "step": 19699 }, { "epoch": 2.1357328707718994, "grad_norm": 4.4281229972839355, "learning_rate": 5.782678079210794e-06, "loss": 0.0291, "step": 19700 }, { "epoch": 2.135841283607979, "grad_norm": 0.3089195787906647, "learning_rate": 5.781952705643407e-06, "loss": 0.0073, "step": 19701 }, { "epoch": 2.135949696444059, "grad_norm": 0.5210570693016052, "learning_rate": 5.781227332076019e-06, "loss": 0.0169, "step": 19702 }, { "epoch": 2.136058109280139, "grad_norm": 0.3424728214740753, "learning_rate": 5.780501958508632e-06, "loss": 0.0104, "step": 19703 }, { "epoch": 2.1361665221162185, "grad_norm": 1.4288979768753052, "learning_rate": 5.779776584941246e-06, "loss": 0.0105, "step": 19704 }, { "epoch": 2.1362749349522985, "grad_norm": 5.451945781707764, "learning_rate": 5.779051211373859e-06, "loss": 0.0116, "step": 19705 }, { "epoch": 2.136383347788378, "grad_norm": 0.4372846782207489, "learning_rate": 5.778325837806471e-06, "loss": 0.0095, "step": 19706 }, { "epoch": 2.136491760624458, "grad_norm": 0.13112282752990723, "learning_rate": 5.777600464239084e-06, "loss": 0.0004, "step": 19707 }, { "epoch": 2.1366001734605375, "grad_norm": 6.234820365905762, "learning_rate": 5.776875090671696e-06, "loss": 0.011, "step": 19708 }, { "epoch": 2.1367085862966175, "grad_norm": 3.1480157375335693, "learning_rate": 5.776149717104309e-06, "loss": 0.0385, "step": 19709 }, { "epoch": 2.1368169991326975, "grad_norm": 4.373185157775879, "learning_rate": 5.775424343536922e-06, "loss": 0.0256, "step": 19710 }, { "epoch": 2.136925411968777, "grad_norm": 0.2867618203163147, "learning_rate": 5.774698969969534e-06, "loss": 0.0099, "step": 19711 }, { "epoch": 2.137033824804857, "grad_norm": 0.7884228825569153, "learning_rate": 5.773973596402147e-06, "loss": 0.0021, "step": 19712 }, { "epoch": 2.1371422376409366, "grad_norm": 0.2833407521247864, "learning_rate": 5.773248222834761e-06, "loss": 0.0108, "step": 19713 }, { "epoch": 2.1372506504770166, "grad_norm": 0.5064643621444702, "learning_rate": 5.772522849267373e-06, "loss": 0.0092, "step": 19714 }, { "epoch": 2.137359063313096, "grad_norm": 0.036103036254644394, "learning_rate": 5.771797475699986e-06, "loss": 0.0004, "step": 19715 }, { "epoch": 2.137467476149176, "grad_norm": 4.693575859069824, "learning_rate": 5.771072102132599e-06, "loss": 0.0463, "step": 19716 }, { "epoch": 2.137575888985256, "grad_norm": 8.227163314819336, "learning_rate": 5.770346728565211e-06, "loss": 0.0159, "step": 19717 }, { "epoch": 2.1376843018213356, "grad_norm": 1.6553493738174438, "learning_rate": 5.769621354997824e-06, "loss": 0.0026, "step": 19718 }, { "epoch": 2.1377927146574156, "grad_norm": 0.08451870083808899, "learning_rate": 5.768895981430437e-06, "loss": 0.0008, "step": 19719 }, { "epoch": 2.137901127493495, "grad_norm": 2.3706045150756836, "learning_rate": 5.768170607863049e-06, "loss": 0.0138, "step": 19720 }, { "epoch": 2.138009540329575, "grad_norm": 0.3829968273639679, "learning_rate": 5.767445234295663e-06, "loss": 0.009, "step": 19721 }, { "epoch": 2.1381179531656547, "grad_norm": 1.3861358165740967, "learning_rate": 5.766719860728276e-06, "loss": 0.0007, "step": 19722 }, { "epoch": 2.1382263660017347, "grad_norm": 0.15541334450244904, "learning_rate": 5.765994487160888e-06, "loss": 0.0003, "step": 19723 }, { "epoch": 2.138334778837814, "grad_norm": 1.5617518424987793, "learning_rate": 5.765269113593501e-06, "loss": 0.015, "step": 19724 }, { "epoch": 2.138443191673894, "grad_norm": 0.23303458094596863, "learning_rate": 5.764543740026114e-06, "loss": 0.0045, "step": 19725 }, { "epoch": 2.138551604509974, "grad_norm": 0.2900071442127228, "learning_rate": 5.7638183664587265e-06, "loss": 0.0069, "step": 19726 }, { "epoch": 2.1386600173460537, "grad_norm": 0.6849187612533569, "learning_rate": 5.7630929928913395e-06, "loss": 0.0252, "step": 19727 }, { "epoch": 2.1387684301821337, "grad_norm": 0.018932398408651352, "learning_rate": 5.762367619323952e-06, "loss": 0.0001, "step": 19728 }, { "epoch": 2.1388768430182132, "grad_norm": 0.11225401610136032, "learning_rate": 5.761642245756565e-06, "loss": 0.0005, "step": 19729 }, { "epoch": 2.138985255854293, "grad_norm": 2.495755910873413, "learning_rate": 5.7609168721891784e-06, "loss": 0.0239, "step": 19730 }, { "epoch": 2.1390936686903728, "grad_norm": 1.4047340154647827, "learning_rate": 5.7601914986217914e-06, "loss": 0.0122, "step": 19731 }, { "epoch": 2.1392020815264527, "grad_norm": 0.3972684144973755, "learning_rate": 5.7594661250544036e-06, "loss": 0.007, "step": 19732 }, { "epoch": 2.1393104943625323, "grad_norm": 0.3344757854938507, "learning_rate": 5.7587407514870166e-06, "loss": 0.0013, "step": 19733 }, { "epoch": 2.1394189071986123, "grad_norm": 0.3116396367549896, "learning_rate": 5.758015377919629e-06, "loss": 0.0071, "step": 19734 }, { "epoch": 2.1395273200346923, "grad_norm": 0.427950918674469, "learning_rate": 5.757290004352242e-06, "loss": 0.0129, "step": 19735 }, { "epoch": 2.139635732870772, "grad_norm": 0.01626378484070301, "learning_rate": 5.756564630784855e-06, "loss": 0.0002, "step": 19736 }, { "epoch": 2.139744145706852, "grad_norm": 0.4579634368419647, "learning_rate": 5.755839257217467e-06, "loss": 0.017, "step": 19737 }, { "epoch": 2.1398525585429313, "grad_norm": 0.026559356600046158, "learning_rate": 5.75511388365008e-06, "loss": 0.0001, "step": 19738 }, { "epoch": 2.1399609713790113, "grad_norm": 0.05650302395224571, "learning_rate": 5.754388510082694e-06, "loss": 0.0004, "step": 19739 }, { "epoch": 2.1400693842150913, "grad_norm": 0.44113436341285706, "learning_rate": 5.753663136515306e-06, "loss": 0.0089, "step": 19740 }, { "epoch": 2.140177797051171, "grad_norm": 0.41689333319664, "learning_rate": 5.752937762947919e-06, "loss": 0.0058, "step": 19741 }, { "epoch": 2.140286209887251, "grad_norm": 0.2775225043296814, "learning_rate": 5.752212389380532e-06, "loss": 0.01, "step": 19742 }, { "epoch": 2.1403946227233304, "grad_norm": 1.3629266023635864, "learning_rate": 5.751487015813144e-06, "loss": 0.006, "step": 19743 }, { "epoch": 2.1405030355594103, "grad_norm": 1.7229764461517334, "learning_rate": 5.750761642245757e-06, "loss": 0.0086, "step": 19744 }, { "epoch": 2.14061144839549, "grad_norm": 0.0562109500169754, "learning_rate": 5.75003626867837e-06, "loss": 0.0003, "step": 19745 }, { "epoch": 2.14071986123157, "grad_norm": 0.483960896730423, "learning_rate": 5.749310895110982e-06, "loss": 0.0178, "step": 19746 }, { "epoch": 2.1408282740676494, "grad_norm": 1.8487813472747803, "learning_rate": 5.748585521543596e-06, "loss": 0.0036, "step": 19747 }, { "epoch": 2.1409366869037294, "grad_norm": 0.05070766806602478, "learning_rate": 5.747860147976209e-06, "loss": 0.0003, "step": 19748 }, { "epoch": 2.1410450997398094, "grad_norm": 2.85735821723938, "learning_rate": 5.747134774408821e-06, "loss": 0.0192, "step": 19749 }, { "epoch": 2.141153512575889, "grad_norm": 0.672768771648407, "learning_rate": 5.746409400841434e-06, "loss": 0.0011, "step": 19750 }, { "epoch": 2.141261925411969, "grad_norm": 1.6316694021224976, "learning_rate": 5.745684027274047e-06, "loss": 0.0137, "step": 19751 }, { "epoch": 2.1413703382480485, "grad_norm": 0.49805212020874023, "learning_rate": 5.744958653706659e-06, "loss": 0.0009, "step": 19752 }, { "epoch": 2.1414787510841284, "grad_norm": 16.052053451538086, "learning_rate": 5.744233280139272e-06, "loss": 0.0437, "step": 19753 }, { "epoch": 2.141587163920208, "grad_norm": 0.3148638606071472, "learning_rate": 5.743507906571884e-06, "loss": 0.0061, "step": 19754 }, { "epoch": 2.141695576756288, "grad_norm": 2.6545536518096924, "learning_rate": 5.742782533004497e-06, "loss": 0.0317, "step": 19755 }, { "epoch": 2.1418039895923675, "grad_norm": 0.23212410509586334, "learning_rate": 5.742057159437111e-06, "loss": 0.0049, "step": 19756 }, { "epoch": 2.1419124024284475, "grad_norm": 0.010457337833940983, "learning_rate": 5.741331785869724e-06, "loss": 0.0001, "step": 19757 }, { "epoch": 2.1420208152645275, "grad_norm": 2.539128541946411, "learning_rate": 5.740606412302336e-06, "loss": 0.0428, "step": 19758 }, { "epoch": 2.142129228100607, "grad_norm": 0.13971081376075745, "learning_rate": 5.739881038734949e-06, "loss": 0.0008, "step": 19759 }, { "epoch": 2.142237640936687, "grad_norm": 0.24609313905239105, "learning_rate": 5.739155665167561e-06, "loss": 0.0053, "step": 19760 }, { "epoch": 2.1423460537727665, "grad_norm": 0.3023563027381897, "learning_rate": 5.738430291600174e-06, "loss": 0.0007, "step": 19761 }, { "epoch": 2.1424544666088465, "grad_norm": 0.6047632098197937, "learning_rate": 5.737704918032787e-06, "loss": 0.0147, "step": 19762 }, { "epoch": 2.142562879444926, "grad_norm": 0.4225057363510132, "learning_rate": 5.7369795444653995e-06, "loss": 0.0095, "step": 19763 }, { "epoch": 2.142671292281006, "grad_norm": 0.009385920129716396, "learning_rate": 5.7362541708980125e-06, "loss": 0.0001, "step": 19764 }, { "epoch": 2.142779705117086, "grad_norm": 7.393016338348389, "learning_rate": 5.735528797330626e-06, "loss": 0.046, "step": 19765 }, { "epoch": 2.1428881179531656, "grad_norm": 0.03487834334373474, "learning_rate": 5.7348034237632385e-06, "loss": 0.0003, "step": 19766 }, { "epoch": 2.1429965307892456, "grad_norm": 0.41584745049476624, "learning_rate": 5.7340780501958515e-06, "loss": 0.0004, "step": 19767 }, { "epoch": 2.143104943625325, "grad_norm": 0.7578663229942322, "learning_rate": 5.7333526766284645e-06, "loss": 0.0252, "step": 19768 }, { "epoch": 2.143213356461405, "grad_norm": 0.0053428588435053825, "learning_rate": 5.732627303061077e-06, "loss": 0.0, "step": 19769 }, { "epoch": 2.1433217692974846, "grad_norm": 1.6878962516784668, "learning_rate": 5.73190192949369e-06, "loss": 0.0335, "step": 19770 }, { "epoch": 2.1434301821335646, "grad_norm": 0.6306493878364563, "learning_rate": 5.7311765559263026e-06, "loss": 0.0211, "step": 19771 }, { "epoch": 2.1435385949696446, "grad_norm": 2.8863067626953125, "learning_rate": 5.730451182358915e-06, "loss": 0.0417, "step": 19772 }, { "epoch": 2.143647007805724, "grad_norm": 24.027437210083008, "learning_rate": 5.7297258087915286e-06, "loss": 0.0341, "step": 19773 }, { "epoch": 2.143755420641804, "grad_norm": 0.508781373500824, "learning_rate": 5.7290004352241415e-06, "loss": 0.0072, "step": 19774 }, { "epoch": 2.1438638334778837, "grad_norm": 0.07055062800645828, "learning_rate": 5.728275061656754e-06, "loss": 0.0004, "step": 19775 }, { "epoch": 2.1439722463139637, "grad_norm": 0.26050370931625366, "learning_rate": 5.727549688089367e-06, "loss": 0.004, "step": 19776 }, { "epoch": 2.144080659150043, "grad_norm": 1.4034185409545898, "learning_rate": 5.72682431452198e-06, "loss": 0.0285, "step": 19777 }, { "epoch": 2.144189071986123, "grad_norm": 0.04152160882949829, "learning_rate": 5.726098940954592e-06, "loss": 0.0003, "step": 19778 }, { "epoch": 2.1442974848222027, "grad_norm": 0.28390929102897644, "learning_rate": 5.725373567387205e-06, "loss": 0.006, "step": 19779 }, { "epoch": 2.1444058976582827, "grad_norm": 1.5185476541519165, "learning_rate": 5.724648193819817e-06, "loss": 0.0137, "step": 19780 }, { "epoch": 2.1445143104943627, "grad_norm": 4.315116882324219, "learning_rate": 5.72392282025243e-06, "loss": 0.0187, "step": 19781 }, { "epoch": 2.1446227233304422, "grad_norm": 0.3130434453487396, "learning_rate": 5.723197446685044e-06, "loss": 0.007, "step": 19782 }, { "epoch": 2.1447311361665222, "grad_norm": 0.7834062576293945, "learning_rate": 5.722472073117657e-06, "loss": 0.0173, "step": 19783 }, { "epoch": 2.1448395490026018, "grad_norm": 0.30421924591064453, "learning_rate": 5.721746699550269e-06, "loss": 0.0097, "step": 19784 }, { "epoch": 2.1449479618386817, "grad_norm": 0.005893968977034092, "learning_rate": 5.721021325982882e-06, "loss": 0.0001, "step": 19785 }, { "epoch": 2.1450563746747613, "grad_norm": 0.35351186990737915, "learning_rate": 5.720295952415494e-06, "loss": 0.0009, "step": 19786 }, { "epoch": 2.1451647875108413, "grad_norm": 0.16041988134384155, "learning_rate": 5.719570578848107e-06, "loss": 0.0004, "step": 19787 }, { "epoch": 2.1452732003469213, "grad_norm": 0.022221721708774567, "learning_rate": 5.71884520528072e-06, "loss": 0.0002, "step": 19788 }, { "epoch": 2.145381613183001, "grad_norm": 4.386765480041504, "learning_rate": 5.718119831713332e-06, "loss": 0.0129, "step": 19789 }, { "epoch": 2.145490026019081, "grad_norm": 0.6107627749443054, "learning_rate": 5.717394458145946e-06, "loss": 0.0259, "step": 19790 }, { "epoch": 2.1455984388551603, "grad_norm": 3.544851064682007, "learning_rate": 5.716669084578559e-06, "loss": 0.0382, "step": 19791 }, { "epoch": 2.1457068516912403, "grad_norm": 4.098683834075928, "learning_rate": 5.715943711011171e-06, "loss": 0.0175, "step": 19792 }, { "epoch": 2.14581526452732, "grad_norm": 17.657766342163086, "learning_rate": 5.715218337443784e-06, "loss": 0.06, "step": 19793 }, { "epoch": 2.1459236773634, "grad_norm": 0.19571229815483093, "learning_rate": 5.714492963876397e-06, "loss": 0.0004, "step": 19794 }, { "epoch": 2.14603209019948, "grad_norm": 2.100773334503174, "learning_rate": 5.713767590309009e-06, "loss": 0.0344, "step": 19795 }, { "epoch": 2.1461405030355594, "grad_norm": 0.012498950585722923, "learning_rate": 5.713042216741622e-06, "loss": 0.0002, "step": 19796 }, { "epoch": 2.1462489158716394, "grad_norm": 2.1950907707214355, "learning_rate": 5.712316843174235e-06, "loss": 0.0226, "step": 19797 }, { "epoch": 2.146357328707719, "grad_norm": 4.542572498321533, "learning_rate": 5.711591469606847e-06, "loss": 0.0158, "step": 19798 }, { "epoch": 2.146465741543799, "grad_norm": 0.3831157684326172, "learning_rate": 5.710866096039461e-06, "loss": 0.0087, "step": 19799 }, { "epoch": 2.1465741543798784, "grad_norm": 1.708609700202942, "learning_rate": 5.710140722472074e-06, "loss": 0.0218, "step": 19800 }, { "epoch": 2.1466825672159584, "grad_norm": 0.38880184292793274, "learning_rate": 5.709415348904686e-06, "loss": 0.0099, "step": 19801 }, { "epoch": 2.146790980052038, "grad_norm": 0.01692240498960018, "learning_rate": 5.708689975337299e-06, "loss": 0.0001, "step": 19802 }, { "epoch": 2.146899392888118, "grad_norm": 0.06698164343833923, "learning_rate": 5.707964601769912e-06, "loss": 0.0005, "step": 19803 }, { "epoch": 2.147007805724198, "grad_norm": 4.041234493255615, "learning_rate": 5.7072392282025245e-06, "loss": 0.0377, "step": 19804 }, { "epoch": 2.1471162185602775, "grad_norm": 0.7178013324737549, "learning_rate": 5.7065138546351375e-06, "loss": 0.0072, "step": 19805 }, { "epoch": 2.1472246313963574, "grad_norm": 0.11457237601280212, "learning_rate": 5.70578848106775e-06, "loss": 0.0005, "step": 19806 }, { "epoch": 2.147333044232437, "grad_norm": 0.3629632890224457, "learning_rate": 5.705063107500363e-06, "loss": 0.0141, "step": 19807 }, { "epoch": 2.147441457068517, "grad_norm": 0.09150861203670502, "learning_rate": 5.7043377339329764e-06, "loss": 0.0004, "step": 19808 }, { "epoch": 2.1475498699045965, "grad_norm": 0.0063745733350515366, "learning_rate": 5.7036123603655894e-06, "loss": 0.0, "step": 19809 }, { "epoch": 2.1476582827406765, "grad_norm": 0.02386738918721676, "learning_rate": 5.702886986798202e-06, "loss": 0.0002, "step": 19810 }, { "epoch": 2.1477666955767565, "grad_norm": 0.2714862823486328, "learning_rate": 5.7021616132308146e-06, "loss": 0.0055, "step": 19811 }, { "epoch": 2.147875108412836, "grad_norm": 0.13374941051006317, "learning_rate": 5.701436239663427e-06, "loss": 0.0006, "step": 19812 }, { "epoch": 2.147983521248916, "grad_norm": 2.5049281120300293, "learning_rate": 5.70071086609604e-06, "loss": 0.0102, "step": 19813 }, { "epoch": 2.1480919340849955, "grad_norm": 1.1438307762145996, "learning_rate": 5.699985492528653e-06, "loss": 0.0034, "step": 19814 }, { "epoch": 2.1482003469210755, "grad_norm": 2.9231433868408203, "learning_rate": 5.699260118961265e-06, "loss": 0.0171, "step": 19815 }, { "epoch": 2.148308759757155, "grad_norm": 0.3670250475406647, "learning_rate": 5.698534745393879e-06, "loss": 0.0128, "step": 19816 }, { "epoch": 2.148417172593235, "grad_norm": 0.0074653420597314835, "learning_rate": 5.697809371826492e-06, "loss": 0.0001, "step": 19817 }, { "epoch": 2.148525585429315, "grad_norm": 2.6930203437805176, "learning_rate": 5.697083998259104e-06, "loss": 0.0048, "step": 19818 }, { "epoch": 2.1486339982653946, "grad_norm": 0.03091423027217388, "learning_rate": 5.696358624691717e-06, "loss": 0.0001, "step": 19819 }, { "epoch": 2.1487424111014746, "grad_norm": 0.031445760279893875, "learning_rate": 5.69563325112433e-06, "loss": 0.0001, "step": 19820 }, { "epoch": 2.148850823937554, "grad_norm": 1.953981876373291, "learning_rate": 5.694907877556942e-06, "loss": 0.0245, "step": 19821 }, { "epoch": 2.148959236773634, "grad_norm": 1.5047118663787842, "learning_rate": 5.694182503989555e-06, "loss": 0.031, "step": 19822 }, { "epoch": 2.1490676496097136, "grad_norm": 0.11753060668706894, "learning_rate": 5.693457130422168e-06, "loss": 0.0004, "step": 19823 }, { "epoch": 2.1491760624457936, "grad_norm": 0.02992234379053116, "learning_rate": 5.69273175685478e-06, "loss": 0.0002, "step": 19824 }, { "epoch": 2.149284475281873, "grad_norm": 0.0007727328920736909, "learning_rate": 5.692006383287394e-06, "loss": 0.0, "step": 19825 }, { "epoch": 2.149392888117953, "grad_norm": 4.429155349731445, "learning_rate": 5.691281009720007e-06, "loss": 0.0258, "step": 19826 }, { "epoch": 2.149501300954033, "grad_norm": 7.85593318939209, "learning_rate": 5.690555636152619e-06, "loss": 0.0489, "step": 19827 }, { "epoch": 2.1496097137901127, "grad_norm": 0.0029707076027989388, "learning_rate": 5.689830262585232e-06, "loss": 0.0, "step": 19828 }, { "epoch": 2.1497181266261927, "grad_norm": 0.04436967894434929, "learning_rate": 5.689104889017845e-06, "loss": 0.0004, "step": 19829 }, { "epoch": 2.149826539462272, "grad_norm": 0.005946733523160219, "learning_rate": 5.688379515450457e-06, "loss": 0.0, "step": 19830 }, { "epoch": 2.149934952298352, "grad_norm": 0.2974347770214081, "learning_rate": 5.68765414188307e-06, "loss": 0.0198, "step": 19831 }, { "epoch": 2.1500433651344317, "grad_norm": 1.0980799198150635, "learning_rate": 5.686928768315682e-06, "loss": 0.0024, "step": 19832 }, { "epoch": 2.1501517779705117, "grad_norm": 0.4525730311870575, "learning_rate": 5.686203394748295e-06, "loss": 0.0161, "step": 19833 }, { "epoch": 2.1502601908065917, "grad_norm": 0.036089807748794556, "learning_rate": 5.685478021180909e-06, "loss": 0.0001, "step": 19834 }, { "epoch": 2.1503686036426712, "grad_norm": 4.7272257804870605, "learning_rate": 5.684752647613522e-06, "loss": 0.0331, "step": 19835 }, { "epoch": 2.1504770164787512, "grad_norm": 0.6041103601455688, "learning_rate": 5.684027274046134e-06, "loss": 0.0197, "step": 19836 }, { "epoch": 2.1505854293148308, "grad_norm": 0.017016984522342682, "learning_rate": 5.683301900478747e-06, "loss": 0.0001, "step": 19837 }, { "epoch": 2.1506938421509108, "grad_norm": 0.5067773461341858, "learning_rate": 5.682576526911359e-06, "loss": 0.0096, "step": 19838 }, { "epoch": 2.1508022549869903, "grad_norm": 0.5150078535079956, "learning_rate": 5.681851153343972e-06, "loss": 0.0098, "step": 19839 }, { "epoch": 2.1509106678230703, "grad_norm": 3.124582529067993, "learning_rate": 5.681125779776585e-06, "loss": 0.0133, "step": 19840 }, { "epoch": 2.1510190806591503, "grad_norm": 1.3498765230178833, "learning_rate": 5.6804004062091975e-06, "loss": 0.0178, "step": 19841 }, { "epoch": 2.15112749349523, "grad_norm": 0.6707476377487183, "learning_rate": 5.679675032641811e-06, "loss": 0.0088, "step": 19842 }, { "epoch": 2.15123590633131, "grad_norm": 3.735339879989624, "learning_rate": 5.678949659074424e-06, "loss": 0.0284, "step": 19843 }, { "epoch": 2.1513443191673893, "grad_norm": 24.02906036376953, "learning_rate": 5.6782242855070365e-06, "loss": 0.0189, "step": 19844 }, { "epoch": 2.1514527320034693, "grad_norm": 0.04127297177910805, "learning_rate": 5.6774989119396495e-06, "loss": 0.0004, "step": 19845 }, { "epoch": 2.151561144839549, "grad_norm": 13.647181510925293, "learning_rate": 5.6767735383722625e-06, "loss": 0.0071, "step": 19846 }, { "epoch": 2.151669557675629, "grad_norm": 2.30971622467041, "learning_rate": 5.676048164804875e-06, "loss": 0.015, "step": 19847 }, { "epoch": 2.1517779705117084, "grad_norm": 0.5461514592170715, "learning_rate": 5.675322791237488e-06, "loss": 0.0109, "step": 19848 }, { "epoch": 2.1518863833477884, "grad_norm": 3.904449224472046, "learning_rate": 5.674597417670101e-06, "loss": 0.038, "step": 19849 }, { "epoch": 2.1519947961838684, "grad_norm": 0.21581530570983887, "learning_rate": 5.673872044102713e-06, "loss": 0.002, "step": 19850 }, { "epoch": 2.152103209019948, "grad_norm": 0.4078790843486786, "learning_rate": 5.6731466705353266e-06, "loss": 0.0155, "step": 19851 }, { "epoch": 2.152211621856028, "grad_norm": 0.28237539529800415, "learning_rate": 5.6724212969679396e-06, "loss": 0.006, "step": 19852 }, { "epoch": 2.1523200346921074, "grad_norm": 0.3078811466693878, "learning_rate": 5.671695923400552e-06, "loss": 0.0053, "step": 19853 }, { "epoch": 2.1524284475281874, "grad_norm": 4.925608158111572, "learning_rate": 5.670970549833165e-06, "loss": 0.0205, "step": 19854 }, { "epoch": 2.152536860364267, "grad_norm": 0.39780765771865845, "learning_rate": 5.670245176265778e-06, "loss": 0.0284, "step": 19855 }, { "epoch": 2.152645273200347, "grad_norm": 0.06531796604394913, "learning_rate": 5.66951980269839e-06, "loss": 0.0003, "step": 19856 }, { "epoch": 2.152753686036427, "grad_norm": 0.6979219913482666, "learning_rate": 5.668794429131003e-06, "loss": 0.0134, "step": 19857 }, { "epoch": 2.1528620988725065, "grad_norm": 0.002246657619252801, "learning_rate": 5.668069055563615e-06, "loss": 0.0, "step": 19858 }, { "epoch": 2.1529705117085864, "grad_norm": 0.022069094702601433, "learning_rate": 5.667343681996228e-06, "loss": 0.0001, "step": 19859 }, { "epoch": 2.153078924544666, "grad_norm": 14.77212142944336, "learning_rate": 5.666618308428842e-06, "loss": 0.0163, "step": 19860 }, { "epoch": 2.153187337380746, "grad_norm": 0.7007558345794678, "learning_rate": 5.665892934861455e-06, "loss": 0.0169, "step": 19861 }, { "epoch": 2.1532957502168255, "grad_norm": 0.3523569107055664, "learning_rate": 5.665167561294067e-06, "loss": 0.0075, "step": 19862 }, { "epoch": 2.1534041630529055, "grad_norm": 0.7819287180900574, "learning_rate": 5.66444218772668e-06, "loss": 0.0131, "step": 19863 }, { "epoch": 2.1535125758889855, "grad_norm": 3.0001814365386963, "learning_rate": 5.663716814159292e-06, "loss": 0.0416, "step": 19864 }, { "epoch": 2.153620988725065, "grad_norm": 6.483752727508545, "learning_rate": 5.662991440591905e-06, "loss": 0.0059, "step": 19865 }, { "epoch": 2.153729401561145, "grad_norm": 0.37991854548454285, "learning_rate": 5.662266067024518e-06, "loss": 0.0077, "step": 19866 }, { "epoch": 2.1538378143972245, "grad_norm": 9.485010147094727, "learning_rate": 5.66154069345713e-06, "loss": 0.0325, "step": 19867 }, { "epoch": 2.1539462272333045, "grad_norm": 1.283368468284607, "learning_rate": 5.660815319889744e-06, "loss": 0.0379, "step": 19868 }, { "epoch": 2.154054640069384, "grad_norm": 0.06336421519517899, "learning_rate": 5.660089946322357e-06, "loss": 0.0004, "step": 19869 }, { "epoch": 2.154163052905464, "grad_norm": 0.015211893245577812, "learning_rate": 5.659364572754969e-06, "loss": 0.0002, "step": 19870 }, { "epoch": 2.1542714657415436, "grad_norm": 1.6708507537841797, "learning_rate": 5.658639199187582e-06, "loss": 0.0024, "step": 19871 }, { "epoch": 2.1543798785776236, "grad_norm": 0.5367584824562073, "learning_rate": 5.657913825620195e-06, "loss": 0.0135, "step": 19872 }, { "epoch": 2.1544882914137036, "grad_norm": 5.452234745025635, "learning_rate": 5.657188452052807e-06, "loss": 0.0554, "step": 19873 }, { "epoch": 2.154596704249783, "grad_norm": 0.23104283213615417, "learning_rate": 5.65646307848542e-06, "loss": 0.0043, "step": 19874 }, { "epoch": 2.154705117085863, "grad_norm": 0.049779679626226425, "learning_rate": 5.655737704918033e-06, "loss": 0.0004, "step": 19875 }, { "epoch": 2.1548135299219426, "grad_norm": 0.008428255096077919, "learning_rate": 5.655012331350645e-06, "loss": 0.0001, "step": 19876 }, { "epoch": 2.1549219427580226, "grad_norm": 0.5571353435516357, "learning_rate": 5.654286957783259e-06, "loss": 0.0272, "step": 19877 }, { "epoch": 2.155030355594102, "grad_norm": 0.4717535078525543, "learning_rate": 5.653561584215872e-06, "loss": 0.023, "step": 19878 }, { "epoch": 2.155138768430182, "grad_norm": 8.040499687194824, "learning_rate": 5.652836210648484e-06, "loss": 0.0229, "step": 19879 }, { "epoch": 2.1552471812662617, "grad_norm": 3.499561071395874, "learning_rate": 5.652110837081097e-06, "loss": 0.0457, "step": 19880 }, { "epoch": 2.1553555941023417, "grad_norm": 9.762560844421387, "learning_rate": 5.65138546351371e-06, "loss": 0.0216, "step": 19881 }, { "epoch": 2.1554640069384217, "grad_norm": 0.3120262920856476, "learning_rate": 5.6506600899463225e-06, "loss": 0.0041, "step": 19882 }, { "epoch": 2.155572419774501, "grad_norm": 0.33059433102607727, "learning_rate": 5.6499347163789355e-06, "loss": 0.0085, "step": 19883 }, { "epoch": 2.155680832610581, "grad_norm": 0.12512534856796265, "learning_rate": 5.649209342811548e-06, "loss": 0.0007, "step": 19884 }, { "epoch": 2.1557892454466607, "grad_norm": 2.2240958213806152, "learning_rate": 5.6484839692441615e-06, "loss": 0.0065, "step": 19885 }, { "epoch": 2.1558976582827407, "grad_norm": 2.6766676902770996, "learning_rate": 5.6477585956767745e-06, "loss": 0.0095, "step": 19886 }, { "epoch": 2.1560060711188207, "grad_norm": 0.7412114143371582, "learning_rate": 5.6470332221093874e-06, "loss": 0.0097, "step": 19887 }, { "epoch": 2.1561144839549002, "grad_norm": 0.05988201126456261, "learning_rate": 5.646307848542e-06, "loss": 0.0005, "step": 19888 }, { "epoch": 2.1562228967909802, "grad_norm": 0.01652725785970688, "learning_rate": 5.6455824749746126e-06, "loss": 0.0001, "step": 19889 }, { "epoch": 2.1563313096270598, "grad_norm": 0.44035497307777405, "learning_rate": 5.644857101407225e-06, "loss": 0.0095, "step": 19890 }, { "epoch": 2.1564397224631398, "grad_norm": 0.7857059240341187, "learning_rate": 5.644131727839838e-06, "loss": 0.0128, "step": 19891 }, { "epoch": 2.1565481352992193, "grad_norm": 0.4965176284313202, "learning_rate": 5.643406354272451e-06, "loss": 0.0142, "step": 19892 }, { "epoch": 2.1566565481352993, "grad_norm": 0.09512735158205032, "learning_rate": 5.642680980705063e-06, "loss": 0.001, "step": 19893 }, { "epoch": 2.156764960971379, "grad_norm": 0.5304806232452393, "learning_rate": 5.641955607137677e-06, "loss": 0.0096, "step": 19894 }, { "epoch": 2.156873373807459, "grad_norm": 0.5398410558700562, "learning_rate": 5.64123023357029e-06, "loss": 0.0143, "step": 19895 }, { "epoch": 2.156981786643539, "grad_norm": 0.04513397440314293, "learning_rate": 5.640504860002902e-06, "loss": 0.0005, "step": 19896 }, { "epoch": 2.1570901994796183, "grad_norm": 0.026283258572220802, "learning_rate": 5.639779486435515e-06, "loss": 0.0002, "step": 19897 }, { "epoch": 2.1571986123156983, "grad_norm": 0.09921180456876755, "learning_rate": 5.639054112868128e-06, "loss": 0.0005, "step": 19898 }, { "epoch": 2.157307025151778, "grad_norm": 3.7497620582580566, "learning_rate": 5.63832873930074e-06, "loss": 0.0326, "step": 19899 }, { "epoch": 2.157415437987858, "grad_norm": 0.46145808696746826, "learning_rate": 5.637603365733353e-06, "loss": 0.0155, "step": 19900 }, { "epoch": 2.1575238508239374, "grad_norm": 0.7423365116119385, "learning_rate": 5.636877992165966e-06, "loss": 0.0332, "step": 19901 }, { "epoch": 2.1576322636600174, "grad_norm": 0.4604150652885437, "learning_rate": 5.636152618598578e-06, "loss": 0.0149, "step": 19902 }, { "epoch": 2.157740676496097, "grad_norm": 2.672034502029419, "learning_rate": 5.635427245031192e-06, "loss": 0.0096, "step": 19903 }, { "epoch": 2.157849089332177, "grad_norm": 0.3958072364330292, "learning_rate": 5.634701871463805e-06, "loss": 0.0073, "step": 19904 }, { "epoch": 2.157957502168257, "grad_norm": 0.7467373013496399, "learning_rate": 5.633976497896417e-06, "loss": 0.0108, "step": 19905 }, { "epoch": 2.1580659150043364, "grad_norm": 0.6164219379425049, "learning_rate": 5.63325112432903e-06, "loss": 0.0198, "step": 19906 }, { "epoch": 2.1581743278404164, "grad_norm": 0.5591186881065369, "learning_rate": 5.632525750761643e-06, "loss": 0.0172, "step": 19907 }, { "epoch": 2.158282740676496, "grad_norm": 3.6153156757354736, "learning_rate": 5.631800377194255e-06, "loss": 0.0751, "step": 19908 }, { "epoch": 2.158391153512576, "grad_norm": 0.9453498125076294, "learning_rate": 5.631075003626868e-06, "loss": 0.0009, "step": 19909 }, { "epoch": 2.1584995663486555, "grad_norm": 0.5181261897087097, "learning_rate": 5.63034963005948e-06, "loss": 0.0104, "step": 19910 }, { "epoch": 2.1586079791847355, "grad_norm": 0.23782643675804138, "learning_rate": 5.629624256492094e-06, "loss": 0.003, "step": 19911 }, { "epoch": 2.1587163920208154, "grad_norm": 0.11560646444559097, "learning_rate": 5.628898882924707e-06, "loss": 0.0003, "step": 19912 }, { "epoch": 2.158824804856895, "grad_norm": 0.6647229790687561, "learning_rate": 5.62817350935732e-06, "loss": 0.0132, "step": 19913 }, { "epoch": 2.158933217692975, "grad_norm": 0.027853913605213165, "learning_rate": 5.627448135789932e-06, "loss": 0.0002, "step": 19914 }, { "epoch": 2.1590416305290545, "grad_norm": 0.05692464858293533, "learning_rate": 5.626722762222545e-06, "loss": 0.0003, "step": 19915 }, { "epoch": 2.1591500433651345, "grad_norm": 0.4657498002052307, "learning_rate": 5.625997388655157e-06, "loss": 0.0061, "step": 19916 }, { "epoch": 2.159258456201214, "grad_norm": 0.3830072283744812, "learning_rate": 5.62527201508777e-06, "loss": 0.0126, "step": 19917 }, { "epoch": 2.159366869037294, "grad_norm": 0.29296645522117615, "learning_rate": 5.624546641520383e-06, "loss": 0.0063, "step": 19918 }, { "epoch": 2.159475281873374, "grad_norm": 0.46191805601119995, "learning_rate": 5.6238212679529955e-06, "loss": 0.0017, "step": 19919 }, { "epoch": 2.1595836947094535, "grad_norm": 1.7421761751174927, "learning_rate": 5.623095894385609e-06, "loss": 0.0055, "step": 19920 }, { "epoch": 2.1596921075455335, "grad_norm": 0.043418705463409424, "learning_rate": 5.622370520818222e-06, "loss": 0.0002, "step": 19921 }, { "epoch": 2.159800520381613, "grad_norm": 0.42599716782569885, "learning_rate": 5.6216451472508345e-06, "loss": 0.0005, "step": 19922 }, { "epoch": 2.159908933217693, "grad_norm": 0.06630189716815948, "learning_rate": 5.6209197736834475e-06, "loss": 0.0002, "step": 19923 }, { "epoch": 2.1600173460537726, "grad_norm": 0.32622864842414856, "learning_rate": 5.6201944001160605e-06, "loss": 0.0013, "step": 19924 }, { "epoch": 2.1601257588898526, "grad_norm": 0.4452314078807831, "learning_rate": 5.619469026548673e-06, "loss": 0.0191, "step": 19925 }, { "epoch": 2.160234171725932, "grad_norm": 0.9853234887123108, "learning_rate": 5.618743652981286e-06, "loss": 0.0064, "step": 19926 }, { "epoch": 2.160342584562012, "grad_norm": 0.3461754322052002, "learning_rate": 5.618018279413899e-06, "loss": 0.0058, "step": 19927 }, { "epoch": 2.160450997398092, "grad_norm": 0.7496806383132935, "learning_rate": 5.617292905846511e-06, "loss": 0.0208, "step": 19928 }, { "epoch": 2.1605594102341716, "grad_norm": 0.12381187081336975, "learning_rate": 5.6165675322791246e-06, "loss": 0.0004, "step": 19929 }, { "epoch": 2.1606678230702516, "grad_norm": 2.5761635303497314, "learning_rate": 5.6158421587117376e-06, "loss": 0.0167, "step": 19930 }, { "epoch": 2.160776235906331, "grad_norm": 1.0012948513031006, "learning_rate": 5.61511678514435e-06, "loss": 0.0209, "step": 19931 }, { "epoch": 2.160884648742411, "grad_norm": 0.9922166466712952, "learning_rate": 5.614391411576963e-06, "loss": 0.0042, "step": 19932 }, { "epoch": 2.1609930615784907, "grad_norm": 0.38578253984451294, "learning_rate": 5.613666038009576e-06, "loss": 0.0097, "step": 19933 }, { "epoch": 2.1611014744145707, "grad_norm": 3.2617859840393066, "learning_rate": 5.612940664442188e-06, "loss": 0.0413, "step": 19934 }, { "epoch": 2.1612098872506507, "grad_norm": 0.2770134210586548, "learning_rate": 5.612215290874801e-06, "loss": 0.0055, "step": 19935 }, { "epoch": 2.16131830008673, "grad_norm": 26.933521270751953, "learning_rate": 5.611489917307413e-06, "loss": 0.0089, "step": 19936 }, { "epoch": 2.16142671292281, "grad_norm": 2.90187406539917, "learning_rate": 5.610764543740027e-06, "loss": 0.034, "step": 19937 }, { "epoch": 2.1615351257588897, "grad_norm": 2.5450081825256348, "learning_rate": 5.61003917017264e-06, "loss": 0.0066, "step": 19938 }, { "epoch": 2.1616435385949697, "grad_norm": 0.7261947989463806, "learning_rate": 5.609313796605253e-06, "loss": 0.0201, "step": 19939 }, { "epoch": 2.1617519514310493, "grad_norm": 3.477560520172119, "learning_rate": 5.608588423037865e-06, "loss": 0.0802, "step": 19940 }, { "epoch": 2.1618603642671292, "grad_norm": 0.1052531823515892, "learning_rate": 5.607863049470478e-06, "loss": 0.0005, "step": 19941 }, { "epoch": 2.1619687771032092, "grad_norm": 0.2381080687046051, "learning_rate": 5.60713767590309e-06, "loss": 0.0048, "step": 19942 }, { "epoch": 2.1620771899392888, "grad_norm": 0.344615638256073, "learning_rate": 5.606412302335703e-06, "loss": 0.0111, "step": 19943 }, { "epoch": 2.1621856027753688, "grad_norm": 0.005659440997987986, "learning_rate": 5.605686928768316e-06, "loss": 0.0, "step": 19944 }, { "epoch": 2.1622940156114483, "grad_norm": 1.5771162509918213, "learning_rate": 5.604961555200928e-06, "loss": 0.0022, "step": 19945 }, { "epoch": 2.1624024284475283, "grad_norm": 0.27649885416030884, "learning_rate": 5.604236181633542e-06, "loss": 0.0055, "step": 19946 }, { "epoch": 2.162510841283608, "grad_norm": 0.023284822702407837, "learning_rate": 5.603510808066155e-06, "loss": 0.0001, "step": 19947 }, { "epoch": 2.162619254119688, "grad_norm": 0.32912132143974304, "learning_rate": 5.602785434498767e-06, "loss": 0.0112, "step": 19948 }, { "epoch": 2.1627276669557673, "grad_norm": 0.028720585629343987, "learning_rate": 5.60206006093138e-06, "loss": 0.0002, "step": 19949 }, { "epoch": 2.1628360797918473, "grad_norm": 0.40575939416885376, "learning_rate": 5.601334687363993e-06, "loss": 0.0011, "step": 19950 }, { "epoch": 2.1629444926279273, "grad_norm": 0.2999471426010132, "learning_rate": 5.600609313796605e-06, "loss": 0.0101, "step": 19951 }, { "epoch": 2.163052905464007, "grad_norm": 0.980527937412262, "learning_rate": 5.599883940229218e-06, "loss": 0.003, "step": 19952 }, { "epoch": 2.163161318300087, "grad_norm": 6.624985694885254, "learning_rate": 5.599158566661831e-06, "loss": 0.0409, "step": 19953 }, { "epoch": 2.1632697311361664, "grad_norm": 2.252124547958374, "learning_rate": 5.598433193094443e-06, "loss": 0.0061, "step": 19954 }, { "epoch": 2.1633781439722464, "grad_norm": 0.3544516861438751, "learning_rate": 5.597707819527057e-06, "loss": 0.0007, "step": 19955 }, { "epoch": 2.163486556808326, "grad_norm": 0.024835826829075813, "learning_rate": 5.59698244595967e-06, "loss": 0.0001, "step": 19956 }, { "epoch": 2.163594969644406, "grad_norm": 2.0102977752685547, "learning_rate": 5.596257072392282e-06, "loss": 0.0218, "step": 19957 }, { "epoch": 2.163703382480486, "grad_norm": 0.2125648409128189, "learning_rate": 5.595531698824895e-06, "loss": 0.0008, "step": 19958 }, { "epoch": 2.1638117953165654, "grad_norm": 0.28808706998825073, "learning_rate": 5.594806325257508e-06, "loss": 0.0013, "step": 19959 }, { "epoch": 2.1639202081526454, "grad_norm": 0.04555010423064232, "learning_rate": 5.5940809516901205e-06, "loss": 0.0003, "step": 19960 }, { "epoch": 2.164028620988725, "grad_norm": 3.9627492427825928, "learning_rate": 5.5933555781227335e-06, "loss": 0.0112, "step": 19961 }, { "epoch": 2.164137033824805, "grad_norm": 0.0404200404882431, "learning_rate": 5.592630204555346e-06, "loss": 0.0001, "step": 19962 }, { "epoch": 2.1642454466608845, "grad_norm": 0.7654624581336975, "learning_rate": 5.5919048309879595e-06, "loss": 0.0154, "step": 19963 }, { "epoch": 2.1643538594969645, "grad_norm": 2.0289080142974854, "learning_rate": 5.5911794574205725e-06, "loss": 0.0293, "step": 19964 }, { "epoch": 2.1644622723330444, "grad_norm": 10.270027160644531, "learning_rate": 5.5904540838531854e-06, "loss": 0.0352, "step": 19965 }, { "epoch": 2.164570685169124, "grad_norm": 2.7607102394104004, "learning_rate": 5.589728710285798e-06, "loss": 0.0035, "step": 19966 }, { "epoch": 2.164679098005204, "grad_norm": 0.3016752600669861, "learning_rate": 5.589003336718411e-06, "loss": 0.0066, "step": 19967 }, { "epoch": 2.1647875108412835, "grad_norm": 3.622978925704956, "learning_rate": 5.588277963151023e-06, "loss": 0.0072, "step": 19968 }, { "epoch": 2.1648959236773635, "grad_norm": 0.979893684387207, "learning_rate": 5.587552589583636e-06, "loss": 0.0127, "step": 19969 }, { "epoch": 2.165004336513443, "grad_norm": 4.177050590515137, "learning_rate": 5.586827216016249e-06, "loss": 0.0072, "step": 19970 }, { "epoch": 2.165112749349523, "grad_norm": 0.40960493683815, "learning_rate": 5.586101842448861e-06, "loss": 0.0165, "step": 19971 }, { "epoch": 2.1652211621856026, "grad_norm": 1.8587372303009033, "learning_rate": 5.585376468881475e-06, "loss": 0.0105, "step": 19972 }, { "epoch": 2.1653295750216826, "grad_norm": 10.191601753234863, "learning_rate": 5.584651095314088e-06, "loss": 0.0374, "step": 19973 }, { "epoch": 2.1654379878577625, "grad_norm": 1.4417933225631714, "learning_rate": 5.5839257217467e-06, "loss": 0.0055, "step": 19974 }, { "epoch": 2.165546400693842, "grad_norm": 4.564650535583496, "learning_rate": 5.583200348179313e-06, "loss": 0.0244, "step": 19975 }, { "epoch": 2.165654813529922, "grad_norm": 0.11553888767957687, "learning_rate": 5.582474974611926e-06, "loss": 0.0002, "step": 19976 }, { "epoch": 2.1657632263660016, "grad_norm": 0.3914635181427002, "learning_rate": 5.581749601044538e-06, "loss": 0.0134, "step": 19977 }, { "epoch": 2.1658716392020816, "grad_norm": 18.531904220581055, "learning_rate": 5.581024227477151e-06, "loss": 0.1004, "step": 19978 }, { "epoch": 2.165980052038161, "grad_norm": 3.1947593688964844, "learning_rate": 5.580298853909764e-06, "loss": 0.0144, "step": 19979 }, { "epoch": 2.166088464874241, "grad_norm": 0.050377700477838516, "learning_rate": 5.579573480342377e-06, "loss": 0.0002, "step": 19980 }, { "epoch": 2.166196877710321, "grad_norm": 0.15725702047348022, "learning_rate": 5.57884810677499e-06, "loss": 0.0004, "step": 19981 }, { "epoch": 2.1663052905464006, "grad_norm": 0.7912248969078064, "learning_rate": 5.578122733207603e-06, "loss": 0.0262, "step": 19982 }, { "epoch": 2.1664137033824806, "grad_norm": 0.47815871238708496, "learning_rate": 5.577397359640215e-06, "loss": 0.0128, "step": 19983 }, { "epoch": 2.16652211621856, "grad_norm": 0.3904455900192261, "learning_rate": 5.576671986072828e-06, "loss": 0.023, "step": 19984 }, { "epoch": 2.16663052905464, "grad_norm": 2.4174275398254395, "learning_rate": 5.575946612505441e-06, "loss": 0.0316, "step": 19985 }, { "epoch": 2.1667389418907197, "grad_norm": 0.31825900077819824, "learning_rate": 5.575221238938053e-06, "loss": 0.0085, "step": 19986 }, { "epoch": 2.1668473547267997, "grad_norm": 0.49989208579063416, "learning_rate": 5.574495865370666e-06, "loss": 0.0107, "step": 19987 }, { "epoch": 2.1669557675628797, "grad_norm": 0.015488605946302414, "learning_rate": 5.573770491803278e-06, "loss": 0.0001, "step": 19988 }, { "epoch": 2.167064180398959, "grad_norm": 2.1040701866149902, "learning_rate": 5.573045118235892e-06, "loss": 0.0664, "step": 19989 }, { "epoch": 2.167172593235039, "grad_norm": 0.2582719922065735, "learning_rate": 5.572319744668505e-06, "loss": 0.0053, "step": 19990 }, { "epoch": 2.1672810060711187, "grad_norm": 0.41403794288635254, "learning_rate": 5.571594371101118e-06, "loss": 0.0175, "step": 19991 }, { "epoch": 2.1673894189071987, "grad_norm": 3.1534016132354736, "learning_rate": 5.57086899753373e-06, "loss": 0.0294, "step": 19992 }, { "epoch": 2.1674978317432783, "grad_norm": 0.3211759626865387, "learning_rate": 5.570143623966343e-06, "loss": 0.0075, "step": 19993 }, { "epoch": 2.1676062445793582, "grad_norm": 0.08667892962694168, "learning_rate": 5.569418250398955e-06, "loss": 0.0002, "step": 19994 }, { "epoch": 2.167714657415438, "grad_norm": 0.003427034243941307, "learning_rate": 5.568692876831568e-06, "loss": 0.0, "step": 19995 }, { "epoch": 2.1678230702515178, "grad_norm": 0.8060097694396973, "learning_rate": 5.567967503264181e-06, "loss": 0.0023, "step": 19996 }, { "epoch": 2.1679314830875978, "grad_norm": 3.153965711593628, "learning_rate": 5.5672421296967935e-06, "loss": 0.0466, "step": 19997 }, { "epoch": 2.1680398959236773, "grad_norm": 0.3273395597934723, "learning_rate": 5.566516756129407e-06, "loss": 0.0139, "step": 19998 }, { "epoch": 2.1681483087597573, "grad_norm": 0.026068653911352158, "learning_rate": 5.56579138256202e-06, "loss": 0.0002, "step": 19999 }, { "epoch": 2.168256721595837, "grad_norm": 4.117761135101318, "learning_rate": 5.5650660089946325e-06, "loss": 0.0132, "step": 20000 }, { "epoch": 2.168365134431917, "grad_norm": 0.48431816697120667, "learning_rate": 5.5643406354272455e-06, "loss": 0.0009, "step": 20001 }, { "epoch": 2.1684735472679963, "grad_norm": 6.019316673278809, "learning_rate": 5.5636152618598585e-06, "loss": 0.0632, "step": 20002 }, { "epoch": 2.1685819601040763, "grad_norm": 1.4654486179351807, "learning_rate": 5.562889888292471e-06, "loss": 0.0366, "step": 20003 }, { "epoch": 2.1686903729401563, "grad_norm": 3.2159080505371094, "learning_rate": 5.562164514725084e-06, "loss": 0.0246, "step": 20004 }, { "epoch": 2.168798785776236, "grad_norm": 5.2976555824279785, "learning_rate": 5.561439141157697e-06, "loss": 0.0797, "step": 20005 }, { "epoch": 2.168907198612316, "grad_norm": 0.20190830528736115, "learning_rate": 5.5607137675903104e-06, "loss": 0.015, "step": 20006 }, { "epoch": 2.1690156114483954, "grad_norm": 0.4171167314052582, "learning_rate": 5.5599883940229226e-06, "loss": 0.0141, "step": 20007 }, { "epoch": 2.1691240242844754, "grad_norm": 0.03398554027080536, "learning_rate": 5.5592630204555356e-06, "loss": 0.0002, "step": 20008 }, { "epoch": 2.169232437120555, "grad_norm": 7.8478922843933105, "learning_rate": 5.558537646888148e-06, "loss": 0.0049, "step": 20009 }, { "epoch": 2.169340849956635, "grad_norm": 0.36636632680892944, "learning_rate": 5.557812273320761e-06, "loss": 0.0127, "step": 20010 }, { "epoch": 2.169449262792715, "grad_norm": 0.3025059998035431, "learning_rate": 5.557086899753374e-06, "loss": 0.0067, "step": 20011 }, { "epoch": 2.1695576756287944, "grad_norm": 2.468874931335449, "learning_rate": 5.556361526185986e-06, "loss": 0.0305, "step": 20012 }, { "epoch": 2.1696660884648744, "grad_norm": 1.3889080286026, "learning_rate": 5.555636152618599e-06, "loss": 0.0071, "step": 20013 }, { "epoch": 2.169774501300954, "grad_norm": 0.2801375985145569, "learning_rate": 5.554910779051211e-06, "loss": 0.0061, "step": 20014 }, { "epoch": 2.169882914137034, "grad_norm": 2.3194258213043213, "learning_rate": 5.554185405483825e-06, "loss": 0.0315, "step": 20015 }, { "epoch": 2.1699913269731135, "grad_norm": 0.04636982083320618, "learning_rate": 5.553460031916438e-06, "loss": 0.0001, "step": 20016 }, { "epoch": 2.1700997398091935, "grad_norm": 0.9734249114990234, "learning_rate": 5.552734658349051e-06, "loss": 0.011, "step": 20017 }, { "epoch": 2.170208152645273, "grad_norm": 0.29998552799224854, "learning_rate": 5.552009284781663e-06, "loss": 0.0099, "step": 20018 }, { "epoch": 2.170316565481353, "grad_norm": 0.06948786973953247, "learning_rate": 5.551283911214276e-06, "loss": 0.0005, "step": 20019 }, { "epoch": 2.170424978317433, "grad_norm": 0.035833586007356644, "learning_rate": 5.550558537646888e-06, "loss": 0.0003, "step": 20020 }, { "epoch": 2.1705333911535125, "grad_norm": 0.9680562019348145, "learning_rate": 5.549833164079501e-06, "loss": 0.0031, "step": 20021 }, { "epoch": 2.1706418039895925, "grad_norm": 5.038597106933594, "learning_rate": 5.549107790512114e-06, "loss": 0.0236, "step": 20022 }, { "epoch": 2.170750216825672, "grad_norm": 0.17206069827079773, "learning_rate": 5.548382416944726e-06, "loss": 0.0008, "step": 20023 }, { "epoch": 2.170858629661752, "grad_norm": 1.3532809019088745, "learning_rate": 5.54765704337734e-06, "loss": 0.0279, "step": 20024 }, { "epoch": 2.1709670424978316, "grad_norm": 0.700810432434082, "learning_rate": 5.546931669809953e-06, "loss": 0.0067, "step": 20025 }, { "epoch": 2.1710754553339116, "grad_norm": 0.4829959273338318, "learning_rate": 5.546206296242565e-06, "loss": 0.0233, "step": 20026 }, { "epoch": 2.171183868169991, "grad_norm": 5.635878562927246, "learning_rate": 5.545480922675178e-06, "loss": 0.0129, "step": 20027 }, { "epoch": 2.171292281006071, "grad_norm": 0.05008725821971893, "learning_rate": 5.544755549107791e-06, "loss": 0.0004, "step": 20028 }, { "epoch": 2.171400693842151, "grad_norm": 7.10658073425293, "learning_rate": 5.544030175540403e-06, "loss": 0.0086, "step": 20029 }, { "epoch": 2.1715091066782306, "grad_norm": 0.2807462215423584, "learning_rate": 5.543304801973016e-06, "loss": 0.0064, "step": 20030 }, { "epoch": 2.1716175195143106, "grad_norm": 2.9479663372039795, "learning_rate": 5.542579428405629e-06, "loss": 0.0213, "step": 20031 }, { "epoch": 2.17172593235039, "grad_norm": 1.1711695194244385, "learning_rate": 5.541854054838243e-06, "loss": 0.0021, "step": 20032 }, { "epoch": 2.17183434518647, "grad_norm": 3.8040122985839844, "learning_rate": 5.541128681270855e-06, "loss": 0.012, "step": 20033 }, { "epoch": 2.17194275802255, "grad_norm": 1.8488941192626953, "learning_rate": 5.540403307703468e-06, "loss": 0.0532, "step": 20034 }, { "epoch": 2.1720511708586296, "grad_norm": 1.7357381582260132, "learning_rate": 5.53967793413608e-06, "loss": 0.0521, "step": 20035 }, { "epoch": 2.1721595836947096, "grad_norm": 0.005493017379194498, "learning_rate": 5.538952560568693e-06, "loss": 0.0001, "step": 20036 }, { "epoch": 2.172267996530789, "grad_norm": 8.061955451965332, "learning_rate": 5.538227187001306e-06, "loss": 0.0174, "step": 20037 }, { "epoch": 2.172376409366869, "grad_norm": 0.35261693596839905, "learning_rate": 5.5375018134339185e-06, "loss": 0.0117, "step": 20038 }, { "epoch": 2.1724848222029487, "grad_norm": 0.3170216381549835, "learning_rate": 5.5367764398665315e-06, "loss": 0.0111, "step": 20039 }, { "epoch": 2.1725932350390287, "grad_norm": 13.638904571533203, "learning_rate": 5.536051066299144e-06, "loss": 0.0466, "step": 20040 }, { "epoch": 2.172701647875108, "grad_norm": 0.26959189772605896, "learning_rate": 5.5353256927317575e-06, "loss": 0.0056, "step": 20041 }, { "epoch": 2.172810060711188, "grad_norm": 0.5520949959754944, "learning_rate": 5.5346003191643705e-06, "loss": 0.0119, "step": 20042 }, { "epoch": 2.172918473547268, "grad_norm": 0.3268657922744751, "learning_rate": 5.5338749455969835e-06, "loss": 0.0102, "step": 20043 }, { "epoch": 2.1730268863833477, "grad_norm": 1.5305850505828857, "learning_rate": 5.533149572029596e-06, "loss": 0.0063, "step": 20044 }, { "epoch": 2.1731352992194277, "grad_norm": 0.4629986882209778, "learning_rate": 5.532424198462209e-06, "loss": 0.0096, "step": 20045 }, { "epoch": 2.1732437120555073, "grad_norm": 0.008067634887993336, "learning_rate": 5.531698824894821e-06, "loss": 0.0001, "step": 20046 }, { "epoch": 2.1733521248915872, "grad_norm": 7.92170524597168, "learning_rate": 5.530973451327434e-06, "loss": 0.0233, "step": 20047 }, { "epoch": 2.173460537727667, "grad_norm": 4.698319435119629, "learning_rate": 5.530248077760047e-06, "loss": 0.0326, "step": 20048 }, { "epoch": 2.1735689505637468, "grad_norm": 0.07554815709590912, "learning_rate": 5.529522704192659e-06, "loss": 0.0007, "step": 20049 }, { "epoch": 2.1736773633998263, "grad_norm": 9.869378089904785, "learning_rate": 5.528797330625273e-06, "loss": 0.031, "step": 20050 }, { "epoch": 2.1737857762359063, "grad_norm": 0.18273642659187317, "learning_rate": 5.528071957057886e-06, "loss": 0.0106, "step": 20051 }, { "epoch": 2.1738941890719863, "grad_norm": 0.20723609626293182, "learning_rate": 5.527346583490498e-06, "loss": 0.0006, "step": 20052 }, { "epoch": 2.174002601908066, "grad_norm": 0.3447647988796234, "learning_rate": 5.526621209923111e-06, "loss": 0.0113, "step": 20053 }, { "epoch": 2.174111014744146, "grad_norm": 9.549973487854004, "learning_rate": 5.525895836355724e-06, "loss": 0.0327, "step": 20054 }, { "epoch": 2.1742194275802254, "grad_norm": 0.07938386499881744, "learning_rate": 5.525170462788336e-06, "loss": 0.0004, "step": 20055 }, { "epoch": 2.1743278404163053, "grad_norm": 0.2332126647233963, "learning_rate": 5.524445089220949e-06, "loss": 0.0053, "step": 20056 }, { "epoch": 2.174436253252385, "grad_norm": 6.580618858337402, "learning_rate": 5.523719715653562e-06, "loss": 0.0147, "step": 20057 }, { "epoch": 2.174544666088465, "grad_norm": 1.3123104572296143, "learning_rate": 5.522994342086176e-06, "loss": 0.0084, "step": 20058 }, { "epoch": 2.174653078924545, "grad_norm": 0.38040730357170105, "learning_rate": 5.522268968518788e-06, "loss": 0.0106, "step": 20059 }, { "epoch": 2.1747614917606244, "grad_norm": 5.4182233810424805, "learning_rate": 5.521543594951401e-06, "loss": 0.0449, "step": 20060 }, { "epoch": 2.1748699045967044, "grad_norm": 0.4621778726577759, "learning_rate": 5.520818221384013e-06, "loss": 0.0184, "step": 20061 }, { "epoch": 2.174978317432784, "grad_norm": 0.9774367809295654, "learning_rate": 5.520092847816626e-06, "loss": 0.0252, "step": 20062 }, { "epoch": 2.175086730268864, "grad_norm": 0.10417844355106354, "learning_rate": 5.519367474249239e-06, "loss": 0.001, "step": 20063 }, { "epoch": 2.1751951431049434, "grad_norm": 0.5216735601425171, "learning_rate": 5.518642100681851e-06, "loss": 0.0116, "step": 20064 }, { "epoch": 2.1753035559410234, "grad_norm": 0.3859753906726837, "learning_rate": 5.517916727114464e-06, "loss": 0.0196, "step": 20065 }, { "epoch": 2.1754119687771034, "grad_norm": 1.1028836965560913, "learning_rate": 5.517191353547076e-06, "loss": 0.0226, "step": 20066 }, { "epoch": 2.175520381613183, "grad_norm": 4.994176864624023, "learning_rate": 5.51646597997969e-06, "loss": 0.0283, "step": 20067 }, { "epoch": 2.175628794449263, "grad_norm": 1.4775190353393555, "learning_rate": 5.515740606412303e-06, "loss": 0.0056, "step": 20068 }, { "epoch": 2.1757372072853425, "grad_norm": 0.2675562798976898, "learning_rate": 5.515015232844916e-06, "loss": 0.0055, "step": 20069 }, { "epoch": 2.1758456201214225, "grad_norm": 2.9432449340820312, "learning_rate": 5.514289859277528e-06, "loss": 0.0151, "step": 20070 }, { "epoch": 2.175954032957502, "grad_norm": 0.19791743159294128, "learning_rate": 5.513564485710141e-06, "loss": 0.0009, "step": 20071 }, { "epoch": 2.176062445793582, "grad_norm": 2.1737117767333984, "learning_rate": 5.512839112142753e-06, "loss": 0.0113, "step": 20072 }, { "epoch": 2.1761708586296615, "grad_norm": 0.244228795170784, "learning_rate": 5.512113738575366e-06, "loss": 0.0027, "step": 20073 }, { "epoch": 2.1762792714657415, "grad_norm": 8.748398780822754, "learning_rate": 5.511388365007979e-06, "loss": 0.0108, "step": 20074 }, { "epoch": 2.1763876843018215, "grad_norm": 0.05089885741472244, "learning_rate": 5.5106629914405915e-06, "loss": 0.0004, "step": 20075 }, { "epoch": 2.176496097137901, "grad_norm": 0.4152260720729828, "learning_rate": 5.509937617873205e-06, "loss": 0.0064, "step": 20076 }, { "epoch": 2.176604509973981, "grad_norm": 0.19271469116210938, "learning_rate": 5.509212244305818e-06, "loss": 0.0006, "step": 20077 }, { "epoch": 2.1767129228100606, "grad_norm": 13.093521118164062, "learning_rate": 5.5084868707384305e-06, "loss": 0.0119, "step": 20078 }, { "epoch": 2.1768213356461406, "grad_norm": 0.19602185487747192, "learning_rate": 5.5077614971710435e-06, "loss": 0.001, "step": 20079 }, { "epoch": 2.17692974848222, "grad_norm": 0.2785298228263855, "learning_rate": 5.5070361236036565e-06, "loss": 0.0061, "step": 20080 }, { "epoch": 2.1770381613183, "grad_norm": 1.9365243911743164, "learning_rate": 5.506310750036269e-06, "loss": 0.0142, "step": 20081 }, { "epoch": 2.17714657415438, "grad_norm": 0.5252853631973267, "learning_rate": 5.505585376468882e-06, "loss": 0.0166, "step": 20082 }, { "epoch": 2.1772549869904596, "grad_norm": 0.2415846735239029, "learning_rate": 5.504860002901495e-06, "loss": 0.0105, "step": 20083 }, { "epoch": 2.1773633998265396, "grad_norm": 0.4386086165904999, "learning_rate": 5.5041346293341084e-06, "loss": 0.0117, "step": 20084 }, { "epoch": 2.177471812662619, "grad_norm": 3.349133014678955, "learning_rate": 5.503409255766721e-06, "loss": 0.0426, "step": 20085 }, { "epoch": 2.177580225498699, "grad_norm": 0.31517863273620605, "learning_rate": 5.5026838821993336e-06, "loss": 0.0207, "step": 20086 }, { "epoch": 2.1776886383347787, "grad_norm": 0.48625484108924866, "learning_rate": 5.501958508631946e-06, "loss": 0.0131, "step": 20087 }, { "epoch": 2.1777970511708586, "grad_norm": 1.9367557764053345, "learning_rate": 5.501233135064559e-06, "loss": 0.0665, "step": 20088 }, { "epoch": 2.1779054640069386, "grad_norm": 0.2539334297180176, "learning_rate": 5.500507761497172e-06, "loss": 0.0058, "step": 20089 }, { "epoch": 2.178013876843018, "grad_norm": 0.20950515568256378, "learning_rate": 5.499782387929784e-06, "loss": 0.0002, "step": 20090 }, { "epoch": 2.178122289679098, "grad_norm": 0.28698664903640747, "learning_rate": 5.499057014362397e-06, "loss": 0.0109, "step": 20091 }, { "epoch": 2.1782307025151777, "grad_norm": 10.218147277832031, "learning_rate": 5.498331640795009e-06, "loss": 0.043, "step": 20092 }, { "epoch": 2.1783391153512577, "grad_norm": 6.289623737335205, "learning_rate": 5.497606267227623e-06, "loss": 0.0384, "step": 20093 }, { "epoch": 2.1784475281873372, "grad_norm": 12.207889556884766, "learning_rate": 5.496880893660236e-06, "loss": 0.0282, "step": 20094 }, { "epoch": 2.178555941023417, "grad_norm": 3.207000255584717, "learning_rate": 5.496155520092849e-06, "loss": 0.0428, "step": 20095 }, { "epoch": 2.1786643538594967, "grad_norm": 0.44213852286338806, "learning_rate": 5.495430146525461e-06, "loss": 0.0176, "step": 20096 }, { "epoch": 2.1787727666955767, "grad_norm": 0.648968517780304, "learning_rate": 5.494704772958074e-06, "loss": 0.0061, "step": 20097 }, { "epoch": 2.1788811795316567, "grad_norm": 1.4189023971557617, "learning_rate": 5.493979399390686e-06, "loss": 0.0403, "step": 20098 }, { "epoch": 2.1789895923677363, "grad_norm": 0.40957579016685486, "learning_rate": 5.493254025823299e-06, "loss": 0.0117, "step": 20099 }, { "epoch": 2.1790980052038162, "grad_norm": 0.12760378420352936, "learning_rate": 5.492528652255912e-06, "loss": 0.0009, "step": 20100 }, { "epoch": 2.179206418039896, "grad_norm": 10.306217193603516, "learning_rate": 5.491803278688526e-06, "loss": 0.0378, "step": 20101 }, { "epoch": 2.1793148308759758, "grad_norm": 0.5233250856399536, "learning_rate": 5.491077905121138e-06, "loss": 0.0073, "step": 20102 }, { "epoch": 2.1794232437120553, "grad_norm": 0.18138465285301208, "learning_rate": 5.490352531553751e-06, "loss": 0.0009, "step": 20103 }, { "epoch": 2.1795316565481353, "grad_norm": 0.05384506285190582, "learning_rate": 5.489627157986363e-06, "loss": 0.0007, "step": 20104 }, { "epoch": 2.1796400693842153, "grad_norm": 1.3717339038848877, "learning_rate": 5.488901784418976e-06, "loss": 0.053, "step": 20105 }, { "epoch": 2.179748482220295, "grad_norm": 0.03946724161505699, "learning_rate": 5.488176410851589e-06, "loss": 0.0004, "step": 20106 }, { "epoch": 2.179856895056375, "grad_norm": 0.578273594379425, "learning_rate": 5.487451037284201e-06, "loss": 0.0134, "step": 20107 }, { "epoch": 2.1799653078924544, "grad_norm": 7.671894550323486, "learning_rate": 5.486725663716814e-06, "loss": 0.03, "step": 20108 }, { "epoch": 2.1800737207285343, "grad_norm": 0.849594235420227, "learning_rate": 5.486000290149427e-06, "loss": 0.0043, "step": 20109 }, { "epoch": 2.180182133564614, "grad_norm": 0.311251699924469, "learning_rate": 5.485274916582041e-06, "loss": 0.0009, "step": 20110 }, { "epoch": 2.180290546400694, "grad_norm": 2.875084161758423, "learning_rate": 5.484549543014653e-06, "loss": 0.0141, "step": 20111 }, { "epoch": 2.180398959236774, "grad_norm": 0.15968436002731323, "learning_rate": 5.483824169447266e-06, "loss": 0.0009, "step": 20112 }, { "epoch": 2.1805073720728534, "grad_norm": 0.08500535041093826, "learning_rate": 5.483098795879878e-06, "loss": 0.0002, "step": 20113 }, { "epoch": 2.1806157849089334, "grad_norm": 0.36673423647880554, "learning_rate": 5.482373422312491e-06, "loss": 0.0067, "step": 20114 }, { "epoch": 2.180724197745013, "grad_norm": 0.4110293686389923, "learning_rate": 5.481648048745104e-06, "loss": 0.0196, "step": 20115 }, { "epoch": 2.180832610581093, "grad_norm": 4.799147129058838, "learning_rate": 5.4809226751777165e-06, "loss": 0.0141, "step": 20116 }, { "epoch": 2.1809410234171724, "grad_norm": 0.47127753496170044, "learning_rate": 5.4801973016103295e-06, "loss": 0.013, "step": 20117 }, { "epoch": 2.1810494362532524, "grad_norm": 3.6696412563323975, "learning_rate": 5.479471928042942e-06, "loss": 0.0571, "step": 20118 }, { "epoch": 2.181157849089332, "grad_norm": 1.5652666091918945, "learning_rate": 5.4787465544755555e-06, "loss": 0.0102, "step": 20119 }, { "epoch": 2.181266261925412, "grad_norm": 4.384006023406982, "learning_rate": 5.4780211809081685e-06, "loss": 0.0067, "step": 20120 }, { "epoch": 2.181374674761492, "grad_norm": 0.03617612645030022, "learning_rate": 5.4772958073407815e-06, "loss": 0.0003, "step": 20121 }, { "epoch": 2.1814830875975715, "grad_norm": 1.1967451572418213, "learning_rate": 5.476570433773394e-06, "loss": 0.0036, "step": 20122 }, { "epoch": 2.1815915004336515, "grad_norm": 0.12411791831254959, "learning_rate": 5.475845060206007e-06, "loss": 0.0007, "step": 20123 }, { "epoch": 2.181699913269731, "grad_norm": 2.342869758605957, "learning_rate": 5.475119686638619e-06, "loss": 0.0413, "step": 20124 }, { "epoch": 2.181808326105811, "grad_norm": 6.252336025238037, "learning_rate": 5.474394313071232e-06, "loss": 0.0281, "step": 20125 }, { "epoch": 2.1819167389418905, "grad_norm": 4.355550289154053, "learning_rate": 5.473668939503845e-06, "loss": 0.0297, "step": 20126 }, { "epoch": 2.1820251517779705, "grad_norm": 3.1005284786224365, "learning_rate": 5.4729435659364586e-06, "loss": 0.0083, "step": 20127 }, { "epoch": 2.1821335646140505, "grad_norm": 0.275631844997406, "learning_rate": 5.472218192369071e-06, "loss": 0.0022, "step": 20128 }, { "epoch": 2.18224197745013, "grad_norm": 0.09633314609527588, "learning_rate": 5.471492818801684e-06, "loss": 0.0007, "step": 20129 }, { "epoch": 2.18235039028621, "grad_norm": 0.10183578729629517, "learning_rate": 5.470767445234297e-06, "loss": 0.0009, "step": 20130 }, { "epoch": 2.1824588031222896, "grad_norm": 0.02536061964929104, "learning_rate": 5.470042071666909e-06, "loss": 0.0003, "step": 20131 }, { "epoch": 2.1825672159583696, "grad_norm": 0.28489503264427185, "learning_rate": 5.469316698099522e-06, "loss": 0.0063, "step": 20132 }, { "epoch": 2.182675628794449, "grad_norm": 2.297903299331665, "learning_rate": 5.468591324532134e-06, "loss": 0.0225, "step": 20133 }, { "epoch": 2.182784041630529, "grad_norm": 5.46718168258667, "learning_rate": 5.467865950964747e-06, "loss": 0.0104, "step": 20134 }, { "epoch": 2.182892454466609, "grad_norm": 0.5642750859260559, "learning_rate": 5.46714057739736e-06, "loss": 0.0084, "step": 20135 }, { "epoch": 2.1830008673026886, "grad_norm": 0.2980636954307556, "learning_rate": 5.466415203829974e-06, "loss": 0.0065, "step": 20136 }, { "epoch": 2.1831092801387686, "grad_norm": 0.05106163024902344, "learning_rate": 5.465689830262586e-06, "loss": 0.0003, "step": 20137 }, { "epoch": 2.183217692974848, "grad_norm": 0.5667401552200317, "learning_rate": 5.464964456695199e-06, "loss": 0.0164, "step": 20138 }, { "epoch": 2.183326105810928, "grad_norm": 3.9652647972106934, "learning_rate": 5.464239083127811e-06, "loss": 0.0183, "step": 20139 }, { "epoch": 2.1834345186470077, "grad_norm": 0.5866153836250305, "learning_rate": 5.463513709560424e-06, "loss": 0.0109, "step": 20140 }, { "epoch": 2.1835429314830876, "grad_norm": 0.5360987782478333, "learning_rate": 5.462788335993037e-06, "loss": 0.0177, "step": 20141 }, { "epoch": 2.183651344319167, "grad_norm": 2.1634514331817627, "learning_rate": 5.462062962425649e-06, "loss": 0.0677, "step": 20142 }, { "epoch": 2.183759757155247, "grad_norm": 3.925783395767212, "learning_rate": 5.461337588858262e-06, "loss": 0.0085, "step": 20143 }, { "epoch": 2.183868169991327, "grad_norm": 0.28206348419189453, "learning_rate": 5.460612215290874e-06, "loss": 0.0064, "step": 20144 }, { "epoch": 2.1839765828274067, "grad_norm": 0.4395803213119507, "learning_rate": 5.459886841723488e-06, "loss": 0.0119, "step": 20145 }, { "epoch": 2.1840849956634867, "grad_norm": 0.2633824646472931, "learning_rate": 5.459161468156101e-06, "loss": 0.0055, "step": 20146 }, { "epoch": 2.1841934084995662, "grad_norm": 1.9815864562988281, "learning_rate": 5.458436094588714e-06, "loss": 0.0051, "step": 20147 }, { "epoch": 2.184301821335646, "grad_norm": 0.03244728222489357, "learning_rate": 5.457710721021326e-06, "loss": 0.0003, "step": 20148 }, { "epoch": 2.1844102341717258, "grad_norm": 0.29448777437210083, "learning_rate": 5.456985347453939e-06, "loss": 0.0103, "step": 20149 }, { "epoch": 2.1845186470078057, "grad_norm": 0.36444151401519775, "learning_rate": 5.456259973886551e-06, "loss": 0.0008, "step": 20150 }, { "epoch": 2.1846270598438857, "grad_norm": 0.1428447663784027, "learning_rate": 5.455534600319164e-06, "loss": 0.0011, "step": 20151 }, { "epoch": 2.1847354726799653, "grad_norm": 0.5368098616600037, "learning_rate": 5.454809226751777e-06, "loss": 0.0165, "step": 20152 }, { "epoch": 2.1848438855160452, "grad_norm": 1.2934861183166504, "learning_rate": 5.454083853184391e-06, "loss": 0.0012, "step": 20153 }, { "epoch": 2.184952298352125, "grad_norm": 0.1729782372713089, "learning_rate": 5.453358479617003e-06, "loss": 0.0009, "step": 20154 }, { "epoch": 2.1850607111882048, "grad_norm": 3.7233715057373047, "learning_rate": 5.452633106049616e-06, "loss": 0.0349, "step": 20155 }, { "epoch": 2.1851691240242843, "grad_norm": 0.4266663193702698, "learning_rate": 5.451907732482229e-06, "loss": 0.0063, "step": 20156 }, { "epoch": 2.1852775368603643, "grad_norm": 0.5548827052116394, "learning_rate": 5.4511823589148415e-06, "loss": 0.0335, "step": 20157 }, { "epoch": 2.1853859496964443, "grad_norm": 0.5319133996963501, "learning_rate": 5.4504569853474545e-06, "loss": 0.0179, "step": 20158 }, { "epoch": 2.185494362532524, "grad_norm": 0.20355355739593506, "learning_rate": 5.449731611780067e-06, "loss": 0.0013, "step": 20159 }, { "epoch": 2.185602775368604, "grad_norm": 1.0232099294662476, "learning_rate": 5.44900623821268e-06, "loss": 0.0246, "step": 20160 }, { "epoch": 2.1857111882046834, "grad_norm": 0.048397839069366455, "learning_rate": 5.448280864645293e-06, "loss": 0.0004, "step": 20161 }, { "epoch": 2.1858196010407633, "grad_norm": 0.781857967376709, "learning_rate": 5.4475554910779064e-06, "loss": 0.0237, "step": 20162 }, { "epoch": 2.185928013876843, "grad_norm": 1.5020097494125366, "learning_rate": 5.446830117510519e-06, "loss": 0.0173, "step": 20163 }, { "epoch": 2.186036426712923, "grad_norm": 1.5340163707733154, "learning_rate": 5.446104743943132e-06, "loss": 0.0128, "step": 20164 }, { "epoch": 2.1861448395490024, "grad_norm": 0.024238357320427895, "learning_rate": 5.445379370375744e-06, "loss": 0.0002, "step": 20165 }, { "epoch": 2.1862532523850824, "grad_norm": 0.35872510075569153, "learning_rate": 5.444653996808357e-06, "loss": 0.0207, "step": 20166 }, { "epoch": 2.1863616652211624, "grad_norm": 0.32980427145957947, "learning_rate": 5.44392862324097e-06, "loss": 0.0117, "step": 20167 }, { "epoch": 2.186470078057242, "grad_norm": 0.31471315026283264, "learning_rate": 5.443203249673582e-06, "loss": 0.0058, "step": 20168 }, { "epoch": 2.186578490893322, "grad_norm": 0.585372269153595, "learning_rate": 5.442477876106195e-06, "loss": 0.0018, "step": 20169 }, { "epoch": 2.1866869037294014, "grad_norm": 0.008927423506975174, "learning_rate": 5.441752502538807e-06, "loss": 0.0001, "step": 20170 }, { "epoch": 2.1867953165654814, "grad_norm": 0.016851508989930153, "learning_rate": 5.441027128971421e-06, "loss": 0.0001, "step": 20171 }, { "epoch": 2.186903729401561, "grad_norm": 0.38701891899108887, "learning_rate": 5.440301755404034e-06, "loss": 0.0101, "step": 20172 }, { "epoch": 2.187012142237641, "grad_norm": 0.22638028860092163, "learning_rate": 5.439576381836647e-06, "loss": 0.0008, "step": 20173 }, { "epoch": 2.1871205550737205, "grad_norm": 0.3557416498661041, "learning_rate": 5.438851008269259e-06, "loss": 0.008, "step": 20174 }, { "epoch": 2.1872289679098005, "grad_norm": 4.71360445022583, "learning_rate": 5.438125634701872e-06, "loss": 0.0248, "step": 20175 }, { "epoch": 2.1873373807458805, "grad_norm": 0.01676151715219021, "learning_rate": 5.437400261134484e-06, "loss": 0.0002, "step": 20176 }, { "epoch": 2.18744579358196, "grad_norm": 0.031200923025608063, "learning_rate": 5.436674887567097e-06, "loss": 0.0002, "step": 20177 }, { "epoch": 2.18755420641804, "grad_norm": 0.015199653804302216, "learning_rate": 5.43594951399971e-06, "loss": 0.0001, "step": 20178 }, { "epoch": 2.1876626192541195, "grad_norm": 0.29884228110313416, "learning_rate": 5.435224140432324e-06, "loss": 0.0103, "step": 20179 }, { "epoch": 2.1877710320901995, "grad_norm": 0.29977431893348694, "learning_rate": 5.434498766864936e-06, "loss": 0.0105, "step": 20180 }, { "epoch": 2.1878794449262795, "grad_norm": 0.38442686200141907, "learning_rate": 5.433773393297549e-06, "loss": 0.0111, "step": 20181 }, { "epoch": 2.187987857762359, "grad_norm": 0.28364551067352295, "learning_rate": 5.433048019730162e-06, "loss": 0.0066, "step": 20182 }, { "epoch": 2.188096270598439, "grad_norm": 1.927245855331421, "learning_rate": 5.432322646162774e-06, "loss": 0.0023, "step": 20183 }, { "epoch": 2.1882046834345186, "grad_norm": 1.0742055177688599, "learning_rate": 5.431597272595387e-06, "loss": 0.0018, "step": 20184 }, { "epoch": 2.1883130962705986, "grad_norm": 3.437962055206299, "learning_rate": 5.430871899027999e-06, "loss": 0.0327, "step": 20185 }, { "epoch": 2.188421509106678, "grad_norm": 6.618170261383057, "learning_rate": 5.430146525460612e-06, "loss": 0.009, "step": 20186 }, { "epoch": 2.188529921942758, "grad_norm": 3.5364627838134766, "learning_rate": 5.429421151893225e-06, "loss": 0.0327, "step": 20187 }, { "epoch": 2.1886383347788376, "grad_norm": 0.49499496817588806, "learning_rate": 5.428695778325839e-06, "loss": 0.0008, "step": 20188 }, { "epoch": 2.1887467476149176, "grad_norm": 0.4186769127845764, "learning_rate": 5.427970404758451e-06, "loss": 0.0014, "step": 20189 }, { "epoch": 2.1888551604509976, "grad_norm": 0.24956361949443817, "learning_rate": 5.427245031191064e-06, "loss": 0.0174, "step": 20190 }, { "epoch": 2.188963573287077, "grad_norm": 0.4202609062194824, "learning_rate": 5.426519657623676e-06, "loss": 0.0166, "step": 20191 }, { "epoch": 2.189071986123157, "grad_norm": 0.11993635445833206, "learning_rate": 5.425794284056289e-06, "loss": 0.0006, "step": 20192 }, { "epoch": 2.1891803989592367, "grad_norm": 0.36849215626716614, "learning_rate": 5.425068910488902e-06, "loss": 0.0198, "step": 20193 }, { "epoch": 2.1892888117953166, "grad_norm": 0.33600497245788574, "learning_rate": 5.4243435369215145e-06, "loss": 0.0126, "step": 20194 }, { "epoch": 2.189397224631396, "grad_norm": 0.7375635504722595, "learning_rate": 5.4236181633541275e-06, "loss": 0.0152, "step": 20195 }, { "epoch": 2.189505637467476, "grad_norm": 1.43595290184021, "learning_rate": 5.422892789786741e-06, "loss": 0.0086, "step": 20196 }, { "epoch": 2.1896140503035557, "grad_norm": 0.46967804431915283, "learning_rate": 5.4221674162193535e-06, "loss": 0.0113, "step": 20197 }, { "epoch": 2.1897224631396357, "grad_norm": 0.09270527213811874, "learning_rate": 5.4214420426519665e-06, "loss": 0.0008, "step": 20198 }, { "epoch": 2.1898308759757157, "grad_norm": 0.019393540918827057, "learning_rate": 5.4207166690845795e-06, "loss": 0.0002, "step": 20199 }, { "epoch": 2.1899392888117952, "grad_norm": 0.030205918475985527, "learning_rate": 5.419991295517192e-06, "loss": 0.0001, "step": 20200 }, { "epoch": 2.190047701647875, "grad_norm": 0.3719600439071655, "learning_rate": 5.419265921949805e-06, "loss": 0.0104, "step": 20201 }, { "epoch": 2.1901561144839548, "grad_norm": 0.040488697588443756, "learning_rate": 5.418540548382417e-06, "loss": 0.0002, "step": 20202 }, { "epoch": 2.1902645273200347, "grad_norm": 0.16963835060596466, "learning_rate": 5.41781517481503e-06, "loss": 0.0007, "step": 20203 }, { "epoch": 2.1903729401561143, "grad_norm": 0.5745999813079834, "learning_rate": 5.417089801247643e-06, "loss": 0.0202, "step": 20204 }, { "epoch": 2.1904813529921943, "grad_norm": 0.2963152527809143, "learning_rate": 5.4163644276802566e-06, "loss": 0.0099, "step": 20205 }, { "epoch": 2.1905897658282742, "grad_norm": 3.7889349460601807, "learning_rate": 5.415639054112869e-06, "loss": 0.0467, "step": 20206 }, { "epoch": 2.190698178664354, "grad_norm": 0.354810893535614, "learning_rate": 5.414913680545482e-06, "loss": 0.006, "step": 20207 }, { "epoch": 2.1908065915004338, "grad_norm": 0.2631891965866089, "learning_rate": 5.414188306978095e-06, "loss": 0.0249, "step": 20208 }, { "epoch": 2.1909150043365133, "grad_norm": 5.067934036254883, "learning_rate": 5.413462933410707e-06, "loss": 0.0017, "step": 20209 }, { "epoch": 2.1910234171725933, "grad_norm": 2.025986909866333, "learning_rate": 5.41273755984332e-06, "loss": 0.0155, "step": 20210 }, { "epoch": 2.191131830008673, "grad_norm": 2.4071884155273438, "learning_rate": 5.412012186275932e-06, "loss": 0.0048, "step": 20211 }, { "epoch": 2.191240242844753, "grad_norm": 0.17127899825572968, "learning_rate": 5.411286812708545e-06, "loss": 0.0008, "step": 20212 }, { "epoch": 2.191348655680833, "grad_norm": 0.3628973364830017, "learning_rate": 5.410561439141158e-06, "loss": 0.0087, "step": 20213 }, { "epoch": 2.1914570685169124, "grad_norm": 0.280850887298584, "learning_rate": 5.409836065573772e-06, "loss": 0.006, "step": 20214 }, { "epoch": 2.1915654813529923, "grad_norm": 0.03686103597283363, "learning_rate": 5.409110692006384e-06, "loss": 0.0001, "step": 20215 }, { "epoch": 2.191673894189072, "grad_norm": 0.014141010120511055, "learning_rate": 5.408385318438997e-06, "loss": 0.0002, "step": 20216 }, { "epoch": 2.191782307025152, "grad_norm": 1.6105338335037231, "learning_rate": 5.407659944871609e-06, "loss": 0.0064, "step": 20217 }, { "epoch": 2.1918907198612314, "grad_norm": 0.24141664803028107, "learning_rate": 5.406934571304222e-06, "loss": 0.0053, "step": 20218 }, { "epoch": 2.1919991326973114, "grad_norm": 2.272592782974243, "learning_rate": 5.406209197736835e-06, "loss": 0.0088, "step": 20219 }, { "epoch": 2.192107545533391, "grad_norm": 0.3158288598060608, "learning_rate": 5.405483824169447e-06, "loss": 0.0061, "step": 20220 }, { "epoch": 2.192215958369471, "grad_norm": 1.2316590547561646, "learning_rate": 5.40475845060206e-06, "loss": 0.0217, "step": 20221 }, { "epoch": 2.192324371205551, "grad_norm": 0.012129740789532661, "learning_rate": 5.404033077034674e-06, "loss": 0.0001, "step": 20222 }, { "epoch": 2.1924327840416304, "grad_norm": 0.6111424565315247, "learning_rate": 5.403307703467286e-06, "loss": 0.0014, "step": 20223 }, { "epoch": 2.1925411968777104, "grad_norm": 0.7569278478622437, "learning_rate": 5.402582329899899e-06, "loss": 0.0073, "step": 20224 }, { "epoch": 2.19264960971379, "grad_norm": 2.291480541229248, "learning_rate": 5.401856956332512e-06, "loss": 0.0215, "step": 20225 }, { "epoch": 2.19275802254987, "grad_norm": 0.2905811667442322, "learning_rate": 5.401131582765124e-06, "loss": 0.0009, "step": 20226 }, { "epoch": 2.1928664353859495, "grad_norm": 0.34908369183540344, "learning_rate": 5.400406209197737e-06, "loss": 0.0087, "step": 20227 }, { "epoch": 2.1929748482220295, "grad_norm": 0.24824592471122742, "learning_rate": 5.39968083563035e-06, "loss": 0.005, "step": 20228 }, { "epoch": 2.1930832610581095, "grad_norm": 0.02220078557729721, "learning_rate": 5.398955462062962e-06, "loss": 0.0002, "step": 20229 }, { "epoch": 2.193191673894189, "grad_norm": 0.11451408267021179, "learning_rate": 5.398230088495575e-06, "loss": 0.0003, "step": 20230 }, { "epoch": 2.193300086730269, "grad_norm": 0.2781483232975006, "learning_rate": 5.397504714928189e-06, "loss": 0.0099, "step": 20231 }, { "epoch": 2.1934084995663485, "grad_norm": 0.6161382794380188, "learning_rate": 5.396779341360801e-06, "loss": 0.0009, "step": 20232 }, { "epoch": 2.1935169124024285, "grad_norm": 1.730941891670227, "learning_rate": 5.396053967793414e-06, "loss": 0.0068, "step": 20233 }, { "epoch": 2.193625325238508, "grad_norm": 0.17788493633270264, "learning_rate": 5.395328594226027e-06, "loss": 0.0006, "step": 20234 }, { "epoch": 2.193733738074588, "grad_norm": 0.005814144387841225, "learning_rate": 5.3946032206586395e-06, "loss": 0.0, "step": 20235 }, { "epoch": 2.193842150910668, "grad_norm": 0.28810232877731323, "learning_rate": 5.3938778470912525e-06, "loss": 0.0074, "step": 20236 }, { "epoch": 2.1939505637467476, "grad_norm": 0.11278919875621796, "learning_rate": 5.393152473523865e-06, "loss": 0.0006, "step": 20237 }, { "epoch": 2.1940589765828276, "grad_norm": 0.020089294761419296, "learning_rate": 5.392427099956478e-06, "loss": 0.0001, "step": 20238 }, { "epoch": 2.194167389418907, "grad_norm": 0.007801306899636984, "learning_rate": 5.391701726389091e-06, "loss": 0.0001, "step": 20239 }, { "epoch": 2.194275802254987, "grad_norm": 3.3367247581481934, "learning_rate": 5.3909763528217045e-06, "loss": 0.0352, "step": 20240 }, { "epoch": 2.1943842150910666, "grad_norm": 0.1846388578414917, "learning_rate": 5.390250979254317e-06, "loss": 0.0003, "step": 20241 }, { "epoch": 2.1944926279271466, "grad_norm": 0.8496453166007996, "learning_rate": 5.38952560568693e-06, "loss": 0.0111, "step": 20242 }, { "epoch": 2.194601040763226, "grad_norm": 10.208159446716309, "learning_rate": 5.388800232119542e-06, "loss": 0.0267, "step": 20243 }, { "epoch": 2.194709453599306, "grad_norm": 2.56147837638855, "learning_rate": 5.388074858552155e-06, "loss": 0.0389, "step": 20244 }, { "epoch": 2.194817866435386, "grad_norm": 0.035997215658426285, "learning_rate": 5.387349484984768e-06, "loss": 0.0002, "step": 20245 }, { "epoch": 2.1949262792714657, "grad_norm": 2.739211320877075, "learning_rate": 5.38662411141738e-06, "loss": 0.0121, "step": 20246 }, { "epoch": 2.1950346921075456, "grad_norm": 0.28395265340805054, "learning_rate": 5.385898737849993e-06, "loss": 0.0061, "step": 20247 }, { "epoch": 2.195143104943625, "grad_norm": 4.202998161315918, "learning_rate": 5.385173364282607e-06, "loss": 0.0094, "step": 20248 }, { "epoch": 2.195251517779705, "grad_norm": 0.21493689715862274, "learning_rate": 5.384447990715219e-06, "loss": 0.0005, "step": 20249 }, { "epoch": 2.1953599306157847, "grad_norm": 0.48538970947265625, "learning_rate": 5.383722617147832e-06, "loss": 0.0143, "step": 20250 }, { "epoch": 2.1954683434518647, "grad_norm": 0.14375965297222137, "learning_rate": 5.382997243580445e-06, "loss": 0.001, "step": 20251 }, { "epoch": 2.1955767562879447, "grad_norm": 0.3726048469543457, "learning_rate": 5.382271870013057e-06, "loss": 0.0132, "step": 20252 }, { "epoch": 2.1956851691240242, "grad_norm": 0.3433443009853363, "learning_rate": 5.38154649644567e-06, "loss": 0.0114, "step": 20253 }, { "epoch": 2.195793581960104, "grad_norm": 0.011737901717424393, "learning_rate": 5.380821122878283e-06, "loss": 0.0001, "step": 20254 }, { "epoch": 2.1959019947961838, "grad_norm": 3.7148518562316895, "learning_rate": 5.380095749310895e-06, "loss": 0.0227, "step": 20255 }, { "epoch": 2.1960104076322637, "grad_norm": 0.023309269919991493, "learning_rate": 5.379370375743508e-06, "loss": 0.0002, "step": 20256 }, { "epoch": 2.1961188204683433, "grad_norm": 0.2012387067079544, "learning_rate": 5.378645002176122e-06, "loss": 0.0003, "step": 20257 }, { "epoch": 2.1962272333044233, "grad_norm": 0.1913452297449112, "learning_rate": 5.377919628608734e-06, "loss": 0.0036, "step": 20258 }, { "epoch": 2.1963356461405033, "grad_norm": 1.4685403108596802, "learning_rate": 5.377194255041347e-06, "loss": 0.0209, "step": 20259 }, { "epoch": 2.196444058976583, "grad_norm": 2.784914016723633, "learning_rate": 5.37646888147396e-06, "loss": 0.0505, "step": 20260 }, { "epoch": 2.1965524718126628, "grad_norm": 5.113440036773682, "learning_rate": 5.375743507906572e-06, "loss": 0.0112, "step": 20261 }, { "epoch": 2.1966608846487423, "grad_norm": 1.5719643831253052, "learning_rate": 5.375018134339185e-06, "loss": 0.0203, "step": 20262 }, { "epoch": 2.1967692974848223, "grad_norm": 0.028498541563749313, "learning_rate": 5.374292760771797e-06, "loss": 0.0002, "step": 20263 }, { "epoch": 2.196877710320902, "grad_norm": 0.11976940184831619, "learning_rate": 5.37356738720441e-06, "loss": 0.0005, "step": 20264 }, { "epoch": 2.196986123156982, "grad_norm": 0.013803220354020596, "learning_rate": 5.372842013637023e-06, "loss": 0.0, "step": 20265 }, { "epoch": 2.1970945359930614, "grad_norm": 0.33954548835754395, "learning_rate": 5.372116640069637e-06, "loss": 0.0085, "step": 20266 }, { "epoch": 2.1972029488291414, "grad_norm": 2.6790647506713867, "learning_rate": 5.371391266502249e-06, "loss": 0.0295, "step": 20267 }, { "epoch": 2.1973113616652213, "grad_norm": 8.576167106628418, "learning_rate": 5.370665892934862e-06, "loss": 0.0667, "step": 20268 }, { "epoch": 2.197419774501301, "grad_norm": 0.3261581361293793, "learning_rate": 5.369940519367474e-06, "loss": 0.0076, "step": 20269 }, { "epoch": 2.197528187337381, "grad_norm": 0.023921851068735123, "learning_rate": 5.369215145800087e-06, "loss": 0.0001, "step": 20270 }, { "epoch": 2.1976366001734604, "grad_norm": 0.4454280734062195, "learning_rate": 5.3684897722327e-06, "loss": 0.0196, "step": 20271 }, { "epoch": 2.1977450130095404, "grad_norm": 0.11390959471464157, "learning_rate": 5.3677643986653125e-06, "loss": 0.0006, "step": 20272 }, { "epoch": 2.19785342584562, "grad_norm": 2.260293483734131, "learning_rate": 5.3670390250979255e-06, "loss": 0.0092, "step": 20273 }, { "epoch": 2.1979618386817, "grad_norm": 0.4864802956581116, "learning_rate": 5.366313651530539e-06, "loss": 0.0014, "step": 20274 }, { "epoch": 2.19807025151778, "grad_norm": 0.2111419141292572, "learning_rate": 5.3655882779631515e-06, "loss": 0.0161, "step": 20275 }, { "epoch": 2.1981786643538594, "grad_norm": 4.304671764373779, "learning_rate": 5.3648629043957645e-06, "loss": 0.0136, "step": 20276 }, { "epoch": 2.1982870771899394, "grad_norm": 3.022830009460449, "learning_rate": 5.3641375308283775e-06, "loss": 0.0191, "step": 20277 }, { "epoch": 2.198395490026019, "grad_norm": 0.30740803480148315, "learning_rate": 5.36341215726099e-06, "loss": 0.0118, "step": 20278 }, { "epoch": 2.198503902862099, "grad_norm": 0.03335585072636604, "learning_rate": 5.362686783693603e-06, "loss": 0.0003, "step": 20279 }, { "epoch": 2.1986123156981785, "grad_norm": 0.021036623045802116, "learning_rate": 5.361961410126216e-06, "loss": 0.0002, "step": 20280 }, { "epoch": 2.1987207285342585, "grad_norm": 17.688718795776367, "learning_rate": 5.361236036558828e-06, "loss": 0.0204, "step": 20281 }, { "epoch": 2.1988291413703385, "grad_norm": 0.014296396635472775, "learning_rate": 5.360510662991441e-06, "loss": 0.0001, "step": 20282 }, { "epoch": 2.198937554206418, "grad_norm": 0.30332112312316895, "learning_rate": 5.3597852894240546e-06, "loss": 0.0069, "step": 20283 }, { "epoch": 2.199045967042498, "grad_norm": 0.08850651979446411, "learning_rate": 5.359059915856667e-06, "loss": 0.0003, "step": 20284 }, { "epoch": 2.1991543798785775, "grad_norm": 8.481900215148926, "learning_rate": 5.35833454228928e-06, "loss": 0.018, "step": 20285 }, { "epoch": 2.1992627927146575, "grad_norm": 0.3507422208786011, "learning_rate": 5.357609168721893e-06, "loss": 0.0008, "step": 20286 }, { "epoch": 2.199371205550737, "grad_norm": 0.4639527201652527, "learning_rate": 5.356883795154505e-06, "loss": 0.0254, "step": 20287 }, { "epoch": 2.199479618386817, "grad_norm": 5.273908615112305, "learning_rate": 5.356158421587118e-06, "loss": 0.0384, "step": 20288 }, { "epoch": 2.1995880312228966, "grad_norm": 0.43007132411003113, "learning_rate": 5.35543304801973e-06, "loss": 0.011, "step": 20289 }, { "epoch": 2.1996964440589766, "grad_norm": 0.8328319191932678, "learning_rate": 5.354707674452343e-06, "loss": 0.0014, "step": 20290 }, { "epoch": 2.1998048568950566, "grad_norm": 0.10288352519273758, "learning_rate": 5.353982300884957e-06, "loss": 0.0006, "step": 20291 }, { "epoch": 2.199913269731136, "grad_norm": 0.44750213623046875, "learning_rate": 5.35325692731757e-06, "loss": 0.0133, "step": 20292 }, { "epoch": 2.200021682567216, "grad_norm": 0.9148734211921692, "learning_rate": 5.352531553750182e-06, "loss": 0.0011, "step": 20293 }, { "epoch": 2.2001300954032956, "grad_norm": 0.8741341233253479, "learning_rate": 5.351806180182795e-06, "loss": 0.002, "step": 20294 }, { "epoch": 2.2002385082393756, "grad_norm": 0.26098111271858215, "learning_rate": 5.351080806615407e-06, "loss": 0.0006, "step": 20295 }, { "epoch": 2.200346921075455, "grad_norm": 3.8387372493743896, "learning_rate": 5.35035543304802e-06, "loss": 0.0438, "step": 20296 }, { "epoch": 2.200455333911535, "grad_norm": 0.007436476182192564, "learning_rate": 5.349630059480633e-06, "loss": 0.0, "step": 20297 }, { "epoch": 2.200563746747615, "grad_norm": 3.5096004009246826, "learning_rate": 5.348904685913245e-06, "loss": 0.02, "step": 20298 }, { "epoch": 2.2006721595836947, "grad_norm": 0.011369478888809681, "learning_rate": 5.348179312345858e-06, "loss": 0.0001, "step": 20299 }, { "epoch": 2.2007805724197746, "grad_norm": 0.31182730197906494, "learning_rate": 5.347453938778472e-06, "loss": 0.01, "step": 20300 }, { "epoch": 2.200888985255854, "grad_norm": 0.3044022023677826, "learning_rate": 5.346728565211084e-06, "loss": 0.0066, "step": 20301 }, { "epoch": 2.200997398091934, "grad_norm": 0.10611431300640106, "learning_rate": 5.346003191643697e-06, "loss": 0.0001, "step": 20302 }, { "epoch": 2.2011058109280137, "grad_norm": 0.4389767348766327, "learning_rate": 5.34527781807631e-06, "loss": 0.0155, "step": 20303 }, { "epoch": 2.2012142237640937, "grad_norm": 1.7148314714431763, "learning_rate": 5.344552444508922e-06, "loss": 0.0142, "step": 20304 }, { "epoch": 2.2013226366001737, "grad_norm": 5.013864994049072, "learning_rate": 5.343827070941535e-06, "loss": 0.0606, "step": 20305 }, { "epoch": 2.2014310494362532, "grad_norm": 0.3910040259361267, "learning_rate": 5.343101697374148e-06, "loss": 0.007, "step": 20306 }, { "epoch": 2.201539462272333, "grad_norm": 0.8710557818412781, "learning_rate": 5.34237632380676e-06, "loss": 0.0007, "step": 20307 }, { "epoch": 2.2016478751084128, "grad_norm": 0.8729866743087769, "learning_rate": 5.341650950239373e-06, "loss": 0.0054, "step": 20308 }, { "epoch": 2.2017562879444927, "grad_norm": 1.2315428256988525, "learning_rate": 5.340925576671987e-06, "loss": 0.0039, "step": 20309 }, { "epoch": 2.2018647007805723, "grad_norm": 1.6943598985671997, "learning_rate": 5.340200203104599e-06, "loss": 0.0185, "step": 20310 }, { "epoch": 2.2019731136166523, "grad_norm": 0.2119205892086029, "learning_rate": 5.339474829537212e-06, "loss": 0.0008, "step": 20311 }, { "epoch": 2.202081526452732, "grad_norm": 0.5674225091934204, "learning_rate": 5.338749455969825e-06, "loss": 0.0164, "step": 20312 }, { "epoch": 2.202189939288812, "grad_norm": 1.4675270318984985, "learning_rate": 5.3380240824024375e-06, "loss": 0.004, "step": 20313 }, { "epoch": 2.202298352124892, "grad_norm": 0.042132068425416946, "learning_rate": 5.3372987088350505e-06, "loss": 0.0002, "step": 20314 }, { "epoch": 2.2024067649609713, "grad_norm": 8.50533676147461, "learning_rate": 5.336573335267663e-06, "loss": 0.0171, "step": 20315 }, { "epoch": 2.2025151777970513, "grad_norm": 1.715161681175232, "learning_rate": 5.335847961700276e-06, "loss": 0.0142, "step": 20316 }, { "epoch": 2.202623590633131, "grad_norm": 0.2709387242794037, "learning_rate": 5.3351225881328895e-06, "loss": 0.0005, "step": 20317 }, { "epoch": 2.202732003469211, "grad_norm": 0.5109561681747437, "learning_rate": 5.3343972145655025e-06, "loss": 0.0132, "step": 20318 }, { "epoch": 2.2028404163052904, "grad_norm": 0.09116122871637344, "learning_rate": 5.333671840998115e-06, "loss": 0.0003, "step": 20319 }, { "epoch": 2.2029488291413704, "grad_norm": 0.3152007460594177, "learning_rate": 5.332946467430728e-06, "loss": 0.0082, "step": 20320 }, { "epoch": 2.20305724197745, "grad_norm": 2.111921548843384, "learning_rate": 5.33222109386334e-06, "loss": 0.0046, "step": 20321 }, { "epoch": 2.20316565481353, "grad_norm": 0.5925354361534119, "learning_rate": 5.331495720295953e-06, "loss": 0.0125, "step": 20322 }, { "epoch": 2.20327406764961, "grad_norm": 0.49297571182250977, "learning_rate": 5.330770346728566e-06, "loss": 0.0127, "step": 20323 }, { "epoch": 2.2033824804856894, "grad_norm": 0.015459584072232246, "learning_rate": 5.330044973161178e-06, "loss": 0.0001, "step": 20324 }, { "epoch": 2.2034908933217694, "grad_norm": 0.07055884599685669, "learning_rate": 5.329319599593791e-06, "loss": 0.0002, "step": 20325 }, { "epoch": 2.203599306157849, "grad_norm": 4.357708930969238, "learning_rate": 5.328594226026405e-06, "loss": 0.0296, "step": 20326 }, { "epoch": 2.203707718993929, "grad_norm": 0.010800429619848728, "learning_rate": 5.327868852459017e-06, "loss": 0.0001, "step": 20327 }, { "epoch": 2.203816131830009, "grad_norm": 0.48824435472488403, "learning_rate": 5.32714347889163e-06, "loss": 0.0072, "step": 20328 }, { "epoch": 2.2039245446660884, "grad_norm": 2.774087905883789, "learning_rate": 5.326418105324243e-06, "loss": 0.0167, "step": 20329 }, { "epoch": 2.2040329575021684, "grad_norm": 0.3415467143058777, "learning_rate": 5.325692731756855e-06, "loss": 0.0005, "step": 20330 }, { "epoch": 2.204141370338248, "grad_norm": 5.406301021575928, "learning_rate": 5.324967358189468e-06, "loss": 0.0243, "step": 20331 }, { "epoch": 2.204249783174328, "grad_norm": 0.024476956576108932, "learning_rate": 5.324241984622081e-06, "loss": 0.0002, "step": 20332 }, { "epoch": 2.2043581960104075, "grad_norm": 0.2843034863471985, "learning_rate": 5.323516611054693e-06, "loss": 0.0092, "step": 20333 }, { "epoch": 2.2044666088464875, "grad_norm": 0.34883224964141846, "learning_rate": 5.322791237487306e-06, "loss": 0.0104, "step": 20334 }, { "epoch": 2.204575021682567, "grad_norm": 0.0024285470135509968, "learning_rate": 5.32206586391992e-06, "loss": 0.0, "step": 20335 }, { "epoch": 2.204683434518647, "grad_norm": 0.031109653413295746, "learning_rate": 5.321340490352532e-06, "loss": 0.0002, "step": 20336 }, { "epoch": 2.204791847354727, "grad_norm": 0.32117539644241333, "learning_rate": 5.320615116785145e-06, "loss": 0.0098, "step": 20337 }, { "epoch": 2.2049002601908065, "grad_norm": 0.5638725757598877, "learning_rate": 5.319889743217758e-06, "loss": 0.0172, "step": 20338 }, { "epoch": 2.2050086730268865, "grad_norm": 0.04234912618994713, "learning_rate": 5.31916436965037e-06, "loss": 0.0002, "step": 20339 }, { "epoch": 2.205117085862966, "grad_norm": 5.138408660888672, "learning_rate": 5.318438996082983e-06, "loss": 0.1191, "step": 20340 }, { "epoch": 2.205225498699046, "grad_norm": 0.7771652936935425, "learning_rate": 5.317713622515595e-06, "loss": 0.0017, "step": 20341 }, { "epoch": 2.2053339115351256, "grad_norm": 0.06855457276105881, "learning_rate": 5.316988248948208e-06, "loss": 0.0002, "step": 20342 }, { "epoch": 2.2054423243712056, "grad_norm": 0.022262604907155037, "learning_rate": 5.316262875380822e-06, "loss": 0.0001, "step": 20343 }, { "epoch": 2.205550737207285, "grad_norm": 1.3615599870681763, "learning_rate": 5.315537501813435e-06, "loss": 0.0242, "step": 20344 }, { "epoch": 2.205659150043365, "grad_norm": 4.214639663696289, "learning_rate": 5.314812128246047e-06, "loss": 0.0316, "step": 20345 }, { "epoch": 2.205767562879445, "grad_norm": 0.37499919533729553, "learning_rate": 5.31408675467866e-06, "loss": 0.012, "step": 20346 }, { "epoch": 2.2058759757155246, "grad_norm": 2.6635851860046387, "learning_rate": 5.313361381111272e-06, "loss": 0.0286, "step": 20347 }, { "epoch": 2.2059843885516046, "grad_norm": 0.47625359892845154, "learning_rate": 5.312636007543885e-06, "loss": 0.0133, "step": 20348 }, { "epoch": 2.206092801387684, "grad_norm": 0.40863049030303955, "learning_rate": 5.311910633976498e-06, "loss": 0.0084, "step": 20349 }, { "epoch": 2.206201214223764, "grad_norm": 1.7855653762817383, "learning_rate": 5.3111852604091105e-06, "loss": 0.0449, "step": 20350 }, { "epoch": 2.2063096270598437, "grad_norm": 0.27895858883857727, "learning_rate": 5.3104598868417235e-06, "loss": 0.0052, "step": 20351 }, { "epoch": 2.2064180398959237, "grad_norm": 0.6468175053596497, "learning_rate": 5.309734513274337e-06, "loss": 0.0007, "step": 20352 }, { "epoch": 2.2065264527320037, "grad_norm": 0.05395802482962608, "learning_rate": 5.3090091397069495e-06, "loss": 0.0003, "step": 20353 }, { "epoch": 2.206634865568083, "grad_norm": 6.835383415222168, "learning_rate": 5.3082837661395625e-06, "loss": 0.0363, "step": 20354 }, { "epoch": 2.206743278404163, "grad_norm": 0.8174788951873779, "learning_rate": 5.3075583925721755e-06, "loss": 0.0022, "step": 20355 }, { "epoch": 2.2068516912402427, "grad_norm": 0.3046223819255829, "learning_rate": 5.306833019004788e-06, "loss": 0.005, "step": 20356 }, { "epoch": 2.2069601040763227, "grad_norm": 2.927368640899658, "learning_rate": 5.306107645437401e-06, "loss": 0.0103, "step": 20357 }, { "epoch": 2.2070685169124022, "grad_norm": 0.018898455426096916, "learning_rate": 5.305382271870014e-06, "loss": 0.0001, "step": 20358 }, { "epoch": 2.2071769297484822, "grad_norm": 0.6104328036308289, "learning_rate": 5.304656898302626e-06, "loss": 0.0143, "step": 20359 }, { "epoch": 2.207285342584562, "grad_norm": 0.46527940034866333, "learning_rate": 5.303931524735239e-06, "loss": 0.0124, "step": 20360 }, { "epoch": 2.2073937554206418, "grad_norm": 0.3119174540042877, "learning_rate": 5.3032061511678526e-06, "loss": 0.0091, "step": 20361 }, { "epoch": 2.2075021682567217, "grad_norm": 0.41519445180892944, "learning_rate": 5.302480777600465e-06, "loss": 0.0102, "step": 20362 }, { "epoch": 2.2076105810928013, "grad_norm": 8.787557601928711, "learning_rate": 5.301755404033078e-06, "loss": 0.0178, "step": 20363 }, { "epoch": 2.2077189939288813, "grad_norm": 3.2758395671844482, "learning_rate": 5.301030030465691e-06, "loss": 0.0157, "step": 20364 }, { "epoch": 2.207827406764961, "grad_norm": 2.9508824348449707, "learning_rate": 5.300304656898303e-06, "loss": 0.0022, "step": 20365 }, { "epoch": 2.207935819601041, "grad_norm": 0.3937433660030365, "learning_rate": 5.299579283330916e-06, "loss": 0.0106, "step": 20366 }, { "epoch": 2.2080442324371203, "grad_norm": 0.02942260541021824, "learning_rate": 5.298853909763528e-06, "loss": 0.0002, "step": 20367 }, { "epoch": 2.2081526452732003, "grad_norm": 0.5396056771278381, "learning_rate": 5.298128536196141e-06, "loss": 0.0136, "step": 20368 }, { "epoch": 2.2082610581092803, "grad_norm": 5.51884651184082, "learning_rate": 5.297403162628755e-06, "loss": 0.0369, "step": 20369 }, { "epoch": 2.20836947094536, "grad_norm": 0.3825657367706299, "learning_rate": 5.296677789061368e-06, "loss": 0.0201, "step": 20370 }, { "epoch": 2.20847788378144, "grad_norm": 0.4762805700302124, "learning_rate": 5.29595241549398e-06, "loss": 0.0081, "step": 20371 }, { "epoch": 2.2085862966175194, "grad_norm": 5.2990241050720215, "learning_rate": 5.295227041926593e-06, "loss": 0.0419, "step": 20372 }, { "epoch": 2.2086947094535994, "grad_norm": 1.9063847064971924, "learning_rate": 5.294501668359205e-06, "loss": 0.0228, "step": 20373 }, { "epoch": 2.208803122289679, "grad_norm": 0.5145139694213867, "learning_rate": 5.293776294791818e-06, "loss": 0.0234, "step": 20374 }, { "epoch": 2.208911535125759, "grad_norm": 1.7568129301071167, "learning_rate": 5.293050921224431e-06, "loss": 0.0032, "step": 20375 }, { "epoch": 2.209019947961839, "grad_norm": 0.5357517004013062, "learning_rate": 5.292325547657043e-06, "loss": 0.018, "step": 20376 }, { "epoch": 2.2091283607979184, "grad_norm": 0.44577890634536743, "learning_rate": 5.291600174089656e-06, "loss": 0.0085, "step": 20377 }, { "epoch": 2.2092367736339984, "grad_norm": 5.975278377532959, "learning_rate": 5.29087480052227e-06, "loss": 0.0374, "step": 20378 }, { "epoch": 2.209345186470078, "grad_norm": 6.9576263427734375, "learning_rate": 5.290149426954882e-06, "loss": 0.0327, "step": 20379 }, { "epoch": 2.209453599306158, "grad_norm": 0.3540419936180115, "learning_rate": 5.289424053387495e-06, "loss": 0.0058, "step": 20380 }, { "epoch": 2.2095620121422375, "grad_norm": 0.03871946781873703, "learning_rate": 5.288698679820108e-06, "loss": 0.0001, "step": 20381 }, { "epoch": 2.2096704249783174, "grad_norm": 4.008090972900391, "learning_rate": 5.28797330625272e-06, "loss": 0.0197, "step": 20382 }, { "epoch": 2.2097788378143974, "grad_norm": 0.019848424941301346, "learning_rate": 5.287247932685333e-06, "loss": 0.0001, "step": 20383 }, { "epoch": 2.209887250650477, "grad_norm": 0.37166091799736023, "learning_rate": 5.286522559117946e-06, "loss": 0.0115, "step": 20384 }, { "epoch": 2.209995663486557, "grad_norm": 0.6371952891349792, "learning_rate": 5.285797185550558e-06, "loss": 0.0178, "step": 20385 }, { "epoch": 2.2101040763226365, "grad_norm": 0.6681990027427673, "learning_rate": 5.285071811983172e-06, "loss": 0.0013, "step": 20386 }, { "epoch": 2.2102124891587165, "grad_norm": 0.006739713717252016, "learning_rate": 5.284346438415785e-06, "loss": 0.0001, "step": 20387 }, { "epoch": 2.210320901994796, "grad_norm": 0.02508142776787281, "learning_rate": 5.283621064848397e-06, "loss": 0.0002, "step": 20388 }, { "epoch": 2.210429314830876, "grad_norm": 0.4800083339214325, "learning_rate": 5.28289569128101e-06, "loss": 0.0058, "step": 20389 }, { "epoch": 2.2105377276669556, "grad_norm": 0.12503862380981445, "learning_rate": 5.282170317713623e-06, "loss": 0.0006, "step": 20390 }, { "epoch": 2.2106461405030355, "grad_norm": 1.6879124641418457, "learning_rate": 5.2814449441462355e-06, "loss": 0.0034, "step": 20391 }, { "epoch": 2.2107545533391155, "grad_norm": 22.698041915893555, "learning_rate": 5.2807195705788485e-06, "loss": 0.0315, "step": 20392 }, { "epoch": 2.210862966175195, "grad_norm": 0.009535077027976513, "learning_rate": 5.279994197011461e-06, "loss": 0.0001, "step": 20393 }, { "epoch": 2.210971379011275, "grad_norm": 19.406496047973633, "learning_rate": 5.279268823444074e-06, "loss": 0.0201, "step": 20394 }, { "epoch": 2.2110797918473546, "grad_norm": 0.891856849193573, "learning_rate": 5.2785434498766875e-06, "loss": 0.0013, "step": 20395 }, { "epoch": 2.2111882046834346, "grad_norm": 0.05705218017101288, "learning_rate": 5.2778180763093005e-06, "loss": 0.0004, "step": 20396 }, { "epoch": 2.211296617519514, "grad_norm": 0.6425046920776367, "learning_rate": 5.277092702741913e-06, "loss": 0.0107, "step": 20397 }, { "epoch": 2.211405030355594, "grad_norm": 0.31603533029556274, "learning_rate": 5.276367329174526e-06, "loss": 0.0011, "step": 20398 }, { "epoch": 2.211513443191674, "grad_norm": 0.41171273589134216, "learning_rate": 5.275641955607138e-06, "loss": 0.0081, "step": 20399 }, { "epoch": 2.2116218560277536, "grad_norm": 0.3905247747898102, "learning_rate": 5.274916582039751e-06, "loss": 0.0195, "step": 20400 }, { "epoch": 2.2117302688638336, "grad_norm": 1.6217763423919678, "learning_rate": 5.274191208472364e-06, "loss": 0.0161, "step": 20401 }, { "epoch": 2.211838681699913, "grad_norm": 0.026938334107398987, "learning_rate": 5.273465834904976e-06, "loss": 0.0002, "step": 20402 }, { "epoch": 2.211947094535993, "grad_norm": 6.086272716522217, "learning_rate": 5.272740461337589e-06, "loss": 0.0679, "step": 20403 }, { "epoch": 2.2120555073720727, "grad_norm": 0.4477560222148895, "learning_rate": 5.272015087770203e-06, "loss": 0.0149, "step": 20404 }, { "epoch": 2.2121639202081527, "grad_norm": 12.211484909057617, "learning_rate": 5.271289714202815e-06, "loss": 0.0084, "step": 20405 }, { "epoch": 2.2122723330442327, "grad_norm": 0.973783016204834, "learning_rate": 5.270564340635428e-06, "loss": 0.0292, "step": 20406 }, { "epoch": 2.212380745880312, "grad_norm": 0.6567264199256897, "learning_rate": 5.269838967068041e-06, "loss": 0.0009, "step": 20407 }, { "epoch": 2.212489158716392, "grad_norm": 0.3508073389530182, "learning_rate": 5.269113593500653e-06, "loss": 0.0008, "step": 20408 }, { "epoch": 2.2125975715524717, "grad_norm": 21.955263137817383, "learning_rate": 5.268388219933266e-06, "loss": 0.0157, "step": 20409 }, { "epoch": 2.2127059843885517, "grad_norm": 0.675423264503479, "learning_rate": 5.267662846365879e-06, "loss": 0.0125, "step": 20410 }, { "epoch": 2.2128143972246312, "grad_norm": 1.0430787801742554, "learning_rate": 5.266937472798491e-06, "loss": 0.0258, "step": 20411 }, { "epoch": 2.2129228100607112, "grad_norm": 0.04454924166202545, "learning_rate": 5.266212099231105e-06, "loss": 0.0001, "step": 20412 }, { "epoch": 2.2130312228967908, "grad_norm": 0.2930539846420288, "learning_rate": 5.265486725663718e-06, "loss": 0.0067, "step": 20413 }, { "epoch": 2.2131396357328708, "grad_norm": 0.007284688297659159, "learning_rate": 5.26476135209633e-06, "loss": 0.0001, "step": 20414 }, { "epoch": 2.2132480485689507, "grad_norm": 0.492418497800827, "learning_rate": 5.264035978528943e-06, "loss": 0.0254, "step": 20415 }, { "epoch": 2.2133564614050303, "grad_norm": 0.2624034583568573, "learning_rate": 5.263310604961556e-06, "loss": 0.0174, "step": 20416 }, { "epoch": 2.2134648742411103, "grad_norm": 0.1989908367395401, "learning_rate": 5.262585231394168e-06, "loss": 0.0005, "step": 20417 }, { "epoch": 2.21357328707719, "grad_norm": 0.1367732584476471, "learning_rate": 5.261859857826781e-06, "loss": 0.0006, "step": 20418 }, { "epoch": 2.21368169991327, "grad_norm": 0.024716027081012726, "learning_rate": 5.261134484259393e-06, "loss": 0.0002, "step": 20419 }, { "epoch": 2.2137901127493493, "grad_norm": 0.007279240991920233, "learning_rate": 5.260409110692006e-06, "loss": 0.0, "step": 20420 }, { "epoch": 2.2138985255854293, "grad_norm": 0.007201684173196554, "learning_rate": 5.25968373712462e-06, "loss": 0.0, "step": 20421 }, { "epoch": 2.2140069384215093, "grad_norm": 1.0388959646224976, "learning_rate": 5.258958363557233e-06, "loss": 0.0102, "step": 20422 }, { "epoch": 2.214115351257589, "grad_norm": 0.35153883695602417, "learning_rate": 5.258232989989845e-06, "loss": 0.0062, "step": 20423 }, { "epoch": 2.214223764093669, "grad_norm": 0.44149234890937805, "learning_rate": 5.257507616422458e-06, "loss": 0.0119, "step": 20424 }, { "epoch": 2.2143321769297484, "grad_norm": 1.9237589836120605, "learning_rate": 5.25678224285507e-06, "loss": 0.0234, "step": 20425 }, { "epoch": 2.2144405897658284, "grad_norm": 0.2913525104522705, "learning_rate": 5.256056869287683e-06, "loss": 0.005, "step": 20426 }, { "epoch": 2.214549002601908, "grad_norm": 2.598041296005249, "learning_rate": 5.255331495720296e-06, "loss": 0.0626, "step": 20427 }, { "epoch": 2.214657415437988, "grad_norm": 0.9136008620262146, "learning_rate": 5.2546061221529085e-06, "loss": 0.0025, "step": 20428 }, { "epoch": 2.214765828274068, "grad_norm": 3.6999542713165283, "learning_rate": 5.2538807485855215e-06, "loss": 0.0108, "step": 20429 }, { "epoch": 2.2148742411101474, "grad_norm": 0.26454007625579834, "learning_rate": 5.253155375018135e-06, "loss": 0.0054, "step": 20430 }, { "epoch": 2.2149826539462274, "grad_norm": 0.2638658881187439, "learning_rate": 5.2524300014507475e-06, "loss": 0.0055, "step": 20431 }, { "epoch": 2.215091066782307, "grad_norm": 0.4396326541900635, "learning_rate": 5.2517046278833605e-06, "loss": 0.0069, "step": 20432 }, { "epoch": 2.215199479618387, "grad_norm": 0.0648333951830864, "learning_rate": 5.2509792543159735e-06, "loss": 0.0002, "step": 20433 }, { "epoch": 2.2153078924544665, "grad_norm": 2.066728115081787, "learning_rate": 5.250253880748586e-06, "loss": 0.0164, "step": 20434 }, { "epoch": 2.2154163052905465, "grad_norm": 4.153767108917236, "learning_rate": 5.249528507181199e-06, "loss": 0.0267, "step": 20435 }, { "epoch": 2.215524718126626, "grad_norm": 0.08018651604652405, "learning_rate": 5.248803133613812e-06, "loss": 0.0001, "step": 20436 }, { "epoch": 2.215633130962706, "grad_norm": 5.256144046783447, "learning_rate": 5.248077760046424e-06, "loss": 0.0032, "step": 20437 }, { "epoch": 2.215741543798786, "grad_norm": 0.018671168014407158, "learning_rate": 5.247352386479038e-06, "loss": 0.0001, "step": 20438 }, { "epoch": 2.2158499566348655, "grad_norm": 0.5923002362251282, "learning_rate": 5.246627012911651e-06, "loss": 0.0185, "step": 20439 }, { "epoch": 2.2159583694709455, "grad_norm": 0.4084080457687378, "learning_rate": 5.245901639344263e-06, "loss": 0.0181, "step": 20440 }, { "epoch": 2.216066782307025, "grad_norm": 0.07477279007434845, "learning_rate": 5.245176265776876e-06, "loss": 0.0004, "step": 20441 }, { "epoch": 2.216175195143105, "grad_norm": 0.34060555696487427, "learning_rate": 5.244450892209489e-06, "loss": 0.0004, "step": 20442 }, { "epoch": 2.2162836079791846, "grad_norm": 0.0073736608028411865, "learning_rate": 5.243725518642101e-06, "loss": 0.0, "step": 20443 }, { "epoch": 2.2163920208152645, "grad_norm": 2.353142023086548, "learning_rate": 5.243000145074714e-06, "loss": 0.0286, "step": 20444 }, { "epoch": 2.2165004336513445, "grad_norm": 0.0025627752766013145, "learning_rate": 5.242274771507326e-06, "loss": 0.0, "step": 20445 }, { "epoch": 2.216608846487424, "grad_norm": 0.46073606610298157, "learning_rate": 5.241549397939939e-06, "loss": 0.0155, "step": 20446 }, { "epoch": 2.216717259323504, "grad_norm": 0.6298935413360596, "learning_rate": 5.240824024372553e-06, "loss": 0.0008, "step": 20447 }, { "epoch": 2.2168256721595836, "grad_norm": 0.26674506068229675, "learning_rate": 5.240098650805166e-06, "loss": 0.0055, "step": 20448 }, { "epoch": 2.2169340849956636, "grad_norm": 3.2201411724090576, "learning_rate": 5.239373277237778e-06, "loss": 0.0217, "step": 20449 }, { "epoch": 2.217042497831743, "grad_norm": 0.688940167427063, "learning_rate": 5.238647903670391e-06, "loss": 0.0195, "step": 20450 }, { "epoch": 2.217150910667823, "grad_norm": 0.24643440544605255, "learning_rate": 5.237922530103003e-06, "loss": 0.005, "step": 20451 }, { "epoch": 2.217259323503903, "grad_norm": 0.6350074410438538, "learning_rate": 5.237197156535616e-06, "loss": 0.0159, "step": 20452 }, { "epoch": 2.2173677363399826, "grad_norm": 0.046435873955488205, "learning_rate": 5.236471782968229e-06, "loss": 0.0002, "step": 20453 }, { "epoch": 2.2174761491760626, "grad_norm": 0.4143119156360626, "learning_rate": 5.235746409400841e-06, "loss": 0.0077, "step": 20454 }, { "epoch": 2.217584562012142, "grad_norm": 0.767697274684906, "learning_rate": 5.235021035833454e-06, "loss": 0.0241, "step": 20455 }, { "epoch": 2.217692974848222, "grad_norm": 5.198317527770996, "learning_rate": 5.234295662266068e-06, "loss": 0.004, "step": 20456 }, { "epoch": 2.2178013876843017, "grad_norm": 0.27284055948257446, "learning_rate": 5.23357028869868e-06, "loss": 0.0054, "step": 20457 }, { "epoch": 2.2179098005203817, "grad_norm": 0.005698029883205891, "learning_rate": 5.232844915131293e-06, "loss": 0.0, "step": 20458 }, { "epoch": 2.218018213356461, "grad_norm": 0.004759716335684061, "learning_rate": 5.232119541563906e-06, "loss": 0.0, "step": 20459 }, { "epoch": 2.218126626192541, "grad_norm": 0.41609612107276917, "learning_rate": 5.231394167996518e-06, "loss": 0.0148, "step": 20460 }, { "epoch": 2.218235039028621, "grad_norm": 0.0026046147104352713, "learning_rate": 5.230668794429131e-06, "loss": 0.0, "step": 20461 }, { "epoch": 2.2183434518647007, "grad_norm": 3.646296501159668, "learning_rate": 5.229943420861744e-06, "loss": 0.0532, "step": 20462 }, { "epoch": 2.2184518647007807, "grad_norm": 0.3389638364315033, "learning_rate": 5.229218047294356e-06, "loss": 0.0009, "step": 20463 }, { "epoch": 2.2185602775368602, "grad_norm": 0.10854252427816391, "learning_rate": 5.22849267372697e-06, "loss": 0.0006, "step": 20464 }, { "epoch": 2.2186686903729402, "grad_norm": 0.5573577284812927, "learning_rate": 5.227767300159583e-06, "loss": 0.0162, "step": 20465 }, { "epoch": 2.2187771032090198, "grad_norm": 0.014535052701830864, "learning_rate": 5.227041926592195e-06, "loss": 0.0001, "step": 20466 }, { "epoch": 2.2188855160450998, "grad_norm": 0.00935112964361906, "learning_rate": 5.226316553024808e-06, "loss": 0.0001, "step": 20467 }, { "epoch": 2.2189939288811793, "grad_norm": 0.7109652757644653, "learning_rate": 5.225591179457421e-06, "loss": 0.0382, "step": 20468 }, { "epoch": 2.2191023417172593, "grad_norm": 6.340170383453369, "learning_rate": 5.2248658058900335e-06, "loss": 0.0097, "step": 20469 }, { "epoch": 2.2192107545533393, "grad_norm": 0.32293039560317993, "learning_rate": 5.2241404323226465e-06, "loss": 0.0066, "step": 20470 }, { "epoch": 2.219319167389419, "grad_norm": 7.605950355529785, "learning_rate": 5.223415058755259e-06, "loss": 0.0394, "step": 20471 }, { "epoch": 2.219427580225499, "grad_norm": 2.8601269721984863, "learning_rate": 5.222689685187872e-06, "loss": 0.0286, "step": 20472 }, { "epoch": 2.2195359930615783, "grad_norm": 0.2700861990451813, "learning_rate": 5.2219643116204855e-06, "loss": 0.0058, "step": 20473 }, { "epoch": 2.2196444058976583, "grad_norm": 0.07261102646589279, "learning_rate": 5.2212389380530985e-06, "loss": 0.0002, "step": 20474 }, { "epoch": 2.2197528187337383, "grad_norm": 0.005692827049642801, "learning_rate": 5.220513564485711e-06, "loss": 0.0, "step": 20475 }, { "epoch": 2.219861231569818, "grad_norm": 0.40505585074424744, "learning_rate": 5.219788190918324e-06, "loss": 0.0142, "step": 20476 }, { "epoch": 2.219969644405898, "grad_norm": 1.0530178546905518, "learning_rate": 5.219062817350936e-06, "loss": 0.0134, "step": 20477 }, { "epoch": 2.2200780572419774, "grad_norm": 2.39886474609375, "learning_rate": 5.218337443783549e-06, "loss": 0.0149, "step": 20478 }, { "epoch": 2.2201864700780574, "grad_norm": 0.7285446524620056, "learning_rate": 5.217612070216162e-06, "loss": 0.0065, "step": 20479 }, { "epoch": 2.220294882914137, "grad_norm": 0.08478853851556778, "learning_rate": 5.216886696648774e-06, "loss": 0.0004, "step": 20480 }, { "epoch": 2.220403295750217, "grad_norm": 0.20266568660736084, "learning_rate": 5.216161323081388e-06, "loss": 0.001, "step": 20481 }, { "epoch": 2.2205117085862964, "grad_norm": 12.695405960083008, "learning_rate": 5.215435949514001e-06, "loss": 0.0338, "step": 20482 }, { "epoch": 2.2206201214223764, "grad_norm": 0.07891647517681122, "learning_rate": 5.214710575946613e-06, "loss": 0.0002, "step": 20483 }, { "epoch": 2.2207285342584564, "grad_norm": 0.3522714376449585, "learning_rate": 5.213985202379226e-06, "loss": 0.0059, "step": 20484 }, { "epoch": 2.220836947094536, "grad_norm": 0.24371390044689178, "learning_rate": 5.213259828811839e-06, "loss": 0.0051, "step": 20485 }, { "epoch": 2.220945359930616, "grad_norm": 0.3145529627799988, "learning_rate": 5.212534455244451e-06, "loss": 0.0103, "step": 20486 }, { "epoch": 2.2210537727666955, "grad_norm": 0.06411752104759216, "learning_rate": 5.211809081677064e-06, "loss": 0.0004, "step": 20487 }, { "epoch": 2.2211621856027755, "grad_norm": 0.030303847044706345, "learning_rate": 5.211083708109677e-06, "loss": 0.0002, "step": 20488 }, { "epoch": 2.221270598438855, "grad_norm": 0.32078564167022705, "learning_rate": 5.210358334542289e-06, "loss": 0.0057, "step": 20489 }, { "epoch": 2.221379011274935, "grad_norm": 0.37338536977767944, "learning_rate": 5.209632960974903e-06, "loss": 0.0122, "step": 20490 }, { "epoch": 2.2214874241110145, "grad_norm": 0.26342153549194336, "learning_rate": 5.208907587407516e-06, "loss": 0.0055, "step": 20491 }, { "epoch": 2.2215958369470945, "grad_norm": 1.044064998626709, "learning_rate": 5.208182213840128e-06, "loss": 0.0657, "step": 20492 }, { "epoch": 2.2217042497831745, "grad_norm": 10.95128345489502, "learning_rate": 5.207456840272741e-06, "loss": 0.0413, "step": 20493 }, { "epoch": 2.221812662619254, "grad_norm": 0.05839632824063301, "learning_rate": 5.206731466705354e-06, "loss": 0.0002, "step": 20494 }, { "epoch": 2.221921075455334, "grad_norm": 0.27055418491363525, "learning_rate": 5.206006093137966e-06, "loss": 0.0059, "step": 20495 }, { "epoch": 2.2220294882914136, "grad_norm": 0.27488794922828674, "learning_rate": 5.205280719570579e-06, "loss": 0.0061, "step": 20496 }, { "epoch": 2.2221379011274935, "grad_norm": 0.30020833015441895, "learning_rate": 5.204555346003191e-06, "loss": 0.0005, "step": 20497 }, { "epoch": 2.222246313963573, "grad_norm": 2.4704582691192627, "learning_rate": 5.203829972435804e-06, "loss": 0.0272, "step": 20498 }, { "epoch": 2.222354726799653, "grad_norm": 8.159133911132812, "learning_rate": 5.203104598868418e-06, "loss": 0.0479, "step": 20499 }, { "epoch": 2.222463139635733, "grad_norm": 2.63373064994812, "learning_rate": 5.202379225301031e-06, "loss": 0.0078, "step": 20500 }, { "epoch": 2.2225715524718126, "grad_norm": 1.7715791463851929, "learning_rate": 5.201653851733643e-06, "loss": 0.0207, "step": 20501 }, { "epoch": 2.2226799653078926, "grad_norm": 1.236171007156372, "learning_rate": 5.200928478166256e-06, "loss": 0.0071, "step": 20502 }, { "epoch": 2.222788378143972, "grad_norm": 1.1486234664916992, "learning_rate": 5.200203104598868e-06, "loss": 0.0339, "step": 20503 }, { "epoch": 2.222896790980052, "grad_norm": 2.505139112472534, "learning_rate": 5.199477731031481e-06, "loss": 0.0494, "step": 20504 }, { "epoch": 2.2230052038161316, "grad_norm": 11.241273880004883, "learning_rate": 5.198752357464094e-06, "loss": 0.0491, "step": 20505 }, { "epoch": 2.2231136166522116, "grad_norm": 0.30457451939582825, "learning_rate": 5.1980269838967065e-06, "loss": 0.0005, "step": 20506 }, { "epoch": 2.2232220294882916, "grad_norm": 0.20487500727176666, "learning_rate": 5.19730161032932e-06, "loss": 0.0003, "step": 20507 }, { "epoch": 2.223330442324371, "grad_norm": 1.492323637008667, "learning_rate": 5.196576236761933e-06, "loss": 0.0501, "step": 20508 }, { "epoch": 2.223438855160451, "grad_norm": 0.004234327934682369, "learning_rate": 5.1958508631945455e-06, "loss": 0.0, "step": 20509 }, { "epoch": 2.2235472679965307, "grad_norm": 0.3338342010974884, "learning_rate": 5.1951254896271585e-06, "loss": 0.0008, "step": 20510 }, { "epoch": 2.2236556808326107, "grad_norm": 3.948437213897705, "learning_rate": 5.1944001160597715e-06, "loss": 0.0399, "step": 20511 }, { "epoch": 2.22376409366869, "grad_norm": 2.60585880279541, "learning_rate": 5.193674742492384e-06, "loss": 0.0049, "step": 20512 }, { "epoch": 2.22387250650477, "grad_norm": 0.4867062568664551, "learning_rate": 5.192949368924997e-06, "loss": 0.0179, "step": 20513 }, { "epoch": 2.2239809193408497, "grad_norm": 0.3172338902950287, "learning_rate": 5.19222399535761e-06, "loss": 0.0129, "step": 20514 }, { "epoch": 2.2240893321769297, "grad_norm": 0.7452172636985779, "learning_rate": 5.191498621790222e-06, "loss": 0.001, "step": 20515 }, { "epoch": 2.2241977450130097, "grad_norm": 0.004871101584285498, "learning_rate": 5.190773248222836e-06, "loss": 0.0, "step": 20516 }, { "epoch": 2.2243061578490892, "grad_norm": 0.5113406181335449, "learning_rate": 5.190047874655449e-06, "loss": 0.0162, "step": 20517 }, { "epoch": 2.2244145706851692, "grad_norm": 1.2126140594482422, "learning_rate": 5.189322501088061e-06, "loss": 0.0144, "step": 20518 }, { "epoch": 2.2245229835212488, "grad_norm": 4.240746021270752, "learning_rate": 5.188597127520674e-06, "loss": 0.0413, "step": 20519 }, { "epoch": 2.2246313963573288, "grad_norm": 0.39358317852020264, "learning_rate": 5.187871753953287e-06, "loss": 0.0091, "step": 20520 }, { "epoch": 2.2247398091934083, "grad_norm": 0.3505766987800598, "learning_rate": 5.187146380385899e-06, "loss": 0.0113, "step": 20521 }, { "epoch": 2.2248482220294883, "grad_norm": 1.5046120882034302, "learning_rate": 5.186421006818512e-06, "loss": 0.0306, "step": 20522 }, { "epoch": 2.2249566348655683, "grad_norm": 0.36380186676979065, "learning_rate": 5.185695633251124e-06, "loss": 0.0228, "step": 20523 }, { "epoch": 2.225065047701648, "grad_norm": 0.48116442561149597, "learning_rate": 5.184970259683737e-06, "loss": 0.0193, "step": 20524 }, { "epoch": 2.225173460537728, "grad_norm": 2.289567708969116, "learning_rate": 5.184244886116351e-06, "loss": 0.01, "step": 20525 }, { "epoch": 2.2252818733738073, "grad_norm": 0.33962392807006836, "learning_rate": 5.183519512548964e-06, "loss": 0.0094, "step": 20526 }, { "epoch": 2.2253902862098873, "grad_norm": 9.287471771240234, "learning_rate": 5.182794138981576e-06, "loss": 0.0214, "step": 20527 }, { "epoch": 2.225498699045967, "grad_norm": 3.4096570014953613, "learning_rate": 5.182068765414189e-06, "loss": 0.0177, "step": 20528 }, { "epoch": 2.225607111882047, "grad_norm": 0.6461564898490906, "learning_rate": 5.181343391846801e-06, "loss": 0.0235, "step": 20529 }, { "epoch": 2.225715524718127, "grad_norm": 0.045417629182338715, "learning_rate": 5.180618018279414e-06, "loss": 0.0006, "step": 20530 }, { "epoch": 2.2258239375542064, "grad_norm": 2.3131015300750732, "learning_rate": 5.179892644712027e-06, "loss": 0.0438, "step": 20531 }, { "epoch": 2.2259323503902864, "grad_norm": 0.3447069823741913, "learning_rate": 5.179167271144639e-06, "loss": 0.0104, "step": 20532 }, { "epoch": 2.226040763226366, "grad_norm": 1.9846652746200562, "learning_rate": 5.178441897577253e-06, "loss": 0.0439, "step": 20533 }, { "epoch": 2.226149176062446, "grad_norm": 0.7409297227859497, "learning_rate": 5.177716524009866e-06, "loss": 0.0195, "step": 20534 }, { "epoch": 2.2262575888985254, "grad_norm": 3.2521708011627197, "learning_rate": 5.176991150442478e-06, "loss": 0.0052, "step": 20535 }, { "epoch": 2.2263660017346054, "grad_norm": 0.15797245502471924, "learning_rate": 5.176265776875091e-06, "loss": 0.0007, "step": 20536 }, { "epoch": 2.226474414570685, "grad_norm": 2.892707347869873, "learning_rate": 5.175540403307704e-06, "loss": 0.0184, "step": 20537 }, { "epoch": 2.226582827406765, "grad_norm": 0.28139936923980713, "learning_rate": 5.174815029740316e-06, "loss": 0.006, "step": 20538 }, { "epoch": 2.226691240242845, "grad_norm": 0.027573708444833755, "learning_rate": 5.174089656172929e-06, "loss": 0.0002, "step": 20539 }, { "epoch": 2.2267996530789245, "grad_norm": 9.321603775024414, "learning_rate": 5.173364282605542e-06, "loss": 0.0159, "step": 20540 }, { "epoch": 2.2269080659150045, "grad_norm": 0.11307333409786224, "learning_rate": 5.1726389090381544e-06, "loss": 0.0002, "step": 20541 }, { "epoch": 2.227016478751084, "grad_norm": 3.414673328399658, "learning_rate": 5.171913535470768e-06, "loss": 0.0104, "step": 20542 }, { "epoch": 2.227124891587164, "grad_norm": 0.5629209280014038, "learning_rate": 5.171188161903381e-06, "loss": 0.0126, "step": 20543 }, { "epoch": 2.2272333044232435, "grad_norm": 0.10023543238639832, "learning_rate": 5.170462788335993e-06, "loss": 0.0006, "step": 20544 }, { "epoch": 2.2273417172593235, "grad_norm": 2.683830499649048, "learning_rate": 5.169737414768606e-06, "loss": 0.0128, "step": 20545 }, { "epoch": 2.2274501300954035, "grad_norm": 1.4282000064849854, "learning_rate": 5.169012041201219e-06, "loss": 0.0429, "step": 20546 }, { "epoch": 2.227558542931483, "grad_norm": 0.2855638265609741, "learning_rate": 5.1682866676338315e-06, "loss": 0.0062, "step": 20547 }, { "epoch": 2.227666955767563, "grad_norm": 11.590166091918945, "learning_rate": 5.1675612940664445e-06, "loss": 0.0853, "step": 20548 }, { "epoch": 2.2277753686036426, "grad_norm": 0.010101287625730038, "learning_rate": 5.166835920499057e-06, "loss": 0.0001, "step": 20549 }, { "epoch": 2.2278837814397225, "grad_norm": 3.75089430809021, "learning_rate": 5.16611054693167e-06, "loss": 0.0211, "step": 20550 }, { "epoch": 2.227992194275802, "grad_norm": 0.22664949297904968, "learning_rate": 5.1653851733642835e-06, "loss": 0.0117, "step": 20551 }, { "epoch": 2.228100607111882, "grad_norm": 0.273480087518692, "learning_rate": 5.1646597997968965e-06, "loss": 0.0056, "step": 20552 }, { "epoch": 2.228209019947962, "grad_norm": 3.0342109203338623, "learning_rate": 5.163934426229509e-06, "loss": 0.0142, "step": 20553 }, { "epoch": 2.2283174327840416, "grad_norm": 0.2690584659576416, "learning_rate": 5.163209052662122e-06, "loss": 0.0059, "step": 20554 }, { "epoch": 2.2284258456201216, "grad_norm": 0.43708139657974243, "learning_rate": 5.162483679094734e-06, "loss": 0.0017, "step": 20555 }, { "epoch": 2.228534258456201, "grad_norm": 0.7887773513793945, "learning_rate": 5.161758305527347e-06, "loss": 0.001, "step": 20556 }, { "epoch": 2.228642671292281, "grad_norm": 0.4270191192626953, "learning_rate": 5.16103293195996e-06, "loss": 0.0119, "step": 20557 }, { "epoch": 2.2287510841283606, "grad_norm": 1.268916130065918, "learning_rate": 5.160307558392572e-06, "loss": 0.0159, "step": 20558 }, { "epoch": 2.2288594969644406, "grad_norm": 0.06234219670295715, "learning_rate": 5.159582184825186e-06, "loss": 0.0007, "step": 20559 }, { "epoch": 2.22896790980052, "grad_norm": 0.37715065479278564, "learning_rate": 5.158856811257799e-06, "loss": 0.0204, "step": 20560 }, { "epoch": 2.2290763226366, "grad_norm": 0.05978221073746681, "learning_rate": 5.158131437690411e-06, "loss": 0.0003, "step": 20561 }, { "epoch": 2.22918473547268, "grad_norm": 0.11111203581094742, "learning_rate": 5.157406064123024e-06, "loss": 0.0009, "step": 20562 }, { "epoch": 2.2292931483087597, "grad_norm": 0.16314475238323212, "learning_rate": 5.156680690555637e-06, "loss": 0.0011, "step": 20563 }, { "epoch": 2.2294015611448397, "grad_norm": 0.48947080969810486, "learning_rate": 5.155955316988249e-06, "loss": 0.0179, "step": 20564 }, { "epoch": 2.229509973980919, "grad_norm": 0.354254812002182, "learning_rate": 5.155229943420862e-06, "loss": 0.0215, "step": 20565 }, { "epoch": 2.229618386816999, "grad_norm": 0.32748252153396606, "learning_rate": 5.154504569853475e-06, "loss": 0.0006, "step": 20566 }, { "epoch": 2.2297267996530787, "grad_norm": 1.6406052112579346, "learning_rate": 5.153779196286087e-06, "loss": 0.024, "step": 20567 }, { "epoch": 2.2298352124891587, "grad_norm": 0.5197947025299072, "learning_rate": 5.153053822718701e-06, "loss": 0.0114, "step": 20568 }, { "epoch": 2.2299436253252387, "grad_norm": 0.09473980218172073, "learning_rate": 5.152328449151314e-06, "loss": 0.0003, "step": 20569 }, { "epoch": 2.2300520381613183, "grad_norm": 3.0391151905059814, "learning_rate": 5.151603075583926e-06, "loss": 0.0306, "step": 20570 }, { "epoch": 2.2301604509973982, "grad_norm": 0.3416268527507782, "learning_rate": 5.150877702016539e-06, "loss": 0.0269, "step": 20571 }, { "epoch": 2.2302688638334778, "grad_norm": 1.8959347009658813, "learning_rate": 5.150152328449152e-06, "loss": 0.0019, "step": 20572 }, { "epoch": 2.2303772766695578, "grad_norm": 0.2667589485645294, "learning_rate": 5.149426954881764e-06, "loss": 0.0051, "step": 20573 }, { "epoch": 2.2304856895056373, "grad_norm": 0.2868347764015198, "learning_rate": 5.148701581314377e-06, "loss": 0.0014, "step": 20574 }, { "epoch": 2.2305941023417173, "grad_norm": 0.22183439135551453, "learning_rate": 5.147976207746989e-06, "loss": 0.0009, "step": 20575 }, { "epoch": 2.2307025151777973, "grad_norm": 0.0037924584466964006, "learning_rate": 5.147250834179602e-06, "loss": 0.0, "step": 20576 }, { "epoch": 2.230810928013877, "grad_norm": 3.695343494415283, "learning_rate": 5.146525460612216e-06, "loss": 0.018, "step": 20577 }, { "epoch": 2.230919340849957, "grad_norm": 0.2973632216453552, "learning_rate": 5.145800087044829e-06, "loss": 0.0067, "step": 20578 }, { "epoch": 2.2310277536860363, "grad_norm": 0.06876819580793381, "learning_rate": 5.145074713477441e-06, "loss": 0.0004, "step": 20579 }, { "epoch": 2.2311361665221163, "grad_norm": 3.3192250728607178, "learning_rate": 5.144349339910054e-06, "loss": 0.0246, "step": 20580 }, { "epoch": 2.231244579358196, "grad_norm": 0.43043646216392517, "learning_rate": 5.143623966342666e-06, "loss": 0.0111, "step": 20581 }, { "epoch": 2.231352992194276, "grad_norm": 0.33418968319892883, "learning_rate": 5.142898592775279e-06, "loss": 0.0114, "step": 20582 }, { "epoch": 2.2314614050303554, "grad_norm": 0.32559993863105774, "learning_rate": 5.142173219207892e-06, "loss": 0.0122, "step": 20583 }, { "epoch": 2.2315698178664354, "grad_norm": 1.828424096107483, "learning_rate": 5.1414478456405045e-06, "loss": 0.0076, "step": 20584 }, { "epoch": 2.2316782307025154, "grad_norm": 1.1497918367385864, "learning_rate": 5.140722472073118e-06, "loss": 0.0429, "step": 20585 }, { "epoch": 2.231786643538595, "grad_norm": 0.34353023767471313, "learning_rate": 5.139997098505731e-06, "loss": 0.009, "step": 20586 }, { "epoch": 2.231895056374675, "grad_norm": 0.4657134711742401, "learning_rate": 5.1392717249383435e-06, "loss": 0.0174, "step": 20587 }, { "epoch": 2.2320034692107544, "grad_norm": 0.336302250623703, "learning_rate": 5.1385463513709565e-06, "loss": 0.0132, "step": 20588 }, { "epoch": 2.2321118820468344, "grad_norm": 0.28560197353363037, "learning_rate": 5.1378209778035695e-06, "loss": 0.0066, "step": 20589 }, { "epoch": 2.232220294882914, "grad_norm": 3.2114906311035156, "learning_rate": 5.137095604236182e-06, "loss": 0.0059, "step": 20590 }, { "epoch": 2.232328707718994, "grad_norm": 0.35957151651382446, "learning_rate": 5.136370230668795e-06, "loss": 0.0133, "step": 20591 }, { "epoch": 2.232437120555074, "grad_norm": 0.6391287446022034, "learning_rate": 5.135644857101408e-06, "loss": 0.0239, "step": 20592 }, { "epoch": 2.2325455333911535, "grad_norm": 3.1131906509399414, "learning_rate": 5.13491948353402e-06, "loss": 0.0654, "step": 20593 }, { "epoch": 2.2326539462272335, "grad_norm": 0.02199387736618519, "learning_rate": 5.134194109966634e-06, "loss": 0.0002, "step": 20594 }, { "epoch": 2.232762359063313, "grad_norm": 0.33353444933891296, "learning_rate": 5.133468736399247e-06, "loss": 0.001, "step": 20595 }, { "epoch": 2.232870771899393, "grad_norm": 4.294694423675537, "learning_rate": 5.132743362831859e-06, "loss": 0.0689, "step": 20596 }, { "epoch": 2.2329791847354725, "grad_norm": 0.2797444462776184, "learning_rate": 5.132017989264472e-06, "loss": 0.0062, "step": 20597 }, { "epoch": 2.2330875975715525, "grad_norm": 1.7088879346847534, "learning_rate": 5.131292615697085e-06, "loss": 0.0296, "step": 20598 }, { "epoch": 2.2331960104076325, "grad_norm": 7.640664577484131, "learning_rate": 5.130567242129697e-06, "loss": 0.0224, "step": 20599 }, { "epoch": 2.233304423243712, "grad_norm": 0.30261948704719543, "learning_rate": 5.12984186856231e-06, "loss": 0.0058, "step": 20600 }, { "epoch": 2.233412836079792, "grad_norm": 0.9644873738288879, "learning_rate": 5.129116494994922e-06, "loss": 0.0196, "step": 20601 }, { "epoch": 2.2335212489158716, "grad_norm": 2.2806878089904785, "learning_rate": 5.128391121427536e-06, "loss": 0.0833, "step": 20602 }, { "epoch": 2.2336296617519515, "grad_norm": 0.24876734614372253, "learning_rate": 5.127665747860149e-06, "loss": 0.0056, "step": 20603 }, { "epoch": 2.233738074588031, "grad_norm": 2.8768515586853027, "learning_rate": 5.126940374292762e-06, "loss": 0.0206, "step": 20604 }, { "epoch": 2.233846487424111, "grad_norm": 0.4644954204559326, "learning_rate": 5.126215000725374e-06, "loss": 0.0094, "step": 20605 }, { "epoch": 2.2339549002601906, "grad_norm": 2.38468337059021, "learning_rate": 5.125489627157987e-06, "loss": 0.0652, "step": 20606 }, { "epoch": 2.2340633130962706, "grad_norm": 0.10007017105817795, "learning_rate": 5.124764253590599e-06, "loss": 0.0101, "step": 20607 }, { "epoch": 2.2341717259323506, "grad_norm": 1.6042593717575073, "learning_rate": 5.124038880023212e-06, "loss": 0.0609, "step": 20608 }, { "epoch": 2.23428013876843, "grad_norm": 0.26495271921157837, "learning_rate": 5.123313506455825e-06, "loss": 0.0062, "step": 20609 }, { "epoch": 2.23438855160451, "grad_norm": 1.110094666481018, "learning_rate": 5.122588132888437e-06, "loss": 0.0037, "step": 20610 }, { "epoch": 2.2344969644405897, "grad_norm": 0.3670090138912201, "learning_rate": 5.121862759321051e-06, "loss": 0.0125, "step": 20611 }, { "epoch": 2.2346053772766696, "grad_norm": 11.159114837646484, "learning_rate": 5.121137385753664e-06, "loss": 0.0333, "step": 20612 }, { "epoch": 2.234713790112749, "grad_norm": 0.01698426343500614, "learning_rate": 5.120412012186276e-06, "loss": 0.0001, "step": 20613 }, { "epoch": 2.234822202948829, "grad_norm": 0.5428294539451599, "learning_rate": 5.119686638618889e-06, "loss": 0.0174, "step": 20614 }, { "epoch": 2.2349306157849087, "grad_norm": 0.1336628645658493, "learning_rate": 5.118961265051502e-06, "loss": 0.0012, "step": 20615 }, { "epoch": 2.2350390286209887, "grad_norm": 0.4076758623123169, "learning_rate": 5.118235891484114e-06, "loss": 0.0147, "step": 20616 }, { "epoch": 2.2351474414570687, "grad_norm": 0.5022906064987183, "learning_rate": 5.117510517916727e-06, "loss": 0.0109, "step": 20617 }, { "epoch": 2.235255854293148, "grad_norm": 1.1356749534606934, "learning_rate": 5.11678514434934e-06, "loss": 0.0141, "step": 20618 }, { "epoch": 2.235364267129228, "grad_norm": 4.734116077423096, "learning_rate": 5.1160597707819524e-06, "loss": 0.0431, "step": 20619 }, { "epoch": 2.2354726799653077, "grad_norm": 0.017113089561462402, "learning_rate": 5.115334397214566e-06, "loss": 0.0002, "step": 20620 }, { "epoch": 2.2355810928013877, "grad_norm": 0.2950685918331146, "learning_rate": 5.114609023647179e-06, "loss": 0.0061, "step": 20621 }, { "epoch": 2.2356895056374677, "grad_norm": 1.4697842597961426, "learning_rate": 5.113883650079791e-06, "loss": 0.0209, "step": 20622 }, { "epoch": 2.2357979184735473, "grad_norm": 2.3624000549316406, "learning_rate": 5.113158276512404e-06, "loss": 0.0187, "step": 20623 }, { "epoch": 2.2359063313096272, "grad_norm": 2.0775911808013916, "learning_rate": 5.112432902945017e-06, "loss": 0.0233, "step": 20624 }, { "epoch": 2.236014744145707, "grad_norm": 0.28913575410842896, "learning_rate": 5.1117075293776295e-06, "loss": 0.0101, "step": 20625 }, { "epoch": 2.2361231569817868, "grad_norm": 1.567888617515564, "learning_rate": 5.1109821558102425e-06, "loss": 0.003, "step": 20626 }, { "epoch": 2.2362315698178663, "grad_norm": 3.0161426067352295, "learning_rate": 5.110256782242855e-06, "loss": 0.0183, "step": 20627 }, { "epoch": 2.2363399826539463, "grad_norm": 1.693634033203125, "learning_rate": 5.1095314086754685e-06, "loss": 0.0398, "step": 20628 }, { "epoch": 2.236448395490026, "grad_norm": 0.2962680160999298, "learning_rate": 5.1088060351080815e-06, "loss": 0.0105, "step": 20629 }, { "epoch": 2.236556808326106, "grad_norm": 1.0905392169952393, "learning_rate": 5.1080806615406945e-06, "loss": 0.0348, "step": 20630 }, { "epoch": 2.236665221162186, "grad_norm": 0.4277307093143463, "learning_rate": 5.107355287973307e-06, "loss": 0.0094, "step": 20631 }, { "epoch": 2.2367736339982653, "grad_norm": 2.936424970626831, "learning_rate": 5.10662991440592e-06, "loss": 0.0322, "step": 20632 }, { "epoch": 2.2368820468343453, "grad_norm": 0.6097865700721741, "learning_rate": 5.105904540838532e-06, "loss": 0.0294, "step": 20633 }, { "epoch": 2.236990459670425, "grad_norm": 0.5054635405540466, "learning_rate": 5.105179167271145e-06, "loss": 0.0136, "step": 20634 }, { "epoch": 2.237098872506505, "grad_norm": 2.0983593463897705, "learning_rate": 5.104453793703758e-06, "loss": 0.0071, "step": 20635 }, { "epoch": 2.2372072853425844, "grad_norm": 0.2758704721927643, "learning_rate": 5.10372842013637e-06, "loss": 0.0083, "step": 20636 }, { "epoch": 2.2373156981786644, "grad_norm": 0.02478797174990177, "learning_rate": 5.103003046568984e-06, "loss": 0.0003, "step": 20637 }, { "epoch": 2.237424111014744, "grad_norm": 2.934674024581909, "learning_rate": 5.102277673001597e-06, "loss": 0.0108, "step": 20638 }, { "epoch": 2.237532523850824, "grad_norm": 1.7751528024673462, "learning_rate": 5.101552299434209e-06, "loss": 0.0216, "step": 20639 }, { "epoch": 2.237640936686904, "grad_norm": 0.522290825843811, "learning_rate": 5.100826925866822e-06, "loss": 0.0031, "step": 20640 }, { "epoch": 2.2377493495229834, "grad_norm": 0.5433348417282104, "learning_rate": 5.100101552299435e-06, "loss": 0.0153, "step": 20641 }, { "epoch": 2.2378577623590634, "grad_norm": 0.22247852385044098, "learning_rate": 5.099376178732047e-06, "loss": 0.0171, "step": 20642 }, { "epoch": 2.237966175195143, "grad_norm": 1.2548857927322388, "learning_rate": 5.09865080516466e-06, "loss": 0.022, "step": 20643 }, { "epoch": 2.238074588031223, "grad_norm": 0.541542112827301, "learning_rate": 5.097925431597273e-06, "loss": 0.0219, "step": 20644 }, { "epoch": 2.2381830008673025, "grad_norm": 2.406400680541992, "learning_rate": 5.097200058029885e-06, "loss": 0.015, "step": 20645 }, { "epoch": 2.2382914137033825, "grad_norm": 4.908295154571533, "learning_rate": 5.096474684462499e-06, "loss": 0.0039, "step": 20646 }, { "epoch": 2.2383998265394625, "grad_norm": 1.1308079957962036, "learning_rate": 5.095749310895112e-06, "loss": 0.0204, "step": 20647 }, { "epoch": 2.238508239375542, "grad_norm": 0.48189887404441833, "learning_rate": 5.095023937327724e-06, "loss": 0.0144, "step": 20648 }, { "epoch": 2.238616652211622, "grad_norm": 0.42389941215515137, "learning_rate": 5.094298563760337e-06, "loss": 0.0084, "step": 20649 }, { "epoch": 2.2387250650477015, "grad_norm": 0.20480641722679138, "learning_rate": 5.09357319019295e-06, "loss": 0.0141, "step": 20650 }, { "epoch": 2.2388334778837815, "grad_norm": 0.32065561413764954, "learning_rate": 5.092847816625562e-06, "loss": 0.0067, "step": 20651 }, { "epoch": 2.238941890719861, "grad_norm": 0.4012676477432251, "learning_rate": 5.092122443058175e-06, "loss": 0.0077, "step": 20652 }, { "epoch": 2.239050303555941, "grad_norm": 0.07565521448850632, "learning_rate": 5.091397069490787e-06, "loss": 0.0007, "step": 20653 }, { "epoch": 2.239158716392021, "grad_norm": 2.32248592376709, "learning_rate": 5.090671695923401e-06, "loss": 0.0045, "step": 20654 }, { "epoch": 2.2392671292281006, "grad_norm": 0.3391744792461395, "learning_rate": 5.089946322356014e-06, "loss": 0.011, "step": 20655 }, { "epoch": 2.2393755420641805, "grad_norm": 0.3377268314361572, "learning_rate": 5.089220948788627e-06, "loss": 0.0206, "step": 20656 }, { "epoch": 2.23948395490026, "grad_norm": 0.6270862817764282, "learning_rate": 5.088495575221239e-06, "loss": 0.0071, "step": 20657 }, { "epoch": 2.23959236773634, "grad_norm": 4.0925517082214355, "learning_rate": 5.087770201653852e-06, "loss": 0.0276, "step": 20658 }, { "epoch": 2.2397007805724196, "grad_norm": 0.3027787208557129, "learning_rate": 5.0870448280864644e-06, "loss": 0.0009, "step": 20659 }, { "epoch": 2.2398091934084996, "grad_norm": 5.579124927520752, "learning_rate": 5.086319454519077e-06, "loss": 0.0091, "step": 20660 }, { "epoch": 2.239917606244579, "grad_norm": 0.39445069432258606, "learning_rate": 5.08559408095169e-06, "loss": 0.0063, "step": 20661 }, { "epoch": 2.240026019080659, "grad_norm": 2.2042791843414307, "learning_rate": 5.0848687073843026e-06, "loss": 0.0018, "step": 20662 }, { "epoch": 2.240134431916739, "grad_norm": 1.784169316291809, "learning_rate": 5.084143333816916e-06, "loss": 0.0078, "step": 20663 }, { "epoch": 2.2402428447528187, "grad_norm": 0.27073585987091064, "learning_rate": 5.083417960249529e-06, "loss": 0.0056, "step": 20664 }, { "epoch": 2.2403512575888986, "grad_norm": 2.894887924194336, "learning_rate": 5.0826925866821415e-06, "loss": 0.0106, "step": 20665 }, { "epoch": 2.240459670424978, "grad_norm": 0.9300939440727234, "learning_rate": 5.0819672131147545e-06, "loss": 0.0078, "step": 20666 }, { "epoch": 2.240568083261058, "grad_norm": 6.498395919799805, "learning_rate": 5.0812418395473675e-06, "loss": 0.0285, "step": 20667 }, { "epoch": 2.2406764960971377, "grad_norm": 0.28943517804145813, "learning_rate": 5.08051646597998e-06, "loss": 0.0066, "step": 20668 }, { "epoch": 2.2407849089332177, "grad_norm": 5.512923717498779, "learning_rate": 5.079791092412593e-06, "loss": 0.041, "step": 20669 }, { "epoch": 2.2408933217692977, "grad_norm": 0.21947580575942993, "learning_rate": 5.079065718845206e-06, "loss": 0.0044, "step": 20670 }, { "epoch": 2.241001734605377, "grad_norm": 12.23426628112793, "learning_rate": 5.078340345277818e-06, "loss": 0.029, "step": 20671 }, { "epoch": 2.241110147441457, "grad_norm": 6.907658576965332, "learning_rate": 5.077614971710432e-06, "loss": 0.03, "step": 20672 }, { "epoch": 2.2412185602775367, "grad_norm": 0.32986098527908325, "learning_rate": 5.076889598143045e-06, "loss": 0.0056, "step": 20673 }, { "epoch": 2.2413269731136167, "grad_norm": 0.8471727967262268, "learning_rate": 5.076164224575657e-06, "loss": 0.0066, "step": 20674 }, { "epoch": 2.2414353859496963, "grad_norm": 2.2925989627838135, "learning_rate": 5.07543885100827e-06, "loss": 0.0259, "step": 20675 }, { "epoch": 2.2415437987857763, "grad_norm": 0.3610804080963135, "learning_rate": 5.074713477440883e-06, "loss": 0.0068, "step": 20676 }, { "epoch": 2.2416522116218562, "grad_norm": 0.37541723251342773, "learning_rate": 5.073988103873495e-06, "loss": 0.0131, "step": 20677 }, { "epoch": 2.241760624457936, "grad_norm": 0.016995470970869064, "learning_rate": 5.073262730306108e-06, "loss": 0.0001, "step": 20678 }, { "epoch": 2.2418690372940158, "grad_norm": 15.257994651794434, "learning_rate": 5.07253735673872e-06, "loss": 0.0152, "step": 20679 }, { "epoch": 2.2419774501300953, "grad_norm": 0.8931142091751099, "learning_rate": 5.071811983171334e-06, "loss": 0.0089, "step": 20680 }, { "epoch": 2.2420858629661753, "grad_norm": 0.36691784858703613, "learning_rate": 5.071086609603947e-06, "loss": 0.0066, "step": 20681 }, { "epoch": 2.242194275802255, "grad_norm": 1.1760036945343018, "learning_rate": 5.07036123603656e-06, "loss": 0.0348, "step": 20682 }, { "epoch": 2.242302688638335, "grad_norm": 0.020328402519226074, "learning_rate": 5.069635862469172e-06, "loss": 0.0001, "step": 20683 }, { "epoch": 2.2424111014744144, "grad_norm": 2.1939518451690674, "learning_rate": 5.068910488901785e-06, "loss": 0.015, "step": 20684 }, { "epoch": 2.2425195143104943, "grad_norm": 5.815206050872803, "learning_rate": 5.068185115334397e-06, "loss": 0.0029, "step": 20685 }, { "epoch": 2.2426279271465743, "grad_norm": 1.4210401773452759, "learning_rate": 5.06745974176701e-06, "loss": 0.0046, "step": 20686 }, { "epoch": 2.242736339982654, "grad_norm": 0.07987097650766373, "learning_rate": 5.066734368199623e-06, "loss": 0.0006, "step": 20687 }, { "epoch": 2.242844752818734, "grad_norm": 0.03689919784665108, "learning_rate": 5.066008994632235e-06, "loss": 0.0004, "step": 20688 }, { "epoch": 2.2429531656548134, "grad_norm": 0.26207852363586426, "learning_rate": 5.065283621064849e-06, "loss": 0.0163, "step": 20689 }, { "epoch": 2.2430615784908934, "grad_norm": 0.3974672853946686, "learning_rate": 5.064558247497462e-06, "loss": 0.0154, "step": 20690 }, { "epoch": 2.243169991326973, "grad_norm": 0.42761659622192383, "learning_rate": 5.063832873930074e-06, "loss": 0.0067, "step": 20691 }, { "epoch": 2.243278404163053, "grad_norm": 0.27333369851112366, "learning_rate": 5.063107500362687e-06, "loss": 0.0063, "step": 20692 }, { "epoch": 2.243386816999133, "grad_norm": 0.3049550950527191, "learning_rate": 5.0623821267953e-06, "loss": 0.0053, "step": 20693 }, { "epoch": 2.2434952298352124, "grad_norm": 0.8652254343032837, "learning_rate": 5.061656753227912e-06, "loss": 0.0022, "step": 20694 }, { "epoch": 2.2436036426712924, "grad_norm": 2.1996958255767822, "learning_rate": 5.060931379660525e-06, "loss": 0.0349, "step": 20695 }, { "epoch": 2.243712055507372, "grad_norm": 0.3699399530887604, "learning_rate": 5.060206006093138e-06, "loss": 0.0068, "step": 20696 }, { "epoch": 2.243820468343452, "grad_norm": 2.7557361125946045, "learning_rate": 5.059480632525752e-06, "loss": 0.0344, "step": 20697 }, { "epoch": 2.2439288811795315, "grad_norm": 22.7434024810791, "learning_rate": 5.058755258958364e-06, "loss": 0.0332, "step": 20698 }, { "epoch": 2.2440372940156115, "grad_norm": 0.4404805302619934, "learning_rate": 5.058029885390977e-06, "loss": 0.006, "step": 20699 }, { "epoch": 2.2441457068516915, "grad_norm": 3.61672043800354, "learning_rate": 5.057304511823589e-06, "loss": 0.0182, "step": 20700 }, { "epoch": 2.244254119687771, "grad_norm": 5.740355014801025, "learning_rate": 5.056579138256202e-06, "loss": 0.0249, "step": 20701 }, { "epoch": 2.244362532523851, "grad_norm": 0.30796536803245544, "learning_rate": 5.055853764688815e-06, "loss": 0.011, "step": 20702 }, { "epoch": 2.2444709453599305, "grad_norm": 0.1766539365053177, "learning_rate": 5.0551283911214275e-06, "loss": 0.0009, "step": 20703 }, { "epoch": 2.2445793581960105, "grad_norm": 0.07148716598749161, "learning_rate": 5.0544030175540405e-06, "loss": 0.0003, "step": 20704 }, { "epoch": 2.24468777103209, "grad_norm": 0.07154372334480286, "learning_rate": 5.053677643986653e-06, "loss": 0.0113, "step": 20705 }, { "epoch": 2.24479618386817, "grad_norm": 0.2645902633666992, "learning_rate": 5.0529522704192665e-06, "loss": 0.0055, "step": 20706 }, { "epoch": 2.2449045967042496, "grad_norm": 0.29646751284599304, "learning_rate": 5.0522268968518795e-06, "loss": 0.0105, "step": 20707 }, { "epoch": 2.2450130095403296, "grad_norm": 0.3179917633533478, "learning_rate": 5.0515015232844925e-06, "loss": 0.0132, "step": 20708 }, { "epoch": 2.2451214223764095, "grad_norm": 1.2777315378189087, "learning_rate": 5.050776149717105e-06, "loss": 0.028, "step": 20709 }, { "epoch": 2.245229835212489, "grad_norm": 0.5070168375968933, "learning_rate": 5.050050776149718e-06, "loss": 0.0173, "step": 20710 }, { "epoch": 2.245338248048569, "grad_norm": 0.3345490097999573, "learning_rate": 5.04932540258233e-06, "loss": 0.0109, "step": 20711 }, { "epoch": 2.2454466608846486, "grad_norm": 0.6487373113632202, "learning_rate": 5.048600029014943e-06, "loss": 0.0082, "step": 20712 }, { "epoch": 2.2455550737207286, "grad_norm": 0.7988582849502563, "learning_rate": 5.047874655447556e-06, "loss": 0.0087, "step": 20713 }, { "epoch": 2.245663486556808, "grad_norm": 0.663754940032959, "learning_rate": 5.047149281880168e-06, "loss": 0.0358, "step": 20714 }, { "epoch": 2.245771899392888, "grad_norm": 0.906876266002655, "learning_rate": 5.046423908312782e-06, "loss": 0.0147, "step": 20715 }, { "epoch": 2.245880312228968, "grad_norm": 0.06558161973953247, "learning_rate": 5.045698534745395e-06, "loss": 0.0006, "step": 20716 }, { "epoch": 2.2459887250650477, "grad_norm": 1.0516313314437866, "learning_rate": 5.044973161178007e-06, "loss": 0.051, "step": 20717 }, { "epoch": 2.2460971379011276, "grad_norm": 0.6965117454528809, "learning_rate": 5.04424778761062e-06, "loss": 0.0264, "step": 20718 }, { "epoch": 2.246205550737207, "grad_norm": 1.6385200023651123, "learning_rate": 5.043522414043233e-06, "loss": 0.021, "step": 20719 }, { "epoch": 2.246313963573287, "grad_norm": 0.2857458293437958, "learning_rate": 5.042797040475845e-06, "loss": 0.0014, "step": 20720 }, { "epoch": 2.2464223764093667, "grad_norm": 2.110583782196045, "learning_rate": 5.042071666908458e-06, "loss": 0.0454, "step": 20721 }, { "epoch": 2.2465307892454467, "grad_norm": 0.19411367177963257, "learning_rate": 5.041346293341071e-06, "loss": 0.0014, "step": 20722 }, { "epoch": 2.2466392020815267, "grad_norm": 1.671119213104248, "learning_rate": 5.040620919773685e-06, "loss": 0.0044, "step": 20723 }, { "epoch": 2.246747614917606, "grad_norm": 0.4105513095855713, "learning_rate": 5.039895546206297e-06, "loss": 0.0079, "step": 20724 }, { "epoch": 2.246856027753686, "grad_norm": 0.9666100144386292, "learning_rate": 5.03917017263891e-06, "loss": 0.037, "step": 20725 }, { "epoch": 2.2469644405897657, "grad_norm": 0.28071868419647217, "learning_rate": 5.038444799071522e-06, "loss": 0.0015, "step": 20726 }, { "epoch": 2.2470728534258457, "grad_norm": 2.4357564449310303, "learning_rate": 5.037719425504135e-06, "loss": 0.0087, "step": 20727 }, { "epoch": 2.2471812662619253, "grad_norm": 0.27765119075775146, "learning_rate": 5.036994051936748e-06, "loss": 0.0116, "step": 20728 }, { "epoch": 2.2472896790980053, "grad_norm": 0.8576947450637817, "learning_rate": 5.03626867836936e-06, "loss": 0.009, "step": 20729 }, { "epoch": 2.247398091934085, "grad_norm": 0.3153248131275177, "learning_rate": 5.035543304801973e-06, "loss": 0.0104, "step": 20730 }, { "epoch": 2.247506504770165, "grad_norm": 0.948584258556366, "learning_rate": 5.034817931234585e-06, "loss": 0.0039, "step": 20731 }, { "epoch": 2.2476149176062448, "grad_norm": 0.4189906120300293, "learning_rate": 5.034092557667199e-06, "loss": 0.0013, "step": 20732 }, { "epoch": 2.2477233304423243, "grad_norm": 0.38773781061172485, "learning_rate": 5.033367184099812e-06, "loss": 0.0117, "step": 20733 }, { "epoch": 2.2478317432784043, "grad_norm": 0.3884389102458954, "learning_rate": 5.032641810532425e-06, "loss": 0.0012, "step": 20734 }, { "epoch": 2.247940156114484, "grad_norm": 1.5680103302001953, "learning_rate": 5.031916436965037e-06, "loss": 0.0283, "step": 20735 }, { "epoch": 2.248048568950564, "grad_norm": 0.38002780079841614, "learning_rate": 5.03119106339765e-06, "loss": 0.0009, "step": 20736 }, { "epoch": 2.2481569817866434, "grad_norm": 0.35132795572280884, "learning_rate": 5.0304656898302624e-06, "loss": 0.0061, "step": 20737 }, { "epoch": 2.2482653946227233, "grad_norm": 0.3821853995323181, "learning_rate": 5.0297403162628754e-06, "loss": 0.0018, "step": 20738 }, { "epoch": 2.248373807458803, "grad_norm": 0.6321619749069214, "learning_rate": 5.029014942695488e-06, "loss": 0.0218, "step": 20739 }, { "epoch": 2.248482220294883, "grad_norm": 0.9138286709785461, "learning_rate": 5.0282895691281006e-06, "loss": 0.0088, "step": 20740 }, { "epoch": 2.248590633130963, "grad_norm": 0.4464658200740814, "learning_rate": 5.027564195560714e-06, "loss": 0.0184, "step": 20741 }, { "epoch": 2.2486990459670424, "grad_norm": 0.25818175077438354, "learning_rate": 5.026838821993327e-06, "loss": 0.0014, "step": 20742 }, { "epoch": 2.2488074588031224, "grad_norm": 0.008806874044239521, "learning_rate": 5.02611344842594e-06, "loss": 0.0001, "step": 20743 }, { "epoch": 2.248915871639202, "grad_norm": 0.012982550077140331, "learning_rate": 5.0253880748585525e-06, "loss": 0.0001, "step": 20744 }, { "epoch": 2.249024284475282, "grad_norm": 3.599250555038452, "learning_rate": 5.0246627012911655e-06, "loss": 0.0062, "step": 20745 }, { "epoch": 2.249132697311362, "grad_norm": 0.05214928463101387, "learning_rate": 5.023937327723778e-06, "loss": 0.0006, "step": 20746 }, { "epoch": 2.2492411101474414, "grad_norm": 5.78755521774292, "learning_rate": 5.023211954156391e-06, "loss": 0.0233, "step": 20747 }, { "epoch": 2.2493495229835214, "grad_norm": 0.08628672361373901, "learning_rate": 5.022486580589004e-06, "loss": 0.0006, "step": 20748 }, { "epoch": 2.249457935819601, "grad_norm": 0.24753332138061523, "learning_rate": 5.0217612070216175e-06, "loss": 0.0056, "step": 20749 }, { "epoch": 2.249566348655681, "grad_norm": 0.39235642552375793, "learning_rate": 5.02103583345423e-06, "loss": 0.0069, "step": 20750 }, { "epoch": 2.2496747614917605, "grad_norm": 0.4345691502094269, "learning_rate": 5.020310459886843e-06, "loss": 0.0089, "step": 20751 }, { "epoch": 2.2497831743278405, "grad_norm": 1.0933901071548462, "learning_rate": 5.019585086319455e-06, "loss": 0.0164, "step": 20752 }, { "epoch": 2.24989158716392, "grad_norm": 0.3867237865924835, "learning_rate": 5.018859712752068e-06, "loss": 0.0116, "step": 20753 }, { "epoch": 2.25, "grad_norm": 6.078932285308838, "learning_rate": 5.018134339184681e-06, "loss": 0.0122, "step": 20754 }, { "epoch": 2.25010841283608, "grad_norm": 0.5926681756973267, "learning_rate": 5.017408965617293e-06, "loss": 0.0065, "step": 20755 }, { "epoch": 2.2502168256721595, "grad_norm": 1.3996930122375488, "learning_rate": 5.016683592049906e-06, "loss": 0.0163, "step": 20756 }, { "epoch": 2.2503252385082395, "grad_norm": 0.27925625443458557, "learning_rate": 5.015958218482518e-06, "loss": 0.0061, "step": 20757 }, { "epoch": 2.250433651344319, "grad_norm": 0.04839608445763588, "learning_rate": 5.015232844915132e-06, "loss": 0.0004, "step": 20758 }, { "epoch": 2.250542064180399, "grad_norm": 0.30748456716537476, "learning_rate": 5.014507471347745e-06, "loss": 0.0065, "step": 20759 }, { "epoch": 2.2506504770164786, "grad_norm": 0.36955711245536804, "learning_rate": 5.013782097780358e-06, "loss": 0.0105, "step": 20760 }, { "epoch": 2.2507588898525586, "grad_norm": 0.2781207263469696, "learning_rate": 5.01305672421297e-06, "loss": 0.0014, "step": 20761 }, { "epoch": 2.250867302688638, "grad_norm": 1.0029525756835938, "learning_rate": 5.012331350645583e-06, "loss": 0.002, "step": 20762 }, { "epoch": 2.250975715524718, "grad_norm": 1.8254485130310059, "learning_rate": 5.011605977078195e-06, "loss": 0.0128, "step": 20763 }, { "epoch": 2.251084128360798, "grad_norm": 0.33893904089927673, "learning_rate": 5.010880603510808e-06, "loss": 0.0147, "step": 20764 }, { "epoch": 2.2511925411968776, "grad_norm": 0.6932057738304138, "learning_rate": 5.010155229943421e-06, "loss": 0.0184, "step": 20765 }, { "epoch": 2.2513009540329576, "grad_norm": 0.18626578152179718, "learning_rate": 5.009429856376033e-06, "loss": 0.001, "step": 20766 }, { "epoch": 2.251409366869037, "grad_norm": 0.3774231970310211, "learning_rate": 5.008704482808647e-06, "loss": 0.015, "step": 20767 }, { "epoch": 2.251517779705117, "grad_norm": 14.511615753173828, "learning_rate": 5.00797910924126e-06, "loss": 0.0397, "step": 20768 }, { "epoch": 2.251626192541197, "grad_norm": 0.27765700221061707, "learning_rate": 5.007253735673873e-06, "loss": 0.0068, "step": 20769 }, { "epoch": 2.2517346053772767, "grad_norm": 0.5680525302886963, "learning_rate": 5.006528362106485e-06, "loss": 0.0183, "step": 20770 }, { "epoch": 2.2518430182133566, "grad_norm": 0.2782040536403656, "learning_rate": 5.005802988539098e-06, "loss": 0.0053, "step": 20771 }, { "epoch": 2.251951431049436, "grad_norm": 6.006999492645264, "learning_rate": 5.00507761497171e-06, "loss": 0.0125, "step": 20772 }, { "epoch": 2.252059843885516, "grad_norm": 3.358154296875, "learning_rate": 5.004352241404323e-06, "loss": 0.0078, "step": 20773 }, { "epoch": 2.2521682567215957, "grad_norm": 0.4601897597312927, "learning_rate": 5.003626867836936e-06, "loss": 0.0137, "step": 20774 }, { "epoch": 2.2522766695576757, "grad_norm": 0.3824455142021179, "learning_rate": 5.00290149426955e-06, "loss": 0.0117, "step": 20775 }, { "epoch": 2.2523850823937552, "grad_norm": 0.06431163847446442, "learning_rate": 5.002176120702162e-06, "loss": 0.0004, "step": 20776 }, { "epoch": 2.252493495229835, "grad_norm": 2.083686351776123, "learning_rate": 5.001450747134775e-06, "loss": 0.0346, "step": 20777 }, { "epoch": 2.252601908065915, "grad_norm": 0.7689279317855835, "learning_rate": 5.000725373567387e-06, "loss": 0.0022, "step": 20778 }, { "epoch": 2.2527103209019947, "grad_norm": 0.2939516305923462, "learning_rate": 5e-06, "loss": 0.0066, "step": 20779 }, { "epoch": 2.2528187337380747, "grad_norm": 2.595827341079712, "learning_rate": 4.999274626432613e-06, "loss": 0.015, "step": 20780 }, { "epoch": 2.2529271465741543, "grad_norm": 0.4752836227416992, "learning_rate": 4.9985492528652255e-06, "loss": 0.0175, "step": 20781 }, { "epoch": 2.2530355594102343, "grad_norm": 0.07325462996959686, "learning_rate": 4.997823879297839e-06, "loss": 0.0005, "step": 20782 }, { "epoch": 2.253143972246314, "grad_norm": 4.4232258796691895, "learning_rate": 4.9970985057304515e-06, "loss": 0.0483, "step": 20783 }, { "epoch": 2.253252385082394, "grad_norm": 0.011719624511897564, "learning_rate": 4.9963731321630645e-06, "loss": 0.0001, "step": 20784 }, { "epoch": 2.2533607979184733, "grad_norm": 0.4199327528476715, "learning_rate": 4.995647758595677e-06, "loss": 0.0142, "step": 20785 }, { "epoch": 2.2534692107545533, "grad_norm": 0.04045351222157478, "learning_rate": 4.9949223850282905e-06, "loss": 0.0003, "step": 20786 }, { "epoch": 2.2535776235906333, "grad_norm": 0.19621194899082184, "learning_rate": 4.994197011460903e-06, "loss": 0.0039, "step": 20787 }, { "epoch": 2.253686036426713, "grad_norm": 0.26245370507240295, "learning_rate": 4.993471637893516e-06, "loss": 0.0003, "step": 20788 }, { "epoch": 2.253794449262793, "grad_norm": 0.29449689388275146, "learning_rate": 4.992746264326128e-06, "loss": 0.0061, "step": 20789 }, { "epoch": 2.2539028620988724, "grad_norm": 0.030307114124298096, "learning_rate": 4.992020890758742e-06, "loss": 0.0002, "step": 20790 }, { "epoch": 2.2540112749349523, "grad_norm": 0.17585039138793945, "learning_rate": 4.991295517191354e-06, "loss": 0.0034, "step": 20791 }, { "epoch": 2.2541196877710323, "grad_norm": 0.2339983433485031, "learning_rate": 4.990570143623967e-06, "loss": 0.0052, "step": 20792 }, { "epoch": 2.254228100607112, "grad_norm": 0.28351548314094543, "learning_rate": 4.98984477005658e-06, "loss": 0.0047, "step": 20793 }, { "epoch": 2.254336513443192, "grad_norm": 4.6230692863464355, "learning_rate": 4.989119396489192e-06, "loss": 0.018, "step": 20794 }, { "epoch": 2.2544449262792714, "grad_norm": 0.03487059473991394, "learning_rate": 4.988394022921806e-06, "loss": 0.0003, "step": 20795 }, { "epoch": 2.2545533391153514, "grad_norm": 2.019109010696411, "learning_rate": 4.987668649354418e-06, "loss": 0.0398, "step": 20796 }, { "epoch": 2.254661751951431, "grad_norm": 0.053496383130550385, "learning_rate": 4.986943275787031e-06, "loss": 0.0003, "step": 20797 }, { "epoch": 2.254770164787511, "grad_norm": 0.15247805416584015, "learning_rate": 4.986217902219643e-06, "loss": 0.0012, "step": 20798 }, { "epoch": 2.2548785776235905, "grad_norm": 0.7976097464561462, "learning_rate": 4.985492528652257e-06, "loss": 0.02, "step": 20799 }, { "epoch": 2.2549869904596704, "grad_norm": 0.25508973002433777, "learning_rate": 4.984767155084869e-06, "loss": 0.0136, "step": 20800 }, { "epoch": 2.2550954032957504, "grad_norm": 6.491713523864746, "learning_rate": 4.984041781517482e-06, "loss": 0.0476, "step": 20801 }, { "epoch": 2.25520381613183, "grad_norm": 0.1052839532494545, "learning_rate": 4.983316407950094e-06, "loss": 0.0003, "step": 20802 }, { "epoch": 2.25531222896791, "grad_norm": 0.11340653896331787, "learning_rate": 4.982591034382708e-06, "loss": 0.0002, "step": 20803 }, { "epoch": 2.2554206418039895, "grad_norm": 0.17563645541667938, "learning_rate": 4.98186566081532e-06, "loss": 0.0036, "step": 20804 }, { "epoch": 2.2555290546400695, "grad_norm": 0.5466635227203369, "learning_rate": 4.981140287247933e-06, "loss": 0.0094, "step": 20805 }, { "epoch": 2.255637467476149, "grad_norm": 1.9945818185806274, "learning_rate": 4.980414913680546e-06, "loss": 0.0364, "step": 20806 }, { "epoch": 2.255745880312229, "grad_norm": 0.35962608456611633, "learning_rate": 4.979689540113158e-06, "loss": 0.0068, "step": 20807 }, { "epoch": 2.2558542931483085, "grad_norm": 0.04499334469437599, "learning_rate": 4.978964166545772e-06, "loss": 0.0002, "step": 20808 }, { "epoch": 2.2559627059843885, "grad_norm": 0.028701938688755035, "learning_rate": 4.978238792978384e-06, "loss": 0.0003, "step": 20809 }, { "epoch": 2.2560711188204685, "grad_norm": 0.09500668197870255, "learning_rate": 4.977513419410997e-06, "loss": 0.0007, "step": 20810 }, { "epoch": 2.256179531656548, "grad_norm": 0.11612053960561752, "learning_rate": 4.976788045843609e-06, "loss": 0.0006, "step": 20811 }, { "epoch": 2.256287944492628, "grad_norm": 0.45336246490478516, "learning_rate": 4.976062672276223e-06, "loss": 0.0093, "step": 20812 }, { "epoch": 2.2563963573287076, "grad_norm": 0.5384462475776672, "learning_rate": 4.975337298708835e-06, "loss": 0.0054, "step": 20813 }, { "epoch": 2.2565047701647876, "grad_norm": 0.5871866345405579, "learning_rate": 4.974611925141448e-06, "loss": 0.0072, "step": 20814 }, { "epoch": 2.2566131830008676, "grad_norm": 7.258268356323242, "learning_rate": 4.9738865515740604e-06, "loss": 0.0234, "step": 20815 }, { "epoch": 2.256721595836947, "grad_norm": 4.45327091217041, "learning_rate": 4.973161178006674e-06, "loss": 0.0256, "step": 20816 }, { "epoch": 2.256830008673027, "grad_norm": 0.5833793878555298, "learning_rate": 4.972435804439286e-06, "loss": 0.0116, "step": 20817 }, { "epoch": 2.2569384215091066, "grad_norm": 0.2592276632785797, "learning_rate": 4.971710430871899e-06, "loss": 0.0009, "step": 20818 }, { "epoch": 2.2570468343451866, "grad_norm": 5.988441467285156, "learning_rate": 4.970985057304512e-06, "loss": 0.0522, "step": 20819 }, { "epoch": 2.257155247181266, "grad_norm": 0.11216667294502258, "learning_rate": 4.9702596837371245e-06, "loss": 0.0005, "step": 20820 }, { "epoch": 2.257263660017346, "grad_norm": 3.833787441253662, "learning_rate": 4.969534310169738e-06, "loss": 0.0255, "step": 20821 }, { "epoch": 2.2573720728534257, "grad_norm": 0.031941212713718414, "learning_rate": 4.9688089366023505e-06, "loss": 0.0004, "step": 20822 }, { "epoch": 2.2574804856895057, "grad_norm": 0.39223214983940125, "learning_rate": 4.9680835630349635e-06, "loss": 0.0098, "step": 20823 }, { "epoch": 2.2575888985255856, "grad_norm": 0.3501564860343933, "learning_rate": 4.967358189467576e-06, "loss": 0.0008, "step": 20824 }, { "epoch": 2.257697311361665, "grad_norm": 0.1758488118648529, "learning_rate": 4.9666328159001895e-06, "loss": 0.0034, "step": 20825 }, { "epoch": 2.257805724197745, "grad_norm": 3.215062141418457, "learning_rate": 4.965907442332802e-06, "loss": 0.0547, "step": 20826 }, { "epoch": 2.2579141370338247, "grad_norm": 0.3285597264766693, "learning_rate": 4.965182068765415e-06, "loss": 0.0085, "step": 20827 }, { "epoch": 2.2580225498699047, "grad_norm": 1.9337236881256104, "learning_rate": 4.964456695198027e-06, "loss": 0.0303, "step": 20828 }, { "epoch": 2.2581309627059842, "grad_norm": 11.817338943481445, "learning_rate": 4.963731321630641e-06, "loss": 0.0149, "step": 20829 }, { "epoch": 2.258239375542064, "grad_norm": 0.6317055225372314, "learning_rate": 4.963005948063253e-06, "loss": 0.0085, "step": 20830 }, { "epoch": 2.2583477883781438, "grad_norm": 0.013789305463433266, "learning_rate": 4.962280574495866e-06, "loss": 0.0001, "step": 20831 }, { "epoch": 2.2584562012142237, "grad_norm": 0.5997750759124756, "learning_rate": 4.961555200928479e-06, "loss": 0.0199, "step": 20832 }, { "epoch": 2.2585646140503037, "grad_norm": 0.8320649266242981, "learning_rate": 4.960829827361092e-06, "loss": 0.0118, "step": 20833 }, { "epoch": 2.2586730268863833, "grad_norm": 0.0231182724237442, "learning_rate": 4.960104453793705e-06, "loss": 0.0002, "step": 20834 }, { "epoch": 2.2587814397224633, "grad_norm": 2.3531572818756104, "learning_rate": 4.959379080226317e-06, "loss": 0.016, "step": 20835 }, { "epoch": 2.258889852558543, "grad_norm": 0.31828007102012634, "learning_rate": 4.95865370665893e-06, "loss": 0.0015, "step": 20836 }, { "epoch": 2.258998265394623, "grad_norm": 0.7986395955085754, "learning_rate": 4.957928333091542e-06, "loss": 0.0235, "step": 20837 }, { "epoch": 2.2591066782307023, "grad_norm": 1.8791455030441284, "learning_rate": 4.957202959524156e-06, "loss": 0.0012, "step": 20838 }, { "epoch": 2.2592150910667823, "grad_norm": 0.0678570494055748, "learning_rate": 4.956477585956768e-06, "loss": 0.0008, "step": 20839 }, { "epoch": 2.259323503902862, "grad_norm": 0.10701631754636765, "learning_rate": 4.955752212389381e-06, "loss": 0.0005, "step": 20840 }, { "epoch": 2.259431916738942, "grad_norm": 0.09631455689668655, "learning_rate": 4.955026838821994e-06, "loss": 0.0005, "step": 20841 }, { "epoch": 2.259540329575022, "grad_norm": 0.34661048650741577, "learning_rate": 4.954301465254607e-06, "loss": 0.005, "step": 20842 }, { "epoch": 2.2596487424111014, "grad_norm": 0.35408246517181396, "learning_rate": 4.953576091687219e-06, "loss": 0.0126, "step": 20843 }, { "epoch": 2.2597571552471813, "grad_norm": 2.333855390548706, "learning_rate": 4.952850718119832e-06, "loss": 0.0331, "step": 20844 }, { "epoch": 2.259865568083261, "grad_norm": 0.42665234208106995, "learning_rate": 4.952125344552445e-06, "loss": 0.0079, "step": 20845 }, { "epoch": 2.259973980919341, "grad_norm": 0.2533227503299713, "learning_rate": 4.951399970985058e-06, "loss": 0.0012, "step": 20846 }, { "epoch": 2.260082393755421, "grad_norm": 0.3046484887599945, "learning_rate": 4.950674597417671e-06, "loss": 0.0198, "step": 20847 }, { "epoch": 2.2601908065915004, "grad_norm": 0.12135916203260422, "learning_rate": 4.949949223850283e-06, "loss": 0.0009, "step": 20848 }, { "epoch": 2.2602992194275804, "grad_norm": 0.45167821645736694, "learning_rate": 4.949223850282896e-06, "loss": 0.0178, "step": 20849 }, { "epoch": 2.26040763226366, "grad_norm": 1.7471842765808105, "learning_rate": 4.948498476715508e-06, "loss": 0.0374, "step": 20850 }, { "epoch": 2.26051604509974, "grad_norm": 0.3629082143306732, "learning_rate": 4.947773103148122e-06, "loss": 0.0123, "step": 20851 }, { "epoch": 2.2606244579358195, "grad_norm": 0.6814035177230835, "learning_rate": 4.947047729580734e-06, "loss": 0.0017, "step": 20852 }, { "epoch": 2.2607328707718994, "grad_norm": 0.6040683388710022, "learning_rate": 4.946322356013347e-06, "loss": 0.0014, "step": 20853 }, { "epoch": 2.260841283607979, "grad_norm": 0.03162248060107231, "learning_rate": 4.94559698244596e-06, "loss": 0.0003, "step": 20854 }, { "epoch": 2.260949696444059, "grad_norm": 0.8049901723861694, "learning_rate": 4.944871608878573e-06, "loss": 0.0081, "step": 20855 }, { "epoch": 2.261058109280139, "grad_norm": 4.377529144287109, "learning_rate": 4.9441462353111854e-06, "loss": 0.0134, "step": 20856 }, { "epoch": 2.2611665221162185, "grad_norm": 0.03125278651714325, "learning_rate": 4.943420861743798e-06, "loss": 0.0002, "step": 20857 }, { "epoch": 2.2612749349522985, "grad_norm": 1.0920376777648926, "learning_rate": 4.942695488176411e-06, "loss": 0.0102, "step": 20858 }, { "epoch": 2.261383347788378, "grad_norm": 0.3795563280582428, "learning_rate": 4.941970114609024e-06, "loss": 0.0116, "step": 20859 }, { "epoch": 2.261491760624458, "grad_norm": 0.21970897912979126, "learning_rate": 4.941244741041637e-06, "loss": 0.0047, "step": 20860 }, { "epoch": 2.2616001734605375, "grad_norm": 0.4611298143863678, "learning_rate": 4.9405193674742495e-06, "loss": 0.0183, "step": 20861 }, { "epoch": 2.2617085862966175, "grad_norm": 4.37768030166626, "learning_rate": 4.9397939939068625e-06, "loss": 0.0302, "step": 20862 }, { "epoch": 2.261816999132697, "grad_norm": 0.5772616267204285, "learning_rate": 4.939068620339475e-06, "loss": 0.0169, "step": 20863 }, { "epoch": 2.261925411968777, "grad_norm": 0.1327335387468338, "learning_rate": 4.9383432467720885e-06, "loss": 0.0005, "step": 20864 }, { "epoch": 2.262033824804857, "grad_norm": 0.37733563780784607, "learning_rate": 4.937617873204701e-06, "loss": 0.0067, "step": 20865 }, { "epoch": 2.2621422376409366, "grad_norm": 3.274681568145752, "learning_rate": 4.936892499637314e-06, "loss": 0.0375, "step": 20866 }, { "epoch": 2.2622506504770166, "grad_norm": 0.2959144115447998, "learning_rate": 4.936167126069927e-06, "loss": 0.0055, "step": 20867 }, { "epoch": 2.262359063313096, "grad_norm": 0.6242707967758179, "learning_rate": 4.93544175250254e-06, "loss": 0.0224, "step": 20868 }, { "epoch": 2.262467476149176, "grad_norm": 0.11152011156082153, "learning_rate": 4.934716378935152e-06, "loss": 0.0007, "step": 20869 }, { "epoch": 2.262575888985256, "grad_norm": 0.02176504395902157, "learning_rate": 4.933991005367765e-06, "loss": 0.0002, "step": 20870 }, { "epoch": 2.2626843018213356, "grad_norm": 0.6967645883560181, "learning_rate": 4.933265631800378e-06, "loss": 0.0016, "step": 20871 }, { "epoch": 2.2627927146574156, "grad_norm": 0.030813708901405334, "learning_rate": 4.932540258232991e-06, "loss": 0.0001, "step": 20872 }, { "epoch": 2.262901127493495, "grad_norm": 0.42319440841674805, "learning_rate": 4.931814884665604e-06, "loss": 0.0215, "step": 20873 }, { "epoch": 2.263009540329575, "grad_norm": 0.42841285467147827, "learning_rate": 4.931089511098216e-06, "loss": 0.0119, "step": 20874 }, { "epoch": 2.2631179531656547, "grad_norm": 0.21324804425239563, "learning_rate": 4.930364137530829e-06, "loss": 0.0043, "step": 20875 }, { "epoch": 2.2632263660017347, "grad_norm": 1.5161967277526855, "learning_rate": 4.929638763963441e-06, "loss": 0.0295, "step": 20876 }, { "epoch": 2.263334778837814, "grad_norm": 0.8237083554267883, "learning_rate": 4.928913390396055e-06, "loss": 0.0147, "step": 20877 }, { "epoch": 2.263443191673894, "grad_norm": 0.01288480032235384, "learning_rate": 4.928188016828667e-06, "loss": 0.0001, "step": 20878 }, { "epoch": 2.263551604509974, "grad_norm": 1.4881513118743896, "learning_rate": 4.92746264326128e-06, "loss": 0.0453, "step": 20879 }, { "epoch": 2.2636600173460537, "grad_norm": 0.0520111620426178, "learning_rate": 4.926737269693893e-06, "loss": 0.0005, "step": 20880 }, { "epoch": 2.2637684301821337, "grad_norm": 0.04111223667860031, "learning_rate": 4.926011896126506e-06, "loss": 0.0002, "step": 20881 }, { "epoch": 2.2638768430182132, "grad_norm": 6.191829681396484, "learning_rate": 4.925286522559118e-06, "loss": 0.0594, "step": 20882 }, { "epoch": 2.263985255854293, "grad_norm": 0.017591675743460655, "learning_rate": 4.924561148991731e-06, "loss": 0.0001, "step": 20883 }, { "epoch": 2.2640936686903728, "grad_norm": 2.2435550689697266, "learning_rate": 4.923835775424344e-06, "loss": 0.0319, "step": 20884 }, { "epoch": 2.2642020815264527, "grad_norm": 0.04423583671450615, "learning_rate": 4.923110401856957e-06, "loss": 0.0001, "step": 20885 }, { "epoch": 2.2643104943625323, "grad_norm": 5.141350269317627, "learning_rate": 4.92238502828957e-06, "loss": 0.0157, "step": 20886 }, { "epoch": 2.2644189071986123, "grad_norm": 0.031448669731616974, "learning_rate": 4.921659654722182e-06, "loss": 0.0002, "step": 20887 }, { "epoch": 2.2645273200346923, "grad_norm": 0.22697456181049347, "learning_rate": 4.920934281154795e-06, "loss": 0.0118, "step": 20888 }, { "epoch": 2.264635732870772, "grad_norm": 0.040576327592134476, "learning_rate": 4.920208907587407e-06, "loss": 0.0002, "step": 20889 }, { "epoch": 2.264744145706852, "grad_norm": 0.41846638917922974, "learning_rate": 4.919483534020021e-06, "loss": 0.0112, "step": 20890 }, { "epoch": 2.2648525585429313, "grad_norm": 11.14435863494873, "learning_rate": 4.918758160452633e-06, "loss": 0.04, "step": 20891 }, { "epoch": 2.2649609713790113, "grad_norm": 0.29823535680770874, "learning_rate": 4.918032786885246e-06, "loss": 0.0066, "step": 20892 }, { "epoch": 2.2650693842150913, "grad_norm": 0.2881760001182556, "learning_rate": 4.917307413317859e-06, "loss": 0.0189, "step": 20893 }, { "epoch": 2.265177797051171, "grad_norm": 0.32860904932022095, "learning_rate": 4.916582039750472e-06, "loss": 0.0059, "step": 20894 }, { "epoch": 2.265286209887251, "grad_norm": 0.40933671593666077, "learning_rate": 4.9158566661830844e-06, "loss": 0.0089, "step": 20895 }, { "epoch": 2.2653946227233304, "grad_norm": 6.495588779449463, "learning_rate": 4.915131292615697e-06, "loss": 0.0226, "step": 20896 }, { "epoch": 2.2655030355594103, "grad_norm": 1.8562272787094116, "learning_rate": 4.91440591904831e-06, "loss": 0.0207, "step": 20897 }, { "epoch": 2.26561144839549, "grad_norm": 0.005780939478427172, "learning_rate": 4.913680545480923e-06, "loss": 0.0001, "step": 20898 }, { "epoch": 2.26571986123157, "grad_norm": 0.26829099655151367, "learning_rate": 4.912955171913536e-06, "loss": 0.0009, "step": 20899 }, { "epoch": 2.2658282740676494, "grad_norm": 0.05792427435517311, "learning_rate": 4.9122297983461485e-06, "loss": 0.0004, "step": 20900 }, { "epoch": 2.2659366869037294, "grad_norm": 1.0846638679504395, "learning_rate": 4.9115044247787615e-06, "loss": 0.0042, "step": 20901 }, { "epoch": 2.2660450997398094, "grad_norm": 0.7326708436012268, "learning_rate": 4.910779051211374e-06, "loss": 0.0011, "step": 20902 }, { "epoch": 2.266153512575889, "grad_norm": 0.5005408525466919, "learning_rate": 4.9100536776439875e-06, "loss": 0.0185, "step": 20903 }, { "epoch": 2.266261925411969, "grad_norm": 0.3672700524330139, "learning_rate": 4.9093283040766e-06, "loss": 0.0093, "step": 20904 }, { "epoch": 2.2663703382480485, "grad_norm": 0.07568134367465973, "learning_rate": 4.908602930509213e-06, "loss": 0.0006, "step": 20905 }, { "epoch": 2.2664787510841284, "grad_norm": 0.1030244305729866, "learning_rate": 4.907877556941826e-06, "loss": 0.0005, "step": 20906 }, { "epoch": 2.266587163920208, "grad_norm": 0.6086670756340027, "learning_rate": 4.907152183374439e-06, "loss": 0.016, "step": 20907 }, { "epoch": 2.266695576756288, "grad_norm": 0.14633190631866455, "learning_rate": 4.906426809807051e-06, "loss": 0.0006, "step": 20908 }, { "epoch": 2.2668039895923675, "grad_norm": 0.36414751410484314, "learning_rate": 4.905701436239664e-06, "loss": 0.0111, "step": 20909 }, { "epoch": 2.2669124024284475, "grad_norm": 0.7969059944152832, "learning_rate": 4.904976062672277e-06, "loss": 0.0025, "step": 20910 }, { "epoch": 2.2670208152645275, "grad_norm": 0.4696660339832306, "learning_rate": 4.90425068910489e-06, "loss": 0.0235, "step": 20911 }, { "epoch": 2.267129228100607, "grad_norm": 0.3280228078365326, "learning_rate": 4.903525315537503e-06, "loss": 0.0131, "step": 20912 }, { "epoch": 2.267237640936687, "grad_norm": 1.3819899559020996, "learning_rate": 4.902799941970115e-06, "loss": 0.0194, "step": 20913 }, { "epoch": 2.2673460537727665, "grad_norm": 0.7956563830375671, "learning_rate": 4.902074568402728e-06, "loss": 0.0225, "step": 20914 }, { "epoch": 2.2674544666088465, "grad_norm": 0.18048200011253357, "learning_rate": 4.90134919483534e-06, "loss": 0.0008, "step": 20915 }, { "epoch": 2.2675628794449265, "grad_norm": 2.4566304683685303, "learning_rate": 4.900623821267954e-06, "loss": 0.0041, "step": 20916 }, { "epoch": 2.267671292281006, "grad_norm": 8.471168518066406, "learning_rate": 4.899898447700566e-06, "loss": 0.0136, "step": 20917 }, { "epoch": 2.267779705117086, "grad_norm": 0.46926283836364746, "learning_rate": 4.899173074133179e-06, "loss": 0.0172, "step": 20918 }, { "epoch": 2.2678881179531656, "grad_norm": 1.3078707456588745, "learning_rate": 4.898447700565792e-06, "loss": 0.0037, "step": 20919 }, { "epoch": 2.2679965307892456, "grad_norm": 1.275559663772583, "learning_rate": 4.897722326998405e-06, "loss": 0.0045, "step": 20920 }, { "epoch": 2.268104943625325, "grad_norm": 2.4170053005218506, "learning_rate": 4.896996953431017e-06, "loss": 0.0158, "step": 20921 }, { "epoch": 2.268213356461405, "grad_norm": 2.907521963119507, "learning_rate": 4.89627157986363e-06, "loss": 0.0165, "step": 20922 }, { "epoch": 2.2683217692974846, "grad_norm": 0.29257795214653015, "learning_rate": 4.895546206296243e-06, "loss": 0.0007, "step": 20923 }, { "epoch": 2.2684301821335646, "grad_norm": 0.17016154527664185, "learning_rate": 4.894820832728856e-06, "loss": 0.0004, "step": 20924 }, { "epoch": 2.2685385949696446, "grad_norm": 0.8376771211624146, "learning_rate": 4.894095459161469e-06, "loss": 0.0022, "step": 20925 }, { "epoch": 2.268647007805724, "grad_norm": 0.013369744643568993, "learning_rate": 4.893370085594081e-06, "loss": 0.0001, "step": 20926 }, { "epoch": 2.268755420641804, "grad_norm": 0.5046089887619019, "learning_rate": 4.892644712026694e-06, "loss": 0.014, "step": 20927 }, { "epoch": 2.2688638334778837, "grad_norm": 3.2693750858306885, "learning_rate": 4.891919338459306e-06, "loss": 0.0048, "step": 20928 }, { "epoch": 2.2689722463139637, "grad_norm": 0.029038293287158012, "learning_rate": 4.89119396489192e-06, "loss": 0.0002, "step": 20929 }, { "epoch": 2.269080659150043, "grad_norm": 0.7862911820411682, "learning_rate": 4.890468591324532e-06, "loss": 0.0245, "step": 20930 }, { "epoch": 2.269189071986123, "grad_norm": 2.382903575897217, "learning_rate": 4.889743217757145e-06, "loss": 0.0151, "step": 20931 }, { "epoch": 2.2692974848222027, "grad_norm": 2.3228070735931396, "learning_rate": 4.889017844189758e-06, "loss": 0.0317, "step": 20932 }, { "epoch": 2.2694058976582827, "grad_norm": 0.28217336535453796, "learning_rate": 4.888292470622371e-06, "loss": 0.006, "step": 20933 }, { "epoch": 2.2695143104943627, "grad_norm": 0.03732747584581375, "learning_rate": 4.8875670970549834e-06, "loss": 0.0001, "step": 20934 }, { "epoch": 2.2696227233304422, "grad_norm": 0.2515510022640228, "learning_rate": 4.886841723487596e-06, "loss": 0.0169, "step": 20935 }, { "epoch": 2.2697311361665222, "grad_norm": 0.1522296667098999, "learning_rate": 4.886116349920209e-06, "loss": 0.0004, "step": 20936 }, { "epoch": 2.2698395490026018, "grad_norm": 0.047748345881700516, "learning_rate": 4.885390976352822e-06, "loss": 0.0002, "step": 20937 }, { "epoch": 2.2699479618386817, "grad_norm": 0.30409353971481323, "learning_rate": 4.884665602785435e-06, "loss": 0.0066, "step": 20938 }, { "epoch": 2.2700563746747617, "grad_norm": 8.081624031066895, "learning_rate": 4.8839402292180475e-06, "loss": 0.0459, "step": 20939 }, { "epoch": 2.2701647875108413, "grad_norm": 0.504074215888977, "learning_rate": 4.8832148556506605e-06, "loss": 0.0211, "step": 20940 }, { "epoch": 2.2702732003469213, "grad_norm": 1.200881838798523, "learning_rate": 4.8824894820832735e-06, "loss": 0.0131, "step": 20941 }, { "epoch": 2.270381613183001, "grad_norm": 1.9177602529525757, "learning_rate": 4.8817641085158865e-06, "loss": 0.0516, "step": 20942 }, { "epoch": 2.270490026019081, "grad_norm": 0.45654296875, "learning_rate": 4.881038734948499e-06, "loss": 0.0093, "step": 20943 }, { "epoch": 2.2705984388551603, "grad_norm": 2.1832616329193115, "learning_rate": 4.880313361381112e-06, "loss": 0.0078, "step": 20944 }, { "epoch": 2.2707068516912403, "grad_norm": 3.3252031803131104, "learning_rate": 4.879587987813725e-06, "loss": 0.0532, "step": 20945 }, { "epoch": 2.27081526452732, "grad_norm": 0.10997062176465988, "learning_rate": 4.878862614246338e-06, "loss": 0.0008, "step": 20946 }, { "epoch": 2.2709236773634, "grad_norm": 5.109654903411865, "learning_rate": 4.87813724067895e-06, "loss": 0.0448, "step": 20947 }, { "epoch": 2.27103209019948, "grad_norm": 3.6776273250579834, "learning_rate": 4.877411867111563e-06, "loss": 0.0354, "step": 20948 }, { "epoch": 2.2711405030355594, "grad_norm": 4.485005855560303, "learning_rate": 4.876686493544176e-06, "loss": 0.0374, "step": 20949 }, { "epoch": 2.2712489158716394, "grad_norm": 4.019374847412109, "learning_rate": 4.875961119976789e-06, "loss": 0.0404, "step": 20950 }, { "epoch": 2.271357328707719, "grad_norm": 5.046272277832031, "learning_rate": 4.875235746409402e-06, "loss": 0.0091, "step": 20951 }, { "epoch": 2.271465741543799, "grad_norm": 0.5415126085281372, "learning_rate": 4.874510372842014e-06, "loss": 0.0119, "step": 20952 }, { "epoch": 2.2715741543798784, "grad_norm": 7.064064979553223, "learning_rate": 4.873784999274627e-06, "loss": 0.0195, "step": 20953 }, { "epoch": 2.2716825672159584, "grad_norm": 0.6043034195899963, "learning_rate": 4.87305962570724e-06, "loss": 0.021, "step": 20954 }, { "epoch": 2.271790980052038, "grad_norm": 3.706646680831909, "learning_rate": 4.872334252139853e-06, "loss": 0.0505, "step": 20955 }, { "epoch": 2.271899392888118, "grad_norm": 3.014105796813965, "learning_rate": 4.871608878572465e-06, "loss": 0.0164, "step": 20956 }, { "epoch": 2.272007805724198, "grad_norm": 0.7146115899085999, "learning_rate": 4.870883505005078e-06, "loss": 0.0029, "step": 20957 }, { "epoch": 2.2721162185602775, "grad_norm": 0.06413476914167404, "learning_rate": 4.870158131437691e-06, "loss": 0.0005, "step": 20958 }, { "epoch": 2.2722246313963574, "grad_norm": 9.31694507598877, "learning_rate": 4.869432757870304e-06, "loss": 0.0139, "step": 20959 }, { "epoch": 2.272333044232437, "grad_norm": 0.8487656712532043, "learning_rate": 4.868707384302916e-06, "loss": 0.0015, "step": 20960 }, { "epoch": 2.272441457068517, "grad_norm": 10.318671226501465, "learning_rate": 4.867982010735529e-06, "loss": 0.0089, "step": 20961 }, { "epoch": 2.272549869904597, "grad_norm": 0.0243097934871912, "learning_rate": 4.867256637168142e-06, "loss": 0.0002, "step": 20962 }, { "epoch": 2.2726582827406765, "grad_norm": 0.16151489317417145, "learning_rate": 4.866531263600755e-06, "loss": 0.0006, "step": 20963 }, { "epoch": 2.2727666955767565, "grad_norm": 0.18291139602661133, "learning_rate": 4.865805890033368e-06, "loss": 0.0005, "step": 20964 }, { "epoch": 2.272875108412836, "grad_norm": 11.241981506347656, "learning_rate": 4.86508051646598e-06, "loss": 0.0288, "step": 20965 }, { "epoch": 2.272983521248916, "grad_norm": 0.5925241112709045, "learning_rate": 4.864355142898593e-06, "loss": 0.0165, "step": 20966 }, { "epoch": 2.2730919340849955, "grad_norm": 0.3163120448589325, "learning_rate": 4.863629769331206e-06, "loss": 0.0111, "step": 20967 }, { "epoch": 2.2732003469210755, "grad_norm": 5.640134811401367, "learning_rate": 4.862904395763819e-06, "loss": 0.0098, "step": 20968 }, { "epoch": 2.273308759757155, "grad_norm": 0.3755457103252411, "learning_rate": 4.862179022196431e-06, "loss": 0.0102, "step": 20969 }, { "epoch": 2.273417172593235, "grad_norm": 0.22624769806861877, "learning_rate": 4.861453648629044e-06, "loss": 0.0011, "step": 20970 }, { "epoch": 2.273525585429315, "grad_norm": 0.732428252696991, "learning_rate": 4.860728275061657e-06, "loss": 0.026, "step": 20971 }, { "epoch": 2.2736339982653946, "grad_norm": 0.3575694262981415, "learning_rate": 4.86000290149427e-06, "loss": 0.0128, "step": 20972 }, { "epoch": 2.2737424111014746, "grad_norm": 0.6234016418457031, "learning_rate": 4.8592775279268824e-06, "loss": 0.0119, "step": 20973 }, { "epoch": 2.273850823937554, "grad_norm": 0.4312925636768341, "learning_rate": 4.8585521543594954e-06, "loss": 0.0115, "step": 20974 }, { "epoch": 2.273959236773634, "grad_norm": 8.065876960754395, "learning_rate": 4.857826780792108e-06, "loss": 0.0106, "step": 20975 }, { "epoch": 2.2740676496097136, "grad_norm": 0.5503265857696533, "learning_rate": 4.857101407224721e-06, "loss": 0.0197, "step": 20976 }, { "epoch": 2.2741760624457936, "grad_norm": 7.18247127532959, "learning_rate": 4.856376033657334e-06, "loss": 0.0391, "step": 20977 }, { "epoch": 2.274284475281873, "grad_norm": 0.5464597940444946, "learning_rate": 4.8556506600899465e-06, "loss": 0.0178, "step": 20978 }, { "epoch": 2.274392888117953, "grad_norm": 1.065399408340454, "learning_rate": 4.8549252865225595e-06, "loss": 0.0368, "step": 20979 }, { "epoch": 2.274501300954033, "grad_norm": 2.6443605422973633, "learning_rate": 4.8541999129551725e-06, "loss": 0.0164, "step": 20980 }, { "epoch": 2.2746097137901127, "grad_norm": 5.962192535400391, "learning_rate": 4.8534745393877855e-06, "loss": 0.0444, "step": 20981 }, { "epoch": 2.2747181266261927, "grad_norm": 0.33756759762763977, "learning_rate": 4.852749165820398e-06, "loss": 0.012, "step": 20982 }, { "epoch": 2.274826539462272, "grad_norm": 2.7613117694854736, "learning_rate": 4.852023792253011e-06, "loss": 0.0077, "step": 20983 }, { "epoch": 2.274934952298352, "grad_norm": 0.3914181590080261, "learning_rate": 4.851298418685624e-06, "loss": 0.0123, "step": 20984 }, { "epoch": 2.2750433651344317, "grad_norm": 0.23793286085128784, "learning_rate": 4.850573045118237e-06, "loss": 0.0051, "step": 20985 }, { "epoch": 2.2751517779705117, "grad_norm": 0.2578059136867523, "learning_rate": 4.849847671550849e-06, "loss": 0.0007, "step": 20986 }, { "epoch": 2.2752601908065913, "grad_norm": 4.148782730102539, "learning_rate": 4.849122297983462e-06, "loss": 0.0024, "step": 20987 }, { "epoch": 2.2753686036426712, "grad_norm": 0.8951556086540222, "learning_rate": 4.848396924416075e-06, "loss": 0.0015, "step": 20988 }, { "epoch": 2.2754770164787512, "grad_norm": 0.08352427184581757, "learning_rate": 4.847671550848688e-06, "loss": 0.0008, "step": 20989 }, { "epoch": 2.2755854293148308, "grad_norm": 0.300092488527298, "learning_rate": 4.846946177281301e-06, "loss": 0.0104, "step": 20990 }, { "epoch": 2.2756938421509108, "grad_norm": 0.4284166395664215, "learning_rate": 4.846220803713913e-06, "loss": 0.0104, "step": 20991 }, { "epoch": 2.2758022549869903, "grad_norm": 0.731846272945404, "learning_rate": 4.845495430146526e-06, "loss": 0.0019, "step": 20992 }, { "epoch": 2.2759106678230703, "grad_norm": 5.911718368530273, "learning_rate": 4.844770056579139e-06, "loss": 0.0263, "step": 20993 }, { "epoch": 2.2760190806591503, "grad_norm": 0.05210364982485771, "learning_rate": 4.844044683011752e-06, "loss": 0.0003, "step": 20994 }, { "epoch": 2.27612749349523, "grad_norm": 0.06234565004706383, "learning_rate": 4.843319309444364e-06, "loss": 0.0004, "step": 20995 }, { "epoch": 2.27623590633131, "grad_norm": 0.6320251226425171, "learning_rate": 4.842593935876977e-06, "loss": 0.0206, "step": 20996 }, { "epoch": 2.2763443191673893, "grad_norm": 0.27216213941574097, "learning_rate": 4.84186856230959e-06, "loss": 0.0016, "step": 20997 }, { "epoch": 2.2764527320034693, "grad_norm": 0.9737802147865295, "learning_rate": 4.841143188742203e-06, "loss": 0.0105, "step": 20998 }, { "epoch": 2.276561144839549, "grad_norm": 0.4877966344356537, "learning_rate": 4.840417815174815e-06, "loss": 0.0065, "step": 20999 }, { "epoch": 2.276669557675629, "grad_norm": 0.3069250285625458, "learning_rate": 4.839692441607428e-06, "loss": 0.0064, "step": 21000 }, { "epoch": 2.2767779705117084, "grad_norm": 0.44849714636802673, "learning_rate": 4.838967068040041e-06, "loss": 0.0018, "step": 21001 }, { "epoch": 2.2768863833477884, "grad_norm": 4.413142681121826, "learning_rate": 4.838241694472654e-06, "loss": 0.0015, "step": 21002 }, { "epoch": 2.2769947961838684, "grad_norm": 0.4328429400920868, "learning_rate": 4.837516320905267e-06, "loss": 0.0164, "step": 21003 }, { "epoch": 2.277103209019948, "grad_norm": 0.301123708486557, "learning_rate": 4.836790947337879e-06, "loss": 0.011, "step": 21004 }, { "epoch": 2.277211621856028, "grad_norm": 0.3471684753894806, "learning_rate": 4.836065573770492e-06, "loss": 0.0009, "step": 21005 }, { "epoch": 2.2773200346921074, "grad_norm": 0.4687420725822449, "learning_rate": 4.835340200203105e-06, "loss": 0.0019, "step": 21006 }, { "epoch": 2.2774284475281874, "grad_norm": 0.43411892652511597, "learning_rate": 4.834614826635718e-06, "loss": 0.0084, "step": 21007 }, { "epoch": 2.277536860364267, "grad_norm": 3.728550672531128, "learning_rate": 4.83388945306833e-06, "loss": 0.0375, "step": 21008 }, { "epoch": 2.277645273200347, "grad_norm": 0.13857094943523407, "learning_rate": 4.833164079500943e-06, "loss": 0.0007, "step": 21009 }, { "epoch": 2.2777536860364265, "grad_norm": 0.3329733908176422, "learning_rate": 4.832438705933556e-06, "loss": 0.0124, "step": 21010 }, { "epoch": 2.2778620988725065, "grad_norm": 0.6261627078056335, "learning_rate": 4.831713332366169e-06, "loss": 0.0075, "step": 21011 }, { "epoch": 2.2779705117085864, "grad_norm": 0.29477912187576294, "learning_rate": 4.8309879587987814e-06, "loss": 0.0063, "step": 21012 }, { "epoch": 2.278078924544666, "grad_norm": 6.742488384246826, "learning_rate": 4.8302625852313944e-06, "loss": 0.0184, "step": 21013 }, { "epoch": 2.278187337380746, "grad_norm": 2.947492837905884, "learning_rate": 4.829537211664007e-06, "loss": 0.0126, "step": 21014 }, { "epoch": 2.2782957502168255, "grad_norm": 0.2444964051246643, "learning_rate": 4.82881183809662e-06, "loss": 0.005, "step": 21015 }, { "epoch": 2.2784041630529055, "grad_norm": 2.4247899055480957, "learning_rate": 4.828086464529233e-06, "loss": 0.0688, "step": 21016 }, { "epoch": 2.2785125758889855, "grad_norm": 0.5782901644706726, "learning_rate": 4.8273610909618455e-06, "loss": 0.0026, "step": 21017 }, { "epoch": 2.278620988725065, "grad_norm": 0.6666333675384521, "learning_rate": 4.8266357173944585e-06, "loss": 0.0013, "step": 21018 }, { "epoch": 2.278729401561145, "grad_norm": 0.2636747658252716, "learning_rate": 4.8259103438270715e-06, "loss": 0.0054, "step": 21019 }, { "epoch": 2.2788378143972245, "grad_norm": 2.851353406906128, "learning_rate": 4.8251849702596845e-06, "loss": 0.0063, "step": 21020 }, { "epoch": 2.2789462272333045, "grad_norm": 0.06002472713589668, "learning_rate": 4.824459596692297e-06, "loss": 0.0003, "step": 21021 }, { "epoch": 2.279054640069384, "grad_norm": 0.5368044972419739, "learning_rate": 4.82373422312491e-06, "loss": 0.0177, "step": 21022 }, { "epoch": 2.279163052905464, "grad_norm": 0.2568035423755646, "learning_rate": 4.823008849557523e-06, "loss": 0.0057, "step": 21023 }, { "epoch": 2.2792714657415436, "grad_norm": 0.43788763880729675, "learning_rate": 4.822283475990136e-06, "loss": 0.0102, "step": 21024 }, { "epoch": 2.2793798785776236, "grad_norm": 10.594358444213867, "learning_rate": 4.821558102422748e-06, "loss": 0.0169, "step": 21025 }, { "epoch": 2.2794882914137036, "grad_norm": 0.4502723217010498, "learning_rate": 4.820832728855361e-06, "loss": 0.0122, "step": 21026 }, { "epoch": 2.279596704249783, "grad_norm": 6.1999592781066895, "learning_rate": 4.820107355287974e-06, "loss": 0.0089, "step": 21027 }, { "epoch": 2.279705117085863, "grad_norm": 1.6599708795547485, "learning_rate": 4.819381981720587e-06, "loss": 0.0388, "step": 21028 }, { "epoch": 2.2798135299219426, "grad_norm": 5.326535224914551, "learning_rate": 4.8186566081532e-06, "loss": 0.0266, "step": 21029 }, { "epoch": 2.2799219427580226, "grad_norm": 0.25656604766845703, "learning_rate": 4.817931234585812e-06, "loss": 0.0052, "step": 21030 }, { "epoch": 2.280030355594102, "grad_norm": 0.22001071274280548, "learning_rate": 4.817205861018425e-06, "loss": 0.0168, "step": 21031 }, { "epoch": 2.280138768430182, "grad_norm": 0.02408275008201599, "learning_rate": 4.816480487451038e-06, "loss": 0.0002, "step": 21032 }, { "epoch": 2.2802471812662617, "grad_norm": 2.5082404613494873, "learning_rate": 4.815755113883651e-06, "loss": 0.0303, "step": 21033 }, { "epoch": 2.2803555941023417, "grad_norm": 0.47328948974609375, "learning_rate": 4.815029740316263e-06, "loss": 0.027, "step": 21034 }, { "epoch": 2.2804640069384217, "grad_norm": 23.871614456176758, "learning_rate": 4.814304366748876e-06, "loss": 0.006, "step": 21035 }, { "epoch": 2.280572419774501, "grad_norm": 0.09007410705089569, "learning_rate": 4.813578993181489e-06, "loss": 0.0005, "step": 21036 }, { "epoch": 2.280680832610581, "grad_norm": 3.268141269683838, "learning_rate": 4.812853619614102e-06, "loss": 0.0172, "step": 21037 }, { "epoch": 2.2807892454466607, "grad_norm": 0.2566940486431122, "learning_rate": 4.812128246046714e-06, "loss": 0.0057, "step": 21038 }, { "epoch": 2.2808976582827407, "grad_norm": 0.005499746650457382, "learning_rate": 4.811402872479327e-06, "loss": 0.0, "step": 21039 }, { "epoch": 2.2810060711188207, "grad_norm": 1.348004698753357, "learning_rate": 4.81067749891194e-06, "loss": 0.0021, "step": 21040 }, { "epoch": 2.2811144839549002, "grad_norm": 2.5892269611358643, "learning_rate": 4.809952125344553e-06, "loss": 0.0556, "step": 21041 }, { "epoch": 2.2812228967909802, "grad_norm": 0.02793111652135849, "learning_rate": 4.809226751777166e-06, "loss": 0.0003, "step": 21042 }, { "epoch": 2.2813313096270598, "grad_norm": 3.5975825786590576, "learning_rate": 4.808501378209778e-06, "loss": 0.0061, "step": 21043 }, { "epoch": 2.2814397224631398, "grad_norm": 0.10427161306142807, "learning_rate": 4.807776004642391e-06, "loss": 0.0007, "step": 21044 }, { "epoch": 2.2815481352992193, "grad_norm": 0.3289969563484192, "learning_rate": 4.807050631075004e-06, "loss": 0.0045, "step": 21045 }, { "epoch": 2.2816565481352993, "grad_norm": 0.4719569683074951, "learning_rate": 4.806325257507617e-06, "loss": 0.0128, "step": 21046 }, { "epoch": 2.281764960971379, "grad_norm": 1.7521188259124756, "learning_rate": 4.805599883940229e-06, "loss": 0.0517, "step": 21047 }, { "epoch": 2.281873373807459, "grad_norm": 0.012339563108980656, "learning_rate": 4.804874510372842e-06, "loss": 0.0001, "step": 21048 }, { "epoch": 2.281981786643539, "grad_norm": 0.45561689138412476, "learning_rate": 4.804149136805455e-06, "loss": 0.0188, "step": 21049 }, { "epoch": 2.2820901994796183, "grad_norm": 0.26587504148483276, "learning_rate": 4.803423763238068e-06, "loss": 0.0055, "step": 21050 }, { "epoch": 2.2821986123156983, "grad_norm": 15.329549789428711, "learning_rate": 4.8026983896706804e-06, "loss": 0.0121, "step": 21051 }, { "epoch": 2.282307025151778, "grad_norm": 0.05963995307683945, "learning_rate": 4.8019730161032934e-06, "loss": 0.0004, "step": 21052 }, { "epoch": 2.282415437987858, "grad_norm": 1.508682131767273, "learning_rate": 4.801247642535906e-06, "loss": 0.0081, "step": 21053 }, { "epoch": 2.2825238508239374, "grad_norm": 3.0004820823669434, "learning_rate": 4.800522268968519e-06, "loss": 0.0068, "step": 21054 }, { "epoch": 2.2826322636600174, "grad_norm": 0.13214997947216034, "learning_rate": 4.799796895401132e-06, "loss": 0.0009, "step": 21055 }, { "epoch": 2.282740676496097, "grad_norm": 0.35170072317123413, "learning_rate": 4.7990715218337445e-06, "loss": 0.01, "step": 21056 }, { "epoch": 2.282849089332177, "grad_norm": 0.012420304119586945, "learning_rate": 4.7983461482663575e-06, "loss": 0.0001, "step": 21057 }, { "epoch": 2.282957502168257, "grad_norm": 0.28987523913383484, "learning_rate": 4.7976207746989705e-06, "loss": 0.0009, "step": 21058 }, { "epoch": 2.2830659150043364, "grad_norm": 7.899535655975342, "learning_rate": 4.7968954011315835e-06, "loss": 0.0098, "step": 21059 }, { "epoch": 2.2831743278404164, "grad_norm": 6.601770401000977, "learning_rate": 4.796170027564196e-06, "loss": 0.0187, "step": 21060 }, { "epoch": 2.283282740676496, "grad_norm": 0.2610720694065094, "learning_rate": 4.795444653996809e-06, "loss": 0.0057, "step": 21061 }, { "epoch": 2.283391153512576, "grad_norm": 5.076845645904541, "learning_rate": 4.794719280429422e-06, "loss": 0.0205, "step": 21062 }, { "epoch": 2.283499566348656, "grad_norm": 0.34864741563796997, "learning_rate": 4.793993906862035e-06, "loss": 0.0112, "step": 21063 }, { "epoch": 2.2836079791847355, "grad_norm": 2.154634952545166, "learning_rate": 4.793268533294647e-06, "loss": 0.0432, "step": 21064 }, { "epoch": 2.2837163920208154, "grad_norm": 0.3293570578098297, "learning_rate": 4.79254315972726e-06, "loss": 0.0084, "step": 21065 }, { "epoch": 2.283824804856895, "grad_norm": 0.7190919518470764, "learning_rate": 4.791817786159873e-06, "loss": 0.0021, "step": 21066 }, { "epoch": 2.283933217692975, "grad_norm": 0.20461976528167725, "learning_rate": 4.791092412592486e-06, "loss": 0.0006, "step": 21067 }, { "epoch": 2.2840416305290545, "grad_norm": 0.10433042794466019, "learning_rate": 4.790367039025099e-06, "loss": 0.0004, "step": 21068 }, { "epoch": 2.2841500433651345, "grad_norm": 0.25195470452308655, "learning_rate": 4.789641665457711e-06, "loss": 0.0058, "step": 21069 }, { "epoch": 2.284258456201214, "grad_norm": 0.4403536319732666, "learning_rate": 4.788916291890324e-06, "loss": 0.0163, "step": 21070 }, { "epoch": 2.284366869037294, "grad_norm": 0.24804091453552246, "learning_rate": 4.788190918322937e-06, "loss": 0.0117, "step": 21071 }, { "epoch": 2.284475281873374, "grad_norm": 0.30120155215263367, "learning_rate": 4.78746554475555e-06, "loss": 0.0078, "step": 21072 }, { "epoch": 2.2845836947094535, "grad_norm": 0.4425756335258484, "learning_rate": 4.786740171188162e-06, "loss": 0.0022, "step": 21073 }, { "epoch": 2.2846921075455335, "grad_norm": 0.4885333180427551, "learning_rate": 4.786014797620775e-06, "loss": 0.0128, "step": 21074 }, { "epoch": 2.284800520381613, "grad_norm": 4.8154706954956055, "learning_rate": 4.785289424053388e-06, "loss": 0.0476, "step": 21075 }, { "epoch": 2.284908933217693, "grad_norm": 0.5870641469955444, "learning_rate": 4.784564050486001e-06, "loss": 0.0048, "step": 21076 }, { "epoch": 2.2850173460537726, "grad_norm": 0.538770854473114, "learning_rate": 4.783838676918613e-06, "loss": 0.002, "step": 21077 }, { "epoch": 2.2851257588898526, "grad_norm": 1.994882583618164, "learning_rate": 4.783113303351226e-06, "loss": 0.0029, "step": 21078 }, { "epoch": 2.285234171725932, "grad_norm": 0.8069579601287842, "learning_rate": 4.782387929783839e-06, "loss": 0.0136, "step": 21079 }, { "epoch": 2.285342584562012, "grad_norm": 0.4869379997253418, "learning_rate": 4.781662556216452e-06, "loss": 0.0145, "step": 21080 }, { "epoch": 2.285450997398092, "grad_norm": 1.5273839235305786, "learning_rate": 4.780937182649065e-06, "loss": 0.002, "step": 21081 }, { "epoch": 2.2855594102341716, "grad_norm": 0.24841506779193878, "learning_rate": 4.780211809081677e-06, "loss": 0.0054, "step": 21082 }, { "epoch": 2.2856678230702516, "grad_norm": 4.296579360961914, "learning_rate": 4.77948643551429e-06, "loss": 0.0206, "step": 21083 }, { "epoch": 2.285776235906331, "grad_norm": 1.7740602493286133, "learning_rate": 4.778761061946903e-06, "loss": 0.001, "step": 21084 }, { "epoch": 2.285884648742411, "grad_norm": 1.275517225265503, "learning_rate": 4.778035688379516e-06, "loss": 0.0024, "step": 21085 }, { "epoch": 2.285993061578491, "grad_norm": 0.3563556373119354, "learning_rate": 4.777310314812128e-06, "loss": 0.0189, "step": 21086 }, { "epoch": 2.2861014744145707, "grad_norm": 0.009459039196372032, "learning_rate": 4.776584941244741e-06, "loss": 0.0001, "step": 21087 }, { "epoch": 2.2862098872506507, "grad_norm": 6.627235412597656, "learning_rate": 4.775859567677354e-06, "loss": 0.0204, "step": 21088 }, { "epoch": 2.28631830008673, "grad_norm": 0.13421329855918884, "learning_rate": 4.775134194109967e-06, "loss": 0.0005, "step": 21089 }, { "epoch": 2.28642671292281, "grad_norm": 0.028177062049508095, "learning_rate": 4.7744088205425794e-06, "loss": 0.0002, "step": 21090 }, { "epoch": 2.2865351257588897, "grad_norm": 0.2794404625892639, "learning_rate": 4.7736834469751924e-06, "loss": 0.0061, "step": 21091 }, { "epoch": 2.2866435385949697, "grad_norm": 2.279470205307007, "learning_rate": 4.7729580734078054e-06, "loss": 0.0411, "step": 21092 }, { "epoch": 2.2867519514310493, "grad_norm": 0.08627624809741974, "learning_rate": 4.772232699840418e-06, "loss": 0.0004, "step": 21093 }, { "epoch": 2.2868603642671292, "grad_norm": 3.7394802570343018, "learning_rate": 4.771507326273031e-06, "loss": 0.0072, "step": 21094 }, { "epoch": 2.2869687771032092, "grad_norm": 3.6947290897369385, "learning_rate": 4.7707819527056435e-06, "loss": 0.0292, "step": 21095 }, { "epoch": 2.2870771899392888, "grad_norm": 0.01700494810938835, "learning_rate": 4.7700565791382565e-06, "loss": 0.0001, "step": 21096 }, { "epoch": 2.2871856027753688, "grad_norm": 0.5781917572021484, "learning_rate": 4.7693312055708695e-06, "loss": 0.0008, "step": 21097 }, { "epoch": 2.2872940156114483, "grad_norm": 5.223631858825684, "learning_rate": 4.7686058320034825e-06, "loss": 0.0222, "step": 21098 }, { "epoch": 2.2874024284475283, "grad_norm": 5.553798675537109, "learning_rate": 4.767880458436095e-06, "loss": 0.0383, "step": 21099 }, { "epoch": 2.287510841283608, "grad_norm": 0.8196737170219421, "learning_rate": 4.767155084868708e-06, "loss": 0.0011, "step": 21100 }, { "epoch": 2.287619254119688, "grad_norm": 0.6903930306434631, "learning_rate": 4.766429711301321e-06, "loss": 0.0199, "step": 21101 }, { "epoch": 2.2877276669557673, "grad_norm": 0.3099585473537445, "learning_rate": 4.765704337733934e-06, "loss": 0.0115, "step": 21102 }, { "epoch": 2.2878360797918473, "grad_norm": 3.5587894916534424, "learning_rate": 4.764978964166546e-06, "loss": 0.0169, "step": 21103 }, { "epoch": 2.2879444926279273, "grad_norm": 2.507084846496582, "learning_rate": 4.764253590599159e-06, "loss": 0.0394, "step": 21104 }, { "epoch": 2.288052905464007, "grad_norm": 8.416221618652344, "learning_rate": 4.763528217031772e-06, "loss": 0.0527, "step": 21105 }, { "epoch": 2.288161318300087, "grad_norm": 0.8878189921379089, "learning_rate": 4.762802843464385e-06, "loss": 0.0147, "step": 21106 }, { "epoch": 2.2882697311361664, "grad_norm": 0.5477927923202515, "learning_rate": 4.762077469896998e-06, "loss": 0.0101, "step": 21107 }, { "epoch": 2.2883781439722464, "grad_norm": 0.018690211698412895, "learning_rate": 4.76135209632961e-06, "loss": 0.0002, "step": 21108 }, { "epoch": 2.2884865568083264, "grad_norm": 7.3747124671936035, "learning_rate": 4.760626722762223e-06, "loss": 0.0354, "step": 21109 }, { "epoch": 2.288594969644406, "grad_norm": 0.30130910873413086, "learning_rate": 4.759901349194836e-06, "loss": 0.0063, "step": 21110 }, { "epoch": 2.288703382480486, "grad_norm": 0.2742677628993988, "learning_rate": 4.759175975627449e-06, "loss": 0.006, "step": 21111 }, { "epoch": 2.2888117953165654, "grad_norm": 0.22635875642299652, "learning_rate": 4.758450602060061e-06, "loss": 0.0047, "step": 21112 }, { "epoch": 2.2889202081526454, "grad_norm": 0.0664505586028099, "learning_rate": 4.757725228492674e-06, "loss": 0.0002, "step": 21113 }, { "epoch": 2.289028620988725, "grad_norm": 0.30768445134162903, "learning_rate": 4.756999854925287e-06, "loss": 0.0051, "step": 21114 }, { "epoch": 2.289137033824805, "grad_norm": 0.47358620166778564, "learning_rate": 4.7562744813579e-06, "loss": 0.0269, "step": 21115 }, { "epoch": 2.2892454466608845, "grad_norm": 0.2011914998292923, "learning_rate": 4.755549107790512e-06, "loss": 0.0005, "step": 21116 }, { "epoch": 2.2893538594969645, "grad_norm": 0.3200116753578186, "learning_rate": 4.754823734223125e-06, "loss": 0.011, "step": 21117 }, { "epoch": 2.2894622723330444, "grad_norm": 0.006020719651132822, "learning_rate": 4.754098360655738e-06, "loss": 0.0001, "step": 21118 }, { "epoch": 2.289570685169124, "grad_norm": 4.137603282928467, "learning_rate": 4.753372987088351e-06, "loss": 0.0281, "step": 21119 }, { "epoch": 2.289679098005204, "grad_norm": 3.302597999572754, "learning_rate": 4.752647613520964e-06, "loss": 0.0087, "step": 21120 }, { "epoch": 2.2897875108412835, "grad_norm": 0.30626827478408813, "learning_rate": 4.751922239953576e-06, "loss": 0.0071, "step": 21121 }, { "epoch": 2.2898959236773635, "grad_norm": 0.046891409903764725, "learning_rate": 4.751196866386189e-06, "loss": 0.0004, "step": 21122 }, { "epoch": 2.290004336513443, "grad_norm": 2.749279022216797, "learning_rate": 4.750471492818802e-06, "loss": 0.0035, "step": 21123 }, { "epoch": 2.290112749349523, "grad_norm": 2.6151294708251953, "learning_rate": 4.749746119251415e-06, "loss": 0.0437, "step": 21124 }, { "epoch": 2.2902211621856026, "grad_norm": 7.195089817047119, "learning_rate": 4.749020745684027e-06, "loss": 0.0042, "step": 21125 }, { "epoch": 2.2903295750216826, "grad_norm": 0.4642372131347656, "learning_rate": 4.74829537211664e-06, "loss": 0.0241, "step": 21126 }, { "epoch": 2.2904379878577625, "grad_norm": 0.33753740787506104, "learning_rate": 4.747569998549253e-06, "loss": 0.0121, "step": 21127 }, { "epoch": 2.290546400693842, "grad_norm": 0.035706304013729095, "learning_rate": 4.746844624981866e-06, "loss": 0.0001, "step": 21128 }, { "epoch": 2.290654813529922, "grad_norm": 0.40705573558807373, "learning_rate": 4.7461192514144784e-06, "loss": 0.0113, "step": 21129 }, { "epoch": 2.2907632263660016, "grad_norm": 0.04941271245479584, "learning_rate": 4.7453938778470914e-06, "loss": 0.0002, "step": 21130 }, { "epoch": 2.2908716392020816, "grad_norm": 0.048663586378097534, "learning_rate": 4.7446685042797044e-06, "loss": 0.0002, "step": 21131 }, { "epoch": 2.290980052038161, "grad_norm": 0.09572973102331161, "learning_rate": 4.743943130712317e-06, "loss": 0.0116, "step": 21132 }, { "epoch": 2.291088464874241, "grad_norm": 2.0867862701416016, "learning_rate": 4.74321775714493e-06, "loss": 0.0234, "step": 21133 }, { "epoch": 2.2911968777103207, "grad_norm": 0.012066199444234371, "learning_rate": 4.7424923835775426e-06, "loss": 0.0001, "step": 21134 }, { "epoch": 2.2913052905464006, "grad_norm": 0.22413581609725952, "learning_rate": 4.7417670100101555e-06, "loss": 0.0172, "step": 21135 }, { "epoch": 2.2914137033824806, "grad_norm": 2.1349549293518066, "learning_rate": 4.7410416364427685e-06, "loss": 0.0153, "step": 21136 }, { "epoch": 2.29152211621856, "grad_norm": 0.7656572461128235, "learning_rate": 4.7403162628753815e-06, "loss": 0.0123, "step": 21137 }, { "epoch": 2.29163052905464, "grad_norm": 1.8241996765136719, "learning_rate": 4.739590889307994e-06, "loss": 0.0105, "step": 21138 }, { "epoch": 2.2917389418907197, "grad_norm": 1.05353844165802, "learning_rate": 4.738865515740607e-06, "loss": 0.0119, "step": 21139 }, { "epoch": 2.2918473547267997, "grad_norm": 0.31552883982658386, "learning_rate": 4.73814014217322e-06, "loss": 0.0067, "step": 21140 }, { "epoch": 2.2919557675628797, "grad_norm": 0.5234491229057312, "learning_rate": 4.737414768605833e-06, "loss": 0.0067, "step": 21141 }, { "epoch": 2.292064180398959, "grad_norm": 1.5023562908172607, "learning_rate": 4.736689395038445e-06, "loss": 0.0164, "step": 21142 }, { "epoch": 2.292172593235039, "grad_norm": 5.231907367706299, "learning_rate": 4.735964021471058e-06, "loss": 0.0066, "step": 21143 }, { "epoch": 2.2922810060711187, "grad_norm": 0.30327853560447693, "learning_rate": 4.735238647903671e-06, "loss": 0.0077, "step": 21144 }, { "epoch": 2.2923894189071987, "grad_norm": 4.099362850189209, "learning_rate": 4.734513274336284e-06, "loss": 0.0289, "step": 21145 }, { "epoch": 2.2924978317432783, "grad_norm": 1.7395068407058716, "learning_rate": 4.733787900768897e-06, "loss": 0.0412, "step": 21146 }, { "epoch": 2.2926062445793582, "grad_norm": 0.27946025133132935, "learning_rate": 4.733062527201509e-06, "loss": 0.0045, "step": 21147 }, { "epoch": 2.292714657415438, "grad_norm": 0.8738684058189392, "learning_rate": 4.732337153634122e-06, "loss": 0.0012, "step": 21148 }, { "epoch": 2.2928230702515178, "grad_norm": 2.08066463470459, "learning_rate": 4.731611780066735e-06, "loss": 0.0036, "step": 21149 }, { "epoch": 2.2929314830875978, "grad_norm": 1.1443532705307007, "learning_rate": 4.730886406499348e-06, "loss": 0.0123, "step": 21150 }, { "epoch": 2.2930398959236773, "grad_norm": 0.4621550142765045, "learning_rate": 4.73016103293196e-06, "loss": 0.0178, "step": 21151 }, { "epoch": 2.2931483087597573, "grad_norm": 0.2897356152534485, "learning_rate": 4.729435659364573e-06, "loss": 0.0007, "step": 21152 }, { "epoch": 2.293256721595837, "grad_norm": 1.694453477859497, "learning_rate": 4.728710285797186e-06, "loss": 0.0263, "step": 21153 }, { "epoch": 2.293365134431917, "grad_norm": 0.03799473121762276, "learning_rate": 4.727984912229799e-06, "loss": 0.0003, "step": 21154 }, { "epoch": 2.2934735472679963, "grad_norm": 0.07820922881364822, "learning_rate": 4.727259538662411e-06, "loss": 0.0002, "step": 21155 }, { "epoch": 2.2935819601040763, "grad_norm": 0.01530348602682352, "learning_rate": 4.726534165095024e-06, "loss": 0.0001, "step": 21156 }, { "epoch": 2.293690372940156, "grad_norm": 0.5845162272453308, "learning_rate": 4.725808791527637e-06, "loss": 0.0194, "step": 21157 }, { "epoch": 2.293798785776236, "grad_norm": 0.09586907923221588, "learning_rate": 4.72508341796025e-06, "loss": 0.0005, "step": 21158 }, { "epoch": 2.293907198612316, "grad_norm": 2.449018716812134, "learning_rate": 4.724358044392863e-06, "loss": 0.0162, "step": 21159 }, { "epoch": 2.2940156114483954, "grad_norm": 1.3132593631744385, "learning_rate": 4.723632670825475e-06, "loss": 0.0174, "step": 21160 }, { "epoch": 2.2941240242844754, "grad_norm": 9.717999458312988, "learning_rate": 4.722907297258088e-06, "loss": 0.0158, "step": 21161 }, { "epoch": 2.294232437120555, "grad_norm": 0.5794721245765686, "learning_rate": 4.722181923690701e-06, "loss": 0.0071, "step": 21162 }, { "epoch": 2.294340849956635, "grad_norm": 0.023725608363747597, "learning_rate": 4.721456550123314e-06, "loss": 0.0002, "step": 21163 }, { "epoch": 2.294449262792715, "grad_norm": 0.2973283529281616, "learning_rate": 4.720731176555926e-06, "loss": 0.0234, "step": 21164 }, { "epoch": 2.2945576756287944, "grad_norm": 0.28394588828086853, "learning_rate": 4.720005802988539e-06, "loss": 0.0107, "step": 21165 }, { "epoch": 2.2946660884648744, "grad_norm": 0.31153959035873413, "learning_rate": 4.719280429421152e-06, "loss": 0.0117, "step": 21166 }, { "epoch": 2.294774501300954, "grad_norm": 0.06709957867860794, "learning_rate": 4.718555055853765e-06, "loss": 0.0006, "step": 21167 }, { "epoch": 2.294882914137034, "grad_norm": 0.9099000096321106, "learning_rate": 4.7178296822863775e-06, "loss": 0.0014, "step": 21168 }, { "epoch": 2.2949913269731135, "grad_norm": 6.079733371734619, "learning_rate": 4.7171043087189904e-06, "loss": 0.0084, "step": 21169 }, { "epoch": 2.2950997398091935, "grad_norm": 0.2530122995376587, "learning_rate": 4.7163789351516034e-06, "loss": 0.0052, "step": 21170 }, { "epoch": 2.295208152645273, "grad_norm": 0.47499457001686096, "learning_rate": 4.715653561584216e-06, "loss": 0.0011, "step": 21171 }, { "epoch": 2.295316565481353, "grad_norm": 0.030653825029730797, "learning_rate": 4.714928188016829e-06, "loss": 0.0002, "step": 21172 }, { "epoch": 2.295424978317433, "grad_norm": 0.373277872800827, "learning_rate": 4.7142028144494416e-06, "loss": 0.0117, "step": 21173 }, { "epoch": 2.2955333911535125, "grad_norm": 0.2800602912902832, "learning_rate": 4.7134774408820545e-06, "loss": 0.0046, "step": 21174 }, { "epoch": 2.2956418039895925, "grad_norm": 1.3820964097976685, "learning_rate": 4.7127520673146675e-06, "loss": 0.0118, "step": 21175 }, { "epoch": 2.295750216825672, "grad_norm": 0.07295884191989899, "learning_rate": 4.7120266937472805e-06, "loss": 0.0004, "step": 21176 }, { "epoch": 2.295858629661752, "grad_norm": 0.005803894717246294, "learning_rate": 4.711301320179893e-06, "loss": 0.0, "step": 21177 }, { "epoch": 2.2959670424978316, "grad_norm": 0.01904573105275631, "learning_rate": 4.710575946612506e-06, "loss": 0.0001, "step": 21178 }, { "epoch": 2.2960754553339116, "grad_norm": 2.824758529663086, "learning_rate": 4.709850573045119e-06, "loss": 0.0137, "step": 21179 }, { "epoch": 2.296183868169991, "grad_norm": 0.3609868884086609, "learning_rate": 4.709125199477732e-06, "loss": 0.011, "step": 21180 }, { "epoch": 2.296292281006071, "grad_norm": 0.5301554799079895, "learning_rate": 4.708399825910344e-06, "loss": 0.016, "step": 21181 }, { "epoch": 2.296400693842151, "grad_norm": 0.4389302134513855, "learning_rate": 4.707674452342957e-06, "loss": 0.0014, "step": 21182 }, { "epoch": 2.2965091066782306, "grad_norm": 6.141799449920654, "learning_rate": 4.70694907877557e-06, "loss": 0.0445, "step": 21183 }, { "epoch": 2.2966175195143106, "grad_norm": 0.35500261187553406, "learning_rate": 4.706223705208183e-06, "loss": 0.009, "step": 21184 }, { "epoch": 2.29672593235039, "grad_norm": 6.049380302429199, "learning_rate": 4.705498331640796e-06, "loss": 0.0083, "step": 21185 }, { "epoch": 2.29683434518647, "grad_norm": 4.1925554275512695, "learning_rate": 4.704772958073408e-06, "loss": 0.0287, "step": 21186 }, { "epoch": 2.29694275802255, "grad_norm": 1.8266448974609375, "learning_rate": 4.704047584506021e-06, "loss": 0.0146, "step": 21187 }, { "epoch": 2.2970511708586296, "grad_norm": 12.078694343566895, "learning_rate": 4.703322210938634e-06, "loss": 0.0155, "step": 21188 }, { "epoch": 2.2971595836947096, "grad_norm": 1.9516628980636597, "learning_rate": 4.702596837371247e-06, "loss": 0.0055, "step": 21189 }, { "epoch": 2.297267996530789, "grad_norm": 0.20709043741226196, "learning_rate": 4.701871463803859e-06, "loss": 0.0158, "step": 21190 }, { "epoch": 2.297376409366869, "grad_norm": 0.4790665805339813, "learning_rate": 4.701146090236472e-06, "loss": 0.0096, "step": 21191 }, { "epoch": 2.2974848222029487, "grad_norm": 0.2780882716178894, "learning_rate": 4.700420716669085e-06, "loss": 0.006, "step": 21192 }, { "epoch": 2.2975932350390287, "grad_norm": 0.4058596193790436, "learning_rate": 4.699695343101698e-06, "loss": 0.0075, "step": 21193 }, { "epoch": 2.297701647875108, "grad_norm": 0.012781690806150436, "learning_rate": 4.69896996953431e-06, "loss": 0.0001, "step": 21194 }, { "epoch": 2.297810060711188, "grad_norm": 0.3438863754272461, "learning_rate": 4.698244595966923e-06, "loss": 0.0113, "step": 21195 }, { "epoch": 2.297918473547268, "grad_norm": 0.4499231278896332, "learning_rate": 4.697519222399536e-06, "loss": 0.001, "step": 21196 }, { "epoch": 2.2980268863833477, "grad_norm": 3.5912458896636963, "learning_rate": 4.696793848832149e-06, "loss": 0.0323, "step": 21197 }, { "epoch": 2.2981352992194277, "grad_norm": 0.028232641518115997, "learning_rate": 4.696068475264762e-06, "loss": 0.0003, "step": 21198 }, { "epoch": 2.2982437120555073, "grad_norm": 0.24322570860385895, "learning_rate": 4.695343101697374e-06, "loss": 0.0052, "step": 21199 }, { "epoch": 2.2983521248915872, "grad_norm": 2.091224193572998, "learning_rate": 4.694617728129987e-06, "loss": 0.0232, "step": 21200 }, { "epoch": 2.298460537727667, "grad_norm": 1.9787471294403076, "learning_rate": 4.6938923545626e-06, "loss": 0.0304, "step": 21201 }, { "epoch": 2.2985689505637468, "grad_norm": 0.012185762636363506, "learning_rate": 4.693166980995213e-06, "loss": 0.0001, "step": 21202 }, { "epoch": 2.2986773633998263, "grad_norm": 0.24955907464027405, "learning_rate": 4.692441607427825e-06, "loss": 0.0054, "step": 21203 }, { "epoch": 2.2987857762359063, "grad_norm": 3.1173603534698486, "learning_rate": 4.691716233860438e-06, "loss": 0.0269, "step": 21204 }, { "epoch": 2.2988941890719863, "grad_norm": 0.02198204956948757, "learning_rate": 4.690990860293051e-06, "loss": 0.0002, "step": 21205 }, { "epoch": 2.299002601908066, "grad_norm": 0.24435065686702728, "learning_rate": 4.690265486725664e-06, "loss": 0.0048, "step": 21206 }, { "epoch": 2.299111014744146, "grad_norm": 4.107675552368164, "learning_rate": 4.6895401131582765e-06, "loss": 0.0297, "step": 21207 }, { "epoch": 2.2992194275802254, "grad_norm": 0.22859062254428864, "learning_rate": 4.6888147395908894e-06, "loss": 0.0007, "step": 21208 }, { "epoch": 2.2993278404163053, "grad_norm": 1.364479660987854, "learning_rate": 4.6880893660235024e-06, "loss": 0.0088, "step": 21209 }, { "epoch": 2.2994362532523853, "grad_norm": 0.03171594440937042, "learning_rate": 4.6873639924561154e-06, "loss": 0.0003, "step": 21210 }, { "epoch": 2.299544666088465, "grad_norm": 0.12813526391983032, "learning_rate": 4.686638618888728e-06, "loss": 0.0005, "step": 21211 }, { "epoch": 2.299653078924545, "grad_norm": 0.2713565528392792, "learning_rate": 4.6859132453213406e-06, "loss": 0.006, "step": 21212 }, { "epoch": 2.2997614917606244, "grad_norm": 0.24689224362373352, "learning_rate": 4.6851878717539535e-06, "loss": 0.0051, "step": 21213 }, { "epoch": 2.2998699045967044, "grad_norm": 9.08408260345459, "learning_rate": 4.6844624981865665e-06, "loss": 0.0718, "step": 21214 }, { "epoch": 2.299978317432784, "grad_norm": 0.7166666984558105, "learning_rate": 4.6837371246191795e-06, "loss": 0.0055, "step": 21215 }, { "epoch": 2.300086730268864, "grad_norm": 0.501785159111023, "learning_rate": 4.683011751051792e-06, "loss": 0.0129, "step": 21216 }, { "epoch": 2.3001951431049434, "grad_norm": 0.14458586275577545, "learning_rate": 4.682286377484405e-06, "loss": 0.001, "step": 21217 }, { "epoch": 2.3003035559410234, "grad_norm": 0.6740038990974426, "learning_rate": 4.681561003917018e-06, "loss": 0.0083, "step": 21218 }, { "epoch": 2.3004119687771034, "grad_norm": 0.5195993781089783, "learning_rate": 4.680835630349631e-06, "loss": 0.0144, "step": 21219 }, { "epoch": 2.300520381613183, "grad_norm": 0.5102576017379761, "learning_rate": 4.680110256782243e-06, "loss": 0.0266, "step": 21220 }, { "epoch": 2.300628794449263, "grad_norm": 0.1778556853532791, "learning_rate": 4.679384883214856e-06, "loss": 0.0004, "step": 21221 }, { "epoch": 2.3007372072853425, "grad_norm": 4.936927318572998, "learning_rate": 4.678659509647469e-06, "loss": 0.0425, "step": 21222 }, { "epoch": 2.3008456201214225, "grad_norm": 0.07041122764348984, "learning_rate": 4.677934136080082e-06, "loss": 0.0006, "step": 21223 }, { "epoch": 2.300954032957502, "grad_norm": 0.01933419145643711, "learning_rate": 4.677208762512695e-06, "loss": 0.0002, "step": 21224 }, { "epoch": 2.301062445793582, "grad_norm": 0.02410690113902092, "learning_rate": 4.676483388945307e-06, "loss": 0.0002, "step": 21225 }, { "epoch": 2.3011708586296615, "grad_norm": 5.179582595825195, "learning_rate": 4.67575801537792e-06, "loss": 0.0083, "step": 21226 }, { "epoch": 2.3012792714657415, "grad_norm": 0.33945178985595703, "learning_rate": 4.675032641810533e-06, "loss": 0.0113, "step": 21227 }, { "epoch": 2.3013876843018215, "grad_norm": 0.04779110848903656, "learning_rate": 4.674307268243146e-06, "loss": 0.0002, "step": 21228 }, { "epoch": 2.301496097137901, "grad_norm": 0.3025842308998108, "learning_rate": 4.673581894675758e-06, "loss": 0.0046, "step": 21229 }, { "epoch": 2.301604509973981, "grad_norm": 0.5413687229156494, "learning_rate": 4.672856521108371e-06, "loss": 0.0213, "step": 21230 }, { "epoch": 2.3017129228100606, "grad_norm": 0.2105586677789688, "learning_rate": 4.672131147540984e-06, "loss": 0.0083, "step": 21231 }, { "epoch": 2.3018213356461406, "grad_norm": 9.962273597717285, "learning_rate": 4.671405773973597e-06, "loss": 0.0198, "step": 21232 }, { "epoch": 2.3019297484822205, "grad_norm": 1.748721718788147, "learning_rate": 4.670680400406209e-06, "loss": 0.0372, "step": 21233 }, { "epoch": 2.3020381613183, "grad_norm": 0.33158934116363525, "learning_rate": 4.669955026838822e-06, "loss": 0.011, "step": 21234 }, { "epoch": 2.30214657415438, "grad_norm": 0.02880619652569294, "learning_rate": 4.669229653271435e-06, "loss": 0.0002, "step": 21235 }, { "epoch": 2.3022549869904596, "grad_norm": 2.7008209228515625, "learning_rate": 4.668504279704048e-06, "loss": 0.0261, "step": 21236 }, { "epoch": 2.3023633998265396, "grad_norm": 0.2968038320541382, "learning_rate": 4.667778906136661e-06, "loss": 0.0066, "step": 21237 }, { "epoch": 2.302471812662619, "grad_norm": 4.920863151550293, "learning_rate": 4.667053532569273e-06, "loss": 0.0175, "step": 21238 }, { "epoch": 2.302580225498699, "grad_norm": 0.7812303304672241, "learning_rate": 4.666328159001886e-06, "loss": 0.0289, "step": 21239 }, { "epoch": 2.3026886383347787, "grad_norm": 3.7427279949188232, "learning_rate": 4.665602785434499e-06, "loss": 0.0222, "step": 21240 }, { "epoch": 2.3027970511708586, "grad_norm": 0.26418671011924744, "learning_rate": 4.664877411867112e-06, "loss": 0.0057, "step": 21241 }, { "epoch": 2.3029054640069386, "grad_norm": 4.680810451507568, "learning_rate": 4.664152038299724e-06, "loss": 0.0141, "step": 21242 }, { "epoch": 2.303013876843018, "grad_norm": 4.284382343292236, "learning_rate": 4.663426664732337e-06, "loss": 0.0123, "step": 21243 }, { "epoch": 2.303122289679098, "grad_norm": 4.725853443145752, "learning_rate": 4.66270129116495e-06, "loss": 0.0086, "step": 21244 }, { "epoch": 2.3032307025151777, "grad_norm": 4.6114726066589355, "learning_rate": 4.661975917597563e-06, "loss": 0.012, "step": 21245 }, { "epoch": 2.3033391153512577, "grad_norm": 0.32903897762298584, "learning_rate": 4.6612505440301755e-06, "loss": 0.0125, "step": 21246 }, { "epoch": 2.3034475281873372, "grad_norm": 0.4686161279678345, "learning_rate": 4.6605251704627884e-06, "loss": 0.0121, "step": 21247 }, { "epoch": 2.303555941023417, "grad_norm": 0.003635890781879425, "learning_rate": 4.6597997968954014e-06, "loss": 0.0, "step": 21248 }, { "epoch": 2.3036643538594967, "grad_norm": 0.19642440974712372, "learning_rate": 4.6590744233280144e-06, "loss": 0.0009, "step": 21249 }, { "epoch": 2.3037727666955767, "grad_norm": 0.5568901300430298, "learning_rate": 4.658349049760627e-06, "loss": 0.0262, "step": 21250 }, { "epoch": 2.3038811795316567, "grad_norm": 4.230791091918945, "learning_rate": 4.6576236761932396e-06, "loss": 0.0179, "step": 21251 }, { "epoch": 2.3039895923677363, "grad_norm": 2.0963046550750732, "learning_rate": 4.6568983026258526e-06, "loss": 0.0223, "step": 21252 }, { "epoch": 2.3040980052038162, "grad_norm": 2.048257827758789, "learning_rate": 4.6561729290584655e-06, "loss": 0.0229, "step": 21253 }, { "epoch": 2.304206418039896, "grad_norm": 0.3569498658180237, "learning_rate": 4.6554475554910785e-06, "loss": 0.0096, "step": 21254 }, { "epoch": 2.3043148308759758, "grad_norm": 0.9277382493019104, "learning_rate": 4.654722181923691e-06, "loss": 0.0144, "step": 21255 }, { "epoch": 2.3044232437120553, "grad_norm": 0.021627169102430344, "learning_rate": 4.653996808356304e-06, "loss": 0.0002, "step": 21256 }, { "epoch": 2.3045316565481353, "grad_norm": 0.029984906315803528, "learning_rate": 4.653271434788917e-06, "loss": 0.0004, "step": 21257 }, { "epoch": 2.3046400693842153, "grad_norm": 0.23149968683719635, "learning_rate": 4.65254606122153e-06, "loss": 0.0041, "step": 21258 }, { "epoch": 2.304748482220295, "grad_norm": 0.5226972699165344, "learning_rate": 4.651820687654142e-06, "loss": 0.0192, "step": 21259 }, { "epoch": 2.304856895056375, "grad_norm": 0.18504515290260315, "learning_rate": 4.651095314086755e-06, "loss": 0.0006, "step": 21260 }, { "epoch": 2.3049653078924544, "grad_norm": 0.3363723158836365, "learning_rate": 4.650369940519368e-06, "loss": 0.0076, "step": 21261 }, { "epoch": 2.3050737207285343, "grad_norm": 3.0827395915985107, "learning_rate": 4.649644566951981e-06, "loss": 0.0298, "step": 21262 }, { "epoch": 2.305182133564614, "grad_norm": 0.279331237077713, "learning_rate": 4.648919193384594e-06, "loss": 0.0064, "step": 21263 }, { "epoch": 2.305290546400694, "grad_norm": 1.7398761510849, "learning_rate": 4.648193819817206e-06, "loss": 0.0076, "step": 21264 }, { "epoch": 2.305398959236774, "grad_norm": 0.0715380311012268, "learning_rate": 4.647468446249819e-06, "loss": 0.0116, "step": 21265 }, { "epoch": 2.3055073720728534, "grad_norm": 0.3772493898868561, "learning_rate": 4.646743072682432e-06, "loss": 0.0104, "step": 21266 }, { "epoch": 2.3056157849089334, "grad_norm": 0.11639764159917831, "learning_rate": 4.646017699115045e-06, "loss": 0.0003, "step": 21267 }, { "epoch": 2.305724197745013, "grad_norm": 5.211041450500488, "learning_rate": 4.645292325547657e-06, "loss": 0.0295, "step": 21268 }, { "epoch": 2.305832610581093, "grad_norm": 6.105491638183594, "learning_rate": 4.64456695198027e-06, "loss": 0.0494, "step": 21269 }, { "epoch": 2.3059410234171724, "grad_norm": 0.8311750888824463, "learning_rate": 4.643841578412883e-06, "loss": 0.0072, "step": 21270 }, { "epoch": 2.3060494362532524, "grad_norm": 0.04731784760951996, "learning_rate": 4.643116204845496e-06, "loss": 0.0003, "step": 21271 }, { "epoch": 2.306157849089332, "grad_norm": 3.352881669998169, "learning_rate": 4.642390831278108e-06, "loss": 0.0345, "step": 21272 }, { "epoch": 2.306266261925412, "grad_norm": 0.6025393009185791, "learning_rate": 4.641665457710721e-06, "loss": 0.0068, "step": 21273 }, { "epoch": 2.306374674761492, "grad_norm": 0.11133553087711334, "learning_rate": 4.640940084143334e-06, "loss": 0.0005, "step": 21274 }, { "epoch": 2.3064830875975715, "grad_norm": 0.3107091784477234, "learning_rate": 4.640214710575947e-06, "loss": 0.0115, "step": 21275 }, { "epoch": 2.3065915004336515, "grad_norm": 2.3770699501037598, "learning_rate": 4.63948933700856e-06, "loss": 0.0228, "step": 21276 }, { "epoch": 2.306699913269731, "grad_norm": 2.1763975620269775, "learning_rate": 4.638763963441172e-06, "loss": 0.0187, "step": 21277 }, { "epoch": 2.306808326105811, "grad_norm": 0.25745466351509094, "learning_rate": 4.638038589873785e-06, "loss": 0.0056, "step": 21278 }, { "epoch": 2.3069167389418905, "grad_norm": 0.02392379753291607, "learning_rate": 4.637313216306398e-06, "loss": 0.0001, "step": 21279 }, { "epoch": 2.3070251517779705, "grad_norm": 0.3858535885810852, "learning_rate": 4.636587842739011e-06, "loss": 0.0272, "step": 21280 }, { "epoch": 2.30713356461405, "grad_norm": 5.958409309387207, "learning_rate": 4.635862469171623e-06, "loss": 0.0652, "step": 21281 }, { "epoch": 2.30724197745013, "grad_norm": 0.6662626266479492, "learning_rate": 4.635137095604236e-06, "loss": 0.0199, "step": 21282 }, { "epoch": 2.30735039028621, "grad_norm": 0.299981951713562, "learning_rate": 4.634411722036849e-06, "loss": 0.0107, "step": 21283 }, { "epoch": 2.3074588031222896, "grad_norm": 0.03448094427585602, "learning_rate": 4.633686348469462e-06, "loss": 0.0002, "step": 21284 }, { "epoch": 2.3075672159583696, "grad_norm": 6.373435020446777, "learning_rate": 4.6329609749020745e-06, "loss": 0.0469, "step": 21285 }, { "epoch": 2.307675628794449, "grad_norm": 0.41659680008888245, "learning_rate": 4.6322356013346875e-06, "loss": 0.0165, "step": 21286 }, { "epoch": 2.307784041630529, "grad_norm": 2.655951738357544, "learning_rate": 4.6315102277673004e-06, "loss": 0.0131, "step": 21287 }, { "epoch": 2.307892454466609, "grad_norm": 9.904047966003418, "learning_rate": 4.6307848541999134e-06, "loss": 0.0231, "step": 21288 }, { "epoch": 2.3080008673026886, "grad_norm": 0.31177452206611633, "learning_rate": 4.630059480632526e-06, "loss": 0.0107, "step": 21289 }, { "epoch": 2.3081092801387686, "grad_norm": 2.254582166671753, "learning_rate": 4.6293341070651386e-06, "loss": 0.0129, "step": 21290 }, { "epoch": 2.308217692974848, "grad_norm": 1.504347562789917, "learning_rate": 4.6286087334977516e-06, "loss": 0.0028, "step": 21291 }, { "epoch": 2.308326105810928, "grad_norm": 3.285816192626953, "learning_rate": 4.6278833599303645e-06, "loss": 0.0184, "step": 21292 }, { "epoch": 2.3084345186470077, "grad_norm": 0.2797955572605133, "learning_rate": 4.6271579863629775e-06, "loss": 0.0056, "step": 21293 }, { "epoch": 2.3085429314830876, "grad_norm": 1.1557042598724365, "learning_rate": 4.62643261279559e-06, "loss": 0.0019, "step": 21294 }, { "epoch": 2.308651344319167, "grad_norm": 0.016934221610426903, "learning_rate": 4.625707239228203e-06, "loss": 0.0002, "step": 21295 }, { "epoch": 2.308759757155247, "grad_norm": 4.8729681968688965, "learning_rate": 4.624981865660816e-06, "loss": 0.0042, "step": 21296 }, { "epoch": 2.308868169991327, "grad_norm": 4.468373775482178, "learning_rate": 4.624256492093429e-06, "loss": 0.0215, "step": 21297 }, { "epoch": 2.3089765828274067, "grad_norm": 0.31813451647758484, "learning_rate": 4.623531118526041e-06, "loss": 0.0123, "step": 21298 }, { "epoch": 2.3090849956634867, "grad_norm": 0.3713372051715851, "learning_rate": 4.622805744958654e-06, "loss": 0.011, "step": 21299 }, { "epoch": 2.3091934084995662, "grad_norm": 18.48579216003418, "learning_rate": 4.622080371391267e-06, "loss": 0.015, "step": 21300 }, { "epoch": 2.309301821335646, "grad_norm": 0.28925973176956177, "learning_rate": 4.62135499782388e-06, "loss": 0.0105, "step": 21301 }, { "epoch": 2.3094102341717258, "grad_norm": 0.37356120347976685, "learning_rate": 4.620629624256493e-06, "loss": 0.0132, "step": 21302 }, { "epoch": 2.3095186470078057, "grad_norm": 0.3584221303462982, "learning_rate": 4.619904250689105e-06, "loss": 0.0118, "step": 21303 }, { "epoch": 2.3096270598438853, "grad_norm": 0.2915603220462799, "learning_rate": 4.619178877121718e-06, "loss": 0.0057, "step": 21304 }, { "epoch": 2.3097354726799653, "grad_norm": 0.013122748583555222, "learning_rate": 4.618453503554331e-06, "loss": 0.0001, "step": 21305 }, { "epoch": 2.3098438855160452, "grad_norm": 0.003866532351821661, "learning_rate": 4.617728129986944e-06, "loss": 0.0, "step": 21306 }, { "epoch": 2.309952298352125, "grad_norm": 4.157104015350342, "learning_rate": 4.617002756419556e-06, "loss": 0.006, "step": 21307 }, { "epoch": 2.3100607111882048, "grad_norm": 0.1771191656589508, "learning_rate": 4.616277382852169e-06, "loss": 0.0137, "step": 21308 }, { "epoch": 2.3101691240242843, "grad_norm": 4.3465657234191895, "learning_rate": 4.615552009284782e-06, "loss": 0.014, "step": 21309 }, { "epoch": 2.3102775368603643, "grad_norm": 18.84769630432129, "learning_rate": 4.614826635717395e-06, "loss": 0.0332, "step": 21310 }, { "epoch": 2.3103859496964443, "grad_norm": 3.2613141536712646, "learning_rate": 4.614101262150007e-06, "loss": 0.009, "step": 21311 }, { "epoch": 2.310494362532524, "grad_norm": 0.2901074290275574, "learning_rate": 4.61337588858262e-06, "loss": 0.0061, "step": 21312 }, { "epoch": 2.310602775368604, "grad_norm": 21.967411041259766, "learning_rate": 4.612650515015233e-06, "loss": 0.0366, "step": 21313 }, { "epoch": 2.3107111882046834, "grad_norm": 0.589698851108551, "learning_rate": 4.611925141447846e-06, "loss": 0.0012, "step": 21314 }, { "epoch": 2.3108196010407633, "grad_norm": 0.32443422079086304, "learning_rate": 4.611199767880459e-06, "loss": 0.0118, "step": 21315 }, { "epoch": 2.310928013876843, "grad_norm": 1.2863225936889648, "learning_rate": 4.610474394313071e-06, "loss": 0.0081, "step": 21316 }, { "epoch": 2.311036426712923, "grad_norm": 2.4828684329986572, "learning_rate": 4.609749020745684e-06, "loss": 0.0126, "step": 21317 }, { "epoch": 2.3111448395490024, "grad_norm": 3.265960931777954, "learning_rate": 4.609023647178297e-06, "loss": 0.0041, "step": 21318 }, { "epoch": 2.3112532523850824, "grad_norm": 2.9032397270202637, "learning_rate": 4.60829827361091e-06, "loss": 0.0491, "step": 21319 }, { "epoch": 2.3113616652211624, "grad_norm": 0.38906046748161316, "learning_rate": 4.607572900043522e-06, "loss": 0.0112, "step": 21320 }, { "epoch": 2.311470078057242, "grad_norm": 1.7485907077789307, "learning_rate": 4.606847526476135e-06, "loss": 0.013, "step": 21321 }, { "epoch": 2.311578490893322, "grad_norm": 0.42924636602401733, "learning_rate": 4.606122152908748e-06, "loss": 0.0156, "step": 21322 }, { "epoch": 2.3116869037294014, "grad_norm": 0.14302895963191986, "learning_rate": 4.605396779341361e-06, "loss": 0.0005, "step": 21323 }, { "epoch": 2.3117953165654814, "grad_norm": 0.24693283438682556, "learning_rate": 4.6046714057739735e-06, "loss": 0.005, "step": 21324 }, { "epoch": 2.311903729401561, "grad_norm": 0.36254897713661194, "learning_rate": 4.6039460322065865e-06, "loss": 0.0109, "step": 21325 }, { "epoch": 2.312012142237641, "grad_norm": 0.3184802234172821, "learning_rate": 4.6032206586391994e-06, "loss": 0.0104, "step": 21326 }, { "epoch": 2.3121205550737205, "grad_norm": 0.02697720006108284, "learning_rate": 4.6024952850718124e-06, "loss": 0.0002, "step": 21327 }, { "epoch": 2.3122289679098005, "grad_norm": 1.4446837902069092, "learning_rate": 4.6017699115044254e-06, "loss": 0.0297, "step": 21328 }, { "epoch": 2.3123373807458805, "grad_norm": 2.338836193084717, "learning_rate": 4.6010445379370376e-06, "loss": 0.011, "step": 21329 }, { "epoch": 2.31244579358196, "grad_norm": 0.22561423480510712, "learning_rate": 4.6003191643696506e-06, "loss": 0.0048, "step": 21330 }, { "epoch": 2.31255420641804, "grad_norm": 1.5503548383712769, "learning_rate": 4.5995937908022635e-06, "loss": 0.0102, "step": 21331 }, { "epoch": 2.3126626192541195, "grad_norm": 0.27253255248069763, "learning_rate": 4.5988684172348765e-06, "loss": 0.0125, "step": 21332 }, { "epoch": 2.3127710320901995, "grad_norm": 0.01715712808072567, "learning_rate": 4.598143043667489e-06, "loss": 0.0001, "step": 21333 }, { "epoch": 2.3128794449262795, "grad_norm": 0.2667329013347626, "learning_rate": 4.5974176701001025e-06, "loss": 0.0056, "step": 21334 }, { "epoch": 2.312987857762359, "grad_norm": 0.23425275087356567, "learning_rate": 4.596692296532715e-06, "loss": 0.0012, "step": 21335 }, { "epoch": 2.313096270598439, "grad_norm": 0.522738516330719, "learning_rate": 4.595966922965328e-06, "loss": 0.0204, "step": 21336 }, { "epoch": 2.3132046834345186, "grad_norm": 0.04059123247861862, "learning_rate": 4.59524154939794e-06, "loss": 0.0002, "step": 21337 }, { "epoch": 2.3133130962705986, "grad_norm": 2.004977226257324, "learning_rate": 4.594516175830553e-06, "loss": 0.0366, "step": 21338 }, { "epoch": 2.313421509106678, "grad_norm": 0.2937481701374054, "learning_rate": 4.593790802263166e-06, "loss": 0.0066, "step": 21339 }, { "epoch": 2.313529921942758, "grad_norm": 0.003781524719670415, "learning_rate": 4.593065428695779e-06, "loss": 0.0, "step": 21340 }, { "epoch": 2.3136383347788376, "grad_norm": 0.0477183572947979, "learning_rate": 4.592340055128392e-06, "loss": 0.0002, "step": 21341 }, { "epoch": 2.3137467476149176, "grad_norm": 0.036048196256160736, "learning_rate": 4.591614681561004e-06, "loss": 0.0002, "step": 21342 }, { "epoch": 2.3138551604509976, "grad_norm": 0.5618035197257996, "learning_rate": 4.590889307993618e-06, "loss": 0.0095, "step": 21343 }, { "epoch": 2.313963573287077, "grad_norm": 3.9398789405822754, "learning_rate": 4.59016393442623e-06, "loss": 0.0179, "step": 21344 }, { "epoch": 2.314071986123157, "grad_norm": 0.27188482880592346, "learning_rate": 4.589438560858843e-06, "loss": 0.0061, "step": 21345 }, { "epoch": 2.3141803989592367, "grad_norm": 1.1739115715026855, "learning_rate": 4.588713187291455e-06, "loss": 0.0087, "step": 21346 }, { "epoch": 2.3142888117953166, "grad_norm": 0.5577757954597473, "learning_rate": 4.587987813724069e-06, "loss": 0.0144, "step": 21347 }, { "epoch": 2.314397224631396, "grad_norm": 2.536346435546875, "learning_rate": 4.587262440156681e-06, "loss": 0.0105, "step": 21348 }, { "epoch": 2.314505637467476, "grad_norm": 0.45289137959480286, "learning_rate": 4.586537066589294e-06, "loss": 0.0109, "step": 21349 }, { "epoch": 2.3146140503035557, "grad_norm": 0.3020683825016022, "learning_rate": 4.585811693021906e-06, "loss": 0.0092, "step": 21350 }, { "epoch": 2.3147224631396357, "grad_norm": 4.721759796142578, "learning_rate": 4.585086319454519e-06, "loss": 0.0253, "step": 21351 }, { "epoch": 2.3148308759757157, "grad_norm": 0.3824593722820282, "learning_rate": 4.584360945887132e-06, "loss": 0.0069, "step": 21352 }, { "epoch": 2.3149392888117952, "grad_norm": 1.811544418334961, "learning_rate": 4.583635572319745e-06, "loss": 0.0227, "step": 21353 }, { "epoch": 2.315047701647875, "grad_norm": 0.40926557779312134, "learning_rate": 4.582910198752358e-06, "loss": 0.0165, "step": 21354 }, { "epoch": 2.3151561144839548, "grad_norm": 0.4645478427410126, "learning_rate": 4.58218482518497e-06, "loss": 0.0135, "step": 21355 }, { "epoch": 2.3152645273200347, "grad_norm": 0.24698130786418915, "learning_rate": 4.581459451617584e-06, "loss": 0.0046, "step": 21356 }, { "epoch": 2.3153729401561147, "grad_norm": 0.31940099596977234, "learning_rate": 4.580734078050196e-06, "loss": 0.0115, "step": 21357 }, { "epoch": 2.3154813529921943, "grad_norm": 0.0038009253330528736, "learning_rate": 4.580008704482809e-06, "loss": 0.0, "step": 21358 }, { "epoch": 2.3155897658282742, "grad_norm": 0.04893561452627182, "learning_rate": 4.579283330915421e-06, "loss": 0.0002, "step": 21359 }, { "epoch": 2.315698178664354, "grad_norm": 0.48151808977127075, "learning_rate": 4.578557957348035e-06, "loss": 0.0073, "step": 21360 }, { "epoch": 2.3158065915004338, "grad_norm": 0.4318545162677765, "learning_rate": 4.577832583780647e-06, "loss": 0.004, "step": 21361 }, { "epoch": 2.3159150043365133, "grad_norm": 0.5316488742828369, "learning_rate": 4.57710721021326e-06, "loss": 0.0159, "step": 21362 }, { "epoch": 2.3160234171725933, "grad_norm": 0.17964957654476166, "learning_rate": 4.5763818366458725e-06, "loss": 0.0005, "step": 21363 }, { "epoch": 2.316131830008673, "grad_norm": 5.040848255157471, "learning_rate": 4.5756564630784855e-06, "loss": 0.0128, "step": 21364 }, { "epoch": 2.316240242844753, "grad_norm": 0.007294190116226673, "learning_rate": 4.5749310895110984e-06, "loss": 0.0001, "step": 21365 }, { "epoch": 2.316348655680833, "grad_norm": 0.6195653676986694, "learning_rate": 4.5742057159437114e-06, "loss": 0.0091, "step": 21366 }, { "epoch": 2.3164570685169124, "grad_norm": 0.39944663643836975, "learning_rate": 4.5734803423763244e-06, "loss": 0.0134, "step": 21367 }, { "epoch": 2.3165654813529923, "grad_norm": 1.9808909893035889, "learning_rate": 4.5727549688089366e-06, "loss": 0.0283, "step": 21368 }, { "epoch": 2.316673894189072, "grad_norm": 11.542243003845215, "learning_rate": 4.57202959524155e-06, "loss": 0.0327, "step": 21369 }, { "epoch": 2.316782307025152, "grad_norm": 3.3867690563201904, "learning_rate": 4.5713042216741626e-06, "loss": 0.0188, "step": 21370 }, { "epoch": 2.3168907198612314, "grad_norm": 1.145829677581787, "learning_rate": 4.5705788481067755e-06, "loss": 0.0123, "step": 21371 }, { "epoch": 2.3169991326973114, "grad_norm": 0.8378096222877502, "learning_rate": 4.569853474539388e-06, "loss": 0.0017, "step": 21372 }, { "epoch": 2.317107545533391, "grad_norm": 0.2562037408351898, "learning_rate": 4.5691281009720015e-06, "loss": 0.0041, "step": 21373 }, { "epoch": 2.317215958369471, "grad_norm": 0.04668453708291054, "learning_rate": 4.568402727404614e-06, "loss": 0.0002, "step": 21374 }, { "epoch": 2.317324371205551, "grad_norm": 0.5276867747306824, "learning_rate": 4.567677353837227e-06, "loss": 0.0131, "step": 21375 }, { "epoch": 2.3174327840416304, "grad_norm": 10.924412727355957, "learning_rate": 4.566951980269839e-06, "loss": 0.0055, "step": 21376 }, { "epoch": 2.3175411968777104, "grad_norm": 1.7678414583206177, "learning_rate": 4.566226606702452e-06, "loss": 0.0017, "step": 21377 }, { "epoch": 2.31764960971379, "grad_norm": 0.9337078332901001, "learning_rate": 4.565501233135065e-06, "loss": 0.0076, "step": 21378 }, { "epoch": 2.31775802254987, "grad_norm": 0.011370211839675903, "learning_rate": 4.564775859567678e-06, "loss": 0.0001, "step": 21379 }, { "epoch": 2.31786643538595, "grad_norm": 8.339864730834961, "learning_rate": 4.564050486000291e-06, "loss": 0.062, "step": 21380 }, { "epoch": 2.3179748482220295, "grad_norm": 0.7062599658966064, "learning_rate": 4.563325112432903e-06, "loss": 0.0342, "step": 21381 }, { "epoch": 2.3180832610581095, "grad_norm": 0.7848585247993469, "learning_rate": 4.562599738865517e-06, "loss": 0.0123, "step": 21382 }, { "epoch": 2.318191673894189, "grad_norm": 0.9305947422981262, "learning_rate": 4.561874365298129e-06, "loss": 0.0202, "step": 21383 }, { "epoch": 2.318300086730269, "grad_norm": 0.592914879322052, "learning_rate": 4.561148991730742e-06, "loss": 0.0129, "step": 21384 }, { "epoch": 2.3184084995663485, "grad_norm": 0.18659359216690063, "learning_rate": 4.560423618163354e-06, "loss": 0.0008, "step": 21385 }, { "epoch": 2.3185169124024285, "grad_norm": 0.4359987676143646, "learning_rate": 4.559698244595968e-06, "loss": 0.0115, "step": 21386 }, { "epoch": 2.318625325238508, "grad_norm": 1.878544807434082, "learning_rate": 4.55897287102858e-06, "loss": 0.0017, "step": 21387 }, { "epoch": 2.318733738074588, "grad_norm": 0.020058467984199524, "learning_rate": 4.558247497461193e-06, "loss": 0.0001, "step": 21388 }, { "epoch": 2.318842150910668, "grad_norm": 3.8699333667755127, "learning_rate": 4.557522123893805e-06, "loss": 0.0497, "step": 21389 }, { "epoch": 2.3189505637467476, "grad_norm": 0.10791300237178802, "learning_rate": 4.556796750326418e-06, "loss": 0.0003, "step": 21390 }, { "epoch": 2.3190589765828276, "grad_norm": 3.1220991611480713, "learning_rate": 4.556071376759031e-06, "loss": 0.0213, "step": 21391 }, { "epoch": 2.319167389418907, "grad_norm": 0.30108341574668884, "learning_rate": 4.555346003191644e-06, "loss": 0.0066, "step": 21392 }, { "epoch": 2.319275802254987, "grad_norm": 0.38836368918418884, "learning_rate": 4.554620629624257e-06, "loss": 0.007, "step": 21393 }, { "epoch": 2.3193842150910666, "grad_norm": 9.050440788269043, "learning_rate": 4.553895256056869e-06, "loss": 0.0134, "step": 21394 }, { "epoch": 2.3194926279271466, "grad_norm": 6.166985988616943, "learning_rate": 4.553169882489483e-06, "loss": 0.059, "step": 21395 }, { "epoch": 2.319601040763226, "grad_norm": 4.8786845207214355, "learning_rate": 4.552444508922095e-06, "loss": 0.0071, "step": 21396 }, { "epoch": 2.319709453599306, "grad_norm": 0.4982666075229645, "learning_rate": 4.551719135354708e-06, "loss": 0.0083, "step": 21397 }, { "epoch": 2.319817866435386, "grad_norm": 0.03773815184831619, "learning_rate": 4.55099376178732e-06, "loss": 0.0002, "step": 21398 }, { "epoch": 2.3199262792714657, "grad_norm": 0.0077078030444681644, "learning_rate": 4.550268388219934e-06, "loss": 0.0001, "step": 21399 }, { "epoch": 2.3200346921075456, "grad_norm": 5.23555850982666, "learning_rate": 4.549543014652546e-06, "loss": 0.0106, "step": 21400 }, { "epoch": 2.320143104943625, "grad_norm": 0.45292091369628906, "learning_rate": 4.548817641085159e-06, "loss": 0.0078, "step": 21401 }, { "epoch": 2.320251517779705, "grad_norm": 0.5732051730155945, "learning_rate": 4.5480922675177715e-06, "loss": 0.0179, "step": 21402 }, { "epoch": 2.3203599306157847, "grad_norm": 6.051377773284912, "learning_rate": 4.5473668939503845e-06, "loss": 0.0556, "step": 21403 }, { "epoch": 2.3204683434518647, "grad_norm": 0.031417299062013626, "learning_rate": 4.5466415203829975e-06, "loss": 0.0002, "step": 21404 }, { "epoch": 2.3205767562879447, "grad_norm": 6.771427631378174, "learning_rate": 4.5459161468156104e-06, "loss": 0.0134, "step": 21405 }, { "epoch": 2.3206851691240242, "grad_norm": 0.010334222577512264, "learning_rate": 4.5451907732482234e-06, "loss": 0.0001, "step": 21406 }, { "epoch": 2.320793581960104, "grad_norm": 19.633811950683594, "learning_rate": 4.5444653996808356e-06, "loss": 0.0323, "step": 21407 }, { "epoch": 2.3209019947961838, "grad_norm": 7.939202308654785, "learning_rate": 4.543740026113449e-06, "loss": 0.0225, "step": 21408 }, { "epoch": 2.3210104076322637, "grad_norm": 0.40294578671455383, "learning_rate": 4.5430146525460616e-06, "loss": 0.0116, "step": 21409 }, { "epoch": 2.3211188204683433, "grad_norm": 2.1766908168792725, "learning_rate": 4.5422892789786745e-06, "loss": 0.0062, "step": 21410 }, { "epoch": 2.3212272333044233, "grad_norm": 0.019472476094961166, "learning_rate": 4.541563905411287e-06, "loss": 0.0001, "step": 21411 }, { "epoch": 2.3213356461405033, "grad_norm": 0.0412672683596611, "learning_rate": 4.5408385318439005e-06, "loss": 0.0003, "step": 21412 }, { "epoch": 2.321444058976583, "grad_norm": 3.220292091369629, "learning_rate": 4.540113158276513e-06, "loss": 0.0362, "step": 21413 }, { "epoch": 2.3215524718126628, "grad_norm": 2.616194009780884, "learning_rate": 4.539387784709126e-06, "loss": 0.0165, "step": 21414 }, { "epoch": 2.3216608846487423, "grad_norm": 0.3371627628803253, "learning_rate": 4.538662411141738e-06, "loss": 0.0072, "step": 21415 }, { "epoch": 2.3217692974848223, "grad_norm": 3.1945641040802, "learning_rate": 4.537937037574351e-06, "loss": 0.027, "step": 21416 }, { "epoch": 2.321877710320902, "grad_norm": 0.22097167372703552, "learning_rate": 4.537211664006964e-06, "loss": 0.0006, "step": 21417 }, { "epoch": 2.321986123156982, "grad_norm": 0.010610738769173622, "learning_rate": 4.536486290439577e-06, "loss": 0.0001, "step": 21418 }, { "epoch": 2.3220945359930614, "grad_norm": 0.24619680643081665, "learning_rate": 4.53576091687219e-06, "loss": 0.0048, "step": 21419 }, { "epoch": 2.3222029488291414, "grad_norm": 0.09591612219810486, "learning_rate": 4.535035543304802e-06, "loss": 0.0003, "step": 21420 }, { "epoch": 2.3223113616652213, "grad_norm": 0.44501829147338867, "learning_rate": 4.534310169737416e-06, "loss": 0.0113, "step": 21421 }, { "epoch": 2.322419774501301, "grad_norm": 0.08446703106164932, "learning_rate": 4.533584796170028e-06, "loss": 0.0004, "step": 21422 }, { "epoch": 2.322528187337381, "grad_norm": 0.8062124252319336, "learning_rate": 4.532859422602641e-06, "loss": 0.0076, "step": 21423 }, { "epoch": 2.3226366001734604, "grad_norm": 2.8976707458496094, "learning_rate": 4.532134049035253e-06, "loss": 0.0118, "step": 21424 }, { "epoch": 2.3227450130095404, "grad_norm": 3.5962984561920166, "learning_rate": 4.531408675467867e-06, "loss": 0.011, "step": 21425 }, { "epoch": 2.32285342584562, "grad_norm": 0.3212483525276184, "learning_rate": 4.530683301900479e-06, "loss": 0.0044, "step": 21426 }, { "epoch": 2.3229618386817, "grad_norm": 0.051526471972465515, "learning_rate": 4.529957928333092e-06, "loss": 0.0004, "step": 21427 }, { "epoch": 2.3230702515177795, "grad_norm": 0.08400838077068329, "learning_rate": 4.529232554765704e-06, "loss": 0.0005, "step": 21428 }, { "epoch": 2.3231786643538594, "grad_norm": 2.5630390644073486, "learning_rate": 4.528507181198318e-06, "loss": 0.0178, "step": 21429 }, { "epoch": 2.3232870771899394, "grad_norm": 0.47292253375053406, "learning_rate": 4.52778180763093e-06, "loss": 0.0204, "step": 21430 }, { "epoch": 2.323395490026019, "grad_norm": 0.9825146198272705, "learning_rate": 4.527056434063543e-06, "loss": 0.0091, "step": 21431 }, { "epoch": 2.323503902862099, "grad_norm": 0.1156364232301712, "learning_rate": 4.526331060496156e-06, "loss": 0.0005, "step": 21432 }, { "epoch": 2.3236123156981785, "grad_norm": 2.6676671504974365, "learning_rate": 4.525605686928768e-06, "loss": 0.0194, "step": 21433 }, { "epoch": 2.3237207285342585, "grad_norm": 1.4006017446517944, "learning_rate": 4.524880313361382e-06, "loss": 0.0063, "step": 21434 }, { "epoch": 2.3238291413703385, "grad_norm": 0.3629562556743622, "learning_rate": 4.524154939793994e-06, "loss": 0.0073, "step": 21435 }, { "epoch": 2.323937554206418, "grad_norm": 1.4915145635604858, "learning_rate": 4.523429566226607e-06, "loss": 0.0468, "step": 21436 }, { "epoch": 2.324045967042498, "grad_norm": 0.15939612686634064, "learning_rate": 4.522704192659219e-06, "loss": 0.0005, "step": 21437 }, { "epoch": 2.3241543798785775, "grad_norm": 6.023175239562988, "learning_rate": 4.521978819091833e-06, "loss": 0.0357, "step": 21438 }, { "epoch": 2.3242627927146575, "grad_norm": 0.617872953414917, "learning_rate": 4.521253445524445e-06, "loss": 0.0098, "step": 21439 }, { "epoch": 2.324371205550737, "grad_norm": 0.5574345588684082, "learning_rate": 4.520528071957058e-06, "loss": 0.0151, "step": 21440 }, { "epoch": 2.324479618386817, "grad_norm": 0.28612375259399414, "learning_rate": 4.5198026983896705e-06, "loss": 0.0061, "step": 21441 }, { "epoch": 2.3245880312228966, "grad_norm": 0.2804093360900879, "learning_rate": 4.519077324822284e-06, "loss": 0.0093, "step": 21442 }, { "epoch": 2.3246964440589766, "grad_norm": 3.8138835430145264, "learning_rate": 4.5183519512548965e-06, "loss": 0.0259, "step": 21443 }, { "epoch": 2.3248048568950566, "grad_norm": 3.8701677322387695, "learning_rate": 4.5176265776875094e-06, "loss": 0.0217, "step": 21444 }, { "epoch": 2.324913269731136, "grad_norm": 0.47345107793807983, "learning_rate": 4.5169012041201224e-06, "loss": 0.0165, "step": 21445 }, { "epoch": 2.325021682567216, "grad_norm": 0.32737091183662415, "learning_rate": 4.516175830552735e-06, "loss": 0.014, "step": 21446 }, { "epoch": 2.3251300954032956, "grad_norm": 2.4642324447631836, "learning_rate": 4.515450456985348e-06, "loss": 0.0215, "step": 21447 }, { "epoch": 2.3252385082393756, "grad_norm": 0.18632306158542633, "learning_rate": 4.5147250834179606e-06, "loss": 0.0004, "step": 21448 }, { "epoch": 2.325346921075455, "grad_norm": 0.026804368942975998, "learning_rate": 4.5139997098505735e-06, "loss": 0.0002, "step": 21449 }, { "epoch": 2.325455333911535, "grad_norm": 1.6949821710586548, "learning_rate": 4.513274336283186e-06, "loss": 0.0062, "step": 21450 }, { "epoch": 2.3255637467476147, "grad_norm": 1.0328627824783325, "learning_rate": 4.5125489627157995e-06, "loss": 0.0287, "step": 21451 }, { "epoch": 2.3256721595836947, "grad_norm": 0.3428414463996887, "learning_rate": 4.511823589148412e-06, "loss": 0.0059, "step": 21452 }, { "epoch": 2.3257805724197746, "grad_norm": 0.2618289887905121, "learning_rate": 4.511098215581025e-06, "loss": 0.0054, "step": 21453 }, { "epoch": 2.325888985255854, "grad_norm": 2.282637119293213, "learning_rate": 4.510372842013638e-06, "loss": 0.0474, "step": 21454 }, { "epoch": 2.325997398091934, "grad_norm": 0.33095479011535645, "learning_rate": 4.509647468446251e-06, "loss": 0.0058, "step": 21455 }, { "epoch": 2.3261058109280137, "grad_norm": 0.9815440773963928, "learning_rate": 4.508922094878863e-06, "loss": 0.0115, "step": 21456 }, { "epoch": 2.3262142237640937, "grad_norm": 0.026543082669377327, "learning_rate": 4.508196721311476e-06, "loss": 0.0003, "step": 21457 }, { "epoch": 2.3263226366001737, "grad_norm": 0.2633027136325836, "learning_rate": 4.507471347744089e-06, "loss": 0.0057, "step": 21458 }, { "epoch": 2.3264310494362532, "grad_norm": 0.1188574954867363, "learning_rate": 4.506745974176701e-06, "loss": 0.0004, "step": 21459 }, { "epoch": 2.326539462272333, "grad_norm": 0.5222315192222595, "learning_rate": 4.506020600609315e-06, "loss": 0.014, "step": 21460 }, { "epoch": 2.3266478751084128, "grad_norm": 2.277099609375, "learning_rate": 4.505295227041927e-06, "loss": 0.0221, "step": 21461 }, { "epoch": 2.3267562879444927, "grad_norm": 0.31095758080482483, "learning_rate": 4.50456985347454e-06, "loss": 0.0122, "step": 21462 }, { "epoch": 2.3268647007805723, "grad_norm": 0.537693440914154, "learning_rate": 4.503844479907152e-06, "loss": 0.0011, "step": 21463 }, { "epoch": 2.3269731136166523, "grad_norm": 5.149623394012451, "learning_rate": 4.503119106339766e-06, "loss": 0.0038, "step": 21464 }, { "epoch": 2.327081526452732, "grad_norm": 4.174661159515381, "learning_rate": 4.502393732772378e-06, "loss": 0.016, "step": 21465 }, { "epoch": 2.327189939288812, "grad_norm": 4.805771350860596, "learning_rate": 4.501668359204991e-06, "loss": 0.0153, "step": 21466 }, { "epoch": 2.327298352124892, "grad_norm": 0.2005947083234787, "learning_rate": 4.500942985637604e-06, "loss": 0.0013, "step": 21467 }, { "epoch": 2.3274067649609713, "grad_norm": 1.2812734842300415, "learning_rate": 4.500217612070217e-06, "loss": 0.0137, "step": 21468 }, { "epoch": 2.3275151777970513, "grad_norm": 0.3816736936569214, "learning_rate": 4.499492238502829e-06, "loss": 0.0095, "step": 21469 }, { "epoch": 2.327623590633131, "grad_norm": 0.2661404013633728, "learning_rate": 4.498766864935442e-06, "loss": 0.0055, "step": 21470 }, { "epoch": 2.327732003469211, "grad_norm": 0.24404621124267578, "learning_rate": 4.498041491368055e-06, "loss": 0.0011, "step": 21471 }, { "epoch": 2.3278404163052904, "grad_norm": 2.3389432430267334, "learning_rate": 4.497316117800667e-06, "loss": 0.0561, "step": 21472 }, { "epoch": 2.3279488291413704, "grad_norm": 0.0019259697292000055, "learning_rate": 4.496590744233281e-06, "loss": 0.0, "step": 21473 }, { "epoch": 2.32805724197745, "grad_norm": 2.522625207901001, "learning_rate": 4.495865370665893e-06, "loss": 0.0321, "step": 21474 }, { "epoch": 2.32816565481353, "grad_norm": 0.37206366658210754, "learning_rate": 4.495139997098506e-06, "loss": 0.0048, "step": 21475 }, { "epoch": 2.32827406764961, "grad_norm": 3.277501106262207, "learning_rate": 4.494414623531118e-06, "loss": 0.018, "step": 21476 }, { "epoch": 2.3283824804856894, "grad_norm": 2.395846128463745, "learning_rate": 4.493689249963732e-06, "loss": 0.0366, "step": 21477 }, { "epoch": 2.3284908933217694, "grad_norm": 0.5952560901641846, "learning_rate": 4.492963876396344e-06, "loss": 0.0135, "step": 21478 }, { "epoch": 2.328599306157849, "grad_norm": 0.6779518127441406, "learning_rate": 4.492238502828957e-06, "loss": 0.0281, "step": 21479 }, { "epoch": 2.328707718993929, "grad_norm": 0.28185561299324036, "learning_rate": 4.49151312926157e-06, "loss": 0.0057, "step": 21480 }, { "epoch": 2.328816131830009, "grad_norm": 0.026708904653787613, "learning_rate": 4.490787755694183e-06, "loss": 0.0001, "step": 21481 }, { "epoch": 2.3289245446660884, "grad_norm": 0.4226400852203369, "learning_rate": 4.4900623821267955e-06, "loss": 0.0118, "step": 21482 }, { "epoch": 2.3290329575021684, "grad_norm": 0.5750205516815186, "learning_rate": 4.4893370085594084e-06, "loss": 0.0016, "step": 21483 }, { "epoch": 2.329141370338248, "grad_norm": 0.02216571941971779, "learning_rate": 4.4886116349920214e-06, "loss": 0.0002, "step": 21484 }, { "epoch": 2.329249783174328, "grad_norm": 0.01068202219903469, "learning_rate": 4.487886261424634e-06, "loss": 0.0001, "step": 21485 }, { "epoch": 2.3293581960104075, "grad_norm": 0.27809593081474304, "learning_rate": 4.487160887857247e-06, "loss": 0.006, "step": 21486 }, { "epoch": 2.3294666088464875, "grad_norm": 0.8881129026412964, "learning_rate": 4.4864355142898596e-06, "loss": 0.0152, "step": 21487 }, { "epoch": 2.329575021682567, "grad_norm": 0.3403133451938629, "learning_rate": 4.4857101407224726e-06, "loss": 0.0073, "step": 21488 }, { "epoch": 2.329683434518647, "grad_norm": 0.29828232526779175, "learning_rate": 4.484984767155085e-06, "loss": 0.0068, "step": 21489 }, { "epoch": 2.329791847354727, "grad_norm": 0.29557478427886963, "learning_rate": 4.4842593935876985e-06, "loss": 0.0052, "step": 21490 }, { "epoch": 2.3299002601908065, "grad_norm": 0.4703063666820526, "learning_rate": 4.483534020020311e-06, "loss": 0.0171, "step": 21491 }, { "epoch": 2.3300086730268865, "grad_norm": 0.326567143201828, "learning_rate": 4.482808646452924e-06, "loss": 0.0071, "step": 21492 }, { "epoch": 2.330117085862966, "grad_norm": 13.03668212890625, "learning_rate": 4.482083272885537e-06, "loss": 0.0235, "step": 21493 }, { "epoch": 2.330225498699046, "grad_norm": 0.43625932931900024, "learning_rate": 4.48135789931815e-06, "loss": 0.0167, "step": 21494 }, { "epoch": 2.3303339115351256, "grad_norm": 2.321981430053711, "learning_rate": 4.480632525750762e-06, "loss": 0.0464, "step": 21495 }, { "epoch": 2.3304423243712056, "grad_norm": 0.304368793964386, "learning_rate": 4.479907152183375e-06, "loss": 0.0112, "step": 21496 }, { "epoch": 2.330550737207285, "grad_norm": 0.5262745022773743, "learning_rate": 4.479181778615988e-06, "loss": 0.0124, "step": 21497 }, { "epoch": 2.330659150043365, "grad_norm": 4.323756694793701, "learning_rate": 4.4784564050486e-06, "loss": 0.0194, "step": 21498 }, { "epoch": 2.330767562879445, "grad_norm": 0.15754373371601105, "learning_rate": 4.477731031481214e-06, "loss": 0.0012, "step": 21499 }, { "epoch": 2.3308759757155246, "grad_norm": 0.21972672641277313, "learning_rate": 4.477005657913826e-06, "loss": 0.001, "step": 21500 }, { "epoch": 2.3309843885516046, "grad_norm": 0.34504854679107666, "learning_rate": 4.476280284346439e-06, "loss": 0.0057, "step": 21501 }, { "epoch": 2.331092801387684, "grad_norm": 0.09951137006282806, "learning_rate": 4.475554910779051e-06, "loss": 0.0008, "step": 21502 }, { "epoch": 2.331201214223764, "grad_norm": 0.01905568689107895, "learning_rate": 4.474829537211665e-06, "loss": 0.0001, "step": 21503 }, { "epoch": 2.331309627059844, "grad_norm": 5.15036678314209, "learning_rate": 4.474104163644277e-06, "loss": 0.019, "step": 21504 }, { "epoch": 2.3314180398959237, "grad_norm": 1.5859553813934326, "learning_rate": 4.47337879007689e-06, "loss": 0.0357, "step": 21505 }, { "epoch": 2.3315264527320037, "grad_norm": 0.09483890235424042, "learning_rate": 4.472653416509503e-06, "loss": 0.0004, "step": 21506 }, { "epoch": 2.331634865568083, "grad_norm": 0.005196619313210249, "learning_rate": 4.471928042942116e-06, "loss": 0.0, "step": 21507 }, { "epoch": 2.331743278404163, "grad_norm": 4.483259677886963, "learning_rate": 4.471202669374728e-06, "loss": 0.0194, "step": 21508 }, { "epoch": 2.3318516912402427, "grad_norm": 0.650220513343811, "learning_rate": 4.470477295807341e-06, "loss": 0.0166, "step": 21509 }, { "epoch": 2.3319601040763227, "grad_norm": 0.22190698981285095, "learning_rate": 4.469751922239954e-06, "loss": 0.0011, "step": 21510 }, { "epoch": 2.3320685169124022, "grad_norm": 0.05682738870382309, "learning_rate": 4.469026548672566e-06, "loss": 0.0004, "step": 21511 }, { "epoch": 2.3321769297484822, "grad_norm": 0.20238390564918518, "learning_rate": 4.46830117510518e-06, "loss": 0.0009, "step": 21512 }, { "epoch": 2.332285342584562, "grad_norm": 0.298943430185318, "learning_rate": 4.467575801537792e-06, "loss": 0.0062, "step": 21513 }, { "epoch": 2.3323937554206418, "grad_norm": 0.4175874888896942, "learning_rate": 4.466850427970405e-06, "loss": 0.0056, "step": 21514 }, { "epoch": 2.3325021682567217, "grad_norm": 0.8731280565261841, "learning_rate": 4.466125054403017e-06, "loss": 0.0163, "step": 21515 }, { "epoch": 2.3326105810928013, "grad_norm": 3.1024882793426514, "learning_rate": 4.465399680835631e-06, "loss": 0.0042, "step": 21516 }, { "epoch": 2.3327189939288813, "grad_norm": 2.922096014022827, "learning_rate": 4.464674307268243e-06, "loss": 0.0111, "step": 21517 }, { "epoch": 2.332827406764961, "grad_norm": 2.670328378677368, "learning_rate": 4.463948933700856e-06, "loss": 0.0274, "step": 21518 }, { "epoch": 2.332935819601041, "grad_norm": 0.6331780552864075, "learning_rate": 4.463223560133469e-06, "loss": 0.0085, "step": 21519 }, { "epoch": 2.3330442324371203, "grad_norm": 2.7316601276397705, "learning_rate": 4.462498186566082e-06, "loss": 0.0233, "step": 21520 }, { "epoch": 2.3331526452732003, "grad_norm": 0.3590830862522125, "learning_rate": 4.4617728129986945e-06, "loss": 0.0061, "step": 21521 }, { "epoch": 2.3332610581092803, "grad_norm": 5.32441520690918, "learning_rate": 4.4610474394313075e-06, "loss": 0.0163, "step": 21522 }, { "epoch": 2.33336947094536, "grad_norm": 0.21864484250545502, "learning_rate": 4.4603220658639204e-06, "loss": 0.0045, "step": 21523 }, { "epoch": 2.33347788378144, "grad_norm": 0.1512618362903595, "learning_rate": 4.459596692296533e-06, "loss": 0.0025, "step": 21524 }, { "epoch": 2.3335862966175194, "grad_norm": 2.7501394748687744, "learning_rate": 4.458871318729146e-06, "loss": 0.0047, "step": 21525 }, { "epoch": 2.3336947094535994, "grad_norm": 1.4048277139663696, "learning_rate": 4.4581459451617586e-06, "loss": 0.0017, "step": 21526 }, { "epoch": 2.3338031222896793, "grad_norm": 0.6347392797470093, "learning_rate": 4.4574205715943716e-06, "loss": 0.0209, "step": 21527 }, { "epoch": 2.333911535125759, "grad_norm": 0.33160150051116943, "learning_rate": 4.456695198026984e-06, "loss": 0.0071, "step": 21528 }, { "epoch": 2.334019947961839, "grad_norm": 1.0145211219787598, "learning_rate": 4.4559698244595975e-06, "loss": 0.0028, "step": 21529 }, { "epoch": 2.3341283607979184, "grad_norm": 0.5092436075210571, "learning_rate": 4.45524445089221e-06, "loss": 0.0136, "step": 21530 }, { "epoch": 2.3342367736339984, "grad_norm": 2.968608856201172, "learning_rate": 4.454519077324823e-06, "loss": 0.009, "step": 21531 }, { "epoch": 2.334345186470078, "grad_norm": 0.031108012422919273, "learning_rate": 4.453793703757436e-06, "loss": 0.0001, "step": 21532 }, { "epoch": 2.334453599306158, "grad_norm": 3.150139331817627, "learning_rate": 4.453068330190049e-06, "loss": 0.0052, "step": 21533 }, { "epoch": 2.3345620121422375, "grad_norm": 4.366488456726074, "learning_rate": 4.452342956622661e-06, "loss": 0.0538, "step": 21534 }, { "epoch": 2.3346704249783174, "grad_norm": 4.177661418914795, "learning_rate": 4.451617583055274e-06, "loss": 0.0263, "step": 21535 }, { "epoch": 2.3347788378143974, "grad_norm": 0.07900417596101761, "learning_rate": 4.450892209487887e-06, "loss": 0.0003, "step": 21536 }, { "epoch": 2.334887250650477, "grad_norm": 1.5091348886489868, "learning_rate": 4.4501668359205e-06, "loss": 0.0342, "step": 21537 }, { "epoch": 2.334995663486557, "grad_norm": 3.9936110973358154, "learning_rate": 4.449441462353113e-06, "loss": 0.0158, "step": 21538 }, { "epoch": 2.3351040763226365, "grad_norm": 5.706121921539307, "learning_rate": 4.448716088785725e-06, "loss": 0.0604, "step": 21539 }, { "epoch": 2.3352124891587165, "grad_norm": 3.2856569290161133, "learning_rate": 4.447990715218338e-06, "loss": 0.0193, "step": 21540 }, { "epoch": 2.335320901994796, "grad_norm": 8.033005714416504, "learning_rate": 4.44726534165095e-06, "loss": 0.0567, "step": 21541 }, { "epoch": 2.335429314830876, "grad_norm": 8.948765754699707, "learning_rate": 4.446539968083564e-06, "loss": 0.0383, "step": 21542 }, { "epoch": 2.3355377276669556, "grad_norm": 0.050550635904073715, "learning_rate": 4.445814594516176e-06, "loss": 0.0002, "step": 21543 }, { "epoch": 2.3356461405030355, "grad_norm": 0.03708961233496666, "learning_rate": 4.445089220948789e-06, "loss": 0.0001, "step": 21544 }, { "epoch": 2.3357545533391155, "grad_norm": 1.8236819505691528, "learning_rate": 4.444363847381402e-06, "loss": 0.0027, "step": 21545 }, { "epoch": 2.335862966175195, "grad_norm": 0.6368491649627686, "learning_rate": 4.443638473814015e-06, "loss": 0.002, "step": 21546 }, { "epoch": 2.335971379011275, "grad_norm": 0.016502197831869125, "learning_rate": 4.442913100246627e-06, "loss": 0.0001, "step": 21547 }, { "epoch": 2.3360797918473546, "grad_norm": 0.35979416966438293, "learning_rate": 4.44218772667924e-06, "loss": 0.0065, "step": 21548 }, { "epoch": 2.3361882046834346, "grad_norm": 0.2372852861881256, "learning_rate": 4.441462353111853e-06, "loss": 0.0183, "step": 21549 }, { "epoch": 2.336296617519514, "grad_norm": 2.7903435230255127, "learning_rate": 4.440736979544466e-06, "loss": 0.0362, "step": 21550 }, { "epoch": 2.336405030355594, "grad_norm": 0.01818171702325344, "learning_rate": 4.440011605977079e-06, "loss": 0.0002, "step": 21551 }, { "epoch": 2.3365134431916736, "grad_norm": 1.879419207572937, "learning_rate": 4.439286232409691e-06, "loss": 0.0182, "step": 21552 }, { "epoch": 2.3366218560277536, "grad_norm": 1.049736499786377, "learning_rate": 4.438560858842304e-06, "loss": 0.014, "step": 21553 }, { "epoch": 2.3367302688638336, "grad_norm": 0.7072759866714478, "learning_rate": 4.437835485274916e-06, "loss": 0.0175, "step": 21554 }, { "epoch": 2.336838681699913, "grad_norm": 5.089236736297607, "learning_rate": 4.43711011170753e-06, "loss": 0.0155, "step": 21555 }, { "epoch": 2.336947094535993, "grad_norm": 0.5076316595077515, "learning_rate": 4.436384738140142e-06, "loss": 0.002, "step": 21556 }, { "epoch": 2.3370555073720727, "grad_norm": 0.29281795024871826, "learning_rate": 4.435659364572755e-06, "loss": 0.0019, "step": 21557 }, { "epoch": 2.3371639202081527, "grad_norm": 3.370000123977661, "learning_rate": 4.434933991005368e-06, "loss": 0.0092, "step": 21558 }, { "epoch": 2.3372723330442327, "grad_norm": 2.7123961448669434, "learning_rate": 4.434208617437981e-06, "loss": 0.1099, "step": 21559 }, { "epoch": 2.337380745880312, "grad_norm": 0.2545771896839142, "learning_rate": 4.4334832438705935e-06, "loss": 0.0055, "step": 21560 }, { "epoch": 2.337489158716392, "grad_norm": 0.02986067160964012, "learning_rate": 4.4327578703032065e-06, "loss": 0.0002, "step": 21561 }, { "epoch": 2.3375975715524717, "grad_norm": 0.4259830713272095, "learning_rate": 4.4320324967358194e-06, "loss": 0.0062, "step": 21562 }, { "epoch": 2.3377059843885517, "grad_norm": 4.240977764129639, "learning_rate": 4.4313071231684324e-06, "loss": 0.0444, "step": 21563 }, { "epoch": 2.3378143972246312, "grad_norm": 0.2812720537185669, "learning_rate": 4.4305817496010454e-06, "loss": 0.0055, "step": 21564 }, { "epoch": 2.3379228100607112, "grad_norm": 22.325098037719727, "learning_rate": 4.4298563760336576e-06, "loss": 0.0321, "step": 21565 }, { "epoch": 2.3380312228967908, "grad_norm": 0.03741564601659775, "learning_rate": 4.4291310024662706e-06, "loss": 0.0002, "step": 21566 }, { "epoch": 2.3381396357328708, "grad_norm": 1.8525314331054688, "learning_rate": 4.428405628898883e-06, "loss": 0.0445, "step": 21567 }, { "epoch": 2.3382480485689507, "grad_norm": 0.03879471495747566, "learning_rate": 4.4276802553314965e-06, "loss": 0.0003, "step": 21568 }, { "epoch": 2.3383564614050303, "grad_norm": 0.5794922709465027, "learning_rate": 4.426954881764109e-06, "loss": 0.0147, "step": 21569 }, { "epoch": 2.3384648742411103, "grad_norm": 3.90019154548645, "learning_rate": 4.426229508196722e-06, "loss": 0.0112, "step": 21570 }, { "epoch": 2.33857328707719, "grad_norm": 0.03608286380767822, "learning_rate": 4.425504134629335e-06, "loss": 0.0003, "step": 21571 }, { "epoch": 2.33868169991327, "grad_norm": 0.45263200998306274, "learning_rate": 4.424778761061948e-06, "loss": 0.0058, "step": 21572 }, { "epoch": 2.3387901127493493, "grad_norm": 0.45848220586776733, "learning_rate": 4.42405338749456e-06, "loss": 0.0108, "step": 21573 }, { "epoch": 2.3388985255854293, "grad_norm": 0.6426528692245483, "learning_rate": 4.423328013927173e-06, "loss": 0.0058, "step": 21574 }, { "epoch": 2.339006938421509, "grad_norm": 0.7541071772575378, "learning_rate": 4.422602640359786e-06, "loss": 0.0109, "step": 21575 }, { "epoch": 2.339115351257589, "grad_norm": 0.9183974266052246, "learning_rate": 4.421877266792399e-06, "loss": 0.0027, "step": 21576 }, { "epoch": 2.339223764093669, "grad_norm": 0.004836175125092268, "learning_rate": 4.421151893225012e-06, "loss": 0.0, "step": 21577 }, { "epoch": 2.3393321769297484, "grad_norm": 0.05510638654232025, "learning_rate": 4.420426519657624e-06, "loss": 0.0005, "step": 21578 }, { "epoch": 2.3394405897658284, "grad_norm": 3.259559154510498, "learning_rate": 4.419701146090237e-06, "loss": 0.032, "step": 21579 }, { "epoch": 2.339549002601908, "grad_norm": 0.4785129129886627, "learning_rate": 4.418975772522849e-06, "loss": 0.019, "step": 21580 }, { "epoch": 2.339657415437988, "grad_norm": 0.037030644714832306, "learning_rate": 4.418250398955463e-06, "loss": 0.0002, "step": 21581 }, { "epoch": 2.339765828274068, "grad_norm": 0.5171322226524353, "learning_rate": 4.417525025388075e-06, "loss": 0.0221, "step": 21582 }, { "epoch": 2.3398742411101474, "grad_norm": 0.4564957916736603, "learning_rate": 4.416799651820688e-06, "loss": 0.019, "step": 21583 }, { "epoch": 2.3399826539462274, "grad_norm": 0.14412377774715424, "learning_rate": 4.416074278253301e-06, "loss": 0.0007, "step": 21584 }, { "epoch": 2.340091066782307, "grad_norm": 0.021288782358169556, "learning_rate": 4.415348904685914e-06, "loss": 0.0002, "step": 21585 }, { "epoch": 2.340199479618387, "grad_norm": 0.04132671281695366, "learning_rate": 4.414623531118526e-06, "loss": 0.0002, "step": 21586 }, { "epoch": 2.3403078924544665, "grad_norm": 0.587801456451416, "learning_rate": 4.413898157551139e-06, "loss": 0.02, "step": 21587 }, { "epoch": 2.3404163052905465, "grad_norm": 0.5773687958717346, "learning_rate": 4.413172783983752e-06, "loss": 0.0036, "step": 21588 }, { "epoch": 2.340524718126626, "grad_norm": 8.558241844177246, "learning_rate": 4.412447410416365e-06, "loss": 0.0383, "step": 21589 }, { "epoch": 2.340633130962706, "grad_norm": 0.2931949198246002, "learning_rate": 4.411722036848978e-06, "loss": 0.0042, "step": 21590 }, { "epoch": 2.340741543798786, "grad_norm": 2.6475727558135986, "learning_rate": 4.41099666328159e-06, "loss": 0.0576, "step": 21591 }, { "epoch": 2.3408499566348655, "grad_norm": 0.1598452776670456, "learning_rate": 4.410271289714203e-06, "loss": 0.0134, "step": 21592 }, { "epoch": 2.3409583694709455, "grad_norm": 0.016122378408908844, "learning_rate": 4.409545916146815e-06, "loss": 0.0002, "step": 21593 }, { "epoch": 2.341066782307025, "grad_norm": 5.321364402770996, "learning_rate": 4.408820542579429e-06, "loss": 0.0245, "step": 21594 }, { "epoch": 2.341175195143105, "grad_norm": 0.060545697808265686, "learning_rate": 4.408095169012041e-06, "loss": 0.0004, "step": 21595 }, { "epoch": 2.3412836079791846, "grad_norm": 0.5639581084251404, "learning_rate": 4.407369795444654e-06, "loss": 0.0126, "step": 21596 }, { "epoch": 2.3413920208152645, "grad_norm": 4.885937213897705, "learning_rate": 4.406644421877267e-06, "loss": 0.0327, "step": 21597 }, { "epoch": 2.341500433651344, "grad_norm": 7.7975640296936035, "learning_rate": 4.40591904830988e-06, "loss": 0.0432, "step": 21598 }, { "epoch": 2.341608846487424, "grad_norm": 0.35600951313972473, "learning_rate": 4.4051936747424925e-06, "loss": 0.0064, "step": 21599 }, { "epoch": 2.341717259323504, "grad_norm": 0.019726835191249847, "learning_rate": 4.4044683011751055e-06, "loss": 0.0002, "step": 21600 }, { "epoch": 2.3418256721595836, "grad_norm": 2.1078293323516846, "learning_rate": 4.4037429276077184e-06, "loss": 0.032, "step": 21601 }, { "epoch": 2.3419340849956636, "grad_norm": 1.8908289670944214, "learning_rate": 4.4030175540403314e-06, "loss": 0.0595, "step": 21602 }, { "epoch": 2.342042497831743, "grad_norm": 0.2883327007293701, "learning_rate": 4.4022921804729444e-06, "loss": 0.0031, "step": 21603 }, { "epoch": 2.342150910667823, "grad_norm": 10.542290687561035, "learning_rate": 4.4015668069055566e-06, "loss": 0.0216, "step": 21604 }, { "epoch": 2.342259323503903, "grad_norm": 0.2174888253211975, "learning_rate": 4.4008414333381696e-06, "loss": 0.0033, "step": 21605 }, { "epoch": 2.3423677363399826, "grad_norm": 0.5152731537818909, "learning_rate": 4.400116059770782e-06, "loss": 0.0179, "step": 21606 }, { "epoch": 2.3424761491760626, "grad_norm": 0.31123536825180054, "learning_rate": 4.3993906862033955e-06, "loss": 0.0111, "step": 21607 }, { "epoch": 2.342584562012142, "grad_norm": 0.2882225215435028, "learning_rate": 4.398665312636008e-06, "loss": 0.0028, "step": 21608 }, { "epoch": 2.342692974848222, "grad_norm": 0.20279966294765472, "learning_rate": 4.397939939068621e-06, "loss": 0.0005, "step": 21609 }, { "epoch": 2.3428013876843017, "grad_norm": 0.46178382635116577, "learning_rate": 4.397214565501234e-06, "loss": 0.0056, "step": 21610 }, { "epoch": 2.3429098005203817, "grad_norm": 0.28674858808517456, "learning_rate": 4.396489191933847e-06, "loss": 0.0045, "step": 21611 }, { "epoch": 2.343018213356461, "grad_norm": 0.37759873270988464, "learning_rate": 4.395763818366459e-06, "loss": 0.0112, "step": 21612 }, { "epoch": 2.343126626192541, "grad_norm": 2.390428304672241, "learning_rate": 4.395038444799072e-06, "loss": 0.0294, "step": 21613 }, { "epoch": 2.343235039028621, "grad_norm": 0.011225232854485512, "learning_rate": 4.394313071231685e-06, "loss": 0.0001, "step": 21614 }, { "epoch": 2.3433434518647007, "grad_norm": 2.508401393890381, "learning_rate": 4.393587697664298e-06, "loss": 0.0177, "step": 21615 }, { "epoch": 2.3434518647007807, "grad_norm": 4.973217487335205, "learning_rate": 4.392862324096911e-06, "loss": 0.0394, "step": 21616 }, { "epoch": 2.3435602775368602, "grad_norm": 2.8615455627441406, "learning_rate": 4.392136950529523e-06, "loss": 0.0038, "step": 21617 }, { "epoch": 2.3436686903729402, "grad_norm": 0.09620007872581482, "learning_rate": 4.391411576962136e-06, "loss": 0.0004, "step": 21618 }, { "epoch": 2.3437771032090198, "grad_norm": 0.24619925022125244, "learning_rate": 4.390686203394748e-06, "loss": 0.0013, "step": 21619 }, { "epoch": 2.3438855160450998, "grad_norm": 8.771409034729004, "learning_rate": 4.389960829827362e-06, "loss": 0.0272, "step": 21620 }, { "epoch": 2.3439939288811793, "grad_norm": 0.5277196764945984, "learning_rate": 4.389235456259974e-06, "loss": 0.0073, "step": 21621 }, { "epoch": 2.3441023417172593, "grad_norm": 0.31143492460250854, "learning_rate": 4.388510082692587e-06, "loss": 0.012, "step": 21622 }, { "epoch": 2.3442107545533393, "grad_norm": 6.401997089385986, "learning_rate": 4.3877847091252e-06, "loss": 0.0113, "step": 21623 }, { "epoch": 2.344319167389419, "grad_norm": 16.466569900512695, "learning_rate": 4.387059335557813e-06, "loss": 0.0976, "step": 21624 }, { "epoch": 2.344427580225499, "grad_norm": 16.761371612548828, "learning_rate": 4.386333961990425e-06, "loss": 0.0247, "step": 21625 }, { "epoch": 2.3445359930615783, "grad_norm": 0.27437055110931396, "learning_rate": 4.385608588423038e-06, "loss": 0.0094, "step": 21626 }, { "epoch": 2.3446444058976583, "grad_norm": 0.45731931924819946, "learning_rate": 4.384883214855651e-06, "loss": 0.0136, "step": 21627 }, { "epoch": 2.3447528187337383, "grad_norm": 0.18318477272987366, "learning_rate": 4.384157841288264e-06, "loss": 0.0119, "step": 21628 }, { "epoch": 2.344861231569818, "grad_norm": 0.4021977186203003, "learning_rate": 4.383432467720877e-06, "loss": 0.0146, "step": 21629 }, { "epoch": 2.344969644405898, "grad_norm": 0.8820748329162598, "learning_rate": 4.382707094153489e-06, "loss": 0.0023, "step": 21630 }, { "epoch": 2.3450780572419774, "grad_norm": 0.39622190594673157, "learning_rate": 4.381981720586102e-06, "loss": 0.0122, "step": 21631 }, { "epoch": 2.3451864700780574, "grad_norm": 6.779007434844971, "learning_rate": 4.381256347018715e-06, "loss": 0.0242, "step": 21632 }, { "epoch": 2.345294882914137, "grad_norm": 0.047186996787786484, "learning_rate": 4.380530973451328e-06, "loss": 0.0002, "step": 21633 }, { "epoch": 2.345403295750217, "grad_norm": 0.37580662965774536, "learning_rate": 4.37980559988394e-06, "loss": 0.0141, "step": 21634 }, { "epoch": 2.3455117085862964, "grad_norm": 13.115087509155273, "learning_rate": 4.379080226316553e-06, "loss": 0.0235, "step": 21635 }, { "epoch": 2.3456201214223764, "grad_norm": 0.1225990504026413, "learning_rate": 4.378354852749166e-06, "loss": 0.0004, "step": 21636 }, { "epoch": 2.3457285342584564, "grad_norm": 0.4174521863460541, "learning_rate": 4.377629479181779e-06, "loss": 0.0153, "step": 21637 }, { "epoch": 2.345836947094536, "grad_norm": 0.6123141050338745, "learning_rate": 4.3769041056143915e-06, "loss": 0.0194, "step": 21638 }, { "epoch": 2.345945359930616, "grad_norm": 0.04129338636994362, "learning_rate": 4.3761787320470045e-06, "loss": 0.0002, "step": 21639 }, { "epoch": 2.3460537727666955, "grad_norm": 0.3630930781364441, "learning_rate": 4.3754533584796175e-06, "loss": 0.0077, "step": 21640 }, { "epoch": 2.3461621856027755, "grad_norm": 0.31944748759269714, "learning_rate": 4.3747279849122304e-06, "loss": 0.0078, "step": 21641 }, { "epoch": 2.346270598438855, "grad_norm": 0.03062070719897747, "learning_rate": 4.3740026113448434e-06, "loss": 0.0002, "step": 21642 }, { "epoch": 2.346379011274935, "grad_norm": 1.292138695716858, "learning_rate": 4.3732772377774556e-06, "loss": 0.0614, "step": 21643 }, { "epoch": 2.3464874241110145, "grad_norm": 0.39118844270706177, "learning_rate": 4.3725518642100686e-06, "loss": 0.0009, "step": 21644 }, { "epoch": 2.3465958369470945, "grad_norm": 0.38253241777420044, "learning_rate": 4.3718264906426816e-06, "loss": 0.0083, "step": 21645 }, { "epoch": 2.3467042497831745, "grad_norm": 7.327284336090088, "learning_rate": 4.3711011170752945e-06, "loss": 0.0491, "step": 21646 }, { "epoch": 2.346812662619254, "grad_norm": 0.2814166247844696, "learning_rate": 4.370375743507907e-06, "loss": 0.0058, "step": 21647 }, { "epoch": 2.346921075455334, "grad_norm": 1.880455493927002, "learning_rate": 4.36965036994052e-06, "loss": 0.0242, "step": 21648 }, { "epoch": 2.3470294882914136, "grad_norm": 10.928449630737305, "learning_rate": 4.368924996373133e-06, "loss": 0.0123, "step": 21649 }, { "epoch": 2.3471379011274935, "grad_norm": 2.9309499263763428, "learning_rate": 4.368199622805746e-06, "loss": 0.0275, "step": 21650 }, { "epoch": 2.3472463139635735, "grad_norm": 0.26026976108551025, "learning_rate": 4.367474249238358e-06, "loss": 0.0056, "step": 21651 }, { "epoch": 2.347354726799653, "grad_norm": 0.030457448214292526, "learning_rate": 4.366748875670971e-06, "loss": 0.0002, "step": 21652 }, { "epoch": 2.347463139635733, "grad_norm": 0.007796773687005043, "learning_rate": 4.366023502103584e-06, "loss": 0.0001, "step": 21653 }, { "epoch": 2.3475715524718126, "grad_norm": 2.521594285964966, "learning_rate": 4.365298128536197e-06, "loss": 0.0042, "step": 21654 }, { "epoch": 2.3476799653078926, "grad_norm": 0.11074406653642654, "learning_rate": 4.36457275496881e-06, "loss": 0.0005, "step": 21655 }, { "epoch": 2.347788378143972, "grad_norm": 20.417240142822266, "learning_rate": 4.363847381401422e-06, "loss": 0.031, "step": 21656 }, { "epoch": 2.347896790980052, "grad_norm": 0.756392240524292, "learning_rate": 4.363122007834035e-06, "loss": 0.0152, "step": 21657 }, { "epoch": 2.3480052038161316, "grad_norm": 1.0093580484390259, "learning_rate": 4.362396634266648e-06, "loss": 0.0165, "step": 21658 }, { "epoch": 2.3481136166522116, "grad_norm": 0.6845592260360718, "learning_rate": 4.361671260699261e-06, "loss": 0.0082, "step": 21659 }, { "epoch": 2.3482220294882916, "grad_norm": 8.688634872436523, "learning_rate": 4.360945887131873e-06, "loss": 0.0191, "step": 21660 }, { "epoch": 2.348330442324371, "grad_norm": 2.24920654296875, "learning_rate": 4.360220513564486e-06, "loss": 0.0142, "step": 21661 }, { "epoch": 2.348438855160451, "grad_norm": 1.1890454292297363, "learning_rate": 4.359495139997099e-06, "loss": 0.0344, "step": 21662 }, { "epoch": 2.3485472679965307, "grad_norm": 0.3188456892967224, "learning_rate": 4.358769766429712e-06, "loss": 0.0138, "step": 21663 }, { "epoch": 2.3486556808326107, "grad_norm": 0.8583667874336243, "learning_rate": 4.358044392862324e-06, "loss": 0.0221, "step": 21664 }, { "epoch": 2.34876409366869, "grad_norm": 0.36978602409362793, "learning_rate": 4.357319019294937e-06, "loss": 0.0005, "step": 21665 }, { "epoch": 2.34887250650477, "grad_norm": 0.4112319052219391, "learning_rate": 4.35659364572755e-06, "loss": 0.0086, "step": 21666 }, { "epoch": 2.3489809193408497, "grad_norm": 0.29283374547958374, "learning_rate": 4.355868272160163e-06, "loss": 0.01, "step": 21667 }, { "epoch": 2.3490893321769297, "grad_norm": 3.072643518447876, "learning_rate": 4.355142898592776e-06, "loss": 0.02, "step": 21668 }, { "epoch": 2.3491977450130097, "grad_norm": 0.22817392647266388, "learning_rate": 4.354417525025388e-06, "loss": 0.003, "step": 21669 }, { "epoch": 2.3493061578490892, "grad_norm": 0.2652459144592285, "learning_rate": 4.353692151458001e-06, "loss": 0.0058, "step": 21670 }, { "epoch": 2.3494145706851692, "grad_norm": 0.39853963255882263, "learning_rate": 4.352966777890614e-06, "loss": 0.0199, "step": 21671 }, { "epoch": 2.3495229835212488, "grad_norm": 0.18082523345947266, "learning_rate": 4.352241404323227e-06, "loss": 0.0008, "step": 21672 }, { "epoch": 2.3496313963573288, "grad_norm": 0.25596287846565247, "learning_rate": 4.351516030755839e-06, "loss": 0.0052, "step": 21673 }, { "epoch": 2.3497398091934087, "grad_norm": 0.04781431704759598, "learning_rate": 4.350790657188452e-06, "loss": 0.0002, "step": 21674 }, { "epoch": 2.3498482220294883, "grad_norm": 0.02929656021296978, "learning_rate": 4.350065283621065e-06, "loss": 0.0002, "step": 21675 }, { "epoch": 2.3499566348655683, "grad_norm": 0.08657565712928772, "learning_rate": 4.349339910053678e-06, "loss": 0.0006, "step": 21676 }, { "epoch": 2.350065047701648, "grad_norm": 0.04118872433900833, "learning_rate": 4.3486145364862905e-06, "loss": 0.0005, "step": 21677 }, { "epoch": 2.350173460537728, "grad_norm": 0.608988881111145, "learning_rate": 4.3478891629189035e-06, "loss": 0.0158, "step": 21678 }, { "epoch": 2.3502818733738073, "grad_norm": 0.04345173016190529, "learning_rate": 4.3471637893515165e-06, "loss": 0.0003, "step": 21679 }, { "epoch": 2.3503902862098873, "grad_norm": 2.662888288497925, "learning_rate": 4.3464384157841294e-06, "loss": 0.0075, "step": 21680 }, { "epoch": 2.350498699045967, "grad_norm": 3.9186179637908936, "learning_rate": 4.3457130422167424e-06, "loss": 0.034, "step": 21681 }, { "epoch": 2.350607111882047, "grad_norm": 0.23033857345581055, "learning_rate": 4.344987668649355e-06, "loss": 0.0037, "step": 21682 }, { "epoch": 2.350715524718127, "grad_norm": 2.8237030506134033, "learning_rate": 4.3442622950819676e-06, "loss": 0.0399, "step": 21683 }, { "epoch": 2.3508239375542064, "grad_norm": 0.655120849609375, "learning_rate": 4.3435369215145806e-06, "loss": 0.0299, "step": 21684 }, { "epoch": 2.3509323503902864, "grad_norm": 0.019745368510484695, "learning_rate": 4.3428115479471935e-06, "loss": 0.0002, "step": 21685 }, { "epoch": 2.351040763226366, "grad_norm": 0.026574529707431793, "learning_rate": 4.342086174379806e-06, "loss": 0.0002, "step": 21686 }, { "epoch": 2.351149176062446, "grad_norm": 0.2988024950027466, "learning_rate": 4.341360800812419e-06, "loss": 0.0109, "step": 21687 }, { "epoch": 2.3512575888985254, "grad_norm": 1.5056395530700684, "learning_rate": 4.340635427245032e-06, "loss": 0.0172, "step": 21688 }, { "epoch": 2.3513660017346054, "grad_norm": 0.1937456727027893, "learning_rate": 4.339910053677645e-06, "loss": 0.0012, "step": 21689 }, { "epoch": 2.351474414570685, "grad_norm": 0.3104315400123596, "learning_rate": 4.339184680110257e-06, "loss": 0.0106, "step": 21690 }, { "epoch": 2.351582827406765, "grad_norm": 0.6314812898635864, "learning_rate": 4.33845930654287e-06, "loss": 0.0037, "step": 21691 }, { "epoch": 2.351691240242845, "grad_norm": 0.3396681547164917, "learning_rate": 4.337733932975483e-06, "loss": 0.0131, "step": 21692 }, { "epoch": 2.3517996530789245, "grad_norm": 9.640142440795898, "learning_rate": 4.337008559408096e-06, "loss": 0.0121, "step": 21693 }, { "epoch": 2.3519080659150045, "grad_norm": 2.3196091651916504, "learning_rate": 4.336283185840709e-06, "loss": 0.0179, "step": 21694 }, { "epoch": 2.352016478751084, "grad_norm": 0.02856661193072796, "learning_rate": 4.335557812273321e-06, "loss": 0.0002, "step": 21695 }, { "epoch": 2.352124891587164, "grad_norm": 0.8565413951873779, "learning_rate": 4.334832438705934e-06, "loss": 0.0071, "step": 21696 }, { "epoch": 2.3522333044232435, "grad_norm": 3.7912545204162598, "learning_rate": 4.334107065138547e-06, "loss": 0.0439, "step": 21697 }, { "epoch": 2.3523417172593235, "grad_norm": 3.6167075634002686, "learning_rate": 4.33338169157116e-06, "loss": 0.0637, "step": 21698 }, { "epoch": 2.352450130095403, "grad_norm": 1.255942702293396, "learning_rate": 4.332656318003772e-06, "loss": 0.0479, "step": 21699 }, { "epoch": 2.352558542931483, "grad_norm": 2.7886123657226562, "learning_rate": 4.331930944436385e-06, "loss": 0.0256, "step": 21700 }, { "epoch": 2.352666955767563, "grad_norm": 0.10005686432123184, "learning_rate": 4.331205570868998e-06, "loss": 0.0012, "step": 21701 }, { "epoch": 2.3527753686036426, "grad_norm": 0.038052212446928024, "learning_rate": 4.330480197301611e-06, "loss": 0.0003, "step": 21702 }, { "epoch": 2.3528837814397225, "grad_norm": 0.07541245222091675, "learning_rate": 4.329754823734223e-06, "loss": 0.0005, "step": 21703 }, { "epoch": 2.352992194275802, "grad_norm": 0.05898277834057808, "learning_rate": 4.329029450166836e-06, "loss": 0.0007, "step": 21704 }, { "epoch": 2.353100607111882, "grad_norm": 0.06963599473237991, "learning_rate": 4.328304076599449e-06, "loss": 0.0003, "step": 21705 }, { "epoch": 2.353209019947962, "grad_norm": 0.2724533975124359, "learning_rate": 4.327578703032062e-06, "loss": 0.0063, "step": 21706 }, { "epoch": 2.3533174327840416, "grad_norm": 0.5170632004737854, "learning_rate": 4.326853329464675e-06, "loss": 0.0009, "step": 21707 }, { "epoch": 2.3534258456201216, "grad_norm": 0.8217616677284241, "learning_rate": 4.326127955897287e-06, "loss": 0.0141, "step": 21708 }, { "epoch": 2.353534258456201, "grad_norm": 1.3301409482955933, "learning_rate": 4.3254025823299e-06, "loss": 0.0347, "step": 21709 }, { "epoch": 2.353642671292281, "grad_norm": 0.4897925555706024, "learning_rate": 4.324677208762513e-06, "loss": 0.0139, "step": 21710 }, { "epoch": 2.3537510841283606, "grad_norm": 0.08215504884719849, "learning_rate": 4.323951835195126e-06, "loss": 0.0004, "step": 21711 }, { "epoch": 2.3538594969644406, "grad_norm": 0.5706449747085571, "learning_rate": 4.323226461627738e-06, "loss": 0.0013, "step": 21712 }, { "epoch": 2.35396790980052, "grad_norm": 0.28066661953926086, "learning_rate": 4.322501088060351e-06, "loss": 0.014, "step": 21713 }, { "epoch": 2.3540763226366, "grad_norm": 3.6078038215637207, "learning_rate": 4.321775714492964e-06, "loss": 0.0099, "step": 21714 }, { "epoch": 2.35418473547268, "grad_norm": 0.1784188598394394, "learning_rate": 4.321050340925577e-06, "loss": 0.0006, "step": 21715 }, { "epoch": 2.3542931483087597, "grad_norm": 0.5668298006057739, "learning_rate": 4.3203249673581895e-06, "loss": 0.0079, "step": 21716 }, { "epoch": 2.3544015611448397, "grad_norm": 0.7094850540161133, "learning_rate": 4.3195995937908025e-06, "loss": 0.025, "step": 21717 }, { "epoch": 2.354509973980919, "grad_norm": 0.25604262948036194, "learning_rate": 4.3188742202234155e-06, "loss": 0.0051, "step": 21718 }, { "epoch": 2.354618386816999, "grad_norm": 0.1416490525007248, "learning_rate": 4.3181488466560284e-06, "loss": 0.0006, "step": 21719 }, { "epoch": 2.3547267996530787, "grad_norm": 0.29315897822380066, "learning_rate": 4.3174234730886414e-06, "loss": 0.0056, "step": 21720 }, { "epoch": 2.3548352124891587, "grad_norm": 0.2435895800590515, "learning_rate": 4.316698099521254e-06, "loss": 0.0013, "step": 21721 }, { "epoch": 2.3549436253252383, "grad_norm": 1.722508430480957, "learning_rate": 4.3159727259538666e-06, "loss": 0.0392, "step": 21722 }, { "epoch": 2.3550520381613183, "grad_norm": 1.9147762060165405, "learning_rate": 4.3152473523864796e-06, "loss": 0.0014, "step": 21723 }, { "epoch": 2.3551604509973982, "grad_norm": 2.8460144996643066, "learning_rate": 4.3145219788190926e-06, "loss": 0.0357, "step": 21724 }, { "epoch": 2.3552688638334778, "grad_norm": 8.621965408325195, "learning_rate": 4.313796605251705e-06, "loss": 0.007, "step": 21725 }, { "epoch": 2.3553772766695578, "grad_norm": 0.5239985585212708, "learning_rate": 4.313071231684318e-06, "loss": 0.0011, "step": 21726 }, { "epoch": 2.3554856895056373, "grad_norm": 0.30953413248062134, "learning_rate": 4.312345858116931e-06, "loss": 0.0008, "step": 21727 }, { "epoch": 2.3555941023417173, "grad_norm": 8.446361541748047, "learning_rate": 4.311620484549544e-06, "loss": 0.019, "step": 21728 }, { "epoch": 2.3557025151777973, "grad_norm": 5.88297700881958, "learning_rate": 4.310895110982156e-06, "loss": 0.048, "step": 21729 }, { "epoch": 2.355810928013877, "grad_norm": 0.3334820866584778, "learning_rate": 4.310169737414769e-06, "loss": 0.0096, "step": 21730 }, { "epoch": 2.355919340849957, "grad_norm": 3.997487783432007, "learning_rate": 4.309444363847382e-06, "loss": 0.0227, "step": 21731 }, { "epoch": 2.3560277536860363, "grad_norm": 0.6415311098098755, "learning_rate": 4.308718990279995e-06, "loss": 0.0074, "step": 21732 }, { "epoch": 2.3561361665221163, "grad_norm": 1.5403681993484497, "learning_rate": 4.307993616712608e-06, "loss": 0.031, "step": 21733 }, { "epoch": 2.356244579358196, "grad_norm": 0.10682515799999237, "learning_rate": 4.30726824314522e-06, "loss": 0.0004, "step": 21734 }, { "epoch": 2.356352992194276, "grad_norm": 0.14452145993709564, "learning_rate": 4.306542869577833e-06, "loss": 0.0008, "step": 21735 }, { "epoch": 2.3564614050303554, "grad_norm": 2.7437970638275146, "learning_rate": 4.305817496010446e-06, "loss": 0.0125, "step": 21736 }, { "epoch": 2.3565698178664354, "grad_norm": 0.3028256595134735, "learning_rate": 4.305092122443059e-06, "loss": 0.0045, "step": 21737 }, { "epoch": 2.3566782307025154, "grad_norm": 0.3299317955970764, "learning_rate": 4.304366748875671e-06, "loss": 0.0012, "step": 21738 }, { "epoch": 2.356786643538595, "grad_norm": 0.36225375533103943, "learning_rate": 4.303641375308284e-06, "loss": 0.0024, "step": 21739 }, { "epoch": 2.356895056374675, "grad_norm": 4.058661937713623, "learning_rate": 4.302916001740897e-06, "loss": 0.0035, "step": 21740 }, { "epoch": 2.3570034692107544, "grad_norm": 0.49367624521255493, "learning_rate": 4.30219062817351e-06, "loss": 0.0061, "step": 21741 }, { "epoch": 2.3571118820468344, "grad_norm": 0.26894450187683105, "learning_rate": 4.301465254606122e-06, "loss": 0.0043, "step": 21742 }, { "epoch": 2.357220294882914, "grad_norm": 0.39877554774284363, "learning_rate": 4.300739881038735e-06, "loss": 0.0133, "step": 21743 }, { "epoch": 2.357328707718994, "grad_norm": 0.013286935165524483, "learning_rate": 4.300014507471348e-06, "loss": 0.0001, "step": 21744 }, { "epoch": 2.3574371205550735, "grad_norm": 1.5763952732086182, "learning_rate": 4.299289133903961e-06, "loss": 0.0423, "step": 21745 }, { "epoch": 2.3575455333911535, "grad_norm": 4.378077507019043, "learning_rate": 4.298563760336574e-06, "loss": 0.0283, "step": 21746 }, { "epoch": 2.3576539462272335, "grad_norm": 0.2624047100543976, "learning_rate": 4.297838386769186e-06, "loss": 0.003, "step": 21747 }, { "epoch": 2.357762359063313, "grad_norm": 1.6155599355697632, "learning_rate": 4.297113013201799e-06, "loss": 0.0637, "step": 21748 }, { "epoch": 2.357870771899393, "grad_norm": 0.7689622044563293, "learning_rate": 4.296387639634412e-06, "loss": 0.0194, "step": 21749 }, { "epoch": 2.3579791847354725, "grad_norm": 0.7792785167694092, "learning_rate": 4.295662266067025e-06, "loss": 0.0158, "step": 21750 }, { "epoch": 2.3580875975715525, "grad_norm": 0.3705865144729614, "learning_rate": 4.294936892499637e-06, "loss": 0.0118, "step": 21751 }, { "epoch": 2.3581960104076325, "grad_norm": 0.2592123746871948, "learning_rate": 4.29421151893225e-06, "loss": 0.0057, "step": 21752 }, { "epoch": 2.358304423243712, "grad_norm": 5.254272937774658, "learning_rate": 4.293486145364863e-06, "loss": 0.0238, "step": 21753 }, { "epoch": 2.358412836079792, "grad_norm": 0.08907636255025864, "learning_rate": 4.292760771797476e-06, "loss": 0.0007, "step": 21754 }, { "epoch": 2.3585212489158716, "grad_norm": 1.0819170475006104, "learning_rate": 4.2920353982300885e-06, "loss": 0.0228, "step": 21755 }, { "epoch": 2.3586296617519515, "grad_norm": 0.22688093781471252, "learning_rate": 4.2913100246627015e-06, "loss": 0.0006, "step": 21756 }, { "epoch": 2.358738074588031, "grad_norm": 3.1758670806884766, "learning_rate": 4.2905846510953145e-06, "loss": 0.0353, "step": 21757 }, { "epoch": 2.358846487424111, "grad_norm": 0.5239842534065247, "learning_rate": 4.2898592775279275e-06, "loss": 0.0101, "step": 21758 }, { "epoch": 2.3589549002601906, "grad_norm": 0.7833097577095032, "learning_rate": 4.2891339039605404e-06, "loss": 0.0196, "step": 21759 }, { "epoch": 2.3590633130962706, "grad_norm": 0.36591729521751404, "learning_rate": 4.288408530393153e-06, "loss": 0.0113, "step": 21760 }, { "epoch": 2.3591717259323506, "grad_norm": 0.00661002891138196, "learning_rate": 4.2876831568257656e-06, "loss": 0.0001, "step": 21761 }, { "epoch": 2.35928013876843, "grad_norm": 0.5422443151473999, "learning_rate": 4.2869577832583786e-06, "loss": 0.012, "step": 21762 }, { "epoch": 2.35938855160451, "grad_norm": 0.026424551382660866, "learning_rate": 4.2862324096909916e-06, "loss": 0.0002, "step": 21763 }, { "epoch": 2.3594969644405897, "grad_norm": 12.77332878112793, "learning_rate": 4.285507036123604e-06, "loss": 0.0281, "step": 21764 }, { "epoch": 2.3596053772766696, "grad_norm": 0.3492121994495392, "learning_rate": 4.284781662556217e-06, "loss": 0.0174, "step": 21765 }, { "epoch": 2.359713790112749, "grad_norm": 0.034534502774477005, "learning_rate": 4.28405628898883e-06, "loss": 0.0003, "step": 21766 }, { "epoch": 2.359822202948829, "grad_norm": 8.24135971069336, "learning_rate": 4.283330915421443e-06, "loss": 0.0285, "step": 21767 }, { "epoch": 2.3599306157849087, "grad_norm": 24.7541446685791, "learning_rate": 4.282605541854055e-06, "loss": 0.0298, "step": 21768 }, { "epoch": 2.3600390286209887, "grad_norm": 0.35613375902175903, "learning_rate": 4.281880168286668e-06, "loss": 0.0025, "step": 21769 }, { "epoch": 2.3601474414570687, "grad_norm": 0.39023759961128235, "learning_rate": 4.281154794719281e-06, "loss": 0.0069, "step": 21770 }, { "epoch": 2.360255854293148, "grad_norm": 0.5106133818626404, "learning_rate": 4.280429421151894e-06, "loss": 0.0019, "step": 21771 }, { "epoch": 2.360364267129228, "grad_norm": 1.3802549839019775, "learning_rate": 4.279704047584507e-06, "loss": 0.0152, "step": 21772 }, { "epoch": 2.3604726799653077, "grad_norm": 2.2587177753448486, "learning_rate": 4.278978674017119e-06, "loss": 0.0176, "step": 21773 }, { "epoch": 2.3605810928013877, "grad_norm": 0.026227271184325218, "learning_rate": 4.278253300449732e-06, "loss": 0.0002, "step": 21774 }, { "epoch": 2.3606895056374677, "grad_norm": 0.06996733695268631, "learning_rate": 4.277527926882345e-06, "loss": 0.0002, "step": 21775 }, { "epoch": 2.3607979184735473, "grad_norm": 1.0732470750808716, "learning_rate": 4.276802553314958e-06, "loss": 0.064, "step": 21776 }, { "epoch": 2.3609063313096272, "grad_norm": 0.30038565397262573, "learning_rate": 4.27607717974757e-06, "loss": 0.0066, "step": 21777 }, { "epoch": 2.361014744145707, "grad_norm": 0.33868539333343506, "learning_rate": 4.275351806180183e-06, "loss": 0.0098, "step": 21778 }, { "epoch": 2.3611231569817868, "grad_norm": 0.12577572464942932, "learning_rate": 4.274626432612796e-06, "loss": 0.0011, "step": 21779 }, { "epoch": 2.3612315698178663, "grad_norm": 0.6044354438781738, "learning_rate": 4.273901059045409e-06, "loss": 0.0314, "step": 21780 }, { "epoch": 2.3613399826539463, "grad_norm": 0.5575655698776245, "learning_rate": 4.273175685478021e-06, "loss": 0.0181, "step": 21781 }, { "epoch": 2.361448395490026, "grad_norm": 1.5507938861846924, "learning_rate": 4.272450311910634e-06, "loss": 0.0156, "step": 21782 }, { "epoch": 2.361556808326106, "grad_norm": 0.4660388231277466, "learning_rate": 4.271724938343247e-06, "loss": 0.0103, "step": 21783 }, { "epoch": 2.361665221162186, "grad_norm": 0.5251153707504272, "learning_rate": 4.27099956477586e-06, "loss": 0.0091, "step": 21784 }, { "epoch": 2.3617736339982653, "grad_norm": 0.32097116112709045, "learning_rate": 4.270274191208473e-06, "loss": 0.0019, "step": 21785 }, { "epoch": 2.3618820468343453, "grad_norm": 0.08077023923397064, "learning_rate": 4.269548817641085e-06, "loss": 0.0005, "step": 21786 }, { "epoch": 2.361990459670425, "grad_norm": 0.8206114172935486, "learning_rate": 4.268823444073698e-06, "loss": 0.0149, "step": 21787 }, { "epoch": 2.362098872506505, "grad_norm": 0.08890638500452042, "learning_rate": 4.268098070506311e-06, "loss": 0.0006, "step": 21788 }, { "epoch": 2.3622072853425844, "grad_norm": 0.015715371817350388, "learning_rate": 4.267372696938924e-06, "loss": 0.0001, "step": 21789 }, { "epoch": 2.3623156981786644, "grad_norm": 4.3235039710998535, "learning_rate": 4.266647323371536e-06, "loss": 0.03, "step": 21790 }, { "epoch": 2.362424111014744, "grad_norm": 0.015963923186063766, "learning_rate": 4.265921949804149e-06, "loss": 0.0002, "step": 21791 }, { "epoch": 2.362532523850824, "grad_norm": 0.10408841073513031, "learning_rate": 4.265196576236762e-06, "loss": 0.0006, "step": 21792 }, { "epoch": 2.362640936686904, "grad_norm": 0.014085201546549797, "learning_rate": 4.264471202669375e-06, "loss": 0.0001, "step": 21793 }, { "epoch": 2.3627493495229834, "grad_norm": 0.0600346103310585, "learning_rate": 4.2637458291019875e-06, "loss": 0.0004, "step": 21794 }, { "epoch": 2.3628577623590634, "grad_norm": 0.6118005514144897, "learning_rate": 4.2630204555346005e-06, "loss": 0.0045, "step": 21795 }, { "epoch": 2.362966175195143, "grad_norm": 4.0981645584106445, "learning_rate": 4.2622950819672135e-06, "loss": 0.0292, "step": 21796 }, { "epoch": 2.363074588031223, "grad_norm": 0.19403591752052307, "learning_rate": 4.2615697083998265e-06, "loss": 0.0103, "step": 21797 }, { "epoch": 2.363183000867303, "grad_norm": 0.5016394853591919, "learning_rate": 4.2608443348324394e-06, "loss": 0.0041, "step": 21798 }, { "epoch": 2.3632914137033825, "grad_norm": 2.322113037109375, "learning_rate": 4.260118961265052e-06, "loss": 0.0033, "step": 21799 }, { "epoch": 2.3633998265394625, "grad_norm": 0.15954455733299255, "learning_rate": 4.259393587697665e-06, "loss": 0.0011, "step": 21800 }, { "epoch": 2.363508239375542, "grad_norm": 0.026057064533233643, "learning_rate": 4.2586682141302776e-06, "loss": 0.0001, "step": 21801 }, { "epoch": 2.363616652211622, "grad_norm": 0.6548587083816528, "learning_rate": 4.2579428405628906e-06, "loss": 0.024, "step": 21802 }, { "epoch": 2.3637250650477015, "grad_norm": 0.7182145118713379, "learning_rate": 4.257217466995503e-06, "loss": 0.0209, "step": 21803 }, { "epoch": 2.3638334778837815, "grad_norm": 0.284214586019516, "learning_rate": 4.256492093428116e-06, "loss": 0.0061, "step": 21804 }, { "epoch": 2.363941890719861, "grad_norm": 2.2486908435821533, "learning_rate": 4.255766719860729e-06, "loss": 0.0098, "step": 21805 }, { "epoch": 2.364050303555941, "grad_norm": 0.025523865595459938, "learning_rate": 4.255041346293342e-06, "loss": 0.0002, "step": 21806 }, { "epoch": 2.364158716392021, "grad_norm": 1.4216045141220093, "learning_rate": 4.254315972725954e-06, "loss": 0.0108, "step": 21807 }, { "epoch": 2.3642671292281006, "grad_norm": 6.4292216300964355, "learning_rate": 4.253590599158567e-06, "loss": 0.0094, "step": 21808 }, { "epoch": 2.3643755420641805, "grad_norm": 0.4295961558818817, "learning_rate": 4.25286522559118e-06, "loss": 0.0087, "step": 21809 }, { "epoch": 2.36448395490026, "grad_norm": 0.05146860331296921, "learning_rate": 4.252139852023793e-06, "loss": 0.0004, "step": 21810 }, { "epoch": 2.36459236773634, "grad_norm": 1.2600570917129517, "learning_rate": 4.251414478456406e-06, "loss": 0.0018, "step": 21811 }, { "epoch": 2.3647007805724196, "grad_norm": 0.030951526015996933, "learning_rate": 4.250689104889018e-06, "loss": 0.0003, "step": 21812 }, { "epoch": 2.3648091934084996, "grad_norm": 0.2384272962808609, "learning_rate": 4.249963731321631e-06, "loss": 0.0048, "step": 21813 }, { "epoch": 2.364917606244579, "grad_norm": 0.3728167712688446, "learning_rate": 4.249238357754244e-06, "loss": 0.0128, "step": 21814 }, { "epoch": 2.365026019080659, "grad_norm": 1.0859057903289795, "learning_rate": 4.248512984186857e-06, "loss": 0.0037, "step": 21815 }, { "epoch": 2.365134431916739, "grad_norm": 0.31961727142333984, "learning_rate": 4.247787610619469e-06, "loss": 0.0066, "step": 21816 }, { "epoch": 2.3652428447528187, "grad_norm": 7.54025411605835, "learning_rate": 4.247062237052082e-06, "loss": 0.0528, "step": 21817 }, { "epoch": 2.3653512575888986, "grad_norm": 5.48227071762085, "learning_rate": 4.246336863484695e-06, "loss": 0.0204, "step": 21818 }, { "epoch": 2.365459670424978, "grad_norm": 7.392144203186035, "learning_rate": 4.245611489917308e-06, "loss": 0.015, "step": 21819 }, { "epoch": 2.365568083261058, "grad_norm": 0.21207523345947266, "learning_rate": 4.24488611634992e-06, "loss": 0.0099, "step": 21820 }, { "epoch": 2.365676496097138, "grad_norm": 3.0685458183288574, "learning_rate": 4.244160742782533e-06, "loss": 0.0282, "step": 21821 }, { "epoch": 2.3657849089332177, "grad_norm": 0.36603212356567383, "learning_rate": 4.243435369215146e-06, "loss": 0.01, "step": 21822 }, { "epoch": 2.3658933217692977, "grad_norm": 1.6792863607406616, "learning_rate": 4.242709995647759e-06, "loss": 0.0177, "step": 21823 }, { "epoch": 2.366001734605377, "grad_norm": 1.5689724683761597, "learning_rate": 4.241984622080372e-06, "loss": 0.0099, "step": 21824 }, { "epoch": 2.366110147441457, "grad_norm": 0.06060651317238808, "learning_rate": 4.241259248512984e-06, "loss": 0.0005, "step": 21825 }, { "epoch": 2.3662185602775367, "grad_norm": 0.7109883427619934, "learning_rate": 4.240533874945597e-06, "loss": 0.0114, "step": 21826 }, { "epoch": 2.3663269731136167, "grad_norm": 0.30773696303367615, "learning_rate": 4.23980850137821e-06, "loss": 0.0115, "step": 21827 }, { "epoch": 2.3664353859496963, "grad_norm": 2.0353615283966064, "learning_rate": 4.239083127810823e-06, "loss": 0.006, "step": 21828 }, { "epoch": 2.3665437987857763, "grad_norm": 0.5133705735206604, "learning_rate": 4.238357754243435e-06, "loss": 0.0249, "step": 21829 }, { "epoch": 2.3666522116218562, "grad_norm": 0.400684118270874, "learning_rate": 4.237632380676048e-06, "loss": 0.0129, "step": 21830 }, { "epoch": 2.366760624457936, "grad_norm": 0.20290282368659973, "learning_rate": 4.236907007108661e-06, "loss": 0.0038, "step": 21831 }, { "epoch": 2.3668690372940158, "grad_norm": 0.3070948123931885, "learning_rate": 4.236181633541274e-06, "loss": 0.0067, "step": 21832 }, { "epoch": 2.3669774501300953, "grad_norm": 0.41157108545303345, "learning_rate": 4.2354562599738865e-06, "loss": 0.0129, "step": 21833 }, { "epoch": 2.3670858629661753, "grad_norm": 0.33943915367126465, "learning_rate": 4.2347308864064995e-06, "loss": 0.0179, "step": 21834 }, { "epoch": 2.367194275802255, "grad_norm": 1.1690067052841187, "learning_rate": 4.2340055128391125e-06, "loss": 0.0127, "step": 21835 }, { "epoch": 2.367302688638335, "grad_norm": 1.697861671447754, "learning_rate": 4.2332801392717255e-06, "loss": 0.0027, "step": 21836 }, { "epoch": 2.3674111014744144, "grad_norm": 0.3201459050178528, "learning_rate": 4.2325547657043384e-06, "loss": 0.0079, "step": 21837 }, { "epoch": 2.3675195143104943, "grad_norm": 0.2473313957452774, "learning_rate": 4.231829392136951e-06, "loss": 0.0149, "step": 21838 }, { "epoch": 2.3676279271465743, "grad_norm": 0.30359768867492676, "learning_rate": 4.231104018569564e-06, "loss": 0.0092, "step": 21839 }, { "epoch": 2.367736339982654, "grad_norm": 0.9016638994216919, "learning_rate": 4.2303786450021766e-06, "loss": 0.048, "step": 21840 }, { "epoch": 2.367844752818734, "grad_norm": 0.3607722818851471, "learning_rate": 4.2296532714347896e-06, "loss": 0.0128, "step": 21841 }, { "epoch": 2.3679531656548134, "grad_norm": 0.3236115872859955, "learning_rate": 4.228927897867402e-06, "loss": 0.0123, "step": 21842 }, { "epoch": 2.3680615784908934, "grad_norm": 0.12726372480392456, "learning_rate": 4.228202524300015e-06, "loss": 0.0005, "step": 21843 }, { "epoch": 2.368169991326973, "grad_norm": 0.3082689046859741, "learning_rate": 4.227477150732628e-06, "loss": 0.0056, "step": 21844 }, { "epoch": 2.368278404163053, "grad_norm": 0.4381304979324341, "learning_rate": 4.226751777165241e-06, "loss": 0.0087, "step": 21845 }, { "epoch": 2.3683868169991324, "grad_norm": 0.04541455954313278, "learning_rate": 4.226026403597853e-06, "loss": 0.0003, "step": 21846 }, { "epoch": 2.3684952298352124, "grad_norm": 0.27428993582725525, "learning_rate": 4.225301030030466e-06, "loss": 0.0049, "step": 21847 }, { "epoch": 2.3686036426712924, "grad_norm": 0.007182741072028875, "learning_rate": 4.224575656463079e-06, "loss": 0.0001, "step": 21848 }, { "epoch": 2.368712055507372, "grad_norm": 0.46761927008628845, "learning_rate": 4.223850282895692e-06, "loss": 0.0048, "step": 21849 }, { "epoch": 2.368820468343452, "grad_norm": 0.3383944034576416, "learning_rate": 4.223124909328305e-06, "loss": 0.01, "step": 21850 }, { "epoch": 2.3689288811795315, "grad_norm": 0.05690980330109596, "learning_rate": 4.222399535760917e-06, "loss": 0.0003, "step": 21851 }, { "epoch": 2.3690372940156115, "grad_norm": 0.024255039170384407, "learning_rate": 4.22167416219353e-06, "loss": 0.0001, "step": 21852 }, { "epoch": 2.3691457068516915, "grad_norm": 3.3542098999023438, "learning_rate": 4.220948788626143e-06, "loss": 0.0035, "step": 21853 }, { "epoch": 2.369254119687771, "grad_norm": 0.4632072448730469, "learning_rate": 4.220223415058756e-06, "loss": 0.0242, "step": 21854 }, { "epoch": 2.369362532523851, "grad_norm": 4.144337177276611, "learning_rate": 4.219498041491368e-06, "loss": 0.0147, "step": 21855 }, { "epoch": 2.3694709453599305, "grad_norm": 0.020721467211842537, "learning_rate": 4.218772667923981e-06, "loss": 0.0002, "step": 21856 }, { "epoch": 2.3695793581960105, "grad_norm": 0.3227936923503876, "learning_rate": 4.218047294356594e-06, "loss": 0.0006, "step": 21857 }, { "epoch": 2.36968777103209, "grad_norm": 0.03676936775445938, "learning_rate": 4.217321920789207e-06, "loss": 0.0003, "step": 21858 }, { "epoch": 2.36979618386817, "grad_norm": 0.33479300141334534, "learning_rate": 4.216596547221819e-06, "loss": 0.0123, "step": 21859 }, { "epoch": 2.3699045967042496, "grad_norm": 1.9829542636871338, "learning_rate": 4.215871173654432e-06, "loss": 0.0107, "step": 21860 }, { "epoch": 2.3700130095403296, "grad_norm": 3.1199135780334473, "learning_rate": 4.215145800087045e-06, "loss": 0.0421, "step": 21861 }, { "epoch": 2.3701214223764095, "grad_norm": 0.3510631024837494, "learning_rate": 4.214420426519658e-06, "loss": 0.0165, "step": 21862 }, { "epoch": 2.370229835212489, "grad_norm": 0.493073970079422, "learning_rate": 4.213695052952271e-06, "loss": 0.0204, "step": 21863 }, { "epoch": 2.370338248048569, "grad_norm": 0.2983129620552063, "learning_rate": 4.212969679384883e-06, "loss": 0.0061, "step": 21864 }, { "epoch": 2.3704466608846486, "grad_norm": 0.21569141745567322, "learning_rate": 4.212244305817496e-06, "loss": 0.0043, "step": 21865 }, { "epoch": 2.3705550737207286, "grad_norm": 0.07976055890321732, "learning_rate": 4.211518932250109e-06, "loss": 0.0005, "step": 21866 }, { "epoch": 2.370663486556808, "grad_norm": 0.08868443965911865, "learning_rate": 4.210793558682722e-06, "loss": 0.0006, "step": 21867 }, { "epoch": 2.370771899392888, "grad_norm": 1.371364712715149, "learning_rate": 4.210068185115334e-06, "loss": 0.0565, "step": 21868 }, { "epoch": 2.3708803122289677, "grad_norm": 0.09124567359685898, "learning_rate": 4.209342811547947e-06, "loss": 0.0004, "step": 21869 }, { "epoch": 2.3709887250650477, "grad_norm": 5.902400016784668, "learning_rate": 4.20861743798056e-06, "loss": 0.0132, "step": 21870 }, { "epoch": 2.3710971379011276, "grad_norm": 0.06438999623060226, "learning_rate": 4.207892064413173e-06, "loss": 0.0003, "step": 21871 }, { "epoch": 2.371205550737207, "grad_norm": 0.06651973724365234, "learning_rate": 4.2071666908457855e-06, "loss": 0.0005, "step": 21872 }, { "epoch": 2.371313963573287, "grad_norm": 0.06465131044387817, "learning_rate": 4.2064413172783985e-06, "loss": 0.0005, "step": 21873 }, { "epoch": 2.3714223764093667, "grad_norm": 0.5738983154296875, "learning_rate": 4.2057159437110115e-06, "loss": 0.0147, "step": 21874 }, { "epoch": 2.3715307892454467, "grad_norm": 1.0071191787719727, "learning_rate": 4.2049905701436245e-06, "loss": 0.0109, "step": 21875 }, { "epoch": 2.3716392020815267, "grad_norm": 0.6082918643951416, "learning_rate": 4.2042651965762375e-06, "loss": 0.0245, "step": 21876 }, { "epoch": 2.371747614917606, "grad_norm": 0.6485486626625061, "learning_rate": 4.20353982300885e-06, "loss": 0.0153, "step": 21877 }, { "epoch": 2.371856027753686, "grad_norm": 0.3841436803340912, "learning_rate": 4.202814449441463e-06, "loss": 0.0143, "step": 21878 }, { "epoch": 2.3719644405897657, "grad_norm": 1.869713306427002, "learning_rate": 4.2020890758740756e-06, "loss": 0.0076, "step": 21879 }, { "epoch": 2.3720728534258457, "grad_norm": 0.8339244723320007, "learning_rate": 4.2013637023066886e-06, "loss": 0.0016, "step": 21880 }, { "epoch": 2.3721812662619253, "grad_norm": 0.04204796627163887, "learning_rate": 4.200638328739301e-06, "loss": 0.0003, "step": 21881 }, { "epoch": 2.3722896790980053, "grad_norm": 0.23841902613639832, "learning_rate": 4.199912955171914e-06, "loss": 0.0048, "step": 21882 }, { "epoch": 2.372398091934085, "grad_norm": 8.175320625305176, "learning_rate": 4.199187581604527e-06, "loss": 0.0436, "step": 21883 }, { "epoch": 2.372506504770165, "grad_norm": 0.22249245643615723, "learning_rate": 4.19846220803714e-06, "loss": 0.0044, "step": 21884 }, { "epoch": 2.3726149176062448, "grad_norm": 0.17653577029705048, "learning_rate": 4.197736834469752e-06, "loss": 0.0008, "step": 21885 }, { "epoch": 2.3727233304423243, "grad_norm": 0.08386433869600296, "learning_rate": 4.197011460902365e-06, "loss": 0.0006, "step": 21886 }, { "epoch": 2.3728317432784043, "grad_norm": 0.26353132724761963, "learning_rate": 4.196286087334978e-06, "loss": 0.003, "step": 21887 }, { "epoch": 2.372940156114484, "grad_norm": 0.370211660861969, "learning_rate": 4.195560713767591e-06, "loss": 0.0052, "step": 21888 }, { "epoch": 2.373048568950564, "grad_norm": 0.3472067713737488, "learning_rate": 4.194835340200204e-06, "loss": 0.0118, "step": 21889 }, { "epoch": 2.3731569817866434, "grad_norm": 0.02275586873292923, "learning_rate": 4.194109966632816e-06, "loss": 0.0002, "step": 21890 }, { "epoch": 2.3732653946227233, "grad_norm": 0.9450488090515137, "learning_rate": 4.193384593065429e-06, "loss": 0.0288, "step": 21891 }, { "epoch": 2.373373807458803, "grad_norm": 0.06000290811061859, "learning_rate": 4.192659219498042e-06, "loss": 0.0005, "step": 21892 }, { "epoch": 2.373482220294883, "grad_norm": 0.2749706208705902, "learning_rate": 4.191933845930655e-06, "loss": 0.0057, "step": 21893 }, { "epoch": 2.373590633130963, "grad_norm": 0.3324931263923645, "learning_rate": 4.191208472363267e-06, "loss": 0.011, "step": 21894 }, { "epoch": 2.3736990459670424, "grad_norm": 0.10022931545972824, "learning_rate": 4.19048309879588e-06, "loss": 0.0004, "step": 21895 }, { "epoch": 2.3738074588031224, "grad_norm": 0.3528941869735718, "learning_rate": 4.189757725228493e-06, "loss": 0.0117, "step": 21896 }, { "epoch": 2.373915871639202, "grad_norm": 0.6958497166633606, "learning_rate": 4.189032351661106e-06, "loss": 0.0057, "step": 21897 }, { "epoch": 2.374024284475282, "grad_norm": 0.6026723980903625, "learning_rate": 4.188306978093718e-06, "loss": 0.0131, "step": 21898 }, { "epoch": 2.374132697311362, "grad_norm": 0.2523888051509857, "learning_rate": 4.187581604526331e-06, "loss": 0.0047, "step": 21899 }, { "epoch": 2.3742411101474414, "grad_norm": 0.11186806112527847, "learning_rate": 4.186856230958944e-06, "loss": 0.0006, "step": 21900 }, { "epoch": 2.3743495229835214, "grad_norm": 3.346254825592041, "learning_rate": 4.186130857391557e-06, "loss": 0.0057, "step": 21901 }, { "epoch": 2.374457935819601, "grad_norm": 0.017064671963453293, "learning_rate": 4.18540548382417e-06, "loss": 0.0002, "step": 21902 }, { "epoch": 2.374566348655681, "grad_norm": 0.42708301544189453, "learning_rate": 4.184680110256782e-06, "loss": 0.0126, "step": 21903 }, { "epoch": 2.3746747614917605, "grad_norm": 0.7307039499282837, "learning_rate": 4.183954736689395e-06, "loss": 0.0013, "step": 21904 }, { "epoch": 2.3747831743278405, "grad_norm": 0.21257972717285156, "learning_rate": 4.183229363122008e-06, "loss": 0.001, "step": 21905 }, { "epoch": 2.37489158716392, "grad_norm": 0.0100612323731184, "learning_rate": 4.182503989554621e-06, "loss": 0.0001, "step": 21906 }, { "epoch": 2.375, "grad_norm": 1.4999163150787354, "learning_rate": 4.181778615987233e-06, "loss": 0.0149, "step": 21907 }, { "epoch": 2.37510841283608, "grad_norm": 0.027969328686594963, "learning_rate": 4.181053242419846e-06, "loss": 0.0003, "step": 21908 }, { "epoch": 2.3752168256721595, "grad_norm": 0.35581573843955994, "learning_rate": 4.180327868852459e-06, "loss": 0.0071, "step": 21909 }, { "epoch": 2.3753252385082395, "grad_norm": 9.707934379577637, "learning_rate": 4.179602495285072e-06, "loss": 0.0443, "step": 21910 }, { "epoch": 2.375433651344319, "grad_norm": 1.653437614440918, "learning_rate": 4.1788771217176845e-06, "loss": 0.0353, "step": 21911 }, { "epoch": 2.375542064180399, "grad_norm": 5.691119194030762, "learning_rate": 4.1781517481502975e-06, "loss": 0.0172, "step": 21912 }, { "epoch": 2.3756504770164786, "grad_norm": 0.3113323450088501, "learning_rate": 4.1774263745829105e-06, "loss": 0.0056, "step": 21913 }, { "epoch": 2.3757588898525586, "grad_norm": 0.025889893993735313, "learning_rate": 4.1767010010155235e-06, "loss": 0.0001, "step": 21914 }, { "epoch": 2.375867302688638, "grad_norm": 0.04720645025372505, "learning_rate": 4.1759756274481365e-06, "loss": 0.0003, "step": 21915 }, { "epoch": 2.375975715524718, "grad_norm": 0.3052423596382141, "learning_rate": 4.175250253880749e-06, "loss": 0.0059, "step": 21916 }, { "epoch": 2.376084128360798, "grad_norm": 0.8259450793266296, "learning_rate": 4.174524880313362e-06, "loss": 0.007, "step": 21917 }, { "epoch": 2.3761925411968776, "grad_norm": 0.7496940493583679, "learning_rate": 4.173799506745975e-06, "loss": 0.0167, "step": 21918 }, { "epoch": 2.3763009540329576, "grad_norm": 0.64179527759552, "learning_rate": 4.1730741331785876e-06, "loss": 0.0247, "step": 21919 }, { "epoch": 2.376409366869037, "grad_norm": 0.4654991924762726, "learning_rate": 4.1723487596112e-06, "loss": 0.0088, "step": 21920 }, { "epoch": 2.376517779705117, "grad_norm": 0.2725600600242615, "learning_rate": 4.171623386043813e-06, "loss": 0.0057, "step": 21921 }, { "epoch": 2.376626192541197, "grad_norm": 2.751335382461548, "learning_rate": 4.170898012476426e-06, "loss": 0.0358, "step": 21922 }, { "epoch": 2.3767346053772767, "grad_norm": 0.06568542867898941, "learning_rate": 4.170172638909039e-06, "loss": 0.0005, "step": 21923 }, { "epoch": 2.3768430182133566, "grad_norm": 0.33811020851135254, "learning_rate": 4.169447265341651e-06, "loss": 0.0111, "step": 21924 }, { "epoch": 2.376951431049436, "grad_norm": 0.4323902130126953, "learning_rate": 4.168721891774264e-06, "loss": 0.0082, "step": 21925 }, { "epoch": 2.377059843885516, "grad_norm": 0.22937917709350586, "learning_rate": 4.167996518206877e-06, "loss": 0.0041, "step": 21926 }, { "epoch": 2.3771682567215957, "grad_norm": 0.3434474468231201, "learning_rate": 4.16727114463949e-06, "loss": 0.0095, "step": 21927 }, { "epoch": 2.3772766695576757, "grad_norm": 0.4789629876613617, "learning_rate": 4.166545771072103e-06, "loss": 0.0013, "step": 21928 }, { "epoch": 2.3773850823937552, "grad_norm": 0.08547046035528183, "learning_rate": 4.165820397504715e-06, "loss": 0.0006, "step": 21929 }, { "epoch": 2.377493495229835, "grad_norm": 0.3318701982498169, "learning_rate": 4.165095023937328e-06, "loss": 0.0106, "step": 21930 }, { "epoch": 2.377601908065915, "grad_norm": 0.8067450523376465, "learning_rate": 4.164369650369941e-06, "loss": 0.0028, "step": 21931 }, { "epoch": 2.3777103209019947, "grad_norm": 0.6980790495872498, "learning_rate": 4.163644276802554e-06, "loss": 0.0151, "step": 21932 }, { "epoch": 2.3778187337380747, "grad_norm": 0.12007148563861847, "learning_rate": 4.162918903235166e-06, "loss": 0.0004, "step": 21933 }, { "epoch": 2.3779271465741543, "grad_norm": 0.007129628211259842, "learning_rate": 4.162193529667779e-06, "loss": 0.0001, "step": 21934 }, { "epoch": 2.3780355594102343, "grad_norm": 0.023175077512860298, "learning_rate": 4.161468156100392e-06, "loss": 0.0002, "step": 21935 }, { "epoch": 2.378143972246314, "grad_norm": 1.4517790079116821, "learning_rate": 4.160742782533005e-06, "loss": 0.0029, "step": 21936 }, { "epoch": 2.378252385082394, "grad_norm": 0.030753780156373978, "learning_rate": 4.160017408965617e-06, "loss": 0.0003, "step": 21937 }, { "epoch": 2.3783607979184733, "grad_norm": 2.0052173137664795, "learning_rate": 4.15929203539823e-06, "loss": 0.0055, "step": 21938 }, { "epoch": 2.3784692107545533, "grad_norm": 0.24693986773490906, "learning_rate": 4.158566661830843e-06, "loss": 0.0052, "step": 21939 }, { "epoch": 2.3785776235906333, "grad_norm": 2.464583158493042, "learning_rate": 4.157841288263456e-06, "loss": 0.0699, "step": 21940 }, { "epoch": 2.378686036426713, "grad_norm": 0.2736243009567261, "learning_rate": 4.157115914696069e-06, "loss": 0.0062, "step": 21941 }, { "epoch": 2.378794449262793, "grad_norm": 0.23931340873241425, "learning_rate": 4.156390541128681e-06, "loss": 0.0047, "step": 21942 }, { "epoch": 2.3789028620988724, "grad_norm": 0.13311389088630676, "learning_rate": 4.155665167561294e-06, "loss": 0.0117, "step": 21943 }, { "epoch": 2.3790112749349523, "grad_norm": 0.4952923357486725, "learning_rate": 4.154939793993907e-06, "loss": 0.0093, "step": 21944 }, { "epoch": 2.3791196877710323, "grad_norm": 5.18237829208374, "learning_rate": 4.15421442042652e-06, "loss": 0.0112, "step": 21945 }, { "epoch": 2.379228100607112, "grad_norm": 2.5847537517547607, "learning_rate": 4.153489046859132e-06, "loss": 0.0644, "step": 21946 }, { "epoch": 2.379336513443192, "grad_norm": 0.028687188401818275, "learning_rate": 4.152763673291745e-06, "loss": 0.0002, "step": 21947 }, { "epoch": 2.3794449262792714, "grad_norm": 0.31165966391563416, "learning_rate": 4.152038299724358e-06, "loss": 0.0066, "step": 21948 }, { "epoch": 2.3795533391153514, "grad_norm": 0.6672695875167847, "learning_rate": 4.151312926156971e-06, "loss": 0.0125, "step": 21949 }, { "epoch": 2.379661751951431, "grad_norm": 0.04516826197504997, "learning_rate": 4.1505875525895835e-06, "loss": 0.0004, "step": 21950 }, { "epoch": 2.379770164787511, "grad_norm": 0.18431754410266876, "learning_rate": 4.1498621790221965e-06, "loss": 0.0007, "step": 21951 }, { "epoch": 2.3798785776235905, "grad_norm": 0.7610020041465759, "learning_rate": 4.1491368054548095e-06, "loss": 0.0224, "step": 21952 }, { "epoch": 2.3799869904596704, "grad_norm": 2.0951027870178223, "learning_rate": 4.1484114318874225e-06, "loss": 0.0221, "step": 21953 }, { "epoch": 2.3800954032957504, "grad_norm": 3.921710252761841, "learning_rate": 4.1476860583200355e-06, "loss": 0.0141, "step": 21954 }, { "epoch": 2.38020381613183, "grad_norm": 7.463258266448975, "learning_rate": 4.146960684752648e-06, "loss": 0.0087, "step": 21955 }, { "epoch": 2.38031222896791, "grad_norm": 0.41979581117630005, "learning_rate": 4.146235311185261e-06, "loss": 0.0119, "step": 21956 }, { "epoch": 2.3804206418039895, "grad_norm": 0.006790009327232838, "learning_rate": 4.145509937617874e-06, "loss": 0.0001, "step": 21957 }, { "epoch": 2.3805290546400695, "grad_norm": 3.2938919067382812, "learning_rate": 4.1447845640504866e-06, "loss": 0.0289, "step": 21958 }, { "epoch": 2.380637467476149, "grad_norm": 0.5652727484703064, "learning_rate": 4.144059190483099e-06, "loss": 0.0148, "step": 21959 }, { "epoch": 2.380745880312229, "grad_norm": 0.22561530768871307, "learning_rate": 4.143333816915712e-06, "loss": 0.0008, "step": 21960 }, { "epoch": 2.3808542931483085, "grad_norm": 2.3314414024353027, "learning_rate": 4.142608443348325e-06, "loss": 0.0212, "step": 21961 }, { "epoch": 2.3809627059843885, "grad_norm": 0.015733573585748672, "learning_rate": 4.141883069780938e-06, "loss": 0.0002, "step": 21962 }, { "epoch": 2.3810711188204685, "grad_norm": 0.29401248693466187, "learning_rate": 4.14115769621355e-06, "loss": 0.0056, "step": 21963 }, { "epoch": 2.381179531656548, "grad_norm": 2.1966941356658936, "learning_rate": 4.140432322646163e-06, "loss": 0.0181, "step": 21964 }, { "epoch": 2.381287944492628, "grad_norm": 0.2001904994249344, "learning_rate": 4.139706949078776e-06, "loss": 0.0004, "step": 21965 }, { "epoch": 2.3813963573287076, "grad_norm": 0.9766460657119751, "learning_rate": 4.138981575511389e-06, "loss": 0.0004, "step": 21966 }, { "epoch": 2.3815047701647876, "grad_norm": 0.07163898646831512, "learning_rate": 4.138256201944002e-06, "loss": 0.0002, "step": 21967 }, { "epoch": 2.3816131830008676, "grad_norm": 0.6514934301376343, "learning_rate": 4.137530828376614e-06, "loss": 0.0216, "step": 21968 }, { "epoch": 2.381721595836947, "grad_norm": 0.9425267577171326, "learning_rate": 4.136805454809228e-06, "loss": 0.0235, "step": 21969 }, { "epoch": 2.381830008673027, "grad_norm": 3.854013204574585, "learning_rate": 4.13608008124184e-06, "loss": 0.0168, "step": 21970 }, { "epoch": 2.3819384215091066, "grad_norm": 0.7625384330749512, "learning_rate": 4.135354707674453e-06, "loss": 0.0283, "step": 21971 }, { "epoch": 2.3820468343451866, "grad_norm": 0.30286791920661926, "learning_rate": 4.134629334107065e-06, "loss": 0.0104, "step": 21972 }, { "epoch": 2.382155247181266, "grad_norm": 7.107115268707275, "learning_rate": 4.133903960539678e-06, "loss": 0.0234, "step": 21973 }, { "epoch": 2.382263660017346, "grad_norm": 0.4121883511543274, "learning_rate": 4.133178586972291e-06, "loss": 0.008, "step": 21974 }, { "epoch": 2.3823720728534257, "grad_norm": 0.568435549736023, "learning_rate": 4.132453213404904e-06, "loss": 0.0132, "step": 21975 }, { "epoch": 2.3824804856895057, "grad_norm": 0.6281443238258362, "learning_rate": 4.131727839837516e-06, "loss": 0.016, "step": 21976 }, { "epoch": 2.3825888985255856, "grad_norm": 0.5582748651504517, "learning_rate": 4.131002466270129e-06, "loss": 0.017, "step": 21977 }, { "epoch": 2.382697311361665, "grad_norm": 0.3629351258277893, "learning_rate": 4.130277092702742e-06, "loss": 0.0054, "step": 21978 }, { "epoch": 2.382805724197745, "grad_norm": 0.23113220930099487, "learning_rate": 4.129551719135355e-06, "loss": 0.0048, "step": 21979 }, { "epoch": 2.3829141370338247, "grad_norm": 0.2974255084991455, "learning_rate": 4.128826345567968e-06, "loss": 0.0114, "step": 21980 }, { "epoch": 2.3830225498699047, "grad_norm": 0.37745100259780884, "learning_rate": 4.12810097200058e-06, "loss": 0.0128, "step": 21981 }, { "epoch": 2.3831309627059842, "grad_norm": 0.037257809191942215, "learning_rate": 4.127375598433194e-06, "loss": 0.0003, "step": 21982 }, { "epoch": 2.383239375542064, "grad_norm": 0.579034686088562, "learning_rate": 4.126650224865806e-06, "loss": 0.016, "step": 21983 }, { "epoch": 2.3833477883781438, "grad_norm": 0.40352585911750793, "learning_rate": 4.125924851298419e-06, "loss": 0.0005, "step": 21984 }, { "epoch": 2.3834562012142237, "grad_norm": 2.4805314540863037, "learning_rate": 4.125199477731031e-06, "loss": 0.0078, "step": 21985 }, { "epoch": 2.3835646140503037, "grad_norm": 0.45439091324806213, "learning_rate": 4.124474104163644e-06, "loss": 0.0081, "step": 21986 }, { "epoch": 2.3836730268863833, "grad_norm": 7.183080196380615, "learning_rate": 4.123748730596257e-06, "loss": 0.0077, "step": 21987 }, { "epoch": 2.3837814397224633, "grad_norm": 0.2505318224430084, "learning_rate": 4.12302335702887e-06, "loss": 0.0048, "step": 21988 }, { "epoch": 2.383889852558543, "grad_norm": 0.04889696463942528, "learning_rate": 4.1222979834614825e-06, "loss": 0.0002, "step": 21989 }, { "epoch": 2.383998265394623, "grad_norm": 1.448278784751892, "learning_rate": 4.1215726098940955e-06, "loss": 0.0458, "step": 21990 }, { "epoch": 2.3841066782307023, "grad_norm": 0.32197895646095276, "learning_rate": 4.1208472363267085e-06, "loss": 0.0109, "step": 21991 }, { "epoch": 2.3842150910667823, "grad_norm": 6.397243976593018, "learning_rate": 4.1201218627593215e-06, "loss": 0.0583, "step": 21992 }, { "epoch": 2.384323503902862, "grad_norm": 0.7207032442092896, "learning_rate": 4.1193964891919345e-06, "loss": 0.0295, "step": 21993 }, { "epoch": 2.384431916738942, "grad_norm": 0.30838826298713684, "learning_rate": 4.118671115624547e-06, "loss": 0.0107, "step": 21994 }, { "epoch": 2.384540329575022, "grad_norm": 1.1365344524383545, "learning_rate": 4.1179457420571604e-06, "loss": 0.0102, "step": 21995 }, { "epoch": 2.3846487424111014, "grad_norm": 0.05569150298833847, "learning_rate": 4.117220368489773e-06, "loss": 0.0004, "step": 21996 }, { "epoch": 2.3847571552471813, "grad_norm": 1.1483814716339111, "learning_rate": 4.1164949949223856e-06, "loss": 0.0054, "step": 21997 }, { "epoch": 2.384865568083261, "grad_norm": 0.24307367205619812, "learning_rate": 4.115769621354998e-06, "loss": 0.0046, "step": 21998 }, { "epoch": 2.384973980919341, "grad_norm": 0.28310462832450867, "learning_rate": 4.115044247787611e-06, "loss": 0.0069, "step": 21999 }, { "epoch": 2.385082393755421, "grad_norm": 0.9806922674179077, "learning_rate": 4.114318874220224e-06, "loss": 0.0104, "step": 22000 }, { "epoch": 2.3851908065915004, "grad_norm": 3.6236844062805176, "learning_rate": 4.113593500652837e-06, "loss": 0.024, "step": 22001 }, { "epoch": 2.3852992194275804, "grad_norm": 0.5803298354148865, "learning_rate": 4.112868127085449e-06, "loss": 0.0306, "step": 22002 }, { "epoch": 2.38540763226366, "grad_norm": 3.3660478591918945, "learning_rate": 4.112142753518062e-06, "loss": 0.0535, "step": 22003 }, { "epoch": 2.38551604509974, "grad_norm": 0.018328247591853142, "learning_rate": 4.111417379950675e-06, "loss": 0.0002, "step": 22004 }, { "epoch": 2.3856244579358195, "grad_norm": 5.824560642242432, "learning_rate": 4.110692006383288e-06, "loss": 0.0317, "step": 22005 }, { "epoch": 2.3857328707718994, "grad_norm": 3.267488718032837, "learning_rate": 4.109966632815901e-06, "loss": 0.0353, "step": 22006 }, { "epoch": 2.385841283607979, "grad_norm": 0.009033305570483208, "learning_rate": 4.109241259248513e-06, "loss": 0.0001, "step": 22007 }, { "epoch": 2.385949696444059, "grad_norm": 0.3757498562335968, "learning_rate": 4.108515885681127e-06, "loss": 0.009, "step": 22008 }, { "epoch": 2.386058109280139, "grad_norm": 0.11071480810642242, "learning_rate": 4.107790512113739e-06, "loss": 0.0004, "step": 22009 }, { "epoch": 2.3861665221162185, "grad_norm": 0.34631818532943726, "learning_rate": 4.107065138546352e-06, "loss": 0.0129, "step": 22010 }, { "epoch": 2.3862749349522985, "grad_norm": 6.370347499847412, "learning_rate": 4.106339764978964e-06, "loss": 0.0568, "step": 22011 }, { "epoch": 2.386383347788378, "grad_norm": 1.4711103439331055, "learning_rate": 4.105614391411577e-06, "loss": 0.0008, "step": 22012 }, { "epoch": 2.386491760624458, "grad_norm": 0.13020311295986176, "learning_rate": 4.10488901784419e-06, "loss": 0.0007, "step": 22013 }, { "epoch": 2.3866001734605375, "grad_norm": 8.356855392456055, "learning_rate": 4.104163644276803e-06, "loss": 0.0489, "step": 22014 }, { "epoch": 2.3867085862966175, "grad_norm": 0.02954714186489582, "learning_rate": 4.103438270709415e-06, "loss": 0.0002, "step": 22015 }, { "epoch": 2.386816999132697, "grad_norm": 0.39902812242507935, "learning_rate": 4.102712897142028e-06, "loss": 0.0214, "step": 22016 }, { "epoch": 2.386925411968777, "grad_norm": 0.1505589485168457, "learning_rate": 4.101987523574641e-06, "loss": 0.0007, "step": 22017 }, { "epoch": 2.387033824804857, "grad_norm": 4.133419036865234, "learning_rate": 4.101262150007254e-06, "loss": 0.0572, "step": 22018 }, { "epoch": 2.3871422376409366, "grad_norm": 0.10614771395921707, "learning_rate": 4.100536776439867e-06, "loss": 0.0006, "step": 22019 }, { "epoch": 2.3872506504770166, "grad_norm": 0.025770513340830803, "learning_rate": 4.099811402872479e-06, "loss": 0.0002, "step": 22020 }, { "epoch": 2.387359063313096, "grad_norm": 7.6367363929748535, "learning_rate": 4.099086029305093e-06, "loss": 0.0157, "step": 22021 }, { "epoch": 2.387467476149176, "grad_norm": 1.0089046955108643, "learning_rate": 4.098360655737705e-06, "loss": 0.0283, "step": 22022 }, { "epoch": 2.387575888985256, "grad_norm": 0.49082791805267334, "learning_rate": 4.097635282170318e-06, "loss": 0.0208, "step": 22023 }, { "epoch": 2.3876843018213356, "grad_norm": 0.3897392451763153, "learning_rate": 4.09690990860293e-06, "loss": 0.001, "step": 22024 }, { "epoch": 2.3877927146574156, "grad_norm": 0.020382553339004517, "learning_rate": 4.096184535035543e-06, "loss": 0.0002, "step": 22025 }, { "epoch": 2.387901127493495, "grad_norm": 2.100018262863159, "learning_rate": 4.095459161468156e-06, "loss": 0.0025, "step": 22026 }, { "epoch": 2.388009540329575, "grad_norm": 0.3151892423629761, "learning_rate": 4.094733787900769e-06, "loss": 0.0071, "step": 22027 }, { "epoch": 2.3881179531656547, "grad_norm": 2.169071674346924, "learning_rate": 4.0940084143333815e-06, "loss": 0.0179, "step": 22028 }, { "epoch": 2.3882263660017347, "grad_norm": 0.07479439675807953, "learning_rate": 4.0932830407659945e-06, "loss": 0.0003, "step": 22029 }, { "epoch": 2.388334778837814, "grad_norm": 0.31542813777923584, "learning_rate": 4.0925576671986075e-06, "loss": 0.0055, "step": 22030 }, { "epoch": 2.388443191673894, "grad_norm": 0.35916638374328613, "learning_rate": 4.0918322936312205e-06, "loss": 0.0082, "step": 22031 }, { "epoch": 2.388551604509974, "grad_norm": 0.11869063973426819, "learning_rate": 4.0911069200638335e-06, "loss": 0.0005, "step": 22032 }, { "epoch": 2.3886600173460537, "grad_norm": 0.19186845421791077, "learning_rate": 4.090381546496446e-06, "loss": 0.0004, "step": 22033 }, { "epoch": 2.3887684301821337, "grad_norm": 0.3176618218421936, "learning_rate": 4.0896561729290594e-06, "loss": 0.0076, "step": 22034 }, { "epoch": 2.3888768430182132, "grad_norm": 17.47640037536621, "learning_rate": 4.088930799361672e-06, "loss": 0.0224, "step": 22035 }, { "epoch": 2.388985255854293, "grad_norm": 0.29597312211990356, "learning_rate": 4.088205425794285e-06, "loss": 0.0106, "step": 22036 }, { "epoch": 2.3890936686903728, "grad_norm": 0.2954713702201843, "learning_rate": 4.087480052226897e-06, "loss": 0.0129, "step": 22037 }, { "epoch": 2.3892020815264527, "grad_norm": 0.03079473413527012, "learning_rate": 4.0867546786595106e-06, "loss": 0.0002, "step": 22038 }, { "epoch": 2.3893104943625323, "grad_norm": 6.6756463050842285, "learning_rate": 4.086029305092123e-06, "loss": 0.0258, "step": 22039 }, { "epoch": 2.3894189071986123, "grad_norm": 10.192560195922852, "learning_rate": 4.085303931524736e-06, "loss": 0.0104, "step": 22040 }, { "epoch": 2.3895273200346923, "grad_norm": 7.984895706176758, "learning_rate": 4.084578557957348e-06, "loss": 0.0592, "step": 22041 }, { "epoch": 2.389635732870772, "grad_norm": 1.336544156074524, "learning_rate": 4.083853184389961e-06, "loss": 0.0385, "step": 22042 }, { "epoch": 2.389744145706852, "grad_norm": 0.29345792531967163, "learning_rate": 4.083127810822574e-06, "loss": 0.0067, "step": 22043 }, { "epoch": 2.3898525585429313, "grad_norm": 2.323369264602661, "learning_rate": 4.082402437255187e-06, "loss": 0.0677, "step": 22044 }, { "epoch": 2.3899609713790113, "grad_norm": 0.30135542154312134, "learning_rate": 4.0816770636878e-06, "loss": 0.0094, "step": 22045 }, { "epoch": 2.3900693842150913, "grad_norm": 2.726196527481079, "learning_rate": 4.080951690120412e-06, "loss": 0.042, "step": 22046 }, { "epoch": 2.390177797051171, "grad_norm": 0.8627971410751343, "learning_rate": 4.080226316553026e-06, "loss": 0.0015, "step": 22047 }, { "epoch": 2.390286209887251, "grad_norm": 2.3836982250213623, "learning_rate": 4.079500942985638e-06, "loss": 0.0275, "step": 22048 }, { "epoch": 2.3903946227233304, "grad_norm": 1.1955090761184692, "learning_rate": 4.078775569418251e-06, "loss": 0.0468, "step": 22049 }, { "epoch": 2.3905030355594103, "grad_norm": 0.28075748682022095, "learning_rate": 4.078050195850863e-06, "loss": 0.006, "step": 22050 }, { "epoch": 2.39061144839549, "grad_norm": 0.024303600192070007, "learning_rate": 4.077324822283477e-06, "loss": 0.0002, "step": 22051 }, { "epoch": 2.39071986123157, "grad_norm": 0.7765716314315796, "learning_rate": 4.076599448716089e-06, "loss": 0.0072, "step": 22052 }, { "epoch": 2.3908282740676494, "grad_norm": 0.04934629052877426, "learning_rate": 4.075874075148702e-06, "loss": 0.0005, "step": 22053 }, { "epoch": 2.3909366869037294, "grad_norm": 1.0955952405929565, "learning_rate": 4.075148701581314e-06, "loss": 0.0084, "step": 22054 }, { "epoch": 2.3910450997398094, "grad_norm": 1.535392165184021, "learning_rate": 4.074423328013927e-06, "loss": 0.0343, "step": 22055 }, { "epoch": 2.391153512575889, "grad_norm": 0.05038481950759888, "learning_rate": 4.07369795444654e-06, "loss": 0.0004, "step": 22056 }, { "epoch": 2.391261925411969, "grad_norm": 1.6332517862319946, "learning_rate": 4.072972580879153e-06, "loss": 0.0097, "step": 22057 }, { "epoch": 2.3913703382480485, "grad_norm": 4.117869853973389, "learning_rate": 4.072247207311766e-06, "loss": 0.0134, "step": 22058 }, { "epoch": 2.3914787510841284, "grad_norm": 0.26940596103668213, "learning_rate": 4.071521833744378e-06, "loss": 0.0064, "step": 22059 }, { "epoch": 2.391587163920208, "grad_norm": 0.2836626172065735, "learning_rate": 4.070796460176992e-06, "loss": 0.0088, "step": 22060 }, { "epoch": 2.391695576756288, "grad_norm": 0.5681830048561096, "learning_rate": 4.070071086609604e-06, "loss": 0.0168, "step": 22061 }, { "epoch": 2.3918039895923675, "grad_norm": 0.1376207023859024, "learning_rate": 4.069345713042217e-06, "loss": 0.0011, "step": 22062 }, { "epoch": 2.3919124024284475, "grad_norm": 0.05877835303544998, "learning_rate": 4.068620339474829e-06, "loss": 0.0005, "step": 22063 }, { "epoch": 2.3920208152645275, "grad_norm": 0.21785664558410645, "learning_rate": 4.067894965907443e-06, "loss": 0.0141, "step": 22064 }, { "epoch": 2.392129228100607, "grad_norm": 0.052334535866975784, "learning_rate": 4.067169592340055e-06, "loss": 0.0003, "step": 22065 }, { "epoch": 2.392237640936687, "grad_norm": 2.841600179672241, "learning_rate": 4.066444218772668e-06, "loss": 0.0235, "step": 22066 }, { "epoch": 2.3923460537727665, "grad_norm": 0.05164163559675217, "learning_rate": 4.065718845205281e-06, "loss": 0.0004, "step": 22067 }, { "epoch": 2.3924544666088465, "grad_norm": 0.2458251565694809, "learning_rate": 4.0649934716378935e-06, "loss": 0.0051, "step": 22068 }, { "epoch": 2.3925628794449265, "grad_norm": 0.24629150331020355, "learning_rate": 4.0642680980705065e-06, "loss": 0.0056, "step": 22069 }, { "epoch": 2.392671292281006, "grad_norm": 0.4042702317237854, "learning_rate": 4.0635427245031195e-06, "loss": 0.0013, "step": 22070 }, { "epoch": 2.392779705117086, "grad_norm": 0.25434601306915283, "learning_rate": 4.0628173509357325e-06, "loss": 0.0057, "step": 22071 }, { "epoch": 2.3928881179531656, "grad_norm": 3.3495585918426514, "learning_rate": 4.062091977368345e-06, "loss": 0.0384, "step": 22072 }, { "epoch": 2.3929965307892456, "grad_norm": 0.24932129681110382, "learning_rate": 4.0613666038009584e-06, "loss": 0.0042, "step": 22073 }, { "epoch": 2.393104943625325, "grad_norm": 5.970222473144531, "learning_rate": 4.060641230233571e-06, "loss": 0.0106, "step": 22074 }, { "epoch": 2.393213356461405, "grad_norm": 1.0840901136398315, "learning_rate": 4.059915856666184e-06, "loss": 0.0031, "step": 22075 }, { "epoch": 2.3933217692974846, "grad_norm": 0.2763279676437378, "learning_rate": 4.059190483098796e-06, "loss": 0.0066, "step": 22076 }, { "epoch": 2.3934301821335646, "grad_norm": 0.481142520904541, "learning_rate": 4.0584651095314096e-06, "loss": 0.0146, "step": 22077 }, { "epoch": 2.3935385949696446, "grad_norm": 3.417123556137085, "learning_rate": 4.057739735964022e-06, "loss": 0.0193, "step": 22078 }, { "epoch": 2.393647007805724, "grad_norm": 1.0017168521881104, "learning_rate": 4.057014362396635e-06, "loss": 0.0171, "step": 22079 }, { "epoch": 2.393755420641804, "grad_norm": 0.5307424068450928, "learning_rate": 4.056288988829248e-06, "loss": 0.0159, "step": 22080 }, { "epoch": 2.3938638334778837, "grad_norm": 0.3663407564163208, "learning_rate": 4.05556361526186e-06, "loss": 0.0092, "step": 22081 }, { "epoch": 2.3939722463139637, "grad_norm": 0.6658740043640137, "learning_rate": 4.054838241694473e-06, "loss": 0.0208, "step": 22082 }, { "epoch": 2.394080659150043, "grad_norm": 2.702465057373047, "learning_rate": 4.054112868127086e-06, "loss": 0.0614, "step": 22083 }, { "epoch": 2.394189071986123, "grad_norm": 0.05356170982122421, "learning_rate": 4.053387494559699e-06, "loss": 0.0004, "step": 22084 }, { "epoch": 2.3942974848222027, "grad_norm": 0.08608498424291611, "learning_rate": 4.052662120992311e-06, "loss": 0.0004, "step": 22085 }, { "epoch": 2.3944058976582827, "grad_norm": 1.0073535442352295, "learning_rate": 4.051936747424925e-06, "loss": 0.0012, "step": 22086 }, { "epoch": 2.3945143104943627, "grad_norm": 2.254058599472046, "learning_rate": 4.051211373857537e-06, "loss": 0.0471, "step": 22087 }, { "epoch": 2.3946227233304422, "grad_norm": 0.33663633465766907, "learning_rate": 4.05048600029015e-06, "loss": 0.0155, "step": 22088 }, { "epoch": 2.3947311361665222, "grad_norm": 0.34974709153175354, "learning_rate": 4.049760626722762e-06, "loss": 0.0119, "step": 22089 }, { "epoch": 2.3948395490026018, "grad_norm": 0.37778863310813904, "learning_rate": 4.049035253155376e-06, "loss": 0.0213, "step": 22090 }, { "epoch": 2.3949479618386817, "grad_norm": 0.2178962230682373, "learning_rate": 4.048309879587988e-06, "loss": 0.0009, "step": 22091 }, { "epoch": 2.3950563746747617, "grad_norm": 0.4542938470840454, "learning_rate": 4.047584506020601e-06, "loss": 0.0204, "step": 22092 }, { "epoch": 2.3951647875108413, "grad_norm": 0.8659388422966003, "learning_rate": 4.046859132453214e-06, "loss": 0.0333, "step": 22093 }, { "epoch": 2.3952732003469213, "grad_norm": 0.5680803060531616, "learning_rate": 4.046133758885826e-06, "loss": 0.0138, "step": 22094 }, { "epoch": 2.395381613183001, "grad_norm": 3.4650051593780518, "learning_rate": 4.045408385318439e-06, "loss": 0.0238, "step": 22095 }, { "epoch": 2.395490026019081, "grad_norm": 0.025147346779704094, "learning_rate": 4.044683011751052e-06, "loss": 0.0002, "step": 22096 }, { "epoch": 2.3955984388551603, "grad_norm": 0.74495929479599, "learning_rate": 4.043957638183665e-06, "loss": 0.0108, "step": 22097 }, { "epoch": 2.3957068516912403, "grad_norm": 0.4264942407608032, "learning_rate": 4.043232264616277e-06, "loss": 0.0116, "step": 22098 }, { "epoch": 2.39581526452732, "grad_norm": 0.30914583802223206, "learning_rate": 4.042506891048891e-06, "loss": 0.0008, "step": 22099 }, { "epoch": 2.3959236773634, "grad_norm": 0.08403702080249786, "learning_rate": 4.041781517481503e-06, "loss": 0.0007, "step": 22100 }, { "epoch": 2.39603209019948, "grad_norm": 0.14836689829826355, "learning_rate": 4.041056143914116e-06, "loss": 0.0009, "step": 22101 }, { "epoch": 2.3961405030355594, "grad_norm": 0.5043883323669434, "learning_rate": 4.040330770346728e-06, "loss": 0.0179, "step": 22102 }, { "epoch": 2.3962489158716394, "grad_norm": 0.28111812472343445, "learning_rate": 4.039605396779342e-06, "loss": 0.0159, "step": 22103 }, { "epoch": 2.396357328707719, "grad_norm": 6.099238395690918, "learning_rate": 4.038880023211954e-06, "loss": 0.0265, "step": 22104 }, { "epoch": 2.396465741543799, "grad_norm": 0.03319516405463219, "learning_rate": 4.038154649644567e-06, "loss": 0.0004, "step": 22105 }, { "epoch": 2.3965741543798784, "grad_norm": 0.35133498907089233, "learning_rate": 4.03742927607718e-06, "loss": 0.007, "step": 22106 }, { "epoch": 2.3966825672159584, "grad_norm": 4.3575286865234375, "learning_rate": 4.0367039025097925e-06, "loss": 0.0045, "step": 22107 }, { "epoch": 2.396790980052038, "grad_norm": 0.2395043969154358, "learning_rate": 4.0359785289424055e-06, "loss": 0.0048, "step": 22108 }, { "epoch": 2.396899392888118, "grad_norm": 1.4867087602615356, "learning_rate": 4.0352531553750185e-06, "loss": 0.0135, "step": 22109 }, { "epoch": 2.397007805724198, "grad_norm": 0.31160399317741394, "learning_rate": 4.0345277818076315e-06, "loss": 0.0061, "step": 22110 }, { "epoch": 2.3971162185602775, "grad_norm": 0.48068368434906006, "learning_rate": 4.033802408240244e-06, "loss": 0.0229, "step": 22111 }, { "epoch": 2.3972246313963574, "grad_norm": 0.5089395046234131, "learning_rate": 4.0330770346728575e-06, "loss": 0.0107, "step": 22112 }, { "epoch": 2.397333044232437, "grad_norm": 0.5641557574272156, "learning_rate": 4.03235166110547e-06, "loss": 0.0141, "step": 22113 }, { "epoch": 2.397441457068517, "grad_norm": 9.976240158081055, "learning_rate": 4.031626287538083e-06, "loss": 0.0558, "step": 22114 }, { "epoch": 2.397549869904597, "grad_norm": 0.20946986973285675, "learning_rate": 4.030900913970695e-06, "loss": 0.001, "step": 22115 }, { "epoch": 2.3976582827406765, "grad_norm": 0.44380372762680054, "learning_rate": 4.0301755404033086e-06, "loss": 0.0116, "step": 22116 }, { "epoch": 2.3977666955767565, "grad_norm": 0.35470443964004517, "learning_rate": 4.029450166835921e-06, "loss": 0.0085, "step": 22117 }, { "epoch": 2.397875108412836, "grad_norm": 5.101950168609619, "learning_rate": 4.028724793268534e-06, "loss": 0.0186, "step": 22118 }, { "epoch": 2.397983521248916, "grad_norm": 0.524870753288269, "learning_rate": 4.027999419701147e-06, "loss": 0.0116, "step": 22119 }, { "epoch": 2.3980919340849955, "grad_norm": 4.9482903480529785, "learning_rate": 4.027274046133759e-06, "loss": 0.0091, "step": 22120 }, { "epoch": 2.3982003469210755, "grad_norm": 0.31212708353996277, "learning_rate": 4.026548672566372e-06, "loss": 0.0104, "step": 22121 }, { "epoch": 2.398308759757155, "grad_norm": 6.550745010375977, "learning_rate": 4.025823298998985e-06, "loss": 0.0525, "step": 22122 }, { "epoch": 2.398417172593235, "grad_norm": 0.03400200977921486, "learning_rate": 4.025097925431598e-06, "loss": 0.0002, "step": 22123 }, { "epoch": 2.398525585429315, "grad_norm": 33.35018539428711, "learning_rate": 4.02437255186421e-06, "loss": 0.0295, "step": 22124 }, { "epoch": 2.3986339982653946, "grad_norm": 9.162666320800781, "learning_rate": 4.023647178296824e-06, "loss": 0.0238, "step": 22125 }, { "epoch": 2.3987424111014746, "grad_norm": 0.2683402895927429, "learning_rate": 4.022921804729436e-06, "loss": 0.006, "step": 22126 }, { "epoch": 2.398850823937554, "grad_norm": 2.4631030559539795, "learning_rate": 4.022196431162049e-06, "loss": 0.0454, "step": 22127 }, { "epoch": 2.398959236773634, "grad_norm": 0.28085070848464966, "learning_rate": 4.021471057594661e-06, "loss": 0.0067, "step": 22128 }, { "epoch": 2.3990676496097136, "grad_norm": 0.045105062425136566, "learning_rate": 4.020745684027275e-06, "loss": 0.0003, "step": 22129 }, { "epoch": 2.3991760624457936, "grad_norm": 0.7739133238792419, "learning_rate": 4.020020310459887e-06, "loss": 0.0029, "step": 22130 }, { "epoch": 2.399284475281873, "grad_norm": 0.5432129502296448, "learning_rate": 4.0192949368925e-06, "loss": 0.0215, "step": 22131 }, { "epoch": 2.399392888117953, "grad_norm": 0.459739089012146, "learning_rate": 4.018569563325113e-06, "loss": 0.0075, "step": 22132 }, { "epoch": 2.399501300954033, "grad_norm": 0.44158118963241577, "learning_rate": 4.017844189757726e-06, "loss": 0.0071, "step": 22133 }, { "epoch": 2.3996097137901127, "grad_norm": 0.5031440258026123, "learning_rate": 4.017118816190338e-06, "loss": 0.0238, "step": 22134 }, { "epoch": 2.3997181266261927, "grad_norm": 0.22195424139499664, "learning_rate": 4.016393442622951e-06, "loss": 0.0005, "step": 22135 }, { "epoch": 2.399826539462272, "grad_norm": 0.2594869136810303, "learning_rate": 4.015668069055564e-06, "loss": 0.0012, "step": 22136 }, { "epoch": 2.399934952298352, "grad_norm": 0.5402153730392456, "learning_rate": 4.014942695488176e-06, "loss": 0.0132, "step": 22137 }, { "epoch": 2.4000433651344317, "grad_norm": 2.985050678253174, "learning_rate": 4.01421732192079e-06, "loss": 0.0204, "step": 22138 }, { "epoch": 2.4001517779705117, "grad_norm": 4.819338798522949, "learning_rate": 4.013491948353402e-06, "loss": 0.0126, "step": 22139 }, { "epoch": 2.4002601908065913, "grad_norm": 0.5287173986434937, "learning_rate": 4.012766574786015e-06, "loss": 0.0144, "step": 22140 }, { "epoch": 2.4003686036426712, "grad_norm": 3.4083058834075928, "learning_rate": 4.012041201218627e-06, "loss": 0.0284, "step": 22141 }, { "epoch": 2.4004770164787512, "grad_norm": 2.4413130283355713, "learning_rate": 4.011315827651241e-06, "loss": 0.0035, "step": 22142 }, { "epoch": 2.4005854293148308, "grad_norm": 0.09260539710521698, "learning_rate": 4.010590454083853e-06, "loss": 0.0005, "step": 22143 }, { "epoch": 2.4006938421509108, "grad_norm": 0.1980893462896347, "learning_rate": 4.009865080516466e-06, "loss": 0.0015, "step": 22144 }, { "epoch": 2.4008022549869903, "grad_norm": 0.41836944222450256, "learning_rate": 4.009139706949079e-06, "loss": 0.0114, "step": 22145 }, { "epoch": 2.4009106678230703, "grad_norm": 0.18281790614128113, "learning_rate": 4.008414333381692e-06, "loss": 0.0002, "step": 22146 }, { "epoch": 2.4010190806591503, "grad_norm": 0.07117577642202377, "learning_rate": 4.0076889598143045e-06, "loss": 0.0004, "step": 22147 }, { "epoch": 2.40112749349523, "grad_norm": 13.482060432434082, "learning_rate": 4.0069635862469175e-06, "loss": 0.0575, "step": 22148 }, { "epoch": 2.40123590633131, "grad_norm": 0.03950149193406105, "learning_rate": 4.0062382126795305e-06, "loss": 0.0003, "step": 22149 }, { "epoch": 2.4013443191673893, "grad_norm": 0.307344913482666, "learning_rate": 4.005512839112143e-06, "loss": 0.0103, "step": 22150 }, { "epoch": 2.4014527320034693, "grad_norm": 0.12473554164171219, "learning_rate": 4.0047874655447565e-06, "loss": 0.0006, "step": 22151 }, { "epoch": 2.401561144839549, "grad_norm": 0.4819739758968353, "learning_rate": 4.004062091977369e-06, "loss": 0.0068, "step": 22152 }, { "epoch": 2.401669557675629, "grad_norm": 4.722630500793457, "learning_rate": 4.003336718409982e-06, "loss": 0.0073, "step": 22153 }, { "epoch": 2.4017779705117084, "grad_norm": 2.2090890407562256, "learning_rate": 4.002611344842594e-06, "loss": 0.0135, "step": 22154 }, { "epoch": 2.4018863833477884, "grad_norm": 0.5172630548477173, "learning_rate": 4.0018859712752076e-06, "loss": 0.012, "step": 22155 }, { "epoch": 2.4019947961838684, "grad_norm": 4.920921802520752, "learning_rate": 4.00116059770782e-06, "loss": 0.0085, "step": 22156 }, { "epoch": 2.402103209019948, "grad_norm": 0.35367992520332336, "learning_rate": 4.000435224140433e-06, "loss": 0.0084, "step": 22157 }, { "epoch": 2.402211621856028, "grad_norm": 0.8006511926651001, "learning_rate": 3.999709850573046e-06, "loss": 0.025, "step": 22158 }, { "epoch": 2.4023200346921074, "grad_norm": 0.2345573455095291, "learning_rate": 3.998984477005659e-06, "loss": 0.0009, "step": 22159 }, { "epoch": 2.4024284475281874, "grad_norm": 2.9553346633911133, "learning_rate": 3.998259103438271e-06, "loss": 0.0206, "step": 22160 }, { "epoch": 2.402536860364267, "grad_norm": 0.11966150999069214, "learning_rate": 3.997533729870884e-06, "loss": 0.0005, "step": 22161 }, { "epoch": 2.402645273200347, "grad_norm": 2.134538173675537, "learning_rate": 3.996808356303497e-06, "loss": 0.0026, "step": 22162 }, { "epoch": 2.4027536860364265, "grad_norm": 0.042763013392686844, "learning_rate": 3.996082982736109e-06, "loss": 0.0004, "step": 22163 }, { "epoch": 2.4028620988725065, "grad_norm": 4.724079608917236, "learning_rate": 3.995357609168723e-06, "loss": 0.0064, "step": 22164 }, { "epoch": 2.4029705117085864, "grad_norm": 1.5816465616226196, "learning_rate": 3.994632235601335e-06, "loss": 0.0203, "step": 22165 }, { "epoch": 2.403078924544666, "grad_norm": 2.4051246643066406, "learning_rate": 3.993906862033948e-06, "loss": 0.0121, "step": 22166 }, { "epoch": 2.403187337380746, "grad_norm": 3.084043264389038, "learning_rate": 3.99318148846656e-06, "loss": 0.0111, "step": 22167 }, { "epoch": 2.4032957502168255, "grad_norm": 0.2910090386867523, "learning_rate": 3.992456114899174e-06, "loss": 0.0101, "step": 22168 }, { "epoch": 2.4034041630529055, "grad_norm": 0.6067379117012024, "learning_rate": 3.991730741331786e-06, "loss": 0.0025, "step": 22169 }, { "epoch": 2.4035125758889855, "grad_norm": 1.4725910425186157, "learning_rate": 3.991005367764399e-06, "loss": 0.045, "step": 22170 }, { "epoch": 2.403620988725065, "grad_norm": 0.025799179449677467, "learning_rate": 3.990279994197012e-06, "loss": 0.0002, "step": 22171 }, { "epoch": 2.403729401561145, "grad_norm": 0.003078545443713665, "learning_rate": 3.989554620629625e-06, "loss": 0.0, "step": 22172 }, { "epoch": 2.4038378143972245, "grad_norm": 0.20928585529327393, "learning_rate": 3.988829247062237e-06, "loss": 0.0012, "step": 22173 }, { "epoch": 2.4039462272333045, "grad_norm": 2.770969867706299, "learning_rate": 3.98810387349485e-06, "loss": 0.0343, "step": 22174 }, { "epoch": 2.404054640069384, "grad_norm": 0.3214908838272095, "learning_rate": 3.987378499927463e-06, "loss": 0.0091, "step": 22175 }, { "epoch": 2.404163052905464, "grad_norm": 1.7250643968582153, "learning_rate": 3.986653126360075e-06, "loss": 0.0345, "step": 22176 }, { "epoch": 2.4042714657415436, "grad_norm": 0.29221081733703613, "learning_rate": 3.985927752792689e-06, "loss": 0.0062, "step": 22177 }, { "epoch": 2.4043798785776236, "grad_norm": 0.6900039315223694, "learning_rate": 3.985202379225301e-06, "loss": 0.002, "step": 22178 }, { "epoch": 2.4044882914137036, "grad_norm": 0.25761523842811584, "learning_rate": 3.984477005657914e-06, "loss": 0.0056, "step": 22179 }, { "epoch": 2.404596704249783, "grad_norm": 0.28611141443252563, "learning_rate": 3.983751632090526e-06, "loss": 0.0009, "step": 22180 }, { "epoch": 2.404705117085863, "grad_norm": 0.19414712488651276, "learning_rate": 3.98302625852314e-06, "loss": 0.0003, "step": 22181 }, { "epoch": 2.4048135299219426, "grad_norm": 0.2610871195793152, "learning_rate": 3.982300884955752e-06, "loss": 0.0058, "step": 22182 }, { "epoch": 2.4049219427580226, "grad_norm": 1.1653292179107666, "learning_rate": 3.981575511388365e-06, "loss": 0.0212, "step": 22183 }, { "epoch": 2.405030355594102, "grad_norm": 15.284582138061523, "learning_rate": 3.980850137820978e-06, "loss": 0.0176, "step": 22184 }, { "epoch": 2.405138768430182, "grad_norm": 1.6152499914169312, "learning_rate": 3.980124764253591e-06, "loss": 0.0215, "step": 22185 }, { "epoch": 2.4052471812662617, "grad_norm": 0.587763249874115, "learning_rate": 3.9793993906862035e-06, "loss": 0.024, "step": 22186 }, { "epoch": 2.4053555941023417, "grad_norm": 2.696316957473755, "learning_rate": 3.9786740171188165e-06, "loss": 0.0096, "step": 22187 }, { "epoch": 2.4054640069384217, "grad_norm": 0.005174139980226755, "learning_rate": 3.9779486435514295e-06, "loss": 0.0001, "step": 22188 }, { "epoch": 2.405572419774501, "grad_norm": 5.6952409744262695, "learning_rate": 3.977223269984042e-06, "loss": 0.0469, "step": 22189 }, { "epoch": 2.405680832610581, "grad_norm": 2.6119563579559326, "learning_rate": 3.9764978964166555e-06, "loss": 0.032, "step": 22190 }, { "epoch": 2.4057892454466607, "grad_norm": 0.7873996496200562, "learning_rate": 3.975772522849268e-06, "loss": 0.023, "step": 22191 }, { "epoch": 2.4058976582827407, "grad_norm": 0.6038787364959717, "learning_rate": 3.975047149281881e-06, "loss": 0.0156, "step": 22192 }, { "epoch": 2.4060060711188207, "grad_norm": 0.1977546364068985, "learning_rate": 3.974321775714493e-06, "loss": 0.0119, "step": 22193 }, { "epoch": 2.4061144839549002, "grad_norm": 0.8693003058433533, "learning_rate": 3.9735964021471066e-06, "loss": 0.0093, "step": 22194 }, { "epoch": 2.4062228967909802, "grad_norm": 0.08529335260391235, "learning_rate": 3.972871028579719e-06, "loss": 0.0002, "step": 22195 }, { "epoch": 2.4063313096270598, "grad_norm": 1.7152769565582275, "learning_rate": 3.972145655012332e-06, "loss": 0.0018, "step": 22196 }, { "epoch": 2.4064397224631398, "grad_norm": 3.180475950241089, "learning_rate": 3.971420281444945e-06, "loss": 0.0043, "step": 22197 }, { "epoch": 2.4065481352992193, "grad_norm": 10.255881309509277, "learning_rate": 3.970694907877558e-06, "loss": 0.0514, "step": 22198 }, { "epoch": 2.4066565481352993, "grad_norm": 0.3183547556400299, "learning_rate": 3.96996953431017e-06, "loss": 0.0088, "step": 22199 }, { "epoch": 2.406764960971379, "grad_norm": 0.4400164484977722, "learning_rate": 3.969244160742783e-06, "loss": 0.0164, "step": 22200 }, { "epoch": 2.406873373807459, "grad_norm": 0.2751586437225342, "learning_rate": 3.968518787175396e-06, "loss": 0.0062, "step": 22201 }, { "epoch": 2.406981786643539, "grad_norm": 0.07596290111541748, "learning_rate": 3.967793413608008e-06, "loss": 0.0005, "step": 22202 }, { "epoch": 2.4070901994796183, "grad_norm": 0.026587052270770073, "learning_rate": 3.967068040040622e-06, "loss": 0.0002, "step": 22203 }, { "epoch": 2.4071986123156983, "grad_norm": 8.069404602050781, "learning_rate": 3.966342666473234e-06, "loss": 0.0133, "step": 22204 }, { "epoch": 2.407307025151778, "grad_norm": 0.06124461442232132, "learning_rate": 3.965617292905847e-06, "loss": 0.0006, "step": 22205 }, { "epoch": 2.407415437987858, "grad_norm": 3.487590789794922, "learning_rate": 3.964891919338459e-06, "loss": 0.0386, "step": 22206 }, { "epoch": 2.4075238508239374, "grad_norm": 7.271755218505859, "learning_rate": 3.964166545771073e-06, "loss": 0.0744, "step": 22207 }, { "epoch": 2.4076322636600174, "grad_norm": 14.491471290588379, "learning_rate": 3.963441172203685e-06, "loss": 0.0159, "step": 22208 }, { "epoch": 2.407740676496097, "grad_norm": 0.027786338701844215, "learning_rate": 3.962715798636298e-06, "loss": 0.0002, "step": 22209 }, { "epoch": 2.407849089332177, "grad_norm": 7.976202011108398, "learning_rate": 3.961990425068911e-06, "loss": 0.024, "step": 22210 }, { "epoch": 2.407957502168257, "grad_norm": 0.02877216972410679, "learning_rate": 3.961265051501524e-06, "loss": 0.0002, "step": 22211 }, { "epoch": 2.4080659150043364, "grad_norm": 0.030994458124041557, "learning_rate": 3.960539677934136e-06, "loss": 0.0003, "step": 22212 }, { "epoch": 2.4081743278404164, "grad_norm": 3.2776038646698, "learning_rate": 3.959814304366749e-06, "loss": 0.0289, "step": 22213 }, { "epoch": 2.408282740676496, "grad_norm": 0.20453877747058868, "learning_rate": 3.959088930799362e-06, "loss": 0.0011, "step": 22214 }, { "epoch": 2.408391153512576, "grad_norm": 0.2532554566860199, "learning_rate": 3.958363557231974e-06, "loss": 0.0009, "step": 22215 }, { "epoch": 2.408499566348656, "grad_norm": 0.4698158800601959, "learning_rate": 3.957638183664588e-06, "loss": 0.0238, "step": 22216 }, { "epoch": 2.4086079791847355, "grad_norm": 0.4547037184238434, "learning_rate": 3.9569128100972e-06, "loss": 0.0154, "step": 22217 }, { "epoch": 2.4087163920208154, "grad_norm": 2.8427369594573975, "learning_rate": 3.956187436529813e-06, "loss": 0.0387, "step": 22218 }, { "epoch": 2.408824804856895, "grad_norm": 0.33830881118774414, "learning_rate": 3.955462062962425e-06, "loss": 0.0098, "step": 22219 }, { "epoch": 2.408933217692975, "grad_norm": 0.3722172975540161, "learning_rate": 3.954736689395039e-06, "loss": 0.0013, "step": 22220 }, { "epoch": 2.4090416305290545, "grad_norm": 7.972840785980225, "learning_rate": 3.954011315827651e-06, "loss": 0.0049, "step": 22221 }, { "epoch": 2.4091500433651345, "grad_norm": 1.0689566135406494, "learning_rate": 3.953285942260264e-06, "loss": 0.0078, "step": 22222 }, { "epoch": 2.409258456201214, "grad_norm": 0.5144177079200745, "learning_rate": 3.952560568692877e-06, "loss": 0.032, "step": 22223 }, { "epoch": 2.409366869037294, "grad_norm": 0.29295361042022705, "learning_rate": 3.95183519512549e-06, "loss": 0.0081, "step": 22224 }, { "epoch": 2.409475281873374, "grad_norm": 0.27924177050590515, "learning_rate": 3.9511098215581025e-06, "loss": 0.0062, "step": 22225 }, { "epoch": 2.4095836947094535, "grad_norm": 0.04975214600563049, "learning_rate": 3.9503844479907155e-06, "loss": 0.0005, "step": 22226 }, { "epoch": 2.4096921075455335, "grad_norm": 0.4792538583278656, "learning_rate": 3.9496590744233285e-06, "loss": 0.0117, "step": 22227 }, { "epoch": 2.409800520381613, "grad_norm": 3.6715688705444336, "learning_rate": 3.9489337008559415e-06, "loss": 0.0257, "step": 22228 }, { "epoch": 2.409908933217693, "grad_norm": 0.5052276849746704, "learning_rate": 3.9482083272885545e-06, "loss": 0.0073, "step": 22229 }, { "epoch": 2.4100173460537726, "grad_norm": 0.39370378851890564, "learning_rate": 3.947482953721167e-06, "loss": 0.0054, "step": 22230 }, { "epoch": 2.4101257588898526, "grad_norm": 0.5428100228309631, "learning_rate": 3.94675758015378e-06, "loss": 0.0285, "step": 22231 }, { "epoch": 2.410234171725932, "grad_norm": 4.442086219787598, "learning_rate": 3.946032206586392e-06, "loss": 0.0632, "step": 22232 }, { "epoch": 2.410342584562012, "grad_norm": 0.5834507346153259, "learning_rate": 3.9453068330190056e-06, "loss": 0.0034, "step": 22233 }, { "epoch": 2.410450997398092, "grad_norm": 5.10817289352417, "learning_rate": 3.944581459451618e-06, "loss": 0.0126, "step": 22234 }, { "epoch": 2.4105594102341716, "grad_norm": 0.3118027448654175, "learning_rate": 3.943856085884231e-06, "loss": 0.0108, "step": 22235 }, { "epoch": 2.4106678230702516, "grad_norm": 0.5187774896621704, "learning_rate": 3.943130712316844e-06, "loss": 0.0146, "step": 22236 }, { "epoch": 2.410776235906331, "grad_norm": 0.1702585220336914, "learning_rate": 3.942405338749457e-06, "loss": 0.0169, "step": 22237 }, { "epoch": 2.410884648742411, "grad_norm": 0.25559213757514954, "learning_rate": 3.941679965182069e-06, "loss": 0.0126, "step": 22238 }, { "epoch": 2.410993061578491, "grad_norm": 12.84177017211914, "learning_rate": 3.940954591614682e-06, "loss": 0.0327, "step": 22239 }, { "epoch": 2.4111014744145707, "grad_norm": 9.034550666809082, "learning_rate": 3.940229218047295e-06, "loss": 0.0093, "step": 22240 }, { "epoch": 2.4112098872506507, "grad_norm": 0.4085535705089569, "learning_rate": 3.939503844479908e-06, "loss": 0.0024, "step": 22241 }, { "epoch": 2.41131830008673, "grad_norm": 0.30478283762931824, "learning_rate": 3.938778470912521e-06, "loss": 0.0102, "step": 22242 }, { "epoch": 2.41142671292281, "grad_norm": 6.733109951019287, "learning_rate": 3.938053097345133e-06, "loss": 0.0109, "step": 22243 }, { "epoch": 2.4115351257588897, "grad_norm": 0.3612305223941803, "learning_rate": 3.937327723777746e-06, "loss": 0.0124, "step": 22244 }, { "epoch": 2.4116435385949697, "grad_norm": 0.054144859313964844, "learning_rate": 3.936602350210358e-06, "loss": 0.0003, "step": 22245 }, { "epoch": 2.4117519514310493, "grad_norm": 0.021831190213561058, "learning_rate": 3.935876976642972e-06, "loss": 0.0002, "step": 22246 }, { "epoch": 2.4118603642671292, "grad_norm": 0.27183306217193604, "learning_rate": 3.935151603075584e-06, "loss": 0.0072, "step": 22247 }, { "epoch": 2.4119687771032092, "grad_norm": 0.026445159688591957, "learning_rate": 3.934426229508197e-06, "loss": 0.0002, "step": 22248 }, { "epoch": 2.4120771899392888, "grad_norm": 0.36174291372299194, "learning_rate": 3.93370085594081e-06, "loss": 0.0119, "step": 22249 }, { "epoch": 2.4121856027753688, "grad_norm": 0.02769686095416546, "learning_rate": 3.932975482373423e-06, "loss": 0.0002, "step": 22250 }, { "epoch": 2.4122940156114483, "grad_norm": 6.315385341644287, "learning_rate": 3.932250108806035e-06, "loss": 0.0159, "step": 22251 }, { "epoch": 2.4124024284475283, "grad_norm": 6.281881809234619, "learning_rate": 3.931524735238648e-06, "loss": 0.0323, "step": 22252 }, { "epoch": 2.412510841283608, "grad_norm": 0.17481087148189545, "learning_rate": 3.930799361671261e-06, "loss": 0.0003, "step": 22253 }, { "epoch": 2.412619254119688, "grad_norm": 1.748573899269104, "learning_rate": 3.930073988103874e-06, "loss": 0.0026, "step": 22254 }, { "epoch": 2.4127276669557673, "grad_norm": 2.318024158477783, "learning_rate": 3.929348614536487e-06, "loss": 0.0011, "step": 22255 }, { "epoch": 2.4128360797918473, "grad_norm": 0.24834689497947693, "learning_rate": 3.928623240969099e-06, "loss": 0.0052, "step": 22256 }, { "epoch": 2.4129444926279273, "grad_norm": 0.025898482650518417, "learning_rate": 3.927897867401712e-06, "loss": 0.0002, "step": 22257 }, { "epoch": 2.413052905464007, "grad_norm": 0.022347137331962585, "learning_rate": 3.927172493834324e-06, "loss": 0.0002, "step": 22258 }, { "epoch": 2.413161318300087, "grad_norm": 0.3532410264015198, "learning_rate": 3.926447120266938e-06, "loss": 0.0119, "step": 22259 }, { "epoch": 2.4132697311361664, "grad_norm": 6.060338973999023, "learning_rate": 3.92572174669955e-06, "loss": 0.0124, "step": 22260 }, { "epoch": 2.4133781439722464, "grad_norm": 1.3488515615463257, "learning_rate": 3.924996373132163e-06, "loss": 0.0072, "step": 22261 }, { "epoch": 2.4134865568083264, "grad_norm": 0.4255391061306, "learning_rate": 3.924270999564776e-06, "loss": 0.0062, "step": 22262 }, { "epoch": 2.413594969644406, "grad_norm": 0.25270840525627136, "learning_rate": 3.923545625997389e-06, "loss": 0.0055, "step": 22263 }, { "epoch": 2.413703382480486, "grad_norm": 0.2794405221939087, "learning_rate": 3.9228202524300015e-06, "loss": 0.0112, "step": 22264 }, { "epoch": 2.4138117953165654, "grad_norm": 0.5320809483528137, "learning_rate": 3.9220948788626145e-06, "loss": 0.0164, "step": 22265 }, { "epoch": 2.4139202081526454, "grad_norm": 0.2147904485464096, "learning_rate": 3.9213695052952275e-06, "loss": 0.0005, "step": 22266 }, { "epoch": 2.414028620988725, "grad_norm": 0.20327827334403992, "learning_rate": 3.9206441317278405e-06, "loss": 0.0008, "step": 22267 }, { "epoch": 2.414137033824805, "grad_norm": 1.367196798324585, "learning_rate": 3.9199187581604535e-06, "loss": 0.0617, "step": 22268 }, { "epoch": 2.4142454466608845, "grad_norm": 2.688295364379883, "learning_rate": 3.919193384593066e-06, "loss": 0.0028, "step": 22269 }, { "epoch": 2.4143538594969645, "grad_norm": 5.345158576965332, "learning_rate": 3.918468011025679e-06, "loss": 0.0285, "step": 22270 }, { "epoch": 2.4144622723330444, "grad_norm": 0.1384878307580948, "learning_rate": 3.917742637458291e-06, "loss": 0.0004, "step": 22271 }, { "epoch": 2.414570685169124, "grad_norm": 0.18931636214256287, "learning_rate": 3.917017263890905e-06, "loss": 0.0008, "step": 22272 }, { "epoch": 2.414679098005204, "grad_norm": 0.24427440762519836, "learning_rate": 3.916291890323517e-06, "loss": 0.0046, "step": 22273 }, { "epoch": 2.4147875108412835, "grad_norm": 0.8123215436935425, "learning_rate": 3.91556651675613e-06, "loss": 0.0174, "step": 22274 }, { "epoch": 2.4148959236773635, "grad_norm": 6.331643104553223, "learning_rate": 3.914841143188743e-06, "loss": 0.0214, "step": 22275 }, { "epoch": 2.415004336513443, "grad_norm": 0.31362655758857727, "learning_rate": 3.914115769621356e-06, "loss": 0.0012, "step": 22276 }, { "epoch": 2.415112749349523, "grad_norm": 0.5530914664268494, "learning_rate": 3.913390396053968e-06, "loss": 0.0179, "step": 22277 }, { "epoch": 2.4152211621856026, "grad_norm": 0.3644321858882904, "learning_rate": 3.912665022486581e-06, "loss": 0.0127, "step": 22278 }, { "epoch": 2.4153295750216826, "grad_norm": 0.37446123361587524, "learning_rate": 3.911939648919194e-06, "loss": 0.0012, "step": 22279 }, { "epoch": 2.4154379878577625, "grad_norm": 1.6258207559585571, "learning_rate": 3.911214275351807e-06, "loss": 0.0198, "step": 22280 }, { "epoch": 2.415546400693842, "grad_norm": 5.916715621948242, "learning_rate": 3.91048890178442e-06, "loss": 0.0367, "step": 22281 }, { "epoch": 2.415654813529922, "grad_norm": 0.31395578384399414, "learning_rate": 3.909763528217032e-06, "loss": 0.0196, "step": 22282 }, { "epoch": 2.4157632263660016, "grad_norm": 1.1953991651535034, "learning_rate": 3.909038154649645e-06, "loss": 0.0178, "step": 22283 }, { "epoch": 2.4158716392020816, "grad_norm": 0.7406049966812134, "learning_rate": 3.908312781082257e-06, "loss": 0.0077, "step": 22284 }, { "epoch": 2.415980052038161, "grad_norm": 0.37474215030670166, "learning_rate": 3.907587407514871e-06, "loss": 0.0007, "step": 22285 }, { "epoch": 2.416088464874241, "grad_norm": 3.1666135787963867, "learning_rate": 3.906862033947483e-06, "loss": 0.0064, "step": 22286 }, { "epoch": 2.4161968777103207, "grad_norm": 0.2613852620124817, "learning_rate": 3.906136660380096e-06, "loss": 0.0055, "step": 22287 }, { "epoch": 2.4163052905464006, "grad_norm": 0.033179257065057755, "learning_rate": 3.905411286812709e-06, "loss": 0.0002, "step": 22288 }, { "epoch": 2.4164137033824806, "grad_norm": 2.9745919704437256, "learning_rate": 3.904685913245322e-06, "loss": 0.0231, "step": 22289 }, { "epoch": 2.41652211621856, "grad_norm": 0.25551125407218933, "learning_rate": 3.903960539677934e-06, "loss": 0.005, "step": 22290 }, { "epoch": 2.41663052905464, "grad_norm": 0.35086339712142944, "learning_rate": 3.903235166110547e-06, "loss": 0.0214, "step": 22291 }, { "epoch": 2.4167389418907197, "grad_norm": 11.968118667602539, "learning_rate": 3.90250979254316e-06, "loss": 0.0381, "step": 22292 }, { "epoch": 2.4168473547267997, "grad_norm": 0.07309329509735107, "learning_rate": 3.901784418975773e-06, "loss": 0.0002, "step": 22293 }, { "epoch": 2.4169557675628797, "grad_norm": 0.25916993618011475, "learning_rate": 3.901059045408386e-06, "loss": 0.0042, "step": 22294 }, { "epoch": 2.417064180398959, "grad_norm": 3.887526035308838, "learning_rate": 3.900333671840998e-06, "loss": 0.0377, "step": 22295 }, { "epoch": 2.417172593235039, "grad_norm": 0.018874455243349075, "learning_rate": 3.899608298273611e-06, "loss": 0.0001, "step": 22296 }, { "epoch": 2.4172810060711187, "grad_norm": 0.025649087503552437, "learning_rate": 3.898882924706223e-06, "loss": 0.0002, "step": 22297 }, { "epoch": 2.4173894189071987, "grad_norm": 0.43943339586257935, "learning_rate": 3.898157551138837e-06, "loss": 0.0062, "step": 22298 }, { "epoch": 2.4174978317432783, "grad_norm": 0.5681672096252441, "learning_rate": 3.897432177571449e-06, "loss": 0.0201, "step": 22299 }, { "epoch": 2.4176062445793582, "grad_norm": 4.266138076782227, "learning_rate": 3.896706804004062e-06, "loss": 0.0177, "step": 22300 }, { "epoch": 2.417714657415438, "grad_norm": 0.3185221254825592, "learning_rate": 3.895981430436675e-06, "loss": 0.0107, "step": 22301 }, { "epoch": 2.4178230702515178, "grad_norm": 0.342083215713501, "learning_rate": 3.895256056869288e-06, "loss": 0.02, "step": 22302 }, { "epoch": 2.4179314830875978, "grad_norm": 10.646883010864258, "learning_rate": 3.8945306833019005e-06, "loss": 0.088, "step": 22303 }, { "epoch": 2.4180398959236773, "grad_norm": 0.5726080536842346, "learning_rate": 3.8938053097345135e-06, "loss": 0.0055, "step": 22304 }, { "epoch": 2.4181483087597573, "grad_norm": 0.3041679263114929, "learning_rate": 3.8930799361671265e-06, "loss": 0.0066, "step": 22305 }, { "epoch": 2.418256721595837, "grad_norm": 0.074103944003582, "learning_rate": 3.8923545625997395e-06, "loss": 0.0003, "step": 22306 }, { "epoch": 2.418365134431917, "grad_norm": 0.031902510672807693, "learning_rate": 3.8916291890323525e-06, "loss": 0.0001, "step": 22307 }, { "epoch": 2.4184735472679963, "grad_norm": 12.613115310668945, "learning_rate": 3.890903815464965e-06, "loss": 0.0404, "step": 22308 }, { "epoch": 2.4185819601040763, "grad_norm": 0.22921225428581238, "learning_rate": 3.890178441897578e-06, "loss": 0.0046, "step": 22309 }, { "epoch": 2.418690372940156, "grad_norm": 6.12469482421875, "learning_rate": 3.88945306833019e-06, "loss": 0.0219, "step": 22310 }, { "epoch": 2.418798785776236, "grad_norm": 0.014201942831277847, "learning_rate": 3.888727694762804e-06, "loss": 0.0001, "step": 22311 }, { "epoch": 2.418907198612316, "grad_norm": 0.310506671667099, "learning_rate": 3.888002321195416e-06, "loss": 0.0071, "step": 22312 }, { "epoch": 2.4190156114483954, "grad_norm": 0.6109163165092468, "learning_rate": 3.887276947628029e-06, "loss": 0.0097, "step": 22313 }, { "epoch": 2.4191240242844754, "grad_norm": 1.608110785484314, "learning_rate": 3.886551574060642e-06, "loss": 0.0257, "step": 22314 }, { "epoch": 2.419232437120555, "grad_norm": 1.0093778371810913, "learning_rate": 3.885826200493255e-06, "loss": 0.0016, "step": 22315 }, { "epoch": 2.419340849956635, "grad_norm": 0.0326215885579586, "learning_rate": 3.885100826925867e-06, "loss": 0.0003, "step": 22316 }, { "epoch": 2.419449262792715, "grad_norm": 0.006141590885818005, "learning_rate": 3.88437545335848e-06, "loss": 0.0001, "step": 22317 }, { "epoch": 2.4195576756287944, "grad_norm": 0.24911916255950928, "learning_rate": 3.883650079791093e-06, "loss": 0.0008, "step": 22318 }, { "epoch": 2.4196660884648744, "grad_norm": 0.9554508924484253, "learning_rate": 3.882924706223706e-06, "loss": 0.0114, "step": 22319 }, { "epoch": 2.419774501300954, "grad_norm": 0.30546170473098755, "learning_rate": 3.882199332656319e-06, "loss": 0.0071, "step": 22320 }, { "epoch": 2.419882914137034, "grad_norm": 0.02210344932973385, "learning_rate": 3.881473959088931e-06, "loss": 0.0001, "step": 22321 }, { "epoch": 2.4199913269731135, "grad_norm": 1.4511659145355225, "learning_rate": 3.880748585521544e-06, "loss": 0.0227, "step": 22322 }, { "epoch": 2.4200997398091935, "grad_norm": 1.1189244985580444, "learning_rate": 3.880023211954156e-06, "loss": 0.0069, "step": 22323 }, { "epoch": 2.420208152645273, "grad_norm": 3.306259870529175, "learning_rate": 3.87929783838677e-06, "loss": 0.0176, "step": 22324 }, { "epoch": 2.420316565481353, "grad_norm": 4.705131530761719, "learning_rate": 3.878572464819382e-06, "loss": 0.0386, "step": 22325 }, { "epoch": 2.420424978317433, "grad_norm": 0.43733495473861694, "learning_rate": 3.877847091251995e-06, "loss": 0.0106, "step": 22326 }, { "epoch": 2.4205333911535125, "grad_norm": 0.48668235540390015, "learning_rate": 3.877121717684608e-06, "loss": 0.0061, "step": 22327 }, { "epoch": 2.4206418039895925, "grad_norm": 0.22925205528736115, "learning_rate": 3.876396344117221e-06, "loss": 0.0045, "step": 22328 }, { "epoch": 2.420750216825672, "grad_norm": 4.061736583709717, "learning_rate": 3.875670970549833e-06, "loss": 0.0162, "step": 22329 }, { "epoch": 2.420858629661752, "grad_norm": 0.3327421545982361, "learning_rate": 3.874945596982446e-06, "loss": 0.0081, "step": 22330 }, { "epoch": 2.4209670424978316, "grad_norm": 0.5886279344558716, "learning_rate": 3.874220223415059e-06, "loss": 0.0186, "step": 22331 }, { "epoch": 2.4210754553339116, "grad_norm": 4.897385120391846, "learning_rate": 3.873494849847672e-06, "loss": 0.14, "step": 22332 }, { "epoch": 2.421183868169991, "grad_norm": 3.4008378982543945, "learning_rate": 3.872769476280285e-06, "loss": 0.0714, "step": 22333 }, { "epoch": 2.421292281006071, "grad_norm": 12.65810489654541, "learning_rate": 3.872044102712897e-06, "loss": 0.0118, "step": 22334 }, { "epoch": 2.421400693842151, "grad_norm": 0.30077338218688965, "learning_rate": 3.87131872914551e-06, "loss": 0.0069, "step": 22335 }, { "epoch": 2.4215091066782306, "grad_norm": 0.07297509908676147, "learning_rate": 3.870593355578123e-06, "loss": 0.0005, "step": 22336 }, { "epoch": 2.4216175195143106, "grad_norm": 3.9889516830444336, "learning_rate": 3.869867982010736e-06, "loss": 0.0285, "step": 22337 }, { "epoch": 2.42172593235039, "grad_norm": 3.2433176040649414, "learning_rate": 3.869142608443348e-06, "loss": 0.0129, "step": 22338 }, { "epoch": 2.42183434518647, "grad_norm": 0.46073117852211, "learning_rate": 3.868417234875961e-06, "loss": 0.0157, "step": 22339 }, { "epoch": 2.42194275802255, "grad_norm": 4.024775981903076, "learning_rate": 3.867691861308574e-06, "loss": 0.0028, "step": 22340 }, { "epoch": 2.4220511708586296, "grad_norm": 0.5879705548286438, "learning_rate": 3.866966487741187e-06, "loss": 0.0132, "step": 22341 }, { "epoch": 2.4221595836947096, "grad_norm": 0.5096485018730164, "learning_rate": 3.8662411141737995e-06, "loss": 0.0144, "step": 22342 }, { "epoch": 2.422267996530789, "grad_norm": 0.2871236205101013, "learning_rate": 3.8655157406064125e-06, "loss": 0.0067, "step": 22343 }, { "epoch": 2.422376409366869, "grad_norm": 0.019134225323796272, "learning_rate": 3.8647903670390255e-06, "loss": 0.0001, "step": 22344 }, { "epoch": 2.4224848222029487, "grad_norm": 0.054912250488996506, "learning_rate": 3.8640649934716385e-06, "loss": 0.0003, "step": 22345 }, { "epoch": 2.4225932350390287, "grad_norm": 0.0908312126994133, "learning_rate": 3.8633396199042515e-06, "loss": 0.0003, "step": 22346 }, { "epoch": 2.422701647875108, "grad_norm": 1.436048150062561, "learning_rate": 3.862614246336864e-06, "loss": 0.0153, "step": 22347 }, { "epoch": 2.422810060711188, "grad_norm": 0.28891944885253906, "learning_rate": 3.861888872769477e-06, "loss": 0.0175, "step": 22348 }, { "epoch": 2.422918473547268, "grad_norm": 0.4530794620513916, "learning_rate": 3.86116349920209e-06, "loss": 0.0108, "step": 22349 }, { "epoch": 2.4230268863833477, "grad_norm": 3.6668701171875, "learning_rate": 3.860438125634703e-06, "loss": 0.0198, "step": 22350 }, { "epoch": 2.4231352992194277, "grad_norm": 0.37917250394821167, "learning_rate": 3.859712752067315e-06, "loss": 0.0137, "step": 22351 }, { "epoch": 2.4232437120555073, "grad_norm": 7.8165669441223145, "learning_rate": 3.858987378499928e-06, "loss": 0.0513, "step": 22352 }, { "epoch": 2.4233521248915872, "grad_norm": 0.4699743092060089, "learning_rate": 3.858262004932541e-06, "loss": 0.0113, "step": 22353 }, { "epoch": 2.423460537727667, "grad_norm": 0.3004825711250305, "learning_rate": 3.857536631365154e-06, "loss": 0.0012, "step": 22354 }, { "epoch": 2.4235689505637468, "grad_norm": 0.23452070355415344, "learning_rate": 3.856811257797766e-06, "loss": 0.004, "step": 22355 }, { "epoch": 2.4236773633998263, "grad_norm": 0.41941744089126587, "learning_rate": 3.856085884230379e-06, "loss": 0.0084, "step": 22356 }, { "epoch": 2.4237857762359063, "grad_norm": 0.38502737879753113, "learning_rate": 3.855360510662992e-06, "loss": 0.0003, "step": 22357 }, { "epoch": 2.4238941890719863, "grad_norm": 0.033507056534290314, "learning_rate": 3.854635137095605e-06, "loss": 0.0003, "step": 22358 }, { "epoch": 2.424002601908066, "grad_norm": 0.06275113672018051, "learning_rate": 3.853909763528218e-06, "loss": 0.0003, "step": 22359 }, { "epoch": 2.424111014744146, "grad_norm": 2.737943649291992, "learning_rate": 3.85318438996083e-06, "loss": 0.0055, "step": 22360 }, { "epoch": 2.4242194275802254, "grad_norm": 0.031962864100933075, "learning_rate": 3.852459016393443e-06, "loss": 0.0002, "step": 22361 }, { "epoch": 2.4243278404163053, "grad_norm": 5.259678840637207, "learning_rate": 3.851733642826056e-06, "loss": 0.016, "step": 22362 }, { "epoch": 2.4244362532523853, "grad_norm": 0.049248140305280685, "learning_rate": 3.851008269258669e-06, "loss": 0.0003, "step": 22363 }, { "epoch": 2.424544666088465, "grad_norm": 0.008041427470743656, "learning_rate": 3.850282895691281e-06, "loss": 0.0001, "step": 22364 }, { "epoch": 2.424653078924545, "grad_norm": 3.187504529953003, "learning_rate": 3.849557522123894e-06, "loss": 0.0383, "step": 22365 }, { "epoch": 2.4247614917606244, "grad_norm": 5.421907424926758, "learning_rate": 3.848832148556507e-06, "loss": 0.0462, "step": 22366 }, { "epoch": 2.4248699045967044, "grad_norm": 0.48444169759750366, "learning_rate": 3.84810677498912e-06, "loss": 0.0129, "step": 22367 }, { "epoch": 2.424978317432784, "grad_norm": 0.40287700295448303, "learning_rate": 3.847381401421732e-06, "loss": 0.0016, "step": 22368 }, { "epoch": 2.425086730268864, "grad_norm": 0.42028918862342834, "learning_rate": 3.846656027854345e-06, "loss": 0.0192, "step": 22369 }, { "epoch": 2.4251951431049434, "grad_norm": 0.06405062973499298, "learning_rate": 3.845930654286958e-06, "loss": 0.0004, "step": 22370 }, { "epoch": 2.4253035559410234, "grad_norm": 0.19142283499240875, "learning_rate": 3.845205280719571e-06, "loss": 0.0007, "step": 22371 }, { "epoch": 2.4254119687771034, "grad_norm": 0.24416399002075195, "learning_rate": 3.844479907152184e-06, "loss": 0.0004, "step": 22372 }, { "epoch": 2.425520381613183, "grad_norm": 0.2924717962741852, "learning_rate": 3.843754533584796e-06, "loss": 0.0068, "step": 22373 }, { "epoch": 2.425628794449263, "grad_norm": 0.7977061867713928, "learning_rate": 3.843029160017409e-06, "loss": 0.0013, "step": 22374 }, { "epoch": 2.4257372072853425, "grad_norm": 0.23097744584083557, "learning_rate": 3.842303786450022e-06, "loss": 0.0009, "step": 22375 }, { "epoch": 2.4258456201214225, "grad_norm": 0.3018742799758911, "learning_rate": 3.841578412882635e-06, "loss": 0.0094, "step": 22376 }, { "epoch": 2.425954032957502, "grad_norm": 0.5009616613388062, "learning_rate": 3.840853039315247e-06, "loss": 0.0129, "step": 22377 }, { "epoch": 2.426062445793582, "grad_norm": 9.704023361206055, "learning_rate": 3.84012766574786e-06, "loss": 0.0073, "step": 22378 }, { "epoch": 2.4261708586296615, "grad_norm": 0.2652173936367035, "learning_rate": 3.839402292180473e-06, "loss": 0.0006, "step": 22379 }, { "epoch": 2.4262792714657415, "grad_norm": 0.3512474000453949, "learning_rate": 3.838676918613086e-06, "loss": 0.0118, "step": 22380 }, { "epoch": 2.4263876843018215, "grad_norm": 0.3348715305328369, "learning_rate": 3.8379515450456985e-06, "loss": 0.0205, "step": 22381 }, { "epoch": 2.426496097137901, "grad_norm": 3.618257999420166, "learning_rate": 3.8372261714783115e-06, "loss": 0.0211, "step": 22382 }, { "epoch": 2.426604509973981, "grad_norm": 0.11923248320817947, "learning_rate": 3.8365007979109245e-06, "loss": 0.0006, "step": 22383 }, { "epoch": 2.4267129228100606, "grad_norm": 3.730747938156128, "learning_rate": 3.8357754243435375e-06, "loss": 0.0091, "step": 22384 }, { "epoch": 2.4268213356461406, "grad_norm": 0.03275632858276367, "learning_rate": 3.8350500507761505e-06, "loss": 0.0002, "step": 22385 }, { "epoch": 2.4269297484822205, "grad_norm": 0.42900350689888, "learning_rate": 3.834324677208763e-06, "loss": 0.0095, "step": 22386 }, { "epoch": 2.4270381613183, "grad_norm": 0.3237597942352295, "learning_rate": 3.833599303641376e-06, "loss": 0.0072, "step": 22387 }, { "epoch": 2.42714657415438, "grad_norm": 3.78416109085083, "learning_rate": 3.832873930073989e-06, "loss": 0.0324, "step": 22388 }, { "epoch": 2.4272549869904596, "grad_norm": 0.5561227202415466, "learning_rate": 3.832148556506602e-06, "loss": 0.0068, "step": 22389 }, { "epoch": 2.4273633998265396, "grad_norm": 6.142908573150635, "learning_rate": 3.831423182939214e-06, "loss": 0.0374, "step": 22390 }, { "epoch": 2.427471812662619, "grad_norm": 0.6444147229194641, "learning_rate": 3.830697809371827e-06, "loss": 0.0009, "step": 22391 }, { "epoch": 2.427580225498699, "grad_norm": 0.14258310198783875, "learning_rate": 3.82997243580444e-06, "loss": 0.0006, "step": 22392 }, { "epoch": 2.4276886383347787, "grad_norm": 0.31710878014564514, "learning_rate": 3.829247062237053e-06, "loss": 0.0145, "step": 22393 }, { "epoch": 2.4277970511708586, "grad_norm": 0.44267138838768005, "learning_rate": 3.828521688669665e-06, "loss": 0.0011, "step": 22394 }, { "epoch": 2.4279054640069386, "grad_norm": 6.095732688903809, "learning_rate": 3.827796315102278e-06, "loss": 0.0198, "step": 22395 }, { "epoch": 2.428013876843018, "grad_norm": 12.19839096069336, "learning_rate": 3.827070941534891e-06, "loss": 0.0075, "step": 22396 }, { "epoch": 2.428122289679098, "grad_norm": 0.21197326481342316, "learning_rate": 3.826345567967504e-06, "loss": 0.0108, "step": 22397 }, { "epoch": 2.4282307025151777, "grad_norm": 0.0056747752241790295, "learning_rate": 3.825620194400117e-06, "loss": 0.0001, "step": 22398 }, { "epoch": 2.4283391153512577, "grad_norm": 2.354918956756592, "learning_rate": 3.824894820832729e-06, "loss": 0.013, "step": 22399 }, { "epoch": 2.4284475281873372, "grad_norm": 0.8339073061943054, "learning_rate": 3.824169447265342e-06, "loss": 0.0017, "step": 22400 }, { "epoch": 2.428555941023417, "grad_norm": 8.326529502868652, "learning_rate": 3.823444073697955e-06, "loss": 0.0087, "step": 22401 }, { "epoch": 2.4286643538594967, "grad_norm": 0.28998520970344543, "learning_rate": 3.822718700130568e-06, "loss": 0.0105, "step": 22402 }, { "epoch": 2.4287727666955767, "grad_norm": 0.0060836439952254295, "learning_rate": 3.82199332656318e-06, "loss": 0.0001, "step": 22403 }, { "epoch": 2.4288811795316567, "grad_norm": 2.04010009765625, "learning_rate": 3.821267952995793e-06, "loss": 0.025, "step": 22404 }, { "epoch": 2.4289895923677363, "grad_norm": 0.31278136372566223, "learning_rate": 3.820542579428406e-06, "loss": 0.0057, "step": 22405 }, { "epoch": 2.4290980052038162, "grad_norm": 0.5496999621391296, "learning_rate": 3.819817205861019e-06, "loss": 0.0278, "step": 22406 }, { "epoch": 2.429206418039896, "grad_norm": 1.6856402158737183, "learning_rate": 3.819091832293631e-06, "loss": 0.0436, "step": 22407 }, { "epoch": 2.4293148308759758, "grad_norm": 0.2962588667869568, "learning_rate": 3.818366458726244e-06, "loss": 0.0111, "step": 22408 }, { "epoch": 2.4294232437120553, "grad_norm": 0.3229468762874603, "learning_rate": 3.817641085158857e-06, "loss": 0.0071, "step": 22409 }, { "epoch": 2.4295316565481353, "grad_norm": 0.44137275218963623, "learning_rate": 3.81691571159147e-06, "loss": 0.0061, "step": 22410 }, { "epoch": 2.4296400693842153, "grad_norm": 1.40290105342865, "learning_rate": 3.816190338024083e-06, "loss": 0.0085, "step": 22411 }, { "epoch": 2.429748482220295, "grad_norm": 0.34251686930656433, "learning_rate": 3.815464964456695e-06, "loss": 0.0071, "step": 22412 }, { "epoch": 2.429856895056375, "grad_norm": 0.2872021198272705, "learning_rate": 3.814739590889308e-06, "loss": 0.0065, "step": 22413 }, { "epoch": 2.4299653078924544, "grad_norm": 0.31702154874801636, "learning_rate": 3.8140142173219213e-06, "loss": 0.0117, "step": 22414 }, { "epoch": 2.4300737207285343, "grad_norm": 0.21364928781986237, "learning_rate": 3.813288843754534e-06, "loss": 0.0038, "step": 22415 }, { "epoch": 2.430182133564614, "grad_norm": 3.6483030319213867, "learning_rate": 3.812563470187147e-06, "loss": 0.0108, "step": 22416 }, { "epoch": 2.430290546400694, "grad_norm": 0.5441153645515442, "learning_rate": 3.8118380966197594e-06, "loss": 0.0084, "step": 22417 }, { "epoch": 2.430398959236774, "grad_norm": 0.39776819944381714, "learning_rate": 3.811112723052372e-06, "loss": 0.0142, "step": 22418 }, { "epoch": 2.4305073720728534, "grad_norm": 0.49111348390579224, "learning_rate": 3.8103873494849854e-06, "loss": 0.0251, "step": 22419 }, { "epoch": 2.4306157849089334, "grad_norm": 3.427762508392334, "learning_rate": 3.809661975917598e-06, "loss": 0.0173, "step": 22420 }, { "epoch": 2.430724197745013, "grad_norm": 0.043492525815963745, "learning_rate": 3.8089366023502105e-06, "loss": 0.0002, "step": 22421 }, { "epoch": 2.430832610581093, "grad_norm": 0.537604808807373, "learning_rate": 3.808211228782823e-06, "loss": 0.0092, "step": 22422 }, { "epoch": 2.4309410234171724, "grad_norm": 0.3578377068042755, "learning_rate": 3.8074858552154365e-06, "loss": 0.0009, "step": 22423 }, { "epoch": 2.4310494362532524, "grad_norm": 0.30762556195259094, "learning_rate": 3.806760481648049e-06, "loss": 0.0067, "step": 22424 }, { "epoch": 2.431157849089332, "grad_norm": 5.641323566436768, "learning_rate": 3.8060351080806616e-06, "loss": 0.0253, "step": 22425 }, { "epoch": 2.431266261925412, "grad_norm": 0.26105648279190063, "learning_rate": 3.8053097345132746e-06, "loss": 0.0046, "step": 22426 }, { "epoch": 2.431374674761492, "grad_norm": 1.6260732412338257, "learning_rate": 3.8045843609458876e-06, "loss": 0.0501, "step": 22427 }, { "epoch": 2.4314830875975715, "grad_norm": 0.3633093237876892, "learning_rate": 3.8038589873785e-06, "loss": 0.0114, "step": 22428 }, { "epoch": 2.4315915004336515, "grad_norm": 0.3795836865901947, "learning_rate": 3.803133613811113e-06, "loss": 0.0246, "step": 22429 }, { "epoch": 2.431699913269731, "grad_norm": 0.005989669356495142, "learning_rate": 3.8024082402437257e-06, "loss": 0.0, "step": 22430 }, { "epoch": 2.431808326105811, "grad_norm": 0.003453349694609642, "learning_rate": 3.8016828666763387e-06, "loss": 0.0, "step": 22431 }, { "epoch": 2.4319167389418905, "grad_norm": 0.015174579806625843, "learning_rate": 3.8009574931089517e-06, "loss": 0.0001, "step": 22432 }, { "epoch": 2.4320251517779705, "grad_norm": 0.30599746108055115, "learning_rate": 3.8002321195415643e-06, "loss": 0.0103, "step": 22433 }, { "epoch": 2.43213356461405, "grad_norm": 0.2738645374774933, "learning_rate": 3.799506745974177e-06, "loss": 0.001, "step": 22434 }, { "epoch": 2.43224197745013, "grad_norm": 1.7830785512924194, "learning_rate": 3.7987813724067894e-06, "loss": 0.0593, "step": 22435 }, { "epoch": 2.43235039028621, "grad_norm": 5.779397010803223, "learning_rate": 3.798055998839403e-06, "loss": 0.0058, "step": 22436 }, { "epoch": 2.4324588031222896, "grad_norm": 3.3802802562713623, "learning_rate": 3.7973306252720154e-06, "loss": 0.0256, "step": 22437 }, { "epoch": 2.4325672159583696, "grad_norm": 0.4090479910373688, "learning_rate": 3.796605251704628e-06, "loss": 0.0256, "step": 22438 }, { "epoch": 2.432675628794449, "grad_norm": 0.5423702001571655, "learning_rate": 3.795879878137241e-06, "loss": 0.0152, "step": 22439 }, { "epoch": 2.432784041630529, "grad_norm": 0.6096758246421814, "learning_rate": 3.795154504569854e-06, "loss": 0.0273, "step": 22440 }, { "epoch": 2.432892454466609, "grad_norm": 0.09816938638687134, "learning_rate": 3.7944291310024665e-06, "loss": 0.0004, "step": 22441 }, { "epoch": 2.4330008673026886, "grad_norm": 0.7509626746177673, "learning_rate": 3.7937037574350795e-06, "loss": 0.0409, "step": 22442 }, { "epoch": 2.4331092801387686, "grad_norm": 0.010283516719937325, "learning_rate": 3.792978383867692e-06, "loss": 0.0001, "step": 22443 }, { "epoch": 2.433217692974848, "grad_norm": 2.5765140056610107, "learning_rate": 3.792253010300305e-06, "loss": 0.0179, "step": 22444 }, { "epoch": 2.433326105810928, "grad_norm": 0.10656508803367615, "learning_rate": 3.791527636732918e-06, "loss": 0.0004, "step": 22445 }, { "epoch": 2.4334345186470077, "grad_norm": 0.22823520004749298, "learning_rate": 3.7908022631655306e-06, "loss": 0.0045, "step": 22446 }, { "epoch": 2.4335429314830876, "grad_norm": 0.44819343090057373, "learning_rate": 3.790076889598143e-06, "loss": 0.009, "step": 22447 }, { "epoch": 2.433651344319167, "grad_norm": 0.25894230604171753, "learning_rate": 3.7893515160307557e-06, "loss": 0.0056, "step": 22448 }, { "epoch": 2.433759757155247, "grad_norm": 0.03855106607079506, "learning_rate": 3.788626142463369e-06, "loss": 0.0002, "step": 22449 }, { "epoch": 2.433868169991327, "grad_norm": 0.01434917189180851, "learning_rate": 3.7879007688959817e-06, "loss": 0.0002, "step": 22450 }, { "epoch": 2.4339765828274067, "grad_norm": 0.11299757659435272, "learning_rate": 3.7871753953285943e-06, "loss": 0.0005, "step": 22451 }, { "epoch": 2.4340849956634867, "grad_norm": 4.670754909515381, "learning_rate": 3.7864500217612073e-06, "loss": 0.0018, "step": 22452 }, { "epoch": 2.4341934084995662, "grad_norm": 0.06199193745851517, "learning_rate": 3.7857246481938203e-06, "loss": 0.0001, "step": 22453 }, { "epoch": 2.434301821335646, "grad_norm": 1.3774447441101074, "learning_rate": 3.784999274626433e-06, "loss": 0.0471, "step": 22454 }, { "epoch": 2.4344102341717258, "grad_norm": 0.4295729994773865, "learning_rate": 3.784273901059046e-06, "loss": 0.0059, "step": 22455 }, { "epoch": 2.4345186470078057, "grad_norm": 2.5712497234344482, "learning_rate": 3.7835485274916584e-06, "loss": 0.0525, "step": 22456 }, { "epoch": 2.4346270598438853, "grad_norm": 0.3732893764972687, "learning_rate": 3.7828231539242714e-06, "loss": 0.0073, "step": 22457 }, { "epoch": 2.4347354726799653, "grad_norm": 1.1544631719589233, "learning_rate": 3.7820977803568844e-06, "loss": 0.0024, "step": 22458 }, { "epoch": 2.4348438855160452, "grad_norm": 4.878114700317383, "learning_rate": 3.781372406789497e-06, "loss": 0.0774, "step": 22459 }, { "epoch": 2.434952298352125, "grad_norm": 5.663516044616699, "learning_rate": 3.7806470332221095e-06, "loss": 0.0154, "step": 22460 }, { "epoch": 2.4350607111882048, "grad_norm": 0.29020804166793823, "learning_rate": 3.779921659654722e-06, "loss": 0.0103, "step": 22461 }, { "epoch": 2.4351691240242843, "grad_norm": 0.08474424481391907, "learning_rate": 3.7791962860873355e-06, "loss": 0.0005, "step": 22462 }, { "epoch": 2.4352775368603643, "grad_norm": 0.05183691531419754, "learning_rate": 3.778470912519948e-06, "loss": 0.0004, "step": 22463 }, { "epoch": 2.4353859496964443, "grad_norm": 0.27916568517684937, "learning_rate": 3.7777455389525606e-06, "loss": 0.0062, "step": 22464 }, { "epoch": 2.435494362532524, "grad_norm": 0.18369971215724945, "learning_rate": 3.7770201653851736e-06, "loss": 0.0005, "step": 22465 }, { "epoch": 2.435602775368604, "grad_norm": 0.00947277806699276, "learning_rate": 3.7762947918177866e-06, "loss": 0.0001, "step": 22466 }, { "epoch": 2.4357111882046834, "grad_norm": 4.90693998336792, "learning_rate": 3.775569418250399e-06, "loss": 0.0374, "step": 22467 }, { "epoch": 2.4358196010407633, "grad_norm": 2.043924570083618, "learning_rate": 3.774844044683012e-06, "loss": 0.0242, "step": 22468 }, { "epoch": 2.435928013876843, "grad_norm": 0.30422717332839966, "learning_rate": 3.7741186711156247e-06, "loss": 0.0104, "step": 22469 }, { "epoch": 2.436036426712923, "grad_norm": 0.03606529161334038, "learning_rate": 3.7733932975482377e-06, "loss": 0.0004, "step": 22470 }, { "epoch": 2.4361448395490024, "grad_norm": 0.2910779118537903, "learning_rate": 3.7726679239808507e-06, "loss": 0.0104, "step": 22471 }, { "epoch": 2.4362532523850824, "grad_norm": 0.5880201458930969, "learning_rate": 3.7719425504134633e-06, "loss": 0.0293, "step": 22472 }, { "epoch": 2.4363616652211624, "grad_norm": 0.3252009153366089, "learning_rate": 3.771217176846076e-06, "loss": 0.0145, "step": 22473 }, { "epoch": 2.436470078057242, "grad_norm": 26.676443099975586, "learning_rate": 3.7704918032786884e-06, "loss": 0.047, "step": 22474 }, { "epoch": 2.436578490893322, "grad_norm": 0.3683246076107025, "learning_rate": 3.769766429711302e-06, "loss": 0.0079, "step": 22475 }, { "epoch": 2.4366869037294014, "grad_norm": 2.2974767684936523, "learning_rate": 3.7690410561439144e-06, "loss": 0.0036, "step": 22476 }, { "epoch": 2.4367953165654814, "grad_norm": 0.29281482100486755, "learning_rate": 3.768315682576527e-06, "loss": 0.0059, "step": 22477 }, { "epoch": 2.436903729401561, "grad_norm": 0.22275041043758392, "learning_rate": 3.76759030900914e-06, "loss": 0.0118, "step": 22478 }, { "epoch": 2.437012142237641, "grad_norm": 0.5109204649925232, "learning_rate": 3.766864935441753e-06, "loss": 0.0168, "step": 22479 }, { "epoch": 2.4371205550737205, "grad_norm": 0.024575425311923027, "learning_rate": 3.7661395618743655e-06, "loss": 0.0002, "step": 22480 }, { "epoch": 2.4372289679098005, "grad_norm": 5.984236717224121, "learning_rate": 3.7654141883069785e-06, "loss": 0.0574, "step": 22481 }, { "epoch": 2.4373373807458805, "grad_norm": 3.868706464767456, "learning_rate": 3.764688814739591e-06, "loss": 0.0083, "step": 22482 }, { "epoch": 2.43744579358196, "grad_norm": 0.03555744141340256, "learning_rate": 3.763963441172204e-06, "loss": 0.0002, "step": 22483 }, { "epoch": 2.43755420641804, "grad_norm": 8.850034713745117, "learning_rate": 3.763238067604817e-06, "loss": 0.0453, "step": 22484 }, { "epoch": 2.4376626192541195, "grad_norm": 0.3330426812171936, "learning_rate": 3.7625126940374296e-06, "loss": 0.0061, "step": 22485 }, { "epoch": 2.4377710320901995, "grad_norm": 3.316049098968506, "learning_rate": 3.761787320470042e-06, "loss": 0.0141, "step": 22486 }, { "epoch": 2.4378794449262795, "grad_norm": 4.921291351318359, "learning_rate": 3.7610619469026547e-06, "loss": 0.0028, "step": 22487 }, { "epoch": 2.437987857762359, "grad_norm": 0.07801483571529388, "learning_rate": 3.760336573335268e-06, "loss": 0.0003, "step": 22488 }, { "epoch": 2.438096270598439, "grad_norm": 0.5568594336509705, "learning_rate": 3.7596111997678807e-06, "loss": 0.0009, "step": 22489 }, { "epoch": 2.4382046834345186, "grad_norm": 3.4218249320983887, "learning_rate": 3.7588858262004933e-06, "loss": 0.0324, "step": 22490 }, { "epoch": 2.4383130962705986, "grad_norm": 0.2535405457019806, "learning_rate": 3.7581604526331063e-06, "loss": 0.0013, "step": 22491 }, { "epoch": 2.438421509106678, "grad_norm": 0.32271692156791687, "learning_rate": 3.7574350790657193e-06, "loss": 0.0073, "step": 22492 }, { "epoch": 2.438529921942758, "grad_norm": 1.1679468154907227, "learning_rate": 3.756709705498332e-06, "loss": 0.0358, "step": 22493 }, { "epoch": 2.4386383347788376, "grad_norm": 0.03622952476143837, "learning_rate": 3.755984331930945e-06, "loss": 0.0003, "step": 22494 }, { "epoch": 2.4387467476149176, "grad_norm": 0.36420485377311707, "learning_rate": 3.7552589583635574e-06, "loss": 0.005, "step": 22495 }, { "epoch": 2.4388551604509976, "grad_norm": 0.840951681137085, "learning_rate": 3.7545335847961704e-06, "loss": 0.0019, "step": 22496 }, { "epoch": 2.438963573287077, "grad_norm": 1.6012036800384521, "learning_rate": 3.7538082112287834e-06, "loss": 0.0039, "step": 22497 }, { "epoch": 2.439071986123157, "grad_norm": 0.15178336203098297, "learning_rate": 3.753082837661396e-06, "loss": 0.0005, "step": 22498 }, { "epoch": 2.4391803989592367, "grad_norm": 0.3094596266746521, "learning_rate": 3.7523574640940085e-06, "loss": 0.0061, "step": 22499 }, { "epoch": 2.4392888117953166, "grad_norm": 0.18106146156787872, "learning_rate": 3.751632090526621e-06, "loss": 0.0007, "step": 22500 }, { "epoch": 2.439397224631396, "grad_norm": 0.32070109248161316, "learning_rate": 3.7509067169592345e-06, "loss": 0.0071, "step": 22501 }, { "epoch": 2.439505637467476, "grad_norm": 0.2504616379737854, "learning_rate": 3.750181343391847e-06, "loss": 0.0038, "step": 22502 }, { "epoch": 2.4396140503035557, "grad_norm": 0.14670822024345398, "learning_rate": 3.7494559698244596e-06, "loss": 0.0005, "step": 22503 }, { "epoch": 2.4397224631396357, "grad_norm": 0.13351117074489594, "learning_rate": 3.7487305962570726e-06, "loss": 0.0106, "step": 22504 }, { "epoch": 2.4398308759757157, "grad_norm": 0.3070044219493866, "learning_rate": 3.7480052226896856e-06, "loss": 0.0068, "step": 22505 }, { "epoch": 2.4399392888117952, "grad_norm": 0.4525096118450165, "learning_rate": 3.747279849122298e-06, "loss": 0.0194, "step": 22506 }, { "epoch": 2.440047701647875, "grad_norm": 1.4402427673339844, "learning_rate": 3.746554475554911e-06, "loss": 0.0781, "step": 22507 }, { "epoch": 2.4401561144839548, "grad_norm": 0.3178398311138153, "learning_rate": 3.7458291019875237e-06, "loss": 0.0071, "step": 22508 }, { "epoch": 2.4402645273200347, "grad_norm": 0.22733458876609802, "learning_rate": 3.7451037284201367e-06, "loss": 0.0044, "step": 22509 }, { "epoch": 2.4403729401561147, "grad_norm": 0.4347049593925476, "learning_rate": 3.7443783548527497e-06, "loss": 0.017, "step": 22510 }, { "epoch": 2.4404813529921943, "grad_norm": 8.913148880004883, "learning_rate": 3.7436529812853623e-06, "loss": 0.0263, "step": 22511 }, { "epoch": 2.4405897658282742, "grad_norm": 0.5941928625106812, "learning_rate": 3.742927607717975e-06, "loss": 0.0007, "step": 22512 }, { "epoch": 2.440698178664354, "grad_norm": 0.02950751781463623, "learning_rate": 3.7422022341505874e-06, "loss": 0.0001, "step": 22513 }, { "epoch": 2.4408065915004338, "grad_norm": 0.021343151107430458, "learning_rate": 3.741476860583201e-06, "loss": 0.0002, "step": 22514 }, { "epoch": 2.4409150043365133, "grad_norm": 0.1419929563999176, "learning_rate": 3.7407514870158134e-06, "loss": 0.0008, "step": 22515 }, { "epoch": 2.4410234171725933, "grad_norm": 0.27825111150741577, "learning_rate": 3.740026113448426e-06, "loss": 0.0226, "step": 22516 }, { "epoch": 2.441131830008673, "grad_norm": 0.45888280868530273, "learning_rate": 3.739300739881039e-06, "loss": 0.0168, "step": 22517 }, { "epoch": 2.441240242844753, "grad_norm": 5.230423450469971, "learning_rate": 3.738575366313652e-06, "loss": 0.0302, "step": 22518 }, { "epoch": 2.441348655680833, "grad_norm": 0.07617504894733429, "learning_rate": 3.7378499927462645e-06, "loss": 0.0004, "step": 22519 }, { "epoch": 2.4414570685169124, "grad_norm": 0.44813698530197144, "learning_rate": 3.7371246191788775e-06, "loss": 0.0168, "step": 22520 }, { "epoch": 2.4415654813529923, "grad_norm": 0.052879903465509415, "learning_rate": 3.73639924561149e-06, "loss": 0.0004, "step": 22521 }, { "epoch": 2.441673894189072, "grad_norm": 2.149784564971924, "learning_rate": 3.735673872044103e-06, "loss": 0.0142, "step": 22522 }, { "epoch": 2.441782307025152, "grad_norm": 1.2961353063583374, "learning_rate": 3.734948498476716e-06, "loss": 0.0075, "step": 22523 }, { "epoch": 2.4418907198612314, "grad_norm": 1.8595224618911743, "learning_rate": 3.7342231249093286e-06, "loss": 0.0183, "step": 22524 }, { "epoch": 2.4419991326973114, "grad_norm": 0.4950758218765259, "learning_rate": 3.733497751341941e-06, "loss": 0.0103, "step": 22525 }, { "epoch": 2.442107545533391, "grad_norm": 0.30698880553245544, "learning_rate": 3.7327723777745537e-06, "loss": 0.0071, "step": 22526 }, { "epoch": 2.442215958369471, "grad_norm": 1.0525623559951782, "learning_rate": 3.732047004207167e-06, "loss": 0.0045, "step": 22527 }, { "epoch": 2.442324371205551, "grad_norm": 0.3006189465522766, "learning_rate": 3.7313216306397797e-06, "loss": 0.0108, "step": 22528 }, { "epoch": 2.4424327840416304, "grad_norm": 0.31113240122795105, "learning_rate": 3.7305962570723923e-06, "loss": 0.0068, "step": 22529 }, { "epoch": 2.4425411968777104, "grad_norm": 0.008971734903752804, "learning_rate": 3.7298708835050053e-06, "loss": 0.0001, "step": 22530 }, { "epoch": 2.44264960971379, "grad_norm": 16.72980308532715, "learning_rate": 3.7291455099376183e-06, "loss": 0.0227, "step": 22531 }, { "epoch": 2.44275802254987, "grad_norm": 0.04417884349822998, "learning_rate": 3.728420136370231e-06, "loss": 0.0002, "step": 22532 }, { "epoch": 2.44286643538595, "grad_norm": 0.05321647971868515, "learning_rate": 3.727694762802844e-06, "loss": 0.0004, "step": 22533 }, { "epoch": 2.4429748482220295, "grad_norm": 1.640022873878479, "learning_rate": 3.7269693892354564e-06, "loss": 0.0138, "step": 22534 }, { "epoch": 2.4430832610581095, "grad_norm": 0.4680992066860199, "learning_rate": 3.7262440156680694e-06, "loss": 0.0167, "step": 22535 }, { "epoch": 2.443191673894189, "grad_norm": 0.30998218059539795, "learning_rate": 3.7255186421006824e-06, "loss": 0.0064, "step": 22536 }, { "epoch": 2.443300086730269, "grad_norm": 0.2731340825557709, "learning_rate": 3.724793268533295e-06, "loss": 0.006, "step": 22537 }, { "epoch": 2.4434084995663485, "grad_norm": 0.02959788590669632, "learning_rate": 3.7240678949659075e-06, "loss": 0.0002, "step": 22538 }, { "epoch": 2.4435169124024285, "grad_norm": 0.25769177079200745, "learning_rate": 3.723342521398521e-06, "loss": 0.0035, "step": 22539 }, { "epoch": 2.443625325238508, "grad_norm": 2.8521857261657715, "learning_rate": 3.7226171478311335e-06, "loss": 0.0055, "step": 22540 }, { "epoch": 2.443733738074588, "grad_norm": 0.4100390374660492, "learning_rate": 3.721891774263746e-06, "loss": 0.0065, "step": 22541 }, { "epoch": 2.443842150910668, "grad_norm": 0.06779515743255615, "learning_rate": 3.7211664006963586e-06, "loss": 0.0005, "step": 22542 }, { "epoch": 2.4439505637467476, "grad_norm": 0.16866843402385712, "learning_rate": 3.7204410271289716e-06, "loss": 0.0008, "step": 22543 }, { "epoch": 2.4440589765828276, "grad_norm": 0.014837208203971386, "learning_rate": 3.7197156535615846e-06, "loss": 0.0001, "step": 22544 }, { "epoch": 2.444167389418907, "grad_norm": 3.094475507736206, "learning_rate": 3.718990279994197e-06, "loss": 0.0144, "step": 22545 }, { "epoch": 2.444275802254987, "grad_norm": 2.659982204437256, "learning_rate": 3.71826490642681e-06, "loss": 0.0366, "step": 22546 }, { "epoch": 2.4443842150910666, "grad_norm": 0.07284661382436752, "learning_rate": 3.7175395328594227e-06, "loss": 0.0002, "step": 22547 }, { "epoch": 2.4444926279271466, "grad_norm": 5.613092422485352, "learning_rate": 3.7168141592920357e-06, "loss": 0.0185, "step": 22548 }, { "epoch": 2.444601040763226, "grad_norm": 1.8577467203140259, "learning_rate": 3.7160887857246487e-06, "loss": 0.0026, "step": 22549 }, { "epoch": 2.444709453599306, "grad_norm": 0.029065921902656555, "learning_rate": 3.7153634121572613e-06, "loss": 0.0003, "step": 22550 }, { "epoch": 2.444817866435386, "grad_norm": 0.7959532737731934, "learning_rate": 3.714638038589874e-06, "loss": 0.0214, "step": 22551 }, { "epoch": 2.4449262792714657, "grad_norm": 0.3054358959197998, "learning_rate": 3.7139126650224873e-06, "loss": 0.0111, "step": 22552 }, { "epoch": 2.4450346921075456, "grad_norm": 12.09798526763916, "learning_rate": 3.7131872914551e-06, "loss": 0.0127, "step": 22553 }, { "epoch": 2.445143104943625, "grad_norm": 0.24091218411922455, "learning_rate": 3.7124619178877124e-06, "loss": 0.0056, "step": 22554 }, { "epoch": 2.445251517779705, "grad_norm": 0.21001645922660828, "learning_rate": 3.711736544320325e-06, "loss": 0.0029, "step": 22555 }, { "epoch": 2.4453599306157847, "grad_norm": 0.022244509309530258, "learning_rate": 3.711011170752938e-06, "loss": 0.0001, "step": 22556 }, { "epoch": 2.4454683434518647, "grad_norm": 0.4586678445339203, "learning_rate": 3.710285797185551e-06, "loss": 0.0317, "step": 22557 }, { "epoch": 2.4455767562879447, "grad_norm": 0.45964354276657104, "learning_rate": 3.7095604236181635e-06, "loss": 0.0111, "step": 22558 }, { "epoch": 2.4456851691240242, "grad_norm": 0.49599558115005493, "learning_rate": 3.7088350500507765e-06, "loss": 0.0017, "step": 22559 }, { "epoch": 2.445793581960104, "grad_norm": 0.46515974402427673, "learning_rate": 3.708109676483389e-06, "loss": 0.0202, "step": 22560 }, { "epoch": 2.4459019947961838, "grad_norm": 0.4691818356513977, "learning_rate": 3.707384302916002e-06, "loss": 0.0073, "step": 22561 }, { "epoch": 2.4460104076322637, "grad_norm": 0.2912658154964447, "learning_rate": 3.706658929348615e-06, "loss": 0.0066, "step": 22562 }, { "epoch": 2.4461188204683433, "grad_norm": 3.9860901832580566, "learning_rate": 3.7059335557812276e-06, "loss": 0.0218, "step": 22563 }, { "epoch": 2.4462272333044233, "grad_norm": 0.3715255558490753, "learning_rate": 3.70520818221384e-06, "loss": 0.0069, "step": 22564 }, { "epoch": 2.4463356461405033, "grad_norm": 0.4100814461708069, "learning_rate": 3.7044828086464536e-06, "loss": 0.0126, "step": 22565 }, { "epoch": 2.446444058976583, "grad_norm": 9.330424308776855, "learning_rate": 3.703757435079066e-06, "loss": 0.0437, "step": 22566 }, { "epoch": 2.4465524718126628, "grad_norm": 0.25215405225753784, "learning_rate": 3.7030320615116787e-06, "loss": 0.005, "step": 22567 }, { "epoch": 2.4466608846487423, "grad_norm": 0.33917951583862305, "learning_rate": 3.7023066879442913e-06, "loss": 0.011, "step": 22568 }, { "epoch": 2.4467692974848223, "grad_norm": 0.00508015975356102, "learning_rate": 3.7015813143769043e-06, "loss": 0.0001, "step": 22569 }, { "epoch": 2.446877710320902, "grad_norm": 0.24314554035663605, "learning_rate": 3.7008559408095173e-06, "loss": 0.0042, "step": 22570 }, { "epoch": 2.446986123156982, "grad_norm": 0.5633642673492432, "learning_rate": 3.70013056724213e-06, "loss": 0.0239, "step": 22571 }, { "epoch": 2.4470945359930614, "grad_norm": 0.27786725759506226, "learning_rate": 3.699405193674743e-06, "loss": 0.006, "step": 22572 }, { "epoch": 2.4472029488291414, "grad_norm": 0.36300620436668396, "learning_rate": 3.6986798201073554e-06, "loss": 0.0292, "step": 22573 }, { "epoch": 2.4473113616652213, "grad_norm": 0.2150481790304184, "learning_rate": 3.6979544465399684e-06, "loss": 0.003, "step": 22574 }, { "epoch": 2.447419774501301, "grad_norm": 0.03133077919483185, "learning_rate": 3.6972290729725814e-06, "loss": 0.0002, "step": 22575 }, { "epoch": 2.447528187337381, "grad_norm": 0.4555591642856598, "learning_rate": 3.696503699405194e-06, "loss": 0.0172, "step": 22576 }, { "epoch": 2.4476366001734604, "grad_norm": 0.26563191413879395, "learning_rate": 3.6957783258378065e-06, "loss": 0.0042, "step": 22577 }, { "epoch": 2.4477450130095404, "grad_norm": 0.42303264141082764, "learning_rate": 3.69505295227042e-06, "loss": 0.0129, "step": 22578 }, { "epoch": 2.44785342584562, "grad_norm": 0.03283488377928734, "learning_rate": 3.6943275787030325e-06, "loss": 0.0003, "step": 22579 }, { "epoch": 2.4479618386817, "grad_norm": 0.09701194614171982, "learning_rate": 3.693602205135645e-06, "loss": 0.0007, "step": 22580 }, { "epoch": 2.4480702515177795, "grad_norm": 0.007124695461243391, "learning_rate": 3.6928768315682576e-06, "loss": 0.0001, "step": 22581 }, { "epoch": 2.4481786643538594, "grad_norm": 0.20901554822921753, "learning_rate": 3.6921514580008706e-06, "loss": 0.0038, "step": 22582 }, { "epoch": 2.4482870771899394, "grad_norm": 0.22491545975208282, "learning_rate": 3.6914260844334836e-06, "loss": 0.0043, "step": 22583 }, { "epoch": 2.448395490026019, "grad_norm": 0.0474262610077858, "learning_rate": 3.690700710866096e-06, "loss": 0.0002, "step": 22584 }, { "epoch": 2.448503902862099, "grad_norm": 0.4472677409648895, "learning_rate": 3.689975337298709e-06, "loss": 0.0156, "step": 22585 }, { "epoch": 2.4486123156981785, "grad_norm": 1.642472267150879, "learning_rate": 3.6892499637313217e-06, "loss": 0.0013, "step": 22586 }, { "epoch": 2.4487207285342585, "grad_norm": 0.37791383266448975, "learning_rate": 3.6885245901639347e-06, "loss": 0.0087, "step": 22587 }, { "epoch": 2.4488291413703385, "grad_norm": 0.035987820476293564, "learning_rate": 3.6877992165965477e-06, "loss": 0.0002, "step": 22588 }, { "epoch": 2.448937554206418, "grad_norm": 0.005790719296783209, "learning_rate": 3.6870738430291603e-06, "loss": 0.0, "step": 22589 }, { "epoch": 2.449045967042498, "grad_norm": 3.901658296585083, "learning_rate": 3.686348469461773e-06, "loss": 0.0226, "step": 22590 }, { "epoch": 2.4491543798785775, "grad_norm": 0.7738184332847595, "learning_rate": 3.6856230958943863e-06, "loss": 0.0267, "step": 22591 }, { "epoch": 2.4492627927146575, "grad_norm": 0.0033805128186941147, "learning_rate": 3.684897722326999e-06, "loss": 0.0, "step": 22592 }, { "epoch": 2.449371205550737, "grad_norm": 0.23040138185024261, "learning_rate": 3.6841723487596114e-06, "loss": 0.0012, "step": 22593 }, { "epoch": 2.449479618386817, "grad_norm": 5.6696038246154785, "learning_rate": 3.683446975192224e-06, "loss": 0.0182, "step": 22594 }, { "epoch": 2.4495880312228966, "grad_norm": 0.020801911130547523, "learning_rate": 3.682721601624837e-06, "loss": 0.0001, "step": 22595 }, { "epoch": 2.4496964440589766, "grad_norm": 0.4051383435726166, "learning_rate": 3.68199622805745e-06, "loss": 0.0229, "step": 22596 }, { "epoch": 2.4498048568950566, "grad_norm": 5.366341590881348, "learning_rate": 3.6812708544900625e-06, "loss": 0.0095, "step": 22597 }, { "epoch": 2.449913269731136, "grad_norm": 0.26130715012550354, "learning_rate": 3.6805454809226755e-06, "loss": 0.0057, "step": 22598 }, { "epoch": 2.450021682567216, "grad_norm": 0.041529618203639984, "learning_rate": 3.679820107355288e-06, "loss": 0.0001, "step": 22599 }, { "epoch": 2.4501300954032956, "grad_norm": 0.08216926455497742, "learning_rate": 3.679094733787901e-06, "loss": 0.011, "step": 22600 }, { "epoch": 2.4502385082393756, "grad_norm": 0.01999558135867119, "learning_rate": 3.678369360220514e-06, "loss": 0.0002, "step": 22601 }, { "epoch": 2.450346921075455, "grad_norm": 1.3848785161972046, "learning_rate": 3.6776439866531266e-06, "loss": 0.0302, "step": 22602 }, { "epoch": 2.450455333911535, "grad_norm": 1.2896277904510498, "learning_rate": 3.676918613085739e-06, "loss": 0.0015, "step": 22603 }, { "epoch": 2.4505637467476147, "grad_norm": 0.27991756796836853, "learning_rate": 3.6761932395183526e-06, "loss": 0.006, "step": 22604 }, { "epoch": 2.4506721595836947, "grad_norm": 3.670563220977783, "learning_rate": 3.675467865950965e-06, "loss": 0.0088, "step": 22605 }, { "epoch": 2.4507805724197746, "grad_norm": 0.5011534094810486, "learning_rate": 3.6747424923835777e-06, "loss": 0.0189, "step": 22606 }, { "epoch": 2.450888985255854, "grad_norm": 22.696901321411133, "learning_rate": 3.6740171188161903e-06, "loss": 0.0172, "step": 22607 }, { "epoch": 2.450997398091934, "grad_norm": 0.08057615906000137, "learning_rate": 3.6732917452488033e-06, "loss": 0.0006, "step": 22608 }, { "epoch": 2.4511058109280137, "grad_norm": 0.016682369634509087, "learning_rate": 3.6725663716814163e-06, "loss": 0.0001, "step": 22609 }, { "epoch": 2.4512142237640937, "grad_norm": 0.39819100499153137, "learning_rate": 3.671840998114029e-06, "loss": 0.0134, "step": 22610 }, { "epoch": 2.4513226366001737, "grad_norm": 2.6775217056274414, "learning_rate": 3.671115624546642e-06, "loss": 0.0283, "step": 22611 }, { "epoch": 2.4514310494362532, "grad_norm": 0.032328344881534576, "learning_rate": 3.6703902509792544e-06, "loss": 0.0001, "step": 22612 }, { "epoch": 2.451539462272333, "grad_norm": 0.21543051302433014, "learning_rate": 3.669664877411868e-06, "loss": 0.0034, "step": 22613 }, { "epoch": 2.4516478751084128, "grad_norm": 0.26749715209007263, "learning_rate": 3.6689395038444804e-06, "loss": 0.0053, "step": 22614 }, { "epoch": 2.4517562879444927, "grad_norm": 0.07731439173221588, "learning_rate": 3.668214130277093e-06, "loss": 0.0004, "step": 22615 }, { "epoch": 2.4518647007805723, "grad_norm": 8.004830360412598, "learning_rate": 3.6674887567097055e-06, "loss": 0.0769, "step": 22616 }, { "epoch": 2.4519731136166523, "grad_norm": 1.8175575733184814, "learning_rate": 3.666763383142319e-06, "loss": 0.0261, "step": 22617 }, { "epoch": 2.452081526452732, "grad_norm": 2.109011173248291, "learning_rate": 3.6660380095749315e-06, "loss": 0.0122, "step": 22618 }, { "epoch": 2.452189939288812, "grad_norm": 0.032259754836559296, "learning_rate": 3.665312636007544e-06, "loss": 0.0001, "step": 22619 }, { "epoch": 2.452298352124892, "grad_norm": 2.5900168418884277, "learning_rate": 3.6645872624401566e-06, "loss": 0.0271, "step": 22620 }, { "epoch": 2.4524067649609713, "grad_norm": 2.863680124282837, "learning_rate": 3.6638618888727696e-06, "loss": 0.0073, "step": 22621 }, { "epoch": 2.4525151777970513, "grad_norm": 0.3557952046394348, "learning_rate": 3.6631365153053826e-06, "loss": 0.0114, "step": 22622 }, { "epoch": 2.452623590633131, "grad_norm": 0.35251384973526, "learning_rate": 3.662411141737995e-06, "loss": 0.0103, "step": 22623 }, { "epoch": 2.452732003469211, "grad_norm": 0.6648775935173035, "learning_rate": 3.661685768170608e-06, "loss": 0.0137, "step": 22624 }, { "epoch": 2.4528404163052904, "grad_norm": 5.774086952209473, "learning_rate": 3.6609603946032207e-06, "loss": 0.0153, "step": 22625 }, { "epoch": 2.4529488291413704, "grad_norm": 0.8544409275054932, "learning_rate": 3.660235021035834e-06, "loss": 0.013, "step": 22626 }, { "epoch": 2.45305724197745, "grad_norm": 0.276721328496933, "learning_rate": 3.6595096474684467e-06, "loss": 0.0062, "step": 22627 }, { "epoch": 2.45316565481353, "grad_norm": 1.3622331619262695, "learning_rate": 3.6587842739010593e-06, "loss": 0.0025, "step": 22628 }, { "epoch": 2.45327406764961, "grad_norm": 0.30925488471984863, "learning_rate": 3.658058900333672e-06, "loss": 0.0108, "step": 22629 }, { "epoch": 2.4533824804856894, "grad_norm": 0.4354085624217987, "learning_rate": 3.6573335267662853e-06, "loss": 0.0085, "step": 22630 }, { "epoch": 2.4534908933217694, "grad_norm": 0.03194787725806236, "learning_rate": 3.656608153198898e-06, "loss": 0.0003, "step": 22631 }, { "epoch": 2.453599306157849, "grad_norm": 1.356549620628357, "learning_rate": 3.6558827796315104e-06, "loss": 0.0704, "step": 22632 }, { "epoch": 2.453707718993929, "grad_norm": 11.563766479492188, "learning_rate": 3.655157406064123e-06, "loss": 0.0295, "step": 22633 }, { "epoch": 2.453816131830009, "grad_norm": 1.2781785726547241, "learning_rate": 3.6544320324967364e-06, "loss": 0.0114, "step": 22634 }, { "epoch": 2.4539245446660884, "grad_norm": 0.015025227330625057, "learning_rate": 3.653706658929349e-06, "loss": 0.0001, "step": 22635 }, { "epoch": 2.4540329575021684, "grad_norm": 2.375202178955078, "learning_rate": 3.6529812853619615e-06, "loss": 0.0089, "step": 22636 }, { "epoch": 2.454141370338248, "grad_norm": 22.944917678833008, "learning_rate": 3.6522559117945745e-06, "loss": 0.0225, "step": 22637 }, { "epoch": 2.454249783174328, "grad_norm": 24.22258758544922, "learning_rate": 3.651530538227187e-06, "loss": 0.0454, "step": 22638 }, { "epoch": 2.4543581960104075, "grad_norm": 11.914081573486328, "learning_rate": 3.6508051646598005e-06, "loss": 0.0525, "step": 22639 }, { "epoch": 2.4544666088464875, "grad_norm": 0.2549349367618561, "learning_rate": 3.650079791092413e-06, "loss": 0.001, "step": 22640 }, { "epoch": 2.454575021682567, "grad_norm": 0.058057259768247604, "learning_rate": 3.6493544175250256e-06, "loss": 0.0004, "step": 22641 }, { "epoch": 2.454683434518647, "grad_norm": 0.016964562237262726, "learning_rate": 3.648629043957638e-06, "loss": 0.0001, "step": 22642 }, { "epoch": 2.454791847354727, "grad_norm": 2.2522945404052734, "learning_rate": 3.6479036703902516e-06, "loss": 0.0461, "step": 22643 }, { "epoch": 2.4549002601908065, "grad_norm": 0.30628886818885803, "learning_rate": 3.647178296822864e-06, "loss": 0.0117, "step": 22644 }, { "epoch": 2.4550086730268865, "grad_norm": 0.0038514002226293087, "learning_rate": 3.6464529232554767e-06, "loss": 0.0, "step": 22645 }, { "epoch": 2.455117085862966, "grad_norm": 6.962734222412109, "learning_rate": 3.6457275496880893e-06, "loss": 0.049, "step": 22646 }, { "epoch": 2.455225498699046, "grad_norm": 0.42844927310943604, "learning_rate": 3.6450021761207027e-06, "loss": 0.0174, "step": 22647 }, { "epoch": 2.4553339115351256, "grad_norm": 5.838751316070557, "learning_rate": 3.6442768025533153e-06, "loss": 0.03, "step": 22648 }, { "epoch": 2.4554423243712056, "grad_norm": 0.09768293052911758, "learning_rate": 3.643551428985928e-06, "loss": 0.0006, "step": 22649 }, { "epoch": 2.455550737207285, "grad_norm": 0.5796429514884949, "learning_rate": 3.642826055418541e-06, "loss": 0.001, "step": 22650 }, { "epoch": 2.455659150043365, "grad_norm": 3.5065274238586426, "learning_rate": 3.6421006818511534e-06, "loss": 0.0055, "step": 22651 }, { "epoch": 2.455767562879445, "grad_norm": 0.2474394589662552, "learning_rate": 3.641375308283767e-06, "loss": 0.001, "step": 22652 }, { "epoch": 2.4558759757155246, "grad_norm": 0.0328308530151844, "learning_rate": 3.6406499347163794e-06, "loss": 0.0002, "step": 22653 }, { "epoch": 2.4559843885516046, "grad_norm": 0.7302895188331604, "learning_rate": 3.639924561148992e-06, "loss": 0.0122, "step": 22654 }, { "epoch": 2.456092801387684, "grad_norm": 1.2289845943450928, "learning_rate": 3.6391991875816045e-06, "loss": 0.003, "step": 22655 }, { "epoch": 2.456201214223764, "grad_norm": 1.8821630477905273, "learning_rate": 3.638473814014218e-06, "loss": 0.046, "step": 22656 }, { "epoch": 2.456309627059844, "grad_norm": 0.4724309742450714, "learning_rate": 3.6377484404468305e-06, "loss": 0.0251, "step": 22657 }, { "epoch": 2.4564180398959237, "grad_norm": 0.3457261323928833, "learning_rate": 3.637023066879443e-06, "loss": 0.0114, "step": 22658 }, { "epoch": 2.4565264527320037, "grad_norm": 1.1224117279052734, "learning_rate": 3.6362976933120556e-06, "loss": 0.0189, "step": 22659 }, { "epoch": 2.456634865568083, "grad_norm": 0.30109426379203796, "learning_rate": 3.635572319744669e-06, "loss": 0.0105, "step": 22660 }, { "epoch": 2.456743278404163, "grad_norm": 7.826532363891602, "learning_rate": 3.6348469461772816e-06, "loss": 0.009, "step": 22661 }, { "epoch": 2.4568516912402427, "grad_norm": 0.6431021690368652, "learning_rate": 3.6341215726098946e-06, "loss": 0.0234, "step": 22662 }, { "epoch": 2.4569601040763227, "grad_norm": 6.255131721496582, "learning_rate": 3.633396199042507e-06, "loss": 0.0251, "step": 22663 }, { "epoch": 2.4570685169124022, "grad_norm": 0.01981120929121971, "learning_rate": 3.6326708254751197e-06, "loss": 0.0002, "step": 22664 }, { "epoch": 2.4571769297484822, "grad_norm": 0.7247825860977173, "learning_rate": 3.631945451907733e-06, "loss": 0.0061, "step": 22665 }, { "epoch": 2.457285342584562, "grad_norm": 0.42162370681762695, "learning_rate": 3.6312200783403457e-06, "loss": 0.0118, "step": 22666 }, { "epoch": 2.4573937554206418, "grad_norm": 16.29182243347168, "learning_rate": 3.6304947047729583e-06, "loss": 0.0329, "step": 22667 }, { "epoch": 2.4575021682567217, "grad_norm": 0.28569766879081726, "learning_rate": 3.629769331205571e-06, "loss": 0.0012, "step": 22668 }, { "epoch": 2.4576105810928013, "grad_norm": 0.5285099148750305, "learning_rate": 3.6290439576381843e-06, "loss": 0.0438, "step": 22669 }, { "epoch": 2.4577189939288813, "grad_norm": 3.1021976470947266, "learning_rate": 3.628318584070797e-06, "loss": 0.0128, "step": 22670 }, { "epoch": 2.457827406764961, "grad_norm": 0.2968982458114624, "learning_rate": 3.6275932105034094e-06, "loss": 0.0066, "step": 22671 }, { "epoch": 2.457935819601041, "grad_norm": 0.8986104130744934, "learning_rate": 3.626867836936022e-06, "loss": 0.006, "step": 22672 }, { "epoch": 2.4580442324371203, "grad_norm": 0.22936081886291504, "learning_rate": 3.6261424633686354e-06, "loss": 0.0153, "step": 22673 }, { "epoch": 2.4581526452732003, "grad_norm": 0.02936922013759613, "learning_rate": 3.625417089801248e-06, "loss": 0.0001, "step": 22674 }, { "epoch": 2.4582610581092803, "grad_norm": 15.40759563446045, "learning_rate": 3.624691716233861e-06, "loss": 0.0178, "step": 22675 }, { "epoch": 2.45836947094536, "grad_norm": 0.3055073916912079, "learning_rate": 3.6239663426664735e-06, "loss": 0.0105, "step": 22676 }, { "epoch": 2.45847788378144, "grad_norm": 0.33559906482696533, "learning_rate": 3.623240969099086e-06, "loss": 0.0085, "step": 22677 }, { "epoch": 2.4585862966175194, "grad_norm": 3.3942692279815674, "learning_rate": 3.6225155955316995e-06, "loss": 0.007, "step": 22678 }, { "epoch": 2.4586947094535994, "grad_norm": 0.2258489429950714, "learning_rate": 3.621790221964312e-06, "loss": 0.0036, "step": 22679 }, { "epoch": 2.4588031222896793, "grad_norm": 25.28606414794922, "learning_rate": 3.6210648483969246e-06, "loss": 0.025, "step": 22680 }, { "epoch": 2.458911535125759, "grad_norm": 0.7104772925376892, "learning_rate": 3.620339474829537e-06, "loss": 0.0286, "step": 22681 }, { "epoch": 2.459019947961839, "grad_norm": 0.0062218173407018185, "learning_rate": 3.6196141012621506e-06, "loss": 0.0001, "step": 22682 }, { "epoch": 2.4591283607979184, "grad_norm": 0.3141399025917053, "learning_rate": 3.618888727694763e-06, "loss": 0.0109, "step": 22683 }, { "epoch": 2.4592367736339984, "grad_norm": 4.160799980163574, "learning_rate": 3.6181633541273757e-06, "loss": 0.0141, "step": 22684 }, { "epoch": 2.459345186470078, "grad_norm": 1.840024709701538, "learning_rate": 3.6174379805599883e-06, "loss": 0.0149, "step": 22685 }, { "epoch": 2.459453599306158, "grad_norm": 0.6761808395385742, "learning_rate": 3.6167126069926017e-06, "loss": 0.0276, "step": 22686 }, { "epoch": 2.4595620121422375, "grad_norm": 0.06687746196985245, "learning_rate": 3.6159872334252143e-06, "loss": 0.0006, "step": 22687 }, { "epoch": 2.4596704249783174, "grad_norm": 2.264995574951172, "learning_rate": 3.6152618598578273e-06, "loss": 0.0103, "step": 22688 }, { "epoch": 2.4597788378143974, "grad_norm": 0.28008997440338135, "learning_rate": 3.61453648629044e-06, "loss": 0.0058, "step": 22689 }, { "epoch": 2.459887250650477, "grad_norm": 0.2892979085445404, "learning_rate": 3.6138111127230524e-06, "loss": 0.0105, "step": 22690 }, { "epoch": 2.459995663486557, "grad_norm": 0.44848018884658813, "learning_rate": 3.613085739155666e-06, "loss": 0.0226, "step": 22691 }, { "epoch": 2.4601040763226365, "grad_norm": 0.5115422606468201, "learning_rate": 3.6123603655882784e-06, "loss": 0.0147, "step": 22692 }, { "epoch": 2.4602124891587165, "grad_norm": 5.175701141357422, "learning_rate": 3.611634992020891e-06, "loss": 0.0214, "step": 22693 }, { "epoch": 2.460320901994796, "grad_norm": 0.017842212691903114, "learning_rate": 3.6109096184535035e-06, "loss": 0.0001, "step": 22694 }, { "epoch": 2.460429314830876, "grad_norm": 0.3317299783229828, "learning_rate": 3.610184244886117e-06, "loss": 0.0114, "step": 22695 }, { "epoch": 2.4605377276669556, "grad_norm": 0.48983561992645264, "learning_rate": 3.6094588713187295e-06, "loss": 0.01, "step": 22696 }, { "epoch": 2.4606461405030355, "grad_norm": 0.030532557517290115, "learning_rate": 3.608733497751342e-06, "loss": 0.0001, "step": 22697 }, { "epoch": 2.4607545533391155, "grad_norm": 0.027734873816370964, "learning_rate": 3.6080081241839546e-06, "loss": 0.0001, "step": 22698 }, { "epoch": 2.460862966175195, "grad_norm": 0.022840920835733414, "learning_rate": 3.607282750616568e-06, "loss": 0.0002, "step": 22699 }, { "epoch": 2.460971379011275, "grad_norm": 0.02332298457622528, "learning_rate": 3.6065573770491806e-06, "loss": 0.0001, "step": 22700 }, { "epoch": 2.4610797918473546, "grad_norm": 0.0024197062011808157, "learning_rate": 3.6058320034817936e-06, "loss": 0.0, "step": 22701 }, { "epoch": 2.4611882046834346, "grad_norm": 0.012166500091552734, "learning_rate": 3.605106629914406e-06, "loss": 0.0, "step": 22702 }, { "epoch": 2.461296617519514, "grad_norm": 0.010271148756146431, "learning_rate": 3.6043812563470187e-06, "loss": 0.0, "step": 22703 }, { "epoch": 2.461405030355594, "grad_norm": 0.006668302696198225, "learning_rate": 3.603655882779632e-06, "loss": 0.0001, "step": 22704 }, { "epoch": 2.4615134431916736, "grad_norm": 11.573408126831055, "learning_rate": 3.6029305092122447e-06, "loss": 0.0172, "step": 22705 }, { "epoch": 2.4616218560277536, "grad_norm": 0.4826631247997284, "learning_rate": 3.6022051356448573e-06, "loss": 0.0326, "step": 22706 }, { "epoch": 2.4617302688638336, "grad_norm": 1.702402949333191, "learning_rate": 3.60147976207747e-06, "loss": 0.0051, "step": 22707 }, { "epoch": 2.461838681699913, "grad_norm": 0.008802651427686214, "learning_rate": 3.6007543885100833e-06, "loss": 0.0001, "step": 22708 }, { "epoch": 2.461947094535993, "grad_norm": 0.47345802187919617, "learning_rate": 3.600029014942696e-06, "loss": 0.0108, "step": 22709 }, { "epoch": 2.4620555073720727, "grad_norm": 0.3795774579048157, "learning_rate": 3.5993036413753084e-06, "loss": 0.0101, "step": 22710 }, { "epoch": 2.4621639202081527, "grad_norm": 5.945162296295166, "learning_rate": 3.5985782678079214e-06, "loss": 0.0107, "step": 22711 }, { "epoch": 2.4622723330442327, "grad_norm": 4.262836933135986, "learning_rate": 3.5978528942405344e-06, "loss": 0.0278, "step": 22712 }, { "epoch": 2.462380745880312, "grad_norm": 0.43719786405563354, "learning_rate": 3.597127520673147e-06, "loss": 0.0163, "step": 22713 }, { "epoch": 2.462489158716392, "grad_norm": 3.056889057159424, "learning_rate": 3.59640214710576e-06, "loss": 0.0237, "step": 22714 }, { "epoch": 2.4625975715524717, "grad_norm": 0.30161792039871216, "learning_rate": 3.5956767735383725e-06, "loss": 0.0073, "step": 22715 }, { "epoch": 2.4627059843885517, "grad_norm": 1.0001449584960938, "learning_rate": 3.594951399970985e-06, "loss": 0.0141, "step": 22716 }, { "epoch": 2.4628143972246312, "grad_norm": 0.03217961639165878, "learning_rate": 3.5942260264035985e-06, "loss": 0.0002, "step": 22717 }, { "epoch": 2.4629228100607112, "grad_norm": 0.004027911461889744, "learning_rate": 3.593500652836211e-06, "loss": 0.0, "step": 22718 }, { "epoch": 2.4630312228967908, "grad_norm": 0.41156435012817383, "learning_rate": 3.5927752792688236e-06, "loss": 0.0119, "step": 22719 }, { "epoch": 2.4631396357328708, "grad_norm": 0.945364773273468, "learning_rate": 3.592049905701436e-06, "loss": 0.0256, "step": 22720 }, { "epoch": 2.4632480485689507, "grad_norm": 4.2454071044921875, "learning_rate": 3.5913245321340496e-06, "loss": 0.0107, "step": 22721 }, { "epoch": 2.4633564614050303, "grad_norm": 0.034140150994062424, "learning_rate": 3.590599158566662e-06, "loss": 0.0002, "step": 22722 }, { "epoch": 2.4634648742411103, "grad_norm": 0.2508711516857147, "learning_rate": 3.5898737849992747e-06, "loss": 0.0056, "step": 22723 }, { "epoch": 2.46357328707719, "grad_norm": 0.8449491262435913, "learning_rate": 3.5891484114318877e-06, "loss": 0.0026, "step": 22724 }, { "epoch": 2.46368169991327, "grad_norm": 4.569653511047363, "learning_rate": 3.5884230378645007e-06, "loss": 0.0464, "step": 22725 }, { "epoch": 2.4637901127493493, "grad_norm": 0.5523778200149536, "learning_rate": 3.5876976642971133e-06, "loss": 0.018, "step": 22726 }, { "epoch": 2.4638985255854293, "grad_norm": 0.5239619016647339, "learning_rate": 3.5869722907297263e-06, "loss": 0.0142, "step": 22727 }, { "epoch": 2.464006938421509, "grad_norm": 11.849227905273438, "learning_rate": 3.586246917162339e-06, "loss": 0.0191, "step": 22728 }, { "epoch": 2.464115351257589, "grad_norm": 0.08555877953767776, "learning_rate": 3.585521543594952e-06, "loss": 0.0005, "step": 22729 }, { "epoch": 2.464223764093669, "grad_norm": 0.21913930773735046, "learning_rate": 3.584796170027565e-06, "loss": 0.0046, "step": 22730 }, { "epoch": 2.4643321769297484, "grad_norm": 3.8275370597839355, "learning_rate": 3.5840707964601774e-06, "loss": 0.0178, "step": 22731 }, { "epoch": 2.4644405897658284, "grad_norm": 0.35300880670547485, "learning_rate": 3.58334542289279e-06, "loss": 0.0155, "step": 22732 }, { "epoch": 2.464549002601908, "grad_norm": 2.9985105991363525, "learning_rate": 3.5826200493254025e-06, "loss": 0.0076, "step": 22733 }, { "epoch": 2.464657415437988, "grad_norm": 16.422075271606445, "learning_rate": 3.581894675758016e-06, "loss": 0.0339, "step": 22734 }, { "epoch": 2.464765828274068, "grad_norm": 0.008951113559305668, "learning_rate": 3.5811693021906285e-06, "loss": 0.0001, "step": 22735 }, { "epoch": 2.4648742411101474, "grad_norm": 0.7076128125190735, "learning_rate": 3.580443928623241e-06, "loss": 0.0262, "step": 22736 }, { "epoch": 2.4649826539462274, "grad_norm": 0.010840101167559624, "learning_rate": 3.579718555055854e-06, "loss": 0.0001, "step": 22737 }, { "epoch": 2.465091066782307, "grad_norm": 1.221838116645813, "learning_rate": 3.578993181488467e-06, "loss": 0.0064, "step": 22738 }, { "epoch": 2.465199479618387, "grad_norm": 1.294614553451538, "learning_rate": 3.5782678079210796e-06, "loss": 0.0016, "step": 22739 }, { "epoch": 2.4653078924544665, "grad_norm": 0.32687997817993164, "learning_rate": 3.5775424343536926e-06, "loss": 0.0119, "step": 22740 }, { "epoch": 2.4654163052905465, "grad_norm": 1.9045029878616333, "learning_rate": 3.576817060786305e-06, "loss": 0.0029, "step": 22741 }, { "epoch": 2.465524718126626, "grad_norm": 0.09721871465444565, "learning_rate": 3.576091687218918e-06, "loss": 0.0004, "step": 22742 }, { "epoch": 2.465633130962706, "grad_norm": 0.5378553867340088, "learning_rate": 3.575366313651531e-06, "loss": 0.0127, "step": 22743 }, { "epoch": 2.465741543798786, "grad_norm": 6.77162504196167, "learning_rate": 3.5746409400841437e-06, "loss": 0.006, "step": 22744 }, { "epoch": 2.4658499566348655, "grad_norm": 0.39263367652893066, "learning_rate": 3.5739155665167563e-06, "loss": 0.0128, "step": 22745 }, { "epoch": 2.4659583694709455, "grad_norm": 0.35855650901794434, "learning_rate": 3.573190192949369e-06, "loss": 0.0096, "step": 22746 }, { "epoch": 2.466066782307025, "grad_norm": 0.31162405014038086, "learning_rate": 3.5724648193819823e-06, "loss": 0.0101, "step": 22747 }, { "epoch": 2.466175195143105, "grad_norm": 0.1667972207069397, "learning_rate": 3.571739445814595e-06, "loss": 0.0005, "step": 22748 }, { "epoch": 2.4662836079791846, "grad_norm": 3.187558650970459, "learning_rate": 3.5710140722472074e-06, "loss": 0.0109, "step": 22749 }, { "epoch": 2.4663920208152645, "grad_norm": 0.0126413619145751, "learning_rate": 3.5702886986798204e-06, "loss": 0.0002, "step": 22750 }, { "epoch": 2.466500433651344, "grad_norm": 0.03185054287314415, "learning_rate": 3.5695633251124334e-06, "loss": 0.0001, "step": 22751 }, { "epoch": 2.466608846487424, "grad_norm": 0.2754228115081787, "learning_rate": 3.568837951545046e-06, "loss": 0.0062, "step": 22752 }, { "epoch": 2.466717259323504, "grad_norm": 3.3193581104278564, "learning_rate": 3.568112577977659e-06, "loss": 0.0068, "step": 22753 }, { "epoch": 2.4668256721595836, "grad_norm": 0.4888777732849121, "learning_rate": 3.5673872044102715e-06, "loss": 0.018, "step": 22754 }, { "epoch": 2.4669340849956636, "grad_norm": 0.2781293988227844, "learning_rate": 3.5666618308428845e-06, "loss": 0.0087, "step": 22755 }, { "epoch": 2.467042497831743, "grad_norm": 0.01698794774711132, "learning_rate": 3.5659364572754975e-06, "loss": 0.0002, "step": 22756 }, { "epoch": 2.467150910667823, "grad_norm": 0.417673259973526, "learning_rate": 3.56521108370811e-06, "loss": 0.0007, "step": 22757 }, { "epoch": 2.467259323503903, "grad_norm": 0.4075380265712738, "learning_rate": 3.5644857101407226e-06, "loss": 0.0123, "step": 22758 }, { "epoch": 2.4673677363399826, "grad_norm": 0.7036474943161011, "learning_rate": 3.563760336573335e-06, "loss": 0.0216, "step": 22759 }, { "epoch": 2.4674761491760626, "grad_norm": 0.32421237230300903, "learning_rate": 3.5630349630059486e-06, "loss": 0.0062, "step": 22760 }, { "epoch": 2.467584562012142, "grad_norm": 2.347766399383545, "learning_rate": 3.562309589438561e-06, "loss": 0.005, "step": 22761 }, { "epoch": 2.467692974848222, "grad_norm": 0.48816388845443726, "learning_rate": 3.5615842158711737e-06, "loss": 0.0004, "step": 22762 }, { "epoch": 2.4678013876843017, "grad_norm": 15.875360488891602, "learning_rate": 3.5608588423037867e-06, "loss": 0.0433, "step": 22763 }, { "epoch": 2.4679098005203817, "grad_norm": 0.5839839577674866, "learning_rate": 3.5601334687363997e-06, "loss": 0.0316, "step": 22764 }, { "epoch": 2.468018213356461, "grad_norm": 0.2454850822687149, "learning_rate": 3.5594080951690123e-06, "loss": 0.0051, "step": 22765 }, { "epoch": 2.468126626192541, "grad_norm": 0.4111265540122986, "learning_rate": 3.5586827216016253e-06, "loss": 0.0097, "step": 22766 }, { "epoch": 2.468235039028621, "grad_norm": 0.6959185004234314, "learning_rate": 3.557957348034238e-06, "loss": 0.0255, "step": 22767 }, { "epoch": 2.4683434518647007, "grad_norm": 0.2437952309846878, "learning_rate": 3.557231974466851e-06, "loss": 0.0229, "step": 22768 }, { "epoch": 2.4684518647007807, "grad_norm": 0.015018484555184841, "learning_rate": 3.556506600899464e-06, "loss": 0.0001, "step": 22769 }, { "epoch": 2.4685602775368602, "grad_norm": 0.3933238685131073, "learning_rate": 3.5557812273320764e-06, "loss": 0.0007, "step": 22770 }, { "epoch": 2.4686686903729402, "grad_norm": 0.3007928729057312, "learning_rate": 3.555055853764689e-06, "loss": 0.0101, "step": 22771 }, { "epoch": 2.4687771032090198, "grad_norm": 0.3925832211971283, "learning_rate": 3.5543304801973015e-06, "loss": 0.0093, "step": 22772 }, { "epoch": 2.4688855160450998, "grad_norm": 0.36623460054397583, "learning_rate": 3.553605106629915e-06, "loss": 0.0205, "step": 22773 }, { "epoch": 2.4689939288811793, "grad_norm": 5.145102500915527, "learning_rate": 3.5528797330625275e-06, "loss": 0.0113, "step": 22774 }, { "epoch": 2.4691023417172593, "grad_norm": 1.7317613363265991, "learning_rate": 3.55215435949514e-06, "loss": 0.0607, "step": 22775 }, { "epoch": 2.4692107545533393, "grad_norm": 0.025527572259306908, "learning_rate": 3.551428985927753e-06, "loss": 0.0001, "step": 22776 }, { "epoch": 2.469319167389419, "grad_norm": 7.197799205780029, "learning_rate": 3.550703612360366e-06, "loss": 0.0304, "step": 22777 }, { "epoch": 2.469427580225499, "grad_norm": 0.6496888995170593, "learning_rate": 3.5499782387929786e-06, "loss": 0.0102, "step": 22778 }, { "epoch": 2.4695359930615783, "grad_norm": 0.18838916718959808, "learning_rate": 3.5492528652255916e-06, "loss": 0.0007, "step": 22779 }, { "epoch": 2.4696444058976583, "grad_norm": 7.365302085876465, "learning_rate": 3.548527491658204e-06, "loss": 0.0136, "step": 22780 }, { "epoch": 2.4697528187337383, "grad_norm": 0.036875542253255844, "learning_rate": 3.547802118090817e-06, "loss": 0.0003, "step": 22781 }, { "epoch": 2.469861231569818, "grad_norm": 0.9002631306648254, "learning_rate": 3.54707674452343e-06, "loss": 0.0139, "step": 22782 }, { "epoch": 2.469969644405898, "grad_norm": 0.22610969841480255, "learning_rate": 3.5463513709560427e-06, "loss": 0.0045, "step": 22783 }, { "epoch": 2.4700780572419774, "grad_norm": 0.40975135564804077, "learning_rate": 3.5456259973886553e-06, "loss": 0.0165, "step": 22784 }, { "epoch": 2.4701864700780574, "grad_norm": 0.00874768290668726, "learning_rate": 3.544900623821268e-06, "loss": 0.0001, "step": 22785 }, { "epoch": 2.470294882914137, "grad_norm": 0.24781917035579681, "learning_rate": 3.5441752502538813e-06, "loss": 0.001, "step": 22786 }, { "epoch": 2.470403295750217, "grad_norm": 0.15839357674121857, "learning_rate": 3.543449876686494e-06, "loss": 0.0004, "step": 22787 }, { "epoch": 2.4705117085862964, "grad_norm": 0.3776807188987732, "learning_rate": 3.5427245031191064e-06, "loss": 0.0109, "step": 22788 }, { "epoch": 2.4706201214223764, "grad_norm": 1.7971254587173462, "learning_rate": 3.5419991295517194e-06, "loss": 0.0275, "step": 22789 }, { "epoch": 2.4707285342584564, "grad_norm": 0.12751057744026184, "learning_rate": 3.5412737559843324e-06, "loss": 0.0005, "step": 22790 }, { "epoch": 2.470836947094536, "grad_norm": 1.9318927526474, "learning_rate": 3.540548382416945e-06, "loss": 0.032, "step": 22791 }, { "epoch": 2.470945359930616, "grad_norm": 0.0029089057352393866, "learning_rate": 3.539823008849558e-06, "loss": 0.0, "step": 22792 }, { "epoch": 2.4710537727666955, "grad_norm": 0.005804647691547871, "learning_rate": 3.5390976352821705e-06, "loss": 0.0, "step": 22793 }, { "epoch": 2.4711621856027755, "grad_norm": 0.2428751140832901, "learning_rate": 3.5383722617147835e-06, "loss": 0.0053, "step": 22794 }, { "epoch": 2.471270598438855, "grad_norm": 0.008522504940629005, "learning_rate": 3.5376468881473965e-06, "loss": 0.0, "step": 22795 }, { "epoch": 2.471379011274935, "grad_norm": 0.12917464971542358, "learning_rate": 3.536921514580009e-06, "loss": 0.0002, "step": 22796 }, { "epoch": 2.4714874241110145, "grad_norm": 0.2387380450963974, "learning_rate": 3.5361961410126216e-06, "loss": 0.0048, "step": 22797 }, { "epoch": 2.4715958369470945, "grad_norm": 0.017067542299628258, "learning_rate": 3.535470767445234e-06, "loss": 0.0001, "step": 22798 }, { "epoch": 2.4717042497831745, "grad_norm": 0.022817131131887436, "learning_rate": 3.5347453938778476e-06, "loss": 0.0002, "step": 22799 }, { "epoch": 2.471812662619254, "grad_norm": 0.3188517689704895, "learning_rate": 3.53402002031046e-06, "loss": 0.0108, "step": 22800 }, { "epoch": 2.471921075455334, "grad_norm": 1.8732913732528687, "learning_rate": 3.5332946467430728e-06, "loss": 0.0095, "step": 22801 }, { "epoch": 2.4720294882914136, "grad_norm": 0.38330453634262085, "learning_rate": 3.5325692731756857e-06, "loss": 0.0085, "step": 22802 }, { "epoch": 2.4721379011274935, "grad_norm": 0.009759304113686085, "learning_rate": 3.5318438996082987e-06, "loss": 0.0001, "step": 22803 }, { "epoch": 2.4722463139635735, "grad_norm": 0.6379774212837219, "learning_rate": 3.5311185260409113e-06, "loss": 0.0013, "step": 22804 }, { "epoch": 2.472354726799653, "grad_norm": 0.34045010805130005, "learning_rate": 3.5303931524735243e-06, "loss": 0.0122, "step": 22805 }, { "epoch": 2.472463139635733, "grad_norm": 10.954633712768555, "learning_rate": 3.529667778906137e-06, "loss": 0.0489, "step": 22806 }, { "epoch": 2.4725715524718126, "grad_norm": 0.27362579107284546, "learning_rate": 3.52894240533875e-06, "loss": 0.0151, "step": 22807 }, { "epoch": 2.4726799653078926, "grad_norm": 0.4070073962211609, "learning_rate": 3.528217031771363e-06, "loss": 0.0074, "step": 22808 }, { "epoch": 2.472788378143972, "grad_norm": 0.02713371440768242, "learning_rate": 3.5274916582039754e-06, "loss": 0.0001, "step": 22809 }, { "epoch": 2.472896790980052, "grad_norm": 3.1514153480529785, "learning_rate": 3.526766284636588e-06, "loss": 0.0613, "step": 22810 }, { "epoch": 2.4730052038161316, "grad_norm": 0.18242815136909485, "learning_rate": 3.5260409110692005e-06, "loss": 0.0007, "step": 22811 }, { "epoch": 2.4731136166522116, "grad_norm": 0.28633370995521545, "learning_rate": 3.525315537501814e-06, "loss": 0.0097, "step": 22812 }, { "epoch": 2.4732220294882916, "grad_norm": 1.3306584358215332, "learning_rate": 3.5245901639344265e-06, "loss": 0.0062, "step": 22813 }, { "epoch": 2.473330442324371, "grad_norm": 2.8716063499450684, "learning_rate": 3.523864790367039e-06, "loss": 0.0166, "step": 22814 }, { "epoch": 2.473438855160451, "grad_norm": 0.02306467294692993, "learning_rate": 3.523139416799652e-06, "loss": 0.0001, "step": 22815 }, { "epoch": 2.4735472679965307, "grad_norm": 0.015137456357479095, "learning_rate": 3.522414043232265e-06, "loss": 0.0001, "step": 22816 }, { "epoch": 2.4736556808326107, "grad_norm": 0.08028700947761536, "learning_rate": 3.5216886696648776e-06, "loss": 0.0005, "step": 22817 }, { "epoch": 2.47376409366869, "grad_norm": 0.39381369948387146, "learning_rate": 3.5209632960974906e-06, "loss": 0.0119, "step": 22818 }, { "epoch": 2.47387250650477, "grad_norm": 0.032170943915843964, "learning_rate": 3.520237922530103e-06, "loss": 0.0002, "step": 22819 }, { "epoch": 2.4739809193408497, "grad_norm": 1.8388490676879883, "learning_rate": 3.519512548962716e-06, "loss": 0.0607, "step": 22820 }, { "epoch": 2.4740893321769297, "grad_norm": 6.616471290588379, "learning_rate": 3.518787175395329e-06, "loss": 0.0542, "step": 22821 }, { "epoch": 2.4741977450130097, "grad_norm": 0.03933209925889969, "learning_rate": 3.5180618018279417e-06, "loss": 0.0001, "step": 22822 }, { "epoch": 2.4743061578490892, "grad_norm": 0.34361913800239563, "learning_rate": 3.5173364282605543e-06, "loss": 0.0201, "step": 22823 }, { "epoch": 2.4744145706851692, "grad_norm": 0.33443883061408997, "learning_rate": 3.516611054693167e-06, "loss": 0.0121, "step": 22824 }, { "epoch": 2.4745229835212488, "grad_norm": 0.008904246613383293, "learning_rate": 3.5158856811257803e-06, "loss": 0.0, "step": 22825 }, { "epoch": 2.4746313963573288, "grad_norm": 0.30601510405540466, "learning_rate": 3.515160307558393e-06, "loss": 0.0135, "step": 22826 }, { "epoch": 2.4747398091934087, "grad_norm": 0.020866576582193375, "learning_rate": 3.5144349339910054e-06, "loss": 0.0003, "step": 22827 }, { "epoch": 2.4748482220294883, "grad_norm": 0.007448721677064896, "learning_rate": 3.5137095604236184e-06, "loss": 0.0001, "step": 22828 }, { "epoch": 2.4749566348655683, "grad_norm": 0.30547356605529785, "learning_rate": 3.5129841868562314e-06, "loss": 0.0062, "step": 22829 }, { "epoch": 2.475065047701648, "grad_norm": 11.597681999206543, "learning_rate": 3.512258813288844e-06, "loss": 0.0241, "step": 22830 }, { "epoch": 2.475173460537728, "grad_norm": 10.859224319458008, "learning_rate": 3.511533439721457e-06, "loss": 0.0141, "step": 22831 }, { "epoch": 2.4752818733738073, "grad_norm": 0.5508835315704346, "learning_rate": 3.5108080661540695e-06, "loss": 0.0177, "step": 22832 }, { "epoch": 2.4753902862098873, "grad_norm": 3.5614376068115234, "learning_rate": 3.5100826925866825e-06, "loss": 0.0232, "step": 22833 }, { "epoch": 2.475498699045967, "grad_norm": 0.04408225789666176, "learning_rate": 3.5093573190192955e-06, "loss": 0.0003, "step": 22834 }, { "epoch": 2.475607111882047, "grad_norm": 1.1797609329223633, "learning_rate": 3.508631945451908e-06, "loss": 0.0387, "step": 22835 }, { "epoch": 2.475715524718127, "grad_norm": 2.663343667984009, "learning_rate": 3.5079065718845206e-06, "loss": 0.0153, "step": 22836 }, { "epoch": 2.4758239375542064, "grad_norm": 0.2831685245037079, "learning_rate": 3.507181198317134e-06, "loss": 0.0049, "step": 22837 }, { "epoch": 2.4759323503902864, "grad_norm": 0.456162691116333, "learning_rate": 3.5064558247497466e-06, "loss": 0.0009, "step": 22838 }, { "epoch": 2.476040763226366, "grad_norm": 0.6392841935157776, "learning_rate": 3.505730451182359e-06, "loss": 0.0061, "step": 22839 }, { "epoch": 2.476149176062446, "grad_norm": 0.39284747838974, "learning_rate": 3.5050050776149718e-06, "loss": 0.0048, "step": 22840 }, { "epoch": 2.4762575888985254, "grad_norm": 0.695188045501709, "learning_rate": 3.5042797040475847e-06, "loss": 0.0237, "step": 22841 }, { "epoch": 2.4763660017346054, "grad_norm": 0.1542217880487442, "learning_rate": 3.5035543304801977e-06, "loss": 0.0004, "step": 22842 }, { "epoch": 2.476474414570685, "grad_norm": 0.10831990838050842, "learning_rate": 3.5028289569128103e-06, "loss": 0.0004, "step": 22843 }, { "epoch": 2.476582827406765, "grad_norm": 0.03898804634809494, "learning_rate": 3.5021035833454233e-06, "loss": 0.0001, "step": 22844 }, { "epoch": 2.476691240242845, "grad_norm": 6.468360900878906, "learning_rate": 3.501378209778036e-06, "loss": 0.0755, "step": 22845 }, { "epoch": 2.4767996530789245, "grad_norm": 0.438547283411026, "learning_rate": 3.500652836210649e-06, "loss": 0.009, "step": 22846 }, { "epoch": 2.4769080659150045, "grad_norm": 0.6207109093666077, "learning_rate": 3.499927462643262e-06, "loss": 0.0299, "step": 22847 }, { "epoch": 2.477016478751084, "grad_norm": 0.2874473035335541, "learning_rate": 3.4992020890758744e-06, "loss": 0.0096, "step": 22848 }, { "epoch": 2.477124891587164, "grad_norm": 0.2257784903049469, "learning_rate": 3.498476715508487e-06, "loss": 0.0046, "step": 22849 }, { "epoch": 2.4772333044232435, "grad_norm": 0.05424002185463905, "learning_rate": 3.4977513419411004e-06, "loss": 0.0003, "step": 22850 }, { "epoch": 2.4773417172593235, "grad_norm": 3.0790717601776123, "learning_rate": 3.497025968373713e-06, "loss": 0.0035, "step": 22851 }, { "epoch": 2.477450130095403, "grad_norm": 0.0034003325272351503, "learning_rate": 3.4963005948063255e-06, "loss": 0.0, "step": 22852 }, { "epoch": 2.477558542931483, "grad_norm": 0.4426601529121399, "learning_rate": 3.495575221238938e-06, "loss": 0.0085, "step": 22853 }, { "epoch": 2.477666955767563, "grad_norm": 2.1339073181152344, "learning_rate": 3.494849847671551e-06, "loss": 0.0415, "step": 22854 }, { "epoch": 2.4777753686036426, "grad_norm": 0.08691106736660004, "learning_rate": 3.494124474104164e-06, "loss": 0.0002, "step": 22855 }, { "epoch": 2.4778837814397225, "grad_norm": 0.23711931705474854, "learning_rate": 3.4933991005367766e-06, "loss": 0.0046, "step": 22856 }, { "epoch": 2.477992194275802, "grad_norm": 0.38514211773872375, "learning_rate": 3.4926737269693896e-06, "loss": 0.0066, "step": 22857 }, { "epoch": 2.478100607111882, "grad_norm": 1.8054814338684082, "learning_rate": 3.491948353402002e-06, "loss": 0.0491, "step": 22858 }, { "epoch": 2.478209019947962, "grad_norm": 0.4702838361263275, "learning_rate": 3.491222979834615e-06, "loss": 0.0121, "step": 22859 }, { "epoch": 2.4783174327840416, "grad_norm": 0.10798002779483795, "learning_rate": 3.490497606267228e-06, "loss": 0.0005, "step": 22860 }, { "epoch": 2.4784258456201216, "grad_norm": 0.24083276093006134, "learning_rate": 3.4897722326998407e-06, "loss": 0.0007, "step": 22861 }, { "epoch": 2.478534258456201, "grad_norm": 0.0053581353276968, "learning_rate": 3.4890468591324533e-06, "loss": 0.0001, "step": 22862 }, { "epoch": 2.478642671292281, "grad_norm": 4.312280654907227, "learning_rate": 3.4883214855650667e-06, "loss": 0.0079, "step": 22863 }, { "epoch": 2.4787510841283606, "grad_norm": 28.38786506652832, "learning_rate": 3.4875961119976793e-06, "loss": 0.058, "step": 22864 }, { "epoch": 2.4788594969644406, "grad_norm": 0.11624592542648315, "learning_rate": 3.486870738430292e-06, "loss": 0.0003, "step": 22865 }, { "epoch": 2.47896790980052, "grad_norm": 0.25227171182632446, "learning_rate": 3.4861453648629044e-06, "loss": 0.0011, "step": 22866 }, { "epoch": 2.4790763226366, "grad_norm": 0.018728181719779968, "learning_rate": 3.4854199912955174e-06, "loss": 0.0001, "step": 22867 }, { "epoch": 2.47918473547268, "grad_norm": 0.3375597298145294, "learning_rate": 3.4846946177281304e-06, "loss": 0.012, "step": 22868 }, { "epoch": 2.4792931483087597, "grad_norm": 9.40742301940918, "learning_rate": 3.483969244160743e-06, "loss": 0.0091, "step": 22869 }, { "epoch": 2.4794015611448397, "grad_norm": 0.3911983370780945, "learning_rate": 3.483243870593356e-06, "loss": 0.0015, "step": 22870 }, { "epoch": 2.479509973980919, "grad_norm": 0.35632285475730896, "learning_rate": 3.4825184970259685e-06, "loss": 0.0062, "step": 22871 }, { "epoch": 2.479618386816999, "grad_norm": 0.22833481431007385, "learning_rate": 3.4817931234585815e-06, "loss": 0.0046, "step": 22872 }, { "epoch": 2.4797267996530787, "grad_norm": 0.24142274260520935, "learning_rate": 3.4810677498911945e-06, "loss": 0.0084, "step": 22873 }, { "epoch": 2.4798352124891587, "grad_norm": 0.013011142611503601, "learning_rate": 3.480342376323807e-06, "loss": 0.0001, "step": 22874 }, { "epoch": 2.4799436253252383, "grad_norm": 0.42683273553848267, "learning_rate": 3.4796170027564196e-06, "loss": 0.0161, "step": 22875 }, { "epoch": 2.4800520381613183, "grad_norm": 0.05877288058400154, "learning_rate": 3.478891629189033e-06, "loss": 0.0005, "step": 22876 }, { "epoch": 2.4801604509973982, "grad_norm": 2.2075700759887695, "learning_rate": 3.4781662556216456e-06, "loss": 0.0587, "step": 22877 }, { "epoch": 2.4802688638334778, "grad_norm": 0.3676646053791046, "learning_rate": 3.477440882054258e-06, "loss": 0.0074, "step": 22878 }, { "epoch": 2.4803772766695578, "grad_norm": 0.60124272108078, "learning_rate": 3.4767155084868708e-06, "loss": 0.0171, "step": 22879 }, { "epoch": 2.4804856895056373, "grad_norm": 0.1114938035607338, "learning_rate": 3.4759901349194837e-06, "loss": 0.0007, "step": 22880 }, { "epoch": 2.4805941023417173, "grad_norm": 0.422536164522171, "learning_rate": 3.4752647613520967e-06, "loss": 0.0011, "step": 22881 }, { "epoch": 2.4807025151777973, "grad_norm": 0.2534453868865967, "learning_rate": 3.4745393877847093e-06, "loss": 0.0055, "step": 22882 }, { "epoch": 2.480810928013877, "grad_norm": 0.11466923356056213, "learning_rate": 3.4738140142173223e-06, "loss": 0.0006, "step": 22883 }, { "epoch": 2.480919340849957, "grad_norm": 0.3221324384212494, "learning_rate": 3.473088640649935e-06, "loss": 0.0117, "step": 22884 }, { "epoch": 2.4810277536860363, "grad_norm": 10.391755104064941, "learning_rate": 3.472363267082548e-06, "loss": 0.0483, "step": 22885 }, { "epoch": 2.4811361665221163, "grad_norm": 1.032242774963379, "learning_rate": 3.471637893515161e-06, "loss": 0.0026, "step": 22886 }, { "epoch": 2.481244579358196, "grad_norm": 0.7145978808403015, "learning_rate": 3.4709125199477734e-06, "loss": 0.0278, "step": 22887 }, { "epoch": 2.481352992194276, "grad_norm": 0.014294926077127457, "learning_rate": 3.470187146380386e-06, "loss": 0.0001, "step": 22888 }, { "epoch": 2.4814614050303554, "grad_norm": 9.930357933044434, "learning_rate": 3.4694617728129994e-06, "loss": 0.0366, "step": 22889 }, { "epoch": 2.4815698178664354, "grad_norm": 0.43870288133621216, "learning_rate": 3.468736399245612e-06, "loss": 0.0162, "step": 22890 }, { "epoch": 2.4816782307025154, "grad_norm": 0.1059817522764206, "learning_rate": 3.4680110256782245e-06, "loss": 0.0005, "step": 22891 }, { "epoch": 2.481786643538595, "grad_norm": 0.03759028762578964, "learning_rate": 3.467285652110837e-06, "loss": 0.0002, "step": 22892 }, { "epoch": 2.481895056374675, "grad_norm": 0.49576306343078613, "learning_rate": 3.46656027854345e-06, "loss": 0.002, "step": 22893 }, { "epoch": 2.4820034692107544, "grad_norm": 0.14718808233737946, "learning_rate": 3.465834904976063e-06, "loss": 0.0124, "step": 22894 }, { "epoch": 2.4821118820468344, "grad_norm": 3.579749822616577, "learning_rate": 3.4651095314086756e-06, "loss": 0.0195, "step": 22895 }, { "epoch": 2.482220294882914, "grad_norm": 1.9809523820877075, "learning_rate": 3.4643841578412886e-06, "loss": 0.0118, "step": 22896 }, { "epoch": 2.482328707718994, "grad_norm": 0.25136056542396545, "learning_rate": 3.463658784273901e-06, "loss": 0.0011, "step": 22897 }, { "epoch": 2.4824371205550735, "grad_norm": 0.2494712769985199, "learning_rate": 3.462933410706514e-06, "loss": 0.0053, "step": 22898 }, { "epoch": 2.4825455333911535, "grad_norm": 0.021534424275159836, "learning_rate": 3.462208037139127e-06, "loss": 0.0002, "step": 22899 }, { "epoch": 2.4826539462272335, "grad_norm": 0.43169891834259033, "learning_rate": 3.4614826635717397e-06, "loss": 0.0156, "step": 22900 }, { "epoch": 2.482762359063313, "grad_norm": 4.72859001159668, "learning_rate": 3.4607572900043523e-06, "loss": 0.0364, "step": 22901 }, { "epoch": 2.482870771899393, "grad_norm": 0.41578537225723267, "learning_rate": 3.4600319164369657e-06, "loss": 0.0076, "step": 22902 }, { "epoch": 2.4829791847354725, "grad_norm": 1.3697903156280518, "learning_rate": 3.4593065428695783e-06, "loss": 0.0009, "step": 22903 }, { "epoch": 2.4830875975715525, "grad_norm": 0.034923676401376724, "learning_rate": 3.458581169302191e-06, "loss": 0.0003, "step": 22904 }, { "epoch": 2.4831960104076325, "grad_norm": 4.337749481201172, "learning_rate": 3.4578557957348034e-06, "loss": 0.0174, "step": 22905 }, { "epoch": 2.483304423243712, "grad_norm": 2.617602586746216, "learning_rate": 3.4571304221674164e-06, "loss": 0.0401, "step": 22906 }, { "epoch": 2.483412836079792, "grad_norm": 0.4705427885055542, "learning_rate": 3.4564050486000294e-06, "loss": 0.0235, "step": 22907 }, { "epoch": 2.4835212489158716, "grad_norm": 10.402480125427246, "learning_rate": 3.455679675032642e-06, "loss": 0.0324, "step": 22908 }, { "epoch": 2.4836296617519515, "grad_norm": 8.865385055541992, "learning_rate": 3.454954301465255e-06, "loss": 0.0053, "step": 22909 }, { "epoch": 2.483738074588031, "grad_norm": 0.27293094992637634, "learning_rate": 3.4542289278978675e-06, "loss": 0.0055, "step": 22910 }, { "epoch": 2.483846487424111, "grad_norm": 0.2855783700942993, "learning_rate": 3.4535035543304805e-06, "loss": 0.0114, "step": 22911 }, { "epoch": 2.4839549002601906, "grad_norm": 0.2631775438785553, "learning_rate": 3.4527781807630935e-06, "loss": 0.0055, "step": 22912 }, { "epoch": 2.4840633130962706, "grad_norm": 12.102896690368652, "learning_rate": 3.452052807195706e-06, "loss": 0.008, "step": 22913 }, { "epoch": 2.4841717259323506, "grad_norm": 1.1916561126708984, "learning_rate": 3.4513274336283186e-06, "loss": 0.0024, "step": 22914 }, { "epoch": 2.48428013876843, "grad_norm": 0.14807063341140747, "learning_rate": 3.450602060060932e-06, "loss": 0.0007, "step": 22915 }, { "epoch": 2.48438855160451, "grad_norm": 0.02701374888420105, "learning_rate": 3.4498766864935446e-06, "loss": 0.0002, "step": 22916 }, { "epoch": 2.4844969644405897, "grad_norm": 0.01701933890581131, "learning_rate": 3.449151312926157e-06, "loss": 0.0001, "step": 22917 }, { "epoch": 2.4846053772766696, "grad_norm": 0.48872682452201843, "learning_rate": 3.4484259393587698e-06, "loss": 0.0176, "step": 22918 }, { "epoch": 2.484713790112749, "grad_norm": 19.376888275146484, "learning_rate": 3.4477005657913828e-06, "loss": 0.029, "step": 22919 }, { "epoch": 2.484822202948829, "grad_norm": 0.06589343398809433, "learning_rate": 3.4469751922239957e-06, "loss": 0.0003, "step": 22920 }, { "epoch": 2.4849306157849087, "grad_norm": 0.005790641065686941, "learning_rate": 3.4462498186566083e-06, "loss": 0.0, "step": 22921 }, { "epoch": 2.4850390286209887, "grad_norm": 2.1066489219665527, "learning_rate": 3.4455244450892213e-06, "loss": 0.0086, "step": 22922 }, { "epoch": 2.4851474414570687, "grad_norm": 0.4190969169139862, "learning_rate": 3.444799071521834e-06, "loss": 0.0095, "step": 22923 }, { "epoch": 2.485255854293148, "grad_norm": 0.2665720582008362, "learning_rate": 3.444073697954447e-06, "loss": 0.0053, "step": 22924 }, { "epoch": 2.485364267129228, "grad_norm": 0.4218137860298157, "learning_rate": 3.44334832438706e-06, "loss": 0.0096, "step": 22925 }, { "epoch": 2.4854726799653077, "grad_norm": 4.176661014556885, "learning_rate": 3.4426229508196724e-06, "loss": 0.0639, "step": 22926 }, { "epoch": 2.4855810928013877, "grad_norm": 0.48941197991371155, "learning_rate": 3.441897577252285e-06, "loss": 0.0104, "step": 22927 }, { "epoch": 2.4856895056374677, "grad_norm": 0.1061115711927414, "learning_rate": 3.4411722036848984e-06, "loss": 0.0004, "step": 22928 }, { "epoch": 2.4857979184735473, "grad_norm": 0.24865667521953583, "learning_rate": 3.440446830117511e-06, "loss": 0.0049, "step": 22929 }, { "epoch": 2.4859063313096272, "grad_norm": 0.05583154037594795, "learning_rate": 3.4397214565501235e-06, "loss": 0.0004, "step": 22930 }, { "epoch": 2.486014744145707, "grad_norm": 1.1985366344451904, "learning_rate": 3.438996082982736e-06, "loss": 0.0024, "step": 22931 }, { "epoch": 2.4861231569817868, "grad_norm": 0.32027021050453186, "learning_rate": 3.4382707094153495e-06, "loss": 0.0197, "step": 22932 }, { "epoch": 2.4862315698178663, "grad_norm": 1.7801653146743774, "learning_rate": 3.437545335847962e-06, "loss": 0.0017, "step": 22933 }, { "epoch": 2.4863399826539463, "grad_norm": 0.5952564477920532, "learning_rate": 3.4368199622805746e-06, "loss": 0.0215, "step": 22934 }, { "epoch": 2.486448395490026, "grad_norm": 0.19685183465480804, "learning_rate": 3.4360945887131876e-06, "loss": 0.001, "step": 22935 }, { "epoch": 2.486556808326106, "grad_norm": 0.30234628915786743, "learning_rate": 3.4353692151458e-06, "loss": 0.0047, "step": 22936 }, { "epoch": 2.486665221162186, "grad_norm": 0.37712904810905457, "learning_rate": 3.434643841578413e-06, "loss": 0.0209, "step": 22937 }, { "epoch": 2.4867736339982653, "grad_norm": 6.209924221038818, "learning_rate": 3.433918468011026e-06, "loss": 0.0206, "step": 22938 }, { "epoch": 2.4868820468343453, "grad_norm": 0.29150211811065674, "learning_rate": 3.4331930944436387e-06, "loss": 0.01, "step": 22939 }, { "epoch": 2.486990459670425, "grad_norm": 0.1570671945810318, "learning_rate": 3.4324677208762513e-06, "loss": 0.0007, "step": 22940 }, { "epoch": 2.487098872506505, "grad_norm": 3.2786505222320557, "learning_rate": 3.4317423473088647e-06, "loss": 0.0251, "step": 22941 }, { "epoch": 2.4872072853425844, "grad_norm": 0.3112802505493164, "learning_rate": 3.4310169737414773e-06, "loss": 0.0068, "step": 22942 }, { "epoch": 2.4873156981786644, "grad_norm": 4.189012050628662, "learning_rate": 3.43029160017409e-06, "loss": 0.0364, "step": 22943 }, { "epoch": 2.487424111014744, "grad_norm": 7.389951705932617, "learning_rate": 3.4295662266067024e-06, "loss": 0.0362, "step": 22944 }, { "epoch": 2.487532523850824, "grad_norm": 5.308841705322266, "learning_rate": 3.428840853039316e-06, "loss": 0.0193, "step": 22945 }, { "epoch": 2.487640936686904, "grad_norm": 0.024463269859552383, "learning_rate": 3.4281154794719284e-06, "loss": 0.0002, "step": 22946 }, { "epoch": 2.4877493495229834, "grad_norm": 0.02227555587887764, "learning_rate": 3.427390105904541e-06, "loss": 0.0002, "step": 22947 }, { "epoch": 2.4878577623590634, "grad_norm": 3.282927989959717, "learning_rate": 3.426664732337154e-06, "loss": 0.0229, "step": 22948 }, { "epoch": 2.487966175195143, "grad_norm": 9.793598175048828, "learning_rate": 3.4259393587697665e-06, "loss": 0.0383, "step": 22949 }, { "epoch": 2.488074588031223, "grad_norm": 0.6049354076385498, "learning_rate": 3.4252139852023795e-06, "loss": 0.0244, "step": 22950 }, { "epoch": 2.488183000867303, "grad_norm": 0.031456656754016876, "learning_rate": 3.4244886116349925e-06, "loss": 0.0001, "step": 22951 }, { "epoch": 2.4882914137033825, "grad_norm": 2.4023220539093018, "learning_rate": 3.423763238067605e-06, "loss": 0.0488, "step": 22952 }, { "epoch": 2.4883998265394625, "grad_norm": 0.33889132738113403, "learning_rate": 3.4230378645002176e-06, "loss": 0.0135, "step": 22953 }, { "epoch": 2.488508239375542, "grad_norm": 0.3055971562862396, "learning_rate": 3.422312490932831e-06, "loss": 0.0072, "step": 22954 }, { "epoch": 2.488616652211622, "grad_norm": 0.01028948649764061, "learning_rate": 3.4215871173654436e-06, "loss": 0.0001, "step": 22955 }, { "epoch": 2.4887250650477015, "grad_norm": 0.25360554456710815, "learning_rate": 3.420861743798056e-06, "loss": 0.0051, "step": 22956 }, { "epoch": 2.4888334778837815, "grad_norm": 0.6206071376800537, "learning_rate": 3.4201363702306688e-06, "loss": 0.0143, "step": 22957 }, { "epoch": 2.488941890719861, "grad_norm": 24.897640228271484, "learning_rate": 3.419410996663282e-06, "loss": 0.0407, "step": 22958 }, { "epoch": 2.489050303555941, "grad_norm": 0.28046971559524536, "learning_rate": 3.4186856230958947e-06, "loss": 0.006, "step": 22959 }, { "epoch": 2.489158716392021, "grad_norm": 0.5182101726531982, "learning_rate": 3.4179602495285073e-06, "loss": 0.0141, "step": 22960 }, { "epoch": 2.4892671292281006, "grad_norm": 0.3336910307407379, "learning_rate": 3.4172348759611203e-06, "loss": 0.006, "step": 22961 }, { "epoch": 2.4893755420641805, "grad_norm": 0.13761238753795624, "learning_rate": 3.416509502393733e-06, "loss": 0.0005, "step": 22962 }, { "epoch": 2.48948395490026, "grad_norm": 0.08279262483119965, "learning_rate": 3.415784128826346e-06, "loss": 0.0109, "step": 22963 }, { "epoch": 2.48959236773634, "grad_norm": 0.3276473581790924, "learning_rate": 3.415058755258959e-06, "loss": 0.0115, "step": 22964 }, { "epoch": 2.4897007805724196, "grad_norm": 0.04736987501382828, "learning_rate": 3.4143333816915714e-06, "loss": 0.0002, "step": 22965 }, { "epoch": 2.4898091934084996, "grad_norm": 19.898330688476562, "learning_rate": 3.413608008124184e-06, "loss": 0.0381, "step": 22966 }, { "epoch": 2.489917606244579, "grad_norm": 0.20819085836410522, "learning_rate": 3.4128826345567974e-06, "loss": 0.0033, "step": 22967 }, { "epoch": 2.490026019080659, "grad_norm": 3.834042549133301, "learning_rate": 3.41215726098941e-06, "loss": 0.0256, "step": 22968 }, { "epoch": 2.490134431916739, "grad_norm": 6.756174564361572, "learning_rate": 3.4114318874220225e-06, "loss": 0.0137, "step": 22969 }, { "epoch": 2.4902428447528187, "grad_norm": 8.55062198638916, "learning_rate": 3.410706513854635e-06, "loss": 0.0312, "step": 22970 }, { "epoch": 2.4903512575888986, "grad_norm": 0.025288265198469162, "learning_rate": 3.4099811402872485e-06, "loss": 0.0002, "step": 22971 }, { "epoch": 2.490459670424978, "grad_norm": 0.318920373916626, "learning_rate": 3.409255766719861e-06, "loss": 0.0072, "step": 22972 }, { "epoch": 2.490568083261058, "grad_norm": 3.8029959201812744, "learning_rate": 3.4085303931524736e-06, "loss": 0.0041, "step": 22973 }, { "epoch": 2.490676496097138, "grad_norm": 0.27863964438438416, "learning_rate": 3.4078050195850866e-06, "loss": 0.006, "step": 22974 }, { "epoch": 2.4907849089332177, "grad_norm": 0.2551056742668152, "learning_rate": 3.407079646017699e-06, "loss": 0.0005, "step": 22975 }, { "epoch": 2.4908933217692977, "grad_norm": 0.6287941932678223, "learning_rate": 3.406354272450312e-06, "loss": 0.011, "step": 22976 }, { "epoch": 2.491001734605377, "grad_norm": 0.5244781970977783, "learning_rate": 3.405628898882925e-06, "loss": 0.0057, "step": 22977 }, { "epoch": 2.491110147441457, "grad_norm": 0.8584980964660645, "learning_rate": 3.4049035253155378e-06, "loss": 0.0151, "step": 22978 }, { "epoch": 2.4912185602775367, "grad_norm": 4.535395622253418, "learning_rate": 3.4041781517481503e-06, "loss": 0.027, "step": 22979 }, { "epoch": 2.4913269731136167, "grad_norm": 4.046253204345703, "learning_rate": 3.4034527781807637e-06, "loss": 0.0187, "step": 22980 }, { "epoch": 2.4914353859496963, "grad_norm": 0.4281499981880188, "learning_rate": 3.4027274046133763e-06, "loss": 0.0163, "step": 22981 }, { "epoch": 2.4915437987857763, "grad_norm": 0.023623475804924965, "learning_rate": 3.402002031045989e-06, "loss": 0.0002, "step": 22982 }, { "epoch": 2.4916522116218562, "grad_norm": 4.568417072296143, "learning_rate": 3.4012766574786014e-06, "loss": 0.0175, "step": 22983 }, { "epoch": 2.491760624457936, "grad_norm": 1.6596423387527466, "learning_rate": 3.400551283911215e-06, "loss": 0.0422, "step": 22984 }, { "epoch": 2.4918690372940158, "grad_norm": 0.08350665122270584, "learning_rate": 3.3998259103438274e-06, "loss": 0.0003, "step": 22985 }, { "epoch": 2.4919774501300953, "grad_norm": 0.0067809708416461945, "learning_rate": 3.39910053677644e-06, "loss": 0.0001, "step": 22986 }, { "epoch": 2.4920858629661753, "grad_norm": 0.2909640371799469, "learning_rate": 3.398375163209053e-06, "loss": 0.0059, "step": 22987 }, { "epoch": 2.492194275802255, "grad_norm": 0.33370885252952576, "learning_rate": 3.3976497896416655e-06, "loss": 0.0055, "step": 22988 }, { "epoch": 2.492302688638335, "grad_norm": 0.22061939537525177, "learning_rate": 3.3969244160742785e-06, "loss": 0.0116, "step": 22989 }, { "epoch": 2.4924111014744144, "grad_norm": 0.44169285893440247, "learning_rate": 3.3961990425068915e-06, "loss": 0.0014, "step": 22990 }, { "epoch": 2.4925195143104943, "grad_norm": 2.14898681640625, "learning_rate": 3.395473668939504e-06, "loss": 0.0053, "step": 22991 }, { "epoch": 2.4926279271465743, "grad_norm": 0.1650063693523407, "learning_rate": 3.3947482953721167e-06, "loss": 0.0006, "step": 22992 }, { "epoch": 2.492736339982654, "grad_norm": 0.42023104429244995, "learning_rate": 3.39402292180473e-06, "loss": 0.0226, "step": 22993 }, { "epoch": 2.492844752818734, "grad_norm": 2.717668056488037, "learning_rate": 3.3932975482373426e-06, "loss": 0.005, "step": 22994 }, { "epoch": 2.4929531656548134, "grad_norm": 0.23321090638637543, "learning_rate": 3.392572174669955e-06, "loss": 0.0145, "step": 22995 }, { "epoch": 2.4930615784908934, "grad_norm": 0.3756047487258911, "learning_rate": 3.3918468011025678e-06, "loss": 0.0122, "step": 22996 }, { "epoch": 2.493169991326973, "grad_norm": 0.478718101978302, "learning_rate": 3.391121427535181e-06, "loss": 0.0125, "step": 22997 }, { "epoch": 2.493278404163053, "grad_norm": 2.8134877681732178, "learning_rate": 3.3903960539677937e-06, "loss": 0.0559, "step": 22998 }, { "epoch": 2.4933868169991324, "grad_norm": 14.824816703796387, "learning_rate": 3.3896706804004063e-06, "loss": 0.0308, "step": 22999 }, { "epoch": 2.4934952298352124, "grad_norm": 2.056875705718994, "learning_rate": 3.3889453068330193e-06, "loss": 0.0042, "step": 23000 }, { "epoch": 2.4936036426712924, "grad_norm": 2.0653915405273438, "learning_rate": 3.388219933265632e-06, "loss": 0.0199, "step": 23001 }, { "epoch": 2.493712055507372, "grad_norm": 2.723787546157837, "learning_rate": 3.387494559698245e-06, "loss": 0.05, "step": 23002 }, { "epoch": 2.493820468343452, "grad_norm": 0.4624936878681183, "learning_rate": 3.386769186130858e-06, "loss": 0.0161, "step": 23003 }, { "epoch": 2.4939288811795315, "grad_norm": 0.27854689955711365, "learning_rate": 3.3860438125634704e-06, "loss": 0.006, "step": 23004 }, { "epoch": 2.4940372940156115, "grad_norm": 0.3378041386604309, "learning_rate": 3.385318438996083e-06, "loss": 0.01, "step": 23005 }, { "epoch": 2.4941457068516915, "grad_norm": 0.6466711163520813, "learning_rate": 3.3845930654286964e-06, "loss": 0.0018, "step": 23006 }, { "epoch": 2.494254119687771, "grad_norm": 0.2784169018268585, "learning_rate": 3.383867691861309e-06, "loss": 0.0143, "step": 23007 }, { "epoch": 2.494362532523851, "grad_norm": 6.972652912139893, "learning_rate": 3.3831423182939215e-06, "loss": 0.0041, "step": 23008 }, { "epoch": 2.4944709453599305, "grad_norm": 0.24255290627479553, "learning_rate": 3.382416944726534e-06, "loss": 0.0053, "step": 23009 }, { "epoch": 2.4945793581960105, "grad_norm": 0.5246735215187073, "learning_rate": 3.3816915711591475e-06, "loss": 0.0166, "step": 23010 }, { "epoch": 2.49468777103209, "grad_norm": 0.262738436460495, "learning_rate": 3.38096619759176e-06, "loss": 0.0118, "step": 23011 }, { "epoch": 2.49479618386817, "grad_norm": 0.0333026647567749, "learning_rate": 3.3802408240243726e-06, "loss": 0.0003, "step": 23012 }, { "epoch": 2.4949045967042496, "grad_norm": 0.028545591980218887, "learning_rate": 3.3795154504569856e-06, "loss": 0.0002, "step": 23013 }, { "epoch": 2.4950130095403296, "grad_norm": 3.8009660243988037, "learning_rate": 3.378790076889598e-06, "loss": 0.0075, "step": 23014 }, { "epoch": 2.4951214223764095, "grad_norm": 0.3601698875427246, "learning_rate": 3.378064703322211e-06, "loss": 0.0113, "step": 23015 }, { "epoch": 2.495229835212489, "grad_norm": 2.3183705806732178, "learning_rate": 3.377339329754824e-06, "loss": 0.0489, "step": 23016 }, { "epoch": 2.495338248048569, "grad_norm": 1.912192702293396, "learning_rate": 3.3766139561874368e-06, "loss": 0.0073, "step": 23017 }, { "epoch": 2.4954466608846486, "grad_norm": 4.2010297775268555, "learning_rate": 3.3758885826200493e-06, "loss": 0.0207, "step": 23018 }, { "epoch": 2.4955550737207286, "grad_norm": 0.3563844859600067, "learning_rate": 3.3751632090526627e-06, "loss": 0.0134, "step": 23019 }, { "epoch": 2.495663486556808, "grad_norm": 0.011577019467949867, "learning_rate": 3.3744378354852753e-06, "loss": 0.0001, "step": 23020 }, { "epoch": 2.495771899392888, "grad_norm": 2.9022817611694336, "learning_rate": 3.373712461917888e-06, "loss": 0.0088, "step": 23021 }, { "epoch": 2.4958803122289677, "grad_norm": 0.3007173240184784, "learning_rate": 3.3729870883505004e-06, "loss": 0.0092, "step": 23022 }, { "epoch": 2.4959887250650477, "grad_norm": 0.3068695068359375, "learning_rate": 3.372261714783114e-06, "loss": 0.0072, "step": 23023 }, { "epoch": 2.4960971379011276, "grad_norm": 0.04319939389824867, "learning_rate": 3.3715363412157264e-06, "loss": 0.0002, "step": 23024 }, { "epoch": 2.496205550737207, "grad_norm": 1.536226749420166, "learning_rate": 3.370810967648339e-06, "loss": 0.0099, "step": 23025 }, { "epoch": 2.496313963573287, "grad_norm": 0.34551435708999634, "learning_rate": 3.370085594080952e-06, "loss": 0.0112, "step": 23026 }, { "epoch": 2.4964223764093667, "grad_norm": 2.999812126159668, "learning_rate": 3.369360220513565e-06, "loss": 0.0195, "step": 23027 }, { "epoch": 2.4965307892454467, "grad_norm": 0.6836624145507812, "learning_rate": 3.3686348469461775e-06, "loss": 0.0007, "step": 23028 }, { "epoch": 2.4966392020815267, "grad_norm": 0.37304261326789856, "learning_rate": 3.3679094733787905e-06, "loss": 0.0015, "step": 23029 }, { "epoch": 2.496747614917606, "grad_norm": 0.059770617634058, "learning_rate": 3.367184099811403e-06, "loss": 0.0002, "step": 23030 }, { "epoch": 2.496856027753686, "grad_norm": 4.4477338790893555, "learning_rate": 3.3664587262440157e-06, "loss": 0.0181, "step": 23031 }, { "epoch": 2.4969644405897657, "grad_norm": 0.07296739518642426, "learning_rate": 3.365733352676629e-06, "loss": 0.0002, "step": 23032 }, { "epoch": 2.4970728534258457, "grad_norm": 0.2525405287742615, "learning_rate": 3.3650079791092416e-06, "loss": 0.0009, "step": 23033 }, { "epoch": 2.4971812662619253, "grad_norm": 2.0743227005004883, "learning_rate": 3.364282605541854e-06, "loss": 0.0294, "step": 23034 }, { "epoch": 2.4972896790980053, "grad_norm": 0.36142870783805847, "learning_rate": 3.3635572319744668e-06, "loss": 0.0213, "step": 23035 }, { "epoch": 2.497398091934085, "grad_norm": 13.132616996765137, "learning_rate": 3.36283185840708e-06, "loss": 0.0365, "step": 23036 }, { "epoch": 2.497506504770165, "grad_norm": 0.4439569115638733, "learning_rate": 3.3621064848396928e-06, "loss": 0.0122, "step": 23037 }, { "epoch": 2.4976149176062448, "grad_norm": 1.3707764148712158, "learning_rate": 3.3613811112723053e-06, "loss": 0.0026, "step": 23038 }, { "epoch": 2.4977233304423243, "grad_norm": 0.006111684255301952, "learning_rate": 3.3606557377049183e-06, "loss": 0.0, "step": 23039 }, { "epoch": 2.4978317432784043, "grad_norm": 0.1812531054019928, "learning_rate": 3.3599303641375313e-06, "loss": 0.0005, "step": 23040 }, { "epoch": 2.497940156114484, "grad_norm": 0.29522866010665894, "learning_rate": 3.359204990570144e-06, "loss": 0.0061, "step": 23041 }, { "epoch": 2.498048568950564, "grad_norm": 0.29296156764030457, "learning_rate": 3.358479617002757e-06, "loss": 0.0068, "step": 23042 }, { "epoch": 2.4981569817866434, "grad_norm": 0.2851600646972656, "learning_rate": 3.3577542434353694e-06, "loss": 0.0102, "step": 23043 }, { "epoch": 2.4982653946227233, "grad_norm": 0.2910722494125366, "learning_rate": 3.357028869867982e-06, "loss": 0.0066, "step": 23044 }, { "epoch": 2.498373807458803, "grad_norm": 0.7280150651931763, "learning_rate": 3.3563034963005954e-06, "loss": 0.0093, "step": 23045 }, { "epoch": 2.498482220294883, "grad_norm": 20.45765495300293, "learning_rate": 3.355578122733208e-06, "loss": 0.016, "step": 23046 }, { "epoch": 2.498590633130963, "grad_norm": 0.2701508402824402, "learning_rate": 3.3548527491658205e-06, "loss": 0.0055, "step": 23047 }, { "epoch": 2.4986990459670424, "grad_norm": 0.284266859292984, "learning_rate": 3.354127375598433e-06, "loss": 0.0057, "step": 23048 }, { "epoch": 2.4988074588031224, "grad_norm": 0.31640738248825073, "learning_rate": 3.3534020020310465e-06, "loss": 0.0105, "step": 23049 }, { "epoch": 2.498915871639202, "grad_norm": 0.44736576080322266, "learning_rate": 3.352676628463659e-06, "loss": 0.0247, "step": 23050 }, { "epoch": 2.499024284475282, "grad_norm": 0.014303125441074371, "learning_rate": 3.3519512548962717e-06, "loss": 0.0001, "step": 23051 }, { "epoch": 2.499132697311362, "grad_norm": 1.2822597026824951, "learning_rate": 3.3512258813288846e-06, "loss": 0.0421, "step": 23052 }, { "epoch": 2.4992411101474414, "grad_norm": 0.14508144557476044, "learning_rate": 3.3505005077614976e-06, "loss": 0.0005, "step": 23053 }, { "epoch": 2.4993495229835214, "grad_norm": 0.27945825457572937, "learning_rate": 3.34977513419411e-06, "loss": 0.0062, "step": 23054 }, { "epoch": 2.499457935819601, "grad_norm": 3.099372386932373, "learning_rate": 3.349049760626723e-06, "loss": 0.0395, "step": 23055 }, { "epoch": 2.499566348655681, "grad_norm": 0.5129455327987671, "learning_rate": 3.3483243870593358e-06, "loss": 0.0221, "step": 23056 }, { "epoch": 2.4996747614917605, "grad_norm": 0.44769155979156494, "learning_rate": 3.3475990134919483e-06, "loss": 0.0058, "step": 23057 }, { "epoch": 2.4997831743278405, "grad_norm": 0.03278939798474312, "learning_rate": 3.3468736399245617e-06, "loss": 0.0003, "step": 23058 }, { "epoch": 2.49989158716392, "grad_norm": 3.2993221282958984, "learning_rate": 3.3461482663571743e-06, "loss": 0.0202, "step": 23059 }, { "epoch": 2.5, "grad_norm": 0.30196914076805115, "learning_rate": 3.345422892789787e-06, "loss": 0.0051, "step": 23060 }, { "epoch": 2.50010841283608, "grad_norm": 6.736860752105713, "learning_rate": 3.3446975192223994e-06, "loss": 0.0258, "step": 23061 }, { "epoch": 2.5002168256721595, "grad_norm": 1.2360788583755493, "learning_rate": 3.343972145655013e-06, "loss": 0.011, "step": 23062 }, { "epoch": 2.5003252385082395, "grad_norm": 0.087367944419384, "learning_rate": 3.3432467720876254e-06, "loss": 0.0001, "step": 23063 }, { "epoch": 2.500433651344319, "grad_norm": 0.42580917477607727, "learning_rate": 3.342521398520238e-06, "loss": 0.0117, "step": 23064 }, { "epoch": 2.500542064180399, "grad_norm": 0.26440924406051636, "learning_rate": 3.341796024952851e-06, "loss": 0.0194, "step": 23065 }, { "epoch": 2.500650477016479, "grad_norm": 4.119454860687256, "learning_rate": 3.341070651385464e-06, "loss": 0.036, "step": 23066 }, { "epoch": 2.5007588898525586, "grad_norm": 3.745840072631836, "learning_rate": 3.3403452778180765e-06, "loss": 0.0277, "step": 23067 }, { "epoch": 2.500867302688638, "grad_norm": 0.41625458002090454, "learning_rate": 3.3396199042506895e-06, "loss": 0.0064, "step": 23068 }, { "epoch": 2.500975715524718, "grad_norm": 0.38329431414604187, "learning_rate": 3.338894530683302e-06, "loss": 0.0221, "step": 23069 }, { "epoch": 2.501084128360798, "grad_norm": 3.25821590423584, "learning_rate": 3.3381691571159147e-06, "loss": 0.0173, "step": 23070 }, { "epoch": 2.5011925411968776, "grad_norm": 0.04619240015745163, "learning_rate": 3.337443783548528e-06, "loss": 0.0002, "step": 23071 }, { "epoch": 2.5013009540329576, "grad_norm": 0.021550627425312996, "learning_rate": 3.3367184099811406e-06, "loss": 0.0002, "step": 23072 }, { "epoch": 2.501409366869037, "grad_norm": 0.5099130272865295, "learning_rate": 3.335993036413753e-06, "loss": 0.0111, "step": 23073 }, { "epoch": 2.501517779705117, "grad_norm": 0.10070312768220901, "learning_rate": 3.3352676628463658e-06, "loss": 0.0005, "step": 23074 }, { "epoch": 2.501626192541197, "grad_norm": 0.037461452186107635, "learning_rate": 3.334542289278979e-06, "loss": 0.0002, "step": 23075 }, { "epoch": 2.5017346053772767, "grad_norm": 1.0409239530563354, "learning_rate": 3.3338169157115918e-06, "loss": 0.0094, "step": 23076 }, { "epoch": 2.501843018213356, "grad_norm": 7.269076347351074, "learning_rate": 3.3330915421442043e-06, "loss": 0.0228, "step": 23077 }, { "epoch": 2.501951431049436, "grad_norm": 0.4388476312160492, "learning_rate": 3.3323661685768173e-06, "loss": 0.0194, "step": 23078 }, { "epoch": 2.502059843885516, "grad_norm": 0.0201970674097538, "learning_rate": 3.3316407950094303e-06, "loss": 0.0002, "step": 23079 }, { "epoch": 2.5021682567215957, "grad_norm": 0.22409726679325104, "learning_rate": 3.330915421442043e-06, "loss": 0.0047, "step": 23080 }, { "epoch": 2.5022766695576757, "grad_norm": 3.948608636856079, "learning_rate": 3.330190047874656e-06, "loss": 0.0494, "step": 23081 }, { "epoch": 2.5023850823937552, "grad_norm": 0.011074128560721874, "learning_rate": 3.3294646743072684e-06, "loss": 0.0001, "step": 23082 }, { "epoch": 2.502493495229835, "grad_norm": 0.023641327396035194, "learning_rate": 3.328739300739881e-06, "loss": 0.0002, "step": 23083 }, { "epoch": 2.502601908065915, "grad_norm": 0.6932910084724426, "learning_rate": 3.3280139271724944e-06, "loss": 0.027, "step": 23084 }, { "epoch": 2.5027103209019947, "grad_norm": 0.1399305909872055, "learning_rate": 3.327288553605107e-06, "loss": 0.0005, "step": 23085 }, { "epoch": 2.5028187337380747, "grad_norm": 5.585461616516113, "learning_rate": 3.3265631800377195e-06, "loss": 0.0335, "step": 23086 }, { "epoch": 2.5029271465741543, "grad_norm": 0.5421668887138367, "learning_rate": 3.325837806470332e-06, "loss": 0.0007, "step": 23087 }, { "epoch": 2.5030355594102343, "grad_norm": 0.634850025177002, "learning_rate": 3.3251124329029455e-06, "loss": 0.0087, "step": 23088 }, { "epoch": 2.503143972246314, "grad_norm": 0.01407573837786913, "learning_rate": 3.324387059335558e-06, "loss": 0.0002, "step": 23089 }, { "epoch": 2.503252385082394, "grad_norm": 7.655965328216553, "learning_rate": 3.3236616857681707e-06, "loss": 0.0427, "step": 23090 }, { "epoch": 2.5033607979184733, "grad_norm": 0.027864176779985428, "learning_rate": 3.3229363122007836e-06, "loss": 0.0002, "step": 23091 }, { "epoch": 2.5034692107545533, "grad_norm": 0.2890171408653259, "learning_rate": 3.3222109386333966e-06, "loss": 0.006, "step": 23092 }, { "epoch": 2.5035776235906333, "grad_norm": 0.04415145516395569, "learning_rate": 3.321485565066009e-06, "loss": 0.0004, "step": 23093 }, { "epoch": 2.503686036426713, "grad_norm": 0.030897481366991997, "learning_rate": 3.320760191498622e-06, "loss": 0.0002, "step": 23094 }, { "epoch": 2.503794449262793, "grad_norm": 1.74907648563385, "learning_rate": 3.3200348179312348e-06, "loss": 0.0131, "step": 23095 }, { "epoch": 2.5039028620988724, "grad_norm": 0.5138731002807617, "learning_rate": 3.3193094443638473e-06, "loss": 0.024, "step": 23096 }, { "epoch": 2.5040112749349523, "grad_norm": 0.05355880782008171, "learning_rate": 3.3185840707964607e-06, "loss": 0.0004, "step": 23097 }, { "epoch": 2.5041196877710323, "grad_norm": 0.35730427503585815, "learning_rate": 3.3178586972290733e-06, "loss": 0.0212, "step": 23098 }, { "epoch": 2.504228100607112, "grad_norm": 1.5774097442626953, "learning_rate": 3.317133323661686e-06, "loss": 0.007, "step": 23099 }, { "epoch": 2.5043365134431914, "grad_norm": 0.243635892868042, "learning_rate": 3.3164079500942984e-06, "loss": 0.0004, "step": 23100 }, { "epoch": 2.5044449262792714, "grad_norm": 1.8893475532531738, "learning_rate": 3.315682576526912e-06, "loss": 0.0286, "step": 23101 }, { "epoch": 2.5045533391153514, "grad_norm": 0.2610318958759308, "learning_rate": 3.3149572029595244e-06, "loss": 0.0009, "step": 23102 }, { "epoch": 2.504661751951431, "grad_norm": 0.3657059669494629, "learning_rate": 3.314231829392137e-06, "loss": 0.0093, "step": 23103 }, { "epoch": 2.504770164787511, "grad_norm": 1.9843136072158813, "learning_rate": 3.31350645582475e-06, "loss": 0.0399, "step": 23104 }, { "epoch": 2.5048785776235905, "grad_norm": 0.3512237071990967, "learning_rate": 3.312781082257363e-06, "loss": 0.006, "step": 23105 }, { "epoch": 2.5049869904596704, "grad_norm": 1.770538091659546, "learning_rate": 3.3120557086899755e-06, "loss": 0.0101, "step": 23106 }, { "epoch": 2.5050954032957504, "grad_norm": 0.0373910516500473, "learning_rate": 3.3113303351225885e-06, "loss": 0.0003, "step": 23107 }, { "epoch": 2.50520381613183, "grad_norm": 0.34144386649131775, "learning_rate": 3.310604961555201e-06, "loss": 0.0069, "step": 23108 }, { "epoch": 2.50531222896791, "grad_norm": 2.467845916748047, "learning_rate": 3.3098795879878137e-06, "loss": 0.0402, "step": 23109 }, { "epoch": 2.5054206418039895, "grad_norm": 9.700549125671387, "learning_rate": 3.309154214420427e-06, "loss": 0.0074, "step": 23110 }, { "epoch": 2.5055290546400695, "grad_norm": 0.8498884439468384, "learning_rate": 3.3084288408530396e-06, "loss": 0.0047, "step": 23111 }, { "epoch": 2.505637467476149, "grad_norm": 0.023346496745944023, "learning_rate": 3.307703467285652e-06, "loss": 0.0002, "step": 23112 }, { "epoch": 2.505745880312229, "grad_norm": 0.3662380874156952, "learning_rate": 3.3069780937182648e-06, "loss": 0.001, "step": 23113 }, { "epoch": 2.5058542931483085, "grad_norm": 0.6024681329727173, "learning_rate": 3.306252720150878e-06, "loss": 0.0134, "step": 23114 }, { "epoch": 2.5059627059843885, "grad_norm": 0.05759004130959511, "learning_rate": 3.3055273465834908e-06, "loss": 0.0003, "step": 23115 }, { "epoch": 2.5060711188204685, "grad_norm": 0.07959553599357605, "learning_rate": 3.3048019730161033e-06, "loss": 0.0005, "step": 23116 }, { "epoch": 2.506179531656548, "grad_norm": 4.8510565757751465, "learning_rate": 3.3040765994487163e-06, "loss": 0.0205, "step": 23117 }, { "epoch": 2.506287944492628, "grad_norm": 0.030828800052404404, "learning_rate": 3.3033512258813293e-06, "loss": 0.0003, "step": 23118 }, { "epoch": 2.5063963573287076, "grad_norm": 0.2894619405269623, "learning_rate": 3.302625852313942e-06, "loss": 0.0107, "step": 23119 }, { "epoch": 2.5065047701647876, "grad_norm": 0.3778625428676605, "learning_rate": 3.301900478746555e-06, "loss": 0.011, "step": 23120 }, { "epoch": 2.5066131830008676, "grad_norm": 0.011071687564253807, "learning_rate": 3.3011751051791674e-06, "loss": 0.0001, "step": 23121 }, { "epoch": 2.506721595836947, "grad_norm": 0.018058614805340767, "learning_rate": 3.30044973161178e-06, "loss": 0.0002, "step": 23122 }, { "epoch": 2.5068300086730266, "grad_norm": 0.45299527049064636, "learning_rate": 3.2997243580443934e-06, "loss": 0.0068, "step": 23123 }, { "epoch": 2.5069384215091066, "grad_norm": 0.05220426991581917, "learning_rate": 3.298998984477006e-06, "loss": 0.0005, "step": 23124 }, { "epoch": 2.5070468343451866, "grad_norm": 12.016409873962402, "learning_rate": 3.2982736109096185e-06, "loss": 0.0284, "step": 23125 }, { "epoch": 2.507155247181266, "grad_norm": 0.05058474466204643, "learning_rate": 3.297548237342231e-06, "loss": 0.0002, "step": 23126 }, { "epoch": 2.507263660017346, "grad_norm": 0.25644993782043457, "learning_rate": 3.2968228637748445e-06, "loss": 0.0057, "step": 23127 }, { "epoch": 2.5073720728534257, "grad_norm": 9.025336265563965, "learning_rate": 3.296097490207457e-06, "loss": 0.0365, "step": 23128 }, { "epoch": 2.5074804856895057, "grad_norm": 0.2931850552558899, "learning_rate": 3.2953721166400697e-06, "loss": 0.0062, "step": 23129 }, { "epoch": 2.5075888985255856, "grad_norm": 0.061902403831481934, "learning_rate": 3.2946467430726826e-06, "loss": 0.0005, "step": 23130 }, { "epoch": 2.507697311361665, "grad_norm": 0.030154846608638763, "learning_rate": 3.2939213695052956e-06, "loss": 0.0002, "step": 23131 }, { "epoch": 2.507805724197745, "grad_norm": 3.5744190216064453, "learning_rate": 3.293195995937908e-06, "loss": 0.0265, "step": 23132 }, { "epoch": 2.5079141370338247, "grad_norm": 8.337486267089844, "learning_rate": 3.292470622370521e-06, "loss": 0.0225, "step": 23133 }, { "epoch": 2.5080225498699047, "grad_norm": 0.753197193145752, "learning_rate": 3.2917452488031338e-06, "loss": 0.0029, "step": 23134 }, { "epoch": 2.5081309627059842, "grad_norm": 0.037957143038511276, "learning_rate": 3.2910198752357468e-06, "loss": 0.0004, "step": 23135 }, { "epoch": 2.508239375542064, "grad_norm": 3.2294812202453613, "learning_rate": 3.2902945016683597e-06, "loss": 0.0033, "step": 23136 }, { "epoch": 2.5083477883781438, "grad_norm": 0.2987314462661743, "learning_rate": 3.2895691281009723e-06, "loss": 0.0058, "step": 23137 }, { "epoch": 2.5084562012142237, "grad_norm": 0.47735896706581116, "learning_rate": 3.288843754533585e-06, "loss": 0.0092, "step": 23138 }, { "epoch": 2.5085646140503037, "grad_norm": 2.5114505290985107, "learning_rate": 3.2881183809661974e-06, "loss": 0.0054, "step": 23139 }, { "epoch": 2.5086730268863833, "grad_norm": 0.3105463683605194, "learning_rate": 3.287393007398811e-06, "loss": 0.0096, "step": 23140 }, { "epoch": 2.5087814397224633, "grad_norm": 0.036410484462976456, "learning_rate": 3.2866676338314234e-06, "loss": 0.0002, "step": 23141 }, { "epoch": 2.508889852558543, "grad_norm": 0.09709229320287704, "learning_rate": 3.285942260264036e-06, "loss": 0.0003, "step": 23142 }, { "epoch": 2.508998265394623, "grad_norm": 0.2585115134716034, "learning_rate": 3.285216886696649e-06, "loss": 0.0055, "step": 23143 }, { "epoch": 2.5091066782307028, "grad_norm": 0.34334099292755127, "learning_rate": 3.284491513129262e-06, "loss": 0.0078, "step": 23144 }, { "epoch": 2.5092150910667823, "grad_norm": 0.4048043489456177, "learning_rate": 3.2837661395618745e-06, "loss": 0.0009, "step": 23145 }, { "epoch": 2.509323503902862, "grad_norm": 4.524026393890381, "learning_rate": 3.2830407659944875e-06, "loss": 0.0205, "step": 23146 }, { "epoch": 2.509431916738942, "grad_norm": 0.8040323853492737, "learning_rate": 3.2823153924271e-06, "loss": 0.0077, "step": 23147 }, { "epoch": 2.509540329575022, "grad_norm": 0.43256688117980957, "learning_rate": 3.281590018859713e-06, "loss": 0.0092, "step": 23148 }, { "epoch": 2.5096487424111014, "grad_norm": 3.0147907733917236, "learning_rate": 3.280864645292326e-06, "loss": 0.0474, "step": 23149 }, { "epoch": 2.5097571552471813, "grad_norm": 0.1360773742198944, "learning_rate": 3.2801392717249386e-06, "loss": 0.0004, "step": 23150 }, { "epoch": 2.509865568083261, "grad_norm": 0.0704663097858429, "learning_rate": 3.2794138981575512e-06, "loss": 0.0003, "step": 23151 }, { "epoch": 2.509973980919341, "grad_norm": 0.3393322229385376, "learning_rate": 3.2786885245901638e-06, "loss": 0.0056, "step": 23152 }, { "epoch": 2.510082393755421, "grad_norm": 0.03863222897052765, "learning_rate": 3.277963151022777e-06, "loss": 0.0002, "step": 23153 }, { "epoch": 2.5101908065915004, "grad_norm": 6.321266174316406, "learning_rate": 3.2772377774553898e-06, "loss": 0.0193, "step": 23154 }, { "epoch": 2.5102992194275804, "grad_norm": 0.36422550678253174, "learning_rate": 3.2765124038880023e-06, "loss": 0.012, "step": 23155 }, { "epoch": 2.51040763226366, "grad_norm": 0.5364894866943359, "learning_rate": 3.2757870303206153e-06, "loss": 0.0197, "step": 23156 }, { "epoch": 2.51051604509974, "grad_norm": 0.005215618759393692, "learning_rate": 3.2750616567532283e-06, "loss": 0.0, "step": 23157 }, { "epoch": 2.5106244579358195, "grad_norm": 0.00499748857691884, "learning_rate": 3.274336283185841e-06, "loss": 0.0, "step": 23158 }, { "epoch": 2.5107328707718994, "grad_norm": 0.08728621155023575, "learning_rate": 3.273610909618454e-06, "loss": 0.0005, "step": 23159 }, { "epoch": 2.510841283607979, "grad_norm": 0.03858339414000511, "learning_rate": 3.2728855360510664e-06, "loss": 0.0003, "step": 23160 }, { "epoch": 2.510949696444059, "grad_norm": 23.229581832885742, "learning_rate": 3.2721601624836794e-06, "loss": 0.0204, "step": 23161 }, { "epoch": 2.511058109280139, "grad_norm": 4.522210121154785, "learning_rate": 3.2714347889162924e-06, "loss": 0.0121, "step": 23162 }, { "epoch": 2.5111665221162185, "grad_norm": 1.5326091051101685, "learning_rate": 3.270709415348905e-06, "loss": 0.0263, "step": 23163 }, { "epoch": 2.5112749349522985, "grad_norm": 0.3104635775089264, "learning_rate": 3.2699840417815175e-06, "loss": 0.005, "step": 23164 }, { "epoch": 2.511383347788378, "grad_norm": 0.013100679963827133, "learning_rate": 3.26925866821413e-06, "loss": 0.0001, "step": 23165 }, { "epoch": 2.511491760624458, "grad_norm": 0.2999315857887268, "learning_rate": 3.2685332946467435e-06, "loss": 0.0107, "step": 23166 }, { "epoch": 2.511600173460538, "grad_norm": 10.091351509094238, "learning_rate": 3.267807921079356e-06, "loss": 0.0259, "step": 23167 }, { "epoch": 2.5117085862966175, "grad_norm": 0.8143491148948669, "learning_rate": 3.2670825475119687e-06, "loss": 0.0016, "step": 23168 }, { "epoch": 2.511816999132697, "grad_norm": 0.014848975464701653, "learning_rate": 3.2663571739445817e-06, "loss": 0.0001, "step": 23169 }, { "epoch": 2.511925411968777, "grad_norm": 0.017223436385393143, "learning_rate": 3.2656318003771946e-06, "loss": 0.0002, "step": 23170 }, { "epoch": 2.512033824804857, "grad_norm": 0.021383531391620636, "learning_rate": 3.264906426809807e-06, "loss": 0.0002, "step": 23171 }, { "epoch": 2.5121422376409366, "grad_norm": 3.9418399333953857, "learning_rate": 3.26418105324242e-06, "loss": 0.0168, "step": 23172 }, { "epoch": 2.5122506504770166, "grad_norm": 0.05785088241100311, "learning_rate": 3.2634556796750328e-06, "loss": 0.0005, "step": 23173 }, { "epoch": 2.512359063313096, "grad_norm": 0.27932727336883545, "learning_rate": 3.2627303061076458e-06, "loss": 0.0061, "step": 23174 }, { "epoch": 2.512467476149176, "grad_norm": 0.37688472867012024, "learning_rate": 3.2620049325402587e-06, "loss": 0.0189, "step": 23175 }, { "epoch": 2.512575888985256, "grad_norm": 5.50198221206665, "learning_rate": 3.2612795589728713e-06, "loss": 0.0327, "step": 23176 }, { "epoch": 2.5126843018213356, "grad_norm": 0.09869750589132309, "learning_rate": 3.260554185405484e-06, "loss": 0.0115, "step": 23177 }, { "epoch": 2.512792714657415, "grad_norm": 1.178990125656128, "learning_rate": 3.2598288118380964e-06, "loss": 0.0069, "step": 23178 }, { "epoch": 2.512901127493495, "grad_norm": 0.3877546489238739, "learning_rate": 3.25910343827071e-06, "loss": 0.0101, "step": 23179 }, { "epoch": 2.513009540329575, "grad_norm": 0.901317298412323, "learning_rate": 3.2583780647033224e-06, "loss": 0.004, "step": 23180 }, { "epoch": 2.5131179531656547, "grad_norm": 1.1270816326141357, "learning_rate": 3.257652691135935e-06, "loss": 0.0073, "step": 23181 }, { "epoch": 2.5132263660017347, "grad_norm": 5.368185043334961, "learning_rate": 3.256927317568548e-06, "loss": 0.04, "step": 23182 }, { "epoch": 2.513334778837814, "grad_norm": 2.693768262863159, "learning_rate": 3.256201944001161e-06, "loss": 0.0444, "step": 23183 }, { "epoch": 2.513443191673894, "grad_norm": 1.1461503505706787, "learning_rate": 3.2554765704337735e-06, "loss": 0.0007, "step": 23184 }, { "epoch": 2.513551604509974, "grad_norm": 7.0841593742370605, "learning_rate": 3.2547511968663865e-06, "loss": 0.0231, "step": 23185 }, { "epoch": 2.5136600173460537, "grad_norm": 2.564789056777954, "learning_rate": 3.254025823298999e-06, "loss": 0.0103, "step": 23186 }, { "epoch": 2.5137684301821337, "grad_norm": 7.566295623779297, "learning_rate": 3.253300449731612e-06, "loss": 0.0186, "step": 23187 }, { "epoch": 2.5138768430182132, "grad_norm": 0.2555995583534241, "learning_rate": 3.252575076164225e-06, "loss": 0.0004, "step": 23188 }, { "epoch": 2.513985255854293, "grad_norm": 19.905452728271484, "learning_rate": 3.2518497025968376e-06, "loss": 0.0097, "step": 23189 }, { "epoch": 2.514093668690373, "grad_norm": 1.6034208536148071, "learning_rate": 3.2511243290294502e-06, "loss": 0.016, "step": 23190 }, { "epoch": 2.5142020815264527, "grad_norm": 0.25079581141471863, "learning_rate": 3.2503989554620628e-06, "loss": 0.005, "step": 23191 }, { "epoch": 2.5143104943625323, "grad_norm": 8.74908447265625, "learning_rate": 3.249673581894676e-06, "loss": 0.0043, "step": 23192 }, { "epoch": 2.5144189071986123, "grad_norm": 2.343656539916992, "learning_rate": 3.2489482083272888e-06, "loss": 0.0372, "step": 23193 }, { "epoch": 2.5145273200346923, "grad_norm": 2.1824023723602295, "learning_rate": 3.2482228347599013e-06, "loss": 0.0088, "step": 23194 }, { "epoch": 2.514635732870772, "grad_norm": 0.41336944699287415, "learning_rate": 3.2474974611925143e-06, "loss": 0.0113, "step": 23195 }, { "epoch": 2.514744145706852, "grad_norm": 0.10808432847261429, "learning_rate": 3.2467720876251273e-06, "loss": 0.0005, "step": 23196 }, { "epoch": 2.5148525585429313, "grad_norm": 2.5604279041290283, "learning_rate": 3.24604671405774e-06, "loss": 0.0112, "step": 23197 }, { "epoch": 2.5149609713790113, "grad_norm": 4.348392963409424, "learning_rate": 3.245321340490353e-06, "loss": 0.0156, "step": 23198 }, { "epoch": 2.5150693842150913, "grad_norm": 0.030715828761458397, "learning_rate": 3.2445959669229654e-06, "loss": 0.0002, "step": 23199 }, { "epoch": 2.515177797051171, "grad_norm": 0.6078766584396362, "learning_rate": 3.2438705933555784e-06, "loss": 0.0114, "step": 23200 }, { "epoch": 2.5152862098872504, "grad_norm": 4.15394926071167, "learning_rate": 3.2431452197881914e-06, "loss": 0.0201, "step": 23201 }, { "epoch": 2.5153946227233304, "grad_norm": 0.38632896542549133, "learning_rate": 3.242419846220804e-06, "loss": 0.0111, "step": 23202 }, { "epoch": 2.5155030355594103, "grad_norm": 0.42166683077812195, "learning_rate": 3.2416944726534165e-06, "loss": 0.0112, "step": 23203 }, { "epoch": 2.51561144839549, "grad_norm": 0.12003850191831589, "learning_rate": 3.240969099086029e-06, "loss": 0.0006, "step": 23204 }, { "epoch": 2.51571986123157, "grad_norm": 0.12936682999134064, "learning_rate": 3.2402437255186425e-06, "loss": 0.0005, "step": 23205 }, { "epoch": 2.5158282740676494, "grad_norm": 2.2002556324005127, "learning_rate": 3.239518351951255e-06, "loss": 0.0063, "step": 23206 }, { "epoch": 2.5159366869037294, "grad_norm": 0.5226560235023499, "learning_rate": 3.2387929783838677e-06, "loss": 0.0147, "step": 23207 }, { "epoch": 2.5160450997398094, "grad_norm": 0.9161612391471863, "learning_rate": 3.2380676048164807e-06, "loss": 0.0018, "step": 23208 }, { "epoch": 2.516153512575889, "grad_norm": 0.6763577461242676, "learning_rate": 3.2373422312490936e-06, "loss": 0.0275, "step": 23209 }, { "epoch": 2.516261925411969, "grad_norm": 0.415175199508667, "learning_rate": 3.2366168576817062e-06, "loss": 0.0009, "step": 23210 }, { "epoch": 2.5163703382480485, "grad_norm": 0.2680475115776062, "learning_rate": 3.235891484114319e-06, "loss": 0.0062, "step": 23211 }, { "epoch": 2.5164787510841284, "grad_norm": 0.4590831995010376, "learning_rate": 3.2351661105469318e-06, "loss": 0.0103, "step": 23212 }, { "epoch": 2.5165871639202084, "grad_norm": 0.16427937150001526, "learning_rate": 3.2344407369795448e-06, "loss": 0.0005, "step": 23213 }, { "epoch": 2.516695576756288, "grad_norm": 0.48426857590675354, "learning_rate": 3.2337153634121578e-06, "loss": 0.0136, "step": 23214 }, { "epoch": 2.5168039895923675, "grad_norm": 0.509768545627594, "learning_rate": 3.2329899898447703e-06, "loss": 0.011, "step": 23215 }, { "epoch": 2.5169124024284475, "grad_norm": 0.582148551940918, "learning_rate": 3.232264616277383e-06, "loss": 0.0122, "step": 23216 }, { "epoch": 2.5170208152645275, "grad_norm": 3.7511684894561768, "learning_rate": 3.2315392427099955e-06, "loss": 0.0179, "step": 23217 }, { "epoch": 2.517129228100607, "grad_norm": 0.080184206366539, "learning_rate": 3.230813869142609e-06, "loss": 0.0005, "step": 23218 }, { "epoch": 2.517237640936687, "grad_norm": 2.664860248565674, "learning_rate": 3.2300884955752214e-06, "loss": 0.0246, "step": 23219 }, { "epoch": 2.5173460537727665, "grad_norm": 0.6290633082389832, "learning_rate": 3.229363122007834e-06, "loss": 0.0327, "step": 23220 }, { "epoch": 2.5174544666088465, "grad_norm": 0.2420826256275177, "learning_rate": 3.228637748440447e-06, "loss": 0.005, "step": 23221 }, { "epoch": 2.5175628794449265, "grad_norm": 0.40445321798324585, "learning_rate": 3.22791237487306e-06, "loss": 0.0078, "step": 23222 }, { "epoch": 2.517671292281006, "grad_norm": 0.29693397879600525, "learning_rate": 3.2271870013056725e-06, "loss": 0.0057, "step": 23223 }, { "epoch": 2.5177797051170856, "grad_norm": 0.02092788740992546, "learning_rate": 3.2264616277382855e-06, "loss": 0.0002, "step": 23224 }, { "epoch": 2.5178881179531656, "grad_norm": 0.08604058623313904, "learning_rate": 3.225736254170898e-06, "loss": 0.0004, "step": 23225 }, { "epoch": 2.5179965307892456, "grad_norm": 0.35122913122177124, "learning_rate": 3.2250108806035115e-06, "loss": 0.0105, "step": 23226 }, { "epoch": 2.518104943625325, "grad_norm": 0.27514320611953735, "learning_rate": 3.224285507036124e-06, "loss": 0.0052, "step": 23227 }, { "epoch": 2.518213356461405, "grad_norm": 0.6525145173072815, "learning_rate": 3.2235601334687367e-06, "loss": 0.001, "step": 23228 }, { "epoch": 2.5183217692974846, "grad_norm": 3.928239583969116, "learning_rate": 3.2228347599013492e-06, "loss": 0.0089, "step": 23229 }, { "epoch": 2.5184301821335646, "grad_norm": 0.5612985491752625, "learning_rate": 3.2221093863339626e-06, "loss": 0.008, "step": 23230 }, { "epoch": 2.5185385949696446, "grad_norm": 21.60059928894043, "learning_rate": 3.221384012766575e-06, "loss": 0.0134, "step": 23231 }, { "epoch": 2.518647007805724, "grad_norm": 0.7659663558006287, "learning_rate": 3.2206586391991878e-06, "loss": 0.0134, "step": 23232 }, { "epoch": 2.518755420641804, "grad_norm": 1.417610764503479, "learning_rate": 3.2199332656318003e-06, "loss": 0.0029, "step": 23233 }, { "epoch": 2.5188638334778837, "grad_norm": 0.008743386715650558, "learning_rate": 3.2192078920644133e-06, "loss": 0.0001, "step": 23234 }, { "epoch": 2.5189722463139637, "grad_norm": 0.29477718472480774, "learning_rate": 3.2184825184970263e-06, "loss": 0.0108, "step": 23235 }, { "epoch": 2.519080659150043, "grad_norm": 0.33536508679389954, "learning_rate": 3.217757144929639e-06, "loss": 0.0074, "step": 23236 }, { "epoch": 2.519189071986123, "grad_norm": 0.3112638592720032, "learning_rate": 3.217031771362252e-06, "loss": 0.0067, "step": 23237 }, { "epoch": 2.5192974848222027, "grad_norm": 0.3017804026603699, "learning_rate": 3.2163063977948644e-06, "loss": 0.0109, "step": 23238 }, { "epoch": 2.5194058976582827, "grad_norm": 0.11275855451822281, "learning_rate": 3.215581024227478e-06, "loss": 0.0003, "step": 23239 }, { "epoch": 2.5195143104943627, "grad_norm": 0.3685021996498108, "learning_rate": 3.2148556506600904e-06, "loss": 0.0209, "step": 23240 }, { "epoch": 2.5196227233304422, "grad_norm": 0.8777637481689453, "learning_rate": 3.214130277092703e-06, "loss": 0.0121, "step": 23241 }, { "epoch": 2.5197311361665222, "grad_norm": 0.17726421356201172, "learning_rate": 3.2134049035253156e-06, "loss": 0.0003, "step": 23242 }, { "epoch": 2.5198395490026018, "grad_norm": 0.4525105953216553, "learning_rate": 3.212679529957929e-06, "loss": 0.016, "step": 23243 }, { "epoch": 2.5199479618386817, "grad_norm": 0.26688435673713684, "learning_rate": 3.2119541563905415e-06, "loss": 0.0058, "step": 23244 }, { "epoch": 2.5200563746747617, "grad_norm": 0.5522749423980713, "learning_rate": 3.211228782823154e-06, "loss": 0.0109, "step": 23245 }, { "epoch": 2.5201647875108413, "grad_norm": 9.669875144958496, "learning_rate": 3.2105034092557667e-06, "loss": 0.0046, "step": 23246 }, { "epoch": 2.520273200346921, "grad_norm": 7.983185291290283, "learning_rate": 3.2097780356883797e-06, "loss": 0.0216, "step": 23247 }, { "epoch": 2.520381613183001, "grad_norm": 0.5559180378913879, "learning_rate": 3.2090526621209926e-06, "loss": 0.0228, "step": 23248 }, { "epoch": 2.520490026019081, "grad_norm": 0.2824536859989166, "learning_rate": 3.2083272885536052e-06, "loss": 0.001, "step": 23249 }, { "epoch": 2.5205984388551603, "grad_norm": 0.5154474973678589, "learning_rate": 3.207601914986218e-06, "loss": 0.0074, "step": 23250 }, { "epoch": 2.5207068516912403, "grad_norm": 0.29844099283218384, "learning_rate": 3.2068765414188308e-06, "loss": 0.0098, "step": 23251 }, { "epoch": 2.52081526452732, "grad_norm": 0.38712945580482483, "learning_rate": 3.206151167851444e-06, "loss": 0.0072, "step": 23252 }, { "epoch": 2.5209236773634, "grad_norm": 0.35414907336235046, "learning_rate": 3.2054257942840568e-06, "loss": 0.0149, "step": 23253 }, { "epoch": 2.52103209019948, "grad_norm": 0.3202141225337982, "learning_rate": 3.2047004207166693e-06, "loss": 0.0108, "step": 23254 }, { "epoch": 2.5211405030355594, "grad_norm": 3.0616161823272705, "learning_rate": 3.203975047149282e-06, "loss": 0.0297, "step": 23255 }, { "epoch": 2.5212489158716394, "grad_norm": 0.1502254158258438, "learning_rate": 3.2032496735818953e-06, "loss": 0.0007, "step": 23256 }, { "epoch": 2.521357328707719, "grad_norm": 0.23431053757667542, "learning_rate": 3.202524300014508e-06, "loss": 0.0005, "step": 23257 }, { "epoch": 2.521465741543799, "grad_norm": 0.4394215941429138, "learning_rate": 3.2017989264471204e-06, "loss": 0.0187, "step": 23258 }, { "epoch": 2.5215741543798784, "grad_norm": 0.29262661933898926, "learning_rate": 3.201073552879733e-06, "loss": 0.0108, "step": 23259 }, { "epoch": 2.5216825672159584, "grad_norm": 12.061201095581055, "learning_rate": 3.200348179312346e-06, "loss": 0.021, "step": 23260 }, { "epoch": 2.521790980052038, "grad_norm": 0.2028231918811798, "learning_rate": 3.199622805744959e-06, "loss": 0.0004, "step": 23261 }, { "epoch": 2.521899392888118, "grad_norm": 3.0726113319396973, "learning_rate": 3.1988974321775715e-06, "loss": 0.0095, "step": 23262 }, { "epoch": 2.522007805724198, "grad_norm": 1.196112036705017, "learning_rate": 3.1981720586101845e-06, "loss": 0.0011, "step": 23263 }, { "epoch": 2.5221162185602775, "grad_norm": 0.21347692608833313, "learning_rate": 3.197446685042797e-06, "loss": 0.0004, "step": 23264 }, { "epoch": 2.5222246313963574, "grad_norm": 0.5512649416923523, "learning_rate": 3.1967213114754105e-06, "loss": 0.0159, "step": 23265 }, { "epoch": 2.522333044232437, "grad_norm": 1.5264045000076294, "learning_rate": 3.195995937908023e-06, "loss": 0.0278, "step": 23266 }, { "epoch": 2.522441457068517, "grad_norm": 0.5949415564537048, "learning_rate": 3.1952705643406357e-06, "loss": 0.0173, "step": 23267 }, { "epoch": 2.522549869904597, "grad_norm": 16.08868408203125, "learning_rate": 3.1945451907732482e-06, "loss": 0.0036, "step": 23268 }, { "epoch": 2.5226582827406765, "grad_norm": 0.2717268764972687, "learning_rate": 3.1938198172058616e-06, "loss": 0.0007, "step": 23269 }, { "epoch": 2.522766695576756, "grad_norm": 0.31041914224624634, "learning_rate": 3.193094443638474e-06, "loss": 0.0172, "step": 23270 }, { "epoch": 2.522875108412836, "grad_norm": 0.26875799894332886, "learning_rate": 3.1923690700710868e-06, "loss": 0.0059, "step": 23271 }, { "epoch": 2.522983521248916, "grad_norm": 0.2152652144432068, "learning_rate": 3.1916436965036993e-06, "loss": 0.0037, "step": 23272 }, { "epoch": 2.5230919340849955, "grad_norm": 0.2872317433357239, "learning_rate": 3.1909183229363123e-06, "loss": 0.0061, "step": 23273 }, { "epoch": 2.5232003469210755, "grad_norm": 0.29280832409858704, "learning_rate": 3.1901929493689253e-06, "loss": 0.0068, "step": 23274 }, { "epoch": 2.523308759757155, "grad_norm": 0.32892951369285583, "learning_rate": 3.1894675758015383e-06, "loss": 0.0007, "step": 23275 }, { "epoch": 2.523417172593235, "grad_norm": 0.06120367348194122, "learning_rate": 3.188742202234151e-06, "loss": 0.01, "step": 23276 }, { "epoch": 2.523525585429315, "grad_norm": 0.43951183557510376, "learning_rate": 3.1880168286667634e-06, "loss": 0.0104, "step": 23277 }, { "epoch": 2.5236339982653946, "grad_norm": 0.1380816400051117, "learning_rate": 3.187291455099377e-06, "loss": 0.0017, "step": 23278 }, { "epoch": 2.5237424111014746, "grad_norm": 0.012177466414868832, "learning_rate": 3.1865660815319894e-06, "loss": 0.0001, "step": 23279 }, { "epoch": 2.523850823937554, "grad_norm": 3.726759433746338, "learning_rate": 3.185840707964602e-06, "loss": 0.028, "step": 23280 }, { "epoch": 2.523959236773634, "grad_norm": 0.02273515984416008, "learning_rate": 3.1851153343972146e-06, "loss": 0.0002, "step": 23281 }, { "epoch": 2.5240676496097136, "grad_norm": 2.3414714336395264, "learning_rate": 3.184389960829828e-06, "loss": 0.0053, "step": 23282 }, { "epoch": 2.5241760624457936, "grad_norm": 0.5703186988830566, "learning_rate": 3.1836645872624405e-06, "loss": 0.0118, "step": 23283 }, { "epoch": 2.524284475281873, "grad_norm": 6.289626121520996, "learning_rate": 3.182939213695053e-06, "loss": 0.1032, "step": 23284 }, { "epoch": 2.524392888117953, "grad_norm": 0.15516752004623413, "learning_rate": 3.1822138401276657e-06, "loss": 0.0005, "step": 23285 }, { "epoch": 2.524501300954033, "grad_norm": 2.2721073627471924, "learning_rate": 3.1814884665602787e-06, "loss": 0.0229, "step": 23286 }, { "epoch": 2.5246097137901127, "grad_norm": 0.27353307604789734, "learning_rate": 3.1807630929928917e-06, "loss": 0.0056, "step": 23287 }, { "epoch": 2.5247181266261927, "grad_norm": 0.03552240878343582, "learning_rate": 3.1800377194255046e-06, "loss": 0.0002, "step": 23288 }, { "epoch": 2.524826539462272, "grad_norm": 0.03758182004094124, "learning_rate": 3.179312345858117e-06, "loss": 0.0002, "step": 23289 }, { "epoch": 2.524934952298352, "grad_norm": 0.7740437388420105, "learning_rate": 3.1785869722907298e-06, "loss": 0.018, "step": 23290 }, { "epoch": 2.525043365134432, "grad_norm": 0.29487839341163635, "learning_rate": 3.177861598723343e-06, "loss": 0.0046, "step": 23291 }, { "epoch": 2.5251517779705117, "grad_norm": 1.5791999101638794, "learning_rate": 3.1771362251559558e-06, "loss": 0.0078, "step": 23292 }, { "epoch": 2.5252601908065913, "grad_norm": 0.7534072399139404, "learning_rate": 3.1764108515885683e-06, "loss": 0.0281, "step": 23293 }, { "epoch": 2.5253686036426712, "grad_norm": 0.3460448086261749, "learning_rate": 3.175685478021181e-06, "loss": 0.0111, "step": 23294 }, { "epoch": 2.5254770164787512, "grad_norm": 1.9133176803588867, "learning_rate": 3.1749601044537943e-06, "loss": 0.0116, "step": 23295 }, { "epoch": 2.5255854293148308, "grad_norm": 0.36202147603034973, "learning_rate": 3.174234730886407e-06, "loss": 0.0009, "step": 23296 }, { "epoch": 2.5256938421509108, "grad_norm": 3.6131017208099365, "learning_rate": 3.1735093573190194e-06, "loss": 0.0094, "step": 23297 }, { "epoch": 2.5258022549869903, "grad_norm": 0.36462074518203735, "learning_rate": 3.172783983751632e-06, "loss": 0.0142, "step": 23298 }, { "epoch": 2.5259106678230703, "grad_norm": 0.03502733260393143, "learning_rate": 3.172058610184245e-06, "loss": 0.0003, "step": 23299 }, { "epoch": 2.5260190806591503, "grad_norm": 4.646998405456543, "learning_rate": 3.171333236616858e-06, "loss": 0.0205, "step": 23300 }, { "epoch": 2.52612749349523, "grad_norm": 21.52530860900879, "learning_rate": 3.170607863049471e-06, "loss": 0.0431, "step": 23301 }, { "epoch": 2.52623590633131, "grad_norm": 0.2536146342754364, "learning_rate": 3.1698824894820835e-06, "loss": 0.0044, "step": 23302 }, { "epoch": 2.5263443191673893, "grad_norm": 3.4857850074768066, "learning_rate": 3.169157115914696e-06, "loss": 0.019, "step": 23303 }, { "epoch": 2.5264527320034693, "grad_norm": 0.31114545464515686, "learning_rate": 3.1684317423473095e-06, "loss": 0.009, "step": 23304 }, { "epoch": 2.526561144839549, "grad_norm": 0.31219482421875, "learning_rate": 3.167706368779922e-06, "loss": 0.0068, "step": 23305 }, { "epoch": 2.526669557675629, "grad_norm": 0.03327617421746254, "learning_rate": 3.1669809952125347e-06, "loss": 0.0003, "step": 23306 }, { "epoch": 2.5267779705117084, "grad_norm": 0.034669406712055206, "learning_rate": 3.1662556216451472e-06, "loss": 0.0003, "step": 23307 }, { "epoch": 2.5268863833477884, "grad_norm": 1.638047456741333, "learning_rate": 3.1655302480777606e-06, "loss": 0.0178, "step": 23308 }, { "epoch": 2.5269947961838684, "grad_norm": 2.0876970291137695, "learning_rate": 3.164804874510373e-06, "loss": 0.0561, "step": 23309 }, { "epoch": 2.527103209019948, "grad_norm": 0.21995514631271362, "learning_rate": 3.1640795009429858e-06, "loss": 0.013, "step": 23310 }, { "epoch": 2.527211621856028, "grad_norm": 0.4582537114620209, "learning_rate": 3.1633541273755983e-06, "loss": 0.0161, "step": 23311 }, { "epoch": 2.5273200346921074, "grad_norm": 8.301465034484863, "learning_rate": 3.1626287538082113e-06, "loss": 0.007, "step": 23312 }, { "epoch": 2.5274284475281874, "grad_norm": 8.143798828125, "learning_rate": 3.1619033802408243e-06, "loss": 0.0382, "step": 23313 }, { "epoch": 2.5275368603642674, "grad_norm": 7.88559103012085, "learning_rate": 3.1611780066734373e-06, "loss": 0.0357, "step": 23314 }, { "epoch": 2.527645273200347, "grad_norm": 2.159623384475708, "learning_rate": 3.16045263310605e-06, "loss": 0.0135, "step": 23315 }, { "epoch": 2.5277536860364265, "grad_norm": 0.10424283891916275, "learning_rate": 3.1597272595386624e-06, "loss": 0.0124, "step": 23316 }, { "epoch": 2.5278620988725065, "grad_norm": 0.4143771529197693, "learning_rate": 3.159001885971276e-06, "loss": 0.0106, "step": 23317 }, { "epoch": 2.5279705117085864, "grad_norm": 0.012174504809081554, "learning_rate": 3.1582765124038884e-06, "loss": 0.0, "step": 23318 }, { "epoch": 2.528078924544666, "grad_norm": 0.0038126970175653696, "learning_rate": 3.157551138836501e-06, "loss": 0.0, "step": 23319 }, { "epoch": 2.528187337380746, "grad_norm": 0.9538012146949768, "learning_rate": 3.1568257652691136e-06, "loss": 0.0046, "step": 23320 }, { "epoch": 2.5282957502168255, "grad_norm": 5.210875988006592, "learning_rate": 3.156100391701727e-06, "loss": 0.0455, "step": 23321 }, { "epoch": 2.5284041630529055, "grad_norm": 0.446172297000885, "learning_rate": 3.1553750181343395e-06, "loss": 0.0112, "step": 23322 }, { "epoch": 2.5285125758889855, "grad_norm": 7.733118534088135, "learning_rate": 3.154649644566952e-06, "loss": 0.0255, "step": 23323 }, { "epoch": 2.528620988725065, "grad_norm": 0.2853318154811859, "learning_rate": 3.1539242709995647e-06, "loss": 0.0044, "step": 23324 }, { "epoch": 2.5287294015611446, "grad_norm": 0.006088569760322571, "learning_rate": 3.1531988974321777e-06, "loss": 0.0001, "step": 23325 }, { "epoch": 2.5288378143972245, "grad_norm": 1.0802890062332153, "learning_rate": 3.1524735238647907e-06, "loss": 0.0211, "step": 23326 }, { "epoch": 2.5289462272333045, "grad_norm": 0.5940074920654297, "learning_rate": 3.1517481502974036e-06, "loss": 0.0212, "step": 23327 }, { "epoch": 2.529054640069384, "grad_norm": 0.02429656870663166, "learning_rate": 3.1510227767300162e-06, "loss": 0.0001, "step": 23328 }, { "epoch": 2.529163052905464, "grad_norm": 4.93675422668457, "learning_rate": 3.1502974031626288e-06, "loss": 0.0058, "step": 23329 }, { "epoch": 2.5292714657415436, "grad_norm": 4.627369403839111, "learning_rate": 3.149572029595242e-06, "loss": 0.0473, "step": 23330 }, { "epoch": 2.5293798785776236, "grad_norm": 0.40251463651657104, "learning_rate": 3.1488466560278548e-06, "loss": 0.0142, "step": 23331 }, { "epoch": 2.5294882914137036, "grad_norm": 0.02054276131093502, "learning_rate": 3.1481212824604673e-06, "loss": 0.0002, "step": 23332 }, { "epoch": 2.529596704249783, "grad_norm": 0.7367944717407227, "learning_rate": 3.14739590889308e-06, "loss": 0.014, "step": 23333 }, { "epoch": 2.529705117085863, "grad_norm": 0.9734072089195251, "learning_rate": 3.1466705353256933e-06, "loss": 0.0068, "step": 23334 }, { "epoch": 2.5298135299219426, "grad_norm": 0.2698630094528198, "learning_rate": 3.145945161758306e-06, "loss": 0.0059, "step": 23335 }, { "epoch": 2.5299219427580226, "grad_norm": 2.6096959114074707, "learning_rate": 3.1452197881909184e-06, "loss": 0.0454, "step": 23336 }, { "epoch": 2.5300303555941026, "grad_norm": 16.6141300201416, "learning_rate": 3.1444944146235314e-06, "loss": 0.0341, "step": 23337 }, { "epoch": 2.530138768430182, "grad_norm": 0.5267683863639832, "learning_rate": 3.1437690410561444e-06, "loss": 0.0161, "step": 23338 }, { "epoch": 2.5302471812662617, "grad_norm": 0.24827487766742706, "learning_rate": 3.143043667488757e-06, "loss": 0.0051, "step": 23339 }, { "epoch": 2.5303555941023417, "grad_norm": 1.8724021911621094, "learning_rate": 3.14231829392137e-06, "loss": 0.011, "step": 23340 }, { "epoch": 2.5304640069384217, "grad_norm": 0.6518787145614624, "learning_rate": 3.1415929203539825e-06, "loss": 0.0006, "step": 23341 }, { "epoch": 2.530572419774501, "grad_norm": 0.4764142334461212, "learning_rate": 3.140867546786595e-06, "loss": 0.0016, "step": 23342 }, { "epoch": 2.530680832610581, "grad_norm": 5.237224578857422, "learning_rate": 3.1401421732192085e-06, "loss": 0.0192, "step": 23343 }, { "epoch": 2.5307892454466607, "grad_norm": 0.6576162576675415, "learning_rate": 3.139416799651821e-06, "loss": 0.0119, "step": 23344 }, { "epoch": 2.5308976582827407, "grad_norm": 0.3953620493412018, "learning_rate": 3.1386914260844337e-06, "loss": 0.0146, "step": 23345 }, { "epoch": 2.5310060711188207, "grad_norm": 0.2256128489971161, "learning_rate": 3.1379660525170462e-06, "loss": 0.0143, "step": 23346 }, { "epoch": 2.5311144839549002, "grad_norm": 6.356628894805908, "learning_rate": 3.1372406789496596e-06, "loss": 0.0294, "step": 23347 }, { "epoch": 2.53122289679098, "grad_norm": 0.01240715105086565, "learning_rate": 3.136515305382272e-06, "loss": 0.0001, "step": 23348 }, { "epoch": 2.5313313096270598, "grad_norm": 6.254459381103516, "learning_rate": 3.1357899318148848e-06, "loss": 0.0406, "step": 23349 }, { "epoch": 2.5314397224631398, "grad_norm": 0.002272038022056222, "learning_rate": 3.1350645582474978e-06, "loss": 0.0, "step": 23350 }, { "epoch": 2.5315481352992193, "grad_norm": 0.2776361405849457, "learning_rate": 3.1343391846801108e-06, "loss": 0.0059, "step": 23351 }, { "epoch": 2.5316565481352993, "grad_norm": 7.930047988891602, "learning_rate": 3.1336138111127233e-06, "loss": 0.0189, "step": 23352 }, { "epoch": 2.531764960971379, "grad_norm": 0.08931468427181244, "learning_rate": 3.1328884375453363e-06, "loss": 0.0005, "step": 23353 }, { "epoch": 2.531873373807459, "grad_norm": 3.564628839492798, "learning_rate": 3.132163063977949e-06, "loss": 0.0195, "step": 23354 }, { "epoch": 2.531981786643539, "grad_norm": 0.4198225736618042, "learning_rate": 3.1314376904105614e-06, "loss": 0.01, "step": 23355 }, { "epoch": 2.5320901994796183, "grad_norm": 5.883105754852295, "learning_rate": 3.130712316843175e-06, "loss": 0.0196, "step": 23356 }, { "epoch": 2.5321986123156983, "grad_norm": 0.20737333595752716, "learning_rate": 3.1299869432757874e-06, "loss": 0.0034, "step": 23357 }, { "epoch": 2.532307025151778, "grad_norm": 0.019382774829864502, "learning_rate": 3.1292615697084e-06, "loss": 0.0001, "step": 23358 }, { "epoch": 2.532415437987858, "grad_norm": 2.1329240798950195, "learning_rate": 3.1285361961410126e-06, "loss": 0.0256, "step": 23359 }, { "epoch": 2.532523850823938, "grad_norm": 0.15694765746593475, "learning_rate": 3.127810822573626e-06, "loss": 0.0002, "step": 23360 }, { "epoch": 2.5326322636600174, "grad_norm": 0.4421863853931427, "learning_rate": 3.1270854490062385e-06, "loss": 0.0117, "step": 23361 }, { "epoch": 2.532740676496097, "grad_norm": 0.2898736298084259, "learning_rate": 3.126360075438851e-06, "loss": 0.0002, "step": 23362 }, { "epoch": 2.532849089332177, "grad_norm": 0.5280722379684448, "learning_rate": 3.125634701871464e-06, "loss": 0.0111, "step": 23363 }, { "epoch": 2.532957502168257, "grad_norm": 0.48115476965904236, "learning_rate": 3.124909328304077e-06, "loss": 0.0072, "step": 23364 }, { "epoch": 2.5330659150043364, "grad_norm": 2.3585333824157715, "learning_rate": 3.1241839547366897e-06, "loss": 0.0406, "step": 23365 }, { "epoch": 2.5331743278404164, "grad_norm": 0.014613045379519463, "learning_rate": 3.1234585811693026e-06, "loss": 0.0001, "step": 23366 }, { "epoch": 2.533282740676496, "grad_norm": 0.6193884611129761, "learning_rate": 3.1227332076019152e-06, "loss": 0.0166, "step": 23367 }, { "epoch": 2.533391153512576, "grad_norm": 2.686070680618286, "learning_rate": 3.1220078340345278e-06, "loss": 0.0141, "step": 23368 }, { "epoch": 2.533499566348656, "grad_norm": 1.5936545133590698, "learning_rate": 3.121282460467141e-06, "loss": 0.001, "step": 23369 }, { "epoch": 2.5336079791847355, "grad_norm": 0.3790408968925476, "learning_rate": 3.1205570868997538e-06, "loss": 0.0128, "step": 23370 }, { "epoch": 2.533716392020815, "grad_norm": 0.7010996341705322, "learning_rate": 3.1198317133323663e-06, "loss": 0.0346, "step": 23371 }, { "epoch": 2.533824804856895, "grad_norm": 3.34883451461792, "learning_rate": 3.119106339764979e-06, "loss": 0.0186, "step": 23372 }, { "epoch": 2.533933217692975, "grad_norm": 0.46646448969841003, "learning_rate": 3.1183809661975923e-06, "loss": 0.0105, "step": 23373 }, { "epoch": 2.5340416305290545, "grad_norm": 0.4700627326965332, "learning_rate": 3.117655592630205e-06, "loss": 0.0112, "step": 23374 }, { "epoch": 2.5341500433651345, "grad_norm": 0.27415943145751953, "learning_rate": 3.1169302190628174e-06, "loss": 0.0055, "step": 23375 }, { "epoch": 2.534258456201214, "grad_norm": 3.4773099422454834, "learning_rate": 3.1162048454954304e-06, "loss": 0.0216, "step": 23376 }, { "epoch": 2.534366869037294, "grad_norm": 3.4198334217071533, "learning_rate": 3.1154794719280434e-06, "loss": 0.027, "step": 23377 }, { "epoch": 2.534475281873374, "grad_norm": 0.4003036320209503, "learning_rate": 3.114754098360656e-06, "loss": 0.0044, "step": 23378 }, { "epoch": 2.5345836947094535, "grad_norm": 0.3669659495353699, "learning_rate": 3.114028724793269e-06, "loss": 0.0108, "step": 23379 }, { "epoch": 2.5346921075455335, "grad_norm": 0.29098349809646606, "learning_rate": 3.1133033512258815e-06, "loss": 0.0056, "step": 23380 }, { "epoch": 2.534800520381613, "grad_norm": 7.01312780380249, "learning_rate": 3.112577977658494e-06, "loss": 0.0313, "step": 23381 }, { "epoch": 2.534908933217693, "grad_norm": 1.4255430698394775, "learning_rate": 3.1118526040911075e-06, "loss": 0.0422, "step": 23382 }, { "epoch": 2.5350173460537726, "grad_norm": 0.28553611040115356, "learning_rate": 3.11112723052372e-06, "loss": 0.006, "step": 23383 }, { "epoch": 2.5351257588898526, "grad_norm": 0.027909476310014725, "learning_rate": 3.1104018569563327e-06, "loss": 0.0002, "step": 23384 }, { "epoch": 2.535234171725932, "grad_norm": 0.35226935148239136, "learning_rate": 3.1096764833889452e-06, "loss": 0.0091, "step": 23385 }, { "epoch": 2.535342584562012, "grad_norm": 0.29001447558403015, "learning_rate": 3.1089511098215586e-06, "loss": 0.006, "step": 23386 }, { "epoch": 2.535450997398092, "grad_norm": 0.04224088415503502, "learning_rate": 3.1082257362541712e-06, "loss": 0.0002, "step": 23387 }, { "epoch": 2.5355594102341716, "grad_norm": 0.06693077832460403, "learning_rate": 3.1075003626867838e-06, "loss": 0.0003, "step": 23388 }, { "epoch": 2.5356678230702516, "grad_norm": 15.561963081359863, "learning_rate": 3.1067749891193968e-06, "loss": 0.0471, "step": 23389 }, { "epoch": 2.535776235906331, "grad_norm": 5.1540117263793945, "learning_rate": 3.1060496155520098e-06, "loss": 0.0122, "step": 23390 }, { "epoch": 2.535884648742411, "grad_norm": 0.04192650318145752, "learning_rate": 3.1053242419846223e-06, "loss": 0.0002, "step": 23391 }, { "epoch": 2.535993061578491, "grad_norm": 1.3517684936523438, "learning_rate": 3.1045988684172353e-06, "loss": 0.0073, "step": 23392 }, { "epoch": 2.5361014744145707, "grad_norm": 0.4419611096382141, "learning_rate": 3.103873494849848e-06, "loss": 0.0158, "step": 23393 }, { "epoch": 2.53620988725065, "grad_norm": 0.09221594780683517, "learning_rate": 3.1031481212824605e-06, "loss": 0.0004, "step": 23394 }, { "epoch": 2.53631830008673, "grad_norm": 0.30264726281166077, "learning_rate": 3.102422747715074e-06, "loss": 0.0103, "step": 23395 }, { "epoch": 2.53642671292281, "grad_norm": 0.4365938901901245, "learning_rate": 3.1016973741476864e-06, "loss": 0.0171, "step": 23396 }, { "epoch": 2.5365351257588897, "grad_norm": 0.3200567662715912, "learning_rate": 3.100972000580299e-06, "loss": 0.0109, "step": 23397 }, { "epoch": 2.5366435385949697, "grad_norm": 0.11995067447423935, "learning_rate": 3.1002466270129116e-06, "loss": 0.0006, "step": 23398 }, { "epoch": 2.5367519514310493, "grad_norm": 0.2918887734413147, "learning_rate": 3.099521253445525e-06, "loss": 0.006, "step": 23399 }, { "epoch": 2.5368603642671292, "grad_norm": 0.5324808359146118, "learning_rate": 3.0987958798781375e-06, "loss": 0.0196, "step": 23400 }, { "epoch": 2.5369687771032092, "grad_norm": 1.0359013080596924, "learning_rate": 3.09807050631075e-06, "loss": 0.0138, "step": 23401 }, { "epoch": 2.5370771899392888, "grad_norm": 0.2609364688396454, "learning_rate": 3.097345132743363e-06, "loss": 0.0055, "step": 23402 }, { "epoch": 2.5371856027753688, "grad_norm": 0.34827566146850586, "learning_rate": 3.096619759175976e-06, "loss": 0.0375, "step": 23403 }, { "epoch": 2.5372940156114483, "grad_norm": 0.09313598275184631, "learning_rate": 3.0958943856085887e-06, "loss": 0.0002, "step": 23404 }, { "epoch": 2.5374024284475283, "grad_norm": 0.3155261278152466, "learning_rate": 3.0951690120412017e-06, "loss": 0.0064, "step": 23405 }, { "epoch": 2.537510841283608, "grad_norm": 0.26286089420318604, "learning_rate": 3.0944436384738142e-06, "loss": 0.0055, "step": 23406 }, { "epoch": 2.537619254119688, "grad_norm": 0.36783695220947266, "learning_rate": 3.0937182649064268e-06, "loss": 0.0113, "step": 23407 }, { "epoch": 2.5377276669557673, "grad_norm": 0.2558611035346985, "learning_rate": 3.09299289133904e-06, "loss": 0.0121, "step": 23408 }, { "epoch": 2.5378360797918473, "grad_norm": 0.29978057742118835, "learning_rate": 3.0922675177716528e-06, "loss": 0.0039, "step": 23409 }, { "epoch": 2.5379444926279273, "grad_norm": 0.007976983673870564, "learning_rate": 3.0915421442042653e-06, "loss": 0.0001, "step": 23410 }, { "epoch": 2.538052905464007, "grad_norm": 0.029945511370897293, "learning_rate": 3.090816770636878e-06, "loss": 0.0001, "step": 23411 }, { "epoch": 2.538161318300087, "grad_norm": 0.5356831550598145, "learning_rate": 3.0900913970694913e-06, "loss": 0.0181, "step": 23412 }, { "epoch": 2.5382697311361664, "grad_norm": 3.405167818069458, "learning_rate": 3.089366023502104e-06, "loss": 0.0372, "step": 23413 }, { "epoch": 2.5383781439722464, "grad_norm": 2.0815398693084717, "learning_rate": 3.0886406499347164e-06, "loss": 0.0385, "step": 23414 }, { "epoch": 2.5384865568083264, "grad_norm": 0.20798996090888977, "learning_rate": 3.0879152763673294e-06, "loss": 0.0033, "step": 23415 }, { "epoch": 2.538594969644406, "grad_norm": 0.2302447110414505, "learning_rate": 3.0871899027999424e-06, "loss": 0.0033, "step": 23416 }, { "epoch": 2.5387033824804854, "grad_norm": 0.013544144108891487, "learning_rate": 3.086464529232555e-06, "loss": 0.0001, "step": 23417 }, { "epoch": 2.5388117953165654, "grad_norm": 0.04372250288724899, "learning_rate": 3.085739155665168e-06, "loss": 0.0003, "step": 23418 }, { "epoch": 2.5389202081526454, "grad_norm": 0.5110712647438049, "learning_rate": 3.0850137820977806e-06, "loss": 0.0157, "step": 23419 }, { "epoch": 2.539028620988725, "grad_norm": 0.3098068833351135, "learning_rate": 3.084288408530393e-06, "loss": 0.0114, "step": 23420 }, { "epoch": 2.539137033824805, "grad_norm": 0.019717061892151833, "learning_rate": 3.0835630349630065e-06, "loss": 0.0001, "step": 23421 }, { "epoch": 2.5392454466608845, "grad_norm": 0.0013206896837800741, "learning_rate": 3.082837661395619e-06, "loss": 0.0, "step": 23422 }, { "epoch": 2.5393538594969645, "grad_norm": 6.6007981300354, "learning_rate": 3.0821122878282317e-06, "loss": 0.026, "step": 23423 }, { "epoch": 2.5394622723330444, "grad_norm": 0.08596822619438171, "learning_rate": 3.0813869142608442e-06, "loss": 0.0004, "step": 23424 }, { "epoch": 2.539570685169124, "grad_norm": 1.4420087337493896, "learning_rate": 3.0806615406934576e-06, "loss": 0.011, "step": 23425 }, { "epoch": 2.539679098005204, "grad_norm": 0.2789021134376526, "learning_rate": 3.0799361671260702e-06, "loss": 0.0061, "step": 23426 }, { "epoch": 2.5397875108412835, "grad_norm": 0.33286052942276, "learning_rate": 3.0792107935586828e-06, "loss": 0.0069, "step": 23427 }, { "epoch": 2.5398959236773635, "grad_norm": 0.004420100711286068, "learning_rate": 3.0784854199912958e-06, "loss": 0.0, "step": 23428 }, { "epoch": 2.540004336513443, "grad_norm": 0.027815377339720726, "learning_rate": 3.0777600464239088e-06, "loss": 0.0001, "step": 23429 }, { "epoch": 2.540112749349523, "grad_norm": 0.2777765095233917, "learning_rate": 3.0770346728565213e-06, "loss": 0.0053, "step": 23430 }, { "epoch": 2.5402211621856026, "grad_norm": 0.4056756794452667, "learning_rate": 3.0763092992891343e-06, "loss": 0.0081, "step": 23431 }, { "epoch": 2.5403295750216826, "grad_norm": 5.895418643951416, "learning_rate": 3.075583925721747e-06, "loss": 0.0191, "step": 23432 }, { "epoch": 2.5404379878577625, "grad_norm": 0.463105171918869, "learning_rate": 3.07485855215436e-06, "loss": 0.0202, "step": 23433 }, { "epoch": 2.540546400693842, "grad_norm": 0.19216829538345337, "learning_rate": 3.074133178586973e-06, "loss": 0.0027, "step": 23434 }, { "epoch": 2.540654813529922, "grad_norm": 0.2819229066371918, "learning_rate": 3.0734078050195854e-06, "loss": 0.0038, "step": 23435 }, { "epoch": 2.5407632263660016, "grad_norm": 1.5148305892944336, "learning_rate": 3.072682431452198e-06, "loss": 0.0025, "step": 23436 }, { "epoch": 2.5408716392020816, "grad_norm": 1.0098960399627686, "learning_rate": 3.0719570578848106e-06, "loss": 0.0004, "step": 23437 }, { "epoch": 2.5409800520381616, "grad_norm": 0.01778276078402996, "learning_rate": 3.071231684317424e-06, "loss": 0.0, "step": 23438 }, { "epoch": 2.541088464874241, "grad_norm": 0.30767199397087097, "learning_rate": 3.0705063107500365e-06, "loss": 0.0072, "step": 23439 }, { "epoch": 2.5411968777103207, "grad_norm": 0.3017749488353729, "learning_rate": 3.069780937182649e-06, "loss": 0.011, "step": 23440 }, { "epoch": 2.5413052905464006, "grad_norm": 0.026778893545269966, "learning_rate": 3.069055563615262e-06, "loss": 0.0002, "step": 23441 }, { "epoch": 2.5414137033824806, "grad_norm": 0.5655446648597717, "learning_rate": 3.068330190047875e-06, "loss": 0.0157, "step": 23442 }, { "epoch": 2.54152211621856, "grad_norm": 0.008476078510284424, "learning_rate": 3.0676048164804877e-06, "loss": 0.0001, "step": 23443 }, { "epoch": 2.54163052905464, "grad_norm": 0.055035825818777084, "learning_rate": 3.0668794429131007e-06, "loss": 0.0003, "step": 23444 }, { "epoch": 2.5417389418907197, "grad_norm": 2.5458593368530273, "learning_rate": 3.0661540693457132e-06, "loss": 0.0316, "step": 23445 }, { "epoch": 2.5418473547267997, "grad_norm": 0.3865117132663727, "learning_rate": 3.0654286957783262e-06, "loss": 0.0112, "step": 23446 }, { "epoch": 2.5419557675628797, "grad_norm": 0.3359050452709198, "learning_rate": 3.064703322210939e-06, "loss": 0.0142, "step": 23447 }, { "epoch": 2.542064180398959, "grad_norm": 0.743471622467041, "learning_rate": 3.0639779486435518e-06, "loss": 0.0007, "step": 23448 }, { "epoch": 2.542172593235039, "grad_norm": 0.5303520560264587, "learning_rate": 3.0632525750761643e-06, "loss": 0.0145, "step": 23449 }, { "epoch": 2.5422810060711187, "grad_norm": 3.9508025646209717, "learning_rate": 3.062527201508777e-06, "loss": 0.0334, "step": 23450 }, { "epoch": 2.5423894189071987, "grad_norm": 0.24262475967407227, "learning_rate": 3.0618018279413903e-06, "loss": 0.0051, "step": 23451 }, { "epoch": 2.5424978317432783, "grad_norm": 0.0026394478045403957, "learning_rate": 3.061076454374003e-06, "loss": 0.0, "step": 23452 }, { "epoch": 2.5426062445793582, "grad_norm": 0.029338262975215912, "learning_rate": 3.0603510808066155e-06, "loss": 0.0002, "step": 23453 }, { "epoch": 2.542714657415438, "grad_norm": 2.6445438861846924, "learning_rate": 3.0596257072392284e-06, "loss": 0.0365, "step": 23454 }, { "epoch": 2.5428230702515178, "grad_norm": 0.08331630378961563, "learning_rate": 3.0589003336718414e-06, "loss": 0.0003, "step": 23455 }, { "epoch": 2.5429314830875978, "grad_norm": 0.2614711821079254, "learning_rate": 3.058174960104454e-06, "loss": 0.003, "step": 23456 }, { "epoch": 2.5430398959236773, "grad_norm": 0.5578105449676514, "learning_rate": 3.057449586537067e-06, "loss": 0.0152, "step": 23457 }, { "epoch": 2.5431483087597573, "grad_norm": 0.019070714712142944, "learning_rate": 3.0567242129696796e-06, "loss": 0.0002, "step": 23458 }, { "epoch": 2.543256721595837, "grad_norm": 0.04652559012174606, "learning_rate": 3.0559988394022925e-06, "loss": 0.0003, "step": 23459 }, { "epoch": 2.543365134431917, "grad_norm": 1.907235860824585, "learning_rate": 3.0552734658349055e-06, "loss": 0.0658, "step": 23460 }, { "epoch": 2.543473547267997, "grad_norm": 6.792980194091797, "learning_rate": 3.054548092267518e-06, "loss": 0.0301, "step": 23461 }, { "epoch": 2.5435819601040763, "grad_norm": 1.067470669746399, "learning_rate": 3.0538227187001307e-06, "loss": 0.0016, "step": 23462 }, { "epoch": 2.543690372940156, "grad_norm": 0.8896917700767517, "learning_rate": 3.0530973451327432e-06, "loss": 0.0263, "step": 23463 }, { "epoch": 2.543798785776236, "grad_norm": 0.5089352130889893, "learning_rate": 3.0523719715653567e-06, "loss": 0.0215, "step": 23464 }, { "epoch": 2.543907198612316, "grad_norm": 0.10079484432935715, "learning_rate": 3.0516465979979692e-06, "loss": 0.0004, "step": 23465 }, { "epoch": 2.5440156114483954, "grad_norm": 0.3978927731513977, "learning_rate": 3.0509212244305818e-06, "loss": 0.0067, "step": 23466 }, { "epoch": 2.5441240242844754, "grad_norm": 0.296938419342041, "learning_rate": 3.0501958508631948e-06, "loss": 0.0062, "step": 23467 }, { "epoch": 2.544232437120555, "grad_norm": 0.3235166370868683, "learning_rate": 3.0494704772958078e-06, "loss": 0.0062, "step": 23468 }, { "epoch": 2.544340849956635, "grad_norm": 4.574334144592285, "learning_rate": 3.0487451037284203e-06, "loss": 0.0207, "step": 23469 }, { "epoch": 2.544449262792715, "grad_norm": 0.3625386953353882, "learning_rate": 3.0480197301610333e-06, "loss": 0.0097, "step": 23470 }, { "epoch": 2.5445576756287944, "grad_norm": 9.08122730255127, "learning_rate": 3.047294356593646e-06, "loss": 0.0735, "step": 23471 }, { "epoch": 2.544666088464874, "grad_norm": 0.25377967953681946, "learning_rate": 3.046568983026259e-06, "loss": 0.0052, "step": 23472 }, { "epoch": 2.544774501300954, "grad_norm": 1.0455738306045532, "learning_rate": 3.045843609458872e-06, "loss": 0.0136, "step": 23473 }, { "epoch": 2.544882914137034, "grad_norm": 5.661748886108398, "learning_rate": 3.0451182358914844e-06, "loss": 0.0559, "step": 23474 }, { "epoch": 2.5449913269731135, "grad_norm": 0.022181572392582893, "learning_rate": 3.044392862324097e-06, "loss": 0.0001, "step": 23475 }, { "epoch": 2.5450997398091935, "grad_norm": 2.7911903858184814, "learning_rate": 3.0436674887567096e-06, "loss": 0.0235, "step": 23476 }, { "epoch": 2.545208152645273, "grad_norm": 0.04895082861185074, "learning_rate": 3.042942115189323e-06, "loss": 0.0004, "step": 23477 }, { "epoch": 2.545316565481353, "grad_norm": 6.277468204498291, "learning_rate": 3.0422167416219356e-06, "loss": 0.0235, "step": 23478 }, { "epoch": 2.545424978317433, "grad_norm": 0.1575033813714981, "learning_rate": 3.041491368054548e-06, "loss": 0.0111, "step": 23479 }, { "epoch": 2.5455333911535125, "grad_norm": 0.27375727891921997, "learning_rate": 3.040765994487161e-06, "loss": 0.0058, "step": 23480 }, { "epoch": 2.5456418039895925, "grad_norm": 0.041033752262592316, "learning_rate": 3.040040620919774e-06, "loss": 0.0002, "step": 23481 }, { "epoch": 2.545750216825672, "grad_norm": 0.010990172624588013, "learning_rate": 3.0393152473523867e-06, "loss": 0.0001, "step": 23482 }, { "epoch": 2.545858629661752, "grad_norm": 0.5455094575881958, "learning_rate": 3.0385898737849997e-06, "loss": 0.0109, "step": 23483 }, { "epoch": 2.545967042497832, "grad_norm": 0.05826394632458687, "learning_rate": 3.0378645002176122e-06, "loss": 0.0004, "step": 23484 }, { "epoch": 2.5460754553339116, "grad_norm": 0.3805408477783203, "learning_rate": 3.0371391266502252e-06, "loss": 0.0088, "step": 23485 }, { "epoch": 2.546183868169991, "grad_norm": 0.006053881254047155, "learning_rate": 3.036413753082838e-06, "loss": 0.0001, "step": 23486 }, { "epoch": 2.546292281006071, "grad_norm": 0.4899826645851135, "learning_rate": 3.0356883795154508e-06, "loss": 0.0097, "step": 23487 }, { "epoch": 2.546400693842151, "grad_norm": 5.481777667999268, "learning_rate": 3.0349630059480633e-06, "loss": 0.0165, "step": 23488 }, { "epoch": 2.5465091066782306, "grad_norm": 1.2260299921035767, "learning_rate": 3.034237632380676e-06, "loss": 0.0188, "step": 23489 }, { "epoch": 2.5466175195143106, "grad_norm": 0.4233882427215576, "learning_rate": 3.0335122588132893e-06, "loss": 0.0096, "step": 23490 }, { "epoch": 2.54672593235039, "grad_norm": 0.6528691053390503, "learning_rate": 3.032786885245902e-06, "loss": 0.0195, "step": 23491 }, { "epoch": 2.54683434518647, "grad_norm": 12.107983589172363, "learning_rate": 3.0320615116785145e-06, "loss": 0.0172, "step": 23492 }, { "epoch": 2.54694275802255, "grad_norm": 0.17375032603740692, "learning_rate": 3.0313361381111274e-06, "loss": 0.0098, "step": 23493 }, { "epoch": 2.5470511708586296, "grad_norm": 0.029319239780306816, "learning_rate": 3.0306107645437404e-06, "loss": 0.0001, "step": 23494 }, { "epoch": 2.547159583694709, "grad_norm": 0.7631980180740356, "learning_rate": 3.029885390976353e-06, "loss": 0.0096, "step": 23495 }, { "epoch": 2.547267996530789, "grad_norm": 0.21570852398872375, "learning_rate": 3.029160017408966e-06, "loss": 0.0002, "step": 23496 }, { "epoch": 2.547376409366869, "grad_norm": 0.0581030510365963, "learning_rate": 3.0284346438415786e-06, "loss": 0.0003, "step": 23497 }, { "epoch": 2.5474848222029487, "grad_norm": 3.1538243293762207, "learning_rate": 3.0277092702741915e-06, "loss": 0.0646, "step": 23498 }, { "epoch": 2.5475932350390287, "grad_norm": 0.3163089454174042, "learning_rate": 3.0269838967068045e-06, "loss": 0.0118, "step": 23499 }, { "epoch": 2.547701647875108, "grad_norm": 1.4655935764312744, "learning_rate": 3.026258523139417e-06, "loss": 0.0357, "step": 23500 }, { "epoch": 2.547810060711188, "grad_norm": 0.4655853807926178, "learning_rate": 3.0255331495720297e-06, "loss": 0.0172, "step": 23501 }, { "epoch": 2.547918473547268, "grad_norm": 3.188016176223755, "learning_rate": 3.0248077760046422e-06, "loss": 0.0173, "step": 23502 }, { "epoch": 2.5480268863833477, "grad_norm": 0.013789938762784004, "learning_rate": 3.0240824024372557e-06, "loss": 0.0002, "step": 23503 }, { "epoch": 2.5481352992194277, "grad_norm": 0.1443873792886734, "learning_rate": 3.0233570288698682e-06, "loss": 0.0004, "step": 23504 }, { "epoch": 2.5482437120555073, "grad_norm": 0.23954671621322632, "learning_rate": 3.022631655302481e-06, "loss": 0.0045, "step": 23505 }, { "epoch": 2.5483521248915872, "grad_norm": 0.11586197465658188, "learning_rate": 3.0219062817350938e-06, "loss": 0.0015, "step": 23506 }, { "epoch": 2.5484605377276672, "grad_norm": 0.7637877464294434, "learning_rate": 3.0211809081677068e-06, "loss": 0.004, "step": 23507 }, { "epoch": 2.5485689505637468, "grad_norm": 6.844396114349365, "learning_rate": 3.0204555346003193e-06, "loss": 0.0266, "step": 23508 }, { "epoch": 2.5486773633998263, "grad_norm": 0.5669618844985962, "learning_rate": 3.0197301610329323e-06, "loss": 0.0092, "step": 23509 }, { "epoch": 2.5487857762359063, "grad_norm": 0.3421328365802765, "learning_rate": 3.019004787465545e-06, "loss": 0.0125, "step": 23510 }, { "epoch": 2.5488941890719863, "grad_norm": 0.32065144181251526, "learning_rate": 3.018279413898158e-06, "loss": 0.0067, "step": 23511 }, { "epoch": 2.549002601908066, "grad_norm": 0.3687966465950012, "learning_rate": 3.017554040330771e-06, "loss": 0.0073, "step": 23512 }, { "epoch": 2.549111014744146, "grad_norm": 0.08797018975019455, "learning_rate": 3.0168286667633834e-06, "loss": 0.0005, "step": 23513 }, { "epoch": 2.5492194275802254, "grad_norm": 6.353999614715576, "learning_rate": 3.016103293195996e-06, "loss": 0.031, "step": 23514 }, { "epoch": 2.5493278404163053, "grad_norm": 0.028239399194717407, "learning_rate": 3.0153779196286086e-06, "loss": 0.0001, "step": 23515 }, { "epoch": 2.5494362532523853, "grad_norm": 0.014551844447851181, "learning_rate": 3.014652546061222e-06, "loss": 0.0001, "step": 23516 }, { "epoch": 2.549544666088465, "grad_norm": 0.01556532271206379, "learning_rate": 3.0139271724938346e-06, "loss": 0.0001, "step": 23517 }, { "epoch": 2.5496530789245444, "grad_norm": 0.43684038519859314, "learning_rate": 3.013201798926447e-06, "loss": 0.0008, "step": 23518 }, { "epoch": 2.5497614917606244, "grad_norm": 3.734346628189087, "learning_rate": 3.01247642535906e-06, "loss": 0.03, "step": 23519 }, { "epoch": 2.5498699045967044, "grad_norm": 3.8541784286499023, "learning_rate": 3.011751051791673e-06, "loss": 0.0061, "step": 23520 }, { "epoch": 2.549978317432784, "grad_norm": 0.49274981021881104, "learning_rate": 3.0110256782242857e-06, "loss": 0.01, "step": 23521 }, { "epoch": 2.550086730268864, "grad_norm": 0.42269256711006165, "learning_rate": 3.0103003046568987e-06, "loss": 0.0078, "step": 23522 }, { "epoch": 2.5501951431049434, "grad_norm": 0.8352177143096924, "learning_rate": 3.0095749310895112e-06, "loss": 0.0128, "step": 23523 }, { "epoch": 2.5503035559410234, "grad_norm": 0.7338119745254517, "learning_rate": 3.0088495575221242e-06, "loss": 0.0176, "step": 23524 }, { "epoch": 2.5504119687771034, "grad_norm": 4.611193656921387, "learning_rate": 3.008124183954737e-06, "loss": 0.0431, "step": 23525 }, { "epoch": 2.550520381613183, "grad_norm": 1.921676516532898, "learning_rate": 3.0073988103873498e-06, "loss": 0.0239, "step": 23526 }, { "epoch": 2.550628794449263, "grad_norm": 7.056471347808838, "learning_rate": 3.0066734368199623e-06, "loss": 0.0143, "step": 23527 }, { "epoch": 2.5507372072853425, "grad_norm": 0.61433345079422, "learning_rate": 3.0059480632525758e-06, "loss": 0.0059, "step": 23528 }, { "epoch": 2.5508456201214225, "grad_norm": 0.1312824785709381, "learning_rate": 3.0052226896851883e-06, "loss": 0.0004, "step": 23529 }, { "epoch": 2.550954032957502, "grad_norm": 1.657477617263794, "learning_rate": 3.004497316117801e-06, "loss": 0.0062, "step": 23530 }, { "epoch": 2.551062445793582, "grad_norm": 4.794061183929443, "learning_rate": 3.0037719425504135e-06, "loss": 0.0276, "step": 23531 }, { "epoch": 2.5511708586296615, "grad_norm": 0.007171884179115295, "learning_rate": 3.0030465689830264e-06, "loss": 0.0001, "step": 23532 }, { "epoch": 2.5512792714657415, "grad_norm": 0.3977459669113159, "learning_rate": 3.0023211954156394e-06, "loss": 0.0153, "step": 23533 }, { "epoch": 2.5513876843018215, "grad_norm": 0.2226577252149582, "learning_rate": 3.001595821848252e-06, "loss": 0.0022, "step": 23534 }, { "epoch": 2.551496097137901, "grad_norm": 4.072218894958496, "learning_rate": 3.000870448280865e-06, "loss": 0.016, "step": 23535 }, { "epoch": 2.551604509973981, "grad_norm": 7.599752426147461, "learning_rate": 3.0001450747134776e-06, "loss": 0.0191, "step": 23536 }, { "epoch": 2.5517129228100606, "grad_norm": 0.3976300358772278, "learning_rate": 2.9994197011460906e-06, "loss": 0.0144, "step": 23537 }, { "epoch": 2.5518213356461406, "grad_norm": 0.248207688331604, "learning_rate": 2.9986943275787035e-06, "loss": 0.005, "step": 23538 }, { "epoch": 2.5519297484822205, "grad_norm": 0.5643009543418884, "learning_rate": 2.997968954011316e-06, "loss": 0.029, "step": 23539 }, { "epoch": 2.5520381613183, "grad_norm": 2.1603734493255615, "learning_rate": 2.9972435804439287e-06, "loss": 0.0034, "step": 23540 }, { "epoch": 2.5521465741543796, "grad_norm": 2.7761285305023193, "learning_rate": 2.996518206876542e-06, "loss": 0.0092, "step": 23541 }, { "epoch": 2.5522549869904596, "grad_norm": 0.22471001744270325, "learning_rate": 2.9957928333091547e-06, "loss": 0.0007, "step": 23542 }, { "epoch": 2.5523633998265396, "grad_norm": 0.39058756828308105, "learning_rate": 2.9950674597417672e-06, "loss": 0.0122, "step": 23543 }, { "epoch": 2.552471812662619, "grad_norm": 0.7380124926567078, "learning_rate": 2.99434208617438e-06, "loss": 0.0238, "step": 23544 }, { "epoch": 2.552580225498699, "grad_norm": 0.47430217266082764, "learning_rate": 2.9936167126069928e-06, "loss": 0.0059, "step": 23545 }, { "epoch": 2.5526886383347787, "grad_norm": 0.03989981487393379, "learning_rate": 2.9928913390396058e-06, "loss": 0.0002, "step": 23546 }, { "epoch": 2.5527970511708586, "grad_norm": 0.234640434384346, "learning_rate": 2.9921659654722183e-06, "loss": 0.0004, "step": 23547 }, { "epoch": 2.5529054640069386, "grad_norm": 21.89449119567871, "learning_rate": 2.9914405919048313e-06, "loss": 0.0726, "step": 23548 }, { "epoch": 2.553013876843018, "grad_norm": 0.4242292046546936, "learning_rate": 2.990715218337444e-06, "loss": 0.0056, "step": 23549 }, { "epoch": 2.553122289679098, "grad_norm": 0.2695962190628052, "learning_rate": 2.989989844770057e-06, "loss": 0.0056, "step": 23550 }, { "epoch": 2.5532307025151777, "grad_norm": 0.27116742730140686, "learning_rate": 2.98926447120267e-06, "loss": 0.0061, "step": 23551 }, { "epoch": 2.5533391153512577, "grad_norm": 0.05357835441827774, "learning_rate": 2.9885390976352824e-06, "loss": 0.0003, "step": 23552 }, { "epoch": 2.5534475281873372, "grad_norm": 0.2465561032295227, "learning_rate": 2.987813724067895e-06, "loss": 0.0038, "step": 23553 }, { "epoch": 2.553555941023417, "grad_norm": 3.683969497680664, "learning_rate": 2.9870883505005084e-06, "loss": 0.0398, "step": 23554 }, { "epoch": 2.5536643538594967, "grad_norm": 0.5634563565254211, "learning_rate": 2.986362976933121e-06, "loss": 0.0006, "step": 23555 }, { "epoch": 2.5537727666955767, "grad_norm": 0.2713857889175415, "learning_rate": 2.9856376033657336e-06, "loss": 0.0056, "step": 23556 }, { "epoch": 2.5538811795316567, "grad_norm": 0.4276990592479706, "learning_rate": 2.984912229798346e-06, "loss": 0.0135, "step": 23557 }, { "epoch": 2.5539895923677363, "grad_norm": 0.5596051216125488, "learning_rate": 2.984186856230959e-06, "loss": 0.0094, "step": 23558 }, { "epoch": 2.5540980052038162, "grad_norm": 1.3945388793945312, "learning_rate": 2.983461482663572e-06, "loss": 0.0667, "step": 23559 }, { "epoch": 2.554206418039896, "grad_norm": 4.08774471282959, "learning_rate": 2.9827361090961847e-06, "loss": 0.0265, "step": 23560 }, { "epoch": 2.5543148308759758, "grad_norm": 0.5848839282989502, "learning_rate": 2.9820107355287977e-06, "loss": 0.0066, "step": 23561 }, { "epoch": 2.5544232437120558, "grad_norm": 0.3815280497074127, "learning_rate": 2.9812853619614102e-06, "loss": 0.0073, "step": 23562 }, { "epoch": 2.5545316565481353, "grad_norm": 0.05316554009914398, "learning_rate": 2.9805599883940232e-06, "loss": 0.0003, "step": 23563 }, { "epoch": 2.554640069384215, "grad_norm": 0.3875303864479065, "learning_rate": 2.9798346148266362e-06, "loss": 0.0065, "step": 23564 }, { "epoch": 2.554748482220295, "grad_norm": 0.33234089612960815, "learning_rate": 2.9791092412592488e-06, "loss": 0.0011, "step": 23565 }, { "epoch": 2.554856895056375, "grad_norm": 0.008490410633385181, "learning_rate": 2.9783838676918613e-06, "loss": 0.0, "step": 23566 }, { "epoch": 2.5549653078924544, "grad_norm": 0.27149420976638794, "learning_rate": 2.9776584941244748e-06, "loss": 0.0057, "step": 23567 }, { "epoch": 2.5550737207285343, "grad_norm": 0.11182645708322525, "learning_rate": 2.9769331205570873e-06, "loss": 0.0005, "step": 23568 }, { "epoch": 2.555182133564614, "grad_norm": 1.3171765804290771, "learning_rate": 2.9762077469897e-06, "loss": 0.0068, "step": 23569 }, { "epoch": 2.555290546400694, "grad_norm": 0.11966285109519958, "learning_rate": 2.9754823734223125e-06, "loss": 0.0007, "step": 23570 }, { "epoch": 2.555398959236774, "grad_norm": 3.4109349250793457, "learning_rate": 2.9747569998549255e-06, "loss": 0.0225, "step": 23571 }, { "epoch": 2.5555073720728534, "grad_norm": 0.5221417546272278, "learning_rate": 2.9740316262875384e-06, "loss": 0.0178, "step": 23572 }, { "epoch": 2.5556157849089334, "grad_norm": 1.0415105819702148, "learning_rate": 2.973306252720151e-06, "loss": 0.0045, "step": 23573 }, { "epoch": 2.555724197745013, "grad_norm": 0.8902451395988464, "learning_rate": 2.972580879152764e-06, "loss": 0.003, "step": 23574 }, { "epoch": 2.555832610581093, "grad_norm": 0.743640661239624, "learning_rate": 2.9718555055853766e-06, "loss": 0.0305, "step": 23575 }, { "epoch": 2.5559410234171724, "grad_norm": 0.0459156297147274, "learning_rate": 2.9711301320179896e-06, "loss": 0.0003, "step": 23576 }, { "epoch": 2.5560494362532524, "grad_norm": 0.02108822576701641, "learning_rate": 2.9704047584506025e-06, "loss": 0.0001, "step": 23577 }, { "epoch": 2.556157849089332, "grad_norm": 0.3147999942302704, "learning_rate": 2.969679384883215e-06, "loss": 0.0219, "step": 23578 }, { "epoch": 2.556266261925412, "grad_norm": 0.3506223261356354, "learning_rate": 2.9689540113158277e-06, "loss": 0.0174, "step": 23579 }, { "epoch": 2.556374674761492, "grad_norm": 0.22821921110153198, "learning_rate": 2.968228637748441e-06, "loss": 0.0042, "step": 23580 }, { "epoch": 2.5564830875975715, "grad_norm": 2.5598514080047607, "learning_rate": 2.9675032641810537e-06, "loss": 0.0301, "step": 23581 }, { "epoch": 2.5565915004336515, "grad_norm": 0.25856608152389526, "learning_rate": 2.9667778906136662e-06, "loss": 0.0055, "step": 23582 }, { "epoch": 2.556699913269731, "grad_norm": 1.3849949836730957, "learning_rate": 2.966052517046279e-06, "loss": 0.0029, "step": 23583 }, { "epoch": 2.556808326105811, "grad_norm": 7.325212001800537, "learning_rate": 2.9653271434788918e-06, "loss": 0.0096, "step": 23584 }, { "epoch": 2.556916738941891, "grad_norm": 2.9022257328033447, "learning_rate": 2.9646017699115048e-06, "loss": 0.0508, "step": 23585 }, { "epoch": 2.5570251517779705, "grad_norm": 0.019593920558691025, "learning_rate": 2.9638763963441173e-06, "loss": 0.0001, "step": 23586 }, { "epoch": 2.55713356461405, "grad_norm": 1.492832064628601, "learning_rate": 2.9631510227767303e-06, "loss": 0.0535, "step": 23587 }, { "epoch": 2.55724197745013, "grad_norm": 0.6474186182022095, "learning_rate": 2.962425649209343e-06, "loss": 0.0138, "step": 23588 }, { "epoch": 2.55735039028621, "grad_norm": 18.60161590576172, "learning_rate": 2.961700275641956e-06, "loss": 0.0382, "step": 23589 }, { "epoch": 2.5574588031222896, "grad_norm": 4.138789176940918, "learning_rate": 2.960974902074569e-06, "loss": 0.0573, "step": 23590 }, { "epoch": 2.5575672159583696, "grad_norm": 0.26727351546287537, "learning_rate": 2.9602495285071814e-06, "loss": 0.0058, "step": 23591 }, { "epoch": 2.557675628794449, "grad_norm": 0.7694294452667236, "learning_rate": 2.959524154939794e-06, "loss": 0.0123, "step": 23592 }, { "epoch": 2.557784041630529, "grad_norm": 0.27623409032821655, "learning_rate": 2.9587987813724074e-06, "loss": 0.0068, "step": 23593 }, { "epoch": 2.557892454466609, "grad_norm": 0.03929838538169861, "learning_rate": 2.95807340780502e-06, "loss": 0.0003, "step": 23594 }, { "epoch": 2.5580008673026886, "grad_norm": 4.902840614318848, "learning_rate": 2.9573480342376326e-06, "loss": 0.0456, "step": 23595 }, { "epoch": 2.5581092801387686, "grad_norm": 0.3767344653606415, "learning_rate": 2.956622660670245e-06, "loss": 0.0081, "step": 23596 }, { "epoch": 2.558217692974848, "grad_norm": 0.06733808666467667, "learning_rate": 2.955897287102858e-06, "loss": 0.001, "step": 23597 }, { "epoch": 2.558326105810928, "grad_norm": 0.031823430210351944, "learning_rate": 2.955171913535471e-06, "loss": 0.0002, "step": 23598 }, { "epoch": 2.5584345186470077, "grad_norm": 0.004969261586666107, "learning_rate": 2.9544465399680837e-06, "loss": 0.0, "step": 23599 }, { "epoch": 2.5585429314830876, "grad_norm": 0.08354882150888443, "learning_rate": 2.9537211664006967e-06, "loss": 0.0102, "step": 23600 }, { "epoch": 2.558651344319167, "grad_norm": 0.24443504214286804, "learning_rate": 2.9529957928333092e-06, "loss": 0.005, "step": 23601 }, { "epoch": 2.558759757155247, "grad_norm": 0.008650699630379677, "learning_rate": 2.9522704192659222e-06, "loss": 0.0001, "step": 23602 }, { "epoch": 2.558868169991327, "grad_norm": 0.5567919015884399, "learning_rate": 2.9515450456985352e-06, "loss": 0.0089, "step": 23603 }, { "epoch": 2.5589765828274067, "grad_norm": 0.017548153176903725, "learning_rate": 2.9508196721311478e-06, "loss": 0.0001, "step": 23604 }, { "epoch": 2.5590849956634867, "grad_norm": 0.4902097284793854, "learning_rate": 2.9500942985637603e-06, "loss": 0.0223, "step": 23605 }, { "epoch": 2.5591934084995662, "grad_norm": 0.008950677700340748, "learning_rate": 2.9493689249963738e-06, "loss": 0.0001, "step": 23606 }, { "epoch": 2.559301821335646, "grad_norm": 0.03861493244767189, "learning_rate": 2.9486435514289863e-06, "loss": 0.0003, "step": 23607 }, { "epoch": 2.559410234171726, "grad_norm": 1.0465178489685059, "learning_rate": 2.947918177861599e-06, "loss": 0.0482, "step": 23608 }, { "epoch": 2.5595186470078057, "grad_norm": 0.3980157673358917, "learning_rate": 2.9471928042942115e-06, "loss": 0.023, "step": 23609 }, { "epoch": 2.5596270598438853, "grad_norm": 0.2550002336502075, "learning_rate": 2.9464674307268245e-06, "loss": 0.0056, "step": 23610 }, { "epoch": 2.5597354726799653, "grad_norm": 0.3262057304382324, "learning_rate": 2.9457420571594374e-06, "loss": 0.0006, "step": 23611 }, { "epoch": 2.5598438855160452, "grad_norm": 0.04804697632789612, "learning_rate": 2.94501668359205e-06, "loss": 0.0003, "step": 23612 }, { "epoch": 2.559952298352125, "grad_norm": 0.6324830055236816, "learning_rate": 2.944291310024663e-06, "loss": 0.0175, "step": 23613 }, { "epoch": 2.5600607111882048, "grad_norm": 2.489008903503418, "learning_rate": 2.9435659364572756e-06, "loss": 0.0116, "step": 23614 }, { "epoch": 2.5601691240242843, "grad_norm": 3.219369649887085, "learning_rate": 2.9428405628898886e-06, "loss": 0.0356, "step": 23615 }, { "epoch": 2.5602775368603643, "grad_norm": 0.09795625507831573, "learning_rate": 2.9421151893225015e-06, "loss": 0.0006, "step": 23616 }, { "epoch": 2.5603859496964443, "grad_norm": 0.4762553572654724, "learning_rate": 2.941389815755114e-06, "loss": 0.0214, "step": 23617 }, { "epoch": 2.560494362532524, "grad_norm": 0.01179470494389534, "learning_rate": 2.9406644421877267e-06, "loss": 0.0001, "step": 23618 }, { "epoch": 2.5606027753686034, "grad_norm": 0.2684260308742523, "learning_rate": 2.93993906862034e-06, "loss": 0.0057, "step": 23619 }, { "epoch": 2.5607111882046834, "grad_norm": 7.848029613494873, "learning_rate": 2.9392136950529527e-06, "loss": 0.0033, "step": 23620 }, { "epoch": 2.5608196010407633, "grad_norm": 0.2892102599143982, "learning_rate": 2.9384883214855652e-06, "loss": 0.0064, "step": 23621 }, { "epoch": 2.560928013876843, "grad_norm": 4.324213981628418, "learning_rate": 2.937762947918178e-06, "loss": 0.0424, "step": 23622 }, { "epoch": 2.561036426712923, "grad_norm": 0.3598194718360901, "learning_rate": 2.937037574350791e-06, "loss": 0.0073, "step": 23623 }, { "epoch": 2.5611448395490024, "grad_norm": 0.06636818498373032, "learning_rate": 2.9363122007834038e-06, "loss": 0.0003, "step": 23624 }, { "epoch": 2.5612532523850824, "grad_norm": 0.43627408146858215, "learning_rate": 2.9355868272160163e-06, "loss": 0.0264, "step": 23625 }, { "epoch": 2.5613616652211624, "grad_norm": 0.2504236698150635, "learning_rate": 2.9348614536486293e-06, "loss": 0.0054, "step": 23626 }, { "epoch": 2.561470078057242, "grad_norm": 0.04648130014538765, "learning_rate": 2.934136080081242e-06, "loss": 0.0003, "step": 23627 }, { "epoch": 2.561578490893322, "grad_norm": 1.1191164255142212, "learning_rate": 2.933410706513855e-06, "loss": 0.0666, "step": 23628 }, { "epoch": 2.5616869037294014, "grad_norm": 2.355530023574829, "learning_rate": 2.932685332946468e-06, "loss": 0.0214, "step": 23629 }, { "epoch": 2.5617953165654814, "grad_norm": 0.0028769108466804028, "learning_rate": 2.9319599593790805e-06, "loss": 0.0, "step": 23630 }, { "epoch": 2.5619037294015614, "grad_norm": 0.3214128017425537, "learning_rate": 2.931234585811693e-06, "loss": 0.0104, "step": 23631 }, { "epoch": 2.562012142237641, "grad_norm": 0.13552813231945038, "learning_rate": 2.9305092122443064e-06, "loss": 0.0008, "step": 23632 }, { "epoch": 2.5621205550737205, "grad_norm": 0.35102787613868713, "learning_rate": 2.929783838676919e-06, "loss": 0.0052, "step": 23633 }, { "epoch": 2.5622289679098005, "grad_norm": 0.31083622574806213, "learning_rate": 2.9290584651095316e-06, "loss": 0.0068, "step": 23634 }, { "epoch": 2.5623373807458805, "grad_norm": 0.34193718433380127, "learning_rate": 2.928333091542144e-06, "loss": 0.0212, "step": 23635 }, { "epoch": 2.56244579358196, "grad_norm": 0.24261116981506348, "learning_rate": 2.9276077179747575e-06, "loss": 0.0051, "step": 23636 }, { "epoch": 2.56255420641804, "grad_norm": 0.004114930983632803, "learning_rate": 2.92688234440737e-06, "loss": 0.0, "step": 23637 }, { "epoch": 2.5626626192541195, "grad_norm": 3.2330880165100098, "learning_rate": 2.9261569708399827e-06, "loss": 0.0116, "step": 23638 }, { "epoch": 2.5627710320901995, "grad_norm": 0.2843109965324402, "learning_rate": 2.9254315972725957e-06, "loss": 0.0066, "step": 23639 }, { "epoch": 2.5628794449262795, "grad_norm": 0.7533779144287109, "learning_rate": 2.9247062237052082e-06, "loss": 0.0301, "step": 23640 }, { "epoch": 2.562987857762359, "grad_norm": 0.07652702182531357, "learning_rate": 2.9239808501378212e-06, "loss": 0.0004, "step": 23641 }, { "epoch": 2.5630962705984386, "grad_norm": 1.0020760297775269, "learning_rate": 2.9232554765704342e-06, "loss": 0.0081, "step": 23642 }, { "epoch": 2.5632046834345186, "grad_norm": 0.0339432917535305, "learning_rate": 2.9225301030030468e-06, "loss": 0.0003, "step": 23643 }, { "epoch": 2.5633130962705986, "grad_norm": 3.703305721282959, "learning_rate": 2.9218047294356594e-06, "loss": 0.0313, "step": 23644 }, { "epoch": 2.563421509106678, "grad_norm": 1.5303698778152466, "learning_rate": 2.9210793558682728e-06, "loss": 0.0413, "step": 23645 }, { "epoch": 2.563529921942758, "grad_norm": 0.03758803382515907, "learning_rate": 2.9203539823008853e-06, "loss": 0.0002, "step": 23646 }, { "epoch": 2.5636383347788376, "grad_norm": 0.5259932279586792, "learning_rate": 2.919628608733498e-06, "loss": 0.0075, "step": 23647 }, { "epoch": 2.5637467476149176, "grad_norm": 0.46826279163360596, "learning_rate": 2.9189032351661105e-06, "loss": 0.0161, "step": 23648 }, { "epoch": 2.5638551604509976, "grad_norm": 0.3210201859474182, "learning_rate": 2.918177861598724e-06, "loss": 0.0107, "step": 23649 }, { "epoch": 2.563963573287077, "grad_norm": 0.3080066740512848, "learning_rate": 2.9174524880313364e-06, "loss": 0.0118, "step": 23650 }, { "epoch": 2.564071986123157, "grad_norm": 0.013368749059736729, "learning_rate": 2.916727114463949e-06, "loss": 0.0002, "step": 23651 }, { "epoch": 2.5641803989592367, "grad_norm": 0.03149671480059624, "learning_rate": 2.916001740896562e-06, "loss": 0.0002, "step": 23652 }, { "epoch": 2.5642888117953166, "grad_norm": 3.041177749633789, "learning_rate": 2.9152763673291746e-06, "loss": 0.0115, "step": 23653 }, { "epoch": 2.5643972246313966, "grad_norm": 0.008694582618772984, "learning_rate": 2.9145509937617876e-06, "loss": 0.0001, "step": 23654 }, { "epoch": 2.564505637467476, "grad_norm": 0.1546722799539566, "learning_rate": 2.9138256201944006e-06, "loss": 0.0003, "step": 23655 }, { "epoch": 2.5646140503035557, "grad_norm": 0.03355282545089722, "learning_rate": 2.913100246627013e-06, "loss": 0.0003, "step": 23656 }, { "epoch": 2.5647224631396357, "grad_norm": 0.018019549548625946, "learning_rate": 2.9123748730596257e-06, "loss": 0.0002, "step": 23657 }, { "epoch": 2.5648308759757157, "grad_norm": 1.8840563297271729, "learning_rate": 2.911649499492239e-06, "loss": 0.0022, "step": 23658 }, { "epoch": 2.5649392888117952, "grad_norm": 0.6488246321678162, "learning_rate": 2.9109241259248517e-06, "loss": 0.0014, "step": 23659 }, { "epoch": 2.565047701647875, "grad_norm": 0.015595304779708385, "learning_rate": 2.9101987523574642e-06, "loss": 0.0002, "step": 23660 }, { "epoch": 2.5651561144839548, "grad_norm": 0.344926655292511, "learning_rate": 2.909473378790077e-06, "loss": 0.0172, "step": 23661 }, { "epoch": 2.5652645273200347, "grad_norm": 0.40810808539390564, "learning_rate": 2.9087480052226902e-06, "loss": 0.0016, "step": 23662 }, { "epoch": 2.5653729401561147, "grad_norm": 4.346673965454102, "learning_rate": 2.9080226316553028e-06, "loss": 0.0296, "step": 23663 }, { "epoch": 2.5654813529921943, "grad_norm": 0.04564981907606125, "learning_rate": 2.9072972580879153e-06, "loss": 0.0003, "step": 23664 }, { "epoch": 2.565589765828274, "grad_norm": 0.029445840045809746, "learning_rate": 2.9065718845205283e-06, "loss": 0.0002, "step": 23665 }, { "epoch": 2.565698178664354, "grad_norm": 8.866666793823242, "learning_rate": 2.905846510953141e-06, "loss": 0.0164, "step": 23666 }, { "epoch": 2.5658065915004338, "grad_norm": 0.006995337083935738, "learning_rate": 2.905121137385754e-06, "loss": 0.0001, "step": 23667 }, { "epoch": 2.5659150043365133, "grad_norm": 0.23113927245140076, "learning_rate": 2.904395763818367e-06, "loss": 0.0015, "step": 23668 }, { "epoch": 2.5660234171725933, "grad_norm": 2.23677921295166, "learning_rate": 2.9036703902509795e-06, "loss": 0.0428, "step": 23669 }, { "epoch": 2.566131830008673, "grad_norm": 0.2980794906616211, "learning_rate": 2.902945016683592e-06, "loss": 0.0065, "step": 23670 }, { "epoch": 2.566240242844753, "grad_norm": 0.7388077974319458, "learning_rate": 2.9022196431162054e-06, "loss": 0.0104, "step": 23671 }, { "epoch": 2.566348655680833, "grad_norm": 0.6754819750785828, "learning_rate": 2.901494269548818e-06, "loss": 0.0081, "step": 23672 }, { "epoch": 2.5664570685169124, "grad_norm": 0.013281757943332195, "learning_rate": 2.9007688959814306e-06, "loss": 0.0002, "step": 23673 }, { "epoch": 2.5665654813529923, "grad_norm": 0.003783261636272073, "learning_rate": 2.900043522414043e-06, "loss": 0.0001, "step": 23674 }, { "epoch": 2.566673894189072, "grad_norm": 0.032974354922771454, "learning_rate": 2.8993181488466565e-06, "loss": 0.0003, "step": 23675 }, { "epoch": 2.566782307025152, "grad_norm": 0.6785123944282532, "learning_rate": 2.898592775279269e-06, "loss": 0.0374, "step": 23676 }, { "epoch": 2.5668907198612314, "grad_norm": 1.7136863470077515, "learning_rate": 2.8978674017118817e-06, "loss": 0.0079, "step": 23677 }, { "epoch": 2.5669991326973114, "grad_norm": 3.6030383110046387, "learning_rate": 2.8971420281444947e-06, "loss": 0.049, "step": 23678 }, { "epoch": 2.567107545533391, "grad_norm": 0.7467876672744751, "learning_rate": 2.8964166545771072e-06, "loss": 0.0064, "step": 23679 }, { "epoch": 2.567215958369471, "grad_norm": 0.29814860224723816, "learning_rate": 2.8956912810097202e-06, "loss": 0.0004, "step": 23680 }, { "epoch": 2.567324371205551, "grad_norm": 2.0785343647003174, "learning_rate": 2.8949659074423332e-06, "loss": 0.0036, "step": 23681 }, { "epoch": 2.5674327840416304, "grad_norm": 0.2618868350982666, "learning_rate": 2.894240533874946e-06, "loss": 0.0044, "step": 23682 }, { "epoch": 2.5675411968777104, "grad_norm": 0.40364858508110046, "learning_rate": 2.8935151603075584e-06, "loss": 0.0159, "step": 23683 }, { "epoch": 2.56764960971379, "grad_norm": 0.2862148880958557, "learning_rate": 2.8927897867401718e-06, "loss": 0.0045, "step": 23684 }, { "epoch": 2.56775802254987, "grad_norm": 3.0144741535186768, "learning_rate": 2.8920644131727843e-06, "loss": 0.0379, "step": 23685 }, { "epoch": 2.56786643538595, "grad_norm": 0.26133254170417786, "learning_rate": 2.891339039605397e-06, "loss": 0.0055, "step": 23686 }, { "epoch": 2.5679748482220295, "grad_norm": 0.03158629685640335, "learning_rate": 2.8906136660380095e-06, "loss": 0.0002, "step": 23687 }, { "epoch": 2.568083261058109, "grad_norm": 0.40722742676734924, "learning_rate": 2.889888292470623e-06, "loss": 0.0179, "step": 23688 }, { "epoch": 2.568191673894189, "grad_norm": 0.3340989053249359, "learning_rate": 2.8891629189032355e-06, "loss": 0.0069, "step": 23689 }, { "epoch": 2.568300086730269, "grad_norm": 0.3100265562534332, "learning_rate": 2.888437545335848e-06, "loss": 0.0096, "step": 23690 }, { "epoch": 2.5684084995663485, "grad_norm": 0.6294353604316711, "learning_rate": 2.887712171768461e-06, "loss": 0.0222, "step": 23691 }, { "epoch": 2.5685169124024285, "grad_norm": 4.863686561584473, "learning_rate": 2.8869867982010736e-06, "loss": 0.0354, "step": 23692 }, { "epoch": 2.568625325238508, "grad_norm": 0.6715256571769714, "learning_rate": 2.8862614246336866e-06, "loss": 0.0167, "step": 23693 }, { "epoch": 2.568733738074588, "grad_norm": 1.3025795221328735, "learning_rate": 2.8855360510662996e-06, "loss": 0.0438, "step": 23694 }, { "epoch": 2.568842150910668, "grad_norm": 0.42260244488716125, "learning_rate": 2.884810677498912e-06, "loss": 0.0013, "step": 23695 }, { "epoch": 2.5689505637467476, "grad_norm": 0.03535572066903114, "learning_rate": 2.8840853039315247e-06, "loss": 0.0002, "step": 23696 }, { "epoch": 2.5690589765828276, "grad_norm": 0.3654862642288208, "learning_rate": 2.883359930364138e-06, "loss": 0.0054, "step": 23697 }, { "epoch": 2.569167389418907, "grad_norm": 0.5049494504928589, "learning_rate": 2.8826345567967507e-06, "loss": 0.0024, "step": 23698 }, { "epoch": 2.569275802254987, "grad_norm": 0.1347772479057312, "learning_rate": 2.8819091832293632e-06, "loss": 0.0007, "step": 23699 }, { "epoch": 2.5693842150910666, "grad_norm": 2.3274178504943848, "learning_rate": 2.881183809661976e-06, "loss": 0.0331, "step": 23700 }, { "epoch": 2.5694926279271466, "grad_norm": 0.20286305248737335, "learning_rate": 2.8804584360945892e-06, "loss": 0.0007, "step": 23701 }, { "epoch": 2.569601040763226, "grad_norm": 0.37192168831825256, "learning_rate": 2.8797330625272018e-06, "loss": 0.0111, "step": 23702 }, { "epoch": 2.569709453599306, "grad_norm": 0.010725626721978188, "learning_rate": 2.8790076889598144e-06, "loss": 0.0001, "step": 23703 }, { "epoch": 2.569817866435386, "grad_norm": 0.33754798769950867, "learning_rate": 2.8782823153924273e-06, "loss": 0.0094, "step": 23704 }, { "epoch": 2.5699262792714657, "grad_norm": 6.321171283721924, "learning_rate": 2.87755694182504e-06, "loss": 0.0442, "step": 23705 }, { "epoch": 2.5700346921075456, "grad_norm": 0.003921720199286938, "learning_rate": 2.876831568257653e-06, "loss": 0.0, "step": 23706 }, { "epoch": 2.570143104943625, "grad_norm": 9.495054244995117, "learning_rate": 2.876106194690266e-06, "loss": 0.0077, "step": 23707 }, { "epoch": 2.570251517779705, "grad_norm": 0.4224109649658203, "learning_rate": 2.8753808211228785e-06, "loss": 0.0131, "step": 23708 }, { "epoch": 2.570359930615785, "grad_norm": 0.287018746137619, "learning_rate": 2.874655447555491e-06, "loss": 0.0004, "step": 23709 }, { "epoch": 2.5704683434518647, "grad_norm": 5.875545024871826, "learning_rate": 2.8739300739881044e-06, "loss": 0.0116, "step": 23710 }, { "epoch": 2.5705767562879442, "grad_norm": 0.47362738847732544, "learning_rate": 2.873204700420717e-06, "loss": 0.0258, "step": 23711 }, { "epoch": 2.5706851691240242, "grad_norm": 5.98503303527832, "learning_rate": 2.8724793268533296e-06, "loss": 0.0187, "step": 23712 }, { "epoch": 2.570793581960104, "grad_norm": 2.2947824001312256, "learning_rate": 2.871753953285942e-06, "loss": 0.0357, "step": 23713 }, { "epoch": 2.5709019947961838, "grad_norm": 0.3922789394855499, "learning_rate": 2.8710285797185556e-06, "loss": 0.0174, "step": 23714 }, { "epoch": 2.5710104076322637, "grad_norm": 0.7645882964134216, "learning_rate": 2.870303206151168e-06, "loss": 0.0282, "step": 23715 }, { "epoch": 2.5711188204683433, "grad_norm": 0.023494955152273178, "learning_rate": 2.8695778325837807e-06, "loss": 0.0002, "step": 23716 }, { "epoch": 2.5712272333044233, "grad_norm": 3.8259077072143555, "learning_rate": 2.8688524590163937e-06, "loss": 0.0463, "step": 23717 }, { "epoch": 2.5713356461405033, "grad_norm": 0.012407478876411915, "learning_rate": 2.8681270854490062e-06, "loss": 0.0001, "step": 23718 }, { "epoch": 2.571444058976583, "grad_norm": 4.1143107414245605, "learning_rate": 2.8674017118816192e-06, "loss": 0.0525, "step": 23719 }, { "epoch": 2.5715524718126628, "grad_norm": 0.14869354665279388, "learning_rate": 2.8666763383142322e-06, "loss": 0.0007, "step": 23720 }, { "epoch": 2.5716608846487423, "grad_norm": 1.8181942701339722, "learning_rate": 2.865950964746845e-06, "loss": 0.0149, "step": 23721 }, { "epoch": 2.5717692974848223, "grad_norm": 0.29126718640327454, "learning_rate": 2.8652255911794574e-06, "loss": 0.0065, "step": 23722 }, { "epoch": 2.571877710320902, "grad_norm": 0.3705812692642212, "learning_rate": 2.8645002176120708e-06, "loss": 0.0122, "step": 23723 }, { "epoch": 2.571986123156982, "grad_norm": 0.36642026901245117, "learning_rate": 2.8637748440446833e-06, "loss": 0.001, "step": 23724 }, { "epoch": 2.5720945359930614, "grad_norm": 10.153177261352539, "learning_rate": 2.863049470477296e-06, "loss": 0.0344, "step": 23725 }, { "epoch": 2.5722029488291414, "grad_norm": 1.8494994640350342, "learning_rate": 2.8623240969099085e-06, "loss": 0.0279, "step": 23726 }, { "epoch": 2.5723113616652213, "grad_norm": 1.1190332174301147, "learning_rate": 2.861598723342522e-06, "loss": 0.0017, "step": 23727 }, { "epoch": 2.572419774501301, "grad_norm": 3.2528350353240967, "learning_rate": 2.8608733497751345e-06, "loss": 0.0099, "step": 23728 }, { "epoch": 2.572528187337381, "grad_norm": 1.1211490631103516, "learning_rate": 2.860147976207747e-06, "loss": 0.0616, "step": 23729 }, { "epoch": 2.5726366001734604, "grad_norm": 0.04520006850361824, "learning_rate": 2.85942260264036e-06, "loss": 0.0004, "step": 23730 }, { "epoch": 2.5727450130095404, "grad_norm": 0.02460782788693905, "learning_rate": 2.858697229072973e-06, "loss": 0.0002, "step": 23731 }, { "epoch": 2.5728534258456204, "grad_norm": 0.25713932514190674, "learning_rate": 2.8579718555055856e-06, "loss": 0.0052, "step": 23732 }, { "epoch": 2.5729618386817, "grad_norm": 5.565371513366699, "learning_rate": 2.8572464819381986e-06, "loss": 0.057, "step": 23733 }, { "epoch": 2.5730702515177795, "grad_norm": 0.7401449680328369, "learning_rate": 2.856521108370811e-06, "loss": 0.017, "step": 23734 }, { "epoch": 2.5731786643538594, "grad_norm": 0.026918167248368263, "learning_rate": 2.8557957348034237e-06, "loss": 0.0002, "step": 23735 }, { "epoch": 2.5732870771899394, "grad_norm": 0.5837629437446594, "learning_rate": 2.855070361236037e-06, "loss": 0.0181, "step": 23736 }, { "epoch": 2.573395490026019, "grad_norm": 0.07406444847583771, "learning_rate": 2.8543449876686497e-06, "loss": 0.0006, "step": 23737 }, { "epoch": 2.573503902862099, "grad_norm": 0.18249092996120453, "learning_rate": 2.8536196141012622e-06, "loss": 0.0003, "step": 23738 }, { "epoch": 2.5736123156981785, "grad_norm": 4.009397029876709, "learning_rate": 2.852894240533875e-06, "loss": 0.0297, "step": 23739 }, { "epoch": 2.5737207285342585, "grad_norm": 0.7001670598983765, "learning_rate": 2.8521688669664882e-06, "loss": 0.0077, "step": 23740 }, { "epoch": 2.5738291413703385, "grad_norm": 0.5099695920944214, "learning_rate": 2.851443493399101e-06, "loss": 0.0128, "step": 23741 }, { "epoch": 2.573937554206418, "grad_norm": 0.3285531997680664, "learning_rate": 2.8507181198317134e-06, "loss": 0.0117, "step": 23742 }, { "epoch": 2.574045967042498, "grad_norm": 1.5859856605529785, "learning_rate": 2.8499927462643263e-06, "loss": 0.0252, "step": 23743 }, { "epoch": 2.5741543798785775, "grad_norm": 2.7493672370910645, "learning_rate": 2.8492673726969393e-06, "loss": 0.0619, "step": 23744 }, { "epoch": 2.5742627927146575, "grad_norm": 4.613476753234863, "learning_rate": 2.848541999129552e-06, "loss": 0.0138, "step": 23745 }, { "epoch": 2.574371205550737, "grad_norm": 2.3810203075408936, "learning_rate": 2.847816625562165e-06, "loss": 0.0056, "step": 23746 }, { "epoch": 2.574479618386817, "grad_norm": 2.742396354675293, "learning_rate": 2.8470912519947775e-06, "loss": 0.0195, "step": 23747 }, { "epoch": 2.5745880312228966, "grad_norm": 0.3580067455768585, "learning_rate": 2.84636587842739e-06, "loss": 0.0091, "step": 23748 }, { "epoch": 2.5746964440589766, "grad_norm": 0.073177769780159, "learning_rate": 2.8456405048600034e-06, "loss": 0.0007, "step": 23749 }, { "epoch": 2.5748048568950566, "grad_norm": 0.04327726736664772, "learning_rate": 2.844915131292616e-06, "loss": 0.0004, "step": 23750 }, { "epoch": 2.574913269731136, "grad_norm": 0.08467087894678116, "learning_rate": 2.8441897577252286e-06, "loss": 0.0004, "step": 23751 }, { "epoch": 2.575021682567216, "grad_norm": 0.8153766989707947, "learning_rate": 2.843464384157841e-06, "loss": 0.0013, "step": 23752 }, { "epoch": 2.5751300954032956, "grad_norm": 17.66324234008789, "learning_rate": 2.8427390105904546e-06, "loss": 0.0399, "step": 23753 }, { "epoch": 2.5752385082393756, "grad_norm": 0.2948354482650757, "learning_rate": 2.842013637023067e-06, "loss": 0.0105, "step": 23754 }, { "epoch": 2.5753469210754556, "grad_norm": 0.2492978572845459, "learning_rate": 2.8412882634556797e-06, "loss": 0.0038, "step": 23755 }, { "epoch": 2.575455333911535, "grad_norm": 0.3628610074520111, "learning_rate": 2.8405628898882927e-06, "loss": 0.0235, "step": 23756 }, { "epoch": 2.5755637467476147, "grad_norm": 0.4981310963630676, "learning_rate": 2.8398375163209057e-06, "loss": 0.0099, "step": 23757 }, { "epoch": 2.5756721595836947, "grad_norm": 2.509946823120117, "learning_rate": 2.8391121427535182e-06, "loss": 0.0077, "step": 23758 }, { "epoch": 2.5757805724197746, "grad_norm": 0.4277655780315399, "learning_rate": 2.8383867691861312e-06, "loss": 0.0225, "step": 23759 }, { "epoch": 2.575888985255854, "grad_norm": 5.535107135772705, "learning_rate": 2.837661395618744e-06, "loss": 0.0156, "step": 23760 }, { "epoch": 2.575997398091934, "grad_norm": 0.28727659583091736, "learning_rate": 2.8369360220513564e-06, "loss": 0.0056, "step": 23761 }, { "epoch": 2.5761058109280137, "grad_norm": 0.4273018538951874, "learning_rate": 2.8362106484839698e-06, "loss": 0.008, "step": 23762 }, { "epoch": 2.5762142237640937, "grad_norm": 0.3095416724681854, "learning_rate": 2.8354852749165823e-06, "loss": 0.0098, "step": 23763 }, { "epoch": 2.5763226366001737, "grad_norm": 3.4676249027252197, "learning_rate": 2.834759901349195e-06, "loss": 0.0098, "step": 23764 }, { "epoch": 2.5764310494362532, "grad_norm": 0.580783486366272, "learning_rate": 2.8340345277818075e-06, "loss": 0.0187, "step": 23765 }, { "epoch": 2.5765394622723328, "grad_norm": 0.25060707330703735, "learning_rate": 2.833309154214421e-06, "loss": 0.0039, "step": 23766 }, { "epoch": 2.5766478751084128, "grad_norm": 0.2581164538860321, "learning_rate": 2.8325837806470335e-06, "loss": 0.003, "step": 23767 }, { "epoch": 2.5767562879444927, "grad_norm": 0.5848572850227356, "learning_rate": 2.831858407079646e-06, "loss": 0.0199, "step": 23768 }, { "epoch": 2.5768647007805723, "grad_norm": 0.06644902378320694, "learning_rate": 2.831133033512259e-06, "loss": 0.0004, "step": 23769 }, { "epoch": 2.5769731136166523, "grad_norm": 0.04657191038131714, "learning_rate": 2.830407659944872e-06, "loss": 0.0005, "step": 23770 }, { "epoch": 2.577081526452732, "grad_norm": 0.3851189911365509, "learning_rate": 2.8296822863774846e-06, "loss": 0.0094, "step": 23771 }, { "epoch": 2.577189939288812, "grad_norm": 1.025667667388916, "learning_rate": 2.8289569128100976e-06, "loss": 0.0472, "step": 23772 }, { "epoch": 2.577298352124892, "grad_norm": 0.36172258853912354, "learning_rate": 2.82823153924271e-06, "loss": 0.0013, "step": 23773 }, { "epoch": 2.5774067649609713, "grad_norm": 1.8883830308914185, "learning_rate": 2.8275061656753227e-06, "loss": 0.0152, "step": 23774 }, { "epoch": 2.5775151777970513, "grad_norm": 0.04360644146800041, "learning_rate": 2.826780792107936e-06, "loss": 0.0002, "step": 23775 }, { "epoch": 2.577623590633131, "grad_norm": 1.600974678993225, "learning_rate": 2.8260554185405487e-06, "loss": 0.0391, "step": 23776 }, { "epoch": 2.577732003469211, "grad_norm": 0.5081615447998047, "learning_rate": 2.8253300449731612e-06, "loss": 0.0128, "step": 23777 }, { "epoch": 2.577840416305291, "grad_norm": 1.5919190645217896, "learning_rate": 2.824604671405774e-06, "loss": 0.0017, "step": 23778 }, { "epoch": 2.5779488291413704, "grad_norm": 0.6135822534561157, "learning_rate": 2.8238792978383872e-06, "loss": 0.0133, "step": 23779 }, { "epoch": 2.57805724197745, "grad_norm": 44.239288330078125, "learning_rate": 2.823153924271e-06, "loss": 0.0309, "step": 23780 }, { "epoch": 2.57816565481353, "grad_norm": 0.3796609342098236, "learning_rate": 2.8224285507036124e-06, "loss": 0.0111, "step": 23781 }, { "epoch": 2.57827406764961, "grad_norm": 0.5952367186546326, "learning_rate": 2.8217031771362253e-06, "loss": 0.0133, "step": 23782 }, { "epoch": 2.5783824804856894, "grad_norm": 3.1794965267181396, "learning_rate": 2.8209778035688383e-06, "loss": 0.0261, "step": 23783 }, { "epoch": 2.5784908933217694, "grad_norm": 0.4469555616378784, "learning_rate": 2.820252430001451e-06, "loss": 0.0071, "step": 23784 }, { "epoch": 2.578599306157849, "grad_norm": 8.583294868469238, "learning_rate": 2.819527056434064e-06, "loss": 0.017, "step": 23785 }, { "epoch": 2.578707718993929, "grad_norm": 5.345860481262207, "learning_rate": 2.8188016828666765e-06, "loss": 0.039, "step": 23786 }, { "epoch": 2.578816131830009, "grad_norm": 0.5515323281288147, "learning_rate": 2.818076309299289e-06, "loss": 0.0128, "step": 23787 }, { "epoch": 2.5789245446660884, "grad_norm": 11.893803596496582, "learning_rate": 2.8173509357319024e-06, "loss": 0.0386, "step": 23788 }, { "epoch": 2.579032957502168, "grad_norm": 0.02507000043988228, "learning_rate": 2.816625562164515e-06, "loss": 0.0002, "step": 23789 }, { "epoch": 2.579141370338248, "grad_norm": 0.06833715736865997, "learning_rate": 2.8159001885971276e-06, "loss": 0.0002, "step": 23790 }, { "epoch": 2.579249783174328, "grad_norm": 2.6138687133789062, "learning_rate": 2.81517481502974e-06, "loss": 0.0283, "step": 23791 }, { "epoch": 2.5793581960104075, "grad_norm": 0.20447272062301636, "learning_rate": 2.8144494414623536e-06, "loss": 0.0039, "step": 23792 }, { "epoch": 2.5794666088464875, "grad_norm": 0.8710947632789612, "learning_rate": 2.813724067894966e-06, "loss": 0.045, "step": 23793 }, { "epoch": 2.579575021682567, "grad_norm": 4.226682662963867, "learning_rate": 2.8129986943275787e-06, "loss": 0.0375, "step": 23794 }, { "epoch": 2.579683434518647, "grad_norm": 0.48534467816352844, "learning_rate": 2.8122733207601917e-06, "loss": 0.0103, "step": 23795 }, { "epoch": 2.579791847354727, "grad_norm": 0.028925329446792603, "learning_rate": 2.8115479471928047e-06, "loss": 0.0002, "step": 23796 }, { "epoch": 2.5799002601908065, "grad_norm": 0.597315788269043, "learning_rate": 2.8108225736254172e-06, "loss": 0.0064, "step": 23797 }, { "epoch": 2.5800086730268865, "grad_norm": 1.8582701683044434, "learning_rate": 2.8100972000580302e-06, "loss": 0.0019, "step": 23798 }, { "epoch": 2.580117085862966, "grad_norm": 0.723312497138977, "learning_rate": 2.809371826490643e-06, "loss": 0.0123, "step": 23799 }, { "epoch": 2.580225498699046, "grad_norm": 0.21591247618198395, "learning_rate": 2.8086464529232554e-06, "loss": 0.0009, "step": 23800 }, { "epoch": 2.580333911535126, "grad_norm": 0.002939252881333232, "learning_rate": 2.8079210793558688e-06, "loss": 0.0, "step": 23801 }, { "epoch": 2.5804423243712056, "grad_norm": 1.7804173231124878, "learning_rate": 2.8071957057884813e-06, "loss": 0.0213, "step": 23802 }, { "epoch": 2.580550737207285, "grad_norm": 2.1773946285247803, "learning_rate": 2.806470332221094e-06, "loss": 0.0106, "step": 23803 }, { "epoch": 2.580659150043365, "grad_norm": 0.006108584813773632, "learning_rate": 2.8057449586537065e-06, "loss": 0.0001, "step": 23804 }, { "epoch": 2.580767562879445, "grad_norm": 0.055951885879039764, "learning_rate": 2.80501958508632e-06, "loss": 0.0003, "step": 23805 }, { "epoch": 2.5808759757155246, "grad_norm": 0.35125696659088135, "learning_rate": 2.8042942115189325e-06, "loss": 0.0088, "step": 23806 }, { "epoch": 2.5809843885516046, "grad_norm": 1.9195204973220825, "learning_rate": 2.803568837951545e-06, "loss": 0.0626, "step": 23807 }, { "epoch": 2.581092801387684, "grad_norm": 0.3267524540424347, "learning_rate": 2.802843464384158e-06, "loss": 0.0084, "step": 23808 }, { "epoch": 2.581201214223764, "grad_norm": 0.02749807760119438, "learning_rate": 2.802118090816771e-06, "loss": 0.0003, "step": 23809 }, { "epoch": 2.581309627059844, "grad_norm": 0.021288927644491196, "learning_rate": 2.8013927172493836e-06, "loss": 0.0002, "step": 23810 }, { "epoch": 2.5814180398959237, "grad_norm": 0.8041740655899048, "learning_rate": 2.8006673436819966e-06, "loss": 0.0095, "step": 23811 }, { "epoch": 2.581526452732003, "grad_norm": 0.44937753677368164, "learning_rate": 2.799941970114609e-06, "loss": 0.0058, "step": 23812 }, { "epoch": 2.581634865568083, "grad_norm": 1.9240086078643799, "learning_rate": 2.7992165965472217e-06, "loss": 0.0281, "step": 23813 }, { "epoch": 2.581743278404163, "grad_norm": 0.21760496497154236, "learning_rate": 2.798491222979835e-06, "loss": 0.0008, "step": 23814 }, { "epoch": 2.5818516912402427, "grad_norm": 3.3168275356292725, "learning_rate": 2.7977658494124477e-06, "loss": 0.0054, "step": 23815 }, { "epoch": 2.5819601040763227, "grad_norm": 0.4436746835708618, "learning_rate": 2.7970404758450602e-06, "loss": 0.0085, "step": 23816 }, { "epoch": 2.5820685169124022, "grad_norm": 0.4267423450946808, "learning_rate": 2.796315102277673e-06, "loss": 0.0126, "step": 23817 }, { "epoch": 2.5821769297484822, "grad_norm": 0.34538018703460693, "learning_rate": 2.7955897287102862e-06, "loss": 0.0135, "step": 23818 }, { "epoch": 2.582285342584562, "grad_norm": 0.5284640192985535, "learning_rate": 2.794864355142899e-06, "loss": 0.0024, "step": 23819 }, { "epoch": 2.5823937554206418, "grad_norm": 1.1731404066085815, "learning_rate": 2.7941389815755114e-06, "loss": 0.0117, "step": 23820 }, { "epoch": 2.5825021682567217, "grad_norm": 0.031633827835321426, "learning_rate": 2.7934136080081244e-06, "loss": 0.0002, "step": 23821 }, { "epoch": 2.5826105810928013, "grad_norm": 2.408799886703491, "learning_rate": 2.7926882344407373e-06, "loss": 0.0172, "step": 23822 }, { "epoch": 2.5827189939288813, "grad_norm": 0.8493029475212097, "learning_rate": 2.79196286087335e-06, "loss": 0.0139, "step": 23823 }, { "epoch": 2.582827406764961, "grad_norm": 1.6793285608291626, "learning_rate": 2.791237487305963e-06, "loss": 0.0043, "step": 23824 }, { "epoch": 2.582935819601041, "grad_norm": 0.3721349835395813, "learning_rate": 2.7905121137385755e-06, "loss": 0.0011, "step": 23825 }, { "epoch": 2.5830442324371203, "grad_norm": 0.9878978729248047, "learning_rate": 2.7897867401711885e-06, "loss": 0.0417, "step": 23826 }, { "epoch": 2.5831526452732003, "grad_norm": 3.0972230434417725, "learning_rate": 2.7890613666038014e-06, "loss": 0.075, "step": 23827 }, { "epoch": 2.5832610581092803, "grad_norm": 0.030744630843400955, "learning_rate": 2.788335993036414e-06, "loss": 0.0003, "step": 23828 }, { "epoch": 2.58336947094536, "grad_norm": 0.020911576226353645, "learning_rate": 2.7876106194690266e-06, "loss": 0.0003, "step": 23829 }, { "epoch": 2.58347788378144, "grad_norm": 4.906507968902588, "learning_rate": 2.786885245901639e-06, "loss": 0.0439, "step": 23830 }, { "epoch": 2.5835862966175194, "grad_norm": 0.7768819332122803, "learning_rate": 2.7861598723342526e-06, "loss": 0.0015, "step": 23831 }, { "epoch": 2.5836947094535994, "grad_norm": 2.4724133014678955, "learning_rate": 2.785434498766865e-06, "loss": 0.0072, "step": 23832 }, { "epoch": 2.5838031222896793, "grad_norm": 1.6289963722229004, "learning_rate": 2.7847091251994777e-06, "loss": 0.0044, "step": 23833 }, { "epoch": 2.583911535125759, "grad_norm": 0.2918621003627777, "learning_rate": 2.7839837516320907e-06, "loss": 0.0065, "step": 23834 }, { "epoch": 2.5840199479618384, "grad_norm": 0.07342614978551865, "learning_rate": 2.7832583780647037e-06, "loss": 0.0106, "step": 23835 }, { "epoch": 2.5841283607979184, "grad_norm": 0.3847367763519287, "learning_rate": 2.7825330044973162e-06, "loss": 0.0103, "step": 23836 }, { "epoch": 2.5842367736339984, "grad_norm": 13.798686027526855, "learning_rate": 2.7818076309299292e-06, "loss": 0.0105, "step": 23837 }, { "epoch": 2.584345186470078, "grad_norm": 0.3120936453342438, "learning_rate": 2.781082257362542e-06, "loss": 0.0067, "step": 23838 }, { "epoch": 2.584453599306158, "grad_norm": 0.5235326290130615, "learning_rate": 2.7803568837951552e-06, "loss": 0.0156, "step": 23839 }, { "epoch": 2.5845620121422375, "grad_norm": 0.07380319386720657, "learning_rate": 2.7796315102277678e-06, "loss": 0.0004, "step": 23840 }, { "epoch": 2.5846704249783174, "grad_norm": 0.4854026436805725, "learning_rate": 2.7789061366603803e-06, "loss": 0.01, "step": 23841 }, { "epoch": 2.5847788378143974, "grad_norm": 0.04154573008418083, "learning_rate": 2.778180763092993e-06, "loss": 0.0004, "step": 23842 }, { "epoch": 2.584887250650477, "grad_norm": 0.2610097825527191, "learning_rate": 2.7774553895256055e-06, "loss": 0.0043, "step": 23843 }, { "epoch": 2.584995663486557, "grad_norm": 0.37764817476272583, "learning_rate": 2.776730015958219e-06, "loss": 0.0079, "step": 23844 }, { "epoch": 2.5851040763226365, "grad_norm": 6.443953514099121, "learning_rate": 2.7760046423908315e-06, "loss": 0.0471, "step": 23845 }, { "epoch": 2.5852124891587165, "grad_norm": 0.051922716200351715, "learning_rate": 2.775279268823444e-06, "loss": 0.0004, "step": 23846 }, { "epoch": 2.585320901994796, "grad_norm": 2.0029592514038086, "learning_rate": 2.774553895256057e-06, "loss": 0.0179, "step": 23847 }, { "epoch": 2.585429314830876, "grad_norm": 1.2376004457473755, "learning_rate": 2.77382852168867e-06, "loss": 0.0039, "step": 23848 }, { "epoch": 2.5855377276669556, "grad_norm": 1.9565743207931519, "learning_rate": 2.7731031481212826e-06, "loss": 0.0036, "step": 23849 }, { "epoch": 2.5856461405030355, "grad_norm": 0.23265734314918518, "learning_rate": 2.7723777745538956e-06, "loss": 0.0046, "step": 23850 }, { "epoch": 2.5857545533391155, "grad_norm": 0.029600419104099274, "learning_rate": 2.771652400986508e-06, "loss": 0.0003, "step": 23851 }, { "epoch": 2.585862966175195, "grad_norm": 0.40478846430778503, "learning_rate": 2.7709270274191215e-06, "loss": 0.0066, "step": 23852 }, { "epoch": 2.585971379011275, "grad_norm": 0.03016933612525463, "learning_rate": 2.770201653851734e-06, "loss": 0.0004, "step": 23853 }, { "epoch": 2.5860797918473546, "grad_norm": 3.7831192016601562, "learning_rate": 2.7694762802843467e-06, "loss": 0.0089, "step": 23854 }, { "epoch": 2.5861882046834346, "grad_norm": 0.3220193684101105, "learning_rate": 2.7687509067169593e-06, "loss": 0.0123, "step": 23855 }, { "epoch": 2.5862966175195146, "grad_norm": 0.2840648591518402, "learning_rate": 2.768025533149572e-06, "loss": 0.0009, "step": 23856 }, { "epoch": 2.586405030355594, "grad_norm": 1.501720905303955, "learning_rate": 2.7673001595821852e-06, "loss": 0.0107, "step": 23857 }, { "epoch": 2.5865134431916736, "grad_norm": 0.1417866051197052, "learning_rate": 2.766574786014798e-06, "loss": 0.0006, "step": 23858 }, { "epoch": 2.5866218560277536, "grad_norm": 0.9423249363899231, "learning_rate": 2.7658494124474104e-06, "loss": 0.0162, "step": 23859 }, { "epoch": 2.5867302688638336, "grad_norm": 0.01660919561982155, "learning_rate": 2.7651240388800234e-06, "loss": 0.0002, "step": 23860 }, { "epoch": 2.586838681699913, "grad_norm": 0.02329528145492077, "learning_rate": 2.7643986653126363e-06, "loss": 0.0002, "step": 23861 }, { "epoch": 2.586947094535993, "grad_norm": 2.6740145683288574, "learning_rate": 2.763673291745249e-06, "loss": 0.0486, "step": 23862 }, { "epoch": 2.5870555073720727, "grad_norm": 4.414543628692627, "learning_rate": 2.762947918177862e-06, "loss": 0.0665, "step": 23863 }, { "epoch": 2.5871639202081527, "grad_norm": 0.25114187598228455, "learning_rate": 2.7622225446104745e-06, "loss": 0.0013, "step": 23864 }, { "epoch": 2.5872723330442327, "grad_norm": 0.15786129236221313, "learning_rate": 2.761497171043088e-06, "loss": 0.0022, "step": 23865 }, { "epoch": 2.587380745880312, "grad_norm": 0.8326967358589172, "learning_rate": 2.7607717974757005e-06, "loss": 0.0104, "step": 23866 }, { "epoch": 2.587489158716392, "grad_norm": 0.3552645146846771, "learning_rate": 2.760046423908313e-06, "loss": 0.0048, "step": 23867 }, { "epoch": 2.5875975715524717, "grad_norm": 0.4855128526687622, "learning_rate": 2.7593210503409256e-06, "loss": 0.0103, "step": 23868 }, { "epoch": 2.5877059843885517, "grad_norm": 0.2924824059009552, "learning_rate": 2.758595676773538e-06, "loss": 0.0063, "step": 23869 }, { "epoch": 2.5878143972246312, "grad_norm": 0.7061929106712341, "learning_rate": 2.7578703032061516e-06, "loss": 0.018, "step": 23870 }, { "epoch": 2.5879228100607112, "grad_norm": 0.5652488470077515, "learning_rate": 2.757144929638764e-06, "loss": 0.0129, "step": 23871 }, { "epoch": 2.5880312228967908, "grad_norm": 0.4767044186592102, "learning_rate": 2.7564195560713767e-06, "loss": 0.0196, "step": 23872 }, { "epoch": 2.5881396357328708, "grad_norm": 0.035041723400354385, "learning_rate": 2.7556941825039897e-06, "loss": 0.0003, "step": 23873 }, { "epoch": 2.5882480485689507, "grad_norm": 3.484910249710083, "learning_rate": 2.7549688089366027e-06, "loss": 0.0032, "step": 23874 }, { "epoch": 2.5883564614050303, "grad_norm": 6.0790557861328125, "learning_rate": 2.7542434353692152e-06, "loss": 0.0579, "step": 23875 }, { "epoch": 2.5884648742411103, "grad_norm": 1.1357887983322144, "learning_rate": 2.7535180618018282e-06, "loss": 0.0336, "step": 23876 }, { "epoch": 2.58857328707719, "grad_norm": 0.3268447518348694, "learning_rate": 2.752792688234441e-06, "loss": 0.0117, "step": 23877 }, { "epoch": 2.58868169991327, "grad_norm": 2.184965133666992, "learning_rate": 2.7520673146670542e-06, "loss": 0.0251, "step": 23878 }, { "epoch": 2.58879011274935, "grad_norm": 0.32042357325553894, "learning_rate": 2.7513419410996668e-06, "loss": 0.0156, "step": 23879 }, { "epoch": 2.5888985255854293, "grad_norm": 0.7405324578285217, "learning_rate": 2.7506165675322794e-06, "loss": 0.029, "step": 23880 }, { "epoch": 2.589006938421509, "grad_norm": 0.33238130807876587, "learning_rate": 2.749891193964892e-06, "loss": 0.0082, "step": 23881 }, { "epoch": 2.589115351257589, "grad_norm": 5.270463943481445, "learning_rate": 2.7491658203975045e-06, "loss": 0.0389, "step": 23882 }, { "epoch": 2.589223764093669, "grad_norm": 0.7809132933616638, "learning_rate": 2.748440446830118e-06, "loss": 0.0374, "step": 23883 }, { "epoch": 2.5893321769297484, "grad_norm": 8.123881340026855, "learning_rate": 2.7477150732627305e-06, "loss": 0.0156, "step": 23884 }, { "epoch": 2.5894405897658284, "grad_norm": 0.5217363238334656, "learning_rate": 2.746989699695343e-06, "loss": 0.029, "step": 23885 }, { "epoch": 2.589549002601908, "grad_norm": 0.5751108527183533, "learning_rate": 2.746264326127956e-06, "loss": 0.0163, "step": 23886 }, { "epoch": 2.589657415437988, "grad_norm": 0.05563070625066757, "learning_rate": 2.745538952560569e-06, "loss": 0.0003, "step": 23887 }, { "epoch": 2.589765828274068, "grad_norm": 0.7070625424385071, "learning_rate": 2.7448135789931816e-06, "loss": 0.0281, "step": 23888 }, { "epoch": 2.5898742411101474, "grad_norm": 0.30190354585647583, "learning_rate": 2.7440882054257946e-06, "loss": 0.0066, "step": 23889 }, { "epoch": 2.5899826539462274, "grad_norm": 2.190324306488037, "learning_rate": 2.743362831858407e-06, "loss": 0.0058, "step": 23890 }, { "epoch": 2.590091066782307, "grad_norm": 0.043455079197883606, "learning_rate": 2.7426374582910206e-06, "loss": 0.0004, "step": 23891 }, { "epoch": 2.590199479618387, "grad_norm": 0.3432423770427704, "learning_rate": 2.741912084723633e-06, "loss": 0.0117, "step": 23892 }, { "epoch": 2.5903078924544665, "grad_norm": 0.28897401690483093, "learning_rate": 2.7411867111562457e-06, "loss": 0.0059, "step": 23893 }, { "epoch": 2.5904163052905465, "grad_norm": 1.8731151819229126, "learning_rate": 2.7404613375888583e-06, "loss": 0.0153, "step": 23894 }, { "epoch": 2.590524718126626, "grad_norm": 0.11704897880554199, "learning_rate": 2.739735964021471e-06, "loss": 0.0011, "step": 23895 }, { "epoch": 2.590633130962706, "grad_norm": 0.9408062696456909, "learning_rate": 2.7390105904540842e-06, "loss": 0.005, "step": 23896 }, { "epoch": 2.590741543798786, "grad_norm": 0.38044559955596924, "learning_rate": 2.738285216886697e-06, "loss": 0.0054, "step": 23897 }, { "epoch": 2.5908499566348655, "grad_norm": 0.1221236065030098, "learning_rate": 2.7375598433193094e-06, "loss": 0.0007, "step": 23898 }, { "epoch": 2.5909583694709455, "grad_norm": 0.6867281794548035, "learning_rate": 2.7368344697519224e-06, "loss": 0.0201, "step": 23899 }, { "epoch": 2.591066782307025, "grad_norm": 0.2233501821756363, "learning_rate": 2.7361090961845353e-06, "loss": 0.0017, "step": 23900 }, { "epoch": 2.591175195143105, "grad_norm": 0.06755711138248444, "learning_rate": 2.7353837226171483e-06, "loss": 0.0009, "step": 23901 }, { "epoch": 2.591283607979185, "grad_norm": 0.08407146483659744, "learning_rate": 2.734658349049761e-06, "loss": 0.011, "step": 23902 }, { "epoch": 2.5913920208152645, "grad_norm": 0.4959128797054291, "learning_rate": 2.7339329754823735e-06, "loss": 0.0078, "step": 23903 }, { "epoch": 2.591500433651344, "grad_norm": 0.23558075726032257, "learning_rate": 2.733207601914987e-06, "loss": 0.0053, "step": 23904 }, { "epoch": 2.591608846487424, "grad_norm": 1.3860199451446533, "learning_rate": 2.7324822283475995e-06, "loss": 0.0094, "step": 23905 }, { "epoch": 2.591717259323504, "grad_norm": 0.3400975465774536, "learning_rate": 2.731756854780212e-06, "loss": 0.0049, "step": 23906 }, { "epoch": 2.5918256721595836, "grad_norm": 4.502228736877441, "learning_rate": 2.7310314812128246e-06, "loss": 0.0078, "step": 23907 }, { "epoch": 2.5919340849956636, "grad_norm": 0.6128430962562561, "learning_rate": 2.730306107645437e-06, "loss": 0.0126, "step": 23908 }, { "epoch": 2.592042497831743, "grad_norm": 2.227977752685547, "learning_rate": 2.7295807340780506e-06, "loss": 0.0198, "step": 23909 }, { "epoch": 2.592150910667823, "grad_norm": 2.5075273513793945, "learning_rate": 2.728855360510663e-06, "loss": 0.0027, "step": 23910 }, { "epoch": 2.592259323503903, "grad_norm": 0.08887455612421036, "learning_rate": 2.7281299869432757e-06, "loss": 0.0007, "step": 23911 }, { "epoch": 2.5923677363399826, "grad_norm": 0.3675498068332672, "learning_rate": 2.7274046133758887e-06, "loss": 0.022, "step": 23912 }, { "epoch": 2.592476149176062, "grad_norm": 0.2969605028629303, "learning_rate": 2.7266792398085017e-06, "loss": 0.0004, "step": 23913 }, { "epoch": 2.592584562012142, "grad_norm": 0.08981356024742126, "learning_rate": 2.7259538662411147e-06, "loss": 0.0008, "step": 23914 }, { "epoch": 2.592692974848222, "grad_norm": 0.9242825508117676, "learning_rate": 2.7252284926737272e-06, "loss": 0.0011, "step": 23915 }, { "epoch": 2.5928013876843017, "grad_norm": 0.478223979473114, "learning_rate": 2.72450311910634e-06, "loss": 0.0152, "step": 23916 }, { "epoch": 2.5929098005203817, "grad_norm": 0.24762384593486786, "learning_rate": 2.7237777455389532e-06, "loss": 0.004, "step": 23917 }, { "epoch": 2.593018213356461, "grad_norm": 0.5684338808059692, "learning_rate": 2.723052371971566e-06, "loss": 0.0121, "step": 23918 }, { "epoch": 2.593126626192541, "grad_norm": 0.4534953534603119, "learning_rate": 2.7223269984041784e-06, "loss": 0.0173, "step": 23919 }, { "epoch": 2.593235039028621, "grad_norm": 0.25519663095474243, "learning_rate": 2.721601624836791e-06, "loss": 0.0052, "step": 23920 }, { "epoch": 2.5933434518647007, "grad_norm": 0.30245479941368103, "learning_rate": 2.7208762512694035e-06, "loss": 0.0066, "step": 23921 }, { "epoch": 2.5934518647007807, "grad_norm": 0.04244064912199974, "learning_rate": 2.720150877702017e-06, "loss": 0.0004, "step": 23922 }, { "epoch": 2.5935602775368602, "grad_norm": 4.953568935394287, "learning_rate": 2.7194255041346295e-06, "loss": 0.0308, "step": 23923 }, { "epoch": 2.5936686903729402, "grad_norm": 0.07163671404123306, "learning_rate": 2.718700130567242e-06, "loss": 0.0114, "step": 23924 }, { "epoch": 2.59377710320902, "grad_norm": 6.325626850128174, "learning_rate": 2.717974756999855e-06, "loss": 0.0187, "step": 23925 }, { "epoch": 2.5938855160450998, "grad_norm": 0.47898775339126587, "learning_rate": 2.717249383432468e-06, "loss": 0.0161, "step": 23926 }, { "epoch": 2.5939939288811793, "grad_norm": 0.010000636801123619, "learning_rate": 2.716524009865081e-06, "loss": 0.0001, "step": 23927 }, { "epoch": 2.5941023417172593, "grad_norm": 2.3074991703033447, "learning_rate": 2.7157986362976936e-06, "loss": 0.0104, "step": 23928 }, { "epoch": 2.5942107545533393, "grad_norm": 0.00510050542652607, "learning_rate": 2.715073262730306e-06, "loss": 0.0001, "step": 23929 }, { "epoch": 2.594319167389419, "grad_norm": 0.38902825117111206, "learning_rate": 2.7143478891629196e-06, "loss": 0.0106, "step": 23930 }, { "epoch": 2.594427580225499, "grad_norm": 0.38563990592956543, "learning_rate": 2.713622515595532e-06, "loss": 0.0242, "step": 23931 }, { "epoch": 2.5945359930615783, "grad_norm": 3.948434591293335, "learning_rate": 2.7128971420281447e-06, "loss": 0.0121, "step": 23932 }, { "epoch": 2.5946444058976583, "grad_norm": 0.3154412508010864, "learning_rate": 2.7121717684607573e-06, "loss": 0.0104, "step": 23933 }, { "epoch": 2.5947528187337383, "grad_norm": 1.6780004501342773, "learning_rate": 2.7114463948933707e-06, "loss": 0.0027, "step": 23934 }, { "epoch": 2.594861231569818, "grad_norm": 0.38327693939208984, "learning_rate": 2.7107210213259832e-06, "loss": 0.0116, "step": 23935 }, { "epoch": 2.5949696444058974, "grad_norm": 0.3926045298576355, "learning_rate": 2.709995647758596e-06, "loss": 0.0091, "step": 23936 }, { "epoch": 2.5950780572419774, "grad_norm": 0.17220328748226166, "learning_rate": 2.7092702741912084e-06, "loss": 0.0013, "step": 23937 }, { "epoch": 2.5951864700780574, "grad_norm": 0.046795520931482315, "learning_rate": 2.7085449006238214e-06, "loss": 0.0005, "step": 23938 }, { "epoch": 2.595294882914137, "grad_norm": 0.04307274520397186, "learning_rate": 2.7078195270564344e-06, "loss": 0.0004, "step": 23939 }, { "epoch": 2.595403295750217, "grad_norm": 5.288890838623047, "learning_rate": 2.7070941534890473e-06, "loss": 0.0338, "step": 23940 }, { "epoch": 2.5955117085862964, "grad_norm": 0.4248071014881134, "learning_rate": 2.70636877992166e-06, "loss": 0.0087, "step": 23941 }, { "epoch": 2.5956201214223764, "grad_norm": 0.2753952443599701, "learning_rate": 2.7056434063542725e-06, "loss": 0.006, "step": 23942 }, { "epoch": 2.5957285342584564, "grad_norm": 0.25715893507003784, "learning_rate": 2.704918032786886e-06, "loss": 0.0051, "step": 23943 }, { "epoch": 2.595836947094536, "grad_norm": 1.990564227104187, "learning_rate": 2.7041926592194985e-06, "loss": 0.0038, "step": 23944 }, { "epoch": 2.595945359930616, "grad_norm": 0.13598991930484772, "learning_rate": 2.703467285652111e-06, "loss": 0.001, "step": 23945 }, { "epoch": 2.5960537727666955, "grad_norm": 0.020065834745764732, "learning_rate": 2.7027419120847236e-06, "loss": 0.0003, "step": 23946 }, { "epoch": 2.5961621856027755, "grad_norm": 0.7950169444084167, "learning_rate": 2.702016538517337e-06, "loss": 0.0024, "step": 23947 }, { "epoch": 2.5962705984388554, "grad_norm": 4.813104629516602, "learning_rate": 2.7012911649499496e-06, "loss": 0.0199, "step": 23948 }, { "epoch": 2.596379011274935, "grad_norm": 0.08874932676553726, "learning_rate": 2.700565791382562e-06, "loss": 0.0004, "step": 23949 }, { "epoch": 2.5964874241110145, "grad_norm": 13.08582878112793, "learning_rate": 2.699840417815175e-06, "loss": 0.0106, "step": 23950 }, { "epoch": 2.5965958369470945, "grad_norm": 0.7015530467033386, "learning_rate": 2.6991150442477877e-06, "loss": 0.0154, "step": 23951 }, { "epoch": 2.5967042497831745, "grad_norm": 1.4977164268493652, "learning_rate": 2.6983896706804007e-06, "loss": 0.0031, "step": 23952 }, { "epoch": 2.596812662619254, "grad_norm": 0.09498118609189987, "learning_rate": 2.6976642971130137e-06, "loss": 0.0005, "step": 23953 }, { "epoch": 2.596921075455334, "grad_norm": 0.4221920371055603, "learning_rate": 2.6969389235456262e-06, "loss": 0.0125, "step": 23954 }, { "epoch": 2.5970294882914136, "grad_norm": 0.33334770798683167, "learning_rate": 2.696213549978239e-06, "loss": 0.0108, "step": 23955 }, { "epoch": 2.5971379011274935, "grad_norm": 0.07404866814613342, "learning_rate": 2.6954881764108522e-06, "loss": 0.0008, "step": 23956 }, { "epoch": 2.5972463139635735, "grad_norm": 0.6506964564323425, "learning_rate": 2.694762802843465e-06, "loss": 0.0225, "step": 23957 }, { "epoch": 2.597354726799653, "grad_norm": 3.200460910797119, "learning_rate": 2.6940374292760774e-06, "loss": 0.0397, "step": 23958 }, { "epoch": 2.5974631396357326, "grad_norm": 2.5817368030548096, "learning_rate": 2.69331205570869e-06, "loss": 0.0052, "step": 23959 }, { "epoch": 2.5975715524718126, "grad_norm": 0.5487369894981384, "learning_rate": 2.6925866821413033e-06, "loss": 0.0019, "step": 23960 }, { "epoch": 2.5976799653078926, "grad_norm": 0.04982106387615204, "learning_rate": 2.691861308573916e-06, "loss": 0.0004, "step": 23961 }, { "epoch": 2.597788378143972, "grad_norm": 9.331910133361816, "learning_rate": 2.6911359350065285e-06, "loss": 0.0202, "step": 23962 }, { "epoch": 2.597896790980052, "grad_norm": 12.9475679397583, "learning_rate": 2.6904105614391415e-06, "loss": 0.0181, "step": 23963 }, { "epoch": 2.5980052038161316, "grad_norm": 0.626211404800415, "learning_rate": 2.689685187871754e-06, "loss": 0.0205, "step": 23964 }, { "epoch": 2.5981136166522116, "grad_norm": 6.69188117980957, "learning_rate": 2.688959814304367e-06, "loss": 0.0252, "step": 23965 }, { "epoch": 2.5982220294882916, "grad_norm": 0.4415017366409302, "learning_rate": 2.68823444073698e-06, "loss": 0.0171, "step": 23966 }, { "epoch": 2.598330442324371, "grad_norm": 2.2190186977386475, "learning_rate": 2.6875090671695926e-06, "loss": 0.0504, "step": 23967 }, { "epoch": 2.598438855160451, "grad_norm": 0.023278770968317986, "learning_rate": 2.686783693602205e-06, "loss": 0.0002, "step": 23968 }, { "epoch": 2.5985472679965307, "grad_norm": 0.39871785044670105, "learning_rate": 2.6860583200348186e-06, "loss": 0.014, "step": 23969 }, { "epoch": 2.5986556808326107, "grad_norm": 1.4328131675720215, "learning_rate": 2.685332946467431e-06, "loss": 0.0135, "step": 23970 }, { "epoch": 2.59876409366869, "grad_norm": 0.03217967599630356, "learning_rate": 2.6846075729000437e-06, "loss": 0.0003, "step": 23971 }, { "epoch": 2.59887250650477, "grad_norm": 0.6541270017623901, "learning_rate": 2.6838821993326563e-06, "loss": 0.0098, "step": 23972 }, { "epoch": 2.5989809193408497, "grad_norm": 0.014582715928554535, "learning_rate": 2.6831568257652697e-06, "loss": 0.0002, "step": 23973 }, { "epoch": 2.5990893321769297, "grad_norm": 6.254929542541504, "learning_rate": 2.6824314521978822e-06, "loss": 0.069, "step": 23974 }, { "epoch": 2.5991977450130097, "grad_norm": 1.8956964015960693, "learning_rate": 2.681706078630495e-06, "loss": 0.0662, "step": 23975 }, { "epoch": 2.5993061578490892, "grad_norm": 0.4692115783691406, "learning_rate": 2.680980705063108e-06, "loss": 0.02, "step": 23976 }, { "epoch": 2.5994145706851692, "grad_norm": 0.09958070516586304, "learning_rate": 2.6802553314957204e-06, "loss": 0.0005, "step": 23977 }, { "epoch": 2.5995229835212488, "grad_norm": 0.39439335465431213, "learning_rate": 2.6795299579283334e-06, "loss": 0.0131, "step": 23978 }, { "epoch": 2.5996313963573288, "grad_norm": 4.528048515319824, "learning_rate": 2.6788045843609463e-06, "loss": 0.0083, "step": 23979 }, { "epoch": 2.5997398091934087, "grad_norm": 1.5675289630889893, "learning_rate": 2.678079210793559e-06, "loss": 0.0042, "step": 23980 }, { "epoch": 2.5998482220294883, "grad_norm": 0.5564650893211365, "learning_rate": 2.6773538372261715e-06, "loss": 0.0169, "step": 23981 }, { "epoch": 2.599956634865568, "grad_norm": 1.8572890758514404, "learning_rate": 2.676628463658785e-06, "loss": 0.0136, "step": 23982 }, { "epoch": 2.600065047701648, "grad_norm": 3.236924409866333, "learning_rate": 2.6759030900913975e-06, "loss": 0.021, "step": 23983 }, { "epoch": 2.600173460537728, "grad_norm": 0.2621361017227173, "learning_rate": 2.67517771652401e-06, "loss": 0.005, "step": 23984 }, { "epoch": 2.6002818733738073, "grad_norm": 1.4607309103012085, "learning_rate": 2.6744523429566226e-06, "loss": 0.0016, "step": 23985 }, { "epoch": 2.6003902862098873, "grad_norm": 0.19977523386478424, "learning_rate": 2.673726969389236e-06, "loss": 0.001, "step": 23986 }, { "epoch": 2.600498699045967, "grad_norm": 2.387665271759033, "learning_rate": 2.6730015958218486e-06, "loss": 0.0318, "step": 23987 }, { "epoch": 2.600607111882047, "grad_norm": 0.4583219289779663, "learning_rate": 2.672276222254461e-06, "loss": 0.0069, "step": 23988 }, { "epoch": 2.600715524718127, "grad_norm": 3.7995879650115967, "learning_rate": 2.671550848687074e-06, "loss": 0.0173, "step": 23989 }, { "epoch": 2.6008239375542064, "grad_norm": 0.5174383521080017, "learning_rate": 2.6708254751196867e-06, "loss": 0.0249, "step": 23990 }, { "epoch": 2.6009323503902864, "grad_norm": 0.3132256269454956, "learning_rate": 2.6701001015522997e-06, "loss": 0.007, "step": 23991 }, { "epoch": 2.601040763226366, "grad_norm": 24.352737426757812, "learning_rate": 2.6693747279849127e-06, "loss": 0.0331, "step": 23992 }, { "epoch": 2.601149176062446, "grad_norm": 0.6240548491477966, "learning_rate": 2.6686493544175252e-06, "loss": 0.0143, "step": 23993 }, { "epoch": 2.6012575888985254, "grad_norm": 0.33127474784851074, "learning_rate": 2.667923980850138e-06, "loss": 0.0109, "step": 23994 }, { "epoch": 2.6013660017346054, "grad_norm": 7.498038291931152, "learning_rate": 2.6671986072827512e-06, "loss": 0.0049, "step": 23995 }, { "epoch": 2.601474414570685, "grad_norm": 3.473155975341797, "learning_rate": 2.666473233715364e-06, "loss": 0.0281, "step": 23996 }, { "epoch": 2.601582827406765, "grad_norm": 0.49511703848838806, "learning_rate": 2.6657478601479764e-06, "loss": 0.0011, "step": 23997 }, { "epoch": 2.601691240242845, "grad_norm": 2.781597375869751, "learning_rate": 2.665022486580589e-06, "loss": 0.0288, "step": 23998 }, { "epoch": 2.6017996530789245, "grad_norm": 2.065556764602661, "learning_rate": 2.6642971130132023e-06, "loss": 0.0485, "step": 23999 }, { "epoch": 2.6019080659150045, "grad_norm": 0.06778482347726822, "learning_rate": 2.663571739445815e-06, "loss": 0.0003, "step": 24000 }, { "epoch": 2.602016478751084, "grad_norm": 0.10389605909585953, "learning_rate": 2.6628463658784275e-06, "loss": 0.0007, "step": 24001 }, { "epoch": 2.602124891587164, "grad_norm": 2.55889630317688, "learning_rate": 2.6621209923110405e-06, "loss": 0.0405, "step": 24002 }, { "epoch": 2.602233304423244, "grad_norm": 0.4249579608440399, "learning_rate": 2.661395618743653e-06, "loss": 0.0128, "step": 24003 }, { "epoch": 2.6023417172593235, "grad_norm": 1.2801812887191772, "learning_rate": 2.660670245176266e-06, "loss": 0.0008, "step": 24004 }, { "epoch": 2.602450130095403, "grad_norm": 0.32845374941825867, "learning_rate": 2.659944871608879e-06, "loss": 0.0115, "step": 24005 }, { "epoch": 2.602558542931483, "grad_norm": 0.32173219323158264, "learning_rate": 2.6592194980414916e-06, "loss": 0.0125, "step": 24006 }, { "epoch": 2.602666955767563, "grad_norm": 0.4219622313976288, "learning_rate": 2.658494124474104e-06, "loss": 0.001, "step": 24007 }, { "epoch": 2.6027753686036426, "grad_norm": 3.648986577987671, "learning_rate": 2.6577687509067176e-06, "loss": 0.0163, "step": 24008 }, { "epoch": 2.6028837814397225, "grad_norm": 3.5683512687683105, "learning_rate": 2.65704337733933e-06, "loss": 0.0296, "step": 24009 }, { "epoch": 2.602992194275802, "grad_norm": 0.025769861415028572, "learning_rate": 2.6563180037719427e-06, "loss": 0.0002, "step": 24010 }, { "epoch": 2.603100607111882, "grad_norm": 0.20023544132709503, "learning_rate": 2.6555926302045553e-06, "loss": 0.0006, "step": 24011 }, { "epoch": 2.603209019947962, "grad_norm": 0.2357342392206192, "learning_rate": 2.6548672566371687e-06, "loss": 0.0022, "step": 24012 }, { "epoch": 2.6033174327840416, "grad_norm": 0.005410626996308565, "learning_rate": 2.6541418830697812e-06, "loss": 0.0001, "step": 24013 }, { "epoch": 2.6034258456201216, "grad_norm": 5.329833507537842, "learning_rate": 2.653416509502394e-06, "loss": 0.0391, "step": 24014 }, { "epoch": 2.603534258456201, "grad_norm": 1.7835770845413208, "learning_rate": 2.652691135935007e-06, "loss": 0.007, "step": 24015 }, { "epoch": 2.603642671292281, "grad_norm": 0.06947585195302963, "learning_rate": 2.6519657623676194e-06, "loss": 0.0006, "step": 24016 }, { "epoch": 2.6037510841283606, "grad_norm": 0.1329183429479599, "learning_rate": 2.6512403888002324e-06, "loss": 0.0009, "step": 24017 }, { "epoch": 2.6038594969644406, "grad_norm": 6.007593631744385, "learning_rate": 2.6505150152328453e-06, "loss": 0.0212, "step": 24018 }, { "epoch": 2.60396790980052, "grad_norm": 1.1263929605484009, "learning_rate": 2.649789641665458e-06, "loss": 0.0178, "step": 24019 }, { "epoch": 2.6040763226366, "grad_norm": 0.6690580248832703, "learning_rate": 2.6490642680980705e-06, "loss": 0.0146, "step": 24020 }, { "epoch": 2.60418473547268, "grad_norm": 0.3282928168773651, "learning_rate": 2.648338894530684e-06, "loss": 0.0097, "step": 24021 }, { "epoch": 2.6042931483087597, "grad_norm": 1.0399845838546753, "learning_rate": 2.6476135209632965e-06, "loss": 0.0207, "step": 24022 }, { "epoch": 2.6044015611448397, "grad_norm": 0.01073443703353405, "learning_rate": 2.646888147395909e-06, "loss": 0.0001, "step": 24023 }, { "epoch": 2.604509973980919, "grad_norm": 0.43783897161483765, "learning_rate": 2.6461627738285216e-06, "loss": 0.0144, "step": 24024 }, { "epoch": 2.604618386816999, "grad_norm": 2.0612995624542236, "learning_rate": 2.645437400261135e-06, "loss": 0.0137, "step": 24025 }, { "epoch": 2.604726799653079, "grad_norm": 1.4865453243255615, "learning_rate": 2.6447120266937476e-06, "loss": 0.0298, "step": 24026 }, { "epoch": 2.6048352124891587, "grad_norm": 2.2925539016723633, "learning_rate": 2.64398665312636e-06, "loss": 0.02, "step": 24027 }, { "epoch": 2.6049436253252383, "grad_norm": 0.1781015247106552, "learning_rate": 2.643261279558973e-06, "loss": 0.0006, "step": 24028 }, { "epoch": 2.6050520381613183, "grad_norm": 0.6215021014213562, "learning_rate": 2.642535905991586e-06, "loss": 0.0165, "step": 24029 }, { "epoch": 2.6051604509973982, "grad_norm": 0.2577485740184784, "learning_rate": 2.6418105324241987e-06, "loss": 0.0052, "step": 24030 }, { "epoch": 2.6052688638334778, "grad_norm": 0.1634209156036377, "learning_rate": 2.6410851588568117e-06, "loss": 0.0003, "step": 24031 }, { "epoch": 2.6053772766695578, "grad_norm": 0.20081618428230286, "learning_rate": 2.6403597852894243e-06, "loss": 0.0029, "step": 24032 }, { "epoch": 2.6054856895056373, "grad_norm": 0.06448076665401459, "learning_rate": 2.639634411722037e-06, "loss": 0.0004, "step": 24033 }, { "epoch": 2.6055941023417173, "grad_norm": 0.30630868673324585, "learning_rate": 2.6389090381546502e-06, "loss": 0.0068, "step": 24034 }, { "epoch": 2.6057025151777973, "grad_norm": 0.5510169267654419, "learning_rate": 2.638183664587263e-06, "loss": 0.0171, "step": 24035 }, { "epoch": 2.605810928013877, "grad_norm": 0.5943373441696167, "learning_rate": 2.6374582910198754e-06, "loss": 0.0119, "step": 24036 }, { "epoch": 2.605919340849957, "grad_norm": 3.860116720199585, "learning_rate": 2.636732917452488e-06, "loss": 0.0568, "step": 24037 }, { "epoch": 2.6060277536860363, "grad_norm": 0.4369298219680786, "learning_rate": 2.6360075438851013e-06, "loss": 0.0163, "step": 24038 }, { "epoch": 2.6061361665221163, "grad_norm": 0.582685649394989, "learning_rate": 2.635282170317714e-06, "loss": 0.021, "step": 24039 }, { "epoch": 2.606244579358196, "grad_norm": 11.600761413574219, "learning_rate": 2.6345567967503265e-06, "loss": 0.0255, "step": 24040 }, { "epoch": 2.606352992194276, "grad_norm": 16.045305252075195, "learning_rate": 2.6338314231829395e-06, "loss": 0.079, "step": 24041 }, { "epoch": 2.6064614050303554, "grad_norm": 0.47489625215530396, "learning_rate": 2.6331060496155525e-06, "loss": 0.0106, "step": 24042 }, { "epoch": 2.6065698178664354, "grad_norm": 0.027385618537664413, "learning_rate": 2.632380676048165e-06, "loss": 0.0001, "step": 24043 }, { "epoch": 2.6066782307025154, "grad_norm": 0.5525198578834534, "learning_rate": 2.631655302480778e-06, "loss": 0.0165, "step": 24044 }, { "epoch": 2.606786643538595, "grad_norm": 0.2508545517921448, "learning_rate": 2.6309299289133906e-06, "loss": 0.0051, "step": 24045 }, { "epoch": 2.606895056374675, "grad_norm": 0.10052888840436935, "learning_rate": 2.630204555346003e-06, "loss": 0.0006, "step": 24046 }, { "epoch": 2.6070034692107544, "grad_norm": 0.3831496238708496, "learning_rate": 2.6294791817786166e-06, "loss": 0.0144, "step": 24047 }, { "epoch": 2.6071118820468344, "grad_norm": 0.6377076506614685, "learning_rate": 2.628753808211229e-06, "loss": 0.001, "step": 24048 }, { "epoch": 2.6072202948829144, "grad_norm": 0.007322638761252165, "learning_rate": 2.6280284346438417e-06, "loss": 0.0001, "step": 24049 }, { "epoch": 2.607328707718994, "grad_norm": 0.1962238848209381, "learning_rate": 2.6273030610764543e-06, "loss": 0.0012, "step": 24050 }, { "epoch": 2.6074371205550735, "grad_norm": 0.3943076729774475, "learning_rate": 2.6265776875090677e-06, "loss": 0.0063, "step": 24051 }, { "epoch": 2.6075455333911535, "grad_norm": 0.35338011384010315, "learning_rate": 2.6258523139416802e-06, "loss": 0.0091, "step": 24052 }, { "epoch": 2.6076539462272335, "grad_norm": 0.03102336823940277, "learning_rate": 2.625126940374293e-06, "loss": 0.0003, "step": 24053 }, { "epoch": 2.607762359063313, "grad_norm": 3.5688369274139404, "learning_rate": 2.624401566806906e-06, "loss": 0.0039, "step": 24054 }, { "epoch": 2.607870771899393, "grad_norm": 0.26430168747901917, "learning_rate": 2.623676193239519e-06, "loss": 0.0055, "step": 24055 }, { "epoch": 2.6079791847354725, "grad_norm": 0.5524054765701294, "learning_rate": 2.6229508196721314e-06, "loss": 0.011, "step": 24056 }, { "epoch": 2.6080875975715525, "grad_norm": 0.34197738766670227, "learning_rate": 2.6222254461047444e-06, "loss": 0.0116, "step": 24057 }, { "epoch": 2.6081960104076325, "grad_norm": 0.32056286931037903, "learning_rate": 2.621500072537357e-06, "loss": 0.0159, "step": 24058 }, { "epoch": 2.608304423243712, "grad_norm": 0.0030439472757279873, "learning_rate": 2.6207746989699695e-06, "loss": 0.0, "step": 24059 }, { "epoch": 2.6084128360797916, "grad_norm": 0.2947607934474945, "learning_rate": 2.620049325402583e-06, "loss": 0.0108, "step": 24060 }, { "epoch": 2.6085212489158716, "grad_norm": 0.08471276611089706, "learning_rate": 2.6193239518351955e-06, "loss": 0.0003, "step": 24061 }, { "epoch": 2.6086296617519515, "grad_norm": 0.07002165913581848, "learning_rate": 2.618598578267808e-06, "loss": 0.0006, "step": 24062 }, { "epoch": 2.608738074588031, "grad_norm": 0.5845566391944885, "learning_rate": 2.6178732047004206e-06, "loss": 0.0148, "step": 24063 }, { "epoch": 2.608846487424111, "grad_norm": 0.17437224090099335, "learning_rate": 2.617147831133034e-06, "loss": 0.0012, "step": 24064 }, { "epoch": 2.6089549002601906, "grad_norm": 0.022781074047088623, "learning_rate": 2.6164224575656466e-06, "loss": 0.0001, "step": 24065 }, { "epoch": 2.6090633130962706, "grad_norm": 0.10813024640083313, "learning_rate": 2.615697083998259e-06, "loss": 0.0006, "step": 24066 }, { "epoch": 2.6091717259323506, "grad_norm": 0.039296455681324005, "learning_rate": 2.614971710430872e-06, "loss": 0.0004, "step": 24067 }, { "epoch": 2.60928013876843, "grad_norm": 0.02392936870455742, "learning_rate": 2.614246336863485e-06, "loss": 0.0002, "step": 24068 }, { "epoch": 2.60938855160451, "grad_norm": 1.1229302883148193, "learning_rate": 2.6135209632960977e-06, "loss": 0.021, "step": 24069 }, { "epoch": 2.6094969644405897, "grad_norm": 1.838950753211975, "learning_rate": 2.6127955897287107e-06, "loss": 0.0123, "step": 24070 }, { "epoch": 2.6096053772766696, "grad_norm": 0.41553226113319397, "learning_rate": 2.6120702161613233e-06, "loss": 0.0073, "step": 24071 }, { "epoch": 2.6097137901127496, "grad_norm": 0.019852513447403908, "learning_rate": 2.611344842593936e-06, "loss": 0.0002, "step": 24072 }, { "epoch": 2.609822202948829, "grad_norm": 2.6457111835479736, "learning_rate": 2.6106194690265492e-06, "loss": 0.0164, "step": 24073 }, { "epoch": 2.6099306157849087, "grad_norm": 4.8710432052612305, "learning_rate": 2.609894095459162e-06, "loss": 0.0088, "step": 24074 }, { "epoch": 2.6100390286209887, "grad_norm": 2.1976029872894287, "learning_rate": 2.6091687218917744e-06, "loss": 0.0296, "step": 24075 }, { "epoch": 2.6101474414570687, "grad_norm": 7.631202697753906, "learning_rate": 2.608443348324387e-06, "loss": 0.0213, "step": 24076 }, { "epoch": 2.610255854293148, "grad_norm": 0.9650010466575623, "learning_rate": 2.6077179747570003e-06, "loss": 0.0122, "step": 24077 }, { "epoch": 2.610364267129228, "grad_norm": 5.928183078765869, "learning_rate": 2.606992601189613e-06, "loss": 0.051, "step": 24078 }, { "epoch": 2.6104726799653077, "grad_norm": 0.4977576732635498, "learning_rate": 2.6062672276222255e-06, "loss": 0.0157, "step": 24079 }, { "epoch": 2.6105810928013877, "grad_norm": 0.32237356901168823, "learning_rate": 2.6055418540548385e-06, "loss": 0.0109, "step": 24080 }, { "epoch": 2.6106895056374677, "grad_norm": 2.5894033908843994, "learning_rate": 2.6048164804874515e-06, "loss": 0.0299, "step": 24081 }, { "epoch": 2.6107979184735473, "grad_norm": 23.536088943481445, "learning_rate": 2.604091106920064e-06, "loss": 0.0107, "step": 24082 }, { "epoch": 2.610906331309627, "grad_norm": 0.4689217209815979, "learning_rate": 2.603365733352677e-06, "loss": 0.0152, "step": 24083 }, { "epoch": 2.611014744145707, "grad_norm": 0.1931040734052658, "learning_rate": 2.6026403597852896e-06, "loss": 0.0019, "step": 24084 }, { "epoch": 2.6111231569817868, "grad_norm": 0.2843218147754669, "learning_rate": 2.601914986217902e-06, "loss": 0.0008, "step": 24085 }, { "epoch": 2.6112315698178663, "grad_norm": 0.40306350588798523, "learning_rate": 2.6011896126505156e-06, "loss": 0.0104, "step": 24086 }, { "epoch": 2.6113399826539463, "grad_norm": 0.8403815031051636, "learning_rate": 2.600464239083128e-06, "loss": 0.0264, "step": 24087 }, { "epoch": 2.611448395490026, "grad_norm": 2.20367431640625, "learning_rate": 2.5997388655157407e-06, "loss": 0.0327, "step": 24088 }, { "epoch": 2.611556808326106, "grad_norm": 0.015942905098199844, "learning_rate": 2.5990134919483533e-06, "loss": 0.0001, "step": 24089 }, { "epoch": 2.611665221162186, "grad_norm": 0.011530717834830284, "learning_rate": 2.5982881183809667e-06, "loss": 0.0001, "step": 24090 }, { "epoch": 2.6117736339982653, "grad_norm": 0.041708510369062424, "learning_rate": 2.5975627448135793e-06, "loss": 0.0003, "step": 24091 }, { "epoch": 2.6118820468343453, "grad_norm": 0.3035925626754761, "learning_rate": 2.596837371246192e-06, "loss": 0.0041, "step": 24092 }, { "epoch": 2.611990459670425, "grad_norm": 0.2912806570529938, "learning_rate": 2.596111997678805e-06, "loss": 0.0116, "step": 24093 }, { "epoch": 2.612098872506505, "grad_norm": 0.49738210439682007, "learning_rate": 2.595386624111418e-06, "loss": 0.0173, "step": 24094 }, { "epoch": 2.612207285342585, "grad_norm": 0.03076137602329254, "learning_rate": 2.5946612505440304e-06, "loss": 0.0003, "step": 24095 }, { "epoch": 2.6123156981786644, "grad_norm": 0.6904426217079163, "learning_rate": 2.5939358769766434e-06, "loss": 0.0231, "step": 24096 }, { "epoch": 2.612424111014744, "grad_norm": 0.5952132344245911, "learning_rate": 2.593210503409256e-06, "loss": 0.025, "step": 24097 }, { "epoch": 2.612532523850824, "grad_norm": 0.018337121233344078, "learning_rate": 2.5924851298418685e-06, "loss": 0.0001, "step": 24098 }, { "epoch": 2.612640936686904, "grad_norm": 0.8754390478134155, "learning_rate": 2.591759756274482e-06, "loss": 0.0254, "step": 24099 }, { "epoch": 2.6127493495229834, "grad_norm": 2.4538075923919678, "learning_rate": 2.5910343827070945e-06, "loss": 0.03, "step": 24100 }, { "epoch": 2.6128577623590634, "grad_norm": 0.24525845050811768, "learning_rate": 2.590309009139707e-06, "loss": 0.0008, "step": 24101 }, { "epoch": 2.612966175195143, "grad_norm": 0.034554366022348404, "learning_rate": 2.5895836355723196e-06, "loss": 0.0002, "step": 24102 }, { "epoch": 2.613074588031223, "grad_norm": 0.2931655943393707, "learning_rate": 2.588858262004933e-06, "loss": 0.0061, "step": 24103 }, { "epoch": 2.613183000867303, "grad_norm": 0.3808499872684479, "learning_rate": 2.5881328884375456e-06, "loss": 0.009, "step": 24104 }, { "epoch": 2.6132914137033825, "grad_norm": 6.094422817230225, "learning_rate": 2.587407514870158e-06, "loss": 0.0285, "step": 24105 }, { "epoch": 2.613399826539462, "grad_norm": 1.1152602434158325, "learning_rate": 2.586682141302771e-06, "loss": 0.0064, "step": 24106 }, { "epoch": 2.613508239375542, "grad_norm": 4.164022445678711, "learning_rate": 2.585956767735384e-06, "loss": 0.0378, "step": 24107 }, { "epoch": 2.613616652211622, "grad_norm": 0.6498210430145264, "learning_rate": 2.5852313941679967e-06, "loss": 0.0016, "step": 24108 }, { "epoch": 2.6137250650477015, "grad_norm": 0.3312827944755554, "learning_rate": 2.5845060206006097e-06, "loss": 0.0116, "step": 24109 }, { "epoch": 2.6138334778837815, "grad_norm": 0.7376329898834229, "learning_rate": 2.5837806470332223e-06, "loss": 0.0267, "step": 24110 }, { "epoch": 2.613941890719861, "grad_norm": 0.30916523933410645, "learning_rate": 2.583055273465835e-06, "loss": 0.0104, "step": 24111 }, { "epoch": 2.614050303555941, "grad_norm": 0.27976539731025696, "learning_rate": 2.5823298998984482e-06, "loss": 0.0009, "step": 24112 }, { "epoch": 2.614158716392021, "grad_norm": 0.33799999952316284, "learning_rate": 2.581604526331061e-06, "loss": 0.0113, "step": 24113 }, { "epoch": 2.6142671292281006, "grad_norm": 2.052123546600342, "learning_rate": 2.5808791527636734e-06, "loss": 0.001, "step": 24114 }, { "epoch": 2.6143755420641805, "grad_norm": 0.34984290599823, "learning_rate": 2.580153779196286e-06, "loss": 0.0075, "step": 24115 }, { "epoch": 2.61448395490026, "grad_norm": 1.5397870540618896, "learning_rate": 2.5794284056288994e-06, "loss": 0.0286, "step": 24116 }, { "epoch": 2.61459236773634, "grad_norm": 0.394821435213089, "learning_rate": 2.578703032061512e-06, "loss": 0.0007, "step": 24117 }, { "epoch": 2.6147007805724196, "grad_norm": 0.005297617521136999, "learning_rate": 2.5779776584941245e-06, "loss": 0.0, "step": 24118 }, { "epoch": 2.6148091934084996, "grad_norm": 0.5262302756309509, "learning_rate": 2.5772522849267375e-06, "loss": 0.014, "step": 24119 }, { "epoch": 2.614917606244579, "grad_norm": 0.49706000089645386, "learning_rate": 2.5765269113593505e-06, "loss": 0.0006, "step": 24120 }, { "epoch": 2.615026019080659, "grad_norm": 3.8471312522888184, "learning_rate": 2.575801537791963e-06, "loss": 0.0169, "step": 24121 }, { "epoch": 2.615134431916739, "grad_norm": 0.1449054628610611, "learning_rate": 2.575076164224576e-06, "loss": 0.0007, "step": 24122 }, { "epoch": 2.6152428447528187, "grad_norm": 0.04543660208582878, "learning_rate": 2.5743507906571886e-06, "loss": 0.0005, "step": 24123 }, { "epoch": 2.6153512575888986, "grad_norm": 0.3216787874698639, "learning_rate": 2.573625417089801e-06, "loss": 0.0212, "step": 24124 }, { "epoch": 2.615459670424978, "grad_norm": 9.847452163696289, "learning_rate": 2.5729000435224146e-06, "loss": 0.0402, "step": 24125 }, { "epoch": 2.615568083261058, "grad_norm": 0.024792784824967384, "learning_rate": 2.572174669955027e-06, "loss": 0.0002, "step": 24126 }, { "epoch": 2.615676496097138, "grad_norm": 0.06437648832798004, "learning_rate": 2.5714492963876397e-06, "loss": 0.0004, "step": 24127 }, { "epoch": 2.6157849089332177, "grad_norm": 4.040193557739258, "learning_rate": 2.5707239228202523e-06, "loss": 0.0488, "step": 24128 }, { "epoch": 2.6158933217692972, "grad_norm": 1.0681483745574951, "learning_rate": 2.5699985492528657e-06, "loss": 0.0225, "step": 24129 }, { "epoch": 2.616001734605377, "grad_norm": 0.027499018236994743, "learning_rate": 2.5692731756854783e-06, "loss": 0.0001, "step": 24130 }, { "epoch": 2.616110147441457, "grad_norm": 0.3946141302585602, "learning_rate": 2.568547802118091e-06, "loss": 0.0057, "step": 24131 }, { "epoch": 2.6162185602775367, "grad_norm": 0.35604459047317505, "learning_rate": 2.567822428550704e-06, "loss": 0.0083, "step": 24132 }, { "epoch": 2.6163269731136167, "grad_norm": 0.04038945212960243, "learning_rate": 2.567097054983317e-06, "loss": 0.0003, "step": 24133 }, { "epoch": 2.6164353859496963, "grad_norm": 0.00711127370595932, "learning_rate": 2.5663716814159294e-06, "loss": 0.0001, "step": 24134 }, { "epoch": 2.6165437987857763, "grad_norm": 0.3113410770893097, "learning_rate": 2.5656463078485424e-06, "loss": 0.0109, "step": 24135 }, { "epoch": 2.6166522116218562, "grad_norm": 0.39512863755226135, "learning_rate": 2.564920934281155e-06, "loss": 0.0012, "step": 24136 }, { "epoch": 2.616760624457936, "grad_norm": 0.3640568256378174, "learning_rate": 2.564195560713768e-06, "loss": 0.0052, "step": 24137 }, { "epoch": 2.6168690372940158, "grad_norm": 0.7710219025611877, "learning_rate": 2.563470187146381e-06, "loss": 0.0356, "step": 24138 }, { "epoch": 2.6169774501300953, "grad_norm": 4.8604888916015625, "learning_rate": 2.5627448135789935e-06, "loss": 0.0083, "step": 24139 }, { "epoch": 2.6170858629661753, "grad_norm": 0.007173067890107632, "learning_rate": 2.562019440011606e-06, "loss": 0.0001, "step": 24140 }, { "epoch": 2.617194275802255, "grad_norm": 7.3423919677734375, "learning_rate": 2.5612940664442186e-06, "loss": 0.0115, "step": 24141 }, { "epoch": 2.617302688638335, "grad_norm": 0.05550263077020645, "learning_rate": 2.560568692876832e-06, "loss": 0.0005, "step": 24142 }, { "epoch": 2.6174111014744144, "grad_norm": 0.09086678922176361, "learning_rate": 2.5598433193094446e-06, "loss": 0.0002, "step": 24143 }, { "epoch": 2.6175195143104943, "grad_norm": 7.224320411682129, "learning_rate": 2.559117945742057e-06, "loss": 0.013, "step": 24144 }, { "epoch": 2.6176279271465743, "grad_norm": 1.080780029296875, "learning_rate": 2.55839257217467e-06, "loss": 0.04, "step": 24145 }, { "epoch": 2.617736339982654, "grad_norm": 0.6363536715507507, "learning_rate": 2.557667198607283e-06, "loss": 0.0012, "step": 24146 }, { "epoch": 2.617844752818734, "grad_norm": 0.022004257887601852, "learning_rate": 2.5569418250398957e-06, "loss": 0.0001, "step": 24147 }, { "epoch": 2.6179531656548134, "grad_norm": 0.11270797252655029, "learning_rate": 2.5562164514725087e-06, "loss": 0.0003, "step": 24148 }, { "epoch": 2.6180615784908934, "grad_norm": 15.68703556060791, "learning_rate": 2.5554910779051213e-06, "loss": 0.0091, "step": 24149 }, { "epoch": 2.6181699913269734, "grad_norm": 0.2410867065191269, "learning_rate": 2.5547657043377343e-06, "loss": 0.0048, "step": 24150 }, { "epoch": 2.618278404163053, "grad_norm": 5.7499470710754395, "learning_rate": 2.5540403307703472e-06, "loss": 0.016, "step": 24151 }, { "epoch": 2.6183868169991324, "grad_norm": 0.47586292028427124, "learning_rate": 2.55331495720296e-06, "loss": 0.0128, "step": 24152 }, { "epoch": 2.6184952298352124, "grad_norm": 0.011142507195472717, "learning_rate": 2.5525895836355724e-06, "loss": 0.0001, "step": 24153 }, { "epoch": 2.6186036426712924, "grad_norm": 0.03459232300519943, "learning_rate": 2.551864210068185e-06, "loss": 0.0003, "step": 24154 }, { "epoch": 2.618712055507372, "grad_norm": 2.791355848312378, "learning_rate": 2.5511388365007984e-06, "loss": 0.0066, "step": 24155 }, { "epoch": 2.618820468343452, "grad_norm": 0.41575440764427185, "learning_rate": 2.550413462933411e-06, "loss": 0.0117, "step": 24156 }, { "epoch": 2.6189288811795315, "grad_norm": 2.8237805366516113, "learning_rate": 2.5496880893660235e-06, "loss": 0.0338, "step": 24157 }, { "epoch": 2.6190372940156115, "grad_norm": 9.899212837219238, "learning_rate": 2.5489627157986365e-06, "loss": 0.016, "step": 24158 }, { "epoch": 2.6191457068516915, "grad_norm": 1.347694993019104, "learning_rate": 2.5482373422312495e-06, "loss": 0.0023, "step": 24159 }, { "epoch": 2.619254119687771, "grad_norm": 0.04065981134772301, "learning_rate": 2.547511968663862e-06, "loss": 0.0003, "step": 24160 }, { "epoch": 2.619362532523851, "grad_norm": 0.1280875951051712, "learning_rate": 2.546786595096475e-06, "loss": 0.0006, "step": 24161 }, { "epoch": 2.6194709453599305, "grad_norm": 9.569982528686523, "learning_rate": 2.5460612215290876e-06, "loss": 0.0211, "step": 24162 }, { "epoch": 2.6195793581960105, "grad_norm": 0.5303994417190552, "learning_rate": 2.5453358479617006e-06, "loss": 0.0073, "step": 24163 }, { "epoch": 2.61968777103209, "grad_norm": 5.477656364440918, "learning_rate": 2.5446104743943136e-06, "loss": 0.007, "step": 24164 }, { "epoch": 2.61979618386817, "grad_norm": 0.028205309063196182, "learning_rate": 2.543885100826926e-06, "loss": 0.0003, "step": 24165 }, { "epoch": 2.6199045967042496, "grad_norm": 3.7720139026641846, "learning_rate": 2.5431597272595387e-06, "loss": 0.0077, "step": 24166 }, { "epoch": 2.6200130095403296, "grad_norm": 1.84148108959198, "learning_rate": 2.5424343536921513e-06, "loss": 0.0204, "step": 24167 }, { "epoch": 2.6201214223764095, "grad_norm": 0.15915893018245697, "learning_rate": 2.5417089801247647e-06, "loss": 0.0009, "step": 24168 }, { "epoch": 2.620229835212489, "grad_norm": 0.34217286109924316, "learning_rate": 2.5409836065573773e-06, "loss": 0.0059, "step": 24169 }, { "epoch": 2.620338248048569, "grad_norm": 0.6081218123435974, "learning_rate": 2.54025823298999e-06, "loss": 0.0185, "step": 24170 }, { "epoch": 2.6204466608846486, "grad_norm": 1.6891577243804932, "learning_rate": 2.539532859422603e-06, "loss": 0.0112, "step": 24171 }, { "epoch": 2.6205550737207286, "grad_norm": 0.1424889862537384, "learning_rate": 2.538807485855216e-06, "loss": 0.0027, "step": 24172 }, { "epoch": 2.6206634865568086, "grad_norm": 0.4925822615623474, "learning_rate": 2.5380821122878284e-06, "loss": 0.0166, "step": 24173 }, { "epoch": 2.620771899392888, "grad_norm": 6.262006759643555, "learning_rate": 2.5373567387204414e-06, "loss": 0.0505, "step": 24174 }, { "epoch": 2.6208803122289677, "grad_norm": 1.599839687347412, "learning_rate": 2.536631365153054e-06, "loss": 0.0169, "step": 24175 }, { "epoch": 2.6209887250650477, "grad_norm": 5.720184803009033, "learning_rate": 2.535905991585667e-06, "loss": 0.0221, "step": 24176 }, { "epoch": 2.6210971379011276, "grad_norm": 0.3172101080417633, "learning_rate": 2.53518061801828e-06, "loss": 0.006, "step": 24177 }, { "epoch": 2.621205550737207, "grad_norm": 8.444334983825684, "learning_rate": 2.5344552444508925e-06, "loss": 0.0252, "step": 24178 }, { "epoch": 2.621313963573287, "grad_norm": 0.4208373725414276, "learning_rate": 2.533729870883505e-06, "loss": 0.001, "step": 24179 }, { "epoch": 2.6214223764093667, "grad_norm": 0.044323358684778214, "learning_rate": 2.5330044973161176e-06, "loss": 0.0003, "step": 24180 }, { "epoch": 2.6215307892454467, "grad_norm": 3.146709680557251, "learning_rate": 2.532279123748731e-06, "loss": 0.017, "step": 24181 }, { "epoch": 2.6216392020815267, "grad_norm": 0.3627013862133026, "learning_rate": 2.5315537501813436e-06, "loss": 0.0084, "step": 24182 }, { "epoch": 2.621747614917606, "grad_norm": 0.39906230568885803, "learning_rate": 2.530828376613956e-06, "loss": 0.0109, "step": 24183 }, { "epoch": 2.621856027753686, "grad_norm": 1.4168004989624023, "learning_rate": 2.530103003046569e-06, "loss": 0.0039, "step": 24184 }, { "epoch": 2.6219644405897657, "grad_norm": 0.046653855592012405, "learning_rate": 2.529377629479182e-06, "loss": 0.0003, "step": 24185 }, { "epoch": 2.6220728534258457, "grad_norm": 0.022421734407544136, "learning_rate": 2.5286522559117947e-06, "loss": 0.0001, "step": 24186 }, { "epoch": 2.6221812662619253, "grad_norm": 0.6252373456954956, "learning_rate": 2.5279268823444077e-06, "loss": 0.0129, "step": 24187 }, { "epoch": 2.6222896790980053, "grad_norm": 0.9783214330673218, "learning_rate": 2.5272015087770203e-06, "loss": 0.0028, "step": 24188 }, { "epoch": 2.622398091934085, "grad_norm": 15.044177055358887, "learning_rate": 2.5264761352096333e-06, "loss": 0.0342, "step": 24189 }, { "epoch": 2.622506504770165, "grad_norm": 0.2551839351654053, "learning_rate": 2.5257507616422462e-06, "loss": 0.0052, "step": 24190 }, { "epoch": 2.6226149176062448, "grad_norm": 0.4551573097705841, "learning_rate": 2.525025388074859e-06, "loss": 0.0086, "step": 24191 }, { "epoch": 2.6227233304423243, "grad_norm": 1.3319721221923828, "learning_rate": 2.5243000145074714e-06, "loss": 0.01, "step": 24192 }, { "epoch": 2.6228317432784043, "grad_norm": 0.016716836020350456, "learning_rate": 2.523574640940084e-06, "loss": 0.0001, "step": 24193 }, { "epoch": 2.622940156114484, "grad_norm": 0.29900503158569336, "learning_rate": 2.5228492673726974e-06, "loss": 0.0065, "step": 24194 }, { "epoch": 2.623048568950564, "grad_norm": 0.230332151055336, "learning_rate": 2.52212389380531e-06, "loss": 0.0169, "step": 24195 }, { "epoch": 2.623156981786644, "grad_norm": 5.485071182250977, "learning_rate": 2.5213985202379225e-06, "loss": 0.0143, "step": 24196 }, { "epoch": 2.6232653946227233, "grad_norm": 0.3530135452747345, "learning_rate": 2.5206731466705355e-06, "loss": 0.0053, "step": 24197 }, { "epoch": 2.623373807458803, "grad_norm": 0.01935829408466816, "learning_rate": 2.5199477731031485e-06, "loss": 0.0002, "step": 24198 }, { "epoch": 2.623482220294883, "grad_norm": 5.812739372253418, "learning_rate": 2.519222399535761e-06, "loss": 0.0171, "step": 24199 }, { "epoch": 2.623590633130963, "grad_norm": 0.36721786856651306, "learning_rate": 2.518497025968374e-06, "loss": 0.0065, "step": 24200 }, { "epoch": 2.6236990459670424, "grad_norm": 5.2116265296936035, "learning_rate": 2.5177716524009866e-06, "loss": 0.0669, "step": 24201 }, { "epoch": 2.6238074588031224, "grad_norm": 0.2973146438598633, "learning_rate": 2.5170462788335996e-06, "loss": 0.0106, "step": 24202 }, { "epoch": 2.623915871639202, "grad_norm": 0.004848301876336336, "learning_rate": 2.5163209052662126e-06, "loss": 0.0001, "step": 24203 }, { "epoch": 2.624024284475282, "grad_norm": 0.014626632444560528, "learning_rate": 2.515595531698825e-06, "loss": 0.0002, "step": 24204 }, { "epoch": 2.624132697311362, "grad_norm": 0.1788141280412674, "learning_rate": 2.5148701581314377e-06, "loss": 0.0006, "step": 24205 }, { "epoch": 2.6242411101474414, "grad_norm": 0.9513466954231262, "learning_rate": 2.5141447845640503e-06, "loss": 0.0067, "step": 24206 }, { "epoch": 2.624349522983521, "grad_norm": 0.024981871247291565, "learning_rate": 2.5134194109966637e-06, "loss": 0.0003, "step": 24207 }, { "epoch": 2.624457935819601, "grad_norm": 9.130691528320312, "learning_rate": 2.5126940374292763e-06, "loss": 0.0178, "step": 24208 }, { "epoch": 2.624566348655681, "grad_norm": 0.025899527594447136, "learning_rate": 2.511968663861889e-06, "loss": 0.0002, "step": 24209 }, { "epoch": 2.6246747614917605, "grad_norm": 0.2150902897119522, "learning_rate": 2.511243290294502e-06, "loss": 0.0042, "step": 24210 }, { "epoch": 2.6247831743278405, "grad_norm": 0.17505863308906555, "learning_rate": 2.510517916727115e-06, "loss": 0.0005, "step": 24211 }, { "epoch": 2.62489158716392, "grad_norm": 0.04363631084561348, "learning_rate": 2.5097925431597274e-06, "loss": 0.0004, "step": 24212 }, { "epoch": 2.625, "grad_norm": 4.987347602844238, "learning_rate": 2.5090671695923404e-06, "loss": 0.0054, "step": 24213 }, { "epoch": 2.62510841283608, "grad_norm": 6.723959922790527, "learning_rate": 2.508341796024953e-06, "loss": 0.071, "step": 24214 }, { "epoch": 2.6252168256721595, "grad_norm": 4.8968000411987305, "learning_rate": 2.507616422457566e-06, "loss": 0.0231, "step": 24215 }, { "epoch": 2.6253252385082395, "grad_norm": 1.3704187870025635, "learning_rate": 2.506891048890179e-06, "loss": 0.0165, "step": 24216 }, { "epoch": 2.625433651344319, "grad_norm": 0.5483052134513855, "learning_rate": 2.5061656753227915e-06, "loss": 0.0151, "step": 24217 }, { "epoch": 2.625542064180399, "grad_norm": 3.9882516860961914, "learning_rate": 2.505440301755404e-06, "loss": 0.0522, "step": 24218 }, { "epoch": 2.625650477016479, "grad_norm": 1.766066074371338, "learning_rate": 2.5047149281880166e-06, "loss": 0.0266, "step": 24219 }, { "epoch": 2.6257588898525586, "grad_norm": 0.4003036916255951, "learning_rate": 2.50398955462063e-06, "loss": 0.0115, "step": 24220 }, { "epoch": 2.625867302688638, "grad_norm": 0.15925228595733643, "learning_rate": 2.5032641810532426e-06, "loss": 0.0006, "step": 24221 }, { "epoch": 2.625975715524718, "grad_norm": 4.111842632293701, "learning_rate": 2.502538807485855e-06, "loss": 0.019, "step": 24222 }, { "epoch": 2.626084128360798, "grad_norm": 0.45663338899612427, "learning_rate": 2.501813433918468e-06, "loss": 0.0134, "step": 24223 }, { "epoch": 2.6261925411968776, "grad_norm": 0.45147597789764404, "learning_rate": 2.501088060351081e-06, "loss": 0.007, "step": 24224 }, { "epoch": 2.6263009540329576, "grad_norm": 0.004491782747209072, "learning_rate": 2.5003626867836937e-06, "loss": 0.0, "step": 24225 }, { "epoch": 2.626409366869037, "grad_norm": 0.038406115025281906, "learning_rate": 2.4996373132163067e-06, "loss": 0.0002, "step": 24226 }, { "epoch": 2.626517779705117, "grad_norm": 0.551034152507782, "learning_rate": 2.4989119396489197e-06, "loss": 0.0159, "step": 24227 }, { "epoch": 2.626626192541197, "grad_norm": 5.137353897094727, "learning_rate": 2.4981865660815323e-06, "loss": 0.0113, "step": 24228 }, { "epoch": 2.6267346053772767, "grad_norm": 6.01547384262085, "learning_rate": 2.4974611925141452e-06, "loss": 0.0081, "step": 24229 }, { "epoch": 2.626843018213356, "grad_norm": 0.39741256833076477, "learning_rate": 2.496735818946758e-06, "loss": 0.0224, "step": 24230 }, { "epoch": 2.626951431049436, "grad_norm": 0.06332150846719742, "learning_rate": 2.496010445379371e-06, "loss": 0.0005, "step": 24231 }, { "epoch": 2.627059843885516, "grad_norm": 0.3365873098373413, "learning_rate": 2.4952850718119834e-06, "loss": 0.0117, "step": 24232 }, { "epoch": 2.6271682567215957, "grad_norm": 0.0044496627524495125, "learning_rate": 2.494559698244596e-06, "loss": 0.0, "step": 24233 }, { "epoch": 2.6272766695576757, "grad_norm": 0.03510963171720505, "learning_rate": 2.493834324677209e-06, "loss": 0.0002, "step": 24234 }, { "epoch": 2.6273850823937552, "grad_norm": 0.9121500253677368, "learning_rate": 2.4931089511098215e-06, "loss": 0.0012, "step": 24235 }, { "epoch": 2.627493495229835, "grad_norm": 0.05583900958299637, "learning_rate": 2.4923835775424345e-06, "loss": 0.0005, "step": 24236 }, { "epoch": 2.627601908065915, "grad_norm": 0.3125274181365967, "learning_rate": 2.491658203975047e-06, "loss": 0.0108, "step": 24237 }, { "epoch": 2.6277103209019947, "grad_norm": 0.1620914489030838, "learning_rate": 2.49093283040766e-06, "loss": 0.0003, "step": 24238 }, { "epoch": 2.6278187337380747, "grad_norm": 0.060308318585157394, "learning_rate": 2.490207456840273e-06, "loss": 0.0003, "step": 24239 }, { "epoch": 2.6279271465741543, "grad_norm": 0.2605665624141693, "learning_rate": 2.489482083272886e-06, "loss": 0.0052, "step": 24240 }, { "epoch": 2.6280355594102343, "grad_norm": 0.34216001629829407, "learning_rate": 2.4887567097054986e-06, "loss": 0.0115, "step": 24241 }, { "epoch": 2.628143972246314, "grad_norm": 0.3603576123714447, "learning_rate": 2.4880313361381116e-06, "loss": 0.0088, "step": 24242 }, { "epoch": 2.628252385082394, "grad_norm": 0.029124578461050987, "learning_rate": 2.487305962570724e-06, "loss": 0.0002, "step": 24243 }, { "epoch": 2.6283607979184733, "grad_norm": 0.03305904567241669, "learning_rate": 2.486580589003337e-06, "loss": 0.0003, "step": 24244 }, { "epoch": 2.6284692107545533, "grad_norm": 0.259236216545105, "learning_rate": 2.4858552154359497e-06, "loss": 0.0042, "step": 24245 }, { "epoch": 2.6285776235906333, "grad_norm": 3.533000946044922, "learning_rate": 2.4851298418685623e-06, "loss": 0.0832, "step": 24246 }, { "epoch": 2.628686036426713, "grad_norm": 1.8688645362854004, "learning_rate": 2.4844044683011753e-06, "loss": 0.0048, "step": 24247 }, { "epoch": 2.628794449262793, "grad_norm": 0.10543859004974365, "learning_rate": 2.483679094733788e-06, "loss": 0.0007, "step": 24248 }, { "epoch": 2.6289028620988724, "grad_norm": 0.6734381318092346, "learning_rate": 2.482953721166401e-06, "loss": 0.0114, "step": 24249 }, { "epoch": 2.6290112749349523, "grad_norm": 0.051496490836143494, "learning_rate": 2.4822283475990134e-06, "loss": 0.0004, "step": 24250 }, { "epoch": 2.6291196877710323, "grad_norm": 0.050584036856889725, "learning_rate": 2.4815029740316264e-06, "loss": 0.0003, "step": 24251 }, { "epoch": 2.629228100607112, "grad_norm": 0.2979300022125244, "learning_rate": 2.4807776004642394e-06, "loss": 0.0111, "step": 24252 }, { "epoch": 2.6293365134431914, "grad_norm": 0.2901332974433899, "learning_rate": 2.4800522268968524e-06, "loss": 0.001, "step": 24253 }, { "epoch": 2.6294449262792714, "grad_norm": 0.23335689306259155, "learning_rate": 2.479326853329465e-06, "loss": 0.0046, "step": 24254 }, { "epoch": 2.6295533391153514, "grad_norm": 0.44645166397094727, "learning_rate": 2.478601479762078e-06, "loss": 0.0067, "step": 24255 }, { "epoch": 2.629661751951431, "grad_norm": 2.476071834564209, "learning_rate": 2.4778761061946905e-06, "loss": 0.0296, "step": 24256 }, { "epoch": 2.629770164787511, "grad_norm": 0.475180983543396, "learning_rate": 2.4771507326273035e-06, "loss": 0.0189, "step": 24257 }, { "epoch": 2.6298785776235905, "grad_norm": 0.043666016310453415, "learning_rate": 2.476425359059916e-06, "loss": 0.0003, "step": 24258 }, { "epoch": 2.6299869904596704, "grad_norm": 0.5641672015190125, "learning_rate": 2.475699985492529e-06, "loss": 0.0149, "step": 24259 }, { "epoch": 2.6300954032957504, "grad_norm": 0.7962584495544434, "learning_rate": 2.4749746119251416e-06, "loss": 0.0101, "step": 24260 }, { "epoch": 2.63020381613183, "grad_norm": 0.3634353578090668, "learning_rate": 2.474249238357754e-06, "loss": 0.0011, "step": 24261 }, { "epoch": 2.63031222896791, "grad_norm": 2.2252888679504395, "learning_rate": 2.473523864790367e-06, "loss": 0.0723, "step": 24262 }, { "epoch": 2.6304206418039895, "grad_norm": 0.3799766004085541, "learning_rate": 2.47279849122298e-06, "loss": 0.011, "step": 24263 }, { "epoch": 2.6305290546400695, "grad_norm": 1.4323347806930542, "learning_rate": 2.4720731176555927e-06, "loss": 0.0008, "step": 24264 }, { "epoch": 2.630637467476149, "grad_norm": 11.076300621032715, "learning_rate": 2.4713477440882057e-06, "loss": 0.0269, "step": 24265 }, { "epoch": 2.630745880312229, "grad_norm": 0.32566213607788086, "learning_rate": 2.4706223705208187e-06, "loss": 0.005, "step": 24266 }, { "epoch": 2.6308542931483085, "grad_norm": 4.150562286376953, "learning_rate": 2.4698969969534313e-06, "loss": 0.0052, "step": 24267 }, { "epoch": 2.6309627059843885, "grad_norm": 0.35612431168556213, "learning_rate": 2.4691716233860443e-06, "loss": 0.012, "step": 24268 }, { "epoch": 2.6310711188204685, "grad_norm": 0.20335325598716736, "learning_rate": 2.468446249818657e-06, "loss": 0.0209, "step": 24269 }, { "epoch": 2.631179531656548, "grad_norm": 0.6326009035110474, "learning_rate": 2.46772087625127e-06, "loss": 0.0167, "step": 24270 }, { "epoch": 2.631287944492628, "grad_norm": 0.38297131657600403, "learning_rate": 2.4669955026838824e-06, "loss": 0.0055, "step": 24271 }, { "epoch": 2.6313963573287076, "grad_norm": 0.32722780108451843, "learning_rate": 2.4662701291164954e-06, "loss": 0.0109, "step": 24272 }, { "epoch": 2.6315047701647876, "grad_norm": 0.03997565805912018, "learning_rate": 2.465544755549108e-06, "loss": 0.0002, "step": 24273 }, { "epoch": 2.6316131830008676, "grad_norm": 0.34733039140701294, "learning_rate": 2.4648193819817205e-06, "loss": 0.0056, "step": 24274 }, { "epoch": 2.631721595836947, "grad_norm": 0.020104622468352318, "learning_rate": 2.4640940084143335e-06, "loss": 0.0002, "step": 24275 }, { "epoch": 2.6318300086730266, "grad_norm": 0.4852427542209625, "learning_rate": 2.4633686348469465e-06, "loss": 0.012, "step": 24276 }, { "epoch": 2.6319384215091066, "grad_norm": 0.018428420647978783, "learning_rate": 2.462643261279559e-06, "loss": 0.0001, "step": 24277 }, { "epoch": 2.6320468343451866, "grad_norm": 9.957509994506836, "learning_rate": 2.461917887712172e-06, "loss": 0.0207, "step": 24278 }, { "epoch": 2.632155247181266, "grad_norm": 0.029710784554481506, "learning_rate": 2.461192514144785e-06, "loss": 0.0002, "step": 24279 }, { "epoch": 2.632263660017346, "grad_norm": 1.8883626461029053, "learning_rate": 2.4604671405773976e-06, "loss": 0.0461, "step": 24280 }, { "epoch": 2.6323720728534257, "grad_norm": 0.289333313703537, "learning_rate": 2.4597417670100106e-06, "loss": 0.0097, "step": 24281 }, { "epoch": 2.6324804856895057, "grad_norm": 0.10042571276426315, "learning_rate": 2.459016393442623e-06, "loss": 0.0002, "step": 24282 }, { "epoch": 2.6325888985255856, "grad_norm": 0.2617857754230499, "learning_rate": 2.458291019875236e-06, "loss": 0.0045, "step": 24283 }, { "epoch": 2.632697311361665, "grad_norm": 10.395092010498047, "learning_rate": 2.4575656463078487e-06, "loss": 0.0121, "step": 24284 }, { "epoch": 2.632805724197745, "grad_norm": 0.6374582648277283, "learning_rate": 2.4568402727404617e-06, "loss": 0.0223, "step": 24285 }, { "epoch": 2.6329141370338247, "grad_norm": 0.8194262385368347, "learning_rate": 2.4561148991730743e-06, "loss": 0.0228, "step": 24286 }, { "epoch": 2.6330225498699047, "grad_norm": 0.21896153688430786, "learning_rate": 2.455389525605687e-06, "loss": 0.0038, "step": 24287 }, { "epoch": 2.6331309627059842, "grad_norm": 0.1174774169921875, "learning_rate": 2.4546641520383e-06, "loss": 0.0007, "step": 24288 }, { "epoch": 2.633239375542064, "grad_norm": 0.548514187335968, "learning_rate": 2.453938778470913e-06, "loss": 0.012, "step": 24289 }, { "epoch": 2.6333477883781438, "grad_norm": 0.3249624967575073, "learning_rate": 2.4532134049035254e-06, "loss": 0.0116, "step": 24290 }, { "epoch": 2.6334562012142237, "grad_norm": 7.863321781158447, "learning_rate": 2.4524880313361384e-06, "loss": 0.0114, "step": 24291 }, { "epoch": 2.6335646140503037, "grad_norm": 0.5857914686203003, "learning_rate": 2.4517626577687514e-06, "loss": 0.0155, "step": 24292 }, { "epoch": 2.6336730268863833, "grad_norm": 0.3560102880001068, "learning_rate": 2.451037284201364e-06, "loss": 0.0127, "step": 24293 }, { "epoch": 2.6337814397224633, "grad_norm": 0.03842625021934509, "learning_rate": 2.450311910633977e-06, "loss": 0.0004, "step": 24294 }, { "epoch": 2.633889852558543, "grad_norm": 0.37948426604270935, "learning_rate": 2.4495865370665895e-06, "loss": 0.0009, "step": 24295 }, { "epoch": 2.633998265394623, "grad_norm": 0.6326211094856262, "learning_rate": 2.4488611634992025e-06, "loss": 0.0068, "step": 24296 }, { "epoch": 2.6341066782307028, "grad_norm": 7.192636966705322, "learning_rate": 2.448135789931815e-06, "loss": 0.0083, "step": 24297 }, { "epoch": 2.6342150910667823, "grad_norm": 0.2976383864879608, "learning_rate": 2.447410416364428e-06, "loss": 0.0061, "step": 24298 }, { "epoch": 2.634323503902862, "grad_norm": 0.3846468925476074, "learning_rate": 2.4466850427970406e-06, "loss": 0.0212, "step": 24299 }, { "epoch": 2.634431916738942, "grad_norm": 0.06970666348934174, "learning_rate": 2.445959669229653e-06, "loss": 0.0003, "step": 24300 }, { "epoch": 2.634540329575022, "grad_norm": 0.5358455777168274, "learning_rate": 2.445234295662266e-06, "loss": 0.0072, "step": 24301 }, { "epoch": 2.6346487424111014, "grad_norm": 8.509318351745605, "learning_rate": 2.444508922094879e-06, "loss": 0.0317, "step": 24302 }, { "epoch": 2.6347571552471813, "grad_norm": 0.265017569065094, "learning_rate": 2.4437835485274917e-06, "loss": 0.0054, "step": 24303 }, { "epoch": 2.634865568083261, "grad_norm": 2.098806858062744, "learning_rate": 2.4430581749601047e-06, "loss": 0.0605, "step": 24304 }, { "epoch": 2.634973980919341, "grad_norm": 0.02941187098622322, "learning_rate": 2.4423328013927177e-06, "loss": 0.0003, "step": 24305 }, { "epoch": 2.635082393755421, "grad_norm": 0.27966198325157166, "learning_rate": 2.4416074278253303e-06, "loss": 0.0058, "step": 24306 }, { "epoch": 2.6351908065915004, "grad_norm": 0.1891005039215088, "learning_rate": 2.4408820542579433e-06, "loss": 0.0153, "step": 24307 }, { "epoch": 2.6352992194275804, "grad_norm": 0.39146530628204346, "learning_rate": 2.440156680690556e-06, "loss": 0.0165, "step": 24308 }, { "epoch": 2.63540763226366, "grad_norm": 0.0048580351285636425, "learning_rate": 2.439431307123169e-06, "loss": 0.0, "step": 24309 }, { "epoch": 2.63551604509974, "grad_norm": 0.38549041748046875, "learning_rate": 2.4387059335557814e-06, "loss": 0.006, "step": 24310 }, { "epoch": 2.6356244579358195, "grad_norm": 3.7506327629089355, "learning_rate": 2.4379805599883944e-06, "loss": 0.0147, "step": 24311 }, { "epoch": 2.6357328707718994, "grad_norm": 0.37272781133651733, "learning_rate": 2.437255186421007e-06, "loss": 0.0121, "step": 24312 }, { "epoch": 2.635841283607979, "grad_norm": 0.08967893570661545, "learning_rate": 2.43652981285362e-06, "loss": 0.0004, "step": 24313 }, { "epoch": 2.635949696444059, "grad_norm": 2.529690742492676, "learning_rate": 2.4358044392862325e-06, "loss": 0.0339, "step": 24314 }, { "epoch": 2.636058109280139, "grad_norm": 4.875465393066406, "learning_rate": 2.4350790657188455e-06, "loss": 0.0216, "step": 24315 }, { "epoch": 2.6361665221162185, "grad_norm": 0.7840391993522644, "learning_rate": 2.434353692151458e-06, "loss": 0.0229, "step": 24316 }, { "epoch": 2.6362749349522985, "grad_norm": 0.773803174495697, "learning_rate": 2.433628318584071e-06, "loss": 0.0007, "step": 24317 }, { "epoch": 2.636383347788378, "grad_norm": 4.289712905883789, "learning_rate": 2.432902945016684e-06, "loss": 0.0475, "step": 24318 }, { "epoch": 2.636491760624458, "grad_norm": 0.01484172884374857, "learning_rate": 2.4321775714492966e-06, "loss": 0.0001, "step": 24319 }, { "epoch": 2.636600173460538, "grad_norm": 13.86800479888916, "learning_rate": 2.4314521978819096e-06, "loss": 0.0112, "step": 24320 }, { "epoch": 2.6367085862966175, "grad_norm": 0.02283240295946598, "learning_rate": 2.430726824314522e-06, "loss": 0.0001, "step": 24321 }, { "epoch": 2.636816999132697, "grad_norm": 0.3182550072669983, "learning_rate": 2.430001450747135e-06, "loss": 0.011, "step": 24322 }, { "epoch": 2.636925411968777, "grad_norm": 3.1013364791870117, "learning_rate": 2.4292760771797477e-06, "loss": 0.0305, "step": 24323 }, { "epoch": 2.637033824804857, "grad_norm": 0.2520797550678253, "learning_rate": 2.4285507036123607e-06, "loss": 0.0049, "step": 24324 }, { "epoch": 2.6371422376409366, "grad_norm": 0.5928369164466858, "learning_rate": 2.4278253300449733e-06, "loss": 0.0134, "step": 24325 }, { "epoch": 2.6372506504770166, "grad_norm": 1.076944351196289, "learning_rate": 2.4270999564775863e-06, "loss": 0.0272, "step": 24326 }, { "epoch": 2.637359063313096, "grad_norm": 0.3759317696094513, "learning_rate": 2.426374582910199e-06, "loss": 0.0101, "step": 24327 }, { "epoch": 2.637467476149176, "grad_norm": 0.5166822075843811, "learning_rate": 2.425649209342812e-06, "loss": 0.0161, "step": 24328 }, { "epoch": 2.637575888985256, "grad_norm": 0.39379024505615234, "learning_rate": 2.4249238357754244e-06, "loss": 0.0091, "step": 24329 }, { "epoch": 2.6376843018213356, "grad_norm": 0.7044351696968079, "learning_rate": 2.4241984622080374e-06, "loss": 0.0007, "step": 24330 }, { "epoch": 2.637792714657415, "grad_norm": 0.3655183017253876, "learning_rate": 2.4234730886406504e-06, "loss": 0.0118, "step": 24331 }, { "epoch": 2.637901127493495, "grad_norm": 0.3081328272819519, "learning_rate": 2.422747715073263e-06, "loss": 0.0056, "step": 24332 }, { "epoch": 2.638009540329575, "grad_norm": 3.768345594406128, "learning_rate": 2.422022341505876e-06, "loss": 0.0398, "step": 24333 }, { "epoch": 2.6381179531656547, "grad_norm": 0.09640059620141983, "learning_rate": 2.4212969679384885e-06, "loss": 0.0004, "step": 24334 }, { "epoch": 2.6382263660017347, "grad_norm": 0.5528615117073059, "learning_rate": 2.4205715943711015e-06, "loss": 0.0017, "step": 24335 }, { "epoch": 2.638334778837814, "grad_norm": 1.6979060173034668, "learning_rate": 2.419846220803714e-06, "loss": 0.0019, "step": 24336 }, { "epoch": 2.638443191673894, "grad_norm": 0.16179826855659485, "learning_rate": 2.419120847236327e-06, "loss": 0.0003, "step": 24337 }, { "epoch": 2.638551604509974, "grad_norm": 0.3155231475830078, "learning_rate": 2.4183954736689396e-06, "loss": 0.0066, "step": 24338 }, { "epoch": 2.6386600173460537, "grad_norm": 0.3157414495944977, "learning_rate": 2.4176701001015526e-06, "loss": 0.0109, "step": 24339 }, { "epoch": 2.6387684301821337, "grad_norm": 0.8040774464607239, "learning_rate": 2.416944726534165e-06, "loss": 0.0081, "step": 24340 }, { "epoch": 2.6388768430182132, "grad_norm": 22.194625854492188, "learning_rate": 2.416219352966778e-06, "loss": 0.0316, "step": 24341 }, { "epoch": 2.638985255854293, "grad_norm": 0.08580157160758972, "learning_rate": 2.4154939793993907e-06, "loss": 0.0105, "step": 24342 }, { "epoch": 2.639093668690373, "grad_norm": 0.7564467191696167, "learning_rate": 2.4147686058320037e-06, "loss": 0.0292, "step": 24343 }, { "epoch": 2.6392020815264527, "grad_norm": 0.34956100583076477, "learning_rate": 2.4140432322646167e-06, "loss": 0.0061, "step": 24344 }, { "epoch": 2.6393104943625323, "grad_norm": 5.085893154144287, "learning_rate": 2.4133178586972293e-06, "loss": 0.042, "step": 24345 }, { "epoch": 2.6394189071986123, "grad_norm": 0.06021181493997574, "learning_rate": 2.4125924851298423e-06, "loss": 0.0003, "step": 24346 }, { "epoch": 2.6395273200346923, "grad_norm": 0.44895777106285095, "learning_rate": 2.411867111562455e-06, "loss": 0.0008, "step": 24347 }, { "epoch": 2.639635732870772, "grad_norm": 0.38044899702072144, "learning_rate": 2.411141737995068e-06, "loss": 0.0052, "step": 24348 }, { "epoch": 2.639744145706852, "grad_norm": 0.2474626898765564, "learning_rate": 2.4104163644276804e-06, "loss": 0.0049, "step": 24349 }, { "epoch": 2.6398525585429313, "grad_norm": 0.2909284234046936, "learning_rate": 2.4096909908602934e-06, "loss": 0.0016, "step": 24350 }, { "epoch": 2.6399609713790113, "grad_norm": 0.010024100542068481, "learning_rate": 2.408965617292906e-06, "loss": 0.0001, "step": 24351 }, { "epoch": 2.6400693842150913, "grad_norm": 0.04396943747997284, "learning_rate": 2.408240243725519e-06, "loss": 0.0004, "step": 24352 }, { "epoch": 2.640177797051171, "grad_norm": 2.0414164066314697, "learning_rate": 2.4075148701581315e-06, "loss": 0.0666, "step": 24353 }, { "epoch": 2.6402862098872504, "grad_norm": 0.8643946051597595, "learning_rate": 2.4067894965907445e-06, "loss": 0.0064, "step": 24354 }, { "epoch": 2.6403946227233304, "grad_norm": 0.11795343458652496, "learning_rate": 2.406064123023357e-06, "loss": 0.0003, "step": 24355 }, { "epoch": 2.6405030355594103, "grad_norm": 12.635720252990723, "learning_rate": 2.40533874945597e-06, "loss": 0.0217, "step": 24356 }, { "epoch": 2.64061144839549, "grad_norm": 0.8145784735679626, "learning_rate": 2.404613375888583e-06, "loss": 0.0019, "step": 24357 }, { "epoch": 2.64071986123157, "grad_norm": 0.018742635846138, "learning_rate": 2.4038880023211956e-06, "loss": 0.0001, "step": 24358 }, { "epoch": 2.6408282740676494, "grad_norm": 21.70797348022461, "learning_rate": 2.4031626287538086e-06, "loss": 0.047, "step": 24359 }, { "epoch": 2.6409366869037294, "grad_norm": 0.2949884235858917, "learning_rate": 2.402437255186421e-06, "loss": 0.0037, "step": 24360 }, { "epoch": 2.6410450997398094, "grad_norm": 0.4547203481197357, "learning_rate": 2.401711881619034e-06, "loss": 0.0165, "step": 24361 }, { "epoch": 2.641153512575889, "grad_norm": 17.36528205871582, "learning_rate": 2.4009865080516467e-06, "loss": 0.0285, "step": 24362 }, { "epoch": 2.641261925411969, "grad_norm": 1.3376264572143555, "learning_rate": 2.4002611344842597e-06, "loss": 0.0518, "step": 24363 }, { "epoch": 2.6413703382480485, "grad_norm": 2.6901743412017822, "learning_rate": 2.3995357609168723e-06, "loss": 0.019, "step": 24364 }, { "epoch": 2.6414787510841284, "grad_norm": 0.030454715713858604, "learning_rate": 2.3988103873494853e-06, "loss": 0.0002, "step": 24365 }, { "epoch": 2.6415871639202084, "grad_norm": 0.3674614131450653, "learning_rate": 2.398085013782098e-06, "loss": 0.0006, "step": 24366 }, { "epoch": 2.641695576756288, "grad_norm": 0.08261851221323013, "learning_rate": 2.397359640214711e-06, "loss": 0.0104, "step": 24367 }, { "epoch": 2.6418039895923675, "grad_norm": 1.3177947998046875, "learning_rate": 2.3966342666473234e-06, "loss": 0.0061, "step": 24368 }, { "epoch": 2.6419124024284475, "grad_norm": 0.013315379619598389, "learning_rate": 2.3959088930799364e-06, "loss": 0.0001, "step": 24369 }, { "epoch": 2.6420208152645275, "grad_norm": 0.3547419309616089, "learning_rate": 2.3951835195125494e-06, "loss": 0.0072, "step": 24370 }, { "epoch": 2.642129228100607, "grad_norm": 0.49567389488220215, "learning_rate": 2.394458145945162e-06, "loss": 0.0021, "step": 24371 }, { "epoch": 2.642237640936687, "grad_norm": 0.3006029427051544, "learning_rate": 2.393732772377775e-06, "loss": 0.0102, "step": 24372 }, { "epoch": 2.6423460537727665, "grad_norm": 4.01624870300293, "learning_rate": 2.3930073988103875e-06, "loss": 0.0184, "step": 24373 }, { "epoch": 2.6424544666088465, "grad_norm": 0.3818841278553009, "learning_rate": 2.3922820252430005e-06, "loss": 0.0065, "step": 24374 }, { "epoch": 2.6425628794449265, "grad_norm": 1.5355514287948608, "learning_rate": 2.391556651675613e-06, "loss": 0.0548, "step": 24375 }, { "epoch": 2.642671292281006, "grad_norm": 1.4970078468322754, "learning_rate": 2.390831278108226e-06, "loss": 0.0381, "step": 24376 }, { "epoch": 2.6427797051170856, "grad_norm": 0.016512244939804077, "learning_rate": 2.3901059045408386e-06, "loss": 0.0002, "step": 24377 }, { "epoch": 2.6428881179531656, "grad_norm": 0.3361040949821472, "learning_rate": 2.3893805309734516e-06, "loss": 0.0073, "step": 24378 }, { "epoch": 2.6429965307892456, "grad_norm": 0.31537148356437683, "learning_rate": 2.388655157406064e-06, "loss": 0.0114, "step": 24379 }, { "epoch": 2.643104943625325, "grad_norm": 0.23976533114910126, "learning_rate": 2.387929783838677e-06, "loss": 0.0046, "step": 24380 }, { "epoch": 2.643213356461405, "grad_norm": 0.026847653090953827, "learning_rate": 2.3872044102712897e-06, "loss": 0.0003, "step": 24381 }, { "epoch": 2.6433217692974846, "grad_norm": 1.8328020572662354, "learning_rate": 2.3864790367039027e-06, "loss": 0.007, "step": 24382 }, { "epoch": 2.6434301821335646, "grad_norm": 1.2931753396987915, "learning_rate": 2.3857536631365157e-06, "loss": 0.0081, "step": 24383 }, { "epoch": 2.6435385949696446, "grad_norm": 0.33089539408683777, "learning_rate": 2.3850282895691283e-06, "loss": 0.0126, "step": 24384 }, { "epoch": 2.643647007805724, "grad_norm": 0.07772835344076157, "learning_rate": 2.3843029160017413e-06, "loss": 0.0006, "step": 24385 }, { "epoch": 2.643755420641804, "grad_norm": 0.7379076480865479, "learning_rate": 2.383577542434354e-06, "loss": 0.0013, "step": 24386 }, { "epoch": 2.6438638334778837, "grad_norm": 0.3017495572566986, "learning_rate": 2.382852168866967e-06, "loss": 0.0071, "step": 24387 }, { "epoch": 2.6439722463139637, "grad_norm": 8.554425239562988, "learning_rate": 2.3821267952995794e-06, "loss": 0.0345, "step": 24388 }, { "epoch": 2.644080659150043, "grad_norm": 1.2891546487808228, "learning_rate": 2.3814014217321924e-06, "loss": 0.0354, "step": 24389 }, { "epoch": 2.644189071986123, "grad_norm": 0.7285446524620056, "learning_rate": 2.380676048164805e-06, "loss": 0.0164, "step": 24390 }, { "epoch": 2.6442974848222027, "grad_norm": 0.01759781315922737, "learning_rate": 2.379950674597418e-06, "loss": 0.0002, "step": 24391 }, { "epoch": 2.6444058976582827, "grad_norm": 1.6404023170471191, "learning_rate": 2.3792253010300305e-06, "loss": 0.0018, "step": 24392 }, { "epoch": 2.6445143104943627, "grad_norm": 0.004780495073646307, "learning_rate": 2.3784999274626435e-06, "loss": 0.0, "step": 24393 }, { "epoch": 2.6446227233304422, "grad_norm": 0.5062457919120789, "learning_rate": 2.377774553895256e-06, "loss": 0.0008, "step": 24394 }, { "epoch": 2.6447311361665222, "grad_norm": 0.026730278506875038, "learning_rate": 2.377049180327869e-06, "loss": 0.0002, "step": 24395 }, { "epoch": 2.6448395490026018, "grad_norm": 1.6487177610397339, "learning_rate": 2.376323806760482e-06, "loss": 0.0113, "step": 24396 }, { "epoch": 2.6449479618386817, "grad_norm": 0.7204189300537109, "learning_rate": 2.3755984331930946e-06, "loss": 0.0212, "step": 24397 }, { "epoch": 2.6450563746747617, "grad_norm": 5.2846760749816895, "learning_rate": 2.3748730596257076e-06, "loss": 0.0396, "step": 24398 }, { "epoch": 2.6451647875108413, "grad_norm": 0.02763494849205017, "learning_rate": 2.37414768605832e-06, "loss": 0.0002, "step": 24399 }, { "epoch": 2.645273200346921, "grad_norm": 0.3250112533569336, "learning_rate": 2.373422312490933e-06, "loss": 0.0042, "step": 24400 }, { "epoch": 2.645381613183001, "grad_norm": 0.2918451428413391, "learning_rate": 2.3726969389235457e-06, "loss": 0.0046, "step": 24401 }, { "epoch": 2.645490026019081, "grad_norm": 0.7282121181488037, "learning_rate": 2.3719715653561587e-06, "loss": 0.0253, "step": 24402 }, { "epoch": 2.6455984388551603, "grad_norm": 4.390984535217285, "learning_rate": 2.3712461917887713e-06, "loss": 0.0028, "step": 24403 }, { "epoch": 2.6457068516912403, "grad_norm": 0.3613073229789734, "learning_rate": 2.3705208182213843e-06, "loss": 0.0107, "step": 24404 }, { "epoch": 2.64581526452732, "grad_norm": 0.3011612892150879, "learning_rate": 2.369795444653997e-06, "loss": 0.0096, "step": 24405 }, { "epoch": 2.6459236773634, "grad_norm": 0.4439345598220825, "learning_rate": 2.36907007108661e-06, "loss": 0.0115, "step": 24406 }, { "epoch": 2.64603209019948, "grad_norm": 3.374025821685791, "learning_rate": 2.3683446975192224e-06, "loss": 0.0261, "step": 24407 }, { "epoch": 2.6461405030355594, "grad_norm": 4.674144268035889, "learning_rate": 2.3676193239518354e-06, "loss": 0.0037, "step": 24408 }, { "epoch": 2.6462489158716394, "grad_norm": 0.12998418509960175, "learning_rate": 2.3668939503844484e-06, "loss": 0.0016, "step": 24409 }, { "epoch": 2.646357328707719, "grad_norm": 0.04824768006801605, "learning_rate": 2.366168576817061e-06, "loss": 0.0003, "step": 24410 }, { "epoch": 2.646465741543799, "grad_norm": 3.017827033996582, "learning_rate": 2.365443203249674e-06, "loss": 0.0163, "step": 24411 }, { "epoch": 2.6465741543798784, "grad_norm": 1.3003123998641968, "learning_rate": 2.3647178296822865e-06, "loss": 0.0078, "step": 24412 }, { "epoch": 2.6466825672159584, "grad_norm": 0.11546196788549423, "learning_rate": 2.3639924561148995e-06, "loss": 0.0005, "step": 24413 }, { "epoch": 2.646790980052038, "grad_norm": 0.5230143070220947, "learning_rate": 2.363267082547512e-06, "loss": 0.0212, "step": 24414 }, { "epoch": 2.646899392888118, "grad_norm": 0.3892141878604889, "learning_rate": 2.362541708980125e-06, "loss": 0.0144, "step": 24415 }, { "epoch": 2.647007805724198, "grad_norm": 4.624504089355469, "learning_rate": 2.3618163354127376e-06, "loss": 0.0206, "step": 24416 }, { "epoch": 2.6471162185602775, "grad_norm": 0.041018031537532806, "learning_rate": 2.3610909618453506e-06, "loss": 0.0003, "step": 24417 }, { "epoch": 2.6472246313963574, "grad_norm": 2.7176408767700195, "learning_rate": 2.360365588277963e-06, "loss": 0.0151, "step": 24418 }, { "epoch": 2.647333044232437, "grad_norm": 0.30811527371406555, "learning_rate": 2.359640214710576e-06, "loss": 0.0016, "step": 24419 }, { "epoch": 2.647441457068517, "grad_norm": 10.466268539428711, "learning_rate": 2.3589148411431887e-06, "loss": 0.0085, "step": 24420 }, { "epoch": 2.647549869904597, "grad_norm": 0.01206228043884039, "learning_rate": 2.3581894675758017e-06, "loss": 0.0001, "step": 24421 }, { "epoch": 2.6476582827406765, "grad_norm": 0.28869590163230896, "learning_rate": 2.3574640940084147e-06, "loss": 0.0058, "step": 24422 }, { "epoch": 2.647766695576756, "grad_norm": 0.38331836462020874, "learning_rate": 2.3567387204410273e-06, "loss": 0.0081, "step": 24423 }, { "epoch": 2.647875108412836, "grad_norm": 0.11650841683149338, "learning_rate": 2.3560133468736403e-06, "loss": 0.0003, "step": 24424 }, { "epoch": 2.647983521248916, "grad_norm": 0.14167162775993347, "learning_rate": 2.355287973306253e-06, "loss": 0.0005, "step": 24425 }, { "epoch": 2.6480919340849955, "grad_norm": 3.1762197017669678, "learning_rate": 2.354562599738866e-06, "loss": 0.0044, "step": 24426 }, { "epoch": 2.6482003469210755, "grad_norm": 0.42351818084716797, "learning_rate": 2.3538372261714784e-06, "loss": 0.0185, "step": 24427 }, { "epoch": 2.648308759757155, "grad_norm": 0.13044139742851257, "learning_rate": 2.3531118526040914e-06, "loss": 0.0003, "step": 24428 }, { "epoch": 2.648417172593235, "grad_norm": 0.007270528934895992, "learning_rate": 2.352386479036704e-06, "loss": 0.0001, "step": 24429 }, { "epoch": 2.648525585429315, "grad_norm": 0.05636081472039223, "learning_rate": 2.351661105469317e-06, "loss": 0.0004, "step": 24430 }, { "epoch": 2.6486339982653946, "grad_norm": 0.07815972715616226, "learning_rate": 2.3509357319019295e-06, "loss": 0.0005, "step": 24431 }, { "epoch": 2.6487424111014746, "grad_norm": 0.7814497351646423, "learning_rate": 2.3502103583345425e-06, "loss": 0.0017, "step": 24432 }, { "epoch": 2.648850823937554, "grad_norm": 0.32637152075767517, "learning_rate": 2.349484984767155e-06, "loss": 0.0086, "step": 24433 }, { "epoch": 2.648959236773634, "grad_norm": 0.37539124488830566, "learning_rate": 2.348759611199768e-06, "loss": 0.0116, "step": 24434 }, { "epoch": 2.6490676496097136, "grad_norm": 0.005529380403459072, "learning_rate": 2.348034237632381e-06, "loss": 0.0001, "step": 24435 }, { "epoch": 2.6491760624457936, "grad_norm": 0.26609405875205994, "learning_rate": 2.3473088640649936e-06, "loss": 0.0056, "step": 24436 }, { "epoch": 2.649284475281873, "grad_norm": 0.5693661570549011, "learning_rate": 2.3465834904976066e-06, "loss": 0.0009, "step": 24437 }, { "epoch": 2.649392888117953, "grad_norm": 6.4268412590026855, "learning_rate": 2.345858116930219e-06, "loss": 0.0087, "step": 24438 }, { "epoch": 2.649501300954033, "grad_norm": 1.4208985567092896, "learning_rate": 2.345132743362832e-06, "loss": 0.0282, "step": 24439 }, { "epoch": 2.6496097137901127, "grad_norm": 0.009460322558879852, "learning_rate": 2.3444073697954447e-06, "loss": 0.0, "step": 24440 }, { "epoch": 2.6497181266261927, "grad_norm": 0.8140618205070496, "learning_rate": 2.3436819962280577e-06, "loss": 0.0021, "step": 24441 }, { "epoch": 2.649826539462272, "grad_norm": 0.03886709362268448, "learning_rate": 2.3429566226606703e-06, "loss": 0.0003, "step": 24442 }, { "epoch": 2.649934952298352, "grad_norm": 0.5043224692344666, "learning_rate": 2.3422312490932833e-06, "loss": 0.0119, "step": 24443 }, { "epoch": 2.650043365134432, "grad_norm": 0.618647038936615, "learning_rate": 2.341505875525896e-06, "loss": 0.0016, "step": 24444 }, { "epoch": 2.6501517779705117, "grad_norm": 0.7148666977882385, "learning_rate": 2.340780501958509e-06, "loss": 0.0127, "step": 24445 }, { "epoch": 2.6502601908065913, "grad_norm": 0.015281247906386852, "learning_rate": 2.3400551283911214e-06, "loss": 0.0001, "step": 24446 }, { "epoch": 2.6503686036426712, "grad_norm": 0.3222813010215759, "learning_rate": 2.3393297548237344e-06, "loss": 0.012, "step": 24447 }, { "epoch": 2.6504770164787512, "grad_norm": 0.1175292581319809, "learning_rate": 2.3386043812563474e-06, "loss": 0.0107, "step": 24448 }, { "epoch": 2.6505854293148308, "grad_norm": 0.008409762755036354, "learning_rate": 2.33787900768896e-06, "loss": 0.0001, "step": 24449 }, { "epoch": 2.6506938421509108, "grad_norm": 0.43866682052612305, "learning_rate": 2.337153634121573e-06, "loss": 0.0133, "step": 24450 }, { "epoch": 2.6508022549869903, "grad_norm": 0.016096940264105797, "learning_rate": 2.3364282605541855e-06, "loss": 0.0001, "step": 24451 }, { "epoch": 2.6509106678230703, "grad_norm": 3.2644505500793457, "learning_rate": 2.3357028869867985e-06, "loss": 0.0186, "step": 24452 }, { "epoch": 2.6510190806591503, "grad_norm": 0.6370330452919006, "learning_rate": 2.334977513419411e-06, "loss": 0.0139, "step": 24453 }, { "epoch": 2.65112749349523, "grad_norm": 0.3326026499271393, "learning_rate": 2.334252139852024e-06, "loss": 0.0071, "step": 24454 }, { "epoch": 2.65123590633131, "grad_norm": 0.40287894010543823, "learning_rate": 2.3335267662846366e-06, "loss": 0.0092, "step": 24455 }, { "epoch": 2.6513443191673893, "grad_norm": 0.4738742709159851, "learning_rate": 2.3328013927172496e-06, "loss": 0.0098, "step": 24456 }, { "epoch": 2.6514527320034693, "grad_norm": 0.03973744809627533, "learning_rate": 2.332076019149862e-06, "loss": 0.0002, "step": 24457 }, { "epoch": 2.651561144839549, "grad_norm": 1.1192946434020996, "learning_rate": 2.331350645582475e-06, "loss": 0.0333, "step": 24458 }, { "epoch": 2.651669557675629, "grad_norm": 0.3682730495929718, "learning_rate": 2.3306252720150877e-06, "loss": 0.0059, "step": 24459 }, { "epoch": 2.6517779705117084, "grad_norm": 0.2694711685180664, "learning_rate": 2.3298998984477007e-06, "loss": 0.0055, "step": 24460 }, { "epoch": 2.6518863833477884, "grad_norm": 0.4467030465602875, "learning_rate": 2.3291745248803137e-06, "loss": 0.0084, "step": 24461 }, { "epoch": 2.6519947961838684, "grad_norm": 0.25094538927078247, "learning_rate": 2.3284491513129263e-06, "loss": 0.005, "step": 24462 }, { "epoch": 2.652103209019948, "grad_norm": 0.5120723843574524, "learning_rate": 2.3277237777455393e-06, "loss": 0.0093, "step": 24463 }, { "epoch": 2.652211621856028, "grad_norm": 14.992355346679688, "learning_rate": 2.326998404178152e-06, "loss": 0.0139, "step": 24464 }, { "epoch": 2.6523200346921074, "grad_norm": 0.22652016580104828, "learning_rate": 2.326273030610765e-06, "loss": 0.0048, "step": 24465 }, { "epoch": 2.6524284475281874, "grad_norm": 0.2675471901893616, "learning_rate": 2.3255476570433774e-06, "loss": 0.0056, "step": 24466 }, { "epoch": 2.6525368603642674, "grad_norm": 0.03092825971543789, "learning_rate": 2.3248222834759904e-06, "loss": 0.0003, "step": 24467 }, { "epoch": 2.652645273200347, "grad_norm": 0.6925280094146729, "learning_rate": 2.324096909908603e-06, "loss": 0.0065, "step": 24468 }, { "epoch": 2.6527536860364265, "grad_norm": 3.666724920272827, "learning_rate": 2.323371536341216e-06, "loss": 0.0173, "step": 24469 }, { "epoch": 2.6528620988725065, "grad_norm": 1.7182562351226807, "learning_rate": 2.3226461627738285e-06, "loss": 0.0011, "step": 24470 }, { "epoch": 2.6529705117085864, "grad_norm": 0.042301688343286514, "learning_rate": 2.3219207892064415e-06, "loss": 0.0002, "step": 24471 }, { "epoch": 2.653078924544666, "grad_norm": 0.5258288383483887, "learning_rate": 2.321195415639054e-06, "loss": 0.012, "step": 24472 }, { "epoch": 2.653187337380746, "grad_norm": 0.28799229860305786, "learning_rate": 2.320470042071667e-06, "loss": 0.0066, "step": 24473 }, { "epoch": 2.6532957502168255, "grad_norm": 0.34362301230430603, "learning_rate": 2.31974466850428e-06, "loss": 0.0132, "step": 24474 }, { "epoch": 2.6534041630529055, "grad_norm": 0.39531537890434265, "learning_rate": 2.3190192949368926e-06, "loss": 0.0104, "step": 24475 }, { "epoch": 2.6535125758889855, "grad_norm": 0.05900146812200546, "learning_rate": 2.3182939213695056e-06, "loss": 0.0005, "step": 24476 }, { "epoch": 2.653620988725065, "grad_norm": 0.03528836742043495, "learning_rate": 2.317568547802118e-06, "loss": 0.0003, "step": 24477 }, { "epoch": 2.6537294015611446, "grad_norm": 0.2355594038963318, "learning_rate": 2.316843174234731e-06, "loss": 0.0047, "step": 24478 }, { "epoch": 2.6538378143972245, "grad_norm": 5.794697284698486, "learning_rate": 2.3161178006673437e-06, "loss": 0.0215, "step": 24479 }, { "epoch": 2.6539462272333045, "grad_norm": 0.01120763923972845, "learning_rate": 2.3153924270999567e-06, "loss": 0.0001, "step": 24480 }, { "epoch": 2.654054640069384, "grad_norm": 3.8893654346466064, "learning_rate": 2.3146670535325693e-06, "loss": 0.037, "step": 24481 }, { "epoch": 2.654163052905464, "grad_norm": 0.04382000118494034, "learning_rate": 2.3139416799651823e-06, "loss": 0.0004, "step": 24482 }, { "epoch": 2.6542714657415436, "grad_norm": 0.10410213470458984, "learning_rate": 2.313216306397795e-06, "loss": 0.0005, "step": 24483 }, { "epoch": 2.6543798785776236, "grad_norm": 2.082728624343872, "learning_rate": 2.312490932830408e-06, "loss": 0.0079, "step": 24484 }, { "epoch": 2.6544882914137036, "grad_norm": 0.05109503120183945, "learning_rate": 2.3117655592630204e-06, "loss": 0.0003, "step": 24485 }, { "epoch": 2.654596704249783, "grad_norm": 0.2667815089225769, "learning_rate": 2.3110401856956334e-06, "loss": 0.0009, "step": 24486 }, { "epoch": 2.654705117085863, "grad_norm": 0.3658978044986725, "learning_rate": 2.3103148121282464e-06, "loss": 0.0064, "step": 24487 }, { "epoch": 2.6548135299219426, "grad_norm": 2.5056238174438477, "learning_rate": 2.309589438560859e-06, "loss": 0.0257, "step": 24488 }, { "epoch": 2.6549219427580226, "grad_norm": 0.02969766966998577, "learning_rate": 2.308864064993472e-06, "loss": 0.0002, "step": 24489 }, { "epoch": 2.6550303555941026, "grad_norm": 0.27533403038978577, "learning_rate": 2.3081386914260845e-06, "loss": 0.0054, "step": 24490 }, { "epoch": 2.655138768430182, "grad_norm": 4.978143215179443, "learning_rate": 2.3074133178586975e-06, "loss": 0.0292, "step": 24491 }, { "epoch": 2.6552471812662617, "grad_norm": 1.8338176012039185, "learning_rate": 2.30668794429131e-06, "loss": 0.0065, "step": 24492 }, { "epoch": 2.6553555941023417, "grad_norm": 7.860265254974365, "learning_rate": 2.305962570723923e-06, "loss": 0.0344, "step": 24493 }, { "epoch": 2.6554640069384217, "grad_norm": 0.17106710374355316, "learning_rate": 2.3052371971565356e-06, "loss": 0.003, "step": 24494 }, { "epoch": 2.655572419774501, "grad_norm": 0.05160845071077347, "learning_rate": 2.3045118235891486e-06, "loss": 0.0003, "step": 24495 }, { "epoch": 2.655680832610581, "grad_norm": 0.286943644285202, "learning_rate": 2.303786450021761e-06, "loss": 0.01, "step": 24496 }, { "epoch": 2.6557892454466607, "grad_norm": 0.02577861025929451, "learning_rate": 2.303061076454374e-06, "loss": 0.0002, "step": 24497 }, { "epoch": 2.6558976582827407, "grad_norm": 0.028391322121024132, "learning_rate": 2.3023357028869867e-06, "loss": 0.0002, "step": 24498 }, { "epoch": 2.6560060711188207, "grad_norm": 0.025529181584715843, "learning_rate": 2.3016103293195997e-06, "loss": 0.0002, "step": 24499 }, { "epoch": 2.6561144839549002, "grad_norm": 4.685073375701904, "learning_rate": 2.3008849557522127e-06, "loss": 0.0334, "step": 24500 }, { "epoch": 2.65622289679098, "grad_norm": 0.09967987984418869, "learning_rate": 2.3001595821848253e-06, "loss": 0.0002, "step": 24501 }, { "epoch": 2.6563313096270598, "grad_norm": 0.2628497779369354, "learning_rate": 2.2994342086174383e-06, "loss": 0.0053, "step": 24502 }, { "epoch": 2.6564397224631398, "grad_norm": 0.21013087034225464, "learning_rate": 2.2987088350500513e-06, "loss": 0.0041, "step": 24503 }, { "epoch": 2.6565481352992193, "grad_norm": 0.006963090039789677, "learning_rate": 2.297983461482664e-06, "loss": 0.0001, "step": 24504 }, { "epoch": 2.6566565481352993, "grad_norm": 0.13413874804973602, "learning_rate": 2.2972580879152764e-06, "loss": 0.0011, "step": 24505 }, { "epoch": 2.656764960971379, "grad_norm": 0.7469648718833923, "learning_rate": 2.2965327143478894e-06, "loss": 0.0172, "step": 24506 }, { "epoch": 2.656873373807459, "grad_norm": 0.4964732229709625, "learning_rate": 2.295807340780502e-06, "loss": 0.0133, "step": 24507 }, { "epoch": 2.656981786643539, "grad_norm": 9.944560050964355, "learning_rate": 2.295081967213115e-06, "loss": 0.0657, "step": 24508 }, { "epoch": 2.6570901994796183, "grad_norm": 3.714804172515869, "learning_rate": 2.2943565936457275e-06, "loss": 0.0105, "step": 24509 }, { "epoch": 2.6571986123156983, "grad_norm": 0.21237413585186005, "learning_rate": 2.2936312200783405e-06, "loss": 0.0035, "step": 24510 }, { "epoch": 2.657307025151778, "grad_norm": 0.006787415128201246, "learning_rate": 2.292905846510953e-06, "loss": 0.0001, "step": 24511 }, { "epoch": 2.657415437987858, "grad_norm": 0.40553995966911316, "learning_rate": 2.292180472943566e-06, "loss": 0.0009, "step": 24512 }, { "epoch": 2.657523850823938, "grad_norm": 3.5857131481170654, "learning_rate": 2.291455099376179e-06, "loss": 0.03, "step": 24513 }, { "epoch": 2.6576322636600174, "grad_norm": 0.013265966437757015, "learning_rate": 2.290729725808792e-06, "loss": 0.0001, "step": 24514 }, { "epoch": 2.657740676496097, "grad_norm": 0.31077367067337036, "learning_rate": 2.2900043522414046e-06, "loss": 0.0006, "step": 24515 }, { "epoch": 2.657849089332177, "grad_norm": 0.07062135636806488, "learning_rate": 2.2892789786740176e-06, "loss": 0.0002, "step": 24516 }, { "epoch": 2.657957502168257, "grad_norm": 0.3630552887916565, "learning_rate": 2.28855360510663e-06, "loss": 0.0094, "step": 24517 }, { "epoch": 2.6580659150043364, "grad_norm": 0.9582531452178955, "learning_rate": 2.2878282315392427e-06, "loss": 0.0021, "step": 24518 }, { "epoch": 2.6581743278404164, "grad_norm": 0.04459157958626747, "learning_rate": 2.2871028579718557e-06, "loss": 0.0004, "step": 24519 }, { "epoch": 2.658282740676496, "grad_norm": 0.4099633991718292, "learning_rate": 2.2863774844044683e-06, "loss": 0.0005, "step": 24520 }, { "epoch": 2.658391153512576, "grad_norm": 0.4294472932815552, "learning_rate": 2.2856521108370813e-06, "loss": 0.0088, "step": 24521 }, { "epoch": 2.658499566348656, "grad_norm": 0.015273061580955982, "learning_rate": 2.284926737269694e-06, "loss": 0.0001, "step": 24522 }, { "epoch": 2.6586079791847355, "grad_norm": 0.24741487205028534, "learning_rate": 2.284201363702307e-06, "loss": 0.0049, "step": 24523 }, { "epoch": 2.658716392020815, "grad_norm": 0.03687175363302231, "learning_rate": 2.2834759901349194e-06, "loss": 0.0004, "step": 24524 }, { "epoch": 2.658824804856895, "grad_norm": 0.019994640722870827, "learning_rate": 2.2827506165675324e-06, "loss": 0.0001, "step": 24525 }, { "epoch": 2.658933217692975, "grad_norm": 2.026399850845337, "learning_rate": 2.2820252430001454e-06, "loss": 0.0162, "step": 24526 }, { "epoch": 2.6590416305290545, "grad_norm": 0.020063811913132668, "learning_rate": 2.2812998694327584e-06, "loss": 0.0001, "step": 24527 }, { "epoch": 2.6591500433651345, "grad_norm": 1.0524017810821533, "learning_rate": 2.280574495865371e-06, "loss": 0.003, "step": 24528 }, { "epoch": 2.659258456201214, "grad_norm": 4.792993545532227, "learning_rate": 2.279849122297984e-06, "loss": 0.0291, "step": 24529 }, { "epoch": 2.659366869037294, "grad_norm": 0.7813266515731812, "learning_rate": 2.2791237487305965e-06, "loss": 0.0157, "step": 24530 }, { "epoch": 2.659475281873374, "grad_norm": 0.3079252541065216, "learning_rate": 2.278398375163209e-06, "loss": 0.0065, "step": 24531 }, { "epoch": 2.6595836947094535, "grad_norm": 0.023699603974819183, "learning_rate": 2.277673001595822e-06, "loss": 0.0001, "step": 24532 }, { "epoch": 2.6596921075455335, "grad_norm": 11.153450012207031, "learning_rate": 2.2769476280284346e-06, "loss": 0.0427, "step": 24533 }, { "epoch": 2.659800520381613, "grad_norm": 6.8318071365356445, "learning_rate": 2.2762222544610476e-06, "loss": 0.0491, "step": 24534 }, { "epoch": 2.659908933217693, "grad_norm": 0.3501530885696411, "learning_rate": 2.27549688089366e-06, "loss": 0.0064, "step": 24535 }, { "epoch": 2.6600173460537726, "grad_norm": 0.3967343270778656, "learning_rate": 2.274771507326273e-06, "loss": 0.014, "step": 24536 }, { "epoch": 2.6601257588898526, "grad_norm": 7.268258571624756, "learning_rate": 2.2740461337588857e-06, "loss": 0.0467, "step": 24537 }, { "epoch": 2.660234171725932, "grad_norm": 2.922305107116699, "learning_rate": 2.2733207601914987e-06, "loss": 0.0015, "step": 24538 }, { "epoch": 2.660342584562012, "grad_norm": 0.18687286972999573, "learning_rate": 2.2725953866241117e-06, "loss": 0.0005, "step": 24539 }, { "epoch": 2.660450997398092, "grad_norm": 0.2956313490867615, "learning_rate": 2.2718700130567247e-06, "loss": 0.0061, "step": 24540 }, { "epoch": 2.6605594102341716, "grad_norm": 0.010591285303235054, "learning_rate": 2.2711446394893373e-06, "loss": 0.0001, "step": 24541 }, { "epoch": 2.6606678230702516, "grad_norm": 15.114876747131348, "learning_rate": 2.2704192659219503e-06, "loss": 0.0221, "step": 24542 }, { "epoch": 2.660776235906331, "grad_norm": 0.017488868907094002, "learning_rate": 2.269693892354563e-06, "loss": 0.0001, "step": 24543 }, { "epoch": 2.660884648742411, "grad_norm": 0.31245940923690796, "learning_rate": 2.2689685187871754e-06, "loss": 0.0117, "step": 24544 }, { "epoch": 2.660993061578491, "grad_norm": 0.24671176075935364, "learning_rate": 2.2682431452197884e-06, "loss": 0.005, "step": 24545 }, { "epoch": 2.6611014744145707, "grad_norm": 0.22512643039226532, "learning_rate": 2.267517771652401e-06, "loss": 0.0046, "step": 24546 }, { "epoch": 2.66120988725065, "grad_norm": 0.3427180349826813, "learning_rate": 2.266792398085014e-06, "loss": 0.0007, "step": 24547 }, { "epoch": 2.66131830008673, "grad_norm": 12.217548370361328, "learning_rate": 2.2660670245176265e-06, "loss": 0.0177, "step": 24548 }, { "epoch": 2.66142671292281, "grad_norm": 0.3481563627719879, "learning_rate": 2.2653416509502395e-06, "loss": 0.0067, "step": 24549 }, { "epoch": 2.6615351257588897, "grad_norm": 2.142822742462158, "learning_rate": 2.264616277382852e-06, "loss": 0.049, "step": 24550 }, { "epoch": 2.6616435385949697, "grad_norm": 0.415959894657135, "learning_rate": 2.263890903815465e-06, "loss": 0.0121, "step": 24551 }, { "epoch": 2.6617519514310493, "grad_norm": 0.014151289127767086, "learning_rate": 2.263165530248078e-06, "loss": 0.0001, "step": 24552 }, { "epoch": 2.6618603642671292, "grad_norm": 3.184122085571289, "learning_rate": 2.262440156680691e-06, "loss": 0.0354, "step": 24553 }, { "epoch": 2.6619687771032092, "grad_norm": 0.6993269324302673, "learning_rate": 2.2617147831133036e-06, "loss": 0.0047, "step": 24554 }, { "epoch": 2.6620771899392888, "grad_norm": 0.04878286272287369, "learning_rate": 2.2609894095459166e-06, "loss": 0.0003, "step": 24555 }, { "epoch": 2.6621856027753688, "grad_norm": 0.00963995698839426, "learning_rate": 2.260264035978529e-06, "loss": 0.0001, "step": 24556 }, { "epoch": 2.6622940156114483, "grad_norm": 1.919052004814148, "learning_rate": 2.259538662411142e-06, "loss": 0.0655, "step": 24557 }, { "epoch": 2.6624024284475283, "grad_norm": 0.38722822070121765, "learning_rate": 2.2588132888437547e-06, "loss": 0.0069, "step": 24558 }, { "epoch": 2.662510841283608, "grad_norm": 0.3268873691558838, "learning_rate": 2.2580879152763673e-06, "loss": 0.0113, "step": 24559 }, { "epoch": 2.662619254119688, "grad_norm": 0.02422862872481346, "learning_rate": 2.2573625417089803e-06, "loss": 0.0001, "step": 24560 }, { "epoch": 2.6627276669557673, "grad_norm": 0.8411363363265991, "learning_rate": 2.256637168141593e-06, "loss": 0.0205, "step": 24561 }, { "epoch": 2.6628360797918473, "grad_norm": 0.009195312857627869, "learning_rate": 2.255911794574206e-06, "loss": 0.0001, "step": 24562 }, { "epoch": 2.6629444926279273, "grad_norm": 2.3220443725585938, "learning_rate": 2.255186421006819e-06, "loss": 0.0307, "step": 24563 }, { "epoch": 2.663052905464007, "grad_norm": 0.025380684062838554, "learning_rate": 2.2544610474394314e-06, "loss": 0.0002, "step": 24564 }, { "epoch": 2.663161318300087, "grad_norm": 0.010281933471560478, "learning_rate": 2.2537356738720444e-06, "loss": 0.0001, "step": 24565 }, { "epoch": 2.6632697311361664, "grad_norm": 0.4185151159763336, "learning_rate": 2.2530103003046574e-06, "loss": 0.0086, "step": 24566 }, { "epoch": 2.6633781439722464, "grad_norm": 0.631328284740448, "learning_rate": 2.25228492673727e-06, "loss": 0.002, "step": 24567 }, { "epoch": 2.6634865568083264, "grad_norm": 2.6905314922332764, "learning_rate": 2.251559553169883e-06, "loss": 0.0137, "step": 24568 }, { "epoch": 2.663594969644406, "grad_norm": 4.719631671905518, "learning_rate": 2.2508341796024955e-06, "loss": 0.0332, "step": 24569 }, { "epoch": 2.6637033824804854, "grad_norm": 6.2201948165893555, "learning_rate": 2.2501088060351085e-06, "loss": 0.0231, "step": 24570 }, { "epoch": 2.6638117953165654, "grad_norm": 0.5231245160102844, "learning_rate": 2.249383432467721e-06, "loss": 0.0147, "step": 24571 }, { "epoch": 2.6639202081526454, "grad_norm": 6.789490222930908, "learning_rate": 2.2486580589003336e-06, "loss": 0.0112, "step": 24572 }, { "epoch": 2.664028620988725, "grad_norm": 0.2599494159221649, "learning_rate": 2.2479326853329466e-06, "loss": 0.0007, "step": 24573 }, { "epoch": 2.664137033824805, "grad_norm": 0.4451512098312378, "learning_rate": 2.247207311765559e-06, "loss": 0.0111, "step": 24574 }, { "epoch": 2.6642454466608845, "grad_norm": 10.259806632995605, "learning_rate": 2.246481938198172e-06, "loss": 0.0194, "step": 24575 }, { "epoch": 2.6643538594969645, "grad_norm": 1.211397647857666, "learning_rate": 2.245756564630785e-06, "loss": 0.0316, "step": 24576 }, { "epoch": 2.6644622723330444, "grad_norm": 2.021982431411743, "learning_rate": 2.2450311910633977e-06, "loss": 0.0352, "step": 24577 }, { "epoch": 2.664570685169124, "grad_norm": 12.824447631835938, "learning_rate": 2.2443058174960107e-06, "loss": 0.0413, "step": 24578 }, { "epoch": 2.664679098005204, "grad_norm": 0.3837340772151947, "learning_rate": 2.2435804439286237e-06, "loss": 0.0122, "step": 24579 }, { "epoch": 2.6647875108412835, "grad_norm": 0.16593037545681, "learning_rate": 2.2428550703612363e-06, "loss": 0.0003, "step": 24580 }, { "epoch": 2.6648959236773635, "grad_norm": 2.164466619491577, "learning_rate": 2.2421296967938493e-06, "loss": 0.0026, "step": 24581 }, { "epoch": 2.665004336513443, "grad_norm": 4.2455973625183105, "learning_rate": 2.241404323226462e-06, "loss": 0.0541, "step": 24582 }, { "epoch": 2.665112749349523, "grad_norm": 2.5034918785095215, "learning_rate": 2.240678949659075e-06, "loss": 0.0138, "step": 24583 }, { "epoch": 2.6652211621856026, "grad_norm": 0.020691171288490295, "learning_rate": 2.2399535760916874e-06, "loss": 0.0001, "step": 24584 }, { "epoch": 2.6653295750216826, "grad_norm": 0.2999187707901001, "learning_rate": 2.2392282025243e-06, "loss": 0.0067, "step": 24585 }, { "epoch": 2.6654379878577625, "grad_norm": 0.34254980087280273, "learning_rate": 2.238502828956913e-06, "loss": 0.0137, "step": 24586 }, { "epoch": 2.665546400693842, "grad_norm": 0.4763297736644745, "learning_rate": 2.2377774553895255e-06, "loss": 0.0079, "step": 24587 }, { "epoch": 2.665654813529922, "grad_norm": 10.799531936645508, "learning_rate": 2.2370520818221385e-06, "loss": 0.0157, "step": 24588 }, { "epoch": 2.6657632263660016, "grad_norm": 15.491630554199219, "learning_rate": 2.2363267082547515e-06, "loss": 0.0341, "step": 24589 }, { "epoch": 2.6658716392020816, "grad_norm": 5.290739059448242, "learning_rate": 2.235601334687364e-06, "loss": 0.0232, "step": 24590 }, { "epoch": 2.6659800520381616, "grad_norm": 0.003338687354698777, "learning_rate": 2.234875961119977e-06, "loss": 0.0, "step": 24591 }, { "epoch": 2.666088464874241, "grad_norm": 1.7463151216506958, "learning_rate": 2.23415058755259e-06, "loss": 0.0481, "step": 24592 }, { "epoch": 2.6661968777103207, "grad_norm": 0.027150483801960945, "learning_rate": 2.2334252139852026e-06, "loss": 0.0002, "step": 24593 }, { "epoch": 2.6663052905464006, "grad_norm": 0.45456215739250183, "learning_rate": 2.2326998404178156e-06, "loss": 0.0018, "step": 24594 }, { "epoch": 2.6664137033824806, "grad_norm": 1.3043296337127686, "learning_rate": 2.231974466850428e-06, "loss": 0.0012, "step": 24595 }, { "epoch": 2.66652211621856, "grad_norm": 0.19214676320552826, "learning_rate": 2.231249093283041e-06, "loss": 0.0003, "step": 24596 }, { "epoch": 2.66663052905464, "grad_norm": 3.0435047149658203, "learning_rate": 2.2305237197156537e-06, "loss": 0.0174, "step": 24597 }, { "epoch": 2.6667389418907197, "grad_norm": 0.32956185936927795, "learning_rate": 2.2297983461482663e-06, "loss": 0.0089, "step": 24598 }, { "epoch": 2.6668473547267997, "grad_norm": 1.4212769269943237, "learning_rate": 2.2290729725808793e-06, "loss": 0.0037, "step": 24599 }, { "epoch": 2.6669557675628797, "grad_norm": 0.07681853324174881, "learning_rate": 2.228347599013492e-06, "loss": 0.0003, "step": 24600 }, { "epoch": 2.667064180398959, "grad_norm": 0.0706746056675911, "learning_rate": 2.227622225446105e-06, "loss": 0.0003, "step": 24601 }, { "epoch": 2.667172593235039, "grad_norm": 24.644573211669922, "learning_rate": 2.226896851878718e-06, "loss": 0.0283, "step": 24602 }, { "epoch": 2.6672810060711187, "grad_norm": 0.49982354044914246, "learning_rate": 2.2261714783113304e-06, "loss": 0.0015, "step": 24603 }, { "epoch": 2.6673894189071987, "grad_norm": 0.009769032709300518, "learning_rate": 2.2254461047439434e-06, "loss": 0.0001, "step": 24604 }, { "epoch": 2.6674978317432783, "grad_norm": 0.4280654191970825, "learning_rate": 2.2247207311765564e-06, "loss": 0.013, "step": 24605 }, { "epoch": 2.6676062445793582, "grad_norm": 0.039989542216062546, "learning_rate": 2.223995357609169e-06, "loss": 0.0004, "step": 24606 }, { "epoch": 2.667714657415438, "grad_norm": 0.06500076502561569, "learning_rate": 2.223269984041782e-06, "loss": 0.0003, "step": 24607 }, { "epoch": 2.6678230702515178, "grad_norm": 2.9084911346435547, "learning_rate": 2.2225446104743945e-06, "loss": 0.0117, "step": 24608 }, { "epoch": 2.6679314830875978, "grad_norm": 1.866645336151123, "learning_rate": 2.2218192369070075e-06, "loss": 0.0064, "step": 24609 }, { "epoch": 2.6680398959236773, "grad_norm": 0.7897723317146301, "learning_rate": 2.22109386333962e-06, "loss": 0.0166, "step": 24610 }, { "epoch": 2.6681483087597573, "grad_norm": 3.8776638507843018, "learning_rate": 2.220368489772233e-06, "loss": 0.0249, "step": 24611 }, { "epoch": 2.668256721595837, "grad_norm": 12.69300365447998, "learning_rate": 2.2196431162048456e-06, "loss": 0.0342, "step": 24612 }, { "epoch": 2.668365134431917, "grad_norm": 0.4679158329963684, "learning_rate": 2.218917742637458e-06, "loss": 0.0019, "step": 24613 }, { "epoch": 2.668473547267997, "grad_norm": 0.0313728041946888, "learning_rate": 2.218192369070071e-06, "loss": 0.0003, "step": 24614 }, { "epoch": 2.6685819601040763, "grad_norm": 0.39050430059432983, "learning_rate": 2.217466995502684e-06, "loss": 0.0107, "step": 24615 }, { "epoch": 2.668690372940156, "grad_norm": 0.3063514828681946, "learning_rate": 2.2167416219352967e-06, "loss": 0.0091, "step": 24616 }, { "epoch": 2.668798785776236, "grad_norm": 3.5040199756622314, "learning_rate": 2.2160162483679097e-06, "loss": 0.016, "step": 24617 }, { "epoch": 2.668907198612316, "grad_norm": 0.5186737179756165, "learning_rate": 2.2152908748005227e-06, "loss": 0.0105, "step": 24618 }, { "epoch": 2.6690156114483954, "grad_norm": 0.3178355097770691, "learning_rate": 2.2145655012331353e-06, "loss": 0.0114, "step": 24619 }, { "epoch": 2.6691240242844754, "grad_norm": 1.864490270614624, "learning_rate": 2.2138401276657483e-06, "loss": 0.0126, "step": 24620 }, { "epoch": 2.669232437120555, "grad_norm": 0.24086220562458038, "learning_rate": 2.213114754098361e-06, "loss": 0.0051, "step": 24621 }, { "epoch": 2.669340849956635, "grad_norm": 0.5624645352363586, "learning_rate": 2.212389380530974e-06, "loss": 0.0196, "step": 24622 }, { "epoch": 2.669449262792715, "grad_norm": 0.46369341015815735, "learning_rate": 2.2116640069635864e-06, "loss": 0.0024, "step": 24623 }, { "epoch": 2.6695576756287944, "grad_norm": 0.30744484066963196, "learning_rate": 2.2109386333961994e-06, "loss": 0.009, "step": 24624 }, { "epoch": 2.669666088464874, "grad_norm": 3.0266153812408447, "learning_rate": 2.210213259828812e-06, "loss": 0.0229, "step": 24625 }, { "epoch": 2.669774501300954, "grad_norm": 2.0274224281311035, "learning_rate": 2.2094878862614245e-06, "loss": 0.0012, "step": 24626 }, { "epoch": 2.669882914137034, "grad_norm": 4.3849334716796875, "learning_rate": 2.2087625126940375e-06, "loss": 0.0114, "step": 24627 }, { "epoch": 2.6699913269731135, "grad_norm": 0.4566013514995575, "learning_rate": 2.2080371391266505e-06, "loss": 0.0103, "step": 24628 }, { "epoch": 2.6700997398091935, "grad_norm": 0.31569933891296387, "learning_rate": 2.207311765559263e-06, "loss": 0.0072, "step": 24629 }, { "epoch": 2.670208152645273, "grad_norm": 0.04559200629591942, "learning_rate": 2.206586391991876e-06, "loss": 0.0001, "step": 24630 }, { "epoch": 2.670316565481353, "grad_norm": 0.022144371643662453, "learning_rate": 2.205861018424489e-06, "loss": 0.0002, "step": 24631 }, { "epoch": 2.670424978317433, "grad_norm": 0.41355881094932556, "learning_rate": 2.2051356448571016e-06, "loss": 0.0081, "step": 24632 }, { "epoch": 2.6705333911535125, "grad_norm": 0.008475523442029953, "learning_rate": 2.2044102712897146e-06, "loss": 0.0001, "step": 24633 }, { "epoch": 2.6706418039895925, "grad_norm": 0.3754418194293976, "learning_rate": 2.203684897722327e-06, "loss": 0.0114, "step": 24634 }, { "epoch": 2.670750216825672, "grad_norm": 0.09504657238721848, "learning_rate": 2.20295952415494e-06, "loss": 0.0106, "step": 24635 }, { "epoch": 2.670858629661752, "grad_norm": 0.4142114520072937, "learning_rate": 2.2022341505875527e-06, "loss": 0.0075, "step": 24636 }, { "epoch": 2.670967042497832, "grad_norm": 0.3672944903373718, "learning_rate": 2.2015087770201657e-06, "loss": 0.0079, "step": 24637 }, { "epoch": 2.6710754553339116, "grad_norm": 5.177220821380615, "learning_rate": 2.2007834034527783e-06, "loss": 0.0523, "step": 24638 }, { "epoch": 2.671183868169991, "grad_norm": 0.40452301502227783, "learning_rate": 2.200058029885391e-06, "loss": 0.0194, "step": 24639 }, { "epoch": 2.671292281006071, "grad_norm": 0.2983510196208954, "learning_rate": 2.199332656318004e-06, "loss": 0.0098, "step": 24640 }, { "epoch": 2.671400693842151, "grad_norm": 0.333820104598999, "learning_rate": 2.198607282750617e-06, "loss": 0.0117, "step": 24641 }, { "epoch": 2.6715091066782306, "grad_norm": 1.949753761291504, "learning_rate": 2.1978819091832294e-06, "loss": 0.0277, "step": 24642 }, { "epoch": 2.6716175195143106, "grad_norm": 1.100682020187378, "learning_rate": 2.1971565356158424e-06, "loss": 0.037, "step": 24643 }, { "epoch": 2.67172593235039, "grad_norm": 1.3351718187332153, "learning_rate": 2.1964311620484554e-06, "loss": 0.0018, "step": 24644 }, { "epoch": 2.67183434518647, "grad_norm": 1.448302149772644, "learning_rate": 2.195705788481068e-06, "loss": 0.0029, "step": 24645 }, { "epoch": 2.67194275802255, "grad_norm": 0.0240489412099123, "learning_rate": 2.194980414913681e-06, "loss": 0.0001, "step": 24646 }, { "epoch": 2.6720511708586296, "grad_norm": 1.0799798965454102, "learning_rate": 2.1942550413462935e-06, "loss": 0.001, "step": 24647 }, { "epoch": 2.672159583694709, "grad_norm": 0.1400100141763687, "learning_rate": 2.1935296677789065e-06, "loss": 0.0008, "step": 24648 }, { "epoch": 2.672267996530789, "grad_norm": 0.0879703238606453, "learning_rate": 2.192804294211519e-06, "loss": 0.0005, "step": 24649 }, { "epoch": 2.672376409366869, "grad_norm": 0.08190866559743881, "learning_rate": 2.192078920644132e-06, "loss": 0.0003, "step": 24650 }, { "epoch": 2.6724848222029487, "grad_norm": 0.28211480379104614, "learning_rate": 2.1913535470767446e-06, "loss": 0.0062, "step": 24651 }, { "epoch": 2.6725932350390287, "grad_norm": 1.6214309930801392, "learning_rate": 2.1906281735093576e-06, "loss": 0.0186, "step": 24652 }, { "epoch": 2.672701647875108, "grad_norm": 0.5030799508094788, "learning_rate": 2.18990279994197e-06, "loss": 0.015, "step": 24653 }, { "epoch": 2.672810060711188, "grad_norm": 0.42527174949645996, "learning_rate": 2.189177426374583e-06, "loss": 0.0159, "step": 24654 }, { "epoch": 2.672918473547268, "grad_norm": 3.086273431777954, "learning_rate": 2.1884520528071957e-06, "loss": 0.0244, "step": 24655 }, { "epoch": 2.6730268863833477, "grad_norm": 0.7882347106933594, "learning_rate": 2.1877266792398087e-06, "loss": 0.0015, "step": 24656 }, { "epoch": 2.6731352992194277, "grad_norm": 0.52587890625, "learning_rate": 2.1870013056724217e-06, "loss": 0.0151, "step": 24657 }, { "epoch": 2.6732437120555073, "grad_norm": 11.39619255065918, "learning_rate": 2.1862759321050343e-06, "loss": 0.0192, "step": 24658 }, { "epoch": 2.6733521248915872, "grad_norm": 0.2478422075510025, "learning_rate": 2.1855505585376473e-06, "loss": 0.0037, "step": 24659 }, { "epoch": 2.6734605377276672, "grad_norm": 0.04834861308336258, "learning_rate": 2.18482518497026e-06, "loss": 0.0001, "step": 24660 }, { "epoch": 2.6735689505637468, "grad_norm": 0.00859770830720663, "learning_rate": 2.184099811402873e-06, "loss": 0.0001, "step": 24661 }, { "epoch": 2.6736773633998263, "grad_norm": 0.5314750075340271, "learning_rate": 2.1833744378354854e-06, "loss": 0.0154, "step": 24662 }, { "epoch": 2.6737857762359063, "grad_norm": 6.014034748077393, "learning_rate": 2.1826490642680984e-06, "loss": 0.0145, "step": 24663 }, { "epoch": 2.6738941890719863, "grad_norm": 0.02508598007261753, "learning_rate": 2.181923690700711e-06, "loss": 0.0001, "step": 24664 }, { "epoch": 2.674002601908066, "grad_norm": 0.02420504204928875, "learning_rate": 2.181198317133324e-06, "loss": 0.0003, "step": 24665 }, { "epoch": 2.674111014744146, "grad_norm": 0.42726370692253113, "learning_rate": 2.1804729435659365e-06, "loss": 0.0224, "step": 24666 }, { "epoch": 2.6742194275802254, "grad_norm": 0.26811474561691284, "learning_rate": 2.1797475699985495e-06, "loss": 0.0056, "step": 24667 }, { "epoch": 2.6743278404163053, "grad_norm": 0.050486303865909576, "learning_rate": 2.179022196431162e-06, "loss": 0.0002, "step": 24668 }, { "epoch": 2.6744362532523853, "grad_norm": 8.289793014526367, "learning_rate": 2.178296822863775e-06, "loss": 0.053, "step": 24669 }, { "epoch": 2.674544666088465, "grad_norm": 0.25819870829582214, "learning_rate": 2.177571449296388e-06, "loss": 0.0052, "step": 24670 }, { "epoch": 2.6746530789245444, "grad_norm": 0.47075676918029785, "learning_rate": 2.1768460757290006e-06, "loss": 0.0146, "step": 24671 }, { "epoch": 2.6747614917606244, "grad_norm": 0.06506816297769547, "learning_rate": 2.1761207021616136e-06, "loss": 0.0111, "step": 24672 }, { "epoch": 2.6748699045967044, "grad_norm": 0.548947274684906, "learning_rate": 2.175395328594226e-06, "loss": 0.002, "step": 24673 }, { "epoch": 2.674978317432784, "grad_norm": 0.38030144572257996, "learning_rate": 2.174669955026839e-06, "loss": 0.0103, "step": 24674 }, { "epoch": 2.675086730268864, "grad_norm": 0.2890810966491699, "learning_rate": 2.1739445814594517e-06, "loss": 0.006, "step": 24675 }, { "epoch": 2.6751951431049434, "grad_norm": 0.6039587259292603, "learning_rate": 2.1732192078920647e-06, "loss": 0.0151, "step": 24676 }, { "epoch": 2.6753035559410234, "grad_norm": 0.26719367504119873, "learning_rate": 2.1724938343246773e-06, "loss": 0.0055, "step": 24677 }, { "epoch": 2.6754119687771034, "grad_norm": 7.45485258102417, "learning_rate": 2.1717684607572903e-06, "loss": 0.0952, "step": 24678 }, { "epoch": 2.675520381613183, "grad_norm": 4.99827766418457, "learning_rate": 2.171043087189903e-06, "loss": 0.0038, "step": 24679 }, { "epoch": 2.675628794449263, "grad_norm": 8.284239768981934, "learning_rate": 2.170317713622516e-06, "loss": 0.0911, "step": 24680 }, { "epoch": 2.6757372072853425, "grad_norm": 0.011270696297287941, "learning_rate": 2.1695923400551284e-06, "loss": 0.0001, "step": 24681 }, { "epoch": 2.6758456201214225, "grad_norm": 15.173064231872559, "learning_rate": 2.1688669664877414e-06, "loss": 0.0288, "step": 24682 }, { "epoch": 2.675954032957502, "grad_norm": 0.4548802673816681, "learning_rate": 2.1681415929203544e-06, "loss": 0.0084, "step": 24683 }, { "epoch": 2.676062445793582, "grad_norm": 6.609391689300537, "learning_rate": 2.167416219352967e-06, "loss": 0.0072, "step": 24684 }, { "epoch": 2.6761708586296615, "grad_norm": 0.5517024993896484, "learning_rate": 2.16669084578558e-06, "loss": 0.0171, "step": 24685 }, { "epoch": 2.6762792714657415, "grad_norm": 0.3692467212677002, "learning_rate": 2.1659654722181925e-06, "loss": 0.0006, "step": 24686 }, { "epoch": 2.6763876843018215, "grad_norm": 0.022069204598665237, "learning_rate": 2.1652400986508055e-06, "loss": 0.0001, "step": 24687 }, { "epoch": 2.676496097137901, "grad_norm": 5.368709564208984, "learning_rate": 2.164514725083418e-06, "loss": 0.0153, "step": 24688 }, { "epoch": 2.676604509973981, "grad_norm": 6.035297870635986, "learning_rate": 2.163789351516031e-06, "loss": 0.0482, "step": 24689 }, { "epoch": 2.6767129228100606, "grad_norm": 4.788644313812256, "learning_rate": 2.1630639779486436e-06, "loss": 0.0209, "step": 24690 }, { "epoch": 2.6768213356461406, "grad_norm": 4.779858589172363, "learning_rate": 2.1623386043812566e-06, "loss": 0.0159, "step": 24691 }, { "epoch": 2.6769297484822205, "grad_norm": 22.234304428100586, "learning_rate": 2.161613230813869e-06, "loss": 0.0263, "step": 24692 }, { "epoch": 2.6770381613183, "grad_norm": 3.7345688343048096, "learning_rate": 2.160887857246482e-06, "loss": 0.0241, "step": 24693 }, { "epoch": 2.6771465741543796, "grad_norm": 0.05421177297830582, "learning_rate": 2.1601624836790947e-06, "loss": 0.0002, "step": 24694 }, { "epoch": 2.6772549869904596, "grad_norm": 7.424501419067383, "learning_rate": 2.1594371101117077e-06, "loss": 0.0501, "step": 24695 }, { "epoch": 2.6773633998265396, "grad_norm": 0.3437649607658386, "learning_rate": 2.1587117365443207e-06, "loss": 0.0125, "step": 24696 }, { "epoch": 2.677471812662619, "grad_norm": 0.5680751204490662, "learning_rate": 2.1579863629769333e-06, "loss": 0.0258, "step": 24697 }, { "epoch": 2.677580225498699, "grad_norm": 0.22833143174648285, "learning_rate": 2.1572609894095463e-06, "loss": 0.0151, "step": 24698 }, { "epoch": 2.6776886383347787, "grad_norm": 0.014393451623618603, "learning_rate": 2.156535615842159e-06, "loss": 0.0002, "step": 24699 }, { "epoch": 2.6777970511708586, "grad_norm": 3.4138967990875244, "learning_rate": 2.155810242274772e-06, "loss": 0.0265, "step": 24700 }, { "epoch": 2.6779054640069386, "grad_norm": 6.192888259887695, "learning_rate": 2.1550848687073844e-06, "loss": 0.0442, "step": 24701 }, { "epoch": 2.678013876843018, "grad_norm": 0.3598376214504242, "learning_rate": 2.1543594951399974e-06, "loss": 0.0071, "step": 24702 }, { "epoch": 2.678122289679098, "grad_norm": 1.8776695728302002, "learning_rate": 2.15363412157261e-06, "loss": 0.0174, "step": 24703 }, { "epoch": 2.6782307025151777, "grad_norm": 0.1707005500793457, "learning_rate": 2.152908748005223e-06, "loss": 0.0004, "step": 24704 }, { "epoch": 2.6783391153512577, "grad_norm": 0.9071829319000244, "learning_rate": 2.1521833744378355e-06, "loss": 0.0025, "step": 24705 }, { "epoch": 2.6784475281873372, "grad_norm": 0.24917708337306976, "learning_rate": 2.1514580008704485e-06, "loss": 0.0055, "step": 24706 }, { "epoch": 2.678555941023417, "grad_norm": 0.47396889328956604, "learning_rate": 2.150732627303061e-06, "loss": 0.0174, "step": 24707 }, { "epoch": 2.6786643538594967, "grad_norm": 0.00885815266519785, "learning_rate": 2.150007253735674e-06, "loss": 0.0001, "step": 24708 }, { "epoch": 2.6787727666955767, "grad_norm": 0.3892066776752472, "learning_rate": 2.149281880168287e-06, "loss": 0.0241, "step": 24709 }, { "epoch": 2.6788811795316567, "grad_norm": 1.047417402267456, "learning_rate": 2.1485565066008996e-06, "loss": 0.0069, "step": 24710 }, { "epoch": 2.6789895923677363, "grad_norm": 0.7057492733001709, "learning_rate": 2.1478311330335126e-06, "loss": 0.0195, "step": 24711 }, { "epoch": 2.6790980052038162, "grad_norm": 0.5234174132347107, "learning_rate": 2.147105759466125e-06, "loss": 0.0106, "step": 24712 }, { "epoch": 2.679206418039896, "grad_norm": 2.7568066120147705, "learning_rate": 2.146380385898738e-06, "loss": 0.0302, "step": 24713 }, { "epoch": 2.6793148308759758, "grad_norm": 0.2624637186527252, "learning_rate": 2.1456550123313507e-06, "loss": 0.0056, "step": 24714 }, { "epoch": 2.6794232437120558, "grad_norm": 13.3624906539917, "learning_rate": 2.1449296387639637e-06, "loss": 0.0205, "step": 24715 }, { "epoch": 2.6795316565481353, "grad_norm": 2.0285096168518066, "learning_rate": 2.1442042651965763e-06, "loss": 0.009, "step": 24716 }, { "epoch": 2.679640069384215, "grad_norm": 0.6170318722724915, "learning_rate": 2.1434788916291893e-06, "loss": 0.0123, "step": 24717 }, { "epoch": 2.679748482220295, "grad_norm": 0.07611437886953354, "learning_rate": 2.142753518061802e-06, "loss": 0.0005, "step": 24718 }, { "epoch": 2.679856895056375, "grad_norm": 0.24096441268920898, "learning_rate": 2.142028144494415e-06, "loss": 0.0048, "step": 24719 }, { "epoch": 2.6799653078924544, "grad_norm": 3.2675297260284424, "learning_rate": 2.1413027709270274e-06, "loss": 0.0218, "step": 24720 }, { "epoch": 2.6800737207285343, "grad_norm": 0.2731646001338959, "learning_rate": 2.1405773973596404e-06, "loss": 0.0047, "step": 24721 }, { "epoch": 2.680182133564614, "grad_norm": 0.0773051530122757, "learning_rate": 2.1398520237922534e-06, "loss": 0.0003, "step": 24722 }, { "epoch": 2.680290546400694, "grad_norm": 0.308633416891098, "learning_rate": 2.139126650224866e-06, "loss": 0.0114, "step": 24723 }, { "epoch": 2.680398959236774, "grad_norm": 0.23316921293735504, "learning_rate": 2.138401276657479e-06, "loss": 0.016, "step": 24724 }, { "epoch": 2.6805073720728534, "grad_norm": 0.5062664151191711, "learning_rate": 2.1376759030900915e-06, "loss": 0.0119, "step": 24725 }, { "epoch": 2.6806157849089334, "grad_norm": 3.192791700363159, "learning_rate": 2.1369505295227045e-06, "loss": 0.0395, "step": 24726 }, { "epoch": 2.680724197745013, "grad_norm": 0.224795863032341, "learning_rate": 2.136225155955317e-06, "loss": 0.0038, "step": 24727 }, { "epoch": 2.680832610581093, "grad_norm": 0.5975081324577332, "learning_rate": 2.13549978238793e-06, "loss": 0.0177, "step": 24728 }, { "epoch": 2.6809410234171724, "grad_norm": 11.77889633178711, "learning_rate": 2.1347744088205426e-06, "loss": 0.028, "step": 24729 }, { "epoch": 2.6810494362532524, "grad_norm": 0.8339849710464478, "learning_rate": 2.1340490352531556e-06, "loss": 0.0065, "step": 24730 }, { "epoch": 2.681157849089332, "grad_norm": 0.15008355677127838, "learning_rate": 2.133323661685768e-06, "loss": 0.002, "step": 24731 }, { "epoch": 2.681266261925412, "grad_norm": 0.9664345979690552, "learning_rate": 2.132598288118381e-06, "loss": 0.0123, "step": 24732 }, { "epoch": 2.681374674761492, "grad_norm": 0.8800044059753418, "learning_rate": 2.1318729145509937e-06, "loss": 0.0007, "step": 24733 }, { "epoch": 2.6814830875975715, "grad_norm": 0.7546797394752502, "learning_rate": 2.1311475409836067e-06, "loss": 0.019, "step": 24734 }, { "epoch": 2.6815915004336515, "grad_norm": 0.30496013164520264, "learning_rate": 2.1304221674162197e-06, "loss": 0.0034, "step": 24735 }, { "epoch": 2.681699913269731, "grad_norm": 0.8491067886352539, "learning_rate": 2.1296967938488323e-06, "loss": 0.0187, "step": 24736 }, { "epoch": 2.681808326105811, "grad_norm": 5.332104206085205, "learning_rate": 2.1289714202814453e-06, "loss": 0.0265, "step": 24737 }, { "epoch": 2.681916738941891, "grad_norm": 9.9874849319458, "learning_rate": 2.128246046714058e-06, "loss": 0.017, "step": 24738 }, { "epoch": 2.6820251517779705, "grad_norm": 0.027603410184383392, "learning_rate": 2.127520673146671e-06, "loss": 0.0002, "step": 24739 }, { "epoch": 2.68213356461405, "grad_norm": 0.26577451825141907, "learning_rate": 2.1267952995792834e-06, "loss": 0.0086, "step": 24740 }, { "epoch": 2.68224197745013, "grad_norm": 0.26839494705200195, "learning_rate": 2.1260699260118964e-06, "loss": 0.0055, "step": 24741 }, { "epoch": 2.68235039028621, "grad_norm": 0.00727357342839241, "learning_rate": 2.125344552444509e-06, "loss": 0.0001, "step": 24742 }, { "epoch": 2.6824588031222896, "grad_norm": 0.1637095957994461, "learning_rate": 2.124619178877122e-06, "loss": 0.0023, "step": 24743 }, { "epoch": 2.6825672159583696, "grad_norm": 7.9707794189453125, "learning_rate": 2.1238938053097345e-06, "loss": 0.0234, "step": 24744 }, { "epoch": 2.682675628794449, "grad_norm": 0.43873730301856995, "learning_rate": 2.1231684317423475e-06, "loss": 0.0098, "step": 24745 }, { "epoch": 2.682784041630529, "grad_norm": 0.40331053733825684, "learning_rate": 2.12244305817496e-06, "loss": 0.0134, "step": 24746 }, { "epoch": 2.682892454466609, "grad_norm": 0.5474538207054138, "learning_rate": 2.121717684607573e-06, "loss": 0.026, "step": 24747 }, { "epoch": 2.6830008673026886, "grad_norm": 2.9373600482940674, "learning_rate": 2.120992311040186e-06, "loss": 0.0323, "step": 24748 }, { "epoch": 2.6831092801387686, "grad_norm": 0.04174415022134781, "learning_rate": 2.1202669374727986e-06, "loss": 0.0002, "step": 24749 }, { "epoch": 2.683217692974848, "grad_norm": 0.035576168447732925, "learning_rate": 2.1195415639054116e-06, "loss": 0.0003, "step": 24750 }, { "epoch": 2.683326105810928, "grad_norm": 0.24813631176948547, "learning_rate": 2.118816190338024e-06, "loss": 0.0137, "step": 24751 }, { "epoch": 2.6834345186470077, "grad_norm": 0.4798569679260254, "learning_rate": 2.118090816770637e-06, "loss": 0.0087, "step": 24752 }, { "epoch": 2.6835429314830876, "grad_norm": 0.05090942978858948, "learning_rate": 2.1173654432032497e-06, "loss": 0.0003, "step": 24753 }, { "epoch": 2.683651344319167, "grad_norm": 2.5303711891174316, "learning_rate": 2.1166400696358627e-06, "loss": 0.0098, "step": 24754 }, { "epoch": 2.683759757155247, "grad_norm": 0.013027963228523731, "learning_rate": 2.1159146960684753e-06, "loss": 0.0001, "step": 24755 }, { "epoch": 2.683868169991327, "grad_norm": 0.5006658434867859, "learning_rate": 2.1151893225010883e-06, "loss": 0.0131, "step": 24756 }, { "epoch": 2.6839765828274067, "grad_norm": 12.343523979187012, "learning_rate": 2.114463948933701e-06, "loss": 0.0554, "step": 24757 }, { "epoch": 2.6840849956634867, "grad_norm": 0.003975770901888609, "learning_rate": 2.113738575366314e-06, "loss": 0.0, "step": 24758 }, { "epoch": 2.6841934084995662, "grad_norm": 4.550819396972656, "learning_rate": 2.1130132017989264e-06, "loss": 0.0386, "step": 24759 }, { "epoch": 2.684301821335646, "grad_norm": 12.568496704101562, "learning_rate": 2.1122878282315394e-06, "loss": 0.0112, "step": 24760 }, { "epoch": 2.684410234171726, "grad_norm": 0.12350708991289139, "learning_rate": 2.1115624546641524e-06, "loss": 0.0098, "step": 24761 }, { "epoch": 2.6845186470078057, "grad_norm": 1.161120057106018, "learning_rate": 2.110837081096765e-06, "loss": 0.0448, "step": 24762 }, { "epoch": 2.6846270598438853, "grad_norm": 0.4759390354156494, "learning_rate": 2.110111707529378e-06, "loss": 0.0014, "step": 24763 }, { "epoch": 2.6847354726799653, "grad_norm": 4.539470195770264, "learning_rate": 2.1093863339619905e-06, "loss": 0.0197, "step": 24764 }, { "epoch": 2.6848438855160452, "grad_norm": 0.3357287347316742, "learning_rate": 2.1086609603946035e-06, "loss": 0.0175, "step": 24765 }, { "epoch": 2.684952298352125, "grad_norm": 0.6247153878211975, "learning_rate": 2.107935586827216e-06, "loss": 0.0115, "step": 24766 }, { "epoch": 2.6850607111882048, "grad_norm": 0.2813856303691864, "learning_rate": 2.107210213259829e-06, "loss": 0.0051, "step": 24767 }, { "epoch": 2.6851691240242843, "grad_norm": 0.581969141960144, "learning_rate": 2.1064848396924416e-06, "loss": 0.0101, "step": 24768 }, { "epoch": 2.6852775368603643, "grad_norm": 5.433154582977295, "learning_rate": 2.1057594661250546e-06, "loss": 0.0517, "step": 24769 }, { "epoch": 2.6853859496964443, "grad_norm": 0.028274092823266983, "learning_rate": 2.105034092557667e-06, "loss": 0.0002, "step": 24770 }, { "epoch": 2.685494362532524, "grad_norm": 0.05406472459435463, "learning_rate": 2.10430871899028e-06, "loss": 0.0003, "step": 24771 }, { "epoch": 2.6856027753686034, "grad_norm": 0.26628756523132324, "learning_rate": 2.1035833454228927e-06, "loss": 0.0056, "step": 24772 }, { "epoch": 2.6857111882046834, "grad_norm": 0.21278029680252075, "learning_rate": 2.1028579718555057e-06, "loss": 0.0004, "step": 24773 }, { "epoch": 2.6858196010407633, "grad_norm": 0.5376214981079102, "learning_rate": 2.1021325982881187e-06, "loss": 0.0066, "step": 24774 }, { "epoch": 2.685928013876843, "grad_norm": 0.044729266315698624, "learning_rate": 2.1014072247207313e-06, "loss": 0.0003, "step": 24775 }, { "epoch": 2.686036426712923, "grad_norm": 0.010121839120984077, "learning_rate": 2.1006818511533443e-06, "loss": 0.0001, "step": 24776 }, { "epoch": 2.6861448395490024, "grad_norm": 0.5614484548568726, "learning_rate": 2.099956477585957e-06, "loss": 0.028, "step": 24777 }, { "epoch": 2.6862532523850824, "grad_norm": 0.10420166701078415, "learning_rate": 2.09923110401857e-06, "loss": 0.0006, "step": 24778 }, { "epoch": 2.6863616652211624, "grad_norm": 1.3267972469329834, "learning_rate": 2.0985057304511824e-06, "loss": 0.0369, "step": 24779 }, { "epoch": 2.686470078057242, "grad_norm": 0.37987658381462097, "learning_rate": 2.0977803568837954e-06, "loss": 0.0138, "step": 24780 }, { "epoch": 2.686578490893322, "grad_norm": 0.284512460231781, "learning_rate": 2.097054983316408e-06, "loss": 0.0061, "step": 24781 }, { "epoch": 2.6866869037294014, "grad_norm": 1.8802226781845093, "learning_rate": 2.096329609749021e-06, "loss": 0.0095, "step": 24782 }, { "epoch": 2.6867953165654814, "grad_norm": 0.3280951678752899, "learning_rate": 2.0956042361816335e-06, "loss": 0.0102, "step": 24783 }, { "epoch": 2.6869037294015614, "grad_norm": 0.29410508275032043, "learning_rate": 2.0948788626142465e-06, "loss": 0.0061, "step": 24784 }, { "epoch": 2.687012142237641, "grad_norm": 0.40904751420021057, "learning_rate": 2.094153489046859e-06, "loss": 0.0105, "step": 24785 }, { "epoch": 2.6871205550737205, "grad_norm": 0.03322479501366615, "learning_rate": 2.093428115479472e-06, "loss": 0.0002, "step": 24786 }, { "epoch": 2.6872289679098005, "grad_norm": 5.510204792022705, "learning_rate": 2.092702741912085e-06, "loss": 0.0327, "step": 24787 }, { "epoch": 2.6873373807458805, "grad_norm": 0.26325783133506775, "learning_rate": 2.0919773683446976e-06, "loss": 0.0058, "step": 24788 }, { "epoch": 2.68744579358196, "grad_norm": 0.08016100525856018, "learning_rate": 2.0912519947773106e-06, "loss": 0.0005, "step": 24789 }, { "epoch": 2.68755420641804, "grad_norm": 1.1447489261627197, "learning_rate": 2.090526621209923e-06, "loss": 0.0104, "step": 24790 }, { "epoch": 2.6876626192541195, "grad_norm": 3.840074062347412, "learning_rate": 2.089801247642536e-06, "loss": 0.0137, "step": 24791 }, { "epoch": 2.6877710320901995, "grad_norm": 0.27405139803886414, "learning_rate": 2.0890758740751487e-06, "loss": 0.0055, "step": 24792 }, { "epoch": 2.6878794449262795, "grad_norm": 0.6046793460845947, "learning_rate": 2.0883505005077617e-06, "loss": 0.015, "step": 24793 }, { "epoch": 2.687987857762359, "grad_norm": 0.3370850384235382, "learning_rate": 2.0876251269403743e-06, "loss": 0.0094, "step": 24794 }, { "epoch": 2.6880962705984386, "grad_norm": 0.507087230682373, "learning_rate": 2.0868997533729873e-06, "loss": 0.016, "step": 24795 }, { "epoch": 2.6882046834345186, "grad_norm": 19.843719482421875, "learning_rate": 2.0861743798056e-06, "loss": 0.0339, "step": 24796 }, { "epoch": 2.6883130962705986, "grad_norm": 0.042588818818330765, "learning_rate": 2.085449006238213e-06, "loss": 0.0004, "step": 24797 }, { "epoch": 2.688421509106678, "grad_norm": 0.47079193592071533, "learning_rate": 2.0847236326708254e-06, "loss": 0.0149, "step": 24798 }, { "epoch": 2.688529921942758, "grad_norm": 4.433749675750732, "learning_rate": 2.0839982591034384e-06, "loss": 0.0022, "step": 24799 }, { "epoch": 2.6886383347788376, "grad_norm": 0.8079765439033508, "learning_rate": 2.0832728855360514e-06, "loss": 0.023, "step": 24800 }, { "epoch": 2.6887467476149176, "grad_norm": 0.778160572052002, "learning_rate": 2.082547511968664e-06, "loss": 0.0142, "step": 24801 }, { "epoch": 2.6888551604509976, "grad_norm": 6.947807788848877, "learning_rate": 2.081822138401277e-06, "loss": 0.0043, "step": 24802 }, { "epoch": 2.688963573287077, "grad_norm": 4.781380653381348, "learning_rate": 2.0810967648338895e-06, "loss": 0.0077, "step": 24803 }, { "epoch": 2.689071986123157, "grad_norm": 0.045089878141880035, "learning_rate": 2.0803713912665025e-06, "loss": 0.0002, "step": 24804 }, { "epoch": 2.6891803989592367, "grad_norm": 0.25335147976875305, "learning_rate": 2.079646017699115e-06, "loss": 0.005, "step": 24805 }, { "epoch": 2.6892888117953166, "grad_norm": 0.007602374069392681, "learning_rate": 2.078920644131728e-06, "loss": 0.0001, "step": 24806 }, { "epoch": 2.6893972246313966, "grad_norm": 0.46958282589912415, "learning_rate": 2.0781952705643406e-06, "loss": 0.0047, "step": 24807 }, { "epoch": 2.689505637467476, "grad_norm": 0.38595497608184814, "learning_rate": 2.0774698969969536e-06, "loss": 0.0094, "step": 24808 }, { "epoch": 2.6896140503035557, "grad_norm": 1.2912894487380981, "learning_rate": 2.076744523429566e-06, "loss": 0.0318, "step": 24809 }, { "epoch": 2.6897224631396357, "grad_norm": 0.11840865761041641, "learning_rate": 2.076019149862179e-06, "loss": 0.0003, "step": 24810 }, { "epoch": 2.6898308759757157, "grad_norm": 3.5375382900238037, "learning_rate": 2.0752937762947917e-06, "loss": 0.0216, "step": 24811 }, { "epoch": 2.6899392888117952, "grad_norm": 4.401638507843018, "learning_rate": 2.0745684027274047e-06, "loss": 0.051, "step": 24812 }, { "epoch": 2.690047701647875, "grad_norm": 2.6135213375091553, "learning_rate": 2.0738430291600177e-06, "loss": 0.0023, "step": 24813 }, { "epoch": 2.6901561144839548, "grad_norm": 2.0033774375915527, "learning_rate": 2.0731176555926303e-06, "loss": 0.0449, "step": 24814 }, { "epoch": 2.6902645273200347, "grad_norm": 1.1202590465545654, "learning_rate": 2.0723922820252433e-06, "loss": 0.0306, "step": 24815 }, { "epoch": 2.6903729401561147, "grad_norm": 0.2804369032382965, "learning_rate": 2.071666908457856e-06, "loss": 0.0058, "step": 24816 }, { "epoch": 2.6904813529921943, "grad_norm": 0.6491496562957764, "learning_rate": 2.070941534890469e-06, "loss": 0.0141, "step": 24817 }, { "epoch": 2.690589765828274, "grad_norm": 0.009121943265199661, "learning_rate": 2.0702161613230814e-06, "loss": 0.0001, "step": 24818 }, { "epoch": 2.690698178664354, "grad_norm": 0.7093496918678284, "learning_rate": 2.0694907877556944e-06, "loss": 0.0234, "step": 24819 }, { "epoch": 2.6908065915004338, "grad_norm": 0.3521670401096344, "learning_rate": 2.068765414188307e-06, "loss": 0.0116, "step": 24820 }, { "epoch": 2.6909150043365133, "grad_norm": 0.018163859844207764, "learning_rate": 2.06804004062092e-06, "loss": 0.0001, "step": 24821 }, { "epoch": 2.6910234171725933, "grad_norm": 5.265886306762695, "learning_rate": 2.0673146670535325e-06, "loss": 0.0224, "step": 24822 }, { "epoch": 2.691131830008673, "grad_norm": 0.07519183307886124, "learning_rate": 2.0665892934861455e-06, "loss": 0.0004, "step": 24823 }, { "epoch": 2.691240242844753, "grad_norm": 0.2879883050918579, "learning_rate": 2.065863919918758e-06, "loss": 0.009, "step": 24824 }, { "epoch": 2.691348655680833, "grad_norm": 6.317279815673828, "learning_rate": 2.065138546351371e-06, "loss": 0.0103, "step": 24825 }, { "epoch": 2.6914570685169124, "grad_norm": 0.07164432853460312, "learning_rate": 2.064413172783984e-06, "loss": 0.0002, "step": 24826 }, { "epoch": 2.6915654813529923, "grad_norm": 0.029458913952112198, "learning_rate": 2.063687799216597e-06, "loss": 0.0002, "step": 24827 }, { "epoch": 2.691673894189072, "grad_norm": 3.195535659790039, "learning_rate": 2.0629624256492096e-06, "loss": 0.0341, "step": 24828 }, { "epoch": 2.691782307025152, "grad_norm": 18.053312301635742, "learning_rate": 2.062237052081822e-06, "loss": 0.0234, "step": 24829 }, { "epoch": 2.6918907198612314, "grad_norm": 0.12235461920499802, "learning_rate": 2.061511678514435e-06, "loss": 0.0005, "step": 24830 }, { "epoch": 2.6919991326973114, "grad_norm": 4.667356014251709, "learning_rate": 2.0607863049470477e-06, "loss": 0.0263, "step": 24831 }, { "epoch": 2.692107545533391, "grad_norm": 0.26033177971839905, "learning_rate": 2.0600609313796607e-06, "loss": 0.0051, "step": 24832 }, { "epoch": 2.692215958369471, "grad_norm": 19.11716651916504, "learning_rate": 2.0593355578122733e-06, "loss": 0.0355, "step": 24833 }, { "epoch": 2.692324371205551, "grad_norm": 2.467613458633423, "learning_rate": 2.0586101842448863e-06, "loss": 0.004, "step": 24834 }, { "epoch": 2.6924327840416304, "grad_norm": 0.5734017491340637, "learning_rate": 2.057884810677499e-06, "loss": 0.0296, "step": 24835 }, { "epoch": 2.6925411968777104, "grad_norm": 0.3586733043193817, "learning_rate": 2.057159437110112e-06, "loss": 0.0129, "step": 24836 }, { "epoch": 2.69264960971379, "grad_norm": 0.13290221989154816, "learning_rate": 2.0564340635427244e-06, "loss": 0.0006, "step": 24837 }, { "epoch": 2.69275802254987, "grad_norm": 1.5081661939620972, "learning_rate": 2.0557086899753374e-06, "loss": 0.0066, "step": 24838 }, { "epoch": 2.69286643538595, "grad_norm": 4.608724594116211, "learning_rate": 2.0549833164079504e-06, "loss": 0.0117, "step": 24839 }, { "epoch": 2.6929748482220295, "grad_norm": 0.5521579384803772, "learning_rate": 2.0542579428405634e-06, "loss": 0.0059, "step": 24840 }, { "epoch": 2.693083261058109, "grad_norm": 9.507414817810059, "learning_rate": 2.053532569273176e-06, "loss": 0.0199, "step": 24841 }, { "epoch": 2.693191673894189, "grad_norm": 0.036337364464998245, "learning_rate": 2.0528071957057885e-06, "loss": 0.0002, "step": 24842 }, { "epoch": 2.693300086730269, "grad_norm": 0.02369946800172329, "learning_rate": 2.0520818221384015e-06, "loss": 0.0002, "step": 24843 }, { "epoch": 2.6934084995663485, "grad_norm": 0.34247666597366333, "learning_rate": 2.051356448571014e-06, "loss": 0.0123, "step": 24844 }, { "epoch": 2.6935169124024285, "grad_norm": 5.400280475616455, "learning_rate": 2.050631075003627e-06, "loss": 0.0834, "step": 24845 }, { "epoch": 2.693625325238508, "grad_norm": 3.0392229557037354, "learning_rate": 2.0499057014362396e-06, "loss": 0.042, "step": 24846 }, { "epoch": 2.693733738074588, "grad_norm": 0.2622097432613373, "learning_rate": 2.0491803278688526e-06, "loss": 0.0056, "step": 24847 }, { "epoch": 2.693842150910668, "grad_norm": 0.570523202419281, "learning_rate": 2.048454954301465e-06, "loss": 0.0079, "step": 24848 }, { "epoch": 2.6939505637467476, "grad_norm": 1.7705309391021729, "learning_rate": 2.047729580734078e-06, "loss": 0.0186, "step": 24849 }, { "epoch": 2.6940589765828276, "grad_norm": 0.11171811074018478, "learning_rate": 2.0470042071666908e-06, "loss": 0.0007, "step": 24850 }, { "epoch": 2.694167389418907, "grad_norm": 0.27106335759162903, "learning_rate": 2.0462788335993037e-06, "loss": 0.0056, "step": 24851 }, { "epoch": 2.694275802254987, "grad_norm": 2.663902521133423, "learning_rate": 2.0455534600319167e-06, "loss": 0.0363, "step": 24852 }, { "epoch": 2.6943842150910666, "grad_norm": 0.7569859623908997, "learning_rate": 2.0448280864645297e-06, "loss": 0.022, "step": 24853 }, { "epoch": 2.6944926279271466, "grad_norm": 2.3689568042755127, "learning_rate": 2.0441027128971423e-06, "loss": 0.0215, "step": 24854 }, { "epoch": 2.694601040763226, "grad_norm": 0.19422687590122223, "learning_rate": 2.0433773393297553e-06, "loss": 0.0005, "step": 24855 }, { "epoch": 2.694709453599306, "grad_norm": 0.8386445641517639, "learning_rate": 2.042651965762368e-06, "loss": 0.014, "step": 24856 }, { "epoch": 2.694817866435386, "grad_norm": 1.0479313135147095, "learning_rate": 2.0419265921949804e-06, "loss": 0.0222, "step": 24857 }, { "epoch": 2.6949262792714657, "grad_norm": 1.3461275100708008, "learning_rate": 2.0412012186275934e-06, "loss": 0.0045, "step": 24858 }, { "epoch": 2.6950346921075456, "grad_norm": 6.625476360321045, "learning_rate": 2.040475845060206e-06, "loss": 0.0172, "step": 24859 }, { "epoch": 2.695143104943625, "grad_norm": 0.475344717502594, "learning_rate": 2.039750471492819e-06, "loss": 0.0097, "step": 24860 }, { "epoch": 2.695251517779705, "grad_norm": 0.31282877922058105, "learning_rate": 2.0390250979254315e-06, "loss": 0.0053, "step": 24861 }, { "epoch": 2.695359930615785, "grad_norm": 1.035107135772705, "learning_rate": 2.0382997243580445e-06, "loss": 0.0474, "step": 24862 }, { "epoch": 2.6954683434518647, "grad_norm": 3.661681652069092, "learning_rate": 2.037574350790657e-06, "loss": 0.0763, "step": 24863 }, { "epoch": 2.6955767562879442, "grad_norm": 0.011294164694845676, "learning_rate": 2.03684897722327e-06, "loss": 0.0001, "step": 24864 }, { "epoch": 2.6956851691240242, "grad_norm": 0.12867894768714905, "learning_rate": 2.036123603655883e-06, "loss": 0.0003, "step": 24865 }, { "epoch": 2.695793581960104, "grad_norm": 0.08071818947792053, "learning_rate": 2.035398230088496e-06, "loss": 0.0005, "step": 24866 }, { "epoch": 2.6959019947961838, "grad_norm": 0.05710937827825546, "learning_rate": 2.0346728565211086e-06, "loss": 0.0003, "step": 24867 }, { "epoch": 2.6960104076322637, "grad_norm": 0.0023787443060427904, "learning_rate": 2.0339474829537216e-06, "loss": 0.0, "step": 24868 }, { "epoch": 2.6961188204683433, "grad_norm": 0.35156458616256714, "learning_rate": 2.033222109386334e-06, "loss": 0.0104, "step": 24869 }, { "epoch": 2.6962272333044233, "grad_norm": 0.031834669411182404, "learning_rate": 2.0324967358189467e-06, "loss": 0.0004, "step": 24870 }, { "epoch": 2.6963356461405033, "grad_norm": 0.3618593215942383, "learning_rate": 2.0317713622515597e-06, "loss": 0.0114, "step": 24871 }, { "epoch": 2.696444058976583, "grad_norm": 0.26272761821746826, "learning_rate": 2.0310459886841723e-06, "loss": 0.0059, "step": 24872 }, { "epoch": 2.6965524718126628, "grad_norm": 0.2871510088443756, "learning_rate": 2.0303206151167853e-06, "loss": 0.0059, "step": 24873 }, { "epoch": 2.6966608846487423, "grad_norm": 0.26588955521583557, "learning_rate": 2.029595241549398e-06, "loss": 0.0057, "step": 24874 }, { "epoch": 2.6967692974848223, "grad_norm": 1.4035676717758179, "learning_rate": 2.028869867982011e-06, "loss": 0.0715, "step": 24875 }, { "epoch": 2.696877710320902, "grad_norm": 5.369513034820557, "learning_rate": 2.028144494414624e-06, "loss": 0.0055, "step": 24876 }, { "epoch": 2.696986123156982, "grad_norm": 0.2675173282623291, "learning_rate": 2.0274191208472364e-06, "loss": 0.0047, "step": 24877 }, { "epoch": 2.6970945359930614, "grad_norm": 0.26338398456573486, "learning_rate": 2.0266937472798494e-06, "loss": 0.0057, "step": 24878 }, { "epoch": 2.6972029488291414, "grad_norm": 0.06888160109519958, "learning_rate": 2.0259683737124624e-06, "loss": 0.0006, "step": 24879 }, { "epoch": 2.6973113616652213, "grad_norm": 0.7257556319236755, "learning_rate": 2.025243000145075e-06, "loss": 0.0046, "step": 24880 }, { "epoch": 2.697419774501301, "grad_norm": 0.25661879777908325, "learning_rate": 2.024517626577688e-06, "loss": 0.0055, "step": 24881 }, { "epoch": 2.697528187337381, "grad_norm": 2.2229554653167725, "learning_rate": 2.0237922530103005e-06, "loss": 0.0488, "step": 24882 }, { "epoch": 2.6976366001734604, "grad_norm": 0.02868053875863552, "learning_rate": 2.023066879442913e-06, "loss": 0.0002, "step": 24883 }, { "epoch": 2.6977450130095404, "grad_norm": 0.5124155879020691, "learning_rate": 2.022341505875526e-06, "loss": 0.0129, "step": 24884 }, { "epoch": 2.6978534258456204, "grad_norm": 0.4914494454860687, "learning_rate": 2.0216161323081386e-06, "loss": 0.0014, "step": 24885 }, { "epoch": 2.6979618386817, "grad_norm": 20.95517349243164, "learning_rate": 2.0208907587407516e-06, "loss": 0.0135, "step": 24886 }, { "epoch": 2.6980702515177795, "grad_norm": 0.054284170269966125, "learning_rate": 2.020165385173364e-06, "loss": 0.0004, "step": 24887 }, { "epoch": 2.6981786643538594, "grad_norm": 0.31783634424209595, "learning_rate": 2.019440011605977e-06, "loss": 0.0086, "step": 24888 }, { "epoch": 2.6982870771899394, "grad_norm": 1.3330087661743164, "learning_rate": 2.01871463803859e-06, "loss": 0.0015, "step": 24889 }, { "epoch": 2.698395490026019, "grad_norm": 0.03707537427544594, "learning_rate": 2.0179892644712027e-06, "loss": 0.0003, "step": 24890 }, { "epoch": 2.698503902862099, "grad_norm": 0.07829725742340088, "learning_rate": 2.0172638909038157e-06, "loss": 0.001, "step": 24891 }, { "epoch": 2.6986123156981785, "grad_norm": 0.02501795068383217, "learning_rate": 2.0165385173364287e-06, "loss": 0.0003, "step": 24892 }, { "epoch": 2.6987207285342585, "grad_norm": 0.056098077446222305, "learning_rate": 2.0158131437690413e-06, "loss": 0.0004, "step": 24893 }, { "epoch": 2.6988291413703385, "grad_norm": 0.2906002700328827, "learning_rate": 2.0150877702016543e-06, "loss": 0.0047, "step": 24894 }, { "epoch": 2.698937554206418, "grad_norm": 0.26003503799438477, "learning_rate": 2.014362396634267e-06, "loss": 0.0057, "step": 24895 }, { "epoch": 2.699045967042498, "grad_norm": 9.539328575134277, "learning_rate": 2.0136370230668794e-06, "loss": 0.0342, "step": 24896 }, { "epoch": 2.6991543798785775, "grad_norm": 0.178102046251297, "learning_rate": 2.0129116494994924e-06, "loss": 0.0028, "step": 24897 }, { "epoch": 2.6992627927146575, "grad_norm": 14.479920387268066, "learning_rate": 2.012186275932105e-06, "loss": 0.0505, "step": 24898 }, { "epoch": 2.699371205550737, "grad_norm": 0.01967986300587654, "learning_rate": 2.011460902364718e-06, "loss": 0.0002, "step": 24899 }, { "epoch": 2.699479618386817, "grad_norm": 0.02336605079472065, "learning_rate": 2.0107355287973305e-06, "loss": 0.0001, "step": 24900 }, { "epoch": 2.6995880312228966, "grad_norm": 0.0401264950633049, "learning_rate": 2.0100101552299435e-06, "loss": 0.0003, "step": 24901 }, { "epoch": 2.6996964440589766, "grad_norm": 0.3784106969833374, "learning_rate": 2.0092847816625565e-06, "loss": 0.0086, "step": 24902 }, { "epoch": 2.6998048568950566, "grad_norm": 0.5555821061134338, "learning_rate": 2.008559408095169e-06, "loss": 0.0143, "step": 24903 }, { "epoch": 2.699913269731136, "grad_norm": 0.21581557393074036, "learning_rate": 2.007834034527782e-06, "loss": 0.0036, "step": 24904 }, { "epoch": 2.700021682567216, "grad_norm": 1.5027778148651123, "learning_rate": 2.007108660960395e-06, "loss": 0.0037, "step": 24905 }, { "epoch": 2.7001300954032956, "grad_norm": 0.2550964653491974, "learning_rate": 2.0063832873930076e-06, "loss": 0.0155, "step": 24906 }, { "epoch": 2.7002385082393756, "grad_norm": 0.5284244418144226, "learning_rate": 2.0056579138256206e-06, "loss": 0.001, "step": 24907 }, { "epoch": 2.7003469210754556, "grad_norm": 0.3101631700992584, "learning_rate": 2.004932540258233e-06, "loss": 0.0069, "step": 24908 }, { "epoch": 2.700455333911535, "grad_norm": 1.0795172452926636, "learning_rate": 2.004207166690846e-06, "loss": 0.0188, "step": 24909 }, { "epoch": 2.7005637467476147, "grad_norm": 0.3453894853591919, "learning_rate": 2.0034817931234587e-06, "loss": 0.0202, "step": 24910 }, { "epoch": 2.7006721595836947, "grad_norm": 0.05793718248605728, "learning_rate": 2.0027564195560713e-06, "loss": 0.0002, "step": 24911 }, { "epoch": 2.7007805724197746, "grad_norm": 0.03168398141860962, "learning_rate": 2.0020310459886843e-06, "loss": 0.0002, "step": 24912 }, { "epoch": 2.700888985255854, "grad_norm": 3.627794027328491, "learning_rate": 2.001305672421297e-06, "loss": 0.0391, "step": 24913 }, { "epoch": 2.700997398091934, "grad_norm": 0.6044625639915466, "learning_rate": 2.00058029885391e-06, "loss": 0.0112, "step": 24914 }, { "epoch": 2.7011058109280137, "grad_norm": 0.004748641513288021, "learning_rate": 1.999854925286523e-06, "loss": 0.0001, "step": 24915 }, { "epoch": 2.7012142237640937, "grad_norm": 1.7811332941055298, "learning_rate": 1.9991295517191354e-06, "loss": 0.0102, "step": 24916 }, { "epoch": 2.7013226366001737, "grad_norm": 0.6301067471504211, "learning_rate": 1.9984041781517484e-06, "loss": 0.0139, "step": 24917 }, { "epoch": 2.7014310494362532, "grad_norm": 3.990626811981201, "learning_rate": 1.9976788045843614e-06, "loss": 0.0096, "step": 24918 }, { "epoch": 2.7015394622723328, "grad_norm": 0.1665152609348297, "learning_rate": 1.996953431016974e-06, "loss": 0.0006, "step": 24919 }, { "epoch": 2.7016478751084128, "grad_norm": 0.00973310973495245, "learning_rate": 1.996228057449587e-06, "loss": 0.0001, "step": 24920 }, { "epoch": 2.7017562879444927, "grad_norm": 0.6414057612419128, "learning_rate": 1.9955026838821995e-06, "loss": 0.0056, "step": 24921 }, { "epoch": 2.7018647007805723, "grad_norm": 0.44853541254997253, "learning_rate": 1.9947773103148125e-06, "loss": 0.0079, "step": 24922 }, { "epoch": 2.7019731136166523, "grad_norm": 1.4298945665359497, "learning_rate": 1.994051936747425e-06, "loss": 0.0423, "step": 24923 }, { "epoch": 2.702081526452732, "grad_norm": 8.961173057556152, "learning_rate": 1.9933265631800376e-06, "loss": 0.0118, "step": 24924 }, { "epoch": 2.702189939288812, "grad_norm": 0.3651478588581085, "learning_rate": 1.9926011896126506e-06, "loss": 0.0064, "step": 24925 }, { "epoch": 2.702298352124892, "grad_norm": 0.03328653424978256, "learning_rate": 1.991875816045263e-06, "loss": 0.0003, "step": 24926 }, { "epoch": 2.7024067649609713, "grad_norm": 1.722813606262207, "learning_rate": 1.991150442477876e-06, "loss": 0.0451, "step": 24927 }, { "epoch": 2.7025151777970513, "grad_norm": 4.96359395980835, "learning_rate": 1.990425068910489e-06, "loss": 0.0205, "step": 24928 }, { "epoch": 2.702623590633131, "grad_norm": 0.027471622452139854, "learning_rate": 1.9896996953431017e-06, "loss": 0.0001, "step": 24929 }, { "epoch": 2.702732003469211, "grad_norm": 0.010630825534462929, "learning_rate": 1.9889743217757147e-06, "loss": 0.0001, "step": 24930 }, { "epoch": 2.702840416305291, "grad_norm": 0.0664733275771141, "learning_rate": 1.9882489482083277e-06, "loss": 0.0006, "step": 24931 }, { "epoch": 2.7029488291413704, "grad_norm": 1.0967705249786377, "learning_rate": 1.9875235746409403e-06, "loss": 0.0012, "step": 24932 }, { "epoch": 2.70305724197745, "grad_norm": 2.4026119709014893, "learning_rate": 1.9867982010735533e-06, "loss": 0.034, "step": 24933 }, { "epoch": 2.70316565481353, "grad_norm": 0.9354049563407898, "learning_rate": 1.986072827506166e-06, "loss": 0.0274, "step": 24934 }, { "epoch": 2.70327406764961, "grad_norm": 0.024913448840379715, "learning_rate": 1.985347453938779e-06, "loss": 0.0002, "step": 24935 }, { "epoch": 2.7033824804856894, "grad_norm": 0.05099092796444893, "learning_rate": 1.9846220803713914e-06, "loss": 0.0002, "step": 24936 }, { "epoch": 2.7034908933217694, "grad_norm": 4.24196195602417, "learning_rate": 1.983896706804004e-06, "loss": 0.0075, "step": 24937 }, { "epoch": 2.703599306157849, "grad_norm": 0.1997576355934143, "learning_rate": 1.983171333236617e-06, "loss": 0.0011, "step": 24938 }, { "epoch": 2.703707718993929, "grad_norm": 0.3797779083251953, "learning_rate": 1.9824459596692295e-06, "loss": 0.0008, "step": 24939 }, { "epoch": 2.703816131830009, "grad_norm": 0.8985378742218018, "learning_rate": 1.9817205861018425e-06, "loss": 0.0108, "step": 24940 }, { "epoch": 2.7039245446660884, "grad_norm": 0.4579298496246338, "learning_rate": 1.9809952125344555e-06, "loss": 0.0121, "step": 24941 }, { "epoch": 2.704032957502168, "grad_norm": 0.05921012908220291, "learning_rate": 1.980269838967068e-06, "loss": 0.0001, "step": 24942 }, { "epoch": 2.704141370338248, "grad_norm": 0.040603917092084885, "learning_rate": 1.979544465399681e-06, "loss": 0.0005, "step": 24943 }, { "epoch": 2.704249783174328, "grad_norm": 0.3314601182937622, "learning_rate": 1.978819091832294e-06, "loss": 0.007, "step": 24944 }, { "epoch": 2.7043581960104075, "grad_norm": 0.03600176423788071, "learning_rate": 1.9780937182649066e-06, "loss": 0.0003, "step": 24945 }, { "epoch": 2.7044666088464875, "grad_norm": 0.30516543984413147, "learning_rate": 1.9773683446975196e-06, "loss": 0.0097, "step": 24946 }, { "epoch": 2.704575021682567, "grad_norm": 0.1976819634437561, "learning_rate": 1.976642971130132e-06, "loss": 0.0004, "step": 24947 }, { "epoch": 2.704683434518647, "grad_norm": 0.37186887860298157, "learning_rate": 1.975917597562745e-06, "loss": 0.0071, "step": 24948 }, { "epoch": 2.704791847354727, "grad_norm": 4.510384559631348, "learning_rate": 1.9751922239953577e-06, "loss": 0.0176, "step": 24949 }, { "epoch": 2.7049002601908065, "grad_norm": 0.6046504974365234, "learning_rate": 1.9744668504279707e-06, "loss": 0.0066, "step": 24950 }, { "epoch": 2.7050086730268865, "grad_norm": 0.31170105934143066, "learning_rate": 1.9737414768605833e-06, "loss": 0.0109, "step": 24951 }, { "epoch": 2.705117085862966, "grad_norm": 1.4685633182525635, "learning_rate": 1.973016103293196e-06, "loss": 0.03, "step": 24952 }, { "epoch": 2.705225498699046, "grad_norm": 0.05537665635347366, "learning_rate": 1.972290729725809e-06, "loss": 0.0005, "step": 24953 }, { "epoch": 2.705333911535126, "grad_norm": 0.24593141674995422, "learning_rate": 1.971565356158422e-06, "loss": 0.0051, "step": 24954 }, { "epoch": 2.7054423243712056, "grad_norm": 3.130537986755371, "learning_rate": 1.9708399825910344e-06, "loss": 0.0321, "step": 24955 }, { "epoch": 2.705550737207285, "grad_norm": 0.4167810082435608, "learning_rate": 1.9701146090236474e-06, "loss": 0.0113, "step": 24956 }, { "epoch": 2.705659150043365, "grad_norm": 0.2681776285171509, "learning_rate": 1.9693892354562604e-06, "loss": 0.0047, "step": 24957 }, { "epoch": 2.705767562879445, "grad_norm": 2.0001308917999268, "learning_rate": 1.968663861888873e-06, "loss": 0.0088, "step": 24958 }, { "epoch": 2.7058759757155246, "grad_norm": 0.522556722164154, "learning_rate": 1.967938488321486e-06, "loss": 0.0096, "step": 24959 }, { "epoch": 2.7059843885516046, "grad_norm": 0.017282476648688316, "learning_rate": 1.9672131147540985e-06, "loss": 0.0001, "step": 24960 }, { "epoch": 2.706092801387684, "grad_norm": 2.9983913898468018, "learning_rate": 1.9664877411867115e-06, "loss": 0.0282, "step": 24961 }, { "epoch": 2.706201214223764, "grad_norm": 2.4888181686401367, "learning_rate": 1.965762367619324e-06, "loss": 0.008, "step": 24962 }, { "epoch": 2.706309627059844, "grad_norm": 17.197879791259766, "learning_rate": 1.965036994051937e-06, "loss": 0.0413, "step": 24963 }, { "epoch": 2.7064180398959237, "grad_norm": 9.348347663879395, "learning_rate": 1.9643116204845496e-06, "loss": 0.0225, "step": 24964 }, { "epoch": 2.706526452732003, "grad_norm": 0.886148989200592, "learning_rate": 1.963586246917162e-06, "loss": 0.0025, "step": 24965 }, { "epoch": 2.706634865568083, "grad_norm": 3.943871021270752, "learning_rate": 1.962860873349775e-06, "loss": 0.0029, "step": 24966 }, { "epoch": 2.706743278404163, "grad_norm": 0.7970097661018372, "learning_rate": 1.962135499782388e-06, "loss": 0.0072, "step": 24967 }, { "epoch": 2.7068516912402427, "grad_norm": 0.03707154467701912, "learning_rate": 1.9614101262150008e-06, "loss": 0.0003, "step": 24968 }, { "epoch": 2.7069601040763227, "grad_norm": 0.019481567665934563, "learning_rate": 1.9606847526476137e-06, "loss": 0.0001, "step": 24969 }, { "epoch": 2.7070685169124022, "grad_norm": 0.4771023690700531, "learning_rate": 1.9599593790802267e-06, "loss": 0.0173, "step": 24970 }, { "epoch": 2.7071769297484822, "grad_norm": 1.0183866024017334, "learning_rate": 1.9592340055128393e-06, "loss": 0.0195, "step": 24971 }, { "epoch": 2.707285342584562, "grad_norm": 0.22558651864528656, "learning_rate": 1.9585086319454523e-06, "loss": 0.0005, "step": 24972 }, { "epoch": 2.7073937554206418, "grad_norm": 0.6897515058517456, "learning_rate": 1.957783258378065e-06, "loss": 0.0234, "step": 24973 }, { "epoch": 2.7075021682567217, "grad_norm": 0.06760232150554657, "learning_rate": 1.957057884810678e-06, "loss": 0.0004, "step": 24974 }, { "epoch": 2.7076105810928013, "grad_norm": 0.6720261573791504, "learning_rate": 1.9563325112432904e-06, "loss": 0.0121, "step": 24975 }, { "epoch": 2.7077189939288813, "grad_norm": 0.022632043808698654, "learning_rate": 1.9556071376759034e-06, "loss": 0.0002, "step": 24976 }, { "epoch": 2.707827406764961, "grad_norm": 5.685403823852539, "learning_rate": 1.954881764108516e-06, "loss": 0.025, "step": 24977 }, { "epoch": 2.707935819601041, "grad_norm": 0.445902943611145, "learning_rate": 1.9541563905411285e-06, "loss": 0.011, "step": 24978 }, { "epoch": 2.7080442324371203, "grad_norm": 0.04551459476351738, "learning_rate": 1.9534310169737415e-06, "loss": 0.0005, "step": 24979 }, { "epoch": 2.7081526452732003, "grad_norm": 2.9462430477142334, "learning_rate": 1.9527056434063545e-06, "loss": 0.0064, "step": 24980 }, { "epoch": 2.7082610581092803, "grad_norm": 0.6307958364486694, "learning_rate": 1.951980269838967e-06, "loss": 0.0019, "step": 24981 }, { "epoch": 2.70836947094536, "grad_norm": 0.3615648150444031, "learning_rate": 1.95125489627158e-06, "loss": 0.0115, "step": 24982 }, { "epoch": 2.70847788378144, "grad_norm": 0.4208696782588959, "learning_rate": 1.950529522704193e-06, "loss": 0.0008, "step": 24983 }, { "epoch": 2.7085862966175194, "grad_norm": 0.054143983870744705, "learning_rate": 1.9498041491368056e-06, "loss": 0.0002, "step": 24984 }, { "epoch": 2.7086947094535994, "grad_norm": 0.49720844626426697, "learning_rate": 1.9490787755694186e-06, "loss": 0.0149, "step": 24985 }, { "epoch": 2.7088031222896793, "grad_norm": 0.2835942804813385, "learning_rate": 1.948353402002031e-06, "loss": 0.0051, "step": 24986 }, { "epoch": 2.708911535125759, "grad_norm": 0.011371615342795849, "learning_rate": 1.947628028434644e-06, "loss": 0.0001, "step": 24987 }, { "epoch": 2.7090199479618384, "grad_norm": 0.5032346248626709, "learning_rate": 1.9469026548672567e-06, "loss": 0.0015, "step": 24988 }, { "epoch": 2.7091283607979184, "grad_norm": 4.374862194061279, "learning_rate": 1.9461772812998697e-06, "loss": 0.0457, "step": 24989 }, { "epoch": 2.7092367736339984, "grad_norm": 1.2403677701950073, "learning_rate": 1.9454519077324823e-06, "loss": 0.0155, "step": 24990 }, { "epoch": 2.709345186470078, "grad_norm": 0.2594350576400757, "learning_rate": 1.944726534165095e-06, "loss": 0.005, "step": 24991 }, { "epoch": 2.709453599306158, "grad_norm": 0.6609346866607666, "learning_rate": 1.944001160597708e-06, "loss": 0.014, "step": 24992 }, { "epoch": 2.7095620121422375, "grad_norm": 1.2007732391357422, "learning_rate": 1.943275787030321e-06, "loss": 0.0207, "step": 24993 }, { "epoch": 2.7096704249783174, "grad_norm": 8.139336585998535, "learning_rate": 1.9425504134629334e-06, "loss": 0.0642, "step": 24994 }, { "epoch": 2.7097788378143974, "grad_norm": 0.36266806721687317, "learning_rate": 1.9418250398955464e-06, "loss": 0.0069, "step": 24995 }, { "epoch": 2.709887250650477, "grad_norm": 0.06085456535220146, "learning_rate": 1.9410996663281594e-06, "loss": 0.0005, "step": 24996 }, { "epoch": 2.709995663486557, "grad_norm": 0.5872800946235657, "learning_rate": 1.940374292760772e-06, "loss": 0.0156, "step": 24997 }, { "epoch": 2.7101040763226365, "grad_norm": 2.2797460556030273, "learning_rate": 1.939648919193385e-06, "loss": 0.0015, "step": 24998 }, { "epoch": 2.7102124891587165, "grad_norm": 22.349525451660156, "learning_rate": 1.9389235456259975e-06, "loss": 0.0244, "step": 24999 }, { "epoch": 2.710320901994796, "grad_norm": 0.6004514098167419, "learning_rate": 1.9381981720586105e-06, "loss": 0.0273, "step": 25000 }, { "epoch": 2.710429314830876, "grad_norm": 0.019587963819503784, "learning_rate": 1.937472798491223e-06, "loss": 0.0002, "step": 25001 }, { "epoch": 2.7105377276669556, "grad_norm": 0.5378715395927429, "learning_rate": 1.936747424923836e-06, "loss": 0.0121, "step": 25002 }, { "epoch": 2.7106461405030355, "grad_norm": 2.3261559009552, "learning_rate": 1.9360220513564486e-06, "loss": 0.0424, "step": 25003 }, { "epoch": 2.7107545533391155, "grad_norm": 0.0330820269882679, "learning_rate": 1.9352966777890616e-06, "loss": 0.0003, "step": 25004 }, { "epoch": 2.710862966175195, "grad_norm": 0.35609039664268494, "learning_rate": 1.934571304221674e-06, "loss": 0.0113, "step": 25005 }, { "epoch": 2.710971379011275, "grad_norm": 0.23257821798324585, "learning_rate": 1.933845930654287e-06, "loss": 0.0033, "step": 25006 }, { "epoch": 2.7110797918473546, "grad_norm": 0.5430325865745544, "learning_rate": 1.9331205570868998e-06, "loss": 0.0105, "step": 25007 }, { "epoch": 2.7111882046834346, "grad_norm": 2.706888198852539, "learning_rate": 1.9323951835195127e-06, "loss": 0.0034, "step": 25008 }, { "epoch": 2.7112966175195146, "grad_norm": 0.07769420742988586, "learning_rate": 1.9316698099521257e-06, "loss": 0.0004, "step": 25009 }, { "epoch": 2.711405030355594, "grad_norm": 2.685053825378418, "learning_rate": 1.9309444363847383e-06, "loss": 0.0095, "step": 25010 }, { "epoch": 2.7115134431916736, "grad_norm": 0.014347865246236324, "learning_rate": 1.9302190628173513e-06, "loss": 0.0001, "step": 25011 }, { "epoch": 2.7116218560277536, "grad_norm": 0.011060080491006374, "learning_rate": 1.929493689249964e-06, "loss": 0.0001, "step": 25012 }, { "epoch": 2.7117302688638336, "grad_norm": 0.348976731300354, "learning_rate": 1.928768315682577e-06, "loss": 0.0057, "step": 25013 }, { "epoch": 2.711838681699913, "grad_norm": 1.022605538368225, "learning_rate": 1.9280429421151894e-06, "loss": 0.0106, "step": 25014 }, { "epoch": 2.711947094535993, "grad_norm": 0.509655237197876, "learning_rate": 1.9273175685478024e-06, "loss": 0.0101, "step": 25015 }, { "epoch": 2.7120555073720727, "grad_norm": 0.2581643760204315, "learning_rate": 1.926592194980415e-06, "loss": 0.0058, "step": 25016 }, { "epoch": 2.7121639202081527, "grad_norm": 0.24543496966362, "learning_rate": 1.925866821413028e-06, "loss": 0.0033, "step": 25017 }, { "epoch": 2.7122723330442327, "grad_norm": 0.018366824835538864, "learning_rate": 1.9251414478456405e-06, "loss": 0.0001, "step": 25018 }, { "epoch": 2.712380745880312, "grad_norm": 0.22767437994480133, "learning_rate": 1.9244160742782535e-06, "loss": 0.0042, "step": 25019 }, { "epoch": 2.712489158716392, "grad_norm": 14.327319145202637, "learning_rate": 1.923690700710866e-06, "loss": 0.0256, "step": 25020 }, { "epoch": 2.7125975715524717, "grad_norm": 0.455532968044281, "learning_rate": 1.922965327143479e-06, "loss": 0.012, "step": 25021 }, { "epoch": 2.7127059843885517, "grad_norm": 0.5655941367149353, "learning_rate": 1.922239953576092e-06, "loss": 0.0162, "step": 25022 }, { "epoch": 2.7128143972246312, "grad_norm": 0.5419785380363464, "learning_rate": 1.9215145800087046e-06, "loss": 0.012, "step": 25023 }, { "epoch": 2.7129228100607112, "grad_norm": 0.8326308727264404, "learning_rate": 1.9207892064413176e-06, "loss": 0.0083, "step": 25024 }, { "epoch": 2.7130312228967908, "grad_norm": 0.5675312280654907, "learning_rate": 1.92006383287393e-06, "loss": 0.0211, "step": 25025 }, { "epoch": 2.7131396357328708, "grad_norm": 0.5090830326080322, "learning_rate": 1.919338459306543e-06, "loss": 0.0256, "step": 25026 }, { "epoch": 2.7132480485689507, "grad_norm": 0.4522538483142853, "learning_rate": 1.9186130857391558e-06, "loss": 0.0111, "step": 25027 }, { "epoch": 2.7133564614050303, "grad_norm": 0.00579209066927433, "learning_rate": 1.9178877121717687e-06, "loss": 0.0001, "step": 25028 }, { "epoch": 2.7134648742411103, "grad_norm": 7.955624580383301, "learning_rate": 1.9171623386043813e-06, "loss": 0.0159, "step": 25029 }, { "epoch": 2.71357328707719, "grad_norm": 0.344955712556839, "learning_rate": 1.9164369650369943e-06, "loss": 0.0125, "step": 25030 }, { "epoch": 2.71368169991327, "grad_norm": 0.03973645716905594, "learning_rate": 1.915711591469607e-06, "loss": 0.0002, "step": 25031 }, { "epoch": 2.71379011274935, "grad_norm": 0.0824180617928505, "learning_rate": 1.91498621790222e-06, "loss": 0.0003, "step": 25032 }, { "epoch": 2.7138985255854293, "grad_norm": 0.23261722922325134, "learning_rate": 1.9142608443348324e-06, "loss": 0.0044, "step": 25033 }, { "epoch": 2.714006938421509, "grad_norm": 2.875101327896118, "learning_rate": 1.9135354707674454e-06, "loss": 0.0239, "step": 25034 }, { "epoch": 2.714115351257589, "grad_norm": 0.4161398708820343, "learning_rate": 1.9128100972000584e-06, "loss": 0.016, "step": 25035 }, { "epoch": 2.714223764093669, "grad_norm": 4.6310648918151855, "learning_rate": 1.912084723632671e-06, "loss": 0.0465, "step": 25036 }, { "epoch": 2.7143321769297484, "grad_norm": 2.439934015274048, "learning_rate": 1.911359350065284e-06, "loss": 0.0151, "step": 25037 }, { "epoch": 2.7144405897658284, "grad_norm": 0.06845409423112869, "learning_rate": 1.9106339764978965e-06, "loss": 0.0003, "step": 25038 }, { "epoch": 2.714549002601908, "grad_norm": 0.24329492449760437, "learning_rate": 1.9099086029305095e-06, "loss": 0.0004, "step": 25039 }, { "epoch": 2.714657415437988, "grad_norm": 0.02834407053887844, "learning_rate": 1.909183229363122e-06, "loss": 0.0002, "step": 25040 }, { "epoch": 2.714765828274068, "grad_norm": 0.44725993275642395, "learning_rate": 1.908457855795735e-06, "loss": 0.0084, "step": 25041 }, { "epoch": 2.7148742411101474, "grad_norm": 3.753711700439453, "learning_rate": 1.9077324822283476e-06, "loss": 0.0193, "step": 25042 }, { "epoch": 2.7149826539462274, "grad_norm": 1.9813560247421265, "learning_rate": 1.9070071086609606e-06, "loss": 0.0387, "step": 25043 }, { "epoch": 2.715091066782307, "grad_norm": 0.28160786628723145, "learning_rate": 1.9062817350935734e-06, "loss": 0.0041, "step": 25044 }, { "epoch": 2.715199479618387, "grad_norm": 0.24082142114639282, "learning_rate": 1.905556361526186e-06, "loss": 0.0046, "step": 25045 }, { "epoch": 2.7153078924544665, "grad_norm": 0.4139566719532013, "learning_rate": 1.904830987958799e-06, "loss": 0.0079, "step": 25046 }, { "epoch": 2.7154163052905465, "grad_norm": 0.321583092212677, "learning_rate": 1.9041056143914115e-06, "loss": 0.0065, "step": 25047 }, { "epoch": 2.715524718126626, "grad_norm": 0.2695273458957672, "learning_rate": 1.9033802408240245e-06, "loss": 0.0062, "step": 25048 }, { "epoch": 2.715633130962706, "grad_norm": 0.008645846508443356, "learning_rate": 1.9026548672566373e-06, "loss": 0.0001, "step": 25049 }, { "epoch": 2.715741543798786, "grad_norm": 0.509970486164093, "learning_rate": 1.90192949368925e-06, "loss": 0.0097, "step": 25050 }, { "epoch": 2.7158499566348655, "grad_norm": 5.3110527992248535, "learning_rate": 1.9012041201218629e-06, "loss": 0.052, "step": 25051 }, { "epoch": 2.7159583694709455, "grad_norm": 4.606271743774414, "learning_rate": 1.9004787465544759e-06, "loss": 0.0018, "step": 25052 }, { "epoch": 2.716066782307025, "grad_norm": 6.291174411773682, "learning_rate": 1.8997533729870884e-06, "loss": 0.037, "step": 25053 }, { "epoch": 2.716175195143105, "grad_norm": 0.05129244923591614, "learning_rate": 1.8990279994197014e-06, "loss": 0.0003, "step": 25054 }, { "epoch": 2.716283607979185, "grad_norm": 0.05552646890282631, "learning_rate": 1.898302625852314e-06, "loss": 0.0004, "step": 25055 }, { "epoch": 2.7163920208152645, "grad_norm": 6.004804611206055, "learning_rate": 1.897577252284927e-06, "loss": 0.0071, "step": 25056 }, { "epoch": 2.716500433651344, "grad_norm": 0.04587296396493912, "learning_rate": 1.8968518787175397e-06, "loss": 0.0005, "step": 25057 }, { "epoch": 2.716608846487424, "grad_norm": 0.07402254641056061, "learning_rate": 1.8961265051501525e-06, "loss": 0.0002, "step": 25058 }, { "epoch": 2.716717259323504, "grad_norm": 0.2475772202014923, "learning_rate": 1.8954011315827653e-06, "loss": 0.0005, "step": 25059 }, { "epoch": 2.7168256721595836, "grad_norm": 0.33374086022377014, "learning_rate": 1.8946757580153779e-06, "loss": 0.0106, "step": 25060 }, { "epoch": 2.7169340849956636, "grad_norm": 5.12117862701416, "learning_rate": 1.8939503844479909e-06, "loss": 0.0221, "step": 25061 }, { "epoch": 2.717042497831743, "grad_norm": 0.34749913215637207, "learning_rate": 1.8932250108806036e-06, "loss": 0.007, "step": 25062 }, { "epoch": 2.717150910667823, "grad_norm": 0.4261645972728729, "learning_rate": 1.8924996373132164e-06, "loss": 0.0094, "step": 25063 }, { "epoch": 2.717259323503903, "grad_norm": 0.4800998270511627, "learning_rate": 1.8917742637458292e-06, "loss": 0.0052, "step": 25064 }, { "epoch": 2.7173677363399826, "grad_norm": 0.4628121852874756, "learning_rate": 1.8910488901784422e-06, "loss": 0.0146, "step": 25065 }, { "epoch": 2.717476149176062, "grad_norm": 0.15489259362220764, "learning_rate": 1.8903235166110548e-06, "loss": 0.0004, "step": 25066 }, { "epoch": 2.717584562012142, "grad_norm": 0.2269565463066101, "learning_rate": 1.8895981430436677e-06, "loss": 0.0003, "step": 25067 }, { "epoch": 2.717692974848222, "grad_norm": 0.34057295322418213, "learning_rate": 1.8888727694762803e-06, "loss": 0.0093, "step": 25068 }, { "epoch": 2.7178013876843017, "grad_norm": 0.25437477231025696, "learning_rate": 1.8881473959088933e-06, "loss": 0.0008, "step": 25069 }, { "epoch": 2.7179098005203817, "grad_norm": 0.347911536693573, "learning_rate": 1.887422022341506e-06, "loss": 0.0089, "step": 25070 }, { "epoch": 2.718018213356461, "grad_norm": 0.33975398540496826, "learning_rate": 1.8866966487741189e-06, "loss": 0.0067, "step": 25071 }, { "epoch": 2.718126626192541, "grad_norm": 0.08350338786840439, "learning_rate": 1.8859712752067316e-06, "loss": 0.0006, "step": 25072 }, { "epoch": 2.718235039028621, "grad_norm": 3.568037509918213, "learning_rate": 1.8852459016393442e-06, "loss": 0.0372, "step": 25073 }, { "epoch": 2.7183434518647007, "grad_norm": 0.23665587604045868, "learning_rate": 1.8845205280719572e-06, "loss": 0.0048, "step": 25074 }, { "epoch": 2.7184518647007807, "grad_norm": 0.03341132402420044, "learning_rate": 1.88379515450457e-06, "loss": 0.0001, "step": 25075 }, { "epoch": 2.7185602775368602, "grad_norm": 0.37291449308395386, "learning_rate": 1.8830697809371828e-06, "loss": 0.0008, "step": 25076 }, { "epoch": 2.7186686903729402, "grad_norm": 0.38618728518486023, "learning_rate": 1.8823444073697955e-06, "loss": 0.0128, "step": 25077 }, { "epoch": 2.71877710320902, "grad_norm": 0.17510855197906494, "learning_rate": 1.8816190338024085e-06, "loss": 0.0005, "step": 25078 }, { "epoch": 2.7188855160450998, "grad_norm": 0.23919948935508728, "learning_rate": 1.880893660235021e-06, "loss": 0.0047, "step": 25079 }, { "epoch": 2.7189939288811793, "grad_norm": 0.03283480182290077, "learning_rate": 1.880168286667634e-06, "loss": 0.0002, "step": 25080 }, { "epoch": 2.7191023417172593, "grad_norm": 1.7593094110488892, "learning_rate": 1.8794429131002466e-06, "loss": 0.0785, "step": 25081 }, { "epoch": 2.7192107545533393, "grad_norm": 0.21499699354171753, "learning_rate": 1.8787175395328596e-06, "loss": 0.0022, "step": 25082 }, { "epoch": 2.719319167389419, "grad_norm": 0.020787011831998825, "learning_rate": 1.8779921659654724e-06, "loss": 0.0, "step": 25083 }, { "epoch": 2.719427580225499, "grad_norm": 0.0471009686589241, "learning_rate": 1.8772667923980852e-06, "loss": 0.0003, "step": 25084 }, { "epoch": 2.7195359930615783, "grad_norm": 0.7077249884605408, "learning_rate": 1.876541418830698e-06, "loss": 0.007, "step": 25085 }, { "epoch": 2.7196444058976583, "grad_norm": 0.5002357959747314, "learning_rate": 1.8758160452633105e-06, "loss": 0.0148, "step": 25086 }, { "epoch": 2.7197528187337383, "grad_norm": 1.2366920709609985, "learning_rate": 1.8750906716959235e-06, "loss": 0.0081, "step": 25087 }, { "epoch": 2.719861231569818, "grad_norm": 0.6061545610427856, "learning_rate": 1.8743652981285363e-06, "loss": 0.0014, "step": 25088 }, { "epoch": 2.7199696444058974, "grad_norm": 0.974092423915863, "learning_rate": 1.873639924561149e-06, "loss": 0.0317, "step": 25089 }, { "epoch": 2.7200780572419774, "grad_norm": 0.1688869595527649, "learning_rate": 1.8729145509937619e-06, "loss": 0.0002, "step": 25090 }, { "epoch": 2.7201864700780574, "grad_norm": 0.24656470119953156, "learning_rate": 1.8721891774263749e-06, "loss": 0.0044, "step": 25091 }, { "epoch": 2.720294882914137, "grad_norm": 0.05901091545820236, "learning_rate": 1.8714638038589874e-06, "loss": 0.0002, "step": 25092 }, { "epoch": 2.720403295750217, "grad_norm": 0.4051488935947418, "learning_rate": 1.8707384302916004e-06, "loss": 0.006, "step": 25093 }, { "epoch": 2.7205117085862964, "grad_norm": 1.991930603981018, "learning_rate": 1.870013056724213e-06, "loss": 0.0084, "step": 25094 }, { "epoch": 2.7206201214223764, "grad_norm": 0.09491671621799469, "learning_rate": 1.869287683156826e-06, "loss": 0.0004, "step": 25095 }, { "epoch": 2.7207285342584564, "grad_norm": 0.04130179435014725, "learning_rate": 1.8685623095894387e-06, "loss": 0.0005, "step": 25096 }, { "epoch": 2.720836947094536, "grad_norm": 2.2204058170318604, "learning_rate": 1.8678369360220515e-06, "loss": 0.026, "step": 25097 }, { "epoch": 2.720945359930616, "grad_norm": 0.3189457058906555, "learning_rate": 1.8671115624546643e-06, "loss": 0.0127, "step": 25098 }, { "epoch": 2.7210537727666955, "grad_norm": 0.8152689933776855, "learning_rate": 1.8663861888872769e-06, "loss": 0.0203, "step": 25099 }, { "epoch": 2.7211621856027755, "grad_norm": 0.13768137991428375, "learning_rate": 1.8656608153198899e-06, "loss": 0.0013, "step": 25100 }, { "epoch": 2.7212705984388554, "grad_norm": 0.013482720591127872, "learning_rate": 1.8649354417525026e-06, "loss": 0.0001, "step": 25101 }, { "epoch": 2.721379011274935, "grad_norm": 0.6163846850395203, "learning_rate": 1.8642100681851154e-06, "loss": 0.0125, "step": 25102 }, { "epoch": 2.7214874241110145, "grad_norm": 0.04227416217327118, "learning_rate": 1.8634846946177282e-06, "loss": 0.0003, "step": 25103 }, { "epoch": 2.7215958369470945, "grad_norm": 0.3105993866920471, "learning_rate": 1.8627593210503412e-06, "loss": 0.0211, "step": 25104 }, { "epoch": 2.7217042497831745, "grad_norm": 1.0146677494049072, "learning_rate": 1.8620339474829538e-06, "loss": 0.0142, "step": 25105 }, { "epoch": 2.721812662619254, "grad_norm": 3.93505597114563, "learning_rate": 1.8613085739155667e-06, "loss": 0.011, "step": 25106 }, { "epoch": 2.721921075455334, "grad_norm": 0.01584823802113533, "learning_rate": 1.8605832003481793e-06, "loss": 0.0001, "step": 25107 }, { "epoch": 2.7220294882914136, "grad_norm": 0.032336898148059845, "learning_rate": 1.8598578267807923e-06, "loss": 0.0002, "step": 25108 }, { "epoch": 2.7221379011274935, "grad_norm": 0.5629099607467651, "learning_rate": 1.859132453213405e-06, "loss": 0.0166, "step": 25109 }, { "epoch": 2.7222463139635735, "grad_norm": 0.3426053822040558, "learning_rate": 1.8584070796460179e-06, "loss": 0.0087, "step": 25110 }, { "epoch": 2.722354726799653, "grad_norm": 2.1316640377044678, "learning_rate": 1.8576817060786306e-06, "loss": 0.0255, "step": 25111 }, { "epoch": 2.7224631396357326, "grad_norm": 2.855275869369507, "learning_rate": 1.8569563325112436e-06, "loss": 0.0394, "step": 25112 }, { "epoch": 2.7225715524718126, "grad_norm": 0.25071513652801514, "learning_rate": 1.8562309589438562e-06, "loss": 0.0052, "step": 25113 }, { "epoch": 2.7226799653078926, "grad_norm": 0.29517823457717896, "learning_rate": 1.855505585376469e-06, "loss": 0.0051, "step": 25114 }, { "epoch": 2.722788378143972, "grad_norm": 0.02834513410925865, "learning_rate": 1.8547802118090818e-06, "loss": 0.0003, "step": 25115 }, { "epoch": 2.722896790980052, "grad_norm": 0.750114381313324, "learning_rate": 1.8540548382416945e-06, "loss": 0.0154, "step": 25116 }, { "epoch": 2.7230052038161316, "grad_norm": 0.25053900480270386, "learning_rate": 1.8533294646743075e-06, "loss": 0.0051, "step": 25117 }, { "epoch": 2.7231136166522116, "grad_norm": 0.7869358062744141, "learning_rate": 1.85260409110692e-06, "loss": 0.0011, "step": 25118 }, { "epoch": 2.7232220294882916, "grad_norm": 0.6988786458969116, "learning_rate": 1.851878717539533e-06, "loss": 0.0145, "step": 25119 }, { "epoch": 2.723330442324371, "grad_norm": 0.028598180040717125, "learning_rate": 1.8511533439721456e-06, "loss": 0.0003, "step": 25120 }, { "epoch": 2.723438855160451, "grad_norm": 0.5456317663192749, "learning_rate": 1.8504279704047586e-06, "loss": 0.0026, "step": 25121 }, { "epoch": 2.7235472679965307, "grad_norm": 0.01681423746049404, "learning_rate": 1.8497025968373714e-06, "loss": 0.0001, "step": 25122 }, { "epoch": 2.7236556808326107, "grad_norm": 4.317503929138184, "learning_rate": 1.8489772232699842e-06, "loss": 0.0466, "step": 25123 }, { "epoch": 2.72376409366869, "grad_norm": 0.052030377089977264, "learning_rate": 1.848251849702597e-06, "loss": 0.0004, "step": 25124 }, { "epoch": 2.72387250650477, "grad_norm": 0.38074883818626404, "learning_rate": 1.84752647613521e-06, "loss": 0.01, "step": 25125 }, { "epoch": 2.7239809193408497, "grad_norm": 0.4203784465789795, "learning_rate": 1.8468011025678225e-06, "loss": 0.0128, "step": 25126 }, { "epoch": 2.7240893321769297, "grad_norm": 6.829375743865967, "learning_rate": 1.8460757290004353e-06, "loss": 0.004, "step": 25127 }, { "epoch": 2.7241977450130097, "grad_norm": 1.2644298076629639, "learning_rate": 1.845350355433048e-06, "loss": 0.0024, "step": 25128 }, { "epoch": 2.7243061578490892, "grad_norm": 8.122445106506348, "learning_rate": 1.8446249818656609e-06, "loss": 0.0041, "step": 25129 }, { "epoch": 2.7244145706851692, "grad_norm": 0.007974406704306602, "learning_rate": 1.8438996082982739e-06, "loss": 0.0, "step": 25130 }, { "epoch": 2.7245229835212488, "grad_norm": 0.48828184604644775, "learning_rate": 1.8431742347308864e-06, "loss": 0.0029, "step": 25131 }, { "epoch": 2.7246313963573288, "grad_norm": 1.424953579902649, "learning_rate": 1.8424488611634994e-06, "loss": 0.0411, "step": 25132 }, { "epoch": 2.7247398091934087, "grad_norm": 0.005940148141235113, "learning_rate": 1.841723487596112e-06, "loss": 0.0001, "step": 25133 }, { "epoch": 2.7248482220294883, "grad_norm": 1.8499637842178345, "learning_rate": 1.840998114028725e-06, "loss": 0.0136, "step": 25134 }, { "epoch": 2.724956634865568, "grad_norm": 0.6252298355102539, "learning_rate": 1.8402727404613378e-06, "loss": 0.0143, "step": 25135 }, { "epoch": 2.725065047701648, "grad_norm": 0.23303748667240143, "learning_rate": 1.8395473668939505e-06, "loss": 0.0007, "step": 25136 }, { "epoch": 2.725173460537728, "grad_norm": 3.0009403228759766, "learning_rate": 1.8388219933265633e-06, "loss": 0.0329, "step": 25137 }, { "epoch": 2.7252818733738073, "grad_norm": 0.03155949339270592, "learning_rate": 1.8380966197591763e-06, "loss": 0.0004, "step": 25138 }, { "epoch": 2.7253902862098873, "grad_norm": 0.397568941116333, "learning_rate": 1.8373712461917889e-06, "loss": 0.0153, "step": 25139 }, { "epoch": 2.725498699045967, "grad_norm": 2.1415209770202637, "learning_rate": 1.8366458726244016e-06, "loss": 0.0021, "step": 25140 }, { "epoch": 2.725607111882047, "grad_norm": 0.5180460214614868, "learning_rate": 1.8359204990570144e-06, "loss": 0.007, "step": 25141 }, { "epoch": 2.725715524718127, "grad_norm": 1.7418913841247559, "learning_rate": 1.8351951254896272e-06, "loss": 0.002, "step": 25142 }, { "epoch": 2.7258239375542064, "grad_norm": 0.11815356463193893, "learning_rate": 1.8344697519222402e-06, "loss": 0.0003, "step": 25143 }, { "epoch": 2.7259323503902864, "grad_norm": 0.05172688513994217, "learning_rate": 1.8337443783548528e-06, "loss": 0.0002, "step": 25144 }, { "epoch": 2.726040763226366, "grad_norm": 0.23708580434322357, "learning_rate": 1.8330190047874658e-06, "loss": 0.0163, "step": 25145 }, { "epoch": 2.726149176062446, "grad_norm": 7.924106597900391, "learning_rate": 1.8322936312200783e-06, "loss": 0.0402, "step": 25146 }, { "epoch": 2.7262575888985254, "grad_norm": 2.696622133255005, "learning_rate": 1.8315682576526913e-06, "loss": 0.0272, "step": 25147 }, { "epoch": 2.7263660017346054, "grad_norm": 0.35313230752944946, "learning_rate": 1.830842884085304e-06, "loss": 0.0113, "step": 25148 }, { "epoch": 2.726474414570685, "grad_norm": 5.713111877441406, "learning_rate": 1.830117510517917e-06, "loss": 0.0033, "step": 25149 }, { "epoch": 2.726582827406765, "grad_norm": 0.28693318367004395, "learning_rate": 1.8293921369505296e-06, "loss": 0.006, "step": 25150 }, { "epoch": 2.726691240242845, "grad_norm": 0.9940080046653748, "learning_rate": 1.8286667633831426e-06, "loss": 0.0041, "step": 25151 }, { "epoch": 2.7267996530789245, "grad_norm": 4.359802722930908, "learning_rate": 1.8279413898157552e-06, "loss": 0.0143, "step": 25152 }, { "epoch": 2.7269080659150045, "grad_norm": 0.013235604390501976, "learning_rate": 1.8272160162483682e-06, "loss": 0.0001, "step": 25153 }, { "epoch": 2.727016478751084, "grad_norm": 0.2395198792219162, "learning_rate": 1.8264906426809808e-06, "loss": 0.0023, "step": 25154 }, { "epoch": 2.727124891587164, "grad_norm": 5.2093400955200195, "learning_rate": 1.8257652691135935e-06, "loss": 0.0239, "step": 25155 }, { "epoch": 2.727233304423244, "grad_norm": 7.804704189300537, "learning_rate": 1.8250398955462065e-06, "loss": 0.0235, "step": 25156 }, { "epoch": 2.7273417172593235, "grad_norm": 0.6568722128868103, "learning_rate": 1.824314521978819e-06, "loss": 0.0222, "step": 25157 }, { "epoch": 2.727450130095403, "grad_norm": 3.4563496112823486, "learning_rate": 1.823589148411432e-06, "loss": 0.0066, "step": 25158 }, { "epoch": 2.727558542931483, "grad_norm": 0.2524796426296234, "learning_rate": 1.8228637748440447e-06, "loss": 0.0004, "step": 25159 }, { "epoch": 2.727666955767563, "grad_norm": 2.5130231380462646, "learning_rate": 1.8221384012766576e-06, "loss": 0.0241, "step": 25160 }, { "epoch": 2.7277753686036426, "grad_norm": 3.5871806144714355, "learning_rate": 1.8214130277092704e-06, "loss": 0.0496, "step": 25161 }, { "epoch": 2.7278837814397225, "grad_norm": 5.490583896636963, "learning_rate": 1.8206876541418834e-06, "loss": 0.0226, "step": 25162 }, { "epoch": 2.727992194275802, "grad_norm": 1.441887617111206, "learning_rate": 1.819962280574496e-06, "loss": 0.0178, "step": 25163 }, { "epoch": 2.728100607111882, "grad_norm": 0.1339145004749298, "learning_rate": 1.819236907007109e-06, "loss": 0.001, "step": 25164 }, { "epoch": 2.728209019947962, "grad_norm": 0.28924673795700073, "learning_rate": 1.8185115334397215e-06, "loss": 0.0088, "step": 25165 }, { "epoch": 2.7283174327840416, "grad_norm": 13.260416030883789, "learning_rate": 1.8177861598723345e-06, "loss": 0.0328, "step": 25166 }, { "epoch": 2.7284258456201216, "grad_norm": 0.34356796741485596, "learning_rate": 1.8170607863049473e-06, "loss": 0.013, "step": 25167 }, { "epoch": 2.728534258456201, "grad_norm": 0.39279019832611084, "learning_rate": 1.8163354127375599e-06, "loss": 0.0144, "step": 25168 }, { "epoch": 2.728642671292281, "grad_norm": 1.6020344495773315, "learning_rate": 1.8156100391701729e-06, "loss": 0.012, "step": 25169 }, { "epoch": 2.7287510841283606, "grad_norm": 0.0029190550558269024, "learning_rate": 1.8148846656027854e-06, "loss": 0.0, "step": 25170 }, { "epoch": 2.7288594969644406, "grad_norm": 0.3782493472099304, "learning_rate": 1.8141592920353984e-06, "loss": 0.0125, "step": 25171 }, { "epoch": 2.72896790980052, "grad_norm": 2.606952667236328, "learning_rate": 1.813433918468011e-06, "loss": 0.0048, "step": 25172 }, { "epoch": 2.7290763226366, "grad_norm": 0.3081909418106079, "learning_rate": 1.812708544900624e-06, "loss": 0.0106, "step": 25173 }, { "epoch": 2.72918473547268, "grad_norm": 0.026003669947385788, "learning_rate": 1.8119831713332368e-06, "loss": 0.0003, "step": 25174 }, { "epoch": 2.7292931483087597, "grad_norm": 0.3352874517440796, "learning_rate": 1.8112577977658497e-06, "loss": 0.0066, "step": 25175 }, { "epoch": 2.7294015611448397, "grad_norm": 0.46376681327819824, "learning_rate": 1.8105324241984623e-06, "loss": 0.0207, "step": 25176 }, { "epoch": 2.729509973980919, "grad_norm": 0.032603468745946884, "learning_rate": 1.8098070506310753e-06, "loss": 0.0003, "step": 25177 }, { "epoch": 2.729618386816999, "grad_norm": 1.379299521446228, "learning_rate": 1.8090816770636879e-06, "loss": 0.0258, "step": 25178 }, { "epoch": 2.729726799653079, "grad_norm": 1.7265790700912476, "learning_rate": 1.8083563034963009e-06, "loss": 0.0675, "step": 25179 }, { "epoch": 2.7298352124891587, "grad_norm": 0.4807529151439667, "learning_rate": 1.8076309299289136e-06, "loss": 0.0132, "step": 25180 }, { "epoch": 2.7299436253252383, "grad_norm": 0.07510331273078918, "learning_rate": 1.8069055563615262e-06, "loss": 0.0004, "step": 25181 }, { "epoch": 2.7300520381613183, "grad_norm": 0.9669332504272461, "learning_rate": 1.8061801827941392e-06, "loss": 0.0103, "step": 25182 }, { "epoch": 2.7301604509973982, "grad_norm": 0.15487441420555115, "learning_rate": 1.8054548092267518e-06, "loss": 0.001, "step": 25183 }, { "epoch": 2.7302688638334778, "grad_norm": 1.2524837255477905, "learning_rate": 1.8047294356593648e-06, "loss": 0.0148, "step": 25184 }, { "epoch": 2.7303772766695578, "grad_norm": 0.2566341161727905, "learning_rate": 1.8040040620919773e-06, "loss": 0.0051, "step": 25185 }, { "epoch": 2.7304856895056373, "grad_norm": 3.7328598499298096, "learning_rate": 1.8032786885245903e-06, "loss": 0.0229, "step": 25186 }, { "epoch": 2.7305941023417173, "grad_norm": 0.021945949643850327, "learning_rate": 1.802553314957203e-06, "loss": 0.0001, "step": 25187 }, { "epoch": 2.7307025151777973, "grad_norm": 0.48618850111961365, "learning_rate": 1.801827941389816e-06, "loss": 0.0073, "step": 25188 }, { "epoch": 2.730810928013877, "grad_norm": 2.734938859939575, "learning_rate": 1.8011025678224286e-06, "loss": 0.021, "step": 25189 }, { "epoch": 2.730919340849957, "grad_norm": 0.011419015005230904, "learning_rate": 1.8003771942550416e-06, "loss": 0.0001, "step": 25190 }, { "epoch": 2.7310277536860363, "grad_norm": 0.5440430641174316, "learning_rate": 1.7996518206876542e-06, "loss": 0.0017, "step": 25191 }, { "epoch": 2.7311361665221163, "grad_norm": 0.0800648182630539, "learning_rate": 1.7989264471202672e-06, "loss": 0.0008, "step": 25192 }, { "epoch": 2.731244579358196, "grad_norm": 0.27446940541267395, "learning_rate": 1.79820107355288e-06, "loss": 0.0058, "step": 25193 }, { "epoch": 2.731352992194276, "grad_norm": 1.1518746614456177, "learning_rate": 1.7974756999854925e-06, "loss": 0.016, "step": 25194 }, { "epoch": 2.7314614050303554, "grad_norm": 0.32908377051353455, "learning_rate": 1.7967503264181055e-06, "loss": 0.0068, "step": 25195 }, { "epoch": 2.7315698178664354, "grad_norm": 0.3344210386276245, "learning_rate": 1.796024952850718e-06, "loss": 0.0083, "step": 25196 }, { "epoch": 2.7316782307025154, "grad_norm": 1.0813138484954834, "learning_rate": 1.795299579283331e-06, "loss": 0.0303, "step": 25197 }, { "epoch": 2.731786643538595, "grad_norm": 0.2829112708568573, "learning_rate": 1.7945742057159439e-06, "loss": 0.006, "step": 25198 }, { "epoch": 2.731895056374675, "grad_norm": 3.8021578788757324, "learning_rate": 1.7938488321485566e-06, "loss": 0.0078, "step": 25199 }, { "epoch": 2.7320034692107544, "grad_norm": 7.0379438400268555, "learning_rate": 1.7931234585811694e-06, "loss": 0.0395, "step": 25200 }, { "epoch": 2.7321118820468344, "grad_norm": 0.07677656412124634, "learning_rate": 1.7923980850137824e-06, "loss": 0.0006, "step": 25201 }, { "epoch": 2.7322202948829144, "grad_norm": 0.003509253729134798, "learning_rate": 1.791672711446395e-06, "loss": 0.0, "step": 25202 }, { "epoch": 2.732328707718994, "grad_norm": 0.7973737716674805, "learning_rate": 1.790947337879008e-06, "loss": 0.0323, "step": 25203 }, { "epoch": 2.7324371205550735, "grad_norm": 2.573509931564331, "learning_rate": 1.7902219643116205e-06, "loss": 0.0017, "step": 25204 }, { "epoch": 2.7325455333911535, "grad_norm": 0.11281712353229523, "learning_rate": 1.7894965907442335e-06, "loss": 0.0007, "step": 25205 }, { "epoch": 2.7326539462272335, "grad_norm": 0.011905600316822529, "learning_rate": 1.7887712171768463e-06, "loss": 0.0001, "step": 25206 }, { "epoch": 2.732762359063313, "grad_norm": 0.27654585242271423, "learning_rate": 1.788045843609459e-06, "loss": 0.0056, "step": 25207 }, { "epoch": 2.732870771899393, "grad_norm": 0.28760966658592224, "learning_rate": 1.7873204700420719e-06, "loss": 0.0091, "step": 25208 }, { "epoch": 2.7329791847354725, "grad_norm": 0.5089738965034485, "learning_rate": 1.7865950964746844e-06, "loss": 0.0011, "step": 25209 }, { "epoch": 2.7330875975715525, "grad_norm": 0.02823060192167759, "learning_rate": 1.7858697229072974e-06, "loss": 0.0002, "step": 25210 }, { "epoch": 2.7331960104076325, "grad_norm": 0.2605358362197876, "learning_rate": 1.7851443493399102e-06, "loss": 0.0057, "step": 25211 }, { "epoch": 2.733304423243712, "grad_norm": 1.8675357103347778, "learning_rate": 1.784418975772523e-06, "loss": 0.037, "step": 25212 }, { "epoch": 2.7334128360797916, "grad_norm": 5.343391418457031, "learning_rate": 1.7836936022051358e-06, "loss": 0.0084, "step": 25213 }, { "epoch": 2.7335212489158716, "grad_norm": 2.1225674152374268, "learning_rate": 1.7829682286377487e-06, "loss": 0.0126, "step": 25214 }, { "epoch": 2.7336296617519515, "grad_norm": 0.6821061372756958, "learning_rate": 1.7822428550703613e-06, "loss": 0.0189, "step": 25215 }, { "epoch": 2.733738074588031, "grad_norm": 0.012180954217910767, "learning_rate": 1.7815174815029743e-06, "loss": 0.0001, "step": 25216 }, { "epoch": 2.733846487424111, "grad_norm": 0.39550143480300903, "learning_rate": 1.7807921079355869e-06, "loss": 0.0109, "step": 25217 }, { "epoch": 2.7339549002601906, "grad_norm": 2.9327023029327393, "learning_rate": 1.7800667343681999e-06, "loss": 0.0168, "step": 25218 }, { "epoch": 2.7340633130962706, "grad_norm": 0.03590264916419983, "learning_rate": 1.7793413608008126e-06, "loss": 0.0003, "step": 25219 }, { "epoch": 2.7341717259323506, "grad_norm": 0.2695813775062561, "learning_rate": 1.7786159872334254e-06, "loss": 0.0055, "step": 25220 }, { "epoch": 2.73428013876843, "grad_norm": 0.04112718626856804, "learning_rate": 1.7778906136660382e-06, "loss": 0.0004, "step": 25221 }, { "epoch": 2.73438855160451, "grad_norm": 0.518963098526001, "learning_rate": 1.7771652400986508e-06, "loss": 0.0112, "step": 25222 }, { "epoch": 2.7344969644405897, "grad_norm": 20.38504409790039, "learning_rate": 1.7764398665312638e-06, "loss": 0.0105, "step": 25223 }, { "epoch": 2.7346053772766696, "grad_norm": 2.2523741722106934, "learning_rate": 1.7757144929638765e-06, "loss": 0.0038, "step": 25224 }, { "epoch": 2.7347137901127496, "grad_norm": 0.33955004811286926, "learning_rate": 1.7749891193964893e-06, "loss": 0.0071, "step": 25225 }, { "epoch": 2.734822202948829, "grad_norm": 1.7462323904037476, "learning_rate": 1.774263745829102e-06, "loss": 0.0355, "step": 25226 }, { "epoch": 2.7349306157849087, "grad_norm": 0.5477951765060425, "learning_rate": 1.773538372261715e-06, "loss": 0.0221, "step": 25227 }, { "epoch": 2.7350390286209887, "grad_norm": 1.5455079078674316, "learning_rate": 1.7728129986943277e-06, "loss": 0.0516, "step": 25228 }, { "epoch": 2.7351474414570687, "grad_norm": 0.4178415536880493, "learning_rate": 1.7720876251269406e-06, "loss": 0.0044, "step": 25229 }, { "epoch": 2.735255854293148, "grad_norm": 0.0337580144405365, "learning_rate": 1.7713622515595532e-06, "loss": 0.0002, "step": 25230 }, { "epoch": 2.735364267129228, "grad_norm": 2.380390167236328, "learning_rate": 1.7706368779921662e-06, "loss": 0.0031, "step": 25231 }, { "epoch": 2.7354726799653077, "grad_norm": 0.8394823670387268, "learning_rate": 1.769911504424779e-06, "loss": 0.0013, "step": 25232 }, { "epoch": 2.7355810928013877, "grad_norm": 0.04270828887820244, "learning_rate": 1.7691861308573918e-06, "loss": 0.0003, "step": 25233 }, { "epoch": 2.7356895056374677, "grad_norm": 0.2029409557580948, "learning_rate": 1.7684607572900045e-06, "loss": 0.003, "step": 25234 }, { "epoch": 2.7357979184735473, "grad_norm": 0.15949180722236633, "learning_rate": 1.767735383722617e-06, "loss": 0.0008, "step": 25235 }, { "epoch": 2.735906331309627, "grad_norm": 0.5427221655845642, "learning_rate": 1.76701001015523e-06, "loss": 0.0123, "step": 25236 }, { "epoch": 2.736014744145707, "grad_norm": 0.048497915267944336, "learning_rate": 1.7662846365878429e-06, "loss": 0.0007, "step": 25237 }, { "epoch": 2.7361231569817868, "grad_norm": 0.26575928926467896, "learning_rate": 1.7655592630204556e-06, "loss": 0.0009, "step": 25238 }, { "epoch": 2.7362315698178663, "grad_norm": 5.782347679138184, "learning_rate": 1.7648338894530684e-06, "loss": 0.003, "step": 25239 }, { "epoch": 2.7363399826539463, "grad_norm": 5.7654032707214355, "learning_rate": 1.7641085158856814e-06, "loss": 0.0626, "step": 25240 }, { "epoch": 2.736448395490026, "grad_norm": 0.38875463604927063, "learning_rate": 1.763383142318294e-06, "loss": 0.0008, "step": 25241 }, { "epoch": 2.736556808326106, "grad_norm": 0.29973599314689636, "learning_rate": 1.762657768750907e-06, "loss": 0.0066, "step": 25242 }, { "epoch": 2.736665221162186, "grad_norm": 0.011387591250240803, "learning_rate": 1.7619323951835195e-06, "loss": 0.0001, "step": 25243 }, { "epoch": 2.7367736339982653, "grad_norm": 1.074380874633789, "learning_rate": 1.7612070216161325e-06, "loss": 0.0524, "step": 25244 }, { "epoch": 2.7368820468343453, "grad_norm": 33.310813903808594, "learning_rate": 1.7604816480487453e-06, "loss": 0.0152, "step": 25245 }, { "epoch": 2.736990459670425, "grad_norm": 0.45454463362693787, "learning_rate": 1.759756274481358e-06, "loss": 0.0102, "step": 25246 }, { "epoch": 2.737098872506505, "grad_norm": 1.4201701879501343, "learning_rate": 1.7590309009139709e-06, "loss": 0.0028, "step": 25247 }, { "epoch": 2.737207285342585, "grad_norm": 0.4232000410556793, "learning_rate": 1.7583055273465834e-06, "loss": 0.0326, "step": 25248 }, { "epoch": 2.7373156981786644, "grad_norm": 0.24346713721752167, "learning_rate": 1.7575801537791964e-06, "loss": 0.0014, "step": 25249 }, { "epoch": 2.737424111014744, "grad_norm": 0.050538256764411926, "learning_rate": 1.7568547802118092e-06, "loss": 0.0005, "step": 25250 }, { "epoch": 2.737532523850824, "grad_norm": 0.07742182165384293, "learning_rate": 1.756129406644422e-06, "loss": 0.0005, "step": 25251 }, { "epoch": 2.737640936686904, "grad_norm": 2.972092628479004, "learning_rate": 1.7554040330770348e-06, "loss": 0.0047, "step": 25252 }, { "epoch": 2.7377493495229834, "grad_norm": 0.37664735317230225, "learning_rate": 1.7546786595096478e-06, "loss": 0.0134, "step": 25253 }, { "epoch": 2.7378577623590634, "grad_norm": 0.8404189348220825, "learning_rate": 1.7539532859422603e-06, "loss": 0.0269, "step": 25254 }, { "epoch": 2.737966175195143, "grad_norm": 0.5048941373825073, "learning_rate": 1.7532279123748733e-06, "loss": 0.0242, "step": 25255 }, { "epoch": 2.738074588031223, "grad_norm": 3.920484781265259, "learning_rate": 1.7525025388074859e-06, "loss": 0.0123, "step": 25256 }, { "epoch": 2.738183000867303, "grad_norm": 0.5095969438552856, "learning_rate": 1.7517771652400989e-06, "loss": 0.0109, "step": 25257 }, { "epoch": 2.7382914137033825, "grad_norm": 1.1598976850509644, "learning_rate": 1.7510517916727116e-06, "loss": 0.0085, "step": 25258 }, { "epoch": 2.738399826539462, "grad_norm": 0.0013738250127062201, "learning_rate": 1.7503264181053244e-06, "loss": 0.0, "step": 25259 }, { "epoch": 2.738508239375542, "grad_norm": 2.822366952896118, "learning_rate": 1.7496010445379372e-06, "loss": 0.042, "step": 25260 }, { "epoch": 2.738616652211622, "grad_norm": 5.1369309425354, "learning_rate": 1.7488756709705502e-06, "loss": 0.0386, "step": 25261 }, { "epoch": 2.7387250650477015, "grad_norm": 3.353808641433716, "learning_rate": 1.7481502974031628e-06, "loss": 0.035, "step": 25262 }, { "epoch": 2.7388334778837815, "grad_norm": 6.232070446014404, "learning_rate": 1.7474249238357755e-06, "loss": 0.0157, "step": 25263 }, { "epoch": 2.738941890719861, "grad_norm": 0.016572441905736923, "learning_rate": 1.7466995502683883e-06, "loss": 0.0001, "step": 25264 }, { "epoch": 2.739050303555941, "grad_norm": 0.5299282670021057, "learning_rate": 1.745974176701001e-06, "loss": 0.0167, "step": 25265 }, { "epoch": 2.739158716392021, "grad_norm": 0.3937252163887024, "learning_rate": 1.745248803133614e-06, "loss": 0.0072, "step": 25266 }, { "epoch": 2.7392671292281006, "grad_norm": 0.015542566776275635, "learning_rate": 1.7445234295662267e-06, "loss": 0.0001, "step": 25267 }, { "epoch": 2.7393755420641805, "grad_norm": 0.00790027529001236, "learning_rate": 1.7437980559988396e-06, "loss": 0.0001, "step": 25268 }, { "epoch": 2.73948395490026, "grad_norm": 3.7191598415374756, "learning_rate": 1.7430726824314522e-06, "loss": 0.0287, "step": 25269 }, { "epoch": 2.73959236773634, "grad_norm": 0.03481355309486389, "learning_rate": 1.7423473088640652e-06, "loss": 0.0002, "step": 25270 }, { "epoch": 2.7397007805724196, "grad_norm": 3.4122068881988525, "learning_rate": 1.741621935296678e-06, "loss": 0.0494, "step": 25271 }, { "epoch": 2.7398091934084996, "grad_norm": 0.013072025030851364, "learning_rate": 1.7408965617292908e-06, "loss": 0.0001, "step": 25272 }, { "epoch": 2.739917606244579, "grad_norm": 2.565690517425537, "learning_rate": 1.7401711881619035e-06, "loss": 0.0184, "step": 25273 }, { "epoch": 2.740026019080659, "grad_norm": 2.261693000793457, "learning_rate": 1.7394458145945165e-06, "loss": 0.0236, "step": 25274 }, { "epoch": 2.740134431916739, "grad_norm": 0.08403625339269638, "learning_rate": 1.738720441027129e-06, "loss": 0.0004, "step": 25275 }, { "epoch": 2.7402428447528187, "grad_norm": 5.153285980224609, "learning_rate": 1.7379950674597419e-06, "loss": 0.0225, "step": 25276 }, { "epoch": 2.7403512575888986, "grad_norm": 0.03408143296837807, "learning_rate": 1.7372696938923547e-06, "loss": 0.0004, "step": 25277 }, { "epoch": 2.740459670424978, "grad_norm": 0.013071433641016483, "learning_rate": 1.7365443203249674e-06, "loss": 0.0001, "step": 25278 }, { "epoch": 2.740568083261058, "grad_norm": 0.009564117528498173, "learning_rate": 1.7358189467575804e-06, "loss": 0.0001, "step": 25279 }, { "epoch": 2.740676496097138, "grad_norm": 0.026214685291051865, "learning_rate": 1.735093573190193e-06, "loss": 0.0003, "step": 25280 }, { "epoch": 2.7407849089332177, "grad_norm": 1.701058030128479, "learning_rate": 1.734368199622806e-06, "loss": 0.028, "step": 25281 }, { "epoch": 2.7408933217692972, "grad_norm": 0.003768333699554205, "learning_rate": 1.7336428260554185e-06, "loss": 0.0, "step": 25282 }, { "epoch": 2.741001734605377, "grad_norm": 0.0659169927239418, "learning_rate": 1.7329174524880315e-06, "loss": 0.0006, "step": 25283 }, { "epoch": 2.741110147441457, "grad_norm": 0.6833950281143188, "learning_rate": 1.7321920789206443e-06, "loss": 0.011, "step": 25284 }, { "epoch": 2.7412185602775367, "grad_norm": 0.7538287043571472, "learning_rate": 1.731466705353257e-06, "loss": 0.0319, "step": 25285 }, { "epoch": 2.7413269731136167, "grad_norm": 2.886180877685547, "learning_rate": 1.7307413317858699e-06, "loss": 0.0221, "step": 25286 }, { "epoch": 2.7414353859496963, "grad_norm": 0.47304069995880127, "learning_rate": 1.7300159582184829e-06, "loss": 0.0199, "step": 25287 }, { "epoch": 2.7415437987857763, "grad_norm": 0.0774763748049736, "learning_rate": 1.7292905846510954e-06, "loss": 0.0004, "step": 25288 }, { "epoch": 2.7416522116218562, "grad_norm": 2.8811328411102295, "learning_rate": 1.7285652110837082e-06, "loss": 0.0641, "step": 25289 }, { "epoch": 2.741760624457936, "grad_norm": 0.3236852288246155, "learning_rate": 1.727839837516321e-06, "loss": 0.0064, "step": 25290 }, { "epoch": 2.7418690372940158, "grad_norm": 4.553393840789795, "learning_rate": 1.7271144639489338e-06, "loss": 0.0355, "step": 25291 }, { "epoch": 2.7419774501300953, "grad_norm": 0.3881481885910034, "learning_rate": 1.7263890903815468e-06, "loss": 0.0122, "step": 25292 }, { "epoch": 2.7420858629661753, "grad_norm": 0.0464845672249794, "learning_rate": 1.7256637168141593e-06, "loss": 0.0005, "step": 25293 }, { "epoch": 2.742194275802255, "grad_norm": 0.06247614324092865, "learning_rate": 1.7249383432467723e-06, "loss": 0.0003, "step": 25294 }, { "epoch": 2.742302688638335, "grad_norm": 0.0171115193516016, "learning_rate": 1.7242129696793849e-06, "loss": 0.0001, "step": 25295 }, { "epoch": 2.7424111014744144, "grad_norm": 0.006037819664925337, "learning_rate": 1.7234875961119979e-06, "loss": 0.0001, "step": 25296 }, { "epoch": 2.7425195143104943, "grad_norm": 0.41672199964523315, "learning_rate": 1.7227622225446106e-06, "loss": 0.0249, "step": 25297 }, { "epoch": 2.7426279271465743, "grad_norm": 0.3173283040523529, "learning_rate": 1.7220368489772234e-06, "loss": 0.0071, "step": 25298 }, { "epoch": 2.742736339982654, "grad_norm": 0.27807602286338806, "learning_rate": 1.7213114754098362e-06, "loss": 0.0101, "step": 25299 }, { "epoch": 2.742844752818734, "grad_norm": 3.169736623764038, "learning_rate": 1.7205861018424492e-06, "loss": 0.0142, "step": 25300 }, { "epoch": 2.7429531656548134, "grad_norm": 0.012041043490171432, "learning_rate": 1.7198607282750618e-06, "loss": 0.0001, "step": 25301 }, { "epoch": 2.7430615784908934, "grad_norm": 14.672094345092773, "learning_rate": 1.7191353547076748e-06, "loss": 0.0409, "step": 25302 }, { "epoch": 2.7431699913269734, "grad_norm": 6.529744625091553, "learning_rate": 1.7184099811402873e-06, "loss": 0.0335, "step": 25303 }, { "epoch": 2.743278404163053, "grad_norm": 0.6620420217514038, "learning_rate": 1.7176846075729e-06, "loss": 0.0023, "step": 25304 }, { "epoch": 2.7433868169991324, "grad_norm": 2.323192596435547, "learning_rate": 1.716959234005513e-06, "loss": 0.0373, "step": 25305 }, { "epoch": 2.7434952298352124, "grad_norm": 2.910219430923462, "learning_rate": 1.7162338604381257e-06, "loss": 0.0041, "step": 25306 }, { "epoch": 2.7436036426712924, "grad_norm": 0.6776041984558105, "learning_rate": 1.7155084868707386e-06, "loss": 0.0223, "step": 25307 }, { "epoch": 2.743712055507372, "grad_norm": 1.0256472826004028, "learning_rate": 1.7147831133033512e-06, "loss": 0.005, "step": 25308 }, { "epoch": 2.743820468343452, "grad_norm": 0.04792948067188263, "learning_rate": 1.7140577397359642e-06, "loss": 0.0001, "step": 25309 }, { "epoch": 2.7439288811795315, "grad_norm": 0.8751067519187927, "learning_rate": 1.713332366168577e-06, "loss": 0.0162, "step": 25310 }, { "epoch": 2.7440372940156115, "grad_norm": 0.3287906348705292, "learning_rate": 1.7126069926011898e-06, "loss": 0.0047, "step": 25311 }, { "epoch": 2.7441457068516915, "grad_norm": 0.006173382978886366, "learning_rate": 1.7118816190338025e-06, "loss": 0.0001, "step": 25312 }, { "epoch": 2.744254119687771, "grad_norm": 0.2983395755290985, "learning_rate": 1.7111562454664155e-06, "loss": 0.0097, "step": 25313 }, { "epoch": 2.744362532523851, "grad_norm": 3.1961379051208496, "learning_rate": 1.710430871899028e-06, "loss": 0.0266, "step": 25314 }, { "epoch": 2.7444709453599305, "grad_norm": 0.5273038744926453, "learning_rate": 1.709705498331641e-06, "loss": 0.0145, "step": 25315 }, { "epoch": 2.7445793581960105, "grad_norm": 1.203693151473999, "learning_rate": 1.7089801247642537e-06, "loss": 0.0549, "step": 25316 }, { "epoch": 2.74468777103209, "grad_norm": 0.023778650909662247, "learning_rate": 1.7082547511968664e-06, "loss": 0.0002, "step": 25317 }, { "epoch": 2.74479618386817, "grad_norm": 1.219496250152588, "learning_rate": 1.7075293776294794e-06, "loss": 0.028, "step": 25318 }, { "epoch": 2.7449045967042496, "grad_norm": 0.19006817042827606, "learning_rate": 1.706804004062092e-06, "loss": 0.0023, "step": 25319 }, { "epoch": 2.7450130095403296, "grad_norm": 0.06737025082111359, "learning_rate": 1.706078630494705e-06, "loss": 0.0003, "step": 25320 }, { "epoch": 2.7451214223764095, "grad_norm": 12.30846881866455, "learning_rate": 1.7053532569273175e-06, "loss": 0.0158, "step": 25321 }, { "epoch": 2.745229835212489, "grad_norm": 0.014456667006015778, "learning_rate": 1.7046278833599305e-06, "loss": 0.0001, "step": 25322 }, { "epoch": 2.745338248048569, "grad_norm": 0.2507464587688446, "learning_rate": 1.7039025097925433e-06, "loss": 0.001, "step": 25323 }, { "epoch": 2.7454466608846486, "grad_norm": 0.04545225575566292, "learning_rate": 1.703177136225156e-06, "loss": 0.0003, "step": 25324 }, { "epoch": 2.7455550737207286, "grad_norm": 0.0801095962524414, "learning_rate": 1.7024517626577689e-06, "loss": 0.0006, "step": 25325 }, { "epoch": 2.7456634865568086, "grad_norm": 0.036302220076322556, "learning_rate": 1.7017263890903819e-06, "loss": 0.0004, "step": 25326 }, { "epoch": 2.745771899392888, "grad_norm": 0.44273513555526733, "learning_rate": 1.7010010155229944e-06, "loss": 0.0121, "step": 25327 }, { "epoch": 2.7458803122289677, "grad_norm": 0.26748162508010864, "learning_rate": 1.7002756419556074e-06, "loss": 0.0055, "step": 25328 }, { "epoch": 2.7459887250650477, "grad_norm": 0.20790812373161316, "learning_rate": 1.69955026838822e-06, "loss": 0.0018, "step": 25329 }, { "epoch": 2.7460971379011276, "grad_norm": 0.9354374408721924, "learning_rate": 1.6988248948208328e-06, "loss": 0.0035, "step": 25330 }, { "epoch": 2.746205550737207, "grad_norm": 0.5239246487617493, "learning_rate": 1.6980995212534458e-06, "loss": 0.0239, "step": 25331 }, { "epoch": 2.746313963573287, "grad_norm": 0.029507296159863472, "learning_rate": 1.6973741476860583e-06, "loss": 0.0001, "step": 25332 }, { "epoch": 2.7464223764093667, "grad_norm": 0.40952977538108826, "learning_rate": 1.6966487741186713e-06, "loss": 0.0009, "step": 25333 }, { "epoch": 2.7465307892454467, "grad_norm": 0.7722920775413513, "learning_rate": 1.6959234005512839e-06, "loss": 0.0122, "step": 25334 }, { "epoch": 2.7466392020815267, "grad_norm": 2.040559768676758, "learning_rate": 1.6951980269838969e-06, "loss": 0.0162, "step": 25335 }, { "epoch": 2.746747614917606, "grad_norm": 0.427797794342041, "learning_rate": 1.6944726534165097e-06, "loss": 0.0041, "step": 25336 }, { "epoch": 2.746856027753686, "grad_norm": 1.0903120040893555, "learning_rate": 1.6937472798491224e-06, "loss": 0.0079, "step": 25337 }, { "epoch": 2.7469644405897657, "grad_norm": 0.9177744388580322, "learning_rate": 1.6930219062817352e-06, "loss": 0.0187, "step": 25338 }, { "epoch": 2.7470728534258457, "grad_norm": 16.131074905395508, "learning_rate": 1.6922965327143482e-06, "loss": 0.0062, "step": 25339 }, { "epoch": 2.7471812662619253, "grad_norm": 0.2844977378845215, "learning_rate": 1.6915711591469608e-06, "loss": 0.0062, "step": 25340 }, { "epoch": 2.7472896790980053, "grad_norm": 4.439621925354004, "learning_rate": 1.6908457855795738e-06, "loss": 0.0119, "step": 25341 }, { "epoch": 2.747398091934085, "grad_norm": 0.9194732904434204, "learning_rate": 1.6901204120121863e-06, "loss": 0.0111, "step": 25342 }, { "epoch": 2.747506504770165, "grad_norm": 0.035496972501277924, "learning_rate": 1.689395038444799e-06, "loss": 0.0004, "step": 25343 }, { "epoch": 2.7476149176062448, "grad_norm": 0.5374487638473511, "learning_rate": 1.688669664877412e-06, "loss": 0.0112, "step": 25344 }, { "epoch": 2.7477233304423243, "grad_norm": 0.07108078896999359, "learning_rate": 1.6879442913100247e-06, "loss": 0.0002, "step": 25345 }, { "epoch": 2.7478317432784043, "grad_norm": 0.24555423855781555, "learning_rate": 1.6872189177426377e-06, "loss": 0.005, "step": 25346 }, { "epoch": 2.747940156114484, "grad_norm": 2.919717311859131, "learning_rate": 1.6864935441752502e-06, "loss": 0.0205, "step": 25347 }, { "epoch": 2.748048568950564, "grad_norm": 0.03906330093741417, "learning_rate": 1.6857681706078632e-06, "loss": 0.0004, "step": 25348 }, { "epoch": 2.748156981786644, "grad_norm": 3.6970198154449463, "learning_rate": 1.685042797040476e-06, "loss": 0.0462, "step": 25349 }, { "epoch": 2.7482653946227233, "grad_norm": 0.261094868183136, "learning_rate": 1.6843174234730888e-06, "loss": 0.0057, "step": 25350 }, { "epoch": 2.748373807458803, "grad_norm": 0.05567502975463867, "learning_rate": 1.6835920499057015e-06, "loss": 0.0006, "step": 25351 }, { "epoch": 2.748482220294883, "grad_norm": 5.697032928466797, "learning_rate": 1.6828666763383145e-06, "loss": 0.0173, "step": 25352 }, { "epoch": 2.748590633130963, "grad_norm": 8.990586280822754, "learning_rate": 1.682141302770927e-06, "loss": 0.0103, "step": 25353 }, { "epoch": 2.7486990459670424, "grad_norm": 0.04849429428577423, "learning_rate": 1.68141592920354e-06, "loss": 0.0005, "step": 25354 }, { "epoch": 2.7488074588031224, "grad_norm": 0.3631434738636017, "learning_rate": 1.6806905556361527e-06, "loss": 0.007, "step": 25355 }, { "epoch": 2.748915871639202, "grad_norm": 0.46016886830329895, "learning_rate": 1.6799651820687656e-06, "loss": 0.0009, "step": 25356 }, { "epoch": 2.749024284475282, "grad_norm": 0.47628331184387207, "learning_rate": 1.6792398085013784e-06, "loss": 0.016, "step": 25357 }, { "epoch": 2.749132697311362, "grad_norm": 0.013545072637498379, "learning_rate": 1.678514434933991e-06, "loss": 0.0001, "step": 25358 }, { "epoch": 2.7492411101474414, "grad_norm": 0.03780851885676384, "learning_rate": 1.677789061366604e-06, "loss": 0.0003, "step": 25359 }, { "epoch": 2.749349522983521, "grad_norm": 0.07502427697181702, "learning_rate": 1.6770636877992166e-06, "loss": 0.0003, "step": 25360 }, { "epoch": 2.749457935819601, "grad_norm": 0.123802050948143, "learning_rate": 1.6763383142318295e-06, "loss": 0.0008, "step": 25361 }, { "epoch": 2.749566348655681, "grad_norm": 5.398662567138672, "learning_rate": 1.6756129406644423e-06, "loss": 0.0925, "step": 25362 }, { "epoch": 2.7496747614917605, "grad_norm": 0.6765477657318115, "learning_rate": 1.674887567097055e-06, "loss": 0.0182, "step": 25363 }, { "epoch": 2.7497831743278405, "grad_norm": 1.7057609558105469, "learning_rate": 1.6741621935296679e-06, "loss": 0.0316, "step": 25364 }, { "epoch": 2.74989158716392, "grad_norm": 0.009835311211645603, "learning_rate": 1.6734368199622809e-06, "loss": 0.0001, "step": 25365 }, { "epoch": 2.75, "grad_norm": 0.29353412985801697, "learning_rate": 1.6727114463948934e-06, "loss": 0.0014, "step": 25366 }, { "epoch": 2.75010841283608, "grad_norm": 0.048183053731918335, "learning_rate": 1.6719860728275064e-06, "loss": 0.0003, "step": 25367 }, { "epoch": 2.7502168256721595, "grad_norm": 1.9802581071853638, "learning_rate": 1.671260699260119e-06, "loss": 0.008, "step": 25368 }, { "epoch": 2.7503252385082395, "grad_norm": 0.3144179880619049, "learning_rate": 1.670535325692732e-06, "loss": 0.0107, "step": 25369 }, { "epoch": 2.750433651344319, "grad_norm": 0.07178111374378204, "learning_rate": 1.6698099521253448e-06, "loss": 0.0004, "step": 25370 }, { "epoch": 2.750542064180399, "grad_norm": 8.157685279846191, "learning_rate": 1.6690845785579573e-06, "loss": 0.0264, "step": 25371 }, { "epoch": 2.750650477016479, "grad_norm": 6.57093620300293, "learning_rate": 1.6683592049905703e-06, "loss": 0.0418, "step": 25372 }, { "epoch": 2.7507588898525586, "grad_norm": 2.0498602390289307, "learning_rate": 1.6676338314231829e-06, "loss": 0.0074, "step": 25373 }, { "epoch": 2.750867302688638, "grad_norm": 0.11329961568117142, "learning_rate": 1.6669084578557959e-06, "loss": 0.012, "step": 25374 }, { "epoch": 2.750975715524718, "grad_norm": 0.2589273452758789, "learning_rate": 1.6661830842884087e-06, "loss": 0.0038, "step": 25375 }, { "epoch": 2.751084128360798, "grad_norm": 0.4929102063179016, "learning_rate": 1.6654577107210214e-06, "loss": 0.0098, "step": 25376 }, { "epoch": 2.7511925411968776, "grad_norm": 3.3084142208099365, "learning_rate": 1.6647323371536342e-06, "loss": 0.0327, "step": 25377 }, { "epoch": 2.7513009540329576, "grad_norm": 4.481612682342529, "learning_rate": 1.6640069635862472e-06, "loss": 0.0225, "step": 25378 }, { "epoch": 2.751409366869037, "grad_norm": 0.3324376940727234, "learning_rate": 1.6632815900188598e-06, "loss": 0.0013, "step": 25379 }, { "epoch": 2.751517779705117, "grad_norm": 0.46062445640563965, "learning_rate": 1.6625562164514728e-06, "loss": 0.0017, "step": 25380 }, { "epoch": 2.751626192541197, "grad_norm": 0.030861148610711098, "learning_rate": 1.6618308428840853e-06, "loss": 0.0002, "step": 25381 }, { "epoch": 2.7517346053772767, "grad_norm": 0.05732245370745659, "learning_rate": 1.6611054693166983e-06, "loss": 0.0003, "step": 25382 }, { "epoch": 2.751843018213356, "grad_norm": 3.678102493286133, "learning_rate": 1.660380095749311e-06, "loss": 0.0216, "step": 25383 }, { "epoch": 2.751951431049436, "grad_norm": 0.6303097009658813, "learning_rate": 1.6596547221819237e-06, "loss": 0.0188, "step": 25384 }, { "epoch": 2.752059843885516, "grad_norm": 0.5122010111808777, "learning_rate": 1.6589293486145367e-06, "loss": 0.0186, "step": 25385 }, { "epoch": 2.7521682567215957, "grad_norm": 1.5272256135940552, "learning_rate": 1.6582039750471492e-06, "loss": 0.0294, "step": 25386 }, { "epoch": 2.7522766695576757, "grad_norm": 8.16340160369873, "learning_rate": 1.6574786014797622e-06, "loss": 0.0297, "step": 25387 }, { "epoch": 2.7523850823937552, "grad_norm": 0.4101642072200775, "learning_rate": 1.656753227912375e-06, "loss": 0.0049, "step": 25388 }, { "epoch": 2.752493495229835, "grad_norm": 0.09911569207906723, "learning_rate": 1.6560278543449878e-06, "loss": 0.0006, "step": 25389 }, { "epoch": 2.752601908065915, "grad_norm": 0.2744187116622925, "learning_rate": 1.6553024807776005e-06, "loss": 0.0053, "step": 25390 }, { "epoch": 2.7527103209019947, "grad_norm": 0.3121148645877838, "learning_rate": 1.6545771072102135e-06, "loss": 0.0069, "step": 25391 }, { "epoch": 2.7528187337380747, "grad_norm": 0.10349640250205994, "learning_rate": 1.653851733642826e-06, "loss": 0.0009, "step": 25392 }, { "epoch": 2.7529271465741543, "grad_norm": 0.007578105200082064, "learning_rate": 1.653126360075439e-06, "loss": 0.0001, "step": 25393 }, { "epoch": 2.7530355594102343, "grad_norm": 5.7389631271362305, "learning_rate": 1.6524009865080517e-06, "loss": 0.0037, "step": 25394 }, { "epoch": 2.753143972246314, "grad_norm": 0.37425872683525085, "learning_rate": 1.6516756129406647e-06, "loss": 0.02, "step": 25395 }, { "epoch": 2.753252385082394, "grad_norm": 0.032187074422836304, "learning_rate": 1.6509502393732774e-06, "loss": 0.0001, "step": 25396 }, { "epoch": 2.7533607979184733, "grad_norm": 0.017623290419578552, "learning_rate": 1.65022486580589e-06, "loss": 0.0001, "step": 25397 }, { "epoch": 2.7534692107545533, "grad_norm": 0.487799733877182, "learning_rate": 1.649499492238503e-06, "loss": 0.0081, "step": 25398 }, { "epoch": 2.7535776235906333, "grad_norm": 19.717361450195312, "learning_rate": 1.6487741186711156e-06, "loss": 0.0431, "step": 25399 }, { "epoch": 2.753686036426713, "grad_norm": 0.0626668632030487, "learning_rate": 1.6480487451037285e-06, "loss": 0.0007, "step": 25400 }, { "epoch": 2.753794449262793, "grad_norm": 0.03556669503450394, "learning_rate": 1.6473233715363413e-06, "loss": 0.0003, "step": 25401 }, { "epoch": 2.7539028620988724, "grad_norm": 0.23480840027332306, "learning_rate": 1.646597997968954e-06, "loss": 0.0003, "step": 25402 }, { "epoch": 2.7540112749349523, "grad_norm": 0.42609748244285583, "learning_rate": 1.6458726244015669e-06, "loss": 0.0117, "step": 25403 }, { "epoch": 2.7541196877710323, "grad_norm": 0.4094564616680145, "learning_rate": 1.6451472508341799e-06, "loss": 0.0134, "step": 25404 }, { "epoch": 2.754228100607112, "grad_norm": 12.879173278808594, "learning_rate": 1.6444218772667924e-06, "loss": 0.0319, "step": 25405 }, { "epoch": 2.7543365134431914, "grad_norm": 9.701581954956055, "learning_rate": 1.6436965036994054e-06, "loss": 0.0789, "step": 25406 }, { "epoch": 2.7544449262792714, "grad_norm": 6.373251438140869, "learning_rate": 1.642971130132018e-06, "loss": 0.0121, "step": 25407 }, { "epoch": 2.7545533391153514, "grad_norm": 5.860119819641113, "learning_rate": 1.642245756564631e-06, "loss": 0.0064, "step": 25408 }, { "epoch": 2.754661751951431, "grad_norm": 0.03615442290902138, "learning_rate": 1.6415203829972438e-06, "loss": 0.0002, "step": 25409 }, { "epoch": 2.754770164787511, "grad_norm": 2.5467934608459473, "learning_rate": 1.6407950094298565e-06, "loss": 0.0355, "step": 25410 }, { "epoch": 2.7548785776235905, "grad_norm": 9.588028907775879, "learning_rate": 1.6400696358624693e-06, "loss": 0.0178, "step": 25411 }, { "epoch": 2.7549869904596704, "grad_norm": 0.3089215159416199, "learning_rate": 1.6393442622950819e-06, "loss": 0.0064, "step": 25412 }, { "epoch": 2.7550954032957504, "grad_norm": 0.009145737625658512, "learning_rate": 1.6386188887276949e-06, "loss": 0.0001, "step": 25413 }, { "epoch": 2.75520381613183, "grad_norm": 13.034049034118652, "learning_rate": 1.6378935151603077e-06, "loss": 0.1, "step": 25414 }, { "epoch": 2.75531222896791, "grad_norm": 0.8731902837753296, "learning_rate": 1.6371681415929204e-06, "loss": 0.0088, "step": 25415 }, { "epoch": 2.7554206418039895, "grad_norm": 2.0522191524505615, "learning_rate": 1.6364427680255332e-06, "loss": 0.0194, "step": 25416 }, { "epoch": 2.7555290546400695, "grad_norm": 3.6768698692321777, "learning_rate": 1.6357173944581462e-06, "loss": 0.0252, "step": 25417 }, { "epoch": 2.755637467476149, "grad_norm": 0.38618239760398865, "learning_rate": 1.6349920208907588e-06, "loss": 0.0106, "step": 25418 }, { "epoch": 2.755745880312229, "grad_norm": 0.015144034288823605, "learning_rate": 1.6342666473233718e-06, "loss": 0.0001, "step": 25419 }, { "epoch": 2.7558542931483085, "grad_norm": 0.5952349901199341, "learning_rate": 1.6335412737559843e-06, "loss": 0.0226, "step": 25420 }, { "epoch": 2.7559627059843885, "grad_norm": 7.269068241119385, "learning_rate": 1.6328159001885973e-06, "loss": 0.0328, "step": 25421 }, { "epoch": 2.7560711188204685, "grad_norm": 5.238999843597412, "learning_rate": 1.63209052662121e-06, "loss": 0.0522, "step": 25422 }, { "epoch": 2.756179531656548, "grad_norm": 0.14461499452590942, "learning_rate": 1.6313651530538229e-06, "loss": 0.0017, "step": 25423 }, { "epoch": 2.756287944492628, "grad_norm": 0.7078723907470703, "learning_rate": 1.6306397794864357e-06, "loss": 0.023, "step": 25424 }, { "epoch": 2.7563963573287076, "grad_norm": 11.237922668457031, "learning_rate": 1.6299144059190482e-06, "loss": 0.022, "step": 25425 }, { "epoch": 2.7565047701647876, "grad_norm": 5.953726768493652, "learning_rate": 1.6291890323516612e-06, "loss": 0.0212, "step": 25426 }, { "epoch": 2.7566131830008676, "grad_norm": 0.15321433544158936, "learning_rate": 1.628463658784274e-06, "loss": 0.0005, "step": 25427 }, { "epoch": 2.756721595836947, "grad_norm": 0.593916118144989, "learning_rate": 1.6277382852168868e-06, "loss": 0.0171, "step": 25428 }, { "epoch": 2.7568300086730266, "grad_norm": 0.11350487172603607, "learning_rate": 1.6270129116494996e-06, "loss": 0.0006, "step": 25429 }, { "epoch": 2.7569384215091066, "grad_norm": 0.29333117604255676, "learning_rate": 1.6262875380821125e-06, "loss": 0.0002, "step": 25430 }, { "epoch": 2.7570468343451866, "grad_norm": 0.5648664832115173, "learning_rate": 1.6255621645147251e-06, "loss": 0.0166, "step": 25431 }, { "epoch": 2.757155247181266, "grad_norm": 0.07514407485723495, "learning_rate": 1.624836790947338e-06, "loss": 0.0005, "step": 25432 }, { "epoch": 2.757263660017346, "grad_norm": 2.0376973152160645, "learning_rate": 1.6241114173799507e-06, "loss": 0.0036, "step": 25433 }, { "epoch": 2.7573720728534257, "grad_norm": 3.3418045043945312, "learning_rate": 1.6233860438125637e-06, "loss": 0.0151, "step": 25434 }, { "epoch": 2.7574804856895057, "grad_norm": 0.5623524188995361, "learning_rate": 1.6226606702451764e-06, "loss": 0.008, "step": 25435 }, { "epoch": 2.7575888985255856, "grad_norm": 0.6546926498413086, "learning_rate": 1.6219352966777892e-06, "loss": 0.0113, "step": 25436 }, { "epoch": 2.757697311361665, "grad_norm": 4.201855659484863, "learning_rate": 1.621209923110402e-06, "loss": 0.0312, "step": 25437 }, { "epoch": 2.757805724197745, "grad_norm": 0.35230758786201477, "learning_rate": 1.6204845495430146e-06, "loss": 0.0059, "step": 25438 }, { "epoch": 2.7579141370338247, "grad_norm": 2.770414113998413, "learning_rate": 1.6197591759756275e-06, "loss": 0.0254, "step": 25439 }, { "epoch": 2.7580225498699047, "grad_norm": 0.3935827910900116, "learning_rate": 1.6190338024082403e-06, "loss": 0.0101, "step": 25440 }, { "epoch": 2.7581309627059842, "grad_norm": 6.458397388458252, "learning_rate": 1.6183084288408531e-06, "loss": 0.0103, "step": 25441 }, { "epoch": 2.758239375542064, "grad_norm": 5.806882858276367, "learning_rate": 1.6175830552734659e-06, "loss": 0.0109, "step": 25442 }, { "epoch": 2.7583477883781438, "grad_norm": 0.04777884483337402, "learning_rate": 1.6168576817060789e-06, "loss": 0.0003, "step": 25443 }, { "epoch": 2.7584562012142237, "grad_norm": 12.806316375732422, "learning_rate": 1.6161323081386914e-06, "loss": 0.0133, "step": 25444 }, { "epoch": 2.7585646140503037, "grad_norm": 0.07535925507545471, "learning_rate": 1.6154069345713044e-06, "loss": 0.0005, "step": 25445 }, { "epoch": 2.7586730268863833, "grad_norm": 4.957490921020508, "learning_rate": 1.614681561003917e-06, "loss": 0.0106, "step": 25446 }, { "epoch": 2.7587814397224633, "grad_norm": 0.20830650627613068, "learning_rate": 1.61395618743653e-06, "loss": 0.003, "step": 25447 }, { "epoch": 2.758889852558543, "grad_norm": 0.6289817094802856, "learning_rate": 1.6132308138691428e-06, "loss": 0.0232, "step": 25448 }, { "epoch": 2.758998265394623, "grad_norm": 0.6930777430534363, "learning_rate": 1.6125054403017558e-06, "loss": 0.022, "step": 25449 }, { "epoch": 2.7591066782307028, "grad_norm": 0.015385139733552933, "learning_rate": 1.6117800667343683e-06, "loss": 0.0001, "step": 25450 }, { "epoch": 2.7592150910667823, "grad_norm": 0.3719545602798462, "learning_rate": 1.6110546931669813e-06, "loss": 0.007, "step": 25451 }, { "epoch": 2.759323503902862, "grad_norm": 0.6320871710777283, "learning_rate": 1.6103293195995939e-06, "loss": 0.0077, "step": 25452 }, { "epoch": 2.759431916738942, "grad_norm": 0.7505658268928528, "learning_rate": 1.6096039460322067e-06, "loss": 0.0122, "step": 25453 }, { "epoch": 2.759540329575022, "grad_norm": 0.3381418287754059, "learning_rate": 1.6088785724648194e-06, "loss": 0.0008, "step": 25454 }, { "epoch": 2.7596487424111014, "grad_norm": 16.841293334960938, "learning_rate": 1.6081531988974322e-06, "loss": 0.0135, "step": 25455 }, { "epoch": 2.7597571552471813, "grad_norm": 0.27841976284980774, "learning_rate": 1.6074278253300452e-06, "loss": 0.0005, "step": 25456 }, { "epoch": 2.759865568083261, "grad_norm": 2.0880837440490723, "learning_rate": 1.6067024517626578e-06, "loss": 0.0022, "step": 25457 }, { "epoch": 2.759973980919341, "grad_norm": 0.5551320314407349, "learning_rate": 1.6059770781952708e-06, "loss": 0.0185, "step": 25458 }, { "epoch": 2.760082393755421, "grad_norm": 3.1017448902130127, "learning_rate": 1.6052517046278833e-06, "loss": 0.0058, "step": 25459 }, { "epoch": 2.7601908065915004, "grad_norm": 0.767362117767334, "learning_rate": 1.6045263310604963e-06, "loss": 0.0125, "step": 25460 }, { "epoch": 2.7602992194275804, "grad_norm": 0.3448871672153473, "learning_rate": 1.603800957493109e-06, "loss": 0.0102, "step": 25461 }, { "epoch": 2.76040763226366, "grad_norm": 0.5044782161712646, "learning_rate": 1.603075583925722e-06, "loss": 0.0108, "step": 25462 }, { "epoch": 2.76051604509974, "grad_norm": 0.09008151292800903, "learning_rate": 1.6023502103583347e-06, "loss": 0.0007, "step": 25463 }, { "epoch": 2.7606244579358195, "grad_norm": 0.09506948292255402, "learning_rate": 1.6016248367909477e-06, "loss": 0.0004, "step": 25464 }, { "epoch": 2.7607328707718994, "grad_norm": 1.7208102941513062, "learning_rate": 1.6008994632235602e-06, "loss": 0.031, "step": 25465 }, { "epoch": 2.760841283607979, "grad_norm": 1.978306531906128, "learning_rate": 1.600174089656173e-06, "loss": 0.0473, "step": 25466 }, { "epoch": 2.760949696444059, "grad_norm": 0.20182651281356812, "learning_rate": 1.5994487160887858e-06, "loss": 0.0007, "step": 25467 }, { "epoch": 2.761058109280139, "grad_norm": 0.19944928586483002, "learning_rate": 1.5987233425213986e-06, "loss": 0.0003, "step": 25468 }, { "epoch": 2.7611665221162185, "grad_norm": 0.11608857661485672, "learning_rate": 1.5979979689540115e-06, "loss": 0.0002, "step": 25469 }, { "epoch": 2.7612749349522985, "grad_norm": 0.3282092213630676, "learning_rate": 1.5972725953866241e-06, "loss": 0.0117, "step": 25470 }, { "epoch": 2.761383347788378, "grad_norm": 0.5148298740386963, "learning_rate": 1.596547221819237e-06, "loss": 0.0148, "step": 25471 }, { "epoch": 2.761491760624458, "grad_norm": 4.757750511169434, "learning_rate": 1.5958218482518497e-06, "loss": 0.0074, "step": 25472 }, { "epoch": 2.761600173460538, "grad_norm": 3.5720109939575195, "learning_rate": 1.5950964746844627e-06, "loss": 0.0373, "step": 25473 }, { "epoch": 2.7617085862966175, "grad_norm": 0.4270031154155731, "learning_rate": 1.5943711011170754e-06, "loss": 0.0086, "step": 25474 }, { "epoch": 2.761816999132697, "grad_norm": 10.166360855102539, "learning_rate": 1.5936457275496884e-06, "loss": 0.0474, "step": 25475 }, { "epoch": 2.761925411968777, "grad_norm": 3.1044843196868896, "learning_rate": 1.592920353982301e-06, "loss": 0.0376, "step": 25476 }, { "epoch": 2.762033824804857, "grad_norm": 0.08129686117172241, "learning_rate": 1.592194980414914e-06, "loss": 0.0005, "step": 25477 }, { "epoch": 2.7621422376409366, "grad_norm": 2.503176212310791, "learning_rate": 1.5914696068475266e-06, "loss": 0.0192, "step": 25478 }, { "epoch": 2.7622506504770166, "grad_norm": 1.6130447387695312, "learning_rate": 1.5907442332801393e-06, "loss": 0.0039, "step": 25479 }, { "epoch": 2.762359063313096, "grad_norm": 0.562445342540741, "learning_rate": 1.5900188597127523e-06, "loss": 0.0163, "step": 25480 }, { "epoch": 2.762467476149176, "grad_norm": 0.6630721688270569, "learning_rate": 1.5892934861453649e-06, "loss": 0.009, "step": 25481 }, { "epoch": 2.762575888985256, "grad_norm": 0.6359907984733582, "learning_rate": 1.5885681125779779e-06, "loss": 0.0017, "step": 25482 }, { "epoch": 2.7626843018213356, "grad_norm": 3.9588751792907715, "learning_rate": 1.5878427390105904e-06, "loss": 0.0339, "step": 25483 }, { "epoch": 2.762792714657415, "grad_norm": 4.640188694000244, "learning_rate": 1.5871173654432034e-06, "loss": 0.0109, "step": 25484 }, { "epoch": 2.762901127493495, "grad_norm": 1.0214985609054565, "learning_rate": 1.586391991875816e-06, "loss": 0.006, "step": 25485 }, { "epoch": 2.763009540329575, "grad_norm": 0.05877748504281044, "learning_rate": 1.585666618308429e-06, "loss": 0.0004, "step": 25486 }, { "epoch": 2.7631179531656547, "grad_norm": 0.09650330245494843, "learning_rate": 1.5849412447410418e-06, "loss": 0.0002, "step": 25487 }, { "epoch": 2.7632263660017347, "grad_norm": 0.2111673355102539, "learning_rate": 1.5842158711736548e-06, "loss": 0.0025, "step": 25488 }, { "epoch": 2.763334778837814, "grad_norm": 0.2581583559513092, "learning_rate": 1.5834904976062673e-06, "loss": 0.0052, "step": 25489 }, { "epoch": 2.763443191673894, "grad_norm": 0.08438844233751297, "learning_rate": 1.5827651240388803e-06, "loss": 0.0004, "step": 25490 }, { "epoch": 2.763551604509974, "grad_norm": 0.698135495185852, "learning_rate": 1.5820397504714929e-06, "loss": 0.0018, "step": 25491 }, { "epoch": 2.7636600173460537, "grad_norm": 0.014951521530747414, "learning_rate": 1.5813143769041057e-06, "loss": 0.0001, "step": 25492 }, { "epoch": 2.7637684301821337, "grad_norm": 2.129409074783325, "learning_rate": 1.5805890033367187e-06, "loss": 0.0017, "step": 25493 }, { "epoch": 2.7638768430182132, "grad_norm": 0.31869712471961975, "learning_rate": 1.5798636297693312e-06, "loss": 0.0052, "step": 25494 }, { "epoch": 2.763985255854293, "grad_norm": 0.38185033202171326, "learning_rate": 1.5791382562019442e-06, "loss": 0.0105, "step": 25495 }, { "epoch": 2.764093668690373, "grad_norm": 0.33329084515571594, "learning_rate": 1.5784128826345568e-06, "loss": 0.004, "step": 25496 }, { "epoch": 2.7642020815264527, "grad_norm": 0.6542454957962036, "learning_rate": 1.5776875090671698e-06, "loss": 0.0162, "step": 25497 }, { "epoch": 2.7643104943625323, "grad_norm": 0.34702226519584656, "learning_rate": 1.5769621354997823e-06, "loss": 0.0052, "step": 25498 }, { "epoch": 2.7644189071986123, "grad_norm": 5.975630760192871, "learning_rate": 1.5762367619323953e-06, "loss": 0.023, "step": 25499 }, { "epoch": 2.7645273200346923, "grad_norm": 0.4972856938838959, "learning_rate": 1.5755113883650081e-06, "loss": 0.0103, "step": 25500 }, { "epoch": 2.764635732870772, "grad_norm": 0.2560569941997528, "learning_rate": 1.574786014797621e-06, "loss": 0.0053, "step": 25501 }, { "epoch": 2.764744145706852, "grad_norm": 0.4501240849494934, "learning_rate": 1.5740606412302337e-06, "loss": 0.0166, "step": 25502 }, { "epoch": 2.7648525585429313, "grad_norm": 0.955910325050354, "learning_rate": 1.5733352676628467e-06, "loss": 0.0076, "step": 25503 }, { "epoch": 2.7649609713790113, "grad_norm": 0.6631568074226379, "learning_rate": 1.5726098940954592e-06, "loss": 0.0329, "step": 25504 }, { "epoch": 2.7650693842150913, "grad_norm": 0.5381293892860413, "learning_rate": 1.5718845205280722e-06, "loss": 0.0075, "step": 25505 }, { "epoch": 2.765177797051171, "grad_norm": 5.917121410369873, "learning_rate": 1.571159146960685e-06, "loss": 0.0393, "step": 25506 }, { "epoch": 2.7652862098872504, "grad_norm": 12.256048202514648, "learning_rate": 1.5704337733932976e-06, "loss": 0.0346, "step": 25507 }, { "epoch": 2.7653946227233304, "grad_norm": 0.12963035702705383, "learning_rate": 1.5697083998259105e-06, "loss": 0.0006, "step": 25508 }, { "epoch": 2.7655030355594103, "grad_norm": 0.019948547706007957, "learning_rate": 1.5689830262585231e-06, "loss": 0.0002, "step": 25509 }, { "epoch": 2.76561144839549, "grad_norm": 0.39364373683929443, "learning_rate": 1.568257652691136e-06, "loss": 0.0127, "step": 25510 }, { "epoch": 2.76571986123157, "grad_norm": 0.5819142460823059, "learning_rate": 1.5675322791237489e-06, "loss": 0.0121, "step": 25511 }, { "epoch": 2.7658282740676494, "grad_norm": 7.317320823669434, "learning_rate": 1.5668069055563617e-06, "loss": 0.0304, "step": 25512 }, { "epoch": 2.7659366869037294, "grad_norm": 0.6009922027587891, "learning_rate": 1.5660815319889744e-06, "loss": 0.0284, "step": 25513 }, { "epoch": 2.7660450997398094, "grad_norm": 0.01870127208530903, "learning_rate": 1.5653561584215874e-06, "loss": 0.0002, "step": 25514 }, { "epoch": 2.766153512575889, "grad_norm": 0.024717116728425026, "learning_rate": 1.5646307848542e-06, "loss": 0.0002, "step": 25515 }, { "epoch": 2.766261925411969, "grad_norm": 0.022901268675923347, "learning_rate": 1.563905411286813e-06, "loss": 0.0001, "step": 25516 }, { "epoch": 2.7663703382480485, "grad_norm": 0.25927793979644775, "learning_rate": 1.5631800377194256e-06, "loss": 0.0053, "step": 25517 }, { "epoch": 2.7664787510841284, "grad_norm": 0.3207792639732361, "learning_rate": 1.5624546641520385e-06, "loss": 0.011, "step": 25518 }, { "epoch": 2.7665871639202084, "grad_norm": 0.19893738627433777, "learning_rate": 1.5617292905846513e-06, "loss": 0.0007, "step": 25519 }, { "epoch": 2.766695576756288, "grad_norm": 0.5398233532905579, "learning_rate": 1.5610039170172639e-06, "loss": 0.0109, "step": 25520 }, { "epoch": 2.7668039895923675, "grad_norm": 3.4845352172851562, "learning_rate": 1.5602785434498769e-06, "loss": 0.0529, "step": 25521 }, { "epoch": 2.7669124024284475, "grad_norm": 0.36459508538246155, "learning_rate": 1.5595531698824894e-06, "loss": 0.0046, "step": 25522 }, { "epoch": 2.7670208152645275, "grad_norm": 0.6653763651847839, "learning_rate": 1.5588277963151024e-06, "loss": 0.0264, "step": 25523 }, { "epoch": 2.767129228100607, "grad_norm": 0.02346370555460453, "learning_rate": 1.5581024227477152e-06, "loss": 0.0002, "step": 25524 }, { "epoch": 2.767237640936687, "grad_norm": 0.02219361811876297, "learning_rate": 1.557377049180328e-06, "loss": 0.0001, "step": 25525 }, { "epoch": 2.7673460537727665, "grad_norm": 2.356433153152466, "learning_rate": 1.5566516756129408e-06, "loss": 0.022, "step": 25526 }, { "epoch": 2.7674544666088465, "grad_norm": 0.01615304872393608, "learning_rate": 1.5559263020455538e-06, "loss": 0.0001, "step": 25527 }, { "epoch": 2.7675628794449265, "grad_norm": 0.35181304812431335, "learning_rate": 1.5552009284781663e-06, "loss": 0.0108, "step": 25528 }, { "epoch": 2.767671292281006, "grad_norm": 0.031195851042866707, "learning_rate": 1.5544755549107793e-06, "loss": 0.0003, "step": 25529 }, { "epoch": 2.7677797051170856, "grad_norm": 0.03580149635672569, "learning_rate": 1.5537501813433919e-06, "loss": 0.0002, "step": 25530 }, { "epoch": 2.7678881179531656, "grad_norm": 0.00798735674470663, "learning_rate": 1.5530248077760049e-06, "loss": 0.0, "step": 25531 }, { "epoch": 2.7679965307892456, "grad_norm": 0.8352863788604736, "learning_rate": 1.5522994342086177e-06, "loss": 0.0131, "step": 25532 }, { "epoch": 2.768104943625325, "grad_norm": 3.124338388442993, "learning_rate": 1.5515740606412302e-06, "loss": 0.064, "step": 25533 }, { "epoch": 2.768213356461405, "grad_norm": 0.7313713431358337, "learning_rate": 1.5508486870738432e-06, "loss": 0.0215, "step": 25534 }, { "epoch": 2.7683217692974846, "grad_norm": 0.07479842752218246, "learning_rate": 1.5501233135064558e-06, "loss": 0.0009, "step": 25535 }, { "epoch": 2.7684301821335646, "grad_norm": 0.07558014988899231, "learning_rate": 1.5493979399390688e-06, "loss": 0.0002, "step": 25536 }, { "epoch": 2.7685385949696446, "grad_norm": 0.342375248670578, "learning_rate": 1.5486725663716816e-06, "loss": 0.0099, "step": 25537 }, { "epoch": 2.768647007805724, "grad_norm": 0.04679543152451515, "learning_rate": 1.5479471928042943e-06, "loss": 0.0002, "step": 25538 }, { "epoch": 2.768755420641804, "grad_norm": 0.01223582774400711, "learning_rate": 1.5472218192369071e-06, "loss": 0.0001, "step": 25539 }, { "epoch": 2.7688638334778837, "grad_norm": 0.5121011137962341, "learning_rate": 1.54649644566952e-06, "loss": 0.0156, "step": 25540 }, { "epoch": 2.7689722463139637, "grad_norm": 0.009342371486127377, "learning_rate": 1.5457710721021327e-06, "loss": 0.0001, "step": 25541 }, { "epoch": 2.769080659150043, "grad_norm": 0.49562937021255493, "learning_rate": 1.5450456985347457e-06, "loss": 0.0122, "step": 25542 }, { "epoch": 2.769189071986123, "grad_norm": 2.225144147872925, "learning_rate": 1.5443203249673582e-06, "loss": 0.035, "step": 25543 }, { "epoch": 2.7692974848222027, "grad_norm": 5.005491733551025, "learning_rate": 1.5435949513999712e-06, "loss": 0.0181, "step": 25544 }, { "epoch": 2.7694058976582827, "grad_norm": 0.38417571783065796, "learning_rate": 1.542869577832584e-06, "loss": 0.0095, "step": 25545 }, { "epoch": 2.7695143104943627, "grad_norm": 0.17161710560321808, "learning_rate": 1.5421442042651966e-06, "loss": 0.0008, "step": 25546 }, { "epoch": 2.7696227233304422, "grad_norm": 0.33569273352622986, "learning_rate": 1.5414188306978096e-06, "loss": 0.0101, "step": 25547 }, { "epoch": 2.7697311361665222, "grad_norm": 0.5545384287834167, "learning_rate": 1.5406934571304221e-06, "loss": 0.0133, "step": 25548 }, { "epoch": 2.7698395490026018, "grad_norm": 3.740445852279663, "learning_rate": 1.5399680835630351e-06, "loss": 0.0179, "step": 25549 }, { "epoch": 2.7699479618386817, "grad_norm": 0.2654224932193756, "learning_rate": 1.5392427099956479e-06, "loss": 0.0055, "step": 25550 }, { "epoch": 2.7700563746747617, "grad_norm": 0.31348705291748047, "learning_rate": 1.5385173364282607e-06, "loss": 0.0067, "step": 25551 }, { "epoch": 2.7701647875108413, "grad_norm": 4.284783363342285, "learning_rate": 1.5377919628608734e-06, "loss": 0.1014, "step": 25552 }, { "epoch": 2.770273200346921, "grad_norm": 0.0999884232878685, "learning_rate": 1.5370665892934864e-06, "loss": 0.0003, "step": 25553 }, { "epoch": 2.770381613183001, "grad_norm": 0.020155392587184906, "learning_rate": 1.536341215726099e-06, "loss": 0.0002, "step": 25554 }, { "epoch": 2.770490026019081, "grad_norm": 7.579788684844971, "learning_rate": 1.535615842158712e-06, "loss": 0.0145, "step": 25555 }, { "epoch": 2.7705984388551603, "grad_norm": 0.487804114818573, "learning_rate": 1.5348904685913246e-06, "loss": 0.0166, "step": 25556 }, { "epoch": 2.7707068516912403, "grad_norm": 0.21001505851745605, "learning_rate": 1.5341650950239375e-06, "loss": 0.0032, "step": 25557 }, { "epoch": 2.77081526452732, "grad_norm": 0.36445534229278564, "learning_rate": 1.5334397214565503e-06, "loss": 0.0084, "step": 25558 }, { "epoch": 2.7709236773634, "grad_norm": 0.15780608355998993, "learning_rate": 1.5327143478891631e-06, "loss": 0.0006, "step": 25559 }, { "epoch": 2.77103209019948, "grad_norm": 0.055876392871141434, "learning_rate": 1.5319889743217759e-06, "loss": 0.0004, "step": 25560 }, { "epoch": 2.7711405030355594, "grad_norm": 0.6206076145172119, "learning_rate": 1.5312636007543885e-06, "loss": 0.0074, "step": 25561 }, { "epoch": 2.7712489158716394, "grad_norm": 0.01749921590089798, "learning_rate": 1.5305382271870014e-06, "loss": 0.0002, "step": 25562 }, { "epoch": 2.771357328707719, "grad_norm": 0.30301156640052795, "learning_rate": 1.5298128536196142e-06, "loss": 0.0063, "step": 25563 }, { "epoch": 2.771465741543799, "grad_norm": 0.34984543919563293, "learning_rate": 1.529087480052227e-06, "loss": 0.0127, "step": 25564 }, { "epoch": 2.7715741543798784, "grad_norm": 3.0620834827423096, "learning_rate": 1.5283621064848398e-06, "loss": 0.0484, "step": 25565 }, { "epoch": 2.7716825672159584, "grad_norm": 3.9861414432525635, "learning_rate": 1.5276367329174528e-06, "loss": 0.0103, "step": 25566 }, { "epoch": 2.771790980052038, "grad_norm": 0.20152005553245544, "learning_rate": 1.5269113593500653e-06, "loss": 0.0008, "step": 25567 }, { "epoch": 2.771899392888118, "grad_norm": 0.4633238613605499, "learning_rate": 1.5261859857826783e-06, "loss": 0.0246, "step": 25568 }, { "epoch": 2.772007805724198, "grad_norm": 0.34741196036338806, "learning_rate": 1.5254606122152909e-06, "loss": 0.0183, "step": 25569 }, { "epoch": 2.7721162185602775, "grad_norm": 2.7204787731170654, "learning_rate": 1.5247352386479039e-06, "loss": 0.0159, "step": 25570 }, { "epoch": 2.7722246313963574, "grad_norm": 0.05292452499270439, "learning_rate": 1.5240098650805167e-06, "loss": 0.0004, "step": 25571 }, { "epoch": 2.772333044232437, "grad_norm": 0.02717828005552292, "learning_rate": 1.5232844915131294e-06, "loss": 0.0002, "step": 25572 }, { "epoch": 2.772441457068517, "grad_norm": 0.9115967750549316, "learning_rate": 1.5225591179457422e-06, "loss": 0.0381, "step": 25573 }, { "epoch": 2.772549869904597, "grad_norm": 0.012140716426074505, "learning_rate": 1.5218337443783548e-06, "loss": 0.0001, "step": 25574 }, { "epoch": 2.7726582827406765, "grad_norm": 2.1518304347991943, "learning_rate": 1.5211083708109678e-06, "loss": 0.0179, "step": 25575 }, { "epoch": 2.772766695576756, "grad_norm": 6.762992858886719, "learning_rate": 1.5203829972435806e-06, "loss": 0.0512, "step": 25576 }, { "epoch": 2.772875108412836, "grad_norm": 0.027432294562458992, "learning_rate": 1.5196576236761933e-06, "loss": 0.0003, "step": 25577 }, { "epoch": 2.772983521248916, "grad_norm": 0.7822034955024719, "learning_rate": 1.5189322501088061e-06, "loss": 0.0109, "step": 25578 }, { "epoch": 2.7730919340849955, "grad_norm": 0.5336945652961731, "learning_rate": 1.518206876541419e-06, "loss": 0.0009, "step": 25579 }, { "epoch": 2.7732003469210755, "grad_norm": 0.006019476801156998, "learning_rate": 1.5174815029740317e-06, "loss": 0.0001, "step": 25580 }, { "epoch": 2.773308759757155, "grad_norm": 2.2583565711975098, "learning_rate": 1.5167561294066447e-06, "loss": 0.0057, "step": 25581 }, { "epoch": 2.773417172593235, "grad_norm": 0.04230494052171707, "learning_rate": 1.5160307558392572e-06, "loss": 0.0003, "step": 25582 }, { "epoch": 2.773525585429315, "grad_norm": 0.3068939447402954, "learning_rate": 1.5153053822718702e-06, "loss": 0.0112, "step": 25583 }, { "epoch": 2.7736339982653946, "grad_norm": 0.08693677186965942, "learning_rate": 1.514580008704483e-06, "loss": 0.0002, "step": 25584 }, { "epoch": 2.7737424111014746, "grad_norm": 1.6929916143417358, "learning_rate": 1.5138546351370958e-06, "loss": 0.0017, "step": 25585 }, { "epoch": 2.773850823937554, "grad_norm": 0.012363271787762642, "learning_rate": 1.5131292615697086e-06, "loss": 0.0001, "step": 25586 }, { "epoch": 2.773959236773634, "grad_norm": 0.020051708444952965, "learning_rate": 1.5124038880023211e-06, "loss": 0.0001, "step": 25587 }, { "epoch": 2.7740676496097136, "grad_norm": 0.2031787484884262, "learning_rate": 1.5116785144349341e-06, "loss": 0.0028, "step": 25588 }, { "epoch": 2.7741760624457936, "grad_norm": 0.19306102395057678, "learning_rate": 1.5109531408675469e-06, "loss": 0.0006, "step": 25589 }, { "epoch": 2.774284475281873, "grad_norm": 9.848982810974121, "learning_rate": 1.5102277673001597e-06, "loss": 0.009, "step": 25590 }, { "epoch": 2.774392888117953, "grad_norm": 8.453859329223633, "learning_rate": 1.5095023937327724e-06, "loss": 0.0852, "step": 25591 }, { "epoch": 2.774501300954033, "grad_norm": 4.498454570770264, "learning_rate": 1.5087770201653854e-06, "loss": 0.0367, "step": 25592 }, { "epoch": 2.7746097137901127, "grad_norm": 3.188802480697632, "learning_rate": 1.508051646597998e-06, "loss": 0.0067, "step": 25593 }, { "epoch": 2.7747181266261927, "grad_norm": 0.06590424478054047, "learning_rate": 1.507326273030611e-06, "loss": 0.0003, "step": 25594 }, { "epoch": 2.774826539462272, "grad_norm": 0.03938847780227661, "learning_rate": 1.5066008994632236e-06, "loss": 0.0003, "step": 25595 }, { "epoch": 2.774934952298352, "grad_norm": 0.2568212151527405, "learning_rate": 1.5058755258958366e-06, "loss": 0.0055, "step": 25596 }, { "epoch": 2.775043365134432, "grad_norm": 18.112924575805664, "learning_rate": 1.5051501523284493e-06, "loss": 0.0494, "step": 25597 }, { "epoch": 2.7751517779705117, "grad_norm": 9.82761287689209, "learning_rate": 1.5044247787610621e-06, "loss": 0.006, "step": 25598 }, { "epoch": 2.7752601908065913, "grad_norm": 7.68851900100708, "learning_rate": 1.5036994051936749e-06, "loss": 0.0421, "step": 25599 }, { "epoch": 2.7753686036426712, "grad_norm": 0.2720091640949249, "learning_rate": 1.5029740316262879e-06, "loss": 0.0055, "step": 25600 }, { "epoch": 2.7754770164787512, "grad_norm": 4.189598560333252, "learning_rate": 1.5022486580589004e-06, "loss": 0.0241, "step": 25601 }, { "epoch": 2.7755854293148308, "grad_norm": 0.5852919816970825, "learning_rate": 1.5015232844915132e-06, "loss": 0.0097, "step": 25602 }, { "epoch": 2.7756938421509108, "grad_norm": 0.2791751027107239, "learning_rate": 1.500797910924126e-06, "loss": 0.0061, "step": 25603 }, { "epoch": 2.7758022549869903, "grad_norm": 0.3782155215740204, "learning_rate": 1.5000725373567388e-06, "loss": 0.0115, "step": 25604 }, { "epoch": 2.7759106678230703, "grad_norm": 0.11153329163789749, "learning_rate": 1.4993471637893518e-06, "loss": 0.0003, "step": 25605 }, { "epoch": 2.7760190806591503, "grad_norm": 0.21126742660999298, "learning_rate": 1.4986217902219643e-06, "loss": 0.0148, "step": 25606 }, { "epoch": 2.77612749349523, "grad_norm": 0.02670935168862343, "learning_rate": 1.4978964166545773e-06, "loss": 0.0001, "step": 25607 }, { "epoch": 2.77623590633131, "grad_norm": 0.3442491888999939, "learning_rate": 1.49717104308719e-06, "loss": 0.0109, "step": 25608 }, { "epoch": 2.7763443191673893, "grad_norm": 0.6955574154853821, "learning_rate": 1.4964456695198029e-06, "loss": 0.0103, "step": 25609 }, { "epoch": 2.7764527320034693, "grad_norm": 0.08549024164676666, "learning_rate": 1.4957202959524157e-06, "loss": 0.0005, "step": 25610 }, { "epoch": 2.776561144839549, "grad_norm": 5.9825119972229, "learning_rate": 1.4949949223850284e-06, "loss": 0.0537, "step": 25611 }, { "epoch": 2.776669557675629, "grad_norm": 0.009987263940274715, "learning_rate": 1.4942695488176412e-06, "loss": 0.0001, "step": 25612 }, { "epoch": 2.7767779705117084, "grad_norm": 0.062025487422943115, "learning_rate": 1.4935441752502542e-06, "loss": 0.0005, "step": 25613 }, { "epoch": 2.7768863833477884, "grad_norm": 0.460726797580719, "learning_rate": 1.4928188016828668e-06, "loss": 0.0068, "step": 25614 }, { "epoch": 2.7769947961838684, "grad_norm": 0.4477218687534332, "learning_rate": 1.4920934281154796e-06, "loss": 0.0154, "step": 25615 }, { "epoch": 2.777103209019948, "grad_norm": 0.04309292882680893, "learning_rate": 1.4913680545480923e-06, "loss": 0.0002, "step": 25616 }, { "epoch": 2.777211621856028, "grad_norm": 0.2829902470111847, "learning_rate": 1.4906426809807051e-06, "loss": 0.0048, "step": 25617 }, { "epoch": 2.7773200346921074, "grad_norm": 0.2696223258972168, "learning_rate": 1.4899173074133181e-06, "loss": 0.0055, "step": 25618 }, { "epoch": 2.7774284475281874, "grad_norm": 3.8127965927124023, "learning_rate": 1.4891919338459307e-06, "loss": 0.0352, "step": 25619 }, { "epoch": 2.7775368603642674, "grad_norm": 0.07641997933387756, "learning_rate": 1.4884665602785437e-06, "loss": 0.0002, "step": 25620 }, { "epoch": 2.777645273200347, "grad_norm": 5.7320027351379395, "learning_rate": 1.4877411867111562e-06, "loss": 0.0209, "step": 25621 }, { "epoch": 2.7777536860364265, "grad_norm": 1.9067753553390503, "learning_rate": 1.4870158131437692e-06, "loss": 0.0061, "step": 25622 }, { "epoch": 2.7778620988725065, "grad_norm": 11.058025360107422, "learning_rate": 1.486290439576382e-06, "loss": 0.0097, "step": 25623 }, { "epoch": 2.7779705117085864, "grad_norm": 3.3406219482421875, "learning_rate": 1.4855650660089948e-06, "loss": 0.0076, "step": 25624 }, { "epoch": 2.778078924544666, "grad_norm": 0.02464851550757885, "learning_rate": 1.4848396924416076e-06, "loss": 0.0001, "step": 25625 }, { "epoch": 2.778187337380746, "grad_norm": 0.245451420545578, "learning_rate": 1.4841143188742205e-06, "loss": 0.0014, "step": 25626 }, { "epoch": 2.7782957502168255, "grad_norm": 0.27202239632606506, "learning_rate": 1.4833889453068331e-06, "loss": 0.0058, "step": 25627 }, { "epoch": 2.7784041630529055, "grad_norm": 6.146726131439209, "learning_rate": 1.4826635717394459e-06, "loss": 0.0041, "step": 25628 }, { "epoch": 2.7785125758889855, "grad_norm": 0.3203810453414917, "learning_rate": 1.4819381981720587e-06, "loss": 0.011, "step": 25629 }, { "epoch": 2.778620988725065, "grad_norm": 0.292485773563385, "learning_rate": 1.4812128246046715e-06, "loss": 0.0061, "step": 25630 }, { "epoch": 2.7787294015611446, "grad_norm": 1.610004186630249, "learning_rate": 1.4804874510372844e-06, "loss": 0.0419, "step": 25631 }, { "epoch": 2.7788378143972245, "grad_norm": 0.061434682458639145, "learning_rate": 1.479762077469897e-06, "loss": 0.0002, "step": 25632 }, { "epoch": 2.7789462272333045, "grad_norm": 0.272195965051651, "learning_rate": 1.47903670390251e-06, "loss": 0.0009, "step": 25633 }, { "epoch": 2.779054640069384, "grad_norm": 0.12587207555770874, "learning_rate": 1.4783113303351226e-06, "loss": 0.0002, "step": 25634 }, { "epoch": 2.779163052905464, "grad_norm": 0.3114974796772003, "learning_rate": 1.4775859567677356e-06, "loss": 0.009, "step": 25635 }, { "epoch": 2.7792714657415436, "grad_norm": 0.39811912178993225, "learning_rate": 1.4768605832003483e-06, "loss": 0.0144, "step": 25636 }, { "epoch": 2.7793798785776236, "grad_norm": 0.14423249661922455, "learning_rate": 1.4761352096329611e-06, "loss": 0.0004, "step": 25637 }, { "epoch": 2.7794882914137036, "grad_norm": 0.7229858040809631, "learning_rate": 1.4754098360655739e-06, "loss": 0.0259, "step": 25638 }, { "epoch": 2.779596704249783, "grad_norm": 0.3793623745441437, "learning_rate": 1.4746844624981869e-06, "loss": 0.0106, "step": 25639 }, { "epoch": 2.779705117085863, "grad_norm": 0.5487468838691711, "learning_rate": 1.4739590889307994e-06, "loss": 0.0154, "step": 25640 }, { "epoch": 2.7798135299219426, "grad_norm": 0.46971261501312256, "learning_rate": 1.4732337153634122e-06, "loss": 0.0139, "step": 25641 }, { "epoch": 2.7799219427580226, "grad_norm": 0.14954626560211182, "learning_rate": 1.472508341796025e-06, "loss": 0.0005, "step": 25642 }, { "epoch": 2.7800303555941026, "grad_norm": 19.160358428955078, "learning_rate": 1.4717829682286378e-06, "loss": 0.0279, "step": 25643 }, { "epoch": 2.780138768430182, "grad_norm": 0.3377257287502289, "learning_rate": 1.4710575946612508e-06, "loss": 0.0114, "step": 25644 }, { "epoch": 2.7802471812662617, "grad_norm": 0.010217149741947651, "learning_rate": 1.4703322210938633e-06, "loss": 0.0001, "step": 25645 }, { "epoch": 2.7803555941023417, "grad_norm": 0.4883235991001129, "learning_rate": 1.4696068475264763e-06, "loss": 0.0114, "step": 25646 }, { "epoch": 2.7804640069384217, "grad_norm": 0.007883859798312187, "learning_rate": 1.468881473959089e-06, "loss": 0.0001, "step": 25647 }, { "epoch": 2.780572419774501, "grad_norm": 0.06044185161590576, "learning_rate": 1.4681561003917019e-06, "loss": 0.0003, "step": 25648 }, { "epoch": 2.780680832610581, "grad_norm": 0.046369586139917374, "learning_rate": 1.4674307268243147e-06, "loss": 0.0002, "step": 25649 }, { "epoch": 2.7807892454466607, "grad_norm": 4.1702961921691895, "learning_rate": 1.4667053532569274e-06, "loss": 0.0119, "step": 25650 }, { "epoch": 2.7808976582827407, "grad_norm": 0.8605561852455139, "learning_rate": 1.4659799796895402e-06, "loss": 0.0165, "step": 25651 }, { "epoch": 2.7810060711188207, "grad_norm": 7.22545862197876, "learning_rate": 1.4652546061221532e-06, "loss": 0.034, "step": 25652 }, { "epoch": 2.7811144839549002, "grad_norm": 0.33437973260879517, "learning_rate": 1.4645292325547658e-06, "loss": 0.0099, "step": 25653 }, { "epoch": 2.78122289679098, "grad_norm": 8.504288673400879, "learning_rate": 1.4638038589873788e-06, "loss": 0.0188, "step": 25654 }, { "epoch": 2.7813313096270598, "grad_norm": 0.011898205615580082, "learning_rate": 1.4630784854199913e-06, "loss": 0.0001, "step": 25655 }, { "epoch": 2.7814397224631398, "grad_norm": 0.33755096793174744, "learning_rate": 1.4623531118526041e-06, "loss": 0.0114, "step": 25656 }, { "epoch": 2.7815481352992193, "grad_norm": 0.41044577956199646, "learning_rate": 1.4616277382852171e-06, "loss": 0.0097, "step": 25657 }, { "epoch": 2.7816565481352993, "grad_norm": 2.209979295730591, "learning_rate": 1.4609023647178297e-06, "loss": 0.0372, "step": 25658 }, { "epoch": 2.781764960971379, "grad_norm": 0.37397873401641846, "learning_rate": 1.4601769911504427e-06, "loss": 0.0117, "step": 25659 }, { "epoch": 2.781873373807459, "grad_norm": 0.2169976532459259, "learning_rate": 1.4594516175830552e-06, "loss": 0.0033, "step": 25660 }, { "epoch": 2.781981786643539, "grad_norm": 0.38135266304016113, "learning_rate": 1.4587262440156682e-06, "loss": 0.0088, "step": 25661 }, { "epoch": 2.7820901994796183, "grad_norm": 0.8497214317321777, "learning_rate": 1.458000870448281e-06, "loss": 0.0015, "step": 25662 }, { "epoch": 2.7821986123156983, "grad_norm": 0.42718613147735596, "learning_rate": 1.4572754968808938e-06, "loss": 0.0164, "step": 25663 }, { "epoch": 2.782307025151778, "grad_norm": 2.062587261199951, "learning_rate": 1.4565501233135066e-06, "loss": 0.042, "step": 25664 }, { "epoch": 2.782415437987858, "grad_norm": 0.4454301595687866, "learning_rate": 1.4558247497461196e-06, "loss": 0.0006, "step": 25665 }, { "epoch": 2.782523850823938, "grad_norm": 0.47389471530914307, "learning_rate": 1.4550993761787321e-06, "loss": 0.0274, "step": 25666 }, { "epoch": 2.7826322636600174, "grad_norm": 0.0218901839107275, "learning_rate": 1.4543740026113451e-06, "loss": 0.0002, "step": 25667 }, { "epoch": 2.782740676496097, "grad_norm": 0.024077335372567177, "learning_rate": 1.4536486290439577e-06, "loss": 0.0001, "step": 25668 }, { "epoch": 2.782849089332177, "grad_norm": 3.758044719696045, "learning_rate": 1.4529232554765705e-06, "loss": 0.008, "step": 25669 }, { "epoch": 2.782957502168257, "grad_norm": 0.028630372136831284, "learning_rate": 1.4521978819091834e-06, "loss": 0.0001, "step": 25670 }, { "epoch": 2.7830659150043364, "grad_norm": 0.014726524241268635, "learning_rate": 1.451472508341796e-06, "loss": 0.0001, "step": 25671 }, { "epoch": 2.7831743278404164, "grad_norm": 2.3334078788757324, "learning_rate": 1.450747134774409e-06, "loss": 0.0507, "step": 25672 }, { "epoch": 2.783282740676496, "grad_norm": 0.40127336978912354, "learning_rate": 1.4500217612070216e-06, "loss": 0.0183, "step": 25673 }, { "epoch": 2.783391153512576, "grad_norm": 0.03724152222275734, "learning_rate": 1.4492963876396346e-06, "loss": 0.0003, "step": 25674 }, { "epoch": 2.783499566348656, "grad_norm": 0.22731803357601166, "learning_rate": 1.4485710140722473e-06, "loss": 0.0032, "step": 25675 }, { "epoch": 2.7836079791847355, "grad_norm": 0.30307674407958984, "learning_rate": 1.4478456405048601e-06, "loss": 0.0061, "step": 25676 }, { "epoch": 2.783716392020815, "grad_norm": 0.2624310553073883, "learning_rate": 1.447120266937473e-06, "loss": 0.0053, "step": 25677 }, { "epoch": 2.783824804856895, "grad_norm": 0.34146279096603394, "learning_rate": 1.4463948933700859e-06, "loss": 0.0113, "step": 25678 }, { "epoch": 2.783933217692975, "grad_norm": 6.5432586669921875, "learning_rate": 1.4456695198026985e-06, "loss": 0.0276, "step": 25679 }, { "epoch": 2.7840416305290545, "grad_norm": 0.3987199068069458, "learning_rate": 1.4449441462353114e-06, "loss": 0.0113, "step": 25680 }, { "epoch": 2.7841500433651345, "grad_norm": 0.5071910619735718, "learning_rate": 1.444218772667924e-06, "loss": 0.0226, "step": 25681 }, { "epoch": 2.784258456201214, "grad_norm": 0.8033077120780945, "learning_rate": 1.4434933991005368e-06, "loss": 0.0016, "step": 25682 }, { "epoch": 2.784366869037294, "grad_norm": 0.007037763018161058, "learning_rate": 1.4427680255331498e-06, "loss": 0.0001, "step": 25683 }, { "epoch": 2.784475281873374, "grad_norm": 0.30471596121788025, "learning_rate": 1.4420426519657623e-06, "loss": 0.0062, "step": 25684 }, { "epoch": 2.7845836947094535, "grad_norm": 0.3970364034175873, "learning_rate": 1.4413172783983753e-06, "loss": 0.0006, "step": 25685 }, { "epoch": 2.7846921075455335, "grad_norm": 0.33002468943595886, "learning_rate": 1.440591904830988e-06, "loss": 0.0117, "step": 25686 }, { "epoch": 2.784800520381613, "grad_norm": 0.13172443211078644, "learning_rate": 1.4398665312636009e-06, "loss": 0.0112, "step": 25687 }, { "epoch": 2.784908933217693, "grad_norm": 0.04350026696920395, "learning_rate": 1.4391411576962137e-06, "loss": 0.0002, "step": 25688 }, { "epoch": 2.7850173460537726, "grad_norm": 3.659569263458252, "learning_rate": 1.4384157841288265e-06, "loss": 0.0163, "step": 25689 }, { "epoch": 2.7851257588898526, "grad_norm": 0.8224115371704102, "learning_rate": 1.4376904105614392e-06, "loss": 0.0031, "step": 25690 }, { "epoch": 2.785234171725932, "grad_norm": 0.2970806658267975, "learning_rate": 1.4369650369940522e-06, "loss": 0.0062, "step": 25691 }, { "epoch": 2.785342584562012, "grad_norm": 0.007709212601184845, "learning_rate": 1.4362396634266648e-06, "loss": 0.0001, "step": 25692 }, { "epoch": 2.785450997398092, "grad_norm": 0.20962470769882202, "learning_rate": 1.4355142898592778e-06, "loss": 0.0009, "step": 25693 }, { "epoch": 2.7855594102341716, "grad_norm": 0.7345011830329895, "learning_rate": 1.4347889162918903e-06, "loss": 0.0127, "step": 25694 }, { "epoch": 2.7856678230702516, "grad_norm": 1.724869966506958, "learning_rate": 1.4340635427245031e-06, "loss": 0.0202, "step": 25695 }, { "epoch": 2.785776235906331, "grad_norm": 0.28234726190567017, "learning_rate": 1.4333381691571161e-06, "loss": 0.0055, "step": 25696 }, { "epoch": 2.785884648742411, "grad_norm": 0.30330318212509155, "learning_rate": 1.4326127955897287e-06, "loss": 0.0059, "step": 25697 }, { "epoch": 2.785993061578491, "grad_norm": 0.01992129348218441, "learning_rate": 1.4318874220223417e-06, "loss": 0.0002, "step": 25698 }, { "epoch": 2.7861014744145707, "grad_norm": 1.1526654958724976, "learning_rate": 1.4311620484549542e-06, "loss": 0.0021, "step": 25699 }, { "epoch": 2.78620988725065, "grad_norm": 27.4650821685791, "learning_rate": 1.4304366748875672e-06, "loss": 0.031, "step": 25700 }, { "epoch": 2.78631830008673, "grad_norm": 0.3089533746242523, "learning_rate": 1.42971130132018e-06, "loss": 0.0063, "step": 25701 }, { "epoch": 2.78642671292281, "grad_norm": 0.44131219387054443, "learning_rate": 1.4289859277527928e-06, "loss": 0.0141, "step": 25702 }, { "epoch": 2.7865351257588897, "grad_norm": 0.35804054141044617, "learning_rate": 1.4282605541854056e-06, "loss": 0.0229, "step": 25703 }, { "epoch": 2.7866435385949697, "grad_norm": 0.07026337087154388, "learning_rate": 1.4275351806180186e-06, "loss": 0.0003, "step": 25704 }, { "epoch": 2.7867519514310493, "grad_norm": 0.9074443578720093, "learning_rate": 1.4268098070506311e-06, "loss": 0.0017, "step": 25705 }, { "epoch": 2.7868603642671292, "grad_norm": 4.776157855987549, "learning_rate": 1.4260844334832441e-06, "loss": 0.0493, "step": 25706 }, { "epoch": 2.7869687771032092, "grad_norm": 0.44366225600242615, "learning_rate": 1.4253590599158567e-06, "loss": 0.0168, "step": 25707 }, { "epoch": 2.7870771899392888, "grad_norm": 0.7358449101448059, "learning_rate": 1.4246336863484697e-06, "loss": 0.0115, "step": 25708 }, { "epoch": 2.7871856027753688, "grad_norm": 0.12620984017848969, "learning_rate": 1.4239083127810824e-06, "loss": 0.0004, "step": 25709 }, { "epoch": 2.7872940156114483, "grad_norm": 0.3059351444244385, "learning_rate": 1.423182939213695e-06, "loss": 0.0166, "step": 25710 }, { "epoch": 2.7874024284475283, "grad_norm": 0.04735717922449112, "learning_rate": 1.422457565646308e-06, "loss": 0.0004, "step": 25711 }, { "epoch": 2.787510841283608, "grad_norm": 0.007401233538985252, "learning_rate": 1.4217321920789206e-06, "loss": 0.0001, "step": 25712 }, { "epoch": 2.787619254119688, "grad_norm": 0.4056544601917267, "learning_rate": 1.4210068185115336e-06, "loss": 0.0112, "step": 25713 }, { "epoch": 2.7877276669557673, "grad_norm": 9.886877059936523, "learning_rate": 1.4202814449441463e-06, "loss": 0.0059, "step": 25714 }, { "epoch": 2.7878360797918473, "grad_norm": 8.574557304382324, "learning_rate": 1.4195560713767591e-06, "loss": 0.0093, "step": 25715 }, { "epoch": 2.7879444926279273, "grad_norm": 0.3849795460700989, "learning_rate": 1.418830697809372e-06, "loss": 0.0108, "step": 25716 }, { "epoch": 2.788052905464007, "grad_norm": 7.742315292358398, "learning_rate": 1.4181053242419849e-06, "loss": 0.0283, "step": 25717 }, { "epoch": 2.788161318300087, "grad_norm": 0.4946001172065735, "learning_rate": 1.4173799506745975e-06, "loss": 0.0167, "step": 25718 }, { "epoch": 2.7882697311361664, "grad_norm": 2.551898717880249, "learning_rate": 1.4166545771072104e-06, "loss": 0.0252, "step": 25719 }, { "epoch": 2.7883781439722464, "grad_norm": 0.28954485058784485, "learning_rate": 1.415929203539823e-06, "loss": 0.0063, "step": 25720 }, { "epoch": 2.7884865568083264, "grad_norm": 0.35954996943473816, "learning_rate": 1.415203829972436e-06, "loss": 0.0156, "step": 25721 }, { "epoch": 2.788594969644406, "grad_norm": 0.08914770931005478, "learning_rate": 1.4144784564050488e-06, "loss": 0.0003, "step": 25722 }, { "epoch": 2.7887033824804854, "grad_norm": 17.730632781982422, "learning_rate": 1.4137530828376613e-06, "loss": 0.0287, "step": 25723 }, { "epoch": 2.7888117953165654, "grad_norm": 1.1830313205718994, "learning_rate": 1.4130277092702743e-06, "loss": 0.0101, "step": 25724 }, { "epoch": 2.7889202081526454, "grad_norm": 9.83400821685791, "learning_rate": 1.412302335702887e-06, "loss": 0.0378, "step": 25725 }, { "epoch": 2.789028620988725, "grad_norm": 0.913827657699585, "learning_rate": 1.4115769621355e-06, "loss": 0.0415, "step": 25726 }, { "epoch": 2.789137033824805, "grad_norm": 0.011793190613389015, "learning_rate": 1.4108515885681127e-06, "loss": 0.0001, "step": 25727 }, { "epoch": 2.7892454466608845, "grad_norm": 2.5867106914520264, "learning_rate": 1.4101262150007255e-06, "loss": 0.0077, "step": 25728 }, { "epoch": 2.7893538594969645, "grad_norm": 0.427176296710968, "learning_rate": 1.4094008414333382e-06, "loss": 0.0131, "step": 25729 }, { "epoch": 2.7894622723330444, "grad_norm": 1.0663244724273682, "learning_rate": 1.4086754678659512e-06, "loss": 0.0022, "step": 25730 }, { "epoch": 2.789570685169124, "grad_norm": 0.05573081970214844, "learning_rate": 1.4079500942985638e-06, "loss": 0.0005, "step": 25731 }, { "epoch": 2.789679098005204, "grad_norm": 0.28159263730049133, "learning_rate": 1.4072247207311768e-06, "loss": 0.006, "step": 25732 }, { "epoch": 2.7897875108412835, "grad_norm": 0.055311188101768494, "learning_rate": 1.4064993471637893e-06, "loss": 0.0005, "step": 25733 }, { "epoch": 2.7898959236773635, "grad_norm": 0.01750335842370987, "learning_rate": 1.4057739735964023e-06, "loss": 0.0003, "step": 25734 }, { "epoch": 2.790004336513443, "grad_norm": 0.012555324472486973, "learning_rate": 1.4050486000290151e-06, "loss": 0.0002, "step": 25735 }, { "epoch": 2.790112749349523, "grad_norm": 0.5431213974952698, "learning_rate": 1.4043232264616277e-06, "loss": 0.0152, "step": 25736 }, { "epoch": 2.7902211621856026, "grad_norm": 0.1797875016927719, "learning_rate": 1.4035978528942407e-06, "loss": 0.0019, "step": 25737 }, { "epoch": 2.7903295750216826, "grad_norm": 0.06394197046756744, "learning_rate": 1.4028724793268532e-06, "loss": 0.0004, "step": 25738 }, { "epoch": 2.7904379878577625, "grad_norm": 0.5928369760513306, "learning_rate": 1.4021471057594662e-06, "loss": 0.0178, "step": 25739 }, { "epoch": 2.790546400693842, "grad_norm": 1.6805810928344727, "learning_rate": 1.401421732192079e-06, "loss": 0.0257, "step": 25740 }, { "epoch": 2.790654813529922, "grad_norm": 4.301839828491211, "learning_rate": 1.4006963586246918e-06, "loss": 0.0382, "step": 25741 }, { "epoch": 2.7907632263660016, "grad_norm": 1.0261037349700928, "learning_rate": 1.3999709850573046e-06, "loss": 0.0175, "step": 25742 }, { "epoch": 2.7908716392020816, "grad_norm": 0.26907333731651306, "learning_rate": 1.3992456114899176e-06, "loss": 0.0055, "step": 25743 }, { "epoch": 2.7909800520381616, "grad_norm": 7.232100963592529, "learning_rate": 1.3985202379225301e-06, "loss": 0.0137, "step": 25744 }, { "epoch": 2.791088464874241, "grad_norm": 4.232330799102783, "learning_rate": 1.3977948643551431e-06, "loss": 0.0021, "step": 25745 }, { "epoch": 2.7911968777103207, "grad_norm": 0.42115917801856995, "learning_rate": 1.3970694907877557e-06, "loss": 0.0129, "step": 25746 }, { "epoch": 2.7913052905464006, "grad_norm": 2.1304121017456055, "learning_rate": 1.3963441172203687e-06, "loss": 0.043, "step": 25747 }, { "epoch": 2.7914137033824806, "grad_norm": 0.014331316575407982, "learning_rate": 1.3956187436529815e-06, "loss": 0.0001, "step": 25748 }, { "epoch": 2.79152211621856, "grad_norm": 0.3159028887748718, "learning_rate": 1.3948933700855942e-06, "loss": 0.0097, "step": 25749 }, { "epoch": 2.79163052905464, "grad_norm": 0.920872688293457, "learning_rate": 1.394167996518207e-06, "loss": 0.0207, "step": 25750 }, { "epoch": 2.7917389418907197, "grad_norm": 1.414328694343567, "learning_rate": 1.3934426229508196e-06, "loss": 0.0115, "step": 25751 }, { "epoch": 2.7918473547267997, "grad_norm": 2.9441025257110596, "learning_rate": 1.3927172493834326e-06, "loss": 0.0133, "step": 25752 }, { "epoch": 2.7919557675628797, "grad_norm": 0.4731731414794922, "learning_rate": 1.3919918758160453e-06, "loss": 0.0123, "step": 25753 }, { "epoch": 2.792064180398959, "grad_norm": 0.0995238795876503, "learning_rate": 1.3912665022486581e-06, "loss": 0.0004, "step": 25754 }, { "epoch": 2.792172593235039, "grad_norm": 0.15808022022247314, "learning_rate": 1.390541128681271e-06, "loss": 0.0008, "step": 25755 }, { "epoch": 2.7922810060711187, "grad_norm": 0.636561930179596, "learning_rate": 1.3898157551138839e-06, "loss": 0.0115, "step": 25756 }, { "epoch": 2.7923894189071987, "grad_norm": 0.3589608371257782, "learning_rate": 1.3890903815464965e-06, "loss": 0.0119, "step": 25757 }, { "epoch": 2.7924978317432783, "grad_norm": 0.513865053653717, "learning_rate": 1.3883650079791094e-06, "loss": 0.0149, "step": 25758 }, { "epoch": 2.7926062445793582, "grad_norm": 0.28712552785873413, "learning_rate": 1.387639634411722e-06, "loss": 0.0055, "step": 25759 }, { "epoch": 2.792714657415438, "grad_norm": 3.5311226844787598, "learning_rate": 1.386914260844335e-06, "loss": 0.0403, "step": 25760 }, { "epoch": 2.7928230702515178, "grad_norm": 0.03359971195459366, "learning_rate": 1.3861888872769478e-06, "loss": 0.0003, "step": 25761 }, { "epoch": 2.7929314830875978, "grad_norm": 0.4299006164073944, "learning_rate": 1.3854635137095608e-06, "loss": 0.0028, "step": 25762 }, { "epoch": 2.7930398959236773, "grad_norm": 1.811159372329712, "learning_rate": 1.3847381401421733e-06, "loss": 0.06, "step": 25763 }, { "epoch": 2.7931483087597573, "grad_norm": 0.6471660733222961, "learning_rate": 1.384012766574786e-06, "loss": 0.007, "step": 25764 }, { "epoch": 2.793256721595837, "grad_norm": 0.050052110105752945, "learning_rate": 1.383287393007399e-06, "loss": 0.0005, "step": 25765 }, { "epoch": 2.793365134431917, "grad_norm": 0.7424392700195312, "learning_rate": 1.3825620194400117e-06, "loss": 0.023, "step": 25766 }, { "epoch": 2.793473547267997, "grad_norm": 3.1639134883880615, "learning_rate": 1.3818366458726245e-06, "loss": 0.0033, "step": 25767 }, { "epoch": 2.7935819601040763, "grad_norm": 0.31017738580703735, "learning_rate": 1.3811112723052372e-06, "loss": 0.0102, "step": 25768 }, { "epoch": 2.793690372940156, "grad_norm": 0.14335393905639648, "learning_rate": 1.3803858987378502e-06, "loss": 0.0003, "step": 25769 }, { "epoch": 2.793798785776236, "grad_norm": 0.41863560676574707, "learning_rate": 1.3796605251704628e-06, "loss": 0.0227, "step": 25770 }, { "epoch": 2.793907198612316, "grad_norm": 0.09088950604200363, "learning_rate": 1.3789351516030758e-06, "loss": 0.0003, "step": 25771 }, { "epoch": 2.7940156114483954, "grad_norm": 3.0027451515197754, "learning_rate": 1.3782097780356883e-06, "loss": 0.0131, "step": 25772 }, { "epoch": 2.7941240242844754, "grad_norm": 0.5085262656211853, "learning_rate": 1.3774844044683013e-06, "loss": 0.0077, "step": 25773 }, { "epoch": 2.794232437120555, "grad_norm": 0.017859064042568207, "learning_rate": 1.3767590309009141e-06, "loss": 0.0001, "step": 25774 }, { "epoch": 2.794340849956635, "grad_norm": 0.5049726963043213, "learning_rate": 1.3760336573335271e-06, "loss": 0.0057, "step": 25775 }, { "epoch": 2.794449262792715, "grad_norm": 0.6264957785606384, "learning_rate": 1.3753082837661397e-06, "loss": 0.0118, "step": 25776 }, { "epoch": 2.7945576756287944, "grad_norm": 0.07841300219297409, "learning_rate": 1.3745829101987522e-06, "loss": 0.0005, "step": 25777 }, { "epoch": 2.794666088464874, "grad_norm": 0.4537389874458313, "learning_rate": 1.3738575366313652e-06, "loss": 0.0016, "step": 25778 }, { "epoch": 2.794774501300954, "grad_norm": 6.661357402801514, "learning_rate": 1.373132163063978e-06, "loss": 0.0209, "step": 25779 }, { "epoch": 2.794882914137034, "grad_norm": 0.5000702142715454, "learning_rate": 1.3724067894965908e-06, "loss": 0.0111, "step": 25780 }, { "epoch": 2.7949913269731135, "grad_norm": 1.357345700263977, "learning_rate": 1.3716814159292036e-06, "loss": 0.0021, "step": 25781 }, { "epoch": 2.7950997398091935, "grad_norm": 3.173768997192383, "learning_rate": 1.3709560423618166e-06, "loss": 0.0136, "step": 25782 }, { "epoch": 2.795208152645273, "grad_norm": 0.03323764726519585, "learning_rate": 1.3702306687944291e-06, "loss": 0.0003, "step": 25783 }, { "epoch": 2.795316565481353, "grad_norm": 0.686489462852478, "learning_rate": 1.3695052952270421e-06, "loss": 0.0145, "step": 25784 }, { "epoch": 2.795424978317433, "grad_norm": 3.969329833984375, "learning_rate": 1.3687799216596547e-06, "loss": 0.0277, "step": 25785 }, { "epoch": 2.7955333911535125, "grad_norm": 0.015801258385181427, "learning_rate": 1.3680545480922677e-06, "loss": 0.0001, "step": 25786 }, { "epoch": 2.7956418039895925, "grad_norm": 0.49095290899276733, "learning_rate": 1.3673291745248805e-06, "loss": 0.0145, "step": 25787 }, { "epoch": 2.795750216825672, "grad_norm": 1.0891317129135132, "learning_rate": 1.3666038009574934e-06, "loss": 0.0344, "step": 25788 }, { "epoch": 2.795858629661752, "grad_norm": 0.031305622309446335, "learning_rate": 1.365878427390106e-06, "loss": 0.0001, "step": 25789 }, { "epoch": 2.795967042497832, "grad_norm": 5.047876358032227, "learning_rate": 1.3651530538227186e-06, "loss": 0.0276, "step": 25790 }, { "epoch": 2.7960754553339116, "grad_norm": 2.9083662033081055, "learning_rate": 1.3644276802553316e-06, "loss": 0.0253, "step": 25791 }, { "epoch": 2.796183868169991, "grad_norm": 1.8514668941497803, "learning_rate": 1.3637023066879443e-06, "loss": 0.0204, "step": 25792 }, { "epoch": 2.796292281006071, "grad_norm": 0.05066680908203125, "learning_rate": 1.3629769331205573e-06, "loss": 0.0006, "step": 25793 }, { "epoch": 2.796400693842151, "grad_norm": 0.6319814324378967, "learning_rate": 1.36225155955317e-06, "loss": 0.0284, "step": 25794 }, { "epoch": 2.7965091066782306, "grad_norm": 3.7209818363189697, "learning_rate": 1.361526185985783e-06, "loss": 0.0308, "step": 25795 }, { "epoch": 2.7966175195143106, "grad_norm": 0.05850348249077797, "learning_rate": 1.3608008124183955e-06, "loss": 0.0004, "step": 25796 }, { "epoch": 2.79672593235039, "grad_norm": 0.40339726209640503, "learning_rate": 1.3600754388510085e-06, "loss": 0.0104, "step": 25797 }, { "epoch": 2.79683434518647, "grad_norm": 17.283775329589844, "learning_rate": 1.359350065283621e-06, "loss": 0.0315, "step": 25798 }, { "epoch": 2.79694275802255, "grad_norm": 0.444508820772171, "learning_rate": 1.358624691716234e-06, "loss": 0.0077, "step": 25799 }, { "epoch": 2.7970511708586296, "grad_norm": 0.058334339410066605, "learning_rate": 1.3578993181488468e-06, "loss": 0.0004, "step": 25800 }, { "epoch": 2.797159583694709, "grad_norm": 0.08055780082941055, "learning_rate": 1.3571739445814598e-06, "loss": 0.0002, "step": 25801 }, { "epoch": 2.797267996530789, "grad_norm": 0.30602675676345825, "learning_rate": 1.3564485710140723e-06, "loss": 0.0065, "step": 25802 }, { "epoch": 2.797376409366869, "grad_norm": 0.3362657129764557, "learning_rate": 1.3557231974466853e-06, "loss": 0.0009, "step": 25803 }, { "epoch": 2.7974848222029487, "grad_norm": 6.010425567626953, "learning_rate": 1.354997823879298e-06, "loss": 0.0086, "step": 25804 }, { "epoch": 2.7975932350390287, "grad_norm": 1.1398450136184692, "learning_rate": 1.3542724503119107e-06, "loss": 0.0467, "step": 25805 }, { "epoch": 2.797701647875108, "grad_norm": 0.24108344316482544, "learning_rate": 1.3535470767445237e-06, "loss": 0.0008, "step": 25806 }, { "epoch": 2.797810060711188, "grad_norm": 1.0025683641433716, "learning_rate": 1.3528217031771362e-06, "loss": 0.0019, "step": 25807 }, { "epoch": 2.797918473547268, "grad_norm": 0.052329499274492264, "learning_rate": 1.3520963296097492e-06, "loss": 0.0004, "step": 25808 }, { "epoch": 2.7980268863833477, "grad_norm": 0.4143204092979431, "learning_rate": 1.3513709560423618e-06, "loss": 0.0075, "step": 25809 }, { "epoch": 2.7981352992194277, "grad_norm": 0.3157113790512085, "learning_rate": 1.3506455824749748e-06, "loss": 0.0114, "step": 25810 }, { "epoch": 2.7982437120555073, "grad_norm": 1.5288431644439697, "learning_rate": 1.3499202089075876e-06, "loss": 0.0131, "step": 25811 }, { "epoch": 2.7983521248915872, "grad_norm": 0.2775511145591736, "learning_rate": 1.3491948353402003e-06, "loss": 0.006, "step": 25812 }, { "epoch": 2.7984605377276672, "grad_norm": 0.40224489569664, "learning_rate": 1.3484694617728131e-06, "loss": 0.0026, "step": 25813 }, { "epoch": 2.7985689505637468, "grad_norm": 4.768337726593018, "learning_rate": 1.3477440882054261e-06, "loss": 0.02, "step": 25814 }, { "epoch": 2.7986773633998263, "grad_norm": 0.07568053156137466, "learning_rate": 1.3470187146380387e-06, "loss": 0.0004, "step": 25815 }, { "epoch": 2.7987857762359063, "grad_norm": 7.567126750946045, "learning_rate": 1.3462933410706517e-06, "loss": 0.0248, "step": 25816 }, { "epoch": 2.7988941890719863, "grad_norm": 0.4105978012084961, "learning_rate": 1.3455679675032642e-06, "loss": 0.0142, "step": 25817 }, { "epoch": 2.799002601908066, "grad_norm": 2.6355783939361572, "learning_rate": 1.344842593935877e-06, "loss": 0.0192, "step": 25818 }, { "epoch": 2.799111014744146, "grad_norm": 0.6062750220298767, "learning_rate": 1.34411722036849e-06, "loss": 0.0113, "step": 25819 }, { "epoch": 2.7992194275802254, "grad_norm": 0.46577706933021545, "learning_rate": 1.3433918468011026e-06, "loss": 0.0243, "step": 25820 }, { "epoch": 2.7993278404163053, "grad_norm": 0.26246127486228943, "learning_rate": 1.3426664732337156e-06, "loss": 0.005, "step": 25821 }, { "epoch": 2.7994362532523853, "grad_norm": 1.7866532802581787, "learning_rate": 1.3419410996663281e-06, "loss": 0.0547, "step": 25822 }, { "epoch": 2.799544666088465, "grad_norm": 0.6437374949455261, "learning_rate": 1.3412157260989411e-06, "loss": 0.027, "step": 25823 }, { "epoch": 2.7996530789245444, "grad_norm": 3.0924630165100098, "learning_rate": 1.340490352531554e-06, "loss": 0.027, "step": 25824 }, { "epoch": 2.7997614917606244, "grad_norm": 0.14679889380931854, "learning_rate": 1.3397649789641667e-06, "loss": 0.0006, "step": 25825 }, { "epoch": 2.7998699045967044, "grad_norm": 1.349509596824646, "learning_rate": 1.3390396053967795e-06, "loss": 0.0506, "step": 25826 }, { "epoch": 2.799978317432784, "grad_norm": 1.2536184787750244, "learning_rate": 1.3383142318293924e-06, "loss": 0.0365, "step": 25827 }, { "epoch": 2.800086730268864, "grad_norm": 0.034185152500867844, "learning_rate": 1.337588858262005e-06, "loss": 0.0002, "step": 25828 }, { "epoch": 2.8001951431049434, "grad_norm": 0.6770361661911011, "learning_rate": 1.336863484694618e-06, "loss": 0.0191, "step": 25829 }, { "epoch": 2.8003035559410234, "grad_norm": 1.4877384901046753, "learning_rate": 1.3361381111272306e-06, "loss": 0.0095, "step": 25830 }, { "epoch": 2.8004119687771034, "grad_norm": 0.4500296413898468, "learning_rate": 1.3354127375598433e-06, "loss": 0.0078, "step": 25831 }, { "epoch": 2.800520381613183, "grad_norm": 0.29523104429244995, "learning_rate": 1.3346873639924563e-06, "loss": 0.0058, "step": 25832 }, { "epoch": 2.800628794449263, "grad_norm": 0.34886759519577026, "learning_rate": 1.333961990425069e-06, "loss": 0.0041, "step": 25833 }, { "epoch": 2.8007372072853425, "grad_norm": 0.30329927802085876, "learning_rate": 1.333236616857682e-06, "loss": 0.0058, "step": 25834 }, { "epoch": 2.8008456201214225, "grad_norm": 0.039084434509277344, "learning_rate": 1.3325112432902945e-06, "loss": 0.0002, "step": 25835 }, { "epoch": 2.800954032957502, "grad_norm": 0.016286935657262802, "learning_rate": 1.3317858697229075e-06, "loss": 0.0001, "step": 25836 }, { "epoch": 2.801062445793582, "grad_norm": 0.027128741145133972, "learning_rate": 1.3310604961555202e-06, "loss": 0.0002, "step": 25837 }, { "epoch": 2.8011708586296615, "grad_norm": 9.344284057617188, "learning_rate": 1.330335122588133e-06, "loss": 0.0063, "step": 25838 }, { "epoch": 2.8012792714657415, "grad_norm": 0.031716324388980865, "learning_rate": 1.3296097490207458e-06, "loss": 0.0002, "step": 25839 }, { "epoch": 2.8013876843018215, "grad_norm": 0.04297155886888504, "learning_rate": 1.3288843754533588e-06, "loss": 0.0003, "step": 25840 }, { "epoch": 2.801496097137901, "grad_norm": 0.0207809004932642, "learning_rate": 1.3281590018859713e-06, "loss": 0.0001, "step": 25841 }, { "epoch": 2.801604509973981, "grad_norm": 3.889247179031372, "learning_rate": 1.3274336283185843e-06, "loss": 0.0244, "step": 25842 }, { "epoch": 2.8017129228100606, "grad_norm": 0.9530799984931946, "learning_rate": 1.326708254751197e-06, "loss": 0.0056, "step": 25843 }, { "epoch": 2.8018213356461406, "grad_norm": 0.23939158022403717, "learning_rate": 1.3259828811838097e-06, "loss": 0.0037, "step": 25844 }, { "epoch": 2.8019297484822205, "grad_norm": 2.330441951751709, "learning_rate": 1.3252575076164227e-06, "loss": 0.0269, "step": 25845 }, { "epoch": 2.8020381613183, "grad_norm": 4.1052422523498535, "learning_rate": 1.3245321340490352e-06, "loss": 0.0238, "step": 25846 }, { "epoch": 2.8021465741543796, "grad_norm": 0.2858547568321228, "learning_rate": 1.3238067604816482e-06, "loss": 0.0057, "step": 25847 }, { "epoch": 2.8022549869904596, "grad_norm": 0.3603481948375702, "learning_rate": 1.3230813869142608e-06, "loss": 0.0086, "step": 25848 }, { "epoch": 2.8023633998265396, "grad_norm": 0.5549299716949463, "learning_rate": 1.3223560133468738e-06, "loss": 0.0166, "step": 25849 }, { "epoch": 2.802471812662619, "grad_norm": 0.43792203068733215, "learning_rate": 1.3216306397794866e-06, "loss": 0.0011, "step": 25850 }, { "epoch": 2.802580225498699, "grad_norm": 0.7413634061813354, "learning_rate": 1.3209052662120993e-06, "loss": 0.0063, "step": 25851 }, { "epoch": 2.8026886383347787, "grad_norm": 0.00871249195188284, "learning_rate": 1.3201798926447121e-06, "loss": 0.0001, "step": 25852 }, { "epoch": 2.8027970511708586, "grad_norm": 0.2995991110801697, "learning_rate": 1.3194545190773251e-06, "loss": 0.0058, "step": 25853 }, { "epoch": 2.8029054640069386, "grad_norm": 0.8971850275993347, "learning_rate": 1.3187291455099377e-06, "loss": 0.0067, "step": 25854 }, { "epoch": 2.803013876843018, "grad_norm": 0.28868019580841064, "learning_rate": 1.3180037719425507e-06, "loss": 0.0061, "step": 25855 }, { "epoch": 2.803122289679098, "grad_norm": 0.4608621597290039, "learning_rate": 1.3172783983751632e-06, "loss": 0.0158, "step": 25856 }, { "epoch": 2.8032307025151777, "grad_norm": 3.547639846801758, "learning_rate": 1.3165530248077762e-06, "loss": 0.0415, "step": 25857 }, { "epoch": 2.8033391153512577, "grad_norm": 8.54391860961914, "learning_rate": 1.315827651240389e-06, "loss": 0.0682, "step": 25858 }, { "epoch": 2.8034475281873372, "grad_norm": 0.7460916042327881, "learning_rate": 1.3151022776730016e-06, "loss": 0.0058, "step": 25859 }, { "epoch": 2.803555941023417, "grad_norm": 6.024717807769775, "learning_rate": 1.3143769041056146e-06, "loss": 0.0063, "step": 25860 }, { "epoch": 2.8036643538594967, "grad_norm": 0.5472590327262878, "learning_rate": 1.3136515305382271e-06, "loss": 0.0196, "step": 25861 }, { "epoch": 2.8037727666955767, "grad_norm": 5.501288414001465, "learning_rate": 1.3129261569708401e-06, "loss": 0.0083, "step": 25862 }, { "epoch": 2.8038811795316567, "grad_norm": 0.8781877756118774, "learning_rate": 1.312200783403453e-06, "loss": 0.016, "step": 25863 }, { "epoch": 2.8039895923677363, "grad_norm": 2.64680814743042, "learning_rate": 1.3114754098360657e-06, "loss": 0.0059, "step": 25864 }, { "epoch": 2.8040980052038162, "grad_norm": 0.5405292510986328, "learning_rate": 1.3107500362686785e-06, "loss": 0.0171, "step": 25865 }, { "epoch": 2.804206418039896, "grad_norm": 0.9608420729637146, "learning_rate": 1.3100246627012915e-06, "loss": 0.0018, "step": 25866 }, { "epoch": 2.8043148308759758, "grad_norm": 0.325714111328125, "learning_rate": 1.309299289133904e-06, "loss": 0.0066, "step": 25867 }, { "epoch": 2.8044232437120558, "grad_norm": 3.0906317234039307, "learning_rate": 1.308573915566517e-06, "loss": 0.0067, "step": 25868 }, { "epoch": 2.8045316565481353, "grad_norm": 0.020596804097294807, "learning_rate": 1.3078485419991296e-06, "loss": 0.0001, "step": 25869 }, { "epoch": 2.804640069384215, "grad_norm": 0.006639882922172546, "learning_rate": 1.3071231684317426e-06, "loss": 0.0001, "step": 25870 }, { "epoch": 2.804748482220295, "grad_norm": 0.03623807057738304, "learning_rate": 1.3063977948643553e-06, "loss": 0.0002, "step": 25871 }, { "epoch": 2.804856895056375, "grad_norm": 0.019005924463272095, "learning_rate": 1.305672421296968e-06, "loss": 0.0002, "step": 25872 }, { "epoch": 2.8049653078924544, "grad_norm": 2.5026967525482178, "learning_rate": 1.304947047729581e-06, "loss": 0.0142, "step": 25873 }, { "epoch": 2.8050737207285343, "grad_norm": 0.5700302720069885, "learning_rate": 1.3042216741621935e-06, "loss": 0.0121, "step": 25874 }, { "epoch": 2.805182133564614, "grad_norm": 3.789088487625122, "learning_rate": 1.3034963005948065e-06, "loss": 0.0317, "step": 25875 }, { "epoch": 2.805290546400694, "grad_norm": 0.13590124249458313, "learning_rate": 1.3027709270274192e-06, "loss": 0.0007, "step": 25876 }, { "epoch": 2.805398959236774, "grad_norm": 0.8313271999359131, "learning_rate": 1.302045553460032e-06, "loss": 0.0127, "step": 25877 }, { "epoch": 2.8055073720728534, "grad_norm": 0.00657907547429204, "learning_rate": 1.3013201798926448e-06, "loss": 0.0001, "step": 25878 }, { "epoch": 2.8056157849089334, "grad_norm": 1.3784857988357544, "learning_rate": 1.3005948063252578e-06, "loss": 0.0351, "step": 25879 }, { "epoch": 2.805724197745013, "grad_norm": 0.5440210700035095, "learning_rate": 1.2998694327578704e-06, "loss": 0.0147, "step": 25880 }, { "epoch": 2.805832610581093, "grad_norm": 0.484643816947937, "learning_rate": 1.2991440591904833e-06, "loss": 0.0064, "step": 25881 }, { "epoch": 2.8059410234171724, "grad_norm": 0.9813010692596436, "learning_rate": 1.298418685623096e-06, "loss": 0.0082, "step": 25882 }, { "epoch": 2.8060494362532524, "grad_norm": 0.3684186637401581, "learning_rate": 1.297693312055709e-06, "loss": 0.0121, "step": 25883 }, { "epoch": 2.806157849089332, "grad_norm": 0.0017040565144270658, "learning_rate": 1.2969679384883217e-06, "loss": 0.0, "step": 25884 }, { "epoch": 2.806266261925412, "grad_norm": 0.32677891850471497, "learning_rate": 1.2962425649209342e-06, "loss": 0.0115, "step": 25885 }, { "epoch": 2.806374674761492, "grad_norm": 0.06066551059484482, "learning_rate": 1.2955171913535472e-06, "loss": 0.0004, "step": 25886 }, { "epoch": 2.8064830875975715, "grad_norm": 0.021074878051877022, "learning_rate": 1.2947918177861598e-06, "loss": 0.0001, "step": 25887 }, { "epoch": 2.8065915004336515, "grad_norm": 2.2772319316864014, "learning_rate": 1.2940664442187728e-06, "loss": 0.0107, "step": 25888 }, { "epoch": 2.806699913269731, "grad_norm": 0.03484388813376427, "learning_rate": 1.2933410706513856e-06, "loss": 0.0002, "step": 25889 }, { "epoch": 2.806808326105811, "grad_norm": 0.5253612399101257, "learning_rate": 1.2926156970839983e-06, "loss": 0.0112, "step": 25890 }, { "epoch": 2.806916738941891, "grad_norm": 0.8157866597175598, "learning_rate": 1.2918903235166111e-06, "loss": 0.0155, "step": 25891 }, { "epoch": 2.8070251517779705, "grad_norm": 0.4569360613822937, "learning_rate": 1.2911649499492241e-06, "loss": 0.0067, "step": 25892 }, { "epoch": 2.80713356461405, "grad_norm": 0.01077092532068491, "learning_rate": 1.2904395763818367e-06, "loss": 0.0001, "step": 25893 }, { "epoch": 2.80724197745013, "grad_norm": 0.16409356892108917, "learning_rate": 1.2897142028144497e-06, "loss": 0.0014, "step": 25894 }, { "epoch": 2.80735039028621, "grad_norm": 0.7527342438697815, "learning_rate": 1.2889888292470622e-06, "loss": 0.0038, "step": 25895 }, { "epoch": 2.8074588031222896, "grad_norm": 2.6911399364471436, "learning_rate": 1.2882634556796752e-06, "loss": 0.0384, "step": 25896 }, { "epoch": 2.8075672159583696, "grad_norm": 0.9303727746009827, "learning_rate": 1.287538082112288e-06, "loss": 0.0009, "step": 25897 }, { "epoch": 2.807675628794449, "grad_norm": 0.5784949064254761, "learning_rate": 1.2868127085449006e-06, "loss": 0.012, "step": 25898 }, { "epoch": 2.807784041630529, "grad_norm": 0.15563298761844635, "learning_rate": 1.2860873349775136e-06, "loss": 0.0004, "step": 25899 }, { "epoch": 2.807892454466609, "grad_norm": 0.33907550573349, "learning_rate": 1.2853619614101261e-06, "loss": 0.0084, "step": 25900 }, { "epoch": 2.8080008673026886, "grad_norm": 3.626452922821045, "learning_rate": 1.2846365878427391e-06, "loss": 0.01, "step": 25901 }, { "epoch": 2.8081092801387686, "grad_norm": 0.35079675912857056, "learning_rate": 1.283911214275352e-06, "loss": 0.001, "step": 25902 }, { "epoch": 2.808217692974848, "grad_norm": 0.03198689967393875, "learning_rate": 1.2831858407079647e-06, "loss": 0.0003, "step": 25903 }, { "epoch": 2.808326105810928, "grad_norm": 0.2811620235443115, "learning_rate": 1.2824604671405775e-06, "loss": 0.0055, "step": 25904 }, { "epoch": 2.8084345186470077, "grad_norm": 0.03915004804730415, "learning_rate": 1.2817350935731905e-06, "loss": 0.0003, "step": 25905 }, { "epoch": 2.8085429314830876, "grad_norm": 3.2612407207489014, "learning_rate": 1.281009720005803e-06, "loss": 0.0037, "step": 25906 }, { "epoch": 2.808651344319167, "grad_norm": 0.06525731086730957, "learning_rate": 1.280284346438416e-06, "loss": 0.0005, "step": 25907 }, { "epoch": 2.808759757155247, "grad_norm": 0.10634011775255203, "learning_rate": 1.2795589728710286e-06, "loss": 0.0108, "step": 25908 }, { "epoch": 2.808868169991327, "grad_norm": 0.563805103302002, "learning_rate": 1.2788335993036416e-06, "loss": 0.0131, "step": 25909 }, { "epoch": 2.8089765828274067, "grad_norm": 2.420820713043213, "learning_rate": 1.2781082257362543e-06, "loss": 0.0334, "step": 25910 }, { "epoch": 2.8090849956634867, "grad_norm": 0.6050880551338196, "learning_rate": 1.2773828521688671e-06, "loss": 0.013, "step": 25911 }, { "epoch": 2.8091934084995662, "grad_norm": 0.31050410866737366, "learning_rate": 1.27665747860148e-06, "loss": 0.0017, "step": 25912 }, { "epoch": 2.809301821335646, "grad_norm": 0.36171194911003113, "learning_rate": 1.2759321050340925e-06, "loss": 0.006, "step": 25913 }, { "epoch": 2.809410234171726, "grad_norm": 0.45041659474372864, "learning_rate": 1.2752067314667055e-06, "loss": 0.0147, "step": 25914 }, { "epoch": 2.8095186470078057, "grad_norm": 0.4094514548778534, "learning_rate": 1.2744813578993182e-06, "loss": 0.0151, "step": 25915 }, { "epoch": 2.8096270598438853, "grad_norm": 0.16196177899837494, "learning_rate": 1.273755984331931e-06, "loss": 0.0012, "step": 25916 }, { "epoch": 2.8097354726799653, "grad_norm": 1.8357288837432861, "learning_rate": 1.2730306107645438e-06, "loss": 0.0548, "step": 25917 }, { "epoch": 2.8098438855160452, "grad_norm": 0.5069132447242737, "learning_rate": 1.2723052371971568e-06, "loss": 0.0048, "step": 25918 }, { "epoch": 2.809952298352125, "grad_norm": 0.03862934187054634, "learning_rate": 1.2715798636297694e-06, "loss": 0.0002, "step": 25919 }, { "epoch": 2.8100607111882048, "grad_norm": 0.5140506625175476, "learning_rate": 1.2708544900623823e-06, "loss": 0.0179, "step": 25920 }, { "epoch": 2.8101691240242843, "grad_norm": 0.6302440762519836, "learning_rate": 1.270129116494995e-06, "loss": 0.0212, "step": 25921 }, { "epoch": 2.8102775368603643, "grad_norm": 0.6596958041191101, "learning_rate": 1.269403742927608e-06, "loss": 0.0007, "step": 25922 }, { "epoch": 2.8103859496964443, "grad_norm": 0.4105204641819, "learning_rate": 1.2686783693602207e-06, "loss": 0.0052, "step": 25923 }, { "epoch": 2.810494362532524, "grad_norm": 0.44317182898521423, "learning_rate": 1.2679529957928335e-06, "loss": 0.0009, "step": 25924 }, { "epoch": 2.8106027753686034, "grad_norm": 0.16957621276378632, "learning_rate": 1.2672276222254462e-06, "loss": 0.0002, "step": 25925 }, { "epoch": 2.8107111882046834, "grad_norm": 0.2683674097061157, "learning_rate": 1.2665022486580588e-06, "loss": 0.0025, "step": 25926 }, { "epoch": 2.8108196010407633, "grad_norm": 0.3230401575565338, "learning_rate": 1.2657768750906718e-06, "loss": 0.0107, "step": 25927 }, { "epoch": 2.810928013876843, "grad_norm": 0.022459786385297775, "learning_rate": 1.2650515015232846e-06, "loss": 0.0002, "step": 25928 }, { "epoch": 2.811036426712923, "grad_norm": 0.0199188943952322, "learning_rate": 1.2643261279558974e-06, "loss": 0.0002, "step": 25929 }, { "epoch": 2.8111448395490024, "grad_norm": 0.45827382802963257, "learning_rate": 1.2636007543885101e-06, "loss": 0.0168, "step": 25930 }, { "epoch": 2.8112532523850824, "grad_norm": 0.27828773856163025, "learning_rate": 1.2628753808211231e-06, "loss": 0.006, "step": 25931 }, { "epoch": 2.8113616652211624, "grad_norm": 0.6189507842063904, "learning_rate": 1.2621500072537357e-06, "loss": 0.001, "step": 25932 }, { "epoch": 2.811470078057242, "grad_norm": 2.2090723514556885, "learning_rate": 1.2614246336863487e-06, "loss": 0.0392, "step": 25933 }, { "epoch": 2.811578490893322, "grad_norm": 0.8768791556358337, "learning_rate": 1.2606992601189612e-06, "loss": 0.0008, "step": 25934 }, { "epoch": 2.8116869037294014, "grad_norm": 0.30547717213630676, "learning_rate": 1.2599738865515742e-06, "loss": 0.0051, "step": 25935 }, { "epoch": 2.8117953165654814, "grad_norm": 0.019592581316828728, "learning_rate": 1.259248512984187e-06, "loss": 0.0002, "step": 25936 }, { "epoch": 2.8119037294015614, "grad_norm": 0.49755486845970154, "learning_rate": 1.2585231394167998e-06, "loss": 0.0091, "step": 25937 }, { "epoch": 2.812012142237641, "grad_norm": 0.6043359637260437, "learning_rate": 1.2577977658494126e-06, "loss": 0.0161, "step": 25938 }, { "epoch": 2.8121205550737205, "grad_norm": 0.3229300379753113, "learning_rate": 1.2570723922820251e-06, "loss": 0.0011, "step": 25939 }, { "epoch": 2.8122289679098005, "grad_norm": 0.2818516492843628, "learning_rate": 1.2563470187146381e-06, "loss": 0.0058, "step": 25940 }, { "epoch": 2.8123373807458805, "grad_norm": 0.13292953372001648, "learning_rate": 1.255621645147251e-06, "loss": 0.0008, "step": 25941 }, { "epoch": 2.81244579358196, "grad_norm": 1.0531418323516846, "learning_rate": 1.2548962715798637e-06, "loss": 0.0142, "step": 25942 }, { "epoch": 2.81255420641804, "grad_norm": 0.7720537781715393, "learning_rate": 1.2541708980124765e-06, "loss": 0.0133, "step": 25943 }, { "epoch": 2.8126626192541195, "grad_norm": 0.15896141529083252, "learning_rate": 1.2534455244450895e-06, "loss": 0.0004, "step": 25944 }, { "epoch": 2.8127710320901995, "grad_norm": 12.860953330993652, "learning_rate": 1.252720150877702e-06, "loss": 0.0399, "step": 25945 }, { "epoch": 2.8128794449262795, "grad_norm": 5.984926223754883, "learning_rate": 1.251994777310315e-06, "loss": 0.0499, "step": 25946 }, { "epoch": 2.812987857762359, "grad_norm": 0.4673023223876953, "learning_rate": 1.2512694037429276e-06, "loss": 0.0157, "step": 25947 }, { "epoch": 2.8130962705984386, "grad_norm": 6.030889987945557, "learning_rate": 1.2505440301755406e-06, "loss": 0.0347, "step": 25948 }, { "epoch": 2.8132046834345186, "grad_norm": 0.5283048152923584, "learning_rate": 1.2498186566081533e-06, "loss": 0.0144, "step": 25949 }, { "epoch": 2.8133130962705986, "grad_norm": 9.013643264770508, "learning_rate": 1.2490932830407661e-06, "loss": 0.0215, "step": 25950 }, { "epoch": 2.813421509106678, "grad_norm": 0.36152154207229614, "learning_rate": 1.248367909473379e-06, "loss": 0.0072, "step": 25951 }, { "epoch": 2.813529921942758, "grad_norm": 21.920194625854492, "learning_rate": 1.2476425359059917e-06, "loss": 0.0261, "step": 25952 }, { "epoch": 2.8136383347788376, "grad_norm": 0.293753445148468, "learning_rate": 1.2469171623386045e-06, "loss": 0.0062, "step": 25953 }, { "epoch": 2.8137467476149176, "grad_norm": 2.078784227371216, "learning_rate": 1.2461917887712172e-06, "loss": 0.0033, "step": 25954 }, { "epoch": 2.8138551604509976, "grad_norm": 0.01731720007956028, "learning_rate": 1.24546641520383e-06, "loss": 0.0001, "step": 25955 }, { "epoch": 2.813963573287077, "grad_norm": 0.6128981113433838, "learning_rate": 1.244741041636443e-06, "loss": 0.0034, "step": 25956 }, { "epoch": 2.814071986123157, "grad_norm": 1.027201771736145, "learning_rate": 1.2440156680690558e-06, "loss": 0.002, "step": 25957 }, { "epoch": 2.8141803989592367, "grad_norm": 4.999894142150879, "learning_rate": 1.2432902945016686e-06, "loss": 0.054, "step": 25958 }, { "epoch": 2.8142888117953166, "grad_norm": 0.8742612600326538, "learning_rate": 1.2425649209342811e-06, "loss": 0.0278, "step": 25959 }, { "epoch": 2.8143972246313966, "grad_norm": 0.42755022644996643, "learning_rate": 1.241839547366894e-06, "loss": 0.0136, "step": 25960 }, { "epoch": 2.814505637467476, "grad_norm": 2.722628355026245, "learning_rate": 1.2411141737995067e-06, "loss": 0.0287, "step": 25961 }, { "epoch": 2.8146140503035557, "grad_norm": 0.8993141055107117, "learning_rate": 1.2403888002321197e-06, "loss": 0.0225, "step": 25962 }, { "epoch": 2.8147224631396357, "grad_norm": 0.6865113377571106, "learning_rate": 1.2396634266647325e-06, "loss": 0.012, "step": 25963 }, { "epoch": 2.8148308759757157, "grad_norm": 0.1903982311487198, "learning_rate": 1.2389380530973452e-06, "loss": 0.0005, "step": 25964 }, { "epoch": 2.8149392888117952, "grad_norm": 0.04329701513051987, "learning_rate": 1.238212679529958e-06, "loss": 0.0003, "step": 25965 }, { "epoch": 2.815047701647875, "grad_norm": 1.2588145732879639, "learning_rate": 1.2374873059625708e-06, "loss": 0.0006, "step": 25966 }, { "epoch": 2.8151561144839548, "grad_norm": 5.18508243560791, "learning_rate": 1.2367619323951836e-06, "loss": 0.0235, "step": 25967 }, { "epoch": 2.8152645273200347, "grad_norm": 0.45766618847846985, "learning_rate": 1.2360365588277964e-06, "loss": 0.0166, "step": 25968 }, { "epoch": 2.8153729401561147, "grad_norm": 0.03130052611231804, "learning_rate": 1.2353111852604093e-06, "loss": 0.0002, "step": 25969 }, { "epoch": 2.8154813529921943, "grad_norm": 1.298875093460083, "learning_rate": 1.2345858116930221e-06, "loss": 0.0123, "step": 25970 }, { "epoch": 2.815589765828274, "grad_norm": 0.6575329303741455, "learning_rate": 1.233860438125635e-06, "loss": 0.011, "step": 25971 }, { "epoch": 2.815698178664354, "grad_norm": 0.2214212417602539, "learning_rate": 1.2331350645582477e-06, "loss": 0.0165, "step": 25972 }, { "epoch": 2.8158065915004338, "grad_norm": 0.008833940140902996, "learning_rate": 1.2324096909908602e-06, "loss": 0.0001, "step": 25973 }, { "epoch": 2.8159150043365133, "grad_norm": 0.1791965514421463, "learning_rate": 1.2316843174234732e-06, "loss": 0.0026, "step": 25974 }, { "epoch": 2.8160234171725933, "grad_norm": 0.003050739411264658, "learning_rate": 1.230958943856086e-06, "loss": 0.0, "step": 25975 }, { "epoch": 2.816131830008673, "grad_norm": 0.14233185350894928, "learning_rate": 1.2302335702886988e-06, "loss": 0.0003, "step": 25976 }, { "epoch": 2.816240242844753, "grad_norm": 0.3182896375656128, "learning_rate": 1.2295081967213116e-06, "loss": 0.0007, "step": 25977 }, { "epoch": 2.816348655680833, "grad_norm": 4.299213886260986, "learning_rate": 1.2287828231539244e-06, "loss": 0.0157, "step": 25978 }, { "epoch": 2.8164570685169124, "grad_norm": 0.3005332052707672, "learning_rate": 1.2280574495865371e-06, "loss": 0.0045, "step": 25979 }, { "epoch": 2.8165654813529923, "grad_norm": 0.19407157599925995, "learning_rate": 1.22733207601915e-06, "loss": 0.0029, "step": 25980 }, { "epoch": 2.816673894189072, "grad_norm": 6.0212602615356445, "learning_rate": 1.2266067024517627e-06, "loss": 0.054, "step": 25981 }, { "epoch": 2.816782307025152, "grad_norm": 0.02020614966750145, "learning_rate": 1.2258813288843757e-06, "loss": 0.0001, "step": 25982 }, { "epoch": 2.8168907198612314, "grad_norm": 0.3247832953929901, "learning_rate": 1.2251559553169885e-06, "loss": 0.0065, "step": 25983 }, { "epoch": 2.8169991326973114, "grad_norm": 0.007927480153739452, "learning_rate": 1.2244305817496012e-06, "loss": 0.0001, "step": 25984 }, { "epoch": 2.817107545533391, "grad_norm": 2.2770378589630127, "learning_rate": 1.223705208182214e-06, "loss": 0.0035, "step": 25985 }, { "epoch": 2.817215958369471, "grad_norm": 3.14780855178833, "learning_rate": 1.2229798346148266e-06, "loss": 0.1038, "step": 25986 }, { "epoch": 2.817324371205551, "grad_norm": 0.0319535955786705, "learning_rate": 1.2222544610474396e-06, "loss": 0.0002, "step": 25987 }, { "epoch": 2.8174327840416304, "grad_norm": 0.42938923835754395, "learning_rate": 1.2215290874800524e-06, "loss": 0.0083, "step": 25988 }, { "epoch": 2.8175411968777104, "grad_norm": 3.023149251937866, "learning_rate": 1.2208037139126651e-06, "loss": 0.0171, "step": 25989 }, { "epoch": 2.81764960971379, "grad_norm": 0.2914571166038513, "learning_rate": 1.220078340345278e-06, "loss": 0.0039, "step": 25990 }, { "epoch": 2.81775802254987, "grad_norm": 0.5637169480323792, "learning_rate": 1.2193529667778907e-06, "loss": 0.0068, "step": 25991 }, { "epoch": 2.81786643538595, "grad_norm": 0.7596518397331238, "learning_rate": 1.2186275932105035e-06, "loss": 0.0156, "step": 25992 }, { "epoch": 2.8179748482220295, "grad_norm": 2.6509013175964355, "learning_rate": 1.2179022196431162e-06, "loss": 0.0445, "step": 25993 }, { "epoch": 2.818083261058109, "grad_norm": 0.5359886288642883, "learning_rate": 1.217176846075729e-06, "loss": 0.0064, "step": 25994 }, { "epoch": 2.818191673894189, "grad_norm": 0.20978803932666779, "learning_rate": 1.216451472508342e-06, "loss": 0.0014, "step": 25995 }, { "epoch": 2.818300086730269, "grad_norm": 0.26748237013816833, "learning_rate": 1.2157260989409548e-06, "loss": 0.0003, "step": 25996 }, { "epoch": 2.8184084995663485, "grad_norm": 0.41739025712013245, "learning_rate": 1.2150007253735676e-06, "loss": 0.0089, "step": 25997 }, { "epoch": 2.8185169124024285, "grad_norm": 0.3052252233028412, "learning_rate": 1.2142753518061804e-06, "loss": 0.0065, "step": 25998 }, { "epoch": 2.818625325238508, "grad_norm": 3.397925615310669, "learning_rate": 1.2135499782387931e-06, "loss": 0.0064, "step": 25999 }, { "epoch": 2.818733738074588, "grad_norm": 11.85004711151123, "learning_rate": 1.212824604671406e-06, "loss": 0.0354, "step": 26000 }, { "epoch": 2.818842150910668, "grad_norm": 0.9147818684577942, "learning_rate": 1.2120992311040187e-06, "loss": 0.0141, "step": 26001 }, { "epoch": 2.8189505637467476, "grad_norm": 0.036821186542510986, "learning_rate": 1.2113738575366315e-06, "loss": 0.0003, "step": 26002 }, { "epoch": 2.8190589765828276, "grad_norm": 0.36269789934158325, "learning_rate": 1.2106484839692442e-06, "loss": 0.0125, "step": 26003 }, { "epoch": 2.819167389418907, "grad_norm": 4.500275611877441, "learning_rate": 1.209923110401857e-06, "loss": 0.0071, "step": 26004 }, { "epoch": 2.819275802254987, "grad_norm": 0.1262136846780777, "learning_rate": 1.2091977368344698e-06, "loss": 0.0014, "step": 26005 }, { "epoch": 2.8193842150910666, "grad_norm": 0.3011736571788788, "learning_rate": 1.2084723632670826e-06, "loss": 0.006, "step": 26006 }, { "epoch": 2.8194926279271466, "grad_norm": 0.1695507913827896, "learning_rate": 1.2077469896996954e-06, "loss": 0.0008, "step": 26007 }, { "epoch": 2.819601040763226, "grad_norm": 0.3350319564342499, "learning_rate": 1.2070216161323083e-06, "loss": 0.0065, "step": 26008 }, { "epoch": 2.819709453599306, "grad_norm": 5.943168640136719, "learning_rate": 1.2062962425649211e-06, "loss": 0.021, "step": 26009 }, { "epoch": 2.819817866435386, "grad_norm": 0.47237974405288696, "learning_rate": 1.205570868997534e-06, "loss": 0.0177, "step": 26010 }, { "epoch": 2.8199262792714657, "grad_norm": 1.742029070854187, "learning_rate": 1.2048454954301467e-06, "loss": 0.0303, "step": 26011 }, { "epoch": 2.8200346921075456, "grad_norm": 6.1039838790893555, "learning_rate": 1.2041201218627595e-06, "loss": 0.0325, "step": 26012 }, { "epoch": 2.820143104943625, "grad_norm": 1.073999285697937, "learning_rate": 1.2033947482953722e-06, "loss": 0.0171, "step": 26013 }, { "epoch": 2.820251517779705, "grad_norm": 0.5962185859680176, "learning_rate": 1.202669374727985e-06, "loss": 0.0087, "step": 26014 }, { "epoch": 2.820359930615785, "grad_norm": 0.6263849139213562, "learning_rate": 1.2019440011605978e-06, "loss": 0.0128, "step": 26015 }, { "epoch": 2.8204683434518647, "grad_norm": 32.66645431518555, "learning_rate": 1.2012186275932106e-06, "loss": 0.0229, "step": 26016 }, { "epoch": 2.8205767562879442, "grad_norm": 2.0432894229888916, "learning_rate": 1.2004932540258234e-06, "loss": 0.0025, "step": 26017 }, { "epoch": 2.8206851691240242, "grad_norm": 0.21592845022678375, "learning_rate": 1.1997678804584361e-06, "loss": 0.0026, "step": 26018 }, { "epoch": 2.820793581960104, "grad_norm": 0.06023794785141945, "learning_rate": 1.199042506891049e-06, "loss": 0.0005, "step": 26019 }, { "epoch": 2.8209019947961838, "grad_norm": 0.33058738708496094, "learning_rate": 1.1983171333236617e-06, "loss": 0.0124, "step": 26020 }, { "epoch": 2.8210104076322637, "grad_norm": 1.715242862701416, "learning_rate": 1.1975917597562747e-06, "loss": 0.0286, "step": 26021 }, { "epoch": 2.8211188204683433, "grad_norm": 2.168365240097046, "learning_rate": 1.1968663861888875e-06, "loss": 0.0033, "step": 26022 }, { "epoch": 2.8212272333044233, "grad_norm": 0.21035929024219513, "learning_rate": 1.1961410126215002e-06, "loss": 0.0006, "step": 26023 }, { "epoch": 2.8213356461405033, "grad_norm": 0.008866354823112488, "learning_rate": 1.195415639054113e-06, "loss": 0.0001, "step": 26024 }, { "epoch": 2.821444058976583, "grad_norm": 2.6270828247070312, "learning_rate": 1.1946902654867258e-06, "loss": 0.0318, "step": 26025 }, { "epoch": 2.8215524718126628, "grad_norm": 13.59367847442627, "learning_rate": 1.1939648919193386e-06, "loss": 0.0212, "step": 26026 }, { "epoch": 2.8216608846487423, "grad_norm": 3.3945586681365967, "learning_rate": 1.1932395183519514e-06, "loss": 0.0023, "step": 26027 }, { "epoch": 2.8217692974848223, "grad_norm": 0.07403525710105896, "learning_rate": 1.1925141447845641e-06, "loss": 0.0002, "step": 26028 }, { "epoch": 2.821877710320902, "grad_norm": 0.2271614968776703, "learning_rate": 1.191788771217177e-06, "loss": 0.0006, "step": 26029 }, { "epoch": 2.821986123156982, "grad_norm": 1.0565671920776367, "learning_rate": 1.1910633976497897e-06, "loss": 0.0479, "step": 26030 }, { "epoch": 2.8220945359930614, "grad_norm": 0.6652761101722717, "learning_rate": 1.1903380240824025e-06, "loss": 0.0112, "step": 26031 }, { "epoch": 2.8222029488291414, "grad_norm": 0.035976752638816833, "learning_rate": 1.1896126505150152e-06, "loss": 0.0003, "step": 26032 }, { "epoch": 2.8223113616652213, "grad_norm": 0.20627427101135254, "learning_rate": 1.188887276947628e-06, "loss": 0.0009, "step": 26033 }, { "epoch": 2.822419774501301, "grad_norm": 0.49204733967781067, "learning_rate": 1.188161903380241e-06, "loss": 0.004, "step": 26034 }, { "epoch": 2.822528187337381, "grad_norm": 0.7179019451141357, "learning_rate": 1.1874365298128538e-06, "loss": 0.0075, "step": 26035 }, { "epoch": 2.8226366001734604, "grad_norm": 0.0479542538523674, "learning_rate": 1.1867111562454666e-06, "loss": 0.0004, "step": 26036 }, { "epoch": 2.8227450130095404, "grad_norm": 0.02773152105510235, "learning_rate": 1.1859857826780794e-06, "loss": 0.0001, "step": 26037 }, { "epoch": 2.8228534258456204, "grad_norm": 0.01883992739021778, "learning_rate": 1.1852604091106921e-06, "loss": 0.0001, "step": 26038 }, { "epoch": 2.8229618386817, "grad_norm": 0.41474905610084534, "learning_rate": 1.184535035543305e-06, "loss": 0.013, "step": 26039 }, { "epoch": 2.8230702515177795, "grad_norm": 0.38296306133270264, "learning_rate": 1.1838096619759177e-06, "loss": 0.0119, "step": 26040 }, { "epoch": 2.8231786643538594, "grad_norm": 0.433931440114975, "learning_rate": 1.1830842884085305e-06, "loss": 0.0258, "step": 26041 }, { "epoch": 2.8232870771899394, "grad_norm": 0.7806884050369263, "learning_rate": 1.1823589148411432e-06, "loss": 0.0078, "step": 26042 }, { "epoch": 2.823395490026019, "grad_norm": 2.6413683891296387, "learning_rate": 1.181633541273756e-06, "loss": 0.0273, "step": 26043 }, { "epoch": 2.823503902862099, "grad_norm": 0.18529696762561798, "learning_rate": 1.1809081677063688e-06, "loss": 0.0025, "step": 26044 }, { "epoch": 2.8236123156981785, "grad_norm": 0.6126941442489624, "learning_rate": 1.1801827941389816e-06, "loss": 0.0038, "step": 26045 }, { "epoch": 2.8237207285342585, "grad_norm": 0.022649595513939857, "learning_rate": 1.1794574205715944e-06, "loss": 0.0002, "step": 26046 }, { "epoch": 2.8238291413703385, "grad_norm": 0.12512922286987305, "learning_rate": 1.1787320470042074e-06, "loss": 0.0007, "step": 26047 }, { "epoch": 2.823937554206418, "grad_norm": 0.024675466120243073, "learning_rate": 1.1780066734368201e-06, "loss": 0.0002, "step": 26048 }, { "epoch": 2.824045967042498, "grad_norm": 0.011350058019161224, "learning_rate": 1.177281299869433e-06, "loss": 0.0001, "step": 26049 }, { "epoch": 2.8241543798785775, "grad_norm": 2.010762929916382, "learning_rate": 1.1765559263020457e-06, "loss": 0.0226, "step": 26050 }, { "epoch": 2.8242627927146575, "grad_norm": 0.0316823273897171, "learning_rate": 1.1758305527346585e-06, "loss": 0.0002, "step": 26051 }, { "epoch": 2.824371205550737, "grad_norm": 0.07211800664663315, "learning_rate": 1.1751051791672712e-06, "loss": 0.0009, "step": 26052 }, { "epoch": 2.824479618386817, "grad_norm": 4.291441440582275, "learning_rate": 1.174379805599884e-06, "loss": 0.0223, "step": 26053 }, { "epoch": 2.8245880312228966, "grad_norm": 0.38676491379737854, "learning_rate": 1.1736544320324968e-06, "loss": 0.0078, "step": 26054 }, { "epoch": 2.8246964440589766, "grad_norm": 0.12787435948848724, "learning_rate": 1.1729290584651096e-06, "loss": 0.0009, "step": 26055 }, { "epoch": 2.8248048568950566, "grad_norm": 0.27588361501693726, "learning_rate": 1.1722036848977224e-06, "loss": 0.0057, "step": 26056 }, { "epoch": 2.824913269731136, "grad_norm": 0.027839554473757744, "learning_rate": 1.1714783113303351e-06, "loss": 0.0002, "step": 26057 }, { "epoch": 2.825021682567216, "grad_norm": 4.890411853790283, "learning_rate": 1.170752937762948e-06, "loss": 0.0096, "step": 26058 }, { "epoch": 2.8251300954032956, "grad_norm": 0.03243577107787132, "learning_rate": 1.1700275641955607e-06, "loss": 0.0002, "step": 26059 }, { "epoch": 2.8252385082393756, "grad_norm": 0.3340650200843811, "learning_rate": 1.1693021906281737e-06, "loss": 0.0051, "step": 26060 }, { "epoch": 2.8253469210754556, "grad_norm": 0.49863168597221375, "learning_rate": 1.1685768170607865e-06, "loss": 0.0187, "step": 26061 }, { "epoch": 2.825455333911535, "grad_norm": 0.007313520647585392, "learning_rate": 1.1678514434933992e-06, "loss": 0.0001, "step": 26062 }, { "epoch": 2.8255637467476147, "grad_norm": 0.21736754477024078, "learning_rate": 1.167126069926012e-06, "loss": 0.0007, "step": 26063 }, { "epoch": 2.8256721595836947, "grad_norm": 3.386765718460083, "learning_rate": 1.1664006963586248e-06, "loss": 0.0692, "step": 26064 }, { "epoch": 2.8257805724197746, "grad_norm": 0.2861095666885376, "learning_rate": 1.1656753227912376e-06, "loss": 0.0096, "step": 26065 }, { "epoch": 2.825888985255854, "grad_norm": 2.529909610748291, "learning_rate": 1.1649499492238504e-06, "loss": 0.0016, "step": 26066 }, { "epoch": 2.825997398091934, "grad_norm": 0.30649033188819885, "learning_rate": 1.1642245756564631e-06, "loss": 0.0051, "step": 26067 }, { "epoch": 2.8261058109280137, "grad_norm": 0.2239232212305069, "learning_rate": 1.163499202089076e-06, "loss": 0.0032, "step": 26068 }, { "epoch": 2.8262142237640937, "grad_norm": 0.048805203288793564, "learning_rate": 1.1627738285216887e-06, "loss": 0.0004, "step": 26069 }, { "epoch": 2.8263226366001737, "grad_norm": 0.30263346433639526, "learning_rate": 1.1620484549543015e-06, "loss": 0.0164, "step": 26070 }, { "epoch": 2.8264310494362532, "grad_norm": 1.9543349742889404, "learning_rate": 1.1613230813869143e-06, "loss": 0.0526, "step": 26071 }, { "epoch": 2.8265394622723328, "grad_norm": 1.425916075706482, "learning_rate": 1.160597707819527e-06, "loss": 0.012, "step": 26072 }, { "epoch": 2.8266478751084128, "grad_norm": 1.2059282064437866, "learning_rate": 1.15987233425214e-06, "loss": 0.022, "step": 26073 }, { "epoch": 2.8267562879444927, "grad_norm": 0.029643677175045013, "learning_rate": 1.1591469606847528e-06, "loss": 0.0003, "step": 26074 }, { "epoch": 2.8268647007805723, "grad_norm": 0.08954375982284546, "learning_rate": 1.1584215871173656e-06, "loss": 0.0002, "step": 26075 }, { "epoch": 2.8269731136166523, "grad_norm": 0.16449840366840363, "learning_rate": 1.1576962135499784e-06, "loss": 0.0022, "step": 26076 }, { "epoch": 2.827081526452732, "grad_norm": 0.0362575463950634, "learning_rate": 1.1569708399825911e-06, "loss": 0.0002, "step": 26077 }, { "epoch": 2.827189939288812, "grad_norm": 0.28448250889778137, "learning_rate": 1.156245466415204e-06, "loss": 0.0049, "step": 26078 }, { "epoch": 2.827298352124892, "grad_norm": 5.189523696899414, "learning_rate": 1.1555200928478167e-06, "loss": 0.0572, "step": 26079 }, { "epoch": 2.8274067649609713, "grad_norm": 0.04001917690038681, "learning_rate": 1.1547947192804295e-06, "loss": 0.0004, "step": 26080 }, { "epoch": 2.8275151777970513, "grad_norm": 0.291016161441803, "learning_rate": 1.1540693457130423e-06, "loss": 0.0008, "step": 26081 }, { "epoch": 2.827623590633131, "grad_norm": 0.037018317729234695, "learning_rate": 1.153343972145655e-06, "loss": 0.0003, "step": 26082 }, { "epoch": 2.827732003469211, "grad_norm": 0.25294628739356995, "learning_rate": 1.1526185985782678e-06, "loss": 0.015, "step": 26083 }, { "epoch": 2.827840416305291, "grad_norm": 15.885415077209473, "learning_rate": 1.1518932250108806e-06, "loss": 0.0214, "step": 26084 }, { "epoch": 2.8279488291413704, "grad_norm": 0.3134453594684601, "learning_rate": 1.1511678514434934e-06, "loss": 0.0008, "step": 26085 }, { "epoch": 2.82805724197745, "grad_norm": 0.058450374752283096, "learning_rate": 1.1504424778761064e-06, "loss": 0.0003, "step": 26086 }, { "epoch": 2.82816565481353, "grad_norm": 2.266237258911133, "learning_rate": 1.1497171043087191e-06, "loss": 0.0346, "step": 26087 }, { "epoch": 2.82827406764961, "grad_norm": 0.5932954549789429, "learning_rate": 1.148991730741332e-06, "loss": 0.0135, "step": 26088 }, { "epoch": 2.8283824804856894, "grad_norm": 0.2536148428916931, "learning_rate": 1.1482663571739447e-06, "loss": 0.0008, "step": 26089 }, { "epoch": 2.8284908933217694, "grad_norm": 0.5751407742500305, "learning_rate": 1.1475409836065575e-06, "loss": 0.0172, "step": 26090 }, { "epoch": 2.828599306157849, "grad_norm": 0.34553858637809753, "learning_rate": 1.1468156100391702e-06, "loss": 0.0074, "step": 26091 }, { "epoch": 2.828707718993929, "grad_norm": 0.06225435808300972, "learning_rate": 1.146090236471783e-06, "loss": 0.0005, "step": 26092 }, { "epoch": 2.828816131830009, "grad_norm": 0.3175200819969177, "learning_rate": 1.145364862904396e-06, "loss": 0.0068, "step": 26093 }, { "epoch": 2.8289245446660884, "grad_norm": 0.2774107754230499, "learning_rate": 1.1446394893370088e-06, "loss": 0.0006, "step": 26094 }, { "epoch": 2.829032957502168, "grad_norm": 0.021442381665110588, "learning_rate": 1.1439141157696214e-06, "loss": 0.0002, "step": 26095 }, { "epoch": 2.829141370338248, "grad_norm": 0.36546316742897034, "learning_rate": 1.1431887422022341e-06, "loss": 0.0008, "step": 26096 }, { "epoch": 2.829249783174328, "grad_norm": 3.49936580657959, "learning_rate": 1.142463368634847e-06, "loss": 0.0036, "step": 26097 }, { "epoch": 2.8293581960104075, "grad_norm": 3.1190433502197266, "learning_rate": 1.1417379950674597e-06, "loss": 0.0393, "step": 26098 }, { "epoch": 2.8294666088464875, "grad_norm": 0.5554168224334717, "learning_rate": 1.1410126215000727e-06, "loss": 0.0193, "step": 26099 }, { "epoch": 2.829575021682567, "grad_norm": 0.17332646250724792, "learning_rate": 1.1402872479326855e-06, "loss": 0.0007, "step": 26100 }, { "epoch": 2.829683434518647, "grad_norm": 0.5628015398979187, "learning_rate": 1.1395618743652982e-06, "loss": 0.0177, "step": 26101 }, { "epoch": 2.829791847354727, "grad_norm": 6.901909828186035, "learning_rate": 1.138836500797911e-06, "loss": 0.0168, "step": 26102 }, { "epoch": 2.8299002601908065, "grad_norm": 0.28645384311676025, "learning_rate": 1.1381111272305238e-06, "loss": 0.0063, "step": 26103 }, { "epoch": 2.8300086730268865, "grad_norm": 5.303163051605225, "learning_rate": 1.1373857536631366e-06, "loss": 0.0155, "step": 26104 }, { "epoch": 2.830117085862966, "grad_norm": 0.3145933747291565, "learning_rate": 1.1366603800957494e-06, "loss": 0.0012, "step": 26105 }, { "epoch": 2.830225498699046, "grad_norm": 2.7621946334838867, "learning_rate": 1.1359350065283624e-06, "loss": 0.0065, "step": 26106 }, { "epoch": 2.830333911535126, "grad_norm": 0.42811086773872375, "learning_rate": 1.1352096329609751e-06, "loss": 0.0069, "step": 26107 }, { "epoch": 2.8304423243712056, "grad_norm": 0.49517813324928284, "learning_rate": 1.1344842593935877e-06, "loss": 0.0011, "step": 26108 }, { "epoch": 2.830550737207285, "grad_norm": 0.6100776791572571, "learning_rate": 1.1337588858262005e-06, "loss": 0.0055, "step": 26109 }, { "epoch": 2.830659150043365, "grad_norm": 0.5531506538391113, "learning_rate": 1.1330335122588133e-06, "loss": 0.0185, "step": 26110 }, { "epoch": 2.830767562879445, "grad_norm": 0.015494463965296745, "learning_rate": 1.132308138691426e-06, "loss": 0.0002, "step": 26111 }, { "epoch": 2.8308759757155246, "grad_norm": 0.8907744288444519, "learning_rate": 1.131582765124039e-06, "loss": 0.0299, "step": 26112 }, { "epoch": 2.8309843885516046, "grad_norm": 2.2563138008117676, "learning_rate": 1.1308573915566518e-06, "loss": 0.0475, "step": 26113 }, { "epoch": 2.831092801387684, "grad_norm": 0.1858765333890915, "learning_rate": 1.1301320179892646e-06, "loss": 0.0009, "step": 26114 }, { "epoch": 2.831201214223764, "grad_norm": 0.4687403738498688, "learning_rate": 1.1294066444218774e-06, "loss": 0.0045, "step": 26115 }, { "epoch": 2.831309627059844, "grad_norm": 0.05271361395716667, "learning_rate": 1.1286812708544901e-06, "loss": 0.0004, "step": 26116 }, { "epoch": 2.8314180398959237, "grad_norm": 1.5089669227600098, "learning_rate": 1.127955897287103e-06, "loss": 0.0237, "step": 26117 }, { "epoch": 2.831526452732003, "grad_norm": 0.26784762740135193, "learning_rate": 1.1272305237197157e-06, "loss": 0.0016, "step": 26118 }, { "epoch": 2.831634865568083, "grad_norm": 0.35241425037384033, "learning_rate": 1.1265051501523287e-06, "loss": 0.0116, "step": 26119 }, { "epoch": 2.831743278404163, "grad_norm": 0.025277357548475266, "learning_rate": 1.1257797765849415e-06, "loss": 0.0002, "step": 26120 }, { "epoch": 2.8318516912402427, "grad_norm": 0.453214168548584, "learning_rate": 1.1250544030175542e-06, "loss": 0.0088, "step": 26121 }, { "epoch": 2.8319601040763227, "grad_norm": 0.5151692628860474, "learning_rate": 1.1243290294501668e-06, "loss": 0.0149, "step": 26122 }, { "epoch": 2.8320685169124022, "grad_norm": 0.2657291293144226, "learning_rate": 1.1236036558827796e-06, "loss": 0.0051, "step": 26123 }, { "epoch": 2.8321769297484822, "grad_norm": 2.519650459289551, "learning_rate": 1.1228782823153926e-06, "loss": 0.0277, "step": 26124 }, { "epoch": 2.832285342584562, "grad_norm": 0.6034983396530151, "learning_rate": 1.1221529087480054e-06, "loss": 0.0188, "step": 26125 }, { "epoch": 2.8323937554206418, "grad_norm": 1.4111183881759644, "learning_rate": 1.1214275351806181e-06, "loss": 0.0072, "step": 26126 }, { "epoch": 2.8325021682567217, "grad_norm": 0.5067136883735657, "learning_rate": 1.120702161613231e-06, "loss": 0.0068, "step": 26127 }, { "epoch": 2.8326105810928013, "grad_norm": 1.93429434299469, "learning_rate": 1.1199767880458437e-06, "loss": 0.0267, "step": 26128 }, { "epoch": 2.8327189939288813, "grad_norm": 0.47628578543663025, "learning_rate": 1.1192514144784565e-06, "loss": 0.0197, "step": 26129 }, { "epoch": 2.832827406764961, "grad_norm": 0.8997023701667786, "learning_rate": 1.1185260409110693e-06, "loss": 0.0318, "step": 26130 }, { "epoch": 2.832935819601041, "grad_norm": 0.16453158855438232, "learning_rate": 1.117800667343682e-06, "loss": 0.0004, "step": 26131 }, { "epoch": 2.8330442324371203, "grad_norm": 0.5547502040863037, "learning_rate": 1.117075293776295e-06, "loss": 0.0066, "step": 26132 }, { "epoch": 2.8331526452732003, "grad_norm": 0.04151759669184685, "learning_rate": 1.1163499202089078e-06, "loss": 0.0003, "step": 26133 }, { "epoch": 2.8332610581092803, "grad_norm": 2.137763023376465, "learning_rate": 1.1156245466415206e-06, "loss": 0.0331, "step": 26134 }, { "epoch": 2.83336947094536, "grad_norm": 1.874929666519165, "learning_rate": 1.1148991730741331e-06, "loss": 0.0047, "step": 26135 }, { "epoch": 2.83347788378144, "grad_norm": 9.029585838317871, "learning_rate": 1.114173799506746e-06, "loss": 0.019, "step": 26136 }, { "epoch": 2.8335862966175194, "grad_norm": 7.943870544433594, "learning_rate": 1.113448425939359e-06, "loss": 0.0096, "step": 26137 }, { "epoch": 2.8336947094535994, "grad_norm": 0.4042827785015106, "learning_rate": 1.1127230523719717e-06, "loss": 0.0133, "step": 26138 }, { "epoch": 2.8338031222896793, "grad_norm": 14.705334663391113, "learning_rate": 1.1119976788045845e-06, "loss": 0.0159, "step": 26139 }, { "epoch": 2.833911535125759, "grad_norm": 0.07275837659835815, "learning_rate": 1.1112723052371973e-06, "loss": 0.0005, "step": 26140 }, { "epoch": 2.8340199479618384, "grad_norm": 6.599242210388184, "learning_rate": 1.11054693166981e-06, "loss": 0.0189, "step": 26141 }, { "epoch": 2.8341283607979184, "grad_norm": 0.06017167121171951, "learning_rate": 1.1098215581024228e-06, "loss": 0.0001, "step": 26142 }, { "epoch": 2.8342367736339984, "grad_norm": 0.5124980807304382, "learning_rate": 1.1090961845350356e-06, "loss": 0.0102, "step": 26143 }, { "epoch": 2.834345186470078, "grad_norm": 2.7350752353668213, "learning_rate": 1.1083708109676484e-06, "loss": 0.0219, "step": 26144 }, { "epoch": 2.834453599306158, "grad_norm": 0.13015452027320862, "learning_rate": 1.1076454374002614e-06, "loss": 0.0006, "step": 26145 }, { "epoch": 2.8345620121422375, "grad_norm": 0.042747389525175095, "learning_rate": 1.1069200638328741e-06, "loss": 0.0003, "step": 26146 }, { "epoch": 2.8346704249783174, "grad_norm": 6.66795539855957, "learning_rate": 1.106194690265487e-06, "loss": 0.064, "step": 26147 }, { "epoch": 2.8347788378143974, "grad_norm": 3.0990118980407715, "learning_rate": 1.1054693166980997e-06, "loss": 0.025, "step": 26148 }, { "epoch": 2.834887250650477, "grad_norm": 0.34646645188331604, "learning_rate": 1.1047439431307123e-06, "loss": 0.015, "step": 26149 }, { "epoch": 2.834995663486557, "grad_norm": 0.3282785415649414, "learning_rate": 1.1040185695633252e-06, "loss": 0.0093, "step": 26150 }, { "epoch": 2.8351040763226365, "grad_norm": 31.03059196472168, "learning_rate": 1.103293195995938e-06, "loss": 0.0211, "step": 26151 }, { "epoch": 2.8352124891587165, "grad_norm": 0.31150832772254944, "learning_rate": 1.1025678224285508e-06, "loss": 0.0012, "step": 26152 }, { "epoch": 2.835320901994796, "grad_norm": 0.019706469029188156, "learning_rate": 1.1018424488611636e-06, "loss": 0.0002, "step": 26153 }, { "epoch": 2.835429314830876, "grad_norm": 0.03792212903499603, "learning_rate": 1.1011170752937764e-06, "loss": 0.0002, "step": 26154 }, { "epoch": 2.8355377276669556, "grad_norm": 1.2153751850128174, "learning_rate": 1.1003917017263891e-06, "loss": 0.0008, "step": 26155 }, { "epoch": 2.8356461405030355, "grad_norm": 0.40041521191596985, "learning_rate": 1.099666328159002e-06, "loss": 0.0008, "step": 26156 }, { "epoch": 2.8357545533391155, "grad_norm": 0.4744660258293152, "learning_rate": 1.0989409545916147e-06, "loss": 0.0027, "step": 26157 }, { "epoch": 2.835862966175195, "grad_norm": 5.1492462158203125, "learning_rate": 1.0982155810242277e-06, "loss": 0.0276, "step": 26158 }, { "epoch": 2.835971379011275, "grad_norm": 0.28082478046417236, "learning_rate": 1.0974902074568405e-06, "loss": 0.0063, "step": 26159 }, { "epoch": 2.8360797918473546, "grad_norm": 3.850425958633423, "learning_rate": 1.0967648338894532e-06, "loss": 0.0334, "step": 26160 }, { "epoch": 2.8361882046834346, "grad_norm": 2.4208106994628906, "learning_rate": 1.096039460322066e-06, "loss": 0.0034, "step": 26161 }, { "epoch": 2.8362966175195146, "grad_norm": 0.45328083634376526, "learning_rate": 1.0953140867546788e-06, "loss": 0.0125, "step": 26162 }, { "epoch": 2.836405030355594, "grad_norm": 0.5185182094573975, "learning_rate": 1.0945887131872916e-06, "loss": 0.0146, "step": 26163 }, { "epoch": 2.8365134431916736, "grad_norm": 0.26733139157295227, "learning_rate": 1.0938633396199044e-06, "loss": 0.005, "step": 26164 }, { "epoch": 2.8366218560277536, "grad_norm": 0.9278663396835327, "learning_rate": 1.0931379660525171e-06, "loss": 0.0038, "step": 26165 }, { "epoch": 2.8367302688638336, "grad_norm": 0.13353632390499115, "learning_rate": 1.09241259248513e-06, "loss": 0.0003, "step": 26166 }, { "epoch": 2.836838681699913, "grad_norm": 0.3267732560634613, "learning_rate": 1.0916872189177427e-06, "loss": 0.0102, "step": 26167 }, { "epoch": 2.836947094535993, "grad_norm": 0.1027117520570755, "learning_rate": 1.0909618453503555e-06, "loss": 0.0003, "step": 26168 }, { "epoch": 2.8370555073720727, "grad_norm": 6.531855583190918, "learning_rate": 1.0902364717829683e-06, "loss": 0.0544, "step": 26169 }, { "epoch": 2.8371639202081527, "grad_norm": 0.33749204874038696, "learning_rate": 1.089511098215581e-06, "loss": 0.0021, "step": 26170 }, { "epoch": 2.8372723330442327, "grad_norm": 20.18706703186035, "learning_rate": 1.088785724648194e-06, "loss": 0.0275, "step": 26171 }, { "epoch": 2.837380745880312, "grad_norm": 4.004715919494629, "learning_rate": 1.0880603510808068e-06, "loss": 0.0266, "step": 26172 }, { "epoch": 2.837489158716392, "grad_norm": 0.6442971229553223, "learning_rate": 1.0873349775134196e-06, "loss": 0.0159, "step": 26173 }, { "epoch": 2.8375975715524717, "grad_norm": 0.24137607216835022, "learning_rate": 1.0866096039460324e-06, "loss": 0.0009, "step": 26174 }, { "epoch": 2.8377059843885517, "grad_norm": 7.064743995666504, "learning_rate": 1.0858842303786451e-06, "loss": 0.0513, "step": 26175 }, { "epoch": 2.8378143972246312, "grad_norm": 8.589614868164062, "learning_rate": 1.085158856811258e-06, "loss": 0.0134, "step": 26176 }, { "epoch": 2.8379228100607112, "grad_norm": 3.248277425765991, "learning_rate": 1.0844334832438707e-06, "loss": 0.0368, "step": 26177 }, { "epoch": 2.8380312228967908, "grad_norm": 0.06123238056898117, "learning_rate": 1.0837081096764835e-06, "loss": 0.0003, "step": 26178 }, { "epoch": 2.8381396357328708, "grad_norm": 0.4623488187789917, "learning_rate": 1.0829827361090963e-06, "loss": 0.0065, "step": 26179 }, { "epoch": 2.8382480485689507, "grad_norm": 0.24687901139259338, "learning_rate": 1.082257362541709e-06, "loss": 0.0041, "step": 26180 }, { "epoch": 2.8383564614050303, "grad_norm": 0.019321365281939507, "learning_rate": 1.0815319889743218e-06, "loss": 0.0002, "step": 26181 }, { "epoch": 2.8384648742411103, "grad_norm": 3.763270378112793, "learning_rate": 1.0808066154069346e-06, "loss": 0.032, "step": 26182 }, { "epoch": 2.83857328707719, "grad_norm": 0.016348252072930336, "learning_rate": 1.0800812418395474e-06, "loss": 0.0001, "step": 26183 }, { "epoch": 2.83868169991327, "grad_norm": 0.23759636282920837, "learning_rate": 1.0793558682721604e-06, "loss": 0.0157, "step": 26184 }, { "epoch": 2.83879011274935, "grad_norm": 0.8623987436294556, "learning_rate": 1.0786304947047731e-06, "loss": 0.0072, "step": 26185 }, { "epoch": 2.8388985255854293, "grad_norm": 0.5513914823532104, "learning_rate": 1.077905121137386e-06, "loss": 0.0014, "step": 26186 }, { "epoch": 2.839006938421509, "grad_norm": 0.04910672828555107, "learning_rate": 1.0771797475699987e-06, "loss": 0.0002, "step": 26187 }, { "epoch": 2.839115351257589, "grad_norm": 0.08039109408855438, "learning_rate": 1.0764543740026115e-06, "loss": 0.0004, "step": 26188 }, { "epoch": 2.839223764093669, "grad_norm": 0.14990627765655518, "learning_rate": 1.0757290004352243e-06, "loss": 0.0003, "step": 26189 }, { "epoch": 2.8393321769297484, "grad_norm": 0.04274178668856621, "learning_rate": 1.075003626867837e-06, "loss": 0.0004, "step": 26190 }, { "epoch": 2.8394405897658284, "grad_norm": 3.921299934387207, "learning_rate": 1.0742782533004498e-06, "loss": 0.0051, "step": 26191 }, { "epoch": 2.839549002601908, "grad_norm": 0.06498093158006668, "learning_rate": 1.0735528797330626e-06, "loss": 0.0004, "step": 26192 }, { "epoch": 2.839657415437988, "grad_norm": 0.053465090692043304, "learning_rate": 1.0728275061656754e-06, "loss": 0.0005, "step": 26193 }, { "epoch": 2.839765828274068, "grad_norm": 1.586195468902588, "learning_rate": 1.0721021325982881e-06, "loss": 0.0026, "step": 26194 }, { "epoch": 2.8398742411101474, "grad_norm": 0.8987025022506714, "learning_rate": 1.071376759030901e-06, "loss": 0.0206, "step": 26195 }, { "epoch": 2.8399826539462274, "grad_norm": 3.0342094898223877, "learning_rate": 1.0706513854635137e-06, "loss": 0.0294, "step": 26196 }, { "epoch": 2.840091066782307, "grad_norm": 0.008589412085711956, "learning_rate": 1.0699260118961267e-06, "loss": 0.0001, "step": 26197 }, { "epoch": 2.840199479618387, "grad_norm": 11.809232711791992, "learning_rate": 1.0692006383287395e-06, "loss": 0.0273, "step": 26198 }, { "epoch": 2.8403078924544665, "grad_norm": 0.37724968791007996, "learning_rate": 1.0684752647613523e-06, "loss": 0.0014, "step": 26199 }, { "epoch": 2.8404163052905465, "grad_norm": 8.902891159057617, "learning_rate": 1.067749891193965e-06, "loss": 0.0234, "step": 26200 }, { "epoch": 2.840524718126626, "grad_norm": 0.23886342346668243, "learning_rate": 1.0670245176265778e-06, "loss": 0.0039, "step": 26201 }, { "epoch": 2.840633130962706, "grad_norm": 0.08255455642938614, "learning_rate": 1.0662991440591906e-06, "loss": 0.0005, "step": 26202 }, { "epoch": 2.840741543798786, "grad_norm": 1.7910927534103394, "learning_rate": 1.0655737704918034e-06, "loss": 0.0065, "step": 26203 }, { "epoch": 2.8408499566348655, "grad_norm": 3.2818305492401123, "learning_rate": 1.0648483969244161e-06, "loss": 0.0661, "step": 26204 }, { "epoch": 2.8409583694709455, "grad_norm": 0.23893728852272034, "learning_rate": 1.064123023357029e-06, "loss": 0.0037, "step": 26205 }, { "epoch": 2.841066782307025, "grad_norm": 0.012608770281076431, "learning_rate": 1.0633976497896417e-06, "loss": 0.0001, "step": 26206 }, { "epoch": 2.841175195143105, "grad_norm": 0.5768680572509766, "learning_rate": 1.0626722762222545e-06, "loss": 0.0064, "step": 26207 }, { "epoch": 2.841283607979185, "grad_norm": 1.7885202169418335, "learning_rate": 1.0619469026548673e-06, "loss": 0.0092, "step": 26208 }, { "epoch": 2.8413920208152645, "grad_norm": 0.07450304925441742, "learning_rate": 1.06122152908748e-06, "loss": 0.0004, "step": 26209 }, { "epoch": 2.841500433651344, "grad_norm": 0.02266210876405239, "learning_rate": 1.060496155520093e-06, "loss": 0.0002, "step": 26210 }, { "epoch": 2.841608846487424, "grad_norm": 0.22770167887210846, "learning_rate": 1.0597707819527058e-06, "loss": 0.0031, "step": 26211 }, { "epoch": 2.841717259323504, "grad_norm": 2.1959316730499268, "learning_rate": 1.0590454083853186e-06, "loss": 0.0423, "step": 26212 }, { "epoch": 2.8418256721595836, "grad_norm": 0.29814788699150085, "learning_rate": 1.0583200348179314e-06, "loss": 0.0055, "step": 26213 }, { "epoch": 2.8419340849956636, "grad_norm": 3.9667422771453857, "learning_rate": 1.0575946612505441e-06, "loss": 0.051, "step": 26214 }, { "epoch": 2.842042497831743, "grad_norm": 0.3378980755805969, "learning_rate": 1.056869287683157e-06, "loss": 0.0066, "step": 26215 }, { "epoch": 2.842150910667823, "grad_norm": 0.34914273023605347, "learning_rate": 1.0561439141157697e-06, "loss": 0.0029, "step": 26216 }, { "epoch": 2.842259323503903, "grad_norm": 2.4143929481506348, "learning_rate": 1.0554185405483825e-06, "loss": 0.0354, "step": 26217 }, { "epoch": 2.8423677363399826, "grad_norm": 0.08855347335338593, "learning_rate": 1.0546931669809953e-06, "loss": 0.0006, "step": 26218 }, { "epoch": 2.842476149176062, "grad_norm": 3.3488290309906006, "learning_rate": 1.053967793413608e-06, "loss": 0.0017, "step": 26219 }, { "epoch": 2.842584562012142, "grad_norm": 0.38365596532821655, "learning_rate": 1.0532424198462208e-06, "loss": 0.0064, "step": 26220 }, { "epoch": 2.842692974848222, "grad_norm": 1.1090874671936035, "learning_rate": 1.0525170462788336e-06, "loss": 0.0515, "step": 26221 }, { "epoch": 2.8428013876843017, "grad_norm": 0.4336073696613312, "learning_rate": 1.0517916727114464e-06, "loss": 0.0128, "step": 26222 }, { "epoch": 2.8429098005203817, "grad_norm": 1.1774991750717163, "learning_rate": 1.0510662991440594e-06, "loss": 0.0156, "step": 26223 }, { "epoch": 2.843018213356461, "grad_norm": 0.7398813366889954, "learning_rate": 1.0503409255766721e-06, "loss": 0.023, "step": 26224 }, { "epoch": 2.843126626192541, "grad_norm": 6.475842475891113, "learning_rate": 1.049615552009285e-06, "loss": 0.0189, "step": 26225 }, { "epoch": 2.843235039028621, "grad_norm": 0.019570661708712578, "learning_rate": 1.0488901784418977e-06, "loss": 0.0002, "step": 26226 }, { "epoch": 2.8433434518647007, "grad_norm": 1.2438102960586548, "learning_rate": 1.0481648048745105e-06, "loss": 0.0052, "step": 26227 }, { "epoch": 2.8434518647007807, "grad_norm": 0.7312417030334473, "learning_rate": 1.0474394313071233e-06, "loss": 0.003, "step": 26228 }, { "epoch": 2.8435602775368602, "grad_norm": 0.6002578139305115, "learning_rate": 1.046714057739736e-06, "loss": 0.0112, "step": 26229 }, { "epoch": 2.8436686903729402, "grad_norm": 0.5922995805740356, "learning_rate": 1.0459886841723488e-06, "loss": 0.0104, "step": 26230 }, { "epoch": 2.84377710320902, "grad_norm": 7.560028553009033, "learning_rate": 1.0452633106049616e-06, "loss": 0.0288, "step": 26231 }, { "epoch": 2.8438855160450998, "grad_norm": 1.8629862070083618, "learning_rate": 1.0445379370375744e-06, "loss": 0.015, "step": 26232 }, { "epoch": 2.8439939288811793, "grad_norm": 0.3496670722961426, "learning_rate": 1.0438125634701871e-06, "loss": 0.0125, "step": 26233 }, { "epoch": 2.8441023417172593, "grad_norm": 0.5969536304473877, "learning_rate": 1.0430871899028e-06, "loss": 0.025, "step": 26234 }, { "epoch": 2.8442107545533393, "grad_norm": 0.5821090340614319, "learning_rate": 1.0423618163354127e-06, "loss": 0.0129, "step": 26235 }, { "epoch": 2.844319167389419, "grad_norm": 0.3218841254711151, "learning_rate": 1.0416364427680257e-06, "loss": 0.0068, "step": 26236 }, { "epoch": 2.844427580225499, "grad_norm": 2.853457450866699, "learning_rate": 1.0409110692006385e-06, "loss": 0.0595, "step": 26237 }, { "epoch": 2.8445359930615783, "grad_norm": 0.4041990041732788, "learning_rate": 1.0401856956332513e-06, "loss": 0.0105, "step": 26238 }, { "epoch": 2.8446444058976583, "grad_norm": 0.04788807034492493, "learning_rate": 1.039460322065864e-06, "loss": 0.0004, "step": 26239 }, { "epoch": 2.8447528187337383, "grad_norm": 0.3560905158519745, "learning_rate": 1.0387349484984768e-06, "loss": 0.0051, "step": 26240 }, { "epoch": 2.844861231569818, "grad_norm": 2.2036304473876953, "learning_rate": 1.0380095749310896e-06, "loss": 0.0166, "step": 26241 }, { "epoch": 2.8449696444058974, "grad_norm": 0.5786071419715881, "learning_rate": 1.0372842013637024e-06, "loss": 0.0128, "step": 26242 }, { "epoch": 2.8450780572419774, "grad_norm": 0.5862796902656555, "learning_rate": 1.0365588277963151e-06, "loss": 0.0121, "step": 26243 }, { "epoch": 2.8451864700780574, "grad_norm": 0.7905644774436951, "learning_rate": 1.035833454228928e-06, "loss": 0.0158, "step": 26244 }, { "epoch": 2.845294882914137, "grad_norm": 0.021181689575314522, "learning_rate": 1.0351080806615407e-06, "loss": 0.0001, "step": 26245 }, { "epoch": 2.845403295750217, "grad_norm": 1.3864861726760864, "learning_rate": 1.0343827070941535e-06, "loss": 0.0268, "step": 26246 }, { "epoch": 2.8455117085862964, "grad_norm": 0.37350931763648987, "learning_rate": 1.0336573335267663e-06, "loss": 0.0059, "step": 26247 }, { "epoch": 2.8456201214223764, "grad_norm": 0.5462680459022522, "learning_rate": 1.032931959959379e-06, "loss": 0.0115, "step": 26248 }, { "epoch": 2.8457285342584564, "grad_norm": 0.2521395981311798, "learning_rate": 1.032206586391992e-06, "loss": 0.005, "step": 26249 }, { "epoch": 2.845836947094536, "grad_norm": 3.5240728855133057, "learning_rate": 1.0314812128246048e-06, "loss": 0.0123, "step": 26250 }, { "epoch": 2.845945359930616, "grad_norm": 0.014785516075789928, "learning_rate": 1.0307558392572176e-06, "loss": 0.0001, "step": 26251 }, { "epoch": 2.8460537727666955, "grad_norm": 1.0638227462768555, "learning_rate": 1.0300304656898304e-06, "loss": 0.0147, "step": 26252 }, { "epoch": 2.8461621856027755, "grad_norm": 0.5637222528457642, "learning_rate": 1.0293050921224431e-06, "loss": 0.0155, "step": 26253 }, { "epoch": 2.8462705984388554, "grad_norm": 0.6153116822242737, "learning_rate": 1.028579718555056e-06, "loss": 0.0063, "step": 26254 }, { "epoch": 2.846379011274935, "grad_norm": 0.2880745530128479, "learning_rate": 1.0278543449876687e-06, "loss": 0.0044, "step": 26255 }, { "epoch": 2.8464874241110145, "grad_norm": 0.03335176408290863, "learning_rate": 1.0271289714202817e-06, "loss": 0.0001, "step": 26256 }, { "epoch": 2.8465958369470945, "grad_norm": 3.6386377811431885, "learning_rate": 1.0264035978528943e-06, "loss": 0.034, "step": 26257 }, { "epoch": 2.8467042497831745, "grad_norm": 0.18345892429351807, "learning_rate": 1.025678224285507e-06, "loss": 0.0022, "step": 26258 }, { "epoch": 2.846812662619254, "grad_norm": 0.28773385286331177, "learning_rate": 1.0249528507181198e-06, "loss": 0.0004, "step": 26259 }, { "epoch": 2.846921075455334, "grad_norm": 0.5460666418075562, "learning_rate": 1.0242274771507326e-06, "loss": 0.001, "step": 26260 }, { "epoch": 2.8470294882914136, "grad_norm": 0.06480684876441956, "learning_rate": 1.0235021035833454e-06, "loss": 0.0004, "step": 26261 }, { "epoch": 2.8471379011274935, "grad_norm": 22.185562133789062, "learning_rate": 1.0227767300159584e-06, "loss": 0.0064, "step": 26262 }, { "epoch": 2.8472463139635735, "grad_norm": 0.3162105083465576, "learning_rate": 1.0220513564485711e-06, "loss": 0.0049, "step": 26263 }, { "epoch": 2.847354726799653, "grad_norm": 3.6326074600219727, "learning_rate": 1.021325982881184e-06, "loss": 0.0127, "step": 26264 }, { "epoch": 2.8474631396357326, "grad_norm": 0.3935931622982025, "learning_rate": 1.0206006093137967e-06, "loss": 0.0033, "step": 26265 }, { "epoch": 2.8475715524718126, "grad_norm": 3.7635176181793213, "learning_rate": 1.0198752357464095e-06, "loss": 0.0406, "step": 26266 }, { "epoch": 2.8476799653078926, "grad_norm": 0.061082467436790466, "learning_rate": 1.0191498621790223e-06, "loss": 0.0002, "step": 26267 }, { "epoch": 2.847788378143972, "grad_norm": 0.9240368604660034, "learning_rate": 1.018424488611635e-06, "loss": 0.0019, "step": 26268 }, { "epoch": 2.847896790980052, "grad_norm": 0.05373254790902138, "learning_rate": 1.017699115044248e-06, "loss": 0.0002, "step": 26269 }, { "epoch": 2.8480052038161316, "grad_norm": 0.19487117230892181, "learning_rate": 1.0169737414768608e-06, "loss": 0.0009, "step": 26270 }, { "epoch": 2.8481136166522116, "grad_norm": 0.04425765201449394, "learning_rate": 1.0162483679094734e-06, "loss": 0.0003, "step": 26271 }, { "epoch": 2.8482220294882916, "grad_norm": 0.415731281042099, "learning_rate": 1.0155229943420862e-06, "loss": 0.0139, "step": 26272 }, { "epoch": 2.848330442324371, "grad_norm": 0.0389874242246151, "learning_rate": 1.014797620774699e-06, "loss": 0.0002, "step": 26273 }, { "epoch": 2.848438855160451, "grad_norm": 2.5451643466949463, "learning_rate": 1.014072247207312e-06, "loss": 0.0265, "step": 26274 }, { "epoch": 2.8485472679965307, "grad_norm": 1.2934999465942383, "learning_rate": 1.0133468736399247e-06, "loss": 0.0574, "step": 26275 }, { "epoch": 2.8486556808326107, "grad_norm": 0.43302440643310547, "learning_rate": 1.0126215000725375e-06, "loss": 0.013, "step": 26276 }, { "epoch": 2.84876409366869, "grad_norm": 0.29702863097190857, "learning_rate": 1.0118961265051503e-06, "loss": 0.0098, "step": 26277 }, { "epoch": 2.84887250650477, "grad_norm": 0.2407456338405609, "learning_rate": 1.011170752937763e-06, "loss": 0.0009, "step": 26278 }, { "epoch": 2.8489809193408497, "grad_norm": 2.842885971069336, "learning_rate": 1.0104453793703758e-06, "loss": 0.0083, "step": 26279 }, { "epoch": 2.8490893321769297, "grad_norm": 0.5438457727432251, "learning_rate": 1.0097200058029886e-06, "loss": 0.0197, "step": 26280 }, { "epoch": 2.8491977450130097, "grad_norm": 0.07011923938989639, "learning_rate": 1.0089946322356014e-06, "loss": 0.0004, "step": 26281 }, { "epoch": 2.8493061578490892, "grad_norm": 0.01853717677295208, "learning_rate": 1.0082692586682144e-06, "loss": 0.0002, "step": 26282 }, { "epoch": 2.8494145706851692, "grad_norm": 0.03042122721672058, "learning_rate": 1.0075438851008271e-06, "loss": 0.0003, "step": 26283 }, { "epoch": 2.8495229835212488, "grad_norm": 3.6488122940063477, "learning_rate": 1.0068185115334397e-06, "loss": 0.0132, "step": 26284 }, { "epoch": 2.8496313963573288, "grad_norm": 0.0061838459223508835, "learning_rate": 1.0060931379660525e-06, "loss": 0.0001, "step": 26285 }, { "epoch": 2.8497398091934087, "grad_norm": 0.7646732330322266, "learning_rate": 1.0053677643986653e-06, "loss": 0.0148, "step": 26286 }, { "epoch": 2.8498482220294883, "grad_norm": 0.6034162044525146, "learning_rate": 1.0046423908312783e-06, "loss": 0.0166, "step": 26287 }, { "epoch": 2.849956634865568, "grad_norm": 3.073544502258301, "learning_rate": 1.003917017263891e-06, "loss": 0.0144, "step": 26288 }, { "epoch": 2.850065047701648, "grad_norm": 0.022099455818533897, "learning_rate": 1.0031916436965038e-06, "loss": 0.0002, "step": 26289 }, { "epoch": 2.850173460537728, "grad_norm": 0.5533838272094727, "learning_rate": 1.0024662701291166e-06, "loss": 0.0159, "step": 26290 }, { "epoch": 2.8502818733738073, "grad_norm": 0.28401753306388855, "learning_rate": 1.0017408965617294e-06, "loss": 0.0057, "step": 26291 }, { "epoch": 2.8503902862098873, "grad_norm": 0.5932480096817017, "learning_rate": 1.0010155229943421e-06, "loss": 0.0129, "step": 26292 }, { "epoch": 2.850498699045967, "grad_norm": 0.014375552535057068, "learning_rate": 1.000290149426955e-06, "loss": 0.0002, "step": 26293 }, { "epoch": 2.850607111882047, "grad_norm": 0.5884516835212708, "learning_rate": 9.995647758595677e-07, "loss": 0.0186, "step": 26294 }, { "epoch": 2.850715524718127, "grad_norm": 3.08807373046875, "learning_rate": 9.988394022921807e-07, "loss": 0.0349, "step": 26295 }, { "epoch": 2.8508239375542064, "grad_norm": 0.044718749821186066, "learning_rate": 9.981140287247935e-07, "loss": 0.0004, "step": 26296 }, { "epoch": 2.8509323503902864, "grad_norm": 0.06264950335025787, "learning_rate": 9.973886551574063e-07, "loss": 0.0004, "step": 26297 }, { "epoch": 2.851040763226366, "grad_norm": 2.999037504196167, "learning_rate": 9.966632815900188e-07, "loss": 0.0243, "step": 26298 }, { "epoch": 2.851149176062446, "grad_norm": 0.3303090035915375, "learning_rate": 9.959379080226316e-07, "loss": 0.006, "step": 26299 }, { "epoch": 2.8512575888985254, "grad_norm": 0.006625500973314047, "learning_rate": 9.952125344552446e-07, "loss": 0.0001, "step": 26300 }, { "epoch": 2.8513660017346054, "grad_norm": 0.0930553749203682, "learning_rate": 9.944871608878574e-07, "loss": 0.0002, "step": 26301 }, { "epoch": 2.851474414570685, "grad_norm": 0.8117078542709351, "learning_rate": 9.937617873204701e-07, "loss": 0.0017, "step": 26302 }, { "epoch": 2.851582827406765, "grad_norm": 0.34762874245643616, "learning_rate": 9.93036413753083e-07, "loss": 0.0101, "step": 26303 }, { "epoch": 2.851691240242845, "grad_norm": 1.4136080741882324, "learning_rate": 9.923110401856957e-07, "loss": 0.0366, "step": 26304 }, { "epoch": 2.8517996530789245, "grad_norm": 0.02240760438144207, "learning_rate": 9.915856666183085e-07, "loss": 0.0001, "step": 26305 }, { "epoch": 2.8519080659150045, "grad_norm": 0.010715150274336338, "learning_rate": 9.908602930509213e-07, "loss": 0.0001, "step": 26306 }, { "epoch": 2.852016478751084, "grad_norm": 0.5159674882888794, "learning_rate": 9.90134919483534e-07, "loss": 0.0087, "step": 26307 }, { "epoch": 2.852124891587164, "grad_norm": 2.8828163146972656, "learning_rate": 9.89409545916147e-07, "loss": 0.036, "step": 26308 }, { "epoch": 2.852233304423244, "grad_norm": 0.04549247771501541, "learning_rate": 9.886841723487598e-07, "loss": 0.0002, "step": 26309 }, { "epoch": 2.8523417172593235, "grad_norm": 0.540680468082428, "learning_rate": 9.879587987813726e-07, "loss": 0.0153, "step": 26310 }, { "epoch": 2.852450130095403, "grad_norm": 0.025634722784161568, "learning_rate": 9.872334252139854e-07, "loss": 0.0002, "step": 26311 }, { "epoch": 2.852558542931483, "grad_norm": 0.7289854288101196, "learning_rate": 9.86508051646598e-07, "loss": 0.0251, "step": 26312 }, { "epoch": 2.852666955767563, "grad_norm": 0.36734697222709656, "learning_rate": 9.85782678079211e-07, "loss": 0.0081, "step": 26313 }, { "epoch": 2.8527753686036426, "grad_norm": 1.0295829772949219, "learning_rate": 9.850573045118237e-07, "loss": 0.0088, "step": 26314 }, { "epoch": 2.8528837814397225, "grad_norm": 0.1310955137014389, "learning_rate": 9.843319309444365e-07, "loss": 0.0005, "step": 26315 }, { "epoch": 2.852992194275802, "grad_norm": 0.02287720888853073, "learning_rate": 9.836065573770493e-07, "loss": 0.0002, "step": 26316 }, { "epoch": 2.853100607111882, "grad_norm": 0.06188666820526123, "learning_rate": 9.82881183809662e-07, "loss": 0.0004, "step": 26317 }, { "epoch": 2.853209019947962, "grad_norm": 0.2679468095302582, "learning_rate": 9.821558102422748e-07, "loss": 0.0052, "step": 26318 }, { "epoch": 2.8533174327840416, "grad_norm": 3.55051326751709, "learning_rate": 9.814304366748876e-07, "loss": 0.0143, "step": 26319 }, { "epoch": 2.8534258456201216, "grad_norm": 0.3225504457950592, "learning_rate": 9.807050631075004e-07, "loss": 0.0108, "step": 26320 }, { "epoch": 2.853534258456201, "grad_norm": 0.2468317449092865, "learning_rate": 9.799796895401134e-07, "loss": 0.0011, "step": 26321 }, { "epoch": 2.853642671292281, "grad_norm": 7.661245346069336, "learning_rate": 9.792543159727261e-07, "loss": 0.0196, "step": 26322 }, { "epoch": 2.8537510841283606, "grad_norm": 7.925021171569824, "learning_rate": 9.78528942405339e-07, "loss": 0.006, "step": 26323 }, { "epoch": 2.8538594969644406, "grad_norm": 1.1128690242767334, "learning_rate": 9.778035688379517e-07, "loss": 0.0155, "step": 26324 }, { "epoch": 2.85396790980052, "grad_norm": 0.24512147903442383, "learning_rate": 9.770781952705643e-07, "loss": 0.005, "step": 26325 }, { "epoch": 2.8540763226366, "grad_norm": 0.19951865077018738, "learning_rate": 9.763528217031773e-07, "loss": 0.0007, "step": 26326 }, { "epoch": 2.85418473547268, "grad_norm": 0.895820140838623, "learning_rate": 9.7562744813579e-07, "loss": 0.0349, "step": 26327 }, { "epoch": 2.8542931483087597, "grad_norm": 0.5019001960754395, "learning_rate": 9.749020745684028e-07, "loss": 0.0151, "step": 26328 }, { "epoch": 2.8544015611448397, "grad_norm": 0.40403321385383606, "learning_rate": 9.741767010010156e-07, "loss": 0.0098, "step": 26329 }, { "epoch": 2.854509973980919, "grad_norm": 0.21488793194293976, "learning_rate": 9.734513274336284e-07, "loss": 0.0007, "step": 26330 }, { "epoch": 2.854618386816999, "grad_norm": 0.6128299832344055, "learning_rate": 9.727259538662412e-07, "loss": 0.0173, "step": 26331 }, { "epoch": 2.854726799653079, "grad_norm": 1.728043794631958, "learning_rate": 9.72000580298854e-07, "loss": 0.0142, "step": 26332 }, { "epoch": 2.8548352124891587, "grad_norm": 0.4809100925922394, "learning_rate": 9.712752067314667e-07, "loss": 0.0011, "step": 26333 }, { "epoch": 2.8549436253252383, "grad_norm": 0.6613264083862305, "learning_rate": 9.705498331640797e-07, "loss": 0.0119, "step": 26334 }, { "epoch": 2.8550520381613183, "grad_norm": 0.8656670451164246, "learning_rate": 9.698244595966925e-07, "loss": 0.0222, "step": 26335 }, { "epoch": 2.8551604509973982, "grad_norm": 0.14926989376544952, "learning_rate": 9.690990860293053e-07, "loss": 0.001, "step": 26336 }, { "epoch": 2.8552688638334778, "grad_norm": 0.25412335991859436, "learning_rate": 9.68373712461918e-07, "loss": 0.0053, "step": 26337 }, { "epoch": 2.8553772766695578, "grad_norm": 3.6934385299682617, "learning_rate": 9.676483388945308e-07, "loss": 0.0375, "step": 26338 }, { "epoch": 2.8554856895056373, "grad_norm": 5.8703694343566895, "learning_rate": 9.669229653271436e-07, "loss": 0.0396, "step": 26339 }, { "epoch": 2.8555941023417173, "grad_norm": 0.8856619596481323, "learning_rate": 9.661975917597564e-07, "loss": 0.0188, "step": 26340 }, { "epoch": 2.8557025151777973, "grad_norm": 0.033202338963747025, "learning_rate": 9.654722181923692e-07, "loss": 0.0002, "step": 26341 }, { "epoch": 2.855810928013877, "grad_norm": 0.8774973750114441, "learning_rate": 9.64746844624982e-07, "loss": 0.0171, "step": 26342 }, { "epoch": 2.855919340849957, "grad_norm": 0.32491475343704224, "learning_rate": 9.640214710575947e-07, "loss": 0.0113, "step": 26343 }, { "epoch": 2.8560277536860363, "grad_norm": 0.6092706918716431, "learning_rate": 9.632960974902075e-07, "loss": 0.0131, "step": 26344 }, { "epoch": 2.8561361665221163, "grad_norm": 2.2540652751922607, "learning_rate": 9.625707239228203e-07, "loss": 0.0223, "step": 26345 }, { "epoch": 2.856244579358196, "grad_norm": 0.17324885725975037, "learning_rate": 9.61845350355433e-07, "loss": 0.001, "step": 26346 }, { "epoch": 2.856352992194276, "grad_norm": 0.06345782428979874, "learning_rate": 9.61119976788046e-07, "loss": 0.0005, "step": 26347 }, { "epoch": 2.8564614050303554, "grad_norm": 0.5663777589797974, "learning_rate": 9.603946032206588e-07, "loss": 0.0124, "step": 26348 }, { "epoch": 2.8565698178664354, "grad_norm": 0.5793936848640442, "learning_rate": 9.596692296532716e-07, "loss": 0.0209, "step": 26349 }, { "epoch": 2.8566782307025154, "grad_norm": 0.4104759395122528, "learning_rate": 9.589438560858844e-07, "loss": 0.0114, "step": 26350 }, { "epoch": 2.856786643538595, "grad_norm": 0.498798668384552, "learning_rate": 9.582184825184971e-07, "loss": 0.0073, "step": 26351 }, { "epoch": 2.856895056374675, "grad_norm": 5.379360675811768, "learning_rate": 9.5749310895111e-07, "loss": 0.0307, "step": 26352 }, { "epoch": 2.8570034692107544, "grad_norm": 0.7626525163650513, "learning_rate": 9.567677353837227e-07, "loss": 0.0217, "step": 26353 }, { "epoch": 2.8571118820468344, "grad_norm": 2.1456971168518066, "learning_rate": 9.560423618163355e-07, "loss": 0.0414, "step": 26354 }, { "epoch": 2.8572202948829144, "grad_norm": 0.02425609901547432, "learning_rate": 9.553169882489483e-07, "loss": 0.0003, "step": 26355 }, { "epoch": 2.857328707718994, "grad_norm": 0.8430508971214294, "learning_rate": 9.54591614681561e-07, "loss": 0.0118, "step": 26356 }, { "epoch": 2.8574371205550735, "grad_norm": 3.8698689937591553, "learning_rate": 9.538662411141738e-07, "loss": 0.0327, "step": 26357 }, { "epoch": 2.8575455333911535, "grad_norm": 0.5552389025688171, "learning_rate": 9.531408675467867e-07, "loss": 0.0069, "step": 26358 }, { "epoch": 2.8576539462272335, "grad_norm": 0.5648808479309082, "learning_rate": 9.524154939793995e-07, "loss": 0.0115, "step": 26359 }, { "epoch": 2.857762359063313, "grad_norm": 4.528321266174316, "learning_rate": 9.516901204120123e-07, "loss": 0.0715, "step": 26360 }, { "epoch": 2.857870771899393, "grad_norm": 6.888574600219727, "learning_rate": 9.50964746844625e-07, "loss": 0.0235, "step": 26361 }, { "epoch": 2.8579791847354725, "grad_norm": 9.661482810974121, "learning_rate": 9.502393732772379e-07, "loss": 0.079, "step": 26362 }, { "epoch": 2.8580875975715525, "grad_norm": 3.9436094760894775, "learning_rate": 9.495139997098507e-07, "loss": 0.0016, "step": 26363 }, { "epoch": 2.8581960104076325, "grad_norm": 10.204529762268066, "learning_rate": 9.487886261424635e-07, "loss": 0.0217, "step": 26364 }, { "epoch": 2.858304423243712, "grad_norm": 0.3587174117565155, "learning_rate": 9.480632525750763e-07, "loss": 0.0125, "step": 26365 }, { "epoch": 2.8584128360797916, "grad_norm": 0.06996674090623856, "learning_rate": 9.473378790076889e-07, "loss": 0.0004, "step": 26366 }, { "epoch": 2.8585212489158716, "grad_norm": 0.27436864376068115, "learning_rate": 9.466125054403018e-07, "loss": 0.0058, "step": 26367 }, { "epoch": 2.8586296617519515, "grad_norm": 0.8640151023864746, "learning_rate": 9.458871318729146e-07, "loss": 0.0253, "step": 26368 }, { "epoch": 2.858738074588031, "grad_norm": 0.3085244596004486, "learning_rate": 9.451617583055274e-07, "loss": 0.0003, "step": 26369 }, { "epoch": 2.858846487424111, "grad_norm": 0.0034783589653670788, "learning_rate": 9.444363847381402e-07, "loss": 0.0, "step": 26370 }, { "epoch": 2.8589549002601906, "grad_norm": 1.0416321754455566, "learning_rate": 9.43711011170753e-07, "loss": 0.0177, "step": 26371 }, { "epoch": 2.8590633130962706, "grad_norm": 0.8694385886192322, "learning_rate": 9.429856376033658e-07, "loss": 0.0067, "step": 26372 }, { "epoch": 2.8591717259323506, "grad_norm": 7.192287921905518, "learning_rate": 9.422602640359786e-07, "loss": 0.008, "step": 26373 }, { "epoch": 2.85928013876843, "grad_norm": 2.583261251449585, "learning_rate": 9.415348904685914e-07, "loss": 0.0189, "step": 26374 }, { "epoch": 2.85938855160451, "grad_norm": 0.01966070383787155, "learning_rate": 9.408095169012043e-07, "loss": 0.0001, "step": 26375 }, { "epoch": 2.8594969644405897, "grad_norm": 1.218512773513794, "learning_rate": 9.40084143333817e-07, "loss": 0.006, "step": 26376 }, { "epoch": 2.8596053772766696, "grad_norm": 0.12119026482105255, "learning_rate": 9.393587697664298e-07, "loss": 0.0003, "step": 26377 }, { "epoch": 2.8597137901127496, "grad_norm": 0.6752580404281616, "learning_rate": 9.386333961990426e-07, "loss": 0.0076, "step": 26378 }, { "epoch": 2.859822202948829, "grad_norm": 3.417092800140381, "learning_rate": 9.379080226316553e-07, "loss": 0.0377, "step": 26379 }, { "epoch": 2.8599306157849087, "grad_norm": 0.4196426272392273, "learning_rate": 9.371826490642682e-07, "loss": 0.0144, "step": 26380 }, { "epoch": 2.8600390286209887, "grad_norm": 0.5830623507499695, "learning_rate": 9.364572754968809e-07, "loss": 0.0163, "step": 26381 }, { "epoch": 2.8601474414570687, "grad_norm": 0.6048155426979065, "learning_rate": 9.357319019294937e-07, "loss": 0.0005, "step": 26382 }, { "epoch": 2.860255854293148, "grad_norm": 0.2985745668411255, "learning_rate": 9.350065283621065e-07, "loss": 0.0112, "step": 26383 }, { "epoch": 2.860364267129228, "grad_norm": 4.249625205993652, "learning_rate": 9.342811547947194e-07, "loss": 0.01, "step": 26384 }, { "epoch": 2.8604726799653077, "grad_norm": 0.931174635887146, "learning_rate": 9.335557812273322e-07, "loss": 0.0139, "step": 26385 }, { "epoch": 2.8605810928013877, "grad_norm": 3.5670251846313477, "learning_rate": 9.328304076599449e-07, "loss": 0.0064, "step": 26386 }, { "epoch": 2.8606895056374677, "grad_norm": 14.942955017089844, "learning_rate": 9.321050340925577e-07, "loss": 0.0118, "step": 26387 }, { "epoch": 2.8607979184735473, "grad_norm": 0.29837095737457275, "learning_rate": 9.313796605251706e-07, "loss": 0.0057, "step": 26388 }, { "epoch": 2.860906331309627, "grad_norm": 8.891910552978516, "learning_rate": 9.306542869577834e-07, "loss": 0.048, "step": 26389 }, { "epoch": 2.861014744145707, "grad_norm": 10.859992980957031, "learning_rate": 9.299289133903962e-07, "loss": 0.0277, "step": 26390 }, { "epoch": 2.8611231569817868, "grad_norm": 0.49678856134414673, "learning_rate": 9.292035398230089e-07, "loss": 0.0049, "step": 26391 }, { "epoch": 2.8612315698178663, "grad_norm": 0.28461551666259766, "learning_rate": 9.284781662556218e-07, "loss": 0.0009, "step": 26392 }, { "epoch": 2.8613399826539463, "grad_norm": 7.312432765960693, "learning_rate": 9.277527926882345e-07, "loss": 0.0191, "step": 26393 }, { "epoch": 2.861448395490026, "grad_norm": 0.0075343698263168335, "learning_rate": 9.270274191208473e-07, "loss": 0.0001, "step": 26394 }, { "epoch": 2.861556808326106, "grad_norm": 0.2941370904445648, "learning_rate": 9.2630204555346e-07, "loss": 0.0075, "step": 26395 }, { "epoch": 2.861665221162186, "grad_norm": 0.06322634220123291, "learning_rate": 9.255766719860728e-07, "loss": 0.0002, "step": 26396 }, { "epoch": 2.8617736339982653, "grad_norm": 0.5721311569213867, "learning_rate": 9.248512984186857e-07, "loss": 0.0146, "step": 26397 }, { "epoch": 2.8618820468343453, "grad_norm": 0.9180817008018494, "learning_rate": 9.241259248512985e-07, "loss": 0.0256, "step": 26398 }, { "epoch": 2.861990459670425, "grad_norm": 0.4334319829940796, "learning_rate": 9.234005512839113e-07, "loss": 0.0048, "step": 26399 }, { "epoch": 2.862098872506505, "grad_norm": 0.3004440367221832, "learning_rate": 9.22675177716524e-07, "loss": 0.0062, "step": 26400 }, { "epoch": 2.862207285342585, "grad_norm": 0.06017310172319412, "learning_rate": 9.219498041491369e-07, "loss": 0.0004, "step": 26401 }, { "epoch": 2.8623156981786644, "grad_norm": 0.7303315997123718, "learning_rate": 9.212244305817497e-07, "loss": 0.028, "step": 26402 }, { "epoch": 2.862424111014744, "grad_norm": 0.3046872317790985, "learning_rate": 9.204990570143625e-07, "loss": 0.0093, "step": 26403 }, { "epoch": 2.862532523850824, "grad_norm": 0.05223039165139198, "learning_rate": 9.197736834469753e-07, "loss": 0.0003, "step": 26404 }, { "epoch": 2.862640936686904, "grad_norm": 0.20180481672286987, "learning_rate": 9.190483098795882e-07, "loss": 0.0008, "step": 26405 }, { "epoch": 2.8627493495229834, "grad_norm": 0.30295151472091675, "learning_rate": 9.183229363122008e-07, "loss": 0.0061, "step": 26406 }, { "epoch": 2.8628577623590634, "grad_norm": 0.018574275076389313, "learning_rate": 9.175975627448136e-07, "loss": 0.0001, "step": 26407 }, { "epoch": 2.862966175195143, "grad_norm": 2.347625970840454, "learning_rate": 9.168721891774264e-07, "loss": 0.0425, "step": 26408 }, { "epoch": 2.863074588031223, "grad_norm": 6.866395473480225, "learning_rate": 9.161468156100392e-07, "loss": 0.0144, "step": 26409 }, { "epoch": 2.863183000867303, "grad_norm": 0.257906973361969, "learning_rate": 9.15421442042652e-07, "loss": 0.0032, "step": 26410 }, { "epoch": 2.8632914137033825, "grad_norm": 0.24593649804592133, "learning_rate": 9.146960684752648e-07, "loss": 0.005, "step": 26411 }, { "epoch": 2.863399826539462, "grad_norm": 0.09065951406955719, "learning_rate": 9.139706949078776e-07, "loss": 0.0005, "step": 26412 }, { "epoch": 2.863508239375542, "grad_norm": 0.3814290463924408, "learning_rate": 9.132453213404904e-07, "loss": 0.0122, "step": 26413 }, { "epoch": 2.863616652211622, "grad_norm": 0.2447022944688797, "learning_rate": 9.125199477731033e-07, "loss": 0.0019, "step": 26414 }, { "epoch": 2.8637250650477015, "grad_norm": 19.289281845092773, "learning_rate": 9.11794574205716e-07, "loss": 0.0174, "step": 26415 }, { "epoch": 2.8638334778837815, "grad_norm": 2.5127294063568115, "learning_rate": 9.110692006383288e-07, "loss": 0.0393, "step": 26416 }, { "epoch": 2.863941890719861, "grad_norm": 0.26747435331344604, "learning_rate": 9.103438270709417e-07, "loss": 0.0056, "step": 26417 }, { "epoch": 2.864050303555941, "grad_norm": 0.2917187809944153, "learning_rate": 9.096184535035545e-07, "loss": 0.0063, "step": 26418 }, { "epoch": 2.864158716392021, "grad_norm": 2.5388808250427246, "learning_rate": 9.088930799361673e-07, "loss": 0.0338, "step": 26419 }, { "epoch": 2.8642671292281006, "grad_norm": 1.4344017505645752, "learning_rate": 9.081677063687799e-07, "loss": 0.0071, "step": 26420 }, { "epoch": 2.8643755420641805, "grad_norm": 0.6537011861801147, "learning_rate": 9.074423328013927e-07, "loss": 0.013, "step": 26421 }, { "epoch": 2.86448395490026, "grad_norm": 0.13843928277492523, "learning_rate": 9.067169592340055e-07, "loss": 0.0006, "step": 26422 }, { "epoch": 2.86459236773634, "grad_norm": 0.4859420657157898, "learning_rate": 9.059915856666184e-07, "loss": 0.0073, "step": 26423 }, { "epoch": 2.8647007805724196, "grad_norm": 1.921973705291748, "learning_rate": 9.052662120992312e-07, "loss": 0.0249, "step": 26424 }, { "epoch": 2.8648091934084996, "grad_norm": 1.3624744415283203, "learning_rate": 9.045408385318439e-07, "loss": 0.0008, "step": 26425 }, { "epoch": 2.864917606244579, "grad_norm": 0.6275467276573181, "learning_rate": 9.038154649644568e-07, "loss": 0.0217, "step": 26426 }, { "epoch": 2.865026019080659, "grad_norm": 0.2807895243167877, "learning_rate": 9.030900913970696e-07, "loss": 0.0061, "step": 26427 }, { "epoch": 2.865134431916739, "grad_norm": 0.33596116304397583, "learning_rate": 9.023647178296824e-07, "loss": 0.0091, "step": 26428 }, { "epoch": 2.8652428447528187, "grad_norm": 10.378881454467773, "learning_rate": 9.016393442622952e-07, "loss": 0.0164, "step": 26429 }, { "epoch": 2.8653512575888986, "grad_norm": 0.033621251583099365, "learning_rate": 9.00913970694908e-07, "loss": 0.0002, "step": 26430 }, { "epoch": 2.865459670424978, "grad_norm": 1.329072117805481, "learning_rate": 9.001885971275208e-07, "loss": 0.0056, "step": 26431 }, { "epoch": 2.865568083261058, "grad_norm": 0.26451173424720764, "learning_rate": 8.994632235601336e-07, "loss": 0.0054, "step": 26432 }, { "epoch": 2.865676496097138, "grad_norm": 0.299782395362854, "learning_rate": 8.987378499927463e-07, "loss": 0.0096, "step": 26433 }, { "epoch": 2.8657849089332177, "grad_norm": 2.1014292240142822, "learning_rate": 8.98012476425359e-07, "loss": 0.019, "step": 26434 }, { "epoch": 2.8658933217692972, "grad_norm": 1.4305250644683838, "learning_rate": 8.972871028579719e-07, "loss": 0.034, "step": 26435 }, { "epoch": 2.866001734605377, "grad_norm": 0.7998849749565125, "learning_rate": 8.965617292905847e-07, "loss": 0.0113, "step": 26436 }, { "epoch": 2.866110147441457, "grad_norm": 2.453050374984741, "learning_rate": 8.958363557231975e-07, "loss": 0.0283, "step": 26437 }, { "epoch": 2.8662185602775367, "grad_norm": 0.7737055420875549, "learning_rate": 8.951109821558103e-07, "loss": 0.0144, "step": 26438 }, { "epoch": 2.8663269731136167, "grad_norm": 2.269442081451416, "learning_rate": 8.943856085884232e-07, "loss": 0.0173, "step": 26439 }, { "epoch": 2.8664353859496963, "grad_norm": 0.34096506237983704, "learning_rate": 8.936602350210359e-07, "loss": 0.0104, "step": 26440 }, { "epoch": 2.8665437987857763, "grad_norm": 4.383372783660889, "learning_rate": 8.929348614536487e-07, "loss": 0.0099, "step": 26441 }, { "epoch": 2.8666522116218562, "grad_norm": 17.118755340576172, "learning_rate": 8.922094878862615e-07, "loss": 0.0178, "step": 26442 }, { "epoch": 2.866760624457936, "grad_norm": 0.49758464097976685, "learning_rate": 8.914841143188744e-07, "loss": 0.0062, "step": 26443 }, { "epoch": 2.8668690372940158, "grad_norm": 3.700110912322998, "learning_rate": 8.907587407514872e-07, "loss": 0.0048, "step": 26444 }, { "epoch": 2.8669774501300953, "grad_norm": 2.993025064468384, "learning_rate": 8.900333671840999e-07, "loss": 0.0496, "step": 26445 }, { "epoch": 2.8670858629661753, "grad_norm": 0.4964566230773926, "learning_rate": 8.893079936167127e-07, "loss": 0.0084, "step": 26446 }, { "epoch": 2.867194275802255, "grad_norm": 0.5773710012435913, "learning_rate": 8.885826200493254e-07, "loss": 0.0179, "step": 26447 }, { "epoch": 2.867302688638335, "grad_norm": 4.041895389556885, "learning_rate": 8.878572464819383e-07, "loss": 0.035, "step": 26448 }, { "epoch": 2.8674111014744144, "grad_norm": 0.037673044949769974, "learning_rate": 8.87131872914551e-07, "loss": 0.0002, "step": 26449 }, { "epoch": 2.8675195143104943, "grad_norm": 2.9188504219055176, "learning_rate": 8.864064993471638e-07, "loss": 0.0127, "step": 26450 }, { "epoch": 2.8676279271465743, "grad_norm": 1.2523835897445679, "learning_rate": 8.856811257797766e-07, "loss": 0.002, "step": 26451 }, { "epoch": 2.867736339982654, "grad_norm": 0.3799697458744049, "learning_rate": 8.849557522123895e-07, "loss": 0.0102, "step": 26452 }, { "epoch": 2.867844752818734, "grad_norm": 0.1765667051076889, "learning_rate": 8.842303786450023e-07, "loss": 0.0008, "step": 26453 }, { "epoch": 2.8679531656548134, "grad_norm": 0.6381243467330933, "learning_rate": 8.83505005077615e-07, "loss": 0.0308, "step": 26454 }, { "epoch": 2.8680615784908934, "grad_norm": 0.12976056337356567, "learning_rate": 8.827796315102278e-07, "loss": 0.0007, "step": 26455 }, { "epoch": 2.8681699913269734, "grad_norm": 3.518901824951172, "learning_rate": 8.820542579428407e-07, "loss": 0.004, "step": 26456 }, { "epoch": 2.868278404163053, "grad_norm": 0.10207273811101913, "learning_rate": 8.813288843754535e-07, "loss": 0.0006, "step": 26457 }, { "epoch": 2.8683868169991324, "grad_norm": 0.4638838469982147, "learning_rate": 8.806035108080663e-07, "loss": 0.0069, "step": 26458 }, { "epoch": 2.8684952298352124, "grad_norm": 0.30769315361976624, "learning_rate": 8.79878137240679e-07, "loss": 0.0057, "step": 26459 }, { "epoch": 2.8686036426712924, "grad_norm": 0.19797135889530182, "learning_rate": 8.791527636732917e-07, "loss": 0.0027, "step": 26460 }, { "epoch": 2.868712055507372, "grad_norm": 0.5025455355644226, "learning_rate": 8.784273901059046e-07, "loss": 0.0116, "step": 26461 }, { "epoch": 2.868820468343452, "grad_norm": 1.8338088989257812, "learning_rate": 8.777020165385174e-07, "loss": 0.0049, "step": 26462 }, { "epoch": 2.8689288811795315, "grad_norm": 0.26936882734298706, "learning_rate": 8.769766429711302e-07, "loss": 0.0056, "step": 26463 }, { "epoch": 2.8690372940156115, "grad_norm": 1.5456721782684326, "learning_rate": 8.762512694037429e-07, "loss": 0.0544, "step": 26464 }, { "epoch": 2.8691457068516915, "grad_norm": 0.016201719641685486, "learning_rate": 8.755258958363558e-07, "loss": 0.0001, "step": 26465 }, { "epoch": 2.869254119687771, "grad_norm": 6.120363235473633, "learning_rate": 8.748005222689686e-07, "loss": 0.0115, "step": 26466 }, { "epoch": 2.869362532523851, "grad_norm": 0.13382962346076965, "learning_rate": 8.740751487015814e-07, "loss": 0.0006, "step": 26467 }, { "epoch": 2.8694709453599305, "grad_norm": 1.0164040327072144, "learning_rate": 8.733497751341942e-07, "loss": 0.0014, "step": 26468 }, { "epoch": 2.8695793581960105, "grad_norm": 0.17547139525413513, "learning_rate": 8.72624401566807e-07, "loss": 0.0013, "step": 26469 }, { "epoch": 2.86968777103209, "grad_norm": 4.866440773010254, "learning_rate": 8.718990279994198e-07, "loss": 0.0353, "step": 26470 }, { "epoch": 2.86979618386817, "grad_norm": 3.0774941444396973, "learning_rate": 8.711736544320326e-07, "loss": 0.016, "step": 26471 }, { "epoch": 2.8699045967042496, "grad_norm": 0.536572277545929, "learning_rate": 8.704482808646454e-07, "loss": 0.0114, "step": 26472 }, { "epoch": 2.8700130095403296, "grad_norm": 0.01193462498486042, "learning_rate": 8.697229072972583e-07, "loss": 0.0002, "step": 26473 }, { "epoch": 2.8701214223764095, "grad_norm": 0.2696419656276703, "learning_rate": 8.689975337298709e-07, "loss": 0.0056, "step": 26474 }, { "epoch": 2.870229835212489, "grad_norm": 0.25848668813705444, "learning_rate": 8.682721601624837e-07, "loss": 0.0052, "step": 26475 }, { "epoch": 2.870338248048569, "grad_norm": 0.08661430329084396, "learning_rate": 8.675467865950965e-07, "loss": 0.0003, "step": 26476 }, { "epoch": 2.8704466608846486, "grad_norm": 0.9576258063316345, "learning_rate": 8.668214130277093e-07, "loss": 0.0176, "step": 26477 }, { "epoch": 2.8705550737207286, "grad_norm": 7.866965293884277, "learning_rate": 8.660960394603222e-07, "loss": 0.0138, "step": 26478 }, { "epoch": 2.8706634865568086, "grad_norm": 0.4045913517475128, "learning_rate": 8.653706658929349e-07, "loss": 0.0101, "step": 26479 }, { "epoch": 2.870771899392888, "grad_norm": 6.898595333099365, "learning_rate": 8.646452923255477e-07, "loss": 0.0052, "step": 26480 }, { "epoch": 2.8708803122289677, "grad_norm": 0.7234919667243958, "learning_rate": 8.639199187581605e-07, "loss": 0.0164, "step": 26481 }, { "epoch": 2.8709887250650477, "grad_norm": 1.1727678775787354, "learning_rate": 8.631945451907734e-07, "loss": 0.049, "step": 26482 }, { "epoch": 2.8710971379011276, "grad_norm": 0.1961609274148941, "learning_rate": 8.624691716233862e-07, "loss": 0.0103, "step": 26483 }, { "epoch": 2.871205550737207, "grad_norm": 0.007990557700395584, "learning_rate": 8.617437980559989e-07, "loss": 0.0001, "step": 26484 }, { "epoch": 2.871313963573287, "grad_norm": 3.2991838455200195, "learning_rate": 8.610184244886117e-07, "loss": 0.0322, "step": 26485 }, { "epoch": 2.8714223764093667, "grad_norm": 2.1673920154571533, "learning_rate": 8.602930509212246e-07, "loss": 0.033, "step": 26486 }, { "epoch": 2.8715307892454467, "grad_norm": 0.565194845199585, "learning_rate": 8.595676773538374e-07, "loss": 0.0075, "step": 26487 }, { "epoch": 2.8716392020815267, "grad_norm": 3.456860303878784, "learning_rate": 8.5884230378645e-07, "loss": 0.0224, "step": 26488 }, { "epoch": 2.871747614917606, "grad_norm": 1.5889992713928223, "learning_rate": 8.581169302190628e-07, "loss": 0.0123, "step": 26489 }, { "epoch": 2.871856027753686, "grad_norm": 1.2524806261062622, "learning_rate": 8.573915566516756e-07, "loss": 0.0457, "step": 26490 }, { "epoch": 2.8719644405897657, "grad_norm": 0.31709402799606323, "learning_rate": 8.566661830842885e-07, "loss": 0.0024, "step": 26491 }, { "epoch": 2.8720728534258457, "grad_norm": 0.6931890845298767, "learning_rate": 8.559408095169013e-07, "loss": 0.0024, "step": 26492 }, { "epoch": 2.8721812662619253, "grad_norm": 0.2693997025489807, "learning_rate": 8.55215435949514e-07, "loss": 0.0057, "step": 26493 }, { "epoch": 2.8722896790980053, "grad_norm": 0.020753072574734688, "learning_rate": 8.544900623821268e-07, "loss": 0.0002, "step": 26494 }, { "epoch": 2.872398091934085, "grad_norm": 5.230224132537842, "learning_rate": 8.537646888147397e-07, "loss": 0.04, "step": 26495 }, { "epoch": 2.872506504770165, "grad_norm": 9.44154167175293, "learning_rate": 8.530393152473525e-07, "loss": 0.0263, "step": 26496 }, { "epoch": 2.8726149176062448, "grad_norm": 0.7563526630401611, "learning_rate": 8.523139416799653e-07, "loss": 0.0188, "step": 26497 }, { "epoch": 2.8727233304423243, "grad_norm": 0.1417699009180069, "learning_rate": 8.51588568112578e-07, "loss": 0.0004, "step": 26498 }, { "epoch": 2.8728317432784043, "grad_norm": 2.4215757846832275, "learning_rate": 8.508631945451909e-07, "loss": 0.0259, "step": 26499 }, { "epoch": 2.872940156114484, "grad_norm": 0.828424334526062, "learning_rate": 8.501378209778037e-07, "loss": 0.024, "step": 26500 }, { "epoch": 2.873048568950564, "grad_norm": 0.48330551385879517, "learning_rate": 8.494124474104164e-07, "loss": 0.009, "step": 26501 }, { "epoch": 2.873156981786644, "grad_norm": 0.01271114032715559, "learning_rate": 8.486870738430292e-07, "loss": 0.0001, "step": 26502 }, { "epoch": 2.8732653946227233, "grad_norm": 6.4779372215271, "learning_rate": 8.479617002756419e-07, "loss": 0.0173, "step": 26503 }, { "epoch": 2.873373807458803, "grad_norm": 5.822052955627441, "learning_rate": 8.472363267082548e-07, "loss": 0.0285, "step": 26504 }, { "epoch": 2.873482220294883, "grad_norm": 0.02648274041712284, "learning_rate": 8.465109531408676e-07, "loss": 0.0003, "step": 26505 }, { "epoch": 2.873590633130963, "grad_norm": 5.183481216430664, "learning_rate": 8.457855795734804e-07, "loss": 0.0379, "step": 26506 }, { "epoch": 2.8736990459670424, "grad_norm": 2.224331855773926, "learning_rate": 8.450602060060932e-07, "loss": 0.0687, "step": 26507 }, { "epoch": 2.8738074588031224, "grad_norm": 0.6606422066688538, "learning_rate": 8.44334832438706e-07, "loss": 0.0113, "step": 26508 }, { "epoch": 2.873915871639202, "grad_norm": 0.05949496105313301, "learning_rate": 8.436094588713188e-07, "loss": 0.0002, "step": 26509 }, { "epoch": 2.874024284475282, "grad_norm": 4.375072002410889, "learning_rate": 8.428840853039316e-07, "loss": 0.0448, "step": 26510 }, { "epoch": 2.874132697311362, "grad_norm": 6.463255882263184, "learning_rate": 8.421587117365444e-07, "loss": 0.0404, "step": 26511 }, { "epoch": 2.8742411101474414, "grad_norm": 2.449949026107788, "learning_rate": 8.414333381691573e-07, "loss": 0.0028, "step": 26512 }, { "epoch": 2.874349522983521, "grad_norm": 0.029788527637720108, "learning_rate": 8.4070796460177e-07, "loss": 0.0002, "step": 26513 }, { "epoch": 2.874457935819601, "grad_norm": 0.025381838902831078, "learning_rate": 8.399825910343828e-07, "loss": 0.0002, "step": 26514 }, { "epoch": 2.874566348655681, "grad_norm": 0.03475675731897354, "learning_rate": 8.392572174669955e-07, "loss": 0.0002, "step": 26515 }, { "epoch": 2.8746747614917605, "grad_norm": 2.9739816188812256, "learning_rate": 8.385318438996083e-07, "loss": 0.0339, "step": 26516 }, { "epoch": 2.8747831743278405, "grad_norm": 0.6383759379386902, "learning_rate": 8.378064703322212e-07, "loss": 0.0158, "step": 26517 }, { "epoch": 2.87489158716392, "grad_norm": 5.134884357452393, "learning_rate": 8.370810967648339e-07, "loss": 0.0404, "step": 26518 }, { "epoch": 2.875, "grad_norm": 0.10164103657007217, "learning_rate": 8.363557231974467e-07, "loss": 0.0006, "step": 26519 }, { "epoch": 2.87510841283608, "grad_norm": 0.020075641572475433, "learning_rate": 8.356303496300595e-07, "loss": 0.0002, "step": 26520 }, { "epoch": 2.8752168256721595, "grad_norm": 0.044603146612644196, "learning_rate": 8.349049760626724e-07, "loss": 0.0003, "step": 26521 }, { "epoch": 2.8753252385082395, "grad_norm": 0.47325149178504944, "learning_rate": 8.341796024952852e-07, "loss": 0.0241, "step": 26522 }, { "epoch": 2.875433651344319, "grad_norm": 0.3797799050807953, "learning_rate": 8.334542289278979e-07, "loss": 0.01, "step": 26523 }, { "epoch": 2.875542064180399, "grad_norm": 0.2591218650341034, "learning_rate": 8.327288553605107e-07, "loss": 0.0163, "step": 26524 }, { "epoch": 2.875650477016479, "grad_norm": 0.30213290452957153, "learning_rate": 8.320034817931236e-07, "loss": 0.0097, "step": 26525 }, { "epoch": 2.8757588898525586, "grad_norm": 0.3998976945877075, "learning_rate": 8.312781082257364e-07, "loss": 0.0135, "step": 26526 }, { "epoch": 2.875867302688638, "grad_norm": 0.2529999613761902, "learning_rate": 8.305527346583492e-07, "loss": 0.0053, "step": 26527 }, { "epoch": 2.875975715524718, "grad_norm": 0.31266677379608154, "learning_rate": 8.298273610909618e-07, "loss": 0.0036, "step": 26528 }, { "epoch": 2.876084128360798, "grad_norm": 5.879187107086182, "learning_rate": 8.291019875235746e-07, "loss": 0.0149, "step": 26529 }, { "epoch": 2.8761925411968776, "grad_norm": 0.32691630721092224, "learning_rate": 8.283766139561875e-07, "loss": 0.0008, "step": 26530 }, { "epoch": 2.8763009540329576, "grad_norm": 0.8879266381263733, "learning_rate": 8.276512403888003e-07, "loss": 0.0259, "step": 26531 }, { "epoch": 2.876409366869037, "grad_norm": 2.2420194149017334, "learning_rate": 8.26925866821413e-07, "loss": 0.0217, "step": 26532 }, { "epoch": 2.876517779705117, "grad_norm": 0.39556559920310974, "learning_rate": 8.262004932540258e-07, "loss": 0.014, "step": 26533 }, { "epoch": 2.876626192541197, "grad_norm": 0.2874811291694641, "learning_rate": 8.254751196866387e-07, "loss": 0.0028, "step": 26534 }, { "epoch": 2.8767346053772767, "grad_norm": 4.461484432220459, "learning_rate": 8.247497461192515e-07, "loss": 0.0186, "step": 26535 }, { "epoch": 2.876843018213356, "grad_norm": 1.7430133819580078, "learning_rate": 8.240243725518643e-07, "loss": 0.0169, "step": 26536 }, { "epoch": 2.876951431049436, "grad_norm": 2.1656925678253174, "learning_rate": 8.23298998984477e-07, "loss": 0.0466, "step": 26537 }, { "epoch": 2.877059843885516, "grad_norm": 0.25157445669174194, "learning_rate": 8.225736254170899e-07, "loss": 0.0047, "step": 26538 }, { "epoch": 2.8771682567215957, "grad_norm": 0.3360652029514313, "learning_rate": 8.218482518497027e-07, "loss": 0.0018, "step": 26539 }, { "epoch": 2.8772766695576757, "grad_norm": 2.62516713142395, "learning_rate": 8.211228782823155e-07, "loss": 0.0101, "step": 26540 }, { "epoch": 2.8773850823937552, "grad_norm": 0.273762583732605, "learning_rate": 8.203975047149283e-07, "loss": 0.0054, "step": 26541 }, { "epoch": 2.877493495229835, "grad_norm": 0.058758992701768875, "learning_rate": 8.196721311475409e-07, "loss": 0.0003, "step": 26542 }, { "epoch": 2.877601908065915, "grad_norm": 1.0763473510742188, "learning_rate": 8.189467575801538e-07, "loss": 0.0051, "step": 26543 }, { "epoch": 2.8777103209019947, "grad_norm": 0.2730772793292999, "learning_rate": 8.182213840127666e-07, "loss": 0.0055, "step": 26544 }, { "epoch": 2.8778187337380747, "grad_norm": 0.050455547869205475, "learning_rate": 8.174960104453794e-07, "loss": 0.0004, "step": 26545 }, { "epoch": 2.8779271465741543, "grad_norm": 0.10902534425258636, "learning_rate": 8.167706368779922e-07, "loss": 0.0004, "step": 26546 }, { "epoch": 2.8780355594102343, "grad_norm": 0.3821215033531189, "learning_rate": 8.16045263310605e-07, "loss": 0.0068, "step": 26547 }, { "epoch": 2.878143972246314, "grad_norm": 13.216461181640625, "learning_rate": 8.153198897432178e-07, "loss": 0.0239, "step": 26548 }, { "epoch": 2.878252385082394, "grad_norm": 1.53477942943573, "learning_rate": 8.145945161758306e-07, "loss": 0.0022, "step": 26549 }, { "epoch": 2.8783607979184733, "grad_norm": 0.8457403182983398, "learning_rate": 8.138691426084434e-07, "loss": 0.0204, "step": 26550 }, { "epoch": 2.8784692107545533, "grad_norm": 0.17759515345096588, "learning_rate": 8.131437690410563e-07, "loss": 0.0004, "step": 26551 }, { "epoch": 2.8785776235906333, "grad_norm": 0.01877489872276783, "learning_rate": 8.12418395473669e-07, "loss": 0.0002, "step": 26552 }, { "epoch": 2.878686036426713, "grad_norm": 0.4480845630168915, "learning_rate": 8.116930219062818e-07, "loss": 0.0156, "step": 26553 }, { "epoch": 2.878794449262793, "grad_norm": 1.8006647825241089, "learning_rate": 8.109676483388946e-07, "loss": 0.0412, "step": 26554 }, { "epoch": 2.8789028620988724, "grad_norm": 5.867341041564941, "learning_rate": 8.102422747715073e-07, "loss": 0.0177, "step": 26555 }, { "epoch": 2.8790112749349523, "grad_norm": 1.8474905490875244, "learning_rate": 8.095169012041202e-07, "loss": 0.0175, "step": 26556 }, { "epoch": 2.8791196877710323, "grad_norm": 6.16566276550293, "learning_rate": 8.087915276367329e-07, "loss": 0.0215, "step": 26557 }, { "epoch": 2.879228100607112, "grad_norm": 10.676828384399414, "learning_rate": 8.080661540693457e-07, "loss": 0.018, "step": 26558 }, { "epoch": 2.8793365134431914, "grad_norm": 0.20735561847686768, "learning_rate": 8.073407805019585e-07, "loss": 0.003, "step": 26559 }, { "epoch": 2.8794449262792714, "grad_norm": 0.05281442776322365, "learning_rate": 8.066154069345714e-07, "loss": 0.0003, "step": 26560 }, { "epoch": 2.8795533391153514, "grad_norm": 3.7204291820526123, "learning_rate": 8.058900333671842e-07, "loss": 0.0168, "step": 26561 }, { "epoch": 2.879661751951431, "grad_norm": 0.12843436002731323, "learning_rate": 8.051646597997969e-07, "loss": 0.0008, "step": 26562 }, { "epoch": 2.879770164787511, "grad_norm": 2.8933777809143066, "learning_rate": 8.044392862324097e-07, "loss": 0.014, "step": 26563 }, { "epoch": 2.8798785776235905, "grad_norm": 0.7039316296577454, "learning_rate": 8.037139126650226e-07, "loss": 0.0014, "step": 26564 }, { "epoch": 2.8799869904596704, "grad_norm": 3.139475107192993, "learning_rate": 8.029885390976354e-07, "loss": 0.0152, "step": 26565 }, { "epoch": 2.8800954032957504, "grad_norm": 0.024763768538832664, "learning_rate": 8.022631655302482e-07, "loss": 0.0003, "step": 26566 }, { "epoch": 2.88020381613183, "grad_norm": 0.020491408184170723, "learning_rate": 8.01537791962861e-07, "loss": 0.0002, "step": 26567 }, { "epoch": 2.88031222896791, "grad_norm": 0.7746114730834961, "learning_rate": 8.008124183954738e-07, "loss": 0.0134, "step": 26568 }, { "epoch": 2.8804206418039895, "grad_norm": 1.7678114175796509, "learning_rate": 8.000870448280865e-07, "loss": 0.0105, "step": 26569 }, { "epoch": 2.8805290546400695, "grad_norm": 1.959883689880371, "learning_rate": 7.993616712606993e-07, "loss": 0.0592, "step": 26570 }, { "epoch": 2.880637467476149, "grad_norm": 2.1705689430236816, "learning_rate": 7.986362976933121e-07, "loss": 0.0243, "step": 26571 }, { "epoch": 2.880745880312229, "grad_norm": 0.44335347414016724, "learning_rate": 7.979109241259248e-07, "loss": 0.0008, "step": 26572 }, { "epoch": 2.8808542931483085, "grad_norm": 1.6438210010528564, "learning_rate": 7.971855505585377e-07, "loss": 0.0258, "step": 26573 }, { "epoch": 2.8809627059843885, "grad_norm": 1.4255601167678833, "learning_rate": 7.964601769911505e-07, "loss": 0.009, "step": 26574 }, { "epoch": 2.8810711188204685, "grad_norm": 0.4313328266143799, "learning_rate": 7.957348034237633e-07, "loss": 0.0169, "step": 26575 }, { "epoch": 2.881179531656548, "grad_norm": 0.3659345805644989, "learning_rate": 7.950094298563762e-07, "loss": 0.0071, "step": 26576 }, { "epoch": 2.881287944492628, "grad_norm": 0.5795327425003052, "learning_rate": 7.942840562889889e-07, "loss": 0.0262, "step": 26577 }, { "epoch": 2.8813963573287076, "grad_norm": 0.5116375088691711, "learning_rate": 7.935586827216017e-07, "loss": 0.0068, "step": 26578 }, { "epoch": 2.8815047701647876, "grad_norm": 0.31846949458122253, "learning_rate": 7.928333091542145e-07, "loss": 0.0067, "step": 26579 }, { "epoch": 2.8816131830008676, "grad_norm": 0.16926604509353638, "learning_rate": 7.921079355868274e-07, "loss": 0.0006, "step": 26580 }, { "epoch": 2.881721595836947, "grad_norm": 0.531649112701416, "learning_rate": 7.913825620194402e-07, "loss": 0.0106, "step": 26581 }, { "epoch": 2.8818300086730266, "grad_norm": 6.842693328857422, "learning_rate": 7.906571884520528e-07, "loss": 0.0488, "step": 26582 }, { "epoch": 2.8819384215091066, "grad_norm": 3.2237861156463623, "learning_rate": 7.899318148846656e-07, "loss": 0.0305, "step": 26583 }, { "epoch": 2.8820468343451866, "grad_norm": 4.993194580078125, "learning_rate": 7.892064413172784e-07, "loss": 0.0252, "step": 26584 }, { "epoch": 2.882155247181266, "grad_norm": 0.4585001766681671, "learning_rate": 7.884810677498912e-07, "loss": 0.0011, "step": 26585 }, { "epoch": 2.882263660017346, "grad_norm": 1.7127991914749146, "learning_rate": 7.877556941825041e-07, "loss": 0.0047, "step": 26586 }, { "epoch": 2.8823720728534257, "grad_norm": 0.6782295107841492, "learning_rate": 7.870303206151168e-07, "loss": 0.0087, "step": 26587 }, { "epoch": 2.8824804856895057, "grad_norm": 0.5349688529968262, "learning_rate": 7.863049470477296e-07, "loss": 0.0045, "step": 26588 }, { "epoch": 2.8825888985255856, "grad_norm": 0.2725645899772644, "learning_rate": 7.855795734803425e-07, "loss": 0.0053, "step": 26589 }, { "epoch": 2.882697311361665, "grad_norm": 1.7121368646621704, "learning_rate": 7.848541999129553e-07, "loss": 0.0134, "step": 26590 }, { "epoch": 2.882805724197745, "grad_norm": 3.8148701190948486, "learning_rate": 7.84128826345568e-07, "loss": 0.0571, "step": 26591 }, { "epoch": 2.8829141370338247, "grad_norm": 0.729526698589325, "learning_rate": 7.834034527781808e-07, "loss": 0.0024, "step": 26592 }, { "epoch": 2.8830225498699047, "grad_norm": 0.756659746170044, "learning_rate": 7.826780792107937e-07, "loss": 0.017, "step": 26593 }, { "epoch": 2.8831309627059842, "grad_norm": 0.296075701713562, "learning_rate": 7.819527056434065e-07, "loss": 0.0094, "step": 26594 }, { "epoch": 2.883239375542064, "grad_norm": 0.593928873538971, "learning_rate": 7.812273320760193e-07, "loss": 0.0078, "step": 26595 }, { "epoch": 2.8833477883781438, "grad_norm": 2.66312837600708, "learning_rate": 7.805019585086319e-07, "loss": 0.0606, "step": 26596 }, { "epoch": 2.8834562012142237, "grad_norm": 0.3485087752342224, "learning_rate": 7.797765849412447e-07, "loss": 0.0115, "step": 26597 }, { "epoch": 2.8835646140503037, "grad_norm": 0.26049554347991943, "learning_rate": 7.790512113738576e-07, "loss": 0.0051, "step": 26598 }, { "epoch": 2.8836730268863833, "grad_norm": 0.36340728402137756, "learning_rate": 7.783258378064704e-07, "loss": 0.0014, "step": 26599 }, { "epoch": 2.8837814397224633, "grad_norm": 0.4503013789653778, "learning_rate": 7.776004642390832e-07, "loss": 0.0116, "step": 26600 }, { "epoch": 2.883889852558543, "grad_norm": 0.5432088375091553, "learning_rate": 7.768750906716959e-07, "loss": 0.0077, "step": 26601 }, { "epoch": 2.883998265394623, "grad_norm": 3.223132371902466, "learning_rate": 7.761497171043088e-07, "loss": 0.0075, "step": 26602 }, { "epoch": 2.8841066782307028, "grad_norm": 0.012571732513606548, "learning_rate": 7.754243435369216e-07, "loss": 0.0002, "step": 26603 }, { "epoch": 2.8842150910667823, "grad_norm": 0.0024337044451385736, "learning_rate": 7.746989699695344e-07, "loss": 0.0, "step": 26604 }, { "epoch": 2.884323503902862, "grad_norm": 2.859513521194458, "learning_rate": 7.739735964021472e-07, "loss": 0.0147, "step": 26605 }, { "epoch": 2.884431916738942, "grad_norm": 0.22105489671230316, "learning_rate": 7.7324822283476e-07, "loss": 0.0006, "step": 26606 }, { "epoch": 2.884540329575022, "grad_norm": 7.717920303344727, "learning_rate": 7.725228492673728e-07, "loss": 0.0198, "step": 26607 }, { "epoch": 2.8846487424111014, "grad_norm": 0.2590269148349762, "learning_rate": 7.717974756999856e-07, "loss": 0.0056, "step": 26608 }, { "epoch": 2.8847571552471813, "grad_norm": 0.1015593484044075, "learning_rate": 7.710721021325983e-07, "loss": 0.0004, "step": 26609 }, { "epoch": 2.884865568083261, "grad_norm": 0.18647481501102448, "learning_rate": 7.703467285652111e-07, "loss": 0.001, "step": 26610 }, { "epoch": 2.884973980919341, "grad_norm": 0.396388977766037, "learning_rate": 7.696213549978239e-07, "loss": 0.0392, "step": 26611 }, { "epoch": 2.885082393755421, "grad_norm": 0.2590457797050476, "learning_rate": 7.688959814304367e-07, "loss": 0.0025, "step": 26612 }, { "epoch": 2.8851908065915004, "grad_norm": 0.019808823242783546, "learning_rate": 7.681706078630495e-07, "loss": 0.0002, "step": 26613 }, { "epoch": 2.8852992194275804, "grad_norm": 1.7163586616516113, "learning_rate": 7.674452342956623e-07, "loss": 0.0111, "step": 26614 }, { "epoch": 2.88540763226366, "grad_norm": 0.3247603178024292, "learning_rate": 7.667198607282752e-07, "loss": 0.0098, "step": 26615 }, { "epoch": 2.88551604509974, "grad_norm": 0.01881432719528675, "learning_rate": 7.659944871608879e-07, "loss": 0.0001, "step": 26616 }, { "epoch": 2.8856244579358195, "grad_norm": 0.5370020270347595, "learning_rate": 7.652691135935007e-07, "loss": 0.0269, "step": 26617 }, { "epoch": 2.8857328707718994, "grad_norm": 0.41194745898246765, "learning_rate": 7.645437400261135e-07, "loss": 0.0248, "step": 26618 }, { "epoch": 2.885841283607979, "grad_norm": 0.3295844495296478, "learning_rate": 7.638183664587264e-07, "loss": 0.0068, "step": 26619 }, { "epoch": 2.885949696444059, "grad_norm": 0.03258275240659714, "learning_rate": 7.630929928913392e-07, "loss": 0.0002, "step": 26620 }, { "epoch": 2.886058109280139, "grad_norm": 8.778557777404785, "learning_rate": 7.623676193239519e-07, "loss": 0.0473, "step": 26621 }, { "epoch": 2.8861665221162185, "grad_norm": 0.03625031188130379, "learning_rate": 7.616422457565647e-07, "loss": 0.0003, "step": 26622 }, { "epoch": 2.8862749349522985, "grad_norm": 0.03664959594607353, "learning_rate": 7.609168721891774e-07, "loss": 0.0002, "step": 26623 }, { "epoch": 2.886383347788378, "grad_norm": 0.275655597448349, "learning_rate": 7.601914986217903e-07, "loss": 0.0057, "step": 26624 }, { "epoch": 2.886491760624458, "grad_norm": 0.06369320303201675, "learning_rate": 7.594661250544031e-07, "loss": 0.0003, "step": 26625 }, { "epoch": 2.886600173460538, "grad_norm": 0.5151777863502502, "learning_rate": 7.587407514870158e-07, "loss": 0.0098, "step": 26626 }, { "epoch": 2.8867085862966175, "grad_norm": 2.9659478664398193, "learning_rate": 7.580153779196286e-07, "loss": 0.0245, "step": 26627 }, { "epoch": 2.886816999132697, "grad_norm": 0.16553446650505066, "learning_rate": 7.572900043522415e-07, "loss": 0.0012, "step": 26628 }, { "epoch": 2.886925411968777, "grad_norm": 0.11157019436359406, "learning_rate": 7.565646307848543e-07, "loss": 0.0008, "step": 26629 }, { "epoch": 2.887033824804857, "grad_norm": 0.7589637637138367, "learning_rate": 7.558392572174671e-07, "loss": 0.0224, "step": 26630 }, { "epoch": 2.8871422376409366, "grad_norm": 0.10428358614444733, "learning_rate": 7.551138836500798e-07, "loss": 0.0005, "step": 26631 }, { "epoch": 2.8872506504770166, "grad_norm": 0.4783613085746765, "learning_rate": 7.543885100826927e-07, "loss": 0.0092, "step": 26632 }, { "epoch": 2.887359063313096, "grad_norm": 5.7320427894592285, "learning_rate": 7.536631365153055e-07, "loss": 0.0297, "step": 26633 }, { "epoch": 2.887467476149176, "grad_norm": 0.2395680844783783, "learning_rate": 7.529377629479183e-07, "loss": 0.0158, "step": 26634 }, { "epoch": 2.887575888985256, "grad_norm": 0.4246188700199127, "learning_rate": 7.522123893805311e-07, "loss": 0.0127, "step": 26635 }, { "epoch": 2.8876843018213356, "grad_norm": 0.015048532746732235, "learning_rate": 7.514870158131439e-07, "loss": 0.0001, "step": 26636 }, { "epoch": 2.887792714657415, "grad_norm": 0.5237489938735962, "learning_rate": 7.507616422457566e-07, "loss": 0.0124, "step": 26637 }, { "epoch": 2.887901127493495, "grad_norm": 0.8434062004089355, "learning_rate": 7.500362686783694e-07, "loss": 0.0187, "step": 26638 }, { "epoch": 2.888009540329575, "grad_norm": 0.12935076653957367, "learning_rate": 7.493108951109822e-07, "loss": 0.0008, "step": 26639 }, { "epoch": 2.8881179531656547, "grad_norm": 0.6455774307250977, "learning_rate": 7.48585521543595e-07, "loss": 0.0104, "step": 26640 }, { "epoch": 2.8882263660017347, "grad_norm": 0.30152878165245056, "learning_rate": 7.478601479762078e-07, "loss": 0.0005, "step": 26641 }, { "epoch": 2.888334778837814, "grad_norm": 3.429365873336792, "learning_rate": 7.471347744088206e-07, "loss": 0.0336, "step": 26642 }, { "epoch": 2.888443191673894, "grad_norm": 0.6856226325035095, "learning_rate": 7.464094008414334e-07, "loss": 0.0024, "step": 26643 }, { "epoch": 2.888551604509974, "grad_norm": 1.2355622053146362, "learning_rate": 7.456840272740462e-07, "loss": 0.0173, "step": 26644 }, { "epoch": 2.8886600173460537, "grad_norm": 0.6729373335838318, "learning_rate": 7.449586537066591e-07, "loss": 0.0015, "step": 26645 }, { "epoch": 2.8887684301821337, "grad_norm": 0.011565795168280602, "learning_rate": 7.442332801392718e-07, "loss": 0.0001, "step": 26646 }, { "epoch": 2.8888768430182132, "grad_norm": 0.10305266827344894, "learning_rate": 7.435079065718846e-07, "loss": 0.0007, "step": 26647 }, { "epoch": 2.888985255854293, "grad_norm": 0.7304809093475342, "learning_rate": 7.427825330044974e-07, "loss": 0.0238, "step": 26648 }, { "epoch": 2.889093668690373, "grad_norm": 0.06169062480330467, "learning_rate": 7.420571594371103e-07, "loss": 0.0004, "step": 26649 }, { "epoch": 2.8892020815264527, "grad_norm": 0.33385175466537476, "learning_rate": 7.413317858697229e-07, "loss": 0.0114, "step": 26650 }, { "epoch": 2.8893104943625323, "grad_norm": 0.031742699444293976, "learning_rate": 7.406064123023357e-07, "loss": 0.0002, "step": 26651 }, { "epoch": 2.8894189071986123, "grad_norm": 5.489563465118408, "learning_rate": 7.398810387349485e-07, "loss": 0.0405, "step": 26652 }, { "epoch": 2.8895273200346923, "grad_norm": 0.3666776716709137, "learning_rate": 7.391556651675613e-07, "loss": 0.0112, "step": 26653 }, { "epoch": 2.889635732870772, "grad_norm": 0.4130234718322754, "learning_rate": 7.384302916001742e-07, "loss": 0.007, "step": 26654 }, { "epoch": 2.889744145706852, "grad_norm": 0.2842872738838196, "learning_rate": 7.377049180327869e-07, "loss": 0.0067, "step": 26655 }, { "epoch": 2.8898525585429313, "grad_norm": 4.843404769897461, "learning_rate": 7.369795444653997e-07, "loss": 0.0231, "step": 26656 }, { "epoch": 2.8899609713790113, "grad_norm": 0.08330411463975906, "learning_rate": 7.362541708980125e-07, "loss": 0.0003, "step": 26657 }, { "epoch": 2.8900693842150913, "grad_norm": 0.602274477481842, "learning_rate": 7.355287973306254e-07, "loss": 0.0115, "step": 26658 }, { "epoch": 2.890177797051171, "grad_norm": 0.03003743849694729, "learning_rate": 7.348034237632382e-07, "loss": 0.0003, "step": 26659 }, { "epoch": 2.8902862098872504, "grad_norm": 0.26377490162849426, "learning_rate": 7.340780501958509e-07, "loss": 0.0046, "step": 26660 }, { "epoch": 2.8903946227233304, "grad_norm": 0.28504055738449097, "learning_rate": 7.333526766284637e-07, "loss": 0.0029, "step": 26661 }, { "epoch": 2.8905030355594103, "grad_norm": 0.02491183578968048, "learning_rate": 7.326273030610766e-07, "loss": 0.0002, "step": 26662 }, { "epoch": 2.89061144839549, "grad_norm": 5.198561191558838, "learning_rate": 7.319019294936894e-07, "loss": 0.0083, "step": 26663 }, { "epoch": 2.89071986123157, "grad_norm": 3.654458522796631, "learning_rate": 7.311765559263021e-07, "loss": 0.0127, "step": 26664 }, { "epoch": 2.8908282740676494, "grad_norm": 1.9770352840423584, "learning_rate": 7.304511823589148e-07, "loss": 0.0255, "step": 26665 }, { "epoch": 2.8909366869037294, "grad_norm": 0.9391716122627258, "learning_rate": 7.297258087915276e-07, "loss": 0.0274, "step": 26666 }, { "epoch": 2.8910450997398094, "grad_norm": 0.45003461837768555, "learning_rate": 7.290004352241405e-07, "loss": 0.0121, "step": 26667 }, { "epoch": 2.891153512575889, "grad_norm": 0.37373363971710205, "learning_rate": 7.282750616567533e-07, "loss": 0.0092, "step": 26668 }, { "epoch": 2.891261925411969, "grad_norm": 0.011387423612177372, "learning_rate": 7.275496880893661e-07, "loss": 0.0001, "step": 26669 }, { "epoch": 2.8913703382480485, "grad_norm": 27.351022720336914, "learning_rate": 7.268243145219788e-07, "loss": 0.013, "step": 26670 }, { "epoch": 2.8914787510841284, "grad_norm": 0.27043968439102173, "learning_rate": 7.260989409545917e-07, "loss": 0.005, "step": 26671 }, { "epoch": 2.8915871639202084, "grad_norm": 4.549018859863281, "learning_rate": 7.253735673872045e-07, "loss": 0.0274, "step": 26672 }, { "epoch": 2.891695576756288, "grad_norm": 0.7169545888900757, "learning_rate": 7.246481938198173e-07, "loss": 0.0143, "step": 26673 }, { "epoch": 2.8918039895923675, "grad_norm": 0.3774793744087219, "learning_rate": 7.239228202524301e-07, "loss": 0.0103, "step": 26674 }, { "epoch": 2.8919124024284475, "grad_norm": 0.5245838761329651, "learning_rate": 7.231974466850429e-07, "loss": 0.0066, "step": 26675 }, { "epoch": 2.8920208152645275, "grad_norm": 0.03043449856340885, "learning_rate": 7.224720731176557e-07, "loss": 0.0002, "step": 26676 }, { "epoch": 2.892129228100607, "grad_norm": 0.5127766728401184, "learning_rate": 7.217466995502684e-07, "loss": 0.0194, "step": 26677 }, { "epoch": 2.892237640936687, "grad_norm": 0.6489497423171997, "learning_rate": 7.210213259828812e-07, "loss": 0.0232, "step": 26678 }, { "epoch": 2.8923460537727665, "grad_norm": 3.917125940322876, "learning_rate": 7.20295952415494e-07, "loss": 0.0105, "step": 26679 }, { "epoch": 2.8924544666088465, "grad_norm": 6.441152572631836, "learning_rate": 7.195705788481068e-07, "loss": 0.0477, "step": 26680 }, { "epoch": 2.8925628794449265, "grad_norm": 0.38942763209342957, "learning_rate": 7.188452052807196e-07, "loss": 0.0095, "step": 26681 }, { "epoch": 2.892671292281006, "grad_norm": 4.232438087463379, "learning_rate": 7.181198317133324e-07, "loss": 0.1156, "step": 26682 }, { "epoch": 2.8927797051170856, "grad_norm": 0.3361518085002899, "learning_rate": 7.173944581459452e-07, "loss": 0.0091, "step": 26683 }, { "epoch": 2.8928881179531656, "grad_norm": 0.5284318923950195, "learning_rate": 7.166690845785581e-07, "loss": 0.0157, "step": 26684 }, { "epoch": 2.8929965307892456, "grad_norm": 8.97850227355957, "learning_rate": 7.159437110111708e-07, "loss": 0.0164, "step": 26685 }, { "epoch": 2.893104943625325, "grad_norm": 0.07458318024873734, "learning_rate": 7.152183374437836e-07, "loss": 0.0006, "step": 26686 }, { "epoch": 2.893213356461405, "grad_norm": 0.5376396775245667, "learning_rate": 7.144929638763964e-07, "loss": 0.0138, "step": 26687 }, { "epoch": 2.8933217692974846, "grad_norm": 0.774830162525177, "learning_rate": 7.137675903090093e-07, "loss": 0.0026, "step": 26688 }, { "epoch": 2.8934301821335646, "grad_norm": 2.093132257461548, "learning_rate": 7.130422167416221e-07, "loss": 0.0015, "step": 26689 }, { "epoch": 2.8935385949696446, "grad_norm": 0.3901956081390381, "learning_rate": 7.123168431742348e-07, "loss": 0.0034, "step": 26690 }, { "epoch": 2.893647007805724, "grad_norm": 7.927067279815674, "learning_rate": 7.115914696068475e-07, "loss": 0.0203, "step": 26691 }, { "epoch": 2.893755420641804, "grad_norm": 0.38653838634490967, "learning_rate": 7.108660960394603e-07, "loss": 0.0066, "step": 26692 }, { "epoch": 2.8938638334778837, "grad_norm": 2.933142900466919, "learning_rate": 7.101407224720732e-07, "loss": 0.0122, "step": 26693 }, { "epoch": 2.8939722463139637, "grad_norm": 0.02954084798693657, "learning_rate": 7.09415348904686e-07, "loss": 0.0002, "step": 26694 }, { "epoch": 2.894080659150043, "grad_norm": 1.9574851989746094, "learning_rate": 7.086899753372987e-07, "loss": 0.0034, "step": 26695 }, { "epoch": 2.894189071986123, "grad_norm": 1.0184890031814575, "learning_rate": 7.079646017699115e-07, "loss": 0.0124, "step": 26696 }, { "epoch": 2.8942974848222027, "grad_norm": 0.023495087400078773, "learning_rate": 7.072392282025244e-07, "loss": 0.0002, "step": 26697 }, { "epoch": 2.8944058976582827, "grad_norm": 8.75085735321045, "learning_rate": 7.065138546351372e-07, "loss": 0.0125, "step": 26698 }, { "epoch": 2.8945143104943627, "grad_norm": 0.8624395132064819, "learning_rate": 7.0578848106775e-07, "loss": 0.0012, "step": 26699 }, { "epoch": 2.8946227233304422, "grad_norm": 0.7794644236564636, "learning_rate": 7.050631075003627e-07, "loss": 0.0114, "step": 26700 }, { "epoch": 2.8947311361665222, "grad_norm": 0.1549532413482666, "learning_rate": 7.043377339329756e-07, "loss": 0.0012, "step": 26701 }, { "epoch": 2.8948395490026018, "grad_norm": 0.06512372940778732, "learning_rate": 7.036123603655884e-07, "loss": 0.0001, "step": 26702 }, { "epoch": 2.8949479618386817, "grad_norm": 0.5558295845985413, "learning_rate": 7.028869867982012e-07, "loss": 0.0062, "step": 26703 }, { "epoch": 2.8950563746747617, "grad_norm": 0.010366051457822323, "learning_rate": 7.021616132308138e-07, "loss": 0.0001, "step": 26704 }, { "epoch": 2.8951647875108413, "grad_norm": 0.24332505464553833, "learning_rate": 7.014362396634266e-07, "loss": 0.0052, "step": 26705 }, { "epoch": 2.895273200346921, "grad_norm": 0.40348488092422485, "learning_rate": 7.007108660960395e-07, "loss": 0.011, "step": 26706 }, { "epoch": 2.895381613183001, "grad_norm": 0.8974972367286682, "learning_rate": 6.999854925286523e-07, "loss": 0.009, "step": 26707 }, { "epoch": 2.895490026019081, "grad_norm": 4.062755107879639, "learning_rate": 6.992601189612651e-07, "loss": 0.0031, "step": 26708 }, { "epoch": 2.8955984388551603, "grad_norm": 0.3137669563293457, "learning_rate": 6.985347453938778e-07, "loss": 0.0054, "step": 26709 }, { "epoch": 2.8957068516912403, "grad_norm": 0.08171537518501282, "learning_rate": 6.978093718264907e-07, "loss": 0.0006, "step": 26710 }, { "epoch": 2.89581526452732, "grad_norm": 0.47957342863082886, "learning_rate": 6.970839982591035e-07, "loss": 0.0172, "step": 26711 }, { "epoch": 2.8959236773634, "grad_norm": 4.462275981903076, "learning_rate": 6.963586246917163e-07, "loss": 0.0251, "step": 26712 }, { "epoch": 2.89603209019948, "grad_norm": 0.3535711467266083, "learning_rate": 6.956332511243291e-07, "loss": 0.0101, "step": 26713 }, { "epoch": 2.8961405030355594, "grad_norm": 0.00897825974971056, "learning_rate": 6.949078775569419e-07, "loss": 0.0001, "step": 26714 }, { "epoch": 2.8962489158716394, "grad_norm": 4.352555274963379, "learning_rate": 6.941825039895547e-07, "loss": 0.0201, "step": 26715 }, { "epoch": 2.896357328707719, "grad_norm": 0.34953758120536804, "learning_rate": 6.934571304221675e-07, "loss": 0.0109, "step": 26716 }, { "epoch": 2.896465741543799, "grad_norm": 0.32312798500061035, "learning_rate": 6.927317568547804e-07, "loss": 0.001, "step": 26717 }, { "epoch": 2.8965741543798784, "grad_norm": 0.17247818410396576, "learning_rate": 6.92006383287393e-07, "loss": 0.0017, "step": 26718 }, { "epoch": 2.8966825672159584, "grad_norm": 0.2984272539615631, "learning_rate": 6.912810097200058e-07, "loss": 0.0049, "step": 26719 }, { "epoch": 2.896790980052038, "grad_norm": 0.29892635345458984, "learning_rate": 6.905556361526186e-07, "loss": 0.0062, "step": 26720 }, { "epoch": 2.896899392888118, "grad_norm": 0.41102731227874756, "learning_rate": 6.898302625852314e-07, "loss": 0.0128, "step": 26721 }, { "epoch": 2.897007805724198, "grad_norm": 0.43900826573371887, "learning_rate": 6.891048890178442e-07, "loss": 0.0091, "step": 26722 }, { "epoch": 2.8971162185602775, "grad_norm": 7.764829635620117, "learning_rate": 6.883795154504571e-07, "loss": 0.0259, "step": 26723 }, { "epoch": 2.8972246313963574, "grad_norm": 0.5541992783546448, "learning_rate": 6.876541418830698e-07, "loss": 0.0136, "step": 26724 }, { "epoch": 2.897333044232437, "grad_norm": 0.009667943231761456, "learning_rate": 6.869287683156826e-07, "loss": 0.0001, "step": 26725 }, { "epoch": 2.897441457068517, "grad_norm": 1.722187876701355, "learning_rate": 6.862033947482954e-07, "loss": 0.017, "step": 26726 }, { "epoch": 2.897549869904597, "grad_norm": 0.5346180200576782, "learning_rate": 6.854780211809083e-07, "loss": 0.0014, "step": 26727 }, { "epoch": 2.8976582827406765, "grad_norm": 2.703126907348633, "learning_rate": 6.847526476135211e-07, "loss": 0.0263, "step": 26728 }, { "epoch": 2.897766695576756, "grad_norm": 0.09892116487026215, "learning_rate": 6.840272740461338e-07, "loss": 0.0004, "step": 26729 }, { "epoch": 2.897875108412836, "grad_norm": 0.10968609154224396, "learning_rate": 6.833019004787467e-07, "loss": 0.0007, "step": 26730 }, { "epoch": 2.897983521248916, "grad_norm": 0.9241652488708496, "learning_rate": 6.825765269113593e-07, "loss": 0.0075, "step": 26731 }, { "epoch": 2.8980919340849955, "grad_norm": 0.028846925124526024, "learning_rate": 6.818511533439722e-07, "loss": 0.0001, "step": 26732 }, { "epoch": 2.8982003469210755, "grad_norm": 0.02762412652373314, "learning_rate": 6.81125779776585e-07, "loss": 0.0002, "step": 26733 }, { "epoch": 2.898308759757155, "grad_norm": 0.03139851987361908, "learning_rate": 6.804004062091977e-07, "loss": 0.0002, "step": 26734 }, { "epoch": 2.898417172593235, "grad_norm": 1.1308751106262207, "learning_rate": 6.796750326418105e-07, "loss": 0.0023, "step": 26735 }, { "epoch": 2.898525585429315, "grad_norm": 6.93812370300293, "learning_rate": 6.789496590744234e-07, "loss": 0.01, "step": 26736 }, { "epoch": 2.8986339982653946, "grad_norm": 2.6425108909606934, "learning_rate": 6.782242855070362e-07, "loss": 0.0344, "step": 26737 }, { "epoch": 2.8987424111014746, "grad_norm": 1.5520985126495361, "learning_rate": 6.77498911939649e-07, "loss": 0.0089, "step": 26738 }, { "epoch": 2.898850823937554, "grad_norm": 1.1268589496612549, "learning_rate": 6.767735383722618e-07, "loss": 0.009, "step": 26739 }, { "epoch": 2.898959236773634, "grad_norm": 0.31382107734680176, "learning_rate": 6.760481648048746e-07, "loss": 0.0067, "step": 26740 }, { "epoch": 2.8990676496097136, "grad_norm": 0.03001388907432556, "learning_rate": 6.753227912374874e-07, "loss": 0.0002, "step": 26741 }, { "epoch": 2.8991760624457936, "grad_norm": 0.31620487570762634, "learning_rate": 6.745974176701002e-07, "loss": 0.0152, "step": 26742 }, { "epoch": 2.899284475281873, "grad_norm": 0.8532934188842773, "learning_rate": 6.738720441027131e-07, "loss": 0.0236, "step": 26743 }, { "epoch": 2.899392888117953, "grad_norm": 0.3049905300140381, "learning_rate": 6.731466705353258e-07, "loss": 0.0058, "step": 26744 }, { "epoch": 2.899501300954033, "grad_norm": 0.10774997621774673, "learning_rate": 6.724212969679385e-07, "loss": 0.0007, "step": 26745 }, { "epoch": 2.8996097137901127, "grad_norm": 0.07197680324316025, "learning_rate": 6.716959234005513e-07, "loss": 0.0004, "step": 26746 }, { "epoch": 2.8997181266261927, "grad_norm": 0.004095737356692553, "learning_rate": 6.709705498331641e-07, "loss": 0.0, "step": 26747 }, { "epoch": 2.899826539462272, "grad_norm": 3.518071413040161, "learning_rate": 6.70245176265777e-07, "loss": 0.031, "step": 26748 }, { "epoch": 2.899934952298352, "grad_norm": 0.3297313153743744, "learning_rate": 6.695198026983897e-07, "loss": 0.0109, "step": 26749 }, { "epoch": 2.900043365134432, "grad_norm": 0.2733338475227356, "learning_rate": 6.687944291310025e-07, "loss": 0.0052, "step": 26750 }, { "epoch": 2.9001517779705117, "grad_norm": 0.5331010818481445, "learning_rate": 6.680690555636153e-07, "loss": 0.0138, "step": 26751 }, { "epoch": 2.9002601908065913, "grad_norm": 0.3909582495689392, "learning_rate": 6.673436819962282e-07, "loss": 0.0052, "step": 26752 }, { "epoch": 2.9003686036426712, "grad_norm": 0.16761843860149384, "learning_rate": 6.66618308428841e-07, "loss": 0.0003, "step": 26753 }, { "epoch": 2.9004770164787512, "grad_norm": 0.055829089134931564, "learning_rate": 6.658929348614537e-07, "loss": 0.0001, "step": 26754 }, { "epoch": 2.9005854293148308, "grad_norm": 0.27775946259498596, "learning_rate": 6.651675612940665e-07, "loss": 0.0059, "step": 26755 }, { "epoch": 2.9006938421509108, "grad_norm": 0.01447684969753027, "learning_rate": 6.644421877266794e-07, "loss": 0.0001, "step": 26756 }, { "epoch": 2.9008022549869903, "grad_norm": 0.36738014221191406, "learning_rate": 6.637168141592922e-07, "loss": 0.0071, "step": 26757 }, { "epoch": 2.9009106678230703, "grad_norm": 2.4754536151885986, "learning_rate": 6.629914405919048e-07, "loss": 0.0022, "step": 26758 }, { "epoch": 2.9010190806591503, "grad_norm": 11.18359375, "learning_rate": 6.622660670245176e-07, "loss": 0.013, "step": 26759 }, { "epoch": 2.90112749349523, "grad_norm": 9.411130905151367, "learning_rate": 6.615406934571304e-07, "loss": 0.0064, "step": 26760 }, { "epoch": 2.90123590633131, "grad_norm": 0.05642382800579071, "learning_rate": 6.608153198897433e-07, "loss": 0.0003, "step": 26761 }, { "epoch": 2.9013443191673893, "grad_norm": 0.036655355244874954, "learning_rate": 6.600899463223561e-07, "loss": 0.0002, "step": 26762 }, { "epoch": 2.9014527320034693, "grad_norm": 0.5802568793296814, "learning_rate": 6.593645727549688e-07, "loss": 0.0084, "step": 26763 }, { "epoch": 2.901561144839549, "grad_norm": 0.6321269869804382, "learning_rate": 6.586391991875816e-07, "loss": 0.0157, "step": 26764 }, { "epoch": 2.901669557675629, "grad_norm": 0.2517196536064148, "learning_rate": 6.579138256201945e-07, "loss": 0.0052, "step": 26765 }, { "epoch": 2.9017779705117084, "grad_norm": 0.5011228322982788, "learning_rate": 6.571884520528073e-07, "loss": 0.0005, "step": 26766 }, { "epoch": 2.9018863833477884, "grad_norm": 0.05687669664621353, "learning_rate": 6.564630784854201e-07, "loss": 0.0003, "step": 26767 }, { "epoch": 2.9019947961838684, "grad_norm": 0.3875727355480194, "learning_rate": 6.557377049180328e-07, "loss": 0.0012, "step": 26768 }, { "epoch": 2.902103209019948, "grad_norm": 0.017144376412034035, "learning_rate": 6.550123313506457e-07, "loss": 0.0002, "step": 26769 }, { "epoch": 2.902211621856028, "grad_norm": 0.6352483630180359, "learning_rate": 6.542869577832585e-07, "loss": 0.0163, "step": 26770 }, { "epoch": 2.9023200346921074, "grad_norm": 1.8322932720184326, "learning_rate": 6.535615842158713e-07, "loss": 0.0055, "step": 26771 }, { "epoch": 2.9024284475281874, "grad_norm": 5.722673416137695, "learning_rate": 6.52836210648484e-07, "loss": 0.0188, "step": 26772 }, { "epoch": 2.9025368603642674, "grad_norm": 1.1304420232772827, "learning_rate": 6.521108370810967e-07, "loss": 0.0027, "step": 26773 }, { "epoch": 2.902645273200347, "grad_norm": 0.06704362481832504, "learning_rate": 6.513854635137096e-07, "loss": 0.0004, "step": 26774 }, { "epoch": 2.9027536860364265, "grad_norm": 0.48609229922294617, "learning_rate": 6.506600899463224e-07, "loss": 0.0098, "step": 26775 }, { "epoch": 2.9028620988725065, "grad_norm": 0.45594480633735657, "learning_rate": 6.499347163789352e-07, "loss": 0.0108, "step": 26776 }, { "epoch": 2.9029705117085864, "grad_norm": 2.6660056114196777, "learning_rate": 6.49209342811548e-07, "loss": 0.0081, "step": 26777 }, { "epoch": 2.903078924544666, "grad_norm": 0.07806291431188583, "learning_rate": 6.484839692441608e-07, "loss": 0.0006, "step": 26778 }, { "epoch": 2.903187337380746, "grad_norm": 0.058499909937381744, "learning_rate": 6.477585956767736e-07, "loss": 0.0005, "step": 26779 }, { "epoch": 2.9032957502168255, "grad_norm": 3.6965415477752686, "learning_rate": 6.470332221093864e-07, "loss": 0.0035, "step": 26780 }, { "epoch": 2.9034041630529055, "grad_norm": 0.3209834098815918, "learning_rate": 6.463078485419992e-07, "loss": 0.0004, "step": 26781 }, { "epoch": 2.9035125758889855, "grad_norm": 2.815675973892212, "learning_rate": 6.455824749746121e-07, "loss": 0.0191, "step": 26782 }, { "epoch": 2.903620988725065, "grad_norm": 0.010815545916557312, "learning_rate": 6.448571014072248e-07, "loss": 0.0001, "step": 26783 }, { "epoch": 2.9037294015611446, "grad_norm": 0.23784545063972473, "learning_rate": 6.441317278398376e-07, "loss": 0.0038, "step": 26784 }, { "epoch": 2.9038378143972245, "grad_norm": 0.2964246869087219, "learning_rate": 6.434063542724503e-07, "loss": 0.0062, "step": 26785 }, { "epoch": 2.9039462272333045, "grad_norm": 0.058122921735048294, "learning_rate": 6.426809807050631e-07, "loss": 0.0005, "step": 26786 }, { "epoch": 2.904054640069384, "grad_norm": 0.041462525725364685, "learning_rate": 6.41955607137676e-07, "loss": 0.0003, "step": 26787 }, { "epoch": 2.904163052905464, "grad_norm": 0.020359115675091743, "learning_rate": 6.412302335702887e-07, "loss": 0.0002, "step": 26788 }, { "epoch": 2.9042714657415436, "grad_norm": 5.748144626617432, "learning_rate": 6.405048600029015e-07, "loss": 0.0298, "step": 26789 }, { "epoch": 2.9043798785776236, "grad_norm": 0.9780769944190979, "learning_rate": 6.397794864355143e-07, "loss": 0.003, "step": 26790 }, { "epoch": 2.9044882914137036, "grad_norm": 1.1088281869888306, "learning_rate": 6.390541128681272e-07, "loss": 0.049, "step": 26791 }, { "epoch": 2.904596704249783, "grad_norm": 0.025546446442604065, "learning_rate": 6.3832873930074e-07, "loss": 0.0001, "step": 26792 }, { "epoch": 2.904705117085863, "grad_norm": 0.4171523153781891, "learning_rate": 6.376033657333527e-07, "loss": 0.0066, "step": 26793 }, { "epoch": 2.9048135299219426, "grad_norm": 1.7983288764953613, "learning_rate": 6.368779921659655e-07, "loss": 0.0488, "step": 26794 }, { "epoch": 2.9049219427580226, "grad_norm": 3.0291860103607178, "learning_rate": 6.361526185985784e-07, "loss": 0.0285, "step": 26795 }, { "epoch": 2.9050303555941026, "grad_norm": 0.9042936563491821, "learning_rate": 6.354272450311912e-07, "loss": 0.0065, "step": 26796 }, { "epoch": 2.905138768430182, "grad_norm": 2.524238109588623, "learning_rate": 6.34701871463804e-07, "loss": 0.0432, "step": 26797 }, { "epoch": 2.9052471812662617, "grad_norm": 2.4321815967559814, "learning_rate": 6.339764978964167e-07, "loss": 0.025, "step": 26798 }, { "epoch": 2.9053555941023417, "grad_norm": 0.2304864376783371, "learning_rate": 6.332511243290294e-07, "loss": 0.0009, "step": 26799 }, { "epoch": 2.9054640069384217, "grad_norm": 2.8180148601531982, "learning_rate": 6.325257507616423e-07, "loss": 0.0295, "step": 26800 }, { "epoch": 2.905572419774501, "grad_norm": 2.5420749187469482, "learning_rate": 6.318003771942551e-07, "loss": 0.0573, "step": 26801 }, { "epoch": 2.905680832610581, "grad_norm": 2.3987669944763184, "learning_rate": 6.310750036268678e-07, "loss": 0.0439, "step": 26802 }, { "epoch": 2.9057892454466607, "grad_norm": 9.479519844055176, "learning_rate": 6.303496300594806e-07, "loss": 0.0359, "step": 26803 }, { "epoch": 2.9058976582827407, "grad_norm": 0.4344705641269684, "learning_rate": 6.296242564920935e-07, "loss": 0.0043, "step": 26804 }, { "epoch": 2.9060060711188207, "grad_norm": 6.478896141052246, "learning_rate": 6.288988829247063e-07, "loss": 0.0147, "step": 26805 }, { "epoch": 2.9061144839549002, "grad_norm": 0.30813953280448914, "learning_rate": 6.281735093573191e-07, "loss": 0.0111, "step": 26806 }, { "epoch": 2.90622289679098, "grad_norm": 0.7566632032394409, "learning_rate": 6.274481357899318e-07, "loss": 0.0121, "step": 26807 }, { "epoch": 2.9063313096270598, "grad_norm": 0.043949734419584274, "learning_rate": 6.267227622225447e-07, "loss": 0.0004, "step": 26808 }, { "epoch": 2.9064397224631398, "grad_norm": 1.5142648220062256, "learning_rate": 6.259973886551575e-07, "loss": 0.0339, "step": 26809 }, { "epoch": 2.9065481352992193, "grad_norm": 0.029029957950115204, "learning_rate": 6.252720150877703e-07, "loss": 0.0001, "step": 26810 }, { "epoch": 2.9066565481352993, "grad_norm": 0.2205880582332611, "learning_rate": 6.245466415203831e-07, "loss": 0.0008, "step": 26811 }, { "epoch": 2.906764960971379, "grad_norm": 0.3615623712539673, "learning_rate": 6.238212679529958e-07, "loss": 0.0066, "step": 26812 }, { "epoch": 2.906873373807459, "grad_norm": 0.2085309624671936, "learning_rate": 6.230958943856086e-07, "loss": 0.0005, "step": 26813 }, { "epoch": 2.906981786643539, "grad_norm": 0.3045007288455963, "learning_rate": 6.223705208182215e-07, "loss": 0.0062, "step": 26814 }, { "epoch": 2.9070901994796183, "grad_norm": 0.431007981300354, "learning_rate": 6.216451472508343e-07, "loss": 0.0102, "step": 26815 }, { "epoch": 2.9071986123156983, "grad_norm": 2.8673524856567383, "learning_rate": 6.20919773683447e-07, "loss": 0.0103, "step": 26816 }, { "epoch": 2.907307025151778, "grad_norm": 1.5676326751708984, "learning_rate": 6.201944001160598e-07, "loss": 0.0204, "step": 26817 }, { "epoch": 2.907415437987858, "grad_norm": 0.37180280685424805, "learning_rate": 6.194690265486726e-07, "loss": 0.0103, "step": 26818 }, { "epoch": 2.907523850823938, "grad_norm": 0.018140634521842003, "learning_rate": 6.187436529812854e-07, "loss": 0.0001, "step": 26819 }, { "epoch": 2.9076322636600174, "grad_norm": 0.8762046098709106, "learning_rate": 6.180182794138982e-07, "loss": 0.0203, "step": 26820 }, { "epoch": 2.907740676496097, "grad_norm": 0.36272045969963074, "learning_rate": 6.172929058465111e-07, "loss": 0.0065, "step": 26821 }, { "epoch": 2.907849089332177, "grad_norm": 0.43627429008483887, "learning_rate": 6.165675322791238e-07, "loss": 0.0107, "step": 26822 }, { "epoch": 2.907957502168257, "grad_norm": 0.4031775891780853, "learning_rate": 6.158421587117366e-07, "loss": 0.0122, "step": 26823 }, { "epoch": 2.9080659150043364, "grad_norm": 0.7367070913314819, "learning_rate": 6.151167851443494e-07, "loss": 0.0318, "step": 26824 }, { "epoch": 2.9081743278404164, "grad_norm": 0.036068130284547806, "learning_rate": 6.143914115769622e-07, "loss": 0.0002, "step": 26825 }, { "epoch": 2.908282740676496, "grad_norm": 0.45122721791267395, "learning_rate": 6.13666038009575e-07, "loss": 0.0111, "step": 26826 }, { "epoch": 2.908391153512576, "grad_norm": 0.3464856445789337, "learning_rate": 6.129406644421878e-07, "loss": 0.0068, "step": 26827 }, { "epoch": 2.908499566348656, "grad_norm": 0.007011635694652796, "learning_rate": 6.122152908748006e-07, "loss": 0.0001, "step": 26828 }, { "epoch": 2.9086079791847355, "grad_norm": 1.6463241577148438, "learning_rate": 6.114899173074133e-07, "loss": 0.0552, "step": 26829 }, { "epoch": 2.908716392020815, "grad_norm": 0.3982383608818054, "learning_rate": 6.107645437400262e-07, "loss": 0.0105, "step": 26830 }, { "epoch": 2.908824804856895, "grad_norm": 0.43083855509757996, "learning_rate": 6.10039170172639e-07, "loss": 0.0106, "step": 26831 }, { "epoch": 2.908933217692975, "grad_norm": 0.9036775231361389, "learning_rate": 6.093137966052517e-07, "loss": 0.0169, "step": 26832 }, { "epoch": 2.9090416305290545, "grad_norm": 0.011430120095610619, "learning_rate": 6.085884230378645e-07, "loss": 0.0001, "step": 26833 }, { "epoch": 2.9091500433651345, "grad_norm": 0.3416607975959778, "learning_rate": 6.078630494704774e-07, "loss": 0.0078, "step": 26834 }, { "epoch": 2.909258456201214, "grad_norm": 15.178486824035645, "learning_rate": 6.071376759030902e-07, "loss": 0.0242, "step": 26835 }, { "epoch": 2.909366869037294, "grad_norm": 0.055779747664928436, "learning_rate": 6.06412302335703e-07, "loss": 0.0005, "step": 26836 }, { "epoch": 2.909475281873374, "grad_norm": 0.015150906518101692, "learning_rate": 6.056869287683157e-07, "loss": 0.0001, "step": 26837 }, { "epoch": 2.9095836947094535, "grad_norm": 0.5876439213752747, "learning_rate": 6.049615552009285e-07, "loss": 0.0085, "step": 26838 }, { "epoch": 2.9096921075455335, "grad_norm": 0.31619948148727417, "learning_rate": 6.042361816335413e-07, "loss": 0.0102, "step": 26839 }, { "epoch": 2.909800520381613, "grad_norm": 0.01231870986521244, "learning_rate": 6.035108080661542e-07, "loss": 0.0001, "step": 26840 }, { "epoch": 2.909908933217693, "grad_norm": 0.062167342752218246, "learning_rate": 6.02785434498767e-07, "loss": 0.0003, "step": 26841 }, { "epoch": 2.9100173460537726, "grad_norm": 0.6386220455169678, "learning_rate": 6.020600609313797e-07, "loss": 0.0102, "step": 26842 }, { "epoch": 2.9101257588898526, "grad_norm": 9.946582794189453, "learning_rate": 6.013346873639925e-07, "loss": 0.0223, "step": 26843 }, { "epoch": 2.910234171725932, "grad_norm": 0.9222648739814758, "learning_rate": 6.006093137966053e-07, "loss": 0.0237, "step": 26844 }, { "epoch": 2.910342584562012, "grad_norm": 3.398428201675415, "learning_rate": 5.998839402292181e-07, "loss": 0.0114, "step": 26845 }, { "epoch": 2.910450997398092, "grad_norm": 0.03728437423706055, "learning_rate": 5.991585666618308e-07, "loss": 0.0003, "step": 26846 }, { "epoch": 2.9105594102341716, "grad_norm": 3.9730381965637207, "learning_rate": 5.984331930944437e-07, "loss": 0.0502, "step": 26847 }, { "epoch": 2.9106678230702516, "grad_norm": 0.890957772731781, "learning_rate": 5.977078195270565e-07, "loss": 0.0328, "step": 26848 }, { "epoch": 2.910776235906331, "grad_norm": 0.8050689697265625, "learning_rate": 5.969824459596693e-07, "loss": 0.021, "step": 26849 }, { "epoch": 2.910884648742411, "grad_norm": 0.35380080342292786, "learning_rate": 5.962570723922821e-07, "loss": 0.001, "step": 26850 }, { "epoch": 2.910993061578491, "grad_norm": 0.011118141934275627, "learning_rate": 5.955316988248948e-07, "loss": 0.0001, "step": 26851 }, { "epoch": 2.9111014744145707, "grad_norm": 0.5084323883056641, "learning_rate": 5.948063252575076e-07, "loss": 0.0153, "step": 26852 }, { "epoch": 2.91120988725065, "grad_norm": 0.7436836957931519, "learning_rate": 5.940809516901205e-07, "loss": 0.0257, "step": 26853 }, { "epoch": 2.91131830008673, "grad_norm": 0.03445734456181526, "learning_rate": 5.933555781227333e-07, "loss": 0.0003, "step": 26854 }, { "epoch": 2.91142671292281, "grad_norm": 0.05592799186706543, "learning_rate": 5.926302045553461e-07, "loss": 0.0004, "step": 26855 }, { "epoch": 2.9115351257588897, "grad_norm": 0.24426835775375366, "learning_rate": 5.919048309879588e-07, "loss": 0.0043, "step": 26856 }, { "epoch": 2.9116435385949697, "grad_norm": 10.588077545166016, "learning_rate": 5.911794574205716e-07, "loss": 0.0377, "step": 26857 }, { "epoch": 2.9117519514310493, "grad_norm": 0.6272902488708496, "learning_rate": 5.904540838531844e-07, "loss": 0.0095, "step": 26858 }, { "epoch": 2.9118603642671292, "grad_norm": 0.028894057497382164, "learning_rate": 5.897287102857972e-07, "loss": 0.0002, "step": 26859 }, { "epoch": 2.9119687771032092, "grad_norm": 0.08151844143867493, "learning_rate": 5.890033367184101e-07, "loss": 0.0006, "step": 26860 }, { "epoch": 2.9120771899392888, "grad_norm": 0.03422674536705017, "learning_rate": 5.882779631510228e-07, "loss": 0.0003, "step": 26861 }, { "epoch": 2.9121856027753688, "grad_norm": 0.4484417736530304, "learning_rate": 5.875525895836356e-07, "loss": 0.013, "step": 26862 }, { "epoch": 2.9122940156114483, "grad_norm": 3.552384614944458, "learning_rate": 5.868272160162484e-07, "loss": 0.007, "step": 26863 }, { "epoch": 2.9124024284475283, "grad_norm": 4.874019622802734, "learning_rate": 5.861018424488612e-07, "loss": 0.0179, "step": 26864 }, { "epoch": 2.912510841283608, "grad_norm": 0.18303759396076202, "learning_rate": 5.85376468881474e-07, "loss": 0.0028, "step": 26865 }, { "epoch": 2.912619254119688, "grad_norm": 0.07250576466321945, "learning_rate": 5.846510953140868e-07, "loss": 0.0005, "step": 26866 }, { "epoch": 2.9127276669557673, "grad_norm": 0.35936617851257324, "learning_rate": 5.839257217466996e-07, "loss": 0.0008, "step": 26867 }, { "epoch": 2.9128360797918473, "grad_norm": 0.05880061164498329, "learning_rate": 5.832003481793124e-07, "loss": 0.0003, "step": 26868 }, { "epoch": 2.9129444926279273, "grad_norm": 0.4842328727245331, "learning_rate": 5.824749746119252e-07, "loss": 0.0076, "step": 26869 }, { "epoch": 2.913052905464007, "grad_norm": 6.20100736618042, "learning_rate": 5.81749601044538e-07, "loss": 0.0207, "step": 26870 }, { "epoch": 2.913161318300087, "grad_norm": 0.2995074391365051, "learning_rate": 5.810242274771507e-07, "loss": 0.0022, "step": 26871 }, { "epoch": 2.9132697311361664, "grad_norm": 0.2569020986557007, "learning_rate": 5.802988539097635e-07, "loss": 0.0051, "step": 26872 }, { "epoch": 2.9133781439722464, "grad_norm": 0.4367125332355499, "learning_rate": 5.795734803423764e-07, "loss": 0.0089, "step": 26873 }, { "epoch": 2.9134865568083264, "grad_norm": 0.026880476623773575, "learning_rate": 5.788481067749892e-07, "loss": 0.0002, "step": 26874 }, { "epoch": 2.913594969644406, "grad_norm": 0.8007621765136719, "learning_rate": 5.78122733207602e-07, "loss": 0.022, "step": 26875 }, { "epoch": 2.9137033824804854, "grad_norm": 15.946907043457031, "learning_rate": 5.773973596402147e-07, "loss": 0.0311, "step": 26876 }, { "epoch": 2.9138117953165654, "grad_norm": 0.15912863612174988, "learning_rate": 5.766719860728275e-07, "loss": 0.0007, "step": 26877 }, { "epoch": 2.9139202081526454, "grad_norm": 0.23367491364479065, "learning_rate": 5.759466125054403e-07, "loss": 0.0008, "step": 26878 }, { "epoch": 2.914028620988725, "grad_norm": 0.024215416982769966, "learning_rate": 5.752212389380532e-07, "loss": 0.0002, "step": 26879 }, { "epoch": 2.914137033824805, "grad_norm": 0.4927472174167633, "learning_rate": 5.74495865370666e-07, "loss": 0.0115, "step": 26880 }, { "epoch": 2.9142454466608845, "grad_norm": 2.1652398109436035, "learning_rate": 5.737704918032787e-07, "loss": 0.0027, "step": 26881 }, { "epoch": 2.9143538594969645, "grad_norm": 0.05636603757739067, "learning_rate": 5.730451182358915e-07, "loss": 0.0003, "step": 26882 }, { "epoch": 2.9144622723330444, "grad_norm": 0.2769264280796051, "learning_rate": 5.723197446685044e-07, "loss": 0.0127, "step": 26883 }, { "epoch": 2.914570685169124, "grad_norm": 0.039456725120544434, "learning_rate": 5.715943711011171e-07, "loss": 0.0004, "step": 26884 }, { "epoch": 2.914679098005204, "grad_norm": 0.26007649302482605, "learning_rate": 5.708689975337299e-07, "loss": 0.0051, "step": 26885 }, { "epoch": 2.9147875108412835, "grad_norm": 9.247635841369629, "learning_rate": 5.701436239663427e-07, "loss": 0.0135, "step": 26886 }, { "epoch": 2.9148959236773635, "grad_norm": 0.09698233753442764, "learning_rate": 5.694182503989555e-07, "loss": 0.0004, "step": 26887 }, { "epoch": 2.915004336513443, "grad_norm": 0.15221069753170013, "learning_rate": 5.686928768315683e-07, "loss": 0.0008, "step": 26888 }, { "epoch": 2.915112749349523, "grad_norm": 3.0730772018432617, "learning_rate": 5.679675032641812e-07, "loss": 0.0243, "step": 26889 }, { "epoch": 2.9152211621856026, "grad_norm": 6.128891468048096, "learning_rate": 5.672421296967938e-07, "loss": 0.0038, "step": 26890 }, { "epoch": 2.9153295750216826, "grad_norm": 8.507440567016602, "learning_rate": 5.665167561294066e-07, "loss": 0.0408, "step": 26891 }, { "epoch": 2.9154379878577625, "grad_norm": 0.0669964849948883, "learning_rate": 5.657913825620195e-07, "loss": 0.0005, "step": 26892 }, { "epoch": 2.915546400693842, "grad_norm": 0.9119313359260559, "learning_rate": 5.650660089946323e-07, "loss": 0.0065, "step": 26893 }, { "epoch": 2.915654813529922, "grad_norm": 2.134413480758667, "learning_rate": 5.643406354272451e-07, "loss": 0.0225, "step": 26894 }, { "epoch": 2.9157632263660016, "grad_norm": 0.040566787123680115, "learning_rate": 5.636152618598578e-07, "loss": 0.0003, "step": 26895 }, { "epoch": 2.9158716392020816, "grad_norm": 9.454720497131348, "learning_rate": 5.628898882924707e-07, "loss": 0.0111, "step": 26896 }, { "epoch": 2.9159800520381616, "grad_norm": 4.109323501586914, "learning_rate": 5.621645147250834e-07, "loss": 0.0087, "step": 26897 }, { "epoch": 2.916088464874241, "grad_norm": 2.18416166305542, "learning_rate": 5.614391411576963e-07, "loss": 0.0534, "step": 26898 }, { "epoch": 2.9161968777103207, "grad_norm": 0.07558974623680115, "learning_rate": 5.607137675903091e-07, "loss": 0.0004, "step": 26899 }, { "epoch": 2.9163052905464006, "grad_norm": 26.920963287353516, "learning_rate": 5.599883940229218e-07, "loss": 0.0273, "step": 26900 }, { "epoch": 2.9164137033824806, "grad_norm": 0.777206540107727, "learning_rate": 5.592630204555346e-07, "loss": 0.0204, "step": 26901 }, { "epoch": 2.91652211621856, "grad_norm": 0.3396759033203125, "learning_rate": 5.585376468881475e-07, "loss": 0.0111, "step": 26902 }, { "epoch": 2.91663052905464, "grad_norm": 0.3733619153499603, "learning_rate": 5.578122733207603e-07, "loss": 0.0063, "step": 26903 }, { "epoch": 2.9167389418907197, "grad_norm": 1.9229894876480103, "learning_rate": 5.57086899753373e-07, "loss": 0.0189, "step": 26904 }, { "epoch": 2.9168473547267997, "grad_norm": 0.7110551595687866, "learning_rate": 5.563615261859858e-07, "loss": 0.0164, "step": 26905 }, { "epoch": 2.9169557675628797, "grad_norm": 2.3494372367858887, "learning_rate": 5.556361526185986e-07, "loss": 0.0539, "step": 26906 }, { "epoch": 2.917064180398959, "grad_norm": 0.3514930307865143, "learning_rate": 5.549107790512114e-07, "loss": 0.0084, "step": 26907 }, { "epoch": 2.917172593235039, "grad_norm": 0.010348140262067318, "learning_rate": 5.541854054838242e-07, "loss": 0.0001, "step": 26908 }, { "epoch": 2.9172810060711187, "grad_norm": 1.948936104774475, "learning_rate": 5.534600319164371e-07, "loss": 0.0328, "step": 26909 }, { "epoch": 2.9173894189071987, "grad_norm": 0.4543253779411316, "learning_rate": 5.527346583490498e-07, "loss": 0.0084, "step": 26910 }, { "epoch": 2.9174978317432783, "grad_norm": 4.32456111907959, "learning_rate": 5.520092847816626e-07, "loss": 0.0113, "step": 26911 }, { "epoch": 2.9176062445793582, "grad_norm": 0.2612761855125427, "learning_rate": 5.512839112142754e-07, "loss": 0.0056, "step": 26912 }, { "epoch": 2.917714657415438, "grad_norm": 0.01724115014076233, "learning_rate": 5.505585376468882e-07, "loss": 0.0002, "step": 26913 }, { "epoch": 2.9178230702515178, "grad_norm": 1.6691653728485107, "learning_rate": 5.49833164079501e-07, "loss": 0.0123, "step": 26914 }, { "epoch": 2.9179314830875978, "grad_norm": 0.035173285752534866, "learning_rate": 5.491077905121138e-07, "loss": 0.0003, "step": 26915 }, { "epoch": 2.9180398959236773, "grad_norm": 0.020201165229082108, "learning_rate": 5.483824169447266e-07, "loss": 0.0002, "step": 26916 }, { "epoch": 2.9181483087597573, "grad_norm": 0.6000863313674927, "learning_rate": 5.476570433773394e-07, "loss": 0.0198, "step": 26917 }, { "epoch": 2.918256721595837, "grad_norm": 0.03180611506104469, "learning_rate": 5.469316698099522e-07, "loss": 0.0002, "step": 26918 }, { "epoch": 2.918365134431917, "grad_norm": 0.2906405031681061, "learning_rate": 5.46206296242565e-07, "loss": 0.0008, "step": 26919 }, { "epoch": 2.918473547267997, "grad_norm": 0.3652644753456116, "learning_rate": 5.454809226751777e-07, "loss": 0.0009, "step": 26920 }, { "epoch": 2.9185819601040763, "grad_norm": 0.5338402390480042, "learning_rate": 5.447555491077905e-07, "loss": 0.0108, "step": 26921 }, { "epoch": 2.918690372940156, "grad_norm": 0.005558626726269722, "learning_rate": 5.440301755404034e-07, "loss": 0.0001, "step": 26922 }, { "epoch": 2.918798785776236, "grad_norm": 0.005497022066265345, "learning_rate": 5.433048019730162e-07, "loss": 0.0001, "step": 26923 }, { "epoch": 2.918907198612316, "grad_norm": 0.30007466673851013, "learning_rate": 5.42579428405629e-07, "loss": 0.0061, "step": 26924 }, { "epoch": 2.9190156114483954, "grad_norm": 2.9290080070495605, "learning_rate": 5.418540548382417e-07, "loss": 0.0015, "step": 26925 }, { "epoch": 2.9191240242844754, "grad_norm": 0.2843779921531677, "learning_rate": 5.411286812708545e-07, "loss": 0.0062, "step": 26926 }, { "epoch": 2.919232437120555, "grad_norm": 1.3956609964370728, "learning_rate": 5.404033077034673e-07, "loss": 0.0386, "step": 26927 }, { "epoch": 2.919340849956635, "grad_norm": 0.040838517248630524, "learning_rate": 5.396779341360802e-07, "loss": 0.0003, "step": 26928 }, { "epoch": 2.919449262792715, "grad_norm": 0.5333268046379089, "learning_rate": 5.38952560568693e-07, "loss": 0.0072, "step": 26929 }, { "epoch": 2.9195576756287944, "grad_norm": 2.31058669090271, "learning_rate": 5.382271870013057e-07, "loss": 0.0128, "step": 26930 }, { "epoch": 2.919666088464874, "grad_norm": 0.6537516117095947, "learning_rate": 5.375018134339185e-07, "loss": 0.0231, "step": 26931 }, { "epoch": 2.919774501300954, "grad_norm": 0.043256524950265884, "learning_rate": 5.367764398665313e-07, "loss": 0.0003, "step": 26932 }, { "epoch": 2.919882914137034, "grad_norm": 4.308847904205322, "learning_rate": 5.360510662991441e-07, "loss": 0.051, "step": 26933 }, { "epoch": 2.9199913269731135, "grad_norm": 0.020504271611571312, "learning_rate": 5.353256927317569e-07, "loss": 0.0002, "step": 26934 }, { "epoch": 2.9200997398091935, "grad_norm": 0.39851993322372437, "learning_rate": 5.346003191643697e-07, "loss": 0.0055, "step": 26935 }, { "epoch": 2.920208152645273, "grad_norm": 0.0322856679558754, "learning_rate": 5.338749455969825e-07, "loss": 0.0003, "step": 26936 }, { "epoch": 2.920316565481353, "grad_norm": 0.10242652893066406, "learning_rate": 5.331495720295953e-07, "loss": 0.0007, "step": 26937 }, { "epoch": 2.920424978317433, "grad_norm": 0.005772436037659645, "learning_rate": 5.324241984622081e-07, "loss": 0.0001, "step": 26938 }, { "epoch": 2.9205333911535125, "grad_norm": 0.15501902997493744, "learning_rate": 5.316988248948209e-07, "loss": 0.0004, "step": 26939 }, { "epoch": 2.9206418039895925, "grad_norm": 0.3234317898750305, "learning_rate": 5.309734513274336e-07, "loss": 0.0069, "step": 26940 }, { "epoch": 2.920750216825672, "grad_norm": 0.45010435581207275, "learning_rate": 5.302480777600465e-07, "loss": 0.0119, "step": 26941 }, { "epoch": 2.920858629661752, "grad_norm": 0.2849528193473816, "learning_rate": 5.295227041926593e-07, "loss": 0.0057, "step": 26942 }, { "epoch": 2.920967042497832, "grad_norm": 0.2899708151817322, "learning_rate": 5.287973306252721e-07, "loss": 0.0008, "step": 26943 }, { "epoch": 2.9210754553339116, "grad_norm": 0.20162174105644226, "learning_rate": 5.280719570578849e-07, "loss": 0.0007, "step": 26944 }, { "epoch": 2.921183868169991, "grad_norm": 0.4987686276435852, "learning_rate": 5.273465834904976e-07, "loss": 0.0036, "step": 26945 }, { "epoch": 2.921292281006071, "grad_norm": 0.3028342127799988, "learning_rate": 5.266212099231104e-07, "loss": 0.0037, "step": 26946 }, { "epoch": 2.921400693842151, "grad_norm": 0.024320269003510475, "learning_rate": 5.258958363557232e-07, "loss": 0.0002, "step": 26947 }, { "epoch": 2.9215091066782306, "grad_norm": 0.3074396252632141, "learning_rate": 5.251704627883361e-07, "loss": 0.0004, "step": 26948 }, { "epoch": 2.9216175195143106, "grad_norm": 0.47894713282585144, "learning_rate": 5.244450892209488e-07, "loss": 0.0063, "step": 26949 }, { "epoch": 2.92172593235039, "grad_norm": 7.618108749389648, "learning_rate": 5.237197156535616e-07, "loss": 0.023, "step": 26950 }, { "epoch": 2.92183434518647, "grad_norm": 0.1928359419107437, "learning_rate": 5.229943420861744e-07, "loss": 0.0009, "step": 26951 }, { "epoch": 2.92194275802255, "grad_norm": 8.222987174987793, "learning_rate": 5.222689685187872e-07, "loss": 0.0097, "step": 26952 }, { "epoch": 2.9220511708586296, "grad_norm": 1.0212920904159546, "learning_rate": 5.215435949514e-07, "loss": 0.0372, "step": 26953 }, { "epoch": 2.922159583694709, "grad_norm": 0.009154007770121098, "learning_rate": 5.208182213840128e-07, "loss": 0.0001, "step": 26954 }, { "epoch": 2.922267996530789, "grad_norm": 0.7625236511230469, "learning_rate": 5.200928478166256e-07, "loss": 0.0254, "step": 26955 }, { "epoch": 2.922376409366869, "grad_norm": 0.5847442150115967, "learning_rate": 5.193674742492384e-07, "loss": 0.0195, "step": 26956 }, { "epoch": 2.9224848222029487, "grad_norm": 0.554356575012207, "learning_rate": 5.186421006818512e-07, "loss": 0.0083, "step": 26957 }, { "epoch": 2.9225932350390287, "grad_norm": 0.26685160398483276, "learning_rate": 5.17916727114464e-07, "loss": 0.005, "step": 26958 }, { "epoch": 2.922701647875108, "grad_norm": 0.0170625988394022, "learning_rate": 5.171913535470767e-07, "loss": 0.0002, "step": 26959 }, { "epoch": 2.922810060711188, "grad_norm": 2.549311876296997, "learning_rate": 5.164659799796895e-07, "loss": 0.0085, "step": 26960 }, { "epoch": 2.922918473547268, "grad_norm": 0.02530457265675068, "learning_rate": 5.157406064123024e-07, "loss": 0.0002, "step": 26961 }, { "epoch": 2.9230268863833477, "grad_norm": 6.750770092010498, "learning_rate": 5.150152328449152e-07, "loss": 0.0477, "step": 26962 }, { "epoch": 2.9231352992194277, "grad_norm": 1.3991174697875977, "learning_rate": 5.14289859277528e-07, "loss": 0.0344, "step": 26963 }, { "epoch": 2.9232437120555073, "grad_norm": 1.8665962219238281, "learning_rate": 5.135644857101408e-07, "loss": 0.0384, "step": 26964 }, { "epoch": 2.9233521248915872, "grad_norm": 0.08634988218545914, "learning_rate": 5.128391121427535e-07, "loss": 0.0121, "step": 26965 }, { "epoch": 2.9234605377276672, "grad_norm": 1.5764808654785156, "learning_rate": 5.121137385753663e-07, "loss": 0.0313, "step": 26966 }, { "epoch": 2.9235689505637468, "grad_norm": 0.32405248284339905, "learning_rate": 5.113883650079792e-07, "loss": 0.0107, "step": 26967 }, { "epoch": 2.9236773633998263, "grad_norm": 0.584286093711853, "learning_rate": 5.10662991440592e-07, "loss": 0.0234, "step": 26968 }, { "epoch": 2.9237857762359063, "grad_norm": 0.43814724683761597, "learning_rate": 5.099376178732047e-07, "loss": 0.0214, "step": 26969 }, { "epoch": 2.9238941890719863, "grad_norm": 0.5477572083473206, "learning_rate": 5.092122443058175e-07, "loss": 0.006, "step": 26970 }, { "epoch": 2.924002601908066, "grad_norm": 0.5508407950401306, "learning_rate": 5.084868707384304e-07, "loss": 0.0135, "step": 26971 }, { "epoch": 2.924111014744146, "grad_norm": 0.03420896455645561, "learning_rate": 5.077614971710431e-07, "loss": 0.0003, "step": 26972 }, { "epoch": 2.9242194275802254, "grad_norm": 0.011043921113014221, "learning_rate": 5.07036123603656e-07, "loss": 0.0001, "step": 26973 }, { "epoch": 2.9243278404163053, "grad_norm": 0.3745231032371521, "learning_rate": 5.063107500362687e-07, "loss": 0.0129, "step": 26974 }, { "epoch": 2.9244362532523853, "grad_norm": 0.2596266269683838, "learning_rate": 5.055853764688815e-07, "loss": 0.0059, "step": 26975 }, { "epoch": 2.924544666088465, "grad_norm": 1.2334399223327637, "learning_rate": 5.048600029014943e-07, "loss": 0.0195, "step": 26976 }, { "epoch": 2.9246530789245444, "grad_norm": 8.612034797668457, "learning_rate": 5.041346293341072e-07, "loss": 0.0368, "step": 26977 }, { "epoch": 2.9247614917606244, "grad_norm": 0.5464527606964111, "learning_rate": 5.034092557667199e-07, "loss": 0.0069, "step": 26978 }, { "epoch": 2.9248699045967044, "grad_norm": 10.549970626831055, "learning_rate": 5.026838821993326e-07, "loss": 0.0096, "step": 26979 }, { "epoch": 2.924978317432784, "grad_norm": 0.12648586928844452, "learning_rate": 5.019585086319455e-07, "loss": 0.0004, "step": 26980 }, { "epoch": 2.925086730268864, "grad_norm": 5.70389461517334, "learning_rate": 5.012331350645583e-07, "loss": 0.0082, "step": 26981 }, { "epoch": 2.9251951431049434, "grad_norm": 0.6773084998130798, "learning_rate": 5.005077614971711e-07, "loss": 0.0143, "step": 26982 }, { "epoch": 2.9253035559410234, "grad_norm": 0.01681503839790821, "learning_rate": 4.997823879297839e-07, "loss": 0.0001, "step": 26983 }, { "epoch": 2.9254119687771034, "grad_norm": 0.01316705159842968, "learning_rate": 4.990570143623967e-07, "loss": 0.0001, "step": 26984 }, { "epoch": 2.925520381613183, "grad_norm": 1.594053030014038, "learning_rate": 4.983316407950094e-07, "loss": 0.0037, "step": 26985 }, { "epoch": 2.925628794449263, "grad_norm": 7.56292724609375, "learning_rate": 4.976062672276223e-07, "loss": 0.025, "step": 26986 }, { "epoch": 2.9257372072853425, "grad_norm": 0.9197974801063538, "learning_rate": 4.968808936602351e-07, "loss": 0.027, "step": 26987 }, { "epoch": 2.9258456201214225, "grad_norm": 8.894043922424316, "learning_rate": 4.961555200928479e-07, "loss": 0.0312, "step": 26988 }, { "epoch": 2.925954032957502, "grad_norm": 8.254021644592285, "learning_rate": 4.954301465254606e-07, "loss": 0.0061, "step": 26989 }, { "epoch": 2.926062445793582, "grad_norm": 0.031346581876277924, "learning_rate": 4.947047729580735e-07, "loss": 0.0003, "step": 26990 }, { "epoch": 2.9261708586296615, "grad_norm": 3.633303165435791, "learning_rate": 4.939793993906863e-07, "loss": 0.058, "step": 26991 }, { "epoch": 2.9262792714657415, "grad_norm": 0.06770490854978561, "learning_rate": 4.93254025823299e-07, "loss": 0.0004, "step": 26992 }, { "epoch": 2.9263876843018215, "grad_norm": 0.2864015996456146, "learning_rate": 4.925286522559119e-07, "loss": 0.0061, "step": 26993 }, { "epoch": 2.926496097137901, "grad_norm": 3.1383168697357178, "learning_rate": 4.918032786885246e-07, "loss": 0.012, "step": 26994 }, { "epoch": 2.926604509973981, "grad_norm": 0.41860929131507874, "learning_rate": 4.910779051211374e-07, "loss": 0.0112, "step": 26995 }, { "epoch": 2.9267129228100606, "grad_norm": 1.9935914278030396, "learning_rate": 4.903525315537502e-07, "loss": 0.0566, "step": 26996 }, { "epoch": 2.9268213356461406, "grad_norm": 1.9957756996154785, "learning_rate": 4.896271579863631e-07, "loss": 0.0149, "step": 26997 }, { "epoch": 2.9269297484822205, "grad_norm": 0.2266562581062317, "learning_rate": 4.889017844189759e-07, "loss": 0.0005, "step": 26998 }, { "epoch": 2.9270381613183, "grad_norm": 0.42002150416374207, "learning_rate": 4.881764108515886e-07, "loss": 0.0007, "step": 26999 }, { "epoch": 2.9271465741543796, "grad_norm": 0.10428163409233093, "learning_rate": 4.874510372842014e-07, "loss": 0.0003, "step": 27000 }, { "epoch": 2.9272549869904596, "grad_norm": 0.12942196428775787, "learning_rate": 4.867256637168142e-07, "loss": 0.0005, "step": 27001 }, { "epoch": 2.9273633998265396, "grad_norm": 0.036874037235975266, "learning_rate": 4.86000290149427e-07, "loss": 0.0003, "step": 27002 }, { "epoch": 2.927471812662619, "grad_norm": 0.4221324622631073, "learning_rate": 4.852749165820399e-07, "loss": 0.0105, "step": 27003 }, { "epoch": 2.927580225498699, "grad_norm": 0.3949218690395355, "learning_rate": 4.845495430146526e-07, "loss": 0.0079, "step": 27004 }, { "epoch": 2.9276886383347787, "grad_norm": 0.15420275926589966, "learning_rate": 4.838241694472654e-07, "loss": 0.0007, "step": 27005 }, { "epoch": 2.9277970511708586, "grad_norm": 0.881712794303894, "learning_rate": 4.830987958798782e-07, "loss": 0.0011, "step": 27006 }, { "epoch": 2.9279054640069386, "grad_norm": 0.4014188349246979, "learning_rate": 4.82373422312491e-07, "loss": 0.0126, "step": 27007 }, { "epoch": 2.928013876843018, "grad_norm": 1.3948265314102173, "learning_rate": 4.816480487451037e-07, "loss": 0.0382, "step": 27008 }, { "epoch": 2.928122289679098, "grad_norm": 0.28461042046546936, "learning_rate": 4.809226751777165e-07, "loss": 0.0055, "step": 27009 }, { "epoch": 2.9282307025151777, "grad_norm": 0.5967814922332764, "learning_rate": 4.801973016103294e-07, "loss": 0.0208, "step": 27010 }, { "epoch": 2.9283391153512577, "grad_norm": 0.1647510975599289, "learning_rate": 4.794719280429422e-07, "loss": 0.0101, "step": 27011 }, { "epoch": 2.9284475281873372, "grad_norm": 0.019799770787358284, "learning_rate": 4.78746554475555e-07, "loss": 0.0002, "step": 27012 }, { "epoch": 2.928555941023417, "grad_norm": 0.408250093460083, "learning_rate": 4.780211809081677e-07, "loss": 0.0067, "step": 27013 }, { "epoch": 2.9286643538594967, "grad_norm": 1.9098392724990845, "learning_rate": 4.772958073407805e-07, "loss": 0.0213, "step": 27014 }, { "epoch": 2.9287727666955767, "grad_norm": 0.012959023006260395, "learning_rate": 4.7657043377339335e-07, "loss": 0.0002, "step": 27015 }, { "epoch": 2.9288811795316567, "grad_norm": 2.0379421710968018, "learning_rate": 4.7584506020600613e-07, "loss": 0.0111, "step": 27016 }, { "epoch": 2.9289895923677363, "grad_norm": 0.2654326856136322, "learning_rate": 4.7511968663861896e-07, "loss": 0.0071, "step": 27017 }, { "epoch": 2.9290980052038162, "grad_norm": 0.573223888874054, "learning_rate": 4.7439431307123174e-07, "loss": 0.0232, "step": 27018 }, { "epoch": 2.929206418039896, "grad_norm": 0.47894760966300964, "learning_rate": 4.7366893950384447e-07, "loss": 0.0057, "step": 27019 }, { "epoch": 2.9293148308759758, "grad_norm": 0.06755619496107101, "learning_rate": 4.729435659364573e-07, "loss": 0.0005, "step": 27020 }, { "epoch": 2.9294232437120558, "grad_norm": 0.34496060013771057, "learning_rate": 4.722181923690701e-07, "loss": 0.0048, "step": 27021 }, { "epoch": 2.9295316565481353, "grad_norm": 0.0468946173787117, "learning_rate": 4.714928188016829e-07, "loss": 0.0004, "step": 27022 }, { "epoch": 2.929640069384215, "grad_norm": 2.196981191635132, "learning_rate": 4.707674452342957e-07, "loss": 0.0184, "step": 27023 }, { "epoch": 2.929748482220295, "grad_norm": 2.3589999675750732, "learning_rate": 4.700420716669085e-07, "loss": 0.0405, "step": 27024 }, { "epoch": 2.929856895056375, "grad_norm": 6.429651737213135, "learning_rate": 4.693166980995213e-07, "loss": 0.0229, "step": 27025 }, { "epoch": 2.9299653078924544, "grad_norm": 0.07010669261217117, "learning_rate": 4.685913245321341e-07, "loss": 0.0007, "step": 27026 }, { "epoch": 2.9300737207285343, "grad_norm": 0.5758549571037292, "learning_rate": 4.6786595096474686e-07, "loss": 0.0081, "step": 27027 }, { "epoch": 2.930182133564614, "grad_norm": 0.4611220955848694, "learning_rate": 4.671405773973597e-07, "loss": 0.0129, "step": 27028 }, { "epoch": 2.930290546400694, "grad_norm": 0.4071516990661621, "learning_rate": 4.6641520382997247e-07, "loss": 0.0225, "step": 27029 }, { "epoch": 2.930398959236774, "grad_norm": 7.733534812927246, "learning_rate": 4.656898302625853e-07, "loss": 0.0418, "step": 27030 }, { "epoch": 2.9305073720728534, "grad_norm": 0.7276251316070557, "learning_rate": 4.649644566951981e-07, "loss": 0.0212, "step": 27031 }, { "epoch": 2.9306157849089334, "grad_norm": 1.6137795448303223, "learning_rate": 4.642390831278109e-07, "loss": 0.0345, "step": 27032 }, { "epoch": 2.930724197745013, "grad_norm": 0.1554100215435028, "learning_rate": 4.6351370956042363e-07, "loss": 0.0006, "step": 27033 }, { "epoch": 2.930832610581093, "grad_norm": 0.8856426477432251, "learning_rate": 4.627883359930364e-07, "loss": 0.035, "step": 27034 }, { "epoch": 2.9309410234171724, "grad_norm": 0.41855087876319885, "learning_rate": 4.6206296242564924e-07, "loss": 0.0061, "step": 27035 }, { "epoch": 2.9310494362532524, "grad_norm": 0.42033112049102783, "learning_rate": 4.61337588858262e-07, "loss": 0.0169, "step": 27036 }, { "epoch": 2.931157849089332, "grad_norm": 0.16354414820671082, "learning_rate": 4.6061221529087485e-07, "loss": 0.0009, "step": 27037 }, { "epoch": 2.931266261925412, "grad_norm": 2.6723713874816895, "learning_rate": 4.5988684172348763e-07, "loss": 0.0405, "step": 27038 }, { "epoch": 2.931374674761492, "grad_norm": 0.29570478200912476, "learning_rate": 4.591614681561004e-07, "loss": 0.0061, "step": 27039 }, { "epoch": 2.9314830875975715, "grad_norm": 0.025356626138091087, "learning_rate": 4.584360945887132e-07, "loss": 0.0002, "step": 27040 }, { "epoch": 2.9315915004336515, "grad_norm": 0.033547282218933105, "learning_rate": 4.57710721021326e-07, "loss": 0.0002, "step": 27041 }, { "epoch": 2.931699913269731, "grad_norm": 0.15775878727436066, "learning_rate": 4.569853474539388e-07, "loss": 0.0106, "step": 27042 }, { "epoch": 2.931808326105811, "grad_norm": 3.209155559539795, "learning_rate": 4.5625997388655163e-07, "loss": 0.0052, "step": 27043 }, { "epoch": 2.931916738941891, "grad_norm": 0.07431504130363464, "learning_rate": 4.555346003191644e-07, "loss": 0.0006, "step": 27044 }, { "epoch": 2.9320251517779705, "grad_norm": 5.30570650100708, "learning_rate": 4.5480922675177724e-07, "loss": 0.0093, "step": 27045 }, { "epoch": 2.93213356461405, "grad_norm": 0.4410160779953003, "learning_rate": 4.5408385318438997e-07, "loss": 0.0074, "step": 27046 }, { "epoch": 2.93224197745013, "grad_norm": 0.9069732427597046, "learning_rate": 4.5335847961700275e-07, "loss": 0.026, "step": 27047 }, { "epoch": 2.93235039028621, "grad_norm": 6.934176445007324, "learning_rate": 4.526331060496156e-07, "loss": 0.012, "step": 27048 }, { "epoch": 2.9324588031222896, "grad_norm": 0.048159755766391754, "learning_rate": 4.519077324822284e-07, "loss": 0.0005, "step": 27049 }, { "epoch": 2.9325672159583696, "grad_norm": 2.855475425720215, "learning_rate": 4.511823589148412e-07, "loss": 0.0185, "step": 27050 }, { "epoch": 2.932675628794449, "grad_norm": 0.009796824306249619, "learning_rate": 4.50456985347454e-07, "loss": 0.0001, "step": 27051 }, { "epoch": 2.932784041630529, "grad_norm": 0.17634405195713043, "learning_rate": 4.497316117800668e-07, "loss": 0.0005, "step": 27052 }, { "epoch": 2.932892454466609, "grad_norm": 2.523702383041382, "learning_rate": 4.490062382126795e-07, "loss": 0.0302, "step": 27053 }, { "epoch": 2.9330008673026886, "grad_norm": 0.6213107705116272, "learning_rate": 4.4828086464529236e-07, "loss": 0.0083, "step": 27054 }, { "epoch": 2.9331092801387686, "grad_norm": 4.3949737548828125, "learning_rate": 4.4755549107790513e-07, "loss": 0.0154, "step": 27055 }, { "epoch": 2.933217692974848, "grad_norm": 0.31798040866851807, "learning_rate": 4.4683011751051797e-07, "loss": 0.0067, "step": 27056 }, { "epoch": 2.933326105810928, "grad_norm": 0.2601669430732727, "learning_rate": 4.4610474394313075e-07, "loss": 0.0054, "step": 27057 }, { "epoch": 2.9334345186470077, "grad_norm": 0.5003427863121033, "learning_rate": 4.453793703757436e-07, "loss": 0.0203, "step": 27058 }, { "epoch": 2.9335429314830876, "grad_norm": 0.7397051453590393, "learning_rate": 4.4465399680835636e-07, "loss": 0.018, "step": 27059 }, { "epoch": 2.933651344319167, "grad_norm": 0.09604526311159134, "learning_rate": 4.4392862324096913e-07, "loss": 0.0007, "step": 27060 }, { "epoch": 2.933759757155247, "grad_norm": 3.1772541999816895, "learning_rate": 4.432032496735819e-07, "loss": 0.0413, "step": 27061 }, { "epoch": 2.933868169991327, "grad_norm": 18.7031192779541, "learning_rate": 4.4247787610619474e-07, "loss": 0.0218, "step": 27062 }, { "epoch": 2.9339765828274067, "grad_norm": 0.05003669112920761, "learning_rate": 4.417525025388075e-07, "loss": 0.0003, "step": 27063 }, { "epoch": 2.9340849956634867, "grad_norm": 0.2554817497730255, "learning_rate": 4.4102712897142035e-07, "loss": 0.0054, "step": 27064 }, { "epoch": 2.9341934084995662, "grad_norm": 0.8740665912628174, "learning_rate": 4.4030175540403313e-07, "loss": 0.021, "step": 27065 }, { "epoch": 2.934301821335646, "grad_norm": 9.284013748168945, "learning_rate": 4.3957638183664586e-07, "loss": 0.0307, "step": 27066 }, { "epoch": 2.934410234171726, "grad_norm": 5.260746955871582, "learning_rate": 4.388510082692587e-07, "loss": 0.0383, "step": 27067 }, { "epoch": 2.9345186470078057, "grad_norm": 0.3129081428050995, "learning_rate": 4.3812563470187147e-07, "loss": 0.0065, "step": 27068 }, { "epoch": 2.9346270598438853, "grad_norm": 0.7456008791923523, "learning_rate": 4.374002611344843e-07, "loss": 0.0141, "step": 27069 }, { "epoch": 2.9347354726799653, "grad_norm": 0.6872161626815796, "learning_rate": 4.366748875670971e-07, "loss": 0.0171, "step": 27070 }, { "epoch": 2.9348438855160452, "grad_norm": 0.6082123517990112, "learning_rate": 4.359495139997099e-07, "loss": 0.0087, "step": 27071 }, { "epoch": 2.934952298352125, "grad_norm": 0.8659010529518127, "learning_rate": 4.352241404323227e-07, "loss": 0.0136, "step": 27072 }, { "epoch": 2.9350607111882048, "grad_norm": 0.12861193716526031, "learning_rate": 4.3449876686493547e-07, "loss": 0.0008, "step": 27073 }, { "epoch": 2.9351691240242843, "grad_norm": 0.282266765832901, "learning_rate": 4.3377339329754825e-07, "loss": 0.0052, "step": 27074 }, { "epoch": 2.9352775368603643, "grad_norm": 2.631819725036621, "learning_rate": 4.330480197301611e-07, "loss": 0.0235, "step": 27075 }, { "epoch": 2.9353859496964443, "grad_norm": 3.2472758293151855, "learning_rate": 4.3232264616277386e-07, "loss": 0.0232, "step": 27076 }, { "epoch": 2.935494362532524, "grad_norm": 0.3039340078830719, "learning_rate": 4.315972725953867e-07, "loss": 0.0058, "step": 27077 }, { "epoch": 2.9356027753686034, "grad_norm": 0.6631530523300171, "learning_rate": 4.3087189902799947e-07, "loss": 0.0169, "step": 27078 }, { "epoch": 2.9357111882046834, "grad_norm": 0.54188072681427, "learning_rate": 4.301465254606123e-07, "loss": 0.0007, "step": 27079 }, { "epoch": 2.9358196010407633, "grad_norm": 0.03632272779941559, "learning_rate": 4.29421151893225e-07, "loss": 0.0003, "step": 27080 }, { "epoch": 2.935928013876843, "grad_norm": 0.771677553653717, "learning_rate": 4.286957783258378e-07, "loss": 0.0128, "step": 27081 }, { "epoch": 2.936036426712923, "grad_norm": 0.24037933349609375, "learning_rate": 4.2797040475845064e-07, "loss": 0.0026, "step": 27082 }, { "epoch": 2.9361448395490024, "grad_norm": 21.30837631225586, "learning_rate": 4.272450311910634e-07, "loss": 0.0494, "step": 27083 }, { "epoch": 2.9362532523850824, "grad_norm": 0.11816111207008362, "learning_rate": 4.2651965762367625e-07, "loss": 0.0006, "step": 27084 }, { "epoch": 2.9363616652211624, "grad_norm": 0.6086007356643677, "learning_rate": 4.25794284056289e-07, "loss": 0.0257, "step": 27085 }, { "epoch": 2.936470078057242, "grad_norm": 0.15038400888442993, "learning_rate": 4.2506891048890186e-07, "loss": 0.0007, "step": 27086 }, { "epoch": 2.936578490893322, "grad_norm": 0.3671441078186035, "learning_rate": 4.243435369215146e-07, "loss": 0.0073, "step": 27087 }, { "epoch": 2.9366869037294014, "grad_norm": 0.6527271866798401, "learning_rate": 4.236181633541274e-07, "loss": 0.0006, "step": 27088 }, { "epoch": 2.9367953165654814, "grad_norm": 0.053726792335510254, "learning_rate": 4.228927897867402e-07, "loss": 0.0004, "step": 27089 }, { "epoch": 2.9369037294015614, "grad_norm": 4.736895561218262, "learning_rate": 4.22167416219353e-07, "loss": 0.0318, "step": 27090 }, { "epoch": 2.937012142237641, "grad_norm": 0.4188246428966522, "learning_rate": 4.214420426519658e-07, "loss": 0.0075, "step": 27091 }, { "epoch": 2.9371205550737205, "grad_norm": 11.669656753540039, "learning_rate": 4.2071666908457863e-07, "loss": 0.018, "step": 27092 }, { "epoch": 2.9372289679098005, "grad_norm": 3.4462029933929443, "learning_rate": 4.199912955171914e-07, "loss": 0.004, "step": 27093 }, { "epoch": 2.9373373807458805, "grad_norm": 31.000932693481445, "learning_rate": 4.1926592194980414e-07, "loss": 0.0126, "step": 27094 }, { "epoch": 2.93744579358196, "grad_norm": 0.8996662497520447, "learning_rate": 4.1854054838241697e-07, "loss": 0.0076, "step": 27095 }, { "epoch": 2.93755420641804, "grad_norm": 2.5619983673095703, "learning_rate": 4.1781517481502975e-07, "loss": 0.0071, "step": 27096 }, { "epoch": 2.9376626192541195, "grad_norm": 0.03577321395277977, "learning_rate": 4.170898012476426e-07, "loss": 0.0003, "step": 27097 }, { "epoch": 2.9377710320901995, "grad_norm": 0.3939119577407837, "learning_rate": 4.1636442768025536e-07, "loss": 0.0044, "step": 27098 }, { "epoch": 2.9378794449262795, "grad_norm": 3.72833514213562, "learning_rate": 4.156390541128682e-07, "loss": 0.0357, "step": 27099 }, { "epoch": 2.937987857762359, "grad_norm": 0.42203548550605774, "learning_rate": 4.149136805454809e-07, "loss": 0.0231, "step": 27100 }, { "epoch": 2.9380962705984386, "grad_norm": 1.3832296133041382, "learning_rate": 4.1418830697809375e-07, "loss": 0.0205, "step": 27101 }, { "epoch": 2.9382046834345186, "grad_norm": 4.142504692077637, "learning_rate": 4.134629334107065e-07, "loss": 0.0432, "step": 27102 }, { "epoch": 2.9383130962705986, "grad_norm": 0.051906049251556396, "learning_rate": 4.1273755984331936e-07, "loss": 0.0002, "step": 27103 }, { "epoch": 2.938421509106678, "grad_norm": 2.0111379623413086, "learning_rate": 4.1201218627593214e-07, "loss": 0.0043, "step": 27104 }, { "epoch": 2.938529921942758, "grad_norm": 0.158009335398674, "learning_rate": 4.1128681270854497e-07, "loss": 0.0005, "step": 27105 }, { "epoch": 2.9386383347788376, "grad_norm": 0.7868201732635498, "learning_rate": 4.1056143914115775e-07, "loss": 0.0114, "step": 27106 }, { "epoch": 2.9387467476149176, "grad_norm": 0.34641900658607483, "learning_rate": 4.0983606557377047e-07, "loss": 0.0005, "step": 27107 }, { "epoch": 2.9388551604509976, "grad_norm": 0.02850499376654625, "learning_rate": 4.091106920063833e-07, "loss": 0.0003, "step": 27108 }, { "epoch": 2.938963573287077, "grad_norm": 0.5941617488861084, "learning_rate": 4.083853184389961e-07, "loss": 0.013, "step": 27109 }, { "epoch": 2.939071986123157, "grad_norm": 2.9808132648468018, "learning_rate": 4.076599448716089e-07, "loss": 0.0681, "step": 27110 }, { "epoch": 2.9391803989592367, "grad_norm": 0.8025034666061401, "learning_rate": 4.069345713042217e-07, "loss": 0.0257, "step": 27111 }, { "epoch": 2.9392888117953166, "grad_norm": 2.3881285190582275, "learning_rate": 4.062091977368345e-07, "loss": 0.0226, "step": 27112 }, { "epoch": 2.9393972246313966, "grad_norm": 2.152644395828247, "learning_rate": 4.054838241694473e-07, "loss": 0.007, "step": 27113 }, { "epoch": 2.939505637467476, "grad_norm": 3.8472254276275635, "learning_rate": 4.047584506020601e-07, "loss": 0.0267, "step": 27114 }, { "epoch": 2.9396140503035557, "grad_norm": 0.22840610146522522, "learning_rate": 4.0403307703467286e-07, "loss": 0.003, "step": 27115 }, { "epoch": 2.9397224631396357, "grad_norm": 3.2920124530792236, "learning_rate": 4.033077034672857e-07, "loss": 0.0485, "step": 27116 }, { "epoch": 2.9398308759757157, "grad_norm": 0.2755591571331024, "learning_rate": 4.0258232989989847e-07, "loss": 0.0056, "step": 27117 }, { "epoch": 2.9399392888117952, "grad_norm": 0.050681617110967636, "learning_rate": 4.018569563325113e-07, "loss": 0.0003, "step": 27118 }, { "epoch": 2.940047701647875, "grad_norm": 0.7664209008216858, "learning_rate": 4.011315827651241e-07, "loss": 0.0185, "step": 27119 }, { "epoch": 2.9401561144839548, "grad_norm": 0.05482024699449539, "learning_rate": 4.004062091977369e-07, "loss": 0.0003, "step": 27120 }, { "epoch": 2.9402645273200347, "grad_norm": 0.7122483253479004, "learning_rate": 3.9968083563034964e-07, "loss": 0.0212, "step": 27121 }, { "epoch": 2.9403729401561147, "grad_norm": 8.357519149780273, "learning_rate": 3.989554620629624e-07, "loss": 0.0049, "step": 27122 }, { "epoch": 2.9404813529921943, "grad_norm": 0.4719240665435791, "learning_rate": 3.9823008849557525e-07, "loss": 0.0115, "step": 27123 }, { "epoch": 2.940589765828274, "grad_norm": 0.09375964105129242, "learning_rate": 3.975047149281881e-07, "loss": 0.0006, "step": 27124 }, { "epoch": 2.940698178664354, "grad_norm": 0.009832649491727352, "learning_rate": 3.9677934136080086e-07, "loss": 0.0001, "step": 27125 }, { "epoch": 2.9408065915004338, "grad_norm": 0.6902312636375427, "learning_rate": 3.960539677934137e-07, "loss": 0.0112, "step": 27126 }, { "epoch": 2.9409150043365133, "grad_norm": 1.8235642910003662, "learning_rate": 3.953285942260264e-07, "loss": 0.0494, "step": 27127 }, { "epoch": 2.9410234171725933, "grad_norm": 0.29906874895095825, "learning_rate": 3.946032206586392e-07, "loss": 0.0061, "step": 27128 }, { "epoch": 2.941131830008673, "grad_norm": 4.163552761077881, "learning_rate": 3.9387784709125203e-07, "loss": 0.0362, "step": 27129 }, { "epoch": 2.941240242844753, "grad_norm": 0.30089619755744934, "learning_rate": 3.931524735238648e-07, "loss": 0.0103, "step": 27130 }, { "epoch": 2.941348655680833, "grad_norm": 0.3496907949447632, "learning_rate": 3.9242709995647764e-07, "loss": 0.0083, "step": 27131 }, { "epoch": 2.9414570685169124, "grad_norm": 1.8489775657653809, "learning_rate": 3.917017263890904e-07, "loss": 0.0014, "step": 27132 }, { "epoch": 2.9415654813529923, "grad_norm": 0.6470617651939392, "learning_rate": 3.9097635282170325e-07, "loss": 0.002, "step": 27133 }, { "epoch": 2.941673894189072, "grad_norm": 0.4659605324268341, "learning_rate": 3.9025097925431597e-07, "loss": 0.0114, "step": 27134 }, { "epoch": 2.941782307025152, "grad_norm": 0.39262309670448303, "learning_rate": 3.895256056869288e-07, "loss": 0.0145, "step": 27135 }, { "epoch": 2.9418907198612314, "grad_norm": 0.5396705269813538, "learning_rate": 3.888002321195416e-07, "loss": 0.0066, "step": 27136 }, { "epoch": 2.9419991326973114, "grad_norm": 1.6666597127914429, "learning_rate": 3.880748585521544e-07, "loss": 0.0494, "step": 27137 }, { "epoch": 2.942107545533391, "grad_norm": 0.6176185011863708, "learning_rate": 3.873494849847672e-07, "loss": 0.0109, "step": 27138 }, { "epoch": 2.942215958369471, "grad_norm": 0.9811189770698547, "learning_rate": 3.8662411141738e-07, "loss": 0.0431, "step": 27139 }, { "epoch": 2.942324371205551, "grad_norm": 0.12982626259326935, "learning_rate": 3.858987378499928e-07, "loss": 0.0004, "step": 27140 }, { "epoch": 2.9424327840416304, "grad_norm": 1.8693714141845703, "learning_rate": 3.8517336428260553e-07, "loss": 0.0477, "step": 27141 }, { "epoch": 2.9425411968777104, "grad_norm": 0.007501685060560703, "learning_rate": 3.8444799071521836e-07, "loss": 0.0, "step": 27142 }, { "epoch": 2.94264960971379, "grad_norm": 1.5649415254592896, "learning_rate": 3.8372261714783114e-07, "loss": 0.0305, "step": 27143 }, { "epoch": 2.94275802254987, "grad_norm": 0.818666398525238, "learning_rate": 3.8299724358044397e-07, "loss": 0.0369, "step": 27144 }, { "epoch": 2.94286643538595, "grad_norm": 0.5898008942604065, "learning_rate": 3.8227187001305675e-07, "loss": 0.0023, "step": 27145 }, { "epoch": 2.9429748482220295, "grad_norm": 0.08206147700548172, "learning_rate": 3.815464964456696e-07, "loss": 0.0005, "step": 27146 }, { "epoch": 2.943083261058109, "grad_norm": 0.0120886554941535, "learning_rate": 3.8082112287828236e-07, "loss": 0.0001, "step": 27147 }, { "epoch": 2.943191673894189, "grad_norm": 0.02289474941790104, "learning_rate": 3.8009574931089514e-07, "loss": 0.0002, "step": 27148 }, { "epoch": 2.943300086730269, "grad_norm": 0.038566380739212036, "learning_rate": 3.793703757435079e-07, "loss": 0.0002, "step": 27149 }, { "epoch": 2.9434084995663485, "grad_norm": 0.3209592401981354, "learning_rate": 3.7864500217612075e-07, "loss": 0.0066, "step": 27150 }, { "epoch": 2.9435169124024285, "grad_norm": 0.5091565251350403, "learning_rate": 3.7791962860873353e-07, "loss": 0.0187, "step": 27151 }, { "epoch": 2.943625325238508, "grad_norm": 5.654763698577881, "learning_rate": 3.7719425504134636e-07, "loss": 0.0489, "step": 27152 }, { "epoch": 2.943733738074588, "grad_norm": 0.3558136820793152, "learning_rate": 3.7646888147395914e-07, "loss": 0.0062, "step": 27153 }, { "epoch": 2.943842150910668, "grad_norm": 0.9759047627449036, "learning_rate": 3.7574350790657197e-07, "loss": 0.0286, "step": 27154 }, { "epoch": 2.9439505637467476, "grad_norm": 1.8334866762161255, "learning_rate": 3.750181343391847e-07, "loss": 0.011, "step": 27155 }, { "epoch": 2.9440589765828276, "grad_norm": 2.4213032722473145, "learning_rate": 3.742927607717975e-07, "loss": 0.0168, "step": 27156 }, { "epoch": 2.944167389418907, "grad_norm": 9.115866661071777, "learning_rate": 3.735673872044103e-07, "loss": 0.0365, "step": 27157 }, { "epoch": 2.944275802254987, "grad_norm": 0.2652527987957001, "learning_rate": 3.728420136370231e-07, "loss": 0.005, "step": 27158 }, { "epoch": 2.9443842150910666, "grad_norm": 0.6293290853500366, "learning_rate": 3.721166400696359e-07, "loss": 0.0167, "step": 27159 }, { "epoch": 2.9444926279271466, "grad_norm": 0.6457228660583496, "learning_rate": 3.713912665022487e-07, "loss": 0.0099, "step": 27160 }, { "epoch": 2.944601040763226, "grad_norm": 0.5505417585372925, "learning_rate": 3.7066589293486147e-07, "loss": 0.0009, "step": 27161 }, { "epoch": 2.944709453599306, "grad_norm": 0.28278329968452454, "learning_rate": 3.6994051936747425e-07, "loss": 0.0062, "step": 27162 }, { "epoch": 2.944817866435386, "grad_norm": 0.5700225830078125, "learning_rate": 3.692151458000871e-07, "loss": 0.0127, "step": 27163 }, { "epoch": 2.9449262792714657, "grad_norm": 0.8399396538734436, "learning_rate": 3.6848977223269986e-07, "loss": 0.0278, "step": 27164 }, { "epoch": 2.9450346921075456, "grad_norm": 0.3808349072933197, "learning_rate": 3.677643986653127e-07, "loss": 0.0094, "step": 27165 }, { "epoch": 2.945143104943625, "grad_norm": 2.0454862117767334, "learning_rate": 3.6703902509792547e-07, "loss": 0.0461, "step": 27166 }, { "epoch": 2.945251517779705, "grad_norm": 2.8600082397460938, "learning_rate": 3.663136515305383e-07, "loss": 0.0213, "step": 27167 }, { "epoch": 2.945359930615785, "grad_norm": 3.0997817516326904, "learning_rate": 3.6558827796315103e-07, "loss": 0.0217, "step": 27168 }, { "epoch": 2.9454683434518647, "grad_norm": 5.827300071716309, "learning_rate": 3.648629043957638e-07, "loss": 0.0221, "step": 27169 }, { "epoch": 2.9455767562879442, "grad_norm": 0.017791016027331352, "learning_rate": 3.6413753082837664e-07, "loss": 0.0001, "step": 27170 }, { "epoch": 2.9456851691240242, "grad_norm": 0.013679016381502151, "learning_rate": 3.634121572609894e-07, "loss": 0.0002, "step": 27171 }, { "epoch": 2.945793581960104, "grad_norm": 4.364699840545654, "learning_rate": 3.6268678369360225e-07, "loss": 0.0126, "step": 27172 }, { "epoch": 2.9459019947961838, "grad_norm": 4.7659502029418945, "learning_rate": 3.6196141012621503e-07, "loss": 0.0169, "step": 27173 }, { "epoch": 2.9460104076322637, "grad_norm": 0.0539126843214035, "learning_rate": 3.6123603655882786e-07, "loss": 0.0004, "step": 27174 }, { "epoch": 2.9461188204683433, "grad_norm": 0.006748898420482874, "learning_rate": 3.605106629914406e-07, "loss": 0.0001, "step": 27175 }, { "epoch": 2.9462272333044233, "grad_norm": 0.09564793109893799, "learning_rate": 3.597852894240534e-07, "loss": 0.0005, "step": 27176 }, { "epoch": 2.9463356461405033, "grad_norm": 0.012977696023881435, "learning_rate": 3.590599158566662e-07, "loss": 0.0001, "step": 27177 }, { "epoch": 2.946444058976583, "grad_norm": 2.177616596221924, "learning_rate": 3.5833454228927903e-07, "loss": 0.0092, "step": 27178 }, { "epoch": 2.9465524718126628, "grad_norm": 1.157444715499878, "learning_rate": 3.576091687218918e-07, "loss": 0.0118, "step": 27179 }, { "epoch": 2.9466608846487423, "grad_norm": 0.873485267162323, "learning_rate": 3.5688379515450464e-07, "loss": 0.0135, "step": 27180 }, { "epoch": 2.9467692974848223, "grad_norm": 0.047703538089990616, "learning_rate": 3.561584215871174e-07, "loss": 0.0003, "step": 27181 }, { "epoch": 2.946877710320902, "grad_norm": 0.3324185609817505, "learning_rate": 3.5543304801973014e-07, "loss": 0.0117, "step": 27182 }, { "epoch": 2.946986123156982, "grad_norm": 0.1842227727174759, "learning_rate": 3.54707674452343e-07, "loss": 0.0007, "step": 27183 }, { "epoch": 2.9470945359930614, "grad_norm": 0.01884007826447487, "learning_rate": 3.5398230088495575e-07, "loss": 0.0001, "step": 27184 }, { "epoch": 2.9472029488291414, "grad_norm": 2.5935540199279785, "learning_rate": 3.532569273175686e-07, "loss": 0.0526, "step": 27185 }, { "epoch": 2.9473113616652213, "grad_norm": 0.2979026138782501, "learning_rate": 3.5253155375018136e-07, "loss": 0.0061, "step": 27186 }, { "epoch": 2.947419774501301, "grad_norm": 2.767510414123535, "learning_rate": 3.518061801827942e-07, "loss": 0.0171, "step": 27187 }, { "epoch": 2.947528187337381, "grad_norm": 0.16561679542064667, "learning_rate": 3.510808066154069e-07, "loss": 0.0006, "step": 27188 }, { "epoch": 2.9476366001734604, "grad_norm": 2.123383045196533, "learning_rate": 3.5035543304801975e-07, "loss": 0.0449, "step": 27189 }, { "epoch": 2.9477450130095404, "grad_norm": 0.04156927391886711, "learning_rate": 3.4963005948063253e-07, "loss": 0.0003, "step": 27190 }, { "epoch": 2.9478534258456204, "grad_norm": 0.27469968795776367, "learning_rate": 3.4890468591324536e-07, "loss": 0.0064, "step": 27191 }, { "epoch": 2.9479618386817, "grad_norm": 0.3458973169326782, "learning_rate": 3.4817931234585814e-07, "loss": 0.0101, "step": 27192 }, { "epoch": 2.9480702515177795, "grad_norm": 4.679110527038574, "learning_rate": 3.4745393877847097e-07, "loss": 0.0146, "step": 27193 }, { "epoch": 2.9481786643538594, "grad_norm": 0.519190788269043, "learning_rate": 3.4672856521108375e-07, "loss": 0.0092, "step": 27194 }, { "epoch": 2.9482870771899394, "grad_norm": 1.9237284660339355, "learning_rate": 3.460031916436965e-07, "loss": 0.0221, "step": 27195 }, { "epoch": 2.948395490026019, "grad_norm": 0.4604342579841614, "learning_rate": 3.452778180763093e-07, "loss": 0.0027, "step": 27196 }, { "epoch": 2.948503902862099, "grad_norm": 3.645200729370117, "learning_rate": 3.445524445089221e-07, "loss": 0.0261, "step": 27197 }, { "epoch": 2.9486123156981785, "grad_norm": 0.27447202801704407, "learning_rate": 3.438270709415349e-07, "loss": 0.0057, "step": 27198 }, { "epoch": 2.9487207285342585, "grad_norm": 0.02867092192173004, "learning_rate": 3.431016973741477e-07, "loss": 0.0003, "step": 27199 }, { "epoch": 2.9488291413703385, "grad_norm": 0.2022133618593216, "learning_rate": 3.4237632380676053e-07, "loss": 0.0006, "step": 27200 }, { "epoch": 2.948937554206418, "grad_norm": 19.99180030822754, "learning_rate": 3.4165095023937336e-07, "loss": 0.0444, "step": 27201 }, { "epoch": 2.949045967042498, "grad_norm": 3.193622350692749, "learning_rate": 3.409255766719861e-07, "loss": 0.0126, "step": 27202 }, { "epoch": 2.9491543798785775, "grad_norm": 0.43293505907058716, "learning_rate": 3.4020020310459887e-07, "loss": 0.0056, "step": 27203 }, { "epoch": 2.9492627927146575, "grad_norm": 0.03882741928100586, "learning_rate": 3.394748295372117e-07, "loss": 0.0003, "step": 27204 }, { "epoch": 2.949371205550737, "grad_norm": 0.7144907116889954, "learning_rate": 3.387494559698245e-07, "loss": 0.0152, "step": 27205 }, { "epoch": 2.949479618386817, "grad_norm": 0.13709455728530884, "learning_rate": 3.380240824024373e-07, "loss": 0.0007, "step": 27206 }, { "epoch": 2.9495880312228966, "grad_norm": 0.2674829661846161, "learning_rate": 3.372987088350501e-07, "loss": 0.0056, "step": 27207 }, { "epoch": 2.9496964440589766, "grad_norm": 5.248715877532959, "learning_rate": 3.365733352676629e-07, "loss": 0.0683, "step": 27208 }, { "epoch": 2.9498048568950566, "grad_norm": 2.1511476039886475, "learning_rate": 3.3584796170027564e-07, "loss": 0.0329, "step": 27209 }, { "epoch": 2.949913269731136, "grad_norm": 5.38840389251709, "learning_rate": 3.351225881328885e-07, "loss": 0.012, "step": 27210 }, { "epoch": 2.950021682567216, "grad_norm": 0.5321046113967896, "learning_rate": 3.3439721456550125e-07, "loss": 0.0201, "step": 27211 }, { "epoch": 2.9501300954032956, "grad_norm": 2.1083245277404785, "learning_rate": 3.336718409981141e-07, "loss": 0.0221, "step": 27212 }, { "epoch": 2.9502385082393756, "grad_norm": 0.29573455452919006, "learning_rate": 3.3294646743072686e-07, "loss": 0.006, "step": 27213 }, { "epoch": 2.9503469210754556, "grad_norm": 10.3342924118042, "learning_rate": 3.322210938633397e-07, "loss": 0.0261, "step": 27214 }, { "epoch": 2.950455333911535, "grad_norm": 0.5221237540245056, "learning_rate": 3.314957202959524e-07, "loss": 0.012, "step": 27215 }, { "epoch": 2.9505637467476147, "grad_norm": 0.0581408329308033, "learning_rate": 3.307703467285652e-07, "loss": 0.0003, "step": 27216 }, { "epoch": 2.9506721595836947, "grad_norm": 1.0022940635681152, "learning_rate": 3.3004497316117803e-07, "loss": 0.0067, "step": 27217 }, { "epoch": 2.9507805724197746, "grad_norm": 14.473556518554688, "learning_rate": 3.293195995937908e-07, "loss": 0.0086, "step": 27218 }, { "epoch": 2.950888985255854, "grad_norm": 4.708042621612549, "learning_rate": 3.2859422602640364e-07, "loss": 0.0061, "step": 27219 }, { "epoch": 2.950997398091934, "grad_norm": 0.5390117168426514, "learning_rate": 3.278688524590164e-07, "loss": 0.0081, "step": 27220 }, { "epoch": 2.9511058109280137, "grad_norm": 0.5636038184165955, "learning_rate": 3.2714347889162925e-07, "loss": 0.0088, "step": 27221 }, { "epoch": 2.9512142237640937, "grad_norm": 0.5289714932441711, "learning_rate": 3.26418105324242e-07, "loss": 0.002, "step": 27222 }, { "epoch": 2.9513226366001737, "grad_norm": 0.056605808436870575, "learning_rate": 3.256927317568548e-07, "loss": 0.0002, "step": 27223 }, { "epoch": 2.9514310494362532, "grad_norm": 4.624154090881348, "learning_rate": 3.249673581894676e-07, "loss": 0.0439, "step": 27224 }, { "epoch": 2.9515394622723328, "grad_norm": 0.04080117493867874, "learning_rate": 3.242419846220804e-07, "loss": 0.0002, "step": 27225 }, { "epoch": 2.9516478751084128, "grad_norm": 1.1833041906356812, "learning_rate": 3.235166110546932e-07, "loss": 0.0256, "step": 27226 }, { "epoch": 2.9517562879444927, "grad_norm": 0.042914874851703644, "learning_rate": 3.2279123748730603e-07, "loss": 0.0003, "step": 27227 }, { "epoch": 2.9518647007805723, "grad_norm": 0.20226232707500458, "learning_rate": 3.220658639199188e-07, "loss": 0.0012, "step": 27228 }, { "epoch": 2.9519731136166523, "grad_norm": 0.043157611042261124, "learning_rate": 3.2134049035253153e-07, "loss": 0.0004, "step": 27229 }, { "epoch": 2.952081526452732, "grad_norm": 0.4211382567882538, "learning_rate": 3.2061511678514437e-07, "loss": 0.006, "step": 27230 }, { "epoch": 2.952189939288812, "grad_norm": 16.571094512939453, "learning_rate": 3.1988974321775714e-07, "loss": 0.0289, "step": 27231 }, { "epoch": 2.952298352124892, "grad_norm": 0.4724059998989105, "learning_rate": 3.1916436965037e-07, "loss": 0.0214, "step": 27232 }, { "epoch": 2.9524067649609713, "grad_norm": 0.5730600953102112, "learning_rate": 3.1843899608298275e-07, "loss": 0.0179, "step": 27233 }, { "epoch": 2.9525151777970513, "grad_norm": 0.012244471348822117, "learning_rate": 3.177136225155956e-07, "loss": 0.0001, "step": 27234 }, { "epoch": 2.952623590633131, "grad_norm": 4.167415618896484, "learning_rate": 3.1698824894820836e-07, "loss": 0.0036, "step": 27235 }, { "epoch": 2.952732003469211, "grad_norm": 2.1492395401000977, "learning_rate": 3.1626287538082114e-07, "loss": 0.036, "step": 27236 }, { "epoch": 2.952840416305291, "grad_norm": 4.056890964508057, "learning_rate": 3.155375018134339e-07, "loss": 0.01, "step": 27237 }, { "epoch": 2.9529488291413704, "grad_norm": 0.3957253098487854, "learning_rate": 3.1481212824604675e-07, "loss": 0.0116, "step": 27238 }, { "epoch": 2.95305724197745, "grad_norm": 0.3131093382835388, "learning_rate": 3.1408675467865953e-07, "loss": 0.011, "step": 27239 }, { "epoch": 2.95316565481353, "grad_norm": 0.4467858374118805, "learning_rate": 3.1336138111127236e-07, "loss": 0.0069, "step": 27240 }, { "epoch": 2.95327406764961, "grad_norm": 0.10810906440019608, "learning_rate": 3.1263600754388514e-07, "loss": 0.0005, "step": 27241 }, { "epoch": 2.9533824804856894, "grad_norm": 0.7947126626968384, "learning_rate": 3.119106339764979e-07, "loss": 0.0306, "step": 27242 }, { "epoch": 2.9534908933217694, "grad_norm": 0.2712346613407135, "learning_rate": 3.1118526040911075e-07, "loss": 0.0055, "step": 27243 }, { "epoch": 2.953599306157849, "grad_norm": 0.015511669218540192, "learning_rate": 3.104598868417235e-07, "loss": 0.0002, "step": 27244 }, { "epoch": 2.953707718993929, "grad_norm": 0.01934026926755905, "learning_rate": 3.097345132743363e-07, "loss": 0.0002, "step": 27245 }, { "epoch": 2.953816131830009, "grad_norm": 0.0780174508690834, "learning_rate": 3.090091397069491e-07, "loss": 0.0007, "step": 27246 }, { "epoch": 2.9539245446660884, "grad_norm": 0.013422518037259579, "learning_rate": 3.082837661395619e-07, "loss": 0.0002, "step": 27247 }, { "epoch": 2.954032957502168, "grad_norm": 0.24828597903251648, "learning_rate": 3.075583925721747e-07, "loss": 0.005, "step": 27248 }, { "epoch": 2.954141370338248, "grad_norm": 0.679573118686676, "learning_rate": 3.068330190047875e-07, "loss": 0.0045, "step": 27249 }, { "epoch": 2.954249783174328, "grad_norm": 6.065178871154785, "learning_rate": 3.061076454374003e-07, "loss": 0.0447, "step": 27250 }, { "epoch": 2.9543581960104075, "grad_norm": 0.17714223265647888, "learning_rate": 3.053822718700131e-07, "loss": 0.0003, "step": 27251 }, { "epoch": 2.9544666088464875, "grad_norm": 3.501443862915039, "learning_rate": 3.0465689830262587e-07, "loss": 0.0592, "step": 27252 }, { "epoch": 2.954575021682567, "grad_norm": 0.07758061587810516, "learning_rate": 3.039315247352387e-07, "loss": 0.0007, "step": 27253 }, { "epoch": 2.954683434518647, "grad_norm": 0.6382249593734741, "learning_rate": 3.032061511678515e-07, "loss": 0.0169, "step": 27254 }, { "epoch": 2.954791847354727, "grad_norm": 0.013143797405064106, "learning_rate": 3.0248077760046426e-07, "loss": 0.0002, "step": 27255 }, { "epoch": 2.9549002601908065, "grad_norm": 0.5940830111503601, "learning_rate": 3.017554040330771e-07, "loss": 0.0022, "step": 27256 }, { "epoch": 2.9550086730268865, "grad_norm": 0.2685435116291046, "learning_rate": 3.0103003046568987e-07, "loss": 0.0056, "step": 27257 }, { "epoch": 2.955117085862966, "grad_norm": 0.7129663825035095, "learning_rate": 3.0030465689830264e-07, "loss": 0.0078, "step": 27258 }, { "epoch": 2.955225498699046, "grad_norm": 1.742891550064087, "learning_rate": 2.995792833309154e-07, "loss": 0.0256, "step": 27259 }, { "epoch": 2.955333911535126, "grad_norm": 3.863985300064087, "learning_rate": 2.9885390976352826e-07, "loss": 0.0246, "step": 27260 }, { "epoch": 2.9554423243712056, "grad_norm": 1.928301453590393, "learning_rate": 2.9812853619614103e-07, "loss": 0.0172, "step": 27261 }, { "epoch": 2.955550737207285, "grad_norm": 0.03873310983181, "learning_rate": 2.974031626287538e-07, "loss": 0.0003, "step": 27262 }, { "epoch": 2.955659150043365, "grad_norm": 0.4593258202075958, "learning_rate": 2.9667778906136664e-07, "loss": 0.0006, "step": 27263 }, { "epoch": 2.955767562879445, "grad_norm": 0.3483181595802307, "learning_rate": 2.959524154939794e-07, "loss": 0.0012, "step": 27264 }, { "epoch": 2.9558759757155246, "grad_norm": 0.2441454529762268, "learning_rate": 2.952270419265922e-07, "loss": 0.0018, "step": 27265 }, { "epoch": 2.9559843885516046, "grad_norm": 0.2349405288696289, "learning_rate": 2.9450166835920503e-07, "loss": 0.0035, "step": 27266 }, { "epoch": 2.956092801387684, "grad_norm": 0.08942872285842896, "learning_rate": 2.937762947918178e-07, "loss": 0.0006, "step": 27267 }, { "epoch": 2.956201214223764, "grad_norm": 0.28293225169181824, "learning_rate": 2.930509212244306e-07, "loss": 0.0059, "step": 27268 }, { "epoch": 2.956309627059844, "grad_norm": 10.173566818237305, "learning_rate": 2.923255476570434e-07, "loss": 0.0146, "step": 27269 }, { "epoch": 2.9564180398959237, "grad_norm": 0.1649039089679718, "learning_rate": 2.916001740896562e-07, "loss": 0.0011, "step": 27270 }, { "epoch": 2.956526452732003, "grad_norm": 4.022253513336182, "learning_rate": 2.90874800522269e-07, "loss": 0.0072, "step": 27271 }, { "epoch": 2.956634865568083, "grad_norm": 0.5270164608955383, "learning_rate": 2.9014942695488176e-07, "loss": 0.0017, "step": 27272 }, { "epoch": 2.956743278404163, "grad_norm": 1.0564371347427368, "learning_rate": 2.894240533874946e-07, "loss": 0.0262, "step": 27273 }, { "epoch": 2.9568516912402427, "grad_norm": 2.9009780883789062, "learning_rate": 2.8869867982010737e-07, "loss": 0.0202, "step": 27274 }, { "epoch": 2.9569601040763227, "grad_norm": 0.09780984371900558, "learning_rate": 2.8797330625272015e-07, "loss": 0.0005, "step": 27275 }, { "epoch": 2.9570685169124022, "grad_norm": 0.01749797724187374, "learning_rate": 2.87247932685333e-07, "loss": 0.0001, "step": 27276 }, { "epoch": 2.9571769297484822, "grad_norm": 0.3881690204143524, "learning_rate": 2.8652255911794576e-07, "loss": 0.0072, "step": 27277 }, { "epoch": 2.957285342584562, "grad_norm": 0.4610026478767395, "learning_rate": 2.8579718555055854e-07, "loss": 0.006, "step": 27278 }, { "epoch": 2.9573937554206418, "grad_norm": 0.05383443832397461, "learning_rate": 2.8507181198317137e-07, "loss": 0.0002, "step": 27279 }, { "epoch": 2.9575021682567217, "grad_norm": 0.1432124823331833, "learning_rate": 2.8434643841578415e-07, "loss": 0.0013, "step": 27280 }, { "epoch": 2.9576105810928013, "grad_norm": 0.014293316751718521, "learning_rate": 2.836210648483969e-07, "loss": 0.0001, "step": 27281 }, { "epoch": 2.9577189939288813, "grad_norm": 0.28862178325653076, "learning_rate": 2.8289569128100976e-07, "loss": 0.0061, "step": 27282 }, { "epoch": 2.957827406764961, "grad_norm": 0.9768004417419434, "learning_rate": 2.8217031771362253e-07, "loss": 0.0049, "step": 27283 }, { "epoch": 2.957935819601041, "grad_norm": 0.19477301836013794, "learning_rate": 2.8144494414623537e-07, "loss": 0.0027, "step": 27284 }, { "epoch": 2.9580442324371203, "grad_norm": 1.8732794523239136, "learning_rate": 2.8071957057884815e-07, "loss": 0.0033, "step": 27285 }, { "epoch": 2.9581526452732003, "grad_norm": 0.5072726011276245, "learning_rate": 2.799941970114609e-07, "loss": 0.0017, "step": 27286 }, { "epoch": 2.9582610581092803, "grad_norm": 0.46533021330833435, "learning_rate": 2.7926882344407376e-07, "loss": 0.0005, "step": 27287 }, { "epoch": 2.95836947094536, "grad_norm": 3.2060232162475586, "learning_rate": 2.785434498766865e-07, "loss": 0.038, "step": 27288 }, { "epoch": 2.95847788378144, "grad_norm": 7.965768814086914, "learning_rate": 2.778180763092993e-07, "loss": 0.0024, "step": 27289 }, { "epoch": 2.9585862966175194, "grad_norm": 0.32958030700683594, "learning_rate": 2.770927027419121e-07, "loss": 0.0016, "step": 27290 }, { "epoch": 2.9586947094535994, "grad_norm": 0.506517231464386, "learning_rate": 2.763673291745249e-07, "loss": 0.0198, "step": 27291 }, { "epoch": 2.9588031222896793, "grad_norm": 0.7529715895652771, "learning_rate": 2.756419556071377e-07, "loss": 0.0235, "step": 27292 }, { "epoch": 2.958911535125759, "grad_norm": 0.025285033509135246, "learning_rate": 2.749165820397505e-07, "loss": 0.0002, "step": 27293 }, { "epoch": 2.9590199479618384, "grad_norm": 1.3817821741104126, "learning_rate": 2.741912084723633e-07, "loss": 0.0534, "step": 27294 }, { "epoch": 2.9591283607979184, "grad_norm": 0.3504832088947296, "learning_rate": 2.734658349049761e-07, "loss": 0.0089, "step": 27295 }, { "epoch": 2.9592367736339984, "grad_norm": 0.5759375691413879, "learning_rate": 2.7274046133758887e-07, "loss": 0.0079, "step": 27296 }, { "epoch": 2.959345186470078, "grad_norm": 0.3157768249511719, "learning_rate": 2.720150877702017e-07, "loss": 0.0066, "step": 27297 }, { "epoch": 2.959453599306158, "grad_norm": 0.5234743356704712, "learning_rate": 2.712897142028145e-07, "loss": 0.0088, "step": 27298 }, { "epoch": 2.9595620121422375, "grad_norm": 0.0349605455994606, "learning_rate": 2.7056434063542726e-07, "loss": 0.0003, "step": 27299 }, { "epoch": 2.9596704249783174, "grad_norm": 0.7516613006591797, "learning_rate": 2.698389670680401e-07, "loss": 0.0358, "step": 27300 }, { "epoch": 2.9597788378143974, "grad_norm": 1.6441763639450073, "learning_rate": 2.6911359350065287e-07, "loss": 0.0884, "step": 27301 }, { "epoch": 2.959887250650477, "grad_norm": 0.2125227302312851, "learning_rate": 2.6838821993326565e-07, "loss": 0.0013, "step": 27302 }, { "epoch": 2.959995663486557, "grad_norm": 0.3310636579990387, "learning_rate": 2.676628463658784e-07, "loss": 0.0007, "step": 27303 }, { "epoch": 2.9601040763226365, "grad_norm": 2.0773119926452637, "learning_rate": 2.6693747279849126e-07, "loss": 0.0552, "step": 27304 }, { "epoch": 2.9602124891587165, "grad_norm": 0.18902218341827393, "learning_rate": 2.6621209923110404e-07, "loss": 0.0008, "step": 27305 }, { "epoch": 2.960320901994796, "grad_norm": 0.2688678801059723, "learning_rate": 2.654867256637168e-07, "loss": 0.0056, "step": 27306 }, { "epoch": 2.960429314830876, "grad_norm": 0.08126211911439896, "learning_rate": 2.6476135209632965e-07, "loss": 0.0007, "step": 27307 }, { "epoch": 2.9605377276669556, "grad_norm": 0.413677453994751, "learning_rate": 2.640359785289424e-07, "loss": 0.0133, "step": 27308 }, { "epoch": 2.9606461405030355, "grad_norm": 0.025180023163557053, "learning_rate": 2.633106049615552e-07, "loss": 0.0002, "step": 27309 }, { "epoch": 2.9607545533391155, "grad_norm": 0.34278228878974915, "learning_rate": 2.6258523139416804e-07, "loss": 0.0092, "step": 27310 }, { "epoch": 2.960862966175195, "grad_norm": 0.046858225017786026, "learning_rate": 2.618598578267808e-07, "loss": 0.0002, "step": 27311 }, { "epoch": 2.960971379011275, "grad_norm": 4.138914108276367, "learning_rate": 2.611344842593936e-07, "loss": 0.0466, "step": 27312 }, { "epoch": 2.9610797918473546, "grad_norm": 0.6155410408973694, "learning_rate": 2.604091106920064e-07, "loss": 0.0152, "step": 27313 }, { "epoch": 2.9611882046834346, "grad_norm": 14.006261825561523, "learning_rate": 2.596837371246192e-07, "loss": 0.0173, "step": 27314 }, { "epoch": 2.9612966175195146, "grad_norm": 3.061047315597534, "learning_rate": 2.58958363557232e-07, "loss": 0.0134, "step": 27315 }, { "epoch": 2.961405030355594, "grad_norm": 0.592581570148468, "learning_rate": 2.5823298998984476e-07, "loss": 0.0204, "step": 27316 }, { "epoch": 2.9615134431916736, "grad_norm": 15.633810997009277, "learning_rate": 2.575076164224576e-07, "loss": 0.0343, "step": 27317 }, { "epoch": 2.9616218560277536, "grad_norm": 0.6966272592544556, "learning_rate": 2.567822428550704e-07, "loss": 0.0159, "step": 27318 }, { "epoch": 2.9617302688638336, "grad_norm": 0.024614647030830383, "learning_rate": 2.5605686928768315e-07, "loss": 0.0002, "step": 27319 }, { "epoch": 2.961838681699913, "grad_norm": 0.02875337563455105, "learning_rate": 2.55331495720296e-07, "loss": 0.0002, "step": 27320 }, { "epoch": 2.961947094535993, "grad_norm": 0.5841279625892639, "learning_rate": 2.5460612215290876e-07, "loss": 0.0019, "step": 27321 }, { "epoch": 2.9620555073720727, "grad_norm": 0.21224716305732727, "learning_rate": 2.5388074858552154e-07, "loss": 0.0028, "step": 27322 }, { "epoch": 2.9621639202081527, "grad_norm": 1.5146905183792114, "learning_rate": 2.5315537501813437e-07, "loss": 0.0033, "step": 27323 }, { "epoch": 2.9622723330442327, "grad_norm": 0.013171960599720478, "learning_rate": 2.5243000145074715e-07, "loss": 0.0001, "step": 27324 }, { "epoch": 2.962380745880312, "grad_norm": 0.07643485069274902, "learning_rate": 2.5170462788335993e-07, "loss": 0.0004, "step": 27325 }, { "epoch": 2.962489158716392, "grad_norm": 0.020672926679253578, "learning_rate": 2.5097925431597276e-07, "loss": 0.0001, "step": 27326 }, { "epoch": 2.9625975715524717, "grad_norm": 0.11069754511117935, "learning_rate": 2.5025388074858554e-07, "loss": 0.0005, "step": 27327 }, { "epoch": 2.9627059843885517, "grad_norm": 0.014657439664006233, "learning_rate": 2.4952850718119837e-07, "loss": 0.0002, "step": 27328 }, { "epoch": 2.9628143972246312, "grad_norm": 0.3813193738460541, "learning_rate": 2.4880313361381115e-07, "loss": 0.0127, "step": 27329 }, { "epoch": 2.9629228100607112, "grad_norm": 0.8840283155441284, "learning_rate": 2.480777600464239e-07, "loss": 0.0015, "step": 27330 }, { "epoch": 2.9630312228967908, "grad_norm": 0.2602372169494629, "learning_rate": 2.4735238647903676e-07, "loss": 0.0009, "step": 27331 }, { "epoch": 2.9631396357328708, "grad_norm": 0.10128722339868546, "learning_rate": 2.466270129116495e-07, "loss": 0.0003, "step": 27332 }, { "epoch": 2.9632480485689507, "grad_norm": 0.025264622643589973, "learning_rate": 2.459016393442623e-07, "loss": 0.0002, "step": 27333 }, { "epoch": 2.9633564614050303, "grad_norm": 0.07801955938339233, "learning_rate": 2.451762657768751e-07, "loss": 0.0004, "step": 27334 }, { "epoch": 2.9634648742411103, "grad_norm": 1.2200777530670166, "learning_rate": 2.444508922094879e-07, "loss": 0.004, "step": 27335 }, { "epoch": 2.96357328707719, "grad_norm": 0.018543411046266556, "learning_rate": 2.437255186421007e-07, "loss": 0.0002, "step": 27336 }, { "epoch": 2.96368169991327, "grad_norm": 1.6037696599960327, "learning_rate": 2.430001450747135e-07, "loss": 0.0308, "step": 27337 }, { "epoch": 2.96379011274935, "grad_norm": 1.2315263748168945, "learning_rate": 2.422747715073263e-07, "loss": 0.0018, "step": 27338 }, { "epoch": 2.9638985255854293, "grad_norm": 0.23905380070209503, "learning_rate": 2.415493979399391e-07, "loss": 0.0036, "step": 27339 }, { "epoch": 2.964006938421509, "grad_norm": 7.344292163848877, "learning_rate": 2.4082402437255187e-07, "loss": 0.0386, "step": 27340 }, { "epoch": 2.964115351257589, "grad_norm": 0.44947564601898193, "learning_rate": 2.400986508051647e-07, "loss": 0.0136, "step": 27341 }, { "epoch": 2.964223764093669, "grad_norm": 0.8155444264411926, "learning_rate": 2.393732772377775e-07, "loss": 0.0268, "step": 27342 }, { "epoch": 2.9643321769297484, "grad_norm": 0.4952370822429657, "learning_rate": 2.3864790367039026e-07, "loss": 0.0154, "step": 27343 }, { "epoch": 2.9644405897658284, "grad_norm": 5.294250965118408, "learning_rate": 2.3792253010300307e-07, "loss": 0.0292, "step": 27344 }, { "epoch": 2.964549002601908, "grad_norm": 18.882978439331055, "learning_rate": 2.3719715653561587e-07, "loss": 0.0098, "step": 27345 }, { "epoch": 2.964657415437988, "grad_norm": 0.010250741615891457, "learning_rate": 2.3647178296822865e-07, "loss": 0.0001, "step": 27346 }, { "epoch": 2.964765828274068, "grad_norm": 0.01631268858909607, "learning_rate": 2.3574640940084145e-07, "loss": 0.0002, "step": 27347 }, { "epoch": 2.9648742411101474, "grad_norm": 4.351093292236328, "learning_rate": 2.3502103583345426e-07, "loss": 0.0262, "step": 27348 }, { "epoch": 2.9649826539462274, "grad_norm": 6.808309555053711, "learning_rate": 2.3429566226606704e-07, "loss": 0.0565, "step": 27349 }, { "epoch": 2.965091066782307, "grad_norm": 0.6738161444664001, "learning_rate": 2.3357028869867984e-07, "loss": 0.015, "step": 27350 }, { "epoch": 2.965199479618387, "grad_norm": 1.7458264827728271, "learning_rate": 2.3284491513129265e-07, "loss": 0.0049, "step": 27351 }, { "epoch": 2.9653078924544665, "grad_norm": 0.02014414221048355, "learning_rate": 2.3211954156390545e-07, "loss": 0.0001, "step": 27352 }, { "epoch": 2.9654163052905465, "grad_norm": 4.202301979064941, "learning_rate": 2.313941679965182e-07, "loss": 0.0295, "step": 27353 }, { "epoch": 2.965524718126626, "grad_norm": 0.32256677746772766, "learning_rate": 2.30668794429131e-07, "loss": 0.0106, "step": 27354 }, { "epoch": 2.965633130962706, "grad_norm": 0.026888811960816383, "learning_rate": 2.2994342086174382e-07, "loss": 0.0001, "step": 27355 }, { "epoch": 2.965741543798786, "grad_norm": 2.544161558151245, "learning_rate": 2.292180472943566e-07, "loss": 0.0149, "step": 27356 }, { "epoch": 2.9658499566348655, "grad_norm": 0.4520837962627411, "learning_rate": 2.284926737269694e-07, "loss": 0.0096, "step": 27357 }, { "epoch": 2.9659583694709455, "grad_norm": 0.6777588129043579, "learning_rate": 2.277673001595822e-07, "loss": 0.0112, "step": 27358 }, { "epoch": 2.966066782307025, "grad_norm": 9.547950744628906, "learning_rate": 2.2704192659219498e-07, "loss": 0.0262, "step": 27359 }, { "epoch": 2.966175195143105, "grad_norm": 0.8060020208358765, "learning_rate": 2.263165530248078e-07, "loss": 0.0172, "step": 27360 }, { "epoch": 2.966283607979185, "grad_norm": 0.7693025469779968, "learning_rate": 2.255911794574206e-07, "loss": 0.0144, "step": 27361 }, { "epoch": 2.9663920208152645, "grad_norm": 0.3093617856502533, "learning_rate": 2.248658058900334e-07, "loss": 0.0034, "step": 27362 }, { "epoch": 2.966500433651344, "grad_norm": 0.5874759554862976, "learning_rate": 2.2414043232264618e-07, "loss": 0.0148, "step": 27363 }, { "epoch": 2.966608846487424, "grad_norm": 3.886354684829712, "learning_rate": 2.2341505875525898e-07, "loss": 0.0477, "step": 27364 }, { "epoch": 2.966717259323504, "grad_norm": 0.06648100912570953, "learning_rate": 2.226896851878718e-07, "loss": 0.0003, "step": 27365 }, { "epoch": 2.9668256721595836, "grad_norm": 0.40856942534446716, "learning_rate": 2.2196431162048457e-07, "loss": 0.0036, "step": 27366 }, { "epoch": 2.9669340849956636, "grad_norm": 0.285157710313797, "learning_rate": 2.2123893805309737e-07, "loss": 0.001, "step": 27367 }, { "epoch": 2.967042497831743, "grad_norm": 0.5050784349441528, "learning_rate": 2.2051356448571018e-07, "loss": 0.0095, "step": 27368 }, { "epoch": 2.967150910667823, "grad_norm": 0.47912365198135376, "learning_rate": 2.1978819091832293e-07, "loss": 0.0054, "step": 27369 }, { "epoch": 2.967259323503903, "grad_norm": 0.312375009059906, "learning_rate": 2.1906281735093573e-07, "loss": 0.0032, "step": 27370 }, { "epoch": 2.9673677363399826, "grad_norm": 0.35280588269233704, "learning_rate": 2.1833744378354854e-07, "loss": 0.0055, "step": 27371 }, { "epoch": 2.967476149176062, "grad_norm": 0.036607954651117325, "learning_rate": 2.1761207021616134e-07, "loss": 0.0003, "step": 27372 }, { "epoch": 2.967584562012142, "grad_norm": 0.09690646082162857, "learning_rate": 2.1688669664877412e-07, "loss": 0.0004, "step": 27373 }, { "epoch": 2.967692974848222, "grad_norm": 0.6411353349685669, "learning_rate": 2.1616132308138693e-07, "loss": 0.0076, "step": 27374 }, { "epoch": 2.9678013876843017, "grad_norm": 1.9077948331832886, "learning_rate": 2.1543594951399973e-07, "loss": 0.0708, "step": 27375 }, { "epoch": 2.9679098005203817, "grad_norm": 0.03707750886678696, "learning_rate": 2.147105759466125e-07, "loss": 0.0003, "step": 27376 }, { "epoch": 2.968018213356461, "grad_norm": 0.01001048181205988, "learning_rate": 2.1398520237922532e-07, "loss": 0.0001, "step": 27377 }, { "epoch": 2.968126626192541, "grad_norm": 6.0821123123168945, "learning_rate": 2.1325982881183812e-07, "loss": 0.0447, "step": 27378 }, { "epoch": 2.968235039028621, "grad_norm": 0.010519202798604965, "learning_rate": 2.1253445524445093e-07, "loss": 0.0001, "step": 27379 }, { "epoch": 2.9683434518647007, "grad_norm": 8.629142761230469, "learning_rate": 2.118090816770637e-07, "loss": 0.0353, "step": 27380 }, { "epoch": 2.9684518647007807, "grad_norm": 0.4677008092403412, "learning_rate": 2.110837081096765e-07, "loss": 0.001, "step": 27381 }, { "epoch": 2.9685602775368602, "grad_norm": 5.433442115783691, "learning_rate": 2.1035833454228932e-07, "loss": 0.0265, "step": 27382 }, { "epoch": 2.9686686903729402, "grad_norm": 0.3960169851779938, "learning_rate": 2.0963296097490207e-07, "loss": 0.0081, "step": 27383 }, { "epoch": 2.96877710320902, "grad_norm": 0.38976985216140747, "learning_rate": 2.0890758740751487e-07, "loss": 0.0096, "step": 27384 }, { "epoch": 2.9688855160450998, "grad_norm": 4.447598457336426, "learning_rate": 2.0818221384012768e-07, "loss": 0.0103, "step": 27385 }, { "epoch": 2.9689939288811793, "grad_norm": 1.1697410345077515, "learning_rate": 2.0745684027274046e-07, "loss": 0.0361, "step": 27386 }, { "epoch": 2.9691023417172593, "grad_norm": 0.49500903487205505, "learning_rate": 2.0673146670535326e-07, "loss": 0.008, "step": 27387 }, { "epoch": 2.9692107545533393, "grad_norm": 10.09487533569336, "learning_rate": 2.0600609313796607e-07, "loss": 0.0268, "step": 27388 }, { "epoch": 2.969319167389419, "grad_norm": 0.034120459109544754, "learning_rate": 2.0528071957057887e-07, "loss": 0.0003, "step": 27389 }, { "epoch": 2.969427580225499, "grad_norm": 0.3303602337837219, "learning_rate": 2.0455534600319165e-07, "loss": 0.0087, "step": 27390 }, { "epoch": 2.9695359930615783, "grad_norm": 0.006100999657064676, "learning_rate": 2.0382997243580446e-07, "loss": 0.0, "step": 27391 }, { "epoch": 2.9696444058976583, "grad_norm": 0.038490064442157745, "learning_rate": 2.0310459886841726e-07, "loss": 0.0003, "step": 27392 }, { "epoch": 2.9697528187337383, "grad_norm": 0.2930275499820709, "learning_rate": 2.0237922530103004e-07, "loss": 0.0062, "step": 27393 }, { "epoch": 2.969861231569818, "grad_norm": 0.23354263603687286, "learning_rate": 2.0165385173364285e-07, "loss": 0.0033, "step": 27394 }, { "epoch": 2.9699696444058974, "grad_norm": 0.01991429179906845, "learning_rate": 2.0092847816625565e-07, "loss": 0.0002, "step": 27395 }, { "epoch": 2.9700780572419774, "grad_norm": 1.1385345458984375, "learning_rate": 2.0020310459886846e-07, "loss": 0.0177, "step": 27396 }, { "epoch": 2.9701864700780574, "grad_norm": 0.5083715319633484, "learning_rate": 1.994777310314812e-07, "loss": 0.0078, "step": 27397 }, { "epoch": 2.970294882914137, "grad_norm": 2.244704246520996, "learning_rate": 1.9875235746409404e-07, "loss": 0.0302, "step": 27398 }, { "epoch": 2.970403295750217, "grad_norm": 0.2334730178117752, "learning_rate": 1.9802698389670685e-07, "loss": 0.003, "step": 27399 }, { "epoch": 2.9705117085862964, "grad_norm": 0.04854587838053703, "learning_rate": 1.973016103293196e-07, "loss": 0.0004, "step": 27400 }, { "epoch": 2.9706201214223764, "grad_norm": 0.37472161650657654, "learning_rate": 1.965762367619324e-07, "loss": 0.011, "step": 27401 }, { "epoch": 2.9707285342584564, "grad_norm": 0.1725296527147293, "learning_rate": 1.958508631945452e-07, "loss": 0.0009, "step": 27402 }, { "epoch": 2.970836947094536, "grad_norm": 0.20064190030097961, "learning_rate": 1.9512548962715799e-07, "loss": 0.0017, "step": 27403 }, { "epoch": 2.970945359930616, "grad_norm": 1.0947810411453247, "learning_rate": 1.944001160597708e-07, "loss": 0.0265, "step": 27404 }, { "epoch": 2.9710537727666955, "grad_norm": 0.0959242433309555, "learning_rate": 1.936747424923836e-07, "loss": 0.0006, "step": 27405 }, { "epoch": 2.9711621856027755, "grad_norm": 0.7950574159622192, "learning_rate": 1.929493689249964e-07, "loss": 0.0052, "step": 27406 }, { "epoch": 2.9712705984388554, "grad_norm": 0.24825116991996765, "learning_rate": 1.9222399535760918e-07, "loss": 0.005, "step": 27407 }, { "epoch": 2.971379011274935, "grad_norm": 0.8421223163604736, "learning_rate": 1.9149862179022199e-07, "loss": 0.0025, "step": 27408 }, { "epoch": 2.9714874241110145, "grad_norm": 0.016503717750310898, "learning_rate": 1.907732482228348e-07, "loss": 0.0002, "step": 27409 }, { "epoch": 2.9715958369470945, "grad_norm": 4.198681831359863, "learning_rate": 1.9004787465544757e-07, "loss": 0.0251, "step": 27410 }, { "epoch": 2.9717042497831745, "grad_norm": 0.01044346671551466, "learning_rate": 1.8932250108806037e-07, "loss": 0.0001, "step": 27411 }, { "epoch": 2.971812662619254, "grad_norm": 0.7984375953674316, "learning_rate": 1.8859712752067318e-07, "loss": 0.0039, "step": 27412 }, { "epoch": 2.971921075455334, "grad_norm": 7.885728359222412, "learning_rate": 1.8787175395328598e-07, "loss": 0.0609, "step": 27413 }, { "epoch": 2.9720294882914136, "grad_norm": 0.08693587779998779, "learning_rate": 1.8714638038589874e-07, "loss": 0.0004, "step": 27414 }, { "epoch": 2.9721379011274935, "grad_norm": 0.06315040588378906, "learning_rate": 1.8642100681851154e-07, "loss": 0.0008, "step": 27415 }, { "epoch": 2.9722463139635735, "grad_norm": 3.564845323562622, "learning_rate": 1.8569563325112435e-07, "loss": 0.0206, "step": 27416 }, { "epoch": 2.972354726799653, "grad_norm": 0.28403881192207336, "learning_rate": 1.8497025968373713e-07, "loss": 0.0016, "step": 27417 }, { "epoch": 2.9724631396357326, "grad_norm": 1.1771724224090576, "learning_rate": 1.8424488611634993e-07, "loss": 0.006, "step": 27418 }, { "epoch": 2.9725715524718126, "grad_norm": 0.024826841428875923, "learning_rate": 1.8351951254896274e-07, "loss": 0.0004, "step": 27419 }, { "epoch": 2.9726799653078926, "grad_norm": 0.045710187405347824, "learning_rate": 1.8279413898157551e-07, "loss": 0.0003, "step": 27420 }, { "epoch": 2.972788378143972, "grad_norm": 0.38134631514549255, "learning_rate": 1.8206876541418832e-07, "loss": 0.0082, "step": 27421 }, { "epoch": 2.972896790980052, "grad_norm": 0.48068493604660034, "learning_rate": 1.8134339184680113e-07, "loss": 0.0014, "step": 27422 }, { "epoch": 2.9730052038161316, "grad_norm": 4.439300060272217, "learning_rate": 1.8061801827941393e-07, "loss": 0.0086, "step": 27423 }, { "epoch": 2.9731136166522116, "grad_norm": 4.756167888641357, "learning_rate": 1.798926447120267e-07, "loss": 0.0091, "step": 27424 }, { "epoch": 2.9732220294882916, "grad_norm": 0.5417370200157166, "learning_rate": 1.7916727114463951e-07, "loss": 0.001, "step": 27425 }, { "epoch": 2.973330442324371, "grad_norm": 0.2828429341316223, "learning_rate": 1.7844189757725232e-07, "loss": 0.0159, "step": 27426 }, { "epoch": 2.973438855160451, "grad_norm": 0.7578422427177429, "learning_rate": 1.7771652400986507e-07, "loss": 0.009, "step": 27427 }, { "epoch": 2.9735472679965307, "grad_norm": 0.5111480355262756, "learning_rate": 1.7699115044247788e-07, "loss": 0.0103, "step": 27428 }, { "epoch": 2.9736556808326107, "grad_norm": 0.5106937289237976, "learning_rate": 1.7626577687509068e-07, "loss": 0.0086, "step": 27429 }, { "epoch": 2.97376409366869, "grad_norm": 0.0716172605752945, "learning_rate": 1.7554040330770346e-07, "loss": 0.0004, "step": 27430 }, { "epoch": 2.97387250650477, "grad_norm": 0.42620426416397095, "learning_rate": 1.7481502974031627e-07, "loss": 0.0046, "step": 27431 }, { "epoch": 2.9739809193408497, "grad_norm": 0.44448742270469666, "learning_rate": 1.7408965617292907e-07, "loss": 0.0053, "step": 27432 }, { "epoch": 2.9740893321769297, "grad_norm": 0.14145086705684662, "learning_rate": 1.7336428260554188e-07, "loss": 0.0002, "step": 27433 }, { "epoch": 2.9741977450130097, "grad_norm": 0.028648732230067253, "learning_rate": 1.7263890903815465e-07, "loss": 0.0003, "step": 27434 }, { "epoch": 2.9743061578490892, "grad_norm": 0.6739627122879028, "learning_rate": 1.7191353547076746e-07, "loss": 0.0154, "step": 27435 }, { "epoch": 2.9744145706851692, "grad_norm": 0.4675251543521881, "learning_rate": 1.7118816190338026e-07, "loss": 0.005, "step": 27436 }, { "epoch": 2.9745229835212488, "grad_norm": 0.22110304236412048, "learning_rate": 1.7046278833599304e-07, "loss": 0.0029, "step": 27437 }, { "epoch": 2.9746313963573288, "grad_norm": 33.235801696777344, "learning_rate": 1.6973741476860585e-07, "loss": 0.0134, "step": 27438 }, { "epoch": 2.9747398091934087, "grad_norm": 2.860142469406128, "learning_rate": 1.6901204120121865e-07, "loss": 0.0284, "step": 27439 }, { "epoch": 2.9748482220294883, "grad_norm": 2.946986675262451, "learning_rate": 1.6828666763383146e-07, "loss": 0.001, "step": 27440 }, { "epoch": 2.974956634865568, "grad_norm": 0.16826076805591583, "learning_rate": 1.6756129406644424e-07, "loss": 0.0023, "step": 27441 }, { "epoch": 2.975065047701648, "grad_norm": 0.0917089506983757, "learning_rate": 1.6683592049905704e-07, "loss": 0.0007, "step": 27442 }, { "epoch": 2.975173460537728, "grad_norm": 0.5990183353424072, "learning_rate": 1.6611054693166985e-07, "loss": 0.0155, "step": 27443 }, { "epoch": 2.9752818733738073, "grad_norm": 0.5109074115753174, "learning_rate": 1.653851733642826e-07, "loss": 0.0107, "step": 27444 }, { "epoch": 2.9753902862098873, "grad_norm": 0.007819145917892456, "learning_rate": 1.646597997968954e-07, "loss": 0.0, "step": 27445 }, { "epoch": 2.975498699045967, "grad_norm": 0.4755796492099762, "learning_rate": 1.639344262295082e-07, "loss": 0.0096, "step": 27446 }, { "epoch": 2.975607111882047, "grad_norm": 1.5252330303192139, "learning_rate": 1.63209052662121e-07, "loss": 0.0172, "step": 27447 }, { "epoch": 2.975715524718127, "grad_norm": 0.44188395142555237, "learning_rate": 1.624836790947338e-07, "loss": 0.0173, "step": 27448 }, { "epoch": 2.9758239375542064, "grad_norm": 4.814097881317139, "learning_rate": 1.617583055273466e-07, "loss": 0.0277, "step": 27449 }, { "epoch": 2.9759323503902864, "grad_norm": 0.07901988923549652, "learning_rate": 1.610329319599594e-07, "loss": 0.0005, "step": 27450 }, { "epoch": 2.976040763226366, "grad_norm": 20.898067474365234, "learning_rate": 1.6030755839257218e-07, "loss": 0.0475, "step": 27451 }, { "epoch": 2.976149176062446, "grad_norm": 3.6374289989471436, "learning_rate": 1.59582184825185e-07, "loss": 0.0205, "step": 27452 }, { "epoch": 2.9762575888985254, "grad_norm": 3.2165310382843018, "learning_rate": 1.588568112577978e-07, "loss": 0.0371, "step": 27453 }, { "epoch": 2.9763660017346054, "grad_norm": 0.05325718969106674, "learning_rate": 1.5813143769041057e-07, "loss": 0.0003, "step": 27454 }, { "epoch": 2.976474414570685, "grad_norm": 24.694324493408203, "learning_rate": 1.5740606412302338e-07, "loss": 0.0453, "step": 27455 }, { "epoch": 2.976582827406765, "grad_norm": 0.45068350434303284, "learning_rate": 1.5668069055563618e-07, "loss": 0.0021, "step": 27456 }, { "epoch": 2.976691240242845, "grad_norm": 0.24742938578128815, "learning_rate": 1.5595531698824896e-07, "loss": 0.0034, "step": 27457 }, { "epoch": 2.9767996530789245, "grad_norm": 0.2667144238948822, "learning_rate": 1.5522994342086174e-07, "loss": 0.0007, "step": 27458 }, { "epoch": 2.9769080659150045, "grad_norm": 0.29095274209976196, "learning_rate": 1.5450456985347454e-07, "loss": 0.0057, "step": 27459 }, { "epoch": 2.977016478751084, "grad_norm": 8.349061012268066, "learning_rate": 1.5377919628608735e-07, "loss": 0.0341, "step": 27460 }, { "epoch": 2.977124891587164, "grad_norm": 0.1509748250246048, "learning_rate": 1.5305382271870015e-07, "loss": 0.0004, "step": 27461 }, { "epoch": 2.977233304423244, "grad_norm": 6.446709632873535, "learning_rate": 1.5232844915131293e-07, "loss": 0.0215, "step": 27462 }, { "epoch": 2.9773417172593235, "grad_norm": 4.235214710235596, "learning_rate": 1.5160307558392574e-07, "loss": 0.0255, "step": 27463 }, { "epoch": 2.977450130095403, "grad_norm": 0.7269187569618225, "learning_rate": 1.5087770201653854e-07, "loss": 0.0115, "step": 27464 }, { "epoch": 2.977558542931483, "grad_norm": 0.04849454015493393, "learning_rate": 1.5015232844915132e-07, "loss": 0.0003, "step": 27465 }, { "epoch": 2.977666955767563, "grad_norm": 0.06702902913093567, "learning_rate": 1.4942695488176413e-07, "loss": 0.0004, "step": 27466 }, { "epoch": 2.9777753686036426, "grad_norm": 3.940575361251831, "learning_rate": 1.487015813143769e-07, "loss": 0.005, "step": 27467 }, { "epoch": 2.9778837814397225, "grad_norm": 0.5062255263328552, "learning_rate": 1.479762077469897e-07, "loss": 0.0059, "step": 27468 }, { "epoch": 2.977992194275802, "grad_norm": 1.0187033414840698, "learning_rate": 1.4725083417960252e-07, "loss": 0.007, "step": 27469 }, { "epoch": 2.978100607111882, "grad_norm": 0.041257623583078384, "learning_rate": 1.465254606122153e-07, "loss": 0.0002, "step": 27470 }, { "epoch": 2.978209019947962, "grad_norm": 0.006549369543790817, "learning_rate": 1.458000870448281e-07, "loss": 0.0001, "step": 27471 }, { "epoch": 2.9783174327840416, "grad_norm": 2.0268588066101074, "learning_rate": 1.4507471347744088e-07, "loss": 0.0233, "step": 27472 }, { "epoch": 2.9784258456201216, "grad_norm": 0.24168744683265686, "learning_rate": 1.4434933991005368e-07, "loss": 0.0048, "step": 27473 }, { "epoch": 2.978534258456201, "grad_norm": 0.3712782859802246, "learning_rate": 1.436239663426665e-07, "loss": 0.0106, "step": 27474 }, { "epoch": 2.978642671292281, "grad_norm": 0.3651651442050934, "learning_rate": 1.4289859277527927e-07, "loss": 0.0113, "step": 27475 }, { "epoch": 2.9787510841283606, "grad_norm": 11.039864540100098, "learning_rate": 1.4217321920789207e-07, "loss": 0.0328, "step": 27476 }, { "epoch": 2.9788594969644406, "grad_norm": 0.35900574922561646, "learning_rate": 1.4144784564050488e-07, "loss": 0.0154, "step": 27477 }, { "epoch": 2.97896790980052, "grad_norm": 0.4896814823150635, "learning_rate": 1.4072247207311768e-07, "loss": 0.001, "step": 27478 }, { "epoch": 2.9790763226366, "grad_norm": 0.7199988961219788, "learning_rate": 1.3999709850573046e-07, "loss": 0.0078, "step": 27479 }, { "epoch": 2.97918473547268, "grad_norm": 0.03961775079369545, "learning_rate": 1.3927172493834324e-07, "loss": 0.0002, "step": 27480 }, { "epoch": 2.9792931483087597, "grad_norm": 0.01593724824488163, "learning_rate": 1.3854635137095605e-07, "loss": 0.0002, "step": 27481 }, { "epoch": 2.9794015611448397, "grad_norm": 0.03173870965838432, "learning_rate": 1.3782097780356885e-07, "loss": 0.0004, "step": 27482 }, { "epoch": 2.979509973980919, "grad_norm": 0.45774826407432556, "learning_rate": 1.3709560423618166e-07, "loss": 0.0223, "step": 27483 }, { "epoch": 2.979618386816999, "grad_norm": 0.42218002676963806, "learning_rate": 1.3637023066879443e-07, "loss": 0.0079, "step": 27484 }, { "epoch": 2.979726799653079, "grad_norm": 0.24993953108787537, "learning_rate": 1.3564485710140724e-07, "loss": 0.0002, "step": 27485 }, { "epoch": 2.9798352124891587, "grad_norm": 0.011821141466498375, "learning_rate": 1.3491948353402004e-07, "loss": 0.0, "step": 27486 }, { "epoch": 2.9799436253252383, "grad_norm": 0.6480017900466919, "learning_rate": 1.3419410996663282e-07, "loss": 0.0009, "step": 27487 }, { "epoch": 2.9800520381613183, "grad_norm": 0.0654943510890007, "learning_rate": 1.3346873639924563e-07, "loss": 0.0003, "step": 27488 }, { "epoch": 2.9801604509973982, "grad_norm": 0.4225602447986603, "learning_rate": 1.327433628318584e-07, "loss": 0.0129, "step": 27489 }, { "epoch": 2.9802688638334778, "grad_norm": 0.3100707232952118, "learning_rate": 1.320179892644712e-07, "loss": 0.01, "step": 27490 }, { "epoch": 2.9803772766695578, "grad_norm": 0.260210782289505, "learning_rate": 1.3129261569708402e-07, "loss": 0.0042, "step": 27491 }, { "epoch": 2.9804856895056373, "grad_norm": 0.5275264978408813, "learning_rate": 1.305672421296968e-07, "loss": 0.0103, "step": 27492 }, { "epoch": 2.9805941023417173, "grad_norm": 0.5642887353897095, "learning_rate": 1.298418685623096e-07, "loss": 0.0053, "step": 27493 }, { "epoch": 2.9807025151777973, "grad_norm": 3.1861510276794434, "learning_rate": 1.2911649499492238e-07, "loss": 0.0172, "step": 27494 }, { "epoch": 2.980810928013877, "grad_norm": 0.30854517221450806, "learning_rate": 1.283911214275352e-07, "loss": 0.0058, "step": 27495 }, { "epoch": 2.980919340849957, "grad_norm": 6.457443714141846, "learning_rate": 1.27665747860148e-07, "loss": 0.0227, "step": 27496 }, { "epoch": 2.9810277536860363, "grad_norm": 0.3446292579174042, "learning_rate": 1.2694037429276077e-07, "loss": 0.0112, "step": 27497 }, { "epoch": 2.9811361665221163, "grad_norm": 0.5152358412742615, "learning_rate": 1.2621500072537357e-07, "loss": 0.0019, "step": 27498 }, { "epoch": 2.981244579358196, "grad_norm": 0.9774147272109985, "learning_rate": 1.2548962715798638e-07, "loss": 0.0076, "step": 27499 }, { "epoch": 2.981352992194276, "grad_norm": 0.6940634846687317, "learning_rate": 1.2476425359059918e-07, "loss": 0.0056, "step": 27500 }, { "epoch": 2.9814614050303554, "grad_norm": 0.06711342185735703, "learning_rate": 1.2403888002321196e-07, "loss": 0.0004, "step": 27501 }, { "epoch": 2.9815698178664354, "grad_norm": 0.2272326499223709, "learning_rate": 1.2331350645582474e-07, "loss": 0.0032, "step": 27502 }, { "epoch": 2.9816782307025154, "grad_norm": 0.5712870359420776, "learning_rate": 1.2258813288843755e-07, "loss": 0.0168, "step": 27503 }, { "epoch": 2.981786643538595, "grad_norm": 0.3552911579608917, "learning_rate": 1.2186275932105035e-07, "loss": 0.0106, "step": 27504 }, { "epoch": 2.981895056374675, "grad_norm": 1.8157906532287598, "learning_rate": 1.2113738575366316e-07, "loss": 0.0171, "step": 27505 }, { "epoch": 2.9820034692107544, "grad_norm": 6.286378383636475, "learning_rate": 1.2041201218627594e-07, "loss": 0.0246, "step": 27506 }, { "epoch": 2.9821118820468344, "grad_norm": 0.36350640654563904, "learning_rate": 1.1968663861888874e-07, "loss": 0.0005, "step": 27507 }, { "epoch": 2.9822202948829144, "grad_norm": 0.09150068461894989, "learning_rate": 1.1896126505150153e-07, "loss": 0.0004, "step": 27508 }, { "epoch": 2.982328707718994, "grad_norm": 2.9884040355682373, "learning_rate": 1.1823589148411432e-07, "loss": 0.0069, "step": 27509 }, { "epoch": 2.9824371205550735, "grad_norm": 0.9092819094657898, "learning_rate": 1.1751051791672713e-07, "loss": 0.0132, "step": 27510 }, { "epoch": 2.9825455333911535, "grad_norm": 1.3014413118362427, "learning_rate": 1.1678514434933992e-07, "loss": 0.0045, "step": 27511 }, { "epoch": 2.9826539462272335, "grad_norm": 17.335811614990234, "learning_rate": 1.1605977078195273e-07, "loss": 0.0227, "step": 27512 }, { "epoch": 2.982762359063313, "grad_norm": 5.161755561828613, "learning_rate": 1.153343972145655e-07, "loss": 0.0032, "step": 27513 }, { "epoch": 2.982870771899393, "grad_norm": 2.9663844108581543, "learning_rate": 1.146090236471783e-07, "loss": 0.0076, "step": 27514 }, { "epoch": 2.9829791847354725, "grad_norm": 0.5397104620933533, "learning_rate": 1.138836500797911e-07, "loss": 0.0126, "step": 27515 }, { "epoch": 2.9830875975715525, "grad_norm": 0.0634605884552002, "learning_rate": 1.131582765124039e-07, "loss": 0.0005, "step": 27516 }, { "epoch": 2.9831960104076325, "grad_norm": 6.811298847198486, "learning_rate": 1.124329029450167e-07, "loss": 0.0347, "step": 27517 }, { "epoch": 2.983304423243712, "grad_norm": 0.6646113991737366, "learning_rate": 1.1170752937762949e-07, "loss": 0.0025, "step": 27518 }, { "epoch": 2.9834128360797916, "grad_norm": 0.930210292339325, "learning_rate": 1.1098215581024228e-07, "loss": 0.005, "step": 27519 }, { "epoch": 2.9835212489158716, "grad_norm": 0.4157073497772217, "learning_rate": 1.1025678224285509e-07, "loss": 0.0022, "step": 27520 }, { "epoch": 2.9836296617519515, "grad_norm": 0.31598398089408875, "learning_rate": 1.0953140867546787e-07, "loss": 0.0009, "step": 27521 }, { "epoch": 2.983738074588031, "grad_norm": 0.5756309628486633, "learning_rate": 1.0880603510808067e-07, "loss": 0.0163, "step": 27522 }, { "epoch": 2.983846487424111, "grad_norm": 1.8788031339645386, "learning_rate": 1.0808066154069346e-07, "loss": 0.0573, "step": 27523 }, { "epoch": 2.9839549002601906, "grad_norm": 4.955904006958008, "learning_rate": 1.0735528797330626e-07, "loss": 0.0839, "step": 27524 }, { "epoch": 2.9840633130962706, "grad_norm": 0.4019317030906677, "learning_rate": 1.0662991440591906e-07, "loss": 0.0216, "step": 27525 }, { "epoch": 2.9841717259323506, "grad_norm": 0.020936092361807823, "learning_rate": 1.0590454083853185e-07, "loss": 0.0001, "step": 27526 }, { "epoch": 2.98428013876843, "grad_norm": 0.9640170931816101, "learning_rate": 1.0517916727114466e-07, "loss": 0.0218, "step": 27527 }, { "epoch": 2.98438855160451, "grad_norm": 0.1612490564584732, "learning_rate": 1.0445379370375744e-07, "loss": 0.0004, "step": 27528 }, { "epoch": 2.9844969644405897, "grad_norm": 0.6331524848937988, "learning_rate": 1.0372842013637023e-07, "loss": 0.0157, "step": 27529 }, { "epoch": 2.9846053772766696, "grad_norm": 0.02025386318564415, "learning_rate": 1.0300304656898303e-07, "loss": 0.0002, "step": 27530 }, { "epoch": 2.9847137901127496, "grad_norm": 4.956486225128174, "learning_rate": 1.0227767300159583e-07, "loss": 0.0591, "step": 27531 }, { "epoch": 2.984822202948829, "grad_norm": 0.8779656291007996, "learning_rate": 1.0155229943420863e-07, "loss": 0.014, "step": 27532 }, { "epoch": 2.9849306157849087, "grad_norm": 0.013637665659189224, "learning_rate": 1.0082692586682142e-07, "loss": 0.0001, "step": 27533 }, { "epoch": 2.9850390286209887, "grad_norm": 2.3767664432525635, "learning_rate": 1.0010155229943423e-07, "loss": 0.065, "step": 27534 }, { "epoch": 2.9851474414570687, "grad_norm": 0.5323300957679749, "learning_rate": 9.937617873204702e-08, "loss": 0.0006, "step": 27535 }, { "epoch": 2.985255854293148, "grad_norm": 0.3130680024623871, "learning_rate": 9.86508051646598e-08, "loss": 0.0173, "step": 27536 }, { "epoch": 2.985364267129228, "grad_norm": 3.440382480621338, "learning_rate": 9.79254315972726e-08, "loss": 0.0193, "step": 27537 }, { "epoch": 2.9854726799653077, "grad_norm": 0.014736427925527096, "learning_rate": 9.72000580298854e-08, "loss": 0.0001, "step": 27538 }, { "epoch": 2.9855810928013877, "grad_norm": 0.6748586893081665, "learning_rate": 9.64746844624982e-08, "loss": 0.0089, "step": 27539 }, { "epoch": 2.9856895056374677, "grad_norm": 0.004584162961691618, "learning_rate": 9.574931089511099e-08, "loss": 0.0, "step": 27540 }, { "epoch": 2.9857979184735473, "grad_norm": 2.3198795318603516, "learning_rate": 9.502393732772378e-08, "loss": 0.0503, "step": 27541 }, { "epoch": 2.985906331309627, "grad_norm": 0.3246772587299347, "learning_rate": 9.429856376033659e-08, "loss": 0.0093, "step": 27542 }, { "epoch": 2.986014744145707, "grad_norm": 0.012948994524776936, "learning_rate": 9.357319019294937e-08, "loss": 0.0001, "step": 27543 }, { "epoch": 2.9861231569817868, "grad_norm": 0.7313475608825684, "learning_rate": 9.284781662556217e-08, "loss": 0.0112, "step": 27544 }, { "epoch": 2.9862315698178663, "grad_norm": 0.0797668844461441, "learning_rate": 9.212244305817497e-08, "loss": 0.0005, "step": 27545 }, { "epoch": 2.9863399826539463, "grad_norm": 1.790854573249817, "learning_rate": 9.139706949078776e-08, "loss": 0.0092, "step": 27546 }, { "epoch": 2.986448395490026, "grad_norm": 0.11137837916612625, "learning_rate": 9.067169592340056e-08, "loss": 0.0002, "step": 27547 }, { "epoch": 2.986556808326106, "grad_norm": 1.0917943716049194, "learning_rate": 8.994632235601335e-08, "loss": 0.0209, "step": 27548 }, { "epoch": 2.986665221162186, "grad_norm": 0.748464822769165, "learning_rate": 8.922094878862616e-08, "loss": 0.0015, "step": 27549 }, { "epoch": 2.9867736339982653, "grad_norm": 0.2549150288105011, "learning_rate": 8.849557522123894e-08, "loss": 0.0053, "step": 27550 }, { "epoch": 2.9868820468343453, "grad_norm": 3.0642359256744385, "learning_rate": 8.777020165385173e-08, "loss": 0.0198, "step": 27551 }, { "epoch": 2.986990459670425, "grad_norm": 0.042343929409980774, "learning_rate": 8.704482808646454e-08, "loss": 0.0004, "step": 27552 }, { "epoch": 2.987098872506505, "grad_norm": 0.46728670597076416, "learning_rate": 8.631945451907733e-08, "loss": 0.0147, "step": 27553 }, { "epoch": 2.987207285342585, "grad_norm": 0.04097704961895943, "learning_rate": 8.559408095169013e-08, "loss": 0.0003, "step": 27554 }, { "epoch": 2.9873156981786644, "grad_norm": 0.229180708527565, "learning_rate": 8.486870738430292e-08, "loss": 0.0007, "step": 27555 }, { "epoch": 2.987424111014744, "grad_norm": 0.6084461212158203, "learning_rate": 8.414333381691573e-08, "loss": 0.0086, "step": 27556 }, { "epoch": 2.987532523850824, "grad_norm": 0.2709397077560425, "learning_rate": 8.341796024952852e-08, "loss": 0.0046, "step": 27557 }, { "epoch": 2.987640936686904, "grad_norm": 0.22475109994411469, "learning_rate": 8.26925866821413e-08, "loss": 0.0003, "step": 27558 }, { "epoch": 2.9877493495229834, "grad_norm": 0.5322557091712952, "learning_rate": 8.19672131147541e-08, "loss": 0.0136, "step": 27559 }, { "epoch": 2.9878577623590634, "grad_norm": 0.011012554168701172, "learning_rate": 8.12418395473669e-08, "loss": 0.0001, "step": 27560 }, { "epoch": 2.987966175195143, "grad_norm": 0.9664759635925293, "learning_rate": 8.05164659799797e-08, "loss": 0.0122, "step": 27561 }, { "epoch": 2.988074588031223, "grad_norm": 0.4431019127368927, "learning_rate": 7.97910924125925e-08, "loss": 0.0087, "step": 27562 }, { "epoch": 2.988183000867303, "grad_norm": 2.08178973197937, "learning_rate": 7.906571884520529e-08, "loss": 0.0256, "step": 27563 }, { "epoch": 2.9882914137033825, "grad_norm": 0.21520745754241943, "learning_rate": 7.834034527781809e-08, "loss": 0.001, "step": 27564 }, { "epoch": 2.988399826539462, "grad_norm": 0.04928437992930412, "learning_rate": 7.761497171043087e-08, "loss": 0.0003, "step": 27565 }, { "epoch": 2.988508239375542, "grad_norm": 0.6065090298652649, "learning_rate": 7.688959814304367e-08, "loss": 0.0217, "step": 27566 }, { "epoch": 2.988616652211622, "grad_norm": 0.6075150966644287, "learning_rate": 7.616422457565647e-08, "loss": 0.0093, "step": 27567 }, { "epoch": 2.9887250650477015, "grad_norm": 0.11718718707561493, "learning_rate": 7.543885100826927e-08, "loss": 0.0004, "step": 27568 }, { "epoch": 2.9888334778837815, "grad_norm": 1.9749640226364136, "learning_rate": 7.471347744088206e-08, "loss": 0.0336, "step": 27569 }, { "epoch": 2.988941890719861, "grad_norm": 0.1494942307472229, "learning_rate": 7.398810387349486e-08, "loss": 0.0013, "step": 27570 }, { "epoch": 2.989050303555941, "grad_norm": 0.010669298470020294, "learning_rate": 7.326273030610765e-08, "loss": 0.0001, "step": 27571 }, { "epoch": 2.989158716392021, "grad_norm": 0.10718242079019547, "learning_rate": 7.253735673872044e-08, "loss": 0.0009, "step": 27572 }, { "epoch": 2.9892671292281006, "grad_norm": 0.4169631004333496, "learning_rate": 7.181198317133324e-08, "loss": 0.0081, "step": 27573 }, { "epoch": 2.9893755420641805, "grad_norm": 3.295569896697998, "learning_rate": 7.108660960394604e-08, "loss": 0.0736, "step": 27574 }, { "epoch": 2.98948395490026, "grad_norm": 0.058350518345832825, "learning_rate": 7.036123603655884e-08, "loss": 0.0005, "step": 27575 }, { "epoch": 2.98959236773634, "grad_norm": 0.018849041312932968, "learning_rate": 6.963586246917162e-08, "loss": 0.0001, "step": 27576 }, { "epoch": 2.9897007805724196, "grad_norm": 0.052976906299591064, "learning_rate": 6.891048890178443e-08, "loss": 0.0002, "step": 27577 }, { "epoch": 2.9898091934084996, "grad_norm": 1.6822681427001953, "learning_rate": 6.818511533439722e-08, "loss": 0.026, "step": 27578 }, { "epoch": 2.989917606244579, "grad_norm": 0.38237759470939636, "learning_rate": 6.745974176701002e-08, "loss": 0.0007, "step": 27579 }, { "epoch": 2.990026019080659, "grad_norm": 9.669676780700684, "learning_rate": 6.673436819962281e-08, "loss": 0.0136, "step": 27580 }, { "epoch": 2.990134431916739, "grad_norm": 0.1762242317199707, "learning_rate": 6.60089946322356e-08, "loss": 0.0004, "step": 27581 }, { "epoch": 2.9902428447528187, "grad_norm": 0.09033013135194778, "learning_rate": 6.52836210648484e-08, "loss": 0.0002, "step": 27582 }, { "epoch": 2.9903512575888986, "grad_norm": 0.007260526064783335, "learning_rate": 6.455824749746119e-08, "loss": 0.0001, "step": 27583 }, { "epoch": 2.990459670424978, "grad_norm": 0.33701086044311523, "learning_rate": 6.3832873930074e-08, "loss": 0.0082, "step": 27584 }, { "epoch": 2.990568083261058, "grad_norm": 1.7523943185806274, "learning_rate": 6.310750036268679e-08, "loss": 0.0459, "step": 27585 }, { "epoch": 2.990676496097138, "grad_norm": 1.366389274597168, "learning_rate": 6.238212679529959e-08, "loss": 0.0049, "step": 27586 }, { "epoch": 2.9907849089332177, "grad_norm": 1.4491212368011475, "learning_rate": 6.165675322791237e-08, "loss": 0.019, "step": 27587 }, { "epoch": 2.9908933217692972, "grad_norm": 0.05812055617570877, "learning_rate": 6.093137966052518e-08, "loss": 0.0004, "step": 27588 }, { "epoch": 2.991001734605377, "grad_norm": 0.8811498284339905, "learning_rate": 6.020600609313797e-08, "loss": 0.0025, "step": 27589 }, { "epoch": 2.991110147441457, "grad_norm": 1.0701850652694702, "learning_rate": 5.9480632525750766e-08, "loss": 0.0245, "step": 27590 }, { "epoch": 2.9912185602775367, "grad_norm": 0.5676831007003784, "learning_rate": 5.8755258958363565e-08, "loss": 0.0067, "step": 27591 }, { "epoch": 2.9913269731136167, "grad_norm": 0.18290646374225616, "learning_rate": 5.8029885390976363e-08, "loss": 0.0007, "step": 27592 }, { "epoch": 2.9914353859496963, "grad_norm": 0.2847866714000702, "learning_rate": 5.730451182358915e-08, "loss": 0.0008, "step": 27593 }, { "epoch": 2.9915437987857763, "grad_norm": 5.582730770111084, "learning_rate": 5.657913825620195e-08, "loss": 0.0094, "step": 27594 }, { "epoch": 2.9916522116218562, "grad_norm": 0.2677493989467621, "learning_rate": 5.5853764688814746e-08, "loss": 0.0059, "step": 27595 }, { "epoch": 2.991760624457936, "grad_norm": 0.5148899555206299, "learning_rate": 5.5128391121427544e-08, "loss": 0.001, "step": 27596 }, { "epoch": 2.9918690372940158, "grad_norm": 0.09527909755706787, "learning_rate": 5.4403017554040336e-08, "loss": 0.0005, "step": 27597 }, { "epoch": 2.9919774501300953, "grad_norm": 0.6592339873313904, "learning_rate": 5.367764398665313e-08, "loss": 0.0077, "step": 27598 }, { "epoch": 2.9920858629661753, "grad_norm": 1.0386028289794922, "learning_rate": 5.2952270419265927e-08, "loss": 0.0023, "step": 27599 }, { "epoch": 2.992194275802255, "grad_norm": 0.3552175760269165, "learning_rate": 5.222689685187872e-08, "loss": 0.0127, "step": 27600 }, { "epoch": 2.992302688638335, "grad_norm": 1.574221134185791, "learning_rate": 5.150152328449152e-08, "loss": 0.0031, "step": 27601 }, { "epoch": 2.9924111014744144, "grad_norm": 6.513606548309326, "learning_rate": 5.0776149717104316e-08, "loss": 0.0251, "step": 27602 }, { "epoch": 2.9925195143104943, "grad_norm": 0.2772023677825928, "learning_rate": 5.0050776149717114e-08, "loss": 0.0049, "step": 27603 }, { "epoch": 2.9926279271465743, "grad_norm": 0.8627147674560547, "learning_rate": 4.93254025823299e-08, "loss": 0.0017, "step": 27604 }, { "epoch": 2.992736339982654, "grad_norm": 5.4552717208862305, "learning_rate": 4.86000290149427e-08, "loss": 0.0837, "step": 27605 }, { "epoch": 2.992844752818734, "grad_norm": 12.102282524108887, "learning_rate": 4.7874655447555496e-08, "loss": 0.0213, "step": 27606 }, { "epoch": 2.9929531656548134, "grad_norm": 0.5234476923942566, "learning_rate": 4.7149281880168295e-08, "loss": 0.0208, "step": 27607 }, { "epoch": 2.9930615784908934, "grad_norm": 0.23872344195842743, "learning_rate": 4.642390831278109e-08, "loss": 0.0035, "step": 27608 }, { "epoch": 2.9931699913269734, "grad_norm": 2.4864726066589355, "learning_rate": 4.569853474539388e-08, "loss": 0.0387, "step": 27609 }, { "epoch": 2.993278404163053, "grad_norm": 0.6304858922958374, "learning_rate": 4.497316117800668e-08, "loss": 0.0105, "step": 27610 }, { "epoch": 2.9933868169991324, "grad_norm": 0.9577813744544983, "learning_rate": 4.424778761061947e-08, "loss": 0.0006, "step": 27611 }, { "epoch": 2.9934952298352124, "grad_norm": 0.9177289009094238, "learning_rate": 4.352241404323227e-08, "loss": 0.0154, "step": 27612 }, { "epoch": 2.9936036426712924, "grad_norm": 4.2773661613464355, "learning_rate": 4.2797040475845066e-08, "loss": 0.0262, "step": 27613 }, { "epoch": 2.993712055507372, "grad_norm": 2.1548194885253906, "learning_rate": 4.2071666908457865e-08, "loss": 0.0016, "step": 27614 }, { "epoch": 2.993820468343452, "grad_norm": 0.02890777960419655, "learning_rate": 4.134629334107065e-08, "loss": 0.0001, "step": 27615 }, { "epoch": 2.9939288811795315, "grad_norm": 3.286536693572998, "learning_rate": 4.062091977368345e-08, "loss": 0.0209, "step": 27616 }, { "epoch": 2.9940372940156115, "grad_norm": 0.478798508644104, "learning_rate": 3.989554620629625e-08, "loss": 0.0187, "step": 27617 }, { "epoch": 2.9941457068516915, "grad_norm": 0.5490431189537048, "learning_rate": 3.9170172638909046e-08, "loss": 0.0156, "step": 27618 }, { "epoch": 2.994254119687771, "grad_norm": 0.012071646749973297, "learning_rate": 3.844479907152184e-08, "loss": 0.0001, "step": 27619 }, { "epoch": 2.994362532523851, "grad_norm": 3.1258349418640137, "learning_rate": 3.7719425504134636e-08, "loss": 0.0475, "step": 27620 }, { "epoch": 2.9944709453599305, "grad_norm": 0.28457167744636536, "learning_rate": 3.699405193674743e-08, "loss": 0.0062, "step": 27621 }, { "epoch": 2.9945793581960105, "grad_norm": 0.05233331769704819, "learning_rate": 3.626867836936022e-08, "loss": 0.0003, "step": 27622 }, { "epoch": 2.99468777103209, "grad_norm": 7.465119361877441, "learning_rate": 3.554330480197302e-08, "loss": 0.0299, "step": 27623 }, { "epoch": 2.99479618386817, "grad_norm": 5.816355228424072, "learning_rate": 3.481793123458581e-08, "loss": 0.0556, "step": 27624 }, { "epoch": 2.9949045967042496, "grad_norm": 0.01240173727273941, "learning_rate": 3.409255766719861e-08, "loss": 0.0002, "step": 27625 }, { "epoch": 2.9950130095403296, "grad_norm": 3.3313302993774414, "learning_rate": 3.336718409981141e-08, "loss": 0.0061, "step": 27626 }, { "epoch": 2.9951214223764095, "grad_norm": 1.6678375005722046, "learning_rate": 3.26418105324242e-08, "loss": 0.0381, "step": 27627 }, { "epoch": 2.995229835212489, "grad_norm": 0.44072601199150085, "learning_rate": 3.1916436965037e-08, "loss": 0.0117, "step": 27628 }, { "epoch": 2.995338248048569, "grad_norm": 0.5376229882240295, "learning_rate": 3.1191063397649796e-08, "loss": 0.0156, "step": 27629 }, { "epoch": 2.9954466608846486, "grad_norm": 0.6363591551780701, "learning_rate": 3.046568983026259e-08, "loss": 0.0016, "step": 27630 }, { "epoch": 2.9955550737207286, "grad_norm": 0.09160953760147095, "learning_rate": 2.9740316262875383e-08, "loss": 0.0006, "step": 27631 }, { "epoch": 2.9956634865568086, "grad_norm": 0.9962654709815979, "learning_rate": 2.9014942695488182e-08, "loss": 0.0429, "step": 27632 }, { "epoch": 2.995771899392888, "grad_norm": 1.3897994756698608, "learning_rate": 2.8289569128100974e-08, "loss": 0.0297, "step": 27633 }, { "epoch": 2.9958803122289677, "grad_norm": 0.41535401344299316, "learning_rate": 2.7564195560713772e-08, "loss": 0.0102, "step": 27634 }, { "epoch": 2.9959887250650477, "grad_norm": 0.007233140990138054, "learning_rate": 2.6838821993326564e-08, "loss": 0.0001, "step": 27635 }, { "epoch": 2.9960971379011276, "grad_norm": 0.5509997606277466, "learning_rate": 2.611344842593936e-08, "loss": 0.0091, "step": 27636 }, { "epoch": 2.996205550737207, "grad_norm": 0.849916398525238, "learning_rate": 2.5388074858552158e-08, "loss": 0.0179, "step": 27637 }, { "epoch": 2.996313963573287, "grad_norm": 1.574615240097046, "learning_rate": 2.466270129116495e-08, "loss": 0.0339, "step": 27638 }, { "epoch": 2.9964223764093667, "grad_norm": 0.3988099694252014, "learning_rate": 2.3937327723777748e-08, "loss": 0.0079, "step": 27639 }, { "epoch": 2.9965307892454467, "grad_norm": 0.5044518113136292, "learning_rate": 2.3211954156390543e-08, "loss": 0.0006, "step": 27640 }, { "epoch": 2.9966392020815267, "grad_norm": 1.3782902956008911, "learning_rate": 2.248658058900334e-08, "loss": 0.0684, "step": 27641 }, { "epoch": 2.996747614917606, "grad_norm": 0.4125846326351166, "learning_rate": 2.1761207021616134e-08, "loss": 0.0092, "step": 27642 }, { "epoch": 2.996856027753686, "grad_norm": 0.03579667955636978, "learning_rate": 2.1035833454228932e-08, "loss": 0.0002, "step": 27643 }, { "epoch": 2.9969644405897657, "grad_norm": 0.40256962180137634, "learning_rate": 2.0310459886841724e-08, "loss": 0.0108, "step": 27644 }, { "epoch": 2.9970728534258457, "grad_norm": 0.7112043499946594, "learning_rate": 1.9585086319454523e-08, "loss": 0.0223, "step": 27645 }, { "epoch": 2.9971812662619253, "grad_norm": 0.6064226031303406, "learning_rate": 1.8859712752067318e-08, "loss": 0.0203, "step": 27646 }, { "epoch": 2.9972896790980053, "grad_norm": 0.6402047872543335, "learning_rate": 1.813433918468011e-08, "loss": 0.0025, "step": 27647 }, { "epoch": 2.997398091934085, "grad_norm": 4.53947114944458, "learning_rate": 1.7408965617292905e-08, "loss": 0.0321, "step": 27648 }, { "epoch": 2.997506504770165, "grad_norm": 11.220685005187988, "learning_rate": 1.6683592049905704e-08, "loss": 0.0066, "step": 27649 }, { "epoch": 2.9976149176062448, "grad_norm": 0.14401817321777344, "learning_rate": 1.59582184825185e-08, "loss": 0.0009, "step": 27650 }, { "epoch": 2.9977233304423243, "grad_norm": 0.7760694622993469, "learning_rate": 1.5232844915131294e-08, "loss": 0.0082, "step": 27651 }, { "epoch": 2.9978317432784043, "grad_norm": 0.19812174141407013, "learning_rate": 1.4507471347744091e-08, "loss": 0.0006, "step": 27652 }, { "epoch": 2.997940156114484, "grad_norm": 0.04594958946108818, "learning_rate": 1.3782097780356886e-08, "loss": 0.0003, "step": 27653 }, { "epoch": 2.998048568950564, "grad_norm": 0.7742661833763123, "learning_rate": 1.305672421296968e-08, "loss": 0.019, "step": 27654 }, { "epoch": 2.998156981786644, "grad_norm": 0.41609159111976624, "learning_rate": 1.2331350645582475e-08, "loss": 0.0011, "step": 27655 }, { "epoch": 2.9982653946227233, "grad_norm": 3.0448837280273438, "learning_rate": 1.1605977078195272e-08, "loss": 0.0203, "step": 27656 }, { "epoch": 2.998373807458803, "grad_norm": 0.045463088899850845, "learning_rate": 1.0880603510808067e-08, "loss": 0.0004, "step": 27657 }, { "epoch": 2.998482220294883, "grad_norm": 0.01620475761592388, "learning_rate": 1.0155229943420862e-08, "loss": 0.0002, "step": 27658 }, { "epoch": 2.998590633130963, "grad_norm": 0.29070353507995605, "learning_rate": 9.429856376033659e-09, "loss": 0.0056, "step": 27659 }, { "epoch": 2.9986990459670424, "grad_norm": 0.029807642102241516, "learning_rate": 8.704482808646453e-09, "loss": 0.0002, "step": 27660 }, { "epoch": 2.9988074588031224, "grad_norm": 7.483935356140137, "learning_rate": 7.97910924125925e-09, "loss": 0.0582, "step": 27661 }, { "epoch": 2.998915871639202, "grad_norm": 0.3714640438556671, "learning_rate": 7.2537356738720454e-09, "loss": 0.0127, "step": 27662 }, { "epoch": 2.999024284475282, "grad_norm": 0.380293071269989, "learning_rate": 6.52836210648484e-09, "loss": 0.0086, "step": 27663 }, { "epoch": 2.999132697311362, "grad_norm": 8.291632652282715, "learning_rate": 5.802988539097636e-09, "loss": 0.0123, "step": 27664 }, { "epoch": 2.9992411101474414, "grad_norm": 2.507683038711548, "learning_rate": 5.077614971710431e-09, "loss": 0.0846, "step": 27665 }, { "epoch": 2.999349522983521, "grad_norm": 2.3105781078338623, "learning_rate": 4.352241404323226e-09, "loss": 0.0438, "step": 27666 }, { "epoch": 2.999457935819601, "grad_norm": 0.032939646393060684, "learning_rate": 3.6268678369360227e-09, "loss": 0.0003, "step": 27667 }, { "epoch": 2.999566348655681, "grad_norm": 0.4375838041305542, "learning_rate": 2.901494269548818e-09, "loss": 0.009, "step": 27668 }, { "epoch": 2.9996747614917605, "grad_norm": 0.04562785476446152, "learning_rate": 2.176120702161613e-09, "loss": 0.0002, "step": 27669 }, { "epoch": 2.9997831743278405, "grad_norm": 0.0435706302523613, "learning_rate": 1.450747134774409e-09, "loss": 0.0002, "step": 27670 }, { "epoch": 2.99989158716392, "grad_norm": 0.07654446363449097, "learning_rate": 7.253735673872045e-10, "loss": 0.0003, "step": 27671 }, { "epoch": 3.0, "grad_norm": 0.10185680538415909, "learning_rate": 0.0, "loss": 0.0004, "step": 27672 }, { "epoch": 3.0, "step": 27672, "total_flos": 3.465458839619371e+18, "train_loss": 0.02820179975424636, "train_runtime": 8093.4838, "train_samples_per_second": 54.703, "train_steps_per_second": 3.419 } ], "logging_steps": 1.0, "max_steps": 27672, "num_input_tokens_seen": 0, "num_train_epochs": 3, "save_steps": 1000, "stateful_callbacks": { "TrainerControl": { "args": { "should_epoch_stop": false, "should_evaluate": false, "should_log": false, "should_save": true, "should_training_stop": true }, "attributes": {} } }, "total_flos": 3.465458839619371e+18, "train_batch_size": 2, "trial_name": null, "trial_params": null }